From gitlab at gitlab.haskell.org Sun Sep 1 10:06:03 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 01 Sep 2024 06:06:03 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] Make switch/jump tables PIC compatible Message-ID: <66d43c8baa3c4_a11551dd080131fe@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 0962e7ad by Sven Tennie at 2024-09-01T12:05:14+02:00 Make switch/jump tables PIC compatible - - - - - 1 changed file: - compiler/GHC/CmmToAsm/RV64/CodeGen.hs Changes: ===================================== compiler/GHC/CmmToAsm/RV64/CodeGen.hs ===================================== @@ -184,10 +184,93 @@ annExpr e {- debugIsOn -} = ANN (text . show $ e) -- ----------------------------------------------------------------------------- -- Generating a table-branch +-- Note [RISCV64 Jump Tables] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- +-- Jump tables are implemented by generating a table of relative addresses, +-- where each entry is the relative offset to the target block from the first +-- entry / table label (`generateJumpTableForInstr`). Using the jump table means +-- loading the entry's value and jumping to the calculated absolute address +-- (`genSwitch`). +-- +-- For example, this Cmm switch +-- +-- switch [1 .. 10] _s2wn::I64 { +-- case 1 : goto c347; +-- case 2 : goto c348; +-- case 3 : goto c349; +-- case 4 : goto c34a; +-- case 5 : goto c34b; +-- case 6 : goto c34c; +-- case 7 : goto c34d; +-- case 8 : goto c34e; +-- case 9 : goto c34f; +-- case 10 : goto c34g; +-- } // CmmSwitch +-- +-- leads to this jump table in Assembly +-- +-- .section .rodata +-- .balign 8 +-- .Ln34G: +-- .quad 0 +-- .quad .Lc347-(.Ln34G)+0 +-- .quad .Lc348-(.Ln34G)+0 +-- .quad .Lc349-(.Ln34G)+0 +-- .quad .Lc34a-(.Ln34G)+0 +-- .quad .Lc34b-(.Ln34G)+0 +-- .quad .Lc34c-(.Ln34G)+0 +-- .quad .Lc34d-(.Ln34G)+0 +-- .quad .Lc34e-(.Ln34G)+0 +-- .quad .Lc34f-(.Ln34G)+0 +-- .quad .Lc34g-(.Ln34G)+0 +-- +-- and this indexing code where the jump should be done (register t0 contains +-- the index) +-- +-- addi t0, t0, 0 // silly move (ignore it) +-- la t1, .Ln34G // load the table's address +-- sll t0, t0, 3 // index * 8 -> offset in bytes +-- add t0, t0, t1 // address of the table's entry +-- ld t0, 0(t0) // load entry +-- add t0, t0, t1 // relative to absolute address +-- jalr zero, t0, 0 // jump to the block +-- +-- In object code (disassembled) the table looks like +-- +-- 0000000000000000 <.Ln34G>: +-- ... +-- 8: R_RISCV_ADD64 .Lc347 +-- 8: R_RISCV_SUB64 .Ln34G +-- 10: R_RISCV_ADD64 .Lc348 +-- 10: R_RISCV_SUB64 .Ln34G +-- 18: R_RISCV_ADD64 .Lc349 +-- 18: R_RISCV_SUB64 .Ln34G +-- 20: R_RISCV_ADD64 .Lc34a +-- 20: R_RISCV_SUB64 .Ln34G +-- 28: R_RISCV_ADD64 .Lc34b +-- 28: R_RISCV_SUB64 .Ln34G +-- 30: R_RISCV_ADD64 .Lc34c +-- 30: R_RISCV_SUB64 .Ln34G +-- 38: R_RISCV_ADD64 .Lc34d +-- 38: R_RISCV_SUB64 .Ln34G +-- 40: R_RISCV_ADD64 .Lc34e +-- 40: R_RISCV_SUB64 .Ln34G +-- 48: R_RISCV_ADD64 .Lc34f +-- 48: R_RISCV_SUB64 .Ln34G +-- 50: R_RISCV_ADD64 .Lc34g +-- 50: R_RISCV_SUB64 .Ln34G +-- +-- I.e. the relative offset calculations are done by the linker via relocations. +-- This seems to be PIC compatible; at least `scanelf` (pax-utils) does not +-- complain. + + -- | Generate jump to jump table target -- -- The index into the jump table is calulated by evaluating @expr at . The --- corresponding table entry contains the address to jump to. +-- corresponding table entry contains the relative address to jump to (relative +-- to the jump table's first entry / the table's own label). genSwitch :: NCGConfig -> CmmExpr -> SwitchTargets -> NatM InstrBlock genSwitch config expr targets = do (reg, fmt1, e_code) <- getSomeReg indexExpr @@ -205,9 +288,15 @@ genSwitch config expr targets = do `appOL` t_code `appOL` toOL [ COMMENT (ftext "Jump table for switch"), + -- index to offset into the table (relative to tableReg) annExpr expr (SLL (OpReg (formatToWidth fmt1) reg) (OpReg (formatToWidth fmt1) reg) (OpImm (ImmInt 3))), + -- calculate table entry address ADD (OpReg W64 tmp) (OpReg (formatToWidth fmt1) reg) (OpReg (formatToWidth fmt2) tableReg), + -- load table entry (relative offset from tableReg (first entry) to target label) LDRU II64 (OpReg W64 tmp) (OpAddr (AddrRegImm tmp (ImmInt 0))), + -- calculate absolute address of the target label + ADD (OpReg W64 tmp) (OpReg W64 tmp) (OpReg W64 tableReg), + -- prepare jump to target label J_TBL ids (Just lbl) tmp ] return code @@ -226,9 +315,9 @@ genSwitch config expr targets = do -- | Generate jump table data (if required) -- --- Relies on PIC relocations. The idea is to emit one table entry per case. The --- entry is the label of the block to jump to. This will be relocated to be the --- address of the jump target. +-- The idea is to emit one table entry per case. The entry is the relative +-- address of the block to jump to (relative to the table's first entry / +-- table's own label.) The calculation itself is done by the linker. generateJumpTableForInstr :: NCGConfig -> Instr -> @@ -240,7 +329,13 @@ generateJumpTableForInstr config (J_TBL ids (Just lbl) _) = jumpTableEntryRel Nothing = CmmStaticLit (CmmInt 0 (ncgWordWidth config)) jumpTableEntryRel (Just blockid) = - CmmStaticLit (CmmLabel blockLabel) + CmmStaticLit + ( CmmLabelDiffOff + blockLabel + lbl + 0 + (ncgWordWidth config) + ) where blockLabel = blockLbl blockid in Just (CmmData (Section ReadOnlyData lbl) (CmmStaticsRaw lbl jumpTable)) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0962e7ad17221416a5d1c355732cc8256c71f236 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0962e7ad17221416a5d1c355732cc8256c71f236 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 1 12:52:48 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 01 Sep 2024 08:52:48 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] 2 commits: Improve documentation for stack spills Message-ID: <66d463a08238_a1155629440247af@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 45d3d6bd by Sven Tennie at 2024-09-01T14:45:18+02:00 Improve documentation for stack spills - - - - - a38bf454 by Sven Tennie at 2024-09-01T14:52:25+02:00 Reformat stack spill code - - - - - 1 changed file: - compiler/GHC/CmmToAsm/RV64/Instr.hs Changes: ===================================== compiler/GHC/CmmToAsm/RV64/Instr.hs ===================================== @@ -71,6 +71,7 @@ regUsageOfInstr platform instr = case instr of MULTILINE_COMMENT {} -> usage ([], []) PUSH_STACK_FRAME -> usage ([], []) POP_STACK_FRAME -> usage ([], []) + LOCATION {} -> usage ([], []) DELTA {} -> usage ([], []) ADD dst src1 src2 -> usage (regOp src1 ++ regOp src2, regOp dst) MUL dst src1 src2 -> usage (regOp src1 ++ regOp src2, regOp dst) @@ -155,6 +156,7 @@ patchRegsOfInstr instr env = case instr of MULTILINE_COMMENT {} -> instr PUSH_STACK_FRAME -> instr POP_STACK_FRAME -> instr + LOCATION {} -> instr DELTA {} -> instr ADD o1 o2 o3 -> ADD (patchOp o1) (patchOp o2) (patchOp o3) MUL o1 o2 o3 -> MUL (patchOp o1) (patchOp o2) (patchOp o3) @@ -255,23 +257,26 @@ patchJumpInstr instr patchF = _ -> panic $ "patchJumpInstr: " ++ instrCon instr -- ----------------------------------------------------------------------------- --- Note [Spills and Reloads] --- ~~~~~~~~~~~~~~~~~~~~~~~~~ +-- Note [RISCV64 Spills and Reloads] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- -- We reserve @RESERVED_C_STACK_BYTES@ on the C stack for spilling and reloading --- registers. AArch64s maximum displacement for SP relative spills and reloads --- is essentially [-256,255], or [0, 0xFFF]*8 = [0, 32760] for 64bits. +-- registers. The load and store instructions of RISCV64 address with a signed +-- 12-bit immediate + a register; machine stackpointer (sp/x2) in this case. -- --- The @RESERVED_C_STACK_BYTES@ is 16k, so we can't address any location in a --- single instruction. The idea is to use the Inter Procedure 0 (ip) register --- to perform the computations for larger offsets. +-- The @RESERVED_C_STACK_BYTES@ is 16k, so we can't always address into it in a +-- single load/store instruction. There are offsets to sp (not to be confused +-- with STG's SP!) which need a register to be calculated. -- -- Using sp to compute the offset will violate assumptions about the stack pointer -- pointing to the top of the stack during signal handling. As we can't force -- every signal to use its own stack, we have to ensure that the stack pointer -- always points to the top of the stack, and we can't use it for computation. -- +-- So, we reserve one register (ip) for this purpose (and other, unrelated +-- intermediate operations.) See Note [The made-up RISCV64 IP register] --- | An instruction to spill a register into a spill slot. +-- | Generate instructions to spill a register into a spill slot. mkSpillInstr :: (HasCallStack) => NCGConfig -> @@ -294,13 +299,22 @@ mkSpillInstr _config reg delta slot = fmt = case reg of RegReal (RealRegSingle n) | n < d0RegNo -> II64 _ -> FF64 - mkStrSpImm imm = ANN (text "Spill@" <> int (off - delta)) $ STR fmt (OpReg W64 reg) (OpAddr (AddrRegImm spMachReg (ImmInt imm))) - movImmToIp imm = ANN (text "Spill: IP <- " <> int imm) $ MOV ip (OpImm (ImmInt imm)) - addSpToIp = ANN (text "Spill: IP <- SP + IP ") $ ADD ip ip sp - mkStrIp = ANN (text "Spill@" <> int (off - delta)) $ STR fmt (OpReg W64 reg) (OpAddr (AddrReg ipReg)) + mkStrSpImm imm = + ANN (text "Spill@" <> int (off - delta)) + $ STR fmt (OpReg W64 reg) (OpAddr (AddrRegImm spMachReg (ImmInt imm))) + movImmToIp imm = + ANN (text "Spill: IP <- " <> int imm) + $ MOV ip (OpImm (ImmInt imm)) + addSpToIp = + ANN (text "Spill: IP <- SP + IP ") + $ ADD ip ip sp + mkStrIp = + ANN (text "Spill@" <> int (off - delta)) + $ STR fmt (OpReg W64 reg) (OpAddr (AddrReg ipReg)) off = spillSlotToOffset slot +-- | Generate instructions to load a register from a spill slot. mkLoadInstr :: NCGConfig -> -- | register to load @@ -322,10 +336,18 @@ mkLoadInstr _config reg delta slot = fmt = case reg of RegReal (RealRegSingle n) | n < d0RegNo -> II64 _ -> FF64 - mkLdrSpImm imm = ANN (text "Reload@" <> int (off - delta)) $ LDR fmt (OpReg W64 reg) (OpAddr (AddrRegImm spMachReg (ImmInt imm))) - movImmToIp imm = ANN (text "Reload: IP <- " <> int imm) $ MOV ip (OpImm (ImmInt imm)) - addSpToIp = ANN (text "Reload: IP <- SP + IP ") $ ADD ip ip sp - mkLdrIp = ANN (text "Reload@" <> int (off - delta)) $ LDR fmt (OpReg W64 reg) (OpAddr (AddrReg ipReg)) + mkLdrSpImm imm = + ANN (text "Reload@" <> int (off - delta)) + $ LDR fmt (OpReg W64 reg) (OpAddr (AddrRegImm spMachReg (ImmInt imm))) + movImmToIp imm = + ANN (text "Reload: IP <- " <> int imm) + $ MOV ip (OpImm (ImmInt imm)) + addSpToIp = + ANN (text "Reload: IP <- SP + IP ") + $ ADD ip ip sp + mkLdrIp = + ANN (text "Reload@" <> int (off - delta)) + $ LDR fmt (OpReg W64 reg) (OpAddr (AddrReg ipReg)) off = spillSlotToOffset slot View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0962e7ad17221416a5d1c355732cc8256c71f236...a38bf4541dc3a2888da1f03bffe906ff5ff5ac63 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0962e7ad17221416a5d1c355732cc8256c71f236...a38bf4541dc3a2888da1f03bffe906ff5ff5ac63 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 1 14:23:02 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 01 Sep 2024 10:23:02 -0400 Subject: [Git][ghc/ghc][wip/supersven/add-LLVMAS-target] Debug output Message-ID: <66d478c6840f6_25ac708ae30718bc@gitlab.mail> Sven Tennie pushed to branch wip/supersven/add-LLVMAS-target at Glasgow Haskell Compiler / GHC Commits: 33ac182b by Sven Tennie at 2024-09-01T16:22:14+02:00 Debug output - - - - - 1 changed file: - m4/fp_prog_llvm_as_args.m4 Changes: ===================================== m4/fp_prog_llvm_as_args.m4 ===================================== @@ -31,6 +31,11 @@ else AC_MSG_RESULT([no]) AC_MSG_WARN([\$LLVMAS ($LLVMAS) does not support host flags: $LlvmAsArgsHost.]) AC_MSG_WARN([Falling back to no flags (won't be able to build stages for the host architecture with \$LLVMAS).]) + + # TODO: Delete this debug output + DEFAULT_TRIPLE=`clang -print-target-triple` + AC_MSG_WARN([Default target triple: $DEFAULT_TRIPLE]) + # Here LLVMAS cannot assemble for the host. I.e. we won't be able to use it # to build intermediate GHC stages (with host target). This ressembles old # behaviour and is added for backwards compatibility. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/33ac182b6970a4a981b53ecd1100cbf223d122f7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/33ac182b6970a4a981b53ecd1100cbf223d122f7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 1 17:40:21 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 01 Sep 2024 13:40:21 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] 2 commits: Make instruction types strict in their fields Message-ID: <66d4a705504fb_1b48553dcd4039119@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 3b4dc97b by Sven Tennie at 2024-09-01T19:29:30+02:00 Make instruction types strict in their fields This should prevent memory leaks. - - - - - 26c79cfa by Sven Tennie at 2024-09-01T19:30:53+02:00 Use FCVT constructor for all float-related conversions This is closer to the assembly instruction and should reduce confusion. - - - - - 3 changed files: - compiler/GHC/CmmToAsm/RV64/CodeGen.hs - compiler/GHC/CmmToAsm/RV64/Instr.hs - compiler/GHC/CmmToAsm/RV64/Ppr.hs Changes: ===================================== compiler/GHC/CmmToAsm/RV64/CodeGen.hs ===================================== @@ -723,7 +723,7 @@ getRegister' config plat expr = ( \dst -> code `appOL` code_x - `snocOL` annExpr expr (SCVTF (OpReg to dst) (OpReg from reg_x)) -- (Signed ConVerT Float) + `snocOL` annExpr expr (FCVT IntToFloat (OpReg to dst) (OpReg from reg_x)) -- (Signed ConVerT Float) ) MO_SF_Round from to -> pure @@ -731,7 +731,7 @@ getRegister' config plat expr = (floatFormat to) ( \dst -> code - `snocOL` annExpr expr (SCVTF (OpReg to dst) (OpReg from reg)) -- (Signed ConVerT Float) + `snocOL` annExpr expr (FCVT IntToFloat (OpReg to dst) (OpReg from reg)) -- (Signed ConVerT Float) ) -- TODO: Can this case happen? MO_FS_Truncate from to @@ -743,7 +743,7 @@ getRegister' config plat expr = code `snocOL` -- W32 is the smallest width to convert to. Decrease width afterwards. - annExpr expr (FCVTZS (OpReg W32 dst) (OpReg from reg)) + annExpr expr (FCVT FloatToInt (OpReg W32 dst) (OpReg from reg)) `appOL` signExtendAdjustPrecission W32 to dst dst -- (float convert (-> zero) signed) ) MO_FS_Truncate from to -> @@ -752,7 +752,7 @@ getRegister' config plat expr = (intFormat to) ( \dst -> code - `snocOL` annExpr expr (FCVTZS (OpReg to dst) (OpReg from reg)) + `snocOL` annExpr expr (FCVT FloatToInt (OpReg to dst) (OpReg from reg)) `appOL` truncateReg from to dst -- (float convert (-> zero) signed) ) MO_UU_Conv from to @@ -774,7 +774,7 @@ getRegister' config plat expr = `appOL` truncateReg from to dst ) MO_SS_Conv from to -> ss_conv from to reg code - MO_FF_Conv from to -> return $ Any (floatFormat to) (\dst -> code `snocOL` annExpr e (FCVT (OpReg to dst) (OpReg from reg))) + MO_FF_Conv from to -> return $ Any (floatFormat to) (\dst -> code `snocOL` annExpr e (FCVT FloatToFloat (OpReg to dst) (OpReg from reg))) MO_WF_Bitcast w -> return $ Any (floatFormat w) (\dst -> code `snocOL` MOV (OpReg w dst) (OpReg w reg)) MO_FW_Bitcast w -> return $ Any (intFormat w) (\dst -> code `snocOL` MOV (OpReg w dst) (OpReg w reg)) @@ -2205,8 +2205,6 @@ makeFarBranches {- only used when debugging -} _platform statics basic_blocks = LDRU {} -> 1 FENCE {} -> 1 FCVT {} -> 1 - SCVTF {} -> 1 - FCVTZS {} -> 1 FABS {} -> 1 FMA {} -> 1 -- estimate the subsituted size for jumps to lables ===================================== compiler/GHC/CmmToAsm/RV64/Instr.hs ===================================== @@ -1,3 +1,6 @@ +-- All instructions will be rendered eventually. Thus, there's no benefit in +-- being lazy in data types. +{-# LANGUAGE StrictData #-} {-# OPTIONS_GHC -fno-warn-orphans #-} module GHC.CmmToAsm.RV64.Instr where @@ -101,9 +104,7 @@ regUsageOfInstr platform instr = case instr of LDR _ dst src -> usage (regOp src, regOp dst) LDRU _ dst src -> usage (regOp src, regOp dst) FENCE _ _ -> usage ([], []) - FCVT dst src -> usage (regOp src, regOp dst) - SCVTF dst src -> usage (regOp src, regOp dst) - FCVTZS dst src -> usage (regOp src, regOp dst) + FCVT _variant dst src -> usage (regOp src, regOp dst) FABS dst src -> usage (regOp src, regOp dst) FMA _ dst src1 src2 src3 -> usage (regOp src1 ++ regOp src2 ++ regOp src3, regOp dst) @@ -186,9 +187,7 @@ patchRegsOfInstr instr env = case instr of LDR f o1 o2 -> LDR f (patchOp o1) (patchOp o2) LDRU f o1 o2 -> LDRU f (patchOp o1) (patchOp o2) FENCE o1 o2 -> FENCE o1 o2 - FCVT o1 o2 -> FCVT (patchOp o1) (patchOp o2) - SCVTF o1 o2 -> SCVTF (patchOp o1) (patchOp o2) - FCVTZS o1 o2 -> FCVTZS (patchOp o1) (patchOp o2) + FCVT variant o1 o2 -> FCVT variant (patchOp o1) (patchOp o2) FABS o1 o2 -> FABS (patchOp o1) (patchOp o2) FMA s o1 o2 o3 o4 -> FMA s (patchOp o1) (patchOp o2) (patchOp o3) (patchOp o4) @@ -588,14 +587,8 @@ data Instr -- -- Memory barrier. FENCE FenceType FenceType - | -- | Floating point ConVerT - FCVT Operand Operand - | -- | Signed floating point ConVerT - SCVTF Operand Operand - | -- TODO: Same as SCVTF? - - -- | Floating point ConVerT to Zero Signed - FCVTZS Operand Operand + | -- | Floating point conversion + FCVT FcvtVariant Operand Operand | -- | Floating point ABSolute value FABS Operand Operand | -- | Floating-point fused multiply-add instructions @@ -609,6 +602,9 @@ data Instr -- | Operand of a FENCE instruction (@r@, @w@ or @rw@) data FenceType = FenceRead | FenceWrite | FenceReadWrite +-- | Variant of a floating point conversion instruction +data FcvtVariant = FloatToFloat | IntToFloat | FloatToInt + instrCon :: Instr -> String instrCon i = case i of @@ -649,8 +645,6 @@ instrCon i = BCOND {} -> "BCOND" FENCE {} -> "FENCE" FCVT {} -> "FCVT" - SCVTF {} -> "SCVTF" - FCVTZS {} -> "FCVTZS" FABS {} -> "FABS" FMA variant _ _ _ _ -> case variant of ===================================== compiler/GHC/CmmToAsm/RV64/Ppr.hs ===================================== @@ -646,23 +646,23 @@ pprInstr platform instr = case instr of LDRU FF64 o1 o2@(OpAddr (AddrRegImm _ _)) -> op2 (text "\tfld") o1 o2 LDRU f o1 o2 -> pprPanic "Unsupported unsigned load" ((text . show) f <+> pprOp platform o1 <+> pprOp platform o2) FENCE r w -> line $ text "\tfence" <+> pprFenceType r <> char ',' <+> pprFenceType w - FCVT o1@(OpReg W32 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.s.d") o1 o2 - FCVT o1@(OpReg W64 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.d.s") o1 o2 - FCVT o1 o2 -> - pprPanic "RV64.pprInstr - impossible float conversion" + FCVT FloatToFloat o1@(OpReg W32 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.s.d") o1 o2 + FCVT FloatToFloat o1@(OpReg W64 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.d.s") o1 o2 + FCVT FloatToFloat o1 o2 -> + pprPanic "RV64.pprInstr - impossible float to float conversion" $ line (pprOp platform o1 <> text "->" <> pprOp platform o2) - SCVTF o1@(OpReg W32 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.s.w") o1 o2 - SCVTF o1@(OpReg W32 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.s.l") o1 o2 - SCVTF o1@(OpReg W64 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.d.w") o1 o2 - SCVTF o1@(OpReg W64 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.d.l") o1 o2 - SCVTF o1 o2 -> + FCVT IntToFloat o1@(OpReg W32 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.s.w") o1 o2 + FCVT IntToFloat o1@(OpReg W32 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.s.l") o1 o2 + FCVT IntToFloat o1@(OpReg W64 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.d.w") o1 o2 + FCVT IntToFloat o1@(OpReg W64 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.d.l") o1 o2 + FCVT IntToFloat o1 o2 -> pprPanic "RV64.pprInstr - impossible integer to float conversion" $ line (pprOp platform o1 <> text "->" <> pprOp platform o2) - FCVTZS o1@(OpReg W32 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.w.s") o1 o2 - FCVTZS o1@(OpReg W32 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.w.d") o1 o2 - FCVTZS o1@(OpReg W64 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.l.s") o1 o2 - FCVTZS o1@(OpReg W64 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.l.d") o1 o2 - FCVTZS o1 o2 -> + FCVT FloatToInt o1@(OpReg W32 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.w.s") o1 o2 + FCVT FloatToInt o1@(OpReg W32 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.w.d") o1 o2 + FCVT FloatToInt o1@(OpReg W64 _) o2@(OpReg W32 _) -> op2 (text "\tfcvt.l.s") o1 o2 + FCVT FloatToInt o1@(OpReg W64 _) o2@(OpReg W64 _) -> op2 (text "\tfcvt.l.d") o1 o2 + FCVT FloatToInt o1 o2 -> pprPanic "RV64.pprInstr - impossible float to integer conversion" $ line (pprOp platform o1 <> text "->" <> pprOp platform o2) FABS o1 o2 | isSingleOp o2 -> op2 (text "\tfabs.s") o1 o2 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a38bf4541dc3a2888da1f03bffe906ff5ff5ac63...26c79cfa3cee3cf0651f91b7b75d00268e985e91 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a38bf4541dc3a2888da1f03bffe906ff5ff5ac63...26c79cfa3cee3cf0651f91b7b75d00268e985e91 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 1 20:09:19 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Sun, 01 Sep 2024 16:09:19 -0400 Subject: [Git][ghc/ghc][wip/T25029] 35 commits: Extend -reexported-module flag to support module renaming Message-ID: <66d4c9ef55c7c_bef2712eb0c1522e@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25029 at Glasgow Haskell Compiler / GHC Commits: ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 0a3f80bd by Simon Peyton Jones at 2024-09-01T22:04:20+02:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 335366bf by Simon Peyton Jones at 2024-09-01T22:04:20+02:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - d0fe1866 by Simon Peyton Jones at 2024-09-01T22:09:03+02:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - 30 changed files: - .gitignore - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Core/InstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Predicate.hs - compiler/GHC/Core/Type.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Config/Finder.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Errors/Types.hs - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Session.hs - compiler/GHC/Hs/Decls.hs - compiler/GHC/Hs/ImpExp.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/HsToCore.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b655485fe887f3898736c79a068e4e240b846083...d0fe1866df362cc3ed3910d04b1921e818a2379a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b655485fe887f3898736c79a068e4e240b846083...d0fe1866df362cc3ed3910d04b1921e818a2379a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 1 20:59:55 2024 From: gitlab at gitlab.haskell.org (Matthew Craven (@clyring)) Date: Sun, 01 Sep 2024 16:59:55 -0400 Subject: [Git][ghc/ghc][wip/T22935] 5451 commits: [haddock @ 2002-04-04 16:23:43 by simonmar] Message-ID: <66d4d5cb8e5db_bef272f2894181ca@gitlab.mail> Matthew Craven pushed to branch wip/T22935 at Glasgow Haskell Compiler / GHC Commits: 2b39cd94 by Simon Marlow at 2002-04-04T16:23:43+00:00 [haddock @ 2002-04-04 16:23:43 by simonmar] This is Haddock, my stab at a Haskell documentation tool. It's not quite ready for release yet, but I'm putting it in the repository so others can take a look. It uses a locally modified version of the hssource parser, extended with support for GHC extensions and documentation annotations. - - - - - 99ede94f by Simon Marlow at 2002-04-04T16:24:10+00:00 [haddock @ 2002-04-04 16:24:10 by simonmar] forgot one file - - - - - 8363294c by Simon Marlow at 2002-04-05T13:58:15+00:00 [haddock @ 2002-04-05 13:58:15 by simonmar] Remap names in the exported declarations to be "closer" to the current module. eg. if an exported declaration mentions a type 'T' which is imported from module A then re-exported from the current module, then links from the type or indeed the documentation will point to the current module rather than module A. This is to support better hiding: module A won't be referred to in the generated output. - - - - - 1570cbc1 by Simon Marlow at 2002-04-05T13:58:23+00:00 [haddock @ 2002-04-05 13:58:23 by simonmar] update the TODO list - - - - - 3a62f96b by Simon Marlow at 2002-04-05T14:11:51+00:00 [haddock @ 2002-04-05 14:11:51 by simonmar] Fix the anchor for a class declaration - - - - - c5d9a471 by Simon Marlow at 2002-04-05T14:18:41+00:00 [haddock @ 2002-04-05 14:18:41 by simonmar] remove underlines on visited links - - - - - 97280525 by Simon Marlow at 2002-04-05T16:11:47+00:00 [haddock @ 2002-04-05 16:11:47 by simonmar] - Update to generate more correct HTML. - Use our own non-overloaded table combinators, as the overloaded versions were giving me a headache. The improved type safety caught several errors in the HTML generation. - - - - - 9acd3a4d by Simon Marlow at 2002-04-05T16:32:19+00:00 [haddock @ 2002-04-05 16:32:19 by simonmar] Add width property to the title, and add TD.children for the module contents page. - - - - - ec9a0847 by Simon Marlow at 2002-04-08T16:39:56+00:00 [haddock @ 2002-04-08 16:39:56 by simonmar] Fix a problem with exports of the form T(..). - - - - - e4627dc8 by Simon Marlow at 2002-04-08T16:41:38+00:00 [haddock @ 2002-04-08 16:41:37 by simonmar] - Add our own versions of Html & BlockTable for the time being. - Add support for generating an index to the HTML backend - - - - - 2d73fd75 by Simon Marlow at 2002-04-09T11:23:24+00:00 [haddock @ 2002-04-09 11:23:24 by simonmar] Add '-- /' as a synonym for '-- |', for compatibility with IDoc. - - - - - 3675464e by Simon Marlow at 2002-04-09T11:33:55+00:00 [haddock @ 2002-04-09 11:33:54 by simonmar] - add the <...> syntax for marking up URLs in documentation - Make the output for data & class declarations more compact when there aren't any documentation annotations on the individual methods or constructors respectively. - - - - - 5077f5b1 by Simon Marlow at 2002-04-09T11:36:04+00:00 [haddock @ 2002-04-09 11:36:04 by simonmar] Update the TODO list - - - - - 9e83c54d by Simon Marlow at 2002-04-10T10:50:06+00:00 [haddock @ 2002-04-10 10:50:06 by simonmar] Use explicit 'px' suffix on pixel sizes; IE seems to prefer them - - - - - 052de51c by Simon Marlow at 2002-04-10T13:23:13+00:00 [haddock @ 2002-04-10 13:23:13 by simonmar] Lex URLs as a single token to avoid having to escape special characters inside the URL string. - - - - - 47187edb by Simon Marlow at 2002-04-10T13:23:55+00:00 [haddock @ 2002-04-10 13:23:55 by simonmar] Not sure why I made the constructor name for a record declaration into a TyCls name, but change it back into a Var name anyhow. - - - - - 3dc6aa81 by Simon Marlow at 2002-04-10T13:26:10+00:00 [haddock @ 2002-04-10 13:26:09 by simonmar] Lots of changes, including: - add index support to the HTML backend - clean up the renamer, put it into a monad - propogate unresolved names to the top level and report them in a nicer way - various bugfixes - - - - - c2a70a72 by Simon Marlow at 2002-04-10T13:32:39+00:00 [haddock @ 2002-04-10 13:32:39 by simonmar] Skeleton documentation - - - - - 50c98d17 by Simon Marlow at 2002-04-10T13:37:23+00:00 [haddock @ 2002-04-10 13:37:23 by simonmar] Update the TODO list, separate into pre-1.0 and post-1.0 items - - - - - f3778be6 by Simon Marlow at 2002-04-10T14:30:58+00:00 [haddock @ 2002-04-10 14:30:58 by simonmar] Add an introduction - - - - - cfbaf9f7 by Simon Marlow at 2002-04-10T14:59:51+00:00 [haddock @ 2002-04-10 14:59:51 by simonmar] Sort the module tree - - - - - 76bd7b34 by Simon Marlow at 2002-04-10T15:50:11+00:00 [haddock @ 2002-04-10 15:50:10 by simonmar] Generate a little table of contents at the top of the module doc (only if the module actually contains some section headings, though). - - - - - bb8560a1 by Simon Marlow at 2002-04-10T16:10:26+00:00 [haddock @ 2002-04-10 16:10:26 by simonmar] Now we understand (or at least don't barf on) type signatures in patterns such as you might find when scoped type variables are in use. - - - - - 86c2a026 by Simon Marlow at 2002-04-10T16:10:49+00:00 [haddock @ 2002-04-10 16:10:49 by simonmar] more updates - - - - - 1c052b0e by Simon Marlow at 2002-04-10T16:28:05+00:00 [haddock @ 2002-04-10 16:28:05 by simonmar] Parse errors in doc strings are now reported as warnings rather that causing the whole thing to fall over. It still needs cleaning up (the warning is emitted with trace) but this will do for the time being. - - - - - ace03e8f by Simon Marlow at 2002-04-10T16:38:03+00:00 [haddock @ 2002-04-10 16:38:03 by simonmar] update again - - - - - 69006c3e by Simon Marlow at 2002-04-11T13:38:02+00:00 [haddock @ 2002-04-11 13:38:02 by simonmar] mention Opera - - - - - fe9b10f8 by Simon Marlow at 2002-04-11T13:40:31+00:00 [haddock @ 2002-04-11 13:40:30 by simonmar] - copy haddock.css into the same place as the generated HTML - new option: --css <file> specifies the style sheet to use - new option: -o <dir> specifies the directory in which to generate the output. - because Haddock now needs to know where to find its default stylesheet, we have to have a wrapper script and do the haddock-inplace thing (Makefile code copied largely from fptools/happy). - - - - - 106adbbe by Simon Marlow at 2002-04-24T15:12:41+00:00 [haddock @ 2002-04-24 15:12:41 by simonmar] Stop slurping comment lines when we see a row of dashes longer than length 2: these are useful as separators. - - - - - 995d3f9e by Simon Marlow at 2002-04-24T15:14:12+00:00 [haddock @ 2002-04-24 15:14:11 by simonmar] Grok the kind of module headers we use in fptools/libraries, and pass the "portability", "stability", and "maintainer" strings through into the generated HTML. If the module header doesn't match the pattern, then we don't include the info in the HTML. - - - - - e14da136 by Simon Marlow at 2002-04-24T15:16:57+00:00 [haddock @ 2002-04-24 15:16:57 by simonmar] Done module headers now. - - - - - 2ca8dfd4 by Simon Marlow at 2002-04-24T15:57:48+00:00 [haddock @ 2002-04-24 15:57:47 by simonmar] Handle gcons in export lists (a common extension). - - - - - 044cea81 by Simon Marlow at 2002-04-25T14:20:12+00:00 [haddock @ 2002-04-25 14:20:12 by simonmar] Add the little lambda icon - - - - - 63955027 by Simon Marlow at 2002-04-25T14:40:05+00:00 [haddock @ 2002-04-25 14:40:05 by simonmar] - Add support for named chunks of documentation which can be referenced from the export list. - Copy the icon from $libdir to the destination in HTML mode. - - - - - 36e3f913 by Simon Marlow at 2002-04-25T16:48:36+00:00 [haddock @ 2002-04-25 16:48:36 by simonmar] More keyboard bashing - - - - - 7ae18dd0 by Simon Marlow at 2002-04-26T08:43:33+00:00 [haddock @ 2002-04-26 08:43:33 by simonmar] Package util reqd. to compile with 4.08.2 - - - - - bbd5fbab by Simon Marlow at 2002-04-26T10:13:00+00:00 [haddock @ 2002-04-26 10:13:00 by simonmar] Include $(GHC_HAPPY_OPTS) when compiling HsParser - - - - - 31c53d79 by Simon Marlow at 2002-04-26T11:18:57+00:00 [haddock @ 2002-04-26 11:18:56 by simonmar] - support for fundeps (partially contributed by Brett Letner - thanks Brett). - make it build with GHC 4.08.2 - - - - - c415ce76 by Simon Marlow at 2002-04-26T13:15:02+00:00 [haddock @ 2002-04-26 13:15:02 by simonmar] Move the explicit formatting of the little table for the stability/portability/maintainer info from the HTML into the CSS, and remove the explicit table size (just right-align it). - - - - - 520ee21a by Simon Marlow at 2002-04-26T16:01:44+00:00 [haddock @ 2002-04-26 16:01:44 by simonmar] Yet more keyboard bashing - this is pretty much complete now. - - - - - 2ae37179 by Simon Marlow at 2002-04-26T16:02:14+00:00 [haddock @ 2002-04-26 16:02:14 by simonmar] Add a couple of things I forgot about - - - - - b7211e04 by Simon Marlow at 2002-04-29T15:28:12+00:00 [haddock @ 2002-04-29 15:28:12 by simonmar] bugfix for declBinders on a NewTypeDecl - - - - - 640c154a by Simon Marlow at 2002-04-29T15:28:54+00:00 [haddock @ 2002-04-29 15:28:54 by simonmar] Allow '-- |' style annotations on constructors and record fields. - - - - - 393f258a by Simon Marlow at 2002-04-29T15:37:32+00:00 [haddock @ 2002-04-29 15:37:32 by simonmar] syntax fix - - - - - 8a2c2549 by Simon Marlow at 2002-04-29T15:37:48+00:00 [haddock @ 2002-04-29 15:37:48 by simonmar] Add an example - - - - - db88f8a2 by Simon Marlow at 2002-04-29T15:55:46+00:00 [haddock @ 2002-04-29 15:55:46 by simonmar] remove a trace - - - - - 2b0248e0 by Simon Marlow at 2002-04-29T15:56:19+00:00 [haddock @ 2002-04-29 15:56:19 by simonmar] Fix for 'make install' - - - - - 120453a0 by Simon Marlow at 2002-04-29T15:56:39+00:00 [haddock @ 2002-04-29 15:56:39 by simonmar] Install the auxilliary bits - - - - - 950e6dbb by Simon Marlow at 2002-04-29T15:57:30+00:00 [haddock @ 2002-04-29 15:57:30 by simonmar] Add BinDist bits - - - - - 154b9d71 by Simon Marlow at 2002-05-01T11:02:52+00:00 [haddock @ 2002-05-01 11:02:52 by simonmar] update - - - - - ba6c39fa by Simon Marlow at 2002-05-01T11:03:26+00:00 [haddock @ 2002-05-01 11:03:26 by simonmar] Add another item - - - - - bacb5e33 by Simon Marlow at 2002-05-03T08:50:00+00:00 [haddock @ 2002-05-03 08:50:00 by simonmar] Fix some typos. - - - - - 54c87895 by Sven Panne at 2002-05-05T19:40:51+00:00 [haddock @ 2002-05-05 19:40:51 by panne] As a temporary hack/workaround for a bug in GHC's simplifier, don't pass Happy the -c option for generating the parsers in this subdir. Furthermore, disable -O for HaddocParse, too. - - - - - e6c08703 by Simon Marlow at 2002-05-06T09:51:10+00:00 [haddock @ 2002-05-06 09:51:10 by simonmar] Add RPM spec file (thanks to Tom Moertel <tom-rpms at moertel.com>) - - - - - 7b8fa8e7 by Simon Marlow at 2002-05-06T12:29:26+00:00 [haddock @ 2002-05-06 12:29:26 by simonmar] Add missing type signature (a different workaround for the bug in GHC's simplifier). - - - - - cd0e300d by Simon Marlow at 2002-05-06T12:30:09+00:00 [haddock @ 2002-05-06 12:30:09 by simonmar] Remove workaround for simplifier bug in previous revision. - - - - - 687e68fa by Simon Marlow at 2002-05-06T12:32:32+00:00 [haddock @ 2002-05-06 12:32:32 by simonmar] Allow empty data declarations (another GHC extension). - - - - - 8f29f696 by Simon Marlow at 2002-05-06T12:49:21+00:00 [haddock @ 2002-05-06 12:49:21 by simonmar] Fix silly bug in named documentation block lookup. - - - - - 8e0059af by Simon Marlow at 2002-05-06T13:02:42+00:00 [haddock @ 2002-05-06 13:02:42 by simonmar] Add another named chunk with a different name - - - - - 68f8a896 by Simon Marlow at 2002-05-06T13:32:32+00:00 [haddock @ 2002-05-06 13:32:32 by simonmar] Be more lenient about extra paragraph breaks - - - - - 65fc31db by Simon Marlow at 2002-05-07T15:36:36+00:00 [haddock @ 2002-05-07 15:36:36 by simonmar] DocEmpty is a right and left-unit of DocAppend (remove it in the smart constructor). - - - - - adc81078 by Simon Marlow at 2002-05-07T15:37:15+00:00 [haddock @ 2002-05-07 15:37:15 by simonmar] Allow code blocks to be denoted with bird-tracks in addition to [...]. - - - - - 1283a3c1 by Simon Marlow at 2002-05-08T11:21:56+00:00 [haddock @ 2002-05-08 11:21:56 by simonmar] Add a facility for specifying options that affect Haddock's treatment of the module. Options are given at the top of the module in a comma-separated list, beginning with '-- #'. eg. -- # prune, hide, ignore-exports Options currently available, with their meanings: prune: ignore declarations which have no documentation annotations ignore-exports: act as if the export list were not specified (i.e. export everything local to the module). hide: do not include this module in the generated documentation, but propagate any exported definitions to modules which re-export them. There's a slight change in the semantics for re-exporting a full module by giving 'module M' in the export list: if module M does not have the 'hide' option, then the documentation will now just contain a reference to module M rather than the full inlined contents of that module. These features, and some other changes in the pipeline, are the result of discussions between myself and Manuel Chakravarty <chak at cse.unsw.edu.au> (author of IDoc) yesterday. Also: some cleanups, use a Writer monad to collect error messages in some places instead of just printing them with trace. - - - - - a2239cf5 by Simon Marlow at 2002-05-08T11:22:30+00:00 [haddock @ 2002-05-08 11:22:30 by simonmar] Update to test new features. - - - - - 6add955f by Simon Marlow at 2002-05-08T13:37:25+00:00 [haddock @ 2002-05-08 13:37:25 by simonmar] Change the markup for typewriter-font from [...] to @... at . The reasoning is that the '@' symbol is much less likely to be needed than square brackets, and we don't want to have to escape square brackets in code fragments. This will be mildly painful in the short term, but it's better to get the change out of the way as early as possible. - - - - - cda06447 by Simon Marlow at 2002-05-08T13:39:56+00:00 [haddock @ 2002-05-08 13:39:56 by simonmar] Allow nested-style comments to be used as documentation annotations too. eg. {-| ... -} is equivalent to -- | ... An extra space can also be left after the comment opener: {- | ... -}. The only version that isn't allowed is {-# ... -}, because this syntax overlaps with Haskell pragmas; use {- # ... -} instead. - - - - - db23f65e by Simon Marlow at 2002-05-08T14:48:41+00:00 [haddock @ 2002-05-08 14:48:39 by simonmar] Add support for existential quantifiers on constructors. - - - - - adce3794 by Simon Marlow at 2002-05-08T15:43:25+00:00 [haddock @ 2002-05-08 15:43:25 by simonmar] update - - - - - 62a1f436 by Simon Marlow at 2002-05-08T15:44:10+00:00 [haddock @ 2002-05-08 15:44:10 by simonmar] Update to version 0.2 - - - - - f6a24ba3 by Simon Marlow at 2002-05-09T08:48:29+00:00 [haddock @ 2002-05-09 08:48:29 by simonmar] typo - - - - - 9f9522a4 by Simon Marlow at 2002-05-09T10:33:14+00:00 [haddock @ 2002-05-09 10:33:14 by simonmar] oops, left out '/' from the special characters in the last change. - - - - - 14abcb39 by Simon Marlow at 2002-05-09T10:34:44+00:00 [haddock @ 2002-05-09 10:34:44 by simonmar] Fix buglet - - - - - b8d878be by Simon Marlow at 2002-05-09T10:35:00+00:00 [haddock @ 2002-05-09 10:35:00 by simonmar] Give a more useful instance of Show for Module. - - - - - f7bfd626 by Simon Marlow at 2002-05-09T10:37:07+00:00 [haddock @ 2002-05-09 10:37:07 by simonmar] The last commit to Main.lhs broke the delicate balance of laziness which was being used to avoid computing the dependency graph of modules. So I finally bit the bullet and did a proper topological sort of the module graph, which turned out to be easy (stealing the Digraph module from GHC - this really ought to be in the libraries somewhere). - - - - - b481c1d0 by Simon Marlow at 2002-05-09T10:37:25+00:00 [haddock @ 2002-05-09 10:37:25 by simonmar] another item done - - - - - 032e2b42 by Simon Marlow at 2002-05-09T10:44:15+00:00 [haddock @ 2002-05-09 10:44:15 by simonmar] Don't consider a module re-export as having documentation, for the purposes of deciding whether we need a Synopsis section or not. - - - - - 5fb45e92 by Simon Marlow at 2002-05-09T11:10:55+00:00 [haddock @ 2002-05-09 11:10:55 by simonmar] Add a special case for list types in ppHsAType - - - - - 1937e428 by Simon Marlow at 2002-05-09T12:43:06+00:00 [haddock @ 2002-05-09 12:43:06 by simonmar] Type synonyms can accept a ctype on the RHS, to match GHC. - - - - - 0f16ce56 by Simon Marlow at 2002-05-09T12:45:19+00:00 [haddock @ 2002-05-09 12:45:19 by simonmar] Add 'stdcall' keyword - - - - - 29b0d7d2 by Simon Marlow at 2002-05-09T13:35:45+00:00 [haddock @ 2002-05-09 13:35:45 by simonmar] Add System Requirements section - - - - - bf14dddd by Simon Marlow at 2002-05-09T13:36:11+00:00 [haddock @ 2002-05-09 13:36:11 by simonmar] Test existential types, amongst other things - - - - - 502f8f6f by Simon Marlow at 2002-05-09T13:37:35+00:00 [haddock @ 2002-05-09 13:37:35 by simonmar] Print the module name in a doc-string parse error - - - - - ca1f8d49 by Simon Marlow at 2002-05-09T13:38:04+00:00 [haddock @ 2002-05-09 13:38:04 by simonmar] Add dependency - - - - - 8d3d91ff by Simon Marlow at 2002-05-09T15:37:57+00:00 [haddock @ 2002-05-09 15:37:57 by simonmar] Add the changelog/release notes - - - - - f3960959 by Simon Marlow at 2002-05-09T15:47:47+00:00 [haddock @ 2002-05-09 15:47:47 by simonmar] mention the backquote-style of markup - - - - - 089fb6e6 by Simon Marlow at 2002-05-09T15:59:45+00:00 [haddock @ 2002-05-09 15:59:45 by simonmar] update - - - - - bdd3be0b by Simon Marlow at 2002-05-09T15:59:56+00:00 [haddock @ 2002-05-09 15:59:56 by simonmar] Document changes since 0.1 - - - - - 00fc4af8 by Simon Marlow at 2002-05-10T08:22:48+00:00 [haddock @ 2002-05-10 08:22:48 by simonmar] oops, update to version 0.2 - - - - - a8a79041 by Simon Marlow at 2002-05-10T16:05:08+00:00 [haddock @ 2002-05-10 16:05:08 by simonmar] Only include a mini-contents if there are 2 or more sections - - - - - 06653319 by Simon Marlow at 2002-05-13T09:13:12+00:00 [haddock @ 2002-05-13 09:13:12 by simonmar] fix typos - - - - - 1402b19b by Simon Marlow at 2002-05-13T10:14:22+00:00 [haddock @ 2002-05-13 10:14:22 by simonmar] Allow backquote as the right-hand quote as well as the left-hand quote, as suggested by Dean Herrington. Clean up the grammar a litte. - - - - - dcd5320d by Simon Marlow at 2002-05-13T10:44:10+00:00 [haddock @ 2002-05-13 10:44:10 by simonmar] a couple more things, prioritise a bit - - - - - a90130c4 by Simon Marlow at 2002-05-13T15:19:03+00:00 [haddock @ 2002-05-13 15:19:03 by simonmar] Cope with datatypes which have documentation on the constructor but not the type itself, and records which have documentation on the fields but not the constructor. (Thanks to Ross Paterson for pointing out the bugs). - - - - - a774d432 by Simon Marlow at 2002-05-13T15:20:54+00:00 [haddock @ 2002-05-13 15:20:54 by simonmar] Fix one of the record examples - - - - - 2d1d5218 by Simon Marlow at 2002-05-15T12:44:35+00:00 [haddock @ 2002-05-15 12:44:35 by simonmar] Preserve the newline before a bird-track, but only within a paragraph. - - - - - 1554c09a by Simon Marlow at 2002-05-15T13:03:02+00:00 [haddock @ 2002-05-15 13:03:01 by simonmar] Reworking of the internals to support documenting function arguments (the Most Wanted new feature by the punters). The old method of keeping parsed documentation in a Name -> Doc mapping wasn't going to cut it for anntations on type components, where there's no name to attach the documentation to, so I've moved to storing all the documentation in the abstract syntax. Previously some of the documentation was left in the abstract syntax by the parser, but was later extracted into the mapping. In order to avoid having to parameterise the abstract syntax over the type of documentation stored in it, we have to parse the documentation at the same time as we parse the Haskell source (well, I suppose we could store 'Either String Doc' in the HsSyn, but that's clunky). One upshot is that documentation is now parsed eagerly, and documentation parse errors are fatal (but have better line numbers in the error message). The new story simplifies matters for the code that processes the source modules, because we don't have to maintain the extra Name->Doc mapping, and it should improve efficiency a little too. New features: - Function arguments and return values can now have doc annotations. - If you refer to a qualified name in a doc string, eg. 'IO.putStr', then Haddock will emit a hyperlink even if the identifier is not in scope, so you don't have to make sure everything referred to from the documentation is imported. - several bugs & minor infelicities fixed. - - - - - 57344dc3 by Simon Marlow at 2002-05-15T13:03:19+00:00 [haddock @ 2002-05-15 13:03:19 by simonmar] Bump to version 0.3 - - - - - b2791812 by Simon Marlow at 2002-05-15T13:03:41+00:00 [haddock @ 2002-05-15 13:03:41 by simonmar] update - - - - - fead183e by Simon Marlow at 2002-05-15T13:10:15+00:00 [haddock @ 2002-05-15 13:10:15 by simonmar] Rename Foo.hs to Test.hs, and add a Makefile - - - - - b0b1f89f by Simon Marlow at 2002-05-15T13:16:07+00:00 [haddock @ 2002-05-15 13:16:07 by simonmar] - Remove the note about function argument docs not being implemented - Note that qualified identifiers can be used to point to entities that aren't in scope. - - - - - 5665f31a by Simon Marlow at 2002-05-15T13:28:46+00:00 [haddock @ 2002-05-15 13:28:46 by simonmar] Patch to add support for GHC-style primitive strings ".."#, from Ross Paterson. - - - - - 0564505d by Simon Marlow at 2002-05-17T10:51:57+00:00 [haddock @ 2002-05-17 10:51:57 by simonmar] Fix bugs in qualified name handling (A.B.f was returned as B.f) - - - - - 10e7311c by Simon Marlow at 2002-05-21T10:24:52+00:00 [haddock @ 2002-05-21 10:24:52 by simonmar] - Use an alternate tabular layout for datatypes, which is more compact - Fix some problems with the function argument documentation - - - - - 2f91c2a6 by Simon Marlow at 2002-05-21T10:27:40+00:00 [haddock @ 2002-05-21 10:27:40 by simonmar] add a few more test cases - - - - - 01c2ddd2 by Simon Marlow at 2002-05-21T10:28:33+00:00 [haddock @ 2002-05-21 10:28:33 by simonmar] Rearrange a bit, and add support for tabular datatype rendering - - - - - a4e4c5f8 by Simon Marlow at 2002-05-27T09:03:52+00:00 [haddock @ 2002-05-27 09:03:51 by simonmar] Lots of changes: - instances of a class are listed with the class, and instances involving a datatype are listed with that type. Derived instances aren't included at the moment: the calculation to find the instance head for a derived instance is non-trivial. - some formatting changes; use rows with specified height rather than cellspacing in some places. - various fixes (source file links were wrong, amongst others) - - - - - 48722e68 by Simon Marlow at 2002-05-27T12:30:38+00:00 [haddock @ 2002-05-27 12:30:37 by simonmar] - Put function arguments *before* the doc for the function, as suggested by Sven Panne. This looks nicer when the function documentation is long. - Switch to using bold for binders at the definition site, and use underline for keywords. This makes the binder stand out more. - - - - - 657204d2 by Simon Marlow at 2002-05-27T13:19:49+00:00 [haddock @ 2002-05-27 13:19:49 by simonmar] Fix bug: we weren't renaming HsDocCommentNamed in renameDecl - - - - - 592aae66 by Simon Marlow at 2002-05-27T14:10:27+00:00 [haddock @ 2002-05-27 14:10:27 by simonmar] Fix some bugs in the rendering of qualified type signatures. - - - - - 69c8f763 by Simon Marlow at 2002-05-27T14:36:45+00:00 [haddock @ 2002-05-27 14:36:45 by simonmar] warning message tweak - - - - - 16e64e21 by Simon Marlow at 2002-05-27T14:53:53+00:00 [haddock @ 2002-05-27 14:53:53 by simonmar] hyperlinked identifiers should be in <tt> - - - - - 8d5e4783 by Simon Marlow at 2002-05-27T15:56:45+00:00 [haddock @ 2002-05-27 15:56:45 by simonmar] Do something sensible for modules which don't export anything (except instances). - - - - - 9d3ef811 by Simon Marlow at 2002-05-28T10:12:50+00:00 [haddock @ 2002-05-28 10:12:50 by simonmar] Rename the module documentation properly (bug reported by Sven Panne). - - - - - ef03a1cc by Simon Marlow at 2002-05-28T10:13:04+00:00 [haddock @ 2002-05-28 10:13:04 by simonmar] Add some more test cases - - - - - 92baa0e8 by Simon Marlow at 2002-05-28T11:17:55+00:00 [haddock @ 2002-05-28 11:17:55 by simonmar] If an identifier doesn't lex, then just replace it by a DocString. - - - - - a3156213 by Simon Marlow at 2002-05-28T16:16:19+00:00 [haddock @ 2002-05-28 16:16:19 by simonmar] Only link to names in the current module which are actually listed in the documentation. A name may be exported but not present in the documentation if it is exported as part of a 'module M' export specifier. - - - - - 31acf941 by Simon Marlow at 2002-05-28T16:17:11+00:00 [haddock @ 2002-05-28 16:17:11 by simonmar] update - - - - - 7e474ebf by Sigbjorn Finne at 2002-05-28T22:42:08+00:00 [haddock @ 2002-05-28 22:42:08 by sof] Handle lone occurrences of '/', e.g., -- | This/that. [did this in the lexer rather than in the parser, as I couldn't see a way not to introduce an S/R conflict that way.] - - - - - 093f7e53 by Simon Marlow at 2002-05-29T09:09:49+00:00 [haddock @ 2002-05-29 09:09:49 by simonmar] Back out previous change until we can find a better way to do this. - - - - - 9234389c by Simon Marlow at 2002-05-29T13:19:06+00:00 [haddock @ 2002-05-29 13:19:06 by simonmar] Make the markup syntax a little more friendly: - single quotes are now interpreted literally unless they surround a valid Haskell identifier. So for example now there's no need to escape a single quote used as an apostrophe. - text to the right of a bird track is now literal (if you want marked-up text in a code block, use @...@). - - - - - b3333526 by Simon Marlow at 2002-05-29T13:38:51+00:00 [haddock @ 2002-05-29 13:38:51 by simonmar] Document recent changes to markup syntax - - - - - f93641d6 by Simon Marlow at 2002-05-29T15:27:18+00:00 [haddock @ 2002-05-29 15:27:18 by simonmar] Include the instances in abstract data types too - - - - - 613f21e3 by Simon Marlow at 2002-06-03T13:05:58+00:00 [haddock @ 2002-06-03 13:05:57 by simonmar] Allow exporting of individual class methods and record selectors. For these we have to invent the correct type signature, which we do in the simplest possible way (i.e. no context reduction nonsense in the class case). - - - - - 14b36807 by Simon Marlow at 2002-06-03T13:20:00+00:00 [haddock @ 2002-06-03 13:20:00 by simonmar] Fix linking to qualified names again (thanks to Sven Panne for pointing out the bug). - - - - - 95b10eac by Simon Marlow at 2002-06-03T13:46:48+00:00 [haddock @ 2002-06-03 13:46:48 by simonmar] Fix for exporting record selectors from a newtype declaration - - - - - 272f932e by Simon Marlow at 2002-06-03T13:56:38+00:00 [haddock @ 2002-06-03 13:56:38 by simonmar] update to version 0.3 - - - - - 1c0a3bed by Simon Marlow at 2002-06-03T14:05:07+00:00 [haddock @ 2002-06-03 14:05:07 by simonmar] Add changes in version 0.3 - - - - - 145b4626 by Simon Marlow at 2002-06-03T14:12:38+00:00 [haddock @ 2002-06-03 14:12:38 by simonmar] Render class names as proper binders - - - - - 052106b3 by Simon Marlow at 2002-06-03T14:15:10+00:00 [haddock @ 2002-06-03 14:15:10 by simonmar] update, and separate into bugs, features, and cosmetic items. - - - - - 854f4914 by Simon Marlow at 2002-06-03T14:16:13+00:00 [haddock @ 2002-06-03 14:16:13 by simonmar] More test cases - - - - - 466922c8 by Simon Marlow at 2002-06-03T14:16:56+00:00 [haddock @ 2002-06-03 14:16:56 by simonmar] Example from the paper - - - - - 9962a045 by Simon Marlow at 2002-06-03T14:17:49+00:00 [haddock @ 2002-06-03 14:17:49 by simonmar] A debugging version of the style-sheet, which gives some tables coloured backgrounds so we can see what's going on. - - - - - f16b79db by Simon Marlow at 2002-06-03T14:19:46+00:00 [haddock @ 2002-06-03 14:19:46 by simonmar] typo - - - - - 620db27b by Simon Marlow at 2002-06-03T14:48:32+00:00 [haddock @ 2002-06-03 14:48:32 by simonmar] oops, fix markup bugs - - - - - 53fd105c by Simon Marlow at 2002-06-05T09:05:07+00:00 [haddock @ 2002-06-05 09:05:07 by simonmar] Keep foreign imports when there is no export list (bug reported by Sven Panne). - - - - - 6d98989c by Simon Marlow at 2002-06-05T09:12:02+00:00 [haddock @ 2002-06-05 09:12:02 by simonmar] Identifiers in single quotes can be symbol names too (bug reported by Hal Daume). - - - - - 001811e5 by Sven Panne at 2002-06-08T14:03:36+00:00 [haddock @ 2002-06-08 14:03:36 by panne] Tiny workaround for the fact that Haddock currently ignores HsImportSpecs: Let the local_orig_env take precedence. This is no real solution at all, but improves things sometimes, e.g. in my GLUT documentation. :-) - - - - - 504d19c9 by Simon Marlow at 2002-06-11T09:23:25+00:00 [haddock @ 2002-06-11 09:23:25 by simonmar] portability nit - - - - - e13b5af4 by Simon Marlow at 2002-06-20T12:38:07+00:00 [haddock @ 2002-06-20 12:38:07 by simonmar] Empty declaration fixes. - - - - - f467a9b6 by Simon Marlow at 2002-06-20T12:39:02+00:00 [haddock @ 2002-06-20 12:39:01 by simonmar] Add support for a "prologue" - a description for the whole library, placed on the contents page before the module list. - - - - - b8dbfe20 by Simon Marlow at 2002-06-21T12:43:06+00:00 [haddock @ 2002-06-21 12:43:06 by simonmar] When we have a single code block paragraph, don't place it in <pre>..</pre>, just use <tt>..</tt> to avoid generating extra vertical white space in some browsers. - - - - - 4831dbbd by Simon Marlow at 2002-06-21T15:50:42+00:00 [haddock @ 2002-06-21 15:50:42 by simonmar] Add support for reading and writing interface files(!) This turned out to be quite easy, and necessary to get decent hyperlinks between the documentation for separate packages in the libraries. The functionality isn't quite complete yet: for a given package of modules, you'd like to say "the HTML for these modules lives in directory <dir>" (currently they are assumed to be all in the same place). Two new flags: --dump-interface=FILE dump an interface file in FILE --read-interface=FILE read interface from FILE an interface file describes *all* the modules being processed. Only the exported names are kept in the interface: if you re-export a name from a module in another interface the signature won't be copied. This is a compromise to keep the size of the interfaces sensible. Also, I added another useful option: --no-implicit-prelude avoids trying to import the Prelude. Previously this was the default, but now importing the Prelude from elsewhere makes sense if you also read in an interface containing the Prelude module, so Haddock imports the Prelude implicitly according to the Haskell spec. - - - - - d3640a19 by Sven Panne at 2002-06-23T14:54:00+00:00 [haddock @ 2002-06-23 14:54:00 by panne] Make it compile with newer GHCs - - - - - 780c506b by Sven Panne at 2002-06-23T15:44:31+00:00 [haddock @ 2002-06-23 15:44:31 by panne] Cleaned up build root handling and added more docs - - - - - 45290d2e by Simon Marlow at 2002-06-24T14:37:43+00:00 [haddock @ 2002-06-24 14:37:42 by simonmar] When reading an interface, allow a file path offset to be specified which represents the path to the HTML files for the modules specified by that interface. The path may be either relative (to the location of the HTML for this package), or absolute. The syntax is --read-interface=PATH,FILE where PATH is the path to the HTML, and FILE is the filename containing the interface. - - - - - 4e2b9ae6 by Simon Marlow at 2002-07-03T16:01:08+00:00 [haddock @ 2002-07-03 16:01:07 by simonmar] Handle import specs properly, include 'hiding'. Haddock now has a complete implementation of the Haskell module system (more or less; I won't claim it's 100% correct). - - - - - 9a9aa1a8 by Simon Marlow at 2002-07-03T16:18:16+00:00 [haddock @ 2002-07-03 16:18:16 by simonmar] Update - - - - - 560c3026 by Simon Marlow at 2002-07-04T14:56:10+00:00 [haddock @ 2002-07-04 14:56:10 by simonmar] Clean up the code that constructs the exported declarations, and fix a couple of bugs along the way. Now if you import a class hiding one of the methods, then re-export the class, the version in the documentation will correctly have the appropriate method removed. - - - - - 2c26e77d by Simon Marlow at 2002-07-04T15:26:13+00:00 [haddock @ 2002-07-04 15:26:13 by simonmar] More bugfixes to the export handling - - - - - 03e0710d by Simon Marlow at 2002-07-09T10:12:10+00:00 [haddock @ 2002-07-09 10:12:10 by simonmar] Don't require that the list type comes from "Prelude" for it to be treated as special syntax (sometimes it comes from Data.List or maybe even GHC.Base). - - - - - 44f3891a by Simon Marlow at 2002-07-09T10:12:51+00:00 [haddock @ 2002-07-09 10:12:51 by simonmar] commented-out debugging code - - - - - 97280873 by Krasimir Angelov at 2002-07-09T16:33:33+00:00 [haddock @ 2002-07-09 16:33:31 by krasimir] 'Microsoft HTML Help' support - - - - - 3dc04655 by Simon Marlow at 2002-07-10T09:40:56+00:00 [haddock @ 2002-07-10 09:40:56 by simonmar] Fix for rendering of the (->) type constructor, from Ross Paterson. - - - - - c9f149c6 by Simon Marlow at 2002-07-10T10:26:11+00:00 [haddock @ 2002-07-10 10:26:11 by simonmar] Tweaks to the MS Help support: the extra files are now only generated if you ask for them (--ms-help). - - - - - e8acc1e6 by Simon Marlow at 2002-07-10T10:57:10+00:00 [haddock @ 2002-07-10 10:57:10 by simonmar] Document all the new options since 0.3 - - - - - 8bb85544 by Simon Marlow at 2002-07-10T10:58:31+00:00 [haddock @ 2002-07-10 10:58:31 by simonmar] Sort the options a bit - - - - - abc0dd59 by Simon Marlow at 2002-07-15T09:19:38+00:00 [haddock @ 2002-07-15 09:19:38 by simonmar] Fix a bug in mkExportItems when processing a module without an explicit export list. We were placing one copy of a declaration for each binder in the declaration, which for a data type would mean one copy of the whole declaration per constructor or record selector. - - - - - dde65bb9 by Simon Marlow at 2002-07-15T09:54:16+00:00 [haddock @ 2002-07-15 09:54:16 by simonmar] merge rev. 1.35 - - - - - bd7eb8c4 by Simon Marlow at 2002-07-15T10:14:31+00:00 [haddock @ 2002-07-15 10:14:30 by simonmar] Be a bit more liberal in the kind of commenting styles we allow, as suggested by Malcolm Wallace. Mostly this consists of allowing doc comments either side of a separator token. In an export list, a section heading is now allowed before the comma, as well as after it. eg. module M where ( T(..) -- * a section heading , f -- * another section heading , g ) In record fields, doc comments are allowed anywhere (previously a doc-next was allowed only after the comma, and a doc-before was allowed only before the comma). eg. data R = C { -- | describes 'f' f :: Int -- | describes 'g' , g :: Int } - - - - - 8f6dfe34 by Simon Marlow at 2002-07-15T10:21:56+00:00 [haddock @ 2002-07-15 10:21:56 by simonmar] Mention alternative commenting styles. - - - - - fc515bb7 by Simon Marlow at 2002-07-15T16:16:50+00:00 [haddock @ 2002-07-15 16:16:50 by simonmar] Allow multiple sections/subsections before and after a comma in the export list. Also at the same time I made the syntax a little stricter (multiple commas now aren't allowed between export specs). - - - - - 80a97e74 by Simon Marlow at 2002-07-19T09:13:10+00:00 [haddock @ 2002-07-19 09:13:10 by simonmar] Allow special id's ([], (), etc.) to be used in an import declaration. - - - - - a69d7378 by Simon Marlow at 2002-07-19T09:59:02+00:00 [haddock @ 2002-07-19 09:59:02 by simonmar] Allow special id's ([], (), etc.) to be used in an import declarations. - - - - - d205fa60 by Simon Marlow at 2002-07-19T10:00:16+00:00 [haddock @ 2002-07-19 10:00:16 by simonmar] Relax the restrictions which require doc comments to be followed by semi colons - in some cases this isn't necessary. Now you can write module M where { -- | some doc class C where {} } without needing to put a semicolon before the class declaration. - - - - - e9301e14 by Simon Marlow at 2002-07-23T08:24:09+00:00 [haddock @ 2002-07-23 08:24:09 by simonmar] A new TODO list item - - - - - e5d77586 by Simon Marlow at 2002-07-23T08:40:56+00:00 [haddock @ 2002-07-23 08:40:56 by simonmar] - update the acknowledgements - remove the paragraph that described how to use explicit layout with doc comments; it isn't relevant any more. - - - - - 78a94137 by Simon Marlow at 2002-07-23T08:43:02+00:00 [haddock @ 2002-07-23 08:43:02 by simonmar] more tests - - - - - 5c320927 by Simon Marlow at 2002-07-23T08:43:26+00:00 [haddock @ 2002-07-23 08:43:26 by simonmar] Updates for version 0.4 - - - - - 488e99ae by Simon Marlow at 2002-07-23T09:10:46+00:00 [haddock @ 2002-07-23 09:10:46 by simonmar] Fix the %changelog (rpm complained that it wasn't in the right order) - - - - - a77bb373 by Simon Marlow at 2002-07-23T09:12:38+00:00 [haddock @ 2002-07-23 09:12:38 by simonmar] Another item for the TODO list - - - - - f1ec1813 by Simon Marlow at 2002-07-23T10:18:46+00:00 [haddock @ 2002-07-23 10:18:46 by simonmar] Add a version banner when invoked with -v - - - - - 1d44cadf by Simon Marlow at 2002-07-24T09:28:19+00:00 [haddock @ 2002-07-24 09:28:19 by simonmar] Remove ^Ms - - - - - 4d8d5e94 by Simon Marlow at 2002-07-24T09:42:18+00:00 [haddock @ 2002-07-24 09:42:17 by simonmar] Patches to quieten ghc -Wall, from those nice folks at Galois. - - - - - d6edc43e by Simon Marlow at 2002-07-25T14:37:29+00:00 [haddock @ 2002-07-25 14:37:28 by simonmar] Patch to allow simple hyperlinking to an arbitrary location in another module's documentation, from Volker Stolz. Now in a doc comment: #foo# creates <a name="foo"></a> And you can use the form "M\#foo" to hyperlink to the label 'foo' in module 'M'. Note that the backslash is necessary for now. - - - - - b34d18fa by Simon Marlow at 2002-08-02T09:08:22+00:00 [haddock @ 2002-08-02 09:08:22 by simonmar] The <TT> and <PRE> environments seem to use a font that is a little too small in IE. Compensate. (suggestion from Daan Leijen). - - - - - 8106b086 by Simon Marlow at 2002-08-02T09:25:23+00:00 [haddock @ 2002-08-02 09:25:20 by simonmar] Remove <P>..</P> from around list items, to reduce excess whitespace between the items of bulleted and ordered lists. (Suggestion from Daan Leijen). - - - - - c1acff8f by Simon Marlow at 2002-08-05T09:03:49+00:00 [haddock @ 2002-08-05 09:03:49 by simonmar] update - - - - - f968661c by Simon Marlow at 2002-11-11T09:32:57+00:00 [haddock @ 2002-11-11 09:32:57 by simonmar] Fix cut-n-pasto - - - - - 12d02619 by Simon Marlow at 2002-11-13T09:49:46+00:00 [haddock @ 2002-11-13 09:49:46 by simonmar] Small bugfix in the --read-interface option parsing from Brett Letner. - - - - - 30e32d5e by Ross Paterson at 2003-01-16T15:07:57+00:00 [haddock @ 2003-01-16 15:07:57 by ross] Adjust for the new exception libraries (as well as the old ones). - - - - - 871f65df by Sven Panne at 2003-02-20T21:31:40+00:00 [haddock @ 2003-02-20 21:31:40 by panne] * Add varsyms and consyms to index * Exclude empty entries from index - - - - - bc42cc87 by Sven Panne at 2003-02-24T21:26:29+00:00 [haddock @ 2003-02-24 21:26:29 by panne] Don't convert a "newtype" to a single-constructor "data" for non-abstractly exported types, they are quite different regarding strictness/pattern matching. Now a "data" without any constructors is only emitted for an abstractly exported type, regardless if it is actually a "newtype" or a "data". - - - - - 0c2a1d99 by Sven Panne at 2003-03-08T19:02:38+00:00 [haddock @ 2003-03-08 19:02:38 by panne] Fixed some broken/redirected/canonicalized links found by a very picky link checker. - - - - - 25459269 by Sven Panne at 2003-03-09T21:13:43+00:00 [haddock @ 2003-03-09 21:13:43 by panne] Don't append a fragment to non-defining index entries, only documents with a defining occurrence have a name anchor. - - - - - 6be4db86 by Sven Panne at 2003-03-10T21:34:25+00:00 [haddock @ 2003-03-10 21:34:24 by panne] Escape fragments. This fixes e.g. links to operators. - - - - - eb12972c by Ross Paterson at 2003-04-25T10:50:06+00:00 [haddock @ 2003-04-25 10:50:05 by ross] An 80% solution to generating derived instances. A complete solution would duplicate the instance inference logic, but if a type variable occurs as a constructor argument, then we can just propagate the derived class to the variable. But we know nothing of the constraints on any type variables that occur elsewhere. For example, the declarations data Either a b = Left a | Right b deriving (Eq, Ord) data Ptr a = Ptr Addr# deriving (Eq, Ord) newtype IORef a = IORef (STRef RealWorld a) deriving Eq yield the instances (Eq a, Eq b) => Eq (Either a b) (Ord a, Ord b) => Ord (Either a b) Eq (Ptr a) Ord (Ptr a) (??? a) => Eq (IORef a) The last example shows the limits of this local analysis. Note that a type variable may be in both categories: then we know a constraint, but there may be more, or a stronger constraint, e.g. data Tree a = Node a [Tree a] deriving Eq yields (Eq a, ??? a) => Eq (Tree a) - - - - - de886f78 by Simon Marlow at 2003-04-25T11:17:55+00:00 [haddock @ 2003-04-25 11:17:55 by simonmar] Some updates, including moving the derived instance item down to the bottom of the list now that Ross has contributed some code that does the job for common cases. - - - - - 1b52cffd by Simon Marlow at 2003-04-30T14:02:32+00:00 [haddock @ 2003-04-30 14:02:32 by simonmar] When installing on Windows, run cygpath over $(HADDOCKLIB) so that haddock (a mingw program, built by GHC) can understand it. You still need to be in a cygwin environment to run Haddock, because of the shell script wrapper. - - - - - d4f638de by Simon Marlow at 2003-05-06T10:04:47+00:00 [haddock @ 2003-05-06 10:04:47 by simonmar] Catch another case of a paragraph containing just a DocMonospaced that should turn into a DocCodeBlock. - - - - - 4162b2b9 by Simon Marlow at 2003-05-06T10:11:44+00:00 [haddock @ 2003-05-06 10:11:44 by simonmar] Add some more code-block tests. - - - - - 4f5802c8 by Simon Marlow at 2003-05-06T10:14:52+00:00 [haddock @ 2003-05-06 10:14:52 by simonmar] Don't turn a single DocCodeBlock into a DocMonospaced, because that tends to remove the line breaks in the code. - - - - - ef8c45f7 by Simon Marlow at 2003-05-21T15:07:21+00:00 [haddock @ 2003-05-21 15:07:21 by simonmar] Only omit the module contents when there are no section headings at all. - - - - - bcee1e75 by Sigbjorn Finne at 2003-05-30T16:50:45+00:00 [haddock @ 2003-05-30 16:50:45 by sof] cygpath: for now, steer clear of --mixed - - - - - 30567af3 by Sigbjorn Finne at 2003-05-30T17:59:28+00:00 [haddock @ 2003-05-30 17:59:28 by sof] oops, drop test defn from prev commit - - - - - b0856e7d by Simon Marlow at 2003-06-03T09:55:26+00:00 [haddock @ 2003-06-03 09:55:26 by simonmar] Two small fixes to make the output valid HTML 4.01 (transitional). Thanks to Malcolm Wallace for pointing out the problems. - - - - - 70e137ea by Simon Marlow at 2003-07-28T13:30:35+00:00 [haddock @ 2003-07-28 13:30:35 by simonmar] Add tests for a couple of bugs. - - - - - 122bd578 by Simon Marlow at 2003-07-28T13:31:25+00:00 [haddock @ 2003-07-28 13:31:25 by simonmar] Add documentation for anchors. - - - - - 0bd27cb2 by Simon Marlow at 2003-07-28T13:31:46+00:00 [haddock @ 2003-07-28 13:31:46 by simonmar] Update - - - - - 08052d42 by Simon Marlow at 2003-07-28T13:32:12+00:00 [haddock @ 2003-07-28 13:32:12 by simonmar] layout tweak. - - - - - 13942749 by Simon Marlow at 2003-07-28T13:33:03+00:00 [haddock @ 2003-07-28 13:33:03 by simonmar] Differentiate links to types/classes from links to variables/constructors with a prefix ("t:" and "v:" respectively). - - - - - d7f493b9 by Simon Marlow at 2003-07-28T13:35:17+00:00 [haddock @ 2003-07-28 13:35:16 by simonmar] When a module A exports another module's contents via 'module B', then modules which import entities from B re-exported by A should link to B.foo rather than A.foo. See examples/Bug2.hs. - - - - - d94cf705 by Simon Marlow at 2003-07-28T13:36:14+00:00 [haddock @ 2003-07-28 13:36:14 by simonmar] Update to version 0.5 - - - - - dbb776cd by Sven Panne at 2003-07-28T14:02:43+00:00 [haddock @ 2003-07-28 14:02:43 by panne] * Updated to version 0.5 * Automagically generate configure if it is not there - - - - - 6cfeee53 by Simon Marlow at 2003-07-28T14:32:43+00:00 [haddock @ 2003-07-28 14:32:42 by simonmar] Update to avoid using hslibs with GHC >= 5.04 - - - - - a1ce838f by Simon Marlow at 2003-07-28T14:33:37+00:00 [haddock @ 2003-07-28 14:33:37 by simonmar] Update for 0.5 - - - - - c0fe6493 by Simon Marlow at 2003-07-28T14:53:22+00:00 [haddock @ 2003-07-28 14:53:22 by simonmar] Markup fix - - - - - 6ea31596 by Sven Panne at 2003-07-28T16:40:45+00:00 [haddock @ 2003-07-28 16:40:45 by panne] Make it compile with GHC >= 6.01 - - - - - afcd30fc by Simon Marlow at 2003-07-30T15:04:52+00:00 [haddock @ 2003-07-30 15:04:52 by simonmar] Pay attention to import specs when building the the import env, as well as the orig env. This may fix some wrong links in documentation when import specs are being used. - - - - - 17c3137f by Simon Marlow at 2003-07-30T16:05:41+00:00 [haddock @ 2003-07-30 16:05:40 by simonmar] Rename instances based on the import_env for the module in which they are to be displayed. This should give, in many cases, better links for the types and classes mentioned in the instance head. This involves keeping around the import_env in the iface until the end, because instances are not collected up until all the modules have been processed. Fortunately it doesn't seem to affect performance much. Instance heads are now attached to ExportDecls, rather than the HTML backend passing around a separate mapping for instances. This is a cleanup. - - - - - 3d3b5c87 by Sven Panne at 2003-08-04T10:18:24+00:00 [haddock @ 2003-08-04 10:18:24 by panne] Don't print parentheses around one-element contexts - - - - - 9e3f3f2d by Simon Marlow at 2003-08-04T12:59:47+00:00 [haddock @ 2003-08-04 12:59:47 by simonmar] A couple of TODOs. - - - - - e9d8085c by Simon Marlow at 2003-08-05T14:10:31+00:00 [haddock @ 2003-08-05 14:10:31 by simonmar] I'm not sure why, but it seems that the index entries for non-defining occurrences of entities did not have an anchor - the link just pointed to the module. This fixes it. - - - - - ff5c7d6d by Simon Marlow at 2003-08-15T14:42:59+00:00 [haddock @ 2003-08-15 14:42:59 by simonmar] Convert the lexer to Alex, and fix a bug in the process. - - - - - 1aa077bf by Simon Marlow at 2003-08-15T15:00:18+00:00 [haddock @ 2003-08-15 15:00:18 by simonmar] Update - - - - - d3de1e38 by Simon Marlow at 2003-08-15T15:01:03+00:00 [haddock @ 2003-08-15 15:01:03 by simonmar] wibbles - - - - - b40ece3b by Simon Marlow at 2003-08-18T10:04:47+00:00 [haddock @ 2003-08-18 10:04:47 by simonmar] Lex the 'mdo' keyword as 'do'. - - - - - 8f9a1146 by Simon Marlow at 2003-08-18T11:48:24+00:00 [haddock @ 2003-08-18 11:48:24 by simonmar] Two bugs from Sven. - - - - - ea54ebc0 by Simon Marlow at 2003-08-18T11:48:46+00:00 [haddock @ 2003-08-18 11:48:46 by simonmar] Fixes to the new lexer. - - - - - d5f6a4b5 by Simon Marlow at 2003-08-19T09:09:03+00:00 [haddock @ 2003-08-19 09:09:03 by simonmar] Further wibbles to the syntax. - - - - - 6bbdadb7 by Sven Panne at 2003-08-26T18:45:35+00:00 [haddock @ 2003-08-26 18:45:35 by panne] Use autoreconf instead of autoconf - - - - - 32e889cb by Sven Panne at 2003-08-26T19:01:19+00:00 [haddock @ 2003-08-26 19:01:18 by panne] Made option handling a bit more consistent with other tools, in particular: Every program in fptools should output * version info on stdout and terminate successfully when -V or --version * usage info on stdout and terminate successfully when -? or --help * usage info on stderr and terminate unsuccessfully when an unknown option is given. - - - - - 5d156a91 by Sven Panne at 2003-08-26T19:20:55+00:00 [haddock @ 2003-08-26 19:20:55 by panne] Make it *very* clear that we terminate when given a -V/--version flag - - - - - e6577265 by Sven Panne at 2003-08-27T07:50:03+00:00 [haddock @ 2003-08-27 07:50:02 by panne] * Made -D a short option for --dump-interface. * Made -m a short option for --ms-help. * Made -n a short option for --no-implicit-prelude. * Made -c a short option for --css. * Removed DocBook options from executable (they didn't do anything), but mark them as reserved in the docs. Note that the short option for DocBook output is now -S (from SGML) instead of -d. The latter is now a short option for --debug. * The order of the Options in the documentation now matches the order printed by Haddock itself. Note: Although changing the names of options is often a bad idea, I'd really like to make the options for the programs in fptools more consistent and compatible to the ones used in common GNU programs. - - - - - d303ff98 by Simon Marlow at 2003-09-10T08:23:48+00:00 [haddock @ 2003-09-10 08:23:48 by simonmar] Add doc subdir. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 9a70e46a by Simon Marlow at 2003-09-10T08:24:32+00:00 [haddock @ 2003-09-10 08:24:32 by simonmar] Install these files in $(datadir), not $(libdir), since they're architecture independent. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - bbb87e7a by Simon Marlow at 2003-09-10T08:25:31+00:00 [haddock @ 2003-09-10 08:25:31 by simonmar] Haddock's supplementary HTML bits now live in $(datadir), not $(libdir). Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 3587c24b by Simon Marlow at 2003-09-22T10:34:38+00:00 [haddock @ 2003-09-22 10:34:38 by simonmar] Allow installing of docs. - - - - - d510b517 by Sven Panne at 2003-10-11T08:10:44+00:00 [haddock @ 2003-10-11 08:10:44 by panne] Include architecture-independent files in file list - - - - - 187d7618 by Sigbjorn Finne at 2003-10-20T17:19:24+00:00 [haddock @ 2003-10-20 17:19:22 by sof] support for i-parameters + zip comprehensions - - - - - b6c7a273 by Simon Marlow at 2003-11-03T14:24:24+00:00 [haddock @ 2003-11-03 14:24:24 by simonmar] Update TODO file. - - - - - 58513e33 by Simon Marlow at 2003-11-05T11:22:04+00:00 [haddock @ 2003-11-05 11:22:04 by simonmar] Remove the last of the uses of 'trace' to emit warnings, and tidy up a couple of places where duplicate warnings were being emitted. - - - - - 33a78846 by Simon Marlow at 2003-11-05T11:30:53+00:00 [haddock @ 2003-11-05 11:30:52 by simonmar] - Suppress warnings about unknown imported modules by default. - Add a -v/--verbose flag to re-enable these warnings. The general idea is to suppress the "Warning: unknown module: Prelude" warnings which most Haddock users will see every time, and which aren't terribly useful. - - - - - a969de7f by Simon Marlow at 2003-11-05T12:30:28+00:00 [haddock @ 2003-11-05 12:30:28 by simonmar] - Remove the emboldening of index entries for defining locations. This isn't useful, and breaks abstractions. - If an entity is re-exported by a module but the module doesn't include documentation for that entity (perhaps because it is re-exported by 'module M'), then don't attempt to hyperlink to the documentation from the index. Instead, just list that module in the index, to indicate that the entity is exported from there. - - - - - f14ea82a by Simon Marlow at 2003-11-05T15:15:59+00:00 [haddock @ 2003-11-05 15:15:59 by simonmar] Index overhaul: - no more separate type/class and variable/function indices - the index now makes a distinction between different entities with the same name. One example is a type constructor with the same name as a data constructor, but another example is simply a function with the same name exported by two different modules. For example, the index entry for 'catch' now looks like this: catch 1 (Function) Control.Exception 2 (Function) GHC.Exception, Prelude, System.IO, System.IO.Error making it clear that there are two different 'catch'es, but one of them is exported by several modules. - Each index page now has the index contents (A B C ...) at the top. Please let me know if you really hate any of this. - - - - - 01a25ca6 by Simon Marlow at 2003-11-05T15:16:38+00:00 [haddock @ 2003-11-05 15:16:38 by simonmar] Update - - - - - 1a7ccb86 by Simon Marlow at 2003-11-05T17:16:05+00:00 [haddock @ 2003-11-05 17:16:04 by simonmar] Support for generating a single unified index for several packages. --use-index=URL turns off normal index generation, causes Index links to point to URL. --gen-index generates an combined index from the specified interfaces. Currently doesn't work exactly right, because the interfaces don't contain the iface_reexported info. I'll need to fix that up. - - - - - a2bca16d by Simon Marlow at 2003-11-06T10:44:52+00:00 [haddock @ 2003-11-06 10:44:52 by simonmar] Include iface_reexported in the .haddock file. This unfortunately bloats the file (40% for base). If this gets to be a problem we can always apply the dictionary trick that GHC uses for squashing .hi files. - - - - - 0a09c293 by Simon Marlow at 2003-11-06T12:39:47+00:00 [haddock @ 2003-11-06 12:39:46 by simonmar] - Add definition lists, marked up like this: -- | This is a definition list: -- -- [@foo@] The description of @foo at . -- -- [@bar@] The description of @bar at . Cunningly, the [] characters are not treated specially unless a [ is found at the beginning of a paragraph, in which case the ] becomes special in the following text. - Add --use-contents and --gen-contents, along the lines of --use-index and --gen-index added yesterday. Now we can generate a combined index and contents for the whole of the hierarchical libraries, and in theory the index/contents on the system could be updated as new packages are added. - - - - - fe1b3460 by Simon Marlow at 2003-11-06T14:47:36+00:00 [haddock @ 2003-11-06 14:47:36 by simonmar] Remove the 'Parent' button - it is of dubious use, and often points into thin air. - - - - - db6d762f by Simon Marlow at 2003-11-06T16:48:14+00:00 [haddock @ 2003-11-06 16:48:11 by simonmar] - Include the OptHide setting in the interface, so we don't include hidden modules in the combined index/contents. - Add a -k/--package flag to set the package name for the current set of modules. The package name for each module is now shown in the right-hand column of the contents, in a combined contents page. - - - - - 7d71718b by Simon Marlow at 2003-11-06T16:50:28+00:00 [haddock @ 2003-11-06 16:50:28 by simonmar] Add -k/--package docs - - - - - ef43949d by Simon Marlow at 2003-11-06T16:51:23+00:00 [haddock @ 2003-11-06 16:51:23 by simonmar] Bump to 0.6 - - - - - 1c419e06 by Simon Marlow at 2003-11-06T16:51:50+00:00 [haddock @ 2003-11-06 16:51:50 by simonmar] update - - - - - 69422327 by Simon Marlow at 2003-11-10T14:41:06+00:00 [haddock @ 2003-11-10 14:41:05 by simonmar] Re-exporting names from a different package is problematic, because we don't have access to the full documentation for the entity. Currently Haddock just ignores entities with no documentation, but this results in bogus-looking empty documentation for many of the modules in the haskell98 package. So: - the documentation will now just list the name, as a link pointing to the location of the actual documentation. - now we don't attempt to link to these re-exported entities if they are referred to by the current module. Additionally: - If there is no documentation in the current module, include just the Synopsis section (rather than just the documentation section, as it was before). This just looks nicer and was on the TODO list. - - - - - 3c3fc433 by Simon Marlow at 2003-11-10T14:51:59+00:00 [haddock @ 2003-11-10 14:51:59 by simonmar] Fix for getReExports: take into account names which are not visible because they are re-exported from a different package. - - - - - 31c8437b by Simon Marlow at 2003-11-10T15:10:53+00:00 [haddock @ 2003-11-10 15:10:53 by simonmar] Version 0.6 changes - - - - - a7c2430b by Simon Marlow at 2003-11-10T15:15:58+00:00 [haddock @ 2003-11-10 15:15:58 by simonmar] getReExports: one error case that isn't - - - - - 00cc459c by Simon Marlow at 2003-11-10T16:15:19+00:00 [haddock @ 2003-11-10 16:15:18 by simonmar] copyright update - - - - - ca62408d by Simon Marlow at 2003-11-11T09:57:25+00:00 [haddock @ 2003-11-11 09:57:25 by simonmar] Version 0.6 - - - - - 3acbf818 by Simon Marlow at 2003-11-11T12:10:44+00:00 [haddock @ 2003-11-11 12:10:44 by simonmar] Go back to producing just the documentation section, rather than just the synopsis section, for a module with no documentation annotations. One reason is that the synopsis section tries to link each entity to its documentation on the same page. Also, the doc section anchors each entity, and it lists instances which the synopsis doesn't. - - - - - 6c90abc2 by Simon Marlow at 2003-11-12T10:03:39+00:00 [haddock @ 2003-11-12 10:03:39 by simonmar] 2002 -> 2003 - - - - - 090bbc4c by Simon Marlow at 2003-11-28T12:08:00+00:00 [haddock @ 2003-11-28 12:08:00 by simonmar] update - - - - - 8096a832 by Simon Marlow at 2003-11-28T12:09:58+00:00 [haddock @ 2003-11-28 12:09:58 by simonmar] Fix some of the problems with Haddock generating pages that are too wide. Now we only specify 'nowrap' when it is necessary to avoid a code box getting squashed up by the text to the right of it. - - - - - 35294929 by Sven Panne at 2003-12-29T17:16:31+00:00 [haddock @ 2003-12-29 17:16:31 by panne] Updated my email address - - - - - cdb697bf by Simon Marlow at 2004-01-08T10:14:24+00:00 [haddock @ 2004-01-08 10:14:24 by simonmar] Add instructions for using GHC to pre-process source for feeding to Haddock. - - - - - 8dfc491f by Simon Marlow at 2004-01-09T12:45:46+00:00 [haddock @ 2004-01-09 12:45:46 by simonmar] Add -optP-P to example ghc command line. - - - - - ac41b820 by Simon Marlow at 2004-02-03T11:02:03+00:00 [haddock @ 2004-02-03 11:02:03 by simonmar] Fix bug in index generation - - - - - f4e7edcb by Simon Marlow at 2004-02-10T11:51:16+00:00 [haddock @ 2004-02-10 11:51:16 by simonmar] Don't throw away whitespace at the beginning of a line (experimental fix). - - - - - 68e212d2 by Simon Marlow at 2004-02-10T12:10:08+00:00 [haddock @ 2004-02-10 12:10:08 by simonmar] Fix for previous commit: I now realise why the whitespace was stripped from the beginning of the line. Work around it. - - - - - e7d7f2df by Sven Panne at 2004-02-10T18:38:45+00:00 [haddock @ 2004-02-10 18:38:45 by panne] Make Haddock link with the latest relocated monad transformer package - - - - - 992d4225 by Simon Marlow at 2004-02-16T10:21:35+00:00 [haddock @ 2004-02-16 10:21:35 by simonmar] Add a TODO - - - - - 1ac55326 by Simon Marlow at 2004-03-12T11:33:39+00:00 [haddock @ 2004-03-12 11:33:39 by simonmar] Add an item. - - - - - 0478e903 by Simon Marlow at 2004-03-15T12:24:05+00:00 [haddock @ 2004-03-15 12:24:05 by simonmar] Add an item. - - - - - 6f26d21a by Simon Marlow at 2004-03-18T14:21:29+00:00 [haddock @ 2004-03-18 14:21:29 by simonmar] Fix URL - - - - - 19b6bb99 by Simon Marlow at 2004-03-22T14:09:03+00:00 [haddock @ 2004-03-22 14:09:03 by simonmar] getReExports was bogus: we should really look in the import_env to find the documentation for an entity which we are re-exporting without documentation. Suggested by: Ross Paterson (patch modified by me). - - - - - 5c756031 by Simon Marlow at 2004-03-24T09:42:11+00:00 [haddock @ 2004-03-24 09:42:10 by simonmar] hiding bug from Ross Paterson (fixed in rev 1.59 of Main.hs) - - - - - 1b692e6c by Simon Marlow at 2004-03-24T10:10:50+00:00 [haddock @ 2004-03-24 10:10:50 by simonmar] mkExportItems fix & simplification: we should be looking at the actual exported names (calculated earlier) to figure out which subordinates of a declaration are exported. This means that if you export a record, and name its fields separately in the export list, the fields will still be visible in the documentation for the constructor. - - - - - 90e5e294 by Simon Marlow at 2004-03-24T10:12:08+00:00 [haddock @ 2004-03-24 10:12:08 by simonmar] Make restrictCons take into account record field names too (removing a ToDo). - - - - - 2600efa4 by Simon Marlow at 2004-03-24T10:16:17+00:00 [haddock @ 2004-03-24 10:16:17 by simonmar] Record export tests. - - - - - 6a8575c7 by Simon Marlow at 2004-03-25T09:35:14+00:00 [haddock @ 2004-03-25 09:35:14 by simonmar] restrictTo: fix for restricting a newtype with a record field. - - - - - dcf55a8d by Simon Marlow at 2004-03-25T10:01:42+00:00 [haddock @ 2004-03-25 10:01:42 by simonmar] Fix duplicate instance bug - - - - - f49aa758 by Simon Marlow at 2004-03-25T10:02:41+00:00 [haddock @ 2004-03-25 10:02:41 by simonmar] Duplicate instance bug. - - - - - 7b87344c by Simon Marlow at 2004-03-25T10:29:56+00:00 [haddock @ 2004-03-25 10:29:56 by simonmar] If a name is imported from two places, one hidden and one not, choose the unhidden one to link to. Also, when there's only a hidden module to link to, don't try linking to it. - - - - - 40f44d7b by Simon Marlow at 2004-03-25T15:17:24+00:00 [haddock @ 2004-03-25 15:17:23 by simonmar] Add support for collaspible parts of the page, with a +/- button and a bit of JavaScript. Make the instances collapsible, and collapse them by default. This makes documentation with long lists of instances (eg. the Prelude) much easier to read. Maybe we should give other documentation sections the same treatment. - - - - - 9b64dc0f by Simon Marlow at 2004-03-25T15:20:55+00:00 [haddock @ 2004-03-25 15:20:55 by simonmar] Update - - - - - c2fff7f2 by Simon Marlow at 2004-03-25T15:45:10+00:00 [haddock @ 2004-03-25 15:45:10 by simonmar] Eliminate some unnecessary spaces in the HTML rendering - - - - - b7948ff0 by Simon Marlow at 2004-03-25T16:00:37+00:00 [haddock @ 2004-03-25 16:00:36 by simonmar] Remove all that indentation in the generated HTML to keep the file sizes down. - - - - - da2bb4ca by Sven Panne at 2004-03-27T09:57:58+00:00 [haddock @ 2004-03-27 09:57:57 by panne] Added the new-born haddock.js to the build process and the documentation. - - - - - b99e6f8c by Sven Panne at 2004-03-27T10:32:20+00:00 [haddock @ 2004-03-27 10:32:20 by panne] "type" is a required attribute of the "script" element - - - - - 562b185a by Sven Panne at 2004-03-27T12:52:34+00:00 [haddock @ 2004-03-27 12:52:34 by panne] Add a doctype for the contents page, too. - - - - - f6a99c2d by Simon Marlow at 2004-04-14T10:03:25+00:00 [haddock @ 2004-04-14 10:03:25 by simonmar] fix for single-line comment syntax - - - - - de366303 by Simon Marlow at 2004-04-20T13:08:04+00:00 [haddock @ 2004-04-20 13:08:04 by simonmar] Allow a 'type' declaration to include documentation comments. These will be ignored by Haddock, but at least one user (Johannes Waldmann) finds this feature useful, and it's easy to add. - - - - - fd78f51e by Simon Marlow at 2004-05-07T15:14:56+00:00 [haddock @ 2004-05-07 15:14:56 by simonmar] - update copyright - add version to abstract - - - - - 59f53e32 by Sven Panne at 2004-05-09T14:39:53+00:00 [haddock @ 2004-05-09 14:39:53 by panne] Fix the fix for single-line comment syntax, ------------------------------------------- is now a valid comment line again. - - - - - 8b18f2fe by Simon Marlow at 2004-05-10T10:11:51+00:00 [haddock @ 2004-05-10 10:11:51 by simonmar] Update - - - - - 225a491d by Ross Paterson at 2004-05-19T13:10:23+00:00 [haddock @ 2004-05-19 13:10:23 by ross] Make the handling of "deriving" slightly smarter, by ignoring data constructor arguments that are identical to the lhs. Now handles things like data Tree a = Leaf a | Branch (Tree a) (Tree a) deriving ... - - - - - 37588686 by Mike Thomas at 2004-05-21T06:38:14+00:00 [haddock @ 2004-05-21 06:38:14 by mthomas] Windows exe extensions (bin remains for Unix). - - - - - cf2b9152 by Simon Marlow at 2004-05-25T09:34:54+00:00 [haddock @ 2004-05-25 09:34:54 by simonmar] Add some TODO items - - - - - 4d29cdfc by Simon Marlow at 2004-05-25T10:41:46+00:00 [haddock @ 2004-05-25 10:41:46 by simonmar] Complain if -h is used with --gen-index or --gen-contents, because it'll overwrite the new index/contents. - - - - - 2e0771e0 by Mike Thomas at 2004-05-28T20:17:55+00:00 [haddock @ 2004-05-28 20:17:55 by mthomas] Windows: search for templates in executable directory. Unix: Haddock tries cwd first rather than error if no -l arg. - - - - - 8d10bde1 by Sven Panne at 2004-06-05T16:53:34+00:00 [haddock @ 2004-06-05 16:53:34 by panne] Misc. rpm spec file cleanup, including: * make BuildRoot handling more consistent * added default file attributes * consistent defines and tags - - - - - 59974349 by Sven Panne at 2004-06-05T18:01:00+00:00 [haddock @ 2004-06-05 18:01:00 by panne] More rpm spec file cleanup, including: * added some BuildRequires * changed packager to me, so people can complain at the right place :-] * consistently refer to haskell.org instead of www.haskell.org - - - - - b94d4903 by Simon Marlow at 2004-07-01T11:08:58+00:00 [haddock @ 2004-07-01 11:08:57 by simonmar] Update to the +/- buttons: use a resized image rather than a <button>. Still seeing some strange effects in Konqueror, so might need to use a fixed-size image instead. - - - - - d5278f67 by Sven Panne at 2004-07-04T15:15:55+00:00 [haddock @ 2004-07-04 15:15:55 by panne] Install pictures for +/- pictures, too (JPEG is a strange format for graphics like this, I would have expected GIF or PNG here.) Things look fine with Konqueror and Netscape on Linux now, the only downside is that the cursor doesn't change when positioned above the "button". - - - - - 46dec6c5 by Sven Panne at 2004-07-13T17:59:28+00:00 [haddock @ 2004-07-13 17:59:28 by panne] A quote is a valid part of a Haskell identifier, but it would interfere with an ECMA script string delimiter, so escape it there. - - - - - 1d7bc432 by Simon Marlow at 2004-07-22T08:54:06+00:00 [haddock @ 2004-07-22 08:54:06 by simonmar] Add single quote to $ident, so you can say eg. 'foldl'' to refer to foldl' (the longest match rule is our friend). Bug reported by Adrian Hey <ahey at iee.org> - - - - - f183618b by Krasimir Angelov at 2004-07-27T22:59:35+00:00 [haddock @ 2004-07-27 22:58:23 by krasimir] Add basic support for Microsoft HTML Help 2.0 - - - - - d515d0c2 by Krasimir Angelov at 2004-07-27T23:02:36+00:00 [haddock @ 2004-07-27 23:02:36 by krasimir] escape names in the index - - - - - a5f1be23 by Krasimir Angelov at 2004-07-27T23:05:21+00:00 [haddock @ 2004-07-27 23:05:21 by krasimir] Add jsFile, plusFile and minusFile to the file list - - - - - c4fb4881 by Krasimir Angelov at 2004-07-28T22:12:10+00:00 [haddock @ 2004-07-28 22:12:09 by krasimir] bugfix. Move contentsHtmlFile, indexHtmlFile and subIndexHtmlFile functions to HaddockUtil.hs module to make them accessible from HaddockHH2.hs - - - - - 64d30b1d by Krasimir Angelov at 2004-07-30T22:15:47+00:00 [haddock @ 2004-07-30 22:15:45 by krasimir] more stuffs - support for separated compilation of packages - the contents page now uses DHTML TreeView - fixed copyFile bug - - - - - 133c8c5c by Krasimir Angelov at 2004-07-31T12:04:38+00:00 [haddock @ 2004-07-31 12:04:37 by krasimir] make the DHtmlTree in contents page more portable. The +/- buttons are replaced with new images which looks more beatiful. - - - - - 79040963 by Krasimir Angelov at 2004-07-31T13:10:20+00:00 [haddock @ 2004-07-31 13:10:20 by krasimir] Make DHtmlTree compatible with Mozila browser - - - - - 1a55dc90 by Krasimir Angelov at 2004-07-31T14:52:55+00:00 [haddock @ 2004-07-31 14:52:55 by krasimir] fix - - - - - 85ce0237 by Krasimir Angelov at 2004-07-31T14:53:28+00:00 [haddock @ 2004-07-31 14:53:28 by krasimir] HtmlHelp 1.x - - - - - 3c0c53ba by Krasimir Angelov at 2004-07-31T20:35:21+00:00 [haddock @ 2004-07-31 20:35:21 by krasimir] Added support for DevHelp - - - - - d42b5af1 by Krasimir Angelov at 2004-07-31T21:17:51+00:00 [haddock @ 2004-07-31 21:17:51 by krasimir] Document new features in HtmlHelp - - - - - 790fe21e by Krasimir Angelov at 2004-08-01T15:14:02+00:00 [haddock @ 2004-08-01 15:14:02 by krasimir] add missing imports - - - - - fd7cc6bc by Krasimir Angelov at 2004-08-01T19:52:08+00:00 [haddock @ 2004-08-01 19:52:06 by krasimir] fix some bugs. Now I have got the entire libraries documentation in HtmlHelp 2.0 format. - - - - - 94ad7ac8 by Krasimir Angelov at 2004-08-01T19:53:50+00:00 [haddock @ 2004-08-01 19:53:50 by krasimir] I forgot to add the new +/- images - - - - - f0c65388 by Krasimir Angelov at 2004-08-02T16:25:53+00:00 [haddock @ 2004-08-02 16:25:53 by krasimir] Add root node to the table of contents. All modules in tree are not children of the root - - - - - f50bd85d by Sven Panne at 2004-08-02T18:17:46+00:00 [haddock @ 2004-08-02 18:17:46 by panne] Mainly DocBook fixes - - - - - 09527ce3 by Sven Panne at 2004-08-02T20:02:29+00:00 [haddock @ 2004-08-02 20:02:29 by panne] Fixed -o/--odir handling. Generating the output, especially the directory handling, is getting a bit convoluted nowadays... - - - - - c8fbacfa by Sven Panne at 2004-08-02T20:31:13+00:00 [haddock @ 2004-08-02 20:31:13 by panne] Warning police - - - - - 37830bff by Sven Panne at 2004-08-02T20:32:29+00:00 [haddock @ 2004-08-02 20:32:28 by panne] Nuked dead code - - - - - 13847171 by Sven Panne at 2004-08-02T21:12:27+00:00 [haddock @ 2004-08-02 21:12:25 by panne] Use pathJoin instead of low-level list-based manipulation for FilePaths - - - - - c711d61e by Sven Panne at 2004-08-02T21:16:02+00:00 [haddock @ 2004-08-02 21:16:02 by panne] Removed WinDoze CRs - - - - - b1f7dc88 by Sven Panne at 2004-08-03T19:35:59+00:00 [haddock @ 2004-08-03 19:35:59 by panne] Fixed spelling of "http-equiv" attribute - - - - - dd5f394e by Sven Panne at 2004-08-03T19:44:03+00:00 [haddock @ 2004-08-03 19:44:03 by panne] Pacify W3C validator: * Added document encoding (currently UTF-8, not sure if this is completely correct) * Fixed syntax of `id' attributes * Added necessary `alt' attribute for +/- images Small layout improvement: * Added space after +/- images (still not perfect, but better than before) - - - - - 919c47c6 by Sigbjorn Finne at 2004-08-03T19:45:11+00:00 [haddock @ 2004-08-03 19:45:11 by sof] make it compile with <= ghc-6.1 - - - - - 4d6f01d8 by Sigbjorn Finne at 2004-08-03T19:45:30+00:00 [haddock @ 2004-08-03 19:45:30 by sof] ffi wibble - - - - - 4770643a by Sven Panne at 2004-08-03T20:47:46+00:00 [haddock @ 2004-08-03 20:47:46 by panne] Fixed CSS for button style. Note that only "0" is a valid measure without a unit! - - - - - 14aaf2e5 by Sven Panne at 2004-08-03T21:07:59+00:00 [haddock @ 2004-08-03 21:07:58 by panne] Improved spacing of dynamic module tree - - - - - 97c3579a by Simon Marlow at 2004-08-09T11:03:04+00:00 [haddock @ 2004-08-09 11:03:04 by simonmar] Add FormatVersion Patch submitted by: George Russell <ger at informatik.uni-bremen.de> - - - - - af7f8c03 by Simon Marlow at 2004-08-09T11:55:07+00:00 [haddock @ 2004-08-09 11:55:05 by simonmar] Add support for a short description for each module, which is included in the contents. The short description should be given in a "Description: " field of the header. Included in this patch are changes that make the format of the header a little more flexible. From the comments: -- all fields in the header are optional and have the form -- -- [spaces1][field name][spaces] ":" -- [text]"\n" ([spaces2][space][text]"\n" | [spaces]"\n")* -- where each [spaces2] should have [spaces1] as a prefix. -- -- Thus for the key "Description", -- -- > Description : this is a -- > rather long -- > -- > description -- > -- > The module comment starts here -- -- the value will be "this is a .. description" and the rest will begin -- at "The module comment". The header fields must be in the following order: Module, Description, Copyright, License, Maintainer, Stability, Portability. Patches submitted by: George Russell <ger at informatik.uni-bremen.de>, with a few small changes be me, mostly to merge with other recent changes. ToDo: document the module header. - - - - - 7b865ad3 by Simon Marlow at 2004-08-10T14:09:57+00:00 [haddock @ 2004-08-10 14:09:57 by simonmar] Fixes for DevHelp/HtmlHelp following introduction of short module description. - - - - - 814766cd by Simon Marlow at 2004-08-10T14:33:46+00:00 [haddock @ 2004-08-10 14:33:45 by simonmar] Fixes to installation under Windows. - - - - - 39cf9ede by Simon Marlow at 2004-08-12T12:08:23+00:00 [haddock @ 2004-08-12 12:08:23 by simonmar] Avoid using string-gap tricks. - - - - - b6d78551 by Simon Marlow at 2004-08-13T10:53:21+00:00 [haddock @ 2004-08-13 10:53:21 by simonmar] Update - - - - - eaae7417 by Simon Marlow at 2004-08-13T10:53:50+00:00 [haddock @ 2004-08-13 10:53:50 by simonmar] Test for primes in quoted links - - - - - 68c34f06 by Sven Panne at 2004-08-16T19:59:38+00:00 [haddock @ 2004-08-16 19:59:36 by panne] XMLification - - - - - 7f45a6f9 by Sven Panne at 2004-08-18T16:42:54+00:00 [haddock @ 2004-08-18 16:42:54 by panne] Re-added indices + minor fixes - - - - - 8a5dd97c by Sigbjorn Finne at 2004-08-25T17:15:42+00:00 [haddock @ 2004-08-25 17:15:42 by sof] backquote HADDOCK_VERSION defn for <= ghc-6.0.x; believe this is only needed under mingw - - - - - 4b1b42ea by Sven Panne at 2004-08-26T20:08:50+00:00 [haddock @ 2004-08-26 20:08:49 by panne] SGML is dead, long live DocBook XML! Note: The BuildRequires tags in the spec files are still incomplete and the documentation about the DocBook tools needs to be updated, too. Stay tuned... - - - - - 8d52cedb by Sven Panne at 2004-08-26T21:03:19+00:00 [haddock @ 2004-08-26 21:03:19 by panne] Updated BuildRequires tags. Alas, there seems to be no real standard here, so your mileage may vary... At least the current specs should work on SuSE Linux. - - - - - e6982912 by Sigbjorn Finne at 2004-08-30T15:44:59+00:00 [haddock @ 2004-08-30 15:44:59 by sof] escape HADDOCK_VERSION double quotes on all platforms when compiling with <=6.0.x - - - - - b3fbc867 by Simon Marlow at 2004-08-31T13:09:42+00:00 [haddock @ 2004-08-31 13:09:42 by simonmar] Avoid GHC/shell versionitis and create Version.hs - - - - - c359e16a by Sven Panne at 2004-09-05T19:12:33+00:00 [haddock @ 2004-09-05 19:12:32 by panne] * HTML documentation for "foo.xml" goes into directory "foo" again, not "foo-html". This is nicer and consistent with the behaviour for building the docs from SGML. * Disabled building PostScript documentation in the spec files for now, there are some strange issues with the FO->PS conversion for some files which have to be clarified first. - - - - - c68b1eba by Sven Panne at 2004-09-24T07:04:38+00:00 [haddock @ 2004-09-24 07:04:38 by panne] Switched the default state for instances and the module hierarchy to non-collapsed. This can be reversed when we finally use cookies from JavaScript to have a more persistent state. Previously going back and forth in the documentation was simply too annoying because everything was collapsed again and therefore the documentation was not easily navigatable. - - - - - dfb32615 by Simon Marlow at 2004-09-30T08:21:29+00:00 [haddock @ 2004-09-30 08:21:29 by simonmar] Add a feature request - - - - - 45ff783c by Sven Panne at 2004-10-23T19:54:00+00:00 [haddock @ 2004-10-23 19:54:00 by panne] Improved the Cygwin/MinGW chaos a little bit. There is still confusion about host platform vs. target platform... - - - - - 5f644714 by Krasimir Angelov at 2004-10-28T16:01:51+00:00 [haddock @ 2004-10-28 16:01:51 by krasimir] update for ghc-6.3+ - - - - - 92d9753e by Sven Panne at 2004-11-01T16:39:01+00:00 [haddock @ 2004-11-01 16:39:01 by panne] Revert previous commit: It's Network.URI which should be changed, not Haddock. - - - - - 05f70f6e by Simon Marlow at 2005-01-04T16:15:51+00:00 [haddock @ 2005-01-04 16:15:51 by simonmar] parser fix: allow qualified specialids. - - - - - 47870837 by Simon Marlow at 2005-01-04T16:16:54+00:00 [haddock @ 2005-01-04 16:16:54 by simonmar] Add a test - - - - - ff11fc2c by Ross Paterson at 2005-01-10T19:18:22+00:00 [haddock @ 2005-01-10 19:18:22 by ross] Render non-ASCII characters using numeric character references, to simplify charset issues. There's a META tag saying the charset is UTF-8, but GHC outputs characters as raw bytes. Ideally we need an encoding on the input side too, primarily in comments, because source files containing non-ASCII characters aren't portable between locales. - - - - - eba2fc4e by Simon Marlow at 2005-01-11T10:44:37+00:00 [haddock @ 2005-01-11 10:44:37 by simonmar] Remove string gap - - - - - b899a381 by Ross Paterson at 2005-01-13T11:41:33+00:00 [haddock @ 2005-01-13 11:41:33 by ross] recognize SGML-style numeric character references &#ddd; or &#xhhhh; and translate them into Chars. - - - - - 106e3cf0 by Ross Paterson at 2005-01-13T14:43:41+00:00 [haddock @ 2005-01-13 14:43:41 by ross] also allow uppercase X in hexadecimal character references (like SGML) - - - - - e8f54f25 by Ross Paterson at 2005-01-13T14:44:24+00:00 [haddock @ 2005-01-13 14:44:24 by ross] Describe numeric character references. - - - - - 914ccdce by Sven Panne at 2005-01-15T18:44:48+00:00 [haddock @ 2005-01-15 18:44:45 by panne] Make Haddock compile again after the recent base package changed. The Map/Set legacy hell has been factored out, so that all modules can simply use the new non-deprecated interfaces. Probably a lot of things can be improved by a little bit of Map/Set/List algebra, this can be done later if needed. Small note: Currently the list of instances in HTML code is reversed. This will hopefully be fixed later. - - - - - 6ab20e84 by Sven Panne at 2005-01-16T12:18:26+00:00 [haddock @ 2005-01-16 12:18:26 by panne] Trim imports - - - - - efb81da9 by Sven Panne at 2005-01-16T12:58:08+00:00 [haddock @ 2005-01-16 12:58:03 by panne] Correctly handle the new order of arguments for the combining function given to fromListWith. - - - - - e27b5834 by Sven Panne at 2005-01-16T14:14:41+00:00 [haddock @ 2005-01-16 14:14:39 by panne] Data.Map.unions is left-biased. - - - - - dae3cc3e by Sven Panne at 2005-01-16T14:22:44+00:00 [haddock @ 2005-01-16 14:22:44 by panne] Added the last missing "flip" to get identical HTML output as previous versions. - - - - - 951d8408 by Sven Panne at 2005-01-16T14:37:10+00:00 [haddock @ 2005-01-16 14:37:10 by panne] Refactored Text.PrettyPrint legacy hell into a separate module. - - - - - f1c4b892 by Sven Panne at 2005-01-16T15:41:25+00:00 [haddock @ 2005-01-16 15:41:21 by panne] Cleaned up imports and dropped support for GHC < 5.03, it never worked, anyway. - - - - - 60824c6e by Simon Marlow at 2005-01-18T10:02:48+00:00 [haddock @ 2005-01-18 10:02:48 by simonmar] Add a TODO - - - - - a8c82f23 by Krasimir Angelov at 2005-01-28T23:19:39+00:00 [haddock @ 2005-01-28 23:19:39 by krasimir] import Foreign/Foreign.C are required for Windows - - - - - d8450a23 by Simon Marlow at 2005-02-02T16:23:04+00:00 [haddock @ 2005-02-02 16:23:00 by simonmar] Revamp the linking strategy in Haddock. Now name resolution is done in two phases: - first resolve everything to original names, like a Haskell compiler would. - then, figure out the "home" location for every entity, and point all the links to there. The home location is the lowest non-hidden module in the import hierarchy that documents the entity. If there are multiple candidates, one is chosen at random. Also: - Haddock should not generate any HTML with dangling links any more. Unlinked references are just rendered as plain text. - Error reporting is better: if we can't find a link destination for an entity reference, we now emit a warning. - - - - - 1cce71d0 by Simon Marlow at 2005-02-03T13:42:19+00:00 [haddock @ 2005-02-03 13:42:19 by simonmar] - add --ignore-all-exports flag, which behaves as if every module has the ignore-exports attribute (requested by Chris Ryder). - add --hide option to hide a module on the command line. - add --use-package option to get Haddock info for a package from ghc-pkg (largely untested). - remove reexports from the .haddock file, they aren't used any more. - - - - - 767123ef by Ross Paterson at 2005-02-03T16:17:37+00:00 [haddock @ 2005-02-03 16:17:37 by ross] fix typo for < 6.3 - - - - - 0c680c04 by Simon Marlow at 2005-02-04T12:03:31+00:00 [haddock @ 2005-02-04 12:03:31 by simonmar] Fix bug in renameExportItems that meant links in instances weren't being renamed properly. - - - - - ff7abe5f by Simon Marlow at 2005-02-04T12:15:53+00:00 [haddock @ 2005-02-04 12:15:52 by simonmar] Add attribute #not-home, to indicate that the current module should not be considered to be a home module for the each entity it exports, unless there is no other module that exports the entity. - - - - - fc2cfd27 by Simon Marlow at 2005-02-04T12:40:02+00:00 [haddock @ 2005-02-04 12:40:02 by simonmar] Update the documentation w.r.t. home modules and the not-home attribute. - - - - - 26b8ddf7 by Ross Paterson at 2005-02-04T13:36:06+00:00 [haddock @ 2005-02-04 13:36:05 by ross] sort lists of instances by - arity of the type constructors (so higher-kinded instances come first) - name of the class - argument types - - - - - 26bfb19c by Simon Marlow at 2005-02-23T15:57:12+00:00 [haddock @ 2005-02-23 15:57:12 by simonmar] Fix documentation regarding the module attributes. - - - - - 9c3afd02 by Simon Marlow at 2005-02-28T16:18:17+00:00 [haddock @ 2005-02-28 16:18:17 by simonmar] version 0.7 - - - - - a95fd63f by Simon Marlow at 2005-02-28T16:22:08+00:00 [haddock @ 2005-02-28 16:22:08 by simonmar] Attempt to fix the layout of the package names in the contents. Having tried just about everything, the only thing I can get to work reliably is to make the package names line up on a fixed offset from the left margin. This obviously isn't ideal, so anyone else that would like to have a go at improving it is welcome. One option is to remove the +/- buttons from the contents list and go back to a plain table. The contents page now uses CSS for layout rather than tables. It seems that most browsers have different interpretations of CSS layout, so only the simplest things lead to consistent results. - - - - - 905d42f7 by Simon Marlow at 2005-03-01T17:16:42+00:00 [haddock @ 2005-03-01 17:16:40 by simonmar] Another attempt at lining up the package names on the contents page. Now, they line up with Konqueror, and almost line up with Firefox & IE (different layout in each case). - - - - - a0e1d178 by Wolfgang Thaller at 2005-03-09T08:28:39+00:00 [haddock @ 2005-03-09 08:28:39 by wolfgang] Hack haddock's lexer to accept the output from Apple's broken version of cpp (Apple's cpp leaves #pragma set_debug_pwd directives in it's output). - - - - - 9e1eb784 by Simon Marlow at 2005-04-22T14:27:15+00:00 [haddock @ 2005-04-22 14:27:15 by simonmar] Add a TODO item - - - - - 23281f78 by Ross Paterson at 2005-05-18T12:41:59+00:00 [haddock @ 2005-05-18 12:41:59 by ross] fix 3 bugs in --use-package, and document it. - - - - - 00074a68 by Sven Panne at 2005-05-21T12:35:29+00:00 [haddock @ 2005-05-21 12:35:29 by panne] Warning/versionitis police - - - - - 341fa822 by Simon Marlow at 2005-06-15T15:43:21+00:00 [haddock @ 2005-06-15 15:43:21 by simonmar] Allow "licence" as an alternate spelling of "license" - - - - - 3b953f8b by Simon Marlow at 2005-06-16T08:14:12+00:00 [haddock @ 2005-06-16 08:14:12 by simonmar] wibble - - - - - abfd9826 by Simon Marlow at 2005-06-27T14:46:40+00:00 [haddock @ 2005-06-27 14:46:40 by simonmar] name hierarchical HTML files as A-B-C.html instead of A.B.C.html. The old way confused Apache because the extensions are sometimes interpreted as having special meanings. - - - - - a01eea00 by Simon Marlow at 2005-08-04T13:59:40+00:00 [haddock @ 2005-08-04 13:59:40 by simonmar] 0.7 changes - - - - - 170ef87e by Simon Marlow at 2005-08-04T15:08:03+00:00 [haddock @ 2005-08-04 15:08:03 by simonmar] spec file from Jens Peterson - - - - - 7621fde4 by Simon Marlow at 2005-08-04T15:59:30+00:00 [haddock @ 2005-08-04 15:59:30 by simonmar] replace mingw tests with $(Windows) - - - - - a20739bb by Sven Panne at 2005-08-05T07:01:12+00:00 [haddock @ 2005-08-05 07:01:12 by panne] Reverted to previous version (but with bumped version number), the last commit broke RPM building on SuSE systems due to differently named dependencies. As a clarification: All .spec files in the repository have to work at least on SuSE, because that's the system I'm using. And as "Mr. Building Police", I reserve me the right to keep them that way... >:-) It might very well be the case that we need different .spec files for different platforms, so packagers which are unhappy with the current .spec files should contact me, stating the actual problems. - - - - - 4afb15cf by Simon Marlow at 2005-10-05T10:51:45+00:00 [haddock @ 2005-10-05 10:51:45 by simonmar] Add a bug - - - - - 60f69f82 by Simon Marlow at 2005-10-05T12:52:03+00:00 [haddock @ 2005-10-05 12:52:03 by simonmar] Document new behaviour of -s option - - - - - f7e520ca by Simon Marlow at 2005-10-10T15:02:55+00:00 [haddock @ 2005-10-10 15:02:55 by simonmar] extractRecSel: ignore non-record constructors (fixes a crash when using datatypes with a mixture of record and non-record style constructors). - - - - - b2edbedb by Simon Marlow at 2005-10-14T09:44:21+00:00 Start CHANGES for 0.8 - - - - - 21c7ac8d by Simon Marlow at 2005-10-14T23:11:19+00:00 First cut of Cabal build system - - - - - 766cecdd by Simon Marlow at 2005-10-29T08:14:43+00:00 Add configure script and Makefile for the docs Add a separate configure script and build system for building the documentation. The configure and Makefile code is stolen from fptools. This is left as a separate build system so that the main Cabal setup doesn't require a Unix build environment or DocBook XML tools. - - - - - aa36c783 by Duncan Coutts at 2006-01-17T19:29:55+00:00 Add a --wiki=URL flag to add a per-module link to a correspondng wiki page. So each html page gets an extra link (placed next to the source code and contents links) to a corresponding wiki page. The idea is to let readers contribute their own notes, examples etc to the documentation. Also slightly tidy up the code for the --source option. - - - - - e06e2da2 by Simon Marlow at 2006-01-18T09:28:15+00:00 TODO: documnet --wiki - - - - - 17adfda9 by Duncan Coutts at 2006-01-19T20:17:59+00:00 Add an optional wiki link for each top level exported name. In each module, for each "top level" exported entity we add a hyper link to a corresponding wiki page. The link url gets the name of the exported entity as a '#'-style anchor, so if there is an anchor in the page with that name then the users browser should jump directly to it. By "top level" we mean functions, classes, class members and data types (data, type, newtype), but not data constructors, class instances or data type class membership. The link is added at the right of the page and in a small font. Hopefully this is the right balance of visibility/distraction. We also include a link to the wiki base url in the contents and index pages. - - - - - f52324bb by Duncan Coutts at 2006-01-19T20:28:27+00:00 Rewrite pathJoin to only add a path separator when necessary. When the path ends in a file seperator there is no need to add another. Now using "--wiki=http://blah.com/foo/" should do the right thing. (Code snippet adapted from Isaac's FilePath package.) - - - - - 43bb89fa by Duncan Coutts at 2006-01-21T17:15:27+00:00 Teach haddock about line pragmas and add accurate source code links Teach haddock about C and Haskell style line pragmas. Extend the lexer/parser's source location tracking to include the file name as well as line/column. This way each AST item that is tagged with a SrcLoc gets the original file name too. Use this original file name to add source links to each exported item, in the same visual style as the wiki links. Note that the per-export source links are to the defining module rather than whichever module haddock pretends it is exported from. This is what we want for source code links. The source code link URL can also contain the name of the export so one could implement jumping to the actual location of the function in the file if it were linked to an html version of the source rather than just plain text. The name can be selected with the %N wild card. So for linking to the raw source code one might use: --source=http://darcs/haskell.org/foo/%F Or for linking to html syntax highlighted code: --source=http://darcs/haskell.org/foo/%M.html#%N - - - - - edd9f229 by Duncan Coutts at 2006-01-22T00:02:00+00:00 Extend URL variable expansion syntax and add source links to the contents page Like the wiki link on the contents and index page, add a source code link too. Extend the wiki & source URL variable expansion syntax. The original syntax was: %F for the source file name (the .hs version only, not the .lhs or .hs.pp one) %M for the module name (with '.' replaced by '/') The new syntax is: %F or %{FILE} for the original source file name %M or %{MODULE} for the module name (no replacements) %N or %{NAME} for the function/type export name %K or %{KIND} for a type/value flag "t" or "v" with these extensions: %{MODULE/./c} to replace the '.' module seperator with any other char c %{VAR|some text with the % char in it} which means if the VAR is not in use in this URL context then "" else replace the given text with the '%' char replaced by the string value of the VAR. This extension allows us to construct URLs wit optional parts, since the module/file name is not available for the URL in the contents/index pages and the value/type name is not available for the URL at the top level of each module. - - - - - eb3c6ada by Duncan Coutts at 2006-01-23T13:42:34+00:00 Remove the complex substitutions and add more command line flags instead. Instead of incomprehensable URL substitutions like ${MODULE/./-|?m=%} we now use three seperate command line flags for the top level, per-module and per-entity source and wiki links. They are: --source-base, --source-module, --source-entity --comments-base, --comments-module, --comments-entity We leave -s, --source as an alias for --source-module which is how that option behaved previously. The long forms of the substitutions are still available, ${FILE} ${MODULE} etc and the only non-trivial substitution is ${MODULE/./c} to replace the '.' characters in the module name with any other character c. eg ${MODULE/./-} Seperating the source and wiki url flags has the added bonus that they can be turned on or off individually. So users can have per-module links for example without having to also have per-entity links.` - - - - - a2f0f2af by Duncan Coutts at 2006-01-23T13:54:52+00:00 Make the --help output fit in 80 columns. This is a purely cosmetic patch, feel free to ignore it. The only trickery going on is that we don't display the deprecated -s, --source flags in the help message, but we do still accept them. - - - - - 2d3a4b0c by Duncan Coutts at 2006-01-23T14:12:16+00:00 Add documentation for the new --source-* and --comments-* command line options - - - - - 1a82a297 by Simon Marlow at 2006-01-23T17:03:27+00:00 fix markup - - - - - 100d464a by Duncan Coutts at 2006-01-23T18:31:13+00:00 remove a couple TODO items that have been done The --wiki, or rather the --comment-* options are now documented. There is probably no need to have haddock invoke unlit or cpp itself since it can now pick up the line pragmas to get the source locations right. Tools like Cabal will arrange for preprocessors to be run so there is less of a need for tools like haddock to do it themselves. - - - - - 3162fa91 by Simon Marlow at 2006-01-24T14:21:56+00:00 add a test I had lying around - - - - - 98947063 by Simon Marlow at 2006-01-31T13:52:54+00:00 add scabal-version field - - - - - c41876e6 by Neil Mitchell at 2006-02-26T17:48:21+00:00 Add Hoogle output option - - - - - f86fb9c0 by Simon Marlow at 2006-03-08T09:15:20+00:00 add haskell.vim Contributed by Brad Bowman <bsb at bereft.net>, thanks! - - - - - 35d3c511 by benjamin.franksen at 2006-03-03T22:39:54+00:00 fixed libdir (/html was missing) - - - - - 4d08fd7d by Simon Marlow at 2006-03-10T11:13:31+00:00 add PatternGuards extension - - - - - 3f095e70 by Simon Marlow at 2006-03-13T11:40:42+00:00 bug fixes from Brad Bowman - - - - - 8610849d by Sven Panne at 2006-03-19T17:02:56+00:00 Fixed Cabal/RPM build - - - - - 34a994d6 by sven.panne at 2006-04-20T12:39:23+00:00 Avoid pattern guards Due to the use of pattern guards in Haddock, GHC was called with -fglasgow-exts. This in turn enables bang patterns, too, which broke the Haddock build. Removing some unnecessary pattern guards seemed to be the better way of fixing this instead of using a pragma to disable pattern guards. - - - - - bb523f51 by Ross Paterson at 2006-04-24T09:03:25+00:00 extend 'deriving' heuristic a little If an argument of a data constructor has a type variable head, it is irreducible and the same type class can be copied into the constraint. (Formerly we just did this for type variable arguments.) - - - - - dab9fe7a by Simon Marlow at 2006-04-26T10:02:31+00:00 record an idea - - - - - 748b7078 by Simon Marlow at 2006-05-08T08:28:53+00:00 add section about deriving - - - - - 11252ea1 by Simon Marlow at 2006-05-24T15:43:10+00:00 replace a fatal error in lexChar with a parseError - - - - - 382c9411 by Simon Marlow at 2006-05-24T15:45:47+00:00 add a bug - - - - - b79272f5 by Simon Marlow at 2006-05-24T15:46:29+00:00 add a bug report - - - - - 912edf65 by David Waern at 2006-07-10T19:09:23+00:00 Initial modifications -- doesn't compile - - - - - a3c7ba99 by David Waern at 2006-07-11T00:54:19+00:00 More porting work -- doesn't compile - - - - - 0a173d19 by David Waern at 2006-07-11T11:30:03+00:00 Make the repos temporarily compile and illustrate a problem - - - - - bad316de by David Waern at 2006-07-11T15:43:47+00:00 Progress on the porting process - - - - - bbf12d02 by David Waern at 2006-07-11T23:07:44+00:00 More progress on the porting -- first pass starting to shape up - - - - - de580ba2 by David Waern at 2006-07-20T17:48:30+00:00 More progress -- still on phase1 - - - - - 75a917a2 by David Waern at 2006-07-23T18:22:43+00:00 More work on pass1 -- mostly done - - - - - 6697b3f7 by David Waern at 2006-07-23T22:17:40+00:00 More work, started working on the renaming phase -- this code will need a cleanup soon :) - - - - - 82a5bcbb by David Waern at 2006-07-29T16:16:43+00:00 Add instances, build renaming environment, start on the renamer - - - - - c3f8f4f1 by David Waern at 2006-07-29T21:37:48+00:00 Complete the renamer - - - - - 7e00d464 by David Waern at 2006-07-30T21:01:57+00:00 Start porting the Html renderer - - - - - f04ce121 by David Waern at 2006-08-09T20:04:56+00:00 More Html rendering progress - - - - - 20c21b53 by David Waern at 2006-08-10T17:37:47+00:00 More progress - - - - - d7097e0d by David Waern at 2006-08-11T20:31:51+00:00 Cleanup - - - - - a7351e86 by David Waern at 2006-08-12T11:44:47+00:00 Render H98 Data declarations - - - - - 3fb2208e by David Waern at 2006-08-12T17:15:34+00:00 Perfect rendering of Test.hs - - - - - 454fd062 by David Waern at 2006-08-13T21:57:08+00:00 Misc fixes and interface load/save - - - - - 7ef7e7be by David Waern at 2006-08-14T00:56:07+00:00 Some refactoring - - - - - a7d3efef by David Waern at 2006-08-19T20:07:55+00:00 Adapt to latest GHC - - - - - 5fc3c0d7 by David Waern at 2006-08-20T21:28:11+00:00 Move interface read/write to its own module + some cleanup - - - - - 037e011c by David Waern at 2006-08-20T21:38:24+00:00 Small cleanup - - - - - da3a1023 by David Waern at 2006-09-03T16:05:22+00:00 Change mode to BatchCompile to avoid GHC API bug - - - - - 3cc9be3b by David Waern at 2006-09-03T16:06:59+00:00 Starting work on GADT rendering - - - - - 94506037 by David Waern at 2006-09-03T20:02:48+00:00 Compensate for change of export list order in GHC - - - - - c2cec4eb by David Waern at 2006-09-04T20:53:01+00:00 Rename a function - - - - - 9a9735ba by David Waern at 2006-09-05T15:51:21+00:00 Change version number to 2.0 - - - - - 3758a714 by David Waern at 2006-09-05T15:51:49+00:00 Align comment properly - - - - - 68478d9e by David Waern at 2006-09-15T18:03:00+00:00 Remove interface reading/writing code and use the GHC api for creating package environments instead - - - - - d2eedd95 by David Waern at 2006-09-15T18:05:29+00:00 Change the executable name to haddock-ghc-nolib - - - - - fcfbcf66 by David Waern at 2006-09-15T18:05:45+00:00 Small source code cleanup - - - - - d08eb017 by David Waern at 2006-09-15T18:06:21+00:00 Remove handling of --package flag - - - - - b8a4cf53 by David Waern at 2006-09-15T18:07:16+00:00 Remove commented-out code - - - - - bef0a684 by David Waern at 2006-09-15T18:37:57+00:00 Don't warn about missing links to () - - - - - e7d25fd7 by David Waern at 2006-09-15T19:50:49+00:00 Remove Interface and Binary2 modules - - - - - 9894f2a1 by David Waern at 2006-09-15T19:53:43+00:00 Remove debug printing from HaddockHtml - - - - - a0e7455d by David Waern at 2006-09-16T00:16:29+00:00 Comments only - - - - - d5b26fa7 by David Waern at 2006-09-16T00:16:57+00:00 Refactor PackageData creation code and start on building the doc env propery (unfinished) - - - - - 06aaa779 by David Waern at 2006-09-16T00:19:25+00:00 Better comments in Main.hs - - - - - 1a52d1b4 by David Waern at 2006-09-18T22:17:11+00:00 Comments and spacing change - - - - - e5a97767 by David Waern at 2006-09-21T17:02:45+00:00 Remove unnecessary fmapM import in Main - - - - - 9d0f9d3a by David Waern at 2006-09-22T18:07:07+00:00 Make import list in HaddockHtml prettier - - - - - 3452f662 by David Waern at 2006-09-22T18:08:47+00:00 Refactor context rendering - - - - - 12d0a6d0 by David Waern at 2006-09-22T18:09:52+00:00 Do proper HsType rendering (inser parentheses correctly) - - - - - 2c20c2f9 by David Waern at 2006-09-22T18:10:45+00:00 Fix a bug in Main.toHsType - - - - - c5396443 by David Waern at 2006-09-22T18:11:16+00:00 Skip external package modules sort for now - - - - - 3fb95547 by David Waern at 2006-09-22T20:35:40+00:00 Take away trailin "2" on all previously clashing type names - - - - - 2174755f by David Waern at 2006-09-22T20:51:43+00:00 Remove unused imports in Main - - - - - 1e9f7a39 by David Waern at 2006-09-22T20:52:11+00:00 Fix a comment in Main - - - - - 32d9e028 by David Waern at 2006-10-05T16:40:11+00:00 Merge with changes to ghc HEAD - - - - - 3058c8f5 by David Waern at 2006-10-05T16:41:02+00:00 Comment fixes - - - - - b9c217ec by David Waern at 2006-10-05T16:49:59+00:00 Filter out more builtin type constructors from warning messages - - - - - 67e7d252 by David Waern at 2006-10-05T19:38:22+00:00 Refactoring -- better structured pass1 - - - - - cd21c0c1 by David Waern at 2006-10-05T19:44:42+00:00 Remove read/dump interface flags - - - - - 313f9e69 by David Waern at 2006-10-05T19:49:26+00:00 Remove unused pretty printing - - - - - 480f09d1 by David Waern at 2006-12-28T13:22:24+00:00 Update to build with latest GHC HEAD - - - - - 63dccfcb by David Waern at 2007-01-05T01:38:45+00:00 Fixed a bug so that --ghc-flag works correctly - - - - - 3117dadc by David Waern at 2006-12-29T18:53:39+00:00 Automatically get the GHC lib dir - - - - - 9dc84a5c by David Waern at 2006-12-29T19:58:53+00:00 Comments - - - - - 0b0237cc by David Waern at 2007-01-05T16:48:30+00:00 Collect docs based on SrcLoc, syncing with removal of DeclEntity from GHC - - - - - a962c256 by David Waern at 2007-01-05T17:02:47+00:00 Add tabs in haddock.cabal - - - - - 0ca30c97 by David Waern at 2007-01-05T17:04:11+00:00 Add GHCUtils.hs - - - - - c0ab9abe by David Waern at 2007-01-10T11:43:08+00:00 Change package name to haddock-ghc, version 0.1 - - - - - 38e18b27 by David Waern at 2007-01-12T12:03:52+00:00 No binder name for foreign exports - - - - - d18587ab by David Waern at 2007-01-12T12:08:15+00:00 Temp record - - - - - ba6251a0 by David Waern at 2007-01-12T18:27:55+00:00 Remove read/dump-interface (again) - - - - - f4ba2b39 by David Waern at 2007-01-12T18:31:36+00:00 Remove DocOption, use the GHC type - - - - - 511be8bd by David Waern at 2007-01-12T18:32:41+00:00 Use exceptions instead of Either when loading package info - - - - - 0f2144d8 by David Waern at 2007-01-12T18:33:23+00:00 Small type change - - - - - 77507eb7 by David Waern at 2007-01-12T18:33:59+00:00 Remove interface file read/write - - - - - 0ea1e14f by David Waern at 2007-01-17T21:40:26+00:00 Add trace_ppr to GHCUtils - - - - - 3878b493 by David Waern at 2007-01-17T21:40:53+00:00 Sort external package modules and build a doc env - - - - - 8dc323fc by David Waern at 2007-01-17T21:42:41+00:00 Remove comment - - - - - f4c5b097 by David Waern at 2007-01-18T23:22:18+00:00 Add haddock-ghc.cabal and remove ghc option pragma in source file - - - - - da242b2c by David Waern at 2007-01-18T23:22:46+00:00 Remove some tabs - - - - - 288ed096 by David Waern at 2007-01-18T23:39:28+00:00 Moved the defaultErrorHandler to scope only over sortAndCheckModules for now - - - - - 4dd150fe by David Waern at 2007-02-03T21:23:56+00:00 Let restrictCons handle infix constructors - - - - - 97893442 by David Waern at 2007-02-04T16:26:00+00:00 Render infix data constructors - - - - - da89db72 by David Waern at 2007-02-04T16:26:33+00:00 CHange project name to Haddock-GHC - - - - - e93d48af by David Waern at 2007-02-04T16:59:08+00:00 Render infix type constructors properly - - - - - 357bc99b by David Waern at 2007-02-04T17:37:08+00:00 Insert spaces around infix function names - - - - - ab6cfc49 by David Waern at 2007-02-04T17:59:54+00:00 Do not list entities without documentation - - - - - 04249c7e by David Waern at 2007-02-04T19:16:25+00:00 Add GADT support (quite untested) - - - - - 2c223f8d by David Waern at 2007-02-04T19:25:10+00:00 Add package file write/save again! - - - - - b07ed218 by David Waern at 2007-02-04T19:33:02+00:00 Comment out minf_iface based stuff - - - - - 953d1fa7 by David Waern at 2007-02-05T00:12:23+00:00 Solve conflicts - - - - - 593247fc by David Waern at 2007-02-06T19:48:48+00:00 Remove -package flag, GHC's can be used instead - - - - - f658ded2 by David Waern at 2007-02-06T20:50:44+00:00 Start for support of ATs - - - - - 97f9e913 by David Waern at 2007-02-06T20:52:27+00:00 Wibble - - - - - 2ce8e4cf by David Waern at 2007-02-16T12:09:49+00:00 Add the DocOptions change - - - - - dee4a9b5 by David Waern at 2007-03-06T01:24:48+00:00 Wibble - - - - - 7cb99d18 by David Waern at 2007-03-06T01:24:58+00:00 Change version to 2.0 and executable name to haddock - - - - - c5aa02bc by David Waern at 2007-03-08T15:59:49+00:00 Go back to -B flag - - - - - 3a349201 by David Waern at 2007-03-09T13:31:59+00:00 Better exception handling and parsing of GHC flags - - - - - 05a69b71 by David Waern at 2007-03-09T17:45:44+00:00 Remove commented-out DocEntity printing - - - - - 755032cb by davve at dtek.chalmers.se at 2007-03-23T23:30:20+00:00 Remove a file that shouldn't be here - - - - - a7077e5f by davve at dtek.chalmers.se at 2007-03-24T03:58:48+00:00 Remove an import - - - - - 6f55aa8b by davve at dtek.chalmers.se at 2007-03-25T00:46:48+00:00 Start work on Haddock API - - - - - f0199480 by davve at dtek.chalmers.se at 2007-03-25T00:56:36+00:00 Prettify some comments - - - - - f952f9d1 by davve at dtek.chalmers.se at 2007-03-25T00:56:53+00:00 Remove ppr in HaddockTypes - - - - - bc594904 by davve at dtek.chalmers.se at 2007-03-25T00:57:53+00:00 Remove commented out doc env inference - - - - - 11ebf08d by davve at dtek.chalmers.se at 2007-03-25T01:23:25+00:00 De-flatten the namespace - - - - - f696b4bc by davve at dtek.chalmers.se at 2007-03-25T03:21:48+00:00 Add missing stuff to API - - - - - 9a2a04c3 by davve at dtek.chalmers.se at 2007-03-25T03:22:02+00:00 Wibble - - - - - 7d04a6d5 by davve at dtek.chalmers.se at 2007-03-25T03:22:08+00:00 Avoid a GHC bug with parseStaticFlags [] - - - - - 4d2820ba by davve at dtek.chalmers.se at 2007-03-26T04:57:01+00:00 Add fall-through case to mkExportItem - - - - - 6ebc8950 by Stefan O'Rear at 2007-03-26T04:14:53+00:00 Add shebang line to Setup.lhs - - - - - 80966ec5 by davve at dtek.chalmers.se at 2007-03-26T05:24:26+00:00 Fix stupid compile error - - - - - 1ea1385d by davve at dtek.chalmers.se at 2007-04-05T17:19:56+00:00 Do save/read of interface files properly - - - - - 0e4f6541 by David Waern at 2007-04-10T21:08:36+00:00 Add version to ghc dependency - - - - - b0499b63 by David Waern at 2007-04-10T21:37:08+00:00 Change package name to haddock - - - - - 9d50d27e by David Waern at 2007-04-24T00:22:14+00:00 Use filepath package instead of FilePath - - - - - 87c7fcdf by David Waern at 2007-07-10T21:03:04+00:00 Add new package dependencies - - - - - 4768709c by David Waern at 2007-07-11T20:37:11+00:00 Follow changes to record constructor representation - - - - - b9a02fee by Simon Marlow at 2007-05-30T14:00:48+00:00 update to compile with the latest GHC & Cabal - - - - - c0ebdc01 by David Waern at 2007-07-11T21:35:45+00:00 Fix conflicts - - - - - 97f7afd4 by David Waern at 2007-07-11T21:52:38+00:00 Follow changes to the GHC API - - - - - a5b7b58f by David Waern at 2007-07-12T20:36:48+00:00 Call parseStaticFlags before newSession - - - - - f7f50dbc by David Waern at 2007-08-01T21:52:58+00:00 Better indentation in haddock.cabal - - - - - d84e52ad by David Waern at 2007-08-02T00:08:18+00:00 Wibble - - - - - a23f494a by David Waern at 2007-08-02T00:08:24+00:00 Be better at trying to load all module dependencies (debugging) - - - - - ee917f13 by David Waern at 2007-08-03T18:48:08+00:00 Load all targets explicitly (checkModule doesn't chase dependencies anymore) - - - - - 5182d631 by David Waern at 2007-08-16T16:48:55+00:00 Finalize support for links to other packages - - - - - dfd1e3da by David Waern at 2007-08-16T16:51:11+00:00 Fix haddock comment errors in Haddock.Types - - - - - 50c0d83e by David Waern at 2007-08-16T16:51:37+00:00 Remove a debug import - - - - - d84b7c2b by David Waern at 2007-08-16T17:06:30+00:00 Rename PackageData to HaddockPackage - - - - - 3b52cb9f by David Waern at 2007-08-16T17:09:42+00:00 Simplify some comments - - - - - 66fa68d9 by David Waern at 2007-08-16T17:11:38+00:00 Comment the HaddockPackage definition - - - - - 8674c761 by David Waern at 2007-08-16T17:25:54+00:00 Improve code layout in Main - - - - - 571a3a0b by David Waern at 2007-08-16T17:32:13+00:00 Remove explict module imports in Main - - - - - d31b3cb0 by David Waern at 2007-08-16T17:36:23+00:00 Correct comments - - - - - 7f8a9f2b by David Waern at 2007-08-16T17:39:50+00:00 Fix layout problems in Haddock.Types - - - - - 9f421d7f by David Waern at 2007-08-17T11:16:48+00:00 Move options out of Main into Haddock.Options - - - - - 80042b63 by David Waern at 2007-08-17T11:26:59+00:00 Small comment/layout fixes - - - - - b141b982 by David Waern at 2007-08-17T11:28:28+00:00 Change project name from Haddock-GHC to Haddock - - - - - dbeb4a81 by David Waern at 2007-08-17T11:41:05+00:00 Add top module comment to all files - - - - - ce99cc9e by David Waern at 2007-08-17T14:53:04+00:00 Factor out typechecking phase into Haddock.Typecheck - - - - - 6bf75d9e by David Waern at 2007-08-17T16:55:35+00:00 Factor out package code to Haddock.Packages - - - - - b396db37 by David Waern at 2007-08-29T22:40:23+00:00 Major refactoring - - - - - 3d4f95ee by David Waern at 2007-08-29T23:26:24+00:00 Rename HaddockModule to Interface and a few more refactorings - - - - - c55326db by David Waern at 2007-08-29T23:48:03+00:00 Some comment cleanup - - - - - 9a84fc46 by David Waern at 2007-08-29T23:49:29+00:00 Add some modules that I forgot to add earlier - - - - - 4536dce2 by David Waern at 2007-08-29T23:55:24+00:00 Wibble - - - - - 9b7f0206 by David Waern at 2007-08-30T16:03:29+00:00 Wibble - - - - - c52c050a by David Waern at 2007-08-30T16:30:37+00:00 Rename HaddockModule to Interface - - - - - eae2995f by David Waern at 2007-08-30T16:42:59+00:00 Simplify createInterfaces - - - - - 53f99caa by David Waern at 2007-09-29T00:04:31+00:00 Add build-type: Simple to the cabal file - - - - - 0d3103a8 by David Waern at 2007-09-29T00:04:58+00:00 Add containers and array dependency - - - - - 6acf5f30 by David Waern at 2007-09-29T00:13:36+00:00 Prettify the cabal file - - - - - 87c1e378 by David Waern at 2007-09-29T13:16:39+00:00 FIX: consym data headers with more than two variables - - - - - b67fc16a by David Waern at 2007-09-29T14:01:32+00:00 FIX: prefix types used as operators should be quoted - - - - - a8f925bc by David Waern at 2007-09-29T14:02:26+00:00 Use isSymOcc from OccName instead of isConSym - - - - - fc330701 by David Waern at 2007-09-29T14:15:37+00:00 Use isLexConSym/isLexVarSym from OccName - - - - - e4f3dbad by David Waern at 2007-09-29T15:01:08+00:00 FIX: do not quote varsym type operators - - - - - 402207d2 by David Waern at 2007-09-29T15:01:50+00:00 Wibble - - - - - f9d89ef0 by David Waern at 2007-09-29T15:17:40+00:00 Take care when pp tyvars - add parens on syms - - - - - 849e2a77 by David Waern at 2007-10-01T21:56:39+00:00 Go back to using a ModuleMap instead of LookupMod - fixes a bug - - - - - 549dbac6 by David Waern at 2007-10-02T01:05:19+00:00 Improve parsing of doc options - - - - - a36021b8 by David Waern at 2007-10-02T23:05:00+00:00 FIX: double arrows in constructor contexts - - - - - d03bf347 by David Waern at 2007-10-09T16:14:05+00:00 Add a simple test suite - - - - - c252c140 by David Waern at 2007-10-17T16:02:28+00:00 Add --optghc=.. style flag passing to GHC - - - - - cce6c1b3 by David Waern at 2007-10-18T22:03:20+00:00 Add support for --read-interface again - - - - - 33d059c0 by David Waern at 2007-10-18T22:30:18+00:00 Refactoring -- get rid of Haddock.Packages - - - - - f9ed0a4c by David Waern at 2007-10-18T22:34:36+00:00 Name changes - - - - - 8a1c816f by David Waern at 2007-10-20T14:24:23+00:00 Add --ghc-version option - - - - - 4925aaa1 by David Waern at 2007-10-21T14:34:26+00:00 Add some Outputable utils - - - - - 69e7e47f by David Waern at 2007-10-21T14:35:49+00:00 FIX: Ord for OrdName was not comparing modules - - - - - 5a4ae535 by David Waern at 2007-10-21T21:18:48+00:00 Wibble - - - - - 03d48e20 by David Waern at 2007-10-24T15:52:56+00:00 Remove Main from "other modules" - - - - - c66f6d82 by David Waern at 2007-10-24T16:37:18+00:00 Make it possible to run haddock on itself - - - - - 21d156d8 by David Waern at 2007-10-25T14:02:14+00:00 Don't set boot modules as targets - - - - - f8bcf91c by David Waern at 2007-10-31T22:11:17+00:00 Add optimisation flags - - - - - 7ac758f2 by David Waern at 2007-11-04T09:48:28+00:00 Go back to loading only targets (seems to work now) - - - - - 4862aae1 by David Waern at 2007-11-05T22:24:57+00:00 Do full compilation of modules -- temporary fix for GHC API problem - - - - - 697e1517 by David Waern at 2007-11-05T22:25:50+00:00 Don't warn about not being able to link to wired/system/builtin-names - - - - - 892186da by David Waern at 2007-11-06T00:49:21+00:00 Filter out instances with TyCons that are not exported - - - - - 9548314c by David Waern at 2007-11-06T09:37:14+00:00 Wibble - - - - - 5cafd627 by David Waern at 2007-11-08T01:43:07+00:00 Filter out all non-vanilla type sigs - - - - - 04621830 by David Waern at 2007-11-08T01:45:13+00:00 Synch loading of names from .haddock files with GHC's name cache - - - - - 88d37f77 by David Waern at 2007-11-08T01:46:21+00:00 Remove commented-out code - - - - - 6409c911 by David Waern at 2007-11-08T01:56:00+00:00 Small bugfix and cleanup in getDeclFromTyCls - - - - - af59d9c2 by David Waern at 2007-11-08T02:08:44+00:00 Remove OrdName stuff - - - - - 3a615e2e by David Waern at 2007-11-08T02:13:41+00:00 Update runtests.hs following changes to haddock - - - - - 01f3314e by David Waern at 2007-11-08T02:33:01+00:00 Complain if we can't link to wired-in names - - - - - fcafb5d1 by David Waern at 2007-11-09T02:40:16+00:00 Don't exit when there are no file arguments - - - - - 194bc332 by David Waern at 2007-11-09T02:55:37+00:00 Wibble - - - - - dbe4cb55 by David Waern at 2007-11-09T02:56:14+00:00 Wibble - - - - - 82869fda by David Waern at 2007-11-10T17:01:43+00:00 Introduce InstalledInterface structure and add more stuff to the .haddock files We introduce InstalledInterface capturing the part of Interface that is stored in the interface files. We change the ppHtmlContents and ppHtmllIndex to take this structure instead of a partial Interface. We add stuff like the doc map and exported names to the .haddock file (via InstalledInterface). - - - - - d6bb57bf by David Waern at 2007-11-10T17:19:48+00:00 FIX: contents and index should include external package modules when --gen-contents/--gen-index - - - - - e8814716 by David Waern at 2007-11-11T00:29:27+00:00 Remove lDocLinkName and its use in Html backend - - - - - 6f9bd702 by David Waern at 2007-11-11T00:50:57+00:00 Do some refactoring in the html backend This also merges an old patch by Augustsson: Wed Jul 12 19:54:36 CEST 2006 lennart.augustsson at credit-suisse.com * Print type definitions like signatures if given arrows. - - - - - 09d0ce24 by Malcolm.Wallace at 2006-07-20T13:13:57+00:00 mention HsColour in the docs, next to option flags for linking to source code - - - - - 24da6c34 by Malcolm.Wallace at 2006-07-20T13:14:50+00:00 change doc references to CVS to give darcs repository location instead - - - - - 74d52cd6 by David Waern at 2007-11-11T00:55:33+00:00 Update copyright - - - - - fcaa3b4f by Duncan Coutts at 2006-09-08T13:41:00+00:00 Eliminate dep on network by doing a little cut'n'paste haddock depending on the network causes a circular dependency at least if you want to build the network lib with haddock docs. - - - - - 10cc9bda by David Waern at 2007-11-11T02:09:41+00:00 Fix conflicts - - - - - 4e3acd39 by David Waern at 2007-11-11T02:21:19+00:00 Manual merge of a patch from Duncan Coutts that removes the dependency on mtl - - - - - fa9070da by Neil Mitchell at 2006-09-29T15:52:03+00:00 Do not generate an empty table if there are no exports, this fixes a <table></table> tag being generated, which is not valid HTML 4.01 - - - - - d7431c85 by David Waern at 2007-11-11T02:28:50+00:00 Fix conflicts - - - - - f87e8f98 by Simon Marlow at 2006-10-10T11:37:16+00:00 changes for 0.8 - - - - - db929565 by Simon Marlow at 2006-10-10T12:07:12+00:00 fix the name of the source file - - - - - 8220aa4b by Simon Marlow at 2006-10-11T14:17:37+00:00 Rename haddock.js to haddock-util.js haddock.js will be run automatically by Windows when you type 'haddock' if it is found on the PATH, so rename to avoid confusion. Spotted by Adrian Hey. - - - - - 6bccdaa1 by sven.panne at 2006-10-12T15:28:23+00:00 Cabal's sdist does not generate "-src.tar.gz" files, but ".tar.gz" ones - - - - - d3f3fc19 by Simon Marlow at 2006-12-06T16:05:07+00:00 add todo item for --maintainer - - - - - 2da7e269 by Simon Marlow at 2006-12-15T15:52:00+00:00 TODO: do something better about re-exported symbols from another package - - - - - 42d85549 by David Waern at 2007-11-11T02:30:59+00:00 Fix conflicts - - - - - 5e7ef6e5 by Neil Mitchell at 2007-01-11T15:41:15+00:00 Never do spliting index files into many - - - - - f3d4aebe by Neil Mitchell at 2007-01-11T17:07:09+00:00 Add searching on the index page - - - - - bad3ab66 by Neil Mitchell at 2007-01-11T18:17:46+00:00 Delete dead code, now there is only one index page - - - - - cd09eedb by Neil Mitchell at 2007-01-11T18:21:19+00:00 Delete more stuff that is no longer required - - - - - e2806646 by David Waern at 2007-11-11T02:41:53+00:00 Fix conflicts - - - - - a872a823 by Neil Mitchell at 2007-01-11T18:51:43+00:00 Make the index be in case-insensitive alphabetic order - - - - - 8bddd9d7 by Neil Mitchell at 2007-02-06T17:49:12+00:00 Do not create empty tables for data declarations which don't have any constructors, instances or comments. Gets better HTML 4.01 compliance - - - - - 036b8120 by David Waern at 2007-11-11T02:56:58+00:00 Fix conflicts - - - - - f50c1639 by Conal Elliott at 2007-02-14T21:54:00+00:00 added substitution %{FILE///c} - - - - - 402e166a by David Waern at 2007-11-11T03:35:46+00:00 Manual merge of old patch: Sat Apr 21 04:36:43 CEST 2007 Roberto Zunino <zunrob at users.sf.net> * URL expansion for %%, %L, %{LINE} - - - - - 2f264fbd by David Waern at 2007-11-11T03:40:33+00:00 Manual merge of an old patch: Thu Apr 19 20:23:40 CEST 2007 Wolfgang Jeltsch <g9ks157k at acme.softbase.org> * bug fix When Haddock was invoked with the --ignore-all-exports flag but the ignore-exports module attribute wasn't used, hyperlinks weren't created for non-exported names. This fix might not be as clean as one would wish (since --ignore-all-exports now results in ignore_all_exports = True *and* an additional OptIgnoreExports option for every module) but at least the bug seems to be resolved now. - - - - - 7d7ae106 by sven.panne at 2007-09-02T12:18:02+00:00 Install LICENSE in the correct place - - - - - 66eaa924 by David Waern at 2007-11-11T19:02:46+00:00 Fix a bug that made haddock loop - - - - - 4ed47b58 by David Waern at 2007-11-11T19:03:09+00:00 Rename java-script file (this wasn't merge correctly) - - - - - d569534a by David Waern at 2007-11-11T19:06:44+00:00 Don't require -B <ghc-libdir> when no argument files Change readInterfaceFile to take a Maybe Session, to avoid having to pass -B <ghc-libdir> to Haddock when there're no source files to process. This is nice when computing contents/index for external packages. - - - - - 373368bc by Neil Mitchell at 2007-01-11T18:22:44+00:00 Change from tabs to spaces in the ppHtmlIndex function - - - - - 6b063a77 by Neil Mitchell at 2007-01-12T12:17:46+00:00 Rewrite much of the index searching code, previously was too slow to execute on the base library with IE, the new version guarantees less than O(log n) operations be performed, where n is the number in the list (before was always O(n)) - - - - - bfad00b7 by David Waern at 2007-11-11T23:33:53+00:00 Fix conflicts - - - - - cd2dcc09 by Neil Mitchell at 2007-01-12T12:25:01+00:00 Make the max number of results 75 instead of 50, to allow map searching in the base library to work - - - - - 3ae74764 by Neil Mitchell at 2007-01-12T12:58:17+00:00 Make the search box in a form so that enter does the default search - - - - - 142103e5 by David Waern at 2007-11-12T00:03:18+00:00 Merge patch from the old branch: Fri Aug 31 13:21:45 CEST 2007 Duncan Coutts <duncan at haskell.org> * Add category: Development to .cabal file Otherwise it appears on the hackage website in the "Unclassified" category. - - - - - 22ec2ddb by David Waern at 2007-11-25T01:55:29+00:00 A a list of small improvements to the TODO file - - - - - eb0129f4 by Wolfgang Jeltsch at 2007-12-03T23:47:55+00:00 addition of type equality support (at least for HTML generation) - - - - - 816a7e22 by David Waern at 2007-12-08T15:46:26+00:00 Handle class operators correctly when rendering predicates - - - - - 68baaad2 by David Waern at 2007-12-08T16:15:54+00:00 Code layout changes - - - - - 09b77fb4 by David Waern at 2007-12-08T16:16:03+00:00 Handle infix operators correctly in the Type -> HsType translation - - - - - 31c36da2 by David Waern at 2007-12-08T16:24:27+00:00 Add ppLParendTypes/ppLParendType - - - - - b17cc818 by David Waern at 2007-12-08T16:26:12+00:00 Use ppParendType when printing types args in predicates - - - - - ffd1f2cf by David Waern at 2007-12-08T16:45:06+00:00 Fix rendering of instance heads to handle infix operators This is also a refactoring to share this code for rendering predicates. - - - - - ff886d45 by David Waern at 2007-12-08T17:27:46+00:00 Fix rendering of class operators - - - - - e2fcbb9e by David Waern at 2007-12-08T17:59:28+00:00 Fix a bug (use ppTyName instead of ppName to print names in type apps) - - - - - 79a1056e by David Waern at 2007-12-08T21:25:18+00:00 Update tests - - - - - 867741ac by David Waern at 2007-12-08T21:25:49+00:00 Give a diff on test failure - - - - - 7e5eb274 by David Waern at 2008-01-05T14:33:45+00:00 Add DrIFT commands - - - - - 3656454d by David Waern at 2008-01-05T20:26:00+00:00 Add "cabal-version: >= 1.2" to the cabal file - - - - - 77974efc by Simon Marlow at 2007-12-20T09:52:44+00:00 add an item - - - - - f6ac1708 by Simon Marlow at 2007-12-06T14:00:10+00:00 Source links must point to the original module, not the referring module - - - - - eda1d5c9 by David Waern at 2008-01-06T14:40:52+00:00 Manual merge of a patch to the 0.8 branch Thu Dec 6 15:00:10 CET 2007 Simon Marlow <simonmar at microsoft.com> * Source links must point to the original module, not the referring module - - - - - 378f4085 by David Waern at 2008-01-06T16:03:45+00:00 Change stability from stable to experimental - - - - - 8bdafe44 by David Waern at 2008-01-06T16:14:22+00:00 Add haskell.vim (it had been removed somehow) - - - - - ea34d02e by David Waern at 2008-01-06T16:36:57+00:00 Change version to 2.0.0.0 - - - - - 34631ac0 by David Waern at 2008-01-06T16:44:57+00:00 Add missing modules to the cabal file - - - - - 9e142935 by David Waern at 2008-01-06T17:25:42+00:00 Depend on ghc >= 6.8.2 && < 6.9 - - - - - 59f9eeaa by Simon Marlow at 2007-12-20T10:43:04+00:00 add build scripts - - - - - 1c29ae30 by Simon Marlow at 2007-12-20T10:47:07+00:00 update version number - - - - - fe16a3e4 by Simon Marlow at 2007-12-20T10:48:03+00:00 update version - - - - - f688530f by Simon Marlow at 2007-12-20T10:48:29+00:00 doc updates - - - - - ce71b611 by David Waern at 2008-01-07T13:46:32+00:00 Change version in docs and spec - - - - - 03ab8d6f by David Waern at 2008-01-07T13:47:38+00:00 Manually merge over changes to CHANGES for 0.9 - - - - - 39f1b042 by David Waern at 2008-01-07T15:17:41+00:00 Remove the -use-package flag, we don't support it anyway - - - - - 7274a544 by David Waern at 2008-01-07T15:33:05+00:00 Update CHANGES for 2.0.0.0 - - - - - 96594f5d by David Waern at 2008-01-07T15:46:49+00:00 Wibble - - - - - f4c5a4c4 by David Waern at 2008-01-07T15:55:36+00:00 Change url to repo in documentation - - - - - 8a4c77f0 by David Waern at 2008-01-07T16:00:54+00:00 Update CHANGES - - - - - cb3a9288 by David Waern at 2008-01-07T16:02:55+00:00 Documentation fix - - - - - d8e45539 by David Waern at 2008-01-07T16:12:00+00:00 Update docs to say that Haddock accets .lhs files and module names - - - - - 4b5ce824 by David Waern at 2008-01-07T16:12:25+00:00 Document -B option - - - - - 47274262 by David Waern at 2008-01-07T16:23:07+00:00 Update CHANGES - - - - - 7ff314a9 by David Waern at 2008-01-07T16:23:20+00:00 Remove --use-package, --package & --no-implicit.. flags from docs - - - - - 6c3819c0 by David Waern at 2008-01-07T16:23:52+00:00 Remove --no-implicit-prelide flag - - - - - 1b14ae40 by David Waern at 2008-01-07T16:32:26+00:00 Update the "Using literate or pre-processed source" section - - - - - 0117f620 by David Waern at 2008-01-07T16:41:55+00:00 Document the --optghc flag - - - - - 087ab1cf by David Waern at 2008-01-07T16:42:10+00:00 Remove the documenation section on derived instances The problem mentioned there doesn't exist in Haddock 2.0.0.0 - - - - - 7253951e by David Waern at 2008-01-07T16:48:40+00:00 Document OPTIONS_HADDOCK - - - - - 3b6bdcf6 by David Waern at 2008-01-07T16:56:54+00:00 Wibble - - - - - 3025adf9 by David Waern at 2008-01-07T17:08:14+00:00 Wibble - - - - - 5f30f1a0 by David Waern at 2008-01-07T17:15:44+00:00 Change synopsis field to description - - - - - 1673f54b by David Waern at 2008-01-07T17:18:21+00:00 Change my email address in the cabal file - - - - - 55aa9808 by David Waern at 2008-01-07T18:18:02+00:00 Add documentation for readInterfaceFile - - - - - eaea417f by David Waern at 2008-01-07T18:21:30+00:00 Export necessary stuff from Distribution.Haddock - - - - - 7ea18759 by David Waern at 2008-01-07T18:31:49+00:00 Remove dep on Cabal - - - - - 7b79c74e by David Waern at 2008-01-07T18:33:49+00:00 Remove dep on process - - - - - ce3054e6 by David Waern at 2008-01-16T23:01:21+00:00 Add feature-requsts from Henning Thielemann to TODO - - - - - 0c08f1ec by David Waern at 2008-01-16T23:03:02+00:00 Record a bug in TODO - - - - - b04605f3 by David Waern at 2008-01-23T16:59:06+00:00 Add a bug reported by Ross to TODO - - - - - 5b17c030 by David Waern at 2008-01-23T18:05:53+00:00 A a bug report to TODO - - - - - 1c993b0d by David Waern at 2008-01-25T16:30:25+00:00 Accept test output - - - - - c22fc0d0 by David Waern at 2008-01-25T16:34:49+00:00 Accept test output - - - - - 4b795811 by David Waern at 2008-01-25T16:38:37+00:00 Change Hidden.hs (test) to use OPTIONS_HADDOCK - - - - - c124dbd9 by David Waern at 2008-01-25T16:39:23+00:00 Accept test output - - - - - ec6f6eea by David Waern at 2008-01-25T16:42:08+00:00 Add Hidden.html.ref to tests - - - - - 1dc9610c by David Waern at 2008-02-02T20:50:51+00:00 Add a comment about UNPACK bug in TODO - - - - - 2d3f7081 by David Waern at 2008-02-09T22:33:24+00:00 Change the representation of DocNames Ross Paterson reported a bug where links would point to the defining module instead of the "best" module for an identifier (e.g Int pointing to GHC.Base instead of Data.Int). This patch fixes this problem by refactoring the way renamed names are represented. Instead of representing them by: > data DocName = Link Name | NoLink Name they are now represented as such: > data DocName = Documented Name Module | Undocumented Name and the the link-env looks like this: > type LinkEnv = Map Name Module There are several reasons for this. First of all, the bug was caused by changing the module part of Names during the renaming process, without changing the Unique field. This caused names to be overwritten during the loading of .haddock files (which caches names using the NameCache of the GHC session). So we might create new Uniques during renaming to fix this (but I'm not sure that would be problem-free). Instead, we just keep the Name and add the Module where the name is best documented, since it can be useful to keep the original Name around (for e.g. source-code location info and for users of the Haddock API). Also, the names Link/NoLink don't really make sense, since wether to use links or not is entirely up to the users of DocName. In the process of following this change into H.Backends.Html I removed the assumption that binder names are Undocumented (which was just an unnecessary assumption, the OccName is the only thing needed to render these). This will probably make it possible to get rid of the renamer and replace it with a traversal from SYB or Uniplate. Since DocName has changed, InterfaceFile has changed so this patch also increments the file-format version. No backwards-compatibility is implemented. - - - - - 0f28c921 by David Waern at 2008-02-09T23:00:36+00:00 H.GHC.Utils: remove unused imports/exports - - - - - 0c44cad5 by David Waern at 2008-02-10T00:28:13+00:00 H.GHC.Utils: add some functions that were removed by mistake - - - - - e3452f49 by David Waern at 2008-02-10T00:28:48+00:00 Fix some trivial warnings in H.InterfaceFile - - - - - a6d74644 by David Waern at 2008-02-10T00:48:06+00:00 Update the version message to fit in small terminals - - - - - 76c9cd3e by David Waern at 2008-02-10T14:47:39+00:00 Remove bugs from TODO that don't apply anymore since the port - - - - - 5e10e090 by David Waern at 2008-02-10T15:22:47+00:00 Remove bugs from TODO that weren't actual bugs - - - - - fef70878 by David Waern at 2008-02-10T15:23:44+00:00 Remove yet another item from TODO that was not an actual bug - - - - - e1af47b8 by David Waern at 2008-02-11T10:25:57+00:00 Bump the version number to 2.1.0 Since the exported datatype DocName has changed, we need to bump the major version number. Let's also drop the fourth version component, it's not that useful. - - - - - e3be7825 by David Waern at 2008-04-11T14:29:04+00:00 Add a bug to TODO - - - - - cb6574be by David Waern at 2008-04-11T16:00:45+00:00 Use the in-place haddock when running tests - - - - - c6d7af0d by David Waern at 2008-04-11T16:09:16+00:00 Turn off GHC warnings when running tests - - - - - 7f61b546 by David Waern at 2008-04-11T17:24:00+00:00 Add a flag for turning off all warnings - - - - - 883b8422 by David Waern at 2008-04-12T14:02:18+00:00 Fix printing of data binders - - - - - 2a0db8fc by David Waern at 2008-04-12T18:52:46+00:00 Fix missing parenthesis in constructor args bug - - - - - 1b3ac3f9 by David Waern at 2008-04-12T18:57:23+00:00 Simplify test suite and add tests I move all tests into one single directory to simplify things, and add a test for the last bug that was fixed. - - - - - 8f178376 by David Waern at 2008-04-12T19:00:15+00:00 Add a script for copying test output to "expected" output - - - - - 193e3a03 by David Waern at 2008-04-12T19:16:37+00:00 Remove two fixed bugs from TODO - - - - - ddc9130c by David Waern at 2008-04-12T19:37:06+00:00 Update test README - - - - - 956069c0 by David Waern at 2008-05-01T12:16:14+00:00 Update version number in spec and docs - - - - - 5478621c by David Waern at 2008-05-01T12:28:12+00:00 Remove claim of backwards compatibility from docs for readInterfaceFile - - - - - 4a16dea9 by David Waern at 2008-05-01T12:33:04+00:00 Update CHANGES - - - - - 804216fb by David Waern at 2008-05-01T12:43:16+00:00 Add a synopsis - - - - - fd0c84d5 by David Waern at 2008-05-01T12:44:44+00:00 Add Haddock.DocName to the cabal file - - - - - 9f4a7439 by David Waern at 2008-05-01T12:45:53+00:00 Remove -fglasgow-exts and -fasm - - - - - aee7c145 by David Waern at 2008-05-01T12:54:01+00:00 Add LANGUAGE pragmas to source files - - - - - 9a58428b by David Waern at 2008-05-01T12:54:19+00:00 Add extensions to cabal file - - - - - 494f1bee by David Waern at 2008-05-01T13:12:09+00:00 Export DocName in the API - - - - - c938196b by David Waern at 2008-05-01T13:12:19+00:00 Add hide options to some source files - - - - - 236e86af by Neil Mitchell at 2008-06-07T20:45:10+00:00 Rewrite the --hoogle flag support - - - - - 6d910950 by Neil Mitchell at 2008-06-14T10:56:50+00:00 Simplify the newtype/data outputting in Hoogle, as haddock does it automatically - - - - - f87a95a8 by Neil Mitchell at 2008-06-14T12:10:18+00:00 Add initial structure for outputting documentation as well, but does not yet output anything - - - - - 7c3bce54 by Neil Mitchell at 2008-06-14T12:27:07+00:00 Remove <document comment> from the Hoogle output - - - - - 9504a325 by Neil Mitchell at 2008-06-16T06:33:21+00:00 Default to "main" if there is no package, otherwise will clobber hoogle's hoogle info - - - - - 4a794a79 by Neil Mitchell at 2008-06-16T06:53:29+00:00 Change packageName to packageStr, as it better reflects the information stored in it - - - - - 7abc9baf by Neil Mitchell at 2008-06-16T07:09:49+00:00 Add modulePkgInfo to Haddock.GHC.Utils, which gives back package name and version info - - - - - 8ca11514 by Neil Mitchell at 2008-06-16T07:13:48+00:00 Change Hoogle to take the package name and package version separately - - - - - a6da452d by Neil Mitchell at 2008-06-18T11:29:46+00:00 In Hoogle do not list things that are not local to this module - - - - - 974b76b7 by David Waern at 2008-06-19T18:40:13+00:00 Be more consistent with GHC API naming in H.GHC.Utils - - - - - 2facb4eb by David Waern at 2008-06-19T19:03:03+00:00 Update test output - - - - - c501de72 by David Waern at 2008-06-26T20:26:49+00:00 Use ghc-paths to get the lib dir The path can still be overridden using the -B flag. It's not longer required to pass the lib dir to the program that runs the test suite. - - - - - ac4c6836 by David Waern at 2008-06-26T20:33:08+00:00 Update CHANGES - - - - - 9d21c60a by David Waern at 2008-06-26T20:34:53+00:00 Update README - - - - - 741448f0 by David Waern at 2008-06-26T21:12:57+00:00 Improve wording in the help message - - - - - b1b42b11 by David Waern at 2008-06-30T10:16:17+00:00 Rename ForeignType - - - - - 6d6c2b34 by David Waern at 2008-06-30T10:25:09+00:00 Rename TyFamily - - - - - 8d1125ed by David Waern at 2008-06-30T10:37:21+00:00 Rename type patterns - - - - - 7610a4cb by David Waern at 2008-06-30T10:45:07+00:00 Rename associated types - - - - - 8eeba14c by David Waern at 2008-06-30T10:47:41+00:00 Remove the TODO file now that we have a trac - - - - - 1af5b25b by David Waern at 2008-07-02T18:19:28+00:00 Render type family declarations (untested) - - - - - ceb99797 by David Waern at 2008-07-02T18:24:06+00:00 Remove redundant check for summary when rendering data types - - - - - b36a58e0 by David Waern at 2008-07-02T22:01:38+00:00 More support for type families and associated types Now we just need to render the instances - - - - - 78784879 by David Waern at 2008-07-07T22:13:58+00:00 Remove filtering of instances We were filtering out all instances for types with unknown names. This was probably an attempt to filter out instances for internal types. I am removing the filtering for the moment, and will try to fix this properly later. - - - - - 3e758dad by David Waern at 2008-06-30T18:50:30+00:00 Run haddock in-place during testing - - - - - d9dab0ce by David Waern at 2008-07-08T21:04:32+00:00 Remove index.html and doc-index.html from output, they should not be versioned - - - - - 3e6c4681 by David Waern at 2008-07-08T21:06:42+00:00 Update test output following change to instance filtering - - - - - e34a3f14 by David Waern at 2008-07-12T16:48:28+00:00 Stop using the map from exported names to declarations During creation of the interface, we were using two maps: one from exported names to declarations, and one from all defined names in the module to declarations. The first contained subordinate names while the second one didn't. The first map was never used to look up names not defined in the associated module, so if we add subordinate names to the second map, we could use it everywhere. That's that this patch does. This simplifies code because we don't have to pass around two maps everywhere. We now store the map from locally defined things in the interface structure instead of the one from exported names. - - - - - 2e1d2766 by David Waern at 2008-07-12T16:55:21+00:00 Get the all locally defined names from GHC API We previously had some code to compute all locally defined names in a module including subordinate names. We don't need it since we can get the names from modInfoTyThings in the GHC API. - - - - - bf637994 by David Waern at 2008-07-13T13:09:16+00:00 Refactoring in H.Interface.Create We were creating a doc map, a declaration map and a list of entities separately by going through the HsGroup. These structures were all used to build the interface of a module. Instead of doing this, we can start by creating a list of declarations from the HsGroup, then collect the docs directly from this list (instead of using the list of entities), creating a documentation map. We no longer need the Entity data type, and we can store a single map from names to declarations and docs in the interface, instead of the declaration map and the doc map. This way, there is only one place where we filter out the declarations that we don't want, and we can remove a lot of code. Another advantage of this is that we can create the exports directly out of the list of declarations when we export the full module contents. (Previously we did a look up for each name to find the declarations). This is faster and removes another point where we depend on names to identify exported declarations, which is good because it eliminates problems with instances (which don't have names). - - - - - 547e410e by David Waern at 2008-07-13T13:34:51+00:00 Remove FastString import and FSLIT macro in H.I.Create -- they were unused - - - - - 693759d1 by David Waern at 2008-07-13T13:36:23+00:00 Remove unused import from H.I.Create - - - - - cde6e7fb by David Waern at 2008-07-13T13:51:54+00:00 Small touches - - - - - 96de8f1d by David Waern at 2008-07-20T11:21:46+00:00 Preparation for rendering instances as separate declarations We want to be able to render instances as separate declarations. So we remove the Name argument of ExportDecl, since instances are nameless. This patch also contains the first steps needed to gather type family instances and display them in the backend, but the implementation is far from complete. Because of this, we don't actually show the instances yet. - - - - - b0f824fb by David Waern at 2008-07-20T15:53:08+00:00 Follow changes to ExportDecl in Hoogle - - - - - 1192eff3 by Neil Mitchell at 2008-06-26T00:28:10+00:00 Change how the Hoogle backend outputs classes, adding the context in - - - - - 7a0d1464 by Neil Mitchell at 2008-06-26T00:28:46+00:00 Remove the indent utility function from Hoogle backend - - - - - 3361241b by Neil Mitchell at 2008-06-26T09:45:09+00:00 Add support for Hoogle writing ForeignImport/ForeignExport properly - - - - - 795ad3bf by Neil Mitchell at 2008-06-26T12:15:25+00:00 Flesh out the Hoogle code to render documentation - - - - - 23277995 by Neil Mitchell at 2008-06-26T14:56:41+00:00 Fix a bug in the Hoogle backend, unordered lists were being written out <ul>...</u> - - - - - db739b27 by Neil Mitchell at 2008-06-26T15:09:54+00:00 Remove any white space around a <li> element - - - - - f2e6bb8c by Neil Mitchell at 2008-07-10T15:30:47+00:00 Remove the TODO in the Hoogle HTML generation, was already done - - - - - 693ec9a3 by Neil Mitchell at 2008-07-10T15:53:00+00:00 Put brackets round operators in more places in the Hoogle output - - - - - 842313aa by Neil Mitchell at 2008-07-10T16:01:25+00:00 Print type signatures with brackets around the name - - - - - cf93deb0 by David Waern at 2008-07-20T17:04:22+00:00 Bump version number to 2.2.0 - - - - - 30e6a8d1 by David Waern at 2008-07-20T17:04:41+00:00 Resolve conflicts in H.B.Hoogle - - - - - 1f0071c9 by David Waern at 2008-07-23T23:05:01+00:00 Add "all" command to runtests.hs that runs all tests despite failures - - - - - f2723023 by David Waern at 2008-07-23T23:08:39+00:00 Update tests/README - - - - - c0304a11 by David Waern at 2008-07-23T23:21:15+00:00 Be compatible with GHC 6.8.3 The cabal file is converted to use the "new" syntax with explicit Library and Executable sections. We define the __GHC_PATCHLEVEL__ symbol using a conditinal cpp-options field in the cabal file. (Ideally, Cabal would define the symbol for us, like it does for __GLASGOW_HASKELL__). We use these symbols to #ifdef around a small difference between 6.8.2 and 6.8.3. Previously, we only supported GHC 6.8.2 officially but the dependencies field said "ghc <= 6.9". This was just for convenience when testing against the (then compatible) HEAD version of GHC, and was left in the release by mistake. Now, we support both GHC 6.8.2 and 6.8.3 and the dependencies field correctly reflects this. - - - - - 88a5fe71 by David Waern at 2008-07-23T23:54:16+00:00 Depend on the currently available ghc-paths versions only - - - - - 8738d97b by David Waern at 2008-07-24T10:50:44+00:00 FIX haskell/haddock#44: Propagate parenthesis level when printing documented types - - - - - 05339119 by David Waern at 2008-07-24T16:06:18+00:00 Drop unnecessary parenthesis in types, put in by the user We were putting in parenthesis were the user did. Let's remove this since it just clutters up the types. The types are readable anyway since we print parens around infix operators and do not rely on fixity levels. When doing this I discovered that we were relying on user parenthesis when printin types like (a `O` b) c. This patchs fixes this problem so that parenthesis are always inserted around an infix op application in case it is applied to further arguments, or if it's an arguments to a type constructor. Tests are updated. - - - - - b3a99828 by David Waern at 2008-07-24T10:19:43+00:00 Print parenthesis around non-atomic banged types Fixes half of haskell/haddock#44 - - - - - ab5238e0 by David Waern at 2008-07-24T22:07:49+00:00 Add a reference file for the TypeFamilies test - - - - - 1941cc11 by David Waern at 2008-07-25T17:15:53+00:00 Simplify definition of pretty and trace_ppr - - - - - e3bfa33c by David Waern at 2008-07-25T17:18:27+00:00 Warning messages Output a warning when filtering out data/type instances and associated types in instances. We don't show these in the documentation yet, and we need to let the user know. - - - - - 9b85fc89 by David Waern at 2008-07-25T17:45:40+00:00 Doc: Mention Hoogle in the Introduction - - - - - afb2dd60 by David Waern at 2008-07-25T17:49:00+00:00 Doc: update -B description - - - - - 584c0c91 by David Waern at 2008-07-25T18:11:38+00:00 Doc: describe -w flag - - - - - 77619c24 by David Waern at 2008-07-28T12:29:07+00:00 Remove TODO from cabal file - - - - - 96717d5f by David Waern at 2008-07-28T12:29:27+00:00 Support type equality predicates - - - - - c2fd2330 by David Waern at 2008-07-29T19:45:14+00:00 Move unL from H.B.Hoogle to H.GHC.Utils I like Neil's shorter unL better than unLoc from the GHC API. - - - - - c4c3bf6a by David Waern at 2008-07-29T19:47:36+00:00 Do not export ATs when not in list of subitems - - - - - bf9a7b85 by David Waern at 2008-08-03T11:42:59+00:00 Filter out ForeignExports - - - - - df59fcb0 by David Waern at 2008-08-03T14:02:51+00:00 Filter out more declarations The previous refactorings in H.I.Create introduced a few bugs. Filtering of some types of declarations that we don't handle was removed. This patch fixes this. - - - - - 2f8a958b by David Waern at 2008-08-03T15:24:07+00:00 Move reL to H.GHC.Utils so we can use it everywhere - - - - - 8ec15efd by David Waern at 2008-08-03T15:25:00+00:00 Use isVanillaLSig from GHC API instead of home brewn function - - - - - 300f93a2 by David Waern at 2008-08-03T15:25:27+00:00 Filter out separately exported ATs This is a quick and dirty hack to get rid of separately exported ATs. We haven't decided how to handle them yet. No warning message is given. - - - - - 8776d1ec by David Waern at 2008-08-03T16:21:21+00:00 Filter out more declarations and keep only vanilla type sigs in classes - - - - - ea07eada by David Waern at 2008-08-03T16:48:00+00:00 Fix layout - - - - - dd5e8199 by David Waern at 2008-08-03T16:50:52+00:00 Move some utility functions from H.I.Create to H.GHC.Utils - - - - - 4a1dbd72 by David Waern at 2008-08-03T17:39:55+00:00 Do not filter out doc declarations - - - - - 0bc8dca4 by David Waern at 2008-08-03T17:47:26+00:00 Filter out separately exported ATs (take two) - - - - - af970fe8 by David Waern at 2008-08-03T22:39:17+00:00 Update CHANGES - - - - - 5436ad24 by David Waern at 2008-08-03T22:40:20+00:00 Bump version number to 2.2.1 - - - - - d66de448 by David Waern at 2008-08-05T19:00:32+00:00 Remove version restriction on ghc-paths - - - - - 534b1364 by David Waern at 2008-08-05T19:04:35+00:00 Bump version to 2.2.2 and update CHANGES - - - - - 549188ff by David Waern at 2008-08-05T19:16:49+00:00 Fix CHANGES - - - - - 0d156bb4 by Luke Plant at 2008-08-11T15:20:59+00:00 invoking haddock clarification and help - - - - - 748295cc by David Waern at 2008-08-11T18:56:37+00:00 Doc: say that the --hoogle option is functional - - - - - 43301db4 by David Waern at 2008-08-05T19:26:08+00:00 Change ghc version dependency to >= 6.8.2 - - - - - 3e5a53b6 by David Waern at 2008-08-10T22:42:05+00:00 Make H.GHC.Utils build with GHC HEAD - - - - - 7568ace0 by David Waern at 2008-08-11T19:41:54+00:00 Import Control.OldException instead of C.Exception when using ghc >= 6.9 We should really test for base version instead, but I don't currently know which version to test for. - - - - - b71ae991 by David Waern at 2008-08-12T22:40:39+00:00 Make our .haddock file version number depend on the GHC version We need to do this, since our .haddock format can potentially change whenever GHC's version changes (even when only the patchlevel changes). - - - - - 6307ce3f by David Waern at 2008-08-12T22:49:57+00:00 Remove matching on NoteTy in AttachInstances, it has been removed - - - - - 2dbcfd5f by David Waern at 2008-08-12T23:02:02+00:00 Comment out H.GHC.loadPackages - it is unused and doesn't build with ghc >= 6.9 - - - - - c74db5c2 by David Waern at 2008-08-12T23:03:58+00:00 Hide <.> from GHC import in Hoogle only for ghc <= 6.8.3 - - - - - 69a44ebb by David Waern at 2008-08-12T23:11:12+00:00 Follow changes to parseDynamic/StaticFlags - - - - - 5881f3f0 by David Waern at 2008-08-13T21:43:58+00:00 Add __GHC_PATCHLEVEL__ symbol also when building the library - - - - - 8574dc11 by David Waern at 2008-08-13T21:44:17+00:00 Follow move of package string functions from PackageConfig to Module - - - - - c9baa77f by David Waern at 2008-08-13T21:45:29+00:00 Follow extensible exceptions changes - - - - - 9092de15 by David Waern at 2008-08-13T21:46:20+00:00 Update test following Haddock version change - - - - - ebe569a4 by David Waern at 2008-08-13T21:46:54+00:00 Follow changes to parseDynamic- parseStaticFlags in GHC - - - - - b8a5ffd3 by David Waern at 2008-08-13T21:47:36+00:00 Follow changes to Binary in GHC 6.9 - - - - - edfda1cc by David Waern at 2008-08-13T21:50:17+00:00 Change ghc version dependency to >= 6.8.2 && <= 6.9 - - - - - d59be1cf by Neil Mitchell at 2008-08-12T16:02:53+00:00 Output all items, even if they are not defined in this module - ensures map comes from Prelude, not just GHC.Base - - - - - dda93b9f by Neil Mitchell at 2008-08-12T21:37:32+00:00 Add support for type synonyms to Hoogle, was accidentally missing before (woops!) - - - - - b6ee795c by Neil Mitchell at 2008-08-13T14:03:24+00:00 Generalise Hoogle.doc and add a docWith - - - - - 415e1bb2 by Neil Mitchell at 2008-08-13T14:03:46+00:00 Make Hoogle add documentation to a package - - - - - 790a1202 by Neil Mitchell at 2008-08-18T12:52:43+00:00 Use the same method to put out signatures as class methods in the Hoogle backend - - - - - ded37eba by Neil Mitchell at 2008-08-18T12:53:04+00:00 Remove Explicit top-level forall's when pretty-printing signatures - - - - - 6468c722 by Neil Mitchell at 2008-08-20T07:59:13+00:00 Simplify the code by removing not-to-important use of <.> in the Hoogle back end - - - - - 788c3a8b by Neil Mitchell at 2008-08-21T18:20:24+00:00 In the hoogle back end, markup definition lists using <i>, not <b> - - - - - 77d4b000 by Ian Lynagh at 2008-08-14T10:49:14+00:00 Add a Makefile for GHC's build system. Still won't work yet, but we're closer - - - - - 920440d7 by Ian Lynagh at 2008-08-27T18:06:46+00:00 Add haddock.wrapper - - - - - bcda925f by Ian Lynagh at 2008-08-27T18:07:02+00:00 Add a manual Cabal flag to control the ghc-paths dependency - - - - - 04d194e2 by Ian Lynagh at 2008-08-27T20:41:27+00:00 Update extensions in Cabal file Use ScopedTypeVariables instead of PatternSignatures - - - - - 12480043 by Ian Lynagh at 2008-08-27T20:41:55+00:00 Increase the upper bound on the GHC version number - - - - - b1f809a5 by Ian Lynagh at 2008-08-27T21:32:22+00:00 Fix some warnings - - - - - aea0453d by Ian Lynagh at 2008-08-28T14:22:29+00:00 Fixes for using haddock in a GHC build tree - - - - - ad23bf86 by Ian Lynagh at 2008-08-28T21:14:27+00:00 Don't use Cabal wrappers on Windows - - - - - 35858e4c by Ian Lynagh at 2008-08-29T00:07:42+00:00 Fix in-tree haddock on Windows - - - - - c2642066 by Ian Lynagh at 2008-09-03T22:35:53+00:00 follow library changes - - - - - 2eb55d50 by Ian Lynagh at 2008-09-07T18:52:51+00:00 bindist fixes - - - - - 3daa5b59 by Ian Lynagh at 2008-09-10T16:58:18+00:00 We need to tell haddock that its datasubdir is . or it can't find package.conf - - - - - 388fd8c2 by Ian Lynagh at 2008-09-10T19:47:44+00:00 Fix haddock inplace on Windows - - - - - 70a641c1 by Ian Lynagh at 2008-09-10T22:15:44+00:00 Fix installed haddock on Windows - - - - - 83c1e997 by Neil Mitchell at 2008-09-11T10:48:55+00:00 Import GHC.Paths if not IN_GHC_TREE, seems to match the use of GHC.Paths functions much better - - - - - b452519b by Ian Lynagh at 2008-09-12T12:58:24+00:00 Add a LANGUAGE ForeignFunctionInterface pragma - - - - - afbd592c by Ian Lynagh at 2008-09-12T12:59:13+00:00 Wibble imports - - - - - 547ac4ad by Ian Lynagh at 2008-09-14T15:34:22+00:00 Add a "#!/bin/sh" to haddock.wrapper - - - - - f207a807 by Ian Lynagh at 2008-09-15T10:02:32+00:00 Use "exec" when calling haddock in the wrapper - - - - - 2ee68509 by Thomas Schilling at 2008-09-15T09:09:16+00:00 Port Haddock.Interface to new GHC API. This required one bigger change: 'readInterfaceFile' used to take an optional 'Session' argument. This was used to optionally update the name cache of an existing GHC session. This does not work with the new GHC API, because an active session requires the function to return a 'GhcMonad' action, but this is not possible if no session is provided. The solution is to use an argument of functions for reading and updating the name cache and to make the function work for any monad that embeds IO, so it's result type can adapt to the calling context. While refactoring, I tried to make the code a little more self-documenting, mostly turning comments into function names. - - - - - 3bb96431 by Thomas Schilling at 2008-09-15T09:09:37+00:00 Reflect GHC API changes. - - - - - 2e60f714 by Thomas Schilling at 2008-09-15T09:10:37+00:00 Port Haddock.GHC.Typecheck to new GHC API. - - - - - 9cfd4cff by Thomas Schilling at 2008-09-15T09:11:00+00:00 Port Haddock.GHC to new GHC API. - - - - - caffa003 by Thomas Schilling at 2008-09-15T09:11:25+00:00 Port Main to new GHC API. - - - - - 069a4608 by Ian Lynagh at 2008-09-21T11:19:00+00:00 Fix paths used on Windows frmo a GHC tree: There is no whare directory - - - - - 7ceee1f7 by Ian Lynagh at 2008-09-21T12:20:16+00:00 Fix the in-tree haddock on Windows - - - - - 0d486514 by Ian Lynagh at 2008-09-23T18:06:58+00:00 Increase the GHC upper bound from 6.11 to 6.13 - - - - - f092c414 by Neil Mitchell at 2008-09-11T14:56:07+00:00 Do not wrap __ in brackets - - - - - 036bdd13 by Ian Lynagh at 2008-09-28T01:42:35+00:00 Fix building haddock when GhcProfiled=YES - - - - - 01434a89 by David Waern at 2008-09-24T20:24:21+00:00 Add PatternSignatures LANGUAGE pragma to Main and Utils - - - - - 1671a750 by David Waern at 2008-10-02T22:57:25+00:00 For source links, get original module from declaration name instead of environment. Getting it from the environment must have been a remnant from the times when we were using unqualified names (versions 0.x). - - - - - a25dde99 by David Waern at 2008-10-02T22:59:57+00:00 Remove ifaceEnv from Interface - it's no longer used - - - - - 610993da by David Waern at 2008-10-02T23:04:58+00:00 Write a comment about source links for type instance declarations - - - - - 5a96b5d5 by Thomas Schilling at 2008-10-03T10:45:08+00:00 Follow GHC API change of parseModule. - - - - - 5a943ae5 by Ian Lynagh at 2008-10-03T15:56:58+00:00 TAG 2008-10-03 - - - - - 76cdd6ae by Thomas Schilling at 2008-10-08T12:29:50+00:00 Only load modules once when typechecking with GHC. This still doesn't fix the memory leak since the typechecked source is retained and then processed separately. To fix the leak, modules must be processed directly after typechecking. - - - - - 7074d251 by David Waern at 2008-10-09T23:53:54+00:00 Interleave typechecking with interface creation At the same time, we fix a bug where the list of interfaces were processed in the wrong order, when building the links and renaming the interfaces. - - - - - 4b9b2b2d by David Waern at 2008-10-09T23:54:49+00:00 Add some strictness annotations in Interface We add some strictness annotations to the fields of Interface, so that less GHC data is hold on to during processing. - - - - - 22035628 by David Waern at 2008-10-10T20:02:31+00:00 Remove typecheckFiles and MonadUtils import from H.GHC.Typeccheck - - - - - be637ad3 by David Waern at 2008-10-10T20:33:38+00:00 Make Haddock build with GHC 6.8.2 - - - - - 523b3404 by David Waern at 2008-10-10T21:08:09+00:00 Fix documentation for createInterfaces - - - - - e1556702 by David Waern at 2008-10-10T21:26:19+00:00 Hide H.Utils in library - - - - - a8e751c3 by David Waern at 2008-10-10T21:34:59+00:00 Add back .haddock file versioning based on GHC version It was accidentally removed in the patch for GHC 6.8.2 compatibility - - - - - 06fb3c01 by David Waern at 2008-10-10T21:47:15+00:00 Bump version number to 2.3.0 - - - - - ff087fce by David Waern at 2008-10-10T22:35:49+00:00 Add support for DocPic The support for DocPic was merged into the GHC source long ago, but the support in Haddock was forgotten. Thanks Peter Gavin for submitting this fix! - - - - - 3af85bf6 by David Waern at 2008-10-10T23:34:05+00:00 Update tests - - - - - 0966873c by Simon Marlow at 2008-10-10T14:43:04+00:00 no need for handleErrMsg now, we don't throw any ErrMsgs - - - - - f1870de3 by Clemens Fruhwirth at 2008-10-10T13:29:36+00:00 Compile with wrapper but remove it for dist-install - - - - - 7b440dc2 by David Waern at 2008-10-11T14:02:25+00:00 Remove interface from LinksInfo It was there to know the documentation home module when creating a wiki link, but we already know this since we have the DocName. - - - - - e5729e6a by David Waern at 2008-10-15T20:49:18+00:00 Wibble - - - - - b2a8e01a by David Waern at 2008-10-15T21:03:36+00:00 Use type synonyms for declarations and docs in H.I.Create - - - - - be71a15b by David Waern at 2008-10-15T21:12:17+00:00 Comment out unused type family stuff completely - - - - - 91aaf075 by David Waern at 2008-10-15T21:49:04+00:00 Wibble - - - - - 42ba4eb4 by David Waern at 2008-10-15T21:53:53+00:00 Move convenient type synonym to H.Types - - - - - db11b723 by David Waern at 2008-10-15T22:14:07+00:00 Add DeclInfo to H.Types - - - - - 193552b6 by David Waern at 2008-10-15T22:15:01+00:00 Add subordinates with docs to the declaration map The only place in the code where we want the subordinates for a declaration is right after having looked up the declaration in the map. And since we include subordinates in the map, we might as well take the opportunity to store those subordinates that belong to a particular declaration together with that declaration. We also store the documentation for each subordinate. - - - - - 31e6eebc by David Waern at 2008-10-16T17:18:47+00:00 Wibble - - - - - 0dcbd79f by David Waern at 2008-10-16T20:58:42+00:00 Fix haskell/haddock#61 We were not getting docs for re-exported class methods. This was because we were looking up the docs in a map made from the declarations in the current module being rendered. Obviously, re-exported class methods come from another module. Class methods and ATs were the only thing we were looking up using the doc map, everything else we found in the ExporItems. So now I've put subordinate docs in the ExportItem's directly, to make things a bit more consistent. To do this, I added subordinates to the the declarations in the declaration map. This was easy since we were computing subordinates anyway, to store stand-alone in the map. I added a new type synonym 'DeclInfo', which is what we call what is now stored in the map. This little refactoring removes duplicate code to retrieve subordinates and documentation from the HsGroup. - - - - - de47f20a by David Waern at 2008-10-16T22:06:35+00:00 Document function and improve its layout - - - - - e74e625a by Thomas Schilling at 2008-10-20T11:12:57+00:00 Force interface more aggressively. For running Haddock on GHC this reduces memory usage by about 50 MB on a 32 bit system. A heap profile shows total memory usage peak at about 100 MB, but actual usage is at around 300 MB even with compacting GC (+RTS -c). - - - - - b63ac9a1 by David Waern at 2008-10-20T20:25:50+00:00 Make renamer consistent Instead of explicitly making some binders Undocumented, treat all names the same way (that is, try to find a Documented name). - - - - - f6de0bb0 by Ian Lynagh at 2008-09-19T00:54:43+00:00 TAG GHC 6.10 fork - - - - - 74599cd0 by David Waern at 2008-10-20T21:13:24+00:00 Do not save hidden modules in the .haddock file We were saving interfaces of all processed modules including those hidden using {-# OPTIONS_HADDOCK hide #-} in the .haddock file. This caused broken links when generating the index for the libraries that come with GHC. This patch excludes modules with hidden documentation when writing .haddock files. It should fix the above problem. - - - - - 7b6742e9 by David Waern at 2008-10-21T19:54:52+00:00 Do not save hidden modules in the .haddock file (also for ghc >= 6.9) When writing the first patch, I forgot to do the fix in both branches of an #if macro. - - - - - b99b1951 by David Waern at 2008-10-22T20:04:18+00:00 Remove subordinate map and its usage It is not needed now that we store subordinate names in the DeclInfo map. - - - - - da97cddc by David Waern at 2008-10-22T20:11:46+00:00 Tidy up code in H.I.Create a little Remove commented out half-done type instance support, and remove DeclWithDoc synonym. - - - - - 6afa76f3 by David Waern at 2008-10-22T21:17:29+00:00 Fix warnings in H.GHC.Utils - - - - - 171ea1e8 by David Waern at 2008-10-22T21:35:04+00:00 Fix warnings in H.Utils - - - - - c8cb3b91 by David Waern at 2008-10-22T21:36:49+00:00 Wibble - - - - - 767fa06a by David Waern at 2008-10-27T19:59:04+00:00 Make named doc comments into ExportDoc instead of ExportDecl Fixes a crash when processing modules without export lists containing named docs. - - - - - e638bbc6 by David Waern at 2008-11-02T22:21:10+00:00 Add HCAR entry - - - - - 92b4ffcf by David Waern at 2008-11-02T22:44:19+00:00 Update CHANGES - - - - - 84d4da6e by David Waern at 2008-11-03T11:25:04+00:00 Add failing test for template haskell crash - - - - - 2a9cd2b1 by David Waern at 2008-11-04T21:13:44+00:00 Add tests/TH.hs - - - - - 8a59348e by David Waern at 2008-11-04T21:30:26+00:00 TAG 2.3.0 - - - - - 54f70d31 by Thomas Schilling at 2008-10-24T17:04:08+00:00 Enable framed view of the HTML documentation. This patch introduces: - A page that displays the documentation in a framed view. The left side will show a full module index. Clicking a module name will show it in the right frame. If Javascript is enabled, the left side is split again to show the modules at the top and a very short synopsis for the module currently displayed on the right. - Code to generate the mini-synopsis for each module and the mini module index ("index-frames.html"). - CSS rules for the mini-synopsis. - A very small amount of javascript to update the mini-synopsis (but only if inside a frame.) Some perhaps controversial things: - Sharing code was very difficult, so there is a small amount of code duplication. - The amount of generated pages has been doubled, since every module now also gets a mini-synopsis. The overhead should not be too much, but I haven't checked. Alternatively, the mini-synopsis could also be generated using Javascript if we properly annotate the actual synopsis. - - - - - 5d7ea5a6 by David Waern at 2008-11-04T23:20:17+00:00 Follow change to ExportDecl in frames code - - - - - 60e16308 by David Waern at 2008-11-04T23:35:26+00:00 Update CHANGES - - - - - d63fd26d by David Waern at 2008-11-04T23:37:43+00:00 Bump version number - - - - - c1660c39 by David Waern at 2008-11-04T23:44:46+00:00 Update CHANGES - - - - - 995ab384 by David Waern at 2008-11-04T23:55:21+00:00 Remove .ref files from tests/output/ - - - - - 1abbbe75 by David Waern at 2008-11-04T23:57:41+00:00 Output version info before running tests - - - - - 649b182f by David Waern at 2008-11-05T22:45:37+00:00 Add ANNOUNCE message - - - - - c36ae0bb by David Waern at 2008-11-05T23:15:35+00:00 Update ANNOUNCE - - - - - 9c4f3d40 by David Waern at 2008-11-05T23:18:30+00:00 Wibble - - - - - 5aac87ce by David Waern at 2008-11-06T21:07:48+00:00 Depend on base 4.* when using GHC >= 6.9, otherwise 3.* - - - - - b9796a74 by David Waern at 2008-11-06T21:13:40+00:00 Bump version to 2.4.1 and update CHANGES - - - - - d4b26baa by David Waern at 2008-11-06T21:26:33+00:00 Depend on base 4.0.* instead of 4.* - - - - - 2cb0903c by David Waern at 2008-11-06T21:46:53+00:00 Fix warnings in H.B.HH and H.B.HH2 - - - - - e568e89a by David Waern at 2008-11-06T21:47:12+00:00 Fix warnings in Haddock.ModuleTree - - - - - 9dc14fbd by David Waern at 2008-11-06T21:47:52+00:00 Fix warnings in Haddock.Version - - - - - 02ac197c by David Waern at 2008-11-06T21:51:31+00:00 Fix warnings in H.InterfaceFile and H.Options - - - - - 63e7439a by David Waern at 2008-11-06T21:59:45+00:00 Fix warnings in H.GHC.Typecheck - - - - - 4bca5b68 by David Waern at 2008-11-08T13:43:42+00:00 Set HscTarget to HscNothing instead of HscAsm There used to be a bug in the GHC API that prevented us from setting this value. - - - - - 07357aec by David Waern at 2008-11-09T22:27:00+00:00 Re-export NameCache and friends from Distribution.Haddock - - - - - ea554b5a by David Waern at 2008-11-09T23:14:10+00:00 Add Haddock.GHC.Utils to other-modules in library - - - - - 74aecfd7 by David Waern at 2008-11-10T01:18:57+00:00 Export DocName in the library - - - - - 241a58b3 by David Waern at 2008-11-10T01:19:18+00:00 Document the functions in H.DocName - - - - - edc2ef1b by David Waern at 2008-11-10T01:20:52+00:00 Export H.DocName in the library - - - - - 4f588d55 by David Waern at 2008-11-10T01:29:14+00:00 Make DocName an instance of NamedThing - - - - - b4647244 by David Waern at 2008-11-15T22:58:18+00:00 Reflect version bump in test suite - - - - - 4bee8ce2 by David Waern at 2008-11-15T22:58:45+00:00 Update tests For unknown reasons, test output for Bug1 and Test has changed for the better. - - - - - 1690e2f9 by David Waern at 2008-11-15T22:59:33+00:00 Store hidden modules in .haddock files We store documentation for an entity in the 'InstalledInterface' of the definition site module, and never in the same structure for a module which re-exports the entity. So when a client of the Haddock library wants to look up some documentation, he/she might need to access a hidden module. But we currently don't store hidden modules in the .haddock files. So we add the hidden modules and the Haddock options to the .haddock files. The options will be used to filter the module list to obtain the visible modules only, which is necessary for generating the contents and index for installed packages. - - - - - 8add6435 by David Waern at 2008-11-16T14:35:50+00:00 Bump major version number due to .haddock file format change - - - - - 48bfcf82 by David Waern at 2008-11-23T14:32:52+00:00 Update tests to account for version number bump - - - - - 0bbd1738 by David Waern at 2008-11-23T14:33:31+00:00 HADDOCK_DATA_DIR changed to haddock_datadir - - - - - 5088b78c by David Waern at 2008-11-23T17:13:21+00:00 FIX haskell/haddock#45: generate two anchors for each name We generate two anchor tags for each name, one where we don't escape the name and one where we URI-encode it. This is for compatibility between IE and Opera. Test output is updated. - - - - - 5ee5ca3b by Neil Mitchell at 2008-11-27T14:38:11+00:00 Drop HsDocTy annotations, they mess up pretty printing and also have a bracketing bug (#2584) - - - - - 51c014e9 by Roman Cheplyaka at 2008-11-27T22:27:36+00:00 Allow referring to a specific section within a module in a module link Fixes haskell/haddock#65 - - - - - 4094bdc5 by David Waern at 2008-11-28T21:13:33+00:00 Update tests following anchor change - - - - - f89552dd by Thomas Schilling at 2008-11-29T16:16:20+00:00 Haddock really shouldn't try to overwrite files. - - - - - 98127499 by David Waern at 2008-12-07T14:09:15+00:00 Solve conflict - - - - - 319356c5 by David Waern at 2008-10-22T21:16:55+00:00 Add -Wall -Werror to ghc-options - - - - - 3c4968c9 by David Waern at 2008-11-04T23:38:56+00:00 TAG 2.4.0 - - - - - 4b21e003 by David Waern at 2008-11-06T21:14:04+00:00 TAG 2.4.1 - - - - - 8e0cad5c by David Waern at 2008-12-07T14:12:54+00:00 Remove -Werror - - - - - 299d6deb by David Waern at 2008-12-07T14:25:18+00:00 Remove -Wall, we'll focus on warnings after 6.10.2 is out - - - - - 5f4216b6 by David Waern at 2008-12-07T20:58:05+00:00 Resolve conflict properly - - - - - 67d774e7 by Neil Mitchell at 2008-12-15T11:44:26+00:00 Make forall's in constructors explicit, i.e. data Foo = Foo {foo :: Eq a => a} - - - - - 61851792 by Neil Mitchell at 2008-12-18T15:39:39+00:00 Try and find a better package name than "main" for Hoogle, goes wrong when working on an executable rather than a library - - - - - 2fab8554 by David Waern at 2008-12-08T23:19:48+00:00 Make visible names from ExportItems Instead of a complicated calculation of visible names out of GHC's export items, we can get them straight out of the already calculated ExportItems. The ExportItems should represent exactly those items that are visible in an interface. If store all the exported sub-names in ExportDecl instead of only those with documentation, the calculation becomes very simple. So we do this change as well (should perhaps have been a separate patch). This should fix the problem with names from ghc-prim not appearing in the link environment. - - - - - 7caadd8c by Ian Lynagh at 2008-12-11T17:01:04+00:00 Wrap the GHC usage with defaultCleanupHandler This fixes a bug where haddock leaves /tmp/ghc* directories uncleaned. - - - - - 7c9fc9a5 by David Waern at 2009-01-02T21:38:27+00:00 Show re-exported names from external packages again This fixes GHC ticket 2746. In order to also link to the exported subordinate names of a declaration, we need to re-introduce the sub map in the .haddock files. - - - - - 119e4e05 by David Waern at 2009-01-06T23:34:17+00:00 Do not process boot modules We should of course not try to produce documentation for boot modules! The reason this has worked in the past is that the output of "real" modules overwrites the output of boot modules later in the process. However, this causes a subtle link environment problem. So let's get rid of this stupid behaviour. We avoid processing boot modules, but we continue to typecheck them. - - - - - c285b9d2 by David Waern at 2009-01-08T18:03:36+00:00 Export modules also when coming from external packages This seems to have regressed since a refactoring that was part of the 2.3.0 release. - - - - - 24031c17 by David Waern at 2009-01-10T15:26:26+00:00 Change version to 2.4.2 - no need to go to 2.5.0 - - - - - 864d1c3f by David Waern at 2009-01-10T15:35:20+00:00 Update tests to account for version number change - - - - - 524ba886 by David Waern at 2009-01-10T18:29:17+00:00 Add test for Template Haskell splicing - - - - - 05e6e003 by David Waern at 2009-01-10T19:35:42+00:00 Fix Trac haskell/haddock#68: Turn on compilation via C for Template Haskell packages We can't use HscNothing if we need to run code coming from modules inside the processed package during typechecking, which is the case for some packages using Template Haskell. This could be improved, to e.g. use HscInterpreted and HscNothing where possible, instead of using HscC for all modules in the package. - - - - - 2b2bafa1 by David Waern at 2009-01-10T20:22:25+00:00 Only use needsTemplateHaskell when compiling with GHC 6.10.2 or above - - - - - bedc3a93 by Ian Lynagh at 2009-01-11T14:58:41+00:00 Fix the location of INPLACE_PKG_CONF; fixes the build Spotted by Conal Elliott - - - - - 943107c8 by David Waern at 2009-01-20T19:27:39+00:00 Document H.I.Create.collectDocs better - - - - - c6252e37 by David Waern at 2009-01-20T19:29:51+00:00 Fix Trac haskell/haddock#59: TH-generated declarations disappearing This patch was contributed by Joachim Breitner (nomeata). - - - - - 3568a6af by David Waern at 2009-01-21T21:41:48+00:00 Do not indicate that a constructor argument is unboxed We only show the strictness annotation for an unboxed constructor argument. The fact that it is unboxed is an implementation detail and should not be part of the module interface. - - - - - 562a4523 by David Waern at 2009-01-22T18:53:49+00:00 Fix Trac haskell/haddock#50: do not attach docs to pragmas or other kinds of non-declarations We now filter out everything that is not a proper Haskell declaration before collecting the docs and attaching them to declarations. - - - - - 6fdf21c2 by David Waern at 2009-01-22T19:48:09+00:00 Add test for quasi quotation. No reference output yet. - - - - - dc4100fd by David Waern at 2009-01-22T19:57:47+00:00 Improve quasi-quotation test and add reference output - - - - - 908b74bb by David Waern at 2009-01-23T23:22:03+00:00 Filter out separately exported associated types in a smarter way - - - - - f6b42ecb by David Waern at 2009-01-24T16:54:39+00:00 Correct spelling mistake in error message - - - - - 24e4245d by David Waern at 2009-01-24T17:48:03+00:00 Correct comment - - - - - b5e8462f by David Waern at 2009-02-07T13:22:29+00:00 Do not show a subordinate at the top level if its parent is also exported See note in the source code for more info. - - - - - 4b09de57 by David Waern at 2009-02-07T13:53:53+00:00 Update test following change to top level subordinates - - - - - 76379896 by David Waern at 2009-02-07T13:58:04+00:00 Remove html files in the tests/output/ directory which have been accidentally added - - - - - 1a6d8b10 by Joachim Breitner at 2009-02-20T10:29:43+00:00 Typo in comment - - - - - fec367d0 by David Waern at 2009-02-24T20:21:17+00:00 Fix small bug The rule is to prefer type constructors to other things when an identifier in a doc string can refer to multiple things. This stopped working with newer GHC versions (due to a tiny change in the GHC renamer). We implement this rule in the HTML backend for now, instead of fixing it in GHC, since we will move renaming of doc strings to Haddock in the future anyway. - - - - - 9b4172eb by David Waern at 2009-02-25T20:04:38+00:00 Fix bad error handling with newer GHCs When support for GHC 6.10 was added, an error handler was installed only around the typechecking phase. This had the effect that errors thrown during dependency chasing were caught in the top-level exception handler and not printed with enough detail. With this patch we wrap the error handler around all our usage of the Ghc monad. - - - - - de2df363 by Simon Peyton Jones at 2009-02-02T16:47:42+00:00 Hide funTyConName, now exported by TypeRep - - - - - 4d40a29f by Ian Lynagh at 2009-02-12T18:57:49+00:00 Don't build the library when building in the GHC tree - - - - - 1cd0abe4 by Ian Lynagh at 2009-02-13T13:58:53+00:00 Add a ghc.mk - - - - - 3d814eeb by Ian Lynagh at 2009-02-13T18:50:28+00:00 do .depend generation for haddock with the stage1 compiler This is a bit of a hack. We mkdepend with stage1 as if .depend depends on the stage2 compiler then make goes wrong: haddock's .depend gets included, which means that make won't reload until it's built, but we can't build it without the stage2 compiler. We therefore build the stage2 compiler before its .depend file is available, and so compilation fails. - - - - - b55036a4 by Ian Lynagh at 2009-02-25T01:38:13+00:00 Give haddock a wrapper on unix in the new GHC build system - - - - - 9eabfe68 by Ian Lynagh at 2009-02-25T19:21:32+00:00 Create inplace/lib/html in the new GHC build system - - - - - 93af30c7 by Ian Lynagh at 2008-11-07T19:18:23+00:00 TAG GHC 6.10.1 release - - - - - 06e6e34a by Thomas Schilling at 2009-02-24T18:11:00+00:00 Define __GHC_PATCHLEVEL__ for recent version of GHC (stable). - - - - - 680e6ed8 by Thomas Schilling at 2009-02-24T18:12:26+00:00 'needsTemplateHaskell' is not defined in current stable GHC. - - - - - 6c5619df by David Waern at 2009-02-25T22:15:23+00:00 Hide fynTyConName only for recent GHC versions - - - - - 6b2344f1 by Ian Lynagh at 2009-02-26T00:49:56+00:00 Add the module to one of haddocks warnings - - - - - e5d11c70 by David Waern at 2009-02-27T21:37:20+00:00 Bug fix We tried to filter out subordinates that were already exported through their parent. This didn't work properly since we were in some cases looking at the grand-parent and not the parent. We now properly compute all the parent-child relations of a declaration, and use this information to get the parent of a subordinate. We also didn't consider record fields with multiple parents. This is now handled correctly. We don't currently support separately exported associated types. But when we do, they should be handled correctly by this process too. Also slightly improved the warning message that we give when filtering out subordinates. - - - - - 10a79a60 by David Waern at 2009-02-27T22:08:08+00:00 Fix error message conflict The module name is already written in the beginning of the message, as seems to be the convention in Haddock. Perhaps not so clear, but we should change it everywhere in that case. Leaving it as it is for now. - - - - - c5055c7f by David Waern at 2009-02-27T22:15:17+00:00 Shorten warning message - - - - - a72fed3a by David Waern at 2009-02-28T00:53:55+00:00 Do not show package name in warning message - - - - - a5daccb2 by Ian Lynagh at 2009-03-01T14:59:35+00:00 Install haddock in the new GHC build system - - - - - dfdb025c by Ian Lynagh at 2009-03-07T23:56:29+00:00 Relax base dependency to < 4.2, not < 4.1 - - - - - 5769c8b4 by David Waern at 2009-03-21T14:58:52+00:00 Bump .haddock file version number (due to change of format) - - - - - f1b8f67b by David Waern at 2009-03-21T14:59:26+00:00 Define __GHC_PATCHLEVEL__=1 when using ghc-6.10.1 - - - - - 23f78831 by David Waern at 2009-03-21T16:40:52+00:00 Update CHANGES - - - - - 7d2735e9 by David Waern at 2009-03-21T16:50:33+00:00 Update ANNOUNCE - - - - - 0771e00a by David Waern at 2009-03-21T16:54:40+00:00 Update ANNOUNCE, again - - - - - 81a6942a by David Waern at 2009-03-21T17:50:06+00:00 Don't be too verbose in CHANGES - - - - - 29861dcf by David Waern at 2009-03-21T18:03:31+00:00 TAG 2.4.2 - - - - - a585f285 by David Waern at 2009-03-21T19:20:29+00:00 Require Cabal >= 1.2.3 - - - - - 7c611662 by David Waern at 2009-03-21T19:21:48+00:00 TAG 2.4.2 with cabal-version >= 1.2.3 - - - - - 23b7deff by Simon Marlow at 2009-03-20T15:43:42+00:00 new GHC build system: use shell-wrappers macro - - - - - 25f8afe7 by Ian Lynagh at 2009-03-21T19:13:53+00:00 Fix (with a hack?) haddock in teh new build system - - - - - 6a29a37e by David Waern at 2009-03-24T22:10:15+00:00 Remove unnecessary LANGUAGE pragma - - - - - 954da57d by David Waern at 2009-03-24T22:21:23+00:00 Fix warnings in H.B.DevHelp - - - - - 1619f1df by David Waern at 2009-03-26T23:20:44+00:00 -Wall police in H.B.Html - - - - - b211e13b by Simon Marlow at 2009-03-24T13:00:56+00:00 install Haddock's html stuff - - - - - 78e0b107 by David Waern at 2008-12-07T19:58:53+00:00 Add verbosity flag and utils, remove "verbose" flag - - - - - 913dae06 by David Waern at 2008-12-07T20:01:05+00:00 Add some basic "verbose" mode logging in H.Interface - - - - - 1cbff3bf by David Waern at 2009-03-27T00:07:26+00:00 Fix conflicts - - - - - 22f82032 by David Waern at 2009-03-27T21:15:11+00:00 Remove H.GHC.Typecheck - - - - - 81557804 by David Waern at 2009-03-27T21:19:22+00:00 Remove docNameOrig and use getName everywhere instead - - - - - d8267213 by David Waern at 2009-03-27T21:21:46+00:00 Use docNameOcc instead of nameOccName . getName - - - - - 5d55deab by David Waern at 2009-03-27T21:33:04+00:00 Remove H.DocName and put DocName in H.Types - - - - - 8ba72611 by David Waern at 2009-03-27T22:06:26+00:00 Document DocName - - - - - 605f8ca5 by David Waern at 2009-03-27T22:45:21+00:00 -Wall police - - - - - e4da93ae by David Waern at 2009-03-27T23:12:53+00:00 -Wall police in H.B.Hoogle - - - - - bb255519 by David Waern at 2009-03-27T23:41:28+00:00 Define Foldable and Traversable instances for Located - - - - - f1195cfe by David Waern at 2009-03-27T23:51:34+00:00 Wibble - - - - - 23818d7c by David Waern at 2009-03-28T00:03:55+00:00 -Wall police in H.I.Rename - - - - - 0f050d67 by David Waern at 2009-03-28T00:15:15+00:00 -Wall police in H.I.AttachInstances - - - - - 0f3fe038 by David Waern at 2009-03-28T21:09:41+00:00 Wibble - - - - - 275d4865 by David Waern at 2009-03-28T21:27:06+00:00 Layout fix - - - - - 54ff0ef8 by David Waern at 2009-03-28T21:59:07+00:00 -Wall police in H.I.Create - - - - - 7f58b117 by David Waern at 2009-03-28T22:10:19+00:00 -Wall police in H.Interface - - - - - f0c03b44 by David Waern at 2009-03-28T22:22:59+00:00 -Wall police in Main - - - - - 29da355c by David Waern at 2009-03-28T22:23:39+00:00 Turn on -Wall -Werror - - - - - 446d3060 by David Waern at 2009-04-01T20:40:30+00:00 hlint police - - - - - 3867c9fc by David Waern at 2009-04-01T20:48:42+00:00 hlint police - - - - - bd1f1600 by David Waern at 2009-04-01T20:58:02+00:00 hlint police - - - - - e0e90866 by David Waern at 2009-04-05T12:42:53+00:00 Move H.GHC.Utils to H.GhcUtils - - - - - 9cbd426b by David Waern at 2009-04-05T12:57:21+00:00 Remove Haddock.GHC and move its (small) contents to Main - - - - - b5c2cbfd by David Waern at 2009-04-05T13:07:04+00:00 Fix whitespace and stylistic issues in Main - - - - - 3c04aa56 by porges at 2008-12-07T08:22:19+00:00 add unicode output - - - - - 607918da by David Waern at 2009-04-26T15:09:43+00:00 Resolve conflict - - - - - 4bec6b6b by Simon Marlow at 2009-05-13T10:00:31+00:00 fix markup - - - - - 436ad6f4 by Simon Marlow at 2009-03-23T11:54:45+00:00 clean up - - - - - bdcd1398 by Simon Marlow at 2009-03-24T10:36:45+00:00 new GHC build system: add $(exeext) - - - - - 9c0972f3 by Simon Marlow at 2009-03-24T11:04:31+00:00 update for new GHC build system layout - - - - - d0f3f83a by Ian Lynagh at 2009-03-29T15:31:43+00:00 GHC new build system fixes - - - - - 5a8245c2 by Ian Lynagh at 2009-04-04T20:44:23+00:00 Tweak new build system - - - - - 9c6f2d7b by Simon Marlow at 2009-05-13T10:01:27+00:00 add build instructions for GHC - - - - - 66d07c76 by Ian Lynagh at 2009-05-31T00:37:53+00:00 Quote program paths in ghc.mk - - - - - bb7de2cd by Ian Lynagh at 2009-06-03T22:57:55+00:00 Use a bang pattern on an unlifted binding - - - - - 3ad283fc by Ian Lynagh at 2009-06-13T16:17:50+00:00 Include haddock in GHC bindists - - - - - ac447ff4 by David Waern at 2009-06-24T21:07:50+00:00 Delete Haddock.Exception and move contents to Haddock.Types Only a few lines of code that mainly declares a type - why not just put it in Haddock.Types. - - - - - 4464fb9b by David Waern at 2009-06-24T22:23:23+00:00 Add Haddock module headers Add a proper Haddock module header to each module, with a more finegrained copyright. If you feel mis-accreditted, please correct any copyright notice! The maintainer field is set to haddock at projects.haskell.org. Next step is to add a brief description to each module. - - - - - 5f4c95dd by David Waern at 2009-06-24T22:39:44+00:00 Fix spelling error - - - - - 6d074cdb by David Waern at 2009-06-25T21:53:56+00:00 Document Interface and InstalledInterface better - - - - - d0cbd183 by David Waern at 2009-06-27T12:46:46+00:00 Remove misplaced whitespace in H.I.Rename - - - - - fa381c49 by David Waern at 2009-06-27T13:26:03+00:00 Fix haskell/haddock#104 - create output directory if missing - - - - - 91fb77ae by Ian Lynagh at 2009-06-25T15:59:50+00:00 TAG 2009-06-25 - - - - - 0d853f40 by Simon Peyton Jones at 2009-07-02T15:35:22+00:00 Follow extra field in ConDecl - - - - - b201735d by Ian Lynagh at 2009-07-05T16:50:35+00:00 Update Makefile for the new GHC build system - - - - - df6c0092 by Ian Lynagh at 2009-07-05T17:01:13+00:00 Resolve conflicts - - - - - 1066870a by Ian Lynagh at 2009-07-05T17:01:48+00:00 Remove the -Wwarn hack in the GHC build system - - - - - 7e856076 by Ian Lynagh at 2009-07-05T17:17:59+00:00 Fix warnings - - - - - 5d4cd958 by Ian Lynagh at 2009-07-05T19:35:40+00:00 Bump version number Cabal needs to distinguish between haddocks having a --verbose and --verbosity flag - - - - - 6ee07c99 by David Waern at 2009-07-06T20:14:57+00:00 Wibble - - - - - 2308b66f by David Waern at 2009-07-06T20:24:20+00:00 Clearer printing of versions by runtests.hs - - - - - d4b5d9ab by David Waern at 2009-07-06T21:22:42+00:00 Fix (invisible) bug introduced by unicode patch - - - - - 2caca8d8 by David Waern at 2009-07-06T21:44:10+00:00 Use HscAsm instead of HscC when using TH - - - - - 18f3b755 by David Waern at 2009-07-06T22:10:22+00:00 Update HCAR entry (by Janis) - - - - - a72ac9db by David Waern at 2009-07-06T23:01:35+00:00 Follow HsRecTy change with an #if __GLASGOW_HASKEL__ >= 611 - - - - - 549135d2 by David Waern at 2009-07-06T23:11:41+00:00 Remove unused functions from Haddock.Utils - - - - - b450134a by Isaac Dupree at 2009-07-11T14:59:00+00:00 revert to split-index for large indices - remove the search-box, because browsers have search-for-text abilities anyway. - pick 150 items in index as the arbitrary time at which to split it - notice the bug that identifiers starting with non-ASCII characters won't be listed in split-index, but don't bother to fix it yet (see ticket haskell/haddock#116, http://trac.haskell.org/haddock/ticket/116 ) - - - - - 78a5661e by Isaac Dupree at 2009-07-20T15:37:18+00:00 Implement GADT records in HTML backend - - - - - 4e163555 by Isaac Dupree at 2009-07-21T22:03:25+00:00 add test for GADT records - - - - - 79aa4d6e by David Waern at 2009-07-23T20:40:37+00:00 Update test suite following version bump - - - - - 5932c011 by David Waern at 2009-08-02T10:25:39+00:00 Fix documentation bug - - - - - a6970fca by David Waern at 2009-08-12T23:08:53+00:00 Remove support for ghc 6.8.* from .cabal file - - - - - c1695902 by Ian Lynagh at 2009-07-07T13:35:45+00:00 Fix unused import warnings - - - - - fb6df7f9 by Ian Lynagh at 2009-07-16T00:20:31+00:00 Use cProjectVersion directly rather than going through compilerInfo Fixes the build after changes in GHC - - - - - 548cdd66 by Simon Marlow at 2009-07-28T14:27:04+00:00 follow changes in GHC's ForeignType - - - - - 9395aaa0 by David Waern at 2009-08-13T22:17:33+00:00 Switch from PatternSignatures to ScopedTypeVariables in Main - - - - - eebf39bd by David Waern at 2009-08-14T17:14:28+00:00 Version .haddock files made with GHC 6.10.3/4 correclty - - - - - 58f3e735 by David Waern at 2009-08-14T17:19:37+00:00 Support GHC 6.10.* and 6.11.* only - - - - - 5f63cecc by David Waern at 2009-08-14T22:03:20+00:00 Do not version .haddock file based on GHC patchlevel version We require that the instances of Binary that we use from GHC will not change between patchlevel versions. - - - - - d519de9f by David Waern at 2009-08-14T23:50:00+00:00 Update CHANGES - - - - - 35dccf5c by David Waern at 2009-08-14T23:51:38+00:00 Update version number everywhere - - - - - 6d363fea by David Waern at 2009-08-15T09:46:49+00:00 Update ANNOUNCE - - - - - c7ee6bc2 by David Waern at 2009-08-15T09:47:13+00:00 Remove -Werror Forgot that Hackage doesn't like it. - - - - - a125c12b by David Waern at 2009-08-15T09:49:50+00:00 Require Cabal >= 1.6 - - - - - adb2f560 by Isaac Dupree at 2009-08-12T03:47:14+00:00 Cross-Package Documentation version 4 - - - - - 3d6dc04d by David Waern at 2009-08-15T23:42:57+00:00 Put all the IN_GHC_TREE stuff inside getGhcLibDir - - - - - 56624097 by David Waern at 2009-08-15T23:52:03+00:00 Add --print-ghc-libdir - - - - - f15d3ccb by David Waern at 2009-08-16T00:37:52+00:00 Read base.haddock when running tests We can now test cross-package docs. - - - - - 283f0fb9 by David Waern at 2009-08-16T00:50:59+00:00 Update test output - we now have more links - - - - - 673d1004 by David Waern at 2009-08-16T01:26:08+00:00 Read process.haddock when running tests - - - - - 0d127f82 by David Waern at 2009-08-16T01:43:04+00:00 Add a test for cross-package documentation - - - - - f94db967 by Ian Lynagh at 2009-08-16T18:42:44+00:00 Follow GHC build system changes - - - - - 5151278a by Isaac Dupree at 2009-08-16T19:58:05+00:00 make cross-package list types look nicer - - - - - c41e8228 by Isaac Dupree at 2009-08-18T01:47:47+00:00 Haddock.Convert: export more functions This lets us remove some code in Haddock.Interface.AttachInstances - - - - - 2e5fa398 by Isaac Dupree at 2009-08-18T02:11:05+00:00 switch AttachInstances to use synify code It changed an instance from showing ((,) a b) to (a, b) because my synify code is more sophisticated; I hope the latter is a good thing rather than a bad thing aesthetically, here. But this definitely reduces code duplication! - - - - - b8b07123 by Isaac Dupree at 2009-08-18T02:23:31+00:00 Find instances using GHC, which is more complete. In particular, it works cross-package. An intermediate patch also moved the instance-finding into createInterface, but that move turned out not to be necessary, so if we want to do that, it'd go in a separate patch. (Is that possible? Or will we need GHC to have loaded all the modules first, before we can go searching for the instances (e.g. if the modules are recursive or something)?) - - - - - 6959b451 by Isaac Dupree at 2009-08-17T00:37:18+00:00 fix preprocessor conditional sense - - - - - 942823af by Isaac Dupree at 2009-08-16T22:46:48+00:00 remove ghc 6.8 conditionals from Haddock.Interface - - - - - 4b3ad888 by Isaac Dupree at 2009-08-18T20:24:38+00:00 Fix GHC 6.11 build in Haddock.Convert - - - - - 0a89c5ab by Isaac Dupree at 2009-08-23T00:08:58+00:00 hacks to make it compile without fnArgDocsn - - - - - 7b3bed43 by Isaac Dupree at 2009-08-23T03:01:28+00:00 less big-Map-based proper extraction of constructor subdocs - - - - - b21c279a by Isaac Dupree at 2009-08-23T03:02:06+00:00 Html: remove unnecessary+troublesome GHC. qualifications - - - - - 96c97115 by Isaac Dupree at 2009-08-23T03:08:03+00:00 Move doc parsing/lexing into Haddock for ghc>=6.11 - - - - - e1cec02d by Isaac Dupree at 2009-08-23T05:08:14+00:00 get rid of unused DocMap parameter in Html - - - - - 66960c59 by Isaac Dupree at 2009-08-23T05:54:20+00:00 fix horrible named-docs-disappearing bug :-) - - - - - a9d7eff3 by Isaac Dupree at 2009-08-23T06:26:36+00:00 re-implement function-argument docs ..on top of the lexParseRn work. This patch doesn't change the InstalledInterface format, and thus, it does not work cross-package, but that will be easy to add subsequently. - - - - - 8bf6852c by Isaac Dupree at 2009-08-23T07:26:05+00:00 cross-package fnArgDocs. WARNING: changes .haddock binary format While breaking the format, I took the opportunity to unrename the DocMap that's saved to disk, because there's really no reason that we want to know what *another* package's favorite place to link a Name to was. (Is that true? Or might we want to know, someday?) Also, I added instance Binary Map in InterfaceFile. It makes the code a little simpler without changing anything of substance. Also it lets us add another Map hidden inside another Map (fnArgsDocs in instDocMap) without having really-convoluted serialization code. Instances are neat! I don't understand why this change to InterfaceFile seemed to subtly break binary compatibility all by itself, but no matter, I'll just roll it into the greater format-changing patch. Done! - - - - - 30115a64 by Isaac Dupree at 2009-08-23T18:22:47+00:00 Improve behavior for unfindable .haddock - - - - - aa364bda by Isaac Dupree at 2009-08-23T18:28:16+00:00 add comment for FnArgsDoc type - - - - - 49b23a99 by Isaac Dupree at 2009-08-23T21:52:48+00:00 bugfix: restore fnArgDocs for type-synonyms - - - - - f65f9467 by Isaac Dupree at 2009-08-23T22:06:55+00:00 Backends.Hoogle: eliminate warnings - - - - - a292d216 by Isaac Dupree at 2009-08-23T22:10:24+00:00 Haddock.Convert: eliminate warnings - - - - - 5546cd20 by Isaac Dupree at 2009-08-23T22:12:31+00:00 Haddock.Interface.Rename: eliminate warnings - - - - - 0a9798b6 by Isaac Dupree at 2009-08-23T22:18:47+00:00 Main.hs: remove ghc<6.9 conditionals - - - - - e8f9867f by Isaac Dupree at 2009-08-23T22:27:46+00:00 Main.hs: eliminate warnings (except for OldException) - - - - - 61c64247 by Isaac Dupree at 2009-08-23T22:41:01+00:00 move get*LibDir code in Main.hs, to +consistent code, -duplication - - - - - 948f1e69 by Isaac Dupree at 2009-08-23T23:14:26+00:00 Main.hs: OldException->Exception: which eliminates warnings - - - - - 3d5d5e03 by Isaac Dupree at 2009-08-23T23:20:11+00:00 GhcUtils: ghc >= 6.10 - - - - - 2771d657 by Isaac Dupree at 2009-08-23T23:21:55+00:00 InterfaceFile: ghc >= 6.10 - - - - - d9f2b9d1 by Isaac Dupree at 2009-08-23T23:22:58+00:00 Types: ghc >= 6.10 - - - - - ca39210e by Isaac Dupree at 2009-08-23T23:23:26+00:00 ModuleTree: ghc >= 6.10 - - - - - 883c4e59 by Isaac Dupree at 2009-08-23T23:24:04+00:00 Backends.DevHelp: ghc >= 6.10 - - - - - 04667df5 by Isaac Dupree at 2009-08-23T23:24:37+00:00 Backends.Html: ghc >= 6.10 - - - - - a9f7f25f by Isaac Dupree at 2009-08-23T23:25:24+00:00 Utils: ghc >= 6.10 - - - - - b7105022 by Isaac Dupree at 2009-08-23T23:37:47+00:00 eliminate haskell98 dependency, following GHC's example It turns out I/we already had, and it was only a matter of deleting it from the cabal file. - - - - - 292e0911 by Isaac Dupree at 2009-08-24T01:22:44+00:00 refactor out subordinatesWithNoDocs dep of inferenced-decls fix - - - - - c2ed46a2 by Isaac Dupree at 2009-08-24T01:24:03+00:00 Eradicate wrong runtime warning for type-inferenced exported-functions see the long comment in the patch for why I did it this way :-) - - - - - 4ac0b57c by David Waern at 2009-09-04T22:56:20+00:00 Clean up tyThingToHsSynSig a little Factor out noLoc and use the case construct. Also rename the function to tyThingToLHsDecl, since it doesn't just create type signatures. - - - - - 28ab9201 by David Waern at 2009-09-04T22:58:50+00:00 Wibble - - - - - 0d9fe6d0 by David Waern at 2009-09-06T18:39:30+00:00 Add more copyright owners to H.I.AttachInstances - - - - - 122441b1 by David Waern at 2009-09-06T18:44:12+00:00 Style police - - - - - 1fa79463 by David Waern at 2009-09-06T18:57:45+00:00 Move toHsInstHead to Haddock.Convert and call it synifyInstHead - - - - - 0d42a8aa by David Waern at 2009-09-06T21:11:38+00:00 Use colordiff to display test results if available - - - - - ea9d8e03 by Simon Marlow at 2009-08-24T08:46:14+00:00 Follow changes in GHC's interface file format Word32 instead of Int for FastString and Name offsets - - - - - 537e051e by Simon Marlow at 2009-07-29T14:16:53+00:00 define unpackPackageId (it was removed from GHC) - - - - - 50c63aa7 by David Waern at 2009-09-09T23:18:03+00:00 Remove commented-out code - - - - - 511631fe by David Waern at 2009-09-09T23:19:05+00:00 Correct copyright in H.I.ParseModuleHeader - - - - - 898ec768 by David Waern at 2009-09-11T11:22:29+00:00 Use Map.fromList/toList intead of fromAscList/toAscList when serializing Maps This fixes the missing docs problem. The Eq and Ord instances for Name uses the unique number in Name. This number is created at deserialization time by GHC's magic Binary instance for Name, and it is random. Thus, fromAscList can't be used at deserialization time, even though toAscList was used at serialization time. - - - - - 37bec0d5 by Simon Peyton Jones at 2009-09-11T08:28:04+00:00 Track change in HsType - - - - - eb3a97c3 by Ian Lynagh at 2009-09-11T16:07:09+00:00 Allow building with base 4.2 - - - - - bb4205ed by Ian Lynagh at 2009-09-22T13:50:02+00:00 Loosen the GHC dependency - - - - - 5c75deb2 by Ian Lynagh at 2009-09-22T14:08:39+00:00 Fix building with GHC >= 6.12 - - - - - fb131481 by David Waern at 2009-09-11T11:24:48+00:00 Update runtests.hs to work with GHC 6.11 - - - - - ac3a419d by David Waern at 2009-09-11T11:25:14+00:00 Update CrossPackageDocs test - - - - - ec65c3c6 by David Waern at 2009-09-11T11:25:40+00:00 Add reference output for CrossPackageDocs - - - - - 520c2758 by Ian Lynagh at 2009-10-25T17:26:40+00:00 Fix installation in the GHC build system - - - - - 28b3d7df by Ian Lynagh at 2009-11-05T15:57:27+00:00 GHC build system: Make *nix installation work in paths containing spaces - - - - - 5c9bb541 by David Waern at 2009-11-14T11:56:39+00:00 Track change in HsType for the right compiler version - - - - - 905097ce by David Waern at 2009-11-14T12:10:47+00:00 hlint police - - - - - 04920630 by Ian Lynagh at 2009-11-20T13:46:30+00:00 Use defaultObjectTarget rather than HscAsm This fixes haddock when we don't have a native code generator - - - - - 966eb079 by David Waern at 2009-11-15T12:32:21+00:00 Remove commented-out code - - - - - 37f00fc4 by David Waern at 2009-11-22T13:58:48+00:00 Make runtests.hs strip links before diffing Generates easier to read diffs when tests fail. The content of the links is not important anyway since it is not taken into account by the tests. - - - - - 3a9bb8ef by David Waern at 2009-11-22T14:05:06+00:00 Follow findProgramOnPath signature change in runtests.hs - - - - - b26b9e5a by David Waern at 2009-11-22T14:08:40+00:00 Follow removal of GHC.MVar from base in CrossPackageDocs - - - - - f4d90ae4 by David Waern at 2009-11-22T14:48:47+00:00 Make copy.hs strip link contents before copying No more updating of reference files when URLs in links changes. - - - - - 4c9c420d by David Waern at 2009-11-22T15:26:41+00:00 Update test reference output * More links (Int, Float etc) * Stripped link contents - - - - - a62b80e3 by David Waern at 2009-11-23T23:19:39+00:00 Update CrossPackageDocs reference output - Remove GHC.MVar import (removed from base) - Strip link contents - - - - - 43491394 by David Waern at 2009-11-23T23:20:00+00:00 Update test reference files with comments on instances - - - - - 0d370a0b by David Waern at 2009-11-23T23:25:16+00:00 Bump version number - - - - - 2293113e by David Waern at 2009-11-24T20:55:49+00:00 Comments on instances Implementing this was a little trickier than I thought, since we need to match up instances from the renamed syntax with instances represented by InstEnv.Instance. This is due to the current design of Haddock, which matches comments with declarations from the renamed syntax, while getting the list of instances of a class/family directly using the GHC API. - Works for class instances only (Haddock has no support for type family instances yet) - The comments are rendered to the right of the instance head in the HTML output - No change to the .haddock file format - Works for normal user-written instances only. No comments are added on derived or TH-generated instances - - - - - bf586f29 by David Waern at 2009-11-27T22:05:15+00:00 Whitespace police - - - - - b8f03afa by David Waern at 2009-11-27T22:11:46+00:00 Remove bad whitespace and commented-out pieces - - - - - 90b8ee90 by David Waern at 2009-11-27T22:15:04+00:00 Whitespace police - - - - - b5ede900 by David Waern at 2009-11-27T22:15:50+00:00 Whitespace police - - - - - e3fddbfe by David Waern at 2009-11-28T13:37:59+00:00 Remove Name from DocInstance It's not used. - - - - - 9502786c by David Waern at 2009-11-28T13:56:54+00:00 Require at least GHC 6.12 While regression testing Haddock, I found a bug that happens with GHC 6.10.3, but not with GHC 6.12-rc2 (haven't tried 6.10.4). I don't have time to track it down. I think we should just always require the latest major GHC version. The time spent on making Haddock work with older versions is too high compared to the time spent on bugfixing, refactoring and features. - - - - - 8fa688d8 by David Waern at 2009-11-28T15:05:03+00:00 Remove cruft due to compatibility with older GHCs - - - - - 46fbbe9d by David Waern at 2009-11-28T15:07:50+00:00 Add a documentation header to Haddock.Convert - - - - - c3d2cc4a by David Waern at 2009-11-28T15:10:14+00:00 Remove unused H.Utils.FastMutInt2 - - - - - 490aba80 by David Waern at 2009-11-28T15:36:36+00:00 Rename Distribution.Haddock into Documentation.Haddock - - - - - 33ee2397 by David Waern at 2009-11-28T15:36:47+00:00 Fix error message - - - - - a5a3b950 by David Waern at 2009-11-28T16:58:39+00:00 Add a test flag that brings in QuickCheck - - - - - fa049e13 by David Waern at 2009-11-28T19:32:18+00:00 Say that we want quickcheck 2 - - - - - f32b0d9b by David Waern at 2009-11-28T19:32:40+00:00 Add an Arbitrary instance for HsDoc - - - - - da9a8bd7 by David Waern at 2009-11-28T20:15:30+00:00 Rename HsDoc back into Doc - - - - - edb60101 by David Waern at 2009-11-28T22:16:16+00:00 Move H.Interface.Parse/Lex to H.Parse/Lex These are not just used to build Interfaces. - - - - - 0656a9b8 by David Waern at 2009-11-28T23:12:14+00:00 Update version number in test suite - - - - - 5e8c6f4a by David Waern at 2009-12-21T14:12:41+00:00 Improve doc of DocName - - - - - 7868e551 by Ian Lynagh at 2009-09-22T10:43:03+00:00 TAG GHC 6.12-branch created - - - - - 0452a3ea by Ian Lynagh at 2009-12-15T12:46:07+00:00 TAG GHC 6.12.1 release - - - - - 65e9be62 by David Waern at 2009-12-21T16:58:58+00:00 Update CHANGES - - - - - 145cee32 by David Waern at 2009-12-21T16:59:09+00:00 TAG 2.6.0 - - - - - 3c552008 by David Waern at 2009-12-22T17:11:14+00:00 Update ANNOUNCE - - - - - 931f9db4 by David Waern at 2010-01-22T19:57:17+00:00 Convert haddock.vim to use unix newlines - - - - - 4e56588f by David Waern at 2010-01-22T22:11:17+00:00 Remove unnecessary (and inexplicable) uses of nub - - - - - 744bb4d1 by David Waern at 2010-01-22T22:12:14+00:00 Follow move of parser and lexer - - - - - e34bab14 by David Waern at 2010-01-22T22:49:13+00:00 Use findProgramLocation instead of findProgramOnPath in runtests.hs - - - - - 8d39891b by Isaac Dupree at 2010-01-14T18:53:18+00:00 fix html arg-doc off-by-one and silliness - - - - - 9401f2e9 by David Waern at 2010-01-22T22:57:03+00:00 Create a test for function argument docs - - - - - 507a82d7 by David Waern at 2010-01-22T23:24:47+00:00 Put parenthesis around type signature arguments of function type - - - - - 8a305c28 by David Waern at 2010-01-23T17:26:59+00:00 Add reference file for the FunArgs test - - - - - 1309d5e1 by David Waern at 2010-01-24T16:05:08+00:00 Improve FunArg test and update Test.html.ref - - - - - 2990f055 by Yitzchak Gale at 2010-02-14T16:03:46+00:00 Do not generate illegal character in HTML ID attribute. - - - - - c5bcab7a by David Waern at 2010-02-22T22:10:30+00:00 Fix Haddock markup error in comment - - - - - c6416a73 by David Waern at 2010-02-24T22:55:08+00:00 Large additions to the Haddock API Also improved and added more doc comments. - - - - - 57d289d7 by David Waern at 2010-02-24T22:58:02+00:00 Remove unused ifaceLocals - - - - - 80528d93 by David Waern at 2010-02-25T21:05:09+00:00 Add HaddockModInfo to the API - - - - - 82806848 by David Waern at 2010-02-25T21:05:27+00:00 Wibble - - - - - 744cad4c by David Waern at 2010-02-25T23:30:59+00:00 Make it possible to run a single test - - - - - 6a806e4c by David Waern at 2010-03-14T14:19:39+00:00 Bump version number - - - - - a5a8e4a7 by David Waern at 2010-03-14T14:36:35+00:00 Update ANNOUNCE - - - - - 6f05435e by Simon Hengel at 2010-03-15T20:52:42+00:00 Add missing dependencies for 'library' in haddock.cabal - - - - - faefe2bd by David Waern at 2010-03-15T22:29:37+00:00 Solve conflicts - - - - - 9808ad52 by David Waern at 2010-03-15T22:51:21+00:00 Bump version number - - - - - eb0bf60b by David Waern at 2010-03-15T22:52:32+00:00 Update CHANGES - - - - - f95cd891 by David Waern at 2010-03-15T23:01:06+00:00 Add Paths_haddock to other-modules of library - - - - - 65997b0a by David Waern at 2010-03-15T23:14:59+00:00 Update CHANGES - - - - - 7e251731 by David Waern at 2010-03-15T23:15:30+00:00 Bump version number - - - - - c9cd0ddc by David Waern at 2010-03-16T00:28:34+00:00 Fix warning - - - - - 1cac2d93 by Simon Peyton Jones at 2010-01-04T15:22:16+00:00 Fix imports for new location of splitKindFunTys - - - - - 474f26f6 by Simon Peyton Jones at 2010-02-10T14:36:06+00:00 Update Haddock for quasiquotes - - - - - 0dcc06c0 by Simon Peyton Jones at 2010-02-10T10:59:45+00:00 Track changes in HsTyVarBndr - - - - - 2d84733a by Simon Peyton Jones at 2010-02-10T14:52:44+00:00 Track HsSyn chnages - - - - - 9e3adb8b by Ian Lynagh at 2010-02-20T17:09:42+00:00 Resolve conflicts - - - - - a3e72ff8 by Simon Peyton Jones at 2010-03-04T13:05:16+00:00 Track change in HsUtils; and use a nicer function not an internal one - - - - - 27994854 by David Waern at 2010-03-18T22:22:27+00:00 Fix build with GHC 6.12.1 - - - - - 11f6e488 by David Waern at 2010-03-18T22:24:09+00:00 Bump version in test reference files - - - - - 0ef2f11b by David Waern at 2010-03-20T00:56:30+00:00 Fix library part of cabal file when in ghc tree - - - - - 3f6146ff by Mark Lentczner at 2010-03-20T22:30:11+00:00 First, experimental XHTML rendering switch to using the xhtml package copied Html.hs to Xhtml.hs and split into sub-modules under Haddock/Backends/Xhtml and detabify moved footer into div, got ready for iface change headers converted to semantic markup contents in semantic markup summary as semantic markup description in semantic markup, info block in header fixed factored out rendering so during debug it can be readable (see renderToString) - - - - - b8ab329b by Mark Lentczner at 2010-03-20T22:54:01+00:00 apply changes to Html.hs to Xhtml/*.hs incorporate changes that were made between the time Html.hs was copied and split into Xhtml.hs and Xhtml/*.hs includes patchs after "Wibble" (!) through "Fix build with GHC 6.12.1" - - - - - 73df2433 by Ian Lynagh at 2010-03-20T21:56:37+00:00 Follow LazyUniqFM->UniqFM in GHC - - - - - db4f602b by David Waern at 2010-03-29T22:00:01+00:00 Fix build with GHC 6.12 - - - - - d8dca088 by Simon Hengel at 2010-04-02T16:39:55+00:00 Add missing dependencies to cabal file - - - - - e2adc437 by Simon Hengel at 2010-04-02T14:08:40+00:00 Add markup support for interactive examples - - - - - e882ac05 by Simon Hengel at 2010-04-02T14:11:53+00:00 Add tests for interactive examples - - - - - 5a07a6d3 by David Waern at 2010-04-07T17:05:20+00:00 Propagate source positions from Lex.x to Parse.y - - - - - 6493b46f by David Waern at 2010-04-07T21:48:57+00:00 Let runtests.hs die when haddock has not been built - - - - - 5e34423e by David Waern at 2010-04-07T22:01:13+00:00 Make runtests.hs slightly more readable - - - - - 321d59b3 by David Waern at 2010-04-07T22:13:27+00:00 Fix haskell/haddock#75 Add colons to the $ident character set. - - - - - 37b08b8d by David Waern at 2010-04-08T00:32:52+00:00 Fix haskell/haddock#118 Avoid being too greedy when lexing URL markup (<..>), in order to allow multiple URLs on the same line. Do the same thing with <<..>> and #..#. - - - - - df8feac9 by David Waern at 2010-04-08T00:57:33+00:00 Make it easier to add new package deps to test suite This is a hack - we should use Cabal to get the package details instead. - - - - - 1ca6f84b by David Waern at 2010-04-08T01:03:06+00:00 Add ghc-prim to test suite deps - - - - - 27371e3a by Simon Hengel at 2010-04-08T19:26:34+00:00 Let parsing fails on paragraphs that are immediately followed by an example This is more consistent with the way we treat code blocks. - - - - - 83096e4a by David Waern at 2010-04-08T21:20:00+00:00 Improve function name - - - - - 439983ce by David Waern at 2010-04-10T10:46:14+00:00 Fix haskell/haddock#112 No link was generated for 'Addr#' in a doc comment. The reason was simply that the identifier didn't parse. We were using parseIdentifier from the GHC API, with a parser state built from 'defaultDynFlags'. If we pass the dynflags of the module instead, the right options are turned on on while parsing the identifer (in this case -XMagicHash), and the parse succeeds. - - - - - 5c0d35d7 by David Waern at 2010-04-10T10:54:06+00:00 Rename startGhc into withGhc - - - - - dca081fa by Simon Hengel at 2010-04-12T19:09:16+00:00 Add documentation for interactive examples - - - - - c7f26bfa by David Waern at 2010-04-13T00:51:51+00:00 Slight fix to the documentation of examples - - - - - 06eb7c4c by David Waern at 2010-04-13T00:57:05+00:00 Rename Interactive Examples into Examples (and simplify explanation) - - - - - 264830cb by David Waern at 2010-05-10T20:07:27+00:00 Update CHANGES with info about 2.6.1 - - - - - 8e5d4514 by Simon Hengel at 2010-04-18T18:16:54+00:00 Add unit tests for parser - - - - - 68297f40 by David Waern at 2010-05-10T21:53:37+00:00 Improve testsuite README - - - - - f04eb6e4 by David Waern at 2010-05-11T19:14:31+00:00 Re-organise the testsuite structure - - - - - a360f710 by David Waern at 2010-05-11T19:18:03+00:00 Shorten function name - - - - - 1d5dd359 by David Waern at 2010-05-11T21:40:02+00:00 Update runtests.hs following testsuite re-organisation - - - - - ffebe217 by David Waern at 2010-05-11T21:40:10+00:00 Update runtests.hs to use base-4.2.0.1 - - - - - 635de402 by David Waern at 2010-05-11T21:41:11+00:00 Update runparsetests.hs following testsuite reorganisation - - - - - 72137910 by Ian Lynagh at 2010-05-06T20:43:06+00:00 Fix build - - - - - 1a80b76e by Ian Lynagh at 2010-05-06T22:25:29+00:00 Remove redundant import - - - - - 1031a80c by Simon Peyton Jones at 2010-05-07T13:21:09+00:00 Minor wibbles to HsBang stuff - - - - - dd8e7fe5 by Ian Lynagh at 2010-05-08T15:22:00+00:00 GHC build system: Follow "rm" variable changes - - - - - 7f5e6748 by David Waern at 2010-05-13T11:53:02+00:00 Fix build with GHC 6.12.2 - - - - - 7953d4d8 by David Waern at 2010-05-13T18:45:01+00:00 Fixes to comments only - - - - - 8ae8eb64 by David Waern at 2010-05-13T18:57:26+00:00 ModuleMap -> IfaceMap - - - - - 1c3eadc6 by David Waern at 2010-05-13T19:03:13+00:00 Fix whitespace style issues - - - - - e96783c0 by David Waern at 2010-05-13T19:08:53+00:00 Fix comment - - - - - c998a78b by David Waern at 2010-05-13T19:39:00+00:00 Position the module header the same way everywhere Silly, but nice with some consistency :-) - - - - - b48a714e by David Waern at 2010-05-13T19:41:32+00:00 Position of module header, this time in the HTML backends - - - - - f9bfb12e by David Waern at 2010-05-13T19:43:05+00:00 Two newlines between declarations in Main - - - - - 071d44c7 by David Waern at 2010-05-13T19:44:21+00:00 Newlines in Convert - - - - - 036346db by David Waern at 2010-05-13T19:46:47+00:00 Fix a few stylistic issues in H.InterfaceFile - - - - - f0b8379e by David Waern at 2010-05-13T19:47:53+00:00 Add newlines to H.ModuleTree - - - - - 27409f8e by David Waern at 2010-05-13T19:51:10+00:00 Fix stylistic issues in H.Utils - - - - - 24774a11 by David Waern at 2010-05-13T20:00:43+00:00 Structure H.Types better - - - - - 7b6f5e40 by David Waern at 2010-05-13T20:01:04+00:00 Remove bad Arbitrary instance - - - - - fac9f1f6 by David Waern at 2010-05-13T20:05:50+00:00 Get rid of H.Utils.pathJoin and use System.FilePath.joinPath instead - - - - - fe6d00c4 by David Waern at 2010-05-13T20:51:55+00:00 Export a couple of more types from the API - - - - - b2e33a5f by David Waern at 2010-05-13T21:27:51+00:00 Improve doc comment for Interface - - - - - c585f2ce by David Waern at 2010-05-13T21:30:14+00:00 Improve documentation of Haddock.Interface - - - - - e6791db2 by David Waern at 2010-05-13T22:07:35+00:00 Remove meaningless comments - - - - - 7801b390 by David Waern at 2010-05-14T17:53:33+00:00 Remove unused modules - - - - - f813e937 by David Waern at 2010-05-14T17:55:17+00:00 Re-direct compilation output to a temporary directory Also add a flag --no-tmp-comp-dir that can be used to get the old behaviour of writing compilation files to GHC's output directory (default "."). - - - - - e56737ec by David Waern at 2010-05-14T18:06:11+00:00 Wibble - - - - - e40b0447 by David Waern at 2010-05-14T19:01:52+00:00 Move flag evaluation code from Main to Haddock.Options Determining the value of "singular" flags (by e.g. taking the last occurrence of the flag) and other flag evaluation should done in Haddock.Options which is the module that is supposed to define the command line interface. This makes Main a bit easier on the eyes as well. - - - - - 27091f57 by David Waern at 2010-05-14T19:05:10+00:00 Wibble - - - - - c658cf61 by David Waern at 2010-05-14T19:06:49+00:00 Re-order things in Haddock.Options a bit - - - - - 8cfdd342 by David Waern at 2010-05-14T19:20:29+00:00 De-tabify Haddock.Options and fix other whitespace issues - - - - - 0df16b62 by David Waern at 2010-05-14T19:25:07+00:00 Improve comments - - - - - 80b38e2b by David Waern at 2010-05-14T19:26:42+00:00 Whitespace police - - - - - fe580255 by David Waern at 2010-05-14T19:31:23+00:00 Wibbles to comments - - - - - a2b43fad by David Waern at 2010-05-14T20:24:32+00:00 Move some more flag functions to Haddock.Options - - - - - 3f895547 by David Waern at 2010-05-14T20:37:12+00:00 Make renderStep a top-level function in Main - - - - - 5cdca11d by David Waern at 2010-05-14T20:39:27+00:00 Spelling in comment - - - - - ad98d14c by David Waern at 2010-05-14T20:40:26+00:00 Comment fixes - - - - - 0bb9218f by David Waern at 2010-05-14T20:49:01+00:00 Whitespace police - - - - - 0f0a533f by David Waern at 2010-05-15T16:42:29+00:00 Improve description of --dump-interface - - - - - 5b2833ac by David Waern at 2010-05-15T17:16:53+00:00 Document --no-tmp-comp-dir - - - - - 8160b170 by David Waern at 2010-05-15T17:18:59+00:00 Wibble - - - - - 570dbe33 by David Waern at 2010-05-18T21:15:38+00:00 HLint police - - - - - 204e425f by David Waern at 2010-05-18T21:16:30+00:00 HLint police - - - - - 6db657ac by David Waern at 2010-05-18T21:16:37+00:00 Wibble - - - - - b942ccd7 by Simon Marlow at 2010-06-02T08:27:30+00:00 Interrupted disappeared in GHC 6.13 (GHC ticket haskell/haddock#4100) - - - - - 3b94a819 by Simon Marlow at 2010-06-02T08:45:08+00:00 Allow base-4.3 - - - - - c5a1fb7c by Simon Marlow at 2010-06-02T09:03:04+00:00 Fix compilation with GHC 6.13 - - - - - 6181296c by David Waern at 2010-06-08T21:09:05+00:00 Display name of prologue file when parsing it fails - - - - - 7cbc6f60 by Ian Lynagh at 2010-06-13T16:20:25+00:00 Remove redundant imports - - - - - 980c804b by Simon Marlow at 2010-06-22T08:41:50+00:00 isLocalAndTypeInferenced: fix for local module names overlapping package modules - - - - - d74d4a12 by Simon Marlow at 2010-06-23T12:03:27+00:00 Unresolved identifiers in Doc get replaced with DocMonospaced rather than plain strings - - - - - d8546783 by Simon Marlow at 2010-06-30T12:45:17+00:00 LaTeX backend (new options: --latex, --latex-style=<style>) - - - - - 437afa9e by David Waern at 2010-07-01T12:02:44+00:00 Fix a few stylistic whitespace issues in LaTeX backend - - - - - 85bc1fae by David Waern at 2010-07-01T15:42:45+00:00 Make runtest.hs work with GHC 6.12.3 (we should really stop hard coding this) - - - - - 7d2eb86f by David Waern at 2010-07-01T15:43:33+00:00 Update test following Simon's patch to render unresolved names in monospaced font - - - - - 08fcbcd2 by David Waern at 2010-07-01T16:12:18+00:00 Warning police - - - - - d04a8d7a by David Waern at 2010-07-04T14:53:39+00:00 Fix a bug in attachInstances We didn't look for instance docs in all the interfaces of the package. This had the effect of instance docs not always showing up under a declaration. I took the opportunity to clean up the code in H.I.AttachInstances a bit as well. More cleanup is needed, however. - - - - - d10344eb by Simon Hengel at 2010-07-10T09:19:04+00:00 Add missing dependencies to cabal file - - - - - 24090531 by Mark Lentczner at 2010-03-21T04:51:16+00:00 add exports to Xhtml modules - - - - - 84f9a333 by Mark Lentczner at 2010-04-03T19:14:22+00:00 clean up Doc formatting code - add CSS for lists - renderToString now uses showHtml since prettyHtml messes up <pre> sections - - - - - bebccf52 by Mark Lentczner at 2010-04-04T04:51:08+00:00 tweak list css - - - - - 0c2aeb5e by Mark Lentczner at 2010-04-04T06:24:14+00:00 all decls now generate Html not HtmlTable - ppDecl return Html, and so now do all of the functions it calls - added some internal tables to some decls, which is wrong, and will have to be fixed - decl "Box" functions became "Elem" functions to make clear they aren't in a table anymore (see Layout.hs) - docBox went away, as only used in one place (and its days are numbered) - cleaned up logic in a number of places, removed dead code - added maybeDocToHtml which simplified a number of places in the code - - - - - dbf73e6e by Mark Lentczner at 2010-04-05T05:02:43+00:00 clean up processExport and place a div around each decl - - - - - e25b7e9f by Mark Lentczner at 2010-04-10T21:23:21+00:00 data decls are now a sequence of paragraphs, not a table - - - - - 89ee0294 by Mark Lentczner at 2010-04-10T21:29:16+00:00 removed commented out code that can't be maintained - - - - - d466f536 by Mark Lentczner at 2010-04-12T04:56:27+00:00 removed declWithDoc and cleaned up data decls in summary - - - - - ed755832 by Mark Lentczner at 2010-04-12T05:07:53+00:00 merge in markupExample changes - - - - - c36f51fd by Mark Lentczner at 2010-04-25T04:56:37+00:00 made record fields be an unordList, not a table - - - - - ed3a28d6 by Mark Lentczner at 2010-04-25T05:23:28+00:00 fixed surround of instance and constructor tables - - - - - 0e35bbc4 by Mark Lentczner at 2010-04-25T05:36:59+00:00 fix class member boxes in summary - - - - - 5041749b by Mark Lentczner at 2010-04-25T05:38:35+00:00 remove unused bodyBox - - - - - e91724db by Mark Lentczner at 2010-04-25T06:26:10+00:00 fixed javascript quoting/escpaing issue - - - - - f4abbb73 by Mark Lentczner at 2010-05-03T23:04:31+00:00 adjust css for current markup - - - - - e75fec4c by Mark Lentczner at 2010-05-04T06:14:34+00:00 added assoicated types and methods back into class decls - - - - - 84169323 by Mark Lentczner at 2010-05-24T13:13:42+00:00 merge in changes from the big-whitespace cleanup - - - - - 3c1c872e by Mark Lentczner at 2010-06-11T21:03:58+00:00 adjust synopsis and bottom bar spacing - - - - - 3c1f9ef7 by Mark Lentczner at 2010-06-11T21:14:44+00:00 fix missing space in "module" lines in synoposis - - - - - 9a137e6d by Mark Lentczner at 2010-06-11T21:34:08+00:00 changed tt elements to code elements - - - - - 50f71ef1 by Mark Lentczner at 2010-06-11T23:27:46+00:00 factored out ppInstances - - - - - 3b9a9de5 by Mark Lentczner at 2010-06-17T17:36:01+00:00 push single constructors (newtype) onto line with decl - - - - - e0f8f2ec by Mark Lentczner at 2010-06-17T22:20:56+00:00 remove <++> connector - - - - - 56c075dd by Mark Lentczner at 2010-07-13T05:26:21+00:00 change to new page structure - - - - - 04be6ca7 by Mark Lentczner at 2010-07-14T04:21:55+00:00 constructors and args as dl lists, built in Layout.hs - - - - - 65aeafc2 by Mark Lentczner at 2010-07-14T05:38:32+00:00 better interface to subDecls - - - - - 72032189 by Mark Lentczner at 2010-07-14T07:04:10+00:00 made subDecl tables looks just so - - - - - b782eca2 by Mark Lentczner at 2010-07-14T16:00:54+00:00 convert args to SubDecl format - - - - - cc75e98f by Mark Lentczner at 2010-07-14T16:28:53+00:00 convert instances to SubDecl - - - - - 34e2aa5a by Mark Lentczner at 2010-07-14T21:07:32+00:00 removing old table cruft from Layout.hs - - - - - d5810d95 by Mark Lentczner at 2010-07-14T21:54:58+00:00 methods and associated types in new layout scheme - - - - - 65ef9579 by Mark Lentczner at 2010-07-14T23:43:42+00:00 clean up synopsis lists - - - - - e523318f by Mark Lentczner at 2010-07-15T05:02:26+00:00 clean up of anchors - - - - - 1215dfc5 by Mark Lentczner at 2010-07-15T23:53:01+00:00 added two new themes and rough css switcher - - - - - 7f0fd36f by Mark Lentczner at 2010-07-16T04:57:38+00:00 fixed package catpion, added style menu - - - - - 0dd4999c by Mark Lentczner at 2010-07-16T20:12:39+00:00 new output for mini_ pages - - - - - 64b2810b by Mark Lentczner at 2010-07-16T20:58:41+00:00 reformat index-frames - - - - - 3173f555 by Mark Lentczner at 2010-07-16T22:41:53+00:00 convert index to new markup - - - - - b0a4b7c9 by Mark Lentczner at 2010-07-17T04:07:22+00:00 convert index.html to new markup, adjust module markup - - - - - 8261ae1e by Mark Lentczner at 2010-07-17T05:07:29+00:00 classing styling of ancillary pages - - - - - 2a4fb025 by Mark Lentczner at 2010-07-17T05:11:45+00:00 clean up Layout.hs: no more vanillaTable - - - - - 87eec685 by Mark Lentczner at 2010-07-17T05:35:16+00:00 clean up Util.hs - - - - - d304e9b0 by Mark Lentczner at 2010-07-17T05:38:50+00:00 qualify import of XHtml as XHtml - - - - - 7dc05807 by Mark Lentczner at 2010-07-17T06:17:53+00:00 factored out head element generation - - - - - 9cdaec9e by Mark Lentczner at 2010-07-17T06:44:54+00:00 refactored out main page body generation - - - - - 8a51019e by Mark Lentczner at 2010-07-17T06:48:20+00:00 moved footer into only place that used it - - - - - efa479da by Mark Lentczner at 2010-07-17T18:48:30+00:00 styling auxillary pages for tibbe and snappy themes - - - - - 81de5509 by Mark Lentczner at 2010-07-18T04:41:38+00:00 fixed alphabet on index page, and styling of it and packages in module lists - - - - - 20718c1a by Mark Lentczner at 2010-07-18T05:34:29+00:00 cleaned up div functions in Layout.hs - - - - - 60d50453 by Mark Lentczner at 2010-07-18T05:48:39+00:00 added content div to main pages - - - - - ed16561c by Mark Lentczner at 2010-07-18T06:12:22+00:00 add .doc class to documentation blocks - - - - - f5c781b0 by Mark Lentczner at 2010-07-19T05:20:53+00:00 refactoring of anchor ID and fragment handling - - - - - a69a93bf by Mark Lentczner at 2010-07-19T05:35:55+00:00 remove an explicit bold tag - replace with .def class - - - - - d76c7225 by Mark Lentczner at 2010-07-19T06:56:15+00:00 rename Haddock.Backends.Xhtml.Util to Utils - - - - - 5a58c0da by David Waern at 2010-07-21T13:30:54+00:00 Remove trailing whitespace in Haddock.Backends.Xhtml - - - - - 0652aa17 by David Waern at 2010-07-21T13:33:21+00:00 Align a few comments - - - - - 785776c3 by David Waern at 2010-07-21T13:39:04+00:00 Remove trailing whitespace in H.B.X.Decl - - - - - 71a30710 by David Waern at 2010-07-21T13:44:27+00:00 Remove more trailing whitespace - - - - - 38750394 by David Waern at 2010-07-21T13:50:43+00:00 Style police - - - - - 3023d940 by David Waern at 2010-07-21T14:01:22+00:00 Style police in H.B.X.Decl - - - - - df16e9e6 by David Waern at 2010-07-21T14:14:45+00:00 Style police in H.B.X.DocMarkup - - - - - 6020e321 by David Waern at 2010-07-21T14:17:32+00:00 More style police - - - - - 86ad8bf5 by David Waern at 2010-07-21T14:21:02+00:00 Style police in H.B.Xhtml - - - - - aea27d03 by David Waern at 2010-07-21T14:42:03+00:00 Fix warnings in LaTeX backend - - - - - 2aff34a9 by David Waern at 2010-07-21T14:50:46+00:00 Style police in LaTeX backend (mainly more newlines) - - - - - e517162d by David Waern at 2010-07-21T15:05:47+00:00 Doc sections in Main - - - - - b971aa0c by David Waern at 2010-07-21T15:06:17+00:00 Trailing whitespace in Documentation.Haddock - - - - - f11628fb by David Waern at 2010-07-21T15:07:06+00:00 Trailing whitespace in Haddock.Convert - - - - - cbaf284c by David Waern at 2010-07-21T15:08:11+00:00 Style police in Haddock.GhcUtils - - - - - 71feb77b by David Waern at 2010-07-21T15:09:06+00:00 Style police in Haddock.InterfaceFile - - - - - 0a9c80e6 by David Waern at 2010-07-21T15:11:33+00:00 Whitespace police - - - - - 6168376c by David Waern at 2010-07-21T15:16:35+00:00 Style police in Haddock.Utils - - - - - 9fe4dd90 by David Waern at 2010-07-21T15:19:31+00:00 Add -fwarn-tabs - - - - - a000d752 by Mark Lentczner at 2010-07-20T17:25:52+00:00 move CSS Theme functions into Themes.hs - - - - - b52b440f by Mark Lentczner at 2010-07-20T17:29:35+00:00 add Thomas Schilling's theme - - - - - e43fa7e8 by Mark Lentczner at 2010-07-21T04:49:34+00:00 correct icon used with Snappy theme - - - - - ba5092d3 by Mark Lentczner at 2010-07-21T04:56:47+00:00 apply Tibbe's updates to his theme - - - - - 7804eef6 by Mark Lentczner at 2010-07-21T05:15:49+00:00 space between "Style" and the downward triangle - - - - - 7131d4c6 by Mark Lentczner at 2010-07-21T17:43:35+00:00 merge with David's source cleanups - - - - - ee65f1cb by David Waern at 2010-07-22T16:50:46+00:00 Fix a bug where we allowed --hoogle, --latex, etc without input files - - - - - e413ff7a by David Waern at 2010-07-22T17:21:58+00:00 Improve function name - - - - - a0fd14f3 by Simon Marlow at 2010-06-30T15:34:32+00:00 fix warnings - - - - - 31f73d2a by David Waern at 2010-07-22T19:29:41+00:00 Solve conflicts - - - - - d563b4a5 by Simon Marlow at 2010-06-30T15:34:37+00:00 fix warning - - - - - 412b6469 by David Waern at 2010-07-22T19:31:28+00:00 Solve conflict - - - - - 35174b94 by Ian Lynagh at 2010-07-06T17:27:16+00:00 Follow mkPState argument order change - - - - - b5c3585c by Simon Marlow at 2010-07-14T08:49:21+00:00 common up code for instance rendering - - - - - d8009560 by Simon Marlow at 2010-07-14T12:37:11+00:00 fix warnings - - - - - a6d88695 by David Waern at 2010-07-24T15:33:33+00:00 Fix build with ghc < 6.13 - - - - - 94cf9de1 by David Waern at 2010-07-24T15:34:37+00:00 Remove conflict left-over - - - - - 313b15c0 by Mark Lentczner at 2010-07-21T22:09:04+00:00 reorganization of nhaddock.css with tibbe - - - - - 9defed80 by Mark Lentczner at 2010-07-21T22:42:14+00:00 further cleanup of nhaddock.css, float TOC, support aux. pages - - - - - 6d944c1b by Mark Lentczner at 2010-07-22T06:22:23+00:00 remove old HTML backend - - - - - b3e8cba5 by Mark Lentczner at 2010-07-22T06:43:32+00:00 remove --html-help support - it was old, out-of-date, and mostly missing - - - - - d2654a08 by Mark Lentczner at 2010-07-22T21:45:34+00:00 tweaks to nhaddock.css - - - - - f73b285c by Mark Lentczner at 2010-07-23T06:19:35+00:00 command like processing for theme selection The bulk of the change is threadnig the selected theme set through functions in Xhtml.hs so that the selected themes can be used when generating the page output. There isn't much going on in most of these changes, just passing it along. The real work is all done in Themes.hs. - - - - - 8bddc90d by Mark Lentczner at 2010-07-23T06:58:31+00:00 drop --themes support, add named theme support decided that --themes was silly - no one would do that, just use multiple --theme arguments made --theme a synonym for --css and -c made those arguments, if no file is found, look up the argument as the name of a built in theme all of this let's haddock be invoked with "--theme=classic" for example. - - - - - 20cafd4f by Mark Lentczner at 2010-07-23T17:44:29+00:00 rename --default-themes to --built-in-themes - - - - - 0fe41307 by Mark Lentczner at 2010-07-23T18:33:02+00:00 tweaks to theme for info table, headings, and tables - - - - - cba4fee0 by Mark Lentczner at 2010-07-23T19:13:59+00:00 tweaks for dl layout, though still not used - - - - - 463fa294 by Mark Lentczner at 2010-07-23T21:07:19+00:00 tweak look of mini pages, keywords, and preblocks - - - - - 5472fc02 by Mark Lentczner at 2010-07-24T05:36:15+00:00 slide out Synopsis drawer - - - - - 9d5d5de5 by Mark Lentczner at 2010-07-24T06:02:42+00:00 extend package header and footer to edges of page - - - - - a47c91a2 by Mark Lentczner at 2010-07-24T06:28:44+00:00 fields are def lists, tweak css for style menu, mini pages, arguments - - - - - ca20f23b by Mark Lentczner at 2010-07-24T16:55:22+00:00 excisting last vestiges of the --xhtml flag - - - - - 71fb012e by Mark Lentczner at 2010-07-25T18:47:49+00:00 change how collapsing sections are done make whole .caption be the target improve javascript for class toggling have plus/minus images come from .css, not img tags - - - - - c168c8d3 by Mark Lentczner at 2010-07-26T00:32:05+00:00 reorganize files in the html lib data dir - - - - - 93324301 by Mark Lentczner at 2010-07-26T01:27:42+00:00 cleaned up Themes.hs - - - - - ad3b5dd4 by Mark Lentczner at 2010-07-26T02:39:15+00:00 make module list use new collapsers - - - - - 1df9bfc6 by Mark Lentczner at 2010-07-27T19:09:25+00:00 remove Tibbe theme - - - - - 8b9b01b3 by Mark Lentczner at 2010-07-27T20:04:03+00:00 move themes into html dir with .theme and .std-theme extensions - - - - - a7beb965 by Mark Lentczner at 2010-07-27T21:06:34+00:00 give a class to empty dd elements so they can be hidden - - - - - a258c117 by Mark Lentczner at 2010-07-27T21:23:58+00:00 remove custom version of copyFile in Xhtml.hs - - - - - b70dba6e by Mark Lentczner at 2010-07-27T22:12:45+00:00 apply margin changes to pre and headings as per group decision, and small cleanups - - - - - e6f722a2 by Mark Lentczner at 2010-07-28T00:03:12+00:00 make info block and package bar links be floatable by placing them first in the dom tree - - - - - c8278867 by Mark Lentczner at 2010-07-28T19:01:18+00:00 styling source links on declarations - - - - - 88fdc399 by Mark Lentczner at 2010-07-29T01:12:46+00:00 styling tweaks don't generate an empty li for absent style menu in links area update css for Classic and Snappy to handle: dl lists links in package header and in declarations floating of links and info block in package and module headers - - - - - 8a75b213 by Ian Lynagh at 2010-07-30T20:21:46+00:00 Fix build in GHC tree - - - - - ce8e18b3 by Simon Hengel at 2010-08-03T18:37:26+00:00 Adapt paths to data files in cabal file - - - - - 9701a455 by Simon Hengel at 2010-08-07T13:20:27+00:00 Add missing dependency to cabal file - - - - - 01b838d1 by Mark Lentczner at 2010-07-30T20:19:40+00:00 improved synopsis drawer: on click, not hover - - - - - 7b6f3e59 by Mark Lentczner at 2010-07-30T23:38:55+00:00 put the synopsis back in the other themes - - - - - 7b2904c9 by Mark Lentczner at 2010-08-11T11:11:26+00:00 close arrows on expanded synopsis drawer - - - - - ea19e177 by Mark Lentczner at 2010-08-12T21:16:45+00:00 width and font changes removed the max width restrictions on the page as a whole and the synopsis made the main font size smaller (nominally 14pt) and then tweaked most font sizes (relative) to be more consistent - - - - - 5ced00c0 by Mark Lentczner at 2010-08-13T15:09:55+00:00 implemented YUI's CSS font approach - - - - - 2799c548 by Mark Lentczner at 2010-08-13T15:11:59+00:00 adjusted margin to 2em, 1 wasn't enough - - - - - 58f06893 by Mark Lentczner at 2010-08-13T15:48:44+00:00 removed underlining on hover for named anchors headings in interface lost thier a element, no need, just put id on heading css for a elements now only applies to those with href attribute - - - - - 7aced4c4 by Mark Lentczner at 2010-08-13T15:50:22+00:00 more space between elements - - - - - 5a3c1cce by Mark Lentczner at 2010-08-13T16:43:43+00:00 adjusted font sizes of auxilary pages per new scheme - - - - - 487539ef by Mark Lentczner at 2010-08-13T21:43:41+00:00 add Frames button and clean up frames.html - - - - - c1a140b6 by Mark Lentczner at 2010-08-13T22:17:48+00:00 move frames button to js - - - - - b0bdb68e by Mark Lentczner at 2010-08-14T03:44:46+00:00 build style menu in javascript moved to javascript, so as to not polute the content with the style menu removed menu building code in Themes.hs removed onclick in Utils.hs changed text of button in header from "Source code" to "Source" more consistent with links in rest of page - - - - - 43ab7120 by Mark Lentczner at 2010-08-16T15:15:37+00:00 font size and margin tweaks - - - - - c0b68652 by Mark Lentczner at 2010-08-17T18:19:52+00:00 clean up collapser logics javascript code for collapasble sections cleaned up rewrote class utilities in javascript to be more robust refactored utilities for generating collapsable sections made toc be same color as synopsis module list has needed clear attribute in CSS - - - - - 5d573427 by Mark Lentczner at 2010-08-17T23:06:02+00:00 don't collapse entries in module list when clicking on links - - - - - 8c307c4a by Mark Lentczner at 2010-08-17T23:21:43+00:00 add missing data file to .cabal - - - - - 414bcfcf by Mark Lentczner at 2010-08-17T23:28:47+00:00 remove synopsis when in frames - - - - - ba0fa98a by Mark Lentczner at 2010-08-18T16:16:11+00:00 layout tweeks - mini page font size, toc color, etc. - - - - - 63c1bed1 by Mark Lentczner at 2010-08-18T19:50:02+00:00 margin fiddling - - - - - c311c094 by Mark Lentczner at 2010-08-20T01:37:55+00:00 better synopsis handling logic - no flashing - - - - - f1fe5fa8 by Mark Lentczner at 2010-08-20T01:41:06+00:00 fix small layout issues mini frames should have same size top heading give info block dts some padding so they don't collide in some browsers - - - - - 0de84d77 by Mark Lentczner at 2010-08-20T02:13:09+00:00 made style changing and cookies storage robust - - - - - 1ef064f9 by Thomas Schilling at 2010-08-04T13:12:22+00:00 Make synopsis frame behave properly in Firefox. In Firefox, pressing the back button first reverted the synopsis frame, and only clicking the back button a second time would update the main frame. - - - - - dd1c9a94 by Mark Lentczner at 2010-08-21T01:46:19+00:00 remove Snappy theme - - - - - 2353a90d by Mark Lentczner at 2010-08-25T05:16:19+00:00 fix occasional v.scroll bars on pre blocks (I think) - - - - - 459b8bf1 by Simon Hengel at 2010-08-08T10:12:45+00:00 Add createInterfaces' (a more high-level alternative to createInterfaces) to Haddock API - - - - - b1b68675 by David Waern at 2010-08-26T20:31:58+00:00 Follow recent API additions with some refactorings Simon Hegel's patch prompted me to do some refactorings in Main, Haddock.Documentation and Haddock.Interface. - - - - - 264d4d67 by David Waern at 2010-08-26T21:40:59+00:00 Get rid of GhcModule and related cruft We can get everything we need directly from TypecheckedModule. - - - - - 0feacec2 by Mark Lentczner at 2010-08-26T23:44:13+00:00 fixed CSS for ordered lists and def lists in doc blocks - - - - - 2997e0c2 by Mark Lentczner at 2010-08-26T23:45:03+00:00 support both kinds of enumerated lists in doc markup The documentation for Haddock says enumerated lists can use either of (1) first item 2. second item The second form wasn't actually supported - - - - - 5d4ddeec by Mark Lentczner at 2010-08-27T21:29:48+00:00 fix broken header link margins - - - - - 614456ba by Mark Lentczner at 2010-08-27T22:16:19+00:00 fix table of contents CSS - - - - - 03f329a2 by David Waern at 2010-08-28T16:36:09+00:00 Update tests following switch to the Xhtml backend - - - - - ca689fa2 by Mark Lentczner at 2010-08-28T18:25:16+00:00 fix def lists - - - - - 18e1d3d2 by Mark Lentczner at 2010-08-28T18:26:18+00:00 push footer to bottom of window - - - - - b0ab8d82 by David Waern at 2010-08-28T22:04:32+00:00 Whitespace police - - - - - 2d217977 by David Waern at 2010-08-29T12:44:45+00:00 Remove Snappy data files - - - - - 01e27d5f by David Waern at 2010-08-29T13:03:28+00:00 Add source entity path to --read-interface You can now use this flag like this: --read-interface=<html path>,<source entity path>,<.haddock file> By "source entity path" I mean the same thing that is specified with the --source-entity flag. The purpose of this is to be able to specify the source entity path per package, to allow source links to work in the presence of cross-package documentation. When given two arguments or less the --read-interface flag behaves as before. - - - - - 20bf4aaa by David Waern at 2010-08-29T13:11:03+00:00 Naming wibbles - - - - - ad22463f by Mark Lentczner at 2010-08-29T15:14:54+00:00 make portability block be a table - solves layout issues - - - - - 97bd1ae6 by Mark Lentczner at 2010-08-29T15:17:42+00:00 update golden test for Test due to portability box change - - - - - d37e139e by Mark Lentczner at 2010-08-29T17:07:17+00:00 move TOC and Info blocks down 0.5em to improve layout issue w/Test.hs - - - - - acf52501 by David Waern at 2010-08-29T17:32:36+00:00 Allow building with ghc < 6.16 - - - - - 1cb34ed8 by Ian Lynagh at 2010-07-24T23:18:49+00:00 Flatten the dynflags before parsing - - - - - b36845b4 by Ian Lynagh at 2010-07-24T23:26:49+00:00 Follow flattenLanguageFlags -> flattenExtensionFlags rename - - - - - 7f7fcc7e by David Waern at 2010-08-29T17:46:23+00:00 Use flattenExtensionFlags with ghc >= 6.13 only - - - - - 13cf9411 by Ian Lynagh at 2010-08-01T18:09:54+00:00 Make the main haddock script versioned, and make plain "haddock" a symlink - - - - - 495cbff2 by Ian Lynagh at 2010-08-18T18:57:24+00:00 Fix installation in the GHC build system Data-files are now in subdirectories, so we need to handle that - - - - - 88ebab0a by Ian Lynagh at 2010-08-18T19:43:53+00:00 GHC build system: Add all the data files to BINDIST_EXTRAS - - - - - 65837172 by David Waern at 2010-08-29T20:12:34+00:00 Update Test - - - - - 094bbaa2 by David Waern at 2010-08-29T20:55:14+00:00 Revert update to Test - - - - - a881cfb3 by David Waern at 2010-08-31T18:24:15+00:00 Bump version number - - - - - 1fc8a3eb by David Waern at 2010-08-31T22:32:27+00:00 Update ANNOUNCE - - - - - ee1df9d0 by David Waern at 2010-08-31T22:33:11+00:00 Update CHANGES - - - - - 394cc854 by David Waern at 2010-08-31T22:33:23+00:00 Update interface file versioning to work with ghc 6.14/15 - - - - - 7d03b79b by David Waern at 2010-08-31T22:36:00+00:00 Update test output following version change - - - - - a48d82d1 by Mark Lentczner at 2010-09-01T04:29:35+00:00 sort options in doc to match --help output removed --html-help option, as it is no longer supported - - - - - 06561aeb by Mark Lentczner at 2010-09-01T05:29:32+00:00 update options documentation rewrote doc for --html added doc for --theme and --built-in-themes added --use-contents and --gen-contents - - - - - 57dea832 by Mark Lentczner at 2010-09-01T05:31:27+00:00 slight wording change about Frames mode - - - - - fa1f6da3 by David Waern at 2010-09-01T10:57:44+00:00 Update doc configure script to find docbook stylesheets on arch linux - - - - - addff770 by David Waern at 2010-09-01T11:02:29+00:00 Wibble - - - - - 8399006d by David Waern at 2010-09-01T11:19:21+00:00 Replace ghci> with >>> in example syntax - - - - - 35074cf8 by David Waern at 2010-09-01T19:03:27+00:00 Improve docs for --no-tmp-comp-dir - - - - - 0f8f8cfd by David Waern at 2010-09-02T11:22:27+00:00 Add a list of contributors to the user guide Break out everyone thanked in the `Acknowledgements` chapter into a separate contributor list and add everyone from `darcs show authors`. We consider everyone who is thanked to be a contributor as a conservative estimation :-) I have added some more contributors that I know about, who were not in the darcs history, but others may be missing. So please add anyone that you think is missing from the list. - - - - - 42ccf099 by David Waern at 2010-09-02T11:29:22+00:00 Update copyright years in license - - - - - 0d560479 by David Waern at 2010-09-02T11:38:52+00:00 Update release instructions - - - - - 72ab7796 by David Waern at 2010-09-02T19:27:08+00:00 Add a note to ANNOUNCE - - - - - bf9d9c5d by David Waern at 2010-09-02T19:27:48+00:00 H.Utils needs FFI on Win+MinGW - - - - - 048ae44a by Mark Lentczner at 2010-09-04T23:19:47+00:00 make TOC group header identifiers validate - - - - - 8c6faf36 by Simon Michael at 2010-09-22T07:12:34+00:00 add hints for cleaner darcs show authors output - - - - - 9909bd17 by Simon Michael at 2010-09-22T17:58:06+00:00 print haddock coverage info on stdout when generating docs A module's haddockable items are its exports and the module itself. The output is lightly formatted so you can align the :'s and sort for readability. - - - - - 6da72171 by David Waern at 2010-10-03T21:31:24+00:00 Style wibble - - - - - 2f8d8e4d by Tobias Brandt at 2010-08-27T07:01:21+00:00 adding the option to fully qualify identifiers - - - - - 833be6c6 by Tobias Brandt at 2010-08-27T15:50:28+00:00 adding support for local and relative name qualification - - - - - df15c4e9 by Tobias Brandt at 2010-08-27T15:56:37+00:00 corrected qualification help message - - - - - 449e9ce1 by David Waern at 2010-10-16T17:34:30+00:00 Solve conflicts - - - - - 3469bda5 by David Waern at 2010-10-16T18:42:40+00:00 Use "qual" as an abbreviation for qualification instead of "quali" for consistency - - - - - 97c2d728 by David Waern at 2010-10-16T18:47:07+00:00 Style police - - - - - ce14fbea by David Waern at 2010-10-16T21:15:25+00:00 Style police - - - - - fdf29e9d by David Waern at 2010-10-17T00:30:44+00:00 Add a pointer to the style guide - - - - - 8e6b44e8 by rrnewton at 2010-10-24T03:19:28+00:00 Change to index pages: include an 'All' option even when subdividing A-Z. - - - - - 755b131c by David Waern at 2010-11-14T19:39:36+00:00 Bump version - - - - - d0345a04 by David Waern at 2010-11-14T19:41:59+00:00 TAG 2.8.1 - - - - - f6221508 by Simon Peyton Jones at 2010-09-13T09:53:00+00:00 Adapt to minor changes in internal GHC functions - - - - - 1290713d by Ian Lynagh at 2010-09-15T10:37:18+00:00 Remove duplicate Outputable instance for Data.Map.Map - - - - - 87f69eef by Ian Lynagh at 2010-09-21T15:01:10+00:00 Bump GHC dep upper bound - - - - - af36e087 by Ian Lynagh at 2010-09-21T15:12:02+00:00 Fix up __GLASGOW_HASKELL__ tests - - - - - ad67716c by Ian Lynagh at 2010-09-21T20:31:35+00:00 Don't build haddock is HADDOCK_DOCS is NO - - - - - 63b3f1f5 by Ian Lynagh at 2010-09-21T21:39:51+00:00 Fixes for when HADDOCK_DOCS=NO - - - - - e92bfa42 by Ian Lynagh at 2010-09-29T21:15:38+00:00 Fix URL creation on Windows: Use / not \ in URLs. Fixes haskell/haddock#4353 - - - - - 66c55e05 by Ian Lynagh at 2010-09-30T17:03:34+00:00 Tidy up haddock symlink installation In particular, it now doesn't get created if we aren't installing haddock. - - - - - 549b5556 by Ian Lynagh at 2010-10-23T21:17:14+00:00 Follow extension-flattening change in GHC - - - - - d7c2f72b by David Waern at 2010-11-14T20:17:55+00:00 Bump version to 2.8.2 - - - - - 6989a3a9 by David Waern at 2010-11-14T20:26:01+00:00 Solve conflict - - - - - 055c6910 by Ian Lynagh at 2010-09-22T15:36:20+00:00 Bump GHC dep - - - - - c96c0763 by Simon Marlow at 2010-10-27T11:09:44+00:00 follow changes in the GHC API - - - - - 45907129 by David Waern at 2010-11-07T14:00:58+00:00 Update the HCAR entry - - - - - 61940b95 by David Waern at 2010-11-07T14:07:34+00:00 Make the HCAR entry smaller - - - - - aa590b7d by David Waern at 2010-11-14T21:30:59+00:00 Update HCAR entry with November 2010 version - - - - - 587f9847 by David Waern at 2010-11-14T23:48:17+00:00 Require ghc >= 7.0 - - - - - ff5c647c by David Waern at 2010-11-14T23:49:09+00:00 TAG 2.8.2 - - - - - 937fcb4f by David Waern at 2010-11-14T23:49:45+00:00 Solve conflict - - - - - 8e5d0c1a by David Waern at 2010-11-15T21:09:50+00:00 Remove code for ghc < 7 - - - - - 3d47b70a by David Waern at 2010-11-15T21:11:06+00:00 Fix bad merge - - - - - 7f4a0d8a by David Waern at 2010-11-15T21:13:57+00:00 Remove more ghc < 7 code - - - - - 9ee34b50 by David Waern at 2010-11-15T21:31:25+00:00 Match all AsyncExceptions in exception handler - - - - - 42849c70 by David Waern at 2010-11-15T21:35:31+00:00 Just say "internal error" instead of "internal Haddock or GHC error" - - - - - c88c809b by David Waern at 2010-11-15T21:44:19+00:00 Remove docNameOcc under the motto "don't name compositions" - - - - - b798fc7c by David Waern at 2010-11-15T23:27:13+00:00 Wibble - - - - - 2228197e by David Waern at 2010-11-15T23:28:24+00:00 Rename the HCAR entry file - - - - - 8a3f9090 by David Waern at 2010-11-16T00:05:29+00:00 Remove Haskell 2010 extensions from .cabal file - - - - - c7a0c597 by David Waern at 2010-11-16T00:10:28+00:00 Style wibbles - - - - - cde707a5 by David Waern at 2010-11-16T00:12:00+00:00 Remove LANGUAGE ForeignFunctionInterface pragmas - - - - - 1dbda8ed by David Waern at 2010-11-16T00:17:21+00:00 Make a little more use of DoAndIfThenElse - - - - - 4c45ff6e by David Waern at 2010-11-16T00:59:41+00:00 hlint police - - - - - d2feaf09 by David Waern at 2010-11-16T01:14:15+00:00 hlint police - - - - - 99876e97 by David Waern at 2010-11-20T19:06:00+00:00 Haddock documentation updates - - - - - 65ce6987 by David Waern at 2010-11-20T19:42:51+00:00 Follow the style guide closer in Haddock.Types and improve docs - - - - - 28ca304a by tob.brandt at 2010-11-20T17:04:40+00:00 add full qualification for undocumented names - - - - - d61341e3 by David Waern at 2010-11-20T20:04:15+00:00 Re-structure qualification code a little - - - - - 0057e4d6 by David Waern at 2010-11-20T20:07:55+00:00 Re-order functions - - - - - d7279afd by David Waern at 2010-11-21T03:39:54+00:00 Add BangPatterns to alex and happy source files - - - - - 629fe60e by tob.brandt at 2010-11-23T23:35:11+00:00 documentation for qualification - - - - - 37031cee by David Waern at 2010-11-23T21:06:44+00:00 Update CHANGES - don't mention 2.8.2, we won't release it - - - - - f2489e19 by David Waern at 2010-12-01T21:57:11+00:00 Update deps of runtests.hs to work with ghc 7.0.1 - - - - - d3657e9a by David Waern at 2010-12-01T22:04:57+00:00 Make tests compile with ghc 7.0.1 - - - - - a2f09d9b by David Waern at 2010-12-01T22:06:59+00:00 Update tests following version bump - - - - - 50883ebb by David Waern at 2010-12-06T14:09:18+00:00 Update tests following recent changes - - - - - fc2fadeb by David Waern at 2010-12-06T14:17:29+00:00 Add a flag --pretty-html for rendering indented html with newlines - - - - - 30832ef2 by David Waern at 2010-12-06T14:17:35+00:00 Use --pretty-html when running the test suite. Makes it easier to compare output - - - - - a0b81b31 by David Waern at 2010-12-06T14:18:27+00:00 Wibble - - - - - 3aaa23fe by David Waern at 2010-12-06T14:19:29+00:00 Haddockify ppHtml comments - - - - - 24bb24f0 by David Waern at 2010-12-06T14:23:15+00:00 Remove --debug. It was't used, and --verbosity should take its place - - - - - 6bc076e5 by David Waern at 2010-12-06T14:25:37+00:00 Rename golden-tests into html-tests. "golden tests" sounds strange - - - - - 53301e55 by David Waern at 2010-12-06T14:26:26+00:00 QUALI -> QUAL in the description --qual for consistency - - - - - 98b6affb by David Waern at 2010-12-06T21:54:02+00:00 Bump version - - - - - 371bf1b3 by David Waern at 2010-12-06T22:08:55+00:00 Update tests following version bump - - - - - 25be762d by David Waern at 2010-12-06T22:21:03+00:00 Update CHANGES - - - - - 7c7dac71 by David Waern at 2010-12-06T22:33:43+00:00 Update ANNOUNCE - - - - - 30d7a5f2 by Simon Peyton Jones at 2010-11-15T08:38:38+00:00 Alex generates BangPatterns, so make Lex.x accept them (It'd be better for Alex to generate this pragma.) - - - - - 605e8018 by Simon Marlow at 2010-11-17T11:37:24+00:00 Add {-# LANGUAGE BangPatterns #-} to mollify GHC - - - - - a46607ba by David Waern at 2010-12-07T14:08:10+00:00 Solve conflicts - - - - - b28cda66 by David Waern at 2010-12-09T20:41:35+00:00 Docs: Mention that \ is a special character in markup - - - - - a435bfdd by Ian Lynagh at 2010-11-17T14:01:19+00:00 TAG GHC 7.0.1 release - - - - - 5a15a05a by David Waern at 2010-12-11T17:51:19+00:00 Fix indentation problem - - - - - 4232289a by Lennart Kolmodin at 2010-12-17T18:32:03+00:00 Revise haddock.cabal given that we now require ghc-7 default-language should be Haskell2010, slight new semantics for extensions. Rewrite into clearer dependencies of base and Cabal. - - - - - a36302dc by David Waern at 2010-12-19T17:12:37+00:00 Update CHANGES - - - - - 7c8b85b3 by David Waern at 2010-12-19T17:14:24+00:00 Bump version - - - - - cff22813 by Ian Lynagh at 2011-01-05T18:24:27+00:00 Write hoogle output in utf8; fixes GHC build on Windows - - - - - c7e762ea by David Waern at 2011-01-22T00:00:35+00:00 Put title outside doc div when HTML:fying title+prologue Avoids indenting the title, and makes more sense since the title is not a doc string anyway. - - - - - 5f639054 by David Waern at 2011-01-22T16:09:44+00:00 Fix spelling error - contributed by Marco Silva - - - - - c11dce78 by Ian Lynagh at 2011-01-07T02:33:11+00:00 Follow GHC build system changes - - - - - 101cfaf5 by David Waern at 2011-01-08T14:06:44+00:00 Bump version - - - - - af62348b by David Waern at 2011-01-08T14:07:07+00:00 TAG 2.9.2 - - - - - 4d1f6461 by Ian Lynagh at 2011-01-07T23:06:57+00:00 Name the haddock script haddock-ghc-7.0.2 instead of haddock-7.0.2; haskell/haddock#4882 "7.0.2" looked like a haddock version number before - - - - - 8ee4d5d3 by Simon Peyton Jones at 2011-01-10T17:31:12+00:00 Update Haddock to reflect change in hs_tyclds field of HsGroup - - - - - 06f3e3db by Ian Lynagh at 2011-03-03T15:02:37+00:00 TAG GHC 7.0.2 release - - - - - 7de0667d by David Waern at 2011-03-10T22:47:13+00:00 Update CHANGES - - - - - 33a9f1c8 by David Waern at 2011-03-10T22:47:31+00:00 Fix build with ghc 7.0.1 - - - - - 4616f861 by David Waern at 2011-03-10T22:47:50+00:00 TAG 2.9.2-actual - - - - - 0dab5e3c by Simon Hengel at 2011-04-08T15:53:01+00:00 Set shell script for unit tests back to work - - - - - 85c54dee by Simon Hengel at 2011-04-08T16:01:24+00:00 Set unit tests back to work Here "ghci>" was still used instead of ">>>". - - - - - 1cea9b78 by Simon Hengel at 2011-04-08T16:25:36+00:00 Update runtests.hs for GHC 7.0.2 - - - - - 8e5b3bbb by Simon Hengel at 2011-04-08T16:28:49+00:00 Update Haddock version in *.html.ref - - - - - 2545e955 by Simon Hengel at 2011-04-08T17:09:28+00:00 Add support for blank lines in the result of examples Result lines that only contain the string "<BLANKLINE>" are treated as a blank line. - - - - - adf64d2e by Simon Hengel at 2011-04-08T17:36:50+00:00 Add documentation for "support for blank lines in the result of examples" - - - - - c51352ca by David Waern at 2011-05-21T23:57:56+00:00 Improve a haddock comment - - - - - 7419cf2c by David Waern at 2011-05-22T15:41:52+00:00 Use cabal's test suite support to run the test suite This gives up proper dependency tracking of the test script. - - - - - 7770070c by David Waern at 2011-05-22T01:45:44+00:00 We don't need to send DocOptions nor a flag to mkExportItems - - - - - 9d95b7b6 by David Waern at 2011-05-22T21:39:03+00:00 Fix a bug - - - - - 1f93699b by David Waern at 2011-05-22T21:40:21+00:00 Break out fullContentsOf, give it a better name and some documentation The documentation describes how we want this function to eventually behave, once we have fixed a few problems with the current implementation. - - - - - 9a86432f by David Waern at 2011-05-22T21:53:52+00:00 Fix some stylistic issues in mkExportItems - - - - - c271ff0c by David Waern at 2011-05-22T22:09:11+00:00 Indentation - - - - - 93e602b1 by David Waern at 2011-06-10T01:35:31+00:00 Add git commits since switchover: darcs format (followed by a conflict resolution): commit 6f92cdd12d1354dfbd80f8323ca333bea700896a Merge: f420cc4 28df3a1 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Thu May 19 17:54:34 2011 +0100 Merge remote branch 'origin/master' into ghc-generics commit 28df3a119f770fdfe85c687dd73d5f6712b8e7d0 Author: Max Bolingbroke <batterseapower at hotmail.com> Date: Sat May 14 22:37:02 2011 +0100 Unicode fix for getExecDir on Windows commit 89813e729be8bce26765b95419a171a7826f6d70 Merge: 6df3a04 797ab27 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 9 11:55:17 2011 +0100 Merge branch 'ghc-new-co' commit 6df3a040da3dbddee67c6e30a892f87e6b164383 Author: Ian Lynagh <igloo at earth.li> Date: Sun May 8 17:05:50 2011 +0100 Follow changes in SDoc commit f420cc48b9259f0b1afd2438b12f9a2bde57053d Author: Jose Pedro Magalhaes <jpm at cs.uu.nl> Date: Wed May 4 17:31:52 2011 +0200 Adapt haddock to the removal of HsNumTy and TypePat. commit 797ab27bdccf39c73ccad374fea265f124cb52ea Merge: 1d81436 5a91450 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:05:03 2011 +0100 Merge remote branch 'origin/master' into ghc-new-co commit 1d8143659a81cf9611668348e33fd0775c7ab1d2 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:03:46 2011 +0100 Wibbles for ghc-new-co branch commit 5a91450e2ea5a93c70bd3904b022445c9cc82488 Author: Ian Lynagh <igloo at earth.li> Date: Fri Apr 22 00:51:56 2011 +0100 Follow defaultDynFlags change in GHC - - - - - 498da5ae by David Waern at 2011-06-11T00:33:33+00:00 * Merge in git patch from Michal Terepeta >From 6fc71d067738ef4b7de159327bb6dc3d0596be29 Mon Sep 17 00:00:00 2001 From: Michal Terepeta <michal.terepeta at gmail.com> Date: Sat, 14 May 2011 19:18:22 +0200 Subject: [PATCH] Follow the change of TypeSig in GHC. This follows the change in GHC to make TypeSig take a list of names (instead of just one); GHC ticket haskell/haddock#1595. This should also improve the Haddock output in case the user writes a type signature that refers to many names: -- | Some comment.. foo, bar :: ... will now generate the expected output with one signature for both names. - - - - - 094607fe by Ian Lynagh at 2011-06-17T19:10:29+01:00 Fix build - - - - - 8fa35740 by Ian Lynagh at 2011-06-26T21:06:40+01:00 Bump GHC dep to allow 7.2 - - - - - e4d2ca3c by Ian Lynagh at 2011-07-07T23:06:28+01:00 Relax base dep - - - - - b948fde9 by Ian Lynagh at 2011-07-28T16:39:45+01:00 GHC build system: Don't install the datafiles twice - - - - - f82f6d70 by Simon Marlow at 2011-08-11T12:08:15+01:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - b341cc12 by Max Bolingbroke at 2011-08-22T20:25:27+01:00 Fix compilation with no-pred-ty GHC - - - - - 30494581 by Max Bolingbroke at 2011-08-23T10:20:54+01:00 Remaining fixes for PredTy removal - - - - - 0b197138 by Max Bolingbroke at 2011-08-26T08:27:45+01:00 Rename factKind to constraintKind - - - - - a379bec5 by Max Bolingbroke at 2011-09-04T12:54:47+01:00 Deal with change to IParam handling in GHC - - - - - f94e421b by Max Bolingbroke at 2011-09-06T17:34:31+01:00 Adapt Haddock for the ConstraintKind extension changes - - - - - 8821e5cc by Max Bolingbroke at 2011-09-09T08:24:59+01:00 Ignore associated type defaults (just as we ignore default methods) - - - - - 31a0afd4 by Max Bolingbroke at 2011-09-09T09:06:00+01:00 Merge branch 'no-pred-ty' of ssh://darcs.haskell.org/srv/darcs/haddock into no-pred-ty - - - - - dd3b530a by Max Bolingbroke at 2011-09-09T14:10:25+01:00 Merge branch 'no-pred-ty' Conflicts: src/Haddock/Convert.hs - - - - - 5f25ec96 by Max Bolingbroke at 2011-09-09T14:10:40+01:00 Replace FactTuple with ConstraintTuple - - - - - cd30b9cc by David Waern at 2011-09-26T02:17:55+02:00 Bump to version 2.9.3 - - - - - 4fbfd397 by Max Bolingbroke at 2011-09-27T14:55:21+01:00 Follow changes to BinIface Name serialization - - - - - 92257d90 by David Waern at 2011-09-30T23:45:07+02:00 Fix problem with test files not added to distribution tarball - - - - - 00255bda by David Waern at 2011-09-30T23:48:24+02:00 Merge branch 'development' - - - - - 5421264f by David Waern at 2011-10-01T01:25:39+02:00 Merge in darcs patch from Simon Meier: Wed Jun 1 19:41:16 CEST 2011 iridcode at gmail.com * prettier haddock coverage info The new coverage info rendering uses less horizontal space. This reduces the number of unnecessary line-wrappings. Moreover, the most important information, how much has been documented already, is now put up front. Hopefully, this makes it more likely that a library author is bothered by the low coverage of his modules and fixes that issue ;-) - - - - - 07d318ef by David Waern at 2011-10-01T01:34:10+02:00 Use printException instead of deprecated printExceptionAndWarnings - - - - - 40d52ee4 by David Waern at 2011-10-01T01:41:13+02:00 Merge in darcs pach: Mon Apr 11 18:09:54 JST 2011 Liyang HU <haddock at liyang.hu> * Remember collapsed sections in index.html / haddock-util.js - - - - - 279d6dd4 by David Waern at 2011-10-01T01:55:45+02:00 Merge in darcs patch: Joachim Breitner <mail at joachim-breitner.de>**20110619201645 Ignore-this: f6c51228205b0902ad5bfad5040b989a As reported on http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=578301, generating the global index takes much too long if type-level (with lots of auto-generated types) is installed. The patch avoids a quadratic runtime in the subfunction getIfaceIndex of ppHtmlIndex by using a temporary set. Runtime improvement observed here from 25.36s to 2.86s. - - - - - d1612383 by David Waern at 2011-10-01T01:56:48+02:00 Merge branch 'development' - - - - - 347520c1 by David Waern at 2011-10-01T01:56:54+02:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9a0c95e8 by David Waern at 2011-10-01T02:19:10+02:00 Improve .cabal file - - - - - 6967dc64 by Ian Lynagh at 2011-10-01T01:34:06+01:00 Follow changes to ForeignImport/ForeignExport in GHC - - - - - 565cb26b by Simon Marlow at 2011-10-04T00:15:04+02:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - 8b74f512 by David Waern at 2011-10-04T00:18:17+02:00 Requre ghc >= 7.2 - - - - - 271d360c by David Waern at 2011-10-04T00:22:50+02:00 Bump version to 2.9.4 - - - - - 37f3edb0 by David Waern at 2011-10-06T02:30:21+02:00 Add alex and happy to build-tools. - - - - - 7ac2bb6e by David Terei at 2011-10-12T14:02:55-07:00 Add safe haskell indication to haddock output - - - - - 42c91a47 by David Terei at 2011-10-12T14:06:03-07:00 Fix CSS issue with info table not being contained in module header - - - - - 0eddab6c by David Terei at 2011-10-12T14:06:58-07:00 Add safe haskell indication to haddock output - - - - - 3df058eb by David Terei at 2011-10-12T14:07:07-07:00 Fix CSS issue with info table not being contained in module header - - - - - a40a6c3f by David Waern at 2011-10-22T11:29:06+02:00 Bump .haddock file version since the format has changed recently - - - - - 8a6254be by David Waern at 2011-10-22T11:30:42+02:00 Merge branch 'development' - - - - - 642e3e02 by David Waern at 2011-10-23T21:23:39+02:00 Sort import list - - - - - 36371cf8 by David Waern at 2011-10-23T22:48:18+02:00 Remove NEW_GHC_LAYOUT conditional. - - - - - 5604b499 by David Waern at 2011-10-27T00:15:03+02:00 Add --print-ghc-path. - - - - - 463499fa by David Waern at 2011-10-27T00:16:22+02:00 Make testsuite able to find its dependencies automatically. - - - - - a3506172 by Ryan Newton at 2011-11-05T05:59:58-04:00 Improved declNames internal error. Added a case to handle DocD. - - - - - 001b8baf by David Waern at 2011-11-05T20:37:29+01:00 Rename copy.hs -> accept.hs. - - - - - 55d808d3 by David Waern at 2011-11-05T23:30:02+01:00 Fix build. - - - - - deb5c3be by David Waern at 2011-11-06T00:01:47+01:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9b663554 by David Waern at 2011-11-06T00:03:45+01:00 Merge https://github.com/rrnewton/haddock - - - - - 1abb0ff6 by David Waern at 2011-11-06T01:20:37+01:00 Use getDeclMainBinder instead of declNames. - - - - - 4b005c01 by David Waern at 2011-11-06T19:09:53+01:00 Fix build. - - - - - c2c51bc7 by Ian Lynagh at 2011-11-06T23:01:33+00:00 Remove -DNEW_GHC_LAYOUT in ghc.mk - - - - - f847d703 by Jose Pedro Magalhaes at 2011-11-11T09:07:39+00:00 New kind-polymorphic core This big patch implements a kind-polymorphic core for GHC. The current implementation focuses on making sure that all kind-monomorphic programs still work in the new core; it is not yet guaranteed that kind-polymorphic programs (using the new -XPolyKinds flag) will work. For more information, see http://haskell.org/haskellwiki/GHC/Kinds - - - - - 7d7c3b09 by Jose Pedro Magalhaes at 2011-11-16T21:42:22+01:00 Follow changes to tuple sorts in master - - - - - 8430e03e by Simon Peyton Jones at 2011-11-17T10:20:27+00:00 Remove redundant imports - - - - - d1b06832 by Ian Lynagh at 2011-11-19T01:33:21+00:00 Follow GHC build system change to the way we call rm - - - - - 9e2230ed by David Waern at 2011-11-24T15:00:24+01:00 Fix a bug in test runner and get rid of regex-compat dependency. - - - - - 52039b21 by David Waern at 2011-11-24T23:55:36+01:00 Avoid haskell98 dependency in test - - - - - 92e1220d by David Waern at 2011-11-25T00:03:33+01:00 Avoid depency on regex-compat also in accept.hs. - - - - - ddac6b6f by David Waern at 2011-11-25T02:13:38+01:00 Accept test output. - - - - - 5a720455 by David Waern at 2011-11-25T02:16:20+01:00 Some more changes to test scripts. - - - - - 170a9004 by David Waern at 2011-11-25T02:30:41+01:00 Add flag --interface-version. - - - - - d225576c by David Waern at 2011-11-25T02:39:26+01:00 Remove #ifs for older compiler versions. - - - - - f0d0a4f5 by David Waern at 2011-11-26T04:20:12+01:00 Give preference to type over data constructors for doc comment links at renaming time. Previously this was done in the backends. Also, warn when a doc comment refers to something that is in scope but which we don't have the .haddock file for. These changes mean we can make DocIdentifier [a] into DocIdentifier a. - - - - - eef0e776 by David Waern at 2011-11-26T17:01:06+01:00 Allow doc comments to link to out-of-scope things (#78). (A bug that should have been fixed long ago.) - - - - - 565ad529 by David Waern at 2011-11-26T19:56:21+01:00 Update tests. - - - - - fb3ce7b9 by David Waern at 2011-11-26T21:44:28+01:00 Cleanup. - - - - - d0328126 by David Waern at 2011-11-26T22:10:28+01:00 Fix module reference bug. - - - - - c03765f8 by David Waern at 2011-12-03T05:20:20+01:00 Slightly better behaviour on top-levels without type signatures. - Docs don't get attached to the next top-level with signature by mistake. - If there's an export list and the top-level is part of it, its doc comment shows up in the documentation. - - - - - 48461d31 by David Waern at 2011-12-03T05:38:10+01:00 Add a test for Unicode doc comments. - - - - - 549c4b4e by David Waern at 2011-12-03T19:07:55+01:00 Cleanup. - - - - - 7bfecf91 by David Waern at 2011-12-03T20:13:08+01:00 More cleanup. - - - - - 14fab722 by Ian Lynagh at 2011-12-12T21:21:35+00:00 Update dependencies and binaryInterfaceVersion - - - - - 469e6568 by Ian Lynagh at 2011-12-18T12:56:16+00:00 Fix (untested) building from source tarball without alex/happy haddock's .cabal file was declaring that it needed alex and happy to build, but in the GHC source tarballs it doesn't. - - - - - 895c9a8c by David Waern at 2011-12-27T12:57:43+01:00 Go back to having a doc, sub and decl map instead of one big decl map. This setup makes more sense since when we add value bindings to the processed declarations (for type inference), we will have multiple declarations which should share documentation. Also, we already have a separate doc map for instances which we can now merge into the main doc map. Another benefit is that we don't need the DeclInfo type any longer. - - - - - 736767d9 by David Waern at 2011-12-27T13:33:41+01:00 Merge ../../../haddock Conflicts: src/Haddock/InterfaceFile.hs - - - - - 20016f79 by David Waern at 2011-12-27T13:57:23+01:00 Bump version. - - - - - 31f276fb by David Waern at 2011-12-27T13:57:32+01:00 Merge ../ghc/utils/haddock - - - - - 95b367cd by David Waern at 2011-12-27T14:57:29+01:00 Update tests following version bump. - - - - - fa3c94cd by David Waern at 2011-12-27T14:57:51+01:00 Get rid of quite unnecessary use of different lists. - - - - - 9c4d3c54 by David Waern at 2011-12-27T15:26:42+01:00 Cleanup. - - - - - 2caf9f90 by David Waern at 2011-12-27T16:18:05+01:00 Wibbles. - - - - - 3757d09b by David Waern at 2011-12-27T20:50:26+01:00 Complete support for inferring types for top-level bindings. - - - - - 53418734 by David Waern at 2011-12-28T15:02:13+01:00 Minor fixes and cleanup. - - - - - 0c9d0385 by Ian Lynagh at 2012-01-03T18:31:29+00:00 Follow rename of Instance to ClsInst in GHC - - - - - c9bc969a by Simon Hengel at 2012-01-12T21:28:14+01:00 Make sure that generated xhtml is valid (close haskell/haddock#186) Thanks to Phyx. - - - - - 836a0b9a by David Waern at 2012-02-01T02:30:05+01:00 Fix bug introduced in my recent refactoring. - - - - - c7d733eb by David Waern at 2012-02-01T02:30:26+01:00 Cleanup mkMaps and avoid quadratic behaviour. - - - - - da3cda8f by David Waern at 2012-02-01T02:56:56+01:00 Require ghc >= 7.4. - - - - - 83a3287e by David Waern at 2012-02-01T02:57:36+01:00 Update CHANGES. - - - - - 93408f0b by Simon Hengel at 2012-02-04T00:48:04+01:00 Add reference renderings - - - - - 49d00d2c by Simon Hengel at 2012-02-04T00:48:25+01:00 Set unit tests for parser back to work - - - - - eb450980 by Simon Hengel at 2012-02-04T00:49:07+01:00 Add .gitignore - - - - - a841602c by Simon Hengel at 2012-02-04T00:49:16+01:00 Add .ghci file - - - - - 8861199d by Simon Hengel at 2012-02-04T00:49:29+01:00 tests/html-tests/copy.hs: Use mapM_ instead of mapM So we do net get a list of () on stdout when running with runhaskell. - - - - - b477d9b5 by Simon Hengel at 2012-02-04T00:49:46+01:00 Remove index files from golden tests - - - - - 9dbda34e by Simon Hengel at 2012-02-04T00:49:57+01:00 Add /tests/html-tests/tests/*index*.ref to .gitignore - - - - - a9434817 by Simon Hengel at 2012-02-04T00:50:04+01:00 Add DocWarning to Doc The Xhtml backend has special markup for that, Hoogle and LaTeX reuse what we have for DocEmphasis. - - - - - de2fb6fa by Simon Hengel at 2012-02-04T00:50:13+01:00 Add support for module warnings - - - - - 0640920e by Simon Hengel at 2012-02-04T00:50:21+01:00 Add tests for module warnings - - - - - 30ce0d77 by Simon Hengel at 2012-02-04T00:50:29+01:00 Add support for warnings - - - - - bb367960 by Simon Hengel at 2012-02-04T00:50:37+01:00 Add tests for warnings - - - - - 6af1dc2d by Simon Hengel at 2012-02-04T00:50:50+01:00 Expand type signatures in export list (fixes haskell/haddock#192) - - - - - a06cbf25 by Simon Hengel at 2012-02-04T00:51:04+01:00 Expand type signatures for modules without explicit export list - - - - - 57dda796 by Simon Hengel at 2012-02-04T00:51:15+01:00 Remove obsolete TODO - - - - - 270c3253 by David Waern at 2012-02-04T00:51:24+01:00 Fix issues in support for warnings. * Match against local names only. * Simplify (it's OK to map over the warnings). - - - - - 683634bd by David Waern at 2012-02-04T00:55:11+01:00 Some cleanup and make sure we filter warnings through exports. - - - - - 210cb4ca by David Waern at 2012-02-04T03:01:30+01:00 Merge branch 'fix-for-186' of https://github.com/sol/haddock into ghc-7.4 - - - - - e8db9031 by David Waern at 2012-02-04T03:07:51+01:00 Style police. - - - - - 261f9462 by David Waern at 2012-02-04T03:20:16+01:00 Update tests. - - - - - 823cfc7c by David Waern at 2012-02-04T03:21:12+01:00 Use mapM_ in accept.hs as well. - - - - - 873dd619 by David Waern at 2012-02-04T03:21:33+01:00 Remove copy.hs - use accept.hs instead. - - - - - 0e31a14a by David Waern at 2012-02-04T03:47:33+01:00 Use <> instead of mappend. - - - - - 2ff7544f by David Waern at 2012-02-04T03:48:55+01:00 Remove code for older ghc versions. - - - - - dacf2786 by David Waern at 2012-02-04T15:52:51+01:00 Clean up some code from last SoC project. - - - - - 00cbb117 by David Waern at 2012-02-04T21:43:49+01:00 Mostly hlint-inspired cleanup. - - - - - 7dc86cc2 by Simon Peyton Jones at 2012-02-06T09:14:41+00:00 Track changes in HsDecls - - - - - f91f82fe by Ian Lynagh at 2012-02-16T13:40:11+00:00 Follow changes in GHC caused by the CAPI CTYPE pragma - - - - - a0ea6b0b by Ian Lynagh at 2012-02-22T02:26:12+00:00 Follow changes in GHC - - - - - b23b07d1 by Simon Peyton Jones at 2012-03-02T16:36:41+00:00 Follow changes in data representation from the big PolyKinds commit - - - - - 43406022 by Simon Hengel at 2012-03-05T11:18:34+01:00 Save/restore global state for static flags when running GHC actions This is necessary if we want to run createInterfaces (from Documentation.Haddock) multiple times in the same process. - - - - - 9fba16fe by Paolo Capriotti at 2012-03-06T10:57:33+00:00 Update .gitignore. - - - - - a9325044 by Simon Peyton Jones at 2012-03-14T17:35:42+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - fd48065a by Iavor Diatchki at 2012-03-15T22:43:35-07:00 Add support for type-level literals. - - - - - 2e8206dd by Simon Peyton Jones at 2012-03-16T14:18:22+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - 93e13319 by Simon Peyton Jones at 2012-03-17T01:04:05+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock Conflicts: src/Haddock/Convert.hs - - - - - d253fa71 by Iavor Diatchki at 2012-03-19T20:12:18-07:00 Merge remote-tracking branch 'origin/master' into type-nats - - - - - fc40acc8 by Iavor Diatchki at 2012-03-19T20:31:27-07:00 Add a missing case for type literals. - - - - - fd2ad699 by Iavor Diatchki at 2012-03-24T13:28:29-07:00 Rename variable to avoid shadowing warning. - - - - - 9369dd3c by Simon Peyton Jones at 2012-03-26T09:14:23+01:00 Follow refactoring of TyClDecl/HsTyDefn - - - - - 38825ca5 by Simon Peyton Jones at 2012-03-26T09:14:37+01:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - 4324ac0f by David Waern at 2012-04-01T01:51:19+02:00 Disable unicode test. - - - - - 3165b750 by David Waern at 2012-04-01T01:51:34+02:00 Take reader environment directly from TypecheckedSource. - - - - - 213b644c by David Waern at 2012-04-01T01:55:20+02:00 Cleanup. - - - - - 3118b4ba by David Waern at 2012-04-01T02:16:15+02:00 Don't filter out unexported names from the four maps - fixes a regression. - - - - - d6524e17 by David Waern at 2012-04-01T02:40:34+02:00 Fix crash when using --qual. Naughty GHC API! - - - - - ea3c43d8 by Henning Thielemann at 2012-04-01T13:03:07+02:00 add QualOption type for distinction between qualification argument given by the user and the actual qualification for a concrete module - - - - - 5422ff05 by Henning Thielemann at 2012-04-01T16:25:02+02:00 emit an error message when the --qual option is used incorrectly - - - - - 026e3404 by David Waern at 2012-04-01T18:10:30+02:00 Don't crash on unicode strings in doc comments. - - - - - ce006632 by David Waern at 2012-04-01T20:13:35+02:00 Add test for --ignore-all-exports flag/ignore-exports pragma. - - - - - 6e4dd33c by David Waern at 2012-04-01T20:21:03+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.4 - - - - - 734ae124 by Henning Thielemann at 2012-04-01T20:22:10+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - 622f9ba5 by David Waern at 2012-04-01T21:26:13+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 55ce17cb by Henning Thielemann at 2012-04-01T22:03:25+02:00 'abbreviate' qualification style - basic support Currently we ignore the package a module is imported from. This means that a module import would shadow another one with the same module name from a different package. - - - - - c85314ef by David Waern at 2012-04-01T22:05:12+02:00 Check qualification option before processing modules. - - - - - ae4b626c by Henning Thielemann at 2012-04-02T00:19:36+02:00 abbreviated qualification: use Packages.lookupModuleInAllPackages for finding the package that a module belongs to - - - - - 60bdbcf5 by Henning Thielemann at 2012-04-02T00:25:31+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - df44301d by Henning Thielemann at 2012-04-02T00:29:05+02:00 qualification style 'abbreviated' -> 'aliased' - - - - - f4192a64 by David Waern at 2012-04-02T01:05:47+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 7ba09067 by David Terei at 2012-04-04T15:08:21-07:00 Fix reporting of modules safe haskell mode (#5989) - - - - - d0cc33d0 by David Terei at 2012-04-06T15:50:41+01:00 Fix reporting of modules safe haskell mode (#5989) - - - - - 6e3434c5 by Simon Peyton Jones at 2012-04-20T18:37:46+01:00 Track changes in HsSyn - - - - - 22014ed0 by Simon Peyton Jones at 2012-05-11T22:45:15+01:00 Follow changes to LHsTyVarBndrs - - - - - d9a07b24 by David Waern at 2012-05-15T01:46:35+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - a6c4ebc6 by David Waern at 2012-05-16T02:18:32+02:00 Update CHANGES. - - - - - 8e181d29 by David Waern at 2012-05-16T02:27:56+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - e358210d by David Waern at 2012-05-16T02:35:33+02:00 Mention the new aliased --qual mode in CHANGES. - - - - - efd36a28 by David Waern at 2012-05-16T21:33:13+02:00 Bump version number. - - - - - d6b3af14 by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for deprecated record field - - - - - 927f800e by Simon Hengel at 2012-05-17T19:08:20+02:00 Use >>= instead of fmap and join - - - - - 048b41d5 by Simon Hengel at 2012-05-17T19:08:20+02:00 newtype-wrap Doc nodes for things that may have warnings attached - - - - - e3a89fc3 by Simon Hengel at 2012-05-17T19:08:20+02:00 Attach warnings to `Documentation` type - - - - - 5d4cc43d by Simon Hengel at 2012-05-17T19:08:20+02:00 Simplify lookupWarning - - - - - cf8ae69d by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for haskell/haddock#205 - - - - - cb409b19 by Simon Peyton Jones at 2012-05-25T08:30:11+01:00 Follow changes in LHsTyVarBndrs - - - - - 2d5f4179 by Simon Hengel at 2012-05-26T19:21:29+02:00 Add Applicative instance for (GenRnM a) - - - - - e4373060 by Simon Hengel at 2012-05-26T19:21:33+02:00 Use a map for warnings, as suggested by @waern - - - - - 597a68c7 by Simon Hengel at 2012-05-27T08:48:24+02:00 Add an optional label to URLs - - - - - ef1ac7fe by Simon Hengel at 2012-05-27T08:48:24+02:00 Add support for hyperlink labels to parser - - - - - 41f2adce by Simon Hengel at 2012-05-27T08:48:24+02:00 Add golden test for hyperlinks - - - - - 83d5e764 by Simon Hengel at 2012-05-27T08:50:02+02:00 Use LANGUAGE pragmas instead of default-extensions in cabal file - - - - - ddb755e5 by Simon Hengel at 2012-05-27T08:50:02+02:00 Fix typo in comment - - - - - 110676b4 by Simon Hengel at 2012-05-27T08:50:02+02:00 Add a type signature for a where-binding - - - - - 7d9ba2a0 by Ian Lynagh at 2012-06-12T14:38:01+01:00 Follow changes in GHC - - - - - 47c704f2 by Ian Lynagh at 2012-06-12T18:52:16+01:00 Follow changes in GHC - - - - - e1efe1ab by Simon Peyton Jones at 2012-06-13T17:25:29+01:00 Follow changes for the implementation of implicit parameters - - - - - 69abc81c by Ian Lynagh at 2012-06-19T22:52:58+01:00 Follow changes in base - - - - - 9d074a21 by Paolo Capriotti at 2012-06-22T18:26:47+01:00 Use right docMap to get decl documentation. - - - - - e3292ef6 by Ian Lynagh at 2012-07-15T01:31:19+01:00 Follow changes in GHC - - - - - ceae56b0 by Ian Lynagh at 2012-07-16T21:22:48+01:00 Fix haddock following some GHC changes Passing _|_ as the Settings for defaultDynFlags no longer works well enough - - - - - 9df72735 by Paolo Capriotti at 2012-07-19T16:49:32+01:00 Forward port changes from stable. - - - - - 572f5fcf by Ian Lynagh at 2012-07-19T20:38:26+01:00 Merge branch 'master' of darcs.haskell.org:/srv/darcs//haddock - - - - - 9195aca4 by Paolo Capriotti at 2012-07-20T10:27:28+01:00 Update dependencies. - - - - - 33db3923 by Ian Lynagh at 2012-07-20T17:54:43+01:00 Build with GHC 7.7 - - - - - 925a2cea by David Waern at 2012-07-23T16:50:40+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.6 Conflicts: src/Haddock/InterfaceFile.hs - - - - - d710ef97 by David Waern at 2012-07-23T16:52:07+02:00 Bump version number. - - - - - eb0c2f83 by David Waern at 2012-07-23T16:57:58+02:00 Update CHANGES. - - - - - b3f56943 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Hide "internal" instances This fixes haskell/haddock#37 (http://trac.haskell.org/haddock/ticket/37) Precisely, we show an instance iff its class and all the types are exported by non-hidden modules. - - - - - a70aa412 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Tests for hiding instances (#37) - - - - - c0f4aa58 by Simon Hengel at 2012-07-27T13:00:13+03:00 Add an other test for hiding instances (#37) - - - - - a7ed6268 by Ian Lynagh at 2012-08-07T14:48:13+01:00 Follow changes in GHC - - - - - 0ab30d38 by Ian Lynagh at 2012-08-13T22:12:27+01:00 Improve haddock memory usage - - - - - 0eaa4e30 by Ian Lynagh at 2012-08-13T23:58:46+01:00 Improve haddock memory usage - - - - - 659d26cf by Ian Lynagh at 2012-08-14T13:16:48+01:00 Remove some temporary pragmas I accidentally recorded - - - - - d97fceb6 by Simon Hengel at 2012-08-25T13:19:34+02:00 Add missing dependency to library - - - - - 4c910697 by Simon Hengel at 2012-08-28T07:39:14+02:00 Move .ghci to project root - - - - - fc3c601a by Simon Hengel at 2012-08-28T07:39:14+02:00 accept.hs: Ignore some files - - - - - 1af9b984 by Simon Hengel at 2012-08-28T07:40:04+02:00 Update reference renderings (bump version) - - - - - 980dc253 by Simon Hengel at 2012-08-28T07:40:32+02:00 Update reference renderings (remove links for ()) - - - - - 33651dbf by Simon Hengel at 2012-08-28T07:41:50+02:00 Update documentation of `runInteractiveProcess` in reference rendering - - - - - 7ab25078 by David Waern at 2012-09-07T10:38:50+02:00 Merge branch 'hiddenInstances2' of http://github.com/feuerbach/haddock into ghc-7.6 - - - - - c3de3a4b by David Waern at 2012-09-07T14:29:27+02:00 Follow changes in GHC. - - - - - 298c43ac by David Waern at 2012-09-07T14:59:24+02:00 Update CHANGES. - - - - - e797993a by David Waern at 2012-09-07T15:21:30+02:00 Update ANNOUNCE. - - - - - d0b44790 by David Waern at 2012-09-07T15:22:43+02:00 Merge branch 'hidden-instances' into ghc-7.6 - - - - - 41a4adc8 by Simon Hengel at 2012-09-08T12:08:37+02:00 Update doc/README - - - - - 71ad1040 by Simon Hengel at 2012-09-08T12:17:17+02:00 Add documentation for URL labels - - - - - 9bb41afd by Simon Peyton Jones at 2012-09-20T18:14:26+01:00 Follow data type changes in the tc-untouchables branch Relating entirely to SynTyConRhs - - - - - b8139bfa by Simon Hengel at 2012-09-21T14:24:16+02:00 Disable Unicode test for now - - - - - a5fafdd7 by Simon Hengel at 2012-09-21T14:35:45+02:00 Update TypeOperators test for GHC 7.6.1 Type operators can't be used as type variables anymore! - - - - - 6ccf0025 by Simon Hengel at 2012-09-21T16:02:24+02:00 Remove (Monad (Either e)) instance from ref. rendering of CrossPackageDocs I do not really understand why the behavior changed, so I'll open a ticket, so that we can further investigate. - - - - - b5c6c138 by Ian Lynagh at 2012-09-27T02:00:57+01:00 Follow changes in GHC build system - - - - - b98eded0 by David Waern at 2012-09-27T15:37:02+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 76cc2051 by David Waern at 2012-09-27T15:48:19+02:00 Update hidden instances tests. - - - - - aeaa1c59 by David Waern at 2012-09-28T10:21:32+02:00 Make API buildable with GHC 7.6. - - - - - d76be1b0 by Simon Peyton Jones at 2012-09-28T15:57:05+01:00 Merge remote-tracking branch 'origin/master' into tc-untouchables - - - - - a1922af8 by David Waern at 2012-09-28T19:50:20+02:00 Fix spurious superclass constraints bug. - - - - - bc41bdbb by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove old examples - - - - - bed7d3dd by Simon Hengel at 2012-10-01T11:30:51+02:00 Adapt parsetests for GHC 7.6.1 - - - - - dcdb22bb by Simon Hengel at 2012-10-01T11:30:51+02:00 Add test-suite section for parsetests to cabal file + get rid of HUnit dependency - - - - - 1e5263c9 by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove test flag from cabal file This was not really used. - - - - - 4beee98b by David Waern at 2012-09-28T23:42:28+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 11dd2256 by Ian Lynagh at 2012-10-03T16:17:35+01:00 Follow change in GHC build system - - - - - fbd77962 by Simon Hengel at 2012-10-03T18:49:40+02:00 Remove redundant dependency from cabal file - - - - - 09218989 by Simon Hengel at 2012-10-04T16:03:05+02:00 Fix typo - - - - - 93a2d5f9 by Simon Hengel at 2012-10-04T16:11:41+02:00 Remove trailing whitespace from cabal file - - - - - c8b46cd3 by Simon Hengel at 2012-10-04T16:12:17+02:00 Export Haddock's main entry point from library - - - - - b411e77b by Simon Hengel at 2012-10-04T16:29:46+02:00 Depend on library for executable The main motivation for this is to increase build speed. In GHC's source tree the library is not build, but all modules are now required for the executable, so that GHC's validate will now detect build failures for the library. - - - - - f8f0979f by Simon Hengel at 2012-10-05T00:32:57+02:00 Set executable flag for Setup.lhs - - - - - dd045998 by Simon Hengel at 2012-10-07T16:44:06+02:00 Extend rather than set environment when running HTML tests On some platforms (e.g. ppc64) GHC requires gcc in the path. - - - - - 7b39c3ae by Simon Hengel at 2012-10-07T17:05:45+02:00 cross-package test: re-export IsString instead of Monad There is a monad instance for Q, which is not available on platforms that do not have GHCi support. This caused CrossPackageDocs to fail on those platforms. Re-exporting IsString should test the same thing, but it works on all platforms. - - - - - 0700c605 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Fix some warnings - - - - - f78eca79 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Make -Wall proof - - - - - 6beec041 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Use listToMaybe/fromMaybe instead of safeHead/maybe - - - - - 44b8ce86 by Ian Lynagh at 2012-10-08T21:59:46+01:00 Follow changes in GHC - - - - - 6da5f702 by Simon Hengel at 2012-10-09T11:16:19+02:00 Update .ghci - - - - - 9ac1a1b9 by Kazu Yamamoto at 2012-10-09T12:45:31+02:00 Add markup support for properties - - - - - 1944cb42 by Simon Hengel at 2012-10-09T12:45:31+02:00 Simplify lexing/parsing of properties In contrast to what we do for examples, we do not really need to capture the "prompt" here. - - - - - bffd8e62 by Simon Hengel at 2012-10-09T13:40:14+02:00 Add HTML test for properties - - - - - 2fe9c5cb by Simon Hengel at 2012-10-09T13:40:21+02:00 Add unit tests for properties - - - - - 874e361b by Simon Hengel at 2012-10-09T13:40:33+02:00 Bump interface version - - - - - 2506cc37 by Simon Hengel at 2012-10-09T15:15:04+02:00 Fix parser bug - - - - - 743d2b7d by Simon Hengel at 2012-10-09T15:31:06+02:00 Allow to load interface files with compatible versions - - - - - 981a1660 by Simon Hengel at 2012-10-10T10:32:05+02:00 Export more types from Documentation.Haddock (fixes haskell/haddock#216) - - - - - dff7dc76 by Simon Hengel at 2012-10-10T11:15:19+02:00 Update ANNOUNCE and CHANGES - - - - - edd2bb01 by Simon Hengel at 2012-10-10T11:22:50+02:00 Bump version - - - - - 5039163b by Simon Hengel at 2012-10-10T13:56:04+02:00 Fix typo in documentation - - - - - e4ce34da by Simon Hengel at 2012-10-10T14:28:35+02:00 Add documentation for properties - - - - - 9555ebca by Simon Hengel at 2012-10-11T10:49:04+02:00 Remove redundant if-defs, more source documentation - - - - - 87aa67e1 by Simon Hengel at 2012-10-11T12:32:51+02:00 Adapt cabal file - - - - - c44c1dee by Simon Hengel at 2012-10-11T12:41:58+02:00 Require ghc 7.6 - - - - - 8383bc34 by Simon Hengel at 2012-10-11T12:50:24+02:00 Bump version - - - - - 1030eb38 by Simon Hengel at 2012-10-11T12:55:44+02:00 Update ANNOUNCE and CHANGES - - - - - 74955088 by Simon Hengel at 2012-10-12T09:49:31+02:00 Improve note about `binaryInterfaceVersion` (thanks David) - - - - - ee30f6b7 by Simon Hengel at 2012-10-13T13:40:59+02:00 Update version in html tests, rpm spec file, and user manual - - - - - f2861f18 by Simon Hengel at 2012-10-13T14:40:33+02:00 Remove unused MonadFix constraint - - - - - dfdf1a74 by Simon Hengel at 2012-10-13T15:15:38+02:00 Minor code simplification - - - - - 4ecd1e70 by Simon Hengel at 2012-10-13T15:33:43+02:00 Increase code locality - - - - - f7df5cc9 by Simon Hengel at 2012-10-13T16:03:12+02:00 Minor code simplification - - - - - e737eb6e by Simon Hengel at 2012-10-13T19:03:04+02:00 Handle HsExplicitListTy in renameer (fixes haskell/haddock#213) - - - - - c2dc8f17 by Simon Hengel at 2012-10-13T20:46:31+02:00 Better error messages - - - - - 14d48b4c by Simon Hengel at 2012-10-14T00:21:07+02:00 Simplify RnM type - - - - - 6c2cc547 by Simon Hengel at 2012-10-14T00:23:35+02:00 Simplify lookupRn - - - - - bc77ce85 by Simon Hengel at 2012-10-14T01:51:32+02:00 Organize unite tests hierarchically - - - - - 2306d117 by Simon Hengel at 2012-10-14T10:34:58+02:00 Handle more cases in renameType - - - - - 8a864203 by Simon Hengel at 2012-10-14T11:47:59+02:00 Add mini_HiddenInstances.html.ref and mini_HiddenInstancesB.html.ref - - - - - 3a978eca by Simon Hengel at 2012-10-14T11:49:28+02:00 Add /tests/html-tests/output/ to .gitignore - - - - - db18888a by Simon Hengel at 2012-10-14T13:38:21+02:00 Allow haddock markup in deprecation messages - - - - - e7cfee9f by Simon Hengel at 2012-10-14T14:00:23+02:00 If parsing of deprecation message fails, include it verbatim - - - - - 242a85be by Simon Hengel at 2012-10-14T14:13:24+02:00 Add description for PruneWithWarning test - - - - - 43d33df1 by Simon Hengel at 2012-10-14T15:40:53+02:00 Minor formatting change - - - - - 22768c44 by Simon Hengel at 2012-10-14T16:03:43+02:00 Properly handle deprecation messages for re-exported things (fixes haskell/haddock#220) - - - - - cb4b9111 by Simon Hengel at 2012-10-14T17:30:28+02:00 Add build artifacts for documentation to .gitignore - - - - - 854cd8de by Simon Hengel at 2012-10-14T23:34:51+02:00 unit-tests: Improve readability Add IsString instance for (Doc RdrName) + use <> instead of DocAppend. - - - - - c4446d54 by Simon Hengel at 2012-10-14T23:37:21+02:00 unit-tests: Minor refactoring Rename parse to parseParas. - - - - - 04f2703c by Simon Hengel at 2012-10-15T00:36:42+02:00 Fix typo - - - - - 3d109e44 by Simon Hengel at 2012-10-15T10:30:07+02:00 Add description for DeprecatedReExport test - - - - - 84f0985c by Simon Hengel at 2012-10-15T14:54:19+02:00 Move resources to /resources directory - - - - - a5de7ca6 by Simon Hengel at 2012-10-15T15:46:18+02:00 Move HTML tests to directory /html-test/ - - - - - e21f727d by Simon Hengel at 2012-10-15T19:32:42+02:00 Move HTML reference renderings to /html-test/ref/ - - - - - 3a3c6c75 by Simon Hengel at 2012-10-15T19:32:42+02:00 Copy css, images, etc. on accept - - - - - 40ead6dc by Simon Hengel at 2012-10-15T19:32:42+02:00 Move unit tests to /test directory - - - - - 99a28231 by Simon Hengel at 2012-10-15T19:32:42+02:00 Fix Setup.lhs /usr/bin/runhaskell is not installed on all systems. - - - - - 95faf45e by Simon Hengel at 2012-10-15T19:32:42+02:00 Make test management scripts more robust * They are now independent from the current directory, and hence can be called from everywhere * On UNIX/Linux they can now be run as scripts - - - - - 027aaa2d by Simon Hengel at 2012-10-15T19:53:40+02:00 Add 'dev' flag to cabal file, that builds without -O2 That way --disable-optimization can be used, which decreases build time considerably. - - - - - e0266ede by Simon Hengel at 2012-10-15T20:03:43+02:00 Add test case for "spurious superclass constraints bug" - - - - - 52a2aa92 by Simon Hengel at 2012-10-15T20:28:55+02:00 Adapt accept.lhs, so that it ignores more index files - - - - - 53530781 by Simon Hengel at 2012-10-15T20:49:39+02:00 Rename html-test/runtests.lhs to html-test/run.lhs - - - - - 84518797 by Simon Hengel at 2012-10-15T20:49:39+02:00 Move source files for HTML tests to html-test/src - - - - - a911dc6c by Simon Hengel at 2012-10-15T20:49:39+02:00 Adapt output directory for HTML tests - - - - - d3c15857 by Ian Lynagh at 2012-10-16T16:54:43+01:00 Follow dopt->gopt rename - - - - - 956665a5 by Simon Hengel at 2012-10-18T08:42:48+02:00 Update html-test/README - - - - - 903b1029 by Simon Hengel at 2012-10-18T08:50:26+02:00 Use markdown for html-test/README - - - - - 150b4d63 by Ian Lynagh at 2012-10-18T16:36:00+01:00 Follow changes in GHC: 'flags' has been renamed 'generalFlags' - - - - - 41e04ff9 by Simon Hengel at 2012-11-28T09:54:35+01:00 Export missing types from Documentation.Haddock - - - - - 9be59237 by Ian Lynagh at 2012-11-30T23:20:47+00:00 Update dependencies - - - - - e06842f5 by Simon Hengel at 2012-12-07T20:58:05+01:00 Bump version - - - - - e3dbede0 by Simon Hengel at 2012-12-07T20:58:05+01:00 Add missing test files to cabal file (fixes haskell/haddock#230) - - - - - ee0dcca7 by Simon Hengel at 2012-12-07T20:58:05+01:00 Update CHANGES - - - - - 51601bdb by Simon Peyton Jones at 2012-12-19T17:28:35+00:00 Track changes in UNPACK pragma stuff - - - - - f2573bc1 by Richard Eisenberg at 2012-12-21T20:56:25-05:00 Implement overlapping type family instances. An ordered, overlapping type family instance is introduced by 'type instance where', followed by equations. See the new section in the user manual (7.7.2.2) for details. The canonical example is Boolean equality at the type level: type family Equals (a :: k) (b :: k) :: Bool type instance where Equals a a = True Equals a b = False A branched family instance, such as this one, checks its equations in order and applies only the first the matches. As explained in the note [Instance checking within groups] in FamInstEnv.lhs, we must be careful not to simplify, say, (Equals Int b) to False, because b might later unify with Int. This commit includes all of the commits on the overlapping-tyfams branch. SPJ requested that I combine all my commits over the past several months into one monolithic commit. The following GHC repos are affected: ghc, testsuite, utils/haddock, libraries/template-haskell, and libraries/dph. Here are some details for the interested: - The definition of CoAxiom has been moved from TyCon.lhs to a new file CoAxiom.lhs. I made this decision because of the number of definitions necessary to support BranchList. - BranchList is a GADT whose type tracks whether it is a singleton list or not-necessarily-a-singleton-list. The reason I introduced this type is to increase static checking of places where GHC code assumes that a FamInst or CoAxiom is indeed a singleton. This assumption takes place roughly 10 times throughout the code. I was worried that a future change to GHC would invalidate the assumption, and GHC might subtly fail to do the right thing. By explicitly labeling CoAxioms and FamInsts as being Unbranched (singleton) or Branched (not-necessarily-singleton), we make this assumption explicit and checkable. Furthermore, to enforce the accuracy of this label, the list of branches of a CoAxiom or FamInst is stored using a BranchList, whose constructors constrain its type index appropriately. I think that the decision to use BranchList is probably the most controversial decision I made from a code design point of view. Although I provide conversions to/from ordinary lists, it is more efficient to use the brList... functions provided in CoAxiom than always to convert. The use of these functions does not wander far from the core CoAxiom/FamInst logic. BranchLists are motivated and explained in the note [Branched axioms] in CoAxiom.lhs. - The CoAxiom type has changed significantly. You can see the new type in CoAxiom.lhs. It uses a CoAxBranch type to track branches of the CoAxiom. Correspondingly various functions producing and consuming CoAxioms had to change, including the binary layout of interface files. - To get branched axioms to work correctly, it is important to have a notion of type "apartness": two types are apart if they cannot unify, and no substitution of variables can ever get them to unify, even after type family simplification. (This is different than the normal failure to unify because of the type family bit.) This notion in encoded in tcApartTys, in Unify.lhs. Because apartness is finer-grained than unification, the tcUnifyTys now calls tcApartTys. - CoreLinting axioms has been updated, both to reflect the new form of CoAxiom and to enforce the apartness rules of branch application. The formalization of the new rules is in docs/core-spec/core-spec.pdf. - The FamInst type (in types/FamInstEnv.lhs) has changed significantly, paralleling the changes to CoAxiom. Of course, this forced minor changes in many files. - There are several new Notes in FamInstEnv.lhs, including one discussing confluent overlap and why we're not doing it. - lookupFamInstEnv, lookupFamInstEnvConflicts, and lookup_fam_inst_env' (the function that actually does the work) have all been more-or-less completely rewritten. There is a Note [lookup_fam_inst_env' implementation] describing the implementation. One of the changes that affects other files is to change the type of matches from a pair of (FamInst, [Type]) to a new datatype (which now includes the index of the matching branch). This seemed a better design. - The TySynInstD constructor in Template Haskell was updated to use the new datatype TySynEqn. I also bumped the TH version number, requiring changes to DPH cabal files. (That's why the DPH repo has an overlapping-tyfams branch.) - As SPJ requested, I refactored some of the code in HsDecls: * splitting up TyDecl into SynDecl and DataDecl, correspondingly changing HsTyDefn to HsDataDefn (with only one constructor) * splitting FamInstD into TyFamInstD and DataFamInstD and splitting FamInstDecl into DataFamInstDecl and TyFamInstDecl * making the ClsInstD take a ClsInstDecl, for parallelism with InstDecl's other constructors * changing constructor TyFamily into FamDecl * creating a FamilyDecl type that stores the details for a family declaration; this is useful because FamilyDecls can appear in classes but other decls cannot * restricting the associated types and associated type defaults for a * class to be the new, more restrictive types * splitting cid_fam_insts into cid_tyfam_insts and cid_datafam_insts, according to the new types * perhaps one or two more that I'm overlooking None of these changes has far-reaching implications. - The user manual, section 7.7.2.2, is updated to describe the new type family instances. - - - - - f788d0fb by Simon Peyton Jones at 2012-12-23T15:49:58+00:00 Track changes in HsBang - - - - - ca460a0c by Simon Peyton Jones at 2012-12-23T15:50:28+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - f078fea6 by Simon Peyton Jones at 2013-01-02T08:33:13+00:00 Use InstEnv.instanceSig rather than instanceHead (name change) - - - - - 88e41305 by Simon Peyton Jones at 2013-01-14T17:10:27+00:00 Track change to HsBang type - - - - - e1ad4e19 by Kazu Yamamoto at 2013-02-01T11:59:24+09:00 Merge branch 'ghc-7.6' into ghc-7.6-merge-2 Conflicts: haddock.cabal src/Haddock/Interface/AttachInstances.hs src/Haddock/Interface/Create.hs src/Haddock/Interface/LexParseRn.hs src/Haddock/InterfaceFile.hs src/Haddock/Types.hs Only GHC HEAD can compile this. GHC 7.6.x cannot compile this. Some test fail. - - - - - 62bec012 by Kazu Yamamoto at 2013-02-06T11:12:28+09:00 Using tcSplitSigmaTy in instanceHead' (FIXME is resolved.) - - - - - 013fd2e4 by Kazu Yamamoto at 2013-02-06T17:56:21+09:00 Refactoring instanceHead'. - - - - - 3148ce0e by Kazu Yamamoto at 2013-02-07T17:45:10+09:00 Using new syntax in html-test/src/GADTRecords.hs. - - - - - 626dabe7 by Gabor Greif at 2013-02-15T22:42:01+01:00 Typo - - - - - 1eb667ae by Ian Lynagh at 2013-02-16T17:02:07+00:00 Follow changes in base - - - - - 3ef8253a by Ian Lynagh at 2013-03-01T23:23:57+00:00 Follow changes in GHC's build system - - - - - 1a265a3c by Ian Lynagh at 2013-03-03T23:12:07+00:00 Follow changes in GHC build system - - - - - 69941c79 by Max Bolingbroke at 2013-03-10T09:38:28-07:00 Use Alex 3's Unicode support to properly lex source files as UTF-8 Signed-off-by: David Waern <david.waern at gmail.com> - - - - - ea687dad by Simon Peyton Jones at 2013-03-15T14:16:10+00:00 Adapt to tcRnGetInfo returning family instances too This API change was part of the fix to Trac haskell/haddock#4175. But it offers new information to Haddock: the type-family instances, as well as the class instances, of this type. This patch just drops the new information on the floor, but there's an open opportunity to use it in the information that Haddock displays. - - - - - 971a30b0 by Andreas Voellmy at 2013-05-19T20:47:39+01:00 Fix for haskell/haddock#7879. Changed copy of utils/haddock/html/resources/html to use "cp -RL" rather than "cp -R". This allows users to run validate in a build tree, where the build tree was setup using lndir with a relative path to the source directory. - - - - - 31fb7694 by Ian Lynagh at 2013-05-19T20:47:49+01:00 Use "cp -L" when making $(INPLACE_LIB)/latex too - - - - - e9952233 by Simon Hengel at 2013-06-01T18:06:50+02:00 Add -itest to .ghci - - - - - b06873b3 by Mateusz Kowalczyk at 2013-06-01T18:06:50+02:00 Workaround for a failing build with --enable-tests. - - - - - e7858d16 by Simon Hengel at 2013-06-01T19:29:28+02:00 Fix broken test - - - - - 0690acb1 by Richard Eisenberg at 2013-06-21T14:08:25+01:00 Updates to reflect changes in HsDecls to support closed type families. - - - - - 7fd347ec by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 53ed81b6 by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 931c4f4f by Richard Eisenberg at 2013-07-24T13:15:59+01:00 Remove (error "synifyKind") to use WithinType, to allow haddock to process base. - - - - - 55a9c804 by Richard Eisenberg at 2013-08-02T15:54:55+01:00 Changes to reflect changes in GHC's type HsTyVarBndr - - - - - b6e9226c by Mathieu Boespflug at 2013-08-04T10:39:43-07:00 Output Copright and License keys in Xhtml backend. This information is as relevant in the documentation as it is in the source files themselves. Signed-off-by: David Waern <david.waern at gmail.com> - - - - - 4c66028a by David Waern at 2013-08-04T15:27:36-07:00 Bump interface file version. - - - - - 67340163 by David Waern at 2013-08-09T16:12:51-07:00 Update tests. - - - - - 2087569b by Mateusz Kowalczyk at 2013-08-25T09:24:13+02:00 Add spec tests. This adds tests for all elements we can create during regular parsing. This also adds tests for text with unicode in it. - - - - - 97f36a11 by Mateusz Kowalczyk at 2013-08-27T06:59:12+01:00 Fix ticket haskell/haddock#247. I do the same thing that the XHTML backend does: give these no special treatment and just act as if they are regular functions. - - - - - 60681b4f by Mateusz Kowalczyk at 2013-08-27T21:22:48+02:00 LaTeX tests setup - - - - - fa4c27b2 by Mateusz Kowalczyk at 2013-09-02T23:21:43+01:00 Fixes haskell/haddock#253 - - - - - 1a202490 by Mateusz Kowalczyk at 2013-09-03T01:12:50+01:00 Use Hspec instead of nanospec This is motivated by the fact that Haddock tests are not ran by the GHC's ‘validate’ script so we're pretty liberal on dependencies in that area. Full Hspec gives us some nice features such as Quickcheck integration. - - - - - 8cde3b20 by David Luposchainsky at 2013-09-08T07:27:28-05:00 Fix AMP warnings Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - d10661f2 by Herbert Valerio Riedel at 2013-09-11T15:15:01+02:00 Update Git repo URL in `.cabal` file - - - - - 16a44eb5 by Richard Eisenberg at 2013-09-17T09:34:26-04:00 Revision to reflect new role annotation syntax in GHC. - - - - - 4b9833b9 by Herbert Valerio Riedel at 2013-09-18T10:15:28+02:00 Add missing `traverse` method for `GenLocated` As `Traversable` needs at least one of `traverse` or `sequenceA` to be overridden. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - b71fed5d by Simon Hengel at 2013-09-18T22:43:34+02:00 Add test helper - - - - - 4fc1ea86 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#231 - - - - - 435872f6 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#256 We inject -dynamic-too into flags before we run all our actions in the GHC monad. - - - - - b8b24abb by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Add new field to DynFlags - - - - - 49558795 by Simon Hengel at 2013-09-18T22:43:35+02:00 Fallback to ./resources when Cabal data is not found (so that themes are found during development) - - - - - bf79d05c by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Fixes haskell/haddock#5 - - - - - e1baebc2 by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Print missing documentation. Fixes haskell/haddock#258. - - - - - 02ea74de by Austin Seipp at 2013-10-09T10:52:22-05:00 Don't consider StaticFlags when parsing arguments. Instead, discard any static flags before parsing the command line using GHC's DynFlags parser. See http://ghc.haskell.org/trac/ghc/ticket/8276 Based off a patch from Simon Hengel. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 704fd5bb by Simon Hengel at 2013-11-09T00:15:13+01:00 Update HTML tests - - - - - f9fed49e by Simon Hengel at 2013-11-10T18:43:58+01:00 Bump version - - - - - 97ae1999 by Simon Peyton Jones at 2013-11-25T17:25:14+00:00 Track changes in HsSpliceTy data constructor - - - - - 59ad8268 by Simon Peyton Jones at 2014-01-10T18:17:43+00:00 Adapt to small change in Pretty's exports - - - - - 8b12e6aa by Simon Hengel at 2014-01-12T14:48:35-06:00 Some code simplification by using traverse - - - - - fc5ea9a2 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix warnings in test helper - - - - - 6dbb3ba5 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Add ByteString version of Attoparsec - - - - - 968d7774 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 One pass parser and tests. We remove the HTML test as it is no longer necessary. We cover the test case in spec tests and other HTML tests but keeping this around fails: this is because the new parser has different semantics there. In fact, I suspect the original behaviour was a bug that wasn't caught/fixed but simply included as-is during the testing. - - - - - 37a07c9c by Simon Hengel at 2014-01-12T14:48:35-06:00 Rename Haddock.ParseSpec to Haddock.ParserSpec - - - - - f0f68fe9 by Simon Hengel at 2014-01-12T14:48:35-06:00 Don't append newline to parseString input We also check that we have parsed everything with endOfInput. - - - - - 95d60093 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix totality, unicode, examples, paragraph parsing Also simplify specs and parsers while we're at it. Some parsers were made more generic. This commit is a part of GHC pre-merge squash, email fuuzetsu at fuuzetsu.co.uk if you need the full commit history. - - - - - 7d99108c by Simon Hengel at 2014-01-12T14:48:35-06:00 Update acceptance tests - - - - - d1b59640 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Support for bold. Conflicts: src/Haddock/Backends/Hoogle.hs src/Haddock/Interface/Rename.hs src/Haddock/Parser.hs - - - - - 4b412b39 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Allow for headings inside function documentation. LaTeX will treat the h3-h6 headings the same as we'd have to hack the style file heavily otherwise and it would make the headings tiny anyway. Hoogle upstream said they will put in the functionality on their end. Conflicts: src/Haddock/Interface/Rename.hs src/Haddock/Types.hs test/Haddock/ParserSpec.hs - - - - - fdcca428 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Per-module extension flags and language listing. Any extensions that are not enabled by a used language (Haskell2010 &c) will be shown. Furthermore, any implicitly enabled are also going to be shown. While we could eliminate this either by using the GHC API or a dirty hack, I opted not to: if a user doesn't want the implied flags to show, they are recommended to use enable extensions more carefully or individually. Perhaps this will encourage users to not enable the most powerful flags needlessly. Enabled with show-extensions. Conflicts: src/Haddock/InterfaceFile.hs - - - - - 368942a2 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Bump interface version There were some breaking changes over the last few patches so we bump the interface version. This causes a big problem with testing: 1. To generate cross package docs, we first need to generate docs for the package used. 2. To generate package docs with new interface version, we need to use Haddock which has the version bumped. 3. To get Haddock with the version bump, we first need to test cross package docs 4. GOTO 1 So the problem is the chicken and the egg problem. It seems that the only solution would be to generate some interface files on the fly but it is non-trivial. To run this test, you'll have to: * build Haddock without the test (make sure everything else passes) * rebuild the packages used in the test with your shiny new binary making sure they are visible to Haddock * remove the ‘_hidden’ suffix and re-run the tests Note: because the packages currently used for this test are those provided by GHC, it's probably non-trivial to just re-build them. Preferably something less tedious to rebuild should be used and something that is not subject to change. - - - - - 124ae7a9 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow for nesting of paragraphs under lists. The nesting rules are similar to Markdown's with the exception that we can not simply indent the first line of a hard wrapped indented paragraph and have it treated as if it was fully indented. The reason is differences in markup as some of our constructs care about whitespace while others just swallow everything up so it's just a lot easier to not bother with it rather than making arbitrary rules. Note that we now drop trailing for string entities inside of lists. They weren't needed and it makes the output look uniform whether we use a single or double newline between list elements. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - c7913535 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow escaping in URLs and pictures. Some tests were moved under parseString as they weren't about paragraph level markup. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - 32326680 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update documentation. - - - - - fbef6406 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update maintainer - - - - - b40e82f4 by Mateusz Kowalczyk at 2014-01-13T02:39:25-06:00 Fixes haskell/haddock#271 Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - f4eafbf8 by Gergő Érdi at 2014-01-19T15:35:16-06:00 Support for -XPatternSynonyms Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - a8939591 by Austin Seipp at 2014-01-29T08:09:04-06:00 Update CPP check for __GLASGOW_HASKELL__ Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 30d7e9d5 by Gergő Érdi at 2014-01-31T00:15:01+08:00 <+>: Don't insert a space when concatenating empty nodes - - - - - a25ccd4d by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Fix @ code blocks In cases where we had some horizontal space before the closing ‘@’, the parser would not accept the block as a code block and we'd get ugly output. - - - - - 0f67305a by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Update tests This updates tests due to Haddock Trac haskell/haddock#271 fix and due to removal of TypeHoles as an extension from GHC. - - - - - 157322a7 by Gergő Érdi at 2014-01-31T01:03:17+08:00 Handle infix vs prefix names correctly everywhere, by explicitly specifying the context The basic idea is that "a" and "+" are either pretty-printed as "a" and "(+)" or "`a`" and "+" - - - - - aa6d9685 by Mateusz Kowalczyk at 2014-01-30T17:21:50+00:00 Correct whitespace in ‘hidden’ test for <+> change - - - - - 121872f0 by Mateusz Kowalczyk at 2014-02-09T17:59:12+00:00 Document module header. Fixes Haddock Trac haskell/haddock#270. - - - - - e3253746 by Mateusz Kowalczyk at 2014-02-10T21:37:48+00:00 Insert a space between module link and description Fixes Haddock Trac haskell/haddock#277. - - - - - 771d2384 by Mateusz Kowalczyk at 2014-02-10T23:27:21+00:00 Ensure a space between type signature and ‘Source’ This is briefly related to Haddock Trac haskell/haddock#249 and employs effectively the suggested fix _but_ it doesn't actually fix the reported issue. This commit simply makes copying the full line a bit less of a pain. - - - - - 8cda9eff by nand at 2014-02-11T15:48:30+00:00 Add support for type/data families This adds support for type/data families with their respective instances, as well as closed type families and associated type/data families. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 3f22c510 by nand at 2014-02-11T15:53:50+00:00 Improve display of poly-kinded type operators This now displays them as (==) k a b c ... to mirror GHC's behavior, instead of the old (k == a) b c ... which was just wrong. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - effb2d6b by nand at 2014-02-11T15:56:50+00:00 Add test case for PatternSynonyms This just tests various stuff including poly-kinded patterns and operator patterns to make sure the rendering isn't broken. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - b38faf0d by Niklas Haas at 2014-02-13T21:53:32+00:00 Get rid of re-implementation of sortBy I have no idea what this was doing lying around here, and due to the usage of tuples it's actually slower, too. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - ac1e0413 by Mateusz Kowalczyk at 2014-02-13T23:57:16+00:00 Only warn about missing docs when docs are missing This fixes the ‘Missing documentation for…’ message for modules with 100% coverage. - - - - - cae2e36a by Niklas Haas at 2014-02-15T21:56:18+00:00 Add test case for inter-module type/data family instances These should show up in every place where the class is visible, and indeed they do right now. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 8bea5c3a by Mateusz Kowalczyk at 2014-02-19T05:11:34+00:00 Use a bespoke data type to indicate fixity This deals with what I imagine was an ancient TODO and makes it much clearer what the argument actually does rather than having the user chase down the comment. - - - - - 5b52d57c by Niklas Haas at 2014-02-22T21:31:03+01:00 Strip a single leading space from bird tracks (#201) This makes bird tracks in the form > foo > bar > bat parse as if they had been written as >foo >bar >bat ie. without the leading whitespace in front of every line. Ideally we also want to look into how leading whitespace affects code blocks written using the @ @ syntax, which are currently unaffected by this patch. - - - - - 5a1315a5 by Simon Hengel at 2014-02-22T21:55:35+01:00 Turn a source code comment into specs - - - - - 784cfe58 by Mateusz Kowalczyk at 2014-02-23T05:02:22+00:00 Update test case for lifted GADT type rendering The parsing of these seems to have been fixed by GHC folk and it now renders differently. IMHO it now renders in a better way so I'm updating the test to reflect this. - - - - - c3c88c2f by Mateusz Kowalczyk at 2014-02-23T06:37:14+00:00 Don't shadow ‘strip’. -Wall complains - - - - - 293031d8 by Niklas Haas at 2014-02-23T15:21:52+01:00 Make ImplicitParams render correctly (#260) This introduces a new precedence level for single contexts (because implicit param contexts always need parens around them, but other types of contexts don't necessarily, even when alone) - - - - - 4200842d by Niklas Haas at 2014-02-23T15:37:13+01:00 Lower precedence of equality constraints This drops them to the new precedence pREC_CTX, which makes single eqaulity constraints show up as (a ~ b) => ty, in line with GHC's rendering. Additional tests added to make sure other type operators render as intended. Current behavior matches GHC - - - - - b59e3227 by Niklas Haas at 2014-02-23T16:11:22+01:00 Add RankNTypes test case to ImplicitParams.hs This test actually tests what haskell/haddock#260 originally reported - I omitted the RankNTypes scenario from the original fix because I realized it's not relevant to the underlying issue and indeed, this renders as intended now. Still good to have more tests. - - - - - c373dbf7 by Mateusz Kowalczyk at 2014-02-24T06:09:54+00:00 Fix rendering of Contents when links are present Fixes Haddock Trac haskell/haddock#267. - - - - - 9ecb0e56 by Mateusz Kowalczyk at 2014-02-24T06:26:50+00:00 Fix wording in the docs - - - - - 4f4dcd8e by Mateusz Kowalczyk at 2014-02-27T03:00:33+00:00 Change rendering of duplicate record field docs See Haddock Trac haskell/haddock#195. We now change this behaviour to only rendering the documentation attached to the first instance of a duplicate field. Perhaps we could improve this by rendering the first instance that has documentation attached to it but for now, we'll stick with this. - - - - - ad8aa609 by Niklas Haas at 2014-03-08T09:43:26+01:00 Render fixity information Affects functions, type synonyms, type families, class names, data type names, constructors, data families, associated TFs/DFs, type synonyms, pattern synonyms and everything else I could think of. - - - - - 6a39c917 by Niklas Haas at 2014-03-09T07:43:39+01:00 Reorder topDeclElem to move the source/wiki links to the top They appear in the same position due to the float: right attribute but now they're always at the top of the box instead of at the bottom. - - - - - 2d34b3b4 by Niklas Haas at 2014-03-09T07:53:46+01:00 Use optLast instead of listToMaybe for sourceUrls/wikiUrls This lets you override them using eg. cabal haddock --haddock-options, which can come in handy if you want to use a different layout or URL for your source code links than cabal-install generates. - - - - - 0eff4624 by Niklas Haas at 2014-03-09T07:53:46+01:00 Differentiate between TH splices (line-links) and regular names This adds a new type of source code link, to a specific line rather than a specific declaration/name - this is used to link to the location of a TH splice that defines a certain name. Rather hefty changes throughout and still one unresolved issue (the line URLs aren't parsed from the third form of --read-interface which means they're currently restricted to same-interface links). Not sure if this issue is really worth all the hassle, especially since we could just use line links in general. This commit also contains some cleanup/clarification of the types in Haddock.Backends.Xhtml.Decl and shortens some overlong lines in the process. Notably, the Bool parameter was replaced by a Unicode type synonym to help clarify its presence in type signatures. - - - - - 66d6f77b by Niklas Haas at 2014-03-09T20:02:43+01:00 Group similar fixities together Identical fixities declared for the same line should now render using syntax like: infix 4 <, >=, >, <= - - - - - 6587f9f5 by Mateusz Kowalczyk at 2014-03-10T04:24:18+00:00 Update changelog - - - - - 7387ddad by Niklas Haas at 2014-03-11T10:26:04+01:00 Include fixity information in the Interface file This resolves fixity information not appearing across package borders. The binary file version has been increased accordingly. - - - - - ab46ef44 by Niklas Haas at 2014-03-11T10:26:04+01:00 Update changelog - - - - - 565cab6f by Niklas Haas at 2014-03-11T10:26:04+01:00 Update appearance of fixity annotations This moves them in-line with their corresponding lines, similar to a presentation envision by @hvr and described in #ghc. Redundant operator names are also omitted when no ambiguity is present. - - - - - 5d7afd67 by Niklas Haas at 2014-03-11T10:26:05+01:00 Filter family instances of hidden types Currently, this check does not extend to hidden right hand sides, although it probably should hide them in that case. - - - - - ec291b0c by Niklas Haas at 2014-03-11T10:26:05+01:00 Add documentation for --source-entity-line - - - - - 0922e581 by Niklas Haas at 2014-03-11T10:37:32+01:00 Revert "Reorder topDeclElem to move the source/wiki links to the top" This reverts commit 843c42c4179526a2ad3526e4c7d38cbf4d50001d. This change is no longer needed with the new rendering style, and it messes with copy/pasting lines. - - - - - 30618e8b by Mateusz Kowalczyk at 2014-03-11T09:41:07+00:00 Bump version to 2.15.0 - - - - - adf3f1bb by Mateusz Kowalczyk at 2014-03-11T09:41:09+00:00 Fix up some whitespace - - - - - 8905f57d by Niklas Haas at 2014-03-13T19:18:06+00:00 Hide RHS of TFs with non-exported right hand sides Not sure what to do about data families yet, since technically it would not make a lot of sense to display constructors that cannot be used by the user. - - - - - 5c44d5c2 by Niklas Haas at 2014-03-13T19:18:08+00:00 Add UnicodeSyntax alternatives for * and -> I could not find a cleaner way to do this other than checking for string equality with the given built-in types. But seeing as it's actually equivalent to string rewriting in GHC's implementation of UnicodeSyntax, it's probably fitting. - - - - - b04a63e6 by Niklas Haas at 2014-03-13T19:18:10+00:00 Display minimal complete definitions for type classes This corresponds to the new {-# MINIMAL #-} pragma present in GHC 7.8+. I also cleaned up some of the places in which ExportDecl is used to make adding fields easier in the future. Lots of test cases have been updated since they now render with minimality information. - - - - - a4a20b16 by Niklas Haas at 2014-03-13T19:18:12+00:00 Strip links from recently added html tests These were accidentally left there when the tests were originally added - - - - - d624f315 by Mateusz Kowalczyk at 2014-03-13T19:19:31+00:00 Update changelog - - - - - d27a21ac by Mateusz Kowalczyk at 2014-03-13T21:19:07+00:00 Always read in prologue files as UTF8 (#286). - - - - - 54b2fd78 by Mateusz Kowalczyk at 2014-03-13T21:28:09+00:00 Style only - - - - - fa4fe650 by Simon Hengel at 2014-03-15T09:04:18+01:00 Add Fuuzetsu maintainers field in cabal file - - - - - f83484b7 by Niklas Haas at 2014-03-15T18:20:24+00:00 Hide minimal definition for only-method classes Previously this was not covered by the All xs check since here it is not actually an All, rather a single Var n. This also adds the previously missing html-test/src/Minimal.hs. - - - - - 0099d276 by Niklas Haas at 2014-03-15T18:20:26+00:00 Fix issue haskell/haddock#281 This is a regression from the data family instances change. Data instances are now distinguished from regular lists by usage of the new class "inst", and the style has been updated to only apply to those. I've also updated the appropriate test case to test this a bit better, including GADT instances with GADT-style records. - - - - - 1f9687bd by Mateusz Kowalczyk at 2014-03-21T17:48:37+00:00 Please cabal sdist - - - - - 75542693 by Mateusz Kowalczyk at 2014-03-22T16:36:16+00:00 Drop needless --split-objs which slows us down. Involves tiny cleanup of all the dynflag bindings. Fixes haskell/haddock#292. - - - - - 31214dc3 by Herbert Valerio Riedel at 2014-03-23T18:01:01+01:00 Fix a few typos Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 0b73e638 by Mateusz Kowalczyk at 2014-03-31T05:34:36+01:00 Print kind signatures on GADTs - - - - - 2bab42f3 by Mateusz Kowalczyk at 2014-03-31T16:53:25+01:00 Add default for new PlatformConstraints field - - - - - 42647c5f by Mateusz Kowalczyk at 2014-03-31T18:29:04+01:00 Drop leading whitespace in @-style blocks. Fixes haskell/haddock#201. - - - - - 98208294 by Niklas Haas at 2014-03-31T20:09:58+02:00 Crash when exporting record selectors of data family instances This fixes bug haskell/haddock#294. This also fixes a related but never-before-mentioned bug about the display of GADT record selectors with non-polymorphic type signatures. Note: Associated data type constructors fail to show up if nothing is exported that they could be attached to. Exporting any of the data types in the instance head, or the class + data family itself, causes them to show up, but in the absence of either of these, exporting just the associated data type with the constructor itself will result in it being hidden. The only scenario I can come up that would involve this kind of situation involved OverlappingInstances, and even then it can be mitigated by just exporting the class itself, so I'm not going to solve it since the logic would most likely be very complicated. - - - - - 3832d171 by Mateusz Kowalczyk at 2014-04-01T19:07:33+01:00 Make CHANGES consistent with what's now in 2.14.2 - - - - - c386ae89 by Mateusz Kowalczyk at 2014-04-01T19:18:36+01:00 Actually bundle extra spec tests in sdist - - - - - bd57a6d3 by Mateusz Kowalczyk at 2014-04-03T21:13:48+01:00 Update test cases for GHC bug haskell/haddock#8945, Haddock haskell/haddock#188 The order of signature groups has been corrected upstream. Here we add a test case and update some existing test-cases to reflect this change. We remove grouped signature in test cases that we can (Minimal, BugDeprecated &c) so that the test is as self-contained as possible. - - - - - 708b88b1 by Mateusz Kowalczyk at 2014-04-03T21:16:07+01:00 Enforce strict GHC version in cabal file This stops people with 7.6.3 trying to install 2.15.x which clearly won't work. Unfortunately we shipped 2.14.x without realising this. - - - - - 60334f7c by Mateusz Kowalczyk at 2014-04-03T21:19:24+01:00 Initialise some new PlatformConstants fields - - - - - ea77f668 by Mateusz Kowalczyk at 2014-04-11T16:52:23+01:00 We don't actually want unicode here - - - - - 0b651cae by Mateusz Kowalczyk at 2014-04-11T18:13:30+01:00 Parse identifiers with ^ and ⋆ in them. Fixes haskell/haddock#298. - - - - - e8ad0f5f by Mateusz Kowalczyk at 2014-04-11T18:47:41+01:00 Ignore version string during HTML tests. - - - - - de489089 by Mateusz Kowalczyk at 2014-04-11T18:59:30+01:00 Update CHANGES to follow 2.14.3 - - - - - beb464a9 by Gergő Érdi at 2014-04-13T16:31:10+08:00 remove Origin flag from LHsBindsLR - - - - - cb16f07c by Herbert Valerio Riedel at 2014-04-21T17:16:50+02:00 Replace local `die` by new `System.Exit.die` Starting with GHC 7.10, System.Exit exports the new `die` which is essentially the same as Haddock.Util.die, so this commit changes Haddock.Util.die to be a simple re-export of System.Exit.die. See also https://ghc.haskell.org/trac/ghc/ticket/9016 for more details. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 9b9b23c7 by Mateusz Kowalczyk at 2014-05-03T15:40:11+02:00 Disambiguate ‘die’ in test runners. - - - - - 5d28a2b8 by Mateusz Kowalczyk at 2014-05-05T09:19:49+02:00 Prepare modules for parser split. We have to generalise the Doc (now DocH) slightly to remove the dependency on GHC-supplied type. - - - - - d3967ff3 by Mateusz Kowalczyk at 2014-05-05T11:00:41+02:00 Move parser + parser tests out to own package. We move some types out that are necessary as well and then re-export and specialise them in the core Haddock. Reason for moving out spec tests is that if we're working on the parser, we can simply work on that and we can ignore the rest of Haddock. The downside is that it's a little inconvenient if at the end of the day we want to see that everything passes. - - - - - 522a448d by Mateusz Kowalczyk at 2014-05-05T11:14:47+02:00 Move out Show and Eq instances to Types They are much more useful to the users here. - - - - - 11a6f0f2 by Mateusz Kowalczyk at 2014-05-06T13:50:31+02:00 Remove no longer necessary parser error handling. We can now drop some Maybe tests and even lets us strip an error handling monad away in a few places. - - - - - 6992c924 by Mateusz Kowalczyk at 2014-05-14T02:23:55+02:00 Please the GHC build-system. As I can not figure out how to do this properly, if we're in GHC tree, we treat the library as being the same package. If we're not in the tree, we require that the library be installed separately. - - - - - 7a8ad763 by Mateusz Kowalczyk at 2014-05-14T14:50:25+02:00 Update issue tracker URL - - - - - f616c521 by Mateusz Kowalczyk at 2014-05-14T14:53:32+02:00 Update issue tracker URL for haddock-library - - - - - 66580ded by Gergő Érdi at 2014-05-25T14:24:16+08:00 Accomodate change in PatSyn representation - - - - - 0e43b988 by Mateusz Kowalczyk at 2014-05-29T03:15:29+02:00 Revert "Accomodate change in PatSyn representation" This reverts commit 57aa591362d7c8ba21285fccd6a958629a422091. I am reverting this because I pushed it to master when it was meant to stay on a wip-branch. Sorry Gergo and everyone who had trouble due to this. - - - - - e10d7ec8 by Mateusz Kowalczyk at 2014-05-29T03:24:11+02:00 Revert "Revert "Accomodate change in PatSyn representation"" This reverts commit e110e6e70e40eed06c06676fd2e62578da01d295. Apparently as per GHC commit ac2796e6ddbd54c5762c53e2fcf29f20ea162fd5 this was actually intended. Embarrasing for me. - - - - - 5861aca9 by Mateusz Kowalczyk at 2014-06-05T19:49:27+02:00 Clear up highlighting of identifiers with ‘'’s. - - - - - d7cc420f by Simon Peyton Jones at 2014-06-06T12:41:09+01:00 Follow change in patSynSig - - - - - 938b4fd8 by Mateusz Kowalczyk at 2014-06-12T07:24:29+02:00 Slightly update the readme. Style-sheets are no longer a recent thing, dead links, old maintainers, different formats. - - - - - c7799dea by Mateusz Kowalczyk at 2014-06-18T00:05:56+02:00 Update cabal files Update repository urls, use subdir property for haddock-library and use a separate versioning scheme for haddock-library in preparation for release. - - - - - a2750b6a by Simon Hengel at 2014-06-18T11:01:18+08:00 Compatibility with older versions of base and bytestring - - - - - 009b4b03 by Simon Hengel at 2014-06-18T11:14:01+08:00 Enable travis-ci for haddock-library - - - - - 9b5862eb by Simon Hengel at 2014-06-18T11:14:01+08:00 haddock-library: Do not depend on haddock-library in test suite I think you either add src to hs-source-dirs or the library to build-depends. But doing both does not make sense (AFAICT). - - - - - fb1f3279 by Simon Hengel at 2014-06-18T11:49:05+08:00 haddock-library: Use -Wall for specs - - - - - 649340e1 by Mateusz Kowalczyk at 2014-06-18T06:58:54+02:00 Use Travis with multiple GHC versions When using HEAD, we build haddock-library directly from repository as a dependency (and thanks to --enable-tests, the tests get ran anyway). In all other cases, we manually run the tests on haddock-library only and don't test the main project. - - - - - d7eeeec2 by Mateusz Kowalczyk at 2014-06-18T07:49:04+02:00 Comment improvements + few words in cabal file - - - - - 0f8db914 by Simon Hengel at 2014-06-18T13:52:23+08:00 Use doctest to check examples in documentation - - - - - 2888a8dc by Simon Hengel at 2014-06-18T14:16:48+08:00 Remove doctest dependency (so that we can use haddock-library with doctest) - - - - - 626d5e85 by Mateusz Kowalczyk at 2014-06-18T08:41:25+02:00 Travis tweaks - - - - - 41d4f9cc by Mateusz Kowalczyk at 2014-06-18T08:58:43+02:00 Don't actually forget to install specified GHC. - - - - - c6aa512a by John MacFarlane at 2014-06-18T10:43:57-07:00 Removed reliance on LambdaCase (which breaks build with ghc 7.4). - - - - - b9b93b6f by John MacFarlane at 2014-06-18T10:54:56-07:00 Fixed haddock warnings. - - - - - a41b0ab5 by Mateusz Kowalczyk at 2014-06-19T01:20:10+02:00 Update Travis, bump version - - - - - 864bf62a by Mateusz Kowalczyk at 2014-06-25T10:36:54+02:00 Fix anchors. Closes haskell/haddock#308. - - - - - 53df91bb by Mateusz Kowalczyk at 2014-06-25T15:04:49+02:00 Drop DocParagraph from front of headers I can not remember why they were wrapped in paragraphs to begin with and it seems unnecessary now that I test it. Closes haskell/haddock#307. - - - - - 29b5f2fa by Mateusz Kowalczyk at 2014-06-25T15:17:20+02:00 Don't mangle append order for nested lists. The benefit of this is that the ‘top-level’ element of such lists is properly wrapped in <p> tags so any CSS working with these will be applied properly. It also just makes more sense. Pointed out at jgm/pandoc#1346. - - - - - 05cb6e9c by Mateusz Kowalczyk at 2014-06-25T15:19:45+02:00 Bump haddock-library to 1.1.0 for release - - - - - 70feab15 by Iavor Diatchki at 2014-07-01T03:37:07-07:00 Propagate overloading-mode for instance declarations in haddock (#9242) - - - - - d4ca34a7 by Simon Peyton Jones at 2014-07-14T16:23:15+01:00 Adapt to new definition of HsDecls.TyFamEqn This is a knock-on from the refactoring from Trac haskell/haddock#9063. I'll push the corresponding changes to GHC shortly. - - - - - f91e2276 by Edward Z. Yang at 2014-07-21T08:14:19-07:00 Track GHC PackageId to PackageKey renaming. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - b010f9ef by Edward Z. Yang at 2014-07-25T16:28:46-07:00 Track changes for module reexports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - 8b85f9f9 by Mateusz Kowalczyk at 2014-07-28T13:25:43+02:00 Catch mid-line URLs. Fixes haskell/haddock#314. - - - - - 4c613a78 by Edward Z. Yang at 2014-08-05T03:11:00-07:00 Track type signature change of lookupModuleInAllPackages Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - e80b051c by Edward Z. Yang at 2014-08-05T17:34:26+01:00 If GhcProfiled, also build Haddock profiled. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f9cccd29 by Edward Z. Yang at 2014-08-07T14:23:35+01:00 Ignore TAGS files. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 00b3af52 by Mateusz Kowalczyk at 2014-08-08T04:58:19+02:00 Update to attoparsec-0.12.1.1 There seems to be memory and speed improvement. - - - - - 5457dc71 by Mateusz Kowalczyk at 2014-08-08T18:24:02+02:00 Fix forgotten src - - - - - 3520cb04 by Mateusz Kowalczyk at 2014-08-14T20:19:07+01:00 Bump down the version for master to 2.14.4 - - - - - dc98c21b by Mateusz Kowalczyk at 2014-08-14T20:23:27+01:00 Revert "Track type signature change of lookupModuleInAllPackages" This reverts commit d59fec2c9551b5662a3507c0011e32a09a9c118f. - - - - - 3f2038c0 by Mateusz Kowalczyk at 2014-08-14T20:23:31+01:00 Revert "Track changes for module reexports." This reverts commit b99b57c0df072d12b67816b45eca2a03cb1da96d. - - - - - 56d4e49e by Mateusz Kowalczyk at 2014-08-14T20:23:42+01:00 Revert "Track GHC PackageId to PackageKey renaming." This reverts commit 8ac42d3327473939c013551750425cac191ff0fd. - - - - - 726ea3cb by Mateusz Kowalczyk at 2014-08-14T20:23:47+01:00 Revert "Adapt to new definition of HsDecls.TyFamEqn" This reverts commit cb96b4f1ed0462b4a394b9fda6612c3bea9886bd. - - - - - 61a88ff0 by Mateusz Kowalczyk at 2014-08-14T20:23:52+01:00 Revert "Propagate overloading-mode for instance declarations in haddock (#9242)" This reverts commit 8d20ca8d5a9bee73252ff2035ec45f9c03d0820c. - - - - - a32ba674 by Mateusz Kowalczyk at 2014-08-14T20:26:03+01:00 Revert "Disambiguate ‘die’ in test runners." This reverts commit dba02d6df32534aac5d257f2d28596238d248942. - - - - - f335820f by Mateusz Kowalczyk at 2014-08-14T20:26:09+01:00 Revert "Replace local `die` by new `System.Exit.die`" This reverts commit 08aa509ebac58bfb202ea79c7c41291ec280a1c5. - - - - - 107078e4 by Mateusz Kowalczyk at 2014-08-14T20:27:34+01:00 Merge branch 'reverts' This reverts any changes that were made to have Haddock compile with 7.9. When 7.10 release comes, we can simply re-apply all the patches and any patches that occur on ghc-head branch from now on. This allows us to build master with 7.8.3 - - - - - b44b3871 by Mateusz Kowalczyk at 2014-08-15T02:47:40+01:00 Fix haskell/haddock#313 by doing some list munging. I get rid of the Monoid instance because we weren't satisfying the laws. Convenience of having <> didn't outweigh the shock-factor of having it behave badly. - - - - - e1a62cde by Mateusz Kowalczyk at 2014-08-15T02:52:56+01:00 Stop testing haskell/haddock#188. Because the change is in GHC 7.9 and we now work against 7.8.3, this test no longer makes sense. We revert it until 7.10 becomes the standard version. If anything, there should be a test for this in GHC itself. - - - - - 54e8286d by Mateusz Kowalczyk at 2014-08-15T05:31:57+01:00 Add haskell/haddock#313 to CHANGES - - - - - 9df7ad5d by Simon Hengel at 2014-08-20T11:25:32+08:00 Fix warning - - - - - ee2574d6 by Simon Hengel at 2014-08-20T12:07:01+08:00 Fix travis builds - - - - - 384cf2e6 by Simon Hengel at 2014-08-20T12:14:31+08:00 Require GHC 7.8.3 - - - - - d4779863 by Simon Hengel at 2014-08-22T12:14:16+08:00 Move Haddock API to a separate package - - - - - 80f3e0e1 by Simon Hengel at 2014-08-22T14:57:38+08:00 Bump version to 2.15.0 and add version constraints - - - - - 309a94ce by Simon Hengel at 2014-08-22T15:18:06+08:00 Add deprecated compatibility module - - - - - 4d1e4e3f by Luite Stegeman at 2014-08-22T20:46:45+02:00 export things to allow customizing how the Ghc session is run - - - - - 47884591 by Luite Stegeman at 2014-08-22T20:46:51+02:00 ghc 7.8.2 compatibility - - - - - 5ea94e2c by Luite Stegeman at 2014-08-22T22:08:58+02:00 install dependencies for haddock-api on travis - - - - - 9fb845b2 by Mateusz Kowalczyk at 2014-08-23T10:09:34+01:00 Move sources under haddock-api/src - - - - - 85817dc4 by Mateusz Kowalczyk at 2014-08-23T10:10:48+01:00 Remove compat stuff - - - - - 151c6169 by Niklas Haas at 2014-08-24T08:14:10+02:00 Fix extra whitespace on signatures and update all test cases This was long overdue, now running ./accept.lhs on a clean test from master will not generate a bunch of changes. - - - - - d320e0d2 by Niklas Haas at 2014-08-24T08:14:35+02:00 Omit unnecessary foralls and fix haskell/haddock#315 This also fixes haskell/haddock#86. - - - - - bdafe108 by Mateusz Kowalczyk at 2014-08-24T15:06:46+01:00 Update CHANGES - - - - - fafa6d6e by Mateusz Kowalczyk at 2014-08-24T15:14:23+01:00 Delete few unused/irrelevant/badly-place files. - - - - - 3634923d by Duncan Coutts at 2014-08-27T13:49:31+01:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. - - - - - 539a7e70 by Herbert Valerio Riedel at 2014-08-31T11:36:32+02:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 9e3a0e5b by Mateusz Kowalczyk at 2014-08-31T12:54:43+01:00 Bump version in doc - - - - - 4a177525 by Mateusz Kowalczyk at 2014-08-31T13:01:23+01:00 Bump haddock-library version - - - - - f99c1384 by Mateusz Kowalczyk at 2014-08-31T13:05:25+01:00 Remove references to deleted files - - - - - 5e51a247 by Mateusz Kowalczyk at 2014-08-31T14:18:44+01:00 Make the doc parser not complain - - - - - 2cedb49a by Mateusz Kowalczyk at 2014-09-03T03:33:15+01:00 CONTRIBUTING file for issues - - - - - 88027143 by Mateusz Kowalczyk at 2014-09-04T00:46:59+01:00 Mention --print-missing-docs - - - - - 42f6754f by Alan Zimmerman at 2014-09-05T18:13:24-05:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e712719e by Austin Seipp at 2014-09-09T01:03:27-05:00 Fix import of 'empty' due to AMP. Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 71c29755 by Herbert Valerio Riedel at 2014-09-09T17:35:20+02:00 Bump `base` constraint for AMP - - - - - 0bf9f3ed by Mateusz Kowalczyk at 2014-09-12T19:18:32+01:00 Delete stale ANNOUNCE - - - - - cac89ee6 by Krzysztof Gogolewski at 2014-09-14T17:17:09+02:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 4d683426 by Edward Z. Yang at 2014-09-18T13:38:11-07:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 80697fd5 by Herbert Valerio Riedel at 2014-09-19T00:07:52+02:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. - - - - - c015eb70 by Herbert Valerio Riedel at 2014-09-19T00:10:36+02:00 Revert "Followup changes to addition of -fwarn-context-quantification" This reverts commit 4023817d7c0e46db012ba2eea28022626841ca9b temporarily as the respective feature hasn't landed in GHC HEAD yet, but this commit blocks later commits from being referenced in GHC HEAD. - - - - - 38ded784 by Edward Z. Yang at 2014-09-18T15:32:15-07:00 Revert "Revert "Followup changes to addition of -fwarn-context-quantification"" This reverts commit db14fd8ab4fab43694139bc203808b814eafb2dc. It's in HEAD now. - - - - - f55d59c9 by Herbert Valerio Riedel at 2014-09-26T19:18:28+02:00 Revert "Fix import of 'empty' due to AMP." This reverts commit 0cc5bc85e9fca92ab712b68a2ba2c0dd9d3d79f4 since it turns out we don't need to re-export `empty` from Control.Monad after all. - - - - - 467050f1 by David Feuer at 2014-10-09T20:07:36-04:00 Fix improper lazy IO use Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. - - - - - cc47b699 by Edward Z. Yang at 2014-10-09T21:38:19-07:00 Fix use-after-close lazy IO bug Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. Signed-off-by: David Feuer <David.Feuer at gmail.com> Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 87babcbe by Austin Seipp at 2014-10-20T20:05:27-05:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - ab259516 by Austin Seipp at 2014-10-20T20:07:01-05:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - b918093c by Mateusz Kowalczyk at 2014-10-29T03:59:39+00:00 Experimental support for collapsable headers Closes haskell/haddock#335 - - - - - 849db129 by Mateusz Kowalczyk at 2014-10-29T10:07:26+01:00 Experimental support for collapsable headers (cherry picked from commit e2ed3b9d8dfab09f1b1861dbc8e74f08e137ebcc) - - - - - a4cc4789 by Herbert Valerio Riedel at 2014-10-31T11:08:26+01:00 Collapse user-defined section by default (re haskell/haddock#335) - - - - - 9da1b33e by Yuras Shumovich at 2014-10-31T16:11:04-05:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - c625aefc by Austin Seipp at 2014-10-31T19:34:10-05:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - c7738e5e by Simon Hengel at 2014-11-02T07:25:30+08:00 Remove -fobject-code from .ghci (this slows down reloads on modifications) - - - - - d4a86e95 by Simon Hengel at 2014-11-03T09:26:11+08:00 Get rid of StandaloneDeriving - - - - - a974e311 by Simon Hengel at 2014-11-03T09:26:11+08:00 Derive more instances - - - - - 8aa0c4d7 by Simon Hengel at 2014-11-03T09:27:08+08:00 Remove unused language extensions - - - - - 3052d46a by Simon Hengel at 2014-11-03T09:30:46+08:00 Minor refactoring - - - - - 4281d3cb by Simon Hengel at 2014-11-03T09:30:46+08:00 parser: Try to parse definition lists right before text paragraphs - - - - - 8ba12bf9 by Simon Hengel at 2014-11-03T09:34:19+08:00 Add support for markdown links (closes haskell/haddock#336) - - - - - a2f8d747 by Simon Hengel at 2014-11-03T09:34:19+08:00 Allow markdown links at the beginning of a paragraph - - - - - 53b11207 by Simon Hengel at 2014-11-03T09:34:20+08:00 Update documentation - - - - - 652267c6 by Simon Hengel at 2014-11-03T09:34:20+08:00 Add support for markdown images - - - - - 9d667502 by Simon Hengel at 2014-11-03T09:34:20+08:00 Allow an optional colon after the closing bracket of definition lists This is to disambiguate them from markdown links and will be require with a future release. - - - - - 8167fc32 by Mateusz Kowalczyk at 2014-11-04T01:16:51+00:00 whitespace only - - - - - 3da62981 by Mateusz Kowalczyk at 2014-11-04T01:17:31+00:00 Fix re-exports of built-in type families Fixes haskell/haddock#310 - - - - - edc76b34 by Mateusz Kowalczyk at 2014-11-04T02:54:28+00:00 Turn some uses of error into recoverable warnings This should at the very least not abort when something weird happens. It does feel like we should have a type that carries these errors until the end however as the user might not see them unless they are printed at the end. - - - - - 0a137400 by Mateusz Kowalczyk at 2014-11-04T04:09:44+00:00 Fix warnings - - - - - d068fc21 by Mateusz Kowalczyk at 2014-11-04T21:04:07+00:00 Fix parsing of identifiers written in infix way - - - - - 1a9f2f3d by Simon Hengel at 2014-11-08T11:32:42+08:00 Minor code simplification - - - - - 6475e9b1 by Simon Hengel at 2014-11-08T17:28:33+08:00 newtype-wrap parser monad - - - - - dc1ea105 by Herbert Valerio Riedel at 2014-11-15T11:55:43+01:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - fbb1aca4 by Simon Hengel at 2014-11-16T08:51:38+08:00 State intention rather than implementation details in Haddock comment - - - - - 97851ab2 by Simon Hengel at 2014-11-16T10:20:19+08:00 (wip) Add support for @since (closes haskell/haddock#26) - - - - - 34bcd18e by Gergő Érdi at 2014-11-20T22:35:38+08:00 Update Haddock to new pattern synonym type signature syntax - - - - - 304b7dc3 by Jan Stolarek at 2014-11-20T17:48:43+01:00 Follow changes from haskell/haddock#9812 - - - - - 920f9b03 by Richard Eisenberg at 2014-11-20T16:52:50-05:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 0bfe4e78 by Alan Zimmerman at 2014-11-21T11:23:09-06:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 356ed45a by Thomas Winant at 2014-11-28T16:11:22-06:00 Support for PartialTypeSignatures - - - - - 5dc8f3b1 by Gergő Érdi at 2014-11-29T15:39:09+08:00 For pattern synonyms, render "pattern" as a keyword - - - - - fe704480 by Mateusz Kowalczyk at 2014-12-09T03:38:32+00:00 List new module in cabal file - - - - - b9ad5a29 by Mateusz Kowalczyk at 2014-12-10T00:58:24+00:00 Allow the parser to spit out meta-info Currently we only use it only for ‘since’ annotations but with these patches it should be fairly simple to add new attributes if we wish to. Closes haskell/haddock#26. It seems to work fine but due to 7.10 rush I don't have the chance to do more exhaustive testing right now. The way the meta is output (emphasis at the end of the whole comment) is fairly arbitrary and subject to bikeshedding. Note that this makes test for Bug310 fail due to interface version bump: it can't find the docs for base with this interface version so it fails. There is not much we can do to help this because it tests for ’built-in’ identifier, not something we can provide ourselves. - - - - - 765af0e3 by Mateusz Kowalczyk at 2014-12-10T01:17:19+00:00 Update doctest parts of comments - - - - - 8670272b by jpmoresmau at 2014-12-10T01:35:31+00:00 header could contain several lines Closes haskell/haddock#348 - - - - - 4f9ae4f3 by Mateusz Kowalczyk at 2014-12-12T06:22:31+00:00 Revert "Merge branch 'reverts'" This reverts commit 5c93cc347773c7634321edd5f808d5b55b46301f, reversing changes made to 5b81a9e53894d2ae591ca0c6c96199632d39eb06. Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - e974ac94 by Duncan Coutts at 2014-12-12T06:26:11+00:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. Conflicts: haddock-api/src/Haddock.hs - - - - - 2f3a2365 by Herbert Valerio Riedel at 2014-12-12T06:26:51+00:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 1dbd6390 by Alan Zimmerman at 2014-12-12T06:32:07+00:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - bb6ff1f4 by Mateusz Kowalczyk at 2014-12-12T06:35:07+00:00 Bump ‘base’ constraint Follows the similar commit made on ghc-head branch - - - - - 466fe4ab by Krzysztof Gogolewski at 2014-12-12T06:37:42+00:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 97e080c9 by Edward Z. Yang at 2014-12-12T06:39:35+00:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: haddock-api/src/Haddock/ModuleTree.hs - - - - - 20b2af56 by Herbert Valerio Riedel at 2014-12-12T06:42:50+00:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. Conflicts: haddock-library/src/Documentation/Haddock/Parser.hs - - - - - b3ad269d by Austin Seipp at 2014-12-12T06:44:14+00:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 072df0dd by Austin Seipp at 2014-12-12T06:45:01+00:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - dbb9294a by Herbert Valerio Riedel at 2014-12-12T06:46:17+00:00 Collapse user-defined section by default (re haskell/haddock#335) Conflicts: haddock-api/src/Haddock/Backends/Xhtml/DocMarkup.hs - - - - - f23ab545 by Yuras Shumovich at 2014-12-12T06:46:41+00:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - 753a4b67 by Austin Seipp at 2014-12-12T06:46:51+00:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8954e8f5 by Herbert Valerio Riedel at 2014-12-12T06:50:53+00:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - d2b06d61 by Gergő Érdi at 2014-12-12T07:07:30+00:00 Update Haddock to new pattern synonym type signature syntax Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 1ff02426 by Jan Stolarek at 2014-12-12T07:13:24+00:00 Follow changes from haskell/haddock#9812 Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - 06ad7600 by Richard Eisenberg at 2014-12-12T07:13:43+00:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 8fd2aa8b by Alan Zimmerman at 2014-12-12T07:22:25+00:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Backends/LaTeX.hs haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 95c3db98 by Thomas Winant at 2014-12-12T07:35:49+00:00 Support for PartialTypeSignatures Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs haddock-api/src/Haddock/Interface/Create.hs - - - - - 45494428 by Gergő Érdi at 2014-12-12T07:36:18+00:00 For pattern synonyms, render "pattern" as a keyword - - - - - a237e3eb by Mateusz Kowalczyk at 2014-12-12T12:27:13+00:00 Various fixups and bumps for next release - - - - - 22918bcd by Herbert Valerio Riedel at 2014-12-14T10:11:47+01:00 Remove redundant wild-card pattern match (this would otherwise cause a build-failure with `-Werror`) - - - - - 1d6ce947 by Herbert Valerio Riedel at 2014-12-14T10:17:06+01:00 Treat GHC 7.10 the same as GHC 7.9 ...since the current GHC 7.9 is going to become GHC 7.10 real-soon-now anyway - - - - - f434ea89 by Herbert Valerio Riedel at 2014-12-14T18:26:50+01:00 Fixup ghc.mk (follow-up to 1739375eb23342) This makes the GHC build-system aware of the data-files to be copied into the bindist (as haddock.cabal doesn't list those anymore) - - - - - 6fb839eb by Mateusz Kowalczyk at 2014-12-17T09:28:59+00:00 Only keep one Version instead of blindly appending - - - - - 40645489 by Mateusz Kowalczyk at 2014-12-18T07:09:44+00:00 Fix dependency version - - - - - 8b3b927b by Mateusz Kowalczyk at 2014-12-18T07:14:23+00:00 Print missing docs by default Adds --no-print-missing-docs - - - - - 59666694 by Mateusz Kowalczyk at 2014-12-18T07:21:37+00:00 update changelog - - - - - aa6d168e by Mateusz Kowalczyk at 2014-12-18T07:30:58+00:00 Update docs for @since - - - - - 2d7043ee by Luite Stegeman at 2014-12-19T18:29:35-06:00 hide projectVersion from DynFlags since it clashes with Haddock.Version.projectVersion - - - - - aaa70fc0 by Luite Stegeman at 2014-12-22T15:58:43+01:00 Add missing import for standalone haddock-api package - - - - - 9ce01269 by Herbert Valerio Riedel at 2014-12-22T17:48:45+01:00 Reset ghc-head with master's tree (this is an overwriting git merge of master into ghc-head) - - - - - fcd6fec1 by Herbert Valerio Riedel at 2014-12-22T17:51:52+01:00 Bump versions for ghc-7.11 - - - - - 525ec900 by Mateusz Kowalczyk at 2014-12-23T13:36:24+00:00 travis-ci: test with HEAD - - - - - cbf494b5 by Simon Peyton Jones at 2014-12-23T15:22:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 50e01c99 by Mateusz Kowalczyk at 2014-12-29T15:28:47+00:00 Make travis use 7.10.x - - - - - 475e60b0 by Njagi Mwaniki at 2014-12-29T15:30:44+00:00 Turn the README into GitHub Markdown format. Closes haskell/haddock#354 - - - - - 8cacf48e by Luite Stegeman at 2015-01-05T16:25:37+01:00 bump haddock-api ghc dependency to allow release candidate and first release - - - - - 6ed6cf1f by Simon Peyton Jones at 2015-01-06T16:37:47+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 8b484f33 by Simon Peyton Jones at 2015-01-08T15:50:22+00:00 Track naming change in DataCon - - - - - 23c5c0b5 by Alan Zimmerman at 2015-01-16T10:15:11-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e7a5532c by JP Moresmau at 2015-01-22T17:19:03+00:00 Ignore warnings, install Cabal 1.22 - - - - - 86942c84 by jpmoresmau at 2015-01-22T17:19:04+00:00 solve dataDir ambiguity - - - - - 5ceb743e by jpmoresmau at 2015-01-22T19:17:32+00:00 support GHC 7.10: no Safe-Inferred, Foldable instance - - - - - 6a3b3fb5 by Mateusz Kowalczyk at 2015-01-22T19:32:10+00:00 Update test files Test: a correct behaviour for fields comma-separating values. I'm surprised we had no bug open for this. Maybe it affects how haskell/haddock#301 renders now but I doubt. Operators: Seems GHC is giving us a new order for operators, something must have changed on their side again. cc @haasn , this makes the fixity to the side not match the order on the LHS which is a bit unpleasant. Maybe the fixity can be made to match the GHC order? Bug335: We expand examples by default now. Bug310: Now inferred safe. - - - - - 708f8b2f by jpmoresmau at 2015-01-22T19:36:59+00:00 Links to source location of class instance definitions - - - - - 5cf8a6da by Vincent Berthoux at 2015-01-22T19:59:58+00:00 Filter '\r' from comments due to Windows problems. On Windows this was causing newline to be rendered twice in code blocks. Closes haskell/haddock#359, fixes haskell/haddock#356. - - - - - 1749e6f0 by Mateusz Kowalczyk at 2015-01-22T20:31:27+00:00 Changelog only - - - - - c8145f90 by Mateusz Kowalczyk at 2015-01-22T23:34:05+00:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. - - - - - 14248254 by Mateusz Kowalczyk at 2015-01-22T23:43:18+00:00 Sort out some module import warnings - - - - - d8a38989 by Simon Peyton Jones at 2015-01-23T07:10:16-06:00 Track naming change in DataCon (cherry picked from commit 04cf63d0195837ed52075ed7d2676e71831e8a0b) - - - - - d3ac6ae4 by Alan Zimmerman at 2015-01-23T07:17:19-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit d61bbc75890e4eb0ad508b9c2a27b91f691213e6) - - - - - 4c1ffeb0 by Simon Peyton Jones at 2015-02-10T12:10:33+00:00 Track changes in HsSyn for quasi-quotes - - - - - 775d20f7 by Mateusz Kowalczyk at 2015-03-15T08:11:48+01:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. (cherry picked from commit 8e06728afb0784128ab2df0be7a5d7a191d30ff4) - - - - - f9245e72 by Phil Ruffwind at 2015-03-16T04:32:01-04:00 Prevent Synopsis from using up too much horizontal space When long type signatures occur in the Synopsis, the element is stretched beyond the width of the window. Scrollbars don't appear, so it's impossible to read anything when this happens. - - - - - cd8fa415 by Mateusz Kowalczyk at 2015-03-17T21:59:39+00:00 Update changelog Closes haskell/haddock#151 due to 71170fc77962f10d7d001e3b8bc8b92bfeda99bc - - - - - b5248b47 by Ben Gamari at 2015-03-25T17:12:17+00:00 Make the error encountered when a package can't be found more user-friendly Closes haskell/haddock#369 - - - - - b756b772 by Mateusz Kowalczyk at 2015-03-26T16:31:40+00:00 Remove now redundant imports - - - - - 5ea5e8dd by Mateusz Kowalczyk at 2015-03-26T16:45:52+00:00 Update test to account for \r filtering - - - - - 6539bfb3 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Test for anchor defaulting I delete the old tests because it turns out that: * test runner would never put them in scope of each other even with imports so just one would suffice * test runner actually needed some hacking to keep links so in the end we would end up with no anchors making them useless - - - - - 1a01d950 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Clearly default to variables in out of scope case - - - - - 7943abe8 by Mateusz Kowalczyk at 2015-03-27T01:14:11+00:00 Fix Hoogle display of constructors Fixes haskell/haddock#361 - - - - - 6d6e587e by Mateusz Kowalczyk at 2015-03-27T01:45:18+00:00 Fully qualify names in Hoogle instances output Closes haskell/haddock#263 - - - - - 52dac365 by Mateusz Kowalczyk at 2015-03-27T01:55:01+00:00 Update changelog - - - - - ca5af9a8 by Mateusz Kowalczyk at 2015-03-27T02:43:55+00:00 Output method documentation in Hoogle backend One thing of note is that we no longer preserve grouping of methods and print each method on its own line. We could preserve it if no documentation is present for any methods in the group if someone asks for it though. Fixes haskell/haddock#259 - - - - - a33f0c10 by Mateusz Kowalczyk at 2015-03-27T03:04:21+00:00 Don't print instance safety information in Hoogle Fixes haskell/haddock#168 - - - - - df6c935a by Mateusz Kowalczyk at 2015-03-28T00:11:47+00:00 Post-release version bumps and changelog - - - - - dde8f7c0 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Loosen bounds on haddock-* - - - - - de93bf89 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Expand response files in arguments Closes haskell/haddock#285 - - - - - 1f0b0856 by Zejun Wu at 2015-04-26T16:35:35-07:00 Do not insert anchor for section headings in contents box - - - - - 860439d7 by Simon Peyton Jones at 2015-05-01T09:36:47+01:00 Track change in API of TyCon - - - - - a32f3e5f by Adam Gundry at 2015-05-04T15:32:59+01:00 Track API changes to support empty closed type familes - - - - - 77e98bee by Ben Gamari at 2015-05-06T20:17:08+01:00 Ignore doc/haddock.{ps,pdf} - - - - - 663d0204 by Murray Campbell at 2015-05-11T04:47:37-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8bb0dcf5 by Murray Campbell at 2015-05-11T06:35:06-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit 2380f07c430c525b205ce2eae6dab23c8388d899) - - - - - bad900ea by Adam Bergmark at 2015-05-11T15:29:39+01:00 haddock-library: require GHC >= 7.4 `Data.Monoid.<>` was added in base-4.5/GHC-7.4 Closes haskell/haddock#394 Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - daceff85 by Simon Peyton Jones at 2015-05-13T12:04:21+01:00 Track the new location of setRdrNameSpace - - - - - 1937d1c4 by Alan Zimmerman at 2015-05-25T21:27:15+02:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - ee0fb6c2 by Łukasz Hanuszczak at 2015-05-27T11:51:31+02:00 Create simple method for indentation parsing. - - - - - 7d6fcad5 by Łukasz Hanuszczak at 2015-05-27T21:36:13+02:00 Make nested lists count indentation according to first item. - - - - - d6819398 by Łukasz Hanuszczak at 2015-05-27T22:46:13+02:00 Add simple test case for arbitrary-depth list nesting. - - - - - 2929c54d by Łukasz Hanuszczak at 2015-06-03T02:11:31+02:00 Add arbitrary-indent spec test for parser. - - - - - 9a0a9bb0 by Mateusz Kowalczyk at 2015-06-03T05:25:29+01:00 Update docs with info on new list nesting rule Fixes haskell/haddock#278 through commits from PR haskell/haddock#401 - - - - - 12efc92c by Mateusz Kowalczyk at 2015-06-03T05:29:26+01:00 Update some meta data at the top of the docs - - - - - 765ee49f by Bartosz Nitka at 2015-06-07T08:40:59-07:00 Add some Hacking docs for getting started - - - - - 19aaf851 by Bartosz Nitka at 2015-06-07T08:44:30-07:00 Fix markdown - - - - - 2a90cb70 by Mateusz Kowalczyk at 2015-06-08T15:08:36+01:00 Refine hacking instructions slightly - - - - - 0894da6e by Thomas Winant at 2015-06-08T23:47:28-05:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 10a9bb76 by Emanuel Borsboom at 2015-06-12T02:46:23+01:00 Build executable with '-threaded' (fixes haskell/haddock#399) - - - - - 7696b94f by Mateusz Kowalczyk at 2015-06-12T02:59:19+01:00 Update changelog for -threaded Closes haskell/haddock#400 - - - - - d3c118ec by Bartosz Nitka at 2015-06-12T03:00:58+01:00 Fix haddock: internal error: spliceURL UnhelpfulSpan (#207) Inferred type signatures don't have SrcSpans, so let's use the one from the declaration. I've tested this manually on the test-case from haskell/haddock#207, but I got stuck at trying to run the test-suite. - - - - - b67e843b by Mateusz Kowalczyk at 2015-06-12T03:01:50+01:00 Changelog for haskell/haddock#207 Fixes haskell/haddock#207, closes haskell/haddock#402 - - - - - 841d785e by jpmoresmau at 2015-06-12T16:03:16+01:00 Attach to instance location the name that has the same location file Fixes haskell/haddock#383 - - - - - 98791cae by Mateusz Kowalczyk at 2015-06-12T16:08:27+01:00 Update changelog Closes haskell/haddock#398 - - - - - 7c0b5a87 by Phil Ruffwind at 2015-06-12T13:07:25-04:00 Fix alignment of Source links in instance table in Firefox Due to a Firefox bug [1], a combination of 'whitespace: nowrap' on the parent element with 'float: right' on the inner element can cause the floated element to be displaced downwards for no apparent reason. To work around this, the left side is wrapped in its own <span> and set to 'float: left'. As a precautionary measure to prevent the parent element from collapsing entirely, we also add the classic "clearfix" hack. The latter is not strictly needed but it helps prevent bugs if the layout is altered again in the future. Fixes haskell/haddock#384. Remark: line 159 of src/Haddock/Backends/Xhtml/Layout.hs was indented to prevent confusion over the operator precedence of (<+>) vs (<<). [1]: https://bugzilla.mozilla.org/show_bug.cgi?id=488725 - - - - - cfe86e73 by Mateusz Kowalczyk at 2015-06-14T10:49:01+01:00 Update tests for the CSS changes - - - - - 2d4983c1 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding for Haskell source parser module. - - - - - 29548785 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement function for tagging parsed chunks with source spans. - - - - - 6a5e4074 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement simple string chunking based on HsColour library. - - - - - 6e52291f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create basic token classification method. - - - - - da971a27 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Adapt source span tagging to work with current whitespace handling. - - - - - 4feb5a22 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add record accessors to exports of hyperlinker parser module. - - - - - a8cc4e39 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Make parser module export all types and associated accessors. - - - - - fb8d468f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple HTML renderer for parsed source file. - - - - - 80747822 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for specifying the CSS file path in HTML source renderer. - - - - - 994dc1f5 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix identifier recognition in Haskell source parser. - - - - - b1bd0430 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix comment recognition in Haskell source parser. - - - - - 11db85ae by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for recognizing compiler pragmas in source parser. - - - - - 736c7bd3 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding of module for associating tokens with AST names. - - - - - 7e149bc2 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement utility method for extracting variable identifiers from AST. - - - - - 32eb640a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple mechanism for associating tokens with AST names. - - - - - d4eba5bc by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add dummy support for hyperlinking named tokens. - - - - - 2b76141f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matcher bug causing wrong items being hyperlinked. - - - - - 2d48002e by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Constrain elements exported by hyperlinker modules. - - - - - 9715eec6 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for type token recognition. - - - - - 8fa401cb by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for binding token recognition. - - - - - d062400b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement go-to-definition mechanism for local bindings. - - - - - f4dc229b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement module export- and import-list item hyperlinking. - - - - - c9a46d58 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matching to allow parenthesized operators hyperlinking. - - - - - 03aad95a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix weird hyperlinking of parenthesized operators. - - - - - b4694a7d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for type declaration anchors. - - - - - 7358d2d2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for top-level function declaration anchors. - - - - - dfc24b24 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix external anchors to contain HTML file extension. - - - - - a045926c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Refactor the way AST names are handled within detailed tokens. - - - - - c76049b4 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement hyperlinking of imported module names. - - - - - 2d2a1572 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix parsing of single line comments with broken up newlines. - - - - - 11afdcf2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix bug with improper newline handling. - - - - - 8137f104 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issues with escaped newlines in comments. - - - - - 34759b19 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for parsing C preprocessor macros. - - - - - 09f0f847 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for parser module of source hyperlinker. - - - - - 709a8389 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for AST module of source hyperlinker. - - - - - 4df5c227 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add command line option for generating hyperlinked source. - - - - - 7a755ea2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Extend module interface with rich source token stream field. - - - - - 494f4ab1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement source tokenization during interface creation process. - - - - - 5f21c953 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create hyperlinker module and plug it into the Haddock pipeline. - - - - - 0cc8a216 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for providing custom CSS files for hyperlinked source. - - - - - a32bbdc1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for fancy highlighting upon hovering over identifier. - - - - - d16d642a by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make source hyperlinker generate output in apropriate directory. - - - - - ae12953d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create module with hyperlinker utility functions. - - - - - 6d4952c5 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make external hyperlinks point to locations specified by source URLs. - - - - - 8417555d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Rewrite source generation to fixed links and directory structure. - - - - - ce9cec01 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add basic support for cross-package hyperlink generation. - - - - - 7eaf025c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Disable generating hyperlinks for module references. - - - - - a50bf92e by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make Haddock generate source for all interfaces (also hidden ones). - - - - - f5ae2838 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Prevent source parser from throwing exception when lexing fails. - - - - - db9ffbe0 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement workaround for Chrome highlighting issues. - - - - - 0b6b453b by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate correct anchors for data constructors. - - - - - c86d38bc by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate anchors for record field declarations. - - - - - 063abf7f by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issue with hyperlink highlight styling in Chrome browser. - - - - - 880fc611 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking constructor names in patters. - - - - - c9e89b95 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record patterns. - - - - - 17a11996 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record expressions. - - - - - 0eef932d by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Make hyperlinker respect pretty-printer flag and add documentation. - - - - - f87c1776 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Unexpose hyperlinker modules in Cabal configuration. - - - - - 4c9e2b06 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Setup HSpec framework for Haddock API package. - - - - - 4b20cb30 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add basic tests related to comment parsing. - - - - - 6842e919 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add tests related to parsing basic language constructs. - - - - - 87bffb35 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add simple tests for do-notation parsing. - - - - - e7af1841 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add very simple QuickCheck properties for source parser spec. - - - - - c84efcf1 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Create simple test runner for hyperlinker tests. - - - - - 76b90447 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for basic identifier hyperlinking. - - - - - 0fbf4df6 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for operator hyperlinking. - - - - - 731aa039 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for constructor hyperlinking. - - - - - 995a78a2 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for record expressions and patterns hyperlinking. - - - - - 3566875a by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for literal syntax highlighting. - - - - - 68469a35 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add hyperlinker test runner to .cabal and .gitignore files. - - - - - aa946c93 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Adapt hyperlinker test runner to have the same interface as HTML one. - - - - - ce34da16 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Fix hyperlinker test runner file paths and add pretty-printing option. - - - - - 0d7dd65e by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add reference files for hyperlinker test cases. - - - - - efa4a1e0 by Łukasz Hanuszczak at 2015-07-01T00:47:32+02:00 Make hyperlinker test runner strip local links from generated source. - - - - - 3e96e584 by Łukasz Hanuszczak at 2015-07-01T01:14:59+02:00 Create simple script for accepting hyperlinker test case references. - - - - - 526fe610 by Łukasz Hanuszczak at 2015-07-01T01:16:41+02:00 Re-accept hyperlinker test cases with local references stripped out. - - - - - 892e2cb3 by Łukasz Hanuszczak at 2015-07-01T01:22:09+02:00 Fix bug with diffing wrong files in hyperlinker test runner. - - - - - 9ff46039 by Łukasz Hanuszczak at 2015-07-01T18:04:46+02:00 Remove unused dependencies in Haddock API spec configuration. - - - - - 47969c07 by Łukasz Hanuszczak at 2015-07-01T18:32:19+02:00 Add support for hyperlinking synonyms in patterns. - - - - - a73449e0 by Łukasz Hanuszczak at 2015-07-01T18:33:44+02:00 Create test case for hyperlinking @-patterns. - - - - - c2077ed8 by Łukasz Hanuszczak at 2015-07-01T19:06:04+02:00 Add support for hyperlinking universally quantified type variables. - - - - - 68017342 by Łukasz Hanuszczak at 2015-07-01T19:28:32+02:00 Create hyperlinker test case with quantified type variables. - - - - - 51c01a78 by Łukasz Hanuszczak at 2015-07-01T19:34:22+02:00 Add scoped type variables test for polymorphism test case. - - - - - 13181ae2 by Łukasz Hanuszczak at 2015-07-01T19:56:27+02:00 Add record wildcards test for records hyperlinking test case. - - - - - 991b81dd by Łukasz Hanuszczak at 2015-07-01T21:01:42+02:00 Document some functions in XHTML utlity module. - - - - - 98c8dfe5 by Łukasz Hanuszczak at 2015-07-01T22:25:21+02:00 Make hyperlinker render qualified names as one entity. - - - - - 75e13b9b by Łukasz Hanuszczak at 2015-07-01T22:27:38+02:00 Add qualified name test for identifiers hyperlinking test case. - - - - - de1e143f by Łukasz Hanuszczak at 2015-07-02T12:32:59+02:00 Fix crash happening when hyperlinking type family declarations. - - - - - 7a8fb175 by Łukasz Hanuszczak at 2015-07-02T12:47:03+02:00 Add support for anchoring data family constructor declarations. - - - - - 3b404e49 by Łukasz Hanuszczak at 2015-07-02T13:31:05+02:00 Improve support for hyperlinking type families. - - - - - 59eb7143 by Łukasz Hanuszczak at 2015-07-02T13:33:34+02:00 Add hyperlinker test case for checking type and type family declarations. - - - - - d1cda0c0 by Łukasz Hanuszczak at 2015-07-02T13:41:38+02:00 Fix issue with operators being recognized as preprocessor directives. - - - - - da206c9d by Łukasz Hanuszczak at 2015-07-02T17:18:12+02:00 Fix broken tests for parsing and hyperlinking hash operators. - - - - - 53750d1b by Łukasz Hanuszczak at 2015-07-02T18:53:28+02:00 Add support for anchoring signatures in type class declarations. - - - - - 1fa5bb10 by Łukasz Hanuszczak at 2015-07-02T19:04:47+02:00 Make hyperlinker generate anchors only to top-level value bindings. - - - - - a542305c by Łukasz Hanuszczak at 2015-07-02T19:05:58+02:00 Create hyperlinker test case for type classes. - - - - - b0dd4581 by Łukasz Hanuszczak at 2015-07-04T16:28:26+02:00 Update docs with information about source hyperlinking. - - - - - 9795302a by Łukasz Hanuszczak at 2015-07-04T16:52:15+02:00 Update docs on using `--read-interface` option. - - - - - 9acdc002 by Łukasz Hanuszczak at 2015-07-04T17:15:26+02:00 Remove potentially dangerous record access in hyperlinker AST module. - - - - - fb3ab7be by Łukasz Hanuszczak at 2015-07-04T17:40:10+02:00 Make Haddock generate warnings about potential misuse of hyperlinker. - - - - - a324c504 by Łukasz Hanuszczak at 2015-07-04T17:43:22+02:00 Fix incorrect specification of source style option in doc file. - - - - - 3f01a8e4 by Łukasz Hanuszczak at 2015-07-05T17:06:36+02:00 Refactor source path mapping to use modules as indices. - - - - - ac70f5b1 by Łukasz Hanuszczak at 2015-07-05T17:47:34+02:00 Fix bug where not all module interfaces were added to source mapping. - - - - - f5e57da9 by Łukasz Hanuszczak at 2015-07-06T16:39:57+02:00 Extract main hyperlinker types to separate module. - - - - - 43974905 by Łukasz Hanuszczak at 2015-07-06T16:52:13+02:00 Move source paths types to hyperlinker types module. - - - - - 3e236055 by Łukasz Hanuszczak at 2015-07-06T17:06:19+02:00 Add support for hyperlinking modules in import lists. - - - - - 58233d9f by Łukasz Hanuszczak at 2015-07-06T17:26:49+02:00 Add short documentation for hyperlinker source map type. - - - - - 14da016d by Łukasz Hanuszczak at 2015-07-06T18:07:20+02:00 Fix bug with module name being hyperlinked to `Prelude`. - - - - - 8f79db52 by Łukasz Hanuszczak at 2015-07-06T18:23:47+02:00 Fix problem with spec build in Haddock API configuration. - - - - - e7cc056c by Adam Sandberg Eriksson at 2015-07-07T23:22:21+01:00 StrictData: print correct strictness marks - - - - - e8253ca8 by Mateusz Kowalczyk at 2015-07-07T23:58:28+01:00 Update changelog - - - - - 0aba676b by Mateusz Kowalczyk at 2015-07-07T23:58:33+01:00 Relax upper bound on GHC a bit - - - - - 7a595381 by Mateusz Kowalczyk at 2015-07-07T23:58:52+01:00 Delete trailing whitespace - - - - - 50976d5e by Adam Sandberg Eriksson at 2015-07-08T15:03:04+02:00 StrictData: changes in HsBang type - - - - - 83b045fa by Mateusz Kowalczyk at 2015-07-11T14:35:18+01:00 Fix expansion icon for user-collapsible sections Closes haskell/haddock#412 - - - - - b2a3b0d1 by Mateusz Kowalczyk at 2015-07-22T22:03:21+01:00 Make some version changes after 2.16.1 release - - - - - a8294423 by Ben Gamari at 2015-07-27T13:16:07+02:00 Merge pull request haskell/haddock#422 from adamse/adamse-D1033 Merge for GHC D1033 - - - - - c0173f17 by randen at 2015-07-30T14:49:08-07:00 Break the response file by line termination rather than spaces, since spaces may be within the parameters. This simple approach avoids having the need for any quoting and/or escaping (although a newline char will not be possible in a parameter and has no escape mechanism to allow it). - - - - - 47c0ca14 by Alan Zimmerman at 2015-07-31T10:41:52+02:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - 45a9d770 by Mateusz Kowalczyk at 2015-07-31T09:47:43+01:00 Update changelog - - - - - 347a20a3 by Phil Ruffwind at 2015-08-02T23:15:26+01:00 Avoid JavaScript error during page load in non-frame mode In non-frame mode, parent.window.synopsis refers to the synopsis div rather than the nonexistent frame. Unfortunately, the script wrongly assumes that if it exists it must be a frame, leading to an error where it tries to access the nonexistent attribute 'replace' of an undefined value (synopsis.location). Closes haskell/haddock#406 - - - - - 54ebd519 by Phil Ruffwind at 2015-08-02T23:27:10+01:00 Link to the definitions to themselves Currently, the definitions already have an anchor tag that allows URLs with fragment identifiers to locate them, but it is rather inconvenient to obtain such a URL (so-called "permalink") as it would require finding the a link to the corresponding item in the Synopsis or elsewhere. This commit adds hyperlinks to the definitions themselves, allowing users to obtain links to them easily. To preserve the original aesthetics of the definitions, we alter the color of the link so as to be identical to what it was, except it now has a hover effect indicating that it is clickable. Additionally, the anchor now uses the 'id' attribute instead of the (obsolete) 'name' attribute. Closes haskell/haddock#407 - - - - - 02cc8bb7 by Phil Ruffwind at 2015-08-02T23:28:02+01:00 Fix typo in Haddock.Backends.Xhtml.Layout: divSynposis -> divSynopsis Closes haskell/haddock#408 - - - - - 2eb0a458 by Phil Ruffwind at 2015-08-02T23:30:07+01:00 Fix record field alignment when name is too long Change <dl> to <ul> and use display:table rather than floats to layout the record fields. This avoids bug haskell/haddock#301 that occurs whenever the field name gets too long. Slight aesthetic change: the entire cell of the field's source code is now shaded gray rather than just the area where text exists. Fixes haskell/haddock#301. Closes haskell/haddock#421 - - - - - 7abb3402 by Łukasz Hanuszczak at 2015-08-02T23:32:14+01:00 Add some utility definitions for generating line anchors. - - - - - e0b1d79b by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Make hyperlinked source renderer generate line anchors. - - - - - 24dd4c9f by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Re-accept test cases after adding line anchors for each of them. - - - - - 0372cfcb by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Override source line flags when source hyperlinker is enabled. - - - - - a81bcd07 by Mateusz Kowalczyk at 2015-08-02T23:58:25+01:00 Update tests to follow HTML changes - - - - - d2d7426f by Łukasz Hanuszczak at 2015-08-06T20:54:59+02:00 Fix quote syntax for promoted types. - - - - - 668cf029 by Łukasz Hanuszczak at 2015-08-06T21:12:00+02:00 Apply promoted type quoting to type-level consing. - - - - - 89f8e7c6 by Łukasz Hanuszczak at 2015-08-06T21:17:10+02:00 Extend advanced types test case with other examples. - - - - - 86494bca by Łukasz Hanuszczak at 2015-08-06T21:22:06+02:00 Rename advanced types test case and accept new output. - - - - - dbb7c7c0 by Adam Sandberg Eriksson at 2015-08-09T23:01:05+02:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 2a7704fa by Ben Gamari at 2015-08-10T13:18:05+02:00 Merge pull request haskell/haddock#433 from adamse/split-hsbang HsBang is split into HsSrcBang and HsImplBang - - - - - 891954bc by Thomas Miedema at 2015-08-15T14:51:18+02:00 Follow changes in GHC build system - - - - - b55d32ab by Mateusz Kowalczyk at 2015-08-21T18:06:09+01:00 Make Travis use 7.10.2 - - - - - 97348b51 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Move SYB utilities to standalone module. - - - - - 748ec081 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement `everywhere` transformation in SYB module. - - - - - 011cc543 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement generic transformation constructor. - - - - - b9510db2 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Create simple utility module for type specialization. - - - - - 43229fa6 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Make type of type specialization function more general. - - - - - fd844e90 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Add basic HTML test case for checking instance specialization. - - - - - 6ea0ad04 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Make HTML class instance printer take optional signature argument. - - - - - 65aa41b6 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Refactor instance head type to record instead of a meaningless tuple. - - - - - 3fc3bede by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add expandable method section for each class instance declaration. - - - - - 99ceb107 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Move dummy post-family instances for `DocName` to `Types` module. - - - - - e98f4708 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create convenience functions for type specialization module. - - - - - b947552f by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Hook type specialization logic with HTML pretty-printer. - - - - - dcaa8030 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create stub functions for sugaring specialized types. - - - - - fa84bc65 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement list syntax sugaring logic for specialized types. - - - - - e8b05b07 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement tuple syntax sugaring logic for specialized types. - - - - - 68a2e5bc by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Get rid of code duplication in type specialization module. - - - - - 4721c336 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create scaffolding of a framework for renaming specialized types. - - - - - 271b488d by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in missing cases in specialized type renaming function. - - - - - bfa5f2a4 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Remove code duplication in specialized type renamer. - - - - - ea6bd0e8 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Change state of the type renaming monad. - - - - - 77c5496e by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement simple mechanism for generating new type names. - - - - - 91bfb48b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in stub behaviour with actual environment renaming. - - - - - d244517b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix logic behind binder type renaming. - - - - - f3c5e360 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add SYB-like utility function for performing stateful queries. - - - - - eb3f9154 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create function for retrieving free variables from given type. - - - - - a94561d3 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix compilation error caused by incorrect type signature. - - - - - 8bb707cf by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Move `SetName` class definition to types module. - - - - - 5800b13b by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Hook type renamer with instance method HTML pretty-printer. - - - - - 6a480164 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some test cases for type renamer. - - - - - 839842f7 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized signatures refer to original signature declaration. - - - - - 4880f7c9 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized methods be nicely formatted again. - - - - - ab5a6a2e by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Attach source locations to the specialized class methods. - - - - - 43f8a559 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Extend instances test case to also test multi-name type signatures. - - - - - 59bc751c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix tab-based indentation in instances test case. - - - - - c2126815 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Improve placement of instance methods expander button. - - - - - 0a32e287 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add new data type declaration to instance specialization test case. - - - - - 5281af1f by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make type renamer first try single-letter names as alternatives. - - - - - 7d509475 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix type renamer bug with incorrect names being generated. - - - - - 0f35bf7c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some documentation and refactor type specialization module. - - - - - da1d0803 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix another bug where type renamer was generating incorrect names. - - - - - cd39b5cb by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Refactor type renamer to rebinding and pure renaming phases. - - - - - 850251f4 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix unwitting compilation bug. - - - - - e5e9fc01 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Integrate instance specification type into class instance definition. - - - - - 825b0ea0 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer neccessary instance specification type. - - - - - cdba44eb by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix declaration converter to use more appropriate mode for methods. - - - - - bc45c309 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug with types not being specialized at all. - - - - - 5d8e5d89 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug where instance expander was opening wrong section. - - - - - 6001ee41 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix another type renamer bug where not all names were rebound. - - - - - 5f58ce2a by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix yet another renamer bug where some names were not unique. - - - - - 8265e521 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split instance subsection layout method to top-level declarations. - - - - - e5e66298 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Rearrange layout of instance methods in generated documentation. - - - - - a50b4eea by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer used layout method. - - - - - 2ff36ec2 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach section title to the instance methods block. - - - - - 7ac15300 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Add basic tests for associated types in instances test case. - - - - - db0ea2f9 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach associated types information to instance header. - - - - - 71cad4d5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details section contain associated types information. - - - - - deee2809 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Improve look of rendered associated families in instance details. - - - - - 839d13a5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Introduce alternative type for family declarations. - - - - - d397f03f by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details record use new type for family declarations. - - - - - 2b23fe97 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split printer of type family header to separate functions. - - - - - c3498cdc by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Implement HTML renderer for pseudo-family declarations. - - - - - c12bbb04 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Apply type specializer to associated type family declarations. - - - - - 2fd69ff2 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Create helper method for specializing type signatures. - - - - - 475826e7 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Refactor specializer module to be independent from XHTML backend. - - - - - f00b431c by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add some documentation for instance head specializer. - - - - - a9fef2dc by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix bug with missing space in documentation for associated types. - - - - - 50e29056 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with incorrect instance details sections being expanded. - - - - - e6dfdd03 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by adding instance details section. - - - - - 75565b2a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make section identifier of instance details more GHC-independent. - - - - - add0c23e by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Re-accept tests after applying deterministic section identifiers. - - - - - 878f2534 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make identifier generation also architecture-independent. - - - - - 48be69f8 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with instance expander hijacking type hyperlink click. - - - - - 47830c1f by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Get rid of dreadful hashing function for generating identifiers. - - - - - 956cd5af by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Move `InstOrigin` type declaration to more appropriate module. - - - - - bf672ed3 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by changes related to instance expander. - - - - - 8f2a949a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add examples with type operators to the instances test case. - - - - - 64600a84 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add basic support for sugaring infix type operators. - - - - - 747d71b8 by Łukasz Hanuszczak at 2015-08-21T18:22:34+01:00 Add support for sugaring built-in function syntax. - - - - - d4696ffb by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Remove default methods from Hoogle class output. - - - - - bf0e09d7 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add fixity declarations in Hoogle backend output. - - - - - 90e91a51 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix bug with incorrect fixities being generated in Hoogle backend. - - - - - 48f11d35 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve class type family declarations output in Hoogle backend. - - - - - 661e8e8f by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add missing default family equations in Hoogle output. - - - - - e2d64103 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve formatting of class details output in Hoogle backend. - - - - - 490fc377 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix weird-looking Hoogle output for familyless classes. - - - - - ea115b64 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create script file for new HTML test runner. - - - - - 609913d3 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Set default behaviour if no arguments given. - - - - - dc115f67 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for providing optional arguments for test runner. - - - - - d93ec867 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve output of test runner error messages. - - - - - 0be9fe12 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for executing Haddock process in test runner. - - - - - 4e4d00d9 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add GHC path to test runner configuration. - - - - - d67a2086 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make GHC path a test runner command-line argument. - - - - - c810079a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Extend test runner configuration with Haddock arguments. - - - - - fee18845 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor test runner and create stub functions. - - - - - ff7c161f by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make test runner actually run Haddock executable. - - - - - 391f73e6 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with test runner not producing any output files. - - - - - 81a74e2d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Setup skeleton of framework for running tests. - - - - - f8a79ec4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with modules not being found in global search mode. - - - - - 7e700b4d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make Haddock standard output redirection be more configurable. - - - - - 53b4c17a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Incorporate old, ugly functions for comparing output files. - - - - - 8277c8aa by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor architecture of test runner output checking functions. - - - - - 587bb414 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement actual diffing mechanism. - - - - - 9ed2b5e4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve code style to match popular guidelines. - - - - - 14bffaf8 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make it possible to choose alternative diff tool. - - - - - 5cdfb005 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create stub methods for processing test output as XML documents. - - - - - 7ef8e12e by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement link-stripping logic as simple SYB transformation. - - - - - 8a1fcd4f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Incorporate link stripping to output diffing mechanism. - - - - - 37dba2bc by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement footer-stripping logic. - - - - - 9cd52120 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Add missing dependencies in Cabal configuration file. - - - - - e0f83c6e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix issue with output being printed in incorrect order. - - - - - 0a94fbb0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to run tests without generating diff. - - - - - 76a58c6f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor HTML test suite boilerplate to external package. - - - - - af41e6b0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utilities for storing directory configuration. - - - - - d8f0698f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move IO-dependent config of HTML test suite to test package. - - - - - 17369fa0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Enable all compiler warnings in Haddock test package configuration. - - - - - 9d03b47a by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move Haddock runner of HTML test suite to Haddock test package. - - - - - 4b3483c5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 03754194 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience wrappers to simplify in test entry points. - - - - - 27476ab7 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adjust module visibility and items they export. - - - - - c40002ba by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Remove no longer useful test option. - - - - - 55ab2541 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Change extension of test files used for diffing. - - - - - 136bf4e4 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor and simplify XHTML helper module of test package. - - - - - 69f7e3df by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix typo in link stripper of HTML test suite runner. - - - - - 0c3c1c6b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience script for running specific HTML tests. - - - - - 489e1b05 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement utility functions for conditional link stripping. - - - - - 0f985dc3 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt `hypsrc-test` module to work with new testing framework. - - - - - 927406f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement output accepting mechanism in test package. - - - - - 8545715e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utility function for recursive obtaining directory contents. - - - - - cb70381f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 019599b5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix path handling in test runner. - - - - - 399b985b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to specify ignored files for test output. - - - - - 41b3d93d by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt HTML test runner to use new ignoring functionality. - - - - - e2091c8b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix bug with not all test output files being checked. - - - - - b22134f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Specify ignored files for hyperlinker source test runner. - - - - - 3301dfa1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Copy test runner script for hyperlinked source case. - - - - - d39a6dfa by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner invoking Haddock in incorrect mode. - - - - - f32c8ff3 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix path handling in test module loader. - - - - - 10f94ee9 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make test runner ignore test packages with no modules. - - - - - 5dc4239c by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create test runner entry points for LaTeX test suite. - - - - - 58d1f7cf by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with unnecessary checking old test output. - - - - - c7ce76e1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Re-implement test acceptance functionality. - - - - - 13bbabe8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix warning about no longer needed definition. - - - - - 958a99b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Adapt Cabal configuration to execute LaTeX suite with new runner. - - - - - 550ff663 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Setup test suite for Hoogle backend. - - - - - 3aa969c4 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make Hoogle backend create output directory if needed. - - - - - eb085b02 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Add appropriate .gitignore entry and configure Hoogle test suite. - - - - - a50bf915 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner failing when run on multiple test packages. - - - - - bf5368b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create simple test cases for Hoogle backend. - - - - - 6121ba4b by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create helper function for conversion between XML and XHTML. - - - - - cb516061 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Refactor existing code to use XHTML printer instead of XML one. - - - - - e2de8c82 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Improve portability of test runner scripts. - - - - - 9563e774 by Łukasz Hanuszczak at 2015-08-22T23:43:16+02:00 Remove redundant import statement. - - - - - 55353df1 by Łukasz Hanuszczak at 2015-08-24T23:09:20+02:00 Fix bug with accepting to non-existing directory. - - - - - 00a334ca by Łukasz Hanuszczak at 2015-08-24T23:09:47+02:00 Accept output for Hoogle and LaTeX backends. - - - - - 29191d8b by Łukasz Hanuszczak at 2015-08-24T23:14:18+02:00 Get rid of obsolete testing utilities. - - - - - bbb25db3 by Łukasz Hanuszczak at 2015-08-24T23:18:50+02:00 Update sandbox setup guide to work with Haddock test package. - - - - - cfd45248 by Łukasz Hanuszczak at 2015-08-24T23:51:30+02:00 Make Travis aware of Haddock test package. - - - - - 74185b7a by Łukasz Hanuszczak at 2015-08-25T17:41:59+02:00 Fix test suite failure when used with Stack. - - - - - 18769697 by Łukasz Hanuszczak at 2015-08-25T18:02:09+02:00 Add sample Stack setup to the hacking guide. - - - - - 22715eeb by Łukasz Hanuszczak at 2015-08-25T18:04:47+02:00 Fix Markdown formatting of README file. - - - - - b49ec386 by Łukasz Hanuszczak at 2015-08-25T18:13:36+02:00 Setup Haddock executable path in Travis configuration. - - - - - 5d29eb03 by Eric Seidel at 2015-08-30T09:55:58-07:00 account for changes to ipClass - - - - - f111740a by Ben Gamari at 2015-09-02T13:20:37+02:00 Merge pull request haskell/haddock#443 from bgamari/ghc-head account for changes to ipClass - - - - - a2654bf6 by Jan Stolarek at 2015-09-03T01:32:57+02:00 Follow changes from haskell/haddock#6018 - - - - - 2678bafe by Richard Eisenberg at 2015-09-21T12:00:47-04:00 React to refactoring CoAxiom branch lists. - - - - - ebc56e24 by Edward Z. Yang at 2015-09-21T11:53:46-07:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4a8c4198 by Tamar Christina at 2015-09-27T13:59:08+02:00 Create Process: removed PhaseFailed - - - - - 7e99b790 by Oleg Grenrus at 2015-09-27T20:52:10+03:00 Generate docs for orphan instances - - - - - 32e932e2 by Oleg Grenrus at 2015-09-28T07:21:11+03:00 Have source links for orphan instances - - - - - c2eb9f4f by Oleg Grenrus at 2015-09-28T07:24:58+03:00 Print orphan instances header only if required - - - - - ff96f978 by Oleg Grenrus at 2015-09-28T07:40:54+03:00 Add orphan instances link to contents box - - - - - d72490a6 by Oleg Grenrus at 2015-09-28T16:37:44+03:00 Fix orphan instance collapsing - - - - - 25d3dfe5 by Ben Gamari at 2015-10-03T12:38:09+02:00 Merge pull request haskell/haddock#448 from Mistuke/fix-silent-death-of-runInteractive Remove PhaseFailed - - - - - 1e45e43b by Edward Z. Yang at 2015-10-11T13:10:10-07:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b1370ac1 by Adam Gundry at 2015-10-16T16:26:42+01:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 60bef421 by Simon Peyton Jones at 2015-10-26T12:52:36+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 4c1898ca by Simon Peyton Jones at 2015-10-27T14:24:56+00:00 Track change to PatSyn.patSynSig - - - - - 25108e85 by Simon Peyton Jones at 2015-10-27T17:34:18+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - 693643ac by Ben Gamari at 2015-10-28T14:33:06+01:00 Account for Typeable changes The treatment of type families changed. - - - - - cd7c2221 by Simon Peyton Jones at 2015-10-30T13:03:51+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - 712032cb by Herbert Valerio Riedel at 2015-10-31T11:01:45+01:00 Relax upper bound on `base` to allow base-4.9 - - - - - 0bfa0475 by Simon Peyton Jones at 2015-10-31T19:08:13+00:00 More adaption to wildcard-refactor - - - - - 0a3c0cb7 by Simon Peyton Jones at 2015-10-31T22:14:43+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - c4fd4ec9 by Alan Zimmerman at 2015-11-01T11:16:34+01:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 42cdd882 by Matthew Pickering at 2015-11-06T20:02:16+00:00 Change for IEThingWith - - - - - f368b7be by Ben Gamari at 2015-11-11T11:35:51+01:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - e32965b8 by Simon Peyton Jones at 2015-11-13T12:18:17+00:00 Merge with origin/head - - - - - ebcf795a by Edward Z. Yang at 2015-11-13T21:56:27-08:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4e23989f by Simon Peyton Jones at 2015-11-18T11:32:54+00:00 Wibbles to Haddock - - - - - 2289cd4a by Simon Peyton Jones at 2015-11-20T23:12:49+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor - - - - - 695975a6 by Alan Zimmerman at 2015-11-21T21:16:12+02:00 Update to match GHC wip/T11019 - - - - - bbba21e7 by Simon Peyton Jones at 2015-11-23T13:54:31+00:00 merge with origin/ghc-head - - - - - 3d664258 by Simon Peyton Jones at 2015-11-23T17:17:18+00:00 Wibble - - - - - e64cf586 by Herbert Valerio Riedel at 2015-12-05T00:29:55+01:00 Canonicalise Monad instances - - - - - a2de15a7 by Alan Zimmerman at 2015-12-05T17:33:52+02:00 Matching changes for haskell/haddock#11028 - - - - - cc29a3e4 by Alan Zimmerman at 2015-12-05T19:45:33+02:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 95dd15d1 by Richard Eisenberg at 2015-12-11T17:33:39-06:00 Update for type=kinds - - - - - cb5fd9ed by Herbert Valerio Riedel at 2015-12-14T15:07:30+00:00 Bump versions for ghc-7.11 - - - - - 4f286d96 by Simon Peyton Jones at 2015-12-14T15:10:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 13ea2733 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 098df8b8 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track changes in HsSyn for quasi-quotes - - - - - 716a64de by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track change in API of TyCon - - - - - 77a66bca by Adam Gundry at 2015-12-14T15:10:58+00:00 Track API changes to support empty closed type familes - - - - - f2808305 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track the new location of setRdrNameSpace - - - - - ba8b08a4 by Alan Zimmerman at 2015-12-14T15:10:59+00:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - a4ded87e by Thomas Winant at 2015-12-14T15:14:05+00:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 25c78107 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: print correct strictness marks - - - - - 6cbc41c4 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: changes in HsBang type - - - - - ad46821a by Alan Zimmerman at 2015-12-14T15:14:06+00:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - abc0ae5b by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 3308d06c by Thomas Miedema at 2015-12-14T15:14:07+00:00 Follow changes in GHC build system - - - - - 6c763deb by Eric Seidel at 2015-12-14T15:14:07+00:00 account for changes to ipClass - - - - - ae5b4eac by Jan Stolarek at 2015-12-14T15:17:00+00:00 Follow changes from haskell/haddock#6018 - - - - - ffbc40e0 by Richard Eisenberg at 2015-12-14T15:17:02+00:00 React to refactoring CoAxiom branch lists. - - - - - d1f531e9 by Edward Z. Yang at 2015-12-14T15:17:02+00:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 79f73754 by Tamar Christina at 2015-12-14T15:17:02+00:00 Create Process: removed PhaseFailed - - - - - 3d37bebb by Edward Z. Yang at 2015-12-14T15:20:46+00:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 5f8a9e44 by Adam Gundry at 2015-12-14T15:20:48+00:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 79dda70f by Simon Peyton Jones at 2015-12-14T15:26:02+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 959930fb by Simon Peyton Jones at 2015-12-14T15:37:50+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - e18a8df5 by Simon Peyton Jones at 2015-12-14T15:37:52+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - aa35ab52 by Simon Peyton Jones at 2015-12-14T15:40:18+00:00 More adaption to wildcard-refactor - - - - - 8ceef94b by Simon Peyton Jones at 2015-12-14T15:46:04+00:00 Track change to PatSyn.patSynSig - - - - - cd81e83d by Ben Gamari at 2015-12-14T15:46:06+00:00 Account for Typeable changes The treatment of type families changed. - - - - - 63c9117c by Herbert Valerio Riedel at 2015-12-14T15:46:34+00:00 Relax upper bound on `base` to allow base-4.9 - - - - - a484c613 by Alan Zimmerman at 2015-12-14T15:47:46+00:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 2c26fa51 by Matthew Pickering at 2015-12-14T15:47:47+00:00 Change for IEThingWith - - - - - 593baa0f by Ben Gamari at 2015-12-14T15:49:21+00:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - b6b5ca78 by Edward Z. Yang at 2015-12-14T15:49:54+00:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b5b0e072 by Alan Zimmerman at 2015-12-14T15:54:20+00:00 Update to match GHC wip/T11019 - - - - - 14ddeb68 by Simon Peyton Jones at 2015-12-14T15:54:22+00:00 Wibble - - - - - 10a90ad8 by Herbert Valerio Riedel at 2015-12-14T15:54:22+00:00 Canonicalise Monad instances - - - - - ed68ac50 by Alan Zimmerman at 2015-12-14T15:55:48+00:00 Matching changes for haskell/haddock#11028 - - - - - 3f7e5a2d by Alan Zimmerman at 2015-12-14T15:55:49+00:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 6543a73f by Richard Eisenberg at 2015-12-14T15:59:55+00:00 Update for type=kinds - - - - - 193a5c48 by Matthew Pickering at 2015-12-14T18:17:00+00:00 Changes to compile with 8.0 - - - - - add669ec by Matthew Pickering at 2015-12-14T18:47:12+00:00 Warnings - - - - - 223f3fb4 by Ben Gamari at 2015-12-15T23:45:05+01:00 Update for D1200 - - - - - d058388f by Ben Gamari at 2015-12-16T05:40:17-05:00 Types: Add Outputable[Bndr] DocName instances - - - - - 62ecd7fb by Ben Gamari at 2015-12-16T09:23:09-05:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ddbc187a by Ben Gamari at 2015-12-16T17:54:55+01:00 Update for D1200 - - - - - cec83b52 by Ben Gamari at 2015-12-16T17:54:55+01:00 Types: Add Outputable[Bndr] DocName instances - - - - - d12ecc98 by Ben Gamari at 2015-12-16T17:54:55+01:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ada1616f by Ben Gamari at 2015-12-16T17:54:58+01:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a4f0383d by Ben Gamari at 2015-12-16T23:32:38+01:00 Fix Hyperlinker GHC.con_names is now GHC.getConNames - - - - - a10e6849 by Ben Gamari at 2015-12-20T00:54:11+01:00 Merge remote-tracking branch 'mrhania/testing-framework-improvements' into ghc-head - - - - - f078b4fd by Ben Gamari at 2015-12-20T00:59:51+01:00 test: Compatibility with Cabal 1.23 - - - - - 88a511a9 by Ben Gamari at 2015-12-20T01:14:35+01:00 Merge remote-tracking branch 'phadej/orphans' into ghc-head - - - - - 4e250f36 by Ben Gamari at 2015-12-20T01:14:52+01:00 Add html-test for orphan instances output - - - - - 87fffbad by Alan Zimmerman at 2015-12-20T09:50:42+02:00 Update for GHC trac#11258 Adding locations to RdrName in FieldOcc and AmbiguousFieldOcc - - - - - 6b7e51c9 by idontgetoutmuch at 2015-12-20T21:01:47+00:00 Merge pull request haskell/haddock#1 from haskell/ghc-head Ghc head - - - - - 229c1fb5 by Dominic Steinitz at 2015-12-21T07:19:16+00:00 Handle inline math with mathjax. - - - - - 57902d66 by Dominic Steinitz at 2015-12-21T08:07:11+00:00 Fix the documentation for haddock itself. Change notation and add support for inline math. Allow newlines in display math. Add a command line option for the mathjax url (you might want to use a locally installed version). Rebase tests because of extra url and version change. Respond to (some of the) comments. Fix warnings in InterfaceFile.hs - - - - - 0e69f236 by Herbert Valerio Riedel at 2015-12-21T18:30:43+01:00 Fix-up left-over assumptions of GHC 7.12 into GHC 8.0 - - - - - c67f8444 by Simon Peyton Jones at 2015-12-22T16:26:56+00:00 Follow removal of NamedWildCard from HsType - - - - - da40327a by Ben Gamari at 2015-12-23T14:15:28+01:00 html-test/Operators: Clear up ambiguous types For reasons that aren't entirely clear a class with ambiguous types was accepted by GHC <8.0. I've added a functional dependency to clear up this ambiguity. - - - - - 541b7fa4 by Ben Gamari at 2015-12-23T14:18:51+01:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 0febc947 by Ben Gamari at 2015-12-24T00:30:20+01:00 hoogle-test/AssocTypes: Allow AmbiguousTypes GHC 8.0 complains otherwise - - - - - 25810841 by Ben Gamari at 2015-12-24T00:33:18+01:00 OrphanInstances: Accept test output - - - - - 841987f3 by Ben Gamari at 2015-12-25T11:03:11+01:00 Merge remote-tracking branch 'idontgetoutmuch/ghc-head' into ghc-head - - - - - 358391f0 by Ben Gamari at 2015-12-26T10:44:50+01:00 Add missing import - - - - - a8896885 by Ben Gamari at 2015-12-26T10:45:27+01:00 travis: Use Travis containers - - - - - 85e82134 by Herbert Valerio Riedel at 2015-12-30T17:25:39+01:00 tweak version bounds for GHC-8.1 - - - - - 672a5f75 by randen at 2016-01-01T23:45:25-08:00 The Haddock part for fully gcc-like response files " driver/Main.hs * Moved the response file handling into ResponseFile.hs, updating import section as appropriate. * driver/ResponseFile.hs * New file. In anticipation that maybe some day this could be provided by another library, and to make it possible to unit test, this functionality is pulled out of the Main.hs module, and expanded to support the style/format of response files which gcc uses. * The specification for the format of response files which gcc generates and consumes, seems to be best derived from the gcc code itself (libiberty/argv.c), so that is what has been done here. * This is intended to fix haskell/haddock#379 * driver-test/Main.hs * New file for testing code in the driver source tree * driver-test/ResponseFileSpec.hs * Tests, adapted/adopted from the same gcc code where the escaping/unescaping is from, in the hspec style of unit tests * haddock.cabal * Add the driver-test test-suite. Introduces a new library dependency (upon hspec) for the haddock driver target in the haddock.cabal file, but practically, this should not be a problem as the haddock-api tests already depend on hspec. - - - - - 498781df by Ben Gamari at 2016-01-06T13:41:04+01:00 Version bumps and changelog - - - - - 8451e46a by Ben Gamari at 2016-01-06T13:47:17+01:00 Merge remote-tracking branch 'randen/bug468' - - - - - fb2d9181 by Ben Gamari at 2016-01-06T08:14:42-05:00 Add ResponseFile to OtherModules - - - - - 2cb2d2e3 by Ben Gamari at 2016-01-06T14:35:00+01:00 Merge branch 'master' into ghc-head - - - - - 913477d4 by Eric Seidel at 2016-01-11T14:57:57-08:00 deal with un-wiring of IP class - - - - - c557a4b3 by Alan Zimmerman at 2016-01-15T11:14:35+02:00 Update to match wip/T11430 in GHC - - - - - 3e135093 by Alan Zimmerman at 2016-01-16T18:21:59+01:00 Update to match wip/T11430 in GHC - - - - - c48ef2f9 by Ben Gamari at 2016-01-18T09:50:06+01:00 Merge remote-tracking branch 'gridaphobe/ghc-head' into ghc-head - - - - - 9138a1b0 by Eric Seidel at 2016-01-18T12:50:15+01:00 deal with un-wiring of IP class (cherry picked from commit 17388b0f0029d969d79353be7737eb01c7b8dc5f) - - - - - b48c172e by Joachim Breitner at 2016-01-19T00:11:38+01:00 Make sure --mathjax affects all written HTML files This fixes haskell/haddock#475. - - - - - af61fe63 by Ryan Scott at 2016-02-07T23:25:57+01:00 Render */# instead of TYPE 'Lifted/TYPE 'Unlifted (fixes haskell/haddock#473) - - - - - b6458693 by Ben Gamari at 2016-02-07T23:29:27+01:00 Merge pull request haskell/haddock#477 from haskell/issue-475 Make sure --mathjax affects all written HTML files - - - - - adcc0071 by Ben Gamari at 2016-02-07T23:34:52+01:00 Merge branch 'master' into ghc-head - - - - - d0404e61 by Ben Gamari at 2016-02-08T12:46:49+01:00 doc: Switch to Sphinx - - - - - acb153b3 by Ben Gamari at 2016-02-08T12:46:56+01:00 Document --use-unicode flag - - - - - c20bdf1d by Ben Gamari at 2016-02-08T13:41:24+01:00 Fix GHC and haddock-library dependency bounds - - - - - 8d946801 by Ben Gamari at 2016-02-08T14:54:56+01:00 testsuite: Rework handling of output sanitization Previously un-cleaned artifacts were kept as reference output, making it difficult to tell what has changed and causing spurious changes in the version control history. Here we rework this, cleaning the output during acceptance. To accomplish this it was necessary to move to strict I/O to ensure the reference handle was closed before accept attempts to open the reference file. - - - - - c465705d by Ben Gamari at 2016-02-08T15:36:05+01:00 test: Compare on dump For reasons I don't understand the Xml representations differ despite their textual representations being identical. - - - - - 1ec0227a by Ben Gamari at 2016-02-08T15:36:05+01:00 html-test: Accept test output - - - - - eefbd63a by Ben Gamari at 2016-02-08T15:36:08+01:00 hypsrc-test: Accept test output And fix impredicative Polymorphism testcase. - - - - - d1df4372 by Ben Gamari at 2016-02-08T15:40:44+01:00 Merge branch 'fix-up-testsuite' - - - - - 206a3859 by Phil Ruffwind at 2016-02-08T17:51:21+01:00 Move the permalinks to "#" on the right side Since pull request haskell/haddock#407, the identifiers have been permalinked to themselves, but this makes it difficult to copy the identifier by double-clicking. To work around this usability problem, the permalinks are now placed on the far right adjacent to "Source", indicated by "#". Also, 'namedAnchor' now uses 'id' instead of 'name' (which is obsolete). - - - - - 6c89fa03 by Phil Ruffwind at 2016-02-08T17:54:44+01:00 Update tests for previous commit - - - - - effaa832 by Ben Gamari at 2016-02-08T17:56:17+01:00 Merge branch 'anchors-redux' - - - - - 9a2bec90 by Ben Gamari at 2016-02-08T17:58:40+01:00 Use -fprint-unicode-syntax when --use-unicode is enabled This allows GHC to render `*` as its Unicode representation, among other things. - - - - - 28ecac5b by Ben Gamari at 2016-02-11T18:53:03+01:00 Merge pull request haskell/haddock#480 from bgamari/sphinx Move documentation to ReStructuredText - - - - - 222e5920 by Ryan Scott at 2016-02-11T15:42:42-05:00 Collapse type/data family instances by default - - - - - a80ac03b by Ryan Scott at 2016-02-11T20:17:09-05:00 Ensure expanded family instances render correctly - - - - - 7f985231 by Ben Gamari at 2016-02-12T10:04:22+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - d4eda086 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Various cleanups - - - - - 79bee48d by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show kind signatures for type family variables Addresses GHC haskell/haddock#11588. - - - - - b2981d98 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show 'where ...' after closed type family Seems like we should ideally show the actual equations as well but that seems like it would be a fair amount of work - - - - - cfc0e621 by Ben Gamari at 2016-02-18T22:48:12+01:00 Merge pull request haskell/haddock#483 from bgamari/T11588 Fix GHC haskell/haddock#11588 This fixes GHC haskell/haddock#11588: * Show where ... after closed type families * Show kind signatures on type family type variables - - - - - 256e8a0d by Ben Gamari at 2016-02-18T23:15:39+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 32402036 by Richard Eisenberg at 2016-02-24T13:21:44-05:00 Follow-on changes to support RuntimeRep - - - - - 2b1c572d by Matthew Pickering at 2016-03-04T21:04:02+00:00 Remove unused functions - - - - - eb906f50 by Richard Eisenberg at 2016-03-13T21:17:20+01:00 Follow-on changes to support RuntimeRep (cherry picked from commit ab954263a793d8ced734459d6194a5d89214b66c) - - - - - 8c34ef34 by Richard Eisenberg at 2016-03-14T23:47:23-04:00 Changes due to fix for GHC#11648. - - - - - 0e022014 by Richard Eisenberg at 2016-03-15T14:06:45+01:00 Changes due to fix for GHC#11648. (cherry picked from commit bb994de1ab0c76d1aaf1e39c54158db2526d31f1) - - - - - ed3f78ab by Rik Steenkamp at 2016-04-02T22:20:36+01:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 - - - - - d3210042 by Rik Steenkamp at 2016-04-04T15:43:32+02:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 (cherry picked from commit 3ddcbd6b8e6884bd95028381176eb33bee6896fb) - - - - - 236eec90 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 692ee7e0 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. - - - - - 79619f57 by Ben Gamari at 2016-04-10T23:46:22+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. (cherry picked from commit c6d6a18d85e5e2d9bb5904e6919e8a8d7e31c4c5) - - - - - 3358ccb4 by Ben Gamari at 2016-04-10T23:47:27+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 264949b1 by Ben Gamari at 2016-04-16T17:50:23+02:00 Merge pull request haskell/haddock#482 from RyanGlScott/ghc-head Collapse type/data family instances by default - - - - - 478c483a by Ben Gamari at 2016-04-16T17:51:09+02:00 Merge pull request haskell/haddock#489 from mpickering/unused-functions Remove some unused functions - - - - - c94e55f0 by Ryan Scott at 2016-04-16T17:57:54+02:00 Collapse type/data family instances by default (cherry picked from commit 2da130a8db8f995c119b544fad807533236cf088) - - - - - 31e633d3 by Ryan Scott at 2016-04-16T17:58:06+02:00 Ensure expanded family instances render correctly (cherry picked from commit 1338b5d7c32939de6bbc31af0049477e4f847103) - - - - - 03e4d197 by Matthew Pickering at 2016-04-16T17:58:21+02:00 Remove unused functions (cherry picked from commit b89d1c2456bdb2d4208d94ded56155f7088a37d0) - - - - - ed4116f6 by Ben Gamari at 2016-04-20T10:46:57+02:00 ghc: Install files for needed --hyperlinked-source - - - - - 0be999c4 by Ben Gamari at 2016-04-20T11:37:54+02:00 ghc: Install files for needed --hyperlinked-source (cherry picked from commit 5c82c9fc2d21ddaae4a2470f1c375426968f19c6) - - - - - 4d17544c by Simon Peyton Jones at 2016-04-20T12:42:28+01:00 Track change to HsGroup This relates to a big GHC patch for Trac haskell/haddock#11348 - - - - - 1700a50d by Ben Gamari at 2016-05-01T13:19:27+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. - - - - - 0b7c8125 by Ben Gamari at 2016-05-01T13:21:43+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. (cherry picked from commit 609018dd09c4ffe27f9248b2d8b50f6196cd42b9) - - - - - af115ce0 by Ryan Scott at 2016-05-04T22:15:50-04:00 Render Haddocks for derived instances Currently, one can document top-level instance declarations, but derived instances (both those in `deriving` clauses and standalone `deriving` instances) do not enjoy the same privilege. This makes the necessary changes to the Haddock API to enable rendering Haddock comments for derived instances. This is part of a fix for Trac haskell/haddock#11768. - - - - - 76fa1edc by Ben Gamari at 2016-05-10T18:13:25+02:00 haddock-test: A bit of refactoring for debuggability - - - - - 7d4c4b20 by Ben Gamari at 2016-05-10T18:13:25+02:00 Create: Mark a comment as TODO - - - - - 2a6d0c90 by Ben Gamari at 2016-05-10T18:13:25+02:00 html-test: Update reference output - - - - - bd60913d by Ben Gamari at 2016-05-10T18:13:25+02:00 hypsrc-test: Fix reference file path in cabal file It appears the haddock insists on prefixing --hyperlinked-sourcer output with directory which the source appeared in. - - - - - c1548057 by Ben Gamari at 2016-05-10T18:22:12+02:00 doc: Update extra-source-files in Cabal file - - - - - 41d5bae3 by Ben Gamari at 2016-05-10T18:29:21+02:00 Bump versions - - - - - ca75b779 by Ben Gamari at 2016-05-11T16:03:44+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 4e3cfd62 by Ben Gamari at 2016-05-11T16:06:45+02:00 Merge remote-tracking branch 'RyanGlScott/ghc-head' into ghc-head - - - - - a2379970 by Ben Gamari at 2016-05-11T23:15:11+02:00 doc: Add clean targets - - - - - f275212e by Ben Gamari at 2016-05-11T23:15:14+02:00 doc: Add html as an all-target for ghc Otherwise the html documentation won't be installed for binary-dist. - - - - - 388fc0af by Ben Gamari at 2016-05-12T09:49:12+02:00 Update CHANGES - - - - - bad81ad5 by Ben Gamari at 2016-05-12T09:49:38+02:00 Version bump - - - - - c01688a7 by Ben Gamari at 2016-05-12T10:04:58+02:00 Revert "Version bump" This bump was a bit premature. This reverts commit 7b238d9c5be9b07aa2d10df323b5c7b8d1634dc8. - - - - - 7ed05724 by Ben Gamari at 2016-05-12T10:05:33+02:00 doc: Fix GHC clean rule Apparently GHC's build system doesn't permit wildcards in clean paths. - - - - - 5d9611f4 by Ben Gamari at 2016-05-12T17:43:50+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 653566b2 by Ben Gamari at 2016-05-14T09:57:31+02:00 Version bump to 2.17.2 - - - - - b355c439 by Ben Gamari at 2016-05-14T09:57:51+02:00 doc: Use `$(MAKE)` instead of `make` This is necessary to ensure we use gmake. - - - - - 8a18537d by Ben Gamari at 2016-05-14T10:15:45+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - b3290ef1 by Sebastian Meric de Bellefon at 2016-05-14T11:29:47-04:00 Fix haskell/haddock#303. Hide footer when printing The "Produced by Haddock" footer was overlapping the page's body when printing. This patch hides the footer with a css media rule. - - - - - b4a76f89 by Sebastian Meric de Bellefon at 2016-05-15T02:12:46-04:00 Fix haskell/haddock#280. Parsing of module header The initial newlines were counted as indentation spaces, thus disturbing the parsing of next lines - - - - - ba797c9e by Ben Gamari at 2016-05-16T14:53:46+02:00 doc: Vendorize alabaster Sphinx theme Alabaster is now the default sphinx theme and is a significant improvement over the previous default that it's worthproviding it when unavailable (e.g. Sphinx <1.3). - - - - - c9283e44 by Ben Gamari at 2016-05-16T14:55:17+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 1c9ea198 by Sebastian Méric de Bellefon at 2016-05-16T12:30:40-04:00 Merge pull request haskell/haddock#502 from Helkafen/master Fix haskell/haddock#303. Hide footer when printing - - - - - 33631016 by Ben Gamari at 2016-05-16T19:56:11+02:00 Revert "doc: Vendorize alabaster Sphinx theme" This ended up causes far too many issues to be worthwhile. We'll just have to live with inconsistent haddock documentation. This reverts commit cec21957001143794e71bcd9420283df18e7de40. - - - - - 93317d26 by Ben Gamari at 2016-05-16T19:56:11+02:00 cabal: Fix README path - - - - - c8695b22 by Ben Gamari at 2016-05-16T19:58:51+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 0b50eaaa by Ben Gamari at 2016-05-16T21:02:08+02:00 doc: Use whichever theme sphinx deems appropriate - - - - - 857c1c9c by Ben Gamari at 2016-05-16T21:07:08+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 15fc5637 by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Remove redundant imports - - - - - 132ddc6a by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Better debug output For tracking down haskell/haddock#505 - - - - - 2252a149 by Ben Gamari at 2016-05-22T12:43:59+02:00 Don't consider default class ops when looking for decls When we are looking for an operation within a class we don't care about `default`-type declarations. This was the cause of haskell/haddock#505. - - - - - 4886b2ec by Oleg Grenrus at 2016-05-24T16:19:48+03:00 UnfelpfulSpan line number omitted Kind of resolves https://github.com/haskell/haddock/issues/508 - - - - - a4befd36 by Oleg Grenrus at 2016-05-24T16:53:35+03:00 Change Hyperlinked lexer to know about DataKinds ticks - - - - - f45cb52e by David Feuer at 2016-05-24T18:48:53-04:00 Make parser state a newtype Previously, it was `data` wrapping a `Maybe`, which seems a bit silly. Obviously, this can be changed back if anyone wants to add more fields some day. - - - - - 05013dd7 by Sebastian Meric de Bellefon at 2016-05-24T22:03:55-04:00 remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) Frames are a bit broken, ignored by Hackage, and considered obsolete in general. This patch disables frames generation. The mini_*.html files are still used in the synopsis. - - - - - b8163a88 by Ben Gamari at 2016-05-25T14:44:15+02:00 Merge pull request haskell/haddock#507 from bgamari/T505 Fix haskell/haddock#505 - - - - - ea1b30c6 by Sebastian Meric de Bellefon at 2016-05-25T14:17:00-04:00 Update CHANGES - - - - - eddfc258 by Sebastian Méric de Bellefon at 2016-05-25T15:17:40-04:00 Merge pull request haskell/haddock#514 from Helkafen/frames remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) - - - - - 0e506818 by Alex Biehl at 2016-05-26T12:43:09+02:00 Remove misplaced haddock comment - - - - - a07d28c0 by Ben Gamari at 2016-05-27T11:34:59+02:00 Merge pull request haskell/haddock#515 from alexbiehl/master Remove misplaced haddock comment - - - - - 9001d267 by Ben Gamari at 2016-05-27T11:35:46+02:00 Merge pull request haskell/haddock#513 from treeowl/newtype-since Make parser state a newtype - - - - - 74e1a018 by Sebastian Méric de Bellefon at 2016-05-28T17:28:15-04:00 Merge pull request haskell/haddock#504 from Helkafen/issue-280 Fix haskell/haddock#280. Parsing of module header - - - - - 37557f4f by Alan Zimmerman at 2016-05-29T23:36:50+02:00 Matching changes for haskell/haddock#12105 - - - - - 7d09e5d6 by Sebastian Meric de Bellefon at 2016-06-03T18:07:48-04:00 Version bumps (2.17.3, 1.4.2) - - - - - 85b4bc15 by Sebastian Méric de Bellefon at 2016-06-06T18:35:13-04:00 Merge pull request haskell/haddock#521 from Helkafen/master Version bumps (2.17.3, 1.4.2) - - - - - e95f0dee by Sebastian Meric de Bellefon at 2016-06-06T19:11:35-04:00 publish haddock-test library - - - - - 4de40586 by Sebastian Méric de Bellefon at 2016-06-06T20:26:30-04:00 Merge pull request haskell/haddock#512 from phadej/oleg-fixes Fixes for haskell/haddock#508 and haskell/haddock#510 - - - - - ddfd0789 by Dominic Steinitz at 2016-06-09T09:27:28+01:00 Documentation for LaTeX markup. - - - - - 697a503a by Dominic Steinitz at 2016-06-09T09:33:59+01:00 Fix spelling mistake. - - - - - 246f6fff by Dominic Steinitz at 2016-06-09T09:37:15+01:00 Camel case MathJax. - - - - - 4684bd23 by Dominic Steinitz at 2016-06-09T09:44:53+01:00 Fix math typo and add link. - - - - - f20c037c by Simon Peyton Jones at 2016-06-13T18:26:03+01:00 Follow changes to LHsSigWcType - - - - - 0c58996d by Simon Peyton Jones at 2016-06-15T12:56:01+01:00 Follow GHC re-adding FunTy - - - - - 401b5ca7 by Sebastian Méric de Bellefon at 2016-06-15T12:16:47-04:00 Merge pull request haskell/haddock#525 from idontgetoutmuch/master Documentation for LaTeX markup. - - - - - 92d263b7 by Sebastian Méric de Bellefon at 2016-06-15T12:17:29-04:00 Merge pull request haskell/haddock#522 from Helkafen/master publish haddock-test library - - - - - 0953a2ca by Sebastian Meric de Bellefon at 2016-06-16T00:46:46-04:00 Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 65453e14 by Ben Gamari at 2016-06-16T11:16:32+02:00 ocean: Ensure that synopsis fully covers other content Previously MathJax content was being rendered on top of the synopsis due to ambiguous z-ordering. Here we explicitly give the synopsis block a higher z-index to ensure it is rendered on top. Fixes haskell/haddock#531. - - - - - 68e411a1 by Sebastian Méric de Bellefon at 2016-06-16T23:34:39-04:00 Merge pull request haskell/haddock#534 from bgamari/T531 ocean: Ensure that synopsis fully covers other content - - - - - fad6491b by Sebastian Méric de Bellefon at 2016-06-18T23:57:20-04:00 Merge pull request haskell/haddock#533 from Helkafen/master Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 6108e21b by Sebastian Meric de Bellefon at 2016-06-22T23:08:28-04:00 do not create empty src directory Fix haskell/haddock#536. - - - - - 1ef23823 by Sebastian Méric de Bellefon at 2016-06-24T00:04:48-04:00 Merge pull request haskell/haddock#537 from Helkafen/master do not create empty src directory - - - - - 966baa96 by Omari Norman at 2016-06-29T21:59:34-04:00 Add $ as a special character If this character is not escaped, documentation built with Haddock 2.17.2 will fail. This was not an issue with 2.16 series, which causes builds to fail and there is nothing in the docs or error message giving a clue about why builds that used to succeed now don't. - - - - - 324adb60 by Ben Gamari at 2016-07-01T12:18:51+02:00 GhcUtils: Changes for multi-pattern signatures - - - - - d7571675 by Ömer Sinan Ağacan at 2016-07-21T13:30:47+02:00 Add support for unboxed sums - - - - - 29d0907b by Simon Marlow at 2016-07-22T13:55:48+01:00 Disable NFData instances for GHC types when GHC >= 8.2 - - - - - 702d95f3 by Simon Marlow at 2016-08-02T15:57:30+02:00 Disable NFData instances for GHC types when GHC >= 8.0.2 (cherry picked from commit a3309e797c42dae9bccdeb17ce52fcababbaff8a) - - - - - f4fa79c3 by Ben Gamari at 2016-08-07T13:51:18+02:00 ghc.mk: Don't attempt to install html/frames.html The frames business has been removed. - - - - - 9cd63daf by Ben Gamari at 2016-08-07T13:51:40+02:00 Haddock.Types: More precise version guard This allows haddock to be built with GHC 8.0.2 pre-releases. - - - - - f3d7e03f by Mateusz Kowalczyk at 2016-08-29T20:47:45+01:00 Merge pull request haskell/haddock#538 from massysett/master Add $ as a special character - - - - - 16dbf7fd by Bartosz Nitka at 2016-09-20T19:44:04+01:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 - - - - - 7c31c1ff by Bartosz Nitka at 2016-09-27T17:32:22-04:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 (cherry picked from commit 073d899a8f94ddec698f617a38d3420160a7fd0b) - - - - - 33a90dce by Ryan Scott at 2016-09-30T20:53:41-04:00 Haddock changes for T10598 See https://ghc.haskell.org/trac/ghc/ticket/10598 - - - - - 1f32f7cb by Ben Gamari at 2016-10-13T20:01:26-04:00 Update for refactoring of NameCache - - - - - 1678ff2e by Ben Gamari at 2016-11-15T17:42:48-05:00 Bump upper bound on base - - - - - 9262a7c5 by Alan Zimmerman at 2016-12-07T21:14:28+02:00 Match changes in GHC wip/T3384 branch - - - - - ac0eaf1a by Ben Gamari at 2016-12-09T09:48:41-05:00 haddock-api: Don't use stdcall calling convention on 64-bit Windows See GHC haskell/haddock#12890. - - - - - 04afe4f7 by Alan Zimmerman at 2016-12-12T20:07:21+02:00 Matching changes for GHC wip/T12942 - - - - - e1d1701d by Ben Gamari at 2016-12-13T16:50:41-05:00 Bump base upper bound - - - - - 3d3eacd1 by Alan Zimmerman at 2017-01-10T16:59:38+02:00 HsIParamTy now has a Located name - - - - - 7dbceefd by Kyrill Briantsev at 2017-01-12T13:23:50+03:00 Prevent GHC API from doing optimization passes. - - - - - d48d1e33 by Richard Eisenberg at 2017-01-19T08:41:41-05:00 Upstream changes re levity polymorphism - - - - - 40c25ed6 by Alan Zimmerman at 2017-01-26T15:16:18+02:00 Changes to match haskell/haddock#13163 in GHC - - - - - 504f586d by Ben Gamari at 2017-02-02T17:19:37-05:00 Kill remaining static flags - - - - - 49147ea0 by Justus Adam at 2017-03-02T15:33:34+01:00 Adding MDoc to exports of Documentation.Haddock - - - - - 1cfba9b4 by Justus Adam at 2017-03-09T11:41:44+01:00 Also exposing toInstalledIface - - - - - 53f0c0dd by Ben Gamari at 2017-03-09T13:10:08-05:00 Bump for GHC 8.3 - - - - - c7902d2e by Ben Gamari at 2017-03-09T23:46:02-05:00 Bump for GHC 8.2 - - - - - 4f3a74f8 by Ben Gamari at 2017-03-10T10:21:55-05:00 Merge branch 'ghc-head' - - - - - e273b72f by Richard Eisenberg at 2017-03-14T13:34:04-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - 6ec3d436 by Richard Eisenberg at 2017-03-14T15:15:52-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - eee3cda1 by Ben Gamari at 2017-03-15T15:19:59-04:00 Adapt to EnumSet - - - - - 017cf58e by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 736d6773 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 475f84a0 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 13240b53 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - cd16d529 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 3bea97ae by Edward Z. Yang at 2017-03-15T22:50:46-07:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b2b051ce by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 0f082795 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 20ef63c9 by Edward Z. Yang at 2017-03-22T13:48:12-07:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 45692dcb by Edward Z. Yang at 2017-03-22T14:11:25-07:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4eae8caf by Ben Gamari at 2017-03-23T09:25:33-04:00 Merge commit '240bc38b94ed2d0af27333b23392d03eeb615e82' into HEAD - - - - - 0bbe03f5 by Ben Gamari at 2017-03-23T09:27:28-04:00 haddock-api: Bump bound on GHC - - - - - 65f3ac9d by Alex Biehl at 2017-03-23T17:36:11+01:00 Merge pull request haskell/haddock#581 from JustusAdam/master Adding more exports to Documentation.Haddock - - - - - 37d49a47 by Alex Biehl at 2017-03-23T17:39:14+01:00 Merge pull request haskell/haddock#568 from awson/ghc-head Prevent GHC API from doing optimization passes. - - - - - 1ed047e4 by Brian Huffman at 2017-03-23T17:45:58+01:00 Print any user-supplied kind signatures on type parameters. This applies to type parameters on data, newtype, type, and class declarations, and also to forall-bound type vars in type signatures. - - - - - 1b78ca5c by Brian Huffman at 2017-03-23T17:45:58+01:00 Update test suite to expect kind annotations on type parameters. - - - - - a856b162 by Alex Biehl at 2017-03-23T17:49:32+01:00 Include travis build indication badge - - - - - 8e2e2c56 by Ben Gamari at 2017-03-23T17:20:08-04:00 haddock-api: Bump bound on GHC - - - - - 4d2d9995 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 26d6c150b31bc4580ab17cfd07b6e7f9afe10737) - - - - - a650e20f by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 930cfbe58e2e87f5a4d431d89a3c204934e6e858) - - - - - caa282c2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 2067a2d0afa9cef381d26fb7140b67c62f433fc0) - - - - - 49684884 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 0671abfe7e8ceae2269467a30b77ed9d9656e2cc) - - - - - 4dcfeb1a by Edward Z. Yang at 2017-03-23T17:20:08-04:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 3d77b373dd5807d5d956719dd7c849a11534fa6a) - - - - - 74dd19d2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 94610e9b446324f4231fa6ad4c6ac51e4eba8c0e) - - - - - a9b19a23 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 6cc832dfb1de6088a4abcaae62b25a7e944d55c3) - - - - - d3631064 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit bf3c4d72a0fda38561376eac7eda216158783267) - - - - - ef2148fc by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 07b88c5d4e79b87a319fbb08f8ea01dbb41063c1) - - - - - 2f29518b by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 4eb765ca4205c79539d60b7afa9b7e261a4a49fe) - - - - - 37de047d by Phil Ruffwind at 2017-04-03T11:57:14+02:00 Update MathJax URL MathJax is shutting down their CDN: https://www.mathjax.org/cdn-shutting-down/ They recommend migrating to cdnjs. - - - - - e9d24ba8 by David C. Turner at 2017-04-03T14:58:01+02:00 Add highlight for :target to ocean.css - - - - - 4819a202 by Alex Biehl at 2017-04-11T19:36:48+02:00 Allow base-4.10 for haddock-test - - - - - 44cec69c by Alex Biehl at 2017-04-11T19:39:22+02:00 cabal.project for haddock-api, haddock-library and haddock-test - - - - - 935d0f6a by Alex Biehl at 2017-04-11T19:46:29+02:00 Move dist scripts to scripts/ - - - - - 128e150c by Alex Biehl at 2017-04-11T20:34:46+02:00 Add haddock to cabal.project - - - - - cc8e08ea by Alex Biehl at 2017-04-11T20:35:08+02:00 Read files for hyperlinker eagerly This also exposes Documentation.Haddock.Utf8 - - - - - 152dda78 by Alex Biehl at 2017-04-11T20:37:06+02:00 Explicit import list ofr Control.DeepSeq in Haddock.Interface.Create - - - - - 501b33c4 by Kyrill Briantsev at 2017-04-11T21:01:42+02:00 Prevent GHC API from doing optimization passes. - - - - - c9f3f5ff by Alexander Biehl at 2017-04-12T16:36:53+02:00 Add @alexbiehl as maintaner - - - - - 76f214cc by Alex Biehl at 2017-04-13T07:27:18+02:00 Disable doctest with ghc-8.3 Currently doctest doesn't support ghc-head - - - - - 46b4f5fc by Edward Z. Yang at 2017-04-22T20:38:26-07:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f0555235 by Alex Biehl at 2017-04-25T10:08:48+02:00 Travis: Use ghc-8.2.1 on master - - - - - 966ea348 by Alex Biehl at 2017-04-25T10:32:01+02:00 Travis: Verbose cabal output cf. https://travis-ci.org/haskell/haddock/jobs/225512194#L377 - - - - - 36972bcd by Alex Biehl at 2017-04-25T10:40:43+02:00 Use travis_retry for cabal invocations - - - - - b3a09d2c by Alex Biehl at 2017-04-25T17:02:20+02:00 Use new MathJax URL in html-test 18ed871afb82560d5433b2f53e31b4db9353a74e switched to a new MathJax URL but didn't update the tests. - - - - - ae331e5f by Alexander Biehl at 2017-04-25T17:02:20+02:00 Expand signatures for class declarations - - - - - e573c65a by Alexander Biehl at 2017-04-25T17:02:20+02:00 Hoogle: Correctly print classes with associated data types - - - - - 3fc6be9b by Edward Z. Yang at 2017-04-25T17:02:20+02:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit a0c4790e15a2d3fab8d830eee8fcd639fe6d39c9) - - - - - 6725c060 by Herbert Valerio Riedel at 2017-04-25T17:02:20+02:00 `html-test --accept` deltas to reference samples - - - - - 7d444d61 by Alex Biehl at 2017-04-26T07:13:50+02:00 Remove anything related to obsolete frames mode - - - - - b888972c by Alex Biehl at 2017-04-26T07:49:10+02:00 Cherry-picked remaining commits from haddock-2.17.4-release (#603) * Release haddock/haddock-api 2.17.4 and haddock-library 1.4.3 * Set version bounds for haddock-library NB: This allows GHC 8.2.1's base * Set version bounds for haddock & haddock-api The version bounds support GHC 8.2 * Merge (temporary) v2.17.3 branch into v2.17 This allows us to delete the v2.17.3 branch * Fixup changelog * Pin down haddock-api to a single version as otherwise `haddock`'s package version has no proper meaning * fix source-repo spec for haddock-api - - - - - 4161099b by Alex Biehl at 2017-04-26T11:11:20+02:00 Update changelog to reflect news in HEAD - - - - - eed72cb8 by Alex Biehl at 2017-04-26T11:11:20+02:00 Markdownify changelog - - - - - 5815cea1 by Alex Biehl at 2017-04-26T11:32:33+02:00 Bump to 2.18.0 (#605) - - - - - a551d558 by Alex Biehl at 2017-04-29T22:00:25+02:00 Update attoparsec-0.12.1.1 to attoparsec-0.13.1.0 - - - - - ea164a8d by Sergey Vinokurov at 2017-04-29T22:42:36+02:00 Improve error message - - - - - 2e10122f by Alex Biehl at 2017-04-30T10:07:46+02:00 Correctly remember collapsed sections (#608) Now the "collapsed" cookie stores which sections have changed state instead of which are collapsed. - - - - - f9b24d99 by Alex Biehl at 2017-05-01T17:40:36+02:00 Lazily decode docMap and argMap (#610) These are only used in case of a doc reexport so most of the time decoding these is wasted work. - - - - - 2372af62 by Alex Biehl at 2017-05-01T21:59:23+02:00 Fix Binary instance for InstalledInterface (#611) (#610) introduced lazy decoding for docs from InstalledInterface but forgot to remove the original calls to get and put_ - - - - - 6c633c13 by Nathan Collins at 2017-05-11T11:47:55+02:00 Improve documenation of Haddock markup (#614) * Improve documentation of Haddock markup. - document that Haddock supports inferring types top-level functions with without type signatures, but also explain why using this feature is discouraged. Looks like this feature has been around since version 2.0.0.0 in 2008! - rework the "Module description" section: - move the general discussion of field formatting to the section intro and add examples illustrating the prose for multiline fields. - mention that newlines are preserved in some multiline fields, but not in others (I also noticed that commas in the `Copyright` field are not preserved; I'll look into this bug later). - add a subsection for the module description fields documentation, and put the field keywords in code formatting (double back ticks) instead of double quotes, to be consistent with the typesetting of keywords in other parts of the documentation. - mention that "Named chunks" are not supported in the long-form "Module description" documentation. - fix formatting of keywords in the "Module attributes" section. Perhaps these errors were left over from an automatic translation to ReST from some other format as part of the transition to using Sphinx for Haddock documentation? Also, add a missing reference here; it just said "See ?"! - update footnote about special treatment for re-exporting partially imported modules not being implemented. In my tests it's not implemented at all -- I tried re-exporting both `import B hiding (f)` and `import B (a, b)` style partial imports, and in both cases got the same result as with full imports `import B`: I only get a module reference. * Rework the `Controlling the documentation structure` section. My main goal was to better explain how to use Haddock without an export list, since that's my most common use case, but I hope I improved the section overall: - remove the incomplete `Omitting the export list` section and fold it into the other sections. In particular, summarize the differences between using and not using an export list -- i.e. control over what and in what order is documented -- in the section lead. - add "realistic" examples that use the structure markup, both with and without an export list. I wanted a realistic example here to capture how it can be useful to explain the relationship between a group of functions in a section, in addition to documenting their individual APIs. - make it clear that you can associate documentation chunks with documentation sections when you aren't using an export list, and that doing it in the most obvious way -- i.e. with `-- |`, as you can in the export list -- doesn't work without an export list. It took me a while to figure this out the first time, since the docs didn't explain it at all before. - add a "no export list" example to the section header section. - add more cross references. * Add examples of gotchas for markup in `@...@`. I'm not sure this will help anyone, since I think most people first learn about `@...@` by reading other people's Haddocks, but I've documented the mistakes which I've made and then gotten confused by. * Use consistent Capitalization of Titles. Some titles were in usual title caps, and others only had the first word capitalized. I chose making them all use title caps because that seems to make the cross references look better. - - - - - d4734f45 by Ben Gamari at 2017-05-12T20:36:08+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 93883f37 by Alex Biehl at 2017-05-12T21:02:33+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 5b8f179c by Alex Biehl at 2017-05-13T12:48:10+02:00 Consequently use inClass and notInClass in haddock-library (#617) These allow attoparsec to do some clever lookup optimization - - - - - 77984b82 by Doug Wilson at 2017-05-27T17:37:38+02:00 Don't enable compilation for template haskell (#624) This is no longer necessary after ghc commit 53c78be0aab76a3107c4dacbb1d177afacdd37fa - - - - - 5a3de2b4 by Doug Wilson at 2017-05-27T19:54:53+02:00 Improve Syb code (#621) Specialize.hs and Ast.hs are modified to have their Syb code not recurse into Name or Id in HsSyn types. Specialize.hs is refactored to have fewer calls to Syb functions. Syb.hs has some foldl calls replaced with foldl' calls. There is still a lot of performance on the floor of Ast.hs. The RenamedSource is traversed many times, and lookupBySpan is very inefficient. everywhereBut and lookupBySpan dominate the runtime whenever --hyperlinked-source is passed. - - - - - 3d35a949 by Alex Biehl at 2017-05-30T19:01:37+02:00 Clear fixme comment (#625) - - - - - 2a44bd0c by Alex Biehl at 2017-05-30T19:02:12+02:00 Make haddock-library and haddock-api warning free (#626) - - - - - bd1a0e42 by Alex Biehl at 2017-06-01T10:40:33+02:00 Include `driver-test/*.hs` sdist (#630) This lead to haskell/haddock#629. - - - - - 184a3ab6 by Doug Wilson at 2017-06-03T12:02:08+02:00 Disable pattern match warnings (#628) This disables the pattern match checker which can be very expensive in some cases. The disabled warnings include: * Opt_WarnIncompletePatterns * Opt_WarnIncompleteUniPatterns * Opt_WarnIncompletePatternsRecUpd * Opt_WarnOverlappingPatterns - - - - - 0cf68004 by Alex Biehl at 2017-06-03T20:37:28+02:00 Allow user defined signatures for pattern synonyms (#631) - - - - - 7f51a58a by Alex Biehl at 2017-06-04T11:56:38+02:00 Use NameSet for isExported check (#632) - - - - - d8f044a9 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Match new AST as per GHC wip/new-tree-one-param See https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow - - - - - da1254e3 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Rename extension index tags - - - - - 538c7514 by Christiaan Baaij at 2017-06-09T08:26:43+02:00 Haddock support for bundled pattern synonyms (#627) * Haddock support for bundled pattern synonyms * Add fixities to bundled pattern synonyms * Add bundled pattern synonyms to the synopsis * Store bundled pattern fixities in expItemFixities * Add test for bundled pattern synonyms * Stop threading fixities * Include bundled pattern synonyms for re-exported data types Sadly, fixity information isn't found for re-exported data types * Support for pattern synonyms * Modify tests after haskell/haddock#631 * Test some reexport variations * Also lookup bundled pattern synonyms from `InstalledInterface`s * Check isExported for bundled pattern synonyms * Pattern synonym is exported check * Always look for pattern synonyms in the current module Another overlooked cornercase * Account for types named twice in export lists Also introduce a fast function for nubbing on a `Name` and use it throughout the code base. * correct fixities for reexported pattern synonyms * Fuse concatMap and map * Remove obsolete import * Add pattern synonyms to visible exports * Fix test * Remove corner case - - - - - a050bffd by Doug Wilson at 2017-06-21T09:27:33+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#636) There is some performance improvement. GHC compiler: | version | bytes allocated | cpu_seconds --------------------------------- | before | 56057108648 | 41.0 | after | 51592019560 | 35.1 base: | version | bytes allocated | cpu_seconds --------------------------------- | before | 25174011784 | 14.6 | after | 23712637272 | 13.1 Cabal: | version | bytes allocated | cpu_seconds --------------------------------- | before | 18754966920 | 12.6 | after | 18198208864 | 11.6 - - - - - 5d06b871 by Doug Wilson at 2017-06-22T20:23:29+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#639) * Use new function getNameToInstancesIndex instead of tcRnGetInfo There is some significant performance improvement in the ghc testsuite. haddock.base: -23.3% haddock.Cabal: -16.7% haddock.compiler: -19.8% * Remove unused imports - - - - - b11bb73a by Alex Biehl at 2017-06-23T14:44:41+02:00 Lookup fixities for reexports without subordinates (#642) So we agree that reexported declarations which do not have subordinates (for example top-level functions) shouldn't have gotten fixities reexported according to the current logic. I wondered why for example Prelude.($) which is obviously reexported from GHC.Base has fixities attached (c.f. http://hackage.haskell.org/package/base-4.9.1.0/docs/Prelude.html#v:-36-). The reason is this: In mkMaps we lookup all the subordinates of top-level declarations, of course top-level functions don't have subordinates so for them the resulting list is empty. In haskell/haddock#644 I established the invariant that there won't be any empty lists in the subordinate map. Without the patch from haskell/haddock#642 top-level functions now started to fail reexporting their fixities. - - - - - d2a6dad6 by Alex Biehl at 2017-06-23T18:30:45+02:00 Don't include names with empty subordinates in maps (#644) These are unecessary anyway and just blow up interface size - - - - - 69c2aac4 by Alex Biehl at 2017-06-29T19:54:49+02:00 Make per-argument docs for class methods work again (#648) * Make per-argument docs for class methods work again * Test case - - - - - c9448d54 by Bartosz Nitka at 2017-07-02T12:12:01+02:00 Fix haddock: internal error: links: UnhelpfulSpan (#561) * Fix haddock: internal error: links: UnhelpfulSpan This fixes haskell/haddock#554 for me. I believe this is another fall out of `wildcard-refactor`, like haskell/haddock#549. * Comment to clarify why we take the methods name location - - - - - d4f29eb7 by Alex Biehl at 2017-07-03T19:43:04+02:00 Document record fields when DuplicateRecordFields is enabled (#649) - - - - - 9d6e3423 by Yuji Yamamoto at 2017-07-03T22:37:58+02:00 Fix test failures on Windows (#564) * Ignore .stack-work * Fix for windows: use nul instead of /dev/null * Fix for windows: canonicalize line separator * Also normalize osx line endings - - - - - 7d81e8b3 by Yuji Yamamoto at 2017-07-04T16:13:12+02:00 Avoid errors on non UTF-8 Windows (#566) * Avoid errors on non UTF-8 Windows Problem ==== haddock exits with errors like below: `(1)` ``` haddock: internal error: <stderr>: hPutChar: invalid argument (invalid character) ``` `(2)` ``` haddock: internal error: Language\Haskell\HsColour\Anchors.hs: hGetContents: invalid argument (invalid byte sequence) ``` `(1)` is caused by printing [the "bullet" character](http://www.fileformat.info/info/unicode/char/2022/index.htm) onto stderr. For example, this warning contains it: ``` Language\Haskell\HsColour\ANSI.hs:62:10: warning: [-Wmissing-methods] • No explicit implementation for ‘toEnum’ • In the instance declaration for ‘Enum Highlight’ ``` `(2)` is caused when the input file of `readFile` contains some Unicode characters. In the case above, '⇒' is the cause. Environment ---- OS: Windows 10 haddock: 2.17.3 GHC: 8.0.1 Solution ==== Add `hSetEncoding handle utf8` to avoid the errors. Note ==== - I found the detailed causes by these changes for debugging: - https://github.com/haskell/haddock/commit/8f29edb6b02691c1cf4c479f6c6f3f922b35a55b - https://github.com/haskell/haddock/commit/1dd23bf2065a1e1f2c14d0f4abd847c906b4ecb4 - These errors happen even after executing `chcp 65001` on the console. According to the debug code, `hGetEncoding stderr` returns `CP932` regardless of the console encoding. * Avoid 'internal error: <stderr>: hPutChar: invalid argument (invalid character)' non UTF-8 Windows Better solution for 59411754a6db41d17820733c076e6a72bcdbd82b's (1) - - - - - eded67d2 by Alex Biehl at 2017-07-07T19:17:15+02:00 Remove redudant import warning (#651) - - - - - 05114757 by Alex Biehl at 2017-07-08T00:33:12+02:00 Avoid missing home module warning (#652) * Avoid missing home module warning * Update haddock-library.cabal - - - - - e9cfc902 by Bryn Edwards at 2017-07-17T07:51:20+02:00 Fix haskell/haddock#249 (#655) - - - - - eb02792b by Herbert Valerio Riedel at 2017-07-20T09:09:15+02:00 Fix compilation of lib:haddock-library w/ GHC < 8 - - - - - 9200bfbc by Alex Biehl at 2017-07-20T09:20:38+02:00 Prepare 2.18.1 release (#657) - - - - - 46ddd22c by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Tweak haddock-api.cabal for pending release - - - - - 85e33d29 by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Avoid trivial use of LambdaCase otherwise we can't test w/ e.g. GHC 7.4.2 - - - - - 3afb4bfe by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Refactor .cabal to use sub-lib for vendored lib A practical benefit is that we can control the build-depends and also avoid some recompilation between library and test-suite. - - - - - e56a552e by Herbert Valerio Riedel at 2017-07-20T10:17:48+02:00 haddock-api: add changelog pointing to haddock's changelog This addresses https://github.com/haskell/haddock/issues/638#issuecomment-309283297 - - - - - 2222ff0d by Herbert Valerio Riedel at 2017-07-20T10:19:56+02:00 Drop obsolete/misleading `stability: experimental` This .cabal property has long been considered obsolete - - - - - 9b882905 by Alex Biehl at 2017-07-20T11:25:54+02:00 Beef up haddock description (#658) * Beef up haddock description * Handle empty lines - - - - - bb60e95c by Herbert Valerio Riedel at 2017-07-20T12:08:53+02:00 Import @aisamanra's Haddock cheatsheet from https://github.com/aisamanra/haddock-cheatsheet - - - - - 0761e456 by Herbert Valerio Riedel at 2017-07-20T12:12:55+02:00 Add cheatsheet to haddock.cabal - - - - - 2ece0f0f by Herbert Valerio Riedel at 2017-07-20T12:18:38+02:00 Mention new-build in README - - - - - 947b7865 by Herbert Valerio Riedel at 2017-07-20T12:32:16+02:00 Update README Also improves markup and removes/fixes redundant/obsolete parts [skip ci] - - - - - 785e09ad by Alex Biehl at 2017-07-27T07:28:57+02:00 Bump haddock to 2.18.2, haddock-library to 1.4.5 - - - - - e3ff1ca3 by Alex Biehl at 2017-07-31T20:15:32+02:00 Move `DocMarkup` from haddock-api to haddock-library (#659) * Move `DocMarkup` from haddock-api to haddock-library * Move more markup related functions * Markup module * CHANGELOG - - - - - cda7c20c by Alex Biehl at 2017-07-31T20:35:49+02:00 Fixup haddock - - - - - 583b6812 by Alex Biehl at 2017-07-31T21:20:45+02:00 Changelog for haddock-library - - - - - bac6a0eb by Alex Biehl at 2017-07-31T21:50:24+02:00 Prepare haddock-library-1.4.5 release - - - - - 58ce6877 by Moritz Drexl at 2017-08-05T16:44:40+02:00 Fix renaming after instance signature specializing (#660) * rework rename * Add regression test for Bug 613 * update tests * update changelog - - - - - b8137ec8 by Tim Baumann at 2017-08-06T11:33:38+02:00 Fix: Generate pattern signatures for constructors exported as patterns (#663) * Fix pretty-printing of pattern signatures Pattern synonyms can have up to two contexts, both having a different semantic meaning: The first holds the constraints required to perform the matching, the second contains the constraints provided by a successful pattern match. When the first context is empty but the second is not it is necessary to render the first, empty context. * Generate pattern synonym signatures for ctors exported as patterns This fixes haskell/haddock#653. * Simplify extractPatternSyn It is not necessary to generate the simplest type signature since it will be simplified when pretty-printed. * Add changelog entries for PR haskell/haddock#663 * Fix extractPatternSyn error message - - - - - d037086b by Alex Biehl at 2017-08-06T12:43:25+02:00 Bump haddock-library - - - - - 99d7e792 by Alex Biehl at 2017-08-06T12:44:07+02:00 Bump haddock-library in haddock-api - - - - - 94802a5b by Alex Biehl at 2017-08-06T13:18:02+02:00 Provide --show-interface option to dump interfaces (#645) * WIP: Provide --show-interface option to dump interfaces Like ghcs own --show-iface this flag dumps a binary interface file to stdout in a human (and machine) readable fashion. Currently it uses json as output format. * Fill all the jsonNull stubs * Rework Bifunctor instance of DocH, update changelog and documentation * replace changelog, bring DocMarkupH doc back * Update CHANGES.md * Update CHANGES.md * Move Control.Arrow up It would result in unused import if the Bifunctor instance is not generated. - - - - - c662e476 by Ryan Scott at 2017-08-14T21:00:21-04:00 Adapt to haskell/haddock#14060 - - - - - b891eb73 by Alex Biehl at 2017-08-16T08:24:48+02:00 Bifoldable and Bitraversable for DocH and MetaDoc - - - - - 021bb56c by Alex Biehl at 2017-08-16T09:06:40+02:00 Refactoring: Make doc renaming monadic This allows us to later throw warnings if can't find an identifier - - - - - 39fbf022 by Alex Biehl at 2017-08-19T20:35:27+02:00 Hyperlinker: Avoid linear lookup in enrichToken (#669) * Make Span strict in Position * Hyperlinker: Use a proper map to enrich tokens - - - - - e13baedd by Alex Biehl at 2017-08-21T20:05:42+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 27dd6e87 by Alex Biehl at 2017-08-21T22:06:35+02:00 Drop Avails from export list - - - - - 86b247e2 by Alex Biehl at 2017-08-22T08:44:22+02:00 Bump ghc version for haddock-api tests - - - - - d4607ca0 by Alex Biehl at 2017-08-22T08:45:17+02:00 Revert "Drop Avails from export list" This reverts commit a850ba86d88a4fb9c0bd175453a2580e544e3def. - - - - - c9c54c30 by Alex Biehl at 2017-08-22T09:26:01+02:00 IntefaceFile version - - - - - a85b7c02 by Ben Gamari at 2017-08-22T09:29:52-04:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 34e976f5 by Ben Gamari at 2017-08-22T17:40:06+02:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 577abf06 by Ryan Scott at 2017-08-23T14:47:29-04:00 Update for haskell/haddock#14131 - - - - - da68fc55 by Florian Eggenhofer at 2017-08-27T18:21:56+02:00 Generate an index for package content search (#662) Generate an index for package content search - - - - - 39e62302 by Alex Biehl at 2017-08-27T18:50:16+02:00 Content search for haddock html doc - - - - - 91fd6fb2 by Alex Biehl at 2017-08-28T18:39:58+02:00 Fix tests for content search - - - - - b4a3798a by Alex Biehl at 2017-08-28T18:44:08+02:00 Add search button to #page-menu - - - - - 25a7ca65 by Alex Biehl at 2017-08-28T18:47:43+02:00 Load javascript below the fold - - - - - 8d323c1a by Alex Biehl at 2017-08-28T18:49:22+02:00 Accept tests - - - - - c5dac557 by Alex Biehl at 2017-08-28T19:14:55+02:00 Content search css - - - - - 89a5af57 by Paolo Veronelli at 2017-08-29T07:42:13+02:00 Removed `nowrap` for interface method sigs (#674) with nowrap the interfaces method sigs would expand at libitum - - - - - a505f6f7 by Alex Biehl at 2017-08-29T08:05:33+02:00 Include subordinates in content index - - - - - 4bb698c4 by Alexander Biehl at 2017-08-29T11:40:19+02:00 QuickNav: Make docbase configurable - - - - - c783bf44 by Alexander Biehl at 2017-08-29T11:48:36+02:00 QuickNav: Also use baseUrl for doc-index.json request - - - - - 47017510 by Alex Biehl at 2017-08-29T17:56:47+02:00 Fix test fallout (again) - - - - - 924fc318 by Alex Biehl at 2017-08-30T09:24:56+02:00 Write meta.json when generating html output (#676) - - - - - 717dea52 by Alex Biehl at 2017-09-01T09:20:34+02:00 Use relative URL when no docBaseUrl given - - - - - e5d85f3b by Alex Biehl at 2017-09-01T09:35:19+02:00 Add missing js files to data-files (#677) - - - - - 95b9231a by Alex Biehl at 2017-09-01T11:01:36+02:00 Rename "Search" tab to "Quick Jump" - - - - - da0ead0b by Alex Biehl at 2017-09-01T13:03:49+02:00 Make trigger link configurable (#678) QuickNav: Configurable show/hide trigger - - - - - de7da594 by Ben Gamari at 2017-09-05T06:49:55-04:00 Account for "Remember the AvailInfo for each IE" As of GHC commit f609374a55bdcf3b79f3a299104767aae2ffbf21 GHC retains the AvailInfo associated with each IE. @alexbiehl has a patch making proper use of this change, but this is just to keep things building. - - - - - b05cd3b3 by Ben Gamari at 2017-09-14T07:55:07-04:00 Bump upper bound on base - - - - - 79db899e by Herbert Valerio Riedel at 2017-09-21T23:27:52+02:00 Make compatible with Prelude.<> export in GHC 8.4/base-4.11 - - - - - 3405dd52 by Tim Baumann at 2017-09-23T22:02:01+02:00 Add compile step that bundles and compresses JS files (#684) * Add compile step that bundles and compresses JS files Also, manage dependencies on third-party JS libraries using NPM. * Compile JS from TypeScript * Enable 'noImplicitAny' in TypeScript * QuickJump: use JSX syntax * Generate source maps from TypeScript for easier debugging * TypeScript: more accurate type * Separate quick jump css file from ocean theme - - - - - df0b5742 by Alex Biehl at 2017-09-29T21:15:40+02:00 Bump base for haddock-library and haddock-test - - - - - 62b12ea0 by Merijn Verstraaten at 2017-10-04T16:03:13+02:00 Inhibit output of coverage information for hidden modules. (#687) * Inhibit output of coverage information for hidden modules. * Add changelog entry. - - - - - 8daf8bc1 by Alexander Biehl at 2017-10-05T11:27:05+02:00 Don't use subMap in attachInstances - - - - - ad75114e by Alexander Biehl at 2017-10-05T11:27:58+02:00 Revert "Don't use subMap in attachInstances" This reverts commit 3adf5bcb1a6c5326ab33dc77b4aa229a91d91ce9. - - - - - 7d4aa02f by Alex Biehl at 2017-10-08T15:32:28+02:00 Precise Haddock: Use Avails for export resolution (#688) * Use Avails for export resolution * Support reexported modules * Factor out availExportItem * Use avails for fullModuleExports * Don't use subMap in attachInstances * lookupDocs without subMap * Completely remove subMap * Only calculate unqualified modules when explicit export list is given * Refactor * Refine comment * return * Fix * Refactoring * Split avail if declaration is not exported itself * Move avail splitting - - - - - b9b4faa8 by Alex Biehl at 2017-10-08T19:38:21+02:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 43325295 by Alex Biehl at 2017-10-08T20:18:46+02:00 Fix merge fallout - - - - - c6423cc0 by Alex Biehl at 2017-10-08T20:36:12+02:00 Copy QuickJump files over - - - - - 1db587c3 by Tim Baumann at 2017-10-09T18:33:09+02:00 Use <details> element for collapsibles (#690) * Remove unnecessary call to 'collapseSection' The call is unnecessary since there is no corresponding toggle for hiding the section of orphan instances. * Use <details> for collapsibles This makes them work even when JS is disabled. Closes haskell/haddock#560. - - - - - 1b54c64b by Tim Baumann at 2017-10-10T09:50:59+02:00 Quick Jump: Show error when loading 'doc-index.json' failed (#691) - - - - - 910f716d by Veronika Romashkina at 2017-10-24T07:36:20+02:00 Fix tiny typo in docs (#693) - - - - - b21de7e5 by Ryan Scott at 2017-10-24T13:07:15+02:00 Overhaul Haddock's rendering of kind signatures (#681) * Overhaul Haddock's rendering of kind signatures * Strip off kind signatures when specializing As an added bonus, this lets us remove an ugly hack specifically for `(->)`. Yay! * Update due to 0390e4a0f61e37bd1dcc24a36d499e92f2561b67 * @alexbiehl's suggestions * Import injectiveVarsOfBinder from GHC - - - - - 6704405c by Ryan Scott at 2017-10-28T07:10:27+02:00 Fix Haddock rendering of kind-indexed data family instances (#694) - - - - - 470f6b9c by Alex Biehl at 2017-10-30T08:45:51+01:00 Add QuickJump version to meta.json (#696) - - - - - b89eccdf by Alex Biehl at 2017-10-30T10:15:49+01:00 Put Quickjump behind --quickjump flag (#697) - - - - - 3095fb58 by Alex Biehl at 2017-10-30T19:09:06+01:00 Add build command to package.json - - - - - f223fda9 by Alex Biehl at 2017-10-30T19:10:39+01:00 Decrease threshold for fuzzy matching - - - - - 80245dda by Edward Z. Yang at 2017-10-31T20:35:05+01:00 Supported reexported-modules via --reexport flag. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 7e389742 by Alex Biehl at 2017-10-31T20:37:56+01:00 Correct missing title in changelog - - - - - 1a2a1c03 by Alex Biehl at 2017-10-31T20:59:07+01:00 Copy quickjump.css for nicer error messages - - - - - db234bb9 by Alex Biehl at 2017-10-31T21:31:18+01:00 Reexported modules: Report warnings if argument cannot be parsed or ... module cannot be found - - - - - eea8a205 by Carlo Hamalainen at 2017-10-31T21:43:14+01:00 More general type for nameCacheFromGhc. (#539) - - - - - 580eb42a by Alex Biehl at 2017-10-31T21:46:52+01:00 Remote tab - - - - - 0e599498 by Alex Biehl at 2017-10-31T21:48:55+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 7b8539bb by Alex Biehl at 2017-10-31T22:28:34+01:00 fullModuleContents traverses exports in declaration order - - - - - 0c91fbf2 by Alex Biehl at 2017-10-31T22:32:31+01:00 Remove excessive use of list comprehensions - - - - - f7356e02 by Alex Biehl at 2017-11-01T19:11:03+01:00 Make better use of AvailInfo - - - - - f3e512d5 by Alex Biehl at 2017-11-02T12:16:22+01:00 Always return documentation for exported subordinates ... event if they have no documentation (e.g. noDocForDecl) By using the information in the AvailInfo we don't need additional export checks. - - - - - 7cf58898 by Alan Zimmerman at 2017-11-07T08:28:03+02:00 Match changes for Trees that Grow in GHC - - - - - e5105a41 by Alan Zimmerman at 2017-11-08T17:21:58+02:00 Match Trees That Grow - - - - - 55178266 by Alan Zimmerman at 2017-11-11T22:20:31+02:00 Match Trees that Grow in GHC for HsExpr - - - - - 2082ab02 by Ryan Scott at 2017-11-14T15:27:03+01:00 Actually render infix type operators as infix (#703) * Actually render infix type operators as infix * Account for things like `(f :*: g) p`, too - - - - - c52ab7d0 by Alan Zimmerman at 2017-11-14T23:14:26+02:00 Clean up use of PlaceHolder, to match TTG - - - - - 81cc9851 by Moritz Angermann at 2017-11-20T07:52:49+01:00 Declare use of `Paths_haddock` module in other-modules (#705) This was detected by `-Wmissing-home-modules` - - - - - f9d27598 by Moritz Angermann at 2017-11-20T12:47:34+01:00 Drop Paths_haddock from ghc.mk (#707) With haskell/haddock#705 and haskell/haddock#706, the custom addition should not be necessary any more. # Conflicts: # ghc.mk - - - - - f34818dc by Moritz Angermann at 2017-11-20T12:47:59+01:00 Add autogen-modules (#706) > Packages using 'cabal-version: >= 1.25' and the autogenerated module Paths_* must include it also on the 'autogen-modules' field besides 'exposed-modules' and 'other-modules'. This specifies that the module does not come with the package and is generated on setup. Modules built with a custom Setup.hs script also go here to ensure that commands like sdist don't fail. # Conflicts: # haddock.cabal - - - - - bb43a0aa by Ben Gamari at 2017-11-21T15:50:12-05:00 Revert "Clean up use of PlaceHolder, to match TTG" This reverts commit 134a7bb054ea730b13c8629a76232d73e3ace049. - - - - - af9ebb2b by Ben Gamari at 2017-11-21T15:50:14-05:00 Revert "Match Trees that Grow in GHC for HsExpr" This reverts commit 9f054dc365379c66668de6719840918190ae6e44. - - - - - 5d35c3af by Ben Gamari at 2017-11-21T15:50:15-05:00 Revert "Match Trees That Grow" This reverts commit 73a26af844ac50b8bec39de11d64452a6286b00c. - - - - - 99a8e43b by Ben Gamari at 2017-11-21T16:36:06-05:00 Revert "Match changes for Trees that Grow in GHC" This reverts commit 01eeeb048acd2dd05ff6471ae148a97cf0720547. - - - - - c4d650c2 by Ben Gamari at 2017-12-04T15:06:07-05:00 Bump GHC version - - - - - 027b2274 by Ben Gamari at 2017-12-04T17:06:31-05:00 Bump GHC bound to 8.4.* - - - - - 58eaf755 by Alex Biehl at 2017-12-06T15:44:24+01:00 Update changelog - - - - - d68f5584 by Simon Peyton Jones at 2017-12-07T14:39:56+00:00 Track changes to follow Trac haskell/haddock#14529 This tracks the refactoring of HsDecl.ConDecl. - - - - - dc519d6b by Alec Theriault at 2018-01-06T08:20:43-08:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - 8285118c by Alec Theriault at 2018-01-13T12:12:37+01:00 Constructor and pattern synonym argument docs (#709) * Support Haddocks on constructor arguments This is in conjunction with https://phabricator.haskell.org/D4094. Adds support for rendering Haddock's on (non-record) constructor arguments, both for regular and GADT constructors. * Support haddocks on pattern synonym arguments It appears that GHC already parsed these - we just weren't using them. In the process of doing this, I tried to deduplicate some code around handling patterns. * Update the markup guide Add some information about the new support for commenting constructor arguments, and mention pattern synonyms and GADT-style constructors. * Overhaul LaTeX support for data/pattern decls This includes at least * fixing several bugs that resulted in invalid LaTeX * fixing GADT data declaration headers * overhaul handling of record fields * overhaul handling of GADT constructors * overhaul handling of bundled patterns * add support for constructor argument docs * Support GADT record constructors This means changes what existing HTML docs look like. As for LaTeX, looks like GADT records were never even supported. Now they are. * Clean up code/comments Made code/comments consistent between the LaTeX and XHTML backend when possible. * Update changelog * Patch post-rebase regressions * Another post-rebase change We want return values to be documentable on record GADT constructors. - - - - - ca4fabb4 by Alec Theriault at 2018-01-15T17:12:18-08:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 4c472fea by Ryan Scott at 2018-01-19T10:44:02+01:00 Fix haskell/haddock#732 (#733) - - - - - bff14dbd by Alex Biehl at 2018-01-19T15:33:30+01:00 extractDecl: Extract associated types correctly (#736) - - - - - a2a94a73 by Alex Biehl at 2018-01-19T15:34:40+01:00 extractDecl: Extract associated types correctly (#736) - - - - - 26df93dc by Alex Biehl at 2018-01-20T10:18:22+01:00 haddock-api: bump ghc to ^>= 8.4 - - - - - f65aeb1d by Alex Biehl at 2018-01-20T19:18:20+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - 0e721b97 by Alex Biehl at 2018-01-20T19:20:19+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - cb6234f6 by Ben Gamari at 2018-01-26T13:40:55-05:00 Merge remote-tracking branch 'harpocrates/fix/missing-orphan-instances' into ghc-head - - - - - 0fc28554 by Alec Theriault at 2018-02-01T14:58:18+01:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - b9123772 by Alec Theriault at 2018-02-01T14:58:18+01:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 0c12e274 by Ryan Scott at 2018-02-01T14:58:18+01:00 Fix haskell/haddock#548 by rendering datatype kinds more carefully (#702) - - - - - 8876d20b by Alec Theriault at 2018-02-01T14:58:18+01:00 Use the GHC lexer for the Hyperlinker backend (#714) * Start changing to use GHC lexer * better cpp * Change SrcSpan to RealSrcSpan * Remove error * Try to stop too many open files * wip * wip * Revert "wip" This reverts commit b605510a195f26315e3d8ca90e6d95a6737553e1. Conflicts: haddock-api/haddock-api.cabal haddock-api/src/Haddock/Interface.hs * Remove pointless 'caching' * Use dlist rather than lists when finding vars * Use a map rather than list * Delete bogus comment * Rebase followup Things now run using the GHC lexer. There are still - stray debug statements - unnecessary changes w.r.t. master * Cleaned up differences w.r.t. current Haddock HEAD Things are looking good. quasiquotes in particular look beautiful: the TH ones (with Haskell source inside) colour/link their contents too! Haven't yet begun to check for possible performance problems. * Support CPP and top-level pragmas The support for these is hackier - but no more hacky than the existing support. * Tests pass, CPP is better recognized The tests were in some cases altered: I consider the new output to be more correct than the old one.... * Fix shrinking of source without tabs in test * Replace 'Position'/'Span' with GHC counterparts Replaces 'Position' -> 'GHC.RealSrcLoc' and 'Span' -> 'GHC.RealSrcSpan'. * Nits * Forgot entry in .cabal * Update changelog - - - - - 95c6a771 by Alec Theriault at 2018-02-01T14:58:18+01:00 Clickable anchors for headings (#716) See haskell/haddock#579. This just adds an <a> tag around the heading, pointing to the heading itself. - - - - - 21463d28 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump: Matches on function names weight more than matches in ... module names. - - - - - 8023af39 by Alex Biehl at 2018-02-01T14:58:18+01:00 Treat escaped \] better in definition lists (#717) This fixes haskell/haddock#546. - - - - - e4866dc1 by Alex Biehl at 2018-02-01T14:58:18+01:00 Remove scanner, takeWhile1_ already takes care of escaping - - - - - 9bcaa49d by Alex Biehl at 2018-02-01T14:58:18+01:00 Take until line feed - - - - - 01d2af93 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Add simple framework for running parser fixtures (#668) * Add simple framework for running parser fixtures * Compatible with tree-diff-0.0.0.1 * Use parseParas to parse fixtures This allows to test all syntactic constructs available in haddock markup. - - - - - 31128417 by Alec Theriault at 2018-02-01T14:58:18+01:00 Patch flaky parser test (#720) * Patch flaky parser test This test was a great idea, but it doesn't port over too well to using the GHC lexer. GHC rewrites its input a bit - nothing surprising, but we need to guard against those cases for the test. * Change instance head * Change use site - - - - - 9704f214 by Herbert Valerio Riedel at 2018-02-01T14:58:18+01:00 Include secondary LICENSE file in source dist - - - - - 51f25074 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Grid Tables (#718) * Add table examples * Add table types and adopt simple parser Simple parser is done by Giovanni Cappellotto (@potomak) in https://github.com/haskell/haddock/pull/577 It seems to support single fine full tables, so far from full RST-grid tables, but it's good start. Table type support row- and colspans, but obviously parser is lacking. Still TODO: - Latex backend. Should we use multirow package https://ctan.org/pkg/multirow?lang=en? - Hoogle backend: ? * Implement grid-tables * Refactor table parser * Add two ill-examples * Update CHANGES.md * Basic documentation for tables * Fix documentation example - - - - - 670d6200 by Alex Biehl at 2018-02-01T14:58:18+01:00 Add grid table example to cheatsheet (pdf and svg need to be regenerated thought) - - - - - 4262dec9 by Alec Theriault at 2018-02-01T14:58:18+01:00 Fix infinite loop when specializing instance heads (#723) * Fix infinite loop when specializing instance heads The bug can only be triggered from TH, hence why it went un-noticed for so long. * Add test for haskell/haddock#679 and haskell/haddock#710 - - - - - 67ecd803 by Alec Theriault at 2018-02-01T14:58:18+01:00 Filter RTS arguments from 'ghc-options' arguments (#725) This fixes haskell/haddock#666. - - - - - 7db26992 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump Scrollable overlay - - - - - da9ff634 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Hyperlinker: Adjust parser to new PFailed constructor - - - - - 7b7cf8cb by Alexander Biehl at 2018-02-01T14:58:18+01:00 Specialize: Add missing IdP annotations - - - - - 78cd7231 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Convert: Correct pass type - - - - - a2d0f590 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Warning free compilation - - - - - cd861cf3 by Alexander Biehl at 2018-02-01T14:58:18+01:00 hadock-2.19.0 / haddock-api-2.19.0 / haddock-library-1.5.0 - - - - - c6651b72 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Adjust changelogs - - - - - 1e93da0b by Alexander Biehl at 2018-02-01T14:58:18+01:00 haddock-library: Info about breaking changes - - - - - f9b11db8 by Alec Theriault at 2018-02-02T12:36:02+01:00 Properly color pragma contents in hyperlinker The hyperlinker backend now classifies the content of pragmas as 'TkPragma'. That means that in something like '{-# INLINE foo #-}', 'foo' still gets classified as a pragma token. - - - - - c40b0043 by Alec Theriault at 2018-02-02T12:36:02+01:00 Support the new 'ITcolumn_prag' token - - - - - 4a2a4d39 by Alex Biehl at 2018-02-03T12:11:55+01:00 QuickJump: Mitigate encoding problems on Windows - - - - - bb34503a by Alex Biehl at 2018-02-04T18:39:31+01:00 Use withBinaryFile - - - - - 637605bf by Herbert Valerio Riedel at 2018-02-05T09:48:32+01:00 Try GHC 8.4.1 for Travis CI job - - - - - 7abb67e4 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 try harder to build w/ GHC 8.4.1 - - - - - 8255cc98 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 Add `SPDX-License-Identifier` as alised for "license" module header tokens C.f. SPDX 2.1 - Appendix V https://spdx.org/spdx-specification-21-web-version#h.twlc0ztnng3b The tag should appear on its own line in the source file, generally as part of a comment. SPDX-License-Identifier: <SPDX License Expression> Cherry-picked from haskell/haddock#743 - - - - - 267cd23d by Herbert Valerio Riedel at 2018-02-05T10:24:34+01:00 Make test-suite SMP compatible - - - - - 95d4bf40 by Alec Theriault at 2018-02-05T22:01:04+01:00 Hyperlink pattern synonyms and 'module' imports (#744) Links to pattern synonyms are now generated, as well as links from modules in import lists. Fixes haskell/haddock#731. - - - - - 67838dcd by Alec Theriault at 2018-02-06T08:23:36+01:00 Don't warn about missing '~' (#746) This manually filters out '~' from the list of things to warn about. It truly makes no sense to warn on this since '~' has nothing it could link to - it is magical. This fixes haskell/haddock#532. - - - - - ab6c3f9f by Alec Theriault at 2018-02-06T08:24:47+01:00 Don't barf on 'HsSpliceTy' (#745) This handles 'HsSpliceTy's by replacing them with what they expand to. IIUC everything that is happening, 'renameHsSpliceTy' should not be able to fail for the inputs we feed it from GHC. This fixes haskell/haddock#574. - - - - - 92bf95ad by Alex Biehl at 2018-02-06T08:28:23+01:00 Rename: renameHsSpliceTy ttg - - - - - 3130b1e1 by Alex Biehl at 2018-02-06T09:02:14+01:00 Expand SigDs - - - - - c72adae5 by Alex Biehl at 2018-02-06T09:20:51+01:00 fullModuleContents: support named docs - - - - - de2e4dbf by Alex Biehl at 2018-02-06T13:56:17+01:00 Hyperlinker: Also link pattern synonym arguments - - - - - b7c98237 by Alex Biehl at 2018-02-09T18:44:23+01:00 Expand SigD in a better place In https://github.com/haskell/haddock/issues/287 we found that haddock-2.19.0 would miss documentation on class methods with multiples names. This patch uses expandSigDecls in a more sensible place. - - - - - 8f598b27 by Alec Theriault at 2018-02-11T12:29:56+01:00 Add module tooltips to linked identifiers (#753) No more clicking to figure out whether your bytestring is strict or lazy! - - - - - d812e65d by Alec Theriault at 2018-02-11T12:31:44+01:00 Add 'show' option to complement 'hide' (#752) * Add 'show' option to complement 'hide' The behaviour is for flags passed in the command line to override flags in file headers. In the command line, later flags override earlier ones. Fixes haskell/haddock#751 and haskell/haddock#266. * Add a '--show-all' option - - - - - 6676cecb by Alex Biehl at 2018-02-18T11:07:15-05:00 QuickJump: Mitigate encoding problems on Windows (cherry picked from commit 86292c54bfee2343aee84559ec01f1fc68f52231) - - - - - e753dd88 by Alex Biehl at 2018-02-18T17:59:54+01:00 Use withBinaryFile - - - - - 724dc881 by Tamar Christina at 2018-02-19T05:34:49+01:00 Haddock: support splitted include paths. (#689) - - - - - 9b6d6f50 by Alex Biehl at 2018-02-19T05:57:02+01:00 Teach the HTML backend how to render methods with multiple names - - - - - a74aa754 by Alexander Biehl at 2018-02-19T10:04:34+01:00 Hoogle/Latex: Remove use of partial function - - - - - 66d8bb0e by Alec Theriault at 2018-02-25T16:04:01+01:00 Fix file handle leak (#763) (#764) Brought back some mistakenly deleted code for handling encoding and eager reading of files from e0ada1743cb722d2f82498a95b201f3ffb303137. - - - - - bb92d03d by Alex Biehl at 2018-03-02T14:21:23+01:00 Enable running test suite with stock haddock and ghc using ``` $ cabal new-run -- html-test --haddock-path=$(which haddock) --ghc-path=$(which ghc) ``` - - - - - dddb3cb2 by Alex Biehl at 2018-03-02T15:43:21+01:00 Make testsuite work with haddock-1.19.0 release (#766) - - - - - f38636ed by Alec Theriault at 2018-03-02T15:48:36+01:00 Support unicode operators, proper modules Unicode operators are a pretty big thing in Haskell, so supporting linking them seems like it outweighs the cost of the extra machinery to force Attoparsec to look for unicode. Fixes haskell/haddock#458. - - - - - 09d89f7c by Alec Theriault at 2018-03-02T15:48:43+01:00 Remove bang pattern - - - - - d150a687 by Alex Biehl at 2018-03-02T15:48:48+01:00 fix test - - - - - d6fd71a5 by Alex Biehl at 2018-03-02T16:22:38+01:00 haddock-test: Be more explicit which packages to pass We now pass `-hide-all-packages` to haddock when invoking the testsuite. This ensures we don't accidentally pick up any dependencies up through ghc.env files. - - - - - 0932c78c by Alex Biehl at 2018-03-02T17:50:38+01:00 Revert "fix test" This reverts commit 1ac2f9569242f6cb074ba6e577285a4c33ae1197. - - - - - 52516029 by Alex Biehl at 2018-03-02T18:16:50+01:00 Fix Bug548 for real - - - - - 89df9eb5 by Alex Biehl at 2018-03-05T18:28:19+01:00 Hyperlinker: Links for TyOps, class methods and associated types - - - - - d019a4cb by Ryan Scott at 2018-03-06T13:43:56-05:00 Updates for haskell/haddock#13324 - - - - - 6d5a42ce by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump haddock-2.19.0.1, haddock-api-2.19.0.1, haddock-library-1.5.0.1 - - - - - c0e6f380 by Alex Biehl at 2018-03-10T18:25:57+01:00 Update changelogs for haddock-2.19.0.1 and haddock-library-1.5.0.1 - - - - - 500da489 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Update to QC 2.11 - - - - - ce8362e9 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Restore backward-compat with base-4.5 through base-4.8 - - - - - baae4435 by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump lower bound for haddock-library - - - - - 10b7a73e by Alex Biehl at 2018-03-10T18:25:57+01:00 Haddock: Straighten out base bound - - - - - a6096f7b by Alex Biehl at 2018-03-13T08:45:06+01:00 extractDecl: Extract constructor patterns from data family instances (#776) * extractDecl: Allow extraction of data family instance constructors * extractDecl: extract data family instance constructors - - - - - ba4a0744 by Simon Jakobi at 2018-03-14T08:26:42+01:00 Readme: Update GHC version (#778) - - - - - 8de157d4 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for definition lists - - - - - 425b46f9 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for links - - - - - d53945d8 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for inline links - - - - - f1dc7c99 by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Slightly unmangle output - - - - - 0879d31c by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Prevent stdout buffering - - - - - 1f9e5f1b by Simon Jakobi at 2018-03-14T20:39:29+01:00 haddock-library.cabal: Clean up GHC options - - - - - 066b891a by Simon Jakobi at 2018-03-14T20:39:29+01:00 Make a proper definition for the <link> parser - - - - - 573d6ba7 by Alec Theriault at 2018-03-21T09:16:57+01:00 Show where instances are defined (#748) * Indicate source module of instances Above instance, we now also display a link to the module where the instance was defined. This is sometimes helpful in figuring out what to import. * Source module for type/data families too * Remove parens * Accept tests - - - - - 99b5d28b by Alex Biehl at 2018-03-21T09:20:36+01:00 Prepare changelog for next release - - - - - 482d3a93 by Alex Biehl at 2018-03-23T15:57:36+01:00 Useful cost centres, timers and allocation counters (#785) * Add some useful cost-centres for profiling * Add withTiming for each haddock phase Invoking haddock with `--optghc=-ddump-timings` now shows the amount of time spent and the number of allocated bytes for each phase. - - - - - 773b41bb by Alec Theriault at 2018-03-27T08:35:59+02:00 @since includes package name (#749) * Metadoc stores a package name This means that '@since' annotations can be package aware. * Get the package name the right way This should extract the package name for `@since` annotations the right way. I had to move `modulePackageInfo` around to do this and, in the process, I took the liberty to update it. Since it appears that finding the package name is something that can fail, I added a warning for this case. * Silence warnings * Hide package for local 'since' annotations As discussed, this is still the usual case (and we should avoid being noisy for it). Although this commit is large, it is basically only about threading a 'Maybe Package' from 'Haddock.render' all the way to 'Haddock.Backends.Xhtml.DocMarkup.renderMeta'. * Bump binary interface version * Add a '--since-qual' option This controls when to qualify since annotations with the package they come from. The default is always, but I've left an 'external' variant where only those annotations coming from outside of the current package are qualified. * Make ParserSpec work * Make Fixtures work * Use package name even if package version is not available The @since stuff needs only the package name passed in, so it makes sense to not be forced to pass in a version too. - - - - - e42c57bc by Alex Biehl at 2018-03-27T08:42:50+02:00 haddock-2.19.1, haddock-api-2.19.1, haddock-library-1.6.0 - - - - - 8373a529 by Alex Biehl at 2018-03-28T10:17:11+02:00 Bump haddock and haddock-api to 2.20.0 - - - - - 5038eddd by Jack Henahan at 2018-04-03T13:28:12+02:00 Clear search string on hide for haskell/haddock#781 (#789) - - - - - 920ca1eb by Alex Biehl at 2018-04-03T16:35:50+02:00 Travis: Build with ghc-8.4.2 (#793) - - - - - a232f0eb by Alan Zimmerman at 2018-04-07T14:14:32+02:00 Match changes in GHC for D4199 Removing HasSourceText and SourceTextX classes. - - - - - ab85060b by Alan Zimmerman at 2018-04-09T21:20:24+02:00 Match GHC changes for TTG - - - - - 739302b6 by Alan Zimmerman at 2018-04-13T13:31:44+02:00 Match GHC for TTG implemented on HsBinds, D4581 - - - - - 2f56d3cb by Ryan Scott at 2018-04-19T11:42:58-04:00 Bump upper bound on base to < 4.13 See https://ghc.haskell.org/trac/ghc/ticket/15018. - - - - - a49df92a by Alex Biehl at 2018-04-20T07:31:44+02:00 Don't treat fixity signatures like declarations - - - - - d02c103b by Ryan Scott at 2018-04-24T11:20:11-04:00 Add regression test for haskell/haddock#413 Fixes haskell/haddock#413. - - - - - c7577f52 by Ryan Scott at 2018-04-24T13:51:06-07:00 Improve the Hoogle backend's treatment of type families (#808) Fixes parts 1 and 2 of haskell/haddock#806. - - - - - d88f85b1 by Alec Theriault at 2018-04-25T11:24:07-07:00 Replace 'attoparsec' with 'parsec' (#799) * Remove attoparsec with parsec and start fixing failed parses * Make tests pass * Fix encoding issues The Haddock parser no longer needs to worry about bytestrings. All the internal parsing work in haddock-library happens over 'Text'. * Remove attoparsec vendor * Fix stuff broken in 'attoparsec' -> 'parsec' * hyperlinks * codeblocks * examples Pretty much all issues are due to attoparsec's backtracking failure behaviour vs. parsec's non-backtracking failure behaviour. * Fix small TODOs * Missing quote + Haddocks * Better handle spaces before/after paragraphs * Address review comments - - - - - fc25e2fe by Alan Zimmerman at 2018-04-27T15:36:53+02:00 Match changes in GHC for TTG - - - - - 06175f91 by Herbert Valerio Riedel at 2018-05-01T18:11:09+02:00 Merge branch 'ghc-head' with 'ghc-8.4' - - - - - 879caaa8 by Alec Theriault at 2018-05-07T18:53:15-07:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 3e0120cb by Simon Jakobi at 2018-05-07T19:00:18-07:00 Add docs for some DocH constructors (#814) - - - - - 0a32c6db by Alec Theriault at 2018-05-08T02:15:45-07:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 8816e783 by Simon Jakobi at 2018-05-08T10:48:11-07:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - ad60366f by Ryan Scott at 2018-05-10T11:19:47-04:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - 03b7cc3b by Ryan Scott at 2018-05-10T11:24:38-04:00 Wibbles - - - - - b03dd563 by Chaitanya Koparkar at 2018-05-10T11:44:58-04:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 9f298a40 by Ben Gamari at 2018-05-13T17:36:04-04:00 Account for refactoring of LitString - - - - - ea3dabe7 by Ryan Scott at 2018-05-16T09:21:43-04:00 Merge pull request haskell/haddock#826 from haskell/T825 Remove Hoogle backend hack that butchers infix datatype names - - - - - 0d234f7c by Alec Theriault at 2018-05-23T11:29:05+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 15fc9712 by Simon Jakobi at 2018-05-31T04:17:47+02:00 Adjust to new HsDocString internals - - - - - 6f1e19a8 by Ben Gamari at 2018-06-02T16:18:58-04:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 0d0355d9 by Ryan Scott at 2018-06-04T21:26:59-04:00 DerivingVia changes - - - - - 0d93475a by Simon Jakobi at 2018-06-05T19:47:05+02:00 Bump a few dependency bounds (#845) - - - - - 5cbef804 by Alec Theriault at 2018-06-05T19:47:16+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 9094c56f by Alec Theriault at 2018-06-05T22:53:25+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 70188719 by Simon Jakobi at 2018-06-08T22:20:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - 495cd1fc by Chaitanya Koparkar at 2018-06-13T23:01:34+02:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 81088732 by Ben Gamari at 2018-06-13T23:01:34+02:00 Account for refactoring of LitString - - - - - 7baf6587 by Simon Jakobi at 2018-06-13T23:05:08+02:00 Adjust to new HsDocString internals - - - - - bb61464d by Ben Gamari at 2018-06-13T23:05:22+02:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 5d8cb87f by Ryan Scott at 2018-06-13T23:39:30+02:00 DerivingVia changes - - - - - 73d373a3 by Alec Theriault at 2018-06-13T23:39:30+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 4865e254 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Remove `ITtildehsh` token - - - - - b867db54 by Alec Theriault at 2018-06-13T23:39:30+02:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 9598e392 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Add docs for some DocH constructors (#814) - - - - - 8a59035b by Alec Theriault at 2018-06-13T23:39:30+02:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 29350fc8 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - 2590bbd9 by Ryan Scott at 2018-06-13T23:39:30+02:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - a9939fdc by Ryan Scott at 2018-06-13T23:39:30+02:00 Wibbles - - - - - a22f7df4 by Alec Theriault at 2018-06-13T23:39:30+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 8741015d by Simon Jakobi at 2018-06-13T23:39:30+02:00 Bump a few dependency bounds (#845) - - - - - 4791e1cc by Alec Theriault at 2018-06-13T23:39:30+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 311d3216 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - d0577817 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Complete FixitySig and FamilyDecl pattern matches - - - - - 055b3aa7 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Fix redundant import warnings - - - - - f9ce19b1 by Simon Jakobi at 2018-06-13T23:49:52+02:00 html-test: Accept output - - - - - 04604ea7 by Simon Jakobi at 2018-06-13T23:54:37+02:00 Bump bounds on Cabal - - - - - 0713b692 by Simon Jakobi at 2018-06-14T00:00:12+02:00 Merge branch 'ghc-head' into ghc-head-update-3 - - - - - c6a56bfd by Simon Jakobi at 2018-06-14T02:33:27+02:00 Bump ghc bound for haddock-api spec test-suite - - - - - 119d04b2 by Simon Jakobi at 2018-06-14T12:37:48+02:00 Travis: `--allow-newer` for all packages - - - - - 0e876e2c by Alex Biehl at 2018-06-14T15:28:52+02:00 Merge pull request haskell/haddock#857 from sjakobi/ghc-head-update-3 Update ghc-head - - - - - 5be46454 by Alec Theriault at 2018-06-14T21:42:45+02:00 Improved handling of interfaces in 'haddock-test' (#851) This should now work with an inplace GHC where (for instance) HTML directories may not be properly recorded in the package DB. - - - - - 96ab1387 by Vladislav Zavialov at 2018-06-14T17:06:21-04:00 Handle -XStarIsType - - - - - e518f8c4 by Ben Gamari at 2018-06-14T17:48:00-04:00 Revert unintentional reversion of fix of haskell/haddock#548 - - - - - 01b9f96d by Alan Zimmerman at 2018-06-19T11:52:22+02:00 Match changes in GHC for haskell/haddock#14259 - - - - - 7f8c8298 by Ben Gamari at 2018-06-19T18:14:27-04:00 Bump GHC version to 8.6 - - - - - 11c6b5d2 by Ryan Scott at 2018-06-19T23:17:31-04:00 Remove HsEqTy and XEqTy - - - - - b33347c2 by Herbert Valerio Riedel at 2018-06-20T23:14:52+02:00 Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. - - - - - f0d2460e by Herbert Valerio Riedel at 2018-06-20T23:28:46+02:00 Update Travis CI job - - - - - ef239223 by Herbert Valerio Riedel at 2018-06-20T23:32:41+02:00 Drop GHC HEAD from CI and update GHC to 8.4.3 It's a waste of resource to even try to build this branch w/ ghc-head; so let's not do that... - - - - - 41c4a9fa by Ben Gamari at 2018-06-20T18:26:20-04:00 Bump GHC version to 8.7 - - - - - 8be593dc by Herbert Valerio Riedel at 2018-06-21T22:32:15+02:00 Update CI job to use GHC 8.7.* - - - - - b91d334a by Simon Jakobi at 2018-06-30T13:41:38+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section - - - - - f707d848 by Alec Theriault at 2018-07-05T10:43:35-04:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. - - - - - a6d2b8dc by Alec Theriault at 2018-07-06T10:06:32-04:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case - - - - - 13819f71 by Alan Zimmerman at 2018-07-15T19:33:51+02:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 - - - - - c346aa78 by Simon Jakobi at 2018-07-19T12:29:32+02:00 haddock-library: Bump bounds for containers - - - - - 722e733c by Simon Jakobi at 2018-07-19T13:36:45+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] - - - - - f0bd83fd by Alec Theriault at 2018-07-19T14:39:57+02:00 Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. - - - - - 394053a8 by Simon Jakobi at 2018-07-19T14:58:07+02:00 haddock-library: Bump bounds for containers - - - - - 1bda11a2 by Alec Theriault at 2018-07-20T09:04:03+02:00 Add HEAD.hackage overlay (#887) * Add HEAD.hackage overlay * Add HCPKG variable - - - - - c7b4ab45 by Alec Theriault at 2018-07-20T12:01:16+02:00 Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test - - - - - c05d32ad by Alec Theriault at 2018-07-20T12:01:49+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output - - - - - 24b39ee4 by Alec Theriault at 2018-07-20T12:02:16+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. - - - - - cb9d2099 by Simon Jakobi at 2018-07-20T13:39:29+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) - - - - - 133f24f5 by Alec Theriault at 2018-07-20T13:39:29+02:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) - - - - - 11024149 by Alec Theriault at 2018-07-20T13:39:29+02:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) - - - - - de0c139e by Simon Jakobi at 2018-07-20T13:39:29+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) - - - - - 6435e952 by Alec Theriault at 2018-07-20T13:39:29+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) - - - - - 1461af39 by Alec Theriault at 2018-07-20T13:39:29+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) - - - - - 69d3bde1 by Alec Theriault at 2018-07-20T13:49:47+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) - - - - - 6a5c73c7 by Alec Theriault at 2018-07-20T13:50:00+02:00 Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests - - - - - 92ca94c6 by Alec Theriault at 2018-07-20T13:55:36+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) - - - - - 981bc7fa by Simon Jakobi at 2018-07-20T15:06:06+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers - - - - - 27e7c0c5 by Simon Jakobi at 2018-07-20T15:09:05+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 49e1a415 by Simon Jakobi at 2018-07-20T16:02:02+02:00 Update the ghc-8.6 branch (#889) * Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. * README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) * Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) * Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) * tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) * Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. * Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test * Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) * Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) * Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) * Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests * Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 5ca14bed by Simon Jakobi at 2018-07-20T16:05:47+02:00 Revert "Revert "Bump GHC version to 8.6"" That commit didn't belong onto the ghc-8.6 branch. This reverts commit acbaef3b9daf1d2dea10017964bf886e77a8e967. - - - - - 2dd600dd by Simon Jakobi at 2018-07-20T16:18:21+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. - - - - - fceb2422 by Simon Jakobi at 2018-07-20T16:18:21+02:00 outOfScope: Recommend qualifying the identifier - - - - - acea5d23 by Simon Jakobi at 2018-07-20T16:19:35+02:00 outOfScope: Recommend qualifying the identifier (cherry picked from commit 73707ed58d879cc04cb644c5dab88c39ca1465b7) - - - - - 1a83ca55 by Simon Jakobi at 2018-07-20T16:19:35+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. (cherry picked from commit d504a2864a4e1982e142cf88c023e7caeea3b76f) - - - - - 48374451 by Masahiro Sakai at 2018-07-20T17:06:42+02:00 Add # as a special character (#884) '#' has special meaning used for anchors and can be escaped using backslash. Therefore it would be nice to be listed as special characters. - - - - - 5e1a5275 by Alec Theriault at 2018-07-20T23:37:24+02:00 Let `haddock-test` bypass interface version check (#890) This means `haddock-test` might * crash during deserialization * deserialize incorrectly Still - it means things _might_ work where they were previously sure not to. - - - - - 27286754 by Yuji Yamamoto at 2018-07-23T08:16:01+02:00 Avoid "invalid argument (invalid character)" on non-unicode Windows (#892) Steps to reproduce and the error message ==== ``` > stack haddock basement ... snip ... Warning: 'A' is out of scope. Warning: 'haddock: internal error: <stdout>: commitBuffer: invalid argument (invalid character) ``` Environment ==== OS: Windows 10 ver. 1709 haddock: [HEAD of ghc-8.4 when I reproduce the error](https://github.com/haskell/haddock/commit/532b209d127e4cecdbf7e9e3dcf4f653a5605b5a). (I had to use this version to avoid another probrem already fixed in HEAD) GHC: 8.4.3 stack: Version 1.7.1, Git revision 681c800873816c022739ca7ed14755e85a579565 (5807 commits) x86_64 hpack-0.28.2 Related pull request ==== https://github.com/haskell/haddock/pull/566 - - - - - 6729d361 by Alec Theriault at 2018-07-23T13:52:56-07:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). - - - - - 8cf4e6b5 by Ryan Scott at 2018-07-27T11:28:03-04:00 eqTyCon_RDR now lives in TysWiredIn After GHC commit http://git.haskell.org/ghc.git/commit/f265008fb6f70830e7e92ce563f6d83833cef071 - - - - - 1ad251a6 by Alan Zimmerman at 2018-07-30T13:28:09-04:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 (cherry picked from commit e3926b50ab8a7269fd6904b06e881745f08bc5d6) - - - - - 8aea2492 by Richard Eisenberg at 2018-08-02T10:54:17-04:00 Update against new HsImplicitBndrs - - - - - e42cada9 by Alec Theriault at 2018-08-04T17:51:30+02:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 0e852512 by Alex Biehl at 2018-08-06T13:04:02+02:00 Make --package-version optional for --hoogle generation (#899) * Make --package-version optional for --hoogle generation * Import mkVersion * It's makeVersion not mkVersion - - - - - d2abd684 by Noel Bourke at 2018-08-21T09:34:18+02:00 Remove unnecessary backslashes from docs (#908) On https://haskell-haddock.readthedocs.io/en/latest/markup.html#special-characters the backslash and backtick special characters showed up with an extra backslash before them – I think the escaping is not (or no longer) needed for those characters in rst. - - - - - 7a578a9e by Matthew Pickering at 2018-08-21T09:34:50+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - aa3d4db3 by Matthew Pickering at 2018-08-21T09:37:34+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - ede91744 by Alec Theriault at 2018-08-21T09:42:52+02:00 Better test output when Haddock crashes on a test (#902) In particular: we report the tests that crashed seperately from the tests that produced incorrect output. In order for tests to pass (and exit 0), they must not crash and must produce the right output. - - - - - 4a872b84 by Guillaume Bouchard at 2018-08-21T09:45:57+02:00 Fix a typo (#878) - - - - - 4dbf7595 by Ben Sklaroff at 2018-08-21T12:04:09-04:00 Add ITcomment_line_prag token to Hyperlinker Parser This token is necessary for parsing #line pragmas inside nested comments. Reviewers: bgamari Reviewed By: bgamari Differential Revision: https://phabricator.haskell.org/D4935 - - - - - 9170b2a9 by Ben Gamari at 2018-08-21T17:55:15-04:00 Merge pull request haskell/haddock#893 from harpocrates/get-name-to-instances Accumulate explicitly which modules to load for 'attachInstances' - - - - - d57b57cc by Ben Gamari at 2018-08-21T17:59:13-04:00 Merge branch 'ghc-head' of github.com:haskell/haddock into ghc-head - - - - - 14601ca2 by Alec Theriault at 2018-08-21T19:09:37-04:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). (cherry picked from commit 5c7c596c51d69b92164e9ba920157b36ce2b2ec1) - - - - - 438c645e by Matthew Pickering at 2018-08-21T19:12:39-04:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 (cherry picked from commit e6aa8fb47b9477cc5ef5e46097524fe83e080f6d) - - - - - a80c5161 by Alec Theriault at 2018-08-21T22:06:40-07:00 Better rendering of unboxed sums/tuples * adds space after/before the '#' marks * properly reify 'HsSumTy' in 'synifyType' - - - - - 88456cc1 by Alec Theriault at 2018-08-21T22:06:40-07:00 Handle promoted tuples in 'synifyType' When we have a fully applied promoted tuple, we can expand it out properly. - - - - - fd1c1094 by Alec Theriault at 2018-08-21T22:19:34-07:00 Accept test cases - - - - - 6e80d9e0 by Alec Theriault at 2018-08-21T22:24:03-07:00 Merge pull request haskell/haddock#914 from harpocrates/feature/unboxed-stuff Better rendering of unboxed sums, unboxed tuples, promoted tuples. - - - - - 181a23f1 by Ben Gamari at 2018-08-23T15:53:48-04:00 Merge remote-tracking branch 'origin/ghc-8.6' into ghc-8.6 - - - - - 3a18c1d8 by Alec Theriault at 2018-08-27T14:15:25-07:00 Properly synify promoted list types We reconstruct promoted list literals whenever possible. That means that 'synifyType' produces '[Int, Bool, ()] instead of (Int ': (() ': (Bool ': ([] :: [Type])))) - - - - - b4794946 by Alec Theriault at 2018-09-03T07:19:55-07:00 Only look at visible types when synifying a 'HsListTy' The other types are still looked at when considering whether to make a kind signature or not. - - - - - a231fce2 by Alec Theriault at 2018-09-03T07:38:10-07:00 Merge pull request haskell/haddock#922 from harpocrates/promoted-lists Properly synify promoted list types - - - - - 0fdf044e by Ningning Xie at 2018-09-15T10:25:58-04:00 Update according to GHC Core changes - - - - - 7379b115 by Ningning Xie at 2018-09-15T15:40:18-04:00 update dataFullSig to work with Co Quantification This should have been in the previous patch, but wasn't. - - - - - cf84a046 by Alec Theriault at 2018-09-17T20:12:18-07:00 Fix/add to various docs * Add documentation for a bunch of previously undocumented options (fixes haskell/haddock#870) * Extend the documentation of `--hoogle` considerably (see haskell/haddock#807) * Describe how to add docs to `deriving` clauses (fixes haskell/haddock#912) * Fix inaccurate docs about hyperlinking infix identifiers (fixes haskell/haddock#780) - - - - - ae017935 by Alec Theriault at 2018-09-22T08:32:16-07:00 Update Travis - - - - - d95ae753 by Alec Theriault at 2018-09-22T09:34:10-07:00 Accept failing tests Also silence orphan warnings. - - - - - f3e67024 by Alec Theriault at 2018-09-22T09:41:23-07:00 Bump haddock-api-2.21.0, haddock-library-1.7.0 * Update CHANGELOGS * Update new versions in Cabal files * Purge references to ghc-8.4/master branches in README - - - - - 3f136d4a by Alec Theriault at 2018-09-22T10:53:31-07:00 Turn haddock-library into a minor release Fix some version bounds in haddock-library too. - - - - - b9def006 by Alec Theriault at 2018-09-22T13:07:35-07:00 keep cabal.project file - - - - - 4909aca7 by Alec Theriault at 2018-10-16T09:36:30-07:00 Build on 7.4 and 7.8 - - - - - 99d20a28 by Herbert Valerio Riedel at 2018-10-16T18:45:52+02:00 Minor tweak to package description - - - - - a8059618 by Herbert Valerio Riedel at 2018-10-16T18:47:24+02:00 Merge pull request haskell/haddock#945 haddock-api 2.21.0 and haddock-library 1.6.1 release - - - - - 2d9bdfc1 by Alec Theriault at 2018-10-16T10:54:21-07:00 Bump haddock-library to 1.7.0 The 1.6.1 release should've been a major bump, since types in the `Documentation.Haddock.Parser.Monad` module changed. This version makes that module internal (as it morally should be). - - - - - ed340cef by Alec Theriault at 2018-10-16T14:59:13-07:00 Merge branch 'ghc-8.4' into ghc-8.6 - - - - - 2821a8df by Alec Theriault at 2018-10-16T15:14:48-07:00 Merge branch 'ghc-8.6' into ghc-head - - - - - a722dc84 by Alec Theriault at 2018-10-16T16:28:55-07:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 63377496 by Alec Theriault at 2018-10-16T16:39:07-07:00 Update changelog - - - - - 099a0110 by Alec Theriault at 2018-10-16T16:49:28-07:00 Merge pull request haskell/haddock#942 from harpocrates/update-docs Fix & add to documentation - - - - - 0927416f by Alec Theriault at 2018-10-16T16:50:14-07:00 Set UTF-8 encoding before writing files (#934) This should fix haskell/haddock#929, as well as guard against future problems of this sort in other places. Basically replaces 'writeFile' (which selects the users default locale) with 'writeUtf8File' (which always uses utf8). - - - - - 83b7b017 by Alec Theriault at 2018-10-16T17:42:05-07:00 Output pattern synonyms in Hoogle backend (#947) * Output pattern synonyms in Hoogle backend We were previously weren't outputting _any_ pattern synonyms, bundled or not. Now, we output both. Fixes haskell/haddock#946. * Update changelog - - - - - 81e5033d by Alec Theriault at 2018-10-16T18:04:40-07:00 Release `haddock{,-api}-2.22.0` This version will accompany ghc-8.6.2 - - - - - 9661744e by Alex Biehl at 2018-10-18T08:14:32-07:00 Add NewOcean theme And make it the default theme. - - - - - 7ae6d722 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Improve appearance and readability These changes include: - use latest Haskell's logo colors - decrease #content width to improve readability - use nicer font - improve sizes and distances - - - - - 37f8703d by NunoAlexandre at 2018-10-18T08:14:32-07:00 Include custom font in the html head - - - - - 1d5e1d79 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Update html test reference files - - - - - 53b7651f by NunoAlexandre at 2018-10-18T08:14:32-07:00 Make it responsive - It makes small screens taking more space than larger ones - fixes a few issues present in small screens currently - make it look good across different screen sizes. - - - - - 6aa1aeb1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make the style consistent with hackage Several things are addressed here: - better responsive behaviour on the header - better space usage - consistent colors overall - other nit PR comments - - - - - 3a250c5c by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Place the package name before the menu links This supports the expected responsive menu design, where the package name appears above the menu links. - - - - - cae699b3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update html-test reference files The package name element in the package-header is now a div instead of a paragraph, and it is now above the menu ul.links instead of below. - - - - - 2ec7fd2d by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve synopsis style and code - Use CSS3 instead of loading pictures to show "+" and "-" symbols - Drop redundant code - - - - - 0c874c01 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Decrease space between code blocks There was too much space between code blocks as pointed out by reviewers. - - - - - 85568ce2 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Add an initial-scale property to all haddock pages This solves an issue reported about the content looking incredibly small on mobile devices. - - - - - c1538926 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Address several PR comments - Darken text color like done for hackage - Move synopsis to left side - Make table of contents stick to the left on wide screens - Wrap links to avoid page overflow - Improve expand/collapse buttons - Fix issue with content size on mobile devices - Fix issue with font-size on landscape mode - Increase width of the content - Change colors of table of contents and synopsis - Etc - - - - - e6639e5f by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make synopsis scrollable on wide screens When the synopsis is longer than the screen, you can’t see its end and you can't scroll down either, making the content unreachable. - - - - - 1f0591ff by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve information density - Reduce font size - Improve space between and within code blocks - Improve alignments - Improve spacing within sub-blocks - - - - - bf083097 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Minor adjustments Bring in some adjustments made to hackage: - link colors - page header show everything when package title is too long - - - - - 10375fc7 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix responsive triggers overlap issue The min and max width triggers have the same values, which caused the style resolution to take an intersection of both style declarations when the screen resolution had the size of the limts (say 1280px), causing an odd behaviour and look. - - - - - 95ff2f95 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix issue with menu alignment on firefox Reported and described here: https://github.com/haskell/haddock/pull/721#issuecomment-374668869 - - - - - dc86587e by Alex Biehl at 2018-10-18T08:14:32-07:00 Changelog entry for NewOcean - - - - - 27195e47 by Herbert Valerio Riedel at 2018-10-18T08:14:32-07:00 html-test --accept - - - - - 83f4f9c0 by Alex Biehl at 2018-10-18T08:14:32-07:00 Avoid name shadowing - - - - - 231487f1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font to PT Sans Also migrate some general text related changes from hackage. - - - - - 313db81a by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Use 'flex' to fix header alignment - - - - - 5087367b by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Misc of tweaks - Update link colors to hackage scheme - Tune spacing between content elements - Update footer style - Fix and improve code blocks identation - - - - - b08020df by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font in Xhtml.hs to PT Sans - - - - - 78ce06e3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve code blocks styling - Fix and improve spacing - Improve colors and borders - - - - - 81262d20 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make package-header caption backward-compatible The current html generator of this branch wraps the package-header caption as a div, which does not work (without style adjustments) with the old themes. Changing it from div to span does the trick, without needing to adjust the old stylesheets. - - - - - dc4475cb by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update test-suite reference html pages - - - - - 393d35d8 by Alec Theriault at 2018-10-18T08:25:36-07:00 Accept tests - - - - - a94484ba by Alec Theriault at 2018-10-21T10:29:29-07:00 Fix CHANGELOG - - - - - 8797eca3 by Alec Theriault at 2018-10-21T10:36:19-07:00 Update 'data-files' to include NewOcean stuff - - - - - 1ae51e4a by Simon Jakobi at 2018-10-23T11:29:14+02:00 Fix typo in a warning - - - - - 009ad8e8 by Alec Theriault at 2018-10-24T12:47:47-07:00 Update JS dependencies This was done via `npm audit fix`. I think this fixes haskell/haddock#903 along with some more serious vulnerabilities that nobody seems to have noticed. - - - - - 051994db by Alec Theriault at 2018-10-24T17:31:09-07:00 Resurrect the style-switcher This fixes haskell/haddock#810. Looks like things were broken during the quickjump refactor of the JS. For the (git) record: I do not think the style switcher is a good idea. I'm fixing it for the same reason @mzero added it; as an answer to "rumblings from some that they didn't want their pixels changed on bit" - - - - - 2a1d620f by Alec Theriault at 2018-10-24T17:38:07-07:00 Fix copy-pasta error in data-files - - - - - ed5bfb7f by Alec Theriault at 2018-10-24T20:42:14-07:00 Fix the synopsis button Here's these changes are supposed to do: * put the synopsis back on the right side * properly have it on the edge of the screen on wide screens * adjust the background of the synopsis to match the button (otherwise the grey blends in with what is underneath) * get rid of the dotted purple line * the synopsis contents are now scrollable even when in wide screens (this has been a long-standing bug) - - - - - 883fd74b by Alec Theriault at 2018-10-25T20:16:46-07:00 Avoid more conflicts in generated ids (#954) This fixes haskell/haddock#953 by passing more names into the generated ids. - - - - - ea54e331 by Alec Theriault at 2018-10-25T21:07:12-07:00 Don't hide bullets in method docs I think thst CSS was meant only to deal with fields and the effect on bullets was accidental. Fixes haskell/haddock#926. - - - - - 9a14ef4a by Alec Theriault at 2018-10-25T22:02:07-07:00 Indent more things + slightly smaller font - - - - - b9f17e29 by Alec Theriault at 2018-10-25T22:10:01-07:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 096a3cfa by Alec Theriault at 2018-10-25T22:24:38-07:00 Accept HTML output - - - - - 2669517d by Alec Theriault at 2018-10-26T09:02:35-07:00 User manual + stuff for building GHC docs - - - - - 46b27687 by Alec Theriault at 2018-10-26T09:10:59-07:00 Make 'Contents' in NewOcean scrollable This only happens if the contents block on the left is so big that it doesn't fit (vertically) on the page. If that happens, we want it to be scrollable. - - - - - 3443dd94 by Alec Theriault at 2018-10-26T09:36:46-07:00 Revert "Make 'Contents' in NewOcean scrollable" This reverts commit f909ffd8353d6463fd5dd184998a32aa98d5c922. I missed the fact this also forces the 'Contents' to always go down to the bottom of the page. - - - - - ed081424 by Alec Theriault at 2018-10-26T14:22:23-07:00 Avoid some partiality AFAICT this wasn't causing any crashes, but that's mostly because we happen not to be forcing `pkgStr` when it would diverge. We come dangerously close to doing that in `ppHtmlIndex`. Fixes haskell/haddock#569. - - - - - 6a5bec41 by Alec Theriault at 2018-10-27T10:05:04-07:00 Fix documentation in `haddock-api` (#957) * Fix misplaced Haddocks in Haddock itself Haddock should be able to generate documentation for 'haddock-api' again. * Make CI check that documentation can be built. * Add back a doc that is OK - - - - - 5100450a by Matthew Yacavone at 2018-10-27T14:51:38-04:00 More explicit foralls (GHC Proposal 0007) - - - - - 8771a6b0 by Alec Theriault at 2018-11-05T13:58:11-08:00 Only run MathJax on entities with "mathjax" class (#960) Correspondingly, we wrap all inline/diplay math in <span class="mathjax"> ... the math .... </span> This fixes haskell/haddock#959. - - - - - bd7ff5c5 by Alec Theriault at 2018-11-05T15:54:22-08:00 Deduplicate some work in 'AttachInstances' Perf only change: * avoid needlessly union-ing maps * avoid synify-ing instances twice Took this opportunity to add some docs too - - - - - cf99fd8f by Alec Theriault at 2018-11-05T15:54:22-08:00 Specialize some SYB functions Perf only change: * Add a 'SPECIALIZE' pragma to help GHC optimize a 'Data a =>' constraint * Manually specialize the needlessly general type of 'specializeTyVarBndrs' - - - - - 4f91c473 by Alec Theriault at 2018-11-05T15:54:22-08:00 Improve perf of renaming Perf only change: * don't look up type variable names (they're never in the environment) * use a difference list for accumulating missing names * more efficient 'Functor'/'Applicative' instances for 'RnM' - - - - - 4bbab0d4 by Alec Theriault at 2018-11-05T15:54:22-08:00 Faster 'Text' driven parser combinators Perf only change: * use 'getParserState'/'setParserState' to make 'Text'-optimized parser combinators * minimize uses of 'Data.Text.{pack,unpack,cons,snoc}' - - - - - fa430c02 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support hyperlink labels with inline markup The parser for pictures hasn't been properly adjusted yet. - - - - - c1431035 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support (and flatten) inline markup in image links Inline markup is supported in image links but, as per the [commonmark recommendation][0], it is stripped back to a plain text representation. [0]: https://spec.commonmark.org/0.28/#example-547 - - - - - d4ee1ba5 by Alec Theriault at 2018-11-06T12:03:24-08:00 Accept test case - - - - - 8088aeb1 by Alec Theriault at 2018-11-06T12:03:24-08:00 Fix/add to haddock-library test suite - - - - - e78f644d by Alec Theriault at 2018-11-06T13:26:31-08:00 Bump version bounds - - - - - 644335eb by Alec Theriault at 2018-11-06T13:53:30-08:00 Merge pull request haskell/haddock#875 from harpocrates/feature/markup-in-hyperlinks Inline markup in markdown-style links and images - - - - - e173ed0d by Alec Theriault at 2018-11-07T12:37:18-08:00 Fix issues around plus/minus * swap the minimize unicode to something more intuitive * use new unicode expander/collapser for instance lists * address some alignment issues in the "index" page - - - - - b2d92df7 by Alec Theriault at 2018-11-07T13:41:57-08:00 Allow "Contents" summary to scroll in a fixed div In the unfortunate event that the "Contents" summary doesn't fit vertically (like in the "Prelude"), it will be scrollable. - - - - - ca704c23 by Alec Theriault at 2018-11-07T13:45:15-08:00 Accept HTML output changes - - - - - 82c0ec6d by Alec Theriault at 2018-11-07T18:12:54-08:00 overflow-y 'scroll' -> 'auto' - - - - - 571d7657 by Alec Theriault at 2018-11-08T19:44:12-08:00 Clicking on "Contents" navigates to top of page - - - - - 8065a012 by Alec Theriault at 2018-11-08T19:44:17-08:00 Space out functions more Also, functions and data decls now have the same space before and after them. - - - - - cc650ede by Alec Theriault at 2018-11-09T08:13:35-08:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 65f8c17f by Alec Theriault at 2018-11-10T14:04:06-08:00 Update changelog - - - - - 20473847 by Alec Theriault at 2018-11-10T14:21:40-08:00 Replace oplus/ominus expander/collapser icons with triangles - - - - - 16592957 by Alec Theriault at 2018-11-10T14:35:10-08:00 Merge pull request haskell/haddock#949 from haskell/wip/new-ocean Introduce NewOcean theme. - - - - - 357cefe1 by Alec Theriault at 2018-11-10T16:02:13-08:00 Merge branch 'ghc-8.6' into ghc-head - - - - - de612267 by Alec Theriault at 2018-11-11T20:01:21-08:00 Rename 'NewOcean' theme to 'Linuwial' - - - - - 954b5baa by Alec Theriault at 2018-11-12T08:33:18-08:00 Add blockquote styling Matches b71da1feabf33efbbc517ac376bb690b5a604c2f from hackage-server. Fixes haskell/haddock#967. - - - - - d32c0b0b by Fangyi Zhou at 2018-11-12T10:24:13-08:00 Fix some broken links (#15733) Summary: For links in subpackages as well. https://phabricator.haskell.org/D5257 Test Plan: Manually verify links Reviewers: mpickering, bgamari, osa1 Reviewed By: osa1 GHC Trac Issues: haskell/haddock#15733 Differential Revision: https://phabricator.haskell.org/D5262 - - - - - 41098b1f by Alp Mestanogullari at 2018-11-15T22:40:09+01:00 Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change It got introduced in ghc/ghc at ae2c9b40f5b6bf272251d1f4107c60003f541b62. - - - - - c5c1c7e0 by Alec Theriault at 2018-11-15T13:48:13-08:00 Merge pull request haskell/haddock#970 from alpmestan/alp/fix-promotionflag Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change - - - - - 6473d3a4 by Shayan-Najd at 2018-11-23T01:38:49+01:00 [TTG: Handling Source Locations] Foundation and Pat Trac Issues haskell/haddock#15495 This patch removes the ping-pong style from HsPat (only, for now), using the plan laid out at https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow/HandlingSourceLocations (solution A). - the class `HasSrcSpan`, and its functions (e.g., `cL` and `dL`), are introduced - some instances of `HasSrcSpan` are introduced - some constructors `L` are replaced with `cL` - some patterns `L` are replaced with `dL->L` view pattern - some type annotation are necessarily updated (e.g., `Pat p` --> `Pat (GhcPass p)`) - - - - - 7a088dfe by Alec Theriault at 2018-11-26T11:11:28-08:00 More uniform handling of `forall`'s in HTML/LaTeX * don't forget to print explicit `forall`'s when there are arg docs * when printing an explicit `forall`, print all tyvars Fixes haskell/haddock#973 - - - - - d735e570 by Alec Theriault at 2018-12-12T08:42:09-08:00 Fix warnings, accept output * remove redundant imports (only brought to light due to recent work for improving redundant import detection) * fix a bug that was casuing exports to appear in reverse order * fix something in haddock-library that prevented compilation on old GHC's - - - - - a3852f8a by Zejun Wu at 2018-12-14T09:37:47-05:00 Output better debug infromation on internal error in extractDecl This will make investigation of haskell/haddock#979 easier - - - - - 2eccb5b9 by Alec Theriault at 2018-12-17T09:25:10-05:00 Refactor names + unused functions (#982) This commit should not introduce any change in functionality! * consistently use `getOccString` to convert `Name`s to strings * compare names directly when possible (instead of comparing strings) * get rid of unused utility functions - - - - - e82e4df8 by Alec Theriault at 2018-12-20T16:16:30-05:00 Load plugins when compiling each module (#983) * WIP: Load (typechecker) plugins from language pragmas * Revert "Load plugins when starting a GHC session (#905)" This reverts commit 72d82e52f2a6225686d9668790ac33c1d1743193. * Simplify plugin initialization code - - - - - 96e86f38 by Alec Theriault at 2018-12-23T10:23:20-05:00 Properly synify and render promoted type variables (#985) * Synify and render properly promoted type variables Fixes haskell/haddock#923. * Accept output - - - - - 23343345 by Alec Theriault at 2018-12-27T16:39:38-05:00 Remove `haddock-test`'s dep. on `syb` (#987) The functionality is easily inlined into one short function: `gmapEverywhere`. This doesn't warrant pulling in another package. - - - - - d0734f21 by Alec Theriault at 2018-12-27T16:39:52-05:00 Address deprecation warnings in `haddock-test` (#988) Fixes haskell/haddock#885. - - - - - 4d9f144e by mynguyen at 2018-12-30T23:42:26-05:00 Visible kind application haddock update - - - - - ffe0e9ed by Alec Theriault at 2019-01-07T13:55:22-08:00 Print kinded tyvars in constructors for Hoogle (#993) Fixes haskell/haddock#992 - - - - - 2e18b55d by Alec Theriault at 2019-01-10T16:42:45-08:00 Accept new output `GHC.Maybe` -> `Data.Maybe` (#996) Since 53874834b779ad0dfbcde6650069c37926da1b79 in GHC, "GHC.Maybe" is marked as `not-home`. That changes around some test output. - - - - - 055da666 by Gabor Greif at 2019-01-22T14:41:51+01:00 Lone typofix - - - - - 01bb71c9 by Alec Theriault at 2019-01-23T11:46:46-08:00 Keep forall on H98 existential data constructors (#1003) The information about whether or not there is a source-level `forall` is already available on a `ConDecl` (as `con_forall`), so we should use it instead of always assuming `False`! Fixes haskell/haddock#1002. - - - - - f9b9bc0e by Ryan Scott at 2019-01-27T09:28:12-08:00 Fix haskell/haddock#1004 with a pinch of dropForAlls - - - - - 5cfcdd0a by Alec Theriault at 2019-01-28T16:49:57-08:00 Loosen 'QuickCheck' and 'hspec' bounds It looks like the new versions don't cause any breakage and loosening the bounds helps deps fit in one stack resolver. - - - - - 3545d3dd by Alec Theriault at 2019-01-31T01:37:25-08:00 Use `.hie` files for the Hyperlinker backend (#977) # Summary This is a large architectural change to the Hyperlinker. * extract link (and now also type) information from `.hie` instead of doing ad-hoc SYB traversals of the `RenamedSource`. Also adds a superb type-on-hover feature (#715). * re-engineer the lexer to avoid needless string conversions. By going directly through GHC's `P` monad and taking bytestring slices, we avoid a ton of allocation and have better handling of position pragmas and CPP. In terms of performance, the Haddock side of things has gotten _much_ more efficient. Unfortunately, much of this is cancelled out by the increased GHC workload for generating `.hie` files. For the full set of boot libs (including `ghc`-the-library) * the sum of total time went down by 9-10% overall * the sum of total allocations went down by 6-7% # Motivation Haddock is moving towards working entirely over `.hi` and `.hie` files. This change means we no longer need the `RenamedSource` from `TypecheckedModule` (something which is _not_ in `.hi` files). # Details Along the way a bunch of things were fixed: * Cross package (and other) links are now more reliable (#496) * The lexer tries to recover from errors on every line (instead of at CPP boundaries) * `LINE`/`COLUMN` pragmas are taken into account * filter out zero length tokens before rendering * avoid recomputing the `ModuleName`-based `SrcMap` * remove the last use of `Documentation.Haddock.Utf8` (see haskell/haddock#998) * restructure temporary folder logic for `.hi`/`.hie` model - - - - - 2ded3359 by Herbert Valerio Riedel at 2019-02-02T12:06:12+01:00 Update/modernise haddock-library.cabal file - - - - - 62b93451 by Herbert Valerio Riedel at 2019-02-02T12:19:31+01:00 Tentatively declare support for unreleased base-4.13/ghc-8.8 - - - - - 6041e767 by Herbert Valerio Riedel at 2019-02-02T16:04:32+01:00 Normalise LICENSE text w/ cabal's BSD2 template Also, correct the `.cabal` files to advertise `BSD2` instead of the incorrect `BSD3` license. - - - - - 0b459d7f by Alec Theriault at 2019-02-02T18:06:12-08:00 CI: fetch GHC from validate artifact Should help make CI be less broken - - - - - 6b5c07cf by Alec Theriault at 2019-02-02T18:06:12-08:00 Fix some Hyperlinker test suite fallout * Amend `ParserSpec` to match new Hyperlinker API - pass in compiler info - strip out null tokens * Make `hypsrc-test` pass reliably - strip out `local-*` ids - strip out `line-*` ids from the `ClangCppBug` test - re-accept output - - - - - ded34791 by Nathan Collins at 2019-02-02T18:31:23-08:00 Update README instructions for Stack No need to `stack install` Haddock to test it. Indeed, `stack install` changes the `haddock` on user's `PATH` if `~/.local/bin` is on user's `PATH` which may not be desirable when hacking on Haddock. - - - - - 723298c9 by Alec Theriault at 2019-02-03T09:11:05-08:00 Remove `Documentation.Haddock.Utf8` The circumstances under which this module appeared are completely gone. The Hyperlinker backend no longer needs this module (it uses the more efficient `Encoding` module from `ghc`). Why no deprecation? Because this module really shouldn't exist! - It isn't used in `haddock-library`/`haddock-api` anymore - It was copy pasted directly from `utf8-string` - Folks seeking a boot-lib only solution can use `ghc`'s `Encoding` - - - - - 51050006 by Alec Theriault at 2019-02-03T22:58:58-08:00 Miscellaneous improvements to `Convert` (#1020) Now that Haddock is moving towards working entirely over `.hi` and `.hie` files, all declarations and types are going to be synthesized via the `Convert` module. In preparation for this change, here are a bunch of fixes to this module: * Add kind annotations to type variables in `forall`'s whose kind is not `Type`, unless the kind can be inferred from some later use of the variable. See `implicitForAll` and `noKindTyVars` in particular if you wish to dive into this. * Properly detect `HsQualTy` in `synifyType`. This is done by following suit with what GHC's `toIfaceTypeX` does and checking the first argument of `FunTy{} :: Type` to see if it classified as a given/wanted in the typechecker (see `isPredTy`). * Beef up the logic around figuring out when an explicit `forall` is needed. This includes: observing if any of the type variables will need kind signatures, if the inferred type variable order _without_ a forall will still match the one GHC claims, and some other small things. * Add some (not yet used) functionality for default levity polymorphic type signatures. This functionality similar to `fprint-explicit-runtime-reps`. Couple other smaller fixes only worth mentioning: * Show the family result signature only when it isn't `Type` * Fix rendering of implicit parameters in the LaTeX and Hoogle backends * Better handling of the return kind of polykinded H98 data declarations * Class decls produced by `tyThingToLHsDecl` now contain associated type defaults and default method signatures when appropriate * Filter out more `forall`'s in pattern synonyms - - - - - 841980c4 by Oleg Grenrus at 2019-02-04T08:44:25-08:00 Make a fixture of weird parsing of lists (#997) The second example is interesting. If there's a list directly after the header, and that list has deeper structure, the parser is confused: It finds two lists: - One with the first nested element, - everything after it I'm not trying to fix this, as I'm not even sure this is a bug, and not a feature. - - - - - 7315c0c8 by Ryan Scott at 2019-02-04T12:17:56-08:00 Fix haskell/haddock#1015 with dataConUserTyVars (#1022) The central trick in this patch is to use `dataConUserTyVars` instead of `univ_tvs ++ ex_tvs`, which displays the foralls in a GADT constructor in a way that's more faithful to how the user originally wrote it. Fixes haskell/haddock#1015. - - - - - ee0b49a3 by Ryan Scott at 2019-02-04T15:25:17-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. - - - - - 1c850dc8 by Alan Zimmerman at 2019-02-05T21:54:18+02:00 Matching changes in GHC for haskell/haddock#16236 - - - - - ab03c38e by Simon Marlow at 2019-02-06T08:07:33+00:00 Merge pull request haskell/haddock#1014 from hvr/pr/bsd2-normalise Normalise LICENSE text w/ cabal's BSD2 template - - - - - 5a92ccae by Alec Theriault at 2019-02-10T06:21:55-05:00 Merge remote-tracking branch 'gitlab/wip/T16236-2' into ghc-head - - - - - c0485a1d by Alec Theriault at 2019-02-10T03:32:52-08:00 Removes `haddock-test`s dependency on `xml`/`xhtml` (#1027) This means that `html-test`, `latex-test`, `hoogle-test`, and `hypsrc-test` now only depend on GHC boot libs. So we should now be able to build and run these as part of GHC's testsuite. \o/ The reference output has changed very slightly, in three ways: * we don't convert quotes back into `&quot;` as the `xml` lib did * we don't add extra `&nbsp;` as the `xml` lib did * we now remove the entire footer `div` (instead of just emptying it) - - - - - 65a448e3 by Alec Theriault at 2019-02-11T12:27:41-05:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - 360ca937 by Alec Theriault at 2019-02-13T11:36:11-05:00 Clean up logic for guessing `-B` and `--lib` (#1026) Haddock built with the `in-ghc-tree` flag tries harder to find the GHC lib folder and its own resources. This should make it possible to use `in-ghc-tree`-built Haddock without having to specify the `-B` and `--lib` options (just how you can use in-tree GHC without always specifying the `-B` option). The logic to do this relies on `getExecutablePath`, so we only get this auto-detection on platforms where this function works. - - - - - d583e364 by Alec Theriault at 2019-02-16T10:41:22-05:00 Fix tests broken by GHC Changes in 19626218566ea709b5f6f287d3c296b0c4021de2 affected some of the hyperlinker output. Accepted the new output (hovering over a `..` now shows you what that wildcard binds). Also fixed some stray deprecation warnings. - - - - - da0c42cc by Vladislav Zavialov at 2019-02-17T11:39:19+03:00 Parser changes to match !380 - - - - - ab96bed7 by Ryan Scott at 2019-02-18T04:44:08-05:00 Bump ghc version to 8.9 - - - - - 44b7c714 by Alec Theriault at 2019-02-22T05:49:43-08:00 Match GHC changes for T16185 `FunTy` now has an `AnonArgFlag` that indicates whether the arrow is a `t1 => t2` or `t1 -> t2`. This commit shouldn't change any functionality in Haddock. - - - - - 2ee653b1 by Alec Theriault at 2019-02-24T18:53:33-08:00 Update .travis.yml Points to the new GHC CI artifact. - - - - - 90939d71 by Alec Theriault at 2019-02-25T00:42:41-08:00 Support value/type namespaces on identifier links Identifier links can be prefixed with a 'v' or 't' to indicate the value or type namespace of the desired identifier. For example: -- | Some link to a value: v'Data.Functor.Identity' -- -- Some link to a type: t'Data.Functor.Identity' The default is still the type (with a warning about the ambiguity) - - - - - d6ed496c by Alec Theriault at 2019-02-25T00:42:46-08:00 Better identifier parsing * '(<|>)' and '`elem`' now get parsed and rendered properly as links * 'DbModule'/'DbUnitId' now properly get split apart into two links * tuple names now get parsed properly * some more small niceties... The identifier parsing code is more precise and more efficient (although to be fair: it is also longer and in its own module). On the rendering side, we need to pipe through information about backticks/parens/neither all the way through from renaming to the backends. In terms of impact: a total of 35 modules in the entirety of the bootlib + ghc lib docs change. The only "regression" is things like '\0'. These should be changed to @\\0@ (the path by which this previously worked seems accidental). - - - - - 3c3b404c by Alec Theriault at 2019-02-25T22:12:11-08:00 Fix standalone deriving docs Docs on standalone deriving decls for classes with associated types should be associated with the class instance, not the associated type instance. Fixes haskell/haddock#1033 - - - - - d51ef69e by Alec Theriault at 2019-02-26T19:14:59-08:00 Fix bogus identifier defaulting This avoids a situation in which an identifier would get defaulted to a completely different identifier. Prior to this commit, the 'Bug1035' test case would hyperlink 'Foo' into 'Bar'! Fixes haskell/haddock#1035. - - - - - 88cbbdc7 by Ryan Scott at 2019-02-27T10:14:03-05:00 Visible dependent quantification (#16326) changes - - - - - 0dcf6cee by Xia Li-yao at 2019-02-27T21:53:27-05:00 Menu item controlling which instances are expanded/collapsed (#1007) Adds a menu item (like "Quick Jump") for options related to displaying instances. This provides functionality for: * expanding/collapsing all instances on the currently opened page * controlling whether instances are expanded/collapsed by default * controlling whether the state of instances should be "remembered" This new functionality is implemented in Typescript in `details-helper`. The built-in-themes style switcher also got a revamp so that all three of QuickJump, the style switcher, and instance preferences now have the same style and implementation structure. See also: https://mail.haskell.org/pipermail/haskell-cafe/2019-January/130495.html Fixes haskell/haddock#698. Co-authored-by: Lysxia <lysxia at gmail.com> Co-authored-by: Nathan Collins <conathan at galois.com> - - - - - 3828c0fb by Alec Theriault at 2019-02-28T12:42:49-05:00 `--show-interface` should output to stdout. (#1040) Fixes haskell/haddock#864. - - - - - a50f4cda by gbaz at 2019-03-01T07:43:16-08:00 Increase contrast of Linuwal theme (#1037) This is to address the concern that, on less nice and older screens, some of the shades of grey blend in too easily with the white background. * darken the font slightly * darken slightly the grey behind type signatures and such * add a border and round the corners on code blocks * knock the font down by one point - - - - - ab4d41de by Alec Theriault at 2019-03-03T09:23:26-08:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 12f509eb by Ben Gamari at 2019-03-04T22:13:20-05:00 Remove reference to Opt_SplitObjs flag Split-objects has been removed. - - - - - 5b3e4c9a by Ryan Scott at 2019-03-06T19:16:24-05:00 Update html-test output to reflect haskell/haddock#16391 changes - - - - - fc228af1 by Alec Theriault at 2019-03-09T08:29:23-08:00 Match changes for "Stop inferring over-polymorphic kinds" The `hsq_ext` field of `HsQTvs` is now just the implicit variables (instead of also including information about which of these variables are dependent). This commit shouldn't change any functionality in Haddock. - - - - - 6ac109eb by Alec Theriault at 2019-03-09T11:22:55-08:00 Add .hi, .dyn_hi, etc files to .gitignore Fixes haskell/haddock#1030. - - - - - b55f0c05 by Alec Theriault at 2019-03-09T11:22:55-08:00 Better support for default methods in classes * default methods now get rendered differently * default associated types get rendered * fix a forgotten `s/TypeSig/ClassOpSig/` refactor in LaTeX backend * LaTeX backend now renders default method signatures NB: there is still no way to document default class members and the NB: LaTeX backend still crashes on associated types - - - - - 10aea0cf by Alec Theriault at 2019-03-09T11:22:55-08:00 Avoid multi-line `emph` in LaTeX backend `markupWarning` often processes inputs which span across paragraphs. Unfortunately, LaTeX's `emph` is not made to handle this (and will crash). Fixes haskell/haddock#936. - - - - - d22dc2c9 by Alec Theriault at 2019-03-09T11:22:55-08:00 Many LaTeX backend fixes After this commit, we can run with `--latex` on all boot libraries without crashing (although the generated LaTeX still fails to compile in a handful of larger packages like `ghc` and `base`). * Add newlines after all block elements in LaTeX. This is important to prevent the final output from being more an more indented. See the `latext-test/src/Example` test case for a sample of this. * Support associated types in class declarations (but not yet defaults) * Several small issues for producing compiling LaTeX; - avoid empy `\haddockbeginargs` lists (ex: `type family Any`) - properly escape identifiers depending on context (ex: `Int#`) - add `vbox` around `itemize`/`enumerate` (so they can be in tables) * Several spacing fixes: - limit the width of `Pretty`-arranged monospaced code - cut out extra space characters in export lists - only escape spaces if there are _multiple_ spaces - allow type signatures to be multiline (even without docs) * Remove uninteresting and repetitive `main.tex`/`haddock.sty` files from `latex-test` test reference output. Fixes haskell/haddock#935, haskell/haddock#929 (LaTeX docs for `text` build & compile) Fixes haskell/haddock#727, haskell/haddock#930 (I think both are really about type families...) - - - - - 0e6cee00 by Alec Theriault at 2019-03-29T12:11:56-07:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - ce05434d by Alan Zimmerman at 2019-03-29T12:12:11-07:00 Matching changes in GHC for haskell/haddock#16236 (cherry picked from commit 3ee6526d4ae7bf4deb7cd1caf24b3d7355573576) - - - - - d85766b2 by Ben Gamari at 2019-03-29T12:14:04-07:00 Bump GHC to 8.8 - - - - - 5a82cbaf by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Redo ParseModuleHeader - - - - - b9033348 by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - bb55c8f4 by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove outdated `.ghci` files and `scripts` The `.ghci` files are actively annoying when trying to `cabal v2-repl`. As for the `scripts`, the distribution workflow is completely different. - - - - - 5ee244dc by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove obsolete arcanist files + STYLE Now that GHC is hosted on Gitlab, the arcanist files don't make sense anymore. The STYLE file contains nothing more than a dead link too. - - - - - d07c1928 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Redo ParseModuleHeader - - - - - 492762d2 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - af2ac773 by Ryan Scott at 2019-05-14T17:22:13-04:00 Changes for haskell/haddock#16110/#16356 - - - - - 6820ed0d by Alec Theriault at 2019-05-17T08:51:27-07:00 Unbreak haskell/haddock#1004 test case `fail` is no longer part of `Monad`. - - - - - 6bf7be98 by Alec Theriault at 2019-05-17T08:51:27-07:00 Fix haskell/haddock#1063 with better parenthesization logic for contexts The only other change in html/hoogle/hyperlinker output for the boot libraries that this caused is a fix to some Hoogle output for implicit params. ``` $ diff -r _build/docs/ old_docs diff -r _build/docs/html/libraries/base/base.txt old_docs/html/libraries/base/base.txt 13296c13296 < assertError :: (?callStack :: CallStack) => Bool -> a -> a --- > assertError :: ?callStack :: CallStack => Bool -> a -> a ``` - - - - - b5716b61 by Ryan Scott at 2019-05-22T17:24:32-04:00 Match changes with haskell/haddock#14332 - - - - - c115abf6 by Alec Theriault at 2019-05-26T16:01:58-04:00 Remove Haddock's dependency on `Cabal` At this point, Haddock depended on Cabal-the-library solely for a verbosity parser (which misleadingly accepts all sorts of verbosity options that Haddock never uses). Now, the only dependency on Cabal is for `haddock-test` (which uses Cabal to locate the Haddock interface files of a couple boot libraries). - - - - - e5b2d4a3 by Alec Theriault at 2019-05-26T16:16:25-04:00 Regression test: promoted lists in associated types When possible, associated types with promoted lists should use the promoted list literal syntax (instead of repeated applications of ': and '[]). This was fixed in 2122de5473fd5b434af690ff9ccb1a2e58491f8c. Closes haskell/haddock#466, - - - - - cc5ad5d3 by Alec Theriault at 2019-05-26T17:55:54-04:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 4b3301a6 by Alec Theriault at 2019-05-26T17:57:52-04:00 Release haddock-2.23, haddock-library-1.8.0 Tentatively adjust bounds and changelogs for the release to be bundled with GHC 8.8.1. - - - - - 69c7cfce by Matthew Pickering at 2019-05-30T10:54:27+01:00 Update hyperlinker tests for new types in .hie files - - - - - 29b7e738 by Zubin Duggal at 2019-05-30T10:57:51+01:00 update for new way to store hiefile headers - - - - - aeca5d5f by Zubin Duggal at 2019-06-04T18:57:42-04:00 update for new way to store hiefile headers - - - - - ba2ca518 by Ben Gamari at 2019-06-07T23:11:14+00:00 Update test output for introduction of Safe-Inferred - - - - - 3a975a6c by Ryan Scott at 2019-07-03T12:06:27-04:00 Changes for haskell/haddock#15247 - - - - - 0df46555 by Zubin Duggal at 2019-07-22T10:52:50+01:00 Fix haddockHypsrcTest - - - - - 2688686b by Sylvain Henry at 2019-09-12T23:19:39+02:00 Fix for GHC module renaming - - - - - 9ec0f3fc by Alec Theriault at 2019-09-20T03:21:00-04:00 Fix Travis CI, loosen .cabal bounds (#1089) Tentatively for the 2.23 release: * updated Travis CI to work again * tweaked bounds in the `.cabal` files * adjusted `extra-source-files` to properly identify test files - - - - - ca559beb by Matthías Páll Gissurarson at 2019-09-28T12:14:40-04:00 Small change in to facilitate extended typed-holes (#1090) This change has no functional effect on haddock itself, it just changes one pattern to use `_ (` rather than `_(`, so that we may use `_(` as a token for extended typed-holes later. - - - - - 02e28976 by Vladislav Zavialov at 2019-09-28T12:17:45-04:00 Remove spaces around @-patterns (#1093) This is needed to compile `haddock` when [GHC Proposal haskell/haddock#229](https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0229-whitespace-bang-patterns.rst) is implemented. - - - - - 83cbbf55 by Alexis King at 2019-09-30T21:12:42-04:00 Fix the ignore-exports option (#1082) The `ignore-exports` option has been broken since haskell/haddock#688, as mentioned in https://github.com/haskell/haddock/pull/766#issue-172505043. This PR fixes it. - - - - - e127e0ab by Ben Gamari at 2019-10-06T15:12:06-04:00 Fix a few haddock issues - - - - - 3a0f5c89 by Zubin Duggal at 2019-10-07T17:56:13-04:00 Fix crash when there are no srcspans in the file due to CPP - - - - - 339c5ff8 by Alec Theriault at 2019-10-07T17:56:13-04:00 Prefer un-hyperlinked sources to no sources It is possible to fail to extract an HIE ast. This is however not a reason to produce _no_ output - we should still make a colorized HTML page. - - - - - d47ef478 by Alec Theriault at 2019-10-07T17:56:13-04:00 Add a regression test for haskell/haddock#1091 Previously, this input would crash Haddock. - - - - - ed7c8b0f by Alec Theriault at 2019-10-07T20:56:48-04:00 Add Hyperlinker test cases for TH-related stuff Hopefully this will guard against regressions around quasiquotes, TH quotes, and TH splices. - - - - - d00436ab by Andreas Klebinger at 2019-10-21T15:53:03+02:00 Refactor for withTiming changes. - - - - - 4230e712 by Ben Gamari at 2019-10-22T09:36:37-04:00 Merge pull request haskell/haddock#1101 from AndreasPK/withTimingRefactor Refactor for withTiming changes. - - - - - d155c5f4 by Ryan Scott at 2019-10-23T10:37:17-04:00 Reify oversaturated data family instances correctly (#1103) This fixes haskell/haddock#1103 by adapting the corresponding patch for GHC (see https://gitlab.haskell.org/ghc/ghc/issues/17296 and https://gitlab.haskell.org/ghc/ghc/merge_requests/1877). - - - - - 331a5adf by Sebastian Graf at 2019-10-25T17:14:40+02:00 Refactor for OutputableBndrId changes - - - - - 48a490e0 by Ben Gamari at 2019-10-27T10:16:16-04:00 Merge pull request haskell/haddock#1105 from sgraf812/wip/flexible-outputable Refactor for OutputableBndrId changes - - - - - f62a7dfc by Sebastian Graf at 2019-11-01T11:54:16+00:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - d9b242ed by Ryan Scott at 2019-11-03T13:20:03-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. (cherry picked from commit cfd682c5fd03b099a3d78c44f9279faf56a0ac70) - - - - - dfd42406 by Sebastian Graf at 2019-11-04T07:02:14-05:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - 0b15be7c by Ben Gamari at 2019-11-09T13:21:33-05:00 Import isRuntimeRepVar from Type rather than TyCoRep isRuntimeRepVar is not longer exported from TyCoRep due to ghc#17441. - - - - - 091f7283 by Ben Gamari at 2019-11-10T12:47:06-05:00 Bump to GHC 8.10 - - - - - e88c71f2 by Ben Gamari at 2019-11-14T00:22:24-05:00 Merge pull request haskell/haddock#1110 from haskell/wip/T17441 Import isRuntimeRepVar from Type rather than TyCoRep - - - - - 4e0bbc17 by Ben Gamari at 2019-11-14T00:22:45-05:00 Version bumps for GHC 8.11 - - - - - 0e85ceb4 by Ben Gamari at 2019-11-15T11:59:45-05:00 Bump to GHC 8.10 - - - - - 00d6d68b by Ben Gamari at 2019-11-16T18:35:58-05:00 Bump ghc version to 8.11 - - - - - dde1fc3f by Ben Gamari at 2019-11-16T20:40:37-05:00 Drop support for base 4.13 - - - - - f52e331d by Vladislav Zavialov at 2019-11-24T13:02:28+03:00 Update Hyperlinker.Parser.classify to use ITdollar - - - - - 1ad96198 by Vladislav Zavialov at 2019-11-28T16:12:33+03:00 Remove HasSrcSpan (#17494) - - - - - 651afd70 by Herbert Valerio Riedel at 2019-12-08T12:08:16+01:00 Document error-prone conditional definition of instances This can easily trip up people if one isn't aware of it. Usually it's better to avoid this kind of conditionality especially for typeclasses for which there's an compat-package as conditional instances like these tend to fragment the ecosystem into those packages that go the extra mile to provide backward compat via those compat-packages and those that fail to do so. - - - - - b521af56 by Herbert Valerio Riedel at 2019-12-08T12:09:54+01:00 Fix build-failure regression for base < 4.7 The `$>` operator definition is available only since base-4.7 which unfortunately wasn't caught before release to Hackage (but has been fixed up by a metadata-revision) This commit introduces a `CompatPrelude` module which allows to reduce the amount of CPP by ousting it to a central location, i.e. the new `CompatPrelude` module. This pattern also tends to reduce the tricks needed to silence unused import warnings. Addresses haskell/haddock#1119 - - - - - 556c375d by Sylvain Henry at 2020-01-02T19:01:55+01:00 Fix after Iface modules renaming - - - - - bd6c53e5 by Sylvain Henry at 2020-01-07T00:48:48+01:00 hsyl20-modules-renamer - - - - - fb23713b by Ryan Scott at 2020-01-08T07:41:13-05:00 Changes for GHC#17608 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2372 - - - - - 4a4dd382 by Ryan Scott at 2020-01-25T08:08:26-05:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - e782a44d by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename PackageConfig into UnitInfo - - - - - ba3c9f05 by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename lookupPackage - - - - - ab37f9b3 by Ben Gamari at 2020-01-29T13:00:44-05:00 Merge pull request haskell/haddock#1125 from haskell/wip/T17566-take-two Changes for GHC#17566 - - - - - 3ebd5ae0 by Ryan Scott at 2020-01-31T05:56:50-05:00 Merge branch 'wip-hsyl20-package-refactor' into ghc-head - - - - - 602a747e by Richard Eisenberg at 2020-02-04T09:05:43+00:00 Echo GHC's removal of PlaceHolder module This goes with GHC's !2083. - - - - - ccfe5679 by Sylvain Henry at 2020-02-10T10:13:56+01:00 Module hierarchy: runtime (cf haskell/haddock#13009) - - - - - 554914ce by Cale Gibbard at 2020-02-10T16:10:39-05:00 Fix build of haddock in stage1 We have to use the correct version of the GHC API, but the version of the compiler itself doesn't matter. - - - - - 5b6fa2a7 by John Ericson at 2020-02-10T16:18:07-05:00 Noramlize `tested-with` fields in cabal files - - - - - e6eb3ebe by Vladislav Zavialov at 2020-02-16T13:25:26+03:00 No MonadFail/Alternative for P - - - - - 90e181f7 by Ben Gamari at 2020-02-18T14:13:47-05:00 Merge pull request haskell/haddock#1129 from obsidiansystems/wip/fix-stage1-build Fix build of haddock in stage1 - - - - - 93b64636 by Sylvain Henry at 2020-02-19T11:20:27+01:00 Modules: Driver (#13009) - - - - - da4f6c7b by Vladislav Zavialov at 2020-02-22T15:33:02+03:00 Use RealSrcSpan in InstMap - - - - - 479b1b50 by Ben Gamari at 2020-02-23T10:28:13-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 55ecacf0 by Sylvain Henry at 2020-02-25T15:18:27+01:00 Modules: Core (#13009) - - - - - 60867b3b by Vladislav Zavialov at 2020-02-28T15:53:52+03:00 Ignore the BufLoc/BufSpan added in GHC's !2516 - - - - - 1e5506d3 by Sylvain Henry at 2020-03-02T12:32:43+01:00 Modules: Core (#13009) - - - - - 6fb53177 by Richard Eisenberg at 2020-03-09T14:49:40+00:00 Changes in GHC's !1913. - - - - - 30b792ea by Ben Gamari at 2020-03-16T12:45:02-04:00 Merge pull request haskell/haddock#1130 from hsyl20/wip/hsyl20-modules-core2 Modules: Core (#13009) - - - - - cd761ffa by Sylvain Henry at 2020-03-18T15:24:00+01:00 Modules: Types - - - - - b6646486 by Ben Gamari at 2020-03-18T14:42:43-04:00 Merge pull request haskell/haddock#1133 from hsyl20/wip/hsyl20/modules/types Modules: Types - - - - - 9325d734 by Kleidukos at 2020-03-19T12:38:31-04:00 Replace the 'caption' class so that the collapsible sections are shown - - - - - 5e2bb555 by Kleidukos at 2020-03-19T12:38:31-04:00 Force ghc-8.8.3 - - - - - c6fcd0aa by Kleidukos at 2020-03-19T12:38:31-04:00 Update test fixtures - - - - - 5c849cb1 by Sylvain Henry at 2020-03-20T09:34:39+01:00 Modules: Types - - - - - 7f439155 by Alec Theriault at 2020-03-20T20:17:01-04:00 Merge branch 'ghc-8.8' into ghc-8.10 - - - - - b7904e5c by Alina Banerjee at 2020-03-20T20:24:17-04:00 Update parsing to strip whitespace from table cells (#1074) * Update parsing to strip leading & trailing whitespace from table cells * Update fixture data to disallow whitespaces at both ends in table cells * Add test case for whitespaces stripped from both ends of table cells * Update table reference test data for html tests - - - - - b9d60a59 by Alec Theriault at 2020-03-22T11:46:42-04:00 Clean up warnings * unused imports * imports of `Data.List` without import lists * missing `CompatPrelude` file in `.cabal` - - - - - 0c317dbe by Alec Theriault at 2020-03-22T18:46:54-04:00 Fix NPM security warnings This was done by calling `npm audit fix`. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. - - - - - 6e306242 by Alec Theriault at 2020-03-22T20:10:52-04:00 Tentative 2.24 release Adjusted changelogs and versions in `.cabal` files in preparation for the upcoming release bundled with GHC 8.10. - - - - - 1bfb4645 by Ben Gamari at 2020-03-23T16:40:54-04:00 Merge commit '3c2944c037263b426c4fe60a3424c27b852ea71c' into HEAD More changes from the GHC types module refactoring. - - - - - be8c6f3d by Alec Theriault at 2020-03-26T20:10:53-04:00 Update `.travis.yml` to work with GHC 8.10.1 * Regenerated the Travis file with `haskell-ci` * Beef up `.cabal` files with more `tested-with` information - - - - - b025a9c6 by Alec Theriault at 2020-03-26T20:10:53-04:00 Update README Removed some out of date links/info, added some more useful links. * badge to Hackage * update old trac link * `ghc-head` => `ghc-8.10` * `cabal new-*` is now `cabal v2-*` and it should Just Work * `--test-option='--accept'` is the way to accept testsuite output - - - - - 564d889a by Alec Theriault at 2020-03-27T20:34:33-04:00 Fix crash in `haddock-library` on unicode space Our quickcheck tests for `haddock-library` stumbled across an edge case input that was causing Haddock to crash: it was a unicode space character. The root cause of the crash is that we were implicitly assuming that if a space character was not " \t\f\v\r", it would have to be "\n". We fix this by instead defining horizontal space as: any space character that is not '\n'. Fixes haskell/haddock#1142 - - - - - 2d360ba1 by Alec Theriault at 2020-03-27T21:57:32-04:00 Disallow qualified uses of reserved identifiers This a GHC bug (https://gitlab.haskell.org/ghc/ghc/issues/14109) too, but it is a relatively easy fix in Haddock. Note that the fix must live in `haddock-api` instead of `haddock-library` because we can only really decide if an identifier is a reserved one by asking the GHC lexer. Fixes haskell/haddock#952 - - - - - 47ae22ed by Alec Theriault at 2020-03-28T13:36:25-04:00 Remove unused `Haddock.Utils` functions * removed functions in `Haddock.Utils` that were not used anywhere (or exported from the `haddock-api` package) * moved GHC-specific utils from `Haddock.Utils` to `Haddock.GhcUtils` - - - - - c0291245 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use TTG empty extensions to remove some `error`'s None of these error cases should ever have been reachable, so this is just a matter of leveraging the type system to assert this. * Use the `NoExtCon` and `noExtCon` to handle case matches for no extension constructors, instead of throwing an `error`. * Use the extension field of `HsSpliceTy` to ensure that this variant of `HsType` cannot exist in an `HsType DocNameI`. - - - - - 0aff8dc4 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use `unLoc`/`noLoc` from GHC instead of `unL`/`reL` * `unL` is already defined by GHC as `unLoc` * `reL` is already defined by GHC as `noLoc` (in a safer way too!) * Condense `setOutputDir` and add a about exporting from GHC Fixes haskell/haddock#978 - - - - - bf6f2fb7 by Alec Theriault at 2020-03-28T13:36:25-04:00 Cleanup up GHC flags in `.cabal` files * enable more useful warning flags in `haddock-api`, handle the new warnings generated * remove `-fwarn-tabs` (now we'd use `-Wtabs`, but this has been in `-Wall` for a while now) - - - - - c576fbf1 by Alec Theriault at 2020-03-28T13:36:25-04:00 `haddock-library` document header level Document the fact the header level is going to always be between 1 and 6 inclusive. Along the way, I also optimized the parsing code a bit. - - - - - 71bce0ee by Alec Theriault at 2020-03-28T14:26:27-04:00 Disallow links in section headers This is quite straightforward to implement, since we already had a function `docToHtmlNoAnchors` (which we used to generate the link in the sidebar "Contents"). This breaks test `Bug387`, but that test case has aged badly: we now automatically generate anchors for all headings, so manually adding an anchor in a section makes no sense. Nested anchors are, as pointed out in haskell/haddock#1054, disallowed by the HTML standard. Fixes haskell/haddock#1054 - - - - - b461b0ed by Sylvain Henry at 2020-03-30T10:34:23+02:00 Modules: type checker - - - - - cd8cd1ee by Ben Gamari at 2020-03-31T12:45:02-04:00 Merge pull request haskell/haddock#1152 from hsyl20/wip/hsyl20/modules/tc Module renaming - - - - - 5e8f8ea7 by Felix Yan at 2020-04-01T17:58:06-07:00 Allow QuickCheck 2.14 Builds fine and all tests pass. - - - - - dc6b1633 by Sylvain Henry at 2020-04-05T16:43:44+02:00 Module renaming: amend previous patch - - - - - eee2f4ae by Ryan Scott at 2020-04-05T09:04:43-07:00 Fix haskell/haddock#1050 by filtering out invisible AppTy arguments This makes the `synifyType` case for `AppTy` more intelligent by taking into consideration the visibilities of each `AppTy` argument and filtering out any invisible arguments, as they aren't intended to be displayed in the source code. (See haskell/haddock#1050 for an example of what can happen if you fail to filter these out.) Along the way, I noticed that a special `synifyType` case for `AppTy t1 (CoercionTy {})` could be consolidated with the case below it, so I took the opportunity to tidy this up. - - - - - 23eb99e8 by Ben Gamari at 2020-04-07T11:19:58-04:00 Merge pull request haskell/haddock#1154 from hsyl20/wip/hsyl20/modules/tc Module renaming: amend previous patch - - - - - 072d994d by Ryan Scott at 2020-04-07T19:32:47-04:00 Make NoExtCon fields strict These changes are a part of a fix for [GHC#17992](https://gitlab.haskell.org/ghc/ghc/issues/17992). - - - - - d8ebf6c8 by Ignat Insarov at 2020-04-09T21:15:01-04:00 Recode Doc to Json. (#1159) * Recode Doc to Json. * More descriptive field labels. - - - - - 52df4b4e by Sylvain Henry at 2020-04-10T12:39:18+02:00 Module renaming - - - - - d9ab8ec8 by Cale Gibbard at 2020-04-14T11:43:34-04:00 Add instance of XCollectPat for DocNameI - - - - - 323d221d by Cale Gibbard at 2020-04-14T11:43:34-04:00 Rename XCollectPat -> CollectPass - - - - - 2df80867 by Alec Theriault at 2020-04-15T07:30:51-07:00 Prune docstrings that are never rendered When first creating a Haddock interface, trim `ifaceDocMap` and `ifaceArgMap` to not include docstrings that can never appear in the final output. Besides checking with GHC which names are exported, we also need to keep all the docs attached to instance declarations (it is much tougher to detect when an instance is fully private). This change means: * slightly smaller interface files (7% reduction on boot libs) * slightly less work to do processing docstrings that aren't used * no warnings in Haddock's output about private docstrings (see haskell/haddock#1070) I've tested manually that this does not affect any of the boot library generated docs (the only change in output was some small re-ordering in a handful of instance lists). This should mean no docstrings have been incorrectly dropped. - - - - - f49c90cc by Alec Theriault at 2020-04-15T07:30:51-07:00 Don't warn about missing links in miminal sigs When renaming the Haddock interface, never emit warnings when renaming a minimal signature. Also added some documention around `renameInterface`. Minimal signatures intentionally include references to potentially un-exported methods (see the discussion in haskell/haddock#330), so it is expected that they will not always have a link destination. On the principle that warnings should always be resolvable, this shouldn't produce a warning. See haskell/haddock#1070. - - - - - a9eda64d by Ben Gamari at 2020-04-17T09:27:35-04:00 Merge pull request haskell/haddock#1160 from hsyl20/wip/hsyl20/modules/systools Module renaming - - - - - f40d7879 by Cale Gibbard at 2020-04-20T11:30:38-04:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ttg-con-pat - - - - - a50e7753 by Ben Gamari at 2020-04-20T11:36:10-04:00 Merge pull request haskell/haddock#1165 from obsidiansystems/wip/ttg-con-pat Trees that Grow refactor (GHC !2553) - - - - - 6a24795c by Alec Theriault at 2020-04-21T08:06:45-07:00 Fallback to `hiDecl` when `extractDecl` fails Sometimes, the declaration being exported is a subdecl (for instance, a record accessor getting exported at the top-level). For these cases, Haddock has to find a way to produce some synthetic sensible top-level declaration. This is done with `extractDecl`. As is shown by haskell/haddock#1067, this is sometimes impossible to do just at a syntactic level (for instance when the subdecl is re-exported). In these cases, the only sensible thing to do is to try to reify a declaration based on a GHC `TyThing` via `hiDecl`. - - - - - eee1a8b7 by Sylvain Henry at 2020-04-24T15:46:05+02:00 Module structure - - - - - 50b9259c by Iñaki at 2020-04-25T18:38:11-04:00 Add support for custom section anchors (#1179) This allows to have stable anchors for groups, even if the set of groups in the documentation is altered. The syntax for setting the anchor of a group is -- * Group name #desiredAnchor# Which will produce an html anchor of the form '#g:desiredAnchor' Co-authored-by: Iñaki García Etxebarria <git at inaki.blueleaf.cc> - - - - - 4003c97a by Ben Gamari at 2020-04-26T09:35:15-04:00 Merge pull request haskell/haddock#1166 from hsyl20/wip/hsyl20/modules/utils Module structure - - - - - 5206ab60 by Sylvain Henry at 2020-04-27T16:47:39+02:00 Renamed UnitInfo fields - - - - - c32c333b by Sylvain Henry at 2020-04-27T17:32:58+02:00 UnitId has been renamed into Unit - - - - - 3e87db64 by Sylvain Henry at 2020-04-27T17:36:00+02:00 Fix for GHC.Unit.* modules - - - - - ae3323a7 by Ben Gamari at 2020-04-29T12:36:37-04:00 Merge pull request haskell/haddock#1183 from hsyl20/wip/hsyl20/unitid Refactoring of Unit code - - - - - b105564a by Artem Pelenitsyn at 2020-05-03T08:14:10+01:00 add dependency on exceptions because GHC.Exception was boiled down (ghc haskell/haddock#18075) - - - - - 9857eff3 by Zubin Duggal at 2020-05-04T18:48:25+01:00 Atomic update of NameCache in readHieFile - - - - - 86bbb226 by Sylvain Henry at 2020-05-14T16:36:27+02:00 Fix after Config module renaming - - - - - a4bbdbc2 by Gert-Jan Bottu at 2020-05-15T22:09:44+02:00 Explicit Specificity Support for Haddock - - - - - 46199daf by Ben Gamari at 2020-05-19T09:59:56-04:00 Merge pull request haskell/haddock#1192 from hsyl20/hsyl20/modules-config Fix after Config module renaming - - - - - f9a9d2ba by Gert-Jan Bottu at 2020-05-20T16:48:38-04:00 Explicit Specificity Support for Haddock - - - - - 55c5b7ea by Ben Gamari at 2020-05-21T00:32:02-04:00 Merge commit 'a8d7e66da4dcc3b242103271875261604be42d6e' into ghc-head - - - - - a566557f by Cale Gibbard at 2020-05-21T16:02:06-04:00 isBootSummary now produces a result of type IsBootInterface - - - - - ea52f905 by Zubin Duggal at 2020-05-24T17:55:48+01:00 update for hiefile-typeclass-info - - - - - 49ba7a67 by Willem Van Onsem at 2020-05-25T12:23:01-04:00 Use floor over round to calculate the percentage (#1195) If we compile documentation where only a small fraction is undocumented, it is misleading to see 100% coverage - 99% is more intuitive. Fixes haskell/haddock#1194 - - - - - c025ebf1 by Ben Gamari at 2020-05-29T14:32:42-04:00 Merge pull request haskell/haddock#1185 from obsidiansystems/boot-disambig isBootSummary now produces a result of type IsBootInterface - - - - - 74ab9415 by Ben Gamari at 2020-05-29T20:23:39-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - b40be944 by Ben Gamari at 2020-06-03T17:02:31-04:00 testsuite: Update expected output for simplified subsumption - - - - - 624be71c by Ryan Scott at 2020-06-05T12:43:23-04:00 Changes for GHC#18191 See https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3337. - - - - - fbd8f7ce by Sylvain Henry at 2020-06-08T15:31:47+02:00 Fix after unit refactoring - - - - - 743fda4d by Ben Gamari at 2020-06-09T12:09:58-04:00 Merge pull request haskell/haddock#1202 from hsyl20/wip/hsyl20/unitid-ii Fix after unit refactoring - - - - - d07a06a9 by Ryan Scott at 2020-06-13T07:16:55-04:00 Use HsForAllTelescope (GHC#18235) - - - - - 389bb60d by Ben Gamari at 2020-06-13T15:30:52-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - 7a377f5f by Ben Gamari at 2020-06-17T14:53:16-04:00 Merge pull request haskell/haddock#1199 from bgamari/wip/ghc-8.12 haddock: Bounds bumps for GHC 8.12 - - - - - 9fd9e586 by Krzysztof Gogolewski at 2020-06-17T16:09:07-04:00 Adapt Haddock to LinearTypes See ghc/ghc!852. - - - - - 46fe7636 by Ben Gamari at 2020-06-18T14:20:02-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 35a3c9e2 by Zubin Duggal at 2020-06-21T21:19:18+05:30 Use functions exported from HsToCore - - - - - 8abe3928 by Ben Gamari at 2020-06-24T13:53:39-04:00 Merge pull request haskell/haddock#1204 from wz1000/wip/haddock-hstocore Use functions exported from GHC.HsToCore.Docs - - - - - 22f2c937 by Matthías Páll Gissurarson at 2020-06-26T19:07:03+02:00 Adapt Haddock for QualifiedDo - - - - - 3f6208d7 by Vladislav Zavialov at 2020-06-28T14:28:16+03:00 Handle LexicalNegation's ITprefixminus - - - - - 03a19f41 by Sylvain Henry at 2020-07-02T09:37:38+02:00 Rename hsctarget into backend - - - - - ea17ff23 by Andreas Klebinger at 2020-07-02T17:44:18+02:00 Update for UniqFM changes. - - - - - 9872f2f3 by Ben Gamari at 2020-07-09T10:39:19-04:00 Merge pull request haskell/haddock#1209 from AndreasPK/wip/typed_uniqfm Update for UniqFM changes. - - - - - 68f7b668 by Krzysztof Gogolewski at 2020-07-12T18:16:57+02:00 Sync with GHC removing {-# CORE #-} pragma See ghc ticket haskell/haddock#18048 - - - - - eb372681 by Sylvain Henry at 2020-07-20T11:41:30+02:00 Rename hscTarget into backend - - - - - fb7f78bf by Ben Gamari at 2020-07-21T12:15:25-04:00 Merge pull request haskell/haddock#1214 from hsyl20/wip/hsyl20/hadrian/ncg Rename hscTarget into backend - - - - - 1e8f5b56 by Ben Gamari at 2020-07-23T09:11:50-04:00 Merge commit '904dce0cafe0a241dd3ef355775db47fc12f434d' into ghc-head - - - - - d8fd1775 by Zubin Duggal at 2020-07-23T18:46:40+05:30 Update for modular ping pong - - - - - 8416f872 by Ben Gamari at 2020-07-23T09:35:03-04:00 Merge pull request haskell/haddock#1200 from wz1000/wip/wz1000-modular-ping-pong Modular ping pong - - - - - a24a8577 by Ben Gamari at 2020-07-28T15:23:36-04:00 Bump GHC version to 9.0 - - - - - 6a51c9dd by Sylvain Henry at 2020-08-05T18:47:05+02:00 Fix after Outputable refactoring - - - - - c05e1c99 by Ben Gamari at 2020-08-10T14:41:41-04:00 Merge pull request haskell/haddock#1223 from hsyl20/wip/hsyl20/dynflags/exception Fix after Outputable refactoring - - - - - d964f15b by Sylvain Henry at 2020-08-12T11:58:49+02:00 Fix after HomeUnit - - - - - 8e6d5b23 by Ben Gamari at 2020-08-12T14:25:30-04:00 Merge pull request haskell/haddock#1225 from hsyl20/wip/hsyl20/plugins/homeunit Fix after HomeUnit - - - - - 8c7880fe by Sylvain Henry at 2020-08-17T14:13:29+02:00 Remove Ord FastString instance - - - - - 8ea410db by Alex Biehl at 2020-08-19T10:56:32+02:00 Another round of `npm audit fix` (#1228) This should shut down the warnings on Github. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. Last NPM dependency audit happend in d576b2327e2bc117f912fe0a9d595e9ae62614e0 Co-authored-by: Alex Biehl <alex.biehl at target.com> - - - - - 7af6e2a8 by Ben Gamari at 2020-08-31T13:59:34-04:00 Merge pull request haskell/haddock#1226 from hsyl20/wip/hsyl20/fs_ord Remove Ord FastString instance - - - - - ffbc8702 by Alan Zimmerman at 2020-09-07T21:47:41+01:00 Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - a93f1268 by Alan Zimmerman at 2020-09-07T23:11:38+01:00 Merge pull request haskell/haddock#1232 from haskell/wip/T18639-remove-generated-pragma, Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - 1f605d50 by Ben Gamari at 2020-09-14T18:30:01-04:00 Bump GHC version to 9.1 - - - - - 6599df62 by Vladislav Zavialov at 2020-09-18T14:05:15+03:00 Bump base upper bound to 4.16 - - - - - a01b3c43 by Ben Gamari at 2020-09-22T15:41:48-04:00 Update hypsrc-test for QuickLook This appears to be a spurious change. - - - - - e9cc6cac by Vladislav Zavialov at 2020-09-26T21:00:12+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 30e3ca7c by Sylvain Henry at 2020-09-29T11:18:32-04:00 Update for parser (#1234) - - - - - b172f3e3 by Vladislav Zavialov at 2020-09-30T01:01:30+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 0b9c08d3 by Sylvain Henry at 2020-09-30T11:02:33+02:00 Adapt to GHC parser changes - - - - - b9540b7a by Sylvain Henry at 2020-10-12T09:13:38-04:00 Don't pass the HomeUnitId (#1239) - - - - - 34762e80 by HaskellMouse at 2020-10-13T12:58:04+03:00 Changed tests due to unification of `Nat` and `Natural` in the follwing merge request: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3583 - - - - - 256f86b6 by Vladislav Zavialov at 2020-10-15T10:48:03+03:00 Add whitespace in: map ($ v) - - - - - 4a3f711b by Alan Zimmerman at 2020-10-19T08:57:27+01:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled - - - - - 072cdd21 by Alan Zimmerman at 2020-10-21T14:48:28-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 9e09a445 by Alan Zimmerman at 2020-10-21T23:53:34-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 636d7de3 by Sylvain Henry at 2020-10-26T14:31:54-04:00 GHC.Driver.Types refactoring (#1242) - - - - - a597f000 by Ryan Scott at 2020-10-29T04:18:05-04:00 Adapt to the removal of Hs{Boxed,Constraint}Tuple See ghc/ghc!4097 and GHC#18723. - - - - - b96660fb by Ryan Scott at 2020-10-30T04:53:05-04:00 Adapt to HsConDecl{H98,GADT}Details split Needed for GHC#18844. - - - - - c287d82c by Ryan Scott at 2020-10-30T19:35:59-04:00 Adapt to HsOuterTyVarBndrs These changes accompany ghc/ghc!4107, which aims to be a fix for haskell/haddock#16762. - - - - - a34c31a1 by Ryan Scott at 2020-11-13T13:38:34-05:00 Adapt to splitPiTysInvisible being renamed to splitInvisPiTys This is a part of !4434, a fix for GHC#18939. - - - - - 66ea459d by Sylvain Henry at 2020-11-16T10:59:30+01:00 Fix after Plugins moved into HscEnv - - - - - 508556d8 by Ben Gamari at 2020-11-18T15:47:40-05:00 Merge pull request haskell/haddock#1253 from hsyl20/wip/hsyl20/plugins/hscenv Fix after Plugins moved into HscEnv - - - - - 620fec1a by Andreas Klebinger at 2020-11-24T20:51:59+01:00 Update for changes in GHC's Pretty - - - - - 01cc13ab by Richard Eisenberg at 2020-11-25T23:18:35-05:00 Avoid GHC#18932. - - - - - 8d29ba21 by Cale Gibbard at 2020-11-25T23:18:35-05:00 Add type arguments to PrefixCon - - - - - 414d5f87 by Sylvain Henry at 2020-11-30T17:06:04+01:00 DynFlags's unit fields moved to HscEnv - - - - - e356668c by Ben Gamari at 2020-11-30T11:11:37-05:00 Merge pull request haskell/haddock#1258 from hsyl20/wip/hsyl20/hscenv/unitstate Unit fields moved from DynFlags to HscEnv - - - - - 7cf552f1 by Ben Gamari at 2020-12-03T10:31:27-05:00 Merge pull request haskell/haddock#1257 from AndreasPK/wip/andreask/opt_dumps Update for changes in GHC's Pretty - - - - - fc0871c3 by Veronika Romashkina at 2020-12-08T16:35:33+01:00 Fix docs links from Darcs to GitHub in intro (#1262) - - - - - 7059e808 by Veronika Romashkina at 2020-12-08T16:36:16+01:00 Use gender neutral word in docs (#1260) - - - - - 1b16e5ee by Maximilian Tagher at 2020-12-08T16:40:03+01:00 Allow scrolling search results (#1235) Closes https://github.com/haskell/haddock/issues/1231 - - - - - 8a118c01 by dependabot[bot] at 2020-12-08T16:40:25+01:00 Bump bl from 1.2.2 to 1.2.3 in /haddock-api/resources/html (#1255) Bumps [bl](https://github.com/rvagg/bl) from 1.2.2 to 1.2.3. - [Release notes](https://github.com/rvagg/bl/releases) - [Commits](https://github.com/rvagg/bl/compare/v1.2.2...v1.2.3) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - c89ff587 by Xia Li-yao at 2020-12-08T16:42:17+01:00 Allow more characters in anchor following module reference (#1220) - - - - - 14af7d64 by Xia Li-yao at 2020-12-08T16:43:05+01:00 Add dangling changes from branches ghc-8.6 and ghc-8.8 (#1243) * Fix multiple typos and inconsistencies in doc/markup.rst Note: I noticed some overlap with haskell/haddock#1112 from @wygulmage and haskell/haddock#1081 from @parsonsmatt after creating these proposed changes - mea culpa for not looking at the open PRs sooner. * Fix haskell/haddock#1113 If no Signatures, no section of index.html * Change the formatting of missing link destinations The current formatting of the missing link destination does not really help user to understand the reasons of the missing link. To address this, I've changed the formatting in two ways: - the missing link symbol name is now fully qualified. This way you immediately know which haskell module cannot be linked. It is then easier to understand why this module does not have documentation (hidden module or broken documentation). - one line per missing link, that's more readable now that symbol name can be longer due to qualification. For example, before haddock was listing missing symbol such as: ``` could not find link destinations for: Word8 Word16 mapMaybe ``` Now it is listed as: ``` could not find link destinations for: - Data.Word.Word8 - Data.Word.Word16 - Data.Maybe.mapMaybe ``` * Add `--ignore-link-symbol` command line argument This argument can be used multiples time. A missing link to a symbol listed by `--ignore-link-symbol` won't trigger "missing link" warning. * Forbid spaces in anchors (#1148) * Improve error messages with context information (#1060) Co-authored-by: Matt Audesse <matt at mattaudesse.com> Co-authored-by: Mike Pilgrem <mpilgrem at users.noreply.github.com> Co-authored-by: Guillaume Bouchard <guillaume.bouchard at tweag.io> Co-authored-by: Pepe Iborra <pepeiborra at gmail.com> - - - - - 89e3af13 by tomjaguarpaw at 2020-12-08T18:00:04+01:00 Enable two warnings (#1245) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - c3320f8d by Willem Van Onsem at 2020-12-08T18:26:55+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 685df308 by Alex Biehl at 2020-12-08T20:06:26+01:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - be3ec3c0 by Alex Biehl at 2020-12-08T20:06:26+01:00 Import intercalate - - - - - 32c33912 by Matthías Páll Gissurarson at 2020-12-08T21:15:30+01:00 Adapt Haddock for QualifiedDo - - - - - 31696088 by Alex Biehl at 2020-12-08T22:06:02+01:00 Fix haddock-library tests - - - - - fbc0998a by Alex Biehl at 2020-12-08T23:08:23+01:00 Move to GitHub CI (#1266) * Initial version of ci.yml This is a straight copy from Dmitrii Kovanikov's blog post at https://kodimensional.dev/github-actions. Will adapt to haddock in successive commits. * Delete .travis.yml * Modify to only test on ghc-8.10.{1,2} * Use actions/setup-haskell at v1.1.4 * Relax QuickCheck bound on haddock-api * Remove stack matrix for now * Nail down to ghc-8.10 branch for now * Pin index state to 2020-12-08T20:13:44Z for now * Disable macOS and Windows tests for now for speed up - - - - - 5b946b9a by tomjaguarpaw at 2020-12-10T19:01:41+01:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - bc5a408f by dependabot[bot] at 2020-12-10T19:02:16+01:00 Bump ini from 1.3.5 to 1.3.7 in /haddock-api/resources/html (#1269) Bumps [ini](https://github.com/isaacs/ini) from 1.3.5 to 1.3.7. - [Release notes](https://github.com/isaacs/ini/releases) - [Commits](https://github.com/isaacs/ini/compare/v1.3.5...v1.3.7) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - d02995f1 by Andrew Martin at 2020-12-14T16:48:40-05:00 Update for boxed rep - - - - - a381aeff by Ben Gamari at 2020-12-15T15:13:30-05:00 Revert "Enable two warnings (#1245) (#1268)" As this does not build on GHC `master`. This reverts commit 7936692badfe38f23ae95b51fb7bd7c2ff7e9bce. - - - - - a63c0a9e by Ben Gamari at 2020-12-15T15:17:59-05:00 Revert "Update for boxed rep" This reverts commit 4ffb30d8b637ccebecc81ce610f0af451ac8088d. - - - - - 53bfbb29 by Ben Gamari at 2020-12-15T15:37:24-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - bae76a30 by Ben Gamari at 2020-12-16T02:44:42+00:00 Update output for nullary TyConApp optimisation (ghc/ghc!2952) - - - - - 4b733b57 by Krzysztof Gogolewski at 2020-12-16T20:03:14+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. - - - - - ee463bd3 by Ryan Scott at 2020-12-16T16:55:23-05:00 Adapt to HsCoreTy (formerly NewHsTypeX) becoming a type synonym Needed for !4417, the fix for GHC#15706 and GHC#18914. - - - - - ed0b02f8 by tomjaguarpaw at 2020-12-19T10:17:19+00:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - d80bf8f5 by Sylvain Henry at 2020-12-21T10:09:25+01:00 Fix after binder collect changes - - - - - bf4c9d32 by Adam Gundry at 2020-12-23T21:35:01+00:00 Adapt to changes to GlobalRdrElt and AvailInfo Needed for ghc/ghc!4467 - - - - - 37736c4c by John Ericson at 2020-12-28T12:27:02-05:00 Support a new ghc --make node type for parallel backpack upsweep - - - - - 717bdeac by Vladislav Zavialov at 2020-12-29T10:50:02+03:00 Inline and fix getGADTConTypeG The getGADTConTypeG used HsRecTy, which is at odds with GHC issue haskell/haddock#18782. I noticed that getGADTConTypeG was only used in the Hoogle backend. Interestingly, when handling H98 constructors, Hoogle converts RecCon to PrefixCon (see Haddock.Backends.Hoogle.ppCtor). So I changed getGADTConTypeG to handle RecConGADT in the same manner as PrefixConGADT, and after this simplification moved it into the 'where' clause of ppCtor, to the only place where it is used. The practical effect of this change is as follows. Consider this example: data TestH98 = T98 { bar::Int } data TestGADT where TG :: { foo :: Int } -> TestGADT Before this patch, haddock --hoogle used to produce: T98 :: Int -> TestH98 [TG] :: {foo :: Int} -> TestGADT Notice how the record syntax was discarded in T98 but not TG. With this patch, we always produce signatures without record syntax: T98 :: Int -> TestH98 [TG] :: Int -> TestGADT I suspect this might also be a bugfix, as currently Hoogle doesn't seem to render GADT record constructors properly. - - - - - cb1b8c56 by Andreas Abel at 2020-12-30T21:12:37+01:00 Build instructions: haddock-library and -api first! - - - - - b947f6ad by Ben Gamari at 2020-12-31T13:04:19-05:00 Merge pull request haskell/haddock#1281 from obsidiansystems/wip/backpack-j Changes to support -j with backpack - - - - - 120e1cfd by Hécate Moonlight at 2021-01-04T19:54:58+01:00 Merge pull request haskell/haddock#1282 from andreasabel/master Build instructions: haddock-library and -api first! - - - - - fd45e41a by Ben Gamari at 2021-01-05T16:14:31-05:00 Merge remote-tracking branch 'origin/ghc-8.10' into ghc-9.0 - - - - - b471bdec by Ben Gamari at 2021-01-05T16:23:02-05:00 Merge commit '1e56f63c3197e7ca1c1e506e083c2bad25d08793' into ghc-9.0 - - - - - 81cdbc41 by Alex Biehl at 2021-01-09T12:14:41+01:00 Prepare Haddock for being a GHC Plugin - - - - - b646d952 by Alex Biehl at 2021-01-09T12:14:41+01:00 Make Haddock a GHC Plugin - - - - - cc044674 by Alex Biehl at 2021-01-09T12:14:41+01:00 Add -j[n] CLI param to Haddock executable It translates to `--ghcopt=-j[n]` - - - - - 84a04073 by Alex Biehl at 2021-01-09T12:14:41+01:00 Abstract Monad for interface creation I found that when running as a plugin the lookupName function (which runs in Ghc monad) does not work correctly from the typeCheckResultAction hook. Instead, we abstracted the monad used when creating interfaces, so that access to GHC session specific parts is explicit and so that the TcM can provide their (correct) implementation of lookupName. - - - - - 5be2c4f7 by Alex Biehl at 2021-01-09T12:14:41+01:00 Accept tests - - - - - 8cefee9d by Alex Biehl at 2021-01-09T16:10:47+01:00 Add missing dependency for mtl - - - - - 3681f919 by Ben Gamari at 2021-01-13T18:39:25-05:00 Merge remote-tracking branch 'origin/ghc-9.0' into ghc-head - - - - - 33c6b152 by Hécate Moonlight at 2021-01-14T16:04:20+01:00 Merge pull request haskell/haddock#1273 from hsyl20/wip/hsyl20/arrows Fix after binder collect changes - - - - - 70d13e8e by Joachim Breitner at 2021-01-22T19:03:45+01:00 Make haddock more robust to changes to the `Language` data type With the introduction of GHC2021, the `Languages` data type in GHC will grow. In preparation of that (and to avoid changing haddock with each new language), this change makes the code handle extensions to that data type gracefully. (cherry picked from commit c341dd7c9c3fc5ebc83a2d577c5a726f3eb152a5) - - - - - 7d6dd57a by John Ericson at 2021-01-22T22:02:02+00:00 Add `NoGhcTc` instance now that it's not closed - - - - - e5fdaf0a by Alan Zimmerman at 2021-01-23T22:57:44+00:00 Merge pull request haskell/haddock#1293 from obsidiansystems/wip/fix-18936 Add `NoGhcTc` instance now that it's not closed - - - - - 989a1e05 by Oleg Grenrus at 2021-01-24T16:11:46+03:00 Add import list to Data.List - - - - - 368e144a by Ben Gamari at 2021-01-28T22:15:48+01:00 Adapt to "Make PatSyn immutable" - - - - - abe66c21 by Alfredo Di Napoli at 2021-02-01T08:05:35+01:00 Rename pprLogErrMsg to new name - - - - - e600e75c by Hécate Moonlight at 2021-02-05T14:53:00+01:00 Move CI to ghc-9.0 - - - - - dd492961 by Vladislav Zavialov at 2021-02-05T14:53:00+01:00 Update cabal.project and README build instructions - - - - - 31bd292a by Hécate Moonlight at 2021-02-05T15:03:56+01:00 Merge pull request haskell/haddock#1296 from Kleidukos/ghc-9.0 Merge the late additions to ghc-8.10 into ghc-9.0 - - - - - 6388989e by Vladislav Zavialov at 2021-02-05T17:41:57+03:00 Cleanup: fix build warnings - - - - - f99407ef by Daniel Rogozin at 2021-02-05T18:11:48+03:00 type level characters support for haddock (required for haskell/haddock#11342) - - - - - d8c6b26f by Hécate Moonlight at 2021-02-05T17:44:50+01:00 Add a CONTRIBUTING.md file - - - - - 6a01ad98 by Hécate Moonlight at 2021-02-05T17:58:16+01:00 Merge pull request haskell/haddock#1312 from Kleidukos/proper-branch-etiquette Add a CONTRIBUTING.md file - - - - - 955eecc4 by Vladislav Zavialov at 2021-02-05T20:29:00+03:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into ghc-head - - - - - 47b3d6ab by Hécate Moonlight at 2021-02-05T19:09:38+01:00 Amend the CONTRIBUTING.md file - - - - - 23de6137 by Hécate Moonlight at 2021-02-05T19:16:49+01:00 Merge pull request haskell/haddock#1313 from Kleidukos/amend-contributing Amend the CONTRIBUTING.md file - - - - - 69026b59 by Krzysztof Gogolewski at 2021-02-05T23:05:56+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. (cherry picked from commit b4b4d896d2d68d6c48e7db7bfe95c185ca0709cb) - - - - - ea026b78 by Oleg Grenrus at 2021-02-06T17:14:45+01:00 Add import list to Data.List - - - - - 5204326f by Hécate Moonlight at 2021-02-06T17:15:44+01:00 Merge pull request haskell/haddock#1316 from Kleidukos/explicit-imports-to-data-list Add import list to Data.List - - - - - 1f4d2136 by Ben Gamari at 2021-02-06T11:53:31-05:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ghc-head-merge - - - - - 13f0d09a by Ben Gamari at 2021-02-06T11:53:45-05:00 Fix partial record selector warning - - - - - 5c115f7e by Ben Gamari at 2021-02-06T11:55:52-05:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into wip/ghc-head-merge - - - - - b6fd8b75 by Ben Gamari at 2021-02-06T12:01:31-05:00 Merge commit '41964cb2fd54b5a10f8c0f28147015b7d5ad2c02' into wip/ghc-head-merge - - - - - a967194c by Ben Gamari at 2021-02-06T18:30:35-05:00 Merge branch 'wip/ghc-head-merge' into ghc-head - - - - - 1f4c3a91 by MorrowM at 2021-02-07T01:52:33+02:00 Fix search div not scrolling - - - - - 684b1287 by Iñaki García Etxebarria at 2021-02-07T16:13:04+01:00 Add support for labeled module references Support a markdown-style way of annotating module references. For instance -- | [label]("Module.Name#anchor") will create a link that points to the same place as the module reference "Module.Name#anchor" but the text displayed on the link will be "label". - - - - - bdb55a5d by Hécate Moonlight at 2021-02-07T16:18:10+01:00 Merge pull request haskell/haddock#1319 from alexbiehl/alex/compat Backward compat: Add support for labeled module references - - - - - 6ca70991 by Hécate Moonlight at 2021-02-07T16:21:29+01:00 Merge pull request haskell/haddock#1314 from tweag/show-linear-backport Backport haskell/haddock#1238 (linear types) to ghc-9.0 - - - - - d9d73298 by Alex Biehl at 2021-02-07T17:46:25+01:00 Remove dubious parseModLink Instead construct the ModLink value directly when parsing. - - - - - 33b4d020 by Hécate Moonlight at 2021-02-07T17:52:05+01:00 Merge pull request haskell/haddock#1320 from haskell/alex/fix Remove dubious parseModLink - - - - - 54211316 by Hécate Moonlight at 2021-02-07T18:12:07+01:00 Merge pull request haskell/haddock#1318 from MorrowM/ghc-9.0 Fix search div not scrolling - - - - - 19db679e by alexbiehl-gc at 2021-02-07T18:14:46+01:00 Merge pull request haskell/haddock#1317 from bgamari/wip/ghc-head-merge Merge ghc-8.10 into ghc-head - - - - - 6bc1e9e4 by Willem Van Onsem at 2021-02-07T18:25:30+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - c8537cf8 by alexbiehl-gc at 2021-02-07T18:30:40+01:00 Merge pull request haskell/haddock#1322 from haskell/alex/forward-port simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 2d47ae4e by alexbiehl-gc at 2021-02-07T18:39:59+01:00 Merge branch 'ghc-head' into ghc-9.0 - - - - - 849e4733 by Hécate Moonlight at 2021-02-07T18:43:19+01:00 Merge pull request haskell/haddock#1321 from Kleidukos/ghc-9.0 Merge ghc-9.0 into ghc-head - - - - - ee6095d7 by Sylvain Henry at 2021-02-08T11:36:38+01:00 Update for Logger - - - - - 4ad688c9 by Alex Biehl at 2021-02-08T18:11:24+01:00 Merge pull request haskell/haddock#1310 from hsyl20/wip/hsyl20/logger2 Logger refactoring - - - - - 922a9e0e by Ben Gamari at 2021-02-08T12:54:33-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - 991649d2 by Sylvain Henry at 2021-02-09T10:55:17+01:00 Fix to build with HEAD - - - - - a8348dc2 by Hécate Moonlight at 2021-02-09T10:58:51+01:00 Merge pull request haskell/haddock#1327 from hsyl20/wip/hsyl20/logger2 Fix to build with HEAD - - - - - 0abdbca6 by Fendor at 2021-02-09T20:06:15+01:00 Add UnitId to Target record - - - - - d5790a0e by Alex Biehl at 2021-02-11T10:32:32+01:00 Stable sort for (data/newtype) instances - - - - - 8e6036f5 by Alex Biehl at 2021-02-11T10:32:32+01:00 Also make TyLit deterministic - - - - - f76d2945 by Hécate Moonlight at 2021-02-11T11:00:31+01:00 Merge pull request haskell/haddock#1329 from hsyl20/hsyl20/stabe_iface Stable sort for instances - - - - - 5e0469ea by Oleg Grenrus at 2021-02-14T15:28:15+02:00 Add import list to Data.List in Haddock.Interface.Create - - - - - fa57cd24 by Hécate Moonlight at 2021-02-14T17:19:27+01:00 Merge pull request haskell/haddock#1331 from phadej/more-explicit-data-list-imports Add import list to Data.List in Haddock.Interface.Create - - - - - f0cd629c by Hécate Moonlight at 2021-02-21T00:22:01+01:00 Merge pull request haskell/haddock#1311 from fendor/wip/add-targetUnitId-to-target Add UnitId to Target record - - - - - 674ef723 by Joachim Breitner at 2021-02-22T10:39:18+01:00 html-test: Always set language from ghc-9.2 on, the “default” langauge of GHC is expected to change more wildly. To prepare for that (and unblock https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4853), this sets the language for all the test files to `Haskell2010`. This should insolate this test suite against changes to the default. Cherry-picked from https://github.com/haskell/haddock/pull/1341 - - - - - f072d623 by Hécate Moonlight at 2021-02-22T10:56:51+01:00 Merge pull request haskell/haddock#1342 from nomeata/joachim/haskell2010-in-tests-ghc-head html-test: Always set language - - - - - caebbfca by Hécate Moonlight at 2021-02-22T11:53:07+01:00 Clean-up of Interface and Interface.Create's imports and pragmata - - - - - f6caa89d by Hécate Moonlight at 2021-02-22T11:54:58+01:00 Merge pull request haskell/haddock#1345 from Kleidukos/head/fix-interface-imports [ghc-head] Clean-up of Interface and Interface.Create's imports and pragmata - - - - - 7395c9cb by Hécate Moonlight at 2021-02-22T18:44:57+01:00 Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 6e9fb5d5 by Hécate Moonlight at 2021-02-22T18:45:28+01:00 Merge pull request haskell/haddock#1348 from Kleidukos/head/explicit-imports-interface Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 9198b118 by Alan Zimmerman at 2021-02-22T20:04:24+00:00 Context becomes a Maybe in the GHC AST This prevents noLoc's appearing in the ParsedSource. Match the change in GHC. - - - - - 0af20f64 by Hécate Moonlight at 2021-02-23T12:36:12+01:00 Fix the call-site of guessTarget in Interface.hs Explicit the imports from GHC.HsToCore.Docs - - - - - b7886885 by Hécate Moonlight at 2021-02-23T12:37:54+01:00 Merge pull request haskell/haddock#1349 from Kleidukos/fix-interface-guesstarget-call Fix the call-site of guessTarget in Interface.hs - - - - - 9cf041ba by Sylvain Henry at 2021-02-24T11:08:20+01:00 Fix haddockHypsrcTest output in ghc-head - - - - - b194182a by Hécate Moonlight at 2021-02-24T11:12:36+01:00 Merge pull request haskell/haddock#1351 from hsyl20/wip/hsyl20/fix-head Fix haddockHypsrcTest output in ghc-head - - - - - 3ce8b375 by Shayne Fletcher at 2021-03-06T09:55:03-05:00 Add ITproj to parser - - - - - d2abf762 by Ben Gamari at 2021-03-06T19:26:49-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a0f6047d by Andrew Martin at 2021-03-07T11:25:23-05:00 Update for boxed rep - - - - - 6f63c99e by Ben Gamari at 2021-03-10T13:20:21-05:00 Update for "FastString: Use FastMutInt instead of IORef Int" - - - - - e13f01df by Luke Lau at 2021-03-10T15:38:40-05:00 Implement template-haskell's putDoc This catches up to GHC using the new extractTHDocs function, which returns documentation added via the putDoc function (provided it was compiled with Opt_Haddock). Since it's already a map from names -> docs, there's no need to do traversal etc. It also matches the change from the argument map being made an IntMap rather than a Map Int - - - - - 89263d94 by Alan Zimmerman at 2021-03-15T17:15:26+00:00 Match changes in GHC AST for in-tree API Annotations As landed via https://gitlab.haskell.org/ghc/ghc/-/merge_requests/2418 - - - - - 28db1934 by Alan Zimmerman at 2021-03-15T20:40:09+00:00 Change some type family test results. It is not clear to me whether the original was printing incorrectly (since we did not have the TopLevel flag before now), or if this behaviour is expected. For the time being I am assuming the former. - - - - - 7c11c989 by Sylvain Henry at 2021-03-22T10:05:19+01:00 Fix after NameCache changes - - - - - addbde15 by Sylvain Henry at 2021-03-22T10:05:19+01:00 NameCache doesn't store a UniqSupply anymore - - - - - 15ec6cec by Ben Gamari at 2021-03-22T17:53:44-04:00 Bump GHC version to 9.2 - - - - - dbd6aa63 by Hécate Moonlight at 2021-03-24T14:28:36+01:00 Merge pull request haskell/haddock#1365 from hsyl20/wip/hsyl20/iface1 NameCache refactoring - - - - - 2d32da7e by Oleg Grenrus at 2021-03-27T01:12:00+02:00 Specialization of Data.List - - - - - 32b84fa6 by Fendor at 2021-03-27T10:50:17+01:00 Add UnitId to Target record This way we always know to which home-unit a given target belongs to. So far, there only exists a single home-unit at a time, but it enables having multiple home-units at the same time. - - - - - 54bf9f0e by Hécate Moonlight at 2021-03-28T14:08:35+02:00 Merge pull request haskell/haddock#1368 from fendor/target-unit-id-revert Add UnitId to Target record - - - - - 7dea168a by Alan Zimmerman at 2021-03-29T08:45:52+01:00 EPA : Rename ApiAnn to EpAnn - - - - - 72967f65 by Alfredo Di Napoli at 2021-03-29T09:47:01+02:00 pprError changed name in GHC - - - - - 4bc61035 by Alan Zimmerman at 2021-03-29T16:16:27-04:00 EPA : Rename ApiAnn to EpAnn - - - - - 108d031d by Ben Gamari at 2021-03-29T18:49:36-04:00 Merge commit '36418c4f70d7d2b179a77925b3ad5caedb08c9b5' into HEAD - - - - - 1444f700 by Ben Gamari at 2021-03-31T09:18:39-04:00 Merge pull request haskell/haddock#1370 from adinapoli/wip/adinapoli-diag-reason-severity Rename pprError to mkParserErr - - - - - d3087b79 by Ben Gamari at 2021-03-31T11:34:17-04:00 Merge commit 'd8d8024ad6796549a8d3b5512dabf3288d14e30f' into ghc-head - - - - - 170b79e9 by Ben Gamari at 2021-03-31T12:24:56-04:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - db0d6bae by Ben Gamari at 2021-04-10T09:34:35-04:00 Bump GHC version to 9.3 - - - - - a9f2c421 by Alan Zimmerman at 2021-04-19T18:26:46-04:00 Update for EPA changes in GHC (cherry picked from commit cafb48118f7c111020663776845897e225607b41) - - - - - 1ee4b7c7 by Sylvain Henry at 2021-05-11T10:00:06+02:00 Removal of HsVersions.h (#1388) * Update for EPA changes in GHC * Account for HsVersions.h removal Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - 79e819e9 by Hécate Moonlight at 2021-05-11T10:14:47+02:00 Revert "Removal of HsVersions.h (#1388)" This reverts commit 72118896464f94d81f10c52f5d9261efcacc57a6. - - - - - 3dbd3f8b by Alan Zimmerman at 2021-05-11T10:15:17+02:00 Update for EPA changes in GHC - - - - - 2ce80c17 by Sylvain Henry at 2021-05-11T10:15:19+02:00 Account for HsVersions.h removal - - - - - 00e4c918 by Christiaan Baaij at 2021-05-13T08:21:56+02:00 Add Haddock support for the OPAQUE pragma (#1380) - - - - - 8f9049b2 by Hécate Moonlight at 2021-05-13T08:40:22+02:00 fixup! Use GHC 9.2 in CI runner - - - - - 27ddec38 by Alan Zimmerman at 2021-05-13T22:51:20+01:00 EPA: match changes from GHC T19834 - - - - - f8a1d714 by Felix Yan at 2021-05-14T17:10:04+02:00 Allow hspec 2.8 (#1389) All tests are passing. - - - - - df44453b by Divam Narula at 2021-05-20T15:42:42+02:00 Update ref, the variables got renamed. (#1391) This is due to ghc/ghc!5555 which caused a change in ModDetails in case of NoBackend. Now the initModDetails is used to recreate the ModDetails from interface and in-memory ModDetails is not used. - - - - - e46bfc87 by Alan Zimmerman at 2021-05-20T19:05:09+01:00 Remove Maybe from HsQualTy Match changes in GHC for haskell/haddock#19845 - - - - - 79bd7b62 by Shayne Fletcher at 2021-05-22T08:20:39+10:00 FieldOcc: rename extFieldOcc to foExt - - - - - 6ed68c74 by Ben Gamari at 2021-05-21T22:29:30-04:00 Merge commit '3b6a8774bdb543dad59b2618458b07feab8a55e9' into ghc-head - - - - - f9a02d34 by Alfredo Di Napoli at 2021-05-24T13:53:00+02:00 New Parser diagnostics interface - - - - - 392807d0 by Ben Gamari at 2021-05-24T09:57:40-04:00 Merge pull request haskell/haddock#1394 from adinapoli/wip/adinapoli-align-ps-messages Align Haddock to use the new Parser diagnostics interface - - - - - 33023cd8 by Ben Gamari at 2021-05-24T11:19:16-04:00 Revert "Add Haddock support for the OPAQUE pragma (#1380)" This reverts commit a1337c599ef7720b0482a25c55f11794112496dc. The GHC patch associated with this change is not yet ready to be merged. - - - - - 8c005af7 by Ben Simms at 2021-05-28T07:56:20+02:00 CI configuration for ghc-head (#1395) - - - - - 1e947612 by Hécate Moonlight at 2021-05-28T12:27:35+02:00 Use GHC 9.2 in CI runner (#1378) - - - - - e6fa10ab by CGenie at 2021-05-31T09:02:13+02:00 Add page about common errors (#1396) * Update index.rst Common errors page * Create common-errors.rst * Update common-errors.rst * Use GHC 9.2 in CI runner (#1378) * [haddock-api] remove .hspec-failures Co-authored-by: Hécate Moonlight <Kleidukos at users.noreply.github.com> - - - - - abc72a8d by Sylvain Henry at 2021-06-01T10:02:06+02:00 Adapt Haddock to Logger and Parser changes (#1399) - - - - - 91373656 by Zubin Duggal at 2021-06-01T20:45:10+02:00 Update haddockHypSrc tests since we now compute slighly more type info (#1397) - - - - - ed712822 by Marcin Szamotulski at 2021-06-02T08:54:33+02:00 Added myself to contributors - - - - - 49fdbcb7 by Marcin Szamotulski at 2021-06-02T08:57:24+02:00 Document multi component support - - - - - 9ddc8d7d by Hécate Moonlight at 2021-06-02T09:35:55+02:00 Merge pull request haskell/haddock#1379 from coot/coot/document-multi-component-support Document multi component support - - - - - 585b5c5e by Ben Simms at 2021-06-02T19:46:54+02:00 Update CONTRIBUTING.md (#1402) - - - - - 1df4a605 by Ben Simms at 2021-06-02T19:47:14+02:00 Update CONTRIBUTING.md (#1403) - - - - - 58ea43d2 by sheaf at 2021-06-02T22:09:06+02:00 Update Haddock Bug873 to account for renaming - - - - - c5d0ab23 by Vladislav Zavialov at 2021-06-10T13:35:42+03:00 HsToken in FunTy, RecConGADT - - - - - 1ae2f40c by Hécate Moonlight at 2021-06-11T11:19:09+02:00 Update the CI badges - - - - - 6fdc4de2 by Sylvain Henry at 2021-06-28T19:21:17+02:00 Fix mkParserOpts (#1411) - - - - - 18201670 by Alfredo Di Napoli at 2021-07-05T07:55:12+02:00 Rename getErrorMessages Lexer import This commit renames the Lexer import in `Hyperlinker.Parser` from `getErrorMessages` to `getPsErrorMessages` to eliminate the ambiguity with the `getErrorMessages` function defined in `GHC.Types.Error`. - - - - - 23173ca3 by Ben Gamari at 2021-07-07T11:31:44-04:00 Merge pull request haskell/haddock#1413 from adinapoli/wip/adinapoli-issue-19920 Rename getErrorMessages Lexer import - - - - - b3dc4ed8 by Alan Zimmerman at 2021-07-28T22:30:59+01:00 EPA: match changes from GHC T19834 (cherry picked from commit 2fec1b44e0ee7e263286709aa528b4ecb99ac6c2) - - - - - 5f177278 by Ben Gamari at 2021-08-06T01:17:37-04:00 Merge commit '2a966c8ca37' into HEAD - - - - - cdd81d08 by Marcin Szamotulski at 2021-08-08T17:19:06+02:00 coot/multiple packages (ghc-9.2) (#1418) - - - - - be0d71f1 by Marcin Szamotulski at 2021-08-16T08:46:03+02:00 coot/multiple package (ghc-head) (#1419) * FromJSON class Aeson style FromJSON class with Parsec based json parser. * doc-index.json file for multiple packages When creating haddock summary page for multiple packages render doc-index.json file using contents of all found 'doc-index.json' files. * Render doc-index.json When rendering html, render doc-index.json file independently of maybe_index_url option. doc-index.json file is useful now even if maybe_index_url is not `Nothing`. * base url option New `Flag_BaseURL` which configures from where static files are loaded (--base-url). If given and not equal "." static files are not coppied, as this indicates that they are not read from the the directory where we'd copy them. The default value is ".". - - - - - 3b09dbdf by Hécate Moonlight at 2021-10-07T23:26:03+02:00 Update GHC 9.2 to latest pre-release in CI - - - - - 7ac55417 by Zubin Duggal at 2021-10-11T12:10:19+02:00 Enable Haddock tests in GHC windows CI (#1428) * testsuite: strip windows line endings for haddock * hyperlinker: Work around double escaping (#19236) * deterministic SCC - - - - - 1cb81f25 by Andrew Lelechenko at 2021-10-12T15:23:19+02:00 haddock-library does not depend on bytestring or transformers (#1426) - - - - - a890b9aa by sheaf at 2021-10-15T22:19:42+02:00 update haddockHypsrcTest for GHC MR !6705 (#1430) - - - - - 42a55c6c by Sylvain Henry at 2021-10-15T22:20:10+02:00 Fix after PkgQual refactoring (#1429) - - - - - 91659238 by Alan Zimmerman at 2021-10-28T18:57:10+01:00 Update for changes in GHC for branch wip/az/no-srcspan-anno-instances - - - - - acf23e60 by Vladislav Zavialov at 2021-11-05T02:09:47+03:00 Do not use forall as an identifier See GHC ticket haskell/haddock#20609 - - - - - c565db0e by Krzysztof Gogolewski at 2021-11-27T02:42:35+01:00 Update after NoExtCon -> DataConCantHappen rename - - - - - b5f55590 by Artem Pelenitsyn at 2021-11-27T11:14:17+01:00 fix CI for 9.2 (#1436) - - - - - 25cd621e by Matthew Pickering at 2021-12-02T11:46:54+00:00 Update html-test for Data.List revert - - - - - 1d5ff85f by malteneuss at 2021-12-15T07:56:55+01:00 Add hint about inline link issue (#1444) - - - - - 791fde81 by Sylvain Henry at 2021-12-16T09:29:51+01:00 Bump ghc-head (#1445) * Update after NoExtCon -> DataConCantHappen rename * Update html-test for Data.List revert * Fix for new Plugins datatype Co-authored-by: Krzysztof Gogolewski <krzysztof.gogolewski at tweag.io> Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 44236317 by Sylvain Henry at 2021-12-17T09:39:00+01:00 Fix for new Plugins datatype - - - - - 80ada0fa by Hécate Moonlight at 2021-12-17T17:28:48+01:00 Remove ghc-head workflow (#1446) Contributions of GHC glue code are now done on the GHC gitlab, not in the GitHub repo anymore. - - - - - 49e171cd by Matthew Pickering at 2021-12-28T09:47:09+00:00 Remove use of ExtendedModSummary - - - - - 0e91b5ea by askeblad at 2022-01-04T09:18:35+01:00 update URLs - - - - - 9f13c212 by Hécate Moonlight at 2022-02-25T10:19:46+01:00 Fix solver for GHC 9.2 - - - - - 386751a1 by Meng Weng Wong at 2022-02-25T19:19:11+01:00 IDoc link has bitrotted; replaced with web.archive.org cache. (#1454) - - - - - d877cbe6 by Hécate Moonlight at 2022-02-25T19:21:58+01:00 Fix haddock user guide (#1456) - - - - - cc47f036 by Andrew Lelechenko at 2022-03-04T17:29:36+01:00 Allow text-2.0 in haddock-library (#1459) - - - - - 7b3685a3 by malteneuss at 2022-03-07T19:27:24+01:00 Add multi-line style hint to style section (#1460) - - - - - c51088b8 by John Ericson at 2022-03-11T16:46:26+01:00 Fix CollectPass instance to match TTG refactor Companion to GHC !7614 (https://gitlab.haskell.org/ghc/ghc/-/merge_requests/7614) - - - - - b882195b by Vladislav Zavialov at 2022-03-14T20:32:30+01:00 Link to (~) - - - - - 877349b8 by Christiaan Baaij at 2022-03-16T09:20:43+01:00 Add Haddock support for the OPAQUE pragma - - - - - 0ea22721 by askeblad at 2022-03-16T09:44:27+01:00 typos (#1464) - - - - - a6d13da1 by Matthew Pickering at 2022-03-22T13:41:17+00:00 Minimum changes needed for compilation with hi-haddock With hi-haddock, of course there is a much large refactoring of haddock which could be achieved but that is left for a future patch which can implemented at any time independently of GHC. - - - - - e7ac9129 by Matthew Pickering at 2022-03-22T21:17:50+00:00 Update test output - - - - - 6d916214 by Matthew Pickering at 2022-03-24T15:06:26+00:00 Merge branch 'wip/opaque_pragma' into 'ghc-head' Add Haddock support for the OPAQUE pragma See merge request ghc/haddock!2 - - - - - 42208183 by Steve Hart at 2022-03-25T20:43:50+01:00 Fix CI (#1467) * CI: Reinstall GHC with docs CI tests were failing because the GHC preinstalled to the CI environment does not include documentation, which is required for running the Haddock tests. This commit causes the CI workflow to reinstall GHC with docs so that tests can succeed. - - - - - 9676fd79 by Steve Hart at 2022-03-25T21:33:34+01:00 Make links in Synopsis functional again (#1458) Commit e41c1cbe9f0476997eac7b4a3f17cbc6b2262faf added a call to e.preventDefault() when handling click events that reach a toggle element. This prevents the browser from following hyperlinks within the Synopsis section when they are clicked by a user. This commit restores functioning hyperlinks within the Synopsis section by removing the call to e.preventDefault(), as it does not appear to be necessary, and removing it increases the flexibility of the details-helper code. - - - - - d1edd637 by sheaf at 2022-04-01T12:02:02+02:00 Keep track of promotion ticks in HsOpTy Keeping track of promotion ticks in HsOpTy allows us to properly pretty-print promoted constructors such as lists. - - - - - 9dcb2dfc by Jakob Brünker at 2022-04-01T15:46:22+00:00 Add support for \cases See merge request ghc/ghc!7873 - - - - - b0412ee5 by askeblad at 2022-04-06T17:47:57+02:00 spelling errors (#1471) - - - - - 6b18829b by Vladislav Zavialov at 2022-04-06T18:53:58+02:00 Rename [] to List - - - - - 2d046691 by Vladislav Zavialov at 2022-04-07T20:25:54+03:00 HsToken ConDeclGADT con_dcolon - - - - - 90b43da4 by Steve Hart at 2022-04-12T13:29:46+02:00 Parse Markdown links at beginning of line within a paragraph (#1470) * Catch Markdown links at beginning of line within paragraph Per Issue haskell/haddock#774, Markdown links were being parsed as ordinary text when they occurred at the beginning of a line other than the first line of the paragraph. This occurred because the parser was not interpreting a left square bracket as a special character that could delimit special markup. A space character was considered a special character, so, if a space occurred at the beginning of the new line, then the parser would interpret the space by itself and then continue parsing, thereby catching the Markdown link. '\n' was not treated as a special character, so the parser did not catch a Markdown link that may have followed. Note that this will allow for Markdown links that are not surrounded by spaces. For example, the following text includes a Markdown link that will be parsed: Hello, world[label](url) This is consistent with how the parser handles other types of markup. * Remove obsolete documentation hint Commit 6b9aeafddf20efc65d3725c16e3fc43a20aac343 should eliminate the need for the workaround suggested in the documentation. - - - - - 5b08312d by Hécate Moonlight at 2022-04-12T13:36:38+02:00 Force ghc-9.2 in the cabal.project - - - - - 0d0ea349 by dependabot[bot] at 2022-04-12T13:57:41+02:00 Bump path-parse from 1.0.5 to 1.0.7 in /haddock-api/resources/html (#1469) Bumps [path-parse](https://github.com/jbgutierrez/path-parse) from 1.0.5 to 1.0.7. - [Release notes](https://github.com/jbgutierrez/path-parse/releases) - [Commits](https://github.com/jbgutierrez/path-parse/commits/v1.0.7) --- updated-dependencies: - dependency-name: path-parse dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 2b9fc65e by dependabot[bot] at 2022-04-12T13:57:54+02:00 Bump copy-props from 2.0.4 to 2.0.5 in /haddock-api/resources/html (#1468) Bumps [copy-props](https://github.com/gulpjs/copy-props) from 2.0.4 to 2.0.5. - [Release notes](https://github.com/gulpjs/copy-props/releases) - [Changelog](https://github.com/gulpjs/copy-props/blob/master/CHANGELOG.md) - [Commits](https://github.com/gulpjs/copy-props/compare/2.0.4...2.0.5) --- updated-dependencies: - dependency-name: copy-props dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - ea98a6fb by Ben Gamari at 2022-04-23T22:54:37-04:00 Update for GHC 9.4 - - - - - 9e11864f by Ben Gamari at 2022-04-25T16:24:31-04:00 Merge remote-tracking branch 'upstream/ghc-9.2' into ghc-head - - - - - f83cc506 by Ben Gamari at 2022-04-25T17:00:25-04:00 Bump ghc version to 9.5 - - - - - e01c2e7d by Ben Gamari at 2022-04-28T16:19:04-04:00 Revert "Bump ghc-head (#1445)" This reverts commit b29a78ef6926101338f62e84f456dac8659dc9d2. This should not have been merged. - - - - - a2b5ee8c by Ben Gamari at 2022-04-28T16:19:24-04:00 Merge commit '2627a86c' into ghc-head - - - - - 0c6fe4f9 by Ben Gamari at 2022-04-29T10:05:54-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-9.4 - - - - - b6e5cb0a by Ben Gamari at 2022-04-29T11:46:06-04:00 Revert "HsToken ConDeclGADT con_dcolon" This reverts commit 24208496649a02d5f87373052c430ea4a97842c5. - - - - - 15a62888 by Ben Gamari at 2022-04-29T15:12:55-04:00 Bump base upper bound - - - - - 165b9031 by Ben Gamari at 2022-04-29T23:58:38-04:00 Update test output - - - - - e0c3e5da by Phil de Joux at 2022-05-02T14:46:38+02:00 Add hlint action .hlint.yaml with ignores & CPP. (#1475) - - - - - ead1158d by Raphael Das Gupta at 2022-05-02T14:46:48+02:00 fix grammar in docs: "can the" → "can be" (#1477) - - - - - cff97944 by Ben Gamari at 2022-05-02T18:38:56-04:00 Allow base-4.17 - - - - - e4ecb201 by Phil de Joux at 2022-05-03T13:14:55+02:00 Remove unused imports that GHC warned about. (#1480) - - - - - 222890b1 by Phil de Joux at 2022-05-03T13:15:46+02:00 Follow hlint suggestion to remove redundant bang. (#1479) - - - - - 058b671f by Phil de Joux at 2022-05-03T13:34:04+02:00 Follow hlint, remove language pragmas in libs. (#1478) - - - - - 0a645049 by Ben Simms at 2022-05-03T14:19:24+02:00 Keep track of ordered list indexes and render them (#1407) * Keep track of ordered list indexes and render them * Rename some identifiers to clarify - - - - - f0433304 by Norman Ramsey at 2022-05-04T15:13:34-04:00 update for changes in GHC API - - - - - 3740cf71 by Emily Martins at 2022-05-06T18:23:48+02:00 Add link to the readthedocs in cabal description to show on hackage. (cherry picked from commit 52e2d40d47295c02d3181aac0c53028e730f1e3b) - - - - - 5d754f1e by Hécate Moonlight at 2022-05-06T18:44:57+02:00 remove Bug873 - - - - - 968fc267 by Hécate Moonlight at 2022-05-06T18:48:28+02:00 Ignore "Use second" HLint suggestion. It increases laziness. - - - - - 02d14e97 by Jade Lovelace at 2022-05-07T17:42:08+02:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` - - - - - b22b87ed by Artem Pelenitsyn at 2022-05-08T16:19:47+02:00 fix parsing trailing quotes in backticked identifiers (#1408) (#1483) - - - - - 80ae107b by Alex Biehl at 2022-05-08T16:37:16+02:00 Fix "Defined by not used" error (cherry picked from commit 6e02a620a26c3a44f98675dd1b93b08070c36c0a) - - - - - 4c838e84 by Hécate Moonlight at 2022-05-08T16:37:16+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - fc9827b4 by Hécate Moonlight at 2022-05-08T16:40:40+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - b153b555 by Xia Li-yao at 2022-05-20T17:52:42+02:00 Hide synopsis from search when hidden (#1486) Fix haskell/haddock#1451 - - - - - f3e38b85 by Marcin Szamotulski at 2022-05-21T23:32:31+02:00 Allow to hide interfaces when rendering multiple components (#1487) This is useful when one wishes to `--gen-contents` when rendering multiple components, but one does not want to render all modules. This is in particular useful when adding base package. - - - - - f942863b by Marcin Szamotulski at 2022-05-24T08:29:59+02:00 Check if doc-index.json exists before reading it (#1488) - - - - - 31e92982 by Marcin Szamotulski at 2022-05-25T16:22:13+02:00 Version bump 2.26.1 (#1489) * Version bump 2.26.1 We extended format accepted by `--read-interface` option, which requires updating the minor version. * Update documentation of --read-interface option - - - - - 7cc873e0 by sheaf at 2022-05-25T16:42:31+02:00 Updated HaddockHypsrcTest output for record update changes (MR !7981) - - - - - cd196942 by Marcin Szamotulski at 2022-05-25T20:28:47+02:00 Use visibility to decide which interfaces are included in quickjump (#1490) This is also consistent with how html index is build. See haskell/cabal#7669 for rationale behind this decision. - - - - - 00c713c5 by Hécate Moonlight at 2022-05-26T17:09:15+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 2f3039f1 by Hécate Moonlight at 2022-05-26T17:10:59+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 63a5650c by romes at 2022-05-31T12:43:22+01:00 TTG: Match new GHC AST - - - - - dd7d1617 by romes at 2022-06-02T16:11:00+01:00 Update for IE changes in !8228 - - - - - c23aaab7 by cydparser at 2022-06-06T08:48:14+02:00 Fix and improve CI (#1495) * Pin GHC version before creating the freeze file * Use newest action versions * Improve caching * Avoid unnecessarily reinstalling GHC * Use GHC 9.2.2 for CI Co-authored-by: Cyd Wise <cwise at tripshot.com> - - - - - c156fa77 by Hécate Moonlight at 2022-06-06T11:59:35+02:00 Add Mergify configuration (#1496) - - - - - 2dba4188 by Hécate Moonlight at 2022-06-06T16:12:50+02:00 Bump haddock's version in cabal file to 2.26.1 (#1497) - - - - - d7d4b8b9 by Marcin Szamotulski at 2022-06-07T06:09:40+00:00 Render module tree per package in the content page (#1492) * Render module tree per package in the content page When rendering content page for multiple packages it is useful to split the module tree per package. Package names in this patch are inferred from haddock's interface file names. * Write PackageInfo into interface file To keep interface file format backward compatible, instead of using `Binary` instance for `InterfaceFile` we introduce functions to serialise and deserialise, which depends on the interface file version. - - - - - 77765665 by Mike Pilgrem at 2022-06-12T21:57:19+01:00 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - b0e079b0 by mergify[bot] at 2022-06-13T11:49:37+00:00 Merge pull request haskell/haddock#1108 from mpilgrem/fix783 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - 6c0292b1 by Hécate Moonlight at 2022-06-21T17:21:08+02:00 Update the contribution guide - - - - - e413b9fa by dependabot[bot] at 2022-06-21T23:38:19+02:00 Bump shell-quote from 1.6.1 to 1.7.3 in /haddock-api/resources/html (#1500) Bumps [shell-quote](https://github.com/substack/node-shell-quote) from 1.6.1 to 1.7.3. - [Release notes](https://github.com/substack/node-shell-quote/releases) - [Changelog](https://github.com/substack/node-shell-quote/blob/master/CHANGELOG.md) - [Commits](https://github.com/substack/node-shell-quote/compare/1.6.1...1.7.3) --- updated-dependencies: - dependency-name: shell-quote dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 29d0ef70 by romes at 2022-07-06T11:29:39+02:00 TTG: AST Updates for !8308 - - - - - 1bae7c87 by Alan Zimmerman at 2022-07-06T22:50:43+01:00 Match GHC changes for T21805 This brings in a newtype for FieldLabelString - - - - - 6fe8b988 by Phil de Joux at 2022-07-16T20:54:26+00:00 Bump hlint version to 3.4.1, the version with counts. (#1503) Redo the counts available with the --default option. - - - - - 48fb43af by Phil de Joux at 2022-07-19T09:32:55+02:00 Follow hlint suggestion: unused LANGUAGE pragma. (#1504) * Follow hlint suggestion: unused LANGUAGE pragma. * Ignore within modules to pass linting and pass tests. - - - - - c1cf1fa7 by Phil de Joux at 2022-07-24T13:45:59+02:00 Follow hlint suggestion: redundant $. (#1505) * Follow hlint suggestion: redundant $. * Remove $ and surplus blank lines in Operators. - - - - - 74777eb2 by Jade Lovelace at 2022-07-29T11:02:41+01:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` (cherry picked from commit ab53ccf089ea703b767581ac14be0f6c78a7678a) - - - - - faa4cfcf by Hécate Moonlight at 2022-07-29T20:31:20+02:00 Merge pull request haskell/haddock#1516 from duog/9-4-backport-fix-hyperlinks Backport 9-4: Fix hyperlinks to external items and modules (#1482) - - - - - 5d2450f3 by Ben Gamari at 2022-08-05T17:41:15-04:00 Merge remote-tracking branch 'origin/ghc-9.4' - - - - - 63954f73 by Ben Gamari at 2022-08-05T19:08:36-04:00 Clean up build and testsuite for GHC 9.4 - - - - - d4568cb8 by Hécate Moonlight at 2022-08-05T19:10:49-04:00 Bump the versions - - - - - 505583a4 by Ben Gamari at 2022-08-06T13:58:27-04:00 Merge pull request haskell/haddock#1518 from bgamari/wip/ghc-9.4-merge Merge GHC 9.4 into `main` - - - - - 5706f6a4 by Ben Gamari at 2022-08-06T22:57:21-04:00 html-test: Testsuite changes for GHC 9.4.1 - - - - - 5f2a45a2 by Ben Gamari at 2022-08-15T14:33:05-04:00 doc: Fix a few minor ReST issues Sphinx was complaining about too-short title underlines. - - - - - 220e6410 by Ben Gamari at 2022-08-15T14:41:24-04:00 Merge branch 'main' into ghc-head - - - - - fbeb1b02 by Ben Gamari at 2022-08-15T14:45:16-04:00 Updates for GHC 9.5 - - - - - eee562eb by Vladislav Zavialov at 2022-08-15T14:46:13-04:00 HsToken ConDeclGADT con_dcolon - - - - - c5f073db by Ben Gamari at 2022-08-15T16:55:35-04:00 Updates for GHC 9.5 - - - - - 3f7ab242 by Vladislav Zavialov at 2022-08-15T16:55:35-04:00 HsToken ConDeclGADT con_dcolon - - - - - a18e473d by Ben Gamari at 2022-08-16T08:35:19-04:00 Merge branch 'wip/ghc-head-bump' into ghc-head - - - - - af0ff3a4 by M Farkas-Dyck at 2022-09-15T21:16:05+00:00 Disuse `mapLoc`. - - - - - a748fc38 by Matthew Farkas-Dyck at 2022-09-17T10:44:18+00:00 Scrub partiality about `NewOrData`. - - - - - 2758fb6c by John Ericson at 2022-09-18T03:27:37+02:00 Test output changed because of change to `base` Spooky, but I guess that is intended? - - - - - a7eec128 by Torsten Schmits at 2022-09-21T11:06:55+02:00 update tests for the move of tuples to GHC.Tuple.Prim - - - - - 461e7b9d by Ross Paterson at 2022-09-24T22:01:25+00:00 match implementation of GHC proposal haskell/haddock#106 (Define Kinds Without Promotion) - - - - - f7fd77ef by sheaf at 2022-10-17T14:53:01+02:00 Update Haddock for GHC MR !8563 (configuration of diagnostics) - - - - - 3d3e85ab by Vladislav Zavialov at 2022-10-22T23:04:06+03:00 Class layout info - - - - - cbde4cb0 by Simon Peyton Jones at 2022-10-25T23:19:18+01:00 Adapt to Constraint-vs-Type See haskell/haddock#21623 and !8750 - - - - - 7108ba96 by Tom Smeding at 2022-11-01T22:33:23+01:00 Remove outdated footnote about module re-exports The footnote is invalid with GHC 9.2.4 (and possibly earlier): the described behaviour in the main text works fine. - - - - - 206c6bc7 by Hécate Moonlight at 2022-11-01T23:00:46+01:00 Merge pull request haskell/haddock#1534 from tomsmeding/patch-1 - - - - - a57b4c4b by Andrew Lelechenko at 2022-11-21T00:39:52+00:00 Support mtl-2.3 - - - - - e9d62453 by Simon Peyton Jones at 2022-11-25T13:49:12+01:00 Track small API change in TyCon.hs - - - - - eb1c73f7 by Ben Gamari at 2022-12-07T08:46:21-05:00 Update for GhC 9.6 - - - - - 063268dd by Ben Gamari at 2022-12-07T11:26:32-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 4ca722fe by Ben Gamari at 2022-12-08T14:43:26-05:00 Bump bounds to accomodate base-4.18 - - - - - 340b7511 by Vladislav Zavialov at 2022-12-10T12:31:28+00:00 HsToken in HsAppKindTy - - - - - 946226ec by Ben Gamari at 2022-12-13T20:12:56-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - fd8faa66 by Ben Gamari at 2022-12-22T13:44:28-05:00 Bump GHC version to 9.7 - - - - - 2958aa9c by Ben Gamari at 2022-12-22T14:49:16-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 9e0fefd8 by Andrei Borzenkov at 2023-01-30T14:02:04+04:00 Rename () to Unit, Rename (,,...,,) to Tuple<n> - - - - - eb3968b5 by Ben Gamari at 2023-03-10T02:32:43-05:00 Bump versions for ghc-9.6 release - - - - - 4aeead36 by Adam Gundry at 2023-03-23T13:53:47+01:00 Adapt to warning categories changes - - - - - 642d8d60 by sheaf at 2023-03-29T13:35:56+02:00 Adapt to record field refactor This commit adapts to the changes in GHC MR !8686, which overhauls the treatment of record fields in the renamer, adding separate record field namespaces and entirely removing the GreName datatype. - - - - - ac8d4333 by doyougnu at 2023-03-29T11:11:44-04:00 Update UniqMap API - - - - - 7866fc86 by Ben Orchard at 2023-04-20T11:29:33+02:00 update classify with new tokens - - - - - ffcdd683 by Finley McIlwaine at 2023-04-24T09:36:18-06:00 Remove index-state - - - - - 05b70982 by Finley McIlwaine at 2023-04-26T08:16:31-06:00 `renameInterface` space leak fixes - Change logic for accumulation of names for which link warnings will be generated - Change handling of `--ignore-link-symbol` to allow qualified and unqualified names. Added to CHANGES.md - Some formatting changes and comments here and there - - - - - e5697d7c by Finley McIlwaine at 2023-04-27T18:46:36-06:00 Messy things - ghc-debug dependency and instrumentation - cabal.project custom with-compiler - hie.yaml files - traces and such - - - - - 0b8ef80b by Finley McIlwaine at 2023-05-02T18:08:52-06:00 Stop retaining GRE closures GRE closures should never be necessary to Haddock, so we never want to keep them on the heap. Despite that, they are retained by a lot of the data structures that Haddock makes use of. - Attempt to fix that situation by adding strictness to various computations and pruning the `ifaceInstances` field of `Interface` to a much thinner data type. - Removes the `ifaceFamInstances` field, as it was never used. - Move some of the attach instances types (e.g. `SimpleType`) to the types module - - - - - 8bda991b by Finley McIlwaine at 2023-05-08T16:07:51-06:00 Memory usage fixes - Refactor `ifaceDeclMap` to drastically reduce memory footprint. We no longer store all declarations associated with a given name, since we only cared to determine if the only declaration associated with a name was a value declaration. Change the `DeclMap` type to better reflect this. - Drop pre-renaming export items after the renaming step. Since the Hoogle backend used the pre-renamed export items, this isn't trivial. We now generate Hoogle output for exported declarations during the renaming step (if Hoogle output /should/ be generated), and store that with the renamed export item. - Slightly refactor Hoogle backend to handle the above change and allow for early generation of Hoogle output. - Remove the `ifaceRnDocMap` and `ifaceRnArgMap` fields of the `Interface` type, as they were never used. - Remove some unnecessary strictness - Remove a lot of dead code from `Syb` module - - - - - 1611ac0c by Finley McIlwaine at 2023-05-09T11:51:57-06:00 Unify ErrMsgM and IfM - Delete ErrMsgM, stop accumulating warnings in a writer - Make IfM a state monad, print warnings directly to stdout, move IfM type into types module - Drop ErrMsg = String synonym - Unset IORefs from plugin after they are read, preventing unnecessary retention of interfaces - - - - - 42d696ab by Finley McIlwaine at 2023-05-11T15:52:07-06:00 Thunk leak fixes The strictness introduced in this commit was motivated by observing thunk leaks in the eventlog2html output. - Refactor attach instances list comprehension to avoid large intermediate thunks - Refactor some HTML backend list comprehensions to avoid large intermediate thunks - Avoid thunks accumulating in documentation types or documentation parser - A lot of orphan NFData instances to allow us to force documentation values - - - - - 68561cf6 by Finley McIlwaine at 2023-05-11T17:02:10-06:00 Remove GHC debug dep - - - - - 10519e3d by Finley McIlwaine at 2023-05-15T12:40:48-06:00 Force HIE file path Removes a potential retainer of `ModSummary`s - - - - - 1e4a6ec6 by Finley McIlwaine at 2023-05-15T14:20:34-06:00 Re-add index-state, with-compiler, delete hie.yamls - - - - - a2363fe9 by Hécate Moonlight at 2023-05-15T22:45:16+02:00 Merge pull request haskell/haddock#1594 from FinleyMcIlwaine/finley/ghc-9.6-mem-fixes Reduce memory usage - - - - - e8a78383 by Finley McIlwaine at 2023-05-17T12:19:16-06:00 Merge branch ghc-9.6 into ghc-head - - - - - 22e25581 by Finley McIlwaine at 2023-05-17T12:20:23-06:00 Merge branch 'ghc-head' of gitlab.haskell.org:ghc/haddock into ghc-head - - - - - 41bbf0df by Bartłomiej Cieślar at 2023-05-24T08:57:58+02:00 changes to the WarningTxt cases Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - c686ba9b by Hécate Moonlight at 2023-06-01T14:03:02-06:00 Port the remains of Hi-Haddock - - - - - 9d8a85fd by Hécate Moonlight at 2023-06-01T14:03:06-06:00 Stdout for tests - - - - - 36331d07 by Finley McIlwaine at 2023-06-01T14:06:02-06:00 Formatting, organize imports - - - - - a06059b1 by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix empty context confusion in Convert module - - - - - 379346ae by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix associated type families in Hoogle output - - - - - fc6ea7ed by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix test refs Accept several changes in Hoogle tests: Pretty printing logic no longer prints the `(Proxy (Proxy (...))` chain in Bug806 with parentheses. Since this test was only meant to verify that line breaks do not occur, accept the change. `tyThingToLHsDecl` is called for class and data declarations, which ends up "synifying" the type variables and giving unlifted types kind signatures. As a result, type variables of kind `Type -> Type` are now printed with kind signatures in Hoogle output. This could be changed by manually drop kind signatures from class variables in the Hoogle backend if the behavior is deemed unacceptable. Sometimes subordinate declarations are exported separate from their parent declarations (e.g. record selectors). In this case, a type signature is cobbled together for the export item in `extractDecl`. Since this type signature is very manually constructed, it may lack kind signatures of decls constructed from `tyThingToLHsDecl`. An example of this is the `type-sigs` Hoogle test. Change `*` to `Type` in Hoogle test refs. I don't think this will break Hoogle behavior, since it appears to not consider type signatures in search. I have not fully verified this. - - - - - e14b7e58 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix LaTeX backend test refs Changes to GHC pretty printing code have resulted in some differences to Haddock's LaTeX output. - Type variables are printed explicitly quantified in the LinearTypes test - Wildcard types in type family equations are now printed numbered, e.g. _1 _2, in the TypeFamilies3 test - Combined signatures in DefaultSignatures test are now documented as separate signatures - - - - - 41b5b296 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and test source updates - Stop using kind `*` in html test sources - Add TypeOperators where necessary to avoid warnings and future errors - Rename some test modules to match their module names - - - - - c640e2a2 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix missing deprecation warnings on record fields `lookupOccEnv` was used to resolve `OccNames` with warnings attached, but it doesn't look in the record field namespace. Thus, if a record field had a warning attached, it would not resolve and the warning map would not include it. This commit fixes by using `lookupOccEnv_WithFields` instead. - - - - - fad0c462 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and some comments - - - - - 751fd023 by Finley McIlwaine at 2023-06-01T14:11:41-06:00 Accept HTML test diffs All diffs now boil down to the expected differences resulting from declarations being reified from TyThings in hi-haddock. Surface syntax now has much less control over the syntax used in the documentation. - - - - - d835c845 by Finley McIlwaine at 2023-06-01T14:11:45-06:00 Adapt to new load' type - - - - - dcf776c4 by Finley McIlwaine at 2023-06-01T14:13:13-06:00 Update mkWarningMap and moduleWarning - - - - - 8e8432fd by Finley McIlwaine at 2023-06-01T14:28:54-06:00 Revert load' changes - - - - - aeb2982c by Finley McIlwaine at 2023-06-01T14:40:24-06:00 Accept change to Instances test in html-test Link to Data.Tuple instead of GHC.Tuple.Prim - - - - - 8adfdbac by Finley McIlwaine at 2023-06-01T15:53:17-06:00 Reset ghc dep to ^>= 9.6 - - - - - 2b1ce93d by Finley McIlwaine at 2023-06-06T07:50:04-06:00 Update CHANGES.md, user guide, recomp avoidance * Add --trace-args flag for tracing arguments received to standard output * Avoid recompiling due to changes in optimization flags * Update users guide and changes.md - - - - - f3da6676 by Finley McIlwaine at 2023-06-06T14:12:56-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - bf36c467 by Matthew Pickering at 2023-06-07T10:16:09+01:00 Revert back to e16e20d592a6f5d9ed1af17b77fafd6495242345 Neither of these MRs are ready to land yet which causes issues with other MRs which are ready to land and need haddock changes. - - - - - 421510a9 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 atSign has no unicode variant Prior to this change, atSign was defined as follows: atSign unicode = text (if unicode then "@" else "@") Yes, this is the same symbol '\64' and not your font playing tricks on you. Now we define: atSign = char '@' Both the LaTeX and the Xhtml backend are updated accordingly. - - - - - 3785c276 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 LaTeX: fix printing of type variable bindings Consider this type signature: kindOf :: forall {k} (a :: k). Proxy a -> Proxy k Prior to this fix, the LaTeX backend rendered it like this: kindOf :: forall k a. Proxy a -> Proxy k Now we preserve explicit specificity and kind annotations. - - - - - 0febf3a8 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 Add support for invisible binders in type declarations - - - - - 13e33bb3 by Finley McIlwaine at 2023-06-08T07:51:59-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - 3e5340ce by Finley McIlwaine at 2023-06-08T07:54:27-06:00 Add note about stubdir to recompilation docs - - - - - db7e84dc by Finley at 2023-06-08T08:11:03-06:00 Merge pull request haskell/haddock#1597 from haskell/finley/hi-haddock-9.6 hi-haddock for ghc 9.6 - - - - - 4e085d17 by Finley McIlwaine at 2023-06-14T13:41:06-06:00 Replace SYB traversals - - - - - 7b39aec5 by Finley McIlwaine at 2023-06-14T14:20:17-06:00 Test ref accepts, remove unused HaddockClsInst - - - - - df9c2090 by Finley McIlwaine at 2023-06-15T08:02:51-06:00 Use batchMsg for progress reporting during load With hi-haddock as is, there is an awkward silence during the load operation. This commit makes haddock use the default `batchMsg` Messager for progress reporting, and makes the default GHC verbosity level 1, so the user can see what GHC is doing. - - - - - f23679a8 by Hécate Moonlight at 2023-06-15T20:31:53+02:00 Merge pull request haskell/haddock#1600 from haskell/finley/hi-haddock-optim - - - - - a7982192 by Finley McIlwaine at 2023-06-15T15:02:16-06:00 hi-haddock squashed - - - - - c34f0c8d by Finley McIlwaine at 2023-06-15T16:22:03-06:00 Merge remote-tracking branch 'origin/ghc-9.6' into finley/hi-haddock-squashed - - - - - 40452797 by Bartłomiej Cieślar at 2023-06-16T12:26:04+02:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - e58673bf by Ben Gamari at 2023-06-16T09:33:35-04:00 Bump GHC version to 9.8 - - - - - 74bdf972 by Ben Gamari at 2023-06-16T09:36:18-04:00 Merge commit 'fcaaad06770a26d35d4aafd65772dedadf17669c' into ghc-head - - - - - 418ee3dc by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Remove NFData SourceText, IfaceWarnings updates The NFData SourceText instance is now available in GHC Handle mod_iface mi_warns now being IfaceWarnings - - - - - 62f31380 by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Accept Instances.hs test output Due to ghc!10469. - - - - - a8f2fc0e by Ben Gamari at 2023-06-20T15:48:08-04:00 Test fixes for "Fix associated data family doc structure items" Associated data families were being given their own export DocStructureItems, which resulted in them being documented separately from their classes in haddocks. This commit fixes it. - - - - - cb1ac33e by Bartłomiej Cieślar at 2023-06-21T12:56:02-04:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - 9933e10b by Ben Gamari at 2023-06-21T12:56:02-04:00 Bump GHC version to 9.8 - - - - - fe8c18b6 by Ben Gamari at 2023-06-21T15:36:29-04:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - c61a0d5b by Ben Gamari at 2023-06-21T16:10:51-04:00 Bump GHC version to 9.9 - - - - - 0c2a756e by sheaf at 2023-07-07T13:45:12+02:00 Avoid incomplete record update in Haddock Hoogle This commit avoids running into an incomplete record update warning in the Hoogle backend of Haddock. This was only noticed now, because incomplete record updates were broken in GHC 9.6. Now that they are fixed, we have to avoid running into them! - - - - - f9b952a7 by Ben Gamari at 2023-07-21T11:58:05-04:00 Bump base bound to <4.20 For GHC 9.8. - - - - - 1b27e151 by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Check for puns (see ghc#23368) - - - - - 457341fd by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Remove fake exports for (~), List, and Tuple<n> The old reasoning no longer applies, nowadays those names can be mentioned in export lists. - - - - - bf3dcddf by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Fix pretty-printing of Solo and MkSolo - - - - - 7acbf0fd by Sven Tennie at 2023-08-10T19:17:11-04:00 Serialize CmmRetInfo in .rodata The handling of case was missing. - - - - - 0c3136f2 by Sven Tennie at 2023-08-10T19:17:11-04:00 Reference StgRetFun payload by its struct field address This is easier to grasp than relative pointer offsets. - - - - - f68ff313 by Sven Tennie at 2023-08-10T19:17:11-04:00 Better variable name: u -> frame The 'u' was likely introduced by copy'n'paste. - - - - - 0131bb7f by Sven Tennie at 2023-08-10T19:17:11-04:00 Make checkSTACK() public Such that it can also be used in tests. - - - - - 7b6e1e53 by Sven Tennie at 2023-08-10T19:17:11-04:00 Publish stack related fields in DerivedConstants.h These will be used in ghc-heap to decode these parts of the stack. - - - - - 907ed054 by Sven Tennie at 2023-08-10T19:17:11-04:00 ghc-heap: Decode StgStack and its stack frames Previously, ghc-heap could only decode heap closures. The approach is explained in detail in note [Decoding the stack]. - - - - - 6beb6ac2 by Sven Tennie at 2023-08-10T19:17:11-04:00 Remove RetFunType from RetFun stack frame representation It's a technical detail. The single usage is replaced by a predicate. - - - - - 006bb4f3 by Sven Tennie at 2023-08-10T19:17:11-04:00 Better parameter name The call-site uses the term "offset", too. - - - - - d4c2c1af by Sven Tennie at 2023-08-10T19:17:11-04:00 Make closure boxing pure There seems to be no need to do something complicated. However, the strictness of the closure pointer matters, otherwise a thunk gets decoded. - - - - - 8d8426c9 by Sven Tennie at 2023-08-10T19:17:11-04:00 Document entertainGC in test It wasn't obvious why it's there and what its role is. Also, increase the "entertainment level" a bit. I checked in STG and Cmm dumps that this really generates closures (and is not e.g. constant folded away.) - - - - - cc52c358 by Finley McIlwaine at 2023-08-10T19:17:47-04:00 Add -dipe-stats flag This is useful for seeing which info tables have information. - - - - - 261c4acb by Finley McIlwaine at 2023-08-10T19:17:47-04:00 Add -finfo-table-map-with-fallback -finfo-table-map-with-stack The -fno-info-table-map-with-stack flag omits STACK info tables from the info table map, and the -fno-info-table-map-with-fallback flag omits info tables with defaulted source locations from the map. In a test on the Agda codebase the build results were about 7% smaller when both of those types of tables were omitted. Adds a test that verifies that passing each combination of these flags results in the correct output for -dipe-stats, which is disabled for the js backend since profiling is not implemented. This commit also refactors a lot of the logic around extracting info tables from the Cmm results and building the info table map. This commit also fixes some issues in the users guide rst source to fix warnings that were noticed while debugging the documentation for these flags. Fixes #23702 - - - - - d7047e0d by Jaro Reinders at 2023-08-14T04:41:42-04:00 Add changelog entry for specialised Enum Int64/Word64 instances - - - - - 52f5e8fb by cydparser at 2023-08-14T04:42:20-04:00 Fix -ddump-to-file and -ddump-timings interaction (#20316) - - - - - 1274c5d6 by cydparser at 2023-08-14T04:42:20-04:00 Update release notes (#20316) - - - - - 8e699b23 by Matthew Pickering at 2023-08-14T10:44:47-04:00 base: Add changelog entry for CLC #188 This proposal modified the implementations of copyBytes, moveBytes and fillBytes (as detailed in the proposal) https://github.com/haskell/core-libraries-committee/issues/188 - - - - - 026f040a by Matthew Pickering at 2023-08-14T10:45:23-04:00 packaging: Build manpage in separate directory to other documentation We were installing two copies of the manpage: * One useless one in the `share/doc` folder, because we copy the doc/ folder into share/ * The one we deliberately installed into `share/man` etc The solution is to build the manpage into the `manpage` directory when building the bindist, and then just install it separately. Fixes #23707 - - - - - 524c60c8 by Bartłomiej Cieślar at 2023-08-14T13:46:33-04:00 Report deprecated fields bound by record wildcards when used This commit ensures that we emit the appropriate warnings when a deprecated record field bound by a record wildcard is used. For example: module A where data Foo = Foo {x :: Int, y :: Bool, z :: Char} {-# DEPRECATED x "Don't use x" #-} {-# WARNING y "Don't use y" #-} module B where import A foo (Foo {..}) = x This will cause us to emit a "Don't use x" warning, with location the location of the record wildcard. Note that we don't warn about `y`, because it is unused in the RHS of `foo`. Fixes #23382 - - - - - d6130065 by Matthew Pickering at 2023-08-14T13:47:11-04:00 Add zstd suffix to jobs which rely on zstd This was causing some confusion as the job was named simply "x86_64-linux-deb10-validate", which implies a standard configuration rather than any dependency on libzstd. - - - - - e24e44fc by Matthew Pickering at 2023-08-14T13:47:11-04:00 ci: Always run project-version job This is needed for the downstream test-primops pipeline to workout what the version of a bindist produced by a pipeline is. - - - - - f17b9d62 by Matthew Pickering at 2023-08-14T13:47:11-04:00 gen_ci: Rework how jobs-metadata.json is generated * We now represent a job group a triple of Maybes, which makes it easier to work out when jobs are enabled/disabled on certain pipelines. ``` data JobGroup a = StandardTriple { v :: Maybe (NamedJob a) , n :: Maybe (NamedJob a) , r :: Maybe (NamedJob a) } ``` * `jobs-metadata.json` generation is reworked using the following algorithm. - For each pipeline type, find all the platforms we are doing builds for. - Select one build per platform - Zip together the results This way we can choose different pipelines for validate/nightly/release which makes the metadata also useful for validate pipelines. This feature is used by the test-primops downstream CI in order to select the right bindist for testing validate pipelines. This makes it easier to inspect which jobs are going to be enabled on a particular pipeline. - - - - - f9a5563d by Matthew Pickering at 2023-08-14T13:47:11-04:00 gen_ci: Rules rework In particular we now distinguish between whether we are dealing with a Nightly/Release pipeline (which labels don't matter for) and a validate pipeline where labels do matter. The overall goal here is to allow a disjunction of labels for validate pipelines, for example, > Run a job if we have the full-ci label or test-primops label Therefore the "ValidateOnly" rules are treated as a set of disjunctions rather than conjunctions like before. What this means in particular is that if we want to ONLY run a job if a label is set, for example, "FreeBSD" label then we have to override the whole label set. Fixes #23772 - - - - - d54b0c1d by Matthew Pickering at 2023-08-14T13:47:11-04:00 ci: set -e for lint-ci-config scripts - - - - - 994a9b35 by Matthew Pickering at 2023-08-14T13:47:11-04:00 ci: Fix job metadata generation - - - - - e194ed2b by Ben Gamari at 2023-08-15T00:58:09-04:00 users-guide: Note that GHC2021 doesn't include ExplicitNamespaces As noted in #23801. - - - - - d814bda9 by Ben Gamari at 2023-08-15T05:43:53-04:00 users-guide: Support both distutils and packaging As noted in #23818, some old distributions (e.g. Debian 9) only include `distutils` while newer distributions only include `packaging`. Fixes #23818. - - - - - 1726db3f by Ben Gamari at 2023-08-15T05:43:53-04:00 users-guide: Ensure extlinks is compatible with Sphinx <4 The semantics of the `extlinks` attribute annoyingly changed in Sphinx 4. Reflect this in our configuration. See #22690. Fixes #23807. - - - - - 173338cf by Matthew Pickering at 2023-08-15T22:00:24-04:00 ci: Run full-ci on master and release branches Fixes #23737 - - - - - bdab6898 by Andrew Lelechenko at 2023-08-15T22:01:03-04:00 Add @since pragmas for Data.Ord.clamp and GHC.Float.clamp - - - - - 662d351b by Matthew Pickering at 2023-08-16T09:35:04-04:00 ghc-toolchain: Match CPP args with configure script At the moment we need ghc-toolchain to precisely match the output as provided by the normal configure script. The normal configure script (FP_HSCPP_CMD_WITH_ARGS) branches on whether we are using clang or gcc so we match that logic exactly in ghc-toolchain. The old implementation (which checks if certain flags are supported) is better but for now we have to match to catch any potential errors in the configuration. Ticket: #23720 - - - - - 09c6759e by Matthew Pickering at 2023-08-16T09:35:04-04:00 configure: Fix `-Wl,--no-as-needed` check The check was failing because the args supplied by $$1 were quoted which failed because then the C compiler thought they were an input file. Fixes #23720 - - - - - 2129678b by Matthew Pickering at 2023-08-16T09:35:04-04:00 configure: Add flag which turns ghc-toolchain check into error We want to catch these errors in CI, but first we need to a flag which turns this check into an error. - - - - - 6e2aa8e0 by Matthew Pickering at 2023-08-16T09:35:04-04:00 ci: Enable --enable-strict-ghc-toolchain-check for all CI jobs This will cause any CI job to fail if we have a mismatch between what ghc-toolchain reports and what ./configure natively reports. Fixing these kinds of issues is highest priority for 9.10 release. - - - - - 12d39e24 by Rodrigo Mesquita at 2023-08-16T09:35:04-04:00 Pass user-specified options to ghc-toolchain The current user interface to configuring target toolchains is `./configure`. In !9263 we added a new tool to configure target toolchains called `ghc-toolchain`, but the blessed way of creating these toolchains is still through configure. However, we were not passing the user-specified options given with the `./configure` invocation to the ghc-toolchain tool. This commit remedies that by storing the user options and environment variables in USER_* variables, which then get passed to GHC-toolchain. The exception to the rule is the windows bundled toolchain, which overrides the USER_* variables with whatever flags the windows bundled toolchain requires to work. We consider the bundled toolchain to be effectively the user specifying options, since the actual user delegated that configuration work. Closes #23678 - - - - - f7b3c3a0 by Rodrigo Mesquita at 2023-08-16T09:35:04-04:00 ghc-toolchain: Parse javascript and ghcjs as a Arch and OS - - - - - 8a0ae4ee by Rodrigo Mesquita at 2023-08-16T09:35:04-04:00 ghc-toolchain: Fix ranlib option - - - - - 31e9ec96 by Rodrigo Mesquita at 2023-08-16T09:35:04-04:00 Check Link Works with -Werror - - - - - bc1998b3 by Matthew Pickering at 2023-08-16T09:35:04-04:00 Only check for no_compact_unwind support on darwin While writing ghc-toolchain we noticed that the FP_PROG_LD_NO_COMPACT_UNWIND check is subtly wrong. Specifically, we pass -Wl,-no_compact_unwind to cc. However, ld.gold interprets this as -n o_compact_unwind, which is a valid argument. Fixes #23676 - - - - - 0283f36e by Matthew Pickering at 2023-08-16T09:35:04-04:00 Add some javascript special cases to ghc-toolchain On javascript there isn't a choice of toolchain but some of the configure checks were not accurately providing the correct answer. 1. The linker was reported as gnu LD because the --version output mentioned gnu LD. 2. The --target flag makes no sense on javascript but it was just ignored by the linker, so we add a special case to stop ghc-toolchain thinking that emcc supports --target when used as a linker. - - - - - a48ec5f8 by Matthew Pickering at 2023-08-16T09:35:04-04:00 check for emcc in gnu_LD check - - - - - 50df2e69 by Matthew Pickering at 2023-08-16T09:35:04-04:00 Add ldOverrideWhitelist to only default to ldOverride on windows/linux On some platforms - ie darwin, javascript etc we really do not want to allow the user to use any linker other than the default one as this leads to all kinds of bugs. Therefore it is a bit more prudant to add a whitelist which specifies on which platforms it might be possible to use a different linker. - - - - - a669a39c by Matthew Pickering at 2023-08-16T09:35:04-04:00 Fix plaform glob in FPTOOLS_SET_C_LD_FLAGS A normal triple may look like x86_64-unknown-linux but when cross-compiling you get $target set to a quad such as.. aarch64-unknown-linux-gnu Which should also match this check. - - - - - c52b6769 by Matthew Pickering at 2023-08-16T09:35:04-04:00 ghc-toolchain: Pass ld-override onto ghc-toolchain - - - - - 039b484f by Matthew Pickering at 2023-08-16T09:35:04-04:00 ld override: Make whitelist override user given option - - - - - d2b63cbc by Matthew Pickering at 2023-08-16T09:35:05-04:00 ghc-toolchain: Add format mode to normalise differences before diffing. The "format" mode takes an "--input" and "--ouput" target file and formats it. This is intended to be useful on windows where the configure/ghc-toolchain target files can't be diffed very easily because the path separators are different. - - - - - f2b39e4a by Matthew Pickering at 2023-08-16T09:35:05-04:00 ghc-toolchain: Bump ci-images commit to get new ghc-wasm-meta We needed to remove -Wno-unused-command-line-argument from the arguments passed in order for the configure check to report correctly. See https://gitlab.haskell.org/ghc/ghc/-/merge_requests/10976#note_516335 - - - - - 92103830 by Matthew Pickering at 2023-08-16T09:35:05-04:00 configure: MergeObjsCmd - distinguish between empty string and unset variable If `MergeObjsCmd` is explicitly set to the empty string then we should assume that MergeObjs is just not supported. This is especially important for windows where we set MergeObjsCmd to "" in m4/fp_setup_windows_toolchain.m4. - - - - - 3500bb2c by Matthew Pickering at 2023-08-16T09:35:05-04:00 configure: Add proper check to see if object merging works - - - - - 08c9a014 by Matthew Pickering at 2023-08-16T09:35:05-04:00 ghc-toolchain: If MergeObjsCmd is not set, replace setting with Nothing If the user explicitly chooses to not set a MergeObjsCmd then it is correct to use Nothing for tgtMergeObjs field in the Target file. - - - - - c9071d94 by Matthew Pickering at 2023-08-16T09:35:05-04:00 HsCppArgs: Augment the HsCppOptions This is important when we pass -I when setting up the windows toolchain. - - - - - 294a6d80 by Matthew Pickering at 2023-08-16T09:35:05-04:00 Set USER_CPP_ARGS when setting up windows toolchain - - - - - bde4b5d4 by Rodrigo Mesquita at 2023-08-16T09:35:05-04:00 Improve handling of Cc as a fallback - - - - - f4c1c3a3 by Rodrigo Mesquita at 2023-08-16T09:35:05-04:00 ghc-toolchain: Configure Cpp and HsCpp correctly when user specifies flags In ghc-toolchain, we were only /not/ configuring required flags when the user specified any flags at all for the of the HsCpp and Cpp tools. Otherwise, the linker takes into consideration the user specified flags to determine whether to search for a better linker implementation, but already configured the remaining GHC and platform-specific flags regardless of the user options. Other Tools consider the user options as a baseline for further configuration (see `findProgram`), so #23689 is not applicable. Closes #23689 - - - - - bfe4ffac by Matthew Pickering at 2023-08-16T09:35:05-04:00 CPP_ARGS: Put new options after user specified options This matches up with the behaviour of ghc-toolchain, so that the output of both matches. - - - - - a6828173 by Gergő Érdi at 2023-08-16T09:35:41-04:00 If a defaulting plugin made progress, re-zonk wanteds before built-in defaulting Fixes #23821. - - - - - e2b38115 by Sylvain Henry at 2023-08-17T07:54:06-04:00 JS: implement openat(AT_FDCWD...) (#23697) Use `openSync` to implement `openat(AT_FDCWD...)`. - - - - - a975c663 by sheaf at 2023-08-17T07:54:47-04:00 Use unsatisfiable for missing methods w/ defaults When a class instance has an Unsatisfiable constraint in its context and the user has not explicitly provided an implementation of a method, we now always provide a RHS of the form `unsatisfiable @msg`, even if the method has a default definition available. This ensures that, when deferring type errors, users get the appropriate error message instead of a possible runtime loop, if class default methods were defined recursively. Fixes #23816 - - - - - 45ca51e5 by Ben Gamari at 2023-08-17T15:16:41-04:00 ghc-internal: Initial commit of the skeleton - - - - - 88bbf8c5 by Ben Gamari at 2023-08-17T15:16:41-04:00 ghc-experimental: Initial commit - - - - - 664468c0 by Ben Gamari at 2023-08-17T15:17:17-04:00 testsuite/cloneStackLib: Fix incorrect format specifiers - - - - - eaa835bb by Ben Gamari at 2023-08-17T15:17:17-04:00 rts/ipe: Fix const-correctness of IpeBufferListNode Both info tables and the string table should be `const` - - - - - 78f6f6fd by Ben Gamari at 2023-08-17T15:17:17-04:00 nonmoving: Drop dead debugging utilities These are largely superceded by support in the ghc-utils GDB extension. - - - - - 3f6e8f42 by Ben Gamari at 2023-08-17T15:17:17-04:00 nonmoving: Refactor management of mark thread Here we refactor that treatment of the worker thread used by the nonmoving GC for concurrent marking, avoiding creating a new thread with every major GC cycle. As well, the new scheme is considerably easier to reason about, consolidating all state in one place, accessed via a small set of accessors with clear semantics. - - - - - 88c32b7d by Ben Gamari at 2023-08-17T15:17:17-04:00 testsuite: Skip T23221 in nonmoving GC ways This test is very dependent upon GC behavior. - - - - - 381cfaed by Ben Gamari at 2023-08-17T15:17:17-04:00 ghc-heap: Don't expose stack dirty and marking fields These are GC metadata and are not relevant to the end-user. Moreover, they are unstable which makes ghc-heap harder to test than necessary. - - - - - 16828ca5 by Luite Stegeman at 2023-08-21T18:42:53-04:00 bump process submodule to include macOS fix and JS support - - - - - b4d5f6ed by Matthew Pickering at 2023-08-21T18:43:29-04:00 ci: Add support for triggering test-primops pipelines This commit adds 4 ways to trigger testing with test-primops. 1. Applying the ~test-primops label to a validate pipeline. 2. A manually triggered job on a validate pipeline 3. A nightly pipeline job 4. A release pipeline job Fixes #23695 - - - - - 32c50daa by Matthew Pickering at 2023-08-21T18:43:29-04:00 Add test-primops label support The test-primops CI job requires some additional builds in the validation pipeline, so we make sure to enable these jobs when test-primops label is set. - - - - - 73ca8340 by Matthew Pickering at 2023-08-21T18:43:29-04:00 Revert "Aarch ncg: Optimize immediate use for address calculations" This reverts commit 8f3b3b78a8cce3bd463ed175ee933c2aabffc631. See #23793 - - - - - 5546ad9e by Matthew Pickering at 2023-08-21T18:43:29-04:00 Revert "AArch NCG: Pure refactor" This reverts commit 00fb6e6b06598752414a0b9a92840fb6ca61338d. See #23793 - - - - - 02dfcdc2 by Matthew Pickering at 2023-08-21T18:43:29-04:00 Revert "Aarch64 NCG: Use encoded immediates for literals." This reverts commit 40425c5021a9d8eb5e1c1046e2d5fa0a2918f96c. See #23793 ------------------------- Metric Increase: T4801 T5321FD T5321Fun ------------------------- - - - - - 7be4a272 by Matthew Pickering at 2023-08-22T08:55:20+01:00 ci: Remove manually triggered test-ci job This doesn't work on slimmed down pipelines as the needed jobs don't exist. If you want to run test-primops then apply the label. - - - - - 76a4d11b by Jaro Reinders at 2023-08-22T08:08:13-04:00 Remove Ptr example from roles docs - - - - - 069729d3 by Bryan Richter at 2023-08-22T08:08:49-04:00 Guard against duplicate pipelines in forks - - - - - f861423b by Rune K. Svendsen at 2023-08-22T08:09:35-04:00 dump-decls: fix "Ambiguous module name"-error Fixes errors of the following kind, which happen when dump-decls is run on a package that contains a module name that clashes with that of another package. ``` dump-decls: <no location info>: error: Ambiguous module name `System.Console.ANSI.Types': it was found in multiple packages: ansi-terminal-0.11.4 ansi-terminal-types-0.11.5 ``` - - - - - edd8bc43 by Krzysztof Gogolewski at 2023-08-22T12:31:20-04:00 Fix MultiWayIf linearity checking (#23814) Co-authored-by: Thomas BAGREL <thomas.bagrel at tweag.io> - - - - - 4ba088d1 by konsumlamm at 2023-08-22T12:32:02-04:00 Update `Control.Concurrent.*` documentation - - - - - 015886ec by ARATA Mizuki at 2023-08-22T15:13:13-04:00 Support 128-bit SIMD on AArch64 via LLVM backend - - - - - 52a6d868 by Krzysztof Gogolewski at 2023-08-22T15:13:51-04:00 Testsuite cleanup - Remove misleading help text in perf_notes, ways are not metrics - Remove no_print_summary - this was used for Phabricator - In linters tests, run 'git ls-files' just once. Previously, it was called on each has_ls_files() - Add ghc-prim.cabal to gitignore, noticed in #23726 - Remove ghc-prim.cabal, it was accidentally committed in 524c60c8cd - - - - - ab40aa52 by Alan Zimmerman at 2023-08-22T15:14:28-04:00 EPA: Use Introduce [DeclTag] in AnnSortKey The AnnSortKey is used to keep track of the order of declarations for printing when the container has split them apart. This applies to HsValBinds and ClassDecl, ClsInstDecl. When making modifications to the list of declarations, the new order must be captured for when it must be printed. For each list of declarations (binds and sigs for a HsValBind) we can just store the list in order. To recreate the list when printing, we must merge them, and this is what the AnnSortKey records. It used to be indexed by SrcSpan, we now simply index by a marker as to which list to take the next item from. - - - - - e7db36c1 by sheaf at 2023-08-23T08:41:28-04:00 Don't attempt pattern synonym error recovery This commit gets rid of the pattern synonym error recovery mechanism (recoverPSB). The rationale is that the fake pattern synonym binding that the recovery mechanism introduced could lead to undesirable knock-on errors, and it isn't really feasible to conjure up a satisfactory binding as pattern synonyms can be used both in expressions and patterns. See Note [Pattern synonym error recovery] in GHC.Tc.TyCl.PatSyn. It isn't such a big deal to eagerly fail compilation on a pattern synonym that doesn't typecheck anyway. Fixes #23467 - - - - - 6ccd9d65 by Ben Gamari at 2023-08-23T08:42:05-04:00 base: Don't use Data.ByteString.Internals.memcpy This function is now deprecated from `bytestring`. Use `Foreign.Marshal.Utils.copyBytes` instead. Fixes #23880. - - - - - 0bfa0031 by Matthew Pickering at 2023-08-23T13:43:48-04:00 hadrian: Uniformly pass buildOptions to all builders in runBuilder In Builder.hs, runBuilderWith mostly ignores the buildOptions in BuildInfo. This leads to hard to diagnose bugs as any build options you pass with runBuilderWithCmdOptions are ignored for many builders. Solution: Uniformly pass buildOptions to the invocation of cmd. Fixes #23845 - - - - - 9cac8f11 by Matthew Pickering at 2023-08-23T13:43:48-04:00 Abstract windows toolchain setup This commit splits up the windows toolchain setup logic into two functions. * FP_INSTALL_WINDOWS_TOOLCHAIN - deals with downloading the toolchain if it isn't already downloaded * FP_SETUP_WINDOWS_TOOLCHAIN - sets the environment variables to point to the correct place FP_SETUP_WINDOWS_TOOLCHAIN is abstracted from the location of the mingw toolchain and also the eventual location where we will install the toolchain in the installed bindist. This is the first step towards #23608 - - - - - 6c043187 by Matthew Pickering at 2023-08-23T13:43:48-04:00 Generate build.mk for bindists The config.mk.in script was relying on some variables which were supposed to be set by build.mk but therefore never were when used to install a bindist. Specifically * BUILD_PROF_LIBS to determine whether we had profiled libraries or not * DYNAMIC_GHC_PROGRAMS to determine whether we had shared libraries or not Not only were these never set but also not really accurate because you could have shared libaries but still statically linked ghc executable. In addition variables like GhcLibWays were just never used, so those have been deleted from the script. Now instead we generate a build.mk file which just directly specifies which RtsWays we have supplied in the bindist and whether we have DYNAMIC_GHC_PROGRAMS. - - - - - fe23629b by Matthew Pickering at 2023-08-23T13:43:48-04:00 hadrian: Add reloc-binary-dist-* targets This adds a command line option to build a "relocatable" bindist. The bindist is created by first creating a normal bindist and then installing it using the `RelocatableBuild=YES` option. This creates a bindist without any wrapper scripts pointing to the libdir. The motivation for this feature is that we want to ship relocatable bindists on windows and this method is more uniform than the ad-hoc method which lead to bugs such as #23608 and #23476 The relocatable bindist can be built with the "reloc-binary-dist" target and supports the same suffixes as the normal "binary-dist" command to specify the compression style. - - - - - 41cbaf44 by Matthew Pickering at 2023-08-23T13:43:48-04:00 packaging: Fix installation scripts on windows/RelocatableBuild case This includes quite a lot of small fixes which fix the installation makefile to work on windows properly. This also required fixing the RelocatableBuild variable which seemed to have been broken for a long while. Sam helped me a lot writing this patch by providing a windows machine to test the changes. Without him it would have taken ages to tweak everything. Co-authored-by: sheaf <sam.derbyshire at gmail.com> - - - - - 03474456 by Matthew Pickering at 2023-08-23T13:43:48-04:00 ci: Build relocatable bindist on windows We now build the relocatable bindist target on windows, which means we test and distribute the new method of creating a relocatable bindist. - - - - - d0b48113 by Matthew Pickering at 2023-08-23T13:43:48-04:00 hadrian: Add error when trying to build binary-dist target on windows The binary dist produced by `binary-dist` target doesn't work on windows because of the wrapper script the makefile installs. In order to not surprise any packagers we just give an error if someone tries to build the old binary-dist target rather than the reloc-binary-dist target. - - - - - 7cbf9361 by Matthew Pickering at 2023-08-23T13:43:48-04:00 hadrian: Remove query' logic to use tooldir - - - - - 03fad42e by Matthew Pickering at 2023-08-23T13:43:48-04:00 configure: Set WindresCmd directly and removed unused variables For some reason there was an indirection via the Windres variable before setting WindresCmd. That indirection led to #23855. I then also noticed that these other variables were just not used anywhere when trying to work out what the correct condition was for this bit of the configure script. - - - - - c82770f5 by sheaf at 2023-08-23T13:43:48-04:00 Apply shellcheck suggestion to SUBST_TOOLDIR - - - - - 896e35e5 by sheaf at 2023-08-23T13:44:34-04:00 Compute hints from TcSolverReportMsg This commit changes how hints are handled in conjunction with constraint solver report messages. Instead of storing `[GhcHint]` in the TcRnSolverReport error constructor, we compute the hints depending on the underlying TcSolverReportMsg. This disentangles the logic and makes it easier to add new hints for certain errors. - - - - - a05cdaf0 by Alexander Esgen at 2023-08-23T13:45:16-04:00 users-guide: remove note about fatal Haddock parse failures - - - - - 4908d798 by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Introduce Data.Enum - - - - - f59707c7 by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Num.Integer - - - - - b1054053 by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Num - - - - - 6baa481d by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Num.Natural - - - - - 2ac15233 by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Float - - - - - f3c489de by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Real - - - - - 94f59eaa by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Eliminate module reexport in GHC.Exception The metric increase here isn't strictly due to this commit but it's a rather small, incidental change. Metric Increase: T8095 T13386 Metric Decrease: T8095 T13386 T18304 - - - - - be1fc7df by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add disclaimers in internal modules To warn users that these modules are internal and their interfaces may change with little warning. As proposed in Core Libraries Committee #146 [CLC146]. [CLC146]: https://github.com/haskell/core-libraries-committee/issues/146 - - - - - 0326f3f4 by sheaf at 2023-08-23T17:37:29-04:00 Bump Cabal submodule We need to bump the Cabal submodule to include commit ec75950 which fixes an issue with a dodgy import Rep(..) which relied on GHC bug #23570 - - - - - 0504cd08 by Facundo Domínguez at 2023-08-23T17:38:11-04:00 Fix typos in the documentation of Data.OldList.permutations - - - - - 1420b8cb by Antoine Leblanc at 2023-08-24T16:18:17-04:00 Be more eager in TyCon boot validity checking This commit performs boot-file consistency checking for TyCons into checkValidTyCl. This ensures that we eagerly catch any mismatches, which prevents the compiler from seeing these inconsistencies and panicking as a result. See Note [TyCon boot consistency checking] in GHC.Tc.TyCl. Fixes #16127 - - - - - d99c816f by Finley McIlwaine at 2023-08-24T16:18:55-04:00 Refactor estimation of stack info table provenance This commit greatly refactors the way we compute estimated provenance for stack info tables. Previously, this process was done using an entirely separate traversal of the whole Cmm code stream to build the map from info tables to source locations. The separate traversal is now fused with the Cmm code generation pipeline in GHC.Driver.Main. This results in very significant code generation speed ups when -finfo-table-map is enabled. In testing, this patch reduces code generation times by almost 30% with -finfo-table-map and -O0, and 60% with -finfo-table-map and -O1 or -O2 . Fixes #23103 - - - - - d3e0124c by Finley McIlwaine at 2023-08-24T16:18:55-04:00 Add a test checking overhead of -finfo-table-map We want to make sure we don't end up with poor codegen performance resulting from -finfo-table-map again as in #23103. This test adds a performance test tracking total allocations while compiling ExactPrint with -finfo-table-map. - - - - - fcfc1777 by Ben Gamari at 2023-08-25T10:58:16-04:00 llvmGen: Add export list to GHC.Llvm.MetaData - - - - - 5880fff6 by Ben Gamari at 2023-08-25T10:58:16-04:00 llvmGen: Allow LlvmLits in MetaExprs This omission appears to be an oversight. - - - - - 86ce92a2 by Ben Gamari at 2023-08-25T10:58:16-04:00 compiler: Move platform feature predicates to GHC.Driver.DynFlags These are useful in `GHC.Driver.Config.*`. - - - - - a6a38742 by Ben Gamari at 2023-08-25T10:58:16-04:00 llvmGen: Introduce infrastructure for module flag metadata - - - - - e9af2cf3 by Ben Gamari at 2023-08-25T10:58:16-04:00 llvmGen: Don't pass stack alignment via command line As of https://reviews.llvm.org/D103048 LLVM no longer supports the `-stack-alignment=...` flag. Instead this information is passed via a module flag metadata node. This requires dropping support for LLVM 11 and 12. Fixes #23870 - - - - - a936f244 by Alan Zimmerman at 2023-08-25T10:58:56-04:00 EPA: Keep track of "in" token for WarningTxt category A warning can now be written with a category, e.g. {-# WARNInG in "x-c" e "d" #-} Keep track of the location of the 'in' keyword and string, as well as the original SourceText of the label, in case it uses character escapes. - - - - - 3df8a653 by Matthew Pickering at 2023-08-25T17:42:18-04:00 Remove redundant import in InfoTableProv The copyBytes function is provided by the import of Foreign. Fixes #23889 - - - - - d6f807ec by Ben Gamari at 2023-08-25T17:42:54-04:00 gitlab/issue-template: Mention report-a-bug - - - - - 50b9f75d by Artin Ghasivand at 2023-08-26T20:02:50+03:30 Added StandaloneKindSignature examples to replace CUSKs ones - - - - - 2f6309a4 by Vladislav Zavialov at 2023-08-27T03:47:37-04:00 Remove outdated CPP in compiler/* and template-haskell/* The boot compiler was bumped to 9.4 in cebb5819b43. There is no point supporting older GHC versions with CPP. - - - - - 5248fdf7 by Zubin Duggal at 2023-08-28T15:01:09+05:30 testsuite: Add regression test for #23861 Simon says this was fixed by commit 8d68685468d0b6e922332a3ee8c7541efbe46137 Author: sheaf <sam.derbyshire at gmail.com> Date: Fri Aug 4 15:28:45 2023 +0200 Remove zonk in tcVTA - - - - - b6903f4d by Zubin Duggal at 2023-08-28T12:33:58-04:00 testsuite: Add regression test for #23864 Simon says this was fixed by commit 59202c800f2c97c16906120ab2561f6e1556e4af Author: Sebastian Graf <sebastian.graf at kit.edu> Date: Fri Mar 31 17:35:22 2023 +0200 CorePrep: Eliminate EmptyCase and unsafeEqualityProof in CoreToStg instead We eliminate EmptyCase by way of `coreToStg (Case e _ _ []) = coreToStg e` now. The main reason is that it plays far better in conjunction with eta expansion (as we aim to do for arguments in CorePrep, #23083), because we can discard any arguments, `(case e of {}) eta == case e of {}`, whereas in `(e |> co) eta` it's impossible to discard the argument. - - - - - 9eecdf33 by sheaf at 2023-08-28T18:54:06+00:00 Remove ScopedTypeVariables => TypeAbstractions This commit implements [amendment 604](https://github.com/ghc-proposals/ghc-proposals/pull/604/) to [GHC proposal 448](https://github.com/ghc-proposals/ghc-proposals/pull/448) by removing the implication of language extensions ScopedTypeVariables => TypeAbstractions To limit breakage, we now allow type arguments in constructor patterns when both ScopedTypeVariables and TypeApplications are enabled, but we emit a warning notifying the user that this is deprecated behaviour that will go away starting in GHC 9.12. Fixes #23776 - - - - - fadd5b4d by sheaf at 2023-08-28T18:54:06+00:00 .stderr: ScopedTypeVariables =/> TypeAbstractions This commit accepts testsuite changes for the changes in the previous commit, which mean that TypeAbstractions is no longer implied by ScopedTypeVariables. - - - - - 4f5fb500 by Greg Steuck at 2023-08-29T07:55:13-04:00 Repair `codes` test on OpenBSD by explicitly requesting extended RE - - - - - 6bbde581 by Vasily Sterekhov at 2023-08-29T12:06:58-04:00 Add test for #23540 `T23540.hs` makes use of `explainEv` from `HieQueries.hs`, so `explainEv` has been moved to `TestUtils.hs`. - - - - - 257bb3bd by Vasily Sterekhov at 2023-08-29T12:06:58-04:00 Add test for #23120 - - - - - 4f192947 by Vasily Sterekhov at 2023-08-29T12:06:58-04:00 Make some evidence uses reachable by toHie Resolves #23540, #23120 This adds spans to certain expressions in the typechecker and renamer, and lets 'toHie' make use of those spans. Therefore the relevant evidence uses for the following syntax will now show up under the expected nodes in 'HieAst's: - Overloaded literals ('IsString', 'Num', 'Fractional') - Natural patterns and N+k patterns ('Eq', 'Ord', and instances from the overloaded literals being matched on) - Arithmetic sequences ('Enum') - Monadic bind statements ('Monad') - Monadic body statements ('Monad', 'Alternative') - ApplicativeDo ('Applicative', 'Functor') - Overloaded lists ('IsList') Also see Note [Source locations for implicit function calls] In the process of handling overloaded lists I added an extra 'SrcSpan' field to 'VAExpansion' - this allows us to more accurately reconstruct the locations from the renamer in 'rebuildHsApps'. This also happens to fix #23120. See the additions to Note [Looking through HsExpanded] - - - - - fe9fcf9d by Sylvain Henry at 2023-08-29T12:07:50-04:00 ghc-heap: rename C file (fix #23898) - - - - - b60d6576 by Krzysztof Gogolewski at 2023-08-29T12:08:29-04:00 Misc cleanup - Builtin.PrimOps: ReturnsAlg was used only for unboxed tuples. Rename to ReturnsTuple. - Builtin.Utils: use SDoc for a panic message. The comment about <<details unavailable>> was obsoleted by e8d356773b56. - TagCheck: fix wrong logic. It was zipping a list 'args' with its version 'args_cmm' after filtering. - Core.Type: remove an outdated 1999 comment about unlifted polymorphic types - hadrian: remove leftover debugging print - - - - - 3054fd6d by Krzysztof Gogolewski at 2023-08-29T12:09:08-04:00 Add a regression test for #23903 The bug has been fixed by commit bad2f8b8aa8424. - - - - - 21584b12 by Ben Gamari at 2023-08-29T19:52:02-04:00 README: Refer to ghc-hq repository for contributor and governance information - - - - - e542d590 by sheaf at 2023-08-29T19:52:40-04:00 Export setInertSet from GHC.Tc.Solver.Monad We used to export getTcSInerts and setTcSInerts from GHC.Tc.Solver.Monad. These got renamed to getInertSet/setInertSet in e1590ddc. That commit also removed the export of setInertSet, but that function is useful for the GHC API. - - - - - 694ec5b1 by sheaf at 2023-08-30T10:18:32-04:00 Don't bundle children for non-parent Avails We used to bundle all children of the parent Avail with things that aren't the parent, e.g. with class C a where type T a meth :: .. we would bundle the whole Avail (C, T, meth) with all of C, T and meth, instead of only with C. Avoiding this fixes #23570 - - - - - d926380d by Krzysztof Gogolewski at 2023-08-30T10:19:08-04:00 Fix typos - - - - - d07080d2 by Josh Meredith at 2023-08-30T19:42:32-04:00 JS: Implement missing C functions `rename`, `realpath`, and `getcwd` (#23806) - - - - - e2940272 by David Binder at 2023-08-30T19:43:08-04:00 Bump submodules of hpc and hpc-bin to version 0.7.0.0 hpc 0.7.0.0 dropped SafeHaskell safety guarantees in order to simplify compatibility with newer versions of the directory package which dropped all SafeHaskell guarantees. - - - - - 5d56d05c by David Binder at 2023-08-30T19:43:08-04:00 Bump hpc bound in ghc.cabal.in - - - - - 99fff496 by Dominik Schrempf at 2023-08-31T00:04:46-04:00 ghc classes documentation: rm redundant comment - - - - - fe021bab by Dominik Schrempf at 2023-08-31T00:04:46-04:00 prelude documentation: various nits - - - - - 48c84547 by Dominik Schrempf at 2023-08-31T00:04:46-04:00 integer documentation: minor corrections - - - - - 20cd12f4 by Dominik Schrempf at 2023-08-31T00:04:46-04:00 real documentation: nits - - - - - dd39bdc0 by sheaf at 2023-08-31T00:05:27-04:00 Add a test for #21765 This issue (of reporting a constraint as being redundant even though removing it causes typechecking to fail) was fixed in aed1974e. This commit simply adds a regression test. Fixes #21765 - - - - - f1ec3628 by Andrew Lelechenko at 2023-08-31T23:53:30-04:00 Export foldl' from Prelude and bump submodules See https://github.com/haskell/core-libraries-committee/issues/167 for discussion Metric Decrease: T8095 T13386 Metric Increase: T13386 T8095 T8095 ghc/alloc decreased on x86_64, but increased on aarch64. T13386 ghc/alloc decreased on x86_64-windows, but increased on other platforms. Neither has anything to do with `foldl'`, so I conclude that both are flaky. - - - - - 3181b97d by Gergő Érdi at 2023-08-31T23:54:06-04:00 Allow cross-tyvar defaulting proposals from plugins Fixes #23832. - - - - - 495b2241 by Matthew Pickering at 2023-09-01T13:02:07+02:00 Fix issue with duplicate reexported definitions (T23616) When a class method was reexported, it's default methods were also showing up in the generated html page. The simplest and most non-invasive fix is to not look for the default method if we are just exporting the class method.. because the backends are just showing default methods when the whole class is exported. In general it would be worthwhile to rewrite this bit of code I think as the logic and what gets included is split over `lookupDocs` and `availExportDecl` it would be clearer to combine the two. The result of lookupDocs is always just passed to availExportDecl so it seems simpler and more obvious to just write the function directly. - - - - - e4af506e by Sebastian Graf at 2023-09-01T14:29:12-04:00 Clarify Note [GlobalId/LocalId] after CorePrep (#23797) Fixes #23797. - - - - - ac29787c by Sylvain Henry at 2023-09-01T14:30:02-04:00 Fix warning with UNPACK on sum type (#23921) - - - - - 9765ac7b by Zubin Duggal at 2023-09-05T00:37:45-04:00 hadrian: track python dependencies in doc rules - - - - - 1578215f by sheaf at 2023-09-05T00:38:26-04:00 Bump Haddock to fix #23616 This commit updates the Haddock submodule to include the fix to #23616. Fixes #23616 - - - - - 5a2fe35a by David Binder at 2023-09-05T00:39:07-04:00 Fix example in GHC user guide in SafeHaskell section The example given in the SafeHaskell section uses an implementation of Monad which no longer works. This MR removes the non-canonical return instance and adds the necessary instances of Functor and Applicative. - - - - - 291d81ae by Matthew Pickering at 2023-09-05T14:03:10-04:00 driver: Check transitive closure of haskell package dependencies when deciding whether to relink We were previously just checking whether direct package dependencies had been modified. This caused issues when compiling without optimisations as we wouldn't relink the direct dependency if one of its dependenices changed. Fixes #23724 - - - - - 35da0775 by Krzysztof Gogolewski at 2023-09-05T14:03:47-04:00 Re-export GHC.Utils.Panic.Plain from GHC.Utils.Panic Fixes #23930 - - - - - 6551824d by Finley McIlwaine at 2023-09-05T13:06:57-07:00 Remove fake export of `FUN` from Prelude This prevents `data FUN` from being shown at the top of the Prelude docs. Fixes \#23920 on GHC. - - - - - 3930d793 by Jaro Reinders at 2023-09-06T18:42:55-04:00 Make STG rewriter produce updatable closures - - - - - 0104221a by Krzysztof Gogolewski at 2023-09-06T18:43:32-04:00 configure: update message to use hadrian (#22616) - - - - - b34f8586 by Alan Zimmerman at 2023-09-07T10:58:38-04:00 EPA: Incorrect locations for UserTyVar with '@' In T13343.hs, the location for the @ is not within the span of the surrounding UserTyVar. type Bad @v = (forall (v1 :: RuntimeRep) (a1 :: TYPE v). a1) :: TYPE v Widen it so it is captured. Closes #23887 - - - - - 8046f020 by Finley McIlwaine at 2023-09-07T10:59:15-04:00 Bump haddock submodule to fix #23920 Removes the fake export of `FUN` from Prelude. Fixes #23920. Bumps haddock submodule. - - - - - e0aa8c6e by Krzysztof Gogolewski at 2023-09-07T11:00:03-04:00 Fix wrong role in mkSelCo_maybe In the Lint failure in #23938, we start with a coercion Refl :: T a ~R T a, and call mkSelCo (SelTyCon 1 nominal) Refl. The function incorrectly returned Refl :: a ~R a. The returned role should be nominal, according to the SelCo rule: co : (T s1..sn) ~r0 (T t1..tn) r = tyConRole tc r0 i ---------------------------------- SelCo (SelTyCon i r) : si ~r ti In this test case, r is nominal while r0 is representational. - - - - - 1d92f2df by Gergő Érdi at 2023-09-08T04:04:30-04:00 If we have multiple defaulting plugins, then we should zonk in between them after any defaulting has taken place, to avoid a defaulting plugin seeing a metavariable that has already been filled. Fixes #23821. - - - - - eaee4d29 by Gergő Érdi at 2023-09-08T04:04:30-04:00 Improvements to the documentation of defaulting plugins Based on @simonpj's draft and comments in !11117 - - - - - ede3df27 by Alan Zimmerman at 2023-09-08T04:05:06-04:00 EPA: Incorrect span for LWarnDec GhcPs The code (from T23465.hs) {-# WARNInG in "x-c" e "d" #-} e = e gives an incorrect span for the LWarnDecl GhcPs Closes #23892 It also fixes the Test23465/Test23464 mixup - - - - - a0ccef7a by Krzysztof Gogolewski at 2023-09-08T04:05:42-04:00 Valid hole fits: don't suggest unsafeCoerce (#17940) - - - - - 9ab5a448 by Alan Zimmerman at 2023-09-08T18:26:53+01:00 Match changes in wip/az/T23885-unicode-funtycon - - - - - 88b942c4 by Oleg Grenrus at 2023-09-08T19:58:42-04:00 Add warning for badly staged types. Resolves #23829. The stage violation results in out-of-bound names in splices. Technically this is an error, but someone might rely on this!? Internal changes: - we now track stages for TyVars. - thLevel (RunSplice _) = 0, instead of panic, as reifyInstances does in fact rename its argument type, and it can contain variables. - - - - - 9861f787 by Ben Gamari at 2023-09-08T19:59:19-04:00 rts: Fix invalid symbol type I suspect this code is dead since we haven't observed this failing despite the obviously incorrect macro name. - - - - - 03ed6a9a by Ben Gamari at 2023-09-08T19:59:19-04:00 testsuite: Add simple test exercising C11 atomics in GHCi See #22012. - - - - - 1aa5733a by Ben Gamari at 2023-09-08T19:59:19-04:00 rts/RtsSymbols: Add AArch64 outline atomic operations Fixes #22012 by adding the symbols described in https://github.com/llvm/llvm-project/blob/main/llvm/docs/Atomics.rst#libcalls-atomic. Ultimately this would be better addressed by #22011, but this is a first step in the right direction and fixes the immediate symptom. Note that we dropped the `__arch64_cas16` operations as these provided by all platforms's compilers. Also, we don't link directly against the libgcc/compiler-rt definitions but rather provide our own wrappers to work around broken toolchains (e.g. https://bugs.gentoo.org/868018). Generated via https://gitlab.haskell.org/ghc/ghc/-/snippets/5733. - - - - - 8f7d3041 by Matthew Pickering at 2023-09-08T19:59:55-04:00 ci: Build debian12 and fedora38 bindists This adds builds for the latest releases for fedora and debian We build these bindists in nightly and release pipelines. - - - - - a1f0d55c by Felix Leitz at 2023-09-08T20:00:37-04:00 Fix documentation around extension implication for MultiParamTypeClasses/ConstrainedClassMethods. - - - - - 98166389 by Teo Camarasu at 2023-09-12T04:30:54-04:00 docs: move -xn flag beside --nonmoving-gc It makes sense to have these beside each other as they are aliases. - - - - - f367835c by Teo Camarasu at 2023-09-12T04:30:55-04:00 nonmoving: introduce a family of dense allocators Supplement the existing power 2 sized nonmoving allocators with a family of dense allocators up to a configurable threshold. This should reduce waste from rounding up block sizes while keeping the amount of allocator sizes manageable. This patch: - Adds a new configuration option `--nonmoving-dense-allocator-count` to control the amount of these new dense allocators. - Adds some constants to `NonmovingAllocator` in order to keep marking fast with the new allocators. Resolves #23340 - - - - - 2b07bf2e by Teo Camarasu at 2023-09-12T04:30:55-04:00 Add changelog entry for #23340 - - - - - f96fe681 by sheaf at 2023-09-12T04:31:44-04:00 Use printGhciException in run{Stmt, Decls} When evaluating statements in GHCi, we need to use printGhciException instead of the printException function that GHC provides in order to get the appropriate error messages that are customised for ghci use. - - - - - d09b932b by psilospore at 2023-09-12T04:31:44-04:00 T23686: Suggest how to enable Language Extension when in ghci Fixes #23686 - - - - - da30f0be by Matthew Craven at 2023-09-12T04:32:24-04:00 Unarise: Split Rubbish literals in function args Fixes #23914. Also adds a check to STG lint that these args are properly unary or nullary after unarisation - - - - - 261b6747 by Matthew Pickering at 2023-09-12T04:33:04-04:00 darwin: Bump MAXOSX_DEPLOYMENT_TARGET to 10.13 This bumps the minumum supported version to 10.13 (High Sierra) which is 6 years old at this point. Fixes #22938 - - - - - f418f919 by Mario Blažević at 2023-09-12T04:33:45-04:00 Fix TH pretty-printing of nested GADTs, issue #23937 This commit fixes `Language.Haskell.TH.Ppr.pprint` so that it correctly pretty-prints GADTs declarations contained within data family instances. Fixes #23937 - - - - - d7a64753 by John Ericson at 2023-09-12T04:34:20-04:00 Put hadrian non-bootstrap plans through `jq` This makes it possible to review changes with conventional diffing tools. This is picking up where ad8cfed4195b1bbfc15b841f010e75e71f63157d left off. - - - - - ff0a709a by Sylvain Henry at 2023-09-12T08:46:28-04:00 JS: fix some tests - Tests using Setup programs need to pass --with-hc-pkg - Several other fixes See https://gitlab.haskell.org/ghc/ghc/-/wikis/javascript-backend/bug_triage for the current status - - - - - fc86f0e7 by Krzysztof Gogolewski at 2023-09-12T08:47:04-04:00 Fix in-scope set assertion failure (#23918) Patch by Simon - - - - - 21a906c2 by Matthew Pickering at 2023-09-12T17:21:04+02:00 Add -Winconsistent-flags warning The warning fires when inconsistent command line flags are passed. For example: * -dynamic-too and -dynamic * -dynamic-too on windows * -O and --interactive * etc This is on by default and allows users to control whether the warning is displayed and whether it should be an error or not. Fixes #22572 - - - - - dfc4f426 by Krzysztof Gogolewski at 2023-09-12T20:31:35-04:00 Avoid serializing BCOs with the internal interpreter Refs #23919 - - - - - 9217950b by Finley McIlwaine at 2023-09-13T08:06:03-04:00 Fix numa auto configure - - - - - 98e7c1cf by Simon Peyton Jones at 2023-09-13T08:06:40-04:00 Add -fno-cse to T15426 and T18964 This -fno-cse change is to avoid these performance tests depending on flukey CSE stuff. Each contains several independent tests, and we don't want them to interact. See #23925. By killing CSE we expect a 400% increase in T15426, and 100% in T18964. Metric Increase: T15426 T18964 - - - - - 236a134e by Simon Peyton Jones at 2023-09-13T08:06:40-04:00 Tiny refactor canEtaReduceToArity was only called internally, and always with two arguments equal to zero. This patch just specialises the function, and renames it to cantEtaReduceFun. No change in behaviour. - - - - - 56b403c9 by Ben Gamari at 2023-09-13T19:21:36-04:00 spec-constr: Lift argument limit for SPEC-marked functions When the user adds a SPEC argument to a function, they are informing us that they expect the function to be specialised. However, previously this instruction could be preempted by the specialised-argument limit (sc_max_args). Fix this. This fixes #14003. - - - - - 6840012e by Simon Peyton Jones at 2023-09-13T19:22:13-04:00 Fix eta reduction Issue #23922 showed that GHC was bogusly eta-reducing a join point. We should never eta-reduce (\x -> j x) to j, if j is a join point. It is extremly difficult to trigger this bug. It took me 45 mins of trying to make a small tests case, here immortalised as T23922a. - - - - - e5c00092 by Andreas Klebinger at 2023-09-14T08:57:43-04:00 Profiling: Properly escape characters when using `-pj`. There are some ways in which unusual characters like quotes or others can make it into cost centre names. So properly escape these. Fixes #23924 - - - - - ec490578 by Ellie Hermaszewska at 2023-09-14T08:58:24-04:00 Use clearer example variable names for bool eliminator - - - - - 5126a2fe by Sylvain Henry at 2023-09-15T11:18:02-04:00 Add missing int64/word64-to-double/float rules (#23907) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/203 - - - - - 566ef411 by Mario Blažević at 2023-09-15T11:18:43-04:00 Fix and test TH pretty-printing of type operator role declarations This commit fixes and tests `Language.Haskell.TH.Ppr.pprint` so that it correctly pretty-prints `type role` declarations for operator names. Fixes #23954 - - - - - 8e05c54a by Simon Peyton Jones at 2023-09-16T01:42:33-04:00 Use correct FunTyFlag in adjustJoinPointType As the Lint error in #23952 showed, the function adjustJoinPointType was failing to adjust the FunTyFlag when adjusting the type. I don't think this caused the seg-fault reported in the ticket, but it is definitely. This patch fixes it. It is tricky to come up a small test case; Krzysztof came up with this one, but it only triggers a failure in GHC 9.6. - - - - - 778c84b6 by Pierre Le Marre at 2023-09-16T01:43:15-04:00 Update to Unicode 15.1.0 See: https://www.unicode.org/versions/Unicode15.1.0/ - - - - - f9d79a6c by Alan Zimmerman at 2023-09-18T00:00:14-04:00 EPA: track unicode version for unrestrictedFunTyCon Closes #23885 Updates haddock submodule - - - - - 9374f116 by Andrew Lelechenko at 2023-09-18T00:00:54-04:00 Bump parsec submodule to allow text-2.1 and bytestring-0.12 - - - - - 7ca0240e by Ben Gamari at 2023-09-18T15:16:48-04:00 base: Advertise linear time of readFloat As noted in #23538, `readFloat` has runtime that scales nonlinearly in the size of its input. Consequently, its use on untrusted input can be exploited as a denial-of-service vector. Point this out and suggest use of `read` instead. See #23538. - - - - - f3f58f13 by Simon Peyton Jones at 2023-09-18T15:17:24-04:00 Remove dead code GHC.CoreToStg.Prep.canFloat This function never fires, so we can delete it: #23965. - - - - - ccab5b15 by Ben Gamari at 2023-09-18T15:18:02-04:00 base/changelog: Move fix for #23907 to 9.8.1 section Since the fix was backported to 9.8.1 - - - - - 51b57d65 by Matthew Pickering at 2023-09-19T08:44:31-04:00 Add aarch64 alpine bindist This is dynamically linked and makes creating statically linked executables more straightforward. Fixes #23482 - - - - - 02c87213 by Matthew Pickering at 2023-09-19T08:44:31-04:00 Add aarch64-deb11 bindist This adds a debian 11 release job for aarch64. Fixes #22005 - - - - - 8b61dfd6 by Alexis King at 2023-09-19T08:45:13-04:00 Don’t store the async exception masking state in CATCH frames - - - - - 86d2971e by doyougnu at 2023-09-19T19:08:19-04:00 compiler,ghci: error codes link to HF error index closes: #23259 - adds -fprint-error-index-links={auto|always|never} flag - - - - - 5f826c18 by sheaf at 2023-09-19T19:09:03-04:00 Pass quantified tyvars in tcDefaultAssocDecl This commit passes the correct set of quantified type variables written by the user in associated type default declarations for validity checking. This ensures that validity checking of associated type defaults mirrors that of standalone type family instances. Fixes #23768 (see testcase T23734 in subsequent commit) - - - - - aba18424 by sheaf at 2023-09-19T19:09:03-04:00 Avoid panic in mkGADTVars This commit avoids panicking in mkGADTVars when we encounter a type variable as in #23784 that is bound by a user-written forall but not actually used. Fixes #23784 - - - - - a525a92a by sheaf at 2023-09-19T19:09:03-04:00 Adjust reporting of unused tyvars in data FamInsts This commit adjusts the validity checking of data family instances to improve the reporting of unused type variables. See Note [Out of scope tvs in data family instances] in GHC.Tc.Validity. The problem was that, in a situation such as data family D :: Type data instance forall (d :: Type). D = MkD the RHS passed to 'checkFamPatBinders' would be the TyCon app R:D d which mentions the type variable 'd' quantified in the user-written forall. Thus, when computing the set of unused type variables in the RHS of the data family instance, we would find that 'd' is used, and report a strange error message that would say that 'd' is not bound on the LHS. To fix this, we special-case the data-family instance case, manually extracting all the type variables that appear in the arguments of all the data constructores of the data family instance. Fixes #23778 - - - - - 28dd52ee by sheaf at 2023-09-19T19:09:03-04:00 Unused tyvars in FamInst: only report user tyvars This commit changes how we perform some validity checking for coercion axioms to mirror how we handle default declarations for associated type families. This allows us to keep track of whether type variables in type and data family instances were user-written or not, in order to only report the user-written ones in "unused type variable" error messages. Consider for example: {-# LANGUAGE PolyKinds #-} type family F type instance forall a. F = () In this case, we get two quantified type variables, (k :: Type) and (a :: k); the second being user-written, but the first is introduced by the typechecker. We should only report 'a' as being unused, as the user has no idea what 'k' is. Fixes #23734 - - - - - 1eed645c by sheaf at 2023-09-19T19:09:03-04:00 Validity: refactor treatment of data families This commit refactors the reporting of unused type variables in type and data family instances to be more principled. This avoids ad-hoc logic in the treatment of data family instances. - - - - - 35bc506b by John Ericson at 2023-09-19T19:09:40-04:00 Remove `ghc-cabal` It is dead code since the Make build system was removed. I tried to go over every match of `git grep -i ghc-cabal` to find other stray bits. Some of those might be workarounds that can be further removed. - - - - - 665ca116 by John Paul Adrian Glaubitz at 2023-09-19T19:10:39-04:00 Re-add unregisterised build support for sparc and sparc64 Closes #23959 - - - - - 142f8740 by Matthew Pickering at 2023-09-19T19:11:16-04:00 Bump ci-images to use updated version of Alex Fixes #23977 - - - - - fa977034 by John Ericson at 2023-09-21T12:55:25-04:00 Use Cabal 3.10 for Hadrian We need the newer version for `CABAL_FLAG_*` env vars for #17191. - - - - - a5d22cab by John Ericson at 2023-09-21T12:55:25-04:00 hadrian: `need` any `configure` script we will call When the script is changed, we should reconfigure. - - - - - db882b57 by John Ericson at 2023-09-21T12:55:25-04:00 hadrian: Make it easier to debug Cabal configure Right now, output is squashed. This make per-package configure scripts extremely hard to maintain, because we get vague "library is missing" errors when the actually probably is usually completely unrelated except for also involving the C/C++ toolchain. (I can always pass `-VVV` to Hadrian locally, but these errors are subtle and I often cannot reproduce them locally!) `--disable-option-checking` was added back in 75c6e0684dda585c37b4ac254cd7a13537a59a91 but seems to be a bit overkill; if other flags are passed that are not recognized behind the two from Cabal mentioned in the former comment, we *do* want to know about it. - - - - - 7ed65f5a by John Ericson at 2023-09-21T12:55:25-04:00 hadrian: Increase verbosity of certain cabal commands This is a hack to get around the cabal function we're calling *decreasing* the verbosity it passes to another function, which is the stuff we often actually care about. Sigh. Keeping this a separate commit so if this makes things too verbose it is easy to revert. - - - - - a4fde569 by John Ericson at 2023-09-21T12:55:25-04:00 rts: Move most external symbols logic to the configure script This is much more terse because we are programmatically handling the leading underscore. `findPtr` however is still handled in the Cabal file because we need a newer Cabal to pass flags to the configure script automatically. Co-Authored-By: Ben Gamari <ben at well-typed.com> - - - - - 56cc85fb by Andrew Lelechenko at 2023-09-21T12:56:21-04:00 Bump Cabal submodule to allow text-2.1 and bytestring-0.12 - - - - - 0cd6148c by Matthew Pickering at 2023-09-21T12:56:21-04:00 hadrian: Generate Distribution/Fields/Lexer.x before creating a source-dist - - - - - b10ba6a3 by Andrew Lelechenko at 2023-09-21T12:56:21-04:00 Bump hadrian's index-state to upgrade alex at least to 3.2.7.3 - - - - - 11ecc37b by Luite Stegeman at 2023-09-21T12:57:03-04:00 JS: correct file size and times Programs produced by the JavaScript backend were returning incorrect file sizes and modification times, causing cabal related tests to fail. This fixes the problem and adds an additional test that verifies basic file information operations. fixes #23980 - - - - - b35fd2cd by Ben Gamari at 2023-09-21T12:57:39-04:00 gitlab-ci: Drop libiserv from upload_ghc_libs libiserv has been merged into the ghci package. - - - - - 37ad04e8 by Ben Gamari at 2023-09-21T12:58:15-04:00 testsuite: Fix Windows line endings - - - - - 5795b365 by Ben Gamari at 2023-09-21T12:58:15-04:00 testsuite: Use makefile_test - - - - - 15118740 by Ben Gamari at 2023-09-21T12:58:55-04:00 system-cxx-std-lib: Add license and description - - - - - 0208f1d5 by Ben Gamari at 2023-09-21T12:59:33-04:00 gitlab/issue-templates: Rename bug.md -> default.md So that it is visible by default. - - - - - 23cc3f21 by Andrew Lelechenko at 2023-09-21T20:18:11+01:00 Bump submodule text to 2.1 - - - - - b8e4fe23 by Andrew Lelechenko at 2023-09-22T20:05:05-04:00 Bump submodule unix to 2.8.2.1 - - - - - 54b2016e by John Ericson at 2023-09-23T11:40:41-04:00 Move lib{numa,dw} defines to RTS configure Clean up the m4 to handle the auto case always and be more consistent. Also simplify the CPP --- we should always have both headers if we are using libnuma. "side effects" (AC_DEFINE, and AC_SUBST) are removed from the macros to better separate searching from actions taken based on search results. This might seem overkill now, but will make shuffling logic between configure scripts easier later. The macro comments are converted from `dnl` to `#` following the recomendation in https://www.gnu.org/software/autoconf/manual/autoconf-2.71/html_node/Macro-Definitions.html - - - - - d51b601b by John Ericson at 2023-09-23T11:40:50-04:00 Shuffle libzstd configuring between scripts Like the prior commit for libdw and libnuma, `AC_DEFINE` to RTS configure, `AC_SUBST` goes to the top-level configure script, and the documentation of the m4 macro is improved. - - - - - d1425af0 by John Ericson at 2023-09-23T11:41:03-04:00 Move `FP_ARM_OUTLINE_ATOMICS` to RTS configure It is just `AC_DEFINE` it belongs there instead. - - - - - 18de37e4 by John Ericson at 2023-09-23T11:41:03-04:00 Move mmap in the runtime linker check to the RTS configure `AC_DEFINE` should go there instead. - - - - - 74132c2b by Andrew Lelechenko at 2023-09-25T21:56:54-04:00 Elaborate comment on GHC_NO_UNICODE - - - - - de142aa2 by Ben Gamari at 2023-09-26T15:25:03-04:00 gitlab-ci: Mark T22012 as broken on CentOS 7 Due to #23979. - - - - - 6a896ce8 by Teo Camarasu at 2023-09-26T15:25:39-04:00 hadrian: better error for failing to find file's dependencies Resolves #24004 - - - - - d697a6c2 by Stefan Holdermans at 2023-09-26T20:58:37+00:00 Refactor uses of `partitionEithers . map` This patch changes occurences of the idiom `partitionEithers (map f xs)` by the simpler form `partitionWith f xs` where `partitionWith` is the utility function defined in `GHC.Utils.Misc`. Resolves: #23953 - - - - - 8a2968b7 by Stefan Holdermans at 2023-09-26T20:58:37+00:00 Refactor uses of `partitionEithers <$> mapM f xs` This patch changes occurences of the idiom `partitionEithers <$> mapM f xs` by the simpler form `partitionWithM f xs` where `partitionWithM` is a utility function newly added to `GHC.Utils.Misc`. - - - - - 6a27eb97 by Stefan Holdermans at 2023-09-26T20:58:37+00:00 Mark `GHC.Utils.Misc.partitionWithM` as inlineable This patch adds an `INLINEABLE` pragma for `partitionWithM` to ensure that the right-hand side of the definition of this function remains available for specialisation at call sites. - - - - - f1e5245a by David Binder at 2023-09-27T01:19:00-04:00 Add RTS option to supress tix file - - - - - 1f43124f by David Binder at 2023-09-27T01:19:00-04:00 Add expected output to testsuite in test interface-stability/base-exports - - - - - b9d2c354 by David Binder at 2023-09-27T01:19:00-04:00 Expose HpcFlags and getHpcFlags from GHC.RTS.Flags - - - - - 345675c6 by David Binder at 2023-09-27T01:19:00-04:00 Fix expected output of interface-stability test - - - - - 146e1c39 by David Binder at 2023-09-27T01:19:00-04:00 Implement getHpcFlags - - - - - 61ba8e20 by David Binder at 2023-09-27T01:19:00-04:00 Add section in user guide - - - - - ea05f890 by David Binder at 2023-09-27T01:19:01-04:00 Rename --emit-tix-file to --write-tix-file - - - - - cabce2ce by David Binder at 2023-09-27T01:19:01-04:00 Update the golden files for interface stability - - - - - 1dbdb9d0 by Krzysztof Gogolewski at 2023-09-27T01:19:37-04:00 Refactor: introduce stgArgRep The function 'stgArgType' returns the type in STG. But this violates the abstraction: in STG we're supposed to operate on PrimReps. This introduces stgArgRep ty = typePrimRep (stgArgType ty) stgArgRep1 ty = typePrimRep1 (stgArgType ty) stgArgRep_maybe ty = typePrimRep_maybe (stgArgType ty) stgArgType is still directly used for unboxed tuples (should be fixable), FFI and in ticky. - - - - - b02f8042 by Mario Blažević at 2023-09-27T17:33:28-04:00 Fix TH pretty-printer's parenthesization This PR Fixes `Language.Haskell.TH.Ppr.pprint` so it correctly emits parentheses where needed. Fixes #23962, #23968, #23971, and #23986 - - - - - 79104334 by Krzysztof Gogolewski at 2023-09-27T17:34:04-04:00 Add a testcase for #17564 The code in the ticket relied on the behaviour of Derived constraints. Derived constraints were removed in GHC 9.4 and now the code works as expected. - - - - - d7a80143 by sheaf at 2023-09-28T03:25:53-04:00 lint-codes: add new modes of operation This commit adds two new modes of operation to the lint-codes utility: list - list all statically used diagnostic codes outdated - list all outdated diagnostic codes The previous behaviour is now: test - test consistency and coverage of diagnostic codes - - - - - 477d223c by sheaf at 2023-09-28T03:25:53-04:00 lint codes: avoid using git-grep We manually traverse through the filesystem to find the diagnostic codes embedded in .stdout and .stderr files, to avoid any issues with old versions of grep. Fixes #23843 - - - - - a38ae69a by sheaf at 2023-09-28T03:25:53-04:00 lint-codes: add Hadrian targets This commit adds new Hadrian targets: codes, codes:used - list all used diagnostic codes codes:outdated - list outdated diagnostic codes This allows users to easily query GHC for used and outdated diagnostic codes, e.g. hadrian/build -j --flavour=<..> codes will list all used diagnostic codes in the command line by running the lint-codes utility in the "list codes" mode of operation. The diagnostic code consistency and coverage test is still run as usual, through the testsuite: hadrian/build test --only="codes" - - - - - 9cdd629b by Ben Gamari at 2023-09-28T03:26:29-04:00 hadrian: Install LICENSE files in bindists Fixes #23548. - - - - - b8ebf876 by Matthew Craven at 2023-09-28T03:27:05-04:00 Fix visibility when eta-reducing a type lambda Fixes #24014. - - - - - d3874407 by Torsten Schmits at 2023-09-30T16:08:10-04:00 Fix several mistakes around free variables in iface breakpoints Fixes #23612 , #23607, #23998 and #23666. MR: !11026 The fingerprinting logic in `Iface.Recomp` failed lookups when processing decls containing breakpoints for two reasons: * IfaceBreakpoint created binders for free variables instead of expressions * When collecting free names for the dependency analysis for fingerprinting, breakpoint FVs were skipped - - - - - ef5342cd by Simon Peyton Jones at 2023-09-30T16:08:48-04:00 Refactor to combine HsLam and HsLamCase This MR is pure refactoring (#23916): * Combine `HsLam` and `HsLamCase` * Combine `HsCmdLam` and `HsCmdLamCase` This just arranges to treat uniformly \x -> e \case pi -> ei \cases pis -> ie In the exising code base the first is treated differently to the latter two. No change in behaviour. More specifics: * Combine `HsLam` and `HsLamCase` (constructors of `Language.Haskell.Syntax.Expr.HsExpr`) into one data construtor covering * Lambda * `\case` * `\cases` * The new `HsLam` has an argument of type `HsLamVariant` to distinguish the three cases. * Similarly, combine `HsCmdLam` and `HsCmdLamCase` (constructors of `Language.Haskell.Syntax.Expr.HsCmd` ) into one. * Similarly, combine `mkHsLamPV` and `mkHsLamCasePV` (methods of class `DisambECP`) into one. (Thank you Alan Zimmerman.) * Similarly, combine `LambdaExpr` and `LamCaseAlt` (constructors of `Language.Haskell.Syntax.Expr.HsMatchContext`) into one: `LamAlt` with a `HsLamVariant` argument. * Similarly, combine `KappaExpr` and `ArrowLamCaseAlt` (constructors of `Language.Haskell.Syntax.Expr.HsArrowMatchContext`) into one: `ArrowLamAlt` with a `HsLamVariant` argument. * Similarly, combine `PsErrLambdaInPat` and `PsErrLambdaCaseInPat` (constructors of `GHC.Parser.Errors.Ppr.PsError`) into one. * Similarly, combine `PsErrLambdaInPat` and `PsErrLambdaCaseInPat` (constructors of `GHC.Parser.Errors.Ppr.PsError`) into one. * In the same `PsError` data type, combine `PsErrLambdaCmdInFunAppCmd` and `PsErrLambdaCaseCmdInFunAppCmd` into one. * In the same `PsError` data tpye, combine `PsErrLambdaInFunAppExpr` and `PsErrLambdaCaseInFunAppExpr` into one. p* Smilarly combine `ExpectedFunTyLam` and `ExpectedFunTyLamCase` (constructors of `GHC.Tc.Types.Origin.ExpectedFunTyOrigin`) into one. Phew! - - - - - b048bea0 by Andreas Klebinger at 2023-09-30T16:09:24-04:00 Arm: Make ppr methods easier to use by not requiring NCGConfig - - - - - 2adc0508 by Andreas Klebinger at 2023-09-30T16:09:24-04:00 AArch64: Fix broken conditional jumps for offsets >= 1MB Rewrite conditional jump instructions with offsets >= 1MB to use unconditional jumps to avoid overflowing the immediate. Fixes #23746 - - - - - 1424f790 by Alan Zimmerman at 2023-09-30T16:10:00-04:00 EPA: Replace Monoid with NoAnn We currently use the Monoid class as a constraint on Exact Print Annotation functions, so we can use mempty. But this leads to requiring Semigroup instances too, which do not always make sense. Instead, introduce a class NoAnn, with a function noAnn analogous to mempty. Closes #20372 Updates haddock submodule - - - - - c1a3ecde by Ben Gamari at 2023-09-30T16:10:36-04:00 users-guide: Refactor handling of :base-ref: et al. - - - - - bc204783 by Richard Eisenberg at 2023-10-02T14:50:52+02:00 Simplify and correct nasty case in coercion opt This fixes #21062. No test case, because triggering this code seems challenging. - - - - - 9c9ca67e by Andrew Lelechenko at 2023-10-04T05:42:28-04:00 Bump bytestring submodule to 0.12.0.2 - - - - - 4e46dc2b by Andrew Lelechenko at 2023-10-04T05:42:28-04:00 Inline bucket_match - - - - - f6b2751f by Ben Gamari at 2023-10-04T05:43:05-04:00 configure: Fix #21712 again This is a bit of a shot in the dark to fix #24033, which appears to be another instance of #21712. For some reason the ld-override logic *still* appears to be active on Darwin targets (or at least one). Consequently, on misconfigured systems we may choose a non-`ld64` linker. It's a bit unclear exactly what happened in #24033 but ultimately the check added for #21712 was not quite right, checking for the `ghc_host_os` (the value of which depends upon the bootstrap compiler) instead of the target platform. Fix this. Fixes #24033. - - - - - 2f0a101d by Krzysztof Gogolewski at 2023-10-04T05:43:42-04:00 Add a regression test for #24029 - - - - - 8cee3fd7 by sheaf at 2023-10-04T05:44:22-04:00 Fix non-symbolic children lookup of fixity decl The fix for #23664 did not correctly account for non-symbolic names when looking up children of a given parent. This one-line fix changes that. Fixes #24037 - - - - - a4785b33 by Cheng Shao at 2023-10-04T05:44:59-04:00 rts: fix incorrect ticket reference - - - - - e037f459 by Ben Gamari at 2023-10-04T05:45:35-04:00 users-guide: Fix discussion of -Wpartial-fields * fix a few typos * add a new example showing when the warning fires * clarify the existing example * point out -Wincomplete-record-selects Fixes #24049. - - - - - 8ff3134e by Matthew Pickering at 2023-10-05T05:34:58-04:00 Revert "Pass preprocessor options to C compiler when building foreign C files (#16737)" This reverts commit 1c18d3b41f897f34a93669edaebe6069f319f9e2. `-optP` should pass options to the preprocessor, that might be a very different program to the C compiler, so passing the options to the C compiler is likely to result in `-optP` being useless. Fixes #17185 and #21291 - - - - - 8f6010b9 by Ben Gamari at 2023-10-05T05:35:36-04:00 rts/nonmoving: Fix on LLP64 platforms Previously `NONMOVING_SEGMENT_MASK` and friends were defined with the `UL` size suffix. However, this is wrong on LLP64 platforms like Windows, where `long` is 32-bits. Fixes #23003. Fixes #24042. - - - - - f20d02f8 by Andreas Klebinger at 2023-10-05T05:36:14-04:00 Fix isAArch64Bitmask for 32bit immediates. Fixes #23802 - - - - - 63afb701 by Bryan Richter at 2023-10-05T05:36:49-04:00 Work around perf note fetch failure Addresses #24055. - - - - - 242102f4 by Krzysztof Gogolewski at 2023-10-05T05:37:26-04:00 Add a test for #21348 - - - - - 7d390bce by Rewbert at 2023-10-05T05:38:08-04:00 Fixes #24046 - - - - - 69abb171 by Finley McIlwaine at 2023-10-06T14:06:28-07:00 Ensure unconstrained instance dictionaries get IPE info In the `StgRhsCon` case of `GHC.Stg.Debug.collectStgRhs`, we were not coming up with an initial source span based on the span of the binder, which was causing instance dictionaries without dynamic superclass constraints to not have source locations in their IPE info. Now they do. Resolves #24005 - - - - - 390443b7 by Andreas Klebinger at 2023-10-07T10:00:20-04:00 rts: Split up rts/include/stg/MachRegs.h by arch - - - - - 3685942f by Bryan Richter at 2023-10-07T10:00:56-04:00 Actually set hackage index state Or at least, use a version of the cabal command that *claims* to set the index state. Time will tell. - - - - - 46a0e5be by Bryan Richter at 2023-10-07T10:00:56-04:00 Update hackage index state - - - - - d4b037de by Bryan Richter at 2023-10-07T10:00:56-04:00 Ensure hadrian uses CI's hackage index state - - - - - e206be64 by Andrew Lelechenko at 2023-10-08T15:06:14-04:00 Do not use O_NONBLOCK on regular files or block devices CLC proposal https://github.com/haskell/core-libraries-committee/issues/166 - - - - - a06197c4 by David Binder at 2023-10-08T15:06:55-04:00 Update hpc-bin submodule to 0.69 - - - - - ed6785b6 by David Binder at 2023-10-08T15:06:55-04:00 Update Hadrian with correct path to happy file for hpc-bin - - - - - 94066d58 by Alan Zimmerman at 2023-10-09T21:35:53-04:00 EPA: Introduce HasAnnotation class The class is defined as class HasAnnotation e where noAnnSrcSpan :: SrcSpan -> e This generalises noAnnSrcSpan, and allows noLocA :: (HasAnnotation e) => a -> GenLocated e a noLocA = L (noAnnSrcSpan noSrcSpan) - - - - - 8792a1bc by Ben Gamari at 2023-10-09T21:36:29-04:00 Bump unix submodule to v2.8.3.0 - - - - - e96c51cb by Andreas Klebinger at 2023-10-10T16:44:27+01:00 Add a flag -fkeep-auto-rules to optionally keep auto-generated rules around. The motivation for the flag is given in #21917. - - - - - 3ed58cef by Matthew Pickering at 2023-10-10T19:01:22-04:00 hadrian: Add ghcToolchain to tool args list This allows you to load ghc-toolchain and ghc-toolchain-bin into HLS. - - - - - 476c02d4 by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Normalise triple via config.sub We were not normalising the target triple anymore like we did with the old make build system. Fixes #23856 - - - - - 303dd237 by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Add missing vendor normalisation This is copied from m4/ghc_convert_vendor.m4 Towards #23868 - - - - - 838026c9 by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Add loongarch64 to parseArch Towards #23868 - - - - - 1a5bc0b5 by Matthew Pickering at 2023-10-10T19:01:22-04:00 Add same LD hack to ghc-toolchain In the ./configure script, if you pass the `LD` variable then this has the effect of stopping use searching for a linker and hence passing `-fuse-ld=...`. We want to emulate this logic in ghc-toolchain, if a use explicilty specifies `LD` variable then don't add `-fuse-ld=..` with the goal of making ./configure and ghc-toolchain agree on which flags to use when using the C compiler as a linker. This is quite unsavoury as we don't bake the choice of LD into the configuration anywhere but what's important for now is making ghc-toolchain and ./configure agree as much as possible. See #23857 for more discussion - - - - - 42d50b5a by Ben Gamari at 2023-10-10T19:01:22-04:00 ghc-toolchain: Check for C99 support with -std=c99 Previously we failed to try enabling C99 support with `-std=c99`, as `autoconf` attempts. This broke on older compilers (e.g. CentOS 7) which don't enable C99 by default. Fixes #23879. - - - - - da2961af by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Add endianess check using __BYTE_ORDER__ macro In very old toolchains the BYTE_ORDER macro is not set but thankfully the __BYTE_ORDER__ macro can be used instead. - - - - - d8da73cd by Matthew Pickering at 2023-10-10T19:01:22-04:00 configure: AC_PATH_TARGET_TOOL for LD We want to make sure that LD is set to an absolute path in order to be consistent with the `LD=$(command -v ld)` call. The AC_PATH_TARGET_TOOL macro uses the absolute path rather than AC_CHECK_TARGET_TOOL which might use a relative path. - - - - - 171f93cc by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Check whether we need -std=gnu99 for CPP as well In ./configure the C99 flag is passed to the C compiler when used as a C preprocessor. So we also check the same thing in ghc-toolchain. - - - - - 89a0918d by Matthew Pickering at 2023-10-10T19:01:22-04:00 Check for --target linker flag separately to C compiler There are situations where the C compiler doesn't accept `--target` but when used as a linker it does (but doesn't do anything most likely) In particular with old gcc toolchains, the C compiler doesn't support --target but when used as a linker it does. - - - - - 37218329 by Matthew Pickering at 2023-10-10T19:01:22-04:00 Use Cc to compile test file in nopie check We were attempting to use the C compiler, as a linker, to compile a file in the nopie check, but that won't work in general as the flags we pass to the linker might not be compatible with the ones we pass when using the C compiler. - - - - - 9b2dfd21 by Matthew Pickering at 2023-10-10T19:01:22-04:00 configure: Error when ghc-toolchain fails to compile This is a small QOL change as if you are working on ghc-toolchain and it fails to compile then configure will continue and can give you outdated results. - - - - - 1f0de49a by Matthew Pickering at 2023-10-10T19:01:22-04:00 configure: Check whether -no-pie works when the C compiler is used as a linker `-no-pie` is a flag we pass when using the C compiler as a linker (see pieCCLDOpts in GHC.Driver.Session) so we should test whether the C compiler used as a linker supports the flag, rather than just the C compiler. - - - - - 62cd2579 by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Remove javascript special case for --target detection emcc when used as a linker seems to ignore the --target flag, and for consistency with configure which now tests for --target, we remove this special case. - - - - - 0720fde7 by Ben Gamari at 2023-10-10T19:01:22-04:00 toolchain: Don't pass --target to emscripten toolchain As noted in `Note [Don't pass --target to emscripten toolchain]`, emscripten's `emcc` is rather inconsistent with respect to its treatment of the `--target` flag. Avoid this by special-casing this toolchain in the `configure` script and `ghc-toolchain`. Fixes on aspect of #23744. - - - - - 6354e1da by Matthew Pickering at 2023-10-10T19:01:22-04:00 hadrian: Don't pass `--gcc-options` as a --configure-arg to cabal configure Stop passing -gcc-options which mixed together linker flags and non-linker flags. There's no guarantee the C compiler will accept both of these in each mode. - - - - - c00a4bd6 by Ben Gamari at 2023-10-10T19:01:22-04:00 configure: Probe stage0 link flags For consistency with later stages and CC. - - - - - 1f11e7c4 by Sebastian Graf at 2023-10-10T19:01:58-04:00 Stricter Binary.get in GHC.Types.Unit (#23964) I noticed some thunking while looking at Core. This change has very modest, but throughout positive ghc/alloc effect: ``` hard_hole_fits(normal) ghc/alloc 283,057,664 281,620,872 -0.5% geo. mean -0.1% minimum -0.5% maximum +0.0% ``` Fixes #23964. - - - - - a4f1a181 by Bryan Richter at 2023-10-10T19:02:37-04:00 rel_eng/upload.sh cleanups - - - - - 80705335 by doyougnu at 2023-10-10T19:03:18-04:00 ci: add javascript label rule This adds a rule which triggers the javascript job when the "javascript" label is assigned to an MR. - - - - - a2c0fff6 by Matthew Craven at 2023-10-10T19:03:54-04:00 Make 'wWarningFlagsDeps' include every WarningFlag Fixes #24071. - - - - - d055f099 by Jan Hrček at 2023-10-10T19:04:33-04:00 Fix pretty printing of overlap pragmas in TH splices (fixes #24074) - - - - - 0746b868 by Andreas Klebinger at 2023-10-10T19:05:09-04:00 Aarch64 NCG: Use encoded immediates for literals. Try to generate instr x2, <imm> instead of mov x1, lit instr x2, x1 When possible. This get's rid if quite a few redundant mov instructions. I believe this causes a metric decrease for LargeRecords as we reduce register pressure. ------------------------- Metric Decrease: LargeRecord ------------------------- - - - - - 739f4e6f by Andreas Klebinger at 2023-10-10T19:05:09-04:00 AArch NCG: Refactor getRegister' Remove some special cases which can be handled just as well by the generic case. This increases code re-use while also fixing #23749. Since some of the special case wasn't upholding Note [Signed arithmetic on AArch64]. - - - - - 1b213d33 by Andreas Klebinger at 2023-10-10T19:05:09-04:00 Aarch ncg: Optimize immediate use for address calculations When the offset doesn't fit into the immediate we now just reuse the general getRegister' code path which is well optimized to compute the offset into a register instead of a special case for CmmRegOff. This means we generate a lot less code under certain conditions which is why performance metrics for these improve. ------------------------- Metric Decrease: T4801 T5321FD T5321Fun ------------------------- - - - - - b7df0732 by John Ericson at 2023-10-11T16:02:11-04:00 RTS configure: Move over mem management checks These are for heap allocation, a strictly RTS concern. All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. The RTS configure one has a new ``` AC_CHECK_SIZEOF([void *]) ``` that the top-level configure version didn't have, so that `ac_cv_sizeof_void_p` is defined. Once more code is moved over in latter commits, that can go away. Progress towards #17191 - - - - - 41130a65 by John Ericson at 2023-10-11T16:02:11-04:00 RTS configure: Move over `__thread` check This used by (@bgamari thinks) the `GCThread` abstraction in the RTS. All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. Progress towards #17191 - - - - - cc5ec2bd by John Ericson at 2023-10-11T16:02:11-04:00 RTS configure: Move over misc function checks These are for general use in the RTS. All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. Progress towards #17191 - - - - - 809e7c2d by John Ericson at 2023-10-11T16:02:11-04:00 RTS configure: Move over `eventfd` check This check is for the RTS part of the event manager and has a corresponding part in `base`. All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. Progress towards #17191 - - - - - 58f3babf by John Ericson at 2023-10-11T16:02:48-04:00 Split `FP_CHECK_PTHREADS` and move part to RTS configure `NEED_PTHREAD_LIB` is unused since 3609340743c1b25fdfd0e18b1670dac54c8d8623 (part of the make build system), and so is no longer defined. Progress towards #17191 - - - - - e99cf237 by Moritz Angermann at 2023-10-11T16:03:24-04:00 nativeGen: section flags for .text$foo only Commit 3ece9856d157c85511d59f9f862ab351bbd9b38b, was supposed to fix #22834 in !9810. It does however add "xr" indiscriminatly to .text sections even if splitSections is disabled. This leads to the assembler saying: ghc_1.s:7849:0: error: Warning: Ignoring changed section attributes for .text | 7849 | .section .text,"xr" | ^ - - - - - f383a242 by Sylvain Henry at 2023-10-11T16:04:04-04:00 Modularity: pass TempDir instead of DynFlags (#17957) - - - - - 34fc28b0 by John Ericson at 2023-10-12T06:48:28-04:00 Test that functions from `mingwex` are available Ryan wrote these two minimizations, but they never got added to the test suite. See #23309, #23378 Co-Authored-By: Ben Gamari <bgamari.foss at gmail.com> Co-Authored-By: Ryan Scott <ryan.gl.scott at gmail.com> - - - - - bdb54a0e by John Ericson at 2023-10-12T06:48:28-04:00 Do not check for the `mingwex` library in `/configure` See the recent discussion in !10360 --- Cabal will itself check for the library for the packages that need it, and while the autoconf check additionally does some other things like define a `HAS_LIBMINGWEX` C Preprocessor macro, those other things are also unused and unneeded. Progress towards #17191, which aims to get rid of `/configure` entirely. - - - - - 43e814e1 by Ben Gamari at 2023-10-12T06:49:40-04:00 base: Introduce move modules into src The only non-move changes here are whitespace changes to pass the `whitespace` test and a few testsuite adaptations. - - - - - df81536f by Moritz Angermann at 2023-10-12T06:50:16-04:00 [PEi386 linker] Bounds check and null-deref guard We should resonably be able to expect that we won't exceed the number of sections if we assume to be dealing with legal object files. We can however not guarantee that we get some negative values, and while we try to special case most, we should exclude negative indexing into the sections array. We also need to ensure that we do not try to derefences targetSection, if it is NULL, due to the switch statement. - - - - - c74c4f00 by John Ericson at 2023-10-12T10:31:13-04:00 Move apple compat check to RTS configure - - - - - c80778ea by John Ericson at 2023-10-12T10:31:13-04:00 Move clock/timer fun checks to RTS configure Actual library check (which will set the Cabal flag) is left in the top-level configure for now. Progress towards #17191 - - - - - 7f9f2686 by John Ericson at 2023-10-12T10:31:13-04:00 Move visibility and "musttail" annotation checks to the RTS configure All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. Progress towards #17191 - - - - - ffb3efe6 by John Ericson at 2023-10-12T10:31:13-04:00 Move leading underscore checks to RTS configure `CabalLeadingUnderscore` is done via Hadrian already, so we can stop `AC_SUBST`ing it completely. - - - - - 25fa4b02 by John Ericson at 2023-10-12T10:31:13-04:00 Move alloca, fork, const, and big endian checks to RTS configure All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. - - - - - 5170f42a by John Ericson at 2023-10-12T10:31:13-04:00 Move libdl check to RTS configure - - - - - ea7a1447 by John Ericson at 2023-10-12T10:31:13-04:00 Adjust `FP_FIND_LIBFFI` Just set vars, and `AC_SUBST` in top-level configure. Don't define `HAVE_SYSTEM_LIBFFI` because nothing is using it. It hasn't be in used since 3609340743c1b25fdfd0e18b1670dac54c8d8623 (part of the make build system). - - - - - f399812c by John Ericson at 2023-10-12T10:31:13-04:00 Split BFD support to RTS configure The flag is still in the top-level configure, but the other checks (which define various macros --- important) are in the RTS configure. - - - - - f64f44e9 by John Ericson at 2023-10-12T10:31:13-04:00 Split libm check between top level and RTS - - - - - dafc4709 by Moritz Angermann at 2023-10-12T10:31:49-04:00 CgUtils.fixStgRegStmt respect register width This change ensure that the reg + offset computation is always of the same size. Before this we could end up with a 64bit register, and then add a 32bit offset (on 32bit platforms). This not only would fail type sanity checking, but also incorrectly truncate 64bit values into 32bit values silently on 32bit architectures. - - - - - 9e6ef7ba by Matthew Pickering at 2023-10-12T20:35:00-04:00 hadrian: Decrease verbosity of cabal commands In Normal, most tools do not produce output to stdout unless there are error conditions. Reverts 7ed65f5a1bc8e040e318ccff395f53a9bbfd8217 - - - - - 08fc27af by John Ericson at 2023-10-12T20:35:36-04:00 Do not substitute `@...@` for stage-specific values in cabal files `rts` and `ghc-prim` now no longer have a `*.cabal.in` to set Cabal flag defaults; instead manual choices are passed to configure in the usual way. The old way was fundamentally broken, because it meant we were baking these Cabal files for a specific stage. Now we only do stage-agnostic @...@ substitution in cabal files (the GHC version), and so all stage-specific configuration is properly confined to `_build` and the right stage dir. Also `include-ghc-prim` is a flag that no longer exists for `ghc-prim` (it was removed in 835d8ddbbfb11796ea8a03d1806b7cee38ba17a6) so I got rid of it. Co-Authored-By: Matthew Pickering <matthewtpickering at gmail.com> - - - - - a0ac8785 by Sebastian Graf at 2023-10-14T19:17:12-04:00 Fix restarts in .ghcid Using the whole of `hadrian/` restarted in a loop for me. - - - - - fea9ecdb by Sebastian Graf at 2023-10-14T19:17:12-04:00 CorePrep: Refactor FloatingBind (#23442) A drastically improved architecture for local floating in CorePrep that decouples the decision of whether a float is going to be let- or case-bound from how far it can float (out of strict contexts, out of lazy contexts, to top-level). There are a couple of new Notes describing the effort: * `Note [Floating in CorePrep]` for the overview * `Note [BindInfo and FloatInfo]` for the new classification of floats * `Note [Floats and FloatDecision]` for how FloatInfo is used to inform floating decisions This is necessary ground work for proper treatment of Strict fields and unlifted values at top-level. Fixes #23442. NoFib results (omitted = 0.0%): ``` -------------------------------------------------------------------------------- Program Allocs Instrs -------------------------------------------------------------------------------- pretty 0.0% -1.6% scc 0.0% -1.7% -------------------------------------------------------------------------------- Min 0.0% -1.7% Max 0.0% -0.0% Geometric Mean -0.0% -0.0% ``` - - - - - 32523713 by Matthew Pickering at 2023-10-14T19:17:49-04:00 hadrian: Move ghcBinDeps into ghcLibDeps This completes a5227080b57cb51ac34d4c9de1accdf6360b818b, the `ghc-usage.txt` and `ghci-usage.txt` file are also used by the `ghc` library so need to make sure they are present in the libdir even if we are not going to build `ghc-bin`. This also fixes things for cross compilers because the stage2 cross-compiler requires the ghc-usage.txt file, but we are using the stage2 lib folder but not building stage3:exe:ghc-bin so ghc-usage.txt was not being generated. - - - - - ec3c4488 by sheaf at 2023-10-14T19:18:29-04:00 Combine GREs when combining in mkImportOccEnv In `GHC.Rename.Names.mkImportOccEnv`, we sometimes discard one import item in favour of another, as explained in Note [Dealing with imports] in `GHC.Rename.Names`. However, this can cause us to lose track of important parent information. Consider for example #24084: module M1 where { class C a where { type T a } } module M2 ( module M1 ) where { import M1 } module M3 where { import M2 ( C, T ); instance C () where T () = () } When processing the import list of `M3`, we start off (for reasons that are not relevant right now) with two `Avail`s attached to `T`, namely `C(C, T)` and `T(T)`. We combine them in the `combine` function of `mkImportOccEnv`; as described in Note [Dealing with imports] we discard `C(C, T)` in favour of `T(T)`. However, in doing so, we **must not** discard the information want that `C` is the parent of `T`. Indeed, losing track of this information can cause errors when importing, as we could get an error of the form ‘T’ is not a (visible) associated type of class ‘C’ We fix this by combining the two GREs for `T` using `plusGRE`. Fixes #24084 - - - - - 257c2807 by Ilias Tsitsimpis at 2023-10-14T19:19:07-04:00 hadrian: Pass -DNOSMP to C compiler when needed Hadrian passes the -DNOSMP flag to GHC when the target doesn't support SMP, but doesn't pass it to CC as well, leading to the following compilation error on mips64el: | Run Cc (FindCDependencies CDep) Stage1: rts/sm/NonMovingScav.c => _build/stage1/rts/build/c/sm/NonMovingScav.o.d Command line: /usr/bin/mips64el-linux-gnuabi64-gcc -E -MM -MG -MF _build/stage1/rts/build/c/hooks/FlagDefaults.thr_debug_p_o.d -MT _build/stage1/rts/build/c/hooks/FlagDefaults.o -Irts/include -I_build/stage1/rts/build -I_build/stage1/rts/build/include -Irts/include -x c rts/hooks/FlagDefaults.c -Wall -Wextra -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Winline -Wpointer-arith -Wmissing-noreturn -Wnested-externs -Wredundant-decls -Wundef -fno-strict-aliasing -DTHREADED_RTS -DDEBUG -fomit-frame-pointer -O2 -g -Irts -I_build/stage1/rts/build -DDEBUG -fno-omit-frame-pointer -g3 -O0 ===> Command failed with error code: 1 In file included from rts/include/Stg.h:348, from rts/include/Rts.h:38, from rts/hooks/FlagDefaults.c:8: rts/include/stg/SMP.h:416:2: error: #error memory barriers unimplemented on this architecture 416 | #error memory barriers unimplemented on this architecture | ^~~~~ rts/include/stg/SMP.h:440:2: error: #error memory barriers unimplemented on this architecture 440 | #error memory barriers unimplemented on this architecture | ^~~~~ rts/include/stg/SMP.h:464:2: error: #error memory barriers unimplemented on this architecture 464 | #error memory barriers unimplemented on this architecture | ^~~~~ The old make system correctly passed this flag to both GHC and CC [1]. Fix this error by passing -DNOSMP to CC as well. [1] https://gitlab.haskell.org/ghc/ghc/-/blob/00920f176b0235d5bb52a8e054d89a664f8938fe/rts/ghc.mk#L407 Closes #24082 - - - - - 13d3c613 by John Ericson at 2023-10-14T19:19:42-04:00 Users Guide: Drop dead code for Haddock refs to `parallel` I noticed while working on !11451 that `@LIBRARY_parallel_UNIT_ID@` was not substituted. It is dead code -- there is no `parallel-ref` usages and it doesn't look like there ever was (going back to 3e5d0f188d6c8633e55e9ba6c8941c07e459fa4b), so let's delete it. - - - - - fe067577 by Sylvain Henry at 2023-10-18T19:40:25-04:00 Avoid out-of-bound array access in bigNatIsPowerOf2 (fix #24066) bigNatIndex# in the `where` clause wasn't guarded by "bigNatIsZero a". - - - - - cc1625b1 by Sylvain Henry at 2023-10-18T19:40:25-04:00 Bignum: fix right shift of negative BigNat with native backend - - - - - cbe4400d by Sylvain Henry at 2023-10-18T19:40:25-04:00 Rts: expose rtsOutOfBoundsAccess symbol - - - - - 72c7380c by Sylvain Henry at 2023-10-18T19:40:25-04:00 Hadrian: enable `-fcheck-prim-bounds` in validate flavour This allows T24066 to fail when the bug is present. Otherwise the out-of-bound access isn't detected as it happens in ghc-bignum which wasn't compiled with the bounds check. - - - - - f9436990 by John Ericson at 2023-10-18T19:41:01-04:00 Make Hadrian solely responsible for substituting `docs/users_guide/ghc_config.py.in` Fixes #24091 Progress on #23966 Issue #24091 reports that `@ProjectVersion@` is no longer being substituted in the GHC user's guide. I assume this is a recent issue, but I am not sure how it's worked since c1a3ecde720b3bddc2c8616daaa06ee324e602ab; it looks like both Hadrian and configure are trying to substitute the same `.in` file! Now only Hadrian does. That is better anyways; already something that issue #23966 requested. It seems like we were missing some dependencies in Hadrian. (I really, really hate that this is possible!) Hopefully it is fixed now. - - - - - b12df0bb by John Ericson at 2023-10-18T19:41:37-04:00 `ghcversion.h`: No need to cope with undefined `ProjectPatchLevel*` Since 4e6c80197f1cc46dfdef0300de46847c7cfbdcb0, these are guaranteed to be defined. (Guaranteed including a test in the testsuite.) - - - - - 0295375a by John Ericson at 2023-10-18T19:41:37-04:00 Generate `ghcversion.h` from a `.in` file Now that there are no conditional sections (see the previous commit), we can just a do simple substitution rather than pasting it together line by line. Progress on #23966 - - - - - 740a1b85 by Krzysztof Gogolewski at 2023-10-19T11:37:20-04:00 Add a regression test for #24064 - - - - - 921fbf2f by Hécate Moonlight at 2023-10-19T11:37:59-04:00 CLC Proposal #182: Export List from Data.List Proposal link: https://github.com/haskell/core-libraries-committee/issues/182 - - - - - 4f02d3c1 by Sylvain Henry at 2023-10-20T04:01:32-04:00 rts: fix small argument passing on big-endian arch (fix #23387) - - - - - b86243b4 by Sylvain Henry at 2023-10-20T04:02:13-04:00 Interpreter: fix literal alignment on big-endian architectures (fix #19261) Literals weren't correctly aligned on big-endian, despite what the comment said. - - - - - a4b2ec47 by Sylvain Henry at 2023-10-20T04:02:54-04:00 Testsuite: recomp011 and recomp015 are fixed on powerpc These tests have been fixed but not tested and re-enabled on big-endian powerpc (see comments in #11260 and #11323) - - - - - fded7dd4 by Sebastian Graf at 2023-10-20T04:03:30-04:00 CorePrep: Allow floating dictionary applications in -O0 into a Rec (#24102) - - - - - 02efc181 by John Ericson at 2023-10-22T02:48:55-04:00 Move function checks to RTS configure Some of these functions are used in `base` too, but we can copy the checks over to its configure if that's an issue. - - - - - 5f4bccab by John Ericson at 2023-10-22T02:48:55-04:00 Move over a number of C-style checks to RTS configure - - - - - 5cf04f58 by John Ericson at 2023-10-22T02:48:55-04:00 Move/Copy more `AC_DEFINE` to RTS config Only exception is the LLVM version macros, which are used for GHC itself. - - - - - b8ce5dfe by John Ericson at 2023-10-22T02:48:55-04:00 Define `TABLES_NEXT_TO_CODE` in the RTS configure We create a new cabal flag to facilitate this. - - - - - 4a40271e by John Ericson at 2023-10-22T02:48:55-04:00 Configure scripts: `checkOS`: Make a bit more robust `mingw64` and `mingw32` are now both accepted for `OSMinGW32`. This allows us to cope with configs/triples that we haven't normalized extra being what GNU `config.sub` does. - - - - - 16bec0a0 by John Ericson at 2023-10-22T02:48:55-04:00 Generate `ghcplatform.h` from RTS configure We create a new cabal flag to facilitate this. - - - - - 7dfcab2f by John Ericson at 2023-10-22T02:48:55-04:00 Get rid of all mention of `mk/config.h` The RTS configure script is now solely responsible for managing its headers; the top level configure script does not help. - - - - - c1e3719c by Cheng Shao at 2023-10-22T02:49:33-04:00 rts: drop stale mentions of MIN_UPD_SIZE We used to have MIN_UPD_SIZE macro that describes the minimum reserved size for thunks, so that the thunk can be overwritten in place as indirections or blackholes. However, this macro has not been actually defined or used anywhere since a long time ago; StgThunkHeader already reserves a padding word for this purpose. Hence this patch which drops stale mentions of MIN_UPD_SIZE. - - - - - d24b0d85 by Andrew Lelechenko at 2023-10-22T02:50:11-04:00 base changelog: move non-backported entries from 4.19 section to 4.20 Neither !10933 (check https://hackage.haskell.org/package/base-4.19.0.0/docs/src/Text.Read.Lex.html#numberToRangedRational) nor !10189 (check https://hackage.haskell.org/package/base-4.19.0.0/docs/src/Data.List.NonEmpty.html#unzip) were backported to `base-4.19.0.0`. Moving them to `base-4.20.0.0` section. Also minor stylistic changes to other entries, bringing them to a uniform form. - - - - - de78b32a by Alan Zimmerman at 2023-10-23T09:09:41-04:00 EPA Some tweaks to annotations - Fix span for GRHS - Move TrailingAnns from last match to FunBind - Fix GADT 'where' clause span - Capture full range for a CaseAlt Match - - - - - d5a8780d by Simon Hengel at 2023-10-23T09:10:23-04:00 Update primitives.rst - - - - - 4d075924 by Josh Meredith at 2023-10-24T23:04:12+11:00 JS/userguide: add explanation of writing jsbits - - - - - 07ab5cc1 by Cheng Shao at 2023-10-24T15:40:32-04:00 testsuite: increase timeout of ghc-api tests for wasm32 ghc-api tests for wasm32 are more likely to timeout due to the large wasm module sizes, especially when testing with wasm native tail calls, given wasmtime's handling of tail call opcodes are suboptimal at the moment. It makes sense to increase timeout specifically for these tests on wasm32. This doesn't affect other targets, and for wasm32 we don't increase timeout for all tests, so not to risk letting major performance regressions slip through the testsuite. - - - - - 0d6acca5 by Greg Steuck at 2023-10-26T08:44:23-04:00 Explicitly require RLIMIT_AS before use in OSMem.c This is done elsewhere in the source tree. It also suddenly is required on OpenBSD. - - - - - 9408b086 by Sylvain Henry at 2023-10-26T08:45:03-04:00 Modularity: modularize external linker Decouple runLink from DynFlags to allow calling runLink more easily. This is preliminary work for calling Emscripten's linker (emcc) from our JavaScript linker. - - - - - e0f35030 by doyougnu at 2023-10-27T08:41:12-04:00 js: add JStg IR, remove unsaturated constructor - Major step towards #22736 and adding the optimizer in #22261 - - - - - 35587eba by Simon Peyton Jones at 2023-10-27T08:41:48-04:00 Fix a bug in tail calls with ticks See #24078 for the diagnosis. The change affects only the Tick case of occurrence analysis. It's a bit hard to test, so no regression test (yet anyway). - - - - - 9bc5cb92 by Matthew Craven at 2023-10-28T07:06:17-04:00 Teach tag-inference about SeqOp/seq# Fixes the STG/tag-inference analogue of #15226. Co-Authored-By: Simon Peyton Jones <simon.peytonjones at gmail.com> - - - - - 34f06334 by Moritz Angermann at 2023-10-28T07:06:53-04:00 [PEi386] Mask SYM_TYPE_DUP_DISCARD in makeSymbolExtra 48e391952c17ff7eab10b0b1456e3f2a2af28a9b introduced `SYM_TYPE_DUP_DISCARD` to the bitfield. The linker however, failed to mask the `SYM_TYPE_DUP_DISCARD` value. Thus `== SYM_TYPE_CODE` comparisons easily failed. This lead to us relocating DATA lookups (GOT) into E8 (call) and E9 (jump) instructions. - - - - - 5b51b2a2 by Mario Blažević at 2023-10-28T07:07:33-04:00 Fix and test for issue #24111, TH.Ppr output of pattern synonyms - - - - - 723bc352 by Alan Zimmerman at 2023-10-30T20:36:41-04:00 EPA: print doc comments as normal comments And ignore the ones allocated in haddock processing. It does not guarantee that every original haddock-like comment appears in the output, as it discards ones that have no legal attachment point. closes #23459 - - - - - 21b76843 by Simon Peyton Jones at 2023-10-30T20:37:17-04:00 Fix non-termination bug in equality solver constraint left-to-right then right to left, forever. Easily fixed. - - - - - 270867ac by Sebastian Graf at 2023-10-30T20:37:52-04:00 ghc-toolchain: build with `-package-env=-` (#24131) Otherwise globally installed libraries (via `cabal install --lib`) break the build. Fixes #24131. - - - - - 4d08364e by Alan Zimmerman at 2023-10-31T19:46:45+00:00 EPA: match changes in GHC - EPA: Comments in AnchorOperation - EPA: Remove EpaEofComment - - - - - 7a90020f by Krzysztof Gogolewski at 2023-10-31T20:03:37-04:00 docs: fix ScopedTypeVariables example (#24101) The previous example didn't compile. Furthermore, it wasn't demonstrating the point properly. I have changed it to an example which shows that 'a' in the signature must be the same 'a' as in the instance head. - - - - - 49f69f50 by Krzysztof Gogolewski at 2023-10-31T20:04:13-04:00 Fix pretty-printing of type family dependencies "where" should be after the injectivity annotation. - - - - - 73c191c0 by Ben Gamari at 2023-10-31T20:04:49-04:00 gitlab-ci: Bump LLVM bootstrap jobs to Debian 12 As the Debian 10 images have too old an LLVM. Addresses #24056. - - - - - 5b0392e0 by Matthew Pickering at 2023-10-31T20:04:49-04:00 ci: Run aarch64 llvm backend job with "LLVM backend" label This brings it into line with the x86 LLVM backend job. - - - - - 9f9c9227 by Ryan Scott at 2023-11-01T09:19:12-04:00 More robust checking for DataKinds As observed in #22141, GHC was not doing its due diligence in catching code that should require `DataKinds` in order to use. Most notably, it was allowing the use of arbitrary data types in kind contexts without `DataKinds`, e.g., ```hs data Vector :: Nat -> Type -> Type where ``` This patch revamps how GHC tracks `DataKinds`. The full specification is written out in the `DataKinds` section of the GHC User's Guide, and the implementation thereof is described in `Note [Checking for DataKinds]` in `GHC.Tc.Validity`. In brief: * We catch _type_-level `DataKinds` violations in the renamer. See `checkDataKinds` in `GHC.Rename.HsType` and `check_data_kinds` in `GHC.Rename.Pat`. * We catch _kind_-level `DataKinds` violations in the typechecker, as this allows us to catch things that appear beneath type synonyms. (We do *not* want to do this in type-level contexts, as it is perfectly fine for a type synonym to mention something that requires DataKinds while still using the type synonym in a module that doesn't enable DataKinds.) See `checkValidType` in `GHC.Tc.Validity`. * There is now a single `TcRnDataKindsError` that classifies all manner of `DataKinds` violations, both in the renamer and the typechecker. The `NoDataKindsDC` error has been removed, as it has been subsumed by `TcRnDataKindsError`. * I have added `CONSTRAINT` is `isKindTyCon`, which is what checks for illicit uses of data types at the kind level without `DataKinds`. Previously, `isKindTyCon` checked for `Constraint` but not `CONSTRAINT`. This is inconsistent, given that both `Type` and `TYPE` were checked by `isKindTyCon`. Moreover, it thwarted the implementation of the `DataKinds` check in `checkValidType`, since we would expand `Constraint` (which was OK without `DataKinds`) to `CONSTRAINT` (which was _not_ OK without `DataKinds`) and reject it. Now both are allowed. * I have added a flurry of additional test cases that test various corners of `DataKinds` checking. Fixes #22141. - - - - - 575d7690 by Sylvain Henry at 2023-11-01T09:19:53-04:00 JS: fix FFI "wrapper" and "dynamic" Fix codegen and helper functions for "wrapper" and "dynamic" foreign imports. Fix tests: - ffi006 - ffi011 - T2469 - T4038 Related to #22363 - - - - - 81fb8885 by Alan Zimmerman at 2023-11-01T22:23:56-04:00 EPA: Use full range for Anchor This change requires a series of related changes, which must all land at the same time, otherwise all the EPA tests break. * Use the current Anchor end as prior end Use the original anchor location end as the source of truth for calculating print deltas. This allows original spacing to apply in most cases, only changed AST items need initial delta positions. * Add DArrow to TrailingAnn * EPA Introduce HasTrailing in ExactPrint Use [TrailingAnn] in enterAnn and remove it from ExactPrint (LocatedN RdrName) * In HsDo, put TrailingAnns at top of LastStmt * EPA: do not convert comments to deltas when balancing. * EPA: deal with fallout from getMonoBind * EPA fix captureLineSpacing * EPA print any comments in the span before exiting it * EPA: Add comments to AnchorOperation * EPA: remove AnnEofComment, it is no longer used Updates Haddock submodule - - - - - 03e82511 by Rodrigo Mesquita at 2023-11-01T22:24:32-04:00 Fix in docs regarding SSymbol, SNat, SChar (#24119) - - - - - 362cc693 by Matthew Pickering at 2023-11-01T22:25:08-04:00 hadrian: Update bootstrap plans (9.4.6, 9.4.7, 9.6.2, 9.6.3, 9.8.1) Updating the bootstrap plans with more recent GHC versions. - - - - - 00b9b8d3 by Matthew Pickering at 2023-11-01T22:25:08-04:00 ci: Add 9.8.1 bootstrap testing job - - - - - ef3d20f8 by Matthew Pickering at 2023-11-01T22:25:08-04:00 Compatibility with 9.8.1 as boot compiler This fixes several compatability issues when using 9.8.1 as the boot compiler. * An incorrect version guard on the stack decoding logic in ghc-heap * Some ghc-prim bounds need relaxing * ghc is no longer wired in, so we have to remove the -this-unit-id ghc call. Fixes #24077 - - - - - 6755d833 by Jaro Reinders at 2023-11-03T10:54:42+01:00 Add NCG support for common 64bit operations to the x86 backend. These used to be implemented via C calls which was obviously quite bad for performance for operations like simple addition. Co-authored-by: Andreas Klebinger - - - - - 0dfb1fa7 by Vladislav Zavialov at 2023-11-03T14:08:41-04:00 T2T in Expressions (#23738) This patch implements the T2T (term-to-type) transformation in expressions. Given a function with a required type argument vfun :: forall a -> ... the user can now call it as vfun (Maybe Int) instead of vfun (type (Maybe Int)) The Maybe Int argument is parsed and renamed as a term (HsExpr), but then undergoes a conversion to a type (HsType). See the new function expr_to_type in compiler/GHC/Tc/Gen/App.hs and Note [RequiredTypeArguments and the T2T mapping] Left as future work: checking for puns. - - - - - cc1c7c54 by Duncan Coutts at 2023-11-05T00:23:44-04:00 Add a test for I/O managers It tries to cover the cases of multiple threads waiting on the same fd for reading and multiple threads waiting for writing, including wait cancellation by async exceptions. It should work for any I/O manager, in-RTS or in-Haskell. Unfortunately it will not currently work for Windows because it relies on anonymous unix sockets. It could in principle be ported to use Windows named pipes. - - - - - 2e448f98 by Cheng Shao at 2023-11-05T00:23:44-04:00 Skip the IOManager test on wasm32 arch. The test relies on the sockets API which are not (yet) available. - - - - - fe50eb35 by Cheng Shao at 2023-11-05T00:24:20-04:00 compiler: fix eager blackhole symbol in wasm32 NCG - - - - - af771148 by Cheng Shao at 2023-11-05T00:24:20-04:00 testsuite: fix optasm tests for wasm32 - - - - - 1b90735c by Matthew Pickering at 2023-11-05T00:24:20-04:00 testsuite: Add wasm32 to testsuite arches with NCG The compiler --info reports that wasm32 compilers have a NCG, so we should agree with that here. - - - - - db9a6496 by Alan Zimmerman at 2023-11-05T00:24:55-04:00 EPA: make locA a function, not a field name And use it to generalise reLoc The following for the windows pipeline one. 5.5% Metric Increase: T5205 - - - - - 833e250c by Simon Peyton Jones at 2023-11-05T00:25:31-04:00 Update the unification count in wrapUnifierX Omitting this caused type inference to fail in #24146. This was an accidental omision in my refactoring of the equality solver. - - - - - e451139f by Andreas Klebinger at 2023-11-05T00:26:07-04:00 Remove an accidental git conflict marker from a comment. - - - - - e7da0d25 by Alan Zimmerman at 2023-11-05T11:20:31+00:00 EPA: match changes in GHC, l2l cleanup - - - - - 30baac7a by Tobias Haslop at 2023-11-06T10:50:32+00:00 Add laws relating between Foldable/Traversable with their Bi- superclasses See https://github.com/haskell/core-libraries-committee/issues/205 for discussion. This commit also documents that the tuple instances only satisfy the laws up to lazyness, similar to the documentation added in !9512. - - - - - df626f00 by Tobias Haslop at 2023-11-07T02:20:37-05:00 Elaborate on the quantified superclass of Bifunctor This was requested in the comment https://github.com/haskell/core-libraries-committee/issues/93#issuecomment-1597271700 for when Traversable becomes a superclass of Bitraversable, but similarly applies to Functor/Bifunctor, which already are in a superclass relationship. - - - - - 8217acb8 by Alan Zimmerman at 2023-11-07T02:21:12-05:00 EPA: get rid of l2l and friends Replace them with l2l to convert the location la2la to convert a GenLocated thing Updates haddock submodule - - - - - dd88a260 by Luite Stegeman at 2023-11-07T02:21:53-05:00 JS: remove broken newIdents from JStg Monad GHC.JS.JStg.Monad.newIdents was broken, resulting in duplicate identifiers being generated in h$c1, h$c2, ... . This change removes the broken newIdents. - - - - - 455524a2 by Matthew Craven at 2023-11-09T08:41:59-05:00 Create specially-solved DataToTag class Closes #20532. This implements CLC proposal 104: https://github.com/haskell/core-libraries-committee/issues/104 The design is explained in Note [DataToTag overview] in GHC.Tc.Instance.Class. This replaces the existing `dataToTag#` primop. These metric changes are not "real"; they represent Unique-related flukes triggering on a different set of jobs than they did previously. See also #19414. Metric Decrease: T13386 T8095 Metric Increase: T13386 T8095 Co-authored-by: Simon Peyton Jones <simon.peytonjones at gmail.com> - - - - - a05f4554 by Alan Zimmerman at 2023-11-09T08:42:35-05:00 EPA: get rid of glRR and friends in GHC/Parser.y With the HasLoc and HasAnnotation classes, we can replace a number of type-specific helper functions in the parser with polymorphic ones instead Metric Decrease: MultiLayerModulesTH_Make - - - - - 18498538 by Cheng Shao at 2023-11-09T16:58:12+00:00 ci: bump ci-images for wasi-sdk upgrade - - - - - 52c0fc69 by PHO at 2023-11-09T19:16:22-05:00 Don't assume the current locale is *.UTF-8, set the encoding explicitly primops.txt contains Unicode characters: > LC_ALL=C ./genprimopcode --data-decl < ./primops.txt > genprimopcode: <stdin>: hGetContents: invalid argument (cannot decode byte sequence starting from 226) Hadrian must also avoid using readFile' to read primops.txt because it tries to decode the file with a locale-specific encoding. - - - - - 7233b3b1 by PHO at 2023-11-09T19:17:01-05:00 Use '[' instead of '[[' because the latter is a Bash-ism It doesn't work on platforms where /bin/sh is something other than Bash. - - - - - 6dbab180 by Simon Peyton Jones at 2023-11-09T19:17:36-05:00 Add an extra check in kcCheckDeclHeader_sig Fix #24083 by checking for a implicitly-scoped type variable that is not actually bound. See Note [Disconnected type variables] in GHC.Tc.Gen.HsType For some reason, on aarch64-darwin we saw a 2.8% decrease in compiler allocations for MultiLayerModulesTH_Make; but 0.0% on other architectures. Metric Decrease: MultiLayerModulesTH_Make - - - - - 22551364 by Sven Tennie at 2023-11-11T06:35:22-05:00 AArch64: Delete unused LDATA pseudo-instruction Though there were consuming functions for LDATA, there were no producers. Thus, the removed code was "dead". - - - - - 2a0ec8eb by Alan Zimmerman at 2023-11-11T06:35:59-05:00 EPA: harmonise acsa and acsA in GHC/Parser.y With the HasLoc class, we can remove the acsa helper function, using acsA instead. - - - - - 4ceac14d by Alan Zimmerman at 2023-11-11T15:16:41+00:00 EPA: Replace Anchor with EpaLocation Match GHC - - - - - 7ae517a0 by Teo Camarasu at 2023-11-12T08:04:12-05:00 nofib: bump submodule This includes changes that: - fix building a benchmark with HEAD - remove a Makefile-ism that causes errors in bash scripts Resolves #24178 - - - - - 3f0036ec by Alan Zimmerman at 2023-11-12T08:04:47-05:00 EPA: Replace Anchor with EpaLocation An Anchor has a location and an operation, which is either that it is unchanged or that it has moved with a DeltaPos data Anchor = Anchor { anchor :: RealSrcSpan , anchor_op :: AnchorOperation } An EpaLocation also has either a location or a DeltaPos data EpaLocation = EpaSpan !RealSrcSpan !(Strict.Maybe BufSpan) | EpaDelta !DeltaPos ![LEpaComment] Now that we do not care about always having a location in the anchor, we remove Anchor and replace it with EpaLocation We do this with a type alias initially, to ease the transition. The alias will be removed in time. We also have helpers to reconstruct the AnchorOperation from an EpaLocation. This is also temporary. Updates Haddock submodule - - - - - a7492048 by Alan Zimmerman at 2023-11-12T13:43:07+00:00 EPA: get rid of AnchorOperation Now that the Anchor type is an alias for EpaLocation, remove AnchorOperation. Updates haddock submodule - - - - - 0745c34d by Andrew Lelechenko at 2023-11-13T16:25:07-05:00 Add since annotation for showHFloat - - - - - e98051a5 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 Suppress duplicate librares linker warning of new macOS linker Fixes #24167 XCode 15 introduced a new linker which warns on duplicate libraries being linked. To disable this warning, we pass -Wl,-no_warn_duplicate_libraries as suggested by Brad King in CMake issue #25297. This flag isn't necessarily available to other linkers on darwin, so we must only configure it into the CC linker arguments if valid. - - - - - c411c431 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 testsuite: Encoding test witnesses recent iconv bug is fragile A regression in the new iconv() distributed with XCode 15 and MacOS Sonoma causes the test 'encoding004' to fail in the CP936 roundrip. We mark this test as fragile until this is fixed upstream (rather than broken, since previous versions of iconv pass the test) See #24161 - - - - - ce7fe5a9 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 testsuite: Update to LC_ALL=C no longer being ignored in darwin MacOS seems to have fixed an issue where it used to ignore the variable `LC_ALL` in program invocations and default to using Unicode. Since the behaviour seems to be fixed to account for the locale variable, we mark tests that were previously broken in spite of it as fragile (since they now pass in recent macOS distributions) See #24161 - - - - - e6c803f7 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 darwin: Fix single_module is obsolete warning In XCode 15's linker, -single_module is the default and otherwise passing it as a flag results in a warning being raised: ld: warning: -single_module is obsolete This patch fixes this warning by, at configure time, determining whether the linker supports -single_module (which is likely false for all non-darwin linkers, and true for darwin linkers in previous versions of macOS), and using that information at runtime to decide to pass or not the flag in the invocation. Fixes #24168 - - - - - 929ba2f9 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 testsuite: Skip MultiLayerModulesTH_Make on darwin The recent toolchain upgrade on darwin machines resulted in the MultiLayerModulesTH_Make test metrics varying too much from the baseline, ultimately blocking the CI pipelines. This commit skips the test on darwin to temporarily avoid failures due to the environment change in the runners. However, the metrics divergence is being investigated still (tracked in #24177) - - - - - af261ccd by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 configure: check target (not build) understands -no_compact_unwind Previously, we were branching on whether the build system was darwin to shortcut this check, but we really want to branch on whether the target system (which is what we are configuring ld_prog for) is darwin. - - - - - 2125c176 by Luite Stegeman at 2023-11-15T13:19:38-05:00 JS: Fix missing variable declarations The JStg IR update was missing some local variable declarations that were present earlier, causing global variables to be used implicitly (or an error in JavaScript strict mode). This adds the local variable declarations again. - - - - - 99ced73b by Krzysztof Gogolewski at 2023-11-15T13:20:14-05:00 Remove loopy superclass solve mechanism Programs with a -Wloopy-superclass-solve warning will now fail with an error. Fixes #23017 - - - - - 2aff2361 by Zubin Duggal at 2023-11-15T13:20:50-05:00 users-guide: Fix links to libraries from the users-guide. The unit-ids generated in c1a3ecde720b3bddc2c8616daaa06ee324e602ab include the package name, so we don't need to explicitly add it to the links. Fixes #24151 - - - - - 27981fac by Alan Zimmerman at 2023-11-15T13:21:25-05:00 EPA: splitLHsForAllTyInvis does not return ann We did not use the annotations returned from splitLHsForAllTyInvis, so do not return them. - - - - - a6467834 by Krzysztof Gogolewski at 2023-11-15T22:22:59-05:00 Document defaulting of RuntimeReps Fixes #24099 - - - - - 2776920e by Simon Peyton Jones at 2023-11-15T22:23:35-05:00 Second fix to #24083 My earlier fix turns out to be too aggressive for data/type families See wrinkle (DTV1) in Note [Disconnected type variables] - - - - - cee81370 by Sylvain Henry at 2023-11-16T09:57:46-05:00 Fix unusable units and module reexport interaction (#21097) This commit fixes an issue with ModUnusable introduced in df0f148feae. In mkUnusableModuleNameProvidersMap we traverse the list of unusable units and generate ModUnusable origin for all the modules they contain: exposed modules, hidden modules, and also re-exported modules. To do this we have a two-level map: ModuleName -> Unit:ModuleName (aka Module) -> ModuleOrigin So for each module name "M" in broken unit "u" we have: "M" -> u:M -> ModUnusable reason However in the case of module reexports we were using the *target* module as a key. E.g. if "u:M" is a reexport for "X" from unit "o": "M" -> o:X -> ModUnusable reason Case 1: suppose a reexport without module renaming (u:M -> o:M) from unusable unit u: "M" -> o:M -> ModUnusable reason Here it's claiming that the import of M is unusable because a reexport from u is unusable. But if unit o isn't unusable we could also have in the map: "M" -> o:M -> ModOrigin ... Issue: the Semigroup instance of ModuleOrigin doesn't handle the case (ModUnusable <> ModOrigin) Case 2: similarly we could have 2 unusable units reexporting the same module without renaming, say (u:M -> o:M) and (v:M -> o:M) with u and v unusable. It gives: "M" -> o:M -> ModUnusable ... (for u) "M" -> o:M -> ModUnusable ... (for v) Issue: the Semigroup instance of ModuleOrigin doesn't handle the case (ModUnusable <> ModUnusable). This led to #21097, #16996, #11050. To fix this, in this commit we make ModUnusable track whether the module used as key is a reexport or not (for better error messages) and we use the re-export module as key. E.g. if "u:M" is a reexport for "o:X" and u is unusable, we now record: "M" -> u:M -> ModUnusable reason reexported=True So now, we have two cases for a reexport u:M -> o:X: - u unusable: "M" -> u:M -> ModUnusable ... reexported=True - u usable: "M" -> o:X -> ModOrigin ... reexportedFrom=u:M The second case is indexed with o:X because in this case the Semigroup instance of ModOrigin is used to combine valid expositions of a module (directly or via reexports). Note that module lookup functions select usable modules first (those who have a ModOrigin value), so it doesn't matter if we add new ModUnusable entries in the map like this: "M" -> { u:M -> ModUnusable ... reexported=True o:M -> ModOrigin ... } The ModOrigin one will be used. Only if there is no ModOrigin or ModHidden entry will the ModUnusable error be printed. See T21097 for an example printing several reasons why an import is unusable. - - - - - 3e606230 by Krzysztof Gogolewski at 2023-11-16T09:58:22-05:00 Fix IPE test A helper function was defined in a different module than used. To reproduce: ./hadrian/build test --test-root-dirs=testsuite/tests/rts/ipe - - - - - 49f5264b by Andreas Klebinger at 2023-11-16T20:52:11-05:00 Properly compute unpacked sizes for -funpack-small-strict-fields. Use rep size rather than rep count to compute the size. Fixes #22309 - - - - - b4f84e4b by James Henri Haydon at 2023-11-16T20:52:53-05:00 Explicit methods for Alternative Compose Explicitly define some and many in Alternative instance for Data.Functor.Compose Implementation of https://github.com/haskell/core-libraries-committee/issues/181 - - - - - 9bc0dd1f by Ignat Insarov at 2023-11-16T20:53:34-05:00 Add permutations for non-empty lists. Approved by CLC in https://github.com/haskell/core-libraries-committee/issues/68#issuecomment-1221409837 - - - - - 5643ecf9 by Andrew Lelechenko at 2023-11-16T20:53:34-05:00 Update changelog and since annotations for Data.List.NonEmpty.permutations Approved by CLC in https://github.com/haskell/core-libraries-committee/issues/68#issuecomment-1221409837 - - - - - 94ff2134 by Oleg Alexander at 2023-11-16T20:54:15-05:00 Update doc string for traceShow Updated doc string for traceShow. - - - - - faff671a by Luite Stegeman at 2023-11-17T14:12:51+01:00 JS: clean up some foreign imports - - - - - 856e0a4e by Sven Tennie at 2023-11-18T06:54:11-05:00 AArch64: Remove unused instructions As these aren't ever emitted, we don't even know if they work or will ever be used. If one of them is needed in future, we may easily re-add it. Deleted instructions are: - CMN - ANDS - BIC - BICS - EON - ORN - ROR - TST - STP - LDP - DMBSY - - - - - 615441ef by Alan Zimmerman at 2023-11-18T06:54:46-05:00 EPA: Replace Monoid with NoAnn Remove the final Monoid instances in the exact print infrastructure. For Windows CI Metric Decrease: T5205 - - - - - 5a6c49d4 by David Feuer at 2023-11-20T18:53:18-05:00 Speed up stimes in instance Semigroup Endo As discussed at https://github.com/haskell/core-libraries-committee/issues/4 - - - - - cf9da4b3 by Andrew Lelechenko at 2023-11-20T18:53:18-05:00 base: reflect latest changes in the changelog - - - - - 48bf364e by Alan Zimmerman at 2023-11-20T18:53:54-05:00 EPA: Use SrcSpan in EpaSpan This is more natural, since we already need to deal with invalid RealSrcSpans, and that is exactly what SrcSpan.UnhelpfulSpan is for. Updates haddock submodule. - - - - - 97ec37cc by Sebastian Graf at 2023-11-20T18:54:31-05:00 Add regression test for #6070 Fixes #6070. - - - - - e9d5ae41 by Owen Shepherd at 2023-11-21T18:32:23-05:00 chore: Correct typo in the gitlab MR template [skip ci] - - - - - f158a8d0 by Rodrigo Mesquita at 2023-11-21T18:32:59-05:00 Improve error message when reading invalid `.target` files A `.target` file generated by ghc-toolchain or by configure can become invalid if the target representation (`Toolchain.Target`) is changed while the files are not re-generated by calling `./configure` or `ghc-toolchain` again. There is also the issue of hadrian caching the dependencies on `.target` files, which makes parsing fail when reading reading the cached value if the representation has been updated. This patch provides a better error message in both situations, moving away from a terrible `Prelude.read: no parse` error that you would get otherwise. Fixes #24199 - - - - - 955520c6 by Ben Gamari at 2023-11-21T18:33:34-05:00 users guide: Note that QuantifiedConstraints implies ExplicitForAll Fixes #24025. - - - - - 17ec3e97 by Owen Shepherd at 2023-11-22T09:37:28+01:00 fix: Change type signatures in NonEmpty export comments to reflect reality This fixes several typos in the comments of Data.List.NonEmpty export list items. - - - - - 2fd78f9f by Samuel Thibault at 2023-11-22T11:49:13-05:00 Fix the platform string for GNU/Hurd As commited in Cargo https://github.com/haskell/cabal/pull/9434 there is confusion between "gnu" and "hurd". This got fixed in Cargo, we need the converse in Hadrian. Fixes #24180 - - - - - a79960fe by Alan Zimmerman at 2023-11-22T11:49:48-05:00 EPA: Tuple Present no longer has annotation The Present constructor for a Tuple argument will never have an exact print annotation. So make this impossible. - - - - - 121c9ab7 by David Binder at 2023-11-22T21:12:29-05:00 Unify the hpc testsuites The hpc testsuite was split between testsuite/tests/hpc and the submodule libraries/hpc/test. This commit unifies the two testsuites in the GHC repository in the directory testsuite/tests/hpc. - - - - - d2733a05 by Alan Zimmerman at 2023-11-22T21:13:05-05:00 EPA: empty tup_tail has noAnn In Parser.y, the tup_tail rule had the following option | {- empty -} %shift { return [Left noAnn] } Once this works through PostProcess.hs, it means we add an extra Missing constructor if the last item was a comma. Change the annotation type to a Bool to indicate this, and use the EpAnn Anchor for the print location for the others. - - - - - fa576eb8 by Andreas Klebinger at 2023-11-24T08:29:13-05:00 Fix FMA primops generating broken assembly on x86. `genFMA3Code` assumed that we had to take extra precations to avoid overwriting the result of `getNonClobberedReg`. One of these special cases caused a bug resulting in broken assembly. I believe we don't need to hadle these cases specially at all, which means this MR simply deletes the special cases to fix the bug. Fixes #24160 - - - - - 34d86315 by Alan Zimmerman at 2023-11-24T08:29:49-05:00 EPA: Remove parenthesizeHsType This is called from PostProcess.hs, and adds spurious parens. With the looser version of exact printing we had before we could tolerate this, as they would be swallowed by the original at the same place. But with the next change (remove EpAnnNotUsed) they result in duplicates in the output. For Darwin build: Metric Increase: MultiLayerModulesTH_OneShot - - - - - 3ede659d by Vladislav Zavialov at 2023-11-26T06:43:32-05:00 Add name for -Wdeprecated-type-abstractions (#24154) This warning had no name or flag and was triggered unconditionally. Now it is part of -Wcompat. - - - - - 7902ebf8 by Alan Zimmerman at 2023-11-26T06:44:08-05:00 EPA: Remove EpAnnNotUsed We no longer need the EpAnnNotUsed constructor for EpAnn, as we can represent an unused annotation with an anchor having a EpaDelta of zero, and empty comments and annotations. This simplifies code handling annotations considerably. Updates haddock submodule Metric Increase: parsing001 - - - - - 471b2672 by Mario Blažević at 2023-11-26T06:44:48-05:00 Bumped the upper bound of text to <2.2 - - - - - d1bf25c7 by Vladislav Zavialov at 2023-11-26T11:45:49-05:00 Term variable capture (#23740) This patch changes type variable lookup rules (lookupTypeOccRn) and implicit quantification rules (filterInScope) so that variables bound in the term namespace can be captured at the type level {-# LANGUAGE RequiredTypeArguments #-} f1 x = g1 @x -- `x` used in a type application f2 x = g2 (undefined :: x) -- `x` used in a type annotation f3 x = g3 (type x) -- `x` used in an embedded type f4 x = ... where g4 :: x -> x -- `x` used in a type signature g4 = ... This change alone does not allow us to accept examples shown above, but at least it gets them past the renamer. - - - - - da863d15 by Vladislav Zavialov at 2023-11-26T11:46:26-05:00 Update Note [hsScopedTvs and visible foralls] The Note was written before GHC gained support for visible forall in types of terms. Rewrite a few sentences and use a better example. - - - - - b5213542 by Matthew Pickering at 2023-11-27T12:53:59-05:00 testsuite: Add mechanism to collect generic metrics * Generalise the metric logic by adding an additional field which allows you to specify how to query for the actual value. Previously the method of querying the baseline value was abstracted (but always set to the same thing). * This requires rejigging how the stat collection works slightly but now it's more uniform and hopefully simpler. * Introduce some new "generic" helper functions for writing generic stats tests. - collect_size ( deviation, path ) Record the size of the file as a metric - stat_from_file ( metric, deviation, path ) Read a value from the given path, and store that as a metric - collect_generic_stat ( metric, deviation, get_stat) Provide your own `get_stat` function, `lambda way: <Int>`, which can be used to establish the current value of the metric. - collect_generic_stats ( metric_info ): Like collect_generic_stat but provide the whole dictionary of metric definitions. { metric: { deviation: <Int> current: lambda way: <Int> } } * Introduce two new "size" metrics for keeping track of build products. - `size_hello_obj` - The size of `hello.o` from compiling hello.hs - `libdir` - The total size of the `libdir` folder. * Track the number of modules in the AST tests - CountDepsAst - CountDepsParser This lays the infrastructure for #24191 #22256 #17129 - - - - - 7d9a2e44 by ARATA Mizuki at 2023-11-27T12:54:39-05:00 x86: Don't require -mavx2 when using 256-bit floating-point SIMD primitives Fixes #24222 - - - - - 4e5ff6a4 by Alan Zimmerman at 2023-11-27T12:55:15-05:00 EPA: Remove SrcSpanAnn Now that we only have a single constructor for EpAnn, And it uses a SrcSpan for its location, we can do away with SrcSpanAnn completely. It only existed to wrap the original SrcSpan in a location, and provide a place for the exact print annotation. For darwin only: Metric Increase: MultiLayerModulesTH_OneShot Updates haddock submodule - - - - - e05bca39 by Krzysztof Gogolewski at 2023-11-28T08:00:55-05:00 testsuite: don't initialize testdir to '.' The test directory is removed during cleanup, if there's an interrupt that could remove the entire repository. Fixes #24219 - - - - - af881674 by Alan Zimmerman at 2023-11-28T08:01:30-05:00 EPA: Clean up mkScope in Ast.hs Now that we have HasLoc we can get rid of all the custom variants of mkScope For deb10-numa Metric Increase: libdir - - - - - 292983c8 by Ben Gamari at 2023-11-28T22:44:28-05:00 distrib: Rediscover otool and install_name_tool on Darwin In the bindist configure script we must rediscover the `otool` and `install_name_tool`s since they may be different from the build environment. Fixes #24211. - - - - - dfe1c354 by Stefan Schulze Frielinghaus at 2023-11-28T22:45:04-05:00 llvmGen: Align objects in the data section Objects in the data section may be referenced via tagged pointers. Thus, align those objects to a 4- or 8-byte boundary for 32- or 64-bit platforms, respectively. Note, this may need to be reconsidered if objects with a greater natural alignment requirement are emitted as e.g. 128-bit atomics. Fixes #24163. - - - - - f6c486c3 by Matthew Pickering at 2023-11-29T11:08:13-05:00 metrics: Widen libdir and size_hello_obj acceptance window af8816740d9b8759be1a22af8adcb5f13edeb61d shows that the libdir size can fluctuate quite significantly even when the change is quite small. Therefore we widen the acceptance window to 10%. - - - - - 99a6a49c by Alan Zimmerman at 2023-11-29T11:08:49-05:00 EPA: Clean up TC Monad Utils We no longer need the alternative variant of addLocM (addLocMA) nor wrapLocAM, wrapLocSndMA. aarch64-darwin Metric Increase: MultiLayerModulesTH_OneShot deb10-numa-slow Metric Decrease: libdir - - - - - 94fb8d47 by Alan Zimmerman at 2023-11-29T18:10:26+00:00 Match GHC, No comments in EpaDelta for comments - - - - - cbc03fa0 by Sebastian Graf at 2023-11-30T12:37:21-05:00 perf tests: Move comments into new `Note [Sensitivity to unique increment]` (#19414) And additionally to T12545, link from T8095, T13386 to this new Note. - - - - - c7623b22 by Alan Zimmerman at 2023-11-30T12:37:56-05:00 EPA: EpaDelta for comment has no comments EpaLocation is used to position things. It has two constructors, EpaSpan holding a SrcSpan, and EpaDelta with a delta position and a possible list of comments. The comment list is needed because the location in EpaDelta has no absolute information to decide which comments should be emitted before them when printing. But it is also used for specifying the position of a comment. To prevent the absurdity of a comment position having a list of comments in it, we make EpaLocation parameterisable, using comments for the normal case and a constant for within comments. Updates haddock submodule. aarch64-darwin Metric Decrease: MultiLayerModulesTH_OneShot - - - - - bd8acc0c by Krzysztof Gogolewski at 2023-11-30T12:38:32-05:00 Kind-check body of a required forall We now require that in 'forall a -> ty', ty has kind TYPE r for some r. Fixes #24176 - - - - - 010fb784 by Owen Shepherd at 2023-12-03T00:10:09-05:00 docs(NonEmpty/group): Remove incorrect haddock link quotes in code block - - - - - cda9c12d by Owen Shepherd at 2023-12-03T00:10:09-05:00 docs(NonEmpty/group): Remove cycle from group haddock example - - - - - 495265b9 by Owen Shepherd at 2023-12-03T00:10:09-05:00 docs(NonEmpty/group): Use repl haddock syntax in group docs - - - - - d134d1de by Owen Shepherd at 2023-12-03T00:10:09-05:00 docs(NonEmpty/group): Use list [] notation in group haddock - - - - - dfcf629c by Owen Shepherd at 2023-12-03T00:10:10-05:00 docs(NonEmpty/group): Specify final property of group function in haddock - - - - - cad3b734 by Owen Shepherd at 2023-12-03T00:10:10-05:00 fix: Add missing property of List.group - - - - - bad37656 by Matthew Pickering at 2023-12-03T00:10:46-05:00 testsuite: Fix T21097b test with make 4.1 (deb9) cee81370cd6ef256f66035e3116878d4cb82e28b recently added a test which failed on deb9 because the version of make was emitting the recipe failure to stdout rather than stderr. One way to fix this is to be more precise in the test about which part of the output we care about inspecting. - - - - - 5efdf421 by Matthew Pickering at 2023-12-03T00:11:21-05:00 testsuite: Track size of libdir in bytes For consistency it's better if we track all size metrics in bytes. Metric Increase: libdir - - - - - f5eb0f29 by Matthew Pickering at 2023-12-03T00:11:22-05:00 testsuite: Remove rogue trace in testsuite I accidentally left a trace in the generics metric patch. - - - - - d5610737 by Claudio Bley at 2023-12-06T16:13:33-05:00 Only exit ghci in -e mode when :add command fails Previously, when running `ghci -e ':add Sample.hs'` the process would exit with exit code 1 if the file exists and could be loaded. Fixes #24115 - - - - - 0f0c53a5 by Vladislav Zavialov at 2023-12-06T16:14:09-05:00 T2T in Patterns (#23739) This patch implements the T2T (term-to-type) transformation in patterns. Patterns that are checked against a visible forall can now be written without the `type` keyword: \(type t) (x :: t) -> ... -- old \t (x :: t) -> ... -- new The `t` binder is parsed and renamed as a term pattern (Pat), but then undergoes a conversion to a type pattern (HsTyPat). See the new function pat_to_type_pat in compiler/GHC/Tc/Gen/Pat.hs - - - - - 10a1a6c6 by Sebastian Graf at 2023-12-06T16:14:45-05:00 Pmc: Fix SrcLoc and warning for incomplete irrefutable pats (#24234) Before, the source location would point at the surrounding function definition, causing the confusion in #24234. I also took the opportunity to introduce a new `LazyPatCtx :: HsMatchContext _` to make the warning message say "irrefutable pattern" instead of "pattern binding". - - - - - 36b9a38c by Matthew Pickering at 2023-12-06T16:15:21-05:00 libraries: Bump filepath to 1.4.200.1 and unix to 2.8.4.0 Updates filepath submodule Updates unix submodule Fixes #24240 - - - - - 91ff0971 by Matthew Pickering at 2023-12-06T16:15:21-05:00 Submodule linter: Allow references to tags We modify the submodule linter so that if the bumped commit is a specific tag then the commit is accepted. Fixes #24241 - - - - - 86f652dc by Zubin Duggal at 2023-12-06T16:15:21-05:00 hadrian: set -Wno-deprecations for directory and Win32 The filepath bump to 1.4.200.1 introduces a deprecation warning. See https://gitlab.haskell.org/ghc/ghc/-/issues/24240 https://github.com/haskell/filepath/pull/206 - - - - - 7ac6006e by Sylvain Henry at 2023-12-06T16:16:02-05:00 Zap OccInfo on case binders during StgCse #14895 #24233 StgCse can revive dead binders: case foo of dead { Foo x y -> Foo x y; ... } ===> case foo of dead { Foo x y -> dead; ... } -- dead is no longer dead So we must zap occurrence information on case binders. Fix #14895 and #24233 - - - - - 57c391c4 by Sebastian Graf at 2023-12-06T16:16:37-05:00 Cpr: Turn an assertion into a check to deal with some dead code (#23862) See the new `Note [Dead code may contain type confusions]`. Fixes #23862. - - - - - c1c8abf8 by Zubin Duggal at 2023-12-08T02:25:07-05:00 testsuite: add test for #23944 - - - - - 6329d308 by Zubin Duggal at 2023-12-08T02:25:07-05:00 driver: Only run a dynamic-too pipeline if object files are going to be generated Otherwise we run into a panic in hscMaybeWriteIface: "Unexpected DT_Dyn state when writing simple interface" when dynamic-too is enabled We could remove the panic and just write the interface even if the state is `DT_Dyn`, but it seems pointless to run the pipeline twice when `hscMaybeWriteIface` is already designed to write both `hi` and `dyn_hi` files if dynamic-too is enabled. Fixes #23944. - - - - - 28811f88 by Simon Peyton Jones at 2023-12-08T05:47:18-05:00 Improve duplicate elimination in SpecConstr This partially fixes #24229. See the new Note [Pattern duplicate elimination] in SpecConstr - - - - - fec7894f by Simon Peyton Jones at 2023-12-08T05:47:18-05:00 Make SpecConstr deal with casts better This patch does two things, to fix #23209: * It improves SpecConstr so that it no longer quantifies over coercion variables. See Note [SpecConstr and casts] * It improves the rule matcher to deal nicely with the case where the rule does not quantify over coercion variables, but the the template has a cast in it. See Note [Casts in the template] - - - - - 8db8d2fd by Zubin Duggal at 2023-12-08T05:47:54-05:00 driver: Don't lose track of nodes when we fail to resolve cycles The nodes that take part in a cycle should include both hs-boot and hs files, but when we fail to resolve a cycle, we were only counting the nodes from the graph without boot files. Fixes #24196 - - - - - c5b4efd3 by Zubin Duggal at 2023-12-08T05:48:30-05:00 testsuite: Skip MultiLayerModulesTH_OneShot on darwin See #24177 - - - - - fae472a9 by Wendao Lee at 2023-12-08T05:49:12-05:00 docs(Data.Char):Add more detailed descriptions for some functions Related changed function's docs: -GHC.Unicode.isAlpha -GHC.Unicode.isPrint -GHC.Unicode.isAlphaNum Add more details for what the function will return. Co-authored-by: Bodigrim <andrew.lelechenko at gmail.com> - - - - - ca7510e4 by Malik Ammar Faisal at 2023-12-08T05:49:55-05:00 Fix float parsing in GHC Cmm Lexer Add test case for bug #24224 - - - - - d8baa1bd by Simon Peyton Jones at 2023-12-08T15:40:37+00:00 Take care when simplifying unfoldings This MR fixes a very subtle bug exposed by #24242. See Note [Environment for simplLetUnfolding]. I also updated a bunch of Notes on shadowing - - - - - 03ca551d by Simon Peyton Jones at 2023-12-08T15:54:50-05:00 Comments only in FloatIn Relevant to #3458 - - - - - 50c78779 by Simon Peyton Jones at 2023-12-08T15:54:50-05:00 Comments only in SpecConstr - - - - - 9431e195 by Simon Peyton Jones at 2023-12-08T15:54:50-05:00 Add test for #22238 - - - - - d9e4c597 by Vladislav Zavialov at 2023-12-11T04:19:34-05:00 Make forall a keyword (#23719) Before this change, GHC used to accept `forall` as a term-level identifier: -- from constraints-0.13 forall :: forall p. (forall a. Dict (p a)) -> Dict (Forall p) forall d = ... Now it is a parse error. The -Wforall-identifier warning has served its purpose and is now a deprecated no-op. - - - - - 58d56644 by Zubin Duggal at 2023-12-11T04:20:10-05:00 driver: Ensure we actually clear the interactive context before reloading Previously we called discardIC, but immediately after set the session back to an old HscEnv that still contained the IC Partially addresses #24107 Fixes #23405 - - - - - 8e5745a0 by Zubin Duggal at 2023-12-11T04:20:10-05:00 driver: Ensure we force the lookup of old build artifacts before returning the build plan This prevents us from retaining all previous build artifacts in memory until a recompile finishes, instead only retaining the exact artifacts we need. Fixes #24118 - - - - - 105c370c by Zubin Duggal at 2023-12-11T04:20:10-05:00 testsuite: add test for #24118 and #24107 MultiLayerModulesDefsGhci was not able to catch the leak because it uses :l which discards the previous environment. Using :r catches both of these leaks - - - - - e822ff88 by Zubin Duggal at 2023-12-11T04:20:10-05:00 compiler: Add some strictness annotations to ImportSpec and related constructors This prevents us from retaining entire HscEnvs. Force these ImportSpecs when forcing the GlobalRdrEltX Adds an NFData instance for Bag Fixes #24107 - - - - - 522c12a4 by Zubin Duggal at 2023-12-11T04:20:10-05:00 compiler: Force IfGlobalRdrEnv in NFData instance. - - - - - 188b280d by Arnaud Spiwack at 2023-12-11T15:33:31+01:00 LinearTypes => MonoLocalBinds - - - - - 8e0446df by Arnaud Spiwack at 2023-12-11T15:44:28+01:00 Linear let and where bindings For expediency, the initial implementation of linear types in GHC made it so that let and where binders would always be considered unrestricted. This was rather unpleasant, and probably a big obstacle to adoption. At any rate, this was not how the proposal was designed. This patch fixes this infelicity. It was surprisingly difficult to build, which explains, in part, why it took so long to materialise. As of this patch, let or where bindings marked with %1 will be linear (respectively %p for an arbitrary multiplicity p). Unmarked let will infer their multiplicity. Here is a prototypical example of program that used to be rejected and is accepted with this patch: ```haskell f :: A %1 -> B g :: B %1 -> C h :: A %1 -> C h x = g y where y = f x ``` Exceptions: - Recursive let are unrestricted, as there isn't a clear semantics of what a linear recursive binding would be. - Destructive lets with lazy bindings are unrestricted, as their desugaring isn't linear (see also #23461). - (Strict) destructive lets with inferred polymorphic type are unrestricted. Because the desugaring isn't linear (See #18461 down-thread). Closes #18461 and #18739 Co-authored-by: @jackohughes - - - - - effa7e2d by Matthew Craven at 2023-12-12T04:37:20-05:00 Introduce `dataToTagSmall#` primop (closes #21710) ...and use it to generate slightly better code when dataToTag# is used at a "small data type" where there is no need to mess with "is_too_big_tag" or potentially look at an info table. Metric Decrease: T18304 - - - - - 35c7aef6 by Matthew Craven at 2023-12-12T04:37:20-05:00 Fix formatting of Note [alg-alt heap check] - - - - - 7397c784 by Oleg Grenrus at 2023-12-12T04:37:56-05:00 Allow untyped brackets in typed splices and vice versa. Resolves #24190 Apparently the check was essentially always (as far as I can trace back: d0d47ba76f8f0501cf3c4966bc83966ab38cac27), and while it does catch some mismatches, the type-checker will catch them too. OTOH, it prevents writing completely reasonable programs. - - - - - 32d208e1 by Vladislav Zavialov at 2023-12-12T20:41:36+03:00 EPA: Match changes to LHsToken removal - - - - - a3ee3b99 by Moritz Angermann at 2023-12-12T19:50:58-05:00 Drop hard Xcode dependency XCODE_VERSION calls out to `xcodebuild`, which is only available when having `Xcode` installed. The CommandLineTools are not sufficient. To install Xcode, you must have an apple id to download the Xcode.xip from apple. We do not use xcodebuild anywhere in our build explicilty. At best it appears to be a proxy for checking the linker or the compiler. These should rather be done with ``` xcrun ld -version ``` or similar, and not by proxy through Xcode. The CLR should be sufficient for building software on macOS. - - - - - 1c9496e0 by Vladislav Zavialov at 2023-12-12T19:51:34-05:00 docs: update information on RequiredTypeArguments Update the User's Guide and Release Notes to account for the recent progress in the implementation of RequiredTypeArguments. - - - - - d0b17576 by Ben Gamari at 2023-12-13T06:33:37-05:00 rts/eventlog: Fix off-by-one in assertion Previously we failed to account for the NULL terminator `postString` asserted that there is enough room in the buffer for the string. - - - - - a10f9b9b by Ben Gamari at 2023-12-13T06:33:37-05:00 rts/eventlog: Honor result of ensureRoomForVariableEvent is Previously we would keep plugging along, even if isn't enough room for the event. - - - - - 0e0f41c0 by Ben Gamari at 2023-12-13T06:33:37-05:00 rts/eventlog: Avoid truncating event sizes Previously ensureRoomForVariableEvent would truncate the desired size to 16-bits, resulting in #24197. Fixes #24197. - - - - - 64e724c8 by Artin Ghasivand at 2023-12-13T06:34:20-05:00 Remove the "Derived Constraint" argument of TcPluginSolver, docs - - - - - fe6d97dd by Vladislav Zavialov at 2023-12-13T06:34:56-05:00 EPA: Move tokens into GhcPs extension fields (#23447) Summary of changes * Remove Language.Haskell.Syntax.Concrete * Move all tokens into GhcPs extension fields (LHsToken -> EpToken) * Create new TTG extension fields as needed * Drop the MultAnn wrapper Updates the haddock submodule. Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - 8106e695 by Zubin Duggal at 2023-12-13T06:35:34-05:00 testsuite: use copy_files in T23405 This prevents the tree from being dirtied when the file is modified. - - - - - ed0e4099 by Bryan Richter at 2023-12-14T04:30:53-05:00 Document ghc package's PVP-noncompliance This changes nothing, it just makes the status quo explicit. - - - - - 8bef8d9f by Luite Stegeman at 2023-12-14T04:31:33-05:00 JS: Mark spurious CI failures js_fragile(24259) This marks the spurious test failures on the JS platform as js_fragile(24259), so we don't hold up merge requests while fixing the underlying issues. See #24259 - - - - - 1c79526a by Finley McIlwaine at 2023-12-15T12:24:40-08:00 Late plugins - - - - - 000c3302 by Finley McIlwaine at 2023-12-15T12:24:40-08:00 withTiming on LateCCs and late plugins - - - - - be4551ac by Finley McIlwaine at 2023-12-15T12:24:40-08:00 add test for late plugins - - - - - 7c29da9f by Finley McIlwaine at 2023-12-15T12:24:40-08:00 Document late plugins - - - - - 9a52ae46 by Ben Gamari at 2023-12-20T07:07:26-05:00 Fix thunk update ordering Previously we attempted to ensure soundness of concurrent thunk update by synchronizing on the access of the thunk's info table pointer field. This was believed to be sufficient since the indirectee (which may expose a closure allocated by another core) would not be examined until the info table pointer update is complete. However, it turns out that this can result in data races in the presence of multiple threads racing a update a single thunk. For instance, consider this interleaving under the old scheme: Thread A Thread B --------- --------- t=0 Enter t 1 Push update frame 2 Begin evaluation 4 Pause thread 5 t.indirectee=tso 6 Release t.info=BLACKHOLE 7 ... (e.g. GC) 8 Resume thread 9 Finish evaluation 10 Relaxed t.indirectee=x 11 Load t.info 12 Acquire fence 13 Inspect t.indirectee 14 Release t.info=BLACKHOLE Here Thread A enters thunk `t` but is soon paused, resulting in `t` being lazily blackholed at t=6. Then, at t=10 Thread A finishes evaluation and updates `t.indirectee` with a relaxed store. Meanwhile, Thread B enters the blackhole. Under the old scheme this would introduce an acquire-fence but this would only synchronize with Thread A at t=6. Consequently, the result of the evaluation, `x`, is not visible to Thread B, introducing a data race. We fix this by treating the `indirectee` field as we do all other mutable fields. This means we must always access this field with acquire-loads and release-stores. See #23185. - - - - - f4b53538 by Vladislav Zavialov at 2023-12-20T07:08:02-05:00 docs: Fix link to 051-ghc-base-libraries.rst The proposal is no longer available at the previous URL. - - - - - f7e21fab by Matthew Pickering at 2023-12-21T14:57:40+00:00 hadrian: Build all executables in bin/ folder In the end the bindist creation logic copies them all into the bin folder. There is no benefit to building a specific few binaries in the lib/bin folder anymore. This also removes the ad-hoc logic to copy the touchy and unlit executables from stage0 into stage1. It takes <1s to build so we might as well just build it. - - - - - 0038d052 by Zubin Duggal at 2023-12-22T23:28:00-05:00 testsuite: mark jspace as fragile on i386. This test has been flaky for some time and has been failing consistently on i386-linux since 8e0446df landed. See #24261 - - - - - dfd670a0 by Ben Bellick at 2023-12-24T10:10:31-05:00 Deprecate -ddump-json and introduce -fdiagnostics-as-json Addresses #19278 This commit deprecates the underspecified -ddump-json flag and introduces a newer, well-specified flag -fdiagnostics-as-json. Also included is a JSON schema as part of the documentation. The -ddump-json flag will be slated for removal shortly after this merge. - - - - - 609e6225 by Ben Bellick at 2023-12-24T10:10:31-05:00 Deprecate -ddump-json and introduce -fdiagnostics-as-json Addresses #19278 This commit deprecates the underspecified -ddump-json flag and introduces a newer, well-specified flag -fdiagnostics-as-json. Also included is a JSON schema as part of the documentation. The -ddump-json flag will be slated for removal shortly after this merge. - - - - - 865513b2 by Ömer Sinan Ağacan at 2023-12-24T10:11:13-05:00 Fix BNF in user manual 6.6.8.2: formal syntax for instance declarations - - - - - c247b6be by Zubin Duggal at 2023-12-25T16:01:23-05:00 docs: document permissibility of -XOverloadedLabels (#24249) Document the permissibility introduced by https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0170-unrestricted-overloadedlabels.rst - - - - - e5b7eb59 by Ömer Sinan Ağacan at 2023-12-25T16:02:03-05:00 Fix a code block syntax in user manual sec. 6.8.8.6 - - - - - 2db11c08 by Ben Gamari at 2023-12-29T15:35:48-05:00 genSym: Reimplement via CAS on 32-bit platforms Previously the remaining use of the C implementation on 32-bit platforms resulted in a subtle bug, #24261. This was due to the C object (which used the RTS's `atomic_inc64` macro) being compiled without `-threaded` yet later being used in a threaded compiler. Side-step this issue by using the pure Haskell `genSym` implementation on all platforms. This required implementing `fetchAddWord64Addr#` in terms of CAS on 64-bit platforms. - - - - - 19328a8c by Xiaoyan Ren at 2023-12-29T15:36:30-05:00 Do not color the diagnostic code in error messages (#24172) - - - - - 685b467c by Krzysztof Gogolewski at 2023-12-29T15:37:06-05:00 Enforce that bindings of implicit parameters are lifted Fixes #24298 - - - - - bc4d67b7 by Matthew Craven at 2023-12-31T06:15:42-05:00 StgToCmm: Detect some no-op case-continuations ...and generate no code for them. Fixes #24264. - - - - - 5b603139 by Krzysztof Gogolewski at 2023-12-31T06:16:18-05:00 Revert "testsuite: mark jspace as fragile on i386." This reverts commit 0038d052c8c80b4b430bb2aa1c66d5280be1aa95. The atomicity bug should be fixed by !11802. - - - - - d55216ad by Krzysztof Gogolewski at 2024-01-01T12:05:49-05:00 Refactor: store [[PrimRep]] rather than [Type] in STG StgConApp stored a list of types. This list was used exclusively during unarisation of unboxed sums (mkUbxSum). However, this is at a wrong level of abstraction: STG shouldn't be concerned with Haskell types, only PrimReps. Update the code to store a [[PrimRep]]. Also, there's no point in storing this list when we're not dealing with an unboxed sum. - - - - - 8b340bc7 by Ömer Sinan Ağacan at 2024-01-01T12:06:29-05:00 Kind signatures docs: mention that they're allowed in newtypes - - - - - 989bf8e5 by Zubin Duggal at 2024-01-03T20:08:47-05:00 ci: Ensure we use the correct bindist name for the test artifact when generating release ghcup metadata Fixes #24268 - - - - - 89299a89 by Krzysztof Gogolewski at 2024-01-03T20:09:23-05:00 Refactor: remove calls to typePrimRepArgs The function typePrimRepArgs is just a thin wrapper around typePrimRep, adding a VoidRep if the list is empty. However, in StgToByteCode, we were discarding that VoidRep anyway, so there's no point in calling it. - - - - - c7be0c68 by mmzk1526 at 2024-01-03T20:10:07-05:00 Use "-V" for alex version check for better backward compatibility Fixes #24302. In recent versions of alex, "-v" is used for "--verbose" instead of "-version". - - - - - 67dbcc0a by Krzysztof Gogolewski at 2024-01-05T02:07:18-05:00 Fix VoidRep handling in ghci debugger 'go' inside extractSubTerms was giving a bad result given a VoidRep, attempting to round towards the next multiple of 0. I don't understand much about the debugger but the code should be better than it was. Fixes #24306 - - - - - 90ea574e by Krzysztof Gogolewski at 2024-01-05T02:07:54-05:00 VoidRep-related refactor * In GHC.StgToByteCode, replace bcIdPrimId with idPrimRep, bcIdArgRep with idArgRep, atomPrimRep with stgArgRep1. All of them were duplicates. * In GHC.Stg.Unarise, we were converting a PrimRep to a Type and back to PrimRep. Remove the calls to primRepToType and typePrimRep1 which cancel out. * In GHC.STG.Lint, GHC.StgToCmm, GHC.Types.RepType we were filtering out VoidRep from the result of typePrimRep. But typePrimRep never returns VoidRep - remove the filtering. - - - - - eaf72479 by brian at 2024-01-06T23:03:09-05:00 Add unaligned Addr# primops Implements CLC proposal #154: https://github.com/haskell/core-libraries-committee/issues/154 * add unaligned addr primops * add tests * accept tests * add documentation * fix js primops * uncomment in access ops * use Word64 in tests * apply suggestions * remove extra file * move docs * remove random options * use setByteArray# primop * better naming * update base-exports test * add base-exports for other architectures - - - - - d471d445 by Krzysztof Gogolewski at 2024-01-06T23:03:47-05:00 Remove VoidRep from PrimRep, introduce PrimOrVoidRep This introduces data PrimOrVoidRep = VoidRep | NVRep PrimRep changes typePrimRep1 to return PrimOrVoidRep, and adds a new function typePrimRepU to be used when the argument is definitely non-void. Details in Note [VoidRep] in GHC.Types.RepType. Fixes #19520 - - - - - 48720a07 by Matthew Craven at 2024-01-08T18:57:36-05:00 Apply Note [Sensitivity to unique increment] to LargeRecord - - - - - 9e2e180f by Sebastian Graf at 2024-01-08T18:58:13-05:00 Debugging: Add diffUFM for convenient diffing between UniqFMs - - - - - 948f3e35 by Sebastian Graf at 2024-01-08T18:58:13-05:00 Rename Opt_D_dump_stranal to Opt_D_dump_dmdanal ... and Opt_D_dump_str_signatures to Opt_D_dump_dmd_signatures - - - - - 4e217e3e by Sebastian Graf at 2024-01-08T18:58:13-05:00 Deprecate -ddump-stranal and -ddump-str-signatures ... and suggest -ddump-dmdanal and -ddump-dmd-signatures instead - - - - - 6c613c90 by Sebastian Graf at 2024-01-08T18:58:13-05:00 Move testsuite/tests/stranal to testsuite/tests/dmdanal A separate commit so that the rename is obvious to Git(Lab) - - - - - c929f02b by Sebastian Graf at 2024-01-08T18:58:13-05:00 CoreSubst: Stricten `substBndr` and `cloneBndr` Doing so reduced allocations of `cloneBndr` by about 25%. ``` T9233(normal) ghc/alloc 672,488,656 663,083,216 -1.4% GOOD T9675(optasm) ghc/alloc 423,029,256 415,812,200 -1.7% geo. mean -0.1% minimum -1.7% maximum +0.1% ``` Metric Decrease: T9233 - - - - - e3ca78f3 by Krzysztof Gogolewski at 2024-01-10T17:35:59-05:00 Deprecate -Wsemigroup This warning was used to prepare for Semigroup becoming a superclass of Monoid, and for (<>) being exported from Prelude. This happened in GHC 8.4 in 8ae263ceb3566 and feac0a3bc69fd3. The leftover logic for (<>) has been removed in GHC 9.8, 4d29ecdfcc79. Now the warning does nothing at all and can be deprecated. - - - - - 08d14925 by amesgen at 2024-01-10T17:36:42-05:00 WASM metadata: use correct GHC version - - - - - 7a808419 by Xiaoyan Ren at 2024-01-10T17:37:24-05:00 Allow SCC declarations in TH (#24081) - - - - - 28827c51 by Xiaoyan Ren at 2024-01-10T17:37:24-05:00 Fix prettyprinting of SCC pragmas - - - - - ae9cc1a8 by Matthew Craven at 2024-01-10T17:38:01-05:00 Fix loopification in the presence of void arguments This also removes Note [Void arguments in self-recursive tail calls], which was just misleading. It's important to count void args both in the function's arity and at the call site. Fixes #24295. - - - - - b718b145 by Zubin Duggal at 2024-01-10T17:38:36-05:00 testsuite: Teach testsuite driver about c++ sources - - - - - 09cb57ad by Zubin Duggal at 2024-01-10T17:38:36-05:00 driver: Set -DPROFILING when compiling C++ sources with profiling Earlier, we used to pass all preprocessor flags to the c++ compiler. This meant that -DPROFILING was passed to the c++ compiler because it was a part of C++ flags However, this was incorrect and the behaviour was changed in 8ff3134ed4aa323b0199ad683f72165e51a59ab6. See #21291. But that commit exposed this bug where -DPROFILING was no longer being passed when compiling c++ sources. The fix is to explicitly include -DPROFILING in `opt_cxx` when profiling is enabled to ensure we pass the correct options for the way to both C and C++ compilers Fixes #24286 - - - - - 2cf9dd96 by Zubin Duggal at 2024-01-10T17:38:36-05:00 testsuite: rename objcpp -> objcxx To avoid confusion with C Pre Processsor - - - - - af6932d6 by Simon Peyton Jones at 2024-01-10T17:39:12-05:00 Make TYPE and CONSTRAINT not-apart Issue #24279 showed up a bug in the logic in GHC.Core.Unify.unify_ty which is supposed to make TYPE and CONSTRAINT be not-apart. Easily fixed. - - - - - 4a39b5ff by Zubin Duggal at 2024-01-10T17:39:48-05:00 ci: Fix typo in mk_ghcup_metadata.py There was a missing colon in the fix to #24268 in 989bf8e53c08eb22de716901b914b3607bc8dd08 - - - - - 13503451 by Zubin Duggal at 2024-01-10T17:40:24-05:00 release-ci: remove release-x86_64-linux-deb11-release+boot_nonmoving_gc job There is no reason to have this release build or distribute this variation. This configuration is for testing purposes only. - - - - - afca46a4 by Sebastian Graf at 2024-01-10T17:41:00-05:00 Parser: Add a Note detailing why we need happy's `error` to implement layout - - - - - eaf8a06d by Krzysztof Gogolewski at 2024-01-11T00:43:17+01:00 Turn -Wtype-equality-out-of-scope on by default Also remove -Wnoncanonical-{monoid,monad}-instances from -Wcompat, since they are enabled by default. Refresh wcompat-warnings/ test with new -Wcompat warnings. Part of #24267 Co-authored-by: sheaf <sam.derbyshire at gmail.com> - - - - - 42bee5aa by Sebastian Graf at 2024-01-12T21:16:21-05:00 Arity: Require called *exactly once* for eta exp with -fpedantic-bottoms (#24296) In #24296, we had a program in which we eta expanded away an error despite the presence of `-fpedantic-bottoms`. This was caused by turning called *at least once* lambdas into one-shot lambdas, while with `-fpedantic-bottoms` it is only sound to eta expand over lambdas that are called *exactly* once. An example can be found in `Note [Combining arity type with demand info]`. Fixes #24296. - - - - - 7e95f738 by Andreas Klebinger at 2024-01-12T21:16:57-05:00 Aarch64: Enable -mfma by default. Fixes #24311 - - - - - e43788d0 by Jason Shipman at 2024-01-14T12:47:38-05:00 Add more instances for Compose: Fractional, RealFrac, Floating, RealFloat CLC proposal #226 https://github.com/haskell/core-libraries-committee/issues/226 - - - - - ae6d8cd2 by Sebastian Graf at 2024-01-14T12:48:15-05:00 Pmc: COMPLETE pragmas associated with Family TyCons should apply to representation TyCons as well (#24326) Fixes #24326. - - - - - c5fc7304 by sheaf at 2024-01-15T14:15:29-05:00 Use lookupOccRn_maybe in TH.lookupName When looking up a value, we want to be able to find both variables and record fields. So we should not use the lookupSameOccRn_maybe function, as we can't know ahead of time which record field namespace a record field with the given textual name will belong to. Fixes #24293 - - - - - da908790 by Krzysztof Gogolewski at 2024-01-15T14:16:05-05:00 Make the build more strict on documentation errors * Detect undefined labels. This can be tested by adding :ref:`nonexistent` to a documentation rst file; attempting to build docs will fail. Fixed the undefined label in `9.8.1-notes.rst`. * Detect errors. While we have plenty of warnings, we can at least enforce that Sphinx does not report errors. Fixed the error in `required_type_arguments.rst`. Unrelated change: I have documented that the `-dlint` enables `-fcatch-nonexhaustive-cases`, as can be verified by checking `enableDLint`. - - - - - 5077416e by Javier Sagredo at 2024-01-16T15:40:06-05:00 Profiling: Adds an option to not start time profiling at startup Using the functionality provided by d89deeba47ce04a5198a71fa4cbc203fe2c90794, this patch creates a new rts flag `--no-automatic-time-samples` which disables the time profiling when starting a program. It is then expected that the user starts it whenever it is needed. Fixes #24337 - - - - - 5776008c by Matthew Pickering at 2024-01-16T15:40:42-05:00 eventlog: Fix off-by-one error in postIPE We were missing the extra_comma from the calculation of the size of the payload of postIPE. This was causing assertion failures when the event would overflow the buffer by one byte, as ensureRoomForVariable event would report there was enough space for `n` bytes but then we would write `n + 1` bytes into the buffer. Fixes #24287 - - - - - 66dc09b1 by Simon Peyton Jones at 2024-01-16T15:41:18-05:00 Improve SpecConstr (esp nofib/spectral/ansi) This MR makes three improvements to SpecConstr: see #24282 * It fixes an outright (and recently-introduced) bug in `betterPat`, which was wrongly forgetting to compare the lengths of the argument lists. * It enhances ConVal to inclue a boolean for work-free-ness, so that the envt can contain non-work-free constructor applications, so that we can do more: see Note [ConVal work-free-ness] * It rejigs `subsumePats` so that it doesn't reverse the list. This can make a difference because, when patterns overlap, we arbitrarily pick the first. There is no "right" way, but this retains the old pre-subsumePats behaviour, thereby "fixing" the regression in #24282. Nofib results +======================================== | spectral/ansi -21.14% | spectral/hartel/comp_lab_zift -0.12% | spectral/hartel/parstof +0.09% | spectral/last-piece -2.32% | spectral/multiplier +6.03% | spectral/para +0.60% | spectral/simple -0.26% +======================================== | geom mean -0.18% +---------------------------------------- The regression in `multiplier` is sad, but it simply replicates GHC's previous behaviour (e.g. GHC 9.6). - - - - - 65da79b3 by Matthew Pickering at 2024-01-16T15:41:54-05:00 hadrian: Reduce Cabal verbosity The comment claims that `simpleUserHooks` decrease verbosity, and it does, but only for the `postConf` phase. The other phases are too verbose with `-V`. At the moment > 5000 lines of the build log are devoted to output from `cabal copy`. So I take the simple approach and just decrease the verbosity level again. If the output of `postConf` is essential then it would be better to implement our own `UserHooks` which doesn't decrease the verbosity for `postConf`. Fixes #24338 - - - - - 16414d7d by Matthew Pickering at 2024-01-17T10:54:59-05:00 Stop retaining old ModGuts throughout subsequent simplifier phases Each phase of the simplifier typically rewrites the majority of ModGuts, so we want to be able to release the old ModGuts as soon as possible. `name_ppr_ctxt` lives throught the whole optimiser phase and it was retaining a reference to `ModGuts`, so we were failing to release the old `ModGuts` until the end of the phase (potentially doubling peak memory usage for that particular phase). This was discovered using eras profiling (#24332) Fixes #24328 - - - - - 7f0879e1 by Matthew Pickering at 2024-01-17T10:55:35-05:00 Update nofib submodule - - - - - 320454d3 by Cheng Shao at 2024-01-17T23:02:40+00:00 ci: bump ci-images for updated wasm image - - - - - 2eca52b4 by Cheng Shao at 2024-01-17T23:06:44+00:00 base: treat all FDs as "nonblocking" on wasm On posix platforms, when performing read/write on FDs, we check the nonblocking flag first. For FDs without this flag (e.g. stdout), we call fdReady() first, which in turn calls poll() to wait for I/O to be available on that FD. This is problematic for wasm32-wasi: although select()/poll() is supported via the poll_oneoff() wasi syscall, that syscall is rather heavyweight and runtime behavior differs in different wasi implementations. The issue is even worse when targeting browsers, given there's no satisfactory way to implement async I/O as a synchronous syscall, so existing JS polyfills for wasi often give up and simply return ENOSYS. Before we have a proper I/O manager that avoids poll_oneoff() for async I/O on wasm, this patch improves the status quo a lot by merely pretending all FDs are "nonblocking". Read/write on FDs will directly invoke read()/write(), which are much more reliably handled in existing wasi implementations, especially those in browsers. Fixes #23275 and the following test cases: T7773 isEOF001 openFile009 T4808 cgrun025 Approved by CLC proposal #234: https://github.com/haskell/core-libraries-committee/issues/234 - - - - - 83c6c710 by Andrew Lelechenko at 2024-01-18T05:21:49-05:00 base: clarify how to disable warnings about partiality of Data.List.{head,tail} - - - - - c4078f2f by Simon Peyton Jones at 2024-01-18T05:22:25-05:00 Fix four bug in handling of (forall cv. body_ty) These bugs are all described in #24335 It's not easy to provoke the bug, hence no test case. - - - - - 119586ea by Alexis King at 2024-01-19T00:08:00-05:00 Always refresh profiling CCSes after running pending initializers Fixes #24171. - - - - - 9718d970 by Oleg Grenrus at 2024-01-19T00:08:36-05:00 Set default-language: GHC2021 in ghc library Go through compiler/ sources, and remove all BangPatterns (and other GHC2021 enabled extensions in these files). - - - - - 3ef71669 by Matthew Pickering at 2024-01-19T21:55:16-05:00 testsuite: Remove unused have_library function Also remove the hence unused testsuite option `--test-package-db`. Fixes #24342 - - - - - 5b7fa20c by Jade at 2024-01-19T21:55:53-05:00 Fix Spelling in the compiler Tracking: #16591 - - - - - 09875f48 by Matthew Pickering at 2024-01-20T12:20:44-05:00 testsuite: Implement `isInTreeCompiler` in a more robust way Just a small refactoring to avoid redundantly specifying the same strings in two different places. - - - - - 0d12b987 by Jade at 2024-01-20T12:21:20-05:00 Change maintainer email from cvs-ghc at haskell.org to ghc-devs at haskell.org. Fixes #22142 - - - - - eebdd316 by Apoorv Ingle at 2024-01-23T13:49:12+00:00 Changes for haskell/haddock#18324 - - - - - 1fa1c00c by Jade at 2024-01-23T19:17:03-05:00 Enhance Documentation of functions exported by Data.Function This patch aims to improve the documentation of functions exported in Data.Function Tracking: #17929 Fixes: #10065 - - - - - ab47a43d by Jade at 2024-01-23T19:17:39-05:00 Improve documentation of hGetLine. - Add explanation for whether a newline is returned - Add examples Fixes #14804 - - - - - dd4af0e5 by Cheng Shao at 2024-01-23T19:18:17-05:00 Fix genapply for cross-compilation by nuking fragile CPP logic This commit fixes incorrectly built genapply when cross compiling (#24347) by nuking all fragile CPP logic in it from the orbit. All target-specific info are now read from DerivedConstants.h at runtime, see added note for details. Also removes a legacy Makefile and adds haskell language server support for genapply. - - - - - 0cda2b8b by Cheng Shao at 2024-01-23T19:18:17-05:00 rts: enable wasm32 register mapping The wasm backend didn't properly make use of all Cmm global registers due to #24347. Now that it is fixed, this patch re-enables full register mapping for wasm32, and we can now generate smaller & faster wasm modules that doesn't always spill arguments onto the stack. Fixes #22460 #24152. - - - - - 0325a6e5 by Greg Steuck at 2024-01-24T01:29:44-05:00 Avoid utf8 in primops.txt.pp comments They don't make it through readFile' without explicitly setting the encoding. See https://gitlab.haskell.org/ghc/ghc/-/issues/17755 - - - - - 1aaf0bd8 by David Binder at 2024-01-24T01:30:20-05:00 Bump hpc and hpc-bin submodule Bump hpc to 0.7.0.1 Bump hpc-bin to commit d1780eb2 - - - - - e693a4e8 by Ben Gamari at 2024-01-24T01:30:56-05:00 testsuite: Ignore stderr in T8089 Otherwise spurious "Killed: 9" messages to stderr may cause the test to fail. Fixes #24361. - - - - - a40f4ab2 by sheaf at 2024-01-24T14:04:33-05:00 Fix FMA instruction on LLVM We were emitting the wrong instructions for fused multiply-add operations on LLVM: - the instruction name is "llvm.fma.f32" or "llvm.fma.f64", not "fmadd" - LLVM does not support other instructions such as "fmsub"; instead we implement these by flipping signs of some arguments - the instruction is an LLVM intrinsic, which requires handling it like a normal function call instead of a machine instruction Fixes #24223 - - - - - 69abc786 by Andrei Borzenkov at 2024-01-24T14:05:09-05:00 Add changelog entry for renaming tuples from (,,...,,) to Tuple<n> (24291) - - - - - 0ac8f385 by Cheng Shao at 2024-01-25T00:27:48-05:00 compiler: remove unused GHC.Linker module The GHC.Linker module is empty and unused, other than as a hack for the make build system. We can remove it now that make is long gone; the note is moved to GHC.Linker.Loader instead. - - - - - 699da01b by Hécate Moonlight at 2024-01-25T00:28:27-05:00 Clarification for newtype constructors when using `coerce` - - - - - b2d8cd85 by Matt Walker at 2024-01-26T09:50:08-05:00 Fix #24308 Add tests for semicolon separated where clauses - - - - - 0da490a1 by Ben Gamari at 2024-01-26T17:34:41-05:00 hsc2hs: Bump submodule - - - - - 3f442fd2 by Ben Gamari at 2024-01-26T17:34:41-05:00 Bump containers submodule to 0.7 - - - - - 82a1c656 by Sebastian Nagel at 2024-01-29T02:32:40-05:00 base: with{Binary}File{Blocking} only annotates own exceptions Fixes #20886 This ensures that inner, unrelated exceptions are not misleadingly annotated with the opened file. - - - - - 9294a086 by Andreas Klebinger at 2024-01-29T02:33:15-05:00 Fix fma warning when using llvm on aarch64. On aarch64 fma is always on so the +fma flag doesn't exist for that target. Hence no need to try and pass +fma to llvm. Fixes #24379 - - - - - ced2e731 by sheaf at 2024-01-29T17:27:12-05:00 No shadowing warnings for NoFieldSelector fields This commit ensures we don't emit shadowing warnings when a user shadows a field defined with NoFieldSelectors. Fixes #24381 - - - - - 8eeadfad by Patrick at 2024-01-29T17:27:51-05:00 Fix bug wrong span of nested_doc_comment #24378 close #24378 1. Update the start position of span in `nested_doc_comment` correctly. and hence the spans of identifiers of haddoc can be computed correctly. 2. add test `HaddockSpanIssueT24378`. - - - - - a557580f by Alexey Radkov at 2024-01-30T19:41:52-05:00 Fix irrelevant dodgy-foreign-imports warning on import f-pointers by value A test *сс018* is attached (not sure about the naming convention though). Note that without the fix, the test fails with the *dodgy-foreign-imports* warning passed to stderr. The warning disappears after the fix. GHC shouldn't warn on imports of natural function pointers from C by value (which is feasible with CApiFFI), such as ```haskell foreign import capi "cc018.h value f" f :: FunPtr (Int -> IO ()) ``` where ```c void (*f)(int); ``` See a related real-world use-case [here](https://gitlab.com/daniel-casanueva/pcre-light/-/merge_requests/17). There, GHC warns on import of C function pointer `pcre_free`. - - - - - ca99efaf by Alexey Radkov at 2024-01-30T19:41:53-05:00 Rename test cc018 -> T24034 - - - - - 88c38dd5 by Ben Gamari at 2024-01-30T19:42:28-05:00 rts/TraverseHeap.c: Ensure that PosixSource.h is included first - - - - - ca2e919e by Simon Peyton Jones at 2024-01-31T09:29:45+00:00 Make decomposeRuleLhs a bit more clever This fixes #24370 by making decomposeRuleLhs undertand dictionary /functions/ as well as plain /dictionaries/ - - - - - 94ce031d by Teo Camarasu at 2024-02-01T05:49:49-05:00 doc: Add -Dn flag to user guide Resolves #24394 - - - - - 31553b11 by Ben Gamari at 2024-02-01T12:21:29-05:00 cmm: Introduce MO_RelaxedRead In hand-written Cmm it can sometimes be necessary to atomically load from memory deep within an expression (e.g. see the `CHECK_GC` macro). This MachOp provides a convenient way to do so without breaking the expression into multiple statements. - - - - - 0785cf81 by Ben Gamari at 2024-02-01T12:21:29-05:00 codeGen: Use relaxed accesses in ticky bumping - - - - - be423dda by Ben Gamari at 2024-02-01T12:21:29-05:00 base: use atomic write when updating timer manager - - - - - 8a310e35 by Ben Gamari at 2024-02-01T12:21:29-05:00 Use relaxed atomics to manipulate TSO status fields - - - - - d6809ee4 by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Add necessary barriers when manipulating TSO owner - - - - - 39e3ac5d by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Use `switch` to branch on why_blocked This is a semantics-preserving refactoring. - - - - - 515eb33d by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Fix synchronization on thread blocking state We now use a release barrier whenever we update a thread's blocking state. This required widening StgTSO.why_blocked as AArch64 does not support atomic writes on 16-bit values. - - - - - eb38812e by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Fix data race in threadPaused This only affects an assertion in the debug RTS and only needs relaxed ordering. - - - - - 26c48dd6 by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Fix data race in threadStatus# - - - - - 6af43ab4 by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Fix data race in Interpreter's preemption check - - - - - 9502ad3c by Ben Gamari at 2024-02-01T12:21:29-05:00 rts/Messages: Fix data race - - - - - 60802db5 by Ben Gamari at 2024-02-01T12:21:30-05:00 rts/Prof: Fix data race - - - - - ef8ccef5 by Ben Gamari at 2024-02-01T12:21:30-05:00 rts: Use relaxed ordering on dirty/clean info tables updates When changing the dirty/clean state of a mutable object we needn't have any particular ordering. - - - - - 76fe2b75 by Ben Gamari at 2024-02-01T12:21:30-05:00 codeGen: Use relaxed-read in closureInfoPtr - - - - - a6316eb4 by Ben Gamari at 2024-02-01T12:21:30-05:00 STM: Use acquire loads when possible Full sequential consistency is not needed here. - - - - - 6bddfd3d by Ben Gamari at 2024-02-01T12:21:30-05:00 rts: Use fence rather than redundant load Previously we would use an atomic load to ensure acquire ordering. However, we now have `ACQUIRE_FENCE_ON`, which allows us to express this more directly. - - - - - 55c65dbc by Ben Gamari at 2024-02-01T12:21:30-05:00 rts: Fix data races in profiling timer - - - - - 856b5e75 by Ben Gamari at 2024-02-01T12:21:30-05:00 Add Note [C11 memory model] - - - - - 6534da24 by Cheng Shao at 2024-02-01T12:22:07-05:00 compiler: move generic cmm optimization logic in NCG to a standalone module This commit moves GHC.CmmToAsm.cmmToCmm to a standalone module, GHC.Cmm.GenericOpt. The main motivation is enabling this logic to be run in the wasm backend NCG code, which is defined in other modules that's imported by GHC.CmmToAsm, causing a cyclic dependency issue. - - - - - 87e34888 by Cheng Shao at 2024-02-01T12:22:07-05:00 compiler: explicitly disable PIC in wasm32 NCG This commit explicitly disables the ncgPIC flag for the wasm32 target. The wasm backend doesn't support PIC for the time being. - - - - - c6ce242e by Cheng Shao at 2024-02-01T12:22:07-05:00 compiler: enable generic cmm optimizations in wasm backend NCG This commit enables the generic cmm optimizations in other NCGs to be run in the wasm backend as well, followed by a late cmm control-flow optimization pass. The added optimizations do catch some corner cases not handled by the pre-NCG cmm pipeline and are useful in generating smaller CFGs. - - - - - 151dda4e by Andrei Borzenkov at 2024-02-01T12:22:43-05:00 Namespacing for WARNING/DEPRECATED pragmas (#24396) New syntax for WARNING and DEPRECATED pragmas was added, namely namespace specifierss: namespace_spec ::= 'type' | 'data' | {- empty -} warning ::= warning_category namespace_spec namelist strings deprecation ::= namespace_spec namelist strings A new data type was introduced to represent these namespace specifiers: data NamespaceSpecifier = NoSpecifier | TypeNamespaceSpecifier (EpToken "type") | DataNamespaceSpecifier (EpToken "data") Extension field XWarning now contains this NamespaceSpecifier. lookupBindGroupOcc function was changed: it now takes NamespaceSpecifier and checks that the namespace of the found names matches the passed flag. With this change {-# WARNING data D "..." #-} pragma will only affect value namespace and {-# WARNING type D "..." #-} will only affect type namespace. The same logic is applicable to DEPRECATED pragmas. Finding duplicated warnings inside rnSrcWarnDecls now takes into consideration NamespaceSpecifier flag to allow warnings with the same names that refer to different namespaces. - - - - - 38c3afb6 by Bryan Richter at 2024-02-01T12:23:19-05:00 CI: Disable the test-cabal-reinstall job Fixes #24363 - - - - - 27020458 by Matthew Craven at 2024-02-03T01:53:26-05:00 Bump bytestring submodule to something closer to 0.12.1 ...mostly so that 16d6b7e835ffdcf9b894e79f933dd52348dedd0c (which reworks unaligned writes in Builder) and the stuff in https://github.com/haskell/bytestring/pull/631 can see wider testing. The less-terrible code for unaligned writes used in Builder on hosts not known to be ulaigned-friendly also takes less effort for GHC to compile, resulting in a metric decrease for T21839c on some platforms. The metric increase on T21839r is caused by the unrelated commit 750dac33465e7b59100698a330b44de7049a345c. It perhaps warrants further analysis and discussion (see #23822) but is not critical. Metric Decrease: T21839c Metric Increase: T21839r - - - - - cdddeb0f by Rodrigo Mesquita at 2024-02-03T01:54:02-05:00 Work around autotools setting C11 standard in CC/CXX In autoconf >=2.70, C11 is set by default for $CC and $CXX via the -std=...11 flag. In this patch, we split the "-std" flag out of the $CC and $CXX variables, which we traditionally assume to be just the executable name/path, and move it to $CFLAGS/$CXXFLAGS instead. Fixes #24324 - - - - - 5ff7cc26 by Apoorv Ingle at 2024-02-03T13:14:46-06:00 Expand `do` blocks right before typechecking using the `HsExpansion` philosophy. - Fixes #18324 #20020 #23147 #22788 #15598 #22086 #21206 - The change is detailed in - Note [Expanding HsDo with HsExpansion] in `GHC.Tc.Gen.Do` - Note [Doing HsExpansion in the Renamer vs Typechecker] in `GHC.Rename.Expr` expains the rational of doing expansions in type checker as opposed to in the renamer - Adds new datatypes: - `GHC.Hs.Expr.XXExprGhcRn`: new datatype makes this expansion work easier 1. Expansion bits for Expressions, Statements and Patterns in (`ExpandedThingRn`) 2. `PopErrCtxt` a special GhcRn Phase only artifcat to pop the previous error message in the error context stack - `GHC.Basic.Origin` now tracks the reason for expansion in case of Generated This is useful for type checking cf. `GHC.Tc.Gen.Expr.tcExpr` case for `HsLam` - Kills `HsExpansion` and `HsExpanded` as we have inlined them in `XXExprGhcRn` and `XXExprGhcTc` - Ensures warnings such as 1. Pattern match checks 2. Failable patterns 3. non-() return in body statements are preserved - Kill `HsMatchCtxt` in favor of `TcMatchAltChecker` - Testcases: * T18324 T20020 T23147 T22788 T15598 T22086 * T23147b (error message check), * DoubleMatch (match inside a match for pmc check) * pattern-fails (check pattern match with non-refutable pattern, eg. newtype) * Simple-rec (rec statements inside do statment) * T22788 (code snippet from #22788) * DoExpanion1 (Error messages for body statments) * DoExpansion2 (Error messages for bind statements) * DoExpansion3 (Error messages for let statements) Also repoint haddock to the right submodule so that the test (haddockHypsrcTest) pass Metric Increase 'compile_time/bytes allocated': T9020 The testcase is a pathalogical example of a `do`-block with many statements that do nothing. Given that we are expanding the statements into function binds, we will have to bear a (small) 2% cost upfront in the compiler to unroll the statements. - - - - - 0df8ce27 by Vladislav Zavialov at 2024-02-04T03:55:14-05:00 Reduce parser allocations in allocateCommentsP In the most common case, the comment queue is empty, so we can skip the work of processing it. This reduces allocations by about 10% in the parsing001 test. Metric Decrease: MultiLayerModulesRecomp parsing001 - - - - - ae856a82 by Matthew Pickering at 2024-02-05T12:22:39+00:00 ghc-internals fallout - - - - - cfd68290 by Simon Peyton Jones at 2024-02-05T17:58:33-05:00 Stop dropping a case whose binder is demanded This MR fixes #24251. See Note [Case-to-let for strictly-used binders] in GHC.Core.Opt.Simplify.Iteration, plus #24251, for lots of discussion. Final Nofib changes over 0.1%: +----------------------------------------- | imaginary/digits-of-e2 -2.16% | imaginary/rfib -0.15% | real/fluid -0.10% | real/gamteb -1.47% | real/gg -0.20% | real/maillist +0.19% | real/pic -0.23% | real/scs -0.43% | shootout/n-body -0.41% | shootout/spectral-norm -0.12% +======================================== | geom mean -0.05% Pleasingly, overall executable size is down by just over 1%. Compile times (in perf/compiler) wobble around a bit +/- 0.5%, but the geometric mean is -0.1% which seems good. - - - - - e4d137bb by Simon Peyton Jones at 2024-02-05T17:58:33-05:00 Add Note [Bangs in Integer functions] ...to document the bangs in the functions in GHC.Num.Integer - - - - - ce90f12f by Andrei Borzenkov at 2024-02-05T17:59:09-05:00 Hide WARNING/DEPRECATED namespacing under -XExplicitNamespaces (#24396) - - - - - e2ea933f by Simon Peyton Jones at 2024-02-06T10:12:04-05:00 Refactoring in preparation for lazy skolemisation * Make HsMatchContext and HsStmtContext be parameterised over the function name itself, rather than over the pass. See [mc_fun field of FunRhs] in Language.Haskell.Syntax.Expr - Replace types HsMatchContext GhcPs --> HsMatchContextPs HsMatchContext GhcRn --> HsMatchContextRn HsMatchContext GhcTc --> HsMatchContextRn (sic! not Tc) HsStmtContext GhcRn --> HsStmtContextRn - Kill off convertHsMatchCtxt * Split GHC.Tc.Type.BasicTypes.TcSigInfo so that TcCompleteSig (describing a complete user-supplied signature) is its own data type. - Split TcIdSigInfo(CompleteSig, PartialSig) into TcCompleteSig(CSig) TcPartialSig(PSig) - Use TcCompleteSig in tcPolyCheck, CheckGen - Rename types and data constructors: TcIdSigInfo --> TcIdSig TcPatSynInfo(TPSI) --> TcPatSynSig(PatSig) - Shuffle around helper functions: tcSigInfoName (moved to GHC.Tc.Types.BasicTypes) completeSigPolyId_maybe (moved to GHC.Tc.Types.BasicTypes) tcIdSigName (inlined and removed) tcIdSigLoc (introduced) - Rearrange the pattern match in chooseInferredQuantifiers * Rename functions and types: tcMatchesCase --> tcCaseMatches tcMatchesFun --> tcFunBindMatches tcMatchLambda --> tcLambdaMatches tcPats --> tcMatchPats matchActualFunTysRho --> matchActualFunTys matchActualFunTySigma --> matchActualFunTy * Add HasDebugCallStack constraints to: mkBigCoreVarTupTy, mkBigCoreTupTy, boxTy, mkPiTy, mkPiTys, splitAppTys, splitTyConAppNoView_maybe * Use `penv` from the outer context in the inner loop of GHC.Tc.Gen.Pat.tcMultiple * Move tcMkVisFunTy, tcMkInvisFunTy, tcMkScaledFunTys down the file, factor out and export tcMkScaledFunTy. * Move isPatSigCtxt down the file. * Formatting and comments Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - f5d3e03c by Andrei Borzenkov at 2024-02-06T10:12:04-05:00 Lazy skolemisation for @a-binders (#17594) This patch is a preparation for @a-binders implementation. The main changes are: * Skolemisation is now prepared to deal with @binders. See Note [Skolemisation overview] in GHC.Tc.Utils.Unify. Most of the action is in - Utils.Unify.matchExpectedFunTys - Gen.Pat.tcMatchPats - Gen.Expr.tcPolyExprCheck - Gen.Binds.tcPolyCheck Some accompanying refactoring: * I found that funTyConAppTy_maybe was doing a lot of allocation, and rejigged userTypeError_maybe to avoid calling it. - - - - - 532993c8 by Zubin Duggal at 2024-02-06T10:12:41-05:00 driver: Really don't lose track of nodes when we fail to resolve cycles This fixes a bug in 8db8d2fd1c881032b1b360c032b6d9d072c11723, where we could lose track of acyclic components at the start of an unresolved cycle. We now ensure we never loose track of any of these components. As T24275 demonstrates, a "cyclic" SCC might not really be a true SCC: When viewed without boot files, we have a single SCC ``` [REC main:T24275B [main:T24275B {-# SOURCE #-}, main:T24275A {-# SOURCE #-}] main:T24275A [main:T24275A {-# SOURCE #-}]] ``` But with boot files this turns into ``` [NONREC main:T24275B {-# SOURCE #-} [], REC main:T24275B [main:T24275B {-# SOURCE #-}, main:T24275A {-# SOURCE #-}] main:T24275A {-# SOURCE #-} [main:T24275B], NONREC main:T24275A [main:T24275A {-# SOURCE #-}]] ``` Note that this is truly not an SCC, as no nodes are reachable from T24275B.hs-boot. However, we treat this entire group as a single "SCC" because it seems so when we analyse the graph without taking boot files into account. Indeed, we must return a single ResolvedCycle element in the BuildPlan for this as described in Note [Upsweep]. However, since after resolving this is not a true SCC anymore, `findCycle` fails to find a cycle and we have a sub-optimal error message as a result. To handle this, I extended `findCycle` to not assume its input is an SCC, and to try harder to find cycles in its input. Fixes #24275 - - - - - b35dd613 by Zubin Duggal at 2024-02-06T10:13:17-05:00 GHCi: Lookup breakpoint CCs in the correct module We need to look up breakpoint CCs in the module that the breakpoint points to, and not the current module. Fixes #24327 - - - - - b09e6958 by Zubin Duggal at 2024-02-06T10:13:17-05:00 testsuite: Add test for #24327 - - - - - 569b4c10 by doyougnu at 2024-02-07T03:06:26-05:00 ts: add compile_artifact, ignore_extension flag In b521354216f2821e00d75f088d74081d8b236810 the testsuite gained the capability to collect generic metrics. But this assumed that the test was not linking and producing artifacts and we only wanted to track object files, interface files, or build artifacts from the compiler build. However, some backends, such as the JS backend, produce artifacts when compiling, such as the jsexe directory which we want to track. This patch: - tweaks the testsuite to collect generic metrics on any build artifact in the test directory. - expands the exe_extension function to consider windows and adds the ignore_extension flag. - Modifies certain tests to add the ignore_extension flag. Tests such as heaprof002 expect a .ps file, but on windows without ignore_extensions the testsuite will look for foo.exe.ps. Hence the flag. - adds the size_hello_artifact test - - - - - 75a31379 by doyougnu at 2024-02-07T03:06:26-05:00 ts: add wasm_arch, heapprof002 wasm extension - - - - - c9731d6d by Rodrigo Mesquita at 2024-02-07T03:07:03-05:00 Synchronize bindist configure for #24324 In cdddeb0f1280b40cc194028bbaef36e127175c4c, we set up a workaround for #24324 in the in-tree configure script, but forgot to update the bindist configure script accordingly. This updates it. - - - - - d309f4e7 by Matthew Pickering at 2024-02-07T03:07:38-05:00 distrib/configure: Fix typo in CONF_GCC_LINKER_OPTS_STAGE2 variable Instead we were setting CONF_GCC_LINK_OPTS_STAGE2 which meant that we were missing passing `--target` when invoking the linker. Fixes #24414 - - - - - 77db84ab by Ben Gamari at 2024-02-08T00:35:22-05:00 llvmGen: Adapt to allow use of new pass manager. We now must use `-passes` in place of `-O<n>` due to #21936. Closes #21936. - - - - - 3c9ddf97 by Matthew Pickering at 2024-02-08T00:35:59-05:00 testsuite: Mark length001 as fragile on javascript Modifying the timeout multiplier is not a robust way to get this test to reliably fail. Therefore we mark it as fragile until/if javascript ever supports the stack limit. - - - - - 20b702b5 by Matthew Pickering at 2024-02-08T00:35:59-05:00 Javascript: Don't filter out rtsDeps list This logic appears to be incorrect as it would drop any dependency which was not in a direct dependency of the package being linked. In the ghc-internals split this started to cause errors because `ghc-internal` is not a direct dependency of most packages, and hence important symbols to keep which are hard coded into the js runtime were getting dropped. - - - - - 2df96366 by Ben Gamari at 2024-02-08T00:35:59-05:00 base: Cleanup whitespace in cbits - - - - - 44f6557a by Ben Gamari at 2024-02-08T00:35:59-05:00 Move `base` to `ghc-internal` Here we move a good deal of the implementation of `base` into a new package, `ghc-internal` such that it can be evolved independently from the user-visible interfaces of `base`. While we want to isolate implementation from interfaces, naturally, we would like to avoid turning `base` into a mere set of module re-exports. However, this is a non-trivial undertaking for a variety of reasons: * `base` contains numerous known-key and wired-in things, requiring corresponding changes in the compiler * `base` contains a significant amount of C code and corresponding autoconf logic, which is very fragile and difficult to break apart * `base` has numerous import cycles, which are currently dealt with via carefully balanced `hs-boot` files * We must not break existing users To accomplish this migration, I tried the following approaches: * [Split-GHC.Base]: Break apart the GHC.Base knot to allow incremental migration of modules into ghc-internal: this knot is simply too intertwined to be easily pulled apart, especially given the rather tricky import cycles that it contains) * [Move-Core]: Moving the "core" connected component of base (roughly 150 modules) into ghc-internal. While the Haskell side of this seems tractable, the C dependencies are very subtle to break apart. * [Move-Incrementally]: 1. Move all of base into ghc-internal 2. Examine the module structure and begin moving obvious modules (e.g. leaves of the import graph) back into base 3. Examine the modules remaining in ghc-internal, refactor as necessary to facilitate further moves 4. Go to (2) iterate until the cost/benefit of further moves is insufficient to justify continuing 5. Rename the modules moved into ghc-internal to ensure that they don't overlap with those in base 6. For each module moved into ghc-internal, add a shim module to base with the declarations which should be exposed and any requisite Haddocks (thus guaranteeing that base will be insulated from changes in the export lists of modules in ghc-internal Here I am using the [Move-Incrementally] approach, which is empirically the least painful of the unpleasant options above Bumps haddock submodule. Metric Decrease: haddock.Cabal haddock.base Metric Increase: MultiComponentModulesRecomp T16875 size_hello_artifact - - - - - e8fb2451 by Vladislav Zavialov at 2024-02-08T00:36:36-05:00 Haddock comments on infix constructors (#24221) Rewrite the `HasHaddock` instance for `ConDecl GhcPs` to account for infix constructors. This change fixes a Haddock regression (introduced in 19e80b9af252) that affected leading comments on infix data constructor declarations: -- | Docs for infix constructor | Int :* Bool The comment should be associated with the data constructor (:*), not with its left-hand side Int. - - - - - 9060d55b by Ben Gamari at 2024-02-08T00:37:13-05:00 Add os-string as a boot package Introduces `os-string` submodule. This will be necessary for `filepath-1.5`. - - - - - 9d65235a by Ben Gamari at 2024-02-08T00:37:13-05:00 gitignore: Ignore .hadrian_ghci_multi/ - - - - - d7ee12ea by Ben Gamari at 2024-02-08T00:37:13-05:00 hadrian: Set -this-package-name When constructing the GHC flags for a package Hadrian must take care to set `-this-package-name` in addition to `-this-unit-id`. This hasn't broken until now as we have not had any uses of qualified package imports. However, this will change with `filepath-1.5` and the corresponding `unix` bump, breaking `hadrian/multi-ghci`. - - - - - f2dffd2e by Ben Gamari at 2024-02-08T00:37:13-05:00 Bump filepath to 1.5.0.0 Required bumps of the following submodules: * `directory` * `filepath` * `haskeline` * `process` * `unix` * `hsc2hs` * `Win32` * `semaphore-compat` and the addition of `os-string` as a boot package. - - - - - ab533e71 by Matthew Pickering at 2024-02-08T00:37:50-05:00 Use specific clang assembler when compiling with -fllvm There are situations where LLVM will produce assembly which older gcc toolchains can't handle. For example on Deb10, it seems that LLVM >= 13 produces assembly which the default gcc doesn't support. A more robust solution in the long term is to require a specific LLVM compatible assembler when using -fllvm. Fixes #16354 - - - - - c32b6426 by Matthew Pickering at 2024-02-08T00:37:50-05:00 Update CI images with LLVM 15, ghc-9.6.4 and cabal-install-3.10.2.0 - - - - - 5fcd58be by Matthew Pickering at 2024-02-08T00:37:50-05:00 Update bootstrap plans for 9.4.8 and 9.6.4 - - - - - 707a32f5 by Matthew Pickering at 2024-02-08T00:37:50-05:00 Add alpine 3_18 release job This is mainly experimental and future proofing to enable a smooth transition to newer alpine releases once 3_12 is too old. - - - - - c37931b3 by John Ericson at 2024-02-08T06:39:05-05:00 Generate LLVM min/max bound policy via Hadrian Per #23966, I want the top-level configure to only generate configuration data for Hadrian, not do any "real" tasks on its own. This is part of that effort --- one less file generated by it. (It is still done with a `.in` file, so in a future world non-Hadrian also can easily create this file.) Split modules: - GHC.CmmToLlvm.Config - GHC.CmmToLlvm.Version - GHC.CmmToLlvm.Version.Bounds - GHC.CmmToLlvm.Version.Type This also means we can get rid of the silly `unused.h` introduced in !6803 / 7dfcab2f4bcb7206174ea48857df1883d05e97a2 as temporary kludge. Part of #23966 - - - - - 9f987235 by Apoorv Ingle at 2024-02-08T06:39:42-05:00 Enable mdo statements to use HsExpansions Fixes: #24411 Added test T24411 for regression - - - - - f8429266 by Jade at 2024-02-08T14:56:50+01:00 Adjust test for ghc MR !10993 - - - - - 762b2120 by Jade at 2024-02-08T15:17:15+00:00 Improve Monad, Functor & Applicative docs This patch aims to improve the documentation of Functor, Applicative, Monad and related symbols. The main goal is to make it more consistent and make accessible. See also: !10979 (closed) and !10985 (closed) Ticket #17929 Updates haddock submodule - - - - - 151770ca by Josh Meredith at 2024-02-10T14:28:15-05:00 JavaScript codegen: Use GHC's tag inference where JS backend-specific evaluation inference was previously used (#24309) - - - - - 2e880635 by Zubin Duggal at 2024-02-10T14:28:51-05:00 ci: Allow release-hackage-lint to fail Otherwise it blocks the ghcup metadata pipeline from running. - - - - - b0293f78 by Matthew Pickering at 2024-02-10T14:29:28-05:00 rts: eras profiling mode The eras profiling mode is useful for tracking the life-time of closures. When a closure is written, the current era is recorded in the profiling header. This records the era in which the closure was created. * Enable with -he * User mode: Use functions ghc-experimental module GHC.Profiling.Eras to modify the era * Automatically: --automatic-era-increment, increases the user era on major collections * The first era is era 1 * -he<era> can be used with other profiling modes to select a specific era If you just want to record the era but not to perform heap profiling you can use `-he --no-automatic-heap-samples`. https://well-typed.com/blog/2024/01/ghc-eras-profiling/ Fixes #24332 - - - - - be674a2c by Jade at 2024-02-10T14:30:04-05:00 Adjust error message for trailing whitespace in as-pattern. Fixes #22524 - - - - - 53ef83f9 by doyougnu at 2024-02-10T14:30:47-05:00 gitlab: js: add codeowners Fixes: - #24409 Follow on from: - #21078 and MR !9133 - When we added the JS backend this was forgotten. This patch adds the rightful codeowners. - - - - - 8bbe12f2 by Matthew Pickering at 2024-02-10T14:31:23-05:00 Bump CI images so that alpine3_18 image includes clang15 The only changes here are that clang15 is now installed on the alpine-3_18 image. - - - - - df9fd9f7 by Sylvain Henry at 2024-02-12T12:18:42-05:00 JS: handle stored null StablePtr Some Haskell codes unsafely cast StablePtr into ptr to compare against NULL. E.g. in direct-sqlite: if castStablePtrToPtr aggStPtr /= nullPtr then where `aggStPtr` is read (`peek`) from zeroed memory initially. We fix this by giving these StablePtr the same representation as other null pointers. It's safe because StablePtr at offset 0 is unused (for this exact reason). - - - - - 55346ede by Sylvain Henry at 2024-02-12T12:18:42-05:00 JS: disable MergeObjsMode test This isn't implemented for JS backend objects. - - - - - aef587f6 by Sylvain Henry at 2024-02-12T12:18:42-05:00 JS: add support for linking C sources Support linking C sources with JS output of the JavaScript backend. See the added documentation in the users guide. The implementation simply extends the JS linker to use the objects (.o) that were already produced by the emcc compiler and which were filtered out previously. I've also added some options to control the link with C functions (see the documentation about pragmas). With this change I've successfully compiled the direct-sqlite package which embeds the sqlite.c database code. Some wrappers are still required (see the documentation about wrappers) but everything generic enough to be reused for other libraries have been integrated into rts/js/mem.js. - - - - - b71b392f by Sylvain Henry at 2024-02-12T12:18:42-05:00 JS: avoid EMCC logging spurious failure emcc would sometime output messages like: cache:INFO: generating system asset: symbol_lists/424b44514e43d789148e69e4e7d1c7fdc0350b79.json... (this will be cached in "/emsdk/upstream/emscripten/cache/symbol_lists/424b44514e43d789148e69e4e7d1c7fdc0350b79.json" for subsequent builds) cache:INFO: - ok Cf https://github.com/emscripten-core/emscripten/issues/18607 This breaks our tests matching the stderr output. We avoid this by setting EMCC_LOGGING=0 - - - - - ff2c0cc9 by Simon Peyton Jones at 2024-02-12T12:19:17-05:00 Remove a dead comment Just remove an out of date block of commented-out code, and tidy up the relevant Notes. See #8317. - - - - - bedb4f0d by Teo Camarasu at 2024-02-12T18:50:33-05:00 nonmoving: Add support for heap profiling Add support for heap profiling while using the nonmoving collector. We greatly simply the implementation by disabling concurrent collection for GCs when heap profiling is enabled. This entails that the marked objects on the nonmoving heap are exactly the live objects. Note that we match the behaviour for live bytes accounting by taking the size of objects on the nonmoving heap to be that of the segment's block rather than the object itself. Resolves #22221 - - - - - d0d5acb5 by Teo Camarasu at 2024-02-12T18:51:09-05:00 doc: Add requires prof annotation to options that require it Resolves #24421 - - - - - 6d1e2386 by Alan Zimmerman at 2024-02-13T22:00:28+03:00 EPA: Match changes to HsParTy and HsFunTy - - - - - 57bb8c92 by Cheng Shao at 2024-02-13T14:07:49-05:00 deriveConstants: add needed constants for wasm backend This commit adds needed constants to deriveConstants. They are used by RTS code in the wasm backend to support the JSFFI logic. - - - - - 615eb855 by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: make genSym use C-based atomic increment on non-JS 32-bit platforms The pure Haskell implementation causes i386 regression in unrelated work that can be fixed by using C-based atomic increment, see added comment for details. - - - - - a9918891 by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: allow JSFFI for wasm32 This commit allows the javascript calling convention to be used when the target platform is wasm32. - - - - - 8771a53b by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: allow boxed JSVal as a foreign type This commit allows the boxed JSVal type to be used as a foreign argument/result type. - - - - - 053c92b3 by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: ensure ctors have the right priority on wasm32 This commit fixes the priorities of ctors generated by GHC codegen on wasm32, see the referred note for details. - - - - - b7942e0a by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: add JSFFI desugar logic for wasm32 This commit adds JSFFI desugar logic for the wasm backend. - - - - - 2c1dca76 by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: add JavaScriptFFI to supported extension list on wasm32 This commit adds JavaScriptFFI as a supported extension when the target platform is wasm32. - - - - - 9ad0e2b4 by Cheng Shao at 2024-02-13T14:07:49-05:00 rts/ghc-internal: add JSFFI support logic for wasm32 This commit adds rts/ghc-internal logic to support the wasm backend's JSFFI functionality. - - - - - e9ebea66 by Cheng Shao at 2024-02-13T14:07:49-05:00 ghc-internal: fix threadDelay for wasm in browsers This commit fixes broken threadDelay for wasm when it runs in browsers, see added note for detailed explanation. - - - - - f85f3fdb by Cheng Shao at 2024-02-13T14:07:49-05:00 utils: add JSFFI utility code This commit adds JavaScript util code to utils to support the wasm backend's JSFFI functionality: - jsffi/post-link.mjs, a post-linker to process the linked wasm module and emit a small complement JavaScript ESM module to be used with it at runtime - jsffi/prelude.js, a tiny bit of prelude code as the JavaScript side of runtime logic - jsffi/test-runner.mjs, run the jsffi test cases Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - 77e91500 by Cheng Shao at 2024-02-13T14:07:49-05:00 hadrian: distribute jsbits needed for wasm backend's JSFFI support The post-linker.mjs/prelude.js files are now distributed in the bindist libdir, so when using the wasm backend's JSFFI feature, the user wouldn't need to fetch them from a ghc checkout manually. - - - - - c47ba1c3 by Cheng Shao at 2024-02-13T14:07:49-05:00 testsuite: add opts.target_wrapper This commit adds opts.target_wrapper which allows overriding the target wrapper on a per test case basis when testing a cross target. This is used when testing the wasm backend's JSFFI functionality; the rest of the cases are tested using wasmtime, though the jsffi cases are tested using the node.js based test runner. - - - - - 8e048675 by Cheng Shao at 2024-02-13T14:07:49-05:00 testsuite: T22774 should work for wasm JSFFI T22774 works since the wasm backend now supports the JSFFI feature. - - - - - 1d07f9a6 by Cheng Shao at 2024-02-13T14:07:49-05:00 testsuite: add JSFFI test cases for wasm backend This commit adds a few test cases for the wasm backend's JSFFI functionality, as well as a simple README to instruct future contributors to add new test cases. - - - - - b8997080 by Cheng Shao at 2024-02-13T14:07:49-05:00 docs: add documentation for wasm backend JSFFI This commit adds changelog and user facing documentation for the wasm backend's JSFFI feature. - - - - - ffeb000d by David Binder at 2024-02-13T14:08:30-05:00 Add tests from libraries/process/tests and libraries/Win32/tests to GHC These tests were previously part of the libraries, which themselves are submodules of the GHC repository. This commit moves the tests directly to the GHC repository. - - - - - 5a932cf2 by David Binder at 2024-02-13T14:08:30-05:00 Do not execute win32 tests on non-windows runners - - - - - 500d8cb8 by Jade at 2024-02-13T14:09:07-05:00 prevent GHCi (and runghc) from suggesting other symbols when not finding main Fixes: #23996 - - - - - b19ec331 by Cheng Shao at 2024-02-13T14:09:46-05:00 rts: update xxHash to v0.8.2 - - - - - 4a97bdb8 by Cheng Shao at 2024-02-13T14:09:46-05:00 rts: use XXH3_64bits hash on all 64-bit platforms This commit enables XXH3_64bits hash to be used on all 64-bit platforms. Previously it was only enabled on x86_64, so platforms like aarch64 silently falls back to using XXH32 which degrades the hashing function quality. - - - - - ee01de7d by Cheng Shao at 2024-02-13T14:09:46-05:00 rts: define XXH_INLINE_ALL This commit cleans up how we include the xxhash.h header and only define XXH_INLINE_ALL, which is sufficient to inline the xxHash functions without symbol collision. - - - - - 0e01e1db by Alan Zimmerman at 2024-02-14T02:13:22-05:00 EPA: Move EpAnn out of extension points Leaving a few that are too tricky, maybe some other time. Also - remove some unneeded helpers from Parser.y - reduce allocations with strictness annotations Updates haddock submodule Metric Decrease: parsing001 - - - - - de589554 by Andreas Klebinger at 2024-02-14T02:13:59-05:00 Fix ffi callbacks with >6 args and non-64bit args. Check for ptr/int arguments rather than 64-bit width arguments when counting integer register arguments. The old approach broke when we stopped using exclusively W64-sized types to represent sub-word sized integers. Fixes #24314 - - - - - 9c588f19 by Fendor at 2024-02-14T11:05:36+01:00 Adapt to GHC giving better Name's for linking - - - - - 325b7613 by Ben Gamari at 2024-02-14T14:27:45-05:00 rts/EventLog: Place eliminate duplicate strlens Previously many of the `post*` implementations would first compute the length of the event's strings in order to determine the event length. Later we would then end up computing the length yet again in `postString`. Now we instead pass the string length to `postStringLen`, avoiding the repeated work. - - - - - 8aafa51c by Ben Gamari at 2024-02-14T14:27:46-05:00 rts/eventlog: Place upper bound on IPE string field lengths The strings in IPE events may be of unbounded length. Limit the lengths of these fields to 64k characters to ensure that we don't exceed the maximum event length. - - - - - 0e60d52c by Zubin Duggal at 2024-02-14T14:27:46-05:00 rts: drop unused postString function - - - - - d8d1333a by Cheng Shao at 2024-02-14T14:28:23-05:00 compiler/rts: fix wasm unreg regression This commit fixes two wasm unreg regressions caught by a nightly pipeline: - Unknown stg_scheduler_loopzh symbol when compiling scheduler.cmm - Invalid _hs_constructor(101) function name when handling ctor - - - - - 264a4fa9 by Owen Shepherd at 2024-02-15T09:41:06-05:00 feat: Add sortOn to Data.List.NonEmpty Adds `sortOn` to `Data.List.NonEmpty`, and adds comments describing when to use it, compared to `sortWith` or `sortBy . comparing`. The aim is to smooth out the API between `Data.List`, and `Data.List.NonEmpty`. This change has been discussed in the [clc issue](https://github.com/haskell/core-libraries-committee/issues/227). - - - - - b57200de by Fendor at 2024-02-15T09:41:47-05:00 Prefer RdrName over OccName for looking up locations in doc renaming step Looking up by OccName only does not take into account when functions are only imported in a qualified way. Fixes issue #24294 Bump haddock submodule to include regression test - - - - - 8ad02724 by Luite Stegeman at 2024-02-15T17:33:32-05:00 JS: add simple optimizer The simple optimizer reduces the size of the code generated by the JavaScript backend without the complexity and performance penalty of the optimizer in GHCJS. Also see #22736 Metric Decrease: libdir size_hello_artifact - - - - - 20769b36 by Matthew Pickering at 2024-02-15T17:34:07-05:00 base: Expose `--no-automatic-time-samples` in `GHC.RTS.Flags` API This patch builds on 5077416e12cf480fb2048928aa51fa4c8fc22cf1 and modifies the base API to reflect the new RTS flag. CLC proposal #243 - https://github.com/haskell/core-libraries-committee/issues/243 Fixes #24337 - - - - - 778e1db3 by Andrei Borzenkov at 2024-02-16T16:12:07+03:00 Namespace specifiers for fixity signatures - - - - - 08031ada by Teo Camarasu at 2024-02-16T13:37:00-05:00 base: export System.Mem.performBlockingMajorGC The corresponding C function was introduced in ba73a807edbb444c49e0cf21ab2ce89226a77f2e. As part of #22264. Resolves #24228 The CLC proposal was disccused at: https://github.com/haskell/core-libraries-committee/issues/230 Co-authored-by: Ben Gamari <bgamari.foss at gmail.com> - - - - - 1f534c2e by Florian Weimer at 2024-02-16T13:37:42-05:00 Fix C output for modern C initiative GCC 14 on aarch64 rejects the C code written by GHC with this kind of error: error: assignment to ‘ffi_arg’ {aka ‘long unsigned int’} from ‘HsPtr’ {aka ‘void *’} makes integer from pointer without a cast [-Wint-conversion] 68 | *(ffi_arg*)resp = cret; | ^ Add the correct cast. For more information on this see: https://fedoraproject.org/wiki/Changes/PortingToModernC Tested-by: Richard W.M. Jones <rjones at redhat.com> - - - - - 5d3f7862 by Matthew Craven at 2024-02-16T13:38:18-05:00 Bump bytestring submodule to 0.12.1.0 - - - - - 902ebcc2 by Ian-Woo Kim at 2024-02-17T06:01:01-05:00 Add missing BCO handling in scavenge_one. - - - - - 97d26206 by Sylvain Henry at 2024-02-17T06:01:44-05:00 Make cast between words and floats real primops (#24331) First step towards fixing #24331. Replace foreign prim imports with real primops. - - - - - a40e4781 by Sylvain Henry at 2024-02-17T06:01:44-05:00 Perf: add constant folding for bitcast between float and word (#24331) - - - - - 5fd2c00f by Sylvain Henry at 2024-02-17T06:01:44-05:00 Perf: replace stack checks with assertions in casting primops There are RESERVED_STACK_WORDS free words (currently 21) on the stack, so omit the checks. Suggested by Cheng Shao. - - - - - 401dfe7b by Sylvain Henry at 2024-02-17T06:01:44-05:00 Reexport primops from GHC.Float + add deprecation - - - - - 4ab48edb by Ben Gamari at 2024-02-17T06:02:21-05:00 rts/Hash: Don't iterate over chunks if we don't need to free data When freeing a `HashTable` there is no reason to walk over the hash list before freeing it if the user has not given us a `dataFreeFun`. Noticed while looking at #24410. - - - - - bd5a1f91 by Cheng Shao at 2024-02-17T06:03:00-05:00 compiler: add SEQ_CST fence support In addition to existing Acquire/Release fences, this commit adds SEQ_CST fence support to GHC, allowing Cmm code to explicitly emit a fence that enforces total memory ordering. The following logic is added: - The MO_SeqCstFence callish MachOp - The %prim fence_seq_cst() Cmm syntax and the SEQ_CST_FENCE macro in Cmm.h - MO_SeqCstFence lowering logic in every single GHC codegen backend - - - - - 2ce2a493 by Cheng Shao at 2024-02-17T06:03:38-05:00 testsuite: fix hs_try_putmvar002 for targets without pthread.h hs_try_putmvar002 includes pthread.h and doesn't work on targets without this header (e.g. wasm32). It doesn't need to include this header at all. This was previously unnoticed by wasm CI, though recent toolchain upgrade brought in upstream changes that completely removes pthread.h in the single-threaded wasm32-wasi sysroot, therefore we need to handle that change. - - - - - 1fb3974e by Cheng Shao at 2024-02-17T06:03:38-05:00 ci: bump ci-images to use updated wasm image This commit bumps our ci-images revision to use updated wasm image. - - - - - 56e3f097 by Andrew Lelechenko at 2024-02-17T06:04:13-05:00 Bump submodule text to 2.1.1 T17123 allocates less because of improvements to Data.Text.concat in 1a6a06a. Metric Decrease: T17123 - - - - - a7569495 by Cheng Shao at 2024-02-17T06:04:51-05:00 rts: remove redundant rCCCS initialization This commit removes the redundant logic of initializing each Capability's rCCCS to CCS_SYSTEM in initProfiling(). Before initProfiling() is called during RTS startup, each Capability's rCCCS has already been assigned CCS_SYSTEM when they're first initialized. - - - - - 7a0293cc by Ben Gamari at 2024-02-19T07:11:00-05:00 Drop dependence on `touch` This drops GHC's dependence on the `touch` program, instead implementing it within GHC. This eliminates an external dependency and means that we have one fewer program to keep track of in the `configure` script - - - - - 0dbd729e by Andrei Borzenkov at 2024-02-19T07:11:37-05:00 Parser, renamer, type checker for @a-binders (#17594) GHC Proposal 448 introduces binders for invisible type arguments (@a-binders) in various contexts. This patch implements @-binders in lambda patterns and function equations: {-# LANGUAGE TypeAbstractions #-} id1 :: a -> a id1 @t x = x :: t -- @t-binder on the LHS of a function equation higherRank :: (forall a. (Num a, Bounded a) => a -> a) -> (Int8, Int16) higherRank f = (f 42, f 42) ex :: (Int8, Int16) ex = higherRank (\ @a x -> maxBound @a - x ) -- @a-binder in a lambda pattern in an argument -- to a higher-order function Syntax ------ To represent those @-binders in the AST, the list of patterns in Match now uses ArgPat instead of Pat: data Match p body = Match { ... - m_pats :: [LPat p], + m_pats :: [LArgPat p], ... } + data ArgPat pass + = VisPat (XVisPat pass) (LPat pass) + | InvisPat (XInvisPat pass) (HsTyPat (NoGhcTc pass)) + | XArgPat !(XXArgPat pass) The VisPat constructor represents patterns for visible arguments, which include ordinary value-level arguments and required type arguments (neither is prefixed with a @), while InvisPat represents invisible type arguments (prefixed with a @). Parser ------ In the grammar (Parser.y), the lambda and lambda-cases productions of aexp non-terminal were updated to accept argpats instead of apats: aexp : ... - | '\\' apats '->' exp + | '\\' argpats '->' exp ... - | '\\' 'lcases' altslist(apats) + | '\\' 'lcases' altslist(argpats) ... + argpat : apat + | PREFIX_AT atype Function left-hand sides did not require any changes to the grammar, as they were already parsed with productions capable of parsing @-binders. Those binders were being rejected in post-processing (isFunLhs), and now we accept them. In Parser.PostProcess, patterns are constructed with the help of PatBuilder, which is used as an intermediate data structure when disambiguating between FunBind and PatBind. In this patch we define ArgPatBuilder to accompany PatBuilder. ArgPatBuilder is a short-lived data structure produced in isFunLhs and consumed in checkFunBind. Renamer ------- Renaming of @-binders builds upon prior work on type patterns, implemented in 2afbddb0f24, which guarantees proper scoping and shadowing behavior of bound type variables. This patch merely defines rnLArgPatsAndThen to process a mix of visible and invisible patterns: + rnLArgPatsAndThen :: NameMaker -> [LArgPat GhcPs] -> CpsRn [LArgPat GhcRn] + rnLArgPatsAndThen mk = mapM (wrapSrcSpanCps rnArgPatAndThen) where + rnArgPatAndThen (VisPat x p) = ... rnLPatAndThen ... + rnArgPatAndThen (InvisPat _ tp) = ... rnHsTyPat ... Common logic between rnArgPats and rnPats is factored out into the rn_pats_general helper. Type checker ------------ Type-checking of @-binders builds upon prior work on lazy skolemisation, implemented in f5d3e03c56f. This patch extends tcMatchPats to handle @-binders. Now it takes and returns a list of LArgPat rather than LPat: tcMatchPats :: ... - -> [LPat GhcRn] + -> [LArgPat GhcRn] ... - -> TcM ([LPat GhcTc], a) + -> TcM ([LArgPat GhcTc], a) Invisible binders in the Match are matched up with invisible (Specified) foralls in the type. This is done with a new clause in the `loop` worker of tcMatchPats: loop :: [LArgPat GhcRn] -> [ExpPatType] -> TcM ([LArgPat GhcTc], a) loop (L l apat : pats) (ExpForAllPatTy (Bndr tv vis) : pat_tys) ... -- NEW CLAUSE: | InvisPat _ tp <- apat, isSpecifiedForAllTyFlag vis = ... In addition to that, tcMatchPats no longer discards type patterns. This is done by filterOutErasedPats in the desugarer instead. x86_64-linux-deb10-validate+debug_info Metric Increase: MultiLayerModulesTH_OneShot - - - - - 486979b0 by Jade at 2024-02-19T07:12:13-05:00 Add specialized sconcat implementation for Data.Monoid.First and Data.Semigroup.First Approved CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/246 Fixes: #24346 - - - - - 17e309d2 by John Ericson at 2024-02-19T07:12:49-05:00 Fix reST in users guide It appears that aef587f65de642142c1dcba0335a301711aab951 wasn't valid syntax. - - - - - 35b0ad90 by Brandon Chinn at 2024-02-19T07:13:25-05:00 Fix searching for errors in sphinx build - - - - - 4696b966 by Cheng Shao at 2024-02-19T07:14:02-05:00 hadrian: fix wasm backend post linker script permissions The post-link.mjs script was incorrectly copied and installed as a regular data file without executable permission, this commit fixes it. - - - - - a6142e0c by Cheng Shao at 2024-02-19T07:14:40-05:00 testsuite: mark T23540 as fragile on i386 See #24449 for details. - - - - - 249caf0d by Matthew Craven at 2024-02-19T20:36:09-05:00 Add @since annotation to Data.Data.mkConstrTag - - - - - cdd939e7 by Jade at 2024-02-19T20:36:46-05:00 Enhance documentation of Data.Complex - - - - - d04f384f by Ben Gamari at 2024-02-21T04:59:23-05:00 hadrian/bindist: Ensure that phony rules are marked as such Otherwise make may not run the rule if file with the same name as the rule happens to exist. - - - - - efcbad2d by Ben Gamari at 2024-02-21T04:59:23-05:00 hadrian: Generate HSC2HS_EXTRAS variable in bindist installation We must generate the hsc2hs wrapper at bindist installation time since it must contain `--lflag` and `--cflag` arguments which depend upon the installation path. The solution here is to substitute these variables in the configure script (see mk/hsc2hs.in). This is then copied over a dummy wrapper in the install rules. Fixes #24050. - - - - - c540559c by Matthew Pickering at 2024-02-21T04:59:23-05:00 ci: Show --info for installed compiler - - - - - ab9281a2 by Matthew Pickering at 2024-02-21T04:59:23-05:00 configure: Correctly set --target flag for linker opts Previously we were trying to use the FP_CC_SUPPORTS_TARGET with 4 arguments, when it only takes 3 arguments. Instead we need to use the `FP_PROG_CC_LINKER_TARGET` function in order to set the linker flags. Actually fixes #24414 - - - - - 9460d504 by Rodrigo Mesquita at 2024-02-21T04:59:59-05:00 configure: Do not override existing linker flags in FP_LD_NO_FIXUP_CHAINS - - - - - 77629e76 by Andrei Borzenkov at 2024-02-21T05:00:35-05:00 Namespacing for fixity signatures (#14032) Namespace specifiers were added to syntax of fixity signatures: - sigdecl ::= infix prec ops | ... + sigdecl ::= infix prec namespace_spec ops | ... To preserve namespace during renaming MiniFixityEnv type now has separate FastStringEnv fields for names that should be on the term level and for name that should be on the type level. makeMiniFixityEnv function was changed to fill MiniFixityEnv in the right way: - signatures without namespace specifiers fill both fields - signatures with 'data' specifier fill data field only - signatures with 'type' specifier fill type field only Was added helper function lookupMiniFixityEnv that takes care about looking for a name in an appropriate namespace. Updates haddock submodule. Metric Decrease: MultiLayerModulesTH_OneShot - - - - - 84357d11 by Teo Camarasu at 2024-02-21T05:01:11-05:00 rts: only collect live words in nonmoving census when non-concurrent This avoids segfaults when the mutator modifies closures as we examine them. Resolves #24393 - - - - - 9ca56dd3 by Ian-Woo Kim at 2024-02-21T05:01:53-05:00 mutex wrap in refreshProfilingCCSs - - - - - 1387966a by Cheng Shao at 2024-02-21T05:02:32-05:00 rts: remove unused HAVE_C11_ATOMICS macro This commit removes the unused HAVE_C11_ATOMICS macro. We used to have a few places that have fallback paths when HAVE_C11_ATOMICS is not defined, but that is completely redundant, since the FP_CC_SUPPORTS__ATOMICS configure check will fail when the C compiler doesn't support C11 style atomics. There are also many places (e.g. in unreg backend, SMP.h, library cbits, etc) where we unconditionally use C11 style atomics anyway which work in even CentOS 7 (gcc 4.8), the oldest distro we test in our CI, so there's no value in keeping HAVE_C11_ATOMICS. - - - - - 0f40d68f by Andreas Klebinger at 2024-02-21T05:03:09-05:00 RTS: -Ds - make sure incall is non-zero before dereferencing it. Fixes #24445 - - - - - e5886de5 by Ben Gamari at 2024-02-21T05:03:44-05:00 rts/AdjustorPool: Use ExecPage abstraction This is just a minor cleanup I found while reviewing the implementation. - - - - - 826c5b47 by Torsten Schmits at 2024-02-21T13:17:05+01:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 09941666 by Adam Gundry at 2024-02-21T13:53:12+00:00 Define GHC2024 language edition (#24320) See https://github.com/ghc-proposals/ghc-proposals/pull/613. Also fixes #24343 and improves the documentation of language editions. Co-authored-by: Joachim Breitner <mail at joachim-breitner.de> - - - - - 2cff14d5 by Ben Gamari at 2024-02-22T09:35:56-05:00 Bump bounds - - - - - f49376b3 by Ben Gamari at 2024-02-22T09:35:56-05:00 Allow `@since` annotations in export lists Here we extend Haddock to admit `@since` annotations in export lists. These can be attached to most export list items (although not subordinate lists). These annotations supercede the declaration's `@since` annotation in produced Haddocks. - - - - - b5aa93df by Ben Gamari at 2024-02-22T12:09:06-05:00 Allow package-qualified @since declarations - - - - - 8f5957f2 by Ben Gamari at 2024-02-22T13:55:19-05:00 Documentation changes from ghc-internal restructuring Previously many declarations (e.g. `Int`) were declared to have a "home" in `Prelude`. However, now Haddock instead chooses to put these in more specific homes (e.g. `Data.Int`). Given that the "home" decision is driven by heuristics and in general these changes seem quite reasonable I am accepting them: * `Int` moved from `Prelude` to `Data.Int` * `(~)` moved from `Prelude` to `Data.Type.Equality` * `Type` moved from `GHC.Types` to `Data.Kind` * `Maybe` moved from `Prelude` to `Data.Maybe` * `Bool` moved from `Prelude` to `Data.Bool` * `Ordering` moved from `Prelude` to `Data.Ord` As well, more identifiers are now hyperlinked; it's not immediately clear *why*, but it is an improvement nevertheless. - - - - - ec33fec3 by Ben Gamari at 2024-02-22T20:36:24-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 5121a4ed by Ben Gamari at 2024-02-23T06:40:55-05:00 Allow docstrings after exports Here we extend the parser and AST to preserve docstrings following export items. We then extend Haddock to parse `@since` annotations in such docstrings, allowing changes in export structure to be properly documented. Bumps haddock submodule. - - - - - 30cfd251 by Torsten Schmits at 2024-02-24T13:00:42-05:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 0eb2265d by Hécate Moonlight at 2024-02-24T16:02:16-05:00 Improve the synopsis and description of base - - - - - 2e36f5d2 by Jade at 2024-02-24T16:02:51-05:00 Error Messages: Properly align cyclic module error Fixes: #24476 - - - - - bbfb051c by Ben Gamari at 2024-02-24T19:10:23-05:00 Allow docstrings after exports Here we extend the parser and AST to preserve docstrings following export items. We then extend Haddock to parse `@since` annotations in such docstrings, allowing changes in export structure to be properly documented. - - - - - d8d6ad8c by Ben Gamari at 2024-02-24T19:10:23-05:00 ghc-internal: Move modules into GHC.Internal.* namespace Bumps haddock submodule due to testsuite output changes. - - - - - a82af7cd by Ben Gamari at 2024-02-24T19:10:23-05:00 ghc-internal: Rewrite `@since ` to `@since base-` These will be incrementally moved to the export sites in `base` where possible. - - - - - ca3836e1 by Ben Gamari at 2024-02-24T19:10:23-05:00 base: Migrate Haddock `not-home` pragmas from `ghc-internal` This ensures that we do not use `base` stub modules as declarations' homes when not appropriate. - - - - - c8cf3e26 by Ben Gamari at 2024-02-24T19:10:23-05:00 base: Partially freeze exports of GHC.Base Sadly there are still a few module reexports. However, at least we have decoupled from the exports of `GHC.Internal.Base`. - - - - - 272573c6 by Ben Gamari at 2024-02-24T19:10:23-05:00 Move Haddock named chunks - - - - - 2d8a881d by Ben Gamari at 2024-02-24T19:10:23-05:00 Drop GHC.Internal.Data.Int - - - - - 55c4c385 by Ben Gamari at 2024-02-24T19:10:23-05:00 compiler: Fix mention to `GHC....` modules in wasm desugaring Really, these references should be via known-key names anyways. I have fixed the proximate issue here but have opened #24472 to track the additional needed refactoring. - - - - - 64150911 by Ben Gamari at 2024-02-24T19:10:23-05:00 Accept performance shifts from ghc-internal restructure As expected, Haddock now does more work. Less expected is that some other testcases actually get faster, presumably due to less interface file loading. As well, the size_hello_artifact test regressed a bit when debug information is enabled due to debug information for the new stub symbols. Metric Decrease: T12227 T13056 Metric Increase: haddock.Cabal haddock.base MultiLayerModulesTH_OneShot size_hello_artifact - - - - - 317a915b by Ben Gamari at 2024-02-24T19:10:23-05:00 Expose GHC.Wasm.Prim from ghc-experimental Previously this was only exposed from `ghc-internal` which violates our agreement that users shall not rely on things exposed from that package. Fixes #24479. - - - - - 3bbd2bf2 by Ben Gamari at 2024-02-24T19:10:23-05:00 compiler/tc: Small optimisation of evCallStack Don't lookupIds unless we actually need them. - - - - - 3e5c9e3c by Ben Gamari at 2024-02-24T19:10:23-05:00 compiler/tc: Use toException instead of SomeException - - - - - 125714a6 by Ben Gamari at 2024-02-24T19:10:23-05:00 base: Factor out errorBelch This was useful when debugging - - - - - 3d6aae7c by Ben Gamari at 2024-02-24T19:10:23-05:00 base: Clean up imports of GHC.Stack.CloneStack - - - - - 6900306e by Ben Gamari at 2024-02-24T19:10:24-05:00 base: Move PrimMVar to GHC.Internal.MVar - - - - - 28f8a148 by Ben Gamari at 2024-02-24T19:10:24-05:00 base: Move prettyCallStack to GHC.Internal.Stack - - - - - 4892de47 by Ben Gamari at 2024-02-24T19:10:24-05:00 base: Explicit dependency to workaround #24436 Currently `ghc -M` fails to account for `.hs-boot` files correctly, leading to issues with cross-package one-shot builds failing. This currently manifests in `GHC.Exception` due to the boot file for `GHC.Internal.Stack`. Work around this by adding an explicit `import`, ensuring that `GHC.Internal.Stack` is built before `GHC.Exception`. See #24436. - - - - - 294c93a5 by Ben Gamari at 2024-02-24T19:10:24-05:00 base: Use displayException in top-level exception handler Happily this also allows us to eliminate a special case for Deadlock exceptions. Implements [CLC #198](https://github.com/haskell/core-libraries-committee/issues/198). - - - - - 732db81d by Ben Gamari at 2024-02-24T19:12:18-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - cf756a25 by Ben Gamari at 2024-02-24T22:11:53-05:00 rts: Fix symbol references in Wasm RTS - - - - - 4e4d47a0 by Jade at 2024-02-26T15:17:20-05:00 GHCi: Improve response to unloading, loading and reloading modules Fixes #13869 - - - - - f3de8a3c by Zubin Duggal at 2024-02-26T15:17:57-05:00 rel-eng/fetch-gitlab.py: Fix name of aarch64 alpine 3_18 release job - - - - - c71bfdff by Cheng Shao at 2024-02-26T15:18:35-05:00 hadrian/hie-bios: pass -j to hadrian This commit passes -j to hadrian in the hadrian/hie-bios scripts. When the user starts HLS in a fresh clone that has just been configured, it takes quite a while for hie-bios to pick up the ghc flags and start actual indexing, due to the fact that the hadrian build step defaulted to -j1, so -j speeds things up and improve HLS user experience in GHC. Also add -j flag to .ghcid to speed up ghcid, and sets the Windows build root to .hie-bios which also works and unifies with other platforms, the previous build root _hie-bios was missing from .gitignore anyway. - - - - - 50bfdb46 by Cheng Shao at 2024-02-26T15:18:35-05:00 ci: enable parallelism in hadrian/ghci scripts This commit enables parallelism when the hadrian/ghci scripts are called in CI. The time bottleneck is in the hadrian build step, but previously the build step wasn't parallelized. - - - - - 61a78231 by Felix Yan at 2024-02-26T15:19:14-05:00 m4: Correctly detect GCC version When calling as `cc`, GCC does not outputs lowercased "gcc" at least in 13.2.1 version here. ``` $ cc --version cc (GCC) 13.2.1 20230801 ... ``` This fails the check and outputs the confusing message: `configure: $CC is not gcc; assuming it's a reasonably new C compiler` This patch makes it check for upper-cased "GCC" too so that it works correctly: ``` checking version of gcc... 13.2.1 ``` - - - - - 001aa539 by Teo Camarasu at 2024-02-27T13:26:46-05:00 Fix formatting in whereFrom docstring Previously it used markdown syntax rather than Haddock syntax for code quotes - - - - - e8034d15 by Teo Camarasu at 2024-02-27T13:26:46-05:00 Move ClosureType type to ghc-internal - Use ClosureType for InfoProv.ipDesc. - Use ClosureType for CloneStack.closureType. - Now ghc-heap re-exports this type from ghc-internal. See the accompanying CLC proposal: https://github.com/haskell/core-libraries-committee/issues/210 Resolves #22600 - - - - - 3da0a551 by Matthew Craven at 2024-02-27T13:27:22-05:00 StgToJS: Simplify ExprInline constructor of ExprResult Its payload was used only for a small optimization in genAlts, avoiding a few assignments for programs of this form: case NormalDataCon arg1 arg2 of x { NormalDataCon x1 x2 -> ... ; } But when compiling with optimizations, this sort of code is generally eliminated by case-of-known-constructor in Core-to-Core. So it doesn't seem worth tracking and cleaning up again in StgToJS. - - - - - 61bc92cc by Cheng Shao at 2024-02-27T16:58:42-05:00 rts: add missing ccs_mutex guard to internal_dlopen See added comment for details. Closes #24423. - - - - - dd29d3b2 by doyougnu at 2024-02-27T16:59:23-05:00 cg: Remove GHC.Cmm.DataFlow.Collections In pursuit of #15560 and #17957 and generally removing redundancy. - - - - - d3a050d2 by Cheng Shao at 2024-02-27T17:00:00-05:00 utils: remove unused lndir from tree Ever since the removal of the make build system, the in tree lndir hasn't been actually built, so this patch removes it. - - - - - 86bf7010 by Ben Gamari at 2024-02-27T19:28:10-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 74b24a9b by Teo Camarasu at 2024-02-28T16:32:58+00:00 rts: avoid checking bdescr of value outside of Haskell heap In nonmovingTidyWeaks we want to check if the key of a weak pointer lives in the non-moving heap. We do this by checking the flags of the block the key lives in. But we need to be careful with values that live outside the Haskell heap, since they will lack a block descriptor and looking for one may lead to a segfault. In this case we should just accept that it isn't on the non-moving heap. Resolves #24492 - - - - - b4cae4ec by Simon Peyton Jones at 2024-02-29T02:10:08-05:00 In mkDataConRep, ensure the in-scope set is right A small change that fixes #24489 - - - - - 3836a110 by Cheng Shao at 2024-02-29T21:25:45-05:00 testsuite: fix T23540 fragility on 32-bit platforms T23540 is fragile on 32-bit platforms. The root cause is usage of `getEvidenceTreesAtPoint`, which internally relies on `Name`'s `Ord` instance, which is indeterministic. The solution is adding a deterministic `Ord` instance for `EvidenceInfo` and sorting the evidence trees before pretty printing. Fixes #24449. - - - - - 960c8d47 by Teo Camarasu at 2024-02-29T21:26:20-05:00 Reduce AtomicModifyIORef increment count This test leads to a lot of contention when N>2 and becomes very slow. Let's reduce the amount of work we do to compensate. Resolves #24490 - - - - - 2e46c8ad by Matthew Pickering at 2024-03-01T05:48:06-05:00 hadrian: Improve parallelism in binary-dist-dir rule I noticed that the "docs" target was needed after the libraries and executables were built. We can improve the parallelism by needing everything at once so that documentation can be built immediately after a library is built for example. - - - - - cb6c11fe by Matthew Pickering at 2024-03-01T05:48:07-05:00 ci: Bump windows and freebsd boot compilers to 9.6.4 We have previously bumped the docker images to use 9.6.4, but neglected to bump the windows images until now. - - - - - 30f06996 by Matthew Pickering at 2024-03-01T05:48:07-05:00 ci: darwin: Update to 9.6.2 for boot compiler 9.6.4 is currently broken due to #24050 Also update to use LLVM-15 rather than LLVM-11, which is out of date. - - - - - d9d69e12 by Matthew Pickering at 2024-03-01T05:48:07-05:00 Bump minimum bootstrap version to 9.6 - - - - - 67ace1c5 by Matthew Pickering at 2024-03-01T05:48:07-05:00 ci: Enable more documentation building Here we enable documentation building on 1. Darwin: The sphinx toolchain was already installed so we enable html and manpages. 2. Rocky8: Full documentation (toolchain already installed) 3. Alpine: Full documetnation (toolchain already installed) 4. Windows: HTML and manpages (toolchain already installed) Fixes #24465 - - - - - 39583c39 by Matthew Pickering at 2024-03-01T05:48:42-05:00 ci: Bump ci-images to allow updated aarch64-alpine image with llvm15 and clang15 - - - - - d91d00fc by Torsten Schmits at 2024-03-01T15:01:50-05:00 Introduce ListTuplePuns extension This implements Proposal 0475, introducing the `ListTuplePuns` extension which is enabled by default. Disabling this extension makes it invalid to refer to list, tuple and sum type constructors by using built-in syntax like `[Int]`, `(Int, Int)`, `(# Int#, Int# #)` or `(# Int | Int #)`. Instead, this syntax exclusively denotes data constructors for use with `DataKinds`. The conventional way of referring to these data constructors by prefixing them with a single quote (`'(Int, Int)`) is now a parser error. Tuple declarations have been moved to `GHC.Tuple.Prim` and the `Solo` data constructor has been renamed to `MkSolo` (in a previous commit). Unboxed tuples and sums now have real source declarations in `GHC.Types`. Unit and solo types for tuples are now called `Unit`, `Unit#`, `Solo` and `Solo#`. Constraint tuples now have the unambiguous type constructors `CTuple<n>` as well as `CUnit` and `CSolo`, defined in `GHC.Classes` like before. A new parser construct has been added for the unboxed sum data constructor declarations. The type families `Tuple`, `Sum#` etc. that were intended to provide nicer syntax have been omitted from this change set due to inference problems, to be implemented at a later time. See the MR discussion for more info. Updates the submodule utils/haddock. Updates the cabal submodule due to new language extension. Metric Increase: haddock.base Metric Decrease: MultiLayerModulesTH_OneShot size_hello_artifact Proposal document: https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0475-tuple-syntax.rst Merge request: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/8820 Tracking ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/21294 - - - - - bbdb6286 by Sylvain Henry at 2024-03-01T15:01:50-05:00 JS linker: filter unboxed tuples - - - - - dec6d8d3 by Arnaud Spiwack at 2024-03-01T15:02:30-05:00 Improve error messages coming from non-linear patterns This enriched the `CtOrigin` for non-linear patterns to include data of the pattern that created the constraint (which can be quite useful if it occurs nested in a pattern) as well as an explanation why the pattern is non-restricted in (at least in some cases). - - - - - 6612388e by Arnaud Spiwack at 2024-03-01T15:02:30-05:00 Adjust documentation of linear lets according to committee decision - - - - - 1c064ef1 by Cheng Shao at 2024-03-02T17:11:19-05:00 compiler: start deprecating cmmToRawCmmHook cmmToRawCmmHook was added 4 years ago in d561c8f6244f8280a2483e8753c38e39d34c1f01. Its only user is the Asterius project, which has been archived and deprecated in favor of the ghc wasm backend. This patch starts deprecating cmmToRawCmmHook by placing a DEPRECATED pragma, and actual removal shall happen in a future GHC major release if no issue to oppose the deprecation has been raised in the meantime. - - - - - 9b74845f by Andrew Lelechenko at 2024-03-02T17:11:55-05:00 Data.List.NonEmpty.unzip: use WARNING with category instead of DEPRECATED CLC proposal: https://github.com/haskell/core-libraries-committee/issues/258 - - - - - 61bb5ff6 by Finley McIlwaine at 2024-03-04T09:01:40-08:00 add -fprof-late-overloaded and -fprof-late-overloaded-calls * Refactor late cost centre insertion for extensibility * Add two more late cost centre insertion methods that add SCCs to overloaded top level bindings and call sites with dictionary arguments. * Some tests for the basic functionality of the new insertion methods Resolves: #24500 - - - - - 82ccb801 by Andreas Klebinger at 2024-03-04T19:59:14-05:00 x86-ncg: Fix fma codegen when arguments are globals Fix a bug in the x86 ncg where results would be wrong when the desired output register and one of the input registers were the same global. Also adds a tiny optimization to make use of the memory addressing support when convenient. Fixes #24496 - - - - - 18ad1077 by Matthew Pickering at 2024-03-05T14:22:31-05:00 rel_eng: Update hackage docs upload scripts This adds the upload of ghc-internal and ghc-experimental to our scripts which upload packages to hackage. - - - - - bf47c9ba by Matthew Pickering at 2024-03-05T14:22:31-05:00 docs: Remove stray module comment from GHC.Profiling.Eras - - - - - 37d9b340 by Matthew Pickering at 2024-03-05T14:22:31-05:00 Fix ghc-internal cabal file The file mentioned some artifacts relating to the base library. I have renamed these to the new ghc-internal variants. - - - - - 23f2a478 by Matthew Pickering at 2024-03-05T14:22:31-05:00 Fix haddock source links and hyperlinked source There were a few issues with the hackage links: 1. We were using the package id rather than the package name for the package links. This is fixed by now allowing the template to mention %pkg% or %pkgid% and substituing both appropiatly. 2. The `--haddock-base-url` flag is renamed to `--haddock-for-hackage` as the new base link works on a local or remote hackage server. 3. The "src" path including too much stuff, so cross-package source links were broken as the template was getting double expanded. Fixes #24086 - - - - - 2fa336a9 by Ben Gamari at 2024-03-05T14:23:07-05:00 filepath: Bump submodule to 1.5.2.0 - - - - - 31217944 by Ben Gamari at 2024-03-05T14:23:07-05:00 os-string: Bump submodule to 2.0.2 - - - - - 4074a3f2 by Matthew Pickering at 2024-03-05T21:44:35-05:00 base: Reflect new era profiling RTS flags in GHC.RTS.Flags * -he profiling mode * -he profiling selector * --automatic-era-increment CLC proposal #254 - https://github.com/haskell/core-libraries-committee/issues/254 - - - - - a8c0e31b by Sylvain Henry at 2024-03-05T21:45:14-05:00 JS: faster implementation for some numeric primitives (#23597) Use faster implementations for the following primitives in the JS backend by not using JavaScript's BigInt: - plusInt64 - minusInt64 - minusWord64 - timesWord64 - timesInt64 Co-authored-by: Josh Meredith <joshmeredith2008 at gmail.com> - - - - - 21e3f325 by Cheng Shao at 2024-03-05T21:45:52-05:00 rts: add -xr option to control two step allocator reserved space size This patch adds a -xr RTS option to control the size of virtual memory address space reserved by the two step allocator on a 64-bit platform, see added documentation for explanation. Closes #24498. - - - - - dedcf102 by Cheng Shao at 2024-03-06T13:39:04-05:00 rts: expose HeapAlloc.h as public header This commit exposes HeapAlloc.h as a public header. The intention is to expose HEAP_ALLOCED/HEAP_ALLOCED_GC, so they can be used in assertions in other public headers, and they may also be useful for user code. - - - - - d19441d7 by Cheng Shao at 2024-03-06T13:39:04-05:00 rts: assert pointer is indeed heap allocated in Bdescr() This commit adds an assertion to Bdescr() to assert the pointer is indeed heap allocated. This is useful to rule out RTS bugs that attempt to access non-existent block descriptor of a static closure, #24492 being one such example. - - - - - 9a656a04 by Ben Gamari at 2024-03-06T13:39:39-05:00 ghc-experimental: Add dummy dependencies to work around #23942 This is a temporary measure to improve CI reliability until a proper solution is developed. Works around #23942. - - - - - 1e84b924 by Simon Peyton Jones at 2024-03-06T13:39:39-05:00 Three compile perf improvements with deep nesting These were changes are all triggered by #24471. 1. Make GHC.Core.Opt.SetLevels.lvlMFE behave better when there are many free variables. See Note [Large free-variable sets]. 2. Make GHC.Core.Opt.Arity.floatIn a bit lazier in its Cost argument. This benefits the common case where the ArityType turns out to be nullary. See Note [Care with nested expressions] 3. Make GHC.CoreToStg.Prep.cpeArg behave for deeply-nested expressions. See Note [Eta expansion of arguments in CorePrep] wrinkle (EA2). Compile times go down by up to 4.5%, and much more in artificial cases. (Geo mean of compiler/perf changes is -0.4%.) Metric Decrease: CoOpt_Read T10421 T12425 - - - - - c4b13113 by Hécate Moonlight at 2024-03-06T13:40:17-05:00 Use "module" instead of "library" when applicable in base haddocks - - - - - 9cd9efb4 by Vladislav Zavialov at 2024-03-07T13:01:54+03:00 Rephrase error message to say "visible arguments" (#24318) * Main change: make the error message generated by mkFunTysMsg more accurate by changing "value arguments" to "visible arguments". * Refactor: define a new type synonym VisArity and use it instead of Arity in a few places. It might be the case that there other places in the compiler that should talk about visible arguments rather than value arguments, but I haven't tried to find them all, focusing only on the error message reported in the ticket. - - - - - 4b6e76b5 by Patrick at 2024-03-07T22:09:30+08:00 fix haskell/haddock#24493, with module name introduced in hieAst The accompanies haddoc PR with GHC PR https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12153 Two things have been done: 1. Link is introduced to every `X` in `module X where`, since we introduce the module name to HieAst, 2. `{-# LINE 4 "hypsrc-test/src/PositionPragmas.hs" #-}` is added before the `module PositionPragmas where` in ` hypsrc-test/ref/src/PositionPragmas.html `.It ensures only a single hieAst for file `hypsrc-test/src/PositionPragmas.hs` is generated. - - - - - d523a6a7 by Ben Gamari at 2024-03-07T19:40:45-05:00 Bump array submodule - - - - - 7e55003c by Ben Gamari at 2024-03-07T19:40:45-05:00 Bump stm submodule - - - - - 32d337ef by Ben Gamari at 2024-03-07T19:40:45-05:00 Introduce exception context Here we introduce the `ExceptionContext` type and `ExceptionAnnotation` class, allowing dynamically-typed user-defined annotations to be attached to exceptions. CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/199 GHC Proposal: https://github.com/ghc-proposals/ghc-proposals/pull/330 - - - - - 39f3d922 by Ben Gamari at 2024-03-07T19:40:46-05:00 testsuite/interface-stability: Update documentation - - - - - fdea7ada by Ben Gamari at 2024-03-07T19:40:46-05:00 ghc-internal: comment formatting - - - - - 4fba42ef by Ben Gamari at 2024-03-07T19:40:46-05:00 compiler: Default and warn ExceptionContext constraints - - - - - 3886a205 by Ben Gamari at 2024-03-07T19:40:46-05:00 base: Introduce exception backtraces Here we introduce the `Backtraces` type and associated machinery for attaching these via `ExceptionContext`. These has a few compile-time regressions (`T15703` and `T9872d`) due to the additional dependencies in the exception machinery. As well, there is a surprisingly large regression in the `size_hello_artifact` test. This appears to be due to various `Integer` and `Read` bits now being reachable at link-time. I believe it should be possible to avoid this but I have accepted the change for now to get the feature merged. CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/199 GHC Proposal: https://github.com/ghc-proposals/ghc-proposals/pull/330 Metric Increase: T15703 T9872d size_hello_artifact - - - - - 18c5409f by Ben Gamari at 2024-03-07T19:40:46-05:00 users guide: Release notes for exception backtrace work - - - - - f849c5fc by Ben Gamari at 2024-03-07T19:40:46-05:00 compiler: Don't show ExceptionContext of GhcExceptions Most GhcExceptions are user-facing errors and therefore the ExceptionContext has little value. Ideally we would enable it in the DEBUG compiler but I am leaving this for future work. - - - - - dc646e6f by Sylvain Henry at 2024-03-07T19:40:46-05:00 Disable T9930fail for the JS target (cf #19174) - - - - - bfc09760 by Alan Zimmerman at 2024-03-07T19:41:22-05:00 Update showAstData to honour blanking of AnnParen Also tweak rendering of SrcSpan to remove extra blank line. - - - - - 50454a29 by Ben Gamari at 2024-03-08T03:32:42-05:00 ghc-internal: Eliminate GHC.Internal.Data.Kind This was simply reexporting things from `ghc-prim`. Instead reexport these directly from `Data.Kind`. Also add build ordering dependency to work around #23942. - - - - - 38a4b6ab by Ben Gamari at 2024-03-08T03:33:18-05:00 rts: Fix SET_HDR initialization of retainer set This fixes a regression in retainer set profiling introduced by b0293f78cb6acf2540389e22bdda420d0ab874da. Prior to that commit the heap traversal word would be initialized by `SET_HDR` using `LDV_RECORD_CREATE`. However, the commit added a `doingLDVProfiling` check in `LDV_RECORD_CREATE`, meaning that this initialization no longer happened. Given that this initialization was awkwardly indirectly anyways, I have fixed this by explicitly initializating the heap traversal word to `NULL` in `SET_PROF_HDR`. This is equivalent to the previous behavior, but much more direct. Fixes #24513. - - - - - 635abccc by Ben Gamari at 2024-03-08T17:09:06-05:00 Bump ghc version to 9.10 - - - - - 2859a637 by Ben Gamari at 2024-03-08T18:26:47-05:00 base: Use strerror_r instead of strerror As noted by #24344, `strerror` is not necessarily thread-safe. Thankfully, POSIX.1-2001 has long offered `strerror_r`, which is safe to use. Fixes #24344. CLC discussion: https://github.com/haskell/core-libraries-committee/issues/249 - - - - - 5b934048 by Ben Gamari at 2024-03-08T18:50:12-05:00 Bump base upper bound - - - - - b30d134e by Ben Gamari at 2024-03-08T18:50:44-05:00 Testsuite output update - - - - - edb9bf77 by Jade at 2024-03-09T03:39:38-05:00 Error messages: Improve Error messages for Data constructors in type signatures. This patch improves the error messages from invalid type signatures by trying to guess what the user did and suggesting an appropriate fix. Partially fixes: #17879 - - - - - cfb197e3 by Patrick at 2024-03-09T03:40:15-05:00 HieAst: add module name #24493 The main purpose of this is to tuck the module name `xxx` in `module xxx where` into the hieAst. It should fix #24493. The following have been done: 1. Renamed and update the `tcg_doc_hdr :: Maybe (LHsDoc GhcRn)` to `tcg_hdr_info :: (Maybe (LHsDoc GhcRn), Maybe (XRec GhcRn ModuleName))` To store the located module name information. 2. update the `RenamedSource` and `RenamedStuff` with extra `Maybe (XRec GhcRn ModuleName)` located module name information. 3. add test `testsuite/tests/hiefile/should_compile/T24493.hs` to ensure the module name is added and update several relevent tests. 4. accompanied submodule haddoc test update MR in https://gitlab.haskell.org/ghc/haddock/-/merge_requests/53 - - - - - 2341d81e by Vaibhav Sagar at 2024-03-09T03:40:54-05:00 GHC.Utils.Binary: fix a couple of typos - - - - - 5580e1bd by Ben Gamari at 2024-03-09T03:41:30-05:00 rts: Drop .wasm suffix from .prof file names This replicates the behavior on Windows, where `Hi.exe` will produce profiling output named `Hi.prof` instead of `Hi.exe.prof`. While in the area I also fixed the extension-stripping logic, which incorrectly rewrote `Hi.exefoo` to `Hi.foo`. Closes #24515. - - - - - 259495ee by Cheng Shao at 2024-03-09T03:41:30-05:00 testsuite: drop exe extension from .hp & .prof filenames See #24515 for details. - - - - - c477a8d2 by Ben Gamari at 2024-03-09T03:42:05-05:00 rts/linker: Enable GOT support on all platforms There is nothing platform-dependent about our GOT implementation and GOT support is needed by `T24171` on i386. - - - - - 2e592857 by Vladislav Zavialov at 2024-03-09T03:42:41-05:00 Drop outdated comment on TcRnIllformedTypePattern This should have been done in 0f0c53a501b but I missed it. - - - - - c554b4da by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/CloneStack: Bounds check array write - - - - - 15c590a5 by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/CloneStack: Don't expose helper functions in header - - - - - e831ce31 by Ben Gamari at 2024-03-09T09:39:20-05:00 base: Move internals of GHC.InfoProv into GHC.InfoProv.Types Such that we can add new helpers into GHC.InfoProv.Types without breakage. - - - - - 6948e24d by Ben Gamari at 2024-03-09T09:39:20-05:00 rts: Lazily decode IPE tables Previously we would eagerly allocate `InfoTableEnt`s for each info table registered in the info table provenance map. However, this costs considerable memory and initialization time. Instead we now lazily decode these tables. This allows us to use one-third the memory *and* opens the door to taking advantage of sharing opportunities within a module. This required considerable reworking since lookupIPE now must be passed its result buffer. - - - - - 9204a04e by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/IPE: Don't expose helper in header - - - - - 308926ff by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/IPE: Share module_name within a Node This allows us to shave a 64-bit word off of the packed IPE entry size. - - - - - bebdea05 by Ben Gamari at 2024-03-09T09:39:20-05:00 IPE: Expose unit ID in InfoTableProv Here we add the unit ID to the info table provenance structure. - - - - - 6519c9ad by Ben Gamari at 2024-03-09T09:39:35-05:00 rts: Refactor GHC.Stack.CloneStack.decode Don't allocate a Ptr constructor per frame. - - - - - ed0b69dc by Ben Gamari at 2024-03-09T09:39:35-05:00 base: Do not expose whereFrom# from GHC.Exts - - - - - 2b1faea9 by Vladislav Zavialov at 2024-03-09T17:38:21-05:00 docs: Update info on TypeAbstractions * Mention TypeAbstractions in 9.10.1-notes.rst * Set the status to "Experimental". * Add a "Since: GHC 9.x" comment to each section. - - - - - f8b88918 by Ben Gamari at 2024-03-09T21:21:46-05:00 ci-images: Bump Alpine image to bootstrap with 9.8.2 - - - - - 705e6927 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Mark T24171 as fragile due to #24512 I will fix this but not in time for 9.10.1-alpha1 - - - - - c74196e1 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Mark linker_unload_native as fragile In particular this fails on platforms without `dlinfo`. I plan to address this but not before 9.10.1-alpha1. - - - - - f4d87f7a by Ben Gamari at 2024-03-09T21:21:46-05:00 configure: Bump version to 9.10 - - - - - 88df9a5f by Ben Gamari at 2024-03-09T21:21:46-05:00 Bump transformers submodule to 0.6.1.1 - - - - - 8176d5e8 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Increase ulimit for T18623 1 MByte was just too tight and failed intermittently on some platforms (e.g. CentOS 7). Bumping the limit to 8 MByte should provide sufficient headroom. Fixes #23139. - - - - - c74b38a3 by Ben Gamari at 2024-03-09T21:21:46-05:00 base: Bump version to 4.20.0.0 - - - - - b2937fc3 by Ben Gamari at 2024-03-09T21:21:46-05:00 ghc-internal: Set initial version at 9.1001.0 This provides PVP compliance while maintaining a clear correspondence between GHC releases and `ghc-internal` versions. - - - - - 4ae7d868 by Ben Gamari at 2024-03-09T21:21:46-05:00 ghc-prim: Bump version to 0.11.0 - - - - - 50798dc6 by Ben Gamari at 2024-03-09T21:21:46-05:00 template-haskell: Bump version to 2.22.0.0 - - - - - 8564f976 by Ben Gamari at 2024-03-09T21:21:46-05:00 base-exports: Accommodate spurious whitespace changes in 32-bit output It appears that this was - - - - - 9d4f0e98 by Ben Gamari at 2024-03-09T21:21:46-05:00 users-guide: Move exception backtrace relnotes to 9.10 This was previously mistakenly added to the GHC 9.8 release notes. - - - - - 145eae60 by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/rel_eng: Fix name of Rocky8 artifact - - - - - 39c2a630 by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/rel_eng: Fix path of generate_jobs_metadata - - - - - aed034de by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/upload: Rework recompression The old `combine` approach was quite fragile due to use of filename globbing. Moreover, it didn't parallelize well. This refactoring makes the goal more obvious, parallelizes better, and is more robust. - - - - - 9bdf3586 by Ben Gamari at 2024-03-09T21:37:44-05:00 Merge branch 'ghc-9.10' into ghc-head - - - - - cec76981 by Ben Gamari at 2024-03-09T21:54:00-05:00 Bump GHC version to 9.11 - - - - - 4c59feb7 by Ben Gamari at 2024-03-09T22:15:01-05:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - dc207d06 by Ben Gamari at 2024-03-10T08:56:08-04:00 configure: Bump GHC version to 9.11 Bumps haddock submodule. - - - - - 8b2513e8 by Ben Gamari at 2024-03-11T01:20:03-04:00 rts/linker: Don't unload code when profiling is enabled The heap census may contain references (e.g. `Counter.identity`) to static data which must be available when the census is reported at the end of execution. Fixes #24512. - - - - - 7810b4c3 by Ben Gamari at 2024-03-11T01:20:03-04:00 rts/linker: Don't unload native objects when dlinfo isn't available To do so is unsafe as we have no way of identifying references to symbols provided by the object. Fixes #24513. Fixes #23993. - - - - - 0590764c by Ben Gamari at 2024-03-11T01:20:39-04:00 rel_eng/upload: Purge both $rel_name/ and $ver/ This is necessary for prereleases, where GHCup accesses the release via `$ver/` - - - - - b85a4631 by Brandon Chinn at 2024-03-12T19:25:56-04:00 Remove duplicate code normalising slashes - - - - - c91946f9 by Brandon Chinn at 2024-03-12T19:25:56-04:00 Simplify regexes with raw strings - - - - - 1a5f53c6 by Brandon Chinn at 2024-03-12T19:25:57-04:00 Don't normalize backslashes in characters - - - - - 7ea971d3 by Andrei Borzenkov at 2024-03-12T19:26:32-04:00 Fix compiler crash caused by implicit RHS quantification in type synonyms (#24470) - - - - - 39f3ac3e by Cheng Shao at 2024-03-12T19:27:11-04:00 Revert "compiler: make genSym use C-based atomic increment on non-JS 32-bit platforms" This reverts commit 615eb855416ce536e02ed935ecc5a6f25519ae16. It was originally intended to fix #24449, but it was merely sweeping the bug under the rug. 3836a110577b5c9343915fd96c1b2c64217e0082 has properly fixed the fragile test, and we no longer need the C version of genSym. Furthermore, the C implementation causes trouble when compiling with clang that targets i386 due to alignment warning and libatomic linking issue, so it makes sense to revert it. - - - - - e6bfb85c by Cheng Shao at 2024-03-12T19:27:11-04:00 compiler: fix out-of-bound memory access of genSym on 32-bit This commit fixes an unnoticed out-of-bound memory access of genSym on 32-bit. ghc_unique_inc is 32-bit sized/aligned on 32-bit platforms, but we mistakenly treat it as a Word64 pointer in genSym, and therefore will accidentally load 2 garbage higher bytes, or with a small but non-zero chance, overwrite something else in the data section depends on how the linker places the data segments. This regression was introduced in !11802 and fixed here. - - - - - 77171cd1 by Ben Orchard at 2024-03-14T09:00:40-04:00 Note mutability of array and address access primops Without an understanding of immutable vs. mutable memory, the index primop family have a potentially non-intuitive type signature: indexOffAddr :: Addr# -> Int# -> a readOffAddr :: Addr# -> Int# -> State# d -> (# State# d, a #) indexOffAddr# might seem like a free generality improvement, which it certainly is not! This change adds a brief note on mutability expectations for most index/read/write access primops. - - - - - 7da7f8f6 by Alan Zimmerman at 2024-03-14T09:01:15-04:00 EPA: Fix regression discarding comments in contexts Closes #24533 - - - - - 73be65ab by Fendor at 2024-03-19T01:42:53-04:00 Fix sharing of 'IfaceTyConInfo' during core to iface type translation During heap analysis, we noticed that during generation of 'mi_extra_decls' we have lots of duplicates for the instances: * `IfaceTyConInfo NotPromoted IfaceNormalTyCon` * `IfaceTyConInfo IsPromoted IfaceNormalTyCon` which should be shared instead of duplicated. This duplication increased the number of live bytes by around 200MB while loading the agda codebase into GHCi. These instances are created during `CoreToIface` translation, in particular `toIfaceTyCon`. The generated core looks like: toIfaceTyCon = \ tc_sjJw -> case $wtoIfaceTyCon tc_sjJw of { (# ww_sjJz, ww1_sjNL, ww2_sjNM #) -> IfaceTyCon ww_sjJz (IfaceTyConInfo ww1_sjNL ww2_sjNM) } whichs removes causes the sharing to work propery. Adding explicit sharing, with NOINLINE annotations, changes the core to: toIfaceTyCon = \ tc_sjJq -> case $wtoIfaceTyCon tc_sjJq of { (# ww_sjNB, ww1_sjNC #) -> IfaceTyCon ww_sjNB ww1_sjNC } which looks much more like sharing is happening. We confirmed via ghc-debug that all duplications were eliminated and the number of live bytes are noticeably reduced. - - - - - bd8209eb by Alan Zimmerman at 2024-03-19T01:43:28-04:00 EPA: Address more 9.10.1-alpha1 regressions from recent changes Closes #24533 Hopefully for good this time - - - - - 31bf85ee by Fendor at 2024-03-19T14:48:08-04:00 Escape multiple arguments in the settings file Uses responseFile syntax. The issue arises when GHC is installed on windows into a location that has a space, for example the user name is 'Fake User'. The $topdir will also contain a space, consequentially. When we resolve the top dir in the string `-I$topdir/mingw/include`, then `words` will turn this single argument into `-I/C/Users/Fake` and `User/.../mingw/include` which trips up the flag argument parser of various tools such as gcc or clang. We avoid this by escaping the $topdir before replacing it in `initSettngs`. Additionally, we allow to escape spaces and quotation marks for arguments in `settings` file. Add regression test case to count the number of options after variable expansion and argument escaping took place. Additionally, we check that escaped spaces and double quotation marks are correctly parsed. - - - - - f45f700e by Matthew Pickering at 2024-03-19T14:48:44-04:00 Read global package database from settings file Before this patch, the global package database was always assumed to be in libdir </> package.conf.d. This causes issues in GHC's build system because there are sometimes situations where the package database you need to use is not located in the same place as the settings file. * The stage1 compiler needs to use stage1 libraries, so we should set "Global Package DB" for the stage1 compiler to the stage1 package database. * Stage 2 cross compilers need to use stage2 libraries, so likewise, we should set the package database path to `_build/stage2/lib/` * The normal situation is where the stage2 compiler uses stage1 libraries. Then everything lines up. * When installing we have rearranged everything so that the settings file and package database line up properly, so then everything should continue to work as before. In this case we set the relative package db path to `package.conf.d`, so it resolves the same as before. * ghc-pkg needs to be modified as well to look in the settings file fo the package database rather than assuming the global package database location relative to the lib folder. * Cabal/cabal-install will work correctly because they query the global package database using `--print-global-package-db`. A reasonable question is why not generate the "right" settings files in the right places in GHC's build system. In order to do this you would need to engineer wrappers for all executables to point to a specific libdir. There are also situations where the same package db is used by two different compilers with two different settings files (think stage2 cross compiler and stage3 compiler). In short, this 10 line patch allows for some reasonable simplifications in Hadrian at very little cost to anything else. Fixes #24502 - - - - - 4c8f1794 by Matthew Pickering at 2024-03-19T14:48:44-04:00 hadrian: Remove stage1 testsuite wrappers logic Now instead of producing wrappers which pass the global package database argument to ghc and ghc-pkg, we write the location of the correct package database into the settings file so you can just use the intree compiler directly. - - - - - da0d8ba5 by Matthew Craven at 2024-03-19T14:49:20-04:00 Remove unused ghc-internal module "GHC.Internal.Constants" - - - - - b56d2761 by Matthew Craven at 2024-03-19T14:49:20-04:00 CorePrep: Rework lowering of BigNat# literals Don't use bigNatFromWord#, because that's terrible: * We shouldn't have to traverse a linked list at run-time to build a BigNat# literal. That's just silly! * The static List object we have to create is much larger than the actual BigNat#'s contents, bloating code size. * We have to read the corresponding interface file, which causes un-tracked implicit dependencies. (#23942) Instead, encode them into the appropriate platform-dependent sequence of bytes, and generate code that copies these bytes at run-time from an Addr# literal into a new ByteArray#. A ByteArray# literal would be the correct thing to generate, but these are not yet supported; see also #17747. Somewhat surprisingly, this change results in a slight reduction in compiler allocations, averaging around 0.5% on ghc's compiler performance tests, including when compiling programs that contain no bignum literals to begin with. The specific cause of this has not been investigated. Since this lowering no longer reads the interface file for GHC.Num.BigNat, the reasoning in Note [Depend on GHC.Num.Integer] is obsoleted. But the story of un-tracked built-in dependencies remains complex, and Note [Tracking dependencies on primitives] now exists to explain this complexity. Additionally, many empty imports have been modified to refer to this new note and comply with its guidance. Several empty imports necessary for other reasons have also been given brief explanations. Metric Decrease: MultiLayerModulesTH_OneShot - - - - - 349ea330 by Fendor at 2024-03-19T14:50:00-04:00 Eliminate thunk in 'IfaceTyCon' Heap analysis showed that `IfaceTyCon` retains a thunk to `IfaceTyConInfo`, defeating the sharing of the most common instances of `IfaceTyConInfo`. We make sure the indirection is removed by adding bang patterns to `IfaceTyCon`. Experimental results on the agda code base, where the `mi_extra_decls` were read from disk: Before this change, we observe around 8654045 instances of: `IfaceTyCon[Name,THUNK_1_0]` But these thunks almost exclusively point to a shared value! Forcing the thunk a little bit more, leads to `ghc-debug` reporting: `IfaceTyCon[Name:Name,IfaceTyConInfo]` and a noticeable reduction of live bytes (on agda ~10%). - - - - - 594bee0b by Krzysztof Gogolewski at 2024-03-19T14:50:36-04:00 Minor misc cleanups - GHC.HsToCore.Foreign.JavaScript: remove dropRuntimeRepArgs; boxed tuples don't take RuntimeRep args - GHC.HsToCore.Foreign.Call: avoid partial pattern matching - GHC.Stg.Unarise: strengthen the assertion; we can assert that non-rubbish literals are unary rather than just non-void - GHC.Tc.Gen.HsType: make sure the fsLit "literal" rule fires - users_guide/using-warnings.rst: remove -Wforall-identifier, now deprecated and does nothing - users_guide/using.rst: fix formatting - andy_cherry/test.T: remove expect_broken_for(23272...), 23272 is fixed The rest are simple cleanups. - - - - - cf55a54b by Ben Gamari at 2024-03-19T14:51:12-04:00 mk/relpath: Fix quoting Previously there were two instances in this script which lacked proper quoting. This resulted in `relpath` invocations in the binary distribution Makefile producing incorrect results on Windows, leading to confusing failures from `sed` and the production of empty package registrations. Fixes #24538. - - - - - 5ff88389 by Bryan Richter at 2024-03-19T14:51:48-04:00 testsuite: Disable T21336a on wasm - - - - - 60023351 by Ben Gamari at 2024-03-19T22:33:10-04:00 hadrian/bindist: Eliminate extraneous `dirname` invocation Previously we would call `dirname` twice per installed library file. We now instead reuse this result. This helps appreciably on Windows, where processes are quite expensive. - - - - - 616ac300 by Ben Gamari at 2024-03-19T22:33:10-04:00 hadrian: Package mingw toolchain in expected location This fixes #24525, a regression due to 41cbaf44a6ab5eb9fa676d65d32df8377898dc89. Specifically, GHC expects to find the mingw32 toolchain in the binary distribution root. However, after this patch it was packaged in the `lib/` directory. - - - - - de9daade by Ben Gamari at 2024-03-19T22:33:11-04:00 gitlab/rel_eng: More upload.sh tweaks - - - - - 1dfe12db by Ben Gamari at 2024-03-19T22:33:11-04:00 rel_eng: Drop dead prepare_docs codepath - - - - - dd2d748b by Ben Gamari at 2024-03-19T22:33:11-04:00 rel_env/recompress_all: unxz before recompressing Previously we would rather compress the xz *again*, before in addition compressing it with the desired scheme. Fixes #24545. - - - - - 9d936c57 by Ben Gamari at 2024-03-19T22:33:11-04:00 mk-ghcup-metadata: Fix directory of testsuite tarball As reported in #24546, the `dlTest` artifact should be extracted into the `testsuite` directory. - - - - - 6d398066 by Ben Gamari at 2024-03-19T22:33:11-04:00 ghcup-metadata: Don't populate dlOutput unless necessary ghcup can apparently infer the output name of an artifact from its URL. Consequently, we should only include the `dlOutput` field when it would differ from the filename of `dlUri`. Fixes #24547. - - - - - 576f8b7e by Zubin Duggal at 2024-03-19T22:33:46-04:00 Revert "Apply shellcheck suggestion to SUBST_TOOLDIR" This reverts commit c82770f57977a2b5add6e1378f234f8dd6153392. The shellcheck suggestion is spurious and results in SUBST_TOOLDIR being a no-op. `set` sets positional arguments for bash, but we want to set the variable given as the first autoconf argument. Fixes #24542 Metric decreases because the paths in the settings file are now shorter, so we allocate less when we read the settings file. ------------------------- Metric Decrease: T12425 T13035 T9198 ------------------------- - - - - - cdfe6e01 by Fendor at 2024-03-19T22:34:22-04:00 Compact serialisation of IfaceAppArgs In #24563, we identified that IfaceAppArgs serialisation tags each cons cell element with a discriminator byte. These bytes add up quickly, blowing up interface files considerably when '-fwrite-if-simplified-core' is enabled. We compact the serialisation by writing out the length of 'IfaceAppArgs', followed by serialising the elements directly without any discriminator byte. This improvement can decrease the size of some interface files by up to 35%. - - - - - 97a2bb1c by Simon Peyton Jones at 2024-03-20T17:11:29+00:00 Expand untyped splices in tcPolyExprCheck Fixes #24559 - - - - - 5f275176 by Alan Zimmerman at 2024-03-20T22:44:12-04:00 EPA: Clean up Exactprint helper functions a bit - Introduce a helper lens to compose on `EpAnn a` vs `a` versions - Rename some prime versions of functions back to non-prime They were renamed during the rework - - - - - da2a10ce by Vladislav Zavialov at 2024-03-20T22:44:48-04:00 Type operators in promoteOccName (#24570) Type operators differ from term operators in that they are lexically classified as (type) constructors, not as (type) variables. Prior to this change, promoteOccName did not account for this difference, causing a scoping issue that affected RequiredTypeArguments. type (!@#) = Bool f = idee (!@#) -- Not in scope: ‘!@#’ (BUG) Now we have a special case in promoteOccName to account for this. - - - - - 247fc0fa by Preetham Gujjula at 2024-03-21T10:19:18-04:00 docs: Remove mention of non-existent Ord instance for Complex The documentation for Data.Complex says that the Ord instance for Complex Float is deficient, but there is no Ord instance for Complex a. The Eq instance for Complex Float is similarly deficient, so we use that as an example instead. - - - - - 6fafc51e by Andrei Borzenkov at 2024-03-21T10:19:54-04:00 Fix TH handling in `pat_to_type_pat` function (#24571) There was missing case for `SplicePat` in `pat_to_type_at` function, hence patterns with splicing that checked against `forall->` doesn't work properly because they fall into the "illegal pattern" case. Code example that is now accepted: g :: forall a -> () g $([p| a |]) = () - - - - - 52072f8e by Sylvain Henry at 2024-03-21T21:01:59-04:00 Type-check default declarations before deriving clauses (#24566) See added Note and #24566. Default declarations must be type-checked before deriving clauses. - - - - - 7dfdf3d9 by Sylvain Henry at 2024-03-21T21:02:40-04:00 Lexer: small perf changes - Use unsafeChr because we know our values to be valid - Remove some unnecessary use of `ord` (return Word8 values directly) - - - - - 864922ef by Sylvain Henry at 2024-03-21T21:02:40-04:00 JS: fix some comments - - - - - 3e0b2b1f by Sebastian Graf at 2024-03-21T21:03:16-04:00 Simplifier: Re-do dependency analysis in abstractFloats (#24551) In #24551, we abstracted a string literal binding over a type variable, triggering a CoreLint error when that binding floated to top-level. The solution implemented in this patch fixes this by re-doing dependency analysis on a simplified recursive let binding that is about to be type abstracted, in order to find the minimal set of type variables to abstract over. See wrinkle (AB5) of Note [Floating and type abstraction] for more details. Fixes #24551 - - - - - 8a8ac65a by Matthew Craven at 2024-03-23T00:20:52-04:00 Improve toInteger @Word32 on 64-bit platforms On 64-bit platforms, every Word32 fits in an Int, so we can convert to Int# without having to perform the overflow check integerFromWord# uses internally. - - - - - 0c48f2b9 by Apoorv Ingle at 2024-03-23T00:21:28-04:00 Fix for #24552 (see testcase T24552) Fixes for a bug in desugaring pattern synonyms matches, introduced while working on on expanding `do`-blocks in #18324 The `matchWrapper` unecessarily (and incorrectly) filtered out the default wild patterns in a match. Now the wild pattern alternative is simply ignored by the pm check as its origin is `Generated`. The current code now matches the expected semantics according to the language spec. - - - - - b72705e9 by Simon Peyton Jones at 2024-03-23T00:22:04-04:00 Print more info about kinds in error messages This fixes #24553, where GHC unhelpfully said error: [GHC-83865] • Expected kind ‘* -> * -> *’, but ‘Foo’ has kind ‘* -> * -> *’ See Note [Showing invisible bits of types in error messages] - - - - - 8f7cfc7e by Tristan Cacqueray at 2024-03-23T00:22:44-04:00 docs: remove the don't use float hint This hint is outdated, ``Complex Float`` are now specialised, and the heap space suggestion needs more nuance so it should be explained in the unboxed/storable array documentation. - - - - - 5bd8ed53 by Andreas Klebinger at 2024-03-23T16:18:33-04:00 NCG: Fix a bug in jump shortcutting. When checking if a jump has more than one destination account for the possibility of some jumps not being representable by a BlockId. We do so by having isJumpishInstr return a `Maybe BlockId` where Nothing represents non-BlockId jump destinations. Fixes #24507 - - - - - 8d67f247 by Ben Gamari at 2024-03-23T16:19:09-04:00 docs: Drop old release notes, add for 9.12.1 - - - - - 7db8c992 by Cheng Shao at 2024-03-25T13:45:46-04:00 rts: fix clang compilation on aarch64 This patch fixes function prototypes in ARMOutlineAtomicsSymbols.h which causes "error: address argument to atomic operation must be a pointer to _Atomic type" when compiling with clang on aarch64. - - - - - 237194ce by Sylvain Henry at 2024-03-25T13:46:27-04:00 Lexer: fix imports for Alex 3.5.1 (#24583) - - - - - 810660b7 by Cheng Shao at 2024-03-25T22:19:16-04:00 libffi-tarballs: bump libffi-tarballs submodule to libffi 3.4.6 This commit bumps the libffi-tarballs submodule to libffi 3.4.6, which includes numerous upstream libffi fixes, especially https://github.com/libffi/libffi/issues/760. - - - - - d2ba41e8 by Alan Zimmerman at 2024-03-25T22:19:51-04:00 EPA: do not duplicate comments in signature RHS - - - - - 32a8103f by Rodrigo Mesquita at 2024-03-26T21:16:12-04:00 configure: Use LDFLAGS when trying linkers A user may configure `LDFLAGS` but not `LD`. When choosing a linker, we will prefer `ldd`, then `ld.gold`, then `ld.bfd` -- however, we have to check for a working linker. If either of these fail, we try the next in line. However, we were not considering the `$LDFLAGS` when checking if these linkers worked. So we would pick a linker that does not support the current $LDFLAGS and fail further down the line when we used that linker with those flags. Fixes #24565, where `LDFLAGS=-Wl,-z,pack-relative-relocs` is not supported by `ld.gold` but that was being picked still. - - - - - bf65a7c3 by Rodrigo Mesquita at 2024-03-26T21:16:48-04:00 bindist: Clean xattrs of bin and lib at configure time For issue #21506, we started cleaning the extended attributes of binaries and libraries from the bindist *after* they were installed to workaround notarisation (#17418), as part of `make install`. However, the `ghc-toolchain` binary that is now shipped with the bindist must be run at `./configure` time. Since we only cleaned the xattributes of the binaries and libs after they were installed, in some situations users would be unable to run `ghc-toolchain` from the bindist, failing at configure time (#24554). In this commit we move the xattr cleaning logic to the configure script. Fixes #24554 - - - - - cfeb70d3 by Rodrigo Mesquita at 2024-03-26T21:17:24-04:00 Revert "NCG: Fix a bug in jump shortcutting." This reverts commit 5bd8ed53dcefe10b72acb5729789e19ceb22df66. Fixes #24586 - - - - - 13223f6d by Serge S. Gulin at 2024-03-27T07:28:51-04:00 JS: `h$rts_isProfiled` is removed from `profiling` and left its version at `rts/js/config.js` - - - - - 0acfe391 by Alan Zimmerman at 2024-03-27T07:29:27-04:00 EPA: Do not extend declaration range for trailine zero len semi The lexer inserts virtual semicolons having zero width. Do not use them to extend the list span of items in a list. - - - - - cd0fb82f by Alan Zimmerman at 2024-03-27T19:33:08+00:00 EPA: Fix FamDecl range The span was incorrect if opt_datafam_kind_sig was empty - - - - - f8f384a8 by Ben Gamari at 2024-03-29T01:23:03-04:00 Fix type of _get_osfhandle foreign import Fixes #24601. - - - - - 00d3ecf0 by Alan Zimmerman at 2024-03-29T12:19:10+00:00 EPA: Extend StringLiteral range to include trailing commas This goes slightly against the exact printing philosophy where trailing decorations should be in an annotation, but the practicalities of adding it to the WarningTxt environment, and the problems caused by deviating do not make a more principles approach worthwhile. - - - - - efab3649 by brandon s allbery kf8nh at 2024-03-31T20:04:01-04:00 clarify Note [Preproccesing invocations] - - - - - c8a4c050 by Ben Gamari at 2024-04-02T12:50:35-04:00 rts: Fix TSAN_ENABLED CPP guard This should be `#if defined(TSAN_ENABLED)`, not `#if TSAN_ENABLED`, lest we suffer warnings. - - - - - e91dad93 by Cheng Shao at 2024-04-02T12:50:35-04:00 rts: fix errors when compiling with TSAN This commit fixes rts compilation errors when compiling with TSAN: - xxx_FENCE macros are redefined and trigger CPP warnings. - Use SIZEOF_W. WORD_SIZE_IN_BITS is provided by MachDeps.h which Cmm.h doesn't include by default. - - - - - a9ab9455 by Cheng Shao at 2024-04-02T12:50:35-04:00 rts: fix clang-specific errors when compiling with TSAN This commit fixes clang-specific rts compilation errors when compiling with TSAN: - clang doesn't have -Wtsan flag - Fix prototype of ghc_tsan_* helper functions - __tsan_atomic_* functions aren't clang built-ins and sanitizer/tsan_interface_atomic.h needs to be included - On macOS, TSAN runtime library is libclang_rt.tsan_osx_dynamic.dylib, not libtsan. -fsanitize-thread as a link-time flag will take care of linking the TSAN runtime library anyway so remove tsan as an rts extra library - - - - - 865bd717 by Cheng Shao at 2024-04-02T12:50:35-04:00 compiler: fix github link to __tsan_memory_order in a comment - - - - - 07cb627c by Cheng Shao at 2024-04-02T12:50:35-04:00 ci: improve TSAN CI jobs - Run TSAN jobs with +thread_sanitizer_cmm which enables Cmm instrumentation as well. - Run TSAN jobs in deb12 which ships gcc-12, a reasonably recent gcc that @bgamari confirms he's using in #GHC:matrix.org. Ideally we should be using latest clang release for latest improvements in sanitizers, though that's left as future work. - Mark TSAN jobs as manual+allow_failure in validate pipelines. The purpose is to demonstrate that we have indeed at least fixed building of TSAN mode in CI without blocking the patch to land, and once merged other people can begin playing with TSAN using their own dev setups and feature branches. - - - - - a1c18c7b by Andrei Borzenkov at 2024-04-02T12:51:11-04:00 Merge tc_infer_hs_type and tc_hs_type into one function using ExpType philosophy (#24299, #23639) This patch implements refactoring which is a prerequisite to updating kind checking of type patterns. This is a huge simplification of the main worker that checks kind of HsType. It also fixes the issues caused by previous code duplication, e.g. that we didn't add module finalizers from splices in inference mode. - - - - - 817e8936 by Rodrigo Mesquita at 2024-04-02T20:13:05-04:00 th: Hide the Language.Haskell.TH.Lib.Internal module from haddock Fixes #24562 - - - - - b36ee57b by Sylvain Henry at 2024-04-02T20:13:46-04:00 JS: reenable h$appendToHsString optimization (#24495) The optimization introducing h$appendToHsString wasn't kicking in anymore (while it did in 9.8.1) because of the changes introduced in #23270 (7e0c8b3bab30). This patch reenables the optimization by matching on case-expression, as done in Cmm for unpackCString# standard thunks. The test is also T24495 added in the next commits (two commits for ease of backporting to 9.8). - - - - - 527616e9 by Sylvain Henry at 2024-04-02T20:13:46-04:00 JS: fix h$appendToHsString implementation (#24495) h$appendToHsString needs to wrap its argument in an updatable thunk to behave like unpackAppendCString#. Otherwise if a SingleEntry thunk is passed, it is stored as-is in a CONS cell, making the resulting list impossible to deepseq (forcing the thunk doesn't update the contents of the CONS cell)! The added test checks that the optimization kicks in and that h$appendToHsString works as intended. Fix #24495 - - - - - faa30b41 by Simon Peyton Jones at 2024-04-02T20:14:22-04:00 Deal with duplicate tyvars in type declarations GHC was outright crashing before this fix: #24604 - - - - - e0b0c717 by Simon Peyton Jones at 2024-04-02T20:14:58-04:00 Try using MCoercion in exprIsConApp_maybe This is just a simple refactor that makes exprIsConApp_maybe a little bit more direct, simple, and efficient. Metrics: compile_time/bytes allocated geo. mean -0.1% minimum -2.0% maximum -0.0% Not a big gain, but worthwhile given that the code is, if anything, easier to grok. - - - - - 15f4d867 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Initial ./configure support for selecting I/O managers In this patch we just define new CPP vars, but don't yet use them or replace the existing approach. That will follow. The intention here is that every I/O manager can be enabled/disabled at GHC build time (subject to some constraints). More than one I/O manager can be enabled to be built. At least one I/O manager supporting the non-threaded RTS must be enabled as well as at least one supporting the non-threaded RTS. The I/O managers enabled here will become the choices available at runtime at RTS startup (in later patches). The choice can be made with RTS flags. There are separate sets of choices for the threaded and non-threaded RTS ways, because most I/O managers are specific to these ways. Furthermore we must establish a default I/O manager for the threaded and non-threaded RTS. Most I/O managers are platform-specific so there are checks to ensure each one can be enabled on the platform. Such checks are also where (in future) any system dependencies (e.g. libraries) can be checked. The output is a set of CPP flags (in the mk/config.h file), with one flag per named I/O manager: * IOMGR_BUILD_<name> : which ones should be built (some) * IOMGR_DEFAULT_NON_THREADED_<name> : which one is default (exactly one) * IOMGR_DEFAULT_THREADED_<name> : which one is default (exactly one) and a set of derived flags in IOManager.h * IOMGR_ENABLED_<name> : enabled for the current RTS way Note that IOMGR_BUILD_<name> just says that an I/O manager will be built for _some_ RTS way (i.e. threaded or non-threaded). The derived flags IOMGR_ENABLED_<name> in IOManager.h say if each I/O manager is enabled in the "current" RTS way. These are the ones that can be used for conditional compilation of the I/O manager code. Co-authored-by: Pi Delport <pi at well-typed.com> - - - - - 85b0f87a by Duncan Coutts at 2024-04-03T01:27:17-04:00 Change the handling of the RTS flag --io-manager= Now instead of it being just used on Windows to select between the WinIO vs the MIO or Win32-legacy I/O managers, it is now used on all platforms for selecting the I/O manager to use. Right now it remains the case that there is only an actual choice on Windows, but that will change later. Document the --io-manager flag in the user guide. This change is also reflected in the RTS flags types in the base library. Deprecate the export of IoSubSystem from GHC.RTS.Flags with a message to import it from GHC.IO.Subsystem. The way the 'IoSubSystem' is detected also changes. Instead of looking at the RTS flag, there is now a C bool global var in the RTS which gets set on startup when the I/O manager is selected. This bool var says whether the selected I/O manager classifies as "native" on Windows, which in practice means the WinIO I/O manager has been selected. Similarly, the is_io_mng_native_p RTS helper function is re-implemented in terms of the selected I/O manager, rather than based on the RTS flags. We do however remove the ./configure --native-io-manager flag because we're bringing the WinIO/MIO/Win32-legacy choice under the new general scheme for selecting I/O managers, and that new scheme involves no ./configure time user choices, just runtime RTS flag choices. - - - - - 1a8f020f by Duncan Coutts at 2024-04-03T01:27:17-04:00 Convert {init,stop,exit}IOManager to switch style Rather than ad-hoc cpp conitionals on THREADED_RTS and mingw32_HOST_OS, we use a style where we switch on the I/O manager impl, with cases for each I/O manager impl. - - - - - a5bad3d2 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Split up the CapIOManager content by I/O manager Using the new IOMGR_ENABLED_<name> CPP defines. - - - - - 1d36e609 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Convert initIOManagerAfterFork and wakeupIOManager to switch style - - - - - c2f26f36 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move most of waitRead#/Write# from cmm to C Moves it into the IOManager.c where we can follow the new pattern of switching on the selected I/O manager. - - - - - 457705a8 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move most of the delay# impl from cmm to C Moves it into the IOManager.c where we can follow the new pattern of switching on the selected I/O manager. Uses a new IOManager API: syncDelay, following the naming convention of sync* for thread-synchronous I/O & timer/delay operations. As part of porting from cmm to C, we maintain the rule that the why_blocked gets accessed using load acquire and store release atomic memory operations. There was one exception to this rule: in the delay# primop cmm code on posix (not win32), the why_blocked was being updated using a store relaxed, not a store release. I've no idea why. In this convesion I'm playing it safe here and using store release consistently. - - - - - e93058e0 by Duncan Coutts at 2024-04-03T01:27:18-04:00 insertIntoSleepingQueue is no longer public No longer defined in IOManager.h, just a private function in IOManager.c. Since it is no longer called from cmm code, just from syncDelay. It ought to get moved further into the select() I/O manager impl, rather than living in IOManager.c. On the other hand appendToIOBlockedQueue is still called from cmm code in the win32-legacy I/O manager primops async{Read,Write}#, and it is also used by the select() I/O manager. Update the CPP and comments to reflect this. - - - - - 60ce9910 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move anyPendingTimeoutsOrIO impl from .h to .c The implementation is eventually going to need to use more private things, which will drag in unwanted includes into IOManager.h, so it's better to move the impl out of the header file and into the .c file, at the slight cost of it no longer being inline. At the same time, change to the "switch (iomgr_type)" style. - - - - - f70b8108 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Take a simpler approach to gcc warnings in IOManager.c We have lots of functions with conditional implementations for different I/O managers. Some functions, for some I/O managers, naturally have implementations that do nothing or barf. When only one such I/O manager is enabled then the whole function implementation will have an implementation that does nothing or barfs. This then results in warnings from gcc that parameters are unused, or that the function should be marked with attribute noreturn (since barf does not return). The USED_IF_THREADS trick for fine-grained warning supression is fine for just two cases, but an equivalent here would need USED_IF_THE_ONLY_ENABLED_IOMGR_IS_X_OR_Y which would have combinitorial blowup. So we take a coarse grained approach and simply disable these two warnings for the whole file. So we use a GCC pragma, with its handy push/pop support: #pragma GCC diagnostic push #pragma GCC diagnostic ignored "-Wsuggest-attribute=noreturn" #pragma GCC diagnostic ignored "-Wunused-parameter" ... #pragma GCC diagnostic pop - - - - - b48805b9 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add a new trace class for the iomanager It makes sense now for it to be separate from the scheduler class of tracers. Enabled with +RTS -Do. Document the -Do debug flag in the user guide. - - - - - f0c1f862 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Have the throwTo impl go via (new) IOManager APIs rather than directly operating on the IO manager's data structures. Specifically, when thowing an async exception to a thread that is blocked waiting for I/O or waiting for a timer, then we want to cancel that I/O waiting or cancel the timer. Currently this is done directly in removeFromQueues() in RaiseAsync.c. We want it to go via proper APIs both for modularity but also to let us support multiple I/O managers. So add sync{IO,Delay}Cancel, which is the cancellation for the corresponding sync{IO,Delay}. The implementations of these use the usual "switch (iomgr_type)" style. - - - - - 4f9e9c4e by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move awaitEvent into a proper IOManager API and have the scheduler use it. Previously the scheduler calls awaitEvent directly, and awaitEvent is implemented directly in the RTS I/O managers (select, win32). This relies on the old scheme where there's a single active I/O manager for each platform and RTS way. We want to move that to go via an API in IOManager.{h,c} which can then call out to the active I/O manager. Also take the opportunity to split awaitEvent into two. The existing awaitEvent has a bool wait parameter, to say if the call should be blocking or non-blocking. We split this into two separate functions: pollCompletedTimeoutsOrIO and awaitCompletedTimeoutsOrIO. We split them for a few reasons: they have different post-conditions (specifically the await version is supposed to guarantee that there are threads runnable when it completes). Secondly, it is also anticipated that in future I/O managers the implementations of the two cases will be simpler if they are separated. - - - - - 5ad4b30f by Duncan Coutts at 2024-04-03T01:27:18-04:00 Rename awaitEvent in select and win32 I/O managers These are now just called from IOManager.c and are the per-I/O manager backend impls (whereas previously awaitEvent was the entry point). Follow the new naming convention in the IOManager.{h,c} of awaitCompletedTimeoutsOrIO, with the I/O manager's name as a suffix: so awaitCompletedTimeoutsOrIO{Select,Win32}. - - - - - d30c6bc6 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Tidy up a couple things in Select.{h,c} Use the standard #include {Begin,End}Private.h style rather than RTS_PRIVATE on individual decls. And conditionally build the code for the select I/O manager based on the new CPP IOMGR_ENABLED_SELECT rather than on THREADED_RTS. - - - - - 4161f516 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add an IOManager API for scavenging TSO blocked_info When the GC scavenges a TSO it needs to scavenge the tso->blocked_info but the blocked_info is a big union and what lives there depends on the two->why_blocked, which for I/O-related reasons is something that in principle is the responsibility of the I/O manager and not the GC. So the right thing to do is for the GC to ask the I/O manager to sscavenge the blocked_info if it encounters any I/O-related why_blocked reasons. So we add scavengeTSOIOManager in IOManager.{h,c} with the usual style. Now as it happens, right now, there is no special scavenging to do, so the implementation of scavengeTSOIOManager is a fancy no-op. That's because the select I/O manager uses only the fd and target members, which are not GC pointers, and the win32-legacy I/O manager _ought_ to be using GC-managed heap objects for the StgAsyncIOResult but it is actually usingthe C heap, so again no GC pointers. If the win32-legacy were doing this more sensibly, then scavengeTSOIOManager would be the right place to do the GC magic. Future I/O managers will need GC heap objects in the tso->blocked_info and will make use of this functionality. - - - - - 94a87d21 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add I/O manager API notifyIOManagerCapabilitiesChanged Used in setNumCapabilities. It only does anything for MIO on Posix. Previously it always invoked Haskell code, but that code only did anything on non-Windows (and non-JS), and only threaded. That currently effectively means the MIO I/O manager on Posix. So now it only invokes it for the MIO Posix case. - - - - - 3be6d591 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Select an I/O manager early in RTS startup We need to select the I/O manager to use during startup before the per-cap I/O manager initialisation. - - - - - aaa294d0 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Make struct CapIOManager be fully opaque Provide an opaque (forward) definition in Capability.h (since the cap contains a *CapIOManager) and then only provide a full definition in a new file IOManagerInternals.h. This new file is only supposed to be included by the IOManager implementation, not by its users. So that means IOManager.c and individual I/O manager implementations. The posix/Signals.c still needs direct access, but that should be eliminated. Anything that needs direct access either needs to be clearly part of an I/O manager (e.g. the sleect() one) or go via a proper API. - - - - - 877a2a80 by Duncan Coutts at 2024-04-03T01:27:18-04:00 The select() I/O manager does have some global initialisation It's just to make sure an exception CAF is a GC root. - - - - - 9c51473b by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add tracing for the main I/O manager actions Using the new tracer class. Note: The unconditional definition of showIOManager should be compatible with the debugTrace change in 7c7d1f6. Co-authored-by: Pi Delport <pi at well-typed.com> - - - - - c7d3e3a3 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Include the default I/O manager in the +RTS --info output Document the extra +RTS --info output in the user guide - - - - - 8023bad4 by Duncan Coutts at 2024-04-03T01:27:18-04:00 waitRead# / waitWrite# do not work for win32-legacy I/O manager Previously it was unclear that they did not work because the code path was shared with other I/O managers (in particular select()). Following the code carefully shows that what actually happens is that the calling thread would block forever: the thread will be put into the blocked queue, but no other action is scheduled that will ever result in it getting unblocked. It's better to just fail loudly in case anyone accidentally calls it, also it's less confusing code. - - - - - 83a74d20 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Conditionally ignore some GCC warnings Some GCC versions don't know about some warnings, and they complain that we're ignoring unknown warnings. So we try to ignore the warning based on the GCC version. - - - - - 1adc6fa4 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Accept changes to base-exports All the changes are in fact not changes at all. Previously, the IoSubSystem data type was defined in GHC.RTS.Flags and exported from both GHC.RTS.Flags and GHC.IO.SubSystem. Now, the data type is defined in GHC.IO.SubSystem and still exported from both modules. Therefore, the same exports and same instances are still available from both modules. But the base-exports records only the defining module, and so it looks like a change when it is fully compatible. Related: we do add a deprecation to the export of the type via GHC.RTS.Flags, telling people to use the export from GHC.IO.SubSystem. Also the sort order for some unrelated Show instances changed. No idea why. The same changes apply in the other versions, with a few more changes due to sort order weirdness. - - - - - 8d950968 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Accept metric decrease in T12227 I can't think of any good reason that anything in this MR should have changed the number of allocations, up or down. (Yes this is an empty commit.) Metric Decrease: T12227 - - - - - e869605e by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Several improvements to the handling of coercions * Make `mkSymCo` and `mkInstCo` smarter Fixes #23642 * Fix return role of `SelCo` in the coercion optimiser. Fixes #23617 * Make the coercion optimiser `opt_trans_rule` work better for newtypes Fixes #23619 - - - - - 1efd0714 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 FloatOut: improve floating for join point See the new Note [Floating join point bindings]. * Completely get rid of the complicated join_ceiling nonsense, which I have never understood. * Do not float join points at all, except perhaps to top level. * Some refactoring around wantToFloat, to treat Rec and NonRec more uniformly - - - - - 9c00154d by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Improve eta-expansion through call stacks See Note [Eta expanding through CallStacks] in GHC.Core.Opt.Arity This is a one-line change, that fixes an inconsistency - || isCallStackPredTy ty + || isCallStackPredTy ty || isCallStackTy ty - - - - - 95a9a172 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Spelling, layout, pretty-printing only - - - - - bdf1660f by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Improve exprIsConApp_maybe a little Eliminate a redundant case at birth. This sometimes reduces Simplifier iterations. See Note [Case elim in exprIsConApp_maybe]. - - - - - 609cd32c by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Inline GHC.HsToCore.Pmc.Solver.Types.trvVarInfo When exploring compile-time regressions after meddling with the Simplifier, I discovered that GHC.HsToCore.Pmc.Solver.Types.trvVarInfo was very delicately balanced. It's a small, heavily used, overloaded function and it's important that it inlines. By a fluke it was before, but at various times in my journey it stopped doing so. So I just added an INLINE pragma to it; no sense in depending on a delicately-balanced fluke. - - - - - ae24c9bc by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Slight improvement in WorkWrap Ensure that WorkWrap preserves lambda binders, in case of join points. Sadly I have forgotten why I made this change (it was while I was doing a lot of meddling in the Simplifier, but * it does no harm, * it is slightly more efficient, and * presumably it made something better! Anyway I have kept it in a separate commit. - - - - - e9297181 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Use named record fields for the CastIt { ... } data constructor This is a pure refactor - - - - - b4581e23 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Remove a long-commented-out line Pure refactoring - - - - - e026bdf2 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Simplifier improvements This MR started as: allow the simplifer to do more in one pass, arising from places I could see the simplifier taking two iterations where one would do. But it turned into a larger project, because these changes unexpectedly made inlining blow up, especially join points in deeply-nested cases. The main changes are below. There are also many new or rewritten Notes. Avoiding simplifying repeatedly ~~~~~~~~~~~~~~~ See Note [Avoiding simplifying repeatedly] * The SimplEnv now has a seInlineDepth field, which says how deep in unfoldings we are. See Note [Inline depth] in Simplify.Env. Currently used only for the next point: avoiding repeatedly simplifying coercions. * Avoid repeatedly simplifying coercions. see Note [Avoid re-simplifying coercions] in Simplify.Iteration As you'll see from the Note, this makes use of the seInlineDepth. * Allow Simplify.Iteration.simplAuxBind to inline used-once things. This is another part of Note [Post-inline for single-use things], and is really good for reducing simplifier iterations in situations like case K e of { K x -> blah } wher x is used once in blah. * Make GHC.Core.SimpleOpt.exprIsConApp_maybe do some simple case elimination. Note [Case elim in exprIsConApp_maybe] * Improve the case-merge transformation: - Move the main code to `GHC.Core.Utils.mergeCaseAlts`, to join `filterAlts` and friends. See Note [Merge Nested Cases] in GHC.Core.Utils. - Add a new case for `tagToEnum#`; see wrinkle (MC3). - Add a new case to look through join points: see wrinkle (MC4) postInlineUnconditionally ~~~~~~~~~~~~~~~~~~~~~~~~~ * Allow Simplify.Utils.postInlineUnconditionally to inline variables that are used exactly once. See Note [Post-inline for single-use things]. * Do not postInlineUnconditionally join point, ever. Doing so does not reduce allocation, which is the main point, and with join points that are used a lot it can bloat code. See point (1) of Note [Duplicating join points] in GHC.Core.Opt.Simplify.Iteration. * Do not postInlineUnconditionally a strict (demanded) binding. It will not allocate a thunk (it'll turn into a case instead) so again the main point of inlining it doesn't hold. Better to check per-call-site. * Improve occurrence analyis for bottoming function calls, to help postInlineUnconditionally. See Note [Bottoming function calls] in GHC.Core.Opt.OccurAnal Inlining generally ~~~~~~~~~~~~~~~~~~ * In GHC.Core.Opt.Simplify.Utils.interestingCallContext, use RhsCtxt NonRecursive (not BoringCtxt) for a plain-seq case. See Note [Seq is boring] Also, wrinkle (SB1), inline in that `seq` context only for INLINE functions (UnfWhen guidance). * In GHC.Core.Opt.Simplify.Utils.interestingArg, - return ValueArg for OtherCon [c1,c2, ...], but - return NonTrivArg for OtherCon [] This makes a function a little less likely to inline if all we know is that the argument is evaluated, but nothing else. * isConLikeUnfolding is no longer true for OtherCon {}. This propagates to exprIsConLike. Con-like-ness has /positive/ information. Join points ~~~~~~~~~~~ * Be very careful about inlining join points. See these two long Notes Note [Duplicating join points] in GHC.Core.Opt.Simplify.Iteration Note [Inlining join points] in GHC.Core.Opt.Simplify.Inline * When making join points, don't do so if the join point is so small it will immediately be inlined; check uncondInlineJoin. * In GHC.Core.Opt.Simplify.Inline.tryUnfolding, improve the inlining heuristics for join points. In general we /do not/ want to inline join points /even if they are small/. See Note [Duplicating join points] GHC.Core.Opt.Simplify.Iteration. But sometimes we do: see Note [Inlining join points] in GHC.Core.Opt.Simplify.Inline; and the new `isBetterUnfoldingThan` function. * Do not add an unfolding to a join point at birth. This is a tricky one and has a long Note [Do not add unfoldings to join points at birth] It shows up in two places - In `mkDupableAlt` do not add an inlining - (trickier) In `simplLetUnfolding` don't add an unfolding for a fresh join point I am not fully satisifed with this, but it works and is well documented. * In GHC.Core.Unfold.sizeExpr, make jumps small, so that we don't penalise having a non-inlined join point. Performance changes ~~~~~~~~~~~~~~~~~~~ * Binary sizes fall by around 2.6%, according to nofib. * Compile times improve slightly. Here are the figures over 1%. I investiate the biggest differnce in T18304. It's a very small module, just a few hundred nodes. The large percentage difffence is due to a single function that didn't quite inline before, and does now, making code size a bit bigger. I decided gains outweighed the losses. Metrics: compile_time/bytes allocated (changes over +/- 1%) ------------------------------------------------ CoOpt_Singletons(normal) -9.2% GOOD LargeRecord(normal) -23.5% GOOD MultiComponentModulesRecomp(normal) +1.2% MultiLayerModulesTH_OneShot(normal) +4.1% BAD PmSeriesS(normal) -3.8% PmSeriesV(normal) -1.5% T11195(normal) -1.3% T12227(normal) -20.4% GOOD T12545(normal) -3.2% T12707(normal) -2.1% GOOD T13253(normal) -1.2% T13253-spj(normal) +8.1% BAD T13386(normal) -3.1% GOOD T14766(normal) -2.6% GOOD T15164(normal) -1.4% T15304(normal) +1.2% T15630(normal) -8.2% T15630a(normal) NEW T15703(normal) -14.7% GOOD T16577(normal) -2.3% GOOD T17516(normal) -39.7% GOOD T18140(normal) +1.2% T18223(normal) -17.1% GOOD T18282(normal) -5.0% GOOD T18304(normal) +10.8% BAD T18923(normal) -2.9% GOOD T1969(normal) +1.0% T19695(normal) -1.5% T20049(normal) -12.7% GOOD T21839c(normal) -4.1% GOOD T3064(normal) -1.5% T3294(normal) +1.2% BAD T4801(normal) +1.2% T5030(normal) -15.2% GOOD T5321Fun(normal) -2.2% GOOD T6048(optasm) -16.8% GOOD T783(normal) -1.2% T8095(normal) -6.0% GOOD T9630(normal) -4.7% GOOD T9961(normal) +1.9% BAD WWRec(normal) -1.4% info_table_map_perf(normal) -1.3% parsing001(normal) +1.5% geo. mean -2.0% minimum -39.7% maximum +10.8% * Runtimes generally improve. In the testsuite perf/should_run gives: Metrics: runtime/bytes allocated ------------------------------------------ Conversions(normal) -0.3% T13536a(optasm) -41.7% GOOD T4830(normal) -0.1% haddock.Cabal(normal) -0.1% haddock.base(normal) -0.1% haddock.compiler(normal) -0.1% geo. mean -0.8% minimum -41.7% maximum +0.0% * For runtime, nofib is a better test. The news is mostly good. Here are the number more than +/- 0.1%: # bytes allocated ==========================++========== imaginary/digits-of-e1 || -14.40% imaginary/digits-of-e2 || -4.41% imaginary/paraffins || -0.17% imaginary/rfib || -0.15% imaginary/wheel-sieve2 || -0.10% real/compress || -0.47% real/fluid || -0.10% real/fulsom || +0.14% real/gamteb || -1.47% real/gg || -0.20% real/infer || +0.24% real/pic || -0.23% real/prolog || -0.36% real/scs || -0.46% real/smallpt || +4.03% shootout/k-nucleotide || -20.23% shootout/n-body || -0.42% shootout/spectral-norm || -0.13% spectral/boyer2 || -3.80% spectral/constraints || -0.27% spectral/hartel/ida || -0.82% spectral/mate || -20.34% spectral/para || +0.46% spectral/rewrite || +1.30% spectral/sphere || -0.14% ==========================++========== geom mean || -0.59% real/smallpt has a huge nest of local definitions, and I could not pin down a reason for a regression. But there are three big wins! Metric Decrease: CoOpt_Singletons LargeRecord T12227 T12707 T13386 T13536a T14766 T15703 T16577 T17516 T18223 T18282 T18923 T21839c T20049 T5321Fun T5030 T6048 T8095 T9630 T783 Metric Increase: MultiLayerModulesTH_OneShot T13253-spj T18304 T18698a T9961 T3294 - - - - - 27db3c5e by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Testsuite message changes from simplifier improvements - - - - - 271a7812 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Account for bottoming functions in OccurAnal This fixes #24582, a small but long-standing bug - - - - - 0fde229f by Ben Gamari at 2024-04-04T07:04:58-04:00 testsuite: Introduce template-haskell-exports test - - - - - 0c4a9686 by Luite Stegeman at 2024-04-04T07:05:39-04:00 Update correct counter in bumpTickyAllocd - - - - - 5f085d3a by Fendor at 2024-04-04T14:47:33-04:00 Replace `SizedSeq` with `FlatBag` for flattened structure LinkedLists are notoriously memory ineffiecient when all we do is traversing a structure. As 'UnlinkedBCO' has been identified as a data structure that impacts the overall memory usage of GHCi sessions, we avoid linked lists and prefer flattened structure for storing. We introduce a new memory efficient representation of sequential elements that has special support for the cases: * Empty * Singleton * Tuple Elements This improves sharing in the 'Empty' case and avoids the overhead of 'Array' until its constant overhead is justified. - - - - - 82cfe10c by Fendor at 2024-04-04T14:47:33-04:00 Compact FlatBag array representation `Array` contains three additional `Word`'s we do not need in `FlatBag`. Move `FlatBag` to `SmallArray`. Expand the API of SmallArray by `sizeofSmallArray` and add common traversal functions, such as `mapSmallArray` and `foldMapSmallArray`. Additionally, allow users to force the elements of a `SmallArray` via `rnfSmallArray`. - - - - - 36a75b80 by Andrei Borzenkov at 2024-04-04T14:48:10-04:00 Change how invisible patterns represented in haskell syntax and TH AST (#24557) Before this patch: data ArgPat p = InvisPat (LHsType p) | VisPat (LPat p) With this patch: data Pat p = ... | InvisPat (LHsType p) ... And the same transformation in the TH land. The rest of the changes is just updating code to handle new AST and writing tests to check if it is possible to create invalid states using TH. Metric Increase: MultiLayerModulesTH_OneShot - - - - - 28009fbc by Matthew Pickering at 2024-04-04T14:48:46-04:00 Fix off by one error in seekBinNoExpand and seekBin - - - - - 9b9e031b by Ben Gamari at 2024-04-04T21:30:08-04:00 compiler: Allow more types in GHCForeignImportPrim For many, many years `GHCForeignImportPrim` has suffered from the rather restrictive limitation of not allowing any non-trivial types in arguments or results. This limitation was justified by the code generator allegely barfing in the presence of such types. However, this restriction appears to originate well before the NCG rewrite and the new NCG does not appear to have any trouble with such types (see the added `T24598` test). Lift this restriction. Fixes #24598. - - - - - 1324b862 by Alan Zimmerman at 2024-04-04T21:30:44-04:00 EPA: Use EpaLocation not SrcSpan in ForeignDecls This allows us to update them for makeDeltaAst in ghc-exactprint - - - - - 19883a23 by Alan Zimmerman at 2024-04-05T16:58:17-04:00 EPA: Use EpaLocation for RecFieldsDotDot So we can update it to a delta position in makeDeltaAst if needed. - - - - - e8724327 by Matthew Pickering at 2024-04-05T16:58:53-04:00 Remove accidentally committed test.hs - - - - - 88cb3e10 by Fendor at 2024-04-08T09:03:34-04:00 Avoid UArray when indexing is not required `UnlinkedBCO`'s can occur many times in the heap. Each `UnlinkedBCO` references two `UArray`'s but never indexes them. They are only needed to encode the elements into a `ByteArray#`. The three words for the lower bound, upper bound and number of elements are essentially unused, thus we replace `UArray` with a wrapper around `ByteArray#`. This saves us up to three words for each `UnlinkedBCO`. Further, to avoid re-allocating these words for `ResolvedBCO`, we repeat the procedure for `ResolvedBCO` and add custom `Binary` and `Show` instances. For example, agda's repl session has around 360_000 UnlinkedBCO's, so avoiding these three words is already saving us around 8MB residency. - - - - - f2cc1107 by Fendor at 2024-04-08T09:04:11-04:00 Never UNPACK `FastMutInt` for counting z-encoded `FastString`s In `FastStringTable`, we count the number of z-encoded FastStrings that exist in a GHC session. We used to UNPACK the counters to not waste memory, but live retainer analysis showed that we allocate a lot of `FastMutInt`s, retained by `mkFastZString`. We lazily compute the `FastZString`, only incrementing the counter when the `FastZString` is forced. The function `mkFastStringWith` calls `mkZFastString` and boxes the `FastMutInt`, leading to the following core: mkFastStringWith = \ mk_fs _ -> = case stringTable of { FastStringTable _ n_zencs segments# _ -> ... case ((mk_fs (I# ...) (FastMutInt n_zencs)) `cast` <Co:2> :: ...) ... Marking this field as `NOUNPACK` avoids this reboxing, eliminating the allocation of a fresh `FastMutInt` on every `FastString` allocation. - - - - - c6def949 by Matthew Pickering at 2024-04-08T16:06:51-04:00 Force in_multi to avoid retaining entire hsc_env - - - - - fbb91a63 by Fendor at 2024-04-08T16:06:51-04:00 Eliminate name thunk in declaration fingerprinting Thunk analysis showed that we have about 100_000 thunks (in agda and `-fwrite-simplified-core`) pointing to the name of the name decl. Forcing this thunk fixes this issue. The thunk created here is retained by the thunk created by forkM, it is better to eagerly force this because the result (a `Name`) is already retained indirectly via the `IfaceDecl`. - - - - - 3b7b0c1c by Alan Zimmerman at 2024-04-08T16:07:27-04:00 EPA: Use EpaLocation in WarningTxt This allows us to use an EpDelta if needed when using makeDeltaAst. - - - - - 12b997df by Alan Zimmerman at 2024-04-08T16:07:27-04:00 EPA: Move DeltaPos and EpaLocation' into GHC.Types.SrcLoc This allows us to use a NoCommentsLocation for the possibly trailing comma location in a StringLiteral. This in turn allows us to correctly roundtrip via makeDeltaAst. - - - - - 868c8a78 by Fendor at 2024-04-09T08:51:50-04:00 Prefer packed representation for CompiledByteCode As there are many 'CompiledByteCode' objects alive during a GHCi session, representing its element in a more packed manner improves space behaviour at a minimal cost. When running GHCi on the agda codebase, we find around 380 live 'CompiledByteCode' objects. Packing their respective 'UnlinkedByteCode' can save quite some pointers. - - - - - be3bddde by Alan Zimmerman at 2024-04-09T08:52:26-04:00 EPA: Capture all comments in a ClassDecl Hopefully the final fix needed for #24533 - - - - - 3d0806fc by Jade at 2024-04-10T05:39:53-04:00 Validate -main-is flag using parseIdentifier Fixes #24368 - - - - - dd530bb7 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 rts: free error message before returning Fixes a memory leak in rts/linker/PEi386.c - - - - - e008a19a by Alexis King at 2024-04-10T05:40:29-04:00 linker: Avoid linear search when looking up Haskell symbols via dlsym See the primary Note [Looking up symbols in the relevant objects] for a more in-depth explanation. When dynamically loading a Haskell symbol (typical when running a splice or GHCi expression), before this commit we would search for the symbol in all dynamic libraries that were loaded. However, this could be very inefficient when too many packages are loaded (which can happen if there are many package dependencies) because the time to lookup the would be linear in the number of packages loaded. This commit drastically improves symbol loading performance by introducing a mapping from units to the handles of corresponding loaded dlls. These handles are returned by dlopen when we load a dll, and can then be used to look up in a specific dynamic library. Looking up a given Name is now much more precise because we can get lookup its unit in the mapping and lookup the symbol solely in the handles of the dynamic libraries loaded for that unit. In one measurement, the wait time before the expression was executed went from +-38 seconds down to +-2s. This commit also includes Note [Symbols may not be found in pkgs_loaded], explaining the fallback to the old behaviour in case no dll can be found in the unit mapping for a given Name. Fixes #23415 Co-authored-by: Rodrigo Mesquita (@alt-romes) - - - - - dcfaa190 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 rts: Make addDLL a wrapper around loadNativeObj Rewrite the implementation of `addDLL` as a wrapper around the more principled `loadNativeObj` rts linker function. The latter should be preferred while the former is preserved for backwards compatibility. `loadNativeObj` was previously only available on ELF platforms, so this commit further refactors the rts linker to transform loadNativeObj_ELF into loadNativeObj_POSIX, which is available in ELF and MachO platforms. The refactor made it possible to remove the `dl_mutex` mutex in favour of always using `linker_mutex` (rather than a combination of both). Lastly, we implement `loadNativeObj` for Windows too. - - - - - 12931698 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 Use symbol cache in internal interpreter too This commit makes the symbol cache that was used by the external interpreter available for the internal interpreter too. This follows from the analysis in #23415 that suggests the internal interpreter could benefit from this cache too, and that there is no good reason not to have the cache for it too. It also makes it a bit more uniform to have the symbol cache range over both the internal and external interpreter. This commit also refactors the cache into a function which is used by both `lookupSymbol` and also by `lookupSymbolInDLL`, extending the caching logic to `lookupSymbolInDLL` too. - - - - - dccd3ea1 by Ben Gamari at 2024-04-10T05:40:29-04:00 testsuite: Add test for lookupSymbolInNativeObj - - - - - 1b1a92bd by Alan Zimmerman at 2024-04-10T05:41:05-04:00 EPA: Remove unnecessary XRec in CompleteMatchSig The XRec for [LIdP pass] is not needed for exact printing, remove it. - - - - - 6e18ce2b by Ben Gamari at 2024-04-12T08:16:09-04:00 users-guide: Clarify language extension documentation Over the years the users guide's language extension documentation has gone through quite a few refactorings. In the process some of the descriptions have been rendered non-sensical. For instance, the description of `NoImplicitPrelude` actually describes the semantics of `ImplicitPrelude`. To fix this we: * ensure that all extensions are named in their "positive" sense (e.g. `ImplicitPrelude` rather than `NoImplicitPrelude`). * rework the documentation to avoid flag-oriented wording like "enable" and "disable" * ensure that the polarity of the documentation is consistent with reality. Fixes #23895. - - - - - a933aff3 by Zubin Duggal at 2024-04-12T08:16:45-04:00 driver: Make `checkHomeUnitsClosed` faster The implementation of `checkHomeUnitsClosed` was traversing every single path in the unit dependency graph - this grows exponentially and quickly grows to be infeasible on larger unit dependency graphs. Instead we replace this with a faster implementation which follows from the specificiation of the closure property - there is a closure error if there are units which are both are both (transitively) depended upon by home units and (transitively) depend on home units, but are not themselves home units. To compute the set of units required for closure, we first compute the closure of the unit dependency graph, then the transpose of this closure, and find all units that are reachable from the home units in the transpose of the closure. - - - - - 23c3e624 by Andreas Klebinger at 2024-04-12T08:17:21-04:00 RTS: Emit warning when -M < -H Fixes #24487 - - - - - d23afb8c by Ben Gamari at 2024-04-12T08:17:56-04:00 testsuite: Add broken test for CApiFFI with -fprefer-bytecode See #24634. - - - - - a4bb3a51 by Ben Gamari at 2024-04-12T08:18:32-04:00 base: Deprecate GHC.Pack As proposed in #21461. Closes #21540. - - - - - 55eb8c98 by Ben Gamari at 2024-04-12T08:19:08-04:00 ghc-internal: Fix mentions of ghc-internal in deprecation warnings Closes #24609. - - - - - b0fbd181 by Ben Gamari at 2024-04-12T08:19:44-04:00 rts: Implement set_initial_registers for AArch64 Fixes #23680. - - - - - 14c9ec62 by Ben Gamari at 2024-04-12T08:20:20-04:00 ghcup-metadata: Use Debian 9 binaries on Ubuntu 16, 17 Closes #24646. - - - - - 35a1621e by Ben Gamari at 2024-04-12T08:20:55-04:00 Bump unix submodule to 2.8.5.1 Closes #24640. - - - - - a1c24df0 by Finley McIlwaine at 2024-04-12T08:21:31-04:00 Correct default -funfolding-use-threshold in docs - - - - - 0255d03c by Oleg Grenrus at 2024-04-12T08:22:07-04:00 FastString is a __Modified__ UTF-8 - - - - - c3489547 by Matthew Pickering at 2024-04-12T13:13:44-04:00 rts: Improve tracing message when nursery is resized It is sometimes more useful to know how much bigger or smaller the nursery got when it is resized. In particular I am trying to investigate situations where we end up with fragmentation due to the nursery (#24577) - - - - - 5e4f4ba8 by Simon Peyton Jones at 2024-04-12T13:14:20-04:00 Don't generate wrappers for `type data` constructors with StrictData Previously, the logic for checking if a data constructor needs a wrapper or not would take into account whether the constructor's fields have explicit strictness (e.g., `data T = MkT !Int`), but the logic would _not_ take into account whether `StrictData` was enabled. This meant that something like `type data T = MkT Int` would incorrectly generate a wrapper for `MkT` if `StrictData` was enabled, leading to the horrible errors seen in #24620. To fix this, we disable generating wrappers for `type data` constructors altogether. Fixes #24620. Co-authored-by: Ryan Scott <ryan.gl.scott at gmail.com> - - - - - dbdf1995 by Alex Mason at 2024-04-15T15:28:26+10:00 Implements MO_S_Mul2 and MO_U_Mul2 using the UMULH, UMULL and SMULH instructions for AArch64 Also adds a test for MO_S_Mul2 - - - - - 42bd0407 by Teo Camarasu at 2024-04-16T20:06:39-04:00 Make template-haskell a stage1 package Promoting template-haskell from a stage0 to a stage1 package means that we can much more easily refactor template-haskell. We implement this by duplicating the in-tree `template-haskell`. A new `template-haskell-next` library is autogenerated to mirror `template-haskell` `stage1:ghc` to depend on the new interface of the library including the `Binary` instances without adding an explicit dependency on `template-haskell`. This is controlled by the `bootstrap-th` cabal flag When building `template-haskell` modules as part of this vendoring we do not have access to quote syntax, so we cannot use variable quote notation (`'Just`). So we either replace these with hand-written `Name`s or hide the code behind CPP. We can remove the `th_hack` from hadrian, which was required when building stage0 packages using the in-tree `template-haskell` library. For more details see Note [Bootstrapping Template Haskell]. Resolves #23536 Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> Co-Authored-By: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 3d973e47 by Ben Gamari at 2024-04-16T20:07:15-04:00 Bump parsec submodule to 3.1.17.0 - - - - - 9d38bfa0 by Simon Peyton Jones at 2024-04-16T20:07:51-04:00 Clone CoVars in CorePrep This MR addresses #24463. It's all explained in the new Note [Cloning CoVars and TyVars] - - - - - 0fe2b410 by Andreas Klebinger at 2024-04-16T20:08:27-04:00 NCG: Fix a bug where we errounously removed a required jump instruction. Add a new method to the Instruction class to check if we can eliminate a jump in favour of fallthrough control flow. Fixes #24507 - - - - - 9f99126a by Teo Camarasu at 2024-04-16T20:09:04-04:00 Fix documentation preview from doc-tarball job - Include all the .html files and assets in the job artefacts - Include all the .pdf files in the job artefacts - Mark the artefact as an "exposed" artefact meaning it turns up in the UI. Resolves #24651 - - - - - 3a0642ea by Ben Gamari at 2024-04-16T20:09:39-04:00 rts: Ignore EINTR while polling in timerfd itimer implementation While the RTS does attempt to mask signals, it may be that a foreign library unmasks them. This previously caused benign warnings which we now ignore. See #24610. - - - - - 9a53cd3f by Alan Zimmerman at 2024-04-16T20:10:15-04:00 EPA: Add additional comments field to AnnsModule This is used in exact printing to store comments coming after the `where` keyword but before any comments allocated to imports or decls. It is used in ghc-exactprint, see https://github.com/alanz/ghc-exactprint/commit/44bbed311fd8f0d053053fef195bf47c17d34fa7 - - - - - e5c43259 by Bryan Richter at 2024-04-16T20:10:51-04:00 Remove unrunnable FreeBSD CI jobs FreeBSD runner supply is inelastic. Currently there is only one, and it's unavailable because of a hardware issue. - - - - - 914eb49a by Ben Gamari at 2024-04-16T20:11:27-04:00 rel-eng: Fix mktemp usage in recompress-all We need a temporary directory, not a file. - - - - - f30e4984 by Teo Camarasu at 2024-04-16T20:12:03-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 - - - - - d7a3d6b5 by Ben Gamari at 2024-04-16T20:12:39-04:00 template-haskell: Declare TH.Lib.Internal as not-home Rather than `hide`. Closes #24659. - - - - - 5eaa46e7 by Matthew Pickering at 2024-04-19T02:14:55-04:00 testsuite: Rename isCross() predicate to needsTargetWrapper() isCross() was a misnamed because it assumed that all cross targets would provide a target wrapper, but the two most common cross targets (javascript, wasm) don't need a target wrapper. Therefore we rename this predicate to `needsTargetWrapper()` so situations in the testsuite where we can check whether running executables requires a target wrapper or not. - - - - - 55a9d699 by Simon Peyton Jones at 2024-04-19T02:15:32-04:00 Do not float HNFs out of lambdas This MR adjusts SetLevels so that it is less eager to float a HNF (lambda or constructor application) out of a lambda, unless it gets to top level. Data suggests that this change is a small net win: * nofib bytes-allocated falls by -0.09% (but a couple go up) * perf/should_compile bytes-allocated falls by -0.5% * perf/should_run bytes-allocated falls by -0.1% See !12410 for more detail. When fiddling elsewhere, I also found that this patch had a huge positive effect on the (very delicate) test perf/should_run/T21839r But that improvement doesn't show up in this MR by itself. Metric Decrease: MultiLayerModulesRecomp T15703 parsing001 - - - - - f0701585 by Alan Zimmerman at 2024-04-19T02:16:08-04:00 EPA: Fix comments in mkListSyntaxTy0 Also extend the test to confirm. Addresses #24669, 1 of 4 - - - - - b01c01d4 by Serge S. Gulin at 2024-04-19T02:16:51-04:00 JS: set image `x86_64-linux-deb11-emsdk-closure` for build - - - - - c90c6039 by Alan Zimmerman at 2024-04-19T02:17:27-04:00 EPA: Provide correct span for PatBind And remove unused parameter in checkPatBind Contributes to #24669 - - - - - bee54c24 by Krzysztof Gogolewski at 2024-04-19T11:13:00+02:00 Update quantification order following GHC haskell/haddock#23764 - - - - - 2814eb89 by Ben Gamari at 2024-04-19T18:57:05+02:00 hypsrc-test: Fix output of PositionPragmas.html - - - - - 26036f96 by Alan Zimmerman at 2024-04-19T13:11:08-04:00 EPA: Fix span for PatBuilderAppType Include the location of the prefix @ in the span for InVisPat. Also removes unnecessary annotations from HsTP. Contributes to #24669 - - - - - dba03aab by Matthew Craven at 2024-04-19T13:11:44-04:00 testsuite: Give the pre_cmd for mhu-perf more time - - - - - d31fbf6c by Krzysztof Gogolewski at 2024-04-19T21:04:09-04:00 Fix quantification order for a `op` b and a %m -> b Fixes #23764 Implements https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0640-tyop-quantification-order.rst Updates haddock submodule. - - - - - 385cd1c4 by Sebastian Graf at 2024-04-19T21:04:45-04:00 Make `seq#` a magic Id and inline it in CorePrep (#24124) We can save much code and explanation in Tag Inference and StgToCmm by making `seq#` a known-key Magic Id in `GHC.Internal.IO` and inline this definition in CorePrep. See the updated `Note [seq# magic]`. I also implemented a new `Note [Flatten case-bind]` to get better code for otherwise nested case scrutinees. I renamed the contructors of `ArgInfo` to use an `AI` prefix in order to resolve the clash between `type CpeApp = CoreExpr` and the data constructor of `ArgInfo`, as well as fixed typos in `Note [CorePrep invariants]`. Fixes #24252 and #24124. - - - - - 275e41a9 by Jade at 2024-04-20T11:10:40-04:00 Put the newline after errors instead of before them This mainly has consequences for GHCi but also slightly alters how the output of GHC on the commandline looks. Fixes: #22499 - - - - - dd339c7a by Teo Camarasu at 2024-04-20T11:11:16-04:00 Remove unecessary stage0 packages Historically quite a few packages had to be stage0 as they depended on `template-haskell` and that was stage0. In #23536 we made it so that was no longer the case. This allows us to remove a bunch of packages from this list. A few still remain. A new version of `Win32` is required by `semaphore-compat`. Including `Win32` in the stage0 set requires also including `filepath` because otherwise Hadrian's dependency logic gets confused. Once our boot compiler has a newer version of `Win32` all of these will be able to be dropped. Resolves #24652 - - - - - 2f8e3a25 by Alan Zimmerman at 2024-04-20T11:11:52-04:00 EPA: Avoid duplicated comments in splice decls Contributes to #24669 - - - - - c70b9ddb by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: fix typos and namings (fixes #24602) You may noted that I've also changed term of ``` , global "h$vt_double" ||= toJExpr IntV ``` See "IntV" and ``` WaitReadOp -> \[] [fd] -> pure $ PRPrimCall $ returnS (app "h$waidRead" [fd]) ``` See "h$waidRead" - - - - - 3db54f9b by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: trivial checks for variable presence (fixes #24602) - - - - - 777f108f by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: fs module imported twice (by emscripten and by ghc-internal). ghc-internal import wrapped in a closure to prevent conflict with emscripten (fixes #24602) Better solution is to use some JavaScript module system like AMD, CommonJS or even UMD. It will be investigated at other issues. At first glance we should try UMD (See https://github.com/umdjs/umd) - - - - - a45a5712 by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: thread.js requires h$fds and h$fdReady to be declared for static code analysis, minimal code copied from GHCJS (fixes #24602) I've just copied some old pieces of GHCJS from publicly available sources (See https://github.com/Taneb/shims/blob/a6dd0202dcdb86ad63201495b8b5d9763483eb35/src/io.js#L607). Also I didn't put details to h$fds. I took minimal and left only its object initialization: `var h$fds = {};` - - - - - ad90bf12 by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: heap and stack overflows reporting defined as js hard failure (fixes #24602) These errors were treated as a hard failure for browser application. The fix is trivial: just throw error. - - - - - 5962fa52 by Serge S. Gulin at 2024-04-21T16:33:44+03:00 JS: Stubs for code without actual implementation detected by Google Closure Compiler (fixes #24602) These errors were fixed just by introducing stubbed functions with throw for further implementation. - - - - - a0694298 by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Add externs to linker (fixes #24602) After enabling jsdoc and built-in google closure compiler types I was needed to deal with the following: 1. Define NodeJS-environment types. I've just copied minimal set of externs from semi-official repo (see https://github.com/externs/nodejs/blob/6c6882c73efcdceecf42e7ba11f1e3e5c9c041f0/v8/nodejs.js#L8). 2. Define Emscripten-environment types: `HEAP8`. Emscripten already provides some externs in our code but it supposed to be run in some module system. And its definitions do not work well in plain bundle. 3. We have some functions which purpose is to add to functions some contextual information via function properties. These functions should be marked as `modifies` to let google closure compiler remove calls if these functions are not used actually by call graph. Such functions are: `h$o`, `h$sti`, `h$init_closure`, `h$setObjInfo`. 4. STG primitives such as registries and stuff from `GHC.StgToJS`. `dXX` properties were already present at externs generator function but they are started from `7`, not from `1`. This message is related: `// fixme does closure compiler bite us here?` - - - - - e58bb29f by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: added both tests: for size and for correctness (fixes #24602) By some reason MacOS builds add to stderr messages like: Ignoring unexpected archive entry: __.SYMDEF ... However I left stderr to `/dev/null` for compatibility with linux CI builds. - - - - - 909f3a9c by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Disable js linker warning for empty symbol table to make js tests running consistent across environments - - - - - 83eb10da by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Add special preprocessor for js files due of needing to keep jsdoc comments (fixes #24602) Our js files have defined google closure compiler types at jsdoc entries but these jsdoc entries are removed by cpp preprocessor. I considered that reusing them in javascript-backend would be a nice thing. Right now haskell processor uses `-traditional` option to deal with comments and `//` operators. But now there are following compiler options: `-C` and `-CC`. You can read about them at GCC (see https://gcc.gnu.org/onlinedocs/gcc/Preprocessor-Options.html#index-CC) and CLang (see https://clang.llvm.org/docs/ClangCommandLineReference.html#cmdoption-clang-CC). It seems that `-CC` works better for javascript jsdoc than `-traditional`. At least it leaves `/* ... */` comments w/o changes. - - - - - e1cf8dc2 by brandon s allbery kf8nh at 2024-04-22T03:48:26-04:00 fix link in CODEOWNERS It seems that our local Gitlab no longer has documentation for the `CODEOWNERS` file, but the master documentation still does. Use that instead. - - - - - a27c6a49 by Fendor at 2024-04-22T10:13:03+02:00 Adapt to UserData split - - - - - 1efc5a7a by Fendor at 2024-04-22T10:13:03+02:00 Adapt to BinHandle split - - - - - 593f4e04 by Fendor at 2024-04-23T10:19:14-04:00 Add performance regression test for '-fwrite-simplified-core' - - - - - 1ba39b05 by Fendor at 2024-04-23T10:19:14-04:00 Typecheck corebindings lazily during bytecode generation This delays typechecking the corebindings until the bytecode generation happens. We also avoid allocating a thunk that is retained by `unsafeInterleaveIO`. In general, we shouldn't retain values of the hydrated `Type`, as not evaluating the bytecode object keeps it alive. It is better if we retain the unhydrated `IfaceType`. See Note [Hydrating Modules] - - - - - e916fc92 by Alan Zimmerman at 2024-04-23T10:19:50-04:00 EPA: Keep comments in a CaseAlt match The comments now live in the surrounding location, not inside the Match. Make sure we keep them. Closes #24707 - - - - - d2b17f32 by Cheng Shao at 2024-04-23T15:01:22-04:00 driver: force merge objects when building dynamic objects This patch forces the driver to always merge objects when building dynamic objects even when ar -L is supported. It is an oversight of !8887: original rationale of that patch is favoring the relatively cheap ar -L operation over object merging when ar -L is supported, which makes sense but only if we are building static objects! Omitting check for whether we are building dynamic objects will result in broken .so files with undefined reference errors at executable link time when building GHC with llvm-ar. Fixes #22210. - - - - - 209d09f5 by Julian Ospald at 2024-04-23T15:02:03-04:00 Allow non-absolute values for bootstrap GHC variable Fixes #24682 - - - - - 3fff0977 by Matthew Pickering at 2024-04-23T15:02:38-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. - - - - - c62dc317 by Cheng Shao at 2024-04-25T01:32:02-04:00 ghc-bignum: remove obsolete ln script This commit removes an obsolete ln script in ghc-bignum/gmp. See 060251c24ad160264ae8553efecbb8bed2f06360 for its original intention, but it's been obsolete for a long time, especially since the removal of the make build system. Hence the house cleaning. - - - - - 6399d52b by Cheng Shao at 2024-04-25T01:32:02-04:00 ghc-bignum: update gmp to 6.3.0 This patch bumps the gmp-tarballs submodule and updates gmp to 6.3.0. The tarball format is now xz, and gmpsrc.patch has been patched into the tarball so hadrian no longer needs to deal with patching logic when building in-tree GMP. - - - - - 65b4b92f by Cheng Shao at 2024-04-25T01:32:02-04:00 hadrian: remove obsolete Patch logic This commit removes obsolete Patch logic from hadrian, given we no longer need to patch the gmp tarball when building in-tree GMP. - - - - - 71f28958 by Cheng Shao at 2024-04-25T01:32:02-04:00 autoconf: remove obsolete patch detection This commit removes obsolete deletection logic of the patch command from autoconf scripts, given we no longer need to patch anything in the GHC build process. - - - - - daeda834 by Sylvain Henry at 2024-04-25T01:32:43-04:00 JS: correctly handle RUBBISH literals (#24664) - - - - - 8a06ddf6 by Matthew Pickering at 2024-04-25T11:16:16-04:00 Linearise ghc-internal and base build This is achieved by requesting the final package database for ghc-internal, which mandates it is fully built as a dependency of configuring the `base` package. This is at the expense of cross-package parrallelism between ghc-internal and the base package. Fixes #24436 - - - - - 94da9365 by Andrei Borzenkov at 2024-04-25T11:16:54-04:00 Fix tuple puns renaming (24702) Move tuple renaming short cutter from `isBuiltInOcc_maybe` to `isPunOcc_maybe`, so we consider incoming module. I also fixed some hidden bugs that raised after the change was done. - - - - - fa03b1fb by Fendor at 2024-04-26T18:03:13-04:00 Refactor the Binary serialisation interface The goal is simplifiy adding deduplication tables to `ModIface` interface serialisation. We identify two main points of interest that make this difficult: 1. UserData hardcodes what `Binary` instances can have deduplication tables. Moreover, it heavily uses partial functions. 2. GHC.Iface.Binary hardcodes the deduplication tables for 'Name' and 'FastString', making it difficult to add more deduplication. Instead of having a single `UserData` record with fields for all the types that can have deduplication tables, we allow to provide custom serialisers for any `Typeable`. These are wrapped in existentials and stored in a `Map` indexed by their respective `TypeRep`. The `Binary` instance of the type to deduplicate still needs to explicitly look up the decoder via `findUserDataReader` and `findUserDataWriter`, which is no worse than the status-quo. `Map` was chosen as microbenchmarks indicate it is the fastest for a small number of keys (< 10). To generalise the deduplication table serialisation mechanism, we introduce the types `ReaderTable` and `WriterTable` which provide a simple interface that is sufficient to implement a general purpose deduplication mechanism for `writeBinIface` and `readBinIface`. This allows us to provide a list of deduplication tables for serialisation that can be extended more easily, for example for `IfaceTyCon`, see the issue https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for more motivation. In addition to this refactoring, we split `UserData` into `ReaderUserData` and `WriterUserData`, to avoid partial functions and reduce overall memory usage, as we need fewer mutable variables. Bump haddock submodule to accomodate for `UserData` split. ------------------------- Metric Increase: MultiLayerModulesTH_Make MultiLayerModulesRecomp T21839c ------------------------- - - - - - bac57298 by Fendor at 2024-04-26T18:03:13-04:00 Split `BinHandle` into `ReadBinHandle` and `WriteBinHandle` A `BinHandle` contains too much information for reading data. For example, it needs to keep a `FastMutInt` and a `IORef BinData`, when the non-mutable variants would suffice. Additionally, this change has the benefit that anyone can immediately tell whether the `BinHandle` is used for reading or writing. Bump haddock submodule BinHandle split. - - - - - 4d6394dd by Simon Peyton Jones at 2024-04-26T18:03:49-04:00 Fix missing escaping-kind check in tcPatSynSig Note [Escaping kind in type signatures] explains how we deal with escaping kinds in type signatures, e.g. f :: forall r (a :: TYPE r). a where the kind of the body is (TYPE r), but `r` is not in scope outside the forall-type. I had missed this subtlety in tcPatSynSig, leading to #24686. This MR fixes it; and a similar bug in tc_top_lhs_type. (The latter is tested by T24686a.) - - - - - 981c2c2c by Alan Zimmerman at 2024-04-26T18:04:25-04:00 EPA: check-exact: check that the roundtrip reproduces the source Closes #24670 - - - - - a8616747 by Andrew Lelechenko at 2024-04-26T18:05:01-04:00 Document that setEnv is not thread-safe - - - - - 1e41de83 by Bryan Richter at 2024-04-26T18:05:37-04:00 CI: Work around frequent Signal 9 errors - - - - - a6d5f9da by Naïm Favier at 2024-04-27T17:52:40-04:00 ghc-internal: add MonadFix instance for (,) Closes https://gitlab.haskell.org/ghc/ghc/-/issues/24288, implements CLC proposal https://github.com/haskell/core-libraries-committee/issues/238. Adds a MonadFix instance for tuples, permitting value recursion in the "native" writer monad and bringing consistency with the existing instance for transformers's WriterT (and, to a lesser extent, for Solo). - - - - - 64feadcd by Rodrigo Mesquita at 2024-04-27T17:53:16-04:00 bindist: Fix xattr cleaning The original fix (725343aa) was incorrect because it used the shell bracket syntax which is the quoting syntax in autoconf, making the test for existence be incorrect and therefore `xattr` was never run. Fixes #24554 - - - - - e2094df3 by damhiya at 2024-04-28T23:52:00+09:00 Make read accepts binary integer formats CLC proposal : https://github.com/haskell/core-libraries-committee/issues/177 - - - - - c62239b7 by Sylvain Henry at 2024-04-29T10:35:00+02:00 Fix tests for T22229 - - - - - 1c2fd963 by Alan Zimmerman at 2024-04-29T23:17:00-04:00 EPA: Preserve comments in Match Pats Closes #24708 Closes #24715 Closes #24734 - - - - - 4189d17e by Sylvain Henry at 2024-04-29T23:17:42-04:00 LLVM: better unreachable default destination in Switch (#24717) See added note. Co-authored-by: Siddharth Bhat <siddu.druid at gmail.com> - - - - - a3725c88 by Cheng Shao at 2024-04-29T23:18:20-04:00 ci: enable wasm jobs for MRs with wasm label This patch enables wasm jobs for MRs with wasm label. Previously the wasm label didn't actually have any effect on the CI pipeline, and full-ci needed to be applied to run wasm jobs which was a waste of runners when working on the wasm backend, hence the fix here. - - - - - 702f7964 by Matthew Pickering at 2024-04-29T23:18:56-04:00 Make interface files and object files depend on inplace .conf file A potential fix for #24737 - - - - - 728af21e by Cheng Shao at 2024-04-30T05:30:23-04:00 utils: remove obsolete vagrant scripts Vagrantfile has long been removed in !5288. This commit further removes the obsolete vagrant scripts in the tree. - - - - - 36f2c342 by Cheng Shao at 2024-04-30T05:31:00-04:00 Update autoconf scripts Scripts taken from autoconf 948ae97ca5703224bd3eada06b7a69f40dd15a02 - - - - - ecbf22a6 by Ben Gamari at 2024-04-30T05:31:36-04:00 ghcup-metadata: Drop output_name field This is entirely redundant to the filename of the URL. There is no compelling reason to name the downloaded file differently from its source. - - - - - c56d728e by Zubin Duggal at 2024-04-30T22:45:09-04:00 testsuite: Handle exceptions in framework_fail when testdir is not initialised When `framework_fail` is called before initialising testdir, it would fail with an exception reporting the testdir not being initialised instead of the actual failure. Ensure we report the actual reason for the failure instead of failing in this way. One way this can manifest is when trying to run a test that doesn't exist using `--only` - - - - - d5bea4d6 by Alan Zimmerman at 2024-04-30T22:45:45-04:00 EPA: Fix range for GADT decl with sig only Closes #24714 - - - - - 4d78c53c by Sylvain Henry at 2024-05-01T17:23:06-04:00 Fix TH dependencies (#22229) Add a dependency between Syntax and Internal (via module reexport). - - - - - 37e38db4 by Sylvain Henry at 2024-05-01T17:23:06-04:00 Bump haddock submodule - - - - - ca13075c by Sylvain Henry at 2024-05-01T17:23:47-04:00 JS: cleanup to prepare for #24743 - - - - - 40026ac3 by Alan Zimmerman at 2024-05-01T22:45:07-04:00 EPA: Preserve comments for PrefixCon Preserve comments in fun (Con {- c1 -} a b) = undefined Closes #24736 - - - - - 92134789 by Hécate Moonlight at 2024-05-01T22:45:42-04:00 Correct `@since` metadata in HpcFlags It was introduced in base-4.20, not 4.22. Fix #24721 - - - - - a580722e by Cheng Shao at 2024-05-02T08:18:45-04:00 testsuite: fix req_target_smp predicate - - - - - ac9c5f84 by Andreas Klebinger at 2024-05-02T08:18:45-04:00 STM: Remove (unused)coarse grained locking. The STM code had a coarse grained locking mode guarded by #defines that was unused. This commit removes the code. - - - - - 917ef81b by Andreas Klebinger at 2024-05-02T08:18:45-04:00 STM: Be more optimistic when validating in-flight transactions. * Don't lock tvars when performing non-committal validation. * If we encounter a locked tvar don't consider it a failure. This means in-flight validation will only fail if committing at the moment of validation is *guaranteed* to fail. This prevents in-flight validation from failing spuriously if it happens in parallel on multiple threads or parallel to thread comitting. - - - - - 167a56a0 by Alan Zimmerman at 2024-05-02T08:19:22-04:00 EPA: fix span for empty \case(s) In instance SDecide Nat where SZero %~ (SSucc _) = Disproved (\case) Ensure the span for the HsLam covers the full construct. Closes #24748 - - - - - 9bae34d8 by doyougnu at 2024-05-02T15:41:08-04:00 testsuite: expand size testing infrastructure - closes #24191 - adds windows_skip, wasm_skip, wasm_arch, find_so, _find_so - path_from_ghcPkg, collect_size_ghc_pkg, collect_object_size, find_non_inplace functions to testsuite - adds on_windows and req_dynamic_ghc predicate to testsuite The design is to not make the testsuite too smart and simply offload to ghc-pkg for locations of object files and directories. - - - - - b85b1199 by Sylvain Henry at 2024-05-02T15:41:49-04:00 GHCi: support inlining breakpoints (#24712) When a breakpoint is inlined, its context may change (e.g. tyvars in scope). We must take this into account and not used the breakpoint tick index as its sole identifier. Each instance of a breakpoint (even with the same tick index) now gets a different "info" index. We also need to distinguish modules: - tick module: module with the break array (tick counters, status, etc.) - info module: module having the CgBreakInfo (info at occurrence site) - - - - - 649c24b9 by Oleg Grenrus at 2024-05-03T20:45:42-04:00 Expose constructors of SNat, SChar and SSymbol in ghc-internal - - - - - d603f199 by Mikolaj Konarski at 2024-05-03T20:46:19-04:00 Add DCoVarSet to PluginProv (!12037) - - - - - ba480026 by Serge S. Gulin at 2024-05-03T20:47:01-04:00 JS: Enable more efficient packing of string data (fixes #24706) - - - - - be1e60ee by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Track in-scope variables in ruleCheckProgram This small patch fixes #24726, by tracking in-scope variables properly in -drule-check. Not hard to do! - - - - - 58408c77 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Add a couple more HasCallStack constraints in SimpleOpt Just for debugging, no effect on normal code - - - - - 70e245e8 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Add comments to Prep.hs This documentation patch fixes a TODO left over from !12364 - - - - - e5687186 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Use HasDebugCallStack, rather than HasCallStack - - - - - 631cefec by Cheng Shao at 2024-05-03T20:48:17-04:00 driver: always merge objects when possible This patch makes the driver always merge objects with `ld -r` when possible, and only fall back to calling `ar -L` when merge objects command is unavailable. This completely reverts !8887 and !12313, given more fixes in Cabal seems to be needed to avoid breaking certain configurations and the maintainence cost is exceeding the behefits in this case :/ - - - - - 1dacb506 by Ben Gamari at 2024-05-03T20:48:53-04:00 Bump time submodule to 1.14 As requested in #24528. ------------------------- Metric Decrease: ghc_bignum_so rts_so Metric Increase: cabal_syntax_dir rts_so time_dir time_so ------------------------- - - - - - 4941b90e by Ben Gamari at 2024-05-03T20:48:53-04:00 Bump terminfo submodule to current master - - - - - 43d48b44 by Cheng Shao at 2024-05-03T20:49:30-04:00 wasm: use scheduler.postTask() for context switch when available This patch makes use of scheduler.postTask() for JSFFI context switch when it's available. It's a more principled approach than our MessageChannel based setImmediate() implementation, and it's available in latest version of Chromium based browsers. - - - - - 08207501 by Cheng Shao at 2024-05-03T20:50:08-04:00 testsuite: give pre_cmd for mhu-perf 5x time - - - - - bf3d4db0 by Alan Zimmerman at 2024-05-03T20:50:43-04:00 EPA: Preserve comments for pattern synonym sig Closes #24749 - - - - - c49493f2 by Matthew Pickering at 2024-05-04T06:02:57-04:00 tests: Widen acceptance window for dir and so size tests These are testing things which are sometimes out the control of a GHC developer. Therefore we shouldn't fail CI if something about these dependencies change because we can't do anything about it. It is still useful to have these statistics for visualisation in grafana though. Ticket #24759 - - - - - 9562808d by Matthew Pickering at 2024-05-04T06:02:57-04:00 Disable rts_so test It has already manifested large fluctuations and destabilising CI Fixes #24762 - - - - - fc24c5cf by Ryan Scott at 2024-05-04T06:03:33-04:00 unboxedSum{Type,Data}Name: Use GHC.Types as the module Unboxed sum constructors are now defined in the `GHC.Types` module, so if you manually quote an unboxed sum (e.g., `''Sum2#`), you will get a `Name` like: ```hs GHC.Types.Sum2# ``` The `unboxedSumTypeName` function in `template-haskell`, however, mistakenly believes that unboxed sum constructors are defined in `GHC.Prim`, so `unboxedSumTypeName 2` would return an entirely different `Name`: ```hs GHC.Prim.(#|#) ``` This is a problem for Template Haskell users, as it means that they can't be sure which `Name` is the correct one. (Similarly for `unboxedSumDataName`.) This patch fixes the implementations of `unboxedSum{Type,Data}Name` to use `GHC.Types` as the module. For consistency with `unboxedTupleTypeName`, the `unboxedSumTypeName` function now uses the non-punned syntax for unboxed sums (`Sum<N>#`) as the `OccName`. Fixes #24750. - - - - - 7eab4e01 by Alan Zimmerman at 2024-05-04T16:14:55+01:00 EPA: Widen stmtslist to include last semicolon Closes #24754 - - - - - 06f7db40 by Teo Camarasu at 2024-05-05T00:19:38-04:00 doc: Fix type error in hs_try_putmvar example - - - - - af000532 by Moritz Schuler at 2024-05-05T06:30:58-04:00 Fix parsing of module names in CLI arguments closes issue #24732 - - - - - da74e9c9 by Ben Gamari at 2024-05-05T06:31:34-04:00 ghc-platform: Add Setup.hs The Hadrian bootstrapping script relies upon `Setup.hs` to drive its build. Addresses #24761. - - - - - 35d34fde by Alan Zimmerman at 2024-05-05T12:52:40-04:00 EPA: preserve comments in class and data decls Fix checkTyClHdr which was discarding comments. Closes #24755 - - - - - 03c5dfbf by Simon Peyton Jones at 2024-05-05T12:53:15-04:00 Fix a float-out error Ticket #24768 showed that the Simplifier was accidentally destroying a join point. It turned out to be that we were sending a bottoming join point to the top, accidentally abstracting over /other/ join points. Easily fixed. - - - - - adba68e7 by John Ericson at 2024-05-05T19:35:56-04:00 Substitute bindist files with Hadrian not configure The `ghc-toolchain` overhaul will eventually replace all this stuff with something much more cleaned up, but I think it is still worth making this sort of cleanup in the meantime so other untanglings and dead code cleaning can procede. I was able to delete a fair amount of dead code doing this too. `LLVMTarget_CPP` is renamed to / merged with `LLVMTarget` because it wasn't actually turned into a valid CPP identifier. (Original to 1345c7cc42c45e63ab1726a8fd24a7e4d4222467, actually.) Progress on #23966 Co-Authored-By: Sylvain Henry <hsyl20 at gmail.com> - - - - - 18f4ff84 by Alan Zimmerman at 2024-05-05T19:36:32-04:00 EPA: fix mkHsOpTyPV duplicating comments Closes #24753 - - - - - a19201d4 by Matthew Craven at 2024-05-06T19:54:29-04:00 Add test cases for #24664 ...since none are present in the original MR !12463 fixing this issue. - - - - - 46328a49 by Alan Zimmerman at 2024-05-06T19:55:05-04:00 EPA: preserve comments in data decls Closes #24771 - - - - - 3b51995c by Andrei Borzenkov at 2024-05-07T14:39:40-04:00 Rename Solo# data constructor to MkSolo# (#24673) - data Solo# a = (# a #) + data Solo# a = MkSolo# a And `(# foo #)` syntax now becomes just a syntactic sugar for `MkSolo# a`. - - - - - 4d59abf2 by Arsen Arsenović at 2024-05-07T14:40:24-04:00 Add the cmm_cpp_is_gcc predicate to the testsuite A future C-- test called T24474-cmm-override-g0 relies on the GCC-specific behaviour of -g3 implying -dD, which, in turn, leads to it emitting #defines past the preprocessing stage. Clang, at least, does not do this, so the test would fail if ran on Clang. As the behaviour here being tested is ``-optCmmP-g3'' undoing effects of the workaround we apply as a fix for bug #24474, and the workaround was for GCC-specific behaviour, the test needs to be marked as fragile on other compilers. - - - - - 25b0b404 by Arsen Arsenović at 2024-05-07T14:40:24-04:00 Split out the C-- preprocessor, and make it pass -g0 Previously, C-- was processed with the C preprocessor program. This means that it inherited flags passed via -optc. A flag that is somewhat often passed through -optc is -g. At certain -g levels (>=2), GCC starts emitting defines *after* preprocessing, for the purposes of debug info generation. This is not useful for the C-- compiler, and, in fact, causes lexer errors. We can suppress this effect (safely, if supported) via -g0. As a workaround, in older versions of GCC (<=10), GCC only emitted defines if a certain set of -g*3 flags was passed. Newer versions check the debug level. For the former, we filter out those -g*3 flags and, for the latter, we specify -g0 on top of that. As a compatible and effective solution, this change adds a C-- preprocessor distinct from the C compiler and preprocessor, but that keeps its flags. The command line produced for C-- preprocessing now looks like: $pgmCmmP $optCs_without_g3 $g0_if_supported $optCmmP Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/24474 - - - - - 9b4129a5 by Andreas Klebinger at 2024-05-08T13:24:20-04:00 -fprof-late: Only insert cost centres on functions/non-workfree cafs. They are usually useless and doing so for data values comes with a large compile time/code size overhead. Fixes #24103 - - - - - 259b63d3 by Sebastian Graf at 2024-05-08T13:24:57-04:00 Simplifier: Preserve OccInfo on DataAlt fields when case binder is dead (#24770) See the adjusted `Note [DataAlt occ info]`. This change also has a positive repercussion on `Note [Combine case alts: awkward corner]`. Fixes #24770. We now try not to call `dataConRepStrictness` in `adjustFieldsIdInfo` when all fields are lazy anyway, leading to a 2% ghc/alloc decrease in T9675. Metric Decrease: T9675 - - - - - 31b28cdb by Sebastian Graf at 2024-05-08T13:24:57-04:00 Kill seqRule, discard dead seq# in Prep (#24334) Discarding seq#s in Core land via `seqRule` was problematic; see #24334. So instead we discard certain dead, discardable seq#s in Prep now. See the updated `Note [seq# magic]`. This fixes the symptoms of #24334. - - - - - b2682534 by Rodrigo Mesquita at 2024-05-10T01:47:51-04:00 Document NcgImpl methods Fixes #19914 - - - - - 4d3acbcf by Zejun Wu at 2024-05-10T01:48:28-04:00 Make renamer to be more flexible with parens in the LHS of the rules We used to reject LHS like `(f a) b` in RULES and requires it to be written as `f a b`. It will be handy to allow both as the expression may be more readable with extra parens in some cases when infix operator is involved. Espceially when TemplateHaskell is used, extra parens may be added out of user's control and result in "valid" rules being rejected and there are not always ways to workaround it. Fixes #24621 - - - - - ab840ce6 by Ben Gamari at 2024-05-10T01:49:04-04:00 IPE: Eliminate dependency on Read Instead of encoding the closure type as decimal string we now simply represent it as an integer, eliminating the need for `Read` in `GHC.Internal.InfoProv.Types.peekInfoProv`. Closes #24504. ------------------------- Metric Decrease: T24602_perf_size size_hello_artifact ------------------------- - - - - - a9979f55 by Cheng Shao at 2024-05-10T01:49:43-04:00 testsuite: fix testwsdeque with recent clang This patch fixes compilation of testwsdeque.c with recent versions of clang, which will fail with the error below: ``` testwsdeque.c:95:33: error: warning: format specifies type 'long' but the argument has type 'void *' [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~~ ^ testwsdeque.c:95:39: error: warning: format specifies type 'int' but the argument has type 'StgWord' (aka 'unsigned long') [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~ ^~~ | %lu testwsdeque.c:133:42: error: error: incompatible function pointer types passing 'void (void *)' to parameter of type 'OSThreadProc *' (aka 'void *(*)(void *)') [-Wincompatible-function-pointer-types] 133 | createOSThread(&ids[n], "thief", thief, (void*)(StgWord)n); | ^~~~~ /workspace/ghc/_build/stage1/lib/../lib/x86_64-linux-ghc-9.11.20240502/rts-1.0.2/include/rts/OSThreads.h:193:51: error: note: passing argument to parameter 'startProc' here 193 | OSThreadProc *startProc, void *param); | ^ 2 warnings and 1 error generated. ``` - - - - - c2b33fc9 by Rodrigo Mesquita at 2024-05-10T01:50:20-04:00 Rename pre-processor invocation args Small clean up. Uses proper names for the various groups of arguments that make up the pre-processor invocation. - - - - - 2b1af08b by Cheng Shao at 2024-05-10T01:50:55-04:00 ghc-heap: fix typo in ghc-heap cbits - - - - - fc2d6de1 by Jade at 2024-05-10T21:07:16-04:00 Improve performance of Data.List.sort(By) This patch improves the algorithm to sort lists in base. It does so using two strategies: 1) Use a four-way-merge instead of the 'default' two-way-merge. This is able to save comparisons and allocations. 2) Use `(>) a b` over `compare a b == GT` and allow inlining and specialization. This mainly benefits types with a fast (>). Note that this *may* break instances with a *malformed* Ord instance where `a > b` is *not* equal to `compare a b == GT`. CLC proposal: https://github.com/haskell/core-libraries-committee/issues/236 Fixes #24280 ------------------------- Metric Decrease: MultiLayerModulesTH_Make T10421 T13719 T15164 T18698a T18698b T1969 T9872a T9961 T18730 WWRec T12425 T15703 ------------------------- - - - - - 1012e8aa by Matthew Pickering at 2024-05-10T21:07:52-04:00 Revert "ghcup-metadata: Drop output_name field" This reverts commit ecbf22a6ac397a791204590f94c0afa82e29e79f. This breaks the ghcup metadata generation on the nightly jobs. - - - - - daff1e30 by Jannis at 2024-05-12T13:38:35-04:00 Division by constants optimization - - - - - 413217ba by Andreas Klebinger at 2024-05-12T13:39:11-04:00 Tidy: Add flag to expose unfoldings if they take dictionary arguments. Add the flag `-fexpose-overloaded-unfoldings` to be able to control this behaviour. For ghc's boot libraries file size grew by less than 1% when it was enabled. However I refrained from enabling it by default for now. I've also added a section on specialization more broadly to the users guide. ------------------------- Metric Decrease: MultiLayerModulesTH_OneShot Metric Increase: T12425 T13386 hard_hole_fits ------------------------- - - - - - c5d89412 by Zubin Duggal at 2024-05-13T22:19:53-04:00 Don't store a GlobalRdrEnv in `mi_globals` for GHCi. GHCi only needs the `mi_globals` field for modules imported with :module +*SomeModule. It uses this field to make the top level environment in `SomeModule` available to the repl. By default, only the first target in the command line parameters is "star" loaded into GHCi. Other modules have to be manually "star" loaded into the repl. Storing the top level GlobalRdrEnv for each module is very wasteful, especially given that we will most likely never need most of these environments. Instead we store only the information needed to reconstruct the top level environment in a module, which is the `IfaceTopEnv` data structure, consisting of all import statements as well as all top level symbols defined in the module (not taking export lists into account) When a particular module is "star-loaded" into GHCi (as the first commandline target, or via an explicit `:module +*SomeModule`, we reconstruct the top level environment on demand using the `IfaceTopEnv`. - - - - - d65bf4a2 by Fendor at 2024-05-13T22:20:30-04:00 Add perf regression test for `-fwrite-if-simplified-core` - - - - - 2c0f8ddb by Andrei Borzenkov at 2024-05-13T22:21:07-04:00 Improve pattern to type pattern transformation (23739) `pat_to_type_pat` function now can handle more patterns: - TuplePat - ListPat - LitPat - NPat - ConPat Allowing these new constructors in type patterns significantly increases possible shapes of type patterns without `type` keyword. This patch also changes how lookups in `lookupOccRnConstr` are performed, because we need to fall back into types when we didn't find a constructor on data level to perform `ConPat` to type transformation properly. - - - - - be514bb4 by Cheng Shao at 2024-05-13T22:21:43-04:00 hadrian: fix hadrian building with ghc-9.10.1 - - - - - ad38e954 by Cheng Shao at 2024-05-13T22:21:43-04:00 linters: fix lint-whitespace compilation with ghc-9.10.1 - - - - - a593f284 by Andreas Klebinger at 2024-05-15T07:32:10-04:00 Expand the `inline` rule to look through casts/ticks. Fixes #24808 - - - - - b1e0c313 by Cheng Shao at 2024-05-15T07:32:46-04:00 testsuite: bump PartialDownSweep timeout to 5x on wasm32 - - - - - b2227487 by Fendor at 2024-05-15T17:14:06-04:00 Add Eq and Ord instance to `IfaceType` We add an `Ord` instance so that we can store `IfaceType` in a `Data.Map` container. This is required to deduplicate `IfaceType` while writing `.hi` files to disk. Deduplication has many beneficial consequences to both file size and memory usage, as the deduplication enables implicit sharing of values. See issue #24540 for more motivation. The `Ord` instance would be unnecessary if we used a `TrieMap` instead of `Data.Map` for the deduplication process. While in theory this is clerarly the better option, experiments on the agda code base showed that a `TrieMap` implementation has worse run-time performance characteristics. To the change itself, we mostly derive `Eq` and `Ord`. This requires us to change occurrences of `FastString` with `LexicalFastString`, since `FastString` has no `Ord` instance. We change the definition of `IfLclName` to a newtype of `LexicalFastString`, to make such changes in the future easier. Bump haddock submodule for IfLclName changes - - - - - d368f9a6 by Fendor at 2024-05-15T17:14:06-04:00 Move out LiteralMap to avoid cyclic module dependencies - - - - - 2fcc09fd by Fendor at 2024-05-15T17:14:06-04:00 Add deduplication table for `IfaceType` The type `IfaceType` is a highly redundant, tree-like data structure. While benchmarking, we realised that the high redundancy of `IfaceType` causes high memory consumption in GHCi sessions when byte code is embedded into the `.hi` file via `-fwrite-if-simplified-core` or `-fbyte-code-and-object-code`. Loading such `.hi` files from disk introduces many duplicates of memory expensive values in `IfaceType`, such as `IfaceTyCon`, `IfaceTyConApp`, `IA_Arg` and many more. We improve the memory behaviour of GHCi by adding an additional deduplication table for `IfaceType` to the serialisation of `ModIface`, similar to how we deduplicate `Name`s and `FastString`s. When reading the interface file back, the table allows us to automatically share identical values of `IfaceType`. To provide some numbers, we evaluated this patch on the agda code base. We loaded the full library from the `.hi` files, which contained the embedded core expressions (`-fwrite-if-simplified-core`). Before this patch: * Load time: 11.7 s, 2.5 GB maximum residency. After this patch: * Load time: 7.3 s, 1.7 GB maximum residency. This deduplication has the beneficial side effect to additionally reduce the size of the on-disk interface files tremendously. For example, on agda, we reduce the size of `.hi` files (with `-fwrite-if-simplified-core`): * Before: 101 MB on disk * Now: 24 MB on disk This has even a beneficial side effect on the cabal store. We reduce the size of the store on disk: * Before: 341 MB on disk * Now: 310 MB on disk Note, none of the dependencies have been compiled with `-fwrite-if-simplified-core`, but `IfaceType` occurs in multiple locations in a `ModIface`. We also add IfaceType deduplication table to .hie serialisation and refactor .hie file serialisation to use the same infrastrucutre as `putWithTables`. Bump haddock submodule to accomodate for changes to the deduplication table layout and binary interface. - - - - - 36aa7cf1 by Fendor at 2024-05-15T17:14:06-04:00 Add run-time configurability of `.hi` file compression Introduce the flag `-fwrite-if-compression=<n>` which allows to configure the compression level of writing .hi files. The motivation is that some deduplication operations are too expensive for the average use case. Hence, we introduce multiple compression levels with variable impact on performance, but still reduce the memory residency and `.hi` file size on disk considerably. We introduce three compression levels: * `1`: `Normal` mode. This is the least amount of compression. It deduplicates only `Name` and `FastString`s, and is naturally the fastest compression mode. * `2`: `Safe` mode. It has a noticeable impact on .hi file size and is marginally slower than `Normal` mode. In general, it should be safe to always use `Safe` mode. * `3`: `Full` deduplication mode. Deduplicate as much as we can, resulting in minimal .hi files, but at the cost of additional compilation time. Reading .hi files doesn't need to know the initial compression level, and can always deserialise a `ModIface`, as we write out a byte that indicates the next value has been deduplicated. This allows users to experiment with different compression levels for packages, without recompilation of dependencies. Note, the deduplication also has an additional side effect of reduced memory consumption to implicit sharing of deduplicated elements. See https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for example where that matters. ------------------------- Metric Decrease: MultiLayerModulesDefsGhciWithCore T16875 T21839c T24471 hard_hole_fits libdir ------------------------- - - - - - 1e63a6fb by Matthew Pickering at 2024-05-15T17:14:07-04:00 Introduce regression tests for `.hi` file sizes Add regression tests to track how `-fwrite-if-compression` levels affect the size of `.hi` files. - - - - - 639d742b by M Farkas-Dyck at 2024-05-15T17:14:49-04:00 TTG: ApplicativeStatement exist only in Rn and Tc Co-Authored-By: romes <rodrigo.m.mesquita at gmail.com> - - - - - aa7b336b by Jade at 2024-05-15T23:06:17-04:00 Documentation: Improve documentation for symbols exported from System.IO - - - - - c561de8f by Jade at 2024-05-15T23:06:54-04:00 Improve suggestions for language extensions - When suggesting Language extensions, also suggest Extensions which imply them - Suggest ExplicitForAll and GADTSyntax instead of more specific extensions - Rephrase suggestion to include the term 'Extension' - Also moves some flag specific definitions out of Session.hs into Flags.hs (#24478) Fixes: #24477 Fixes: #24448 Fixes: #10893 - - - - - 4c7ae2a1 by Andreas Klebinger at 2024-05-15T23:07:30-04:00 Testsuite: Check if llvm assembler is available for have_llvm - - - - - bc672166 by Torsten Schmits at 2024-05-15T23:08:06-04:00 refactor quadratic search in warnMissingHomeModules - - - - - 7875e8cb by Torsten Schmits at 2024-05-15T23:08:06-04:00 add test that runs MakeDepend on thousands of modules - - - - - b84b91f5 by Adam Gundry at 2024-05-16T15:32:06-04:00 Representation-polymorphic HasField (fixes #22156) This generalises the HasField class to support representation polymorphism, so that instead of type HasField :: forall {k} . k -> Type -> Type -> Constraint we have type HasField :: forall {k} {r_rep} {a_rep} . k -> TYPE r_rep -> TYPE a_rep -> Constraint - - - - - 05285090 by Matthew Pickering at 2024-05-16T15:32:43-04:00 Bump os-string submodule to 2.0.2.2 Closes #24786 - - - - - 886ab43a by Cheng Shao at 2024-05-17T01:34:50-04:00 rts: do not prefetch mark_closure bdescr in non-moving gc when ASSERTS_ENABLED This commit fixes a small an oversight in !12148: the prefetch logic in non-moving GC may trap in debug RTS because it calls Bdescr() for mark_closure which may be a static one. It's fine in non-debug RTS because even invalid bdescr addresses are prefetched, they will not cause segfaults, so this commit implements the most straightforward fix: don't prefetch mark_closure bdescr when assertions are enabled. - - - - - b38dcf39 by Teo Camarasu at 2024-05-17T01:34:50-04:00 rts: Allocate non-moving segments with megablocks Non-moving segments are 8 blocks long and need to be aligned. Previously we serviced allocations by grabbing 15 blocks, finding an aligned 8 block group in it and returning the rest. This proved to lead to high levels of fragmentation as a de-allocating a segment caused an 8 block gap to form, and this could not be reused for allocation. This patch introduces a segment allocator based around using entire megablocks to service segment allocations in bulk. When there are no free segments, we grab an entire megablock and fill it with aligned segments. As the megablock is free, we can easily guarantee alignment. Any unused segments are placed on a free list. It only makes sense to free segments in bulk when all of the segments in a megablock are freeable. After sweeping, we grab the free list, sort it, and find all groups of segments where they cover the megablock and free them. This introduces a period of time when free segments are not available to the mutator, but the risk that this would lead to excessive allocation is low. Right after sweep, we should have an abundance of partially full segments, and this pruning step is relatively quick. In implementing this we drop the logic that kept NONMOVING_MAX_FREE segments on the free list. We also introduce an eventlog event to log the amount of pruned/retained free segments. See Note [Segment allocation strategy] Resolves #24150 ------------------------- Metric Decrease: T13253 T19695 ------------------------- - - - - - 710665bd by Cheng Shao at 2024-05-17T01:35:30-04:00 rts: fix I/O manager compilation errors for win32 target This patch fixes I/O manager compilation errors for win32 target discovered when cross-compiling to win32 using recent clang: ``` rts/win32/ThrIOManager.c:117:7: error: error: call to undeclared function 'is_io_mng_native_p'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 117 | if (is_io_mng_native_p ()) { | ^ | 117 | if (is_io_mng_native_p ()) { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/fs.c:143:28: error: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] 143 | int setErrNoFromWin32Error () { | ^ | void | 143 | int setErrNoFromWin32Error () { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:227:9: error: error: call to undeclared function 'interruptIOManagerEvent'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 227 | interruptIOManagerEvent (); | ^ | 227 | interruptIOManagerEvent (); | ^ rts/win32/ConsoleHandler.c:227:9: error: note: did you mean 'getIOManagerEvent'? | 227 | interruptIOManagerEvent (); | ^ rts/include/rts/IOInterface.h:27:10: error: note: 'getIOManagerEvent' declared here 27 | void * getIOManagerEvent (void); | ^ | 27 | void * getIOManagerEvent (void); | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:196:9: error: error: call to undeclared function 'setThreadLabel'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/win32/ConsoleHandler.c:196:9: error: note: did you mean 'postThreadLabel'? | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/eventlog/EventLog.h:118:6: error: note: 'postThreadLabel' declared here 118 | void postThreadLabel(Capability *cap, | ^ | 118 | void postThreadLabel(Capability *cap, | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) ``` - - - - - 28b9cee0 by Rodrigo Mesquita at 2024-05-17T01:36:05-04:00 configure: Check C99-compat for Cmm preprocessor Fixes #24815 - - - - - 8927e0c3 by Andreas Klebinger at 2024-05-17T01:36:41-04:00 Ensure `tcHasFixedRuntimeRep (# #)` returns True. - - - - - 04179044 by doyougnu at 2024-05-17T09:00:32-04:00 testsuite: make find_so regex less general Closes #24759 Background. In MR !12372 we began tracking shared object files and directories sizes for dependencies. However, this broke release builds because release builds alter the filenames swapping "in-place" for a hash. This was not considered in the MR and thus broke release pipelines. Furthermore, the rts_so test was found to be wildly varying and was therefore disabled in !12561. This commit fixes both of these issues: - fix the rts_so test by making the regex less general, now the rts_so test and all other foo.so tests must match "libHS<some-lib>-<version>-<hash|'in-place>-<ghc>". This prevents the rts_so test from accidentally matching different rts variants such as rts_threaded, which was the cause of the wild swings after !12372. - add logic to match either a hash or the string in-place. This should make the find_so function build agnostic. - - - - - 0962b50d by Andreas Klebinger at 2024-05-17T09:01:08-04:00 TagAnalysis: Treat all bottom ids as tagged during analysis. Ticket #24806 showed that we also need to treat dead end thunks as tagged during the analysis. - - - - - 7eb9f184 by Ben Gamari at 2024-05-17T11:23:37-04:00 Remove haddock submodule In preparation for merge into the GHC, as proposed in #23178. - - - - - 47b14dcc by Fendor at 2024-05-17T11:28:17-04:00 Adapt to `IfLclName` newtype changes (cherry picked from commit a711607e29b925f3d69e27c5fde4ba655c711ff1) - - - - - 6cc6681d by Fendor at 2024-05-17T11:28:17-04:00 Add IfaceType deduplication table to interface file serialisation Although we do not really need it in the interface file serialisation, as the deserialisation uses `getWithUserData`, we need to mirror the structure `getWithUserData` expects. Thus, we write essentially an empty `IfaceType` table at the end of the file, as the interface file doesn't reference `IfaceType`. (cherry picked from commit c9bc29c6a708483d2abc3d8ec9262510ce87ca61) - - - - - b9721206 by Ben Gamari at 2024-05-17T11:30:22-04:00 ghc-tags.yaml: Initial commit - - - - - 074e7d8f by Ben Gamari at 2024-05-17T11:31:29-04:00 fourmolu: Add configuration - - - - - 151b1736 by Ben Gamari at 2024-05-17T11:32:52-04:00 Makefile: Rework for use by haddock developers Previously the Makefile was present only for GHC's old make-based build system. Now since the make-based build system is gone we can use it for more useful ends. - - - - - a7dcf13b by Ben Gamari at 2024-05-17T11:36:14-04:00 Reformat with fourmolu Using previously-added configuration and `fourmolu -i .` Note that we exclude the test-cases (`./{hoogle,html-hypsrc,latex}-test`) as they are sensitive to formatting. - - - - - 0ea6017b by Ben Gamari at 2024-05-17T11:40:04-04:00 Add 'utils/haddock/' from commit 'a7dcf13bfbb97b20e75cc8ce650e2bb628db4660' git-subtree-dir: utils/haddock git-subtree-mainline: 7eb9f1849b1c72a1c61dee88462b4244550406f3 git-subtree-split: a7dcf13bfbb97b20e75cc8ce650e2bb628db4660 - - - - - aba1d304 by Hécate Moonlight at 2024-05-17T11:40:48-04:00 Add exceptions to the dangling notes list - - - - - 527bfbfb by Hécate Moonlight at 2024-05-17T11:40:52-04:00 Add haddock to the whitespace lint ignore list - - - - - 43274677 by Ben Gamari at 2024-05-17T11:41:20-04:00 git-blame-ignore-revs: Ignore haddock reformatting - - - - - 0e679e37 by Fendor at 2024-05-18T00:27:24-04:00 Pass cpp options to the CC builder in hadrian - - - - - bb40244e by Sylvain Henry at 2024-05-18T00:28:06-04:00 JS: fix allocation constant (fix #24746) - - - - - 646d30ab by Jade at 2024-05-18T19:23:31+02:00 Add highlighting for inline-code snippets in haddock - - - - - 64459a3e by Hécate Moonlight at 2024-05-19T08:42:27-04:00 haddock: Add a .readthedocs.yml file for online documentation - - - - - 7d3d9bbf by Serge S. Gulin at 2024-05-19T18:47:05+00:00 Unicode: General Category size test (related #24789) Added trivial size performance test which involves unicode general category usage via `read`. The `read` itself uses general category to detect spaces. The purpose for this test is to measure outcome of applying improvements at General Category representation in code discussed at #24789. - - - - - 8e04efcf by Alan Zimmerman at 2024-05-19T21:29:34-04:00 EPA: Remove redundant code Remove unused epAnnAnns function various cases for showAstData that no longer exist - - - - - 071d7a1e by Rodrigo Mesquita at 2024-05-20T10:55:16-04:00 Improve docs on closed type families in hs-boots Fixes #24776 - - - - - d9e2c119 by Torsten Schmits at 2024-05-20T10:55:52-04:00 Use default deviation for large-project test This new performance test has the purpose of detecting regressions in complexity in relation to the number of modules in a project, so 1% deviation is way too small to avoid false positives. - - - - - 20b0136a by Ben Gamari at 2024-05-22T00:31:39-04:00 ghcup-metadata: Various fixes from 9.10.1 Use Debian 12/x86-64, Debian 10/aarch64, and Debian 11/aarch64 bindists where possible. - - - - - 6838a7c3 by Sylvain Henry at 2024-05-22T00:32:23-04:00 Reverse arguments to stgCallocBytes (fix #24828) - - - - - f50f46c3 by Fendor at 2024-05-22T00:32:59-04:00 Add log messages for Iface serialisation compression level Fix the label of the number of 'IfaceType' entries in the log message. Add log message for the compression level that is used to serialise a an interface file. Adds `Outputable` instance for 'CompressionIFace'. - - - - - 3bad5d55 by Hécate Moonlight at 2024-05-22T00:33:40-04:00 base: Update doctests outputs ghc-internal: Update doctests outputs - - - - - 9317c6fb by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix the testsuites of the haddock-library - Apply all the metadata revisions from Hackage to the cabal file. - Fix the `ParserSpec.hs` file in the `spec` testsuite of haddock-library. - Make `CHANGES.md` an extra-doc-file instead of an extra-source-file. - - - - - 54073b02 by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix parser of @since pragma The testsuite contained tests for annotations of the form `@since foo-bar-0.5.0`, but the parser was written incorrectly. - - - - - ede6ede3 by Matthew Pickering at 2024-05-22T00:34:57-04:00 Fix nightly pages job It seems likely broken by 9f99126a which moved `index.html` from the root folder into `docs/` folder. Fixes #24840 - - - - - b7bcf729 by Cheng Shao at 2024-05-22T00:35:32-04:00 autoconf: remove unused context diff check This patch removes redundant autoconf check for the context diff program given it isn't actually been used anywhere, especially since make removal. - - - - - ea2fe66e by Hécate Moonlight at 2024-05-22T00:36:13-04:00 haddock: Rework the contributing guide - - - - - 0f302a94 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 haddock: Add module relationships diagrams of haddock-api and haddock-library - - - - - d1a9f34f by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add instructions - - - - - b880ee80 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add SVG outputs - - - - - 6d7e6ad8 by Ben Gamari at 2024-05-22T13:40:05-04:00 rts: Fix size of StgOrigThunkInfo frames Previously the entry code of the `stg_orig_thunk` frame failed to account for the size of the profiling header as it hard-coded the frame size. Fix this. Fixes #24809. - - - - - c645fe40 by Fendor at 2024-05-22T13:40:05-04:00 Add regression test T24809 for stg_orig_thunk_info_frame size - - - - - 4181aa40 by Andreas Klebinger at 2024-05-22T13:40:42-04:00 bindists: Check for existence of share folder before trying to copy it. This folder isn't distributed in windows bindists A lack of doing so resulted us copying loads of files twice. - - - - - d216510e by Matthew Pickering at 2024-05-22T13:40:42-04:00 Remove ad-hoc installation of mingw toolchain in relocatable bindists This reverts 616ac30026e8dd7d2ebb98d92dde071eedf5d951 The choice about whether to install mingw is taken in the installation makefile. This is also broken on non-windows systems. The actual issue was the EnableDistroToolchain variable wasn't declared in mk/config.mk and therefore the check to install mingw was failing. - - - - - 7b4c1998 by Cheng Shao at 2024-05-22T21:52:52-04:00 testsuite: fix T17920 for wasm backend T17920 was marked as fragile on wasm before; it can be trivially fixed by avoiding calling variadic printf() in cmm. - - - - - c739383b by Cheng Shao at 2024-05-22T21:53:29-04:00 testsuite: bump T22744 timeout to 5x - - - - - c4c6d714 by Cheng Shao at 2024-05-22T21:54:06-04:00 testsuite: don't attempt to detect host cpu features when testing cross ghc The testsuite driver CPU feature detection logic only detects host CPU and only makes sense when we are not testing a cross GHC. - - - - - 3d9e4ce6 by Simon Peyton Jones at 2024-05-22T21:54:43-04:00 Better skolemisation As #24810 showed, it is (a little) better to skolemise en-bloc, so that Note [Let-bound skolems] fires more often. See Note [Skolemisation en bloc] in GHC.Tc.Utils.Instantiate. - - - - - a3cd3a1d by Ryan Scott at 2024-05-22T21:55:19-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. - - - - - 4bb2a7cc by Hécate Moonlight at 2024-05-22T21:55:59-04:00 [base] Document the memory overhead of ByteArray Add a diagram that shows the constituent parts of a ByteArray and their memory overhead. - - - - - 8b2a016a by Hécate Moonlight at 2024-05-22T21:56:38-04:00 Haddock: Add MR template for Haddock - - - - - ead75532 by Peter Trommler at 2024-05-23T02:28:05-04:00 PPC: Support ELF v2 on powerpc64 big-endian Detect ELF v2 on PowerPC 64-bit systems. Check for `_CALL_ELF` preprocessor macro. Fixes #21191 - - - - - 9d4c10f2 by Hécate Kleidukos at 2024-05-23T02:28:44-04:00 gitlab: Add @Kleidukos to CODEOWNERS for utils/haddock - - - - - 28e64170 by Preetham Gujjula at 2024-05-23T07:20:48-04:00 haddock: Add cabal-fmt to tools for `make style` - - - - - 00126a89 by Andrei Borzenkov at 2024-05-23T07:21:24-04:00 haddock: fix verbosity option parsing - - - - - a3e0b68b by Ryan Hendrickson at 2024-05-23T15:52:03-04:00 base: specify tie-breaking behavior of min, max, and related list/Foldable functions - - - - - bdcc0f37 by doyougnu at 2024-05-24T07:51:18-04:00 cmm: add word <-> double/float bitcast - closes: #25331 This is the last step in the project plan described in #25331. This commit: - adds bitcast operands for x86_64, LLVM, aarch64 - For PPC and i386 we resort to using the cmm implementations - renames conversion MachOps from Conv to Round|Truncate - - - - - f0d257f7 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor Some functions in StgToByteCode were filtering out void arguments. However, StgToByteCode is called after unarisation: the void arguments should have been removed earlier. Instead of filtering out, we assert that the args are non-void. - - - - - 03137fd2 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor `layoutNativeCall` was always called with a `primRepCmmType platform` callback. Hence we can put it inside of `layoutNativeCall` rather than repeat it. - - - - - 27c430f3 by David Binder at 2024-05-24T07:52:38-04:00 haddock: Remove compatibility shims for GHC < 8.4 from haddock-library - - - - - 8dd8a076 by Cheng Shao at 2024-05-24T07:53:14-04:00 compiler: avoid saving foreign call target to local when there are no caller-save GlobalRegs This patch makes the STG->Cmm backend avoid saving foreign call target to local when there are no caller-save GlobalRegs. Since 321941a8ebe25192cdeece723e1058f2f47809ea, when we lower a foreign call, we unconditionally save the foreign call target to a temporary local first, then rely on cmmSink to clean it up later, which only happens with -fcmm-sink (implied by -O) and not in unoptimized code. And this is troublesome for the wasm backend NCG, which needs to infer a foreign call target symbol's type signature from the Cmm call site. Previously, the NCG has been emitting incorrect type signatures for unoptimized code, which happens to work with `wasm-ld` most of the time, but this is never future-proof against upstream toolchain updates, and it causes horrible breakages when LTO objects are included in linker input. Hence this patch. - - - - - 986df1ab by Cheng Shao at 2024-05-24T07:53:14-04:00 testsuite: add callee-no-local regression test - - - - - 52d62e2a by Sylvain Henry at 2024-05-24T07:53:57-04:00 Fix HasCallStack leftovers from !12514 / #24726 - - - - - c5e00c35 by crumbtoo at 2024-05-24T07:54:38-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 - - - - - bd323b0e by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Ensure that CHANGELOG is included in extra-source-files This was missed in the `ghc-internal` split. Closes #24831. - - - - - 1bfd32e8 by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Fix changelog reference to setBacktraceMechanismState (cherry picked from commit b63f7ba01fdfd98a01d2f0dec8d9262b3e595c5d) - - - - - 43e8e4f3 by Sylvain Henry at 2024-05-24T12:16:43-04:00 Float/double unboxed literal support for HexFloatLiterals (fix #22155) - - - - - 4a7f4713 by Fendor at 2024-05-24T12:17:19-04:00 Improve test labels for binary interface file size tests Test labels for binary interface file sizes are hard to read and overly verbose at the same time. Extend the name for the metric title, but shorten it in the actual comparison table. - - - - - 14e554cf by Zubin Duggal at 2024-05-24T12:17:55-04:00 Revert "Fix haskell/haddock#783 Don't show button if --quickjump not present" This reverts commit 7776566531e72c415f66dd3b13da9041c52076aa. - - - - - f56838c3 by Ben Gamari at 2024-05-24T12:17:55-04:00 Fix default hyperlinked sources pattern Previously this didn't include the `%M` token which manifested as broken links to the hyperlinked sources of reexports of declarations defined in other packages. Fixes haddock#1628. (cherry picked from commit 1432bcc943d41736eca491ecec4eb9a6304dab36) - - - - - 42efa62c by Ben Gamari at 2024-05-24T12:17:55-04:00 Make DocPaths a proper data type (cherry picked from commit 7f3a5c4da0023ae47b4c376c9b1ea2d706c94d8c) - - - - - 53d9ceb3 by Ben Gamari at 2024-05-24T12:17:55-04:00 haddock: Bump version to 2.30 (cherry picked from commit 994989ed3d535177e57b778629726aeabe8c7602) - - - - - e4db1112 by Zubin Duggal at 2024-05-24T12:17:55-04:00 haddock-api: allow base 4.20 and ghc 9.11 - - - - - e294f7a2 by PHO at 2024-05-24T12:17:55-04:00 Add a flag "threaded" for building haddock with the threaded RTS GHC isn't guaranteed to have a threaded RTS. There should be a way to build it with the vanilla one. (cherry picked from commit 75a94e010fb5b0236c670d22b04f5472397dc15d) - - - - - 51165bc9 by Andreas Klebinger at 2024-05-25T10:58:03-04:00 Update ticky counter event docs. Add the info about the info table address and json fields. Fixes #23200 - - - - - 98597ad5 by Sylvain Henry at 2024-05-25T10:58:45-04:00 Export extractPromotedList (#24866) This can be useful in plugins. - - - - - 228dcae6 by Teo Camarasu at 2024-05-28T13:12:24+00:00 template-haskell: Move wired-ins to ghc-internal Thus we make `template-haskell` reinstallable and keep it as the public API for Template Haskell. All of the wired-in identifiers are moved to `ghc-internal`. This necessitates also moving much of `ghc-boot-th` into `ghc-internal`. These modules are then re-exported from `ghc-boot-th` and `template-haskell`. To avoid a dependency on `template-haskell` from `lib:ghc`, we instead depend on the TH ASTs via `ghc-boot-th`. As `template-haskell` no longer has special status, we can drop the logic adding an implicit dependency on `template-haskell` when using TH. We can also drop the `template-haskell-next` package, which was previously used when bootstrapping. When bootstrapping, we need to vendor the TH AST modules from `ghc-internal` into `ghc-boot-th`. This is controlled by the `bootstrap` cabal flag as before. See Note [Bootstrapping Template Haskell]. We split out a GHC.Internal.TH.Lift module resolving #24752. This module is only built when not bootstrapping. Resolves #24703 ------------------------- Metric Increase: ghc_boot_th_dir ghc_boot_th_so ------------------------- - - - - - 62dded28 by Teo Camarasu at 2024-05-28T13:12:24+00:00 testsuite: mark tests broken by #24886 Now that `template-haskell` is no longer wired-in. These tests are triggering #24886, and so need to be marked broken. - - - - - 3ca72ad9 by Cheng Shao at 2024-05-30T02:57:06-04:00 rts: fix missing function prototypes in ClosureMacros.h - - - - - e0029e3d by Andreas Klebinger at 2024-05-30T02:57:43-04:00 UnliftedFFITypes: Allow `(# #)` as argument when it's the only argument. This allows representing functions like: int foo(void); to be imported like this: foreign import ccall "a_number_c" c_number :: (# #) -> Int64# Which can be useful when the imported function isn't implicitly stateful. - - - - - d0401335 by Matthew Pickering at 2024-05-30T02:58:19-04:00 ci: Update ci-images commit for fedora38 image The fedora38 nightly job has been failing for quite a while because `diff` was no longer installed. The ci-images bump explicitly installs `diffutils` into these images so hopefully they now pass again. - - - - - 3c97c74a by Jan Hrček at 2024-05-30T02:58:58-04:00 Update exactprint docs - - - - - 77760cd7 by Jan Hrček at 2024-05-30T02:58:58-04:00 Incorporate review feedback - - - - - 87591368 by Jan Hrček at 2024-05-30T02:58:58-04:00 Remove no longer relevant reference to comments - - - - - 05f4f142 by Jan Hrček at 2024-05-30T02:58:59-04:00 Replace outdated code example - - - - - 45a4a5f3 by Andreas Klebinger at 2024-05-30T02:59:34-04:00 Reword error resulting from missing -XBangPatterns. It can be the result of either a bang pattern or strict binding, so now we say so instead of claiming it must be a bang pattern. Fixes #21032 - - - - - e17f2df9 by Cheng Shao at 2024-05-30T03:00:10-04:00 testsuite: bump MultiLayerModulesDefsGhciReload timeout to 10x - - - - - 7a660042 by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - c77a48af by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 05c4fafb by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 5ff83bfc by Sylvain Henry at 2024-05-30T14:43:10-04:00 JS: remove useless h$CLOCK_REALTIME (#23202) - - - - - 95ef2d58 by Matthew Pickering at 2024-05-30T14:43:47-04:00 ghcup-metadata: Fix metadata generation There were some syntax errors in the generation script which were preventing it from running. I have tested this with: ``` nix shell --extra-experimental-features nix-command -f .gitlab/rel_eng -c ghcup-metadata --metadata ghcup-0.0.7.yaml --date="2024-05-27" --pipeline-id=95534 --version=9.11.20240525 ``` which completed successfully. - - - - - 1bc66ee4 by Jakob Bruenker at 2024-05-30T14:44:22-04:00 Add diagrams to Arrows documentation This adds diagrams to the documentation of Arrows, similar to the ones found on https://www.haskell.org/arrows/. It does not add diagrams for ArrowChoice for the time being, mainly because it's not clear to me how to visually distinguish them from the ones for Arrow. Ideally, you might want to do something like highlight the arrows belonging to the same tuple or same Either in common colors, but that's not really possible with unicode. - - - - - d10a1c65 by Matthew Craven at 2024-05-30T23:35:48-04:00 Make UnsafeSNat et al. into pattern synonyms ...so that they do not cause coerce to bypass the nominal role on the corresponding singleton types when they are imported. See Note [Preventing unsafe coercions for singleton types] and the discussion at #23478. This also introduces unsafeWithSNatCo (and analogues for Char and Symbol) so that users can still access the dangerous coercions that importing the real constructors would allow, but only in a very localized way. - - - - - 0958937e by Cheng Shao at 2024-05-30T23:36:25-04:00 hadrian: build C/C++ with split sections when enabled When split sections is enabled, ensure -fsplit-sections is passed to GHC as well when invoking GHC to compile C/C++; and pass -ffunction-sections -fdata-sections to gcc/clang when compiling C/C++ with the hadrian Cc builder. Fixes #23381. - - - - - 02b1f91e by Cheng Shao at 2024-05-30T23:36:25-04:00 driver: build C/C++ with -ffunction-sections -fdata-sections when split sections is enabled When -fsplit-sections is passed to GHC, pass -ffunction-sections -fdata-sections to gcc/clang when building C/C++. Previously, -fsplit-sections was only respected by the NCG/LLVM backends, but not the unregisterised backend; the GHC driver did not pass -fdata-sections and -ffunction-sections to the C compiler, which resulted in excessive executable sizes. Fixes #23381. ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - fd47e2e3 by Cheng Shao at 2024-05-30T23:37:00-04:00 testsuite: mark process005 as fragile on JS - - - - - 34a04ea1 by Matthew Pickering at 2024-05-31T06:08:36-04:00 Add -Wderiving-typeable to -Wall Deriving `Typeable` does nothing, and it hasn't done for a long while. There has also been a warning for a long while which warns you about uselessly deriving it but it wasn't enabled in -Wall. Fixes #24784 - - - - - 75fa7b0b by Matthew Pickering at 2024-05-31T06:08:36-04:00 docs: Fix formatting of changelog entries - - - - - 303c4b33 by Preetham Gujjula at 2024-05-31T06:09:21-04:00 docs: Fix link to injective type families paper Closes #24863 - - - - - df97e9a6 by Ben Gamari at 2024-05-31T06:09:57-04:00 ghc-internal: Fix package description The previous description was inherited from `base` and was inappropriate for `ghc-internal`. Also fix the maintainer and bug reporting fields. Closes #24906. - - - - - bf0737c0 by Cheng Shao at 2024-05-31T06:10:33-04:00 compiler: remove ArchWasm32 special case in cmmDoCmmSwitchPlans This patch removes special consideration for ArchWasm32 in cmmDoCmmSwitchPlans, which means the compiler will now disable cmmImplementSwitchPlans for wasm unreg backend, just like unreg backend of other targets. We enabled it in the past to workaround some compile-time panic in older versions of LLVM, but those panics are no longer present, hence no need to keep this workaround. - - - - - 7eda4bd2 by Cheng Shao at 2024-05-31T15:52:04-04:00 utils: add hie.yaml config file for ghc-config Add hie.yaml to ghc-config project directory so it can be edited using HLS. - - - - - 1e5752f6 by Cheng Shao at 2024-05-31T15:52:05-04:00 hadrian: handle findExecutable "" gracefully hadrian may invoke findExecutable "" at run-time due to a certain program is not found by configure script. Which is fine and findExecutable is supposed to return Nothing in this case. However, on Windows there's a directory bug that throws an exception (see https://github.com/haskell/directory/issues/180), so we might as well use a wrapper for findExecutable and handle exceptions gracefully. - - - - - 4eb5ad09 by Cheng Shao at 2024-05-31T15:52:05-04:00 configure: do not set LLC/OPT/LLVMAS fallback values when FIND_LLVM_PROG fails When configure fails to find LLC/OPT/LLVMAS within supported version range, it used to set "llc"/"opt"/"clang" as fallback values. This behavior is particularly troublesome when the user has llc/opt/clang with other versions in their PATH and run the testsuite, since hadrian will incorrectly assume have_llvm=True and pass that to the testsuite driver, resulting in annoying optllvm test failures (#23186). If configure determines llc/opt/clang wouldn't work, then we shouldn't pretend it'll work at all, and the bindist configure will invoke FIND_LLVM_PROG check again at install time anyway. - - - - - 5f1afdf7 by Sylvain Henry at 2024-05-31T15:52:52-04:00 Introduce UniqueSet and use it to replace 'UniqSet Unique' 'UniqSet Unique' represents a set of uniques as a 'Map Unique Unique', which is wasting space (associated key/value are always the same). Fix #23572 and #23605 - - - - - e0aa42b9 by crumbtoo at 2024-05-31T15:53:33-04:00 Improve template-haskell haddocks Closes #15822 - - - - - ae170155 by Olivier Benz at 2024-06-01T09:35:17-04:00 Bump max LLVM version to 19 (not inclusive) - - - - - 92aa65ea by Matthew Pickering at 2024-06-01T09:35:17-04:00 ci: Update CI images to test LLVM 18 The debian12 image in this commit has llvm 18 installed. - - - - - adb1fe42 by Serge S. Gulin at 2024-06-01T09:35:53-04:00 Unicode: make ucd2haskell build-able again ucd2haskell tool used streamly library which version in cabal was out of date. It is updated to the latest version at hackage with deprecated parts rewritten. Also following fixes were applied to existing code in suppose that from its last run the code structure was changed and now it was required to be up to date with actual folder structures: 1. Ghc module path environment got a suffix with `src`. 2. Generated code got 2.1 `GHC.Internal` prefix for `Data.*`. 2.2 `GHC.Unicode.Internal` swapped on `GHC.Internal.Unicode` according to actual structure. - - - - - ad56fd84 by Jade at 2024-06-01T09:36:29-04:00 Replace 'NB' with 'Note' in error messages - - - - - 6346c669 by Cheng Shao at 2024-06-01T09:37:04-04:00 compiler: fix -ddump-cmm-raw when compiling .cmm This patch fixes missing -ddump-cmm-raw output when compiling .cmm, which is useful for debugging cmm related codegen issues. - - - - - 1c834ad4 by Ryan Scott at 2024-06-01T09:37:40-04:00 Print namespace specifiers in FixitySig's Outputable instance For whatever reason, the `Outputable` instance for `FixitySig` simply did not print out namespace specifiers, leading to the confusing `-ddump-splices` output seen in #24911. This patch corrects this oversight. Fixes #24911. - - - - - cf49fb5f by Sylvain Henry at 2024-06-01T09:38:19-04:00 Configure: display C++ compiler path - - - - - f9c1ae12 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable PIC for in-tree GMP on wasm32 This patch disables PIC for in-tree GMP on wasm32 target. Enabling PIC unconditionally adds undesired code size and runtime overhead for wasm32. - - - - - 1a32f828 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable in-tree gmp fft code path for wasm32 This patch disables in-tree GMP FFT code paths for wasm32 target in order to give up some performance of multiplying very large operands in exchange for reduced code size. - - - - - 06277d56 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: build in-tree GMP with malloc-notreentrant on wasm32 This patch makes hadrian build in-tree GMP with the --enable-alloca=malloc-notreentrant configure option. We will only need malloc-reentrant when we have threaded RTS and SMP support on wasm32, which will take some time to happen, before which we should use malloc-notreentrant to avoid undesired runtime overhead. - - - - - 9f614270 by ARATA Mizuki at 2024-06-02T14:02:35-04:00 Set package include paths when assembling .S files Fixes #24839. Co-authored-by: Sylvain Henry <hsyl20 at gmail.com> - - - - - 4998a6ed by Alex Mason at 2024-06-03T02:09:29-04:00 Improve performance of genericWordQuotRem2Op (#22966) Implements the algorithm from compiler-rt's udiv128by64to64default. This rewrite results in a roughly 24x improvement in runtime on AArch64 (and likely any other arch that uses it). - - - - - ae50a8eb by Cheng Shao at 2024-06-03T02:10:05-04:00 testsuite: mark T7773 as fragile on wasm - - - - - c8ece0df by Fendor at 2024-06-03T19:43:22-04:00 Migrate `Finder` component to `OsPath`, fixed #24616 For each module in a GHCi session, we keep alive one `ModLocation`. A `ModLocation` is fairly inefficiently packed, as `String`s are expensive in memory usage. While benchmarking the agda codebase, we concluded that we keep alive around 11MB of `FilePath`'s, solely retained by `ModLocation`. We provide a more densely packed encoding of `ModLocation`, by moving from `FilePath` to `OsPath`. Further, we migrate the full `Finder` component to `OsPath` to avoid unnecessary transformations. As the `Finder` component is well-encapsulated, this requires only a minimal amount of changes in other modules. We introduce pattern synonym for 'ModLocation' which maintains backwards compatibility and avoids breaking consumers of 'ModLocation'. - - - - - 0cff083a by Cheng Shao at 2024-06-03T19:43:58-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. - - - - - 18f63970 by Sebastian Graf at 2024-06-04T05:05:27-04:00 Parser: Remove unused `apats` rule - - - - - 38757c30 by David Knothe at 2024-06-04T05:05:27-04:00 Implement Or Patterns (#22596) This commit introduces a new language extension, `-XOrPatterns`, as described in GHC Proposal 522. An or-pattern `pat1; ...; patk` succeeds iff one of the patterns `pat1`, ..., `patk` succeed, in this order. See also the summary `Note [Implmentation of OrPatterns]`. Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> - - - - - 395412e8 by Cheng Shao at 2024-06-04T05:06:04-04:00 compiler/ghci/rts: remove stdcall support completely We have formally dropped i386 windows support (#18487) a long time ago. The stdcall foreign call convention is only used by i386 windows, and the legacy logic around it is a significant maintenance burden for future work that adds arm64 windows support (#24603). Therefore, this patch removes stdcall support completely from the compiler as well as the RTS (#24883): - stdcall is still recognized as a FFI calling convention in Haskell syntax. GHC will now unconditionally emit a warning (-Wunsupported-calling-conventions) and treat it as ccall. - Apart from minimum logic to support the parsing and warning logic, all other code paths related to stdcall has been completely stripped from the compiler. - ghci only supports FFI_DEFAULT_ABI and ccall convention from now on. - FFI foreign export adjustor code on all platforms no longer handles the stdcall case and only handles ccall from now on. - The Win32 specific parts of RTS no longer has special code paths for stdcall. This commit is the final nail on the coffin for i386 windows support. Further commits will perform more housecleaning to strip the legacy code paths and pave way for future arm64 windows support. - - - - - d1fe9ab6 by Cheng Shao at 2024-06-04T05:06:04-04:00 rts: remove legacy i386 windows code paths This commit removes some legacy i386 windows related code paths in the RTS, given this target is no longer supported. - - - - - a605e4b2 by Cheng Shao at 2024-06-04T05:06:04-04:00 autoconf: remove i386 windows related logic This commit removes legacy i386 windows logic in autoconf scripts. - - - - - 91e5ac5e by Cheng Shao at 2024-06-04T05:06:04-04:00 llvm-targets: remove i386 windows support This commit removes i386 windows from llvm-targets and the script to generate it. - - - - - 65fe75a4 by Cheng Shao at 2024-06-04T05:06:04-04:00 libraries/utils: remove stdcall related legacy logic This commit removes stdcall related legacy logic in libraries and utils. ccall should be used uniformly for all supported windows hosts from now on. - - - - - d2a83302 by Cheng Shao at 2024-06-04T05:06:04-04:00 testsuite: adapt the testsuite for stdcall removal This patch adjusts test cases to handle the stdcall removal: - Some stdcall usages are replaced with ccall since stdcall doesn't make sense anymore. - We also preserve some stdcall usages, and check in the expected warning messages to ensure GHC always warn about stdcall usages (-Wunsupported-calling-conventions) as expected. - Error code testsuite coverage is slightly improved, -Wunsupported-calling-conventions is now tested. - Obsolete code paths related to i386 windows are also removed. - - - - - cef8f47a by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: minor adjustments for stdcall removal This commit include minor adjustments of documentation related to stdcall removal. - - - - - 54332437 by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: mention i386 Windows removal in 9.12 changelog This commit mentions removal of i386 Windows support and stdcall related change in the 9.12 changelog. - - - - - 2aaea8a1 by Cheng Shao at 2024-06-04T05:06:40-04:00 hadrian: improve user settings documentation This patch adds minor improvements to hadrian user settings documentation: - Add missing `ghc.cpp.opts` case - Remove non-existent `cxx` case - Clarify `cc.c.opts` also works for C++, while `cc.deps.opts` doesn't - Add example of passing configure argument to autoconf packages - - - - - 71010381 by Alex Mason at 2024-06-04T12:09:07-04:00 Add AArch64 CLZ, CTZ, RBIT primop implementations. Adds support for emitting the clz and rbit instructions, which are used by GHC.Prim.clz*#, GHC.Prim.ctz*# and GHC.Prim.bitReverse*#. - - - - - 44e2abfb by Cheng Shao at 2024-06-04T12:09:43-04:00 hadrian: add +text_simdutf flavour transformer to allow building text with simdutf This patch adds a +text_simdutf flavour transformer to hadrian to allow downstream packagers and users that build from source to opt-in simdutf support for text, in order to benefit from SIMD speedup at run-time. It's still disabled by default for the time being. - - - - - 077cb2e1 by Cheng Shao at 2024-06-04T12:09:43-04:00 ci: enable +text_simdutf flavour transformer for wasm jobs This commit enables +text_simdutf flavour transformer for wasm jobs, so text is now built with simdutf support for wasm. - - - - - b23746ad by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Use TemplateHaskellQuotes in instance Lift ByteArray Resolves #24852 - - - - - 3fd25743 by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Mark addrToByteArray as NOINLINE This function should never be inlined in order to keep code size small. - - - - - 98ad1ea5 by Cheng Shao at 2024-06-04T22:51:26-04:00 compiler: remove unused CompilerInfo/LinkerInfo types This patch removes CompilerInfo/LinkerInfo types from the compiler since they aren't actually used anywhere. - - - - - 11795244 by Cheng Shao at 2024-06-05T06:33:17-04:00 rts: remove unused PowerPC/IA64 native adjustor code This commit removes unused PowerPC/IA64 native adjustor code which is never actually enabled by autoconf/hadrian. Fixes #24920. - - - - - 5132754b by Sylvain Henry at 2024-06-05T06:33:57-04:00 RTS: fix warnings with doing*Profiling (#24918) - - - - - accc8c33 by Cheng Shao at 2024-06-05T11:35:36-04:00 hadrian: don't depend on inplace/mingw when --enable-distro-toolchain on Windows - - - - - 6ffbd678 by Cheng Shao at 2024-06-05T11:35:37-04:00 autoconf: normalize paths of some build-time dependencies on Windows This commit applies path normalization via cygpath -m to some build-time dependencies on Windows. Without this logic, the /clang64/bin prefixed msys2-style paths cause the build to fail with --enable-distro-toolchain. - - - - - 075dc6d4 by Cheng Shao at 2024-06-05T11:36:12-04:00 hadrian: remove OSDarwin mention from speedHack This commit removes mentioning of OSDarwin from speedHack, since speedHack is purely for i386 and we no longer support i386 darwin (#24921). - - - - - 83235c4c by Cheng Shao at 2024-06-05T11:36:12-04:00 compiler: remove 32-bit darwin logic This commit removes all 32-bit darwin logic from the compiler, given we no longer support 32-bit apple systems (#24921). Also contains a bit more cleanup of obsolete i386 windows logic. - - - - - 1eb99bc3 by Cheng Shao at 2024-06-05T11:36:12-04:00 rts: remove 32-bit darwin/ios logic This commit removes 32-bit darwin/ios related logic from the rts, given we no longer support them (#24921). - - - - - 24f65892 by Cheng Shao at 2024-06-05T11:36:12-04:00 llvm-targets: remove 32-bit darwin/ios targets This commit removes 32-bit darwin/ios targets from llvm-targets given we no longer support them (#24921). - - - - - ccdbd689 by Cheng Shao at 2024-06-05T11:36:12-04:00 testsuite: remove 32-bit darwin logic This commit removes 32-bit darwin logic from the testsuite given it's no longer supported (#24921). Also contains more cleanup of obsolete i386 windows logic. - - - - - 11d661c4 by Cheng Shao at 2024-06-05T11:36:13-04:00 docs: mention 32-bit darwin/ios removal in 9.12 changelog This commit mentions removal of 32-bit darwin/ios support (#24921) in the 9.12 changelog. - - - - - 7c173310 by Georgi Lyubenov at 2024-06-05T15:17:22-04:00 Add firstA and secondA to Data.Bitraversable Please see https://github.com/haskell/core-libraries-committee/issues/172 for related discussion - - - - - 3b6f9fd1 by Ben Gamari at 2024-06-05T15:17:59-04:00 base: Fix name of changelog Fixes #24899. Also place it under `extra-doc-files` to better reflect its nature and avoid triggering unnecessary recompilation if it changes. - - - - - 1f4d2ef7 by Sebastian Graf at 2024-06-05T15:18:34-04:00 Announce Or-patterns in the release notes for GHC 9.12 (#22596) Leftover from !9229. - - - - - 8650338d by Jan Hrček at 2024-06-06T10:39:24-04:00 Improve haddocks of Language.Haskell.Syntax.Pat.Pat - - - - - 2eee65e1 by Cheng Shao at 2024-06-06T10:40:00-04:00 testsuite: bump T7653 timeout for wasm - - - - - 990fed60 by Sylvain Henry at 2024-06-07T14:45:23-04:00 StgToCmm: refactor opTranslate and friends - Change arguments order to avoid `\args -> ...` lambdas - Fix documentation - Rename StgToCmm options ("big" doesn't mean anything) - - - - - 1afad514 by Sylvain Henry at 2024-06-07T14:45:23-04:00 NCG x86: remove dead code (#5444) Since 6755d833af8c21bbad6585144b10e20ac4a0a1ab this code is dead. - - - - - 595c0894 by Cheng Shao at 2024-06-07T14:45:58-04:00 testsuite: skip objc-hi/objcxx-hi when cross compiling objc-hi/objcxx-hi should be skipped when cross compiling. The existing opsys('darwin') predicate only asserts the host system is darwin but tells us nothing about the target, hence the oversight. - - - - - edfe6140 by qqwy at 2024-06-08T11:23:54-04:00 Replace '?callStack' implicit param with HasCallStack in GHC.Internal.Exception.throw - - - - - 35a64220 by Cheng Shao at 2024-06-08T11:24:30-04:00 rts: cleanup inlining logic This patch removes pre-C11 legacy code paths related to INLINE_HEADER/STATIC_INLINE/EXTERN_INLINE macros, ensure EXTERN_INLINE is treated as static inline in most cases (fixes #24945), and also corrects the comments accordingly. - - - - - 9ea90ed2 by Andrew Lelechenko at 2024-06-08T11:25:06-04:00 CODEOWNERS: add @core-libraries to track base interface changes A low-tech tactical solution for #24919 - - - - - 580fef7b by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update CHANGELOG to reflect current version - - - - - 391ecff5 by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update prologue.txt to reflect package description - - - - - 3dca3b7d by Ben Gamari at 2024-06-09T01:27:57-04:00 compiler: Clarify comment regarding need for MOVABS The comment wasn't clear in stating that it was only applicable to immediate source and memory target operands. - - - - - 6bd850e8 by doyougnu at 2024-06-09T21:02:14-04:00 JS: establish single source of truth for symbols In pursuit of: #22736. This MR moves ad-hoc symbols used throughout the js backend into a single symbols file. Why? First, this cleans up the code by removing ad-hoc strings created on the fly and therefore makes the code more maintainable. Second, it makes it much easier to eventually type these identifiers. - - - - - f3017dd3 by Cheng Shao at 2024-06-09T21:02:49-04:00 rts: replace ad-hoc MYTASK_USE_TLV with proper CC_SUPPORTS_TLS This patch replaces the ad-hoc `MYTASK_USE_TLV` with the `CC_SUPPORTS_TLS` macro. If TLS support is detected by autoconf, then we should use that for managing `myTask` in the threaded RTS. - - - - - e17d7e8c by Ben Gamari at 2024-06-11T05:25:21-04:00 users-guide: Fix stylistic issues in 9.12 release notes - - - - - 8a8a982a by Hugo Peters at 2024-06-11T05:25:57-04:00 fix typo in the simplifier debug output: baling -> bailing - - - - - 16475bb8 by Hécate Moonlight at 2024-06-12T03:07:55-04:00 haddock: Correct the Makefile to take into account Darwin systems - - - - - a2f60da5 by Hécate Kleidukos at 2024-06-12T03:08:35-04:00 haddock: Remove obsolete links to github.com/haskell/haddock in the docs - - - - - de4395cd by qqwy at 2024-06-12T03:09:12-04:00 Add `__GLASGOW_HASKELL_ASSERTS_IGNORED__` as CPP macro name if `-fasserts-ignored is set. This allows users to create their own Control.Exception.assert-like functionality that does something other than raising an `AssertFailed` exception. Fixes #24967 - - - - - 0e9c4dee by Ryan Hendrickson at 2024-06-12T03:09:53-04:00 compiler: add hint to TcRnBadlyStaged message - - - - - 2747cd34 by Simon Peyton Jones at 2024-06-12T12:51:37-04:00 Fix a QuickLook bug This MR fixes the bug exposed by #24676. The problem was that quickLookArg was trying to avoid calling tcInstFun unnecessarily; but it was in fact necessary. But that in turn forced me into a significant refactoring, putting more fields into EValArgQL. Highlights: see Note [Quick Look overview] in GHC.Tc.Gen.App * Instantiation variables are now distinguishable from ordinary unification variables, by level number = QLInstVar. This is treated like "level infinity". See Note [The QLInstVar TcLevel] in GHC.Tc.Utils.TcType. * In `tcApp`, we don't track the instantiation variables in a set Delta any more; instead, we just tell them apart by their level number. * EValArgQL now much more clearly captures the "half-done" state of typechecking an argument, ready for later resumption. See Note [Quick Look at value arguments] in GHC.Tc.Gen.App * Elminated a bogus (never used) fast-path in GHC.Tc.Utils.Instantiate.instCallConstraints See Note [Possible fast path for equality constraints] Many other small refactorings. - - - - - 1b1523b1 by George Thomas at 2024-06-12T12:52:18-04:00 Fix non-compiling extensible record `HasField` example - - - - - 97b141a3 by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Fix hyperlinker source urls (#24907) This fixes a bug introduced by f56838c36235febb224107fa62334ebfe9941aba Links to external modules in the hyperlinker are uniformly generated using splicing the template given to us instead of attempting to construct the url in an ad-hoc manner. - - - - - 954f864c by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Add name anchor to external source urls from documentation page URLs for external source links from documentation pages were missing a splice location for the name. Fixes #24912 - - - - - b0b64177 by Simon Peyton Jones at 2024-06-12T12:53:31-04:00 Prioritise nominal equalities The main payload of this patch is * Prioritise nominal equalities in the constraint solver. This ameliorates the incompleteness of solving for representational constraints over newtypes: see #24887. See (EX2) in Note [Decomposing newtype equalities] in GHC.Tc.Solver.Equality In doing this patch I tripped over some other things that I refactored: * Move `isCoVarType` from `GHC.Core.Type` to `GHC.Core.Predicate` where it seems more at home. * Clarify the "rewrite role" of a constraint. I was very puzzled about what the role of, say `(Eq a)` might be, but see the new Note [The rewrite-role of a constraint]. In doing so I made predTypeEqRel crash when given a non-equality. Usually it expects an equality; but it was being mis-used for the above rewrite-role stuff. - - - - - cb7c1b83 by Liam Goodacre at 2024-06-12T12:54:09-04:00 compiler: missing-deriving-strategies suggested fix Extends the missing-deriving-strategies warning with a suggested fix that includes which deriving strategies were assumed. For info about the warning, see comments for `TcRnNoDerivStratSpecified`, `TcRnNoDerivingClauseStrategySpecified`, & `TcRnNoStandaloneDerivingStrategySpecified`. For info about the suggested fix, see `SuggestExplicitDerivingClauseStrategies` & `SuggestExplicitStandalanoDerivingStrategy`. docs: Rewords missing-deriving-strategies to mention the suggested fix. Resolves #24955 - - - - - 4e36d3a3 by Jan Hrček at 2024-06-12T12:54:48-04:00 Further haddocks improvements in Language.Haskell.Syntax.Pat.Pat - - - - - 558353f4 by Cheng Shao at 2024-06-12T12:55:24-04:00 rts: use page sized mblocks on wasm This patch changes mblock size to page size on wasm. It allows us to simplify our wasi-libc fork, makes it much easier to test third party libc allocators like emmalloc/mimalloc, as well as experimenting with threaded RTS in wasm. - - - - - b3cc5366 by Matthew Pickering at 2024-06-12T23:06:57-04:00 compiler: Make ghc-experimental not wired in If you need to wire in definitions, then place them in ghc-internal and reexport them from ghc-experimental. Ticket #24903 - - - - - 700eeab9 by Hécate Kleidukos at 2024-06-12T23:07:37-04:00 base: Use a more appropriate unicode arrow for the ByteArray diagram This commit rectifies the usage of a unicode arrow in favour of one that doesn't provoke mis-alignment. - - - - - cca7de25 by Matthew Pickering at 2024-06-12T23:08:14-04:00 ghcup-metadata: Fix debian version ranges This was caught by `ghcup-ci` failing and attempting to install a deb12 bindist on deb11. ``` configure: WARNING: m4/prep_target_file.m4: Expecting YES/NO but got in ArSupportsDashL_STAGE0. Defaulting to False. bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by bin/ghc-toolchain-bin) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) ``` Fixes #24974 - - - - - 7b23ce8b by Pierre Le Marre at 2024-06-13T15:35:04-04:00 ucd2haskell: remove Streamly dependency + misc - Remove dead code. - Remove `streamly` dependency. - Process files with `bytestring`. - Replace Unicode files parsers with the corresponding ones from the package `unicode-data-parser`. - Simplify cabal file and rename module - Regenerate `ghc-internal` Unicode files with new header - - - - - 4570319f by Jacco Krijnen at 2024-06-13T15:35:41-04:00 Document how to run haddocks tests (#24976) Also remove ghc 9.7 requirement - - - - - fb629e24 by amesgen at 2024-06-14T00:28:20-04:00 compiler: refactor lower_CmmExpr_Ptr - - - - - def46c8c by amesgen at 2024-06-14T00:28:20-04:00 compiler: handle CmmRegOff in lower_CmmExpr_Ptr - - - - - ce76bf78 by Simon Peyton Jones at 2024-06-14T00:28:56-04:00 Small documentation update in Quick Look - - - - - 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 7a5a81e6 by Matthew Craven at 2024-09-01T16:57:55-04:00 Make pseq robust by using seq# Fixes #23699, fixes #23233, and adds a test for the latter. The implementation of par was near-by to and very similar to that of pseq, and has been adjusted in an analogous way, which also allows the deprecation of par# to resume, resolving #24825. The documentation around Control.Exception.evaluate and Note [seq# magic] has been expanded and improved, fixing #22935. The documentation around GHC.Magic.lazy and Note [lazyId magic] which was relevant to the old implementations of `pseq` and `par` has also been improved, fixing #25210. - - - - - 13 changed files: - .ghcid - + .git-blame-ignore-revs - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/generate-ci/generate-job-metadata - .gitlab/generate-ci/generate-jobs - .gitlab/issue_templates/bug.md → .gitlab/issue_templates/default.md The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7681296196e1075c5b7cf3f9abd303bb53226cd2...7a5a81e6fa3a45114c6b2973b66a31ffb7958b0a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7681296196e1075c5b7cf3f9abd303bb53226cd2...7a5a81e6fa3a45114c6b2973b66a31ffb7958b0a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 05:55:27 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 02 Sep 2024 01:55:27 -0400 Subject: [Git][ghc/ghc][wip/T25029] Add defaulting of equalities Message-ID: <66d5534f8df42_2a4f37e9ee18133fd@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25029 at Glasgow Haskell Compiler / GHC Commits: 9f06a233 by Simon Peyton Jones at 2024-09-02T07:54:57+02:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - 11 changed files: - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Dict.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Utils/Unify.hs - testsuite/tests/indexed-types/should_compile/Simple14.hs - − testsuite/tests/indexed-types/should_compile/Simple14.stderr - testsuite/tests/indexed-types/should_compile/all.T - + testsuite/tests/typecheck/should_compile/T25029.hs - + testsuite/tests/typecheck/should_compile/T25125.hs - testsuite/tests/typecheck/should_compile/all.T - testsuite/tests/typecheck/should_fail/T21338.stderr Changes: ===================================== compiler/GHC/Tc/Solver.hs ===================================== @@ -34,6 +34,7 @@ import GHC.Core.Class import GHC.Core import GHC.Core.DataCon import GHC.Core.Make +import GHC.Core.Coercion( mkNomReflCo ) import GHC.Driver.DynFlags import GHC.Data.FastString import GHC.Data.List.SetOps @@ -50,7 +51,8 @@ import GHC.Tc.Types.Evidence import GHC.Tc.Solver.Solve ( solveSimpleGivens, solveSimpleWanteds ) import GHC.Tc.Solver.Dict ( makeSuperClasses, solveCallStack ) import GHC.Tc.Solver.Rewrite ( rewriteType ) -import GHC.Tc.Utils.Unify ( buildTvImplication ) +import GHC.Tc.Utils.Unify ( buildTvImplication, touchabilityAndShapeTest + , simpleUnifyCheck, UnifyCheckCaller(..) ) import GHC.Tc.Utils.TcMType as TcM import GHC.Tc.Utils.Monad as TcM import GHC.Tc.Zonk.TcType as TcM @@ -85,8 +87,8 @@ import Data.Foldable ( toList, traverse_ ) import Data.List ( partition, intersect ) import Data.List.NonEmpty ( NonEmpty(..), nonEmpty ) import qualified Data.List.NonEmpty as NE -import Data.Monoid (First (First, getFirst)) -import GHC.Data.Maybe ( catMaybes, mapMaybe, runMaybeT, MaybeT ) +import GHC.Data.Maybe ( isJust, mapMaybe, catMaybes ) +import Data.Monoid ( First(..) ) {- ********************************************************************************* @@ -495,76 +497,70 @@ report_unsolved_equalities skol_info_anon skol_tvs tclvl wanted -- | Simplify top-level constraints, but without reporting any unsolved -- constraints nor unsafe overlapping. simplifyTopWanteds :: WantedConstraints -> TcS WantedConstraints - -- See Note [Top-level Defaulting Plan] simplifyTopWanteds wanteds - = do { wc_first_go <- nestTcS (solveWanteds wanteds) - -- This is where the main work happens - ; dflags <- getDynFlags - ; wc_defaulted <- try_tyvar_defaulting dflags wc_first_go - - -- See Note [Implementation of Unsatisfiable constraints] in GHC.Tc.Errors, - -- point (C). - ; useUnsatisfiableGivens wc_defaulted } - where - try_tyvar_defaulting :: DynFlags -> WantedConstraints -> TcS WantedConstraints - try_tyvar_defaulting dflags wc - | isEmptyWC wc - = return wc - | insolubleWC wc - , gopt Opt_PrintExplicitRuntimeReps dflags -- See Note [Defaulting insolubles] - = try_class_defaulting wc - | otherwise - = do { -- Need to zonk first, as the WantedConstraints are not yet zonked. - ; free_tvs <- TcS.zonkTyCoVarsAndFVList (tyCoVarsOfWCList wc) - ; let defaultable_tvs = filter can_default free_tvs - can_default tv - = isTyVar tv - -- Weed out coercion variables. - - && isMetaTyVar tv - -- Weed out runtime-skolems in GHCi, which we definitely - -- shouldn't try to default. - - && not (tv `elemVarSet` nonDefaultableTyVarsOfWC wc) - -- Weed out variables for which defaulting would be unhelpful, - -- e.g. alpha appearing in [W] alpha[conc] ~# rr[sk]. - - ; defaulted <- mapM defaultTyVarTcS defaultable_tvs -- Has unification side effects - ; if or defaulted - then do { wc_residual <- nestTcS (solveWanteds wc) - -- See Note [Must simplify after defaulting] - ; try_class_defaulting wc_residual } - else try_class_defaulting wc } -- No defaulting took place - - try_class_defaulting :: WantedConstraints -> TcS WantedConstraints - try_class_defaulting wc - | isEmptyWC wc || insolubleWC wc -- See Note [Defaulting insolubles] - = try_callstack_defaulting wc - | otherwise -- See Note [When to do type-class defaulting] - = do { something_happened <- applyDefaultingRules wc - -- See Note [Top-level Defaulting Plan] - ; if something_happened - then do { wc_residual <- nestTcS (solveWanteds wc) - ; try_class_defaulting wc_residual } - -- See Note [Overview of implicit CallStacks] in GHC.Tc.Types.Evidence - else try_callstack_defaulting wc } - - try_callstack_defaulting :: WantedConstraints -> TcS WantedConstraints - try_callstack_defaulting wc - = defaultConstraints [defaultCallStack, defaultExceptionContext] wc - --- | If an implication contains a Given of the form @Unsatisfiable msg@, use --- it to solve all Wanteds within the implication. + = do { -- Solve the constraints + wc_first_go <- nestTcS (solveWanteds wanteds) + + -- Now try defaulting: + -- see Note [Top-level Defaulting Plan] + ; tryDefaulting wc_first_go } + +-------------------------- +tryDefaulting :: WantedConstraints -> TcS WantedConstraints +tryDefaulting wc + = do { dflags <- getDynFlags + ; traceTcS "tryDefaulting:before" (ppr wc) + ; wc1 <- tryTyVarDefaulting dflags wc + ; wc2 <- tryConstraintDefaulting wc1 + ; wc3 <- tryTypeClassDefaulting wc2 + ; wc4 <- tryUnsatisfiableGivens wc3 + ; traceTcS "tryDefaulting:after" (ppr wc) + ; return wc4 } + +solveAgainIf :: Bool -> WantedConstraints -> TcS WantedConstraints +-- If the Bool is true, solve the wanted constraints again +-- See Note [Must simplify after defaulting] +solveAgainIf False wc = return wc +solveAgainIf True wc = nestTcS (solveWanteds wc) + +-------------------------- +tryTyVarDefaulting :: DynFlags -> WantedConstraints -> TcS WantedConstraints +tryTyVarDefaulting dflags wc + | isEmptyWC wc + = return wc + | insolubleWC wc + , gopt Opt_PrintExplicitRuntimeReps dflags -- See Note [Defaulting insolubles] + = return wc + | otherwise + = do { -- Need to zonk first, as the WantedConstraints are not yet zonked. + ; free_tvs <- TcS.zonkTyCoVarsAndFVList (tyCoVarsOfWCList wc) + ; let defaultable_tvs = filter can_default free_tvs + can_default tv + = isTyVar tv + -- Weed out coercion variables. + + && isMetaTyVar tv + -- Weed out runtime-skolems in GHCi, which we definitely + -- shouldn't try to default. + + && not (tv `elemVarSet` nonDefaultableTyVarsOfWC wc) + -- Weed out variables for which defaulting would be unhelpful, + -- e.g. alpha appearing in [W] alpha[conc] ~# rr[sk]. + + ; unification_s <- mapM defaultTyVarTcS defaultable_tvs -- Has unification side effects + ; solveAgainIf (or unification_s) wc } + -- solveAgainIf: see Note [Must simplify after defaulting] + +---------------------------- +-- | If an implication contains a Given of the form @Unsatisfiable msg@, +-- use it to solve all Wanteds within the implication. +-- See point (C) in Note [Implementation of Unsatisfiable constraints] in GHC.Tc.Errors. -- -- This does a complete walk over the implication tree. --- --- See point (C) in Note [Implementation of Unsatisfiable constraints] in GHC.Tc.Errors. -useUnsatisfiableGivens :: WantedConstraints -> TcS WantedConstraints -useUnsatisfiableGivens wc = +tryUnsatisfiableGivens :: WantedConstraints -> TcS WantedConstraints +tryUnsatisfiableGivens wc = do { (final_wc, did_work) <- (`runStateT` False) $ go_wc wc - ; if did_work - then nestTcS (solveWanteds final_wc) - else return final_wc } + ; solveAgainIf did_work final_wc } where go_wc (WC { wc_simple = wtds, wc_impl = impls, wc_errors = errs }) = do impls' <- mapMaybeBagM go_impl impls @@ -614,7 +610,7 @@ solveImplicationUsingUnsatGiven go_simple ct = case ctEvidence ct of CtWanted { ctev_pred = pty, ctev_dest = dst } -> do { ev_expr <- unsatisfiableEvExpr unsat_given pty - ; setWantedEvTerm dst True $ EvExpr ev_expr } + ; setWantedEvTerm dst EvNonCanonical $ EvExpr ev_expr } _ -> return () -- | Create an evidence expression for an arbitrary constraint using @@ -696,69 +692,122 @@ This allows us to indirectly box constraints with different representations (such as primitive equality constraints). -} --- | A 'TcS' action which can may default a 'Ct'. -type CtDefaultingStrategy = Ct -> MaybeT TcS () +-- | A 'TcS' action which can may solve a `Ct` +type CtDefaultingStrategy = Ct -> TcS Bool + -- True <=> I solved the constraint + +-------------------------------- +tryConstraintDefaulting :: WantedConstraints -> TcS WantedConstraints +-- See Note [Overview of implicit CallStacks] in GHC.Tc.Types.Evidence +tryConstraintDefaulting wc + | isEmptyWC wc + = return wc + | otherwise + = do { (n_unifs, better_wc) <- reportUnifications (go_wc wc) + -- We may have done unifications; so solve again + ; solveAgainIf (n_unifs > 0) better_wc } + where + go_wc :: WantedConstraints -> TcS WantedConstraints + go_wc wc@(WC { wc_simple = simples, wc_impl = implics }) + = do { mb_simples <- mapMaybeBagM go_simple simples + ; mb_implics <- mapMaybeBagM go_implic implics + ; return (wc { wc_simple = mb_simples, wc_impl = mb_implics }) } + + go_simple :: Ct -> TcS (Maybe Ct) + go_simple ct = do { solved <- tryCtDefaultingStrategy ct + ; if solved then return Nothing + else return (Just ct) } + + go_implic :: Implication -> TcS (Maybe Implication) + -- The Maybe is because solving the CallStack constraint + -- may well allow us to discard the implication entirely + go_implic implic + | isSolvedStatus (ic_status implic) + = return (Just implic) -- Nothing to solve inside here + | otherwise + = do { wanteds <- setEvBindsTcS (ic_binds implic) $ + -- defaultCallStack sets a binding, so + -- we must set the correct binding group + go_wc (ic_wanted implic) + ; setImplicationStatus (implic { ic_wanted = wanteds }) } + +tryCtDefaultingStrategy :: CtDefaultingStrategy +-- The composition of all the CtDefaultingStrategies we want +tryCtDefaultingStrategy + = foldr1 combineStrategies + [ defaultCallStack + , defaultExceptionContext + , defaultEquality ] -- | Default @ExceptionContext@ constraints to @emptyExceptionContext at . defaultExceptionContext :: CtDefaultingStrategy defaultExceptionContext ct - = do { ClassPred cls tys <- pure $ classifyPredType (ctPred ct) - ; Just {} <- pure $ isExceptionContextPred cls tys - ; emptyEC <- Var <$> lift (lookupId emptyExceptionContextName) + | ClassPred cls tys <- classifyPredType (ctPred ct) + , isJust (isExceptionContextPred cls tys) + = do { warnTcS $ TcRnDefaultedExceptionContext (ctLoc ct) + ; empty_ec_id <- lookupId emptyExceptionContextName ; let ev = ctEvidence ct - ; let ev_tm = mkEvCast emptyEC (wrapIP (ctEvPred ev)) - ; lift $ warnTcS $ TcRnDefaultedExceptionContext (ctLoc ct) - ; lift $ setEvBindIfWanted ev False ev_tm - } + ev_tm = mkEvCast (Var empty_ec_id) (wrapIP (ctEvPred ev)) + ; setEvBindIfWanted ev EvCanonical ev_tm + -- EvCanonical: see Note [CallStack and ExecptionContext hack] + -- in GHC.Tc.Solver.Dict + ; return True } + | otherwise + = return False -- | Default any remaining @CallStack@ constraints to empty @CallStack at s. -- See Note [Overview of implicit CallStacks] in GHC.Tc.Types.Evidence defaultCallStack :: CtDefaultingStrategy defaultCallStack ct - = do { ClassPred cls tys <- pure $ classifyPredType (ctPred ct) - ; Just {} <- pure $ isCallStackPred cls tys - ; lift $ solveCallStack (ctEvidence ct) EvCsEmpty - } - -defaultConstraints :: [CtDefaultingStrategy] - -> WantedConstraints - -> TcS WantedConstraints --- See Note [Overview of implicit CallStacks] in GHC.Tc.Types.Evidence -defaultConstraints defaulting_strategies wanteds - | isEmptyWC wanteds = return wanteds + | ClassPred cls tys <- classifyPredType (ctPred ct) + , isJust (isCallStackPred cls tys) + = do { solveCallStack (ctEvidence ct) EvCsEmpty + ; return True } | otherwise - = do simples <- handle_simples (wc_simple wanteds) - mb_implics <- mapBagM handle_implic (wc_impl wanteds) - return (wanteds { wc_simple = simples - , wc_impl = catBagMaybes mb_implics }) + = return False +defaultEquality :: CtDefaultingStrategy +-- See Note [Defaulting equalities] +defaultEquality ct + | EqPred NomEq ty1 ty2 <- classifyPredType (ctPred ct) + , Just tv1 <- getTyVar_maybe ty1 + = do { -- Remember: `ct` may not be zonked; + -- see (DE3) in Note [Defaulting equalities] + z_ty1 <- TcS.zonkTcTyVar tv1 + ; z_ty2 <- TcS.zonkTcType ty2 + ; case getTyVar_maybe z_ty1 of + Just z_tv1 | defaultable z_tv1 z_ty2 + -> do { default_tv z_tv1 z_ty2 + ; return True } + _ -> return False } + | otherwise + = return False where - handle_simples :: Bag Ct -> TcS (Bag Ct) - handle_simples simples - = catBagMaybes <$> mapBagM handle_simple simples - where - handle_simple :: Ct -> TcS (Maybe Ct) - handle_simple ct = go defaulting_strategies - where - go [] = return (Just ct) - go (f:fs) = do - mb <- runMaybeT (f ct) - case mb of - Just () -> return Nothing - Nothing -> go fs - - handle_implic :: Implication -> TcS (Maybe Implication) - -- The Maybe is because solving the CallStack constraint - -- may well allow us to discard the implication entirely - handle_implic implic - | isSolvedStatus (ic_status implic) - = return (Just implic) - | otherwise - = do { wanteds <- setEvBindsTcS (ic_binds implic) $ - -- defaultCallStack sets a binding, so - -- we must set the correct binding group - defaultConstraints defaulting_strategies (ic_wanted implic) - ; setImplicationStatus (implic { ic_wanted = wanteds }) } + defaultable tv1 ty2 + = -- Do the standard unification checks; + -- c.f. uUnfilledVar2 in GHC.Tc.Utils.Unify + -- EXCEPT drop the untouchability test + tyVarKind tv1 `tcEqType` typeKind ty2 + && touchabilityAndShapeTest topTcLevel tv1 ty2 + -- topTcLevel makes the untoucability test vacuous, + -- which is the Whole Point of `defaultEquality` + -- See (DE2) in Note [Defaulting equalities] + && simpleUnifyCheck UC_Defaulting tv1 ty2 + + default_tv tv1 ty2 + = do { unifyTyVar tv1 ty2 -- NB: unifyTyVar adds to the + -- TcS unification counter + ; setEvBindIfWanted (ctEvidence ct) EvCanonical $ + evCoercion (mkNomReflCo ty2) } + +combineStrategies :: CtDefaultingStrategy -> CtDefaultingStrategy -> CtDefaultingStrategy +combineStrategies default1 default2 ct + = do { solved <- default1 ct + ; case solved of + True -> return True -- default1 solved it! + False -> default2 ct -- default1 failed, try default2 + } + {- Note [When to do type-class defaulting] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -787,6 +836,56 @@ Another potential alternative would be to suppress *all* non-insoluble errors if there are *any* insoluble errors, anywhere, but that seems too drastic. +Note [Defaulting equalities] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Consider + f :: forall a. (forall t. (F t ~ Int) => a -> Int) -> Int + + g :: Int + g = f id + +We'll typecheck + id :: forall t. (F t ~ Int) => alpha[1] -> Int +where the `alpha[1]` comes from instantiating `f`. So we'll end up +with the implication constraint + forall[2] t. (F t ~ Int) => alpha[1] ~ Int +And that can't be solved because `alpha` is untouchable under the +equality (F t ~ Int). + +This is tiresome, and gave rise to user complaints: #25125 and #25029. +Moreover, in this case there is no good reason not to unify alpha:=Int. +Doing so solves the constraint, and since `alpha` is not otherwise +constrained, it does no harm. So the new plan is this: + + * For the Wanted constraint + [W] alpha ~ ty + if the only reason for not unifying is untouchability, then during + top-level defaulting, go ahead and unify + +In top-level defaulting, we already do several other somewhat-ad-hoc, +but terribly convenient, unifications. This is just one more. + +Wrinkles: + +(DE1) Note carefully that this does not threaten principal types. + The original worry about unifying untouchable type variables was this: + + data T a where + T1 :: T Bool + f x = case x of T1 -> True + + Should we infer f :: T a -> Bool, or f :: T a -> a. Both are valid, but + neither is more general than the other + +(DE2) We still can't unify if there is a skolem-escape check, or an occurs check, + or it it'd mean unifying a TyVarTv with a non-tyvar. It's only the + "untouchability test" that we lift. We can lift it by saying that the innermost + given equality is at top level. + +(DE3) The contraint we are looking at may not be fully zonked; for example, + an earlier deafaulting might have affected it. So we zonk-on-the fly in + `defaultEquality`. + Note [Don't default in syntactic equalities] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ When there are unsolved syntactic equalities such as @@ -1027,11 +1126,12 @@ last example above. ------------------ simplifyAmbiguityCheck :: Type -> WantedConstraints -> TcM () -simplifyAmbiguityCheck ty wanteds +simplifyAmbiguityCheck ty wc = do { traceTc "simplifyAmbiguityCheck {" $ - text "type = " <+> ppr ty $$ text "wanted = " <+> ppr wanteds + text "type = " <+> ppr ty $$ text "wanted = " <+> ppr wc - ; (final_wc, _) <- runTcS $ useUnsatisfiableGivens =<< solveWanteds wanteds + ; (final_wc, _) <- runTcS $ do { wc1 <- solveWanteds wc + ; tryUnsatisfiableGivens wc1 } -- NB: no defaulting! See Note [No defaulting in the ambiguity check] -- Note: we do still use Unsatisfiable Givens to solve Wanteds, -- see Wrinkle [Ambiguity] under point (C) of @@ -2881,11 +2981,11 @@ setImplicationStatus :: Implication -> TcS (Maybe Implication) -- setting the ic_status field -- Precondition: the ic_status field is not already IC_Solved -- Return Nothing if we can discard the implication altogether -setImplicationStatus implic@(Implic { ic_status = status +setImplicationStatus implic@(Implic { ic_status = old_status , ic_info = info , ic_wanted = wc , ic_given = givens }) - | assertPpr (not (isSolvedStatus status)) (ppr info) $ + | assertPpr (not (isSolvedStatus old_status)) (ppr info) $ -- Precondition: we only set the status if it is not already solved not (isSolvedWC pruned_wc) = do { traceTcS "setImplicationStatus(not-all-solved) {" (ppr implic) @@ -3389,28 +3489,34 @@ The constraint in f's signature is redundant; not used to typecheck be an ambiguous variable in `g`. -} +type UnificationDone = Bool + +noUnification, didUnification :: UnificationDone +noUnification = False +didUnification = True + -- | Like 'defaultTyVar', but in the TcS monad. -defaultTyVarTcS :: TcTyVar -> TcS Bool +defaultTyVarTcS :: TcTyVar -> TcS UnificationDone defaultTyVarTcS the_tv | isTyVarTyVar the_tv -- TyVarTvs should only be unified with a tyvar -- never with a type; c.f. GHC.Tc.Utils.TcMType.defaultTyVar -- and Note [Inferring kinds for type declarations] in GHC.Tc.TyCl - = return False + = return noUnification | isRuntimeRepVar the_tv = do { traceTcS "defaultTyVarTcS RuntimeRep" (ppr the_tv) ; unifyTyVar the_tv liftedRepTy - ; return True } + ; return didUnification } | isLevityVar the_tv = do { traceTcS "defaultTyVarTcS Levity" (ppr the_tv) ; unifyTyVar the_tv liftedDataConTy - ; return True } + ; return didUnification } | isMultiplicityVar the_tv = do { traceTcS "defaultTyVarTcS Multiplicity" (ppr the_tv) ; unifyTyVar the_tv ManyTy - ; return True } + ; return didUnification } | otherwise - = return False -- the common case + = return noUnification -- the common case approximateWC :: Bool -- See Wrinkle (W3) in Note [ApproximateWC] -> WantedConstraints @@ -3681,6 +3787,15 @@ Wrinkle (DP2): Interactions between defaulting mechanisms -} +tryTypeClassDefaulting :: WantedConstraints -> TcS WantedConstraints +tryTypeClassDefaulting wc + | isEmptyWC wc || insolubleWC wc -- See Note [Defaulting insolubles] + = return wc + | otherwise -- See Note [When to do type-class defaulting] + = do { something_happened <- applyDefaultingRules wc + -- See Note [Top-level Defaulting Plan] + ; solveAgainIf something_happened wc } + applyDefaultingRules :: WantedConstraints -> TcS Bool -- True <=> I did some defaulting, by unifying a meta-tyvar -- Input WantedConstraints are not necessarily zonked @@ -3720,8 +3835,10 @@ applyDefaultingRules wanteds ; traceTcS "applyDefaultingRules }" (ppr something_happeneds) ; return $ or something_happeneds || or plugin_defaulted } - where run_defaulting_plugin wanteds p = - do { groups <- runTcPluginTcS (p wanteds) + + where + run_defaulting_plugin wanteds p + = do { groups <- runTcPluginTcS (p wanteds) ; defaultedGroups <- filterM (\g -> disambigMultiGroup wanteds @@ -3737,9 +3854,7 @@ applyDefaultingRules wanteds -- Note [Defaulting plugins]). So we re-zonk to make sure later -- defaulting doesn't try to solve the same metavars. wanteds' <- TcS.zonkWC wanteds - return (wanteds', True) - } - + return (wanteds', True) } findDefaultableGroups :: ( [ClassDefaults] ===================================== compiler/GHC/Tc/Solver/Dict.hs ===================================== @@ -194,7 +194,7 @@ solveCallStack ev ev_cs {- Note [CallStack and ExecptionContext hack] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -It isn't really right thta we treat CallStack and ExceptionContext dictionaries +It isn't really right that we treat CallStack and ExceptionContext dictionaries as canonical, in the sense of Note [Coherence and specialisation: overview]. They definitely are not! ===================================== compiler/GHC/Tc/Solver/Monad.hs ===================================== @@ -819,7 +819,8 @@ data TcSEnv tcs_unified :: IORef Int, -- The number of unification variables we have filled - -- The important thing is whether it is non-zero + -- The important thing is whether it is non-zero, so it + -- could equally well be a Bool instead of an Int. tcs_unif_lvl :: IORef (Maybe TcLevel), -- The Unification Level Flag @@ -1312,6 +1313,9 @@ unifyTyVar tv ty ; TcM.updTcRef (tcs_unified env) (+1) } reportUnifications :: TcS a -> TcS (Int, a) +-- Record how many unifications are done by thing_inside +-- We could return a Bool instead of an Int; +-- all that matters is whether it is no-zero reportUnifications (TcS thing_inside) = TcS $ \ env -> do { inner_unified <- TcM.newTcRef 0 ===================================== compiler/GHC/Tc/Utils/Unify.hs ===================================== @@ -2906,18 +2906,20 @@ data UnifyCheckCaller = UC_OnTheFly -- Called from the on-the-fly unifier | UC_QuickLook -- Called from Quick Look | UC_Solver -- Called from constraint solver + | UC_Defaulting -- Called when doing top-level defaulting simpleUnifyCheck :: UnifyCheckCaller -> TcTyVar -> TcType -> Bool -- simpleUnifyCheck does a fast check: True <=> unification is OK -- If it says 'False' then unification might still be OK, but -- it'll take more work to do -- use the full checkTypeEq -- +-- * Rejects if lhs_tv occurs in rhs_ty (occurs check) -- * Rejects foralls unless -- lhs_tv is RuntimeUnk (used by GHCi debugger) -- or is a QL instantiation variable -- * Rejects a non-concrete type if lhs_tv is concrete -- * Rejects type families unless fam_ok=True --- * Does a level-check for type variables +-- * Does a level-check for type variables, to avoid skolem escape -- -- This function is pretty heavily used, so it's optimised not to allocate simpleUnifyCheck caller lhs_tv rhs @@ -2939,9 +2941,10 @@ simpleUnifyCheck caller lhs_tv rhs -- families, so we let it through there (not very principled, but let's -- see if it bites us) fam_ok = case caller of - UC_Solver -> True - UC_QuickLook -> True - UC_OnTheFly -> False + UC_Solver -> True + UC_QuickLook -> True + UC_OnTheFly -> False + UC_Defaulting -> True go (TyVarTy tv) | lhs_tv == tv = False ===================================== testsuite/tests/indexed-types/should_compile/Simple14.hs ===================================== @@ -1,6 +1,6 @@ {-# LANGUAGE Haskell2010 #-} {-# LANGUAGE TypeFamilies, RankNTypes, FlexibleContexts, ScopedTypeVariables #-} -{-# LANGUAGE AllowAmbiguousTypes #-} +{-# LANGUAGE TypeOperators, AllowAmbiguousTypes #-} module Simple14 where @@ -19,12 +19,11 @@ ntI :: (forall p. EQ_ x y -> p) -> EQ_ x y ntI x = error "ntI" foo :: forall m n. EQ_ (Maybe m) (Maybe n) -foo = ntI (\x -> x `eqE` (eqI :: EQ_ m n)) - --- Alternative --- foo = ntI (\eq -> eq `eqE` (eqI :: EQ_ m n)) +foo = ntI (\eq -> eq `eqE` (eqI :: EQ_ m n)) +-- Aug 2024: this test started passing with the fix to #25029 +-- See Note [Defaulting equalities] in GHC.Tc.Solver -- eq :: EQ_ (Maybe m) (Maybe n) -- Need (Maybe m ~ Maybe n) => EQ_ m n ~ EQ_ zeta zeta --- which reduces to (m~n) => m ~ zeta --- but then we are stuck +-- which reduces to (m~n) => m ~ zeta, but then +-- we were stuck; now we default zeta:=m in tryDefaulting ===================================== testsuite/tests/indexed-types/should_compile/Simple14.stderr deleted ===================================== @@ -1,21 +0,0 @@ - -Simple14.hs:22:27: error: [GHC-83865] - • Couldn't match type ‘z0’ with ‘n’ - Expected: EQ_ z0 z0 - Actual: EQ_ m n - ‘z0’ is untouchable - inside the constraints: Maybe m ~ Maybe n - bound by a type expected by the context: - (Maybe m ~ Maybe n) => EQ_ z0 z0 - at Simple14.hs:22:26-41 - ‘n’ is a rigid type variable bound by - the type signature for: - foo :: forall m n. EQ_ (Maybe m) (Maybe n) - at Simple14.hs:21:1-42 - • In the second argument of ‘eqE’, namely ‘(eqI :: EQ_ m n)’ - In the expression: x `eqE` (eqI :: EQ_ m n) - In the first argument of ‘ntI’, namely - ‘(\ x -> x `eqE` (eqI :: EQ_ m n))’ - • Relevant bindings include - x :: EQ_ (Maybe m) (Maybe n) (bound at Simple14.hs:22:13) - foo :: EQ_ (Maybe m) (Maybe n) (bound at Simple14.hs:22:1) ===================================== testsuite/tests/indexed-types/should_compile/all.T ===================================== @@ -11,7 +11,7 @@ test('Simple10', normal, compile, ['']) test('Simple11', normal, compile, ['']) test('Simple12', normal, compile, ['']) test('Simple13', normal, compile, ['']) -test('Simple14', normal, compile_fail, ['']) +test('Simple14', normal, compile, ['']) test('Simple15', normal, compile, ['']) test('Simple16', normal, compile, ['']) test('Simple17', normal, compile, ['']) ===================================== testsuite/tests/typecheck/should_compile/T25029.hs ===================================== @@ -0,0 +1,28 @@ +{-# LANGUAGE GADTs #-} +{-# LANGUAGE QuantifiedConstraints #-} +{-# LANGUAGE TypeFamilies #-} + +module T25029 where + +import Data.Kind (Type) +import GHC.TypeLits (Nat) + +type family RankedOf2 (f :: Type -> Type) :: Type -> Type +type family PrimalOf2 (f :: Type -> Type) :: Type -> Type + +rrev :: forall r r1 u. (Floating r, Floating r1) + => (forall f. ( RankedOf2 (PrimalOf2 f) ~ PrimalOf2 f + , (forall r2. Floating r2 => Floating (f r2)) +-- , f r1 ~ f u + ) + => f r -> f r1) + -> () +rrev f = () + +-- fails +testSin0RrevPP1 :: () +testSin0RrevPP1 = rrev @Double sin + +-- works +testSin0RrevPP2 :: () +testSin0RrevPP2 = rrev @Double @Double sin ===================================== testsuite/tests/typecheck/should_compile/T25125.hs ===================================== @@ -0,0 +1,27 @@ +{-# LANGUAGE DataKinds #-} +module T25125 where + +import GHC.TypeNats + +newtype NonEmptyText (n :: Nat) = NonEmptyText String + +toT :: NonEmptyText 10 -> String +toT = undefined + +fromT :: forall n. String -> NonEmptyText n +fromT t = undefined + +baz = () + where + validate :: forall n. (1 <= n) => NonEmptyText 10 -> (NonEmptyText n) + validate n = fromT (toT (check n)) + + + -- Giving a type signature works + --check :: forall n. (1 <= n) => NonEmptyText n -> AppM (NonEmptyText n) + check = check2 + -- Eta expanding check works + --check x = check2 x + + check2 :: forall n. (1 <= n) => NonEmptyText n -> (NonEmptyText n) + check2 inputText = undefined ===================================== testsuite/tests/typecheck/should_compile/all.T ===================================== @@ -920,3 +920,5 @@ test('T24810', normal, compile, ['']) test('T24887', normal, compile, ['']) test('T24938a', normal, compile, ['']) test('T25094', normal, compile, ['']) +test('T25029', normal, compile, ['']) +test('T25125', normal, compile, ['']) ===================================== testsuite/tests/typecheck/should_fail/T21338.stderr ===================================== @@ -1,27 +1,4 @@ -T21338.hs:38:24: error: [GHC-83865] - • Couldn't match type ‘flds0’ with ‘flds’ - Expected: NP (K String) flds - Actual: NP (K String) flds0 - ‘flds0’ is untouchable - inside the constraints: All flds0 - bound by a pattern with constructor: - Record :: forall (xs :: [*]). - All xs => - NP (K String) xs -> ConstructorInfo xs, - in a case alternative - at T21338.hs:38:3-11 - ‘flds’ is a rigid type variable bound by - the type signature for: - fieldNames :: forall a (flds :: [*]). NP (K String) flds - at T21338.hs:36:1-57 - • In the second argument of ‘hmap’, namely ‘np’ - In the expression: hmap id np - In a case alternative: Record np -> hmap id np - • Relevant bindings include - np :: NP (K String) flds0 (bound at T21338.hs:38:10) - fieldNames :: NP (K String) flds (bound at T21338.hs:37:1) - T21338.hs:39:8: error: [GHC-95781] • Cannot apply expression of type ‘h0 f0 xs0 -> h0 g0 xs0’ to a visible type argument ‘flds’ View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9f06a23376822bcad98555800734d139fd2de226 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9f06a23376822bcad98555800734d139fd2de226 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 09:20:16 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Mon, 02 Sep 2024 05:20:16 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 3 commits: Fix #25169 using Plan A from the ticket Message-ID: <66d58350b0ed1_22d969d681c291f9@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: e10aa6df by sheaf at 2024-09-02T11:19:53+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 528fc781 by sheaf at 2024-09-02T11:19:53+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 39341768 by sheaf at 2024-09-02T11:19:53+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Expr.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Parser.y - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToAsm/X86/Instr.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs - compiler/GHC/StgToByteCode.hs - compiler/GHC/StgToCmm/ArgRep.hs - compiler/GHC/StgToCmm/Foreign.hs - compiler/GHC/StgToCmm/Layout.hs - hadrian/src/Settings/Packages.hs - libraries/ghc-heap/tests/stack_misc_closures_c.c - rts/AutoApply.h - + rts/CheckVectorSupport.c - + rts/CheckVectorSupport.h - rts/CloneStack.c - rts/HeapStackCheck.cmm - rts/Interpreter.c - + rts/Jumps.h - + rts/Jumps_D.cmm - + rts/Jumps_V16.cmm - + rts/Jumps_V32.cmm - + rts/Jumps_V64.cmm - rts/Printer.c The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b792b70cb22976009486514d578ec10de0c90b63...39341768c8711d650e446bf20811201197eb8a15 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b792b70cb22976009486514d578ec10de0c90b63...39341768c8711d650e446bf20811201197eb8a15 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 09:40:45 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Mon, 02 Sep 2024 05:40:45 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 3 commits: Fix #25169 using Plan A from the ticket Message-ID: <66d5881d5b539_22d969e26303434f@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 3ffa5e27 by sheaf at 2024-09-02T11:40:37+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 203b131a by sheaf at 2024-09-02T11:40:37+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 530ca4af by sheaf at 2024-09-02T11:40:37+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Expr.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Parser.y - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToAsm/X86/Instr.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs - compiler/GHC/StgToByteCode.hs - compiler/GHC/StgToCmm/ArgRep.hs - compiler/GHC/StgToCmm/Foreign.hs - compiler/GHC/StgToCmm/Layout.hs - hadrian/src/Settings/Packages.hs - libraries/ghc-heap/tests/stack_misc_closures_c.c - rts/Apply.cmm - rts/AutoApply.h - + rts/CheckVectorSupport.c - + rts/CheckVectorSupport.h - rts/CloneStack.c - rts/HeapStackCheck.cmm - rts/Interpreter.c - + rts/Jumps.h - + rts/Jumps_D.cmm - + rts/Jumps_V16.cmm - + rts/Jumps_V32.cmm - + rts/Jumps_V64.cmm The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/39341768c8711d650e446bf20811201197eb8a15...530ca4af9e498128a6d27eb29f7ddb6359b6ea09 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/39341768c8711d650e446bf20811201197eb8a15...530ca4af9e498128a6d27eb29f7ddb6359b6ea09 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 09:50:17 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Mon, 02 Sep 2024 05:50:17 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] rts: Declare barf Message-ID: <66d58a59ed693_22d9693ea9f435438@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 0442507f by Ben Gamari at 2024-09-02T11:49:57+02:00 rts: Declare barf As it is used by -dtag-inference-checks. Closes #22066. - - - - - 1 changed file: - rts/include/stg/MiscClosures.h Changes: ===================================== rts/include/stg/MiscClosures.h ===================================== @@ -453,6 +453,9 @@ RTS_FUN_DECL(stg_threadFinished); RTS_FUN_DECL(StgReturn); +// Used by -dtag-inference-checks +RTS_FUN_DECL(barf); + /* ----------------------------------------------------------------------------- PrimOps -------------------------------------------------------------------------- */ View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0442507f541b1979b6c260f4061424045b7bb342 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0442507f541b1979b6c260f4061424045b7bb342 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 09:51:54 2024 From: gitlab at gitlab.haskell.org (Cheng Shao (@TerrorJack)) Date: Mon, 02 Sep 2024 05:51:54 -0400 Subject: [Git][ghc/ghc][wip/buildplan] 16 commits: haddock: include package info with --show-interface Message-ID: <66d58abaa0525_22d9693fcba43657@gitlab.mail> Cheng Shao pushed to branch wip/buildplan at Glasgow Haskell Compiler / GHC Commits: 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - d4aca884 by Cheng Shao at 2024-09-02T11:51:11+02:00 driver: implement --buildplan major mode to extract BuildPlan info from dependency analysis - - - - - 30 changed files: - .gitignore - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - − a.out - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/HsToCore/Usage.hs - compiler/GHC/Iface/Make.hs - compiler/GHC/Iface/Recomp.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/SysTools/Cpp.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Module/Graph.hs - compiler/GHC/Unit/Module/ModIface.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - configure.ac - docs/users_guide/9.12.1-notes.rst - docs/users_guide/expected-undocumented-flags.txt The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/200c6e624b73952dac593886313a6e63d9b92775...d4aca8849941016d23e44749c1197b54045f40b6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/200c6e624b73952dac593886313a6e63d9b92775...d4aca8849941016d23e44749c1197b54045f40b6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 09:56:52 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Mon, 02 Sep 2024 05:56:52 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 3 commits: Fix #25169 using Plan A from the ticket Message-ID: <66d58be47b571_18688adee68463c9@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 8df7cdb4 by sheaf at 2024-09-02T11:56:35+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 7f3a77a9 by sheaf at 2024-09-02T11:56:35+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 03e14e76 by sheaf at 2024-09-02T11:56:35+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Expr.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Parser.y - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToAsm/X86/Instr.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs - compiler/GHC/StgToByteCode.hs - compiler/GHC/StgToCmm/ArgRep.hs - compiler/GHC/StgToCmm/Foreign.hs - compiler/GHC/StgToCmm/Layout.hs - hadrian/src/Settings/Packages.hs - libraries/ghc-heap/tests/stack_misc_closures_c.c - rts/Apply.cmm - rts/AutoApply.h - + rts/CheckVectorSupport.c - + rts/CheckVectorSupport.h - rts/CloneStack.c - rts/HeapStackCheck.cmm - rts/Interpreter.c - + rts/Jumps.h - + rts/Jumps_D.cmm - + rts/Jumps_V16.cmm - + rts/Jumps_V32.cmm - + rts/Jumps_V64.cmm The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0442507f541b1979b6c260f4061424045b7bb342...03e14e76c8c382a16f15c35f2be09bccd418ab21 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0442507f541b1979b6c260f4061424045b7bb342...03e14e76c8c382a16f15c35f2be09bccd418ab21 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 09:58:14 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Mon, 02 Sep 2024 05:58:14 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 12 commits: The X86 SIMD patch. Message-ID: <66d58c36b908a_18688af77d84748b@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 86517c0f by sheaf at 2024-09-02T11:58:04+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - c7f7b018 by sheaf at 2024-09-02T11:58:04+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - da816518 by sheaf at 2024-09-02T11:58:05+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 32f3ecb1 by sheaf at 2024-09-02T11:58:05+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - c359f870 by sheaf at 2024-09-02T11:58:05+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - ae58000b by sheaf at 2024-09-02T11:58:05+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 09a3f0dd by sheaf at 2024-09-02T11:58:05+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 0a1c8364 by sheaf at 2024-09-02T11:58:06+02:00 Add test for C calls & SIMD vectors - - - - - 5355eb5c by sheaf at 2024-09-02T11:58:06+02:00 Add test for #25169 - - - - - 63c5fe85 by sheaf at 2024-09-02T11:58:06+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 8a6c0cab by sheaf at 2024-09-02T11:58:06+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - b8deddd6 by sheaf at 2024-09-02T11:58:06+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Expr.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/03e14e76c8c382a16f15c35f2be09bccd418ab21...b8deddd681b9fd1807905b01f59a51b6ec506ef0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/03e14e76c8c382a16f15c35f2be09bccd418ab21...b8deddd681b9fd1807905b01f59a51b6ec506ef0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 10:19:29 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Mon, 02 Sep 2024 06:19:29 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 12 commits: The X86 SIMD patch. Message-ID: <66d591313b604_18688a458d14546be@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 1cf51cc1 by sheaf at 2024-09-02T12:17:17+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 8052534e by sheaf at 2024-09-02T12:17:21+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 5a8ef415 by sheaf at 2024-09-02T12:17:21+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - a5484cab by sheaf at 2024-09-02T12:17:21+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - dac317f4 by sheaf at 2024-09-02T12:17:21+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 0fe18c4a by sheaf at 2024-09-02T12:17:21+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - c0d8ff6f by sheaf at 2024-09-02T12:17:22+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 76e9f66f by sheaf at 2024-09-02T12:17:22+02:00 Add test for C calls & SIMD vectors - - - - - 94474260 by sheaf at 2024-09-02T12:17:22+02:00 Add test for #25169 - - - - - 98c8015c by sheaf at 2024-09-02T12:17:22+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - d703050f by sheaf at 2024-09-02T12:18:13+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 61691275 by sheaf at 2024-09-02T12:19:17+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Expr.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b8deddd681b9fd1807905b01f59a51b6ec506ef0...61691275d45fd07552cf46da611cdad9c1fd77c7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b8deddd681b9fd1807905b01f59a51b6ec506ef0...61691275d45fd07552cf46da611cdad9c1fd77c7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 12:12:11 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Mon, 02 Sep 2024 08:12:11 -0400 Subject: [Git][ghc/ghc][wip/fltused] Use x86_64-unknown-windows-gnu target for LLVM on Windows Message-ID: <66d5ab9be1f44_12f96914cf4458836@gitlab.mail> sheaf pushed to branch wip/fltused at Glasgow Haskell Compiler / GHC Commits: 7c774ee6 by sheaf at 2024-09-02T14:11:59+02:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 3 changed files: - llvm-targets - m4/ghc_llvm_target.m4 - utils/llvm-targets/gen-data-layout.sh Changes: ===================================== llvm-targets ===================================== @@ -1,4 +1,4 @@ -[("x86_64-unknown-windows", ("e-m:w-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", "x86-64", "")) +[("x86_64-unknown-windows-gnu", ("e-m:w-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", "x86-64", "")) ,("arm-unknown-linux-gnueabi", ("e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64", "arm7tdmi", "+strict-align")) ,("arm-unknown-linux-gnueabihf", ("e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64", "arm1176jzf-s", "+strict-align")) ,("arm-unknown-linux-musleabihf", ("e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64", "arm1176jzf-s", "+strict-align")) ===================================== m4/ghc_llvm_target.m4 ===================================== @@ -14,7 +14,7 @@ AC_DEFUN([GHC_LLVM_TARGET], [ ;; *-mingw32|*-mingw64|*-msys) llvm_target_vendor="unknown" - llvm_target_os="windows" + llvm_target_os="windows-gnu" ;; # apple is a bit about their naming scheme for # aarch64; and clang on macOS doesn't know that ===================================== utils/llvm-targets/gen-data-layout.sh ===================================== @@ -26,7 +26,7 @@ TARGETS=( ######################### # Windows - "x86_64-unknown-windows" + "x86_64-unknown-windows-gnu" ######################### # Linux View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7c774ee639c62e4d144194514218e48a3225afe1 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7c774ee639c62e4d144194514218e48a3225afe1 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 12:23:26 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 02 Sep 2024 08:23:26 -0400 Subject: [Git][ghc/ghc][wip/andreask/setNonBlockingMode] Allow unknown fd device types for setNonBlockingMode. Message-ID: <66d5ae3ea98df_12f9694e16147318c@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/setNonBlockingMode at Glasgow Haskell Compiler / GHC Commits: 2eebba58 by Andreas Klebinger at 2024-09-02T14:04:43+02:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - 7 changed files: - libraries/base/changelog.md - libraries/ghc-internal/src/GHC/Internal/IO/FD.hs - libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/interface-stability/base-exports.stdout-ws-32 Changes: ===================================== libraries/base/changelog.md ===================================== @@ -16,6 +16,7 @@ * Add `firstA` and `secondA` to `Data.Bitraversable`. ([CLC proposal #172](https://github.com/haskell/core-libraries-committee/issues/172)) * Deprecate `GHC.TypeNats.Internal`, `GHC.TypeLits.Internal`, `GHC.ExecutionStack.Internal` ([CLC proposal #217](https://github.com/haskell/core-libraries-committee/issues/217)) * Define `Eq1`, `Ord1`, `Show1` and `Read1` instances for basic `Generic` representation types. ([CLC proposal #273](https://github.com/haskell/core-libraries-committee/issues/273)) + * `setNonBlockingMode` will no longer throw an exception when called on a FD associated with a unknown device type. ([CLC proposal #282](https://github.com/haskell/core-libraries-committee/issues/282)) * Add exception type metadata to default exception handler output. ([CLC proposal #231](https://github.com/haskell/core-libraries-committee/issues/231) and [CLC proposal #261](https://github.com/haskell/core-libraries-committee/issues/261)) ===================================== libraries/ghc-internal/src/GHC/Internal/IO/FD.hs ===================================== @@ -467,8 +467,8 @@ setNonBlockingMode fd set = do -- utilities inspecting fdIsNonBlocking (such as readRawBufferPtr) -- should not be tricked to think otherwise. is_nonblock <- if set then do - (fd_type, _, _) <- fdStat (fdFD fd) - pure $ fd_type /= RegularFile && fd_type /= RawDevice + fd_type <- statGetType_maybe (fdFD fd) + pure $ fd_type /= Just RegularFile && fd_type /= Just RawDevice else pure False setNonBlockingFD (fdFD fd) is_nonblock #if defined(mingw32_HOST_OS) ===================================== libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs ===================================== @@ -140,17 +140,30 @@ fdStat fd = fdType :: FD -> IO IODeviceType fdType fd = do (ty,_,_) <- fdStat fd; return ty +-- | Return a known device type or throw an exception if the device +-- type is unknown. statGetType :: Ptr CStat -> IO IODeviceType statGetType p_stat = do + dev_ty_m <- statGetType_maybe p_stat + case dev_ty_m of + Nothing -> ioError ioe_unknownfiletype + Just dev_ty -> pure dev_ty_m + +-- | Unlike @statGetType@, @statGetType_maybe@ will not throw an exception +-- if the CStat refers to a unknown device type. +-- +-- @since base-4.21.0.0 +statGetType_maybe :: Ptr CStat -> IO (Maybe IODeviceType) +statGetType_maybe p_stat = do c_mode <- st_mode p_stat :: IO CMode case () of - _ | s_isdir c_mode -> return Directory + _ | s_isdir c_mode -> return $ Just Directory | s_isfifo c_mode || s_issock c_mode || s_ischr c_mode - -> return Stream - | s_isreg c_mode -> return RegularFile + -> return $ Just Stream + | s_isreg c_mode -> return $ Just RegularFile -- Q: map char devices to RawDevice too? - | s_isblk c_mode -> return RawDevice - | otherwise -> ioError ioe_unknownfiletype + | s_isblk c_mode -> return $ Just RawDevice + | otherwise -> return Nothing ioe_unknownfiletype :: IOException ioe_unknownfiletype = IOError Nothing UnsupportedOperation "fdType" ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -10549,6 +10549,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10596,6 +10597,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs ===================================== @@ -13590,6 +13590,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -13637,6 +13638,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -10821,6 +10821,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10865,6 +10866,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.CWString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-ws-32 ===================================== @@ -10549,6 +10549,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10596,6 +10597,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2eebba58a534a49558338b507929192bd783d9e6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2eebba58a534a49558338b507929192bd783d9e6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 12:42:50 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 02 Sep 2024 08:42:50 -0400 Subject: [Git][ghc/ghc][wip/andreask/weakly_pinned] Add functions to check for weakly pinned arrays. Message-ID: <66d5b2cab99da_12f9695bc40878378@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/weakly_pinned at Glasgow Haskell Compiler / GHC Commits: 59153543 by Andreas Klebinger at 2024-09-02T14:23:57+02:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - 12 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/ffi.rst - libraries/base/src/GHC/Base.hs - libraries/base/src/GHC/Exts.hs - libraries/ghc-prim/changelog.md - rts/PrimOps.cmm - rts/RtsSymbols.c - rts/include/stg/MiscClosures.h - testsuite/tests/rts/T13894.hs Changes: ===================================== compiler/GHC/Builtin/primops.txt.pp ===================================== @@ -1925,7 +1925,25 @@ primop MutableByteArrayIsPinnedOp "isMutableByteArrayPinned#" GenPrimOp primop ByteArrayIsPinnedOp "isByteArrayPinned#" GenPrimOp ByteArray# -> Int# - {Determine whether a 'ByteArray#' is guaranteed not to move during GC.} + {Determine whether a 'ByteArray#' is guaranteed not to move.} + with out_of_line = True + +primop ByteArrayIsWeaklyPinnedOp "isByteArrayWeaklyPinned#" GenPrimOp + ByteArray# -> Int# + {Similar to 'isByteArrayPinned#'. Weakly pinned byte arrays are allowed + to be copied into compact regions by the user, potentially invalidating + the results of earlier calls to 'byteArrayContents#'. + + See the section `Pinned Byte Arrays` in the user guide for more information. + + This function also returns true for regular pinned bytearrays. + } + with out_of_line = True + +primop MutableByteArrayIsWeaklyPinnedOp "isMutableByteArrayWeaklyPinned#" GenPrimOp + MutableByteArray# s -> Int# + { 'isByteArrayWeaklyPinned#' but for mutable arrays. + } with out_of_line = True primop ByteArrayContents_Char "byteArrayContents#" GenPrimOp ===================================== compiler/GHC/StgToCmm/Prim.hs ===================================== @@ -1668,10 +1668,12 @@ emitPrimOp cfg primop = NewPinnedByteArrayOp_Char -> alwaysExternal NewAlignedPinnedByteArrayOp_Char -> alwaysExternal MutableByteArrayIsPinnedOp -> alwaysExternal + MutableByteArrayIsWeaklyPinnedOp -> alwaysExternal DoubleDecode_2IntOp -> alwaysExternal DoubleDecode_Int64Op -> alwaysExternal FloatDecode_IntOp -> alwaysExternal ByteArrayIsPinnedOp -> alwaysExternal + ByteArrayIsWeaklyPinnedOp -> alwaysExternal ShrinkMutableByteArrayOp_Char -> alwaysExternal ResizeMutableByteArrayOp_Char -> alwaysExternal ShrinkSmallMutableArrayOp_Char -> alwaysExternal ===================================== compiler/GHC/StgToJS/Prim.hs ===================================== @@ -670,6 +670,8 @@ genPrim prof bound ty op = case op of NewAlignedPinnedByteArrayOp_Char -> \[r] [l,_align] -> pure $ PrimInline (newByteArray r l) MutableByteArrayIsPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ ByteArrayIsPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ + ByteArrayIsWeaklyPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ + MutableByteArrayIsWeaklyPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ ByteArrayContents_Char -> \[a,o] [b] -> pure $ PrimInline $ mconcat [a |= b, o |= zero_] MutableByteArrayContents_Char -> \[a,o] [b] -> pure $ PrimInline $ mconcat [a |= b, o |= zero_] ShrinkMutableByteArrayOp_Char -> \[] [a,n] -> pure $ PrimInline $ appS hdShrinkMutableByteArrayStr [a,n] ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -163,6 +163,12 @@ Runtime system ~~~~~~~~~~~~~~~~~~~~ - Usage of deprecated primops is now correctly reported (#19629). +- New primops `isMutableByteArrayWeaklyPinned#` and `isByteArrayWeaklyPinned#` + to allow users to avoid copying large arrays safely when dealing with ffi. + See the users guide for more details on the different kinds of + pinned arrays in 9.12. + + This need for this distinction originally surfaced in https://gitlab.haskell.org/ghc/ghc/-/issues/22255 ``ghc`` library ===================================== docs/users_guide/exts/ffi.rst ===================================== @@ -1114,21 +1114,67 @@ Pinned Byte Arrays A pinned byte array is one that the garbage collector is not allowed to move. Consequently, it has a stable address that can be safely -requested with ``byteArrayContents#``. Not that being pinned doesn't -prevent the byteArray from being gc'ed in the same fashion a regular -byte array would be. +requested with ``byteArrayContents#``. As long as the array remains live +the address returned by ``byteArrayContents#`` will remain valid. Note that +being pinned doesn't prevent the byteArray from being gc'ed in the same fashion +a regular byte array would be if there are no more references to the ``ByteArray#``. There are a handful of primitive functions in :base-ref:`GHC.Exts.` used to enforce or check for pinnedness: ``isByteArrayPinned#``, -``isMutableByteArrayPinned#``, and ``newPinnedByteArray#``. A -byte array can be pinned as a result of three possible causes: +``isMutableByteArrayPinned#``, ``isByteArrayWeaklyPinned#``, +``isMutableByteArrayWeaklyPinned#``, and ``newPinnedByteArray#``. A +byte array can be pinned or weakly pinned as a result of three possible causes: -1. It was allocated by ``newPinnedByteArray#``. -2. It is large. Currently, GHC defines large object to be one +1. It was allocated by ``newPinnedByteArray#``. This results in a regular pinned byte array. +2. It is large, this results in a weakly pinned byte array. Currently, GHC defines large object to be one that is at least as large as 80% of a 4KB block (i.e. at least 3277 bytes). -3. It has been copied into a compact region. The documentation +3. It has been copied into a compact region, resulting in a weakly pinned array. The documentation for ``ghc-compact`` and ``compact`` describes this process. +The difference between a pinned array and a weakly pinned array is simply that +trying to compact a pinned array will result in an exception. Trying to compact +a weakly pinned array will succeed. However the result of earlier +calls to ``byteArrayContents#`` is not updated during compaction, which means +these results will still point to the address where the array was located originally, +and not to the new address inside the compact region. + +This is particularly dangerous when an address to a byte arrays content is stored +inside a datastructure along with a reference to the byte array. +If the data structure is compacted later on the pointer won't be updated but the +reference to the byte array will point to a copy inside the compact region. +A common data type susceptible to this is `ForeignPtr` when used to represent a ByteArray#. + +Here is an example to illustrate this: + +.. code-block:: haskell + + workWithArrayContents :: (ByteArray, Ptr Word8) -> (Ptr Word8 -> IO ()) -> IO () + workWithArrayContents (arr@(ByteArray uarr),ptr) worker = + case () of + _ + -- Conservative but safe + | isByteArrayPinned arr -> keepAliveUnlifted uarr (worker ptr) + -- Potentially dangerous, the program needs to ensures the Ptr points into the array. + | isByteArrayWeaklyPinned arr -> keepAliveUnlifted uarr (worker ptr) + | otherwise -> ... -- Otherwise we can't directly use it for safe FFI calls directly at all. + + main :: IO () + main = do + -- We create a large array, which causes it to be implicitly pinned + arr <- newByteArray 5000 + arr@(ByteArray uarr) <- freezeByteArray arr 0 5000 -- Make it immutable + let ptr = byteArrayContents arr + + -- Compacting a data structure that contains both an array and a ptr to + -- the arrays content's is dangerous and usually the wrong thing to do. + let foo = (arr, ptr) + foo_compacted <- compact foo + + -- This is fine + workWithArrayContents foo do_work + -- This is unsound + workWithArrayContents (getCompact foo_compacted) do_work + .. [1] Prior to GHC 8.10, when passing an ``ArrayArray#`` argument to a foreign function, the foreign function would see a pointer to the ``StgMutArrPtrs`` rather than just the payload. ===================================== libraries/base/src/GHC/Base.hs ===================================== @@ -139,7 +139,8 @@ module GHC.Base ) where import GHC.Internal.Base -import GHC.Prim hiding (dataToTagLarge#, dataToTagSmall#, whereFrom#) +import GHC.Prim hiding (dataToTagLarge#, dataToTagSmall#, whereFrom#, + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned#) -- Hide dataToTagLarge# because it is expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -112,7 +112,9 @@ module GHC.Exts import GHC.Internal.Exts import GHC.Internal.ArrayArray -import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# ) +import GHC.Prim hiding ( + coerce, dataToTagSmall#, dataToTagLarge#, whereFrom#, + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/ghc-prim/changelog.md ===================================== @@ -1,3 +1,12 @@ +## 0.13.0 + +- Shipped with GHC 9.12.1 + +- Add primops that allow users to distinguish weakly pinned byte arrays from unpinned ones. + + isMutableByteArrayWeaklyPinned# :: MutableByteArray# s -> Int# + isByteArrayWeaklyPinned# :: ByteArray# s -> Int# + ## 0.12.0 - Shipped with GHC 9.10.1 ===================================== rts/PrimOps.cmm ===================================== @@ -215,12 +215,29 @@ stg_isByteArrayPinnedzh ( gcptr ba ) return (flags & BF_PINNED != 0); } +stg_isByteArrayWeaklyPinnedzh ( gcptr ba ) +// ByteArray# s -> Int# +{ + W_ bd, flags; + bd = Bdescr(ba); + // See #22255 and the primop docs. + flags = TO_W_(bdescr_flags(bd)); + + return (flags & (BF_PINNED | BF_COMPACT | BF_LARGE) != 0); +} + stg_isMutableByteArrayPinnedzh ( gcptr mba ) // MutableByteArray# s -> Int# { jump stg_isByteArrayPinnedzh(mba); } +stg_isMutableByteArrayWeaklyPinnedzh ( gcptr mba ) +// MutableByteArray# s -> Int# +{ + jump stg_isByteArrayWeaklyPinnedzh(mba); +} + /* Note [LDV profiling and resizing arrays] * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * As far as the LDV profiler is concerned arrays are "inherently used" which ===================================== rts/RtsSymbols.c ===================================== @@ -656,6 +656,8 @@ extern char **environ; SymI_HasDataProto(stg_newAlignedPinnedByteArrayzh) \ SymI_HasDataProto(stg_isByteArrayPinnedzh) \ SymI_HasDataProto(stg_isMutableByteArrayPinnedzh) \ + SymI_HasDataProto(stg_isByteArrayWeaklyPinnedzh) \ + SymI_HasDataProto(stg_isMutableByteArrayWeaklyPinnedzh) \ SymI_HasDataProto(stg_shrinkMutableByteArrayzh) \ SymI_HasDataProto(stg_resizzeMutableByteArrayzh) \ SymI_HasDataProto(stg_shrinkSmallMutableArrayzh) \ ===================================== rts/include/stg/MiscClosures.h ===================================== @@ -454,6 +454,8 @@ RTS_FUN_DECL(stg_newPinnedByteArrayzh); RTS_FUN_DECL(stg_newAlignedPinnedByteArrayzh); RTS_FUN_DECL(stg_isByteArrayPinnedzh); RTS_FUN_DECL(stg_isMutableByteArrayPinnedzh); +RTS_FUN_DECL(stg_isByteArrayWeaklyPinnedzh); +RTS_FUN_DECL(stg_isMutableByteArrayWeaklyPinnedzh); RTS_FUN_DECL(stg_shrinkMutableByteArrayzh); RTS_FUN_DECL(stg_resizzeMutableByteArrayzh); RTS_FUN_DECL(stg_shrinkSmallMutableArrayzh); ===================================== testsuite/tests/rts/T13894.hs ===================================== @@ -6,6 +6,7 @@ import Control.Monad import GHC.Exts +import GHC.Internal.Exts (isMutableByteArrayWeaklyPinned#) import GHC.IO main :: IO () @@ -16,3 +17,10 @@ main = do case isMutableByteArrayPinned# arr# of n# -> (# s1, isTrue# n# #) when pinned $ putStrLn "BAD" + + weakly_pinned <- IO $ \s0 -> + case newByteArray# 1000000# s0 of + (# s1, arr# #) -> + case isMutableByteArrayWeaklyPinned# arr# of + n# -> (# s1, isTrue# n# #) + when (not weakly_pinned) $ putStrLn "BAD" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/59153543c2270b0a6a9a8f8a1ac356f076c4ef8f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/59153543c2270b0a6a9a8f8a1ac356f076c4ef8f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 12:54:36 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 02 Sep 2024 08:54:36 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] store IO actions in the EPS Message-ID: <66d5b58c8491f_12f9697d34e4830ea@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: c28ed0b8 by Torsten Schmits at 2024-09-02T14:54:14+02:00 store IO actions in the EPS - - - - - 18 changed files: - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/Iface/Tidy.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Home/ModInfo.hs - compiler/GHC/Unit/Module/ModDetails.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - ghc/GHCi/Leak.hs Changes: ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,8 +49,8 @@ module GHC.Driver.Main , Messager, batchMsg, batchMultiMsg , HscBackendAction (..), HscRecompStatus (..) , initModDetails - , initWholeCoreBindings - , initWholeCoreBindingsEps + , ensureHomeModuleByteCode + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -277,7 +277,7 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.Foldable (fold) +import Data.Functor ((<&>)) import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad @@ -296,7 +296,6 @@ import System.IO import {-# SOURCE #-} GHC.Driver.Pipeline import Data.Time -import System.IO.Unsafe ( unsafeInterleaveIO ) import GHC.Iface.Env ( trace_if ) import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) @@ -955,25 +954,25 @@ checkObjects dflags mb_old_linkable summary = do -- | Check to see if we can reuse the old linkable, by this point we will -- have just checked that the old interface matches up with the source hash, so -- no need to check that again here -checkByteCode :: ModIface -> ModSummary -> Maybe Linkable -> IO (MaybeValidated Linkable) -checkByteCode iface mod_sum mb_old_linkable = - case mb_old_linkable of - Just old_linkable - | not (linkableIsNativeCodeOnly old_linkable) - -> return $ (UpToDateItem old_linkable) - _ -> loadByteCode iface mod_sum - -loadByteCode :: ModIface -> ModSummary -> IO (MaybeValidated Linkable) +checkByteCode :: + ModIface -> + ModSummary -> + HomeModByteCode -> + IO (MaybeValidated HomeModByteCode) +checkByteCode iface mod_sum = \case + NoHomeModByteCode -> fmap HomeModIfaceCore <$> loadByteCode iface mod_sum + old_bytecode -> return (UpToDateItem old_bytecode) + +loadByteCode :: ModIface -> ModSummary -> IO (MaybeValidated WholeCoreBindings) loadByteCode iface mod_sum = do - let - this_mod = ms_mod mod_sum - if_date = fromJust $ ms_iface_date mod_sum case mi_extra_decls iface of Just extra_decls -> do - let fi = WholeCoreBindings extra_decls this_mod (ms_location mod_sum) - (mi_foreign iface) - return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) + let this_mod = ms_mod mod_sum + wcb = WholeCoreBindings extra_decls this_mod (ms_location mod_sum) + (mi_foreign iface) + return (UpToDateItem wcb) _ -> return $ outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- @@ -996,69 +995,104 @@ initModDetails hsc_env iface = -- in make mode, since this HMI will go into the HPT. genModDetails hsc_env' iface --- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects, --- using the supplied environment for type checking. --- --- The laziness is necessary because this value is stored purely in a --- 'HomeModLinkable' in the home package table, rather than some dedicated --- mutable state that would generate bytecode on demand, so we have to call this --- function even when we don't know that we'll need the bytecode. --- --- In addition, the laziness has to be hidden inside 'LazyBCOs' because --- 'Linkable' is used too generally, so that looking at the constructor to --- decide whether to discard it when linking native code would force the thunk --- otherwise, incurring a significant performance penalty. --- --- This is sound because generateByteCode just depends on things already loaded --- in the interface file. -initWcbWithTcEnv :: +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Hydrate core bindings for a module in the home package table, for which we +-- can obtain a 'ModDetails' with a type env. +ensureHomeModuleByteCode :: HscEnv -> + ModIface -> + ModLocation -> + ModDetails -> + HomeModByteCode -> + Maybe (IO Linkable) +ensureHomeModuleByteCode hsc_env iface location details = \case + NoHomeModByteCode -> + loadIfaceByteCode hsc_env' iface location type_env + HomeModIfaceCore wcb -> + Just (initWholeCoreBindings hsc_env' type_env wcb) + HomeModByteCode bc -> + Just (pure bc) + where + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate Core bindings if the interface contains any, using the supplied +-- type env for typechecking. +loadIfaceByteCode :: HscEnv -> + ModIface -> + ModLocation -> TypeEnv -> - Linkable -> - IO Linkable -initWcbWithTcEnv tc_hsc_env hsc_env type_env (Linkable utc_time this_mod uls) = - Linkable utc_time this_mod <$> mapM go uls + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + initWholeCoreBindings hsc_env type_env <$> iface_core_bindings iface location + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +-- +-- 4. Wrapping the build products in 'Linkable' with the proper modification +-- time obtained from the interface. +initWholeCoreBindings :: HscEnv -> TypeEnv -> WholeCoreBindings -> IO Linkable +initWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + parts <- gen_bytecode core_binds stubs foreign_files + linkable parts where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef type_env - let - tc_hsc_env_with_kv = tc_hsc_env { - hsc_type_env_vars = - knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") tc_hsc_env_with_kv $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons type_env) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + (bcos, fos) <- generateByteCode hsc_env cgi_guts wcb_mod_location + pure $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file wcb_mod_location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time wcb_module parts + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env --- | Hydrate core bindings for a module in the home package table, for which we --- can obtain a 'ModDetails'. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env iface details = - initWcbWithTcEnv (add_iface_to_hpt iface details hsc_env) hsc_env (md_types details) - --- | Hydrate core bindings for a module in the external package state. --- This is used for home modules as well when compiling in oneshot mode. -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable -initWholeCoreBindingsEps hsc_env iface lnk = do - eps <- hscEPS hsc_env - let type_env = fold (lookupModuleEnv (eps_PTT eps) (mi_module iface)) - initWcbWithTcEnv hsc_env hsc_env type_env lnk - - {- Note [ModDetails and --make mode] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -1,8 +1,15 @@ module GHC.Driver.Main where -import GHC.Driver.Env -import GHC.Linker.Types -import GHC.Prelude -import GHC.Unit.Module.ModIface +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1319,11 +1319,12 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- | Add the entries from a BCO linkable to the SPT table, see -- See Note [Grand plan for static forms] in GHC.Iface.Tidy.StaticPtrTable. -addSptEntries :: HscEnv -> Maybe Linkable -> IO () +addSptEntries :: HscEnv -> HomeModByteCode -> IO () addSptEntries hsc_env mlinkable = hscAddSptEntries hsc_env [ spt - | linkable <- maybeToList mlinkable + -- TODO + | HomeModByteCode linkable <- [mlinkable] , bco <- linkableBCOs linkable , spt <- bc_spt_entries bco ] ===================================== compiler/GHC/Driver/Pipeline.hs ===================================== @@ -124,6 +124,7 @@ import Data.List.NonEmpty (NonEmpty(..)) import Data.Time ( getCurrentTime ) import GHC.Iface.Recomp import GHC.Types.Unique.DSet +import GHC.Unit.Module.ModDetails (ModDetails(..)) -- Simpler type synonym for actions in the pipeline monad type P m = TPipelineClass TPhase m @@ -248,8 +249,10 @@ compileOne' mHscMessage (iface, linkable) <- runPipeline (hsc_hooks plugin_hsc_env) pipeline -- See Note [ModDetails and --make mode] details <- initModDetails plugin_hsc_env iface - linkable' <- traverse (initWholeCoreBindings plugin_hsc_env iface details) (homeMod_bytecode linkable) - return $! HomeModInfo iface details (linkable { homeMod_bytecode = linkable' }) + let md_bytecode = + ensureHomeModuleByteCode hsc_env iface (ms_location summary) details + (homeMod_bytecode linkable) + return $! HomeModInfo iface details {md_bytecode} linkable where lcl_dflags = ms_hspp_opts summary location = ms_location summary ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -602,7 +602,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do if gopt Opt_ByteCodeAndObjectCode dflags then do bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return $ emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } + return $ emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } else return emptyHomeModInfoLinkable @@ -619,7 +619,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do final_iface <- mkFullIface hsc_env partial_iface Nothing Nothing NoStubs [] hscMaybeWriteIface logger dflags True final_iface mb_old_iface_hash location bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } , panic "interpreter") + return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } , panic "interpreter") runUnlitPhase :: HscEnv -> FilePath -> FilePath -> IO FilePath ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,6 +506,7 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) @@ -518,13 +520,34 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") + -- TODO can't do that because we use it for + -- fingerprinting. + -- & set_mi_foreign (panic "No mi_foreign in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + -- TODO in @getLinkDeps@ we fall back to bytecode when the HMI + -- doesn't have object code, even if the flag is not given – + -- what's the rule? Should we provide it unconditionally if it + -- exists? + | prefer_bytecode + , Just action <- loadIfaceByteCode hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have @extra_decls@ + -- so @getLinkDeps@ knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,8 +559,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, - eps_PTT = - extendModuleEnv (eps_PTT eps) mod (mkNameEnv new_eps_decls), + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -700,7 +722,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -847,7 +869,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -877,7 +899,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -902,7 +924,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/Iface/Tidy.hs ===================================== @@ -193,6 +193,7 @@ mkBootModDetailsTc logger , md_anns = [] , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing }) where -- Find the LocalIds in the type env that are exported @@ -492,6 +493,7 @@ tidyProgram opts (ModGuts { mg_module = mod , md_exports = exports , md_anns = anns -- are already tidy , md_complete_matches = complete_matches + , md_bytecode = Nothing } ) ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -254,6 +254,7 @@ typecheckIface iface , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } } @@ -470,6 +471,7 @@ typecheckIfacesForMerging mod ifaces tc_env_vars = , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } return (global_type_env, details) @@ -512,6 +514,7 @@ typecheckIfaceForInstantiate nsubst iface , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } -- Note [Resolving never-exported Names] ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -35,7 +35,6 @@ import GHC.Unit.Env import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.Module.ModIface -import GHC.Unit.Module.WholeCoreBindings import GHC.Unit.Module.Deps import GHC.Unit.Module.Graph import GHC.Unit.Home.ModInfo @@ -56,26 +55,20 @@ import Data.List (isSuffixOf) import System.FilePath import System.Directory -import GHC.Driver.Env -import {-# SOURCE #-} GHC.Driver.Main -import Data.Time.Clock -import GHC.Driver.Flags -import GHC.Driver.Session data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function - , ldHscEnv :: !HscEnv + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -266,54 +259,40 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- See Note [Using Byte Code rather than Object Code for Template Haskell] - homeModLinkable :: HomeModInfo -> Maybe Linkable + homeModLinkable :: HomeModInfo -> Maybe (IO Linkable) homeModLinkable hmi = - if ldUseByteCode opts - then homeModInfoByteCode hmi <|> homeModInfoObject hmi - else homeModInfoObject hmi <|> homeModInfoByteCode hmi + let obj = pure <$> homeModInfoObject hmi + bc = evalHomeModByteCode hmi + in if ldUseByteCode opts + then bc <|> obj + else obj <|> bc get_linkable osuf mod -- A home-package module | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) - = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) + = do + lnk <- expectJust "getLinkDeps" (homeModLinkable mod_info) + adjust_linkable lnk | otherwise = do -- It's not in the HPT because we are in one shot mode, -- so use the Finder to get a ModLocation... case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod - | prefer_bytecode = do - Succeeded iface <- ldLoadIface opts (text "load core bindings") mod - case mi_extra_decls iface of - Just extra_decls -> do - t <- getCurrentTime - let - stubs = mi_foreign iface - wcb = WholeCoreBindings extra_decls mod loc stubs - linkable = Linkable t mod (pure (CoreBindings wcb)) - initWholeCoreBindingsEps hsc_env iface linkable - _ -> fallback_no_bytecode loc mod - | otherwise = fallback_no_bytecode loc mod - - fallback_no_bytecode loc mod = do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - - prefer_bytecode = gopt Opt_UseBytecodeRatherThanObjects dflags - - dflags = hsc_dflags hsc_env - - hsc_env = ldHscEnv opts + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do @@ -338,9 +317,6 @@ get_link_deps opts pls maybe_normal_osuf span mods = do DotA fp -> panic ("adjust_ul DotA " ++ show fp) DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,19 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags - , ldHscEnv = hsc_env + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Linker/Types.hs ===================================== @@ -65,7 +65,6 @@ import Data.Time ( UTCTime ) import GHC.Unit.Module.Env import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM -import GHC.Unit.Module.WholeCoreBindings import Data.Maybe (mapMaybe) import Data.List.NonEmpty (NonEmpty, nonEmpty) import qualified Data.List.NonEmpty as NE @@ -284,18 +283,6 @@ data LinkablePart | DotDLL FilePath -- ^ Dynamically linked library file (.so, .dll, .dylib) - | CoreBindings WholeCoreBindings - -- ^ Serialised core which we can turn into BCOs (or object files), or - -- used by some other backend See Note [Interface Files with Core - -- Definitions] - - | LazyBCOs - CompiledByteCode - -- ^ Some BCOs generated on-demand when forced. This is used for - -- WholeCoreBindings, see Note [Interface Files with Core Definitions] - [FilePath] - -- ^ Objects containing foreign stubs and files - | BCOs CompiledByteCode -- ^ A byte-code object, lives only in memory. @@ -308,8 +295,6 @@ instance Outputable LinkablePart where ppr (DotA path) = text "DotA" <+> text path ppr (DotDLL path) = text "DotDLL" <+> text path ppr (BCOs bco) = text "BCOs" <+> ppr bco - ppr (LazyBCOs{}) = text "LazyBCOs" - ppr (CoreBindings {}) = text "CoreBindings" -- | Return true if the linkable only consists of native code (no BCO) linkableIsNativeCodeOnly :: Linkable -> Bool @@ -350,8 +335,6 @@ isNativeCode = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Is the part a native library? (.so/.dll) isNativeLib :: LinkablePart -> Bool @@ -360,8 +343,6 @@ isNativeLib = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Get the FilePath of linkable part (if applicable) linkablePartPath :: LinkablePart -> Maybe FilePath @@ -369,8 +350,6 @@ linkablePartPath = \case DotO fn _ -> Just fn DotA fn -> Just fn DotDLL fn -> Just fn - CoreBindings {} -> Nothing - LazyBCOs {} -> Nothing BCOs {} -> Nothing -- | Return the paths of all object code files (.o, .a, .so) contained in this @@ -380,8 +359,6 @@ linkablePartNativePaths = \case DotO fn _ -> [fn] DotA fn -> [fn] DotDLL fn -> [fn] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Return the paths of all object files (.o) contained in this 'LinkablePart'. @@ -390,8 +367,6 @@ linkablePartObjectPaths = \case DotO fn _ -> [fn] DotA _ -> [] DotDLL _ -> [] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Retrieve the compiled byte-code from the linkable part. @@ -400,7 +375,6 @@ linkablePartObjectPaths = \case linkablePartAllBCOs :: LinkablePart -> [CompiledByteCode] linkablePartAllBCOs = \case BCOs bco -> [bco] - LazyBCOs bcos _ -> [bcos] _ -> [] linkableFilter :: (LinkablePart -> [LinkablePart]) -> Linkable -> Maybe Linkable @@ -413,13 +387,11 @@ linkablePartNative = \case u at DotO {} -> [u] u at DotA {} -> [u] u at DotDLL {} -> [u] - LazyBCOs _ os -> [DotO f ForeignObject | f <- os] _ -> [] linkablePartByteCode :: LinkablePart -> [LinkablePart] linkablePartByteCode = \case u at BCOs {} -> [u] - LazyBCOs bcos _ -> [BCOs bcos] _ -> [] -- | Transform the 'LinkablePart' list in this 'Linkable' to contain only ===================================== compiler/GHC/Runtime/Eval.hs ===================================== @@ -1275,14 +1275,14 @@ showModule mod_summary = let interpreted = case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> isJust (homeModInfoByteCode mod_info) && isNothing (homeModInfoObject mod_info) + Just mod_info -> homeModInfoHasByteCode mod_info && isNothing (homeModInfoObject mod_info) return (showSDoc dflags $ showModMsg dflags interpreted (ModuleNode [] mod_summary)) moduleIsBootOrNotObjectLinkable :: GhcMonad m => ModSummary -> m Bool moduleIsBootOrNotObjectLinkable mod_summary = withSession $ \hsc_env -> case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> return . isNothing $ homeModInfoByteCode mod_info + Just mod_info -> return . not $ homeModInfoHasByteCode mod_info ---------------------------------------------------------------------------- -- RTTI primitives ===================================== compiler/GHC/Runtime/Interpreter.hs ===================================== @@ -113,6 +113,7 @@ import qualified GHC.Exts.Heap as Heap import GHC.Stack.CCS (CostCentre,CostCentreStack) import System.Directory import System.Process +import GHC.Unit.Module.ModDetails (ModDetails(..)) {- Note [Remote GHCi] ~~~~~~~~~~~~~~~~~~ @@ -684,10 +685,13 @@ fromEvalResult (EvalSuccess a) = return a getModBreaks :: HomeModInfo -> ModBreaks getModBreaks hmi - | Just linkable <- homeModInfoByteCode hmi, + | HomeModByteCode linkable <- homeModInfoByteCode hmi, -- The linkable may have 'DotO's as well; only consider BCOs. See #20570. [cbc] <- linkableBCOs linkable = fromMaybe emptyModBreaks (bc_breaks cbc) + | ModDetails {md_bytecode = Just _} <- hm_details hmi + -- TODO + = emptyModBreaks | otherwise = emptyModBreaks -- probably object code ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -45,8 +47,6 @@ type PackageCompleteMatches = CompleteMatches type PackageIfaceTable = ModuleEnv ModIface -- Domain = modules in the imported packages -type PackageTypeTable = ModuleEnv TypeEnv - -- | Constructs an empty PackageIfaceTable emptyPackageIfaceTable :: PackageIfaceTable emptyPackageIfaceTable = emptyModuleEnv @@ -70,7 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv - , eps_PTT = emptyModuleEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -142,7 +142,11 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules - eps_PTT :: !PackageTypeTable, + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules ===================================== compiler/GHC/Unit/Home/ModInfo.hs ===================================== @@ -1,13 +1,19 @@ +{-# LANGUAGE LambdaCase #-} + -- | Info about modules in the "home" unit module GHC.Unit.Home.ModInfo ( HomeModInfo (..) - , HomeModLinkable(..) + , HomeModLinkable (..) + , HomeModByteCode (..) , homeModInfoObject , homeModInfoByteCode + , homeModInfoHasByteCode , emptyHomeModInfoLinkable , justBytecode , justObjects , bytecodeAndObjects + , pureHomeModByteCode + , evalHomeModByteCode , HomePackageTable , emptyHomePackageTable , lookupHpt @@ -44,6 +50,7 @@ import GHC.Utils.Outputable import Data.List (sortOn) import Data.Ord import GHC.Utils.Panic +import GHC.Unit.Module.WholeCoreBindings (WholeCoreBindings) -- | Information about modules in the package being compiled data HomeModInfo = HomeModInfo @@ -73,37 +80,67 @@ data HomeModInfo = HomeModInfo -- 'ModIface' (only). } -homeModInfoByteCode :: HomeModInfo -> Maybe Linkable +homeModInfoByteCode :: HomeModInfo -> HomeModByteCode homeModInfoByteCode = homeMod_bytecode . hm_linkable +homeModInfoHasByteCode :: HomeModInfo -> Bool +homeModInfoHasByteCode hmi = case homeModInfoByteCode hmi of + NoHomeModByteCode -> False + _ -> True + homeModInfoObject :: HomeModInfo -> Maybe Linkable homeModInfoObject = homeMod_object . hm_linkable emptyHomeModInfoLinkable :: HomeModLinkable -emptyHomeModInfoLinkable = HomeModLinkable Nothing Nothing +emptyHomeModInfoLinkable = HomeModLinkable NoHomeModByteCode Nothing -- See Note [Home module build products] -data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !(Maybe Linkable) +data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !HomeModByteCode , homeMod_object :: !(Maybe Linkable) } instance Outputable HomeModLinkable where ppr (HomeModLinkable l1 l2) = ppr l1 $$ ppr l2 -justBytecode :: Linkable -> HomeModLinkable -justBytecode lm = - assertPpr (not (linkableIsNativeCodeOnly lm)) (ppr lm) - $ emptyHomeModInfoLinkable { homeMod_bytecode = Just lm } +justBytecode :: HomeModByteCode -> HomeModLinkable +justBytecode bc = + emptyHomeModInfoLinkable { homeMod_bytecode = bc } justObjects :: Linkable -> HomeModLinkable justObjects lm = assertPpr (linkableIsNativeCodeOnly lm) (ppr lm) $ emptyHomeModInfoLinkable { homeMod_object = Just lm } -bytecodeAndObjects :: Linkable -> Linkable -> HomeModLinkable +bytecodeAndObjects :: HomeModByteCode -> Linkable -> HomeModLinkable bytecodeAndObjects bc o = - assertPpr (not (linkableIsNativeCodeOnly bc) && linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) - (HomeModLinkable (Just bc) (Just o)) - + assertPpr (linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) + (HomeModLinkable bc (Just o)) + +pureHomeModByteCode :: HomeModByteCode -> Maybe Linkable +pureHomeModByteCode = \case + NoHomeModByteCode -> Nothing + HomeModIfaceCore _ -> Nothing + HomeModByteCode l -> Just l + +-- | +evalHomeModByteCode :: HomeModInfo -> Maybe (IO Linkable) +evalHomeModByteCode HomeModInfo {hm_details, hm_linkable} + | HomeModByteCode bc <- homeMod_bytecode hm_linkable + = Just (pure bc) + | otherwise + = md_bytecode hm_details + +data HomeModByteCode = + NoHomeModByteCode + | + HomeModIfaceCore WholeCoreBindings + | + HomeModByteCode Linkable + +instance Outputable HomeModByteCode where + ppr = \case + NoHomeModByteCode -> text "no bytecode" + HomeModIfaceCore _ -> text "dehydrated Core" + HomeModByteCode linkable -> ppr linkable {- Note [Home module build products] ===================================== compiler/GHC/Unit/Module/ModDetails.hs ===================================== @@ -14,6 +14,9 @@ import GHC.Types.DefaultEnv ( DefaultEnv, emptyDefaultEnv ) import GHC.Types.TypeEnv import GHC.Types.Annotations ( Annotation ) +import GHC.Linker.Types (Linkable) +import GHC.Prelude + -- | The 'ModDetails' is essentially a cache for information in the 'ModIface' -- for home modules only. Information relating to packages will be loaded into -- global environments in 'ExternalPackageState'. @@ -40,6 +43,8 @@ data ModDetails = ModDetails , md_complete_matches :: CompleteMatches -- ^ Complete match pragmas for this module + + , md_bytecode :: !(Maybe (IO Linkable)) } -- | Constructs an empty ModDetails @@ -53,4 +58,5 @@ emptyModDetails = ModDetails , md_fam_insts = [] , md_anns = [] , md_complete_matches = [] + , md_bytecode = Nothing } ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -27,6 +27,8 @@ import Data.Maybe (fromMaybe) import System.FilePath (takeExtension) {- +TODO update + Note [Interface Files with Core Definitions] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -87,7 +89,12 @@ data WholeCoreBindings = WholeCoreBindings , wcb_foreign :: IfaceForeign } +instance Outputable WholeCoreBindings where + ppr WholeCoreBindings {wcb_module} = text "iface Core for " <+> ppr wcb_module + {- +TODO update + Note [Foreign stubs and TH bytecode linking] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== ghc/GHCi/Leak.hs ===================================== @@ -50,7 +50,7 @@ getLeakIndicators hsc_env = where mkWeakLinkables :: HomeModLinkable -> IO [Maybe (Weak Linkable)] mkWeakLinkables (HomeModLinkable mbc mo) = - mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [mbc, mo] + mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [pureHomeModByteCode mbc, mo] -- | Look at the LeakIndicators collected by an earlier call to -- `getLeakIndicators`, and print messasges if any of them are still View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c28ed0b8eb515ff6b734e6dad744eef1ff80fa55 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c28ed0b8eb515ff6b734e6dad744eef1ff80fa55 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 13:04:10 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 02 Sep 2024 09:04:10 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] store IO actions in the EPS Message-ID: <66d5b7ca88cef_12f96983133c879a@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 4de4d0f5 by Torsten Schmits at 2024-09-02T15:03:55+02:00 store IO actions in the EPS - - - - - 18 changed files: - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/Iface/Tidy.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Home/ModInfo.hs - compiler/GHC/Unit/Module/ModDetails.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - ghc/GHCi/Leak.hs Changes: ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,8 +49,8 @@ module GHC.Driver.Main , Messager, batchMsg, batchMultiMsg , HscBackendAction (..), HscRecompStatus (..) , initModDetails - , initWholeCoreBindings - , initWholeCoreBindingsEps + , ensureHomeModuleByteCode + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -277,7 +277,7 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.Foldable (fold) +import Data.Functor ((<&>)) import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad @@ -296,7 +296,6 @@ import System.IO import {-# SOURCE #-} GHC.Driver.Pipeline import Data.Time -import System.IO.Unsafe ( unsafeInterleaveIO ) import GHC.Iface.Env ( trace_if ) import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) @@ -868,7 +867,7 @@ hscRecompStatus -- Do need linkable -- 1. Just check whether we have bytecode/object linkables and then -- we will decide if we need them or not. - bc_linkable <- checkByteCode checked_iface mod_summary (homeMod_bytecode old_linkable) + let bc_linkable = checkByteCode checked_iface mod_summary (homeMod_bytecode old_linkable) obj_linkable <- liftIO $ checkObjects lcl_dflags (homeMod_object old_linkable) mod_summary trace_if (hsc_logger hsc_env) (vcat [text "BCO linkable", nest 2 (ppr bc_linkable), text "Object Linkable", ppr obj_linkable]) @@ -955,25 +954,21 @@ checkObjects dflags mb_old_linkable summary = do -- | Check to see if we can reuse the old linkable, by this point we will -- have just checked that the old interface matches up with the source hash, so -- no need to check that again here -checkByteCode :: ModIface -> ModSummary -> Maybe Linkable -> IO (MaybeValidated Linkable) -checkByteCode iface mod_sum mb_old_linkable = - case mb_old_linkable of - Just old_linkable - | not (linkableIsNativeCodeOnly old_linkable) - -> return $ (UpToDateItem old_linkable) - _ -> loadByteCode iface mod_sum - -loadByteCode :: ModIface -> ModSummary -> IO (MaybeValidated Linkable) -loadByteCode iface mod_sum = do - let - this_mod = ms_mod mod_sum - if_date = fromJust $ ms_iface_date mod_sum - case mi_extra_decls iface of - Just extra_decls -> do - let fi = WholeCoreBindings extra_decls this_mod (ms_location mod_sum) - (mi_foreign iface) - return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) - _ -> return $ outOfDateItemBecause MissingBytecode Nothing +checkByteCode :: + ModIface -> + ModSummary -> + HomeModByteCode -> + MaybeValidated HomeModByteCode +checkByteCode iface mod_sum = \case + NoHomeModByteCode -> HomeModIfaceCore <$> loadByteCode iface mod_sum + old_bytecode -> UpToDateItem old_bytecode + +loadByteCode :: ModIface -> ModSummary -> MaybeValidated WholeCoreBindings +loadByteCode iface mod_sum = + case iface_core_bindings iface (ms_location mod_sum) of + Just wcb -> UpToDateItem wcb + Nothing -> outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- @@ -996,69 +991,104 @@ initModDetails hsc_env iface = -- in make mode, since this HMI will go into the HPT. genModDetails hsc_env' iface --- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects, --- using the supplied environment for type checking. --- --- The laziness is necessary because this value is stored purely in a --- 'HomeModLinkable' in the home package table, rather than some dedicated --- mutable state that would generate bytecode on demand, so we have to call this --- function even when we don't know that we'll need the bytecode. --- --- In addition, the laziness has to be hidden inside 'LazyBCOs' because --- 'Linkable' is used too generally, so that looking at the constructor to --- decide whether to discard it when linking native code would force the thunk --- otherwise, incurring a significant performance penalty. --- --- This is sound because generateByteCode just depends on things already loaded --- in the interface file. -initWcbWithTcEnv :: +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Hydrate core bindings for a module in the home package table, for which we +-- can obtain a 'ModDetails' with a type env. +ensureHomeModuleByteCode :: HscEnv -> + ModIface -> + ModLocation -> + ModDetails -> + HomeModByteCode -> + Maybe (IO Linkable) +ensureHomeModuleByteCode hsc_env iface location details = \case + NoHomeModByteCode -> + loadIfaceByteCode hsc_env' iface location type_env + HomeModIfaceCore wcb -> + Just (initWholeCoreBindings hsc_env' type_env wcb) + HomeModByteCode bc -> + Just (pure bc) + where + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate Core bindings if the interface contains any, using the supplied +-- type env for typechecking. +loadIfaceByteCode :: HscEnv -> + ModIface -> + ModLocation -> TypeEnv -> - Linkable -> - IO Linkable -initWcbWithTcEnv tc_hsc_env hsc_env type_env (Linkable utc_time this_mod uls) = - Linkable utc_time this_mod <$> mapM go uls + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + initWholeCoreBindings hsc_env type_env <$> iface_core_bindings iface location + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +-- +-- 4. Wrapping the build products in 'Linkable' with the proper modification +-- time obtained from the interface. +initWholeCoreBindings :: HscEnv -> TypeEnv -> WholeCoreBindings -> IO Linkable +initWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + parts <- gen_bytecode core_binds stubs foreign_files + linkable parts where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef type_env - let - tc_hsc_env_with_kv = tc_hsc_env { - hsc_type_env_vars = - knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") tc_hsc_env_with_kv $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons type_env) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + (bcos, fos) <- generateByteCode hsc_env cgi_guts wcb_mod_location + pure $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file wcb_mod_location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time wcb_module parts + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env --- | Hydrate core bindings for a module in the home package table, for which we --- can obtain a 'ModDetails'. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env iface details = - initWcbWithTcEnv (add_iface_to_hpt iface details hsc_env) hsc_env (md_types details) - --- | Hydrate core bindings for a module in the external package state. --- This is used for home modules as well when compiling in oneshot mode. -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable -initWholeCoreBindingsEps hsc_env iface lnk = do - eps <- hscEPS hsc_env - let type_env = fold (lookupModuleEnv (eps_PTT eps) (mi_module iface)) - initWcbWithTcEnv hsc_env hsc_env type_env lnk - - {- Note [ModDetails and --make mode] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -1,8 +1,15 @@ module GHC.Driver.Main where -import GHC.Driver.Env -import GHC.Linker.Types -import GHC.Prelude -import GHC.Unit.Module.ModIface +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1319,11 +1319,12 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- | Add the entries from a BCO linkable to the SPT table, see -- See Note [Grand plan for static forms] in GHC.Iface.Tidy.StaticPtrTable. -addSptEntries :: HscEnv -> Maybe Linkable -> IO () +addSptEntries :: HscEnv -> HomeModByteCode -> IO () addSptEntries hsc_env mlinkable = hscAddSptEntries hsc_env [ spt - | linkable <- maybeToList mlinkable + -- TODO + | HomeModByteCode linkable <- [mlinkable] , bco <- linkableBCOs linkable , spt <- bc_spt_entries bco ] ===================================== compiler/GHC/Driver/Pipeline.hs ===================================== @@ -124,6 +124,7 @@ import Data.List.NonEmpty (NonEmpty(..)) import Data.Time ( getCurrentTime ) import GHC.Iface.Recomp import GHC.Types.Unique.DSet +import GHC.Unit.Module.ModDetails (ModDetails(..)) -- Simpler type synonym for actions in the pipeline monad type P m = TPipelineClass TPhase m @@ -248,8 +249,10 @@ compileOne' mHscMessage (iface, linkable) <- runPipeline (hsc_hooks plugin_hsc_env) pipeline -- See Note [ModDetails and --make mode] details <- initModDetails plugin_hsc_env iface - linkable' <- traverse (initWholeCoreBindings plugin_hsc_env iface details) (homeMod_bytecode linkable) - return $! HomeModInfo iface details (linkable { homeMod_bytecode = linkable' }) + let md_bytecode = + ensureHomeModuleByteCode hsc_env iface (ms_location summary) details + (homeMod_bytecode linkable) + return $! HomeModInfo iface details {md_bytecode} linkable where lcl_dflags = ms_hspp_opts summary location = ms_location summary ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -602,7 +602,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do if gopt Opt_ByteCodeAndObjectCode dflags then do bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return $ emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } + return $ emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } else return emptyHomeModInfoLinkable @@ -619,7 +619,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do final_iface <- mkFullIface hsc_env partial_iface Nothing Nothing NoStubs [] hscMaybeWriteIface logger dflags True final_iface mb_old_iface_hash location bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } , panic "interpreter") + return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } , panic "interpreter") runUnlitPhase :: HscEnv -> FilePath -> FilePath -> IO FilePath ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,6 +506,7 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) @@ -518,13 +520,34 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") + -- TODO can't do that because we use it for + -- fingerprinting. + -- & set_mi_foreign (panic "No mi_foreign in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + -- TODO in @getLinkDeps@ we fall back to bytecode when the HMI + -- doesn't have object code, even if the flag is not given – + -- what's the rule? Should we provide it unconditionally if it + -- exists? + | prefer_bytecode + , Just action <- loadIfaceByteCode hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have @extra_decls@ + -- so @getLinkDeps@ knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,8 +559,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, - eps_PTT = - extendModuleEnv (eps_PTT eps) mod (mkNameEnv new_eps_decls), + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -700,7 +722,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -847,7 +869,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -877,7 +899,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -902,7 +924,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/Iface/Tidy.hs ===================================== @@ -193,6 +193,7 @@ mkBootModDetailsTc logger , md_anns = [] , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing }) where -- Find the LocalIds in the type env that are exported @@ -492,6 +493,7 @@ tidyProgram opts (ModGuts { mg_module = mod , md_exports = exports , md_anns = anns -- are already tidy , md_complete_matches = complete_matches + , md_bytecode = Nothing } ) ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -254,6 +254,7 @@ typecheckIface iface , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } } @@ -470,6 +471,7 @@ typecheckIfacesForMerging mod ifaces tc_env_vars = , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } return (global_type_env, details) @@ -512,6 +514,7 @@ typecheckIfaceForInstantiate nsubst iface , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } -- Note [Resolving never-exported Names] ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -35,7 +35,6 @@ import GHC.Unit.Env import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.Module.ModIface -import GHC.Unit.Module.WholeCoreBindings import GHC.Unit.Module.Deps import GHC.Unit.Module.Graph import GHC.Unit.Home.ModInfo @@ -56,26 +55,20 @@ import Data.List (isSuffixOf) import System.FilePath import System.Directory -import GHC.Driver.Env -import {-# SOURCE #-} GHC.Driver.Main -import Data.Time.Clock -import GHC.Driver.Flags -import GHC.Driver.Session data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function - , ldHscEnv :: !HscEnv + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -266,54 +259,40 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- See Note [Using Byte Code rather than Object Code for Template Haskell] - homeModLinkable :: HomeModInfo -> Maybe Linkable + homeModLinkable :: HomeModInfo -> Maybe (IO Linkable) homeModLinkable hmi = - if ldUseByteCode opts - then homeModInfoByteCode hmi <|> homeModInfoObject hmi - else homeModInfoObject hmi <|> homeModInfoByteCode hmi + let obj = pure <$> homeModInfoObject hmi + bc = evalHomeModByteCode hmi + in if ldUseByteCode opts + then bc <|> obj + else obj <|> bc get_linkable osuf mod -- A home-package module | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) - = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) + = do + lnk <- expectJust "getLinkDeps" (homeModLinkable mod_info) + adjust_linkable lnk | otherwise = do -- It's not in the HPT because we are in one shot mode, -- so use the Finder to get a ModLocation... case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod - | prefer_bytecode = do - Succeeded iface <- ldLoadIface opts (text "load core bindings") mod - case mi_extra_decls iface of - Just extra_decls -> do - t <- getCurrentTime - let - stubs = mi_foreign iface - wcb = WholeCoreBindings extra_decls mod loc stubs - linkable = Linkable t mod (pure (CoreBindings wcb)) - initWholeCoreBindingsEps hsc_env iface linkable - _ -> fallback_no_bytecode loc mod - | otherwise = fallback_no_bytecode loc mod - - fallback_no_bytecode loc mod = do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - - prefer_bytecode = gopt Opt_UseBytecodeRatherThanObjects dflags - - dflags = hsc_dflags hsc_env - - hsc_env = ldHscEnv opts + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do @@ -338,9 +317,6 @@ get_link_deps opts pls maybe_normal_osuf span mods = do DotA fp -> panic ("adjust_ul DotA " ++ show fp) DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,19 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags - , ldHscEnv = hsc_env + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Linker/Types.hs ===================================== @@ -65,7 +65,6 @@ import Data.Time ( UTCTime ) import GHC.Unit.Module.Env import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM -import GHC.Unit.Module.WholeCoreBindings import Data.Maybe (mapMaybe) import Data.List.NonEmpty (NonEmpty, nonEmpty) import qualified Data.List.NonEmpty as NE @@ -284,18 +283,6 @@ data LinkablePart | DotDLL FilePath -- ^ Dynamically linked library file (.so, .dll, .dylib) - | CoreBindings WholeCoreBindings - -- ^ Serialised core which we can turn into BCOs (or object files), or - -- used by some other backend See Note [Interface Files with Core - -- Definitions] - - | LazyBCOs - CompiledByteCode - -- ^ Some BCOs generated on-demand when forced. This is used for - -- WholeCoreBindings, see Note [Interface Files with Core Definitions] - [FilePath] - -- ^ Objects containing foreign stubs and files - | BCOs CompiledByteCode -- ^ A byte-code object, lives only in memory. @@ -308,8 +295,6 @@ instance Outputable LinkablePart where ppr (DotA path) = text "DotA" <+> text path ppr (DotDLL path) = text "DotDLL" <+> text path ppr (BCOs bco) = text "BCOs" <+> ppr bco - ppr (LazyBCOs{}) = text "LazyBCOs" - ppr (CoreBindings {}) = text "CoreBindings" -- | Return true if the linkable only consists of native code (no BCO) linkableIsNativeCodeOnly :: Linkable -> Bool @@ -350,8 +335,6 @@ isNativeCode = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Is the part a native library? (.so/.dll) isNativeLib :: LinkablePart -> Bool @@ -360,8 +343,6 @@ isNativeLib = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Get the FilePath of linkable part (if applicable) linkablePartPath :: LinkablePart -> Maybe FilePath @@ -369,8 +350,6 @@ linkablePartPath = \case DotO fn _ -> Just fn DotA fn -> Just fn DotDLL fn -> Just fn - CoreBindings {} -> Nothing - LazyBCOs {} -> Nothing BCOs {} -> Nothing -- | Return the paths of all object code files (.o, .a, .so) contained in this @@ -380,8 +359,6 @@ linkablePartNativePaths = \case DotO fn _ -> [fn] DotA fn -> [fn] DotDLL fn -> [fn] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Return the paths of all object files (.o) contained in this 'LinkablePart'. @@ -390,8 +367,6 @@ linkablePartObjectPaths = \case DotO fn _ -> [fn] DotA _ -> [] DotDLL _ -> [] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Retrieve the compiled byte-code from the linkable part. @@ -400,7 +375,6 @@ linkablePartObjectPaths = \case linkablePartAllBCOs :: LinkablePart -> [CompiledByteCode] linkablePartAllBCOs = \case BCOs bco -> [bco] - LazyBCOs bcos _ -> [bcos] _ -> [] linkableFilter :: (LinkablePart -> [LinkablePart]) -> Linkable -> Maybe Linkable @@ -413,13 +387,11 @@ linkablePartNative = \case u at DotO {} -> [u] u at DotA {} -> [u] u at DotDLL {} -> [u] - LazyBCOs _ os -> [DotO f ForeignObject | f <- os] _ -> [] linkablePartByteCode :: LinkablePart -> [LinkablePart] linkablePartByteCode = \case u at BCOs {} -> [u] - LazyBCOs bcos _ -> [BCOs bcos] _ -> [] -- | Transform the 'LinkablePart' list in this 'Linkable' to contain only ===================================== compiler/GHC/Runtime/Eval.hs ===================================== @@ -1275,14 +1275,14 @@ showModule mod_summary = let interpreted = case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> isJust (homeModInfoByteCode mod_info) && isNothing (homeModInfoObject mod_info) + Just mod_info -> homeModInfoHasByteCode mod_info && isNothing (homeModInfoObject mod_info) return (showSDoc dflags $ showModMsg dflags interpreted (ModuleNode [] mod_summary)) moduleIsBootOrNotObjectLinkable :: GhcMonad m => ModSummary -> m Bool moduleIsBootOrNotObjectLinkable mod_summary = withSession $ \hsc_env -> case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> return . isNothing $ homeModInfoByteCode mod_info + Just mod_info -> return . not $ homeModInfoHasByteCode mod_info ---------------------------------------------------------------------------- -- RTTI primitives ===================================== compiler/GHC/Runtime/Interpreter.hs ===================================== @@ -113,6 +113,7 @@ import qualified GHC.Exts.Heap as Heap import GHC.Stack.CCS (CostCentre,CostCentreStack) import System.Directory import System.Process +import GHC.Unit.Module.ModDetails (ModDetails(..)) {- Note [Remote GHCi] ~~~~~~~~~~~~~~~~~~ @@ -684,10 +685,13 @@ fromEvalResult (EvalSuccess a) = return a getModBreaks :: HomeModInfo -> ModBreaks getModBreaks hmi - | Just linkable <- homeModInfoByteCode hmi, + | HomeModByteCode linkable <- homeModInfoByteCode hmi, -- The linkable may have 'DotO's as well; only consider BCOs. See #20570. [cbc] <- linkableBCOs linkable = fromMaybe emptyModBreaks (bc_breaks cbc) + | ModDetails {md_bytecode = Just _} <- hm_details hmi + -- TODO + = emptyModBreaks | otherwise = emptyModBreaks -- probably object code ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -45,8 +47,6 @@ type PackageCompleteMatches = CompleteMatches type PackageIfaceTable = ModuleEnv ModIface -- Domain = modules in the imported packages -type PackageTypeTable = ModuleEnv TypeEnv - -- | Constructs an empty PackageIfaceTable emptyPackageIfaceTable :: PackageIfaceTable emptyPackageIfaceTable = emptyModuleEnv @@ -70,7 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv - , eps_PTT = emptyModuleEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -142,7 +142,11 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules - eps_PTT :: !PackageTypeTable, + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules ===================================== compiler/GHC/Unit/Home/ModInfo.hs ===================================== @@ -1,13 +1,19 @@ +{-# LANGUAGE LambdaCase #-} + -- | Info about modules in the "home" unit module GHC.Unit.Home.ModInfo ( HomeModInfo (..) - , HomeModLinkable(..) + , HomeModLinkable (..) + , HomeModByteCode (..) , homeModInfoObject , homeModInfoByteCode + , homeModInfoHasByteCode , emptyHomeModInfoLinkable , justBytecode , justObjects , bytecodeAndObjects + , pureHomeModByteCode + , evalHomeModByteCode , HomePackageTable , emptyHomePackageTable , lookupHpt @@ -44,6 +50,7 @@ import GHC.Utils.Outputable import Data.List (sortOn) import Data.Ord import GHC.Utils.Panic +import GHC.Unit.Module.WholeCoreBindings (WholeCoreBindings) -- | Information about modules in the package being compiled data HomeModInfo = HomeModInfo @@ -73,37 +80,67 @@ data HomeModInfo = HomeModInfo -- 'ModIface' (only). } -homeModInfoByteCode :: HomeModInfo -> Maybe Linkable +homeModInfoByteCode :: HomeModInfo -> HomeModByteCode homeModInfoByteCode = homeMod_bytecode . hm_linkable +homeModInfoHasByteCode :: HomeModInfo -> Bool +homeModInfoHasByteCode hmi = case homeModInfoByteCode hmi of + NoHomeModByteCode -> False + _ -> True + homeModInfoObject :: HomeModInfo -> Maybe Linkable homeModInfoObject = homeMod_object . hm_linkable emptyHomeModInfoLinkable :: HomeModLinkable -emptyHomeModInfoLinkable = HomeModLinkable Nothing Nothing +emptyHomeModInfoLinkable = HomeModLinkable NoHomeModByteCode Nothing -- See Note [Home module build products] -data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !(Maybe Linkable) +data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !HomeModByteCode , homeMod_object :: !(Maybe Linkable) } instance Outputable HomeModLinkable where ppr (HomeModLinkable l1 l2) = ppr l1 $$ ppr l2 -justBytecode :: Linkable -> HomeModLinkable -justBytecode lm = - assertPpr (not (linkableIsNativeCodeOnly lm)) (ppr lm) - $ emptyHomeModInfoLinkable { homeMod_bytecode = Just lm } +justBytecode :: HomeModByteCode -> HomeModLinkable +justBytecode bc = + emptyHomeModInfoLinkable { homeMod_bytecode = bc } justObjects :: Linkable -> HomeModLinkable justObjects lm = assertPpr (linkableIsNativeCodeOnly lm) (ppr lm) $ emptyHomeModInfoLinkable { homeMod_object = Just lm } -bytecodeAndObjects :: Linkable -> Linkable -> HomeModLinkable +bytecodeAndObjects :: HomeModByteCode -> Linkable -> HomeModLinkable bytecodeAndObjects bc o = - assertPpr (not (linkableIsNativeCodeOnly bc) && linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) - (HomeModLinkable (Just bc) (Just o)) - + assertPpr (linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) + (HomeModLinkable bc (Just o)) + +pureHomeModByteCode :: HomeModByteCode -> Maybe Linkable +pureHomeModByteCode = \case + NoHomeModByteCode -> Nothing + HomeModIfaceCore _ -> Nothing + HomeModByteCode l -> Just l + +-- | +evalHomeModByteCode :: HomeModInfo -> Maybe (IO Linkable) +evalHomeModByteCode HomeModInfo {hm_details, hm_linkable} + | HomeModByteCode bc <- homeMod_bytecode hm_linkable + = Just (pure bc) + | otherwise + = md_bytecode hm_details + +data HomeModByteCode = + NoHomeModByteCode + | + HomeModIfaceCore WholeCoreBindings + | + HomeModByteCode Linkable + +instance Outputable HomeModByteCode where + ppr = \case + NoHomeModByteCode -> text "no bytecode" + HomeModIfaceCore _ -> text "dehydrated Core" + HomeModByteCode linkable -> ppr linkable {- Note [Home module build products] ===================================== compiler/GHC/Unit/Module/ModDetails.hs ===================================== @@ -14,6 +14,9 @@ import GHC.Types.DefaultEnv ( DefaultEnv, emptyDefaultEnv ) import GHC.Types.TypeEnv import GHC.Types.Annotations ( Annotation ) +import GHC.Linker.Types (Linkable) +import GHC.Prelude + -- | The 'ModDetails' is essentially a cache for information in the 'ModIface' -- for home modules only. Information relating to packages will be loaded into -- global environments in 'ExternalPackageState'. @@ -40,6 +43,8 @@ data ModDetails = ModDetails , md_complete_matches :: CompleteMatches -- ^ Complete match pragmas for this module + + , md_bytecode :: !(Maybe (IO Linkable)) } -- | Constructs an empty ModDetails @@ -53,4 +58,5 @@ emptyModDetails = ModDetails , md_fam_insts = [] , md_anns = [] , md_complete_matches = [] + , md_bytecode = Nothing } ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -27,6 +27,8 @@ import Data.Maybe (fromMaybe) import System.FilePath (takeExtension) {- +TODO update + Note [Interface Files with Core Definitions] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -87,7 +89,12 @@ data WholeCoreBindings = WholeCoreBindings , wcb_foreign :: IfaceForeign } +instance Outputable WholeCoreBindings where + ppr WholeCoreBindings {wcb_module} = text "iface Core for " <+> ppr wcb_module + {- +TODO update + Note [Foreign stubs and TH bytecode linking] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== ghc/GHCi/Leak.hs ===================================== @@ -50,7 +50,7 @@ getLeakIndicators hsc_env = where mkWeakLinkables :: HomeModLinkable -> IO [Maybe (Weak Linkable)] mkWeakLinkables (HomeModLinkable mbc mo) = - mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [mbc, mo] + mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [pureHomeModByteCode mbc, mo] -- | Look at the LeakIndicators collected by an earlier call to -- `getLeakIndicators`, and print messasges if any of them are still View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4de4d0f59c2ab1757d13c6be99f7a90d1909651e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4de4d0f59c2ab1757d13c6be99f7a90d1909651e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 14:58:54 2024 From: gitlab at gitlab.haskell.org (Sylvain Henry (@hsyl20)) Date: Mon, 02 Sep 2024 10:58:54 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-0006-Adds-support-for-Hidden-symbols] 23 commits: compiler: Fix pretty printing of ticked prefix constructors (#24237) Message-ID: <66d5d2ae2498c_3f33692eedfc392f9@gitlab.mail> Sylvain Henry pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-0006-Adds-support-for-Hidden-symbols at Glasgow Haskell Compiler / GHC Commits: 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - b1b22c96 by doyougnu at 2024-09-02T16:58:28+02:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 30 changed files: - .gitignore - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/HsToCore/Usage.hs - compiler/GHC/Iface/Make.hs - compiler/GHC/Iface/Recomp.hs - compiler/GHC/Iface/Type.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/StgToJS/Literal.hs - compiler/GHC/SysTools/Cpp.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - compiler/GHC/Types/Id.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Module/ModIface.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - configure.ac - docs/users_guide/9.12.1-notes.rst The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f1cf9751dec367fdb53977d6d25a16cf2d1ed201...b1b22c9651a8cd311817f046bce1c6ce0ba414d2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f1cf9751dec367fdb53977d6d25a16cf2d1ed201...b1b22c9651a8cd311817f046bce1c6ce0ba414d2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 14:59:58 2024 From: gitlab at gitlab.haskell.org (Andrei Borzenkov (@sand-witch)) Date: Mon, 02 Sep 2024 10:59:58 -0400 Subject: [Git][ghc/ghc][wip/sand-witch/tysyn-info-ppr] Update GHCi :info type synonym and type family printing (24459) Message-ID: <66d5d2eea8cbb_3f33692bde503948d@gitlab.mail> Andrei Borzenkov pushed to branch wip/sand-witch/tysyn-info-ppr at Glasgow Haskell Compiler / GHC Commits: 2bbf0329 by Andrei Borzenkov at 2024-09-02T18:57:24+04:00 Update GHCi :info type synonym and type family printing (24459) - Do not print result's kind since we have full kind in SAKS and we display invisible arity using @-binders - Do not suppress significant invisible binders An invisible binder is considered significant when it meets at least one of the following three criteria: - It visibly occurs in the RHS type - It is not followed by a visible binder, so it affects the arity of a type synonym - It is followed by a significant binder, so it affects positioning - - - - - 23 changed files: - compiler/GHC/Core/TyCon.hs - compiler/GHC/Iface/Syntax.hs - compiler/GHC/Iface/Type.hs - testsuite/tests/backpack/should_fail/T19244a.stderr - testsuite/tests/backpack/should_fail/T19244b.stderr - testsuite/tests/backpack/should_fail/bkpfail46.stderr - testsuite/tests/ghci/T18060/T18060.stdout - testsuite/tests/ghci/scripts/ListTuplePunsPpr.stdout - testsuite/tests/ghci/scripts/T15941.stdout - testsuite/tests/ghci/scripts/T19310.stdout - testsuite/tests/ghci/scripts/T21294a.stdout - + testsuite/tests/ghci/scripts/T24459.script - + testsuite/tests/ghci/scripts/T24459.stdout - testsuite/tests/ghci/scripts/T8535.stdout - testsuite/tests/ghci/scripts/T9181.stdout - testsuite/tests/ghci/scripts/all.T - testsuite/tests/ghci/scripts/ghci020.stdout - testsuite/tests/ghci/should_run/T10145.stdout - testsuite/tests/ghci/should_run/T18594.stdout - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/ghc-experimental-exports.stdout - testsuite/tests/interface-stability/template-haskell-exports.stdout - testsuite/tests/rep-poly/RepPolyBackpack3.stderr Changes: ===================================== compiler/GHC/Core/TyCon.hs ===================================== @@ -25,7 +25,8 @@ module GHC.Core.TyCon( mkRequiredTyConBinder, mkAnonTyConBinder, mkAnonTyConBinders, tyConBinderForAllTyFlag, tyConBndrVisForAllTyFlag, isNamedTyConBinder, - isVisibleTyConBinder, isInvisibleTyConBinder, + isVisibleTyConBinder, isInvisSpecTyConBinder, isInvisibleTyConBinder, + isInferredTyConBinder, isVisibleTcbVis, isInvisSpecTcbVis, -- ** Field labels @@ -488,7 +489,7 @@ mkRequiredTyConBinder dep_set tv | tv `elemVarSet` dep_set = mkNamedTyConBinder Required tv | otherwise = mkAnonTyConBinder tv -tyConBinderForAllTyFlag :: TyConBinder -> ForAllTyFlag +tyConBinderForAllTyFlag :: VarBndr a TyConBndrVis -> ForAllTyFlag tyConBinderForAllTyFlag (Bndr _ vis) = tyConBndrVisForAllTyFlag vis tyConBndrVisForAllTyFlag :: TyConBndrVis -> ForAllTyFlag @@ -514,10 +515,22 @@ isInvisSpecTcbVis :: TyConBndrVis -> Bool isInvisSpecTcbVis (NamedTCB Specified) = True isInvisSpecTcbVis _ = False +isInvisInferTcbVis :: TyConBndrVis -> Bool +isInvisInferTcbVis (NamedTCB Inferred) = True +isInvisInferTcbVis _ = False + +isInvisSpecTyConBinder :: VarBndr tv TyConBndrVis -> Bool +-- Works for IfaceTyConBinder too +isInvisSpecTyConBinder (Bndr _ tcb_vis) = isInvisSpecTcbVis tcb_vis + isInvisibleTyConBinder :: VarBndr tv TyConBndrVis -> Bool -- Works for IfaceTyConBinder too isInvisibleTyConBinder tcb = not (isVisibleTyConBinder tcb) +isInferredTyConBinder :: VarBndr var TyConBndrVis -> Bool +-- Works for IfaceTyConBinder too +isInferredTyConBinder (Bndr _ tcb_vis) = isInvisInferTcbVis tcb_vis + -- Build the 'tyConKind' from the binders and the result kind. -- Keep in sync with 'mkTyConKind' in GHC.Iface.Type. mkTyConKind :: [TyConBinder] -> Kind -> Kind ===================================== compiler/GHC/Iface/Syntax.hs ===================================== @@ -98,6 +98,7 @@ import Control.Monad import System.IO.Unsafe import Control.DeepSeq import Data.Proxy +import qualified Data.Set as Set infixl 3 &&& @@ -1051,17 +1052,18 @@ pprIfaceDecl ss (IfaceClass { ifName = clas pprIfaceDecl ss (IfaceSynonym { ifName = tc , ifBinders = binders - , ifSynRhs = mono_ty + , ifSynRhs = poly_ty , ifResKind = res_kind}) = vcat [ pprStandaloneKindSig name_doc (mkIfaceTyConKind binders res_kind) - , hang (text "type" <+> pprIfaceDeclHead suppress_bndr_sig [] ss tc binders <+> equals) - 2 (sep [ pprIfaceForAll tvs, pprIfaceContextArr theta, ppr_tau - , ppUnless (isIfaceLiftedTypeKind res_kind) (dcolon <+> ppr res_kind) ]) + , hang (text "type" <+> decl_head <+> equals) + 2 (sep [ pprIfaceForAll tvs, pprIfaceContextArr theta, ppr_tau ]) ] where - (tvs, theta, tau) = splitIfaceSigmaTy mono_ty + (tvs, theta, tau) = splitIfaceSigmaTy poly_ty name_doc = pprPrefixIfDeclBndr (ss_how_much ss) (occName tc) + decl_head = pprIfaceDeclHeadWith invisibles_suppressor suppress_bndr_sig [] ss tc binders + -- See Note [Printing type abbreviations] in GHC.Iface.Type ppr_tau | tc `hasKey` liftedTypeKindTyConKey || tc `hasKey` unrestrictedFunTyConKey || @@ -1072,6 +1074,10 @@ pprIfaceDecl ss (IfaceSynonym { ifName = tc -- See Note [Suppressing binder signatures] in GHC.Iface.Type suppress_bndr_sig = SuppressBndrSig True + invisibles_suppressor = + suppressIfaceInsignificantInvisibles + (\var -> ifTyConBinderName var `ifTypeVarVisiblyOccurs` tau) + pprIfaceDecl ss (IfaceFamily { ifName = tycon , ifFamFlav = rhs, ifBinders = binders , ifResKind = res_kind @@ -1083,9 +1089,7 @@ pprIfaceDecl ss (IfaceFamily { ifName = tycon | otherwise = vcat [ pprStandaloneKindSig name_doc (mkIfaceTyConKind binders res_kind) - , hang (text "type family" - <+> pprIfaceDeclHead suppress_bndr_sig [] ss tycon binders - <+> pp_inj res_var inj + , hang (text "type family" <+> decl_head <+> pp_inj res_var inj <+> ppShowRhs ss (pp_where rhs)) 2 (ppShowRhs ss (pp_rhs rhs)) $$ @@ -1094,6 +1098,8 @@ pprIfaceDecl ss (IfaceFamily { ifName = tycon where name_doc = pprPrefixIfDeclBndr (ss_how_much ss) (occName tycon) + decl_head = pprIfaceDeclHeadWith invisibles_suppressor suppress_bndr_sig [] ss tycon binders + pp_where (IfaceClosedSynFamilyTyCon {}) = text "where" pp_where _ = empty @@ -1130,6 +1136,16 @@ pprIfaceDecl ss (IfaceFamily { ifName = tycon -- See Note [Suppressing binder signatures] in GHC.Iface.Type suppress_bndr_sig = SuppressBndrSig True + invisibles_suppressor = + suppressIfaceInsignificantInvisibles + (\var -> ifTyConBinderName var `Set.member` mentioned_vars) + + mentioned_vars + | Just{} <- res_var + , Injective injectivity <- inj + = Set.fromList . map (ifTyConBinderName . snd) . filter fst $ zip injectivity binders + | otherwise = mempty + pprIfaceDecl _ (IfacePatSyn { ifName = name, ifPatUnivBndrs = univ_bndrs, ifPatExBndrs = ex_bndrs, ifPatProvCtxt = prov_ctxt, ifPatReqCtxt = req_ctxt, @@ -1239,16 +1255,26 @@ pprIfaceTyConParent IfNoParent pprIfaceTyConParent (IfDataInstance _ tc tys) = pprIfaceTypeApp topPrec tc tys + +-- The most common use case of `pprIfaceDeclHeadWith` pprIfaceDeclHead :: SuppressBndrSig -> IfaceContext -> ShowSub -> Name -> [IfaceTyConBinder] -- of the tycon, for invisible-suppression -> SDoc -pprIfaceDeclHead suppress_sig context ss tc_occ bndrs +pprIfaceDeclHead + = pprIfaceDeclHeadWith (\pk bndrs -> suppressIfaceInvisibles pk bndrs bndrs) + +pprIfaceDeclHeadWith :: (PrintExplicitKinds -> [IfaceTyConBinder] -> [IfaceTyConBinder]) + -> SuppressBndrSig + -> IfaceContext -> ShowSub -> Name + -> [IfaceTyConBinder] -- of the tycon, for invisible-suppression + -> SDoc +pprIfaceDeclHeadWith suppress_invisibles suppress_sig context ss tc_occ bndrs = sdocOption sdocPrintExplicitKinds $ \print_kinds -> sep [ pprIfaceContextArr context , pprPrefixIfDeclBndr (ss_how_much ss) (occName tc_occ) <+> pprIfaceTyConBinders suppress_sig - (suppressIfaceInvisibles (PrintExplicitKinds print_kinds) bndrs bndrs) ] + (suppress_invisibles (PrintExplicitKinds print_kinds) bndrs) ] pprIfaceConDecl :: ShowSub -> Bool -> IfaceTopBndr ===================================== compiler/GHC/Iface/Type.hs ===================================== @@ -56,7 +56,8 @@ module GHC.Iface.Type ( pprIfaceCoTcApp, pprTyTcApp, pprIfacePrefixApp, isIfaceRhoType, - suppressIfaceInvisibles, + suppressIfaceInvisibles, suppressIfaceInsignificantInvisibles, + ifTypeVarVisiblyOccurs, stripIfaceInvisVars, stripInvisArgs, @@ -99,6 +100,8 @@ import Data.Word (Word8) import Control.Arrow (first) import Control.DeepSeq import Control.Monad ((<$!>)) +import Data.List (dropWhileEnd) +import qualified Data.List.NonEmpty as NonEmpty {- ************************************************************************ @@ -623,6 +626,42 @@ suppressIfaceInvisibles (PrintExplicitKinds False) tys xs = suppress tys xs | isInvisibleTyConBinder k = suppress ks xs | otherwise = x : suppress ks xs +-- An invisible binder is considered significant when it meets at least +-- one of the following three criteria: +-- - It visibly occurs in the RHS type +-- - It is not followed by a visible binder, so it +-- affects the arity of a type synonym +-- - It is followed by a significant binder, +-- so it affects positioning +suppressIfaceInsignificantInvisibles :: (IfaceTyConBinder -> Bool) + -> PrintExplicitKinds + -> [IfaceTyConBinder] + -> [IfaceTyConBinder] +suppressIfaceInsignificantInvisibles _ (PrintExplicitKinds True) tys = tys +suppressIfaceInsignificantInvisibles mentioned_var (PrintExplicitKinds False) tys = suppress tys + where + -- Consider this example: + -- type T :: forall k1 k2. Type + -- type T @a @b = b + -- `@a` is not mentioned on the RHS however we can't just + -- drop it because implicit argument positioning matters. + -- + -- Hence just drop the end + mentioned_vars = dropWhileEnd (not . mentioned_var) + + suppress_invisible_groups [] = [] + suppress_invisible_groups [group] = NonEmpty.toList group -- the last group affects arity + suppress_invisible_groups (group : groups) + = applyWhen invis_group mentioned_vars bndrs ++ suppress_invisible_groups groups + where + bndrs = NonEmpty.toList group + invis_group = isInvisibleTyConBinder (NonEmpty.head group) + + suppress + = suppress_invisible_groups -- Filter out insignificant invisible binders + . NonEmpty.groupWith isInvisibleTyConBinder -- Find chunks of @-binders + . filterOut isInferredTyConBinder -- We don't want to display @{binders} + stripIfaceInvisVars :: PrintExplicitKinds -> [IfaceTyConBinder] -> [IfaceTyConBinder] stripIfaceInvisVars (PrintExplicitKinds True) tyvars = tyvars stripIfaceInvisVars (PrintExplicitKinds False) tyvars @@ -663,6 +702,27 @@ ifTypeIsVarFree ty = go ty go_args IA_Nil = True go_args (IA_Arg arg _ args) = go arg && go_args args +ifTypeVarVisiblyOccurs :: IfLclName -> IfaceType -> Bool +-- Returns True if the type contains this name. Doesn't count +-- invisible application +-- Just used to control pretty printing +ifTypeVarVisiblyOccurs name ty = go ty + where + go (IfaceTyVar var) = var == name + go (IfaceFreeTyVar {}) = False + go (IfaceAppTy fun args) = go fun || go_args args + go (IfaceFunTy _ w arg res) = go w || go arg || go res + go (IfaceForAllTy bndr ty) = go (ifaceBndrType (binderVar bndr)) || go ty + go (IfaceTyConApp _ args) = go_args args + go (IfaceTupleTy _ _ args) = go_args args + go (IfaceLitTy _) = False + go (IfaceCastTy {}) = False -- Safe + go (IfaceCoercionTy {}) = False -- Safe + + go_args IA_Nil = False + go_args (IA_Arg arg Required args) = go arg || go_args args + go_args (IA_Arg _arg _ args) = go_args args + {- Note [Substitution on IfaceType] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Substitutions on IfaceType are done only during pretty-printing to ===================================== testsuite/tests/backpack/should_fail/T19244a.stderr ===================================== @@ -17,7 +17,7 @@ T19244a.bkp:22:9: error: [GHC-15843] • Type constructor ‘Key’ has conflicting definitions in the module and its hsig file. Main module: type Key :: * -> Constraint - type Key = GHC.Classes.Ord :: * -> Constraint + type Key = GHC.Classes.Ord Hsig file: type Key :: forall {k}. k -> Constraint class Key k1 The types have different kinds. ===================================== testsuite/tests/backpack/should_fail/T19244b.stderr ===================================== @@ -1,20 +1,20 @@ [1 of 3] Processing user - [1 of 2] Compiling Map[sig] ( user\Map.hsig, nothing ) - [2 of 2] Compiling User ( user\User.hs, nothing ) + [1 of 2] Compiling Map[sig] ( user/Map.hsig, nothing ) + [2 of 2] Compiling User ( user/User.hs, nothing ) [2 of 3] Processing ordmap Instantiating ordmap - [1 of 1] Compiling Map ( ordmap\Map.hs, T19244b.out\ordmap\Map.o ) + [1 of 1] Compiling Map ( ordmap/Map.hs, T19244b.out/ordmap/Map.o ) [3 of 3] Processing main Instantiating main [1 of 1] Including user[Map=ordmap:Map] Instantiating user[Map=ordmap:Map] - [1 of 2] Compiling Map[sig] ( user\Map.hsig, T19244b.out\user\user-GzloW2NeDdA2M0V8qzN4g2\Map.o ) + [1 of 2] Compiling Map[sig] ( user/Map.hsig, T19244b.out/user/user-GzloW2NeDdA2M0V8qzN4g2/Map.o ) T19244b.bkp:11:27: error: [GHC-15843] • Type constructor ‘Key’ has conflicting definitions in the module and its hsig file. Main module: type Key :: * -> Constraint - type Key = GHC.Classes.Ord :: * -> Constraint + type Key = GHC.Classes.Ord Hsig file: type Key :: forall {k}. k -> Constraint class Key k1 The types have different kinds. ===================================== testsuite/tests/backpack/should_fail/bkpfail46.stderr ===================================== @@ -1,20 +1,20 @@ [1 of 3] Processing p - [1 of 2] Compiling A[sig] ( p\A.hsig, nothing ) - [2 of 2] Compiling M ( p\M.hs, nothing ) + [1 of 2] Compiling A[sig] ( p/A.hsig, nothing ) + [2 of 2] Compiling M ( p/M.hs, nothing ) [2 of 3] Processing q Instantiating q - [1 of 1] Compiling A ( q\A.hs, bkpfail46.out\q\A.o ) + [1 of 1] Compiling A ( q/A.hs, bkpfail46.out/q/A.o ) [3 of 3] Processing r Instantiating r [1 of 1] Including p[A=q:A] Instantiating p[A=q:A] - [1 of 2] Compiling A[sig] ( p\A.hsig, bkpfail46.out\p\p-HVmFlcYSefiK5n1aDP1v7x\A.o ) + [1 of 2] Compiling A[sig] ( p/A.hsig, bkpfail46.out/p/p-HVmFlcYSefiK5n1aDP1v7x/A.o ) bkpfail46.bkp:16:9: error: [GHC-15843] • Type constructor ‘K’ has conflicting definitions in the module and its hsig file. Main module: type K :: * -> Constraint - type K a = GHC.Classes.Eq a :: Constraint + type K a = GHC.Classes.Eq a Hsig file: type K :: * -> Constraint class K a Illegal parameterized type synonym in implementation of abstract data. ===================================== testsuite/tests/ghci/T18060/T18060.stdout ===================================== @@ -1,5 +1,5 @@ type (->) :: * -> * -> * -type (->) = FUN Many :: * -> * -> * +type (->) = FUN Many -- Defined in ‘GHC.Types’ infixr -1 -> instance Monoid b => Monoid (a -> b) ===================================== testsuite/tests/ghci/scripts/ListTuplePunsPpr.stdout ===================================== @@ -74,13 +74,13 @@ data Tuple2# a b = (#,#) a b (# , #) :: a -> b -> Tuple2# a b (Int, Int) :: Tuple2 (*) (*) type T :: Tuple2 (*) (*) -type T = (Int, Int) :: Tuple2 (*) (*) +type T = (Int, Int) -- Defined at :19:1 type S :: Solo (*) -type S = MkSolo Int :: Solo (*) +type S = MkSolo Int -- Defined at :20:1 type L :: List (*) -type L = [Int] :: List (*) +type L = [Int] -- Defined at :21:1 f :: Int -> Tuple2 Int Int -> Int (\ (_, _) -> ()) :: Tuple2 a b -> Unit ===================================== testsuite/tests/ghci/scripts/T15941.stdout ===================================== @@ -1,4 +1,3 @@ type T :: * -> * -> * -type T = - (->) @{GHC.Types.LiftedRep} @{GHC.Types.LiftedRep} :: * -> * -> * +type T = (->) @{GHC.Types.LiftedRep} @{GHC.Types.LiftedRep} -- Defined at :2:1 ===================================== testsuite/tests/ghci/scripts/T19310.stdout ===================================== @@ -1,3 +1,3 @@ type T :: * -> * -> * -type T = (->) @{LiftedRep} @{LiftedRep} :: * -> * -> * +type T = (->) @{LiftedRep} @{LiftedRep} -- Defined at :3:1 ===================================== testsuite/tests/ghci/scripts/T21294a.stdout ===================================== @@ -1,5 +1,5 @@ type L0 :: * -> * -type L0 = [] :: * -> * +type L0 = [] -- Defined at :1:1 type L1 :: * -> * type L1 a = [a] ===================================== testsuite/tests/ghci/scripts/T24459.script ===================================== @@ -0,0 +1,50 @@ +:set -XTypeAbstractions + +:{ +import Data.Kind +import Data.Data + +type T7 :: forall a b c. b -> forall d. Type +type T7 @a @b @c f @d = b + +type T6 :: forall k. Type -> Type +type T6 _a = () + +type T5 :: forall a. Type -> Type +type T5 @a _b = Proxy _b + +type T4 :: forall a. Type -> Type +type T4 @a _b = Proxy a + +type T3 :: forall k. k -> Type +type T3 a = Proxy a + +type T2 :: forall k. k -> Type +type T2 @k = Proxy + +type T1 :: forall k. k -> Type +type T1 = Proxy + +type T0 :: forall k. k -> Type +type T0 = Proxy :: forall k. k -> Type +:} + +:i T0 +:i T1 +:i T2 +:i T3 +:i T4 +:i T5 +:i T6 +:i T7 + +:set -fprint-explicit-kinds + +:i T0 +:i T1 +:i T2 +:i T3 +:i T4 +:i T5 +:i T6 +:i T7 ===================================== testsuite/tests/ghci/scripts/T24459.stdout ===================================== @@ -0,0 +1,50 @@ +type T0 :: forall k. k -> * +type T0 = Proxy + -- Defined at :29:1 +type T1 :: forall k. k -> * +type T1 = Proxy + -- Defined at :26:1 +type T2 :: forall k. k -> * +type T2 @k = Proxy + -- Defined at :23:1 +type T3 :: forall k. k -> * +type T3 a = Proxy a + -- Defined at :20:1 +type T4 :: forall {k} (a :: k). * -> * +type T4 @a _b = Proxy a + -- Defined at :17:1 +type T5 :: forall {k} (a :: k). * -> * +type T5 _b = Proxy _b + -- Defined at :14:1 +type T6 :: forall {k} (k1 :: k). * -> * +type T6 _a = () + -- Defined at :11:1 +type T7 :: forall {k} {k1} {k2} (a :: k) b (c :: k1). + b -> forall (d :: k2). * +type T7 @a @b f @d = b + -- Defined at :8:1 +type T0 :: forall k. k -> * +type T0 = Proxy + -- Defined at :29:1 +type T1 :: forall k. k -> * +type T1 = Proxy + -- Defined at :26:1 +type T2 :: forall k. k -> * +type T2 @k = Proxy @{k} + -- Defined at :23:1 +type T3 :: forall k. k -> * +type T3 @k a = Proxy @{k} a + -- Defined at :20:1 +type T4 :: forall {k} (a :: k). * -> * +type T4 @{k} @a _b = Proxy @{k} a + -- Defined at :17:1 +type T5 :: forall {k} (a :: k). * -> * +type T5 @{k} @a _b = Proxy @{*} _b + -- Defined at :14:1 +type T6 :: forall {k} (k1 :: k). * -> * +type T6 @{k} @k1 _a = () + -- Defined at :11:1 +type T7 :: forall {k} {k1} {k2} (a :: k) b (c :: k1). + b -> forall (d :: k2). * +type T7 @{k} @{k1} @{k2} @a @b @c f @d = b + -- Defined at :8:1 ===================================== testsuite/tests/ghci/scripts/T8535.stdout ===================================== @@ -1,5 +1,5 @@ type (->) :: * -> * -> * -type (->) = FUN Many :: * -> * -> * +type (->) = FUN Many -- Defined in ‘GHC.Types’ infixr -1 -> instance Monoid b => Monoid (a -> b) ===================================== testsuite/tests/ghci/scripts/T9181.stdout ===================================== @@ -13,12 +13,10 @@ type (GHC.Internal.Data.Type.Ord.<=) x y = GHC.Internal.TypeError.Assert (x GHC.Internal.Data.Type.Ord.<=? y) (GHC.Internal.Data.Type.Ord.LeErrMsg x y) - :: Constraint type (GHC.Internal.Data.Type.Ord.<=?) :: forall k. k -> k -> Bool type (GHC.Internal.Data.Type.Ord.<=?) m n = GHC.Internal.Data.Type.Ord.OrdCond (GHC.Internal.Data.Type.Ord.Compare m n) True True False - :: Bool type GHC.Internal.TypeLits.AppendSymbol :: GHC.Types.Symbol -> GHC.Types.Symbol -> GHC.Types.Symbol type family GHC.Internal.TypeLits.AppendSymbol a b ===================================== testsuite/tests/ghci/scripts/all.T ===================================== @@ -384,3 +384,4 @@ test('T23686', normal, ghci_script, ['T23686.script']) test('T13869', extra_files(['T13869a.hs', 'T13869b.hs']), ghci_script, ['T13869.script']) test('ListTuplePunsPpr', normal, ghci_script, ['ListTuplePunsPpr.script']) test('ListTuplePunsPprNoAbbrevTuple', [expect_broken(23135), limit_stdout_lines(13)], ghci_script, ['ListTuplePunsPprNoAbbrevTuple.script']) +test('T24459', normal, ghci_script, ['T24459.script']) ===================================== testsuite/tests/ghci/scripts/ghci020.stdout ===================================== @@ -1,5 +1,5 @@ type (->) :: * -> * -> * -type (->) = FUN Many :: * -> * -> * +type (->) = FUN Many -- Defined in ‘GHC.Types’ infixr -1 -> instance Monoid b => Monoid (a -> b) ===================================== testsuite/tests/ghci/should_run/T10145.stdout ===================================== @@ -1,5 +1,5 @@ type (->) :: * -> * -> * -type (->) = FUN Many :: * -> * -> * +type (->) = FUN Many -- Defined in ‘GHC.Types’ infixr -1 -> instance Monoid b => Monoid (a -> b) ===================================== testsuite/tests/ghci/should_run/T18594.stdout ===================================== @@ -1,5 +1,5 @@ type (->) :: * -> * -> * -type (->) = FUN Many :: * -> * -> * +type (->) = FUN Many -- Defined in ‘GHC.Types’ infixr -1 -> instance Monoid b => Monoid (a -> b) ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -1766,27 +1766,27 @@ module Data.Type.Equality where module Data.Type.Ord where -- Safety: Safe type (<) :: forall {t}. t -> t -> Constraint - type (<) x y = GHC.Internal.TypeError.Assert (x t -> Constraint - type (<=) x y = GHC.Internal.TypeError.Assert (x <=? y) (GHC.Internal.Data.Type.Ord.LeErrMsg x y) :: Constraint + type (<=) x y = GHC.Internal.TypeError.Assert (x <=? y) (GHC.Internal.Data.Type.Ord.LeErrMsg x y) type (<=?) :: forall k. k -> k -> GHC.Types.Bool - type (<=?) m n = OrdCond (Compare m n) GHC.Types.True GHC.Types.True GHC.Types.False :: GHC.Types.Bool + type (<=?) m n = OrdCond (Compare m n) GHC.Types.True GHC.Types.True GHC.Types.False type ( k -> GHC.Types.Bool - type () :: forall {t}. t -> t -> Constraint - type (>) x y = GHC.Internal.TypeError.Assert (x >? y) (GHC.Internal.Data.Type.Ord.GtErrMsg x y) :: Constraint + type (>) x y = GHC.Internal.TypeError.Assert (x >? y) (GHC.Internal.Data.Type.Ord.GtErrMsg x y) type (>=) :: forall {t}. t -> t -> Constraint - type (>=) x y = GHC.Internal.TypeError.Assert (x >=? y) (GHC.Internal.Data.Type.Ord.GeErrMsg x y) :: Constraint + type (>=) x y = GHC.Internal.TypeError.Assert (x >=? y) (GHC.Internal.Data.Type.Ord.GeErrMsg x y) type (>=?) :: forall k. k -> k -> GHC.Types.Bool - type (>=?) m n = OrdCond (Compare m n) GHC.Types.False GHC.Types.True GHC.Types.True :: GHC.Types.Bool + type (>=?) m n = OrdCond (Compare m n) GHC.Types.False GHC.Types.True GHC.Types.True type (>?) :: forall k. k -> k -> GHC.Types.Bool - type (>?) m n = OrdCond (Compare m n) GHC.Types.False GHC.Types.False GHC.Types.True :: GHC.Types.Bool + type (>?) m n = OrdCond (Compare m n) GHC.Types.False GHC.Types.False GHC.Types.True type Compare :: forall k. k -> k -> GHC.Types.Ordering type family Compare a b type Max :: forall k. k -> k -> k - type Max m n = OrdCond (Compare m n) n n m :: k + type Max m n = OrdCond (Compare m n) n n m type Min :: forall k. k -> k -> k - type Min m n = OrdCond (Compare m n) m m n :: k + type Min m n = OrdCond (Compare m n) m m n type OrdCond :: forall k. GHC.Types.Ordering -> k -> k -> k -> k type family OrdCond o lt eq gt where forall k (lt :: k) (eq :: k) (gt :: k). OrdCond GHC.Types.LT lt eq gt = lt @@ -3321,7 +3321,7 @@ module GHC.Base where type Levity :: * data Levity = Lifted | Unlifted type LiftedRep :: RuntimeRep - type LiftedRep = BoxedRep Lifted :: RuntimeRep + type LiftedRep = BoxedRep Lifted type List :: * -> * data List a = ... type role MVar# nominal representational @@ -3434,7 +3434,7 @@ module GHC.Base where type TypeLitSort :: * data TypeLitSort = TypeLitSymbol | TypeLitNat | TypeLitChar type UnliftedRep :: RuntimeRep - type UnliftedRep = BoxedRep Unlifted :: RuntimeRep + type UnliftedRep = BoxedRep Unlifted type UnliftedType :: * type UnliftedType = TYPE UnliftedRep type VecCount :: * @@ -3444,7 +3444,7 @@ module GHC.Base where type Void :: * data Void type Void# :: ZeroBitType - type Void# = (# #) :: ZeroBitType + type Void# = (# #) type Weak# :: forall {l :: Levity}. TYPE (BoxedRep l) -> UnliftedType data Weak# a type WithDict :: Constraint -> * -> Constraint @@ -3490,7 +3490,7 @@ module GHC.Base where type WordBox :: TYPE WordRep -> * data WordBox a = MkWordBox a type ZeroBitRep :: RuntimeRep - type ZeroBitRep = TupleRep '[] :: RuntimeRep + type ZeroBitRep = TupleRep '[] type ZeroBitType :: * type ZeroBitType = TYPE ZeroBitRep absentErr :: forall a. a @@ -5519,7 +5519,7 @@ module GHC.Exts where type Levity :: * data Levity = Lifted | Unlifted type LiftedRep :: RuntimeRep - type LiftedRep = BoxedRep Lifted :: RuntimeRep + type LiftedRep = BoxedRep Lifted type List :: * -> * data List a = ... type role MVar# nominal representational @@ -5594,7 +5594,7 @@ module GHC.Exts where TypeLitNat :: GHC.Types.TypeLitSort TypeLitSymbol :: GHC.Types.TypeLitSort type UnliftedRep :: RuntimeRep - type UnliftedRep = BoxedRep Unlifted :: RuntimeRep + type UnliftedRep = BoxedRep Unlifted type UnliftedType :: * type UnliftedType = TYPE UnliftedRep type VecCount :: * @@ -5602,7 +5602,7 @@ module GHC.Exts where type VecElem :: * data VecElem = Int8ElemRep | Int16ElemRep | Int32ElemRep | Int64ElemRep | Word8ElemRep | Word16ElemRep | Word32ElemRep | Word64ElemRep | FloatElemRep | DoubleElemRep type Void# :: ZeroBitType - type Void# = (# #) :: ZeroBitType + type Void# = (# #) type Weak# :: forall {l :: Levity}. TYPE (BoxedRep l) -> UnliftedType data Weak# a type WithDict :: Constraint -> * -> Constraint @@ -5648,7 +5648,7 @@ module GHC.Exts where type WordBox :: TYPE WordRep -> * data WordBox a = MkWordBox a type ZeroBitRep :: RuntimeRep - type ZeroBitRep = TupleRep '[] :: RuntimeRep + type ZeroBitRep = TupleRep '[] type ZeroBitType :: * type ZeroBitType = TYPE ZeroBitRep acosDouble# :: Double# -> Double# @@ -7383,7 +7383,7 @@ module GHC.Generics where type C :: * data C type C1 :: forall {k}. Meta -> (k -> *) -> k -> * - type C1 = M1 C :: Meta -> (k -> *) -> k -> * + type C1 = M1 C type Constructor :: forall {k}. k -> Constraint class Constructor c where conName :: forall k1 (t :: k -> (k1 -> *) -> k1 -> *) (f :: k1 -> *) (a :: k1). t c f a -> [GHC.Types.Char] @@ -7393,7 +7393,7 @@ module GHC.Generics where type D :: * data D type D1 :: forall {k}. Meta -> (k -> *) -> k -> * - type D1 = M1 D :: Meta -> (k -> *) -> k -> * + type D1 = M1 D type Datatype :: forall {k}. k -> Constraint class Datatype d where datatypeName :: forall k1 (t :: k -> (k1 -> *) -> k1 -> *) (f :: k1 -> *) (a :: k1). t d f a -> [GHC.Types.Char] @@ -7440,14 +7440,14 @@ module GHC.Generics where type R :: * data R type Rec0 :: forall {k}. * -> k -> * - type Rec0 = K1 R :: * -> k -> * + type Rec0 = K1 R type role Rec1 representational nominal type Rec1 :: forall k. (k -> *) -> k -> * newtype Rec1 f p = Rec1 {unRec1 :: f p} type S :: * data S type S1 :: forall {k}. Meta -> (k -> *) -> k -> * - type S1 = M1 S :: Meta -> (k -> *) -> k -> * + type S1 = M1 S type Selector :: forall {k}. k -> Constraint class Selector s where selName :: forall k1 (t :: k -> (k1 -> *) -> k1 -> *) (f :: k1 -> *) (a :: k1). t s f a -> [GHC.Types.Char] @@ -7464,24 +7464,24 @@ module GHC.Generics where data U1 p = U1 UAddr :: forall k (p :: k). GHC.Prim.Addr# -> URec (GHC.Internal.Ptr.Ptr ()) p type UAddr :: forall {k}. k -> * - type UAddr = URec (GHC.Internal.Ptr.Ptr ()) :: k -> * + type UAddr = URec (GHC.Internal.Ptr.Ptr ()) UChar :: forall k (p :: k). GHC.Prim.Char# -> URec GHC.Types.Char p type UChar :: forall {k}. k -> * - type UChar = URec GHC.Types.Char :: k -> * + type UChar = URec GHC.Types.Char UDouble :: forall k (p :: k). GHC.Prim.Double# -> URec GHC.Types.Double p type UDouble :: forall {k}. k -> * - type UDouble = URec GHC.Types.Double :: k -> * + type UDouble = URec GHC.Types.Double UFloat :: forall k (p :: k). GHC.Prim.Float# -> URec GHC.Types.Float p type UFloat :: forall {k}. k -> * - type UFloat = URec GHC.Types.Float :: k -> * + type UFloat = URec GHC.Types.Float UInt :: forall k (p :: k). GHC.Prim.Int# -> URec GHC.Types.Int p type UInt :: forall {k}. k -> * - type UInt = URec GHC.Types.Int :: k -> * + type UInt = URec GHC.Types.Int type URec :: forall k. * -> k -> * data family URec a p UWord :: forall k (p :: k). GHC.Prim.Word# -> URec GHC.Types.Word p type UWord :: forall {k}. k -> * - type UWord = URec GHC.Types.Word :: k -> * + type UWord = URec GHC.Types.Word type role V1 phantom type V1 :: forall k. k -> * data V1 p @@ -8569,7 +8569,7 @@ module GHC.Num.BigNat where type BigNat :: * data BigNat = BN# {unBigNat :: BigNat#} type BigNat# :: GHC.Types.UnliftedType - type BigNat# = GHC.Num.WordArray.WordArray# :: GHC.Types.UnliftedType + type BigNat# = GHC.Num.WordArray.WordArray# bigNatAdd :: BigNat# -> BigNat# -> BigNat# bigNatAddWord :: BigNat# -> GHC.Types.Word -> BigNat# bigNatAddWord# :: BigNat# -> GHC.Prim.Word# -> BigNat# @@ -9343,7 +9343,7 @@ module GHC.Stack where type CostCentreStack :: * data CostCentreStack type HasCallStack :: Constraint - type HasCallStack = ?callStack::CallStack :: Constraint + type HasCallStack = ?callStack::CallStack type SrcLoc :: * data SrcLoc = SrcLoc {srcLocPackage :: [GHC.Types.Char], srcLocModule :: [GHC.Types.Char], srcLocFile :: [GHC.Types.Char], srcLocStartLine :: GHC.Types.Int, srcLocStartCol :: GHC.Types.Int, srcLocEndLine :: GHC.Types.Int, srcLocEndCol :: GHC.Types.Int} callStack :: HasCallStack => CallStack @@ -9404,7 +9404,7 @@ module GHC.Stack.Types where type CallStack :: * data CallStack = EmptyCallStack | PushCallStack [GHC.Types.Char] SrcLoc CallStack | FreezeCallStack CallStack type HasCallStack :: Constraint - type HasCallStack = ?callStack::CallStack :: Constraint + type HasCallStack = ?callStack::CallStack type SrcLoc :: * data SrcLoc = SrcLoc {srcLocPackage :: [GHC.Types.Char], srcLocModule :: [GHC.Types.Char], srcLocFile :: [GHC.Types.Char], srcLocStartLine :: GHC.Types.Int, srcLocStartCol :: GHC.Types.Int, srcLocEndLine :: GHC.Types.Int, srcLocEndCol :: GHC.Types.Int} emptyCallStack :: CallStack @@ -9559,9 +9559,9 @@ module GHC.TypeLits where type (-) :: Natural -> Natural -> Natural type family (-) a b type (<=) :: forall {t}. t -> t -> Constraint - type (<=) x y = GHC.Internal.TypeError.Assert (x <=? y) (GHC.Internal.Data.Type.Ord.LeErrMsg x y) :: Constraint + type (<=) x y = GHC.Internal.TypeError.Assert (x <=? y) (GHC.Internal.Data.Type.Ord.LeErrMsg x y) type (<=?) :: forall k. k -> k -> GHC.Types.Bool - type (<=?) m n = GHC.Internal.Data.Type.Ord.OrdCond (GHC.Internal.Data.Type.Ord.Compare m n) GHC.Types.True GHC.Types.True GHC.Types.False :: GHC.Types.Bool + type (<=?) m n = GHC.Internal.Data.Type.Ord.OrdCond (GHC.Internal.Data.Type.Ord.Compare m n) GHC.Types.True GHC.Types.True GHC.Types.False type AppendSymbol :: Symbol -> Symbol -> Symbol type family AppendSymbol a b type CharToNat :: GHC.Types.Char -> Natural @@ -9678,9 +9678,9 @@ module GHC.TypeNats where type (-) :: Natural -> Natural -> Natural type family (-) a b type (<=) :: forall {t}. t -> t -> Constraint - type (<=) x y = GHC.Internal.TypeError.Assert (x <=? y) (GHC.Internal.Data.Type.Ord.LeErrMsg x y) :: Constraint + type (<=) x y = GHC.Internal.TypeError.Assert (x <=? y) (GHC.Internal.Data.Type.Ord.LeErrMsg x y) type (<=?) :: forall k. k -> k -> GHC.Types.Bool - type (<=?) m n = GHC.Internal.Data.Type.Ord.OrdCond (GHC.Internal.Data.Type.Ord.Compare m n) GHC.Types.True GHC.Types.True GHC.Types.False :: GHC.Types.Bool + type (<=?) m n = GHC.Internal.Data.Type.Ord.OrdCond (GHC.Internal.Data.Type.Ord.Compare m n) GHC.Types.True GHC.Types.True GHC.Types.False type CmpNat :: Natural -> Natural -> GHC.Types.Ordering type family CmpNat a b type Div :: Natural -> Natural -> Natural ===================================== testsuite/tests/interface-stability/ghc-experimental-exports.stdout ===================================== @@ -1480,9 +1480,9 @@ module Data.Tuple.Experimental where class a => CSolo a {-# MINIMAL #-} type CTuple0 :: Constraint - type CTuple0 = () :: Constraint :: Constraint + type CTuple0 = () :: Constraint type CTuple1 :: Constraint -> Constraint - type CTuple1 = CSolo :: Constraint -> Constraint + type CTuple1 = CSolo type CTuple10 :: Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint class (a, b, c, d, e, f, g, h, i, j) => CTuple10 a b c d e f g h i j {-# MINIMAL #-} @@ -2637,11 +2637,11 @@ module Data.Tuple.Experimental where type Tuple0 :: * type Tuple0 = () type Tuple0# :: GHC.Types.ZeroBitType - type Tuple0# = (# #) :: GHC.Types.ZeroBitType + type Tuple0# = (# #) type Tuple1 :: * -> * - type Tuple1 = Solo :: * -> * + type Tuple1 = Solo type Tuple1# :: * -> TYPE (GHC.Types.TupleRep '[GHC.Types.LiftedRep]) - type Tuple1# = Solo# :: * -> TYPE (GHC.Types.TupleRep '[GHC.Types.LiftedRep]) + type Tuple1# = Solo# type Tuple10 :: * -> * -> * -> * -> * -> * -> * -> * -> * -> * -> * data Tuple10 a b c d e f g h i j = ... type Tuple10# :: forall (k0 :: GHC.Types.RuntimeRep) (k1 :: GHC.Types.RuntimeRep) (k2 :: GHC.Types.RuntimeRep) (k3 :: GHC.Types.RuntimeRep) (k4 :: GHC.Types.RuntimeRep) (k5 :: GHC.Types.RuntimeRep) (k6 :: GHC.Types.RuntimeRep) (k7 :: GHC.Types.RuntimeRep) (k8 :: GHC.Types.RuntimeRep) (k9 :: GHC.Types.RuntimeRep). TYPE k0 -> TYPE k1 -> TYPE k2 -> TYPE k3 -> TYPE k4 -> TYPE k5 -> TYPE k6 -> TYPE k7 -> TYPE k8 -> TYPE k9 -> TYPE (GHC.Types.TupleRep [k0, k1, k2, k3, k4, k5, k6, k7, k8, k9]) @@ -4345,9 +4345,9 @@ module Prelude.Experimental where class a => CSolo a {-# MINIMAL #-} type CTuple0 :: Constraint - type CTuple0 = () :: Constraint :: Constraint + type CTuple0 = () :: Constraint type CTuple1 :: Constraint -> Constraint - type CTuple1 = CSolo :: Constraint -> Constraint + type CTuple1 = CSolo type CTuple10 :: Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint -> Constraint class (a, b, c, d, e, f, g, h, i, j) => CTuple10 a b c d e f g h i j {-# MINIMAL #-} @@ -6976,11 +6976,11 @@ module Prelude.Experimental where type Tuple0 :: * type Tuple0 = () type Tuple0# :: GHC.Types.ZeroBitType - type Tuple0# = (# #) :: GHC.Types.ZeroBitType + type Tuple0# = (# #) type Tuple1 :: * -> * - type Tuple1 = Solo :: * -> * + type Tuple1 = Solo type Tuple1# :: * -> TYPE (GHC.Types.TupleRep '[GHC.Types.LiftedRep]) - type Tuple1# = Solo# :: * -> TYPE (GHC.Types.TupleRep '[GHC.Types.LiftedRep]) + type Tuple1# = Solo# type Tuple10 :: * -> * -> * -> * -> * -> * -> * -> * -> * -> * -> * data Tuple10 a b c d e f g h i j = ... type Tuple10# :: forall (k0 :: GHC.Types.RuntimeRep) (k1 :: GHC.Types.RuntimeRep) (k2 :: GHC.Types.RuntimeRep) (k3 :: GHC.Types.RuntimeRep) (k4 :: GHC.Types.RuntimeRep) (k5 :: GHC.Types.RuntimeRep) (k6 :: GHC.Types.RuntimeRep) (k7 :: GHC.Types.RuntimeRep) (k8 :: GHC.Types.RuntimeRep) (k9 :: GHC.Types.RuntimeRep). TYPE k0 -> TYPE k1 -> TYPE k2 -> TYPE k3 -> TYPE k4 -> TYPE k5 -> TYPE k6 -> TYPE k7 -> TYPE k8 -> TYPE k9 -> TYPE (GHC.Types.TupleRep [k0, k1, k2, k3, k4, k5, k6, k7, k8, k9]) ===================================== testsuite/tests/interface-stability/template-haskell-exports.stdout ===================================== @@ -31,7 +31,7 @@ module Language.Haskell.TH where type Code :: (* -> *) -> forall (r :: GHC.Types.RuntimeRep). TYPE r -> * newtype Code m a = Code {examineCode :: m (TExp a)} type CodeQ :: forall (r :: GHC.Types.RuntimeRep). TYPE r -> * - type CodeQ = Code Q :: forall (r :: GHC.Types.RuntimeRep). TYPE r -> * + type CodeQ = Code Q type Con :: * data Con = NormalC Name [BangType] | RecC Name [VarBangType] | InfixC BangType Name BangType | ForallC [TyVarBndr Specificity] Cxt Con | GadtC [Name] [BangType] Type | RecGadtC [Name] [VarBangType] Type type ConQ :: * @@ -881,7 +881,7 @@ module Language.Haskell.TH.Lib where type ClauseQ :: * type ClauseQ = GHC.Internal.TH.Syntax.Q GHC.Internal.TH.Syntax.Clause type CodeQ :: forall (r :: GHC.Types.RuntimeRep). TYPE r -> * - type CodeQ = GHC.Internal.TH.Syntax.Code GHC.Internal.TH.Syntax.Q :: forall (r :: GHC.Types.RuntimeRep). TYPE r -> * + type CodeQ = GHC.Internal.TH.Syntax.Code GHC.Internal.TH.Syntax.Q type ConQ :: * type ConQ = GHC.Internal.TH.Syntax.Q GHC.Internal.TH.Syntax.Con type CxtQ :: * @@ -1220,7 +1220,7 @@ module Language.Haskell.TH.Lib.Internal where type ClauseQ :: * type ClauseQ = GHC.Internal.TH.Syntax.Q GHC.Internal.TH.Syntax.Clause type CodeQ :: forall (r :: GHC.Types.RuntimeRep). TYPE r -> * - type CodeQ = GHC.Internal.TH.Syntax.Code GHC.Internal.TH.Syntax.Q :: forall (r :: GHC.Types.RuntimeRep). TYPE r -> * + type CodeQ = GHC.Internal.TH.Syntax.Code GHC.Internal.TH.Syntax.Q type ConQ :: * type ConQ = GHC.Internal.TH.Syntax.Q GHC.Internal.TH.Syntax.Con type CxtQ :: * @@ -2052,8 +2052,8 @@ module Language.Haskell.TH.Syntax where qAddForeignFilePath :: ForeignSrcLang -> GHC.Internal.Base.String -> m () qAddModFinalizer :: Q () -> m () qAddCorePlugin :: GHC.Internal.Base.String -> m () - qGetQ :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => m (GHC.Internal.Maybe.Maybe a) - qPutQ :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => a -> m () + qGetQ :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => m (GHC.Internal.Maybe.Maybe a) + qPutQ :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => a -> m () qIsExtEnabled :: Extension -> m GHC.Types.Bool qExtsEnabled :: m [Extension] qPutDoc :: DocLoc -> GHC.Internal.Base.String -> m () @@ -2135,7 +2135,7 @@ module Language.Haskell.TH.Syntax where falseName :: Name getDoc :: DocLoc -> Q (GHC.Internal.Maybe.Maybe GHC.Internal.Base.String) getPackageRoot :: Q GHC.Internal.IO.FilePath - getQ :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => Q (GHC.Internal.Maybe.Maybe a) + getQ :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => Q (GHC.Internal.Maybe.Maybe a) get_cons_names :: Con -> [Name] hoistCode :: forall (m :: * -> *) (n :: * -> *) (r :: GHC.Types.RuntimeRep) (a :: TYPE r). GHC.Internal.Base.Monad m => (forall x. m x -> n x) -> Code m a -> Code n a isExtEnabled :: Extension -> Q GHC.Types.Bool @@ -2181,7 +2181,7 @@ module Language.Haskell.TH.Syntax where oneName :: Name pkgString :: PkgName -> GHC.Internal.Base.String putDoc :: DocLoc -> GHC.Internal.Base.String -> Q () - putQ :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => a -> Q () + putQ :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => a -> Q () recover :: forall a. Q a -> Q a -> Q a reify :: Name -> Q Info reifyAnnotations :: forall a. GHC.Internal.Data.Data.Data a => AnnLookup -> Q [a] ===================================== testsuite/tests/rep-poly/RepPolyBackpack3.stderr ===================================== @@ -1,19 +1,19 @@ [1 of 3] Processing sig - [1 of 1] Compiling Sig[sig] ( sig\Sig.hsig, nothing ) + [1 of 1] Compiling Sig[sig] ( sig/Sig.hsig, nothing ) [2 of 3] Processing impl Instantiating impl - [1 of 1] Compiling Impl ( impl\Impl.hs, RepPolyBackpack3.out\impl\Impl.o ) + [1 of 1] Compiling Impl ( impl/Impl.hs, RepPolyBackpack3.out/impl/Impl.o ) [3 of 3] Processing main Instantiating main [1 of 1] Including sig[Sig=impl:Impl] Instantiating sig[Sig=impl:Impl] - [1 of 1] Compiling Sig[sig] ( sig\Sig.hsig, RepPolyBackpack3.out\sig\sig-Absk5cIXTXe6UYhGMYGber\Sig.o ) + [1 of 1] Compiling Sig[sig] ( sig/Sig.hsig, RepPolyBackpack3.out/sig/sig-Absk5cIXTXe6UYhGMYGber/Sig.o ) RepPolyBackpack3.bkp:17:5: error: [GHC-15843] • Type constructor ‘Rep’ has conflicting definitions in the module and its hsig file. Main module: type Rep :: GHC.Types.RuntimeRep - type Rep = T :: GHC.Types.RuntimeRep + type Rep = T Hsig file: type Rep :: GHC.Types.RuntimeRep data Rep Illegal implementation of abstract data: Invalid type family ‘T’. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2bbf032989a145ddba354339fca5a1ccf96ae639 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2bbf032989a145ddba354339fca5a1ccf96ae639 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 20:20:07 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 02 Sep 2024 16:20:07 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] store IO actions in the EPS Message-ID: <66d61df728f0_5c9cd4c9d841239b@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 23116a0f by Torsten Schmits at 2024-09-02T22:19:50+02:00 store IO actions in the EPS - - - - - 18 changed files: - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/Iface/Tidy.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Home/ModInfo.hs - compiler/GHC/Unit/Module/ModDetails.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - ghc/GHCi/Leak.hs Changes: ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,8 +49,8 @@ module GHC.Driver.Main , Messager, batchMsg, batchMultiMsg , HscBackendAction (..), HscRecompStatus (..) , initModDetails - , initWholeCoreBindings - , initWholeCoreBindingsEps + , ensureHomeModuleByteCode + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -277,7 +277,7 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.Foldable (fold) +import Data.Functor ((<&>)) import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad @@ -296,7 +296,6 @@ import System.IO import {-# SOURCE #-} GHC.Driver.Pipeline import Data.Time -import System.IO.Unsafe ( unsafeInterleaveIO ) import GHC.Iface.Env ( trace_if ) import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) @@ -868,7 +867,7 @@ hscRecompStatus -- Do need linkable -- 1. Just check whether we have bytecode/object linkables and then -- we will decide if we need them or not. - bc_linkable <- checkByteCode checked_iface mod_summary (homeMod_bytecode old_linkable) + let bc_linkable = checkByteCode checked_iface mod_summary (homeMod_bytecode old_linkable) obj_linkable <- liftIO $ checkObjects lcl_dflags (homeMod_object old_linkable) mod_summary trace_if (hsc_logger hsc_env) (vcat [text "BCO linkable", nest 2 (ppr bc_linkable), text "Object Linkable", ppr obj_linkable]) @@ -955,25 +954,21 @@ checkObjects dflags mb_old_linkable summary = do -- | Check to see if we can reuse the old linkable, by this point we will -- have just checked that the old interface matches up with the source hash, so -- no need to check that again here -checkByteCode :: ModIface -> ModSummary -> Maybe Linkable -> IO (MaybeValidated Linkable) -checkByteCode iface mod_sum mb_old_linkable = - case mb_old_linkable of - Just old_linkable - | not (linkableIsNativeCodeOnly old_linkable) - -> return $ (UpToDateItem old_linkable) - _ -> loadByteCode iface mod_sum - -loadByteCode :: ModIface -> ModSummary -> IO (MaybeValidated Linkable) -loadByteCode iface mod_sum = do - let - this_mod = ms_mod mod_sum - if_date = fromJust $ ms_iface_date mod_sum - case mi_extra_decls iface of - Just extra_decls -> do - let fi = WholeCoreBindings extra_decls this_mod (ms_location mod_sum) - (mi_foreign iface) - return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) - _ -> return $ outOfDateItemBecause MissingBytecode Nothing +checkByteCode :: + ModIface -> + ModSummary -> + HomeModByteCode -> + MaybeValidated HomeModByteCode +checkByteCode iface mod_sum = \case + NoHomeModByteCode -> HomeModIfaceCore <$> loadByteCode iface mod_sum + old_bytecode -> UpToDateItem old_bytecode + +loadByteCode :: ModIface -> ModSummary -> MaybeValidated WholeCoreBindings +loadByteCode iface mod_sum = + case iface_core_bindings iface (ms_location mod_sum) of + Just wcb -> UpToDateItem wcb + Nothing -> outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- @@ -986,79 +981,121 @@ add_iface_to_hpt iface details = -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] -initModDetails :: HscEnv -> ModIface -> IO ModDetails -initModDetails hsc_env iface = +initModDetails :: + HscEnv -> + ModIface -> + ModLocation -> + HomeModByteCode -> + IO ModDetails +initModDetails hsc_env iface location bytecode = fixIO $ \details' -> do let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful -- in make mode, since this HMI will go into the HPT. - genModDetails hsc_env' iface + details <- genModDetails hsc_env' iface + let md_bytecode = + ensureHomeModuleByteCode hsc_env' iface location details bytecode + pure details {md_bytecode} + +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface --- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects, --- using the supplied environment for type checking. --- --- The laziness is necessary because this value is stored purely in a --- 'HomeModLinkable' in the home package table, rather than some dedicated --- mutable state that would generate bytecode on demand, so we have to call this --- function even when we don't know that we'll need the bytecode. --- --- In addition, the laziness has to be hidden inside 'LazyBCOs' because --- 'Linkable' is used too generally, so that looking at the constructor to --- decide whether to discard it when linking native code would force the thunk --- otherwise, incurring a significant performance penalty. --- --- This is sound because generateByteCode just depends on things already loaded --- in the interface file. -initWcbWithTcEnv :: +-- | Hydrate core bindings for a module in the home package table, for which we +-- can obtain a 'ModDetails' with a type env. +ensureHomeModuleByteCode :: HscEnv -> + ModIface -> + ModLocation -> + ModDetails -> + HomeModByteCode -> + Maybe (IO Linkable) +ensureHomeModuleByteCode hsc_env iface location details = \case + NoHomeModByteCode -> + loadIfaceByteCode hsc_env iface location type_env + HomeModIfaceCore wcb -> + Just (initWholeCoreBindings hsc_env type_env wcb) + HomeModByteCode bc -> + Just (pure bc) + where + type_env = md_types details + +-- | Hydrate Core bindings if the interface contains any, using the supplied +-- type env for typechecking. +loadIfaceByteCode :: HscEnv -> + ModIface -> + ModLocation -> TypeEnv -> - Linkable -> - IO Linkable -initWcbWithTcEnv tc_hsc_env hsc_env type_env (Linkable utc_time this_mod uls) = - Linkable utc_time this_mod <$> mapM go uls + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + initWholeCoreBindings hsc_env type_env <$> iface_core_bindings iface location + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +-- +-- 4. Wrapping the build products in 'Linkable' with the proper modification +-- time obtained from the interface. +initWholeCoreBindings :: HscEnv -> TypeEnv -> WholeCoreBindings -> IO Linkable +initWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + parts <- gen_bytecode core_binds stubs foreign_files + linkable parts where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef type_env - let - tc_hsc_env_with_kv = tc_hsc_env { - hsc_type_env_vars = - knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") tc_hsc_env_with_kv $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons type_env) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + (bcos, fos) <- generateByteCode hsc_env cgi_guts wcb_mod_location + pure $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file wcb_mod_location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time wcb_module parts + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env --- | Hydrate core bindings for a module in the home package table, for which we --- can obtain a 'ModDetails'. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env iface details = - initWcbWithTcEnv (add_iface_to_hpt iface details hsc_env) hsc_env (md_types details) - --- | Hydrate core bindings for a module in the external package state. --- This is used for home modules as well when compiling in oneshot mode. -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable -initWholeCoreBindingsEps hsc_env iface lnk = do - eps <- hscEPS hsc_env - let type_env = fold (lookupModuleEnv (eps_PTT eps) (mi_module iface)) - initWcbWithTcEnv hsc_env hsc_env type_env lnk - - {- Note [ModDetails and --make mode] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -1,8 +1,15 @@ module GHC.Driver.Main where -import GHC.Driver.Env -import GHC.Linker.Types -import GHC.Prelude -import GHC.Unit.Module.ModIface +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1319,11 +1319,12 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- | Add the entries from a BCO linkable to the SPT table, see -- See Note [Grand plan for static forms] in GHC.Iface.Tidy.StaticPtrTable. -addSptEntries :: HscEnv -> Maybe Linkable -> IO () +addSptEntries :: HscEnv -> HomeModByteCode -> IO () addSptEntries hsc_env mlinkable = hscAddSptEntries hsc_env [ spt - | linkable <- maybeToList mlinkable + -- TODO + | HomeModByteCode linkable <- [mlinkable] , bco <- linkableBCOs linkable , spt <- bc_spt_entries bco ] ===================================== compiler/GHC/Driver/Pipeline.hs ===================================== @@ -124,6 +124,7 @@ import Data.List.NonEmpty (NonEmpty(..)) import Data.Time ( getCurrentTime ) import GHC.Iface.Recomp import GHC.Types.Unique.DSet +import GHC.Unit.Module.ModDetails (ModDetails(..)) -- Simpler type synonym for actions in the pipeline monad type P m = TPipelineClass TPhase m @@ -247,9 +248,9 @@ compileOne' mHscMessage let pipeline = hscPipeline pipe_env (setDumpPrefix pipe_env plugin_hsc_env, upd_summary, status) (iface, linkable) <- runPipeline (hsc_hooks plugin_hsc_env) pipeline -- See Note [ModDetails and --make mode] - details <- initModDetails plugin_hsc_env iface - linkable' <- traverse (initWholeCoreBindings plugin_hsc_env iface details) (homeMod_bytecode linkable) - return $! HomeModInfo iface details (linkable { homeMod_bytecode = linkable' }) + details <- initModDetails plugin_hsc_env iface (ms_location summary) + (homeMod_bytecode linkable) + return $! HomeModInfo iface details linkable where lcl_dflags = ms_hspp_opts summary location = ms_location summary ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -602,7 +602,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do if gopt Opt_ByteCodeAndObjectCode dflags then do bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return $ emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } + return $ emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } else return emptyHomeModInfoLinkable @@ -619,7 +619,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do final_iface <- mkFullIface hsc_env partial_iface Nothing Nothing NoStubs [] hscMaybeWriteIface logger dflags True final_iface mb_old_iface_hash location bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } , panic "interpreter") + return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } , panic "interpreter") runUnlitPhase :: HscEnv -> FilePath -> FilePath -> IO FilePath ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,6 +506,7 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) @@ -518,13 +520,34 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") + -- TODO can't do that because we use it for + -- fingerprinting. + -- & set_mi_foreign (panic "No mi_foreign in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + -- TODO in @getLinkDeps@ we fall back to bytecode when the HMI + -- doesn't have object code, even if the flag is not given – + -- what's the rule? Should we provide it unconditionally if it + -- exists? + | prefer_bytecode + , Just action <- loadIfaceByteCode hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have @extra_decls@ + -- so @getLinkDeps@ knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,8 +559,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, - eps_PTT = - extendModuleEnv (eps_PTT eps) mod (mkNameEnv new_eps_decls), + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -700,7 +722,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -847,7 +869,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -877,7 +899,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -902,7 +924,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/Iface/Tidy.hs ===================================== @@ -193,6 +193,7 @@ mkBootModDetailsTc logger , md_anns = [] , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing }) where -- Find the LocalIds in the type env that are exported @@ -492,6 +493,7 @@ tidyProgram opts (ModGuts { mg_module = mod , md_exports = exports , md_anns = anns -- are already tidy , md_complete_matches = complete_matches + , md_bytecode = Nothing } ) ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -254,6 +254,7 @@ typecheckIface iface , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } } @@ -470,6 +471,7 @@ typecheckIfacesForMerging mod ifaces tc_env_vars = , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } return (global_type_env, details) @@ -512,6 +514,7 @@ typecheckIfaceForInstantiate nsubst iface , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } -- Note [Resolving never-exported Names] ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -35,7 +35,6 @@ import GHC.Unit.Env import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.Module.ModIface -import GHC.Unit.Module.WholeCoreBindings import GHC.Unit.Module.Deps import GHC.Unit.Module.Graph import GHC.Unit.Home.ModInfo @@ -56,26 +55,20 @@ import Data.List (isSuffixOf) import System.FilePath import System.Directory -import GHC.Driver.Env -import {-# SOURCE #-} GHC.Driver.Main -import Data.Time.Clock -import GHC.Driver.Flags -import GHC.Driver.Session data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function - , ldHscEnv :: !HscEnv + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -266,54 +259,40 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- See Note [Using Byte Code rather than Object Code for Template Haskell] - homeModLinkable :: HomeModInfo -> Maybe Linkable + homeModLinkable :: HomeModInfo -> Maybe (IO Linkable) homeModLinkable hmi = - if ldUseByteCode opts - then homeModInfoByteCode hmi <|> homeModInfoObject hmi - else homeModInfoObject hmi <|> homeModInfoByteCode hmi + let obj = pure <$> homeModInfoObject hmi + bc = evalHomeModByteCode hmi + in if ldUseByteCode opts + then bc <|> obj + else obj <|> bc get_linkable osuf mod -- A home-package module | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) - = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) + = do + lnk <- expectJust "getLinkDeps" (homeModLinkable mod_info) + adjust_linkable lnk | otherwise = do -- It's not in the HPT because we are in one shot mode, -- so use the Finder to get a ModLocation... case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod - | prefer_bytecode = do - Succeeded iface <- ldLoadIface opts (text "load core bindings") mod - case mi_extra_decls iface of - Just extra_decls -> do - t <- getCurrentTime - let - stubs = mi_foreign iface - wcb = WholeCoreBindings extra_decls mod loc stubs - linkable = Linkable t mod (pure (CoreBindings wcb)) - initWholeCoreBindingsEps hsc_env iface linkable - _ -> fallback_no_bytecode loc mod - | otherwise = fallback_no_bytecode loc mod - - fallback_no_bytecode loc mod = do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - - prefer_bytecode = gopt Opt_UseBytecodeRatherThanObjects dflags - - dflags = hsc_dflags hsc_env - - hsc_env = ldHscEnv opts + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do @@ -338,9 +317,6 @@ get_link_deps opts pls maybe_normal_osuf span mods = do DotA fp -> panic ("adjust_ul DotA " ++ show fp) DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,19 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags - , ldHscEnv = hsc_env + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Linker/Types.hs ===================================== @@ -65,7 +65,6 @@ import Data.Time ( UTCTime ) import GHC.Unit.Module.Env import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM -import GHC.Unit.Module.WholeCoreBindings import Data.Maybe (mapMaybe) import Data.List.NonEmpty (NonEmpty, nonEmpty) import qualified Data.List.NonEmpty as NE @@ -284,18 +283,6 @@ data LinkablePart | DotDLL FilePath -- ^ Dynamically linked library file (.so, .dll, .dylib) - | CoreBindings WholeCoreBindings - -- ^ Serialised core which we can turn into BCOs (or object files), or - -- used by some other backend See Note [Interface Files with Core - -- Definitions] - - | LazyBCOs - CompiledByteCode - -- ^ Some BCOs generated on-demand when forced. This is used for - -- WholeCoreBindings, see Note [Interface Files with Core Definitions] - [FilePath] - -- ^ Objects containing foreign stubs and files - | BCOs CompiledByteCode -- ^ A byte-code object, lives only in memory. @@ -308,8 +295,6 @@ instance Outputable LinkablePart where ppr (DotA path) = text "DotA" <+> text path ppr (DotDLL path) = text "DotDLL" <+> text path ppr (BCOs bco) = text "BCOs" <+> ppr bco - ppr (LazyBCOs{}) = text "LazyBCOs" - ppr (CoreBindings {}) = text "CoreBindings" -- | Return true if the linkable only consists of native code (no BCO) linkableIsNativeCodeOnly :: Linkable -> Bool @@ -350,8 +335,6 @@ isNativeCode = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Is the part a native library? (.so/.dll) isNativeLib :: LinkablePart -> Bool @@ -360,8 +343,6 @@ isNativeLib = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Get the FilePath of linkable part (if applicable) linkablePartPath :: LinkablePart -> Maybe FilePath @@ -369,8 +350,6 @@ linkablePartPath = \case DotO fn _ -> Just fn DotA fn -> Just fn DotDLL fn -> Just fn - CoreBindings {} -> Nothing - LazyBCOs {} -> Nothing BCOs {} -> Nothing -- | Return the paths of all object code files (.o, .a, .so) contained in this @@ -380,8 +359,6 @@ linkablePartNativePaths = \case DotO fn _ -> [fn] DotA fn -> [fn] DotDLL fn -> [fn] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Return the paths of all object files (.o) contained in this 'LinkablePart'. @@ -390,8 +367,6 @@ linkablePartObjectPaths = \case DotO fn _ -> [fn] DotA _ -> [] DotDLL _ -> [] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Retrieve the compiled byte-code from the linkable part. @@ -400,7 +375,6 @@ linkablePartObjectPaths = \case linkablePartAllBCOs :: LinkablePart -> [CompiledByteCode] linkablePartAllBCOs = \case BCOs bco -> [bco] - LazyBCOs bcos _ -> [bcos] _ -> [] linkableFilter :: (LinkablePart -> [LinkablePart]) -> Linkable -> Maybe Linkable @@ -413,13 +387,11 @@ linkablePartNative = \case u at DotO {} -> [u] u at DotA {} -> [u] u at DotDLL {} -> [u] - LazyBCOs _ os -> [DotO f ForeignObject | f <- os] _ -> [] linkablePartByteCode :: LinkablePart -> [LinkablePart] linkablePartByteCode = \case u at BCOs {} -> [u] - LazyBCOs bcos _ -> [BCOs bcos] _ -> [] -- | Transform the 'LinkablePart' list in this 'Linkable' to contain only ===================================== compiler/GHC/Runtime/Eval.hs ===================================== @@ -1275,14 +1275,14 @@ showModule mod_summary = let interpreted = case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> isJust (homeModInfoByteCode mod_info) && isNothing (homeModInfoObject mod_info) + Just mod_info -> homeModInfoHasByteCode mod_info && isNothing (homeModInfoObject mod_info) return (showSDoc dflags $ showModMsg dflags interpreted (ModuleNode [] mod_summary)) moduleIsBootOrNotObjectLinkable :: GhcMonad m => ModSummary -> m Bool moduleIsBootOrNotObjectLinkable mod_summary = withSession $ \hsc_env -> case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> return . isNothing $ homeModInfoByteCode mod_info + Just mod_info -> return . not $ homeModInfoHasByteCode mod_info ---------------------------------------------------------------------------- -- RTTI primitives ===================================== compiler/GHC/Runtime/Interpreter.hs ===================================== @@ -113,6 +113,7 @@ import qualified GHC.Exts.Heap as Heap import GHC.Stack.CCS (CostCentre,CostCentreStack) import System.Directory import System.Process +import GHC.Unit.Module.ModDetails (ModDetails(..)) {- Note [Remote GHCi] ~~~~~~~~~~~~~~~~~~ @@ -684,10 +685,13 @@ fromEvalResult (EvalSuccess a) = return a getModBreaks :: HomeModInfo -> ModBreaks getModBreaks hmi - | Just linkable <- homeModInfoByteCode hmi, + | HomeModByteCode linkable <- homeModInfoByteCode hmi, -- The linkable may have 'DotO's as well; only consider BCOs. See #20570. [cbc] <- linkableBCOs linkable = fromMaybe emptyModBreaks (bc_breaks cbc) + | ModDetails {md_bytecode = Just _} <- hm_details hmi + -- TODO + = emptyModBreaks | otherwise = emptyModBreaks -- probably object code ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -45,8 +47,6 @@ type PackageCompleteMatches = CompleteMatches type PackageIfaceTable = ModuleEnv ModIface -- Domain = modules in the imported packages -type PackageTypeTable = ModuleEnv TypeEnv - -- | Constructs an empty PackageIfaceTable emptyPackageIfaceTable :: PackageIfaceTable emptyPackageIfaceTable = emptyModuleEnv @@ -70,7 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv - , eps_PTT = emptyModuleEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -142,7 +142,11 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules - eps_PTT :: !PackageTypeTable, + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules ===================================== compiler/GHC/Unit/Home/ModInfo.hs ===================================== @@ -1,13 +1,19 @@ +{-# LANGUAGE LambdaCase #-} + -- | Info about modules in the "home" unit module GHC.Unit.Home.ModInfo ( HomeModInfo (..) - , HomeModLinkable(..) + , HomeModLinkable (..) + , HomeModByteCode (..) , homeModInfoObject , homeModInfoByteCode + , homeModInfoHasByteCode , emptyHomeModInfoLinkable , justBytecode , justObjects , bytecodeAndObjects + , pureHomeModByteCode + , evalHomeModByteCode , HomePackageTable , emptyHomePackageTable , lookupHpt @@ -44,6 +50,7 @@ import GHC.Utils.Outputable import Data.List (sortOn) import Data.Ord import GHC.Utils.Panic +import GHC.Unit.Module.WholeCoreBindings (WholeCoreBindings) -- | Information about modules in the package being compiled data HomeModInfo = HomeModInfo @@ -73,37 +80,72 @@ data HomeModInfo = HomeModInfo -- 'ModIface' (only). } -homeModInfoByteCode :: HomeModInfo -> Maybe Linkable +homeModInfoByteCode :: HomeModInfo -> HomeModByteCode homeModInfoByteCode = homeMod_bytecode . hm_linkable +homeModInfoHasByteCode :: HomeModInfo -> Bool +homeModInfoHasByteCode hmi = case homeModInfoByteCode hmi of + NoHomeModByteCode -> False + _ -> True + homeModInfoObject :: HomeModInfo -> Maybe Linkable homeModInfoObject = homeMod_object . hm_linkable emptyHomeModInfoLinkable :: HomeModLinkable -emptyHomeModInfoLinkable = HomeModLinkable Nothing Nothing +emptyHomeModInfoLinkable = HomeModLinkable NoHomeModByteCode Nothing -- See Note [Home module build products] -data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !(Maybe Linkable) +data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !HomeModByteCode , homeMod_object :: !(Maybe Linkable) } instance Outputable HomeModLinkable where ppr (HomeModLinkable l1 l2) = ppr l1 $$ ppr l2 -justBytecode :: Linkable -> HomeModLinkable -justBytecode lm = - assertPpr (not (linkableIsNativeCodeOnly lm)) (ppr lm) - $ emptyHomeModInfoLinkable { homeMod_bytecode = Just lm } +justBytecode :: HomeModByteCode -> HomeModLinkable +justBytecode bc = + emptyHomeModInfoLinkable { homeMod_bytecode = bc } justObjects :: Linkable -> HomeModLinkable justObjects lm = assertPpr (linkableIsNativeCodeOnly lm) (ppr lm) $ emptyHomeModInfoLinkable { homeMod_object = Just lm } -bytecodeAndObjects :: Linkable -> Linkable -> HomeModLinkable +bytecodeAndObjects :: HomeModByteCode -> Linkable -> HomeModLinkable bytecodeAndObjects bc o = - assertPpr (not (linkableIsNativeCodeOnly bc) && linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) - (HomeModLinkable (Just bc) (Just o)) - + assertPpr (linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) + (HomeModLinkable bc (Just o)) + +pureHomeModByteCode :: HomeModByteCode -> Maybe Linkable +pureHomeModByteCode = \case + NoHomeModByteCode -> Nothing + HomeModIfaceCore _ -> Nothing + HomeModByteCode l -> Just l + +-- | Obtain the bytecode stored in this 'HomeModInfo', preferring the value in +-- 'HomeModLinkable' that's already in memory before evaluating the 'IO' in +-- 'ModDetails' that hydrates and parses Core loaded from an interface. +-- +-- This should only be called once in the module's lifecycle; afterwards, the +-- bytecode is cached in 'LoaderState'. +evalHomeModByteCode :: HomeModInfo -> Maybe (IO Linkable) +evalHomeModByteCode HomeModInfo {hm_details, hm_linkable} + | HomeModByteCode bc <- homeMod_bytecode hm_linkable + = Just (pure bc) + | otherwise + = md_bytecode hm_details + +data HomeModByteCode = + NoHomeModByteCode + | + HomeModIfaceCore WholeCoreBindings + | + HomeModByteCode Linkable + +instance Outputable HomeModByteCode where + ppr = \case + NoHomeModByteCode -> text "no bytecode" + HomeModIfaceCore _ -> text "dehydrated Core" + HomeModByteCode linkable -> ppr linkable {- Note [Home module build products] ===================================== compiler/GHC/Unit/Module/ModDetails.hs ===================================== @@ -14,6 +14,9 @@ import GHC.Types.DefaultEnv ( DefaultEnv, emptyDefaultEnv ) import GHC.Types.TypeEnv import GHC.Types.Annotations ( Annotation ) +import GHC.Linker.Types (Linkable) +import GHC.Prelude + -- | The 'ModDetails' is essentially a cache for information in the 'ModIface' -- for home modules only. Information relating to packages will be loaded into -- global environments in 'ExternalPackageState'. @@ -40,6 +43,8 @@ data ModDetails = ModDetails , md_complete_matches :: CompleteMatches -- ^ Complete match pragmas for this module + + , md_bytecode :: !(Maybe (IO Linkable)) } -- | Constructs an empty ModDetails @@ -53,4 +58,5 @@ emptyModDetails = ModDetails , md_fam_insts = [] , md_anns = [] , md_complete_matches = [] + , md_bytecode = Nothing } ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -27,6 +27,8 @@ import Data.Maybe (fromMaybe) import System.FilePath (takeExtension) {- +TODO update + Note [Interface Files with Core Definitions] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -87,7 +89,12 @@ data WholeCoreBindings = WholeCoreBindings , wcb_foreign :: IfaceForeign } +instance Outputable WholeCoreBindings where + ppr WholeCoreBindings {wcb_module} = text "iface Core for " <+> ppr wcb_module + {- +TODO update + Note [Foreign stubs and TH bytecode linking] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== ghc/GHCi/Leak.hs ===================================== @@ -50,7 +50,7 @@ getLeakIndicators hsc_env = where mkWeakLinkables :: HomeModLinkable -> IO [Maybe (Weak Linkable)] mkWeakLinkables (HomeModLinkable mbc mo) = - mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [mbc, mo] + mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [pureHomeModByteCode mbc, mo] -- | Look at the LeakIndicators collected by an earlier call to -- `getLeakIndicators`, and print messasges if any of them are still View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/23116a0f4cca366467cff515d488258fdeb8def9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/23116a0f4cca366467cff515d488258fdeb8def9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 20:37:29 2024 From: gitlab at gitlab.haskell.org (Adriaan Leijnse (@aidylns)) Date: Mon, 02 Sep 2024 16:37:29 -0400 Subject: [Git][ghc/ghc][wip/aidylns/ttg-remove-hsunboundvar-via-hshole] 11 commits: JS: add basic support for POSIX *at functions (#25190) Message-ID: <66d6220962c2_2279c91ce1fc13120@gitlab.mail> Adriaan Leijnse pushed to branch wip/aidylns/ttg-remove-hsunboundvar-via-hshole at Glasgow Haskell Compiler / GHC Commits: 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 44c3364c by Adriaan Leijnse at 2024-08-31T23:48:17+02:00 Merge remote-tracking branch 'origin/master' into wip/aidylns/ttg-remove-hsunboundvar-via-hshole-epann - - - - - b2b11063 by Adriaan Leijnse at 2024-09-01T16:22:06+02:00 Use EpAnnHole instead of SrcSpanAnnN - - - - - 3bbd76d6 by Adriaan Leijnse at 2024-09-02T22:36:37+02:00 Remove [Holes] move [Holes in expressions]. - - - - - 30 changed files: - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Hs/Expr.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/Tc/Types/Constraint.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Zonk/Type.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - compiler/Language/Haskell/Syntax/Expr.hs - configure.ac - docs/users_guide/9.12.1-notes.rst - hadrian/src/Packages.hs - hadrian/src/Rules/ToolArgs.hs - hadrian/src/Settings/Default.hs - libraries/directory - + libraries/file-io - libraries/ghc-internal/jsbits/base.js - libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs - rts/linker/PEi386.c - + testsuite/tests/default/T25206.hs - + testsuite/tests/default/T25206.stderr - + testsuite/tests/default/T25206_helper.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a94f1ab74fc3c72d39ce90ac036483a983370945...3bbd76d6d2a946f8b5ec276e9930dbee30dc1988 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a94f1ab74fc3c72d39ce90ac036483a983370945...3bbd76d6d2a946f8b5ec276e9930dbee30dc1988 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 2 20:44:45 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 02 Sep 2024 16:44:45 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] store IO actions in the EPS Message-ID: <66d623bdd0189_2279c930e1fc172ca@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 8baf8f1f by Torsten Schmits at 2024-09-02T22:44:28+02:00 store IO actions in the EPS - - - - - 19 changed files: - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/Iface/Tidy.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Home/ModInfo.hs - compiler/GHC/Unit/Module/ModDetails.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - ghc/GHCi/Leak.hs - testsuite/tests/driver/j-space/jspace.hs Changes: ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,8 +49,7 @@ module GHC.Driver.Main , Messager, batchMsg, batchMultiMsg , HscBackendAction (..), HscRecompStatus (..) , initModDetails - , initWholeCoreBindings - , initWholeCoreBindingsEps + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -277,7 +276,7 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.Foldable (fold) +import Data.Functor ((<&>)) import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad @@ -296,7 +295,6 @@ import System.IO import {-# SOURCE #-} GHC.Driver.Pipeline import Data.Time -import System.IO.Unsafe ( unsafeInterleaveIO ) import GHC.Iface.Env ( trace_if ) import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) @@ -868,7 +866,7 @@ hscRecompStatus -- Do need linkable -- 1. Just check whether we have bytecode/object linkables and then -- we will decide if we need them or not. - bc_linkable <- checkByteCode checked_iface mod_summary (homeMod_bytecode old_linkable) + let bc_linkable = checkByteCode checked_iface mod_summary (homeMod_bytecode old_linkable) obj_linkable <- liftIO $ checkObjects lcl_dflags (homeMod_object old_linkable) mod_summary trace_if (hsc_logger hsc_env) (vcat [text "BCO linkable", nest 2 (ppr bc_linkable), text "Object Linkable", ppr obj_linkable]) @@ -955,25 +953,21 @@ checkObjects dflags mb_old_linkable summary = do -- | Check to see if we can reuse the old linkable, by this point we will -- have just checked that the old interface matches up with the source hash, so -- no need to check that again here -checkByteCode :: ModIface -> ModSummary -> Maybe Linkable -> IO (MaybeValidated Linkable) -checkByteCode iface mod_sum mb_old_linkable = - case mb_old_linkable of - Just old_linkable - | not (linkableIsNativeCodeOnly old_linkable) - -> return $ (UpToDateItem old_linkable) - _ -> loadByteCode iface mod_sum - -loadByteCode :: ModIface -> ModSummary -> IO (MaybeValidated Linkable) -loadByteCode iface mod_sum = do - let - this_mod = ms_mod mod_sum - if_date = fromJust $ ms_iface_date mod_sum - case mi_extra_decls iface of - Just extra_decls -> do - let fi = WholeCoreBindings extra_decls this_mod (ms_location mod_sum) - (mi_foreign iface) - return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) - _ -> return $ outOfDateItemBecause MissingBytecode Nothing +checkByteCode :: + ModIface -> + ModSummary -> + HomeModByteCode -> + MaybeValidated HomeModByteCode +checkByteCode iface mod_sum = \case + NoHomeModByteCode -> HomeModIfaceCore <$> loadByteCode iface mod_sum + old_bytecode -> UpToDateItem old_bytecode + +loadByteCode :: ModIface -> ModSummary -> MaybeValidated WholeCoreBindings +loadByteCode iface mod_sum = + case iface_core_bindings iface (ms_location mod_sum) of + Just wcb -> UpToDateItem wcb + Nothing -> outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- @@ -986,79 +980,122 @@ add_iface_to_hpt iface details = -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] -initModDetails :: HscEnv -> ModIface -> IO ModDetails -initModDetails hsc_env iface = +initModDetails :: + HscEnv -> + ModIface -> + ModLocation -> + HomeModByteCode -> + IO ModDetails +initModDetails hsc_env iface location bytecode = fixIO $ \details' -> do let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful -- in make mode, since this HMI will go into the HPT. - genModDetails hsc_env' iface + details <- genModDetails hsc_env' iface + pure details { + md_bytecode = + ensureHomeModuleByteCode hsc_env' iface location details' bytecode + } + +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface --- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects, --- using the supplied environment for type checking. --- --- The laziness is necessary because this value is stored purely in a --- 'HomeModLinkable' in the home package table, rather than some dedicated --- mutable state that would generate bytecode on demand, so we have to call this --- function even when we don't know that we'll need the bytecode. --- --- In addition, the laziness has to be hidden inside 'LazyBCOs' because --- 'Linkable' is used too generally, so that looking at the constructor to --- decide whether to discard it when linking native code would force the thunk --- otherwise, incurring a significant performance penalty. --- --- This is sound because generateByteCode just depends on things already loaded --- in the interface file. -initWcbWithTcEnv :: +-- | Hydrate core bindings for a module in the home package table, for which we +-- can obtain a 'ModDetails' with a type env. +ensureHomeModuleByteCode :: HscEnv -> + ModIface -> + ModLocation -> + ModDetails -> + HomeModByteCode -> + Maybe (IO Linkable) +ensureHomeModuleByteCode hsc_env iface location details = \case + NoHomeModByteCode -> + loadIfaceByteCode hsc_env iface location type_env + HomeModIfaceCore wcb -> + Just (initWholeCoreBindings hsc_env type_env wcb) + HomeModByteCode bc -> + Just (pure bc) + where + type_env = md_types details + +-- | Hydrate Core bindings if the interface contains any, using the supplied +-- type env for typechecking. +loadIfaceByteCode :: HscEnv -> + ModIface -> + ModLocation -> TypeEnv -> - Linkable -> - IO Linkable -initWcbWithTcEnv tc_hsc_env hsc_env type_env (Linkable utc_time this_mod uls) = - Linkable utc_time this_mod <$> mapM go uls + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + initWholeCoreBindings hsc_env type_env <$> iface_core_bindings iface location + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +-- +-- 4. Wrapping the build products in 'Linkable' with the proper modification +-- time obtained from the interface. +initWholeCoreBindings :: HscEnv -> TypeEnv -> WholeCoreBindings -> IO Linkable +initWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + parts <- gen_bytecode core_binds stubs foreign_files + linkable parts where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef type_env - let - tc_hsc_env_with_kv = tc_hsc_env { - hsc_type_env_vars = - knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") tc_hsc_env_with_kv $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons type_env) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + (bcos, fos) <- generateByteCode hsc_env cgi_guts wcb_mod_location + pure $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file wcb_mod_location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time wcb_module parts + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env --- | Hydrate core bindings for a module in the home package table, for which we --- can obtain a 'ModDetails'. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env iface details = - initWcbWithTcEnv (add_iface_to_hpt iface details hsc_env) hsc_env (md_types details) - --- | Hydrate core bindings for a module in the external package state. --- This is used for home modules as well when compiling in oneshot mode. -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable -initWholeCoreBindingsEps hsc_env iface lnk = do - eps <- hscEPS hsc_env - let type_env = fold (lookupModuleEnv (eps_PTT eps) (mi_module iface)) - initWcbWithTcEnv hsc_env hsc_env type_env lnk - - {- Note [ModDetails and --make mode] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -1,8 +1,15 @@ module GHC.Driver.Main where -import GHC.Driver.Env -import GHC.Linker.Types -import GHC.Prelude -import GHC.Unit.Module.ModIface +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1319,11 +1319,12 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- | Add the entries from a BCO linkable to the SPT table, see -- See Note [Grand plan for static forms] in GHC.Iface.Tidy.StaticPtrTable. -addSptEntries :: HscEnv -> Maybe Linkable -> IO () +addSptEntries :: HscEnv -> HomeModByteCode -> IO () addSptEntries hsc_env mlinkable = hscAddSptEntries hsc_env [ spt - | linkable <- maybeToList mlinkable + -- TODO + | HomeModByteCode linkable <- [mlinkable] , bco <- linkableBCOs linkable , spt <- bc_spt_entries bco ] ===================================== compiler/GHC/Driver/Pipeline.hs ===================================== @@ -247,9 +247,9 @@ compileOne' mHscMessage let pipeline = hscPipeline pipe_env (setDumpPrefix pipe_env plugin_hsc_env, upd_summary, status) (iface, linkable) <- runPipeline (hsc_hooks plugin_hsc_env) pipeline -- See Note [ModDetails and --make mode] - details <- initModDetails plugin_hsc_env iface - linkable' <- traverse (initWholeCoreBindings plugin_hsc_env iface details) (homeMod_bytecode linkable) - return $! HomeModInfo iface details (linkable { homeMod_bytecode = linkable' }) + details <- initModDetails plugin_hsc_env iface (ms_location summary) + (homeMod_bytecode linkable) + return $! HomeModInfo iface details linkable where lcl_dflags = ms_hspp_opts summary location = ms_location summary ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -602,7 +602,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do if gopt Opt_ByteCodeAndObjectCode dflags then do bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return $ emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } + return $ emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } else return emptyHomeModInfoLinkable @@ -619,7 +619,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do final_iface <- mkFullIface hsc_env partial_iface Nothing Nothing NoStubs [] hscMaybeWriteIface logger dflags True final_iface mb_old_iface_hash location bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } , panic "interpreter") + return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } , panic "interpreter") runUnlitPhase :: HscEnv -> FilePath -> FilePath -> IO FilePath ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,6 +506,7 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) @@ -518,13 +520,34 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") + -- TODO can't do that because we use it for + -- fingerprinting. + -- & set_mi_foreign (panic "No mi_foreign in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + -- TODO in @getLinkDeps@ we fall back to bytecode when the HMI + -- doesn't have object code, even if the flag is not given – + -- what's the rule? Should we provide it unconditionally if it + -- exists? + | prefer_bytecode + , Just action <- loadIfaceByteCode hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have @extra_decls@ + -- so @getLinkDeps@ knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,8 +559,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, - eps_PTT = - extendModuleEnv (eps_PTT eps) mod (mkNameEnv new_eps_decls), + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -700,7 +722,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -847,7 +869,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -877,7 +899,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -902,7 +924,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/Iface/Tidy.hs ===================================== @@ -193,6 +193,7 @@ mkBootModDetailsTc logger , md_anns = [] , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing }) where -- Find the LocalIds in the type env that are exported @@ -492,6 +493,7 @@ tidyProgram opts (ModGuts { mg_module = mod , md_exports = exports , md_anns = anns -- are already tidy , md_complete_matches = complete_matches + , md_bytecode = Nothing } ) ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -254,6 +254,7 @@ typecheckIface iface , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } } @@ -470,6 +471,7 @@ typecheckIfacesForMerging mod ifaces tc_env_vars = , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } return (global_type_env, details) @@ -512,6 +514,7 @@ typecheckIfaceForInstantiate nsubst iface , md_anns = anns , md_exports = exports , md_complete_matches = complete_matches + , md_bytecode = Nothing } -- Note [Resolving never-exported Names] ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -35,7 +35,6 @@ import GHC.Unit.Env import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.Module.ModIface -import GHC.Unit.Module.WholeCoreBindings import GHC.Unit.Module.Deps import GHC.Unit.Module.Graph import GHC.Unit.Home.ModInfo @@ -56,26 +55,20 @@ import Data.List (isSuffixOf) import System.FilePath import System.Directory -import GHC.Driver.Env -import {-# SOURCE #-} GHC.Driver.Main -import Data.Time.Clock -import GHC.Driver.Flags -import GHC.Driver.Session data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function - , ldHscEnv :: !HscEnv + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -266,54 +259,40 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- See Note [Using Byte Code rather than Object Code for Template Haskell] - homeModLinkable :: HomeModInfo -> Maybe Linkable + homeModLinkable :: HomeModInfo -> Maybe (IO Linkable) homeModLinkable hmi = - if ldUseByteCode opts - then homeModInfoByteCode hmi <|> homeModInfoObject hmi - else homeModInfoObject hmi <|> homeModInfoByteCode hmi + let obj = pure <$> homeModInfoObject hmi + bc = evalHomeModByteCode hmi + in if ldUseByteCode opts + then bc <|> obj + else obj <|> bc get_linkable osuf mod -- A home-package module | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) - = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) + = do + lnk <- expectJust "getLinkDeps" (homeModLinkable mod_info) + adjust_linkable lnk | otherwise = do -- It's not in the HPT because we are in one shot mode, -- so use the Finder to get a ModLocation... case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod - | prefer_bytecode = do - Succeeded iface <- ldLoadIface opts (text "load core bindings") mod - case mi_extra_decls iface of - Just extra_decls -> do - t <- getCurrentTime - let - stubs = mi_foreign iface - wcb = WholeCoreBindings extra_decls mod loc stubs - linkable = Linkable t mod (pure (CoreBindings wcb)) - initWholeCoreBindingsEps hsc_env iface linkable - _ -> fallback_no_bytecode loc mod - | otherwise = fallback_no_bytecode loc mod - - fallback_no_bytecode loc mod = do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - - prefer_bytecode = gopt Opt_UseBytecodeRatherThanObjects dflags - - dflags = hsc_dflags hsc_env - - hsc_env = ldHscEnv opts + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do @@ -338,9 +317,6 @@ get_link_deps opts pls maybe_normal_osuf span mods = do DotA fp -> panic ("adjust_ul DotA " ++ show fp) DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,19 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags - , ldHscEnv = hsc_env + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Linker/Types.hs ===================================== @@ -65,7 +65,6 @@ import Data.Time ( UTCTime ) import GHC.Unit.Module.Env import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM -import GHC.Unit.Module.WholeCoreBindings import Data.Maybe (mapMaybe) import Data.List.NonEmpty (NonEmpty, nonEmpty) import qualified Data.List.NonEmpty as NE @@ -284,18 +283,6 @@ data LinkablePart | DotDLL FilePath -- ^ Dynamically linked library file (.so, .dll, .dylib) - | CoreBindings WholeCoreBindings - -- ^ Serialised core which we can turn into BCOs (or object files), or - -- used by some other backend See Note [Interface Files with Core - -- Definitions] - - | LazyBCOs - CompiledByteCode - -- ^ Some BCOs generated on-demand when forced. This is used for - -- WholeCoreBindings, see Note [Interface Files with Core Definitions] - [FilePath] - -- ^ Objects containing foreign stubs and files - | BCOs CompiledByteCode -- ^ A byte-code object, lives only in memory. @@ -308,8 +295,6 @@ instance Outputable LinkablePart where ppr (DotA path) = text "DotA" <+> text path ppr (DotDLL path) = text "DotDLL" <+> text path ppr (BCOs bco) = text "BCOs" <+> ppr bco - ppr (LazyBCOs{}) = text "LazyBCOs" - ppr (CoreBindings {}) = text "CoreBindings" -- | Return true if the linkable only consists of native code (no BCO) linkableIsNativeCodeOnly :: Linkable -> Bool @@ -350,8 +335,6 @@ isNativeCode = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Is the part a native library? (.so/.dll) isNativeLib :: LinkablePart -> Bool @@ -360,8 +343,6 @@ isNativeLib = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Get the FilePath of linkable part (if applicable) linkablePartPath :: LinkablePart -> Maybe FilePath @@ -369,8 +350,6 @@ linkablePartPath = \case DotO fn _ -> Just fn DotA fn -> Just fn DotDLL fn -> Just fn - CoreBindings {} -> Nothing - LazyBCOs {} -> Nothing BCOs {} -> Nothing -- | Return the paths of all object code files (.o, .a, .so) contained in this @@ -380,8 +359,6 @@ linkablePartNativePaths = \case DotO fn _ -> [fn] DotA fn -> [fn] DotDLL fn -> [fn] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Return the paths of all object files (.o) contained in this 'LinkablePart'. @@ -390,8 +367,6 @@ linkablePartObjectPaths = \case DotO fn _ -> [fn] DotA _ -> [] DotDLL _ -> [] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Retrieve the compiled byte-code from the linkable part. @@ -400,7 +375,6 @@ linkablePartObjectPaths = \case linkablePartAllBCOs :: LinkablePart -> [CompiledByteCode] linkablePartAllBCOs = \case BCOs bco -> [bco] - LazyBCOs bcos _ -> [bcos] _ -> [] linkableFilter :: (LinkablePart -> [LinkablePart]) -> Linkable -> Maybe Linkable @@ -413,13 +387,11 @@ linkablePartNative = \case u at DotO {} -> [u] u at DotA {} -> [u] u at DotDLL {} -> [u] - LazyBCOs _ os -> [DotO f ForeignObject | f <- os] _ -> [] linkablePartByteCode :: LinkablePart -> [LinkablePart] linkablePartByteCode = \case u at BCOs {} -> [u] - LazyBCOs bcos _ -> [BCOs bcos] _ -> [] -- | Transform the 'LinkablePart' list in this 'Linkable' to contain only ===================================== compiler/GHC/Runtime/Eval.hs ===================================== @@ -1275,14 +1275,14 @@ showModule mod_summary = let interpreted = case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> isJust (homeModInfoByteCode mod_info) && isNothing (homeModInfoObject mod_info) + Just mod_info -> homeModInfoHasByteCode mod_info && isNothing (homeModInfoObject mod_info) return (showSDoc dflags $ showModMsg dflags interpreted (ModuleNode [] mod_summary)) moduleIsBootOrNotObjectLinkable :: GhcMonad m => ModSummary -> m Bool moduleIsBootOrNotObjectLinkable mod_summary = withSession $ \hsc_env -> case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> return . isNothing $ homeModInfoByteCode mod_info + Just mod_info -> return . not $ homeModInfoHasByteCode mod_info ---------------------------------------------------------------------------- -- RTTI primitives ===================================== compiler/GHC/Runtime/Interpreter.hs ===================================== @@ -113,6 +113,7 @@ import qualified GHC.Exts.Heap as Heap import GHC.Stack.CCS (CostCentre,CostCentreStack) import System.Directory import System.Process +import GHC.Unit.Module.ModDetails (ModDetails(..)) {- Note [Remote GHCi] ~~~~~~~~~~~~~~~~~~ @@ -684,10 +685,13 @@ fromEvalResult (EvalSuccess a) = return a getModBreaks :: HomeModInfo -> ModBreaks getModBreaks hmi - | Just linkable <- homeModInfoByteCode hmi, + | HomeModByteCode linkable <- homeModInfoByteCode hmi, -- The linkable may have 'DotO's as well; only consider BCOs. See #20570. [cbc] <- linkableBCOs linkable = fromMaybe emptyModBreaks (bc_breaks cbc) + | ModDetails {md_bytecode = Just _} <- hm_details hmi + -- TODO + = emptyModBreaks | otherwise = emptyModBreaks -- probably object code ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -45,8 +47,6 @@ type PackageCompleteMatches = CompleteMatches type PackageIfaceTable = ModuleEnv ModIface -- Domain = modules in the imported packages -type PackageTypeTable = ModuleEnv TypeEnv - -- | Constructs an empty PackageIfaceTable emptyPackageIfaceTable :: PackageIfaceTable emptyPackageIfaceTable = emptyModuleEnv @@ -70,7 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv - , eps_PTT = emptyModuleEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -142,7 +142,11 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules - eps_PTT :: !PackageTypeTable, + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules ===================================== compiler/GHC/Unit/Home/ModInfo.hs ===================================== @@ -1,13 +1,19 @@ +{-# LANGUAGE LambdaCase #-} + -- | Info about modules in the "home" unit module GHC.Unit.Home.ModInfo ( HomeModInfo (..) - , HomeModLinkable(..) + , HomeModLinkable (..) + , HomeModByteCode (..) , homeModInfoObject , homeModInfoByteCode + , homeModInfoHasByteCode , emptyHomeModInfoLinkable , justBytecode , justObjects , bytecodeAndObjects + , pureHomeModByteCode + , evalHomeModByteCode , HomePackageTable , emptyHomePackageTable , lookupHpt @@ -44,6 +50,7 @@ import GHC.Utils.Outputable import Data.List (sortOn) import Data.Ord import GHC.Utils.Panic +import GHC.Unit.Module.WholeCoreBindings (WholeCoreBindings) -- | Information about modules in the package being compiled data HomeModInfo = HomeModInfo @@ -73,37 +80,72 @@ data HomeModInfo = HomeModInfo -- 'ModIface' (only). } -homeModInfoByteCode :: HomeModInfo -> Maybe Linkable +homeModInfoByteCode :: HomeModInfo -> HomeModByteCode homeModInfoByteCode = homeMod_bytecode . hm_linkable +homeModInfoHasByteCode :: HomeModInfo -> Bool +homeModInfoHasByteCode hmi = case homeModInfoByteCode hmi of + NoHomeModByteCode -> False + _ -> True + homeModInfoObject :: HomeModInfo -> Maybe Linkable homeModInfoObject = homeMod_object . hm_linkable emptyHomeModInfoLinkable :: HomeModLinkable -emptyHomeModInfoLinkable = HomeModLinkable Nothing Nothing +emptyHomeModInfoLinkable = HomeModLinkable NoHomeModByteCode Nothing -- See Note [Home module build products] -data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !(Maybe Linkable) +data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !HomeModByteCode , homeMod_object :: !(Maybe Linkable) } instance Outputable HomeModLinkable where ppr (HomeModLinkable l1 l2) = ppr l1 $$ ppr l2 -justBytecode :: Linkable -> HomeModLinkable -justBytecode lm = - assertPpr (not (linkableIsNativeCodeOnly lm)) (ppr lm) - $ emptyHomeModInfoLinkable { homeMod_bytecode = Just lm } +justBytecode :: HomeModByteCode -> HomeModLinkable +justBytecode bc = + emptyHomeModInfoLinkable { homeMod_bytecode = bc } justObjects :: Linkable -> HomeModLinkable justObjects lm = assertPpr (linkableIsNativeCodeOnly lm) (ppr lm) $ emptyHomeModInfoLinkable { homeMod_object = Just lm } -bytecodeAndObjects :: Linkable -> Linkable -> HomeModLinkable +bytecodeAndObjects :: HomeModByteCode -> Linkable -> HomeModLinkable bytecodeAndObjects bc o = - assertPpr (not (linkableIsNativeCodeOnly bc) && linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) - (HomeModLinkable (Just bc) (Just o)) - + assertPpr (linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) + (HomeModLinkable bc (Just o)) + +pureHomeModByteCode :: HomeModByteCode -> Maybe Linkable +pureHomeModByteCode = \case + NoHomeModByteCode -> Nothing + HomeModIfaceCore _ -> Nothing + HomeModByteCode l -> Just l + +-- | Obtain the bytecode stored in this 'HomeModInfo', preferring the value in +-- 'HomeModLinkable' that's already in memory before evaluating the 'IO' in +-- 'ModDetails' that hydrates and parses Core loaded from an interface. +-- +-- This should only be called once in the module's lifecycle; afterwards, the +-- bytecode is cached in 'LoaderState'. +evalHomeModByteCode :: HomeModInfo -> Maybe (IO Linkable) +evalHomeModByteCode HomeModInfo {hm_details, hm_linkable} + | HomeModByteCode bc <- homeMod_bytecode hm_linkable + = Just (pure bc) + | otherwise + = md_bytecode hm_details + +data HomeModByteCode = + NoHomeModByteCode + | + HomeModIfaceCore WholeCoreBindings + | + HomeModByteCode Linkable + +instance Outputable HomeModByteCode where + ppr = \case + NoHomeModByteCode -> text "no bytecode" + HomeModIfaceCore _ -> text "dehydrated Core" + HomeModByteCode linkable -> ppr linkable {- Note [Home module build products] ===================================== compiler/GHC/Unit/Module/ModDetails.hs ===================================== @@ -14,6 +14,9 @@ import GHC.Types.DefaultEnv ( DefaultEnv, emptyDefaultEnv ) import GHC.Types.TypeEnv import GHC.Types.Annotations ( Annotation ) +import GHC.Linker.Types (Linkable) +import GHC.Prelude + -- | The 'ModDetails' is essentially a cache for information in the 'ModIface' -- for home modules only. Information relating to packages will be loaded into -- global environments in 'ExternalPackageState'. @@ -40,6 +43,8 @@ data ModDetails = ModDetails , md_complete_matches :: CompleteMatches -- ^ Complete match pragmas for this module + + , md_bytecode :: !(Maybe (IO Linkable)) } -- | Constructs an empty ModDetails @@ -53,4 +58,5 @@ emptyModDetails = ModDetails , md_fam_insts = [] , md_anns = [] , md_complete_matches = [] + , md_bytecode = Nothing } ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -27,6 +27,8 @@ import Data.Maybe (fromMaybe) import System.FilePath (takeExtension) {- +TODO update + Note [Interface Files with Core Definitions] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -87,7 +89,12 @@ data WholeCoreBindings = WholeCoreBindings , wcb_foreign :: IfaceForeign } +instance Outputable WholeCoreBindings where + ppr WholeCoreBindings {wcb_module} = text "iface Core for " <+> ppr wcb_module + {- +TODO update + Note [Foreign stubs and TH bytecode linking] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== ghc/GHCi/Leak.hs ===================================== @@ -50,7 +50,7 @@ getLeakIndicators hsc_env = where mkWeakLinkables :: HomeModLinkable -> IO [Maybe (Weak Linkable)] mkWeakLinkables (HomeModLinkable mbc mo) = - mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [mbc, mo] + mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [pureHomeModByteCode mbc, mo] -- | Look at the LeakIndicators collected by an earlier call to -- `getLeakIndicators`, and print messasges if any of them are still ===================================== testsuite/tests/driver/j-space/jspace.hs ===================================== @@ -42,17 +42,17 @@ initGhcM xs = do performGC [ys] <- filter (isPrefixOf (ghcUnitId <> ":GHC.Unit.Module.ModDetails.ModDetails")) . lines <$> readFile "jspace.hp" let (n :: Int) = read (last (words ys)) - -- The output should be 50 * 8 * word_size (i.e. 3600, or 1600 on 32-bit architectures): + -- The output should be 50 * 10 * word_size (i.e. 4000, or 2000 on 32-bit architectures): -- the test contains DEPTH + WIDTH + 2 = 50 modules J, H_0, .., H_DEPTH, W_1, .., W_WIDTH, - -- and each ModDetails contains 1 (info table) + 8 word-sized fields. + -- and each ModDetails contains 1 (info table) + 9 word-sized fields. -- If this number changes DO NOT ACCEPT THE TEST, you have introduced a space leak. -- -- There is some unexplained behaviour where the result is infrequently 3264.. but -- this resisted investigation using ghc-debug so the test actually checks whether there -- are less than 51 live ModDetails which is still a big improvement over before. - when (n > (51 * word_size * 9)) $ do + when (n > (51 * word_size * 10)) $ do putStrLn "Space leak detected by jspace test:" - putStrLn $ (show (n `div` (word_size * 9))) ++ " live ModDetails when <= 51 are expected" + putStrLn $ (show (n `div` (word_size * 10))) ++ " live ModDetails when <= 51 are expected" readFile "jspace.hp" >>= putStrLn exitFailure return () View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8baf8f1f3170aec0248b662e204aee884b0f8305 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8baf8f1f3170aec0248b662e204aee884b0f8305 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 09:54:49 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Tue, 03 Sep 2024 05:54:49 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] Revert "wip: temp file names" Message-ID: <66d6dce9b8065_363f6a102b4c416e5@gitlab.mail> Matthew Pickering pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 5f85ad9f by Matthew Pickering at 2024-09-03T10:53:43+01:00 Revert "wip: temp file names" This reverts commit 43be188ecf5de7b65fe093f7e9bd04228788ef33. - - - - - 2 changed files: - compiler/GHC/Driver/Make.hs - compiler/GHC/Utils/TmpFs.hs Changes: ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2924,22 +2924,19 @@ runParPipelines worker_limit plugin_hsc_env diag_wrapper mHscMessager all_pipeli atomically $ writeTVar stopped_var True wait_log_thread -withLocalTmpFS :: TmpFs -> (TmpFs -> IO a) -> IO a -withLocalTmpFS tmpfs act = do +withLocalTmpFS :: RunMakeM a -> RunMakeM a +withLocalTmpFS act = do let initialiser = do - liftIO $ forkTmpFsFrom tmpfs - finaliser tmpfs_local = do - liftIO $ mergeTmpFsInto tmpfs_local tmpfs + MakeEnv{..} <- ask + lcl_tmpfs <- liftIO $ forkTmpFsFrom (hsc_tmpfs hsc_env) + return $ hsc_env { hsc_tmpfs = lcl_tmpfs } + finaliser lcl_env = do + gbl_env <- ask + liftIO $ mergeTmpFsInto (hsc_tmpfs lcl_env) (hsc_tmpfs (hsc_env gbl_env)) -- Add remaining files which weren't cleaned up into local tmp fs for -- clean-up later. -- Clear the logQueue if this node had it's own log queue - MC.bracket initialiser finaliser act - -withLocalTmpFSMake :: MakeEnv -> (MakeEnv -> IO a) -> IO a -withLocalTmpFSMake env k = - withLocalTmpFS (hsc_tmpfs (hsc_env env)) $ \lcl_tmpfs - -> k (env { hsc_env = (hsc_env env) { hsc_tmpfs = lcl_tmpfs }}) - + MC.bracket initialiser finaliser $ \lcl_hsc_env -> local (\env -> env { hsc_env = lcl_hsc_env}) act -- | Run the given actions and then wait for them all to finish. runAllPipelines :: WorkerLimit -> MakeEnv -> [MakeAction] -> IO () @@ -2961,18 +2958,16 @@ runAllPipelines worker_limit env acts = do runLoop :: (((forall a. IO a -> IO a) -> IO ()) -> IO a) -> MakeEnv -> [MakeAction] -> IO [a] runLoop _ _env [] = return [] runLoop fork_thread env (MakeAction act res_var :acts) = do - - -- withLocalTmpFs has to occur outside of fork to remain deterministic - new_thread <- withLocalTmpFSMake env $ \lcl_env -> + new_thread <- fork_thread $ \unmask -> (do - mres <- (unmask $ run_pipeline lcl_env act) + mres <- (unmask $ run_pipeline (withLocalTmpFS act)) `MC.onException` (putMVar res_var Nothing) -- Defensive: If there's an unhandled exception then still signal the failure. putMVar res_var mres) threads <- runLoop fork_thread env acts return (new_thread : threads) where - run_pipeline :: MakeEnv -> RunMakeM a -> IO (Maybe a) - run_pipeline env p = runMaybeT (runReaderT p env) + run_pipeline :: RunMakeM a -> IO (Maybe a) + run_pipeline p = runMaybeT (runReaderT p env) data MakeAction = forall a . MakeAction !(RunMakeM a) !(MVar (Maybe a)) ===================================== compiler/GHC/Utils/TmpFs.hs ===================================== @@ -10,7 +10,6 @@ module GHC.Utils.TmpFs , emptyPathsToClean , TempFileLifetime(..) , TempDir (..) - , getTempDir , cleanTempDirs , cleanTempFiles , cleanCurrentModuleTempFiles @@ -65,8 +64,6 @@ data TmpFs = TmpFs -- -- Shared with forked TmpFs. - , tmp_dir_prefix :: String - , tmp_files_to_clean :: IORef PathsToClean -- ^ Files to clean (per session or per module) -- @@ -124,7 +121,6 @@ initTmpFs = do , tmp_subdirs_to_clean = subdirs , tmp_dirs_to_clean = dirs , tmp_next_suffix = next - , tmp_dir_prefix = "tmp" } -- | Initialise an empty TmpFs sharing unique numbers and per-process temporary @@ -136,16 +132,11 @@ forkTmpFsFrom :: TmpFs -> IO TmpFs forkTmpFsFrom old = do files <- newIORef emptyPathsToClean subdirs <- newIORef emptyPathsToClean - counter <- newIORef 0 - prefix <- newTempSuffix old - - return $ TmpFs { tmp_files_to_clean = files , tmp_subdirs_to_clean = subdirs , tmp_dirs_to_clean = tmp_dirs_to_clean old - , tmp_next_suffix = counter - , tmp_dir_prefix = prefix + , tmp_next_suffix = tmp_next_suffix old } -- | Merge the first TmpFs into the second. @@ -268,11 +259,9 @@ changeTempFilesLifetime tmpfs lifetime files = do addFilesToClean tmpfs lifetime existing_files -- Return a unique numeric temp file suffix -newTempSuffix :: TmpFs -> IO String -newTempSuffix tmpfs = do - n <- atomicModifyIORef' (tmp_next_suffix tmpfs) $ \n -> (n+1,n) - return $ tmp_dir_prefix tmpfs ++ "_" ++ show n - +newTempSuffix :: TmpFs -> IO Int +newTempSuffix tmpfs = + atomicModifyIORef' (tmp_next_suffix tmpfs) $ \n -> (n+1,n) -- Find a temporary name that doesn't already exist. newTempName :: Logger -> TmpFs -> TempDir -> TempFileLifetime -> Suffix -> IO FilePath @@ -282,8 +271,8 @@ newTempName logger tmpfs tmp_dir lifetime extn where findTempName :: FilePath -> IO FilePath findTempName prefix - = do suffix <- newTempSuffix tmpfs - let filename = prefix ++ suffix <.> extn + = do n <- newTempSuffix tmpfs + let filename = prefix ++ show n <.> extn b <- doesFileExist filename if b then findTempName prefix else do -- clean it up later @@ -306,8 +295,8 @@ newTempSubDir logger tmpfs tmp_dir where findTempDir :: FilePath -> IO FilePath findTempDir prefix - = do suffix <- newTempSuffix tmpfs - let name = prefix ++ suffix + = do n <- newTempSuffix tmpfs + let name = prefix ++ show n b <- doesDirectoryExist name if b then findTempDir prefix else (do @@ -325,8 +314,8 @@ newTempLibName logger tmpfs tmp_dir lifetime extn where findTempName :: FilePath -> String -> IO (FilePath, FilePath, String) findTempName dir prefix - = do suffix <- newTempSuffix tmpfs -- See Note [Deterministic base name] - let libname = prefix ++ suffix + = do n <- newTempSuffix tmpfs -- See Note [Deterministic base name] + let libname = prefix ++ show n filename = dir "lib" ++ libname <.> extn b <- doesFileExist filename if b then findTempName dir prefix @@ -351,8 +340,8 @@ getTempDir logger tmpfs (TempDir tmp_dir) = do mkTempDir :: FilePath -> IO FilePath mkTempDir prefix = do - suffix <- newTempSuffix tmpfs - let our_dir = prefix ++ suffix + n <- newTempSuffix tmpfs + let our_dir = prefix ++ show n -- 1. Speculatively create our new directory. createDirectory our_dir View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5f85ad9fafd3fa885ab888c6b4ea160794cdf0b9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5f85ad9fafd3fa885ab888c6b4ea160794cdf0b9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 12:14:50 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 03 Sep 2024 08:14:50 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 6 commits: simplifier: Fix space leak during demand analysis Message-ID: <66d6fdba787b9_1d45df21500c99674@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 6f239537 by Andreas Klebinger at 2024-09-03T08:14:22-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - 3ad9ee55 by Arsen Arsenović at 2024-09-03T08:14:29-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 86761127 by Torsten Schmits at 2024-09-03T08:14:29-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 29 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Driver/Main.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/ffi.rst - libraries/base/src/GHC/Base.hs - libraries/base/src/GHC/Exts.hs - libraries/ghc-prim/changelog.md - rts/PrimOps.cmm - rts/RtsSymbols.c - rts/include/stg/MiscClosures.h - testsuite/tests/rts/T13894.hs - + testsuite/tests/th/T25209.hs - + testsuite/tests/th/T25209.stderr - testsuite/tests/th/all.T - utils/ghc-toolchain/src/GHC/Toolchain/Tools/Cpp.hs - utils/haddock/.gitignore - utils/haddock/CHANGES.md - utils/haddock/doc/invoking.rst - utils/haddock/haddock-api/src/Haddock.hs - utils/haddock/haddock-api/src/Haddock/Interface.hs - utils/haddock/haddock-api/src/Haddock/Options.hs - utils/haddock/haddock-test/src/Test/Haddock.hs - utils/haddock/haddock-test/src/Test/Haddock/Config.hs - utils/haddock/latex-test/Main.hs Changes: ===================================== compiler/GHC/Builtin/primops.txt.pp ===================================== @@ -1925,7 +1925,25 @@ primop MutableByteArrayIsPinnedOp "isMutableByteArrayPinned#" GenPrimOp primop ByteArrayIsPinnedOp "isByteArrayPinned#" GenPrimOp ByteArray# -> Int# - {Determine whether a 'ByteArray#' is guaranteed not to move during GC.} + {Determine whether a 'ByteArray#' is guaranteed not to move.} + with out_of_line = True + +primop ByteArrayIsWeaklyPinnedOp "isByteArrayWeaklyPinned#" GenPrimOp + ByteArray# -> Int# + {Similar to 'isByteArrayPinned#'. Weakly pinned byte arrays are allowed + to be copied into compact regions by the user, potentially invalidating + the results of earlier calls to 'byteArrayContents#'. + + See the section `Pinned Byte Arrays` in the user guide for more information. + + This function also returns true for regular pinned bytearrays. + } + with out_of_line = True + +primop MutableByteArrayIsWeaklyPinnedOp "isMutableByteArrayWeaklyPinned#" GenPrimOp + MutableByteArray# s -> Int# + { 'isByteArrayWeaklyPinned#' but for mutable arrays. + } with out_of_line = True primop ByteArrayContents_Char "byteArrayContents#" GenPrimOp ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -1,5 +1,4 @@ {-# LANGUAGE LambdaCase #-} - {-# LANGUAGE NondecreasingIndentation #-} {-# LANGUAGE GADTs #-} {-# LANGUAGE MultiWayIf #-} @@ -295,13 +294,13 @@ import Data.Time import System.IO.Unsafe ( unsafeInterleaveIO ) import GHC.Iface.Env ( trace_if ) +import GHC.Platform.Ways import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) import GHC.Types.Unique.FM import GHC.Types.Unique.DFM import GHC.Cmm.Config (CmmConfig) - {- ********************************************************************** %* * Initialisation @@ -990,6 +989,27 @@ initModDetails hsc_env iface = -- in make mode, since this HMI will go into the HPT. genModDetails hsc_env' iface +-- | Modify flags such that objects are compiled for the interpreter's way. +-- This is necessary when building foreign objects for Template Haskell, since +-- those are object code built outside of the pipeline, which means they aren't +-- subject to the mechanism in 'enableCodeGenWhen' that requests dynamic build +-- outputs for dependencies when the interpreter used for TH is dynamic but the +-- main outputs aren't. +-- Furthermore, the HPT only stores one set of objects with different names for +-- bytecode linking in 'HomeModLinkable', so the usual hack for switching +-- between ways in 'get_link_deps' doesn't work. +compile_for_interpreter :: HscEnv -> (HscEnv -> IO a) -> IO a +compile_for_interpreter hsc_env use = + use (hscUpdateFlags update hsc_env) + where + update dflags = dflags { + targetWays_ = adapt_way interpreterDynamic WayDyn $ + adapt_way interpreterProfiled WayProf $ + targetWays_ dflags + } + + adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace -- them with a lazy IO thunk that compiles them to bytecode and foreign objects. -- @@ -2063,9 +2083,10 @@ generateByteCode :: HscEnv -> IO (CompiledByteCode, [FilePath]) generateByteCode hsc_env cgguts mod_location = do (hasStub, comp_bc) <- hscInteractive hsc_env cgguts mod_location - stub_o <- traverse (compileForeign hsc_env LangC) hasStub - foreign_files_o <- traverse (uncurry (compileForeign hsc_env)) (cgi_foreign_files cgguts) - pure (comp_bc, maybeToList stub_o ++ foreign_files_o) + compile_for_interpreter hsc_env $ \ i_env -> do + stub_o <- traverse (compileForeign i_env LangC) hasStub + foreign_files_o <- traverse (uncurry (compileForeign i_env)) (cgi_foreign_files cgguts) + pure (comp_bc, maybeToList stub_o ++ foreign_files_o) generateFreshByteCode :: HscEnv -> ModuleName ===================================== compiler/GHC/StgToCmm/Prim.hs ===================================== @@ -1668,10 +1668,12 @@ emitPrimOp cfg primop = NewPinnedByteArrayOp_Char -> alwaysExternal NewAlignedPinnedByteArrayOp_Char -> alwaysExternal MutableByteArrayIsPinnedOp -> alwaysExternal + MutableByteArrayIsWeaklyPinnedOp -> alwaysExternal DoubleDecode_2IntOp -> alwaysExternal DoubleDecode_Int64Op -> alwaysExternal FloatDecode_IntOp -> alwaysExternal ByteArrayIsPinnedOp -> alwaysExternal + ByteArrayIsWeaklyPinnedOp -> alwaysExternal ShrinkMutableByteArrayOp_Char -> alwaysExternal ResizeMutableByteArrayOp_Char -> alwaysExternal ShrinkSmallMutableArrayOp_Char -> alwaysExternal ===================================== compiler/GHC/StgToJS/Prim.hs ===================================== @@ -670,6 +670,8 @@ genPrim prof bound ty op = case op of NewAlignedPinnedByteArrayOp_Char -> \[r] [l,_align] -> pure $ PrimInline (newByteArray r l) MutableByteArrayIsPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ ByteArrayIsPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ + ByteArrayIsWeaklyPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ + MutableByteArrayIsWeaklyPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ ByteArrayContents_Char -> \[a,o] [b] -> pure $ PrimInline $ mconcat [a |= b, o |= zero_] MutableByteArrayContents_Char -> \[a,o] [b] -> pure $ PrimInline $ mconcat [a |= b, o |= zero_] ShrinkMutableByteArrayOp_Char -> \[] [a,n] -> pure $ PrimInline $ appS hdShrinkMutableByteArrayStr [a,n] ===================================== compiler/GHC/ThToHs.hs ===================================== @@ -1519,7 +1519,7 @@ cvtp (ViewP e p) = do { e' <- cvtl e; p' <- cvtPat p ; wrapParLA gParPat $ ViewPat noAnn e' p'} cvtp (TypeP t) = do { t' <- cvtType t ; return $ EmbTyPat noAnn (mkHsTyPat t') } -cvtp (InvisP t) = do { t' <- cvtType t +cvtp (InvisP t) = do { t' <- parenthesizeHsType appPrec <$> cvtType t ; pure (InvisPat noAnn (mkHsTyPat t'))} cvtp (OrP ps) = do { ps' <- cvtPats ps ; pure (OrPat noExtField ps')} ===================================== compiler/GHC/Types/Demand.hs ===================================== @@ -1913,7 +1913,7 @@ multDmdType :: Card -> DmdType -> DmdType multDmdType n (DmdType fv args) = -- pprTrace "multDmdType" (ppr n $$ ppr fv $$ ppr (multDmdEnv n fv)) $ DmdType (multDmdEnv n fv) - (map (multDmd n) args) + (strictMap (multDmd n) args) peelFV :: DmdType -> Var -> (DmdType, Demand) peelFV (DmdType fv ds) id = -- pprTrace "rfv" (ppr id <+> ppr dmd $$ ppr fv) ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -190,13 +190,9 @@ Even if that wasn't an issue, they are compiled for the session's 'Way', not its associated module's, so the dynamic variant wouldn't be available when building only static outputs. -For now, this doesn't have much of an impact, since we're only supporting -foreign imports initially, which produce very simple objects that can easily be -handled by the linker when 'GHC.Linker.Loader.dynLoadObjs' creates a shared -library from all object file inputs. -However, for more complex circumstances, we should compile foreign stubs -specially for TH according to the interpreter 'Way', or request dynamic products -for TH dependencies like it happens for the conventional case. +To mitigate this, we instead build foreign objects specially for the +interpreter, updating the build flags in 'compile_for_interpreter' to use the +interpreter's way. Problem 4: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -163,6 +163,12 @@ Runtime system ~~~~~~~~~~~~~~~~~~~~ - Usage of deprecated primops is now correctly reported (#19629). +- New primops `isMutableByteArrayWeaklyPinned#` and `isByteArrayWeaklyPinned#` + to allow users to avoid copying large arrays safely when dealing with ffi. + See the users guide for more details on the different kinds of + pinned arrays in 9.12. + + This need for this distinction originally surfaced in https://gitlab.haskell.org/ghc/ghc/-/issues/22255 ``ghc`` library ===================================== docs/users_guide/exts/ffi.rst ===================================== @@ -1114,21 +1114,67 @@ Pinned Byte Arrays A pinned byte array is one that the garbage collector is not allowed to move. Consequently, it has a stable address that can be safely -requested with ``byteArrayContents#``. Not that being pinned doesn't -prevent the byteArray from being gc'ed in the same fashion a regular -byte array would be. +requested with ``byteArrayContents#``. As long as the array remains live +the address returned by ``byteArrayContents#`` will remain valid. Note that +being pinned doesn't prevent the byteArray from being gc'ed in the same fashion +a regular byte array would be if there are no more references to the ``ByteArray#``. There are a handful of primitive functions in :base-ref:`GHC.Exts.` used to enforce or check for pinnedness: ``isByteArrayPinned#``, -``isMutableByteArrayPinned#``, and ``newPinnedByteArray#``. A -byte array can be pinned as a result of three possible causes: +``isMutableByteArrayPinned#``, ``isByteArrayWeaklyPinned#``, +``isMutableByteArrayWeaklyPinned#``, and ``newPinnedByteArray#``. A +byte array can be pinned or weakly pinned as a result of three possible causes: -1. It was allocated by ``newPinnedByteArray#``. -2. It is large. Currently, GHC defines large object to be one +1. It was allocated by ``newPinnedByteArray#``. This results in a regular pinned byte array. +2. It is large, this results in a weakly pinned byte array. Currently, GHC defines large object to be one that is at least as large as 80% of a 4KB block (i.e. at least 3277 bytes). -3. It has been copied into a compact region. The documentation +3. It has been copied into a compact region, resulting in a weakly pinned array. The documentation for ``ghc-compact`` and ``compact`` describes this process. +The difference between a pinned array and a weakly pinned array is simply that +trying to compact a pinned array will result in an exception. Trying to compact +a weakly pinned array will succeed. However the result of earlier +calls to ``byteArrayContents#`` is not updated during compaction, which means +these results will still point to the address where the array was located originally, +and not to the new address inside the compact region. + +This is particularly dangerous when an address to a byte arrays content is stored +inside a datastructure along with a reference to the byte array. +If the data structure is compacted later on the pointer won't be updated but the +reference to the byte array will point to a copy inside the compact region. +A common data type susceptible to this is `ForeignPtr` when used to represent a ByteArray#. + +Here is an example to illustrate this: + +.. code-block:: haskell + + workWithArrayContents :: (ByteArray, Ptr Word8) -> (Ptr Word8 -> IO ()) -> IO () + workWithArrayContents (arr@(ByteArray uarr),ptr) worker = + case () of + _ + -- Conservative but safe + | isByteArrayPinned arr -> keepAliveUnlifted uarr (worker ptr) + -- Potentially dangerous, the program needs to ensures the Ptr points into the array. + | isByteArrayWeaklyPinned arr -> keepAliveUnlifted uarr (worker ptr) + | otherwise -> ... -- Otherwise we can't directly use it for safe FFI calls directly at all. + + main :: IO () + main = do + -- We create a large array, which causes it to be implicitly pinned + arr <- newByteArray 5000 + arr@(ByteArray uarr) <- freezeByteArray arr 0 5000 -- Make it immutable + let ptr = byteArrayContents arr + + -- Compacting a data structure that contains both an array and a ptr to + -- the arrays content's is dangerous and usually the wrong thing to do. + let foo = (arr, ptr) + foo_compacted <- compact foo + + -- This is fine + workWithArrayContents foo do_work + -- This is unsound + workWithArrayContents (getCompact foo_compacted) do_work + .. [1] Prior to GHC 8.10, when passing an ``ArrayArray#`` argument to a foreign function, the foreign function would see a pointer to the ``StgMutArrPtrs`` rather than just the payload. ===================================== libraries/base/src/GHC/Base.hs ===================================== @@ -139,7 +139,8 @@ module GHC.Base ) where import GHC.Internal.Base -import GHC.Prim hiding (dataToTagLarge#, dataToTagSmall#, whereFrom#) +import GHC.Prim hiding (dataToTagLarge#, dataToTagSmall#, whereFrom#, + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned#) -- Hide dataToTagLarge# because it is expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -112,7 +112,9 @@ module GHC.Exts import GHC.Internal.Exts import GHC.Internal.ArrayArray -import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# ) +import GHC.Prim hiding ( + coerce, dataToTagSmall#, dataToTagLarge#, whereFrom#, + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/ghc-prim/changelog.md ===================================== @@ -1,3 +1,12 @@ +## 0.13.0 + +- Shipped with GHC 9.12.1 + +- Add primops that allow users to distinguish weakly pinned byte arrays from unpinned ones. + + isMutableByteArrayWeaklyPinned# :: MutableByteArray# s -> Int# + isByteArrayWeaklyPinned# :: ByteArray# s -> Int# + ## 0.12.0 - Shipped with GHC 9.10.1 ===================================== rts/PrimOps.cmm ===================================== @@ -215,12 +215,29 @@ stg_isByteArrayPinnedzh ( gcptr ba ) return (flags & BF_PINNED != 0); } +stg_isByteArrayWeaklyPinnedzh ( gcptr ba ) +// ByteArray# s -> Int# +{ + W_ bd, flags; + bd = Bdescr(ba); + // See #22255 and the primop docs. + flags = TO_W_(bdescr_flags(bd)); + + return (flags & (BF_PINNED | BF_COMPACT | BF_LARGE) != 0); +} + stg_isMutableByteArrayPinnedzh ( gcptr mba ) // MutableByteArray# s -> Int# { jump stg_isByteArrayPinnedzh(mba); } +stg_isMutableByteArrayWeaklyPinnedzh ( gcptr mba ) +// MutableByteArray# s -> Int# +{ + jump stg_isByteArrayWeaklyPinnedzh(mba); +} + /* Note [LDV profiling and resizing arrays] * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * As far as the LDV profiler is concerned arrays are "inherently used" which ===================================== rts/RtsSymbols.c ===================================== @@ -656,6 +656,8 @@ extern char **environ; SymI_HasDataProto(stg_newAlignedPinnedByteArrayzh) \ SymI_HasDataProto(stg_isByteArrayPinnedzh) \ SymI_HasDataProto(stg_isMutableByteArrayPinnedzh) \ + SymI_HasDataProto(stg_isByteArrayWeaklyPinnedzh) \ + SymI_HasDataProto(stg_isMutableByteArrayWeaklyPinnedzh) \ SymI_HasDataProto(stg_shrinkMutableByteArrayzh) \ SymI_HasDataProto(stg_resizzeMutableByteArrayzh) \ SymI_HasDataProto(stg_shrinkSmallMutableArrayzh) \ ===================================== rts/include/stg/MiscClosures.h ===================================== @@ -454,6 +454,8 @@ RTS_FUN_DECL(stg_newPinnedByteArrayzh); RTS_FUN_DECL(stg_newAlignedPinnedByteArrayzh); RTS_FUN_DECL(stg_isByteArrayPinnedzh); RTS_FUN_DECL(stg_isMutableByteArrayPinnedzh); +RTS_FUN_DECL(stg_isByteArrayWeaklyPinnedzh); +RTS_FUN_DECL(stg_isMutableByteArrayWeaklyPinnedzh); RTS_FUN_DECL(stg_shrinkMutableByteArrayzh); RTS_FUN_DECL(stg_resizzeMutableByteArrayzh); RTS_FUN_DECL(stg_shrinkSmallMutableArrayzh); ===================================== testsuite/tests/rts/T13894.hs ===================================== @@ -6,6 +6,7 @@ import Control.Monad import GHC.Exts +import GHC.Internal.Exts (isMutableByteArrayWeaklyPinned#) import GHC.IO main :: IO () @@ -16,3 +17,10 @@ main = do case isMutableByteArrayPinned# arr# of n# -> (# s1, isTrue# n# #) when pinned $ putStrLn "BAD" + + weakly_pinned <- IO $ \s0 -> + case newByteArray# 1000000# s0 of + (# s1, arr# #) -> + case isMutableByteArrayWeaklyPinned# arr# of + n# -> (# s1, isTrue# n# #) + when (not weakly_pinned) $ putStrLn "BAD" ===================================== testsuite/tests/th/T25209.hs ===================================== @@ -0,0 +1,9 @@ +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE TypeAbstractions #-} +module T25209 where + +import Data.Proxy + +$([d| f :: Proxy a -> Proxy a + f @(a :: k) p = p + |]) ===================================== testsuite/tests/th/T25209.stderr ===================================== @@ -0,0 +1,6 @@ +T25209.hs:(7,2)-(9,7): Splicing declarations + [d| f :: Proxy a -> Proxy a + f @(a :: k) p = p |] + ======> + f :: Proxy a -> Proxy a + f @(a :: k) p = p ===================================== testsuite/tests/th/all.T ===================================== @@ -622,4 +622,5 @@ test('T24572a', normal, compile, ['']) test('T24572b', normal, compile_fail, ['']) test('T24572c', normal, compile_fail, ['']) test('T24572d', normal, compile, ['']) +test('T25209', normal, compile, ['-v0 -ddump-splices -dsuppress-uniques']) test('TH_MultilineStrings', normal, compile_and_run, ['']) ===================================== utils/ghc-toolchain/src/GHC/Toolchain/Tools/Cpp.hs ===================================== @@ -39,10 +39,10 @@ data CmmCpp = CmmCpp { cmmCppProgram :: Program } deriving (Show, Read, Eq, Ord) -checkFlag :: String -> Program -> String -> M () -checkFlag conftest cpp flag = checking ("for "++flag++" support") $ +checkFlag :: String -> Program -> String -> [String] -> M () +checkFlag conftest cpp flag extra_args = checking ("for "++flag++" support") $ -- Werror to ensure that unrecognized warnings result in an error - callProgram cpp ["-Werror", flag, conftest] + callProgram cpp $ ["-Werror", flag, conftest] ++ extra_args -- tryFlag :: String -> Program -> String -> M [String] -- tryFlag conftest cpp flag = -- ([flag] <$ checkFlag conftest cpp flag) <|> return [] @@ -167,7 +167,7 @@ findCmmCpp progOpt cc = checking "for a Cmm preprocessor" $ do cmmCppSupportsG0 <- withTempDir $ \dir -> do let conftest = dir "conftest.c" writeFile conftest "int main(void) {}" - True <$ checkFlag conftest cpp "-g0" <|> return False + True <$ checkFlag conftest cpp "-g0" ["-o", dir "conftest"] <|> return False -- Always add the -E flag to the CPP, regardless of the user options let cmmCppProgram = foldr addFlagIfNew cpp ["-E"] ===================================== utils/haddock/.gitignore ===================================== @@ -11,6 +11,10 @@ /hypsrc-test/one-shot-out/ /latex-test/one-shot-out/ /hoogle-test/one-shot-out/ +/html-test/no-compilation-out/ +/hypsrc-test/no-compilation-out/ +/latex-test/no-compilation-out/ +/hoogle-test/no-compilation-out/ *.o *.hi ===================================== utils/haddock/CHANGES.md ===================================== @@ -3,6 +3,10 @@ * Add incremental mode to support rendering documentation one module at a time. + * The flag `--no-compilation` has been added. This flag causes Haddock to avoid + recompilation of the code when generating documentation by only reading + the `.hi` and `.hie` files, and will throw an error if it can't find them. + * Fix large margin on top of small headings * Include `package_info` with haddock's `--show-interface` option. ===================================== utils/haddock/doc/invoking.rst ===================================== @@ -542,6 +542,13 @@ The following options are available: ``cabal`` uses temporary `response files `_ to pass arguments to Haddock. +.. option:: --no-compilation + + Always :ref:`avoids recompilation`, only loads the + required ``.hi`` and ``.hie`` files. Haddock will throw an error when it can't + find them. This will not check if the input files are out of date. + (This flag implies :option:`--no-tmp-comp-dir`.) + .. option:: --incremental= Use Haddock in :ref:`incremental mode`. Haddock will generate @@ -555,6 +562,8 @@ sources are accepted without the need for the user to do anything. To use the C pre-processor, however, the user must pass the ``-cpp`` option to GHC using :option:`--optghc`. +.. _avoiding-recompilation: + Avoiding recompilation ---------------------- @@ -579,32 +588,15 @@ should write the ``.hi`` and ``.hie`` files by providing the are building your application with ``cabal build``, the default location is in ``dist-newstyle/build/-/ghc-/-0.1.0/build``. -The next step is to ensure that the flags which Haddock passes to GHC will not -trigger recompilation. Unfortunately, this is not very easy to do if you are -invoking Haddock through ``cabal haddock``. Upon ``cabal haddock``, Cabal passes -a ``--optghc="-optP-D__HADDOCK_VERSION__=NNNN"`` (where ``NNNN`` is the Haddock -version number) flag to Haddock, which forwards the ``-optP=...`` flag to GHC -and triggers a recompilation (unless the existing build results were also -created by a ``cabal haddock``). Additionally, Cabal passes a -``--optghc="-stubdir="`` flag to Haddock, which forwards the -``-stubdir=`` flag to GHC and triggers a recompilation since -``-stubdir`` adds a global include directory. Moreover, since the ``stubdir`` -that Cabal passes is a temporary directory, a recompilation is triggered even -for immediately successive invocations. To avoid recompilations due to these -flags, one must manually extract the arguments passed to Haddock by Cabal and -remove the ``--optghc="-optP-D__HADDOCK_VERSION__=NNNN"`` and -``--optghc="-stubdir="`` flags. This can be achieved using the -:option:`--trace-args` flag by invoking ``cabal haddock`` with -``--haddock-option="--trace-args"`` and copying the traced arguments to a script -which makes an equivalent call to Haddock without the aformentioned flags. - -In addition to the above, Cabal passes a temporary directory as ``-hidir`` to -Haddock by default. Obviously, this also triggers a recompilation for every -invocation of ``cabal haddock``, since it will never find the necessary +The next step is to make sure Haddock runs in no-compilation mode by using +the :option:`--no-compilation` flag. In addition, Cabal passes a +temporary directory as ``-hidir`` to Haddock by default. This will cause +``cabal haddock`` to error, since it will never find the necessary interface files in that temporary directory. To remedy this, pass a ``--optghc="-hidir=/path/to/hidir"`` flag to Haddock, where ``/path/to/hidir`` is the path to the directory in which your build process is writing ``.hi`` -files. +files. You can do this by invoking ``cabal haddock`` with +``--haddock-options="--no-compilation --optghc=-hidir --optghc=/path/to/hidir"``. Following the steps above will allow you to take full advantage of "hi-haddock" and generate Haddock documentation from existing build results without requiring ===================================== utils/haddock/haddock-api/src/Haddock.hs ===================================== @@ -166,14 +166,14 @@ haddockWithGhc ghc args = handleTopExceptions $ do qual <- rightOrThrowE (qualification flags) sinceQual <- rightOrThrowE (sinceQualification flags) - let isOneShotMode = isJust (optOneShot flags) + let noCompilation = isJust (optOneShot flags) || Flag_NoCompilation `elem` flags -- Inject dynamic-too into ghc options if the ghc we are using was built with - -- dynamic linking (except when in one-shot mode) + -- dynamic linking (except when not doing any compilation) flags'' <- ghc flags $ do df <- getDynFlags case lookup "GHC Dynamic" (compilerInfo df) of - Just "YES" | not isOneShotMode -> return $ Flag_OptGhc "-dynamic-too" : flags + Just "YES" | not noCompilation -> return $ Flag_OptGhc "-dynamic-too" : flags _ -> return flags -- Inject `-j` into ghc options, if given to Haddock @@ -191,8 +191,7 @@ haddockWithGhc ghc args = handleTopExceptions $ do -- Output dir needs to be set before calling 'depanal' since 'depanal' uses it -- to compute output file names that are stored in the 'DynFlags' of the -- resulting 'ModSummary's. - let withDir | Flag_NoTmpCompDir `elem` flags = id - | isOneShotMode = id + let withDir | Flag_NoTmpCompDir `elem` flags || noCompilation = id | otherwise = withTempOutputDir -- Output warnings about potential misuse of some flags ===================================== utils/haddock/haddock-api/src/Haddock/Interface.hs ===================================== @@ -176,19 +176,21 @@ createIfaces verbosity modules flags instIfaceMap = do dflags <- getSessionDynFlags let dflags' = dflags { ldInputs = map (FileOption "") o_files ++ ldInputs dflags } - _ <- setSessionDynFlags dflags' + dflags'' = if Flag_NoCompilation `elem` flags then dflags' { ghcMode = OneShot } else dflags' + _ <- setSessionDynFlags dflags'' targets <- mapM (\(filePath, _) -> guessTarget filePath Nothing Nothing) hs_srcs setTargets targets (_errs, modGraph) <- depanalE [] False - liftIO $ traceMarkerIO "Load started" - -- Create (if necessary) and load .hi-files. - success <- withTimingM "load'" (const ()) $ - load' noIfaceCache LoadAllTargets mkUnknownDiagnostic (Just batchMsg) modGraph - when (failed success) $ do - out verbosity normal "load' failed" - liftIO exitFailure - liftIO $ traceMarkerIO "Load ended" + -- Create (if necessary) and load .hi-files. With --no-compilation this happens later. + when (Flag_NoCompilation `notElem` flags) $ do + liftIO $ traceMarkerIO "Load started" + success <- withTimingM "load'" (const ()) $ + load' noIfaceCache LoadAllTargets mkUnknownDiagnostic (Just batchMsg) modGraph + when (failed success) $ do + out verbosity normal "load' failed" + liftIO exitFailure + liftIO $ traceMarkerIO "Load ended" -- We topologically sort the module graph including boot files, -- so it should be acylic (hopefully we failed much earlier if this is not the case) @@ -260,6 +262,20 @@ dropErr :: MaybeErr e a -> Maybe a dropErr (Succeeded a) = Just a dropErr (Failed _) = Nothing +loadHiFile :: HscEnv -> Outputable.SDoc -> Module -> IO (ModIface, ([ClsInst], [FamInst])) +loadHiFile hsc_env doc theModule = initIfaceLoad hsc_env $ do + + mod_iface <- loadSysInterface doc theModule + + insts <- initIfaceLcl (mi_semantic_module mod_iface) doc (mi_boot mod_iface) $ do + + new_eps_insts <- mapM tcIfaceInst (mi_insts mod_iface) + new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts mod_iface) + + pure (new_eps_insts, new_eps_fam_insts) + + pure (mod_iface, insts) + processModule :: Verbosity -> ModSummary -> [Flag] -> IfaceMap -> InstIfaceMap -> WarningMap -> Ghc (Maybe Interface) processModule verbosity modSummary flags ifaceMap instIfaceMap warningMap = do out verbosity verbose $ "Checking module " ++ moduleString (ms_mod modSummary) ++ "..." @@ -267,17 +283,19 @@ processModule verbosity modSummary flags ifaceMap instIfaceMap warningMap = do hsc_env <- getSession dflags <- getDynFlags let sDocContext = DynFlags.initSDocContext dflags Outputable.defaultUserStyle - let hmi = case lookupHpt (hsc_HPT hsc_env) (moduleName $ ms_mod modSummary) of - Nothing -> error "processModule: All modules should be loaded into the HPT by this point" - Just x -> x - mod_iface = hm_iface hmi + doc = text "processModule" unit_state = hsc_units hsc_env - cls_insts = instEnvElts . md_insts $ hm_details hmi + (mod_iface, insts) <- if Flag_NoCompilation `elem` flags + then liftIO $ loadHiFile hsc_env doc $ ms_mod modSummary + else + let hmi = case lookupHpt (hsc_HPT hsc_env) (moduleName $ ms_mod modSummary) of + Nothing -> error "processModule: All modules should be loaded into the HPT by this point" + Just x -> x + cls_insts = instEnvElts . md_insts $ hm_details hmi + fam_insts = md_fam_insts $ hm_details hmi - fam_insts = md_fam_insts $ hm_details hmi - - insts = (cls_insts, fam_insts) + in pure (hm_iface hmi, (cls_insts, fam_insts)) !interface <- do logger <- getLogger @@ -363,18 +381,7 @@ createOneShotIface verbosity flags instIfaceMap moduleNameStr = do modifySession $ hscSetFlags dflags hsc_env <- getSession - (iface, insts) <- liftIO $ initIfaceLoad hsc_env $ do - - iface <- loadSysInterface doc $ mkMainModule_ moduleNm - - insts <- initIfaceLcl (mi_semantic_module iface) doc (mi_boot iface) $ do - - new_eps_insts <- mapM tcIfaceInst (mi_insts iface) - new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) - - pure (new_eps_insts, new_eps_fam_insts) - - pure (iface, insts) + (iface, insts) <- liftIO $ loadHiFile hsc_env doc $ mkMainModule_ moduleNm -- Update the DynFlags with the extensions from the source file (as stored in the interface file) -- This is instead of ms_hspp_opts from ModSummary, which is not available in one-shot mode. ===================================== utils/haddock/haddock-api/src/Haddock/Options.hs ===================================== @@ -124,6 +124,7 @@ data Flag | Flag_ParCount (Maybe Int) | Flag_TraceArgs | Flag_OneShot String + | Flag_NoCompilation deriving (Eq, Show) options :: Bool -> [OptDescr Flag] @@ -158,6 +159,11 @@ options backwardsCompat = ["show-interface"] (ReqArg Flag_ShowInterface "FILE") "print the interface in a human readable form" + , Option + [] + ["no-compilation"] + (NoArg Flag_NoCompilation) + "never compile the code, just read the .hi files" , Option [] ["incremental"] ===================================== utils/haddock/haddock-test/src/Test/Haddock.hs ===================================== @@ -10,6 +10,7 @@ module Test.Haddock import Control.Monad import qualified Data.ByteString.Char8 as BS import qualified Data.Map.Strict as Map +import Data.Foldable (for_) import Data.Maybe import GHC.ResponseFile import System.Directory @@ -74,6 +75,7 @@ maybeDiff cfg@(Config{cfgDiffTool = (Just diff)}) files = do runHaddock :: Config c -> IO Bool runHaddock cfg@(Config{..}) = do createEmptyDirectory $ cfgOutDir cfg + createEmptyDirectory $ cfgNoCompilationOutDir cfg createEmptyDirectory $ cfgOneShotOutDir cfg putStrLn "Generating documentation..." @@ -93,41 +95,65 @@ runHaddock cfg@(Config{..}) = do succeeded <- waitForSuccess msg stdout =<< runProcess' cfgHaddockPath pc unless succeeded $ removeDirectoryRecursive (outDir cfgDirConfig tpkg) - if cfgSkipOneShot then pure succeeded else do - let oneShotDir = oneshotOutDir cfgDirConfig tpkg - hiDir = oneShotDir "hi" - hieDir = oneShotDir "hie" + let noCompilationDir = noCompilationOutDir cfgDirConfig tpkg + hiDir = noCompilationDir "hi" + hieDir = noCompilationDir "hie" + + createEmptyDirectory noCompilationDir + createEmptyDirectory hiDir + createEmptyDirectory hieDir + + -- Build .hi files + let pc = + processConfig + { pcArgs = + concat + [ + [ "--make" + , "-haddock" + , "-fwrite-interface" + , "-fwrite-ide-info" + , "-no-keep-o-files" + , "-hidir=" ++ hiDir + , "-hiedir=" ++ hieDir + ] + , tpkgFiles tpkg + ] + , pcEnv = Just cfgEnv + } + let msg = "Failed to run GHC on test package '" ++ tpkgName tpkg ++ "'" + _ <- waitForSuccess msg stdout =<< runProcess' cfgGhcPath pc + + -- Generate documentation with no-compilation flag + let pc = + processConfig + { pcArgs = + concat + [ cfgHaddockArgs + , [ "--odir=" ++ noCompilationDir + , "--optghc=-hidir=" ++ hiDir + , "--optghc=-hiedir=" ++ hieDir + , "--no-compilation" + ] + , tpkgFiles tpkg + ] + , pcEnv = Just cfgEnv + } + + let msg = "Failed to run Haddock in no-compilation mode on test package '" ++ tpkgName tpkg ++ "'" + succeededNC <- waitForSuccess msg stdout =<< runProcess' cfgHaddockPath pc + + -- Generate documentation incrementally + if cfgSkipOneShot then pure (succeeded && succeededNC) else do + let oneShotDir = oneShotOutDir cfgDirConfig tpkg responseFile = hiDir "response-file" createEmptyDirectory oneShotDir - createEmptyDirectory hiDir - createEmptyDirectory hieDir writeFile responseFile $ escapeArgs [ "--odir=" ++ oneShotDir , "--optghc=-hidir=" ++ hiDir , "--optghc=-hiedir=" ++ hieDir ] - -- Build .hi files - let pc' = - processConfig - { pcArgs = - concat - [ - [ "--make" - , "-haddock" - , "-fwrite-interface" - , "-fwrite-ide-info" - , "-no-keep-o-files" - , "-hidir=" ++ hiDir - , "-hiedir=" ++ hieDir - ] - , tpkgFiles tpkg - ] - , pcEnv = Just cfgEnv - } - let msg = "Failed to run GHC on test package '" ++ tpkgName tpkg ++ "'" - _ <- waitForSuccess msg stdout =<< runProcess' cfgGhcPath pc' - files <- filter ((== ".hi") . takeExtension) <$> listDirectory hiDir -- Use the output order of GHC as a simple dependency order filesSorted <- Map.elems . Map.fromList <$> traverse (\file -> (,file) <$> getModificationTime (hiDir file)) files @@ -157,37 +183,30 @@ runHaddock cfg@(Config{..}) = do escapeArgs [ "--read-interface=" ++ srcRef ++ haddockFile ] loop files else pure False - succeeded2 <- loop filesSorted - when succeeded2 $ do + succeededOS <- loop filesSorted + when (succeededNC && succeededOS) $ do removeDirectoryRecursive hiDir removeDirectoryRecursive hieDir - pure succeeded2 + pure (succeeded && succeededNC && succeededOS) let somethingFailed = any not successes pure somethingFailed checkFile :: Config c -> FilePath -> IO CheckResult checkFile cfg file = do - hasRef <- doesFileExist $ refFile dcfg file - if hasRef - then do - mout <- readOut cfg file - mref <- readRef cfg file - case (mout, mref) of - (Just out, Just ref) - | ccfgEqual ccfg out ref -> - if cfgSkipOneShot cfg || dcfgCheckIgnoreOneShot (cfgDirConfig cfg) file - then return Pass - else do - mOneShotOut <- readOneShotOut cfg file - return $ case mOneShotOut of - Just oneShotOut - | ccfgEqual ccfg oneShotOut out -> Pass - | otherwise -> Fail - Nothing -> Error "Failed to parse one-shot input file" - | otherwise -> return Fail - _ -> return $ Error "Failed to parse input files" - else return NoRef + mref <- readRef cfg file + case mref of + Just ref -> do + let checkStep dcfgDir = ccfgEqual ccfg ref <$> readOut cfg dcfgDir file + result <- checkStep dcfgOutDir + resultNC <- if dcfgCheckIgnoreNoCompilation (cfgDirConfig cfg) file + then pure True + else checkStep dcfgNoCompilationOutDir + resultOS <- if cfgSkipOneShot cfg || dcfgCheckIgnoreOneShot (cfgDirConfig cfg) file + then pure True + else checkStep dcfgOneShotOutDir + pure $ if and [result, resultNC, resultOS] then Pass else Fail + Nothing -> return NoRef where ccfg = cfgCheckConfig cfg dcfg = cfgDirConfig cfg @@ -207,59 +226,50 @@ readRef cfg file = dcfg = cfgDirConfig cfg -- | Read (and clean) the test output artifact for a test -readOut :: Config c -> FilePath -> IO (Maybe c) -readOut cfg file = - fmap (ccfgClean ccfg file) . ccfgRead ccfg . BS.unpack - <$> BS.readFile (outFile dcfg file) - where - ccfg = cfgCheckConfig cfg - dcfg = cfgDirConfig cfg - -readOneShotOut :: Config c -> FilePath -> IO (Maybe c) -readOneShotOut cfg file = - fmap (ccfgClean ccfg file) . ccfgRead ccfg . BS.unpack - <$> BS.readFile (oneShotOutFile dcfg file) +readOut :: Config c -> (DirConfig -> FilePath) -> FilePath -> IO c +readOut cfg dcfgDir file = do + res <- fmap (ccfgClean ccfg file) . ccfgRead ccfg . BS.unpack + <$> BS.readFile outFile + case res of + Just out -> return out + Nothing -> error $ "Failed to parse output file: " ++ outFile where ccfg = cfgCheckConfig cfg dcfg = cfgDirConfig cfg + outFile = dcfgDir dcfg file diffFile :: Config c -> FilePath -> FilePath -> IO () diffFile cfg diff file = do - Just out <- readOut cfg file - Just oneShotOut <- readOneShotOut cfg file Just ref <- readRef cfg file - writeFile outFile' $ ccfgDump ccfg out - writeFile oneShotOutFile' $ ccfgDump ccfg oneShotOut - writeFile refFile' $ ccfgDump ccfg ref - - putStrLn $ "Diff for file \"" ++ file ++ "\":" - hFlush stdout - handle <- - runProcess' diff $ - processConfig - { pcArgs = [outFile', refFile'] - , pcStdOut = Just stdout - } - void $ waitForProcess handle - handle' <- - runProcess' diff $ - processConfig - { pcArgs = [oneShotOutFile', outFile'] - , pcStdOut = Just stdout - } - void $ waitForProcess handle' - return () + out <- readOut cfg dcfgOutDir file + noCompilationOut <- readOut cfg dcfgNoCompilationOutDir file + oneShotOut <- readOut cfg dcfgOneShotOutDir file + writeFile (dumpFile "ref") $ ccfgDump ccfg ref + writeFile (dumpFile "out") $ ccfgDump ccfg out + writeFile (dumpFile "oneShot") $ ccfgDump ccfg oneShotOut + writeFile (dumpFile "noCompilation") $ ccfgDump ccfg oneShotOut + + for_ ["out", "oneShot", "noCompilation"] $ \nm -> do + let outFile = dumpFile nm + refFile = dumpFile "ref" + putStrLn $ "Diff for file \"" ++ outFile ++ "\":" + hFlush stdout + handle <- + runProcess' diff $ + processConfig + { pcArgs = [outFile, refFile] + , pcStdOut = Just stdout + } + void $ waitForProcess handle where dcfg = cfgDirConfig cfg ccfg = cfgCheckConfig cfg - outFile' = outFile dcfg file <.> "dump" - oneShotOutFile' = oneShotOutFile dcfg file <.> "dump" - refFile' = outFile dcfg file <.> "ref" <.> "dump" + dumpFile nm = dcfgOutDir dcfg file <.> nm <.> "dump" maybeAcceptFile :: Config c -> FilePath -> CheckResult -> IO CheckResult maybeAcceptFile cfg file result | cfgAccept cfg && result `elem` [NoRef, Fail] = do - Just out <- readOut cfg file + out <- readOut cfg dcfgOutDir file let ref = refFile dcfg file createDirectoryIfMissing True (takeDirectory ref) writeFile ref $ ccfgDump ccfg out @@ -272,14 +282,11 @@ maybeAcceptFile _ _ result = pure result outDir :: DirConfig -> TestPackage -> FilePath outDir dcfg tpkg = dcfgOutDir dcfg tpkgName tpkg -oneshotOutDir :: DirConfig -> TestPackage -> FilePath -oneshotOutDir dcfg tpkg = dcfgOneShotOutDir dcfg tpkgName tpkg - -outFile :: DirConfig -> FilePath -> FilePath -outFile dcfg file = dcfgOutDir dcfg file +oneShotOutDir :: DirConfig -> TestPackage -> FilePath +oneShotOutDir dcfg tpkg = dcfgOneShotOutDir dcfg tpkgName tpkg -oneShotOutFile :: DirConfig -> FilePath -> FilePath -oneShotOutFile dcfg file = dcfgOneShotOutDir dcfg file +noCompilationOutDir :: DirConfig -> TestPackage -> FilePath +noCompilationOutDir dcfg tpkg = dcfgNoCompilationOutDir dcfg tpkgName tpkg refFile :: DirConfig -> FilePath -> FilePath refFile dcfg file = dcfgRefDir dcfg file ===================================== utils/haddock/haddock-test/src/Test/Haddock/Config.hs ===================================== @@ -4,7 +4,7 @@ module Test.Haddock.Config ( TestPackage(..), CheckConfig(..), DirConfig(..), Config(..) , defaultDirConfig - , cfgSrcDir, cfgRefDir, cfgOutDir, cfgResDir, cfgOneShotOutDir + , cfgSrcDir, cfgRefDir, cfgOutDir, cfgResDir, cfgOneShotOutDir, cfgNoCompilationOutDir , parseArgs, checkOpt, loadConfig ) where @@ -58,9 +58,11 @@ data DirConfig = DirConfig , dcfgRefDir :: FilePath , dcfgOutDir :: FilePath , dcfgOneShotOutDir :: FilePath + , dcfgNoCompilationOutDir :: FilePath , dcfgResDir :: FilePath , dcfgCheckIgnore :: FilePath -> Bool , dcfgCheckIgnoreOneShot :: FilePath -> Bool + , dcfgCheckIgnoreNoCompilation :: FilePath -> Bool } @@ -70,9 +72,11 @@ defaultDirConfig baseDir = DirConfig , dcfgRefDir = baseDir "ref" , dcfgOutDir = baseDir "out" , dcfgOneShotOutDir = baseDir "one-shot-out" + , dcfgNoCompilationOutDir = baseDir "no-compilation-out" , dcfgResDir = rootDir "resources" , dcfgCheckIgnore = const False , dcfgCheckIgnoreOneShot = const False + , dcfgCheckIgnoreNoCompilation = const False } where rootDir = baseDir ".." @@ -92,12 +96,13 @@ data Config c = Config } -cfgSrcDir, cfgRefDir, cfgOutDir, cfgResDir, cfgOneShotOutDir :: Config c -> FilePath +cfgSrcDir, cfgRefDir, cfgOutDir, cfgResDir, cfgOneShotOutDir, cfgNoCompilationOutDir :: Config c -> FilePath cfgSrcDir = dcfgSrcDir . cfgDirConfig cfgRefDir = dcfgRefDir . cfgDirConfig cfgOutDir = dcfgOutDir . cfgDirConfig cfgResDir = dcfgResDir . cfgDirConfig cfgOneShotOutDir = dcfgOneShotOutDir . cfgDirConfig +cfgNoCompilationOutDir = dcfgNoCompilationOutDir . cfgDirConfig ===================================== utils/haddock/latex-test/Main.hs ===================================== @@ -23,6 +23,7 @@ dirConfig = (defaultDirConfig $ takeDirectory __FILE__) { dcfgCheckIgnore = (`elem` ["haddock.sty", "main.tex"]) . takeFileName -- Just a discrepancy in output order , dcfgCheckIgnoreOneShot = (`elem` ["ConstructorArgs.tex"]) . takeFileName + , dcfgCheckIgnoreNoCompilation = (`elem` ["ConstructorArgs.tex"]) . takeFileName } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9c0ddce1eab8e4bbb8117328911cac63ae67a9ba...867611279614f3fd44f90c92404a7a08b7a0d0a0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9c0ddce1eab8e4bbb8117328911cac63ae67a9ba...867611279614f3fd44f90c92404a7a08b7a0d0a0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 12:57:12 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Tue, 03 Sep 2024 08:57:12 -0400 Subject: [Git][ghc/ghc][wip/andreask/setNonBlockingMode] Allow unknown fd device types for setNonBlockingMode. Message-ID: <66d707a8852aa_467c6c32d06226f@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/setNonBlockingMode at Glasgow Haskell Compiler / GHC Commits: 7c8f57f0 by Andreas Klebinger at 2024-09-03T14:38:27+02:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - 7 changed files: - libraries/base/changelog.md - libraries/ghc-internal/src/GHC/Internal/IO/FD.hs - libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/interface-stability/base-exports.stdout-ws-32 Changes: ===================================== libraries/base/changelog.md ===================================== @@ -16,6 +16,7 @@ * Add `firstA` and `secondA` to `Data.Bitraversable`. ([CLC proposal #172](https://github.com/haskell/core-libraries-committee/issues/172)) * Deprecate `GHC.TypeNats.Internal`, `GHC.TypeLits.Internal`, `GHC.ExecutionStack.Internal` ([CLC proposal #217](https://github.com/haskell/core-libraries-committee/issues/217)) * Define `Eq1`, `Ord1`, `Show1` and `Read1` instances for basic `Generic` representation types. ([CLC proposal #273](https://github.com/haskell/core-libraries-committee/issues/273)) + * `setNonBlockingMode` will no longer throw an exception when called on a FD associated with a unknown device type. ([CLC proposal #282](https://github.com/haskell/core-libraries-committee/issues/282)) * Add exception type metadata to default exception handler output. ([CLC proposal #231](https://github.com/haskell/core-libraries-committee/issues/231) and [CLC proposal #261](https://github.com/haskell/core-libraries-committee/issues/261)) ===================================== libraries/ghc-internal/src/GHC/Internal/IO/FD.hs ===================================== @@ -467,8 +467,8 @@ setNonBlockingMode fd set = do -- utilities inspecting fdIsNonBlocking (such as readRawBufferPtr) -- should not be tricked to think otherwise. is_nonblock <- if set then do - (fd_type, _, _) <- fdStat (fdFD fd) - pure $ fd_type /= RegularFile && fd_type /= RawDevice + fd_type <- statGetType_maybe (fdFD fd) + pure $ fd_type /= Just RegularFile && fd_type /= Just RawDevice else pure False setNonBlockingFD (fdFD fd) is_nonblock #if defined(mingw32_HOST_OS) ===================================== libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs ===================================== @@ -140,17 +140,30 @@ fdStat fd = fdType :: FD -> IO IODeviceType fdType fd = do (ty,_,_) <- fdStat fd; return ty +-- | Return a known device type or throw an exception if the device +-- type is unknown. statGetType :: Ptr CStat -> IO IODeviceType statGetType p_stat = do + dev_ty_m <- statGetType_maybe p_stat + case dev_ty_m of + Nothing -> ioError ioe_unknownfiletype + Just dev_ty -> pure dev_ty + +-- | Unlike @statGetType@, @statGetType_maybe@ will not throw an exception +-- if the CStat refers to a unknown device type. +-- +-- @since base-4.21.0.0 +statGetType_maybe :: Ptr CStat -> IO (Maybe IODeviceType) +statGetType_maybe p_stat = do c_mode <- st_mode p_stat :: IO CMode case () of - _ | s_isdir c_mode -> return Directory + _ | s_isdir c_mode -> return $ Just Directory | s_isfifo c_mode || s_issock c_mode || s_ischr c_mode - -> return Stream - | s_isreg c_mode -> return RegularFile + -> return $ Just Stream + | s_isreg c_mode -> return $ Just RegularFile -- Q: map char devices to RawDevice too? - | s_isblk c_mode -> return RawDevice - | otherwise -> ioError ioe_unknownfiletype + | s_isblk c_mode -> return $ Just RawDevice + | otherwise -> return Nothing ioe_unknownfiletype :: IOException ioe_unknownfiletype = IOError Nothing UnsupportedOperation "fdType" ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -10549,6 +10549,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10596,6 +10597,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs ===================================== @@ -13590,6 +13590,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -13637,6 +13638,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -10821,6 +10821,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10865,6 +10866,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.CWString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-ws-32 ===================================== @@ -10549,6 +10549,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10596,6 +10597,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7c8f57f012831f69e77205b3dde3a4751de08302 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7c8f57f012831f69e77205b3dde3a4751de08302 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 13:22:17 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 03 Sep 2024 09:22:17 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/romes/25110 Message-ID: <66d70d8937578_467c62b97d869662@gitlab.mail> Rodrigo Mesquita pushed new branch wip/romes/25110 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/romes/25110 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 13:47:38 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Tue, 03 Sep 2024 09:47:38 -0400 Subject: [Git][ghc/ghc][wip/fix-checkClosure] 331 commits: template-haskell: Move wired-ins to ghc-internal Message-ID: <66d7137ad1148_58dbc12a4a8361f4@gitlab.mail> Andreas Klebinger pushed to branch wip/fix-checkClosure at Glasgow Haskell Compiler / GHC Commits: 228dcae6 by Teo Camarasu at 2024-05-28T13:12:24+00:00 template-haskell: Move wired-ins to ghc-internal Thus we make `template-haskell` reinstallable and keep it as the public API for Template Haskell. All of the wired-in identifiers are moved to `ghc-internal`. This necessitates also moving much of `ghc-boot-th` into `ghc-internal`. These modules are then re-exported from `ghc-boot-th` and `template-haskell`. To avoid a dependency on `template-haskell` from `lib:ghc`, we instead depend on the TH ASTs via `ghc-boot-th`. As `template-haskell` no longer has special status, we can drop the logic adding an implicit dependency on `template-haskell` when using TH. We can also drop the `template-haskell-next` package, which was previously used when bootstrapping. When bootstrapping, we need to vendor the TH AST modules from `ghc-internal` into `ghc-boot-th`. This is controlled by the `bootstrap` cabal flag as before. See Note [Bootstrapping Template Haskell]. We split out a GHC.Internal.TH.Lift module resolving #24752. This module is only built when not bootstrapping. Resolves #24703 ------------------------- Metric Increase: ghc_boot_th_dir ghc_boot_th_so ------------------------- - - - - - 62dded28 by Teo Camarasu at 2024-05-28T13:12:24+00:00 testsuite: mark tests broken by #24886 Now that `template-haskell` is no longer wired-in. These tests are triggering #24886, and so need to be marked broken. - - - - - 3ca72ad9 by Cheng Shao at 2024-05-30T02:57:06-04:00 rts: fix missing function prototypes in ClosureMacros.h - - - - - e0029e3d by Andreas Klebinger at 2024-05-30T02:57:43-04:00 UnliftedFFITypes: Allow `(# #)` as argument when it's the only argument. This allows representing functions like: int foo(void); to be imported like this: foreign import ccall "a_number_c" c_number :: (# #) -> Int64# Which can be useful when the imported function isn't implicitly stateful. - - - - - d0401335 by Matthew Pickering at 2024-05-30T02:58:19-04:00 ci: Update ci-images commit for fedora38 image The fedora38 nightly job has been failing for quite a while because `diff` was no longer installed. The ci-images bump explicitly installs `diffutils` into these images so hopefully they now pass again. - - - - - 3c97c74a by Jan Hrček at 2024-05-30T02:58:58-04:00 Update exactprint docs - - - - - 77760cd7 by Jan Hrček at 2024-05-30T02:58:58-04:00 Incorporate review feedback - - - - - 87591368 by Jan Hrček at 2024-05-30T02:58:58-04:00 Remove no longer relevant reference to comments - - - - - 05f4f142 by Jan Hrček at 2024-05-30T02:58:59-04:00 Replace outdated code example - - - - - 45a4a5f3 by Andreas Klebinger at 2024-05-30T02:59:34-04:00 Reword error resulting from missing -XBangPatterns. It can be the result of either a bang pattern or strict binding, so now we say so instead of claiming it must be a bang pattern. Fixes #21032 - - - - - e17f2df9 by Cheng Shao at 2024-05-30T03:00:10-04:00 testsuite: bump MultiLayerModulesDefsGhciReload timeout to 10x - - - - - 7a660042 by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - c77a48af by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 05c4fafb by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 5ff83bfc by Sylvain Henry at 2024-05-30T14:43:10-04:00 JS: remove useless h$CLOCK_REALTIME (#23202) - - - - - 95ef2d58 by Matthew Pickering at 2024-05-30T14:43:47-04:00 ghcup-metadata: Fix metadata generation There were some syntax errors in the generation script which were preventing it from running. I have tested this with: ``` nix shell --extra-experimental-features nix-command -f .gitlab/rel_eng -c ghcup-metadata --metadata ghcup-0.0.7.yaml --date="2024-05-27" --pipeline-id=95534 --version=9.11.20240525 ``` which completed successfully. - - - - - 1bc66ee4 by Jakob Bruenker at 2024-05-30T14:44:22-04:00 Add diagrams to Arrows documentation This adds diagrams to the documentation of Arrows, similar to the ones found on https://www.haskell.org/arrows/. It does not add diagrams for ArrowChoice for the time being, mainly because it's not clear to me how to visually distinguish them from the ones for Arrow. Ideally, you might want to do something like highlight the arrows belonging to the same tuple or same Either in common colors, but that's not really possible with unicode. - - - - - d10a1c65 by Matthew Craven at 2024-05-30T23:35:48-04:00 Make UnsafeSNat et al. into pattern synonyms ...so that they do not cause coerce to bypass the nominal role on the corresponding singleton types when they are imported. See Note [Preventing unsafe coercions for singleton types] and the discussion at #23478. This also introduces unsafeWithSNatCo (and analogues for Char and Symbol) so that users can still access the dangerous coercions that importing the real constructors would allow, but only in a very localized way. - - - - - 0958937e by Cheng Shao at 2024-05-30T23:36:25-04:00 hadrian: build C/C++ with split sections when enabled When split sections is enabled, ensure -fsplit-sections is passed to GHC as well when invoking GHC to compile C/C++; and pass -ffunction-sections -fdata-sections to gcc/clang when compiling C/C++ with the hadrian Cc builder. Fixes #23381. - - - - - 02b1f91e by Cheng Shao at 2024-05-30T23:36:25-04:00 driver: build C/C++ with -ffunction-sections -fdata-sections when split sections is enabled When -fsplit-sections is passed to GHC, pass -ffunction-sections -fdata-sections to gcc/clang when building C/C++. Previously, -fsplit-sections was only respected by the NCG/LLVM backends, but not the unregisterised backend; the GHC driver did not pass -fdata-sections and -ffunction-sections to the C compiler, which resulted in excessive executable sizes. Fixes #23381. ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - fd47e2e3 by Cheng Shao at 2024-05-30T23:37:00-04:00 testsuite: mark process005 as fragile on JS - - - - - 34a04ea1 by Matthew Pickering at 2024-05-31T06:08:36-04:00 Add -Wderiving-typeable to -Wall Deriving `Typeable` does nothing, and it hasn't done for a long while. There has also been a warning for a long while which warns you about uselessly deriving it but it wasn't enabled in -Wall. Fixes #24784 - - - - - 75fa7b0b by Matthew Pickering at 2024-05-31T06:08:36-04:00 docs: Fix formatting of changelog entries - - - - - 303c4b33 by Preetham Gujjula at 2024-05-31T06:09:21-04:00 docs: Fix link to injective type families paper Closes #24863 - - - - - df97e9a6 by Ben Gamari at 2024-05-31T06:09:57-04:00 ghc-internal: Fix package description The previous description was inherited from `base` and was inappropriate for `ghc-internal`. Also fix the maintainer and bug reporting fields. Closes #24906. - - - - - bf0737c0 by Cheng Shao at 2024-05-31T06:10:33-04:00 compiler: remove ArchWasm32 special case in cmmDoCmmSwitchPlans This patch removes special consideration for ArchWasm32 in cmmDoCmmSwitchPlans, which means the compiler will now disable cmmImplementSwitchPlans for wasm unreg backend, just like unreg backend of other targets. We enabled it in the past to workaround some compile-time panic in older versions of LLVM, but those panics are no longer present, hence no need to keep this workaround. - - - - - 7eda4bd2 by Cheng Shao at 2024-05-31T15:52:04-04:00 utils: add hie.yaml config file for ghc-config Add hie.yaml to ghc-config project directory so it can be edited using HLS. - - - - - 1e5752f6 by Cheng Shao at 2024-05-31T15:52:05-04:00 hadrian: handle findExecutable "" gracefully hadrian may invoke findExecutable "" at run-time due to a certain program is not found by configure script. Which is fine and findExecutable is supposed to return Nothing in this case. However, on Windows there's a directory bug that throws an exception (see https://github.com/haskell/directory/issues/180), so we might as well use a wrapper for findExecutable and handle exceptions gracefully. - - - - - 4eb5ad09 by Cheng Shao at 2024-05-31T15:52:05-04:00 configure: do not set LLC/OPT/LLVMAS fallback values when FIND_LLVM_PROG fails When configure fails to find LLC/OPT/LLVMAS within supported version range, it used to set "llc"/"opt"/"clang" as fallback values. This behavior is particularly troublesome when the user has llc/opt/clang with other versions in their PATH and run the testsuite, since hadrian will incorrectly assume have_llvm=True and pass that to the testsuite driver, resulting in annoying optllvm test failures (#23186). If configure determines llc/opt/clang wouldn't work, then we shouldn't pretend it'll work at all, and the bindist configure will invoke FIND_LLVM_PROG check again at install time anyway. - - - - - 5f1afdf7 by Sylvain Henry at 2024-05-31T15:52:52-04:00 Introduce UniqueSet and use it to replace 'UniqSet Unique' 'UniqSet Unique' represents a set of uniques as a 'Map Unique Unique', which is wasting space (associated key/value are always the same). Fix #23572 and #23605 - - - - - e0aa42b9 by crumbtoo at 2024-05-31T15:53:33-04:00 Improve template-haskell haddocks Closes #15822 - - - - - ae170155 by Olivier Benz at 2024-06-01T09:35:17-04:00 Bump max LLVM version to 19 (not inclusive) - - - - - 92aa65ea by Matthew Pickering at 2024-06-01T09:35:17-04:00 ci: Update CI images to test LLVM 18 The debian12 image in this commit has llvm 18 installed. - - - - - adb1fe42 by Serge S. Gulin at 2024-06-01T09:35:53-04:00 Unicode: make ucd2haskell build-able again ucd2haskell tool used streamly library which version in cabal was out of date. It is updated to the latest version at hackage with deprecated parts rewritten. Also following fixes were applied to existing code in suppose that from its last run the code structure was changed and now it was required to be up to date with actual folder structures: 1. Ghc module path environment got a suffix with `src`. 2. Generated code got 2.1 `GHC.Internal` prefix for `Data.*`. 2.2 `GHC.Unicode.Internal` swapped on `GHC.Internal.Unicode` according to actual structure. - - - - - ad56fd84 by Jade at 2024-06-01T09:36:29-04:00 Replace 'NB' with 'Note' in error messages - - - - - 6346c669 by Cheng Shao at 2024-06-01T09:37:04-04:00 compiler: fix -ddump-cmm-raw when compiling .cmm This patch fixes missing -ddump-cmm-raw output when compiling .cmm, which is useful for debugging cmm related codegen issues. - - - - - 1c834ad4 by Ryan Scott at 2024-06-01T09:37:40-04:00 Print namespace specifiers in FixitySig's Outputable instance For whatever reason, the `Outputable` instance for `FixitySig` simply did not print out namespace specifiers, leading to the confusing `-ddump-splices` output seen in #24911. This patch corrects this oversight. Fixes #24911. - - - - - cf49fb5f by Sylvain Henry at 2024-06-01T09:38:19-04:00 Configure: display C++ compiler path - - - - - f9c1ae12 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable PIC for in-tree GMP on wasm32 This patch disables PIC for in-tree GMP on wasm32 target. Enabling PIC unconditionally adds undesired code size and runtime overhead for wasm32. - - - - - 1a32f828 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable in-tree gmp fft code path for wasm32 This patch disables in-tree GMP FFT code paths for wasm32 target in order to give up some performance of multiplying very large operands in exchange for reduced code size. - - - - - 06277d56 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: build in-tree GMP with malloc-notreentrant on wasm32 This patch makes hadrian build in-tree GMP with the --enable-alloca=malloc-notreentrant configure option. We will only need malloc-reentrant when we have threaded RTS and SMP support on wasm32, which will take some time to happen, before which we should use malloc-notreentrant to avoid undesired runtime overhead. - - - - - 9f614270 by ARATA Mizuki at 2024-06-02T14:02:35-04:00 Set package include paths when assembling .S files Fixes #24839. Co-authored-by: Sylvain Henry <hsyl20 at gmail.com> - - - - - 4998a6ed by Alex Mason at 2024-06-03T02:09:29-04:00 Improve performance of genericWordQuotRem2Op (#22966) Implements the algorithm from compiler-rt's udiv128by64to64default. This rewrite results in a roughly 24x improvement in runtime on AArch64 (and likely any other arch that uses it). - - - - - ae50a8eb by Cheng Shao at 2024-06-03T02:10:05-04:00 testsuite: mark T7773 as fragile on wasm - - - - - c8ece0df by Fendor at 2024-06-03T19:43:22-04:00 Migrate `Finder` component to `OsPath`, fixed #24616 For each module in a GHCi session, we keep alive one `ModLocation`. A `ModLocation` is fairly inefficiently packed, as `String`s are expensive in memory usage. While benchmarking the agda codebase, we concluded that we keep alive around 11MB of `FilePath`'s, solely retained by `ModLocation`. We provide a more densely packed encoding of `ModLocation`, by moving from `FilePath` to `OsPath`. Further, we migrate the full `Finder` component to `OsPath` to avoid unnecessary transformations. As the `Finder` component is well-encapsulated, this requires only a minimal amount of changes in other modules. We introduce pattern synonym for 'ModLocation' which maintains backwards compatibility and avoids breaking consumers of 'ModLocation'. - - - - - 0cff083a by Cheng Shao at 2024-06-03T19:43:58-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. - - - - - 18f63970 by Sebastian Graf at 2024-06-04T05:05:27-04:00 Parser: Remove unused `apats` rule - - - - - 38757c30 by David Knothe at 2024-06-04T05:05:27-04:00 Implement Or Patterns (#22596) This commit introduces a new language extension, `-XOrPatterns`, as described in GHC Proposal 522. An or-pattern `pat1; ...; patk` succeeds iff one of the patterns `pat1`, ..., `patk` succeed, in this order. See also the summary `Note [Implmentation of OrPatterns]`. Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> - - - - - 395412e8 by Cheng Shao at 2024-06-04T05:06:04-04:00 compiler/ghci/rts: remove stdcall support completely We have formally dropped i386 windows support (#18487) a long time ago. The stdcall foreign call convention is only used by i386 windows, and the legacy logic around it is a significant maintenance burden for future work that adds arm64 windows support (#24603). Therefore, this patch removes stdcall support completely from the compiler as well as the RTS (#24883): - stdcall is still recognized as a FFI calling convention in Haskell syntax. GHC will now unconditionally emit a warning (-Wunsupported-calling-conventions) and treat it as ccall. - Apart from minimum logic to support the parsing and warning logic, all other code paths related to stdcall has been completely stripped from the compiler. - ghci only supports FFI_DEFAULT_ABI and ccall convention from now on. - FFI foreign export adjustor code on all platforms no longer handles the stdcall case and only handles ccall from now on. - The Win32 specific parts of RTS no longer has special code paths for stdcall. This commit is the final nail on the coffin for i386 windows support. Further commits will perform more housecleaning to strip the legacy code paths and pave way for future arm64 windows support. - - - - - d1fe9ab6 by Cheng Shao at 2024-06-04T05:06:04-04:00 rts: remove legacy i386 windows code paths This commit removes some legacy i386 windows related code paths in the RTS, given this target is no longer supported. - - - - - a605e4b2 by Cheng Shao at 2024-06-04T05:06:04-04:00 autoconf: remove i386 windows related logic This commit removes legacy i386 windows logic in autoconf scripts. - - - - - 91e5ac5e by Cheng Shao at 2024-06-04T05:06:04-04:00 llvm-targets: remove i386 windows support This commit removes i386 windows from llvm-targets and the script to generate it. - - - - - 65fe75a4 by Cheng Shao at 2024-06-04T05:06:04-04:00 libraries/utils: remove stdcall related legacy logic This commit removes stdcall related legacy logic in libraries and utils. ccall should be used uniformly for all supported windows hosts from now on. - - - - - d2a83302 by Cheng Shao at 2024-06-04T05:06:04-04:00 testsuite: adapt the testsuite for stdcall removal This patch adjusts test cases to handle the stdcall removal: - Some stdcall usages are replaced with ccall since stdcall doesn't make sense anymore. - We also preserve some stdcall usages, and check in the expected warning messages to ensure GHC always warn about stdcall usages (-Wunsupported-calling-conventions) as expected. - Error code testsuite coverage is slightly improved, -Wunsupported-calling-conventions is now tested. - Obsolete code paths related to i386 windows are also removed. - - - - - cef8f47a by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: minor adjustments for stdcall removal This commit include minor adjustments of documentation related to stdcall removal. - - - - - 54332437 by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: mention i386 Windows removal in 9.12 changelog This commit mentions removal of i386 Windows support and stdcall related change in the 9.12 changelog. - - - - - 2aaea8a1 by Cheng Shao at 2024-06-04T05:06:40-04:00 hadrian: improve user settings documentation This patch adds minor improvements to hadrian user settings documentation: - Add missing `ghc.cpp.opts` case - Remove non-existent `cxx` case - Clarify `cc.c.opts` also works for C++, while `cc.deps.opts` doesn't - Add example of passing configure argument to autoconf packages - - - - - 71010381 by Alex Mason at 2024-06-04T12:09:07-04:00 Add AArch64 CLZ, CTZ, RBIT primop implementations. Adds support for emitting the clz and rbit instructions, which are used by GHC.Prim.clz*#, GHC.Prim.ctz*# and GHC.Prim.bitReverse*#. - - - - - 44e2abfb by Cheng Shao at 2024-06-04T12:09:43-04:00 hadrian: add +text_simdutf flavour transformer to allow building text with simdutf This patch adds a +text_simdutf flavour transformer to hadrian to allow downstream packagers and users that build from source to opt-in simdutf support for text, in order to benefit from SIMD speedup at run-time. It's still disabled by default for the time being. - - - - - 077cb2e1 by Cheng Shao at 2024-06-04T12:09:43-04:00 ci: enable +text_simdutf flavour transformer for wasm jobs This commit enables +text_simdutf flavour transformer for wasm jobs, so text is now built with simdutf support for wasm. - - - - - b23746ad by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Use TemplateHaskellQuotes in instance Lift ByteArray Resolves #24852 - - - - - 3fd25743 by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Mark addrToByteArray as NOINLINE This function should never be inlined in order to keep code size small. - - - - - 98ad1ea5 by Cheng Shao at 2024-06-04T22:51:26-04:00 compiler: remove unused CompilerInfo/LinkerInfo types This patch removes CompilerInfo/LinkerInfo types from the compiler since they aren't actually used anywhere. - - - - - 11795244 by Cheng Shao at 2024-06-05T06:33:17-04:00 rts: remove unused PowerPC/IA64 native adjustor code This commit removes unused PowerPC/IA64 native adjustor code which is never actually enabled by autoconf/hadrian. Fixes #24920. - - - - - 5132754b by Sylvain Henry at 2024-06-05T06:33:57-04:00 RTS: fix warnings with doing*Profiling (#24918) - - - - - accc8c33 by Cheng Shao at 2024-06-05T11:35:36-04:00 hadrian: don't depend on inplace/mingw when --enable-distro-toolchain on Windows - - - - - 6ffbd678 by Cheng Shao at 2024-06-05T11:35:37-04:00 autoconf: normalize paths of some build-time dependencies on Windows This commit applies path normalization via cygpath -m to some build-time dependencies on Windows. Without this logic, the /clang64/bin prefixed msys2-style paths cause the build to fail with --enable-distro-toolchain. - - - - - 075dc6d4 by Cheng Shao at 2024-06-05T11:36:12-04:00 hadrian: remove OSDarwin mention from speedHack This commit removes mentioning of OSDarwin from speedHack, since speedHack is purely for i386 and we no longer support i386 darwin (#24921). - - - - - 83235c4c by Cheng Shao at 2024-06-05T11:36:12-04:00 compiler: remove 32-bit darwin logic This commit removes all 32-bit darwin logic from the compiler, given we no longer support 32-bit apple systems (#24921). Also contains a bit more cleanup of obsolete i386 windows logic. - - - - - 1eb99bc3 by Cheng Shao at 2024-06-05T11:36:12-04:00 rts: remove 32-bit darwin/ios logic This commit removes 32-bit darwin/ios related logic from the rts, given we no longer support them (#24921). - - - - - 24f65892 by Cheng Shao at 2024-06-05T11:36:12-04:00 llvm-targets: remove 32-bit darwin/ios targets This commit removes 32-bit darwin/ios targets from llvm-targets given we no longer support them (#24921). - - - - - ccdbd689 by Cheng Shao at 2024-06-05T11:36:12-04:00 testsuite: remove 32-bit darwin logic This commit removes 32-bit darwin logic from the testsuite given it's no longer supported (#24921). Also contains more cleanup of obsolete i386 windows logic. - - - - - 11d661c4 by Cheng Shao at 2024-06-05T11:36:13-04:00 docs: mention 32-bit darwin/ios removal in 9.12 changelog This commit mentions removal of 32-bit darwin/ios support (#24921) in the 9.12 changelog. - - - - - 7c173310 by Georgi Lyubenov at 2024-06-05T15:17:22-04:00 Add firstA and secondA to Data.Bitraversable Please see https://github.com/haskell/core-libraries-committee/issues/172 for related discussion - - - - - 3b6f9fd1 by Ben Gamari at 2024-06-05T15:17:59-04:00 base: Fix name of changelog Fixes #24899. Also place it under `extra-doc-files` to better reflect its nature and avoid triggering unnecessary recompilation if it changes. - - - - - 1f4d2ef7 by Sebastian Graf at 2024-06-05T15:18:34-04:00 Announce Or-patterns in the release notes for GHC 9.12 (#22596) Leftover from !9229. - - - - - 8650338d by Jan Hrček at 2024-06-06T10:39:24-04:00 Improve haddocks of Language.Haskell.Syntax.Pat.Pat - - - - - 2eee65e1 by Cheng Shao at 2024-06-06T10:40:00-04:00 testsuite: bump T7653 timeout for wasm - - - - - 990fed60 by Sylvain Henry at 2024-06-07T14:45:23-04:00 StgToCmm: refactor opTranslate and friends - Change arguments order to avoid `\args -> ...` lambdas - Fix documentation - Rename StgToCmm options ("big" doesn't mean anything) - - - - - 1afad514 by Sylvain Henry at 2024-06-07T14:45:23-04:00 NCG x86: remove dead code (#5444) Since 6755d833af8c21bbad6585144b10e20ac4a0a1ab this code is dead. - - - - - 595c0894 by Cheng Shao at 2024-06-07T14:45:58-04:00 testsuite: skip objc-hi/objcxx-hi when cross compiling objc-hi/objcxx-hi should be skipped when cross compiling. The existing opsys('darwin') predicate only asserts the host system is darwin but tells us nothing about the target, hence the oversight. - - - - - edfe6140 by qqwy at 2024-06-08T11:23:54-04:00 Replace '?callStack' implicit param with HasCallStack in GHC.Internal.Exception.throw - - - - - 35a64220 by Cheng Shao at 2024-06-08T11:24:30-04:00 rts: cleanup inlining logic This patch removes pre-C11 legacy code paths related to INLINE_HEADER/STATIC_INLINE/EXTERN_INLINE macros, ensure EXTERN_INLINE is treated as static inline in most cases (fixes #24945), and also corrects the comments accordingly. - - - - - 9ea90ed2 by Andrew Lelechenko at 2024-06-08T11:25:06-04:00 CODEOWNERS: add @core-libraries to track base interface changes A low-tech tactical solution for #24919 - - - - - 580fef7b by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update CHANGELOG to reflect current version - - - - - 391ecff5 by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update prologue.txt to reflect package description - - - - - 3dca3b7d by Ben Gamari at 2024-06-09T01:27:57-04:00 compiler: Clarify comment regarding need for MOVABS The comment wasn't clear in stating that it was only applicable to immediate source and memory target operands. - - - - - 6bd850e8 by doyougnu at 2024-06-09T21:02:14-04:00 JS: establish single source of truth for symbols In pursuit of: #22736. This MR moves ad-hoc symbols used throughout the js backend into a single symbols file. Why? First, this cleans up the code by removing ad-hoc strings created on the fly and therefore makes the code more maintainable. Second, it makes it much easier to eventually type these identifiers. - - - - - f3017dd3 by Cheng Shao at 2024-06-09T21:02:49-04:00 rts: replace ad-hoc MYTASK_USE_TLV with proper CC_SUPPORTS_TLS This patch replaces the ad-hoc `MYTASK_USE_TLV` with the `CC_SUPPORTS_TLS` macro. If TLS support is detected by autoconf, then we should use that for managing `myTask` in the threaded RTS. - - - - - e17d7e8c by Ben Gamari at 2024-06-11T05:25:21-04:00 users-guide: Fix stylistic issues in 9.12 release notes - - - - - 8a8a982a by Hugo Peters at 2024-06-11T05:25:57-04:00 fix typo in the simplifier debug output: baling -> bailing - - - - - 16475bb8 by Hécate Moonlight at 2024-06-12T03:07:55-04:00 haddock: Correct the Makefile to take into account Darwin systems - - - - - a2f60da5 by Hécate Kleidukos at 2024-06-12T03:08:35-04:00 haddock: Remove obsolete links to github.com/haskell/haddock in the docs - - - - - de4395cd by qqwy at 2024-06-12T03:09:12-04:00 Add `__GLASGOW_HASKELL_ASSERTS_IGNORED__` as CPP macro name if `-fasserts-ignored is set. This allows users to create their own Control.Exception.assert-like functionality that does something other than raising an `AssertFailed` exception. Fixes #24967 - - - - - 0e9c4dee by Ryan Hendrickson at 2024-06-12T03:09:53-04:00 compiler: add hint to TcRnBadlyStaged message - - - - - 2747cd34 by Simon Peyton Jones at 2024-06-12T12:51:37-04:00 Fix a QuickLook bug This MR fixes the bug exposed by #24676. The problem was that quickLookArg was trying to avoid calling tcInstFun unnecessarily; but it was in fact necessary. But that in turn forced me into a significant refactoring, putting more fields into EValArgQL. Highlights: see Note [Quick Look overview] in GHC.Tc.Gen.App * Instantiation variables are now distinguishable from ordinary unification variables, by level number = QLInstVar. This is treated like "level infinity". See Note [The QLInstVar TcLevel] in GHC.Tc.Utils.TcType. * In `tcApp`, we don't track the instantiation variables in a set Delta any more; instead, we just tell them apart by their level number. * EValArgQL now much more clearly captures the "half-done" state of typechecking an argument, ready for later resumption. See Note [Quick Look at value arguments] in GHC.Tc.Gen.App * Elminated a bogus (never used) fast-path in GHC.Tc.Utils.Instantiate.instCallConstraints See Note [Possible fast path for equality constraints] Many other small refactorings. - - - - - 1b1523b1 by George Thomas at 2024-06-12T12:52:18-04:00 Fix non-compiling extensible record `HasField` example - - - - - 97b141a3 by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Fix hyperlinker source urls (#24907) This fixes a bug introduced by f56838c36235febb224107fa62334ebfe9941aba Links to external modules in the hyperlinker are uniformly generated using splicing the template given to us instead of attempting to construct the url in an ad-hoc manner. - - - - - 954f864c by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Add name anchor to external source urls from documentation page URLs for external source links from documentation pages were missing a splice location for the name. Fixes #24912 - - - - - b0b64177 by Simon Peyton Jones at 2024-06-12T12:53:31-04:00 Prioritise nominal equalities The main payload of this patch is * Prioritise nominal equalities in the constraint solver. This ameliorates the incompleteness of solving for representational constraints over newtypes: see #24887. See (EX2) in Note [Decomposing newtype equalities] in GHC.Tc.Solver.Equality In doing this patch I tripped over some other things that I refactored: * Move `isCoVarType` from `GHC.Core.Type` to `GHC.Core.Predicate` where it seems more at home. * Clarify the "rewrite role" of a constraint. I was very puzzled about what the role of, say `(Eq a)` might be, but see the new Note [The rewrite-role of a constraint]. In doing so I made predTypeEqRel crash when given a non-equality. Usually it expects an equality; but it was being mis-used for the above rewrite-role stuff. - - - - - cb7c1b83 by Liam Goodacre at 2024-06-12T12:54:09-04:00 compiler: missing-deriving-strategies suggested fix Extends the missing-deriving-strategies warning with a suggested fix that includes which deriving strategies were assumed. For info about the warning, see comments for `TcRnNoDerivStratSpecified`, `TcRnNoDerivingClauseStrategySpecified`, & `TcRnNoStandaloneDerivingStrategySpecified`. For info about the suggested fix, see `SuggestExplicitDerivingClauseStrategies` & `SuggestExplicitStandalanoDerivingStrategy`. docs: Rewords missing-deriving-strategies to mention the suggested fix. Resolves #24955 - - - - - 4e36d3a3 by Jan Hrček at 2024-06-12T12:54:48-04:00 Further haddocks improvements in Language.Haskell.Syntax.Pat.Pat - - - - - 558353f4 by Cheng Shao at 2024-06-12T12:55:24-04:00 rts: use page sized mblocks on wasm This patch changes mblock size to page size on wasm. It allows us to simplify our wasi-libc fork, makes it much easier to test third party libc allocators like emmalloc/mimalloc, as well as experimenting with threaded RTS in wasm. - - - - - b3cc5366 by Matthew Pickering at 2024-06-12T23:06:57-04:00 compiler: Make ghc-experimental not wired in If you need to wire in definitions, then place them in ghc-internal and reexport them from ghc-experimental. Ticket #24903 - - - - - 700eeab9 by Hécate Kleidukos at 2024-06-12T23:07:37-04:00 base: Use a more appropriate unicode arrow for the ByteArray diagram This commit rectifies the usage of a unicode arrow in favour of one that doesn't provoke mis-alignment. - - - - - cca7de25 by Matthew Pickering at 2024-06-12T23:08:14-04:00 ghcup-metadata: Fix debian version ranges This was caught by `ghcup-ci` failing and attempting to install a deb12 bindist on deb11. ``` configure: WARNING: m4/prep_target_file.m4: Expecting YES/NO but got in ArSupportsDashL_STAGE0. Defaulting to False. bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by bin/ghc-toolchain-bin) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) ``` Fixes #24974 - - - - - 7b23ce8b by Pierre Le Marre at 2024-06-13T15:35:04-04:00 ucd2haskell: remove Streamly dependency + misc - Remove dead code. - Remove `streamly` dependency. - Process files with `bytestring`. - Replace Unicode files parsers with the corresponding ones from the package `unicode-data-parser`. - Simplify cabal file and rename module - Regenerate `ghc-internal` Unicode files with new header - - - - - 4570319f by Jacco Krijnen at 2024-06-13T15:35:41-04:00 Document how to run haddocks tests (#24976) Also remove ghc 9.7 requirement - - - - - fb629e24 by amesgen at 2024-06-14T00:28:20-04:00 compiler: refactor lower_CmmExpr_Ptr - - - - - def46c8c by amesgen at 2024-06-14T00:28:20-04:00 compiler: handle CmmRegOff in lower_CmmExpr_Ptr - - - - - ce76bf78 by Simon Peyton Jones at 2024-06-14T00:28:56-04:00 Small documentation update in Quick Look - - - - - 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - e31963ae by Cheng Shao at 2024-09-03T13:47:28+00:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - 23 changed files: - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - CODEOWNERS - compiler/GHC.hs - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/PrimOps.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/Utils.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/ByteCode/Types.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/639aa02898fe4c135ea98f987bc8912aa37c1994...e31963ae797633a47fee69f2e17bdf79562959c9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/639aa02898fe4c135ea98f987bc8912aa37c1994...e31963ae797633a47fee69f2e17bdf79562959c9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 14:20:32 2024 From: gitlab at gitlab.haskell.org (Sylvain Henry (@hsyl20)) Date: Tue, 03 Sep 2024 10:20:32 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-0006-Adds-support-for-Hidden-symbols] RTS linker: add support for hidden symbols (#25191) Message-ID: <66d71b30d1a47_58dbc56c458543ba@gitlab.mail> Sylvain Henry pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-0006-Adds-support-for-Hidden-symbols at Glasgow Haskell Compiler / GHC Commits: f3b98dbc by doyougnu at 2024-09-03T16:19:57+02:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 11 changed files: - rts/Linker.c - rts/LinkerInternals.h - rts/linker/Elf.c - rts/linker/ElfTypes.h - rts/linker/PEi386.c - testsuite/tests/rts/linker/Makefile - + testsuite/tests/rts/linker/T25191.hs - + testsuite/tests/rts/linker/T25191.stdout - + testsuite/tests/rts/linker/T25191_foo1.c - + testsuite/tests/rts/linker/T25191_foo2.c - testsuite/tests/rts/linker/all.T Changes: ===================================== rts/Linker.c ===================================== @@ -232,11 +232,11 @@ static void ghciRemoveSymbolTable(StrHashTable *table, const SymbolName* key, static const char * symbolTypeString (SymType type) { - switch (type & ~SYM_TYPE_DUP_DISCARD) { + switch (type & ~(SYM_TYPE_DUP_DISCARD | SYM_TYPE_HIDDEN)) { case SYM_TYPE_CODE: return "code"; case SYM_TYPE_DATA: return "data"; case SYM_TYPE_INDIRECT_DATA: return "indirect-data"; - default: barf("symbolTypeString: unknown symbol type"); + default: barf("symbolTypeString: unknown symbol type (%d)", type); } } @@ -283,10 +283,19 @@ int ghciInsertSymbolTable( } else if (pinfo->type ^ type) { + if(pinfo->type & SYM_TYPE_HIDDEN) + { + /* The existing symbol is hidden, let's replace it */ + pinfo->value = data; + pinfo->owner = owner; + pinfo->strength = strength; + pinfo->type = type; + return 1; + } /* We were asked to discard the symbol on duplicates, do so quietly. */ - if (!(type & SYM_TYPE_DUP_DISCARD)) + if (!(type & (SYM_TYPE_DUP_DISCARD | SYM_TYPE_HIDDEN))) { - debugBelch("Symbol type mismatch.\n"); + debugBelch("Symbol type mismatch (existing %d, new %d).\n", pinfo->type, type); debugBelch("Symbol %s was defined by %" PATH_FMT " to be a %s symbol.\n", key, obj_name, symbolTypeString(type)); debugBelch(" yet was defined by %" PATH_FMT " to be a %s symbol.\n", ===================================== rts/LinkerInternals.h ===================================== @@ -64,6 +64,8 @@ typedef enum _SymType { SYM_TYPE_DUP_DISCARD = 1 << 3, /* the symbol is a symbol in a BFD import library however if a duplicate is found with a mismatching SymType then discard this one. */ + SYM_TYPE_HIDDEN = 1 << 4, /* the symbol is hidden and should not be exported */ + } SymType; ===================================== rts/linker/Elf.c ===================================== @@ -1073,6 +1073,9 @@ ocGetNames_ELF ( ObjectCode* oc ) } else { sym_type = SYM_TYPE_DATA; } + if(ELF_ST_VISIBILITY(symbol->elf_sym->st_other) == STV_HIDDEN) { + sym_type |= SYM_TYPE_HIDDEN; + } /* And the decision is ... */ ===================================== rts/linker/ElfTypes.h ===================================== @@ -33,6 +33,9 @@ #define Elf_Sym Elf64_Sym #define Elf_Rel Elf64_Rel #define Elf_Rela Elf64_Rela +#if !defined(ELF_ST_VISIBILITY) +#define ELF_ST_VISIBILITY ELF64_ST_VISIBILITY +#endif #if !defined(ELF_ST_TYPE) #define ELF_ST_TYPE ELF64_ST_TYPE #endif @@ -57,6 +60,9 @@ #define Elf_Sym Elf32_Sym #define Elf_Rel Elf32_Rel #define Elf_Rela Elf32_Rela +#if !defined(ELF_ST_VISIBILITY) +#define ELF_ST_VISIBILITY ELF32_ST_VISIBILITY +#endif /* ELF_ST_VISIBILITY */ #if !defined(ELF_ST_TYPE) #define ELF_ST_TYPE ELF32_ST_TYPE #endif /* ELF_ST_TYPE */ ===================================== rts/linker/PEi386.c ===================================== @@ -1891,6 +1891,9 @@ ocGetNames_PEi386 ( ObjectCode* oc ) sname[size-start]='\0'; stgFree(tmp); sname = strdup (sname); + if(secNumber == IMAGE_SYM_UNDEFINED) + type |= SYM_TYPE_HIDDEN; + if (!ghciInsertSymbolTable(oc->fileName, symhash, sname, addr, false, type, oc)) return false; @@ -1905,6 +1908,8 @@ ocGetNames_PEi386 ( ObjectCode* oc ) && (!section || (section && section->kind != SECTIONKIND_IMPORT))) { /* debugBelch("addSymbol %p `%s' Weak:%lld \n", addr, sname, isWeak); */ sname = strdup (sname); + if(secNumber == IMAGE_SYM_UNDEFINED) + type |= SYM_TYPE_HIDDEN; IF_DEBUG(linker_verbose, debugBelch("addSymbol %p `%s'\n", addr, sname)); ASSERT(i < (uint32_t)oc->n_symbols); oc->symbols[i].name = sname; @@ -1936,7 +1941,7 @@ static size_t makeSymbolExtra_PEi386( ObjectCode* oc, uint64_t index STG_UNUSED, size_t s, char* symbol STG_UNUSED, SymType type ) { SymbolExtra *extra; - switch(type & ~SYM_TYPE_DUP_DISCARD) { + switch(type & ~(SYM_TYPE_DUP_DISCARD | SYM_TYPE_HIDDEN)) { case SYM_TYPE_CODE: { // jmp *-14(%rip) extra = m32_alloc(oc->rx_m32, sizeof(SymbolExtra), 8); ===================================== testsuite/tests/rts/linker/Makefile ===================================== @@ -145,3 +145,10 @@ reloc-none: "$(TEST_HC)" load-object.c -o load-object -no-hs-main -debug "$(TEST_HC)" -c reloc-none.c -o reloc-none.o ./load-object reloc-none.o + +.PHONY: T25191 +T25191: + "$(TEST_HC)" -c T25191_foo1.c -o foo1.o -v0 + "$(TEST_HC)" -c T25191_foo2.c -o foo2.o -v0 + "$(TEST_HC)" T25191.hs -v0 + ./T25191 ===================================== testsuite/tests/rts/linker/T25191.hs ===================================== @@ -0,0 +1,29 @@ +{-# LANGUAGE ForeignFunctionInterface, CPP #-} +import Foreign.C.String +import Control.Monad +import System.FilePath +import Foreign.Ptr + +-- Type of paths is different on Windows +#if defined(mingw32_HOST_OS) +type PathString = CWString +withPathString = withCWString +#else +type PathString = CString +withPathString = withCString +#endif + +main = do + initLinker + r1 <- withPathString "foo1.o" loadObj + when (r1 /= 1) $ error "loadObj failed" + r2 <- withPathString "foo2.o" loadObj + when (r2 /= 1) $ error "loadObj failed" + r <- resolveObjs + when (r /= 1) $ error "resolveObj failed" + putStrLn "success" + +foreign import ccall "initLinker" initLinker :: IO () +foreign import ccall "addDLL" addDLL :: PathString -> IO CString +foreign import ccall "loadObj" loadObj :: PathString -> IO Int +foreign import ccall "resolveObjs" resolveObjs :: IO Int ===================================== testsuite/tests/rts/linker/T25191.stdout ===================================== @@ -0,0 +1 @@ +success ===================================== testsuite/tests/rts/linker/T25191_foo1.c ===================================== @@ -0,0 +1,10 @@ +#include + +void __attribute__ ((__visibility__ ("hidden"))) foo(void) { + printf("HIDDEN FOO\n"); +} + +void bar(void) { + printf("BAR\n"); + foo(); +} ===================================== testsuite/tests/rts/linker/T25191_foo2.c ===================================== @@ -0,0 +1,8 @@ +#include + +extern void bar(void); + +void foo(void) { + printf("VISIBLE FOO\n"); + bar(); +} ===================================== testsuite/tests/rts/linker/all.T ===================================== @@ -168,3 +168,11 @@ test('reloc-none', unless(opsys('linux'), skip), req_rts_linker], makefile_test, ['reloc-none']) + +test('T25191', + [req_rts_linker, + extra_files(['T25191_foo1.c','T25191_foo2.c']), + when(opsys('darwin'), expect_broken(25191)), # not supported in the MachO linker yet + when(opsys('mingw32'), expect_broken(25191)) # not supported in the PE linker yet + ], + makefile_test, ['T25191']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f3b98dbcff144af0163d9ebfe5504f6fba8dfd9d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f3b98dbcff144af0163d9ebfe5504f6fba8dfd9d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 14:34:00 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 03 Sep 2024 10:34:00 -0400 Subject: [Git][ghc/ghc][wip/romes/25110] base: Deprecate BCO primops exports from GHC.Exts Message-ID: <66d71e58732a7_1aea528b678536dc@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/25110 at Glasgow Haskell Compiler / GHC Commits: 919de09b by Rodrigo Mesquita at 2024-09-03T15:33:47+01:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - 1 changed file: - libraries/base/src/GHC/Exts.hs Changes: ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -23,6 +23,12 @@ module GHC.Exts -- ** Legacy interface for arrays of arrays module GHC.Internal.ArrayArray, -- * Primitive operations + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.BCO, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.mkApUpd0#, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.newBCO#, module GHC.Prim, module GHC.Prim.Ext, -- ** Running 'RealWorld' state thread @@ -112,7 +118,10 @@ module GHC.Exts import GHC.Internal.Exts import GHC.Internal.ArrayArray -import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# ) +import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# + -- Deprecated + , BCO, mkApUpd0#, newBCO# ) +import qualified GHC.Prim as Prim ( BCO, mkApUpd0#, newBCO# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/919de09bf32869ce495e135f701b38547f0a0029 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/919de09bf32869ce495e135f701b38547f0a0029 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 14:38:19 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Tue, 03 Sep 2024 10:38:19 -0400 Subject: [Git][ghc/ghc][wip/unwire-base] 2 commits: Move Control.Monad.Zip into ghc-internal Message-ID: <66d71f5bb9243_1aea52190a00542e2@gitlab.mail> Matthew Pickering pushed to branch wip/unwire-base at Glasgow Haskell Compiler / GHC Commits: 44505e59 by Matthew Pickering at 2024-09-03T15:37:50+01:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - 8b91e994 by Matthew Pickering at 2024-09-03T15:37:50+01:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 17 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Unit/Types.hs - libraries/base/base.cabal.in - libraries/base/src/Control/Monad/Zip.hs - libraries/base/src/Data/List/NonEmpty.hs - libraries/ghc-internal/ghc-internal.cabal.in - libraries/ghc-internal/src/GHC/Internal/Base.hs - + libraries/ghc-internal/src/GHC/Internal/Control/Monad/Zip.hs - libraries/ghc-internal/src/GHC/Internal/Data/List/NonEmpty.hs - testsuite/tests/backpack/should_compile/bkp16.stderr - testsuite/tests/backpack/should_fail/bkpfail17.stderr - testsuite/tests/backpack/should_fail/bkpfail19.stderr - testsuite/tests/iface/IfaceSharingIfaceType.hs - testsuite/tests/iface/IfaceSharingName.hs - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/db069de5ab77bab6b8a82e540c3f1ff5493829da...8b91e994f202e4a6dfad139be6d7f841879e4a46 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/db069de5ab77bab6b8a82e540c3f1ff5493829da...8b91e994f202e4a6dfad139be6d7f841879e4a46 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 14:40:43 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Tue, 03 Sep 2024 10:40:43 -0400 Subject: [Git][ghc/ghc][wip/temp-files2] Use deterministic names for temporary files Message-ID: <66d71feb4cbe3_1aea522d2b34549db@gitlab.mail> Matthew Pickering pushed to branch wip/temp-files2 at Glasgow Haskell Compiler / GHC Commits: 1c39a2d1 by Matthew Pickering at 2024-09-03T15:40:12+01:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 3 changed files: - compiler/GHC/Driver/Make.hs - compiler/GHC/Utils/TmpFs.hs - testsuite/tests/ghc-e/should_fail/T9930fail.stderr Changes: ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2924,19 +2924,22 @@ runParPipelines worker_limit plugin_hsc_env diag_wrapper mHscMessager all_pipeli atomically $ writeTVar stopped_var True wait_log_thread -withLocalTmpFS :: RunMakeM a -> RunMakeM a -withLocalTmpFS act = do +withLocalTmpFS :: TmpFs -> (TmpFs -> IO a) -> IO a +withLocalTmpFS tmpfs act = do let initialiser = do - MakeEnv{..} <- ask - lcl_tmpfs <- liftIO $ forkTmpFsFrom (hsc_tmpfs hsc_env) - return $ hsc_env { hsc_tmpfs = lcl_tmpfs } - finaliser lcl_env = do - gbl_env <- ask - liftIO $ mergeTmpFsInto (hsc_tmpfs lcl_env) (hsc_tmpfs (hsc_env gbl_env)) + liftIO $ forkTmpFsFrom tmpfs + finaliser tmpfs_local = do + liftIO $ mergeTmpFsInto tmpfs_local tmpfs -- Add remaining files which weren't cleaned up into local tmp fs for -- clean-up later. -- Clear the logQueue if this node had it's own log queue - MC.bracket initialiser finaliser $ \lcl_hsc_env -> local (\env -> env { hsc_env = lcl_hsc_env}) act + MC.bracket initialiser finaliser act + +withLocalTmpFSMake :: MakeEnv -> (MakeEnv -> IO a) -> IO a +withLocalTmpFSMake env k = + withLocalTmpFS (hsc_tmpfs (hsc_env env)) $ \lcl_tmpfs + -> k (env { hsc_env = (hsc_env env) { hsc_tmpfs = lcl_tmpfs }}) + -- | Run the given actions and then wait for them all to finish. runAllPipelines :: WorkerLimit -> MakeEnv -> [MakeAction] -> IO () @@ -2958,16 +2961,18 @@ runAllPipelines worker_limit env acts = do runLoop :: (((forall a. IO a -> IO a) -> IO ()) -> IO a) -> MakeEnv -> [MakeAction] -> IO [a] runLoop _ _env [] = return [] runLoop fork_thread env (MakeAction act res_var :acts) = do - new_thread <- + + -- withLocalTmpFs has to occur outside of fork to remain deterministic + new_thread <- withLocalTmpFSMake env $ \lcl_env -> fork_thread $ \unmask -> (do - mres <- (unmask $ run_pipeline (withLocalTmpFS act)) + mres <- (unmask $ run_pipeline lcl_env act) `MC.onException` (putMVar res_var Nothing) -- Defensive: If there's an unhandled exception then still signal the failure. putMVar res_var mres) threads <- runLoop fork_thread env acts return (new_thread : threads) where - run_pipeline :: RunMakeM a -> IO (Maybe a) - run_pipeline p = runMaybeT (runReaderT p env) + run_pipeline :: MakeEnv -> RunMakeM a -> IO (Maybe a) + run_pipeline env p = runMaybeT (runReaderT p env) data MakeAction = forall a . MakeAction !(RunMakeM a) !(MVar (Maybe a)) ===================================== compiler/GHC/Utils/TmpFs.hs ===================================== @@ -64,6 +64,8 @@ data TmpFs = TmpFs -- -- Shared with forked TmpFs. + , tmp_dir_prefix :: String + , tmp_files_to_clean :: IORef PathsToClean -- ^ Files to clean (per session or per module) -- @@ -121,6 +123,7 @@ initTmpFs = do , tmp_subdirs_to_clean = subdirs , tmp_dirs_to_clean = dirs , tmp_next_suffix = next + , tmp_dir_prefix = "tmp" } -- | Initialise an empty TmpFs sharing unique numbers and per-process temporary @@ -132,11 +135,16 @@ forkTmpFsFrom :: TmpFs -> IO TmpFs forkTmpFsFrom old = do files <- newIORef emptyPathsToClean subdirs <- newIORef emptyPathsToClean + counter <- newIORef 0 + prefix <- newTempSuffix old + + return $ TmpFs { tmp_files_to_clean = files , tmp_subdirs_to_clean = subdirs , tmp_dirs_to_clean = tmp_dirs_to_clean old - , tmp_next_suffix = tmp_next_suffix old + , tmp_next_suffix = counter + , tmp_dir_prefix = prefix } -- | Merge the first TmpFs into the second. @@ -259,9 +267,11 @@ changeTempFilesLifetime tmpfs lifetime files = do addFilesToClean tmpfs lifetime existing_files -- Return a unique numeric temp file suffix -newTempSuffix :: TmpFs -> IO Int -newTempSuffix tmpfs = - atomicModifyIORef' (tmp_next_suffix tmpfs) $ \n -> (n+1,n) +newTempSuffix :: TmpFs -> IO String +newTempSuffix tmpfs = do + n <- atomicModifyIORef' (tmp_next_suffix tmpfs) $ \n -> (n+1,n) + return $ tmp_dir_prefix tmpfs ++ "_" ++ show n + -- Find a temporary name that doesn't already exist. newTempName :: Logger -> TmpFs -> TempDir -> TempFileLifetime -> Suffix -> IO FilePath @@ -271,8 +281,8 @@ newTempName logger tmpfs tmp_dir lifetime extn where findTempName :: FilePath -> IO FilePath findTempName prefix - = do n <- newTempSuffix tmpfs - let filename = prefix ++ show n <.> extn + = do suffix <- newTempSuffix tmpfs + let filename = prefix ++ suffix <.> extn b <- doesFileExist filename if b then findTempName prefix else do -- clean it up later @@ -295,8 +305,8 @@ newTempSubDir logger tmpfs tmp_dir where findTempDir :: FilePath -> IO FilePath findTempDir prefix - = do n <- newTempSuffix tmpfs - let name = prefix ++ show n + = do suffix <- newTempSuffix tmpfs + let name = prefix ++ suffix b <- doesDirectoryExist name if b then findTempDir prefix else (do @@ -314,8 +324,8 @@ newTempLibName logger tmpfs tmp_dir lifetime extn where findTempName :: FilePath -> String -> IO (FilePath, FilePath, String) findTempName dir prefix - = do n <- newTempSuffix tmpfs -- See Note [Deterministic base name] - let libname = prefix ++ show n + = do suffix <- newTempSuffix tmpfs -- See Note [Deterministic base name] + let libname = prefix ++ suffix filename = dir "lib" ++ libname <.> extn b <- doesFileExist filename if b then findTempName dir prefix @@ -340,8 +350,8 @@ getTempDir logger tmpfs (TempDir tmp_dir) = do mkTempDir :: FilePath -> IO FilePath mkTempDir prefix = do - n <- newTempSuffix tmpfs - let our_dir = prefix ++ show n + suffix <- newTempSuffix tmpfs + let our_dir = prefix ++ suffix -- 1. Speculatively create our new directory. createDirectory our_dir @@ -376,6 +386,11 @@ the temporary file no longer contains random information (it used to contain the process id). This is ok, as the temporary directory used contains the pid (see getTempDir). + +In addition to this, multiple threads can race against each other creating temporary +files. Therefore we supply a prefix when creating temporary files, when a thread is +forked, each thread must be given an TmpFs with a unique prefix. This is achieved +by forkTmpFsFrom creating a fresh prefix from the parent TmpFs. -} manyWithTrace :: Logger -> String -> ([FilePath] -> IO ()) -> [FilePath] -> IO () ===================================== testsuite/tests/ghc-e/should_fail/T9930fail.stderr ===================================== @@ -1,18 +1,12 @@ -ghc: Exception: +ghc-9.11.20240830: Exception: default output name would overwrite the input file; must specify -o explicitly Usage: For basic information, try the `--help' option. -Package: ghc-inplace +Package: ghc-9.11-inplace Module: GHC.Utils.Panic Type: GhcException HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO - throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:860:84 in exceptions-0.10.7-inplace:Control.Monad.Catch - onException, called at compiler/GHC/Driver/Make.hs:2974:23 in ghc-9.9-inplace:GHC.Driver.Make - - + bracket, called at compiler/GHC/Driver/Make.hs:2936:3 in ghc-9.11-inplace:GHC.Driver.Make View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1c39a2d14774a242f3bbec752393a215b284c158 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1c39a2d14774a242f3bbec752393a215b284c158 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 17:15:28 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 03 Sep 2024 13:15:28 -0400 Subject: [Git][ghc/ghc][master] Haddock: Add no-compilation flag Message-ID: <66d74430cde56_397d2d51d7406288f@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 9 changed files: - utils/haddock/.gitignore - utils/haddock/CHANGES.md - utils/haddock/doc/invoking.rst - utils/haddock/haddock-api/src/Haddock.hs - utils/haddock/haddock-api/src/Haddock/Interface.hs - utils/haddock/haddock-api/src/Haddock/Options.hs - utils/haddock/haddock-test/src/Test/Haddock.hs - utils/haddock/haddock-test/src/Test/Haddock/Config.hs - utils/haddock/latex-test/Main.hs Changes: ===================================== utils/haddock/.gitignore ===================================== @@ -11,6 +11,10 @@ /hypsrc-test/one-shot-out/ /latex-test/one-shot-out/ /hoogle-test/one-shot-out/ +/html-test/no-compilation-out/ +/hypsrc-test/no-compilation-out/ +/latex-test/no-compilation-out/ +/hoogle-test/no-compilation-out/ *.o *.hi ===================================== utils/haddock/CHANGES.md ===================================== @@ -3,6 +3,10 @@ * Add incremental mode to support rendering documentation one module at a time. + * The flag `--no-compilation` has been added. This flag causes Haddock to avoid + recompilation of the code when generating documentation by only reading + the `.hi` and `.hie` files, and will throw an error if it can't find them. + * Fix large margin on top of small headings * Include `package_info` with haddock's `--show-interface` option. ===================================== utils/haddock/doc/invoking.rst ===================================== @@ -542,6 +542,13 @@ The following options are available: ``cabal`` uses temporary `response files `_ to pass arguments to Haddock. +.. option:: --no-compilation + + Always :ref:`avoids recompilation`, only loads the + required ``.hi`` and ``.hie`` files. Haddock will throw an error when it can't + find them. This will not check if the input files are out of date. + (This flag implies :option:`--no-tmp-comp-dir`.) + .. option:: --incremental= Use Haddock in :ref:`incremental mode`. Haddock will generate @@ -555,6 +562,8 @@ sources are accepted without the need for the user to do anything. To use the C pre-processor, however, the user must pass the ``-cpp`` option to GHC using :option:`--optghc`. +.. _avoiding-recompilation: + Avoiding recompilation ---------------------- @@ -579,32 +588,15 @@ should write the ``.hi`` and ``.hie`` files by providing the are building your application with ``cabal build``, the default location is in ``dist-newstyle/build/-/ghc-/-0.1.0/build``. -The next step is to ensure that the flags which Haddock passes to GHC will not -trigger recompilation. Unfortunately, this is not very easy to do if you are -invoking Haddock through ``cabal haddock``. Upon ``cabal haddock``, Cabal passes -a ``--optghc="-optP-D__HADDOCK_VERSION__=NNNN"`` (where ``NNNN`` is the Haddock -version number) flag to Haddock, which forwards the ``-optP=...`` flag to GHC -and triggers a recompilation (unless the existing build results were also -created by a ``cabal haddock``). Additionally, Cabal passes a -``--optghc="-stubdir="`` flag to Haddock, which forwards the -``-stubdir=`` flag to GHC and triggers a recompilation since -``-stubdir`` adds a global include directory. Moreover, since the ``stubdir`` -that Cabal passes is a temporary directory, a recompilation is triggered even -for immediately successive invocations. To avoid recompilations due to these -flags, one must manually extract the arguments passed to Haddock by Cabal and -remove the ``--optghc="-optP-D__HADDOCK_VERSION__=NNNN"`` and -``--optghc="-stubdir="`` flags. This can be achieved using the -:option:`--trace-args` flag by invoking ``cabal haddock`` with -``--haddock-option="--trace-args"`` and copying the traced arguments to a script -which makes an equivalent call to Haddock without the aformentioned flags. - -In addition to the above, Cabal passes a temporary directory as ``-hidir`` to -Haddock by default. Obviously, this also triggers a recompilation for every -invocation of ``cabal haddock``, since it will never find the necessary +The next step is to make sure Haddock runs in no-compilation mode by using +the :option:`--no-compilation` flag. In addition, Cabal passes a +temporary directory as ``-hidir`` to Haddock by default. This will cause +``cabal haddock`` to error, since it will never find the necessary interface files in that temporary directory. To remedy this, pass a ``--optghc="-hidir=/path/to/hidir"`` flag to Haddock, where ``/path/to/hidir`` is the path to the directory in which your build process is writing ``.hi`` -files. +files. You can do this by invoking ``cabal haddock`` with +``--haddock-options="--no-compilation --optghc=-hidir --optghc=/path/to/hidir"``. Following the steps above will allow you to take full advantage of "hi-haddock" and generate Haddock documentation from existing build results without requiring ===================================== utils/haddock/haddock-api/src/Haddock.hs ===================================== @@ -166,14 +166,14 @@ haddockWithGhc ghc args = handleTopExceptions $ do qual <- rightOrThrowE (qualification flags) sinceQual <- rightOrThrowE (sinceQualification flags) - let isOneShotMode = isJust (optOneShot flags) + let noCompilation = isJust (optOneShot flags) || Flag_NoCompilation `elem` flags -- Inject dynamic-too into ghc options if the ghc we are using was built with - -- dynamic linking (except when in one-shot mode) + -- dynamic linking (except when not doing any compilation) flags'' <- ghc flags $ do df <- getDynFlags case lookup "GHC Dynamic" (compilerInfo df) of - Just "YES" | not isOneShotMode -> return $ Flag_OptGhc "-dynamic-too" : flags + Just "YES" | not noCompilation -> return $ Flag_OptGhc "-dynamic-too" : flags _ -> return flags -- Inject `-j` into ghc options, if given to Haddock @@ -191,8 +191,7 @@ haddockWithGhc ghc args = handleTopExceptions $ do -- Output dir needs to be set before calling 'depanal' since 'depanal' uses it -- to compute output file names that are stored in the 'DynFlags' of the -- resulting 'ModSummary's. - let withDir | Flag_NoTmpCompDir `elem` flags = id - | isOneShotMode = id + let withDir | Flag_NoTmpCompDir `elem` flags || noCompilation = id | otherwise = withTempOutputDir -- Output warnings about potential misuse of some flags ===================================== utils/haddock/haddock-api/src/Haddock/Interface.hs ===================================== @@ -176,19 +176,21 @@ createIfaces verbosity modules flags instIfaceMap = do dflags <- getSessionDynFlags let dflags' = dflags { ldInputs = map (FileOption "") o_files ++ ldInputs dflags } - _ <- setSessionDynFlags dflags' + dflags'' = if Flag_NoCompilation `elem` flags then dflags' { ghcMode = OneShot } else dflags' + _ <- setSessionDynFlags dflags'' targets <- mapM (\(filePath, _) -> guessTarget filePath Nothing Nothing) hs_srcs setTargets targets (_errs, modGraph) <- depanalE [] False - liftIO $ traceMarkerIO "Load started" - -- Create (if necessary) and load .hi-files. - success <- withTimingM "load'" (const ()) $ - load' noIfaceCache LoadAllTargets mkUnknownDiagnostic (Just batchMsg) modGraph - when (failed success) $ do - out verbosity normal "load' failed" - liftIO exitFailure - liftIO $ traceMarkerIO "Load ended" + -- Create (if necessary) and load .hi-files. With --no-compilation this happens later. + when (Flag_NoCompilation `notElem` flags) $ do + liftIO $ traceMarkerIO "Load started" + success <- withTimingM "load'" (const ()) $ + load' noIfaceCache LoadAllTargets mkUnknownDiagnostic (Just batchMsg) modGraph + when (failed success) $ do + out verbosity normal "load' failed" + liftIO exitFailure + liftIO $ traceMarkerIO "Load ended" -- We topologically sort the module graph including boot files, -- so it should be acylic (hopefully we failed much earlier if this is not the case) @@ -260,6 +262,20 @@ dropErr :: MaybeErr e a -> Maybe a dropErr (Succeeded a) = Just a dropErr (Failed _) = Nothing +loadHiFile :: HscEnv -> Outputable.SDoc -> Module -> IO (ModIface, ([ClsInst], [FamInst])) +loadHiFile hsc_env doc theModule = initIfaceLoad hsc_env $ do + + mod_iface <- loadSysInterface doc theModule + + insts <- initIfaceLcl (mi_semantic_module mod_iface) doc (mi_boot mod_iface) $ do + + new_eps_insts <- mapM tcIfaceInst (mi_insts mod_iface) + new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts mod_iface) + + pure (new_eps_insts, new_eps_fam_insts) + + pure (mod_iface, insts) + processModule :: Verbosity -> ModSummary -> [Flag] -> IfaceMap -> InstIfaceMap -> WarningMap -> Ghc (Maybe Interface) processModule verbosity modSummary flags ifaceMap instIfaceMap warningMap = do out verbosity verbose $ "Checking module " ++ moduleString (ms_mod modSummary) ++ "..." @@ -267,17 +283,19 @@ processModule verbosity modSummary flags ifaceMap instIfaceMap warningMap = do hsc_env <- getSession dflags <- getDynFlags let sDocContext = DynFlags.initSDocContext dflags Outputable.defaultUserStyle - let hmi = case lookupHpt (hsc_HPT hsc_env) (moduleName $ ms_mod modSummary) of - Nothing -> error "processModule: All modules should be loaded into the HPT by this point" - Just x -> x - mod_iface = hm_iface hmi + doc = text "processModule" unit_state = hsc_units hsc_env - cls_insts = instEnvElts . md_insts $ hm_details hmi + (mod_iface, insts) <- if Flag_NoCompilation `elem` flags + then liftIO $ loadHiFile hsc_env doc $ ms_mod modSummary + else + let hmi = case lookupHpt (hsc_HPT hsc_env) (moduleName $ ms_mod modSummary) of + Nothing -> error "processModule: All modules should be loaded into the HPT by this point" + Just x -> x + cls_insts = instEnvElts . md_insts $ hm_details hmi + fam_insts = md_fam_insts $ hm_details hmi - fam_insts = md_fam_insts $ hm_details hmi - - insts = (cls_insts, fam_insts) + in pure (hm_iface hmi, (cls_insts, fam_insts)) !interface <- do logger <- getLogger @@ -363,18 +381,7 @@ createOneShotIface verbosity flags instIfaceMap moduleNameStr = do modifySession $ hscSetFlags dflags hsc_env <- getSession - (iface, insts) <- liftIO $ initIfaceLoad hsc_env $ do - - iface <- loadSysInterface doc $ mkMainModule_ moduleNm - - insts <- initIfaceLcl (mi_semantic_module iface) doc (mi_boot iface) $ do - - new_eps_insts <- mapM tcIfaceInst (mi_insts iface) - new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) - - pure (new_eps_insts, new_eps_fam_insts) - - pure (iface, insts) + (iface, insts) <- liftIO $ loadHiFile hsc_env doc $ mkMainModule_ moduleNm -- Update the DynFlags with the extensions from the source file (as stored in the interface file) -- This is instead of ms_hspp_opts from ModSummary, which is not available in one-shot mode. ===================================== utils/haddock/haddock-api/src/Haddock/Options.hs ===================================== @@ -124,6 +124,7 @@ data Flag | Flag_ParCount (Maybe Int) | Flag_TraceArgs | Flag_OneShot String + | Flag_NoCompilation deriving (Eq, Show) options :: Bool -> [OptDescr Flag] @@ -158,6 +159,11 @@ options backwardsCompat = ["show-interface"] (ReqArg Flag_ShowInterface "FILE") "print the interface in a human readable form" + , Option + [] + ["no-compilation"] + (NoArg Flag_NoCompilation) + "never compile the code, just read the .hi files" , Option [] ["incremental"] ===================================== utils/haddock/haddock-test/src/Test/Haddock.hs ===================================== @@ -10,6 +10,7 @@ module Test.Haddock import Control.Monad import qualified Data.ByteString.Char8 as BS import qualified Data.Map.Strict as Map +import Data.Foldable (for_) import Data.Maybe import GHC.ResponseFile import System.Directory @@ -74,6 +75,7 @@ maybeDiff cfg@(Config{cfgDiffTool = (Just diff)}) files = do runHaddock :: Config c -> IO Bool runHaddock cfg@(Config{..}) = do createEmptyDirectory $ cfgOutDir cfg + createEmptyDirectory $ cfgNoCompilationOutDir cfg createEmptyDirectory $ cfgOneShotOutDir cfg putStrLn "Generating documentation..." @@ -93,41 +95,65 @@ runHaddock cfg@(Config{..}) = do succeeded <- waitForSuccess msg stdout =<< runProcess' cfgHaddockPath pc unless succeeded $ removeDirectoryRecursive (outDir cfgDirConfig tpkg) - if cfgSkipOneShot then pure succeeded else do - let oneShotDir = oneshotOutDir cfgDirConfig tpkg - hiDir = oneShotDir "hi" - hieDir = oneShotDir "hie" + let noCompilationDir = noCompilationOutDir cfgDirConfig tpkg + hiDir = noCompilationDir "hi" + hieDir = noCompilationDir "hie" + + createEmptyDirectory noCompilationDir + createEmptyDirectory hiDir + createEmptyDirectory hieDir + + -- Build .hi files + let pc = + processConfig + { pcArgs = + concat + [ + [ "--make" + , "-haddock" + , "-fwrite-interface" + , "-fwrite-ide-info" + , "-no-keep-o-files" + , "-hidir=" ++ hiDir + , "-hiedir=" ++ hieDir + ] + , tpkgFiles tpkg + ] + , pcEnv = Just cfgEnv + } + let msg = "Failed to run GHC on test package '" ++ tpkgName tpkg ++ "'" + _ <- waitForSuccess msg stdout =<< runProcess' cfgGhcPath pc + + -- Generate documentation with no-compilation flag + let pc = + processConfig + { pcArgs = + concat + [ cfgHaddockArgs + , [ "--odir=" ++ noCompilationDir + , "--optghc=-hidir=" ++ hiDir + , "--optghc=-hiedir=" ++ hieDir + , "--no-compilation" + ] + , tpkgFiles tpkg + ] + , pcEnv = Just cfgEnv + } + + let msg = "Failed to run Haddock in no-compilation mode on test package '" ++ tpkgName tpkg ++ "'" + succeededNC <- waitForSuccess msg stdout =<< runProcess' cfgHaddockPath pc + + -- Generate documentation incrementally + if cfgSkipOneShot then pure (succeeded && succeededNC) else do + let oneShotDir = oneShotOutDir cfgDirConfig tpkg responseFile = hiDir "response-file" createEmptyDirectory oneShotDir - createEmptyDirectory hiDir - createEmptyDirectory hieDir writeFile responseFile $ escapeArgs [ "--odir=" ++ oneShotDir , "--optghc=-hidir=" ++ hiDir , "--optghc=-hiedir=" ++ hieDir ] - -- Build .hi files - let pc' = - processConfig - { pcArgs = - concat - [ - [ "--make" - , "-haddock" - , "-fwrite-interface" - , "-fwrite-ide-info" - , "-no-keep-o-files" - , "-hidir=" ++ hiDir - , "-hiedir=" ++ hieDir - ] - , tpkgFiles tpkg - ] - , pcEnv = Just cfgEnv - } - let msg = "Failed to run GHC on test package '" ++ tpkgName tpkg ++ "'" - _ <- waitForSuccess msg stdout =<< runProcess' cfgGhcPath pc' - files <- filter ((== ".hi") . takeExtension) <$> listDirectory hiDir -- Use the output order of GHC as a simple dependency order filesSorted <- Map.elems . Map.fromList <$> traverse (\file -> (,file) <$> getModificationTime (hiDir file)) files @@ -157,37 +183,30 @@ runHaddock cfg@(Config{..}) = do escapeArgs [ "--read-interface=" ++ srcRef ++ haddockFile ] loop files else pure False - succeeded2 <- loop filesSorted - when succeeded2 $ do + succeededOS <- loop filesSorted + when (succeededNC && succeededOS) $ do removeDirectoryRecursive hiDir removeDirectoryRecursive hieDir - pure succeeded2 + pure (succeeded && succeededNC && succeededOS) let somethingFailed = any not successes pure somethingFailed checkFile :: Config c -> FilePath -> IO CheckResult checkFile cfg file = do - hasRef <- doesFileExist $ refFile dcfg file - if hasRef - then do - mout <- readOut cfg file - mref <- readRef cfg file - case (mout, mref) of - (Just out, Just ref) - | ccfgEqual ccfg out ref -> - if cfgSkipOneShot cfg || dcfgCheckIgnoreOneShot (cfgDirConfig cfg) file - then return Pass - else do - mOneShotOut <- readOneShotOut cfg file - return $ case mOneShotOut of - Just oneShotOut - | ccfgEqual ccfg oneShotOut out -> Pass - | otherwise -> Fail - Nothing -> Error "Failed to parse one-shot input file" - | otherwise -> return Fail - _ -> return $ Error "Failed to parse input files" - else return NoRef + mref <- readRef cfg file + case mref of + Just ref -> do + let checkStep dcfgDir = ccfgEqual ccfg ref <$> readOut cfg dcfgDir file + result <- checkStep dcfgOutDir + resultNC <- if dcfgCheckIgnoreNoCompilation (cfgDirConfig cfg) file + then pure True + else checkStep dcfgNoCompilationOutDir + resultOS <- if cfgSkipOneShot cfg || dcfgCheckIgnoreOneShot (cfgDirConfig cfg) file + then pure True + else checkStep dcfgOneShotOutDir + pure $ if and [result, resultNC, resultOS] then Pass else Fail + Nothing -> return NoRef where ccfg = cfgCheckConfig cfg dcfg = cfgDirConfig cfg @@ -207,59 +226,50 @@ readRef cfg file = dcfg = cfgDirConfig cfg -- | Read (and clean) the test output artifact for a test -readOut :: Config c -> FilePath -> IO (Maybe c) -readOut cfg file = - fmap (ccfgClean ccfg file) . ccfgRead ccfg . BS.unpack - <$> BS.readFile (outFile dcfg file) - where - ccfg = cfgCheckConfig cfg - dcfg = cfgDirConfig cfg - -readOneShotOut :: Config c -> FilePath -> IO (Maybe c) -readOneShotOut cfg file = - fmap (ccfgClean ccfg file) . ccfgRead ccfg . BS.unpack - <$> BS.readFile (oneShotOutFile dcfg file) +readOut :: Config c -> (DirConfig -> FilePath) -> FilePath -> IO c +readOut cfg dcfgDir file = do + res <- fmap (ccfgClean ccfg file) . ccfgRead ccfg . BS.unpack + <$> BS.readFile outFile + case res of + Just out -> return out + Nothing -> error $ "Failed to parse output file: " ++ outFile where ccfg = cfgCheckConfig cfg dcfg = cfgDirConfig cfg + outFile = dcfgDir dcfg file diffFile :: Config c -> FilePath -> FilePath -> IO () diffFile cfg diff file = do - Just out <- readOut cfg file - Just oneShotOut <- readOneShotOut cfg file Just ref <- readRef cfg file - writeFile outFile' $ ccfgDump ccfg out - writeFile oneShotOutFile' $ ccfgDump ccfg oneShotOut - writeFile refFile' $ ccfgDump ccfg ref - - putStrLn $ "Diff for file \"" ++ file ++ "\":" - hFlush stdout - handle <- - runProcess' diff $ - processConfig - { pcArgs = [outFile', refFile'] - , pcStdOut = Just stdout - } - void $ waitForProcess handle - handle' <- - runProcess' diff $ - processConfig - { pcArgs = [oneShotOutFile', outFile'] - , pcStdOut = Just stdout - } - void $ waitForProcess handle' - return () + out <- readOut cfg dcfgOutDir file + noCompilationOut <- readOut cfg dcfgNoCompilationOutDir file + oneShotOut <- readOut cfg dcfgOneShotOutDir file + writeFile (dumpFile "ref") $ ccfgDump ccfg ref + writeFile (dumpFile "out") $ ccfgDump ccfg out + writeFile (dumpFile "oneShot") $ ccfgDump ccfg oneShotOut + writeFile (dumpFile "noCompilation") $ ccfgDump ccfg oneShotOut + + for_ ["out", "oneShot", "noCompilation"] $ \nm -> do + let outFile = dumpFile nm + refFile = dumpFile "ref" + putStrLn $ "Diff for file \"" ++ outFile ++ "\":" + hFlush stdout + handle <- + runProcess' diff $ + processConfig + { pcArgs = [outFile, refFile] + , pcStdOut = Just stdout + } + void $ waitForProcess handle where dcfg = cfgDirConfig cfg ccfg = cfgCheckConfig cfg - outFile' = outFile dcfg file <.> "dump" - oneShotOutFile' = oneShotOutFile dcfg file <.> "dump" - refFile' = outFile dcfg file <.> "ref" <.> "dump" + dumpFile nm = dcfgOutDir dcfg file <.> nm <.> "dump" maybeAcceptFile :: Config c -> FilePath -> CheckResult -> IO CheckResult maybeAcceptFile cfg file result | cfgAccept cfg && result `elem` [NoRef, Fail] = do - Just out <- readOut cfg file + out <- readOut cfg dcfgOutDir file let ref = refFile dcfg file createDirectoryIfMissing True (takeDirectory ref) writeFile ref $ ccfgDump ccfg out @@ -272,14 +282,11 @@ maybeAcceptFile _ _ result = pure result outDir :: DirConfig -> TestPackage -> FilePath outDir dcfg tpkg = dcfgOutDir dcfg tpkgName tpkg -oneshotOutDir :: DirConfig -> TestPackage -> FilePath -oneshotOutDir dcfg tpkg = dcfgOneShotOutDir dcfg tpkgName tpkg - -outFile :: DirConfig -> FilePath -> FilePath -outFile dcfg file = dcfgOutDir dcfg file +oneShotOutDir :: DirConfig -> TestPackage -> FilePath +oneShotOutDir dcfg tpkg = dcfgOneShotOutDir dcfg tpkgName tpkg -oneShotOutFile :: DirConfig -> FilePath -> FilePath -oneShotOutFile dcfg file = dcfgOneShotOutDir dcfg file +noCompilationOutDir :: DirConfig -> TestPackage -> FilePath +noCompilationOutDir dcfg tpkg = dcfgNoCompilationOutDir dcfg tpkgName tpkg refFile :: DirConfig -> FilePath -> FilePath refFile dcfg file = dcfgRefDir dcfg file ===================================== utils/haddock/haddock-test/src/Test/Haddock/Config.hs ===================================== @@ -4,7 +4,7 @@ module Test.Haddock.Config ( TestPackage(..), CheckConfig(..), DirConfig(..), Config(..) , defaultDirConfig - , cfgSrcDir, cfgRefDir, cfgOutDir, cfgResDir, cfgOneShotOutDir + , cfgSrcDir, cfgRefDir, cfgOutDir, cfgResDir, cfgOneShotOutDir, cfgNoCompilationOutDir , parseArgs, checkOpt, loadConfig ) where @@ -58,9 +58,11 @@ data DirConfig = DirConfig , dcfgRefDir :: FilePath , dcfgOutDir :: FilePath , dcfgOneShotOutDir :: FilePath + , dcfgNoCompilationOutDir :: FilePath , dcfgResDir :: FilePath , dcfgCheckIgnore :: FilePath -> Bool , dcfgCheckIgnoreOneShot :: FilePath -> Bool + , dcfgCheckIgnoreNoCompilation :: FilePath -> Bool } @@ -70,9 +72,11 @@ defaultDirConfig baseDir = DirConfig , dcfgRefDir = baseDir "ref" , dcfgOutDir = baseDir "out" , dcfgOneShotOutDir = baseDir "one-shot-out" + , dcfgNoCompilationOutDir = baseDir "no-compilation-out" , dcfgResDir = rootDir "resources" , dcfgCheckIgnore = const False , dcfgCheckIgnoreOneShot = const False + , dcfgCheckIgnoreNoCompilation = const False } where rootDir = baseDir ".." @@ -92,12 +96,13 @@ data Config c = Config } -cfgSrcDir, cfgRefDir, cfgOutDir, cfgResDir, cfgOneShotOutDir :: Config c -> FilePath +cfgSrcDir, cfgRefDir, cfgOutDir, cfgResDir, cfgOneShotOutDir, cfgNoCompilationOutDir :: Config c -> FilePath cfgSrcDir = dcfgSrcDir . cfgDirConfig cfgRefDir = dcfgRefDir . cfgDirConfig cfgOutDir = dcfgOutDir . cfgDirConfig cfgResDir = dcfgResDir . cfgDirConfig cfgOneShotOutDir = dcfgOneShotOutDir . cfgDirConfig +cfgNoCompilationOutDir = dcfgNoCompilationOutDir . cfgDirConfig ===================================== utils/haddock/latex-test/Main.hs ===================================== @@ -23,6 +23,7 @@ dirConfig = (defaultDirConfig $ takeDirectory __FILE__) { dcfgCheckIgnore = (`elem` ["haddock.sty", "main.tex"]) . takeFileName -- Just a discrepancy in output order , dcfgCheckIgnoreOneShot = (`elem` ["ConstructorArgs.tex"]) . takeFileName + , dcfgCheckIgnoreNoCompilation = (`elem` ["ConstructorArgs.tex"]) . takeFileName } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1499764f729bfe8d36c317b5ee508e5d422fc494 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1499764f729bfe8d36c317b5ee508e5d422fc494 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 17:16:05 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 03 Sep 2024 13:16:05 -0400 Subject: [Git][ghc/ghc][master] Add functions to check for weakly pinned arrays. Message-ID: <66d74455ca77b_397d2d69760c659a@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - 12 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/ffi.rst - libraries/base/src/GHC/Base.hs - libraries/base/src/GHC/Exts.hs - libraries/ghc-prim/changelog.md - rts/PrimOps.cmm - rts/RtsSymbols.c - rts/include/stg/MiscClosures.h - testsuite/tests/rts/T13894.hs Changes: ===================================== compiler/GHC/Builtin/primops.txt.pp ===================================== @@ -1925,7 +1925,25 @@ primop MutableByteArrayIsPinnedOp "isMutableByteArrayPinned#" GenPrimOp primop ByteArrayIsPinnedOp "isByteArrayPinned#" GenPrimOp ByteArray# -> Int# - {Determine whether a 'ByteArray#' is guaranteed not to move during GC.} + {Determine whether a 'ByteArray#' is guaranteed not to move.} + with out_of_line = True + +primop ByteArrayIsWeaklyPinnedOp "isByteArrayWeaklyPinned#" GenPrimOp + ByteArray# -> Int# + {Similar to 'isByteArrayPinned#'. Weakly pinned byte arrays are allowed + to be copied into compact regions by the user, potentially invalidating + the results of earlier calls to 'byteArrayContents#'. + + See the section `Pinned Byte Arrays` in the user guide for more information. + + This function also returns true for regular pinned bytearrays. + } + with out_of_line = True + +primop MutableByteArrayIsWeaklyPinnedOp "isMutableByteArrayWeaklyPinned#" GenPrimOp + MutableByteArray# s -> Int# + { 'isByteArrayWeaklyPinned#' but for mutable arrays. + } with out_of_line = True primop ByteArrayContents_Char "byteArrayContents#" GenPrimOp ===================================== compiler/GHC/StgToCmm/Prim.hs ===================================== @@ -1668,10 +1668,12 @@ emitPrimOp cfg primop = NewPinnedByteArrayOp_Char -> alwaysExternal NewAlignedPinnedByteArrayOp_Char -> alwaysExternal MutableByteArrayIsPinnedOp -> alwaysExternal + MutableByteArrayIsWeaklyPinnedOp -> alwaysExternal DoubleDecode_2IntOp -> alwaysExternal DoubleDecode_Int64Op -> alwaysExternal FloatDecode_IntOp -> alwaysExternal ByteArrayIsPinnedOp -> alwaysExternal + ByteArrayIsWeaklyPinnedOp -> alwaysExternal ShrinkMutableByteArrayOp_Char -> alwaysExternal ResizeMutableByteArrayOp_Char -> alwaysExternal ShrinkSmallMutableArrayOp_Char -> alwaysExternal ===================================== compiler/GHC/StgToJS/Prim.hs ===================================== @@ -670,6 +670,8 @@ genPrim prof bound ty op = case op of NewAlignedPinnedByteArrayOp_Char -> \[r] [l,_align] -> pure $ PrimInline (newByteArray r l) MutableByteArrayIsPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ ByteArrayIsPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ + ByteArrayIsWeaklyPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ + MutableByteArrayIsWeaklyPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ ByteArrayContents_Char -> \[a,o] [b] -> pure $ PrimInline $ mconcat [a |= b, o |= zero_] MutableByteArrayContents_Char -> \[a,o] [b] -> pure $ PrimInline $ mconcat [a |= b, o |= zero_] ShrinkMutableByteArrayOp_Char -> \[] [a,n] -> pure $ PrimInline $ appS hdShrinkMutableByteArrayStr [a,n] ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -163,6 +163,12 @@ Runtime system ~~~~~~~~~~~~~~~~~~~~ - Usage of deprecated primops is now correctly reported (#19629). +- New primops `isMutableByteArrayWeaklyPinned#` and `isByteArrayWeaklyPinned#` + to allow users to avoid copying large arrays safely when dealing with ffi. + See the users guide for more details on the different kinds of + pinned arrays in 9.12. + + This need for this distinction originally surfaced in https://gitlab.haskell.org/ghc/ghc/-/issues/22255 ``ghc`` library ===================================== docs/users_guide/exts/ffi.rst ===================================== @@ -1114,21 +1114,67 @@ Pinned Byte Arrays A pinned byte array is one that the garbage collector is not allowed to move. Consequently, it has a stable address that can be safely -requested with ``byteArrayContents#``. Not that being pinned doesn't -prevent the byteArray from being gc'ed in the same fashion a regular -byte array would be. +requested with ``byteArrayContents#``. As long as the array remains live +the address returned by ``byteArrayContents#`` will remain valid. Note that +being pinned doesn't prevent the byteArray from being gc'ed in the same fashion +a regular byte array would be if there are no more references to the ``ByteArray#``. There are a handful of primitive functions in :base-ref:`GHC.Exts.` used to enforce or check for pinnedness: ``isByteArrayPinned#``, -``isMutableByteArrayPinned#``, and ``newPinnedByteArray#``. A -byte array can be pinned as a result of three possible causes: +``isMutableByteArrayPinned#``, ``isByteArrayWeaklyPinned#``, +``isMutableByteArrayWeaklyPinned#``, and ``newPinnedByteArray#``. A +byte array can be pinned or weakly pinned as a result of three possible causes: -1. It was allocated by ``newPinnedByteArray#``. -2. It is large. Currently, GHC defines large object to be one +1. It was allocated by ``newPinnedByteArray#``. This results in a regular pinned byte array. +2. It is large, this results in a weakly pinned byte array. Currently, GHC defines large object to be one that is at least as large as 80% of a 4KB block (i.e. at least 3277 bytes). -3. It has been copied into a compact region. The documentation +3. It has been copied into a compact region, resulting in a weakly pinned array. The documentation for ``ghc-compact`` and ``compact`` describes this process. +The difference between a pinned array and a weakly pinned array is simply that +trying to compact a pinned array will result in an exception. Trying to compact +a weakly pinned array will succeed. However the result of earlier +calls to ``byteArrayContents#`` is not updated during compaction, which means +these results will still point to the address where the array was located originally, +and not to the new address inside the compact region. + +This is particularly dangerous when an address to a byte arrays content is stored +inside a datastructure along with a reference to the byte array. +If the data structure is compacted later on the pointer won't be updated but the +reference to the byte array will point to a copy inside the compact region. +A common data type susceptible to this is `ForeignPtr` when used to represent a ByteArray#. + +Here is an example to illustrate this: + +.. code-block:: haskell + + workWithArrayContents :: (ByteArray, Ptr Word8) -> (Ptr Word8 -> IO ()) -> IO () + workWithArrayContents (arr@(ByteArray uarr),ptr) worker = + case () of + _ + -- Conservative but safe + | isByteArrayPinned arr -> keepAliveUnlifted uarr (worker ptr) + -- Potentially dangerous, the program needs to ensures the Ptr points into the array. + | isByteArrayWeaklyPinned arr -> keepAliveUnlifted uarr (worker ptr) + | otherwise -> ... -- Otherwise we can't directly use it for safe FFI calls directly at all. + + main :: IO () + main = do + -- We create a large array, which causes it to be implicitly pinned + arr <- newByteArray 5000 + arr@(ByteArray uarr) <- freezeByteArray arr 0 5000 -- Make it immutable + let ptr = byteArrayContents arr + + -- Compacting a data structure that contains both an array and a ptr to + -- the arrays content's is dangerous and usually the wrong thing to do. + let foo = (arr, ptr) + foo_compacted <- compact foo + + -- This is fine + workWithArrayContents foo do_work + -- This is unsound + workWithArrayContents (getCompact foo_compacted) do_work + .. [1] Prior to GHC 8.10, when passing an ``ArrayArray#`` argument to a foreign function, the foreign function would see a pointer to the ``StgMutArrPtrs`` rather than just the payload. ===================================== libraries/base/src/GHC/Base.hs ===================================== @@ -139,7 +139,8 @@ module GHC.Base ) where import GHC.Internal.Base -import GHC.Prim hiding (dataToTagLarge#, dataToTagSmall#, whereFrom#) +import GHC.Prim hiding (dataToTagLarge#, dataToTagSmall#, whereFrom#, + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned#) -- Hide dataToTagLarge# because it is expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -112,7 +112,9 @@ module GHC.Exts import GHC.Internal.Exts import GHC.Internal.ArrayArray -import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# ) +import GHC.Prim hiding ( + coerce, dataToTagSmall#, dataToTagLarge#, whereFrom#, + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/ghc-prim/changelog.md ===================================== @@ -1,3 +1,12 @@ +## 0.13.0 + +- Shipped with GHC 9.12.1 + +- Add primops that allow users to distinguish weakly pinned byte arrays from unpinned ones. + + isMutableByteArrayWeaklyPinned# :: MutableByteArray# s -> Int# + isByteArrayWeaklyPinned# :: ByteArray# s -> Int# + ## 0.12.0 - Shipped with GHC 9.10.1 ===================================== rts/PrimOps.cmm ===================================== @@ -215,12 +215,29 @@ stg_isByteArrayPinnedzh ( gcptr ba ) return (flags & BF_PINNED != 0); } +stg_isByteArrayWeaklyPinnedzh ( gcptr ba ) +// ByteArray# s -> Int# +{ + W_ bd, flags; + bd = Bdescr(ba); + // See #22255 and the primop docs. + flags = TO_W_(bdescr_flags(bd)); + + return (flags & (BF_PINNED | BF_COMPACT | BF_LARGE) != 0); +} + stg_isMutableByteArrayPinnedzh ( gcptr mba ) // MutableByteArray# s -> Int# { jump stg_isByteArrayPinnedzh(mba); } +stg_isMutableByteArrayWeaklyPinnedzh ( gcptr mba ) +// MutableByteArray# s -> Int# +{ + jump stg_isByteArrayWeaklyPinnedzh(mba); +} + /* Note [LDV profiling and resizing arrays] * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * As far as the LDV profiler is concerned arrays are "inherently used" which ===================================== rts/RtsSymbols.c ===================================== @@ -656,6 +656,8 @@ extern char **environ; SymI_HasDataProto(stg_newAlignedPinnedByteArrayzh) \ SymI_HasDataProto(stg_isByteArrayPinnedzh) \ SymI_HasDataProto(stg_isMutableByteArrayPinnedzh) \ + SymI_HasDataProto(stg_isByteArrayWeaklyPinnedzh) \ + SymI_HasDataProto(stg_isMutableByteArrayWeaklyPinnedzh) \ SymI_HasDataProto(stg_shrinkMutableByteArrayzh) \ SymI_HasDataProto(stg_resizzeMutableByteArrayzh) \ SymI_HasDataProto(stg_shrinkSmallMutableArrayzh) \ ===================================== rts/include/stg/MiscClosures.h ===================================== @@ -454,6 +454,8 @@ RTS_FUN_DECL(stg_newPinnedByteArrayzh); RTS_FUN_DECL(stg_newAlignedPinnedByteArrayzh); RTS_FUN_DECL(stg_isByteArrayPinnedzh); RTS_FUN_DECL(stg_isMutableByteArrayPinnedzh); +RTS_FUN_DECL(stg_isByteArrayWeaklyPinnedzh); +RTS_FUN_DECL(stg_isMutableByteArrayWeaklyPinnedzh); RTS_FUN_DECL(stg_shrinkMutableByteArrayzh); RTS_FUN_DECL(stg_resizzeMutableByteArrayzh); RTS_FUN_DECL(stg_shrinkSmallMutableArrayzh); ===================================== testsuite/tests/rts/T13894.hs ===================================== @@ -6,6 +6,7 @@ import Control.Monad import GHC.Exts +import GHC.Internal.Exts (isMutableByteArrayWeaklyPinned#) import GHC.IO main :: IO () @@ -16,3 +17,10 @@ main = do case isMutableByteArrayPinned# arr# of n# -> (# s1, isTrue# n# #) when pinned $ putStrLn "BAD" + + weakly_pinned <- IO $ \s0 -> + case newByteArray# 1000000# s0 of + (# s1, arr# #) -> + case isMutableByteArrayWeaklyPinned# arr# of + n# -> (# s1, isTrue# n# #) + when (not weakly_pinned) $ putStrLn "BAD" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/768fe6445dd39e21497b86a11fbb2dd92d07e2e8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/768fe6445dd39e21497b86a11fbb2dd92d07e2e8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 17:17:05 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 03 Sep 2024 13:17:05 -0400 Subject: [Git][ghc/ghc][master] ghc-toolchain: Don't leave stranded a.outs when testing for -g0 Message-ID: <66d744915a8d1_397d2d885af4739a8@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 1 changed file: - utils/ghc-toolchain/src/GHC/Toolchain/Tools/Cpp.hs Changes: ===================================== utils/ghc-toolchain/src/GHC/Toolchain/Tools/Cpp.hs ===================================== @@ -39,10 +39,10 @@ data CmmCpp = CmmCpp { cmmCppProgram :: Program } deriving (Show, Read, Eq, Ord) -checkFlag :: String -> Program -> String -> M () -checkFlag conftest cpp flag = checking ("for "++flag++" support") $ +checkFlag :: String -> Program -> String -> [String] -> M () +checkFlag conftest cpp flag extra_args = checking ("for "++flag++" support") $ -- Werror to ensure that unrecognized warnings result in an error - callProgram cpp ["-Werror", flag, conftest] + callProgram cpp $ ["-Werror", flag, conftest] ++ extra_args -- tryFlag :: String -> Program -> String -> M [String] -- tryFlag conftest cpp flag = -- ([flag] <$ checkFlag conftest cpp flag) <|> return [] @@ -167,7 +167,7 @@ findCmmCpp progOpt cc = checking "for a Cmm preprocessor" $ do cmmCppSupportsG0 <- withTempDir $ \dir -> do let conftest = dir "conftest.c" writeFile conftest "int main(void) {}" - True <$ checkFlag conftest cpp "-g0" <|> return False + True <$ checkFlag conftest cpp "-g0" ["-o", dir "conftest"] <|> return False -- Always add the -E flag to the CPP, regardless of the user options let cmmCppProgram = foldr addFlagIfNew cpp ["-E"] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/b16605e7c135f8cfd357a60c7f358132faec6a84 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/b16605e7c135f8cfd357a60c7f358132faec6a84 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 17:17:38 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 03 Sep 2024 13:17:38 -0400 Subject: [Git][ghc/ghc][master] Build foreign objects for TH with interpreter's way when loading from iface Message-ID: <66d744b25cad_397d2da378d475974@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 2 changed files: - compiler/GHC/Driver/Main.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs Changes: ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -1,5 +1,4 @@ {-# LANGUAGE LambdaCase #-} - {-# LANGUAGE NondecreasingIndentation #-} {-# LANGUAGE GADTs #-} {-# LANGUAGE MultiWayIf #-} @@ -295,13 +294,13 @@ import Data.Time import System.IO.Unsafe ( unsafeInterleaveIO ) import GHC.Iface.Env ( trace_if ) +import GHC.Platform.Ways import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) import GHC.Types.Unique.FM import GHC.Types.Unique.DFM import GHC.Cmm.Config (CmmConfig) - {- ********************************************************************** %* * Initialisation @@ -990,6 +989,27 @@ initModDetails hsc_env iface = -- in make mode, since this HMI will go into the HPT. genModDetails hsc_env' iface +-- | Modify flags such that objects are compiled for the interpreter's way. +-- This is necessary when building foreign objects for Template Haskell, since +-- those are object code built outside of the pipeline, which means they aren't +-- subject to the mechanism in 'enableCodeGenWhen' that requests dynamic build +-- outputs for dependencies when the interpreter used for TH is dynamic but the +-- main outputs aren't. +-- Furthermore, the HPT only stores one set of objects with different names for +-- bytecode linking in 'HomeModLinkable', so the usual hack for switching +-- between ways in 'get_link_deps' doesn't work. +compile_for_interpreter :: HscEnv -> (HscEnv -> IO a) -> IO a +compile_for_interpreter hsc_env use = + use (hscUpdateFlags update hsc_env) + where + update dflags = dflags { + targetWays_ = adapt_way interpreterDynamic WayDyn $ + adapt_way interpreterProfiled WayProf $ + targetWays_ dflags + } + + adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace -- them with a lazy IO thunk that compiles them to bytecode and foreign objects. -- @@ -2063,9 +2083,10 @@ generateByteCode :: HscEnv -> IO (CompiledByteCode, [FilePath]) generateByteCode hsc_env cgguts mod_location = do (hasStub, comp_bc) <- hscInteractive hsc_env cgguts mod_location - stub_o <- traverse (compileForeign hsc_env LangC) hasStub - foreign_files_o <- traverse (uncurry (compileForeign hsc_env)) (cgi_foreign_files cgguts) - pure (comp_bc, maybeToList stub_o ++ foreign_files_o) + compile_for_interpreter hsc_env $ \ i_env -> do + stub_o <- traverse (compileForeign i_env LangC) hasStub + foreign_files_o <- traverse (uncurry (compileForeign i_env)) (cgi_foreign_files cgguts) + pure (comp_bc, maybeToList stub_o ++ foreign_files_o) generateFreshByteCode :: HscEnv -> ModuleName ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -190,13 +190,9 @@ Even if that wasn't an issue, they are compiled for the session's 'Way', not its associated module's, so the dynamic variant wouldn't be available when building only static outputs. -For now, this doesn't have much of an impact, since we're only supporting -foreign imports initially, which produce very simple objects that can easily be -handled by the linker when 'GHC.Linker.Loader.dynLoadObjs' creates a shared -library from all object file inputs. -However, for more complex circumstances, we should compile foreign stubs -specially for TH according to the interpreter 'Way', or request dynamic products -for TH dependencies like it happens for the conventional case. +To mitigate this, we instead build foreign objects specially for the +interpreter, updating the build flags in 'compile_for_interpreter' to use the +interpreter's way. Problem 4: View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/83e70b14c70de5e0c59c55deac43b0b9b7b54203 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/83e70b14c70de5e0c59c55deac43b0b9b7b54203 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 17:29:40 2024 From: gitlab at gitlab.haskell.org (Matthew Craven (@clyring)) Date: Tue, 03 Sep 2024 13:29:40 -0400 Subject: [Git][ghc/ghc][wip/T23490-part2] 48 commits: haddock: Build haddock-api and haddock-library using hadrian Message-ID: <66d747849fee7_bda388b2e07598f@gitlab.mail> Matthew Craven pushed to branch wip/T23490-part2 at Glasgow Haskell Compiler / GHC Commits: 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 1e574e65 by Matthew Craven at 2024-09-03T13:20:33-04:00 Add missing images to transformers submodule - - - - - 0bb63b31 by Matthew Craven at 2024-09-03T13:20:33-04:00 Remove reference to non-existent file in haddock.cabal - - - - - a68882e4 by Matthew Craven at 2024-09-03T13:20:33-04:00 Update hackage index state - - - - - a327dd1b by Matthew Craven at 2024-09-03T13:20:33-04:00 Move tests T11462 and T11525 into tests/tcplugins - - - - - 157705e3 by Matthew Craven at 2024-09-03T13:22:56-04:00 Repair the 'build-cabal' hadrian target Fixes #23117. Fixes #23281. Fixes #23490. This required: * Updating the bit-rotted compiler/Setup.hs and its setup-depends * Listing a few recently-added libraries and utilities in cabal.project-reinstall * Setting allow-boot-library-installs to 'True' since Cabal now considers the 'ghc' package itself a boot library for the purposes of this flag Additionally, the allow-newer block in cabal.project-reinstall was removed. This block was probably added because when the libraries/Cabal submodule is too new relative to the cabal-install executable, solving the setup-depends for any package with a custom setup requires building an old Cabal (from Hackage) against the in-tree version of base, and this can fail un-necessarily due to tight version bounds on base. However, the blind allow-newer can also cause the solver to go berserk and choose a stupid build plan that has no business succeeding, and the failures when this happens are dreadfully confusing. (See #23281 and #24363.) Why does setup-depends solving insist on an old version of Cabal? See: https://github.com/haskell/cabal/blob/0a0b33983b0f022b9697f7df3a69358ee9061a89/cabal-install/src/Distribution/Client/ProjectPlanning.hs#L1393-L1410 The right solution here is probably to use the in-tree cabal-install from libraries/Cabal/cabal-install with the build-cabal target rather than whatever the environment happens to provide. But this is left for future work. - - - - - dc86ccc2 by Matthew Craven at 2024-09-03T13:28:18-04:00 Revert "CI: Disable the test-cabal-reinstall job" This reverts commit 38c3afb64d3ffc42f12163c6f0f0d5c414aa8255. - - - - - 30 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - cabal.project-reinstall - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Predicate.hs - compiler/GHC/Core/Type.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Config/Finder.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Errors/Types.hs - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Session.hs - compiler/GHC/Hs/Decls.hs - compiler/GHC/Hs/ImpExp.hs - compiler/GHC/Hs/Pat.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/913e05eaa96aa82bf5aa8baea6f7afd26bf5e348...dc86ccc205c68eb632661894689724dec4174192 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/913e05eaa96aa82bf5aa8baea6f7afd26bf5e348...dc86ccc205c68eb632661894689724dec4174192 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 18:59:14 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Tue, 03 Sep 2024 14:59:14 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] 2 commits: store IO in the EPS Message-ID: <66d75c8225b1a_bda384c598c87552@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: ff2e1919 by Torsten Schmits at 2024-09-03T20:24:14+02:00 store IO in the EPS - - - - - 6ae0c945 by Torsten Schmits at 2024-09-03T20:58:11+02:00 Move lazy bytecode storage from Linkable to HomeModLinkable This shifts the responsibility of handling the laziness properly from the rather generic Linkable to a type that is more specific to the domain that necessitates the laziness. - - - - - 14 changed files: - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Home/ModInfo.hs - ghc/GHCi/Leak.hs Changes: ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -50,7 +50,7 @@ module GHC.Driver.Main , HscBackendAction (..), HscRecompStatus (..) , initModDetails , initWholeCoreBindings - , initWholeCoreBindingsEps + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -277,7 +277,7 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.Foldable (fold) +import Data.Functor ((<&>)) import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad @@ -868,7 +868,7 @@ hscRecompStatus -- Do need linkable -- 1. Just check whether we have bytecode/object linkables and then -- we will decide if we need them or not. - bc_linkable <- checkByteCode checked_iface mod_summary (homeMod_bytecode old_linkable) + let bc_linkable = checkByteCode checked_iface mod_summary (homeMod_bytecode old_linkable) obj_linkable <- liftIO $ checkObjects lcl_dflags (homeMod_object old_linkable) mod_summary trace_if (hsc_logger hsc_env) (vcat [text "BCO linkable", nest 2 (ppr bc_linkable), text "Object Linkable", ppr obj_linkable]) @@ -955,25 +955,21 @@ checkObjects dflags mb_old_linkable summary = do -- | Check to see if we can reuse the old linkable, by this point we will -- have just checked that the old interface matches up with the source hash, so -- no need to check that again here -checkByteCode :: ModIface -> ModSummary -> Maybe Linkable -> IO (MaybeValidated Linkable) -checkByteCode iface mod_sum mb_old_linkable = - case mb_old_linkable of - Just old_linkable - | not (linkableIsNativeCodeOnly old_linkable) - -> return $ (UpToDateItem old_linkable) - _ -> loadByteCode iface mod_sum - -loadByteCode :: ModIface -> ModSummary -> IO (MaybeValidated Linkable) -loadByteCode iface mod_sum = do - let - this_mod = ms_mod mod_sum - if_date = fromJust $ ms_iface_date mod_sum - case mi_extra_decls iface of - Just extra_decls -> do - let fi = WholeCoreBindings extra_decls this_mod (ms_location mod_sum) - (mi_foreign iface) - return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) - _ -> return $ outOfDateItemBecause MissingBytecode Nothing +checkByteCode :: + ModIface -> + ModSummary -> + HomeModByteCode -> + MaybeValidated HomeModByteCode +checkByteCode iface mod_sum = \case + NoHomeModByteCode -> HomeModIfaceCore <$> loadByteCode iface mod_sum + old_bytecode -> UpToDateItem old_bytecode + +loadByteCode :: ModIface -> ModSummary -> MaybeValidated WholeCoreBindings +loadByteCode iface mod_sum = + case iface_core_bindings iface (ms_location mod_sum) of + Just wcb -> UpToDateItem wcb + Nothing -> outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- @@ -996,9 +992,43 @@ initModDetails hsc_env iface = -- in make mode, since this HMI will go into the HPT. genModDetails hsc_env' iface --- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects, --- using the supplied environment for type checking. +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Return an 'IO' that hydrates Core bindings and compiles them to bytecode if +-- the interface contains any, using the supplied type env for typechecking. +-- +-- Unlike 'initWholeCoreBindings', this does not use lazy IO. +-- Instead, the 'IO' is only evaluated (in @get_link_deps@) when it is clear +-- that it will be used immediately (because we're linking TH with +-- @-fprefer-byte-code@ in oneshot mode), and the result is cached in +-- 'LoaderState'. +-- +-- 'initWholeCoreBindings' needs the laziness because it is used to populate +-- 'HomeModInfo', which is done preemptively, in anticipation of downstream +-- modules using the bytecode for TH in make mode, which might never happen. +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + compileWholeCoreBindings hsc_env type_env <$> iface_core_bindings iface location + +-- | If the 'HomeModByteCode' contains Core bindings loaded from an interface, +-- replace them with a lazy IO thunk that compiles them to bytecode and foreign +-- objects, using the supplied environment for type checking. -- -- The laziness is necessary because this value is stored purely in a -- 'HomeModLinkable' in the home package table, rather than some dedicated @@ -1012,53 +1042,97 @@ initModDetails hsc_env iface = -- -- This is sound because generateByteCode just depends on things already loaded -- in the interface file. -initWcbWithTcEnv :: +initWholeCoreBindings :: HscEnv -> + ModIface -> + ModLocation -> + ModDetails -> + HomeModByteCode -> + IO HomeModByteCode +initWholeCoreBindings hsc_env iface location details = \case + NoHomeModByteCode + -- REVIEW this is not necessary, but maybe nice to be safe anyway? + -- If @NoHomeModByteCode@ was returned by @runHscBackendPhase@, it's because + -- @-fprefer@ is off. + | gopt Opt_UseBytecodeRatherThanObjects (hsc_dflags hsc_env) -> + maybe (pure NoHomeModByteCode) defer $ + loadIfaceByteCode hsc_env' iface location type_env + | otherwise -> + pure NoHomeModByteCode + HomeModIfaceCore wcb -> + defer $ compileWholeCoreBindings hsc_env' type_env wcb + HomeModByteCode bc -> + pure (HomeModByteCode bc) + HomeModLazyByteCode bc -> + pure (HomeModLazyByteCode bc) + where + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + + -- Run an IO lazily and wrap its result in a lazy datacon, so that the IO + -- is executed only when 'HomeModLazyByteCode' is pattern-matched and the + -- value inside is forced. + defer = fmap HomeModLazyByteCode . unsafeInterleaveIO + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written or were unavailable due to boot import +-- cycles, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +-- +-- 4. Wrapping the build products in 'Linkable' with the proper modification +-- time obtained from the interface. +compileWholeCoreBindings :: HscEnv -> TypeEnv -> - Linkable -> + WholeCoreBindings -> IO Linkable -initWcbWithTcEnv tc_hsc_env hsc_env type_env (Linkable utc_time this_mod uls) = - Linkable utc_time this_mod <$> mapM go uls +compileWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + parts <- gen_bytecode core_binds stubs foreign_files + linkable parts where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef type_env - let - tc_hsc_env_with_kv = tc_hsc_env { - hsc_type_env_vars = - knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") tc_hsc_env_with_kv $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons type_env) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + (bcos, fos) <- generateByteCode hsc_env cgi_guts wcb_mod_location + pure $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file wcb_mod_location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time wcb_module parts + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env --- | Hydrate core bindings for a module in the home package table, for which we --- can obtain a 'ModDetails'. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env iface details = - initWcbWithTcEnv (add_iface_to_hpt iface details hsc_env) hsc_env (md_types details) - --- | Hydrate core bindings for a module in the external package state. --- This is used for home modules as well when compiling in oneshot mode. -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable -initWholeCoreBindingsEps hsc_env iface lnk = do - eps <- hscEPS hsc_env - let type_env = fold (lookupModuleEnv (eps_PTT eps) (mi_module iface)) - initWcbWithTcEnv hsc_env hsc_env type_env lnk - - {- Note [ModDetails and --make mode] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -1,8 +1,15 @@ module GHC.Driver.Main where -import GHC.Driver.Env -import GHC.Linker.Types -import GHC.Prelude -import GHC.Unit.Module.ModIface +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) -initWholeCoreBindingsEps :: HscEnv -> ModIface -> Linkable -> IO Linkable +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1310,8 +1310,10 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- MP: This is a bit janky, because before you add the entries you have to extend the HPT with the module -- you just compiled. Another option, would be delay adding anything until after upsweep has finished, but I -- am unsure if this is sound (wrt running TH splices for example). - -- This function only does anything if the linkable produced is a BCO, which only happens with the - -- bytecode backend, no need to guard against the backend type additionally. + -- This function only does anything if the linkable produced is a BCO, which + -- used to only happen with the bytecode backend, but with + -- @-fprefer-byte-code@, @HomeModInfo@ has bytecode even when generating + -- object code, see #25230. addSptEntries (hscUpdateHPT (\hpt -> addToHpt hpt (ms_mod_name summary) hmi) hsc_env) (homeModInfoByteCode hmi) @@ -1319,11 +1321,12 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- | Add the entries from a BCO linkable to the SPT table, see -- See Note [Grand plan for static forms] in GHC.Iface.Tidy.StaticPtrTable. -addSptEntries :: HscEnv -> Maybe Linkable -> IO () +addSptEntries :: HscEnv -> HomeModByteCode -> IO () addSptEntries hsc_env mlinkable = hscAddSptEntries hsc_env [ spt - | linkable <- maybeToList mlinkable + -- This ignores lazy bytecode from interfaces, see #25230 + | HomeModByteCode linkable <- [mlinkable] , bco <- linkableBCOs linkable , spt <- bc_spt_entries bco ] ===================================== compiler/GHC/Driver/Pipeline.hs ===================================== @@ -248,7 +248,8 @@ compileOne' mHscMessage (iface, linkable) <- runPipeline (hsc_hooks plugin_hsc_env) pipeline -- See Note [ModDetails and --make mode] details <- initModDetails plugin_hsc_env iface - linkable' <- traverse (initWholeCoreBindings plugin_hsc_env iface details) (homeMod_bytecode linkable) + linkable' <- initWholeCoreBindings plugin_hsc_env iface + (ms_location summary) details (homeMod_bytecode linkable) return $! HomeModInfo iface details (linkable { homeMod_bytecode = linkable' }) where lcl_dflags = ms_hspp_opts summary ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -602,7 +602,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do if gopt Opt_ByteCodeAndObjectCode dflags then do bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return $ emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } + return $ emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } else return emptyHomeModInfoLinkable @@ -619,7 +619,7 @@ runHscBackendPhase pipe_env hsc_env mod_name src_flavour location result = do final_iface <- mkFullIface hsc_env partial_iface Nothing Nothing NoStubs [] hscMaybeWriteIface logger dflags True final_iface mb_old_iface_hash location bc <- generateFreshByteCode hsc_env mod_name (mkCgInteractiveGuts cgguts) mod_location - return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = Just bc } , panic "interpreter") + return ([], final_iface, emptyHomeModInfoLinkable { homeMod_bytecode = HomeModByteCode bc } , panic "interpreter") runUnlitPhase :: HscEnv -> FilePath -> FilePath -> IO FilePath ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,6 +506,7 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) @@ -518,13 +520,34 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") + -- REVIEW can't do that because we use it for + -- fingerprinting. + -- & set_mi_foreign (panic "No mi_foreign in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + -- REVIEW in @getLinkDeps@ we fall back to bytecode when the HMI + -- doesn't have object code, even if the flag is not given – + -- what's the rule? Should we provide it unconditionally if it + -- exists? + | prefer_bytecode + , Just action <- loadIfaceByteCode hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have @extra_decls@ + -- so @getLinkDeps@ knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,8 +559,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, - eps_PTT = - extendModuleEnv (eps_PTT eps) mod (mkNameEnv new_eps_decls), + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -700,7 +722,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -847,7 +869,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -877,7 +899,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -902,7 +924,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -35,7 +35,6 @@ import GHC.Unit.Env import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.Module.ModIface -import GHC.Unit.Module.WholeCoreBindings import GHC.Unit.Module.Deps import GHC.Unit.Module.Graph import GHC.Unit.Home.ModInfo @@ -56,26 +55,20 @@ import Data.List (isSuffixOf) import System.FilePath import System.Directory -import GHC.Driver.Env -import {-# SOURCE #-} GHC.Driver.Main -import Data.Time.Clock -import GHC.Driver.Flags -import GHC.Driver.Session data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function - , ldHscEnv :: !HscEnv + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -269,8 +262,8 @@ get_link_deps opts pls maybe_normal_osuf span mods = do homeModLinkable :: HomeModInfo -> Maybe Linkable homeModLinkable hmi = if ldUseByteCode opts - then homeModInfoByteCode hmi <|> homeModInfoObject hmi - else homeModInfoObject hmi <|> homeModInfoByteCode hmi + then evalHomeModByteCode hmi <|> homeModInfoObject hmi + else homeModInfoObject hmi <|> evalHomeModByteCode hmi get_linkable osuf mod -- A home-package module | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) @@ -281,39 +274,21 @@ get_link_deps opts pls maybe_normal_osuf span mods = do case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod - | prefer_bytecode = do - Succeeded iface <- ldLoadIface opts (text "load core bindings") mod - case mi_extra_decls iface of - Just extra_decls -> do - t <- getCurrentTime - let - stubs = mi_foreign iface - wcb = WholeCoreBindings extra_decls mod loc stubs - linkable = Linkable t mod (pure (CoreBindings wcb)) - initWholeCoreBindingsEps hsc_env iface linkable - _ -> fallback_no_bytecode loc mod - | otherwise = fallback_no_bytecode loc mod - - fallback_no_bytecode loc mod = do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - - prefer_bytecode = gopt Opt_UseBytecodeRatherThanObjects dflags - - dflags = hsc_dflags hsc_env - - hsc_env = ldHscEnv opts + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do @@ -338,9 +313,6 @@ get_link_deps opts pls maybe_normal_osuf span mods = do DotA fp -> panic ("adjust_ul DotA " ++ show fp) DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,19 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags - , ldHscEnv = hsc_env + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Linker/Types.hs ===================================== @@ -65,7 +65,6 @@ import Data.Time ( UTCTime ) import GHC.Unit.Module.Env import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM -import GHC.Unit.Module.WholeCoreBindings import Data.Maybe (mapMaybe) import Data.List.NonEmpty (NonEmpty, nonEmpty) import qualified Data.List.NonEmpty as NE @@ -284,18 +283,6 @@ data LinkablePart | DotDLL FilePath -- ^ Dynamically linked library file (.so, .dll, .dylib) - | CoreBindings WholeCoreBindings - -- ^ Serialised core which we can turn into BCOs (or object files), or - -- used by some other backend See Note [Interface Files with Core - -- Definitions] - - | LazyBCOs - CompiledByteCode - -- ^ Some BCOs generated on-demand when forced. This is used for - -- WholeCoreBindings, see Note [Interface Files with Core Definitions] - [FilePath] - -- ^ Objects containing foreign stubs and files - | BCOs CompiledByteCode -- ^ A byte-code object, lives only in memory. @@ -308,8 +295,6 @@ instance Outputable LinkablePart where ppr (DotA path) = text "DotA" <+> text path ppr (DotDLL path) = text "DotDLL" <+> text path ppr (BCOs bco) = text "BCOs" <+> ppr bco - ppr (LazyBCOs{}) = text "LazyBCOs" - ppr (CoreBindings {}) = text "CoreBindings" -- | Return true if the linkable only consists of native code (no BCO) linkableIsNativeCodeOnly :: Linkable -> Bool @@ -350,8 +335,6 @@ isNativeCode = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Is the part a native library? (.so/.dll) isNativeLib :: LinkablePart -> Bool @@ -360,8 +343,6 @@ isNativeLib = \case DotA {} -> True DotDLL {} -> True BCOs {} -> False - LazyBCOs{} -> False - CoreBindings {} -> False -- | Get the FilePath of linkable part (if applicable) linkablePartPath :: LinkablePart -> Maybe FilePath @@ -369,8 +350,6 @@ linkablePartPath = \case DotO fn _ -> Just fn DotA fn -> Just fn DotDLL fn -> Just fn - CoreBindings {} -> Nothing - LazyBCOs {} -> Nothing BCOs {} -> Nothing -- | Return the paths of all object code files (.o, .a, .so) contained in this @@ -380,8 +359,6 @@ linkablePartNativePaths = \case DotO fn _ -> [fn] DotA fn -> [fn] DotDLL fn -> [fn] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Return the paths of all object files (.o) contained in this 'LinkablePart'. @@ -390,8 +367,6 @@ linkablePartObjectPaths = \case DotO fn _ -> [fn] DotA _ -> [] DotDLL _ -> [] - CoreBindings {} -> [] - LazyBCOs _ fos -> fos BCOs {} -> [] -- | Retrieve the compiled byte-code from the linkable part. @@ -400,7 +375,6 @@ linkablePartObjectPaths = \case linkablePartAllBCOs :: LinkablePart -> [CompiledByteCode] linkablePartAllBCOs = \case BCOs bco -> [bco] - LazyBCOs bcos _ -> [bcos] _ -> [] linkableFilter :: (LinkablePart -> [LinkablePart]) -> Linkable -> Maybe Linkable @@ -413,13 +387,11 @@ linkablePartNative = \case u at DotO {} -> [u] u at DotA {} -> [u] u at DotDLL {} -> [u] - LazyBCOs _ os -> [DotO f ForeignObject | f <- os] _ -> [] linkablePartByteCode :: LinkablePart -> [LinkablePart] linkablePartByteCode = \case u at BCOs {} -> [u] - LazyBCOs bcos _ -> [BCOs bcos] _ -> [] -- | Transform the 'LinkablePart' list in this 'Linkable' to contain only ===================================== compiler/GHC/Runtime/Eval.hs ===================================== @@ -1275,14 +1275,14 @@ showModule mod_summary = let interpreted = case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> isJust (homeModInfoByteCode mod_info) && isNothing (homeModInfoObject mod_info) + Just mod_info -> homeModInfoHasByteCode mod_info && isNothing (homeModInfoObject mod_info) return (showSDoc dflags $ showModMsg dflags interpreted (ModuleNode [] mod_summary)) moduleIsBootOrNotObjectLinkable :: GhcMonad m => ModSummary -> m Bool moduleIsBootOrNotObjectLinkable mod_summary = withSession $ \hsc_env -> case lookupHug (hsc_HUG hsc_env) (ms_unitid mod_summary) (ms_mod_name mod_summary) of Nothing -> panic "missing linkable" - Just mod_info -> return . isNothing $ homeModInfoByteCode mod_info + Just mod_info -> return . not $ homeModInfoHasByteCode mod_info ---------------------------------------------------------------------------- -- RTTI primitives ===================================== compiler/GHC/Runtime/Interpreter.hs ===================================== @@ -684,7 +684,8 @@ fromEvalResult (EvalSuccess a) = return a getModBreaks :: HomeModInfo -> ModBreaks getModBreaks hmi - | Just linkable <- homeModInfoByteCode hmi, + -- This ignores lazy bytecode from interfaces, see #25230 + | HomeModByteCode linkable <- homeModInfoByteCode hmi, -- The linkable may have 'DotO's as well; only consider BCOs. See #20570. [cbc] <- linkableBCOs linkable = fromMaybe emptyModBreaks (bc_breaks cbc) ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -45,8 +47,6 @@ type PackageCompleteMatches = CompleteMatches type PackageIfaceTable = ModuleEnv ModIface -- Domain = modules in the imported packages -type PackageTypeTable = ModuleEnv TypeEnv - -- | Constructs an empty PackageIfaceTable emptyPackageIfaceTable :: PackageIfaceTable emptyPackageIfaceTable = emptyModuleEnv @@ -70,7 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv - , eps_PTT = emptyModuleEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -142,7 +142,11 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules - eps_PTT :: !PackageTypeTable, + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules ===================================== compiler/GHC/Unit/Home/ModInfo.hs ===================================== @@ -1,13 +1,19 @@ +{-# LANGUAGE LambdaCase #-} + -- | Info about modules in the "home" unit module GHC.Unit.Home.ModInfo ( HomeModInfo (..) - , HomeModLinkable(..) + , HomeModLinkable (..) + , HomeModByteCode (..) , homeModInfoObject , homeModInfoByteCode + , homeModInfoHasByteCode , emptyHomeModInfoLinkable , justBytecode , justObjects , bytecodeAndObjects + , pureHomeModByteCode + , evalHomeModByteCode , HomePackageTable , emptyHomePackageTable , lookupHpt @@ -34,6 +40,7 @@ import GHC.Prelude import GHC.Unit.Module.ModIface import GHC.Unit.Module.ModDetails import GHC.Unit.Module +import GHC.Unit.Module.WholeCoreBindings (WholeCoreBindings) import GHC.Linker.Types ( Linkable(..), linkableIsNativeCodeOnly ) @@ -73,37 +80,79 @@ data HomeModInfo = HomeModInfo -- 'ModIface' (only). } -homeModInfoByteCode :: HomeModInfo -> Maybe Linkable +homeModInfoByteCode :: HomeModInfo -> HomeModByteCode homeModInfoByteCode = homeMod_bytecode . hm_linkable +homeModInfoHasByteCode :: HomeModInfo -> Bool +homeModInfoHasByteCode hmi = case homeModInfoByteCode hmi of + NoHomeModByteCode -> False + _ -> True + homeModInfoObject :: HomeModInfo -> Maybe Linkable homeModInfoObject = homeMod_object . hm_linkable emptyHomeModInfoLinkable :: HomeModLinkable -emptyHomeModInfoLinkable = HomeModLinkable Nothing Nothing +emptyHomeModInfoLinkable = HomeModLinkable NoHomeModByteCode Nothing -- See Note [Home module build products] -data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !(Maybe Linkable) +data HomeModLinkable = HomeModLinkable { homeMod_bytecode :: !HomeModByteCode , homeMod_object :: !(Maybe Linkable) } instance Outputable HomeModLinkable where ppr (HomeModLinkable l1 l2) = ppr l1 $$ ppr l2 -justBytecode :: Linkable -> HomeModLinkable -justBytecode lm = - assertPpr (not (linkableIsNativeCodeOnly lm)) (ppr lm) - $ emptyHomeModInfoLinkable { homeMod_bytecode = Just lm } +justBytecode :: HomeModByteCode -> HomeModLinkable +justBytecode bc = + emptyHomeModInfoLinkable { homeMod_bytecode = bc } justObjects :: Linkable -> HomeModLinkable justObjects lm = assertPpr (linkableIsNativeCodeOnly lm) (ppr lm) $ emptyHomeModInfoLinkable { homeMod_object = Just lm } -bytecodeAndObjects :: Linkable -> Linkable -> HomeModLinkable +bytecodeAndObjects :: HomeModByteCode -> Linkable -> HomeModLinkable bytecodeAndObjects bc o = - assertPpr (not (linkableIsNativeCodeOnly bc) && linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) - (HomeModLinkable (Just bc) (Just o)) - + assertPpr (linkableIsNativeCodeOnly o) (ppr bc $$ ppr o) + (HomeModLinkable bc (Just o)) + +pureHomeModByteCode :: HomeModByteCode -> Maybe Linkable +pureHomeModByteCode = \case + NoHomeModByteCode -> Nothing + HomeModIfaceCore _ -> Nothing + HomeModLazyByteCode {} -> Nothing + HomeModByteCode l -> Just l + +-- | Obtain the bytecode stored in this 'HomeModInfo', preferring the value in +-- 'HomeModLinkable' that's already in memory before evaluating the lazy thunk +-- in 'HomeModLazyByteCode' that hydrates and parses Core loaded from an +-- interface. +-- +-- This should only be called once in the module's lifecycle; afterwards, the +-- bytecode is cached in 'LoaderState'. +evalHomeModByteCode :: HomeModInfo -> Maybe Linkable +evalHomeModByteCode HomeModInfo {hm_linkable} + | HomeModByteCode bc <- homeMod_bytecode hm_linkable + = Just bc + | HomeModLazyByteCode bc <- homeMod_bytecode hm_linkable + = Just bc + | otherwise + = Nothing + +data HomeModByteCode = + NoHomeModByteCode + | + HomeModIfaceCore !WholeCoreBindings + | + HomeModByteCode !Linkable + | + HomeModLazyByteCode Linkable + +instance Outputable HomeModByteCode where + ppr = \case + NoHomeModByteCode -> text "no bytecode" + HomeModIfaceCore {} -> text "dehydrated Core" + HomeModByteCode linkable -> ppr linkable + HomeModLazyByteCode {} -> text "lazy bytecode" {- Note [Home module build products] ===================================== ghc/GHCi/Leak.hs ===================================== @@ -50,7 +50,7 @@ getLeakIndicators hsc_env = where mkWeakLinkables :: HomeModLinkable -> IO [Maybe (Weak Linkable)] mkWeakLinkables (HomeModLinkable mbc mo) = - mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [mbc, mo] + mapM (\ln -> traverse (flip mkWeakPtr Nothing <=< evaluate) ln) [pureHomeModByteCode mbc, mo] -- | Look at the LeakIndicators collected by an earlier call to -- `getLeakIndicators`, and print messasges if any of them are still View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8baf8f1f3170aec0248b662e204aee884b0f8305...6ae0c945cdfab81c3f085b65f05c11cd1277f3f8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8baf8f1f3170aec0248b662e204aee884b0f8305...6ae0c945cdfab81c3f085b65f05c11cd1277f3f8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 19:02:56 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Tue, 03 Sep 2024 15:02:56 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] 17 commits: git: remove a.out and include it in .gitignore Message-ID: <66d75d60efb9f_bda384b9880900cd@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 4120f8d6 by Cheng Shao at 2024-09-03T21:00:41+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). Metric Decrease: MultiLayerModules T13701 - - - - - caf03ccd by Torsten Schmits at 2024-09-03T21:02:47+02:00 store IO in the EPS - - - - - 06528df5 by Torsten Schmits at 2024-09-03T21:02:49+02:00 Move lazy bytecode storage from Linkable to HomeModLinkable This shifts the responsibility of handling the laziness properly from the rather generic Linkable to a type that is more specific to the domain that necessitates the laziness. - - - - - 30 changed files: - .gitignore - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - − a.out - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Home/ModInfo.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - configure.ac The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6ae0c945cdfab81c3f085b65f05c11cd1277f3f8...06528df5b1486c52a06a582031e5ca82de011eaa -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6ae0c945cdfab81c3f085b65f05c11cd1277f3f8...06528df5b1486c52a06a582031e5ca82de011eaa You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 20:49:21 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 03 Sep 2024 16:49:21 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 5 commits: Add functions to check for weakly pinned arrays. Message-ID: <66d776511917f_bda38dab20411085d@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 2f10c710 by Cheng Shao at 2024-09-03T16:49:04-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - 50ce9af1 by Sven Tennie at 2024-09-03T16:49:05-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - 19 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/ffi.rst - libraries/base/src/GHC/Base.hs - libraries/base/src/GHC/Exts.hs - libraries/ghc-prim/changelog.md - rts/PrimOps.cmm - rts/RtsSymbols.c - rts/include/stg/MiscClosures.h - rts/sm/Sanity.c - testsuite/tests/rts/T13894.hs - utils/ghc-toolchain/src/GHC/Toolchain/Tools/Cpp.hs Changes: ===================================== compiler/GHC/Builtin/primops.txt.pp ===================================== @@ -1925,7 +1925,25 @@ primop MutableByteArrayIsPinnedOp "isMutableByteArrayPinned#" GenPrimOp primop ByteArrayIsPinnedOp "isByteArrayPinned#" GenPrimOp ByteArray# -> Int# - {Determine whether a 'ByteArray#' is guaranteed not to move during GC.} + {Determine whether a 'ByteArray#' is guaranteed not to move.} + with out_of_line = True + +primop ByteArrayIsWeaklyPinnedOp "isByteArrayWeaklyPinned#" GenPrimOp + ByteArray# -> Int# + {Similar to 'isByteArrayPinned#'. Weakly pinned byte arrays are allowed + to be copied into compact regions by the user, potentially invalidating + the results of earlier calls to 'byteArrayContents#'. + + See the section `Pinned Byte Arrays` in the user guide for more information. + + This function also returns true for regular pinned bytearrays. + } + with out_of_line = True + +primop MutableByteArrayIsWeaklyPinnedOp "isMutableByteArrayWeaklyPinned#" GenPrimOp + MutableByteArray# s -> Int# + { 'isByteArrayWeaklyPinned#' but for mutable arrays. + } with out_of_line = True primop ByteArrayContents_Char "byteArrayContents#" GenPrimOp ===================================== compiler/GHC/CmmToAsm/AArch64/CodeGen.hs ===================================== @@ -1990,9 +1990,13 @@ genCCall target dest_regs arg_regs = do MO_SubIntC _w -> unsupported mop -- Memory Ordering - MO_AcquireFence -> return (unitOL DMBISH) - MO_ReleaseFence -> return (unitOL DMBISH) - MO_SeqCstFence -> return (unitOL DMBISH) + -- Set flags according to their C pendants (stdatomic.h): + -- atomic_thread_fence(memory_order_acquire); // -> dmb ishld + MO_AcquireFence -> return . unitOL $ DMBISH DmbLoad + -- atomic_thread_fence(memory_order_release); // -> dmb ish + MO_ReleaseFence -> return . unitOL $ DMBISH DmbLoadStore + -- atomic_thread_fence(memory_order_seq_cst); // -> dmb ish + MO_SeqCstFence -> return . unitOL $ DMBISH DmbLoadStore MO_Touch -> return nilOL -- Keep variables live (when using interior pointers) -- Prefetch MO_Prefetch_Data _n -> return nilOL -- Prefetch hint. ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -134,7 +134,7 @@ regUsageOfInstr platform instr = case instr of LDAR _ dst src -> usage (regOp src, regOp dst) -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> usage ([], []) + DMBISH _ -> usage ([], []) -- 9. Floating Point Instructions -------------------------------------------- FMOV dst src -> usage (regOp src, regOp dst) @@ -281,7 +281,7 @@ patchRegsOfInstr instr env = case instr of LDAR f o1 o2 -> LDAR f (patchOp o1) (patchOp o2) -- 8. Synchronization Instructions ----------------------------------------- - DMBISH -> DMBISH + DMBISH c -> DMBISH c -- 9. Floating Point Instructions ------------------------------------------ FMOV o1 o2 -> FMOV (patchOp o1) (patchOp o2) @@ -649,7 +649,7 @@ data Instr | BCOND Cond Target -- branch with condition. b. -- 8. Synchronization Instructions ----------------------------------------- - | DMBISH + | DMBISH DMBISHFlags -- 9. Floating Point Instructions -- move to/from general purpose <-> floating, or floating to floating | FMOV Operand Operand @@ -672,6 +672,9 @@ data Instr -- - fnmadd: d = - r1 * r2 - r3 | FMA FMASign Operand Operand Operand Operand +data DMBISHFlags = DmbLoad | DmbLoadStore + deriving (Eq, Show) + instrCon :: Instr -> String instrCon i = case i of ===================================== compiler/GHC/CmmToAsm/AArch64/Ppr.hs ===================================== @@ -527,7 +527,8 @@ pprInstr platform instr = case instr of LDAR _f o1 o2 -> op2 (text "\tldar") o1 o2 -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> line $ text "\tdmb ish" + DMBISH DmbLoadStore -> line $ text "\tdmb ish" + DMBISH DmbLoad -> line $ text "\tdmb ishld" -- 9. Floating Point Instructions -------------------------------------------- FMOV o1 o2 -> op2 (text "\tfmov") o1 o2 ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -1,5 +1,4 @@ {-# LANGUAGE LambdaCase #-} - {-# LANGUAGE NondecreasingIndentation #-} {-# LANGUAGE GADTs #-} {-# LANGUAGE MultiWayIf #-} @@ -295,13 +294,13 @@ import Data.Time import System.IO.Unsafe ( unsafeInterleaveIO ) import GHC.Iface.Env ( trace_if ) +import GHC.Platform.Ways import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) import GHC.Types.Unique.FM import GHC.Types.Unique.DFM import GHC.Cmm.Config (CmmConfig) - {- ********************************************************************** %* * Initialisation @@ -990,6 +989,27 @@ initModDetails hsc_env iface = -- in make mode, since this HMI will go into the HPT. genModDetails hsc_env' iface +-- | Modify flags such that objects are compiled for the interpreter's way. +-- This is necessary when building foreign objects for Template Haskell, since +-- those are object code built outside of the pipeline, which means they aren't +-- subject to the mechanism in 'enableCodeGenWhen' that requests dynamic build +-- outputs for dependencies when the interpreter used for TH is dynamic but the +-- main outputs aren't. +-- Furthermore, the HPT only stores one set of objects with different names for +-- bytecode linking in 'HomeModLinkable', so the usual hack for switching +-- between ways in 'get_link_deps' doesn't work. +compile_for_interpreter :: HscEnv -> (HscEnv -> IO a) -> IO a +compile_for_interpreter hsc_env use = + use (hscUpdateFlags update hsc_env) + where + update dflags = dflags { + targetWays_ = adapt_way interpreterDynamic WayDyn $ + adapt_way interpreterProfiled WayProf $ + targetWays_ dflags + } + + adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace -- them with a lazy IO thunk that compiles them to bytecode and foreign objects. -- @@ -2063,9 +2083,10 @@ generateByteCode :: HscEnv -> IO (CompiledByteCode, [FilePath]) generateByteCode hsc_env cgguts mod_location = do (hasStub, comp_bc) <- hscInteractive hsc_env cgguts mod_location - stub_o <- traverse (compileForeign hsc_env LangC) hasStub - foreign_files_o <- traverse (uncurry (compileForeign hsc_env)) (cgi_foreign_files cgguts) - pure (comp_bc, maybeToList stub_o ++ foreign_files_o) + compile_for_interpreter hsc_env $ \ i_env -> do + stub_o <- traverse (compileForeign i_env LangC) hasStub + foreign_files_o <- traverse (uncurry (compileForeign i_env)) (cgi_foreign_files cgguts) + pure (comp_bc, maybeToList stub_o ++ foreign_files_o) generateFreshByteCode :: HscEnv -> ModuleName ===================================== compiler/GHC/StgToCmm/Prim.hs ===================================== @@ -1668,10 +1668,12 @@ emitPrimOp cfg primop = NewPinnedByteArrayOp_Char -> alwaysExternal NewAlignedPinnedByteArrayOp_Char -> alwaysExternal MutableByteArrayIsPinnedOp -> alwaysExternal + MutableByteArrayIsWeaklyPinnedOp -> alwaysExternal DoubleDecode_2IntOp -> alwaysExternal DoubleDecode_Int64Op -> alwaysExternal FloatDecode_IntOp -> alwaysExternal ByteArrayIsPinnedOp -> alwaysExternal + ByteArrayIsWeaklyPinnedOp -> alwaysExternal ShrinkMutableByteArrayOp_Char -> alwaysExternal ResizeMutableByteArrayOp_Char -> alwaysExternal ShrinkSmallMutableArrayOp_Char -> alwaysExternal ===================================== compiler/GHC/StgToJS/Prim.hs ===================================== @@ -670,6 +670,8 @@ genPrim prof bound ty op = case op of NewAlignedPinnedByteArrayOp_Char -> \[r] [l,_align] -> pure $ PrimInline (newByteArray r l) MutableByteArrayIsPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ ByteArrayIsPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ + ByteArrayIsWeaklyPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ + MutableByteArrayIsWeaklyPinnedOp -> \[r] [_] -> pure $ PrimInline $ r |= one_ ByteArrayContents_Char -> \[a,o] [b] -> pure $ PrimInline $ mconcat [a |= b, o |= zero_] MutableByteArrayContents_Char -> \[a,o] [b] -> pure $ PrimInline $ mconcat [a |= b, o |= zero_] ShrinkMutableByteArrayOp_Char -> \[] [a,n] -> pure $ PrimInline $ appS hdShrinkMutableByteArrayStr [a,n] ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -190,13 +190,9 @@ Even if that wasn't an issue, they are compiled for the session's 'Way', not its associated module's, so the dynamic variant wouldn't be available when building only static outputs. -For now, this doesn't have much of an impact, since we're only supporting -foreign imports initially, which produce very simple objects that can easily be -handled by the linker when 'GHC.Linker.Loader.dynLoadObjs' creates a shared -library from all object file inputs. -However, for more complex circumstances, we should compile foreign stubs -specially for TH according to the interpreter 'Way', or request dynamic products -for TH dependencies like it happens for the conventional case. +To mitigate this, we instead build foreign objects specially for the +interpreter, updating the build flags in 'compile_for_interpreter' to use the +interpreter's way. Problem 4: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -163,6 +163,12 @@ Runtime system ~~~~~~~~~~~~~~~~~~~~ - Usage of deprecated primops is now correctly reported (#19629). +- New primops `isMutableByteArrayWeaklyPinned#` and `isByteArrayWeaklyPinned#` + to allow users to avoid copying large arrays safely when dealing with ffi. + See the users guide for more details on the different kinds of + pinned arrays in 9.12. + + This need for this distinction originally surfaced in https://gitlab.haskell.org/ghc/ghc/-/issues/22255 ``ghc`` library ===================================== docs/users_guide/exts/ffi.rst ===================================== @@ -1114,21 +1114,67 @@ Pinned Byte Arrays A pinned byte array is one that the garbage collector is not allowed to move. Consequently, it has a stable address that can be safely -requested with ``byteArrayContents#``. Not that being pinned doesn't -prevent the byteArray from being gc'ed in the same fashion a regular -byte array would be. +requested with ``byteArrayContents#``. As long as the array remains live +the address returned by ``byteArrayContents#`` will remain valid. Note that +being pinned doesn't prevent the byteArray from being gc'ed in the same fashion +a regular byte array would be if there are no more references to the ``ByteArray#``. There are a handful of primitive functions in :base-ref:`GHC.Exts.` used to enforce or check for pinnedness: ``isByteArrayPinned#``, -``isMutableByteArrayPinned#``, and ``newPinnedByteArray#``. A -byte array can be pinned as a result of three possible causes: +``isMutableByteArrayPinned#``, ``isByteArrayWeaklyPinned#``, +``isMutableByteArrayWeaklyPinned#``, and ``newPinnedByteArray#``. A +byte array can be pinned or weakly pinned as a result of three possible causes: -1. It was allocated by ``newPinnedByteArray#``. -2. It is large. Currently, GHC defines large object to be one +1. It was allocated by ``newPinnedByteArray#``. This results in a regular pinned byte array. +2. It is large, this results in a weakly pinned byte array. Currently, GHC defines large object to be one that is at least as large as 80% of a 4KB block (i.e. at least 3277 bytes). -3. It has been copied into a compact region. The documentation +3. It has been copied into a compact region, resulting in a weakly pinned array. The documentation for ``ghc-compact`` and ``compact`` describes this process. +The difference between a pinned array and a weakly pinned array is simply that +trying to compact a pinned array will result in an exception. Trying to compact +a weakly pinned array will succeed. However the result of earlier +calls to ``byteArrayContents#`` is not updated during compaction, which means +these results will still point to the address where the array was located originally, +and not to the new address inside the compact region. + +This is particularly dangerous when an address to a byte arrays content is stored +inside a datastructure along with a reference to the byte array. +If the data structure is compacted later on the pointer won't be updated but the +reference to the byte array will point to a copy inside the compact region. +A common data type susceptible to this is `ForeignPtr` when used to represent a ByteArray#. + +Here is an example to illustrate this: + +.. code-block:: haskell + + workWithArrayContents :: (ByteArray, Ptr Word8) -> (Ptr Word8 -> IO ()) -> IO () + workWithArrayContents (arr@(ByteArray uarr),ptr) worker = + case () of + _ + -- Conservative but safe + | isByteArrayPinned arr -> keepAliveUnlifted uarr (worker ptr) + -- Potentially dangerous, the program needs to ensures the Ptr points into the array. + | isByteArrayWeaklyPinned arr -> keepAliveUnlifted uarr (worker ptr) + | otherwise -> ... -- Otherwise we can't directly use it for safe FFI calls directly at all. + + main :: IO () + main = do + -- We create a large array, which causes it to be implicitly pinned + arr <- newByteArray 5000 + arr@(ByteArray uarr) <- freezeByteArray arr 0 5000 -- Make it immutable + let ptr = byteArrayContents arr + + -- Compacting a data structure that contains both an array and a ptr to + -- the arrays content's is dangerous and usually the wrong thing to do. + let foo = (arr, ptr) + foo_compacted <- compact foo + + -- This is fine + workWithArrayContents foo do_work + -- This is unsound + workWithArrayContents (getCompact foo_compacted) do_work + .. [1] Prior to GHC 8.10, when passing an ``ArrayArray#`` argument to a foreign function, the foreign function would see a pointer to the ``StgMutArrPtrs`` rather than just the payload. ===================================== libraries/base/src/GHC/Base.hs ===================================== @@ -139,7 +139,8 @@ module GHC.Base ) where import GHC.Internal.Base -import GHC.Prim hiding (dataToTagLarge#, dataToTagSmall#, whereFrom#) +import GHC.Prim hiding (dataToTagLarge#, dataToTagSmall#, whereFrom#, + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned#) -- Hide dataToTagLarge# because it is expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -112,7 +112,9 @@ module GHC.Exts import GHC.Internal.Exts import GHC.Internal.ArrayArray -import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# ) +import GHC.Prim hiding ( + coerce, dataToTagSmall#, dataToTagLarge#, whereFrom#, + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/ghc-prim/changelog.md ===================================== @@ -1,3 +1,12 @@ +## 0.13.0 + +- Shipped with GHC 9.12.1 + +- Add primops that allow users to distinguish weakly pinned byte arrays from unpinned ones. + + isMutableByteArrayWeaklyPinned# :: MutableByteArray# s -> Int# + isByteArrayWeaklyPinned# :: ByteArray# s -> Int# + ## 0.12.0 - Shipped with GHC 9.10.1 ===================================== rts/PrimOps.cmm ===================================== @@ -215,12 +215,29 @@ stg_isByteArrayPinnedzh ( gcptr ba ) return (flags & BF_PINNED != 0); } +stg_isByteArrayWeaklyPinnedzh ( gcptr ba ) +// ByteArray# s -> Int# +{ + W_ bd, flags; + bd = Bdescr(ba); + // See #22255 and the primop docs. + flags = TO_W_(bdescr_flags(bd)); + + return (flags & (BF_PINNED | BF_COMPACT | BF_LARGE) != 0); +} + stg_isMutableByteArrayPinnedzh ( gcptr mba ) // MutableByteArray# s -> Int# { jump stg_isByteArrayPinnedzh(mba); } +stg_isMutableByteArrayWeaklyPinnedzh ( gcptr mba ) +// MutableByteArray# s -> Int# +{ + jump stg_isByteArrayWeaklyPinnedzh(mba); +} + /* Note [LDV profiling and resizing arrays] * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * As far as the LDV profiler is concerned arrays are "inherently used" which ===================================== rts/RtsSymbols.c ===================================== @@ -656,6 +656,8 @@ extern char **environ; SymI_HasDataProto(stg_newAlignedPinnedByteArrayzh) \ SymI_HasDataProto(stg_isByteArrayPinnedzh) \ SymI_HasDataProto(stg_isMutableByteArrayPinnedzh) \ + SymI_HasDataProto(stg_isByteArrayWeaklyPinnedzh) \ + SymI_HasDataProto(stg_isMutableByteArrayWeaklyPinnedzh) \ SymI_HasDataProto(stg_shrinkMutableByteArrayzh) \ SymI_HasDataProto(stg_resizzeMutableByteArrayzh) \ SymI_HasDataProto(stg_shrinkSmallMutableArrayzh) \ ===================================== rts/include/stg/MiscClosures.h ===================================== @@ -454,6 +454,8 @@ RTS_FUN_DECL(stg_newPinnedByteArrayzh); RTS_FUN_DECL(stg_newAlignedPinnedByteArrayzh); RTS_FUN_DECL(stg_isByteArrayPinnedzh); RTS_FUN_DECL(stg_isMutableByteArrayPinnedzh); +RTS_FUN_DECL(stg_isByteArrayWeaklyPinnedzh); +RTS_FUN_DECL(stg_isMutableByteArrayWeaklyPinnedzh); RTS_FUN_DECL(stg_shrinkMutableByteArrayzh); RTS_FUN_DECL(stg_resizzeMutableByteArrayzh); RTS_FUN_DECL(stg_shrinkSmallMutableArrayzh); ===================================== rts/sm/Sanity.c ===================================== @@ -357,7 +357,8 @@ checkClosure( const StgClosure* p ) info = ACQUIRE_LOAD(&p->header.info); if (IS_FORWARDING_PTR(info)) { - barf("checkClosure: found EVACUATED closure %d", info->type); + ASSERT(LOOKS_LIKE_CLOSURE_PTR(info)); + barf("checkClosure: found EVACUATED closure %u", GET_INFO((StgClosure*)UN_FORWARDING_PTR(info))->type); } #if defined(PROFILING) ===================================== testsuite/tests/rts/T13894.hs ===================================== @@ -6,6 +6,7 @@ import Control.Monad import GHC.Exts +import GHC.Internal.Exts (isMutableByteArrayWeaklyPinned#) import GHC.IO main :: IO () @@ -16,3 +17,10 @@ main = do case isMutableByteArrayPinned# arr# of n# -> (# s1, isTrue# n# #) when pinned $ putStrLn "BAD" + + weakly_pinned <- IO $ \s0 -> + case newByteArray# 1000000# s0 of + (# s1, arr# #) -> + case isMutableByteArrayWeaklyPinned# arr# of + n# -> (# s1, isTrue# n# #) + when (not weakly_pinned) $ putStrLn "BAD" ===================================== utils/ghc-toolchain/src/GHC/Toolchain/Tools/Cpp.hs ===================================== @@ -39,10 +39,10 @@ data CmmCpp = CmmCpp { cmmCppProgram :: Program } deriving (Show, Read, Eq, Ord) -checkFlag :: String -> Program -> String -> M () -checkFlag conftest cpp flag = checking ("for "++flag++" support") $ +checkFlag :: String -> Program -> String -> [String] -> M () +checkFlag conftest cpp flag extra_args = checking ("for "++flag++" support") $ -- Werror to ensure that unrecognized warnings result in an error - callProgram cpp ["-Werror", flag, conftest] + callProgram cpp $ ["-Werror", flag, conftest] ++ extra_args -- tryFlag :: String -> Program -> String -> M [String] -- tryFlag conftest cpp flag = -- ([flag] <$ checkFlag conftest cpp flag) <|> return [] @@ -167,7 +167,7 @@ findCmmCpp progOpt cc = checking "for a Cmm preprocessor" $ do cmmCppSupportsG0 <- withTempDir $ \dir -> do let conftest = dir "conftest.c" writeFile conftest "int main(void) {}" - True <$ checkFlag conftest cpp "-g0" <|> return False + True <$ checkFlag conftest cpp "-g0" ["-o", dir "conftest"] <|> return False -- Always add the -E flag to the CPP, regardless of the user options let cmmCppProgram = foldr addFlagIfNew cpp ["-E"] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/867611279614f3fd44f90c92404a7a08b7a0d0a0...50ce9af142800d839d5f988d7db06499dd86115c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/867611279614f3fd44f90c92404a7a08b7a0d0a0...50ce9af142800d839d5f988d7db06499dd86115c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 3 20:52:56 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Tue, 03 Sep 2024 16:52:56 -0400 Subject: [Git][ghc/ghc][wip/T23479] 11 commits: Haddock: Add no-compilation flag Message-ID: <66d7772842781_bda38f86ee8116013@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - df15c3be by Serge S. Gulin at 2024-09-03T23:52:18+03:00 JS: Re-add optimization for literal strings in genApp (fixes 23479 (muted temporary)) Based on https://gitlab.haskell.org/ghc/ghc/-/merge_requests/10588/ - - - - - 55f1d5de by Serge S. Gulin at 2024-09-03T23:52:18+03:00 Use name defined at `GHC.Builtin.Names` - - - - - e90008bf by Serge S. Gulin at 2024-09-03T23:52:18+03:00 Apply 1 suggestion(s) to 1 file(s) Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 6c4d5713 by Serge S. Gulin at 2024-09-03T23:52:18+03:00 Attempt to take 805 for id - - - - - f46aa9c7 by Serge S. Gulin at 2024-09-03T23:52:18+03:00 Attempt to add to basicKnownKeyNames Co-authored-by: Andrei Borzenkov <root at sandwitch.dev> Co-authored-by: Danil Berestov <goosedb at yandex.ru> - - - - - f40ec3d1 by Serge S. Gulin at 2024-09-03T23:52:18+03:00 Naive attempt to add `StgLitArg (LitString bs)` - - - - - f601a3ef by Serge S. Gulin at 2024-09-03T23:52:18+03:00 WIP add logging - - - - - 30 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Driver/Main.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Apply.hs - compiler/GHC/StgToJS/Prim.hs - compiler/GHC/StgToJS/Sinker.hs - compiler/GHC/StgToJS/Symbols.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/ffi.rst - libraries/base/src/GHC/Base.hs - libraries/base/src/GHC/Exts.hs - libraries/ghc-prim/changelog.md - rts/PrimOps.cmm - rts/RtsSymbols.c - rts/include/stg/MiscClosures.h - testsuite/tests/javascript/Makefile - + testsuite/tests/javascript/T23479.hs - + testsuite/tests/javascript/T23479.stdout - testsuite/tests/javascript/all.T - testsuite/tests/rts/T13894.hs - utils/ghc-toolchain/src/GHC/Toolchain/Tools/Cpp.hs - utils/haddock/.gitignore - utils/haddock/CHANGES.md - utils/haddock/doc/invoking.rst - utils/haddock/haddock-api/src/Haddock.hs - utils/haddock/haddock-api/src/Haddock/Interface.hs - utils/haddock/haddock-api/src/Haddock/Options.hs - utils/haddock/haddock-test/src/Test/Haddock.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/982efb5a014d03a2dc923096448908e542df1957...f601a3efabfcdea2c92daa05c8692c1f4831eefb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/982efb5a014d03a2dc923096448908e542df1957...f601a3efabfcdea2c92daa05c8692c1f4831eefb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 07:00:05 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 04 Sep 2024 03:00:05 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 2 commits: rts: fix checkClosure error message Message-ID: <66d805759d4a1_3132270495073111@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: e58eade2 by Cheng Shao at 2024-09-04T02:59:58-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - 17961736 by Sven Tennie at 2024-09-04T02:59:59-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - 4 changed files: - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - rts/sm/Sanity.c Changes: ===================================== compiler/GHC/CmmToAsm/AArch64/CodeGen.hs ===================================== @@ -1990,9 +1990,13 @@ genCCall target dest_regs arg_regs = do MO_SubIntC _w -> unsupported mop -- Memory Ordering - MO_AcquireFence -> return (unitOL DMBISH) - MO_ReleaseFence -> return (unitOL DMBISH) - MO_SeqCstFence -> return (unitOL DMBISH) + -- Set flags according to their C pendants (stdatomic.h): + -- atomic_thread_fence(memory_order_acquire); // -> dmb ishld + MO_AcquireFence -> return . unitOL $ DMBISH DmbLoad + -- atomic_thread_fence(memory_order_release); // -> dmb ish + MO_ReleaseFence -> return . unitOL $ DMBISH DmbLoadStore + -- atomic_thread_fence(memory_order_seq_cst); // -> dmb ish + MO_SeqCstFence -> return . unitOL $ DMBISH DmbLoadStore MO_Touch -> return nilOL -- Keep variables live (when using interior pointers) -- Prefetch MO_Prefetch_Data _n -> return nilOL -- Prefetch hint. ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -134,7 +134,7 @@ regUsageOfInstr platform instr = case instr of LDAR _ dst src -> usage (regOp src, regOp dst) -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> usage ([], []) + DMBISH _ -> usage ([], []) -- 9. Floating Point Instructions -------------------------------------------- FMOV dst src -> usage (regOp src, regOp dst) @@ -281,7 +281,7 @@ patchRegsOfInstr instr env = case instr of LDAR f o1 o2 -> LDAR f (patchOp o1) (patchOp o2) -- 8. Synchronization Instructions ----------------------------------------- - DMBISH -> DMBISH + DMBISH c -> DMBISH c -- 9. Floating Point Instructions ------------------------------------------ FMOV o1 o2 -> FMOV (patchOp o1) (patchOp o2) @@ -649,7 +649,7 @@ data Instr | BCOND Cond Target -- branch with condition. b. -- 8. Synchronization Instructions ----------------------------------------- - | DMBISH + | DMBISH DMBISHFlags -- 9. Floating Point Instructions -- move to/from general purpose <-> floating, or floating to floating | FMOV Operand Operand @@ -672,6 +672,9 @@ data Instr -- - fnmadd: d = - r1 * r2 - r3 | FMA FMASign Operand Operand Operand Operand +data DMBISHFlags = DmbLoad | DmbLoadStore + deriving (Eq, Show) + instrCon :: Instr -> String instrCon i = case i of ===================================== compiler/GHC/CmmToAsm/AArch64/Ppr.hs ===================================== @@ -527,7 +527,8 @@ pprInstr platform instr = case instr of LDAR _f o1 o2 -> op2 (text "\tldar") o1 o2 -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> line $ text "\tdmb ish" + DMBISH DmbLoadStore -> line $ text "\tdmb ish" + DMBISH DmbLoad -> line $ text "\tdmb ishld" -- 9. Floating Point Instructions -------------------------------------------- FMOV o1 o2 -> op2 (text "\tfmov") o1 o2 ===================================== rts/sm/Sanity.c ===================================== @@ -357,7 +357,8 @@ checkClosure( const StgClosure* p ) info = ACQUIRE_LOAD(&p->header.info); if (IS_FORWARDING_PTR(info)) { - barf("checkClosure: found EVACUATED closure %d", info->type); + ASSERT(LOOKS_LIKE_CLOSURE_PTR(info)); + barf("checkClosure: found EVACUATED closure %u", GET_INFO((StgClosure*)UN_FORWARDING_PTR(info))->type); } #if defined(PROFILING) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/50ce9af142800d839d5f988d7db06499dd86115c...1796173632b68c8878976fc04ac9973f16b3619d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/50ce9af142800d839d5f988d7db06499dd86115c...1796173632b68c8878976fc04ac9973f16b3619d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 07:18:05 2024 From: gitlab at gitlab.haskell.org (Hannes Siebenhandl (@fendor)) Date: Wed, 04 Sep 2024 03:18:05 -0400 Subject: [Git][ghc/ghc][wip/perf-ci] 15 commits: JS: add basic support for POSIX *at functions (#25190) Message-ID: <66d809adc9831_17857016fcb04903a@gitlab.mail> Hannes Siebenhandl pushed to branch wip/perf-ci at Glasgow Haskell Compiler / GHC Commits: 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 43c0ddea by Fendor at 2024-09-04T09:17:52+02:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - 83a27852 by Fendor at 2024-09-04T09:17:52+02:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 1172d258 by Fendor at 2024-09-04T09:17:52+02:00 Enable perf profiling for compiler performance tests - - - - - 30 changed files: - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Driver/Main.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - configure.ac - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/ffi.rst - hadrian/src/Packages.hs - hadrian/src/Rules/ToolArgs.hs - hadrian/src/Settings/Default.hs - libraries/base/src/GHC/Base.hs - libraries/base/src/GHC/Exts.hs - libraries/directory - + libraries/file-io - libraries/ghc-internal/jsbits/base.js - libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs - libraries/ghc-prim/changelog.md - rts/PrimOps.cmm - rts/RtsSymbols.c - rts/include/stg/MiscClosures.h - rts/linker/PEi386.c - testsuite/driver/perf_notes.py The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/64618037cfb7b3f07eda50a986389d6fab19f2e1...1172d25805e65e7e3e318371576555955dea326d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/64618037cfb7b3f07eda50a986389d6fab19f2e1...1172d25805e65e7e3e318371576555955dea326d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 08:52:31 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Wed, 04 Sep 2024 04:52:31 -0400 Subject: [Git][ghc/ghc][wip/fltused] LLVM: use -relocation-model=pic on Windows Message-ID: <66d81fcfdbb4_25b5688ab88857d2@gitlab.mail> sheaf pushed to branch wip/fltused at Glasgow Haskell Compiler / GHC Commits: 7bda232a by sheaf at 2024-09-04T10:50:56+02:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - 3 changed files: - compiler/GHC/Driver/Pipeline/Execute.hs - + testsuite/tests/llvm/should_run/T22487.hs - + testsuite/tests/llvm/should_run/all.T Changes: ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -969,13 +969,18 @@ llvmOptions llvm_config dflags = ++ [("", "-target-abi=" ++ abi) | not (null abi) ] where target = platformMisc_llvmTarget $ platformMisc dflags + target_os = platformOS (targetPlatform dflags) Just (LlvmTarget _ mcpu mattr) = lookup target (llvmTargets llvm_config) -- Relocation models - rmodel | gopt Opt_PIC dflags = "pic" - | positionIndependent dflags = "pic" - | ways dflags `hasWay` WayDyn = "dynamic-no-pic" - | otherwise = "static" + rmodel | gopt Opt_PIC dflags + || positionIndependent dflags + || target_os == OSMinGW32 -- #22487: use PIC on (64-bit) Windows + = "pic" + | ways dflags `hasWay` WayDyn + = "dynamic-no-pic" + | otherwise + = "static" platform = targetPlatform dflags arch = platformArch platform ===================================== testsuite/tests/llvm/should_run/T22487.hs ===================================== @@ -0,0 +1,8 @@ + +module Main where + +add :: Double -> Double -> Double +add x y = x + y +{-# NOINLINE add #-} +main = do putStrLn "Hello world!" + print (add 1.0 2.0) ===================================== testsuite/tests/llvm/should_run/all.T ===================================== @@ -0,0 +1,15 @@ + +def f( name, opts ): + opts.only_ways = ['optllvm', 'llvm', 'debugllvm'] + +setTestOpts(f) + +# Apples LLVM Toolchain knows about a `vortex` cpu (and possibly others), that +# the stock LLVM toolchain doesn't know about and will warn about. Let's not +# have test fail just because of processor name differences due to different +# LLVM Toolchains. GHC tries to pass what apple expects (on darwin), but can +# be used with the stock LLVM toolchain as well. +def ignore_llvm_and_vortex( msg ): + return re.sub(r".* is not a recognized processor for this target.*\n",r"",msg) + +test('T22487', [normal, normalise_errmsg_fun(ignore_llvm_and_vortex)], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7bda232ad966e93bb2f44594a0a04c92dd19758f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7bda232ad966e93bb2f44594a0a04c92dd19758f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 09:19:41 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Wed, 04 Sep 2024 05:19:41 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25178 Message-ID: <66d8262d21b4f_25b56835f9589722@gitlab.mail> sheaf pushed new branch wip/T25178 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25178 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 09:46:18 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 04 Sep 2024 05:46:18 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/typo-binds Message-ID: <66d82c6a42dd2_210cd4119ab86555f@gitlab.mail> Matthew Pickering pushed new branch wip/typo-binds at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/typo-binds You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 10:10:41 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 04 Sep 2024 06:10:41 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 14 commits: Haddock: Add no-compilation flag Message-ID: <66d83221e6764_176a388b6b486893@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - aa4fd2b0 by Matthew Pickering at 2024-09-04T09:33:10+01:00 Run on test-abi label - - - - - d47bc1fd by Rodrigo Mesquita at 2024-09-04T11:00:19+01:00 Write a test for object determinism Extend abi_test with object determinism check Standalone run abi test Disable local test on CI Update abi test with decrementing uniques Twekas to script check - - - - - 89d83218 by Rodrigo Mesquita at 2024-09-04T11:10:19+01:00 UniqDSM det uniques + use in Cmm.Info Now for SRTs SRT generation using deterministic uniq supply Back LabelMap with deterministic UDFM TSAN uniq rename hard Revert "TSAN uniq rename hard" This reverts commit 7ca5ab3036c15f38c6d4cbcb616d415958c6bcda. improvements to uniqdsm UniqDSM ProcPoint CmmLayoutStack UniqDet 90% of cpsTop UniqDSM Major progress in using UniqDSM in CmmToAsm and Ncg backends Fix imports Un-back label map with udfm Revert "Un-back label map with udfm" This reverts commit f5d2e4257214a3f7b7d845651e6662c5babfd6a3. Make UDSM oneshot deriving via state Fix -fllvm hang - - - - - bab7730b by Rodrigo Mesquita at 2024-09-04T11:10:19+01:00 Remame uniques straight off stgtocmm, before cmm pipeline WIP Progress Work around LLVM assembler bug! In a really stupid way) Fix ordering of CLabels for IdLabels Local test script tweaks Do uniq renaming before SRTs Revert "Do uniq renaming before SRTs" This reverts commit db38b635d626106e40b3ab18091e0a24046c30c5. Do on CmmGroup Do uniq-renaming pass right at `codeGen` not better Revert "Do uniq-renaming pass right at `codeGen`" This reverts commit 74e9068aaaf736bf815a36bf74a0dde19a074a7a. Reapply "Do uniq renaming before SRTs" This reverts commit 682f89732fc2a95fa011f530c0c6922bf576d229. Try ALSO after SRT Revert "Try ALSO after SRT" This reverts commit c5dd7b426cde768126402aac3f39617ccb99f5c5. Renaming before and after SRTs bc of procs and srts and ... Wait no that was way too slow... cleaner approach, same idea Put deterministic renaming behind a flag Fix Ord CLabel only compare externalNames UniqRnem fixes external names DCmmDecl UniqRenam - - - - - 413f8737 by Rodrigo Mesquita at 2024-09-04T11:10:20+01:00 cmm: Back LabelMap with UDFM Use a deterministic unique map to back the implementation of `LabelMap`. This is necessary towards the goal of object code determinism in #12935. Our intended solution requires renaming uniques in a deterministic order (which will be the order in which they were created), but storing them label map makes us lose this order. Backing it with a UDFM fixes this issue. Introduce back LabelMap non deterministic Use NonDeterministic Label map in multiple passes (TODO: More could be available. Look through Det LabelMap uses again) Use NonDet for CFG More NonDet More explicit Introduce DCmmDecl, start Removing more maps Fix warnings tests Fix tests undo undo - - - - - 38132851 by Rodrigo Mesquita at 2024-09-04T11:10:20+01:00 Refactor ProfilingInfo to preserve Unique information before rendering it Rename Profiling Info now that names are preserved Revert "Rename Profiling Info now that names are preserved" This reverts commit 2dd3da96b7e771ae272791a00d7fb55313401c9e. Revert "Refactor ProfilingInfo to preserve Unique information before rendering it" This reverts commit 8aba0515bb744ca5add6a4c3c9c7760e226e0b31. Performance tweaks Get rid of UniqRenamable class, do it directly Make sure graph is renamed first, info table last Turns out it does matter! Whitespace - - - - - 0adbf0c2 by Rodrigo Mesquita at 2024-09-04T11:10:20+01:00 Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled - - - - - 7eb50739 by Rodrigo Mesquita at 2024-09-04T11:10:20+01:00 distinct-constructor-tables determinism - - - - - f5cb9ac9 by Rodrigo Mesquita at 2024-09-04T11:10:20+01:00 Rename deterministically CmmGroups in generateCgIPEStub - - - - - 8c244e4d by Rodrigo Mesquita at 2024-09-04T11:10:20+01:00 Writing Notes and Cleaning up v1 Fix linters MP entries - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Reg/Graph.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5f85ad9fafd3fa885ab888c6b4ea160794cdf0b9...8c244e4d5376adb99df1eb2008b01a7078193e29 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5f85ad9fafd3fa885ab888c6b4ea160794cdf0b9...8c244e4d5376adb99df1eb2008b01a7078193e29 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 10:45:46 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 04 Sep 2024 06:45:46 -0400 Subject: [Git][ghc/ghc][wip/romes/25110] base: Deprecate BCO primops exports from GHC.Exts Message-ID: <66d83a5a2bf39_176a383411601015aa@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/25110 at Glasgow Haskell Compiler / GHC Commits: 7907bf40 by Rodrigo Mesquita at 2024-09-04T11:45:34+01:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - 3 changed files: - libraries/base/src/GHC/Exts.hs - libraries/ghci/GHCi/CreateBCO.hs - libraries/ghci/ghci.cabal.in Changes: ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -23,6 +23,12 @@ module GHC.Exts -- ** Legacy interface for arrays of arrays module GHC.Internal.ArrayArray, -- * Primitive operations + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.BCO, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.mkApUpd0#, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.newBCO#, module GHC.Prim, module GHC.Prim.Ext, -- ** Running 'RealWorld' state thread @@ -112,7 +118,10 @@ module GHC.Exts import GHC.Internal.Exts import GHC.Internal.ArrayArray -import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# ) +import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# + -- Deprecated + , BCO, mkApUpd0#, newBCO# ) +import qualified GHC.Prim as Prim ( BCO, mkApUpd0#, newBCO# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/ghci/GHCi/CreateBCO.hs ===================================== @@ -6,6 +6,10 @@ {-# LANGUAGE UnboxedTuples #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE CPP #-} +{-# OPTIONS_GHC -Wno-warnings-deprecations #-} +-- TODO We want to import GHC.Internal.Base (BCO, mkApUpd0#, newBCO#) instead +-- of from GHC.Exts when we can require of the bootstrap compiler to have +-- ghc-internal. -- -- (c) The University of Glasgow 2002-2006 ===================================== libraries/ghci/ghci.cabal.in ===================================== @@ -86,9 +86,10 @@ library array == 0.5.*, base >= 4.8 && < 4.21, -- ghc-internal == @ProjectVersionForLib at .* - -- TODO: Use GHC.Internal.Desugar from ghc-internal instead of ignoring - -- the deprecation warning of GHC.Desugar when we require ghc-internal - -- of the bootstrap compiler + -- TODO: Use GHC.Internal.Desugar and GHC.Internal.Base from + -- ghc-internal instead of ignoring the deprecation warning in GHCi.TH + -- and GHCi.CreateBCO when we require ghc-internal of the bootstrap + -- compiler ghc-prim >= 0.5.0 && < 0.12, binary == 0.8.*, bytestring >= 0.10 && < 0.13, View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7907bf403fd1425bf4d3782c8afb6138b1c602d6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7907bf403fd1425bf4d3782c8afb6138b1c602d6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 10:51:36 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Wed, 04 Sep 2024 06:51:36 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 23 commits: Update directory submodule to latest master Message-ID: <66d83bb87a0db_176a38450f7410192a@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 250852e8 by sheaf at 2024-09-04T12:51:07+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - fd854ca8 by sheaf at 2024-09-04T12:51:07+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 4981eff1 by sheaf at 2024-09-04T12:51:23+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - f704d915 by sheaf at 2024-09-04T12:51:26+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - e2a39d94 by sheaf at 2024-09-04T12:51:26+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 7d068fd5 by sheaf at 2024-09-04T12:51:26+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - a0eb9e26 by sheaf at 2024-09-04T12:51:26+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 1d2363a8 by sheaf at 2024-09-04T12:51:27+02:00 Add test for C calls & SIMD vectors - - - - - 16ba9654 by sheaf at 2024-09-04T12:51:27+02:00 Add test for #25169 - - - - - e9267961 by sheaf at 2024-09-04T12:51:27+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 15529ecb by sheaf at 2024-09-04T12:51:27+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - c37faadf by sheaf at 2024-09-04T12:51:27+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/61691275d45fd07552cf46da611cdad9c1fd77c7...c37faadfd5b1e3f484a7b3b83c0e0632ff662b98 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/61691275d45fd07552cf46da611cdad9c1fd77c7...c37faadfd5b1e3f484a7b3b83c0e0632ff662b98 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 11:20:40 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 04 Sep 2024 07:20:40 -0400 Subject: [Git][ghc/ghc][master] rts: fix checkClosure error message Message-ID: <66d8428847666_176a386c7ce4118914@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - 1 changed file: - rts/sm/Sanity.c Changes: ===================================== rts/sm/Sanity.c ===================================== @@ -357,7 +357,8 @@ checkClosure( const StgClosure* p ) info = ACQUIRE_LOAD(&p->header.info); if (IS_FORWARDING_PTR(info)) { - barf("checkClosure: found EVACUATED closure %d", info->type); + ASSERT(LOOKS_LIKE_CLOSURE_PTR(info)); + barf("checkClosure: found EVACUATED closure %u", GET_INFO((StgClosure*)UN_FORWARDING_PTR(info))->type); } #if defined(PROFILING) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0d3bc2fa3a9a8c342ec34bb9d32e493655a4ec69 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0d3bc2fa3a9a8c342ec34bb9d32e493655a4ec69 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 11:21:09 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 04 Sep 2024 07:21:09 -0400 Subject: [Git][ghc/ghc][master] MO_AcquireFence: Less restrictive barrier Message-ID: <66d842a51eb99_176a386c93c81224b6@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - 3 changed files: - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs Changes: ===================================== compiler/GHC/CmmToAsm/AArch64/CodeGen.hs ===================================== @@ -1990,9 +1990,13 @@ genCCall target dest_regs arg_regs = do MO_SubIntC _w -> unsupported mop -- Memory Ordering - MO_AcquireFence -> return (unitOL DMBISH) - MO_ReleaseFence -> return (unitOL DMBISH) - MO_SeqCstFence -> return (unitOL DMBISH) + -- Set flags according to their C pendants (stdatomic.h): + -- atomic_thread_fence(memory_order_acquire); // -> dmb ishld + MO_AcquireFence -> return . unitOL $ DMBISH DmbLoad + -- atomic_thread_fence(memory_order_release); // -> dmb ish + MO_ReleaseFence -> return . unitOL $ DMBISH DmbLoadStore + -- atomic_thread_fence(memory_order_seq_cst); // -> dmb ish + MO_SeqCstFence -> return . unitOL $ DMBISH DmbLoadStore MO_Touch -> return nilOL -- Keep variables live (when using interior pointers) -- Prefetch MO_Prefetch_Data _n -> return nilOL -- Prefetch hint. ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -134,7 +134,7 @@ regUsageOfInstr platform instr = case instr of LDAR _ dst src -> usage (regOp src, regOp dst) -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> usage ([], []) + DMBISH _ -> usage ([], []) -- 9. Floating Point Instructions -------------------------------------------- FMOV dst src -> usage (regOp src, regOp dst) @@ -281,7 +281,7 @@ patchRegsOfInstr instr env = case instr of LDAR f o1 o2 -> LDAR f (patchOp o1) (patchOp o2) -- 8. Synchronization Instructions ----------------------------------------- - DMBISH -> DMBISH + DMBISH c -> DMBISH c -- 9. Floating Point Instructions ------------------------------------------ FMOV o1 o2 -> FMOV (patchOp o1) (patchOp o2) @@ -649,7 +649,7 @@ data Instr | BCOND Cond Target -- branch with condition. b. -- 8. Synchronization Instructions ----------------------------------------- - | DMBISH + | DMBISH DMBISHFlags -- 9. Floating Point Instructions -- move to/from general purpose <-> floating, or floating to floating | FMOV Operand Operand @@ -672,6 +672,9 @@ data Instr -- - fnmadd: d = - r1 * r2 - r3 | FMA FMASign Operand Operand Operand Operand +data DMBISHFlags = DmbLoad | DmbLoadStore + deriving (Eq, Show) + instrCon :: Instr -> String instrCon i = case i of ===================================== compiler/GHC/CmmToAsm/AArch64/Ppr.hs ===================================== @@ -527,7 +527,8 @@ pprInstr platform instr = case instr of LDAR _f o1 o2 -> op2 (text "\tldar") o1 o2 -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> line $ text "\tdmb ish" + DMBISH DmbLoadStore -> line $ text "\tdmb ish" + DMBISH DmbLoad -> line $ text "\tdmb ishld" -- 9. Floating Point Instructions -------------------------------------------- FMOV o1 o2 -> op2 (text "\tfmov") o1 o2 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fb0a4e5cc545d8d995b6631138f40495917e795a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fb0a4e5cc545d8d995b6631138f40495917e795a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 11:52:15 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 04 Sep 2024 07:52:15 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 7 commits: rts: fix checkClosure error message Message-ID: <66d849efcafb4_13bcc7df23c79d0@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - 4b52ea8b by Fendor at 2024-09-04T07:51:37-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - 4d308bf0 by Fendor at 2024-09-04T07:51:37-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 56ea0432 by Fendor at 2024-09-04T07:51:37-04:00 Enable perf profiling for compiler performance tests - - - - - e15cf8b8 by sheaf at 2024-09-04T07:51:46-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - b9c46af5 by Matthew Pickering at 2024-09-04T07:51:46-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 28 changed files: - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Utils/Env.hs - compiler/GHC/Utils/TmpFs.hs - rts/sm/Sanity.c - testsuite/driver/perf_notes.py - testsuite/driver/testglobals.py - testsuite/driver/testlib.py - testsuite/driver/testutil.py - testsuite/tests/ghc-e/should_fail/T9930fail.stderr - testsuite/tests/perf/compiler/all.T - + testsuite/tests/rename/should_fail/T25056.hs - + testsuite/tests/rename/should_fail/T25056.stderr - + testsuite/tests/rename/should_fail/T25056a.hs - + testsuite/tests/rename/should_fail/T25056b.hs - testsuite/tests/rename/should_fail/all.T - testsuite/tests/typecheck/should_fail/T23739b.hs - testsuite/tests/typecheck/should_fail/T23739b.stderr - + testsuite/tests/typecheck/should_fail/T23739c.hs - + testsuite/tests/typecheck/should_fail/T23739c.stderr - testsuite/tests/typecheck/should_fail/all.T Changes: ===================================== .gitlab/generate-ci/gen_ci.hs ===================================== @@ -155,6 +155,7 @@ data BuildConfig , noSplitSections :: Bool , validateNonmovingGc :: Bool , textWithSIMDUTF :: Bool + , testsuiteUsePerf :: Bool } -- Extra arguments to pass to ./configure due to the BuildConfig @@ -216,6 +217,7 @@ vanilla = BuildConfig , noSplitSections = False , validateNonmovingGc = False , textWithSIMDUTF = False + , testsuiteUsePerf = False } splitSectionsBroken :: BuildConfig -> BuildConfig @@ -268,6 +270,9 @@ tsan = vanilla { threadSanitiser = True } noTntc :: BuildConfig noTntc = vanilla { tablesNextToCode = False } +usePerfProfilingTestsuite :: BuildConfig -> BuildConfig +usePerfProfilingTestsuite bc = bc { testsuiteUsePerf = True } + ----------------------------------------------------------------------------- -- Platform specific variables ----------------------------------------------------------------------------- @@ -288,6 +293,9 @@ runnerTag _ _ = error "Invalid arch/opsys" tags :: Arch -> Opsys -> BuildConfig -> [String] tags arch opsys _bc = [runnerTag arch opsys] -- Tag for which runners we can use +runnerPerfTag :: Arch -> Opsys -> String +runnerPerfTag arch sys = runnerTag arch sys ++ "-perf" + -- These names are used to find the docker image so they have to match what is -- in the docker registry. distroName :: LinuxDistro -> String @@ -775,6 +783,7 @@ job arch opsys buildConfig = NamedJob { name = jobName, jobInfo = Job {..} } | validateNonmovingGc buildConfig ] in "RUNTEST_ARGS" =: unwords runtestArgs + , if testsuiteUsePerf buildConfig then "RUNTEST_ARGS" =: "--config perf_path=perf" else mempty ] jobArtifacts = Artifacts @@ -897,6 +906,12 @@ highCompression = addVariable "XZ_OPT" "-9" useHashUnitIds :: Job -> Job useHashUnitIds = addVariable "HADRIAN_ARGS" "--hash-unit-ids" +-- | Change the tag of the job to make sure the job is scheduled on a +-- runner that has the necessary capabilties to run the job with 'perf' +-- profiling counters. +perfProfilingJobTag :: Arch -> Opsys -> Job -> Job +perfProfilingJobTag arch opsys j = j { jobTags = [ runnerPerfTag arch opsys ] } + -- | Mark the validate job to run in fast-ci mode -- This is default way, to enable all jobs you have to apply the `full-ci` label. fastCI :: JobGroup Job -> JobGroup Job @@ -1000,6 +1015,8 @@ debian_x86 = , modifyNightlyJobs allowFailure (modifyValidateJobs (allowFailure . manual) tsan_jobs) , -- Nightly allowed to fail: #22343 modifyNightlyJobs allowFailure (modifyValidateJobs manual (validateBuilds Amd64 (Linux validate_debian) noTntc)) + -- Run the 'perf' profiling nightly job in the release config. + , perfProfilingJob Amd64 (Linux Debian12) releaseConfig , onlyRule LLVMBackend (validateBuilds Amd64 (Linux validate_debian) llvm) , addValidateRule TestPrimops (standardBuilds Amd64 (Linux validate_debian)) @@ -1010,6 +1027,12 @@ debian_x86 = where validate_debian = Debian12 + perfProfilingJob arch sys buildConfig = + -- Rename the job to avoid conflicts + rename (<> "-perf") + $ modifyJobs (perfProfilingJobTag arch sys) + $ disableValidate (validateBuilds arch sys $ usePerfProfilingTestsuite buildConfig) + tsan_jobs = modifyJobs ( addVariable "TSAN_OPTIONS" "suppressions=$CI_PROJECT_DIR/rts/.tsan-suppressions" ===================================== .gitlab/jobs.yaml ===================================== @@ -1791,6 +1791,69 @@ "XZ_OPT": "-9" } }, + "nightly-x86_64-linux-deb12-release-perf": { + "after_script": [ + ".gitlab/ci.sh save_cache", + ".gitlab/ci.sh save_test_output", + ".gitlab/ci.sh clean", + "cat ci_timings" + ], + "allow_failure": false, + "artifacts": { + "expire_in": "8 weeks", + "paths": [ + "ghc-x86_64-linux-deb12-release.tar.xz", + "junit.xml", + "unexpected-test-output.tar.gz" + ], + "reports": { + "junit": "junit.xml" + }, + "when": "always" + }, + "cache": { + "key": "x86_64-linux-deb12-$CACHE_REV", + "paths": [ + "cabal-cache", + "toolchain" + ] + }, + "dependencies": [], + "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "needs": [ + { + "artifacts": false, + "job": "hadrian-ghc-in-ghci" + } + ], + "rules": [ + { + "if": "(\"true\" == \"true\") && ($RELEASE_JOB != \"yes\") && ($NIGHTLY)", + "when": "on_success" + } + ], + "script": [ + "sudo chown ghc:ghc -R .", + ".gitlab/ci.sh setup", + ".gitlab/ci.sh configure", + ".gitlab/ci.sh build_hadrian", + ".gitlab/ci.sh test_hadrian" + ], + "stage": "full-build", + "tags": [ + "x86_64-linux-perf" + ], + "variables": { + "BIGNUM_BACKEND": "gmp", + "BIN_DIST_NAME": "ghc-x86_64-linux-deb12-release", + "BUILD_FLAVOUR": "release", + "CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", + "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", + "RUNTEST_ARGS": " --config perf_path=perf", + "TEST_ENV": "x86_64-linux-deb12-release", + "XZ_OPT": "-9" + } + }, "nightly-x86_64-linux-deb12-unreg-validate": { "after_script": [ ".gitlab/ci.sh save_cache", ===================================== compiler/GHC/CmmToAsm/AArch64/CodeGen.hs ===================================== @@ -1990,9 +1990,13 @@ genCCall target dest_regs arg_regs = do MO_SubIntC _w -> unsupported mop -- Memory Ordering - MO_AcquireFence -> return (unitOL DMBISH) - MO_ReleaseFence -> return (unitOL DMBISH) - MO_SeqCstFence -> return (unitOL DMBISH) + -- Set flags according to their C pendants (stdatomic.h): + -- atomic_thread_fence(memory_order_acquire); // -> dmb ishld + MO_AcquireFence -> return . unitOL $ DMBISH DmbLoad + -- atomic_thread_fence(memory_order_release); // -> dmb ish + MO_ReleaseFence -> return . unitOL $ DMBISH DmbLoadStore + -- atomic_thread_fence(memory_order_seq_cst); // -> dmb ish + MO_SeqCstFence -> return . unitOL $ DMBISH DmbLoadStore MO_Touch -> return nilOL -- Keep variables live (when using interior pointers) -- Prefetch MO_Prefetch_Data _n -> return nilOL -- Prefetch hint. ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -134,7 +134,7 @@ regUsageOfInstr platform instr = case instr of LDAR _ dst src -> usage (regOp src, regOp dst) -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> usage ([], []) + DMBISH _ -> usage ([], []) -- 9. Floating Point Instructions -------------------------------------------- FMOV dst src -> usage (regOp src, regOp dst) @@ -281,7 +281,7 @@ patchRegsOfInstr instr env = case instr of LDAR f o1 o2 -> LDAR f (patchOp o1) (patchOp o2) -- 8. Synchronization Instructions ----------------------------------------- - DMBISH -> DMBISH + DMBISH c -> DMBISH c -- 9. Floating Point Instructions ------------------------------------------ FMOV o1 o2 -> FMOV (patchOp o1) (patchOp o2) @@ -649,7 +649,7 @@ data Instr | BCOND Cond Target -- branch with condition. b. -- 8. Synchronization Instructions ----------------------------------------- - | DMBISH + | DMBISH DMBISHFlags -- 9. Floating Point Instructions -- move to/from general purpose <-> floating, or floating to floating | FMOV Operand Operand @@ -672,6 +672,9 @@ data Instr -- - fnmadd: d = - r1 * r2 - r3 | FMA FMASign Operand Operand Operand Operand +data DMBISHFlags = DmbLoad | DmbLoadStore + deriving (Eq, Show) + instrCon :: Instr -> String instrCon i = case i of ===================================== compiler/GHC/CmmToAsm/AArch64/Ppr.hs ===================================== @@ -527,7 +527,8 @@ pprInstr platform instr = case instr of LDAR _f o1 o2 -> op2 (text "\tldar") o1 o2 -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> line $ text "\tdmb ish" + DMBISH DmbLoadStore -> line $ text "\tdmb ish" + DMBISH DmbLoad -> line $ text "\tdmb ishld" -- 9. Floating Point Instructions -------------------------------------------- FMOV o1 o2 -> op2 (text "\tfmov") o1 o2 ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2924,19 +2924,22 @@ runParPipelines worker_limit plugin_hsc_env diag_wrapper mHscMessager all_pipeli atomically $ writeTVar stopped_var True wait_log_thread -withLocalTmpFS :: RunMakeM a -> RunMakeM a -withLocalTmpFS act = do +withLocalTmpFS :: TmpFs -> (TmpFs -> IO a) -> IO a +withLocalTmpFS tmpfs act = do let initialiser = do - MakeEnv{..} <- ask - lcl_tmpfs <- liftIO $ forkTmpFsFrom (hsc_tmpfs hsc_env) - return $ hsc_env { hsc_tmpfs = lcl_tmpfs } - finaliser lcl_env = do - gbl_env <- ask - liftIO $ mergeTmpFsInto (hsc_tmpfs lcl_env) (hsc_tmpfs (hsc_env gbl_env)) + liftIO $ forkTmpFsFrom tmpfs + finaliser tmpfs_local = do + liftIO $ mergeTmpFsInto tmpfs_local tmpfs -- Add remaining files which weren't cleaned up into local tmp fs for -- clean-up later. -- Clear the logQueue if this node had it's own log queue - MC.bracket initialiser finaliser $ \lcl_hsc_env -> local (\env -> env { hsc_env = lcl_hsc_env}) act + MC.bracket initialiser finaliser act + +withLocalTmpFSMake :: MakeEnv -> (MakeEnv -> IO a) -> IO a +withLocalTmpFSMake env k = + withLocalTmpFS (hsc_tmpfs (hsc_env env)) $ \lcl_tmpfs + -> k (env { hsc_env = (hsc_env env) { hsc_tmpfs = lcl_tmpfs }}) + -- | Run the given actions and then wait for them all to finish. runAllPipelines :: WorkerLimit -> MakeEnv -> [MakeAction] -> IO () @@ -2958,16 +2961,18 @@ runAllPipelines worker_limit env acts = do runLoop :: (((forall a. IO a -> IO a) -> IO ()) -> IO a) -> MakeEnv -> [MakeAction] -> IO [a] runLoop _ _env [] = return [] runLoop fork_thread env (MakeAction act res_var :acts) = do - new_thread <- + + -- withLocalTmpFs has to occur outside of fork to remain deterministic + new_thread <- withLocalTmpFSMake env $ \lcl_env -> fork_thread $ \unmask -> (do - mres <- (unmask $ run_pipeline (withLocalTmpFS act)) + mres <- (unmask $ run_pipeline lcl_env act) `MC.onException` (putMVar res_var Nothing) -- Defensive: If there's an unhandled exception then still signal the failure. putMVar res_var mres) threads <- runLoop fork_thread env acts return (new_thread : threads) where - run_pipeline :: RunMakeM a -> IO (Maybe a) - run_pipeline p = runMaybeT (runReaderT p env) + run_pipeline :: MakeEnv -> RunMakeM a -> IO (Maybe a) + run_pipeline env p = runMaybeT (runReaderT p env) data MakeAction = forall a . MakeAction !(RunMakeM a) !(MVar (Maybe a)) ===================================== compiler/GHC/Rename/Env.hs ===================================== @@ -442,6 +442,7 @@ lookupConstructorInfo con_name ; case info of IAmConLike con_info -> return con_info UnboundGRE -> return $ ConInfo (ConIsData []) ConHasPositionalArgs + IAmTyCon {} -> failIllegalTyCon WL_Constructor con_name _ -> pprPanic "lookupConstructorInfo: not a ConLike" $ vcat [ text "name:" <+> ppr con_name ] } @@ -1035,24 +1036,12 @@ lookupOccRn' which_suggest rdr_name lookupOccRn :: RdrName -> RnM Name lookupOccRn = lookupOccRn' WL_Anything --- lookupOccRnConstr looks up an occurrence of a RdrName and displays --- constructors and pattern synonyms as suggestions if it is not in scope +-- | Look up an occurrence of a 'RdrName'. -- --- There is a fallback to the type level, when the first lookup fails. --- This is required to implement a pat-to-type transformation --- (See Note [Pattern to type (P2T) conversion] in GHC.Tc.Gen.Pat) --- Consider this example: +-- Displays constructors and pattern synonyms as suggestions if +-- it is not in scope. -- --- data VisProxy a where VP :: forall a -> VisProxy a --- --- f :: VisProxy Int -> () --- f (VP Int) = () --- --- Here `Int` is actually a type, but it stays on position where --- we expect a data constructor. --- --- In all other cases we just use this additional lookup for better --- error messaging (See Note [Promotion]). +-- See Note [lookupOccRnConstr] lookupOccRnConstr :: RdrName -> RnM Name lookupOccRnConstr rdr_name = do { mb_gre <- lookupOccRn_maybe rdr_name @@ -1064,6 +1053,28 @@ lookupOccRnConstr rdr_name Just gre -> return $ greName gre Nothing -> reportUnboundName' WL_Constructor rdr_name} } +{- Note [lookupOccRnConstr] +~~~~~~~~~~~~~~~~~~~~~~~~~~~ +lookupOccRnConstr looks up a data constructor or pattern synonym. Simple. + +However, there is a fallback to the type level when the lookup fails. +This is required to implement a pat-to-type transformation +(See Note [Pattern to type (P2T) conversion] in GHC.Tc.Gen.Pat) + +Consider this example: + + data VisProxy a where VP :: forall a -> VisProxy a + + f :: VisProxy Int -> () + f (VP Int) = () + +Here `Int` is actually a type, but it occurs in a position in which we expect +a data constructor. + +In all other cases we just use this additional lookup for better +error messaging (See Note [Promotion]). +-} + -- lookupOccRnRecField looks up an occurrence of a RdrName and displays -- record fields as suggestions if it is not in scope lookupOccRnRecField :: RdrName -> RnM Name ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -539,9 +539,9 @@ rnExpr (ExplicitSum _ alt arity expr) = do { (expr', fvs) <- rnLExpr expr ; return (ExplicitSum noExtField alt arity expr', fvs) } -rnExpr (RecordCon { rcon_con = con_id +rnExpr (RecordCon { rcon_con = con_rdr , rcon_flds = rec_binds@(HsRecFields { rec_dotdot = dd }) }) - = do { con_lname@(L _ con_name) <- lookupLocatedOccRnConstr con_id + = do { con_lname@(L _ con_name) <- lookupLocatedOccRnConstr con_rdr ; (flds, fvs) <- rnHsRecFields (HsRecFieldCon con_name) mk_hs_var rec_binds ; (flds', fvss) <- mapAndUnzipM rn_field flds ; let rec_binds' = HsRecFields { rec_ext = noExtField, rec_flds = flds', rec_dotdot = dd } ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -841,7 +841,7 @@ tc_infer_id id_name AGlobal (AConLike (RealDataCon con)) -> tcInferDataCon con AGlobal (AConLike (PatSynCon ps)) -> tcInferPatSyn id_name ps - (tcTyThingTyCon_maybe -> Just tc) -> failIllegalTyCon WL_Anything tc -- TyCon or TcTyCon + (tcTyThingTyCon_maybe -> Just tc) -> failIllegalTyCon WL_Anything (tyConName tc) ATyVar name _ -> failIllegalTyVal name _ -> failWithTc $ TcRnExpectedValueId thing } ===================================== compiler/GHC/Tc/Utils/Env.hs ===================================== @@ -280,7 +280,7 @@ tcLookupConLike name = do thing <- tcLookupGlobal name case thing of AConLike cl -> return cl - ATyCon tc -> failIllegalTyCon WL_Constructor tc + ATyCon {} -> failIllegalTyCon WL_Constructor name _ -> wrongThingErr WrongThingConLike (AGlobal thing) name tcLookupRecSelParent :: HsRecUpdParent GhcRn -> TcM RecSelParent @@ -353,19 +353,20 @@ instance MonadThings (IOEnv (Env TcGblEnv TcLclEnv)) where lookupThing = tcLookupGlobal -- Illegal term-level use of type things -failIllegalTyCon :: WhatLooking -> TyCon -> TcM a +failIllegalTyCon :: WhatLooking -> Name -> TcM a failIllegalTyVal :: Name -> TcM a (failIllegalTyCon, failIllegalTyVal) = (fail_tycon, fail_tyvar) where - fail_tycon what_looking tc = do + fail_tycon what_looking tc_nm = do gre <- getGlobalRdrEnv - let nm = tyConName tc - pprov = case lookupGRE_Name gre nm of + let mb_gre = lookupGRE_Name gre tc_nm + pprov = case mb_gre of Just gre -> nest 2 (pprNameProvenance gre) Nothing -> empty - err | isClassTyCon tc = ClassTE - | otherwise = TyConTE - fail_with_msg what_looking dataName nm pprov err + err = case greInfo <$> mb_gre of + Just (IAmTyCon ClassFlavour) -> ClassTE + _ -> TyConTE + fail_with_msg what_looking dataName tc_nm pprov err fail_tyvar nm = let pprov = nest 2 (text "bound at" <+> ppr (getSrcLoc nm)) ===================================== compiler/GHC/Utils/TmpFs.hs ===================================== @@ -64,6 +64,8 @@ data TmpFs = TmpFs -- -- Shared with forked TmpFs. + , tmp_dir_prefix :: String + , tmp_files_to_clean :: IORef PathsToClean -- ^ Files to clean (per session or per module) -- @@ -121,6 +123,7 @@ initTmpFs = do , tmp_subdirs_to_clean = subdirs , tmp_dirs_to_clean = dirs , tmp_next_suffix = next + , tmp_dir_prefix = "tmp" } -- | Initialise an empty TmpFs sharing unique numbers and per-process temporary @@ -132,11 +135,16 @@ forkTmpFsFrom :: TmpFs -> IO TmpFs forkTmpFsFrom old = do files <- newIORef emptyPathsToClean subdirs <- newIORef emptyPathsToClean + counter <- newIORef 0 + prefix <- newTempSuffix old + + return $ TmpFs { tmp_files_to_clean = files , tmp_subdirs_to_clean = subdirs , tmp_dirs_to_clean = tmp_dirs_to_clean old - , tmp_next_suffix = tmp_next_suffix old + , tmp_next_suffix = counter + , tmp_dir_prefix = prefix } -- | Merge the first TmpFs into the second. @@ -259,9 +267,11 @@ changeTempFilesLifetime tmpfs lifetime files = do addFilesToClean tmpfs lifetime existing_files -- Return a unique numeric temp file suffix -newTempSuffix :: TmpFs -> IO Int -newTempSuffix tmpfs = - atomicModifyIORef' (tmp_next_suffix tmpfs) $ \n -> (n+1,n) +newTempSuffix :: TmpFs -> IO String +newTempSuffix tmpfs = do + n <- atomicModifyIORef' (tmp_next_suffix tmpfs) $ \n -> (n+1,n) + return $ tmp_dir_prefix tmpfs ++ "_" ++ show n + -- Find a temporary name that doesn't already exist. newTempName :: Logger -> TmpFs -> TempDir -> TempFileLifetime -> Suffix -> IO FilePath @@ -271,8 +281,8 @@ newTempName logger tmpfs tmp_dir lifetime extn where findTempName :: FilePath -> IO FilePath findTempName prefix - = do n <- newTempSuffix tmpfs - let filename = prefix ++ show n <.> extn + = do suffix <- newTempSuffix tmpfs + let filename = prefix ++ suffix <.> extn b <- doesFileExist filename if b then findTempName prefix else do -- clean it up later @@ -295,8 +305,8 @@ newTempSubDir logger tmpfs tmp_dir where findTempDir :: FilePath -> IO FilePath findTempDir prefix - = do n <- newTempSuffix tmpfs - let name = prefix ++ show n + = do suffix <- newTempSuffix tmpfs + let name = prefix ++ suffix b <- doesDirectoryExist name if b then findTempDir prefix else (do @@ -314,8 +324,8 @@ newTempLibName logger tmpfs tmp_dir lifetime extn where findTempName :: FilePath -> String -> IO (FilePath, FilePath, String) findTempName dir prefix - = do n <- newTempSuffix tmpfs -- See Note [Deterministic base name] - let libname = prefix ++ show n + = do suffix <- newTempSuffix tmpfs -- See Note [Deterministic base name] + let libname = prefix ++ suffix filename = dir "lib" ++ libname <.> extn b <- doesFileExist filename if b then findTempName dir prefix @@ -340,8 +350,8 @@ getTempDir logger tmpfs (TempDir tmp_dir) = do mkTempDir :: FilePath -> IO FilePath mkTempDir prefix = do - n <- newTempSuffix tmpfs - let our_dir = prefix ++ show n + suffix <- newTempSuffix tmpfs + let our_dir = prefix ++ suffix -- 1. Speculatively create our new directory. createDirectory our_dir @@ -376,6 +386,11 @@ the temporary file no longer contains random information (it used to contain the process id). This is ok, as the temporary directory used contains the pid (see getTempDir). + +In addition to this, multiple threads can race against each other creating temporary +files. Therefore we supply a prefix when creating temporary files, when a thread is +forked, each thread must be given an TmpFs with a unique prefix. This is achieved +by forkTmpFsFrom creating a fresh prefix from the parent TmpFs. -} manyWithTrace :: Logger -> String -> ([FilePath] -> IO ()) -> [FilePath] -> IO () ===================================== rts/sm/Sanity.c ===================================== @@ -357,7 +357,8 @@ checkClosure( const StgClosure* p ) info = ACQUIRE_LOAD(&p->header.info); if (IS_FORWARDING_PTR(info)) { - barf("checkClosure: found EVACUATED closure %d", info->type); + ASSERT(LOOKS_LIKE_CLOSURE_PTR(info)); + barf("checkClosure: found EVACUATED closure %u", GET_INFO((StgClosure*)UN_FORWARDING_PTR(info))->type); } #if defined(PROFILING) ===================================== testsuite/driver/perf_notes.py ===================================== @@ -128,6 +128,41 @@ AllowedPerfChange = NamedTuple('AllowedPerfChange', ('opts', Dict[str, str]) ]) +class MetricAcceptanceWindow: + """ + A strategy for computing an acceptance window for a metric measurement + given a baseline value. + """ + def get_bounds(self, baseline: float) -> Tuple[float, float]: + raise NotImplemented + def describe(self) -> str: + raise NotImplemented + +class AlwaysAccept(MetricAcceptanceWindow): + def get_bounds(self, baseline: float) -> Tuple[float, float]: + return (-1/0, +1/0) + + def describe(self) -> str: + raise NotImplemented + +class RelativeMetricAcceptanceWindow(MetricAcceptanceWindow): + """ + A MetricAcceptanceWindow which accepts measurements within tol-percent of + the baseline. + """ + def __init__(self, tol: float): + """ Accept any metric within tol-percent of the baseline """ + self.__tol = tol + + def get_bounds(self, baseline: float) -> Tuple[float, float]: + lowerBound = trunc( int(baseline) * ((100 - float(self.__tol))/100)) + upperBound = trunc(0.5 + ceil(int(baseline) * ((100 + float(self.__tol))/100))) + + return (lowerBound, upperBound) + + def describe(self) -> str: + return '+/- %1.1f%%' % (100*self.__tol) + def parse_perf_stat(stat_str: str) -> PerfStat: field_vals = stat_str.strip('\t').split('\t') stat = PerfStat(*field_vals) # type: ignore @@ -558,26 +593,32 @@ def get_commit_metric(gitNoteRef, _commit_metric_cache[cacheKeyA] = baseline_by_cache_key_b return baseline_by_cache_key_b.get(cacheKeyB) -# Check test stats. This prints the results for the user. -# actual: the PerfStat with actual value. -# baseline: the expected Baseline value (this should generally be derived from baseline_metric()) -# tolerance_dev: allowed deviation of the actual value from the expected value. -# allowed_perf_changes: allowed changes in stats. This is a dictionary as returned by get_allowed_perf_changes(). -# force_print: Print stats even if the test stat was in the tolerance range. -# Returns a (MetricChange, pass/fail object) tuple. Passes if the stats are within the expected value ranges. def check_stats_change(actual: PerfStat, baseline: Baseline, - tolerance_dev, + acceptance_window: MetricAcceptanceWindow, allowed_perf_changes: Dict[TestName, List[AllowedPerfChange]] = {}, force_print = False ) -> Tuple[MetricChange, Any]: + """ + Check test stats. This prints the results for the user. + + Parameters: + actual: the PerfStat with actual value + baseline: the expected Baseline value (this should generally be derived + from baseline_metric()) + acceptance_window: allowed deviation of the actual value from the expected + value. + allowed_perf_changes: allowed changes in stats. This is a dictionary as + returned by get_allowed_perf_changes(). + force_print: Print stats even if the test stat was in the tolerance range. + + Returns a (MetricChange, pass/fail object) tuple. Passes if the stats are within the expected value ranges. + """ expected_val = baseline.perfStat.value full_name = actual.test + ' (' + actual.way + ')' - lowerBound = trunc( int(expected_val) * ((100 - float(tolerance_dev))/100)) - upperBound = trunc(0.5 + ceil(int(expected_val) * ((100 + float(tolerance_dev))/100))) - actual_dev = round(((float(actual.value) * 100)/ int(expected_val)) - 100, 1) + lowerBound, upperBound = acceptance_window.get_bounds(expected_val) # Find the direction of change. change = MetricChange.NoChange @@ -613,11 +654,12 @@ def check_stats_change(actual: PerfStat, def display(descr, val, extra): print(descr, str(val).rjust(length), extra) - display(' Expected ' + full_name + ' ' + actual.metric + ':', expected_val, '+/-' + str(tolerance_dev) + '%') + display(' Expected ' + full_name + ' ' + actual.metric + ':', expected_val, acceptance_window.describe()) display(' Lower bound ' + full_name + ' ' + actual.metric + ':', lowerBound, '') display(' Upper bound ' + full_name + ' ' + actual.metric + ':', upperBound, '') display(' Actual ' + full_name + ' ' + actual.metric + ':', actual.value, '') if actual.value != expected_val: + actual_dev = round(((float(actual.value) * 100)/ int(expected_val)) - 100, 1) display(' Deviation ' + full_name + ' ' + actual.metric + ':', actual_dev, '%') return (change, result) ===================================== testsuite/driver/testglobals.py ===================================== @@ -49,6 +49,9 @@ class TestConfig: # Path to Ghostscript self.gs = None # type: Optional[Path] + # Path to Linux `perf` tool + self.perf_path = None # type: Optional[Path] + # Run tests requiring Haddock self.haddock = False @@ -472,6 +475,9 @@ class TestOptions: # The extra hadrian dependencies we need for this particular test self.hadrian_deps = set(["test:ghc"]) # type: Set[str] + # Record these `perf-events` counters when compiling this test, if `perf` is available + self.compiler_perf_counters = [] # type: List[str] + @property def testdir(self) -> Path: if self.testdir_raw is None: ===================================== testsuite/driver/testlib.py ===================================== @@ -3,6 +3,7 @@ # (c) Simon Marlow 2002 # +import csv import io import shutil import os @@ -23,12 +24,13 @@ from testglobals import config, ghc_env, default_testopts, brokens, t, \ TestRun, TestResult, TestOptions, PerfMetric from testutil import strip_quotes, lndir, link_or_copy_file, passed, \ failBecause, testing_metrics, residency_testing_metrics, \ + stable_perf_counters, \ PassFail, badResult, memoize from term_color import Color, colored import testutil from cpu_features import have_cpu_feature import perf_notes as Perf -from perf_notes import MetricChange, PerfStat, StatsException +from perf_notes import MetricChange, PerfStat, StatsException, AlwaysAccept, RelativeMetricAcceptanceWindow extra_src_files = {'T4198': ['exitminus1.c']} # TODO: See #12223 from my_typing import * @@ -752,9 +754,14 @@ def find_so(lib): def find_non_inplace_so(lib): return _find_so(lib,path_from_ghcPkg(lib, "dynamic-library-dirs"),False) -# Define a generic stat test, which computes the statistic by calling the function -# given as the third argument. -def collect_generic_stat ( metric, deviation, get_stat ): + +def collect_generic_stat ( metric, deviation: Optional[int], get_stat: Callable[[WayName], str]): + """ + Define a generic stat test, which computes the statistic by calling the function + given as the third argument. + + If no deviation is given, the test cannot fail, but the metric will be recorded nevertheless. + """ return collect_generic_stats ( { metric: { 'deviation': deviation, 'current': get_stat } } ) def _collect_generic_stat(name : TestName, opts, metric_infos): @@ -801,16 +808,27 @@ def collect_stats(metric='all', deviation=20, static_stats_file=None): def statsFile(comp_test: bool, name: str) -> str: return name + ('.comp' if comp_test else '') + '.stats' +def perfStatsFile(comp_test: bool, name: str) -> str: + return name + ('.comp' if comp_test else '') + '.perf.csv' + # This is an internal function that is used only in the implementation. # 'is_compiler_stats_test' is somewhat of an unfortunate name. # If the boolean is set to true, it indicates that this test is one that # measures the performance numbers of the compiler. # As this is a fairly rare case in the testsuite, it defaults to false to # indicate that it is a 'normal' performance test. -def _collect_stats(name: TestName, opts, metrics, deviation, static_stats_file, is_compiler_stats_test=False): +def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], + static_stats_file: Optional[Union[Path,str]], + is_compiler_stats_test: bool = False, is_compiler_perf_test: bool = False) -> None: if not re.match('^[0-9]*[a-zA-Z][a-zA-Z0-9._-]*$', name): failBecause('This test has an invalid name.') + if is_compiler_perf_test and config.perf_path is None: + # If we are doing a 'perf' run but no 'perf' is configured, + # don't try to read the results. + # This is a bit weird, though. + return + # Normalize metrics to a list of strings. if isinstance(metrics, str): if metrics == 'all': @@ -865,11 +883,47 @@ def _collect_stats(name: TestName, opts, metrics, deviation, static_stats_file, assert val is not None return int(val) + # How to read the result of the performance test + def read_perf_stats_file(way, metric_name): + FIELDS = ['value','unit','event','runtime','percent'] + # Confusingly compile time ghci tests are actually runtime tests, so we have + # to go and look for the name.stats file rather than name.comp.stats file. + compiler_stats_test = is_compiler_stats_test and not (way == "ghci" or way == "ghci-opt") + + perf_stats_file = Path(in_testdir(perfStatsFile(compiler_stats_test, name))) + perf_metrics = {} + try: + perf_csv_lines = perf_stats_file.read_text().splitlines() + # Output looks like: + # """ + # # Started on + # + # ,,,,,... + # """ + # + # Ignore empty lines and lines starting with '#' + perf_csv = [l for l in perf_csv_lines if l and not l.startswith('#')] + + perf_stats_csv_reader = csv.DictReader(perf_csv, fieldnames=FIELDS, delimiter=";", quotechar="\"") + for fields in perf_stats_csv_reader: + perf_metrics[fields['event']] = float(fields['value']) + + except IOError as e: + raise StatsException(str(e)) + + val = perf_metrics[metric_name] + if val is None: + print('Failed to find metric: ', metric_name) + raise StatsException("No such metric") + else: + assert val is not None + return int(val) collect_stat = {} for metric_name in metrics: def action_generator(mn): - return lambda way: read_stats_file(way, mn) + read_stats = read_perf_stats_file if is_compiler_perf_test else read_stats_file + return lambda way: read_stats(way, mn) metric = '{}/{}'.format(tag, metric_name) collect_stat[metric] = { "deviation": deviation , "current": action_generator(metric_name) } @@ -1009,7 +1063,32 @@ def have_thread_sanitizer( ) -> bool: def gcc_as_cmmp() -> bool: return config.cmm_cpp_is_gcc -# --- +# ----- + +def collect_compiler_perf(deviation: Optional[int] = None): + """ + Record stable performance counters using `perf stat` when available. + """ + return [ + _collect_compiler_perf_counters(stable_perf_counters(), deviation) + ] + +def collect_compiler_perf_counters(counters: List[str], deviation: Optional[int] = None): + """ + Record the given event counters using `perf stat` when available. + """ + return [ + _collect_compiler_perf_counters(set(counters), deviation) + ] + +def _collect_compiler_perf_counters(counters: Set[str], deviation: Optional[int] = None): + def f(name, opts): + # Slightly hacky, we need the requested perf_counters in 'simple_run'. + # Thus, we have to globally register these counters + opts.compiler_perf_counters += list(counters) + _collect_stats(name, opts, counters, deviation, False, True, True) + return f + # Note [Measuring residency] # ~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -1062,6 +1141,12 @@ def collect_compiler_residency(tolerance_pct: float): collect_compiler_stats(residency_testing_metrics(), tolerance_pct) ] +def collect_compiler_runtime(tolerance_pct: float): + return [ + collect_compiler_stats('bytes allocated', tolerance_pct), + _collect_compiler_perf_counters(stable_perf_counters()) + ] + # --- def high_memory_usage(name, opts): @@ -1619,7 +1704,7 @@ async def do_test(name: TestName, stdout = stdout_path, stderr = stderr_path, print_output = config.verbose >= 3, - timeout_multiplier = opts.pre_cmd_timeout_multiplier, + timeout_multiplier = opts.pre_cmd_timeout_multiplier ) # If user used expect_broken then don't record failures of pre_cmd @@ -1809,6 +1894,8 @@ async def do_compile(name: TestName, result = await extras_build( way, extra_mods, extra_hc_opts ) if badResult(result): return result + + assert result.hc_opts is not None extra_hc_opts = result.hc_opts result = await simple_build(name, way, extra_hc_opts, should_fail, top_mod, units, should_link, True, **kwargs) @@ -1934,6 +2021,7 @@ async def compile_and_run__(name: TestName, result = await extras_build( way, extra_mods, extra_hc_opts ) if badResult(result): return result + assert result.hc_opts is not None extra_hc_opts = result.hc_opts assert extra_hc_opts is not None @@ -2027,10 +2115,15 @@ def report_stats(name, way, metric, gen_stat): metric_result = passed() perf_change = MetricChange.NewMetric else: + deviation = gen_stat["deviation"] + if deviation: + tolerance_metric = RelativeMetricAcceptanceWindow(deviation) + else: + tolerance_metric = AlwaysAccept() (perf_change, metric_result) = Perf.check_stats_change( perf_stat, baseline, - gen_stat["deviation"], + tolerance_metric, config.allowed_perf_changes, config.verbose >= 4) @@ -2051,9 +2144,14 @@ def report_stats(name, way, metric, gen_stat): # ----------------------------------------------------------------------------- # Build a single-module program -async def extras_build( way, extra_mods, extra_hc_opts ): +async def extras_build(way: WayName, extra_mods, extra_hc_opts) -> PassFail: for mod, opts in extra_mods: - result = await simple_build(mod, way, opts + ' ' + extra_hc_opts, False, None, [], False, False) + result = await simple_build(mod, way, opts + ' ' + extra_hc_opts, + should_fail=False, + top_mod=None, + units=[], + link=False, + addsuf=False) if not (mod.endswith('.hs') or mod.endswith('.lhs')): extra_hc_opts += ' %s' % Path(mod).with_suffix('.o') if badResult(result): @@ -2135,14 +2233,22 @@ async def simple_build(name: Union[TestName, str], flags = ' '.join(get_compiler_flags() + config.way_flags[way]) - cmd = ('cd "{opts.testdir}" && {cmd_prefix} ' + cmd = ('{cmd_prefix} ' '{{compiler}} {to_do} {srcname} {flags} {extra_hc_opts}' ).format(**locals()) if filter_with != '': cmd = cmd + ' | ' + filter_with - exit_code = await runCmd(cmd, None, stdout, stderr, opts.compile_timeout_multiplier) + output_file = perfStatsFile(True, name) + + exit_code = await runCmdPerf( + opts.compiler_perf_counters, + cmd, + output_file, + working_dir=opts.testdir, + stdin=None, stdout=stdout, stderr=stderr, + timeout_multiplier=opts.compile_timeout_multiplier) actual_stderr_path = in_testdir(name, 'comp.stderr') @@ -2162,7 +2268,6 @@ async def simple_build(name: Union[TestName, str], stderr_contents = actual_stderr_path.read_text(encoding='UTF-8', errors='replace') return failBecause('exit code non-0', stderr=stderr_contents) - return passed() # ----------------------------------------------------------------------------- @@ -2214,10 +2319,13 @@ async def simple_run(name: TestName, way: WayName, prog: str, extra_run_opts: st if opts.cmd_wrapper is not None: cmd = opts.cmd_wrapper(cmd) - cmd = 'cd "{opts.testdir}" && {cmd}'.format(**locals()) + output_file = perfStatsFile(False, name) # run the command - exit_code = await runCmd(cmd, stdin_arg, stdout_arg, stderr_arg, opts.run_timeout_multiplier) + exit_code = await runCmdPerf(opts.compiler_perf_counters, cmd, output_file, + working_dir=opts.testdir, + stdin=stdin_arg, stdout=stdout_arg, stderr=stderr_arg, + timeout_multiplier=opts.run_timeout_multiplier) # check the exit code if exit_code != opts.exit_code: @@ -2315,7 +2423,7 @@ async def interpreter_run(name: TestName, cmd = 'cd "{opts.testdir}" && {cmd}'.format(**locals()) - exit_code = await runCmd(cmd, script, stdout, stderr, opts.run_timeout_multiplier) + exit_code = await runCmd(cmd, script, stdout, stderr, timeout_multiplier=opts.run_timeout_multiplier) # split the stdout into compilation/program output split_file(stdout, delimiter, @@ -2973,12 +3081,56 @@ def dump_file(f: Path): except Exception: print('') +# ----------------------------------------------------------------------------- +# Run a program in the interpreter and check its output + +async def runCmdPerf( + perf_counters: List[str], + cmd: str, + output_file: str, + working_dir: Optional[Path] = None, + **kwargs) -> int: + """ + Run a command under `perf stat`, collecting the given counters. + + Returns the exit code and a dictionary of the collected counter values. + + If given a 'working_dir', we generate a command looking like: + + .. code-block:: text + + cd $working_dir && perf stat ... $cmd + + This allows users to find the test directory by looking at the execution logs, + and allows us to write 'perf' output files in the test directory. + """ + if len(perf_counters) == 0 or config.perf_path is None: + if working_dir: + cmd = f"cd \"{working_dir}\" && {cmd}" + + exit_code = await runCmd(cmd, **kwargs) + return exit_code + + perf_cmd_args: List[str] = [str(config.perf_path), 'stat', '-x\\;', '-o', output_file, '-e', ','.join(perf_counters), cmd] + cmd = ' '.join(perf_cmd_args) + if working_dir: + cmd = f"cd \"{working_dir}\" && {cmd}" + + exit_code = await runCmd(cmd, **kwargs) + return exit_code + async def runCmd(cmd: str, stdin: Union[None, Path]=None, stdout: Union[None, Path]=None, stderr: Union[None, int, Path]=None, timeout_multiplier=1.0, print_output=False) -> int: + """ + Run a command enforcing a timeout and returning the exit code. + + The process's working directory is changed to 'working_dir'. + """ + timeout_prog = strip_quotes(config.timeout_prog) timeout = str(int(ceil(config.timeout * timeout_multiplier))) @@ -3000,7 +3152,12 @@ async def runCmd(cmd: str, # Hence it must ultimately be run by a Bourne shell. It's timeout's job # to invoke the Bourne shell - proc = await asyncio.create_subprocess_exec(timeout_prog, timeout, cmd, stdin=stdin_file, stdout=asyncio.subprocess.PIPE, stderr=hStdErr, env=ghc_env) + proc = await asyncio.create_subprocess_exec(timeout_prog, timeout, cmd, + stdin=stdin_file, + stdout=asyncio.subprocess.PIPE, + stderr=hStdErr, + env=ghc_env + ) stdout_buffer, stderr_buffer = await proc.communicate() finally: ===================================== testsuite/driver/testutil.py ===================================== @@ -77,6 +77,11 @@ def lndir(srcdir: Path, dstdir: Path, force_copy=False): def testing_metrics(): return { 'bytes allocated', 'peak_megabytes_allocated', 'max_bytes_used' } +# All performance counters we consider to be stable enough in CI to +# test for. +def stable_perf_counters(): + return { 'instructions:u' } + # Metrics which are testing residency information def residency_testing_metrics(): return { 'peak_megabytes_allocated', 'max_bytes_used' } ===================================== testsuite/tests/ghc-e/should_fail/T9930fail.stderr ===================================== @@ -1,18 +1,12 @@ -ghc: Exception: +ghc-9.11.20240830: Exception: default output name would overwrite the input file; must specify -o explicitly Usage: For basic information, try the `--help' option. -Package: ghc-inplace +Package: ghc-9.11-inplace Module: GHC.Utils.Panic Type: GhcException HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO - throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:860:84 in exceptions-0.10.7-inplace:Control.Monad.Catch - onException, called at compiler/GHC/Driver/Make.hs:2974:23 in ghc-9.9-inplace:GHC.Driver.Make - - + bracket, called at compiler/GHC/Driver/Make.hs:2936:3 in ghc-9.11-inplace:GHC.Driver.Make ===================================== testsuite/tests/perf/compiler/all.T ===================================== @@ -8,7 +8,7 @@ test('T1969', extra_run_opts('+RTS -A64k -RTS'), # The default RESIDENCY_OPTS is 256k and we need higher sampling # frequency. Incurs a slow-down by about 2. - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), only_ways(['normal']), extra_hc_opts('-dcore-lint -static'), @@ -32,14 +32,14 @@ else: test('T3294', [collect_compiler_residency(15), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), conf_3294, ], compile, ['']) test('T4801', - [collect_compiler_stats('bytes allocated',2), + [collect_compiler_runtime(2), only_ways(['normal']), extra_hc_opts('-static'), when(arch('wasm32') and unregisterised(), fragile(23290)) @@ -49,7 +49,7 @@ test('T4801', test('T3064', [collect_compiler_residency(20), - collect_compiler_stats('bytes allocated',2), + collect_compiler_runtime(2), only_ways(['normal']), ], compile, @@ -59,7 +59,7 @@ test('T3064', test('T4007', normal, makefile_test, ['T4007']) test('T5030', - [collect_compiler_stats('bytes allocated', 2), + [collect_compiler_runtime(2), only_ways(['normal']) ], @@ -67,14 +67,14 @@ test('T5030', ['-freduction-depth=300']) test('T5631', - [collect_compiler_stats('bytes allocated',2), + [collect_compiler_runtime(2), only_ways(['normal']) ], compile, ['']) test('parsing001', - [collect_compiler_stats('bytes allocated',2), + [collect_compiler_runtime(2), only_ways(['normal']), ], compile_fail, ['']) @@ -82,27 +82,27 @@ test('parsing001', test('T783', [ only_ways(['normal']), # no optimisation for this one - collect_compiler_stats('bytes allocated',2), + collect_compiler_runtime(2), extra_hc_opts('-static') ], compile,['']) test('T5321Fun', [ only_ways(['normal']), # no optimisation for this one - collect_compiler_stats('bytes allocated',2) + collect_compiler_runtime(2), ], compile,['']) test('T5321FD', [ only_ways(['normal']), # no optimisation for this one - collect_compiler_stats('bytes allocated',2) + collect_compiler_runtime(2), ], compile,['']) test('T5642', [ only_ways(['normal']), normal, - collect_compiler_stats('bytes allocated',2) + collect_compiler_runtime(2), ], compile,['-O']) @@ -114,7 +114,7 @@ test('T5837', test('T6048', [ only_ways(['optasm']), - collect_compiler_stats('bytes allocated',2) + collect_compiler_runtime(2), ], compile,['']) @@ -134,7 +134,7 @@ test('T9675', test('T9872a', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), high_memory_usage ], compile_fail, @@ -142,28 +142,28 @@ test('T9872a', test('T9872b', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), high_memory_usage ], compile_fail, ['']) test('T9872b_defer', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), high_memory_usage ], compile, ['-fdefer-type-errors']) test('T9872c', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), high_memory_usage ], compile_fail, ['']) test('T9872d', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], compile, ['']) @@ -227,14 +227,14 @@ test ('LargeRecord', test('T9961', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], compile, ['-O']) test('T9233', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], multimod_compile, ['T9233', '-v0 -O2 -fno-spec-constr']) @@ -249,14 +249,14 @@ test('T10370', test('T11068', normal, makefile_test, ['T11068']) test('T10547', - [ collect_compiler_stats('bytes allocated', 4), + [ collect_compiler_runtime(4), ], compile_fail, ['-fprint-expanded-synonyms']) test('T12227', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], compile, # Use `-M1G` to prevent memory thrashing with ghc-8.0.1. @@ -264,14 +264,14 @@ test('T12227', test('T12425', [ only_ways(['optasm']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], compile, ['']) test('T12234', [ only_ways(['optasm']), - collect_compiler_stats('bytes allocated', 2), + collect_compiler_runtime(2), ], compile, ['']) @@ -279,14 +279,14 @@ test('T12234', # See Note [Sensitivity to unique increment] in T12545.hs; spread was 4.8% test('T12545', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 10), # + collect_compiler_runtime(10), # ], multimod_compile, ['T12545', '-v0'] ) test('T13035', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), ], compile, [''] ) @@ -299,7 +299,7 @@ test('T13056', ['-O1']) test('T12707', - [ collect_compiler_stats('bytes allocated', 1), + [ collect_compiler_runtime(1), ], compile, ['']) @@ -311,7 +311,7 @@ test('T12707', # to avoid spurious errors. test('T12150', [ only_ways(['optasm']), - collect_compiler_stats('bytes allocated', 2) + collect_compiler_runtime(2) ], compile, ['']) @@ -483,7 +483,7 @@ test('MultiLayerModulesNoCode', ['MultiLayerModulesNoCode.script']) test('MultiComponentModulesRecomp', - [ collect_compiler_stats('bytes allocated', 2), + [ collect_compiler_runtime(2), pre_cmd('$MAKE -s --no-print-directory MultiComponentModulesRecomp'), extra_files(['genMultiComp.py']), compile_timeout_multiplier(5) @@ -492,7 +492,7 @@ test('MultiComponentModulesRecomp', [['unitp%d' % n for n in range(20)], '-fno-code -fwrite-interface -v0']) test('MultiComponentModules', - [ collect_compiler_stats('bytes allocated', 2), + [ collect_compiler_runtime(2), pre_cmd('$PYTHON ./genMultiComp.py'), extra_files(['genMultiComp.py']), compile_timeout_multiplier(5) @@ -565,7 +565,7 @@ test('T14683', test ('T9630', [ collect_compiler_residency(15), - collect_compiler_stats('bytes allocated', 2), + collect_compiler_runtime(2), ], multimod_compile, ['T9630', '-v0 -O']) @@ -611,7 +611,7 @@ test ('T16473', ['-O2 -flate-specialise']) test('T17516', - [ collect_compiler_stats('bytes allocated', 5), + [ collect_compiler_runtime(5), ], multimod_compile, ['T17516', '-O -v0']) @@ -635,13 +635,13 @@ test ('T18140', ['-v0 -O']) test('T10421', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], multimod_compile, ['T10421', '-v0 -O']) test('T10421a', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 10) + collect_compiler_runtime(10) ], multimod_compile, ['T10421a', '-v0 -O']) @@ -700,13 +700,13 @@ test ('T19695', ['-v0 -O2']) test('hard_hole_fits', # Testing multiple hole-fits with lots in scope for #16875 - collect_compiler_stats('bytes allocated', 2), # 1 is 300s, 0.010 is 3s. Without hole-fits it takes 1s + collect_compiler_runtime(2), # 1 is 300s, 0.010 is 3s. Without hole-fits it takes 1s compile, ['-fdefer-type-errors -fno-max-valid-hole-fits -package ghc']) test('T16875', # Testing one hole-fit with a lot in scope for #16875 # This test is very sensitive to environmental differences.. we should fix # that but for now the failure threshold is 4% (see #21557) - collect_compiler_stats('bytes allocated', 4), + collect_compiler_runtime(4), compile, ['-fdefer-type-errors -fno-max-valid-hole-fits -package ghc']) test ('T20261', @@ -720,7 +720,7 @@ test ('T20261', # a compile-time and a run-time performance test test('T21839c', [ collect_compiler_stats('all', 10), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), only_ways(['normal'])], compile, ['-O']) ===================================== testsuite/tests/rename/should_fail/T25056.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE RecordWildCards #-} +module T25056 where + +import T25056b + +foo :: T -> () +foo (T { unT = x }) = x ===================================== testsuite/tests/rename/should_fail/T25056.stderr ===================================== @@ -0,0 +1,5 @@ +T25056.hs:7:10: error: [GHC-01928] + • Illegal term-level use of the type constructor ‘T’ + • imported from ‘T25056b’ at T25056.hs:4:1-14 + (and originally defined in ‘T25056a’ at T25056a.hs:8:1-14) + ===================================== testsuite/tests/rename/should_fail/T25056a.hs ===================================== @@ -0,0 +1,13 @@ +{-# LANGUAGE PatternSynonyms #-} +module T25056a + ( T + , T_(unT) + , pattern T + ) where + +type T = T_ () + +data T_ a = PrivateT { unT_ :: a } + +pattern T :: a -> T_ a +pattern T { unT } <- PrivateT { unT_ = unT } ===================================== testsuite/tests/rename/should_fail/T25056b.hs ===================================== @@ -0,0 +1,3 @@ +module T25056b (T, T_(..)) where + +import T25056a (T, T_(..)) ===================================== testsuite/tests/rename/should_fail/all.T ===================================== @@ -222,6 +222,7 @@ test('T23740g', normal, compile_fail, ['']) test('T23740h', normal, compile_fail, ['']) test('T23740i', req_th, compile_fail, ['']) test('T23740j', normal, compile_fail, ['']) +test('T25056', [extra_files(['T25056a.hs', 'T25056b.hs'])], multimod_compile_fail, ['T25056', '-v0']) test('Or3', normal, compile_fail, ['']) test('T23570', [extra_files(['T23570_aux.hs'])], multimod_compile_fail, ['T23570', '-v0']) test('T23570b', [extra_files(['T23570_aux.hs'])], multimod_compile, ['T23570b', '-v0']) ===================================== testsuite/tests/typecheck/should_fail/T23739b.hs ===================================== @@ -8,7 +8,4 @@ g1 :: Int -> Unit g1 Int = () g2 :: Int -g2 = Int{} - -g3 :: Int -g3 = Int +g2 = Int ===================================== testsuite/tests/typecheck/should_fail/T23739b.stderr ===================================== @@ -6,16 +6,9 @@ T23739b.hs:8:4: error: [GHC-01928] In an equation for ‘g1’: g1 Int = () T23739b.hs:11:6: error: [GHC-01928] - • Illegal term-level use of the type constructor ‘Int’ - • imported from ‘Prelude’ at T23739b.hs:2:8-14 - (and originally defined in ‘GHC.Types’) - • In the expression: Int {} - In an equation for ‘g2’: g2 = Int {} - -T23739b.hs:14:6: error: [GHC-01928] • Illegal term-level use of the type constructor ‘Int’ • imported from ‘Prelude’ at T23739b.hs:2:8-14 (and originally defined in ‘GHC.Types’) • In the expression: Int - In an equation for ‘g3’: g3 = Int + In an equation for ‘g2’: g2 = Int ===================================== testsuite/tests/typecheck/should_fail/T23739c.hs ===================================== @@ -0,0 +1,8 @@ + +module T23739c where + +import Data.Tuple.Experimental +import GHC.TypeLits + +g :: Int +g = Int{} ===================================== testsuite/tests/typecheck/should_fail/T23739c.stderr ===================================== @@ -0,0 +1,7 @@ +T23739c.hs:8:5: error: [GHC-01928] + • Illegal term-level use of the type constructor ‘Int’ + • imported from ‘Prelude’ at T23739c.hs:2:8-14 + (and originally defined in ‘GHC.Types’) + • In the expression: Int {} + In an equation for ‘g’: g = Int {} + ===================================== testsuite/tests/typecheck/should_fail/all.T ===================================== @@ -727,5 +727,6 @@ test('T17594g', normal, compile_fail, ['']) test('T24470a', normal, compile_fail, ['']) test('T24553', normal, compile_fail, ['']) test('T23739b', normal, compile_fail, ['']) +test('T23739c', normal, compile_fail, ['']) test('T24868', normal, compile_fail, ['']) test('T24938', normal, compile_fail, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1796173632b68c8878976fc04ac9973f16b3619d...b9c46af5f341a67b1318f3a44c43f29d5478d510 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1796173632b68c8878976fc04ac9973f16b3619d...b9c46af5f341a67b1318f3a44c43f29d5478d510 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 13:13:32 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Wed, 04 Sep 2024 09:13:32 -0400 Subject: [Git][ghc/ghc][wip/fltused] LLVM: use -relocation-model=pic on Windows Message-ID: <66d85cfc41174_74e0b8aafc668a9@gitlab.mail> sheaf pushed to branch wip/fltused at Glasgow Haskell Compiler / GHC Commits: f18b4797 by sheaf at 2024-09-04T15:13:25+02:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - 4 changed files: - compiler/GHC/Driver/Pipeline/Execute.hs - + testsuite/tests/llvm/should_run/T22487.hs - + testsuite/tests/llvm/should_run/T22487.stdout - + testsuite/tests/llvm/should_run/all.T Changes: ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -969,13 +969,18 @@ llvmOptions llvm_config dflags = ++ [("", "-target-abi=" ++ abi) | not (null abi) ] where target = platformMisc_llvmTarget $ platformMisc dflags + target_os = platformOS (targetPlatform dflags) Just (LlvmTarget _ mcpu mattr) = lookup target (llvmTargets llvm_config) -- Relocation models - rmodel | gopt Opt_PIC dflags = "pic" - | positionIndependent dflags = "pic" - | ways dflags `hasWay` WayDyn = "dynamic-no-pic" - | otherwise = "static" + rmodel | gopt Opt_PIC dflags + || positionIndependent dflags + || target_os == OSMinGW32 -- #22487: use PIC on (64-bit) Windows + = "pic" + | ways dflags `hasWay` WayDyn + = "dynamic-no-pic" + | otherwise + = "static" platform = targetPlatform dflags arch = platformArch platform ===================================== testsuite/tests/llvm/should_run/T22487.hs ===================================== @@ -0,0 +1,8 @@ + +module Main where + +add :: Double -> Double -> Double +add x y = x + y +{-# NOINLINE add #-} +main = do putStrLn "Hello world!" + print (add 1.0 2.0) ===================================== testsuite/tests/llvm/should_run/T22487.stdout ===================================== @@ -0,0 +1,2 @@ +Hello world! +3.0 ===================================== testsuite/tests/llvm/should_run/all.T ===================================== @@ -0,0 +1,15 @@ + +def f( name, opts ): + opts.only_ways = ['optllvm', 'llvm', 'debugllvm'] + +setTestOpts(f) + +# Apples LLVM Toolchain knows about a `vortex` cpu (and possibly others), that +# the stock LLVM toolchain doesn't know about and will warn about. Let's not +# have test fail just because of processor name differences due to different +# LLVM Toolchains. GHC tries to pass what apple expects (on darwin), but can +# be used with the stock LLVM toolchain as well. +def ignore_llvm_and_vortex( msg ): + return re.sub(r".* is not a recognized processor for this target.*\n",r"",msg) + +test('T22487', [normal, normalise_errmsg_fun(ignore_llvm_and_vortex)], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f18b47975adef5f9a9a2c4c33a5e95ca9741ee02 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f18b47975adef5f9a9a2c4c33a5e95ca9741ee02 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 15:25:29 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 04 Sep 2024 11:25:29 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 7 commits: UniqDSM det uniques + use in Cmm.Info Message-ID: <66d87be916936_3af607d0d7c87885@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 7adb8189 by Rodrigo Mesquita at 2024-09-04T16:24:33+01:00 UniqDSM det uniques + use in Cmm.Info Now for SRTs SRT generation using deterministic uniq supply Back LabelMap with deterministic UDFM TSAN uniq rename hard Revert "TSAN uniq rename hard" This reverts commit 7ca5ab3036c15f38c6d4cbcb616d415958c6bcda. improvements to uniqdsm UniqDSM ProcPoint CmmLayoutStack UniqDet 90% of cpsTop UniqDSM Major progress in using UniqDSM in CmmToAsm and Ncg backends Fix imports Un-back label map with udfm Revert "Un-back label map with udfm" This reverts commit f5d2e4257214a3f7b7d845651e6662c5babfd6a3. Make UDSM oneshot deriving via state Fix -fllvm hang - - - - - 02596b5d by Rodrigo Mesquita at 2024-09-04T16:25:04+01:00 Remame uniques straight off stgtocmm, before cmm pipeline WIP Progress Work around LLVM assembler bug! In a really stupid way) Fix ordering of CLabels for IdLabels Local test script tweaks Do uniq renaming before SRTs Revert "Do uniq renaming before SRTs" This reverts commit db38b635d626106e40b3ab18091e0a24046c30c5. Do on CmmGroup Do uniq-renaming pass right at `codeGen` not better Revert "Do uniq-renaming pass right at `codeGen`" This reverts commit 74e9068aaaf736bf815a36bf74a0dde19a074a7a. Reapply "Do uniq renaming before SRTs" This reverts commit 682f89732fc2a95fa011f530c0c6922bf576d229. Try ALSO after SRT Revert "Try ALSO after SRT" This reverts commit c5dd7b426cde768126402aac3f39617ccb99f5c5. Renaming before and after SRTs bc of procs and srts and ... Wait no that was way too slow... cleaner approach, same idea Put deterministic renaming behind a flag Fix Ord CLabel only compare externalNames UniqRnem fixes external names DCmmDecl UniqRenam Refactor ProfilingInfo to preserve Unique information before rendering it Rename Profiling Info now that names are preserved Revert "Rename Profiling Info now that names are preserved" This reverts commit 2dd3da96b7e771ae272791a00d7fb55313401c9e. Revert "Refactor ProfilingInfo to preserve Unique information before rendering it" This reverts commit 8aba0515bb744ca5add6a4c3c9c7760e226e0b31. Performance tweaks Get rid of UniqRenamable class, do it directly Make sure graph is renamed first, info table last Turns out it does matter! Whitespace - - - - - 96f65fa8 by Rodrigo Mesquita at 2024-09-04T16:25:05+01:00 cmm: Back LabelMap with UDFM Use a deterministic unique map to back the implementation of `LabelMap`. This is necessary towards the goal of object code determinism in #12935. Our intended solution requires renaming uniques in a deterministic order (which will be the order in which they were created), but storing them label map makes us lose this order. Backing it with a UDFM fixes this issue. Introduce back LabelMap non deterministic Use NonDeterministic Label map in multiple passes (TODO: More could be available. Look through Det LabelMap uses again) Use NonDet for CFG More NonDet More explicit Introduce DCmmDecl, start Removing more maps - - - - - 59813fd7 by Rodrigo Mesquita at 2024-09-04T16:25:05+01:00 Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled - - - - - a7555a1f by Rodrigo Mesquita at 2024-09-04T16:25:05+01:00 distinct-constructor-tables determinism - - - - - 75cd963d by Rodrigo Mesquita at 2024-09-04T16:25:06+01:00 Rename deterministically CmmGroups in generateCgIPEStub - - - - - 9e260112 by Rodrigo Mesquita at 2024-09-04T16:25:06+01:00 Writing Notes and Cleaning up v1 Fix linters MP entries - - - - - 30 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Reg/Graph.hs - compiler/GHC/CmmToAsm/Reg/Graph/Spill.hs - compiler/GHC/CmmToAsm/Reg/Linear.hs - compiler/GHC/CmmToAsm/Reg/Linear/Base.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8c244e4d5376adb99df1eb2008b01a7078193e29...9e26011233806205d37aba2605f4ed3341f73e36 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8c244e4d5376adb99df1eb2008b01a7078193e29...9e26011233806205d37aba2605f4ed3341f73e36 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 16:18:54 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 04 Sep 2024 12:18:54 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/exception-propagate Message-ID: <66d8886e691f9_3af60754f08899736@gitlab.mail> Matthew Pickering pushed new branch wip/exception-propagate at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/exception-propagate You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 16:24:49 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 04 Sep 2024 12:24:49 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] 2 commits: fix order of arguments Message-ID: <66d889d18040b_3af60742da8899972@gitlab.mail> Matthew Pickering pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: 5a07a053 by Matthew Pickering at 2024-09-04T17:20:04+01:00 fix order of arguments - - - - - b47b517b by Matthew Pickering at 2024-09-04T17:24:37+01:00 pointers - - - - - 3 changed files: - libraries/ghc-internal/src/GHC/Internal/Control/Monad/ST/Imp.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs Changes: ===================================== libraries/ghc-internal/src/GHC/Internal/Control/Monad/ST/Imp.hs ===================================== @@ -55,6 +55,7 @@ fixST k = unsafeIOToST $ do m <- newEmptyMVar ans <- unsafeDupableInterleaveIO (readMVar m `catch` \BlockedIndefinitelyOnMVar -> + -- TODO, should this propagate the error? throwIO NonTermination) result <- unsafeSTToIO (k ans) putMVar m result ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs ===================================== @@ -123,7 +123,7 @@ addFilePathToIOError fun fp ioe catchAndAnnotate :: FilePath -> String -> IO a -> IO a catchAndAnnotate fp s a = catchExceptionNoPropagate @IOError a - (\(ExceptionWithContext c e) -> rethrowIO (ExceptionWithContext c (addFilePathToIOError fp s e))) + (\(ExceptionWithContext c e) -> rethrowIO (ExceptionWithContext c (addFilePathToIOError s fp e))) -- | Specialised 'rethrowIO' to 'IOError', helpful for type inference rethrowError :: ExceptionWithContext IOError -> IO a ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs ===================================== @@ -184,6 +184,7 @@ do_operation fun h act m = do putMVar m h_ case () of _ | Just ioe <- fromException e -> + -- TODO: Stop this adding a callstack ioError (augmentIOError ioe fun h) _ | Just async_ex <- fromException e -> do -- see Note [async] let _ = async_ex :: SomeAsyncException View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/bcc33162465ea3cace930668887ac59f0f6aa9f0...b47b517b4ae17cca113c3462e79708a58941a012 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/bcc33162465ea3cace930668887ac59f0f6aa9f0...b47b517b4ae17cca113c3462e79708a58941a012 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 16:25:43 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 04 Sep 2024 12:25:43 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 6 commits: Deterministic Uniques in the NCG Message-ID: <66d88a0750507_3af60760aba810015e@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 789ab550 by Rodrigo Mesquita at 2024-09-04T17:23:33+01:00 Deterministic Uniques in the NCG See Note [Deterministic Uniques in the NCG] UniqDSM det uniques + use in Cmm.Info Now for SRTs SRT generation using deterministic uniq supply Back LabelMap with deterministic UDFM TSAN uniq rename hard Revert "TSAN uniq rename hard" This reverts commit 7ca5ab3036c15f38c6d4cbcb616d415958c6bcda. improvements to uniqdsm UniqDSM ProcPoint CmmLayoutStack UniqDet 90% of cpsTop UniqDSM Major progress in using UniqDSM in CmmToAsm and Ncg backends Fix imports Un-back label map with udfm Revert "Un-back label map with udfm" This reverts commit f5d2e4257214a3f7b7d845651e6662c5babfd6a3. Make UDSM oneshot deriving via state Fix -fllvm hang - - - - - 417c7937 by Rodrigo Mesquita at 2024-09-04T17:24:07+01:00 Renaming Cmm uniques deterministically See Note [Object determinism] and Note [Renaming uniques deterministically] Remame uniques straight off stgtocmm, before cmm pipeline WIP Progress Work around LLVM assembler bug! In a really stupid way) Fix ordering of CLabels for IdLabels Local test script tweaks Do uniq renaming before SRTs Revert "Do uniq renaming before SRTs" This reverts commit db38b635d626106e40b3ab18091e0a24046c30c5. Do on CmmGroup Do uniq-renaming pass right at `codeGen` not better Revert "Do uniq-renaming pass right at `codeGen`" This reverts commit 74e9068aaaf736bf815a36bf74a0dde19a074a7a. Reapply "Do uniq renaming before SRTs" This reverts commit 682f89732fc2a95fa011f530c0c6922bf576d229. Try ALSO after SRT Revert "Try ALSO after SRT" This reverts commit c5dd7b426cde768126402aac3f39617ccb99f5c5. Renaming before and after SRTs bc of procs and srts and ... Wait no that was way too slow... cleaner approach, same idea Put deterministic renaming behind a flag Fix Ord CLabel only compare externalNames UniqRnem fixes external names DCmmDecl UniqRenam Refactor ProfilingInfo to preserve Unique information before rendering it Rename Profiling Info now that names are preserved Revert "Rename Profiling Info now that names are preserved" This reverts commit 2dd3da96b7e771ae272791a00d7fb55313401c9e. Revert "Refactor ProfilingInfo to preserve Unique information before rendering it" This reverts commit 8aba0515bb744ca5add6a4c3c9c7760e226e0b31. Performance tweaks Get rid of UniqRenamable class, do it directly Make sure graph is renamed first, info table last Turns out it does matter! Whitespace - - - - - c3dad666 by Rodrigo Mesquita at 2024-09-04T17:24:45+01:00 DCmmGroup vs CmmGroup or: Deterministic Info Tables See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and Note [Object determinism] cmm: Back LabelMap with UDFM Use a deterministic unique map to back the implementation of `LabelMap`. This is necessary towards the goal of object code determinism in #12935. Our intended solution requires renaming uniques in a deterministic order (which will be the order in which they were created), but storing them label map makes us lose this order. Backing it with a UDFM fixes this issue. Introduce back LabelMap non deterministic Use NonDeterministic Label map in multiple passes (TODO: More could be available. Look through Det LabelMap uses again) Use NonDet for CFG More NonDet More explicit Introduce DCmmDecl, start Removing more maps - - - - - 2e53ae13 by Rodrigo Mesquita at 2024-09-04T17:25:27+01:00 Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled - - - - - 75901ed5 by Rodrigo Mesquita at 2024-09-04T17:25:27+01:00 distinct-constructor-tables determinism - - - - - 3dc7d298 by Rodrigo Mesquita at 2024-09-04T17:25:27+01:00 Rename deterministically CmmGroups in generateCgIPEStub - - - - - 30 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Reg/Graph.hs - compiler/GHC/CmmToAsm/Reg/Graph/Spill.hs - compiler/GHC/CmmToAsm/Reg/Linear.hs - compiler/GHC/CmmToAsm/Reg/Linear/Base.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9e26011233806205d37aba2605f4ed3341f73e36...3dc7d298060098fdf864dc83e87152521f11d212 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9e26011233806205d37aba2605f4ed3341f73e36...3dc7d298060098fdf864dc83e87152521f11d212 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 16:56:02 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 04 Sep 2024 12:56:02 -0400 Subject: [Git][ghc/ghc][wip/abi-test] 35 commits: isIrrefutableHsPat: look up ConLikes in the HscEnv Message-ID: <66d8912240b36_1ddafa8ad04595df@gitlab.mail> Rodrigo Mesquita pushed to branch wip/abi-test at Glasgow Haskell Compiler / GHC Commits: bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - 605eb7f5 by Matthew Pickering at 2024-09-04T16:55:57+00:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 30 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/Core/Predicate.hs - compiler/GHC/Core/Type.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Hs/Decls.hs - compiler/GHC/Hs/ImpExp.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/HsToCore.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/HsToCore/Usage.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Iface/Ext/Ast.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0a534ae2f249c794530f01c041c494fd75159de0...605eb7f515221e167a9482017fdc8cd5b83d2786 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0a534ae2f249c794530f01c041c494fd75159de0...605eb7f515221e167a9482017fdc8cd5b83d2786 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 4 17:29:07 2024 From: gitlab at gitlab.haskell.org (Matthew Craven (@clyring)) Date: Wed, 04 Sep 2024 13:29:07 -0400 Subject: [Git][ghc/ghc][wip/T23490-part2] 8 commits: rts: fix checkClosure error message Message-ID: <66d898e31606_1ddafa3a4cb0655a7@gitlab.mail> Matthew Craven pushed to branch wip/T23490-part2 at Glasgow Haskell Compiler / GHC Commits: 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - 05b0816d by Matthew Craven at 2024-09-04T13:26:14-04:00 Bump transformers submodule The svg image files mentioned in transformers.cabal were previously not checked in, which broke sdist generation. - - - - - 76e7295c by Matthew Craven at 2024-09-04T13:27:53-04:00 Remove reference to non-existent file in haddock.cabal - - - - - 228e48bc by Matthew Craven at 2024-09-04T13:27:53-04:00 Update hackage index state - - - - - 266a978b by Matthew Craven at 2024-09-04T13:27:53-04:00 Move tests T11462 and T11525 into tests/tcplugins - - - - - 58090b05 by Matthew Craven at 2024-09-04T13:28:24-04:00 Repair the 'build-cabal' hadrian target Fixes #23117. Fixes #23281. Fixes #23490. This required: * Updating the bit-rotted compiler/Setup.hs and its setup-depends * Listing a few recently-added libraries and utilities in cabal.project-reinstall * Setting allow-boot-library-installs to 'True' since Cabal now considers the 'ghc' package itself a boot library for the purposes of this flag Additionally, the allow-newer block in cabal.project-reinstall was removed. This block was probably added because when the libraries/Cabal submodule is too new relative to the cabal-install executable, solving the setup-depends for any package with a custom setup requires building an old Cabal (from Hackage) against the in-tree version of base, and this can fail un-necessarily due to tight version bounds on base. However, the blind allow-newer can also cause the solver to go berserk and choose a stupid build plan that has no business succeeding, and the failures when this happens are dreadfully confusing. (See #23281 and #24363.) Why does setup-depends solving insist on an old version of Cabal? See: https://github.com/haskell/cabal/blob/0a0b33983b0f022b9697f7df3a69358ee9061a89/cabal-install/src/Distribution/Client/ProjectPlanning.hs#L1393-L1410 The right solution here is probably to use the in-tree cabal-install from libraries/Cabal/cabal-install with the build-cabal target rather than whatever the environment happens to provide. But this is left for future work. - - - - - a547477e by Matthew Craven at 2024-09-04T13:28:33-04:00 Revert "CI: Disable the test-cabal-reinstall job" This reverts commit 38c3afb64d3ffc42f12163c6f0f0d5c414aa8255. - - - - - 18 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - cabal.project-reinstall - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/Setup.hs - compiler/ghc.cabal.in - hadrian/cabal.project - libraries/transformers - rts/sm/Sanity.c - testsuite/tests/typecheck/should_compile/T11462.hs → testsuite/tests/tcplugins/T11462.hs - testsuite/tests/typecheck/should_compile/T11462_Plugin.hs → testsuite/tests/tcplugins/T11462_Plugin.hs - testsuite/tests/typecheck/should_compile/T11525.hs → testsuite/tests/tcplugins/T11525.hs - testsuite/tests/typecheck/should_compile/T11525_Plugin.hs → testsuite/tests/tcplugins/T11525_Plugin.hs - testsuite/tests/tcplugins/all.T - testsuite/tests/typecheck/should_compile/all.T - utils/haddock/haddock.cabal Changes: ===================================== .gitlab-ci.yml ===================================== @@ -489,21 +489,16 @@ stack-hadrian-build: # Testing reinstallable ghc codepath #################################### -# As documented on the original ticket #19896, this feature already has a long -# way to go before it can actually be used. Meanwhile, parts of it have -# bit-rotted, possibly related to some Cabal change. The job is disabled for -# now. -# -# test-cabal-reinstall-x86_64-linux-deb10: -# extends: nightly-x86_64-linux-deb10-validate -# stage: full-build -# variables: -# REINSTALL_GHC: "yes" -# BUILD_FLAVOUR: validate -# TEST_ENV: "x86_64-linux-deb10-cabal-install" -# rules: -# - if: $NIGHTLY -# - if: '$CI_MERGE_REQUEST_LABELS =~ /.*test-reinstall.*/' +test-cabal-reinstall-x86_64-linux-deb10: + extends: nightly-x86_64-linux-deb10-validate + stage: full-build + variables: + REINSTALL_GHC: "yes" + BUILD_FLAVOUR: validate + TEST_ENV: "x86_64-linux-deb10-cabal-install" + rules: + - if: $NIGHTLY + - if: '$CI_MERGE_REQUEST_LABELS =~ /.*test-reinstall.*/' ######################################## # Testing ABI is invariant across builds ===================================== .gitlab/ci.sh ===================================== @@ -7,7 +7,7 @@ set -Eeuo pipefail # Configuration: -HACKAGE_INDEX_STATE="2024-05-13T15:04:38Z" +HACKAGE_INDEX_STATE="2024-06-27T02:07:41Z" MIN_HAPPY_VERSION="1.20" MIN_ALEX_VERSION="3.2.6" ===================================== cabal.project-reinstall ===================================== @@ -12,11 +12,13 @@ packages: ./compiler -- ./libraries/deepseq/ ./libraries/directory/ ./libraries/exceptions/ + ./libraries/file-io/ ./libraries/filepath/ -- ./libraries/ghc-bignum/ ./libraries/ghc-boot/ -- ./libraries/ghc-boot-th/ ./libraries/ghc-compact + ./libraries/ghc-experimental ./libraries/ghc-heap ./libraries/ghci -- ./libraries/ghc-prim @@ -25,6 +27,7 @@ packages: ./compiler ./libraries/hpc -- ./libraries/integer-gmp ./libraries/mtl/ + ./libraries/os-string/ ./libraries/parsec/ -- ./libraries/pretty/ ./libraries/process/ @@ -39,7 +42,10 @@ packages: ./compiler ./libraries/Win32/ ./libraries/xhtml/ ./utils/ghc-pkg + ./utils/ghc-toolchain + ./utils/ghc-toolchain/exe ./utils/haddock + ./utils/haddock/haddock-api ./utils/hp2ps ./utils/hpc ./utils/hsc2hs @@ -61,15 +67,10 @@ constraints: ghc +internal-interpreter +dynamic-system-linke, any.pretty installed, any.template-haskell installed -allow-newer: - ghc-paths:Cabal, - *:base, - *:ghc-prim, - tree-diff:time benchmarks: False tests: False -allow-boot-library-installs: False +allow-boot-library-installs: True -- Workaround for https://github.com/haskell/cabal/issues/7297 package * ===================================== compiler/GHC/CmmToAsm/AArch64/CodeGen.hs ===================================== @@ -1990,9 +1990,13 @@ genCCall target dest_regs arg_regs = do MO_SubIntC _w -> unsupported mop -- Memory Ordering - MO_AcquireFence -> return (unitOL DMBISH) - MO_ReleaseFence -> return (unitOL DMBISH) - MO_SeqCstFence -> return (unitOL DMBISH) + -- Set flags according to their C pendants (stdatomic.h): + -- atomic_thread_fence(memory_order_acquire); // -> dmb ishld + MO_AcquireFence -> return . unitOL $ DMBISH DmbLoad + -- atomic_thread_fence(memory_order_release); // -> dmb ish + MO_ReleaseFence -> return . unitOL $ DMBISH DmbLoadStore + -- atomic_thread_fence(memory_order_seq_cst); // -> dmb ish + MO_SeqCstFence -> return . unitOL $ DMBISH DmbLoadStore MO_Touch -> return nilOL -- Keep variables live (when using interior pointers) -- Prefetch MO_Prefetch_Data _n -> return nilOL -- Prefetch hint. ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -134,7 +134,7 @@ regUsageOfInstr platform instr = case instr of LDAR _ dst src -> usage (regOp src, regOp dst) -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> usage ([], []) + DMBISH _ -> usage ([], []) -- 9. Floating Point Instructions -------------------------------------------- FMOV dst src -> usage (regOp src, regOp dst) @@ -281,7 +281,7 @@ patchRegsOfInstr instr env = case instr of LDAR f o1 o2 -> LDAR f (patchOp o1) (patchOp o2) -- 8. Synchronization Instructions ----------------------------------------- - DMBISH -> DMBISH + DMBISH c -> DMBISH c -- 9. Floating Point Instructions ------------------------------------------ FMOV o1 o2 -> FMOV (patchOp o1) (patchOp o2) @@ -649,7 +649,7 @@ data Instr | BCOND Cond Target -- branch with condition. b. -- 8. Synchronization Instructions ----------------------------------------- - | DMBISH + | DMBISH DMBISHFlags -- 9. Floating Point Instructions -- move to/from general purpose <-> floating, or floating to floating | FMOV Operand Operand @@ -672,6 +672,9 @@ data Instr -- - fnmadd: d = - r1 * r2 - r3 | FMA FMASign Operand Operand Operand Operand +data DMBISHFlags = DmbLoad | DmbLoadStore + deriving (Eq, Show) + instrCon :: Instr -> String instrCon i = case i of ===================================== compiler/GHC/CmmToAsm/AArch64/Ppr.hs ===================================== @@ -527,7 +527,8 @@ pprInstr platform instr = case instr of LDAR _f o1 o2 -> op2 (text "\tldar") o1 o2 -- 8. Synchronization Instructions ------------------------------------------- - DMBISH -> line $ text "\tdmb ish" + DMBISH DmbLoadStore -> line $ text "\tdmb ish" + DMBISH DmbLoad -> line $ text "\tdmb ishld" -- 9. Floating Point Instructions -------------------------------------------- FMOV o1 o2 -> op2 (text "\tfmov") o1 o2 ===================================== compiler/Setup.hs ===================================== @@ -1,4 +1,4 @@ -{-# LANGUAGE RecordWildCards #-} +{-# LANGUAGE NamedFieldPuns #-} module Main where import Distribution.Simple @@ -52,10 +52,12 @@ primopIncls = , ("primop-vector-tys-exports.hs-incl", "--primop-vector-tys-exports") , ("primop-vector-tycons.hs-incl" , "--primop-vector-tycons") , ("primop-docs.hs-incl" , "--wired-in-docs") + , ("primop-deprecations.hs-incl" , "--wired-in-deprecations") ] ghcAutogen :: Verbosity -> LocalBuildInfo -> IO () -ghcAutogen verbosity lbi at LocalBuildInfo{..} = do +ghcAutogen verbosity lbi at LocalBuildInfo{pkgDescrFile,withPrograms,componentNameMap} + = do -- Get compiler/ root directory from the cabal file let Just compilerRoot = takeDirectory <$> pkgDescrFile @@ -77,7 +79,7 @@ ghcAutogen verbosity lbi at LocalBuildInfo{..} = do -- Call genprimopcode to generate *.hs-incl forM_ primopIncls $ \(file,command) -> do contents <- readProcess "genprimopcode" [command] primopsStr - rewriteFileEx verbosity (buildDir file) contents + rewriteFileEx verbosity (buildDir lbi file) contents -- Write GHC.Platform.Constants let platformConstantsPath = autogenPackageModulesDir lbi "GHC/Platform/Constants.hs" ===================================== compiler/ghc.cabal.in ===================================== @@ -50,7 +50,7 @@ extra-source-files: custom-setup - setup-depends: base >= 3 && < 5, Cabal >= 1.6 && <3.10, directory, process, filepath, containers + setup-depends: base >= 3 && < 5, Cabal >= 1.6 && <3.14, directory, process, filepath, containers Flag internal-interpreter Description: Build with internal interpreter support. ===================================== hadrian/cabal.project ===================================== @@ -12,7 +12,7 @@ packages: ./ -- This essentially freezes the build plan for hadrian -- It would be wise to keep this up to date with the state set in ci.sh -index-state: 2024-05-13T15:04:38Z +index-state: 2024-06-27T02:07:41Z -- unordered-containers-0.2.20-r1 requires template-haskell < 2.22 -- ghc-9.10 has template-haskell-2.22.0.0 ===================================== libraries/transformers ===================================== @@ -1 +1 @@ -Subproject commit ba3503905dec072acc6515323c884706efd4dbb4 +Subproject commit ac377ff5abfb65d84988a5b9edb2231df1c0cea3 ===================================== rts/sm/Sanity.c ===================================== @@ -357,7 +357,8 @@ checkClosure( const StgClosure* p ) info = ACQUIRE_LOAD(&p->header.info); if (IS_FORWARDING_PTR(info)) { - barf("checkClosure: found EVACUATED closure %d", info->type); + ASSERT(LOOKS_LIKE_CLOSURE_PTR(info)); + barf("checkClosure: found EVACUATED closure %u", GET_INFO((StgClosure*)UN_FORWARDING_PTR(info))->type); } #if defined(PROFILING) ===================================== testsuite/tests/typecheck/should_compile/T11462.hs → testsuite/tests/tcplugins/T11462.hs ===================================== ===================================== testsuite/tests/typecheck/should_compile/T11462_Plugin.hs → testsuite/tests/tcplugins/T11462_Plugin.hs ===================================== ===================================== testsuite/tests/typecheck/should_compile/T11525.hs → testsuite/tests/tcplugins/T11525.hs ===================================== ===================================== testsuite/tests/typecheck/should_compile/T11525_Plugin.hs → testsuite/tests/tcplugins/T11525_Plugin.hs ===================================== ===================================== testsuite/tests/tcplugins/all.T ===================================== @@ -106,3 +106,10 @@ test('TcPlugin_CtId' , [ 'TcPlugin_CtId.hs' , '-dynamic -package ghc' if have_dynamic() else '-package ghc' ] ) + +test('T11462', [js_broken(22261), req_th], multi_compile, + [None, [('T11462_Plugin.hs', '-package ghc'), ('T11462.hs', '')], + '-dynamic' if have_dynamic() else '']) +test('T11525', [js_broken(22261), req_th], multi_compile, + [None, [('T11525_Plugin.hs', '-package ghc'), ('T11525.hs', '')], + '-dynamic' if have_dynamic() else '']) ===================================== testsuite/tests/typecheck/should_compile/all.T ===================================== @@ -491,9 +491,6 @@ test('T10592', normal, compile, ['']) test('T11305', normal, compile, ['']) test('T11254', normal, compile, ['']) test('T11379', normal, compile, ['']) -test('T11462', [js_broken(22261), req_th], multi_compile, - [None, [('T11462_Plugin.hs', '-package ghc'), ('T11462.hs', '')], - '-dynamic' if have_dynamic() else '']) test('T11480', normal, compile, ['']) test('RebindHR', normal, compile, ['']) test('RebindNegate', normal, compile, ['']) @@ -559,9 +556,6 @@ test('T11723', normal, compile, ['']) test('T12987', normal, compile, ['']) test('T11736', normal, compile, ['']) test('T13248', expect_broken(13248), compile, ['']) -test('T11525', [js_broken(22261), req_th], multi_compile, - [None, [('T11525_Plugin.hs', '-package ghc'), ('T11525.hs', '')], - '-dynamic' if have_dynamic() else '']) test('T12923_1', normal, compile, ['']) test('T21208', normal, compile, ['']) test('T12923_2', normal, compile, ['']) ===================================== utils/haddock/haddock.cabal ===================================== @@ -43,7 +43,6 @@ extra-source-files: doc/README.md doc/*.rst doc/conf.py - haddock-api/src/haddock.sh html-test/src/*.hs html-test/ref/*.html hypsrc-test/src/*.hs View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/dc86ccc205c68eb632661894689724dec4174192...a547477e49eb032a0d653c3c88f51fc3eca8171a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/dc86ccc205c68eb632661894689724dec4174192...a547477e49eb032a0d653c3c88f51fc3eca8171a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 00:22:33 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 04 Sep 2024 20:22:33 -0400 Subject: [Git][ghc/ghc][master] 3 commits: testsuite: Add support to capture performance metrics via 'perf' Message-ID: <66d8f9c8f12d7_1250935ce69416128@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - 7 changed files: - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - testsuite/driver/perf_notes.py - testsuite/driver/testglobals.py - testsuite/driver/testlib.py - testsuite/driver/testutil.py - testsuite/tests/perf/compiler/all.T Changes: ===================================== .gitlab/generate-ci/gen_ci.hs ===================================== @@ -155,6 +155,7 @@ data BuildConfig , noSplitSections :: Bool , validateNonmovingGc :: Bool , textWithSIMDUTF :: Bool + , testsuiteUsePerf :: Bool } -- Extra arguments to pass to ./configure due to the BuildConfig @@ -216,6 +217,7 @@ vanilla = BuildConfig , noSplitSections = False , validateNonmovingGc = False , textWithSIMDUTF = False + , testsuiteUsePerf = False } splitSectionsBroken :: BuildConfig -> BuildConfig @@ -268,6 +270,9 @@ tsan = vanilla { threadSanitiser = True } noTntc :: BuildConfig noTntc = vanilla { tablesNextToCode = False } +usePerfProfilingTestsuite :: BuildConfig -> BuildConfig +usePerfProfilingTestsuite bc = bc { testsuiteUsePerf = True } + ----------------------------------------------------------------------------- -- Platform specific variables ----------------------------------------------------------------------------- @@ -288,6 +293,9 @@ runnerTag _ _ = error "Invalid arch/opsys" tags :: Arch -> Opsys -> BuildConfig -> [String] tags arch opsys _bc = [runnerTag arch opsys] -- Tag for which runners we can use +runnerPerfTag :: Arch -> Opsys -> String +runnerPerfTag arch sys = runnerTag arch sys ++ "-perf" + -- These names are used to find the docker image so they have to match what is -- in the docker registry. distroName :: LinuxDistro -> String @@ -775,6 +783,7 @@ job arch opsys buildConfig = NamedJob { name = jobName, jobInfo = Job {..} } | validateNonmovingGc buildConfig ] in "RUNTEST_ARGS" =: unwords runtestArgs + , if testsuiteUsePerf buildConfig then "RUNTEST_ARGS" =: "--config perf_path=perf" else mempty ] jobArtifacts = Artifacts @@ -897,6 +906,12 @@ highCompression = addVariable "XZ_OPT" "-9" useHashUnitIds :: Job -> Job useHashUnitIds = addVariable "HADRIAN_ARGS" "--hash-unit-ids" +-- | Change the tag of the job to make sure the job is scheduled on a +-- runner that has the necessary capabilties to run the job with 'perf' +-- profiling counters. +perfProfilingJobTag :: Arch -> Opsys -> Job -> Job +perfProfilingJobTag arch opsys j = j { jobTags = [ runnerPerfTag arch opsys ] } + -- | Mark the validate job to run in fast-ci mode -- This is default way, to enable all jobs you have to apply the `full-ci` label. fastCI :: JobGroup Job -> JobGroup Job @@ -1000,6 +1015,8 @@ debian_x86 = , modifyNightlyJobs allowFailure (modifyValidateJobs (allowFailure . manual) tsan_jobs) , -- Nightly allowed to fail: #22343 modifyNightlyJobs allowFailure (modifyValidateJobs manual (validateBuilds Amd64 (Linux validate_debian) noTntc)) + -- Run the 'perf' profiling nightly job in the release config. + , perfProfilingJob Amd64 (Linux Debian12) releaseConfig , onlyRule LLVMBackend (validateBuilds Amd64 (Linux validate_debian) llvm) , addValidateRule TestPrimops (standardBuilds Amd64 (Linux validate_debian)) @@ -1010,6 +1027,12 @@ debian_x86 = where validate_debian = Debian12 + perfProfilingJob arch sys buildConfig = + -- Rename the job to avoid conflicts + rename (<> "-perf") + $ modifyJobs (perfProfilingJobTag arch sys) + $ disableValidate (validateBuilds arch sys $ usePerfProfilingTestsuite buildConfig) + tsan_jobs = modifyJobs ( addVariable "TSAN_OPTIONS" "suppressions=$CI_PROJECT_DIR/rts/.tsan-suppressions" ===================================== .gitlab/jobs.yaml ===================================== @@ -1791,6 +1791,69 @@ "XZ_OPT": "-9" } }, + "nightly-x86_64-linux-deb12-release-perf": { + "after_script": [ + ".gitlab/ci.sh save_cache", + ".gitlab/ci.sh save_test_output", + ".gitlab/ci.sh clean", + "cat ci_timings" + ], + "allow_failure": false, + "artifacts": { + "expire_in": "8 weeks", + "paths": [ + "ghc-x86_64-linux-deb12-release.tar.xz", + "junit.xml", + "unexpected-test-output.tar.gz" + ], + "reports": { + "junit": "junit.xml" + }, + "when": "always" + }, + "cache": { + "key": "x86_64-linux-deb12-$CACHE_REV", + "paths": [ + "cabal-cache", + "toolchain" + ] + }, + "dependencies": [], + "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "needs": [ + { + "artifacts": false, + "job": "hadrian-ghc-in-ghci" + } + ], + "rules": [ + { + "if": "(\"true\" == \"true\") && ($RELEASE_JOB != \"yes\") && ($NIGHTLY)", + "when": "on_success" + } + ], + "script": [ + "sudo chown ghc:ghc -R .", + ".gitlab/ci.sh setup", + ".gitlab/ci.sh configure", + ".gitlab/ci.sh build_hadrian", + ".gitlab/ci.sh test_hadrian" + ], + "stage": "full-build", + "tags": [ + "x86_64-linux-perf" + ], + "variables": { + "BIGNUM_BACKEND": "gmp", + "BIN_DIST_NAME": "ghc-x86_64-linux-deb12-release", + "BUILD_FLAVOUR": "release", + "CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", + "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", + "RUNTEST_ARGS": " --config perf_path=perf", + "TEST_ENV": "x86_64-linux-deb12-release", + "XZ_OPT": "-9" + } + }, "nightly-x86_64-linux-deb12-unreg-validate": { "after_script": [ ".gitlab/ci.sh save_cache", ===================================== testsuite/driver/perf_notes.py ===================================== @@ -128,6 +128,41 @@ AllowedPerfChange = NamedTuple('AllowedPerfChange', ('opts', Dict[str, str]) ]) +class MetricAcceptanceWindow: + """ + A strategy for computing an acceptance window for a metric measurement + given a baseline value. + """ + def get_bounds(self, baseline: float) -> Tuple[float, float]: + raise NotImplemented + def describe(self) -> str: + raise NotImplemented + +class AlwaysAccept(MetricAcceptanceWindow): + def get_bounds(self, baseline: float) -> Tuple[float, float]: + return (-1/0, +1/0) + + def describe(self) -> str: + raise NotImplemented + +class RelativeMetricAcceptanceWindow(MetricAcceptanceWindow): + """ + A MetricAcceptanceWindow which accepts measurements within tol-percent of + the baseline. + """ + def __init__(self, tol: float): + """ Accept any metric within tol-percent of the baseline """ + self.__tol = tol + + def get_bounds(self, baseline: float) -> Tuple[float, float]: + lowerBound = trunc( int(baseline) * ((100 - float(self.__tol))/100)) + upperBound = trunc(0.5 + ceil(int(baseline) * ((100 + float(self.__tol))/100))) + + return (lowerBound, upperBound) + + def describe(self) -> str: + return '+/- %1.1f%%' % (100*self.__tol) + def parse_perf_stat(stat_str: str) -> PerfStat: field_vals = stat_str.strip('\t').split('\t') stat = PerfStat(*field_vals) # type: ignore @@ -558,26 +593,32 @@ def get_commit_metric(gitNoteRef, _commit_metric_cache[cacheKeyA] = baseline_by_cache_key_b return baseline_by_cache_key_b.get(cacheKeyB) -# Check test stats. This prints the results for the user. -# actual: the PerfStat with actual value. -# baseline: the expected Baseline value (this should generally be derived from baseline_metric()) -# tolerance_dev: allowed deviation of the actual value from the expected value. -# allowed_perf_changes: allowed changes in stats. This is a dictionary as returned by get_allowed_perf_changes(). -# force_print: Print stats even if the test stat was in the tolerance range. -# Returns a (MetricChange, pass/fail object) tuple. Passes if the stats are within the expected value ranges. def check_stats_change(actual: PerfStat, baseline: Baseline, - tolerance_dev, + acceptance_window: MetricAcceptanceWindow, allowed_perf_changes: Dict[TestName, List[AllowedPerfChange]] = {}, force_print = False ) -> Tuple[MetricChange, Any]: + """ + Check test stats. This prints the results for the user. + + Parameters: + actual: the PerfStat with actual value + baseline: the expected Baseline value (this should generally be derived + from baseline_metric()) + acceptance_window: allowed deviation of the actual value from the expected + value. + allowed_perf_changes: allowed changes in stats. This is a dictionary as + returned by get_allowed_perf_changes(). + force_print: Print stats even if the test stat was in the tolerance range. + + Returns a (MetricChange, pass/fail object) tuple. Passes if the stats are within the expected value ranges. + """ expected_val = baseline.perfStat.value full_name = actual.test + ' (' + actual.way + ')' - lowerBound = trunc( int(expected_val) * ((100 - float(tolerance_dev))/100)) - upperBound = trunc(0.5 + ceil(int(expected_val) * ((100 + float(tolerance_dev))/100))) - actual_dev = round(((float(actual.value) * 100)/ int(expected_val)) - 100, 1) + lowerBound, upperBound = acceptance_window.get_bounds(expected_val) # Find the direction of change. change = MetricChange.NoChange @@ -613,11 +654,12 @@ def check_stats_change(actual: PerfStat, def display(descr, val, extra): print(descr, str(val).rjust(length), extra) - display(' Expected ' + full_name + ' ' + actual.metric + ':', expected_val, '+/-' + str(tolerance_dev) + '%') + display(' Expected ' + full_name + ' ' + actual.metric + ':', expected_val, acceptance_window.describe()) display(' Lower bound ' + full_name + ' ' + actual.metric + ':', lowerBound, '') display(' Upper bound ' + full_name + ' ' + actual.metric + ':', upperBound, '') display(' Actual ' + full_name + ' ' + actual.metric + ':', actual.value, '') if actual.value != expected_val: + actual_dev = round(((float(actual.value) * 100)/ int(expected_val)) - 100, 1) display(' Deviation ' + full_name + ' ' + actual.metric + ':', actual_dev, '%') return (change, result) ===================================== testsuite/driver/testglobals.py ===================================== @@ -49,6 +49,9 @@ class TestConfig: # Path to Ghostscript self.gs = None # type: Optional[Path] + # Path to Linux `perf` tool + self.perf_path = None # type: Optional[Path] + # Run tests requiring Haddock self.haddock = False @@ -472,6 +475,9 @@ class TestOptions: # The extra hadrian dependencies we need for this particular test self.hadrian_deps = set(["test:ghc"]) # type: Set[str] + # Record these `perf-events` counters when compiling this test, if `perf` is available + self.compiler_perf_counters = [] # type: List[str] + @property def testdir(self) -> Path: if self.testdir_raw is None: ===================================== testsuite/driver/testlib.py ===================================== @@ -3,6 +3,7 @@ # (c) Simon Marlow 2002 # +import csv import io import shutil import os @@ -23,12 +24,13 @@ from testglobals import config, ghc_env, default_testopts, brokens, t, \ TestRun, TestResult, TestOptions, PerfMetric from testutil import strip_quotes, lndir, link_or_copy_file, passed, \ failBecause, testing_metrics, residency_testing_metrics, \ + stable_perf_counters, \ PassFail, badResult, memoize from term_color import Color, colored import testutil from cpu_features import have_cpu_feature import perf_notes as Perf -from perf_notes import MetricChange, PerfStat, StatsException +from perf_notes import MetricChange, PerfStat, StatsException, AlwaysAccept, RelativeMetricAcceptanceWindow extra_src_files = {'T4198': ['exitminus1.c']} # TODO: See #12223 from my_typing import * @@ -752,9 +754,14 @@ def find_so(lib): def find_non_inplace_so(lib): return _find_so(lib,path_from_ghcPkg(lib, "dynamic-library-dirs"),False) -# Define a generic stat test, which computes the statistic by calling the function -# given as the third argument. -def collect_generic_stat ( metric, deviation, get_stat ): + +def collect_generic_stat ( metric, deviation: Optional[int], get_stat: Callable[[WayName], str]): + """ + Define a generic stat test, which computes the statistic by calling the function + given as the third argument. + + If no deviation is given, the test cannot fail, but the metric will be recorded nevertheless. + """ return collect_generic_stats ( { metric: { 'deviation': deviation, 'current': get_stat } } ) def _collect_generic_stat(name : TestName, opts, metric_infos): @@ -801,16 +808,27 @@ def collect_stats(metric='all', deviation=20, static_stats_file=None): def statsFile(comp_test: bool, name: str) -> str: return name + ('.comp' if comp_test else '') + '.stats' +def perfStatsFile(comp_test: bool, name: str) -> str: + return name + ('.comp' if comp_test else '') + '.perf.csv' + # This is an internal function that is used only in the implementation. # 'is_compiler_stats_test' is somewhat of an unfortunate name. # If the boolean is set to true, it indicates that this test is one that # measures the performance numbers of the compiler. # As this is a fairly rare case in the testsuite, it defaults to false to # indicate that it is a 'normal' performance test. -def _collect_stats(name: TestName, opts, metrics, deviation, static_stats_file, is_compiler_stats_test=False): +def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], + static_stats_file: Optional[Union[Path,str]], + is_compiler_stats_test: bool = False, is_compiler_perf_test: bool = False) -> None: if not re.match('^[0-9]*[a-zA-Z][a-zA-Z0-9._-]*$', name): failBecause('This test has an invalid name.') + if is_compiler_perf_test and config.perf_path is None: + # If we are doing a 'perf' run but no 'perf' is configured, + # don't try to read the results. + # This is a bit weird, though. + return + # Normalize metrics to a list of strings. if isinstance(metrics, str): if metrics == 'all': @@ -865,11 +883,47 @@ def _collect_stats(name: TestName, opts, metrics, deviation, static_stats_file, assert val is not None return int(val) + # How to read the result of the performance test + def read_perf_stats_file(way, metric_name): + FIELDS = ['value','unit','event','runtime','percent'] + # Confusingly compile time ghci tests are actually runtime tests, so we have + # to go and look for the name.stats file rather than name.comp.stats file. + compiler_stats_test = is_compiler_stats_test and not (way == "ghci" or way == "ghci-opt") + + perf_stats_file = Path(in_testdir(perfStatsFile(compiler_stats_test, name))) + perf_metrics = {} + try: + perf_csv_lines = perf_stats_file.read_text().splitlines() + # Output looks like: + # """ + # # Started on + # + # ,,,,,... + # """ + # + # Ignore empty lines and lines starting with '#' + perf_csv = [l for l in perf_csv_lines if l and not l.startswith('#')] + + perf_stats_csv_reader = csv.DictReader(perf_csv, fieldnames=FIELDS, delimiter=";", quotechar="\"") + for fields in perf_stats_csv_reader: + perf_metrics[fields['event']] = float(fields['value']) + + except IOError as e: + raise StatsException(str(e)) + + val = perf_metrics[metric_name] + if val is None: + print('Failed to find metric: ', metric_name) + raise StatsException("No such metric") + else: + assert val is not None + return int(val) collect_stat = {} for metric_name in metrics: def action_generator(mn): - return lambda way: read_stats_file(way, mn) + read_stats = read_perf_stats_file if is_compiler_perf_test else read_stats_file + return lambda way: read_stats(way, mn) metric = '{}/{}'.format(tag, metric_name) collect_stat[metric] = { "deviation": deviation , "current": action_generator(metric_name) } @@ -1009,7 +1063,32 @@ def have_thread_sanitizer( ) -> bool: def gcc_as_cmmp() -> bool: return config.cmm_cpp_is_gcc -# --- +# ----- + +def collect_compiler_perf(deviation: Optional[int] = None): + """ + Record stable performance counters using `perf stat` when available. + """ + return [ + _collect_compiler_perf_counters(stable_perf_counters(), deviation) + ] + +def collect_compiler_perf_counters(counters: List[str], deviation: Optional[int] = None): + """ + Record the given event counters using `perf stat` when available. + """ + return [ + _collect_compiler_perf_counters(set(counters), deviation) + ] + +def _collect_compiler_perf_counters(counters: Set[str], deviation: Optional[int] = None): + def f(name, opts): + # Slightly hacky, we need the requested perf_counters in 'simple_run'. + # Thus, we have to globally register these counters + opts.compiler_perf_counters += list(counters) + _collect_stats(name, opts, counters, deviation, False, True, True) + return f + # Note [Measuring residency] # ~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -1062,6 +1141,12 @@ def collect_compiler_residency(tolerance_pct: float): collect_compiler_stats(residency_testing_metrics(), tolerance_pct) ] +def collect_compiler_runtime(tolerance_pct: float): + return [ + collect_compiler_stats('bytes allocated', tolerance_pct), + _collect_compiler_perf_counters(stable_perf_counters()) + ] + # --- def high_memory_usage(name, opts): @@ -1619,7 +1704,7 @@ async def do_test(name: TestName, stdout = stdout_path, stderr = stderr_path, print_output = config.verbose >= 3, - timeout_multiplier = opts.pre_cmd_timeout_multiplier, + timeout_multiplier = opts.pre_cmd_timeout_multiplier ) # If user used expect_broken then don't record failures of pre_cmd @@ -1809,6 +1894,8 @@ async def do_compile(name: TestName, result = await extras_build( way, extra_mods, extra_hc_opts ) if badResult(result): return result + + assert result.hc_opts is not None extra_hc_opts = result.hc_opts result = await simple_build(name, way, extra_hc_opts, should_fail, top_mod, units, should_link, True, **kwargs) @@ -1934,6 +2021,7 @@ async def compile_and_run__(name: TestName, result = await extras_build( way, extra_mods, extra_hc_opts ) if badResult(result): return result + assert result.hc_opts is not None extra_hc_opts = result.hc_opts assert extra_hc_opts is not None @@ -2027,10 +2115,15 @@ def report_stats(name, way, metric, gen_stat): metric_result = passed() perf_change = MetricChange.NewMetric else: + deviation = gen_stat["deviation"] + if deviation: + tolerance_metric = RelativeMetricAcceptanceWindow(deviation) + else: + tolerance_metric = AlwaysAccept() (perf_change, metric_result) = Perf.check_stats_change( perf_stat, baseline, - gen_stat["deviation"], + tolerance_metric, config.allowed_perf_changes, config.verbose >= 4) @@ -2051,9 +2144,14 @@ def report_stats(name, way, metric, gen_stat): # ----------------------------------------------------------------------------- # Build a single-module program -async def extras_build( way, extra_mods, extra_hc_opts ): +async def extras_build(way: WayName, extra_mods, extra_hc_opts) -> PassFail: for mod, opts in extra_mods: - result = await simple_build(mod, way, opts + ' ' + extra_hc_opts, False, None, [], False, False) + result = await simple_build(mod, way, opts + ' ' + extra_hc_opts, + should_fail=False, + top_mod=None, + units=[], + link=False, + addsuf=False) if not (mod.endswith('.hs') or mod.endswith('.lhs')): extra_hc_opts += ' %s' % Path(mod).with_suffix('.o') if badResult(result): @@ -2135,14 +2233,22 @@ async def simple_build(name: Union[TestName, str], flags = ' '.join(get_compiler_flags() + config.way_flags[way]) - cmd = ('cd "{opts.testdir}" && {cmd_prefix} ' + cmd = ('{cmd_prefix} ' '{{compiler}} {to_do} {srcname} {flags} {extra_hc_opts}' ).format(**locals()) if filter_with != '': cmd = cmd + ' | ' + filter_with - exit_code = await runCmd(cmd, None, stdout, stderr, opts.compile_timeout_multiplier) + output_file = perfStatsFile(True, name) + + exit_code = await runCmdPerf( + opts.compiler_perf_counters, + cmd, + output_file, + working_dir=opts.testdir, + stdin=None, stdout=stdout, stderr=stderr, + timeout_multiplier=opts.compile_timeout_multiplier) actual_stderr_path = in_testdir(name, 'comp.stderr') @@ -2162,7 +2268,6 @@ async def simple_build(name: Union[TestName, str], stderr_contents = actual_stderr_path.read_text(encoding='UTF-8', errors='replace') return failBecause('exit code non-0', stderr=stderr_contents) - return passed() # ----------------------------------------------------------------------------- @@ -2214,10 +2319,13 @@ async def simple_run(name: TestName, way: WayName, prog: str, extra_run_opts: st if opts.cmd_wrapper is not None: cmd = opts.cmd_wrapper(cmd) - cmd = 'cd "{opts.testdir}" && {cmd}'.format(**locals()) + output_file = perfStatsFile(False, name) # run the command - exit_code = await runCmd(cmd, stdin_arg, stdout_arg, stderr_arg, opts.run_timeout_multiplier) + exit_code = await runCmdPerf(opts.compiler_perf_counters, cmd, output_file, + working_dir=opts.testdir, + stdin=stdin_arg, stdout=stdout_arg, stderr=stderr_arg, + timeout_multiplier=opts.run_timeout_multiplier) # check the exit code if exit_code != opts.exit_code: @@ -2315,7 +2423,7 @@ async def interpreter_run(name: TestName, cmd = 'cd "{opts.testdir}" && {cmd}'.format(**locals()) - exit_code = await runCmd(cmd, script, stdout, stderr, opts.run_timeout_multiplier) + exit_code = await runCmd(cmd, script, stdout, stderr, timeout_multiplier=opts.run_timeout_multiplier) # split the stdout into compilation/program output split_file(stdout, delimiter, @@ -2973,12 +3081,56 @@ def dump_file(f: Path): except Exception: print('') +# ----------------------------------------------------------------------------- +# Run a program in the interpreter and check its output + +async def runCmdPerf( + perf_counters: List[str], + cmd: str, + output_file: str, + working_dir: Optional[Path] = None, + **kwargs) -> int: + """ + Run a command under `perf stat`, collecting the given counters. + + Returns the exit code and a dictionary of the collected counter values. + + If given a 'working_dir', we generate a command looking like: + + .. code-block:: text + + cd $working_dir && perf stat ... $cmd + + This allows users to find the test directory by looking at the execution logs, + and allows us to write 'perf' output files in the test directory. + """ + if len(perf_counters) == 0 or config.perf_path is None: + if working_dir: + cmd = f"cd \"{working_dir}\" && {cmd}" + + exit_code = await runCmd(cmd, **kwargs) + return exit_code + + perf_cmd_args: List[str] = [str(config.perf_path), 'stat', '-x\\;', '-o', output_file, '-e', ','.join(perf_counters), cmd] + cmd = ' '.join(perf_cmd_args) + if working_dir: + cmd = f"cd \"{working_dir}\" && {cmd}" + + exit_code = await runCmd(cmd, **kwargs) + return exit_code + async def runCmd(cmd: str, stdin: Union[None, Path]=None, stdout: Union[None, Path]=None, stderr: Union[None, int, Path]=None, timeout_multiplier=1.0, print_output=False) -> int: + """ + Run a command enforcing a timeout and returning the exit code. + + The process's working directory is changed to 'working_dir'. + """ + timeout_prog = strip_quotes(config.timeout_prog) timeout = str(int(ceil(config.timeout * timeout_multiplier))) @@ -3000,7 +3152,12 @@ async def runCmd(cmd: str, # Hence it must ultimately be run by a Bourne shell. It's timeout's job # to invoke the Bourne shell - proc = await asyncio.create_subprocess_exec(timeout_prog, timeout, cmd, stdin=stdin_file, stdout=asyncio.subprocess.PIPE, stderr=hStdErr, env=ghc_env) + proc = await asyncio.create_subprocess_exec(timeout_prog, timeout, cmd, + stdin=stdin_file, + stdout=asyncio.subprocess.PIPE, + stderr=hStdErr, + env=ghc_env + ) stdout_buffer, stderr_buffer = await proc.communicate() finally: ===================================== testsuite/driver/testutil.py ===================================== @@ -77,6 +77,11 @@ def lndir(srcdir: Path, dstdir: Path, force_copy=False): def testing_metrics(): return { 'bytes allocated', 'peak_megabytes_allocated', 'max_bytes_used' } +# All performance counters we consider to be stable enough in CI to +# test for. +def stable_perf_counters(): + return { 'instructions:u' } + # Metrics which are testing residency information def residency_testing_metrics(): return { 'peak_megabytes_allocated', 'max_bytes_used' } ===================================== testsuite/tests/perf/compiler/all.T ===================================== @@ -8,7 +8,7 @@ test('T1969', extra_run_opts('+RTS -A64k -RTS'), # The default RESIDENCY_OPTS is 256k and we need higher sampling # frequency. Incurs a slow-down by about 2. - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), only_ways(['normal']), extra_hc_opts('-dcore-lint -static'), @@ -32,14 +32,14 @@ else: test('T3294', [collect_compiler_residency(15), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), conf_3294, ], compile, ['']) test('T4801', - [collect_compiler_stats('bytes allocated',2), + [collect_compiler_runtime(2), only_ways(['normal']), extra_hc_opts('-static'), when(arch('wasm32') and unregisterised(), fragile(23290)) @@ -49,7 +49,7 @@ test('T4801', test('T3064', [collect_compiler_residency(20), - collect_compiler_stats('bytes allocated',2), + collect_compiler_runtime(2), only_ways(['normal']), ], compile, @@ -59,7 +59,7 @@ test('T3064', test('T4007', normal, makefile_test, ['T4007']) test('T5030', - [collect_compiler_stats('bytes allocated', 2), + [collect_compiler_runtime(2), only_ways(['normal']) ], @@ -67,14 +67,14 @@ test('T5030', ['-freduction-depth=300']) test('T5631', - [collect_compiler_stats('bytes allocated',2), + [collect_compiler_runtime(2), only_ways(['normal']) ], compile, ['']) test('parsing001', - [collect_compiler_stats('bytes allocated',2), + [collect_compiler_runtime(2), only_ways(['normal']), ], compile_fail, ['']) @@ -82,27 +82,27 @@ test('parsing001', test('T783', [ only_ways(['normal']), # no optimisation for this one - collect_compiler_stats('bytes allocated',2), + collect_compiler_runtime(2), extra_hc_opts('-static') ], compile,['']) test('T5321Fun', [ only_ways(['normal']), # no optimisation for this one - collect_compiler_stats('bytes allocated',2) + collect_compiler_runtime(2), ], compile,['']) test('T5321FD', [ only_ways(['normal']), # no optimisation for this one - collect_compiler_stats('bytes allocated',2) + collect_compiler_runtime(2), ], compile,['']) test('T5642', [ only_ways(['normal']), normal, - collect_compiler_stats('bytes allocated',2) + collect_compiler_runtime(2), ], compile,['-O']) @@ -114,7 +114,7 @@ test('T5837', test('T6048', [ only_ways(['optasm']), - collect_compiler_stats('bytes allocated',2) + collect_compiler_runtime(2), ], compile,['']) @@ -134,7 +134,7 @@ test('T9675', test('T9872a', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), high_memory_usage ], compile_fail, @@ -142,28 +142,28 @@ test('T9872a', test('T9872b', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), high_memory_usage ], compile_fail, ['']) test('T9872b_defer', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), high_memory_usage ], compile, ['-fdefer-type-errors']) test('T9872c', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), high_memory_usage ], compile_fail, ['']) test('T9872d', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], compile, ['']) @@ -227,14 +227,14 @@ test ('LargeRecord', test('T9961', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], compile, ['-O']) test('T9233', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], multimod_compile, ['T9233', '-v0 -O2 -fno-spec-constr']) @@ -249,14 +249,14 @@ test('T10370', test('T11068', normal, makefile_test, ['T11068']) test('T10547', - [ collect_compiler_stats('bytes allocated', 4), + [ collect_compiler_runtime(4), ], compile_fail, ['-fprint-expanded-synonyms']) test('T12227', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], compile, # Use `-M1G` to prevent memory thrashing with ghc-8.0.1. @@ -264,14 +264,14 @@ test('T12227', test('T12425', [ only_ways(['optasm']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], compile, ['']) test('T12234', [ only_ways(['optasm']), - collect_compiler_stats('bytes allocated', 2), + collect_compiler_runtime(2), ], compile, ['']) @@ -279,14 +279,14 @@ test('T12234', # See Note [Sensitivity to unique increment] in T12545.hs; spread was 4.8% test('T12545', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 10), # + collect_compiler_runtime(10), # ], multimod_compile, ['T12545', '-v0'] ) test('T13035', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), ], compile, [''] ) @@ -299,7 +299,7 @@ test('T13056', ['-O1']) test('T12707', - [ collect_compiler_stats('bytes allocated', 1), + [ collect_compiler_runtime(1), ], compile, ['']) @@ -311,7 +311,7 @@ test('T12707', # to avoid spurious errors. test('T12150', [ only_ways(['optasm']), - collect_compiler_stats('bytes allocated', 2) + collect_compiler_runtime(2) ], compile, ['']) @@ -483,7 +483,7 @@ test('MultiLayerModulesNoCode', ['MultiLayerModulesNoCode.script']) test('MultiComponentModulesRecomp', - [ collect_compiler_stats('bytes allocated', 2), + [ collect_compiler_runtime(2), pre_cmd('$MAKE -s --no-print-directory MultiComponentModulesRecomp'), extra_files(['genMultiComp.py']), compile_timeout_multiplier(5) @@ -492,7 +492,7 @@ test('MultiComponentModulesRecomp', [['unitp%d' % n for n in range(20)], '-fno-code -fwrite-interface -v0']) test('MultiComponentModules', - [ collect_compiler_stats('bytes allocated', 2), + [ collect_compiler_runtime(2), pre_cmd('$PYTHON ./genMultiComp.py'), extra_files(['genMultiComp.py']), compile_timeout_multiplier(5) @@ -565,7 +565,7 @@ test('T14683', test ('T9630', [ collect_compiler_residency(15), - collect_compiler_stats('bytes allocated', 2), + collect_compiler_runtime(2), ], multimod_compile, ['T9630', '-v0 -O']) @@ -611,7 +611,7 @@ test ('T16473', ['-O2 -flate-specialise']) test('T17516', - [ collect_compiler_stats('bytes allocated', 5), + [ collect_compiler_runtime(5), ], multimod_compile, ['T17516', '-O -v0']) @@ -635,13 +635,13 @@ test ('T18140', ['-v0 -O']) test('T10421', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 1) + collect_compiler_runtime(1) ], multimod_compile, ['T10421', '-v0 -O']) test('T10421a', [ only_ways(['normal']), - collect_compiler_stats('bytes allocated', 10) + collect_compiler_runtime(10) ], multimod_compile, ['T10421a', '-v0 -O']) @@ -700,13 +700,13 @@ test ('T19695', ['-v0 -O2']) test('hard_hole_fits', # Testing multiple hole-fits with lots in scope for #16875 - collect_compiler_stats('bytes allocated', 2), # 1 is 300s, 0.010 is 3s. Without hole-fits it takes 1s + collect_compiler_runtime(2), # 1 is 300s, 0.010 is 3s. Without hole-fits it takes 1s compile, ['-fdefer-type-errors -fno-max-valid-hole-fits -package ghc']) test('T16875', # Testing one hole-fit with a lot in scope for #16875 # This test is very sensitive to environmental differences.. we should fix # that but for now the failure threshold is 4% (see #21557) - collect_compiler_stats('bytes allocated', 4), + collect_compiler_runtime(4), compile, ['-fdefer-type-errors -fno-max-valid-hole-fits -package ghc']) test ('T20261', @@ -720,7 +720,7 @@ test ('T20261', # a compile-time and a run-time performance test test('T21839c', [ collect_compiler_stats('all', 10), - collect_compiler_stats('bytes allocated', 1), + collect_compiler_runtime(1), only_ways(['normal'])], compile, ['-O']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fb0a4e5cc545d8d995b6631138f40495917e795a...6dfb9471822660c6a90a018c4615407e4a5469ba -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fb0a4e5cc545d8d995b6631138f40495917e795a...6dfb9471822660c6a90a018c4615407e4a5469ba You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 00:23:12 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 04 Sep 2024 20:23:12 -0400 Subject: [Git][ghc/ghc][master] RecordCon lookup: don't allow a TyCon Message-ID: <66d8f9f0a4a8f_1250935c553021143@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 14 changed files: - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Utils/Env.hs - + testsuite/tests/rename/should_fail/T25056.hs - + testsuite/tests/rename/should_fail/T25056.stderr - + testsuite/tests/rename/should_fail/T25056a.hs - + testsuite/tests/rename/should_fail/T25056b.hs - testsuite/tests/rename/should_fail/all.T - testsuite/tests/typecheck/should_fail/T23739b.hs - testsuite/tests/typecheck/should_fail/T23739b.stderr - + testsuite/tests/typecheck/should_fail/T23739c.hs - + testsuite/tests/typecheck/should_fail/T23739c.stderr - testsuite/tests/typecheck/should_fail/all.T Changes: ===================================== compiler/GHC/Rename/Env.hs ===================================== @@ -442,6 +442,7 @@ lookupConstructorInfo con_name ; case info of IAmConLike con_info -> return con_info UnboundGRE -> return $ ConInfo (ConIsData []) ConHasPositionalArgs + IAmTyCon {} -> failIllegalTyCon WL_Constructor con_name _ -> pprPanic "lookupConstructorInfo: not a ConLike" $ vcat [ text "name:" <+> ppr con_name ] } @@ -1035,24 +1036,12 @@ lookupOccRn' which_suggest rdr_name lookupOccRn :: RdrName -> RnM Name lookupOccRn = lookupOccRn' WL_Anything --- lookupOccRnConstr looks up an occurrence of a RdrName and displays --- constructors and pattern synonyms as suggestions if it is not in scope +-- | Look up an occurrence of a 'RdrName'. -- --- There is a fallback to the type level, when the first lookup fails. --- This is required to implement a pat-to-type transformation --- (See Note [Pattern to type (P2T) conversion] in GHC.Tc.Gen.Pat) --- Consider this example: +-- Displays constructors and pattern synonyms as suggestions if +-- it is not in scope. -- --- data VisProxy a where VP :: forall a -> VisProxy a --- --- f :: VisProxy Int -> () --- f (VP Int) = () --- --- Here `Int` is actually a type, but it stays on position where --- we expect a data constructor. --- --- In all other cases we just use this additional lookup for better --- error messaging (See Note [Promotion]). +-- See Note [lookupOccRnConstr] lookupOccRnConstr :: RdrName -> RnM Name lookupOccRnConstr rdr_name = do { mb_gre <- lookupOccRn_maybe rdr_name @@ -1064,6 +1053,28 @@ lookupOccRnConstr rdr_name Just gre -> return $ greName gre Nothing -> reportUnboundName' WL_Constructor rdr_name} } +{- Note [lookupOccRnConstr] +~~~~~~~~~~~~~~~~~~~~~~~~~~~ +lookupOccRnConstr looks up a data constructor or pattern synonym. Simple. + +However, there is a fallback to the type level when the lookup fails. +This is required to implement a pat-to-type transformation +(See Note [Pattern to type (P2T) conversion] in GHC.Tc.Gen.Pat) + +Consider this example: + + data VisProxy a where VP :: forall a -> VisProxy a + + f :: VisProxy Int -> () + f (VP Int) = () + +Here `Int` is actually a type, but it occurs in a position in which we expect +a data constructor. + +In all other cases we just use this additional lookup for better +error messaging (See Note [Promotion]). +-} + -- lookupOccRnRecField looks up an occurrence of a RdrName and displays -- record fields as suggestions if it is not in scope lookupOccRnRecField :: RdrName -> RnM Name ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -539,9 +539,9 @@ rnExpr (ExplicitSum _ alt arity expr) = do { (expr', fvs) <- rnLExpr expr ; return (ExplicitSum noExtField alt arity expr', fvs) } -rnExpr (RecordCon { rcon_con = con_id +rnExpr (RecordCon { rcon_con = con_rdr , rcon_flds = rec_binds@(HsRecFields { rec_dotdot = dd }) }) - = do { con_lname@(L _ con_name) <- lookupLocatedOccRnConstr con_id + = do { con_lname@(L _ con_name) <- lookupLocatedOccRnConstr con_rdr ; (flds, fvs) <- rnHsRecFields (HsRecFieldCon con_name) mk_hs_var rec_binds ; (flds', fvss) <- mapAndUnzipM rn_field flds ; let rec_binds' = HsRecFields { rec_ext = noExtField, rec_flds = flds', rec_dotdot = dd } ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -841,7 +841,7 @@ tc_infer_id id_name AGlobal (AConLike (RealDataCon con)) -> tcInferDataCon con AGlobal (AConLike (PatSynCon ps)) -> tcInferPatSyn id_name ps - (tcTyThingTyCon_maybe -> Just tc) -> failIllegalTyCon WL_Anything tc -- TyCon or TcTyCon + (tcTyThingTyCon_maybe -> Just tc) -> failIllegalTyCon WL_Anything (tyConName tc) ATyVar name _ -> failIllegalTyVal name _ -> failWithTc $ TcRnExpectedValueId thing } ===================================== compiler/GHC/Tc/Utils/Env.hs ===================================== @@ -280,7 +280,7 @@ tcLookupConLike name = do thing <- tcLookupGlobal name case thing of AConLike cl -> return cl - ATyCon tc -> failIllegalTyCon WL_Constructor tc + ATyCon {} -> failIllegalTyCon WL_Constructor name _ -> wrongThingErr WrongThingConLike (AGlobal thing) name tcLookupRecSelParent :: HsRecUpdParent GhcRn -> TcM RecSelParent @@ -353,19 +353,20 @@ instance MonadThings (IOEnv (Env TcGblEnv TcLclEnv)) where lookupThing = tcLookupGlobal -- Illegal term-level use of type things -failIllegalTyCon :: WhatLooking -> TyCon -> TcM a +failIllegalTyCon :: WhatLooking -> Name -> TcM a failIllegalTyVal :: Name -> TcM a (failIllegalTyCon, failIllegalTyVal) = (fail_tycon, fail_tyvar) where - fail_tycon what_looking tc = do + fail_tycon what_looking tc_nm = do gre <- getGlobalRdrEnv - let nm = tyConName tc - pprov = case lookupGRE_Name gre nm of + let mb_gre = lookupGRE_Name gre tc_nm + pprov = case mb_gre of Just gre -> nest 2 (pprNameProvenance gre) Nothing -> empty - err | isClassTyCon tc = ClassTE - | otherwise = TyConTE - fail_with_msg what_looking dataName nm pprov err + err = case greInfo <$> mb_gre of + Just (IAmTyCon ClassFlavour) -> ClassTE + _ -> TyConTE + fail_with_msg what_looking dataName tc_nm pprov err fail_tyvar nm = let pprov = nest 2 (text "bound at" <+> ppr (getSrcLoc nm)) ===================================== testsuite/tests/rename/should_fail/T25056.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE RecordWildCards #-} +module T25056 where + +import T25056b + +foo :: T -> () +foo (T { unT = x }) = x ===================================== testsuite/tests/rename/should_fail/T25056.stderr ===================================== @@ -0,0 +1,5 @@ +T25056.hs:7:10: error: [GHC-01928] + • Illegal term-level use of the type constructor ‘T’ + • imported from ‘T25056b’ at T25056.hs:4:1-14 + (and originally defined in ‘T25056a’ at T25056a.hs:8:1-14) + ===================================== testsuite/tests/rename/should_fail/T25056a.hs ===================================== @@ -0,0 +1,13 @@ +{-# LANGUAGE PatternSynonyms #-} +module T25056a + ( T + , T_(unT) + , pattern T + ) where + +type T = T_ () + +data T_ a = PrivateT { unT_ :: a } + +pattern T :: a -> T_ a +pattern T { unT } <- PrivateT { unT_ = unT } ===================================== testsuite/tests/rename/should_fail/T25056b.hs ===================================== @@ -0,0 +1,3 @@ +module T25056b (T, T_(..)) where + +import T25056a (T, T_(..)) ===================================== testsuite/tests/rename/should_fail/all.T ===================================== @@ -222,6 +222,7 @@ test('T23740g', normal, compile_fail, ['']) test('T23740h', normal, compile_fail, ['']) test('T23740i', req_th, compile_fail, ['']) test('T23740j', normal, compile_fail, ['']) +test('T25056', [extra_files(['T25056a.hs', 'T25056b.hs'])], multimod_compile_fail, ['T25056', '-v0']) test('Or3', normal, compile_fail, ['']) test('T23570', [extra_files(['T23570_aux.hs'])], multimod_compile_fail, ['T23570', '-v0']) test('T23570b', [extra_files(['T23570_aux.hs'])], multimod_compile, ['T23570b', '-v0']) ===================================== testsuite/tests/typecheck/should_fail/T23739b.hs ===================================== @@ -8,7 +8,4 @@ g1 :: Int -> Unit g1 Int = () g2 :: Int -g2 = Int{} - -g3 :: Int -g3 = Int +g2 = Int ===================================== testsuite/tests/typecheck/should_fail/T23739b.stderr ===================================== @@ -6,16 +6,9 @@ T23739b.hs:8:4: error: [GHC-01928] In an equation for ‘g1’: g1 Int = () T23739b.hs:11:6: error: [GHC-01928] - • Illegal term-level use of the type constructor ‘Int’ - • imported from ‘Prelude’ at T23739b.hs:2:8-14 - (and originally defined in ‘GHC.Types’) - • In the expression: Int {} - In an equation for ‘g2’: g2 = Int {} - -T23739b.hs:14:6: error: [GHC-01928] • Illegal term-level use of the type constructor ‘Int’ • imported from ‘Prelude’ at T23739b.hs:2:8-14 (and originally defined in ‘GHC.Types’) • In the expression: Int - In an equation for ‘g3’: g3 = Int + In an equation for ‘g2’: g2 = Int ===================================== testsuite/tests/typecheck/should_fail/T23739c.hs ===================================== @@ -0,0 +1,8 @@ + +module T23739c where + +import Data.Tuple.Experimental +import GHC.TypeLits + +g :: Int +g = Int{} ===================================== testsuite/tests/typecheck/should_fail/T23739c.stderr ===================================== @@ -0,0 +1,7 @@ +T23739c.hs:8:5: error: [GHC-01928] + • Illegal term-level use of the type constructor ‘Int’ + • imported from ‘Prelude’ at T23739c.hs:2:8-14 + (and originally defined in ‘GHC.Types’) + • In the expression: Int {} + In an equation for ‘g’: g = Int {} + ===================================== testsuite/tests/typecheck/should_fail/all.T ===================================== @@ -727,5 +727,6 @@ test('T17594g', normal, compile_fail, ['']) test('T24470a', normal, compile_fail, ['']) test('T24553', normal, compile_fail, ['']) test('T23739b', normal, compile_fail, ['']) +test('T23739c', normal, compile_fail, ['']) test('T24868', normal, compile_fail, ['']) test('T24938', normal, compile_fail, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/da306610b9e58cfb7cf2530ebeec7ee8ad17183a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/da306610b9e58cfb7cf2530ebeec7ee8ad17183a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 00:23:44 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 04 Sep 2024 20:23:44 -0400 Subject: [Git][ghc/ghc][master] Use deterministic names for temporary files Message-ID: <66d8fa10e3a36_1250935e60c8257d0@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 3 changed files: - compiler/GHC/Driver/Make.hs - compiler/GHC/Utils/TmpFs.hs - testsuite/tests/ghc-e/should_fail/T9930fail.stderr Changes: ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2924,19 +2924,22 @@ runParPipelines worker_limit plugin_hsc_env diag_wrapper mHscMessager all_pipeli atomically $ writeTVar stopped_var True wait_log_thread -withLocalTmpFS :: RunMakeM a -> RunMakeM a -withLocalTmpFS act = do +withLocalTmpFS :: TmpFs -> (TmpFs -> IO a) -> IO a +withLocalTmpFS tmpfs act = do let initialiser = do - MakeEnv{..} <- ask - lcl_tmpfs <- liftIO $ forkTmpFsFrom (hsc_tmpfs hsc_env) - return $ hsc_env { hsc_tmpfs = lcl_tmpfs } - finaliser lcl_env = do - gbl_env <- ask - liftIO $ mergeTmpFsInto (hsc_tmpfs lcl_env) (hsc_tmpfs (hsc_env gbl_env)) + liftIO $ forkTmpFsFrom tmpfs + finaliser tmpfs_local = do + liftIO $ mergeTmpFsInto tmpfs_local tmpfs -- Add remaining files which weren't cleaned up into local tmp fs for -- clean-up later. -- Clear the logQueue if this node had it's own log queue - MC.bracket initialiser finaliser $ \lcl_hsc_env -> local (\env -> env { hsc_env = lcl_hsc_env}) act + MC.bracket initialiser finaliser act + +withLocalTmpFSMake :: MakeEnv -> (MakeEnv -> IO a) -> IO a +withLocalTmpFSMake env k = + withLocalTmpFS (hsc_tmpfs (hsc_env env)) $ \lcl_tmpfs + -> k (env { hsc_env = (hsc_env env) { hsc_tmpfs = lcl_tmpfs }}) + -- | Run the given actions and then wait for them all to finish. runAllPipelines :: WorkerLimit -> MakeEnv -> [MakeAction] -> IO () @@ -2958,16 +2961,18 @@ runAllPipelines worker_limit env acts = do runLoop :: (((forall a. IO a -> IO a) -> IO ()) -> IO a) -> MakeEnv -> [MakeAction] -> IO [a] runLoop _ _env [] = return [] runLoop fork_thread env (MakeAction act res_var :acts) = do - new_thread <- + + -- withLocalTmpFs has to occur outside of fork to remain deterministic + new_thread <- withLocalTmpFSMake env $ \lcl_env -> fork_thread $ \unmask -> (do - mres <- (unmask $ run_pipeline (withLocalTmpFS act)) + mres <- (unmask $ run_pipeline lcl_env act) `MC.onException` (putMVar res_var Nothing) -- Defensive: If there's an unhandled exception then still signal the failure. putMVar res_var mres) threads <- runLoop fork_thread env acts return (new_thread : threads) where - run_pipeline :: RunMakeM a -> IO (Maybe a) - run_pipeline p = runMaybeT (runReaderT p env) + run_pipeline :: MakeEnv -> RunMakeM a -> IO (Maybe a) + run_pipeline env p = runMaybeT (runReaderT p env) data MakeAction = forall a . MakeAction !(RunMakeM a) !(MVar (Maybe a)) ===================================== compiler/GHC/Utils/TmpFs.hs ===================================== @@ -64,6 +64,8 @@ data TmpFs = TmpFs -- -- Shared with forked TmpFs. + , tmp_dir_prefix :: String + , tmp_files_to_clean :: IORef PathsToClean -- ^ Files to clean (per session or per module) -- @@ -121,6 +123,7 @@ initTmpFs = do , tmp_subdirs_to_clean = subdirs , tmp_dirs_to_clean = dirs , tmp_next_suffix = next + , tmp_dir_prefix = "tmp" } -- | Initialise an empty TmpFs sharing unique numbers and per-process temporary @@ -132,11 +135,16 @@ forkTmpFsFrom :: TmpFs -> IO TmpFs forkTmpFsFrom old = do files <- newIORef emptyPathsToClean subdirs <- newIORef emptyPathsToClean + counter <- newIORef 0 + prefix <- newTempSuffix old + + return $ TmpFs { tmp_files_to_clean = files , tmp_subdirs_to_clean = subdirs , tmp_dirs_to_clean = tmp_dirs_to_clean old - , tmp_next_suffix = tmp_next_suffix old + , tmp_next_suffix = counter + , tmp_dir_prefix = prefix } -- | Merge the first TmpFs into the second. @@ -259,9 +267,11 @@ changeTempFilesLifetime tmpfs lifetime files = do addFilesToClean tmpfs lifetime existing_files -- Return a unique numeric temp file suffix -newTempSuffix :: TmpFs -> IO Int -newTempSuffix tmpfs = - atomicModifyIORef' (tmp_next_suffix tmpfs) $ \n -> (n+1,n) +newTempSuffix :: TmpFs -> IO String +newTempSuffix tmpfs = do + n <- atomicModifyIORef' (tmp_next_suffix tmpfs) $ \n -> (n+1,n) + return $ tmp_dir_prefix tmpfs ++ "_" ++ show n + -- Find a temporary name that doesn't already exist. newTempName :: Logger -> TmpFs -> TempDir -> TempFileLifetime -> Suffix -> IO FilePath @@ -271,8 +281,8 @@ newTempName logger tmpfs tmp_dir lifetime extn where findTempName :: FilePath -> IO FilePath findTempName prefix - = do n <- newTempSuffix tmpfs - let filename = prefix ++ show n <.> extn + = do suffix <- newTempSuffix tmpfs + let filename = prefix ++ suffix <.> extn b <- doesFileExist filename if b then findTempName prefix else do -- clean it up later @@ -295,8 +305,8 @@ newTempSubDir logger tmpfs tmp_dir where findTempDir :: FilePath -> IO FilePath findTempDir prefix - = do n <- newTempSuffix tmpfs - let name = prefix ++ show n + = do suffix <- newTempSuffix tmpfs + let name = prefix ++ suffix b <- doesDirectoryExist name if b then findTempDir prefix else (do @@ -314,8 +324,8 @@ newTempLibName logger tmpfs tmp_dir lifetime extn where findTempName :: FilePath -> String -> IO (FilePath, FilePath, String) findTempName dir prefix - = do n <- newTempSuffix tmpfs -- See Note [Deterministic base name] - let libname = prefix ++ show n + = do suffix <- newTempSuffix tmpfs -- See Note [Deterministic base name] + let libname = prefix ++ suffix filename = dir "lib" ++ libname <.> extn b <- doesFileExist filename if b then findTempName dir prefix @@ -340,8 +350,8 @@ getTempDir logger tmpfs (TempDir tmp_dir) = do mkTempDir :: FilePath -> IO FilePath mkTempDir prefix = do - n <- newTempSuffix tmpfs - let our_dir = prefix ++ show n + suffix <- newTempSuffix tmpfs + let our_dir = prefix ++ suffix -- 1. Speculatively create our new directory. createDirectory our_dir @@ -376,6 +386,11 @@ the temporary file no longer contains random information (it used to contain the process id). This is ok, as the temporary directory used contains the pid (see getTempDir). + +In addition to this, multiple threads can race against each other creating temporary +files. Therefore we supply a prefix when creating temporary files, when a thread is +forked, each thread must be given an TmpFs with a unique prefix. This is achieved +by forkTmpFsFrom creating a fresh prefix from the parent TmpFs. -} manyWithTrace :: Logger -> String -> ([FilePath] -> IO ()) -> [FilePath] -> IO () ===================================== testsuite/tests/ghc-e/should_fail/T9930fail.stderr ===================================== @@ -1,18 +1,12 @@ -ghc: Exception: +ghc-9.11.20240830: Exception: default output name would overwrite the input file; must specify -o explicitly Usage: For basic information, try the `--help' option. -Package: ghc-inplace +Package: ghc-9.11-inplace Module: GHC.Utils.Panic Type: GhcException HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO - throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:860:84 in exceptions-0.10.7-inplace:Control.Monad.Catch - onException, called at compiler/GHC/Driver/Make.hs:2974:23 in ghc-9.9-inplace:GHC.Driver.Make - - + bracket, called at compiler/GHC/Driver/Make.hs:2936:3 in ghc-9.11-inplace:GHC.Driver.Make View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9c354beba3e03f56f9a6345f7607a04b55a3318f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9c354beba3e03f56f9a6345f7607a04b55a3318f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 05:26:43 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 05 Sep 2024 01:26:43 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 9 commits: testsuite: Add support to capture performance metrics via 'perf' Message-ID: <66d941132cf84_12509315327fc3564c@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 5d63345d by Hécate Kleidukos at 2024-09-05T01:26:38-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - d350293d by Hécate Kleidukos at 2024-09-05T01:26:38-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - e7b3dae9 by Matthew Pickering at 2024-09-05T01:26:38-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 4ccc25bb by Matthew Pickering at 2024-09-05T01:26:38-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/Driver/Make.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Utils/Env.hs - compiler/GHC/Utils/TmpFs.hs - hadrian/ghci-multi-cabal.in - hadrian/src/Rules/ToolArgs.hs - testsuite/driver/perf_notes.py - testsuite/driver/testglobals.py - testsuite/driver/testlib.py - testsuite/driver/testutil.py - testsuite/tests/ghc-e/should_fail/T9930fail.stderr - testsuite/tests/perf/compiler/all.T - + testsuite/tests/rename/should_fail/T25056.hs - + testsuite/tests/rename/should_fail/T25056.stderr - + testsuite/tests/rename/should_fail/T25056a.hs - + testsuite/tests/rename/should_fail/T25056b.hs - testsuite/tests/rename/should_fail/all.T - testsuite/tests/typecheck/should_fail/T23739b.hs - testsuite/tests/typecheck/should_fail/T23739b.stderr - + testsuite/tests/typecheck/should_fail/T23739c.hs - + testsuite/tests/typecheck/should_fail/T23739c.stderr - testsuite/tests/typecheck/should_fail/all.T - utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b9c46af5f341a67b1318f3a44c43f29d5478d510...4ccc25bbc3d6f35d22e7f327a1aab9f0456fcbdf -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b9c46af5f341a67b1318f3a44c43f29d5478d510...4ccc25bbc3d6f35d22e7f327a1aab9f0456fcbdf You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 07:21:58 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 05 Sep 2024 03:21:58 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/supersven/fix-new-c-warnings Message-ID: <66d95c161e5b5_20a9512dc864715ed@gitlab.mail> Sven Tennie pushed new branch wip/supersven/fix-new-c-warnings at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/supersven/fix-new-c-warnings You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 07:31:54 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 05 Sep 2024 03:31:54 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] 2 commits: Fix regression in RISCV64 RTS linker Message-ID: <66d95e6ac3de_20a951438564734a1@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: f208f182 by Sven Tennie at 2024-09-05T07:25:26+00:00 Fix regression in RISCV64 RTS linker The check was non-sense. - - - - - 22fc1a09 by Sven Tennie at 2024-09-05T07:31:36+00:00 Adjust test timings for small machines The reference was a Lichee Pi 4a RISCV64 machine. - - - - - 3 changed files: - rts/linker/elf_reloc_riscv64.c - testsuite/tests/concurrent/should_run/async001.hs - testsuite/tests/concurrent/should_run/conc024.hs Changes: ===================================== rts/linker/elf_reloc_riscv64.c ===================================== @@ -118,7 +118,7 @@ int32_t decodeAddendRISCV64(Section *section STG_UNUSED, // Make sure that V can be represented as an N bit signed integer. void checkInt(inst_t *loc, int32_t v, int n) { - if (v != signExtend32(v, n)) { + if (!isInt(n, v)) { barf("Relocation at 0x%x is out of range. value: 0x%x (%d), " "sign-extended value: 0x%x (%d), max bits 0x%x (%d)\n", *loc, v, v, signExtend32(v, n), signExtend32(v, n), n, n); ===================================== testsuite/tests/concurrent/should_run/async001.hs ===================================== @@ -9,11 +9,11 @@ import System.IO.Unsafe main = do let x = unsafePerformIO $ - (do threadDelay 1000000; return 42) + (do threadDelay 3000000; return 42) `onException` return () t <- forkIO $ do evaluate x; return () - threadDelay 2000 + threadDelay 6000 killThread t print x `E.catch` \e -> putStrLn ("main caught: " ++ show (e::SomeException)) ===================================== testsuite/tests/concurrent/should_run/conc024.hs ===================================== @@ -11,5 +11,5 @@ main = do id <- myThreadId forkIO (catch (do m <- newEmptyMVar; takeMVar m) (\e -> throwTo id (e::SomeException))) - catch (do yield; performGC; threadDelay 1000000) + catch (do yield; performGC; threadDelay 2000000) (\e -> print (e::SomeException)) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/26c79cfa3cee3cf0651f91b7b75d00268e985e91...22fc1a09dda3e9373f83c16a05d4a72636ae65e4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/26c79cfa3cee3cf0651f91b7b75d00268e985e91...22fc1a09dda3e9373f83c16a05d4a72636ae65e4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 07:42:54 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 05 Sep 2024 03:42:54 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] 38 commits: JS: add basic support for POSIX *at functions (#25190) Message-ID: <66d960febe7b5_319fba8aaac1668c@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 3a767377 by Sven Tennie at 2024-09-05T09:41:21+02:00 Add RISCV64 Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 0371de6d by Sven Tennie at 2024-09-05T09:41:21+02:00 async001: Adjust for slower computers Increase the delay a bit to be able to run this test on slower computers (e.g. RISCV64 LicheePi 4a.) - - - - - 8d7f8882 by Sven Tennie at 2024-09-05T09:41:21+02:00 Add RTS linker for RISCV64 This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - c82ab120 by Sven Tennie at 2024-09-05T09:41:21+02:00 Ignore divbyzero test for RISCV64 The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - 1d6e0f7c by Sven Tennie at 2024-09-05T09:41:21+02:00 Enable MulMayOflo_full test for RISCV64 It works and thus can be tested. - - - - - 0b2ba0da by Sven Tennie at 2024-09-05T09:41:21+02:00 LibffiAdjustor: Ensure code caches are flushed (RISCV64) RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - c7c15820 by Sven Tennie at 2024-09-05T09:41:21+02:00 Add additional linker symbols for builtins (RISCV64) We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 41237342 by Sven Tennie at 2024-09-05T09:41:21+02:00 Add GHCi support for RISCV64 As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - c9f6a91a by Sven Tennie at 2024-09-05T09:41:21+02:00 Set codeowners of the RISCV64 NCG - - - - - 1769afef by Sven Tennie at 2024-09-05T09:41:21+02:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. - - - - - fa0313d7 by Sven Tennie at 2024-09-05T09:41:21+02:00 WIP: FIXUP for master Will be squashed later. For now it's convenient to have a separate commit. - - - - - 6a790662 by Sven Tennie at 2024-09-05T09:41:21+02:00 RISCV64 has rtsLinker Likely a mistake during merge. - - - - - 6fdedddc by Sven Tennie at 2024-09-05T09:41:21+02:00 Make switch/jump tables PIC compatible - - - - - 0680e603 by Sven Tennie at 2024-09-05T09:41:21+02:00 Improve documentation for stack spills - - - - - f0c7c300 by Sven Tennie at 2024-09-05T09:41:22+02:00 Reformat stack spill code - - - - - ac90700a by Sven Tennie at 2024-09-05T09:41:22+02:00 Make instruction types strict in their fields This should prevent memory leaks. - - - - - c3beff6c by Sven Tennie at 2024-09-05T09:41:22+02:00 Use FCVT constructor for all float-related conversions This is closer to the assembly instruction and should reduce confusion. - - - - - 230bb190 by Sven Tennie at 2024-09-05T09:41:22+02:00 Fix regression in RISCV64 RTS linker The check was non-sense. - - - - - 98685673 by Sven Tennie at 2024-09-05T09:41:22+02:00 Adjust test timings for small machines The reference was a Lichee Pi 4a RISCV64 machine. - - - - - 18 changed files: - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/22fc1a09dda3e9373f83c16a05d4a72636ae65e4...98685673cd9cd4a8bc438187f05cd64429f23ab3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/22fc1a09dda3e9373f83c16a05d4a72636ae65e4...98685673cd9cd4a8bc438187f05cd64429f23ab3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 09:22:44 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 05 Sep 2024 05:22:44 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 4 commits: DCmmGroup vs CmmGroup or: Deterministic Info Tables Message-ID: <66d9786431413_3bd3eed4aa8977b5@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: b756c6b7 by Rodrigo Mesquita at 2024-09-05T10:22:10+01:00 DCmmGroup vs CmmGroup or: Deterministic Info Tables See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and Note [Object determinism] cmm: Back LabelMap with UDFM Use a deterministic unique map to back the implementation of `LabelMap`. This is necessary towards the goal of object code determinism in #12935. Our intended solution requires renaming uniques in a deterministic order (which will be the order in which they were created), but storing them label map makes us lose this order. Backing it with a UDFM fixes this issue. Introduce back LabelMap non deterministic Use NonDeterministic Label map in multiple passes (TODO: More could be available. Look through Det LabelMap uses again) Use NonDet for CFG More NonDet More explicit Introduce DCmmDecl, start Removing more maps - - - - - 776543ea by Rodrigo Mesquita at 2024-09-05T10:22:22+01:00 Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled - - - - - 879263c1 by Rodrigo Mesquita at 2024-09-05T10:22:22+01:00 distinct-constructor-tables determinism - - - - - f446aa3c by Rodrigo Mesquita at 2024-09-05T10:22:22+01:00 Rename deterministically CmmGroups in generateCgIPEStub - - - - - 20 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ThreadSanitizer.hs - compiler/GHC/CmmToAsm/Reg/Liveness.hs - compiler/GHC/Driver/GenerateCgIPEStub.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Stg/Debug.hs - compiler/GHC/StgToCmm.hs - compiler/GHC/StgToCmm/CgUtils.hs - compiler/GHC/StgToCmm/Monad.hs - compiler/GHC/StgToCmm/Utils.hs - compiler/GHC/Types/IPE.hs - compiler/GHC/Types/Name.hs - compiler/GHC/Types/Unique/DFM.hs - compiler/GHC/Utils/Outputable.hs - testsuite/tests/regalloc/regalloc_unit_tests.hs - testsuite/tests/wasm/should_run/control-flow/LoadCmmGroup.hs Changes: ===================================== compiler/GHC/Cmm.hs ===================================== @@ -12,22 +12,28 @@ module GHC.Cmm ( -- * Cmm top-level datatypes + DCmmGroup, CmmProgram, CmmGroup, CmmGroupSRTs, RawCmmGroup, GenCmmGroup, - CmmDecl, CmmDeclSRTs, GenCmmDecl(..), - CmmDataDecl, cmmDataDeclCmmDecl, - CmmGraph, GenCmmGraph(..), + CmmDecl, DCmmDecl, CmmDeclSRTs, GenCmmDecl(..), + CmmDataDecl, cmmDataDeclCmmDecl, DCmmGraph, + CmmGraph, GenCmmGraph, GenGenCmmGraph(..), toBlockMap, revPostorder, toBlockList, CmmBlock, RawCmmDecl, Section(..), SectionType(..), GenCmmStatics(..), type CmmStatics, type RawCmmStatics, CmmStatic(..), SectionProtection(..), sectionProtection, + DWrap(..), unDeterm, removeDeterm, removeDetermDecl, removeDetermGraph, + -- ** Blocks containing lists GenBasicBlock(..), blockId, ListGraph(..), pprBBlock, -- * Info Tables - CmmTopInfo(..), CmmStackInfo(..), CmmInfoTable(..), topInfoTable, + GenCmmTopInfo(..) + , DCmmTopInfo + , CmmTopInfo + , CmmStackInfo(..), CmmInfoTable(..), topInfoTable, topInfoTableD, ClosureTypeInfo(..), ProfilingInfo(..), ConstrDescription, @@ -74,6 +80,8 @@ import qualified Data.ByteString as BS type CmmProgram = [CmmGroup] type GenCmmGroup d h g = [GenCmmDecl d h g] +-- | Cmm group after STG generation +type DCmmGroup = GenCmmGroup CmmStatics DCmmTopInfo DCmmGraph -- | Cmm group before SRT generation type CmmGroup = GenCmmGroup CmmStatics CmmTopInfo CmmGraph -- | Cmm group with SRTs @@ -117,6 +125,7 @@ instance (OutputableP Platform d, OutputableP Platform info, OutputableP Platfor => OutputableP Platform (GenCmmDecl d info i) where pdoc = pprTop +type DCmmDecl = GenCmmDecl CmmStatics DCmmTopInfo DCmmGraph type CmmDecl = GenCmmDecl CmmStatics CmmTopInfo CmmGraph type CmmDeclSRTs = GenCmmDecl RawCmmStatics CmmTopInfo CmmGraph type CmmDataDecl = GenCmmDataDecl CmmStatics @@ -139,7 +148,11 @@ type RawCmmDecl ----------------------------------------------------------------------------- type CmmGraph = GenCmmGraph CmmNode -data GenCmmGraph n = CmmGraph { g_entry :: BlockId, g_graph :: Graph n C C } +type DCmmGraph = GenGenCmmGraph DWrap CmmNode + +type GenCmmGraph n = GenGenCmmGraph LabelMap n + +data GenGenCmmGraph s n = CmmGraph { g_entry :: BlockId, g_graph :: Graph' s Block n C C } type CmmBlock = Block CmmNode C C instance OutputableP Platform CmmGraph where @@ -171,8 +184,16 @@ toBlockList g = mapElems $ toBlockMap g -- | CmmTopInfo is attached to each CmmDecl (see defn of CmmGroup), and contains -- the extra info (beyond the executable code) that belongs to that CmmDecl. -data CmmTopInfo = TopInfo { info_tbls :: LabelMap CmmInfoTable - , stack_info :: CmmStackInfo } +data GenCmmTopInfo f = TopInfo { info_tbls :: f CmmInfoTable + , stack_info :: CmmStackInfo } + +newtype DWrap a = DWrap [(BlockId, a)] + +unDeterm :: DWrap a -> [(BlockId, a)] +unDeterm (DWrap f) = f + +type DCmmTopInfo = GenCmmTopInfo DWrap +type CmmTopInfo = GenCmmTopInfo LabelMap instance OutputableP Platform CmmTopInfo where pdoc = pprTopInfo @@ -182,7 +203,12 @@ pprTopInfo platform (TopInfo {info_tbls=info_tbl, stack_info=stack_info}) = vcat [text "info_tbls: " <> pdoc platform info_tbl, text "stack_info: " <> ppr stack_info] -topInfoTable :: GenCmmDecl a CmmTopInfo (GenCmmGraph n) -> Maybe CmmInfoTable +topInfoTableD :: GenCmmDecl a DCmmTopInfo (GenGenCmmGraph s n) -> Maybe CmmInfoTable +topInfoTableD (CmmProc infos _ _ g) = case (info_tbls infos) of + DWrap xs -> lookup (g_entry g) xs +topInfoTableD _ = Nothing + +topInfoTable :: GenCmmDecl a CmmTopInfo (GenGenCmmGraph s n) -> Maybe CmmInfoTable topInfoTable (CmmProc infos _ _ g) = mapLookup (g_entry g) (info_tbls infos) topInfoTable _ = Nothing @@ -237,6 +263,7 @@ data ProfilingInfo = NoProfilingInfo | ProfilingInfo ByteString ByteString -- closure_type, closure_desc deriving (Eq, Ord) + ----------------------------------------------------------------------------- -- Static Data ----------------------------------------------------------------------------- @@ -328,6 +355,58 @@ instance OutputableP Platform (GenCmmStatics a) where type CmmStatics = GenCmmStatics 'False type RawCmmStatics = GenCmmStatics 'True +{- +----------------------------------------------------------------------------- +-- Deterministic Cmm / Info Tables +----------------------------------------------------------------------------- + +Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Consulting Note [Object determinism] one will learn that in order to produce +deterministic objects just after cmm is produced we perform a renaming pass which +provides fresh uniques for all unique-able things in the input Cmm. + +After this point, we use a deterministic unique supply (an incrementing counter) +so any resulting labels which make their way into object code have a deterministic name. + +A key assumption to this process is that the input is deterministic modulo the uniques +and the order that bindings appear in the definitions is the same. + +CmmGroup uses LabelMap in two places: + +* In CmmProc for info tables +* In CmmGraph for the blocks of the graph + +LabelMap is not a deterministic strucutre, so traversing a LabelMap can process +elements in different order (depending on the given uniques). + +Therefore before we do the renaming we need to use a deterministic strucutre, one +which we can traverse in a guaranteed order. A list does the job perfectly. + +Once the renaming happens it is converted back into a LabelMap, which is now deterministic +due to the uniques being generated and assigned in a deterministic manner. + +-} + +-- Converting out of deterministic Cmm + +removeDeterm :: DCmmGroup -> CmmGroup +removeDeterm = map removeDetermDecl + +removeDetermDecl :: DCmmDecl -> CmmDecl +removeDetermDecl (CmmProc h e r g) = CmmProc (removeDetermTop h) e r (removeDetermGraph g) +removeDetermDecl (CmmData a b) = CmmData a b + +removeDetermTop :: DCmmTopInfo -> CmmTopInfo +removeDetermTop (TopInfo a b) = TopInfo (mapFromList $ unDeterm a) b + +removeDetermGraph :: DCmmGraph -> CmmGraph +removeDetermGraph (CmmGraph x y) = + let y' = case y of + GMany a (DWrap b) c -> GMany a (mapFromList b) c + in CmmGraph x y' + -- ----------------------------------------------------------------------------- -- Basic blocks consisting of lists ===================================== compiler/GHC/Cmm/Dataflow/Graph.hs ===================================== @@ -26,10 +26,10 @@ import GHC.Cmm.Dataflow.Block import Data.Kind -- | A (possibly empty) collection of closed/closed blocks -type Body n = LabelMap (Block n C C) +type Body s n = Body' s Block n -- | @Body@ abstracted over @block@ -type Body' block (n :: Extensibility -> Extensibility -> Type) = LabelMap (block n C C) +type Body' s block (n :: Extensibility -> Extensibility -> Type) = s (block n C C) ------------------------------- -- | Gives access to the anchor points for @@ -46,13 +46,13 @@ instance NonLocal n => NonLocal (Block n) where successors (BlockCC _ _ n) = successors n -emptyBody :: Body' block n +emptyBody :: Body' LabelMap block n emptyBody = mapEmpty -bodyList :: Body' block n -> [(Label,block n C C)] +bodyList :: Body' LabelMap block n -> [(Label,block n C C)] bodyList body = mapToList body -bodyToBlockList :: Body n -> [Block n C C] +bodyToBlockList :: Body LabelMap n -> [Block n C C] bodyToBlockList body = mapElems body addBlock @@ -72,18 +72,18 @@ addBlock block body = mapAlter add lbl body -- O/C, C/O, C/C). A graph open at the entry has a single, -- distinguished, anonymous entry point; if a graph is closed at the -- entry, its entry point(s) are supplied by a context. -type Graph = Graph' Block +type Graph = Graph' LabelMap Block -- | @Graph'@ is abstracted over the block type, so that we can build -- graphs of annotated blocks for example (Compiler.Hoopl.Dataflow -- needs this). -data Graph' block (n :: Extensibility -> Extensibility -> Type) e x where - GNil :: Graph' block n O O - GUnit :: block n O O -> Graph' block n O O +data Graph' s block (n :: Extensibility -> Extensibility -> Type) e x where + GNil :: Graph' s block n O O + GUnit :: block n O O -> Graph' s block n O O GMany :: MaybeO e (block n O C) - -> Body' block n + -> Body' s block n -> MaybeO x (block n C O) - -> Graph' block n e x + -> Graph' s block n e x -- ----------------------------------------------------------------------------- @@ -91,26 +91,27 @@ data Graph' block (n :: Extensibility -> Extensibility -> Type) e x where -- | Maps over all nodes in a graph. mapGraph :: (forall e x. n e x -> n' e x) -> Graph n e x -> Graph n' e x -mapGraph f = mapGraphBlocks (mapBlock f) +mapGraph f = mapGraphBlocks mapMap (mapBlock f) -- | Function 'mapGraphBlocks' enables a change of representation of blocks, -- nodes, or both. It lifts a polymorphic block transform into a polymorphic -- graph transform. When the block representation stabilizes, a similar -- function should be provided for blocks. -mapGraphBlocks :: forall block n block' n' e x . - (forall e x . block n e x -> block' n' e x) - -> (Graph' block n e x -> Graph' block' n' e x) +mapGraphBlocks :: forall s block n block' n' e x . + (forall a b . (a -> b) -> s a -> s b) + -> (forall e x . block n e x -> block' n' e x) + -> (Graph' s block n e x -> Graph' s block' n' e x) -mapGraphBlocks f = map - where map :: Graph' block n e x -> Graph' block' n' e x +mapGraphBlocks f g = map + where map :: Graph' s block n e x -> Graph' s block' n' e x map GNil = GNil - map (GUnit b) = GUnit (f b) - map (GMany e b x) = GMany (fmap f e) (mapMap f b) (fmap f x) + map (GUnit b) = GUnit (g b) + map (GMany e b x) = GMany (fmap g e) (f g b) (fmap g x) -- ----------------------------------------------------------------------------- -- Extracting Labels from graphs -labelsDefined :: forall block n e x . NonLocal (block n) => Graph' block n e x +labelsDefined :: forall block n e x . NonLocal (block n) => Graph' LabelMap block n e x -> LabelSet labelsDefined GNil = setEmpty labelsDefined (GUnit{}) = setEmpty ===================================== compiler/GHC/Cmm/Graph.hs ===================================== @@ -73,12 +73,12 @@ data CgStmt | CgLast (CmmNode O C) | CgFork BlockId CmmAGraph CmmTickScope -flattenCmmAGraph :: BlockId -> CmmAGraphScoped -> CmmGraph +flattenCmmAGraph :: BlockId -> CmmAGraphScoped -> DCmmGraph flattenCmmAGraph id (stmts_t, tscope) = CmmGraph { g_entry = id, g_graph = GMany NothingO body NothingO } where - body = foldr addBlock emptyBody $ flatten id stmts_t tscope [] + body = DWrap [(entryLabel b, b) | b <- flatten id stmts_t tscope [] ] -- -- flatten: given an entry label and a CmmAGraph, make a list of blocks. @@ -169,13 +169,13 @@ outOfLine :: BlockId -> CmmAGraphScoped -> CmmAGraph outOfLine l (c,s) = unitOL (CgFork l c s) -- | allocate a fresh label for the entry point -lgraphOfAGraph :: CmmAGraphScoped -> UniqDSM CmmGraph +lgraphOfAGraph :: CmmAGraphScoped -> UniqDSM DCmmGraph lgraphOfAGraph g = do u <- getUniqueDSM return (labelAGraph (mkBlockId u) g) -- | use the given BlockId as the label of the entry point -labelAGraph :: BlockId -> CmmAGraphScoped -> CmmGraph +labelAGraph :: BlockId -> CmmAGraphScoped -> DCmmGraph labelAGraph lbl ag = flattenCmmAGraph lbl ag ---------- No-ops ===================================== compiler/GHC/Cmm/LayoutStack.hs ===================================== @@ -282,12 +282,18 @@ layout cfg procpoints liveness entry entry_args final_stackmaps final_sp_high bl where (updfr, cont_info) = collectContInfo blocks - init_stackmap = mapSingleton entry StackMap{ sm_sp = entry_args - , sm_args = entry_args - , sm_ret_off = updfr - , sm_regs = emptyUFM - } - + init_stackmap = mapSingleton entry + StackMap{ sm_sp = entry_args + , sm_args = entry_args + , sm_ret_off = updfr + , sm_regs = emptyUFM + } + + go :: [Block CmmNode C C] + -> LabelMap StackMap + -> StackLoc + -> [CmmBlock] + -> UniqDSM (LabelMap StackMap, StackLoc, [CmmBlock]) go [] acc_stackmaps acc_hwm acc_blocks = return (acc_stackmaps, acc_hwm, acc_blocks) @@ -1180,7 +1186,7 @@ lowerSafeForeignCall profile block copyout <*> mkLast jump, tscp) - case toBlockList graph' of + case toBlockList (removeDetermGraph graph') of [one] -> let (_, middle', last) = blockSplit one in return (blockJoin entry (middle `blockAppend` middle') last) _ -> panic "lowerSafeForeignCall0" ===================================== compiler/GHC/Cmm/Parser.y ===================================== @@ -1575,7 +1575,7 @@ parseCmmFile :: CmmParserConfig -> Module -> HomeUnit -> FilePath - -> IO (Messages PsMessage, Messages PsMessage, Maybe (CmmGroup, [InfoProvEnt])) + -> IO (Messages PsMessage, Messages PsMessage, Maybe (DCmmGroup, [InfoProvEnt])) parseCmmFile cmmpConfig this_mod home_unit filename = do buf <- hGetStringBuffer filename let @@ -1595,7 +1595,7 @@ parseCmmFile cmmpConfig this_mod home_unit filename = do ((), cmm) <- getCmm $ unEC code "global" (initEnv (pdProfile pdConfig)) [] >> return () -- See Note [Mapping Info Tables to Source Positions] (IPE Maps) let used_info - | do_ipe = map (cmmInfoTableToInfoProvEnt this_mod) (mapMaybe topInfoTable cmm) + | do_ipe = map (cmmInfoTableToInfoProvEnt this_mod) (mapMaybe topInfoTableD cmm) | otherwise = [] where do_ipe = stgToCmmInfoTableMap $ cmmpStgToCmmConfig cmmpConfig ===================================== compiler/GHC/Cmm/ThreadSanitizer.hs ===================================== @@ -19,6 +19,7 @@ import GHC.Types.Basic import GHC.Types.ForeignCall import GHC.Types.Unique import GHC.Types.Unique.Supply +import GHC.Cmm.Dataflow.Label import Data.Maybe (fromMaybe) @@ -29,7 +30,7 @@ data Env = Env { platform :: Platform annotateTSAN :: Platform -> CmmGraph -> UniqSM CmmGraph annotateTSAN platform graph = do env <- Env platform <$> getUniqueSupplyM - return $ modifyGraph (mapGraphBlocks (annotateBlock env)) graph + return $ modifyGraph (mapGraphBlocks mapMap (annotateBlock env)) graph mapBlockList :: (forall e' x'. n e' x' -> Block n e' x') -> Block n e x -> Block n e x ===================================== compiler/GHC/CmmToAsm/Reg/Liveness.hs ===================================== @@ -14,7 +14,7 @@ module GHC.CmmToAsm.Reg.Liveness ( RegSet, RegMap, emptyRegMap, - BlockMap, mapEmpty, + BlockMap, LiveCmmDecl, InstrSR (..), LiveInstr (..), @@ -260,7 +260,7 @@ instance OutputableP Platform LiveInfo where = (pdoc env mb_static) $$ text "# entryIds = " <> ppr entryIds $$ text "# liveVRegsOnEntry = " <> ppr liveVRegsOnEntry - $$ text "# liveSlotsOnEntry = " <> text (show liveSlotsOnEntry) + $$ text "# liveSlotsOnEntry = " <> ppr liveSlotsOnEntry ===================================== compiler/GHC/Driver/GenerateCgIPEStub.hs ===================================== @@ -20,7 +20,7 @@ import GHC.Driver.Env.Types (HscEnv) import GHC.Driver.Flags (GeneralFlag (..), DumpFlag(Opt_D_ipe_stats)) import GHC.Driver.DynFlags (gopt, targetPlatform) import GHC.Driver.Config.StgToCmm -import GHC.Driver.Config.Cmm +import GHC.Driver.Config.Cmm ( initCmmConfig ) import GHC.Prelude import GHC.Runtime.Heap.Layout (isStackRep) import GHC.Settings (platformTablesNextToCode) @@ -36,6 +36,7 @@ import GHC.Unit.Module (moduleNameString) import qualified GHC.Utils.Logger as Logger import GHC.Utils.Outputable (ppr) import GHC.Types.Unique.DSM +import GHC.Cmm.UniqueRenamer {- Note [Stacktraces from Info Table Provenance Entries (IPE based stack unwinding)] @@ -199,9 +200,10 @@ generateCgIPEStub , Map CmmInfoTable (Maybe IpeSourceLocation) , IPEStats , DUniqSupply + , DetUniqFM ) -> Stream IO CmmGroupSRTs CmmCgInfos -generateCgIPEStub hsc_env this_mod denv (nonCaffySet, moduleLFInfos, infoTablesWithTickishes, initStats, dus) = do +generateCgIPEStub hsc_env this_mod denv (nonCaffySet, moduleLFInfos, infoTablesWithTickishes, initStats, dus, detRnEnv) = do let dflags = hsc_dflags hsc_env platform = targetPlatform dflags logger = hsc_logger hsc_env @@ -213,7 +215,9 @@ generateCgIPEStub hsc_env this_mod denv (nonCaffySet, moduleLFInfos, infoTablesW let denv' = denv {provInfoTables = Map.mapKeys cit_lbl infoTablesWithTickishes} ((mIpeStub, ipeCmmGroup), _) = runC (initStgToCmmConfig dflags this_mod) fstate cgState $ getCmm (initInfoTableProv initStats (Map.keys infoTablesWithTickishes) denv') - (_, _, ipeCmmGroupSRTs) <- liftIO $ cmmPipeline logger cmm_cfg (emptySRT this_mod) dus ipeCmmGroup + (_detRnEnv', rn_ipeCmmGroup) = detRenameCmmGroup detRnEnv ipeCmmGroup + + (_, _, ipeCmmGroupSRTs) <- liftIO $ cmmPipeline logger cmm_cfg (emptySRT this_mod) dus rn_ipeCmmGroup Stream.yield ipeCmmGroupSRTs ipeStub <- ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -214,6 +214,7 @@ import GHC.Cmm.Info.Build import GHC.Cmm.Pipeline import GHC.Cmm.Info import GHC.Cmm.Parser +import GHC.Cmm.UniqueRenamer import GHC.Unit import GHC.Unit.Env @@ -299,7 +300,6 @@ import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) import GHC.Types.Unique.FM import GHC.Types.Unique.DFM -import GHC.Types.Unique.DSM import GHC.Cmm.Config (CmmConfig) {- ********************************************************************** @@ -2120,12 +2120,14 @@ hscCompileCmmFile hsc_env original_filename filename output_filename = runHsc hs mod_name = mkModuleName $ "Cmm$" ++ original_filename cmm_mod = mkHomeModule home_unit mod_name cmmpConfig = initCmmParserConfig dflags - (cmm, ipe_ents) <- ioMsgMaybe + (dcmm, ipe_ents) <- ioMsgMaybe $ do (warns,errs,cmm) <- withTiming logger (text "ParseCmm"<+>brackets (text filename)) (\_ -> ()) $ parseCmmFile cmmpConfig cmm_mod home_unit filename let msgs = warns `unionMessages` errs return (GhcPsMessage <$> msgs, cmm) + -- Probably need to rename cmm here + let cmm = removeDeterm dcmm liftIO $ do putDumpFileMaybe logger Opt_D_dump_cmm_verbose_by_proc "Parsed Cmm" FormatCMM (pdoc platform cmm) @@ -2210,11 +2212,11 @@ doCodeGen hsc_env this_mod denv data_tycons putDumpFileMaybe logger Opt_D_dump_stg_final "Final STG:" FormatSTG (pprGenStgTopBindings stg_ppr_opts stg_binds_w_fvs) - let stg_to_cmm dflags mod = case stgToCmmHook hooks of - Nothing -> StgToCmm.codeGen logger tmpfs (initStgToCmmConfig dflags mod) - Just h -> h (initStgToCmmConfig dflags mod) + let stg_to_cmm dflags mod a b c d e = case stgToCmmHook hooks of + Nothing -> StgToCmm.codeGen logger tmpfs (initStgToCmmConfig dflags mod) a b c d e + Just h -> (,emptyDetUFM) <$> h (initStgToCmmConfig dflags mod) a b c d e - let cmm_stream :: Stream IO CmmGroup ModuleLFInfos + let cmm_stream :: Stream IO CmmGroup (ModuleLFInfos, DetUniqFM) -- See Note [Forcing of stg_binds] cmm_stream = stg_binds_w_fvs `seqList` {-# SCC "StgToCmm" #-} stg_to_cmm dflags this_mod denv data_tycons cost_centre_info stg_binds_w_fvs hpc_info @@ -2236,11 +2238,11 @@ doCodeGen hsc_env this_mod denv data_tycons pipeline_stream :: Stream IO CmmGroupSRTs CmmCgInfos pipeline_stream = do - ((mod_srt_info, ipes, ipe_stats, dus), lf_infos) <- + ((mod_srt_info, ipes, ipe_stats, dus), (lf_infos, detRnEnv)) <- {-# SCC "cmmPipeline" #-} Stream.mapAccumL_ (pipeline_action logger cmm_config) (emptySRT this_mod, M.empty, mempty, initDUniqSupply 'u' 1) ppr_stream1 let nonCaffySet = srtMapNonCAFs (moduleSRTMap mod_srt_info) - cmmCgInfos <- generateCgIPEStub hsc_env this_mod denv (nonCaffySet, lf_infos, ipes, ipe_stats, dus) + cmmCgInfos <- generateCgIPEStub hsc_env this_mod denv (nonCaffySet, lf_infos, ipes, ipe_stats, dus, detRnEnv) return cmmCgInfos pipeline_action ===================================== compiler/GHC/Stg/Debug.hs ===================================== @@ -11,6 +11,7 @@ import GHC.Prelude import GHC.Stg.Syntax +import GHC.Types.Unique.DFM import GHC.Types.Id import GHC.Types.Tickish import GHC.Core.DataCon @@ -166,13 +167,13 @@ numberDataCon dc ts = do env <- lift get mcc <- asks rSpan let !mbest_span = (\(SpanWithLabel rss l) -> (rss, l)) <$> (selectTick ts <|> mcc) - let !dcMap' = alterUniqMap (maybe (Just ((0, mbest_span) :| [] )) - (\xs@((k, _):|_) -> Just $! ((k + 1, mbest_span) `NE.cons` xs))) (provDC env) dc + let !dcMap' = alterUDFM (maybe (Just (dc, (0, mbest_span) :| [] )) + (\(_dc, xs@((k, _):|_)) -> Just $! (dc, (k + 1, mbest_span) `NE.cons` xs))) (provDC env) dc lift $ put (env { provDC = dcMap' }) - let r = lookupUniqMap dcMap' dc + let r = lookupUDFM dcMap' dc return $ case r of Nothing -> NoNumber - Just res -> Numbered (fst (NE.head res)) + Just (_, res) -> Numbered (fst (NE.head res)) selectTick :: [StgTickish] -> Maybe SpanWithLabel selectTick [] = Nothing ===================================== compiler/GHC/StgToCmm.hs ===================================== @@ -43,6 +43,7 @@ import GHC.Types.Id.Info import GHC.Types.RepType import GHC.Types.Basic import GHC.Types.Var.Set ( isEmptyDVarSet ) +import GHC.Types.Unique.DFM import GHC.Types.Unique.FM import GHC.Types.Name.Env @@ -60,7 +61,6 @@ import GHC.Utils.TmpFs import GHC.Data.Stream import GHC.Data.OrdList -import GHC.Types.Unique.Map import Control.Monad (when,void, forM_) import GHC.Utils.Misc @@ -77,10 +77,11 @@ codeGen :: Logger -> CollectedCCs -- (Local/global) cost-centres needing declaring/registering. -> [CgStgTopBinding] -- Bindings to convert -> HpcInfo - -> Stream IO CmmGroup ModuleLFInfos -- Output as a stream, so codegen can + -> Stream IO CmmGroup (ModuleLFInfos, DetUniqFM) + -- Output as a stream, so codegen can -- be interleaved with output -codeGen logger tmpfs cfg (InfoTableProvMap (UniqMap denv) _ _) data_tycons +codeGen logger tmpfs cfg (InfoTableProvMap denv _ _) data_tycons cost_centre_info stg_binds hpc_info = do { -- cg: run the code generator, and yield the resulting CmmGroup -- Using an IORef to store the state is a bit crude, but otherwise @@ -103,8 +104,8 @@ codeGen logger tmpfs cfg (InfoTableProvMap (UniqMap denv) _ _) data_tycons -- renaming uniques deterministically. -- See Note [Object determinism] if stgToCmmObjectDeterminism cfg - then detRenameUniques rnm0 cmm -- The yielded cmm will already be renamed. - else (rnm0, cmm) + then detRenameCmmGroup rnm0 cmm -- The yielded cmm will already be renamed. + else (rnm0, removeDeterm cmm) -- NB. stub-out cgs_tops and cgs_stmts. This fixes -- a big space leak. DO NOT REMOVE! @@ -135,7 +136,7 @@ codeGen logger tmpfs cfg (InfoTableProvMap (UniqMap denv) _ _) data_tycons -- Emit special info tables for everything used in this module -- This will only do something if `-fdistinct-info-tables` is turned on. - ; mapM_ (\(dc, ns) -> forM_ ns $ \(k, _ss) -> cg (cgDataCon (UsageSite (stgToCmmThisModule cfg) k) dc)) (nonDetEltsUFM denv) + ; mapM_ (\(dc, ns) -> forM_ ns $ \(k, _ss) -> cg (cgDataCon (UsageSite (stgToCmmThisModule cfg) k) dc)) (eltsUDFM denv) ; final_state <- liftIO (readIORef cgref) ; let cg_id_infos = cgs_binds final_state @@ -156,7 +157,7 @@ codeGen logger tmpfs cfg (InfoTableProvMap (UniqMap denv) _ _) data_tycons ; rn_mapping <- liftIO (readIORef uniqRnRef) ; liftIO $ debugTraceMsg logger 3 (text "DetRnM mapping:" <+> ppr rn_mapping) - ; return generatedInfo + ; return (generatedInfo, rn_mapping) } {- ===================================== compiler/GHC/StgToCmm/CgUtils.hs ===================================== @@ -26,6 +26,7 @@ import GHC.Cmm.Dataflow.Graph import GHC.Cmm.Utils import GHC.Cmm.CLabel import GHC.Utils.Panic +import GHC.Cmm.Dataflow.Label -- ----------------------------------------------------------------------------- -- Information about global registers @@ -132,7 +133,7 @@ fixStgRegisters :: Platform -> RawCmmDecl -> RawCmmDecl fixStgRegisters _ top@(CmmData _ _) = top fixStgRegisters platform (CmmProc info lbl live graph) = - let graph' = modifyGraph (mapGraphBlocks (fixStgRegBlock platform)) graph + let graph' = modifyGraph (mapGraphBlocks mapMap (fixStgRegBlock platform)) graph in CmmProc info lbl live graph' fixStgRegBlock :: Platform -> Block CmmNode e x -> Block CmmNode e x ===================================== compiler/GHC/StgToCmm/Monad.hs ===================================== @@ -76,7 +76,6 @@ import GHC.StgToCmm.Sequel import GHC.Cmm.Graph as CmmGraph import GHC.Cmm.BlockId import GHC.Cmm.CLabel -import GHC.Cmm.Dataflow.Label import GHC.Runtime.Heap.Layout import GHC.Unit import GHC.Types.Id @@ -285,7 +284,7 @@ data CgState = MkCgState { cgs_stmts :: CmmAGraph, -- Current procedure - cgs_tops :: OrdList CmmDecl, + cgs_tops :: OrdList DCmmDecl, -- Other procedures and data blocks in this compilation unit -- Both are ordered only so that we can -- reduce forward references, when it's easy to do so @@ -744,7 +743,7 @@ emit ag = do { state <- getState ; setState $ state { cgs_stmts = cgs_stmts state CmmGraph.<*> ag } } -emitDecl :: CmmDecl -> FCode () +emitDecl :: DCmmDecl -> FCode () emitDecl decl = do { state <- getState ; setState $ state { cgs_tops = cgs_tops state `snocOL` decl } } @@ -787,16 +786,16 @@ emitProc :: Maybe CmmInfoTable -> CLabel -> [GlobalReg] -> CmmAGraphScoped emitProc mb_info lbl live blocks offset do_layout = do { l <- newBlockId ; let - blks :: CmmGraph + blks :: DCmmGraph blks = labelAGraph l blocks - infos | Just info <- mb_info = mapSingleton (g_entry blks) info - | otherwise = mapEmpty + infos | Just info <- mb_info = [((g_entry blks), info)] + | otherwise = [] sinfo = StackInfo { arg_space = offset , do_layout = do_layout } - tinfo = TopInfo { info_tbls = infos + tinfo = TopInfo { info_tbls = DWrap infos , stack_info=sinfo} proc_block = CmmProc tinfo lbl live blks @@ -804,7 +803,7 @@ emitProc mb_info lbl live blocks offset do_layout ; state <- getState ; setState $ state { cgs_tops = cgs_tops state `snocOL` proc_block } } -getCmm :: FCode a -> FCode (a, CmmGroup) +getCmm :: FCode a -> FCode (a, DCmmGroup) -- Get all the CmmTops (there should be no stmts) -- Return a single Cmm which may be split from other Cmms by -- object splitting (at a later stage) @@ -880,7 +879,7 @@ mkCmmCall f results actuals updfr_off -- ---------------------------------------------------------------------------- -- turn CmmAGraph into CmmGraph, for making a new proc. -aGraphToGraph :: CmmAGraphScoped -> FCode CmmGraph +aGraphToGraph :: CmmAGraphScoped -> FCode DCmmGraph aGraphToGraph stmts = do { l <- newBlockId ; return (labelAGraph l stmts) } ===================================== compiler/GHC/StgToCmm/Utils.hs ===================================== @@ -90,7 +90,7 @@ import GHC.Types.Unique.Map import Data.Maybe import qualified Data.List.NonEmpty as NE import GHC.Core.DataCon -import GHC.Types.Unique.FM +import GHC.Types.Unique.DFM import GHC.Data.Maybe import Control.Monad import qualified Data.Map.Strict as Map @@ -673,7 +673,7 @@ pprIPEStats (IPEStats{..}) = -- for stack info tables skipped during 'generateCgIPEStub'. As the fold -- progresses, counts of tables per closure type will be accumulated. convertInfoProvMap :: StgToCmmConfig -> Module -> InfoTableProvMap -> IPEStats -> [CmmInfoTable] -> (IPEStats, [InfoProvEnt]) -convertInfoProvMap cfg this_mod (InfoTableProvMap (UniqMap dcenv) denv infoTableToSourceLocationMap) initStats cmits = +convertInfoProvMap cfg this_mod (InfoTableProvMap dcenv denv infoTableToSourceLocationMap) initStats cmits = foldl' convertInfoProvMap' (initStats, []) cmits where convertInfoProvMap' :: (IPEStats, [InfoProvEnt]) -> CmmInfoTable -> (IPEStats, [InfoProvEnt]) @@ -694,7 +694,7 @@ convertInfoProvMap cfg this_mod (InfoTableProvMap (UniqMap dcenv) denv infoTable lookupDataConMap = (closureIpeStats cn,) <$> do UsageSite _ n <- hasIdLabelInfo cl >>= getConInfoTableLocation -- This is a bit grimy, relies on the DataCon and Name having the same Unique, which they do - (dc, ns) <- hasHaskellName cl >>= lookupUFM_Directly dcenv . getUnique + (dc, ns) <- hasHaskellName cl >>= lookupUDFM_Directly dcenv . getUnique -- Lookup is linear but lists will be small (< 100) return $ (InfoProvEnt cl cn (tyString (dataConTyCon dc)) this_mod (join $ lookup n (NE.toList ns))) ===================================== compiler/GHC/Types/IPE.hs ===================================== @@ -13,6 +13,7 @@ import GHC.Data.FastString import GHC.Types.SrcLoc import GHC.Core.DataCon +import GHC.Types.Unique.DFM import GHC.Types.Unique.Map import GHC.Core.Type import Data.List.NonEmpty @@ -38,7 +39,7 @@ type ClosureMap = UniqMap Name -- The binding -- the constructor was used at, if possible and a string which names -- the source location. This is the same information as is the payload -- for the 'GHC.Core.SourceNote' constructor. -type DCMap = UniqMap DataCon (NonEmpty (Int, Maybe IpeSourceLocation)) +type DCMap = UniqDFM DataCon (DataCon, NonEmpty (Int, Maybe IpeSourceLocation)) type InfoTableToSourceLocationMap = Map.Map CLabel (Maybe IpeSourceLocation) @@ -49,4 +50,4 @@ data InfoTableProvMap = InfoTableProvMap } emptyInfoTableProvMap :: InfoTableProvMap -emptyInfoTableProvMap = InfoTableProvMap emptyUniqMap emptyUniqMap Map.empty +emptyInfoTableProvMap = InfoTableProvMap emptyUDFM emptyUniqMap Map.empty ===================================== compiler/GHC/Types/Name.hs ===================================== @@ -716,9 +716,9 @@ pprName name@(Name {n_sort = sort, n_uniq = uniq, n_occ = occ}) {-# SPECIALISE pprName :: Name -> SDoc #-} {-# SPECIALISE pprName :: Name -> HLine #-} -- see Note [SPECIALIZE to HDoc] in GHC.Utils.Outputable --- | Print fully qualified name (with unit-id, module and unique) +-- | Print fully qualified name (with unit-id, module but no unique) pprFullName :: Module -> Name -> SDoc -pprFullName this_mod Name{n_sort = sort, n_uniq = uniq, n_occ = occ} = +pprFullName this_mod Name{n_sort = sort, n_occ = occ} = let mod = case sort of WiredIn m _ _ -> m External m -> m @@ -727,8 +727,6 @@ pprFullName this_mod Name{n_sort = sort, n_uniq = uniq, n_occ = occ} = in ftext (unitIdFS (moduleUnitId mod)) <> colon <> ftext (moduleNameFS $ moduleName mod) <> dot <> ftext (occNameFS occ) - <> char '_' <> pprUniqueAlways uniq - -- | Print a ticky ticky styled name -- ===================================== compiler/GHC/Types/Unique/DFM.hs ===================================== @@ -43,10 +43,10 @@ module GHC.Types.Unique.DFM ( mapMaybeUDFM, mapMUDFM, plusUDFM, - plusUDFM_C, + plusUDFM_C, plusUDFM_CK, lookupUDFM, lookupUDFM_Directly, elemUDFM, - foldUDFM, + foldUDFM, foldWithKeyUDFM, eltsUDFM, filterUDFM, filterUDFM_Directly, isNullUDFM, @@ -56,6 +56,7 @@ module GHC.Types.Unique.DFM ( equalKeysUDFM, minusUDFM, listToUDFM, listToUDFM_Directly, + listToUDFM_C_Directly, udfmMinusUFM, ufmMinusUDFM, partitionUDFM, udfmRestrictKeys, @@ -224,6 +225,12 @@ addListToUDFM_Directly_C addListToUDFM_Directly_C f = foldl' (\m (k, v) -> addToUDFM_C_Directly f m k v) {-# INLINEABLE addListToUDFM_Directly_C #-} +-- | Like 'addListToUDFM_Directly_C' but also passes the unique key to the combine function +addListToUDFM_Directly_CK + :: (Unique -> elt -> elt -> elt) -> UniqDFM key elt -> [(Unique,elt)] -> UniqDFM key elt +addListToUDFM_Directly_CK f = foldl' (\m (k, v) -> addToUDFM_C_Directly (f k) m k v) +{-# INLINEABLE addListToUDFM_Directly_CK #-} + delFromUDFM :: Uniquable key => UniqDFM key elt -> key -> UniqDFM key elt delFromUDFM (UDFM m i) k = UDFM (M.delete (getKey $ getUnique k) m) i @@ -234,6 +241,15 @@ plusUDFM_C f udfml@(UDFM _ i) udfmr@(UDFM _ j) | i > j = insertUDFMIntoLeft_C f udfml udfmr | otherwise = insertUDFMIntoLeft_C f udfmr udfml +-- | Like 'plusUDFM_C' but the combine function also receives the unique key +plusUDFM_CK :: (Unique -> elt -> elt -> elt) -> UniqDFM key elt -> UniqDFM key elt -> UniqDFM key elt +plusUDFM_CK f udfml@(UDFM _ i) udfmr@(UDFM _ j) + -- we will use the upper bound on the tag as a proxy for the set size, + -- to insert the smaller one into the bigger one + | i > j = insertUDFMIntoLeft_CK f udfml udfmr + | otherwise = insertUDFMIntoLeft_CK f udfmr udfml + + -- Note [Overflow on plusUDFM] -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~ -- There are multiple ways of implementing plusUDFM. @@ -282,6 +298,12 @@ insertUDFMIntoLeft_C insertUDFMIntoLeft_C f udfml udfmr = addListToUDFM_Directly_C f udfml $ udfmToList udfmr +-- | Like 'insertUDFMIntoLeft_C', but the merge function also receives the unique key +insertUDFMIntoLeft_CK + :: (Unique -> elt -> elt -> elt) -> UniqDFM key elt -> UniqDFM key elt -> UniqDFM key elt +insertUDFMIntoLeft_CK f udfml udfmr = + addListToUDFM_Directly_CK f udfml $ udfmToList udfmr + lookupUDFM :: Uniquable key => UniqDFM key elt -> key -> Maybe elt lookupUDFM (UDFM m _i) k = taggedFst `fmap` M.lookup (getKey $ getUnique k) m @@ -298,6 +320,12 @@ foldUDFM :: (elt -> a -> a) -> a -> UniqDFM key elt -> a -- This INLINE prevents a regression in !10568 foldUDFM k z m = foldr k z (eltsUDFM m) +-- | Like 'foldUDFM' but the function also receives a key +foldWithKeyUDFM :: (Unique -> elt -> a -> a) -> a -> UniqDFM key elt -> a +{-# INLINE foldWithKeyUDFM #-} +-- This INLINE was copied from foldUDFM +foldWithKeyUDFM k z m = foldr (uncurry k) z (udfmToList m) + -- | Performs a nondeterministic strict fold over the UniqDFM. -- It's O(n), same as the corresponding function on `UniqFM`. -- If you use this please provide a justification why it doesn't introduce @@ -397,6 +425,9 @@ listToUDFM = foldl' (\m (k, v) -> addToUDFM m k v) emptyUDFM listToUDFM_Directly :: [(Unique, elt)] -> UniqDFM key elt listToUDFM_Directly = foldl' (\m (u, v) -> addToUDFM_Directly m u v) emptyUDFM +listToUDFM_C_Directly :: (elt -> elt -> elt) -> [(Unique, elt)] -> UniqDFM key elt +listToUDFM_C_Directly f = foldl' (\m (u, v) -> addToUDFM_C_Directly f m u v) emptyUDFM + -- | Apply a function to a particular element adjustUDFM :: Uniquable key => (elt -> elt) -> UniqDFM key elt -> key -> UniqDFM key elt adjustUDFM f (UDFM m i) k = UDFM (M.adjust (fmap f) (getKey $ getUnique k) m) i ===================================== compiler/GHC/Utils/Outputable.hs ===================================== @@ -142,6 +142,7 @@ import Data.Int import qualified Data.IntMap as IM import Data.Set (Set) import qualified Data.Set as Set +import qualified Data.IntSet as IntSet import qualified GHC.Data.Word64Set as Word64Set import Data.String import Data.Word @@ -991,6 +992,9 @@ instance (Outputable a) => Outputable (Set a) where instance Outputable Word64Set.Word64Set where ppr s = braces (pprWithCommas ppr (Word64Set.toList s)) +instance Outputable IntSet.IntSet where + ppr s = braces (pprWithCommas ppr (IntSet.toList s)) + instance (Outputable a, Outputable b) => Outputable (a, b) where ppr (x,y) = parens (sep [ppr x <> comma, ppr y]) ===================================== testsuite/tests/regalloc/regalloc_unit_tests.hs ===================================== @@ -137,7 +137,9 @@ compileCmmForRegAllocStats logger home_unit dflags cmmFile ncgImplF us = do -- parse the cmm file and output any warnings or errors let fake_mod = mkHomeModule home_unit (mkModuleName "fake") cmmpConfig = initCmmParserConfig dflags - (warnings, errors, parsedCmm) <- parseCmmFile cmmpConfig fake_mod home_unit cmmFile + (warnings, errors, dparsedCmm) <- parseCmmFile cmmpConfig fake_mod home_unit cmmFile + + let parsedCmm = removeDeterm (fst (fromJust dparsedCmm)) -- print parser errors or warnings let !diag_opts = initDiagOpts dflags ===================================== testsuite/tests/wasm/should_run/control-flow/LoadCmmGroup.hs ===================================== @@ -120,7 +120,7 @@ slurpCmm hsc_env filename = runHsc hsc_env $ do $ parseCmmFile cmmpConfig cmm_mod home_unit filename let msgs = warns `unionMessages` errs return (GhcPsMessage <$> msgs, cmm) - return cmm + return (removeDeterm cmm) collectAll :: Monad m => Stream m a b -> m ([a], b) collectAll = gobble . runStream View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3dc7d298060098fdf864dc83e87152521f11d212...f446aa3cdaccc492d3ac1b07593e536b3d9590ef -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3dc7d298060098fdf864dc83e87152521f11d212...f446aa3cdaccc492d3ac1b07593e536b3d9590ef You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 10:08:21 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 05 Sep 2024 06:08:21 -0400 Subject: [Git][ghc/ghc][wip/romes/25110] base: Deprecate BCO primops exports from GHC.Exts Message-ID: <66d9831529d2a_3bd3ee3f3ff41098cb@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/25110 at Glasgow Haskell Compiler / GHC Commits: 99984e36 by Rodrigo Mesquita at 2024-09-05T11:08:11+01:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - 4 changed files: - libraries/base/src/GHC/Exts.hs - libraries/ghci/GHCi/CreateBCO.hs - libraries/ghci/ghci.cabal.in - + testsuite/tests/module/T21752.stderr Changes: ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -23,6 +23,12 @@ module GHC.Exts -- ** Legacy interface for arrays of arrays module GHC.Internal.ArrayArray, -- * Primitive operations + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.BCO, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.mkApUpd0#, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.newBCO#, module GHC.Prim, module GHC.Prim.Ext, -- ** Running 'RealWorld' state thread @@ -112,7 +118,10 @@ module GHC.Exts import GHC.Internal.Exts import GHC.Internal.ArrayArray -import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# ) +import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom# + -- Deprecated + , BCO, mkApUpd0#, newBCO# ) +import qualified GHC.Prim as Prim ( BCO, mkApUpd0#, newBCO# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/ghci/GHCi/CreateBCO.hs ===================================== @@ -6,6 +6,10 @@ {-# LANGUAGE UnboxedTuples #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE CPP #-} +{-# OPTIONS_GHC -Wno-warnings-deprecations #-} +-- TODO We want to import GHC.Internal.Base (BCO, mkApUpd0#, newBCO#) instead +-- of from GHC.Exts when we can require of the bootstrap compiler to have +-- ghc-internal. -- -- (c) The University of Glasgow 2002-2006 ===================================== libraries/ghci/ghci.cabal.in ===================================== @@ -86,9 +86,10 @@ library array == 0.5.*, base >= 4.8 && < 4.21, -- ghc-internal == @ProjectVersionForLib at .* - -- TODO: Use GHC.Internal.Desugar from ghc-internal instead of ignoring - -- the deprecation warning of GHC.Desugar when we require ghc-internal - -- of the bootstrap compiler + -- TODO: Use GHC.Internal.Desugar and GHC.Internal.Base from + -- ghc-internal instead of ignoring the deprecation warning in GHCi.TH + -- and GHCi.CreateBCO when we require ghc-internal of the bootstrap + -- compiler ghc-prim >= 0.5.0 && < 0.12, binary == 0.8.*, bytestring >= 0.10 && < 0.13, ===================================== testsuite/tests/module/T21752.stderr ===================================== @@ -0,0 +1,32 @@ +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘newBCO#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘newBCO#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘mkApUpd0#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘mkApUpd0#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of type constructor or class ‘BCO’ + (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of type constructor or class ‘BCO’ + (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/99984e36f346ff59968d9e33b64c0add2a412356 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/99984e36f346ff59968d9e33b64c0add2a412356 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 10:31:56 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 05 Sep 2024 06:31:56 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] Apply typo/grammar suggestions Message-ID: <66d9889c836da_3bd3ee57bc3c111542@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 5856aa6e by Sven Tennie at 2024-09-05T10:31:53+00:00 Apply typo/grammar suggestions - - - - - 3 changed files: - compiler/GHC/CmmToAsm/RV64/CodeGen.hs - compiler/GHC/CmmToAsm/RV64/Instr.hs - compiler/GHC/CmmToAsm/RV64/Ppr.hs Changes: ===================================== compiler/GHC/CmmToAsm/RV64/CodeGen.hs ===================================== @@ -163,7 +163,7 @@ ann doc instr {- debugIsOn -} = ANN doc instr -- going back to the exact CmmExpr representation can be laborious and adds -- indirections to find the matches that lead to the assembly. -- --- An improvement oculd be to have +-- An improvement could be to have -- -- (pprExpr genericPlatform e) <> parens (text. show e) -- ===================================== compiler/GHC/CmmToAsm/RV64/Instr.hs ===================================== @@ -267,7 +267,7 @@ patchJumpInstr instr patchF = -- single load/store instruction. There are offsets to sp (not to be confused -- with STG's SP!) which need a register to be calculated. -- --- Using sp to compute the offset will violate assumptions about the stack pointer +-- Using sp to compute the offset would violate assumptions about the stack pointer -- pointing to the top of the stack during signal handling. As we can't force -- every signal to use its own stack, we have to ensure that the stack pointer -- always points to the top of the stack, and we can't use it for computation. ===================================== compiler/GHC/CmmToAsm/RV64/Ppr.hs ===================================== @@ -222,7 +222,7 @@ pprGloblDecl platform lbl -- (PLT)s, and thus the lookup wouldn't point to the function, but into the -- jump table. -- --- Fun fact: The LLVMMangler exists to patch this issue su on the LLVM side as +-- Fun fact: The LLVMMangler exists to patch this issue on the LLVM side as -- well. pprLabelType' :: (IsLine doc) => Platform -> CLabel -> doc pprLabelType' platform lbl = View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5856aa6ec7f5673f227558d4378aecba1b692dd0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5856aa6ec7f5673f227558d4378aecba1b692dd0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 10:37:14 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 05 Sep 2024 06:37:14 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 4 commits: Silence x-partial in Haddock.Backends.Xhtml Message-ID: <66d989da12d23_3bd3ee5a47b8115281@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 46021a1b by Hécate Kleidukos at 2024-09-05T06:37:04-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 23a4eed5 by Hécate Kleidukos at 2024-09-05T06:37:04-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 79bc287a by Matthew Pickering at 2024-09-05T06:37:04-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 07ba0d21 by Matthew Pickering at 2024-09-05T06:37:05-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 8 changed files: - .gitlab/ci.sh - hadrian/ghci-multi-cabal.in - hadrian/src/Rules/ToolArgs.hs - utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Layout.hs - utils/haddock/haddock-api/src/Haddock/Utils/Json.hs Changes: ===================================== .gitlab/ci.sh ===================================== @@ -753,7 +753,7 @@ function check_interfaces(){ } function abi_test() { - for i in {1..20}; do info "iteration $i"; run_abi_test; done + for i in {1..10}; do info "iteration $i"; run_abi_test; done } function run_abi_test() { @@ -761,8 +761,8 @@ function run_abi_test() { fail "HC not set" fi mkdir -p out - OUT="$PWD/out/run1" DIR=$(mktemp -d XXXX-looooooooong) cabal_abi_test -O0 - OUT="$PWD/out/run2" DIR=$(mktemp -d XXXX-short) cabal_abi_test -O0 + OUT="$PWD/out/run1" DIR=$(mktemp -d XXXX-looooooooong) cabal_abi_test -O1 -haddock + OUT="$PWD/out/run2" DIR=$(mktemp -d XXXX-short) cabal_abi_test -O1 -haddock -dunique-increment=-1 -dinitial-unique=16777215 check_interfaces out/run1 out/run2 abis "Mismatched ABI hash" check_interfaces out/run1 out/run2 interfaces "Mismatched interface hashes" } ===================================== hadrian/ghci-multi-cabal.in ===================================== @@ -8,6 +8,6 @@ if [[ $(printf "9.4.0\n%s\n" $($RUN_GHC --numeric-version) | sort -uV | head -n set -e export TOOL_OUTPUT=.hadrian_ghci_multi/ghci_args # Replace newlines with spaces, as these otherwise break the ghci invocation on windows. -CABFLAGS=-v0 "hadrian/build-cabal" multi:ghc --build-root=.hadrian_ghci_multi --flavour=ghc-in-ghci $HADRIAN_ARGS +CABFLAGS=-v0 "hadrian/build-cabal" multi --build-root=.hadrian_ghci_multi --flavour=ghc-in-ghci $HADRIAN_ARGS GHC_FLAGS="$GHC_FLAGS $(cat $TOOL_OUTPUT | tr '\n\r' ' ')" $RUN_GHC --interactive $GHC_FLAGS $@ -fno-code -fwrite-interface -O0 +RTS -A128m ===================================== hadrian/src/Rules/ToolArgs.hs ===================================== @@ -16,7 +16,9 @@ import System.Directory (canonicalizePath) import System.Environment (lookupEnv) import qualified Data.Set as Set import Oracles.ModuleFiles +import Oracles.Setting import Utilities +import Data.Version.Extra -- | @tool:@ is used by tooling in order to get the arguments necessary -- to set up a GHC API session which can compile modules from GHC. When @@ -85,7 +87,16 @@ multiSetup pkg_s = do need (srcs ++ gens) let rexp m = ["-reexported-module", m] let hidir = root "interfaces" pkgPath p - writeFile' (resp_file root p) (intercalate "\n" (arg_list + ghcVersion <- ghcVersionStage stage0InTree + let ghc_wired_in = readVersion ghcVersion < makeVersion [9,8,1] + ghc_package_id = "-package-id ghc-" ++ ghcVersion + normalise_ghc = if ghc_wired_in then normalisePackageIds else id + normalisePackageIds :: [String] -> [String] + normalisePackageIds ((isPrefixOf ghc_package_id -> True) : xs) = "-package-id" : "ghc" : xs + normalisePackageIds (x:xs) = x : normalisePackageIds xs + normalisePackageIds [] = [] + + writeFile' (resp_file root p) (intercalate "\n" (normalise_ghc arg_list ++ modules cd ++ concatMap rexp (reexportModules cd) ++ ["-outputdir", hidir])) @@ -150,7 +161,9 @@ toolTargets = [ cabalSyntax , ghcHeap , ghci , ghcPkg -- # executable - -- , haddock -- # depends on ghc library + , haddock -- # depends on ghc library + , haddockApi + , haddockLibrary , hsc2hs -- # executable , hpc , hpcBin -- # executable ===================================== utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs ===================================== @@ -26,20 +26,20 @@ import Data.Foldable (toList) import Data.List (sort) import Data.List.NonEmpty (NonEmpty (..)) import qualified Data.Map as Map -import Data.Maybe +import qualified Data.Maybe as Maybe import GHC hiding (fromMaybeContext) import GHC.Core.Type (Specificity (..)) import GHC.Data.FastString (unpackFS) import GHC.Types.Name (getOccString, nameOccName, tidyNameOcc) import GHC.Types.Name.Occurrence import GHC.Types.Name.Reader (rdrNameOcc) +import GHC.Utils.Ppr hiding (Doc, quote) +import qualified GHC.Utils.Ppr as Pretty import System.Directory import System.FilePath import Prelude hiding ((<>)) import Documentation.Haddock.Markup -import GHC.Utils.Ppr hiding (Doc, quote) -import qualified GHC.Utils.Ppr as Pretty import Haddock.Doc (combineDocumentation) import Haddock.GhcUtils import Haddock.Types @@ -90,7 +90,7 @@ ppLaTeX ppLaTeX title packageStr visible_ifaces odir prologue maybe_style libdir = do createDirectoryIfMissing True odir - when (isNothing maybe_style) $ + when (Maybe.isNothing maybe_style) $ copyFile (libdir "latex" haddockSty) (odir haddockSty) ppLaTeXTop title packageStr odir prologue maybe_style visible_ifaces mapM_ (ppLaTeXModule title odir) visible_ifaces @@ -139,7 +139,7 @@ ppLaTeXTop doctitle packageStr odir prologue maybe_style ifaces = do mods = sort (map (moduleBasename . ifaceMod) ifaces) - filename = odir (fromMaybe "haddock" packageStr <.> "tex") + filename = odir (Maybe.fromMaybe "haddock" packageStr <.> "tex") writeUtf8File filename (show tex) @@ -174,7 +174,7 @@ ppLaTeXModule _title odir iface = do ] description = - (fromMaybe empty . documentationToLaTeX . ifaceRnDoc) iface + (Maybe.fromMaybe empty . documentationToLaTeX . ifaceRnDoc) iface body = processExports exports -- @@ -201,7 +201,7 @@ exportListItem in sep (punctuate comma [leader <+> ppDocBinder name | name <- names]) <> case subdocs of [] -> empty - _ -> parens (sep (punctuate comma (mapMaybe go subdocs))) + _ -> parens (sep (punctuate comma (Maybe.mapMaybe go subdocs))) exportListItem (ExportNoDecl y []) = ppDocBinder y exportListItem (ExportNoDecl y subs) = @@ -368,7 +368,7 @@ ppFamDecl associated doc instances decl unicode = (if null body then Nothing else Just (vcat body)) $$ instancesBit where - body = catMaybes [familyEqns, documentationToLaTeX doc] + body = Maybe.catMaybes [familyEqns, documentationToLaTeX doc] whereBit = case fdInfo (tcdFam decl) of ClosedTypeFamily _ -> keyword "where" @@ -544,7 +544,7 @@ ppTypeOrFunSig typ (doc, argDocs) (pref1, pref2, sep0) unicode text "\\haddockbeginargs" $$ vcat (map (uncurry (<->)) (ppSubSigLike unicode typ argDocs [] sep0)) $$ text "\\end{tabulary}\\par" - $$ fromMaybe empty (documentationToLaTeX doc) + $$ Maybe.fromMaybe empty (documentationToLaTeX doc) -- | This splits up a type signature along @->@ and adds docs (when they exist) -- to the arguments. The output is a list of (leader/seperator, argument and @@ -741,7 +741,7 @@ ppClassDecl hdr = ppClassHdr False lctxt (unLoc lname) ltyvars lfds - body = catMaybes [documentationToLaTeX doc, body_] + body = Maybe.catMaybes [documentationToLaTeX doc, body_] body_ | null lsigs, null ats, null at_defs = Nothing @@ -764,9 +764,13 @@ ppClassDecl | L _ (ClassOpSig _ is_def lnames typ) <- lsigs , let doc | is_def = noDocForDecl - | otherwise = lookupAnySubdoc (head names) subdocs + | otherwise = lookupAnySubdoc firstName subdocs names = map (cleanName . unLoc) lnames leader = if is_def then Just (keyword "default") else Nothing + firstName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd ] -- N.B. taking just the first name is ok. Signatures with multiple -- names are expanded so that each name gets its own signature. @@ -853,7 +857,7 @@ ppDataDecl pats instances subdocs doc dataDecl unicode = where cons = dd_cons (tcdDataDefn dataDecl) - body = catMaybes [doc >>= documentationToLaTeX, constrBit, patternBit] + body = Maybe.catMaybes [doc >>= documentationToLaTeX, constrBit, patternBit] (whereBit, leaders) | null cons @@ -1031,7 +1035,11 @@ ppSideBySideField subdocs unicode (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc $ head names) subdocs >>= fmap _doc . combineDocumentation . fst + mbDoc = lookup (foExt $ unLoc name) subdocs >>= fmap _doc . combineDocumentation . fst + name = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -- | Pretty-print a bundled pattern synonym ppSideBySidePat @@ -1157,7 +1165,7 @@ ppContextNoLocsMaybe cxt unicode = Just $ pp_hs_context cxt unicode ppContextNoArrow :: HsContext DocNameI -> Bool -> LaTeX ppContextNoArrow cxt unicode = - fromMaybe empty $ + Maybe.fromMaybe empty $ ppContextNoLocsMaybe (map unLoc cxt) unicode ppContextNoLocs :: [HsType DocNameI] -> Bool -> LaTeX ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml.hs ===================================== @@ -6,6 +6,7 @@ {-# LANGUAGE TupleSections #-} {-# LANGUAGE TypeApplications #-} {-# LANGUAGE TypeFamilies #-} +{-# OPTIONS_GHC -Wwarn=x-partial #-} -- | -- Module : Haddock.Backends.Html ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs ===================================== @@ -10,7 +10,7 @@ ----------------------------------------------------------------------------- -- | --- Module : Haddock.Backends.Html.Decl +-- Module : Haddock.Backends.Xhtml.Decl -- Copyright : (c) Simon Marlow 2003-2006, -- David Waern 2006-2009, -- Mark Lentczner 2010 @@ -28,7 +28,7 @@ import Data.Foldable (toList) import Data.List (intersperse, sort) import Data.List.NonEmpty (NonEmpty (..)) import qualified Data.Map as Map -import Data.Maybe +import qualified Data.Maybe as Maybe import GHC hiding (LexicalFixity (..), fromMaybeContext) import GHC.Core.Type (Specificity (..)) import GHC.Data.BooleanFormula @@ -279,13 +279,17 @@ ppTypeOrFunSig qual emptyCtxts | summary = pref1 - | Map.null argDocs = topDeclElem links loc splice docnames pref1 +++ docSection curname pkg qual doc + | Map.null argDocs = topDeclElem links loc splice docName pref1 +++ docSection curname pkg qual doc | otherwise = - topDeclElem links loc splice docnames pref2 + topDeclElem links loc splice docName pref2 +++ subArguments pkg qual (ppSubSigLike unicode qual typ argDocs [] sep emptyCtxts) +++ docSection curname pkg qual doc where - curname = getName <$> listToMaybe docnames + curname = getName <$> Maybe.listToMaybe docnames + docName = + case Maybe.listToMaybe docnames of + Nothing -> error "No docnames. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -- | This splits up a type signature along @->@ and adds docs (when they exist) -- to the arguments. @@ -489,11 +493,15 @@ ppSimpleSig -> HsSigType DocNameI -> Html ppSimpleSig links splice unicode qual emptyCtxts loc names typ = - topDeclElem' names $ ppTypeSig True occNames ppTyp unicode + topDeclElem' docName $ ppTypeSig True occNames ppTyp unicode where topDeclElem' = topDeclElem links loc splice ppTyp = ppSigType unicode qual emptyCtxts typ occNames = map getOccName names + docName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -------------------------------------------------------------------------------- @@ -530,13 +538,13 @@ ppFamDecl summary associated links instances fixities loc doc decl splice unicod curname = Just $ getName docname header_ = - topDeclElem links loc splice [docname] $ + topDeclElem links loc splice docname $ ppFamHeader summary associated decl unicode qual <+> ppFixities fixities qual instancesBit | FamilyDecl{fdInfo = ClosedTypeFamily mb_eqns} <- decl , not summary = - subEquations pkg qual $ map (ppFamDeclEqn . unLoc) $ fromMaybe [] mb_eqns + subEquations pkg qual $ map (ppFamDeclEqn . unLoc) $ Maybe.fromMaybe [] mb_eqns | otherwise = ppInstances links (OriginFamily docname) instances splice unicode pkg qual @@ -706,7 +714,7 @@ ppLContextNoArrow c u q h = ppContextNoArrow (unLoc c) u q h ppContextNoArrow :: HsContext DocNameI -> Unicode -> Qualification -> HideEmptyContexts -> Html ppContextNoArrow cxt unicode qual emptyCtxts = - fromMaybe noHtml $ + Maybe.fromMaybe noHtml $ ppContextNoLocsMaybe (map unLoc cxt) unicode qual emptyCtxts ppContextNoLocs :: [HsType DocNameI] -> Unicode -> Qualification -> HideEmptyContexts -> Html @@ -790,9 +798,9 @@ ppShortClassDecl pkg qual = if not (any isUserLSig sigs) && null ats - then (if summary then id else topDeclElem links loc splice [nm]) hdr + then (if summary then id else topDeclElem links loc splice nm) hdr else - (if summary then id else topDeclElem links loc splice [nm]) (hdr <+> keyword "where") + (if summary then id else topDeclElem links loc splice nm) (hdr <+> keyword "where") +++ shortSubDecls False ( [ ppAssocType summary links doc at [] splice unicode pkg qual | at <- ats, let doc = lookupAnySubdoc (unL $ fdLName $ unL at) subdocs @@ -814,8 +822,12 @@ ppShortClassDecl pkg qual | L _ (ClassOpSig _ False lnames typ) <- sigs - , let doc = lookupAnySubdoc (head names) subdocs - names = map unLoc lnames + , let names = map unLoc lnames + subdocName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd + doc = lookupAnySubdoc subdocName subdocs ] -- FIXME: is taking just the first name ok? Is it possible that -- there are different subdocs for different names in a single @@ -876,8 +888,8 @@ ppClassDecl sigs = map unLoc lsigs classheader - | any isUserLSig lsigs = topDeclElem links loc splice [nm] (hdr unicode qual <+> keyword "where" <+> fixs) - | otherwise = topDeclElem links loc splice [nm] (hdr unicode qual <+> fixs) + | any isUserLSig lsigs = topDeclElem links loc splice nm (hdr unicode qual <+> keyword "where" <+> fixs) + | otherwise = topDeclElem links loc splice nm (hdr unicode qual <+> fixs) -- Only the fixity relevant to the class header fixs = ppFixities [f | f@(n, _) <- fixities, n == unLoc lname] qual @@ -890,7 +902,7 @@ ppClassDecl atBit = subAssociatedTypes [ ppAssocType summary links doc at subfixs splice unicode pkg qual - <+> subDefaults (maybeToList defTys) + <+> subDefaults (Maybe.maybeToList defTys) | at <- ats , let name = unLoc . fdLName $ unLoc at doc = lookupAnySubdoc name subdocs @@ -941,7 +953,7 @@ ppClassDecl unicode pkg qual - <+> subDefaults (maybeToList defSigs) + <+> subDefaults (Maybe.maybeToList defSigs) | ClassOpSig _ False lnames typ <- sigs , name <- map unLoc lnames , let doc = lookupAnySubdoc name subdocs @@ -1111,7 +1123,7 @@ ppInstanceAssocTys -> [DocInstance DocNameI] -> [Html] ppInstanceAssocTys links splice unicode qual orphan insts = - maybeToList $ + Maybe.maybeToList $ subTableSrc Nothing qual links True $ zipWith mkInstHead @@ -1137,10 +1149,14 @@ ppInstanceSigs links splice unicode qual sigs = do L _ rtyp = dropWildCards typ -- Instance methods signatures are synified and thus don't have a useful -- SrcSpan value. Use the methods name location instead. - return $ ppSimpleSig links splice unicode qual HideEmptyContexts (getLocA $ head lnames) names rtyp + let lname = + case Maybe.listToMaybe lnames of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd + return $ ppSimpleSig links splice unicode qual HideEmptyContexts (getLocA lname) names rtyp lookupAnySubdoc :: Eq id1 => id1 -> [(id1, DocForDecl id2)] -> DocForDecl id2 -lookupAnySubdoc n = fromMaybe noDocForDecl . lookup n +lookupAnySubdoc n = Maybe.fromMaybe noDocForDecl . lookup n instanceId :: InstOrigin DocName -> Int -> Bool -> InstHead DocNameI -> String instanceId origin no orphan ihd = @@ -1256,7 +1272,7 @@ ppDataDecl ConDeclGADT{} -> False header_ = - topDeclElem links loc splice [docname] $ + topDeclElem links loc splice docname $ ppDataHeader summary dataDecl unicode qual <+> whereBit <+> fix fix = ppFixities (filter (\(n, _) -> n == docname) fixities) qual @@ -1531,7 +1547,10 @@ ppSideBySideField subdocs unicode qual (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc $ head names) subdocs >>= combineDocumentation . fst + mbDoc = lookup (foExt $ unLoc declName) subdocs >>= combineDocumentation . fst + declName = case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd ppShortField :: Bool -> Unicode -> Qualification -> ConDeclField DocNameI -> Html ppShortField summary unicode qual (ConDeclField _ names ltype _) = ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Layout.hs ===================================== @@ -311,9 +311,9 @@ declElem = paragraph ! [theclass "src"] -- a box for top level documented names -- it adds a source and wiki link at the right hand side of the box -topDeclElem :: LinksInfo -> SrcSpan -> Bool -> [DocName] -> Html -> Html -topDeclElem lnks loc splice names html = - declElem << (html <+> (links lnks loc splice Nothing $ head names)) +topDeclElem :: LinksInfo -> SrcSpan -> Bool -> DocName -> Html -> Html +topDeclElem lnks loc splice name html = + declElem << (html <+> links lnks loc splice Nothing name) -- FIXME: is it ok to simply take the first name? ===================================== utils/haddock/haddock-api/src/Haddock/Utils/Json.hs ===================================== @@ -371,10 +371,9 @@ instance FromJSON Char where parseJSONList v = typeMismatch "String" v parseChar :: String -> Parser Char -parseChar t = - if length t == 1 - then pure $ head t - else prependContext "Char" $ fail "expected a string of length 1" +parseChar [c] = pure c +parseChar [] = prependContext "Char" $ fail "expected a string of length 1, got an empty string" +parseChar (_ : _) = prependContext "Char" $ fail "expected a string of length 1, got a longer string" parseRealFloat :: RealFloat a => String -> Value -> Parser a parseRealFloat _ (Number s) = pure $ realToFrac s View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4ccc25bbc3d6f35d22e7f327a1aab9f0456fcbdf...07ba0d216f0a91353db2125b7c44a19219c80213 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4ccc25bbc3d6f35d22e7f327a1aab9f0456fcbdf...07ba0d216f0a91353db2125b7c44a19219c80213 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 13:20:20 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 05 Sep 2024 09:20:20 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 6 commits: Deterministic Uniques in the NCG Message-ID: <66d9b014fac4_ccd296310a0284fe@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: b9fcb459 by Rodrigo Mesquita at 2024-09-05T11:11:18+01:00 Deterministic Uniques in the NCG See Note [Deterministic Uniques in the NCG] UniqDSM det uniques + use in Cmm.Info Now for SRTs SRT generation using deterministic uniq supply Back LabelMap with deterministic UDFM TSAN uniq rename hard Revert "TSAN uniq rename hard" This reverts commit 7ca5ab3036c15f38c6d4cbcb616d415958c6bcda. improvements to uniqdsm UniqDSM ProcPoint CmmLayoutStack UniqDet 90% of cpsTop UniqDSM Major progress in using UniqDSM in CmmToAsm and Ncg backends Fix imports Un-back label map with udfm Revert "Un-back label map with udfm" This reverts commit f5d2e4257214a3f7b7d845651e6662c5babfd6a3. Make UDSM oneshot deriving via state Fix -fllvm hang - - - - - 516769b7 by Rodrigo Mesquita at 2024-09-05T14:10:12+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - beb30330 by Rodrigo Mesquita at 2024-09-05T14:16:42+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - 901a7329 by Rodrigo Mesquita at 2024-09-05T14:16:42+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 5c72dda0 by Rodrigo Mesquita at 2024-09-05T14:16:42+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 8cef1175 by Rodrigo Mesquita at 2024-09-05T14:16:42+01:00 determinism: Rename CmmGroups in generateCgIPEStub Make sure to also deterministically rename the IPE Cmm (as per Note [Renaming uniques deterministically]) to guarantee deterministic objects when IPE information is requested. - - - - - 30 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Reg/Graph.hs - compiler/GHC/CmmToAsm/Reg/Graph/Spill.hs - compiler/GHC/CmmToAsm/Reg/Linear.hs - compiler/GHC/CmmToAsm/Reg/Linear/Base.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f446aa3cdaccc492d3ac1b07593e536b3d9590ef...8cef11755bce1655a1fe8384a8eb8b9897a26a8a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f446aa3cdaccc492d3ac1b07593e536b3d9590ef...8cef11755bce1655a1fe8384a8eb8b9897a26a8a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 14:25:33 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 05 Sep 2024 10:25:33 -0400 Subject: [Git][ghc/ghc][wip/supersven/fix-new-c-warnings] Fix C warnings (#25237) Message-ID: <66d9bf5d1e925_ccd299f5fb0393eb@gitlab.mail> Sven Tennie pushed to branch wip/supersven/fix-new-c-warnings at Glasgow Haskell Compiler / GHC Commits: 93c2a3ac by Sven Tennie at 2024-09-05T16:25:19+02:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 2 changed files: - libraries/ghc-heap/tests/stack_misc_closures_c.c - testsuite/tests/rts/pause-resume/list_threads_and_misc_roots_c.c Changes: ===================================== libraries/ghc-heap/tests/stack_misc_closures_c.c ===================================== @@ -193,7 +193,7 @@ void create_any_bco_frame(Capability *cap, StgStack *stack, StgWord w) { StgWord bcoSizeWords = sizeofW(StgBCO) + sizeofW(StgLargeBitmap) + sizeofW(StgWord); StgBCO *bco = (StgBCO *)allocate(cap, bcoSizeWords); - SET_HDR(bco, &stg_BCO_info, CCS_MAIN); + SET_HDR(bco, (StgInfoTable*) &stg_BCO_info, CCS_MAIN); c->payload[0] = (StgClosure *)bco; bco->size = bcoSizeWords; ===================================== testsuite/tests/rts/pause-resume/list_threads_and_misc_roots_c.c ===================================== @@ -27,10 +27,10 @@ void checkGcRoots(void) rts_listThreads(&collectTSOsCallback, NULL); for (int i = 0; i < tsoCount; i++) { - StgTSO *tso = UNTAG_CLOSURE(tsos[i]); + StgClosure *tso = UNTAG_CLOSURE((StgClosure*) tsos[i]); if (get_itbl(tso)->type != TSO) { - fprintf(stderr, "tso returned a non-TSO type %zu at index %i\n", + fprintf(stderr, "tso returned a non-TSO type %u at index %i\n", tso->header.info->type, i); exit(1); @@ -44,7 +44,7 @@ void checkGcRoots(void) StgClosure *root = UNTAG_CLOSURE(miscRoots[i]); if (get_itbl(root)->type == TSO) { - fprintf(stderr, "rts_listThreads unexpectedly returned an TSO type at index %i (TSO=%zu)\n", i, TSO); + fprintf(stderr, "rts_listThreads unexpectedly returned an TSO type at index %i (TSO=%d)\n", i, TSO); exit(1); } } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/93c2a3ac3bafd879b6819c763b45edab5da58d0d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/93c2a3ac3bafd879b6819c763b45edab5da58d0d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 14:57:49 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 05 Sep 2024 10:57:49 -0400 Subject: [Git][ghc/ghc][master] 3 commits: Silence x-partial in Haddock.Backends.Xhtml Message-ID: <66d9c6edef910_ccd29b67e0c482a@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7 changed files: - hadrian/ghci-multi-cabal.in - hadrian/src/Rules/ToolArgs.hs - utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Layout.hs - utils/haddock/haddock-api/src/Haddock/Utils/Json.hs Changes: ===================================== hadrian/ghci-multi-cabal.in ===================================== @@ -8,6 +8,6 @@ if [[ $(printf "9.4.0\n%s\n" $($RUN_GHC --numeric-version) | sort -uV | head -n set -e export TOOL_OUTPUT=.hadrian_ghci_multi/ghci_args # Replace newlines with spaces, as these otherwise break the ghci invocation on windows. -CABFLAGS=-v0 "hadrian/build-cabal" multi:ghc --build-root=.hadrian_ghci_multi --flavour=ghc-in-ghci $HADRIAN_ARGS +CABFLAGS=-v0 "hadrian/build-cabal" multi --build-root=.hadrian_ghci_multi --flavour=ghc-in-ghci $HADRIAN_ARGS GHC_FLAGS="$GHC_FLAGS $(cat $TOOL_OUTPUT | tr '\n\r' ' ')" $RUN_GHC --interactive $GHC_FLAGS $@ -fno-code -fwrite-interface -O0 +RTS -A128m ===================================== hadrian/src/Rules/ToolArgs.hs ===================================== @@ -16,7 +16,9 @@ import System.Directory (canonicalizePath) import System.Environment (lookupEnv) import qualified Data.Set as Set import Oracles.ModuleFiles +import Oracles.Setting import Utilities +import Data.Version.Extra -- | @tool:@ is used by tooling in order to get the arguments necessary -- to set up a GHC API session which can compile modules from GHC. When @@ -85,7 +87,16 @@ multiSetup pkg_s = do need (srcs ++ gens) let rexp m = ["-reexported-module", m] let hidir = root "interfaces" pkgPath p - writeFile' (resp_file root p) (intercalate "\n" (arg_list + ghcVersion <- ghcVersionStage stage0InTree + let ghc_wired_in = readVersion ghcVersion < makeVersion [9,8,1] + ghc_package_id = "-package-id ghc-" ++ ghcVersion + normalise_ghc = if ghc_wired_in then normalisePackageIds else id + normalisePackageIds :: [String] -> [String] + normalisePackageIds ((isPrefixOf ghc_package_id -> True) : xs) = "-package-id" : "ghc" : xs + normalisePackageIds (x:xs) = x : normalisePackageIds xs + normalisePackageIds [] = [] + + writeFile' (resp_file root p) (intercalate "\n" (normalise_ghc arg_list ++ modules cd ++ concatMap rexp (reexportModules cd) ++ ["-outputdir", hidir])) @@ -150,7 +161,9 @@ toolTargets = [ cabalSyntax , ghcHeap , ghci , ghcPkg -- # executable - -- , haddock -- # depends on ghc library + , haddock -- # depends on ghc library + , haddockApi + , haddockLibrary , hsc2hs -- # executable , hpc , hpcBin -- # executable ===================================== utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs ===================================== @@ -26,20 +26,20 @@ import Data.Foldable (toList) import Data.List (sort) import Data.List.NonEmpty (NonEmpty (..)) import qualified Data.Map as Map -import Data.Maybe +import qualified Data.Maybe as Maybe import GHC hiding (fromMaybeContext) import GHC.Core.Type (Specificity (..)) import GHC.Data.FastString (unpackFS) import GHC.Types.Name (getOccString, nameOccName, tidyNameOcc) import GHC.Types.Name.Occurrence import GHC.Types.Name.Reader (rdrNameOcc) +import GHC.Utils.Ppr hiding (Doc, quote) +import qualified GHC.Utils.Ppr as Pretty import System.Directory import System.FilePath import Prelude hiding ((<>)) import Documentation.Haddock.Markup -import GHC.Utils.Ppr hiding (Doc, quote) -import qualified GHC.Utils.Ppr as Pretty import Haddock.Doc (combineDocumentation) import Haddock.GhcUtils import Haddock.Types @@ -90,7 +90,7 @@ ppLaTeX ppLaTeX title packageStr visible_ifaces odir prologue maybe_style libdir = do createDirectoryIfMissing True odir - when (isNothing maybe_style) $ + when (Maybe.isNothing maybe_style) $ copyFile (libdir "latex" haddockSty) (odir haddockSty) ppLaTeXTop title packageStr odir prologue maybe_style visible_ifaces mapM_ (ppLaTeXModule title odir) visible_ifaces @@ -139,7 +139,7 @@ ppLaTeXTop doctitle packageStr odir prologue maybe_style ifaces = do mods = sort (map (moduleBasename . ifaceMod) ifaces) - filename = odir (fromMaybe "haddock" packageStr <.> "tex") + filename = odir (Maybe.fromMaybe "haddock" packageStr <.> "tex") writeUtf8File filename (show tex) @@ -174,7 +174,7 @@ ppLaTeXModule _title odir iface = do ] description = - (fromMaybe empty . documentationToLaTeX . ifaceRnDoc) iface + (Maybe.fromMaybe empty . documentationToLaTeX . ifaceRnDoc) iface body = processExports exports -- @@ -201,7 +201,7 @@ exportListItem in sep (punctuate comma [leader <+> ppDocBinder name | name <- names]) <> case subdocs of [] -> empty - _ -> parens (sep (punctuate comma (mapMaybe go subdocs))) + _ -> parens (sep (punctuate comma (Maybe.mapMaybe go subdocs))) exportListItem (ExportNoDecl y []) = ppDocBinder y exportListItem (ExportNoDecl y subs) = @@ -368,7 +368,7 @@ ppFamDecl associated doc instances decl unicode = (if null body then Nothing else Just (vcat body)) $$ instancesBit where - body = catMaybes [familyEqns, documentationToLaTeX doc] + body = Maybe.catMaybes [familyEqns, documentationToLaTeX doc] whereBit = case fdInfo (tcdFam decl) of ClosedTypeFamily _ -> keyword "where" @@ -544,7 +544,7 @@ ppTypeOrFunSig typ (doc, argDocs) (pref1, pref2, sep0) unicode text "\\haddockbeginargs" $$ vcat (map (uncurry (<->)) (ppSubSigLike unicode typ argDocs [] sep0)) $$ text "\\end{tabulary}\\par" - $$ fromMaybe empty (documentationToLaTeX doc) + $$ Maybe.fromMaybe empty (documentationToLaTeX doc) -- | This splits up a type signature along @->@ and adds docs (when they exist) -- to the arguments. The output is a list of (leader/seperator, argument and @@ -741,7 +741,7 @@ ppClassDecl hdr = ppClassHdr False lctxt (unLoc lname) ltyvars lfds - body = catMaybes [documentationToLaTeX doc, body_] + body = Maybe.catMaybes [documentationToLaTeX doc, body_] body_ | null lsigs, null ats, null at_defs = Nothing @@ -764,9 +764,13 @@ ppClassDecl | L _ (ClassOpSig _ is_def lnames typ) <- lsigs , let doc | is_def = noDocForDecl - | otherwise = lookupAnySubdoc (head names) subdocs + | otherwise = lookupAnySubdoc firstName subdocs names = map (cleanName . unLoc) lnames leader = if is_def then Just (keyword "default") else Nothing + firstName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd ] -- N.B. taking just the first name is ok. Signatures with multiple -- names are expanded so that each name gets its own signature. @@ -853,7 +857,7 @@ ppDataDecl pats instances subdocs doc dataDecl unicode = where cons = dd_cons (tcdDataDefn dataDecl) - body = catMaybes [doc >>= documentationToLaTeX, constrBit, patternBit] + body = Maybe.catMaybes [doc >>= documentationToLaTeX, constrBit, patternBit] (whereBit, leaders) | null cons @@ -1031,7 +1035,11 @@ ppSideBySideField subdocs unicode (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc $ head names) subdocs >>= fmap _doc . combineDocumentation . fst + mbDoc = lookup (foExt $ unLoc name) subdocs >>= fmap _doc . combineDocumentation . fst + name = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -- | Pretty-print a bundled pattern synonym ppSideBySidePat @@ -1157,7 +1165,7 @@ ppContextNoLocsMaybe cxt unicode = Just $ pp_hs_context cxt unicode ppContextNoArrow :: HsContext DocNameI -> Bool -> LaTeX ppContextNoArrow cxt unicode = - fromMaybe empty $ + Maybe.fromMaybe empty $ ppContextNoLocsMaybe (map unLoc cxt) unicode ppContextNoLocs :: [HsType DocNameI] -> Bool -> LaTeX ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml.hs ===================================== @@ -6,6 +6,7 @@ {-# LANGUAGE TupleSections #-} {-# LANGUAGE TypeApplications #-} {-# LANGUAGE TypeFamilies #-} +{-# OPTIONS_GHC -Wwarn=x-partial #-} -- | -- Module : Haddock.Backends.Html ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs ===================================== @@ -10,7 +10,7 @@ ----------------------------------------------------------------------------- -- | --- Module : Haddock.Backends.Html.Decl +-- Module : Haddock.Backends.Xhtml.Decl -- Copyright : (c) Simon Marlow 2003-2006, -- David Waern 2006-2009, -- Mark Lentczner 2010 @@ -28,7 +28,7 @@ import Data.Foldable (toList) import Data.List (intersperse, sort) import Data.List.NonEmpty (NonEmpty (..)) import qualified Data.Map as Map -import Data.Maybe +import qualified Data.Maybe as Maybe import GHC hiding (LexicalFixity (..), fromMaybeContext) import GHC.Core.Type (Specificity (..)) import GHC.Data.BooleanFormula @@ -279,13 +279,17 @@ ppTypeOrFunSig qual emptyCtxts | summary = pref1 - | Map.null argDocs = topDeclElem links loc splice docnames pref1 +++ docSection curname pkg qual doc + | Map.null argDocs = topDeclElem links loc splice docName pref1 +++ docSection curname pkg qual doc | otherwise = - topDeclElem links loc splice docnames pref2 + topDeclElem links loc splice docName pref2 +++ subArguments pkg qual (ppSubSigLike unicode qual typ argDocs [] sep emptyCtxts) +++ docSection curname pkg qual doc where - curname = getName <$> listToMaybe docnames + curname = getName <$> Maybe.listToMaybe docnames + docName = + case Maybe.listToMaybe docnames of + Nothing -> error "No docnames. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -- | This splits up a type signature along @->@ and adds docs (when they exist) -- to the arguments. @@ -489,11 +493,15 @@ ppSimpleSig -> HsSigType DocNameI -> Html ppSimpleSig links splice unicode qual emptyCtxts loc names typ = - topDeclElem' names $ ppTypeSig True occNames ppTyp unicode + topDeclElem' docName $ ppTypeSig True occNames ppTyp unicode where topDeclElem' = topDeclElem links loc splice ppTyp = ppSigType unicode qual emptyCtxts typ occNames = map getOccName names + docName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -------------------------------------------------------------------------------- @@ -530,13 +538,13 @@ ppFamDecl summary associated links instances fixities loc doc decl splice unicod curname = Just $ getName docname header_ = - topDeclElem links loc splice [docname] $ + topDeclElem links loc splice docname $ ppFamHeader summary associated decl unicode qual <+> ppFixities fixities qual instancesBit | FamilyDecl{fdInfo = ClosedTypeFamily mb_eqns} <- decl , not summary = - subEquations pkg qual $ map (ppFamDeclEqn . unLoc) $ fromMaybe [] mb_eqns + subEquations pkg qual $ map (ppFamDeclEqn . unLoc) $ Maybe.fromMaybe [] mb_eqns | otherwise = ppInstances links (OriginFamily docname) instances splice unicode pkg qual @@ -706,7 +714,7 @@ ppLContextNoArrow c u q h = ppContextNoArrow (unLoc c) u q h ppContextNoArrow :: HsContext DocNameI -> Unicode -> Qualification -> HideEmptyContexts -> Html ppContextNoArrow cxt unicode qual emptyCtxts = - fromMaybe noHtml $ + Maybe.fromMaybe noHtml $ ppContextNoLocsMaybe (map unLoc cxt) unicode qual emptyCtxts ppContextNoLocs :: [HsType DocNameI] -> Unicode -> Qualification -> HideEmptyContexts -> Html @@ -790,9 +798,9 @@ ppShortClassDecl pkg qual = if not (any isUserLSig sigs) && null ats - then (if summary then id else topDeclElem links loc splice [nm]) hdr + then (if summary then id else topDeclElem links loc splice nm) hdr else - (if summary then id else topDeclElem links loc splice [nm]) (hdr <+> keyword "where") + (if summary then id else topDeclElem links loc splice nm) (hdr <+> keyword "where") +++ shortSubDecls False ( [ ppAssocType summary links doc at [] splice unicode pkg qual | at <- ats, let doc = lookupAnySubdoc (unL $ fdLName $ unL at) subdocs @@ -814,8 +822,12 @@ ppShortClassDecl pkg qual | L _ (ClassOpSig _ False lnames typ) <- sigs - , let doc = lookupAnySubdoc (head names) subdocs - names = map unLoc lnames + , let names = map unLoc lnames + subdocName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd + doc = lookupAnySubdoc subdocName subdocs ] -- FIXME: is taking just the first name ok? Is it possible that -- there are different subdocs for different names in a single @@ -876,8 +888,8 @@ ppClassDecl sigs = map unLoc lsigs classheader - | any isUserLSig lsigs = topDeclElem links loc splice [nm] (hdr unicode qual <+> keyword "where" <+> fixs) - | otherwise = topDeclElem links loc splice [nm] (hdr unicode qual <+> fixs) + | any isUserLSig lsigs = topDeclElem links loc splice nm (hdr unicode qual <+> keyword "where" <+> fixs) + | otherwise = topDeclElem links loc splice nm (hdr unicode qual <+> fixs) -- Only the fixity relevant to the class header fixs = ppFixities [f | f@(n, _) <- fixities, n == unLoc lname] qual @@ -890,7 +902,7 @@ ppClassDecl atBit = subAssociatedTypes [ ppAssocType summary links doc at subfixs splice unicode pkg qual - <+> subDefaults (maybeToList defTys) + <+> subDefaults (Maybe.maybeToList defTys) | at <- ats , let name = unLoc . fdLName $ unLoc at doc = lookupAnySubdoc name subdocs @@ -941,7 +953,7 @@ ppClassDecl unicode pkg qual - <+> subDefaults (maybeToList defSigs) + <+> subDefaults (Maybe.maybeToList defSigs) | ClassOpSig _ False lnames typ <- sigs , name <- map unLoc lnames , let doc = lookupAnySubdoc name subdocs @@ -1111,7 +1123,7 @@ ppInstanceAssocTys -> [DocInstance DocNameI] -> [Html] ppInstanceAssocTys links splice unicode qual orphan insts = - maybeToList $ + Maybe.maybeToList $ subTableSrc Nothing qual links True $ zipWith mkInstHead @@ -1137,10 +1149,14 @@ ppInstanceSigs links splice unicode qual sigs = do L _ rtyp = dropWildCards typ -- Instance methods signatures are synified and thus don't have a useful -- SrcSpan value. Use the methods name location instead. - return $ ppSimpleSig links splice unicode qual HideEmptyContexts (getLocA $ head lnames) names rtyp + let lname = + case Maybe.listToMaybe lnames of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd + return $ ppSimpleSig links splice unicode qual HideEmptyContexts (getLocA lname) names rtyp lookupAnySubdoc :: Eq id1 => id1 -> [(id1, DocForDecl id2)] -> DocForDecl id2 -lookupAnySubdoc n = fromMaybe noDocForDecl . lookup n +lookupAnySubdoc n = Maybe.fromMaybe noDocForDecl . lookup n instanceId :: InstOrigin DocName -> Int -> Bool -> InstHead DocNameI -> String instanceId origin no orphan ihd = @@ -1256,7 +1272,7 @@ ppDataDecl ConDeclGADT{} -> False header_ = - topDeclElem links loc splice [docname] $ + topDeclElem links loc splice docname $ ppDataHeader summary dataDecl unicode qual <+> whereBit <+> fix fix = ppFixities (filter (\(n, _) -> n == docname) fixities) qual @@ -1531,7 +1547,10 @@ ppSideBySideField subdocs unicode qual (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc $ head names) subdocs >>= combineDocumentation . fst + mbDoc = lookup (foExt $ unLoc declName) subdocs >>= combineDocumentation . fst + declName = case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd ppShortField :: Bool -> Unicode -> Qualification -> ConDeclField DocNameI -> Html ppShortField summary unicode qual (ConDeclField _ names ltype _) = ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Layout.hs ===================================== @@ -311,9 +311,9 @@ declElem = paragraph ! [theclass "src"] -- a box for top level documented names -- it adds a source and wiki link at the right hand side of the box -topDeclElem :: LinksInfo -> SrcSpan -> Bool -> [DocName] -> Html -> Html -topDeclElem lnks loc splice names html = - declElem << (html <+> (links lnks loc splice Nothing $ head names)) +topDeclElem :: LinksInfo -> SrcSpan -> Bool -> DocName -> Html -> Html +topDeclElem lnks loc splice name html = + declElem << (html <+> links lnks loc splice Nothing name) -- FIXME: is it ok to simply take the first name? ===================================== utils/haddock/haddock-api/src/Haddock/Utils/Json.hs ===================================== @@ -371,10 +371,9 @@ instance FromJSON Char where parseJSONList v = typeMismatch "String" v parseChar :: String -> Parser Char -parseChar t = - if length t == 1 - then pure $ head t - else prependContext "Char" $ fail "expected a string of length 1" +parseChar [c] = pure c +parseChar [] = prependContext "Char" $ fail "expected a string of length 1, got an empty string" +parseChar (_ : _) = prependContext "Char" $ fail "expected a string of length 1, got a longer string" parseRealFloat :: RealFloat a => String -> Value -> Parser a parseRealFloat _ (Number s) = pure $ realToFrac s View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9c354beba3e03f56f9a6345f7607a04b55a3318f...6cac9eb8a598b4954934c64789aa5bdfef5128a7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9c354beba3e03f56f9a6345f7607a04b55a3318f...6cac9eb8a598b4954934c64789aa5bdfef5128a7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 14:58:26 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 05 Sep 2024 10:58:26 -0400 Subject: [Git][ghc/ghc][master] ci: Beef up determinism interface test Message-ID: <66d9c71236f4c_ccd29bbc6dc51469@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 1 changed file: - .gitlab/ci.sh Changes: ===================================== .gitlab/ci.sh ===================================== @@ -753,7 +753,7 @@ function check_interfaces(){ } function abi_test() { - for i in {1..20}; do info "iteration $i"; run_abi_test; done + for i in {1..10}; do info "iteration $i"; run_abi_test; done } function run_abi_test() { @@ -761,8 +761,8 @@ function run_abi_test() { fail "HC not set" fi mkdir -p out - OUT="$PWD/out/run1" DIR=$(mktemp -d XXXX-looooooooong) cabal_abi_test -O0 - OUT="$PWD/out/run2" DIR=$(mktemp -d XXXX-short) cabal_abi_test -O0 + OUT="$PWD/out/run1" DIR=$(mktemp -d XXXX-looooooooong) cabal_abi_test -O1 -haddock + OUT="$PWD/out/run2" DIR=$(mktemp -d XXXX-short) cabal_abi_test -O1 -haddock -dunique-increment=-1 -dinitial-unique=16777215 check_interfaces out/run1 out/run2 abis "Mismatched ABI hash" check_interfaces out/run1 out/run2 interfaces "Mismatched interface hashes" } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7d84df86d8c639a9ef442593d4a8c017a2046b92 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7d84df86d8c639a9ef442593d4a8c017a2046b92 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 15:24:12 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 05 Sep 2024 11:24:12 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 19 commits: rts: fix checkClosure error message Message-ID: <66d9cd1cc1e69_3cb9a2d8e50250ca@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 4d5513cc by Matthew Pickering at 2024-09-05T16:21:46+01:00 Run on test-abi label - - - - - 60afd1a2 by Rodrigo Mesquita at 2024-09-05T16:23:00+01:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - e64683cd by Rodrigo Mesquita at 2024-09-05T16:24:01+01:00 Deterministic Uniques in the NCG See Note [Deterministic Uniques in the NCG] UniqDSM det uniques + use in Cmm.Info Now for SRTs SRT generation using deterministic uniq supply Back LabelMap with deterministic UDFM TSAN uniq rename hard Revert "TSAN uniq rename hard" This reverts commit 7ca5ab3036c15f38c6d4cbcb616d415958c6bcda. improvements to uniqdsm UniqDSM ProcPoint CmmLayoutStack UniqDet 90% of cpsTop UniqDSM Major progress in using UniqDSM in CmmToAsm and Ncg backends Fix imports Un-back label map with udfm Revert "Un-back label map with udfm" This reverts commit f5d2e4257214a3f7b7d845651e6662c5babfd6a3. Make UDSM oneshot deriving via state Fix -fllvm hang - - - - - ea273aaa by Rodrigo Mesquita at 2024-09-05T16:24:01+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - f35cd9d7 by Rodrigo Mesquita at 2024-09-05T16:24:02+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - 2dbda747 by Rodrigo Mesquita at 2024-09-05T16:24:02+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 5306bdf2 by Rodrigo Mesquita at 2024-09-05T16:24:02+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 0e766d8e by Rodrigo Mesquita at 2024-09-05T16:24:02+01:00 determinism: Rename CmmGroups in generateCgIPEStub Make sure to also deterministically rename the IPE Cmm (as per Note [Renaming uniques deterministically]) to guarantee deterministic objects when IPE information is requested. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8cef11755bce1655a1fe8384a8eb8b9897a26a8a...0e766d8eb501a9f37802d119777117e361e09649 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8cef11755bce1655a1fe8384a8eb8b9897a26a8a...0e766d8eb501a9f37802d119777117e361e09649 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 20:19:10 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Thu, 05 Sep 2024 16:19:10 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] Link bytecode from interface-stored core bindings in oneshot mode Message-ID: <66da123e99622_1dd1142b032c197f6@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 966c7100 by Cheng Shao at 2024-09-05T22:11:21+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). Metric Decrease: MultiLayerModules T13701 - - - - - 20 changed files: - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/A.hs - + testsuite/tests/bytecode/T25090/B.hs - + testsuite/tests/bytecode/T25090/C.hs - + testsuite/tests/bytecode/T25090/C.hs-boot - + testsuite/tests/bytecode/T25090/D.hs - + testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/T25090-debug.stderr - + testsuite/tests/bytecode/T25090/T25090.stdout - + testsuite/tests/bytecode/T25090/all.T Changes: ===================================== compiler/GHC/CoreToIface.hs ===================================== @@ -602,7 +602,7 @@ toIfaceTopBind b = in (top_bndr, rhs') -- The sharing behaviour is currently disabled due to #22807, and relies on - -- finished #220056 to be re-enabled. + -- finished #20056 to be re-enabled. disabledDueTo22807 = True already_has_unfolding b = not disabledDueTo22807 @@ -774,8 +774,8 @@ outside of the hs-boot loop. Note [Interface File with Core: Sharing RHSs] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -IMPORTANT: This optimisation is currently disabled due to #22027, it can be - re-enabled once #220056 is implemented. +IMPORTANT: This optimisation is currently disabled due to #22807, it can be + re-enabled once #22056 is implemented. In order to avoid duplicating definitions for bindings which already have unfoldings we do some minor headstands to avoid serialising the RHS of a definition if it has ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -260,7 +260,6 @@ outputForeignStubs Maybe FilePath) -- C file created outputForeignStubs logger tmpfs dflags unit_state mod location stubs = do - let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location stub_c <- newTempName logger tmpfs (tmpDir dflags) TFL_CurrentModule "c" case stubs of @@ -276,8 +275,6 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs stub_h_output_d = pprCode h_code stub_h_output_w = showSDoc dflags stub_h_output_d - createDirectoryIfMissing True (takeDirectory stub_h) - putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export header file" FormatC @@ -299,9 +296,20 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs | platformMisc_libFFI $ platformMisc dflags = "#include \"rts/ghc_ffi.h\"\n" | otherwise = "" - stub_h_file_exists - <- outputForeignStubs_help stub_h stub_h_output_w - ("#include \n" ++ cplusplus_hdr) cplusplus_ftr + stub_h_file_exists <- + if null stub_h_output_w + then pure False + else do + -- The header path is computed from the module source path, which + -- does not exist when loading interface core bindings for Template + -- Haskell. + -- The header is only generated for foreign exports. + -- Since those aren't supported for TH with bytecode, we can skip + -- this here for now. + let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location + createDirectoryIfMissing True (takeDirectory stub_h) + outputForeignStubs_help stub_h stub_h_output_w + ("#include \n" ++ cplusplus_hdr) cplusplus_ftr putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export stubs" FormatC stub_c_output_d ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,6 +49,7 @@ module GHC.Driver.Main , HscBackendAction (..), HscRecompStatus (..) , initModDetails , initWholeCoreBindings + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -105,6 +106,7 @@ module GHC.Driver.Main , showModuleIndex , hscAddSptEntries , writeInterfaceOnlyMode + , loadByteCode ) where import GHC.Prelude @@ -274,7 +276,8 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.List ( nub, isPrefixOf, partition ) +import Data.Functor ((<&>)) +import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad import Data.IORef @@ -970,19 +973,23 @@ loadByteCode iface mod_sum = do (mi_foreign iface) return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) _ -> return $ outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- +add_iface_to_hpt :: ModIface -> ModDetails -> HscEnv -> HscEnv +add_iface_to_hpt iface details = + hscUpdateHPT $ \ hpt -> + addToHpt hpt (moduleName (mi_module iface)) + (HomeModInfo iface details emptyHomeModInfoLinkable) -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] initModDetails :: HscEnv -> ModIface -> IO ModDetails initModDetails hsc_env iface = fixIO $ \details' -> do - let act hpt = addToHpt hpt (moduleName $ mi_module iface) - (HomeModInfo iface details' emptyHomeModInfoLinkable) - let !hsc_env' = hscUpdateHPT act hsc_env + let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful @@ -1010,8 +1017,52 @@ compile_for_interpreter hsc_env use = adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Return an 'IO' that hydrates Core bindings and compiles them to bytecode if +-- the interface contains any, using the supplied type env for typechecking. +-- +-- Unlike 'initWholeCoreBindings', this does not use lazy IO. +-- Instead, the 'IO' is only evaluated (in @get_link_deps@) when it is clear +-- that it will be used immediately (because we're linking TH with +-- @-fprefer-byte-code@ in oneshot mode), and the result is cached in +-- 'LoaderState'. +-- +-- 'initWholeCoreBindings' needs the laziness because it is used to populate +-- 'HomeModInfo', which is done preemptively, in anticipation of downstream +-- modules using the bytecode for TH in make mode, which might never happen. +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + compile <$> iface_core_bindings iface location + where + compile decls = do + (bcos, fos) <- compileWholeCoreBindings hsc_env type_env decls + linkable $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time (mi_module iface) parts + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects. +-- them with a lazy IO thunk that compiles them to bytecode and foreign objects, +-- using the supplied environment for type checking. -- -- The laziness is necessary because this value is stored purely in a -- 'HomeModLinkable' in the home package table, rather than some dedicated @@ -1025,29 +1076,71 @@ compile_for_interpreter hsc_env use = -- -- This is sound because generateByteCode just depends on things already loaded -- in the interface file. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env mod_iface details (Linkable utc_time this_mod uls) = +initWholeCoreBindings :: + HscEnv -> + ModIface -> + ModDetails -> + Linkable -> + IO Linkable +initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = Linkable utc_time this_mod <$> mapM go uls where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef (md_types details) - let act hpt = addToHpt hpt (moduleName $ mi_module mod_iface) - (HomeModInfo mod_iface details emptyHomeModInfoLinkable) - kv = knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - hsc_env' = hscUpdateHPT act hsc_env { hsc_type_env_vars = kv } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") hsc_env' $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons (md_types details)) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + go = \case + CoreBindings wcb -> do + ~(bco, fos) <- unsafeInterleaveIO $ + compileWholeCoreBindings hsc_env' type_env wcb + pure (LazyBCOs bco fos) + l -> pure l + + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written or were unavailable due to boot import +-- cycles, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +compileWholeCoreBindings :: + HscEnv -> + TypeEnv -> + WholeCoreBindings -> + IO (CompiledByteCode, [FilePath]) +compileWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + gen_bytecode core_binds stubs foreign_files + where + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + generateByteCode hsc_env cgi_guts wcb_mod_location + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -0,0 +1,15 @@ +module GHC.Driver.Main where + +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) + +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1310,8 +1310,10 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- MP: This is a bit janky, because before you add the entries you have to extend the HPT with the module -- you just compiled. Another option, would be delay adding anything until after upsweep has finished, but I -- am unsure if this is sound (wrt running TH splices for example). - -- This function only does anything if the linkable produced is a BCO, which only happens with the - -- bytecode backend, no need to guard against the backend type additionally. + -- This function only does anything if the linkable produced is a BCO, which + -- used to only happen with the bytecode backend, but with + -- @-fprefer-byte-code@, @HomeModInfo@ has bytecode even when generating + -- object code, see #25230. addSptEntries (hscUpdateHPT (\hpt -> addToHpt hpt (ms_mod_name summary) hmi) hsc_env) (homeModInfoByteCode hmi) @@ -2985,7 +2987,7 @@ which can be checked easily using ghc-debug. a reference to the entire HscEnv, if we are not careful the HscEnv will contain the HomePackageTable at the time the interface was loaded and it will never be released. - Where? dontLeakTheHPT in GHC.Iface.Load + Where? dontLeakTheHUG in GHC.Iface.Load 2. No KnotVars are live at the end of upsweep (#20491) Why? KnotVars contains an old stale reference to the TypeEnv for modules ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,12 +506,14 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) ; new_eps_rules <- tcIfaceRules ignore_prags (mi_rules iface) ; new_eps_anns <- tcIfaceAnnotations (mi_anns iface) ; new_eps_complete_matches <- tcIfaceCompleteMatches (mi_complete_matches iface) + ; purged_hsc_env <- getTopEnv ; let final_iface = iface & set_mi_decls (panic "No mi_decls in PIT") @@ -518,13 +521,27 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + | prefer_bytecode + , Just action <- loadIfaceByteCode purged_hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have 'extra_decls' + -- so 'get_link_deps' knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,6 +553,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -569,7 +587,7 @@ loadInterface doc_str mod from {- Note [Loading your own hi-boot file] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Generally speaking, when compiling module M, we should not -load M.hi boot into the EPS. After all, we are very shortly +load M.hi-boot into the EPS. After all, we are very shortly going to have full information about M. Moreover, see Note [Do not update EPS with your own hi-boot] in GHC.Iface.Recomp. @@ -698,7 +716,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -845,7 +863,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -875,7 +893,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -900,7 +918,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -905,11 +905,11 @@ tcTopIfaceBindings :: IORef TypeEnv -> [IfaceBindingX IfaceMaybeRhs IfaceTopBndr -> IfL [CoreBind] tcTopIfaceBindings ty_var ver_decls = do - int <- mapM tcTopBinders ver_decls + int <- mapM tcTopBinders ver_decls let all_ids :: [Id] = concatMap toList int liftIO $ modifyIORef ty_var (flip extendTypeEnvList (map AnId all_ids)) - extendIfaceIdEnv all_ids $ mapM (tc_iface_bindings) int + extendIfaceIdEnv all_ids $ mapM tc_iface_bindings int tcTopBinders :: IfaceBindingX a IfaceTopBndrInfo -> IfL (IfaceBindingX a Id) tcTopBinders = traverse mk_top_id ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -60,16 +60,16 @@ import System.Directory data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -275,21 +275,21 @@ get_link_deps opts pls maybe_normal_osuf span mods = do case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod = do { - -- ...and then find the linkable for it - mb_lnk <- findObjectLinkableMaybe mod loc ; - case mb_lnk of { - Nothing -> no_obj mod ; - Just lnk -> adjust_linkable lnk - }} + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,18 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -68,6 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -139,6 +142,12 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), + eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules eps_fam_inst_env :: !PackageFamInstEnv,-- ^ The total 'FamInstEnv' accumulated ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -28,12 +28,12 @@ import System.FilePath (takeExtension) {- Note [Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A interface file can optionally contain the definitions of all core bindings, this is enabled by the flag `-fwrite-if-simplified-core`. This provides everything needed in addition to the normal ModIface and ModDetails -to restart compilation after typechecking to generate bytecode. The `fi_bindings` field +to restart compilation after typechecking to generate bytecode. The `wcb_bindings` field is stored in the normal interface file and the other fields populated whilst loading the interface file. @@ -62,8 +62,50 @@ after whatever simplification the user requested has been performed. So the simp of the interface file agree with the optimisation level as reported by the interface file. +The lifecycle differs beyond laziness depending on the provenance of a module. +In all cases, the main consumer for interface bytecode is 'get_link_deps', which +traverses a splice's or GHCi expression's dependencies and collects the needed +build artifacts, which can be objects or bytecode, depending on the build +settings. + +1. In make mode, all eligible modules are part of the dependency graph. + Their interfaces are loaded unconditionally and in dependency order by the + compilation manager, and each module's bytecode is prepared before its + dependents are compiled, in one of two ways: + + - If the interface file for a module is missing or out of sync with its + source, it is recompiled and bytecode is generated directly and + immediately, not involving 'WholeCoreBindings' (in 'runHscBackendPhase'). + + - If the interface file is up to date, no compilation is performed, and a + lazy thunk generating bytecode from interface Core bindings is created in + 'compileOne'', which will only be compiled if a downstream module contains + a splice that depends on it, as described above. + + In both cases, the bytecode 'Linkable' is stored in a 'HomeModLinkable' in + the Home Unit Graph, lazy or not. + +2. In oneshot mode, which compiles individual modules without a shared home unit + graph, a previously compiled module is not reprocessed as described for make + mode above. + When 'get_link_deps' encounters a dependency on a local module, it requests + its bytecode from the External Package State, who loads the interface + on-demand. + + Since the EPS stores interfaces for all package dependencies in addition to + local modules in oneshot mode, it has a substantial memory footprint. + We try to curtail that by extracting important data into specialized fields + in the EPS, and retaining only a few fields of 'ModIface' by overwriting the + others with bottom values. + + In order to avoid keeping around all of the interface's components needed for + compiling bytecode, we instead store an IO action in 'eps_iface_bytecode'. + When 'get_link_deps' evaluates this action, the result is not retained in the + EPS, but stored in 'LoaderState', where it may eventually get evicted to free + up the memory. + Note [Size of Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ How much overhead does `-fwrite-if-simplified-core` add to a typical interface file? As an experiment I compiled the `Cabal` library and `ghc` library (Aug 22) with ===================================== testsuite/tests/bytecode/T25090/A.hs ===================================== @@ -0,0 +1,7 @@ +{-# language TemplateHaskell #-} +module Main where + +import D + +main :: IO () +main = putStrLn (show ($splc :: Int)) ===================================== testsuite/tests/bytecode/T25090/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} C (C) + +data B = B C ===================================== testsuite/tests/bytecode/T25090/C.hs ===================================== @@ -0,0 +1,8 @@ +module C where + +import B + +data C = C Int + +b :: B +b = B (C 2024) ===================================== testsuite/tests/bytecode/T25090/C.hs-boot ===================================== @@ -0,0 +1,3 @@ +module C where + +data C ===================================== testsuite/tests/bytecode/T25090/D.hs ===================================== @@ -0,0 +1,12 @@ +module D where + +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +import B +import C + +splc :: ExpQ +splc = + lift @_ @Int num + where + B (C num) = b ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -0,0 +1,19 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +T25090a: + $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot + $(TEST_HC) -c -fbyte-code-and-object-code B.hs + $(TEST_HC) -c -fbyte-code-and-object-code C.hs + $(TEST_HC) -c -fbyte-code-and-object-code D.hs + echo "bad" > B.o + echo "bad" > C.o + echo "bad" > C.o-boot + $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code D.o A.o -o exe + ./exe + +T25090b: + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 + ./exe ===================================== testsuite/tests/bytecode/T25090/T25090-debug.stderr ===================================== @@ -0,0 +1,6 @@ +WARNING: + loadInterface + C + Call stack: + CallStack (from HasCallStack): + warnPprTrace, called at compiler/GHC/Iface/Load.hs:: in :GHC.Iface.Load ===================================== testsuite/tests/bytecode/T25090/T25090.stdout ===================================== @@ -0,0 +1 @@ +2024 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -0,0 +1,19 @@ +# This test compiles the boot file separately from its source file, which causes +# a debug assertion warning. +# Since this appears to be intentional according to the Note [Loading your own hi-boot file], +# the warning is added to the expected stderr for debugged builds. +def test_T25090(name): + assert_warn_spec = {'stderr': 'T25090-debug.stderr'} + extra_specs = assert_warn_spec if name == 'T25090a' and compiler_debugged() else {} + return test(name, + [extra_files(['A.hs', 'B.hs', 'C.hs-boot', 'C.hs', 'D.hs']), + req_th, + js_skip, + use_specs(dict(stdout = 'T25090.stdout', **extra_specs)), + ], + makefile_test, + []) + +test_T25090('T25090a') + +test_T25090('T25090b') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/966c71003e0339384f993a41cfad97ff2198287b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/966c71003e0339384f993a41cfad97ff2198287b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 20:30:17 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Thu, 05 Sep 2024 16:30:17 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] 12 commits: rts: fix checkClosure error message Message-ID: <66da14d98e55c_1dd1144b19f021731@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 55cf9c14 by Cheng Shao at 2024-09-05T22:29:56+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). Metric Decrease: MultiLayerModules T13701 - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Utils/Env.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - compiler/GHC/Utils/TmpFs.hs - hadrian/ghci-multi-cabal.in - hadrian/src/Rules/ToolArgs.hs - rts/sm/Sanity.c - testsuite/driver/perf_notes.py - testsuite/driver/testglobals.py - testsuite/driver/testlib.py - testsuite/driver/testutil.py - + testsuite/tests/bytecode/T25090/A.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/966c71003e0339384f993a41cfad97ff2198287b...55cf9c141df43c2355789db62c0e15f053e890b1 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/966c71003e0339384f993a41cfad97ff2198287b...55cf9c141df43c2355789db62c0e15f053e890b1 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 5 22:50:41 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Thu, 05 Sep 2024 18:50:41 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] Link bytecode from interface-stored core bindings in oneshot mode Message-ID: <66da35c1a0cb7_1dd114a66760305ef@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: a14b588f by Cheng Shao at 2024-09-06T00:50:29+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). Metric Decrease: MultiLayerModules T13701 - - - - - 20 changed files: - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/A.hs - + testsuite/tests/bytecode/T25090/B.hs - + testsuite/tests/bytecode/T25090/C.hs - + testsuite/tests/bytecode/T25090/C.hs-boot - + testsuite/tests/bytecode/T25090/D.hs - + testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/T25090-debug.stderr - + testsuite/tests/bytecode/T25090/T25090.stdout - + testsuite/tests/bytecode/T25090/all.T Changes: ===================================== compiler/GHC/CoreToIface.hs ===================================== @@ -602,7 +602,7 @@ toIfaceTopBind b = in (top_bndr, rhs') -- The sharing behaviour is currently disabled due to #22807, and relies on - -- finished #220056 to be re-enabled. + -- finished #20056 to be re-enabled. disabledDueTo22807 = True already_has_unfolding b = not disabledDueTo22807 @@ -774,8 +774,8 @@ outside of the hs-boot loop. Note [Interface File with Core: Sharing RHSs] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -IMPORTANT: This optimisation is currently disabled due to #22027, it can be - re-enabled once #220056 is implemented. +IMPORTANT: This optimisation is currently disabled due to #22807, it can be + re-enabled once #22056 is implemented. In order to avoid duplicating definitions for bindings which already have unfoldings we do some minor headstands to avoid serialising the RHS of a definition if it has ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -260,7 +260,6 @@ outputForeignStubs Maybe FilePath) -- C file created outputForeignStubs logger tmpfs dflags unit_state mod location stubs = do - let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location stub_c <- newTempName logger tmpfs (tmpDir dflags) TFL_CurrentModule "c" case stubs of @@ -276,8 +275,6 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs stub_h_output_d = pprCode h_code stub_h_output_w = showSDoc dflags stub_h_output_d - createDirectoryIfMissing True (takeDirectory stub_h) - putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export header file" FormatC @@ -299,9 +296,20 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs | platformMisc_libFFI $ platformMisc dflags = "#include \"rts/ghc_ffi.h\"\n" | otherwise = "" - stub_h_file_exists - <- outputForeignStubs_help stub_h stub_h_output_w - ("#include \n" ++ cplusplus_hdr) cplusplus_ftr + stub_h_file_exists <- + if null stub_h_output_w + then pure False + else do + -- The header path is computed from the module source path, which + -- does not exist when loading interface core bindings for Template + -- Haskell. + -- The header is only generated for foreign exports. + -- Since those aren't supported for TH with bytecode, we can skip + -- this here for now. + let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location + createDirectoryIfMissing True (takeDirectory stub_h) + outputForeignStubs_help stub_h stub_h_output_w + ("#include \n" ++ cplusplus_hdr) cplusplus_ftr putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export stubs" FormatC stub_c_output_d ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,6 +49,7 @@ module GHC.Driver.Main , HscBackendAction (..), HscRecompStatus (..) , initModDetails , initWholeCoreBindings + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -105,6 +106,7 @@ module GHC.Driver.Main , showModuleIndex , hscAddSptEntries , writeInterfaceOnlyMode + , loadByteCode ) where import GHC.Prelude @@ -274,7 +276,8 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.List ( nub, isPrefixOf, partition ) +import Data.Functor ((<&>)) +import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad import Data.IORef @@ -970,19 +973,23 @@ loadByteCode iface mod_sum = do (mi_foreign iface) return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) _ -> return $ outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- +add_iface_to_hpt :: ModIface -> ModDetails -> HscEnv -> HscEnv +add_iface_to_hpt iface details = + hscUpdateHPT $ \ hpt -> + addToHpt hpt (moduleName (mi_module iface)) + (HomeModInfo iface details emptyHomeModInfoLinkable) -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] initModDetails :: HscEnv -> ModIface -> IO ModDetails initModDetails hsc_env iface = fixIO $ \details' -> do - let act hpt = addToHpt hpt (moduleName $ mi_module iface) - (HomeModInfo iface details' emptyHomeModInfoLinkable) - let !hsc_env' = hscUpdateHPT act hsc_env + let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful @@ -1010,8 +1017,52 @@ compile_for_interpreter hsc_env use = adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Return an 'IO' that hydrates Core bindings and compiles them to bytecode if +-- the interface contains any, using the supplied type env for typechecking. +-- +-- Unlike 'initWholeCoreBindings', this does not use lazy IO. +-- Instead, the 'IO' is only evaluated (in @get_link_deps@) when it is clear +-- that it will be used immediately (because we're linking TH with +-- @-fprefer-byte-code@ in oneshot mode), and the result is cached in +-- 'LoaderState'. +-- +-- 'initWholeCoreBindings' needs the laziness because it is used to populate +-- 'HomeModInfo', which is done preemptively, in anticipation of downstream +-- modules using the bytecode for TH in make mode, which might never happen. +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + compile <$> iface_core_bindings iface location + where + compile decls = do + (bcos, fos) <- compileWholeCoreBindings hsc_env type_env decls + linkable $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time (mi_module iface) parts + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects. +-- them with a lazy IO thunk that compiles them to bytecode and foreign objects, +-- using the supplied environment for type checking. -- -- The laziness is necessary because this value is stored purely in a -- 'HomeModLinkable' in the home package table, rather than some dedicated @@ -1025,29 +1076,71 @@ compile_for_interpreter hsc_env use = -- -- This is sound because generateByteCode just depends on things already loaded -- in the interface file. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env mod_iface details (Linkable utc_time this_mod uls) = +initWholeCoreBindings :: + HscEnv -> + ModIface -> + ModDetails -> + Linkable -> + IO Linkable +initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = Linkable utc_time this_mod <$> mapM go uls where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef (md_types details) - let act hpt = addToHpt hpt (moduleName $ mi_module mod_iface) - (HomeModInfo mod_iface details emptyHomeModInfoLinkable) - kv = knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - hsc_env' = hscUpdateHPT act hsc_env { hsc_type_env_vars = kv } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") hsc_env' $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons (md_types details)) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + go = \case + CoreBindings wcb -> do + ~(bco, fos) <- unsafeInterleaveIO $ + compileWholeCoreBindings hsc_env' type_env wcb + pure (LazyBCOs bco fos) + l -> pure l + + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written or were unavailable due to boot import +-- cycles, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +compileWholeCoreBindings :: + HscEnv -> + TypeEnv -> + WholeCoreBindings -> + IO (CompiledByteCode, [FilePath]) +compileWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + gen_bytecode core_binds stubs foreign_files + where + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + generateByteCode hsc_env cgi_guts wcb_mod_location + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -0,0 +1,15 @@ +module GHC.Driver.Main where + +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) + +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1310,8 +1310,10 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- MP: This is a bit janky, because before you add the entries you have to extend the HPT with the module -- you just compiled. Another option, would be delay adding anything until after upsweep has finished, but I -- am unsure if this is sound (wrt running TH splices for example). - -- This function only does anything if the linkable produced is a BCO, which only happens with the - -- bytecode backend, no need to guard against the backend type additionally. + -- This function only does anything if the linkable produced is a BCO, which + -- used to only happen with the bytecode backend, but with + -- @-fprefer-byte-code@, @HomeModInfo@ has bytecode even when generating + -- object code, see #25230. addSptEntries (hscUpdateHPT (\hpt -> addToHpt hpt (ms_mod_name summary) hmi) hsc_env) (homeModInfoByteCode hmi) @@ -2990,7 +2992,7 @@ which can be checked easily using ghc-debug. a reference to the entire HscEnv, if we are not careful the HscEnv will contain the HomePackageTable at the time the interface was loaded and it will never be released. - Where? dontLeakTheHPT in GHC.Iface.Load + Where? dontLeakTheHUG in GHC.Iface.Load 2. No KnotVars are live at the end of upsweep (#20491) Why? KnotVars contains an old stale reference to the TypeEnv for modules ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,12 +506,14 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) ; new_eps_rules <- tcIfaceRules ignore_prags (mi_rules iface) ; new_eps_anns <- tcIfaceAnnotations (mi_anns iface) ; new_eps_complete_matches <- tcIfaceCompleteMatches (mi_complete_matches iface) + ; purged_hsc_env <- getTopEnv ; let final_iface = iface & set_mi_decls (panic "No mi_decls in PIT") @@ -518,13 +521,27 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + | prefer_bytecode + , Just action <- loadIfaceByteCode purged_hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have 'extra_decls' + -- so 'get_link_deps' knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,6 +553,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -569,7 +587,7 @@ loadInterface doc_str mod from {- Note [Loading your own hi-boot file] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Generally speaking, when compiling module M, we should not -load M.hi boot into the EPS. After all, we are very shortly +load M.hi-boot into the EPS. After all, we are very shortly going to have full information about M. Moreover, see Note [Do not update EPS with your own hi-boot] in GHC.Iface.Recomp. @@ -698,7 +716,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -845,7 +863,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -875,7 +893,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -900,7 +918,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -905,11 +905,11 @@ tcTopIfaceBindings :: IORef TypeEnv -> [IfaceBindingX IfaceMaybeRhs IfaceTopBndr -> IfL [CoreBind] tcTopIfaceBindings ty_var ver_decls = do - int <- mapM tcTopBinders ver_decls + int <- mapM tcTopBinders ver_decls let all_ids :: [Id] = concatMap toList int liftIO $ modifyIORef ty_var (flip extendTypeEnvList (map AnId all_ids)) - extendIfaceIdEnv all_ids $ mapM (tc_iface_bindings) int + extendIfaceIdEnv all_ids $ mapM tc_iface_bindings int tcTopBinders :: IfaceBindingX a IfaceTopBndrInfo -> IfL (IfaceBindingX a Id) tcTopBinders = traverse mk_top_id ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -60,16 +60,16 @@ import System.Directory data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -275,21 +275,21 @@ get_link_deps opts pls maybe_normal_osuf span mods = do case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod = do { - -- ...and then find the linkable for it - mb_lnk <- findObjectLinkableMaybe mod loc ; - case mb_lnk of { - Nothing -> no_obj mod ; - Just lnk -> adjust_linkable lnk - }} + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,18 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -68,6 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -139,6 +142,12 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), + eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules eps_fam_inst_env :: !PackageFamInstEnv,-- ^ The total 'FamInstEnv' accumulated ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -28,12 +28,12 @@ import System.FilePath (takeExtension) {- Note [Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A interface file can optionally contain the definitions of all core bindings, this is enabled by the flag `-fwrite-if-simplified-core`. This provides everything needed in addition to the normal ModIface and ModDetails -to restart compilation after typechecking to generate bytecode. The `fi_bindings` field +to restart compilation after typechecking to generate bytecode. The `wcb_bindings` field is stored in the normal interface file and the other fields populated whilst loading the interface file. @@ -62,8 +62,50 @@ after whatever simplification the user requested has been performed. So the simp of the interface file agree with the optimisation level as reported by the interface file. +The lifecycle differs beyond laziness depending on the provenance of a module. +In all cases, the main consumer for interface bytecode is 'get_link_deps', which +traverses a splice's or GHCi expression's dependencies and collects the needed +build artifacts, which can be objects or bytecode, depending on the build +settings. + +1. In make mode, all eligible modules are part of the dependency graph. + Their interfaces are loaded unconditionally and in dependency order by the + compilation manager, and each module's bytecode is prepared before its + dependents are compiled, in one of two ways: + + - If the interface file for a module is missing or out of sync with its + source, it is recompiled and bytecode is generated directly and + immediately, not involving 'WholeCoreBindings' (in 'runHscBackendPhase'). + + - If the interface file is up to date, no compilation is performed, and a + lazy thunk generating bytecode from interface Core bindings is created in + 'compileOne'', which will only be compiled if a downstream module contains + a splice that depends on it, as described above. + + In both cases, the bytecode 'Linkable' is stored in a 'HomeModLinkable' in + the Home Unit Graph, lazy or not. + +2. In oneshot mode, which compiles individual modules without a shared home unit + graph, a previously compiled module is not reprocessed as described for make + mode above. + When 'get_link_deps' encounters a dependency on a local module, it requests + its bytecode from the External Package State, who loads the interface + on-demand. + + Since the EPS stores interfaces for all package dependencies in addition to + local modules in oneshot mode, it has a substantial memory footprint. + We try to curtail that by extracting important data into specialized fields + in the EPS, and retaining only a few fields of 'ModIface' by overwriting the + others with bottom values. + + In order to avoid keeping around all of the interface's components needed for + compiling bytecode, we instead store an IO action in 'eps_iface_bytecode'. + When 'get_link_deps' evaluates this action, the result is not retained in the + EPS, but stored in 'LoaderState', where it may eventually get evicted to free + up the memory. + Note [Size of Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ How much overhead does `-fwrite-if-simplified-core` add to a typical interface file? As an experiment I compiled the `Cabal` library and `ghc` library (Aug 22) with ===================================== testsuite/tests/bytecode/T25090/A.hs ===================================== @@ -0,0 +1,7 @@ +{-# language TemplateHaskell #-} +module Main where + +import D + +main :: IO () +main = putStrLn (show ($splc :: Int)) ===================================== testsuite/tests/bytecode/T25090/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} C (C) + +data B = B C ===================================== testsuite/tests/bytecode/T25090/C.hs ===================================== @@ -0,0 +1,8 @@ +module C where + +import B + +data C = C Int + +b :: B +b = B (C 2024) ===================================== testsuite/tests/bytecode/T25090/C.hs-boot ===================================== @@ -0,0 +1,3 @@ +module C where + +data C ===================================== testsuite/tests/bytecode/T25090/D.hs ===================================== @@ -0,0 +1,12 @@ +module D where + +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +import B +import C + +splc :: ExpQ +splc = + lift @_ @Int num + where + B (C num) = b ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -0,0 +1,16 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +T25090a: + $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot + $(TEST_HC) -c -fbyte-code-and-object-code B.hs + $(TEST_HC) -c -fbyte-code-and-object-code C.hs + $(TEST_HC) -c -fbyte-code-and-object-code D.hs + $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code D.o C.o B.o A.o -o exe + ./exe + +T25090b: + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 + ./exe ===================================== testsuite/tests/bytecode/T25090/T25090-debug.stderr ===================================== @@ -0,0 +1,6 @@ +WARNING: + loadInterface + C + Call stack: + CallStack (from HasCallStack): + warnPprTrace, called at compiler/GHC/Iface/Load.hs:: in :GHC.Iface.Load ===================================== testsuite/tests/bytecode/T25090/T25090.stdout ===================================== @@ -0,0 +1 @@ +2024 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -0,0 +1,19 @@ +# This test compiles the boot file separately from its source file, which causes +# a debug assertion warning. +# Since this appears to be intentional according to the Note [Loading your own hi-boot file], +# the warning is added to the expected stderr for debugged builds. +def test_T25090(name): + assert_warn_spec = {'stderr': 'T25090-debug.stderr'} + extra_specs = assert_warn_spec if name == 'T25090a' and compiler_debugged() else {} + return test(name, + [extra_files(['A.hs', 'B.hs', 'C.hs-boot', 'C.hs', 'D.hs']), + req_th, + js_skip, + use_specs(dict(stdout = 'T25090.stdout', **extra_specs)), + ], + makefile_test, + []) + +test_T25090('T25090a') + +test_T25090('T25090b') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a14b588f20e2bd5821e9af671ea72790c27054d0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a14b588f20e2bd5821e9af671ea72790c27054d0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 05:47:31 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 06 Sep 2024 01:47:31 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/supersven/ghc-master-riscv-ncg_SAVE Message-ID: <66da97732e05c_1dd11418dfc9c350c9@gitlab.mail> Sven Tennie pushed new branch wip/supersven/ghc-master-riscv-ncg_SAVE at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/supersven/ghc-master-riscv-ncg_SAVE You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 07:21:27 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 06 Sep 2024 03:21:27 -0400 Subject: [Git][ghc/ghc][wip/supersven/AArch64_takeRegRegMoveInst] 4672 commits: [haddock @ 2002-04-04 16:23:43 by simonmar] Message-ID: <66daad77c865_256ec7159a7849179@gitlab.mail> Sven Tennie pushed to branch wip/supersven/AArch64_takeRegRegMoveInst at Glasgow Haskell Compiler / GHC Commits: 2b39cd94 by Simon Marlow at 2002-04-04T16:23:43+00:00 [haddock @ 2002-04-04 16:23:43 by simonmar] This is Haddock, my stab at a Haskell documentation tool. It's not quite ready for release yet, but I'm putting it in the repository so others can take a look. It uses a locally modified version of the hssource parser, extended with support for GHC extensions and documentation annotations. - - - - - 99ede94f by Simon Marlow at 2002-04-04T16:24:10+00:00 [haddock @ 2002-04-04 16:24:10 by simonmar] forgot one file - - - - - 8363294c by Simon Marlow at 2002-04-05T13:58:15+00:00 [haddock @ 2002-04-05 13:58:15 by simonmar] Remap names in the exported declarations to be "closer" to the current module. eg. if an exported declaration mentions a type 'T' which is imported from module A then re-exported from the current module, then links from the type or indeed the documentation will point to the current module rather than module A. This is to support better hiding: module A won't be referred to in the generated output. - - - - - 1570cbc1 by Simon Marlow at 2002-04-05T13:58:23+00:00 [haddock @ 2002-04-05 13:58:23 by simonmar] update the TODO list - - - - - 3a62f96b by Simon Marlow at 2002-04-05T14:11:51+00:00 [haddock @ 2002-04-05 14:11:51 by simonmar] Fix the anchor for a class declaration - - - - - c5d9a471 by Simon Marlow at 2002-04-05T14:18:41+00:00 [haddock @ 2002-04-05 14:18:41 by simonmar] remove underlines on visited links - - - - - 97280525 by Simon Marlow at 2002-04-05T16:11:47+00:00 [haddock @ 2002-04-05 16:11:47 by simonmar] - Update to generate more correct HTML. - Use our own non-overloaded table combinators, as the overloaded versions were giving me a headache. The improved type safety caught several errors in the HTML generation. - - - - - 9acd3a4d by Simon Marlow at 2002-04-05T16:32:19+00:00 [haddock @ 2002-04-05 16:32:19 by simonmar] Add width property to the title, and add TD.children for the module contents page. - - - - - ec9a0847 by Simon Marlow at 2002-04-08T16:39:56+00:00 [haddock @ 2002-04-08 16:39:56 by simonmar] Fix a problem with exports of the form T(..). - - - - - e4627dc8 by Simon Marlow at 2002-04-08T16:41:38+00:00 [haddock @ 2002-04-08 16:41:37 by simonmar] - Add our own versions of Html & BlockTable for the time being. - Add support for generating an index to the HTML backend - - - - - 2d73fd75 by Simon Marlow at 2002-04-09T11:23:24+00:00 [haddock @ 2002-04-09 11:23:24 by simonmar] Add '-- /' as a synonym for '-- |', for compatibility with IDoc. - - - - - 3675464e by Simon Marlow at 2002-04-09T11:33:55+00:00 [haddock @ 2002-04-09 11:33:54 by simonmar] - add the <...> syntax for marking up URLs in documentation - Make the output for data & class declarations more compact when there aren't any documentation annotations on the individual methods or constructors respectively. - - - - - 5077f5b1 by Simon Marlow at 2002-04-09T11:36:04+00:00 [haddock @ 2002-04-09 11:36:04 by simonmar] Update the TODO list - - - - - 9e83c54d by Simon Marlow at 2002-04-10T10:50:06+00:00 [haddock @ 2002-04-10 10:50:06 by simonmar] Use explicit 'px' suffix on pixel sizes; IE seems to prefer them - - - - - 052de51c by Simon Marlow at 2002-04-10T13:23:13+00:00 [haddock @ 2002-04-10 13:23:13 by simonmar] Lex URLs as a single token to avoid having to escape special characters inside the URL string. - - - - - 47187edb by Simon Marlow at 2002-04-10T13:23:55+00:00 [haddock @ 2002-04-10 13:23:55 by simonmar] Not sure why I made the constructor name for a record declaration into a TyCls name, but change it back into a Var name anyhow. - - - - - 3dc6aa81 by Simon Marlow at 2002-04-10T13:26:10+00:00 [haddock @ 2002-04-10 13:26:09 by simonmar] Lots of changes, including: - add index support to the HTML backend - clean up the renamer, put it into a monad - propogate unresolved names to the top level and report them in a nicer way - various bugfixes - - - - - c2a70a72 by Simon Marlow at 2002-04-10T13:32:39+00:00 [haddock @ 2002-04-10 13:32:39 by simonmar] Skeleton documentation - - - - - 50c98d17 by Simon Marlow at 2002-04-10T13:37:23+00:00 [haddock @ 2002-04-10 13:37:23 by simonmar] Update the TODO list, separate into pre-1.0 and post-1.0 items - - - - - f3778be6 by Simon Marlow at 2002-04-10T14:30:58+00:00 [haddock @ 2002-04-10 14:30:58 by simonmar] Add an introduction - - - - - cfbaf9f7 by Simon Marlow at 2002-04-10T14:59:51+00:00 [haddock @ 2002-04-10 14:59:51 by simonmar] Sort the module tree - - - - - 76bd7b34 by Simon Marlow at 2002-04-10T15:50:11+00:00 [haddock @ 2002-04-10 15:50:10 by simonmar] Generate a little table of contents at the top of the module doc (only if the module actually contains some section headings, though). - - - - - bb8560a1 by Simon Marlow at 2002-04-10T16:10:26+00:00 [haddock @ 2002-04-10 16:10:26 by simonmar] Now we understand (or at least don't barf on) type signatures in patterns such as you might find when scoped type variables are in use. - - - - - 86c2a026 by Simon Marlow at 2002-04-10T16:10:49+00:00 [haddock @ 2002-04-10 16:10:49 by simonmar] more updates - - - - - 1c052b0e by Simon Marlow at 2002-04-10T16:28:05+00:00 [haddock @ 2002-04-10 16:28:05 by simonmar] Parse errors in doc strings are now reported as warnings rather that causing the whole thing to fall over. It still needs cleaning up (the warning is emitted with trace) but this will do for the time being. - - - - - ace03e8f by Simon Marlow at 2002-04-10T16:38:03+00:00 [haddock @ 2002-04-10 16:38:03 by simonmar] update again - - - - - 69006c3e by Simon Marlow at 2002-04-11T13:38:02+00:00 [haddock @ 2002-04-11 13:38:02 by simonmar] mention Opera - - - - - fe9b10f8 by Simon Marlow at 2002-04-11T13:40:31+00:00 [haddock @ 2002-04-11 13:40:30 by simonmar] - copy haddock.css into the same place as the generated HTML - new option: --css <file> specifies the style sheet to use - new option: -o <dir> specifies the directory in which to generate the output. - because Haddock now needs to know where to find its default stylesheet, we have to have a wrapper script and do the haddock-inplace thing (Makefile code copied largely from fptools/happy). - - - - - 106adbbe by Simon Marlow at 2002-04-24T15:12:41+00:00 [haddock @ 2002-04-24 15:12:41 by simonmar] Stop slurping comment lines when we see a row of dashes longer than length 2: these are useful as separators. - - - - - 995d3f9e by Simon Marlow at 2002-04-24T15:14:12+00:00 [haddock @ 2002-04-24 15:14:11 by simonmar] Grok the kind of module headers we use in fptools/libraries, and pass the "portability", "stability", and "maintainer" strings through into the generated HTML. If the module header doesn't match the pattern, then we don't include the info in the HTML. - - - - - e14da136 by Simon Marlow at 2002-04-24T15:16:57+00:00 [haddock @ 2002-04-24 15:16:57 by simonmar] Done module headers now. - - - - - 2ca8dfd4 by Simon Marlow at 2002-04-24T15:57:48+00:00 [haddock @ 2002-04-24 15:57:47 by simonmar] Handle gcons in export lists (a common extension). - - - - - 044cea81 by Simon Marlow at 2002-04-25T14:20:12+00:00 [haddock @ 2002-04-25 14:20:12 by simonmar] Add the little lambda icon - - - - - 63955027 by Simon Marlow at 2002-04-25T14:40:05+00:00 [haddock @ 2002-04-25 14:40:05 by simonmar] - Add support for named chunks of documentation which can be referenced from the export list. - Copy the icon from $libdir to the destination in HTML mode. - - - - - 36e3f913 by Simon Marlow at 2002-04-25T16:48:36+00:00 [haddock @ 2002-04-25 16:48:36 by simonmar] More keyboard bashing - - - - - 7ae18dd0 by Simon Marlow at 2002-04-26T08:43:33+00:00 [haddock @ 2002-04-26 08:43:33 by simonmar] Package util reqd. to compile with 4.08.2 - - - - - bbd5fbab by Simon Marlow at 2002-04-26T10:13:00+00:00 [haddock @ 2002-04-26 10:13:00 by simonmar] Include $(GHC_HAPPY_OPTS) when compiling HsParser - - - - - 31c53d79 by Simon Marlow at 2002-04-26T11:18:57+00:00 [haddock @ 2002-04-26 11:18:56 by simonmar] - support for fundeps (partially contributed by Brett Letner - thanks Brett). - make it build with GHC 4.08.2 - - - - - c415ce76 by Simon Marlow at 2002-04-26T13:15:02+00:00 [haddock @ 2002-04-26 13:15:02 by simonmar] Move the explicit formatting of the little table for the stability/portability/maintainer info from the HTML into the CSS, and remove the explicit table size (just right-align it). - - - - - 520ee21a by Simon Marlow at 2002-04-26T16:01:44+00:00 [haddock @ 2002-04-26 16:01:44 by simonmar] Yet more keyboard bashing - this is pretty much complete now. - - - - - 2ae37179 by Simon Marlow at 2002-04-26T16:02:14+00:00 [haddock @ 2002-04-26 16:02:14 by simonmar] Add a couple of things I forgot about - - - - - b7211e04 by Simon Marlow at 2002-04-29T15:28:12+00:00 [haddock @ 2002-04-29 15:28:12 by simonmar] bugfix for declBinders on a NewTypeDecl - - - - - 640c154a by Simon Marlow at 2002-04-29T15:28:54+00:00 [haddock @ 2002-04-29 15:28:54 by simonmar] Allow '-- |' style annotations on constructors and record fields. - - - - - 393f258a by Simon Marlow at 2002-04-29T15:37:32+00:00 [haddock @ 2002-04-29 15:37:32 by simonmar] syntax fix - - - - - 8a2c2549 by Simon Marlow at 2002-04-29T15:37:48+00:00 [haddock @ 2002-04-29 15:37:48 by simonmar] Add an example - - - - - db88f8a2 by Simon Marlow at 2002-04-29T15:55:46+00:00 [haddock @ 2002-04-29 15:55:46 by simonmar] remove a trace - - - - - 2b0248e0 by Simon Marlow at 2002-04-29T15:56:19+00:00 [haddock @ 2002-04-29 15:56:19 by simonmar] Fix for 'make install' - - - - - 120453a0 by Simon Marlow at 2002-04-29T15:56:39+00:00 [haddock @ 2002-04-29 15:56:39 by simonmar] Install the auxilliary bits - - - - - 950e6dbb by Simon Marlow at 2002-04-29T15:57:30+00:00 [haddock @ 2002-04-29 15:57:30 by simonmar] Add BinDist bits - - - - - 154b9d71 by Simon Marlow at 2002-05-01T11:02:52+00:00 [haddock @ 2002-05-01 11:02:52 by simonmar] update - - - - - ba6c39fa by Simon Marlow at 2002-05-01T11:03:26+00:00 [haddock @ 2002-05-01 11:03:26 by simonmar] Add another item - - - - - bacb5e33 by Simon Marlow at 2002-05-03T08:50:00+00:00 [haddock @ 2002-05-03 08:50:00 by simonmar] Fix some typos. - - - - - 54c87895 by Sven Panne at 2002-05-05T19:40:51+00:00 [haddock @ 2002-05-05 19:40:51 by panne] As a temporary hack/workaround for a bug in GHC's simplifier, don't pass Happy the -c option for generating the parsers in this subdir. Furthermore, disable -O for HaddocParse, too. - - - - - e6c08703 by Simon Marlow at 2002-05-06T09:51:10+00:00 [haddock @ 2002-05-06 09:51:10 by simonmar] Add RPM spec file (thanks to Tom Moertel <tom-rpms at moertel.com>) - - - - - 7b8fa8e7 by Simon Marlow at 2002-05-06T12:29:26+00:00 [haddock @ 2002-05-06 12:29:26 by simonmar] Add missing type signature (a different workaround for the bug in GHC's simplifier). - - - - - cd0e300d by Simon Marlow at 2002-05-06T12:30:09+00:00 [haddock @ 2002-05-06 12:30:09 by simonmar] Remove workaround for simplifier bug in previous revision. - - - - - 687e68fa by Simon Marlow at 2002-05-06T12:32:32+00:00 [haddock @ 2002-05-06 12:32:32 by simonmar] Allow empty data declarations (another GHC extension). - - - - - 8f29f696 by Simon Marlow at 2002-05-06T12:49:21+00:00 [haddock @ 2002-05-06 12:49:21 by simonmar] Fix silly bug in named documentation block lookup. - - - - - 8e0059af by Simon Marlow at 2002-05-06T13:02:42+00:00 [haddock @ 2002-05-06 13:02:42 by simonmar] Add another named chunk with a different name - - - - - 68f8a896 by Simon Marlow at 2002-05-06T13:32:32+00:00 [haddock @ 2002-05-06 13:32:32 by simonmar] Be more lenient about extra paragraph breaks - - - - - 65fc31db by Simon Marlow at 2002-05-07T15:36:36+00:00 [haddock @ 2002-05-07 15:36:36 by simonmar] DocEmpty is a right and left-unit of DocAppend (remove it in the smart constructor). - - - - - adc81078 by Simon Marlow at 2002-05-07T15:37:15+00:00 [haddock @ 2002-05-07 15:37:15 by simonmar] Allow code blocks to be denoted with bird-tracks in addition to [...]. - - - - - 1283a3c1 by Simon Marlow at 2002-05-08T11:21:56+00:00 [haddock @ 2002-05-08 11:21:56 by simonmar] Add a facility for specifying options that affect Haddock's treatment of the module. Options are given at the top of the module in a comma-separated list, beginning with '-- #'. eg. -- # prune, hide, ignore-exports Options currently available, with their meanings: prune: ignore declarations which have no documentation annotations ignore-exports: act as if the export list were not specified (i.e. export everything local to the module). hide: do not include this module in the generated documentation, but propagate any exported definitions to modules which re-export them. There's a slight change in the semantics for re-exporting a full module by giving 'module M' in the export list: if module M does not have the 'hide' option, then the documentation will now just contain a reference to module M rather than the full inlined contents of that module. These features, and some other changes in the pipeline, are the result of discussions between myself and Manuel Chakravarty <chak at cse.unsw.edu.au> (author of IDoc) yesterday. Also: some cleanups, use a Writer monad to collect error messages in some places instead of just printing them with trace. - - - - - a2239cf5 by Simon Marlow at 2002-05-08T11:22:30+00:00 [haddock @ 2002-05-08 11:22:30 by simonmar] Update to test new features. - - - - - 6add955f by Simon Marlow at 2002-05-08T13:37:25+00:00 [haddock @ 2002-05-08 13:37:25 by simonmar] Change the markup for typewriter-font from [...] to @... at . The reasoning is that the '@' symbol is much less likely to be needed than square brackets, and we don't want to have to escape square brackets in code fragments. This will be mildly painful in the short term, but it's better to get the change out of the way as early as possible. - - - - - cda06447 by Simon Marlow at 2002-05-08T13:39:56+00:00 [haddock @ 2002-05-08 13:39:56 by simonmar] Allow nested-style comments to be used as documentation annotations too. eg. {-| ... -} is equivalent to -- | ... An extra space can also be left after the comment opener: {- | ... -}. The only version that isn't allowed is {-# ... -}, because this syntax overlaps with Haskell pragmas; use {- # ... -} instead. - - - - - db23f65e by Simon Marlow at 2002-05-08T14:48:41+00:00 [haddock @ 2002-05-08 14:48:39 by simonmar] Add support for existential quantifiers on constructors. - - - - - adce3794 by Simon Marlow at 2002-05-08T15:43:25+00:00 [haddock @ 2002-05-08 15:43:25 by simonmar] update - - - - - 62a1f436 by Simon Marlow at 2002-05-08T15:44:10+00:00 [haddock @ 2002-05-08 15:44:10 by simonmar] Update to version 0.2 - - - - - f6a24ba3 by Simon Marlow at 2002-05-09T08:48:29+00:00 [haddock @ 2002-05-09 08:48:29 by simonmar] typo - - - - - 9f9522a4 by Simon Marlow at 2002-05-09T10:33:14+00:00 [haddock @ 2002-05-09 10:33:14 by simonmar] oops, left out '/' from the special characters in the last change. - - - - - 14abcb39 by Simon Marlow at 2002-05-09T10:34:44+00:00 [haddock @ 2002-05-09 10:34:44 by simonmar] Fix buglet - - - - - b8d878be by Simon Marlow at 2002-05-09T10:35:00+00:00 [haddock @ 2002-05-09 10:35:00 by simonmar] Give a more useful instance of Show for Module. - - - - - f7bfd626 by Simon Marlow at 2002-05-09T10:37:07+00:00 [haddock @ 2002-05-09 10:37:07 by simonmar] The last commit to Main.lhs broke the delicate balance of laziness which was being used to avoid computing the dependency graph of modules. So I finally bit the bullet and did a proper topological sort of the module graph, which turned out to be easy (stealing the Digraph module from GHC - this really ought to be in the libraries somewhere). - - - - - b481c1d0 by Simon Marlow at 2002-05-09T10:37:25+00:00 [haddock @ 2002-05-09 10:37:25 by simonmar] another item done - - - - - 032e2b42 by Simon Marlow at 2002-05-09T10:44:15+00:00 [haddock @ 2002-05-09 10:44:15 by simonmar] Don't consider a module re-export as having documentation, for the purposes of deciding whether we need a Synopsis section or not. - - - - - 5fb45e92 by Simon Marlow at 2002-05-09T11:10:55+00:00 [haddock @ 2002-05-09 11:10:55 by simonmar] Add a special case for list types in ppHsAType - - - - - 1937e428 by Simon Marlow at 2002-05-09T12:43:06+00:00 [haddock @ 2002-05-09 12:43:06 by simonmar] Type synonyms can accept a ctype on the RHS, to match GHC. - - - - - 0f16ce56 by Simon Marlow at 2002-05-09T12:45:19+00:00 [haddock @ 2002-05-09 12:45:19 by simonmar] Add 'stdcall' keyword - - - - - 29b0d7d2 by Simon Marlow at 2002-05-09T13:35:45+00:00 [haddock @ 2002-05-09 13:35:45 by simonmar] Add System Requirements section - - - - - bf14dddd by Simon Marlow at 2002-05-09T13:36:11+00:00 [haddock @ 2002-05-09 13:36:11 by simonmar] Test existential types, amongst other things - - - - - 502f8f6f by Simon Marlow at 2002-05-09T13:37:35+00:00 [haddock @ 2002-05-09 13:37:35 by simonmar] Print the module name in a doc-string parse error - - - - - ca1f8d49 by Simon Marlow at 2002-05-09T13:38:04+00:00 [haddock @ 2002-05-09 13:38:04 by simonmar] Add dependency - - - - - 8d3d91ff by Simon Marlow at 2002-05-09T15:37:57+00:00 [haddock @ 2002-05-09 15:37:57 by simonmar] Add the changelog/release notes - - - - - f3960959 by Simon Marlow at 2002-05-09T15:47:47+00:00 [haddock @ 2002-05-09 15:47:47 by simonmar] mention the backquote-style of markup - - - - - 089fb6e6 by Simon Marlow at 2002-05-09T15:59:45+00:00 [haddock @ 2002-05-09 15:59:45 by simonmar] update - - - - - bdd3be0b by Simon Marlow at 2002-05-09T15:59:56+00:00 [haddock @ 2002-05-09 15:59:56 by simonmar] Document changes since 0.1 - - - - - 00fc4af8 by Simon Marlow at 2002-05-10T08:22:48+00:00 [haddock @ 2002-05-10 08:22:48 by simonmar] oops, update to version 0.2 - - - - - a8a79041 by Simon Marlow at 2002-05-10T16:05:08+00:00 [haddock @ 2002-05-10 16:05:08 by simonmar] Only include a mini-contents if there are 2 or more sections - - - - - 06653319 by Simon Marlow at 2002-05-13T09:13:12+00:00 [haddock @ 2002-05-13 09:13:12 by simonmar] fix typos - - - - - 1402b19b by Simon Marlow at 2002-05-13T10:14:22+00:00 [haddock @ 2002-05-13 10:14:22 by simonmar] Allow backquote as the right-hand quote as well as the left-hand quote, as suggested by Dean Herrington. Clean up the grammar a litte. - - - - - dcd5320d by Simon Marlow at 2002-05-13T10:44:10+00:00 [haddock @ 2002-05-13 10:44:10 by simonmar] a couple more things, prioritise a bit - - - - - a90130c4 by Simon Marlow at 2002-05-13T15:19:03+00:00 [haddock @ 2002-05-13 15:19:03 by simonmar] Cope with datatypes which have documentation on the constructor but not the type itself, and records which have documentation on the fields but not the constructor. (Thanks to Ross Paterson for pointing out the bugs). - - - - - a774d432 by Simon Marlow at 2002-05-13T15:20:54+00:00 [haddock @ 2002-05-13 15:20:54 by simonmar] Fix one of the record examples - - - - - 2d1d5218 by Simon Marlow at 2002-05-15T12:44:35+00:00 [haddock @ 2002-05-15 12:44:35 by simonmar] Preserve the newline before a bird-track, but only within a paragraph. - - - - - 1554c09a by Simon Marlow at 2002-05-15T13:03:02+00:00 [haddock @ 2002-05-15 13:03:01 by simonmar] Reworking of the internals to support documenting function arguments (the Most Wanted new feature by the punters). The old method of keeping parsed documentation in a Name -> Doc mapping wasn't going to cut it for anntations on type components, where there's no name to attach the documentation to, so I've moved to storing all the documentation in the abstract syntax. Previously some of the documentation was left in the abstract syntax by the parser, but was later extracted into the mapping. In order to avoid having to parameterise the abstract syntax over the type of documentation stored in it, we have to parse the documentation at the same time as we parse the Haskell source (well, I suppose we could store 'Either String Doc' in the HsSyn, but that's clunky). One upshot is that documentation is now parsed eagerly, and documentation parse errors are fatal (but have better line numbers in the error message). The new story simplifies matters for the code that processes the source modules, because we don't have to maintain the extra Name->Doc mapping, and it should improve efficiency a little too. New features: - Function arguments and return values can now have doc annotations. - If you refer to a qualified name in a doc string, eg. 'IO.putStr', then Haddock will emit a hyperlink even if the identifier is not in scope, so you don't have to make sure everything referred to from the documentation is imported. - several bugs & minor infelicities fixed. - - - - - 57344dc3 by Simon Marlow at 2002-05-15T13:03:19+00:00 [haddock @ 2002-05-15 13:03:19 by simonmar] Bump to version 0.3 - - - - - b2791812 by Simon Marlow at 2002-05-15T13:03:41+00:00 [haddock @ 2002-05-15 13:03:41 by simonmar] update - - - - - fead183e by Simon Marlow at 2002-05-15T13:10:15+00:00 [haddock @ 2002-05-15 13:10:15 by simonmar] Rename Foo.hs to Test.hs, and add a Makefile - - - - - b0b1f89f by Simon Marlow at 2002-05-15T13:16:07+00:00 [haddock @ 2002-05-15 13:16:07 by simonmar] - Remove the note about function argument docs not being implemented - Note that qualified identifiers can be used to point to entities that aren't in scope. - - - - - 5665f31a by Simon Marlow at 2002-05-15T13:28:46+00:00 [haddock @ 2002-05-15 13:28:46 by simonmar] Patch to add support for GHC-style primitive strings ".."#, from Ross Paterson. - - - - - 0564505d by Simon Marlow at 2002-05-17T10:51:57+00:00 [haddock @ 2002-05-17 10:51:57 by simonmar] Fix bugs in qualified name handling (A.B.f was returned as B.f) - - - - - 10e7311c by Simon Marlow at 2002-05-21T10:24:52+00:00 [haddock @ 2002-05-21 10:24:52 by simonmar] - Use an alternate tabular layout for datatypes, which is more compact - Fix some problems with the function argument documentation - - - - - 2f91c2a6 by Simon Marlow at 2002-05-21T10:27:40+00:00 [haddock @ 2002-05-21 10:27:40 by simonmar] add a few more test cases - - - - - 01c2ddd2 by Simon Marlow at 2002-05-21T10:28:33+00:00 [haddock @ 2002-05-21 10:28:33 by simonmar] Rearrange a bit, and add support for tabular datatype rendering - - - - - a4e4c5f8 by Simon Marlow at 2002-05-27T09:03:52+00:00 [haddock @ 2002-05-27 09:03:51 by simonmar] Lots of changes: - instances of a class are listed with the class, and instances involving a datatype are listed with that type. Derived instances aren't included at the moment: the calculation to find the instance head for a derived instance is non-trivial. - some formatting changes; use rows with specified height rather than cellspacing in some places. - various fixes (source file links were wrong, amongst others) - - - - - 48722e68 by Simon Marlow at 2002-05-27T12:30:38+00:00 [haddock @ 2002-05-27 12:30:37 by simonmar] - Put function arguments *before* the doc for the function, as suggested by Sven Panne. This looks nicer when the function documentation is long. - Switch to using bold for binders at the definition site, and use underline for keywords. This makes the binder stand out more. - - - - - 657204d2 by Simon Marlow at 2002-05-27T13:19:49+00:00 [haddock @ 2002-05-27 13:19:49 by simonmar] Fix bug: we weren't renaming HsDocCommentNamed in renameDecl - - - - - 592aae66 by Simon Marlow at 2002-05-27T14:10:27+00:00 [haddock @ 2002-05-27 14:10:27 by simonmar] Fix some bugs in the rendering of qualified type signatures. - - - - - 69c8f763 by Simon Marlow at 2002-05-27T14:36:45+00:00 [haddock @ 2002-05-27 14:36:45 by simonmar] warning message tweak - - - - - 16e64e21 by Simon Marlow at 2002-05-27T14:53:53+00:00 [haddock @ 2002-05-27 14:53:53 by simonmar] hyperlinked identifiers should be in <tt> - - - - - 8d5e4783 by Simon Marlow at 2002-05-27T15:56:45+00:00 [haddock @ 2002-05-27 15:56:45 by simonmar] Do something sensible for modules which don't export anything (except instances). - - - - - 9d3ef811 by Simon Marlow at 2002-05-28T10:12:50+00:00 [haddock @ 2002-05-28 10:12:50 by simonmar] Rename the module documentation properly (bug reported by Sven Panne). - - - - - ef03a1cc by Simon Marlow at 2002-05-28T10:13:04+00:00 [haddock @ 2002-05-28 10:13:04 by simonmar] Add some more test cases - - - - - 92baa0e8 by Simon Marlow at 2002-05-28T11:17:55+00:00 [haddock @ 2002-05-28 11:17:55 by simonmar] If an identifier doesn't lex, then just replace it by a DocString. - - - - - a3156213 by Simon Marlow at 2002-05-28T16:16:19+00:00 [haddock @ 2002-05-28 16:16:19 by simonmar] Only link to names in the current module which are actually listed in the documentation. A name may be exported but not present in the documentation if it is exported as part of a 'module M' export specifier. - - - - - 31acf941 by Simon Marlow at 2002-05-28T16:17:11+00:00 [haddock @ 2002-05-28 16:17:11 by simonmar] update - - - - - 7e474ebf by Sigbjorn Finne at 2002-05-28T22:42:08+00:00 [haddock @ 2002-05-28 22:42:08 by sof] Handle lone occurrences of '/', e.g., -- | This/that. [did this in the lexer rather than in the parser, as I couldn't see a way not to introduce an S/R conflict that way.] - - - - - 093f7e53 by Simon Marlow at 2002-05-29T09:09:49+00:00 [haddock @ 2002-05-29 09:09:49 by simonmar] Back out previous change until we can find a better way to do this. - - - - - 9234389c by Simon Marlow at 2002-05-29T13:19:06+00:00 [haddock @ 2002-05-29 13:19:06 by simonmar] Make the markup syntax a little more friendly: - single quotes are now interpreted literally unless they surround a valid Haskell identifier. So for example now there's no need to escape a single quote used as an apostrophe. - text to the right of a bird track is now literal (if you want marked-up text in a code block, use @...@). - - - - - b3333526 by Simon Marlow at 2002-05-29T13:38:51+00:00 [haddock @ 2002-05-29 13:38:51 by simonmar] Document recent changes to markup syntax - - - - - f93641d6 by Simon Marlow at 2002-05-29T15:27:18+00:00 [haddock @ 2002-05-29 15:27:18 by simonmar] Include the instances in abstract data types too - - - - - 613f21e3 by Simon Marlow at 2002-06-03T13:05:58+00:00 [haddock @ 2002-06-03 13:05:57 by simonmar] Allow exporting of individual class methods and record selectors. For these we have to invent the correct type signature, which we do in the simplest possible way (i.e. no context reduction nonsense in the class case). - - - - - 14b36807 by Simon Marlow at 2002-06-03T13:20:00+00:00 [haddock @ 2002-06-03 13:20:00 by simonmar] Fix linking to qualified names again (thanks to Sven Panne for pointing out the bug). - - - - - 95b10eac by Simon Marlow at 2002-06-03T13:46:48+00:00 [haddock @ 2002-06-03 13:46:48 by simonmar] Fix for exporting record selectors from a newtype declaration - - - - - 272f932e by Simon Marlow at 2002-06-03T13:56:38+00:00 [haddock @ 2002-06-03 13:56:38 by simonmar] update to version 0.3 - - - - - 1c0a3bed by Simon Marlow at 2002-06-03T14:05:07+00:00 [haddock @ 2002-06-03 14:05:07 by simonmar] Add changes in version 0.3 - - - - - 145b4626 by Simon Marlow at 2002-06-03T14:12:38+00:00 [haddock @ 2002-06-03 14:12:38 by simonmar] Render class names as proper binders - - - - - 052106b3 by Simon Marlow at 2002-06-03T14:15:10+00:00 [haddock @ 2002-06-03 14:15:10 by simonmar] update, and separate into bugs, features, and cosmetic items. - - - - - 854f4914 by Simon Marlow at 2002-06-03T14:16:13+00:00 [haddock @ 2002-06-03 14:16:13 by simonmar] More test cases - - - - - 466922c8 by Simon Marlow at 2002-06-03T14:16:56+00:00 [haddock @ 2002-06-03 14:16:56 by simonmar] Example from the paper - - - - - 9962a045 by Simon Marlow at 2002-06-03T14:17:49+00:00 [haddock @ 2002-06-03 14:17:49 by simonmar] A debugging version of the style-sheet, which gives some tables coloured backgrounds so we can see what's going on. - - - - - f16b79db by Simon Marlow at 2002-06-03T14:19:46+00:00 [haddock @ 2002-06-03 14:19:46 by simonmar] typo - - - - - 620db27b by Simon Marlow at 2002-06-03T14:48:32+00:00 [haddock @ 2002-06-03 14:48:32 by simonmar] oops, fix markup bugs - - - - - 53fd105c by Simon Marlow at 2002-06-05T09:05:07+00:00 [haddock @ 2002-06-05 09:05:07 by simonmar] Keep foreign imports when there is no export list (bug reported by Sven Panne). - - - - - 6d98989c by Simon Marlow at 2002-06-05T09:12:02+00:00 [haddock @ 2002-06-05 09:12:02 by simonmar] Identifiers in single quotes can be symbol names too (bug reported by Hal Daume). - - - - - 001811e5 by Sven Panne at 2002-06-08T14:03:36+00:00 [haddock @ 2002-06-08 14:03:36 by panne] Tiny workaround for the fact that Haddock currently ignores HsImportSpecs: Let the local_orig_env take precedence. This is no real solution at all, but improves things sometimes, e.g. in my GLUT documentation. :-) - - - - - 504d19c9 by Simon Marlow at 2002-06-11T09:23:25+00:00 [haddock @ 2002-06-11 09:23:25 by simonmar] portability nit - - - - - e13b5af4 by Simon Marlow at 2002-06-20T12:38:07+00:00 [haddock @ 2002-06-20 12:38:07 by simonmar] Empty declaration fixes. - - - - - f467a9b6 by Simon Marlow at 2002-06-20T12:39:02+00:00 [haddock @ 2002-06-20 12:39:01 by simonmar] Add support for a "prologue" - a description for the whole library, placed on the contents page before the module list. - - - - - b8dbfe20 by Simon Marlow at 2002-06-21T12:43:06+00:00 [haddock @ 2002-06-21 12:43:06 by simonmar] When we have a single code block paragraph, don't place it in <pre>..</pre>, just use <tt>..</tt> to avoid generating extra vertical white space in some browsers. - - - - - 4831dbbd by Simon Marlow at 2002-06-21T15:50:42+00:00 [haddock @ 2002-06-21 15:50:42 by simonmar] Add support for reading and writing interface files(!) This turned out to be quite easy, and necessary to get decent hyperlinks between the documentation for separate packages in the libraries. The functionality isn't quite complete yet: for a given package of modules, you'd like to say "the HTML for these modules lives in directory <dir>" (currently they are assumed to be all in the same place). Two new flags: --dump-interface=FILE dump an interface file in FILE --read-interface=FILE read interface from FILE an interface file describes *all* the modules being processed. Only the exported names are kept in the interface: if you re-export a name from a module in another interface the signature won't be copied. This is a compromise to keep the size of the interfaces sensible. Also, I added another useful option: --no-implicit-prelude avoids trying to import the Prelude. Previously this was the default, but now importing the Prelude from elsewhere makes sense if you also read in an interface containing the Prelude module, so Haddock imports the Prelude implicitly according to the Haskell spec. - - - - - d3640a19 by Sven Panne at 2002-06-23T14:54:00+00:00 [haddock @ 2002-06-23 14:54:00 by panne] Make it compile with newer GHCs - - - - - 780c506b by Sven Panne at 2002-06-23T15:44:31+00:00 [haddock @ 2002-06-23 15:44:31 by panne] Cleaned up build root handling and added more docs - - - - - 45290d2e by Simon Marlow at 2002-06-24T14:37:43+00:00 [haddock @ 2002-06-24 14:37:42 by simonmar] When reading an interface, allow a file path offset to be specified which represents the path to the HTML files for the modules specified by that interface. The path may be either relative (to the location of the HTML for this package), or absolute. The syntax is --read-interface=PATH,FILE where PATH is the path to the HTML, and FILE is the filename containing the interface. - - - - - 4e2b9ae6 by Simon Marlow at 2002-07-03T16:01:08+00:00 [haddock @ 2002-07-03 16:01:07 by simonmar] Handle import specs properly, include 'hiding'. Haddock now has a complete implementation of the Haskell module system (more or less; I won't claim it's 100% correct). - - - - - 9a9aa1a8 by Simon Marlow at 2002-07-03T16:18:16+00:00 [haddock @ 2002-07-03 16:18:16 by simonmar] Update - - - - - 560c3026 by Simon Marlow at 2002-07-04T14:56:10+00:00 [haddock @ 2002-07-04 14:56:10 by simonmar] Clean up the code that constructs the exported declarations, and fix a couple of bugs along the way. Now if you import a class hiding one of the methods, then re-export the class, the version in the documentation will correctly have the appropriate method removed. - - - - - 2c26e77d by Simon Marlow at 2002-07-04T15:26:13+00:00 [haddock @ 2002-07-04 15:26:13 by simonmar] More bugfixes to the export handling - - - - - 03e0710d by Simon Marlow at 2002-07-09T10:12:10+00:00 [haddock @ 2002-07-09 10:12:10 by simonmar] Don't require that the list type comes from "Prelude" for it to be treated as special syntax (sometimes it comes from Data.List or maybe even GHC.Base). - - - - - 44f3891a by Simon Marlow at 2002-07-09T10:12:51+00:00 [haddock @ 2002-07-09 10:12:51 by simonmar] commented-out debugging code - - - - - 97280873 by Krasimir Angelov at 2002-07-09T16:33:33+00:00 [haddock @ 2002-07-09 16:33:31 by krasimir] 'Microsoft HTML Help' support - - - - - 3dc04655 by Simon Marlow at 2002-07-10T09:40:56+00:00 [haddock @ 2002-07-10 09:40:56 by simonmar] Fix for rendering of the (->) type constructor, from Ross Paterson. - - - - - c9f149c6 by Simon Marlow at 2002-07-10T10:26:11+00:00 [haddock @ 2002-07-10 10:26:11 by simonmar] Tweaks to the MS Help support: the extra files are now only generated if you ask for them (--ms-help). - - - - - e8acc1e6 by Simon Marlow at 2002-07-10T10:57:10+00:00 [haddock @ 2002-07-10 10:57:10 by simonmar] Document all the new options since 0.3 - - - - - 8bb85544 by Simon Marlow at 2002-07-10T10:58:31+00:00 [haddock @ 2002-07-10 10:58:31 by simonmar] Sort the options a bit - - - - - abc0dd59 by Simon Marlow at 2002-07-15T09:19:38+00:00 [haddock @ 2002-07-15 09:19:38 by simonmar] Fix a bug in mkExportItems when processing a module without an explicit export list. We were placing one copy of a declaration for each binder in the declaration, which for a data type would mean one copy of the whole declaration per constructor or record selector. - - - - - dde65bb9 by Simon Marlow at 2002-07-15T09:54:16+00:00 [haddock @ 2002-07-15 09:54:16 by simonmar] merge rev. 1.35 - - - - - bd7eb8c4 by Simon Marlow at 2002-07-15T10:14:31+00:00 [haddock @ 2002-07-15 10:14:30 by simonmar] Be a bit more liberal in the kind of commenting styles we allow, as suggested by Malcolm Wallace. Mostly this consists of allowing doc comments either side of a separator token. In an export list, a section heading is now allowed before the comma, as well as after it. eg. module M where ( T(..) -- * a section heading , f -- * another section heading , g ) In record fields, doc comments are allowed anywhere (previously a doc-next was allowed only after the comma, and a doc-before was allowed only before the comma). eg. data R = C { -- | describes 'f' f :: Int -- | describes 'g' , g :: Int } - - - - - 8f6dfe34 by Simon Marlow at 2002-07-15T10:21:56+00:00 [haddock @ 2002-07-15 10:21:56 by simonmar] Mention alternative commenting styles. - - - - - fc515bb7 by Simon Marlow at 2002-07-15T16:16:50+00:00 [haddock @ 2002-07-15 16:16:50 by simonmar] Allow multiple sections/subsections before and after a comma in the export list. Also at the same time I made the syntax a little stricter (multiple commas now aren't allowed between export specs). - - - - - 80a97e74 by Simon Marlow at 2002-07-19T09:13:10+00:00 [haddock @ 2002-07-19 09:13:10 by simonmar] Allow special id's ([], (), etc.) to be used in an import declaration. - - - - - a69d7378 by Simon Marlow at 2002-07-19T09:59:02+00:00 [haddock @ 2002-07-19 09:59:02 by simonmar] Allow special id's ([], (), etc.) to be used in an import declarations. - - - - - d205fa60 by Simon Marlow at 2002-07-19T10:00:16+00:00 [haddock @ 2002-07-19 10:00:16 by simonmar] Relax the restrictions which require doc comments to be followed by semi colons - in some cases this isn't necessary. Now you can write module M where { -- | some doc class C where {} } without needing to put a semicolon before the class declaration. - - - - - e9301e14 by Simon Marlow at 2002-07-23T08:24:09+00:00 [haddock @ 2002-07-23 08:24:09 by simonmar] A new TODO list item - - - - - e5d77586 by Simon Marlow at 2002-07-23T08:40:56+00:00 [haddock @ 2002-07-23 08:40:56 by simonmar] - update the acknowledgements - remove the paragraph that described how to use explicit layout with doc comments; it isn't relevant any more. - - - - - 78a94137 by Simon Marlow at 2002-07-23T08:43:02+00:00 [haddock @ 2002-07-23 08:43:02 by simonmar] more tests - - - - - 5c320927 by Simon Marlow at 2002-07-23T08:43:26+00:00 [haddock @ 2002-07-23 08:43:26 by simonmar] Updates for version 0.4 - - - - - 488e99ae by Simon Marlow at 2002-07-23T09:10:46+00:00 [haddock @ 2002-07-23 09:10:46 by simonmar] Fix the %changelog (rpm complained that it wasn't in the right order) - - - - - a77bb373 by Simon Marlow at 2002-07-23T09:12:38+00:00 [haddock @ 2002-07-23 09:12:38 by simonmar] Another item for the TODO list - - - - - f1ec1813 by Simon Marlow at 2002-07-23T10:18:46+00:00 [haddock @ 2002-07-23 10:18:46 by simonmar] Add a version banner when invoked with -v - - - - - 1d44cadf by Simon Marlow at 2002-07-24T09:28:19+00:00 [haddock @ 2002-07-24 09:28:19 by simonmar] Remove ^Ms - - - - - 4d8d5e94 by Simon Marlow at 2002-07-24T09:42:18+00:00 [haddock @ 2002-07-24 09:42:17 by simonmar] Patches to quieten ghc -Wall, from those nice folks at Galois. - - - - - d6edc43e by Simon Marlow at 2002-07-25T14:37:29+00:00 [haddock @ 2002-07-25 14:37:28 by simonmar] Patch to allow simple hyperlinking to an arbitrary location in another module's documentation, from Volker Stolz. Now in a doc comment: #foo# creates <a name="foo"></a> And you can use the form "M\#foo" to hyperlink to the label 'foo' in module 'M'. Note that the backslash is necessary for now. - - - - - b34d18fa by Simon Marlow at 2002-08-02T09:08:22+00:00 [haddock @ 2002-08-02 09:08:22 by simonmar] The <TT> and <PRE> environments seem to use a font that is a little too small in IE. Compensate. (suggestion from Daan Leijen). - - - - - 8106b086 by Simon Marlow at 2002-08-02T09:25:23+00:00 [haddock @ 2002-08-02 09:25:20 by simonmar] Remove <P>..</P> from around list items, to reduce excess whitespace between the items of bulleted and ordered lists. (Suggestion from Daan Leijen). - - - - - c1acff8f by Simon Marlow at 2002-08-05T09:03:49+00:00 [haddock @ 2002-08-05 09:03:49 by simonmar] update - - - - - f968661c by Simon Marlow at 2002-11-11T09:32:57+00:00 [haddock @ 2002-11-11 09:32:57 by simonmar] Fix cut-n-pasto - - - - - 12d02619 by Simon Marlow at 2002-11-13T09:49:46+00:00 [haddock @ 2002-11-13 09:49:46 by simonmar] Small bugfix in the --read-interface option parsing from Brett Letner. - - - - - 30e32d5e by Ross Paterson at 2003-01-16T15:07:57+00:00 [haddock @ 2003-01-16 15:07:57 by ross] Adjust for the new exception libraries (as well as the old ones). - - - - - 871f65df by Sven Panne at 2003-02-20T21:31:40+00:00 [haddock @ 2003-02-20 21:31:40 by panne] * Add varsyms and consyms to index * Exclude empty entries from index - - - - - bc42cc87 by Sven Panne at 2003-02-24T21:26:29+00:00 [haddock @ 2003-02-24 21:26:29 by panne] Don't convert a "newtype" to a single-constructor "data" for non-abstractly exported types, they are quite different regarding strictness/pattern matching. Now a "data" without any constructors is only emitted for an abstractly exported type, regardless if it is actually a "newtype" or a "data". - - - - - 0c2a1d99 by Sven Panne at 2003-03-08T19:02:38+00:00 [haddock @ 2003-03-08 19:02:38 by panne] Fixed some broken/redirected/canonicalized links found by a very picky link checker. - - - - - 25459269 by Sven Panne at 2003-03-09T21:13:43+00:00 [haddock @ 2003-03-09 21:13:43 by panne] Don't append a fragment to non-defining index entries, only documents with a defining occurrence have a name anchor. - - - - - 6be4db86 by Sven Panne at 2003-03-10T21:34:25+00:00 [haddock @ 2003-03-10 21:34:24 by panne] Escape fragments. This fixes e.g. links to operators. - - - - - eb12972c by Ross Paterson at 2003-04-25T10:50:06+00:00 [haddock @ 2003-04-25 10:50:05 by ross] An 80% solution to generating derived instances. A complete solution would duplicate the instance inference logic, but if a type variable occurs as a constructor argument, then we can just propagate the derived class to the variable. But we know nothing of the constraints on any type variables that occur elsewhere. For example, the declarations data Either a b = Left a | Right b deriving (Eq, Ord) data Ptr a = Ptr Addr# deriving (Eq, Ord) newtype IORef a = IORef (STRef RealWorld a) deriving Eq yield the instances (Eq a, Eq b) => Eq (Either a b) (Ord a, Ord b) => Ord (Either a b) Eq (Ptr a) Ord (Ptr a) (??? a) => Eq (IORef a) The last example shows the limits of this local analysis. Note that a type variable may be in both categories: then we know a constraint, but there may be more, or a stronger constraint, e.g. data Tree a = Node a [Tree a] deriving Eq yields (Eq a, ??? a) => Eq (Tree a) - - - - - de886f78 by Simon Marlow at 2003-04-25T11:17:55+00:00 [haddock @ 2003-04-25 11:17:55 by simonmar] Some updates, including moving the derived instance item down to the bottom of the list now that Ross has contributed some code that does the job for common cases. - - - - - 1b52cffd by Simon Marlow at 2003-04-30T14:02:32+00:00 [haddock @ 2003-04-30 14:02:32 by simonmar] When installing on Windows, run cygpath over $(HADDOCKLIB) so that haddock (a mingw program, built by GHC) can understand it. You still need to be in a cygwin environment to run Haddock, because of the shell script wrapper. - - - - - d4f638de by Simon Marlow at 2003-05-06T10:04:47+00:00 [haddock @ 2003-05-06 10:04:47 by simonmar] Catch another case of a paragraph containing just a DocMonospaced that should turn into a DocCodeBlock. - - - - - 4162b2b9 by Simon Marlow at 2003-05-06T10:11:44+00:00 [haddock @ 2003-05-06 10:11:44 by simonmar] Add some more code-block tests. - - - - - 4f5802c8 by Simon Marlow at 2003-05-06T10:14:52+00:00 [haddock @ 2003-05-06 10:14:52 by simonmar] Don't turn a single DocCodeBlock into a DocMonospaced, because that tends to remove the line breaks in the code. - - - - - ef8c45f7 by Simon Marlow at 2003-05-21T15:07:21+00:00 [haddock @ 2003-05-21 15:07:21 by simonmar] Only omit the module contents when there are no section headings at all. - - - - - bcee1e75 by Sigbjorn Finne at 2003-05-30T16:50:45+00:00 [haddock @ 2003-05-30 16:50:45 by sof] cygpath: for now, steer clear of --mixed - - - - - 30567af3 by Sigbjorn Finne at 2003-05-30T17:59:28+00:00 [haddock @ 2003-05-30 17:59:28 by sof] oops, drop test defn from prev commit - - - - - b0856e7d by Simon Marlow at 2003-06-03T09:55:26+00:00 [haddock @ 2003-06-03 09:55:26 by simonmar] Two small fixes to make the output valid HTML 4.01 (transitional). Thanks to Malcolm Wallace for pointing out the problems. - - - - - 70e137ea by Simon Marlow at 2003-07-28T13:30:35+00:00 [haddock @ 2003-07-28 13:30:35 by simonmar] Add tests for a couple of bugs. - - - - - 122bd578 by Simon Marlow at 2003-07-28T13:31:25+00:00 [haddock @ 2003-07-28 13:31:25 by simonmar] Add documentation for anchors. - - - - - 0bd27cb2 by Simon Marlow at 2003-07-28T13:31:46+00:00 [haddock @ 2003-07-28 13:31:46 by simonmar] Update - - - - - 08052d42 by Simon Marlow at 2003-07-28T13:32:12+00:00 [haddock @ 2003-07-28 13:32:12 by simonmar] layout tweak. - - - - - 13942749 by Simon Marlow at 2003-07-28T13:33:03+00:00 [haddock @ 2003-07-28 13:33:03 by simonmar] Differentiate links to types/classes from links to variables/constructors with a prefix ("t:" and "v:" respectively). - - - - - d7f493b9 by Simon Marlow at 2003-07-28T13:35:17+00:00 [haddock @ 2003-07-28 13:35:16 by simonmar] When a module A exports another module's contents via 'module B', then modules which import entities from B re-exported by A should link to B.foo rather than A.foo. See examples/Bug2.hs. - - - - - d94cf705 by Simon Marlow at 2003-07-28T13:36:14+00:00 [haddock @ 2003-07-28 13:36:14 by simonmar] Update to version 0.5 - - - - - dbb776cd by Sven Panne at 2003-07-28T14:02:43+00:00 [haddock @ 2003-07-28 14:02:43 by panne] * Updated to version 0.5 * Automagically generate configure if it is not there - - - - - 6cfeee53 by Simon Marlow at 2003-07-28T14:32:43+00:00 [haddock @ 2003-07-28 14:32:42 by simonmar] Update to avoid using hslibs with GHC >= 5.04 - - - - - a1ce838f by Simon Marlow at 2003-07-28T14:33:37+00:00 [haddock @ 2003-07-28 14:33:37 by simonmar] Update for 0.5 - - - - - c0fe6493 by Simon Marlow at 2003-07-28T14:53:22+00:00 [haddock @ 2003-07-28 14:53:22 by simonmar] Markup fix - - - - - 6ea31596 by Sven Panne at 2003-07-28T16:40:45+00:00 [haddock @ 2003-07-28 16:40:45 by panne] Make it compile with GHC >= 6.01 - - - - - afcd30fc by Simon Marlow at 2003-07-30T15:04:52+00:00 [haddock @ 2003-07-30 15:04:52 by simonmar] Pay attention to import specs when building the the import env, as well as the orig env. This may fix some wrong links in documentation when import specs are being used. - - - - - 17c3137f by Simon Marlow at 2003-07-30T16:05:41+00:00 [haddock @ 2003-07-30 16:05:40 by simonmar] Rename instances based on the import_env for the module in which they are to be displayed. This should give, in many cases, better links for the types and classes mentioned in the instance head. This involves keeping around the import_env in the iface until the end, because instances are not collected up until all the modules have been processed. Fortunately it doesn't seem to affect performance much. Instance heads are now attached to ExportDecls, rather than the HTML backend passing around a separate mapping for instances. This is a cleanup. - - - - - 3d3b5c87 by Sven Panne at 2003-08-04T10:18:24+00:00 [haddock @ 2003-08-04 10:18:24 by panne] Don't print parentheses around one-element contexts - - - - - 9e3f3f2d by Simon Marlow at 2003-08-04T12:59:47+00:00 [haddock @ 2003-08-04 12:59:47 by simonmar] A couple of TODOs. - - - - - e9d8085c by Simon Marlow at 2003-08-05T14:10:31+00:00 [haddock @ 2003-08-05 14:10:31 by simonmar] I'm not sure why, but it seems that the index entries for non-defining occurrences of entities did not have an anchor - the link just pointed to the module. This fixes it. - - - - - ff5c7d6d by Simon Marlow at 2003-08-15T14:42:59+00:00 [haddock @ 2003-08-15 14:42:59 by simonmar] Convert the lexer to Alex, and fix a bug in the process. - - - - - 1aa077bf by Simon Marlow at 2003-08-15T15:00:18+00:00 [haddock @ 2003-08-15 15:00:18 by simonmar] Update - - - - - d3de1e38 by Simon Marlow at 2003-08-15T15:01:03+00:00 [haddock @ 2003-08-15 15:01:03 by simonmar] wibbles - - - - - b40ece3b by Simon Marlow at 2003-08-18T10:04:47+00:00 [haddock @ 2003-08-18 10:04:47 by simonmar] Lex the 'mdo' keyword as 'do'. - - - - - 8f9a1146 by Simon Marlow at 2003-08-18T11:48:24+00:00 [haddock @ 2003-08-18 11:48:24 by simonmar] Two bugs from Sven. - - - - - ea54ebc0 by Simon Marlow at 2003-08-18T11:48:46+00:00 [haddock @ 2003-08-18 11:48:46 by simonmar] Fixes to the new lexer. - - - - - d5f6a4b5 by Simon Marlow at 2003-08-19T09:09:03+00:00 [haddock @ 2003-08-19 09:09:03 by simonmar] Further wibbles to the syntax. - - - - - 6bbdadb7 by Sven Panne at 2003-08-26T18:45:35+00:00 [haddock @ 2003-08-26 18:45:35 by panne] Use autoreconf instead of autoconf - - - - - 32e889cb by Sven Panne at 2003-08-26T19:01:19+00:00 [haddock @ 2003-08-26 19:01:18 by panne] Made option handling a bit more consistent with other tools, in particular: Every program in fptools should output * version info on stdout and terminate successfully when -V or --version * usage info on stdout and terminate successfully when -? or --help * usage info on stderr and terminate unsuccessfully when an unknown option is given. - - - - - 5d156a91 by Sven Panne at 2003-08-26T19:20:55+00:00 [haddock @ 2003-08-26 19:20:55 by panne] Make it *very* clear that we terminate when given a -V/--version flag - - - - - e6577265 by Sven Panne at 2003-08-27T07:50:03+00:00 [haddock @ 2003-08-27 07:50:02 by panne] * Made -D a short option for --dump-interface. * Made -m a short option for --ms-help. * Made -n a short option for --no-implicit-prelude. * Made -c a short option for --css. * Removed DocBook options from executable (they didn't do anything), but mark them as reserved in the docs. Note that the short option for DocBook output is now -S (from SGML) instead of -d. The latter is now a short option for --debug. * The order of the Options in the documentation now matches the order printed by Haddock itself. Note: Although changing the names of options is often a bad idea, I'd really like to make the options for the programs in fptools more consistent and compatible to the ones used in common GNU programs. - - - - - d303ff98 by Simon Marlow at 2003-09-10T08:23:48+00:00 [haddock @ 2003-09-10 08:23:48 by simonmar] Add doc subdir. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 9a70e46a by Simon Marlow at 2003-09-10T08:24:32+00:00 [haddock @ 2003-09-10 08:24:32 by simonmar] Install these files in $(datadir), not $(libdir), since they're architecture independent. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - bbb87e7a by Simon Marlow at 2003-09-10T08:25:31+00:00 [haddock @ 2003-09-10 08:25:31 by simonmar] Haddock's supplementary HTML bits now live in $(datadir), not $(libdir). Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 3587c24b by Simon Marlow at 2003-09-22T10:34:38+00:00 [haddock @ 2003-09-22 10:34:38 by simonmar] Allow installing of docs. - - - - - d510b517 by Sven Panne at 2003-10-11T08:10:44+00:00 [haddock @ 2003-10-11 08:10:44 by panne] Include architecture-independent files in file list - - - - - 187d7618 by Sigbjorn Finne at 2003-10-20T17:19:24+00:00 [haddock @ 2003-10-20 17:19:22 by sof] support for i-parameters + zip comprehensions - - - - - b6c7a273 by Simon Marlow at 2003-11-03T14:24:24+00:00 [haddock @ 2003-11-03 14:24:24 by simonmar] Update TODO file. - - - - - 58513e33 by Simon Marlow at 2003-11-05T11:22:04+00:00 [haddock @ 2003-11-05 11:22:04 by simonmar] Remove the last of the uses of 'trace' to emit warnings, and tidy up a couple of places where duplicate warnings were being emitted. - - - - - 33a78846 by Simon Marlow at 2003-11-05T11:30:53+00:00 [haddock @ 2003-11-05 11:30:52 by simonmar] - Suppress warnings about unknown imported modules by default. - Add a -v/--verbose flag to re-enable these warnings. The general idea is to suppress the "Warning: unknown module: Prelude" warnings which most Haddock users will see every time, and which aren't terribly useful. - - - - - a969de7f by Simon Marlow at 2003-11-05T12:30:28+00:00 [haddock @ 2003-11-05 12:30:28 by simonmar] - Remove the emboldening of index entries for defining locations. This isn't useful, and breaks abstractions. - If an entity is re-exported by a module but the module doesn't include documentation for that entity (perhaps because it is re-exported by 'module M'), then don't attempt to hyperlink to the documentation from the index. Instead, just list that module in the index, to indicate that the entity is exported from there. - - - - - f14ea82a by Simon Marlow at 2003-11-05T15:15:59+00:00 [haddock @ 2003-11-05 15:15:59 by simonmar] Index overhaul: - no more separate type/class and variable/function indices - the index now makes a distinction between different entities with the same name. One example is a type constructor with the same name as a data constructor, but another example is simply a function with the same name exported by two different modules. For example, the index entry for 'catch' now looks like this: catch 1 (Function) Control.Exception 2 (Function) GHC.Exception, Prelude, System.IO, System.IO.Error making it clear that there are two different 'catch'es, but one of them is exported by several modules. - Each index page now has the index contents (A B C ...) at the top. Please let me know if you really hate any of this. - - - - - 01a25ca6 by Simon Marlow at 2003-11-05T15:16:38+00:00 [haddock @ 2003-11-05 15:16:38 by simonmar] Update - - - - - 1a7ccb86 by Simon Marlow at 2003-11-05T17:16:05+00:00 [haddock @ 2003-11-05 17:16:04 by simonmar] Support for generating a single unified index for several packages. --use-index=URL turns off normal index generation, causes Index links to point to URL. --gen-index generates an combined index from the specified interfaces. Currently doesn't work exactly right, because the interfaces don't contain the iface_reexported info. I'll need to fix that up. - - - - - a2bca16d by Simon Marlow at 2003-11-06T10:44:52+00:00 [haddock @ 2003-11-06 10:44:52 by simonmar] Include iface_reexported in the .haddock file. This unfortunately bloats the file (40% for base). If this gets to be a problem we can always apply the dictionary trick that GHC uses for squashing .hi files. - - - - - 0a09c293 by Simon Marlow at 2003-11-06T12:39:47+00:00 [haddock @ 2003-11-06 12:39:46 by simonmar] - Add definition lists, marked up like this: -- | This is a definition list: -- -- [@foo@] The description of @foo at . -- -- [@bar@] The description of @bar at . Cunningly, the [] characters are not treated specially unless a [ is found at the beginning of a paragraph, in which case the ] becomes special in the following text. - Add --use-contents and --gen-contents, along the lines of --use-index and --gen-index added yesterday. Now we can generate a combined index and contents for the whole of the hierarchical libraries, and in theory the index/contents on the system could be updated as new packages are added. - - - - - fe1b3460 by Simon Marlow at 2003-11-06T14:47:36+00:00 [haddock @ 2003-11-06 14:47:36 by simonmar] Remove the 'Parent' button - it is of dubious use, and often points into thin air. - - - - - db6d762f by Simon Marlow at 2003-11-06T16:48:14+00:00 [haddock @ 2003-11-06 16:48:11 by simonmar] - Include the OptHide setting in the interface, so we don't include hidden modules in the combined index/contents. - Add a -k/--package flag to set the package name for the current set of modules. The package name for each module is now shown in the right-hand column of the contents, in a combined contents page. - - - - - 7d71718b by Simon Marlow at 2003-11-06T16:50:28+00:00 [haddock @ 2003-11-06 16:50:28 by simonmar] Add -k/--package docs - - - - - ef43949d by Simon Marlow at 2003-11-06T16:51:23+00:00 [haddock @ 2003-11-06 16:51:23 by simonmar] Bump to 0.6 - - - - - 1c419e06 by Simon Marlow at 2003-11-06T16:51:50+00:00 [haddock @ 2003-11-06 16:51:50 by simonmar] update - - - - - 69422327 by Simon Marlow at 2003-11-10T14:41:06+00:00 [haddock @ 2003-11-10 14:41:05 by simonmar] Re-exporting names from a different package is problematic, because we don't have access to the full documentation for the entity. Currently Haddock just ignores entities with no documentation, but this results in bogus-looking empty documentation for many of the modules in the haskell98 package. So: - the documentation will now just list the name, as a link pointing to the location of the actual documentation. - now we don't attempt to link to these re-exported entities if they are referred to by the current module. Additionally: - If there is no documentation in the current module, include just the Synopsis section (rather than just the documentation section, as it was before). This just looks nicer and was on the TODO list. - - - - - 3c3fc433 by Simon Marlow at 2003-11-10T14:51:59+00:00 [haddock @ 2003-11-10 14:51:59 by simonmar] Fix for getReExports: take into account names which are not visible because they are re-exported from a different package. - - - - - 31c8437b by Simon Marlow at 2003-11-10T15:10:53+00:00 [haddock @ 2003-11-10 15:10:53 by simonmar] Version 0.6 changes - - - - - a7c2430b by Simon Marlow at 2003-11-10T15:15:58+00:00 [haddock @ 2003-11-10 15:15:58 by simonmar] getReExports: one error case that isn't - - - - - 00cc459c by Simon Marlow at 2003-11-10T16:15:19+00:00 [haddock @ 2003-11-10 16:15:18 by simonmar] copyright update - - - - - ca62408d by Simon Marlow at 2003-11-11T09:57:25+00:00 [haddock @ 2003-11-11 09:57:25 by simonmar] Version 0.6 - - - - - 3acbf818 by Simon Marlow at 2003-11-11T12:10:44+00:00 [haddock @ 2003-11-11 12:10:44 by simonmar] Go back to producing just the documentation section, rather than just the synopsis section, for a module with no documentation annotations. One reason is that the synopsis section tries to link each entity to its documentation on the same page. Also, the doc section anchors each entity, and it lists instances which the synopsis doesn't. - - - - - 6c90abc2 by Simon Marlow at 2003-11-12T10:03:39+00:00 [haddock @ 2003-11-12 10:03:39 by simonmar] 2002 -> 2003 - - - - - 090bbc4c by Simon Marlow at 2003-11-28T12:08:00+00:00 [haddock @ 2003-11-28 12:08:00 by simonmar] update - - - - - 8096a832 by Simon Marlow at 2003-11-28T12:09:58+00:00 [haddock @ 2003-11-28 12:09:58 by simonmar] Fix some of the problems with Haddock generating pages that are too wide. Now we only specify 'nowrap' when it is necessary to avoid a code box getting squashed up by the text to the right of it. - - - - - 35294929 by Sven Panne at 2003-12-29T17:16:31+00:00 [haddock @ 2003-12-29 17:16:31 by panne] Updated my email address - - - - - cdb697bf by Simon Marlow at 2004-01-08T10:14:24+00:00 [haddock @ 2004-01-08 10:14:24 by simonmar] Add instructions for using GHC to pre-process source for feeding to Haddock. - - - - - 8dfc491f by Simon Marlow at 2004-01-09T12:45:46+00:00 [haddock @ 2004-01-09 12:45:46 by simonmar] Add -optP-P to example ghc command line. - - - - - ac41b820 by Simon Marlow at 2004-02-03T11:02:03+00:00 [haddock @ 2004-02-03 11:02:03 by simonmar] Fix bug in index generation - - - - - f4e7edcb by Simon Marlow at 2004-02-10T11:51:16+00:00 [haddock @ 2004-02-10 11:51:16 by simonmar] Don't throw away whitespace at the beginning of a line (experimental fix). - - - - - 68e212d2 by Simon Marlow at 2004-02-10T12:10:08+00:00 [haddock @ 2004-02-10 12:10:08 by simonmar] Fix for previous commit: I now realise why the whitespace was stripped from the beginning of the line. Work around it. - - - - - e7d7f2df by Sven Panne at 2004-02-10T18:38:45+00:00 [haddock @ 2004-02-10 18:38:45 by panne] Make Haddock link with the latest relocated monad transformer package - - - - - 992d4225 by Simon Marlow at 2004-02-16T10:21:35+00:00 [haddock @ 2004-02-16 10:21:35 by simonmar] Add a TODO - - - - - 1ac55326 by Simon Marlow at 2004-03-12T11:33:39+00:00 [haddock @ 2004-03-12 11:33:39 by simonmar] Add an item. - - - - - 0478e903 by Simon Marlow at 2004-03-15T12:24:05+00:00 [haddock @ 2004-03-15 12:24:05 by simonmar] Add an item. - - - - - 6f26d21a by Simon Marlow at 2004-03-18T14:21:29+00:00 [haddock @ 2004-03-18 14:21:29 by simonmar] Fix URL - - - - - 19b6bb99 by Simon Marlow at 2004-03-22T14:09:03+00:00 [haddock @ 2004-03-22 14:09:03 by simonmar] getReExports was bogus: we should really look in the import_env to find the documentation for an entity which we are re-exporting without documentation. Suggested by: Ross Paterson (patch modified by me). - - - - - 5c756031 by Simon Marlow at 2004-03-24T09:42:11+00:00 [haddock @ 2004-03-24 09:42:10 by simonmar] hiding bug from Ross Paterson (fixed in rev 1.59 of Main.hs) - - - - - 1b692e6c by Simon Marlow at 2004-03-24T10:10:50+00:00 [haddock @ 2004-03-24 10:10:50 by simonmar] mkExportItems fix & simplification: we should be looking at the actual exported names (calculated earlier) to figure out which subordinates of a declaration are exported. This means that if you export a record, and name its fields separately in the export list, the fields will still be visible in the documentation for the constructor. - - - - - 90e5e294 by Simon Marlow at 2004-03-24T10:12:08+00:00 [haddock @ 2004-03-24 10:12:08 by simonmar] Make restrictCons take into account record field names too (removing a ToDo). - - - - - 2600efa4 by Simon Marlow at 2004-03-24T10:16:17+00:00 [haddock @ 2004-03-24 10:16:17 by simonmar] Record export tests. - - - - - 6a8575c7 by Simon Marlow at 2004-03-25T09:35:14+00:00 [haddock @ 2004-03-25 09:35:14 by simonmar] restrictTo: fix for restricting a newtype with a record field. - - - - - dcf55a8d by Simon Marlow at 2004-03-25T10:01:42+00:00 [haddock @ 2004-03-25 10:01:42 by simonmar] Fix duplicate instance bug - - - - - f49aa758 by Simon Marlow at 2004-03-25T10:02:41+00:00 [haddock @ 2004-03-25 10:02:41 by simonmar] Duplicate instance bug. - - - - - 7b87344c by Simon Marlow at 2004-03-25T10:29:56+00:00 [haddock @ 2004-03-25 10:29:56 by simonmar] If a name is imported from two places, one hidden and one not, choose the unhidden one to link to. Also, when there's only a hidden module to link to, don't try linking to it. - - - - - 40f44d7b by Simon Marlow at 2004-03-25T15:17:24+00:00 [haddock @ 2004-03-25 15:17:23 by simonmar] Add support for collaspible parts of the page, with a +/- button and a bit of JavaScript. Make the instances collapsible, and collapse them by default. This makes documentation with long lists of instances (eg. the Prelude) much easier to read. Maybe we should give other documentation sections the same treatment. - - - - - 9b64dc0f by Simon Marlow at 2004-03-25T15:20:55+00:00 [haddock @ 2004-03-25 15:20:55 by simonmar] Update - - - - - c2fff7f2 by Simon Marlow at 2004-03-25T15:45:10+00:00 [haddock @ 2004-03-25 15:45:10 by simonmar] Eliminate some unnecessary spaces in the HTML rendering - - - - - b7948ff0 by Simon Marlow at 2004-03-25T16:00:37+00:00 [haddock @ 2004-03-25 16:00:36 by simonmar] Remove all that indentation in the generated HTML to keep the file sizes down. - - - - - da2bb4ca by Sven Panne at 2004-03-27T09:57:58+00:00 [haddock @ 2004-03-27 09:57:57 by panne] Added the new-born haddock.js to the build process and the documentation. - - - - - b99e6f8c by Sven Panne at 2004-03-27T10:32:20+00:00 [haddock @ 2004-03-27 10:32:20 by panne] "type" is a required attribute of the "script" element - - - - - 562b185a by Sven Panne at 2004-03-27T12:52:34+00:00 [haddock @ 2004-03-27 12:52:34 by panne] Add a doctype for the contents page, too. - - - - - f6a99c2d by Simon Marlow at 2004-04-14T10:03:25+00:00 [haddock @ 2004-04-14 10:03:25 by simonmar] fix for single-line comment syntax - - - - - de366303 by Simon Marlow at 2004-04-20T13:08:04+00:00 [haddock @ 2004-04-20 13:08:04 by simonmar] Allow a 'type' declaration to include documentation comments. These will be ignored by Haddock, but at least one user (Johannes Waldmann) finds this feature useful, and it's easy to add. - - - - - fd78f51e by Simon Marlow at 2004-05-07T15:14:56+00:00 [haddock @ 2004-05-07 15:14:56 by simonmar] - update copyright - add version to abstract - - - - - 59f53e32 by Sven Panne at 2004-05-09T14:39:53+00:00 [haddock @ 2004-05-09 14:39:53 by panne] Fix the fix for single-line comment syntax, ------------------------------------------- is now a valid comment line again. - - - - - 8b18f2fe by Simon Marlow at 2004-05-10T10:11:51+00:00 [haddock @ 2004-05-10 10:11:51 by simonmar] Update - - - - - 225a491d by Ross Paterson at 2004-05-19T13:10:23+00:00 [haddock @ 2004-05-19 13:10:23 by ross] Make the handling of "deriving" slightly smarter, by ignoring data constructor arguments that are identical to the lhs. Now handles things like data Tree a = Leaf a | Branch (Tree a) (Tree a) deriving ... - - - - - 37588686 by Mike Thomas at 2004-05-21T06:38:14+00:00 [haddock @ 2004-05-21 06:38:14 by mthomas] Windows exe extensions (bin remains for Unix). - - - - - cf2b9152 by Simon Marlow at 2004-05-25T09:34:54+00:00 [haddock @ 2004-05-25 09:34:54 by simonmar] Add some TODO items - - - - - 4d29cdfc by Simon Marlow at 2004-05-25T10:41:46+00:00 [haddock @ 2004-05-25 10:41:46 by simonmar] Complain if -h is used with --gen-index or --gen-contents, because it'll overwrite the new index/contents. - - - - - 2e0771e0 by Mike Thomas at 2004-05-28T20:17:55+00:00 [haddock @ 2004-05-28 20:17:55 by mthomas] Windows: search for templates in executable directory. Unix: Haddock tries cwd first rather than error if no -l arg. - - - - - 8d10bde1 by Sven Panne at 2004-06-05T16:53:34+00:00 [haddock @ 2004-06-05 16:53:34 by panne] Misc. rpm spec file cleanup, including: * make BuildRoot handling more consistent * added default file attributes * consistent defines and tags - - - - - 59974349 by Sven Panne at 2004-06-05T18:01:00+00:00 [haddock @ 2004-06-05 18:01:00 by panne] More rpm spec file cleanup, including: * added some BuildRequires * changed packager to me, so people can complain at the right place :-] * consistently refer to haskell.org instead of www.haskell.org - - - - - b94d4903 by Simon Marlow at 2004-07-01T11:08:58+00:00 [haddock @ 2004-07-01 11:08:57 by simonmar] Update to the +/- buttons: use a resized image rather than a <button>. Still seeing some strange effects in Konqueror, so might need to use a fixed-size image instead. - - - - - d5278f67 by Sven Panne at 2004-07-04T15:15:55+00:00 [haddock @ 2004-07-04 15:15:55 by panne] Install pictures for +/- pictures, too (JPEG is a strange format for graphics like this, I would have expected GIF or PNG here.) Things look fine with Konqueror and Netscape on Linux now, the only downside is that the cursor doesn't change when positioned above the "button". - - - - - 46dec6c5 by Sven Panne at 2004-07-13T17:59:28+00:00 [haddock @ 2004-07-13 17:59:28 by panne] A quote is a valid part of a Haskell identifier, but it would interfere with an ECMA script string delimiter, so escape it there. - - - - - 1d7bc432 by Simon Marlow at 2004-07-22T08:54:06+00:00 [haddock @ 2004-07-22 08:54:06 by simonmar] Add single quote to $ident, so you can say eg. 'foldl'' to refer to foldl' (the longest match rule is our friend). Bug reported by Adrian Hey <ahey at iee.org> - - - - - f183618b by Krasimir Angelov at 2004-07-27T22:59:35+00:00 [haddock @ 2004-07-27 22:58:23 by krasimir] Add basic support for Microsoft HTML Help 2.0 - - - - - d515d0c2 by Krasimir Angelov at 2004-07-27T23:02:36+00:00 [haddock @ 2004-07-27 23:02:36 by krasimir] escape names in the index - - - - - a5f1be23 by Krasimir Angelov at 2004-07-27T23:05:21+00:00 [haddock @ 2004-07-27 23:05:21 by krasimir] Add jsFile, plusFile and minusFile to the file list - - - - - c4fb4881 by Krasimir Angelov at 2004-07-28T22:12:10+00:00 [haddock @ 2004-07-28 22:12:09 by krasimir] bugfix. Move contentsHtmlFile, indexHtmlFile and subIndexHtmlFile functions to HaddockUtil.hs module to make them accessible from HaddockHH2.hs - - - - - 64d30b1d by Krasimir Angelov at 2004-07-30T22:15:47+00:00 [haddock @ 2004-07-30 22:15:45 by krasimir] more stuffs - support for separated compilation of packages - the contents page now uses DHTML TreeView - fixed copyFile bug - - - - - 133c8c5c by Krasimir Angelov at 2004-07-31T12:04:38+00:00 [haddock @ 2004-07-31 12:04:37 by krasimir] make the DHtmlTree in contents page more portable. The +/- buttons are replaced with new images which looks more beatiful. - - - - - 79040963 by Krasimir Angelov at 2004-07-31T13:10:20+00:00 [haddock @ 2004-07-31 13:10:20 by krasimir] Make DHtmlTree compatible with Mozila browser - - - - - 1a55dc90 by Krasimir Angelov at 2004-07-31T14:52:55+00:00 [haddock @ 2004-07-31 14:52:55 by krasimir] fix - - - - - 85ce0237 by Krasimir Angelov at 2004-07-31T14:53:28+00:00 [haddock @ 2004-07-31 14:53:28 by krasimir] HtmlHelp 1.x - - - - - 3c0c53ba by Krasimir Angelov at 2004-07-31T20:35:21+00:00 [haddock @ 2004-07-31 20:35:21 by krasimir] Added support for DevHelp - - - - - d42b5af1 by Krasimir Angelov at 2004-07-31T21:17:51+00:00 [haddock @ 2004-07-31 21:17:51 by krasimir] Document new features in HtmlHelp - - - - - 790fe21e by Krasimir Angelov at 2004-08-01T15:14:02+00:00 [haddock @ 2004-08-01 15:14:02 by krasimir] add missing imports - - - - - fd7cc6bc by Krasimir Angelov at 2004-08-01T19:52:08+00:00 [haddock @ 2004-08-01 19:52:06 by krasimir] fix some bugs. Now I have got the entire libraries documentation in HtmlHelp 2.0 format. - - - - - 94ad7ac8 by Krasimir Angelov at 2004-08-01T19:53:50+00:00 [haddock @ 2004-08-01 19:53:50 by krasimir] I forgot to add the new +/- images - - - - - f0c65388 by Krasimir Angelov at 2004-08-02T16:25:53+00:00 [haddock @ 2004-08-02 16:25:53 by krasimir] Add root node to the table of contents. All modules in tree are not children of the root - - - - - f50bd85d by Sven Panne at 2004-08-02T18:17:46+00:00 [haddock @ 2004-08-02 18:17:46 by panne] Mainly DocBook fixes - - - - - 09527ce3 by Sven Panne at 2004-08-02T20:02:29+00:00 [haddock @ 2004-08-02 20:02:29 by panne] Fixed -o/--odir handling. Generating the output, especially the directory handling, is getting a bit convoluted nowadays... - - - - - c8fbacfa by Sven Panne at 2004-08-02T20:31:13+00:00 [haddock @ 2004-08-02 20:31:13 by panne] Warning police - - - - - 37830bff by Sven Panne at 2004-08-02T20:32:29+00:00 [haddock @ 2004-08-02 20:32:28 by panne] Nuked dead code - - - - - 13847171 by Sven Panne at 2004-08-02T21:12:27+00:00 [haddock @ 2004-08-02 21:12:25 by panne] Use pathJoin instead of low-level list-based manipulation for FilePaths - - - - - c711d61e by Sven Panne at 2004-08-02T21:16:02+00:00 [haddock @ 2004-08-02 21:16:02 by panne] Removed WinDoze CRs - - - - - b1f7dc88 by Sven Panne at 2004-08-03T19:35:59+00:00 [haddock @ 2004-08-03 19:35:59 by panne] Fixed spelling of "http-equiv" attribute - - - - - dd5f394e by Sven Panne at 2004-08-03T19:44:03+00:00 [haddock @ 2004-08-03 19:44:03 by panne] Pacify W3C validator: * Added document encoding (currently UTF-8, not sure if this is completely correct) * Fixed syntax of `id' attributes * Added necessary `alt' attribute for +/- images Small layout improvement: * Added space after +/- images (still not perfect, but better than before) - - - - - 919c47c6 by Sigbjorn Finne at 2004-08-03T19:45:11+00:00 [haddock @ 2004-08-03 19:45:11 by sof] make it compile with <= ghc-6.1 - - - - - 4d6f01d8 by Sigbjorn Finne at 2004-08-03T19:45:30+00:00 [haddock @ 2004-08-03 19:45:30 by sof] ffi wibble - - - - - 4770643a by Sven Panne at 2004-08-03T20:47:46+00:00 [haddock @ 2004-08-03 20:47:46 by panne] Fixed CSS for button style. Note that only "0" is a valid measure without a unit! - - - - - 14aaf2e5 by Sven Panne at 2004-08-03T21:07:59+00:00 [haddock @ 2004-08-03 21:07:58 by panne] Improved spacing of dynamic module tree - - - - - 97c3579a by Simon Marlow at 2004-08-09T11:03:04+00:00 [haddock @ 2004-08-09 11:03:04 by simonmar] Add FormatVersion Patch submitted by: George Russell <ger at informatik.uni-bremen.de> - - - - - af7f8c03 by Simon Marlow at 2004-08-09T11:55:07+00:00 [haddock @ 2004-08-09 11:55:05 by simonmar] Add support for a short description for each module, which is included in the contents. The short description should be given in a "Description: " field of the header. Included in this patch are changes that make the format of the header a little more flexible. From the comments: -- all fields in the header are optional and have the form -- -- [spaces1][field name][spaces] ":" -- [text]"\n" ([spaces2][space][text]"\n" | [spaces]"\n")* -- where each [spaces2] should have [spaces1] as a prefix. -- -- Thus for the key "Description", -- -- > Description : this is a -- > rather long -- > -- > description -- > -- > The module comment starts here -- -- the value will be "this is a .. description" and the rest will begin -- at "The module comment". The header fields must be in the following order: Module, Description, Copyright, License, Maintainer, Stability, Portability. Patches submitted by: George Russell <ger at informatik.uni-bremen.de>, with a few small changes be me, mostly to merge with other recent changes. ToDo: document the module header. - - - - - 7b865ad3 by Simon Marlow at 2004-08-10T14:09:57+00:00 [haddock @ 2004-08-10 14:09:57 by simonmar] Fixes for DevHelp/HtmlHelp following introduction of short module description. - - - - - 814766cd by Simon Marlow at 2004-08-10T14:33:46+00:00 [haddock @ 2004-08-10 14:33:45 by simonmar] Fixes to installation under Windows. - - - - - 39cf9ede by Simon Marlow at 2004-08-12T12:08:23+00:00 [haddock @ 2004-08-12 12:08:23 by simonmar] Avoid using string-gap tricks. - - - - - b6d78551 by Simon Marlow at 2004-08-13T10:53:21+00:00 [haddock @ 2004-08-13 10:53:21 by simonmar] Update - - - - - eaae7417 by Simon Marlow at 2004-08-13T10:53:50+00:00 [haddock @ 2004-08-13 10:53:50 by simonmar] Test for primes in quoted links - - - - - 68c34f06 by Sven Panne at 2004-08-16T19:59:38+00:00 [haddock @ 2004-08-16 19:59:36 by panne] XMLification - - - - - 7f45a6f9 by Sven Panne at 2004-08-18T16:42:54+00:00 [haddock @ 2004-08-18 16:42:54 by panne] Re-added indices + minor fixes - - - - - 8a5dd97c by Sigbjorn Finne at 2004-08-25T17:15:42+00:00 [haddock @ 2004-08-25 17:15:42 by sof] backquote HADDOCK_VERSION defn for <= ghc-6.0.x; believe this is only needed under mingw - - - - - 4b1b42ea by Sven Panne at 2004-08-26T20:08:50+00:00 [haddock @ 2004-08-26 20:08:49 by panne] SGML is dead, long live DocBook XML! Note: The BuildRequires tags in the spec files are still incomplete and the documentation about the DocBook tools needs to be updated, too. Stay tuned... - - - - - 8d52cedb by Sven Panne at 2004-08-26T21:03:19+00:00 [haddock @ 2004-08-26 21:03:19 by panne] Updated BuildRequires tags. Alas, there seems to be no real standard here, so your mileage may vary... At least the current specs should work on SuSE Linux. - - - - - e6982912 by Sigbjorn Finne at 2004-08-30T15:44:59+00:00 [haddock @ 2004-08-30 15:44:59 by sof] escape HADDOCK_VERSION double quotes on all platforms when compiling with <=6.0.x - - - - - b3fbc867 by Simon Marlow at 2004-08-31T13:09:42+00:00 [haddock @ 2004-08-31 13:09:42 by simonmar] Avoid GHC/shell versionitis and create Version.hs - - - - - c359e16a by Sven Panne at 2004-09-05T19:12:33+00:00 [haddock @ 2004-09-05 19:12:32 by panne] * HTML documentation for "foo.xml" goes into directory "foo" again, not "foo-html". This is nicer and consistent with the behaviour for building the docs from SGML. * Disabled building PostScript documentation in the spec files for now, there are some strange issues with the FO->PS conversion for some files which have to be clarified first. - - - - - c68b1eba by Sven Panne at 2004-09-24T07:04:38+00:00 [haddock @ 2004-09-24 07:04:38 by panne] Switched the default state for instances and the module hierarchy to non-collapsed. This can be reversed when we finally use cookies from JavaScript to have a more persistent state. Previously going back and forth in the documentation was simply too annoying because everything was collapsed again and therefore the documentation was not easily navigatable. - - - - - dfb32615 by Simon Marlow at 2004-09-30T08:21:29+00:00 [haddock @ 2004-09-30 08:21:29 by simonmar] Add a feature request - - - - - 45ff783c by Sven Panne at 2004-10-23T19:54:00+00:00 [haddock @ 2004-10-23 19:54:00 by panne] Improved the Cygwin/MinGW chaos a little bit. There is still confusion about host platform vs. target platform... - - - - - 5f644714 by Krasimir Angelov at 2004-10-28T16:01:51+00:00 [haddock @ 2004-10-28 16:01:51 by krasimir] update for ghc-6.3+ - - - - - 92d9753e by Sven Panne at 2004-11-01T16:39:01+00:00 [haddock @ 2004-11-01 16:39:01 by panne] Revert previous commit: It's Network.URI which should be changed, not Haddock. - - - - - 05f70f6e by Simon Marlow at 2005-01-04T16:15:51+00:00 [haddock @ 2005-01-04 16:15:51 by simonmar] parser fix: allow qualified specialids. - - - - - 47870837 by Simon Marlow at 2005-01-04T16:16:54+00:00 [haddock @ 2005-01-04 16:16:54 by simonmar] Add a test - - - - - ff11fc2c by Ross Paterson at 2005-01-10T19:18:22+00:00 [haddock @ 2005-01-10 19:18:22 by ross] Render non-ASCII characters using numeric character references, to simplify charset issues. There's a META tag saying the charset is UTF-8, but GHC outputs characters as raw bytes. Ideally we need an encoding on the input side too, primarily in comments, because source files containing non-ASCII characters aren't portable between locales. - - - - - eba2fc4e by Simon Marlow at 2005-01-11T10:44:37+00:00 [haddock @ 2005-01-11 10:44:37 by simonmar] Remove string gap - - - - - b899a381 by Ross Paterson at 2005-01-13T11:41:33+00:00 [haddock @ 2005-01-13 11:41:33 by ross] recognize SGML-style numeric character references &#ddd; or &#xhhhh; and translate them into Chars. - - - - - 106e3cf0 by Ross Paterson at 2005-01-13T14:43:41+00:00 [haddock @ 2005-01-13 14:43:41 by ross] also allow uppercase X in hexadecimal character references (like SGML) - - - - - e8f54f25 by Ross Paterson at 2005-01-13T14:44:24+00:00 [haddock @ 2005-01-13 14:44:24 by ross] Describe numeric character references. - - - - - 914ccdce by Sven Panne at 2005-01-15T18:44:48+00:00 [haddock @ 2005-01-15 18:44:45 by panne] Make Haddock compile again after the recent base package changed. The Map/Set legacy hell has been factored out, so that all modules can simply use the new non-deprecated interfaces. Probably a lot of things can be improved by a little bit of Map/Set/List algebra, this can be done later if needed. Small note: Currently the list of instances in HTML code is reversed. This will hopefully be fixed later. - - - - - 6ab20e84 by Sven Panne at 2005-01-16T12:18:26+00:00 [haddock @ 2005-01-16 12:18:26 by panne] Trim imports - - - - - efb81da9 by Sven Panne at 2005-01-16T12:58:08+00:00 [haddock @ 2005-01-16 12:58:03 by panne] Correctly handle the new order of arguments for the combining function given to fromListWith. - - - - - e27b5834 by Sven Panne at 2005-01-16T14:14:41+00:00 [haddock @ 2005-01-16 14:14:39 by panne] Data.Map.unions is left-biased. - - - - - dae3cc3e by Sven Panne at 2005-01-16T14:22:44+00:00 [haddock @ 2005-01-16 14:22:44 by panne] Added the last missing "flip" to get identical HTML output as previous versions. - - - - - 951d8408 by Sven Panne at 2005-01-16T14:37:10+00:00 [haddock @ 2005-01-16 14:37:10 by panne] Refactored Text.PrettyPrint legacy hell into a separate module. - - - - - f1c4b892 by Sven Panne at 2005-01-16T15:41:25+00:00 [haddock @ 2005-01-16 15:41:21 by panne] Cleaned up imports and dropped support for GHC < 5.03, it never worked, anyway. - - - - - 60824c6e by Simon Marlow at 2005-01-18T10:02:48+00:00 [haddock @ 2005-01-18 10:02:48 by simonmar] Add a TODO - - - - - a8c82f23 by Krasimir Angelov at 2005-01-28T23:19:39+00:00 [haddock @ 2005-01-28 23:19:39 by krasimir] import Foreign/Foreign.C are required for Windows - - - - - d8450a23 by Simon Marlow at 2005-02-02T16:23:04+00:00 [haddock @ 2005-02-02 16:23:00 by simonmar] Revamp the linking strategy in Haddock. Now name resolution is done in two phases: - first resolve everything to original names, like a Haskell compiler would. - then, figure out the "home" location for every entity, and point all the links to there. The home location is the lowest non-hidden module in the import hierarchy that documents the entity. If there are multiple candidates, one is chosen at random. Also: - Haddock should not generate any HTML with dangling links any more. Unlinked references are just rendered as plain text. - Error reporting is better: if we can't find a link destination for an entity reference, we now emit a warning. - - - - - 1cce71d0 by Simon Marlow at 2005-02-03T13:42:19+00:00 [haddock @ 2005-02-03 13:42:19 by simonmar] - add --ignore-all-exports flag, which behaves as if every module has the ignore-exports attribute (requested by Chris Ryder). - add --hide option to hide a module on the command line. - add --use-package option to get Haddock info for a package from ghc-pkg (largely untested). - remove reexports from the .haddock file, they aren't used any more. - - - - - 767123ef by Ross Paterson at 2005-02-03T16:17:37+00:00 [haddock @ 2005-02-03 16:17:37 by ross] fix typo for < 6.3 - - - - - 0c680c04 by Simon Marlow at 2005-02-04T12:03:31+00:00 [haddock @ 2005-02-04 12:03:31 by simonmar] Fix bug in renameExportItems that meant links in instances weren't being renamed properly. - - - - - ff7abe5f by Simon Marlow at 2005-02-04T12:15:53+00:00 [haddock @ 2005-02-04 12:15:52 by simonmar] Add attribute #not-home, to indicate that the current module should not be considered to be a home module for the each entity it exports, unless there is no other module that exports the entity. - - - - - fc2cfd27 by Simon Marlow at 2005-02-04T12:40:02+00:00 [haddock @ 2005-02-04 12:40:02 by simonmar] Update the documentation w.r.t. home modules and the not-home attribute. - - - - - 26b8ddf7 by Ross Paterson at 2005-02-04T13:36:06+00:00 [haddock @ 2005-02-04 13:36:05 by ross] sort lists of instances by - arity of the type constructors (so higher-kinded instances come first) - name of the class - argument types - - - - - 26bfb19c by Simon Marlow at 2005-02-23T15:57:12+00:00 [haddock @ 2005-02-23 15:57:12 by simonmar] Fix documentation regarding the module attributes. - - - - - 9c3afd02 by Simon Marlow at 2005-02-28T16:18:17+00:00 [haddock @ 2005-02-28 16:18:17 by simonmar] version 0.7 - - - - - a95fd63f by Simon Marlow at 2005-02-28T16:22:08+00:00 [haddock @ 2005-02-28 16:22:08 by simonmar] Attempt to fix the layout of the package names in the contents. Having tried just about everything, the only thing I can get to work reliably is to make the package names line up on a fixed offset from the left margin. This obviously isn't ideal, so anyone else that would like to have a go at improving it is welcome. One option is to remove the +/- buttons from the contents list and go back to a plain table. The contents page now uses CSS for layout rather than tables. It seems that most browsers have different interpretations of CSS layout, so only the simplest things lead to consistent results. - - - - - 905d42f7 by Simon Marlow at 2005-03-01T17:16:42+00:00 [haddock @ 2005-03-01 17:16:40 by simonmar] Another attempt at lining up the package names on the contents page. Now, they line up with Konqueror, and almost line up with Firefox & IE (different layout in each case). - - - - - a0e1d178 by Wolfgang Thaller at 2005-03-09T08:28:39+00:00 [haddock @ 2005-03-09 08:28:39 by wolfgang] Hack haddock's lexer to accept the output from Apple's broken version of cpp (Apple's cpp leaves #pragma set_debug_pwd directives in it's output). - - - - - 9e1eb784 by Simon Marlow at 2005-04-22T14:27:15+00:00 [haddock @ 2005-04-22 14:27:15 by simonmar] Add a TODO item - - - - - 23281f78 by Ross Paterson at 2005-05-18T12:41:59+00:00 [haddock @ 2005-05-18 12:41:59 by ross] fix 3 bugs in --use-package, and document it. - - - - - 00074a68 by Sven Panne at 2005-05-21T12:35:29+00:00 [haddock @ 2005-05-21 12:35:29 by panne] Warning/versionitis police - - - - - 341fa822 by Simon Marlow at 2005-06-15T15:43:21+00:00 [haddock @ 2005-06-15 15:43:21 by simonmar] Allow "licence" as an alternate spelling of "license" - - - - - 3b953f8b by Simon Marlow at 2005-06-16T08:14:12+00:00 [haddock @ 2005-06-16 08:14:12 by simonmar] wibble - - - - - abfd9826 by Simon Marlow at 2005-06-27T14:46:40+00:00 [haddock @ 2005-06-27 14:46:40 by simonmar] name hierarchical HTML files as A-B-C.html instead of A.B.C.html. The old way confused Apache because the extensions are sometimes interpreted as having special meanings. - - - - - a01eea00 by Simon Marlow at 2005-08-04T13:59:40+00:00 [haddock @ 2005-08-04 13:59:40 by simonmar] 0.7 changes - - - - - 170ef87e by Simon Marlow at 2005-08-04T15:08:03+00:00 [haddock @ 2005-08-04 15:08:03 by simonmar] spec file from Jens Peterson - - - - - 7621fde4 by Simon Marlow at 2005-08-04T15:59:30+00:00 [haddock @ 2005-08-04 15:59:30 by simonmar] replace mingw tests with $(Windows) - - - - - a20739bb by Sven Panne at 2005-08-05T07:01:12+00:00 [haddock @ 2005-08-05 07:01:12 by panne] Reverted to previous version (but with bumped version number), the last commit broke RPM building on SuSE systems due to differently named dependencies. As a clarification: All .spec files in the repository have to work at least on SuSE, because that's the system I'm using. And as "Mr. Building Police", I reserve me the right to keep them that way... >:-) It might very well be the case that we need different .spec files for different platforms, so packagers which are unhappy with the current .spec files should contact me, stating the actual problems. - - - - - 4afb15cf by Simon Marlow at 2005-10-05T10:51:45+00:00 [haddock @ 2005-10-05 10:51:45 by simonmar] Add a bug - - - - - 60f69f82 by Simon Marlow at 2005-10-05T12:52:03+00:00 [haddock @ 2005-10-05 12:52:03 by simonmar] Document new behaviour of -s option - - - - - f7e520ca by Simon Marlow at 2005-10-10T15:02:55+00:00 [haddock @ 2005-10-10 15:02:55 by simonmar] extractRecSel: ignore non-record constructors (fixes a crash when using datatypes with a mixture of record and non-record style constructors). - - - - - b2edbedb by Simon Marlow at 2005-10-14T09:44:21+00:00 Start CHANGES for 0.8 - - - - - 21c7ac8d by Simon Marlow at 2005-10-14T23:11:19+00:00 First cut of Cabal build system - - - - - 766cecdd by Simon Marlow at 2005-10-29T08:14:43+00:00 Add configure script and Makefile for the docs Add a separate configure script and build system for building the documentation. The configure and Makefile code is stolen from fptools. This is left as a separate build system so that the main Cabal setup doesn't require a Unix build environment or DocBook XML tools. - - - - - aa36c783 by Duncan Coutts at 2006-01-17T19:29:55+00:00 Add a --wiki=URL flag to add a per-module link to a correspondng wiki page. So each html page gets an extra link (placed next to the source code and contents links) to a corresponding wiki page. The idea is to let readers contribute their own notes, examples etc to the documentation. Also slightly tidy up the code for the --source option. - - - - - e06e2da2 by Simon Marlow at 2006-01-18T09:28:15+00:00 TODO: documnet --wiki - - - - - 17adfda9 by Duncan Coutts at 2006-01-19T20:17:59+00:00 Add an optional wiki link for each top level exported name. In each module, for each "top level" exported entity we add a hyper link to a corresponding wiki page. The link url gets the name of the exported entity as a '#'-style anchor, so if there is an anchor in the page with that name then the users browser should jump directly to it. By "top level" we mean functions, classes, class members and data types (data, type, newtype), but not data constructors, class instances or data type class membership. The link is added at the right of the page and in a small font. Hopefully this is the right balance of visibility/distraction. We also include a link to the wiki base url in the contents and index pages. - - - - - f52324bb by Duncan Coutts at 2006-01-19T20:28:27+00:00 Rewrite pathJoin to only add a path separator when necessary. When the path ends in a file seperator there is no need to add another. Now using "--wiki=http://blah.com/foo/" should do the right thing. (Code snippet adapted from Isaac's FilePath package.) - - - - - 43bb89fa by Duncan Coutts at 2006-01-21T17:15:27+00:00 Teach haddock about line pragmas and add accurate source code links Teach haddock about C and Haskell style line pragmas. Extend the lexer/parser's source location tracking to include the file name as well as line/column. This way each AST item that is tagged with a SrcLoc gets the original file name too. Use this original file name to add source links to each exported item, in the same visual style as the wiki links. Note that the per-export source links are to the defining module rather than whichever module haddock pretends it is exported from. This is what we want for source code links. The source code link URL can also contain the name of the export so one could implement jumping to the actual location of the function in the file if it were linked to an html version of the source rather than just plain text. The name can be selected with the %N wild card. So for linking to the raw source code one might use: --source=http://darcs/haskell.org/foo/%F Or for linking to html syntax highlighted code: --source=http://darcs/haskell.org/foo/%M.html#%N - - - - - edd9f229 by Duncan Coutts at 2006-01-22T00:02:00+00:00 Extend URL variable expansion syntax and add source links to the contents page Like the wiki link on the contents and index page, add a source code link too. Extend the wiki & source URL variable expansion syntax. The original syntax was: %F for the source file name (the .hs version only, not the .lhs or .hs.pp one) %M for the module name (with '.' replaced by '/') The new syntax is: %F or %{FILE} for the original source file name %M or %{MODULE} for the module name (no replacements) %N or %{NAME} for the function/type export name %K or %{KIND} for a type/value flag "t" or "v" with these extensions: %{MODULE/./c} to replace the '.' module seperator with any other char c %{VAR|some text with the % char in it} which means if the VAR is not in use in this URL context then "" else replace the given text with the '%' char replaced by the string value of the VAR. This extension allows us to construct URLs wit optional parts, since the module/file name is not available for the URL in the contents/index pages and the value/type name is not available for the URL at the top level of each module. - - - - - eb3c6ada by Duncan Coutts at 2006-01-23T13:42:34+00:00 Remove the complex substitutions and add more command line flags instead. Instead of incomprehensable URL substitutions like ${MODULE/./-|?m=%} we now use three seperate command line flags for the top level, per-module and per-entity source and wiki links. They are: --source-base, --source-module, --source-entity --comments-base, --comments-module, --comments-entity We leave -s, --source as an alias for --source-module which is how that option behaved previously. The long forms of the substitutions are still available, ${FILE} ${MODULE} etc and the only non-trivial substitution is ${MODULE/./c} to replace the '.' characters in the module name with any other character c. eg ${MODULE/./-} Seperating the source and wiki url flags has the added bonus that they can be turned on or off individually. So users can have per-module links for example without having to also have per-entity links.` - - - - - a2f0f2af by Duncan Coutts at 2006-01-23T13:54:52+00:00 Make the --help output fit in 80 columns. This is a purely cosmetic patch, feel free to ignore it. The only trickery going on is that we don't display the deprecated -s, --source flags in the help message, but we do still accept them. - - - - - 2d3a4b0c by Duncan Coutts at 2006-01-23T14:12:16+00:00 Add documentation for the new --source-* and --comments-* command line options - - - - - 1a82a297 by Simon Marlow at 2006-01-23T17:03:27+00:00 fix markup - - - - - 100d464a by Duncan Coutts at 2006-01-23T18:31:13+00:00 remove a couple TODO items that have been done The --wiki, or rather the --comment-* options are now documented. There is probably no need to have haddock invoke unlit or cpp itself since it can now pick up the line pragmas to get the source locations right. Tools like Cabal will arrange for preprocessors to be run so there is less of a need for tools like haddock to do it themselves. - - - - - 3162fa91 by Simon Marlow at 2006-01-24T14:21:56+00:00 add a test I had lying around - - - - - 98947063 by Simon Marlow at 2006-01-31T13:52:54+00:00 add scabal-version field - - - - - c41876e6 by Neil Mitchell at 2006-02-26T17:48:21+00:00 Add Hoogle output option - - - - - f86fb9c0 by Simon Marlow at 2006-03-08T09:15:20+00:00 add haskell.vim Contributed by Brad Bowman <bsb at bereft.net>, thanks! - - - - - 35d3c511 by benjamin.franksen at 2006-03-03T22:39:54+00:00 fixed libdir (/html was missing) - - - - - 4d08fd7d by Simon Marlow at 2006-03-10T11:13:31+00:00 add PatternGuards extension - - - - - 3f095e70 by Simon Marlow at 2006-03-13T11:40:42+00:00 bug fixes from Brad Bowman - - - - - 8610849d by Sven Panne at 2006-03-19T17:02:56+00:00 Fixed Cabal/RPM build - - - - - 34a994d6 by sven.panne at 2006-04-20T12:39:23+00:00 Avoid pattern guards Due to the use of pattern guards in Haddock, GHC was called with -fglasgow-exts. This in turn enables bang patterns, too, which broke the Haddock build. Removing some unnecessary pattern guards seemed to be the better way of fixing this instead of using a pragma to disable pattern guards. - - - - - bb523f51 by Ross Paterson at 2006-04-24T09:03:25+00:00 extend 'deriving' heuristic a little If an argument of a data constructor has a type variable head, it is irreducible and the same type class can be copied into the constraint. (Formerly we just did this for type variable arguments.) - - - - - dab9fe7a by Simon Marlow at 2006-04-26T10:02:31+00:00 record an idea - - - - - 748b7078 by Simon Marlow at 2006-05-08T08:28:53+00:00 add section about deriving - - - - - 11252ea1 by Simon Marlow at 2006-05-24T15:43:10+00:00 replace a fatal error in lexChar with a parseError - - - - - 382c9411 by Simon Marlow at 2006-05-24T15:45:47+00:00 add a bug - - - - - b79272f5 by Simon Marlow at 2006-05-24T15:46:29+00:00 add a bug report - - - - - 912edf65 by David Waern at 2006-07-10T19:09:23+00:00 Initial modifications -- doesn't compile - - - - - a3c7ba99 by David Waern at 2006-07-11T00:54:19+00:00 More porting work -- doesn't compile - - - - - 0a173d19 by David Waern at 2006-07-11T11:30:03+00:00 Make the repos temporarily compile and illustrate a problem - - - - - bad316de by David Waern at 2006-07-11T15:43:47+00:00 Progress on the porting process - - - - - bbf12d02 by David Waern at 2006-07-11T23:07:44+00:00 More progress on the porting -- first pass starting to shape up - - - - - de580ba2 by David Waern at 2006-07-20T17:48:30+00:00 More progress -- still on phase1 - - - - - 75a917a2 by David Waern at 2006-07-23T18:22:43+00:00 More work on pass1 -- mostly done - - - - - 6697b3f7 by David Waern at 2006-07-23T22:17:40+00:00 More work, started working on the renaming phase -- this code will need a cleanup soon :) - - - - - 82a5bcbb by David Waern at 2006-07-29T16:16:43+00:00 Add instances, build renaming environment, start on the renamer - - - - - c3f8f4f1 by David Waern at 2006-07-29T21:37:48+00:00 Complete the renamer - - - - - 7e00d464 by David Waern at 2006-07-30T21:01:57+00:00 Start porting the Html renderer - - - - - f04ce121 by David Waern at 2006-08-09T20:04:56+00:00 More Html rendering progress - - - - - 20c21b53 by David Waern at 2006-08-10T17:37:47+00:00 More progress - - - - - d7097e0d by David Waern at 2006-08-11T20:31:51+00:00 Cleanup - - - - - a7351e86 by David Waern at 2006-08-12T11:44:47+00:00 Render H98 Data declarations - - - - - 3fb2208e by David Waern at 2006-08-12T17:15:34+00:00 Perfect rendering of Test.hs - - - - - 454fd062 by David Waern at 2006-08-13T21:57:08+00:00 Misc fixes and interface load/save - - - - - 7ef7e7be by David Waern at 2006-08-14T00:56:07+00:00 Some refactoring - - - - - a7d3efef by David Waern at 2006-08-19T20:07:55+00:00 Adapt to latest GHC - - - - - 5fc3c0d7 by David Waern at 2006-08-20T21:28:11+00:00 Move interface read/write to its own module + some cleanup - - - - - 037e011c by David Waern at 2006-08-20T21:38:24+00:00 Small cleanup - - - - - da3a1023 by David Waern at 2006-09-03T16:05:22+00:00 Change mode to BatchCompile to avoid GHC API bug - - - - - 3cc9be3b by David Waern at 2006-09-03T16:06:59+00:00 Starting work on GADT rendering - - - - - 94506037 by David Waern at 2006-09-03T20:02:48+00:00 Compensate for change of export list order in GHC - - - - - c2cec4eb by David Waern at 2006-09-04T20:53:01+00:00 Rename a function - - - - - 9a9735ba by David Waern at 2006-09-05T15:51:21+00:00 Change version number to 2.0 - - - - - 3758a714 by David Waern at 2006-09-05T15:51:49+00:00 Align comment properly - - - - - 68478d9e by David Waern at 2006-09-15T18:03:00+00:00 Remove interface reading/writing code and use the GHC api for creating package environments instead - - - - - d2eedd95 by David Waern at 2006-09-15T18:05:29+00:00 Change the executable name to haddock-ghc-nolib - - - - - fcfbcf66 by David Waern at 2006-09-15T18:05:45+00:00 Small source code cleanup - - - - - d08eb017 by David Waern at 2006-09-15T18:06:21+00:00 Remove handling of --package flag - - - - - b8a4cf53 by David Waern at 2006-09-15T18:07:16+00:00 Remove commented-out code - - - - - bef0a684 by David Waern at 2006-09-15T18:37:57+00:00 Don't warn about missing links to () - - - - - e7d25fd7 by David Waern at 2006-09-15T19:50:49+00:00 Remove Interface and Binary2 modules - - - - - 9894f2a1 by David Waern at 2006-09-15T19:53:43+00:00 Remove debug printing from HaddockHtml - - - - - a0e7455d by David Waern at 2006-09-16T00:16:29+00:00 Comments only - - - - - d5b26fa7 by David Waern at 2006-09-16T00:16:57+00:00 Refactor PackageData creation code and start on building the doc env propery (unfinished) - - - - - 06aaa779 by David Waern at 2006-09-16T00:19:25+00:00 Better comments in Main.hs - - - - - 1a52d1b4 by David Waern at 2006-09-18T22:17:11+00:00 Comments and spacing change - - - - - e5a97767 by David Waern at 2006-09-21T17:02:45+00:00 Remove unnecessary fmapM import in Main - - - - - 9d0f9d3a by David Waern at 2006-09-22T18:07:07+00:00 Make import list in HaddockHtml prettier - - - - - 3452f662 by David Waern at 2006-09-22T18:08:47+00:00 Refactor context rendering - - - - - 12d0a6d0 by David Waern at 2006-09-22T18:09:52+00:00 Do proper HsType rendering (inser parentheses correctly) - - - - - 2c20c2f9 by David Waern at 2006-09-22T18:10:45+00:00 Fix a bug in Main.toHsType - - - - - c5396443 by David Waern at 2006-09-22T18:11:16+00:00 Skip external package modules sort for now - - - - - 3fb95547 by David Waern at 2006-09-22T20:35:40+00:00 Take away trailin "2" on all previously clashing type names - - - - - 2174755f by David Waern at 2006-09-22T20:51:43+00:00 Remove unused imports in Main - - - - - 1e9f7a39 by David Waern at 2006-09-22T20:52:11+00:00 Fix a comment in Main - - - - - 32d9e028 by David Waern at 2006-10-05T16:40:11+00:00 Merge with changes to ghc HEAD - - - - - 3058c8f5 by David Waern at 2006-10-05T16:41:02+00:00 Comment fixes - - - - - b9c217ec by David Waern at 2006-10-05T16:49:59+00:00 Filter out more builtin type constructors from warning messages - - - - - 67e7d252 by David Waern at 2006-10-05T19:38:22+00:00 Refactoring -- better structured pass1 - - - - - cd21c0c1 by David Waern at 2006-10-05T19:44:42+00:00 Remove read/dump interface flags - - - - - 313f9e69 by David Waern at 2006-10-05T19:49:26+00:00 Remove unused pretty printing - - - - - 480f09d1 by David Waern at 2006-12-28T13:22:24+00:00 Update to build with latest GHC HEAD - - - - - 63dccfcb by David Waern at 2007-01-05T01:38:45+00:00 Fixed a bug so that --ghc-flag works correctly - - - - - 3117dadc by David Waern at 2006-12-29T18:53:39+00:00 Automatically get the GHC lib dir - - - - - 9dc84a5c by David Waern at 2006-12-29T19:58:53+00:00 Comments - - - - - 0b0237cc by David Waern at 2007-01-05T16:48:30+00:00 Collect docs based on SrcLoc, syncing with removal of DeclEntity from GHC - - - - - a962c256 by David Waern at 2007-01-05T17:02:47+00:00 Add tabs in haddock.cabal - - - - - 0ca30c97 by David Waern at 2007-01-05T17:04:11+00:00 Add GHCUtils.hs - - - - - c0ab9abe by David Waern at 2007-01-10T11:43:08+00:00 Change package name to haddock-ghc, version 0.1 - - - - - 38e18b27 by David Waern at 2007-01-12T12:03:52+00:00 No binder name for foreign exports - - - - - d18587ab by David Waern at 2007-01-12T12:08:15+00:00 Temp record - - - - - ba6251a0 by David Waern at 2007-01-12T18:27:55+00:00 Remove read/dump-interface (again) - - - - - f4ba2b39 by David Waern at 2007-01-12T18:31:36+00:00 Remove DocOption, use the GHC type - - - - - 511be8bd by David Waern at 2007-01-12T18:32:41+00:00 Use exceptions instead of Either when loading package info - - - - - 0f2144d8 by David Waern at 2007-01-12T18:33:23+00:00 Small type change - - - - - 77507eb7 by David Waern at 2007-01-12T18:33:59+00:00 Remove interface file read/write - - - - - 0ea1e14f by David Waern at 2007-01-17T21:40:26+00:00 Add trace_ppr to GHCUtils - - - - - 3878b493 by David Waern at 2007-01-17T21:40:53+00:00 Sort external package modules and build a doc env - - - - - 8dc323fc by David Waern at 2007-01-17T21:42:41+00:00 Remove comment - - - - - f4c5b097 by David Waern at 2007-01-18T23:22:18+00:00 Add haddock-ghc.cabal and remove ghc option pragma in source file - - - - - da242b2c by David Waern at 2007-01-18T23:22:46+00:00 Remove some tabs - - - - - 288ed096 by David Waern at 2007-01-18T23:39:28+00:00 Moved the defaultErrorHandler to scope only over sortAndCheckModules for now - - - - - 4dd150fe by David Waern at 2007-02-03T21:23:56+00:00 Let restrictCons handle infix constructors - - - - - 97893442 by David Waern at 2007-02-04T16:26:00+00:00 Render infix data constructors - - - - - da89db72 by David Waern at 2007-02-04T16:26:33+00:00 CHange project name to Haddock-GHC - - - - - e93d48af by David Waern at 2007-02-04T16:59:08+00:00 Render infix type constructors properly - - - - - 357bc99b by David Waern at 2007-02-04T17:37:08+00:00 Insert spaces around infix function names - - - - - ab6cfc49 by David Waern at 2007-02-04T17:59:54+00:00 Do not list entities without documentation - - - - - 04249c7e by David Waern at 2007-02-04T19:16:25+00:00 Add GADT support (quite untested) - - - - - 2c223f8d by David Waern at 2007-02-04T19:25:10+00:00 Add package file write/save again! - - - - - b07ed218 by David Waern at 2007-02-04T19:33:02+00:00 Comment out minf_iface based stuff - - - - - 953d1fa7 by David Waern at 2007-02-05T00:12:23+00:00 Solve conflicts - - - - - 593247fc by David Waern at 2007-02-06T19:48:48+00:00 Remove -package flag, GHC's can be used instead - - - - - f658ded2 by David Waern at 2007-02-06T20:50:44+00:00 Start for support of ATs - - - - - 97f9e913 by David Waern at 2007-02-06T20:52:27+00:00 Wibble - - - - - 2ce8e4cf by David Waern at 2007-02-16T12:09:49+00:00 Add the DocOptions change - - - - - dee4a9b5 by David Waern at 2007-03-06T01:24:48+00:00 Wibble - - - - - 7cb99d18 by David Waern at 2007-03-06T01:24:58+00:00 Change version to 2.0 and executable name to haddock - - - - - c5aa02bc by David Waern at 2007-03-08T15:59:49+00:00 Go back to -B flag - - - - - 3a349201 by David Waern at 2007-03-09T13:31:59+00:00 Better exception handling and parsing of GHC flags - - - - - 05a69b71 by David Waern at 2007-03-09T17:45:44+00:00 Remove commented-out DocEntity printing - - - - - 755032cb by davve at dtek.chalmers.se at 2007-03-23T23:30:20+00:00 Remove a file that shouldn't be here - - - - - a7077e5f by davve at dtek.chalmers.se at 2007-03-24T03:58:48+00:00 Remove an import - - - - - 6f55aa8b by davve at dtek.chalmers.se at 2007-03-25T00:46:48+00:00 Start work on Haddock API - - - - - f0199480 by davve at dtek.chalmers.se at 2007-03-25T00:56:36+00:00 Prettify some comments - - - - - f952f9d1 by davve at dtek.chalmers.se at 2007-03-25T00:56:53+00:00 Remove ppr in HaddockTypes - - - - - bc594904 by davve at dtek.chalmers.se at 2007-03-25T00:57:53+00:00 Remove commented out doc env inference - - - - - 11ebf08d by davve at dtek.chalmers.se at 2007-03-25T01:23:25+00:00 De-flatten the namespace - - - - - f696b4bc by davve at dtek.chalmers.se at 2007-03-25T03:21:48+00:00 Add missing stuff to API - - - - - 9a2a04c3 by davve at dtek.chalmers.se at 2007-03-25T03:22:02+00:00 Wibble - - - - - 7d04a6d5 by davve at dtek.chalmers.se at 2007-03-25T03:22:08+00:00 Avoid a GHC bug with parseStaticFlags [] - - - - - 4d2820ba by davve at dtek.chalmers.se at 2007-03-26T04:57:01+00:00 Add fall-through case to mkExportItem - - - - - 6ebc8950 by Stefan O'Rear at 2007-03-26T04:14:53+00:00 Add shebang line to Setup.lhs - - - - - 80966ec5 by davve at dtek.chalmers.se at 2007-03-26T05:24:26+00:00 Fix stupid compile error - - - - - 1ea1385d by davve at dtek.chalmers.se at 2007-04-05T17:19:56+00:00 Do save/read of interface files properly - - - - - 0e4f6541 by David Waern at 2007-04-10T21:08:36+00:00 Add version to ghc dependency - - - - - b0499b63 by David Waern at 2007-04-10T21:37:08+00:00 Change package name to haddock - - - - - 9d50d27e by David Waern at 2007-04-24T00:22:14+00:00 Use filepath package instead of FilePath - - - - - 87c7fcdf by David Waern at 2007-07-10T21:03:04+00:00 Add new package dependencies - - - - - 4768709c by David Waern at 2007-07-11T20:37:11+00:00 Follow changes to record constructor representation - - - - - b9a02fee by Simon Marlow at 2007-05-30T14:00:48+00:00 update to compile with the latest GHC & Cabal - - - - - c0ebdc01 by David Waern at 2007-07-11T21:35:45+00:00 Fix conflicts - - - - - 97f7afd4 by David Waern at 2007-07-11T21:52:38+00:00 Follow changes to the GHC API - - - - - a5b7b58f by David Waern at 2007-07-12T20:36:48+00:00 Call parseStaticFlags before newSession - - - - - f7f50dbc by David Waern at 2007-08-01T21:52:58+00:00 Better indentation in haddock.cabal - - - - - d84e52ad by David Waern at 2007-08-02T00:08:18+00:00 Wibble - - - - - a23f494a by David Waern at 2007-08-02T00:08:24+00:00 Be better at trying to load all module dependencies (debugging) - - - - - ee917f13 by David Waern at 2007-08-03T18:48:08+00:00 Load all targets explicitly (checkModule doesn't chase dependencies anymore) - - - - - 5182d631 by David Waern at 2007-08-16T16:48:55+00:00 Finalize support for links to other packages - - - - - dfd1e3da by David Waern at 2007-08-16T16:51:11+00:00 Fix haddock comment errors in Haddock.Types - - - - - 50c0d83e by David Waern at 2007-08-16T16:51:37+00:00 Remove a debug import - - - - - d84b7c2b by David Waern at 2007-08-16T17:06:30+00:00 Rename PackageData to HaddockPackage - - - - - 3b52cb9f by David Waern at 2007-08-16T17:09:42+00:00 Simplify some comments - - - - - 66fa68d9 by David Waern at 2007-08-16T17:11:38+00:00 Comment the HaddockPackage definition - - - - - 8674c761 by David Waern at 2007-08-16T17:25:54+00:00 Improve code layout in Main - - - - - 571a3a0b by David Waern at 2007-08-16T17:32:13+00:00 Remove explict module imports in Main - - - - - d31b3cb0 by David Waern at 2007-08-16T17:36:23+00:00 Correct comments - - - - - 7f8a9f2b by David Waern at 2007-08-16T17:39:50+00:00 Fix layout problems in Haddock.Types - - - - - 9f421d7f by David Waern at 2007-08-17T11:16:48+00:00 Move options out of Main into Haddock.Options - - - - - 80042b63 by David Waern at 2007-08-17T11:26:59+00:00 Small comment/layout fixes - - - - - b141b982 by David Waern at 2007-08-17T11:28:28+00:00 Change project name from Haddock-GHC to Haddock - - - - - dbeb4a81 by David Waern at 2007-08-17T11:41:05+00:00 Add top module comment to all files - - - - - ce99cc9e by David Waern at 2007-08-17T14:53:04+00:00 Factor out typechecking phase into Haddock.Typecheck - - - - - 6bf75d9e by David Waern at 2007-08-17T16:55:35+00:00 Factor out package code to Haddock.Packages - - - - - b396db37 by David Waern at 2007-08-29T22:40:23+00:00 Major refactoring - - - - - 3d4f95ee by David Waern at 2007-08-29T23:26:24+00:00 Rename HaddockModule to Interface and a few more refactorings - - - - - c55326db by David Waern at 2007-08-29T23:48:03+00:00 Some comment cleanup - - - - - 9a84fc46 by David Waern at 2007-08-29T23:49:29+00:00 Add some modules that I forgot to add earlier - - - - - 4536dce2 by David Waern at 2007-08-29T23:55:24+00:00 Wibble - - - - - 9b7f0206 by David Waern at 2007-08-30T16:03:29+00:00 Wibble - - - - - c52c050a by David Waern at 2007-08-30T16:30:37+00:00 Rename HaddockModule to Interface - - - - - eae2995f by David Waern at 2007-08-30T16:42:59+00:00 Simplify createInterfaces - - - - - 53f99caa by David Waern at 2007-09-29T00:04:31+00:00 Add build-type: Simple to the cabal file - - - - - 0d3103a8 by David Waern at 2007-09-29T00:04:58+00:00 Add containers and array dependency - - - - - 6acf5f30 by David Waern at 2007-09-29T00:13:36+00:00 Prettify the cabal file - - - - - 87c1e378 by David Waern at 2007-09-29T13:16:39+00:00 FIX: consym data headers with more than two variables - - - - - b67fc16a by David Waern at 2007-09-29T14:01:32+00:00 FIX: prefix types used as operators should be quoted - - - - - a8f925bc by David Waern at 2007-09-29T14:02:26+00:00 Use isSymOcc from OccName instead of isConSym - - - - - fc330701 by David Waern at 2007-09-29T14:15:37+00:00 Use isLexConSym/isLexVarSym from OccName - - - - - e4f3dbad by David Waern at 2007-09-29T15:01:08+00:00 FIX: do not quote varsym type operators - - - - - 402207d2 by David Waern at 2007-09-29T15:01:50+00:00 Wibble - - - - - f9d89ef0 by David Waern at 2007-09-29T15:17:40+00:00 Take care when pp tyvars - add parens on syms - - - - - 849e2a77 by David Waern at 2007-10-01T21:56:39+00:00 Go back to using a ModuleMap instead of LookupMod - fixes a bug - - - - - 549dbac6 by David Waern at 2007-10-02T01:05:19+00:00 Improve parsing of doc options - - - - - a36021b8 by David Waern at 2007-10-02T23:05:00+00:00 FIX: double arrows in constructor contexts - - - - - d03bf347 by David Waern at 2007-10-09T16:14:05+00:00 Add a simple test suite - - - - - c252c140 by David Waern at 2007-10-17T16:02:28+00:00 Add --optghc=.. style flag passing to GHC - - - - - cce6c1b3 by David Waern at 2007-10-18T22:03:20+00:00 Add support for --read-interface again - - - - - 33d059c0 by David Waern at 2007-10-18T22:30:18+00:00 Refactoring -- get rid of Haddock.Packages - - - - - f9ed0a4c by David Waern at 2007-10-18T22:34:36+00:00 Name changes - - - - - 8a1c816f by David Waern at 2007-10-20T14:24:23+00:00 Add --ghc-version option - - - - - 4925aaa1 by David Waern at 2007-10-21T14:34:26+00:00 Add some Outputable utils - - - - - 69e7e47f by David Waern at 2007-10-21T14:35:49+00:00 FIX: Ord for OrdName was not comparing modules - - - - - 5a4ae535 by David Waern at 2007-10-21T21:18:48+00:00 Wibble - - - - - 03d48e20 by David Waern at 2007-10-24T15:52:56+00:00 Remove Main from "other modules" - - - - - c66f6d82 by David Waern at 2007-10-24T16:37:18+00:00 Make it possible to run haddock on itself - - - - - 21d156d8 by David Waern at 2007-10-25T14:02:14+00:00 Don't set boot modules as targets - - - - - f8bcf91c by David Waern at 2007-10-31T22:11:17+00:00 Add optimisation flags - - - - - 7ac758f2 by David Waern at 2007-11-04T09:48:28+00:00 Go back to loading only targets (seems to work now) - - - - - 4862aae1 by David Waern at 2007-11-05T22:24:57+00:00 Do full compilation of modules -- temporary fix for GHC API problem - - - - - 697e1517 by David Waern at 2007-11-05T22:25:50+00:00 Don't warn about not being able to link to wired/system/builtin-names - - - - - 892186da by David Waern at 2007-11-06T00:49:21+00:00 Filter out instances with TyCons that are not exported - - - - - 9548314c by David Waern at 2007-11-06T09:37:14+00:00 Wibble - - - - - 5cafd627 by David Waern at 2007-11-08T01:43:07+00:00 Filter out all non-vanilla type sigs - - - - - 04621830 by David Waern at 2007-11-08T01:45:13+00:00 Synch loading of names from .haddock files with GHC's name cache - - - - - 88d37f77 by David Waern at 2007-11-08T01:46:21+00:00 Remove commented-out code - - - - - 6409c911 by David Waern at 2007-11-08T01:56:00+00:00 Small bugfix and cleanup in getDeclFromTyCls - - - - - af59d9c2 by David Waern at 2007-11-08T02:08:44+00:00 Remove OrdName stuff - - - - - 3a615e2e by David Waern at 2007-11-08T02:13:41+00:00 Update runtests.hs following changes to haddock - - - - - 01f3314e by David Waern at 2007-11-08T02:33:01+00:00 Complain if we can't link to wired-in names - - - - - fcafb5d1 by David Waern at 2007-11-09T02:40:16+00:00 Don't exit when there are no file arguments - - - - - 194bc332 by David Waern at 2007-11-09T02:55:37+00:00 Wibble - - - - - dbe4cb55 by David Waern at 2007-11-09T02:56:14+00:00 Wibble - - - - - 82869fda by David Waern at 2007-11-10T17:01:43+00:00 Introduce InstalledInterface structure and add more stuff to the .haddock files We introduce InstalledInterface capturing the part of Interface that is stored in the interface files. We change the ppHtmlContents and ppHtmllIndex to take this structure instead of a partial Interface. We add stuff like the doc map and exported names to the .haddock file (via InstalledInterface). - - - - - d6bb57bf by David Waern at 2007-11-10T17:19:48+00:00 FIX: contents and index should include external package modules when --gen-contents/--gen-index - - - - - e8814716 by David Waern at 2007-11-11T00:29:27+00:00 Remove lDocLinkName and its use in Html backend - - - - - 6f9bd702 by David Waern at 2007-11-11T00:50:57+00:00 Do some refactoring in the html backend This also merges an old patch by Augustsson: Wed Jul 12 19:54:36 CEST 2006 lennart.augustsson at credit-suisse.com * Print type definitions like signatures if given arrows. - - - - - 09d0ce24 by Malcolm.Wallace at 2006-07-20T13:13:57+00:00 mention HsColour in the docs, next to option flags for linking to source code - - - - - 24da6c34 by Malcolm.Wallace at 2006-07-20T13:14:50+00:00 change doc references to CVS to give darcs repository location instead - - - - - 74d52cd6 by David Waern at 2007-11-11T00:55:33+00:00 Update copyright - - - - - fcaa3b4f by Duncan Coutts at 2006-09-08T13:41:00+00:00 Eliminate dep on network by doing a little cut'n'paste haddock depending on the network causes a circular dependency at least if you want to build the network lib with haddock docs. - - - - - 10cc9bda by David Waern at 2007-11-11T02:09:41+00:00 Fix conflicts - - - - - 4e3acd39 by David Waern at 2007-11-11T02:21:19+00:00 Manual merge of a patch from Duncan Coutts that removes the dependency on mtl - - - - - fa9070da by Neil Mitchell at 2006-09-29T15:52:03+00:00 Do not generate an empty table if there are no exports, this fixes a <table></table> tag being generated, which is not valid HTML 4.01 - - - - - d7431c85 by David Waern at 2007-11-11T02:28:50+00:00 Fix conflicts - - - - - f87e8f98 by Simon Marlow at 2006-10-10T11:37:16+00:00 changes for 0.8 - - - - - db929565 by Simon Marlow at 2006-10-10T12:07:12+00:00 fix the name of the source file - - - - - 8220aa4b by Simon Marlow at 2006-10-11T14:17:37+00:00 Rename haddock.js to haddock-util.js haddock.js will be run automatically by Windows when you type 'haddock' if it is found on the PATH, so rename to avoid confusion. Spotted by Adrian Hey. - - - - - 6bccdaa1 by sven.panne at 2006-10-12T15:28:23+00:00 Cabal's sdist does not generate "-src.tar.gz" files, but ".tar.gz" ones - - - - - d3f3fc19 by Simon Marlow at 2006-12-06T16:05:07+00:00 add todo item for --maintainer - - - - - 2da7e269 by Simon Marlow at 2006-12-15T15:52:00+00:00 TODO: do something better about re-exported symbols from another package - - - - - 42d85549 by David Waern at 2007-11-11T02:30:59+00:00 Fix conflicts - - - - - 5e7ef6e5 by Neil Mitchell at 2007-01-11T15:41:15+00:00 Never do spliting index files into many - - - - - f3d4aebe by Neil Mitchell at 2007-01-11T17:07:09+00:00 Add searching on the index page - - - - - bad3ab66 by Neil Mitchell at 2007-01-11T18:17:46+00:00 Delete dead code, now there is only one index page - - - - - cd09eedb by Neil Mitchell at 2007-01-11T18:21:19+00:00 Delete more stuff that is no longer required - - - - - e2806646 by David Waern at 2007-11-11T02:41:53+00:00 Fix conflicts - - - - - a872a823 by Neil Mitchell at 2007-01-11T18:51:43+00:00 Make the index be in case-insensitive alphabetic order - - - - - 8bddd9d7 by Neil Mitchell at 2007-02-06T17:49:12+00:00 Do not create empty tables for data declarations which don't have any constructors, instances or comments. Gets better HTML 4.01 compliance - - - - - 036b8120 by David Waern at 2007-11-11T02:56:58+00:00 Fix conflicts - - - - - f50c1639 by Conal Elliott at 2007-02-14T21:54:00+00:00 added substitution %{FILE///c} - - - - - 402e166a by David Waern at 2007-11-11T03:35:46+00:00 Manual merge of old patch: Sat Apr 21 04:36:43 CEST 2007 Roberto Zunino <zunrob at users.sf.net> * URL expansion for %%, %L, %{LINE} - - - - - 2f264fbd by David Waern at 2007-11-11T03:40:33+00:00 Manual merge of an old patch: Thu Apr 19 20:23:40 CEST 2007 Wolfgang Jeltsch <g9ks157k at acme.softbase.org> * bug fix When Haddock was invoked with the --ignore-all-exports flag but the ignore-exports module attribute wasn't used, hyperlinks weren't created for non-exported names. This fix might not be as clean as one would wish (since --ignore-all-exports now results in ignore_all_exports = True *and* an additional OptIgnoreExports option for every module) but at least the bug seems to be resolved now. - - - - - 7d7ae106 by sven.panne at 2007-09-02T12:18:02+00:00 Install LICENSE in the correct place - - - - - 66eaa924 by David Waern at 2007-11-11T19:02:46+00:00 Fix a bug that made haddock loop - - - - - 4ed47b58 by David Waern at 2007-11-11T19:03:09+00:00 Rename java-script file (this wasn't merge correctly) - - - - - d569534a by David Waern at 2007-11-11T19:06:44+00:00 Don't require -B <ghc-libdir> when no argument files Change readInterfaceFile to take a Maybe Session, to avoid having to pass -B <ghc-libdir> to Haddock when there're no source files to process. This is nice when computing contents/index for external packages. - - - - - 373368bc by Neil Mitchell at 2007-01-11T18:22:44+00:00 Change from tabs to spaces in the ppHtmlIndex function - - - - - 6b063a77 by Neil Mitchell at 2007-01-12T12:17:46+00:00 Rewrite much of the index searching code, previously was too slow to execute on the base library with IE, the new version guarantees less than O(log n) operations be performed, where n is the number in the list (before was always O(n)) - - - - - bfad00b7 by David Waern at 2007-11-11T23:33:53+00:00 Fix conflicts - - - - - cd2dcc09 by Neil Mitchell at 2007-01-12T12:25:01+00:00 Make the max number of results 75 instead of 50, to allow map searching in the base library to work - - - - - 3ae74764 by Neil Mitchell at 2007-01-12T12:58:17+00:00 Make the search box in a form so that enter does the default search - - - - - 142103e5 by David Waern at 2007-11-12T00:03:18+00:00 Merge patch from the old branch: Fri Aug 31 13:21:45 CEST 2007 Duncan Coutts <duncan at haskell.org> * Add category: Development to .cabal file Otherwise it appears on the hackage website in the "Unclassified" category. - - - - - 22ec2ddb by David Waern at 2007-11-25T01:55:29+00:00 A a list of small improvements to the TODO file - - - - - eb0129f4 by Wolfgang Jeltsch at 2007-12-03T23:47:55+00:00 addition of type equality support (at least for HTML generation) - - - - - 816a7e22 by David Waern at 2007-12-08T15:46:26+00:00 Handle class operators correctly when rendering predicates - - - - - 68baaad2 by David Waern at 2007-12-08T16:15:54+00:00 Code layout changes - - - - - 09b77fb4 by David Waern at 2007-12-08T16:16:03+00:00 Handle infix operators correctly in the Type -> HsType translation - - - - - 31c36da2 by David Waern at 2007-12-08T16:24:27+00:00 Add ppLParendTypes/ppLParendType - - - - - b17cc818 by David Waern at 2007-12-08T16:26:12+00:00 Use ppParendType when printing types args in predicates - - - - - ffd1f2cf by David Waern at 2007-12-08T16:45:06+00:00 Fix rendering of instance heads to handle infix operators This is also a refactoring to share this code for rendering predicates. - - - - - ff886d45 by David Waern at 2007-12-08T17:27:46+00:00 Fix rendering of class operators - - - - - e2fcbb9e by David Waern at 2007-12-08T17:59:28+00:00 Fix a bug (use ppTyName instead of ppName to print names in type apps) - - - - - 79a1056e by David Waern at 2007-12-08T21:25:18+00:00 Update tests - - - - - 867741ac by David Waern at 2007-12-08T21:25:49+00:00 Give a diff on test failure - - - - - 7e5eb274 by David Waern at 2008-01-05T14:33:45+00:00 Add DrIFT commands - - - - - 3656454d by David Waern at 2008-01-05T20:26:00+00:00 Add "cabal-version: >= 1.2" to the cabal file - - - - - 77974efc by Simon Marlow at 2007-12-20T09:52:44+00:00 add an item - - - - - f6ac1708 by Simon Marlow at 2007-12-06T14:00:10+00:00 Source links must point to the original module, not the referring module - - - - - eda1d5c9 by David Waern at 2008-01-06T14:40:52+00:00 Manual merge of a patch to the 0.8 branch Thu Dec 6 15:00:10 CET 2007 Simon Marlow <simonmar at microsoft.com> * Source links must point to the original module, not the referring module - - - - - 378f4085 by David Waern at 2008-01-06T16:03:45+00:00 Change stability from stable to experimental - - - - - 8bdafe44 by David Waern at 2008-01-06T16:14:22+00:00 Add haskell.vim (it had been removed somehow) - - - - - ea34d02e by David Waern at 2008-01-06T16:36:57+00:00 Change version to 2.0.0.0 - - - - - 34631ac0 by David Waern at 2008-01-06T16:44:57+00:00 Add missing modules to the cabal file - - - - - 9e142935 by David Waern at 2008-01-06T17:25:42+00:00 Depend on ghc >= 6.8.2 && < 6.9 - - - - - 59f9eeaa by Simon Marlow at 2007-12-20T10:43:04+00:00 add build scripts - - - - - 1c29ae30 by Simon Marlow at 2007-12-20T10:47:07+00:00 update version number - - - - - fe16a3e4 by Simon Marlow at 2007-12-20T10:48:03+00:00 update version - - - - - f688530f by Simon Marlow at 2007-12-20T10:48:29+00:00 doc updates - - - - - ce71b611 by David Waern at 2008-01-07T13:46:32+00:00 Change version in docs and spec - - - - - 03ab8d6f by David Waern at 2008-01-07T13:47:38+00:00 Manually merge over changes to CHANGES for 0.9 - - - - - 39f1b042 by David Waern at 2008-01-07T15:17:41+00:00 Remove the -use-package flag, we don't support it anyway - - - - - 7274a544 by David Waern at 2008-01-07T15:33:05+00:00 Update CHANGES for 2.0.0.0 - - - - - 96594f5d by David Waern at 2008-01-07T15:46:49+00:00 Wibble - - - - - f4c5a4c4 by David Waern at 2008-01-07T15:55:36+00:00 Change url to repo in documentation - - - - - 8a4c77f0 by David Waern at 2008-01-07T16:00:54+00:00 Update CHANGES - - - - - cb3a9288 by David Waern at 2008-01-07T16:02:55+00:00 Documentation fix - - - - - d8e45539 by David Waern at 2008-01-07T16:12:00+00:00 Update docs to say that Haddock accets .lhs files and module names - - - - - 4b5ce824 by David Waern at 2008-01-07T16:12:25+00:00 Document -B option - - - - - 47274262 by David Waern at 2008-01-07T16:23:07+00:00 Update CHANGES - - - - - 7ff314a9 by David Waern at 2008-01-07T16:23:20+00:00 Remove --use-package, --package & --no-implicit.. flags from docs - - - - - 6c3819c0 by David Waern at 2008-01-07T16:23:52+00:00 Remove --no-implicit-prelide flag - - - - - 1b14ae40 by David Waern at 2008-01-07T16:32:26+00:00 Update the "Using literate or pre-processed source" section - - - - - 0117f620 by David Waern at 2008-01-07T16:41:55+00:00 Document the --optghc flag - - - - - 087ab1cf by David Waern at 2008-01-07T16:42:10+00:00 Remove the documenation section on derived instances The problem mentioned there doesn't exist in Haddock 2.0.0.0 - - - - - 7253951e by David Waern at 2008-01-07T16:48:40+00:00 Document OPTIONS_HADDOCK - - - - - 3b6bdcf6 by David Waern at 2008-01-07T16:56:54+00:00 Wibble - - - - - 3025adf9 by David Waern at 2008-01-07T17:08:14+00:00 Wibble - - - - - 5f30f1a0 by David Waern at 2008-01-07T17:15:44+00:00 Change synopsis field to description - - - - - 1673f54b by David Waern at 2008-01-07T17:18:21+00:00 Change my email address in the cabal file - - - - - 55aa9808 by David Waern at 2008-01-07T18:18:02+00:00 Add documentation for readInterfaceFile - - - - - eaea417f by David Waern at 2008-01-07T18:21:30+00:00 Export necessary stuff from Distribution.Haddock - - - - - 7ea18759 by David Waern at 2008-01-07T18:31:49+00:00 Remove dep on Cabal - - - - - 7b79c74e by David Waern at 2008-01-07T18:33:49+00:00 Remove dep on process - - - - - ce3054e6 by David Waern at 2008-01-16T23:01:21+00:00 Add feature-requsts from Henning Thielemann to TODO - - - - - 0c08f1ec by David Waern at 2008-01-16T23:03:02+00:00 Record a bug in TODO - - - - - b04605f3 by David Waern at 2008-01-23T16:59:06+00:00 Add a bug reported by Ross to TODO - - - - - 5b17c030 by David Waern at 2008-01-23T18:05:53+00:00 A a bug report to TODO - - - - - 1c993b0d by David Waern at 2008-01-25T16:30:25+00:00 Accept test output - - - - - c22fc0d0 by David Waern at 2008-01-25T16:34:49+00:00 Accept test output - - - - - 4b795811 by David Waern at 2008-01-25T16:38:37+00:00 Change Hidden.hs (test) to use OPTIONS_HADDOCK - - - - - c124dbd9 by David Waern at 2008-01-25T16:39:23+00:00 Accept test output - - - - - ec6f6eea by David Waern at 2008-01-25T16:42:08+00:00 Add Hidden.html.ref to tests - - - - - 1dc9610c by David Waern at 2008-02-02T20:50:51+00:00 Add a comment about UNPACK bug in TODO - - - - - 2d3f7081 by David Waern at 2008-02-09T22:33:24+00:00 Change the representation of DocNames Ross Paterson reported a bug where links would point to the defining module instead of the "best" module for an identifier (e.g Int pointing to GHC.Base instead of Data.Int). This patch fixes this problem by refactoring the way renamed names are represented. Instead of representing them by: > data DocName = Link Name | NoLink Name they are now represented as such: > data DocName = Documented Name Module | Undocumented Name and the the link-env looks like this: > type LinkEnv = Map Name Module There are several reasons for this. First of all, the bug was caused by changing the module part of Names during the renaming process, without changing the Unique field. This caused names to be overwritten during the loading of .haddock files (which caches names using the NameCache of the GHC session). So we might create new Uniques during renaming to fix this (but I'm not sure that would be problem-free). Instead, we just keep the Name and add the Module where the name is best documented, since it can be useful to keep the original Name around (for e.g. source-code location info and for users of the Haddock API). Also, the names Link/NoLink don't really make sense, since wether to use links or not is entirely up to the users of DocName. In the process of following this change into H.Backends.Html I removed the assumption that binder names are Undocumented (which was just an unnecessary assumption, the OccName is the only thing needed to render these). This will probably make it possible to get rid of the renamer and replace it with a traversal from SYB or Uniplate. Since DocName has changed, InterfaceFile has changed so this patch also increments the file-format version. No backwards-compatibility is implemented. - - - - - 0f28c921 by David Waern at 2008-02-09T23:00:36+00:00 H.GHC.Utils: remove unused imports/exports - - - - - 0c44cad5 by David Waern at 2008-02-10T00:28:13+00:00 H.GHC.Utils: add some functions that were removed by mistake - - - - - e3452f49 by David Waern at 2008-02-10T00:28:48+00:00 Fix some trivial warnings in H.InterfaceFile - - - - - a6d74644 by David Waern at 2008-02-10T00:48:06+00:00 Update the version message to fit in small terminals - - - - - 76c9cd3e by David Waern at 2008-02-10T14:47:39+00:00 Remove bugs from TODO that don't apply anymore since the port - - - - - 5e10e090 by David Waern at 2008-02-10T15:22:47+00:00 Remove bugs from TODO that weren't actual bugs - - - - - fef70878 by David Waern at 2008-02-10T15:23:44+00:00 Remove yet another item from TODO that was not an actual bug - - - - - e1af47b8 by David Waern at 2008-02-11T10:25:57+00:00 Bump the version number to 2.1.0 Since the exported datatype DocName has changed, we need to bump the major version number. Let's also drop the fourth version component, it's not that useful. - - - - - e3be7825 by David Waern at 2008-04-11T14:29:04+00:00 Add a bug to TODO - - - - - cb6574be by David Waern at 2008-04-11T16:00:45+00:00 Use the in-place haddock when running tests - - - - - c6d7af0d by David Waern at 2008-04-11T16:09:16+00:00 Turn off GHC warnings when running tests - - - - - 7f61b546 by David Waern at 2008-04-11T17:24:00+00:00 Add a flag for turning off all warnings - - - - - 883b8422 by David Waern at 2008-04-12T14:02:18+00:00 Fix printing of data binders - - - - - 2a0db8fc by David Waern at 2008-04-12T18:52:46+00:00 Fix missing parenthesis in constructor args bug - - - - - 1b3ac3f9 by David Waern at 2008-04-12T18:57:23+00:00 Simplify test suite and add tests I move all tests into one single directory to simplify things, and add a test for the last bug that was fixed. - - - - - 8f178376 by David Waern at 2008-04-12T19:00:15+00:00 Add a script for copying test output to "expected" output - - - - - 193e3a03 by David Waern at 2008-04-12T19:16:37+00:00 Remove two fixed bugs from TODO - - - - - ddc9130c by David Waern at 2008-04-12T19:37:06+00:00 Update test README - - - - - 956069c0 by David Waern at 2008-05-01T12:16:14+00:00 Update version number in spec and docs - - - - - 5478621c by David Waern at 2008-05-01T12:28:12+00:00 Remove claim of backwards compatibility from docs for readInterfaceFile - - - - - 4a16dea9 by David Waern at 2008-05-01T12:33:04+00:00 Update CHANGES - - - - - 804216fb by David Waern at 2008-05-01T12:43:16+00:00 Add a synopsis - - - - - fd0c84d5 by David Waern at 2008-05-01T12:44:44+00:00 Add Haddock.DocName to the cabal file - - - - - 9f4a7439 by David Waern at 2008-05-01T12:45:53+00:00 Remove -fglasgow-exts and -fasm - - - - - aee7c145 by David Waern at 2008-05-01T12:54:01+00:00 Add LANGUAGE pragmas to source files - - - - - 9a58428b by David Waern at 2008-05-01T12:54:19+00:00 Add extensions to cabal file - - - - - 494f1bee by David Waern at 2008-05-01T13:12:09+00:00 Export DocName in the API - - - - - c938196b by David Waern at 2008-05-01T13:12:19+00:00 Add hide options to some source files - - - - - 236e86af by Neil Mitchell at 2008-06-07T20:45:10+00:00 Rewrite the --hoogle flag support - - - - - 6d910950 by Neil Mitchell at 2008-06-14T10:56:50+00:00 Simplify the newtype/data outputting in Hoogle, as haddock does it automatically - - - - - f87a95a8 by Neil Mitchell at 2008-06-14T12:10:18+00:00 Add initial structure for outputting documentation as well, but does not yet output anything - - - - - 7c3bce54 by Neil Mitchell at 2008-06-14T12:27:07+00:00 Remove <document comment> from the Hoogle output - - - - - 9504a325 by Neil Mitchell at 2008-06-16T06:33:21+00:00 Default to "main" if there is no package, otherwise will clobber hoogle's hoogle info - - - - - 4a794a79 by Neil Mitchell at 2008-06-16T06:53:29+00:00 Change packageName to packageStr, as it better reflects the information stored in it - - - - - 7abc9baf by Neil Mitchell at 2008-06-16T07:09:49+00:00 Add modulePkgInfo to Haddock.GHC.Utils, which gives back package name and version info - - - - - 8ca11514 by Neil Mitchell at 2008-06-16T07:13:48+00:00 Change Hoogle to take the package name and package version separately - - - - - a6da452d by Neil Mitchell at 2008-06-18T11:29:46+00:00 In Hoogle do not list things that are not local to this module - - - - - 974b76b7 by David Waern at 2008-06-19T18:40:13+00:00 Be more consistent with GHC API naming in H.GHC.Utils - - - - - 2facb4eb by David Waern at 2008-06-19T19:03:03+00:00 Update test output - - - - - c501de72 by David Waern at 2008-06-26T20:26:49+00:00 Use ghc-paths to get the lib dir The path can still be overridden using the -B flag. It's not longer required to pass the lib dir to the program that runs the test suite. - - - - - ac4c6836 by David Waern at 2008-06-26T20:33:08+00:00 Update CHANGES - - - - - 9d21c60a by David Waern at 2008-06-26T20:34:53+00:00 Update README - - - - - 741448f0 by David Waern at 2008-06-26T21:12:57+00:00 Improve wording in the help message - - - - - b1b42b11 by David Waern at 2008-06-30T10:16:17+00:00 Rename ForeignType - - - - - 6d6c2b34 by David Waern at 2008-06-30T10:25:09+00:00 Rename TyFamily - - - - - 8d1125ed by David Waern at 2008-06-30T10:37:21+00:00 Rename type patterns - - - - - 7610a4cb by David Waern at 2008-06-30T10:45:07+00:00 Rename associated types - - - - - 8eeba14c by David Waern at 2008-06-30T10:47:41+00:00 Remove the TODO file now that we have a trac - - - - - 1af5b25b by David Waern at 2008-07-02T18:19:28+00:00 Render type family declarations (untested) - - - - - ceb99797 by David Waern at 2008-07-02T18:24:06+00:00 Remove redundant check for summary when rendering data types - - - - - b36a58e0 by David Waern at 2008-07-02T22:01:38+00:00 More support for type families and associated types Now we just need to render the instances - - - - - 78784879 by David Waern at 2008-07-07T22:13:58+00:00 Remove filtering of instances We were filtering out all instances for types with unknown names. This was probably an attempt to filter out instances for internal types. I am removing the filtering for the moment, and will try to fix this properly later. - - - - - 3e758dad by David Waern at 2008-06-30T18:50:30+00:00 Run haddock in-place during testing - - - - - d9dab0ce by David Waern at 2008-07-08T21:04:32+00:00 Remove index.html and doc-index.html from output, they should not be versioned - - - - - 3e6c4681 by David Waern at 2008-07-08T21:06:42+00:00 Update test output following change to instance filtering - - - - - e34a3f14 by David Waern at 2008-07-12T16:48:28+00:00 Stop using the map from exported names to declarations During creation of the interface, we were using two maps: one from exported names to declarations, and one from all defined names in the module to declarations. The first contained subordinate names while the second one didn't. The first map was never used to look up names not defined in the associated module, so if we add subordinate names to the second map, we could use it everywhere. That's that this patch does. This simplifies code because we don't have to pass around two maps everywhere. We now store the map from locally defined things in the interface structure instead of the one from exported names. - - - - - 2e1d2766 by David Waern at 2008-07-12T16:55:21+00:00 Get the all locally defined names from GHC API We previously had some code to compute all locally defined names in a module including subordinate names. We don't need it since we can get the names from modInfoTyThings in the GHC API. - - - - - bf637994 by David Waern at 2008-07-13T13:09:16+00:00 Refactoring in H.Interface.Create We were creating a doc map, a declaration map and a list of entities separately by going through the HsGroup. These structures were all used to build the interface of a module. Instead of doing this, we can start by creating a list of declarations from the HsGroup, then collect the docs directly from this list (instead of using the list of entities), creating a documentation map. We no longer need the Entity data type, and we can store a single map from names to declarations and docs in the interface, instead of the declaration map and the doc map. This way, there is only one place where we filter out the declarations that we don't want, and we can remove a lot of code. Another advantage of this is that we can create the exports directly out of the list of declarations when we export the full module contents. (Previously we did a look up for each name to find the declarations). This is faster and removes another point where we depend on names to identify exported declarations, which is good because it eliminates problems with instances (which don't have names). - - - - - 547e410e by David Waern at 2008-07-13T13:34:51+00:00 Remove FastString import and FSLIT macro in H.I.Create -- they were unused - - - - - 693759d1 by David Waern at 2008-07-13T13:36:23+00:00 Remove unused import from H.I.Create - - - - - cde6e7fb by David Waern at 2008-07-13T13:51:54+00:00 Small touches - - - - - 96de8f1d by David Waern at 2008-07-20T11:21:46+00:00 Preparation for rendering instances as separate declarations We want to be able to render instances as separate declarations. So we remove the Name argument of ExportDecl, since instances are nameless. This patch also contains the first steps needed to gather type family instances and display them in the backend, but the implementation is far from complete. Because of this, we don't actually show the instances yet. - - - - - b0f824fb by David Waern at 2008-07-20T15:53:08+00:00 Follow changes to ExportDecl in Hoogle - - - - - 1192eff3 by Neil Mitchell at 2008-06-26T00:28:10+00:00 Change how the Hoogle backend outputs classes, adding the context in - - - - - 7a0d1464 by Neil Mitchell at 2008-06-26T00:28:46+00:00 Remove the indent utility function from Hoogle backend - - - - - 3361241b by Neil Mitchell at 2008-06-26T09:45:09+00:00 Add support for Hoogle writing ForeignImport/ForeignExport properly - - - - - 795ad3bf by Neil Mitchell at 2008-06-26T12:15:25+00:00 Flesh out the Hoogle code to render documentation - - - - - 23277995 by Neil Mitchell at 2008-06-26T14:56:41+00:00 Fix a bug in the Hoogle backend, unordered lists were being written out <ul>...</u> - - - - - db739b27 by Neil Mitchell at 2008-06-26T15:09:54+00:00 Remove any white space around a <li> element - - - - - f2e6bb8c by Neil Mitchell at 2008-07-10T15:30:47+00:00 Remove the TODO in the Hoogle HTML generation, was already done - - - - - 693ec9a3 by Neil Mitchell at 2008-07-10T15:53:00+00:00 Put brackets round operators in more places in the Hoogle output - - - - - 842313aa by Neil Mitchell at 2008-07-10T16:01:25+00:00 Print type signatures with brackets around the name - - - - - cf93deb0 by David Waern at 2008-07-20T17:04:22+00:00 Bump version number to 2.2.0 - - - - - 30e6a8d1 by David Waern at 2008-07-20T17:04:41+00:00 Resolve conflicts in H.B.Hoogle - - - - - 1f0071c9 by David Waern at 2008-07-23T23:05:01+00:00 Add "all" command to runtests.hs that runs all tests despite failures - - - - - f2723023 by David Waern at 2008-07-23T23:08:39+00:00 Update tests/README - - - - - c0304a11 by David Waern at 2008-07-23T23:21:15+00:00 Be compatible with GHC 6.8.3 The cabal file is converted to use the "new" syntax with explicit Library and Executable sections. We define the __GHC_PATCHLEVEL__ symbol using a conditinal cpp-options field in the cabal file. (Ideally, Cabal would define the symbol for us, like it does for __GLASGOW_HASKELL__). We use these symbols to #ifdef around a small difference between 6.8.2 and 6.8.3. Previously, we only supported GHC 6.8.2 officially but the dependencies field said "ghc <= 6.9". This was just for convenience when testing against the (then compatible) HEAD version of GHC, and was left in the release by mistake. Now, we support both GHC 6.8.2 and 6.8.3 and the dependencies field correctly reflects this. - - - - - 88a5fe71 by David Waern at 2008-07-23T23:54:16+00:00 Depend on the currently available ghc-paths versions only - - - - - 8738d97b by David Waern at 2008-07-24T10:50:44+00:00 FIX haskell/haddock#44: Propagate parenthesis level when printing documented types - - - - - 05339119 by David Waern at 2008-07-24T16:06:18+00:00 Drop unnecessary parenthesis in types, put in by the user We were putting in parenthesis were the user did. Let's remove this since it just clutters up the types. The types are readable anyway since we print parens around infix operators and do not rely on fixity levels. When doing this I discovered that we were relying on user parenthesis when printin types like (a `O` b) c. This patchs fixes this problem so that parenthesis are always inserted around an infix op application in case it is applied to further arguments, or if it's an arguments to a type constructor. Tests are updated. - - - - - b3a99828 by David Waern at 2008-07-24T10:19:43+00:00 Print parenthesis around non-atomic banged types Fixes half of haskell/haddock#44 - - - - - ab5238e0 by David Waern at 2008-07-24T22:07:49+00:00 Add a reference file for the TypeFamilies test - - - - - 1941cc11 by David Waern at 2008-07-25T17:15:53+00:00 Simplify definition of pretty and trace_ppr - - - - - e3bfa33c by David Waern at 2008-07-25T17:18:27+00:00 Warning messages Output a warning when filtering out data/type instances and associated types in instances. We don't show these in the documentation yet, and we need to let the user know. - - - - - 9b85fc89 by David Waern at 2008-07-25T17:45:40+00:00 Doc: Mention Hoogle in the Introduction - - - - - afb2dd60 by David Waern at 2008-07-25T17:49:00+00:00 Doc: update -B description - - - - - 584c0c91 by David Waern at 2008-07-25T18:11:38+00:00 Doc: describe -w flag - - - - - 77619c24 by David Waern at 2008-07-28T12:29:07+00:00 Remove TODO from cabal file - - - - - 96717d5f by David Waern at 2008-07-28T12:29:27+00:00 Support type equality predicates - - - - - c2fd2330 by David Waern at 2008-07-29T19:45:14+00:00 Move unL from H.B.Hoogle to H.GHC.Utils I like Neil's shorter unL better than unLoc from the GHC API. - - - - - c4c3bf6a by David Waern at 2008-07-29T19:47:36+00:00 Do not export ATs when not in list of subitems - - - - - bf9a7b85 by David Waern at 2008-08-03T11:42:59+00:00 Filter out ForeignExports - - - - - df59fcb0 by David Waern at 2008-08-03T14:02:51+00:00 Filter out more declarations The previous refactorings in H.I.Create introduced a few bugs. Filtering of some types of declarations that we don't handle was removed. This patch fixes this. - - - - - 2f8a958b by David Waern at 2008-08-03T15:24:07+00:00 Move reL to H.GHC.Utils so we can use it everywhere - - - - - 8ec15efd by David Waern at 2008-08-03T15:25:00+00:00 Use isVanillaLSig from GHC API instead of home brewn function - - - - - 300f93a2 by David Waern at 2008-08-03T15:25:27+00:00 Filter out separately exported ATs This is a quick and dirty hack to get rid of separately exported ATs. We haven't decided how to handle them yet. No warning message is given. - - - - - 8776d1ec by David Waern at 2008-08-03T16:21:21+00:00 Filter out more declarations and keep only vanilla type sigs in classes - - - - - ea07eada by David Waern at 2008-08-03T16:48:00+00:00 Fix layout - - - - - dd5e8199 by David Waern at 2008-08-03T16:50:52+00:00 Move some utility functions from H.I.Create to H.GHC.Utils - - - - - 4a1dbd72 by David Waern at 2008-08-03T17:39:55+00:00 Do not filter out doc declarations - - - - - 0bc8dca4 by David Waern at 2008-08-03T17:47:26+00:00 Filter out separately exported ATs (take two) - - - - - af970fe8 by David Waern at 2008-08-03T22:39:17+00:00 Update CHANGES - - - - - 5436ad24 by David Waern at 2008-08-03T22:40:20+00:00 Bump version number to 2.2.1 - - - - - d66de448 by David Waern at 2008-08-05T19:00:32+00:00 Remove version restriction on ghc-paths - - - - - 534b1364 by David Waern at 2008-08-05T19:04:35+00:00 Bump version to 2.2.2 and update CHANGES - - - - - 549188ff by David Waern at 2008-08-05T19:16:49+00:00 Fix CHANGES - - - - - 0d156bb4 by Luke Plant at 2008-08-11T15:20:59+00:00 invoking haddock clarification and help - - - - - 748295cc by David Waern at 2008-08-11T18:56:37+00:00 Doc: say that the --hoogle option is functional - - - - - 43301db4 by David Waern at 2008-08-05T19:26:08+00:00 Change ghc version dependency to >= 6.8.2 - - - - - 3e5a53b6 by David Waern at 2008-08-10T22:42:05+00:00 Make H.GHC.Utils build with GHC HEAD - - - - - 7568ace0 by David Waern at 2008-08-11T19:41:54+00:00 Import Control.OldException instead of C.Exception when using ghc >= 6.9 We should really test for base version instead, but I don't currently know which version to test for. - - - - - b71ae991 by David Waern at 2008-08-12T22:40:39+00:00 Make our .haddock file version number depend on the GHC version We need to do this, since our .haddock format can potentially change whenever GHC's version changes (even when only the patchlevel changes). - - - - - 6307ce3f by David Waern at 2008-08-12T22:49:57+00:00 Remove matching on NoteTy in AttachInstances, it has been removed - - - - - 2dbcfd5f by David Waern at 2008-08-12T23:02:02+00:00 Comment out H.GHC.loadPackages - it is unused and doesn't build with ghc >= 6.9 - - - - - c74db5c2 by David Waern at 2008-08-12T23:03:58+00:00 Hide <.> from GHC import in Hoogle only for ghc <= 6.8.3 - - - - - 69a44ebb by David Waern at 2008-08-12T23:11:12+00:00 Follow changes to parseDynamic/StaticFlags - - - - - 5881f3f0 by David Waern at 2008-08-13T21:43:58+00:00 Add __GHC_PATCHLEVEL__ symbol also when building the library - - - - - 8574dc11 by David Waern at 2008-08-13T21:44:17+00:00 Follow move of package string functions from PackageConfig to Module - - - - - c9baa77f by David Waern at 2008-08-13T21:45:29+00:00 Follow extensible exceptions changes - - - - - 9092de15 by David Waern at 2008-08-13T21:46:20+00:00 Update test following Haddock version change - - - - - ebe569a4 by David Waern at 2008-08-13T21:46:54+00:00 Follow changes to parseDynamic- parseStaticFlags in GHC - - - - - b8a5ffd3 by David Waern at 2008-08-13T21:47:36+00:00 Follow changes to Binary in GHC 6.9 - - - - - edfda1cc by David Waern at 2008-08-13T21:50:17+00:00 Change ghc version dependency to >= 6.8.2 && <= 6.9 - - - - - d59be1cf by Neil Mitchell at 2008-08-12T16:02:53+00:00 Output all items, even if they are not defined in this module - ensures map comes from Prelude, not just GHC.Base - - - - - dda93b9f by Neil Mitchell at 2008-08-12T21:37:32+00:00 Add support for type synonyms to Hoogle, was accidentally missing before (woops!) - - - - - b6ee795c by Neil Mitchell at 2008-08-13T14:03:24+00:00 Generalise Hoogle.doc and add a docWith - - - - - 415e1bb2 by Neil Mitchell at 2008-08-13T14:03:46+00:00 Make Hoogle add documentation to a package - - - - - 790a1202 by Neil Mitchell at 2008-08-18T12:52:43+00:00 Use the same method to put out signatures as class methods in the Hoogle backend - - - - - ded37eba by Neil Mitchell at 2008-08-18T12:53:04+00:00 Remove Explicit top-level forall's when pretty-printing signatures - - - - - 6468c722 by Neil Mitchell at 2008-08-20T07:59:13+00:00 Simplify the code by removing not-to-important use of <.> in the Hoogle back end - - - - - 788c3a8b by Neil Mitchell at 2008-08-21T18:20:24+00:00 In the hoogle back end, markup definition lists using <i>, not <b> - - - - - 77d4b000 by Ian Lynagh at 2008-08-14T10:49:14+00:00 Add a Makefile for GHC's build system. Still won't work yet, but we're closer - - - - - 920440d7 by Ian Lynagh at 2008-08-27T18:06:46+00:00 Add haddock.wrapper - - - - - bcda925f by Ian Lynagh at 2008-08-27T18:07:02+00:00 Add a manual Cabal flag to control the ghc-paths dependency - - - - - 04d194e2 by Ian Lynagh at 2008-08-27T20:41:27+00:00 Update extensions in Cabal file Use ScopedTypeVariables instead of PatternSignatures - - - - - 12480043 by Ian Lynagh at 2008-08-27T20:41:55+00:00 Increase the upper bound on the GHC version number - - - - - b1f809a5 by Ian Lynagh at 2008-08-27T21:32:22+00:00 Fix some warnings - - - - - aea0453d by Ian Lynagh at 2008-08-28T14:22:29+00:00 Fixes for using haddock in a GHC build tree - - - - - ad23bf86 by Ian Lynagh at 2008-08-28T21:14:27+00:00 Don't use Cabal wrappers on Windows - - - - - 35858e4c by Ian Lynagh at 2008-08-29T00:07:42+00:00 Fix in-tree haddock on Windows - - - - - c2642066 by Ian Lynagh at 2008-09-03T22:35:53+00:00 follow library changes - - - - - 2eb55d50 by Ian Lynagh at 2008-09-07T18:52:51+00:00 bindist fixes - - - - - 3daa5b59 by Ian Lynagh at 2008-09-10T16:58:18+00:00 We need to tell haddock that its datasubdir is . or it can't find package.conf - - - - - 388fd8c2 by Ian Lynagh at 2008-09-10T19:47:44+00:00 Fix haddock inplace on Windows - - - - - 70a641c1 by Ian Lynagh at 2008-09-10T22:15:44+00:00 Fix installed haddock on Windows - - - - - 83c1e997 by Neil Mitchell at 2008-09-11T10:48:55+00:00 Import GHC.Paths if not IN_GHC_TREE, seems to match the use of GHC.Paths functions much better - - - - - b452519b by Ian Lynagh at 2008-09-12T12:58:24+00:00 Add a LANGUAGE ForeignFunctionInterface pragma - - - - - afbd592c by Ian Lynagh at 2008-09-12T12:59:13+00:00 Wibble imports - - - - - 547ac4ad by Ian Lynagh at 2008-09-14T15:34:22+00:00 Add a "#!/bin/sh" to haddock.wrapper - - - - - f207a807 by Ian Lynagh at 2008-09-15T10:02:32+00:00 Use "exec" when calling haddock in the wrapper - - - - - 2ee68509 by Thomas Schilling at 2008-09-15T09:09:16+00:00 Port Haddock.Interface to new GHC API. This required one bigger change: 'readInterfaceFile' used to take an optional 'Session' argument. This was used to optionally update the name cache of an existing GHC session. This does not work with the new GHC API, because an active session requires the function to return a 'GhcMonad' action, but this is not possible if no session is provided. The solution is to use an argument of functions for reading and updating the name cache and to make the function work for any monad that embeds IO, so it's result type can adapt to the calling context. While refactoring, I tried to make the code a little more self-documenting, mostly turning comments into function names. - - - - - 3bb96431 by Thomas Schilling at 2008-09-15T09:09:37+00:00 Reflect GHC API changes. - - - - - 2e60f714 by Thomas Schilling at 2008-09-15T09:10:37+00:00 Port Haddock.GHC.Typecheck to new GHC API. - - - - - 9cfd4cff by Thomas Schilling at 2008-09-15T09:11:00+00:00 Port Haddock.GHC to new GHC API. - - - - - caffa003 by Thomas Schilling at 2008-09-15T09:11:25+00:00 Port Main to new GHC API. - - - - - 069a4608 by Ian Lynagh at 2008-09-21T11:19:00+00:00 Fix paths used on Windows frmo a GHC tree: There is no whare directory - - - - - 7ceee1f7 by Ian Lynagh at 2008-09-21T12:20:16+00:00 Fix the in-tree haddock on Windows - - - - - 0d486514 by Ian Lynagh at 2008-09-23T18:06:58+00:00 Increase the GHC upper bound from 6.11 to 6.13 - - - - - f092c414 by Neil Mitchell at 2008-09-11T14:56:07+00:00 Do not wrap __ in brackets - - - - - 036bdd13 by Ian Lynagh at 2008-09-28T01:42:35+00:00 Fix building haddock when GhcProfiled=YES - - - - - 01434a89 by David Waern at 2008-09-24T20:24:21+00:00 Add PatternSignatures LANGUAGE pragma to Main and Utils - - - - - 1671a750 by David Waern at 2008-10-02T22:57:25+00:00 For source links, get original module from declaration name instead of environment. Getting it from the environment must have been a remnant from the times when we were using unqualified names (versions 0.x). - - - - - a25dde99 by David Waern at 2008-10-02T22:59:57+00:00 Remove ifaceEnv from Interface - it's no longer used - - - - - 610993da by David Waern at 2008-10-02T23:04:58+00:00 Write a comment about source links for type instance declarations - - - - - 5a96b5d5 by Thomas Schilling at 2008-10-03T10:45:08+00:00 Follow GHC API change of parseModule. - - - - - 5a943ae5 by Ian Lynagh at 2008-10-03T15:56:58+00:00 TAG 2008-10-03 - - - - - 76cdd6ae by Thomas Schilling at 2008-10-08T12:29:50+00:00 Only load modules once when typechecking with GHC. This still doesn't fix the memory leak since the typechecked source is retained and then processed separately. To fix the leak, modules must be processed directly after typechecking. - - - - - 7074d251 by David Waern at 2008-10-09T23:53:54+00:00 Interleave typechecking with interface creation At the same time, we fix a bug where the list of interfaces were processed in the wrong order, when building the links and renaming the interfaces. - - - - - 4b9b2b2d by David Waern at 2008-10-09T23:54:49+00:00 Add some strictness annotations in Interface We add some strictness annotations to the fields of Interface, so that less GHC data is hold on to during processing. - - - - - 22035628 by David Waern at 2008-10-10T20:02:31+00:00 Remove typecheckFiles and MonadUtils import from H.GHC.Typeccheck - - - - - be637ad3 by David Waern at 2008-10-10T20:33:38+00:00 Make Haddock build with GHC 6.8.2 - - - - - 523b3404 by David Waern at 2008-10-10T21:08:09+00:00 Fix documentation for createInterfaces - - - - - e1556702 by David Waern at 2008-10-10T21:26:19+00:00 Hide H.Utils in library - - - - - a8e751c3 by David Waern at 2008-10-10T21:34:59+00:00 Add back .haddock file versioning based on GHC version It was accidentally removed in the patch for GHC 6.8.2 compatibility - - - - - 06fb3c01 by David Waern at 2008-10-10T21:47:15+00:00 Bump version number to 2.3.0 - - - - - ff087fce by David Waern at 2008-10-10T22:35:49+00:00 Add support for DocPic The support for DocPic was merged into the GHC source long ago, but the support in Haddock was forgotten. Thanks Peter Gavin for submitting this fix! - - - - - 3af85bf6 by David Waern at 2008-10-10T23:34:05+00:00 Update tests - - - - - 0966873c by Simon Marlow at 2008-10-10T14:43:04+00:00 no need for handleErrMsg now, we don't throw any ErrMsgs - - - - - f1870de3 by Clemens Fruhwirth at 2008-10-10T13:29:36+00:00 Compile with wrapper but remove it for dist-install - - - - - 7b440dc2 by David Waern at 2008-10-11T14:02:25+00:00 Remove interface from LinksInfo It was there to know the documentation home module when creating a wiki link, but we already know this since we have the DocName. - - - - - e5729e6a by David Waern at 2008-10-15T20:49:18+00:00 Wibble - - - - - b2a8e01a by David Waern at 2008-10-15T21:03:36+00:00 Use type synonyms for declarations and docs in H.I.Create - - - - - be71a15b by David Waern at 2008-10-15T21:12:17+00:00 Comment out unused type family stuff completely - - - - - 91aaf075 by David Waern at 2008-10-15T21:49:04+00:00 Wibble - - - - - 42ba4eb4 by David Waern at 2008-10-15T21:53:53+00:00 Move convenient type synonym to H.Types - - - - - db11b723 by David Waern at 2008-10-15T22:14:07+00:00 Add DeclInfo to H.Types - - - - - 193552b6 by David Waern at 2008-10-15T22:15:01+00:00 Add subordinates with docs to the declaration map The only place in the code where we want the subordinates for a declaration is right after having looked up the declaration in the map. And since we include subordinates in the map, we might as well take the opportunity to store those subordinates that belong to a particular declaration together with that declaration. We also store the documentation for each subordinate. - - - - - 31e6eebc by David Waern at 2008-10-16T17:18:47+00:00 Wibble - - - - - 0dcbd79f by David Waern at 2008-10-16T20:58:42+00:00 Fix haskell/haddock#61 We were not getting docs for re-exported class methods. This was because we were looking up the docs in a map made from the declarations in the current module being rendered. Obviously, re-exported class methods come from another module. Class methods and ATs were the only thing we were looking up using the doc map, everything else we found in the ExporItems. So now I've put subordinate docs in the ExportItem's directly, to make things a bit more consistent. To do this, I added subordinates to the the declarations in the declaration map. This was easy since we were computing subordinates anyway, to store stand-alone in the map. I added a new type synonym 'DeclInfo', which is what we call what is now stored in the map. This little refactoring removes duplicate code to retrieve subordinates and documentation from the HsGroup. - - - - - de47f20a by David Waern at 2008-10-16T22:06:35+00:00 Document function and improve its layout - - - - - e74e625a by Thomas Schilling at 2008-10-20T11:12:57+00:00 Force interface more aggressively. For running Haddock on GHC this reduces memory usage by about 50 MB on a 32 bit system. A heap profile shows total memory usage peak at about 100 MB, but actual usage is at around 300 MB even with compacting GC (+RTS -c). - - - - - b63ac9a1 by David Waern at 2008-10-20T20:25:50+00:00 Make renamer consistent Instead of explicitly making some binders Undocumented, treat all names the same way (that is, try to find a Documented name). - - - - - f6de0bb0 by Ian Lynagh at 2008-09-19T00:54:43+00:00 TAG GHC 6.10 fork - - - - - 74599cd0 by David Waern at 2008-10-20T21:13:24+00:00 Do not save hidden modules in the .haddock file We were saving interfaces of all processed modules including those hidden using {-# OPTIONS_HADDOCK hide #-} in the .haddock file. This caused broken links when generating the index for the libraries that come with GHC. This patch excludes modules with hidden documentation when writing .haddock files. It should fix the above problem. - - - - - 7b6742e9 by David Waern at 2008-10-21T19:54:52+00:00 Do not save hidden modules in the .haddock file (also for ghc >= 6.9) When writing the first patch, I forgot to do the fix in both branches of an #if macro. - - - - - b99b1951 by David Waern at 2008-10-22T20:04:18+00:00 Remove subordinate map and its usage It is not needed now that we store subordinate names in the DeclInfo map. - - - - - da97cddc by David Waern at 2008-10-22T20:11:46+00:00 Tidy up code in H.I.Create a little Remove commented out half-done type instance support, and remove DeclWithDoc synonym. - - - - - 6afa76f3 by David Waern at 2008-10-22T21:17:29+00:00 Fix warnings in H.GHC.Utils - - - - - 171ea1e8 by David Waern at 2008-10-22T21:35:04+00:00 Fix warnings in H.Utils - - - - - c8cb3b91 by David Waern at 2008-10-22T21:36:49+00:00 Wibble - - - - - 767fa06a by David Waern at 2008-10-27T19:59:04+00:00 Make named doc comments into ExportDoc instead of ExportDecl Fixes a crash when processing modules without export lists containing named docs. - - - - - e638bbc6 by David Waern at 2008-11-02T22:21:10+00:00 Add HCAR entry - - - - - 92b4ffcf by David Waern at 2008-11-02T22:44:19+00:00 Update CHANGES - - - - - 84d4da6e by David Waern at 2008-11-03T11:25:04+00:00 Add failing test for template haskell crash - - - - - 2a9cd2b1 by David Waern at 2008-11-04T21:13:44+00:00 Add tests/TH.hs - - - - - 8a59348e by David Waern at 2008-11-04T21:30:26+00:00 TAG 2.3.0 - - - - - 54f70d31 by Thomas Schilling at 2008-10-24T17:04:08+00:00 Enable framed view of the HTML documentation. This patch introduces: - A page that displays the documentation in a framed view. The left side will show a full module index. Clicking a module name will show it in the right frame. If Javascript is enabled, the left side is split again to show the modules at the top and a very short synopsis for the module currently displayed on the right. - Code to generate the mini-synopsis for each module and the mini module index ("index-frames.html"). - CSS rules for the mini-synopsis. - A very small amount of javascript to update the mini-synopsis (but only if inside a frame.) Some perhaps controversial things: - Sharing code was very difficult, so there is a small amount of code duplication. - The amount of generated pages has been doubled, since every module now also gets a mini-synopsis. The overhead should not be too much, but I haven't checked. Alternatively, the mini-synopsis could also be generated using Javascript if we properly annotate the actual synopsis. - - - - - 5d7ea5a6 by David Waern at 2008-11-04T23:20:17+00:00 Follow change to ExportDecl in frames code - - - - - 60e16308 by David Waern at 2008-11-04T23:35:26+00:00 Update CHANGES - - - - - d63fd26d by David Waern at 2008-11-04T23:37:43+00:00 Bump version number - - - - - c1660c39 by David Waern at 2008-11-04T23:44:46+00:00 Update CHANGES - - - - - 995ab384 by David Waern at 2008-11-04T23:55:21+00:00 Remove .ref files from tests/output/ - - - - - 1abbbe75 by David Waern at 2008-11-04T23:57:41+00:00 Output version info before running tests - - - - - 649b182f by David Waern at 2008-11-05T22:45:37+00:00 Add ANNOUNCE message - - - - - c36ae0bb by David Waern at 2008-11-05T23:15:35+00:00 Update ANNOUNCE - - - - - 9c4f3d40 by David Waern at 2008-11-05T23:18:30+00:00 Wibble - - - - - 5aac87ce by David Waern at 2008-11-06T21:07:48+00:00 Depend on base 4.* when using GHC >= 6.9, otherwise 3.* - - - - - b9796a74 by David Waern at 2008-11-06T21:13:40+00:00 Bump version to 2.4.1 and update CHANGES - - - - - d4b26baa by David Waern at 2008-11-06T21:26:33+00:00 Depend on base 4.0.* instead of 4.* - - - - - 2cb0903c by David Waern at 2008-11-06T21:46:53+00:00 Fix warnings in H.B.HH and H.B.HH2 - - - - - e568e89a by David Waern at 2008-11-06T21:47:12+00:00 Fix warnings in Haddock.ModuleTree - - - - - 9dc14fbd by David Waern at 2008-11-06T21:47:52+00:00 Fix warnings in Haddock.Version - - - - - 02ac197c by David Waern at 2008-11-06T21:51:31+00:00 Fix warnings in H.InterfaceFile and H.Options - - - - - 63e7439a by David Waern at 2008-11-06T21:59:45+00:00 Fix warnings in H.GHC.Typecheck - - - - - 4bca5b68 by David Waern at 2008-11-08T13:43:42+00:00 Set HscTarget to HscNothing instead of HscAsm There used to be a bug in the GHC API that prevented us from setting this value. - - - - - 07357aec by David Waern at 2008-11-09T22:27:00+00:00 Re-export NameCache and friends from Distribution.Haddock - - - - - ea554b5a by David Waern at 2008-11-09T23:14:10+00:00 Add Haddock.GHC.Utils to other-modules in library - - - - - 74aecfd7 by David Waern at 2008-11-10T01:18:57+00:00 Export DocName in the library - - - - - 241a58b3 by David Waern at 2008-11-10T01:19:18+00:00 Document the functions in H.DocName - - - - - edc2ef1b by David Waern at 2008-11-10T01:20:52+00:00 Export H.DocName in the library - - - - - 4f588d55 by David Waern at 2008-11-10T01:29:14+00:00 Make DocName an instance of NamedThing - - - - - b4647244 by David Waern at 2008-11-15T22:58:18+00:00 Reflect version bump in test suite - - - - - 4bee8ce2 by David Waern at 2008-11-15T22:58:45+00:00 Update tests For unknown reasons, test output for Bug1 and Test has changed for the better. - - - - - 1690e2f9 by David Waern at 2008-11-15T22:59:33+00:00 Store hidden modules in .haddock files We store documentation for an entity in the 'InstalledInterface' of the definition site module, and never in the same structure for a module which re-exports the entity. So when a client of the Haddock library wants to look up some documentation, he/she might need to access a hidden module. But we currently don't store hidden modules in the .haddock files. So we add the hidden modules and the Haddock options to the .haddock files. The options will be used to filter the module list to obtain the visible modules only, which is necessary for generating the contents and index for installed packages. - - - - - 8add6435 by David Waern at 2008-11-16T14:35:50+00:00 Bump major version number due to .haddock file format change - - - - - 48bfcf82 by David Waern at 2008-11-23T14:32:52+00:00 Update tests to account for version number bump - - - - - 0bbd1738 by David Waern at 2008-11-23T14:33:31+00:00 HADDOCK_DATA_DIR changed to haddock_datadir - - - - - 5088b78c by David Waern at 2008-11-23T17:13:21+00:00 FIX haskell/haddock#45: generate two anchors for each name We generate two anchor tags for each name, one where we don't escape the name and one where we URI-encode it. This is for compatibility between IE and Opera. Test output is updated. - - - - - 5ee5ca3b by Neil Mitchell at 2008-11-27T14:38:11+00:00 Drop HsDocTy annotations, they mess up pretty printing and also have a bracketing bug (#2584) - - - - - 51c014e9 by Roman Cheplyaka at 2008-11-27T22:27:36+00:00 Allow referring to a specific section within a module in a module link Fixes haskell/haddock#65 - - - - - 4094bdc5 by David Waern at 2008-11-28T21:13:33+00:00 Update tests following anchor change - - - - - f89552dd by Thomas Schilling at 2008-11-29T16:16:20+00:00 Haddock really shouldn't try to overwrite files. - - - - - 98127499 by David Waern at 2008-12-07T14:09:15+00:00 Solve conflict - - - - - 319356c5 by David Waern at 2008-10-22T21:16:55+00:00 Add -Wall -Werror to ghc-options - - - - - 3c4968c9 by David Waern at 2008-11-04T23:38:56+00:00 TAG 2.4.0 - - - - - 4b21e003 by David Waern at 2008-11-06T21:14:04+00:00 TAG 2.4.1 - - - - - 8e0cad5c by David Waern at 2008-12-07T14:12:54+00:00 Remove -Werror - - - - - 299d6deb by David Waern at 2008-12-07T14:25:18+00:00 Remove -Wall, we'll focus on warnings after 6.10.2 is out - - - - - 5f4216b6 by David Waern at 2008-12-07T20:58:05+00:00 Resolve conflict properly - - - - - 67d774e7 by Neil Mitchell at 2008-12-15T11:44:26+00:00 Make forall's in constructors explicit, i.e. data Foo = Foo {foo :: Eq a => a} - - - - - 61851792 by Neil Mitchell at 2008-12-18T15:39:39+00:00 Try and find a better package name than "main" for Hoogle, goes wrong when working on an executable rather than a library - - - - - 2fab8554 by David Waern at 2008-12-08T23:19:48+00:00 Make visible names from ExportItems Instead of a complicated calculation of visible names out of GHC's export items, we can get them straight out of the already calculated ExportItems. The ExportItems should represent exactly those items that are visible in an interface. If store all the exported sub-names in ExportDecl instead of only those with documentation, the calculation becomes very simple. So we do this change as well (should perhaps have been a separate patch). This should fix the problem with names from ghc-prim not appearing in the link environment. - - - - - 7caadd8c by Ian Lynagh at 2008-12-11T17:01:04+00:00 Wrap the GHC usage with defaultCleanupHandler This fixes a bug where haddock leaves /tmp/ghc* directories uncleaned. - - - - - 7c9fc9a5 by David Waern at 2009-01-02T21:38:27+00:00 Show re-exported names from external packages again This fixes GHC ticket 2746. In order to also link to the exported subordinate names of a declaration, we need to re-introduce the sub map in the .haddock files. - - - - - 119e4e05 by David Waern at 2009-01-06T23:34:17+00:00 Do not process boot modules We should of course not try to produce documentation for boot modules! The reason this has worked in the past is that the output of "real" modules overwrites the output of boot modules later in the process. However, this causes a subtle link environment problem. So let's get rid of this stupid behaviour. We avoid processing boot modules, but we continue to typecheck them. - - - - - c285b9d2 by David Waern at 2009-01-08T18:03:36+00:00 Export modules also when coming from external packages This seems to have regressed since a refactoring that was part of the 2.3.0 release. - - - - - 24031c17 by David Waern at 2009-01-10T15:26:26+00:00 Change version to 2.4.2 - no need to go to 2.5.0 - - - - - 864d1c3f by David Waern at 2009-01-10T15:35:20+00:00 Update tests to account for version number change - - - - - 524ba886 by David Waern at 2009-01-10T18:29:17+00:00 Add test for Template Haskell splicing - - - - - 05e6e003 by David Waern at 2009-01-10T19:35:42+00:00 Fix Trac haskell/haddock#68: Turn on compilation via C for Template Haskell packages We can't use HscNothing if we need to run code coming from modules inside the processed package during typechecking, which is the case for some packages using Template Haskell. This could be improved, to e.g. use HscInterpreted and HscNothing where possible, instead of using HscC for all modules in the package. - - - - - 2b2bafa1 by David Waern at 2009-01-10T20:22:25+00:00 Only use needsTemplateHaskell when compiling with GHC 6.10.2 or above - - - - - bedc3a93 by Ian Lynagh at 2009-01-11T14:58:41+00:00 Fix the location of INPLACE_PKG_CONF; fixes the build Spotted by Conal Elliott - - - - - 943107c8 by David Waern at 2009-01-20T19:27:39+00:00 Document H.I.Create.collectDocs better - - - - - c6252e37 by David Waern at 2009-01-20T19:29:51+00:00 Fix Trac haskell/haddock#59: TH-generated declarations disappearing This patch was contributed by Joachim Breitner (nomeata). - - - - - 3568a6af by David Waern at 2009-01-21T21:41:48+00:00 Do not indicate that a constructor argument is unboxed We only show the strictness annotation for an unboxed constructor argument. The fact that it is unboxed is an implementation detail and should not be part of the module interface. - - - - - 562a4523 by David Waern at 2009-01-22T18:53:49+00:00 Fix Trac haskell/haddock#50: do not attach docs to pragmas or other kinds of non-declarations We now filter out everything that is not a proper Haskell declaration before collecting the docs and attaching them to declarations. - - - - - 6fdf21c2 by David Waern at 2009-01-22T19:48:09+00:00 Add test for quasi quotation. No reference output yet. - - - - - dc4100fd by David Waern at 2009-01-22T19:57:47+00:00 Improve quasi-quotation test and add reference output - - - - - 908b74bb by David Waern at 2009-01-23T23:22:03+00:00 Filter out separately exported associated types in a smarter way - - - - - f6b42ecb by David Waern at 2009-01-24T16:54:39+00:00 Correct spelling mistake in error message - - - - - 24e4245d by David Waern at 2009-01-24T17:48:03+00:00 Correct comment - - - - - b5e8462f by David Waern at 2009-02-07T13:22:29+00:00 Do not show a subordinate at the top level if its parent is also exported See note in the source code for more info. - - - - - 4b09de57 by David Waern at 2009-02-07T13:53:53+00:00 Update test following change to top level subordinates - - - - - 76379896 by David Waern at 2009-02-07T13:58:04+00:00 Remove html files in the tests/output/ directory which have been accidentally added - - - - - 1a6d8b10 by Joachim Breitner at 2009-02-20T10:29:43+00:00 Typo in comment - - - - - fec367d0 by David Waern at 2009-02-24T20:21:17+00:00 Fix small bug The rule is to prefer type constructors to other things when an identifier in a doc string can refer to multiple things. This stopped working with newer GHC versions (due to a tiny change in the GHC renamer). We implement this rule in the HTML backend for now, instead of fixing it in GHC, since we will move renaming of doc strings to Haddock in the future anyway. - - - - - 9b4172eb by David Waern at 2009-02-25T20:04:38+00:00 Fix bad error handling with newer GHCs When support for GHC 6.10 was added, an error handler was installed only around the typechecking phase. This had the effect that errors thrown during dependency chasing were caught in the top-level exception handler and not printed with enough detail. With this patch we wrap the error handler around all our usage of the Ghc monad. - - - - - de2df363 by Simon Peyton Jones at 2009-02-02T16:47:42+00:00 Hide funTyConName, now exported by TypeRep - - - - - 4d40a29f by Ian Lynagh at 2009-02-12T18:57:49+00:00 Don't build the library when building in the GHC tree - - - - - 1cd0abe4 by Ian Lynagh at 2009-02-13T13:58:53+00:00 Add a ghc.mk - - - - - 3d814eeb by Ian Lynagh at 2009-02-13T18:50:28+00:00 do .depend generation for haddock with the stage1 compiler This is a bit of a hack. We mkdepend with stage1 as if .depend depends on the stage2 compiler then make goes wrong: haddock's .depend gets included, which means that make won't reload until it's built, but we can't build it without the stage2 compiler. We therefore build the stage2 compiler before its .depend file is available, and so compilation fails. - - - - - b55036a4 by Ian Lynagh at 2009-02-25T01:38:13+00:00 Give haddock a wrapper on unix in the new GHC build system - - - - - 9eabfe68 by Ian Lynagh at 2009-02-25T19:21:32+00:00 Create inplace/lib/html in the new GHC build system - - - - - 93af30c7 by Ian Lynagh at 2008-11-07T19:18:23+00:00 TAG GHC 6.10.1 release - - - - - 06e6e34a by Thomas Schilling at 2009-02-24T18:11:00+00:00 Define __GHC_PATCHLEVEL__ for recent version of GHC (stable). - - - - - 680e6ed8 by Thomas Schilling at 2009-02-24T18:12:26+00:00 'needsTemplateHaskell' is not defined in current stable GHC. - - - - - 6c5619df by David Waern at 2009-02-25T22:15:23+00:00 Hide fynTyConName only for recent GHC versions - - - - - 6b2344f1 by Ian Lynagh at 2009-02-26T00:49:56+00:00 Add the module to one of haddocks warnings - - - - - e5d11c70 by David Waern at 2009-02-27T21:37:20+00:00 Bug fix We tried to filter out subordinates that were already exported through their parent. This didn't work properly since we were in some cases looking at the grand-parent and not the parent. We now properly compute all the parent-child relations of a declaration, and use this information to get the parent of a subordinate. We also didn't consider record fields with multiple parents. This is now handled correctly. We don't currently support separately exported associated types. But when we do, they should be handled correctly by this process too. Also slightly improved the warning message that we give when filtering out subordinates. - - - - - 10a79a60 by David Waern at 2009-02-27T22:08:08+00:00 Fix error message conflict The module name is already written in the beginning of the message, as seems to be the convention in Haddock. Perhaps not so clear, but we should change it everywhere in that case. Leaving it as it is for now. - - - - - c5055c7f by David Waern at 2009-02-27T22:15:17+00:00 Shorten warning message - - - - - a72fed3a by David Waern at 2009-02-28T00:53:55+00:00 Do not show package name in warning message - - - - - a5daccb2 by Ian Lynagh at 2009-03-01T14:59:35+00:00 Install haddock in the new GHC build system - - - - - dfdb025c by Ian Lynagh at 2009-03-07T23:56:29+00:00 Relax base dependency to < 4.2, not < 4.1 - - - - - 5769c8b4 by David Waern at 2009-03-21T14:58:52+00:00 Bump .haddock file version number (due to change of format) - - - - - f1b8f67b by David Waern at 2009-03-21T14:59:26+00:00 Define __GHC_PATCHLEVEL__=1 when using ghc-6.10.1 - - - - - 23f78831 by David Waern at 2009-03-21T16:40:52+00:00 Update CHANGES - - - - - 7d2735e9 by David Waern at 2009-03-21T16:50:33+00:00 Update ANNOUNCE - - - - - 0771e00a by David Waern at 2009-03-21T16:54:40+00:00 Update ANNOUNCE, again - - - - - 81a6942a by David Waern at 2009-03-21T17:50:06+00:00 Don't be too verbose in CHANGES - - - - - 29861dcf by David Waern at 2009-03-21T18:03:31+00:00 TAG 2.4.2 - - - - - a585f285 by David Waern at 2009-03-21T19:20:29+00:00 Require Cabal >= 1.2.3 - - - - - 7c611662 by David Waern at 2009-03-21T19:21:48+00:00 TAG 2.4.2 with cabal-version >= 1.2.3 - - - - - 23b7deff by Simon Marlow at 2009-03-20T15:43:42+00:00 new GHC build system: use shell-wrappers macro - - - - - 25f8afe7 by Ian Lynagh at 2009-03-21T19:13:53+00:00 Fix (with a hack?) haddock in teh new build system - - - - - 6a29a37e by David Waern at 2009-03-24T22:10:15+00:00 Remove unnecessary LANGUAGE pragma - - - - - 954da57d by David Waern at 2009-03-24T22:21:23+00:00 Fix warnings in H.B.DevHelp - - - - - 1619f1df by David Waern at 2009-03-26T23:20:44+00:00 -Wall police in H.B.Html - - - - - b211e13b by Simon Marlow at 2009-03-24T13:00:56+00:00 install Haddock's html stuff - - - - - 78e0b107 by David Waern at 2008-12-07T19:58:53+00:00 Add verbosity flag and utils, remove "verbose" flag - - - - - 913dae06 by David Waern at 2008-12-07T20:01:05+00:00 Add some basic "verbose" mode logging in H.Interface - - - - - 1cbff3bf by David Waern at 2009-03-27T00:07:26+00:00 Fix conflicts - - - - - 22f82032 by David Waern at 2009-03-27T21:15:11+00:00 Remove H.GHC.Typecheck - - - - - 81557804 by David Waern at 2009-03-27T21:19:22+00:00 Remove docNameOrig and use getName everywhere instead - - - - - d8267213 by David Waern at 2009-03-27T21:21:46+00:00 Use docNameOcc instead of nameOccName . getName - - - - - 5d55deab by David Waern at 2009-03-27T21:33:04+00:00 Remove H.DocName and put DocName in H.Types - - - - - 8ba72611 by David Waern at 2009-03-27T22:06:26+00:00 Document DocName - - - - - 605f8ca5 by David Waern at 2009-03-27T22:45:21+00:00 -Wall police - - - - - e4da93ae by David Waern at 2009-03-27T23:12:53+00:00 -Wall police in H.B.Hoogle - - - - - bb255519 by David Waern at 2009-03-27T23:41:28+00:00 Define Foldable and Traversable instances for Located - - - - - f1195cfe by David Waern at 2009-03-27T23:51:34+00:00 Wibble - - - - - 23818d7c by David Waern at 2009-03-28T00:03:55+00:00 -Wall police in H.I.Rename - - - - - 0f050d67 by David Waern at 2009-03-28T00:15:15+00:00 -Wall police in H.I.AttachInstances - - - - - 0f3fe038 by David Waern at 2009-03-28T21:09:41+00:00 Wibble - - - - - 275d4865 by David Waern at 2009-03-28T21:27:06+00:00 Layout fix - - - - - 54ff0ef8 by David Waern at 2009-03-28T21:59:07+00:00 -Wall police in H.I.Create - - - - - 7f58b117 by David Waern at 2009-03-28T22:10:19+00:00 -Wall police in H.Interface - - - - - f0c03b44 by David Waern at 2009-03-28T22:22:59+00:00 -Wall police in Main - - - - - 29da355c by David Waern at 2009-03-28T22:23:39+00:00 Turn on -Wall -Werror - - - - - 446d3060 by David Waern at 2009-04-01T20:40:30+00:00 hlint police - - - - - 3867c9fc by David Waern at 2009-04-01T20:48:42+00:00 hlint police - - - - - bd1f1600 by David Waern at 2009-04-01T20:58:02+00:00 hlint police - - - - - e0e90866 by David Waern at 2009-04-05T12:42:53+00:00 Move H.GHC.Utils to H.GhcUtils - - - - - 9cbd426b by David Waern at 2009-04-05T12:57:21+00:00 Remove Haddock.GHC and move its (small) contents to Main - - - - - b5c2cbfd by David Waern at 2009-04-05T13:07:04+00:00 Fix whitespace and stylistic issues in Main - - - - - 3c04aa56 by porges at 2008-12-07T08:22:19+00:00 add unicode output - - - - - 607918da by David Waern at 2009-04-26T15:09:43+00:00 Resolve conflict - - - - - 4bec6b6b by Simon Marlow at 2009-05-13T10:00:31+00:00 fix markup - - - - - 436ad6f4 by Simon Marlow at 2009-03-23T11:54:45+00:00 clean up - - - - - bdcd1398 by Simon Marlow at 2009-03-24T10:36:45+00:00 new GHC build system: add $(exeext) - - - - - 9c0972f3 by Simon Marlow at 2009-03-24T11:04:31+00:00 update for new GHC build system layout - - - - - d0f3f83a by Ian Lynagh at 2009-03-29T15:31:43+00:00 GHC new build system fixes - - - - - 5a8245c2 by Ian Lynagh at 2009-04-04T20:44:23+00:00 Tweak new build system - - - - - 9c6f2d7b by Simon Marlow at 2009-05-13T10:01:27+00:00 add build instructions for GHC - - - - - 66d07c76 by Ian Lynagh at 2009-05-31T00:37:53+00:00 Quote program paths in ghc.mk - - - - - bb7de2cd by Ian Lynagh at 2009-06-03T22:57:55+00:00 Use a bang pattern on an unlifted binding - - - - - 3ad283fc by Ian Lynagh at 2009-06-13T16:17:50+00:00 Include haddock in GHC bindists - - - - - ac447ff4 by David Waern at 2009-06-24T21:07:50+00:00 Delete Haddock.Exception and move contents to Haddock.Types Only a few lines of code that mainly declares a type - why not just put it in Haddock.Types. - - - - - 4464fb9b by David Waern at 2009-06-24T22:23:23+00:00 Add Haddock module headers Add a proper Haddock module header to each module, with a more finegrained copyright. If you feel mis-accreditted, please correct any copyright notice! The maintainer field is set to haddock at projects.haskell.org. Next step is to add a brief description to each module. - - - - - 5f4c95dd by David Waern at 2009-06-24T22:39:44+00:00 Fix spelling error - - - - - 6d074cdb by David Waern at 2009-06-25T21:53:56+00:00 Document Interface and InstalledInterface better - - - - - d0cbd183 by David Waern at 2009-06-27T12:46:46+00:00 Remove misplaced whitespace in H.I.Rename - - - - - fa381c49 by David Waern at 2009-06-27T13:26:03+00:00 Fix haskell/haddock#104 - create output directory if missing - - - - - 91fb77ae by Ian Lynagh at 2009-06-25T15:59:50+00:00 TAG 2009-06-25 - - - - - 0d853f40 by Simon Peyton Jones at 2009-07-02T15:35:22+00:00 Follow extra field in ConDecl - - - - - b201735d by Ian Lynagh at 2009-07-05T16:50:35+00:00 Update Makefile for the new GHC build system - - - - - df6c0092 by Ian Lynagh at 2009-07-05T17:01:13+00:00 Resolve conflicts - - - - - 1066870a by Ian Lynagh at 2009-07-05T17:01:48+00:00 Remove the -Wwarn hack in the GHC build system - - - - - 7e856076 by Ian Lynagh at 2009-07-05T17:17:59+00:00 Fix warnings - - - - - 5d4cd958 by Ian Lynagh at 2009-07-05T19:35:40+00:00 Bump version number Cabal needs to distinguish between haddocks having a --verbose and --verbosity flag - - - - - 6ee07c99 by David Waern at 2009-07-06T20:14:57+00:00 Wibble - - - - - 2308b66f by David Waern at 2009-07-06T20:24:20+00:00 Clearer printing of versions by runtests.hs - - - - - d4b5d9ab by David Waern at 2009-07-06T21:22:42+00:00 Fix (invisible) bug introduced by unicode patch - - - - - 2caca8d8 by David Waern at 2009-07-06T21:44:10+00:00 Use HscAsm instead of HscC when using TH - - - - - 18f3b755 by David Waern at 2009-07-06T22:10:22+00:00 Update HCAR entry (by Janis) - - - - - a72ac9db by David Waern at 2009-07-06T23:01:35+00:00 Follow HsRecTy change with an #if __GLASGOW_HASKEL__ >= 611 - - - - - 549135d2 by David Waern at 2009-07-06T23:11:41+00:00 Remove unused functions from Haddock.Utils - - - - - b450134a by Isaac Dupree at 2009-07-11T14:59:00+00:00 revert to split-index for large indices - remove the search-box, because browsers have search-for-text abilities anyway. - pick 150 items in index as the arbitrary time at which to split it - notice the bug that identifiers starting with non-ASCII characters won't be listed in split-index, but don't bother to fix it yet (see ticket haskell/haddock#116, http://trac.haskell.org/haddock/ticket/116 ) - - - - - 78a5661e by Isaac Dupree at 2009-07-20T15:37:18+00:00 Implement GADT records in HTML backend - - - - - 4e163555 by Isaac Dupree at 2009-07-21T22:03:25+00:00 add test for GADT records - - - - - 79aa4d6e by David Waern at 2009-07-23T20:40:37+00:00 Update test suite following version bump - - - - - 5932c011 by David Waern at 2009-08-02T10:25:39+00:00 Fix documentation bug - - - - - a6970fca by David Waern at 2009-08-12T23:08:53+00:00 Remove support for ghc 6.8.* from .cabal file - - - - - c1695902 by Ian Lynagh at 2009-07-07T13:35:45+00:00 Fix unused import warnings - - - - - fb6df7f9 by Ian Lynagh at 2009-07-16T00:20:31+00:00 Use cProjectVersion directly rather than going through compilerInfo Fixes the build after changes in GHC - - - - - 548cdd66 by Simon Marlow at 2009-07-28T14:27:04+00:00 follow changes in GHC's ForeignType - - - - - 9395aaa0 by David Waern at 2009-08-13T22:17:33+00:00 Switch from PatternSignatures to ScopedTypeVariables in Main - - - - - eebf39bd by David Waern at 2009-08-14T17:14:28+00:00 Version .haddock files made with GHC 6.10.3/4 correclty - - - - - 58f3e735 by David Waern at 2009-08-14T17:19:37+00:00 Support GHC 6.10.* and 6.11.* only - - - - - 5f63cecc by David Waern at 2009-08-14T22:03:20+00:00 Do not version .haddock file based on GHC patchlevel version We require that the instances of Binary that we use from GHC will not change between patchlevel versions. - - - - - d519de9f by David Waern at 2009-08-14T23:50:00+00:00 Update CHANGES - - - - - 35dccf5c by David Waern at 2009-08-14T23:51:38+00:00 Update version number everywhere - - - - - 6d363fea by David Waern at 2009-08-15T09:46:49+00:00 Update ANNOUNCE - - - - - c7ee6bc2 by David Waern at 2009-08-15T09:47:13+00:00 Remove -Werror Forgot that Hackage doesn't like it. - - - - - a125c12b by David Waern at 2009-08-15T09:49:50+00:00 Require Cabal >= 1.6 - - - - - adb2f560 by Isaac Dupree at 2009-08-12T03:47:14+00:00 Cross-Package Documentation version 4 - - - - - 3d6dc04d by David Waern at 2009-08-15T23:42:57+00:00 Put all the IN_GHC_TREE stuff inside getGhcLibDir - - - - - 56624097 by David Waern at 2009-08-15T23:52:03+00:00 Add --print-ghc-libdir - - - - - f15d3ccb by David Waern at 2009-08-16T00:37:52+00:00 Read base.haddock when running tests We can now test cross-package docs. - - - - - 283f0fb9 by David Waern at 2009-08-16T00:50:59+00:00 Update test output - we now have more links - - - - - 673d1004 by David Waern at 2009-08-16T01:26:08+00:00 Read process.haddock when running tests - - - - - 0d127f82 by David Waern at 2009-08-16T01:43:04+00:00 Add a test for cross-package documentation - - - - - f94db967 by Ian Lynagh at 2009-08-16T18:42:44+00:00 Follow GHC build system changes - - - - - 5151278a by Isaac Dupree at 2009-08-16T19:58:05+00:00 make cross-package list types look nicer - - - - - c41e8228 by Isaac Dupree at 2009-08-18T01:47:47+00:00 Haddock.Convert: export more functions This lets us remove some code in Haddock.Interface.AttachInstances - - - - - 2e5fa398 by Isaac Dupree at 2009-08-18T02:11:05+00:00 switch AttachInstances to use synify code It changed an instance from showing ((,) a b) to (a, b) because my synify code is more sophisticated; I hope the latter is a good thing rather than a bad thing aesthetically, here. But this definitely reduces code duplication! - - - - - b8b07123 by Isaac Dupree at 2009-08-18T02:23:31+00:00 Find instances using GHC, which is more complete. In particular, it works cross-package. An intermediate patch also moved the instance-finding into createInterface, but that move turned out not to be necessary, so if we want to do that, it'd go in a separate patch. (Is that possible? Or will we need GHC to have loaded all the modules first, before we can go searching for the instances (e.g. if the modules are recursive or something)?) - - - - - 6959b451 by Isaac Dupree at 2009-08-17T00:37:18+00:00 fix preprocessor conditional sense - - - - - 942823af by Isaac Dupree at 2009-08-16T22:46:48+00:00 remove ghc 6.8 conditionals from Haddock.Interface - - - - - 4b3ad888 by Isaac Dupree at 2009-08-18T20:24:38+00:00 Fix GHC 6.11 build in Haddock.Convert - - - - - 0a89c5ab by Isaac Dupree at 2009-08-23T00:08:58+00:00 hacks to make it compile without fnArgDocsn - - - - - 7b3bed43 by Isaac Dupree at 2009-08-23T03:01:28+00:00 less big-Map-based proper extraction of constructor subdocs - - - - - b21c279a by Isaac Dupree at 2009-08-23T03:02:06+00:00 Html: remove unnecessary+troublesome GHC. qualifications - - - - - 96c97115 by Isaac Dupree at 2009-08-23T03:08:03+00:00 Move doc parsing/lexing into Haddock for ghc>=6.11 - - - - - e1cec02d by Isaac Dupree at 2009-08-23T05:08:14+00:00 get rid of unused DocMap parameter in Html - - - - - 66960c59 by Isaac Dupree at 2009-08-23T05:54:20+00:00 fix horrible named-docs-disappearing bug :-) - - - - - a9d7eff3 by Isaac Dupree at 2009-08-23T06:26:36+00:00 re-implement function-argument docs ..on top of the lexParseRn work. This patch doesn't change the InstalledInterface format, and thus, it does not work cross-package, but that will be easy to add subsequently. - - - - - 8bf6852c by Isaac Dupree at 2009-08-23T07:26:05+00:00 cross-package fnArgDocs. WARNING: changes .haddock binary format While breaking the format, I took the opportunity to unrename the DocMap that's saved to disk, because there's really no reason that we want to know what *another* package's favorite place to link a Name to was. (Is that true? Or might we want to know, someday?) Also, I added instance Binary Map in InterfaceFile. It makes the code a little simpler without changing anything of substance. Also it lets us add another Map hidden inside another Map (fnArgsDocs in instDocMap) without having really-convoluted serialization code. Instances are neat! I don't understand why this change to InterfaceFile seemed to subtly break binary compatibility all by itself, but no matter, I'll just roll it into the greater format-changing patch. Done! - - - - - 30115a64 by Isaac Dupree at 2009-08-23T18:22:47+00:00 Improve behavior for unfindable .haddock - - - - - aa364bda by Isaac Dupree at 2009-08-23T18:28:16+00:00 add comment for FnArgsDoc type - - - - - 49b23a99 by Isaac Dupree at 2009-08-23T21:52:48+00:00 bugfix: restore fnArgDocs for type-synonyms - - - - - f65f9467 by Isaac Dupree at 2009-08-23T22:06:55+00:00 Backends.Hoogle: eliminate warnings - - - - - a292d216 by Isaac Dupree at 2009-08-23T22:10:24+00:00 Haddock.Convert: eliminate warnings - - - - - 5546cd20 by Isaac Dupree at 2009-08-23T22:12:31+00:00 Haddock.Interface.Rename: eliminate warnings - - - - - 0a9798b6 by Isaac Dupree at 2009-08-23T22:18:47+00:00 Main.hs: remove ghc<6.9 conditionals - - - - - e8f9867f by Isaac Dupree at 2009-08-23T22:27:46+00:00 Main.hs: eliminate warnings (except for OldException) - - - - - 61c64247 by Isaac Dupree at 2009-08-23T22:41:01+00:00 move get*LibDir code in Main.hs, to +consistent code, -duplication - - - - - 948f1e69 by Isaac Dupree at 2009-08-23T23:14:26+00:00 Main.hs: OldException->Exception: which eliminates warnings - - - - - 3d5d5e03 by Isaac Dupree at 2009-08-23T23:20:11+00:00 GhcUtils: ghc >= 6.10 - - - - - 2771d657 by Isaac Dupree at 2009-08-23T23:21:55+00:00 InterfaceFile: ghc >= 6.10 - - - - - d9f2b9d1 by Isaac Dupree at 2009-08-23T23:22:58+00:00 Types: ghc >= 6.10 - - - - - ca39210e by Isaac Dupree at 2009-08-23T23:23:26+00:00 ModuleTree: ghc >= 6.10 - - - - - 883c4e59 by Isaac Dupree at 2009-08-23T23:24:04+00:00 Backends.DevHelp: ghc >= 6.10 - - - - - 04667df5 by Isaac Dupree at 2009-08-23T23:24:37+00:00 Backends.Html: ghc >= 6.10 - - - - - a9f7f25f by Isaac Dupree at 2009-08-23T23:25:24+00:00 Utils: ghc >= 6.10 - - - - - b7105022 by Isaac Dupree at 2009-08-23T23:37:47+00:00 eliminate haskell98 dependency, following GHC's example It turns out I/we already had, and it was only a matter of deleting it from the cabal file. - - - - - 292e0911 by Isaac Dupree at 2009-08-24T01:22:44+00:00 refactor out subordinatesWithNoDocs dep of inferenced-decls fix - - - - - c2ed46a2 by Isaac Dupree at 2009-08-24T01:24:03+00:00 Eradicate wrong runtime warning for type-inferenced exported-functions see the long comment in the patch for why I did it this way :-) - - - - - 4ac0b57c by David Waern at 2009-09-04T22:56:20+00:00 Clean up tyThingToHsSynSig a little Factor out noLoc and use the case construct. Also rename the function to tyThingToLHsDecl, since it doesn't just create type signatures. - - - - - 28ab9201 by David Waern at 2009-09-04T22:58:50+00:00 Wibble - - - - - 0d9fe6d0 by David Waern at 2009-09-06T18:39:30+00:00 Add more copyright owners to H.I.AttachInstances - - - - - 122441b1 by David Waern at 2009-09-06T18:44:12+00:00 Style police - - - - - 1fa79463 by David Waern at 2009-09-06T18:57:45+00:00 Move toHsInstHead to Haddock.Convert and call it synifyInstHead - - - - - 0d42a8aa by David Waern at 2009-09-06T21:11:38+00:00 Use colordiff to display test results if available - - - - - ea9d8e03 by Simon Marlow at 2009-08-24T08:46:14+00:00 Follow changes in GHC's interface file format Word32 instead of Int for FastString and Name offsets - - - - - 537e051e by Simon Marlow at 2009-07-29T14:16:53+00:00 define unpackPackageId (it was removed from GHC) - - - - - 50c63aa7 by David Waern at 2009-09-09T23:18:03+00:00 Remove commented-out code - - - - - 511631fe by David Waern at 2009-09-09T23:19:05+00:00 Correct copyright in H.I.ParseModuleHeader - - - - - 898ec768 by David Waern at 2009-09-11T11:22:29+00:00 Use Map.fromList/toList intead of fromAscList/toAscList when serializing Maps This fixes the missing docs problem. The Eq and Ord instances for Name uses the unique number in Name. This number is created at deserialization time by GHC's magic Binary instance for Name, and it is random. Thus, fromAscList can't be used at deserialization time, even though toAscList was used at serialization time. - - - - - 37bec0d5 by Simon Peyton Jones at 2009-09-11T08:28:04+00:00 Track change in HsType - - - - - eb3a97c3 by Ian Lynagh at 2009-09-11T16:07:09+00:00 Allow building with base 4.2 - - - - - bb4205ed by Ian Lynagh at 2009-09-22T13:50:02+00:00 Loosen the GHC dependency - - - - - 5c75deb2 by Ian Lynagh at 2009-09-22T14:08:39+00:00 Fix building with GHC >= 6.12 - - - - - fb131481 by David Waern at 2009-09-11T11:24:48+00:00 Update runtests.hs to work with GHC 6.11 - - - - - ac3a419d by David Waern at 2009-09-11T11:25:14+00:00 Update CrossPackageDocs test - - - - - ec65c3c6 by David Waern at 2009-09-11T11:25:40+00:00 Add reference output for CrossPackageDocs - - - - - 520c2758 by Ian Lynagh at 2009-10-25T17:26:40+00:00 Fix installation in the GHC build system - - - - - 28b3d7df by Ian Lynagh at 2009-11-05T15:57:27+00:00 GHC build system: Make *nix installation work in paths containing spaces - - - - - 5c9bb541 by David Waern at 2009-11-14T11:56:39+00:00 Track change in HsType for the right compiler version - - - - - 905097ce by David Waern at 2009-11-14T12:10:47+00:00 hlint police - - - - - 04920630 by Ian Lynagh at 2009-11-20T13:46:30+00:00 Use defaultObjectTarget rather than HscAsm This fixes haddock when we don't have a native code generator - - - - - 966eb079 by David Waern at 2009-11-15T12:32:21+00:00 Remove commented-out code - - - - - 37f00fc4 by David Waern at 2009-11-22T13:58:48+00:00 Make runtests.hs strip links before diffing Generates easier to read diffs when tests fail. The content of the links is not important anyway since it is not taken into account by the tests. - - - - - 3a9bb8ef by David Waern at 2009-11-22T14:05:06+00:00 Follow findProgramOnPath signature change in runtests.hs - - - - - b26b9e5a by David Waern at 2009-11-22T14:08:40+00:00 Follow removal of GHC.MVar from base in CrossPackageDocs - - - - - f4d90ae4 by David Waern at 2009-11-22T14:48:47+00:00 Make copy.hs strip link contents before copying No more updating of reference files when URLs in links changes. - - - - - 4c9c420d by David Waern at 2009-11-22T15:26:41+00:00 Update test reference output * More links (Int, Float etc) * Stripped link contents - - - - - a62b80e3 by David Waern at 2009-11-23T23:19:39+00:00 Update CrossPackageDocs reference output - Remove GHC.MVar import (removed from base) - Strip link contents - - - - - 43491394 by David Waern at 2009-11-23T23:20:00+00:00 Update test reference files with comments on instances - - - - - 0d370a0b by David Waern at 2009-11-23T23:25:16+00:00 Bump version number - - - - - 2293113e by David Waern at 2009-11-24T20:55:49+00:00 Comments on instances Implementing this was a little trickier than I thought, since we need to match up instances from the renamed syntax with instances represented by InstEnv.Instance. This is due to the current design of Haddock, which matches comments with declarations from the renamed syntax, while getting the list of instances of a class/family directly using the GHC API. - Works for class instances only (Haddock has no support for type family instances yet) - The comments are rendered to the right of the instance head in the HTML output - No change to the .haddock file format - Works for normal user-written instances only. No comments are added on derived or TH-generated instances - - - - - bf586f29 by David Waern at 2009-11-27T22:05:15+00:00 Whitespace police - - - - - b8f03afa by David Waern at 2009-11-27T22:11:46+00:00 Remove bad whitespace and commented-out pieces - - - - - 90b8ee90 by David Waern at 2009-11-27T22:15:04+00:00 Whitespace police - - - - - b5ede900 by David Waern at 2009-11-27T22:15:50+00:00 Whitespace police - - - - - e3fddbfe by David Waern at 2009-11-28T13:37:59+00:00 Remove Name from DocInstance It's not used. - - - - - 9502786c by David Waern at 2009-11-28T13:56:54+00:00 Require at least GHC 6.12 While regression testing Haddock, I found a bug that happens with GHC 6.10.3, but not with GHC 6.12-rc2 (haven't tried 6.10.4). I don't have time to track it down. I think we should just always require the latest major GHC version. The time spent on making Haddock work with older versions is too high compared to the time spent on bugfixing, refactoring and features. - - - - - 8fa688d8 by David Waern at 2009-11-28T15:05:03+00:00 Remove cruft due to compatibility with older GHCs - - - - - 46fbbe9d by David Waern at 2009-11-28T15:07:50+00:00 Add a documentation header to Haddock.Convert - - - - - c3d2cc4a by David Waern at 2009-11-28T15:10:14+00:00 Remove unused H.Utils.FastMutInt2 - - - - - 490aba80 by David Waern at 2009-11-28T15:36:36+00:00 Rename Distribution.Haddock into Documentation.Haddock - - - - - 33ee2397 by David Waern at 2009-11-28T15:36:47+00:00 Fix error message - - - - - a5a3b950 by David Waern at 2009-11-28T16:58:39+00:00 Add a test flag that brings in QuickCheck - - - - - fa049e13 by David Waern at 2009-11-28T19:32:18+00:00 Say that we want quickcheck 2 - - - - - f32b0d9b by David Waern at 2009-11-28T19:32:40+00:00 Add an Arbitrary instance for HsDoc - - - - - da9a8bd7 by David Waern at 2009-11-28T20:15:30+00:00 Rename HsDoc back into Doc - - - - - edb60101 by David Waern at 2009-11-28T22:16:16+00:00 Move H.Interface.Parse/Lex to H.Parse/Lex These are not just used to build Interfaces. - - - - - 0656a9b8 by David Waern at 2009-11-28T23:12:14+00:00 Update version number in test suite - - - - - 5e8c6f4a by David Waern at 2009-12-21T14:12:41+00:00 Improve doc of DocName - - - - - 7868e551 by Ian Lynagh at 2009-09-22T10:43:03+00:00 TAG GHC 6.12-branch created - - - - - 0452a3ea by Ian Lynagh at 2009-12-15T12:46:07+00:00 TAG GHC 6.12.1 release - - - - - 65e9be62 by David Waern at 2009-12-21T16:58:58+00:00 Update CHANGES - - - - - 145cee32 by David Waern at 2009-12-21T16:59:09+00:00 TAG 2.6.0 - - - - - 3c552008 by David Waern at 2009-12-22T17:11:14+00:00 Update ANNOUNCE - - - - - 931f9db4 by David Waern at 2010-01-22T19:57:17+00:00 Convert haddock.vim to use unix newlines - - - - - 4e56588f by David Waern at 2010-01-22T22:11:17+00:00 Remove unnecessary (and inexplicable) uses of nub - - - - - 744bb4d1 by David Waern at 2010-01-22T22:12:14+00:00 Follow move of parser and lexer - - - - - e34bab14 by David Waern at 2010-01-22T22:49:13+00:00 Use findProgramLocation instead of findProgramOnPath in runtests.hs - - - - - 8d39891b by Isaac Dupree at 2010-01-14T18:53:18+00:00 fix html arg-doc off-by-one and silliness - - - - - 9401f2e9 by David Waern at 2010-01-22T22:57:03+00:00 Create a test for function argument docs - - - - - 507a82d7 by David Waern at 2010-01-22T23:24:47+00:00 Put parenthesis around type signature arguments of function type - - - - - 8a305c28 by David Waern at 2010-01-23T17:26:59+00:00 Add reference file for the FunArgs test - - - - - 1309d5e1 by David Waern at 2010-01-24T16:05:08+00:00 Improve FunArg test and update Test.html.ref - - - - - 2990f055 by Yitzchak Gale at 2010-02-14T16:03:46+00:00 Do not generate illegal character in HTML ID attribute. - - - - - c5bcab7a by David Waern at 2010-02-22T22:10:30+00:00 Fix Haddock markup error in comment - - - - - c6416a73 by David Waern at 2010-02-24T22:55:08+00:00 Large additions to the Haddock API Also improved and added more doc comments. - - - - - 57d289d7 by David Waern at 2010-02-24T22:58:02+00:00 Remove unused ifaceLocals - - - - - 80528d93 by David Waern at 2010-02-25T21:05:09+00:00 Add HaddockModInfo to the API - - - - - 82806848 by David Waern at 2010-02-25T21:05:27+00:00 Wibble - - - - - 744cad4c by David Waern at 2010-02-25T23:30:59+00:00 Make it possible to run a single test - - - - - 6a806e4c by David Waern at 2010-03-14T14:19:39+00:00 Bump version number - - - - - a5a8e4a7 by David Waern at 2010-03-14T14:36:35+00:00 Update ANNOUNCE - - - - - 6f05435e by Simon Hengel at 2010-03-15T20:52:42+00:00 Add missing dependencies for 'library' in haddock.cabal - - - - - faefe2bd by David Waern at 2010-03-15T22:29:37+00:00 Solve conflicts - - - - - 9808ad52 by David Waern at 2010-03-15T22:51:21+00:00 Bump version number - - - - - eb0bf60b by David Waern at 2010-03-15T22:52:32+00:00 Update CHANGES - - - - - f95cd891 by David Waern at 2010-03-15T23:01:06+00:00 Add Paths_haddock to other-modules of library - - - - - 65997b0a by David Waern at 2010-03-15T23:14:59+00:00 Update CHANGES - - - - - 7e251731 by David Waern at 2010-03-15T23:15:30+00:00 Bump version number - - - - - c9cd0ddc by David Waern at 2010-03-16T00:28:34+00:00 Fix warning - - - - - 1cac2d93 by Simon Peyton Jones at 2010-01-04T15:22:16+00:00 Fix imports for new location of splitKindFunTys - - - - - 474f26f6 by Simon Peyton Jones at 2010-02-10T14:36:06+00:00 Update Haddock for quasiquotes - - - - - 0dcc06c0 by Simon Peyton Jones at 2010-02-10T10:59:45+00:00 Track changes in HsTyVarBndr - - - - - 2d84733a by Simon Peyton Jones at 2010-02-10T14:52:44+00:00 Track HsSyn chnages - - - - - 9e3adb8b by Ian Lynagh at 2010-02-20T17:09:42+00:00 Resolve conflicts - - - - - a3e72ff8 by Simon Peyton Jones at 2010-03-04T13:05:16+00:00 Track change in HsUtils; and use a nicer function not an internal one - - - - - 27994854 by David Waern at 2010-03-18T22:22:27+00:00 Fix build with GHC 6.12.1 - - - - - 11f6e488 by David Waern at 2010-03-18T22:24:09+00:00 Bump version in test reference files - - - - - 0ef2f11b by David Waern at 2010-03-20T00:56:30+00:00 Fix library part of cabal file when in ghc tree - - - - - 3f6146ff by Mark Lentczner at 2010-03-20T22:30:11+00:00 First, experimental XHTML rendering switch to using the xhtml package copied Html.hs to Xhtml.hs and split into sub-modules under Haddock/Backends/Xhtml and detabify moved footer into div, got ready for iface change headers converted to semantic markup contents in semantic markup summary as semantic markup description in semantic markup, info block in header fixed factored out rendering so during debug it can be readable (see renderToString) - - - - - b8ab329b by Mark Lentczner at 2010-03-20T22:54:01+00:00 apply changes to Html.hs to Xhtml/*.hs incorporate changes that were made between the time Html.hs was copied and split into Xhtml.hs and Xhtml/*.hs includes patchs after "Wibble" (!) through "Fix build with GHC 6.12.1" - - - - - 73df2433 by Ian Lynagh at 2010-03-20T21:56:37+00:00 Follow LazyUniqFM->UniqFM in GHC - - - - - db4f602b by David Waern at 2010-03-29T22:00:01+00:00 Fix build with GHC 6.12 - - - - - d8dca088 by Simon Hengel at 2010-04-02T16:39:55+00:00 Add missing dependencies to cabal file - - - - - e2adc437 by Simon Hengel at 2010-04-02T14:08:40+00:00 Add markup support for interactive examples - - - - - e882ac05 by Simon Hengel at 2010-04-02T14:11:53+00:00 Add tests for interactive examples - - - - - 5a07a6d3 by David Waern at 2010-04-07T17:05:20+00:00 Propagate source positions from Lex.x to Parse.y - - - - - 6493b46f by David Waern at 2010-04-07T21:48:57+00:00 Let runtests.hs die when haddock has not been built - - - - - 5e34423e by David Waern at 2010-04-07T22:01:13+00:00 Make runtests.hs slightly more readable - - - - - 321d59b3 by David Waern at 2010-04-07T22:13:27+00:00 Fix haskell/haddock#75 Add colons to the $ident character set. - - - - - 37b08b8d by David Waern at 2010-04-08T00:32:52+00:00 Fix haskell/haddock#118 Avoid being too greedy when lexing URL markup (<..>), in order to allow multiple URLs on the same line. Do the same thing with <<..>> and #..#. - - - - - df8feac9 by David Waern at 2010-04-08T00:57:33+00:00 Make it easier to add new package deps to test suite This is a hack - we should use Cabal to get the package details instead. - - - - - 1ca6f84b by David Waern at 2010-04-08T01:03:06+00:00 Add ghc-prim to test suite deps - - - - - 27371e3a by Simon Hengel at 2010-04-08T19:26:34+00:00 Let parsing fails on paragraphs that are immediately followed by an example This is more consistent with the way we treat code blocks. - - - - - 83096e4a by David Waern at 2010-04-08T21:20:00+00:00 Improve function name - - - - - 439983ce by David Waern at 2010-04-10T10:46:14+00:00 Fix haskell/haddock#112 No link was generated for 'Addr#' in a doc comment. The reason was simply that the identifier didn't parse. We were using parseIdentifier from the GHC API, with a parser state built from 'defaultDynFlags'. If we pass the dynflags of the module instead, the right options are turned on on while parsing the identifer (in this case -XMagicHash), and the parse succeeds. - - - - - 5c0d35d7 by David Waern at 2010-04-10T10:54:06+00:00 Rename startGhc into withGhc - - - - - dca081fa by Simon Hengel at 2010-04-12T19:09:16+00:00 Add documentation for interactive examples - - - - - c7f26bfa by David Waern at 2010-04-13T00:51:51+00:00 Slight fix to the documentation of examples - - - - - 06eb7c4c by David Waern at 2010-04-13T00:57:05+00:00 Rename Interactive Examples into Examples (and simplify explanation) - - - - - 264830cb by David Waern at 2010-05-10T20:07:27+00:00 Update CHANGES with info about 2.6.1 - - - - - 8e5d4514 by Simon Hengel at 2010-04-18T18:16:54+00:00 Add unit tests for parser - - - - - 68297f40 by David Waern at 2010-05-10T21:53:37+00:00 Improve testsuite README - - - - - f04eb6e4 by David Waern at 2010-05-11T19:14:31+00:00 Re-organise the testsuite structure - - - - - a360f710 by David Waern at 2010-05-11T19:18:03+00:00 Shorten function name - - - - - 1d5dd359 by David Waern at 2010-05-11T21:40:02+00:00 Update runtests.hs following testsuite re-organisation - - - - - ffebe217 by David Waern at 2010-05-11T21:40:10+00:00 Update runtests.hs to use base-4.2.0.1 - - - - - 635de402 by David Waern at 2010-05-11T21:41:11+00:00 Update runparsetests.hs following testsuite reorganisation - - - - - 72137910 by Ian Lynagh at 2010-05-06T20:43:06+00:00 Fix build - - - - - 1a80b76e by Ian Lynagh at 2010-05-06T22:25:29+00:00 Remove redundant import - - - - - 1031a80c by Simon Peyton Jones at 2010-05-07T13:21:09+00:00 Minor wibbles to HsBang stuff - - - - - dd8e7fe5 by Ian Lynagh at 2010-05-08T15:22:00+00:00 GHC build system: Follow "rm" variable changes - - - - - 7f5e6748 by David Waern at 2010-05-13T11:53:02+00:00 Fix build with GHC 6.12.2 - - - - - 7953d4d8 by David Waern at 2010-05-13T18:45:01+00:00 Fixes to comments only - - - - - 8ae8eb64 by David Waern at 2010-05-13T18:57:26+00:00 ModuleMap -> IfaceMap - - - - - 1c3eadc6 by David Waern at 2010-05-13T19:03:13+00:00 Fix whitespace style issues - - - - - e96783c0 by David Waern at 2010-05-13T19:08:53+00:00 Fix comment - - - - - c998a78b by David Waern at 2010-05-13T19:39:00+00:00 Position the module header the same way everywhere Silly, but nice with some consistency :-) - - - - - b48a714e by David Waern at 2010-05-13T19:41:32+00:00 Position of module header, this time in the HTML backends - - - - - f9bfb12e by David Waern at 2010-05-13T19:43:05+00:00 Two newlines between declarations in Main - - - - - 071d44c7 by David Waern at 2010-05-13T19:44:21+00:00 Newlines in Convert - - - - - 036346db by David Waern at 2010-05-13T19:46:47+00:00 Fix a few stylistic issues in H.InterfaceFile - - - - - f0b8379e by David Waern at 2010-05-13T19:47:53+00:00 Add newlines to H.ModuleTree - - - - - 27409f8e by David Waern at 2010-05-13T19:51:10+00:00 Fix stylistic issues in H.Utils - - - - - 24774a11 by David Waern at 2010-05-13T20:00:43+00:00 Structure H.Types better - - - - - 7b6f5e40 by David Waern at 2010-05-13T20:01:04+00:00 Remove bad Arbitrary instance - - - - - fac9f1f6 by David Waern at 2010-05-13T20:05:50+00:00 Get rid of H.Utils.pathJoin and use System.FilePath.joinPath instead - - - - - fe6d00c4 by David Waern at 2010-05-13T20:51:55+00:00 Export a couple of more types from the API - - - - - b2e33a5f by David Waern at 2010-05-13T21:27:51+00:00 Improve doc comment for Interface - - - - - c585f2ce by David Waern at 2010-05-13T21:30:14+00:00 Improve documentation of Haddock.Interface - - - - - e6791db2 by David Waern at 2010-05-13T22:07:35+00:00 Remove meaningless comments - - - - - 7801b390 by David Waern at 2010-05-14T17:53:33+00:00 Remove unused modules - - - - - f813e937 by David Waern at 2010-05-14T17:55:17+00:00 Re-direct compilation output to a temporary directory Also add a flag --no-tmp-comp-dir that can be used to get the old behaviour of writing compilation files to GHC's output directory (default "."). - - - - - e56737ec by David Waern at 2010-05-14T18:06:11+00:00 Wibble - - - - - e40b0447 by David Waern at 2010-05-14T19:01:52+00:00 Move flag evaluation code from Main to Haddock.Options Determining the value of "singular" flags (by e.g. taking the last occurrence of the flag) and other flag evaluation should done in Haddock.Options which is the module that is supposed to define the command line interface. This makes Main a bit easier on the eyes as well. - - - - - 27091f57 by David Waern at 2010-05-14T19:05:10+00:00 Wibble - - - - - c658cf61 by David Waern at 2010-05-14T19:06:49+00:00 Re-order things in Haddock.Options a bit - - - - - 8cfdd342 by David Waern at 2010-05-14T19:20:29+00:00 De-tabify Haddock.Options and fix other whitespace issues - - - - - 0df16b62 by David Waern at 2010-05-14T19:25:07+00:00 Improve comments - - - - - 80b38e2b by David Waern at 2010-05-14T19:26:42+00:00 Whitespace police - - - - - fe580255 by David Waern at 2010-05-14T19:31:23+00:00 Wibbles to comments - - - - - a2b43fad by David Waern at 2010-05-14T20:24:32+00:00 Move some more flag functions to Haddock.Options - - - - - 3f895547 by David Waern at 2010-05-14T20:37:12+00:00 Make renderStep a top-level function in Main - - - - - 5cdca11d by David Waern at 2010-05-14T20:39:27+00:00 Spelling in comment - - - - - ad98d14c by David Waern at 2010-05-14T20:40:26+00:00 Comment fixes - - - - - 0bb9218f by David Waern at 2010-05-14T20:49:01+00:00 Whitespace police - - - - - 0f0a533f by David Waern at 2010-05-15T16:42:29+00:00 Improve description of --dump-interface - - - - - 5b2833ac by David Waern at 2010-05-15T17:16:53+00:00 Document --no-tmp-comp-dir - - - - - 8160b170 by David Waern at 2010-05-15T17:18:59+00:00 Wibble - - - - - 570dbe33 by David Waern at 2010-05-18T21:15:38+00:00 HLint police - - - - - 204e425f by David Waern at 2010-05-18T21:16:30+00:00 HLint police - - - - - 6db657ac by David Waern at 2010-05-18T21:16:37+00:00 Wibble - - - - - b942ccd7 by Simon Marlow at 2010-06-02T08:27:30+00:00 Interrupted disappeared in GHC 6.13 (GHC ticket haskell/haddock#4100) - - - - - 3b94a819 by Simon Marlow at 2010-06-02T08:45:08+00:00 Allow base-4.3 - - - - - c5a1fb7c by Simon Marlow at 2010-06-02T09:03:04+00:00 Fix compilation with GHC 6.13 - - - - - 6181296c by David Waern at 2010-06-08T21:09:05+00:00 Display name of prologue file when parsing it fails - - - - - 7cbc6f60 by Ian Lynagh at 2010-06-13T16:20:25+00:00 Remove redundant imports - - - - - 980c804b by Simon Marlow at 2010-06-22T08:41:50+00:00 isLocalAndTypeInferenced: fix for local module names overlapping package modules - - - - - d74d4a12 by Simon Marlow at 2010-06-23T12:03:27+00:00 Unresolved identifiers in Doc get replaced with DocMonospaced rather than plain strings - - - - - d8546783 by Simon Marlow at 2010-06-30T12:45:17+00:00 LaTeX backend (new options: --latex, --latex-style=<style>) - - - - - 437afa9e by David Waern at 2010-07-01T12:02:44+00:00 Fix a few stylistic whitespace issues in LaTeX backend - - - - - 85bc1fae by David Waern at 2010-07-01T15:42:45+00:00 Make runtest.hs work with GHC 6.12.3 (we should really stop hard coding this) - - - - - 7d2eb86f by David Waern at 2010-07-01T15:43:33+00:00 Update test following Simon's patch to render unresolved names in monospaced font - - - - - 08fcbcd2 by David Waern at 2010-07-01T16:12:18+00:00 Warning police - - - - - d04a8d7a by David Waern at 2010-07-04T14:53:39+00:00 Fix a bug in attachInstances We didn't look for instance docs in all the interfaces of the package. This had the effect of instance docs not always showing up under a declaration. I took the opportunity to clean up the code in H.I.AttachInstances a bit as well. More cleanup is needed, however. - - - - - d10344eb by Simon Hengel at 2010-07-10T09:19:04+00:00 Add missing dependencies to cabal file - - - - - 24090531 by Mark Lentczner at 2010-03-21T04:51:16+00:00 add exports to Xhtml modules - - - - - 84f9a333 by Mark Lentczner at 2010-04-03T19:14:22+00:00 clean up Doc formatting code - add CSS for lists - renderToString now uses showHtml since prettyHtml messes up <pre> sections - - - - - bebccf52 by Mark Lentczner at 2010-04-04T04:51:08+00:00 tweak list css - - - - - 0c2aeb5e by Mark Lentczner at 2010-04-04T06:24:14+00:00 all decls now generate Html not HtmlTable - ppDecl return Html, and so now do all of the functions it calls - added some internal tables to some decls, which is wrong, and will have to be fixed - decl "Box" functions became "Elem" functions to make clear they aren't in a table anymore (see Layout.hs) - docBox went away, as only used in one place (and its days are numbered) - cleaned up logic in a number of places, removed dead code - added maybeDocToHtml which simplified a number of places in the code - - - - - dbf73e6e by Mark Lentczner at 2010-04-05T05:02:43+00:00 clean up processExport and place a div around each decl - - - - - e25b7e9f by Mark Lentczner at 2010-04-10T21:23:21+00:00 data decls are now a sequence of paragraphs, not a table - - - - - 89ee0294 by Mark Lentczner at 2010-04-10T21:29:16+00:00 removed commented out code that can't be maintained - - - - - d466f536 by Mark Lentczner at 2010-04-12T04:56:27+00:00 removed declWithDoc and cleaned up data decls in summary - - - - - ed755832 by Mark Lentczner at 2010-04-12T05:07:53+00:00 merge in markupExample changes - - - - - c36f51fd by Mark Lentczner at 2010-04-25T04:56:37+00:00 made record fields be an unordList, not a table - - - - - ed3a28d6 by Mark Lentczner at 2010-04-25T05:23:28+00:00 fixed surround of instance and constructor tables - - - - - 0e35bbc4 by Mark Lentczner at 2010-04-25T05:36:59+00:00 fix class member boxes in summary - - - - - 5041749b by Mark Lentczner at 2010-04-25T05:38:35+00:00 remove unused bodyBox - - - - - e91724db by Mark Lentczner at 2010-04-25T06:26:10+00:00 fixed javascript quoting/escpaing issue - - - - - f4abbb73 by Mark Lentczner at 2010-05-03T23:04:31+00:00 adjust css for current markup - - - - - e75fec4c by Mark Lentczner at 2010-05-04T06:14:34+00:00 added assoicated types and methods back into class decls - - - - - 84169323 by Mark Lentczner at 2010-05-24T13:13:42+00:00 merge in changes from the big-whitespace cleanup - - - - - 3c1c872e by Mark Lentczner at 2010-06-11T21:03:58+00:00 adjust synopsis and bottom bar spacing - - - - - 3c1f9ef7 by Mark Lentczner at 2010-06-11T21:14:44+00:00 fix missing space in "module" lines in synoposis - - - - - 9a137e6d by Mark Lentczner at 2010-06-11T21:34:08+00:00 changed tt elements to code elements - - - - - 50f71ef1 by Mark Lentczner at 2010-06-11T23:27:46+00:00 factored out ppInstances - - - - - 3b9a9de5 by Mark Lentczner at 2010-06-17T17:36:01+00:00 push single constructors (newtype) onto line with decl - - - - - e0f8f2ec by Mark Lentczner at 2010-06-17T22:20:56+00:00 remove <++> connector - - - - - 56c075dd by Mark Lentczner at 2010-07-13T05:26:21+00:00 change to new page structure - - - - - 04be6ca7 by Mark Lentczner at 2010-07-14T04:21:55+00:00 constructors and args as dl lists, built in Layout.hs - - - - - 65aeafc2 by Mark Lentczner at 2010-07-14T05:38:32+00:00 better interface to subDecls - - - - - 72032189 by Mark Lentczner at 2010-07-14T07:04:10+00:00 made subDecl tables looks just so - - - - - b782eca2 by Mark Lentczner at 2010-07-14T16:00:54+00:00 convert args to SubDecl format - - - - - cc75e98f by Mark Lentczner at 2010-07-14T16:28:53+00:00 convert instances to SubDecl - - - - - 34e2aa5a by Mark Lentczner at 2010-07-14T21:07:32+00:00 removing old table cruft from Layout.hs - - - - - d5810d95 by Mark Lentczner at 2010-07-14T21:54:58+00:00 methods and associated types in new layout scheme - - - - - 65ef9579 by Mark Lentczner at 2010-07-14T23:43:42+00:00 clean up synopsis lists - - - - - e523318f by Mark Lentczner at 2010-07-15T05:02:26+00:00 clean up of anchors - - - - - 1215dfc5 by Mark Lentczner at 2010-07-15T23:53:01+00:00 added two new themes and rough css switcher - - - - - 7f0fd36f by Mark Lentczner at 2010-07-16T04:57:38+00:00 fixed package catpion, added style menu - - - - - 0dd4999c by Mark Lentczner at 2010-07-16T20:12:39+00:00 new output for mini_ pages - - - - - 64b2810b by Mark Lentczner at 2010-07-16T20:58:41+00:00 reformat index-frames - - - - - 3173f555 by Mark Lentczner at 2010-07-16T22:41:53+00:00 convert index to new markup - - - - - b0a4b7c9 by Mark Lentczner at 2010-07-17T04:07:22+00:00 convert index.html to new markup, adjust module markup - - - - - 8261ae1e by Mark Lentczner at 2010-07-17T05:07:29+00:00 classing styling of ancillary pages - - - - - 2a4fb025 by Mark Lentczner at 2010-07-17T05:11:45+00:00 clean up Layout.hs: no more vanillaTable - - - - - 87eec685 by Mark Lentczner at 2010-07-17T05:35:16+00:00 clean up Util.hs - - - - - d304e9b0 by Mark Lentczner at 2010-07-17T05:38:50+00:00 qualify import of XHtml as XHtml - - - - - 7dc05807 by Mark Lentczner at 2010-07-17T06:17:53+00:00 factored out head element generation - - - - - 9cdaec9e by Mark Lentczner at 2010-07-17T06:44:54+00:00 refactored out main page body generation - - - - - 8a51019e by Mark Lentczner at 2010-07-17T06:48:20+00:00 moved footer into only place that used it - - - - - efa479da by Mark Lentczner at 2010-07-17T18:48:30+00:00 styling auxillary pages for tibbe and snappy themes - - - - - 81de5509 by Mark Lentczner at 2010-07-18T04:41:38+00:00 fixed alphabet on index page, and styling of it and packages in module lists - - - - - 20718c1a by Mark Lentczner at 2010-07-18T05:34:29+00:00 cleaned up div functions in Layout.hs - - - - - 60d50453 by Mark Lentczner at 2010-07-18T05:48:39+00:00 added content div to main pages - - - - - ed16561c by Mark Lentczner at 2010-07-18T06:12:22+00:00 add .doc class to documentation blocks - - - - - f5c781b0 by Mark Lentczner at 2010-07-19T05:20:53+00:00 refactoring of anchor ID and fragment handling - - - - - a69a93bf by Mark Lentczner at 2010-07-19T05:35:55+00:00 remove an explicit bold tag - replace with .def class - - - - - d76c7225 by Mark Lentczner at 2010-07-19T06:56:15+00:00 rename Haddock.Backends.Xhtml.Util to Utils - - - - - 5a58c0da by David Waern at 2010-07-21T13:30:54+00:00 Remove trailing whitespace in Haddock.Backends.Xhtml - - - - - 0652aa17 by David Waern at 2010-07-21T13:33:21+00:00 Align a few comments - - - - - 785776c3 by David Waern at 2010-07-21T13:39:04+00:00 Remove trailing whitespace in H.B.X.Decl - - - - - 71a30710 by David Waern at 2010-07-21T13:44:27+00:00 Remove more trailing whitespace - - - - - 38750394 by David Waern at 2010-07-21T13:50:43+00:00 Style police - - - - - 3023d940 by David Waern at 2010-07-21T14:01:22+00:00 Style police in H.B.X.Decl - - - - - df16e9e6 by David Waern at 2010-07-21T14:14:45+00:00 Style police in H.B.X.DocMarkup - - - - - 6020e321 by David Waern at 2010-07-21T14:17:32+00:00 More style police - - - - - 86ad8bf5 by David Waern at 2010-07-21T14:21:02+00:00 Style police in H.B.Xhtml - - - - - aea27d03 by David Waern at 2010-07-21T14:42:03+00:00 Fix warnings in LaTeX backend - - - - - 2aff34a9 by David Waern at 2010-07-21T14:50:46+00:00 Style police in LaTeX backend (mainly more newlines) - - - - - e517162d by David Waern at 2010-07-21T15:05:47+00:00 Doc sections in Main - - - - - b971aa0c by David Waern at 2010-07-21T15:06:17+00:00 Trailing whitespace in Documentation.Haddock - - - - - f11628fb by David Waern at 2010-07-21T15:07:06+00:00 Trailing whitespace in Haddock.Convert - - - - - cbaf284c by David Waern at 2010-07-21T15:08:11+00:00 Style police in Haddock.GhcUtils - - - - - 71feb77b by David Waern at 2010-07-21T15:09:06+00:00 Style police in Haddock.InterfaceFile - - - - - 0a9c80e6 by David Waern at 2010-07-21T15:11:33+00:00 Whitespace police - - - - - 6168376c by David Waern at 2010-07-21T15:16:35+00:00 Style police in Haddock.Utils - - - - - 9fe4dd90 by David Waern at 2010-07-21T15:19:31+00:00 Add -fwarn-tabs - - - - - a000d752 by Mark Lentczner at 2010-07-20T17:25:52+00:00 move CSS Theme functions into Themes.hs - - - - - b52b440f by Mark Lentczner at 2010-07-20T17:29:35+00:00 add Thomas Schilling's theme - - - - - e43fa7e8 by Mark Lentczner at 2010-07-21T04:49:34+00:00 correct icon used with Snappy theme - - - - - ba5092d3 by Mark Lentczner at 2010-07-21T04:56:47+00:00 apply Tibbe's updates to his theme - - - - - 7804eef6 by Mark Lentczner at 2010-07-21T05:15:49+00:00 space between "Style" and the downward triangle - - - - - 7131d4c6 by Mark Lentczner at 2010-07-21T17:43:35+00:00 merge with David's source cleanups - - - - - ee65f1cb by David Waern at 2010-07-22T16:50:46+00:00 Fix a bug where we allowed --hoogle, --latex, etc without input files - - - - - e413ff7a by David Waern at 2010-07-22T17:21:58+00:00 Improve function name - - - - - a0fd14f3 by Simon Marlow at 2010-06-30T15:34:32+00:00 fix warnings - - - - - 31f73d2a by David Waern at 2010-07-22T19:29:41+00:00 Solve conflicts - - - - - d563b4a5 by Simon Marlow at 2010-06-30T15:34:37+00:00 fix warning - - - - - 412b6469 by David Waern at 2010-07-22T19:31:28+00:00 Solve conflict - - - - - 35174b94 by Ian Lynagh at 2010-07-06T17:27:16+00:00 Follow mkPState argument order change - - - - - b5c3585c by Simon Marlow at 2010-07-14T08:49:21+00:00 common up code for instance rendering - - - - - d8009560 by Simon Marlow at 2010-07-14T12:37:11+00:00 fix warnings - - - - - a6d88695 by David Waern at 2010-07-24T15:33:33+00:00 Fix build with ghc < 6.13 - - - - - 94cf9de1 by David Waern at 2010-07-24T15:34:37+00:00 Remove conflict left-over - - - - - 313b15c0 by Mark Lentczner at 2010-07-21T22:09:04+00:00 reorganization of nhaddock.css with tibbe - - - - - 9defed80 by Mark Lentczner at 2010-07-21T22:42:14+00:00 further cleanup of nhaddock.css, float TOC, support aux. pages - - - - - 6d944c1b by Mark Lentczner at 2010-07-22T06:22:23+00:00 remove old HTML backend - - - - - b3e8cba5 by Mark Lentczner at 2010-07-22T06:43:32+00:00 remove --html-help support - it was old, out-of-date, and mostly missing - - - - - d2654a08 by Mark Lentczner at 2010-07-22T21:45:34+00:00 tweaks to nhaddock.css - - - - - f73b285c by Mark Lentczner at 2010-07-23T06:19:35+00:00 command like processing for theme selection The bulk of the change is threadnig the selected theme set through functions in Xhtml.hs so that the selected themes can be used when generating the page output. There isn't much going on in most of these changes, just passing it along. The real work is all done in Themes.hs. - - - - - 8bddc90d by Mark Lentczner at 2010-07-23T06:58:31+00:00 drop --themes support, add named theme support decided that --themes was silly - no one would do that, just use multiple --theme arguments made --theme a synonym for --css and -c made those arguments, if no file is found, look up the argument as the name of a built in theme all of this let's haddock be invoked with "--theme=classic" for example. - - - - - 20cafd4f by Mark Lentczner at 2010-07-23T17:44:29+00:00 rename --default-themes to --built-in-themes - - - - - 0fe41307 by Mark Lentczner at 2010-07-23T18:33:02+00:00 tweaks to theme for info table, headings, and tables - - - - - cba4fee0 by Mark Lentczner at 2010-07-23T19:13:59+00:00 tweaks for dl layout, though still not used - - - - - 463fa294 by Mark Lentczner at 2010-07-23T21:07:19+00:00 tweak look of mini pages, keywords, and preblocks - - - - - 5472fc02 by Mark Lentczner at 2010-07-24T05:36:15+00:00 slide out Synopsis drawer - - - - - 9d5d5de5 by Mark Lentczner at 2010-07-24T06:02:42+00:00 extend package header and footer to edges of page - - - - - a47c91a2 by Mark Lentczner at 2010-07-24T06:28:44+00:00 fields are def lists, tweak css for style menu, mini pages, arguments - - - - - ca20f23b by Mark Lentczner at 2010-07-24T16:55:22+00:00 excisting last vestiges of the --xhtml flag - - - - - 71fb012e by Mark Lentczner at 2010-07-25T18:47:49+00:00 change how collapsing sections are done make whole .caption be the target improve javascript for class toggling have plus/minus images come from .css, not img tags - - - - - c168c8d3 by Mark Lentczner at 2010-07-26T00:32:05+00:00 reorganize files in the html lib data dir - - - - - 93324301 by Mark Lentczner at 2010-07-26T01:27:42+00:00 cleaned up Themes.hs - - - - - ad3b5dd4 by Mark Lentczner at 2010-07-26T02:39:15+00:00 make module list use new collapsers - - - - - 1df9bfc6 by Mark Lentczner at 2010-07-27T19:09:25+00:00 remove Tibbe theme - - - - - 8b9b01b3 by Mark Lentczner at 2010-07-27T20:04:03+00:00 move themes into html dir with .theme and .std-theme extensions - - - - - a7beb965 by Mark Lentczner at 2010-07-27T21:06:34+00:00 give a class to empty dd elements so they can be hidden - - - - - a258c117 by Mark Lentczner at 2010-07-27T21:23:58+00:00 remove custom version of copyFile in Xhtml.hs - - - - - b70dba6e by Mark Lentczner at 2010-07-27T22:12:45+00:00 apply margin changes to pre and headings as per group decision, and small cleanups - - - - - e6f722a2 by Mark Lentczner at 2010-07-28T00:03:12+00:00 make info block and package bar links be floatable by placing them first in the dom tree - - - - - c8278867 by Mark Lentczner at 2010-07-28T19:01:18+00:00 styling source links on declarations - - - - - 88fdc399 by Mark Lentczner at 2010-07-29T01:12:46+00:00 styling tweaks don't generate an empty li for absent style menu in links area update css for Classic and Snappy to handle: dl lists links in package header and in declarations floating of links and info block in package and module headers - - - - - 8a75b213 by Ian Lynagh at 2010-07-30T20:21:46+00:00 Fix build in GHC tree - - - - - ce8e18b3 by Simon Hengel at 2010-08-03T18:37:26+00:00 Adapt paths to data files in cabal file - - - - - 9701a455 by Simon Hengel at 2010-08-07T13:20:27+00:00 Add missing dependency to cabal file - - - - - 01b838d1 by Mark Lentczner at 2010-07-30T20:19:40+00:00 improved synopsis drawer: on click, not hover - - - - - 7b6f3e59 by Mark Lentczner at 2010-07-30T23:38:55+00:00 put the synopsis back in the other themes - - - - - 7b2904c9 by Mark Lentczner at 2010-08-11T11:11:26+00:00 close arrows on expanded synopsis drawer - - - - - ea19e177 by Mark Lentczner at 2010-08-12T21:16:45+00:00 width and font changes removed the max width restrictions on the page as a whole and the synopsis made the main font size smaller (nominally 14pt) and then tweaked most font sizes (relative) to be more consistent - - - - - 5ced00c0 by Mark Lentczner at 2010-08-13T15:09:55+00:00 implemented YUI's CSS font approach - - - - - 2799c548 by Mark Lentczner at 2010-08-13T15:11:59+00:00 adjusted margin to 2em, 1 wasn't enough - - - - - 58f06893 by Mark Lentczner at 2010-08-13T15:48:44+00:00 removed underlining on hover for named anchors headings in interface lost thier a element, no need, just put id on heading css for a elements now only applies to those with href attribute - - - - - 7aced4c4 by Mark Lentczner at 2010-08-13T15:50:22+00:00 more space between elements - - - - - 5a3c1cce by Mark Lentczner at 2010-08-13T16:43:43+00:00 adjusted font sizes of auxilary pages per new scheme - - - - - 487539ef by Mark Lentczner at 2010-08-13T21:43:41+00:00 add Frames button and clean up frames.html - - - - - c1a140b6 by Mark Lentczner at 2010-08-13T22:17:48+00:00 move frames button to js - - - - - b0bdb68e by Mark Lentczner at 2010-08-14T03:44:46+00:00 build style menu in javascript moved to javascript, so as to not polute the content with the style menu removed menu building code in Themes.hs removed onclick in Utils.hs changed text of button in header from "Source code" to "Source" more consistent with links in rest of page - - - - - 43ab7120 by Mark Lentczner at 2010-08-16T15:15:37+00:00 font size and margin tweaks - - - - - c0b68652 by Mark Lentczner at 2010-08-17T18:19:52+00:00 clean up collapser logics javascript code for collapasble sections cleaned up rewrote class utilities in javascript to be more robust refactored utilities for generating collapsable sections made toc be same color as synopsis module list has needed clear attribute in CSS - - - - - 5d573427 by Mark Lentczner at 2010-08-17T23:06:02+00:00 don't collapse entries in module list when clicking on links - - - - - 8c307c4a by Mark Lentczner at 2010-08-17T23:21:43+00:00 add missing data file to .cabal - - - - - 414bcfcf by Mark Lentczner at 2010-08-17T23:28:47+00:00 remove synopsis when in frames - - - - - ba0fa98a by Mark Lentczner at 2010-08-18T16:16:11+00:00 layout tweeks - mini page font size, toc color, etc. - - - - - 63c1bed1 by Mark Lentczner at 2010-08-18T19:50:02+00:00 margin fiddling - - - - - c311c094 by Mark Lentczner at 2010-08-20T01:37:55+00:00 better synopsis handling logic - no flashing - - - - - f1fe5fa8 by Mark Lentczner at 2010-08-20T01:41:06+00:00 fix small layout issues mini frames should have same size top heading give info block dts some padding so they don't collide in some browsers - - - - - 0de84d77 by Mark Lentczner at 2010-08-20T02:13:09+00:00 made style changing and cookies storage robust - - - - - 1ef064f9 by Thomas Schilling at 2010-08-04T13:12:22+00:00 Make synopsis frame behave properly in Firefox. In Firefox, pressing the back button first reverted the synopsis frame, and only clicking the back button a second time would update the main frame. - - - - - dd1c9a94 by Mark Lentczner at 2010-08-21T01:46:19+00:00 remove Snappy theme - - - - - 2353a90d by Mark Lentczner at 2010-08-25T05:16:19+00:00 fix occasional v.scroll bars on pre blocks (I think) - - - - - 459b8bf1 by Simon Hengel at 2010-08-08T10:12:45+00:00 Add createInterfaces' (a more high-level alternative to createInterfaces) to Haddock API - - - - - b1b68675 by David Waern at 2010-08-26T20:31:58+00:00 Follow recent API additions with some refactorings Simon Hegel's patch prompted me to do some refactorings in Main, Haddock.Documentation and Haddock.Interface. - - - - - 264d4d67 by David Waern at 2010-08-26T21:40:59+00:00 Get rid of GhcModule and related cruft We can get everything we need directly from TypecheckedModule. - - - - - 0feacec2 by Mark Lentczner at 2010-08-26T23:44:13+00:00 fixed CSS for ordered lists and def lists in doc blocks - - - - - 2997e0c2 by Mark Lentczner at 2010-08-26T23:45:03+00:00 support both kinds of enumerated lists in doc markup The documentation for Haddock says enumerated lists can use either of (1) first item 2. second item The second form wasn't actually supported - - - - - 5d4ddeec by Mark Lentczner at 2010-08-27T21:29:48+00:00 fix broken header link margins - - - - - 614456ba by Mark Lentczner at 2010-08-27T22:16:19+00:00 fix table of contents CSS - - - - - 03f329a2 by David Waern at 2010-08-28T16:36:09+00:00 Update tests following switch to the Xhtml backend - - - - - ca689fa2 by Mark Lentczner at 2010-08-28T18:25:16+00:00 fix def lists - - - - - 18e1d3d2 by Mark Lentczner at 2010-08-28T18:26:18+00:00 push footer to bottom of window - - - - - b0ab8d82 by David Waern at 2010-08-28T22:04:32+00:00 Whitespace police - - - - - 2d217977 by David Waern at 2010-08-29T12:44:45+00:00 Remove Snappy data files - - - - - 01e27d5f by David Waern at 2010-08-29T13:03:28+00:00 Add source entity path to --read-interface You can now use this flag like this: --read-interface=<html path>,<source entity path>,<.haddock file> By "source entity path" I mean the same thing that is specified with the --source-entity flag. The purpose of this is to be able to specify the source entity path per package, to allow source links to work in the presence of cross-package documentation. When given two arguments or less the --read-interface flag behaves as before. - - - - - 20bf4aaa by David Waern at 2010-08-29T13:11:03+00:00 Naming wibbles - - - - - ad22463f by Mark Lentczner at 2010-08-29T15:14:54+00:00 make portability block be a table - solves layout issues - - - - - 97bd1ae6 by Mark Lentczner at 2010-08-29T15:17:42+00:00 update golden test for Test due to portability box change - - - - - d37e139e by Mark Lentczner at 2010-08-29T17:07:17+00:00 move TOC and Info blocks down 0.5em to improve layout issue w/Test.hs - - - - - acf52501 by David Waern at 2010-08-29T17:32:36+00:00 Allow building with ghc < 6.16 - - - - - 1cb34ed8 by Ian Lynagh at 2010-07-24T23:18:49+00:00 Flatten the dynflags before parsing - - - - - b36845b4 by Ian Lynagh at 2010-07-24T23:26:49+00:00 Follow flattenLanguageFlags -> flattenExtensionFlags rename - - - - - 7f7fcc7e by David Waern at 2010-08-29T17:46:23+00:00 Use flattenExtensionFlags with ghc >= 6.13 only - - - - - 13cf9411 by Ian Lynagh at 2010-08-01T18:09:54+00:00 Make the main haddock script versioned, and make plain "haddock" a symlink - - - - - 495cbff2 by Ian Lynagh at 2010-08-18T18:57:24+00:00 Fix installation in the GHC build system Data-files are now in subdirectories, so we need to handle that - - - - - 88ebab0a by Ian Lynagh at 2010-08-18T19:43:53+00:00 GHC build system: Add all the data files to BINDIST_EXTRAS - - - - - 65837172 by David Waern at 2010-08-29T20:12:34+00:00 Update Test - - - - - 094bbaa2 by David Waern at 2010-08-29T20:55:14+00:00 Revert update to Test - - - - - a881cfb3 by David Waern at 2010-08-31T18:24:15+00:00 Bump version number - - - - - 1fc8a3eb by David Waern at 2010-08-31T22:32:27+00:00 Update ANNOUNCE - - - - - ee1df9d0 by David Waern at 2010-08-31T22:33:11+00:00 Update CHANGES - - - - - 394cc854 by David Waern at 2010-08-31T22:33:23+00:00 Update interface file versioning to work with ghc 6.14/15 - - - - - 7d03b79b by David Waern at 2010-08-31T22:36:00+00:00 Update test output following version change - - - - - a48d82d1 by Mark Lentczner at 2010-09-01T04:29:35+00:00 sort options in doc to match --help output removed --html-help option, as it is no longer supported - - - - - 06561aeb by Mark Lentczner at 2010-09-01T05:29:32+00:00 update options documentation rewrote doc for --html added doc for --theme and --built-in-themes added --use-contents and --gen-contents - - - - - 57dea832 by Mark Lentczner at 2010-09-01T05:31:27+00:00 slight wording change about Frames mode - - - - - fa1f6da3 by David Waern at 2010-09-01T10:57:44+00:00 Update doc configure script to find docbook stylesheets on arch linux - - - - - addff770 by David Waern at 2010-09-01T11:02:29+00:00 Wibble - - - - - 8399006d by David Waern at 2010-09-01T11:19:21+00:00 Replace ghci> with >>> in example syntax - - - - - 35074cf8 by David Waern at 2010-09-01T19:03:27+00:00 Improve docs for --no-tmp-comp-dir - - - - - 0f8f8cfd by David Waern at 2010-09-02T11:22:27+00:00 Add a list of contributors to the user guide Break out everyone thanked in the `Acknowledgements` chapter into a separate contributor list and add everyone from `darcs show authors`. We consider everyone who is thanked to be a contributor as a conservative estimation :-) I have added some more contributors that I know about, who were not in the darcs history, but others may be missing. So please add anyone that you think is missing from the list. - - - - - 42ccf099 by David Waern at 2010-09-02T11:29:22+00:00 Update copyright years in license - - - - - 0d560479 by David Waern at 2010-09-02T11:38:52+00:00 Update release instructions - - - - - 72ab7796 by David Waern at 2010-09-02T19:27:08+00:00 Add a note to ANNOUNCE - - - - - bf9d9c5d by David Waern at 2010-09-02T19:27:48+00:00 H.Utils needs FFI on Win+MinGW - - - - - 048ae44a by Mark Lentczner at 2010-09-04T23:19:47+00:00 make TOC group header identifiers validate - - - - - 8c6faf36 by Simon Michael at 2010-09-22T07:12:34+00:00 add hints for cleaner darcs show authors output - - - - - 9909bd17 by Simon Michael at 2010-09-22T17:58:06+00:00 print haddock coverage info on stdout when generating docs A module's haddockable items are its exports and the module itself. The output is lightly formatted so you can align the :'s and sort for readability. - - - - - 6da72171 by David Waern at 2010-10-03T21:31:24+00:00 Style wibble - - - - - 2f8d8e4d by Tobias Brandt at 2010-08-27T07:01:21+00:00 adding the option to fully qualify identifiers - - - - - 833be6c6 by Tobias Brandt at 2010-08-27T15:50:28+00:00 adding support for local and relative name qualification - - - - - df15c4e9 by Tobias Brandt at 2010-08-27T15:56:37+00:00 corrected qualification help message - - - - - 449e9ce1 by David Waern at 2010-10-16T17:34:30+00:00 Solve conflicts - - - - - 3469bda5 by David Waern at 2010-10-16T18:42:40+00:00 Use "qual" as an abbreviation for qualification instead of "quali" for consistency - - - - - 97c2d728 by David Waern at 2010-10-16T18:47:07+00:00 Style police - - - - - ce14fbea by David Waern at 2010-10-16T21:15:25+00:00 Style police - - - - - fdf29e9d by David Waern at 2010-10-17T00:30:44+00:00 Add a pointer to the style guide - - - - - 8e6b44e8 by rrnewton at 2010-10-24T03:19:28+00:00 Change to index pages: include an 'All' option even when subdividing A-Z. - - - - - 755b131c by David Waern at 2010-11-14T19:39:36+00:00 Bump version - - - - - d0345a04 by David Waern at 2010-11-14T19:41:59+00:00 TAG 2.8.1 - - - - - f6221508 by Simon Peyton Jones at 2010-09-13T09:53:00+00:00 Adapt to minor changes in internal GHC functions - - - - - 1290713d by Ian Lynagh at 2010-09-15T10:37:18+00:00 Remove duplicate Outputable instance for Data.Map.Map - - - - - 87f69eef by Ian Lynagh at 2010-09-21T15:01:10+00:00 Bump GHC dep upper bound - - - - - af36e087 by Ian Lynagh at 2010-09-21T15:12:02+00:00 Fix up __GLASGOW_HASKELL__ tests - - - - - ad67716c by Ian Lynagh at 2010-09-21T20:31:35+00:00 Don't build haddock is HADDOCK_DOCS is NO - - - - - 63b3f1f5 by Ian Lynagh at 2010-09-21T21:39:51+00:00 Fixes for when HADDOCK_DOCS=NO - - - - - e92bfa42 by Ian Lynagh at 2010-09-29T21:15:38+00:00 Fix URL creation on Windows: Use / not \ in URLs. Fixes haskell/haddock#4353 - - - - - 66c55e05 by Ian Lynagh at 2010-09-30T17:03:34+00:00 Tidy up haddock symlink installation In particular, it now doesn't get created if we aren't installing haddock. - - - - - 549b5556 by Ian Lynagh at 2010-10-23T21:17:14+00:00 Follow extension-flattening change in GHC - - - - - d7c2f72b by David Waern at 2010-11-14T20:17:55+00:00 Bump version to 2.8.2 - - - - - 6989a3a9 by David Waern at 2010-11-14T20:26:01+00:00 Solve conflict - - - - - 055c6910 by Ian Lynagh at 2010-09-22T15:36:20+00:00 Bump GHC dep - - - - - c96c0763 by Simon Marlow at 2010-10-27T11:09:44+00:00 follow changes in the GHC API - - - - - 45907129 by David Waern at 2010-11-07T14:00:58+00:00 Update the HCAR entry - - - - - 61940b95 by David Waern at 2010-11-07T14:07:34+00:00 Make the HCAR entry smaller - - - - - aa590b7d by David Waern at 2010-11-14T21:30:59+00:00 Update HCAR entry with November 2010 version - - - - - 587f9847 by David Waern at 2010-11-14T23:48:17+00:00 Require ghc >= 7.0 - - - - - ff5c647c by David Waern at 2010-11-14T23:49:09+00:00 TAG 2.8.2 - - - - - 937fcb4f by David Waern at 2010-11-14T23:49:45+00:00 Solve conflict - - - - - 8e5d0c1a by David Waern at 2010-11-15T21:09:50+00:00 Remove code for ghc < 7 - - - - - 3d47b70a by David Waern at 2010-11-15T21:11:06+00:00 Fix bad merge - - - - - 7f4a0d8a by David Waern at 2010-11-15T21:13:57+00:00 Remove more ghc < 7 code - - - - - 9ee34b50 by David Waern at 2010-11-15T21:31:25+00:00 Match all AsyncExceptions in exception handler - - - - - 42849c70 by David Waern at 2010-11-15T21:35:31+00:00 Just say "internal error" instead of "internal Haddock or GHC error" - - - - - c88c809b by David Waern at 2010-11-15T21:44:19+00:00 Remove docNameOcc under the motto "don't name compositions" - - - - - b798fc7c by David Waern at 2010-11-15T23:27:13+00:00 Wibble - - - - - 2228197e by David Waern at 2010-11-15T23:28:24+00:00 Rename the HCAR entry file - - - - - 8a3f9090 by David Waern at 2010-11-16T00:05:29+00:00 Remove Haskell 2010 extensions from .cabal file - - - - - c7a0c597 by David Waern at 2010-11-16T00:10:28+00:00 Style wibbles - - - - - cde707a5 by David Waern at 2010-11-16T00:12:00+00:00 Remove LANGUAGE ForeignFunctionInterface pragmas - - - - - 1dbda8ed by David Waern at 2010-11-16T00:17:21+00:00 Make a little more use of DoAndIfThenElse - - - - - 4c45ff6e by David Waern at 2010-11-16T00:59:41+00:00 hlint police - - - - - d2feaf09 by David Waern at 2010-11-16T01:14:15+00:00 hlint police - - - - - 99876e97 by David Waern at 2010-11-20T19:06:00+00:00 Haddock documentation updates - - - - - 65ce6987 by David Waern at 2010-11-20T19:42:51+00:00 Follow the style guide closer in Haddock.Types and improve docs - - - - - 28ca304a by tob.brandt at 2010-11-20T17:04:40+00:00 add full qualification for undocumented names - - - - - d61341e3 by David Waern at 2010-11-20T20:04:15+00:00 Re-structure qualification code a little - - - - - 0057e4d6 by David Waern at 2010-11-20T20:07:55+00:00 Re-order functions - - - - - d7279afd by David Waern at 2010-11-21T03:39:54+00:00 Add BangPatterns to alex and happy source files - - - - - 629fe60e by tob.brandt at 2010-11-23T23:35:11+00:00 documentation for qualification - - - - - 37031cee by David Waern at 2010-11-23T21:06:44+00:00 Update CHANGES - don't mention 2.8.2, we won't release it - - - - - f2489e19 by David Waern at 2010-12-01T21:57:11+00:00 Update deps of runtests.hs to work with ghc 7.0.1 - - - - - d3657e9a by David Waern at 2010-12-01T22:04:57+00:00 Make tests compile with ghc 7.0.1 - - - - - a2f09d9b by David Waern at 2010-12-01T22:06:59+00:00 Update tests following version bump - - - - - 50883ebb by David Waern at 2010-12-06T14:09:18+00:00 Update tests following recent changes - - - - - fc2fadeb by David Waern at 2010-12-06T14:17:29+00:00 Add a flag --pretty-html for rendering indented html with newlines - - - - - 30832ef2 by David Waern at 2010-12-06T14:17:35+00:00 Use --pretty-html when running the test suite. Makes it easier to compare output - - - - - a0b81b31 by David Waern at 2010-12-06T14:18:27+00:00 Wibble - - - - - 3aaa23fe by David Waern at 2010-12-06T14:19:29+00:00 Haddockify ppHtml comments - - - - - 24bb24f0 by David Waern at 2010-12-06T14:23:15+00:00 Remove --debug. It was't used, and --verbosity should take its place - - - - - 6bc076e5 by David Waern at 2010-12-06T14:25:37+00:00 Rename golden-tests into html-tests. "golden tests" sounds strange - - - - - 53301e55 by David Waern at 2010-12-06T14:26:26+00:00 QUALI -> QUAL in the description --qual for consistency - - - - - 98b6affb by David Waern at 2010-12-06T21:54:02+00:00 Bump version - - - - - 371bf1b3 by David Waern at 2010-12-06T22:08:55+00:00 Update tests following version bump - - - - - 25be762d by David Waern at 2010-12-06T22:21:03+00:00 Update CHANGES - - - - - 7c7dac71 by David Waern at 2010-12-06T22:33:43+00:00 Update ANNOUNCE - - - - - 30d7a5f2 by Simon Peyton Jones at 2010-11-15T08:38:38+00:00 Alex generates BangPatterns, so make Lex.x accept them (It'd be better for Alex to generate this pragma.) - - - - - 605e8018 by Simon Marlow at 2010-11-17T11:37:24+00:00 Add {-# LANGUAGE BangPatterns #-} to mollify GHC - - - - - a46607ba by David Waern at 2010-12-07T14:08:10+00:00 Solve conflicts - - - - - b28cda66 by David Waern at 2010-12-09T20:41:35+00:00 Docs: Mention that \ is a special character in markup - - - - - a435bfdd by Ian Lynagh at 2010-11-17T14:01:19+00:00 TAG GHC 7.0.1 release - - - - - 5a15a05a by David Waern at 2010-12-11T17:51:19+00:00 Fix indentation problem - - - - - 4232289a by Lennart Kolmodin at 2010-12-17T18:32:03+00:00 Revise haddock.cabal given that we now require ghc-7 default-language should be Haskell2010, slight new semantics for extensions. Rewrite into clearer dependencies of base and Cabal. - - - - - a36302dc by David Waern at 2010-12-19T17:12:37+00:00 Update CHANGES - - - - - 7c8b85b3 by David Waern at 2010-12-19T17:14:24+00:00 Bump version - - - - - cff22813 by Ian Lynagh at 2011-01-05T18:24:27+00:00 Write hoogle output in utf8; fixes GHC build on Windows - - - - - c7e762ea by David Waern at 2011-01-22T00:00:35+00:00 Put title outside doc div when HTML:fying title+prologue Avoids indenting the title, and makes more sense since the title is not a doc string anyway. - - - - - 5f639054 by David Waern at 2011-01-22T16:09:44+00:00 Fix spelling error - contributed by Marco Silva - - - - - c11dce78 by Ian Lynagh at 2011-01-07T02:33:11+00:00 Follow GHC build system changes - - - - - 101cfaf5 by David Waern at 2011-01-08T14:06:44+00:00 Bump version - - - - - af62348b by David Waern at 2011-01-08T14:07:07+00:00 TAG 2.9.2 - - - - - 4d1f6461 by Ian Lynagh at 2011-01-07T23:06:57+00:00 Name the haddock script haddock-ghc-7.0.2 instead of haddock-7.0.2; haskell/haddock#4882 "7.0.2" looked like a haddock version number before - - - - - 8ee4d5d3 by Simon Peyton Jones at 2011-01-10T17:31:12+00:00 Update Haddock to reflect change in hs_tyclds field of HsGroup - - - - - 06f3e3db by Ian Lynagh at 2011-03-03T15:02:37+00:00 TAG GHC 7.0.2 release - - - - - 7de0667d by David Waern at 2011-03-10T22:47:13+00:00 Update CHANGES - - - - - 33a9f1c8 by David Waern at 2011-03-10T22:47:31+00:00 Fix build with ghc 7.0.1 - - - - - 4616f861 by David Waern at 2011-03-10T22:47:50+00:00 TAG 2.9.2-actual - - - - - 0dab5e3c by Simon Hengel at 2011-04-08T15:53:01+00:00 Set shell script for unit tests back to work - - - - - 85c54dee by Simon Hengel at 2011-04-08T16:01:24+00:00 Set unit tests back to work Here "ghci>" was still used instead of ">>>". - - - - - 1cea9b78 by Simon Hengel at 2011-04-08T16:25:36+00:00 Update runtests.hs for GHC 7.0.2 - - - - - 8e5b3bbb by Simon Hengel at 2011-04-08T16:28:49+00:00 Update Haddock version in *.html.ref - - - - - 2545e955 by Simon Hengel at 2011-04-08T17:09:28+00:00 Add support for blank lines in the result of examples Result lines that only contain the string "<BLANKLINE>" are treated as a blank line. - - - - - adf64d2e by Simon Hengel at 2011-04-08T17:36:50+00:00 Add documentation for "support for blank lines in the result of examples" - - - - - c51352ca by David Waern at 2011-05-21T23:57:56+00:00 Improve a haddock comment - - - - - 7419cf2c by David Waern at 2011-05-22T15:41:52+00:00 Use cabal's test suite support to run the test suite This gives up proper dependency tracking of the test script. - - - - - 7770070c by David Waern at 2011-05-22T01:45:44+00:00 We don't need to send DocOptions nor a flag to mkExportItems - - - - - 9d95b7b6 by David Waern at 2011-05-22T21:39:03+00:00 Fix a bug - - - - - 1f93699b by David Waern at 2011-05-22T21:40:21+00:00 Break out fullContentsOf, give it a better name and some documentation The documentation describes how we want this function to eventually behave, once we have fixed a few problems with the current implementation. - - - - - 9a86432f by David Waern at 2011-05-22T21:53:52+00:00 Fix some stylistic issues in mkExportItems - - - - - c271ff0c by David Waern at 2011-05-22T22:09:11+00:00 Indentation - - - - - 93e602b1 by David Waern at 2011-06-10T01:35:31+00:00 Add git commits since switchover: darcs format (followed by a conflict resolution): commit 6f92cdd12d1354dfbd80f8323ca333bea700896a Merge: f420cc4 28df3a1 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Thu May 19 17:54:34 2011 +0100 Merge remote branch 'origin/master' into ghc-generics commit 28df3a119f770fdfe85c687dd73d5f6712b8e7d0 Author: Max Bolingbroke <batterseapower at hotmail.com> Date: Sat May 14 22:37:02 2011 +0100 Unicode fix for getExecDir on Windows commit 89813e729be8bce26765b95419a171a7826f6d70 Merge: 6df3a04 797ab27 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 9 11:55:17 2011 +0100 Merge branch 'ghc-new-co' commit 6df3a040da3dbddee67c6e30a892f87e6b164383 Author: Ian Lynagh <igloo at earth.li> Date: Sun May 8 17:05:50 2011 +0100 Follow changes in SDoc commit f420cc48b9259f0b1afd2438b12f9a2bde57053d Author: Jose Pedro Magalhaes <jpm at cs.uu.nl> Date: Wed May 4 17:31:52 2011 +0200 Adapt haddock to the removal of HsNumTy and TypePat. commit 797ab27bdccf39c73ccad374fea265f124cb52ea Merge: 1d81436 5a91450 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:05:03 2011 +0100 Merge remote branch 'origin/master' into ghc-new-co commit 1d8143659a81cf9611668348e33fd0775c7ab1d2 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:03:46 2011 +0100 Wibbles for ghc-new-co branch commit 5a91450e2ea5a93c70bd3904b022445c9cc82488 Author: Ian Lynagh <igloo at earth.li> Date: Fri Apr 22 00:51:56 2011 +0100 Follow defaultDynFlags change in GHC - - - - - 498da5ae by David Waern at 2011-06-11T00:33:33+00:00 * Merge in git patch from Michal Terepeta >From 6fc71d067738ef4b7de159327bb6dc3d0596be29 Mon Sep 17 00:00:00 2001 From: Michal Terepeta <michal.terepeta at gmail.com> Date: Sat, 14 May 2011 19:18:22 +0200 Subject: [PATCH] Follow the change of TypeSig in GHC. This follows the change in GHC to make TypeSig take a list of names (instead of just one); GHC ticket haskell/haddock#1595. This should also improve the Haddock output in case the user writes a type signature that refers to many names: -- | Some comment.. foo, bar :: ... will now generate the expected output with one signature for both names. - - - - - 094607fe by Ian Lynagh at 2011-06-17T19:10:29+01:00 Fix build - - - - - 8fa35740 by Ian Lynagh at 2011-06-26T21:06:40+01:00 Bump GHC dep to allow 7.2 - - - - - e4d2ca3c by Ian Lynagh at 2011-07-07T23:06:28+01:00 Relax base dep - - - - - b948fde9 by Ian Lynagh at 2011-07-28T16:39:45+01:00 GHC build system: Don't install the datafiles twice - - - - - f82f6d70 by Simon Marlow at 2011-08-11T12:08:15+01:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - b341cc12 by Max Bolingbroke at 2011-08-22T20:25:27+01:00 Fix compilation with no-pred-ty GHC - - - - - 30494581 by Max Bolingbroke at 2011-08-23T10:20:54+01:00 Remaining fixes for PredTy removal - - - - - 0b197138 by Max Bolingbroke at 2011-08-26T08:27:45+01:00 Rename factKind to constraintKind - - - - - a379bec5 by Max Bolingbroke at 2011-09-04T12:54:47+01:00 Deal with change to IParam handling in GHC - - - - - f94e421b by Max Bolingbroke at 2011-09-06T17:34:31+01:00 Adapt Haddock for the ConstraintKind extension changes - - - - - 8821e5cc by Max Bolingbroke at 2011-09-09T08:24:59+01:00 Ignore associated type defaults (just as we ignore default methods) - - - - - 31a0afd4 by Max Bolingbroke at 2011-09-09T09:06:00+01:00 Merge branch 'no-pred-ty' of ssh://darcs.haskell.org/srv/darcs/haddock into no-pred-ty - - - - - dd3b530a by Max Bolingbroke at 2011-09-09T14:10:25+01:00 Merge branch 'no-pred-ty' Conflicts: src/Haddock/Convert.hs - - - - - 5f25ec96 by Max Bolingbroke at 2011-09-09T14:10:40+01:00 Replace FactTuple with ConstraintTuple - - - - - cd30b9cc by David Waern at 2011-09-26T02:17:55+02:00 Bump to version 2.9.3 - - - - - 4fbfd397 by Max Bolingbroke at 2011-09-27T14:55:21+01:00 Follow changes to BinIface Name serialization - - - - - 92257d90 by David Waern at 2011-09-30T23:45:07+02:00 Fix problem with test files not added to distribution tarball - - - - - 00255bda by David Waern at 2011-09-30T23:48:24+02:00 Merge branch 'development' - - - - - 5421264f by David Waern at 2011-10-01T01:25:39+02:00 Merge in darcs patch from Simon Meier: Wed Jun 1 19:41:16 CEST 2011 iridcode at gmail.com * prettier haddock coverage info The new coverage info rendering uses less horizontal space. This reduces the number of unnecessary line-wrappings. Moreover, the most important information, how much has been documented already, is now put up front. Hopefully, this makes it more likely that a library author is bothered by the low coverage of his modules and fixes that issue ;-) - - - - - 07d318ef by David Waern at 2011-10-01T01:34:10+02:00 Use printException instead of deprecated printExceptionAndWarnings - - - - - 40d52ee4 by David Waern at 2011-10-01T01:41:13+02:00 Merge in darcs pach: Mon Apr 11 18:09:54 JST 2011 Liyang HU <haddock at liyang.hu> * Remember collapsed sections in index.html / haddock-util.js - - - - - 279d6dd4 by David Waern at 2011-10-01T01:55:45+02:00 Merge in darcs patch: Joachim Breitner <mail at joachim-breitner.de>**20110619201645 Ignore-this: f6c51228205b0902ad5bfad5040b989a As reported on http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=578301, generating the global index takes much too long if type-level (with lots of auto-generated types) is installed. The patch avoids a quadratic runtime in the subfunction getIfaceIndex of ppHtmlIndex by using a temporary set. Runtime improvement observed here from 25.36s to 2.86s. - - - - - d1612383 by David Waern at 2011-10-01T01:56:48+02:00 Merge branch 'development' - - - - - 347520c1 by David Waern at 2011-10-01T01:56:54+02:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9a0c95e8 by David Waern at 2011-10-01T02:19:10+02:00 Improve .cabal file - - - - - 6967dc64 by Ian Lynagh at 2011-10-01T01:34:06+01:00 Follow changes to ForeignImport/ForeignExport in GHC - - - - - 565cb26b by Simon Marlow at 2011-10-04T00:15:04+02:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - 8b74f512 by David Waern at 2011-10-04T00:18:17+02:00 Requre ghc >= 7.2 - - - - - 271d360c by David Waern at 2011-10-04T00:22:50+02:00 Bump version to 2.9.4 - - - - - 37f3edb0 by David Waern at 2011-10-06T02:30:21+02:00 Add alex and happy to build-tools. - - - - - 7ac2bb6e by David Terei at 2011-10-12T14:02:55-07:00 Add safe haskell indication to haddock output - - - - - 42c91a47 by David Terei at 2011-10-12T14:06:03-07:00 Fix CSS issue with info table not being contained in module header - - - - - 0eddab6c by David Terei at 2011-10-12T14:06:58-07:00 Add safe haskell indication to haddock output - - - - - 3df058eb by David Terei at 2011-10-12T14:07:07-07:00 Fix CSS issue with info table not being contained in module header - - - - - a40a6c3f by David Waern at 2011-10-22T11:29:06+02:00 Bump .haddock file version since the format has changed recently - - - - - 8a6254be by David Waern at 2011-10-22T11:30:42+02:00 Merge branch 'development' - - - - - 642e3e02 by David Waern at 2011-10-23T21:23:39+02:00 Sort import list - - - - - 36371cf8 by David Waern at 2011-10-23T22:48:18+02:00 Remove NEW_GHC_LAYOUT conditional. - - - - - 5604b499 by David Waern at 2011-10-27T00:15:03+02:00 Add --print-ghc-path. - - - - - 463499fa by David Waern at 2011-10-27T00:16:22+02:00 Make testsuite able to find its dependencies automatically. - - - - - a3506172 by Ryan Newton at 2011-11-05T05:59:58-04:00 Improved declNames internal error. Added a case to handle DocD. - - - - - 001b8baf by David Waern at 2011-11-05T20:37:29+01:00 Rename copy.hs -> accept.hs. - - - - - 55d808d3 by David Waern at 2011-11-05T23:30:02+01:00 Fix build. - - - - - deb5c3be by David Waern at 2011-11-06T00:01:47+01:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9b663554 by David Waern at 2011-11-06T00:03:45+01:00 Merge https://github.com/rrnewton/haddock - - - - - 1abb0ff6 by David Waern at 2011-11-06T01:20:37+01:00 Use getDeclMainBinder instead of declNames. - - - - - 4b005c01 by David Waern at 2011-11-06T19:09:53+01:00 Fix build. - - - - - c2c51bc7 by Ian Lynagh at 2011-11-06T23:01:33+00:00 Remove -DNEW_GHC_LAYOUT in ghc.mk - - - - - f847d703 by Jose Pedro Magalhaes at 2011-11-11T09:07:39+00:00 New kind-polymorphic core This big patch implements a kind-polymorphic core for GHC. The current implementation focuses on making sure that all kind-monomorphic programs still work in the new core; it is not yet guaranteed that kind-polymorphic programs (using the new -XPolyKinds flag) will work. For more information, see http://haskell.org/haskellwiki/GHC/Kinds - - - - - 7d7c3b09 by Jose Pedro Magalhaes at 2011-11-16T21:42:22+01:00 Follow changes to tuple sorts in master - - - - - 8430e03e by Simon Peyton Jones at 2011-11-17T10:20:27+00:00 Remove redundant imports - - - - - d1b06832 by Ian Lynagh at 2011-11-19T01:33:21+00:00 Follow GHC build system change to the way we call rm - - - - - 9e2230ed by David Waern at 2011-11-24T15:00:24+01:00 Fix a bug in test runner and get rid of regex-compat dependency. - - - - - 52039b21 by David Waern at 2011-11-24T23:55:36+01:00 Avoid haskell98 dependency in test - - - - - 92e1220d by David Waern at 2011-11-25T00:03:33+01:00 Avoid depency on regex-compat also in accept.hs. - - - - - ddac6b6f by David Waern at 2011-11-25T02:13:38+01:00 Accept test output. - - - - - 5a720455 by David Waern at 2011-11-25T02:16:20+01:00 Some more changes to test scripts. - - - - - 170a9004 by David Waern at 2011-11-25T02:30:41+01:00 Add flag --interface-version. - - - - - d225576c by David Waern at 2011-11-25T02:39:26+01:00 Remove #ifs for older compiler versions. - - - - - f0d0a4f5 by David Waern at 2011-11-26T04:20:12+01:00 Give preference to type over data constructors for doc comment links at renaming time. Previously this was done in the backends. Also, warn when a doc comment refers to something that is in scope but which we don't have the .haddock file for. These changes mean we can make DocIdentifier [a] into DocIdentifier a. - - - - - eef0e776 by David Waern at 2011-11-26T17:01:06+01:00 Allow doc comments to link to out-of-scope things (#78). (A bug that should have been fixed long ago.) - - - - - 565ad529 by David Waern at 2011-11-26T19:56:21+01:00 Update tests. - - - - - fb3ce7b9 by David Waern at 2011-11-26T21:44:28+01:00 Cleanup. - - - - - d0328126 by David Waern at 2011-11-26T22:10:28+01:00 Fix module reference bug. - - - - - c03765f8 by David Waern at 2011-12-03T05:20:20+01:00 Slightly better behaviour on top-levels without type signatures. - Docs don't get attached to the next top-level with signature by mistake. - If there's an export list and the top-level is part of it, its doc comment shows up in the documentation. - - - - - 48461d31 by David Waern at 2011-12-03T05:38:10+01:00 Add a test for Unicode doc comments. - - - - - 549c4b4e by David Waern at 2011-12-03T19:07:55+01:00 Cleanup. - - - - - 7bfecf91 by David Waern at 2011-12-03T20:13:08+01:00 More cleanup. - - - - - 14fab722 by Ian Lynagh at 2011-12-12T21:21:35+00:00 Update dependencies and binaryInterfaceVersion - - - - - 469e6568 by Ian Lynagh at 2011-12-18T12:56:16+00:00 Fix (untested) building from source tarball without alex/happy haddock's .cabal file was declaring that it needed alex and happy to build, but in the GHC source tarballs it doesn't. - - - - - 895c9a8c by David Waern at 2011-12-27T12:57:43+01:00 Go back to having a doc, sub and decl map instead of one big decl map. This setup makes more sense since when we add value bindings to the processed declarations (for type inference), we will have multiple declarations which should share documentation. Also, we already have a separate doc map for instances which we can now merge into the main doc map. Another benefit is that we don't need the DeclInfo type any longer. - - - - - 736767d9 by David Waern at 2011-12-27T13:33:41+01:00 Merge ../../../haddock Conflicts: src/Haddock/InterfaceFile.hs - - - - - 20016f79 by David Waern at 2011-12-27T13:57:23+01:00 Bump version. - - - - - 31f276fb by David Waern at 2011-12-27T13:57:32+01:00 Merge ../ghc/utils/haddock - - - - - 95b367cd by David Waern at 2011-12-27T14:57:29+01:00 Update tests following version bump. - - - - - fa3c94cd by David Waern at 2011-12-27T14:57:51+01:00 Get rid of quite unnecessary use of different lists. - - - - - 9c4d3c54 by David Waern at 2011-12-27T15:26:42+01:00 Cleanup. - - - - - 2caf9f90 by David Waern at 2011-12-27T16:18:05+01:00 Wibbles. - - - - - 3757d09b by David Waern at 2011-12-27T20:50:26+01:00 Complete support for inferring types for top-level bindings. - - - - - 53418734 by David Waern at 2011-12-28T15:02:13+01:00 Minor fixes and cleanup. - - - - - 0c9d0385 by Ian Lynagh at 2012-01-03T18:31:29+00:00 Follow rename of Instance to ClsInst in GHC - - - - - c9bc969a by Simon Hengel at 2012-01-12T21:28:14+01:00 Make sure that generated xhtml is valid (close haskell/haddock#186) Thanks to Phyx. - - - - - 836a0b9a by David Waern at 2012-02-01T02:30:05+01:00 Fix bug introduced in my recent refactoring. - - - - - c7d733eb by David Waern at 2012-02-01T02:30:26+01:00 Cleanup mkMaps and avoid quadratic behaviour. - - - - - da3cda8f by David Waern at 2012-02-01T02:56:56+01:00 Require ghc >= 7.4. - - - - - 83a3287e by David Waern at 2012-02-01T02:57:36+01:00 Update CHANGES. - - - - - 93408f0b by Simon Hengel at 2012-02-04T00:48:04+01:00 Add reference renderings - - - - - 49d00d2c by Simon Hengel at 2012-02-04T00:48:25+01:00 Set unit tests for parser back to work - - - - - eb450980 by Simon Hengel at 2012-02-04T00:49:07+01:00 Add .gitignore - - - - - a841602c by Simon Hengel at 2012-02-04T00:49:16+01:00 Add .ghci file - - - - - 8861199d by Simon Hengel at 2012-02-04T00:49:29+01:00 tests/html-tests/copy.hs: Use mapM_ instead of mapM So we do net get a list of () on stdout when running with runhaskell. - - - - - b477d9b5 by Simon Hengel at 2012-02-04T00:49:46+01:00 Remove index files from golden tests - - - - - 9dbda34e by Simon Hengel at 2012-02-04T00:49:57+01:00 Add /tests/html-tests/tests/*index*.ref to .gitignore - - - - - a9434817 by Simon Hengel at 2012-02-04T00:50:04+01:00 Add DocWarning to Doc The Xhtml backend has special markup for that, Hoogle and LaTeX reuse what we have for DocEmphasis. - - - - - de2fb6fa by Simon Hengel at 2012-02-04T00:50:13+01:00 Add support for module warnings - - - - - 0640920e by Simon Hengel at 2012-02-04T00:50:21+01:00 Add tests for module warnings - - - - - 30ce0d77 by Simon Hengel at 2012-02-04T00:50:29+01:00 Add support for warnings - - - - - bb367960 by Simon Hengel at 2012-02-04T00:50:37+01:00 Add tests for warnings - - - - - 6af1dc2d by Simon Hengel at 2012-02-04T00:50:50+01:00 Expand type signatures in export list (fixes haskell/haddock#192) - - - - - a06cbf25 by Simon Hengel at 2012-02-04T00:51:04+01:00 Expand type signatures for modules without explicit export list - - - - - 57dda796 by Simon Hengel at 2012-02-04T00:51:15+01:00 Remove obsolete TODO - - - - - 270c3253 by David Waern at 2012-02-04T00:51:24+01:00 Fix issues in support for warnings. * Match against local names only. * Simplify (it's OK to map over the warnings). - - - - - 683634bd by David Waern at 2012-02-04T00:55:11+01:00 Some cleanup and make sure we filter warnings through exports. - - - - - 210cb4ca by David Waern at 2012-02-04T03:01:30+01:00 Merge branch 'fix-for-186' of https://github.com/sol/haddock into ghc-7.4 - - - - - e8db9031 by David Waern at 2012-02-04T03:07:51+01:00 Style police. - - - - - 261f9462 by David Waern at 2012-02-04T03:20:16+01:00 Update tests. - - - - - 823cfc7c by David Waern at 2012-02-04T03:21:12+01:00 Use mapM_ in accept.hs as well. - - - - - 873dd619 by David Waern at 2012-02-04T03:21:33+01:00 Remove copy.hs - use accept.hs instead. - - - - - 0e31a14a by David Waern at 2012-02-04T03:47:33+01:00 Use <> instead of mappend. - - - - - 2ff7544f by David Waern at 2012-02-04T03:48:55+01:00 Remove code for older ghc versions. - - - - - dacf2786 by David Waern at 2012-02-04T15:52:51+01:00 Clean up some code from last SoC project. - - - - - 00cbb117 by David Waern at 2012-02-04T21:43:49+01:00 Mostly hlint-inspired cleanup. - - - - - 7dc86cc2 by Simon Peyton Jones at 2012-02-06T09:14:41+00:00 Track changes in HsDecls - - - - - f91f82fe by Ian Lynagh at 2012-02-16T13:40:11+00:00 Follow changes in GHC caused by the CAPI CTYPE pragma - - - - - a0ea6b0b by Ian Lynagh at 2012-02-22T02:26:12+00:00 Follow changes in GHC - - - - - b23b07d1 by Simon Peyton Jones at 2012-03-02T16:36:41+00:00 Follow changes in data representation from the big PolyKinds commit - - - - - 43406022 by Simon Hengel at 2012-03-05T11:18:34+01:00 Save/restore global state for static flags when running GHC actions This is necessary if we want to run createInterfaces (from Documentation.Haddock) multiple times in the same process. - - - - - 9fba16fe by Paolo Capriotti at 2012-03-06T10:57:33+00:00 Update .gitignore. - - - - - a9325044 by Simon Peyton Jones at 2012-03-14T17:35:42+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - fd48065a by Iavor Diatchki at 2012-03-15T22:43:35-07:00 Add support for type-level literals. - - - - - 2e8206dd by Simon Peyton Jones at 2012-03-16T14:18:22+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - 93e13319 by Simon Peyton Jones at 2012-03-17T01:04:05+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock Conflicts: src/Haddock/Convert.hs - - - - - d253fa71 by Iavor Diatchki at 2012-03-19T20:12:18-07:00 Merge remote-tracking branch 'origin/master' into type-nats - - - - - fc40acc8 by Iavor Diatchki at 2012-03-19T20:31:27-07:00 Add a missing case for type literals. - - - - - fd2ad699 by Iavor Diatchki at 2012-03-24T13:28:29-07:00 Rename variable to avoid shadowing warning. - - - - - 9369dd3c by Simon Peyton Jones at 2012-03-26T09:14:23+01:00 Follow refactoring of TyClDecl/HsTyDefn - - - - - 38825ca5 by Simon Peyton Jones at 2012-03-26T09:14:37+01:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - 4324ac0f by David Waern at 2012-04-01T01:51:19+02:00 Disable unicode test. - - - - - 3165b750 by David Waern at 2012-04-01T01:51:34+02:00 Take reader environment directly from TypecheckedSource. - - - - - 213b644c by David Waern at 2012-04-01T01:55:20+02:00 Cleanup. - - - - - 3118b4ba by David Waern at 2012-04-01T02:16:15+02:00 Don't filter out unexported names from the four maps - fixes a regression. - - - - - d6524e17 by David Waern at 2012-04-01T02:40:34+02:00 Fix crash when using --qual. Naughty GHC API! - - - - - ea3c43d8 by Henning Thielemann at 2012-04-01T13:03:07+02:00 add QualOption type for distinction between qualification argument given by the user and the actual qualification for a concrete module - - - - - 5422ff05 by Henning Thielemann at 2012-04-01T16:25:02+02:00 emit an error message when the --qual option is used incorrectly - - - - - 026e3404 by David Waern at 2012-04-01T18:10:30+02:00 Don't crash on unicode strings in doc comments. - - - - - ce006632 by David Waern at 2012-04-01T20:13:35+02:00 Add test for --ignore-all-exports flag/ignore-exports pragma. - - - - - 6e4dd33c by David Waern at 2012-04-01T20:21:03+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.4 - - - - - 734ae124 by Henning Thielemann at 2012-04-01T20:22:10+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - 622f9ba5 by David Waern at 2012-04-01T21:26:13+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 55ce17cb by Henning Thielemann at 2012-04-01T22:03:25+02:00 'abbreviate' qualification style - basic support Currently we ignore the package a module is imported from. This means that a module import would shadow another one with the same module name from a different package. - - - - - c85314ef by David Waern at 2012-04-01T22:05:12+02:00 Check qualification option before processing modules. - - - - - ae4b626c by Henning Thielemann at 2012-04-02T00:19:36+02:00 abbreviated qualification: use Packages.lookupModuleInAllPackages for finding the package that a module belongs to - - - - - 60bdbcf5 by Henning Thielemann at 2012-04-02T00:25:31+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - df44301d by Henning Thielemann at 2012-04-02T00:29:05+02:00 qualification style 'abbreviated' -> 'aliased' - - - - - f4192a64 by David Waern at 2012-04-02T01:05:47+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 7ba09067 by David Terei at 2012-04-04T15:08:21-07:00 Fix reporting of modules safe haskell mode (#5989) - - - - - d0cc33d0 by David Terei at 2012-04-06T15:50:41+01:00 Fix reporting of modules safe haskell mode (#5989) - - - - - 6e3434c5 by Simon Peyton Jones at 2012-04-20T18:37:46+01:00 Track changes in HsSyn - - - - - 22014ed0 by Simon Peyton Jones at 2012-05-11T22:45:15+01:00 Follow changes to LHsTyVarBndrs - - - - - d9a07b24 by David Waern at 2012-05-15T01:46:35+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - a6c4ebc6 by David Waern at 2012-05-16T02:18:32+02:00 Update CHANGES. - - - - - 8e181d29 by David Waern at 2012-05-16T02:27:56+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - e358210d by David Waern at 2012-05-16T02:35:33+02:00 Mention the new aliased --qual mode in CHANGES. - - - - - efd36a28 by David Waern at 2012-05-16T21:33:13+02:00 Bump version number. - - - - - d6b3af14 by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for deprecated record field - - - - - 927f800e by Simon Hengel at 2012-05-17T19:08:20+02:00 Use >>= instead of fmap and join - - - - - 048b41d5 by Simon Hengel at 2012-05-17T19:08:20+02:00 newtype-wrap Doc nodes for things that may have warnings attached - - - - - e3a89fc3 by Simon Hengel at 2012-05-17T19:08:20+02:00 Attach warnings to `Documentation` type - - - - - 5d4cc43d by Simon Hengel at 2012-05-17T19:08:20+02:00 Simplify lookupWarning - - - - - cf8ae69d by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for haskell/haddock#205 - - - - - cb409b19 by Simon Peyton Jones at 2012-05-25T08:30:11+01:00 Follow changes in LHsTyVarBndrs - - - - - 2d5f4179 by Simon Hengel at 2012-05-26T19:21:29+02:00 Add Applicative instance for (GenRnM a) - - - - - e4373060 by Simon Hengel at 2012-05-26T19:21:33+02:00 Use a map for warnings, as suggested by @waern - - - - - 597a68c7 by Simon Hengel at 2012-05-27T08:48:24+02:00 Add an optional label to URLs - - - - - ef1ac7fe by Simon Hengel at 2012-05-27T08:48:24+02:00 Add support for hyperlink labels to parser - - - - - 41f2adce by Simon Hengel at 2012-05-27T08:48:24+02:00 Add golden test for hyperlinks - - - - - 83d5e764 by Simon Hengel at 2012-05-27T08:50:02+02:00 Use LANGUAGE pragmas instead of default-extensions in cabal file - - - - - ddb755e5 by Simon Hengel at 2012-05-27T08:50:02+02:00 Fix typo in comment - - - - - 110676b4 by Simon Hengel at 2012-05-27T08:50:02+02:00 Add a type signature for a where-binding - - - - - 7d9ba2a0 by Ian Lynagh at 2012-06-12T14:38:01+01:00 Follow changes in GHC - - - - - 47c704f2 by Ian Lynagh at 2012-06-12T18:52:16+01:00 Follow changes in GHC - - - - - e1efe1ab by Simon Peyton Jones at 2012-06-13T17:25:29+01:00 Follow changes for the implementation of implicit parameters - - - - - 69abc81c by Ian Lynagh at 2012-06-19T22:52:58+01:00 Follow changes in base - - - - - 9d074a21 by Paolo Capriotti at 2012-06-22T18:26:47+01:00 Use right docMap to get decl documentation. - - - - - e3292ef6 by Ian Lynagh at 2012-07-15T01:31:19+01:00 Follow changes in GHC - - - - - ceae56b0 by Ian Lynagh at 2012-07-16T21:22:48+01:00 Fix haddock following some GHC changes Passing _|_ as the Settings for defaultDynFlags no longer works well enough - - - - - 9df72735 by Paolo Capriotti at 2012-07-19T16:49:32+01:00 Forward port changes from stable. - - - - - 572f5fcf by Ian Lynagh at 2012-07-19T20:38:26+01:00 Merge branch 'master' of darcs.haskell.org:/srv/darcs//haddock - - - - - 9195aca4 by Paolo Capriotti at 2012-07-20T10:27:28+01:00 Update dependencies. - - - - - 33db3923 by Ian Lynagh at 2012-07-20T17:54:43+01:00 Build with GHC 7.7 - - - - - 925a2cea by David Waern at 2012-07-23T16:50:40+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.6 Conflicts: src/Haddock/InterfaceFile.hs - - - - - d710ef97 by David Waern at 2012-07-23T16:52:07+02:00 Bump version number. - - - - - eb0c2f83 by David Waern at 2012-07-23T16:57:58+02:00 Update CHANGES. - - - - - b3f56943 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Hide "internal" instances This fixes haskell/haddock#37 (http://trac.haskell.org/haddock/ticket/37) Precisely, we show an instance iff its class and all the types are exported by non-hidden modules. - - - - - a70aa412 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Tests for hiding instances (#37) - - - - - c0f4aa58 by Simon Hengel at 2012-07-27T13:00:13+03:00 Add an other test for hiding instances (#37) - - - - - a7ed6268 by Ian Lynagh at 2012-08-07T14:48:13+01:00 Follow changes in GHC - - - - - 0ab30d38 by Ian Lynagh at 2012-08-13T22:12:27+01:00 Improve haddock memory usage - - - - - 0eaa4e30 by Ian Lynagh at 2012-08-13T23:58:46+01:00 Improve haddock memory usage - - - - - 659d26cf by Ian Lynagh at 2012-08-14T13:16:48+01:00 Remove some temporary pragmas I accidentally recorded - - - - - d97fceb6 by Simon Hengel at 2012-08-25T13:19:34+02:00 Add missing dependency to library - - - - - 4c910697 by Simon Hengel at 2012-08-28T07:39:14+02:00 Move .ghci to project root - - - - - fc3c601a by Simon Hengel at 2012-08-28T07:39:14+02:00 accept.hs: Ignore some files - - - - - 1af9b984 by Simon Hengel at 2012-08-28T07:40:04+02:00 Update reference renderings (bump version) - - - - - 980dc253 by Simon Hengel at 2012-08-28T07:40:32+02:00 Update reference renderings (remove links for ()) - - - - - 33651dbf by Simon Hengel at 2012-08-28T07:41:50+02:00 Update documentation of `runInteractiveProcess` in reference rendering - - - - - 7ab25078 by David Waern at 2012-09-07T10:38:50+02:00 Merge branch 'hiddenInstances2' of http://github.com/feuerbach/haddock into ghc-7.6 - - - - - c3de3a4b by David Waern at 2012-09-07T14:29:27+02:00 Follow changes in GHC. - - - - - 298c43ac by David Waern at 2012-09-07T14:59:24+02:00 Update CHANGES. - - - - - e797993a by David Waern at 2012-09-07T15:21:30+02:00 Update ANNOUNCE. - - - - - d0b44790 by David Waern at 2012-09-07T15:22:43+02:00 Merge branch 'hidden-instances' into ghc-7.6 - - - - - 41a4adc8 by Simon Hengel at 2012-09-08T12:08:37+02:00 Update doc/README - - - - - 71ad1040 by Simon Hengel at 2012-09-08T12:17:17+02:00 Add documentation for URL labels - - - - - 9bb41afd by Simon Peyton Jones at 2012-09-20T18:14:26+01:00 Follow data type changes in the tc-untouchables branch Relating entirely to SynTyConRhs - - - - - b8139bfa by Simon Hengel at 2012-09-21T14:24:16+02:00 Disable Unicode test for now - - - - - a5fafdd7 by Simon Hengel at 2012-09-21T14:35:45+02:00 Update TypeOperators test for GHC 7.6.1 Type operators can't be used as type variables anymore! - - - - - 6ccf0025 by Simon Hengel at 2012-09-21T16:02:24+02:00 Remove (Monad (Either e)) instance from ref. rendering of CrossPackageDocs I do not really understand why the behavior changed, so I'll open a ticket, so that we can further investigate. - - - - - b5c6c138 by Ian Lynagh at 2012-09-27T02:00:57+01:00 Follow changes in GHC build system - - - - - b98eded0 by David Waern at 2012-09-27T15:37:02+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 76cc2051 by David Waern at 2012-09-27T15:48:19+02:00 Update hidden instances tests. - - - - - aeaa1c59 by David Waern at 2012-09-28T10:21:32+02:00 Make API buildable with GHC 7.6. - - - - - d76be1b0 by Simon Peyton Jones at 2012-09-28T15:57:05+01:00 Merge remote-tracking branch 'origin/master' into tc-untouchables - - - - - a1922af8 by David Waern at 2012-09-28T19:50:20+02:00 Fix spurious superclass constraints bug. - - - - - bc41bdbb by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove old examples - - - - - bed7d3dd by Simon Hengel at 2012-10-01T11:30:51+02:00 Adapt parsetests for GHC 7.6.1 - - - - - dcdb22bb by Simon Hengel at 2012-10-01T11:30:51+02:00 Add test-suite section for parsetests to cabal file + get rid of HUnit dependency - - - - - 1e5263c9 by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove test flag from cabal file This was not really used. - - - - - 4beee98b by David Waern at 2012-09-28T23:42:28+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 11dd2256 by Ian Lynagh at 2012-10-03T16:17:35+01:00 Follow change in GHC build system - - - - - fbd77962 by Simon Hengel at 2012-10-03T18:49:40+02:00 Remove redundant dependency from cabal file - - - - - 09218989 by Simon Hengel at 2012-10-04T16:03:05+02:00 Fix typo - - - - - 93a2d5f9 by Simon Hengel at 2012-10-04T16:11:41+02:00 Remove trailing whitespace from cabal file - - - - - c8b46cd3 by Simon Hengel at 2012-10-04T16:12:17+02:00 Export Haddock's main entry point from library - - - - - b411e77b by Simon Hengel at 2012-10-04T16:29:46+02:00 Depend on library for executable The main motivation for this is to increase build speed. In GHC's source tree the library is not build, but all modules are now required for the executable, so that GHC's validate will now detect build failures for the library. - - - - - f8f0979f by Simon Hengel at 2012-10-05T00:32:57+02:00 Set executable flag for Setup.lhs - - - - - dd045998 by Simon Hengel at 2012-10-07T16:44:06+02:00 Extend rather than set environment when running HTML tests On some platforms (e.g. ppc64) GHC requires gcc in the path. - - - - - 7b39c3ae by Simon Hengel at 2012-10-07T17:05:45+02:00 cross-package test: re-export IsString instead of Monad There is a monad instance for Q, which is not available on platforms that do not have GHCi support. This caused CrossPackageDocs to fail on those platforms. Re-exporting IsString should test the same thing, but it works on all platforms. - - - - - 0700c605 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Fix some warnings - - - - - f78eca79 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Make -Wall proof - - - - - 6beec041 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Use listToMaybe/fromMaybe instead of safeHead/maybe - - - - - 44b8ce86 by Ian Lynagh at 2012-10-08T21:59:46+01:00 Follow changes in GHC - - - - - 6da5f702 by Simon Hengel at 2012-10-09T11:16:19+02:00 Update .ghci - - - - - 9ac1a1b9 by Kazu Yamamoto at 2012-10-09T12:45:31+02:00 Add markup support for properties - - - - - 1944cb42 by Simon Hengel at 2012-10-09T12:45:31+02:00 Simplify lexing/parsing of properties In contrast to what we do for examples, we do not really need to capture the "prompt" here. - - - - - bffd8e62 by Simon Hengel at 2012-10-09T13:40:14+02:00 Add HTML test for properties - - - - - 2fe9c5cb by Simon Hengel at 2012-10-09T13:40:21+02:00 Add unit tests for properties - - - - - 874e361b by Simon Hengel at 2012-10-09T13:40:33+02:00 Bump interface version - - - - - 2506cc37 by Simon Hengel at 2012-10-09T15:15:04+02:00 Fix parser bug - - - - - 743d2b7d by Simon Hengel at 2012-10-09T15:31:06+02:00 Allow to load interface files with compatible versions - - - - - 981a1660 by Simon Hengel at 2012-10-10T10:32:05+02:00 Export more types from Documentation.Haddock (fixes haskell/haddock#216) - - - - - dff7dc76 by Simon Hengel at 2012-10-10T11:15:19+02:00 Update ANNOUNCE and CHANGES - - - - - edd2bb01 by Simon Hengel at 2012-10-10T11:22:50+02:00 Bump version - - - - - 5039163b by Simon Hengel at 2012-10-10T13:56:04+02:00 Fix typo in documentation - - - - - e4ce34da by Simon Hengel at 2012-10-10T14:28:35+02:00 Add documentation for properties - - - - - 9555ebca by Simon Hengel at 2012-10-11T10:49:04+02:00 Remove redundant if-defs, more source documentation - - - - - 87aa67e1 by Simon Hengel at 2012-10-11T12:32:51+02:00 Adapt cabal file - - - - - c44c1dee by Simon Hengel at 2012-10-11T12:41:58+02:00 Require ghc 7.6 - - - - - 8383bc34 by Simon Hengel at 2012-10-11T12:50:24+02:00 Bump version - - - - - 1030eb38 by Simon Hengel at 2012-10-11T12:55:44+02:00 Update ANNOUNCE and CHANGES - - - - - 74955088 by Simon Hengel at 2012-10-12T09:49:31+02:00 Improve note about `binaryInterfaceVersion` (thanks David) - - - - - ee30f6b7 by Simon Hengel at 2012-10-13T13:40:59+02:00 Update version in html tests, rpm spec file, and user manual - - - - - f2861f18 by Simon Hengel at 2012-10-13T14:40:33+02:00 Remove unused MonadFix constraint - - - - - dfdf1a74 by Simon Hengel at 2012-10-13T15:15:38+02:00 Minor code simplification - - - - - 4ecd1e70 by Simon Hengel at 2012-10-13T15:33:43+02:00 Increase code locality - - - - - f7df5cc9 by Simon Hengel at 2012-10-13T16:03:12+02:00 Minor code simplification - - - - - e737eb6e by Simon Hengel at 2012-10-13T19:03:04+02:00 Handle HsExplicitListTy in renameer (fixes haskell/haddock#213) - - - - - c2dc8f17 by Simon Hengel at 2012-10-13T20:46:31+02:00 Better error messages - - - - - 14d48b4c by Simon Hengel at 2012-10-14T00:21:07+02:00 Simplify RnM type - - - - - 6c2cc547 by Simon Hengel at 2012-10-14T00:23:35+02:00 Simplify lookupRn - - - - - bc77ce85 by Simon Hengel at 2012-10-14T01:51:32+02:00 Organize unite tests hierarchically - - - - - 2306d117 by Simon Hengel at 2012-10-14T10:34:58+02:00 Handle more cases in renameType - - - - - 8a864203 by Simon Hengel at 2012-10-14T11:47:59+02:00 Add mini_HiddenInstances.html.ref and mini_HiddenInstancesB.html.ref - - - - - 3a978eca by Simon Hengel at 2012-10-14T11:49:28+02:00 Add /tests/html-tests/output/ to .gitignore - - - - - db18888a by Simon Hengel at 2012-10-14T13:38:21+02:00 Allow haddock markup in deprecation messages - - - - - e7cfee9f by Simon Hengel at 2012-10-14T14:00:23+02:00 If parsing of deprecation message fails, include it verbatim - - - - - 242a85be by Simon Hengel at 2012-10-14T14:13:24+02:00 Add description for PruneWithWarning test - - - - - 43d33df1 by Simon Hengel at 2012-10-14T15:40:53+02:00 Minor formatting change - - - - - 22768c44 by Simon Hengel at 2012-10-14T16:03:43+02:00 Properly handle deprecation messages for re-exported things (fixes haskell/haddock#220) - - - - - cb4b9111 by Simon Hengel at 2012-10-14T17:30:28+02:00 Add build artifacts for documentation to .gitignore - - - - - 854cd8de by Simon Hengel at 2012-10-14T23:34:51+02:00 unit-tests: Improve readability Add IsString instance for (Doc RdrName) + use <> instead of DocAppend. - - - - - c4446d54 by Simon Hengel at 2012-10-14T23:37:21+02:00 unit-tests: Minor refactoring Rename parse to parseParas. - - - - - 04f2703c by Simon Hengel at 2012-10-15T00:36:42+02:00 Fix typo - - - - - 3d109e44 by Simon Hengel at 2012-10-15T10:30:07+02:00 Add description for DeprecatedReExport test - - - - - 84f0985c by Simon Hengel at 2012-10-15T14:54:19+02:00 Move resources to /resources directory - - - - - a5de7ca6 by Simon Hengel at 2012-10-15T15:46:18+02:00 Move HTML tests to directory /html-test/ - - - - - e21f727d by Simon Hengel at 2012-10-15T19:32:42+02:00 Move HTML reference renderings to /html-test/ref/ - - - - - 3a3c6c75 by Simon Hengel at 2012-10-15T19:32:42+02:00 Copy css, images, etc. on accept - - - - - 40ead6dc by Simon Hengel at 2012-10-15T19:32:42+02:00 Move unit tests to /test directory - - - - - 99a28231 by Simon Hengel at 2012-10-15T19:32:42+02:00 Fix Setup.lhs /usr/bin/runhaskell is not installed on all systems. - - - - - 95faf45e by Simon Hengel at 2012-10-15T19:32:42+02:00 Make test management scripts more robust * They are now independent from the current directory, and hence can be called from everywhere * On UNIX/Linux they can now be run as scripts - - - - - 027aaa2d by Simon Hengel at 2012-10-15T19:53:40+02:00 Add 'dev' flag to cabal file, that builds without -O2 That way --disable-optimization can be used, which decreases build time considerably. - - - - - e0266ede by Simon Hengel at 2012-10-15T20:03:43+02:00 Add test case for "spurious superclass constraints bug" - - - - - 52a2aa92 by Simon Hengel at 2012-10-15T20:28:55+02:00 Adapt accept.lhs, so that it ignores more index files - - - - - 53530781 by Simon Hengel at 2012-10-15T20:49:39+02:00 Rename html-test/runtests.lhs to html-test/run.lhs - - - - - 84518797 by Simon Hengel at 2012-10-15T20:49:39+02:00 Move source files for HTML tests to html-test/src - - - - - a911dc6c by Simon Hengel at 2012-10-15T20:49:39+02:00 Adapt output directory for HTML tests - - - - - d3c15857 by Ian Lynagh at 2012-10-16T16:54:43+01:00 Follow dopt->gopt rename - - - - - 956665a5 by Simon Hengel at 2012-10-18T08:42:48+02:00 Update html-test/README - - - - - 903b1029 by Simon Hengel at 2012-10-18T08:50:26+02:00 Use markdown for html-test/README - - - - - 150b4d63 by Ian Lynagh at 2012-10-18T16:36:00+01:00 Follow changes in GHC: 'flags' has been renamed 'generalFlags' - - - - - 41e04ff9 by Simon Hengel at 2012-11-28T09:54:35+01:00 Export missing types from Documentation.Haddock - - - - - 9be59237 by Ian Lynagh at 2012-11-30T23:20:47+00:00 Update dependencies - - - - - e06842f5 by Simon Hengel at 2012-12-07T20:58:05+01:00 Bump version - - - - - e3dbede0 by Simon Hengel at 2012-12-07T20:58:05+01:00 Add missing test files to cabal file (fixes haskell/haddock#230) - - - - - ee0dcca7 by Simon Hengel at 2012-12-07T20:58:05+01:00 Update CHANGES - - - - - 51601bdb by Simon Peyton Jones at 2012-12-19T17:28:35+00:00 Track changes in UNPACK pragma stuff - - - - - f2573bc1 by Richard Eisenberg at 2012-12-21T20:56:25-05:00 Implement overlapping type family instances. An ordered, overlapping type family instance is introduced by 'type instance where', followed by equations. See the new section in the user manual (7.7.2.2) for details. The canonical example is Boolean equality at the type level: type family Equals (a :: k) (b :: k) :: Bool type instance where Equals a a = True Equals a b = False A branched family instance, such as this one, checks its equations in order and applies only the first the matches. As explained in the note [Instance checking within groups] in FamInstEnv.lhs, we must be careful not to simplify, say, (Equals Int b) to False, because b might later unify with Int. This commit includes all of the commits on the overlapping-tyfams branch. SPJ requested that I combine all my commits over the past several months into one monolithic commit. The following GHC repos are affected: ghc, testsuite, utils/haddock, libraries/template-haskell, and libraries/dph. Here are some details for the interested: - The definition of CoAxiom has been moved from TyCon.lhs to a new file CoAxiom.lhs. I made this decision because of the number of definitions necessary to support BranchList. - BranchList is a GADT whose type tracks whether it is a singleton list or not-necessarily-a-singleton-list. The reason I introduced this type is to increase static checking of places where GHC code assumes that a FamInst or CoAxiom is indeed a singleton. This assumption takes place roughly 10 times throughout the code. I was worried that a future change to GHC would invalidate the assumption, and GHC might subtly fail to do the right thing. By explicitly labeling CoAxioms and FamInsts as being Unbranched (singleton) or Branched (not-necessarily-singleton), we make this assumption explicit and checkable. Furthermore, to enforce the accuracy of this label, the list of branches of a CoAxiom or FamInst is stored using a BranchList, whose constructors constrain its type index appropriately. I think that the decision to use BranchList is probably the most controversial decision I made from a code design point of view. Although I provide conversions to/from ordinary lists, it is more efficient to use the brList... functions provided in CoAxiom than always to convert. The use of these functions does not wander far from the core CoAxiom/FamInst logic. BranchLists are motivated and explained in the note [Branched axioms] in CoAxiom.lhs. - The CoAxiom type has changed significantly. You can see the new type in CoAxiom.lhs. It uses a CoAxBranch type to track branches of the CoAxiom. Correspondingly various functions producing and consuming CoAxioms had to change, including the binary layout of interface files. - To get branched axioms to work correctly, it is important to have a notion of type "apartness": two types are apart if they cannot unify, and no substitution of variables can ever get them to unify, even after type family simplification. (This is different than the normal failure to unify because of the type family bit.) This notion in encoded in tcApartTys, in Unify.lhs. Because apartness is finer-grained than unification, the tcUnifyTys now calls tcApartTys. - CoreLinting axioms has been updated, both to reflect the new form of CoAxiom and to enforce the apartness rules of branch application. The formalization of the new rules is in docs/core-spec/core-spec.pdf. - The FamInst type (in types/FamInstEnv.lhs) has changed significantly, paralleling the changes to CoAxiom. Of course, this forced minor changes in many files. - There are several new Notes in FamInstEnv.lhs, including one discussing confluent overlap and why we're not doing it. - lookupFamInstEnv, lookupFamInstEnvConflicts, and lookup_fam_inst_env' (the function that actually does the work) have all been more-or-less completely rewritten. There is a Note [lookup_fam_inst_env' implementation] describing the implementation. One of the changes that affects other files is to change the type of matches from a pair of (FamInst, [Type]) to a new datatype (which now includes the index of the matching branch). This seemed a better design. - The TySynInstD constructor in Template Haskell was updated to use the new datatype TySynEqn. I also bumped the TH version number, requiring changes to DPH cabal files. (That's why the DPH repo has an overlapping-tyfams branch.) - As SPJ requested, I refactored some of the code in HsDecls: * splitting up TyDecl into SynDecl and DataDecl, correspondingly changing HsTyDefn to HsDataDefn (with only one constructor) * splitting FamInstD into TyFamInstD and DataFamInstD and splitting FamInstDecl into DataFamInstDecl and TyFamInstDecl * making the ClsInstD take a ClsInstDecl, for parallelism with InstDecl's other constructors * changing constructor TyFamily into FamDecl * creating a FamilyDecl type that stores the details for a family declaration; this is useful because FamilyDecls can appear in classes but other decls cannot * restricting the associated types and associated type defaults for a * class to be the new, more restrictive types * splitting cid_fam_insts into cid_tyfam_insts and cid_datafam_insts, according to the new types * perhaps one or two more that I'm overlooking None of these changes has far-reaching implications. - The user manual, section 7.7.2.2, is updated to describe the new type family instances. - - - - - f788d0fb by Simon Peyton Jones at 2012-12-23T15:49:58+00:00 Track changes in HsBang - - - - - ca460a0c by Simon Peyton Jones at 2012-12-23T15:50:28+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - f078fea6 by Simon Peyton Jones at 2013-01-02T08:33:13+00:00 Use InstEnv.instanceSig rather than instanceHead (name change) - - - - - 88e41305 by Simon Peyton Jones at 2013-01-14T17:10:27+00:00 Track change to HsBang type - - - - - e1ad4e19 by Kazu Yamamoto at 2013-02-01T11:59:24+09:00 Merge branch 'ghc-7.6' into ghc-7.6-merge-2 Conflicts: haddock.cabal src/Haddock/Interface/AttachInstances.hs src/Haddock/Interface/Create.hs src/Haddock/Interface/LexParseRn.hs src/Haddock/InterfaceFile.hs src/Haddock/Types.hs Only GHC HEAD can compile this. GHC 7.6.x cannot compile this. Some test fail. - - - - - 62bec012 by Kazu Yamamoto at 2013-02-06T11:12:28+09:00 Using tcSplitSigmaTy in instanceHead' (FIXME is resolved.) - - - - - 013fd2e4 by Kazu Yamamoto at 2013-02-06T17:56:21+09:00 Refactoring instanceHead'. - - - - - 3148ce0e by Kazu Yamamoto at 2013-02-07T17:45:10+09:00 Using new syntax in html-test/src/GADTRecords.hs. - - - - - 626dabe7 by Gabor Greif at 2013-02-15T22:42:01+01:00 Typo - - - - - 1eb667ae by Ian Lynagh at 2013-02-16T17:02:07+00:00 Follow changes in base - - - - - 3ef8253a by Ian Lynagh at 2013-03-01T23:23:57+00:00 Follow changes in GHC's build system - - - - - 1a265a3c by Ian Lynagh at 2013-03-03T23:12:07+00:00 Follow changes in GHC build system - - - - - 69941c79 by Max Bolingbroke at 2013-03-10T09:38:28-07:00 Use Alex 3's Unicode support to properly lex source files as UTF-8 Signed-off-by: David Waern <david.waern at gmail.com> - - - - - ea687dad by Simon Peyton Jones at 2013-03-15T14:16:10+00:00 Adapt to tcRnGetInfo returning family instances too This API change was part of the fix to Trac haskell/haddock#4175. But it offers new information to Haddock: the type-family instances, as well as the class instances, of this type. This patch just drops the new information on the floor, but there's an open opportunity to use it in the information that Haddock displays. - - - - - 971a30b0 by Andreas Voellmy at 2013-05-19T20:47:39+01:00 Fix for haskell/haddock#7879. Changed copy of utils/haddock/html/resources/html to use "cp -RL" rather than "cp -R". This allows users to run validate in a build tree, where the build tree was setup using lndir with a relative path to the source directory. - - - - - 31fb7694 by Ian Lynagh at 2013-05-19T20:47:49+01:00 Use "cp -L" when making $(INPLACE_LIB)/latex too - - - - - e9952233 by Simon Hengel at 2013-06-01T18:06:50+02:00 Add -itest to .ghci - - - - - b06873b3 by Mateusz Kowalczyk at 2013-06-01T18:06:50+02:00 Workaround for a failing build with --enable-tests. - - - - - e7858d16 by Simon Hengel at 2013-06-01T19:29:28+02:00 Fix broken test - - - - - 0690acb1 by Richard Eisenberg at 2013-06-21T14:08:25+01:00 Updates to reflect changes in HsDecls to support closed type families. - - - - - 7fd347ec by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 53ed81b6 by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 931c4f4f by Richard Eisenberg at 2013-07-24T13:15:59+01:00 Remove (error "synifyKind") to use WithinType, to allow haddock to process base. - - - - - 55a9c804 by Richard Eisenberg at 2013-08-02T15:54:55+01:00 Changes to reflect changes in GHC's type HsTyVarBndr - - - - - b6e9226c by Mathieu Boespflug at 2013-08-04T10:39:43-07:00 Output Copright and License keys in Xhtml backend. This information is as relevant in the documentation as it is in the source files themselves. Signed-off-by: David Waern <david.waern at gmail.com> - - - - - 4c66028a by David Waern at 2013-08-04T15:27:36-07:00 Bump interface file version. - - - - - 67340163 by David Waern at 2013-08-09T16:12:51-07:00 Update tests. - - - - - 2087569b by Mateusz Kowalczyk at 2013-08-25T09:24:13+02:00 Add spec tests. This adds tests for all elements we can create during regular parsing. This also adds tests for text with unicode in it. - - - - - 97f36a11 by Mateusz Kowalczyk at 2013-08-27T06:59:12+01:00 Fix ticket haskell/haddock#247. I do the same thing that the XHTML backend does: give these no special treatment and just act as if they are regular functions. - - - - - 60681b4f by Mateusz Kowalczyk at 2013-08-27T21:22:48+02:00 LaTeX tests setup - - - - - fa4c27b2 by Mateusz Kowalczyk at 2013-09-02T23:21:43+01:00 Fixes haskell/haddock#253 - - - - - 1a202490 by Mateusz Kowalczyk at 2013-09-03T01:12:50+01:00 Use Hspec instead of nanospec This is motivated by the fact that Haddock tests are not ran by the GHC's ‘validate’ script so we're pretty liberal on dependencies in that area. Full Hspec gives us some nice features such as Quickcheck integration. - - - - - 8cde3b20 by David Luposchainsky at 2013-09-08T07:27:28-05:00 Fix AMP warnings Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - d10661f2 by Herbert Valerio Riedel at 2013-09-11T15:15:01+02:00 Update Git repo URL in `.cabal` file - - - - - 16a44eb5 by Richard Eisenberg at 2013-09-17T09:34:26-04:00 Revision to reflect new role annotation syntax in GHC. - - - - - 4b9833b9 by Herbert Valerio Riedel at 2013-09-18T10:15:28+02:00 Add missing `traverse` method for `GenLocated` As `Traversable` needs at least one of `traverse` or `sequenceA` to be overridden. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - b71fed5d by Simon Hengel at 2013-09-18T22:43:34+02:00 Add test helper - - - - - 4fc1ea86 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#231 - - - - - 435872f6 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#256 We inject -dynamic-too into flags before we run all our actions in the GHC monad. - - - - - b8b24abb by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Add new field to DynFlags - - - - - 49558795 by Simon Hengel at 2013-09-18T22:43:35+02:00 Fallback to ./resources when Cabal data is not found (so that themes are found during development) - - - - - bf79d05c by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Fixes haskell/haddock#5 - - - - - e1baebc2 by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Print missing documentation. Fixes haskell/haddock#258. - - - - - 02ea74de by Austin Seipp at 2013-10-09T10:52:22-05:00 Don't consider StaticFlags when parsing arguments. Instead, discard any static flags before parsing the command line using GHC's DynFlags parser. See http://ghc.haskell.org/trac/ghc/ticket/8276 Based off a patch from Simon Hengel. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 704fd5bb by Simon Hengel at 2013-11-09T00:15:13+01:00 Update HTML tests - - - - - f9fed49e by Simon Hengel at 2013-11-10T18:43:58+01:00 Bump version - - - - - 97ae1999 by Simon Peyton Jones at 2013-11-25T17:25:14+00:00 Track changes in HsSpliceTy data constructor - - - - - 59ad8268 by Simon Peyton Jones at 2014-01-10T18:17:43+00:00 Adapt to small change in Pretty's exports - - - - - 8b12e6aa by Simon Hengel at 2014-01-12T14:48:35-06:00 Some code simplification by using traverse - - - - - fc5ea9a2 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix warnings in test helper - - - - - 6dbb3ba5 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Add ByteString version of Attoparsec - - - - - 968d7774 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 One pass parser and tests. We remove the HTML test as it is no longer necessary. We cover the test case in spec tests and other HTML tests but keeping this around fails: this is because the new parser has different semantics there. In fact, I suspect the original behaviour was a bug that wasn't caught/fixed but simply included as-is during the testing. - - - - - 37a07c9c by Simon Hengel at 2014-01-12T14:48:35-06:00 Rename Haddock.ParseSpec to Haddock.ParserSpec - - - - - f0f68fe9 by Simon Hengel at 2014-01-12T14:48:35-06:00 Don't append newline to parseString input We also check that we have parsed everything with endOfInput. - - - - - 95d60093 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix totality, unicode, examples, paragraph parsing Also simplify specs and parsers while we're at it. Some parsers were made more generic. This commit is a part of GHC pre-merge squash, email fuuzetsu at fuuzetsu.co.uk if you need the full commit history. - - - - - 7d99108c by Simon Hengel at 2014-01-12T14:48:35-06:00 Update acceptance tests - - - - - d1b59640 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Support for bold. Conflicts: src/Haddock/Backends/Hoogle.hs src/Haddock/Interface/Rename.hs src/Haddock/Parser.hs - - - - - 4b412b39 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Allow for headings inside function documentation. LaTeX will treat the h3-h6 headings the same as we'd have to hack the style file heavily otherwise and it would make the headings tiny anyway. Hoogle upstream said they will put in the functionality on their end. Conflicts: src/Haddock/Interface/Rename.hs src/Haddock/Types.hs test/Haddock/ParserSpec.hs - - - - - fdcca428 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Per-module extension flags and language listing. Any extensions that are not enabled by a used language (Haskell2010 &c) will be shown. Furthermore, any implicitly enabled are also going to be shown. While we could eliminate this either by using the GHC API or a dirty hack, I opted not to: if a user doesn't want the implied flags to show, they are recommended to use enable extensions more carefully or individually. Perhaps this will encourage users to not enable the most powerful flags needlessly. Enabled with show-extensions. Conflicts: src/Haddock/InterfaceFile.hs - - - - - 368942a2 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Bump interface version There were some breaking changes over the last few patches so we bump the interface version. This causes a big problem with testing: 1. To generate cross package docs, we first need to generate docs for the package used. 2. To generate package docs with new interface version, we need to use Haddock which has the version bumped. 3. To get Haddock with the version bump, we first need to test cross package docs 4. GOTO 1 So the problem is the chicken and the egg problem. It seems that the only solution would be to generate some interface files on the fly but it is non-trivial. To run this test, you'll have to: * build Haddock without the test (make sure everything else passes) * rebuild the packages used in the test with your shiny new binary making sure they are visible to Haddock * remove the ‘_hidden’ suffix and re-run the tests Note: because the packages currently used for this test are those provided by GHC, it's probably non-trivial to just re-build them. Preferably something less tedious to rebuild should be used and something that is not subject to change. - - - - - 124ae7a9 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow for nesting of paragraphs under lists. The nesting rules are similar to Markdown's with the exception that we can not simply indent the first line of a hard wrapped indented paragraph and have it treated as if it was fully indented. The reason is differences in markup as some of our constructs care about whitespace while others just swallow everything up so it's just a lot easier to not bother with it rather than making arbitrary rules. Note that we now drop trailing for string entities inside of lists. They weren't needed and it makes the output look uniform whether we use a single or double newline between list elements. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - c7913535 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow escaping in URLs and pictures. Some tests were moved under parseString as they weren't about paragraph level markup. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - 32326680 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update documentation. - - - - - fbef6406 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update maintainer - - - - - b40e82f4 by Mateusz Kowalczyk at 2014-01-13T02:39:25-06:00 Fixes haskell/haddock#271 Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - f4eafbf8 by Gergő Érdi at 2014-01-19T15:35:16-06:00 Support for -XPatternSynonyms Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - a8939591 by Austin Seipp at 2014-01-29T08:09:04-06:00 Update CPP check for __GLASGOW_HASKELL__ Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 30d7e9d5 by Gergő Érdi at 2014-01-31T00:15:01+08:00 <+>: Don't insert a space when concatenating empty nodes - - - - - a25ccd4d by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Fix @ code blocks In cases where we had some horizontal space before the closing ‘@’, the parser would not accept the block as a code block and we'd get ugly output. - - - - - 0f67305a by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Update tests This updates tests due to Haddock Trac haskell/haddock#271 fix and due to removal of TypeHoles as an extension from GHC. - - - - - 157322a7 by Gergő Érdi at 2014-01-31T01:03:17+08:00 Handle infix vs prefix names correctly everywhere, by explicitly specifying the context The basic idea is that "a" and "+" are either pretty-printed as "a" and "(+)" or "`a`" and "+" - - - - - aa6d9685 by Mateusz Kowalczyk at 2014-01-30T17:21:50+00:00 Correct whitespace in ‘hidden’ test for <+> change - - - - - 121872f0 by Mateusz Kowalczyk at 2014-02-09T17:59:12+00:00 Document module header. Fixes Haddock Trac haskell/haddock#270. - - - - - e3253746 by Mateusz Kowalczyk at 2014-02-10T21:37:48+00:00 Insert a space between module link and description Fixes Haddock Trac haskell/haddock#277. - - - - - 771d2384 by Mateusz Kowalczyk at 2014-02-10T23:27:21+00:00 Ensure a space between type signature and ‘Source’ This is briefly related to Haddock Trac haskell/haddock#249 and employs effectively the suggested fix _but_ it doesn't actually fix the reported issue. This commit simply makes copying the full line a bit less of a pain. - - - - - 8cda9eff by nand at 2014-02-11T15:48:30+00:00 Add support for type/data families This adds support for type/data families with their respective instances, as well as closed type families and associated type/data families. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 3f22c510 by nand at 2014-02-11T15:53:50+00:00 Improve display of poly-kinded type operators This now displays them as (==) k a b c ... to mirror GHC's behavior, instead of the old (k == a) b c ... which was just wrong. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - effb2d6b by nand at 2014-02-11T15:56:50+00:00 Add test case for PatternSynonyms This just tests various stuff including poly-kinded patterns and operator patterns to make sure the rendering isn't broken. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - b38faf0d by Niklas Haas at 2014-02-13T21:53:32+00:00 Get rid of re-implementation of sortBy I have no idea what this was doing lying around here, and due to the usage of tuples it's actually slower, too. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - ac1e0413 by Mateusz Kowalczyk at 2014-02-13T23:57:16+00:00 Only warn about missing docs when docs are missing This fixes the ‘Missing documentation for…’ message for modules with 100% coverage. - - - - - cae2e36a by Niklas Haas at 2014-02-15T21:56:18+00:00 Add test case for inter-module type/data family instances These should show up in every place where the class is visible, and indeed they do right now. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 8bea5c3a by Mateusz Kowalczyk at 2014-02-19T05:11:34+00:00 Use a bespoke data type to indicate fixity This deals with what I imagine was an ancient TODO and makes it much clearer what the argument actually does rather than having the user chase down the comment. - - - - - 5b52d57c by Niklas Haas at 2014-02-22T21:31:03+01:00 Strip a single leading space from bird tracks (#201) This makes bird tracks in the form > foo > bar > bat parse as if they had been written as >foo >bar >bat ie. without the leading whitespace in front of every line. Ideally we also want to look into how leading whitespace affects code blocks written using the @ @ syntax, which are currently unaffected by this patch. - - - - - 5a1315a5 by Simon Hengel at 2014-02-22T21:55:35+01:00 Turn a source code comment into specs - - - - - 784cfe58 by Mateusz Kowalczyk at 2014-02-23T05:02:22+00:00 Update test case for lifted GADT type rendering The parsing of these seems to have been fixed by GHC folk and it now renders differently. IMHO it now renders in a better way so I'm updating the test to reflect this. - - - - - c3c88c2f by Mateusz Kowalczyk at 2014-02-23T06:37:14+00:00 Don't shadow ‘strip’. -Wall complains - - - - - 293031d8 by Niklas Haas at 2014-02-23T15:21:52+01:00 Make ImplicitParams render correctly (#260) This introduces a new precedence level for single contexts (because implicit param contexts always need parens around them, but other types of contexts don't necessarily, even when alone) - - - - - 4200842d by Niklas Haas at 2014-02-23T15:37:13+01:00 Lower precedence of equality constraints This drops them to the new precedence pREC_CTX, which makes single eqaulity constraints show up as (a ~ b) => ty, in line with GHC's rendering. Additional tests added to make sure other type operators render as intended. Current behavior matches GHC - - - - - b59e3227 by Niklas Haas at 2014-02-23T16:11:22+01:00 Add RankNTypes test case to ImplicitParams.hs This test actually tests what haskell/haddock#260 originally reported - I omitted the RankNTypes scenario from the original fix because I realized it's not relevant to the underlying issue and indeed, this renders as intended now. Still good to have more tests. - - - - - c373dbf7 by Mateusz Kowalczyk at 2014-02-24T06:09:54+00:00 Fix rendering of Contents when links are present Fixes Haddock Trac haskell/haddock#267. - - - - - 9ecb0e56 by Mateusz Kowalczyk at 2014-02-24T06:26:50+00:00 Fix wording in the docs - - - - - 4f4dcd8e by Mateusz Kowalczyk at 2014-02-27T03:00:33+00:00 Change rendering of duplicate record field docs See Haddock Trac haskell/haddock#195. We now change this behaviour to only rendering the documentation attached to the first instance of a duplicate field. Perhaps we could improve this by rendering the first instance that has documentation attached to it but for now, we'll stick with this. - - - - - ad8aa609 by Niklas Haas at 2014-03-08T09:43:26+01:00 Render fixity information Affects functions, type synonyms, type families, class names, data type names, constructors, data families, associated TFs/DFs, type synonyms, pattern synonyms and everything else I could think of. - - - - - 6a39c917 by Niklas Haas at 2014-03-09T07:43:39+01:00 Reorder topDeclElem to move the source/wiki links to the top They appear in the same position due to the float: right attribute but now they're always at the top of the box instead of at the bottom. - - - - - 2d34b3b4 by Niklas Haas at 2014-03-09T07:53:46+01:00 Use optLast instead of listToMaybe for sourceUrls/wikiUrls This lets you override them using eg. cabal haddock --haddock-options, which can come in handy if you want to use a different layout or URL for your source code links than cabal-install generates. - - - - - 0eff4624 by Niklas Haas at 2014-03-09T07:53:46+01:00 Differentiate between TH splices (line-links) and regular names This adds a new type of source code link, to a specific line rather than a specific declaration/name - this is used to link to the location of a TH splice that defines a certain name. Rather hefty changes throughout and still one unresolved issue (the line URLs aren't parsed from the third form of --read-interface which means they're currently restricted to same-interface links). Not sure if this issue is really worth all the hassle, especially since we could just use line links in general. This commit also contains some cleanup/clarification of the types in Haddock.Backends.Xhtml.Decl and shortens some overlong lines in the process. Notably, the Bool parameter was replaced by a Unicode type synonym to help clarify its presence in type signatures. - - - - - 66d6f77b by Niklas Haas at 2014-03-09T20:02:43+01:00 Group similar fixities together Identical fixities declared for the same line should now render using syntax like: infix 4 <, >=, >, <= - - - - - 6587f9f5 by Mateusz Kowalczyk at 2014-03-10T04:24:18+00:00 Update changelog - - - - - 7387ddad by Niklas Haas at 2014-03-11T10:26:04+01:00 Include fixity information in the Interface file This resolves fixity information not appearing across package borders. The binary file version has been increased accordingly. - - - - - ab46ef44 by Niklas Haas at 2014-03-11T10:26:04+01:00 Update changelog - - - - - 565cab6f by Niklas Haas at 2014-03-11T10:26:04+01:00 Update appearance of fixity annotations This moves them in-line with their corresponding lines, similar to a presentation envision by @hvr and described in #ghc. Redundant operator names are also omitted when no ambiguity is present. - - - - - 5d7afd67 by Niklas Haas at 2014-03-11T10:26:05+01:00 Filter family instances of hidden types Currently, this check does not extend to hidden right hand sides, although it probably should hide them in that case. - - - - - ec291b0c by Niklas Haas at 2014-03-11T10:26:05+01:00 Add documentation for --source-entity-line - - - - - 0922e581 by Niklas Haas at 2014-03-11T10:37:32+01:00 Revert "Reorder topDeclElem to move the source/wiki links to the top" This reverts commit 843c42c4179526a2ad3526e4c7d38cbf4d50001d. This change is no longer needed with the new rendering style, and it messes with copy/pasting lines. - - - - - 30618e8b by Mateusz Kowalczyk at 2014-03-11T09:41:07+00:00 Bump version to 2.15.0 - - - - - adf3f1bb by Mateusz Kowalczyk at 2014-03-11T09:41:09+00:00 Fix up some whitespace - - - - - 8905f57d by Niklas Haas at 2014-03-13T19:18:06+00:00 Hide RHS of TFs with non-exported right hand sides Not sure what to do about data families yet, since technically it would not make a lot of sense to display constructors that cannot be used by the user. - - - - - 5c44d5c2 by Niklas Haas at 2014-03-13T19:18:08+00:00 Add UnicodeSyntax alternatives for * and -> I could not find a cleaner way to do this other than checking for string equality with the given built-in types. But seeing as it's actually equivalent to string rewriting in GHC's implementation of UnicodeSyntax, it's probably fitting. - - - - - b04a63e6 by Niklas Haas at 2014-03-13T19:18:10+00:00 Display minimal complete definitions for type classes This corresponds to the new {-# MINIMAL #-} pragma present in GHC 7.8+. I also cleaned up some of the places in which ExportDecl is used to make adding fields easier in the future. Lots of test cases have been updated since they now render with minimality information. - - - - - a4a20b16 by Niklas Haas at 2014-03-13T19:18:12+00:00 Strip links from recently added html tests These were accidentally left there when the tests were originally added - - - - - d624f315 by Mateusz Kowalczyk at 2014-03-13T19:19:31+00:00 Update changelog - - - - - d27a21ac by Mateusz Kowalczyk at 2014-03-13T21:19:07+00:00 Always read in prologue files as UTF8 (#286). - - - - - 54b2fd78 by Mateusz Kowalczyk at 2014-03-13T21:28:09+00:00 Style only - - - - - fa4fe650 by Simon Hengel at 2014-03-15T09:04:18+01:00 Add Fuuzetsu maintainers field in cabal file - - - - - f83484b7 by Niklas Haas at 2014-03-15T18:20:24+00:00 Hide minimal definition for only-method classes Previously this was not covered by the All xs check since here it is not actually an All, rather a single Var n. This also adds the previously missing html-test/src/Minimal.hs. - - - - - 0099d276 by Niklas Haas at 2014-03-15T18:20:26+00:00 Fix issue haskell/haddock#281 This is a regression from the data family instances change. Data instances are now distinguished from regular lists by usage of the new class "inst", and the style has been updated to only apply to those. I've also updated the appropriate test case to test this a bit better, including GADT instances with GADT-style records. - - - - - 1f9687bd by Mateusz Kowalczyk at 2014-03-21T17:48:37+00:00 Please cabal sdist - - - - - 75542693 by Mateusz Kowalczyk at 2014-03-22T16:36:16+00:00 Drop needless --split-objs which slows us down. Involves tiny cleanup of all the dynflag bindings. Fixes haskell/haddock#292. - - - - - 31214dc3 by Herbert Valerio Riedel at 2014-03-23T18:01:01+01:00 Fix a few typos Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 0b73e638 by Mateusz Kowalczyk at 2014-03-31T05:34:36+01:00 Print kind signatures on GADTs - - - - - 2bab42f3 by Mateusz Kowalczyk at 2014-03-31T16:53:25+01:00 Add default for new PlatformConstraints field - - - - - 42647c5f by Mateusz Kowalczyk at 2014-03-31T18:29:04+01:00 Drop leading whitespace in @-style blocks. Fixes haskell/haddock#201. - - - - - 98208294 by Niklas Haas at 2014-03-31T20:09:58+02:00 Crash when exporting record selectors of data family instances This fixes bug haskell/haddock#294. This also fixes a related but never-before-mentioned bug about the display of GADT record selectors with non-polymorphic type signatures. Note: Associated data type constructors fail to show up if nothing is exported that they could be attached to. Exporting any of the data types in the instance head, or the class + data family itself, causes them to show up, but in the absence of either of these, exporting just the associated data type with the constructor itself will result in it being hidden. The only scenario I can come up that would involve this kind of situation involved OverlappingInstances, and even then it can be mitigated by just exporting the class itself, so I'm not going to solve it since the logic would most likely be very complicated. - - - - - 3832d171 by Mateusz Kowalczyk at 2014-04-01T19:07:33+01:00 Make CHANGES consistent with what's now in 2.14.2 - - - - - c386ae89 by Mateusz Kowalczyk at 2014-04-01T19:18:36+01:00 Actually bundle extra spec tests in sdist - - - - - bd57a6d3 by Mateusz Kowalczyk at 2014-04-03T21:13:48+01:00 Update test cases for GHC bug haskell/haddock#8945, Haddock haskell/haddock#188 The order of signature groups has been corrected upstream. Here we add a test case and update some existing test-cases to reflect this change. We remove grouped signature in test cases that we can (Minimal, BugDeprecated &c) so that the test is as self-contained as possible. - - - - - 708b88b1 by Mateusz Kowalczyk at 2014-04-03T21:16:07+01:00 Enforce strict GHC version in cabal file This stops people with 7.6.3 trying to install 2.15.x which clearly won't work. Unfortunately we shipped 2.14.x without realising this. - - - - - 60334f7c by Mateusz Kowalczyk at 2014-04-03T21:19:24+01:00 Initialise some new PlatformConstants fields - - - - - ea77f668 by Mateusz Kowalczyk at 2014-04-11T16:52:23+01:00 We don't actually want unicode here - - - - - 0b651cae by Mateusz Kowalczyk at 2014-04-11T18:13:30+01:00 Parse identifiers with ^ and ⋆ in them. Fixes haskell/haddock#298. - - - - - e8ad0f5f by Mateusz Kowalczyk at 2014-04-11T18:47:41+01:00 Ignore version string during HTML tests. - - - - - de489089 by Mateusz Kowalczyk at 2014-04-11T18:59:30+01:00 Update CHANGES to follow 2.14.3 - - - - - beb464a9 by Gergő Érdi at 2014-04-13T16:31:10+08:00 remove Origin flag from LHsBindsLR - - - - - cb16f07c by Herbert Valerio Riedel at 2014-04-21T17:16:50+02:00 Replace local `die` by new `System.Exit.die` Starting with GHC 7.10, System.Exit exports the new `die` which is essentially the same as Haddock.Util.die, so this commit changes Haddock.Util.die to be a simple re-export of System.Exit.die. See also https://ghc.haskell.org/trac/ghc/ticket/9016 for more details. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 9b9b23c7 by Mateusz Kowalczyk at 2014-05-03T15:40:11+02:00 Disambiguate ‘die’ in test runners. - - - - - 5d28a2b8 by Mateusz Kowalczyk at 2014-05-05T09:19:49+02:00 Prepare modules for parser split. We have to generalise the Doc (now DocH) slightly to remove the dependency on GHC-supplied type. - - - - - d3967ff3 by Mateusz Kowalczyk at 2014-05-05T11:00:41+02:00 Move parser + parser tests out to own package. We move some types out that are necessary as well and then re-export and specialise them in the core Haddock. Reason for moving out spec tests is that if we're working on the parser, we can simply work on that and we can ignore the rest of Haddock. The downside is that it's a little inconvenient if at the end of the day we want to see that everything passes. - - - - - 522a448d by Mateusz Kowalczyk at 2014-05-05T11:14:47+02:00 Move out Show and Eq instances to Types They are much more useful to the users here. - - - - - 11a6f0f2 by Mateusz Kowalczyk at 2014-05-06T13:50:31+02:00 Remove no longer necessary parser error handling. We can now drop some Maybe tests and even lets us strip an error handling monad away in a few places. - - - - - 6992c924 by Mateusz Kowalczyk at 2014-05-14T02:23:55+02:00 Please the GHC build-system. As I can not figure out how to do this properly, if we're in GHC tree, we treat the library as being the same package. If we're not in the tree, we require that the library be installed separately. - - - - - 7a8ad763 by Mateusz Kowalczyk at 2014-05-14T14:50:25+02:00 Update issue tracker URL - - - - - f616c521 by Mateusz Kowalczyk at 2014-05-14T14:53:32+02:00 Update issue tracker URL for haddock-library - - - - - 66580ded by Gergő Érdi at 2014-05-25T14:24:16+08:00 Accomodate change in PatSyn representation - - - - - 0e43b988 by Mateusz Kowalczyk at 2014-05-29T03:15:29+02:00 Revert "Accomodate change in PatSyn representation" This reverts commit 57aa591362d7c8ba21285fccd6a958629a422091. I am reverting this because I pushed it to master when it was meant to stay on a wip-branch. Sorry Gergo and everyone who had trouble due to this. - - - - - e10d7ec8 by Mateusz Kowalczyk at 2014-05-29T03:24:11+02:00 Revert "Revert "Accomodate change in PatSyn representation"" This reverts commit e110e6e70e40eed06c06676fd2e62578da01d295. Apparently as per GHC commit ac2796e6ddbd54c5762c53e2fcf29f20ea162fd5 this was actually intended. Embarrasing for me. - - - - - 5861aca9 by Mateusz Kowalczyk at 2014-06-05T19:49:27+02:00 Clear up highlighting of identifiers with ‘'’s. - - - - - d7cc420f by Simon Peyton Jones at 2014-06-06T12:41:09+01:00 Follow change in patSynSig - - - - - 938b4fd8 by Mateusz Kowalczyk at 2014-06-12T07:24:29+02:00 Slightly update the readme. Style-sheets are no longer a recent thing, dead links, old maintainers, different formats. - - - - - c7799dea by Mateusz Kowalczyk at 2014-06-18T00:05:56+02:00 Update cabal files Update repository urls, use subdir property for haddock-library and use a separate versioning scheme for haddock-library in preparation for release. - - - - - a2750b6a by Simon Hengel at 2014-06-18T11:01:18+08:00 Compatibility with older versions of base and bytestring - - - - - 009b4b03 by Simon Hengel at 2014-06-18T11:14:01+08:00 Enable travis-ci for haddock-library - - - - - 9b5862eb by Simon Hengel at 2014-06-18T11:14:01+08:00 haddock-library: Do not depend on haddock-library in test suite I think you either add src to hs-source-dirs or the library to build-depends. But doing both does not make sense (AFAICT). - - - - - fb1f3279 by Simon Hengel at 2014-06-18T11:49:05+08:00 haddock-library: Use -Wall for specs - - - - - 649340e1 by Mateusz Kowalczyk at 2014-06-18T06:58:54+02:00 Use Travis with multiple GHC versions When using HEAD, we build haddock-library directly from repository as a dependency (and thanks to --enable-tests, the tests get ran anyway). In all other cases, we manually run the tests on haddock-library only and don't test the main project. - - - - - d7eeeec2 by Mateusz Kowalczyk at 2014-06-18T07:49:04+02:00 Comment improvements + few words in cabal file - - - - - 0f8db914 by Simon Hengel at 2014-06-18T13:52:23+08:00 Use doctest to check examples in documentation - - - - - 2888a8dc by Simon Hengel at 2014-06-18T14:16:48+08:00 Remove doctest dependency (so that we can use haddock-library with doctest) - - - - - 626d5e85 by Mateusz Kowalczyk at 2014-06-18T08:41:25+02:00 Travis tweaks - - - - - 41d4f9cc by Mateusz Kowalczyk at 2014-06-18T08:58:43+02:00 Don't actually forget to install specified GHC. - - - - - c6aa512a by John MacFarlane at 2014-06-18T10:43:57-07:00 Removed reliance on LambdaCase (which breaks build with ghc 7.4). - - - - - b9b93b6f by John MacFarlane at 2014-06-18T10:54:56-07:00 Fixed haddock warnings. - - - - - a41b0ab5 by Mateusz Kowalczyk at 2014-06-19T01:20:10+02:00 Update Travis, bump version - - - - - 864bf62a by Mateusz Kowalczyk at 2014-06-25T10:36:54+02:00 Fix anchors. Closes haskell/haddock#308. - - - - - 53df91bb by Mateusz Kowalczyk at 2014-06-25T15:04:49+02:00 Drop DocParagraph from front of headers I can not remember why they were wrapped in paragraphs to begin with and it seems unnecessary now that I test it. Closes haskell/haddock#307. - - - - - 29b5f2fa by Mateusz Kowalczyk at 2014-06-25T15:17:20+02:00 Don't mangle append order for nested lists. The benefit of this is that the ‘top-level’ element of such lists is properly wrapped in <p> tags so any CSS working with these will be applied properly. It also just makes more sense. Pointed out at jgm/pandoc#1346. - - - - - 05cb6e9c by Mateusz Kowalczyk at 2014-06-25T15:19:45+02:00 Bump haddock-library to 1.1.0 for release - - - - - 70feab15 by Iavor Diatchki at 2014-07-01T03:37:07-07:00 Propagate overloading-mode for instance declarations in haddock (#9242) - - - - - d4ca34a7 by Simon Peyton Jones at 2014-07-14T16:23:15+01:00 Adapt to new definition of HsDecls.TyFamEqn This is a knock-on from the refactoring from Trac haskell/haddock#9063. I'll push the corresponding changes to GHC shortly. - - - - - f91e2276 by Edward Z. Yang at 2014-07-21T08:14:19-07:00 Track GHC PackageId to PackageKey renaming. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - b010f9ef by Edward Z. Yang at 2014-07-25T16:28:46-07:00 Track changes for module reexports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - 8b85f9f9 by Mateusz Kowalczyk at 2014-07-28T13:25:43+02:00 Catch mid-line URLs. Fixes haskell/haddock#314. - - - - - 4c613a78 by Edward Z. Yang at 2014-08-05T03:11:00-07:00 Track type signature change of lookupModuleInAllPackages Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - e80b051c by Edward Z. Yang at 2014-08-05T17:34:26+01:00 If GhcProfiled, also build Haddock profiled. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f9cccd29 by Edward Z. Yang at 2014-08-07T14:23:35+01:00 Ignore TAGS files. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 00b3af52 by Mateusz Kowalczyk at 2014-08-08T04:58:19+02:00 Update to attoparsec-0.12.1.1 There seems to be memory and speed improvement. - - - - - 5457dc71 by Mateusz Kowalczyk at 2014-08-08T18:24:02+02:00 Fix forgotten src - - - - - 3520cb04 by Mateusz Kowalczyk at 2014-08-14T20:19:07+01:00 Bump down the version for master to 2.14.4 - - - - - dc98c21b by Mateusz Kowalczyk at 2014-08-14T20:23:27+01:00 Revert "Track type signature change of lookupModuleInAllPackages" This reverts commit d59fec2c9551b5662a3507c0011e32a09a9c118f. - - - - - 3f2038c0 by Mateusz Kowalczyk at 2014-08-14T20:23:31+01:00 Revert "Track changes for module reexports." This reverts commit b99b57c0df072d12b67816b45eca2a03cb1da96d. - - - - - 56d4e49e by Mateusz Kowalczyk at 2014-08-14T20:23:42+01:00 Revert "Track GHC PackageId to PackageKey renaming." This reverts commit 8ac42d3327473939c013551750425cac191ff0fd. - - - - - 726ea3cb by Mateusz Kowalczyk at 2014-08-14T20:23:47+01:00 Revert "Adapt to new definition of HsDecls.TyFamEqn" This reverts commit cb96b4f1ed0462b4a394b9fda6612c3bea9886bd. - - - - - 61a88ff0 by Mateusz Kowalczyk at 2014-08-14T20:23:52+01:00 Revert "Propagate overloading-mode for instance declarations in haddock (#9242)" This reverts commit 8d20ca8d5a9bee73252ff2035ec45f9c03d0820c. - - - - - a32ba674 by Mateusz Kowalczyk at 2014-08-14T20:26:03+01:00 Revert "Disambiguate ‘die’ in test runners." This reverts commit dba02d6df32534aac5d257f2d28596238d248942. - - - - - f335820f by Mateusz Kowalczyk at 2014-08-14T20:26:09+01:00 Revert "Replace local `die` by new `System.Exit.die`" This reverts commit 08aa509ebac58bfb202ea79c7c41291ec280a1c5. - - - - - 107078e4 by Mateusz Kowalczyk at 2014-08-14T20:27:34+01:00 Merge branch 'reverts' This reverts any changes that were made to have Haddock compile with 7.9. When 7.10 release comes, we can simply re-apply all the patches and any patches that occur on ghc-head branch from now on. This allows us to build master with 7.8.3 - - - - - b44b3871 by Mateusz Kowalczyk at 2014-08-15T02:47:40+01:00 Fix haskell/haddock#313 by doing some list munging. I get rid of the Monoid instance because we weren't satisfying the laws. Convenience of having <> didn't outweigh the shock-factor of having it behave badly. - - - - - e1a62cde by Mateusz Kowalczyk at 2014-08-15T02:52:56+01:00 Stop testing haskell/haddock#188. Because the change is in GHC 7.9 and we now work against 7.8.3, this test no longer makes sense. We revert it until 7.10 becomes the standard version. If anything, there should be a test for this in GHC itself. - - - - - 54e8286d by Mateusz Kowalczyk at 2014-08-15T05:31:57+01:00 Add haskell/haddock#313 to CHANGES - - - - - 9df7ad5d by Simon Hengel at 2014-08-20T11:25:32+08:00 Fix warning - - - - - ee2574d6 by Simon Hengel at 2014-08-20T12:07:01+08:00 Fix travis builds - - - - - 384cf2e6 by Simon Hengel at 2014-08-20T12:14:31+08:00 Require GHC 7.8.3 - - - - - d4779863 by Simon Hengel at 2014-08-22T12:14:16+08:00 Move Haddock API to a separate package - - - - - 80f3e0e1 by Simon Hengel at 2014-08-22T14:57:38+08:00 Bump version to 2.15.0 and add version constraints - - - - - 309a94ce by Simon Hengel at 2014-08-22T15:18:06+08:00 Add deprecated compatibility module - - - - - 4d1e4e3f by Luite Stegeman at 2014-08-22T20:46:45+02:00 export things to allow customizing how the Ghc session is run - - - - - 47884591 by Luite Stegeman at 2014-08-22T20:46:51+02:00 ghc 7.8.2 compatibility - - - - - 5ea94e2c by Luite Stegeman at 2014-08-22T22:08:58+02:00 install dependencies for haddock-api on travis - - - - - 9fb845b2 by Mateusz Kowalczyk at 2014-08-23T10:09:34+01:00 Move sources under haddock-api/src - - - - - 85817dc4 by Mateusz Kowalczyk at 2014-08-23T10:10:48+01:00 Remove compat stuff - - - - - 151c6169 by Niklas Haas at 2014-08-24T08:14:10+02:00 Fix extra whitespace on signatures and update all test cases This was long overdue, now running ./accept.lhs on a clean test from master will not generate a bunch of changes. - - - - - d320e0d2 by Niklas Haas at 2014-08-24T08:14:35+02:00 Omit unnecessary foralls and fix haskell/haddock#315 This also fixes haskell/haddock#86. - - - - - bdafe108 by Mateusz Kowalczyk at 2014-08-24T15:06:46+01:00 Update CHANGES - - - - - fafa6d6e by Mateusz Kowalczyk at 2014-08-24T15:14:23+01:00 Delete few unused/irrelevant/badly-place files. - - - - - 3634923d by Duncan Coutts at 2014-08-27T13:49:31+01:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. - - - - - 539a7e70 by Herbert Valerio Riedel at 2014-08-31T11:36:32+02:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 9e3a0e5b by Mateusz Kowalczyk at 2014-08-31T12:54:43+01:00 Bump version in doc - - - - - 4a177525 by Mateusz Kowalczyk at 2014-08-31T13:01:23+01:00 Bump haddock-library version - - - - - f99c1384 by Mateusz Kowalczyk at 2014-08-31T13:05:25+01:00 Remove references to deleted files - - - - - 5e51a247 by Mateusz Kowalczyk at 2014-08-31T14:18:44+01:00 Make the doc parser not complain - - - - - 2cedb49a by Mateusz Kowalczyk at 2014-09-03T03:33:15+01:00 CONTRIBUTING file for issues - - - - - 88027143 by Mateusz Kowalczyk at 2014-09-04T00:46:59+01:00 Mention --print-missing-docs - - - - - 42f6754f by Alan Zimmerman at 2014-09-05T18:13:24-05:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e712719e by Austin Seipp at 2014-09-09T01:03:27-05:00 Fix import of 'empty' due to AMP. Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 71c29755 by Herbert Valerio Riedel at 2014-09-09T17:35:20+02:00 Bump `base` constraint for AMP - - - - - 0bf9f3ed by Mateusz Kowalczyk at 2014-09-12T19:18:32+01:00 Delete stale ANNOUNCE - - - - - cac89ee6 by Krzysztof Gogolewski at 2014-09-14T17:17:09+02:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 4d683426 by Edward Z. Yang at 2014-09-18T13:38:11-07:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 80697fd5 by Herbert Valerio Riedel at 2014-09-19T00:07:52+02:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. - - - - - c015eb70 by Herbert Valerio Riedel at 2014-09-19T00:10:36+02:00 Revert "Followup changes to addition of -fwarn-context-quantification" This reverts commit 4023817d7c0e46db012ba2eea28022626841ca9b temporarily as the respective feature hasn't landed in GHC HEAD yet, but this commit blocks later commits from being referenced in GHC HEAD. - - - - - 38ded784 by Edward Z. Yang at 2014-09-18T15:32:15-07:00 Revert "Revert "Followup changes to addition of -fwarn-context-quantification"" This reverts commit db14fd8ab4fab43694139bc203808b814eafb2dc. It's in HEAD now. - - - - - f55d59c9 by Herbert Valerio Riedel at 2014-09-26T19:18:28+02:00 Revert "Fix import of 'empty' due to AMP." This reverts commit 0cc5bc85e9fca92ab712b68a2ba2c0dd9d3d79f4 since it turns out we don't need to re-export `empty` from Control.Monad after all. - - - - - 467050f1 by David Feuer at 2014-10-09T20:07:36-04:00 Fix improper lazy IO use Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. - - - - - cc47b699 by Edward Z. Yang at 2014-10-09T21:38:19-07:00 Fix use-after-close lazy IO bug Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. Signed-off-by: David Feuer <David.Feuer at gmail.com> Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 87babcbe by Austin Seipp at 2014-10-20T20:05:27-05:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - ab259516 by Austin Seipp at 2014-10-20T20:07:01-05:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - b918093c by Mateusz Kowalczyk at 2014-10-29T03:59:39+00:00 Experimental support for collapsable headers Closes haskell/haddock#335 - - - - - 849db129 by Mateusz Kowalczyk at 2014-10-29T10:07:26+01:00 Experimental support for collapsable headers (cherry picked from commit e2ed3b9d8dfab09f1b1861dbc8e74f08e137ebcc) - - - - - a4cc4789 by Herbert Valerio Riedel at 2014-10-31T11:08:26+01:00 Collapse user-defined section by default (re haskell/haddock#335) - - - - - 9da1b33e by Yuras Shumovich at 2014-10-31T16:11:04-05:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - c625aefc by Austin Seipp at 2014-10-31T19:34:10-05:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - c7738e5e by Simon Hengel at 2014-11-02T07:25:30+08:00 Remove -fobject-code from .ghci (this slows down reloads on modifications) - - - - - d4a86e95 by Simon Hengel at 2014-11-03T09:26:11+08:00 Get rid of StandaloneDeriving - - - - - a974e311 by Simon Hengel at 2014-11-03T09:26:11+08:00 Derive more instances - - - - - 8aa0c4d7 by Simon Hengel at 2014-11-03T09:27:08+08:00 Remove unused language extensions - - - - - 3052d46a by Simon Hengel at 2014-11-03T09:30:46+08:00 Minor refactoring - - - - - 4281d3cb by Simon Hengel at 2014-11-03T09:30:46+08:00 parser: Try to parse definition lists right before text paragraphs - - - - - 8ba12bf9 by Simon Hengel at 2014-11-03T09:34:19+08:00 Add support for markdown links (closes haskell/haddock#336) - - - - - a2f8d747 by Simon Hengel at 2014-11-03T09:34:19+08:00 Allow markdown links at the beginning of a paragraph - - - - - 53b11207 by Simon Hengel at 2014-11-03T09:34:20+08:00 Update documentation - - - - - 652267c6 by Simon Hengel at 2014-11-03T09:34:20+08:00 Add support for markdown images - - - - - 9d667502 by Simon Hengel at 2014-11-03T09:34:20+08:00 Allow an optional colon after the closing bracket of definition lists This is to disambiguate them from markdown links and will be require with a future release. - - - - - 8167fc32 by Mateusz Kowalczyk at 2014-11-04T01:16:51+00:00 whitespace only - - - - - 3da62981 by Mateusz Kowalczyk at 2014-11-04T01:17:31+00:00 Fix re-exports of built-in type families Fixes haskell/haddock#310 - - - - - edc76b34 by Mateusz Kowalczyk at 2014-11-04T02:54:28+00:00 Turn some uses of error into recoverable warnings This should at the very least not abort when something weird happens. It does feel like we should have a type that carries these errors until the end however as the user might not see them unless they are printed at the end. - - - - - 0a137400 by Mateusz Kowalczyk at 2014-11-04T04:09:44+00:00 Fix warnings - - - - - d068fc21 by Mateusz Kowalczyk at 2014-11-04T21:04:07+00:00 Fix parsing of identifiers written in infix way - - - - - 1a9f2f3d by Simon Hengel at 2014-11-08T11:32:42+08:00 Minor code simplification - - - - - 6475e9b1 by Simon Hengel at 2014-11-08T17:28:33+08:00 newtype-wrap parser monad - - - - - dc1ea105 by Herbert Valerio Riedel at 2014-11-15T11:55:43+01:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - fbb1aca4 by Simon Hengel at 2014-11-16T08:51:38+08:00 State intention rather than implementation details in Haddock comment - - - - - 97851ab2 by Simon Hengel at 2014-11-16T10:20:19+08:00 (wip) Add support for @since (closes haskell/haddock#26) - - - - - 34bcd18e by Gergő Érdi at 2014-11-20T22:35:38+08:00 Update Haddock to new pattern synonym type signature syntax - - - - - 304b7dc3 by Jan Stolarek at 2014-11-20T17:48:43+01:00 Follow changes from haskell/haddock#9812 - - - - - 920f9b03 by Richard Eisenberg at 2014-11-20T16:52:50-05:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 0bfe4e78 by Alan Zimmerman at 2014-11-21T11:23:09-06:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 356ed45a by Thomas Winant at 2014-11-28T16:11:22-06:00 Support for PartialTypeSignatures - - - - - 5dc8f3b1 by Gergő Érdi at 2014-11-29T15:39:09+08:00 For pattern synonyms, render "pattern" as a keyword - - - - - fe704480 by Mateusz Kowalczyk at 2014-12-09T03:38:32+00:00 List new module in cabal file - - - - - b9ad5a29 by Mateusz Kowalczyk at 2014-12-10T00:58:24+00:00 Allow the parser to spit out meta-info Currently we only use it only for ‘since’ annotations but with these patches it should be fairly simple to add new attributes if we wish to. Closes haskell/haddock#26. It seems to work fine but due to 7.10 rush I don't have the chance to do more exhaustive testing right now. The way the meta is output (emphasis at the end of the whole comment) is fairly arbitrary and subject to bikeshedding. Note that this makes test for Bug310 fail due to interface version bump: it can't find the docs for base with this interface version so it fails. There is not much we can do to help this because it tests for ’built-in’ identifier, not something we can provide ourselves. - - - - - 765af0e3 by Mateusz Kowalczyk at 2014-12-10T01:17:19+00:00 Update doctest parts of comments - - - - - 8670272b by jpmoresmau at 2014-12-10T01:35:31+00:00 header could contain several lines Closes haskell/haddock#348 - - - - - 4f9ae4f3 by Mateusz Kowalczyk at 2014-12-12T06:22:31+00:00 Revert "Merge branch 'reverts'" This reverts commit 5c93cc347773c7634321edd5f808d5b55b46301f, reversing changes made to 5b81a9e53894d2ae591ca0c6c96199632d39eb06. Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - e974ac94 by Duncan Coutts at 2014-12-12T06:26:11+00:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. Conflicts: haddock-api/src/Haddock.hs - - - - - 2f3a2365 by Herbert Valerio Riedel at 2014-12-12T06:26:51+00:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 1dbd6390 by Alan Zimmerman at 2014-12-12T06:32:07+00:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - bb6ff1f4 by Mateusz Kowalczyk at 2014-12-12T06:35:07+00:00 Bump ‘base’ constraint Follows the similar commit made on ghc-head branch - - - - - 466fe4ab by Krzysztof Gogolewski at 2014-12-12T06:37:42+00:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 97e080c9 by Edward Z. Yang at 2014-12-12T06:39:35+00:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: haddock-api/src/Haddock/ModuleTree.hs - - - - - 20b2af56 by Herbert Valerio Riedel at 2014-12-12T06:42:50+00:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. Conflicts: haddock-library/src/Documentation/Haddock/Parser.hs - - - - - b3ad269d by Austin Seipp at 2014-12-12T06:44:14+00:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 072df0dd by Austin Seipp at 2014-12-12T06:45:01+00:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - dbb9294a by Herbert Valerio Riedel at 2014-12-12T06:46:17+00:00 Collapse user-defined section by default (re haskell/haddock#335) Conflicts: haddock-api/src/Haddock/Backends/Xhtml/DocMarkup.hs - - - - - f23ab545 by Yuras Shumovich at 2014-12-12T06:46:41+00:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - 753a4b67 by Austin Seipp at 2014-12-12T06:46:51+00:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8954e8f5 by Herbert Valerio Riedel at 2014-12-12T06:50:53+00:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - d2b06d61 by Gergő Érdi at 2014-12-12T07:07:30+00:00 Update Haddock to new pattern synonym type signature syntax Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 1ff02426 by Jan Stolarek at 2014-12-12T07:13:24+00:00 Follow changes from haskell/haddock#9812 Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - 06ad7600 by Richard Eisenberg at 2014-12-12T07:13:43+00:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 8fd2aa8b by Alan Zimmerman at 2014-12-12T07:22:25+00:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Backends/LaTeX.hs haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 95c3db98 by Thomas Winant at 2014-12-12T07:35:49+00:00 Support for PartialTypeSignatures Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs haddock-api/src/Haddock/Interface/Create.hs - - - - - 45494428 by Gergő Érdi at 2014-12-12T07:36:18+00:00 For pattern synonyms, render "pattern" as a keyword - - - - - a237e3eb by Mateusz Kowalczyk at 2014-12-12T12:27:13+00:00 Various fixups and bumps for next release - - - - - 22918bcd by Herbert Valerio Riedel at 2014-12-14T10:11:47+01:00 Remove redundant wild-card pattern match (this would otherwise cause a build-failure with `-Werror`) - - - - - 1d6ce947 by Herbert Valerio Riedel at 2014-12-14T10:17:06+01:00 Treat GHC 7.10 the same as GHC 7.9 ...since the current GHC 7.9 is going to become GHC 7.10 real-soon-now anyway - - - - - f434ea89 by Herbert Valerio Riedel at 2014-12-14T18:26:50+01:00 Fixup ghc.mk (follow-up to 1739375eb23342) This makes the GHC build-system aware of the data-files to be copied into the bindist (as haddock.cabal doesn't list those anymore) - - - - - 6fb839eb by Mateusz Kowalczyk at 2014-12-17T09:28:59+00:00 Only keep one Version instead of blindly appending - - - - - 40645489 by Mateusz Kowalczyk at 2014-12-18T07:09:44+00:00 Fix dependency version - - - - - 8b3b927b by Mateusz Kowalczyk at 2014-12-18T07:14:23+00:00 Print missing docs by default Adds --no-print-missing-docs - - - - - 59666694 by Mateusz Kowalczyk at 2014-12-18T07:21:37+00:00 update changelog - - - - - aa6d168e by Mateusz Kowalczyk at 2014-12-18T07:30:58+00:00 Update docs for @since - - - - - 2d7043ee by Luite Stegeman at 2014-12-19T18:29:35-06:00 hide projectVersion from DynFlags since it clashes with Haddock.Version.projectVersion - - - - - aaa70fc0 by Luite Stegeman at 2014-12-22T15:58:43+01:00 Add missing import for standalone haddock-api package - - - - - 9ce01269 by Herbert Valerio Riedel at 2014-12-22T17:48:45+01:00 Reset ghc-head with master's tree (this is an overwriting git merge of master into ghc-head) - - - - - fcd6fec1 by Herbert Valerio Riedel at 2014-12-22T17:51:52+01:00 Bump versions for ghc-7.11 - - - - - 525ec900 by Mateusz Kowalczyk at 2014-12-23T13:36:24+00:00 travis-ci: test with HEAD - - - - - cbf494b5 by Simon Peyton Jones at 2014-12-23T15:22:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 50e01c99 by Mateusz Kowalczyk at 2014-12-29T15:28:47+00:00 Make travis use 7.10.x - - - - - 475e60b0 by Njagi Mwaniki at 2014-12-29T15:30:44+00:00 Turn the README into GitHub Markdown format. Closes haskell/haddock#354 - - - - - 8cacf48e by Luite Stegeman at 2015-01-05T16:25:37+01:00 bump haddock-api ghc dependency to allow release candidate and first release - - - - - 6ed6cf1f by Simon Peyton Jones at 2015-01-06T16:37:47+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 8b484f33 by Simon Peyton Jones at 2015-01-08T15:50:22+00:00 Track naming change in DataCon - - - - - 23c5c0b5 by Alan Zimmerman at 2015-01-16T10:15:11-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e7a5532c by JP Moresmau at 2015-01-22T17:19:03+00:00 Ignore warnings, install Cabal 1.22 - - - - - 86942c84 by jpmoresmau at 2015-01-22T17:19:04+00:00 solve dataDir ambiguity - - - - - 5ceb743e by jpmoresmau at 2015-01-22T19:17:32+00:00 support GHC 7.10: no Safe-Inferred, Foldable instance - - - - - 6a3b3fb5 by Mateusz Kowalczyk at 2015-01-22T19:32:10+00:00 Update test files Test: a correct behaviour for fields comma-separating values. I'm surprised we had no bug open for this. Maybe it affects how haskell/haddock#301 renders now but I doubt. Operators: Seems GHC is giving us a new order for operators, something must have changed on their side again. cc @haasn , this makes the fixity to the side not match the order on the LHS which is a bit unpleasant. Maybe the fixity can be made to match the GHC order? Bug335: We expand examples by default now. Bug310: Now inferred safe. - - - - - 708f8b2f by jpmoresmau at 2015-01-22T19:36:59+00:00 Links to source location of class instance definitions - - - - - 5cf8a6da by Vincent Berthoux at 2015-01-22T19:59:58+00:00 Filter '\r' from comments due to Windows problems. On Windows this was causing newline to be rendered twice in code blocks. Closes haskell/haddock#359, fixes haskell/haddock#356. - - - - - 1749e6f0 by Mateusz Kowalczyk at 2015-01-22T20:31:27+00:00 Changelog only - - - - - c8145f90 by Mateusz Kowalczyk at 2015-01-22T23:34:05+00:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. - - - - - 14248254 by Mateusz Kowalczyk at 2015-01-22T23:43:18+00:00 Sort out some module import warnings - - - - - d8a38989 by Simon Peyton Jones at 2015-01-23T07:10:16-06:00 Track naming change in DataCon (cherry picked from commit 04cf63d0195837ed52075ed7d2676e71831e8a0b) - - - - - d3ac6ae4 by Alan Zimmerman at 2015-01-23T07:17:19-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit d61bbc75890e4eb0ad508b9c2a27b91f691213e6) - - - - - 4c1ffeb0 by Simon Peyton Jones at 2015-02-10T12:10:33+00:00 Track changes in HsSyn for quasi-quotes - - - - - 775d20f7 by Mateusz Kowalczyk at 2015-03-15T08:11:48+01:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. (cherry picked from commit 8e06728afb0784128ab2df0be7a5d7a191d30ff4) - - - - - f9245e72 by Phil Ruffwind at 2015-03-16T04:32:01-04:00 Prevent Synopsis from using up too much horizontal space When long type signatures occur in the Synopsis, the element is stretched beyond the width of the window. Scrollbars don't appear, so it's impossible to read anything when this happens. - - - - - cd8fa415 by Mateusz Kowalczyk at 2015-03-17T21:59:39+00:00 Update changelog Closes haskell/haddock#151 due to 71170fc77962f10d7d001e3b8bc8b92bfeda99bc - - - - - b5248b47 by Ben Gamari at 2015-03-25T17:12:17+00:00 Make the error encountered when a package can't be found more user-friendly Closes haskell/haddock#369 - - - - - b756b772 by Mateusz Kowalczyk at 2015-03-26T16:31:40+00:00 Remove now redundant imports - - - - - 5ea5e8dd by Mateusz Kowalczyk at 2015-03-26T16:45:52+00:00 Update test to account for \r filtering - - - - - 6539bfb3 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Test for anchor defaulting I delete the old tests because it turns out that: * test runner would never put them in scope of each other even with imports so just one would suffice * test runner actually needed some hacking to keep links so in the end we would end up with no anchors making them useless - - - - - 1a01d950 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Clearly default to variables in out of scope case - - - - - 7943abe8 by Mateusz Kowalczyk at 2015-03-27T01:14:11+00:00 Fix Hoogle display of constructors Fixes haskell/haddock#361 - - - - - 6d6e587e by Mateusz Kowalczyk at 2015-03-27T01:45:18+00:00 Fully qualify names in Hoogle instances output Closes haskell/haddock#263 - - - - - 52dac365 by Mateusz Kowalczyk at 2015-03-27T01:55:01+00:00 Update changelog - - - - - ca5af9a8 by Mateusz Kowalczyk at 2015-03-27T02:43:55+00:00 Output method documentation in Hoogle backend One thing of note is that we no longer preserve grouping of methods and print each method on its own line. We could preserve it if no documentation is present for any methods in the group if someone asks for it though. Fixes haskell/haddock#259 - - - - - a33f0c10 by Mateusz Kowalczyk at 2015-03-27T03:04:21+00:00 Don't print instance safety information in Hoogle Fixes haskell/haddock#168 - - - - - df6c935a by Mateusz Kowalczyk at 2015-03-28T00:11:47+00:00 Post-release version bumps and changelog - - - - - dde8f7c0 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Loosen bounds on haddock-* - - - - - de93bf89 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Expand response files in arguments Closes haskell/haddock#285 - - - - - 1f0b0856 by Zejun Wu at 2015-04-26T16:35:35-07:00 Do not insert anchor for section headings in contents box - - - - - 860439d7 by Simon Peyton Jones at 2015-05-01T09:36:47+01:00 Track change in API of TyCon - - - - - a32f3e5f by Adam Gundry at 2015-05-04T15:32:59+01:00 Track API changes to support empty closed type familes - - - - - 77e98bee by Ben Gamari at 2015-05-06T20:17:08+01:00 Ignore doc/haddock.{ps,pdf} - - - - - 663d0204 by Murray Campbell at 2015-05-11T04:47:37-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8bb0dcf5 by Murray Campbell at 2015-05-11T06:35:06-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit 2380f07c430c525b205ce2eae6dab23c8388d899) - - - - - bad900ea by Adam Bergmark at 2015-05-11T15:29:39+01:00 haddock-library: require GHC >= 7.4 `Data.Monoid.<>` was added in base-4.5/GHC-7.4 Closes haskell/haddock#394 Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - daceff85 by Simon Peyton Jones at 2015-05-13T12:04:21+01:00 Track the new location of setRdrNameSpace - - - - - 1937d1c4 by Alan Zimmerman at 2015-05-25T21:27:15+02:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - ee0fb6c2 by Łukasz Hanuszczak at 2015-05-27T11:51:31+02:00 Create simple method for indentation parsing. - - - - - 7d6fcad5 by Łukasz Hanuszczak at 2015-05-27T21:36:13+02:00 Make nested lists count indentation according to first item. - - - - - d6819398 by Łukasz Hanuszczak at 2015-05-27T22:46:13+02:00 Add simple test case for arbitrary-depth list nesting. - - - - - 2929c54d by Łukasz Hanuszczak at 2015-06-03T02:11:31+02:00 Add arbitrary-indent spec test for parser. - - - - - 9a0a9bb0 by Mateusz Kowalczyk at 2015-06-03T05:25:29+01:00 Update docs with info on new list nesting rule Fixes haskell/haddock#278 through commits from PR haskell/haddock#401 - - - - - 12efc92c by Mateusz Kowalczyk at 2015-06-03T05:29:26+01:00 Update some meta data at the top of the docs - - - - - 765ee49f by Bartosz Nitka at 2015-06-07T08:40:59-07:00 Add some Hacking docs for getting started - - - - - 19aaf851 by Bartosz Nitka at 2015-06-07T08:44:30-07:00 Fix markdown - - - - - 2a90cb70 by Mateusz Kowalczyk at 2015-06-08T15:08:36+01:00 Refine hacking instructions slightly - - - - - 0894da6e by Thomas Winant at 2015-06-08T23:47:28-05:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 10a9bb76 by Emanuel Borsboom at 2015-06-12T02:46:23+01:00 Build executable with '-threaded' (fixes haskell/haddock#399) - - - - - 7696b94f by Mateusz Kowalczyk at 2015-06-12T02:59:19+01:00 Update changelog for -threaded Closes haskell/haddock#400 - - - - - d3c118ec by Bartosz Nitka at 2015-06-12T03:00:58+01:00 Fix haddock: internal error: spliceURL UnhelpfulSpan (#207) Inferred type signatures don't have SrcSpans, so let's use the one from the declaration. I've tested this manually on the test-case from haskell/haddock#207, but I got stuck at trying to run the test-suite. - - - - - b67e843b by Mateusz Kowalczyk at 2015-06-12T03:01:50+01:00 Changelog for haskell/haddock#207 Fixes haskell/haddock#207, closes haskell/haddock#402 - - - - - 841d785e by jpmoresmau at 2015-06-12T16:03:16+01:00 Attach to instance location the name that has the same location file Fixes haskell/haddock#383 - - - - - 98791cae by Mateusz Kowalczyk at 2015-06-12T16:08:27+01:00 Update changelog Closes haskell/haddock#398 - - - - - 7c0b5a87 by Phil Ruffwind at 2015-06-12T13:07:25-04:00 Fix alignment of Source links in instance table in Firefox Due to a Firefox bug [1], a combination of 'whitespace: nowrap' on the parent element with 'float: right' on the inner element can cause the floated element to be displaced downwards for no apparent reason. To work around this, the left side is wrapped in its own <span> and set to 'float: left'. As a precautionary measure to prevent the parent element from collapsing entirely, we also add the classic "clearfix" hack. The latter is not strictly needed but it helps prevent bugs if the layout is altered again in the future. Fixes haskell/haddock#384. Remark: line 159 of src/Haddock/Backends/Xhtml/Layout.hs was indented to prevent confusion over the operator precedence of (<+>) vs (<<). [1]: https://bugzilla.mozilla.org/show_bug.cgi?id=488725 - - - - - cfe86e73 by Mateusz Kowalczyk at 2015-06-14T10:49:01+01:00 Update tests for the CSS changes - - - - - 2d4983c1 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding for Haskell source parser module. - - - - - 29548785 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement function for tagging parsed chunks with source spans. - - - - - 6a5e4074 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement simple string chunking based on HsColour library. - - - - - 6e52291f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create basic token classification method. - - - - - da971a27 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Adapt source span tagging to work with current whitespace handling. - - - - - 4feb5a22 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add record accessors to exports of hyperlinker parser module. - - - - - a8cc4e39 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Make parser module export all types and associated accessors. - - - - - fb8d468f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple HTML renderer for parsed source file. - - - - - 80747822 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for specifying the CSS file path in HTML source renderer. - - - - - 994dc1f5 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix identifier recognition in Haskell source parser. - - - - - b1bd0430 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix comment recognition in Haskell source parser. - - - - - 11db85ae by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for recognizing compiler pragmas in source parser. - - - - - 736c7bd3 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding of module for associating tokens with AST names. - - - - - 7e149bc2 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement utility method for extracting variable identifiers from AST. - - - - - 32eb640a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple mechanism for associating tokens with AST names. - - - - - d4eba5bc by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add dummy support for hyperlinking named tokens. - - - - - 2b76141f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matcher bug causing wrong items being hyperlinked. - - - - - 2d48002e by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Constrain elements exported by hyperlinker modules. - - - - - 9715eec6 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for type token recognition. - - - - - 8fa401cb by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for binding token recognition. - - - - - d062400b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement go-to-definition mechanism for local bindings. - - - - - f4dc229b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement module export- and import-list item hyperlinking. - - - - - c9a46d58 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matching to allow parenthesized operators hyperlinking. - - - - - 03aad95a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix weird hyperlinking of parenthesized operators. - - - - - b4694a7d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for type declaration anchors. - - - - - 7358d2d2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for top-level function declaration anchors. - - - - - dfc24b24 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix external anchors to contain HTML file extension. - - - - - a045926c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Refactor the way AST names are handled within detailed tokens. - - - - - c76049b4 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement hyperlinking of imported module names. - - - - - 2d2a1572 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix parsing of single line comments with broken up newlines. - - - - - 11afdcf2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix bug with improper newline handling. - - - - - 8137f104 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issues with escaped newlines in comments. - - - - - 34759b19 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for parsing C preprocessor macros. - - - - - 09f0f847 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for parser module of source hyperlinker. - - - - - 709a8389 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for AST module of source hyperlinker. - - - - - 4df5c227 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add command line option for generating hyperlinked source. - - - - - 7a755ea2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Extend module interface with rich source token stream field. - - - - - 494f4ab1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement source tokenization during interface creation process. - - - - - 5f21c953 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create hyperlinker module and plug it into the Haddock pipeline. - - - - - 0cc8a216 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for providing custom CSS files for hyperlinked source. - - - - - a32bbdc1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for fancy highlighting upon hovering over identifier. - - - - - d16d642a by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make source hyperlinker generate output in apropriate directory. - - - - - ae12953d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create module with hyperlinker utility functions. - - - - - 6d4952c5 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make external hyperlinks point to locations specified by source URLs. - - - - - 8417555d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Rewrite source generation to fixed links and directory structure. - - - - - ce9cec01 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add basic support for cross-package hyperlink generation. - - - - - 7eaf025c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Disable generating hyperlinks for module references. - - - - - a50bf92e by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make Haddock generate source for all interfaces (also hidden ones). - - - - - f5ae2838 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Prevent source parser from throwing exception when lexing fails. - - - - - db9ffbe0 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement workaround for Chrome highlighting issues. - - - - - 0b6b453b by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate correct anchors for data constructors. - - - - - c86d38bc by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate anchors for record field declarations. - - - - - 063abf7f by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issue with hyperlink highlight styling in Chrome browser. - - - - - 880fc611 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking constructor names in patters. - - - - - c9e89b95 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record patterns. - - - - - 17a11996 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record expressions. - - - - - 0eef932d by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Make hyperlinker respect pretty-printer flag and add documentation. - - - - - f87c1776 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Unexpose hyperlinker modules in Cabal configuration. - - - - - 4c9e2b06 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Setup HSpec framework for Haddock API package. - - - - - 4b20cb30 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add basic tests related to comment parsing. - - - - - 6842e919 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add tests related to parsing basic language constructs. - - - - - 87bffb35 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add simple tests for do-notation parsing. - - - - - e7af1841 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add very simple QuickCheck properties for source parser spec. - - - - - c84efcf1 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Create simple test runner for hyperlinker tests. - - - - - 76b90447 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for basic identifier hyperlinking. - - - - - 0fbf4df6 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for operator hyperlinking. - - - - - 731aa039 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for constructor hyperlinking. - - - - - 995a78a2 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for record expressions and patterns hyperlinking. - - - - - 3566875a by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for literal syntax highlighting. - - - - - 68469a35 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add hyperlinker test runner to .cabal and .gitignore files. - - - - - aa946c93 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Adapt hyperlinker test runner to have the same interface as HTML one. - - - - - ce34da16 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Fix hyperlinker test runner file paths and add pretty-printing option. - - - - - 0d7dd65e by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add reference files for hyperlinker test cases. - - - - - efa4a1e0 by Łukasz Hanuszczak at 2015-07-01T00:47:32+02:00 Make hyperlinker test runner strip local links from generated source. - - - - - 3e96e584 by Łukasz Hanuszczak at 2015-07-01T01:14:59+02:00 Create simple script for accepting hyperlinker test case references. - - - - - 526fe610 by Łukasz Hanuszczak at 2015-07-01T01:16:41+02:00 Re-accept hyperlinker test cases with local references stripped out. - - - - - 892e2cb3 by Łukasz Hanuszczak at 2015-07-01T01:22:09+02:00 Fix bug with diffing wrong files in hyperlinker test runner. - - - - - 9ff46039 by Łukasz Hanuszczak at 2015-07-01T18:04:46+02:00 Remove unused dependencies in Haddock API spec configuration. - - - - - 47969c07 by Łukasz Hanuszczak at 2015-07-01T18:32:19+02:00 Add support for hyperlinking synonyms in patterns. - - - - - a73449e0 by Łukasz Hanuszczak at 2015-07-01T18:33:44+02:00 Create test case for hyperlinking @-patterns. - - - - - c2077ed8 by Łukasz Hanuszczak at 2015-07-01T19:06:04+02:00 Add support for hyperlinking universally quantified type variables. - - - - - 68017342 by Łukasz Hanuszczak at 2015-07-01T19:28:32+02:00 Create hyperlinker test case with quantified type variables. - - - - - 51c01a78 by Łukasz Hanuszczak at 2015-07-01T19:34:22+02:00 Add scoped type variables test for polymorphism test case. - - - - - 13181ae2 by Łukasz Hanuszczak at 2015-07-01T19:56:27+02:00 Add record wildcards test for records hyperlinking test case. - - - - - 991b81dd by Łukasz Hanuszczak at 2015-07-01T21:01:42+02:00 Document some functions in XHTML utlity module. - - - - - 98c8dfe5 by Łukasz Hanuszczak at 2015-07-01T22:25:21+02:00 Make hyperlinker render qualified names as one entity. - - - - - 75e13b9b by Łukasz Hanuszczak at 2015-07-01T22:27:38+02:00 Add qualified name test for identifiers hyperlinking test case. - - - - - de1e143f by Łukasz Hanuszczak at 2015-07-02T12:32:59+02:00 Fix crash happening when hyperlinking type family declarations. - - - - - 7a8fb175 by Łukasz Hanuszczak at 2015-07-02T12:47:03+02:00 Add support for anchoring data family constructor declarations. - - - - - 3b404e49 by Łukasz Hanuszczak at 2015-07-02T13:31:05+02:00 Improve support for hyperlinking type families. - - - - - 59eb7143 by Łukasz Hanuszczak at 2015-07-02T13:33:34+02:00 Add hyperlinker test case for checking type and type family declarations. - - - - - d1cda0c0 by Łukasz Hanuszczak at 2015-07-02T13:41:38+02:00 Fix issue with operators being recognized as preprocessor directives. - - - - - da206c9d by Łukasz Hanuszczak at 2015-07-02T17:18:12+02:00 Fix broken tests for parsing and hyperlinking hash operators. - - - - - 53750d1b by Łukasz Hanuszczak at 2015-07-02T18:53:28+02:00 Add support for anchoring signatures in type class declarations. - - - - - 1fa5bb10 by Łukasz Hanuszczak at 2015-07-02T19:04:47+02:00 Make hyperlinker generate anchors only to top-level value bindings. - - - - - a542305c by Łukasz Hanuszczak at 2015-07-02T19:05:58+02:00 Create hyperlinker test case for type classes. - - - - - b0dd4581 by Łukasz Hanuszczak at 2015-07-04T16:28:26+02:00 Update docs with information about source hyperlinking. - - - - - 9795302a by Łukasz Hanuszczak at 2015-07-04T16:52:15+02:00 Update docs on using `--read-interface` option. - - - - - 9acdc002 by Łukasz Hanuszczak at 2015-07-04T17:15:26+02:00 Remove potentially dangerous record access in hyperlinker AST module. - - - - - fb3ab7be by Łukasz Hanuszczak at 2015-07-04T17:40:10+02:00 Make Haddock generate warnings about potential misuse of hyperlinker. - - - - - a324c504 by Łukasz Hanuszczak at 2015-07-04T17:43:22+02:00 Fix incorrect specification of source style option in doc file. - - - - - 3f01a8e4 by Łukasz Hanuszczak at 2015-07-05T17:06:36+02:00 Refactor source path mapping to use modules as indices. - - - - - ac70f5b1 by Łukasz Hanuszczak at 2015-07-05T17:47:34+02:00 Fix bug where not all module interfaces were added to source mapping. - - - - - f5e57da9 by Łukasz Hanuszczak at 2015-07-06T16:39:57+02:00 Extract main hyperlinker types to separate module. - - - - - 43974905 by Łukasz Hanuszczak at 2015-07-06T16:52:13+02:00 Move source paths types to hyperlinker types module. - - - - - 3e236055 by Łukasz Hanuszczak at 2015-07-06T17:06:19+02:00 Add support for hyperlinking modules in import lists. - - - - - 58233d9f by Łukasz Hanuszczak at 2015-07-06T17:26:49+02:00 Add short documentation for hyperlinker source map type. - - - - - 14da016d by Łukasz Hanuszczak at 2015-07-06T18:07:20+02:00 Fix bug with module name being hyperlinked to `Prelude`. - - - - - 8f79db52 by Łukasz Hanuszczak at 2015-07-06T18:23:47+02:00 Fix problem with spec build in Haddock API configuration. - - - - - e7cc056c by Adam Sandberg Eriksson at 2015-07-07T23:22:21+01:00 StrictData: print correct strictness marks - - - - - e8253ca8 by Mateusz Kowalczyk at 2015-07-07T23:58:28+01:00 Update changelog - - - - - 0aba676b by Mateusz Kowalczyk at 2015-07-07T23:58:33+01:00 Relax upper bound on GHC a bit - - - - - 7a595381 by Mateusz Kowalczyk at 2015-07-07T23:58:52+01:00 Delete trailing whitespace - - - - - 50976d5e by Adam Sandberg Eriksson at 2015-07-08T15:03:04+02:00 StrictData: changes in HsBang type - - - - - 83b045fa by Mateusz Kowalczyk at 2015-07-11T14:35:18+01:00 Fix expansion icon for user-collapsible sections Closes haskell/haddock#412 - - - - - b2a3b0d1 by Mateusz Kowalczyk at 2015-07-22T22:03:21+01:00 Make some version changes after 2.16.1 release - - - - - a8294423 by Ben Gamari at 2015-07-27T13:16:07+02:00 Merge pull request haskell/haddock#422 from adamse/adamse-D1033 Merge for GHC D1033 - - - - - c0173f17 by randen at 2015-07-30T14:49:08-07:00 Break the response file by line termination rather than spaces, since spaces may be within the parameters. This simple approach avoids having the need for any quoting and/or escaping (although a newline char will not be possible in a parameter and has no escape mechanism to allow it). - - - - - 47c0ca14 by Alan Zimmerman at 2015-07-31T10:41:52+02:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - 45a9d770 by Mateusz Kowalczyk at 2015-07-31T09:47:43+01:00 Update changelog - - - - - 347a20a3 by Phil Ruffwind at 2015-08-02T23:15:26+01:00 Avoid JavaScript error during page load in non-frame mode In non-frame mode, parent.window.synopsis refers to the synopsis div rather than the nonexistent frame. Unfortunately, the script wrongly assumes that if it exists it must be a frame, leading to an error where it tries to access the nonexistent attribute 'replace' of an undefined value (synopsis.location). Closes haskell/haddock#406 - - - - - 54ebd519 by Phil Ruffwind at 2015-08-02T23:27:10+01:00 Link to the definitions to themselves Currently, the definitions already have an anchor tag that allows URLs with fragment identifiers to locate them, but it is rather inconvenient to obtain such a URL (so-called "permalink") as it would require finding the a link to the corresponding item in the Synopsis or elsewhere. This commit adds hyperlinks to the definitions themselves, allowing users to obtain links to them easily. To preserve the original aesthetics of the definitions, we alter the color of the link so as to be identical to what it was, except it now has a hover effect indicating that it is clickable. Additionally, the anchor now uses the 'id' attribute instead of the (obsolete) 'name' attribute. Closes haskell/haddock#407 - - - - - 02cc8bb7 by Phil Ruffwind at 2015-08-02T23:28:02+01:00 Fix typo in Haddock.Backends.Xhtml.Layout: divSynposis -> divSynopsis Closes haskell/haddock#408 - - - - - 2eb0a458 by Phil Ruffwind at 2015-08-02T23:30:07+01:00 Fix record field alignment when name is too long Change <dl> to <ul> and use display:table rather than floats to layout the record fields. This avoids bug haskell/haddock#301 that occurs whenever the field name gets too long. Slight aesthetic change: the entire cell of the field's source code is now shaded gray rather than just the area where text exists. Fixes haskell/haddock#301. Closes haskell/haddock#421 - - - - - 7abb3402 by Łukasz Hanuszczak at 2015-08-02T23:32:14+01:00 Add some utility definitions for generating line anchors. - - - - - e0b1d79b by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Make hyperlinked source renderer generate line anchors. - - - - - 24dd4c9f by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Re-accept test cases after adding line anchors for each of them. - - - - - 0372cfcb by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Override source line flags when source hyperlinker is enabled. - - - - - a81bcd07 by Mateusz Kowalczyk at 2015-08-02T23:58:25+01:00 Update tests to follow HTML changes - - - - - d2d7426f by Łukasz Hanuszczak at 2015-08-06T20:54:59+02:00 Fix quote syntax for promoted types. - - - - - 668cf029 by Łukasz Hanuszczak at 2015-08-06T21:12:00+02:00 Apply promoted type quoting to type-level consing. - - - - - 89f8e7c6 by Łukasz Hanuszczak at 2015-08-06T21:17:10+02:00 Extend advanced types test case with other examples. - - - - - 86494bca by Łukasz Hanuszczak at 2015-08-06T21:22:06+02:00 Rename advanced types test case and accept new output. - - - - - dbb7c7c0 by Adam Sandberg Eriksson at 2015-08-09T23:01:05+02:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 2a7704fa by Ben Gamari at 2015-08-10T13:18:05+02:00 Merge pull request haskell/haddock#433 from adamse/split-hsbang HsBang is split into HsSrcBang and HsImplBang - - - - - 891954bc by Thomas Miedema at 2015-08-15T14:51:18+02:00 Follow changes in GHC build system - - - - - b55d32ab by Mateusz Kowalczyk at 2015-08-21T18:06:09+01:00 Make Travis use 7.10.2 - - - - - 97348b51 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Move SYB utilities to standalone module. - - - - - 748ec081 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement `everywhere` transformation in SYB module. - - - - - 011cc543 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement generic transformation constructor. - - - - - b9510db2 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Create simple utility module for type specialization. - - - - - 43229fa6 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Make type of type specialization function more general. - - - - - fd844e90 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Add basic HTML test case for checking instance specialization. - - - - - 6ea0ad04 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Make HTML class instance printer take optional signature argument. - - - - - 65aa41b6 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Refactor instance head type to record instead of a meaningless tuple. - - - - - 3fc3bede by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add expandable method section for each class instance declaration. - - - - - 99ceb107 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Move dummy post-family instances for `DocName` to `Types` module. - - - - - e98f4708 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create convenience functions for type specialization module. - - - - - b947552f by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Hook type specialization logic with HTML pretty-printer. - - - - - dcaa8030 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create stub functions for sugaring specialized types. - - - - - fa84bc65 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement list syntax sugaring logic for specialized types. - - - - - e8b05b07 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement tuple syntax sugaring logic for specialized types. - - - - - 68a2e5bc by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Get rid of code duplication in type specialization module. - - - - - 4721c336 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create scaffolding of a framework for renaming specialized types. - - - - - 271b488d by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in missing cases in specialized type renaming function. - - - - - bfa5f2a4 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Remove code duplication in specialized type renamer. - - - - - ea6bd0e8 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Change state of the type renaming monad. - - - - - 77c5496e by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement simple mechanism for generating new type names. - - - - - 91bfb48b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in stub behaviour with actual environment renaming. - - - - - d244517b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix logic behind binder type renaming. - - - - - f3c5e360 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add SYB-like utility function for performing stateful queries. - - - - - eb3f9154 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create function for retrieving free variables from given type. - - - - - a94561d3 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix compilation error caused by incorrect type signature. - - - - - 8bb707cf by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Move `SetName` class definition to types module. - - - - - 5800b13b by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Hook type renamer with instance method HTML pretty-printer. - - - - - 6a480164 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some test cases for type renamer. - - - - - 839842f7 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized signatures refer to original signature declaration. - - - - - 4880f7c9 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized methods be nicely formatted again. - - - - - ab5a6a2e by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Attach source locations to the specialized class methods. - - - - - 43f8a559 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Extend instances test case to also test multi-name type signatures. - - - - - 59bc751c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix tab-based indentation in instances test case. - - - - - c2126815 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Improve placement of instance methods expander button. - - - - - 0a32e287 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add new data type declaration to instance specialization test case. - - - - - 5281af1f by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make type renamer first try single-letter names as alternatives. - - - - - 7d509475 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix type renamer bug with incorrect names being generated. - - - - - 0f35bf7c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some documentation and refactor type specialization module. - - - - - da1d0803 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix another bug where type renamer was generating incorrect names. - - - - - cd39b5cb by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Refactor type renamer to rebinding and pure renaming phases. - - - - - 850251f4 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix unwitting compilation bug. - - - - - e5e9fc01 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Integrate instance specification type into class instance definition. - - - - - 825b0ea0 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer neccessary instance specification type. - - - - - cdba44eb by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix declaration converter to use more appropriate mode for methods. - - - - - bc45c309 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug with types not being specialized at all. - - - - - 5d8e5d89 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug where instance expander was opening wrong section. - - - - - 6001ee41 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix another type renamer bug where not all names were rebound. - - - - - 5f58ce2a by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix yet another renamer bug where some names were not unique. - - - - - 8265e521 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split instance subsection layout method to top-level declarations. - - - - - e5e66298 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Rearrange layout of instance methods in generated documentation. - - - - - a50b4eea by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer used layout method. - - - - - 2ff36ec2 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach section title to the instance methods block. - - - - - 7ac15300 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Add basic tests for associated types in instances test case. - - - - - db0ea2f9 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach associated types information to instance header. - - - - - 71cad4d5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details section contain associated types information. - - - - - deee2809 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Improve look of rendered associated families in instance details. - - - - - 839d13a5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Introduce alternative type for family declarations. - - - - - d397f03f by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details record use new type for family declarations. - - - - - 2b23fe97 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split printer of type family header to separate functions. - - - - - c3498cdc by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Implement HTML renderer for pseudo-family declarations. - - - - - c12bbb04 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Apply type specializer to associated type family declarations. - - - - - 2fd69ff2 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Create helper method for specializing type signatures. - - - - - 475826e7 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Refactor specializer module to be independent from XHTML backend. - - - - - f00b431c by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add some documentation for instance head specializer. - - - - - a9fef2dc by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix bug with missing space in documentation for associated types. - - - - - 50e29056 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with incorrect instance details sections being expanded. - - - - - e6dfdd03 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by adding instance details section. - - - - - 75565b2a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make section identifier of instance details more GHC-independent. - - - - - add0c23e by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Re-accept tests after applying deterministic section identifiers. - - - - - 878f2534 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make identifier generation also architecture-independent. - - - - - 48be69f8 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with instance expander hijacking type hyperlink click. - - - - - 47830c1f by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Get rid of dreadful hashing function for generating identifiers. - - - - - 956cd5af by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Move `InstOrigin` type declaration to more appropriate module. - - - - - bf672ed3 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by changes related to instance expander. - - - - - 8f2a949a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add examples with type operators to the instances test case. - - - - - 64600a84 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add basic support for sugaring infix type operators. - - - - - 747d71b8 by Łukasz Hanuszczak at 2015-08-21T18:22:34+01:00 Add support for sugaring built-in function syntax. - - - - - d4696ffb by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Remove default methods from Hoogle class output. - - - - - bf0e09d7 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add fixity declarations in Hoogle backend output. - - - - - 90e91a51 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix bug with incorrect fixities being generated in Hoogle backend. - - - - - 48f11d35 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve class type family declarations output in Hoogle backend. - - - - - 661e8e8f by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add missing default family equations in Hoogle output. - - - - - e2d64103 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve formatting of class details output in Hoogle backend. - - - - - 490fc377 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix weird-looking Hoogle output for familyless classes. - - - - - ea115b64 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create script file for new HTML test runner. - - - - - 609913d3 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Set default behaviour if no arguments given. - - - - - dc115f67 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for providing optional arguments for test runner. - - - - - d93ec867 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve output of test runner error messages. - - - - - 0be9fe12 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for executing Haddock process in test runner. - - - - - 4e4d00d9 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add GHC path to test runner configuration. - - - - - d67a2086 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make GHC path a test runner command-line argument. - - - - - c810079a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Extend test runner configuration with Haddock arguments. - - - - - fee18845 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor test runner and create stub functions. - - - - - ff7c161f by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make test runner actually run Haddock executable. - - - - - 391f73e6 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with test runner not producing any output files. - - - - - 81a74e2d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Setup skeleton of framework for running tests. - - - - - f8a79ec4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with modules not being found in global search mode. - - - - - 7e700b4d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make Haddock standard output redirection be more configurable. - - - - - 53b4c17a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Incorporate old, ugly functions for comparing output files. - - - - - 8277c8aa by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor architecture of test runner output checking functions. - - - - - 587bb414 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement actual diffing mechanism. - - - - - 9ed2b5e4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve code style to match popular guidelines. - - - - - 14bffaf8 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make it possible to choose alternative diff tool. - - - - - 5cdfb005 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create stub methods for processing test output as XML documents. - - - - - 7ef8e12e by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement link-stripping logic as simple SYB transformation. - - - - - 8a1fcd4f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Incorporate link stripping to output diffing mechanism. - - - - - 37dba2bc by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement footer-stripping logic. - - - - - 9cd52120 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Add missing dependencies in Cabal configuration file. - - - - - e0f83c6e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix issue with output being printed in incorrect order. - - - - - 0a94fbb0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to run tests without generating diff. - - - - - 76a58c6f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor HTML test suite boilerplate to external package. - - - - - af41e6b0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utilities for storing directory configuration. - - - - - d8f0698f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move IO-dependent config of HTML test suite to test package. - - - - - 17369fa0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Enable all compiler warnings in Haddock test package configuration. - - - - - 9d03b47a by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move Haddock runner of HTML test suite to Haddock test package. - - - - - 4b3483c5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 03754194 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience wrappers to simplify in test entry points. - - - - - 27476ab7 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adjust module visibility and items they export. - - - - - c40002ba by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Remove no longer useful test option. - - - - - 55ab2541 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Change extension of test files used for diffing. - - - - - 136bf4e4 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor and simplify XHTML helper module of test package. - - - - - 69f7e3df by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix typo in link stripper of HTML test suite runner. - - - - - 0c3c1c6b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience script for running specific HTML tests. - - - - - 489e1b05 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement utility functions for conditional link stripping. - - - - - 0f985dc3 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt `hypsrc-test` module to work with new testing framework. - - - - - 927406f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement output accepting mechanism in test package. - - - - - 8545715e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utility function for recursive obtaining directory contents. - - - - - cb70381f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 019599b5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix path handling in test runner. - - - - - 399b985b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to specify ignored files for test output. - - - - - 41b3d93d by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt HTML test runner to use new ignoring functionality. - - - - - e2091c8b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix bug with not all test output files being checked. - - - - - b22134f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Specify ignored files for hyperlinker source test runner. - - - - - 3301dfa1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Copy test runner script for hyperlinked source case. - - - - - d39a6dfa by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner invoking Haddock in incorrect mode. - - - - - f32c8ff3 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix path handling in test module loader. - - - - - 10f94ee9 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make test runner ignore test packages with no modules. - - - - - 5dc4239c by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create test runner entry points for LaTeX test suite. - - - - - 58d1f7cf by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with unnecessary checking old test output. - - - - - c7ce76e1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Re-implement test acceptance functionality. - - - - - 13bbabe8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix warning about no longer needed definition. - - - - - 958a99b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Adapt Cabal configuration to execute LaTeX suite with new runner. - - - - - 550ff663 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Setup test suite for Hoogle backend. - - - - - 3aa969c4 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make Hoogle backend create output directory if needed. - - - - - eb085b02 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Add appropriate .gitignore entry and configure Hoogle test suite. - - - - - a50bf915 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner failing when run on multiple test packages. - - - - - bf5368b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create simple test cases for Hoogle backend. - - - - - 6121ba4b by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create helper function for conversion between XML and XHTML. - - - - - cb516061 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Refactor existing code to use XHTML printer instead of XML one. - - - - - e2de8c82 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Improve portability of test runner scripts. - - - - - 9563e774 by Łukasz Hanuszczak at 2015-08-22T23:43:16+02:00 Remove redundant import statement. - - - - - 55353df1 by Łukasz Hanuszczak at 2015-08-24T23:09:20+02:00 Fix bug with accepting to non-existing directory. - - - - - 00a334ca by Łukasz Hanuszczak at 2015-08-24T23:09:47+02:00 Accept output for Hoogle and LaTeX backends. - - - - - 29191d8b by Łukasz Hanuszczak at 2015-08-24T23:14:18+02:00 Get rid of obsolete testing utilities. - - - - - bbb25db3 by Łukasz Hanuszczak at 2015-08-24T23:18:50+02:00 Update sandbox setup guide to work with Haddock test package. - - - - - cfd45248 by Łukasz Hanuszczak at 2015-08-24T23:51:30+02:00 Make Travis aware of Haddock test package. - - - - - 74185b7a by Łukasz Hanuszczak at 2015-08-25T17:41:59+02:00 Fix test suite failure when used with Stack. - - - - - 18769697 by Łukasz Hanuszczak at 2015-08-25T18:02:09+02:00 Add sample Stack setup to the hacking guide. - - - - - 22715eeb by Łukasz Hanuszczak at 2015-08-25T18:04:47+02:00 Fix Markdown formatting of README file. - - - - - b49ec386 by Łukasz Hanuszczak at 2015-08-25T18:13:36+02:00 Setup Haddock executable path in Travis configuration. - - - - - 5d29eb03 by Eric Seidel at 2015-08-30T09:55:58-07:00 account for changes to ipClass - - - - - f111740a by Ben Gamari at 2015-09-02T13:20:37+02:00 Merge pull request haskell/haddock#443 from bgamari/ghc-head account for changes to ipClass - - - - - a2654bf6 by Jan Stolarek at 2015-09-03T01:32:57+02:00 Follow changes from haskell/haddock#6018 - - - - - 2678bafe by Richard Eisenberg at 2015-09-21T12:00:47-04:00 React to refactoring CoAxiom branch lists. - - - - - ebc56e24 by Edward Z. Yang at 2015-09-21T11:53:46-07:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4a8c4198 by Tamar Christina at 2015-09-27T13:59:08+02:00 Create Process: removed PhaseFailed - - - - - 7e99b790 by Oleg Grenrus at 2015-09-27T20:52:10+03:00 Generate docs for orphan instances - - - - - 32e932e2 by Oleg Grenrus at 2015-09-28T07:21:11+03:00 Have source links for orphan instances - - - - - c2eb9f4f by Oleg Grenrus at 2015-09-28T07:24:58+03:00 Print orphan instances header only if required - - - - - ff96f978 by Oleg Grenrus at 2015-09-28T07:40:54+03:00 Add orphan instances link to contents box - - - - - d72490a6 by Oleg Grenrus at 2015-09-28T16:37:44+03:00 Fix orphan instance collapsing - - - - - 25d3dfe5 by Ben Gamari at 2015-10-03T12:38:09+02:00 Merge pull request haskell/haddock#448 from Mistuke/fix-silent-death-of-runInteractive Remove PhaseFailed - - - - - 1e45e43b by Edward Z. Yang at 2015-10-11T13:10:10-07:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b1370ac1 by Adam Gundry at 2015-10-16T16:26:42+01:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 60bef421 by Simon Peyton Jones at 2015-10-26T12:52:36+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 4c1898ca by Simon Peyton Jones at 2015-10-27T14:24:56+00:00 Track change to PatSyn.patSynSig - - - - - 25108e85 by Simon Peyton Jones at 2015-10-27T17:34:18+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - 693643ac by Ben Gamari at 2015-10-28T14:33:06+01:00 Account for Typeable changes The treatment of type families changed. - - - - - cd7c2221 by Simon Peyton Jones at 2015-10-30T13:03:51+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - 712032cb by Herbert Valerio Riedel at 2015-10-31T11:01:45+01:00 Relax upper bound on `base` to allow base-4.9 - - - - - 0bfa0475 by Simon Peyton Jones at 2015-10-31T19:08:13+00:00 More adaption to wildcard-refactor - - - - - 0a3c0cb7 by Simon Peyton Jones at 2015-10-31T22:14:43+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - c4fd4ec9 by Alan Zimmerman at 2015-11-01T11:16:34+01:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 42cdd882 by Matthew Pickering at 2015-11-06T20:02:16+00:00 Change for IEThingWith - - - - - f368b7be by Ben Gamari at 2015-11-11T11:35:51+01:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - e32965b8 by Simon Peyton Jones at 2015-11-13T12:18:17+00:00 Merge with origin/head - - - - - ebcf795a by Edward Z. Yang at 2015-11-13T21:56:27-08:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4e23989f by Simon Peyton Jones at 2015-11-18T11:32:54+00:00 Wibbles to Haddock - - - - - 2289cd4a by Simon Peyton Jones at 2015-11-20T23:12:49+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor - - - - - 695975a6 by Alan Zimmerman at 2015-11-21T21:16:12+02:00 Update to match GHC wip/T11019 - - - - - bbba21e7 by Simon Peyton Jones at 2015-11-23T13:54:31+00:00 merge with origin/ghc-head - - - - - 3d664258 by Simon Peyton Jones at 2015-11-23T17:17:18+00:00 Wibble - - - - - e64cf586 by Herbert Valerio Riedel at 2015-12-05T00:29:55+01:00 Canonicalise Monad instances - - - - - a2de15a7 by Alan Zimmerman at 2015-12-05T17:33:52+02:00 Matching changes for haskell/haddock#11028 - - - - - cc29a3e4 by Alan Zimmerman at 2015-12-05T19:45:33+02:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 95dd15d1 by Richard Eisenberg at 2015-12-11T17:33:39-06:00 Update for type=kinds - - - - - cb5fd9ed by Herbert Valerio Riedel at 2015-12-14T15:07:30+00:00 Bump versions for ghc-7.11 - - - - - 4f286d96 by Simon Peyton Jones at 2015-12-14T15:10:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 13ea2733 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 098df8b8 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track changes in HsSyn for quasi-quotes - - - - - 716a64de by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track change in API of TyCon - - - - - 77a66bca by Adam Gundry at 2015-12-14T15:10:58+00:00 Track API changes to support empty closed type familes - - - - - f2808305 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track the new location of setRdrNameSpace - - - - - ba8b08a4 by Alan Zimmerman at 2015-12-14T15:10:59+00:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - a4ded87e by Thomas Winant at 2015-12-14T15:14:05+00:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 25c78107 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: print correct strictness marks - - - - - 6cbc41c4 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: changes in HsBang type - - - - - ad46821a by Alan Zimmerman at 2015-12-14T15:14:06+00:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - abc0ae5b by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 3308d06c by Thomas Miedema at 2015-12-14T15:14:07+00:00 Follow changes in GHC build system - - - - - 6c763deb by Eric Seidel at 2015-12-14T15:14:07+00:00 account for changes to ipClass - - - - - ae5b4eac by Jan Stolarek at 2015-12-14T15:17:00+00:00 Follow changes from haskell/haddock#6018 - - - - - ffbc40e0 by Richard Eisenberg at 2015-12-14T15:17:02+00:00 React to refactoring CoAxiom branch lists. - - - - - d1f531e9 by Edward Z. Yang at 2015-12-14T15:17:02+00:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 79f73754 by Tamar Christina at 2015-12-14T15:17:02+00:00 Create Process: removed PhaseFailed - - - - - 3d37bebb by Edward Z. Yang at 2015-12-14T15:20:46+00:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 5f8a9e44 by Adam Gundry at 2015-12-14T15:20:48+00:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 79dda70f by Simon Peyton Jones at 2015-12-14T15:26:02+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 959930fb by Simon Peyton Jones at 2015-12-14T15:37:50+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - e18a8df5 by Simon Peyton Jones at 2015-12-14T15:37:52+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - aa35ab52 by Simon Peyton Jones at 2015-12-14T15:40:18+00:00 More adaption to wildcard-refactor - - - - - 8ceef94b by Simon Peyton Jones at 2015-12-14T15:46:04+00:00 Track change to PatSyn.patSynSig - - - - - cd81e83d by Ben Gamari at 2015-12-14T15:46:06+00:00 Account for Typeable changes The treatment of type families changed. - - - - - 63c9117c by Herbert Valerio Riedel at 2015-12-14T15:46:34+00:00 Relax upper bound on `base` to allow base-4.9 - - - - - a484c613 by Alan Zimmerman at 2015-12-14T15:47:46+00:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 2c26fa51 by Matthew Pickering at 2015-12-14T15:47:47+00:00 Change for IEThingWith - - - - - 593baa0f by Ben Gamari at 2015-12-14T15:49:21+00:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - b6b5ca78 by Edward Z. Yang at 2015-12-14T15:49:54+00:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b5b0e072 by Alan Zimmerman at 2015-12-14T15:54:20+00:00 Update to match GHC wip/T11019 - - - - - 14ddeb68 by Simon Peyton Jones at 2015-12-14T15:54:22+00:00 Wibble - - - - - 10a90ad8 by Herbert Valerio Riedel at 2015-12-14T15:54:22+00:00 Canonicalise Monad instances - - - - - ed68ac50 by Alan Zimmerman at 2015-12-14T15:55:48+00:00 Matching changes for haskell/haddock#11028 - - - - - 3f7e5a2d by Alan Zimmerman at 2015-12-14T15:55:49+00:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 6543a73f by Richard Eisenberg at 2015-12-14T15:59:55+00:00 Update for type=kinds - - - - - 193a5c48 by Matthew Pickering at 2015-12-14T18:17:00+00:00 Changes to compile with 8.0 - - - - - add669ec by Matthew Pickering at 2015-12-14T18:47:12+00:00 Warnings - - - - - 223f3fb4 by Ben Gamari at 2015-12-15T23:45:05+01:00 Update for D1200 - - - - - d058388f by Ben Gamari at 2015-12-16T05:40:17-05:00 Types: Add Outputable[Bndr] DocName instances - - - - - 62ecd7fb by Ben Gamari at 2015-12-16T09:23:09-05:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ddbc187a by Ben Gamari at 2015-12-16T17:54:55+01:00 Update for D1200 - - - - - cec83b52 by Ben Gamari at 2015-12-16T17:54:55+01:00 Types: Add Outputable[Bndr] DocName instances - - - - - d12ecc98 by Ben Gamari at 2015-12-16T17:54:55+01:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ada1616f by Ben Gamari at 2015-12-16T17:54:58+01:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a4f0383d by Ben Gamari at 2015-12-16T23:32:38+01:00 Fix Hyperlinker GHC.con_names is now GHC.getConNames - - - - - a10e6849 by Ben Gamari at 2015-12-20T00:54:11+01:00 Merge remote-tracking branch 'mrhania/testing-framework-improvements' into ghc-head - - - - - f078b4fd by Ben Gamari at 2015-12-20T00:59:51+01:00 test: Compatibility with Cabal 1.23 - - - - - 88a511a9 by Ben Gamari at 2015-12-20T01:14:35+01:00 Merge remote-tracking branch 'phadej/orphans' into ghc-head - - - - - 4e250f36 by Ben Gamari at 2015-12-20T01:14:52+01:00 Add html-test for orphan instances output - - - - - 87fffbad by Alan Zimmerman at 2015-12-20T09:50:42+02:00 Update for GHC trac#11258 Adding locations to RdrName in FieldOcc and AmbiguousFieldOcc - - - - - 6b7e51c9 by idontgetoutmuch at 2015-12-20T21:01:47+00:00 Merge pull request haskell/haddock#1 from haskell/ghc-head Ghc head - - - - - 229c1fb5 by Dominic Steinitz at 2015-12-21T07:19:16+00:00 Handle inline math with mathjax. - - - - - 57902d66 by Dominic Steinitz at 2015-12-21T08:07:11+00:00 Fix the documentation for haddock itself. Change notation and add support for inline math. Allow newlines in display math. Add a command line option for the mathjax url (you might want to use a locally installed version). Rebase tests because of extra url and version change. Respond to (some of the) comments. Fix warnings in InterfaceFile.hs - - - - - 0e69f236 by Herbert Valerio Riedel at 2015-12-21T18:30:43+01:00 Fix-up left-over assumptions of GHC 7.12 into GHC 8.0 - - - - - c67f8444 by Simon Peyton Jones at 2015-12-22T16:26:56+00:00 Follow removal of NamedWildCard from HsType - - - - - da40327a by Ben Gamari at 2015-12-23T14:15:28+01:00 html-test/Operators: Clear up ambiguous types For reasons that aren't entirely clear a class with ambiguous types was accepted by GHC <8.0. I've added a functional dependency to clear up this ambiguity. - - - - - 541b7fa4 by Ben Gamari at 2015-12-23T14:18:51+01:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 0febc947 by Ben Gamari at 2015-12-24T00:30:20+01:00 hoogle-test/AssocTypes: Allow AmbiguousTypes GHC 8.0 complains otherwise - - - - - 25810841 by Ben Gamari at 2015-12-24T00:33:18+01:00 OrphanInstances: Accept test output - - - - - 841987f3 by Ben Gamari at 2015-12-25T11:03:11+01:00 Merge remote-tracking branch 'idontgetoutmuch/ghc-head' into ghc-head - - - - - 358391f0 by Ben Gamari at 2015-12-26T10:44:50+01:00 Add missing import - - - - - a8896885 by Ben Gamari at 2015-12-26T10:45:27+01:00 travis: Use Travis containers - - - - - 85e82134 by Herbert Valerio Riedel at 2015-12-30T17:25:39+01:00 tweak version bounds for GHC-8.1 - - - - - 672a5f75 by randen at 2016-01-01T23:45:25-08:00 The Haddock part for fully gcc-like response files " driver/Main.hs * Moved the response file handling into ResponseFile.hs, updating import section as appropriate. * driver/ResponseFile.hs * New file. In anticipation that maybe some day this could be provided by another library, and to make it possible to unit test, this functionality is pulled out of the Main.hs module, and expanded to support the style/format of response files which gcc uses. * The specification for the format of response files which gcc generates and consumes, seems to be best derived from the gcc code itself (libiberty/argv.c), so that is what has been done here. * This is intended to fix haskell/haddock#379 * driver-test/Main.hs * New file for testing code in the driver source tree * driver-test/ResponseFileSpec.hs * Tests, adapted/adopted from the same gcc code where the escaping/unescaping is from, in the hspec style of unit tests * haddock.cabal * Add the driver-test test-suite. Introduces a new library dependency (upon hspec) for the haddock driver target in the haddock.cabal file, but practically, this should not be a problem as the haddock-api tests already depend on hspec. - - - - - 498781df by Ben Gamari at 2016-01-06T13:41:04+01:00 Version bumps and changelog - - - - - 8451e46a by Ben Gamari at 2016-01-06T13:47:17+01:00 Merge remote-tracking branch 'randen/bug468' - - - - - fb2d9181 by Ben Gamari at 2016-01-06T08:14:42-05:00 Add ResponseFile to OtherModules - - - - - 2cb2d2e3 by Ben Gamari at 2016-01-06T14:35:00+01:00 Merge branch 'master' into ghc-head - - - - - 913477d4 by Eric Seidel at 2016-01-11T14:57:57-08:00 deal with un-wiring of IP class - - - - - c557a4b3 by Alan Zimmerman at 2016-01-15T11:14:35+02:00 Update to match wip/T11430 in GHC - - - - - 3e135093 by Alan Zimmerman at 2016-01-16T18:21:59+01:00 Update to match wip/T11430 in GHC - - - - - c48ef2f9 by Ben Gamari at 2016-01-18T09:50:06+01:00 Merge remote-tracking branch 'gridaphobe/ghc-head' into ghc-head - - - - - 9138a1b0 by Eric Seidel at 2016-01-18T12:50:15+01:00 deal with un-wiring of IP class (cherry picked from commit 17388b0f0029d969d79353be7737eb01c7b8dc5f) - - - - - b48c172e by Joachim Breitner at 2016-01-19T00:11:38+01:00 Make sure --mathjax affects all written HTML files This fixes haskell/haddock#475. - - - - - af61fe63 by Ryan Scott at 2016-02-07T23:25:57+01:00 Render */# instead of TYPE 'Lifted/TYPE 'Unlifted (fixes haskell/haddock#473) - - - - - b6458693 by Ben Gamari at 2016-02-07T23:29:27+01:00 Merge pull request haskell/haddock#477 from haskell/issue-475 Make sure --mathjax affects all written HTML files - - - - - adcc0071 by Ben Gamari at 2016-02-07T23:34:52+01:00 Merge branch 'master' into ghc-head - - - - - d0404e61 by Ben Gamari at 2016-02-08T12:46:49+01:00 doc: Switch to Sphinx - - - - - acb153b3 by Ben Gamari at 2016-02-08T12:46:56+01:00 Document --use-unicode flag - - - - - c20bdf1d by Ben Gamari at 2016-02-08T13:41:24+01:00 Fix GHC and haddock-library dependency bounds - - - - - 8d946801 by Ben Gamari at 2016-02-08T14:54:56+01:00 testsuite: Rework handling of output sanitization Previously un-cleaned artifacts were kept as reference output, making it difficult to tell what has changed and causing spurious changes in the version control history. Here we rework this, cleaning the output during acceptance. To accomplish this it was necessary to move to strict I/O to ensure the reference handle was closed before accept attempts to open the reference file. - - - - - c465705d by Ben Gamari at 2016-02-08T15:36:05+01:00 test: Compare on dump For reasons I don't understand the Xml representations differ despite their textual representations being identical. - - - - - 1ec0227a by Ben Gamari at 2016-02-08T15:36:05+01:00 html-test: Accept test output - - - - - eefbd63a by Ben Gamari at 2016-02-08T15:36:08+01:00 hypsrc-test: Accept test output And fix impredicative Polymorphism testcase. - - - - - d1df4372 by Ben Gamari at 2016-02-08T15:40:44+01:00 Merge branch 'fix-up-testsuite' - - - - - 206a3859 by Phil Ruffwind at 2016-02-08T17:51:21+01:00 Move the permalinks to "#" on the right side Since pull request haskell/haddock#407, the identifiers have been permalinked to themselves, but this makes it difficult to copy the identifier by double-clicking. To work around this usability problem, the permalinks are now placed on the far right adjacent to "Source", indicated by "#". Also, 'namedAnchor' now uses 'id' instead of 'name' (which is obsolete). - - - - - 6c89fa03 by Phil Ruffwind at 2016-02-08T17:54:44+01:00 Update tests for previous commit - - - - - effaa832 by Ben Gamari at 2016-02-08T17:56:17+01:00 Merge branch 'anchors-redux' - - - - - 9a2bec90 by Ben Gamari at 2016-02-08T17:58:40+01:00 Use -fprint-unicode-syntax when --use-unicode is enabled This allows GHC to render `*` as its Unicode representation, among other things. - - - - - 28ecac5b by Ben Gamari at 2016-02-11T18:53:03+01:00 Merge pull request haskell/haddock#480 from bgamari/sphinx Move documentation to ReStructuredText - - - - - 222e5920 by Ryan Scott at 2016-02-11T15:42:42-05:00 Collapse type/data family instances by default - - - - - a80ac03b by Ryan Scott at 2016-02-11T20:17:09-05:00 Ensure expanded family instances render correctly - - - - - 7f985231 by Ben Gamari at 2016-02-12T10:04:22+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - d4eda086 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Various cleanups - - - - - 79bee48d by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show kind signatures for type family variables Addresses GHC haskell/haddock#11588. - - - - - b2981d98 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show 'where ...' after closed type family Seems like we should ideally show the actual equations as well but that seems like it would be a fair amount of work - - - - - cfc0e621 by Ben Gamari at 2016-02-18T22:48:12+01:00 Merge pull request haskell/haddock#483 from bgamari/T11588 Fix GHC haskell/haddock#11588 This fixes GHC haskell/haddock#11588: * Show where ... after closed type families * Show kind signatures on type family type variables - - - - - 256e8a0d by Ben Gamari at 2016-02-18T23:15:39+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 32402036 by Richard Eisenberg at 2016-02-24T13:21:44-05:00 Follow-on changes to support RuntimeRep - - - - - 2b1c572d by Matthew Pickering at 2016-03-04T21:04:02+00:00 Remove unused functions - - - - - eb906f50 by Richard Eisenberg at 2016-03-13T21:17:20+01:00 Follow-on changes to support RuntimeRep (cherry picked from commit ab954263a793d8ced734459d6194a5d89214b66c) - - - - - 8c34ef34 by Richard Eisenberg at 2016-03-14T23:47:23-04:00 Changes due to fix for GHC#11648. - - - - - 0e022014 by Richard Eisenberg at 2016-03-15T14:06:45+01:00 Changes due to fix for GHC#11648. (cherry picked from commit bb994de1ab0c76d1aaf1e39c54158db2526d31f1) - - - - - ed3f78ab by Rik Steenkamp at 2016-04-02T22:20:36+01:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 - - - - - d3210042 by Rik Steenkamp at 2016-04-04T15:43:32+02:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 (cherry picked from commit 3ddcbd6b8e6884bd95028381176eb33bee6896fb) - - - - - 236eec90 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 692ee7e0 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. - - - - - 79619f57 by Ben Gamari at 2016-04-10T23:46:22+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. (cherry picked from commit c6d6a18d85e5e2d9bb5904e6919e8a8d7e31c4c5) - - - - - 3358ccb4 by Ben Gamari at 2016-04-10T23:47:27+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 264949b1 by Ben Gamari at 2016-04-16T17:50:23+02:00 Merge pull request haskell/haddock#482 from RyanGlScott/ghc-head Collapse type/data family instances by default - - - - - 478c483a by Ben Gamari at 2016-04-16T17:51:09+02:00 Merge pull request haskell/haddock#489 from mpickering/unused-functions Remove some unused functions - - - - - c94e55f0 by Ryan Scott at 2016-04-16T17:57:54+02:00 Collapse type/data family instances by default (cherry picked from commit 2da130a8db8f995c119b544fad807533236cf088) - - - - - 31e633d3 by Ryan Scott at 2016-04-16T17:58:06+02:00 Ensure expanded family instances render correctly (cherry picked from commit 1338b5d7c32939de6bbc31af0049477e4f847103) - - - - - 03e4d197 by Matthew Pickering at 2016-04-16T17:58:21+02:00 Remove unused functions (cherry picked from commit b89d1c2456bdb2d4208d94ded56155f7088a37d0) - - - - - ed4116f6 by Ben Gamari at 2016-04-20T10:46:57+02:00 ghc: Install files for needed --hyperlinked-source - - - - - 0be999c4 by Ben Gamari at 2016-04-20T11:37:54+02:00 ghc: Install files for needed --hyperlinked-source (cherry picked from commit 5c82c9fc2d21ddaae4a2470f1c375426968f19c6) - - - - - 4d17544c by Simon Peyton Jones at 2016-04-20T12:42:28+01:00 Track change to HsGroup This relates to a big GHC patch for Trac haskell/haddock#11348 - - - - - 1700a50d by Ben Gamari at 2016-05-01T13:19:27+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. - - - - - 0b7c8125 by Ben Gamari at 2016-05-01T13:21:43+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. (cherry picked from commit 609018dd09c4ffe27f9248b2d8b50f6196cd42b9) - - - - - af115ce0 by Ryan Scott at 2016-05-04T22:15:50-04:00 Render Haddocks for derived instances Currently, one can document top-level instance declarations, but derived instances (both those in `deriving` clauses and standalone `deriving` instances) do not enjoy the same privilege. This makes the necessary changes to the Haddock API to enable rendering Haddock comments for derived instances. This is part of a fix for Trac haskell/haddock#11768. - - - - - 76fa1edc by Ben Gamari at 2016-05-10T18:13:25+02:00 haddock-test: A bit of refactoring for debuggability - - - - - 7d4c4b20 by Ben Gamari at 2016-05-10T18:13:25+02:00 Create: Mark a comment as TODO - - - - - 2a6d0c90 by Ben Gamari at 2016-05-10T18:13:25+02:00 html-test: Update reference output - - - - - bd60913d by Ben Gamari at 2016-05-10T18:13:25+02:00 hypsrc-test: Fix reference file path in cabal file It appears the haddock insists on prefixing --hyperlinked-sourcer output with directory which the source appeared in. - - - - - c1548057 by Ben Gamari at 2016-05-10T18:22:12+02:00 doc: Update extra-source-files in Cabal file - - - - - 41d5bae3 by Ben Gamari at 2016-05-10T18:29:21+02:00 Bump versions - - - - - ca75b779 by Ben Gamari at 2016-05-11T16:03:44+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 4e3cfd62 by Ben Gamari at 2016-05-11T16:06:45+02:00 Merge remote-tracking branch 'RyanGlScott/ghc-head' into ghc-head - - - - - a2379970 by Ben Gamari at 2016-05-11T23:15:11+02:00 doc: Add clean targets - - - - - f275212e by Ben Gamari at 2016-05-11T23:15:14+02:00 doc: Add html as an all-target for ghc Otherwise the html documentation won't be installed for binary-dist. - - - - - 388fc0af by Ben Gamari at 2016-05-12T09:49:12+02:00 Update CHANGES - - - - - bad81ad5 by Ben Gamari at 2016-05-12T09:49:38+02:00 Version bump - - - - - c01688a7 by Ben Gamari at 2016-05-12T10:04:58+02:00 Revert "Version bump" This bump was a bit premature. This reverts commit 7b238d9c5be9b07aa2d10df323b5c7b8d1634dc8. - - - - - 7ed05724 by Ben Gamari at 2016-05-12T10:05:33+02:00 doc: Fix GHC clean rule Apparently GHC's build system doesn't permit wildcards in clean paths. - - - - - 5d9611f4 by Ben Gamari at 2016-05-12T17:43:50+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 653566b2 by Ben Gamari at 2016-05-14T09:57:31+02:00 Version bump to 2.17.2 - - - - - b355c439 by Ben Gamari at 2016-05-14T09:57:51+02:00 doc: Use `$(MAKE)` instead of `make` This is necessary to ensure we use gmake. - - - - - 8a18537d by Ben Gamari at 2016-05-14T10:15:45+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - b3290ef1 by Sebastian Meric de Bellefon at 2016-05-14T11:29:47-04:00 Fix haskell/haddock#303. Hide footer when printing The "Produced by Haddock" footer was overlapping the page's body when printing. This patch hides the footer with a css media rule. - - - - - b4a76f89 by Sebastian Meric de Bellefon at 2016-05-15T02:12:46-04:00 Fix haskell/haddock#280. Parsing of module header The initial newlines were counted as indentation spaces, thus disturbing the parsing of next lines - - - - - ba797c9e by Ben Gamari at 2016-05-16T14:53:46+02:00 doc: Vendorize alabaster Sphinx theme Alabaster is now the default sphinx theme and is a significant improvement over the previous default that it's worthproviding it when unavailable (e.g. Sphinx <1.3). - - - - - c9283e44 by Ben Gamari at 2016-05-16T14:55:17+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 1c9ea198 by Sebastian Méric de Bellefon at 2016-05-16T12:30:40-04:00 Merge pull request haskell/haddock#502 from Helkafen/master Fix haskell/haddock#303. Hide footer when printing - - - - - 33631016 by Ben Gamari at 2016-05-16T19:56:11+02:00 Revert "doc: Vendorize alabaster Sphinx theme" This ended up causes far too many issues to be worthwhile. We'll just have to live with inconsistent haddock documentation. This reverts commit cec21957001143794e71bcd9420283df18e7de40. - - - - - 93317d26 by Ben Gamari at 2016-05-16T19:56:11+02:00 cabal: Fix README path - - - - - c8695b22 by Ben Gamari at 2016-05-16T19:58:51+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 0b50eaaa by Ben Gamari at 2016-05-16T21:02:08+02:00 doc: Use whichever theme sphinx deems appropriate - - - - - 857c1c9c by Ben Gamari at 2016-05-16T21:07:08+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 15fc5637 by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Remove redundant imports - - - - - 132ddc6a by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Better debug output For tracking down haskell/haddock#505 - - - - - 2252a149 by Ben Gamari at 2016-05-22T12:43:59+02:00 Don't consider default class ops when looking for decls When we are looking for an operation within a class we don't care about `default`-type declarations. This was the cause of haskell/haddock#505. - - - - - 4886b2ec by Oleg Grenrus at 2016-05-24T16:19:48+03:00 UnfelpfulSpan line number omitted Kind of resolves https://github.com/haskell/haddock/issues/508 - - - - - a4befd36 by Oleg Grenrus at 2016-05-24T16:53:35+03:00 Change Hyperlinked lexer to know about DataKinds ticks - - - - - f45cb52e by David Feuer at 2016-05-24T18:48:53-04:00 Make parser state a newtype Previously, it was `data` wrapping a `Maybe`, which seems a bit silly. Obviously, this can be changed back if anyone wants to add more fields some day. - - - - - 05013dd7 by Sebastian Meric de Bellefon at 2016-05-24T22:03:55-04:00 remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) Frames are a bit broken, ignored by Hackage, and considered obsolete in general. This patch disables frames generation. The mini_*.html files are still used in the synopsis. - - - - - b8163a88 by Ben Gamari at 2016-05-25T14:44:15+02:00 Merge pull request haskell/haddock#507 from bgamari/T505 Fix haskell/haddock#505 - - - - - ea1b30c6 by Sebastian Meric de Bellefon at 2016-05-25T14:17:00-04:00 Update CHANGES - - - - - eddfc258 by Sebastian Méric de Bellefon at 2016-05-25T15:17:40-04:00 Merge pull request haskell/haddock#514 from Helkafen/frames remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) - - - - - 0e506818 by Alex Biehl at 2016-05-26T12:43:09+02:00 Remove misplaced haddock comment - - - - - a07d28c0 by Ben Gamari at 2016-05-27T11:34:59+02:00 Merge pull request haskell/haddock#515 from alexbiehl/master Remove misplaced haddock comment - - - - - 9001d267 by Ben Gamari at 2016-05-27T11:35:46+02:00 Merge pull request haskell/haddock#513 from treeowl/newtype-since Make parser state a newtype - - - - - 74e1a018 by Sebastian Méric de Bellefon at 2016-05-28T17:28:15-04:00 Merge pull request haskell/haddock#504 from Helkafen/issue-280 Fix haskell/haddock#280. Parsing of module header - - - - - 37557f4f by Alan Zimmerman at 2016-05-29T23:36:50+02:00 Matching changes for haskell/haddock#12105 - - - - - 7d09e5d6 by Sebastian Meric de Bellefon at 2016-06-03T18:07:48-04:00 Version bumps (2.17.3, 1.4.2) - - - - - 85b4bc15 by Sebastian Méric de Bellefon at 2016-06-06T18:35:13-04:00 Merge pull request haskell/haddock#521 from Helkafen/master Version bumps (2.17.3, 1.4.2) - - - - - e95f0dee by Sebastian Meric de Bellefon at 2016-06-06T19:11:35-04:00 publish haddock-test library - - - - - 4de40586 by Sebastian Méric de Bellefon at 2016-06-06T20:26:30-04:00 Merge pull request haskell/haddock#512 from phadej/oleg-fixes Fixes for haskell/haddock#508 and haskell/haddock#510 - - - - - ddfd0789 by Dominic Steinitz at 2016-06-09T09:27:28+01:00 Documentation for LaTeX markup. - - - - - 697a503a by Dominic Steinitz at 2016-06-09T09:33:59+01:00 Fix spelling mistake. - - - - - 246f6fff by Dominic Steinitz at 2016-06-09T09:37:15+01:00 Camel case MathJax. - - - - - 4684bd23 by Dominic Steinitz at 2016-06-09T09:44:53+01:00 Fix math typo and add link. - - - - - f20c037c by Simon Peyton Jones at 2016-06-13T18:26:03+01:00 Follow changes to LHsSigWcType - - - - - 0c58996d by Simon Peyton Jones at 2016-06-15T12:56:01+01:00 Follow GHC re-adding FunTy - - - - - 401b5ca7 by Sebastian Méric de Bellefon at 2016-06-15T12:16:47-04:00 Merge pull request haskell/haddock#525 from idontgetoutmuch/master Documentation for LaTeX markup. - - - - - 92d263b7 by Sebastian Méric de Bellefon at 2016-06-15T12:17:29-04:00 Merge pull request haskell/haddock#522 from Helkafen/master publish haddock-test library - - - - - 0953a2ca by Sebastian Meric de Bellefon at 2016-06-16T00:46:46-04:00 Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 65453e14 by Ben Gamari at 2016-06-16T11:16:32+02:00 ocean: Ensure that synopsis fully covers other content Previously MathJax content was being rendered on top of the synopsis due to ambiguous z-ordering. Here we explicitly give the synopsis block a higher z-index to ensure it is rendered on top. Fixes haskell/haddock#531. - - - - - 68e411a1 by Sebastian Méric de Bellefon at 2016-06-16T23:34:39-04:00 Merge pull request haskell/haddock#534 from bgamari/T531 ocean: Ensure that synopsis fully covers other content - - - - - fad6491b by Sebastian Méric de Bellefon at 2016-06-18T23:57:20-04:00 Merge pull request haskell/haddock#533 from Helkafen/master Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 6108e21b by Sebastian Meric de Bellefon at 2016-06-22T23:08:28-04:00 do not create empty src directory Fix haskell/haddock#536. - - - - - 1ef23823 by Sebastian Méric de Bellefon at 2016-06-24T00:04:48-04:00 Merge pull request haskell/haddock#537 from Helkafen/master do not create empty src directory - - - - - 966baa96 by Omari Norman at 2016-06-29T21:59:34-04:00 Add $ as a special character If this character is not escaped, documentation built with Haddock 2.17.2 will fail. This was not an issue with 2.16 series, which causes builds to fail and there is nothing in the docs or error message giving a clue about why builds that used to succeed now don't. - - - - - 324adb60 by Ben Gamari at 2016-07-01T12:18:51+02:00 GhcUtils: Changes for multi-pattern signatures - - - - - d7571675 by Ömer Sinan Ağacan at 2016-07-21T13:30:47+02:00 Add support for unboxed sums - - - - - 29d0907b by Simon Marlow at 2016-07-22T13:55:48+01:00 Disable NFData instances for GHC types when GHC >= 8.2 - - - - - 702d95f3 by Simon Marlow at 2016-08-02T15:57:30+02:00 Disable NFData instances for GHC types when GHC >= 8.0.2 (cherry picked from commit a3309e797c42dae9bccdeb17ce52fcababbaff8a) - - - - - f4fa79c3 by Ben Gamari at 2016-08-07T13:51:18+02:00 ghc.mk: Don't attempt to install html/frames.html The frames business has been removed. - - - - - 9cd63daf by Ben Gamari at 2016-08-07T13:51:40+02:00 Haddock.Types: More precise version guard This allows haddock to be built with GHC 8.0.2 pre-releases. - - - - - f3d7e03f by Mateusz Kowalczyk at 2016-08-29T20:47:45+01:00 Merge pull request haskell/haddock#538 from massysett/master Add $ as a special character - - - - - 16dbf7fd by Bartosz Nitka at 2016-09-20T19:44:04+01:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 - - - - - 7c31c1ff by Bartosz Nitka at 2016-09-27T17:32:22-04:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 (cherry picked from commit 073d899a8f94ddec698f617a38d3420160a7fd0b) - - - - - 33a90dce by Ryan Scott at 2016-09-30T20:53:41-04:00 Haddock changes for T10598 See https://ghc.haskell.org/trac/ghc/ticket/10598 - - - - - 1f32f7cb by Ben Gamari at 2016-10-13T20:01:26-04:00 Update for refactoring of NameCache - - - - - 1678ff2e by Ben Gamari at 2016-11-15T17:42:48-05:00 Bump upper bound on base - - - - - 9262a7c5 by Alan Zimmerman at 2016-12-07T21:14:28+02:00 Match changes in GHC wip/T3384 branch - - - - - ac0eaf1a by Ben Gamari at 2016-12-09T09:48:41-05:00 haddock-api: Don't use stdcall calling convention on 64-bit Windows See GHC haskell/haddock#12890. - - - - - 04afe4f7 by Alan Zimmerman at 2016-12-12T20:07:21+02:00 Matching changes for GHC wip/T12942 - - - - - e1d1701d by Ben Gamari at 2016-12-13T16:50:41-05:00 Bump base upper bound - - - - - 3d3eacd1 by Alan Zimmerman at 2017-01-10T16:59:38+02:00 HsIParamTy now has a Located name - - - - - 7dbceefd by Kyrill Briantsev at 2017-01-12T13:23:50+03:00 Prevent GHC API from doing optimization passes. - - - - - d48d1e33 by Richard Eisenberg at 2017-01-19T08:41:41-05:00 Upstream changes re levity polymorphism - - - - - 40c25ed6 by Alan Zimmerman at 2017-01-26T15:16:18+02:00 Changes to match haskell/haddock#13163 in GHC - - - - - 504f586d by Ben Gamari at 2017-02-02T17:19:37-05:00 Kill remaining static flags - - - - - 49147ea0 by Justus Adam at 2017-03-02T15:33:34+01:00 Adding MDoc to exports of Documentation.Haddock - - - - - 1cfba9b4 by Justus Adam at 2017-03-09T11:41:44+01:00 Also exposing toInstalledIface - - - - - 53f0c0dd by Ben Gamari at 2017-03-09T13:10:08-05:00 Bump for GHC 8.3 - - - - - c7902d2e by Ben Gamari at 2017-03-09T23:46:02-05:00 Bump for GHC 8.2 - - - - - 4f3a74f8 by Ben Gamari at 2017-03-10T10:21:55-05:00 Merge branch 'ghc-head' - - - - - e273b72f by Richard Eisenberg at 2017-03-14T13:34:04-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - 6ec3d436 by Richard Eisenberg at 2017-03-14T15:15:52-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - eee3cda1 by Ben Gamari at 2017-03-15T15:19:59-04:00 Adapt to EnumSet - - - - - 017cf58e by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 736d6773 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 475f84a0 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 13240b53 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - cd16d529 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 3bea97ae by Edward Z. Yang at 2017-03-15T22:50:46-07:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b2b051ce by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 0f082795 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 20ef63c9 by Edward Z. Yang at 2017-03-22T13:48:12-07:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 45692dcb by Edward Z. Yang at 2017-03-22T14:11:25-07:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4eae8caf by Ben Gamari at 2017-03-23T09:25:33-04:00 Merge commit '240bc38b94ed2d0af27333b23392d03eeb615e82' into HEAD - - - - - 0bbe03f5 by Ben Gamari at 2017-03-23T09:27:28-04:00 haddock-api: Bump bound on GHC - - - - - 65f3ac9d by Alex Biehl at 2017-03-23T17:36:11+01:00 Merge pull request haskell/haddock#581 from JustusAdam/master Adding more exports to Documentation.Haddock - - - - - 37d49a47 by Alex Biehl at 2017-03-23T17:39:14+01:00 Merge pull request haskell/haddock#568 from awson/ghc-head Prevent GHC API from doing optimization passes. - - - - - 1ed047e4 by Brian Huffman at 2017-03-23T17:45:58+01:00 Print any user-supplied kind signatures on type parameters. This applies to type parameters on data, newtype, type, and class declarations, and also to forall-bound type vars in type signatures. - - - - - 1b78ca5c by Brian Huffman at 2017-03-23T17:45:58+01:00 Update test suite to expect kind annotations on type parameters. - - - - - a856b162 by Alex Biehl at 2017-03-23T17:49:32+01:00 Include travis build indication badge - - - - - 8e2e2c56 by Ben Gamari at 2017-03-23T17:20:08-04:00 haddock-api: Bump bound on GHC - - - - - 4d2d9995 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 26d6c150b31bc4580ab17cfd07b6e7f9afe10737) - - - - - a650e20f by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 930cfbe58e2e87f5a4d431d89a3c204934e6e858) - - - - - caa282c2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 2067a2d0afa9cef381d26fb7140b67c62f433fc0) - - - - - 49684884 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 0671abfe7e8ceae2269467a30b77ed9d9656e2cc) - - - - - 4dcfeb1a by Edward Z. Yang at 2017-03-23T17:20:08-04:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 3d77b373dd5807d5d956719dd7c849a11534fa6a) - - - - - 74dd19d2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 94610e9b446324f4231fa6ad4c6ac51e4eba8c0e) - - - - - a9b19a23 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 6cc832dfb1de6088a4abcaae62b25a7e944d55c3) - - - - - d3631064 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit bf3c4d72a0fda38561376eac7eda216158783267) - - - - - ef2148fc by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 07b88c5d4e79b87a319fbb08f8ea01dbb41063c1) - - - - - 2f29518b by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 4eb765ca4205c79539d60b7afa9b7e261a4a49fe) - - - - - 37de047d by Phil Ruffwind at 2017-04-03T11:57:14+02:00 Update MathJax URL MathJax is shutting down their CDN: https://www.mathjax.org/cdn-shutting-down/ They recommend migrating to cdnjs. - - - - - e9d24ba8 by David C. Turner at 2017-04-03T14:58:01+02:00 Add highlight for :target to ocean.css - - - - - 4819a202 by Alex Biehl at 2017-04-11T19:36:48+02:00 Allow base-4.10 for haddock-test - - - - - 44cec69c by Alex Biehl at 2017-04-11T19:39:22+02:00 cabal.project for haddock-api, haddock-library and haddock-test - - - - - 935d0f6a by Alex Biehl at 2017-04-11T19:46:29+02:00 Move dist scripts to scripts/ - - - - - 128e150c by Alex Biehl at 2017-04-11T20:34:46+02:00 Add haddock to cabal.project - - - - - cc8e08ea by Alex Biehl at 2017-04-11T20:35:08+02:00 Read files for hyperlinker eagerly This also exposes Documentation.Haddock.Utf8 - - - - - 152dda78 by Alex Biehl at 2017-04-11T20:37:06+02:00 Explicit import list ofr Control.DeepSeq in Haddock.Interface.Create - - - - - 501b33c4 by Kyrill Briantsev at 2017-04-11T21:01:42+02:00 Prevent GHC API from doing optimization passes. - - - - - c9f3f5ff by Alexander Biehl at 2017-04-12T16:36:53+02:00 Add @alexbiehl as maintaner - - - - - 76f214cc by Alex Biehl at 2017-04-13T07:27:18+02:00 Disable doctest with ghc-8.3 Currently doctest doesn't support ghc-head - - - - - 46b4f5fc by Edward Z. Yang at 2017-04-22T20:38:26-07:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f0555235 by Alex Biehl at 2017-04-25T10:08:48+02:00 Travis: Use ghc-8.2.1 on master - - - - - 966ea348 by Alex Biehl at 2017-04-25T10:32:01+02:00 Travis: Verbose cabal output cf. https://travis-ci.org/haskell/haddock/jobs/225512194#L377 - - - - - 36972bcd by Alex Biehl at 2017-04-25T10:40:43+02:00 Use travis_retry for cabal invocations - - - - - b3a09d2c by Alex Biehl at 2017-04-25T17:02:20+02:00 Use new MathJax URL in html-test 18ed871afb82560d5433b2f53e31b4db9353a74e switched to a new MathJax URL but didn't update the tests. - - - - - ae331e5f by Alexander Biehl at 2017-04-25T17:02:20+02:00 Expand signatures for class declarations - - - - - e573c65a by Alexander Biehl at 2017-04-25T17:02:20+02:00 Hoogle: Correctly print classes with associated data types - - - - - 3fc6be9b by Edward Z. Yang at 2017-04-25T17:02:20+02:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit a0c4790e15a2d3fab8d830eee8fcd639fe6d39c9) - - - - - 6725c060 by Herbert Valerio Riedel at 2017-04-25T17:02:20+02:00 `html-test --accept` deltas to reference samples - - - - - 7d444d61 by Alex Biehl at 2017-04-26T07:13:50+02:00 Remove anything related to obsolete frames mode - - - - - b888972c by Alex Biehl at 2017-04-26T07:49:10+02:00 Cherry-picked remaining commits from haddock-2.17.4-release (#603) * Release haddock/haddock-api 2.17.4 and haddock-library 1.4.3 * Set version bounds for haddock-library NB: This allows GHC 8.2.1's base * Set version bounds for haddock & haddock-api The version bounds support GHC 8.2 * Merge (temporary) v2.17.3 branch into v2.17 This allows us to delete the v2.17.3 branch * Fixup changelog * Pin down haddock-api to a single version as otherwise `haddock`'s package version has no proper meaning * fix source-repo spec for haddock-api - - - - - 4161099b by Alex Biehl at 2017-04-26T11:11:20+02:00 Update changelog to reflect news in HEAD - - - - - eed72cb8 by Alex Biehl at 2017-04-26T11:11:20+02:00 Markdownify changelog - - - - - 5815cea1 by Alex Biehl at 2017-04-26T11:32:33+02:00 Bump to 2.18.0 (#605) - - - - - a551d558 by Alex Biehl at 2017-04-29T22:00:25+02:00 Update attoparsec-0.12.1.1 to attoparsec-0.13.1.0 - - - - - ea164a8d by Sergey Vinokurov at 2017-04-29T22:42:36+02:00 Improve error message - - - - - 2e10122f by Alex Biehl at 2017-04-30T10:07:46+02:00 Correctly remember collapsed sections (#608) Now the "collapsed" cookie stores which sections have changed state instead of which are collapsed. - - - - - f9b24d99 by Alex Biehl at 2017-05-01T17:40:36+02:00 Lazily decode docMap and argMap (#610) These are only used in case of a doc reexport so most of the time decoding these is wasted work. - - - - - 2372af62 by Alex Biehl at 2017-05-01T21:59:23+02:00 Fix Binary instance for InstalledInterface (#611) (#610) introduced lazy decoding for docs from InstalledInterface but forgot to remove the original calls to get and put_ - - - - - 6c633c13 by Nathan Collins at 2017-05-11T11:47:55+02:00 Improve documenation of Haddock markup (#614) * Improve documentation of Haddock markup. - document that Haddock supports inferring types top-level functions with without type signatures, but also explain why using this feature is discouraged. Looks like this feature has been around since version 2.0.0.0 in 2008! - rework the "Module description" section: - move the general discussion of field formatting to the section intro and add examples illustrating the prose for multiline fields. - mention that newlines are preserved in some multiline fields, but not in others (I also noticed that commas in the `Copyright` field are not preserved; I'll look into this bug later). - add a subsection for the module description fields documentation, and put the field keywords in code formatting (double back ticks) instead of double quotes, to be consistent with the typesetting of keywords in other parts of the documentation. - mention that "Named chunks" are not supported in the long-form "Module description" documentation. - fix formatting of keywords in the "Module attributes" section. Perhaps these errors were left over from an automatic translation to ReST from some other format as part of the transition to using Sphinx for Haddock documentation? Also, add a missing reference here; it just said "See ?"! - update footnote about special treatment for re-exporting partially imported modules not being implemented. In my tests it's not implemented at all -- I tried re-exporting both `import B hiding (f)` and `import B (a, b)` style partial imports, and in both cases got the same result as with full imports `import B`: I only get a module reference. * Rework the `Controlling the documentation structure` section. My main goal was to better explain how to use Haddock without an export list, since that's my most common use case, but I hope I improved the section overall: - remove the incomplete `Omitting the export list` section and fold it into the other sections. In particular, summarize the differences between using and not using an export list -- i.e. control over what and in what order is documented -- in the section lead. - add "realistic" examples that use the structure markup, both with and without an export list. I wanted a realistic example here to capture how it can be useful to explain the relationship between a group of functions in a section, in addition to documenting their individual APIs. - make it clear that you can associate documentation chunks with documentation sections when you aren't using an export list, and that doing it in the most obvious way -- i.e. with `-- |`, as you can in the export list -- doesn't work without an export list. It took me a while to figure this out the first time, since the docs didn't explain it at all before. - add a "no export list" example to the section header section. - add more cross references. * Add examples of gotchas for markup in `@...@`. I'm not sure this will help anyone, since I think most people first learn about `@...@` by reading other people's Haddocks, but I've documented the mistakes which I've made and then gotten confused by. * Use consistent Capitalization of Titles. Some titles were in usual title caps, and others only had the first word capitalized. I chose making them all use title caps because that seems to make the cross references look better. - - - - - d4734f45 by Ben Gamari at 2017-05-12T20:36:08+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 93883f37 by Alex Biehl at 2017-05-12T21:02:33+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 5b8f179c by Alex Biehl at 2017-05-13T12:48:10+02:00 Consequently use inClass and notInClass in haddock-library (#617) These allow attoparsec to do some clever lookup optimization - - - - - 77984b82 by Doug Wilson at 2017-05-27T17:37:38+02:00 Don't enable compilation for template haskell (#624) This is no longer necessary after ghc commit 53c78be0aab76a3107c4dacbb1d177afacdd37fa - - - - - 5a3de2b4 by Doug Wilson at 2017-05-27T19:54:53+02:00 Improve Syb code (#621) Specialize.hs and Ast.hs are modified to have their Syb code not recurse into Name or Id in HsSyn types. Specialize.hs is refactored to have fewer calls to Syb functions. Syb.hs has some foldl calls replaced with foldl' calls. There is still a lot of performance on the floor of Ast.hs. The RenamedSource is traversed many times, and lookupBySpan is very inefficient. everywhereBut and lookupBySpan dominate the runtime whenever --hyperlinked-source is passed. - - - - - 3d35a949 by Alex Biehl at 2017-05-30T19:01:37+02:00 Clear fixme comment (#625) - - - - - 2a44bd0c by Alex Biehl at 2017-05-30T19:02:12+02:00 Make haddock-library and haddock-api warning free (#626) - - - - - bd1a0e42 by Alex Biehl at 2017-06-01T10:40:33+02:00 Include `driver-test/*.hs` sdist (#630) This lead to haskell/haddock#629. - - - - - 184a3ab6 by Doug Wilson at 2017-06-03T12:02:08+02:00 Disable pattern match warnings (#628) This disables the pattern match checker which can be very expensive in some cases. The disabled warnings include: * Opt_WarnIncompletePatterns * Opt_WarnIncompleteUniPatterns * Opt_WarnIncompletePatternsRecUpd * Opt_WarnOverlappingPatterns - - - - - 0cf68004 by Alex Biehl at 2017-06-03T20:37:28+02:00 Allow user defined signatures for pattern synonyms (#631) - - - - - 7f51a58a by Alex Biehl at 2017-06-04T11:56:38+02:00 Use NameSet for isExported check (#632) - - - - - d8f044a9 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Match new AST as per GHC wip/new-tree-one-param See https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow - - - - - da1254e3 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Rename extension index tags - - - - - 538c7514 by Christiaan Baaij at 2017-06-09T08:26:43+02:00 Haddock support for bundled pattern synonyms (#627) * Haddock support for bundled pattern synonyms * Add fixities to bundled pattern synonyms * Add bundled pattern synonyms to the synopsis * Store bundled pattern fixities in expItemFixities * Add test for bundled pattern synonyms * Stop threading fixities * Include bundled pattern synonyms for re-exported data types Sadly, fixity information isn't found for re-exported data types * Support for pattern synonyms * Modify tests after haskell/haddock#631 * Test some reexport variations * Also lookup bundled pattern synonyms from `InstalledInterface`s * Check isExported for bundled pattern synonyms * Pattern synonym is exported check * Always look for pattern synonyms in the current module Another overlooked cornercase * Account for types named twice in export lists Also introduce a fast function for nubbing on a `Name` and use it throughout the code base. * correct fixities for reexported pattern synonyms * Fuse concatMap and map * Remove obsolete import * Add pattern synonyms to visible exports * Fix test * Remove corner case - - - - - a050bffd by Doug Wilson at 2017-06-21T09:27:33+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#636) There is some performance improvement. GHC compiler: | version | bytes allocated | cpu_seconds --------------------------------- | before | 56057108648 | 41.0 | after | 51592019560 | 35.1 base: | version | bytes allocated | cpu_seconds --------------------------------- | before | 25174011784 | 14.6 | after | 23712637272 | 13.1 Cabal: | version | bytes allocated | cpu_seconds --------------------------------- | before | 18754966920 | 12.6 | after | 18198208864 | 11.6 - - - - - 5d06b871 by Doug Wilson at 2017-06-22T20:23:29+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#639) * Use new function getNameToInstancesIndex instead of tcRnGetInfo There is some significant performance improvement in the ghc testsuite. haddock.base: -23.3% haddock.Cabal: -16.7% haddock.compiler: -19.8% * Remove unused imports - - - - - b11bb73a by Alex Biehl at 2017-06-23T14:44:41+02:00 Lookup fixities for reexports without subordinates (#642) So we agree that reexported declarations which do not have subordinates (for example top-level functions) shouldn't have gotten fixities reexported according to the current logic. I wondered why for example Prelude.($) which is obviously reexported from GHC.Base has fixities attached (c.f. http://hackage.haskell.org/package/base-4.9.1.0/docs/Prelude.html#v:-36-). The reason is this: In mkMaps we lookup all the subordinates of top-level declarations, of course top-level functions don't have subordinates so for them the resulting list is empty. In haskell/haddock#644 I established the invariant that there won't be any empty lists in the subordinate map. Without the patch from haskell/haddock#642 top-level functions now started to fail reexporting their fixities. - - - - - d2a6dad6 by Alex Biehl at 2017-06-23T18:30:45+02:00 Don't include names with empty subordinates in maps (#644) These are unecessary anyway and just blow up interface size - - - - - 69c2aac4 by Alex Biehl at 2017-06-29T19:54:49+02:00 Make per-argument docs for class methods work again (#648) * Make per-argument docs for class methods work again * Test case - - - - - c9448d54 by Bartosz Nitka at 2017-07-02T12:12:01+02:00 Fix haddock: internal error: links: UnhelpfulSpan (#561) * Fix haddock: internal error: links: UnhelpfulSpan This fixes haskell/haddock#554 for me. I believe this is another fall out of `wildcard-refactor`, like haskell/haddock#549. * Comment to clarify why we take the methods name location - - - - - d4f29eb7 by Alex Biehl at 2017-07-03T19:43:04+02:00 Document record fields when DuplicateRecordFields is enabled (#649) - - - - - 9d6e3423 by Yuji Yamamoto at 2017-07-03T22:37:58+02:00 Fix test failures on Windows (#564) * Ignore .stack-work * Fix for windows: use nul instead of /dev/null * Fix for windows: canonicalize line separator * Also normalize osx line endings - - - - - 7d81e8b3 by Yuji Yamamoto at 2017-07-04T16:13:12+02:00 Avoid errors on non UTF-8 Windows (#566) * Avoid errors on non UTF-8 Windows Problem ==== haddock exits with errors like below: `(1)` ``` haddock: internal error: <stderr>: hPutChar: invalid argument (invalid character) ``` `(2)` ``` haddock: internal error: Language\Haskell\HsColour\Anchors.hs: hGetContents: invalid argument (invalid byte sequence) ``` `(1)` is caused by printing [the "bullet" character](http://www.fileformat.info/info/unicode/char/2022/index.htm) onto stderr. For example, this warning contains it: ``` Language\Haskell\HsColour\ANSI.hs:62:10: warning: [-Wmissing-methods] • No explicit implementation for ‘toEnum’ • In the instance declaration for ‘Enum Highlight’ ``` `(2)` is caused when the input file of `readFile` contains some Unicode characters. In the case above, '⇒' is the cause. Environment ---- OS: Windows 10 haddock: 2.17.3 GHC: 8.0.1 Solution ==== Add `hSetEncoding handle utf8` to avoid the errors. Note ==== - I found the detailed causes by these changes for debugging: - https://github.com/haskell/haddock/commit/8f29edb6b02691c1cf4c479f6c6f3f922b35a55b - https://github.com/haskell/haddock/commit/1dd23bf2065a1e1f2c14d0f4abd847c906b4ecb4 - These errors happen even after executing `chcp 65001` on the console. According to the debug code, `hGetEncoding stderr` returns `CP932` regardless of the console encoding. * Avoid 'internal error: <stderr>: hPutChar: invalid argument (invalid character)' non UTF-8 Windows Better solution for 59411754a6db41d17820733c076e6a72bcdbd82b's (1) - - - - - eded67d2 by Alex Biehl at 2017-07-07T19:17:15+02:00 Remove redudant import warning (#651) - - - - - 05114757 by Alex Biehl at 2017-07-08T00:33:12+02:00 Avoid missing home module warning (#652) * Avoid missing home module warning * Update haddock-library.cabal - - - - - e9cfc902 by Bryn Edwards at 2017-07-17T07:51:20+02:00 Fix haskell/haddock#249 (#655) - - - - - eb02792b by Herbert Valerio Riedel at 2017-07-20T09:09:15+02:00 Fix compilation of lib:haddock-library w/ GHC < 8 - - - - - 9200bfbc by Alex Biehl at 2017-07-20T09:20:38+02:00 Prepare 2.18.1 release (#657) - - - - - 46ddd22c by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Tweak haddock-api.cabal for pending release - - - - - 85e33d29 by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Avoid trivial use of LambdaCase otherwise we can't test w/ e.g. GHC 7.4.2 - - - - - 3afb4bfe by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Refactor .cabal to use sub-lib for vendored lib A practical benefit is that we can control the build-depends and also avoid some recompilation between library and test-suite. - - - - - e56a552e by Herbert Valerio Riedel at 2017-07-20T10:17:48+02:00 haddock-api: add changelog pointing to haddock's changelog This addresses https://github.com/haskell/haddock/issues/638#issuecomment-309283297 - - - - - 2222ff0d by Herbert Valerio Riedel at 2017-07-20T10:19:56+02:00 Drop obsolete/misleading `stability: experimental` This .cabal property has long been considered obsolete - - - - - 9b882905 by Alex Biehl at 2017-07-20T11:25:54+02:00 Beef up haddock description (#658) * Beef up haddock description * Handle empty lines - - - - - bb60e95c by Herbert Valerio Riedel at 2017-07-20T12:08:53+02:00 Import @aisamanra's Haddock cheatsheet from https://github.com/aisamanra/haddock-cheatsheet - - - - - 0761e456 by Herbert Valerio Riedel at 2017-07-20T12:12:55+02:00 Add cheatsheet to haddock.cabal - - - - - 2ece0f0f by Herbert Valerio Riedel at 2017-07-20T12:18:38+02:00 Mention new-build in README - - - - - 947b7865 by Herbert Valerio Riedel at 2017-07-20T12:32:16+02:00 Update README Also improves markup and removes/fixes redundant/obsolete parts [skip ci] - - - - - 785e09ad by Alex Biehl at 2017-07-27T07:28:57+02:00 Bump haddock to 2.18.2, haddock-library to 1.4.5 - - - - - e3ff1ca3 by Alex Biehl at 2017-07-31T20:15:32+02:00 Move `DocMarkup` from haddock-api to haddock-library (#659) * Move `DocMarkup` from haddock-api to haddock-library * Move more markup related functions * Markup module * CHANGELOG - - - - - cda7c20c by Alex Biehl at 2017-07-31T20:35:49+02:00 Fixup haddock - - - - - 583b6812 by Alex Biehl at 2017-07-31T21:20:45+02:00 Changelog for haddock-library - - - - - bac6a0eb by Alex Biehl at 2017-07-31T21:50:24+02:00 Prepare haddock-library-1.4.5 release - - - - - 58ce6877 by Moritz Drexl at 2017-08-05T16:44:40+02:00 Fix renaming after instance signature specializing (#660) * rework rename * Add regression test for Bug 613 * update tests * update changelog - - - - - b8137ec8 by Tim Baumann at 2017-08-06T11:33:38+02:00 Fix: Generate pattern signatures for constructors exported as patterns (#663) * Fix pretty-printing of pattern signatures Pattern synonyms can have up to two contexts, both having a different semantic meaning: The first holds the constraints required to perform the matching, the second contains the constraints provided by a successful pattern match. When the first context is empty but the second is not it is necessary to render the first, empty context. * Generate pattern synonym signatures for ctors exported as patterns This fixes haskell/haddock#653. * Simplify extractPatternSyn It is not necessary to generate the simplest type signature since it will be simplified when pretty-printed. * Add changelog entries for PR haskell/haddock#663 * Fix extractPatternSyn error message - - - - - d037086b by Alex Biehl at 2017-08-06T12:43:25+02:00 Bump haddock-library - - - - - 99d7e792 by Alex Biehl at 2017-08-06T12:44:07+02:00 Bump haddock-library in haddock-api - - - - - 94802a5b by Alex Biehl at 2017-08-06T13:18:02+02:00 Provide --show-interface option to dump interfaces (#645) * WIP: Provide --show-interface option to dump interfaces Like ghcs own --show-iface this flag dumps a binary interface file to stdout in a human (and machine) readable fashion. Currently it uses json as output format. * Fill all the jsonNull stubs * Rework Bifunctor instance of DocH, update changelog and documentation * replace changelog, bring DocMarkupH doc back * Update CHANGES.md * Update CHANGES.md * Move Control.Arrow up It would result in unused import if the Bifunctor instance is not generated. - - - - - c662e476 by Ryan Scott at 2017-08-14T21:00:21-04:00 Adapt to haskell/haddock#14060 - - - - - b891eb73 by Alex Biehl at 2017-08-16T08:24:48+02:00 Bifoldable and Bitraversable for DocH and MetaDoc - - - - - 021bb56c by Alex Biehl at 2017-08-16T09:06:40+02:00 Refactoring: Make doc renaming monadic This allows us to later throw warnings if can't find an identifier - - - - - 39fbf022 by Alex Biehl at 2017-08-19T20:35:27+02:00 Hyperlinker: Avoid linear lookup in enrichToken (#669) * Make Span strict in Position * Hyperlinker: Use a proper map to enrich tokens - - - - - e13baedd by Alex Biehl at 2017-08-21T20:05:42+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 27dd6e87 by Alex Biehl at 2017-08-21T22:06:35+02:00 Drop Avails from export list - - - - - 86b247e2 by Alex Biehl at 2017-08-22T08:44:22+02:00 Bump ghc version for haddock-api tests - - - - - d4607ca0 by Alex Biehl at 2017-08-22T08:45:17+02:00 Revert "Drop Avails from export list" This reverts commit a850ba86d88a4fb9c0bd175453a2580e544e3def. - - - - - c9c54c30 by Alex Biehl at 2017-08-22T09:26:01+02:00 IntefaceFile version - - - - - a85b7c02 by Ben Gamari at 2017-08-22T09:29:52-04:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 34e976f5 by Ben Gamari at 2017-08-22T17:40:06+02:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 577abf06 by Ryan Scott at 2017-08-23T14:47:29-04:00 Update for haskell/haddock#14131 - - - - - da68fc55 by Florian Eggenhofer at 2017-08-27T18:21:56+02:00 Generate an index for package content search (#662) Generate an index for package content search - - - - - 39e62302 by Alex Biehl at 2017-08-27T18:50:16+02:00 Content search for haddock html doc - - - - - 91fd6fb2 by Alex Biehl at 2017-08-28T18:39:58+02:00 Fix tests for content search - - - - - b4a3798a by Alex Biehl at 2017-08-28T18:44:08+02:00 Add search button to #page-menu - - - - - 25a7ca65 by Alex Biehl at 2017-08-28T18:47:43+02:00 Load javascript below the fold - - - - - 8d323c1a by Alex Biehl at 2017-08-28T18:49:22+02:00 Accept tests - - - - - c5dac557 by Alex Biehl at 2017-08-28T19:14:55+02:00 Content search css - - - - - 89a5af57 by Paolo Veronelli at 2017-08-29T07:42:13+02:00 Removed `nowrap` for interface method sigs (#674) with nowrap the interfaces method sigs would expand at libitum - - - - - a505f6f7 by Alex Biehl at 2017-08-29T08:05:33+02:00 Include subordinates in content index - - - - - 4bb698c4 by Alexander Biehl at 2017-08-29T11:40:19+02:00 QuickNav: Make docbase configurable - - - - - c783bf44 by Alexander Biehl at 2017-08-29T11:48:36+02:00 QuickNav: Also use baseUrl for doc-index.json request - - - - - 47017510 by Alex Biehl at 2017-08-29T17:56:47+02:00 Fix test fallout (again) - - - - - 924fc318 by Alex Biehl at 2017-08-30T09:24:56+02:00 Write meta.json when generating html output (#676) - - - - - 717dea52 by Alex Biehl at 2017-09-01T09:20:34+02:00 Use relative URL when no docBaseUrl given - - - - - e5d85f3b by Alex Biehl at 2017-09-01T09:35:19+02:00 Add missing js files to data-files (#677) - - - - - 95b9231a by Alex Biehl at 2017-09-01T11:01:36+02:00 Rename "Search" tab to "Quick Jump" - - - - - da0ead0b by Alex Biehl at 2017-09-01T13:03:49+02:00 Make trigger link configurable (#678) QuickNav: Configurable show/hide trigger - - - - - de7da594 by Ben Gamari at 2017-09-05T06:49:55-04:00 Account for "Remember the AvailInfo for each IE" As of GHC commit f609374a55bdcf3b79f3a299104767aae2ffbf21 GHC retains the AvailInfo associated with each IE. @alexbiehl has a patch making proper use of this change, but this is just to keep things building. - - - - - b05cd3b3 by Ben Gamari at 2017-09-14T07:55:07-04:00 Bump upper bound on base - - - - - 79db899e by Herbert Valerio Riedel at 2017-09-21T23:27:52+02:00 Make compatible with Prelude.<> export in GHC 8.4/base-4.11 - - - - - 3405dd52 by Tim Baumann at 2017-09-23T22:02:01+02:00 Add compile step that bundles and compresses JS files (#684) * Add compile step that bundles and compresses JS files Also, manage dependencies on third-party JS libraries using NPM. * Compile JS from TypeScript * Enable 'noImplicitAny' in TypeScript * QuickJump: use JSX syntax * Generate source maps from TypeScript for easier debugging * TypeScript: more accurate type * Separate quick jump css file from ocean theme - - - - - df0b5742 by Alex Biehl at 2017-09-29T21:15:40+02:00 Bump base for haddock-library and haddock-test - - - - - 62b12ea0 by Merijn Verstraaten at 2017-10-04T16:03:13+02:00 Inhibit output of coverage information for hidden modules. (#687) * Inhibit output of coverage information for hidden modules. * Add changelog entry. - - - - - 8daf8bc1 by Alexander Biehl at 2017-10-05T11:27:05+02:00 Don't use subMap in attachInstances - - - - - ad75114e by Alexander Biehl at 2017-10-05T11:27:58+02:00 Revert "Don't use subMap in attachInstances" This reverts commit 3adf5bcb1a6c5326ab33dc77b4aa229a91d91ce9. - - - - - 7d4aa02f by Alex Biehl at 2017-10-08T15:32:28+02:00 Precise Haddock: Use Avails for export resolution (#688) * Use Avails for export resolution * Support reexported modules * Factor out availExportItem * Use avails for fullModuleExports * Don't use subMap in attachInstances * lookupDocs without subMap * Completely remove subMap * Only calculate unqualified modules when explicit export list is given * Refactor * Refine comment * return * Fix * Refactoring * Split avail if declaration is not exported itself * Move avail splitting - - - - - b9b4faa8 by Alex Biehl at 2017-10-08T19:38:21+02:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 43325295 by Alex Biehl at 2017-10-08T20:18:46+02:00 Fix merge fallout - - - - - c6423cc0 by Alex Biehl at 2017-10-08T20:36:12+02:00 Copy QuickJump files over - - - - - 1db587c3 by Tim Baumann at 2017-10-09T18:33:09+02:00 Use <details> element for collapsibles (#690) * Remove unnecessary call to 'collapseSection' The call is unnecessary since there is no corresponding toggle for hiding the section of orphan instances. * Use <details> for collapsibles This makes them work even when JS is disabled. Closes haskell/haddock#560. - - - - - 1b54c64b by Tim Baumann at 2017-10-10T09:50:59+02:00 Quick Jump: Show error when loading 'doc-index.json' failed (#691) - - - - - 910f716d by Veronika Romashkina at 2017-10-24T07:36:20+02:00 Fix tiny typo in docs (#693) - - - - - b21de7e5 by Ryan Scott at 2017-10-24T13:07:15+02:00 Overhaul Haddock's rendering of kind signatures (#681) * Overhaul Haddock's rendering of kind signatures * Strip off kind signatures when specializing As an added bonus, this lets us remove an ugly hack specifically for `(->)`. Yay! * Update due to 0390e4a0f61e37bd1dcc24a36d499e92f2561b67 * @alexbiehl's suggestions * Import injectiveVarsOfBinder from GHC - - - - - 6704405c by Ryan Scott at 2017-10-28T07:10:27+02:00 Fix Haddock rendering of kind-indexed data family instances (#694) - - - - - 470f6b9c by Alex Biehl at 2017-10-30T08:45:51+01:00 Add QuickJump version to meta.json (#696) - - - - - b89eccdf by Alex Biehl at 2017-10-30T10:15:49+01:00 Put Quickjump behind --quickjump flag (#697) - - - - - 3095fb58 by Alex Biehl at 2017-10-30T19:09:06+01:00 Add build command to package.json - - - - - f223fda9 by Alex Biehl at 2017-10-30T19:10:39+01:00 Decrease threshold for fuzzy matching - - - - - 80245dda by Edward Z. Yang at 2017-10-31T20:35:05+01:00 Supported reexported-modules via --reexport flag. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 7e389742 by Alex Biehl at 2017-10-31T20:37:56+01:00 Correct missing title in changelog - - - - - 1a2a1c03 by Alex Biehl at 2017-10-31T20:59:07+01:00 Copy quickjump.css for nicer error messages - - - - - db234bb9 by Alex Biehl at 2017-10-31T21:31:18+01:00 Reexported modules: Report warnings if argument cannot be parsed or ... module cannot be found - - - - - eea8a205 by Carlo Hamalainen at 2017-10-31T21:43:14+01:00 More general type for nameCacheFromGhc. (#539) - - - - - 580eb42a by Alex Biehl at 2017-10-31T21:46:52+01:00 Remote tab - - - - - 0e599498 by Alex Biehl at 2017-10-31T21:48:55+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 7b8539bb by Alex Biehl at 2017-10-31T22:28:34+01:00 fullModuleContents traverses exports in declaration order - - - - - 0c91fbf2 by Alex Biehl at 2017-10-31T22:32:31+01:00 Remove excessive use of list comprehensions - - - - - f7356e02 by Alex Biehl at 2017-11-01T19:11:03+01:00 Make better use of AvailInfo - - - - - f3e512d5 by Alex Biehl at 2017-11-02T12:16:22+01:00 Always return documentation for exported subordinates ... event if they have no documentation (e.g. noDocForDecl) By using the information in the AvailInfo we don't need additional export checks. - - - - - 7cf58898 by Alan Zimmerman at 2017-11-07T08:28:03+02:00 Match changes for Trees that Grow in GHC - - - - - e5105a41 by Alan Zimmerman at 2017-11-08T17:21:58+02:00 Match Trees That Grow - - - - - 55178266 by Alan Zimmerman at 2017-11-11T22:20:31+02:00 Match Trees that Grow in GHC for HsExpr - - - - - 2082ab02 by Ryan Scott at 2017-11-14T15:27:03+01:00 Actually render infix type operators as infix (#703) * Actually render infix type operators as infix * Account for things like `(f :*: g) p`, too - - - - - c52ab7d0 by Alan Zimmerman at 2017-11-14T23:14:26+02:00 Clean up use of PlaceHolder, to match TTG - - - - - 81cc9851 by Moritz Angermann at 2017-11-20T07:52:49+01:00 Declare use of `Paths_haddock` module in other-modules (#705) This was detected by `-Wmissing-home-modules` - - - - - f9d27598 by Moritz Angermann at 2017-11-20T12:47:34+01:00 Drop Paths_haddock from ghc.mk (#707) With haskell/haddock#705 and haskell/haddock#706, the custom addition should not be necessary any more. # Conflicts: # ghc.mk - - - - - f34818dc by Moritz Angermann at 2017-11-20T12:47:59+01:00 Add autogen-modules (#706) > Packages using 'cabal-version: >= 1.25' and the autogenerated module Paths_* must include it also on the 'autogen-modules' field besides 'exposed-modules' and 'other-modules'. This specifies that the module does not come with the package and is generated on setup. Modules built with a custom Setup.hs script also go here to ensure that commands like sdist don't fail. # Conflicts: # haddock.cabal - - - - - bb43a0aa by Ben Gamari at 2017-11-21T15:50:12-05:00 Revert "Clean up use of PlaceHolder, to match TTG" This reverts commit 134a7bb054ea730b13c8629a76232d73e3ace049. - - - - - af9ebb2b by Ben Gamari at 2017-11-21T15:50:14-05:00 Revert "Match Trees that Grow in GHC for HsExpr" This reverts commit 9f054dc365379c66668de6719840918190ae6e44. - - - - - 5d35c3af by Ben Gamari at 2017-11-21T15:50:15-05:00 Revert "Match Trees That Grow" This reverts commit 73a26af844ac50b8bec39de11d64452a6286b00c. - - - - - 99a8e43b by Ben Gamari at 2017-11-21T16:36:06-05:00 Revert "Match changes for Trees that Grow in GHC" This reverts commit 01eeeb048acd2dd05ff6471ae148a97cf0720547. - - - - - c4d650c2 by Ben Gamari at 2017-12-04T15:06:07-05:00 Bump GHC version - - - - - 027b2274 by Ben Gamari at 2017-12-04T17:06:31-05:00 Bump GHC bound to 8.4.* - - - - - 58eaf755 by Alex Biehl at 2017-12-06T15:44:24+01:00 Update changelog - - - - - d68f5584 by Simon Peyton Jones at 2017-12-07T14:39:56+00:00 Track changes to follow Trac haskell/haddock#14529 This tracks the refactoring of HsDecl.ConDecl. - - - - - dc519d6b by Alec Theriault at 2018-01-06T08:20:43-08:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - 8285118c by Alec Theriault at 2018-01-13T12:12:37+01:00 Constructor and pattern synonym argument docs (#709) * Support Haddocks on constructor arguments This is in conjunction with https://phabricator.haskell.org/D4094. Adds support for rendering Haddock's on (non-record) constructor arguments, both for regular and GADT constructors. * Support haddocks on pattern synonym arguments It appears that GHC already parsed these - we just weren't using them. In the process of doing this, I tried to deduplicate some code around handling patterns. * Update the markup guide Add some information about the new support for commenting constructor arguments, and mention pattern synonyms and GADT-style constructors. * Overhaul LaTeX support for data/pattern decls This includes at least * fixing several bugs that resulted in invalid LaTeX * fixing GADT data declaration headers * overhaul handling of record fields * overhaul handling of GADT constructors * overhaul handling of bundled patterns * add support for constructor argument docs * Support GADT record constructors This means changes what existing HTML docs look like. As for LaTeX, looks like GADT records were never even supported. Now they are. * Clean up code/comments Made code/comments consistent between the LaTeX and XHTML backend when possible. * Update changelog * Patch post-rebase regressions * Another post-rebase change We want return values to be documentable on record GADT constructors. - - - - - ca4fabb4 by Alec Theriault at 2018-01-15T17:12:18-08:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 4c472fea by Ryan Scott at 2018-01-19T10:44:02+01:00 Fix haskell/haddock#732 (#733) - - - - - bff14dbd by Alex Biehl at 2018-01-19T15:33:30+01:00 extractDecl: Extract associated types correctly (#736) - - - - - a2a94a73 by Alex Biehl at 2018-01-19T15:34:40+01:00 extractDecl: Extract associated types correctly (#736) - - - - - 26df93dc by Alex Biehl at 2018-01-20T10:18:22+01:00 haddock-api: bump ghc to ^>= 8.4 - - - - - f65aeb1d by Alex Biehl at 2018-01-20T19:18:20+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - 0e721b97 by Alex Biehl at 2018-01-20T19:20:19+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - cb6234f6 by Ben Gamari at 2018-01-26T13:40:55-05:00 Merge remote-tracking branch 'harpocrates/fix/missing-orphan-instances' into ghc-head - - - - - 0fc28554 by Alec Theriault at 2018-02-01T14:58:18+01:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - b9123772 by Alec Theriault at 2018-02-01T14:58:18+01:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 0c12e274 by Ryan Scott at 2018-02-01T14:58:18+01:00 Fix haskell/haddock#548 by rendering datatype kinds more carefully (#702) - - - - - 8876d20b by Alec Theriault at 2018-02-01T14:58:18+01:00 Use the GHC lexer for the Hyperlinker backend (#714) * Start changing to use GHC lexer * better cpp * Change SrcSpan to RealSrcSpan * Remove error * Try to stop too many open files * wip * wip * Revert "wip" This reverts commit b605510a195f26315e3d8ca90e6d95a6737553e1. Conflicts: haddock-api/haddock-api.cabal haddock-api/src/Haddock/Interface.hs * Remove pointless 'caching' * Use dlist rather than lists when finding vars * Use a map rather than list * Delete bogus comment * Rebase followup Things now run using the GHC lexer. There are still - stray debug statements - unnecessary changes w.r.t. master * Cleaned up differences w.r.t. current Haddock HEAD Things are looking good. quasiquotes in particular look beautiful: the TH ones (with Haskell source inside) colour/link their contents too! Haven't yet begun to check for possible performance problems. * Support CPP and top-level pragmas The support for these is hackier - but no more hacky than the existing support. * Tests pass, CPP is better recognized The tests were in some cases altered: I consider the new output to be more correct than the old one.... * Fix shrinking of source without tabs in test * Replace 'Position'/'Span' with GHC counterparts Replaces 'Position' -> 'GHC.RealSrcLoc' and 'Span' -> 'GHC.RealSrcSpan'. * Nits * Forgot entry in .cabal * Update changelog - - - - - 95c6a771 by Alec Theriault at 2018-02-01T14:58:18+01:00 Clickable anchors for headings (#716) See haskell/haddock#579. This just adds an <a> tag around the heading, pointing to the heading itself. - - - - - 21463d28 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump: Matches on function names weight more than matches in ... module names. - - - - - 8023af39 by Alex Biehl at 2018-02-01T14:58:18+01:00 Treat escaped \] better in definition lists (#717) This fixes haskell/haddock#546. - - - - - e4866dc1 by Alex Biehl at 2018-02-01T14:58:18+01:00 Remove scanner, takeWhile1_ already takes care of escaping - - - - - 9bcaa49d by Alex Biehl at 2018-02-01T14:58:18+01:00 Take until line feed - - - - - 01d2af93 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Add simple framework for running parser fixtures (#668) * Add simple framework for running parser fixtures * Compatible with tree-diff-0.0.0.1 * Use parseParas to parse fixtures This allows to test all syntactic constructs available in haddock markup. - - - - - 31128417 by Alec Theriault at 2018-02-01T14:58:18+01:00 Patch flaky parser test (#720) * Patch flaky parser test This test was a great idea, but it doesn't port over too well to using the GHC lexer. GHC rewrites its input a bit - nothing surprising, but we need to guard against those cases for the test. * Change instance head * Change use site - - - - - 9704f214 by Herbert Valerio Riedel at 2018-02-01T14:58:18+01:00 Include secondary LICENSE file in source dist - - - - - 51f25074 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Grid Tables (#718) * Add table examples * Add table types and adopt simple parser Simple parser is done by Giovanni Cappellotto (@potomak) in https://github.com/haskell/haddock/pull/577 It seems to support single fine full tables, so far from full RST-grid tables, but it's good start. Table type support row- and colspans, but obviously parser is lacking. Still TODO: - Latex backend. Should we use multirow package https://ctan.org/pkg/multirow?lang=en? - Hoogle backend: ? * Implement grid-tables * Refactor table parser * Add two ill-examples * Update CHANGES.md * Basic documentation for tables * Fix documentation example - - - - - 670d6200 by Alex Biehl at 2018-02-01T14:58:18+01:00 Add grid table example to cheatsheet (pdf and svg need to be regenerated thought) - - - - - 4262dec9 by Alec Theriault at 2018-02-01T14:58:18+01:00 Fix infinite loop when specializing instance heads (#723) * Fix infinite loop when specializing instance heads The bug can only be triggered from TH, hence why it went un-noticed for so long. * Add test for haskell/haddock#679 and haskell/haddock#710 - - - - - 67ecd803 by Alec Theriault at 2018-02-01T14:58:18+01:00 Filter RTS arguments from 'ghc-options' arguments (#725) This fixes haskell/haddock#666. - - - - - 7db26992 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump Scrollable overlay - - - - - da9ff634 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Hyperlinker: Adjust parser to new PFailed constructor - - - - - 7b7cf8cb by Alexander Biehl at 2018-02-01T14:58:18+01:00 Specialize: Add missing IdP annotations - - - - - 78cd7231 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Convert: Correct pass type - - - - - a2d0f590 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Warning free compilation - - - - - cd861cf3 by Alexander Biehl at 2018-02-01T14:58:18+01:00 hadock-2.19.0 / haddock-api-2.19.0 / haddock-library-1.5.0 - - - - - c6651b72 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Adjust changelogs - - - - - 1e93da0b by Alexander Biehl at 2018-02-01T14:58:18+01:00 haddock-library: Info about breaking changes - - - - - f9b11db8 by Alec Theriault at 2018-02-02T12:36:02+01:00 Properly color pragma contents in hyperlinker The hyperlinker backend now classifies the content of pragmas as 'TkPragma'. That means that in something like '{-# INLINE foo #-}', 'foo' still gets classified as a pragma token. - - - - - c40b0043 by Alec Theriault at 2018-02-02T12:36:02+01:00 Support the new 'ITcolumn_prag' token - - - - - 4a2a4d39 by Alex Biehl at 2018-02-03T12:11:55+01:00 QuickJump: Mitigate encoding problems on Windows - - - - - bb34503a by Alex Biehl at 2018-02-04T18:39:31+01:00 Use withBinaryFile - - - - - 637605bf by Herbert Valerio Riedel at 2018-02-05T09:48:32+01:00 Try GHC 8.4.1 for Travis CI job - - - - - 7abb67e4 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 try harder to build w/ GHC 8.4.1 - - - - - 8255cc98 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 Add `SPDX-License-Identifier` as alised for "license" module header tokens C.f. SPDX 2.1 - Appendix V https://spdx.org/spdx-specification-21-web-version#h.twlc0ztnng3b The tag should appear on its own line in the source file, generally as part of a comment. SPDX-License-Identifier: <SPDX License Expression> Cherry-picked from haskell/haddock#743 - - - - - 267cd23d by Herbert Valerio Riedel at 2018-02-05T10:24:34+01:00 Make test-suite SMP compatible - - - - - 95d4bf40 by Alec Theriault at 2018-02-05T22:01:04+01:00 Hyperlink pattern synonyms and 'module' imports (#744) Links to pattern synonyms are now generated, as well as links from modules in import lists. Fixes haskell/haddock#731. - - - - - 67838dcd by Alec Theriault at 2018-02-06T08:23:36+01:00 Don't warn about missing '~' (#746) This manually filters out '~' from the list of things to warn about. It truly makes no sense to warn on this since '~' has nothing it could link to - it is magical. This fixes haskell/haddock#532. - - - - - ab6c3f9f by Alec Theriault at 2018-02-06T08:24:47+01:00 Don't barf on 'HsSpliceTy' (#745) This handles 'HsSpliceTy's by replacing them with what they expand to. IIUC everything that is happening, 'renameHsSpliceTy' should not be able to fail for the inputs we feed it from GHC. This fixes haskell/haddock#574. - - - - - 92bf95ad by Alex Biehl at 2018-02-06T08:28:23+01:00 Rename: renameHsSpliceTy ttg - - - - - 3130b1e1 by Alex Biehl at 2018-02-06T09:02:14+01:00 Expand SigDs - - - - - c72adae5 by Alex Biehl at 2018-02-06T09:20:51+01:00 fullModuleContents: support named docs - - - - - de2e4dbf by Alex Biehl at 2018-02-06T13:56:17+01:00 Hyperlinker: Also link pattern synonym arguments - - - - - b7c98237 by Alex Biehl at 2018-02-09T18:44:23+01:00 Expand SigD in a better place In https://github.com/haskell/haddock/issues/287 we found that haddock-2.19.0 would miss documentation on class methods with multiples names. This patch uses expandSigDecls in a more sensible place. - - - - - 8f598b27 by Alec Theriault at 2018-02-11T12:29:56+01:00 Add module tooltips to linked identifiers (#753) No more clicking to figure out whether your bytestring is strict or lazy! - - - - - d812e65d by Alec Theriault at 2018-02-11T12:31:44+01:00 Add 'show' option to complement 'hide' (#752) * Add 'show' option to complement 'hide' The behaviour is for flags passed in the command line to override flags in file headers. In the command line, later flags override earlier ones. Fixes haskell/haddock#751 and haskell/haddock#266. * Add a '--show-all' option - - - - - 6676cecb by Alex Biehl at 2018-02-18T11:07:15-05:00 QuickJump: Mitigate encoding problems on Windows (cherry picked from commit 86292c54bfee2343aee84559ec01f1fc68f52231) - - - - - e753dd88 by Alex Biehl at 2018-02-18T17:59:54+01:00 Use withBinaryFile - - - - - 724dc881 by Tamar Christina at 2018-02-19T05:34:49+01:00 Haddock: support splitted include paths. (#689) - - - - - 9b6d6f50 by Alex Biehl at 2018-02-19T05:57:02+01:00 Teach the HTML backend how to render methods with multiple names - - - - - a74aa754 by Alexander Biehl at 2018-02-19T10:04:34+01:00 Hoogle/Latex: Remove use of partial function - - - - - 66d8bb0e by Alec Theriault at 2018-02-25T16:04:01+01:00 Fix file handle leak (#763) (#764) Brought back some mistakenly deleted code for handling encoding and eager reading of files from e0ada1743cb722d2f82498a95b201f3ffb303137. - - - - - bb92d03d by Alex Biehl at 2018-03-02T14:21:23+01:00 Enable running test suite with stock haddock and ghc using ``` $ cabal new-run -- html-test --haddock-path=$(which haddock) --ghc-path=$(which ghc) ``` - - - - - dddb3cb2 by Alex Biehl at 2018-03-02T15:43:21+01:00 Make testsuite work with haddock-1.19.0 release (#766) - - - - - f38636ed by Alec Theriault at 2018-03-02T15:48:36+01:00 Support unicode operators, proper modules Unicode operators are a pretty big thing in Haskell, so supporting linking them seems like it outweighs the cost of the extra machinery to force Attoparsec to look for unicode. Fixes haskell/haddock#458. - - - - - 09d89f7c by Alec Theriault at 2018-03-02T15:48:43+01:00 Remove bang pattern - - - - - d150a687 by Alex Biehl at 2018-03-02T15:48:48+01:00 fix test - - - - - d6fd71a5 by Alex Biehl at 2018-03-02T16:22:38+01:00 haddock-test: Be more explicit which packages to pass We now pass `-hide-all-packages` to haddock when invoking the testsuite. This ensures we don't accidentally pick up any dependencies up through ghc.env files. - - - - - 0932c78c by Alex Biehl at 2018-03-02T17:50:38+01:00 Revert "fix test" This reverts commit 1ac2f9569242f6cb074ba6e577285a4c33ae1197. - - - - - 52516029 by Alex Biehl at 2018-03-02T18:16:50+01:00 Fix Bug548 for real - - - - - 89df9eb5 by Alex Biehl at 2018-03-05T18:28:19+01:00 Hyperlinker: Links for TyOps, class methods and associated types - - - - - d019a4cb by Ryan Scott at 2018-03-06T13:43:56-05:00 Updates for haskell/haddock#13324 - - - - - 6d5a42ce by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump haddock-2.19.0.1, haddock-api-2.19.0.1, haddock-library-1.5.0.1 - - - - - c0e6f380 by Alex Biehl at 2018-03-10T18:25:57+01:00 Update changelogs for haddock-2.19.0.1 and haddock-library-1.5.0.1 - - - - - 500da489 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Update to QC 2.11 - - - - - ce8362e9 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Restore backward-compat with base-4.5 through base-4.8 - - - - - baae4435 by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump lower bound for haddock-library - - - - - 10b7a73e by Alex Biehl at 2018-03-10T18:25:57+01:00 Haddock: Straighten out base bound - - - - - a6096f7b by Alex Biehl at 2018-03-13T08:45:06+01:00 extractDecl: Extract constructor patterns from data family instances (#776) * extractDecl: Allow extraction of data family instance constructors * extractDecl: extract data family instance constructors - - - - - ba4a0744 by Simon Jakobi at 2018-03-14T08:26:42+01:00 Readme: Update GHC version (#778) - - - - - 8de157d4 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for definition lists - - - - - 425b46f9 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for links - - - - - d53945d8 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for inline links - - - - - f1dc7c99 by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Slightly unmangle output - - - - - 0879d31c by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Prevent stdout buffering - - - - - 1f9e5f1b by Simon Jakobi at 2018-03-14T20:39:29+01:00 haddock-library.cabal: Clean up GHC options - - - - - 066b891a by Simon Jakobi at 2018-03-14T20:39:29+01:00 Make a proper definition for the <link> parser - - - - - 573d6ba7 by Alec Theriault at 2018-03-21T09:16:57+01:00 Show where instances are defined (#748) * Indicate source module of instances Above instance, we now also display a link to the module where the instance was defined. This is sometimes helpful in figuring out what to import. * Source module for type/data families too * Remove parens * Accept tests - - - - - 99b5d28b by Alex Biehl at 2018-03-21T09:20:36+01:00 Prepare changelog for next release - - - - - 482d3a93 by Alex Biehl at 2018-03-23T15:57:36+01:00 Useful cost centres, timers and allocation counters (#785) * Add some useful cost-centres for profiling * Add withTiming for each haddock phase Invoking haddock with `--optghc=-ddump-timings` now shows the amount of time spent and the number of allocated bytes for each phase. - - - - - 773b41bb by Alec Theriault at 2018-03-27T08:35:59+02:00 @since includes package name (#749) * Metadoc stores a package name This means that '@since' annotations can be package aware. * Get the package name the right way This should extract the package name for `@since` annotations the right way. I had to move `modulePackageInfo` around to do this and, in the process, I took the liberty to update it. Since it appears that finding the package name is something that can fail, I added a warning for this case. * Silence warnings * Hide package for local 'since' annotations As discussed, this is still the usual case (and we should avoid being noisy for it). Although this commit is large, it is basically only about threading a 'Maybe Package' from 'Haddock.render' all the way to 'Haddock.Backends.Xhtml.DocMarkup.renderMeta'. * Bump binary interface version * Add a '--since-qual' option This controls when to qualify since annotations with the package they come from. The default is always, but I've left an 'external' variant where only those annotations coming from outside of the current package are qualified. * Make ParserSpec work * Make Fixtures work * Use package name even if package version is not available The @since stuff needs only the package name passed in, so it makes sense to not be forced to pass in a version too. - - - - - e42c57bc by Alex Biehl at 2018-03-27T08:42:50+02:00 haddock-2.19.1, haddock-api-2.19.1, haddock-library-1.6.0 - - - - - 8373a529 by Alex Biehl at 2018-03-28T10:17:11+02:00 Bump haddock and haddock-api to 2.20.0 - - - - - 5038eddd by Jack Henahan at 2018-04-03T13:28:12+02:00 Clear search string on hide for haskell/haddock#781 (#789) - - - - - 920ca1eb by Alex Biehl at 2018-04-03T16:35:50+02:00 Travis: Build with ghc-8.4.2 (#793) - - - - - a232f0eb by Alan Zimmerman at 2018-04-07T14:14:32+02:00 Match changes in GHC for D4199 Removing HasSourceText and SourceTextX classes. - - - - - ab85060b by Alan Zimmerman at 2018-04-09T21:20:24+02:00 Match GHC changes for TTG - - - - - 739302b6 by Alan Zimmerman at 2018-04-13T13:31:44+02:00 Match GHC for TTG implemented on HsBinds, D4581 - - - - - 2f56d3cb by Ryan Scott at 2018-04-19T11:42:58-04:00 Bump upper bound on base to < 4.13 See https://ghc.haskell.org/trac/ghc/ticket/15018. - - - - - a49df92a by Alex Biehl at 2018-04-20T07:31:44+02:00 Don't treat fixity signatures like declarations - - - - - d02c103b by Ryan Scott at 2018-04-24T11:20:11-04:00 Add regression test for haskell/haddock#413 Fixes haskell/haddock#413. - - - - - c7577f52 by Ryan Scott at 2018-04-24T13:51:06-07:00 Improve the Hoogle backend's treatment of type families (#808) Fixes parts 1 and 2 of haskell/haddock#806. - - - - - d88f85b1 by Alec Theriault at 2018-04-25T11:24:07-07:00 Replace 'attoparsec' with 'parsec' (#799) * Remove attoparsec with parsec and start fixing failed parses * Make tests pass * Fix encoding issues The Haddock parser no longer needs to worry about bytestrings. All the internal parsing work in haddock-library happens over 'Text'. * Remove attoparsec vendor * Fix stuff broken in 'attoparsec' -> 'parsec' * hyperlinks * codeblocks * examples Pretty much all issues are due to attoparsec's backtracking failure behaviour vs. parsec's non-backtracking failure behaviour. * Fix small TODOs * Missing quote + Haddocks * Better handle spaces before/after paragraphs * Address review comments - - - - - fc25e2fe by Alan Zimmerman at 2018-04-27T15:36:53+02:00 Match changes in GHC for TTG - - - - - 06175f91 by Herbert Valerio Riedel at 2018-05-01T18:11:09+02:00 Merge branch 'ghc-head' with 'ghc-8.4' - - - - - 879caaa8 by Alec Theriault at 2018-05-07T18:53:15-07:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 3e0120cb by Simon Jakobi at 2018-05-07T19:00:18-07:00 Add docs for some DocH constructors (#814) - - - - - 0a32c6db by Alec Theriault at 2018-05-08T02:15:45-07:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 8816e783 by Simon Jakobi at 2018-05-08T10:48:11-07:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - ad60366f by Ryan Scott at 2018-05-10T11:19:47-04:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - 03b7cc3b by Ryan Scott at 2018-05-10T11:24:38-04:00 Wibbles - - - - - b03dd563 by Chaitanya Koparkar at 2018-05-10T11:44:58-04:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 9f298a40 by Ben Gamari at 2018-05-13T17:36:04-04:00 Account for refactoring of LitString - - - - - ea3dabe7 by Ryan Scott at 2018-05-16T09:21:43-04:00 Merge pull request haskell/haddock#826 from haskell/T825 Remove Hoogle backend hack that butchers infix datatype names - - - - - 0d234f7c by Alec Theriault at 2018-05-23T11:29:05+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 15fc9712 by Simon Jakobi at 2018-05-31T04:17:47+02:00 Adjust to new HsDocString internals - - - - - 6f1e19a8 by Ben Gamari at 2018-06-02T16:18:58-04:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 0d0355d9 by Ryan Scott at 2018-06-04T21:26:59-04:00 DerivingVia changes - - - - - 0d93475a by Simon Jakobi at 2018-06-05T19:47:05+02:00 Bump a few dependency bounds (#845) - - - - - 5cbef804 by Alec Theriault at 2018-06-05T19:47:16+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 9094c56f by Alec Theriault at 2018-06-05T22:53:25+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 70188719 by Simon Jakobi at 2018-06-08T22:20:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - 495cd1fc by Chaitanya Koparkar at 2018-06-13T23:01:34+02:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 81088732 by Ben Gamari at 2018-06-13T23:01:34+02:00 Account for refactoring of LitString - - - - - 7baf6587 by Simon Jakobi at 2018-06-13T23:05:08+02:00 Adjust to new HsDocString internals - - - - - bb61464d by Ben Gamari at 2018-06-13T23:05:22+02:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 5d8cb87f by Ryan Scott at 2018-06-13T23:39:30+02:00 DerivingVia changes - - - - - 73d373a3 by Alec Theriault at 2018-06-13T23:39:30+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 4865e254 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Remove `ITtildehsh` token - - - - - b867db54 by Alec Theriault at 2018-06-13T23:39:30+02:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 9598e392 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Add docs for some DocH constructors (#814) - - - - - 8a59035b by Alec Theriault at 2018-06-13T23:39:30+02:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 29350fc8 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - 2590bbd9 by Ryan Scott at 2018-06-13T23:39:30+02:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - a9939fdc by Ryan Scott at 2018-06-13T23:39:30+02:00 Wibbles - - - - - a22f7df4 by Alec Theriault at 2018-06-13T23:39:30+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 8741015d by Simon Jakobi at 2018-06-13T23:39:30+02:00 Bump a few dependency bounds (#845) - - - - - 4791e1cc by Alec Theriault at 2018-06-13T23:39:30+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 311d3216 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - d0577817 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Complete FixitySig and FamilyDecl pattern matches - - - - - 055b3aa7 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Fix redundant import warnings - - - - - f9ce19b1 by Simon Jakobi at 2018-06-13T23:49:52+02:00 html-test: Accept output - - - - - 04604ea7 by Simon Jakobi at 2018-06-13T23:54:37+02:00 Bump bounds on Cabal - - - - - 0713b692 by Simon Jakobi at 2018-06-14T00:00:12+02:00 Merge branch 'ghc-head' into ghc-head-update-3 - - - - - c6a56bfd by Simon Jakobi at 2018-06-14T02:33:27+02:00 Bump ghc bound for haddock-api spec test-suite - - - - - 119d04b2 by Simon Jakobi at 2018-06-14T12:37:48+02:00 Travis: `--allow-newer` for all packages - - - - - 0e876e2c by Alex Biehl at 2018-06-14T15:28:52+02:00 Merge pull request haskell/haddock#857 from sjakobi/ghc-head-update-3 Update ghc-head - - - - - 5be46454 by Alec Theriault at 2018-06-14T21:42:45+02:00 Improved handling of interfaces in 'haddock-test' (#851) This should now work with an inplace GHC where (for instance) HTML directories may not be properly recorded in the package DB. - - - - - 96ab1387 by Vladislav Zavialov at 2018-06-14T17:06:21-04:00 Handle -XStarIsType - - - - - e518f8c4 by Ben Gamari at 2018-06-14T17:48:00-04:00 Revert unintentional reversion of fix of haskell/haddock#548 - - - - - 01b9f96d by Alan Zimmerman at 2018-06-19T11:52:22+02:00 Match changes in GHC for haskell/haddock#14259 - - - - - 7f8c8298 by Ben Gamari at 2018-06-19T18:14:27-04:00 Bump GHC version to 8.6 - - - - - 11c6b5d2 by Ryan Scott at 2018-06-19T23:17:31-04:00 Remove HsEqTy and XEqTy - - - - - b33347c2 by Herbert Valerio Riedel at 2018-06-20T23:14:52+02:00 Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. - - - - - f0d2460e by Herbert Valerio Riedel at 2018-06-20T23:28:46+02:00 Update Travis CI job - - - - - ef239223 by Herbert Valerio Riedel at 2018-06-20T23:32:41+02:00 Drop GHC HEAD from CI and update GHC to 8.4.3 It's a waste of resource to even try to build this branch w/ ghc-head; so let's not do that... - - - - - 41c4a9fa by Ben Gamari at 2018-06-20T18:26:20-04:00 Bump GHC version to 8.7 - - - - - 8be593dc by Herbert Valerio Riedel at 2018-06-21T22:32:15+02:00 Update CI job to use GHC 8.7.* - - - - - b91d334a by Simon Jakobi at 2018-06-30T13:41:38+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section - - - - - f707d848 by Alec Theriault at 2018-07-05T10:43:35-04:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. - - - - - a6d2b8dc by Alec Theriault at 2018-07-06T10:06:32-04:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case - - - - - 13819f71 by Alan Zimmerman at 2018-07-15T19:33:51+02:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 - - - - - c346aa78 by Simon Jakobi at 2018-07-19T12:29:32+02:00 haddock-library: Bump bounds for containers - - - - - 722e733c by Simon Jakobi at 2018-07-19T13:36:45+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] - - - - - f0bd83fd by Alec Theriault at 2018-07-19T14:39:57+02:00 Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. - - - - - 394053a8 by Simon Jakobi at 2018-07-19T14:58:07+02:00 haddock-library: Bump bounds for containers - - - - - 1bda11a2 by Alec Theriault at 2018-07-20T09:04:03+02:00 Add HEAD.hackage overlay (#887) * Add HEAD.hackage overlay * Add HCPKG variable - - - - - c7b4ab45 by Alec Theriault at 2018-07-20T12:01:16+02:00 Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test - - - - - c05d32ad by Alec Theriault at 2018-07-20T12:01:49+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output - - - - - 24b39ee4 by Alec Theriault at 2018-07-20T12:02:16+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. - - - - - cb9d2099 by Simon Jakobi at 2018-07-20T13:39:29+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) - - - - - 133f24f5 by Alec Theriault at 2018-07-20T13:39:29+02:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) - - - - - 11024149 by Alec Theriault at 2018-07-20T13:39:29+02:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) - - - - - de0c139e by Simon Jakobi at 2018-07-20T13:39:29+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) - - - - - 6435e952 by Alec Theriault at 2018-07-20T13:39:29+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) - - - - - 1461af39 by Alec Theriault at 2018-07-20T13:39:29+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) - - - - - 69d3bde1 by Alec Theriault at 2018-07-20T13:49:47+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) - - - - - 6a5c73c7 by Alec Theriault at 2018-07-20T13:50:00+02:00 Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests - - - - - 92ca94c6 by Alec Theriault at 2018-07-20T13:55:36+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) - - - - - 981bc7fa by Simon Jakobi at 2018-07-20T15:06:06+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers - - - - - 27e7c0c5 by Simon Jakobi at 2018-07-20T15:09:05+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 49e1a415 by Simon Jakobi at 2018-07-20T16:02:02+02:00 Update the ghc-8.6 branch (#889) * Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. * README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) * Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) * Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) * tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) * Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. * Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test * Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) * Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) * Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) * Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests * Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 5ca14bed by Simon Jakobi at 2018-07-20T16:05:47+02:00 Revert "Revert "Bump GHC version to 8.6"" That commit didn't belong onto the ghc-8.6 branch. This reverts commit acbaef3b9daf1d2dea10017964bf886e77a8e967. - - - - - 2dd600dd by Simon Jakobi at 2018-07-20T16:18:21+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. - - - - - fceb2422 by Simon Jakobi at 2018-07-20T16:18:21+02:00 outOfScope: Recommend qualifying the identifier - - - - - acea5d23 by Simon Jakobi at 2018-07-20T16:19:35+02:00 outOfScope: Recommend qualifying the identifier (cherry picked from commit 73707ed58d879cc04cb644c5dab88c39ca1465b7) - - - - - 1a83ca55 by Simon Jakobi at 2018-07-20T16:19:35+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. (cherry picked from commit d504a2864a4e1982e142cf88c023e7caeea3b76f) - - - - - 48374451 by Masahiro Sakai at 2018-07-20T17:06:42+02:00 Add # as a special character (#884) '#' has special meaning used for anchors and can be escaped using backslash. Therefore it would be nice to be listed as special characters. - - - - - 5e1a5275 by Alec Theriault at 2018-07-20T23:37:24+02:00 Let `haddock-test` bypass interface version check (#890) This means `haddock-test` might * crash during deserialization * deserialize incorrectly Still - it means things _might_ work where they were previously sure not to. - - - - - 27286754 by Yuji Yamamoto at 2018-07-23T08:16:01+02:00 Avoid "invalid argument (invalid character)" on non-unicode Windows (#892) Steps to reproduce and the error message ==== ``` > stack haddock basement ... snip ... Warning: 'A' is out of scope. Warning: 'haddock: internal error: <stdout>: commitBuffer: invalid argument (invalid character) ``` Environment ==== OS: Windows 10 ver. 1709 haddock: [HEAD of ghc-8.4 when I reproduce the error](https://github.com/haskell/haddock/commit/532b209d127e4cecdbf7e9e3dcf4f653a5605b5a). (I had to use this version to avoid another probrem already fixed in HEAD) GHC: 8.4.3 stack: Version 1.7.1, Git revision 681c800873816c022739ca7ed14755e85a579565 (5807 commits) x86_64 hpack-0.28.2 Related pull request ==== https://github.com/haskell/haddock/pull/566 - - - - - 6729d361 by Alec Theriault at 2018-07-23T13:52:56-07:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). - - - - - 8cf4e6b5 by Ryan Scott at 2018-07-27T11:28:03-04:00 eqTyCon_RDR now lives in TysWiredIn After GHC commit http://git.haskell.org/ghc.git/commit/f265008fb6f70830e7e92ce563f6d83833cef071 - - - - - 1ad251a6 by Alan Zimmerman at 2018-07-30T13:28:09-04:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 (cherry picked from commit e3926b50ab8a7269fd6904b06e881745f08bc5d6) - - - - - 8aea2492 by Richard Eisenberg at 2018-08-02T10:54:17-04:00 Update against new HsImplicitBndrs - - - - - e42cada9 by Alec Theriault at 2018-08-04T17:51:30+02:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 0e852512 by Alex Biehl at 2018-08-06T13:04:02+02:00 Make --package-version optional for --hoogle generation (#899) * Make --package-version optional for --hoogle generation * Import mkVersion * It's makeVersion not mkVersion - - - - - d2abd684 by Noel Bourke at 2018-08-21T09:34:18+02:00 Remove unnecessary backslashes from docs (#908) On https://haskell-haddock.readthedocs.io/en/latest/markup.html#special-characters the backslash and backtick special characters showed up with an extra backslash before them – I think the escaping is not (or no longer) needed for those characters in rst. - - - - - 7a578a9e by Matthew Pickering at 2018-08-21T09:34:50+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - aa3d4db3 by Matthew Pickering at 2018-08-21T09:37:34+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - ede91744 by Alec Theriault at 2018-08-21T09:42:52+02:00 Better test output when Haddock crashes on a test (#902) In particular: we report the tests that crashed seperately from the tests that produced incorrect output. In order for tests to pass (and exit 0), they must not crash and must produce the right output. - - - - - 4a872b84 by Guillaume Bouchard at 2018-08-21T09:45:57+02:00 Fix a typo (#878) - - - - - 4dbf7595 by Ben Sklaroff at 2018-08-21T12:04:09-04:00 Add ITcomment_line_prag token to Hyperlinker Parser This token is necessary for parsing #line pragmas inside nested comments. Reviewers: bgamari Reviewed By: bgamari Differential Revision: https://phabricator.haskell.org/D4935 - - - - - 9170b2a9 by Ben Gamari at 2018-08-21T17:55:15-04:00 Merge pull request haskell/haddock#893 from harpocrates/get-name-to-instances Accumulate explicitly which modules to load for 'attachInstances' - - - - - d57b57cc by Ben Gamari at 2018-08-21T17:59:13-04:00 Merge branch 'ghc-head' of github.com:haskell/haddock into ghc-head - - - - - 14601ca2 by Alec Theriault at 2018-08-21T19:09:37-04:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). (cherry picked from commit 5c7c596c51d69b92164e9ba920157b36ce2b2ec1) - - - - - 438c645e by Matthew Pickering at 2018-08-21T19:12:39-04:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 (cherry picked from commit e6aa8fb47b9477cc5ef5e46097524fe83e080f6d) - - - - - a80c5161 by Alec Theriault at 2018-08-21T22:06:40-07:00 Better rendering of unboxed sums/tuples * adds space after/before the '#' marks * properly reify 'HsSumTy' in 'synifyType' - - - - - 88456cc1 by Alec Theriault at 2018-08-21T22:06:40-07:00 Handle promoted tuples in 'synifyType' When we have a fully applied promoted tuple, we can expand it out properly. - - - - - fd1c1094 by Alec Theriault at 2018-08-21T22:19:34-07:00 Accept test cases - - - - - 6e80d9e0 by Alec Theriault at 2018-08-21T22:24:03-07:00 Merge pull request haskell/haddock#914 from harpocrates/feature/unboxed-stuff Better rendering of unboxed sums, unboxed tuples, promoted tuples. - - - - - 181a23f1 by Ben Gamari at 2018-08-23T15:53:48-04:00 Merge remote-tracking branch 'origin/ghc-8.6' into ghc-8.6 - - - - - 3a18c1d8 by Alec Theriault at 2018-08-27T14:15:25-07:00 Properly synify promoted list types We reconstruct promoted list literals whenever possible. That means that 'synifyType' produces '[Int, Bool, ()] instead of (Int ': (() ': (Bool ': ([] :: [Type])))) - - - - - b4794946 by Alec Theriault at 2018-09-03T07:19:55-07:00 Only look at visible types when synifying a 'HsListTy' The other types are still looked at when considering whether to make a kind signature or not. - - - - - a231fce2 by Alec Theriault at 2018-09-03T07:38:10-07:00 Merge pull request haskell/haddock#922 from harpocrates/promoted-lists Properly synify promoted list types - - - - - 0fdf044e by Ningning Xie at 2018-09-15T10:25:58-04:00 Update according to GHC Core changes - - - - - 7379b115 by Ningning Xie at 2018-09-15T15:40:18-04:00 update dataFullSig to work with Co Quantification This should have been in the previous patch, but wasn't. - - - - - cf84a046 by Alec Theriault at 2018-09-17T20:12:18-07:00 Fix/add to various docs * Add documentation for a bunch of previously undocumented options (fixes haskell/haddock#870) * Extend the documentation of `--hoogle` considerably (see haskell/haddock#807) * Describe how to add docs to `deriving` clauses (fixes haskell/haddock#912) * Fix inaccurate docs about hyperlinking infix identifiers (fixes haskell/haddock#780) - - - - - ae017935 by Alec Theriault at 2018-09-22T08:32:16-07:00 Update Travis - - - - - d95ae753 by Alec Theriault at 2018-09-22T09:34:10-07:00 Accept failing tests Also silence orphan warnings. - - - - - f3e67024 by Alec Theriault at 2018-09-22T09:41:23-07:00 Bump haddock-api-2.21.0, haddock-library-1.7.0 * Update CHANGELOGS * Update new versions in Cabal files * Purge references to ghc-8.4/master branches in README - - - - - 3f136d4a by Alec Theriault at 2018-09-22T10:53:31-07:00 Turn haddock-library into a minor release Fix some version bounds in haddock-library too. - - - - - b9def006 by Alec Theriault at 2018-09-22T13:07:35-07:00 keep cabal.project file - - - - - 4909aca7 by Alec Theriault at 2018-10-16T09:36:30-07:00 Build on 7.4 and 7.8 - - - - - 99d20a28 by Herbert Valerio Riedel at 2018-10-16T18:45:52+02:00 Minor tweak to package description - - - - - a8059618 by Herbert Valerio Riedel at 2018-10-16T18:47:24+02:00 Merge pull request haskell/haddock#945 haddock-api 2.21.0 and haddock-library 1.6.1 release - - - - - 2d9bdfc1 by Alec Theriault at 2018-10-16T10:54:21-07:00 Bump haddock-library to 1.7.0 The 1.6.1 release should've been a major bump, since types in the `Documentation.Haddock.Parser.Monad` module changed. This version makes that module internal (as it morally should be). - - - - - ed340cef by Alec Theriault at 2018-10-16T14:59:13-07:00 Merge branch 'ghc-8.4' into ghc-8.6 - - - - - 2821a8df by Alec Theriault at 2018-10-16T15:14:48-07:00 Merge branch 'ghc-8.6' into ghc-head - - - - - a722dc84 by Alec Theriault at 2018-10-16T16:28:55-07:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 63377496 by Alec Theriault at 2018-10-16T16:39:07-07:00 Update changelog - - - - - 099a0110 by Alec Theriault at 2018-10-16T16:49:28-07:00 Merge pull request haskell/haddock#942 from harpocrates/update-docs Fix & add to documentation - - - - - 0927416f by Alec Theriault at 2018-10-16T16:50:14-07:00 Set UTF-8 encoding before writing files (#934) This should fix haskell/haddock#929, as well as guard against future problems of this sort in other places. Basically replaces 'writeFile' (which selects the users default locale) with 'writeUtf8File' (which always uses utf8). - - - - - 83b7b017 by Alec Theriault at 2018-10-16T17:42:05-07:00 Output pattern synonyms in Hoogle backend (#947) * Output pattern synonyms in Hoogle backend We were previously weren't outputting _any_ pattern synonyms, bundled or not. Now, we output both. Fixes haskell/haddock#946. * Update changelog - - - - - 81e5033d by Alec Theriault at 2018-10-16T18:04:40-07:00 Release `haddock{,-api}-2.22.0` This version will accompany ghc-8.6.2 - - - - - 9661744e by Alex Biehl at 2018-10-18T08:14:32-07:00 Add NewOcean theme And make it the default theme. - - - - - 7ae6d722 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Improve appearance and readability These changes include: - use latest Haskell's logo colors - decrease #content width to improve readability - use nicer font - improve sizes and distances - - - - - 37f8703d by NunoAlexandre at 2018-10-18T08:14:32-07:00 Include custom font in the html head - - - - - 1d5e1d79 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Update html test reference files - - - - - 53b7651f by NunoAlexandre at 2018-10-18T08:14:32-07:00 Make it responsive - It makes small screens taking more space than larger ones - fixes a few issues present in small screens currently - make it look good across different screen sizes. - - - - - 6aa1aeb1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make the style consistent with hackage Several things are addressed here: - better responsive behaviour on the header - better space usage - consistent colors overall - other nit PR comments - - - - - 3a250c5c by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Place the package name before the menu links This supports the expected responsive menu design, where the package name appears above the menu links. - - - - - cae699b3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update html-test reference files The package name element in the package-header is now a div instead of a paragraph, and it is now above the menu ul.links instead of below. - - - - - 2ec7fd2d by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve synopsis style and code - Use CSS3 instead of loading pictures to show "+" and "-" symbols - Drop redundant code - - - - - 0c874c01 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Decrease space between code blocks There was too much space between code blocks as pointed out by reviewers. - - - - - 85568ce2 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Add an initial-scale property to all haddock pages This solves an issue reported about the content looking incredibly small on mobile devices. - - - - - c1538926 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Address several PR comments - Darken text color like done for hackage - Move synopsis to left side - Make table of contents stick to the left on wide screens - Wrap links to avoid page overflow - Improve expand/collapse buttons - Fix issue with content size on mobile devices - Fix issue with font-size on landscape mode - Increase width of the content - Change colors of table of contents and synopsis - Etc - - - - - e6639e5f by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make synopsis scrollable on wide screens When the synopsis is longer than the screen, you can’t see its end and you can't scroll down either, making the content unreachable. - - - - - 1f0591ff by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve information density - Reduce font size - Improve space between and within code blocks - Improve alignments - Improve spacing within sub-blocks - - - - - bf083097 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Minor adjustments Bring in some adjustments made to hackage: - link colors - page header show everything when package title is too long - - - - - 10375fc7 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix responsive triggers overlap issue The min and max width triggers have the same values, which caused the style resolution to take an intersection of both style declarations when the screen resolution had the size of the limts (say 1280px), causing an odd behaviour and look. - - - - - 95ff2f95 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix issue with menu alignment on firefox Reported and described here: https://github.com/haskell/haddock/pull/721#issuecomment-374668869 - - - - - dc86587e by Alex Biehl at 2018-10-18T08:14:32-07:00 Changelog entry for NewOcean - - - - - 27195e47 by Herbert Valerio Riedel at 2018-10-18T08:14:32-07:00 html-test --accept - - - - - 83f4f9c0 by Alex Biehl at 2018-10-18T08:14:32-07:00 Avoid name shadowing - - - - - 231487f1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font to PT Sans Also migrate some general text related changes from hackage. - - - - - 313db81a by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Use 'flex' to fix header alignment - - - - - 5087367b by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Misc of tweaks - Update link colors to hackage scheme - Tune spacing between content elements - Update footer style - Fix and improve code blocks identation - - - - - b08020df by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font in Xhtml.hs to PT Sans - - - - - 78ce06e3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve code blocks styling - Fix and improve spacing - Improve colors and borders - - - - - 81262d20 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make package-header caption backward-compatible The current html generator of this branch wraps the package-header caption as a div, which does not work (without style adjustments) with the old themes. Changing it from div to span does the trick, without needing to adjust the old stylesheets. - - - - - dc4475cb by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update test-suite reference html pages - - - - - 393d35d8 by Alec Theriault at 2018-10-18T08:25:36-07:00 Accept tests - - - - - a94484ba by Alec Theriault at 2018-10-21T10:29:29-07:00 Fix CHANGELOG - - - - - 8797eca3 by Alec Theriault at 2018-10-21T10:36:19-07:00 Update 'data-files' to include NewOcean stuff - - - - - 1ae51e4a by Simon Jakobi at 2018-10-23T11:29:14+02:00 Fix typo in a warning - - - - - 009ad8e8 by Alec Theriault at 2018-10-24T12:47:47-07:00 Update JS dependencies This was done via `npm audit fix`. I think this fixes haskell/haddock#903 along with some more serious vulnerabilities that nobody seems to have noticed. - - - - - 051994db by Alec Theriault at 2018-10-24T17:31:09-07:00 Resurrect the style-switcher This fixes haskell/haddock#810. Looks like things were broken during the quickjump refactor of the JS. For the (git) record: I do not think the style switcher is a good idea. I'm fixing it for the same reason @mzero added it; as an answer to "rumblings from some that they didn't want their pixels changed on bit" - - - - - 2a1d620f by Alec Theriault at 2018-10-24T17:38:07-07:00 Fix copy-pasta error in data-files - - - - - ed5bfb7f by Alec Theriault at 2018-10-24T20:42:14-07:00 Fix the synopsis button Here's these changes are supposed to do: * put the synopsis back on the right side * properly have it on the edge of the screen on wide screens * adjust the background of the synopsis to match the button (otherwise the grey blends in with what is underneath) * get rid of the dotted purple line * the synopsis contents are now scrollable even when in wide screens (this has been a long-standing bug) - - - - - 883fd74b by Alec Theriault at 2018-10-25T20:16:46-07:00 Avoid more conflicts in generated ids (#954) This fixes haskell/haddock#953 by passing more names into the generated ids. - - - - - ea54e331 by Alec Theriault at 2018-10-25T21:07:12-07:00 Don't hide bullets in method docs I think thst CSS was meant only to deal with fields and the effect on bullets was accidental. Fixes haskell/haddock#926. - - - - - 9a14ef4a by Alec Theriault at 2018-10-25T22:02:07-07:00 Indent more things + slightly smaller font - - - - - b9f17e29 by Alec Theriault at 2018-10-25T22:10:01-07:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 096a3cfa by Alec Theriault at 2018-10-25T22:24:38-07:00 Accept HTML output - - - - - 2669517d by Alec Theriault at 2018-10-26T09:02:35-07:00 User manual + stuff for building GHC docs - - - - - 46b27687 by Alec Theriault at 2018-10-26T09:10:59-07:00 Make 'Contents' in NewOcean scrollable This only happens if the contents block on the left is so big that it doesn't fit (vertically) on the page. If that happens, we want it to be scrollable. - - - - - 3443dd94 by Alec Theriault at 2018-10-26T09:36:46-07:00 Revert "Make 'Contents' in NewOcean scrollable" This reverts commit f909ffd8353d6463fd5dd184998a32aa98d5c922. I missed the fact this also forces the 'Contents' to always go down to the bottom of the page. - - - - - ed081424 by Alec Theriault at 2018-10-26T14:22:23-07:00 Avoid some partiality AFAICT this wasn't causing any crashes, but that's mostly because we happen not to be forcing `pkgStr` when it would diverge. We come dangerously close to doing that in `ppHtmlIndex`. Fixes haskell/haddock#569. - - - - - 6a5bec41 by Alec Theriault at 2018-10-27T10:05:04-07:00 Fix documentation in `haddock-api` (#957) * Fix misplaced Haddocks in Haddock itself Haddock should be able to generate documentation for 'haddock-api' again. * Make CI check that documentation can be built. * Add back a doc that is OK - - - - - 5100450a by Matthew Yacavone at 2018-10-27T14:51:38-04:00 More explicit foralls (GHC Proposal 0007) - - - - - 8771a6b0 by Alec Theriault at 2018-11-05T13:58:11-08:00 Only run MathJax on entities with "mathjax" class (#960) Correspondingly, we wrap all inline/diplay math in <span class="mathjax"> ... the math .... </span> This fixes haskell/haddock#959. - - - - - bd7ff5c5 by Alec Theriault at 2018-11-05T15:54:22-08:00 Deduplicate some work in 'AttachInstances' Perf only change: * avoid needlessly union-ing maps * avoid synify-ing instances twice Took this opportunity to add some docs too - - - - - cf99fd8f by Alec Theriault at 2018-11-05T15:54:22-08:00 Specialize some SYB functions Perf only change: * Add a 'SPECIALIZE' pragma to help GHC optimize a 'Data a =>' constraint * Manually specialize the needlessly general type of 'specializeTyVarBndrs' - - - - - 4f91c473 by Alec Theriault at 2018-11-05T15:54:22-08:00 Improve perf of renaming Perf only change: * don't look up type variable names (they're never in the environment) * use a difference list for accumulating missing names * more efficient 'Functor'/'Applicative' instances for 'RnM' - - - - - 4bbab0d4 by Alec Theriault at 2018-11-05T15:54:22-08:00 Faster 'Text' driven parser combinators Perf only change: * use 'getParserState'/'setParserState' to make 'Text'-optimized parser combinators * minimize uses of 'Data.Text.{pack,unpack,cons,snoc}' - - - - - fa430c02 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support hyperlink labels with inline markup The parser for pictures hasn't been properly adjusted yet. - - - - - c1431035 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support (and flatten) inline markup in image links Inline markup is supported in image links but, as per the [commonmark recommendation][0], it is stripped back to a plain text representation. [0]: https://spec.commonmark.org/0.28/#example-547 - - - - - d4ee1ba5 by Alec Theriault at 2018-11-06T12:03:24-08:00 Accept test case - - - - - 8088aeb1 by Alec Theriault at 2018-11-06T12:03:24-08:00 Fix/add to haddock-library test suite - - - - - e78f644d by Alec Theriault at 2018-11-06T13:26:31-08:00 Bump version bounds - - - - - 644335eb by Alec Theriault at 2018-11-06T13:53:30-08:00 Merge pull request haskell/haddock#875 from harpocrates/feature/markup-in-hyperlinks Inline markup in markdown-style links and images - - - - - e173ed0d by Alec Theriault at 2018-11-07T12:37:18-08:00 Fix issues around plus/minus * swap the minimize unicode to something more intuitive * use new unicode expander/collapser for instance lists * address some alignment issues in the "index" page - - - - - b2d92df7 by Alec Theriault at 2018-11-07T13:41:57-08:00 Allow "Contents" summary to scroll in a fixed div In the unfortunate event that the "Contents" summary doesn't fit vertically (like in the "Prelude"), it will be scrollable. - - - - - ca704c23 by Alec Theriault at 2018-11-07T13:45:15-08:00 Accept HTML output changes - - - - - 82c0ec6d by Alec Theriault at 2018-11-07T18:12:54-08:00 overflow-y 'scroll' -> 'auto' - - - - - 571d7657 by Alec Theriault at 2018-11-08T19:44:12-08:00 Clicking on "Contents" navigates to top of page - - - - - 8065a012 by Alec Theriault at 2018-11-08T19:44:17-08:00 Space out functions more Also, functions and data decls now have the same space before and after them. - - - - - cc650ede by Alec Theriault at 2018-11-09T08:13:35-08:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 65f8c17f by Alec Theriault at 2018-11-10T14:04:06-08:00 Update changelog - - - - - 20473847 by Alec Theriault at 2018-11-10T14:21:40-08:00 Replace oplus/ominus expander/collapser icons with triangles - - - - - 16592957 by Alec Theriault at 2018-11-10T14:35:10-08:00 Merge pull request haskell/haddock#949 from haskell/wip/new-ocean Introduce NewOcean theme. - - - - - 357cefe1 by Alec Theriault at 2018-11-10T16:02:13-08:00 Merge branch 'ghc-8.6' into ghc-head - - - - - de612267 by Alec Theriault at 2018-11-11T20:01:21-08:00 Rename 'NewOcean' theme to 'Linuwial' - - - - - 954b5baa by Alec Theriault at 2018-11-12T08:33:18-08:00 Add blockquote styling Matches b71da1feabf33efbbc517ac376bb690b5a604c2f from hackage-server. Fixes haskell/haddock#967. - - - - - d32c0b0b by Fangyi Zhou at 2018-11-12T10:24:13-08:00 Fix some broken links (#15733) Summary: For links in subpackages as well. https://phabricator.haskell.org/D5257 Test Plan: Manually verify links Reviewers: mpickering, bgamari, osa1 Reviewed By: osa1 GHC Trac Issues: haskell/haddock#15733 Differential Revision: https://phabricator.haskell.org/D5262 - - - - - 41098b1f by Alp Mestanogullari at 2018-11-15T22:40:09+01:00 Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change It got introduced in ghc/ghc at ae2c9b40f5b6bf272251d1f4107c60003f541b62. - - - - - c5c1c7e0 by Alec Theriault at 2018-11-15T13:48:13-08:00 Merge pull request haskell/haddock#970 from alpmestan/alp/fix-promotionflag Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change - - - - - 6473d3a4 by Shayan-Najd at 2018-11-23T01:38:49+01:00 [TTG: Handling Source Locations] Foundation and Pat Trac Issues haskell/haddock#15495 This patch removes the ping-pong style from HsPat (only, for now), using the plan laid out at https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow/HandlingSourceLocations (solution A). - the class `HasSrcSpan`, and its functions (e.g., `cL` and `dL`), are introduced - some instances of `HasSrcSpan` are introduced - some constructors `L` are replaced with `cL` - some patterns `L` are replaced with `dL->L` view pattern - some type annotation are necessarily updated (e.g., `Pat p` --> `Pat (GhcPass p)`) - - - - - 7a088dfe by Alec Theriault at 2018-11-26T11:11:28-08:00 More uniform handling of `forall`'s in HTML/LaTeX * don't forget to print explicit `forall`'s when there are arg docs * when printing an explicit `forall`, print all tyvars Fixes haskell/haddock#973 - - - - - d735e570 by Alec Theriault at 2018-12-12T08:42:09-08:00 Fix warnings, accept output * remove redundant imports (only brought to light due to recent work for improving redundant import detection) * fix a bug that was casuing exports to appear in reverse order * fix something in haddock-library that prevented compilation on old GHC's - - - - - a3852f8a by Zejun Wu at 2018-12-14T09:37:47-05:00 Output better debug infromation on internal error in extractDecl This will make investigation of haskell/haddock#979 easier - - - - - 2eccb5b9 by Alec Theriault at 2018-12-17T09:25:10-05:00 Refactor names + unused functions (#982) This commit should not introduce any change in functionality! * consistently use `getOccString` to convert `Name`s to strings * compare names directly when possible (instead of comparing strings) * get rid of unused utility functions - - - - - e82e4df8 by Alec Theriault at 2018-12-20T16:16:30-05:00 Load plugins when compiling each module (#983) * WIP: Load (typechecker) plugins from language pragmas * Revert "Load plugins when starting a GHC session (#905)" This reverts commit 72d82e52f2a6225686d9668790ac33c1d1743193. * Simplify plugin initialization code - - - - - 96e86f38 by Alec Theriault at 2018-12-23T10:23:20-05:00 Properly synify and render promoted type variables (#985) * Synify and render properly promoted type variables Fixes haskell/haddock#923. * Accept output - - - - - 23343345 by Alec Theriault at 2018-12-27T16:39:38-05:00 Remove `haddock-test`'s dep. on `syb` (#987) The functionality is easily inlined into one short function: `gmapEverywhere`. This doesn't warrant pulling in another package. - - - - - d0734f21 by Alec Theriault at 2018-12-27T16:39:52-05:00 Address deprecation warnings in `haddock-test` (#988) Fixes haskell/haddock#885. - - - - - 4d9f144e by mynguyen at 2018-12-30T23:42:26-05:00 Visible kind application haddock update - - - - - ffe0e9ed by Alec Theriault at 2019-01-07T13:55:22-08:00 Print kinded tyvars in constructors for Hoogle (#993) Fixes haskell/haddock#992 - - - - - 2e18b55d by Alec Theriault at 2019-01-10T16:42:45-08:00 Accept new output `GHC.Maybe` -> `Data.Maybe` (#996) Since 53874834b779ad0dfbcde6650069c37926da1b79 in GHC, "GHC.Maybe" is marked as `not-home`. That changes around some test output. - - - - - 055da666 by Gabor Greif at 2019-01-22T14:41:51+01:00 Lone typofix - - - - - 01bb71c9 by Alec Theriault at 2019-01-23T11:46:46-08:00 Keep forall on H98 existential data constructors (#1003) The information about whether or not there is a source-level `forall` is already available on a `ConDecl` (as `con_forall`), so we should use it instead of always assuming `False`! Fixes haskell/haddock#1002. - - - - - f9b9bc0e by Ryan Scott at 2019-01-27T09:28:12-08:00 Fix haskell/haddock#1004 with a pinch of dropForAlls - - - - - 5cfcdd0a by Alec Theriault at 2019-01-28T16:49:57-08:00 Loosen 'QuickCheck' and 'hspec' bounds It looks like the new versions don't cause any breakage and loosening the bounds helps deps fit in one stack resolver. - - - - - 3545d3dd by Alec Theriault at 2019-01-31T01:37:25-08:00 Use `.hie` files for the Hyperlinker backend (#977) # Summary This is a large architectural change to the Hyperlinker. * extract link (and now also type) information from `.hie` instead of doing ad-hoc SYB traversals of the `RenamedSource`. Also adds a superb type-on-hover feature (#715). * re-engineer the lexer to avoid needless string conversions. By going directly through GHC's `P` monad and taking bytestring slices, we avoid a ton of allocation and have better handling of position pragmas and CPP. In terms of performance, the Haddock side of things has gotten _much_ more efficient. Unfortunately, much of this is cancelled out by the increased GHC workload for generating `.hie` files. For the full set of boot libs (including `ghc`-the-library) * the sum of total time went down by 9-10% overall * the sum of total allocations went down by 6-7% # Motivation Haddock is moving towards working entirely over `.hi` and `.hie` files. This change means we no longer need the `RenamedSource` from `TypecheckedModule` (something which is _not_ in `.hi` files). # Details Along the way a bunch of things were fixed: * Cross package (and other) links are now more reliable (#496) * The lexer tries to recover from errors on every line (instead of at CPP boundaries) * `LINE`/`COLUMN` pragmas are taken into account * filter out zero length tokens before rendering * avoid recomputing the `ModuleName`-based `SrcMap` * remove the last use of `Documentation.Haddock.Utf8` (see haskell/haddock#998) * restructure temporary folder logic for `.hi`/`.hie` model - - - - - 2ded3359 by Herbert Valerio Riedel at 2019-02-02T12:06:12+01:00 Update/modernise haddock-library.cabal file - - - - - 62b93451 by Herbert Valerio Riedel at 2019-02-02T12:19:31+01:00 Tentatively declare support for unreleased base-4.13/ghc-8.8 - - - - - 6041e767 by Herbert Valerio Riedel at 2019-02-02T16:04:32+01:00 Normalise LICENSE text w/ cabal's BSD2 template Also, correct the `.cabal` files to advertise `BSD2` instead of the incorrect `BSD3` license. - - - - - 0b459d7f by Alec Theriault at 2019-02-02T18:06:12-08:00 CI: fetch GHC from validate artifact Should help make CI be less broken - - - - - 6b5c07cf by Alec Theriault at 2019-02-02T18:06:12-08:00 Fix some Hyperlinker test suite fallout * Amend `ParserSpec` to match new Hyperlinker API - pass in compiler info - strip out null tokens * Make `hypsrc-test` pass reliably - strip out `local-*` ids - strip out `line-*` ids from the `ClangCppBug` test - re-accept output - - - - - ded34791 by Nathan Collins at 2019-02-02T18:31:23-08:00 Update README instructions for Stack No need to `stack install` Haddock to test it. Indeed, `stack install` changes the `haddock` on user's `PATH` if `~/.local/bin` is on user's `PATH` which may not be desirable when hacking on Haddock. - - - - - 723298c9 by Alec Theriault at 2019-02-03T09:11:05-08:00 Remove `Documentation.Haddock.Utf8` The circumstances under which this module appeared are completely gone. The Hyperlinker backend no longer needs this module (it uses the more efficient `Encoding` module from `ghc`). Why no deprecation? Because this module really shouldn't exist! - It isn't used in `haddock-library`/`haddock-api` anymore - It was copy pasted directly from `utf8-string` - Folks seeking a boot-lib only solution can use `ghc`'s `Encoding` - - - - - 51050006 by Alec Theriault at 2019-02-03T22:58:58-08:00 Miscellaneous improvements to `Convert` (#1020) Now that Haddock is moving towards working entirely over `.hi` and `.hie` files, all declarations and types are going to be synthesized via the `Convert` module. In preparation for this change, here are a bunch of fixes to this module: * Add kind annotations to type variables in `forall`'s whose kind is not `Type`, unless the kind can be inferred from some later use of the variable. See `implicitForAll` and `noKindTyVars` in particular if you wish to dive into this. * Properly detect `HsQualTy` in `synifyType`. This is done by following suit with what GHC's `toIfaceTypeX` does and checking the first argument of `FunTy{} :: Type` to see if it classified as a given/wanted in the typechecker (see `isPredTy`). * Beef up the logic around figuring out when an explicit `forall` is needed. This includes: observing if any of the type variables will need kind signatures, if the inferred type variable order _without_ a forall will still match the one GHC claims, and some other small things. * Add some (not yet used) functionality for default levity polymorphic type signatures. This functionality similar to `fprint-explicit-runtime-reps`. Couple other smaller fixes only worth mentioning: * Show the family result signature only when it isn't `Type` * Fix rendering of implicit parameters in the LaTeX and Hoogle backends * Better handling of the return kind of polykinded H98 data declarations * Class decls produced by `tyThingToLHsDecl` now contain associated type defaults and default method signatures when appropriate * Filter out more `forall`'s in pattern synonyms - - - - - 841980c4 by Oleg Grenrus at 2019-02-04T08:44:25-08:00 Make a fixture of weird parsing of lists (#997) The second example is interesting. If there's a list directly after the header, and that list has deeper structure, the parser is confused: It finds two lists: - One with the first nested element, - everything after it I'm not trying to fix this, as I'm not even sure this is a bug, and not a feature. - - - - - 7315c0c8 by Ryan Scott at 2019-02-04T12:17:56-08:00 Fix haskell/haddock#1015 with dataConUserTyVars (#1022) The central trick in this patch is to use `dataConUserTyVars` instead of `univ_tvs ++ ex_tvs`, which displays the foralls in a GADT constructor in a way that's more faithful to how the user originally wrote it. Fixes haskell/haddock#1015. - - - - - ee0b49a3 by Ryan Scott at 2019-02-04T15:25:17-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. - - - - - 1c850dc8 by Alan Zimmerman at 2019-02-05T21:54:18+02:00 Matching changes in GHC for haskell/haddock#16236 - - - - - ab03c38e by Simon Marlow at 2019-02-06T08:07:33+00:00 Merge pull request haskell/haddock#1014 from hvr/pr/bsd2-normalise Normalise LICENSE text w/ cabal's BSD2 template - - - - - 5a92ccae by Alec Theriault at 2019-02-10T06:21:55-05:00 Merge remote-tracking branch 'gitlab/wip/T16236-2' into ghc-head - - - - - c0485a1d by Alec Theriault at 2019-02-10T03:32:52-08:00 Removes `haddock-test`s dependency on `xml`/`xhtml` (#1027) This means that `html-test`, `latex-test`, `hoogle-test`, and `hypsrc-test` now only depend on GHC boot libs. So we should now be able to build and run these as part of GHC's testsuite. \o/ The reference output has changed very slightly, in three ways: * we don't convert quotes back into `&quot;` as the `xml` lib did * we don't add extra `&nbsp;` as the `xml` lib did * we now remove the entire footer `div` (instead of just emptying it) - - - - - 65a448e3 by Alec Theriault at 2019-02-11T12:27:41-05:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - 360ca937 by Alec Theriault at 2019-02-13T11:36:11-05:00 Clean up logic for guessing `-B` and `--lib` (#1026) Haddock built with the `in-ghc-tree` flag tries harder to find the GHC lib folder and its own resources. This should make it possible to use `in-ghc-tree`-built Haddock without having to specify the `-B` and `--lib` options (just how you can use in-tree GHC without always specifying the `-B` option). The logic to do this relies on `getExecutablePath`, so we only get this auto-detection on platforms where this function works. - - - - - d583e364 by Alec Theriault at 2019-02-16T10:41:22-05:00 Fix tests broken by GHC Changes in 19626218566ea709b5f6f287d3c296b0c4021de2 affected some of the hyperlinker output. Accepted the new output (hovering over a `..` now shows you what that wildcard binds). Also fixed some stray deprecation warnings. - - - - - da0c42cc by Vladislav Zavialov at 2019-02-17T11:39:19+03:00 Parser changes to match !380 - - - - - ab96bed7 by Ryan Scott at 2019-02-18T04:44:08-05:00 Bump ghc version to 8.9 - - - - - 44b7c714 by Alec Theriault at 2019-02-22T05:49:43-08:00 Match GHC changes for T16185 `FunTy` now has an `AnonArgFlag` that indicates whether the arrow is a `t1 => t2` or `t1 -> t2`. This commit shouldn't change any functionality in Haddock. - - - - - 2ee653b1 by Alec Theriault at 2019-02-24T18:53:33-08:00 Update .travis.yml Points to the new GHC CI artifact. - - - - - 90939d71 by Alec Theriault at 2019-02-25T00:42:41-08:00 Support value/type namespaces on identifier links Identifier links can be prefixed with a 'v' or 't' to indicate the value or type namespace of the desired identifier. For example: -- | Some link to a value: v'Data.Functor.Identity' -- -- Some link to a type: t'Data.Functor.Identity' The default is still the type (with a warning about the ambiguity) - - - - - d6ed496c by Alec Theriault at 2019-02-25T00:42:46-08:00 Better identifier parsing * '(<|>)' and '`elem`' now get parsed and rendered properly as links * 'DbModule'/'DbUnitId' now properly get split apart into two links * tuple names now get parsed properly * some more small niceties... The identifier parsing code is more precise and more efficient (although to be fair: it is also longer and in its own module). On the rendering side, we need to pipe through information about backticks/parens/neither all the way through from renaming to the backends. In terms of impact: a total of 35 modules in the entirety of the bootlib + ghc lib docs change. The only "regression" is things like '\0'. These should be changed to @\\0@ (the path by which this previously worked seems accidental). - - - - - 3c3b404c by Alec Theriault at 2019-02-25T22:12:11-08:00 Fix standalone deriving docs Docs on standalone deriving decls for classes with associated types should be associated with the class instance, not the associated type instance. Fixes haskell/haddock#1033 - - - - - d51ef69e by Alec Theriault at 2019-02-26T19:14:59-08:00 Fix bogus identifier defaulting This avoids a situation in which an identifier would get defaulted to a completely different identifier. Prior to this commit, the 'Bug1035' test case would hyperlink 'Foo' into 'Bar'! Fixes haskell/haddock#1035. - - - - - 88cbbdc7 by Ryan Scott at 2019-02-27T10:14:03-05:00 Visible dependent quantification (#16326) changes - - - - - 0dcf6cee by Xia Li-yao at 2019-02-27T21:53:27-05:00 Menu item controlling which instances are expanded/collapsed (#1007) Adds a menu item (like "Quick Jump") for options related to displaying instances. This provides functionality for: * expanding/collapsing all instances on the currently opened page * controlling whether instances are expanded/collapsed by default * controlling whether the state of instances should be "remembered" This new functionality is implemented in Typescript in `details-helper`. The built-in-themes style switcher also got a revamp so that all three of QuickJump, the style switcher, and instance preferences now have the same style and implementation structure. See also: https://mail.haskell.org/pipermail/haskell-cafe/2019-January/130495.html Fixes haskell/haddock#698. Co-authored-by: Lysxia <lysxia at gmail.com> Co-authored-by: Nathan Collins <conathan at galois.com> - - - - - 3828c0fb by Alec Theriault at 2019-02-28T12:42:49-05:00 `--show-interface` should output to stdout. (#1040) Fixes haskell/haddock#864. - - - - - a50f4cda by gbaz at 2019-03-01T07:43:16-08:00 Increase contrast of Linuwal theme (#1037) This is to address the concern that, on less nice and older screens, some of the shades of grey blend in too easily with the white background. * darken the font slightly * darken slightly the grey behind type signatures and such * add a border and round the corners on code blocks * knock the font down by one point - - - - - ab4d41de by Alec Theriault at 2019-03-03T09:23:26-08:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 12f509eb by Ben Gamari at 2019-03-04T22:13:20-05:00 Remove reference to Opt_SplitObjs flag Split-objects has been removed. - - - - - 5b3e4c9a by Ryan Scott at 2019-03-06T19:16:24-05:00 Update html-test output to reflect haskell/haddock#16391 changes - - - - - fc228af1 by Alec Theriault at 2019-03-09T08:29:23-08:00 Match changes for "Stop inferring over-polymorphic kinds" The `hsq_ext` field of `HsQTvs` is now just the implicit variables (instead of also including information about which of these variables are dependent). This commit shouldn't change any functionality in Haddock. - - - - - 6ac109eb by Alec Theriault at 2019-03-09T11:22:55-08:00 Add .hi, .dyn_hi, etc files to .gitignore Fixes haskell/haddock#1030. - - - - - b55f0c05 by Alec Theriault at 2019-03-09T11:22:55-08:00 Better support for default methods in classes * default methods now get rendered differently * default associated types get rendered * fix a forgotten `s/TypeSig/ClassOpSig/` refactor in LaTeX backend * LaTeX backend now renders default method signatures NB: there is still no way to document default class members and the NB: LaTeX backend still crashes on associated types - - - - - 10aea0cf by Alec Theriault at 2019-03-09T11:22:55-08:00 Avoid multi-line `emph` in LaTeX backend `markupWarning` often processes inputs which span across paragraphs. Unfortunately, LaTeX's `emph` is not made to handle this (and will crash). Fixes haskell/haddock#936. - - - - - d22dc2c9 by Alec Theriault at 2019-03-09T11:22:55-08:00 Many LaTeX backend fixes After this commit, we can run with `--latex` on all boot libraries without crashing (although the generated LaTeX still fails to compile in a handful of larger packages like `ghc` and `base`). * Add newlines after all block elements in LaTeX. This is important to prevent the final output from being more an more indented. See the `latext-test/src/Example` test case for a sample of this. * Support associated types in class declarations (but not yet defaults) * Several small issues for producing compiling LaTeX; - avoid empy `\haddockbeginargs` lists (ex: `type family Any`) - properly escape identifiers depending on context (ex: `Int#`) - add `vbox` around `itemize`/`enumerate` (so they can be in tables) * Several spacing fixes: - limit the width of `Pretty`-arranged monospaced code - cut out extra space characters in export lists - only escape spaces if there are _multiple_ spaces - allow type signatures to be multiline (even without docs) * Remove uninteresting and repetitive `main.tex`/`haddock.sty` files from `latex-test` test reference output. Fixes haskell/haddock#935, haskell/haddock#929 (LaTeX docs for `text` build & compile) Fixes haskell/haddock#727, haskell/haddock#930 (I think both are really about type families...) - - - - - 0e6cee00 by Alec Theriault at 2019-03-29T12:11:56-07:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - ce05434d by Alan Zimmerman at 2019-03-29T12:12:11-07:00 Matching changes in GHC for haskell/haddock#16236 (cherry picked from commit 3ee6526d4ae7bf4deb7cd1caf24b3d7355573576) - - - - - d85766b2 by Ben Gamari at 2019-03-29T12:14:04-07:00 Bump GHC to 8.8 - - - - - 5a82cbaf by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Redo ParseModuleHeader - - - - - b9033348 by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - bb55c8f4 by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove outdated `.ghci` files and `scripts` The `.ghci` files are actively annoying when trying to `cabal v2-repl`. As for the `scripts`, the distribution workflow is completely different. - - - - - 5ee244dc by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove obsolete arcanist files + STYLE Now that GHC is hosted on Gitlab, the arcanist files don't make sense anymore. The STYLE file contains nothing more than a dead link too. - - - - - d07c1928 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Redo ParseModuleHeader - - - - - 492762d2 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - af2ac773 by Ryan Scott at 2019-05-14T17:22:13-04:00 Changes for haskell/haddock#16110/#16356 - - - - - 6820ed0d by Alec Theriault at 2019-05-17T08:51:27-07:00 Unbreak haskell/haddock#1004 test case `fail` is no longer part of `Monad`. - - - - - 6bf7be98 by Alec Theriault at 2019-05-17T08:51:27-07:00 Fix haskell/haddock#1063 with better parenthesization logic for contexts The only other change in html/hoogle/hyperlinker output for the boot libraries that this caused is a fix to some Hoogle output for implicit params. ``` $ diff -r _build/docs/ old_docs diff -r _build/docs/html/libraries/base/base.txt old_docs/html/libraries/base/base.txt 13296c13296 < assertError :: (?callStack :: CallStack) => Bool -> a -> a --- > assertError :: ?callStack :: CallStack => Bool -> a -> a ``` - - - - - b5716b61 by Ryan Scott at 2019-05-22T17:24:32-04:00 Match changes with haskell/haddock#14332 - - - - - c115abf6 by Alec Theriault at 2019-05-26T16:01:58-04:00 Remove Haddock's dependency on `Cabal` At this point, Haddock depended on Cabal-the-library solely for a verbosity parser (which misleadingly accepts all sorts of verbosity options that Haddock never uses). Now, the only dependency on Cabal is for `haddock-test` (which uses Cabal to locate the Haddock interface files of a couple boot libraries). - - - - - e5b2d4a3 by Alec Theriault at 2019-05-26T16:16:25-04:00 Regression test: promoted lists in associated types When possible, associated types with promoted lists should use the promoted list literal syntax (instead of repeated applications of ': and '[]). This was fixed in 2122de5473fd5b434af690ff9ccb1a2e58491f8c. Closes haskell/haddock#466, - - - - - cc5ad5d3 by Alec Theriault at 2019-05-26T17:55:54-04:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 4b3301a6 by Alec Theriault at 2019-05-26T17:57:52-04:00 Release haddock-2.23, haddock-library-1.8.0 Tentatively adjust bounds and changelogs for the release to be bundled with GHC 8.8.1. - - - - - 69c7cfce by Matthew Pickering at 2019-05-30T10:54:27+01:00 Update hyperlinker tests for new types in .hie files - - - - - 29b7e738 by Zubin Duggal at 2019-05-30T10:57:51+01:00 update for new way to store hiefile headers - - - - - aeca5d5f by Zubin Duggal at 2019-06-04T18:57:42-04:00 update for new way to store hiefile headers - - - - - ba2ca518 by Ben Gamari at 2019-06-07T23:11:14+00:00 Update test output for introduction of Safe-Inferred - - - - - 3a975a6c by Ryan Scott at 2019-07-03T12:06:27-04:00 Changes for haskell/haddock#15247 - - - - - 0df46555 by Zubin Duggal at 2019-07-22T10:52:50+01:00 Fix haddockHypsrcTest - - - - - 2688686b by Sylvain Henry at 2019-09-12T23:19:39+02:00 Fix for GHC module renaming - - - - - 9ec0f3fc by Alec Theriault at 2019-09-20T03:21:00-04:00 Fix Travis CI, loosen .cabal bounds (#1089) Tentatively for the 2.23 release: * updated Travis CI to work again * tweaked bounds in the `.cabal` files * adjusted `extra-source-files` to properly identify test files - - - - - ca559beb by Matthías Páll Gissurarson at 2019-09-28T12:14:40-04:00 Small change in to facilitate extended typed-holes (#1090) This change has no functional effect on haddock itself, it just changes one pattern to use `_ (` rather than `_(`, so that we may use `_(` as a token for extended typed-holes later. - - - - - 02e28976 by Vladislav Zavialov at 2019-09-28T12:17:45-04:00 Remove spaces around @-patterns (#1093) This is needed to compile `haddock` when [GHC Proposal haskell/haddock#229](https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0229-whitespace-bang-patterns.rst) is implemented. - - - - - 83cbbf55 by Alexis King at 2019-09-30T21:12:42-04:00 Fix the ignore-exports option (#1082) The `ignore-exports` option has been broken since haskell/haddock#688, as mentioned in https://github.com/haskell/haddock/pull/766#issue-172505043. This PR fixes it. - - - - - e127e0ab by Ben Gamari at 2019-10-06T15:12:06-04:00 Fix a few haddock issues - - - - - 3a0f5c89 by Zubin Duggal at 2019-10-07T17:56:13-04:00 Fix crash when there are no srcspans in the file due to CPP - - - - - 339c5ff8 by Alec Theriault at 2019-10-07T17:56:13-04:00 Prefer un-hyperlinked sources to no sources It is possible to fail to extract an HIE ast. This is however not a reason to produce _no_ output - we should still make a colorized HTML page. - - - - - d47ef478 by Alec Theriault at 2019-10-07T17:56:13-04:00 Add a regression test for haskell/haddock#1091 Previously, this input would crash Haddock. - - - - - ed7c8b0f by Alec Theriault at 2019-10-07T20:56:48-04:00 Add Hyperlinker test cases for TH-related stuff Hopefully this will guard against regressions around quasiquotes, TH quotes, and TH splices. - - - - - d00436ab by Andreas Klebinger at 2019-10-21T15:53:03+02:00 Refactor for withTiming changes. - - - - - 4230e712 by Ben Gamari at 2019-10-22T09:36:37-04:00 Merge pull request haskell/haddock#1101 from AndreasPK/withTimingRefactor Refactor for withTiming changes. - - - - - d155c5f4 by Ryan Scott at 2019-10-23T10:37:17-04:00 Reify oversaturated data family instances correctly (#1103) This fixes haskell/haddock#1103 by adapting the corresponding patch for GHC (see https://gitlab.haskell.org/ghc/ghc/issues/17296 and https://gitlab.haskell.org/ghc/ghc/merge_requests/1877). - - - - - 331a5adf by Sebastian Graf at 2019-10-25T17:14:40+02:00 Refactor for OutputableBndrId changes - - - - - 48a490e0 by Ben Gamari at 2019-10-27T10:16:16-04:00 Merge pull request haskell/haddock#1105 from sgraf812/wip/flexible-outputable Refactor for OutputableBndrId changes - - - - - f62a7dfc by Sebastian Graf at 2019-11-01T11:54:16+00:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - d9b242ed by Ryan Scott at 2019-11-03T13:20:03-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. (cherry picked from commit cfd682c5fd03b099a3d78c44f9279faf56a0ac70) - - - - - dfd42406 by Sebastian Graf at 2019-11-04T07:02:14-05:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - 0b15be7c by Ben Gamari at 2019-11-09T13:21:33-05:00 Import isRuntimeRepVar from Type rather than TyCoRep isRuntimeRepVar is not longer exported from TyCoRep due to ghc#17441. - - - - - 091f7283 by Ben Gamari at 2019-11-10T12:47:06-05:00 Bump to GHC 8.10 - - - - - e88c71f2 by Ben Gamari at 2019-11-14T00:22:24-05:00 Merge pull request haskell/haddock#1110 from haskell/wip/T17441 Import isRuntimeRepVar from Type rather than TyCoRep - - - - - 4e0bbc17 by Ben Gamari at 2019-11-14T00:22:45-05:00 Version bumps for GHC 8.11 - - - - - 0e85ceb4 by Ben Gamari at 2019-11-15T11:59:45-05:00 Bump to GHC 8.10 - - - - - 00d6d68b by Ben Gamari at 2019-11-16T18:35:58-05:00 Bump ghc version to 8.11 - - - - - dde1fc3f by Ben Gamari at 2019-11-16T20:40:37-05:00 Drop support for base 4.13 - - - - - f52e331d by Vladislav Zavialov at 2019-11-24T13:02:28+03:00 Update Hyperlinker.Parser.classify to use ITdollar - - - - - 1ad96198 by Vladislav Zavialov at 2019-11-28T16:12:33+03:00 Remove HasSrcSpan (#17494) - - - - - 651afd70 by Herbert Valerio Riedel at 2019-12-08T12:08:16+01:00 Document error-prone conditional definition of instances This can easily trip up people if one isn't aware of it. Usually it's better to avoid this kind of conditionality especially for typeclasses for which there's an compat-package as conditional instances like these tend to fragment the ecosystem into those packages that go the extra mile to provide backward compat via those compat-packages and those that fail to do so. - - - - - b521af56 by Herbert Valerio Riedel at 2019-12-08T12:09:54+01:00 Fix build-failure regression for base < 4.7 The `$>` operator definition is available only since base-4.7 which unfortunately wasn't caught before release to Hackage (but has been fixed up by a metadata-revision) This commit introduces a `CompatPrelude` module which allows to reduce the amount of CPP by ousting it to a central location, i.e. the new `CompatPrelude` module. This pattern also tends to reduce the tricks needed to silence unused import warnings. Addresses haskell/haddock#1119 - - - - - 556c375d by Sylvain Henry at 2020-01-02T19:01:55+01:00 Fix after Iface modules renaming - - - - - bd6c53e5 by Sylvain Henry at 2020-01-07T00:48:48+01:00 hsyl20-modules-renamer - - - - - fb23713b by Ryan Scott at 2020-01-08T07:41:13-05:00 Changes for GHC#17608 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2372 - - - - - 4a4dd382 by Ryan Scott at 2020-01-25T08:08:26-05:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - e782a44d by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename PackageConfig into UnitInfo - - - - - ba3c9f05 by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename lookupPackage - - - - - ab37f9b3 by Ben Gamari at 2020-01-29T13:00:44-05:00 Merge pull request haskell/haddock#1125 from haskell/wip/T17566-take-two Changes for GHC#17566 - - - - - 3ebd5ae0 by Ryan Scott at 2020-01-31T05:56:50-05:00 Merge branch 'wip-hsyl20-package-refactor' into ghc-head - - - - - 602a747e by Richard Eisenberg at 2020-02-04T09:05:43+00:00 Echo GHC's removal of PlaceHolder module This goes with GHC's !2083. - - - - - ccfe5679 by Sylvain Henry at 2020-02-10T10:13:56+01:00 Module hierarchy: runtime (cf haskell/haddock#13009) - - - - - 554914ce by Cale Gibbard at 2020-02-10T16:10:39-05:00 Fix build of haddock in stage1 We have to use the correct version of the GHC API, but the version of the compiler itself doesn't matter. - - - - - 5b6fa2a7 by John Ericson at 2020-02-10T16:18:07-05:00 Noramlize `tested-with` fields in cabal files - - - - - e6eb3ebe by Vladislav Zavialov at 2020-02-16T13:25:26+03:00 No MonadFail/Alternative for P - - - - - 90e181f7 by Ben Gamari at 2020-02-18T14:13:47-05:00 Merge pull request haskell/haddock#1129 from obsidiansystems/wip/fix-stage1-build Fix build of haddock in stage1 - - - - - 93b64636 by Sylvain Henry at 2020-02-19T11:20:27+01:00 Modules: Driver (#13009) - - - - - da4f6c7b by Vladislav Zavialov at 2020-02-22T15:33:02+03:00 Use RealSrcSpan in InstMap - - - - - 479b1b50 by Ben Gamari at 2020-02-23T10:28:13-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 55ecacf0 by Sylvain Henry at 2020-02-25T15:18:27+01:00 Modules: Core (#13009) - - - - - 60867b3b by Vladislav Zavialov at 2020-02-28T15:53:52+03:00 Ignore the BufLoc/BufSpan added in GHC's !2516 - - - - - 1e5506d3 by Sylvain Henry at 2020-03-02T12:32:43+01:00 Modules: Core (#13009) - - - - - 6fb53177 by Richard Eisenberg at 2020-03-09T14:49:40+00:00 Changes in GHC's !1913. - - - - - 30b792ea by Ben Gamari at 2020-03-16T12:45:02-04:00 Merge pull request haskell/haddock#1130 from hsyl20/wip/hsyl20-modules-core2 Modules: Core (#13009) - - - - - cd761ffa by Sylvain Henry at 2020-03-18T15:24:00+01:00 Modules: Types - - - - - b6646486 by Ben Gamari at 2020-03-18T14:42:43-04:00 Merge pull request haskell/haddock#1133 from hsyl20/wip/hsyl20/modules/types Modules: Types - - - - - 9325d734 by Kleidukos at 2020-03-19T12:38:31-04:00 Replace the 'caption' class so that the collapsible sections are shown - - - - - 5e2bb555 by Kleidukos at 2020-03-19T12:38:31-04:00 Force ghc-8.8.3 - - - - - c6fcd0aa by Kleidukos at 2020-03-19T12:38:31-04:00 Update test fixtures - - - - - 5c849cb1 by Sylvain Henry at 2020-03-20T09:34:39+01:00 Modules: Types - - - - - 7f439155 by Alec Theriault at 2020-03-20T20:17:01-04:00 Merge branch 'ghc-8.8' into ghc-8.10 - - - - - b7904e5c by Alina Banerjee at 2020-03-20T20:24:17-04:00 Update parsing to strip whitespace from table cells (#1074) * Update parsing to strip leading & trailing whitespace from table cells * Update fixture data to disallow whitespaces at both ends in table cells * Add test case for whitespaces stripped from both ends of table cells * Update table reference test data for html tests - - - - - b9d60a59 by Alec Theriault at 2020-03-22T11:46:42-04:00 Clean up warnings * unused imports * imports of `Data.List` without import lists * missing `CompatPrelude` file in `.cabal` - - - - - 0c317dbe by Alec Theriault at 2020-03-22T18:46:54-04:00 Fix NPM security warnings This was done by calling `npm audit fix`. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. - - - - - 6e306242 by Alec Theriault at 2020-03-22T20:10:52-04:00 Tentative 2.24 release Adjusted changelogs and versions in `.cabal` files in preparation for the upcoming release bundled with GHC 8.10. - - - - - 1bfb4645 by Ben Gamari at 2020-03-23T16:40:54-04:00 Merge commit '3c2944c037263b426c4fe60a3424c27b852ea71c' into HEAD More changes from the GHC types module refactoring. - - - - - be8c6f3d by Alec Theriault at 2020-03-26T20:10:53-04:00 Update `.travis.yml` to work with GHC 8.10.1 * Regenerated the Travis file with `haskell-ci` * Beef up `.cabal` files with more `tested-with` information - - - - - b025a9c6 by Alec Theriault at 2020-03-26T20:10:53-04:00 Update README Removed some out of date links/info, added some more useful links. * badge to Hackage * update old trac link * `ghc-head` => `ghc-8.10` * `cabal new-*` is now `cabal v2-*` and it should Just Work * `--test-option='--accept'` is the way to accept testsuite output - - - - - 564d889a by Alec Theriault at 2020-03-27T20:34:33-04:00 Fix crash in `haddock-library` on unicode space Our quickcheck tests for `haddock-library` stumbled across an edge case input that was causing Haddock to crash: it was a unicode space character. The root cause of the crash is that we were implicitly assuming that if a space character was not " \t\f\v\r", it would have to be "\n". We fix this by instead defining horizontal space as: any space character that is not '\n'. Fixes haskell/haddock#1142 - - - - - 2d360ba1 by Alec Theriault at 2020-03-27T21:57:32-04:00 Disallow qualified uses of reserved identifiers This a GHC bug (https://gitlab.haskell.org/ghc/ghc/issues/14109) too, but it is a relatively easy fix in Haddock. Note that the fix must live in `haddock-api` instead of `haddock-library` because we can only really decide if an identifier is a reserved one by asking the GHC lexer. Fixes haskell/haddock#952 - - - - - 47ae22ed by Alec Theriault at 2020-03-28T13:36:25-04:00 Remove unused `Haddock.Utils` functions * removed functions in `Haddock.Utils` that were not used anywhere (or exported from the `haddock-api` package) * moved GHC-specific utils from `Haddock.Utils` to `Haddock.GhcUtils` - - - - - c0291245 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use TTG empty extensions to remove some `error`'s None of these error cases should ever have been reachable, so this is just a matter of leveraging the type system to assert this. * Use the `NoExtCon` and `noExtCon` to handle case matches for no extension constructors, instead of throwing an `error`. * Use the extension field of `HsSpliceTy` to ensure that this variant of `HsType` cannot exist in an `HsType DocNameI`. - - - - - 0aff8dc4 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use `unLoc`/`noLoc` from GHC instead of `unL`/`reL` * `unL` is already defined by GHC as `unLoc` * `reL` is already defined by GHC as `noLoc` (in a safer way too!) * Condense `setOutputDir` and add a about exporting from GHC Fixes haskell/haddock#978 - - - - - bf6f2fb7 by Alec Theriault at 2020-03-28T13:36:25-04:00 Cleanup up GHC flags in `.cabal` files * enable more useful warning flags in `haddock-api`, handle the new warnings generated * remove `-fwarn-tabs` (now we'd use `-Wtabs`, but this has been in `-Wall` for a while now) - - - - - c576fbf1 by Alec Theriault at 2020-03-28T13:36:25-04:00 `haddock-library` document header level Document the fact the header level is going to always be between 1 and 6 inclusive. Along the way, I also optimized the parsing code a bit. - - - - - 71bce0ee by Alec Theriault at 2020-03-28T14:26:27-04:00 Disallow links in section headers This is quite straightforward to implement, since we already had a function `docToHtmlNoAnchors` (which we used to generate the link in the sidebar "Contents"). This breaks test `Bug387`, but that test case has aged badly: we now automatically generate anchors for all headings, so manually adding an anchor in a section makes no sense. Nested anchors are, as pointed out in haskell/haddock#1054, disallowed by the HTML standard. Fixes haskell/haddock#1054 - - - - - b461b0ed by Sylvain Henry at 2020-03-30T10:34:23+02:00 Modules: type checker - - - - - cd8cd1ee by Ben Gamari at 2020-03-31T12:45:02-04:00 Merge pull request haskell/haddock#1152 from hsyl20/wip/hsyl20/modules/tc Module renaming - - - - - 5e8f8ea7 by Felix Yan at 2020-04-01T17:58:06-07:00 Allow QuickCheck 2.14 Builds fine and all tests pass. - - - - - dc6b1633 by Sylvain Henry at 2020-04-05T16:43:44+02:00 Module renaming: amend previous patch - - - - - eee2f4ae by Ryan Scott at 2020-04-05T09:04:43-07:00 Fix haskell/haddock#1050 by filtering out invisible AppTy arguments This makes the `synifyType` case for `AppTy` more intelligent by taking into consideration the visibilities of each `AppTy` argument and filtering out any invisible arguments, as they aren't intended to be displayed in the source code. (See haskell/haddock#1050 for an example of what can happen if you fail to filter these out.) Along the way, I noticed that a special `synifyType` case for `AppTy t1 (CoercionTy {})` could be consolidated with the case below it, so I took the opportunity to tidy this up. - - - - - 23eb99e8 by Ben Gamari at 2020-04-07T11:19:58-04:00 Merge pull request haskell/haddock#1154 from hsyl20/wip/hsyl20/modules/tc Module renaming: amend previous patch - - - - - 072d994d by Ryan Scott at 2020-04-07T19:32:47-04:00 Make NoExtCon fields strict These changes are a part of a fix for [GHC#17992](https://gitlab.haskell.org/ghc/ghc/issues/17992). - - - - - d8ebf6c8 by Ignat Insarov at 2020-04-09T21:15:01-04:00 Recode Doc to Json. (#1159) * Recode Doc to Json. * More descriptive field labels. - - - - - 52df4b4e by Sylvain Henry at 2020-04-10T12:39:18+02:00 Module renaming - - - - - d9ab8ec8 by Cale Gibbard at 2020-04-14T11:43:34-04:00 Add instance of XCollectPat for DocNameI - - - - - 323d221d by Cale Gibbard at 2020-04-14T11:43:34-04:00 Rename XCollectPat -> CollectPass - - - - - 2df80867 by Alec Theriault at 2020-04-15T07:30:51-07:00 Prune docstrings that are never rendered When first creating a Haddock interface, trim `ifaceDocMap` and `ifaceArgMap` to not include docstrings that can never appear in the final output. Besides checking with GHC which names are exported, we also need to keep all the docs attached to instance declarations (it is much tougher to detect when an instance is fully private). This change means: * slightly smaller interface files (7% reduction on boot libs) * slightly less work to do processing docstrings that aren't used * no warnings in Haddock's output about private docstrings (see haskell/haddock#1070) I've tested manually that this does not affect any of the boot library generated docs (the only change in output was some small re-ordering in a handful of instance lists). This should mean no docstrings have been incorrectly dropped. - - - - - f49c90cc by Alec Theriault at 2020-04-15T07:30:51-07:00 Don't warn about missing links in miminal sigs When renaming the Haddock interface, never emit warnings when renaming a minimal signature. Also added some documention around `renameInterface`. Minimal signatures intentionally include references to potentially un-exported methods (see the discussion in haskell/haddock#330), so it is expected that they will not always have a link destination. On the principle that warnings should always be resolvable, this shouldn't produce a warning. See haskell/haddock#1070. - - - - - a9eda64d by Ben Gamari at 2020-04-17T09:27:35-04:00 Merge pull request haskell/haddock#1160 from hsyl20/wip/hsyl20/modules/systools Module renaming - - - - - f40d7879 by Cale Gibbard at 2020-04-20T11:30:38-04:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ttg-con-pat - - - - - a50e7753 by Ben Gamari at 2020-04-20T11:36:10-04:00 Merge pull request haskell/haddock#1165 from obsidiansystems/wip/ttg-con-pat Trees that Grow refactor (GHC !2553) - - - - - 6a24795c by Alec Theriault at 2020-04-21T08:06:45-07:00 Fallback to `hiDecl` when `extractDecl` fails Sometimes, the declaration being exported is a subdecl (for instance, a record accessor getting exported at the top-level). For these cases, Haddock has to find a way to produce some synthetic sensible top-level declaration. This is done with `extractDecl`. As is shown by haskell/haddock#1067, this is sometimes impossible to do just at a syntactic level (for instance when the subdecl is re-exported). In these cases, the only sensible thing to do is to try to reify a declaration based on a GHC `TyThing` via `hiDecl`. - - - - - eee1a8b7 by Sylvain Henry at 2020-04-24T15:46:05+02:00 Module structure - - - - - 50b9259c by Iñaki at 2020-04-25T18:38:11-04:00 Add support for custom section anchors (#1179) This allows to have stable anchors for groups, even if the set of groups in the documentation is altered. The syntax for setting the anchor of a group is -- * Group name #desiredAnchor# Which will produce an html anchor of the form '#g:desiredAnchor' Co-authored-by: Iñaki García Etxebarria <git at inaki.blueleaf.cc> - - - - - 4003c97a by Ben Gamari at 2020-04-26T09:35:15-04:00 Merge pull request haskell/haddock#1166 from hsyl20/wip/hsyl20/modules/utils Module structure - - - - - 5206ab60 by Sylvain Henry at 2020-04-27T16:47:39+02:00 Renamed UnitInfo fields - - - - - c32c333b by Sylvain Henry at 2020-04-27T17:32:58+02:00 UnitId has been renamed into Unit - - - - - 3e87db64 by Sylvain Henry at 2020-04-27T17:36:00+02:00 Fix for GHC.Unit.* modules - - - - - ae3323a7 by Ben Gamari at 2020-04-29T12:36:37-04:00 Merge pull request haskell/haddock#1183 from hsyl20/wip/hsyl20/unitid Refactoring of Unit code - - - - - b105564a by Artem Pelenitsyn at 2020-05-03T08:14:10+01:00 add dependency on exceptions because GHC.Exception was boiled down (ghc haskell/haddock#18075) - - - - - 9857eff3 by Zubin Duggal at 2020-05-04T18:48:25+01:00 Atomic update of NameCache in readHieFile - - - - - 86bbb226 by Sylvain Henry at 2020-05-14T16:36:27+02:00 Fix after Config module renaming - - - - - a4bbdbc2 by Gert-Jan Bottu at 2020-05-15T22:09:44+02:00 Explicit Specificity Support for Haddock - - - - - 46199daf by Ben Gamari at 2020-05-19T09:59:56-04:00 Merge pull request haskell/haddock#1192 from hsyl20/hsyl20/modules-config Fix after Config module renaming - - - - - f9a9d2ba by Gert-Jan Bottu at 2020-05-20T16:48:38-04:00 Explicit Specificity Support for Haddock - - - - - 55c5b7ea by Ben Gamari at 2020-05-21T00:32:02-04:00 Merge commit 'a8d7e66da4dcc3b242103271875261604be42d6e' into ghc-head - - - - - a566557f by Cale Gibbard at 2020-05-21T16:02:06-04:00 isBootSummary now produces a result of type IsBootInterface - - - - - ea52f905 by Zubin Duggal at 2020-05-24T17:55:48+01:00 update for hiefile-typeclass-info - - - - - 49ba7a67 by Willem Van Onsem at 2020-05-25T12:23:01-04:00 Use floor over round to calculate the percentage (#1195) If we compile documentation where only a small fraction is undocumented, it is misleading to see 100% coverage - 99% is more intuitive. Fixes haskell/haddock#1194 - - - - - c025ebf1 by Ben Gamari at 2020-05-29T14:32:42-04:00 Merge pull request haskell/haddock#1185 from obsidiansystems/boot-disambig isBootSummary now produces a result of type IsBootInterface - - - - - 74ab9415 by Ben Gamari at 2020-05-29T20:23:39-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - b40be944 by Ben Gamari at 2020-06-03T17:02:31-04:00 testsuite: Update expected output for simplified subsumption - - - - - 624be71c by Ryan Scott at 2020-06-05T12:43:23-04:00 Changes for GHC#18191 See https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3337. - - - - - fbd8f7ce by Sylvain Henry at 2020-06-08T15:31:47+02:00 Fix after unit refactoring - - - - - 743fda4d by Ben Gamari at 2020-06-09T12:09:58-04:00 Merge pull request haskell/haddock#1202 from hsyl20/wip/hsyl20/unitid-ii Fix after unit refactoring - - - - - d07a06a9 by Ryan Scott at 2020-06-13T07:16:55-04:00 Use HsForAllTelescope (GHC#18235) - - - - - 389bb60d by Ben Gamari at 2020-06-13T15:30:52-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - 7a377f5f by Ben Gamari at 2020-06-17T14:53:16-04:00 Merge pull request haskell/haddock#1199 from bgamari/wip/ghc-8.12 haddock: Bounds bumps for GHC 8.12 - - - - - 9fd9e586 by Krzysztof Gogolewski at 2020-06-17T16:09:07-04:00 Adapt Haddock to LinearTypes See ghc/ghc!852. - - - - - 46fe7636 by Ben Gamari at 2020-06-18T14:20:02-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 35a3c9e2 by Zubin Duggal at 2020-06-21T21:19:18+05:30 Use functions exported from HsToCore - - - - - 8abe3928 by Ben Gamari at 2020-06-24T13:53:39-04:00 Merge pull request haskell/haddock#1204 from wz1000/wip/haddock-hstocore Use functions exported from GHC.HsToCore.Docs - - - - - 22f2c937 by Matthías Páll Gissurarson at 2020-06-26T19:07:03+02:00 Adapt Haddock for QualifiedDo - - - - - 3f6208d7 by Vladislav Zavialov at 2020-06-28T14:28:16+03:00 Handle LexicalNegation's ITprefixminus - - - - - 03a19f41 by Sylvain Henry at 2020-07-02T09:37:38+02:00 Rename hsctarget into backend - - - - - ea17ff23 by Andreas Klebinger at 2020-07-02T17:44:18+02:00 Update for UniqFM changes. - - - - - 9872f2f3 by Ben Gamari at 2020-07-09T10:39:19-04:00 Merge pull request haskell/haddock#1209 from AndreasPK/wip/typed_uniqfm Update for UniqFM changes. - - - - - 68f7b668 by Krzysztof Gogolewski at 2020-07-12T18:16:57+02:00 Sync with GHC removing {-# CORE #-} pragma See ghc ticket haskell/haddock#18048 - - - - - eb372681 by Sylvain Henry at 2020-07-20T11:41:30+02:00 Rename hscTarget into backend - - - - - fb7f78bf by Ben Gamari at 2020-07-21T12:15:25-04:00 Merge pull request haskell/haddock#1214 from hsyl20/wip/hsyl20/hadrian/ncg Rename hscTarget into backend - - - - - 1e8f5b56 by Ben Gamari at 2020-07-23T09:11:50-04:00 Merge commit '904dce0cafe0a241dd3ef355775db47fc12f434d' into ghc-head - - - - - d8fd1775 by Zubin Duggal at 2020-07-23T18:46:40+05:30 Update for modular ping pong - - - - - 8416f872 by Ben Gamari at 2020-07-23T09:35:03-04:00 Merge pull request haskell/haddock#1200 from wz1000/wip/wz1000-modular-ping-pong Modular ping pong - - - - - a24a8577 by Ben Gamari at 2020-07-28T15:23:36-04:00 Bump GHC version to 9.0 - - - - - 6a51c9dd by Sylvain Henry at 2020-08-05T18:47:05+02:00 Fix after Outputable refactoring - - - - - c05e1c99 by Ben Gamari at 2020-08-10T14:41:41-04:00 Merge pull request haskell/haddock#1223 from hsyl20/wip/hsyl20/dynflags/exception Fix after Outputable refactoring - - - - - d964f15b by Sylvain Henry at 2020-08-12T11:58:49+02:00 Fix after HomeUnit - - - - - 8e6d5b23 by Ben Gamari at 2020-08-12T14:25:30-04:00 Merge pull request haskell/haddock#1225 from hsyl20/wip/hsyl20/plugins/homeunit Fix after HomeUnit - - - - - 8c7880fe by Sylvain Henry at 2020-08-17T14:13:29+02:00 Remove Ord FastString instance - - - - - 8ea410db by Alex Biehl at 2020-08-19T10:56:32+02:00 Another round of `npm audit fix` (#1228) This should shut down the warnings on Github. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. Last NPM dependency audit happend in d576b2327e2bc117f912fe0a9d595e9ae62614e0 Co-authored-by: Alex Biehl <alex.biehl at target.com> - - - - - 7af6e2a8 by Ben Gamari at 2020-08-31T13:59:34-04:00 Merge pull request haskell/haddock#1226 from hsyl20/wip/hsyl20/fs_ord Remove Ord FastString instance - - - - - ffbc8702 by Alan Zimmerman at 2020-09-07T21:47:41+01:00 Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - a93f1268 by Alan Zimmerman at 2020-09-07T23:11:38+01:00 Merge pull request haskell/haddock#1232 from haskell/wip/T18639-remove-generated-pragma, Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - 1f605d50 by Ben Gamari at 2020-09-14T18:30:01-04:00 Bump GHC version to 9.1 - - - - - 6599df62 by Vladislav Zavialov at 2020-09-18T14:05:15+03:00 Bump base upper bound to 4.16 - - - - - a01b3c43 by Ben Gamari at 2020-09-22T15:41:48-04:00 Update hypsrc-test for QuickLook This appears to be a spurious change. - - - - - e9cc6cac by Vladislav Zavialov at 2020-09-26T21:00:12+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 30e3ca7c by Sylvain Henry at 2020-09-29T11:18:32-04:00 Update for parser (#1234) - - - - - b172f3e3 by Vladislav Zavialov at 2020-09-30T01:01:30+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 0b9c08d3 by Sylvain Henry at 2020-09-30T11:02:33+02:00 Adapt to GHC parser changes - - - - - b9540b7a by Sylvain Henry at 2020-10-12T09:13:38-04:00 Don't pass the HomeUnitId (#1239) - - - - - 34762e80 by HaskellMouse at 2020-10-13T12:58:04+03:00 Changed tests due to unification of `Nat` and `Natural` in the follwing merge request: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3583 - - - - - 256f86b6 by Vladislav Zavialov at 2020-10-15T10:48:03+03:00 Add whitespace in: map ($ v) - - - - - 4a3f711b by Alan Zimmerman at 2020-10-19T08:57:27+01:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled - - - - - 072cdd21 by Alan Zimmerman at 2020-10-21T14:48:28-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 9e09a445 by Alan Zimmerman at 2020-10-21T23:53:34-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 636d7de3 by Sylvain Henry at 2020-10-26T14:31:54-04:00 GHC.Driver.Types refactoring (#1242) - - - - - a597f000 by Ryan Scott at 2020-10-29T04:18:05-04:00 Adapt to the removal of Hs{Boxed,Constraint}Tuple See ghc/ghc!4097 and GHC#18723. - - - - - b96660fb by Ryan Scott at 2020-10-30T04:53:05-04:00 Adapt to HsConDecl{H98,GADT}Details split Needed for GHC#18844. - - - - - c287d82c by Ryan Scott at 2020-10-30T19:35:59-04:00 Adapt to HsOuterTyVarBndrs These changes accompany ghc/ghc!4107, which aims to be a fix for haskell/haddock#16762. - - - - - a34c31a1 by Ryan Scott at 2020-11-13T13:38:34-05:00 Adapt to splitPiTysInvisible being renamed to splitInvisPiTys This is a part of !4434, a fix for GHC#18939. - - - - - 66ea459d by Sylvain Henry at 2020-11-16T10:59:30+01:00 Fix after Plugins moved into HscEnv - - - - - 508556d8 by Ben Gamari at 2020-11-18T15:47:40-05:00 Merge pull request haskell/haddock#1253 from hsyl20/wip/hsyl20/plugins/hscenv Fix after Plugins moved into HscEnv - - - - - 620fec1a by Andreas Klebinger at 2020-11-24T20:51:59+01:00 Update for changes in GHC's Pretty - - - - - 01cc13ab by Richard Eisenberg at 2020-11-25T23:18:35-05:00 Avoid GHC#18932. - - - - - 8d29ba21 by Cale Gibbard at 2020-11-25T23:18:35-05:00 Add type arguments to PrefixCon - - - - - 414d5f87 by Sylvain Henry at 2020-11-30T17:06:04+01:00 DynFlags's unit fields moved to HscEnv - - - - - e356668c by Ben Gamari at 2020-11-30T11:11:37-05:00 Merge pull request haskell/haddock#1258 from hsyl20/wip/hsyl20/hscenv/unitstate Unit fields moved from DynFlags to HscEnv - - - - - 7cf552f1 by Ben Gamari at 2020-12-03T10:31:27-05:00 Merge pull request haskell/haddock#1257 from AndreasPK/wip/andreask/opt_dumps Update for changes in GHC's Pretty - - - - - fc0871c3 by Veronika Romashkina at 2020-12-08T16:35:33+01:00 Fix docs links from Darcs to GitHub in intro (#1262) - - - - - 7059e808 by Veronika Romashkina at 2020-12-08T16:36:16+01:00 Use gender neutral word in docs (#1260) - - - - - 1b16e5ee by Maximilian Tagher at 2020-12-08T16:40:03+01:00 Allow scrolling search results (#1235) Closes https://github.com/haskell/haddock/issues/1231 - - - - - 8a118c01 by dependabot[bot] at 2020-12-08T16:40:25+01:00 Bump bl from 1.2.2 to 1.2.3 in /haddock-api/resources/html (#1255) Bumps [bl](https://github.com/rvagg/bl) from 1.2.2 to 1.2.3. - [Release notes](https://github.com/rvagg/bl/releases) - [Commits](https://github.com/rvagg/bl/compare/v1.2.2...v1.2.3) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - c89ff587 by Xia Li-yao at 2020-12-08T16:42:17+01:00 Allow more characters in anchor following module reference (#1220) - - - - - 14af7d64 by Xia Li-yao at 2020-12-08T16:43:05+01:00 Add dangling changes from branches ghc-8.6 and ghc-8.8 (#1243) * Fix multiple typos and inconsistencies in doc/markup.rst Note: I noticed some overlap with haskell/haddock#1112 from @wygulmage and haskell/haddock#1081 from @parsonsmatt after creating these proposed changes - mea culpa for not looking at the open PRs sooner. * Fix haskell/haddock#1113 If no Signatures, no section of index.html * Change the formatting of missing link destinations The current formatting of the missing link destination does not really help user to understand the reasons of the missing link. To address this, I've changed the formatting in two ways: - the missing link symbol name is now fully qualified. This way you immediately know which haskell module cannot be linked. It is then easier to understand why this module does not have documentation (hidden module or broken documentation). - one line per missing link, that's more readable now that symbol name can be longer due to qualification. For example, before haddock was listing missing symbol such as: ``` could not find link destinations for: Word8 Word16 mapMaybe ``` Now it is listed as: ``` could not find link destinations for: - Data.Word.Word8 - Data.Word.Word16 - Data.Maybe.mapMaybe ``` * Add `--ignore-link-symbol` command line argument This argument can be used multiples time. A missing link to a symbol listed by `--ignore-link-symbol` won't trigger "missing link" warning. * Forbid spaces in anchors (#1148) * Improve error messages with context information (#1060) Co-authored-by: Matt Audesse <matt at mattaudesse.com> Co-authored-by: Mike Pilgrem <mpilgrem at users.noreply.github.com> Co-authored-by: Guillaume Bouchard <guillaume.bouchard at tweag.io> Co-authored-by: Pepe Iborra <pepeiborra at gmail.com> - - - - - 89e3af13 by tomjaguarpaw at 2020-12-08T18:00:04+01:00 Enable two warnings (#1245) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - c3320f8d by Willem Van Onsem at 2020-12-08T18:26:55+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 685df308 by Alex Biehl at 2020-12-08T20:06:26+01:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - be3ec3c0 by Alex Biehl at 2020-12-08T20:06:26+01:00 Import intercalate - - - - - 32c33912 by Matthías Páll Gissurarson at 2020-12-08T21:15:30+01:00 Adapt Haddock for QualifiedDo - - - - - 31696088 by Alex Biehl at 2020-12-08T22:06:02+01:00 Fix haddock-library tests - - - - - fbc0998a by Alex Biehl at 2020-12-08T23:08:23+01:00 Move to GitHub CI (#1266) * Initial version of ci.yml This is a straight copy from Dmitrii Kovanikov's blog post at https://kodimensional.dev/github-actions. Will adapt to haddock in successive commits. * Delete .travis.yml * Modify to only test on ghc-8.10.{1,2} * Use actions/setup-haskell at v1.1.4 * Relax QuickCheck bound on haddock-api * Remove stack matrix for now * Nail down to ghc-8.10 branch for now * Pin index state to 2020-12-08T20:13:44Z for now * Disable macOS and Windows tests for now for speed up - - - - - 5b946b9a by tomjaguarpaw at 2020-12-10T19:01:41+01:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - bc5a408f by dependabot[bot] at 2020-12-10T19:02:16+01:00 Bump ini from 1.3.5 to 1.3.7 in /haddock-api/resources/html (#1269) Bumps [ini](https://github.com/isaacs/ini) from 1.3.5 to 1.3.7. - [Release notes](https://github.com/isaacs/ini/releases) - [Commits](https://github.com/isaacs/ini/compare/v1.3.5...v1.3.7) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - d02995f1 by Andrew Martin at 2020-12-14T16:48:40-05:00 Update for boxed rep - - - - - a381aeff by Ben Gamari at 2020-12-15T15:13:30-05:00 Revert "Enable two warnings (#1245) (#1268)" As this does not build on GHC `master`. This reverts commit 7936692badfe38f23ae95b51fb7bd7c2ff7e9bce. - - - - - a63c0a9e by Ben Gamari at 2020-12-15T15:17:59-05:00 Revert "Update for boxed rep" This reverts commit 4ffb30d8b637ccebecc81ce610f0af451ac8088d. - - - - - 53bfbb29 by Ben Gamari at 2020-12-15T15:37:24-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - bae76a30 by Ben Gamari at 2020-12-16T02:44:42+00:00 Update output for nullary TyConApp optimisation (ghc/ghc!2952) - - - - - 4b733b57 by Krzysztof Gogolewski at 2020-12-16T20:03:14+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. - - - - - ee463bd3 by Ryan Scott at 2020-12-16T16:55:23-05:00 Adapt to HsCoreTy (formerly NewHsTypeX) becoming a type synonym Needed for !4417, the fix for GHC#15706 and GHC#18914. - - - - - ed0b02f8 by tomjaguarpaw at 2020-12-19T10:17:19+00:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - d80bf8f5 by Sylvain Henry at 2020-12-21T10:09:25+01:00 Fix after binder collect changes - - - - - bf4c9d32 by Adam Gundry at 2020-12-23T21:35:01+00:00 Adapt to changes to GlobalRdrElt and AvailInfo Needed for ghc/ghc!4467 - - - - - 37736c4c by John Ericson at 2020-12-28T12:27:02-05:00 Support a new ghc --make node type for parallel backpack upsweep - - - - - 717bdeac by Vladislav Zavialov at 2020-12-29T10:50:02+03:00 Inline and fix getGADTConTypeG The getGADTConTypeG used HsRecTy, which is at odds with GHC issue haskell/haddock#18782. I noticed that getGADTConTypeG was only used in the Hoogle backend. Interestingly, when handling H98 constructors, Hoogle converts RecCon to PrefixCon (see Haddock.Backends.Hoogle.ppCtor). So I changed getGADTConTypeG to handle RecConGADT in the same manner as PrefixConGADT, and after this simplification moved it into the 'where' clause of ppCtor, to the only place where it is used. The practical effect of this change is as follows. Consider this example: data TestH98 = T98 { bar::Int } data TestGADT where TG :: { foo :: Int } -> TestGADT Before this patch, haddock --hoogle used to produce: T98 :: Int -> TestH98 [TG] :: {foo :: Int} -> TestGADT Notice how the record syntax was discarded in T98 but not TG. With this patch, we always produce signatures without record syntax: T98 :: Int -> TestH98 [TG] :: Int -> TestGADT I suspect this might also be a bugfix, as currently Hoogle doesn't seem to render GADT record constructors properly. - - - - - cb1b8c56 by Andreas Abel at 2020-12-30T21:12:37+01:00 Build instructions: haddock-library and -api first! - - - - - b947f6ad by Ben Gamari at 2020-12-31T13:04:19-05:00 Merge pull request haskell/haddock#1281 from obsidiansystems/wip/backpack-j Changes to support -j with backpack - - - - - 120e1cfd by Hécate Moonlight at 2021-01-04T19:54:58+01:00 Merge pull request haskell/haddock#1282 from andreasabel/master Build instructions: haddock-library and -api first! - - - - - fd45e41a by Ben Gamari at 2021-01-05T16:14:31-05:00 Merge remote-tracking branch 'origin/ghc-8.10' into ghc-9.0 - - - - - b471bdec by Ben Gamari at 2021-01-05T16:23:02-05:00 Merge commit '1e56f63c3197e7ca1c1e506e083c2bad25d08793' into ghc-9.0 - - - - - 81cdbc41 by Alex Biehl at 2021-01-09T12:14:41+01:00 Prepare Haddock for being a GHC Plugin - - - - - b646d952 by Alex Biehl at 2021-01-09T12:14:41+01:00 Make Haddock a GHC Plugin - - - - - cc044674 by Alex Biehl at 2021-01-09T12:14:41+01:00 Add -j[n] CLI param to Haddock executable It translates to `--ghcopt=-j[n]` - - - - - 84a04073 by Alex Biehl at 2021-01-09T12:14:41+01:00 Abstract Monad for interface creation I found that when running as a plugin the lookupName function (which runs in Ghc monad) does not work correctly from the typeCheckResultAction hook. Instead, we abstracted the monad used when creating interfaces, so that access to GHC session specific parts is explicit and so that the TcM can provide their (correct) implementation of lookupName. - - - - - 5be2c4f7 by Alex Biehl at 2021-01-09T12:14:41+01:00 Accept tests - - - - - 8cefee9d by Alex Biehl at 2021-01-09T16:10:47+01:00 Add missing dependency for mtl - - - - - 3681f919 by Ben Gamari at 2021-01-13T18:39:25-05:00 Merge remote-tracking branch 'origin/ghc-9.0' into ghc-head - - - - - 33c6b152 by Hécate Moonlight at 2021-01-14T16:04:20+01:00 Merge pull request haskell/haddock#1273 from hsyl20/wip/hsyl20/arrows Fix after binder collect changes - - - - - 70d13e8e by Joachim Breitner at 2021-01-22T19:03:45+01:00 Make haddock more robust to changes to the `Language` data type With the introduction of GHC2021, the `Languages` data type in GHC will grow. In preparation of that (and to avoid changing haddock with each new language), this change makes the code handle extensions to that data type gracefully. (cherry picked from commit c341dd7c9c3fc5ebc83a2d577c5a726f3eb152a5) - - - - - 7d6dd57a by John Ericson at 2021-01-22T22:02:02+00:00 Add `NoGhcTc` instance now that it's not closed - - - - - e5fdaf0a by Alan Zimmerman at 2021-01-23T22:57:44+00:00 Merge pull request haskell/haddock#1293 from obsidiansystems/wip/fix-18936 Add `NoGhcTc` instance now that it's not closed - - - - - 989a1e05 by Oleg Grenrus at 2021-01-24T16:11:46+03:00 Add import list to Data.List - - - - - 368e144a by Ben Gamari at 2021-01-28T22:15:48+01:00 Adapt to "Make PatSyn immutable" - - - - - abe66c21 by Alfredo Di Napoli at 2021-02-01T08:05:35+01:00 Rename pprLogErrMsg to new name - - - - - e600e75c by Hécate Moonlight at 2021-02-05T14:53:00+01:00 Move CI to ghc-9.0 - - - - - dd492961 by Vladislav Zavialov at 2021-02-05T14:53:00+01:00 Update cabal.project and README build instructions - - - - - 31bd292a by Hécate Moonlight at 2021-02-05T15:03:56+01:00 Merge pull request haskell/haddock#1296 from Kleidukos/ghc-9.0 Merge the late additions to ghc-8.10 into ghc-9.0 - - - - - 6388989e by Vladislav Zavialov at 2021-02-05T17:41:57+03:00 Cleanup: fix build warnings - - - - - f99407ef by Daniel Rogozin at 2021-02-05T18:11:48+03:00 type level characters support for haddock (required for haskell/haddock#11342) - - - - - d8c6b26f by Hécate Moonlight at 2021-02-05T17:44:50+01:00 Add a CONTRIBUTING.md file - - - - - 6a01ad98 by Hécate Moonlight at 2021-02-05T17:58:16+01:00 Merge pull request haskell/haddock#1312 from Kleidukos/proper-branch-etiquette Add a CONTRIBUTING.md file - - - - - 955eecc4 by Vladislav Zavialov at 2021-02-05T20:29:00+03:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into ghc-head - - - - - 47b3d6ab by Hécate Moonlight at 2021-02-05T19:09:38+01:00 Amend the CONTRIBUTING.md file - - - - - 23de6137 by Hécate Moonlight at 2021-02-05T19:16:49+01:00 Merge pull request haskell/haddock#1313 from Kleidukos/amend-contributing Amend the CONTRIBUTING.md file - - - - - 69026b59 by Krzysztof Gogolewski at 2021-02-05T23:05:56+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. (cherry picked from commit b4b4d896d2d68d6c48e7db7bfe95c185ca0709cb) - - - - - ea026b78 by Oleg Grenrus at 2021-02-06T17:14:45+01:00 Add import list to Data.List - - - - - 5204326f by Hécate Moonlight at 2021-02-06T17:15:44+01:00 Merge pull request haskell/haddock#1316 from Kleidukos/explicit-imports-to-data-list Add import list to Data.List - - - - - 1f4d2136 by Ben Gamari at 2021-02-06T11:53:31-05:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ghc-head-merge - - - - - 13f0d09a by Ben Gamari at 2021-02-06T11:53:45-05:00 Fix partial record selector warning - - - - - 5c115f7e by Ben Gamari at 2021-02-06T11:55:52-05:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into wip/ghc-head-merge - - - - - b6fd8b75 by Ben Gamari at 2021-02-06T12:01:31-05:00 Merge commit '41964cb2fd54b5a10f8c0f28147015b7d5ad2c02' into wip/ghc-head-merge - - - - - a967194c by Ben Gamari at 2021-02-06T18:30:35-05:00 Merge branch 'wip/ghc-head-merge' into ghc-head - - - - - 1f4c3a91 by MorrowM at 2021-02-07T01:52:33+02:00 Fix search div not scrolling - - - - - 684b1287 by Iñaki García Etxebarria at 2021-02-07T16:13:04+01:00 Add support for labeled module references Support a markdown-style way of annotating module references. For instance -- | [label]("Module.Name#anchor") will create a link that points to the same place as the module reference "Module.Name#anchor" but the text displayed on the link will be "label". - - - - - bdb55a5d by Hécate Moonlight at 2021-02-07T16:18:10+01:00 Merge pull request haskell/haddock#1319 from alexbiehl/alex/compat Backward compat: Add support for labeled module references - - - - - 6ca70991 by Hécate Moonlight at 2021-02-07T16:21:29+01:00 Merge pull request haskell/haddock#1314 from tweag/show-linear-backport Backport haskell/haddock#1238 (linear types) to ghc-9.0 - - - - - d9d73298 by Alex Biehl at 2021-02-07T17:46:25+01:00 Remove dubious parseModLink Instead construct the ModLink value directly when parsing. - - - - - 33b4d020 by Hécate Moonlight at 2021-02-07T17:52:05+01:00 Merge pull request haskell/haddock#1320 from haskell/alex/fix Remove dubious parseModLink - - - - - 54211316 by Hécate Moonlight at 2021-02-07T18:12:07+01:00 Merge pull request haskell/haddock#1318 from MorrowM/ghc-9.0 Fix search div not scrolling - - - - - 19db679e by alexbiehl-gc at 2021-02-07T18:14:46+01:00 Merge pull request haskell/haddock#1317 from bgamari/wip/ghc-head-merge Merge ghc-8.10 into ghc-head - - - - - 6bc1e9e4 by Willem Van Onsem at 2021-02-07T18:25:30+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - c8537cf8 by alexbiehl-gc at 2021-02-07T18:30:40+01:00 Merge pull request haskell/haddock#1322 from haskell/alex/forward-port simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 2d47ae4e by alexbiehl-gc at 2021-02-07T18:39:59+01:00 Merge branch 'ghc-head' into ghc-9.0 - - - - - 849e4733 by Hécate Moonlight at 2021-02-07T18:43:19+01:00 Merge pull request haskell/haddock#1321 from Kleidukos/ghc-9.0 Merge ghc-9.0 into ghc-head - - - - - ee6095d7 by Sylvain Henry at 2021-02-08T11:36:38+01:00 Update for Logger - - - - - 4ad688c9 by Alex Biehl at 2021-02-08T18:11:24+01:00 Merge pull request haskell/haddock#1310 from hsyl20/wip/hsyl20/logger2 Logger refactoring - - - - - 922a9e0e by Ben Gamari at 2021-02-08T12:54:33-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - 991649d2 by Sylvain Henry at 2021-02-09T10:55:17+01:00 Fix to build with HEAD - - - - - a8348dc2 by Hécate Moonlight at 2021-02-09T10:58:51+01:00 Merge pull request haskell/haddock#1327 from hsyl20/wip/hsyl20/logger2 Fix to build with HEAD - - - - - 0abdbca6 by Fendor at 2021-02-09T20:06:15+01:00 Add UnitId to Target record - - - - - d5790a0e by Alex Biehl at 2021-02-11T10:32:32+01:00 Stable sort for (data/newtype) instances - - - - - 8e6036f5 by Alex Biehl at 2021-02-11T10:32:32+01:00 Also make TyLit deterministic - - - - - f76d2945 by Hécate Moonlight at 2021-02-11T11:00:31+01:00 Merge pull request haskell/haddock#1329 from hsyl20/hsyl20/stabe_iface Stable sort for instances - - - - - 5e0469ea by Oleg Grenrus at 2021-02-14T15:28:15+02:00 Add import list to Data.List in Haddock.Interface.Create - - - - - fa57cd24 by Hécate Moonlight at 2021-02-14T17:19:27+01:00 Merge pull request haskell/haddock#1331 from phadej/more-explicit-data-list-imports Add import list to Data.List in Haddock.Interface.Create - - - - - f0cd629c by Hécate Moonlight at 2021-02-21T00:22:01+01:00 Merge pull request haskell/haddock#1311 from fendor/wip/add-targetUnitId-to-target Add UnitId to Target record - - - - - 674ef723 by Joachim Breitner at 2021-02-22T10:39:18+01:00 html-test: Always set language from ghc-9.2 on, the “default” langauge of GHC is expected to change more wildly. To prepare for that (and unblock https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4853), this sets the language for all the test files to `Haskell2010`. This should insolate this test suite against changes to the default. Cherry-picked from https://github.com/haskell/haddock/pull/1341 - - - - - f072d623 by Hécate Moonlight at 2021-02-22T10:56:51+01:00 Merge pull request haskell/haddock#1342 from nomeata/joachim/haskell2010-in-tests-ghc-head html-test: Always set language - - - - - caebbfca by Hécate Moonlight at 2021-02-22T11:53:07+01:00 Clean-up of Interface and Interface.Create's imports and pragmata - - - - - f6caa89d by Hécate Moonlight at 2021-02-22T11:54:58+01:00 Merge pull request haskell/haddock#1345 from Kleidukos/head/fix-interface-imports [ghc-head] Clean-up of Interface and Interface.Create's imports and pragmata - - - - - 7395c9cb by Hécate Moonlight at 2021-02-22T18:44:57+01:00 Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 6e9fb5d5 by Hécate Moonlight at 2021-02-22T18:45:28+01:00 Merge pull request haskell/haddock#1348 from Kleidukos/head/explicit-imports-interface Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 9198b118 by Alan Zimmerman at 2021-02-22T20:04:24+00:00 Context becomes a Maybe in the GHC AST This prevents noLoc's appearing in the ParsedSource. Match the change in GHC. - - - - - 0af20f64 by Hécate Moonlight at 2021-02-23T12:36:12+01:00 Fix the call-site of guessTarget in Interface.hs Explicit the imports from GHC.HsToCore.Docs - - - - - b7886885 by Hécate Moonlight at 2021-02-23T12:37:54+01:00 Merge pull request haskell/haddock#1349 from Kleidukos/fix-interface-guesstarget-call Fix the call-site of guessTarget in Interface.hs - - - - - 9cf041ba by Sylvain Henry at 2021-02-24T11:08:20+01:00 Fix haddockHypsrcTest output in ghc-head - - - - - b194182a by Hécate Moonlight at 2021-02-24T11:12:36+01:00 Merge pull request haskell/haddock#1351 from hsyl20/wip/hsyl20/fix-head Fix haddockHypsrcTest output in ghc-head - - - - - 3ce8b375 by Shayne Fletcher at 2021-03-06T09:55:03-05:00 Add ITproj to parser - - - - - d2abf762 by Ben Gamari at 2021-03-06T19:26:49-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a0f6047d by Andrew Martin at 2021-03-07T11:25:23-05:00 Update for boxed rep - - - - - 6f63c99e by Ben Gamari at 2021-03-10T13:20:21-05:00 Update for "FastString: Use FastMutInt instead of IORef Int" - - - - - e13f01df by Luke Lau at 2021-03-10T15:38:40-05:00 Implement template-haskell's putDoc This catches up to GHC using the new extractTHDocs function, which returns documentation added via the putDoc function (provided it was compiled with Opt_Haddock). Since it's already a map from names -> docs, there's no need to do traversal etc. It also matches the change from the argument map being made an IntMap rather than a Map Int - - - - - 89263d94 by Alan Zimmerman at 2021-03-15T17:15:26+00:00 Match changes in GHC AST for in-tree API Annotations As landed via https://gitlab.haskell.org/ghc/ghc/-/merge_requests/2418 - - - - - 28db1934 by Alan Zimmerman at 2021-03-15T20:40:09+00:00 Change some type family test results. It is not clear to me whether the original was printing incorrectly (since we did not have the TopLevel flag before now), or if this behaviour is expected. For the time being I am assuming the former. - - - - - 7c11c989 by Sylvain Henry at 2021-03-22T10:05:19+01:00 Fix after NameCache changes - - - - - addbde15 by Sylvain Henry at 2021-03-22T10:05:19+01:00 NameCache doesn't store a UniqSupply anymore - - - - - 15ec6cec by Ben Gamari at 2021-03-22T17:53:44-04:00 Bump GHC version to 9.2 - - - - - dbd6aa63 by Hécate Moonlight at 2021-03-24T14:28:36+01:00 Merge pull request haskell/haddock#1365 from hsyl20/wip/hsyl20/iface1 NameCache refactoring - - - - - 2d32da7e by Oleg Grenrus at 2021-03-27T01:12:00+02:00 Specialization of Data.List - - - - - 32b84fa6 by Fendor at 2021-03-27T10:50:17+01:00 Add UnitId to Target record This way we always know to which home-unit a given target belongs to. So far, there only exists a single home-unit at a time, but it enables having multiple home-units at the same time. - - - - - 54bf9f0e by Hécate Moonlight at 2021-03-28T14:08:35+02:00 Merge pull request haskell/haddock#1368 from fendor/target-unit-id-revert Add UnitId to Target record - - - - - 7dea168a by Alan Zimmerman at 2021-03-29T08:45:52+01:00 EPA : Rename ApiAnn to EpAnn - - - - - 72967f65 by Alfredo Di Napoli at 2021-03-29T09:47:01+02:00 pprError changed name in GHC - - - - - 4bc61035 by Alan Zimmerman at 2021-03-29T16:16:27-04:00 EPA : Rename ApiAnn to EpAnn - - - - - 108d031d by Ben Gamari at 2021-03-29T18:49:36-04:00 Merge commit '36418c4f70d7d2b179a77925b3ad5caedb08c9b5' into HEAD - - - - - 1444f700 by Ben Gamari at 2021-03-31T09:18:39-04:00 Merge pull request haskell/haddock#1370 from adinapoli/wip/adinapoli-diag-reason-severity Rename pprError to mkParserErr - - - - - d3087b79 by Ben Gamari at 2021-03-31T11:34:17-04:00 Merge commit 'd8d8024ad6796549a8d3b5512dabf3288d14e30f' into ghc-head - - - - - 170b79e9 by Ben Gamari at 2021-03-31T12:24:56-04:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - db0d6bae by Ben Gamari at 2021-04-10T09:34:35-04:00 Bump GHC version to 9.3 - - - - - a9f2c421 by Alan Zimmerman at 2021-04-19T18:26:46-04:00 Update for EPA changes in GHC (cherry picked from commit cafb48118f7c111020663776845897e225607b41) - - - - - 1ee4b7c7 by Sylvain Henry at 2021-05-11T10:00:06+02:00 Removal of HsVersions.h (#1388) * Update for EPA changes in GHC * Account for HsVersions.h removal Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - 79e819e9 by Hécate Moonlight at 2021-05-11T10:14:47+02:00 Revert "Removal of HsVersions.h (#1388)" This reverts commit 72118896464f94d81f10c52f5d9261efcacc57a6. - - - - - 3dbd3f8b by Alan Zimmerman at 2021-05-11T10:15:17+02:00 Update for EPA changes in GHC - - - - - 2ce80c17 by Sylvain Henry at 2021-05-11T10:15:19+02:00 Account for HsVersions.h removal - - - - - 00e4c918 by Christiaan Baaij at 2021-05-13T08:21:56+02:00 Add Haddock support for the OPAQUE pragma (#1380) - - - - - 8f9049b2 by Hécate Moonlight at 2021-05-13T08:40:22+02:00 fixup! Use GHC 9.2 in CI runner - - - - - 27ddec38 by Alan Zimmerman at 2021-05-13T22:51:20+01:00 EPA: match changes from GHC T19834 - - - - - f8a1d714 by Felix Yan at 2021-05-14T17:10:04+02:00 Allow hspec 2.8 (#1389) All tests are passing. - - - - - df44453b by Divam Narula at 2021-05-20T15:42:42+02:00 Update ref, the variables got renamed. (#1391) This is due to ghc/ghc!5555 which caused a change in ModDetails in case of NoBackend. Now the initModDetails is used to recreate the ModDetails from interface and in-memory ModDetails is not used. - - - - - e46bfc87 by Alan Zimmerman at 2021-05-20T19:05:09+01:00 Remove Maybe from HsQualTy Match changes in GHC for haskell/haddock#19845 - - - - - 79bd7b62 by Shayne Fletcher at 2021-05-22T08:20:39+10:00 FieldOcc: rename extFieldOcc to foExt - - - - - 6ed68c74 by Ben Gamari at 2021-05-21T22:29:30-04:00 Merge commit '3b6a8774bdb543dad59b2618458b07feab8a55e9' into ghc-head - - - - - f9a02d34 by Alfredo Di Napoli at 2021-05-24T13:53:00+02:00 New Parser diagnostics interface - - - - - 392807d0 by Ben Gamari at 2021-05-24T09:57:40-04:00 Merge pull request haskell/haddock#1394 from adinapoli/wip/adinapoli-align-ps-messages Align Haddock to use the new Parser diagnostics interface - - - - - 33023cd8 by Ben Gamari at 2021-05-24T11:19:16-04:00 Revert "Add Haddock support for the OPAQUE pragma (#1380)" This reverts commit a1337c599ef7720b0482a25c55f11794112496dc. The GHC patch associated with this change is not yet ready to be merged. - - - - - 8c005af7 by Ben Simms at 2021-05-28T07:56:20+02:00 CI configuration for ghc-head (#1395) - - - - - 1e947612 by Hécate Moonlight at 2021-05-28T12:27:35+02:00 Use GHC 9.2 in CI runner (#1378) - - - - - e6fa10ab by CGenie at 2021-05-31T09:02:13+02:00 Add page about common errors (#1396) * Update index.rst Common errors page * Create common-errors.rst * Update common-errors.rst * Use GHC 9.2 in CI runner (#1378) * [haddock-api] remove .hspec-failures Co-authored-by: Hécate Moonlight <Kleidukos at users.noreply.github.com> - - - - - abc72a8d by Sylvain Henry at 2021-06-01T10:02:06+02:00 Adapt Haddock to Logger and Parser changes (#1399) - - - - - 91373656 by Zubin Duggal at 2021-06-01T20:45:10+02:00 Update haddockHypSrc tests since we now compute slighly more type info (#1397) - - - - - ed712822 by Marcin Szamotulski at 2021-06-02T08:54:33+02:00 Added myself to contributors - - - - - 49fdbcb7 by Marcin Szamotulski at 2021-06-02T08:57:24+02:00 Document multi component support - - - - - 9ddc8d7d by Hécate Moonlight at 2021-06-02T09:35:55+02:00 Merge pull request haskell/haddock#1379 from coot/coot/document-multi-component-support Document multi component support - - - - - 585b5c5e by Ben Simms at 2021-06-02T19:46:54+02:00 Update CONTRIBUTING.md (#1402) - - - - - 1df4a605 by Ben Simms at 2021-06-02T19:47:14+02:00 Update CONTRIBUTING.md (#1403) - - - - - 58ea43d2 by sheaf at 2021-06-02T22:09:06+02:00 Update Haddock Bug873 to account for renaming - - - - - c5d0ab23 by Vladislav Zavialov at 2021-06-10T13:35:42+03:00 HsToken in FunTy, RecConGADT - - - - - 1ae2f40c by Hécate Moonlight at 2021-06-11T11:19:09+02:00 Update the CI badges - - - - - 6fdc4de2 by Sylvain Henry at 2021-06-28T19:21:17+02:00 Fix mkParserOpts (#1411) - - - - - 18201670 by Alfredo Di Napoli at 2021-07-05T07:55:12+02:00 Rename getErrorMessages Lexer import This commit renames the Lexer import in `Hyperlinker.Parser` from `getErrorMessages` to `getPsErrorMessages` to eliminate the ambiguity with the `getErrorMessages` function defined in `GHC.Types.Error`. - - - - - 23173ca3 by Ben Gamari at 2021-07-07T11:31:44-04:00 Merge pull request haskell/haddock#1413 from adinapoli/wip/adinapoli-issue-19920 Rename getErrorMessages Lexer import - - - - - b3dc4ed8 by Alan Zimmerman at 2021-07-28T22:30:59+01:00 EPA: match changes from GHC T19834 (cherry picked from commit 2fec1b44e0ee7e263286709aa528b4ecb99ac6c2) - - - - - 5f177278 by Ben Gamari at 2021-08-06T01:17:37-04:00 Merge commit '2a966c8ca37' into HEAD - - - - - cdd81d08 by Marcin Szamotulski at 2021-08-08T17:19:06+02:00 coot/multiple packages (ghc-9.2) (#1418) - - - - - be0d71f1 by Marcin Szamotulski at 2021-08-16T08:46:03+02:00 coot/multiple package (ghc-head) (#1419) * FromJSON class Aeson style FromJSON class with Parsec based json parser. * doc-index.json file for multiple packages When creating haddock summary page for multiple packages render doc-index.json file using contents of all found 'doc-index.json' files. * Render doc-index.json When rendering html, render doc-index.json file independently of maybe_index_url option. doc-index.json file is useful now even if maybe_index_url is not `Nothing`. * base url option New `Flag_BaseURL` which configures from where static files are loaded (--base-url). If given and not equal "." static files are not coppied, as this indicates that they are not read from the the directory where we'd copy them. The default value is ".". - - - - - 3b09dbdf by Hécate Moonlight at 2021-10-07T23:26:03+02:00 Update GHC 9.2 to latest pre-release in CI - - - - - 7ac55417 by Zubin Duggal at 2021-10-11T12:10:19+02:00 Enable Haddock tests in GHC windows CI (#1428) * testsuite: strip windows line endings for haddock * hyperlinker: Work around double escaping (#19236) * deterministic SCC - - - - - 1cb81f25 by Andrew Lelechenko at 2021-10-12T15:23:19+02:00 haddock-library does not depend on bytestring or transformers (#1426) - - - - - a890b9aa by sheaf at 2021-10-15T22:19:42+02:00 update haddockHypsrcTest for GHC MR !6705 (#1430) - - - - - 42a55c6c by Sylvain Henry at 2021-10-15T22:20:10+02:00 Fix after PkgQual refactoring (#1429) - - - - - 91659238 by Alan Zimmerman at 2021-10-28T18:57:10+01:00 Update for changes in GHC for branch wip/az/no-srcspan-anno-instances - - - - - acf23e60 by Vladislav Zavialov at 2021-11-05T02:09:47+03:00 Do not use forall as an identifier See GHC ticket haskell/haddock#20609 - - - - - c565db0e by Krzysztof Gogolewski at 2021-11-27T02:42:35+01:00 Update after NoExtCon -> DataConCantHappen rename - - - - - b5f55590 by Artem Pelenitsyn at 2021-11-27T11:14:17+01:00 fix CI for 9.2 (#1436) - - - - - 25cd621e by Matthew Pickering at 2021-12-02T11:46:54+00:00 Update html-test for Data.List revert - - - - - 1d5ff85f by malteneuss at 2021-12-15T07:56:55+01:00 Add hint about inline link issue (#1444) - - - - - 791fde81 by Sylvain Henry at 2021-12-16T09:29:51+01:00 Bump ghc-head (#1445) * Update after NoExtCon -> DataConCantHappen rename * Update html-test for Data.List revert * Fix for new Plugins datatype Co-authored-by: Krzysztof Gogolewski <krzysztof.gogolewski at tweag.io> Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 44236317 by Sylvain Henry at 2021-12-17T09:39:00+01:00 Fix for new Plugins datatype - - - - - 80ada0fa by Hécate Moonlight at 2021-12-17T17:28:48+01:00 Remove ghc-head workflow (#1446) Contributions of GHC glue code are now done on the GHC gitlab, not in the GitHub repo anymore. - - - - - 49e171cd by Matthew Pickering at 2021-12-28T09:47:09+00:00 Remove use of ExtendedModSummary - - - - - 0e91b5ea by askeblad at 2022-01-04T09:18:35+01:00 update URLs - - - - - 9f13c212 by Hécate Moonlight at 2022-02-25T10:19:46+01:00 Fix solver for GHC 9.2 - - - - - 386751a1 by Meng Weng Wong at 2022-02-25T19:19:11+01:00 IDoc link has bitrotted; replaced with web.archive.org cache. (#1454) - - - - - d877cbe6 by Hécate Moonlight at 2022-02-25T19:21:58+01:00 Fix haddock user guide (#1456) - - - - - cc47f036 by Andrew Lelechenko at 2022-03-04T17:29:36+01:00 Allow text-2.0 in haddock-library (#1459) - - - - - 7b3685a3 by malteneuss at 2022-03-07T19:27:24+01:00 Add multi-line style hint to style section (#1460) - - - - - c51088b8 by John Ericson at 2022-03-11T16:46:26+01:00 Fix CollectPass instance to match TTG refactor Companion to GHC !7614 (https://gitlab.haskell.org/ghc/ghc/-/merge_requests/7614) - - - - - b882195b by Vladislav Zavialov at 2022-03-14T20:32:30+01:00 Link to (~) - - - - - 877349b8 by Christiaan Baaij at 2022-03-16T09:20:43+01:00 Add Haddock support for the OPAQUE pragma - - - - - 0ea22721 by askeblad at 2022-03-16T09:44:27+01:00 typos (#1464) - - - - - a6d13da1 by Matthew Pickering at 2022-03-22T13:41:17+00:00 Minimum changes needed for compilation with hi-haddock With hi-haddock, of course there is a much large refactoring of haddock which could be achieved but that is left for a future patch which can implemented at any time independently of GHC. - - - - - e7ac9129 by Matthew Pickering at 2022-03-22T21:17:50+00:00 Update test output - - - - - 6d916214 by Matthew Pickering at 2022-03-24T15:06:26+00:00 Merge branch 'wip/opaque_pragma' into 'ghc-head' Add Haddock support for the OPAQUE pragma See merge request ghc/haddock!2 - - - - - 42208183 by Steve Hart at 2022-03-25T20:43:50+01:00 Fix CI (#1467) * CI: Reinstall GHC with docs CI tests were failing because the GHC preinstalled to the CI environment does not include documentation, which is required for running the Haddock tests. This commit causes the CI workflow to reinstall GHC with docs so that tests can succeed. - - - - - 9676fd79 by Steve Hart at 2022-03-25T21:33:34+01:00 Make links in Synopsis functional again (#1458) Commit e41c1cbe9f0476997eac7b4a3f17cbc6b2262faf added a call to e.preventDefault() when handling click events that reach a toggle element. This prevents the browser from following hyperlinks within the Synopsis section when they are clicked by a user. This commit restores functioning hyperlinks within the Synopsis section by removing the call to e.preventDefault(), as it does not appear to be necessary, and removing it increases the flexibility of the details-helper code. - - - - - d1edd637 by sheaf at 2022-04-01T12:02:02+02:00 Keep track of promotion ticks in HsOpTy Keeping track of promotion ticks in HsOpTy allows us to properly pretty-print promoted constructors such as lists. - - - - - 9dcb2dfc by Jakob Brünker at 2022-04-01T15:46:22+00:00 Add support for \cases See merge request ghc/ghc!7873 - - - - - b0412ee5 by askeblad at 2022-04-06T17:47:57+02:00 spelling errors (#1471) - - - - - 6b18829b by Vladislav Zavialov at 2022-04-06T18:53:58+02:00 Rename [] to List - - - - - 2d046691 by Vladislav Zavialov at 2022-04-07T20:25:54+03:00 HsToken ConDeclGADT con_dcolon - - - - - 90b43da4 by Steve Hart at 2022-04-12T13:29:46+02:00 Parse Markdown links at beginning of line within a paragraph (#1470) * Catch Markdown links at beginning of line within paragraph Per Issue haskell/haddock#774, Markdown links were being parsed as ordinary text when they occurred at the beginning of a line other than the first line of the paragraph. This occurred because the parser was not interpreting a left square bracket as a special character that could delimit special markup. A space character was considered a special character, so, if a space occurred at the beginning of the new line, then the parser would interpret the space by itself and then continue parsing, thereby catching the Markdown link. '\n' was not treated as a special character, so the parser did not catch a Markdown link that may have followed. Note that this will allow for Markdown links that are not surrounded by spaces. For example, the following text includes a Markdown link that will be parsed: Hello, world[label](url) This is consistent with how the parser handles other types of markup. * Remove obsolete documentation hint Commit 6b9aeafddf20efc65d3725c16e3fc43a20aac343 should eliminate the need for the workaround suggested in the documentation. - - - - - 5b08312d by Hécate Moonlight at 2022-04-12T13:36:38+02:00 Force ghc-9.2 in the cabal.project - - - - - 0d0ea349 by dependabot[bot] at 2022-04-12T13:57:41+02:00 Bump path-parse from 1.0.5 to 1.0.7 in /haddock-api/resources/html (#1469) Bumps [path-parse](https://github.com/jbgutierrez/path-parse) from 1.0.5 to 1.0.7. - [Release notes](https://github.com/jbgutierrez/path-parse/releases) - [Commits](https://github.com/jbgutierrez/path-parse/commits/v1.0.7) --- updated-dependencies: - dependency-name: path-parse dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 2b9fc65e by dependabot[bot] at 2022-04-12T13:57:54+02:00 Bump copy-props from 2.0.4 to 2.0.5 in /haddock-api/resources/html (#1468) Bumps [copy-props](https://github.com/gulpjs/copy-props) from 2.0.4 to 2.0.5. - [Release notes](https://github.com/gulpjs/copy-props/releases) - [Changelog](https://github.com/gulpjs/copy-props/blob/master/CHANGELOG.md) - [Commits](https://github.com/gulpjs/copy-props/compare/2.0.4...2.0.5) --- updated-dependencies: - dependency-name: copy-props dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - ea98a6fb by Ben Gamari at 2022-04-23T22:54:37-04:00 Update for GHC 9.4 - - - - - 9e11864f by Ben Gamari at 2022-04-25T16:24:31-04:00 Merge remote-tracking branch 'upstream/ghc-9.2' into ghc-head - - - - - f83cc506 by Ben Gamari at 2022-04-25T17:00:25-04:00 Bump ghc version to 9.5 - - - - - e01c2e7d by Ben Gamari at 2022-04-28T16:19:04-04:00 Revert "Bump ghc-head (#1445)" This reverts commit b29a78ef6926101338f62e84f456dac8659dc9d2. This should not have been merged. - - - - - a2b5ee8c by Ben Gamari at 2022-04-28T16:19:24-04:00 Merge commit '2627a86c' into ghc-head - - - - - 0c6fe4f9 by Ben Gamari at 2022-04-29T10:05:54-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-9.4 - - - - - b6e5cb0a by Ben Gamari at 2022-04-29T11:46:06-04:00 Revert "HsToken ConDeclGADT con_dcolon" This reverts commit 24208496649a02d5f87373052c430ea4a97842c5. - - - - - 15a62888 by Ben Gamari at 2022-04-29T15:12:55-04:00 Bump base upper bound - - - - - 165b9031 by Ben Gamari at 2022-04-29T23:58:38-04:00 Update test output - - - - - e0c3e5da by Phil de Joux at 2022-05-02T14:46:38+02:00 Add hlint action .hlint.yaml with ignores & CPP. (#1475) - - - - - ead1158d by Raphael Das Gupta at 2022-05-02T14:46:48+02:00 fix grammar in docs: "can the" → "can be" (#1477) - - - - - cff97944 by Ben Gamari at 2022-05-02T18:38:56-04:00 Allow base-4.17 - - - - - e4ecb201 by Phil de Joux at 2022-05-03T13:14:55+02:00 Remove unused imports that GHC warned about. (#1480) - - - - - 222890b1 by Phil de Joux at 2022-05-03T13:15:46+02:00 Follow hlint suggestion to remove redundant bang. (#1479) - - - - - 058b671f by Phil de Joux at 2022-05-03T13:34:04+02:00 Follow hlint, remove language pragmas in libs. (#1478) - - - - - 0a645049 by Ben Simms at 2022-05-03T14:19:24+02:00 Keep track of ordered list indexes and render them (#1407) * Keep track of ordered list indexes and render them * Rename some identifiers to clarify - - - - - f0433304 by Norman Ramsey at 2022-05-04T15:13:34-04:00 update for changes in GHC API - - - - - 3740cf71 by Emily Martins at 2022-05-06T18:23:48+02:00 Add link to the readthedocs in cabal description to show on hackage. (cherry picked from commit 52e2d40d47295c02d3181aac0c53028e730f1e3b) - - - - - 5d754f1e by Hécate Moonlight at 2022-05-06T18:44:57+02:00 remove Bug873 - - - - - 968fc267 by Hécate Moonlight at 2022-05-06T18:48:28+02:00 Ignore "Use second" HLint suggestion. It increases laziness. - - - - - 02d14e97 by Jade Lovelace at 2022-05-07T17:42:08+02:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` - - - - - b22b87ed by Artem Pelenitsyn at 2022-05-08T16:19:47+02:00 fix parsing trailing quotes in backticked identifiers (#1408) (#1483) - - - - - 80ae107b by Alex Biehl at 2022-05-08T16:37:16+02:00 Fix "Defined by not used" error (cherry picked from commit 6e02a620a26c3a44f98675dd1b93b08070c36c0a) - - - - - 4c838e84 by Hécate Moonlight at 2022-05-08T16:37:16+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - fc9827b4 by Hécate Moonlight at 2022-05-08T16:40:40+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - b153b555 by Xia Li-yao at 2022-05-20T17:52:42+02:00 Hide synopsis from search when hidden (#1486) Fix haskell/haddock#1451 - - - - - f3e38b85 by Marcin Szamotulski at 2022-05-21T23:32:31+02:00 Allow to hide interfaces when rendering multiple components (#1487) This is useful when one wishes to `--gen-contents` when rendering multiple components, but one does not want to render all modules. This is in particular useful when adding base package. - - - - - f942863b by Marcin Szamotulski at 2022-05-24T08:29:59+02:00 Check if doc-index.json exists before reading it (#1488) - - - - - 31e92982 by Marcin Szamotulski at 2022-05-25T16:22:13+02:00 Version bump 2.26.1 (#1489) * Version bump 2.26.1 We extended format accepted by `--read-interface` option, which requires updating the minor version. * Update documentation of --read-interface option - - - - - 7cc873e0 by sheaf at 2022-05-25T16:42:31+02:00 Updated HaddockHypsrcTest output for record update changes (MR !7981) - - - - - cd196942 by Marcin Szamotulski at 2022-05-25T20:28:47+02:00 Use visibility to decide which interfaces are included in quickjump (#1490) This is also consistent with how html index is build. See haskell/cabal#7669 for rationale behind this decision. - - - - - 00c713c5 by Hécate Moonlight at 2022-05-26T17:09:15+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 2f3039f1 by Hécate Moonlight at 2022-05-26T17:10:59+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 63a5650c by romes at 2022-05-31T12:43:22+01:00 TTG: Match new GHC AST - - - - - dd7d1617 by romes at 2022-06-02T16:11:00+01:00 Update for IE changes in !8228 - - - - - c23aaab7 by cydparser at 2022-06-06T08:48:14+02:00 Fix and improve CI (#1495) * Pin GHC version before creating the freeze file * Use newest action versions * Improve caching * Avoid unnecessarily reinstalling GHC * Use GHC 9.2.2 for CI Co-authored-by: Cyd Wise <cwise at tripshot.com> - - - - - c156fa77 by Hécate Moonlight at 2022-06-06T11:59:35+02:00 Add Mergify configuration (#1496) - - - - - 2dba4188 by Hécate Moonlight at 2022-06-06T16:12:50+02:00 Bump haddock's version in cabal file to 2.26.1 (#1497) - - - - - d7d4b8b9 by Marcin Szamotulski at 2022-06-07T06:09:40+00:00 Render module tree per package in the content page (#1492) * Render module tree per package in the content page When rendering content page for multiple packages it is useful to split the module tree per package. Package names in this patch are inferred from haddock's interface file names. * Write PackageInfo into interface file To keep interface file format backward compatible, instead of using `Binary` instance for `InterfaceFile` we introduce functions to serialise and deserialise, which depends on the interface file version. - - - - - 77765665 by Mike Pilgrem at 2022-06-12T21:57:19+01:00 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - b0e079b0 by mergify[bot] at 2022-06-13T11:49:37+00:00 Merge pull request haskell/haddock#1108 from mpilgrem/fix783 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - 6c0292b1 by Hécate Moonlight at 2022-06-21T17:21:08+02:00 Update the contribution guide - - - - - e413b9fa by dependabot[bot] at 2022-06-21T23:38:19+02:00 Bump shell-quote from 1.6.1 to 1.7.3 in /haddock-api/resources/html (#1500) Bumps [shell-quote](https://github.com/substack/node-shell-quote) from 1.6.1 to 1.7.3. - [Release notes](https://github.com/substack/node-shell-quote/releases) - [Changelog](https://github.com/substack/node-shell-quote/blob/master/CHANGELOG.md) - [Commits](https://github.com/substack/node-shell-quote/compare/1.6.1...1.7.3) --- updated-dependencies: - dependency-name: shell-quote dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 29d0ef70 by romes at 2022-07-06T11:29:39+02:00 TTG: AST Updates for !8308 - - - - - 1bae7c87 by Alan Zimmerman at 2022-07-06T22:50:43+01:00 Match GHC changes for T21805 This brings in a newtype for FieldLabelString - - - - - 6fe8b988 by Phil de Joux at 2022-07-16T20:54:26+00:00 Bump hlint version to 3.4.1, the version with counts. (#1503) Redo the counts available with the --default option. - - - - - 48fb43af by Phil de Joux at 2022-07-19T09:32:55+02:00 Follow hlint suggestion: unused LANGUAGE pragma. (#1504) * Follow hlint suggestion: unused LANGUAGE pragma. * Ignore within modules to pass linting and pass tests. - - - - - c1cf1fa7 by Phil de Joux at 2022-07-24T13:45:59+02:00 Follow hlint suggestion: redundant $. (#1505) * Follow hlint suggestion: redundant $. * Remove $ and surplus blank lines in Operators. - - - - - 74777eb2 by Jade Lovelace at 2022-07-29T11:02:41+01:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` (cherry picked from commit ab53ccf089ea703b767581ac14be0f6c78a7678a) - - - - - faa4cfcf by Hécate Moonlight at 2022-07-29T20:31:20+02:00 Merge pull request haskell/haddock#1516 from duog/9-4-backport-fix-hyperlinks Backport 9-4: Fix hyperlinks to external items and modules (#1482) - - - - - 5d2450f3 by Ben Gamari at 2022-08-05T17:41:15-04:00 Merge remote-tracking branch 'origin/ghc-9.4' - - - - - 63954f73 by Ben Gamari at 2022-08-05T19:08:36-04:00 Clean up build and testsuite for GHC 9.4 - - - - - d4568cb8 by Hécate Moonlight at 2022-08-05T19:10:49-04:00 Bump the versions - - - - - 505583a4 by Ben Gamari at 2022-08-06T13:58:27-04:00 Merge pull request haskell/haddock#1518 from bgamari/wip/ghc-9.4-merge Merge GHC 9.4 into `main` - - - - - 5706f6a4 by Ben Gamari at 2022-08-06T22:57:21-04:00 html-test: Testsuite changes for GHC 9.4.1 - - - - - 5f2a45a2 by Ben Gamari at 2022-08-15T14:33:05-04:00 doc: Fix a few minor ReST issues Sphinx was complaining about too-short title underlines. - - - - - 220e6410 by Ben Gamari at 2022-08-15T14:41:24-04:00 Merge branch 'main' into ghc-head - - - - - fbeb1b02 by Ben Gamari at 2022-08-15T14:45:16-04:00 Updates for GHC 9.5 - - - - - eee562eb by Vladislav Zavialov at 2022-08-15T14:46:13-04:00 HsToken ConDeclGADT con_dcolon - - - - - c5f073db by Ben Gamari at 2022-08-15T16:55:35-04:00 Updates for GHC 9.5 - - - - - 3f7ab242 by Vladislav Zavialov at 2022-08-15T16:55:35-04:00 HsToken ConDeclGADT con_dcolon - - - - - a18e473d by Ben Gamari at 2022-08-16T08:35:19-04:00 Merge branch 'wip/ghc-head-bump' into ghc-head - - - - - af0ff3a4 by M Farkas-Dyck at 2022-09-15T21:16:05+00:00 Disuse `mapLoc`. - - - - - a748fc38 by Matthew Farkas-Dyck at 2022-09-17T10:44:18+00:00 Scrub partiality about `NewOrData`. - - - - - 2758fb6c by John Ericson at 2022-09-18T03:27:37+02:00 Test output changed because of change to `base` Spooky, but I guess that is intended? - - - - - a7eec128 by Torsten Schmits at 2022-09-21T11:06:55+02:00 update tests for the move of tuples to GHC.Tuple.Prim - - - - - 461e7b9d by Ross Paterson at 2022-09-24T22:01:25+00:00 match implementation of GHC proposal haskell/haddock#106 (Define Kinds Without Promotion) - - - - - f7fd77ef by sheaf at 2022-10-17T14:53:01+02:00 Update Haddock for GHC MR !8563 (configuration of diagnostics) - - - - - 3d3e85ab by Vladislav Zavialov at 2022-10-22T23:04:06+03:00 Class layout info - - - - - cbde4cb0 by Simon Peyton Jones at 2022-10-25T23:19:18+01:00 Adapt to Constraint-vs-Type See haskell/haddock#21623 and !8750 - - - - - 7108ba96 by Tom Smeding at 2022-11-01T22:33:23+01:00 Remove outdated footnote about module re-exports The footnote is invalid with GHC 9.2.4 (and possibly earlier): the described behaviour in the main text works fine. - - - - - 206c6bc7 by Hécate Moonlight at 2022-11-01T23:00:46+01:00 Merge pull request haskell/haddock#1534 from tomsmeding/patch-1 - - - - - a57b4c4b by Andrew Lelechenko at 2022-11-21T00:39:52+00:00 Support mtl-2.3 - - - - - e9d62453 by Simon Peyton Jones at 2022-11-25T13:49:12+01:00 Track small API change in TyCon.hs - - - - - eb1c73f7 by Ben Gamari at 2022-12-07T08:46:21-05:00 Update for GhC 9.6 - - - - - 063268dd by Ben Gamari at 2022-12-07T11:26:32-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 4ca722fe by Ben Gamari at 2022-12-08T14:43:26-05:00 Bump bounds to accomodate base-4.18 - - - - - 340b7511 by Vladislav Zavialov at 2022-12-10T12:31:28+00:00 HsToken in HsAppKindTy - - - - - 946226ec by Ben Gamari at 2022-12-13T20:12:56-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - fd8faa66 by Ben Gamari at 2022-12-22T13:44:28-05:00 Bump GHC version to 9.7 - - - - - 2958aa9c by Ben Gamari at 2022-12-22T14:49:16-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 9e0fefd8 by Andrei Borzenkov at 2023-01-30T14:02:04+04:00 Rename () to Unit, Rename (,,...,,) to Tuple<n> - - - - - eb3968b5 by Ben Gamari at 2023-03-10T02:32:43-05:00 Bump versions for ghc-9.6 release - - - - - 4aeead36 by Adam Gundry at 2023-03-23T13:53:47+01:00 Adapt to warning categories changes - - - - - 642d8d60 by sheaf at 2023-03-29T13:35:56+02:00 Adapt to record field refactor This commit adapts to the changes in GHC MR !8686, which overhauls the treatment of record fields in the renamer, adding separate record field namespaces and entirely removing the GreName datatype. - - - - - ac8d4333 by doyougnu at 2023-03-29T11:11:44-04:00 Update UniqMap API - - - - - 7866fc86 by Ben Orchard at 2023-04-20T11:29:33+02:00 update classify with new tokens - - - - - ffcdd683 by Finley McIlwaine at 2023-04-24T09:36:18-06:00 Remove index-state - - - - - 05b70982 by Finley McIlwaine at 2023-04-26T08:16:31-06:00 `renameInterface` space leak fixes - Change logic for accumulation of names for which link warnings will be generated - Change handling of `--ignore-link-symbol` to allow qualified and unqualified names. Added to CHANGES.md - Some formatting changes and comments here and there - - - - - e5697d7c by Finley McIlwaine at 2023-04-27T18:46:36-06:00 Messy things - ghc-debug dependency and instrumentation - cabal.project custom with-compiler - hie.yaml files - traces and such - - - - - 0b8ef80b by Finley McIlwaine at 2023-05-02T18:08:52-06:00 Stop retaining GRE closures GRE closures should never be necessary to Haddock, so we never want to keep them on the heap. Despite that, they are retained by a lot of the data structures that Haddock makes use of. - Attempt to fix that situation by adding strictness to various computations and pruning the `ifaceInstances` field of `Interface` to a much thinner data type. - Removes the `ifaceFamInstances` field, as it was never used. - Move some of the attach instances types (e.g. `SimpleType`) to the types module - - - - - 8bda991b by Finley McIlwaine at 2023-05-08T16:07:51-06:00 Memory usage fixes - Refactor `ifaceDeclMap` to drastically reduce memory footprint. We no longer store all declarations associated with a given name, since we only cared to determine if the only declaration associated with a name was a value declaration. Change the `DeclMap` type to better reflect this. - Drop pre-renaming export items after the renaming step. Since the Hoogle backend used the pre-renamed export items, this isn't trivial. We now generate Hoogle output for exported declarations during the renaming step (if Hoogle output /should/ be generated), and store that with the renamed export item. - Slightly refactor Hoogle backend to handle the above change and allow for early generation of Hoogle output. - Remove the `ifaceRnDocMap` and `ifaceRnArgMap` fields of the `Interface` type, as they were never used. - Remove some unnecessary strictness - Remove a lot of dead code from `Syb` module - - - - - 1611ac0c by Finley McIlwaine at 2023-05-09T11:51:57-06:00 Unify ErrMsgM and IfM - Delete ErrMsgM, stop accumulating warnings in a writer - Make IfM a state monad, print warnings directly to stdout, move IfM type into types module - Drop ErrMsg = String synonym - Unset IORefs from plugin after they are read, preventing unnecessary retention of interfaces - - - - - 42d696ab by Finley McIlwaine at 2023-05-11T15:52:07-06:00 Thunk leak fixes The strictness introduced in this commit was motivated by observing thunk leaks in the eventlog2html output. - Refactor attach instances list comprehension to avoid large intermediate thunks - Refactor some HTML backend list comprehensions to avoid large intermediate thunks - Avoid thunks accumulating in documentation types or documentation parser - A lot of orphan NFData instances to allow us to force documentation values - - - - - 68561cf6 by Finley McIlwaine at 2023-05-11T17:02:10-06:00 Remove GHC debug dep - - - - - 10519e3d by Finley McIlwaine at 2023-05-15T12:40:48-06:00 Force HIE file path Removes a potential retainer of `ModSummary`s - - - - - 1e4a6ec6 by Finley McIlwaine at 2023-05-15T14:20:34-06:00 Re-add index-state, with-compiler, delete hie.yamls - - - - - a2363fe9 by Hécate Moonlight at 2023-05-15T22:45:16+02:00 Merge pull request haskell/haddock#1594 from FinleyMcIlwaine/finley/ghc-9.6-mem-fixes Reduce memory usage - - - - - e8a78383 by Finley McIlwaine at 2023-05-17T12:19:16-06:00 Merge branch ghc-9.6 into ghc-head - - - - - 22e25581 by Finley McIlwaine at 2023-05-17T12:20:23-06:00 Merge branch 'ghc-head' of gitlab.haskell.org:ghc/haddock into ghc-head - - - - - 41bbf0df by Bartłomiej Cieślar at 2023-05-24T08:57:58+02:00 changes to the WarningTxt cases Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - c686ba9b by Hécate Moonlight at 2023-06-01T14:03:02-06:00 Port the remains of Hi-Haddock - - - - - 9d8a85fd by Hécate Moonlight at 2023-06-01T14:03:06-06:00 Stdout for tests - - - - - 36331d07 by Finley McIlwaine at 2023-06-01T14:06:02-06:00 Formatting, organize imports - - - - - a06059b1 by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix empty context confusion in Convert module - - - - - 379346ae by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix associated type families in Hoogle output - - - - - fc6ea7ed by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix test refs Accept several changes in Hoogle tests: Pretty printing logic no longer prints the `(Proxy (Proxy (...))` chain in Bug806 with parentheses. Since this test was only meant to verify that line breaks do not occur, accept the change. `tyThingToLHsDecl` is called for class and data declarations, which ends up "synifying" the type variables and giving unlifted types kind signatures. As a result, type variables of kind `Type -> Type` are now printed with kind signatures in Hoogle output. This could be changed by manually drop kind signatures from class variables in the Hoogle backend if the behavior is deemed unacceptable. Sometimes subordinate declarations are exported separate from their parent declarations (e.g. record selectors). In this case, a type signature is cobbled together for the export item in `extractDecl`. Since this type signature is very manually constructed, it may lack kind signatures of decls constructed from `tyThingToLHsDecl`. An example of this is the `type-sigs` Hoogle test. Change `*` to `Type` in Hoogle test refs. I don't think this will break Hoogle behavior, since it appears to not consider type signatures in search. I have not fully verified this. - - - - - e14b7e58 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix LaTeX backend test refs Changes to GHC pretty printing code have resulted in some differences to Haddock's LaTeX output. - Type variables are printed explicitly quantified in the LinearTypes test - Wildcard types in type family equations are now printed numbered, e.g. _1 _2, in the TypeFamilies3 test - Combined signatures in DefaultSignatures test are now documented as separate signatures - - - - - 41b5b296 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and test source updates - Stop using kind `*` in html test sources - Add TypeOperators where necessary to avoid warnings and future errors - Rename some test modules to match their module names - - - - - c640e2a2 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix missing deprecation warnings on record fields `lookupOccEnv` was used to resolve `OccNames` with warnings attached, but it doesn't look in the record field namespace. Thus, if a record field had a warning attached, it would not resolve and the warning map would not include it. This commit fixes by using `lookupOccEnv_WithFields` instead. - - - - - fad0c462 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and some comments - - - - - 751fd023 by Finley McIlwaine at 2023-06-01T14:11:41-06:00 Accept HTML test diffs All diffs now boil down to the expected differences resulting from declarations being reified from TyThings in hi-haddock. Surface syntax now has much less control over the syntax used in the documentation. - - - - - d835c845 by Finley McIlwaine at 2023-06-01T14:11:45-06:00 Adapt to new load' type - - - - - dcf776c4 by Finley McIlwaine at 2023-06-01T14:13:13-06:00 Update mkWarningMap and moduleWarning - - - - - 8e8432fd by Finley McIlwaine at 2023-06-01T14:28:54-06:00 Revert load' changes - - - - - aeb2982c by Finley McIlwaine at 2023-06-01T14:40:24-06:00 Accept change to Instances test in html-test Link to Data.Tuple instead of GHC.Tuple.Prim - - - - - 8adfdbac by Finley McIlwaine at 2023-06-01T15:53:17-06:00 Reset ghc dep to ^>= 9.6 - - - - - 2b1ce93d by Finley McIlwaine at 2023-06-06T07:50:04-06:00 Update CHANGES.md, user guide, recomp avoidance * Add --trace-args flag for tracing arguments received to standard output * Avoid recompiling due to changes in optimization flags * Update users guide and changes.md - - - - - f3da6676 by Finley McIlwaine at 2023-06-06T14:12:56-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - bf36c467 by Matthew Pickering at 2023-06-07T10:16:09+01:00 Revert back to e16e20d592a6f5d9ed1af17b77fafd6495242345 Neither of these MRs are ready to land yet which causes issues with other MRs which are ready to land and need haddock changes. - - - - - 421510a9 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 atSign has no unicode variant Prior to this change, atSign was defined as follows: atSign unicode = text (if unicode then "@" else "@") Yes, this is the same symbol '\64' and not your font playing tricks on you. Now we define: atSign = char '@' Both the LaTeX and the Xhtml backend are updated accordingly. - - - - - 3785c276 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 LaTeX: fix printing of type variable bindings Consider this type signature: kindOf :: forall {k} (a :: k). Proxy a -> Proxy k Prior to this fix, the LaTeX backend rendered it like this: kindOf :: forall k a. Proxy a -> Proxy k Now we preserve explicit specificity and kind annotations. - - - - - 0febf3a8 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 Add support for invisible binders in type declarations - - - - - 13e33bb3 by Finley McIlwaine at 2023-06-08T07:51:59-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - 3e5340ce by Finley McIlwaine at 2023-06-08T07:54:27-06:00 Add note about stubdir to recompilation docs - - - - - db7e84dc by Finley at 2023-06-08T08:11:03-06:00 Merge pull request haskell/haddock#1597 from haskell/finley/hi-haddock-9.6 hi-haddock for ghc 9.6 - - - - - 4e085d17 by Finley McIlwaine at 2023-06-14T13:41:06-06:00 Replace SYB traversals - - - - - 7b39aec5 by Finley McIlwaine at 2023-06-14T14:20:17-06:00 Test ref accepts, remove unused HaddockClsInst - - - - - df9c2090 by Finley McIlwaine at 2023-06-15T08:02:51-06:00 Use batchMsg for progress reporting during load With hi-haddock as is, there is an awkward silence during the load operation. This commit makes haddock use the default `batchMsg` Messager for progress reporting, and makes the default GHC verbosity level 1, so the user can see what GHC is doing. - - - - - f23679a8 by Hécate Moonlight at 2023-06-15T20:31:53+02:00 Merge pull request haskell/haddock#1600 from haskell/finley/hi-haddock-optim - - - - - a7982192 by Finley McIlwaine at 2023-06-15T15:02:16-06:00 hi-haddock squashed - - - - - c34f0c8d by Finley McIlwaine at 2023-06-15T16:22:03-06:00 Merge remote-tracking branch 'origin/ghc-9.6' into finley/hi-haddock-squashed - - - - - 40452797 by Bartłomiej Cieślar at 2023-06-16T12:26:04+02:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - e58673bf by Ben Gamari at 2023-06-16T09:33:35-04:00 Bump GHC version to 9.8 - - - - - 74bdf972 by Ben Gamari at 2023-06-16T09:36:18-04:00 Merge commit 'fcaaad06770a26d35d4aafd65772dedadf17669c' into ghc-head - - - - - 418ee3dc by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Remove NFData SourceText, IfaceWarnings updates The NFData SourceText instance is now available in GHC Handle mod_iface mi_warns now being IfaceWarnings - - - - - 62f31380 by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Accept Instances.hs test output Due to ghc!10469. - - - - - a8f2fc0e by Ben Gamari at 2023-06-20T15:48:08-04:00 Test fixes for "Fix associated data family doc structure items" Associated data families were being given their own export DocStructureItems, which resulted in them being documented separately from their classes in haddocks. This commit fixes it. - - - - - cb1ac33e by Bartłomiej Cieślar at 2023-06-21T12:56:02-04:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - 9933e10b by Ben Gamari at 2023-06-21T12:56:02-04:00 Bump GHC version to 9.8 - - - - - fe8c18b6 by Ben Gamari at 2023-06-21T15:36:29-04:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - c61a0d5b by Ben Gamari at 2023-06-21T16:10:51-04:00 Bump GHC version to 9.9 - - - - - 0c2a756e by sheaf at 2023-07-07T13:45:12+02:00 Avoid incomplete record update in Haddock Hoogle This commit avoids running into an incomplete record update warning in the Hoogle backend of Haddock. This was only noticed now, because incomplete record updates were broken in GHC 9.6. Now that they are fixed, we have to avoid running into them! - - - - - f9b952a7 by Ben Gamari at 2023-07-21T11:58:05-04:00 Bump base bound to <4.20 For GHC 9.8. - - - - - 1b27e151 by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Check for puns (see ghc#23368) - - - - - 457341fd by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Remove fake exports for (~), List, and Tuple<n> The old reasoning no longer applies, nowadays those names can be mentioned in export lists. - - - - - bf3dcddf by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Fix pretty-printing of Solo and MkSolo - - - - - 495b2241 by Matthew Pickering at 2023-09-01T13:02:07+02:00 Fix issue with duplicate reexported definitions (T23616) When a class method was reexported, it's default methods were also showing up in the generated html page. The simplest and most non-invasive fix is to not look for the default method if we are just exporting the class method.. because the backends are just showing default methods when the whole class is exported. In general it would be worthwhile to rewrite this bit of code I think as the logic and what gets included is split over `lookupDocs` and `availExportDecl` it would be clearer to combine the two. The result of lookupDocs is always just passed to availExportDecl so it seems simpler and more obvious to just write the function directly. - - - - - 6551824d by Finley McIlwaine at 2023-09-05T13:06:57-07:00 Remove fake export of `FUN` from Prelude This prevents `data FUN` from being shown at the top of the Prelude docs. Fixes \#23920 on GHC. - - - - - 9ab5a448 by Alan Zimmerman at 2023-09-08T18:26:53+01:00 Match changes in wip/az/T23885-unicode-funtycon - - - - - 4d08364e by Alan Zimmerman at 2023-10-31T19:46:45+00:00 EPA: match changes in GHC - EPA: Comments in AnchorOperation - EPA: Remove EpaEofComment - - - - - e7da0d25 by Alan Zimmerman at 2023-11-05T11:20:31+00:00 EPA: match changes in GHC, l2l cleanup - - - - - 4ceac14d by Alan Zimmerman at 2023-11-11T15:16:41+00:00 EPA: Replace Anchor with EpaLocation Match GHC - - - - - 94fb8d47 by Alan Zimmerman at 2023-11-29T18:10:26+00:00 Match GHC, No comments in EpaDelta for comments - - - - - 32d208e1 by Vladislav Zavialov at 2023-12-12T20:41:36+03:00 EPA: Match changes to LHsToken removal - - - - - eebdd316 by Apoorv Ingle at 2024-01-23T13:49:12+00:00 Changes for haskell/haddock#18324 - - - - - ae856a82 by Matthew Pickering at 2024-02-05T12:22:39+00:00 ghc-internals fallout - - - - - f8429266 by Jade at 2024-02-08T14:56:50+01:00 Adjust test for ghc MR !10993 - - - - - 6d1e2386 by Alan Zimmerman at 2024-02-13T22:00:28+03:00 EPA: Match changes to HsParTy and HsFunTy - - - - - 9c588f19 by Fendor at 2024-02-14T11:05:36+01:00 Adapt to GHC giving better Name's for linking - - - - - 778e1db3 by Andrei Borzenkov at 2024-02-16T16:12:07+03:00 Namespace specifiers for fixity signatures - - - - - 826c5b47 by Torsten Schmits at 2024-02-21T13:17:05+01:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 2cff14d5 by Ben Gamari at 2024-02-22T09:35:56-05:00 Bump bounds - - - - - f49376b3 by Ben Gamari at 2024-02-22T09:35:56-05:00 Allow `@since` annotations in export lists Here we extend Haddock to admit `@since` annotations in export lists. These can be attached to most export list items (although not subordinate lists). These annotations supercede the declaration's `@since` annotation in produced Haddocks. - - - - - b5aa93df by Ben Gamari at 2024-02-22T12:09:06-05:00 Allow package-qualified @since declarations - - - - - 8f5957f2 by Ben Gamari at 2024-02-22T13:55:19-05:00 Documentation changes from ghc-internal restructuring Previously many declarations (e.g. `Int`) were declared to have a "home" in `Prelude`. However, now Haddock instead chooses to put these in more specific homes (e.g. `Data.Int`). Given that the "home" decision is driven by heuristics and in general these changes seem quite reasonable I am accepting them: * `Int` moved from `Prelude` to `Data.Int` * `(~)` moved from `Prelude` to `Data.Type.Equality` * `Type` moved from `GHC.Types` to `Data.Kind` * `Maybe` moved from `Prelude` to `Data.Maybe` * `Bool` moved from `Prelude` to `Data.Bool` * `Ordering` moved from `Prelude` to `Data.Ord` As well, more identifiers are now hyperlinked; it's not immediately clear *why*, but it is an improvement nevertheless. - - - - - ec33fec3 by Ben Gamari at 2024-02-22T20:36:24-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 30cfd251 by Torsten Schmits at 2024-02-24T13:00:42-05:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 732db81d by Ben Gamari at 2024-02-24T19:12:18-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 86bf7010 by Ben Gamari at 2024-02-27T19:28:10-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - dedcf102 by Cheng Shao at 2024-03-06T13:39:04-05:00 rts: expose HeapAlloc.h as public header This commit exposes HeapAlloc.h as a public header. The intention is to expose HEAP_ALLOCED/HEAP_ALLOCED_GC, so they can be used in assertions in other public headers, and they may also be useful for user code. - - - - - d19441d7 by Cheng Shao at 2024-03-06T13:39:04-05:00 rts: assert pointer is indeed heap allocated in Bdescr() This commit adds an assertion to Bdescr() to assert the pointer is indeed heap allocated. This is useful to rule out RTS bugs that attempt to access non-existent block descriptor of a static closure, #24492 being one such example. - - - - - 9a656a04 by Ben Gamari at 2024-03-06T13:39:39-05:00 ghc-experimental: Add dummy dependencies to work around #23942 This is a temporary measure to improve CI reliability until a proper solution is developed. Works around #23942. - - - - - 1e84b924 by Simon Peyton Jones at 2024-03-06T13:39:39-05:00 Three compile perf improvements with deep nesting These were changes are all triggered by #24471. 1. Make GHC.Core.Opt.SetLevels.lvlMFE behave better when there are many free variables. See Note [Large free-variable sets]. 2. Make GHC.Core.Opt.Arity.floatIn a bit lazier in its Cost argument. This benefits the common case where the ArityType turns out to be nullary. See Note [Care with nested expressions] 3. Make GHC.CoreToStg.Prep.cpeArg behave for deeply-nested expressions. See Note [Eta expansion of arguments in CorePrep] wrinkle (EA2). Compile times go down by up to 4.5%, and much more in artificial cases. (Geo mean of compiler/perf changes is -0.4%.) Metric Decrease: CoOpt_Read T10421 T12425 - - - - - c4b13113 by Hécate Moonlight at 2024-03-06T13:40:17-05:00 Use "module" instead of "library" when applicable in base haddocks - - - - - 9cd9efb4 by Vladislav Zavialov at 2024-03-07T13:01:54+03:00 Rephrase error message to say "visible arguments" (#24318) * Main change: make the error message generated by mkFunTysMsg more accurate by changing "value arguments" to "visible arguments". * Refactor: define a new type synonym VisArity and use it instead of Arity in a few places. It might be the case that there other places in the compiler that should talk about visible arguments rather than value arguments, but I haven't tried to find them all, focusing only on the error message reported in the ticket. - - - - - 4b6e76b5 by Patrick at 2024-03-07T22:09:30+08:00 fix haskell/haddock#24493, with module name introduced in hieAst The accompanies haddoc PR with GHC PR https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12153 Two things have been done: 1. Link is introduced to every `X` in `module X where`, since we introduce the module name to HieAst, 2. `{-# LINE 4 "hypsrc-test/src/PositionPragmas.hs" #-}` is added before the `module PositionPragmas where` in ` hypsrc-test/ref/src/PositionPragmas.html `.It ensures only a single hieAst for file `hypsrc-test/src/PositionPragmas.hs` is generated. - - - - - d523a6a7 by Ben Gamari at 2024-03-07T19:40:45-05:00 Bump array submodule - - - - - 7e55003c by Ben Gamari at 2024-03-07T19:40:45-05:00 Bump stm submodule - - - - - 32d337ef by Ben Gamari at 2024-03-07T19:40:45-05:00 Introduce exception context Here we introduce the `ExceptionContext` type and `ExceptionAnnotation` class, allowing dynamically-typed user-defined annotations to be attached to exceptions. CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/199 GHC Proposal: https://github.com/ghc-proposals/ghc-proposals/pull/330 - - - - - 39f3d922 by Ben Gamari at 2024-03-07T19:40:46-05:00 testsuite/interface-stability: Update documentation - - - - - fdea7ada by Ben Gamari at 2024-03-07T19:40:46-05:00 ghc-internal: comment formatting - - - - - 4fba42ef by Ben Gamari at 2024-03-07T19:40:46-05:00 compiler: Default and warn ExceptionContext constraints - - - - - 3886a205 by Ben Gamari at 2024-03-07T19:40:46-05:00 base: Introduce exception backtraces Here we introduce the `Backtraces` type and associated machinery for attaching these via `ExceptionContext`. These has a few compile-time regressions (`T15703` and `T9872d`) due to the additional dependencies in the exception machinery. As well, there is a surprisingly large regression in the `size_hello_artifact` test. This appears to be due to various `Integer` and `Read` bits now being reachable at link-time. I believe it should be possible to avoid this but I have accepted the change for now to get the feature merged. CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/199 GHC Proposal: https://github.com/ghc-proposals/ghc-proposals/pull/330 Metric Increase: T15703 T9872d size_hello_artifact - - - - - 18c5409f by Ben Gamari at 2024-03-07T19:40:46-05:00 users guide: Release notes for exception backtrace work - - - - - f849c5fc by Ben Gamari at 2024-03-07T19:40:46-05:00 compiler: Don't show ExceptionContext of GhcExceptions Most GhcExceptions are user-facing errors and therefore the ExceptionContext has little value. Ideally we would enable it in the DEBUG compiler but I am leaving this for future work. - - - - - dc646e6f by Sylvain Henry at 2024-03-07T19:40:46-05:00 Disable T9930fail for the JS target (cf #19174) - - - - - bfc09760 by Alan Zimmerman at 2024-03-07T19:41:22-05:00 Update showAstData to honour blanking of AnnParen Also tweak rendering of SrcSpan to remove extra blank line. - - - - - 50454a29 by Ben Gamari at 2024-03-08T03:32:42-05:00 ghc-internal: Eliminate GHC.Internal.Data.Kind This was simply reexporting things from `ghc-prim`. Instead reexport these directly from `Data.Kind`. Also add build ordering dependency to work around #23942. - - - - - 38a4b6ab by Ben Gamari at 2024-03-08T03:33:18-05:00 rts: Fix SET_HDR initialization of retainer set This fixes a regression in retainer set profiling introduced by b0293f78cb6acf2540389e22bdda420d0ab874da. Prior to that commit the heap traversal word would be initialized by `SET_HDR` using `LDV_RECORD_CREATE`. However, the commit added a `doingLDVProfiling` check in `LDV_RECORD_CREATE`, meaning that this initialization no longer happened. Given that this initialization was awkwardly indirectly anyways, I have fixed this by explicitly initializating the heap traversal word to `NULL` in `SET_PROF_HDR`. This is equivalent to the previous behavior, but much more direct. Fixes #24513. - - - - - 635abccc by Ben Gamari at 2024-03-08T17:09:06-05:00 Bump ghc version to 9.10 - - - - - 2859a637 by Ben Gamari at 2024-03-08T18:26:47-05:00 base: Use strerror_r instead of strerror As noted by #24344, `strerror` is not necessarily thread-safe. Thankfully, POSIX.1-2001 has long offered `strerror_r`, which is safe to use. Fixes #24344. CLC discussion: https://github.com/haskell/core-libraries-committee/issues/249 - - - - - 5b934048 by Ben Gamari at 2024-03-08T18:50:12-05:00 Bump base upper bound - - - - - b30d134e by Ben Gamari at 2024-03-08T18:50:44-05:00 Testsuite output update - - - - - edb9bf77 by Jade at 2024-03-09T03:39:38-05:00 Error messages: Improve Error messages for Data constructors in type signatures. This patch improves the error messages from invalid type signatures by trying to guess what the user did and suggesting an appropriate fix. Partially fixes: #17879 - - - - - cfb197e3 by Patrick at 2024-03-09T03:40:15-05:00 HieAst: add module name #24493 The main purpose of this is to tuck the module name `xxx` in `module xxx where` into the hieAst. It should fix #24493. The following have been done: 1. Renamed and update the `tcg_doc_hdr :: Maybe (LHsDoc GhcRn)` to `tcg_hdr_info :: (Maybe (LHsDoc GhcRn), Maybe (XRec GhcRn ModuleName))` To store the located module name information. 2. update the `RenamedSource` and `RenamedStuff` with extra `Maybe (XRec GhcRn ModuleName)` located module name information. 3. add test `testsuite/tests/hiefile/should_compile/T24493.hs` to ensure the module name is added and update several relevent tests. 4. accompanied submodule haddoc test update MR in https://gitlab.haskell.org/ghc/haddock/-/merge_requests/53 - - - - - 2341d81e by Vaibhav Sagar at 2024-03-09T03:40:54-05:00 GHC.Utils.Binary: fix a couple of typos - - - - - 5580e1bd by Ben Gamari at 2024-03-09T03:41:30-05:00 rts: Drop .wasm suffix from .prof file names This replicates the behavior on Windows, where `Hi.exe` will produce profiling output named `Hi.prof` instead of `Hi.exe.prof`. While in the area I also fixed the extension-stripping logic, which incorrectly rewrote `Hi.exefoo` to `Hi.foo`. Closes #24515. - - - - - 259495ee by Cheng Shao at 2024-03-09T03:41:30-05:00 testsuite: drop exe extension from .hp & .prof filenames See #24515 for details. - - - - - c477a8d2 by Ben Gamari at 2024-03-09T03:42:05-05:00 rts/linker: Enable GOT support on all platforms There is nothing platform-dependent about our GOT implementation and GOT support is needed by `T24171` on i386. - - - - - 2e592857 by Vladislav Zavialov at 2024-03-09T03:42:41-05:00 Drop outdated comment on TcRnIllformedTypePattern This should have been done in 0f0c53a501b but I missed it. - - - - - c554b4da by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/CloneStack: Bounds check array write - - - - - 15c590a5 by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/CloneStack: Don't expose helper functions in header - - - - - e831ce31 by Ben Gamari at 2024-03-09T09:39:20-05:00 base: Move internals of GHC.InfoProv into GHC.InfoProv.Types Such that we can add new helpers into GHC.InfoProv.Types without breakage. - - - - - 6948e24d by Ben Gamari at 2024-03-09T09:39:20-05:00 rts: Lazily decode IPE tables Previously we would eagerly allocate `InfoTableEnt`s for each info table registered in the info table provenance map. However, this costs considerable memory and initialization time. Instead we now lazily decode these tables. This allows us to use one-third the memory *and* opens the door to taking advantage of sharing opportunities within a module. This required considerable reworking since lookupIPE now must be passed its result buffer. - - - - - 9204a04e by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/IPE: Don't expose helper in header - - - - - 308926ff by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/IPE: Share module_name within a Node This allows us to shave a 64-bit word off of the packed IPE entry size. - - - - - bebdea05 by Ben Gamari at 2024-03-09T09:39:20-05:00 IPE: Expose unit ID in InfoTableProv Here we add the unit ID to the info table provenance structure. - - - - - 6519c9ad by Ben Gamari at 2024-03-09T09:39:35-05:00 rts: Refactor GHC.Stack.CloneStack.decode Don't allocate a Ptr constructor per frame. - - - - - ed0b69dc by Ben Gamari at 2024-03-09T09:39:35-05:00 base: Do not expose whereFrom# from GHC.Exts - - - - - 2b1faea9 by Vladislav Zavialov at 2024-03-09T17:38:21-05:00 docs: Update info on TypeAbstractions * Mention TypeAbstractions in 9.10.1-notes.rst * Set the status to "Experimental". * Add a "Since: GHC 9.x" comment to each section. - - - - - f8b88918 by Ben Gamari at 2024-03-09T21:21:46-05:00 ci-images: Bump Alpine image to bootstrap with 9.8.2 - - - - - 705e6927 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Mark T24171 as fragile due to #24512 I will fix this but not in time for 9.10.1-alpha1 - - - - - c74196e1 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Mark linker_unload_native as fragile In particular this fails on platforms without `dlinfo`. I plan to address this but not before 9.10.1-alpha1. - - - - - f4d87f7a by Ben Gamari at 2024-03-09T21:21:46-05:00 configure: Bump version to 9.10 - - - - - 88df9a5f by Ben Gamari at 2024-03-09T21:21:46-05:00 Bump transformers submodule to 0.6.1.1 - - - - - 8176d5e8 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Increase ulimit for T18623 1 MByte was just too tight and failed intermittently on some platforms (e.g. CentOS 7). Bumping the limit to 8 MByte should provide sufficient headroom. Fixes #23139. - - - - - c74b38a3 by Ben Gamari at 2024-03-09T21:21:46-05:00 base: Bump version to 4.20.0.0 - - - - - b2937fc3 by Ben Gamari at 2024-03-09T21:21:46-05:00 ghc-internal: Set initial version at 9.1001.0 This provides PVP compliance while maintaining a clear correspondence between GHC releases and `ghc-internal` versions. - - - - - 4ae7d868 by Ben Gamari at 2024-03-09T21:21:46-05:00 ghc-prim: Bump version to 0.11.0 - - - - - 50798dc6 by Ben Gamari at 2024-03-09T21:21:46-05:00 template-haskell: Bump version to 2.22.0.0 - - - - - 8564f976 by Ben Gamari at 2024-03-09T21:21:46-05:00 base-exports: Accommodate spurious whitespace changes in 32-bit output It appears that this was - - - - - 9d4f0e98 by Ben Gamari at 2024-03-09T21:21:46-05:00 users-guide: Move exception backtrace relnotes to 9.10 This was previously mistakenly added to the GHC 9.8 release notes. - - - - - 145eae60 by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/rel_eng: Fix name of Rocky8 artifact - - - - - 39c2a630 by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/rel_eng: Fix path of generate_jobs_metadata - - - - - aed034de by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/upload: Rework recompression The old `combine` approach was quite fragile due to use of filename globbing. Moreover, it didn't parallelize well. This refactoring makes the goal more obvious, parallelizes better, and is more robust. - - - - - 9bdf3586 by Ben Gamari at 2024-03-09T21:37:44-05:00 Merge branch 'ghc-9.10' into ghc-head - - - - - cec76981 by Ben Gamari at 2024-03-09T21:54:00-05:00 Bump GHC version to 9.11 - - - - - 4c59feb7 by Ben Gamari at 2024-03-09T22:15:01-05:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - dc207d06 by Ben Gamari at 2024-03-10T08:56:08-04:00 configure: Bump GHC version to 9.11 Bumps haddock submodule. - - - - - 8b2513e8 by Ben Gamari at 2024-03-11T01:20:03-04:00 rts/linker: Don't unload code when profiling is enabled The heap census may contain references (e.g. `Counter.identity`) to static data which must be available when the census is reported at the end of execution. Fixes #24512. - - - - - 7810b4c3 by Ben Gamari at 2024-03-11T01:20:03-04:00 rts/linker: Don't unload native objects when dlinfo isn't available To do so is unsafe as we have no way of identifying references to symbols provided by the object. Fixes #24513. Fixes #23993. - - - - - 0590764c by Ben Gamari at 2024-03-11T01:20:39-04:00 rel_eng/upload: Purge both $rel_name/ and $ver/ This is necessary for prereleases, where GHCup accesses the release via `$ver/` - - - - - b85a4631 by Brandon Chinn at 2024-03-12T19:25:56-04:00 Remove duplicate code normalising slashes - - - - - c91946f9 by Brandon Chinn at 2024-03-12T19:25:56-04:00 Simplify regexes with raw strings - - - - - 1a5f53c6 by Brandon Chinn at 2024-03-12T19:25:57-04:00 Don't normalize backslashes in characters - - - - - 7ea971d3 by Andrei Borzenkov at 2024-03-12T19:26:32-04:00 Fix compiler crash caused by implicit RHS quantification in type synonyms (#24470) - - - - - 39f3ac3e by Cheng Shao at 2024-03-12T19:27:11-04:00 Revert "compiler: make genSym use C-based atomic increment on non-JS 32-bit platforms" This reverts commit 615eb855416ce536e02ed935ecc5a6f25519ae16. It was originally intended to fix #24449, but it was merely sweeping the bug under the rug. 3836a110577b5c9343915fd96c1b2c64217e0082 has properly fixed the fragile test, and we no longer need the C version of genSym. Furthermore, the C implementation causes trouble when compiling with clang that targets i386 due to alignment warning and libatomic linking issue, so it makes sense to revert it. - - - - - e6bfb85c by Cheng Shao at 2024-03-12T19:27:11-04:00 compiler: fix out-of-bound memory access of genSym on 32-bit This commit fixes an unnoticed out-of-bound memory access of genSym on 32-bit. ghc_unique_inc is 32-bit sized/aligned on 32-bit platforms, but we mistakenly treat it as a Word64 pointer in genSym, and therefore will accidentally load 2 garbage higher bytes, or with a small but non-zero chance, overwrite something else in the data section depends on how the linker places the data segments. This regression was introduced in !11802 and fixed here. - - - - - 77171cd1 by Ben Orchard at 2024-03-14T09:00:40-04:00 Note mutability of array and address access primops Without an understanding of immutable vs. mutable memory, the index primop family have a potentially non-intuitive type signature: indexOffAddr :: Addr# -> Int# -> a readOffAddr :: Addr# -> Int# -> State# d -> (# State# d, a #) indexOffAddr# might seem like a free generality improvement, which it certainly is not! This change adds a brief note on mutability expectations for most index/read/write access primops. - - - - - 7da7f8f6 by Alan Zimmerman at 2024-03-14T09:01:15-04:00 EPA: Fix regression discarding comments in contexts Closes #24533 - - - - - 73be65ab by Fendor at 2024-03-19T01:42:53-04:00 Fix sharing of 'IfaceTyConInfo' during core to iface type translation During heap analysis, we noticed that during generation of 'mi_extra_decls' we have lots of duplicates for the instances: * `IfaceTyConInfo NotPromoted IfaceNormalTyCon` * `IfaceTyConInfo IsPromoted IfaceNormalTyCon` which should be shared instead of duplicated. This duplication increased the number of live bytes by around 200MB while loading the agda codebase into GHCi. These instances are created during `CoreToIface` translation, in particular `toIfaceTyCon`. The generated core looks like: toIfaceTyCon = \ tc_sjJw -> case $wtoIfaceTyCon tc_sjJw of { (# ww_sjJz, ww1_sjNL, ww2_sjNM #) -> IfaceTyCon ww_sjJz (IfaceTyConInfo ww1_sjNL ww2_sjNM) } whichs removes causes the sharing to work propery. Adding explicit sharing, with NOINLINE annotations, changes the core to: toIfaceTyCon = \ tc_sjJq -> case $wtoIfaceTyCon tc_sjJq of { (# ww_sjNB, ww1_sjNC #) -> IfaceTyCon ww_sjNB ww1_sjNC } which looks much more like sharing is happening. We confirmed via ghc-debug that all duplications were eliminated and the number of live bytes are noticeably reduced. - - - - - bd8209eb by Alan Zimmerman at 2024-03-19T01:43:28-04:00 EPA: Address more 9.10.1-alpha1 regressions from recent changes Closes #24533 Hopefully for good this time - - - - - 31bf85ee by Fendor at 2024-03-19T14:48:08-04:00 Escape multiple arguments in the settings file Uses responseFile syntax. The issue arises when GHC is installed on windows into a location that has a space, for example the user name is 'Fake User'. The $topdir will also contain a space, consequentially. When we resolve the top dir in the string `-I$topdir/mingw/include`, then `words` will turn this single argument into `-I/C/Users/Fake` and `User/.../mingw/include` which trips up the flag argument parser of various tools such as gcc or clang. We avoid this by escaping the $topdir before replacing it in `initSettngs`. Additionally, we allow to escape spaces and quotation marks for arguments in `settings` file. Add regression test case to count the number of options after variable expansion and argument escaping took place. Additionally, we check that escaped spaces and double quotation marks are correctly parsed. - - - - - f45f700e by Matthew Pickering at 2024-03-19T14:48:44-04:00 Read global package database from settings file Before this patch, the global package database was always assumed to be in libdir </> package.conf.d. This causes issues in GHC's build system because there are sometimes situations where the package database you need to use is not located in the same place as the settings file. * The stage1 compiler needs to use stage1 libraries, so we should set "Global Package DB" for the stage1 compiler to the stage1 package database. * Stage 2 cross compilers need to use stage2 libraries, so likewise, we should set the package database path to `_build/stage2/lib/` * The normal situation is where the stage2 compiler uses stage1 libraries. Then everything lines up. * When installing we have rearranged everything so that the settings file and package database line up properly, so then everything should continue to work as before. In this case we set the relative package db path to `package.conf.d`, so it resolves the same as before. * ghc-pkg needs to be modified as well to look in the settings file fo the package database rather than assuming the global package database location relative to the lib folder. * Cabal/cabal-install will work correctly because they query the global package database using `--print-global-package-db`. A reasonable question is why not generate the "right" settings files in the right places in GHC's build system. In order to do this you would need to engineer wrappers for all executables to point to a specific libdir. There are also situations where the same package db is used by two different compilers with two different settings files (think stage2 cross compiler and stage3 compiler). In short, this 10 line patch allows for some reasonable simplifications in Hadrian at very little cost to anything else. Fixes #24502 - - - - - 4c8f1794 by Matthew Pickering at 2024-03-19T14:48:44-04:00 hadrian: Remove stage1 testsuite wrappers logic Now instead of producing wrappers which pass the global package database argument to ghc and ghc-pkg, we write the location of the correct package database into the settings file so you can just use the intree compiler directly. - - - - - da0d8ba5 by Matthew Craven at 2024-03-19T14:49:20-04:00 Remove unused ghc-internal module "GHC.Internal.Constants" - - - - - b56d2761 by Matthew Craven at 2024-03-19T14:49:20-04:00 CorePrep: Rework lowering of BigNat# literals Don't use bigNatFromWord#, because that's terrible: * We shouldn't have to traverse a linked list at run-time to build a BigNat# literal. That's just silly! * The static List object we have to create is much larger than the actual BigNat#'s contents, bloating code size. * We have to read the corresponding interface file, which causes un-tracked implicit dependencies. (#23942) Instead, encode them into the appropriate platform-dependent sequence of bytes, and generate code that copies these bytes at run-time from an Addr# literal into a new ByteArray#. A ByteArray# literal would be the correct thing to generate, but these are not yet supported; see also #17747. Somewhat surprisingly, this change results in a slight reduction in compiler allocations, averaging around 0.5% on ghc's compiler performance tests, including when compiling programs that contain no bignum literals to begin with. The specific cause of this has not been investigated. Since this lowering no longer reads the interface file for GHC.Num.BigNat, the reasoning in Note [Depend on GHC.Num.Integer] is obsoleted. But the story of un-tracked built-in dependencies remains complex, and Note [Tracking dependencies on primitives] now exists to explain this complexity. Additionally, many empty imports have been modified to refer to this new note and comply with its guidance. Several empty imports necessary for other reasons have also been given brief explanations. Metric Decrease: MultiLayerModulesTH_OneShot - - - - - 349ea330 by Fendor at 2024-03-19T14:50:00-04:00 Eliminate thunk in 'IfaceTyCon' Heap analysis showed that `IfaceTyCon` retains a thunk to `IfaceTyConInfo`, defeating the sharing of the most common instances of `IfaceTyConInfo`. We make sure the indirection is removed by adding bang patterns to `IfaceTyCon`. Experimental results on the agda code base, where the `mi_extra_decls` were read from disk: Before this change, we observe around 8654045 instances of: `IfaceTyCon[Name,THUNK_1_0]` But these thunks almost exclusively point to a shared value! Forcing the thunk a little bit more, leads to `ghc-debug` reporting: `IfaceTyCon[Name:Name,IfaceTyConInfo]` and a noticeable reduction of live bytes (on agda ~10%). - - - - - 594bee0b by Krzysztof Gogolewski at 2024-03-19T14:50:36-04:00 Minor misc cleanups - GHC.HsToCore.Foreign.JavaScript: remove dropRuntimeRepArgs; boxed tuples don't take RuntimeRep args - GHC.HsToCore.Foreign.Call: avoid partial pattern matching - GHC.Stg.Unarise: strengthen the assertion; we can assert that non-rubbish literals are unary rather than just non-void - GHC.Tc.Gen.HsType: make sure the fsLit "literal" rule fires - users_guide/using-warnings.rst: remove -Wforall-identifier, now deprecated and does nothing - users_guide/using.rst: fix formatting - andy_cherry/test.T: remove expect_broken_for(23272...), 23272 is fixed The rest are simple cleanups. - - - - - cf55a54b by Ben Gamari at 2024-03-19T14:51:12-04:00 mk/relpath: Fix quoting Previously there were two instances in this script which lacked proper quoting. This resulted in `relpath` invocations in the binary distribution Makefile producing incorrect results on Windows, leading to confusing failures from `sed` and the production of empty package registrations. Fixes #24538. - - - - - 5ff88389 by Bryan Richter at 2024-03-19T14:51:48-04:00 testsuite: Disable T21336a on wasm - - - - - 60023351 by Ben Gamari at 2024-03-19T22:33:10-04:00 hadrian/bindist: Eliminate extraneous `dirname` invocation Previously we would call `dirname` twice per installed library file. We now instead reuse this result. This helps appreciably on Windows, where processes are quite expensive. - - - - - 616ac300 by Ben Gamari at 2024-03-19T22:33:10-04:00 hadrian: Package mingw toolchain in expected location This fixes #24525, a regression due to 41cbaf44a6ab5eb9fa676d65d32df8377898dc89. Specifically, GHC expects to find the mingw32 toolchain in the binary distribution root. However, after this patch it was packaged in the `lib/` directory. - - - - - de9daade by Ben Gamari at 2024-03-19T22:33:11-04:00 gitlab/rel_eng: More upload.sh tweaks - - - - - 1dfe12db by Ben Gamari at 2024-03-19T22:33:11-04:00 rel_eng: Drop dead prepare_docs codepath - - - - - dd2d748b by Ben Gamari at 2024-03-19T22:33:11-04:00 rel_env/recompress_all: unxz before recompressing Previously we would rather compress the xz *again*, before in addition compressing it with the desired scheme. Fixes #24545. - - - - - 9d936c57 by Ben Gamari at 2024-03-19T22:33:11-04:00 mk-ghcup-metadata: Fix directory of testsuite tarball As reported in #24546, the `dlTest` artifact should be extracted into the `testsuite` directory. - - - - - 6d398066 by Ben Gamari at 2024-03-19T22:33:11-04:00 ghcup-metadata: Don't populate dlOutput unless necessary ghcup can apparently infer the output name of an artifact from its URL. Consequently, we should only include the `dlOutput` field when it would differ from the filename of `dlUri`. Fixes #24547. - - - - - 576f8b7e by Zubin Duggal at 2024-03-19T22:33:46-04:00 Revert "Apply shellcheck suggestion to SUBST_TOOLDIR" This reverts commit c82770f57977a2b5add6e1378f234f8dd6153392. The shellcheck suggestion is spurious and results in SUBST_TOOLDIR being a no-op. `set` sets positional arguments for bash, but we want to set the variable given as the first autoconf argument. Fixes #24542 Metric decreases because the paths in the settings file are now shorter, so we allocate less when we read the settings file. ------------------------- Metric Decrease: T12425 T13035 T9198 ------------------------- - - - - - cdfe6e01 by Fendor at 2024-03-19T22:34:22-04:00 Compact serialisation of IfaceAppArgs In #24563, we identified that IfaceAppArgs serialisation tags each cons cell element with a discriminator byte. These bytes add up quickly, blowing up interface files considerably when '-fwrite-if-simplified-core' is enabled. We compact the serialisation by writing out the length of 'IfaceAppArgs', followed by serialising the elements directly without any discriminator byte. This improvement can decrease the size of some interface files by up to 35%. - - - - - 97a2bb1c by Simon Peyton Jones at 2024-03-20T17:11:29+00:00 Expand untyped splices in tcPolyExprCheck Fixes #24559 - - - - - 5f275176 by Alan Zimmerman at 2024-03-20T22:44:12-04:00 EPA: Clean up Exactprint helper functions a bit - Introduce a helper lens to compose on `EpAnn a` vs `a` versions - Rename some prime versions of functions back to non-prime They were renamed during the rework - - - - - da2a10ce by Vladislav Zavialov at 2024-03-20T22:44:48-04:00 Type operators in promoteOccName (#24570) Type operators differ from term operators in that they are lexically classified as (type) constructors, not as (type) variables. Prior to this change, promoteOccName did not account for this difference, causing a scoping issue that affected RequiredTypeArguments. type (!@#) = Bool f = idee (!@#) -- Not in scope: ‘!@#’ (BUG) Now we have a special case in promoteOccName to account for this. - - - - - 247fc0fa by Preetham Gujjula at 2024-03-21T10:19:18-04:00 docs: Remove mention of non-existent Ord instance for Complex The documentation for Data.Complex says that the Ord instance for Complex Float is deficient, but there is no Ord instance for Complex a. The Eq instance for Complex Float is similarly deficient, so we use that as an example instead. - - - - - 6fafc51e by Andrei Borzenkov at 2024-03-21T10:19:54-04:00 Fix TH handling in `pat_to_type_pat` function (#24571) There was missing case for `SplicePat` in `pat_to_type_at` function, hence patterns with splicing that checked against `forall->` doesn't work properly because they fall into the "illegal pattern" case. Code example that is now accepted: g :: forall a -> () g $([p| a |]) = () - - - - - 52072f8e by Sylvain Henry at 2024-03-21T21:01:59-04:00 Type-check default declarations before deriving clauses (#24566) See added Note and #24566. Default declarations must be type-checked before deriving clauses. - - - - - 7dfdf3d9 by Sylvain Henry at 2024-03-21T21:02:40-04:00 Lexer: small perf changes - Use unsafeChr because we know our values to be valid - Remove some unnecessary use of `ord` (return Word8 values directly) - - - - - 864922ef by Sylvain Henry at 2024-03-21T21:02:40-04:00 JS: fix some comments - - - - - 3e0b2b1f by Sebastian Graf at 2024-03-21T21:03:16-04:00 Simplifier: Re-do dependency analysis in abstractFloats (#24551) In #24551, we abstracted a string literal binding over a type variable, triggering a CoreLint error when that binding floated to top-level. The solution implemented in this patch fixes this by re-doing dependency analysis on a simplified recursive let binding that is about to be type abstracted, in order to find the minimal set of type variables to abstract over. See wrinkle (AB5) of Note [Floating and type abstraction] for more details. Fixes #24551 - - - - - 8a8ac65a by Matthew Craven at 2024-03-23T00:20:52-04:00 Improve toInteger @Word32 on 64-bit platforms On 64-bit platforms, every Word32 fits in an Int, so we can convert to Int# without having to perform the overflow check integerFromWord# uses internally. - - - - - 0c48f2b9 by Apoorv Ingle at 2024-03-23T00:21:28-04:00 Fix for #24552 (see testcase T24552) Fixes for a bug in desugaring pattern synonyms matches, introduced while working on on expanding `do`-blocks in #18324 The `matchWrapper` unecessarily (and incorrectly) filtered out the default wild patterns in a match. Now the wild pattern alternative is simply ignored by the pm check as its origin is `Generated`. The current code now matches the expected semantics according to the language spec. - - - - - b72705e9 by Simon Peyton Jones at 2024-03-23T00:22:04-04:00 Print more info about kinds in error messages This fixes #24553, where GHC unhelpfully said error: [GHC-83865] • Expected kind ‘* -> * -> *’, but ‘Foo’ has kind ‘* -> * -> *’ See Note [Showing invisible bits of types in error messages] - - - - - 8f7cfc7e by Tristan Cacqueray at 2024-03-23T00:22:44-04:00 docs: remove the don't use float hint This hint is outdated, ``Complex Float`` are now specialised, and the heap space suggestion needs more nuance so it should be explained in the unboxed/storable array documentation. - - - - - 5bd8ed53 by Andreas Klebinger at 2024-03-23T16:18:33-04:00 NCG: Fix a bug in jump shortcutting. When checking if a jump has more than one destination account for the possibility of some jumps not being representable by a BlockId. We do so by having isJumpishInstr return a `Maybe BlockId` where Nothing represents non-BlockId jump destinations. Fixes #24507 - - - - - 8d67f247 by Ben Gamari at 2024-03-23T16:19:09-04:00 docs: Drop old release notes, add for 9.12.1 - - - - - 7db8c992 by Cheng Shao at 2024-03-25T13:45:46-04:00 rts: fix clang compilation on aarch64 This patch fixes function prototypes in ARMOutlineAtomicsSymbols.h which causes "error: address argument to atomic operation must be a pointer to _Atomic type" when compiling with clang on aarch64. - - - - - 237194ce by Sylvain Henry at 2024-03-25T13:46:27-04:00 Lexer: fix imports for Alex 3.5.1 (#24583) - - - - - 810660b7 by Cheng Shao at 2024-03-25T22:19:16-04:00 libffi-tarballs: bump libffi-tarballs submodule to libffi 3.4.6 This commit bumps the libffi-tarballs submodule to libffi 3.4.6, which includes numerous upstream libffi fixes, especially https://github.com/libffi/libffi/issues/760. - - - - - d2ba41e8 by Alan Zimmerman at 2024-03-25T22:19:51-04:00 EPA: do not duplicate comments in signature RHS - - - - - 32a8103f by Rodrigo Mesquita at 2024-03-26T21:16:12-04:00 configure: Use LDFLAGS when trying linkers A user may configure `LDFLAGS` but not `LD`. When choosing a linker, we will prefer `ldd`, then `ld.gold`, then `ld.bfd` -- however, we have to check for a working linker. If either of these fail, we try the next in line. However, we were not considering the `$LDFLAGS` when checking if these linkers worked. So we would pick a linker that does not support the current $LDFLAGS and fail further down the line when we used that linker with those flags. Fixes #24565, where `LDFLAGS=-Wl,-z,pack-relative-relocs` is not supported by `ld.gold` but that was being picked still. - - - - - bf65a7c3 by Rodrigo Mesquita at 2024-03-26T21:16:48-04:00 bindist: Clean xattrs of bin and lib at configure time For issue #21506, we started cleaning the extended attributes of binaries and libraries from the bindist *after* they were installed to workaround notarisation (#17418), as part of `make install`. However, the `ghc-toolchain` binary that is now shipped with the bindist must be run at `./configure` time. Since we only cleaned the xattributes of the binaries and libs after they were installed, in some situations users would be unable to run `ghc-toolchain` from the bindist, failing at configure time (#24554). In this commit we move the xattr cleaning logic to the configure script. Fixes #24554 - - - - - cfeb70d3 by Rodrigo Mesquita at 2024-03-26T21:17:24-04:00 Revert "NCG: Fix a bug in jump shortcutting." This reverts commit 5bd8ed53dcefe10b72acb5729789e19ceb22df66. Fixes #24586 - - - - - 13223f6d by Serge S. Gulin at 2024-03-27T07:28:51-04:00 JS: `h$rts_isProfiled` is removed from `profiling` and left its version at `rts/js/config.js` - - - - - 0acfe391 by Alan Zimmerman at 2024-03-27T07:29:27-04:00 EPA: Do not extend declaration range for trailine zero len semi The lexer inserts virtual semicolons having zero width. Do not use them to extend the list span of items in a list. - - - - - cd0fb82f by Alan Zimmerman at 2024-03-27T19:33:08+00:00 EPA: Fix FamDecl range The span was incorrect if opt_datafam_kind_sig was empty - - - - - f8f384a8 by Ben Gamari at 2024-03-29T01:23:03-04:00 Fix type of _get_osfhandle foreign import Fixes #24601. - - - - - 00d3ecf0 by Alan Zimmerman at 2024-03-29T12:19:10+00:00 EPA: Extend StringLiteral range to include trailing commas This goes slightly against the exact printing philosophy where trailing decorations should be in an annotation, but the practicalities of adding it to the WarningTxt environment, and the problems caused by deviating do not make a more principles approach worthwhile. - - - - - efab3649 by brandon s allbery kf8nh at 2024-03-31T20:04:01-04:00 clarify Note [Preproccesing invocations] - - - - - c8a4c050 by Ben Gamari at 2024-04-02T12:50:35-04:00 rts: Fix TSAN_ENABLED CPP guard This should be `#if defined(TSAN_ENABLED)`, not `#if TSAN_ENABLED`, lest we suffer warnings. - - - - - e91dad93 by Cheng Shao at 2024-04-02T12:50:35-04:00 rts: fix errors when compiling with TSAN This commit fixes rts compilation errors when compiling with TSAN: - xxx_FENCE macros are redefined and trigger CPP warnings. - Use SIZEOF_W. WORD_SIZE_IN_BITS is provided by MachDeps.h which Cmm.h doesn't include by default. - - - - - a9ab9455 by Cheng Shao at 2024-04-02T12:50:35-04:00 rts: fix clang-specific errors when compiling with TSAN This commit fixes clang-specific rts compilation errors when compiling with TSAN: - clang doesn't have -Wtsan flag - Fix prototype of ghc_tsan_* helper functions - __tsan_atomic_* functions aren't clang built-ins and sanitizer/tsan_interface_atomic.h needs to be included - On macOS, TSAN runtime library is libclang_rt.tsan_osx_dynamic.dylib, not libtsan. -fsanitize-thread as a link-time flag will take care of linking the TSAN runtime library anyway so remove tsan as an rts extra library - - - - - 865bd717 by Cheng Shao at 2024-04-02T12:50:35-04:00 compiler: fix github link to __tsan_memory_order in a comment - - - - - 07cb627c by Cheng Shao at 2024-04-02T12:50:35-04:00 ci: improve TSAN CI jobs - Run TSAN jobs with +thread_sanitizer_cmm which enables Cmm instrumentation as well. - Run TSAN jobs in deb12 which ships gcc-12, a reasonably recent gcc that @bgamari confirms he's using in #GHC:matrix.org. Ideally we should be using latest clang release for latest improvements in sanitizers, though that's left as future work. - Mark TSAN jobs as manual+allow_failure in validate pipelines. The purpose is to demonstrate that we have indeed at least fixed building of TSAN mode in CI without blocking the patch to land, and once merged other people can begin playing with TSAN using their own dev setups and feature branches. - - - - - a1c18c7b by Andrei Borzenkov at 2024-04-02T12:51:11-04:00 Merge tc_infer_hs_type and tc_hs_type into one function using ExpType philosophy (#24299, #23639) This patch implements refactoring which is a prerequisite to updating kind checking of type patterns. This is a huge simplification of the main worker that checks kind of HsType. It also fixes the issues caused by previous code duplication, e.g. that we didn't add module finalizers from splices in inference mode. - - - - - 817e8936 by Rodrigo Mesquita at 2024-04-02T20:13:05-04:00 th: Hide the Language.Haskell.TH.Lib.Internal module from haddock Fixes #24562 - - - - - b36ee57b by Sylvain Henry at 2024-04-02T20:13:46-04:00 JS: reenable h$appendToHsString optimization (#24495) The optimization introducing h$appendToHsString wasn't kicking in anymore (while it did in 9.8.1) because of the changes introduced in #23270 (7e0c8b3bab30). This patch reenables the optimization by matching on case-expression, as done in Cmm for unpackCString# standard thunks. The test is also T24495 added in the next commits (two commits for ease of backporting to 9.8). - - - - - 527616e9 by Sylvain Henry at 2024-04-02T20:13:46-04:00 JS: fix h$appendToHsString implementation (#24495) h$appendToHsString needs to wrap its argument in an updatable thunk to behave like unpackAppendCString#. Otherwise if a SingleEntry thunk is passed, it is stored as-is in a CONS cell, making the resulting list impossible to deepseq (forcing the thunk doesn't update the contents of the CONS cell)! The added test checks that the optimization kicks in and that h$appendToHsString works as intended. Fix #24495 - - - - - faa30b41 by Simon Peyton Jones at 2024-04-02T20:14:22-04:00 Deal with duplicate tyvars in type declarations GHC was outright crashing before this fix: #24604 - - - - - e0b0c717 by Simon Peyton Jones at 2024-04-02T20:14:58-04:00 Try using MCoercion in exprIsConApp_maybe This is just a simple refactor that makes exprIsConApp_maybe a little bit more direct, simple, and efficient. Metrics: compile_time/bytes allocated geo. mean -0.1% minimum -2.0% maximum -0.0% Not a big gain, but worthwhile given that the code is, if anything, easier to grok. - - - - - 15f4d867 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Initial ./configure support for selecting I/O managers In this patch we just define new CPP vars, but don't yet use them or replace the existing approach. That will follow. The intention here is that every I/O manager can be enabled/disabled at GHC build time (subject to some constraints). More than one I/O manager can be enabled to be built. At least one I/O manager supporting the non-threaded RTS must be enabled as well as at least one supporting the non-threaded RTS. The I/O managers enabled here will become the choices available at runtime at RTS startup (in later patches). The choice can be made with RTS flags. There are separate sets of choices for the threaded and non-threaded RTS ways, because most I/O managers are specific to these ways. Furthermore we must establish a default I/O manager for the threaded and non-threaded RTS. Most I/O managers are platform-specific so there are checks to ensure each one can be enabled on the platform. Such checks are also where (in future) any system dependencies (e.g. libraries) can be checked. The output is a set of CPP flags (in the mk/config.h file), with one flag per named I/O manager: * IOMGR_BUILD_<name> : which ones should be built (some) * IOMGR_DEFAULT_NON_THREADED_<name> : which one is default (exactly one) * IOMGR_DEFAULT_THREADED_<name> : which one is default (exactly one) and a set of derived flags in IOManager.h * IOMGR_ENABLED_<name> : enabled for the current RTS way Note that IOMGR_BUILD_<name> just says that an I/O manager will be built for _some_ RTS way (i.e. threaded or non-threaded). The derived flags IOMGR_ENABLED_<name> in IOManager.h say if each I/O manager is enabled in the "current" RTS way. These are the ones that can be used for conditional compilation of the I/O manager code. Co-authored-by: Pi Delport <pi at well-typed.com> - - - - - 85b0f87a by Duncan Coutts at 2024-04-03T01:27:17-04:00 Change the handling of the RTS flag --io-manager= Now instead of it being just used on Windows to select between the WinIO vs the MIO or Win32-legacy I/O managers, it is now used on all platforms for selecting the I/O manager to use. Right now it remains the case that there is only an actual choice on Windows, but that will change later. Document the --io-manager flag in the user guide. This change is also reflected in the RTS flags types in the base library. Deprecate the export of IoSubSystem from GHC.RTS.Flags with a message to import it from GHC.IO.Subsystem. The way the 'IoSubSystem' is detected also changes. Instead of looking at the RTS flag, there is now a C bool global var in the RTS which gets set on startup when the I/O manager is selected. This bool var says whether the selected I/O manager classifies as "native" on Windows, which in practice means the WinIO I/O manager has been selected. Similarly, the is_io_mng_native_p RTS helper function is re-implemented in terms of the selected I/O manager, rather than based on the RTS flags. We do however remove the ./configure --native-io-manager flag because we're bringing the WinIO/MIO/Win32-legacy choice under the new general scheme for selecting I/O managers, and that new scheme involves no ./configure time user choices, just runtime RTS flag choices. - - - - - 1a8f020f by Duncan Coutts at 2024-04-03T01:27:17-04:00 Convert {init,stop,exit}IOManager to switch style Rather than ad-hoc cpp conitionals on THREADED_RTS and mingw32_HOST_OS, we use a style where we switch on the I/O manager impl, with cases for each I/O manager impl. - - - - - a5bad3d2 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Split up the CapIOManager content by I/O manager Using the new IOMGR_ENABLED_<name> CPP defines. - - - - - 1d36e609 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Convert initIOManagerAfterFork and wakeupIOManager to switch style - - - - - c2f26f36 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move most of waitRead#/Write# from cmm to C Moves it into the IOManager.c where we can follow the new pattern of switching on the selected I/O manager. - - - - - 457705a8 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move most of the delay# impl from cmm to C Moves it into the IOManager.c where we can follow the new pattern of switching on the selected I/O manager. Uses a new IOManager API: syncDelay, following the naming convention of sync* for thread-synchronous I/O & timer/delay operations. As part of porting from cmm to C, we maintain the rule that the why_blocked gets accessed using load acquire and store release atomic memory operations. There was one exception to this rule: in the delay# primop cmm code on posix (not win32), the why_blocked was being updated using a store relaxed, not a store release. I've no idea why. In this convesion I'm playing it safe here and using store release consistently. - - - - - e93058e0 by Duncan Coutts at 2024-04-03T01:27:18-04:00 insertIntoSleepingQueue is no longer public No longer defined in IOManager.h, just a private function in IOManager.c. Since it is no longer called from cmm code, just from syncDelay. It ought to get moved further into the select() I/O manager impl, rather than living in IOManager.c. On the other hand appendToIOBlockedQueue is still called from cmm code in the win32-legacy I/O manager primops async{Read,Write}#, and it is also used by the select() I/O manager. Update the CPP and comments to reflect this. - - - - - 60ce9910 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move anyPendingTimeoutsOrIO impl from .h to .c The implementation is eventually going to need to use more private things, which will drag in unwanted includes into IOManager.h, so it's better to move the impl out of the header file and into the .c file, at the slight cost of it no longer being inline. At the same time, change to the "switch (iomgr_type)" style. - - - - - f70b8108 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Take a simpler approach to gcc warnings in IOManager.c We have lots of functions with conditional implementations for different I/O managers. Some functions, for some I/O managers, naturally have implementations that do nothing or barf. When only one such I/O manager is enabled then the whole function implementation will have an implementation that does nothing or barfs. This then results in warnings from gcc that parameters are unused, or that the function should be marked with attribute noreturn (since barf does not return). The USED_IF_THREADS trick for fine-grained warning supression is fine for just two cases, but an equivalent here would need USED_IF_THE_ONLY_ENABLED_IOMGR_IS_X_OR_Y which would have combinitorial blowup. So we take a coarse grained approach and simply disable these two warnings for the whole file. So we use a GCC pragma, with its handy push/pop support: #pragma GCC diagnostic push #pragma GCC diagnostic ignored "-Wsuggest-attribute=noreturn" #pragma GCC diagnostic ignored "-Wunused-parameter" ... #pragma GCC diagnostic pop - - - - - b48805b9 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add a new trace class for the iomanager It makes sense now for it to be separate from the scheduler class of tracers. Enabled with +RTS -Do. Document the -Do debug flag in the user guide. - - - - - f0c1f862 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Have the throwTo impl go via (new) IOManager APIs rather than directly operating on the IO manager's data structures. Specifically, when thowing an async exception to a thread that is blocked waiting for I/O or waiting for a timer, then we want to cancel that I/O waiting or cancel the timer. Currently this is done directly in removeFromQueues() in RaiseAsync.c. We want it to go via proper APIs both for modularity but also to let us support multiple I/O managers. So add sync{IO,Delay}Cancel, which is the cancellation for the corresponding sync{IO,Delay}. The implementations of these use the usual "switch (iomgr_type)" style. - - - - - 4f9e9c4e by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move awaitEvent into a proper IOManager API and have the scheduler use it. Previously the scheduler calls awaitEvent directly, and awaitEvent is implemented directly in the RTS I/O managers (select, win32). This relies on the old scheme where there's a single active I/O manager for each platform and RTS way. We want to move that to go via an API in IOManager.{h,c} which can then call out to the active I/O manager. Also take the opportunity to split awaitEvent into two. The existing awaitEvent has a bool wait parameter, to say if the call should be blocking or non-blocking. We split this into two separate functions: pollCompletedTimeoutsOrIO and awaitCompletedTimeoutsOrIO. We split them for a few reasons: they have different post-conditions (specifically the await version is supposed to guarantee that there are threads runnable when it completes). Secondly, it is also anticipated that in future I/O managers the implementations of the two cases will be simpler if they are separated. - - - - - 5ad4b30f by Duncan Coutts at 2024-04-03T01:27:18-04:00 Rename awaitEvent in select and win32 I/O managers These are now just called from IOManager.c and are the per-I/O manager backend impls (whereas previously awaitEvent was the entry point). Follow the new naming convention in the IOManager.{h,c} of awaitCompletedTimeoutsOrIO, with the I/O manager's name as a suffix: so awaitCompletedTimeoutsOrIO{Select,Win32}. - - - - - d30c6bc6 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Tidy up a couple things in Select.{h,c} Use the standard #include {Begin,End}Private.h style rather than RTS_PRIVATE on individual decls. And conditionally build the code for the select I/O manager based on the new CPP IOMGR_ENABLED_SELECT rather than on THREADED_RTS. - - - - - 4161f516 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add an IOManager API for scavenging TSO blocked_info When the GC scavenges a TSO it needs to scavenge the tso->blocked_info but the blocked_info is a big union and what lives there depends on the two->why_blocked, which for I/O-related reasons is something that in principle is the responsibility of the I/O manager and not the GC. So the right thing to do is for the GC to ask the I/O manager to sscavenge the blocked_info if it encounters any I/O-related why_blocked reasons. So we add scavengeTSOIOManager in IOManager.{h,c} with the usual style. Now as it happens, right now, there is no special scavenging to do, so the implementation of scavengeTSOIOManager is a fancy no-op. That's because the select I/O manager uses only the fd and target members, which are not GC pointers, and the win32-legacy I/O manager _ought_ to be using GC-managed heap objects for the StgAsyncIOResult but it is actually usingthe C heap, so again no GC pointers. If the win32-legacy were doing this more sensibly, then scavengeTSOIOManager would be the right place to do the GC magic. Future I/O managers will need GC heap objects in the tso->blocked_info and will make use of this functionality. - - - - - 94a87d21 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add I/O manager API notifyIOManagerCapabilitiesChanged Used in setNumCapabilities. It only does anything for MIO on Posix. Previously it always invoked Haskell code, but that code only did anything on non-Windows (and non-JS), and only threaded. That currently effectively means the MIO I/O manager on Posix. So now it only invokes it for the MIO Posix case. - - - - - 3be6d591 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Select an I/O manager early in RTS startup We need to select the I/O manager to use during startup before the per-cap I/O manager initialisation. - - - - - aaa294d0 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Make struct CapIOManager be fully opaque Provide an opaque (forward) definition in Capability.h (since the cap contains a *CapIOManager) and then only provide a full definition in a new file IOManagerInternals.h. This new file is only supposed to be included by the IOManager implementation, not by its users. So that means IOManager.c and individual I/O manager implementations. The posix/Signals.c still needs direct access, but that should be eliminated. Anything that needs direct access either needs to be clearly part of an I/O manager (e.g. the sleect() one) or go via a proper API. - - - - - 877a2a80 by Duncan Coutts at 2024-04-03T01:27:18-04:00 The select() I/O manager does have some global initialisation It's just to make sure an exception CAF is a GC root. - - - - - 9c51473b by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add tracing for the main I/O manager actions Using the new tracer class. Note: The unconditional definition of showIOManager should be compatible with the debugTrace change in 7c7d1f6. Co-authored-by: Pi Delport <pi at well-typed.com> - - - - - c7d3e3a3 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Include the default I/O manager in the +RTS --info output Document the extra +RTS --info output in the user guide - - - - - 8023bad4 by Duncan Coutts at 2024-04-03T01:27:18-04:00 waitRead# / waitWrite# do not work for win32-legacy I/O manager Previously it was unclear that they did not work because the code path was shared with other I/O managers (in particular select()). Following the code carefully shows that what actually happens is that the calling thread would block forever: the thread will be put into the blocked queue, but no other action is scheduled that will ever result in it getting unblocked. It's better to just fail loudly in case anyone accidentally calls it, also it's less confusing code. - - - - - 83a74d20 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Conditionally ignore some GCC warnings Some GCC versions don't know about some warnings, and they complain that we're ignoring unknown warnings. So we try to ignore the warning based on the GCC version. - - - - - 1adc6fa4 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Accept changes to base-exports All the changes are in fact not changes at all. Previously, the IoSubSystem data type was defined in GHC.RTS.Flags and exported from both GHC.RTS.Flags and GHC.IO.SubSystem. Now, the data type is defined in GHC.IO.SubSystem and still exported from both modules. Therefore, the same exports and same instances are still available from both modules. But the base-exports records only the defining module, and so it looks like a change when it is fully compatible. Related: we do add a deprecation to the export of the type via GHC.RTS.Flags, telling people to use the export from GHC.IO.SubSystem. Also the sort order for some unrelated Show instances changed. No idea why. The same changes apply in the other versions, with a few more changes due to sort order weirdness. - - - - - 8d950968 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Accept metric decrease in T12227 I can't think of any good reason that anything in this MR should have changed the number of allocations, up or down. (Yes this is an empty commit.) Metric Decrease: T12227 - - - - - e869605e by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Several improvements to the handling of coercions * Make `mkSymCo` and `mkInstCo` smarter Fixes #23642 * Fix return role of `SelCo` in the coercion optimiser. Fixes #23617 * Make the coercion optimiser `opt_trans_rule` work better for newtypes Fixes #23619 - - - - - 1efd0714 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 FloatOut: improve floating for join point See the new Note [Floating join point bindings]. * Completely get rid of the complicated join_ceiling nonsense, which I have never understood. * Do not float join points at all, except perhaps to top level. * Some refactoring around wantToFloat, to treat Rec and NonRec more uniformly - - - - - 9c00154d by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Improve eta-expansion through call stacks See Note [Eta expanding through CallStacks] in GHC.Core.Opt.Arity This is a one-line change, that fixes an inconsistency - || isCallStackPredTy ty + || isCallStackPredTy ty || isCallStackTy ty - - - - - 95a9a172 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Spelling, layout, pretty-printing only - - - - - bdf1660f by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Improve exprIsConApp_maybe a little Eliminate a redundant case at birth. This sometimes reduces Simplifier iterations. See Note [Case elim in exprIsConApp_maybe]. - - - - - 609cd32c by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Inline GHC.HsToCore.Pmc.Solver.Types.trvVarInfo When exploring compile-time regressions after meddling with the Simplifier, I discovered that GHC.HsToCore.Pmc.Solver.Types.trvVarInfo was very delicately balanced. It's a small, heavily used, overloaded function and it's important that it inlines. By a fluke it was before, but at various times in my journey it stopped doing so. So I just added an INLINE pragma to it; no sense in depending on a delicately-balanced fluke. - - - - - ae24c9bc by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Slight improvement in WorkWrap Ensure that WorkWrap preserves lambda binders, in case of join points. Sadly I have forgotten why I made this change (it was while I was doing a lot of meddling in the Simplifier, but * it does no harm, * it is slightly more efficient, and * presumably it made something better! Anyway I have kept it in a separate commit. - - - - - e9297181 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Use named record fields for the CastIt { ... } data constructor This is a pure refactor - - - - - b4581e23 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Remove a long-commented-out line Pure refactoring - - - - - e026bdf2 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Simplifier improvements This MR started as: allow the simplifer to do more in one pass, arising from places I could see the simplifier taking two iterations where one would do. But it turned into a larger project, because these changes unexpectedly made inlining blow up, especially join points in deeply-nested cases. The main changes are below. There are also many new or rewritten Notes. Avoiding simplifying repeatedly ~~~~~~~~~~~~~~~ See Note [Avoiding simplifying repeatedly] * The SimplEnv now has a seInlineDepth field, which says how deep in unfoldings we are. See Note [Inline depth] in Simplify.Env. Currently used only for the next point: avoiding repeatedly simplifying coercions. * Avoid repeatedly simplifying coercions. see Note [Avoid re-simplifying coercions] in Simplify.Iteration As you'll see from the Note, this makes use of the seInlineDepth. * Allow Simplify.Iteration.simplAuxBind to inline used-once things. This is another part of Note [Post-inline for single-use things], and is really good for reducing simplifier iterations in situations like case K e of { K x -> blah } wher x is used once in blah. * Make GHC.Core.SimpleOpt.exprIsConApp_maybe do some simple case elimination. Note [Case elim in exprIsConApp_maybe] * Improve the case-merge transformation: - Move the main code to `GHC.Core.Utils.mergeCaseAlts`, to join `filterAlts` and friends. See Note [Merge Nested Cases] in GHC.Core.Utils. - Add a new case for `tagToEnum#`; see wrinkle (MC3). - Add a new case to look through join points: see wrinkle (MC4) postInlineUnconditionally ~~~~~~~~~~~~~~~~~~~~~~~~~ * Allow Simplify.Utils.postInlineUnconditionally to inline variables that are used exactly once. See Note [Post-inline for single-use things]. * Do not postInlineUnconditionally join point, ever. Doing so does not reduce allocation, which is the main point, and with join points that are used a lot it can bloat code. See point (1) of Note [Duplicating join points] in GHC.Core.Opt.Simplify.Iteration. * Do not postInlineUnconditionally a strict (demanded) binding. It will not allocate a thunk (it'll turn into a case instead) so again the main point of inlining it doesn't hold. Better to check per-call-site. * Improve occurrence analyis for bottoming function calls, to help postInlineUnconditionally. See Note [Bottoming function calls] in GHC.Core.Opt.OccurAnal Inlining generally ~~~~~~~~~~~~~~~~~~ * In GHC.Core.Opt.Simplify.Utils.interestingCallContext, use RhsCtxt NonRecursive (not BoringCtxt) for a plain-seq case. See Note [Seq is boring] Also, wrinkle (SB1), inline in that `seq` context only for INLINE functions (UnfWhen guidance). * In GHC.Core.Opt.Simplify.Utils.interestingArg, - return ValueArg for OtherCon [c1,c2, ...], but - return NonTrivArg for OtherCon [] This makes a function a little less likely to inline if all we know is that the argument is evaluated, but nothing else. * isConLikeUnfolding is no longer true for OtherCon {}. This propagates to exprIsConLike. Con-like-ness has /positive/ information. Join points ~~~~~~~~~~~ * Be very careful about inlining join points. See these two long Notes Note [Duplicating join points] in GHC.Core.Opt.Simplify.Iteration Note [Inlining join points] in GHC.Core.Opt.Simplify.Inline * When making join points, don't do so if the join point is so small it will immediately be inlined; check uncondInlineJoin. * In GHC.Core.Opt.Simplify.Inline.tryUnfolding, improve the inlining heuristics for join points. In general we /do not/ want to inline join points /even if they are small/. See Note [Duplicating join points] GHC.Core.Opt.Simplify.Iteration. But sometimes we do: see Note [Inlining join points] in GHC.Core.Opt.Simplify.Inline; and the new `isBetterUnfoldingThan` function. * Do not add an unfolding to a join point at birth. This is a tricky one and has a long Note [Do not add unfoldings to join points at birth] It shows up in two places - In `mkDupableAlt` do not add an inlining - (trickier) In `simplLetUnfolding` don't add an unfolding for a fresh join point I am not fully satisifed with this, but it works and is well documented. * In GHC.Core.Unfold.sizeExpr, make jumps small, so that we don't penalise having a non-inlined join point. Performance changes ~~~~~~~~~~~~~~~~~~~ * Binary sizes fall by around 2.6%, according to nofib. * Compile times improve slightly. Here are the figures over 1%. I investiate the biggest differnce in T18304. It's a very small module, just a few hundred nodes. The large percentage difffence is due to a single function that didn't quite inline before, and does now, making code size a bit bigger. I decided gains outweighed the losses. Metrics: compile_time/bytes allocated (changes over +/- 1%) ------------------------------------------------ CoOpt_Singletons(normal) -9.2% GOOD LargeRecord(normal) -23.5% GOOD MultiComponentModulesRecomp(normal) +1.2% MultiLayerModulesTH_OneShot(normal) +4.1% BAD PmSeriesS(normal) -3.8% PmSeriesV(normal) -1.5% T11195(normal) -1.3% T12227(normal) -20.4% GOOD T12545(normal) -3.2% T12707(normal) -2.1% GOOD T13253(normal) -1.2% T13253-spj(normal) +8.1% BAD T13386(normal) -3.1% GOOD T14766(normal) -2.6% GOOD T15164(normal) -1.4% T15304(normal) +1.2% T15630(normal) -8.2% T15630a(normal) NEW T15703(normal) -14.7% GOOD T16577(normal) -2.3% GOOD T17516(normal) -39.7% GOOD T18140(normal) +1.2% T18223(normal) -17.1% GOOD T18282(normal) -5.0% GOOD T18304(normal) +10.8% BAD T18923(normal) -2.9% GOOD T1969(normal) +1.0% T19695(normal) -1.5% T20049(normal) -12.7% GOOD T21839c(normal) -4.1% GOOD T3064(normal) -1.5% T3294(normal) +1.2% BAD T4801(normal) +1.2% T5030(normal) -15.2% GOOD T5321Fun(normal) -2.2% GOOD T6048(optasm) -16.8% GOOD T783(normal) -1.2% T8095(normal) -6.0% GOOD T9630(normal) -4.7% GOOD T9961(normal) +1.9% BAD WWRec(normal) -1.4% info_table_map_perf(normal) -1.3% parsing001(normal) +1.5% geo. mean -2.0% minimum -39.7% maximum +10.8% * Runtimes generally improve. In the testsuite perf/should_run gives: Metrics: runtime/bytes allocated ------------------------------------------ Conversions(normal) -0.3% T13536a(optasm) -41.7% GOOD T4830(normal) -0.1% haddock.Cabal(normal) -0.1% haddock.base(normal) -0.1% haddock.compiler(normal) -0.1% geo. mean -0.8% minimum -41.7% maximum +0.0% * For runtime, nofib is a better test. The news is mostly good. Here are the number more than +/- 0.1%: # bytes allocated ==========================++========== imaginary/digits-of-e1 || -14.40% imaginary/digits-of-e2 || -4.41% imaginary/paraffins || -0.17% imaginary/rfib || -0.15% imaginary/wheel-sieve2 || -0.10% real/compress || -0.47% real/fluid || -0.10% real/fulsom || +0.14% real/gamteb || -1.47% real/gg || -0.20% real/infer || +0.24% real/pic || -0.23% real/prolog || -0.36% real/scs || -0.46% real/smallpt || +4.03% shootout/k-nucleotide || -20.23% shootout/n-body || -0.42% shootout/spectral-norm || -0.13% spectral/boyer2 || -3.80% spectral/constraints || -0.27% spectral/hartel/ida || -0.82% spectral/mate || -20.34% spectral/para || +0.46% spectral/rewrite || +1.30% spectral/sphere || -0.14% ==========================++========== geom mean || -0.59% real/smallpt has a huge nest of local definitions, and I could not pin down a reason for a regression. But there are three big wins! Metric Decrease: CoOpt_Singletons LargeRecord T12227 T12707 T13386 T13536a T14766 T15703 T16577 T17516 T18223 T18282 T18923 T21839c T20049 T5321Fun T5030 T6048 T8095 T9630 T783 Metric Increase: MultiLayerModulesTH_OneShot T13253-spj T18304 T18698a T9961 T3294 - - - - - 27db3c5e by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Testsuite message changes from simplifier improvements - - - - - 271a7812 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Account for bottoming functions in OccurAnal This fixes #24582, a small but long-standing bug - - - - - 0fde229f by Ben Gamari at 2024-04-04T07:04:58-04:00 testsuite: Introduce template-haskell-exports test - - - - - 0c4a9686 by Luite Stegeman at 2024-04-04T07:05:39-04:00 Update correct counter in bumpTickyAllocd - - - - - 5f085d3a by Fendor at 2024-04-04T14:47:33-04:00 Replace `SizedSeq` with `FlatBag` for flattened structure LinkedLists are notoriously memory ineffiecient when all we do is traversing a structure. As 'UnlinkedBCO' has been identified as a data structure that impacts the overall memory usage of GHCi sessions, we avoid linked lists and prefer flattened structure for storing. We introduce a new memory efficient representation of sequential elements that has special support for the cases: * Empty * Singleton * Tuple Elements This improves sharing in the 'Empty' case and avoids the overhead of 'Array' until its constant overhead is justified. - - - - - 82cfe10c by Fendor at 2024-04-04T14:47:33-04:00 Compact FlatBag array representation `Array` contains three additional `Word`'s we do not need in `FlatBag`. Move `FlatBag` to `SmallArray`. Expand the API of SmallArray by `sizeofSmallArray` and add common traversal functions, such as `mapSmallArray` and `foldMapSmallArray`. Additionally, allow users to force the elements of a `SmallArray` via `rnfSmallArray`. - - - - - 36a75b80 by Andrei Borzenkov at 2024-04-04T14:48:10-04:00 Change how invisible patterns represented in haskell syntax and TH AST (#24557) Before this patch: data ArgPat p = InvisPat (LHsType p) | VisPat (LPat p) With this patch: data Pat p = ... | InvisPat (LHsType p) ... And the same transformation in the TH land. The rest of the changes is just updating code to handle new AST and writing tests to check if it is possible to create invalid states using TH. Metric Increase: MultiLayerModulesTH_OneShot - - - - - 28009fbc by Matthew Pickering at 2024-04-04T14:48:46-04:00 Fix off by one error in seekBinNoExpand and seekBin - - - - - 9b9e031b by Ben Gamari at 2024-04-04T21:30:08-04:00 compiler: Allow more types in GHCForeignImportPrim For many, many years `GHCForeignImportPrim` has suffered from the rather restrictive limitation of not allowing any non-trivial types in arguments or results. This limitation was justified by the code generator allegely barfing in the presence of such types. However, this restriction appears to originate well before the NCG rewrite and the new NCG does not appear to have any trouble with such types (see the added `T24598` test). Lift this restriction. Fixes #24598. - - - - - 1324b862 by Alan Zimmerman at 2024-04-04T21:30:44-04:00 EPA: Use EpaLocation not SrcSpan in ForeignDecls This allows us to update them for makeDeltaAst in ghc-exactprint - - - - - 19883a23 by Alan Zimmerman at 2024-04-05T16:58:17-04:00 EPA: Use EpaLocation for RecFieldsDotDot So we can update it to a delta position in makeDeltaAst if needed. - - - - - e8724327 by Matthew Pickering at 2024-04-05T16:58:53-04:00 Remove accidentally committed test.hs - - - - - 88cb3e10 by Fendor at 2024-04-08T09:03:34-04:00 Avoid UArray when indexing is not required `UnlinkedBCO`'s can occur many times in the heap. Each `UnlinkedBCO` references two `UArray`'s but never indexes them. They are only needed to encode the elements into a `ByteArray#`. The three words for the lower bound, upper bound and number of elements are essentially unused, thus we replace `UArray` with a wrapper around `ByteArray#`. This saves us up to three words for each `UnlinkedBCO`. Further, to avoid re-allocating these words for `ResolvedBCO`, we repeat the procedure for `ResolvedBCO` and add custom `Binary` and `Show` instances. For example, agda's repl session has around 360_000 UnlinkedBCO's, so avoiding these three words is already saving us around 8MB residency. - - - - - f2cc1107 by Fendor at 2024-04-08T09:04:11-04:00 Never UNPACK `FastMutInt` for counting z-encoded `FastString`s In `FastStringTable`, we count the number of z-encoded FastStrings that exist in a GHC session. We used to UNPACK the counters to not waste memory, but live retainer analysis showed that we allocate a lot of `FastMutInt`s, retained by `mkFastZString`. We lazily compute the `FastZString`, only incrementing the counter when the `FastZString` is forced. The function `mkFastStringWith` calls `mkZFastString` and boxes the `FastMutInt`, leading to the following core: mkFastStringWith = \ mk_fs _ -> = case stringTable of { FastStringTable _ n_zencs segments# _ -> ... case ((mk_fs (I# ...) (FastMutInt n_zencs)) `cast` <Co:2> :: ...) ... Marking this field as `NOUNPACK` avoids this reboxing, eliminating the allocation of a fresh `FastMutInt` on every `FastString` allocation. - - - - - c6def949 by Matthew Pickering at 2024-04-08T16:06:51-04:00 Force in_multi to avoid retaining entire hsc_env - - - - - fbb91a63 by Fendor at 2024-04-08T16:06:51-04:00 Eliminate name thunk in declaration fingerprinting Thunk analysis showed that we have about 100_000 thunks (in agda and `-fwrite-simplified-core`) pointing to the name of the name decl. Forcing this thunk fixes this issue. The thunk created here is retained by the thunk created by forkM, it is better to eagerly force this because the result (a `Name`) is already retained indirectly via the `IfaceDecl`. - - - - - 3b7b0c1c by Alan Zimmerman at 2024-04-08T16:07:27-04:00 EPA: Use EpaLocation in WarningTxt This allows us to use an EpDelta if needed when using makeDeltaAst. - - - - - 12b997df by Alan Zimmerman at 2024-04-08T16:07:27-04:00 EPA: Move DeltaPos and EpaLocation' into GHC.Types.SrcLoc This allows us to use a NoCommentsLocation for the possibly trailing comma location in a StringLiteral. This in turn allows us to correctly roundtrip via makeDeltaAst. - - - - - 868c8a78 by Fendor at 2024-04-09T08:51:50-04:00 Prefer packed representation for CompiledByteCode As there are many 'CompiledByteCode' objects alive during a GHCi session, representing its element in a more packed manner improves space behaviour at a minimal cost. When running GHCi on the agda codebase, we find around 380 live 'CompiledByteCode' objects. Packing their respective 'UnlinkedByteCode' can save quite some pointers. - - - - - be3bddde by Alan Zimmerman at 2024-04-09T08:52:26-04:00 EPA: Capture all comments in a ClassDecl Hopefully the final fix needed for #24533 - - - - - 3d0806fc by Jade at 2024-04-10T05:39:53-04:00 Validate -main-is flag using parseIdentifier Fixes #24368 - - - - - dd530bb7 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 rts: free error message before returning Fixes a memory leak in rts/linker/PEi386.c - - - - - e008a19a by Alexis King at 2024-04-10T05:40:29-04:00 linker: Avoid linear search when looking up Haskell symbols via dlsym See the primary Note [Looking up symbols in the relevant objects] for a more in-depth explanation. When dynamically loading a Haskell symbol (typical when running a splice or GHCi expression), before this commit we would search for the symbol in all dynamic libraries that were loaded. However, this could be very inefficient when too many packages are loaded (which can happen if there are many package dependencies) because the time to lookup the would be linear in the number of packages loaded. This commit drastically improves symbol loading performance by introducing a mapping from units to the handles of corresponding loaded dlls. These handles are returned by dlopen when we load a dll, and can then be used to look up in a specific dynamic library. Looking up a given Name is now much more precise because we can get lookup its unit in the mapping and lookup the symbol solely in the handles of the dynamic libraries loaded for that unit. In one measurement, the wait time before the expression was executed went from +-38 seconds down to +-2s. This commit also includes Note [Symbols may not be found in pkgs_loaded], explaining the fallback to the old behaviour in case no dll can be found in the unit mapping for a given Name. Fixes #23415 Co-authored-by: Rodrigo Mesquita (@alt-romes) - - - - - dcfaa190 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 rts: Make addDLL a wrapper around loadNativeObj Rewrite the implementation of `addDLL` as a wrapper around the more principled `loadNativeObj` rts linker function. The latter should be preferred while the former is preserved for backwards compatibility. `loadNativeObj` was previously only available on ELF platforms, so this commit further refactors the rts linker to transform loadNativeObj_ELF into loadNativeObj_POSIX, which is available in ELF and MachO platforms. The refactor made it possible to remove the `dl_mutex` mutex in favour of always using `linker_mutex` (rather than a combination of both). Lastly, we implement `loadNativeObj` for Windows too. - - - - - 12931698 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 Use symbol cache in internal interpreter too This commit makes the symbol cache that was used by the external interpreter available for the internal interpreter too. This follows from the analysis in #23415 that suggests the internal interpreter could benefit from this cache too, and that there is no good reason not to have the cache for it too. It also makes it a bit more uniform to have the symbol cache range over both the internal and external interpreter. This commit also refactors the cache into a function which is used by both `lookupSymbol` and also by `lookupSymbolInDLL`, extending the caching logic to `lookupSymbolInDLL` too. - - - - - dccd3ea1 by Ben Gamari at 2024-04-10T05:40:29-04:00 testsuite: Add test for lookupSymbolInNativeObj - - - - - 1b1a92bd by Alan Zimmerman at 2024-04-10T05:41:05-04:00 EPA: Remove unnecessary XRec in CompleteMatchSig The XRec for [LIdP pass] is not needed for exact printing, remove it. - - - - - 6e18ce2b by Ben Gamari at 2024-04-12T08:16:09-04:00 users-guide: Clarify language extension documentation Over the years the users guide's language extension documentation has gone through quite a few refactorings. In the process some of the descriptions have been rendered non-sensical. For instance, the description of `NoImplicitPrelude` actually describes the semantics of `ImplicitPrelude`. To fix this we: * ensure that all extensions are named in their "positive" sense (e.g. `ImplicitPrelude` rather than `NoImplicitPrelude`). * rework the documentation to avoid flag-oriented wording like "enable" and "disable" * ensure that the polarity of the documentation is consistent with reality. Fixes #23895. - - - - - a933aff3 by Zubin Duggal at 2024-04-12T08:16:45-04:00 driver: Make `checkHomeUnitsClosed` faster The implementation of `checkHomeUnitsClosed` was traversing every single path in the unit dependency graph - this grows exponentially and quickly grows to be infeasible on larger unit dependency graphs. Instead we replace this with a faster implementation which follows from the specificiation of the closure property - there is a closure error if there are units which are both are both (transitively) depended upon by home units and (transitively) depend on home units, but are not themselves home units. To compute the set of units required for closure, we first compute the closure of the unit dependency graph, then the transpose of this closure, and find all units that are reachable from the home units in the transpose of the closure. - - - - - 23c3e624 by Andreas Klebinger at 2024-04-12T08:17:21-04:00 RTS: Emit warning when -M < -H Fixes #24487 - - - - - d23afb8c by Ben Gamari at 2024-04-12T08:17:56-04:00 testsuite: Add broken test for CApiFFI with -fprefer-bytecode See #24634. - - - - - a4bb3a51 by Ben Gamari at 2024-04-12T08:18:32-04:00 base: Deprecate GHC.Pack As proposed in #21461. Closes #21540. - - - - - 55eb8c98 by Ben Gamari at 2024-04-12T08:19:08-04:00 ghc-internal: Fix mentions of ghc-internal in deprecation warnings Closes #24609. - - - - - b0fbd181 by Ben Gamari at 2024-04-12T08:19:44-04:00 rts: Implement set_initial_registers for AArch64 Fixes #23680. - - - - - 14c9ec62 by Ben Gamari at 2024-04-12T08:20:20-04:00 ghcup-metadata: Use Debian 9 binaries on Ubuntu 16, 17 Closes #24646. - - - - - 35a1621e by Ben Gamari at 2024-04-12T08:20:55-04:00 Bump unix submodule to 2.8.5.1 Closes #24640. - - - - - a1c24df0 by Finley McIlwaine at 2024-04-12T08:21:31-04:00 Correct default -funfolding-use-threshold in docs - - - - - 0255d03c by Oleg Grenrus at 2024-04-12T08:22:07-04:00 FastString is a __Modified__ UTF-8 - - - - - c3489547 by Matthew Pickering at 2024-04-12T13:13:44-04:00 rts: Improve tracing message when nursery is resized It is sometimes more useful to know how much bigger or smaller the nursery got when it is resized. In particular I am trying to investigate situations where we end up with fragmentation due to the nursery (#24577) - - - - - 5e4f4ba8 by Simon Peyton Jones at 2024-04-12T13:14:20-04:00 Don't generate wrappers for `type data` constructors with StrictData Previously, the logic for checking if a data constructor needs a wrapper or not would take into account whether the constructor's fields have explicit strictness (e.g., `data T = MkT !Int`), but the logic would _not_ take into account whether `StrictData` was enabled. This meant that something like `type data T = MkT Int` would incorrectly generate a wrapper for `MkT` if `StrictData` was enabled, leading to the horrible errors seen in #24620. To fix this, we disable generating wrappers for `type data` constructors altogether. Fixes #24620. Co-authored-by: Ryan Scott <ryan.gl.scott at gmail.com> - - - - - dbdf1995 by Alex Mason at 2024-04-15T15:28:26+10:00 Implements MO_S_Mul2 and MO_U_Mul2 using the UMULH, UMULL and SMULH instructions for AArch64 Also adds a test for MO_S_Mul2 - - - - - 42bd0407 by Teo Camarasu at 2024-04-16T20:06:39-04:00 Make template-haskell a stage1 package Promoting template-haskell from a stage0 to a stage1 package means that we can much more easily refactor template-haskell. We implement this by duplicating the in-tree `template-haskell`. A new `template-haskell-next` library is autogenerated to mirror `template-haskell` `stage1:ghc` to depend on the new interface of the library including the `Binary` instances without adding an explicit dependency on `template-haskell`. This is controlled by the `bootstrap-th` cabal flag When building `template-haskell` modules as part of this vendoring we do not have access to quote syntax, so we cannot use variable quote notation (`'Just`). So we either replace these with hand-written `Name`s or hide the code behind CPP. We can remove the `th_hack` from hadrian, which was required when building stage0 packages using the in-tree `template-haskell` library. For more details see Note [Bootstrapping Template Haskell]. Resolves #23536 Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> Co-Authored-By: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 3d973e47 by Ben Gamari at 2024-04-16T20:07:15-04:00 Bump parsec submodule to 3.1.17.0 - - - - - 9d38bfa0 by Simon Peyton Jones at 2024-04-16T20:07:51-04:00 Clone CoVars in CorePrep This MR addresses #24463. It's all explained in the new Note [Cloning CoVars and TyVars] - - - - - 0fe2b410 by Andreas Klebinger at 2024-04-16T20:08:27-04:00 NCG: Fix a bug where we errounously removed a required jump instruction. Add a new method to the Instruction class to check if we can eliminate a jump in favour of fallthrough control flow. Fixes #24507 - - - - - 9f99126a by Teo Camarasu at 2024-04-16T20:09:04-04:00 Fix documentation preview from doc-tarball job - Include all the .html files and assets in the job artefacts - Include all the .pdf files in the job artefacts - Mark the artefact as an "exposed" artefact meaning it turns up in the UI. Resolves #24651 - - - - - 3a0642ea by Ben Gamari at 2024-04-16T20:09:39-04:00 rts: Ignore EINTR while polling in timerfd itimer implementation While the RTS does attempt to mask signals, it may be that a foreign library unmasks them. This previously caused benign warnings which we now ignore. See #24610. - - - - - 9a53cd3f by Alan Zimmerman at 2024-04-16T20:10:15-04:00 EPA: Add additional comments field to AnnsModule This is used in exact printing to store comments coming after the `where` keyword but before any comments allocated to imports or decls. It is used in ghc-exactprint, see https://github.com/alanz/ghc-exactprint/commit/44bbed311fd8f0d053053fef195bf47c17d34fa7 - - - - - e5c43259 by Bryan Richter at 2024-04-16T20:10:51-04:00 Remove unrunnable FreeBSD CI jobs FreeBSD runner supply is inelastic. Currently there is only one, and it's unavailable because of a hardware issue. - - - - - 914eb49a by Ben Gamari at 2024-04-16T20:11:27-04:00 rel-eng: Fix mktemp usage in recompress-all We need a temporary directory, not a file. - - - - - f30e4984 by Teo Camarasu at 2024-04-16T20:12:03-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 - - - - - d7a3d6b5 by Ben Gamari at 2024-04-16T20:12:39-04:00 template-haskell: Declare TH.Lib.Internal as not-home Rather than `hide`. Closes #24659. - - - - - 5eaa46e7 by Matthew Pickering at 2024-04-19T02:14:55-04:00 testsuite: Rename isCross() predicate to needsTargetWrapper() isCross() was a misnamed because it assumed that all cross targets would provide a target wrapper, but the two most common cross targets (javascript, wasm) don't need a target wrapper. Therefore we rename this predicate to `needsTargetWrapper()` so situations in the testsuite where we can check whether running executables requires a target wrapper or not. - - - - - 55a9d699 by Simon Peyton Jones at 2024-04-19T02:15:32-04:00 Do not float HNFs out of lambdas This MR adjusts SetLevels so that it is less eager to float a HNF (lambda or constructor application) out of a lambda, unless it gets to top level. Data suggests that this change is a small net win: * nofib bytes-allocated falls by -0.09% (but a couple go up) * perf/should_compile bytes-allocated falls by -0.5% * perf/should_run bytes-allocated falls by -0.1% See !12410 for more detail. When fiddling elsewhere, I also found that this patch had a huge positive effect on the (very delicate) test perf/should_run/T21839r But that improvement doesn't show up in this MR by itself. Metric Decrease: MultiLayerModulesRecomp T15703 parsing001 - - - - - f0701585 by Alan Zimmerman at 2024-04-19T02:16:08-04:00 EPA: Fix comments in mkListSyntaxTy0 Also extend the test to confirm. Addresses #24669, 1 of 4 - - - - - b01c01d4 by Serge S. Gulin at 2024-04-19T02:16:51-04:00 JS: set image `x86_64-linux-deb11-emsdk-closure` for build - - - - - c90c6039 by Alan Zimmerman at 2024-04-19T02:17:27-04:00 EPA: Provide correct span for PatBind And remove unused parameter in checkPatBind Contributes to #24669 - - - - - bee54c24 by Krzysztof Gogolewski at 2024-04-19T11:13:00+02:00 Update quantification order following GHC haskell/haddock#23764 - - - - - 2814eb89 by Ben Gamari at 2024-04-19T18:57:05+02:00 hypsrc-test: Fix output of PositionPragmas.html - - - - - 26036f96 by Alan Zimmerman at 2024-04-19T13:11:08-04:00 EPA: Fix span for PatBuilderAppType Include the location of the prefix @ in the span for InVisPat. Also removes unnecessary annotations from HsTP. Contributes to #24669 - - - - - dba03aab by Matthew Craven at 2024-04-19T13:11:44-04:00 testsuite: Give the pre_cmd for mhu-perf more time - - - - - d31fbf6c by Krzysztof Gogolewski at 2024-04-19T21:04:09-04:00 Fix quantification order for a `op` b and a %m -> b Fixes #23764 Implements https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0640-tyop-quantification-order.rst Updates haddock submodule. - - - - - 385cd1c4 by Sebastian Graf at 2024-04-19T21:04:45-04:00 Make `seq#` a magic Id and inline it in CorePrep (#24124) We can save much code and explanation in Tag Inference and StgToCmm by making `seq#` a known-key Magic Id in `GHC.Internal.IO` and inline this definition in CorePrep. See the updated `Note [seq# magic]`. I also implemented a new `Note [Flatten case-bind]` to get better code for otherwise nested case scrutinees. I renamed the contructors of `ArgInfo` to use an `AI` prefix in order to resolve the clash between `type CpeApp = CoreExpr` and the data constructor of `ArgInfo`, as well as fixed typos in `Note [CorePrep invariants]`. Fixes #24252 and #24124. - - - - - 275e41a9 by Jade at 2024-04-20T11:10:40-04:00 Put the newline after errors instead of before them This mainly has consequences for GHCi but also slightly alters how the output of GHC on the commandline looks. Fixes: #22499 - - - - - dd339c7a by Teo Camarasu at 2024-04-20T11:11:16-04:00 Remove unecessary stage0 packages Historically quite a few packages had to be stage0 as they depended on `template-haskell` and that was stage0. In #23536 we made it so that was no longer the case. This allows us to remove a bunch of packages from this list. A few still remain. A new version of `Win32` is required by `semaphore-compat`. Including `Win32` in the stage0 set requires also including `filepath` because otherwise Hadrian's dependency logic gets confused. Once our boot compiler has a newer version of `Win32` all of these will be able to be dropped. Resolves #24652 - - - - - 2f8e3a25 by Alan Zimmerman at 2024-04-20T11:11:52-04:00 EPA: Avoid duplicated comments in splice decls Contributes to #24669 - - - - - c70b9ddb by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: fix typos and namings (fixes #24602) You may noted that I've also changed term of ``` , global "h$vt_double" ||= toJExpr IntV ``` See "IntV" and ``` WaitReadOp -> \[] [fd] -> pure $ PRPrimCall $ returnS (app "h$waidRead" [fd]) ``` See "h$waidRead" - - - - - 3db54f9b by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: trivial checks for variable presence (fixes #24602) - - - - - 777f108f by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: fs module imported twice (by emscripten and by ghc-internal). ghc-internal import wrapped in a closure to prevent conflict with emscripten (fixes #24602) Better solution is to use some JavaScript module system like AMD, CommonJS or even UMD. It will be investigated at other issues. At first glance we should try UMD (See https://github.com/umdjs/umd) - - - - - a45a5712 by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: thread.js requires h$fds and h$fdReady to be declared for static code analysis, minimal code copied from GHCJS (fixes #24602) I've just copied some old pieces of GHCJS from publicly available sources (See https://github.com/Taneb/shims/blob/a6dd0202dcdb86ad63201495b8b5d9763483eb35/src/io.js#L607). Also I didn't put details to h$fds. I took minimal and left only its object initialization: `var h$fds = {};` - - - - - ad90bf12 by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: heap and stack overflows reporting defined as js hard failure (fixes #24602) These errors were treated as a hard failure for browser application. The fix is trivial: just throw error. - - - - - 5962fa52 by Serge S. Gulin at 2024-04-21T16:33:44+03:00 JS: Stubs for code without actual implementation detected by Google Closure Compiler (fixes #24602) These errors were fixed just by introducing stubbed functions with throw for further implementation. - - - - - a0694298 by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Add externs to linker (fixes #24602) After enabling jsdoc and built-in google closure compiler types I was needed to deal with the following: 1. Define NodeJS-environment types. I've just copied minimal set of externs from semi-official repo (see https://github.com/externs/nodejs/blob/6c6882c73efcdceecf42e7ba11f1e3e5c9c041f0/v8/nodejs.js#L8). 2. Define Emscripten-environment types: `HEAP8`. Emscripten already provides some externs in our code but it supposed to be run in some module system. And its definitions do not work well in plain bundle. 3. We have some functions which purpose is to add to functions some contextual information via function properties. These functions should be marked as `modifies` to let google closure compiler remove calls if these functions are not used actually by call graph. Such functions are: `h$o`, `h$sti`, `h$init_closure`, `h$setObjInfo`. 4. STG primitives such as registries and stuff from `GHC.StgToJS`. `dXX` properties were already present at externs generator function but they are started from `7`, not from `1`. This message is related: `// fixme does closure compiler bite us here?` - - - - - e58bb29f by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: added both tests: for size and for correctness (fixes #24602) By some reason MacOS builds add to stderr messages like: Ignoring unexpected archive entry: __.SYMDEF ... However I left stderr to `/dev/null` for compatibility with linux CI builds. - - - - - 909f3a9c by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Disable js linker warning for empty symbol table to make js tests running consistent across environments - - - - - 83eb10da by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Add special preprocessor for js files due of needing to keep jsdoc comments (fixes #24602) Our js files have defined google closure compiler types at jsdoc entries but these jsdoc entries are removed by cpp preprocessor. I considered that reusing them in javascript-backend would be a nice thing. Right now haskell processor uses `-traditional` option to deal with comments and `//` operators. But now there are following compiler options: `-C` and `-CC`. You can read about them at GCC (see https://gcc.gnu.org/onlinedocs/gcc/Preprocessor-Options.html#index-CC) and CLang (see https://clang.llvm.org/docs/ClangCommandLineReference.html#cmdoption-clang-CC). It seems that `-CC` works better for javascript jsdoc than `-traditional`. At least it leaves `/* ... */` comments w/o changes. - - - - - e1cf8dc2 by brandon s allbery kf8nh at 2024-04-22T03:48:26-04:00 fix link in CODEOWNERS It seems that our local Gitlab no longer has documentation for the `CODEOWNERS` file, but the master documentation still does. Use that instead. - - - - - a27c6a49 by Fendor at 2024-04-22T10:13:03+02:00 Adapt to UserData split - - - - - 1efc5a7a by Fendor at 2024-04-22T10:13:03+02:00 Adapt to BinHandle split - - - - - 593f4e04 by Fendor at 2024-04-23T10:19:14-04:00 Add performance regression test for '-fwrite-simplified-core' - - - - - 1ba39b05 by Fendor at 2024-04-23T10:19:14-04:00 Typecheck corebindings lazily during bytecode generation This delays typechecking the corebindings until the bytecode generation happens. We also avoid allocating a thunk that is retained by `unsafeInterleaveIO`. In general, we shouldn't retain values of the hydrated `Type`, as not evaluating the bytecode object keeps it alive. It is better if we retain the unhydrated `IfaceType`. See Note [Hydrating Modules] - - - - - e916fc92 by Alan Zimmerman at 2024-04-23T10:19:50-04:00 EPA: Keep comments in a CaseAlt match The comments now live in the surrounding location, not inside the Match. Make sure we keep them. Closes #24707 - - - - - d2b17f32 by Cheng Shao at 2024-04-23T15:01:22-04:00 driver: force merge objects when building dynamic objects This patch forces the driver to always merge objects when building dynamic objects even when ar -L is supported. It is an oversight of !8887: original rationale of that patch is favoring the relatively cheap ar -L operation over object merging when ar -L is supported, which makes sense but only if we are building static objects! Omitting check for whether we are building dynamic objects will result in broken .so files with undefined reference errors at executable link time when building GHC with llvm-ar. Fixes #22210. - - - - - 209d09f5 by Julian Ospald at 2024-04-23T15:02:03-04:00 Allow non-absolute values for bootstrap GHC variable Fixes #24682 - - - - - 3fff0977 by Matthew Pickering at 2024-04-23T15:02:38-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. - - - - - c62dc317 by Cheng Shao at 2024-04-25T01:32:02-04:00 ghc-bignum: remove obsolete ln script This commit removes an obsolete ln script in ghc-bignum/gmp. See 060251c24ad160264ae8553efecbb8bed2f06360 for its original intention, but it's been obsolete for a long time, especially since the removal of the make build system. Hence the house cleaning. - - - - - 6399d52b by Cheng Shao at 2024-04-25T01:32:02-04:00 ghc-bignum: update gmp to 6.3.0 This patch bumps the gmp-tarballs submodule and updates gmp to 6.3.0. The tarball format is now xz, and gmpsrc.patch has been patched into the tarball so hadrian no longer needs to deal with patching logic when building in-tree GMP. - - - - - 65b4b92f by Cheng Shao at 2024-04-25T01:32:02-04:00 hadrian: remove obsolete Patch logic This commit removes obsolete Patch logic from hadrian, given we no longer need to patch the gmp tarball when building in-tree GMP. - - - - - 71f28958 by Cheng Shao at 2024-04-25T01:32:02-04:00 autoconf: remove obsolete patch detection This commit removes obsolete deletection logic of the patch command from autoconf scripts, given we no longer need to patch anything in the GHC build process. - - - - - daeda834 by Sylvain Henry at 2024-04-25T01:32:43-04:00 JS: correctly handle RUBBISH literals (#24664) - - - - - 8a06ddf6 by Matthew Pickering at 2024-04-25T11:16:16-04:00 Linearise ghc-internal and base build This is achieved by requesting the final package database for ghc-internal, which mandates it is fully built as a dependency of configuring the `base` package. This is at the expense of cross-package parrallelism between ghc-internal and the base package. Fixes #24436 - - - - - 94da9365 by Andrei Borzenkov at 2024-04-25T11:16:54-04:00 Fix tuple puns renaming (24702) Move tuple renaming short cutter from `isBuiltInOcc_maybe` to `isPunOcc_maybe`, so we consider incoming module. I also fixed some hidden bugs that raised after the change was done. - - - - - fa03b1fb by Fendor at 2024-04-26T18:03:13-04:00 Refactor the Binary serialisation interface The goal is simplifiy adding deduplication tables to `ModIface` interface serialisation. We identify two main points of interest that make this difficult: 1. UserData hardcodes what `Binary` instances can have deduplication tables. Moreover, it heavily uses partial functions. 2. GHC.Iface.Binary hardcodes the deduplication tables for 'Name' and 'FastString', making it difficult to add more deduplication. Instead of having a single `UserData` record with fields for all the types that can have deduplication tables, we allow to provide custom serialisers for any `Typeable`. These are wrapped in existentials and stored in a `Map` indexed by their respective `TypeRep`. The `Binary` instance of the type to deduplicate still needs to explicitly look up the decoder via `findUserDataReader` and `findUserDataWriter`, which is no worse than the status-quo. `Map` was chosen as microbenchmarks indicate it is the fastest for a small number of keys (< 10). To generalise the deduplication table serialisation mechanism, we introduce the types `ReaderTable` and `WriterTable` which provide a simple interface that is sufficient to implement a general purpose deduplication mechanism for `writeBinIface` and `readBinIface`. This allows us to provide a list of deduplication tables for serialisation that can be extended more easily, for example for `IfaceTyCon`, see the issue https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for more motivation. In addition to this refactoring, we split `UserData` into `ReaderUserData` and `WriterUserData`, to avoid partial functions and reduce overall memory usage, as we need fewer mutable variables. Bump haddock submodule to accomodate for `UserData` split. ------------------------- Metric Increase: MultiLayerModulesTH_Make MultiLayerModulesRecomp T21839c ------------------------- - - - - - bac57298 by Fendor at 2024-04-26T18:03:13-04:00 Split `BinHandle` into `ReadBinHandle` and `WriteBinHandle` A `BinHandle` contains too much information for reading data. For example, it needs to keep a `FastMutInt` and a `IORef BinData`, when the non-mutable variants would suffice. Additionally, this change has the benefit that anyone can immediately tell whether the `BinHandle` is used for reading or writing. Bump haddock submodule BinHandle split. - - - - - 4d6394dd by Simon Peyton Jones at 2024-04-26T18:03:49-04:00 Fix missing escaping-kind check in tcPatSynSig Note [Escaping kind in type signatures] explains how we deal with escaping kinds in type signatures, e.g. f :: forall r (a :: TYPE r). a where the kind of the body is (TYPE r), but `r` is not in scope outside the forall-type. I had missed this subtlety in tcPatSynSig, leading to #24686. This MR fixes it; and a similar bug in tc_top_lhs_type. (The latter is tested by T24686a.) - - - - - 981c2c2c by Alan Zimmerman at 2024-04-26T18:04:25-04:00 EPA: check-exact: check that the roundtrip reproduces the source Closes #24670 - - - - - a8616747 by Andrew Lelechenko at 2024-04-26T18:05:01-04:00 Document that setEnv is not thread-safe - - - - - 1e41de83 by Bryan Richter at 2024-04-26T18:05:37-04:00 CI: Work around frequent Signal 9 errors - - - - - a6d5f9da by Naïm Favier at 2024-04-27T17:52:40-04:00 ghc-internal: add MonadFix instance for (,) Closes https://gitlab.haskell.org/ghc/ghc/-/issues/24288, implements CLC proposal https://github.com/haskell/core-libraries-committee/issues/238. Adds a MonadFix instance for tuples, permitting value recursion in the "native" writer monad and bringing consistency with the existing instance for transformers's WriterT (and, to a lesser extent, for Solo). - - - - - 64feadcd by Rodrigo Mesquita at 2024-04-27T17:53:16-04:00 bindist: Fix xattr cleaning The original fix (725343aa) was incorrect because it used the shell bracket syntax which is the quoting syntax in autoconf, making the test for existence be incorrect and therefore `xattr` was never run. Fixes #24554 - - - - - e2094df3 by damhiya at 2024-04-28T23:52:00+09:00 Make read accepts binary integer formats CLC proposal : https://github.com/haskell/core-libraries-committee/issues/177 - - - - - c62239b7 by Sylvain Henry at 2024-04-29T10:35:00+02:00 Fix tests for T22229 - - - - - 1c2fd963 by Alan Zimmerman at 2024-04-29T23:17:00-04:00 EPA: Preserve comments in Match Pats Closes #24708 Closes #24715 Closes #24734 - - - - - 4189d17e by Sylvain Henry at 2024-04-29T23:17:42-04:00 LLVM: better unreachable default destination in Switch (#24717) See added note. Co-authored-by: Siddharth Bhat <siddu.druid at gmail.com> - - - - - a3725c88 by Cheng Shao at 2024-04-29T23:18:20-04:00 ci: enable wasm jobs for MRs with wasm label This patch enables wasm jobs for MRs with wasm label. Previously the wasm label didn't actually have any effect on the CI pipeline, and full-ci needed to be applied to run wasm jobs which was a waste of runners when working on the wasm backend, hence the fix here. - - - - - 702f7964 by Matthew Pickering at 2024-04-29T23:18:56-04:00 Make interface files and object files depend on inplace .conf file A potential fix for #24737 - - - - - 728af21e by Cheng Shao at 2024-04-30T05:30:23-04:00 utils: remove obsolete vagrant scripts Vagrantfile has long been removed in !5288. This commit further removes the obsolete vagrant scripts in the tree. - - - - - 36f2c342 by Cheng Shao at 2024-04-30T05:31:00-04:00 Update autoconf scripts Scripts taken from autoconf 948ae97ca5703224bd3eada06b7a69f40dd15a02 - - - - - ecbf22a6 by Ben Gamari at 2024-04-30T05:31:36-04:00 ghcup-metadata: Drop output_name field This is entirely redundant to the filename of the URL. There is no compelling reason to name the downloaded file differently from its source. - - - - - c56d728e by Zubin Duggal at 2024-04-30T22:45:09-04:00 testsuite: Handle exceptions in framework_fail when testdir is not initialised When `framework_fail` is called before initialising testdir, it would fail with an exception reporting the testdir not being initialised instead of the actual failure. Ensure we report the actual reason for the failure instead of failing in this way. One way this can manifest is when trying to run a test that doesn't exist using `--only` - - - - - d5bea4d6 by Alan Zimmerman at 2024-04-30T22:45:45-04:00 EPA: Fix range for GADT decl with sig only Closes #24714 - - - - - 4d78c53c by Sylvain Henry at 2024-05-01T17:23:06-04:00 Fix TH dependencies (#22229) Add a dependency between Syntax and Internal (via module reexport). - - - - - 37e38db4 by Sylvain Henry at 2024-05-01T17:23:06-04:00 Bump haddock submodule - - - - - ca13075c by Sylvain Henry at 2024-05-01T17:23:47-04:00 JS: cleanup to prepare for #24743 - - - - - 40026ac3 by Alan Zimmerman at 2024-05-01T22:45:07-04:00 EPA: Preserve comments for PrefixCon Preserve comments in fun (Con {- c1 -} a b) = undefined Closes #24736 - - - - - 92134789 by Hécate Moonlight at 2024-05-01T22:45:42-04:00 Correct `@since` metadata in HpcFlags It was introduced in base-4.20, not 4.22. Fix #24721 - - - - - a580722e by Cheng Shao at 2024-05-02T08:18:45-04:00 testsuite: fix req_target_smp predicate - - - - - ac9c5f84 by Andreas Klebinger at 2024-05-02T08:18:45-04:00 STM: Remove (unused)coarse grained locking. The STM code had a coarse grained locking mode guarded by #defines that was unused. This commit removes the code. - - - - - 917ef81b by Andreas Klebinger at 2024-05-02T08:18:45-04:00 STM: Be more optimistic when validating in-flight transactions. * Don't lock tvars when performing non-committal validation. * If we encounter a locked tvar don't consider it a failure. This means in-flight validation will only fail if committing at the moment of validation is *guaranteed* to fail. This prevents in-flight validation from failing spuriously if it happens in parallel on multiple threads or parallel to thread comitting. - - - - - 167a56a0 by Alan Zimmerman at 2024-05-02T08:19:22-04:00 EPA: fix span for empty \case(s) In instance SDecide Nat where SZero %~ (SSucc _) = Disproved (\case) Ensure the span for the HsLam covers the full construct. Closes #24748 - - - - - 9bae34d8 by doyougnu at 2024-05-02T15:41:08-04:00 testsuite: expand size testing infrastructure - closes #24191 - adds windows_skip, wasm_skip, wasm_arch, find_so, _find_so - path_from_ghcPkg, collect_size_ghc_pkg, collect_object_size, find_non_inplace functions to testsuite - adds on_windows and req_dynamic_ghc predicate to testsuite The design is to not make the testsuite too smart and simply offload to ghc-pkg for locations of object files and directories. - - - - - b85b1199 by Sylvain Henry at 2024-05-02T15:41:49-04:00 GHCi: support inlining breakpoints (#24712) When a breakpoint is inlined, its context may change (e.g. tyvars in scope). We must take this into account and not used the breakpoint tick index as its sole identifier. Each instance of a breakpoint (even with the same tick index) now gets a different "info" index. We also need to distinguish modules: - tick module: module with the break array (tick counters, status, etc.) - info module: module having the CgBreakInfo (info at occurrence site) - - - - - 649c24b9 by Oleg Grenrus at 2024-05-03T20:45:42-04:00 Expose constructors of SNat, SChar and SSymbol in ghc-internal - - - - - d603f199 by Mikolaj Konarski at 2024-05-03T20:46:19-04:00 Add DCoVarSet to PluginProv (!12037) - - - - - ba480026 by Serge S. Gulin at 2024-05-03T20:47:01-04:00 JS: Enable more efficient packing of string data (fixes #24706) - - - - - be1e60ee by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Track in-scope variables in ruleCheckProgram This small patch fixes #24726, by tracking in-scope variables properly in -drule-check. Not hard to do! - - - - - 58408c77 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Add a couple more HasCallStack constraints in SimpleOpt Just for debugging, no effect on normal code - - - - - 70e245e8 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Add comments to Prep.hs This documentation patch fixes a TODO left over from !12364 - - - - - e5687186 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Use HasDebugCallStack, rather than HasCallStack - - - - - 631cefec by Cheng Shao at 2024-05-03T20:48:17-04:00 driver: always merge objects when possible This patch makes the driver always merge objects with `ld -r` when possible, and only fall back to calling `ar -L` when merge objects command is unavailable. This completely reverts !8887 and !12313, given more fixes in Cabal seems to be needed to avoid breaking certain configurations and the maintainence cost is exceeding the behefits in this case :/ - - - - - 1dacb506 by Ben Gamari at 2024-05-03T20:48:53-04:00 Bump time submodule to 1.14 As requested in #24528. ------------------------- Metric Decrease: ghc_bignum_so rts_so Metric Increase: cabal_syntax_dir rts_so time_dir time_so ------------------------- - - - - - 4941b90e by Ben Gamari at 2024-05-03T20:48:53-04:00 Bump terminfo submodule to current master - - - - - 43d48b44 by Cheng Shao at 2024-05-03T20:49:30-04:00 wasm: use scheduler.postTask() for context switch when available This patch makes use of scheduler.postTask() for JSFFI context switch when it's available. It's a more principled approach than our MessageChannel based setImmediate() implementation, and it's available in latest version of Chromium based browsers. - - - - - 08207501 by Cheng Shao at 2024-05-03T20:50:08-04:00 testsuite: give pre_cmd for mhu-perf 5x time - - - - - bf3d4db0 by Alan Zimmerman at 2024-05-03T20:50:43-04:00 EPA: Preserve comments for pattern synonym sig Closes #24749 - - - - - c49493f2 by Matthew Pickering at 2024-05-04T06:02:57-04:00 tests: Widen acceptance window for dir and so size tests These are testing things which are sometimes out the control of a GHC developer. Therefore we shouldn't fail CI if something about these dependencies change because we can't do anything about it. It is still useful to have these statistics for visualisation in grafana though. Ticket #24759 - - - - - 9562808d by Matthew Pickering at 2024-05-04T06:02:57-04:00 Disable rts_so test It has already manifested large fluctuations and destabilising CI Fixes #24762 - - - - - fc24c5cf by Ryan Scott at 2024-05-04T06:03:33-04:00 unboxedSum{Type,Data}Name: Use GHC.Types as the module Unboxed sum constructors are now defined in the `GHC.Types` module, so if you manually quote an unboxed sum (e.g., `''Sum2#`), you will get a `Name` like: ```hs GHC.Types.Sum2# ``` The `unboxedSumTypeName` function in `template-haskell`, however, mistakenly believes that unboxed sum constructors are defined in `GHC.Prim`, so `unboxedSumTypeName 2` would return an entirely different `Name`: ```hs GHC.Prim.(#|#) ``` This is a problem for Template Haskell users, as it means that they can't be sure which `Name` is the correct one. (Similarly for `unboxedSumDataName`.) This patch fixes the implementations of `unboxedSum{Type,Data}Name` to use `GHC.Types` as the module. For consistency with `unboxedTupleTypeName`, the `unboxedSumTypeName` function now uses the non-punned syntax for unboxed sums (`Sum<N>#`) as the `OccName`. Fixes #24750. - - - - - 7eab4e01 by Alan Zimmerman at 2024-05-04T16:14:55+01:00 EPA: Widen stmtslist to include last semicolon Closes #24754 - - - - - 06f7db40 by Teo Camarasu at 2024-05-05T00:19:38-04:00 doc: Fix type error in hs_try_putmvar example - - - - - af000532 by Moritz Schuler at 2024-05-05T06:30:58-04:00 Fix parsing of module names in CLI arguments closes issue #24732 - - - - - da74e9c9 by Ben Gamari at 2024-05-05T06:31:34-04:00 ghc-platform: Add Setup.hs The Hadrian bootstrapping script relies upon `Setup.hs` to drive its build. Addresses #24761. - - - - - 35d34fde by Alan Zimmerman at 2024-05-05T12:52:40-04:00 EPA: preserve comments in class and data decls Fix checkTyClHdr which was discarding comments. Closes #24755 - - - - - 03c5dfbf by Simon Peyton Jones at 2024-05-05T12:53:15-04:00 Fix a float-out error Ticket #24768 showed that the Simplifier was accidentally destroying a join point. It turned out to be that we were sending a bottoming join point to the top, accidentally abstracting over /other/ join points. Easily fixed. - - - - - adba68e7 by John Ericson at 2024-05-05T19:35:56-04:00 Substitute bindist files with Hadrian not configure The `ghc-toolchain` overhaul will eventually replace all this stuff with something much more cleaned up, but I think it is still worth making this sort of cleanup in the meantime so other untanglings and dead code cleaning can procede. I was able to delete a fair amount of dead code doing this too. `LLVMTarget_CPP` is renamed to / merged with `LLVMTarget` because it wasn't actually turned into a valid CPP identifier. (Original to 1345c7cc42c45e63ab1726a8fd24a7e4d4222467, actually.) Progress on #23966 Co-Authored-By: Sylvain Henry <hsyl20 at gmail.com> - - - - - 18f4ff84 by Alan Zimmerman at 2024-05-05T19:36:32-04:00 EPA: fix mkHsOpTyPV duplicating comments Closes #24753 - - - - - a19201d4 by Matthew Craven at 2024-05-06T19:54:29-04:00 Add test cases for #24664 ...since none are present in the original MR !12463 fixing this issue. - - - - - 46328a49 by Alan Zimmerman at 2024-05-06T19:55:05-04:00 EPA: preserve comments in data decls Closes #24771 - - - - - 3b51995c by Andrei Borzenkov at 2024-05-07T14:39:40-04:00 Rename Solo# data constructor to MkSolo# (#24673) - data Solo# a = (# a #) + data Solo# a = MkSolo# a And `(# foo #)` syntax now becomes just a syntactic sugar for `MkSolo# a`. - - - - - 4d59abf2 by Arsen Arsenović at 2024-05-07T14:40:24-04:00 Add the cmm_cpp_is_gcc predicate to the testsuite A future C-- test called T24474-cmm-override-g0 relies on the GCC-specific behaviour of -g3 implying -dD, which, in turn, leads to it emitting #defines past the preprocessing stage. Clang, at least, does not do this, so the test would fail if ran on Clang. As the behaviour here being tested is ``-optCmmP-g3'' undoing effects of the workaround we apply as a fix for bug #24474, and the workaround was for GCC-specific behaviour, the test needs to be marked as fragile on other compilers. - - - - - 25b0b404 by Arsen Arsenović at 2024-05-07T14:40:24-04:00 Split out the C-- preprocessor, and make it pass -g0 Previously, C-- was processed with the C preprocessor program. This means that it inherited flags passed via -optc. A flag that is somewhat often passed through -optc is -g. At certain -g levels (>=2), GCC starts emitting defines *after* preprocessing, for the purposes of debug info generation. This is not useful for the C-- compiler, and, in fact, causes lexer errors. We can suppress this effect (safely, if supported) via -g0. As a workaround, in older versions of GCC (<=10), GCC only emitted defines if a certain set of -g*3 flags was passed. Newer versions check the debug level. For the former, we filter out those -g*3 flags and, for the latter, we specify -g0 on top of that. As a compatible and effective solution, this change adds a C-- preprocessor distinct from the C compiler and preprocessor, but that keeps its flags. The command line produced for C-- preprocessing now looks like: $pgmCmmP $optCs_without_g3 $g0_if_supported $optCmmP Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/24474 - - - - - 9b4129a5 by Andreas Klebinger at 2024-05-08T13:24:20-04:00 -fprof-late: Only insert cost centres on functions/non-workfree cafs. They are usually useless and doing so for data values comes with a large compile time/code size overhead. Fixes #24103 - - - - - 259b63d3 by Sebastian Graf at 2024-05-08T13:24:57-04:00 Simplifier: Preserve OccInfo on DataAlt fields when case binder is dead (#24770) See the adjusted `Note [DataAlt occ info]`. This change also has a positive repercussion on `Note [Combine case alts: awkward corner]`. Fixes #24770. We now try not to call `dataConRepStrictness` in `adjustFieldsIdInfo` when all fields are lazy anyway, leading to a 2% ghc/alloc decrease in T9675. Metric Decrease: T9675 - - - - - 31b28cdb by Sebastian Graf at 2024-05-08T13:24:57-04:00 Kill seqRule, discard dead seq# in Prep (#24334) Discarding seq#s in Core land via `seqRule` was problematic; see #24334. So instead we discard certain dead, discardable seq#s in Prep now. See the updated `Note [seq# magic]`. This fixes the symptoms of #24334. - - - - - b2682534 by Rodrigo Mesquita at 2024-05-10T01:47:51-04:00 Document NcgImpl methods Fixes #19914 - - - - - 4d3acbcf by Zejun Wu at 2024-05-10T01:48:28-04:00 Make renamer to be more flexible with parens in the LHS of the rules We used to reject LHS like `(f a) b` in RULES and requires it to be written as `f a b`. It will be handy to allow both as the expression may be more readable with extra parens in some cases when infix operator is involved. Espceially when TemplateHaskell is used, extra parens may be added out of user's control and result in "valid" rules being rejected and there are not always ways to workaround it. Fixes #24621 - - - - - ab840ce6 by Ben Gamari at 2024-05-10T01:49:04-04:00 IPE: Eliminate dependency on Read Instead of encoding the closure type as decimal string we now simply represent it as an integer, eliminating the need for `Read` in `GHC.Internal.InfoProv.Types.peekInfoProv`. Closes #24504. ------------------------- Metric Decrease: T24602_perf_size size_hello_artifact ------------------------- - - - - - a9979f55 by Cheng Shao at 2024-05-10T01:49:43-04:00 testsuite: fix testwsdeque with recent clang This patch fixes compilation of testwsdeque.c with recent versions of clang, which will fail with the error below: ``` testwsdeque.c:95:33: error: warning: format specifies type 'long' but the argument has type 'void *' [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~~ ^ testwsdeque.c:95:39: error: warning: format specifies type 'int' but the argument has type 'StgWord' (aka 'unsigned long') [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~ ^~~ | %lu testwsdeque.c:133:42: error: error: incompatible function pointer types passing 'void (void *)' to parameter of type 'OSThreadProc *' (aka 'void *(*)(void *)') [-Wincompatible-function-pointer-types] 133 | createOSThread(&ids[n], "thief", thief, (void*)(StgWord)n); | ^~~~~ /workspace/ghc/_build/stage1/lib/../lib/x86_64-linux-ghc-9.11.20240502/rts-1.0.2/include/rts/OSThreads.h:193:51: error: note: passing argument to parameter 'startProc' here 193 | OSThreadProc *startProc, void *param); | ^ 2 warnings and 1 error generated. ``` - - - - - c2b33fc9 by Rodrigo Mesquita at 2024-05-10T01:50:20-04:00 Rename pre-processor invocation args Small clean up. Uses proper names for the various groups of arguments that make up the pre-processor invocation. - - - - - 2b1af08b by Cheng Shao at 2024-05-10T01:50:55-04:00 ghc-heap: fix typo in ghc-heap cbits - - - - - fc2d6de1 by Jade at 2024-05-10T21:07:16-04:00 Improve performance of Data.List.sort(By) This patch improves the algorithm to sort lists in base. It does so using two strategies: 1) Use a four-way-merge instead of the 'default' two-way-merge. This is able to save comparisons and allocations. 2) Use `(>) a b` over `compare a b == GT` and allow inlining and specialization. This mainly benefits types with a fast (>). Note that this *may* break instances with a *malformed* Ord instance where `a > b` is *not* equal to `compare a b == GT`. CLC proposal: https://github.com/haskell/core-libraries-committee/issues/236 Fixes #24280 ------------------------- Metric Decrease: MultiLayerModulesTH_Make T10421 T13719 T15164 T18698a T18698b T1969 T9872a T9961 T18730 WWRec T12425 T15703 ------------------------- - - - - - 1012e8aa by Matthew Pickering at 2024-05-10T21:07:52-04:00 Revert "ghcup-metadata: Drop output_name field" This reverts commit ecbf22a6ac397a791204590f94c0afa82e29e79f. This breaks the ghcup metadata generation on the nightly jobs. - - - - - daff1e30 by Jannis at 2024-05-12T13:38:35-04:00 Division by constants optimization - - - - - 413217ba by Andreas Klebinger at 2024-05-12T13:39:11-04:00 Tidy: Add flag to expose unfoldings if they take dictionary arguments. Add the flag `-fexpose-overloaded-unfoldings` to be able to control this behaviour. For ghc's boot libraries file size grew by less than 1% when it was enabled. However I refrained from enabling it by default for now. I've also added a section on specialization more broadly to the users guide. ------------------------- Metric Decrease: MultiLayerModulesTH_OneShot Metric Increase: T12425 T13386 hard_hole_fits ------------------------- - - - - - c5d89412 by Zubin Duggal at 2024-05-13T22:19:53-04:00 Don't store a GlobalRdrEnv in `mi_globals` for GHCi. GHCi only needs the `mi_globals` field for modules imported with :module +*SomeModule. It uses this field to make the top level environment in `SomeModule` available to the repl. By default, only the first target in the command line parameters is "star" loaded into GHCi. Other modules have to be manually "star" loaded into the repl. Storing the top level GlobalRdrEnv for each module is very wasteful, especially given that we will most likely never need most of these environments. Instead we store only the information needed to reconstruct the top level environment in a module, which is the `IfaceTopEnv` data structure, consisting of all import statements as well as all top level symbols defined in the module (not taking export lists into account) When a particular module is "star-loaded" into GHCi (as the first commandline target, or via an explicit `:module +*SomeModule`, we reconstruct the top level environment on demand using the `IfaceTopEnv`. - - - - - d65bf4a2 by Fendor at 2024-05-13T22:20:30-04:00 Add perf regression test for `-fwrite-if-simplified-core` - - - - - 2c0f8ddb by Andrei Borzenkov at 2024-05-13T22:21:07-04:00 Improve pattern to type pattern transformation (23739) `pat_to_type_pat` function now can handle more patterns: - TuplePat - ListPat - LitPat - NPat - ConPat Allowing these new constructors in type patterns significantly increases possible shapes of type patterns without `type` keyword. This patch also changes how lookups in `lookupOccRnConstr` are performed, because we need to fall back into types when we didn't find a constructor on data level to perform `ConPat` to type transformation properly. - - - - - be514bb4 by Cheng Shao at 2024-05-13T22:21:43-04:00 hadrian: fix hadrian building with ghc-9.10.1 - - - - - ad38e954 by Cheng Shao at 2024-05-13T22:21:43-04:00 linters: fix lint-whitespace compilation with ghc-9.10.1 - - - - - a593f284 by Andreas Klebinger at 2024-05-15T07:32:10-04:00 Expand the `inline` rule to look through casts/ticks. Fixes #24808 - - - - - b1e0c313 by Cheng Shao at 2024-05-15T07:32:46-04:00 testsuite: bump PartialDownSweep timeout to 5x on wasm32 - - - - - b2227487 by Fendor at 2024-05-15T17:14:06-04:00 Add Eq and Ord instance to `IfaceType` We add an `Ord` instance so that we can store `IfaceType` in a `Data.Map` container. This is required to deduplicate `IfaceType` while writing `.hi` files to disk. Deduplication has many beneficial consequences to both file size and memory usage, as the deduplication enables implicit sharing of values. See issue #24540 for more motivation. The `Ord` instance would be unnecessary if we used a `TrieMap` instead of `Data.Map` for the deduplication process. While in theory this is clerarly the better option, experiments on the agda code base showed that a `TrieMap` implementation has worse run-time performance characteristics. To the change itself, we mostly derive `Eq` and `Ord`. This requires us to change occurrences of `FastString` with `LexicalFastString`, since `FastString` has no `Ord` instance. We change the definition of `IfLclName` to a newtype of `LexicalFastString`, to make such changes in the future easier. Bump haddock submodule for IfLclName changes - - - - - d368f9a6 by Fendor at 2024-05-15T17:14:06-04:00 Move out LiteralMap to avoid cyclic module dependencies - - - - - 2fcc09fd by Fendor at 2024-05-15T17:14:06-04:00 Add deduplication table for `IfaceType` The type `IfaceType` is a highly redundant, tree-like data structure. While benchmarking, we realised that the high redundancy of `IfaceType` causes high memory consumption in GHCi sessions when byte code is embedded into the `.hi` file via `-fwrite-if-simplified-core` or `-fbyte-code-and-object-code`. Loading such `.hi` files from disk introduces many duplicates of memory expensive values in `IfaceType`, such as `IfaceTyCon`, `IfaceTyConApp`, `IA_Arg` and many more. We improve the memory behaviour of GHCi by adding an additional deduplication table for `IfaceType` to the serialisation of `ModIface`, similar to how we deduplicate `Name`s and `FastString`s. When reading the interface file back, the table allows us to automatically share identical values of `IfaceType`. To provide some numbers, we evaluated this patch on the agda code base. We loaded the full library from the `.hi` files, which contained the embedded core expressions (`-fwrite-if-simplified-core`). Before this patch: * Load time: 11.7 s, 2.5 GB maximum residency. After this patch: * Load time: 7.3 s, 1.7 GB maximum residency. This deduplication has the beneficial side effect to additionally reduce the size of the on-disk interface files tremendously. For example, on agda, we reduce the size of `.hi` files (with `-fwrite-if-simplified-core`): * Before: 101 MB on disk * Now: 24 MB on disk This has even a beneficial side effect on the cabal store. We reduce the size of the store on disk: * Before: 341 MB on disk * Now: 310 MB on disk Note, none of the dependencies have been compiled with `-fwrite-if-simplified-core`, but `IfaceType` occurs in multiple locations in a `ModIface`. We also add IfaceType deduplication table to .hie serialisation and refactor .hie file serialisation to use the same infrastrucutre as `putWithTables`. Bump haddock submodule to accomodate for changes to the deduplication table layout and binary interface. - - - - - 36aa7cf1 by Fendor at 2024-05-15T17:14:06-04:00 Add run-time configurability of `.hi` file compression Introduce the flag `-fwrite-if-compression=<n>` which allows to configure the compression level of writing .hi files. The motivation is that some deduplication operations are too expensive for the average use case. Hence, we introduce multiple compression levels with variable impact on performance, but still reduce the memory residency and `.hi` file size on disk considerably. We introduce three compression levels: * `1`: `Normal` mode. This is the least amount of compression. It deduplicates only `Name` and `FastString`s, and is naturally the fastest compression mode. * `2`: `Safe` mode. It has a noticeable impact on .hi file size and is marginally slower than `Normal` mode. In general, it should be safe to always use `Safe` mode. * `3`: `Full` deduplication mode. Deduplicate as much as we can, resulting in minimal .hi files, but at the cost of additional compilation time. Reading .hi files doesn't need to know the initial compression level, and can always deserialise a `ModIface`, as we write out a byte that indicates the next value has been deduplicated. This allows users to experiment with different compression levels for packages, without recompilation of dependencies. Note, the deduplication also has an additional side effect of reduced memory consumption to implicit sharing of deduplicated elements. See https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for example where that matters. ------------------------- Metric Decrease: MultiLayerModulesDefsGhciWithCore T16875 T21839c T24471 hard_hole_fits libdir ------------------------- - - - - - 1e63a6fb by Matthew Pickering at 2024-05-15T17:14:07-04:00 Introduce regression tests for `.hi` file sizes Add regression tests to track how `-fwrite-if-compression` levels affect the size of `.hi` files. - - - - - 639d742b by M Farkas-Dyck at 2024-05-15T17:14:49-04:00 TTG: ApplicativeStatement exist only in Rn and Tc Co-Authored-By: romes <rodrigo.m.mesquita at gmail.com> - - - - - aa7b336b by Jade at 2024-05-15T23:06:17-04:00 Documentation: Improve documentation for symbols exported from System.IO - - - - - c561de8f by Jade at 2024-05-15T23:06:54-04:00 Improve suggestions for language extensions - When suggesting Language extensions, also suggest Extensions which imply them - Suggest ExplicitForAll and GADTSyntax instead of more specific extensions - Rephrase suggestion to include the term 'Extension' - Also moves some flag specific definitions out of Session.hs into Flags.hs (#24478) Fixes: #24477 Fixes: #24448 Fixes: #10893 - - - - - 4c7ae2a1 by Andreas Klebinger at 2024-05-15T23:07:30-04:00 Testsuite: Check if llvm assembler is available for have_llvm - - - - - bc672166 by Torsten Schmits at 2024-05-15T23:08:06-04:00 refactor quadratic search in warnMissingHomeModules - - - - - 7875e8cb by Torsten Schmits at 2024-05-15T23:08:06-04:00 add test that runs MakeDepend on thousands of modules - - - - - b84b91f5 by Adam Gundry at 2024-05-16T15:32:06-04:00 Representation-polymorphic HasField (fixes #22156) This generalises the HasField class to support representation polymorphism, so that instead of type HasField :: forall {k} . k -> Type -> Type -> Constraint we have type HasField :: forall {k} {r_rep} {a_rep} . k -> TYPE r_rep -> TYPE a_rep -> Constraint - - - - - 05285090 by Matthew Pickering at 2024-05-16T15:32:43-04:00 Bump os-string submodule to 2.0.2.2 Closes #24786 - - - - - 886ab43a by Cheng Shao at 2024-05-17T01:34:50-04:00 rts: do not prefetch mark_closure bdescr in non-moving gc when ASSERTS_ENABLED This commit fixes a small an oversight in !12148: the prefetch logic in non-moving GC may trap in debug RTS because it calls Bdescr() for mark_closure which may be a static one. It's fine in non-debug RTS because even invalid bdescr addresses are prefetched, they will not cause segfaults, so this commit implements the most straightforward fix: don't prefetch mark_closure bdescr when assertions are enabled. - - - - - b38dcf39 by Teo Camarasu at 2024-05-17T01:34:50-04:00 rts: Allocate non-moving segments with megablocks Non-moving segments are 8 blocks long and need to be aligned. Previously we serviced allocations by grabbing 15 blocks, finding an aligned 8 block group in it and returning the rest. This proved to lead to high levels of fragmentation as a de-allocating a segment caused an 8 block gap to form, and this could not be reused for allocation. This patch introduces a segment allocator based around using entire megablocks to service segment allocations in bulk. When there are no free segments, we grab an entire megablock and fill it with aligned segments. As the megablock is free, we can easily guarantee alignment. Any unused segments are placed on a free list. It only makes sense to free segments in bulk when all of the segments in a megablock are freeable. After sweeping, we grab the free list, sort it, and find all groups of segments where they cover the megablock and free them. This introduces a period of time when free segments are not available to the mutator, but the risk that this would lead to excessive allocation is low. Right after sweep, we should have an abundance of partially full segments, and this pruning step is relatively quick. In implementing this we drop the logic that kept NONMOVING_MAX_FREE segments on the free list. We also introduce an eventlog event to log the amount of pruned/retained free segments. See Note [Segment allocation strategy] Resolves #24150 ------------------------- Metric Decrease: T13253 T19695 ------------------------- - - - - - 710665bd by Cheng Shao at 2024-05-17T01:35:30-04:00 rts: fix I/O manager compilation errors for win32 target This patch fixes I/O manager compilation errors for win32 target discovered when cross-compiling to win32 using recent clang: ``` rts/win32/ThrIOManager.c:117:7: error: error: call to undeclared function 'is_io_mng_native_p'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 117 | if (is_io_mng_native_p ()) { | ^ | 117 | if (is_io_mng_native_p ()) { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/fs.c:143:28: error: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] 143 | int setErrNoFromWin32Error () { | ^ | void | 143 | int setErrNoFromWin32Error () { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:227:9: error: error: call to undeclared function 'interruptIOManagerEvent'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 227 | interruptIOManagerEvent (); | ^ | 227 | interruptIOManagerEvent (); | ^ rts/win32/ConsoleHandler.c:227:9: error: note: did you mean 'getIOManagerEvent'? | 227 | interruptIOManagerEvent (); | ^ rts/include/rts/IOInterface.h:27:10: error: note: 'getIOManagerEvent' declared here 27 | void * getIOManagerEvent (void); | ^ | 27 | void * getIOManagerEvent (void); | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:196:9: error: error: call to undeclared function 'setThreadLabel'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/win32/ConsoleHandler.c:196:9: error: note: did you mean 'postThreadLabel'? | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/eventlog/EventLog.h:118:6: error: note: 'postThreadLabel' declared here 118 | void postThreadLabel(Capability *cap, | ^ | 118 | void postThreadLabel(Capability *cap, | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) ``` - - - - - 28b9cee0 by Rodrigo Mesquita at 2024-05-17T01:36:05-04:00 configure: Check C99-compat for Cmm preprocessor Fixes #24815 - - - - - 8927e0c3 by Andreas Klebinger at 2024-05-17T01:36:41-04:00 Ensure `tcHasFixedRuntimeRep (# #)` returns True. - - - - - 04179044 by doyougnu at 2024-05-17T09:00:32-04:00 testsuite: make find_so regex less general Closes #24759 Background. In MR !12372 we began tracking shared object files and directories sizes for dependencies. However, this broke release builds because release builds alter the filenames swapping "in-place" for a hash. This was not considered in the MR and thus broke release pipelines. Furthermore, the rts_so test was found to be wildly varying and was therefore disabled in !12561. This commit fixes both of these issues: - fix the rts_so test by making the regex less general, now the rts_so test and all other foo.so tests must match "libHS<some-lib>-<version>-<hash|'in-place>-<ghc>". This prevents the rts_so test from accidentally matching different rts variants such as rts_threaded, which was the cause of the wild swings after !12372. - add logic to match either a hash or the string in-place. This should make the find_so function build agnostic. - - - - - 0962b50d by Andreas Klebinger at 2024-05-17T09:01:08-04:00 TagAnalysis: Treat all bottom ids as tagged during analysis. Ticket #24806 showed that we also need to treat dead end thunks as tagged during the analysis. - - - - - 7eb9f184 by Ben Gamari at 2024-05-17T11:23:37-04:00 Remove haddock submodule In preparation for merge into the GHC, as proposed in #23178. - - - - - 47b14dcc by Fendor at 2024-05-17T11:28:17-04:00 Adapt to `IfLclName` newtype changes (cherry picked from commit a711607e29b925f3d69e27c5fde4ba655c711ff1) - - - - - 6cc6681d by Fendor at 2024-05-17T11:28:17-04:00 Add IfaceType deduplication table to interface file serialisation Although we do not really need it in the interface file serialisation, as the deserialisation uses `getWithUserData`, we need to mirror the structure `getWithUserData` expects. Thus, we write essentially an empty `IfaceType` table at the end of the file, as the interface file doesn't reference `IfaceType`. (cherry picked from commit c9bc29c6a708483d2abc3d8ec9262510ce87ca61) - - - - - b9721206 by Ben Gamari at 2024-05-17T11:30:22-04:00 ghc-tags.yaml: Initial commit - - - - - 074e7d8f by Ben Gamari at 2024-05-17T11:31:29-04:00 fourmolu: Add configuration - - - - - 151b1736 by Ben Gamari at 2024-05-17T11:32:52-04:00 Makefile: Rework for use by haddock developers Previously the Makefile was present only for GHC's old make-based build system. Now since the make-based build system is gone we can use it for more useful ends. - - - - - a7dcf13b by Ben Gamari at 2024-05-17T11:36:14-04:00 Reformat with fourmolu Using previously-added configuration and `fourmolu -i .` Note that we exclude the test-cases (`./{hoogle,html-hypsrc,latex}-test`) as they are sensitive to formatting. - - - - - 0ea6017b by Ben Gamari at 2024-05-17T11:40:04-04:00 Add 'utils/haddock/' from commit 'a7dcf13bfbb97b20e75cc8ce650e2bb628db4660' git-subtree-dir: utils/haddock git-subtree-mainline: 7eb9f1849b1c72a1c61dee88462b4244550406f3 git-subtree-split: a7dcf13bfbb97b20e75cc8ce650e2bb628db4660 - - - - - aba1d304 by Hécate Moonlight at 2024-05-17T11:40:48-04:00 Add exceptions to the dangling notes list - - - - - 527bfbfb by Hécate Moonlight at 2024-05-17T11:40:52-04:00 Add haddock to the whitespace lint ignore list - - - - - 43274677 by Ben Gamari at 2024-05-17T11:41:20-04:00 git-blame-ignore-revs: Ignore haddock reformatting - - - - - 0e679e37 by Fendor at 2024-05-18T00:27:24-04:00 Pass cpp options to the CC builder in hadrian - - - - - bb40244e by Sylvain Henry at 2024-05-18T00:28:06-04:00 JS: fix allocation constant (fix #24746) - - - - - 646d30ab by Jade at 2024-05-18T19:23:31+02:00 Add highlighting for inline-code snippets in haddock - - - - - 64459a3e by Hécate Moonlight at 2024-05-19T08:42:27-04:00 haddock: Add a .readthedocs.yml file for online documentation - - - - - 7d3d9bbf by Serge S. Gulin at 2024-05-19T18:47:05+00:00 Unicode: General Category size test (related #24789) Added trivial size performance test which involves unicode general category usage via `read`. The `read` itself uses general category to detect spaces. The purpose for this test is to measure outcome of applying improvements at General Category representation in code discussed at #24789. - - - - - 8e04efcf by Alan Zimmerman at 2024-05-19T21:29:34-04:00 EPA: Remove redundant code Remove unused epAnnAnns function various cases for showAstData that no longer exist - - - - - 071d7a1e by Rodrigo Mesquita at 2024-05-20T10:55:16-04:00 Improve docs on closed type families in hs-boots Fixes #24776 - - - - - d9e2c119 by Torsten Schmits at 2024-05-20T10:55:52-04:00 Use default deviation for large-project test This new performance test has the purpose of detecting regressions in complexity in relation to the number of modules in a project, so 1% deviation is way too small to avoid false positives. - - - - - 20b0136a by Ben Gamari at 2024-05-22T00:31:39-04:00 ghcup-metadata: Various fixes from 9.10.1 Use Debian 12/x86-64, Debian 10/aarch64, and Debian 11/aarch64 bindists where possible. - - - - - 6838a7c3 by Sylvain Henry at 2024-05-22T00:32:23-04:00 Reverse arguments to stgCallocBytes (fix #24828) - - - - - f50f46c3 by Fendor at 2024-05-22T00:32:59-04:00 Add log messages for Iface serialisation compression level Fix the label of the number of 'IfaceType' entries in the log message. Add log message for the compression level that is used to serialise a an interface file. Adds `Outputable` instance for 'CompressionIFace'. - - - - - 3bad5d55 by Hécate Moonlight at 2024-05-22T00:33:40-04:00 base: Update doctests outputs ghc-internal: Update doctests outputs - - - - - 9317c6fb by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix the testsuites of the haddock-library - Apply all the metadata revisions from Hackage to the cabal file. - Fix the `ParserSpec.hs` file in the `spec` testsuite of haddock-library. - Make `CHANGES.md` an extra-doc-file instead of an extra-source-file. - - - - - 54073b02 by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix parser of @since pragma The testsuite contained tests for annotations of the form `@since foo-bar-0.5.0`, but the parser was written incorrectly. - - - - - ede6ede3 by Matthew Pickering at 2024-05-22T00:34:57-04:00 Fix nightly pages job It seems likely broken by 9f99126a which moved `index.html` from the root folder into `docs/` folder. Fixes #24840 - - - - - b7bcf729 by Cheng Shao at 2024-05-22T00:35:32-04:00 autoconf: remove unused context diff check This patch removes redundant autoconf check for the context diff program given it isn't actually been used anywhere, especially since make removal. - - - - - ea2fe66e by Hécate Moonlight at 2024-05-22T00:36:13-04:00 haddock: Rework the contributing guide - - - - - 0f302a94 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 haddock: Add module relationships diagrams of haddock-api and haddock-library - - - - - d1a9f34f by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add instructions - - - - - b880ee80 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add SVG outputs - - - - - 6d7e6ad8 by Ben Gamari at 2024-05-22T13:40:05-04:00 rts: Fix size of StgOrigThunkInfo frames Previously the entry code of the `stg_orig_thunk` frame failed to account for the size of the profiling header as it hard-coded the frame size. Fix this. Fixes #24809. - - - - - c645fe40 by Fendor at 2024-05-22T13:40:05-04:00 Add regression test T24809 for stg_orig_thunk_info_frame size - - - - - 4181aa40 by Andreas Klebinger at 2024-05-22T13:40:42-04:00 bindists: Check for existence of share folder before trying to copy it. This folder isn't distributed in windows bindists A lack of doing so resulted us copying loads of files twice. - - - - - d216510e by Matthew Pickering at 2024-05-22T13:40:42-04:00 Remove ad-hoc installation of mingw toolchain in relocatable bindists This reverts 616ac30026e8dd7d2ebb98d92dde071eedf5d951 The choice about whether to install mingw is taken in the installation makefile. This is also broken on non-windows systems. The actual issue was the EnableDistroToolchain variable wasn't declared in mk/config.mk and therefore the check to install mingw was failing. - - - - - 7b4c1998 by Cheng Shao at 2024-05-22T21:52:52-04:00 testsuite: fix T17920 for wasm backend T17920 was marked as fragile on wasm before; it can be trivially fixed by avoiding calling variadic printf() in cmm. - - - - - c739383b by Cheng Shao at 2024-05-22T21:53:29-04:00 testsuite: bump T22744 timeout to 5x - - - - - c4c6d714 by Cheng Shao at 2024-05-22T21:54:06-04:00 testsuite: don't attempt to detect host cpu features when testing cross ghc The testsuite driver CPU feature detection logic only detects host CPU and only makes sense when we are not testing a cross GHC. - - - - - 3d9e4ce6 by Simon Peyton Jones at 2024-05-22T21:54:43-04:00 Better skolemisation As #24810 showed, it is (a little) better to skolemise en-bloc, so that Note [Let-bound skolems] fires more often. See Note [Skolemisation en bloc] in GHC.Tc.Utils.Instantiate. - - - - - a3cd3a1d by Ryan Scott at 2024-05-22T21:55:19-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. - - - - - 4bb2a7cc by Hécate Moonlight at 2024-05-22T21:55:59-04:00 [base] Document the memory overhead of ByteArray Add a diagram that shows the constituent parts of a ByteArray and their memory overhead. - - - - - 8b2a016a by Hécate Moonlight at 2024-05-22T21:56:38-04:00 Haddock: Add MR template for Haddock - - - - - ead75532 by Peter Trommler at 2024-05-23T02:28:05-04:00 PPC: Support ELF v2 on powerpc64 big-endian Detect ELF v2 on PowerPC 64-bit systems. Check for `_CALL_ELF` preprocessor macro. Fixes #21191 - - - - - 9d4c10f2 by Hécate Kleidukos at 2024-05-23T02:28:44-04:00 gitlab: Add @Kleidukos to CODEOWNERS for utils/haddock - - - - - 28e64170 by Preetham Gujjula at 2024-05-23T07:20:48-04:00 haddock: Add cabal-fmt to tools for `make style` - - - - - 00126a89 by Andrei Borzenkov at 2024-05-23T07:21:24-04:00 haddock: fix verbosity option parsing - - - - - a3e0b68b by Ryan Hendrickson at 2024-05-23T15:52:03-04:00 base: specify tie-breaking behavior of min, max, and related list/Foldable functions - - - - - bdcc0f37 by doyougnu at 2024-05-24T07:51:18-04:00 cmm: add word <-> double/float bitcast - closes: #25331 This is the last step in the project plan described in #25331. This commit: - adds bitcast operands for x86_64, LLVM, aarch64 - For PPC and i386 we resort to using the cmm implementations - renames conversion MachOps from Conv to Round|Truncate - - - - - f0d257f7 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor Some functions in StgToByteCode were filtering out void arguments. However, StgToByteCode is called after unarisation: the void arguments should have been removed earlier. Instead of filtering out, we assert that the args are non-void. - - - - - 03137fd2 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor `layoutNativeCall` was always called with a `primRepCmmType platform` callback. Hence we can put it inside of `layoutNativeCall` rather than repeat it. - - - - - 27c430f3 by David Binder at 2024-05-24T07:52:38-04:00 haddock: Remove compatibility shims for GHC < 8.4 from haddock-library - - - - - 8dd8a076 by Cheng Shao at 2024-05-24T07:53:14-04:00 compiler: avoid saving foreign call target to local when there are no caller-save GlobalRegs This patch makes the STG->Cmm backend avoid saving foreign call target to local when there are no caller-save GlobalRegs. Since 321941a8ebe25192cdeece723e1058f2f47809ea, when we lower a foreign call, we unconditionally save the foreign call target to a temporary local first, then rely on cmmSink to clean it up later, which only happens with -fcmm-sink (implied by -O) and not in unoptimized code. And this is troublesome for the wasm backend NCG, which needs to infer a foreign call target symbol's type signature from the Cmm call site. Previously, the NCG has been emitting incorrect type signatures for unoptimized code, which happens to work with `wasm-ld` most of the time, but this is never future-proof against upstream toolchain updates, and it causes horrible breakages when LTO objects are included in linker input. Hence this patch. - - - - - 986df1ab by Cheng Shao at 2024-05-24T07:53:14-04:00 testsuite: add callee-no-local regression test - - - - - 52d62e2a by Sylvain Henry at 2024-05-24T07:53:57-04:00 Fix HasCallStack leftovers from !12514 / #24726 - - - - - c5e00c35 by crumbtoo at 2024-05-24T07:54:38-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 - - - - - bd323b0e by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Ensure that CHANGELOG is included in extra-source-files This was missed in the `ghc-internal` split. Closes #24831. - - - - - 1bfd32e8 by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Fix changelog reference to setBacktraceMechanismState (cherry picked from commit b63f7ba01fdfd98a01d2f0dec8d9262b3e595c5d) - - - - - 43e8e4f3 by Sylvain Henry at 2024-05-24T12:16:43-04:00 Float/double unboxed literal support for HexFloatLiterals (fix #22155) - - - - - 4a7f4713 by Fendor at 2024-05-24T12:17:19-04:00 Improve test labels for binary interface file size tests Test labels for binary interface file sizes are hard to read and overly verbose at the same time. Extend the name for the metric title, but shorten it in the actual comparison table. - - - - - 14e554cf by Zubin Duggal at 2024-05-24T12:17:55-04:00 Revert "Fix haskell/haddock#783 Don't show button if --quickjump not present" This reverts commit 7776566531e72c415f66dd3b13da9041c52076aa. - - - - - f56838c3 by Ben Gamari at 2024-05-24T12:17:55-04:00 Fix default hyperlinked sources pattern Previously this didn't include the `%M` token which manifested as broken links to the hyperlinked sources of reexports of declarations defined in other packages. Fixes haddock#1628. (cherry picked from commit 1432bcc943d41736eca491ecec4eb9a6304dab36) - - - - - 42efa62c by Ben Gamari at 2024-05-24T12:17:55-04:00 Make DocPaths a proper data type (cherry picked from commit 7f3a5c4da0023ae47b4c376c9b1ea2d706c94d8c) - - - - - 53d9ceb3 by Ben Gamari at 2024-05-24T12:17:55-04:00 haddock: Bump version to 2.30 (cherry picked from commit 994989ed3d535177e57b778629726aeabe8c7602) - - - - - e4db1112 by Zubin Duggal at 2024-05-24T12:17:55-04:00 haddock-api: allow base 4.20 and ghc 9.11 - - - - - e294f7a2 by PHO at 2024-05-24T12:17:55-04:00 Add a flag "threaded" for building haddock with the threaded RTS GHC isn't guaranteed to have a threaded RTS. There should be a way to build it with the vanilla one. (cherry picked from commit 75a94e010fb5b0236c670d22b04f5472397dc15d) - - - - - 51165bc9 by Andreas Klebinger at 2024-05-25T10:58:03-04:00 Update ticky counter event docs. Add the info about the info table address and json fields. Fixes #23200 - - - - - 98597ad5 by Sylvain Henry at 2024-05-25T10:58:45-04:00 Export extractPromotedList (#24866) This can be useful in plugins. - - - - - 228dcae6 by Teo Camarasu at 2024-05-28T13:12:24+00:00 template-haskell: Move wired-ins to ghc-internal Thus we make `template-haskell` reinstallable and keep it as the public API for Template Haskell. All of the wired-in identifiers are moved to `ghc-internal`. This necessitates also moving much of `ghc-boot-th` into `ghc-internal`. These modules are then re-exported from `ghc-boot-th` and `template-haskell`. To avoid a dependency on `template-haskell` from `lib:ghc`, we instead depend on the TH ASTs via `ghc-boot-th`. As `template-haskell` no longer has special status, we can drop the logic adding an implicit dependency on `template-haskell` when using TH. We can also drop the `template-haskell-next` package, which was previously used when bootstrapping. When bootstrapping, we need to vendor the TH AST modules from `ghc-internal` into `ghc-boot-th`. This is controlled by the `bootstrap` cabal flag as before. See Note [Bootstrapping Template Haskell]. We split out a GHC.Internal.TH.Lift module resolving #24752. This module is only built when not bootstrapping. Resolves #24703 ------------------------- Metric Increase: ghc_boot_th_dir ghc_boot_th_so ------------------------- - - - - - 62dded28 by Teo Camarasu at 2024-05-28T13:12:24+00:00 testsuite: mark tests broken by #24886 Now that `template-haskell` is no longer wired-in. These tests are triggering #24886, and so need to be marked broken. - - - - - 3ca72ad9 by Cheng Shao at 2024-05-30T02:57:06-04:00 rts: fix missing function prototypes in ClosureMacros.h - - - - - e0029e3d by Andreas Klebinger at 2024-05-30T02:57:43-04:00 UnliftedFFITypes: Allow `(# #)` as argument when it's the only argument. This allows representing functions like: int foo(void); to be imported like this: foreign import ccall "a_number_c" c_number :: (# #) -> Int64# Which can be useful when the imported function isn't implicitly stateful. - - - - - d0401335 by Matthew Pickering at 2024-05-30T02:58:19-04:00 ci: Update ci-images commit for fedora38 image The fedora38 nightly job has been failing for quite a while because `diff` was no longer installed. The ci-images bump explicitly installs `diffutils` into these images so hopefully they now pass again. - - - - - 3c97c74a by Jan Hrček at 2024-05-30T02:58:58-04:00 Update exactprint docs - - - - - 77760cd7 by Jan Hrček at 2024-05-30T02:58:58-04:00 Incorporate review feedback - - - - - 87591368 by Jan Hrček at 2024-05-30T02:58:58-04:00 Remove no longer relevant reference to comments - - - - - 05f4f142 by Jan Hrček at 2024-05-30T02:58:59-04:00 Replace outdated code example - - - - - 45a4a5f3 by Andreas Klebinger at 2024-05-30T02:59:34-04:00 Reword error resulting from missing -XBangPatterns. It can be the result of either a bang pattern or strict binding, so now we say so instead of claiming it must be a bang pattern. Fixes #21032 - - - - - e17f2df9 by Cheng Shao at 2024-05-30T03:00:10-04:00 testsuite: bump MultiLayerModulesDefsGhciReload timeout to 10x - - - - - 7a660042 by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - c77a48af by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 05c4fafb by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 5ff83bfc by Sylvain Henry at 2024-05-30T14:43:10-04:00 JS: remove useless h$CLOCK_REALTIME (#23202) - - - - - 95ef2d58 by Matthew Pickering at 2024-05-30T14:43:47-04:00 ghcup-metadata: Fix metadata generation There were some syntax errors in the generation script which were preventing it from running. I have tested this with: ``` nix shell --extra-experimental-features nix-command -f .gitlab/rel_eng -c ghcup-metadata --metadata ghcup-0.0.7.yaml --date="2024-05-27" --pipeline-id=95534 --version=9.11.20240525 ``` which completed successfully. - - - - - 1bc66ee4 by Jakob Bruenker at 2024-05-30T14:44:22-04:00 Add diagrams to Arrows documentation This adds diagrams to the documentation of Arrows, similar to the ones found on https://www.haskell.org/arrows/. It does not add diagrams for ArrowChoice for the time being, mainly because it's not clear to me how to visually distinguish them from the ones for Arrow. Ideally, you might want to do something like highlight the arrows belonging to the same tuple or same Either in common colors, but that's not really possible with unicode. - - - - - d10a1c65 by Matthew Craven at 2024-05-30T23:35:48-04:00 Make UnsafeSNat et al. into pattern synonyms ...so that they do not cause coerce to bypass the nominal role on the corresponding singleton types when they are imported. See Note [Preventing unsafe coercions for singleton types] and the discussion at #23478. This also introduces unsafeWithSNatCo (and analogues for Char and Symbol) so that users can still access the dangerous coercions that importing the real constructors would allow, but only in a very localized way. - - - - - 0958937e by Cheng Shao at 2024-05-30T23:36:25-04:00 hadrian: build C/C++ with split sections when enabled When split sections is enabled, ensure -fsplit-sections is passed to GHC as well when invoking GHC to compile C/C++; and pass -ffunction-sections -fdata-sections to gcc/clang when compiling C/C++ with the hadrian Cc builder. Fixes #23381. - - - - - 02b1f91e by Cheng Shao at 2024-05-30T23:36:25-04:00 driver: build C/C++ with -ffunction-sections -fdata-sections when split sections is enabled When -fsplit-sections is passed to GHC, pass -ffunction-sections -fdata-sections to gcc/clang when building C/C++. Previously, -fsplit-sections was only respected by the NCG/LLVM backends, but not the unregisterised backend; the GHC driver did not pass -fdata-sections and -ffunction-sections to the C compiler, which resulted in excessive executable sizes. Fixes #23381. ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - fd47e2e3 by Cheng Shao at 2024-05-30T23:37:00-04:00 testsuite: mark process005 as fragile on JS - - - - - 34a04ea1 by Matthew Pickering at 2024-05-31T06:08:36-04:00 Add -Wderiving-typeable to -Wall Deriving `Typeable` does nothing, and it hasn't done for a long while. There has also been a warning for a long while which warns you about uselessly deriving it but it wasn't enabled in -Wall. Fixes #24784 - - - - - 75fa7b0b by Matthew Pickering at 2024-05-31T06:08:36-04:00 docs: Fix formatting of changelog entries - - - - - 303c4b33 by Preetham Gujjula at 2024-05-31T06:09:21-04:00 docs: Fix link to injective type families paper Closes #24863 - - - - - df97e9a6 by Ben Gamari at 2024-05-31T06:09:57-04:00 ghc-internal: Fix package description The previous description was inherited from `base` and was inappropriate for `ghc-internal`. Also fix the maintainer and bug reporting fields. Closes #24906. - - - - - bf0737c0 by Cheng Shao at 2024-05-31T06:10:33-04:00 compiler: remove ArchWasm32 special case in cmmDoCmmSwitchPlans This patch removes special consideration for ArchWasm32 in cmmDoCmmSwitchPlans, which means the compiler will now disable cmmImplementSwitchPlans for wasm unreg backend, just like unreg backend of other targets. We enabled it in the past to workaround some compile-time panic in older versions of LLVM, but those panics are no longer present, hence no need to keep this workaround. - - - - - 7eda4bd2 by Cheng Shao at 2024-05-31T15:52:04-04:00 utils: add hie.yaml config file for ghc-config Add hie.yaml to ghc-config project directory so it can be edited using HLS. - - - - - 1e5752f6 by Cheng Shao at 2024-05-31T15:52:05-04:00 hadrian: handle findExecutable "" gracefully hadrian may invoke findExecutable "" at run-time due to a certain program is not found by configure script. Which is fine and findExecutable is supposed to return Nothing in this case. However, on Windows there's a directory bug that throws an exception (see https://github.com/haskell/directory/issues/180), so we might as well use a wrapper for findExecutable and handle exceptions gracefully. - - - - - 4eb5ad09 by Cheng Shao at 2024-05-31T15:52:05-04:00 configure: do not set LLC/OPT/LLVMAS fallback values when FIND_LLVM_PROG fails When configure fails to find LLC/OPT/LLVMAS within supported version range, it used to set "llc"/"opt"/"clang" as fallback values. This behavior is particularly troublesome when the user has llc/opt/clang with other versions in their PATH and run the testsuite, since hadrian will incorrectly assume have_llvm=True and pass that to the testsuite driver, resulting in annoying optllvm test failures (#23186). If configure determines llc/opt/clang wouldn't work, then we shouldn't pretend it'll work at all, and the bindist configure will invoke FIND_LLVM_PROG check again at install time anyway. - - - - - 5f1afdf7 by Sylvain Henry at 2024-05-31T15:52:52-04:00 Introduce UniqueSet and use it to replace 'UniqSet Unique' 'UniqSet Unique' represents a set of uniques as a 'Map Unique Unique', which is wasting space (associated key/value are always the same). Fix #23572 and #23605 - - - - - e0aa42b9 by crumbtoo at 2024-05-31T15:53:33-04:00 Improve template-haskell haddocks Closes #15822 - - - - - ae170155 by Olivier Benz at 2024-06-01T09:35:17-04:00 Bump max LLVM version to 19 (not inclusive) - - - - - 92aa65ea by Matthew Pickering at 2024-06-01T09:35:17-04:00 ci: Update CI images to test LLVM 18 The debian12 image in this commit has llvm 18 installed. - - - - - adb1fe42 by Serge S. Gulin at 2024-06-01T09:35:53-04:00 Unicode: make ucd2haskell build-able again ucd2haskell tool used streamly library which version in cabal was out of date. It is updated to the latest version at hackage with deprecated parts rewritten. Also following fixes were applied to existing code in suppose that from its last run the code structure was changed and now it was required to be up to date with actual folder structures: 1. Ghc module path environment got a suffix with `src`. 2. Generated code got 2.1 `GHC.Internal` prefix for `Data.*`. 2.2 `GHC.Unicode.Internal` swapped on `GHC.Internal.Unicode` according to actual structure. - - - - - ad56fd84 by Jade at 2024-06-01T09:36:29-04:00 Replace 'NB' with 'Note' in error messages - - - - - 6346c669 by Cheng Shao at 2024-06-01T09:37:04-04:00 compiler: fix -ddump-cmm-raw when compiling .cmm This patch fixes missing -ddump-cmm-raw output when compiling .cmm, which is useful for debugging cmm related codegen issues. - - - - - 1c834ad4 by Ryan Scott at 2024-06-01T09:37:40-04:00 Print namespace specifiers in FixitySig's Outputable instance For whatever reason, the `Outputable` instance for `FixitySig` simply did not print out namespace specifiers, leading to the confusing `-ddump-splices` output seen in #24911. This patch corrects this oversight. Fixes #24911. - - - - - cf49fb5f by Sylvain Henry at 2024-06-01T09:38:19-04:00 Configure: display C++ compiler path - - - - - f9c1ae12 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable PIC for in-tree GMP on wasm32 This patch disables PIC for in-tree GMP on wasm32 target. Enabling PIC unconditionally adds undesired code size and runtime overhead for wasm32. - - - - - 1a32f828 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable in-tree gmp fft code path for wasm32 This patch disables in-tree GMP FFT code paths for wasm32 target in order to give up some performance of multiplying very large operands in exchange for reduced code size. - - - - - 06277d56 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: build in-tree GMP with malloc-notreentrant on wasm32 This patch makes hadrian build in-tree GMP with the --enable-alloca=malloc-notreentrant configure option. We will only need malloc-reentrant when we have threaded RTS and SMP support on wasm32, which will take some time to happen, before which we should use malloc-notreentrant to avoid undesired runtime overhead. - - - - - 9f614270 by ARATA Mizuki at 2024-06-02T14:02:35-04:00 Set package include paths when assembling .S files Fixes #24839. Co-authored-by: Sylvain Henry <hsyl20 at gmail.com> - - - - - 4998a6ed by Alex Mason at 2024-06-03T02:09:29-04:00 Improve performance of genericWordQuotRem2Op (#22966) Implements the algorithm from compiler-rt's udiv128by64to64default. This rewrite results in a roughly 24x improvement in runtime on AArch64 (and likely any other arch that uses it). - - - - - ae50a8eb by Cheng Shao at 2024-06-03T02:10:05-04:00 testsuite: mark T7773 as fragile on wasm - - - - - c8ece0df by Fendor at 2024-06-03T19:43:22-04:00 Migrate `Finder` component to `OsPath`, fixed #24616 For each module in a GHCi session, we keep alive one `ModLocation`. A `ModLocation` is fairly inefficiently packed, as `String`s are expensive in memory usage. While benchmarking the agda codebase, we concluded that we keep alive around 11MB of `FilePath`'s, solely retained by `ModLocation`. We provide a more densely packed encoding of `ModLocation`, by moving from `FilePath` to `OsPath`. Further, we migrate the full `Finder` component to `OsPath` to avoid unnecessary transformations. As the `Finder` component is well-encapsulated, this requires only a minimal amount of changes in other modules. We introduce pattern synonym for 'ModLocation' which maintains backwards compatibility and avoids breaking consumers of 'ModLocation'. - - - - - 0cff083a by Cheng Shao at 2024-06-03T19:43:58-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. - - - - - 18f63970 by Sebastian Graf at 2024-06-04T05:05:27-04:00 Parser: Remove unused `apats` rule - - - - - 38757c30 by David Knothe at 2024-06-04T05:05:27-04:00 Implement Or Patterns (#22596) This commit introduces a new language extension, `-XOrPatterns`, as described in GHC Proposal 522. An or-pattern `pat1; ...; patk` succeeds iff one of the patterns `pat1`, ..., `patk` succeed, in this order. See also the summary `Note [Implmentation of OrPatterns]`. Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> - - - - - 395412e8 by Cheng Shao at 2024-06-04T05:06:04-04:00 compiler/ghci/rts: remove stdcall support completely We have formally dropped i386 windows support (#18487) a long time ago. The stdcall foreign call convention is only used by i386 windows, and the legacy logic around it is a significant maintenance burden for future work that adds arm64 windows support (#24603). Therefore, this patch removes stdcall support completely from the compiler as well as the RTS (#24883): - stdcall is still recognized as a FFI calling convention in Haskell syntax. GHC will now unconditionally emit a warning (-Wunsupported-calling-conventions) and treat it as ccall. - Apart from minimum logic to support the parsing and warning logic, all other code paths related to stdcall has been completely stripped from the compiler. - ghci only supports FFI_DEFAULT_ABI and ccall convention from now on. - FFI foreign export adjustor code on all platforms no longer handles the stdcall case and only handles ccall from now on. - The Win32 specific parts of RTS no longer has special code paths for stdcall. This commit is the final nail on the coffin for i386 windows support. Further commits will perform more housecleaning to strip the legacy code paths and pave way for future arm64 windows support. - - - - - d1fe9ab6 by Cheng Shao at 2024-06-04T05:06:04-04:00 rts: remove legacy i386 windows code paths This commit removes some legacy i386 windows related code paths in the RTS, given this target is no longer supported. - - - - - a605e4b2 by Cheng Shao at 2024-06-04T05:06:04-04:00 autoconf: remove i386 windows related logic This commit removes legacy i386 windows logic in autoconf scripts. - - - - - 91e5ac5e by Cheng Shao at 2024-06-04T05:06:04-04:00 llvm-targets: remove i386 windows support This commit removes i386 windows from llvm-targets and the script to generate it. - - - - - 65fe75a4 by Cheng Shao at 2024-06-04T05:06:04-04:00 libraries/utils: remove stdcall related legacy logic This commit removes stdcall related legacy logic in libraries and utils. ccall should be used uniformly for all supported windows hosts from now on. - - - - - d2a83302 by Cheng Shao at 2024-06-04T05:06:04-04:00 testsuite: adapt the testsuite for stdcall removal This patch adjusts test cases to handle the stdcall removal: - Some stdcall usages are replaced with ccall since stdcall doesn't make sense anymore. - We also preserve some stdcall usages, and check in the expected warning messages to ensure GHC always warn about stdcall usages (-Wunsupported-calling-conventions) as expected. - Error code testsuite coverage is slightly improved, -Wunsupported-calling-conventions is now tested. - Obsolete code paths related to i386 windows are also removed. - - - - - cef8f47a by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: minor adjustments for stdcall removal This commit include minor adjustments of documentation related to stdcall removal. - - - - - 54332437 by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: mention i386 Windows removal in 9.12 changelog This commit mentions removal of i386 Windows support and stdcall related change in the 9.12 changelog. - - - - - 2aaea8a1 by Cheng Shao at 2024-06-04T05:06:40-04:00 hadrian: improve user settings documentation This patch adds minor improvements to hadrian user settings documentation: - Add missing `ghc.cpp.opts` case - Remove non-existent `cxx` case - Clarify `cc.c.opts` also works for C++, while `cc.deps.opts` doesn't - Add example of passing configure argument to autoconf packages - - - - - 71010381 by Alex Mason at 2024-06-04T12:09:07-04:00 Add AArch64 CLZ, CTZ, RBIT primop implementations. Adds support for emitting the clz and rbit instructions, which are used by GHC.Prim.clz*#, GHC.Prim.ctz*# and GHC.Prim.bitReverse*#. - - - - - 44e2abfb by Cheng Shao at 2024-06-04T12:09:43-04:00 hadrian: add +text_simdutf flavour transformer to allow building text with simdutf This patch adds a +text_simdutf flavour transformer to hadrian to allow downstream packagers and users that build from source to opt-in simdutf support for text, in order to benefit from SIMD speedup at run-time. It's still disabled by default for the time being. - - - - - 077cb2e1 by Cheng Shao at 2024-06-04T12:09:43-04:00 ci: enable +text_simdutf flavour transformer for wasm jobs This commit enables +text_simdutf flavour transformer for wasm jobs, so text is now built with simdutf support for wasm. - - - - - b23746ad by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Use TemplateHaskellQuotes in instance Lift ByteArray Resolves #24852 - - - - - 3fd25743 by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Mark addrToByteArray as NOINLINE This function should never be inlined in order to keep code size small. - - - - - 98ad1ea5 by Cheng Shao at 2024-06-04T22:51:26-04:00 compiler: remove unused CompilerInfo/LinkerInfo types This patch removes CompilerInfo/LinkerInfo types from the compiler since they aren't actually used anywhere. - - - - - 11795244 by Cheng Shao at 2024-06-05T06:33:17-04:00 rts: remove unused PowerPC/IA64 native adjustor code This commit removes unused PowerPC/IA64 native adjustor code which is never actually enabled by autoconf/hadrian. Fixes #24920. - - - - - 5132754b by Sylvain Henry at 2024-06-05T06:33:57-04:00 RTS: fix warnings with doing*Profiling (#24918) - - - - - accc8c33 by Cheng Shao at 2024-06-05T11:35:36-04:00 hadrian: don't depend on inplace/mingw when --enable-distro-toolchain on Windows - - - - - 6ffbd678 by Cheng Shao at 2024-06-05T11:35:37-04:00 autoconf: normalize paths of some build-time dependencies on Windows This commit applies path normalization via cygpath -m to some build-time dependencies on Windows. Without this logic, the /clang64/bin prefixed msys2-style paths cause the build to fail with --enable-distro-toolchain. - - - - - 075dc6d4 by Cheng Shao at 2024-06-05T11:36:12-04:00 hadrian: remove OSDarwin mention from speedHack This commit removes mentioning of OSDarwin from speedHack, since speedHack is purely for i386 and we no longer support i386 darwin (#24921). - - - - - 83235c4c by Cheng Shao at 2024-06-05T11:36:12-04:00 compiler: remove 32-bit darwin logic This commit removes all 32-bit darwin logic from the compiler, given we no longer support 32-bit apple systems (#24921). Also contains a bit more cleanup of obsolete i386 windows logic. - - - - - 1eb99bc3 by Cheng Shao at 2024-06-05T11:36:12-04:00 rts: remove 32-bit darwin/ios logic This commit removes 32-bit darwin/ios related logic from the rts, given we no longer support them (#24921). - - - - - 24f65892 by Cheng Shao at 2024-06-05T11:36:12-04:00 llvm-targets: remove 32-bit darwin/ios targets This commit removes 32-bit darwin/ios targets from llvm-targets given we no longer support them (#24921). - - - - - ccdbd689 by Cheng Shao at 2024-06-05T11:36:12-04:00 testsuite: remove 32-bit darwin logic This commit removes 32-bit darwin logic from the testsuite given it's no longer supported (#24921). Also contains more cleanup of obsolete i386 windows logic. - - - - - 11d661c4 by Cheng Shao at 2024-06-05T11:36:13-04:00 docs: mention 32-bit darwin/ios removal in 9.12 changelog This commit mentions removal of 32-bit darwin/ios support (#24921) in the 9.12 changelog. - - - - - 7c173310 by Georgi Lyubenov at 2024-06-05T15:17:22-04:00 Add firstA and secondA to Data.Bitraversable Please see https://github.com/haskell/core-libraries-committee/issues/172 for related discussion - - - - - 3b6f9fd1 by Ben Gamari at 2024-06-05T15:17:59-04:00 base: Fix name of changelog Fixes #24899. Also place it under `extra-doc-files` to better reflect its nature and avoid triggering unnecessary recompilation if it changes. - - - - - 1f4d2ef7 by Sebastian Graf at 2024-06-05T15:18:34-04:00 Announce Or-patterns in the release notes for GHC 9.12 (#22596) Leftover from !9229. - - - - - 8650338d by Jan Hrček at 2024-06-06T10:39:24-04:00 Improve haddocks of Language.Haskell.Syntax.Pat.Pat - - - - - 2eee65e1 by Cheng Shao at 2024-06-06T10:40:00-04:00 testsuite: bump T7653 timeout for wasm - - - - - 990fed60 by Sylvain Henry at 2024-06-07T14:45:23-04:00 StgToCmm: refactor opTranslate and friends - Change arguments order to avoid `\args -> ...` lambdas - Fix documentation - Rename StgToCmm options ("big" doesn't mean anything) - - - - - 1afad514 by Sylvain Henry at 2024-06-07T14:45:23-04:00 NCG x86: remove dead code (#5444) Since 6755d833af8c21bbad6585144b10e20ac4a0a1ab this code is dead. - - - - - 595c0894 by Cheng Shao at 2024-06-07T14:45:58-04:00 testsuite: skip objc-hi/objcxx-hi when cross compiling objc-hi/objcxx-hi should be skipped when cross compiling. The existing opsys('darwin') predicate only asserts the host system is darwin but tells us nothing about the target, hence the oversight. - - - - - edfe6140 by qqwy at 2024-06-08T11:23:54-04:00 Replace '?callStack' implicit param with HasCallStack in GHC.Internal.Exception.throw - - - - - 35a64220 by Cheng Shao at 2024-06-08T11:24:30-04:00 rts: cleanup inlining logic This patch removes pre-C11 legacy code paths related to INLINE_HEADER/STATIC_INLINE/EXTERN_INLINE macros, ensure EXTERN_INLINE is treated as static inline in most cases (fixes #24945), and also corrects the comments accordingly. - - - - - 9ea90ed2 by Andrew Lelechenko at 2024-06-08T11:25:06-04:00 CODEOWNERS: add @core-libraries to track base interface changes A low-tech tactical solution for #24919 - - - - - 580fef7b by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update CHANGELOG to reflect current version - - - - - 391ecff5 by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update prologue.txt to reflect package description - - - - - 3dca3b7d by Ben Gamari at 2024-06-09T01:27:57-04:00 compiler: Clarify comment regarding need for MOVABS The comment wasn't clear in stating that it was only applicable to immediate source and memory target operands. - - - - - 6bd850e8 by doyougnu at 2024-06-09T21:02:14-04:00 JS: establish single source of truth for symbols In pursuit of: #22736. This MR moves ad-hoc symbols used throughout the js backend into a single symbols file. Why? First, this cleans up the code by removing ad-hoc strings created on the fly and therefore makes the code more maintainable. Second, it makes it much easier to eventually type these identifiers. - - - - - f3017dd3 by Cheng Shao at 2024-06-09T21:02:49-04:00 rts: replace ad-hoc MYTASK_USE_TLV with proper CC_SUPPORTS_TLS This patch replaces the ad-hoc `MYTASK_USE_TLV` with the `CC_SUPPORTS_TLS` macro. If TLS support is detected by autoconf, then we should use that for managing `myTask` in the threaded RTS. - - - - - e17d7e8c by Ben Gamari at 2024-06-11T05:25:21-04:00 users-guide: Fix stylistic issues in 9.12 release notes - - - - - 8a8a982a by Hugo Peters at 2024-06-11T05:25:57-04:00 fix typo in the simplifier debug output: baling -> bailing - - - - - 16475bb8 by Hécate Moonlight at 2024-06-12T03:07:55-04:00 haddock: Correct the Makefile to take into account Darwin systems - - - - - a2f60da5 by Hécate Kleidukos at 2024-06-12T03:08:35-04:00 haddock: Remove obsolete links to github.com/haskell/haddock in the docs - - - - - de4395cd by qqwy at 2024-06-12T03:09:12-04:00 Add `__GLASGOW_HASKELL_ASSERTS_IGNORED__` as CPP macro name if `-fasserts-ignored is set. This allows users to create their own Control.Exception.assert-like functionality that does something other than raising an `AssertFailed` exception. Fixes #24967 - - - - - 0e9c4dee by Ryan Hendrickson at 2024-06-12T03:09:53-04:00 compiler: add hint to TcRnBadlyStaged message - - - - - 2747cd34 by Simon Peyton Jones at 2024-06-12T12:51:37-04:00 Fix a QuickLook bug This MR fixes the bug exposed by #24676. The problem was that quickLookArg was trying to avoid calling tcInstFun unnecessarily; but it was in fact necessary. But that in turn forced me into a significant refactoring, putting more fields into EValArgQL. Highlights: see Note [Quick Look overview] in GHC.Tc.Gen.App * Instantiation variables are now distinguishable from ordinary unification variables, by level number = QLInstVar. This is treated like "level infinity". See Note [The QLInstVar TcLevel] in GHC.Tc.Utils.TcType. * In `tcApp`, we don't track the instantiation variables in a set Delta any more; instead, we just tell them apart by their level number. * EValArgQL now much more clearly captures the "half-done" state of typechecking an argument, ready for later resumption. See Note [Quick Look at value arguments] in GHC.Tc.Gen.App * Elminated a bogus (never used) fast-path in GHC.Tc.Utils.Instantiate.instCallConstraints See Note [Possible fast path for equality constraints] Many other small refactorings. - - - - - 1b1523b1 by George Thomas at 2024-06-12T12:52:18-04:00 Fix non-compiling extensible record `HasField` example - - - - - 97b141a3 by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Fix hyperlinker source urls (#24907) This fixes a bug introduced by f56838c36235febb224107fa62334ebfe9941aba Links to external modules in the hyperlinker are uniformly generated using splicing the template given to us instead of attempting to construct the url in an ad-hoc manner. - - - - - 954f864c by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Add name anchor to external source urls from documentation page URLs for external source links from documentation pages were missing a splice location for the name. Fixes #24912 - - - - - b0b64177 by Simon Peyton Jones at 2024-06-12T12:53:31-04:00 Prioritise nominal equalities The main payload of this patch is * Prioritise nominal equalities in the constraint solver. This ameliorates the incompleteness of solving for representational constraints over newtypes: see #24887. See (EX2) in Note [Decomposing newtype equalities] in GHC.Tc.Solver.Equality In doing this patch I tripped over some other things that I refactored: * Move `isCoVarType` from `GHC.Core.Type` to `GHC.Core.Predicate` where it seems more at home. * Clarify the "rewrite role" of a constraint. I was very puzzled about what the role of, say `(Eq a)` might be, but see the new Note [The rewrite-role of a constraint]. In doing so I made predTypeEqRel crash when given a non-equality. Usually it expects an equality; but it was being mis-used for the above rewrite-role stuff. - - - - - cb7c1b83 by Liam Goodacre at 2024-06-12T12:54:09-04:00 compiler: missing-deriving-strategies suggested fix Extends the missing-deriving-strategies warning with a suggested fix that includes which deriving strategies were assumed. For info about the warning, see comments for `TcRnNoDerivStratSpecified`, `TcRnNoDerivingClauseStrategySpecified`, & `TcRnNoStandaloneDerivingStrategySpecified`. For info about the suggested fix, see `SuggestExplicitDerivingClauseStrategies` & `SuggestExplicitStandalanoDerivingStrategy`. docs: Rewords missing-deriving-strategies to mention the suggested fix. Resolves #24955 - - - - - 4e36d3a3 by Jan Hrček at 2024-06-12T12:54:48-04:00 Further haddocks improvements in Language.Haskell.Syntax.Pat.Pat - - - - - 558353f4 by Cheng Shao at 2024-06-12T12:55:24-04:00 rts: use page sized mblocks on wasm This patch changes mblock size to page size on wasm. It allows us to simplify our wasi-libc fork, makes it much easier to test third party libc allocators like emmalloc/mimalloc, as well as experimenting with threaded RTS in wasm. - - - - - b3cc5366 by Matthew Pickering at 2024-06-12T23:06:57-04:00 compiler: Make ghc-experimental not wired in If you need to wire in definitions, then place them in ghc-internal and reexport them from ghc-experimental. Ticket #24903 - - - - - 700eeab9 by Hécate Kleidukos at 2024-06-12T23:07:37-04:00 base: Use a more appropriate unicode arrow for the ByteArray diagram This commit rectifies the usage of a unicode arrow in favour of one that doesn't provoke mis-alignment. - - - - - cca7de25 by Matthew Pickering at 2024-06-12T23:08:14-04:00 ghcup-metadata: Fix debian version ranges This was caught by `ghcup-ci` failing and attempting to install a deb12 bindist on deb11. ``` configure: WARNING: m4/prep_target_file.m4: Expecting YES/NO but got in ArSupportsDashL_STAGE0. Defaulting to False. bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by bin/ghc-toolchain-bin) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) ``` Fixes #24974 - - - - - 7b23ce8b by Pierre Le Marre at 2024-06-13T15:35:04-04:00 ucd2haskell: remove Streamly dependency + misc - Remove dead code. - Remove `streamly` dependency. - Process files with `bytestring`. - Replace Unicode files parsers with the corresponding ones from the package `unicode-data-parser`. - Simplify cabal file and rename module - Regenerate `ghc-internal` Unicode files with new header - - - - - 4570319f by Jacco Krijnen at 2024-06-13T15:35:41-04:00 Document how to run haddocks tests (#24976) Also remove ghc 9.7 requirement - - - - - fb629e24 by amesgen at 2024-06-14T00:28:20-04:00 compiler: refactor lower_CmmExpr_Ptr - - - - - def46c8c by amesgen at 2024-06-14T00:28:20-04:00 compiler: handle CmmRegOff in lower_CmmExpr_Ptr - - - - - ce76bf78 by Simon Peyton Jones at 2024-06-14T00:28:56-04:00 Small documentation update in Quick Look - - - - - 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 7658b654 by Sven Tennie at 2024-09-06T09:19:02+02:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 19 changed files: - + .git-blame-ignore-revs - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - + .gitlab/merge_request_templates/Haddock.md - .gitlab/rel_eng/default.nix - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/README.mkd - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - + .gitlab/rel_eng/recompress-all - .gitlab/rel_eng/upload.sh - .gitmodules - CODEOWNERS The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4111bde9f2aec45b976ba9e5d6bb77f25ca28387...7658b654218f4496d47d62551de0c54e8819438c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4111bde9f2aec45b976ba9e5d6bb77f25ca28387...7658b654218f4496d47d62551de0c54e8819438c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 09:48:25 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 06 Sep 2024 05:48:25 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] 14 commits: Silence x-partial in Haddock.Backends.Xhtml Message-ID: <66dacfe99f5e3_256ec76d9110727af@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - e3e9ef8b by Sven Tennie at 2024-09-06T11:48:02+02:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - b8136473 by Sven Tennie at 2024-09-06T11:48:02+02:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - f47889e2 by Sven Tennie at 2024-09-06T11:48:02+02:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 0020bf0e by Sven Tennie at 2024-09-06T11:48:02+02:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - 2380602c by Sven Tennie at 2024-09-06T11:48:02+02:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - bb9bbe6a by Sven Tennie at 2024-09-06T11:48:02+02:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 257434b9 by Sven Tennie at 2024-09-06T11:48:03+02:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 147a40c4 by Sven Tennie at 2024-09-06T11:48:03+02:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - 40722eeb by Sven Tennie at 2024-09-06T11:48:03+02:00 RISCV64: Set codeowners of the NCG - - - - - 3149b403 by Sven Tennie at 2024-09-06T11:48:03+02:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. - - - - - 11 changed files: - .gitlab/ci.sh - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5856aa6ec7f5673f227558d4378aecba1b692dd0...3149b4035c6b6422f28ee03a660e3b31cf85581f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5856aa6ec7f5673f227558d4378aecba1b692dd0...3149b4035c6b6422f28ee03a660e3b31cf85581f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 11:41:40 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Fri, 06 Sep 2024 07:41:40 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] 2 commits: Link bytecode from interface-stored core bindings in oneshot mode Message-ID: <66daea745d87b_29545431c1e45136d@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: fbe2f90e by Cheng Shao at 2024-09-06T12:49:41+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). Metric Decrease: MultiLayerModules T13701 - - - - - 14017dde by Torsten Schmits at 2024-09-06T13:41:19+02:00 debug - - - - - 20 changed files: - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/A.hs - + testsuite/tests/bytecode/T25090/B.hs - + testsuite/tests/bytecode/T25090/C.hs - + testsuite/tests/bytecode/T25090/C.hs-boot - + testsuite/tests/bytecode/T25090/D.hs - + testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/T25090-debug.stderr - + testsuite/tests/bytecode/T25090/T25090.stdout - + testsuite/tests/bytecode/T25090/all.T Changes: ===================================== compiler/GHC/CoreToIface.hs ===================================== @@ -602,7 +602,7 @@ toIfaceTopBind b = in (top_bndr, rhs') -- The sharing behaviour is currently disabled due to #22807, and relies on - -- finished #220056 to be re-enabled. + -- finished #20056 to be re-enabled. disabledDueTo22807 = True already_has_unfolding b = not disabledDueTo22807 @@ -774,8 +774,8 @@ outside of the hs-boot loop. Note [Interface File with Core: Sharing RHSs] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -IMPORTANT: This optimisation is currently disabled due to #22027, it can be - re-enabled once #220056 is implemented. +IMPORTANT: This optimisation is currently disabled due to #22807, it can be + re-enabled once #22056 is implemented. In order to avoid duplicating definitions for bindings which already have unfoldings we do some minor headstands to avoid serialising the RHS of a definition if it has ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -260,7 +260,6 @@ outputForeignStubs Maybe FilePath) -- C file created outputForeignStubs logger tmpfs dflags unit_state mod location stubs = do - let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location stub_c <- newTempName logger tmpfs (tmpDir dflags) TFL_CurrentModule "c" case stubs of @@ -276,8 +275,6 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs stub_h_output_d = pprCode h_code stub_h_output_w = showSDoc dflags stub_h_output_d - createDirectoryIfMissing True (takeDirectory stub_h) - putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export header file" FormatC @@ -299,9 +296,20 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs | platformMisc_libFFI $ platformMisc dflags = "#include \"rts/ghc_ffi.h\"\n" | otherwise = "" - stub_h_file_exists - <- outputForeignStubs_help stub_h stub_h_output_w - ("#include \n" ++ cplusplus_hdr) cplusplus_ftr + stub_h_file_exists <- + if null stub_h_output_w + then pure False + else do + -- The header path is computed from the module source path, which + -- does not exist when loading interface core bindings for Template + -- Haskell. + -- The header is only generated for foreign exports. + -- Since those aren't supported for TH with bytecode, we can skip + -- this here for now. + let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location + createDirectoryIfMissing True (takeDirectory stub_h) + outputForeignStubs_help stub_h stub_h_output_w + ("#include \n" ++ cplusplus_hdr) cplusplus_ftr putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export stubs" FormatC stub_c_output_d ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,6 +49,7 @@ module GHC.Driver.Main , HscBackendAction (..), HscRecompStatus (..) , initModDetails , initWholeCoreBindings + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -105,6 +106,7 @@ module GHC.Driver.Main , showModuleIndex , hscAddSptEntries , writeInterfaceOnlyMode + , loadByteCode ) where import GHC.Prelude @@ -274,7 +276,8 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.List ( nub, isPrefixOf, partition ) +import Data.Functor ((<&>)) +import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad import Data.IORef @@ -970,19 +973,23 @@ loadByteCode iface mod_sum = do (mi_foreign iface) return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) _ -> return $ outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- +add_iface_to_hpt :: ModIface -> ModDetails -> HscEnv -> HscEnv +add_iface_to_hpt iface details = + hscUpdateHPT $ \ hpt -> + addToHpt hpt (moduleName (mi_module iface)) + (HomeModInfo iface details emptyHomeModInfoLinkable) -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] initModDetails :: HscEnv -> ModIface -> IO ModDetails initModDetails hsc_env iface = fixIO $ \details' -> do - let act hpt = addToHpt hpt (moduleName $ mi_module iface) - (HomeModInfo iface details' emptyHomeModInfoLinkable) - let !hsc_env' = hscUpdateHPT act hsc_env + let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful @@ -1010,8 +1017,52 @@ compile_for_interpreter hsc_env use = adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Return an 'IO' that hydrates Core bindings and compiles them to bytecode if +-- the interface contains any, using the supplied type env for typechecking. +-- +-- Unlike 'initWholeCoreBindings', this does not use lazy IO. +-- Instead, the 'IO' is only evaluated (in @get_link_deps@) when it is clear +-- that it will be used immediately (because we're linking TH with +-- @-fprefer-byte-code@ in oneshot mode), and the result is cached in +-- 'LoaderState'. +-- +-- 'initWholeCoreBindings' needs the laziness because it is used to populate +-- 'HomeModInfo', which is done preemptively, in anticipation of downstream +-- modules using the bytecode for TH in make mode, which might never happen. +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + compile <$> iface_core_bindings iface location + where + compile decls = do + (bcos, fos) <- compileWholeCoreBindings hsc_env type_env decls + linkable $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time (mi_module iface) parts + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects. +-- them with a lazy IO thunk that compiles them to bytecode and foreign objects, +-- using the supplied environment for type checking. -- -- The laziness is necessary because this value is stored purely in a -- 'HomeModLinkable' in the home package table, rather than some dedicated @@ -1025,29 +1076,71 @@ compile_for_interpreter hsc_env use = -- -- This is sound because generateByteCode just depends on things already loaded -- in the interface file. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env mod_iface details (Linkable utc_time this_mod uls) = +initWholeCoreBindings :: + HscEnv -> + ModIface -> + ModDetails -> + Linkable -> + IO Linkable +initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = Linkable utc_time this_mod <$> mapM go uls where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef (md_types details) - let act hpt = addToHpt hpt (moduleName $ mi_module mod_iface) - (HomeModInfo mod_iface details emptyHomeModInfoLinkable) - kv = knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - hsc_env' = hscUpdateHPT act hsc_env { hsc_type_env_vars = kv } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") hsc_env' $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons (md_types details)) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + go = \case + CoreBindings wcb -> do + ~(bco, fos) <- unsafeInterleaveIO $ + compileWholeCoreBindings hsc_env' type_env wcb + pure (LazyBCOs bco fos) + l -> pure l + + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written or were unavailable due to boot import +-- cycles, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +compileWholeCoreBindings :: + HscEnv -> + TypeEnv -> + WholeCoreBindings -> + IO (CompiledByteCode, [FilePath]) +compileWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + gen_bytecode core_binds stubs foreign_files + where + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + generateByteCode hsc_env cgi_guts wcb_mod_location + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -0,0 +1,15 @@ +module GHC.Driver.Main where + +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) + +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1310,8 +1310,10 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- MP: This is a bit janky, because before you add the entries you have to extend the HPT with the module -- you just compiled. Another option, would be delay adding anything until after upsweep has finished, but I -- am unsure if this is sound (wrt running TH splices for example). - -- This function only does anything if the linkable produced is a BCO, which only happens with the - -- bytecode backend, no need to guard against the backend type additionally. + -- This function only does anything if the linkable produced is a BCO, which + -- used to only happen with the bytecode backend, but with + -- @-fprefer-byte-code@, @HomeModInfo@ has bytecode even when generating + -- object code, see #25230. addSptEntries (hscUpdateHPT (\hpt -> addToHpt hpt (ms_mod_name summary) hmi) hsc_env) (homeModInfoByteCode hmi) @@ -2990,7 +2992,7 @@ which can be checked easily using ghc-debug. a reference to the entire HscEnv, if we are not careful the HscEnv will contain the HomePackageTable at the time the interface was loaded and it will never be released. - Where? dontLeakTheHPT in GHC.Iface.Load + Where? dontLeakTheHUG in GHC.Iface.Load 2. No KnotVars are live at the end of upsweep (#20491) Why? KnotVars contains an old stale reference to the TypeEnv for modules ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,12 +506,14 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) ; new_eps_rules <- tcIfaceRules ignore_prags (mi_rules iface) ; new_eps_anns <- tcIfaceAnnotations (mi_anns iface) ; new_eps_complete_matches <- tcIfaceCompleteMatches (mi_complete_matches iface) + ; purged_hsc_env <- getTopEnv ; let final_iface = iface & set_mi_decls (panic "No mi_decls in PIT") @@ -518,13 +521,27 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + | prefer_bytecode + , Just action <- loadIfaceByteCode purged_hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have 'extra_decls' + -- so 'get_link_deps' knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,6 +553,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -569,7 +587,7 @@ loadInterface doc_str mod from {- Note [Loading your own hi-boot file] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Generally speaking, when compiling module M, we should not -load M.hi boot into the EPS. After all, we are very shortly +load M.hi-boot into the EPS. After all, we are very shortly going to have full information about M. Moreover, see Note [Do not update EPS with your own hi-boot] in GHC.Iface.Recomp. @@ -698,7 +716,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -845,7 +863,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -875,7 +893,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -900,7 +918,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -905,11 +905,11 @@ tcTopIfaceBindings :: IORef TypeEnv -> [IfaceBindingX IfaceMaybeRhs IfaceTopBndr -> IfL [CoreBind] tcTopIfaceBindings ty_var ver_decls = do - int <- mapM tcTopBinders ver_decls + int <- mapM tcTopBinders ver_decls let all_ids :: [Id] = concatMap toList int liftIO $ modifyIORef ty_var (flip extendTypeEnvList (map AnId all_ids)) - extendIfaceIdEnv all_ids $ mapM (tc_iface_bindings) int + extendIfaceIdEnv all_ids $ mapM tc_iface_bindings int tcTopBinders :: IfaceBindingX a IfaceTopBndrInfo -> IfL (IfaceBindingX a Id) tcTopBinders = traverse mk_top_id ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -60,16 +60,16 @@ import System.Directory data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -275,21 +275,21 @@ get_link_deps opts pls maybe_normal_osuf span mods = do case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod = do { - -- ...and then find the linkable for it - mb_lnk <- findObjectLinkableMaybe mod loc ; - case mb_lnk of { - Nothing -> no_obj mod ; - Just lnk -> adjust_linkable lnk - }} + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,18 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -68,6 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -139,6 +142,12 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), + eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules eps_fam_inst_env :: !PackageFamInstEnv,-- ^ The total 'FamInstEnv' accumulated ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -28,12 +28,12 @@ import System.FilePath (takeExtension) {- Note [Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A interface file can optionally contain the definitions of all core bindings, this is enabled by the flag `-fwrite-if-simplified-core`. This provides everything needed in addition to the normal ModIface and ModDetails -to restart compilation after typechecking to generate bytecode. The `fi_bindings` field +to restart compilation after typechecking to generate bytecode. The `wcb_bindings` field is stored in the normal interface file and the other fields populated whilst loading the interface file. @@ -62,8 +62,50 @@ after whatever simplification the user requested has been performed. So the simp of the interface file agree with the optimisation level as reported by the interface file. +The lifecycle differs beyond laziness depending on the provenance of a module. +In all cases, the main consumer for interface bytecode is 'get_link_deps', which +traverses a splice's or GHCi expression's dependencies and collects the needed +build artifacts, which can be objects or bytecode, depending on the build +settings. + +1. In make mode, all eligible modules are part of the dependency graph. + Their interfaces are loaded unconditionally and in dependency order by the + compilation manager, and each module's bytecode is prepared before its + dependents are compiled, in one of two ways: + + - If the interface file for a module is missing or out of sync with its + source, it is recompiled and bytecode is generated directly and + immediately, not involving 'WholeCoreBindings' (in 'runHscBackendPhase'). + + - If the interface file is up to date, no compilation is performed, and a + lazy thunk generating bytecode from interface Core bindings is created in + 'compileOne'', which will only be compiled if a downstream module contains + a splice that depends on it, as described above. + + In both cases, the bytecode 'Linkable' is stored in a 'HomeModLinkable' in + the Home Unit Graph, lazy or not. + +2. In oneshot mode, which compiles individual modules without a shared home unit + graph, a previously compiled module is not reprocessed as described for make + mode above. + When 'get_link_deps' encounters a dependency on a local module, it requests + its bytecode from the External Package State, who loads the interface + on-demand. + + Since the EPS stores interfaces for all package dependencies in addition to + local modules in oneshot mode, it has a substantial memory footprint. + We try to curtail that by extracting important data into specialized fields + in the EPS, and retaining only a few fields of 'ModIface' by overwriting the + others with bottom values. + + In order to avoid keeping around all of the interface's components needed for + compiling bytecode, we instead store an IO action in 'eps_iface_bytecode'. + When 'get_link_deps' evaluates this action, the result is not retained in the + EPS, but stored in 'LoaderState', where it may eventually get evicted to free + up the memory. + Note [Size of Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ How much overhead does `-fwrite-if-simplified-core` add to a typical interface file? As an experiment I compiled the `Cabal` library and `ghc` library (Aug 22) with ===================================== testsuite/tests/bytecode/T25090/A.hs ===================================== @@ -0,0 +1,7 @@ +{-# language TemplateHaskell #-} +module Main where + +import D + +main :: IO () +main = putStrLn (show ($splc :: Int)) ===================================== testsuite/tests/bytecode/T25090/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} C (C) + +data B = B C ===================================== testsuite/tests/bytecode/T25090/C.hs ===================================== @@ -0,0 +1,8 @@ +module C where + +import B + +data C = C Int + +b :: B +b = B (C 2024) ===================================== testsuite/tests/bytecode/T25090/C.hs-boot ===================================== @@ -0,0 +1,3 @@ +module C where + +data C ===================================== testsuite/tests/bytecode/T25090/D.hs ===================================== @@ -0,0 +1,12 @@ +module D where + +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +import B +import C + +splc :: ExpQ +splc = + lift @_ @Int num + where + B (C num) = b ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -0,0 +1,32 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +T25090a: + $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot + $(TEST_HC) -c -fbyte-code-and-object-code B.hs + $(TEST_HC) -c -fbyte-code-and-object-code C.hs + $(TEST_HC) -c -fbyte-code-and-object-code D.hs + $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code D.o C.o B.o A.o -o exe + ./exe + +# Verify that the object files aren't linked by clobbering them. +# Windows and Darwin still need those symbols to be defined during linking, so +# we only run this test on Linux. +T25090b: + $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot + $(TEST_HC) -c -fbyte-code-and-object-code B.hs + $(TEST_HC) -c -fbyte-code-and-object-code C.hs + $(TEST_HC) -c -fbyte-code-and-object-code D.hs + echo 'corrupt' > B.o + echo 'corrupt' > C.o + echo 'corrupt' > C.o-boot + echo 'corrupt' > D.o + $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A.o -o exe + ./exe + +T25090c: + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 + ./exe ===================================== testsuite/tests/bytecode/T25090/T25090-debug.stderr ===================================== @@ -0,0 +1,6 @@ +WARNING: + loadInterface + C + Call stack: + CallStack (from HasCallStack): + warnPprTrace, called at compiler/GHC/Iface/Load.hs:: in :GHC.Iface.Load ===================================== testsuite/tests/bytecode/T25090/T25090.stdout ===================================== @@ -0,0 +1 @@ +2024 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -0,0 +1,21 @@ +# This test compiles the boot file separately from its source file, which causes +# a debug assertion warning. +# Since this appears to be intentional according to the Note [Loading your own hi-boot file], +# the warning is added to the expected stderr for debugged builds. +def test_T25090(name): + assert_warn_spec = {'stderr': 'T25090-debug.stderr'} + extra_specs = assert_warn_spec if name == 'T25090a' and compiler_debugged() else {} + # extra_setup = [skip] if name == 'T25090b' and not opsys('linux') else [] + extra_setup = [] + return test(name, + [extra_files(['A.hs', 'B.hs', 'C.hs-boot', 'C.hs', 'D.hs']), + req_th, + js_skip, + use_specs(dict(stdout = 'T25090.stdout', **extra_specs)), + ] + extra_setup, + makefile_test, + []) + +test_T25090('T25090a') +test_T25090('T25090b') +test_T25090('T25090c') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a14b588f20e2bd5821e9af671ea72790c27054d0...14017dde8291f02901d8b6bd506bff02ad1a6049 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a14b588f20e2bd5821e9af671ea72790c27054d0...14017dde8291f02901d8b6bd506bff02ad1a6049 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 11:50:21 2024 From: gitlab at gitlab.haskell.org (Sjoerd Visscher (@trac-sjoerd_visscher)) Date: Fri, 06 Sep 2024 07:50:21 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/finder-boot] 38 commits: compiler: Fix pretty printing of ticked prefix constructors (#24237) Message-ID: <66daec7df64d_2954544bbc20538c4@gitlab.mail> Sjoerd Visscher pushed to branch wip/torsten.schmits/finder-boot at Glasgow Haskell Compiler / GHC Commits: 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5545f976 by Torsten Schmits at 2024-09-06T13:49:59+02:00 add `IsBootInterface` to finder cache keys - - - - - 30 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/HsToCore/Usage.hs - compiler/GHC/Iface/Make.hs - compiler/GHC/Iface/Recomp.hs - compiler/GHC/Iface/Type.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Literal.hs - compiler/GHC/StgToJS/Prim.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e4254f658dc7f2359917edc2fff35fda5a4f08d0...5545f976a0934e1b0ab7543b7879aaf9ba5c9fc3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e4254f658dc7f2359917edc2fff35fda5a4f08d0...5545f976a0934e1b0ab7543b7879aaf9ba5c9fc3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 13:16:35 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Fri, 06 Sep 2024 09:16:35 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 12 commits: The X86 SIMD patch. Message-ID: <66db00b32cf08_1357ced3554376eb@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 0dc192b1 by sheaf at 2024-09-06T15:15:56+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 180e1b0c by sheaf at 2024-09-06T15:16:00+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - df470a18 by sheaf at 2024-09-06T15:16:00+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2e18f32b by sheaf at 2024-09-06T15:16:00+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 9dbd6457 by sheaf at 2024-09-06T15:16:16+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - ea57f352 by sheaf at 2024-09-06T15:16:19+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - a8aa3717 by sheaf at 2024-09-06T15:16:19+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - d27ce84d by sheaf at 2024-09-06T15:16:19+02:00 Add test for C calls & SIMD vectors - - - - - 626cd452 by sheaf at 2024-09-06T15:16:20+02:00 Add test for #25169 - - - - - 7ea046d6 by sheaf at 2024-09-06T15:16:20+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 2591bdff by sheaf at 2024-09-06T15:16:20+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 2770c67d by sheaf at 2024-09-06T15:16:20+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c37faadfd5b1e3f484a7b3b83c0e0632ff662b98...2770c67db97a99934342a95fda73b6c924679984 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c37faadfd5b1e3f484a7b3b83c0e0632ff662b98...2770c67db97a99934342a95fda73b6c924679984 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 13:17:27 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Fri, 06 Sep 2024 09:17:27 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 12 commits: The X86 SIMD patch. Message-ID: <66db00e7de32b_1357cec88c0387b3@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 0dcf6949 by sheaf at 2024-09-06T15:17:16+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - a171f51f by sheaf at 2024-09-06T15:17:17+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 86f25a28 by sheaf at 2024-09-06T15:17:17+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - e2b11cbe by sheaf at 2024-09-06T15:17:17+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - ada10bc9 by sheaf at 2024-09-06T15:17:17+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - ecc38a83 by sheaf at 2024-09-06T15:17:17+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - ad9852b0 by sheaf at 2024-09-06T15:17:17+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - e489ca8a by sheaf at 2024-09-06T15:17:17+02:00 Add test for C calls & SIMD vectors - - - - - fac07580 by sheaf at 2024-09-06T15:17:17+02:00 Add test for #25169 - - - - - d9dcc041 by sheaf at 2024-09-06T15:17:18+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 47f4244b by sheaf at 2024-09-06T15:17:18+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 086ca50e by sheaf at 2024-09-06T15:17:18+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2770c67db97a99934342a95fda73b6c924679984...086ca50e7d242fcc8102df437637496b1dfb4136 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2770c67db97a99934342a95fda73b6c924679984...086ca50e7d242fcc8102df437637496b1dfb4136 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 13:54:30 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Fri, 06 Sep 2024 09:54:30 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] debug Message-ID: <66db0996a35da_1357ce4db804424a9@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 2088616d by Torsten Schmits at 2024-09-06T15:33:22+02:00 debug - - - - - 2 changed files: - testsuite/tests/bytecode/T25090/Makefile - testsuite/tests/bytecode/T25090/all.T Changes: ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -2,13 +2,18 @@ TOP=../../.. include $(TOP)/mk/boilerplate.mk include $(TOP)/mk/test.mk +# Verify that the object files aren't linked by clobbering them. T25090a: $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot $(TEST_HC) -c -fbyte-code-and-object-code B.hs $(TEST_HC) -c -fbyte-code-and-object-code C.hs + echo 'corrupt' > B.o + echo 'corrupt' > C.o + echo 'corrupt' > C.o-boot $(TEST_HC) -c -fbyte-code-and-object-code D.hs + echo 'corrupt' > D.o $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs - $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code D.o C.o B.o A.o -o exe + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A.o -o exe ./exe T25090b: ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -15,5 +15,4 @@ def test_T25090(name): []) test_T25090('T25090a') - test_T25090('T25090b') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2088616d47d89831851c54d36917137ba949fcf9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2088616d47d89831851c54d36917137ba949fcf9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 16:10:39 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 06 Sep 2024 12:10:39 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 6 commits: Silence x-partial in Haddock.Backends.Xhtml Message-ID: <66db297f108d7_1357ced8d6a0667b2@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 5bba677e by Hécate Kleidukos at 2024-09-06T12:10:04-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 24 changed files: - .gitlab/ci.sh - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs - compiler/GHC/Stg/Unarise.hs - compiler/GHC/Stg/Utils.hs - compiler/GHC/StgToCmm/DataCon.hs - compiler/GHC/Types/CostCentre.hs - compiler/ghc.cabal.in - hadrian/ghci-multi-cabal.in - hadrian/src/Rules/ToolArgs.hs - testsuite/tests/codeGen/should_compile/Makefile - + testsuite/tests/codeGen/should_compile/T25166.hs - + testsuite/tests/codeGen/should_compile/T25166.stdout - testsuite/tests/codeGen/should_compile/all.T - + utils/haddock/doc/requirements.txt - utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Layout.hs - utils/haddock/haddock-api/src/Haddock/Utils/Json.hs Changes: ===================================== .gitlab/ci.sh ===================================== @@ -753,7 +753,7 @@ function check_interfaces(){ } function abi_test() { - for i in {1..20}; do info "iteration $i"; run_abi_test; done + for i in {1..10}; do info "iteration $i"; run_abi_test; done } function run_abi_test() { @@ -761,8 +761,8 @@ function run_abi_test() { fail "HC not set" fi mkdir -p out - OUT="$PWD/out/run1" DIR=$(mktemp -d XXXX-looooooooong) cabal_abi_test -O0 - OUT="$PWD/out/run2" DIR=$(mktemp -d XXXX-short) cabal_abi_test -O0 + OUT="$PWD/out/run1" DIR=$(mktemp -d XXXX-looooooooong) cabal_abi_test -O1 -haddock + OUT="$PWD/out/run2" DIR=$(mktemp -d XXXX-short) cabal_abi_test -O1 -haddock -dunique-increment=-1 -dinitial-unique=16777215 check_interfaces out/run1 out/run2 abis "Mismatched ABI hash" check_interfaces out/run1 out/run2 interfaces "Mismatched interface hashes" } ===================================== compiler/GHC/CoreToStg.hs ===================================== @@ -27,7 +27,8 @@ import GHC.Core.DataCon import GHC.Stg.Syntax import GHC.Stg.Debug -import GHC.Stg.Utils +import GHC.Stg.Make +import GHC.Stg.Utils (allowTopLevelConApp) import GHC.Types.RepType import GHC.Types.Id.Make ( coercionTokenId ) @@ -36,16 +37,13 @@ import GHC.Types.Id.Info import GHC.Types.CostCentre import GHC.Types.Tickish import GHC.Types.Var.Env -import GHC.Types.Name ( isExternalName, nameModule_maybe ) +import GHC.Types.Name ( isExternalName ) import GHC.Types.Basic ( Arity, TypeOrConstraint(..) ) import GHC.Types.Literal import GHC.Types.ForeignCall import GHC.Types.IPE -import GHC.Types.Demand ( isAtMostOnceDmd ) -import GHC.Types.SrcLoc ( mkGeneralSrcSpan ) import GHC.Unit.Module -import GHC.Data.FastString import GHC.Platform ( Platform ) import GHC.Platform.Ways import GHC.Builtin.PrimOps @@ -338,10 +336,12 @@ coreToTopStgRhs -> CtsM (CollectedCCs, (Id, StgRhs)) coreToTopStgRhs opts this_mod ccs (bndr, rhs) - = do { new_rhs <- coreToPreStgRhs rhs + = do { new_rhs <- coreToMkStgRhs bndr rhs ; let (stg_rhs, ccs') = - mkTopStgRhs opts this_mod ccs bndr new_rhs + mkTopStgRhs (allowTopLevelConApp (coreToStg_platform opts) (coreToStg_ExternalDynamicRefs opts)) + (coreToStg_AutoSccsOnIndividualCafs opts) + this_mod ccs bndr new_rhs stg_arity = stgRhsArity stg_rhs @@ -372,7 +372,7 @@ coreToTopStgRhs opts this_mod ccs (bndr, rhs) -- coreToStgExpr panics if the input expression is a value lambda. CorePrep -- ensures that value lambdas only exist as the RHS of bindings, which we --- handle with the function coreToPreStgRhs. +-- handle with the function coreToMkStgRhs. coreToStgExpr :: HasDebugCallStack => CoreExpr @@ -685,166 +685,24 @@ coreToStgRhs :: (Id,CoreExpr) -> CtsM StgRhs coreToStgRhs (bndr, rhs) = do - new_rhs <- coreToPreStgRhs rhs + new_rhs <- coreToMkStgRhs bndr rhs return (mkStgRhs bndr new_rhs) --- Represents the RHS of a binding for use with mk(Top)StgRhs. -data PreStgRhs = PreStgRhs [Id] StgExpr Type -- The [Id] is empty for thunks - -- Convert the RHS of a binding from Core to STG. This is a wrapper around -- coreToStgExpr that can handle value lambdas. -coreToPreStgRhs :: HasDebugCallStack => CoreExpr -> CtsM PreStgRhs -coreToPreStgRhs expr - = extendVarEnvCts [ (a, LambdaBound) | a <- args' ] $ - do { body' <- coreToStgExpr body - ; return (PreStgRhs args' body' (exprType body)) } - where - (args, body) = myCollectBinders expr - args' = filterStgBinders args - --- Generate a top-level RHS. Any new cost centres generated for CAFs will be --- appended to `CollectedCCs` argument. -mkTopStgRhs :: CoreToStgOpts -> Module -> CollectedCCs - -> Id -> PreStgRhs -> (StgRhs, CollectedCCs) - -mkTopStgRhs CoreToStgOpts - { coreToStg_platform = platform - , coreToStg_ExternalDynamicRefs = opt_ExternalDynamicRefs - , coreToStg_AutoSccsOnIndividualCafs = opt_AutoSccsOnIndividualCafs - } this_mod ccs bndr (PreStgRhs bndrs rhs typ) - | not (null bndrs) - = -- The list of arguments is non-empty, so not CAF - ( StgRhsClosure noExtFieldSilent - dontCareCCS - ReEntrant - bndrs rhs typ - , ccs ) - - -- After this point we know that `bndrs` is empty, - -- so this is not a function binding - | StgConApp con mn args _ <- unticked_rhs - , -- Dynamic StgConApps are updatable - not (isDllConApp platform opt_ExternalDynamicRefs this_mod con args) - = -- CorePrep does this right, but just to make sure - assertPpr (not (isUnboxedTupleDataCon con || isUnboxedSumDataCon con)) - (ppr bndr $$ ppr con $$ ppr args) - ( StgRhsCon dontCareCCS con mn ticks args typ, ccs ) - - -- Otherwise it's a CAF, see Note [Cost-centre initialization plan]. - | opt_AutoSccsOnIndividualCafs - = ( StgRhsClosure noExtFieldSilent - caf_ccs - upd_flag [] rhs typ - , collectCC caf_cc caf_ccs ccs ) - - | otherwise - = ( StgRhsClosure noExtFieldSilent - all_cafs_ccs - upd_flag [] rhs typ - , ccs ) - - where - (ticks, unticked_rhs) = stripStgTicksTop (not . tickishIsCode) rhs - - upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry - | otherwise = Updatable - - -- CAF cost centres generated for -fcaf-all - caf_cc = mkAutoCC bndr modl - caf_ccs = mkSingletonCCS caf_cc - -- careful: the binder might be :Main.main, - -- which doesn't belong to module mod_name. - -- bug #249, tests prof001, prof002 - modl | Just m <- nameModule_maybe (idName bndr) = m - | otherwise = this_mod - - -- default CAF cost centre - (_, all_cafs_ccs) = getAllCAFsCC this_mod - --- Generate a non-top-level RHS. Cost-centre is always currentCCS, --- see Note [Cost-centre initialization plan]. -mkStgRhs :: Id -> PreStgRhs -> StgRhs -mkStgRhs bndr (PreStgRhs bndrs rhs typ) - | not (null bndrs) - = StgRhsClosure noExtFieldSilent - currentCCS - ReEntrant - bndrs rhs typ - - -- After this point we know that `bndrs` is empty, - -- so this is not a function binding - - | isJoinId bndr -- Must be a nullary join point - = -- It might have /type/ arguments (T18328), - -- so its JoinArity might be >0 - StgRhsClosure noExtFieldSilent - currentCCS - ReEntrant -- ignored for LNE - [] rhs typ - - | StgConApp con mn args _ <- unticked_rhs - = StgRhsCon currentCCS con mn ticks args typ - - | otherwise - = StgRhsClosure noExtFieldSilent - currentCCS - upd_flag [] rhs typ - where - (ticks, unticked_rhs) = stripStgTicksTop (not . tickishIsCode) rhs - - upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry - | otherwise = Updatable - - {- - SDM: disabled. Eval/Apply can't handle functions with arity zero very - well; and making these into simple non-updatable thunks breaks other - assumptions (namely that they will be entered only once). - - upd_flag | isPAP env rhs = ReEntrant - | otherwise = Updatable - --- Detect thunks which will reduce immediately to PAPs, and make them --- non-updatable. This has several advantages: --- --- - the non-updatable thunk behaves exactly like the PAP, --- --- - the thunk is more efficient to enter, because it is --- specialised to the task. --- --- - we save one update frame, one stg_update_PAP, one update --- and lots of PAP_enters. --- --- - in the case where the thunk is top-level, we save building --- a black hole and furthermore the thunk isn't considered to --- be a CAF any more, so it doesn't appear in any SRTs. --- --- We do it here, because the arity information is accurate, and we need --- to do it before the SRT pass to save the SRT entries associated with --- any top-level PAPs. - -isPAP env (StgApp f args) = listLengthCmp args arity == LT -- idArity f > length args - where - arity = stgArity f (lookupBinding env f) -isPAP env _ = False - --} - -{- ToDo: - upd = if isOnceDem dem - then (if isNotTop toplev - then SingleEntry -- HA! Paydirt for "dem" - else - (if debugIsOn then trace "WARNING: SE CAFs unsupported, forcing UPD instead" else id) $ - Updatable) - else Updatable - -- For now we forbid SingleEntry CAFs; they tickle the - -- ASSERT in rts/Storage.c line 215 at newCAF() re mut_link, - -- and I don't understand why. There's only one SE_CAF (well, - -- only one that tickled a great gaping bug in an earlier attempt - -- at ClosureInfo.getEntryConvention) in the whole of nofib, - -- specifically Main.lvl6 in spectral/cryptarithm2. - -- So no great loss. KSW 2000-07. --} +coreToMkStgRhs :: HasDebugCallStack => Id -> CoreExpr -> CtsM MkStgRhs +coreToMkStgRhs bndr expr = do + let (args, body) = myCollectBinders expr + let args' = filterStgBinders args + extendVarEnvCts [ (a, LambdaBound) | a <- args' ] $ do + body' <- coreToStgExpr body + let mk_rhs = MkStgRhs + { rhs_args = args' + , rhs_expr = body' + , rhs_type = exprType body + , rhs_is_join = isJoinId bndr + } + pure mk_rhs -- --------------------------------------------------------------------------- -- A monad for the core-to-STG pass @@ -933,15 +791,6 @@ lookupBinding env v = case lookupVarEnv env v of Just xx -> xx Nothing -> assertPpr (isGlobalId v) (ppr v) ImportBound -getAllCAFsCC :: Module -> (CostCentre, CostCentreStack) -getAllCAFsCC this_mod = - let - span = mkGeneralSrcSpan (mkFastString "") -- XXX do better - all_cafs_cc = mkAllCafsCC this_mod span - all_cafs_ccs = mkSingletonCCS all_cafs_cc - in - (all_cafs_cc, all_cafs_ccs) - -- Misc. filterStgBinders :: [Var] -> [Var] ===================================== compiler/GHC/Driver/Config/Stg/Pipeline.hs ===================================== @@ -7,6 +7,7 @@ import GHC.Prelude import Control.Monad (guard) import GHC.Stg.Pipeline +import GHC.Stg.Utils import GHC.Driver.Config.Diagnostic import GHC.Driver.Config.Stg.Lift @@ -15,15 +16,19 @@ import GHC.Driver.DynFlags -- | Initialize STG pretty-printing options from DynFlags initStgPipelineOpts :: DynFlags -> Bool -> StgPipelineOpts -initStgPipelineOpts dflags for_bytecode = StgPipelineOpts - { stgPipeline_lint = do - guard $ gopt Opt_DoStgLinting dflags - Just $ initDiagOpts dflags - , stgPipeline_pprOpts = initStgPprOpts dflags - , stgPipeline_phases = getStgToDo for_bytecode dflags - , stgPlatform = targetPlatform dflags - , stgPipeline_forBytecode = for_bytecode - } +initStgPipelineOpts dflags for_bytecode = + let !platform = targetPlatform dflags + !ext_dyn_refs = gopt Opt_ExternalDynamicRefs dflags + in StgPipelineOpts + { stgPipeline_lint = do + guard $ gopt Opt_DoStgLinting dflags + Just $ initDiagOpts dflags + , stgPipeline_pprOpts = initStgPprOpts dflags + , stgPipeline_phases = getStgToDo for_bytecode dflags + , stgPlatform = platform + , stgPipeline_forBytecode = for_bytecode + , stgPipeline_allowTopLevelConApp = allowTopLevelConApp platform ext_dyn_refs + } -- | Which Stg-to-Stg passes to run. Depends on flags, ways etc. getStgToDo ===================================== compiler/GHC/Stg/Make.hs ===================================== @@ -0,0 +1,172 @@ +module GHC.Stg.Make + ( MkStgRhs (..) + , mkTopStgRhs + , mkStgRhs + , mkStgRhsCon_maybe + , mkTopStgRhsCon_maybe + ) +where + +import GHC.Prelude +import GHC.Unit.Module + +import GHC.Core.DataCon +import GHC.Core.Type (Type) + +import GHC.Stg.Syntax +import GHC.Stg.Utils (stripStgTicksTop) + +import GHC.Types.Id +import GHC.Types.Name +import GHC.Types.CostCentre +import GHC.Types.Demand ( isAtMostOnceDmd ) +import GHC.Types.Tickish + +-- Represents the RHS of a binding for use with mk(Top)StgRhs and +-- mk(Top)StgRhsCon_maybe. +data MkStgRhs = MkStgRhs + { rhs_args :: [Id] -- ^ Empty for thunks + , rhs_expr :: StgExpr -- ^ RHS expression + , rhs_type :: Type -- ^ RHS type (only used in the JS backend: layering violation) + , rhs_is_join :: !Bool -- ^ Is it a RHS for a join-point? + } + + +-- Generate a top-level RHS. Any new cost centres generated for CAFs will be +-- appended to `CollectedCCs` argument. +mkTopStgRhs :: (Module -> DataCon -> [StgArg] -> Bool) + -> Bool -> Module -> CollectedCCs + -> Id -> MkStgRhs -> (StgRhs, CollectedCCs) +mkTopStgRhs allow_toplevel_con_app opt_AutoSccsOnIndividualCafs this_mod ccs bndr mk_rhs@(MkStgRhs bndrs rhs typ _) + -- try to make a StgRhsCon first + | Just rhs_con <- mkTopStgRhsCon_maybe (allow_toplevel_con_app this_mod) mk_rhs + = ( rhs_con, ccs ) + + | not (null bndrs) + = -- The list of arguments is non-empty, so not CAF + ( StgRhsClosure noExtFieldSilent + dontCareCCS + ReEntrant + bndrs rhs typ + , ccs ) + + -- Otherwise it's a CAF, see Note [Cost-centre initialization plan]. + | opt_AutoSccsOnIndividualCafs + = ( StgRhsClosure noExtFieldSilent + caf_ccs + upd_flag [] rhs typ + , collectCC caf_cc caf_ccs ccs ) + + | otherwise + = ( StgRhsClosure noExtFieldSilent + all_cafs_ccs + upd_flag [] rhs typ + , ccs ) + + where + upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry + | otherwise = Updatable + + -- CAF cost centres generated for -fcaf-all + caf_cc = mkAutoCC bndr modl + caf_ccs = mkSingletonCCS caf_cc + -- careful: the binder might be :Main.main, + -- which doesn't belong to module mod_name. + -- bug #249, tests prof001, prof002 + modl | Just m <- nameModule_maybe (idName bndr) = m + | otherwise = this_mod + + -- default CAF cost centre + (_, all_cafs_ccs) = getAllCAFsCC this_mod + +-- Generate a non-top-level RHS. Cost-centre is always currentCCS, +-- see Note [Cost-centre initialization plan]. +mkStgRhs :: Id -> MkStgRhs -> StgRhs +mkStgRhs bndr mk_rhs@(MkStgRhs bndrs rhs typ is_join) + -- try to make a StgRhsCon first + | Just rhs_con <- mkStgRhsCon_maybe mk_rhs + = rhs_con + + | otherwise + = StgRhsClosure noExtFieldSilent + currentCCS + upd_flag bndrs rhs typ + where + upd_flag | is_join = JumpedTo + | not (null bndrs) = ReEntrant + | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry + | otherwise = Updatable + + {- + SDM: disabled. Eval/Apply can't handle functions with arity zero very + well; and making these into simple non-updatable thunks breaks other + assumptions (namely that they will be entered only once). + + upd_flag | isPAP env rhs = ReEntrant + | otherwise = Updatable + +-- Detect thunks which will reduce immediately to PAPs, and make them +-- non-updatable. This has several advantages: +-- +-- - the non-updatable thunk behaves exactly like the PAP, +-- +-- - the thunk is more efficient to enter, because it is +-- specialised to the task. +-- +-- - we save one update frame, one stg_update_PAP, one update +-- and lots of PAP_enters. +-- +-- - in the case where the thunk is top-level, we save building +-- a black hole and furthermore the thunk isn't considered to +-- be a CAF any more, so it doesn't appear in any SRTs. +-- +-- We do it here, because the arity information is accurate, and we need +-- to do it before the SRT pass to save the SRT entries associated with +-- any top-level PAPs. + +isPAP env (StgApp f args) = listLengthCmp args arity == LT -- idArity f > length args + where + arity = stgArity f (lookupBinding env f) +isPAP env _ = False + +-} + +{- ToDo: + upd = if isOnceDem dem + then (if isNotTop toplev + then SingleEntry -- HA! Paydirt for "dem" + else + (if debugIsOn then trace "WARNING: SE CAFs unsupported, forcing UPD instead" else id) $ + Updatable) + else Updatable + -- For now we forbid SingleEntry CAFs; they tickle the + -- ASSERT in rts/Storage.c line 215 at newCAF() re mut_link, + -- and I don't understand why. There's only one SE_CAF (well, + -- only one that tickled a great gaping bug in an earlier attempt + -- at ClosureInfo.getEntryConvention) in the whole of nofib, + -- specifically Main.lvl6 in spectral/cryptarithm2. + -- So no great loss. KSW 2000-07. +-} + + +-- | Try to make a non top-level StgRhsCon if appropriate +mkStgRhsCon_maybe :: MkStgRhs -> Maybe StgRhs +mkStgRhsCon_maybe (MkStgRhs bndrs rhs typ is_join) + | [] <- bndrs + , not is_join + , (ticks, StgConApp con mn args _) <- stripStgTicksTop (not . tickishIsCode) rhs + = Just (StgRhsCon currentCCS con mn ticks args typ) + + | otherwise = Nothing + + +-- | Try to make a top-level StgRhsCon if appropriate +mkTopStgRhsCon_maybe :: (DataCon -> [StgArg] -> Bool) -> MkStgRhs -> Maybe StgRhs +mkTopStgRhsCon_maybe allow_static_con_app (MkStgRhs bndrs rhs typ is_join) + | [] <- bndrs + , not is_join -- shouldn't happen at top-level + , (ticks, StgConApp con mn args _) <- stripStgTicksTop (not . tickishIsCode) rhs + , allow_static_con_app con args + = Just (StgRhsCon dontCareCCS con mn ticks args typ) + + | otherwise = Nothing ===================================== compiler/GHC/Stg/Pipeline.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Stg.CSE ( stgCse ) import GHC.Stg.Lift ( StgLiftConfig, stgLiftLams ) import GHC.Unit.Module ( Module ) +import GHC.Core.DataCon (DataCon) + import GHC.Utils.Error import GHC.Types.Var import GHC.Types.Var.Set @@ -52,6 +54,12 @@ data StgPipelineOpts = StgPipelineOpts , stgPipeline_pprOpts :: !StgPprOpts , stgPlatform :: !Platform , stgPipeline_forBytecode :: !Bool + + , stgPipeline_allowTopLevelConApp :: Module -> DataCon -> [StgArg] -> Bool + -- ^ Is a top-level (static) StgConApp allowed or not. If not, use dynamic allocation. + -- + -- This is typically used to support dynamic linking on Windows and the + -- -fexternal-dynamic-refs flag. See GHC.Stg.Utils.allowTopLevelConApp. } newtype StgM a = StgM { _unStgM :: ReaderT Char IO a } @@ -136,7 +144,7 @@ stg2stg logger extra_vars opts this_mod binds StgUnarise -> do us <- getUniqueSupplyM liftIO (stg_linter False "Pre-unarise" binds) - let binds' = {-# SCC "StgUnarise" #-} unarise us binds + let binds' = {-# SCC "StgUnarise" #-} unarise us (stgPipeline_allowTopLevelConApp opts this_mod) binds liftIO (dump_when Opt_D_dump_stg_unarised "Unarised STG:" binds') liftIO (stg_linter True "Unarise" binds') return binds' ===================================== compiler/GHC/Stg/Stats.hs ===================================== @@ -46,6 +46,7 @@ data CounterType | ReEntrantBinds Bool{-ditto-} | SingleEntryBinds Bool{-ditto-} | UpdatableBinds Bool{-ditto-} + | JoinPointBinds Bool{-ditto-} deriving (Eq, Ord) type Count = Int @@ -94,6 +95,7 @@ showStgStats prog s (ReEntrantBinds _) = "ReEntrantBindsBinds_Nested " s (SingleEntryBinds _) = "SingleEntryBinds_Nested " s (UpdatableBinds _) = "UpdatableBinds_Nested " + s (JoinPointBinds _) = "JoinPointBinds_Nested " gatherStgStats :: [StgTopBinding] -> StatEnv gatherStgStats binds = combineSEs (map statTopBinding binds) @@ -132,6 +134,7 @@ statRhs top (_, StgRhsClosure _ _ u _ body _) ReEntrant -> ReEntrantBinds top Updatable -> UpdatableBinds top SingleEntry -> SingleEntryBinds top + JumpedTo -> JoinPointBinds top ) {- ===================================== compiler/GHC/Stg/Syntax.hs ===================================== @@ -54,7 +54,6 @@ module GHC.Stg.Syntax ( -- utils stgRhsArity, freeVarsOfRhs, - isDllConApp, stgArgType, stgArgRep, stgArgRep1, @@ -87,17 +86,14 @@ import GHC.Core.Ppr( {- instances -} ) import GHC.Types.ForeignCall ( ForeignCall ) import GHC.Types.Id -import GHC.Types.Name ( isDynLinkName ) import GHC.Types.Tickish ( StgTickish ) import GHC.Types.Var.Set import GHC.Types.Literal ( Literal, literalType ) import GHC.Types.RepType ( typePrimRep, typePrimRep1, typePrimRepU, typePrimRep_maybe ) -import GHC.Unit.Module ( Module ) import GHC.Utils.Outputable import GHC.Utils.Panic.Plain -import GHC.Platform import GHC.Builtin.PrimOps ( PrimOp, PrimCall ) import Data.ByteString ( ByteString ) @@ -138,51 +134,6 @@ data StgArg = StgVarArg Id | StgLitArg Literal --- | Does this constructor application refer to anything in a different --- *Windows* DLL? --- If so, we can't allocate it statically -isDllConApp - :: Platform - -> Bool -- is Opt_ExternalDynamicRefs enabled? - -> Module - -> DataCon - -> [StgArg] - -> Bool -isDllConApp platform ext_dyn_refs this_mod con args - | not ext_dyn_refs = False - | platformOS platform == OSMinGW32 - = isDynLinkName platform this_mod (dataConName con) || any is_dll_arg args - | otherwise = False - where - -- NB: typePrimRep1 is legit because any free variables won't have - -- unlifted type (there are no unlifted things at top level) - is_dll_arg :: StgArg -> Bool - is_dll_arg (StgVarArg v) = isAddrRep (typePrimRep1 (idType v)) - && isDynLinkName platform this_mod (idName v) - is_dll_arg _ = False - --- True of machine addresses; these are the things that don't work across DLLs. --- The key point here is that VoidRep comes out False, so that a top level --- nullary GADT constructor is False for isDllConApp --- --- data T a where --- T1 :: T Int --- --- gives --- --- T1 :: forall a. (a~Int) -> T a --- --- and hence the top-level binding --- --- $WT1 :: T Int --- $WT1 = T1 Int (Coercion (Refl Int)) --- --- The coercion argument here gets VoidRep -isAddrRep :: PrimOrVoidRep -> Bool -isAddrRep (NVRep AddrRep) = True -isAddrRep (NVRep (BoxedRep _)) = True -- FIXME: not true for JavaScript -isAddrRep _ = False - -- | Type of an @StgArg@ -- -- Very half baked because we have lost the type arguments. @@ -721,24 +672,35 @@ UpdateFlag This is also used in @LambdaFormInfo@ in the @ClosureInfo@ module. -A @ReEntrant@ closure may be entered multiple times, but should not be updated -or blackholed. An @Updatable@ closure should be updated after evaluation (and -may be blackholed during evaluation). A @SingleEntry@ closure will only be -entered once, and so need not be updated but may safely be blackholed. -} -data UpdateFlag = ReEntrant | Updatable | SingleEntry +data UpdateFlag + = ReEntrant + -- ^ A @ReEntrant@ closure may be entered multiple times, but should not + -- be updated or blackholed. + | Updatable + -- ^ An @Updatable@ closure should be updated after evaluation (and may be + -- blackholed during evaluation). + | SingleEntry + -- ^ A @SingleEntry@ closure will only be entered once, and so need not be + -- updated but may safely be blackholed. + | JumpedTo + -- ^ A @JumpedTo@ (join-point) closure is entered once or multiple times + -- but has no heap-allocated associated closure. + deriving (Show,Eq) instance Outputable UpdateFlag where ppr u = char $ case u of ReEntrant -> 'r' Updatable -> 'u' SingleEntry -> 's' + JumpedTo -> 'j' isUpdatable :: UpdateFlag -> Bool isUpdatable ReEntrant = False isUpdatable SingleEntry = False isUpdatable Updatable = True +isUpdatable JumpedTo = False {- ************************************************************************ ===================================== compiler/GHC/Stg/Unarise.hs ===================================== @@ -1,6 +1,7 @@ {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE TupleSections #-} +{-# LANGUAGE MultiWayIf #-} {- (c) The GRASP/AQUA Project, Glasgow University, 1992-2012 @@ -401,6 +402,7 @@ import GHC.Utils.Panic import GHC.Types.RepType import GHC.Stg.Syntax import GHC.Stg.Utils +import GHC.Stg.Make import GHC.Core.Type import GHC.Builtin.Types.Prim (intPrimTy) import GHC.Builtin.Types @@ -442,10 +444,14 @@ import Data.List (mapAccumL) -- INVARIANT: OutStgArgs in the range only have NvUnaryTypes -- (i.e. no unboxed tuples, sums or voids) -- -newtype UnariseEnv = UnariseEnv { ue_rho :: (VarEnv UnariseVal) } +data UnariseEnv = UnariseEnv + { ue_rho :: (VarEnv UnariseVal) + , ue_allow_static_conapp :: DataCon -> [StgArg] -> Bool + } -initUnariseEnv :: VarEnv UnariseVal -> UnariseEnv +initUnariseEnv :: VarEnv UnariseVal -> (DataCon -> [StgArg] -> Bool) -> UnariseEnv initUnariseEnv = UnariseEnv + data UnariseVal = MultiVal [OutStgArg] -- MultiVal to tuple. Can be empty list (void). | UnaryVal OutStgArg -- See Note [Renaming during unarisation]. @@ -477,27 +483,57 @@ lookupRho env v = lookupVarEnv (ue_rho env) v -------------------------------------------------------------------------------- -unarise :: UniqSupply -> [StgTopBinding] -> [StgTopBinding] -unarise us binds = initUs_ us (mapM (unariseTopBinding (initUnariseEnv emptyVarEnv)) binds) +unarise :: UniqSupply -> (DataCon -> [StgArg] -> Bool) -> [StgTopBinding] -> [StgTopBinding] +unarise us is_dll_con_app binds = initUs_ us (mapM (unariseTopBinding (initUnariseEnv emptyVarEnv is_dll_con_app)) binds) unariseTopBinding :: UnariseEnv -> StgTopBinding -> UniqSM StgTopBinding unariseTopBinding rho (StgTopLifted bind) - = StgTopLifted <$> unariseBinding rho bind + = StgTopLifted <$> unariseBinding rho True bind unariseTopBinding _ bind at StgTopStringLit{} = return bind -unariseBinding :: UnariseEnv -> StgBinding -> UniqSM StgBinding -unariseBinding rho (StgNonRec x rhs) - = StgNonRec x <$> unariseRhs rho rhs -unariseBinding rho (StgRec xrhss) - = StgRec <$> mapM (\(x, rhs) -> (x,) <$> unariseRhs rho rhs) xrhss +unariseBinding :: UnariseEnv -> Bool -> StgBinding -> UniqSM StgBinding +unariseBinding rho top_level (StgNonRec x rhs) + = StgNonRec x <$> unariseRhs rho top_level rhs +unariseBinding rho top_level (StgRec xrhss) + = StgRec <$> mapM (\(x, rhs) -> (x,) <$> unariseRhs rho top_level rhs) xrhss -unariseRhs :: UnariseEnv -> StgRhs -> UniqSM StgRhs -unariseRhs rho (StgRhsClosure ext ccs update_flag args expr typ) +unariseRhs :: UnariseEnv -> Bool -> StgRhs -> UniqSM StgRhs +unariseRhs rho top_level (StgRhsClosure ext ccs update_flag args expr typ) = do (rho', args1) <- unariseFunArgBinders rho args expr' <- unariseExpr rho' expr - return (StgRhsClosure ext ccs update_flag args1 expr' typ) - -unariseRhs rho (StgRhsCon ccs con mu ts args typ) + -- Unarisation can lead to a StgRhsClosure becoming a StgRhsCon. + -- Hence, we call `mk(Top)StgRhsCon_maybe` rather than just building + -- another `StgRhsClosure`. + -- + -- For example with unboxed sums (#25166): + -- + -- foo = \u [] case (# | _ | #) [(##)] of tag { __DEFAULT -> D [True tag] } + -- + -- ====> {unarisation} + -- + -- foo = D [True 2#] + -- + -- Transforming an appropriate StgRhsClosure into a StgRhsCon is + -- important as top-level StgRhsCon are statically allocated. + -- + let mk_rhs = MkStgRhs + { rhs_args = args1 + , rhs_expr = expr' + , rhs_type = typ + , rhs_is_join = update_flag == JumpedTo + } + if | top_level + , Just rhs_con <- mkTopStgRhsCon_maybe (ue_allow_static_conapp rho) mk_rhs + -> pure rhs_con + + | not top_level + , Just rhs_con <- mkStgRhsCon_maybe mk_rhs + -> pure rhs_con + + | otherwise + -> pure (StgRhsClosure ext ccs update_flag args1 expr' typ) + +unariseRhs rho _top (StgRhsCon ccs con mu ts args typ) = assert (not (isUnboxedTupleDataCon con || isUnboxedSumDataCon con)) return (StgRhsCon ccs con mu ts (unariseConArgs rho args) typ) @@ -576,10 +612,10 @@ unariseExpr rho (StgCase scrut bndr alt_ty alts) -- dead after unarise (checked in GHC.Stg.Lint) unariseExpr rho (StgLet ext bind e) - = StgLet ext <$> unariseBinding rho bind <*> unariseExpr rho e + = StgLet ext <$> unariseBinding rho False bind <*> unariseExpr rho e unariseExpr rho (StgLetNoEscape ext bind e) - = StgLetNoEscape ext <$> unariseBinding rho bind <*> unariseExpr rho e + = StgLetNoEscape ext <$> unariseBinding rho False bind <*> unariseExpr rho e unariseExpr rho (StgTick tick e) = StgTick tick <$> unariseExpr rho e ===================================== compiler/GHC/Stg/Utils.hs ===================================== @@ -9,9 +9,12 @@ module GHC.Stg.Utils , idArgs , mkUnarisedId, mkUnarisedIds + + , allowTopLevelConApp ) where import GHC.Prelude +import GHC.Platform import GHC.Types.Id import GHC.Core.Type @@ -22,6 +25,8 @@ import GHC.Types.Tickish import GHC.Types.Unique.Supply import GHC.Types.RepType +import GHC.Types.Name ( isDynLinkName ) +import GHC.Unit.Module ( Module ) import GHC.Stg.Syntax import GHC.Utils.Outputable @@ -122,3 +127,54 @@ stripStgTicksTopE :: (StgTickish -> Bool) -> GenStgExpr p -> GenStgExpr p stripStgTicksTopE p = go where go (StgTick t e) | p t = go e go other = other + +-- | Do we allow the given top-level (static) ConApp? +allowTopLevelConApp + :: Platform + -> Bool -- is Opt_ExternalDynamicRefs enabled? + -> Module + -> DataCon + -> [StgArg] + -> Bool +allowTopLevelConApp platform ext_dyn_refs this_mod con args + -- we're not using dynamic linking + | not ext_dyn_refs = True + -- if the target OS is Windows, we only allow top-level ConApps if they don't + -- reference external names (Windows DLLs have a problem with static cross-DLL + -- refs) + | platformOS platform == OSMinGW32 = not is_external_con_app + -- otherwise, allowed + -- Sylvain: shouldn't this be False when (ext_dyn_refs && is_external_con_app)? + | otherwise = True + where + is_external_con_app = isDynLinkName platform this_mod (dataConName con) || any is_dll_arg args + + -- NB: typePrimRep1 is legit because any free variables won't have + -- unlifted type (there are no unlifted things at top level) + is_dll_arg :: StgArg -> Bool + is_dll_arg (StgVarArg v) = isAddrRep (typePrimRep1 (idType v)) + && isDynLinkName platform this_mod (idName v) + is_dll_arg _ = False + +-- True of machine addresses; these are the things that don't work across DLLs. +-- The key point here is that VoidRep comes out False, so that a top level +-- nullary GADT constructor is True for allowTopLevelConApp +-- +-- data T a where +-- T1 :: T Int +-- +-- gives +-- +-- T1 :: forall a. (a~Int) -> T a +-- +-- and hence the top-level binding +-- +-- $WT1 :: T Int +-- $WT1 = T1 Int (Coercion (Refl Int)) +-- +-- The coercion argument here gets VoidRep +isAddrRep :: PrimOrVoidRep -> Bool +isAddrRep (NVRep AddrRep) = True +isAddrRep (NVRep (BoxedRep _)) = True -- FIXME: not true for JavaScript +isAddrRep _ = False + ===================================== compiler/GHC/StgToCmm/DataCon.hs ===================================== @@ -19,6 +19,7 @@ import GHC.Prelude import GHC.Platform +import GHC.Stg.Utils (allowTopLevelConApp) import GHC.Stg.Syntax import GHC.Core ( AltCon(..) ) @@ -48,7 +49,6 @@ import GHC.Utils.Panic import GHC.Utils.Misc import GHC.Utils.Monad (mapMaybeM) -import Control.Monad import Data.Char import GHC.StgToCmm.Config (stgToCmmPlatform) import GHC.StgToCmm.TagCheck (checkConArgsStatic, checkConArgsDyn) @@ -90,10 +90,8 @@ cgTopRhsCon cfg id con mn args gen_code = do { profile <- getProfile ; this_mod <- getModuleName - ; when (platformOS platform == OSMinGW32) $ - -- Windows DLLs have a problem with static cross-DLL refs. - massert (not (isDllConApp platform (stgToCmmExtDynRefs cfg) this_mod con (map fromNonVoid args))) - ; assert (args `lengthIs` countConRepArgs con ) return () + ; massert (allowTopLevelConApp platform (stgToCmmExtDynRefs cfg) this_mod con (map fromNonVoid args)) + ; massert (args `lengthIs` countConRepArgs con ) ; checkConArgsStatic (text "TagCheck failed - Top level con") con (map fromNonVoid args) -- LAY IT OUT ; let ===================================== compiler/GHC/Types/CostCentre.hs ===================================== @@ -4,6 +4,7 @@ module GHC.Types.CostCentre ( CostCentre(..), CcName, CCFlavour, mkCafFlavour, mkExprCCFlavour, mkDeclCCFlavour, mkHpcCCFlavour, mkLateCCFlavour, mkCallerCCFlavour, + getAllCAFsCC, pprCostCentre, CostCentreStack, @@ -393,3 +394,13 @@ instance Binary CostCentre where -- ok, because we only need the SrcSpan when declaring the -- CostCentre in the original module, it is not used by importing -- modules. + +getAllCAFsCC :: Module -> (CostCentre, CostCentreStack) +getAllCAFsCC this_mod = + let + span = mkGeneralSrcSpan (mkFastString "") -- XXX do better + all_cafs_cc = mkAllCafsCC this_mod span + all_cafs_ccs = mkSingletonCCS all_cafs_cc + in + (all_cafs_cc, all_cafs_ccs) + ===================================== compiler/ghc.cabal.in ===================================== @@ -702,6 +702,7 @@ Library GHC.Stg.InferTags.Rewrite GHC.Stg.InferTags.TagSig GHC.Stg.InferTags.Types + GHC.Stg.Make GHC.Stg.Pipeline GHC.Stg.Stats GHC.Stg.Subst ===================================== hadrian/ghci-multi-cabal.in ===================================== @@ -8,6 +8,6 @@ if [[ $(printf "9.4.0\n%s\n" $($RUN_GHC --numeric-version) | sort -uV | head -n set -e export TOOL_OUTPUT=.hadrian_ghci_multi/ghci_args # Replace newlines with spaces, as these otherwise break the ghci invocation on windows. -CABFLAGS=-v0 "hadrian/build-cabal" multi:ghc --build-root=.hadrian_ghci_multi --flavour=ghc-in-ghci $HADRIAN_ARGS +CABFLAGS=-v0 "hadrian/build-cabal" multi --build-root=.hadrian_ghci_multi --flavour=ghc-in-ghci $HADRIAN_ARGS GHC_FLAGS="$GHC_FLAGS $(cat $TOOL_OUTPUT | tr '\n\r' ' ')" $RUN_GHC --interactive $GHC_FLAGS $@ -fno-code -fwrite-interface -O0 +RTS -A128m ===================================== hadrian/src/Rules/ToolArgs.hs ===================================== @@ -16,7 +16,9 @@ import System.Directory (canonicalizePath) import System.Environment (lookupEnv) import qualified Data.Set as Set import Oracles.ModuleFiles +import Oracles.Setting import Utilities +import Data.Version.Extra -- | @tool:@ is used by tooling in order to get the arguments necessary -- to set up a GHC API session which can compile modules from GHC. When @@ -85,7 +87,16 @@ multiSetup pkg_s = do need (srcs ++ gens) let rexp m = ["-reexported-module", m] let hidir = root "interfaces" pkgPath p - writeFile' (resp_file root p) (intercalate "\n" (arg_list + ghcVersion <- ghcVersionStage stage0InTree + let ghc_wired_in = readVersion ghcVersion < makeVersion [9,8,1] + ghc_package_id = "-package-id ghc-" ++ ghcVersion + normalise_ghc = if ghc_wired_in then normalisePackageIds else id + normalisePackageIds :: [String] -> [String] + normalisePackageIds ((isPrefixOf ghc_package_id -> True) : xs) = "-package-id" : "ghc" : xs + normalisePackageIds (x:xs) = x : normalisePackageIds xs + normalisePackageIds [] = [] + + writeFile' (resp_file root p) (intercalate "\n" (normalise_ghc arg_list ++ modules cd ++ concatMap rexp (reexportModules cd) ++ ["-outputdir", hidir])) @@ -150,7 +161,9 @@ toolTargets = [ cabalSyntax , ghcHeap , ghci , ghcPkg -- # executable - -- , haddock -- # depends on ghc library + , haddock -- # depends on ghc library + , haddockApi + , haddockLibrary , hsc2hs -- # executable , hpc , hpcBin -- # executable ===================================== testsuite/tests/codeGen/should_compile/Makefile ===================================== @@ -77,3 +77,6 @@ T17648: -fcatch-nonexhaustive-cases T17648.hs -v0 -fforce-recomp '$(TEST_HC)' --show-iface T17648.hi | tr -d '\n\r' | \ grep -F 'f :: T GHC.Types.Int -> () [TagSig' >/dev/null + +T25166: + '$(TEST_HC)' -O2 -dno-typeable-binds -ddump-cmm T25166.hs | awk '/foo_closure/{flag=1}/}]/{flag=0}flag' ===================================== testsuite/tests/codeGen/should_compile/T25166.hs ===================================== @@ -0,0 +1,7 @@ +module Test where + +data A = A | B | C + +data D = D !Bool {-# UNPACK #-} !A + +foo = D True B ===================================== testsuite/tests/codeGen/should_compile/T25166.stdout ===================================== @@ -0,0 +1,6 @@ +[section ""data" . Test.foo_closure" { + Test.foo_closure: + const Test.D_con_info; + const GHC.Types.True_closure+2; + const 2; + const 3; ===================================== testsuite/tests/codeGen/should_compile/all.T ===================================== @@ -139,5 +139,7 @@ test('callee-no-local', [ ['-ddump-cmm-raw'] ) +test('T25166', [req_cmm], makefile_test, []) + # dump Core to ensure that d is defined as: d = D 10## RUBBISH(IntRep) test('T25177', normal, compile, ['-O2 -dno-typeable-binds -ddump-simpl -dsuppress-all -dsuppress-uniques -v0']) ===================================== utils/haddock/doc/requirements.txt ===================================== @@ -0,0 +1 @@ +sphinxcontrib-applehelp ==2.0.0 ===================================== utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs ===================================== @@ -26,20 +26,20 @@ import Data.Foldable (toList) import Data.List (sort) import Data.List.NonEmpty (NonEmpty (..)) import qualified Data.Map as Map -import Data.Maybe +import qualified Data.Maybe as Maybe import GHC hiding (fromMaybeContext) import GHC.Core.Type (Specificity (..)) import GHC.Data.FastString (unpackFS) import GHC.Types.Name (getOccString, nameOccName, tidyNameOcc) import GHC.Types.Name.Occurrence import GHC.Types.Name.Reader (rdrNameOcc) +import GHC.Utils.Ppr hiding (Doc, quote) +import qualified GHC.Utils.Ppr as Pretty import System.Directory import System.FilePath import Prelude hiding ((<>)) import Documentation.Haddock.Markup -import GHC.Utils.Ppr hiding (Doc, quote) -import qualified GHC.Utils.Ppr as Pretty import Haddock.Doc (combineDocumentation) import Haddock.GhcUtils import Haddock.Types @@ -90,7 +90,7 @@ ppLaTeX ppLaTeX title packageStr visible_ifaces odir prologue maybe_style libdir = do createDirectoryIfMissing True odir - when (isNothing maybe_style) $ + when (Maybe.isNothing maybe_style) $ copyFile (libdir "latex" haddockSty) (odir haddockSty) ppLaTeXTop title packageStr odir prologue maybe_style visible_ifaces mapM_ (ppLaTeXModule title odir) visible_ifaces @@ -139,7 +139,7 @@ ppLaTeXTop doctitle packageStr odir prologue maybe_style ifaces = do mods = sort (map (moduleBasename . ifaceMod) ifaces) - filename = odir (fromMaybe "haddock" packageStr <.> "tex") + filename = odir (Maybe.fromMaybe "haddock" packageStr <.> "tex") writeUtf8File filename (show tex) @@ -174,7 +174,7 @@ ppLaTeXModule _title odir iface = do ] description = - (fromMaybe empty . documentationToLaTeX . ifaceRnDoc) iface + (Maybe.fromMaybe empty . documentationToLaTeX . ifaceRnDoc) iface body = processExports exports -- @@ -201,7 +201,7 @@ exportListItem in sep (punctuate comma [leader <+> ppDocBinder name | name <- names]) <> case subdocs of [] -> empty - _ -> parens (sep (punctuate comma (mapMaybe go subdocs))) + _ -> parens (sep (punctuate comma (Maybe.mapMaybe go subdocs))) exportListItem (ExportNoDecl y []) = ppDocBinder y exportListItem (ExportNoDecl y subs) = @@ -368,7 +368,7 @@ ppFamDecl associated doc instances decl unicode = (if null body then Nothing else Just (vcat body)) $$ instancesBit where - body = catMaybes [familyEqns, documentationToLaTeX doc] + body = Maybe.catMaybes [familyEqns, documentationToLaTeX doc] whereBit = case fdInfo (tcdFam decl) of ClosedTypeFamily _ -> keyword "where" @@ -544,7 +544,7 @@ ppTypeOrFunSig typ (doc, argDocs) (pref1, pref2, sep0) unicode text "\\haddockbeginargs" $$ vcat (map (uncurry (<->)) (ppSubSigLike unicode typ argDocs [] sep0)) $$ text "\\end{tabulary}\\par" - $$ fromMaybe empty (documentationToLaTeX doc) + $$ Maybe.fromMaybe empty (documentationToLaTeX doc) -- | This splits up a type signature along @->@ and adds docs (when they exist) -- to the arguments. The output is a list of (leader/seperator, argument and @@ -741,7 +741,7 @@ ppClassDecl hdr = ppClassHdr False lctxt (unLoc lname) ltyvars lfds - body = catMaybes [documentationToLaTeX doc, body_] + body = Maybe.catMaybes [documentationToLaTeX doc, body_] body_ | null lsigs, null ats, null at_defs = Nothing @@ -764,9 +764,13 @@ ppClassDecl | L _ (ClassOpSig _ is_def lnames typ) <- lsigs , let doc | is_def = noDocForDecl - | otherwise = lookupAnySubdoc (head names) subdocs + | otherwise = lookupAnySubdoc firstName subdocs names = map (cleanName . unLoc) lnames leader = if is_def then Just (keyword "default") else Nothing + firstName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd ] -- N.B. taking just the first name is ok. Signatures with multiple -- names are expanded so that each name gets its own signature. @@ -853,7 +857,7 @@ ppDataDecl pats instances subdocs doc dataDecl unicode = where cons = dd_cons (tcdDataDefn dataDecl) - body = catMaybes [doc >>= documentationToLaTeX, constrBit, patternBit] + body = Maybe.catMaybes [doc >>= documentationToLaTeX, constrBit, patternBit] (whereBit, leaders) | null cons @@ -1031,7 +1035,11 @@ ppSideBySideField subdocs unicode (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc $ head names) subdocs >>= fmap _doc . combineDocumentation . fst + mbDoc = lookup (foExt $ unLoc name) subdocs >>= fmap _doc . combineDocumentation . fst + name = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -- | Pretty-print a bundled pattern synonym ppSideBySidePat @@ -1157,7 +1165,7 @@ ppContextNoLocsMaybe cxt unicode = Just $ pp_hs_context cxt unicode ppContextNoArrow :: HsContext DocNameI -> Bool -> LaTeX ppContextNoArrow cxt unicode = - fromMaybe empty $ + Maybe.fromMaybe empty $ ppContextNoLocsMaybe (map unLoc cxt) unicode ppContextNoLocs :: [HsType DocNameI] -> Bool -> LaTeX ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml.hs ===================================== @@ -6,6 +6,7 @@ {-# LANGUAGE TupleSections #-} {-# LANGUAGE TypeApplications #-} {-# LANGUAGE TypeFamilies #-} +{-# OPTIONS_GHC -Wwarn=x-partial #-} -- | -- Module : Haddock.Backends.Html ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs ===================================== @@ -10,7 +10,7 @@ ----------------------------------------------------------------------------- -- | --- Module : Haddock.Backends.Html.Decl +-- Module : Haddock.Backends.Xhtml.Decl -- Copyright : (c) Simon Marlow 2003-2006, -- David Waern 2006-2009, -- Mark Lentczner 2010 @@ -28,7 +28,7 @@ import Data.Foldable (toList) import Data.List (intersperse, sort) import Data.List.NonEmpty (NonEmpty (..)) import qualified Data.Map as Map -import Data.Maybe +import qualified Data.Maybe as Maybe import GHC hiding (LexicalFixity (..), fromMaybeContext) import GHC.Core.Type (Specificity (..)) import GHC.Data.BooleanFormula @@ -279,13 +279,17 @@ ppTypeOrFunSig qual emptyCtxts | summary = pref1 - | Map.null argDocs = topDeclElem links loc splice docnames pref1 +++ docSection curname pkg qual doc + | Map.null argDocs = topDeclElem links loc splice docName pref1 +++ docSection curname pkg qual doc | otherwise = - topDeclElem links loc splice docnames pref2 + topDeclElem links loc splice docName pref2 +++ subArguments pkg qual (ppSubSigLike unicode qual typ argDocs [] sep emptyCtxts) +++ docSection curname pkg qual doc where - curname = getName <$> listToMaybe docnames + curname = getName <$> Maybe.listToMaybe docnames + docName = + case Maybe.listToMaybe docnames of + Nothing -> error "No docnames. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -- | This splits up a type signature along @->@ and adds docs (when they exist) -- to the arguments. @@ -489,11 +493,15 @@ ppSimpleSig -> HsSigType DocNameI -> Html ppSimpleSig links splice unicode qual emptyCtxts loc names typ = - topDeclElem' names $ ppTypeSig True occNames ppTyp unicode + topDeclElem' docName $ ppTypeSig True occNames ppTyp unicode where topDeclElem' = topDeclElem links loc splice ppTyp = ppSigType unicode qual emptyCtxts typ occNames = map getOccName names + docName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd -------------------------------------------------------------------------------- @@ -530,13 +538,13 @@ ppFamDecl summary associated links instances fixities loc doc decl splice unicod curname = Just $ getName docname header_ = - topDeclElem links loc splice [docname] $ + topDeclElem links loc splice docname $ ppFamHeader summary associated decl unicode qual <+> ppFixities fixities qual instancesBit | FamilyDecl{fdInfo = ClosedTypeFamily mb_eqns} <- decl , not summary = - subEquations pkg qual $ map (ppFamDeclEqn . unLoc) $ fromMaybe [] mb_eqns + subEquations pkg qual $ map (ppFamDeclEqn . unLoc) $ Maybe.fromMaybe [] mb_eqns | otherwise = ppInstances links (OriginFamily docname) instances splice unicode pkg qual @@ -706,7 +714,7 @@ ppLContextNoArrow c u q h = ppContextNoArrow (unLoc c) u q h ppContextNoArrow :: HsContext DocNameI -> Unicode -> Qualification -> HideEmptyContexts -> Html ppContextNoArrow cxt unicode qual emptyCtxts = - fromMaybe noHtml $ + Maybe.fromMaybe noHtml $ ppContextNoLocsMaybe (map unLoc cxt) unicode qual emptyCtxts ppContextNoLocs :: [HsType DocNameI] -> Unicode -> Qualification -> HideEmptyContexts -> Html @@ -790,9 +798,9 @@ ppShortClassDecl pkg qual = if not (any isUserLSig sigs) && null ats - then (if summary then id else topDeclElem links loc splice [nm]) hdr + then (if summary then id else topDeclElem links loc splice nm) hdr else - (if summary then id else topDeclElem links loc splice [nm]) (hdr <+> keyword "where") + (if summary then id else topDeclElem links loc splice nm) (hdr <+> keyword "where") +++ shortSubDecls False ( [ ppAssocType summary links doc at [] splice unicode pkg qual | at <- ats, let doc = lookupAnySubdoc (unL $ fdLName $ unL at) subdocs @@ -814,8 +822,12 @@ ppShortClassDecl pkg qual | L _ (ClassOpSig _ False lnames typ) <- sigs - , let doc = lookupAnySubdoc (head names) subdocs - names = map unLoc lnames + , let names = map unLoc lnames + subdocName = + case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd + doc = lookupAnySubdoc subdocName subdocs ] -- FIXME: is taking just the first name ok? Is it possible that -- there are different subdocs for different names in a single @@ -876,8 +888,8 @@ ppClassDecl sigs = map unLoc lsigs classheader - | any isUserLSig lsigs = topDeclElem links loc splice [nm] (hdr unicode qual <+> keyword "where" <+> fixs) - | otherwise = topDeclElem links loc splice [nm] (hdr unicode qual <+> fixs) + | any isUserLSig lsigs = topDeclElem links loc splice nm (hdr unicode qual <+> keyword "where" <+> fixs) + | otherwise = topDeclElem links loc splice nm (hdr unicode qual <+> fixs) -- Only the fixity relevant to the class header fixs = ppFixities [f | f@(n, _) <- fixities, n == unLoc lname] qual @@ -890,7 +902,7 @@ ppClassDecl atBit = subAssociatedTypes [ ppAssocType summary links doc at subfixs splice unicode pkg qual - <+> subDefaults (maybeToList defTys) + <+> subDefaults (Maybe.maybeToList defTys) | at <- ats , let name = unLoc . fdLName $ unLoc at doc = lookupAnySubdoc name subdocs @@ -941,7 +953,7 @@ ppClassDecl unicode pkg qual - <+> subDefaults (maybeToList defSigs) + <+> subDefaults (Maybe.maybeToList defSigs) | ClassOpSig _ False lnames typ <- sigs , name <- map unLoc lnames , let doc = lookupAnySubdoc name subdocs @@ -1111,7 +1123,7 @@ ppInstanceAssocTys -> [DocInstance DocNameI] -> [Html] ppInstanceAssocTys links splice unicode qual orphan insts = - maybeToList $ + Maybe.maybeToList $ subTableSrc Nothing qual links True $ zipWith mkInstHead @@ -1137,10 +1149,14 @@ ppInstanceSigs links splice unicode qual sigs = do L _ rtyp = dropWildCards typ -- Instance methods signatures are synified and thus don't have a useful -- SrcSpan value. Use the methods name location instead. - return $ ppSimpleSig links splice unicode qual HideEmptyContexts (getLocA $ head lnames) names rtyp + let lname = + case Maybe.listToMaybe lnames of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd + return $ ppSimpleSig links splice unicode qual HideEmptyContexts (getLocA lname) names rtyp lookupAnySubdoc :: Eq id1 => id1 -> [(id1, DocForDecl id2)] -> DocForDecl id2 -lookupAnySubdoc n = fromMaybe noDocForDecl . lookup n +lookupAnySubdoc n = Maybe.fromMaybe noDocForDecl . lookup n instanceId :: InstOrigin DocName -> Int -> Bool -> InstHead DocNameI -> String instanceId origin no orphan ihd = @@ -1256,7 +1272,7 @@ ppDataDecl ConDeclGADT{} -> False header_ = - topDeclElem links loc splice [docname] $ + topDeclElem links loc splice docname $ ppDataHeader summary dataDecl unicode qual <+> whereBit <+> fix fix = ppFixities (filter (\(n, _) -> n == docname) fixities) qual @@ -1531,7 +1547,10 @@ ppSideBySideField subdocs unicode qual (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc $ head names) subdocs >>= combineDocumentation . fst + mbDoc = lookup (foExt $ unLoc declName) subdocs >>= combineDocumentation . fst + declName = case Maybe.listToMaybe names of + Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" + Just hd -> hd ppShortField :: Bool -> Unicode -> Qualification -> ConDeclField DocNameI -> Html ppShortField summary unicode qual (ConDeclField _ names ltype _) = ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Layout.hs ===================================== @@ -311,9 +311,9 @@ declElem = paragraph ! [theclass "src"] -- a box for top level documented names -- it adds a source and wiki link at the right hand side of the box -topDeclElem :: LinksInfo -> SrcSpan -> Bool -> [DocName] -> Html -> Html -topDeclElem lnks loc splice names html = - declElem << (html <+> (links lnks loc splice Nothing $ head names)) +topDeclElem :: LinksInfo -> SrcSpan -> Bool -> DocName -> Html -> Html +topDeclElem lnks loc splice name html = + declElem << (html <+> links lnks loc splice Nothing name) -- FIXME: is it ok to simply take the first name? ===================================== utils/haddock/haddock-api/src/Haddock/Utils/Json.hs ===================================== @@ -371,10 +371,9 @@ instance FromJSON Char where parseJSONList v = typeMismatch "String" v parseChar :: String -> Parser Char -parseChar t = - if length t == 1 - then pure $ head t - else prependContext "Char" $ fail "expected a string of length 1" +parseChar [c] = pure c +parseChar [] = prependContext "Char" $ fail "expected a string of length 1, got an empty string" +parseChar (_ : _) = prependContext "Char" $ fail "expected a string of length 1, got a longer string" parseRealFloat :: RealFloat a => String -> Value -> Parser a parseRealFloat _ (Number s) = pure $ realToFrac s View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/07ba0d216f0a91353db2125b7c44a19219c80213...5bba677e1bd93fc514b66283b87fd281fc620402 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/07ba0d216f0a91353db2125b7c44a19219c80213...5bba677e1bd93fc514b66283b87fd281fc620402 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 20:40:47 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 06 Sep 2024 16:40:47 -0400 Subject: [Git][ghc/ghc][master] Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Message-ID: <66db68cf551b6_19b5b610c75032440@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 15 changed files: - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs - compiler/GHC/Stg/Unarise.hs - compiler/GHC/Stg/Utils.hs - compiler/GHC/StgToCmm/DataCon.hs - compiler/GHC/Types/CostCentre.hs - compiler/ghc.cabal.in - testsuite/tests/codeGen/should_compile/Makefile - + testsuite/tests/codeGen/should_compile/T25166.hs - + testsuite/tests/codeGen/should_compile/T25166.stdout - testsuite/tests/codeGen/should_compile/all.T Changes: ===================================== compiler/GHC/CoreToStg.hs ===================================== @@ -27,7 +27,8 @@ import GHC.Core.DataCon import GHC.Stg.Syntax import GHC.Stg.Debug -import GHC.Stg.Utils +import GHC.Stg.Make +import GHC.Stg.Utils (allowTopLevelConApp) import GHC.Types.RepType import GHC.Types.Id.Make ( coercionTokenId ) @@ -36,16 +37,13 @@ import GHC.Types.Id.Info import GHC.Types.CostCentre import GHC.Types.Tickish import GHC.Types.Var.Env -import GHC.Types.Name ( isExternalName, nameModule_maybe ) +import GHC.Types.Name ( isExternalName ) import GHC.Types.Basic ( Arity, TypeOrConstraint(..) ) import GHC.Types.Literal import GHC.Types.ForeignCall import GHC.Types.IPE -import GHC.Types.Demand ( isAtMostOnceDmd ) -import GHC.Types.SrcLoc ( mkGeneralSrcSpan ) import GHC.Unit.Module -import GHC.Data.FastString import GHC.Platform ( Platform ) import GHC.Platform.Ways import GHC.Builtin.PrimOps @@ -338,10 +336,12 @@ coreToTopStgRhs -> CtsM (CollectedCCs, (Id, StgRhs)) coreToTopStgRhs opts this_mod ccs (bndr, rhs) - = do { new_rhs <- coreToPreStgRhs rhs + = do { new_rhs <- coreToMkStgRhs bndr rhs ; let (stg_rhs, ccs') = - mkTopStgRhs opts this_mod ccs bndr new_rhs + mkTopStgRhs (allowTopLevelConApp (coreToStg_platform opts) (coreToStg_ExternalDynamicRefs opts)) + (coreToStg_AutoSccsOnIndividualCafs opts) + this_mod ccs bndr new_rhs stg_arity = stgRhsArity stg_rhs @@ -372,7 +372,7 @@ coreToTopStgRhs opts this_mod ccs (bndr, rhs) -- coreToStgExpr panics if the input expression is a value lambda. CorePrep -- ensures that value lambdas only exist as the RHS of bindings, which we --- handle with the function coreToPreStgRhs. +-- handle with the function coreToMkStgRhs. coreToStgExpr :: HasDebugCallStack => CoreExpr @@ -685,166 +685,24 @@ coreToStgRhs :: (Id,CoreExpr) -> CtsM StgRhs coreToStgRhs (bndr, rhs) = do - new_rhs <- coreToPreStgRhs rhs + new_rhs <- coreToMkStgRhs bndr rhs return (mkStgRhs bndr new_rhs) --- Represents the RHS of a binding for use with mk(Top)StgRhs. -data PreStgRhs = PreStgRhs [Id] StgExpr Type -- The [Id] is empty for thunks - -- Convert the RHS of a binding from Core to STG. This is a wrapper around -- coreToStgExpr that can handle value lambdas. -coreToPreStgRhs :: HasDebugCallStack => CoreExpr -> CtsM PreStgRhs -coreToPreStgRhs expr - = extendVarEnvCts [ (a, LambdaBound) | a <- args' ] $ - do { body' <- coreToStgExpr body - ; return (PreStgRhs args' body' (exprType body)) } - where - (args, body) = myCollectBinders expr - args' = filterStgBinders args - --- Generate a top-level RHS. Any new cost centres generated for CAFs will be --- appended to `CollectedCCs` argument. -mkTopStgRhs :: CoreToStgOpts -> Module -> CollectedCCs - -> Id -> PreStgRhs -> (StgRhs, CollectedCCs) - -mkTopStgRhs CoreToStgOpts - { coreToStg_platform = platform - , coreToStg_ExternalDynamicRefs = opt_ExternalDynamicRefs - , coreToStg_AutoSccsOnIndividualCafs = opt_AutoSccsOnIndividualCafs - } this_mod ccs bndr (PreStgRhs bndrs rhs typ) - | not (null bndrs) - = -- The list of arguments is non-empty, so not CAF - ( StgRhsClosure noExtFieldSilent - dontCareCCS - ReEntrant - bndrs rhs typ - , ccs ) - - -- After this point we know that `bndrs` is empty, - -- so this is not a function binding - | StgConApp con mn args _ <- unticked_rhs - , -- Dynamic StgConApps are updatable - not (isDllConApp platform opt_ExternalDynamicRefs this_mod con args) - = -- CorePrep does this right, but just to make sure - assertPpr (not (isUnboxedTupleDataCon con || isUnboxedSumDataCon con)) - (ppr bndr $$ ppr con $$ ppr args) - ( StgRhsCon dontCareCCS con mn ticks args typ, ccs ) - - -- Otherwise it's a CAF, see Note [Cost-centre initialization plan]. - | opt_AutoSccsOnIndividualCafs - = ( StgRhsClosure noExtFieldSilent - caf_ccs - upd_flag [] rhs typ - , collectCC caf_cc caf_ccs ccs ) - - | otherwise - = ( StgRhsClosure noExtFieldSilent - all_cafs_ccs - upd_flag [] rhs typ - , ccs ) - - where - (ticks, unticked_rhs) = stripStgTicksTop (not . tickishIsCode) rhs - - upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry - | otherwise = Updatable - - -- CAF cost centres generated for -fcaf-all - caf_cc = mkAutoCC bndr modl - caf_ccs = mkSingletonCCS caf_cc - -- careful: the binder might be :Main.main, - -- which doesn't belong to module mod_name. - -- bug #249, tests prof001, prof002 - modl | Just m <- nameModule_maybe (idName bndr) = m - | otherwise = this_mod - - -- default CAF cost centre - (_, all_cafs_ccs) = getAllCAFsCC this_mod - --- Generate a non-top-level RHS. Cost-centre is always currentCCS, --- see Note [Cost-centre initialization plan]. -mkStgRhs :: Id -> PreStgRhs -> StgRhs -mkStgRhs bndr (PreStgRhs bndrs rhs typ) - | not (null bndrs) - = StgRhsClosure noExtFieldSilent - currentCCS - ReEntrant - bndrs rhs typ - - -- After this point we know that `bndrs` is empty, - -- so this is not a function binding - - | isJoinId bndr -- Must be a nullary join point - = -- It might have /type/ arguments (T18328), - -- so its JoinArity might be >0 - StgRhsClosure noExtFieldSilent - currentCCS - ReEntrant -- ignored for LNE - [] rhs typ - - | StgConApp con mn args _ <- unticked_rhs - = StgRhsCon currentCCS con mn ticks args typ - - | otherwise - = StgRhsClosure noExtFieldSilent - currentCCS - upd_flag [] rhs typ - where - (ticks, unticked_rhs) = stripStgTicksTop (not . tickishIsCode) rhs - - upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry - | otherwise = Updatable - - {- - SDM: disabled. Eval/Apply can't handle functions with arity zero very - well; and making these into simple non-updatable thunks breaks other - assumptions (namely that they will be entered only once). - - upd_flag | isPAP env rhs = ReEntrant - | otherwise = Updatable - --- Detect thunks which will reduce immediately to PAPs, and make them --- non-updatable. This has several advantages: --- --- - the non-updatable thunk behaves exactly like the PAP, --- --- - the thunk is more efficient to enter, because it is --- specialised to the task. --- --- - we save one update frame, one stg_update_PAP, one update --- and lots of PAP_enters. --- --- - in the case where the thunk is top-level, we save building --- a black hole and furthermore the thunk isn't considered to --- be a CAF any more, so it doesn't appear in any SRTs. --- --- We do it here, because the arity information is accurate, and we need --- to do it before the SRT pass to save the SRT entries associated with --- any top-level PAPs. - -isPAP env (StgApp f args) = listLengthCmp args arity == LT -- idArity f > length args - where - arity = stgArity f (lookupBinding env f) -isPAP env _ = False - --} - -{- ToDo: - upd = if isOnceDem dem - then (if isNotTop toplev - then SingleEntry -- HA! Paydirt for "dem" - else - (if debugIsOn then trace "WARNING: SE CAFs unsupported, forcing UPD instead" else id) $ - Updatable) - else Updatable - -- For now we forbid SingleEntry CAFs; they tickle the - -- ASSERT in rts/Storage.c line 215 at newCAF() re mut_link, - -- and I don't understand why. There's only one SE_CAF (well, - -- only one that tickled a great gaping bug in an earlier attempt - -- at ClosureInfo.getEntryConvention) in the whole of nofib, - -- specifically Main.lvl6 in spectral/cryptarithm2. - -- So no great loss. KSW 2000-07. --} +coreToMkStgRhs :: HasDebugCallStack => Id -> CoreExpr -> CtsM MkStgRhs +coreToMkStgRhs bndr expr = do + let (args, body) = myCollectBinders expr + let args' = filterStgBinders args + extendVarEnvCts [ (a, LambdaBound) | a <- args' ] $ do + body' <- coreToStgExpr body + let mk_rhs = MkStgRhs + { rhs_args = args' + , rhs_expr = body' + , rhs_type = exprType body + , rhs_is_join = isJoinId bndr + } + pure mk_rhs -- --------------------------------------------------------------------------- -- A monad for the core-to-STG pass @@ -933,15 +791,6 @@ lookupBinding env v = case lookupVarEnv env v of Just xx -> xx Nothing -> assertPpr (isGlobalId v) (ppr v) ImportBound -getAllCAFsCC :: Module -> (CostCentre, CostCentreStack) -getAllCAFsCC this_mod = - let - span = mkGeneralSrcSpan (mkFastString "") -- XXX do better - all_cafs_cc = mkAllCafsCC this_mod span - all_cafs_ccs = mkSingletonCCS all_cafs_cc - in - (all_cafs_cc, all_cafs_ccs) - -- Misc. filterStgBinders :: [Var] -> [Var] ===================================== compiler/GHC/Driver/Config/Stg/Pipeline.hs ===================================== @@ -7,6 +7,7 @@ import GHC.Prelude import Control.Monad (guard) import GHC.Stg.Pipeline +import GHC.Stg.Utils import GHC.Driver.Config.Diagnostic import GHC.Driver.Config.Stg.Lift @@ -15,15 +16,19 @@ import GHC.Driver.DynFlags -- | Initialize STG pretty-printing options from DynFlags initStgPipelineOpts :: DynFlags -> Bool -> StgPipelineOpts -initStgPipelineOpts dflags for_bytecode = StgPipelineOpts - { stgPipeline_lint = do - guard $ gopt Opt_DoStgLinting dflags - Just $ initDiagOpts dflags - , stgPipeline_pprOpts = initStgPprOpts dflags - , stgPipeline_phases = getStgToDo for_bytecode dflags - , stgPlatform = targetPlatform dflags - , stgPipeline_forBytecode = for_bytecode - } +initStgPipelineOpts dflags for_bytecode = + let !platform = targetPlatform dflags + !ext_dyn_refs = gopt Opt_ExternalDynamicRefs dflags + in StgPipelineOpts + { stgPipeline_lint = do + guard $ gopt Opt_DoStgLinting dflags + Just $ initDiagOpts dflags + , stgPipeline_pprOpts = initStgPprOpts dflags + , stgPipeline_phases = getStgToDo for_bytecode dflags + , stgPlatform = platform + , stgPipeline_forBytecode = for_bytecode + , stgPipeline_allowTopLevelConApp = allowTopLevelConApp platform ext_dyn_refs + } -- | Which Stg-to-Stg passes to run. Depends on flags, ways etc. getStgToDo ===================================== compiler/GHC/Stg/Make.hs ===================================== @@ -0,0 +1,172 @@ +module GHC.Stg.Make + ( MkStgRhs (..) + , mkTopStgRhs + , mkStgRhs + , mkStgRhsCon_maybe + , mkTopStgRhsCon_maybe + ) +where + +import GHC.Prelude +import GHC.Unit.Module + +import GHC.Core.DataCon +import GHC.Core.Type (Type) + +import GHC.Stg.Syntax +import GHC.Stg.Utils (stripStgTicksTop) + +import GHC.Types.Id +import GHC.Types.Name +import GHC.Types.CostCentre +import GHC.Types.Demand ( isAtMostOnceDmd ) +import GHC.Types.Tickish + +-- Represents the RHS of a binding for use with mk(Top)StgRhs and +-- mk(Top)StgRhsCon_maybe. +data MkStgRhs = MkStgRhs + { rhs_args :: [Id] -- ^ Empty for thunks + , rhs_expr :: StgExpr -- ^ RHS expression + , rhs_type :: Type -- ^ RHS type (only used in the JS backend: layering violation) + , rhs_is_join :: !Bool -- ^ Is it a RHS for a join-point? + } + + +-- Generate a top-level RHS. Any new cost centres generated for CAFs will be +-- appended to `CollectedCCs` argument. +mkTopStgRhs :: (Module -> DataCon -> [StgArg] -> Bool) + -> Bool -> Module -> CollectedCCs + -> Id -> MkStgRhs -> (StgRhs, CollectedCCs) +mkTopStgRhs allow_toplevel_con_app opt_AutoSccsOnIndividualCafs this_mod ccs bndr mk_rhs@(MkStgRhs bndrs rhs typ _) + -- try to make a StgRhsCon first + | Just rhs_con <- mkTopStgRhsCon_maybe (allow_toplevel_con_app this_mod) mk_rhs + = ( rhs_con, ccs ) + + | not (null bndrs) + = -- The list of arguments is non-empty, so not CAF + ( StgRhsClosure noExtFieldSilent + dontCareCCS + ReEntrant + bndrs rhs typ + , ccs ) + + -- Otherwise it's a CAF, see Note [Cost-centre initialization plan]. + | opt_AutoSccsOnIndividualCafs + = ( StgRhsClosure noExtFieldSilent + caf_ccs + upd_flag [] rhs typ + , collectCC caf_cc caf_ccs ccs ) + + | otherwise + = ( StgRhsClosure noExtFieldSilent + all_cafs_ccs + upd_flag [] rhs typ + , ccs ) + + where + upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry + | otherwise = Updatable + + -- CAF cost centres generated for -fcaf-all + caf_cc = mkAutoCC bndr modl + caf_ccs = mkSingletonCCS caf_cc + -- careful: the binder might be :Main.main, + -- which doesn't belong to module mod_name. + -- bug #249, tests prof001, prof002 + modl | Just m <- nameModule_maybe (idName bndr) = m + | otherwise = this_mod + + -- default CAF cost centre + (_, all_cafs_ccs) = getAllCAFsCC this_mod + +-- Generate a non-top-level RHS. Cost-centre is always currentCCS, +-- see Note [Cost-centre initialization plan]. +mkStgRhs :: Id -> MkStgRhs -> StgRhs +mkStgRhs bndr mk_rhs@(MkStgRhs bndrs rhs typ is_join) + -- try to make a StgRhsCon first + | Just rhs_con <- mkStgRhsCon_maybe mk_rhs + = rhs_con + + | otherwise + = StgRhsClosure noExtFieldSilent + currentCCS + upd_flag bndrs rhs typ + where + upd_flag | is_join = JumpedTo + | not (null bndrs) = ReEntrant + | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry + | otherwise = Updatable + + {- + SDM: disabled. Eval/Apply can't handle functions with arity zero very + well; and making these into simple non-updatable thunks breaks other + assumptions (namely that they will be entered only once). + + upd_flag | isPAP env rhs = ReEntrant + | otherwise = Updatable + +-- Detect thunks which will reduce immediately to PAPs, and make them +-- non-updatable. This has several advantages: +-- +-- - the non-updatable thunk behaves exactly like the PAP, +-- +-- - the thunk is more efficient to enter, because it is +-- specialised to the task. +-- +-- - we save one update frame, one stg_update_PAP, one update +-- and lots of PAP_enters. +-- +-- - in the case where the thunk is top-level, we save building +-- a black hole and furthermore the thunk isn't considered to +-- be a CAF any more, so it doesn't appear in any SRTs. +-- +-- We do it here, because the arity information is accurate, and we need +-- to do it before the SRT pass to save the SRT entries associated with +-- any top-level PAPs. + +isPAP env (StgApp f args) = listLengthCmp args arity == LT -- idArity f > length args + where + arity = stgArity f (lookupBinding env f) +isPAP env _ = False + +-} + +{- ToDo: + upd = if isOnceDem dem + then (if isNotTop toplev + then SingleEntry -- HA! Paydirt for "dem" + else + (if debugIsOn then trace "WARNING: SE CAFs unsupported, forcing UPD instead" else id) $ + Updatable) + else Updatable + -- For now we forbid SingleEntry CAFs; they tickle the + -- ASSERT in rts/Storage.c line 215 at newCAF() re mut_link, + -- and I don't understand why. There's only one SE_CAF (well, + -- only one that tickled a great gaping bug in an earlier attempt + -- at ClosureInfo.getEntryConvention) in the whole of nofib, + -- specifically Main.lvl6 in spectral/cryptarithm2. + -- So no great loss. KSW 2000-07. +-} + + +-- | Try to make a non top-level StgRhsCon if appropriate +mkStgRhsCon_maybe :: MkStgRhs -> Maybe StgRhs +mkStgRhsCon_maybe (MkStgRhs bndrs rhs typ is_join) + | [] <- bndrs + , not is_join + , (ticks, StgConApp con mn args _) <- stripStgTicksTop (not . tickishIsCode) rhs + = Just (StgRhsCon currentCCS con mn ticks args typ) + + | otherwise = Nothing + + +-- | Try to make a top-level StgRhsCon if appropriate +mkTopStgRhsCon_maybe :: (DataCon -> [StgArg] -> Bool) -> MkStgRhs -> Maybe StgRhs +mkTopStgRhsCon_maybe allow_static_con_app (MkStgRhs bndrs rhs typ is_join) + | [] <- bndrs + , not is_join -- shouldn't happen at top-level + , (ticks, StgConApp con mn args _) <- stripStgTicksTop (not . tickishIsCode) rhs + , allow_static_con_app con args + = Just (StgRhsCon dontCareCCS con mn ticks args typ) + + | otherwise = Nothing ===================================== compiler/GHC/Stg/Pipeline.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Stg.CSE ( stgCse ) import GHC.Stg.Lift ( StgLiftConfig, stgLiftLams ) import GHC.Unit.Module ( Module ) +import GHC.Core.DataCon (DataCon) + import GHC.Utils.Error import GHC.Types.Var import GHC.Types.Var.Set @@ -52,6 +54,12 @@ data StgPipelineOpts = StgPipelineOpts , stgPipeline_pprOpts :: !StgPprOpts , stgPlatform :: !Platform , stgPipeline_forBytecode :: !Bool + + , stgPipeline_allowTopLevelConApp :: Module -> DataCon -> [StgArg] -> Bool + -- ^ Is a top-level (static) StgConApp allowed or not. If not, use dynamic allocation. + -- + -- This is typically used to support dynamic linking on Windows and the + -- -fexternal-dynamic-refs flag. See GHC.Stg.Utils.allowTopLevelConApp. } newtype StgM a = StgM { _unStgM :: ReaderT Char IO a } @@ -136,7 +144,7 @@ stg2stg logger extra_vars opts this_mod binds StgUnarise -> do us <- getUniqueSupplyM liftIO (stg_linter False "Pre-unarise" binds) - let binds' = {-# SCC "StgUnarise" #-} unarise us binds + let binds' = {-# SCC "StgUnarise" #-} unarise us (stgPipeline_allowTopLevelConApp opts this_mod) binds liftIO (dump_when Opt_D_dump_stg_unarised "Unarised STG:" binds') liftIO (stg_linter True "Unarise" binds') return binds' ===================================== compiler/GHC/Stg/Stats.hs ===================================== @@ -46,6 +46,7 @@ data CounterType | ReEntrantBinds Bool{-ditto-} | SingleEntryBinds Bool{-ditto-} | UpdatableBinds Bool{-ditto-} + | JoinPointBinds Bool{-ditto-} deriving (Eq, Ord) type Count = Int @@ -94,6 +95,7 @@ showStgStats prog s (ReEntrantBinds _) = "ReEntrantBindsBinds_Nested " s (SingleEntryBinds _) = "SingleEntryBinds_Nested " s (UpdatableBinds _) = "UpdatableBinds_Nested " + s (JoinPointBinds _) = "JoinPointBinds_Nested " gatherStgStats :: [StgTopBinding] -> StatEnv gatherStgStats binds = combineSEs (map statTopBinding binds) @@ -132,6 +134,7 @@ statRhs top (_, StgRhsClosure _ _ u _ body _) ReEntrant -> ReEntrantBinds top Updatable -> UpdatableBinds top SingleEntry -> SingleEntryBinds top + JumpedTo -> JoinPointBinds top ) {- ===================================== compiler/GHC/Stg/Syntax.hs ===================================== @@ -54,7 +54,6 @@ module GHC.Stg.Syntax ( -- utils stgRhsArity, freeVarsOfRhs, - isDllConApp, stgArgType, stgArgRep, stgArgRep1, @@ -87,17 +86,14 @@ import GHC.Core.Ppr( {- instances -} ) import GHC.Types.ForeignCall ( ForeignCall ) import GHC.Types.Id -import GHC.Types.Name ( isDynLinkName ) import GHC.Types.Tickish ( StgTickish ) import GHC.Types.Var.Set import GHC.Types.Literal ( Literal, literalType ) import GHC.Types.RepType ( typePrimRep, typePrimRep1, typePrimRepU, typePrimRep_maybe ) -import GHC.Unit.Module ( Module ) import GHC.Utils.Outputable import GHC.Utils.Panic.Plain -import GHC.Platform import GHC.Builtin.PrimOps ( PrimOp, PrimCall ) import Data.ByteString ( ByteString ) @@ -138,51 +134,6 @@ data StgArg = StgVarArg Id | StgLitArg Literal --- | Does this constructor application refer to anything in a different --- *Windows* DLL? --- If so, we can't allocate it statically -isDllConApp - :: Platform - -> Bool -- is Opt_ExternalDynamicRefs enabled? - -> Module - -> DataCon - -> [StgArg] - -> Bool -isDllConApp platform ext_dyn_refs this_mod con args - | not ext_dyn_refs = False - | platformOS platform == OSMinGW32 - = isDynLinkName platform this_mod (dataConName con) || any is_dll_arg args - | otherwise = False - where - -- NB: typePrimRep1 is legit because any free variables won't have - -- unlifted type (there are no unlifted things at top level) - is_dll_arg :: StgArg -> Bool - is_dll_arg (StgVarArg v) = isAddrRep (typePrimRep1 (idType v)) - && isDynLinkName platform this_mod (idName v) - is_dll_arg _ = False - --- True of machine addresses; these are the things that don't work across DLLs. --- The key point here is that VoidRep comes out False, so that a top level --- nullary GADT constructor is False for isDllConApp --- --- data T a where --- T1 :: T Int --- --- gives --- --- T1 :: forall a. (a~Int) -> T a --- --- and hence the top-level binding --- --- $WT1 :: T Int --- $WT1 = T1 Int (Coercion (Refl Int)) --- --- The coercion argument here gets VoidRep -isAddrRep :: PrimOrVoidRep -> Bool -isAddrRep (NVRep AddrRep) = True -isAddrRep (NVRep (BoxedRep _)) = True -- FIXME: not true for JavaScript -isAddrRep _ = False - -- | Type of an @StgArg@ -- -- Very half baked because we have lost the type arguments. @@ -721,24 +672,35 @@ UpdateFlag This is also used in @LambdaFormInfo@ in the @ClosureInfo@ module. -A @ReEntrant@ closure may be entered multiple times, but should not be updated -or blackholed. An @Updatable@ closure should be updated after evaluation (and -may be blackholed during evaluation). A @SingleEntry@ closure will only be -entered once, and so need not be updated but may safely be blackholed. -} -data UpdateFlag = ReEntrant | Updatable | SingleEntry +data UpdateFlag + = ReEntrant + -- ^ A @ReEntrant@ closure may be entered multiple times, but should not + -- be updated or blackholed. + | Updatable + -- ^ An @Updatable@ closure should be updated after evaluation (and may be + -- blackholed during evaluation). + | SingleEntry + -- ^ A @SingleEntry@ closure will only be entered once, and so need not be + -- updated but may safely be blackholed. + | JumpedTo + -- ^ A @JumpedTo@ (join-point) closure is entered once or multiple times + -- but has no heap-allocated associated closure. + deriving (Show,Eq) instance Outputable UpdateFlag where ppr u = char $ case u of ReEntrant -> 'r' Updatable -> 'u' SingleEntry -> 's' + JumpedTo -> 'j' isUpdatable :: UpdateFlag -> Bool isUpdatable ReEntrant = False isUpdatable SingleEntry = False isUpdatable Updatable = True +isUpdatable JumpedTo = False {- ************************************************************************ ===================================== compiler/GHC/Stg/Unarise.hs ===================================== @@ -1,6 +1,7 @@ {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE TupleSections #-} +{-# LANGUAGE MultiWayIf #-} {- (c) The GRASP/AQUA Project, Glasgow University, 1992-2012 @@ -401,6 +402,7 @@ import GHC.Utils.Panic import GHC.Types.RepType import GHC.Stg.Syntax import GHC.Stg.Utils +import GHC.Stg.Make import GHC.Core.Type import GHC.Builtin.Types.Prim (intPrimTy) import GHC.Builtin.Types @@ -442,10 +444,14 @@ import Data.List (mapAccumL) -- INVARIANT: OutStgArgs in the range only have NvUnaryTypes -- (i.e. no unboxed tuples, sums or voids) -- -newtype UnariseEnv = UnariseEnv { ue_rho :: (VarEnv UnariseVal) } +data UnariseEnv = UnariseEnv + { ue_rho :: (VarEnv UnariseVal) + , ue_allow_static_conapp :: DataCon -> [StgArg] -> Bool + } -initUnariseEnv :: VarEnv UnariseVal -> UnariseEnv +initUnariseEnv :: VarEnv UnariseVal -> (DataCon -> [StgArg] -> Bool) -> UnariseEnv initUnariseEnv = UnariseEnv + data UnariseVal = MultiVal [OutStgArg] -- MultiVal to tuple. Can be empty list (void). | UnaryVal OutStgArg -- See Note [Renaming during unarisation]. @@ -477,27 +483,57 @@ lookupRho env v = lookupVarEnv (ue_rho env) v -------------------------------------------------------------------------------- -unarise :: UniqSupply -> [StgTopBinding] -> [StgTopBinding] -unarise us binds = initUs_ us (mapM (unariseTopBinding (initUnariseEnv emptyVarEnv)) binds) +unarise :: UniqSupply -> (DataCon -> [StgArg] -> Bool) -> [StgTopBinding] -> [StgTopBinding] +unarise us is_dll_con_app binds = initUs_ us (mapM (unariseTopBinding (initUnariseEnv emptyVarEnv is_dll_con_app)) binds) unariseTopBinding :: UnariseEnv -> StgTopBinding -> UniqSM StgTopBinding unariseTopBinding rho (StgTopLifted bind) - = StgTopLifted <$> unariseBinding rho bind + = StgTopLifted <$> unariseBinding rho True bind unariseTopBinding _ bind at StgTopStringLit{} = return bind -unariseBinding :: UnariseEnv -> StgBinding -> UniqSM StgBinding -unariseBinding rho (StgNonRec x rhs) - = StgNonRec x <$> unariseRhs rho rhs -unariseBinding rho (StgRec xrhss) - = StgRec <$> mapM (\(x, rhs) -> (x,) <$> unariseRhs rho rhs) xrhss +unariseBinding :: UnariseEnv -> Bool -> StgBinding -> UniqSM StgBinding +unariseBinding rho top_level (StgNonRec x rhs) + = StgNonRec x <$> unariseRhs rho top_level rhs +unariseBinding rho top_level (StgRec xrhss) + = StgRec <$> mapM (\(x, rhs) -> (x,) <$> unariseRhs rho top_level rhs) xrhss -unariseRhs :: UnariseEnv -> StgRhs -> UniqSM StgRhs -unariseRhs rho (StgRhsClosure ext ccs update_flag args expr typ) +unariseRhs :: UnariseEnv -> Bool -> StgRhs -> UniqSM StgRhs +unariseRhs rho top_level (StgRhsClosure ext ccs update_flag args expr typ) = do (rho', args1) <- unariseFunArgBinders rho args expr' <- unariseExpr rho' expr - return (StgRhsClosure ext ccs update_flag args1 expr' typ) - -unariseRhs rho (StgRhsCon ccs con mu ts args typ) + -- Unarisation can lead to a StgRhsClosure becoming a StgRhsCon. + -- Hence, we call `mk(Top)StgRhsCon_maybe` rather than just building + -- another `StgRhsClosure`. + -- + -- For example with unboxed sums (#25166): + -- + -- foo = \u [] case (# | _ | #) [(##)] of tag { __DEFAULT -> D [True tag] } + -- + -- ====> {unarisation} + -- + -- foo = D [True 2#] + -- + -- Transforming an appropriate StgRhsClosure into a StgRhsCon is + -- important as top-level StgRhsCon are statically allocated. + -- + let mk_rhs = MkStgRhs + { rhs_args = args1 + , rhs_expr = expr' + , rhs_type = typ + , rhs_is_join = update_flag == JumpedTo + } + if | top_level + , Just rhs_con <- mkTopStgRhsCon_maybe (ue_allow_static_conapp rho) mk_rhs + -> pure rhs_con + + | not top_level + , Just rhs_con <- mkStgRhsCon_maybe mk_rhs + -> pure rhs_con + + | otherwise + -> pure (StgRhsClosure ext ccs update_flag args1 expr' typ) + +unariseRhs rho _top (StgRhsCon ccs con mu ts args typ) = assert (not (isUnboxedTupleDataCon con || isUnboxedSumDataCon con)) return (StgRhsCon ccs con mu ts (unariseConArgs rho args) typ) @@ -576,10 +612,10 @@ unariseExpr rho (StgCase scrut bndr alt_ty alts) -- dead after unarise (checked in GHC.Stg.Lint) unariseExpr rho (StgLet ext bind e) - = StgLet ext <$> unariseBinding rho bind <*> unariseExpr rho e + = StgLet ext <$> unariseBinding rho False bind <*> unariseExpr rho e unariseExpr rho (StgLetNoEscape ext bind e) - = StgLetNoEscape ext <$> unariseBinding rho bind <*> unariseExpr rho e + = StgLetNoEscape ext <$> unariseBinding rho False bind <*> unariseExpr rho e unariseExpr rho (StgTick tick e) = StgTick tick <$> unariseExpr rho e ===================================== compiler/GHC/Stg/Utils.hs ===================================== @@ -9,9 +9,12 @@ module GHC.Stg.Utils , idArgs , mkUnarisedId, mkUnarisedIds + + , allowTopLevelConApp ) where import GHC.Prelude +import GHC.Platform import GHC.Types.Id import GHC.Core.Type @@ -22,6 +25,8 @@ import GHC.Types.Tickish import GHC.Types.Unique.Supply import GHC.Types.RepType +import GHC.Types.Name ( isDynLinkName ) +import GHC.Unit.Module ( Module ) import GHC.Stg.Syntax import GHC.Utils.Outputable @@ -122,3 +127,54 @@ stripStgTicksTopE :: (StgTickish -> Bool) -> GenStgExpr p -> GenStgExpr p stripStgTicksTopE p = go where go (StgTick t e) | p t = go e go other = other + +-- | Do we allow the given top-level (static) ConApp? +allowTopLevelConApp + :: Platform + -> Bool -- is Opt_ExternalDynamicRefs enabled? + -> Module + -> DataCon + -> [StgArg] + -> Bool +allowTopLevelConApp platform ext_dyn_refs this_mod con args + -- we're not using dynamic linking + | not ext_dyn_refs = True + -- if the target OS is Windows, we only allow top-level ConApps if they don't + -- reference external names (Windows DLLs have a problem with static cross-DLL + -- refs) + | platformOS platform == OSMinGW32 = not is_external_con_app + -- otherwise, allowed + -- Sylvain: shouldn't this be False when (ext_dyn_refs && is_external_con_app)? + | otherwise = True + where + is_external_con_app = isDynLinkName platform this_mod (dataConName con) || any is_dll_arg args + + -- NB: typePrimRep1 is legit because any free variables won't have + -- unlifted type (there are no unlifted things at top level) + is_dll_arg :: StgArg -> Bool + is_dll_arg (StgVarArg v) = isAddrRep (typePrimRep1 (idType v)) + && isDynLinkName platform this_mod (idName v) + is_dll_arg _ = False + +-- True of machine addresses; these are the things that don't work across DLLs. +-- The key point here is that VoidRep comes out False, so that a top level +-- nullary GADT constructor is True for allowTopLevelConApp +-- +-- data T a where +-- T1 :: T Int +-- +-- gives +-- +-- T1 :: forall a. (a~Int) -> T a +-- +-- and hence the top-level binding +-- +-- $WT1 :: T Int +-- $WT1 = T1 Int (Coercion (Refl Int)) +-- +-- The coercion argument here gets VoidRep +isAddrRep :: PrimOrVoidRep -> Bool +isAddrRep (NVRep AddrRep) = True +isAddrRep (NVRep (BoxedRep _)) = True -- FIXME: not true for JavaScript +isAddrRep _ = False + ===================================== compiler/GHC/StgToCmm/DataCon.hs ===================================== @@ -19,6 +19,7 @@ import GHC.Prelude import GHC.Platform +import GHC.Stg.Utils (allowTopLevelConApp) import GHC.Stg.Syntax import GHC.Core ( AltCon(..) ) @@ -48,7 +49,6 @@ import GHC.Utils.Panic import GHC.Utils.Misc import GHC.Utils.Monad (mapMaybeM) -import Control.Monad import Data.Char import GHC.StgToCmm.Config (stgToCmmPlatform) import GHC.StgToCmm.TagCheck (checkConArgsStatic, checkConArgsDyn) @@ -90,10 +90,8 @@ cgTopRhsCon cfg id con mn args gen_code = do { profile <- getProfile ; this_mod <- getModuleName - ; when (platformOS platform == OSMinGW32) $ - -- Windows DLLs have a problem with static cross-DLL refs. - massert (not (isDllConApp platform (stgToCmmExtDynRefs cfg) this_mod con (map fromNonVoid args))) - ; assert (args `lengthIs` countConRepArgs con ) return () + ; massert (allowTopLevelConApp platform (stgToCmmExtDynRefs cfg) this_mod con (map fromNonVoid args)) + ; massert (args `lengthIs` countConRepArgs con ) ; checkConArgsStatic (text "TagCheck failed - Top level con") con (map fromNonVoid args) -- LAY IT OUT ; let ===================================== compiler/GHC/Types/CostCentre.hs ===================================== @@ -4,6 +4,7 @@ module GHC.Types.CostCentre ( CostCentre(..), CcName, CCFlavour, mkCafFlavour, mkExprCCFlavour, mkDeclCCFlavour, mkHpcCCFlavour, mkLateCCFlavour, mkCallerCCFlavour, + getAllCAFsCC, pprCostCentre, CostCentreStack, @@ -393,3 +394,13 @@ instance Binary CostCentre where -- ok, because we only need the SrcSpan when declaring the -- CostCentre in the original module, it is not used by importing -- modules. + +getAllCAFsCC :: Module -> (CostCentre, CostCentreStack) +getAllCAFsCC this_mod = + let + span = mkGeneralSrcSpan (mkFastString "") -- XXX do better + all_cafs_cc = mkAllCafsCC this_mod span + all_cafs_ccs = mkSingletonCCS all_cafs_cc + in + (all_cafs_cc, all_cafs_ccs) + ===================================== compiler/ghc.cabal.in ===================================== @@ -702,6 +702,7 @@ Library GHC.Stg.InferTags.Rewrite GHC.Stg.InferTags.TagSig GHC.Stg.InferTags.Types + GHC.Stg.Make GHC.Stg.Pipeline GHC.Stg.Stats GHC.Stg.Subst ===================================== testsuite/tests/codeGen/should_compile/Makefile ===================================== @@ -77,3 +77,6 @@ T17648: -fcatch-nonexhaustive-cases T17648.hs -v0 -fforce-recomp '$(TEST_HC)' --show-iface T17648.hi | tr -d '\n\r' | \ grep -F 'f :: T GHC.Types.Int -> () [TagSig' >/dev/null + +T25166: + '$(TEST_HC)' -O2 -dno-typeable-binds -ddump-cmm T25166.hs | awk '/foo_closure/{flag=1}/}]/{flag=0}flag' ===================================== testsuite/tests/codeGen/should_compile/T25166.hs ===================================== @@ -0,0 +1,7 @@ +module Test where + +data A = A | B | C + +data D = D !Bool {-# UNPACK #-} !A + +foo = D True B ===================================== testsuite/tests/codeGen/should_compile/T25166.stdout ===================================== @@ -0,0 +1,6 @@ +[section ""data" . Test.foo_closure" { + Test.foo_closure: + const Test.D_con_info; + const GHC.Types.True_closure+2; + const 2; + const 3; ===================================== testsuite/tests/codeGen/should_compile/all.T ===================================== @@ -139,5 +139,7 @@ test('callee-no-local', [ ['-ddump-cmm-raw'] ) +test('T25166', [req_cmm], makefile_test, []) + # dump Core to ensure that d is defined as: d = D 10## RUBBISH(IntRep) test('T25177', normal, compile, ['-O2 -dno-typeable-binds -ddump-simpl -dsuppress-all -dsuppress-uniques -v0']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5456e02ee97ea2f8b156bdbadd982611274bef0d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5456e02ee97ea2f8b156bdbadd982611274bef0d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 20:41:23 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 06 Sep 2024 16:41:23 -0400 Subject: [Git][ghc/ghc][master] haddock: Add missing requirements.txt for the online manual Message-ID: <66db68f3b2dcc_19b5b61a000435551@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 1 changed file: - + utils/haddock/doc/requirements.txt Changes: ===================================== utils/haddock/doc/requirements.txt ===================================== @@ -0,0 +1 @@ +sphinxcontrib-applehelp ==2.0.0 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/958b45186742ec7266d8bc30e338bea35c872bea -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/958b45186742ec7266d8bc30e338bea35c872bea You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 6 21:28:14 2024 From: gitlab at gitlab.haskell.org (Matthew Craven (@clyring)) Date: Fri, 06 Sep 2024 17:28:14 -0400 Subject: [Git][ghc/ghc][wip/T23490-part2] 17 commits: testsuite: Add support to capture performance metrics via 'perf' Message-ID: <66db73ee8b035_19b5b67736d43886b@gitlab.mail> Matthew Craven pushed to branch wip/T23490-part2 at Glasgow Haskell Compiler / GHC Commits: a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - fe7bbdae by Matthew Craven at 2024-09-06T17:20:44-04:00 Bump transformers submodule The svg image files mentioned in transformers.cabal were previously not checked in, which broke sdist generation. - - - - - 60c9e5e7 by Matthew Craven at 2024-09-06T17:20:44-04:00 Remove reference to non-existent file in haddock.cabal - - - - - 955f0e9c by Matthew Craven at 2024-09-06T17:20:45-04:00 Update hackage index state - - - - - c641d336 by Matthew Craven at 2024-09-06T17:20:45-04:00 Move tests T11462 and T11525 into tests/tcplugins - - - - - 03bbabb8 by Matthew Craven at 2024-09-06T17:21:02-04:00 Repair the 'build-cabal' hadrian target Fixes #23117. Fixes #23281. Fixes #23490. This required: * Updating the bit-rotted compiler/Setup.hs and its setup-depends * Listing a few recently-added libraries and utilities in cabal.project-reinstall * Setting allow-boot-library-installs to 'True' since Cabal now considers the 'ghc' package itself a boot library for the purposes of this flag Additionally, the allow-newer block in cabal.project-reinstall was removed. This block was probably added because when the libraries/Cabal submodule is too new relative to the cabal-install executable, solving the setup-depends for any package with a custom setup requires building an old Cabal (from Hackage) against the in-tree version of base, and this can fail un-necessarily due to tight version bounds on base. However, the blind allow-newer can also cause the solver to go berserk and choose a stupid build plan that has no business succeeding, and the failures when this happens are dreadfully confusing. (See #23281 and #24363.) Why does setup-depends solving insist on an old version of Cabal? See: https://github.com/haskell/cabal/blob/0a0b33983b0f022b9697f7df3a69358ee9061a89/cabal-install/src/Distribution/Client/ProjectPlanning.hs#L1393-L1410 The right solution here is probably to use the in-tree cabal-install from libraries/Cabal/cabal-install with the build-cabal target rather than whatever the environment happens to provide. But this is left for future work. - - - - - 25d52bf9 by Matthew Craven at 2024-09-06T17:26:22-04:00 Revert "CI: Disable the test-cabal-reinstall job" This reverts commit 38c3afb64d3ffc42f12163c6f0f0d5c414aa8255. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - cabal.project-reinstall - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs - compiler/GHC/Stg/Unarise.hs - compiler/GHC/Stg/Utils.hs - compiler/GHC/StgToCmm/DataCon.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Utils/Env.hs - compiler/GHC/Types/CostCentre.hs - compiler/GHC/Utils/TmpFs.hs - compiler/Setup.hs - compiler/ghc.cabal.in - hadrian/cabal.project - hadrian/ghci-multi-cabal.in - hadrian/src/Rules/ToolArgs.hs - libraries/transformers - testsuite/driver/perf_notes.py - testsuite/driver/testglobals.py - testsuite/driver/testlib.py The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a547477e49eb032a0d653c3c88f51fc3eca8171a...25d52bf9c59431642f27a7a9e49e65ea013c4cfb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a547477e49eb032a0d653c3c88f51fc3eca8171a...25d52bf9c59431642f27a7a9e49e65ea013c4cfb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 7 00:20:20 2024 From: gitlab at gitlab.haskell.org (Matthew Craven (@clyring)) Date: Fri, 06 Sep 2024 20:20:20 -0400 Subject: [Git][ghc/ghc][wip/T22935] Fix test outputs Message-ID: <66db9c43f015b_d9d6c402a7c1712b@gitlab.mail> Matthew Craven pushed to branch wip/T22935 at Glasgow Haskell Compiler / GHC Commits: 62ab854c by Matthew Craven at 2024-09-06T20:19:25-04:00 Fix test outputs - - - - - 4 changed files: - + libraries/base/tests/T23233.stderr - libraries/base/tests/T23233.stdout - testsuite/tests/llvm/should_compile/T5681.hs - testsuite/tests/simplCore/should_compile/par01.stderr Changes: ===================================== libraries/base/tests/T23233.stderr ===================================== @@ -0,0 +1,2 @@ +eval p +eval q ===================================== libraries/base/tests/T23233.stdout ===================================== @@ -1,2 +1 @@ -eval p -eval q +SP 3 144 ===================================== testsuite/tests/llvm/should_compile/T5681.hs ===================================== @@ -1,6 +1,10 @@ {-# LANGUAGE Haskell2010 #-} {-# LANGUAGE MagicHash, UnboxedTuples #-} -- Test case for #5681 + +{-# OPTIONS_GHC -Wno-deprecations #-} + -- The use of the (now-deprecated) primop par# is an + -- essential part of the issue that this is meant to test. module Main where import GHC.Exts ===================================== testsuite/tests/simplCore/should_compile/par01.stderr ===================================== @@ -1,16 +1,23 @@ ==================== CorePrep ==================== Result size of CorePrep - = {terms: 22, types: 10, coercions: 0, joins: 0/0} + = {terms: 31, types: 34, coercions: 0, joins: 0/0} Rec { --- RHS size: {terms: 7, types: 3, coercions: 0, joins: 0/0} +-- RHS size: {terms: 16, types: 27, coercions: 0, joins: 0/0} Par01.depth [Occ=LoopBreaker] :: GHC.Types.Int -> GHC.Types.Int [GblId, Arity=1, Str=, Unf=OtherCon []] Par01.depth = \ (d :: GHC.Types.Int) -> - case GHC.Prim.par# @GHC.Types.Int d of { __DEFAULT -> - Par01.depth d + case GHC.Prim.spark# + @GHC.Types.Int @GHC.Prim.RealWorld d GHC.Prim.realWorld# + of + { (# ipv [Occ=Once1], _ [Occ=Dead] #) -> + case Par01.depth d of sat [Occ=Once1] { __DEFAULT -> + case (# ipv, sat #) of { (# _ [Occ=Dead], ipv3 [Occ=Once1] #) -> + ipv3 + } + } } end Rec } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/62ab854c931f1ef50c6607dd930babae5459a9de -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/62ab854c931f1ef50c6607dd930babae5459a9de You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 7 14:26:01 2024 From: gitlab at gitlab.haskell.org (Krzysztof Gogolewski (@monoidal)) Date: Sat, 07 Sep 2024 10:26:01 -0400 Subject: [Git][ghc/ghc][wip/forall-kind-rule] 26 commits: JS: add basic support for POSIX *at functions (#25190) Message-ID: <66dc627976b30_a45ea11f710179e2@gitlab.mail> Krzysztof Gogolewski pushed to branch wip/forall-kind-rule at Glasgow Haskell Compiler / GHC Commits: 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - c8fe9967 by Krzysztof Gogolewski at 2024-09-07T16:25:47+02:00 Limit forall to TYPE r and CONSTRAINT r Fixes #22063. - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs - compiler/GHC/Stg/Unarise.hs - compiler/GHC/Stg/Utils.hs - compiler/GHC/StgToCmm/DataCon.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/GHC/Tc/Utils/Env.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/CostCentre.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ecb4320dc60b56b0ba3ca295b94633289ab927ef...c8fe9967f16535d7515403ad43631ca23aa10913 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ecb4320dc60b56b0ba3ca295b94633289ab927ef...c8fe9967f16535d7515403ad43631ca23aa10913 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 7 17:58:59 2024 From: gitlab at gitlab.haskell.org (Krzysztof Gogolewski (@monoidal)) Date: Sat, 07 Sep 2024 13:58:59 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25243 Message-ID: <66dc9463938ac_3dea7355221065671@gitlab.mail> Krzysztof Gogolewski pushed new branch wip/T25243 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25243 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 08:57:49 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 08 Sep 2024 04:57:49 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] 12 commits: Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Message-ID: <66dd670d74e9e_99646363fbc805e@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - af46239c by Sven Tennie at 2024-09-08T08:57:47+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 9228d765 by Sven Tennie at 2024-09-08T08:57:47+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - ba702986 by Sven Tennie at 2024-09-08T08:57:47+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 26e41e7b by Sven Tennie at 2024-09-08T08:57:47+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - a41e9cdd by Sven Tennie at 2024-09-08T08:57:47+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - fb4ee42b by Sven Tennie at 2024-09-08T08:57:47+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 09bdfc8e by Sven Tennie at 2024-09-08T08:57:47+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 086806b2 by Sven Tennie at 2024-09-08T08:57:47+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - 606000eb by Sven Tennie at 2024-09-08T08:57:47+00:00 RISCV64: Set codeowners of the NCG - - - - - 63feb19f by Sven Tennie at 2024-09-08T08:57:47+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. - - - - - 10 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3149b4035c6b6422f28ee03a660e3b31cf85581f...63feb19ffef71c9dda93b4e8a1e2a6ccdba862cc -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3149b4035c6b6422f28ee03a660e3b31cf85581f...63feb19ffef71c9dda93b4e8a1e2a6ccdba862cc You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 09:58:32 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 08 Sep 2024 05:58:32 -0400 Subject: [Git][ghc/ghc][wip/supersven/AArch64_takeRegRegMoveInst] 3 commits: Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Message-ID: <66dd75482d61c_996465ec720875f@gitlab.mail> Sven Tennie pushed to branch wip/supersven/AArch64_takeRegRegMoveInst at Glasgow Haskell Compiler / GHC Commits: 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 17 changed files: - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs - compiler/GHC/Stg/Unarise.hs - compiler/GHC/Stg/Utils.hs - compiler/GHC/StgToCmm/DataCon.hs - compiler/GHC/Types/CostCentre.hs - compiler/ghc.cabal.in - testsuite/tests/codeGen/should_compile/Makefile - + testsuite/tests/codeGen/should_compile/T25166.hs - + testsuite/tests/codeGen/should_compile/T25166.stdout - testsuite/tests/codeGen/should_compile/all.T - + utils/haddock/doc/requirements.txt Changes: ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -30,6 +30,7 @@ import GHC.Utils.Panic import Data.Maybe (fromMaybe) import GHC.Stack +import GHC.Platform.Reg.Class -- | LR and FP (8 byte each) are the prologue of each stack frame stackFrameHeaderSize :: Int @@ -454,10 +455,22 @@ isMetaInstr instr mkRegRegMoveInstr :: Reg -> Reg -> Instr mkRegRegMoveInstr src dst = ANN (text "Reg->Reg Move: " <> ppr src <> text " -> " <> ppr dst) $ MOV (OpReg W64 dst) (OpReg W64 src) --- | Take the source and destination from this reg -> reg move instruction --- or Nothing if it's not one +-- | Take the source and destination registers from a move instruction of same +-- register class (`RegClass`). +-- +-- The idea is to identify moves that can be eliminated by the register +-- allocator: If the source register serves no special purpose, one could +-- continue using it; saving one move instruction. For this, the register kinds +-- (classes) must be the same (no conversion involved.) takeRegRegMoveInstr :: Instr -> Maybe (Reg,Reg) ---takeRegRegMoveInstr (MOV (OpReg fmt dst) (OpReg fmt' src)) | fmt == fmt' = Just (src, dst) +takeRegRegMoveInstr (MOV (OpReg _fmt dst) (OpReg _fmt' src)) + | classOfReg dst == classOfReg src = pure (src, dst) + where + classOfReg ::Reg -> RegClass + classOfReg reg + = case reg of + RegVirtual vr -> classOfVirtualReg vr + RegReal rr -> classOfRealReg rr takeRegRegMoveInstr _ = Nothing -- | Make an unconditional jump instruction. ===================================== compiler/GHC/CoreToStg.hs ===================================== @@ -27,7 +27,8 @@ import GHC.Core.DataCon import GHC.Stg.Syntax import GHC.Stg.Debug -import GHC.Stg.Utils +import GHC.Stg.Make +import GHC.Stg.Utils (allowTopLevelConApp) import GHC.Types.RepType import GHC.Types.Id.Make ( coercionTokenId ) @@ -36,16 +37,13 @@ import GHC.Types.Id.Info import GHC.Types.CostCentre import GHC.Types.Tickish import GHC.Types.Var.Env -import GHC.Types.Name ( isExternalName, nameModule_maybe ) +import GHC.Types.Name ( isExternalName ) import GHC.Types.Basic ( Arity, TypeOrConstraint(..) ) import GHC.Types.Literal import GHC.Types.ForeignCall import GHC.Types.IPE -import GHC.Types.Demand ( isAtMostOnceDmd ) -import GHC.Types.SrcLoc ( mkGeneralSrcSpan ) import GHC.Unit.Module -import GHC.Data.FastString import GHC.Platform ( Platform ) import GHC.Platform.Ways import GHC.Builtin.PrimOps @@ -338,10 +336,12 @@ coreToTopStgRhs -> CtsM (CollectedCCs, (Id, StgRhs)) coreToTopStgRhs opts this_mod ccs (bndr, rhs) - = do { new_rhs <- coreToPreStgRhs rhs + = do { new_rhs <- coreToMkStgRhs bndr rhs ; let (stg_rhs, ccs') = - mkTopStgRhs opts this_mod ccs bndr new_rhs + mkTopStgRhs (allowTopLevelConApp (coreToStg_platform opts) (coreToStg_ExternalDynamicRefs opts)) + (coreToStg_AutoSccsOnIndividualCafs opts) + this_mod ccs bndr new_rhs stg_arity = stgRhsArity stg_rhs @@ -372,7 +372,7 @@ coreToTopStgRhs opts this_mod ccs (bndr, rhs) -- coreToStgExpr panics if the input expression is a value lambda. CorePrep -- ensures that value lambdas only exist as the RHS of bindings, which we --- handle with the function coreToPreStgRhs. +-- handle with the function coreToMkStgRhs. coreToStgExpr :: HasDebugCallStack => CoreExpr @@ -685,166 +685,24 @@ coreToStgRhs :: (Id,CoreExpr) -> CtsM StgRhs coreToStgRhs (bndr, rhs) = do - new_rhs <- coreToPreStgRhs rhs + new_rhs <- coreToMkStgRhs bndr rhs return (mkStgRhs bndr new_rhs) --- Represents the RHS of a binding for use with mk(Top)StgRhs. -data PreStgRhs = PreStgRhs [Id] StgExpr Type -- The [Id] is empty for thunks - -- Convert the RHS of a binding from Core to STG. This is a wrapper around -- coreToStgExpr that can handle value lambdas. -coreToPreStgRhs :: HasDebugCallStack => CoreExpr -> CtsM PreStgRhs -coreToPreStgRhs expr - = extendVarEnvCts [ (a, LambdaBound) | a <- args' ] $ - do { body' <- coreToStgExpr body - ; return (PreStgRhs args' body' (exprType body)) } - where - (args, body) = myCollectBinders expr - args' = filterStgBinders args - --- Generate a top-level RHS. Any new cost centres generated for CAFs will be --- appended to `CollectedCCs` argument. -mkTopStgRhs :: CoreToStgOpts -> Module -> CollectedCCs - -> Id -> PreStgRhs -> (StgRhs, CollectedCCs) - -mkTopStgRhs CoreToStgOpts - { coreToStg_platform = platform - , coreToStg_ExternalDynamicRefs = opt_ExternalDynamicRefs - , coreToStg_AutoSccsOnIndividualCafs = opt_AutoSccsOnIndividualCafs - } this_mod ccs bndr (PreStgRhs bndrs rhs typ) - | not (null bndrs) - = -- The list of arguments is non-empty, so not CAF - ( StgRhsClosure noExtFieldSilent - dontCareCCS - ReEntrant - bndrs rhs typ - , ccs ) - - -- After this point we know that `bndrs` is empty, - -- so this is not a function binding - | StgConApp con mn args _ <- unticked_rhs - , -- Dynamic StgConApps are updatable - not (isDllConApp platform opt_ExternalDynamicRefs this_mod con args) - = -- CorePrep does this right, but just to make sure - assertPpr (not (isUnboxedTupleDataCon con || isUnboxedSumDataCon con)) - (ppr bndr $$ ppr con $$ ppr args) - ( StgRhsCon dontCareCCS con mn ticks args typ, ccs ) - - -- Otherwise it's a CAF, see Note [Cost-centre initialization plan]. - | opt_AutoSccsOnIndividualCafs - = ( StgRhsClosure noExtFieldSilent - caf_ccs - upd_flag [] rhs typ - , collectCC caf_cc caf_ccs ccs ) - - | otherwise - = ( StgRhsClosure noExtFieldSilent - all_cafs_ccs - upd_flag [] rhs typ - , ccs ) - - where - (ticks, unticked_rhs) = stripStgTicksTop (not . tickishIsCode) rhs - - upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry - | otherwise = Updatable - - -- CAF cost centres generated for -fcaf-all - caf_cc = mkAutoCC bndr modl - caf_ccs = mkSingletonCCS caf_cc - -- careful: the binder might be :Main.main, - -- which doesn't belong to module mod_name. - -- bug #249, tests prof001, prof002 - modl | Just m <- nameModule_maybe (idName bndr) = m - | otherwise = this_mod - - -- default CAF cost centre - (_, all_cafs_ccs) = getAllCAFsCC this_mod - --- Generate a non-top-level RHS. Cost-centre is always currentCCS, --- see Note [Cost-centre initialization plan]. -mkStgRhs :: Id -> PreStgRhs -> StgRhs -mkStgRhs bndr (PreStgRhs bndrs rhs typ) - | not (null bndrs) - = StgRhsClosure noExtFieldSilent - currentCCS - ReEntrant - bndrs rhs typ - - -- After this point we know that `bndrs` is empty, - -- so this is not a function binding - - | isJoinId bndr -- Must be a nullary join point - = -- It might have /type/ arguments (T18328), - -- so its JoinArity might be >0 - StgRhsClosure noExtFieldSilent - currentCCS - ReEntrant -- ignored for LNE - [] rhs typ - - | StgConApp con mn args _ <- unticked_rhs - = StgRhsCon currentCCS con mn ticks args typ - - | otherwise - = StgRhsClosure noExtFieldSilent - currentCCS - upd_flag [] rhs typ - where - (ticks, unticked_rhs) = stripStgTicksTop (not . tickishIsCode) rhs - - upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry - | otherwise = Updatable - - {- - SDM: disabled. Eval/Apply can't handle functions with arity zero very - well; and making these into simple non-updatable thunks breaks other - assumptions (namely that they will be entered only once). - - upd_flag | isPAP env rhs = ReEntrant - | otherwise = Updatable - --- Detect thunks which will reduce immediately to PAPs, and make them --- non-updatable. This has several advantages: --- --- - the non-updatable thunk behaves exactly like the PAP, --- --- - the thunk is more efficient to enter, because it is --- specialised to the task. --- --- - we save one update frame, one stg_update_PAP, one update --- and lots of PAP_enters. --- --- - in the case where the thunk is top-level, we save building --- a black hole and furthermore the thunk isn't considered to --- be a CAF any more, so it doesn't appear in any SRTs. --- --- We do it here, because the arity information is accurate, and we need --- to do it before the SRT pass to save the SRT entries associated with --- any top-level PAPs. - -isPAP env (StgApp f args) = listLengthCmp args arity == LT -- idArity f > length args - where - arity = stgArity f (lookupBinding env f) -isPAP env _ = False - --} - -{- ToDo: - upd = if isOnceDem dem - then (if isNotTop toplev - then SingleEntry -- HA! Paydirt for "dem" - else - (if debugIsOn then trace "WARNING: SE CAFs unsupported, forcing UPD instead" else id) $ - Updatable) - else Updatable - -- For now we forbid SingleEntry CAFs; they tickle the - -- ASSERT in rts/Storage.c line 215 at newCAF() re mut_link, - -- and I don't understand why. There's only one SE_CAF (well, - -- only one that tickled a great gaping bug in an earlier attempt - -- at ClosureInfo.getEntryConvention) in the whole of nofib, - -- specifically Main.lvl6 in spectral/cryptarithm2. - -- So no great loss. KSW 2000-07. --} +coreToMkStgRhs :: HasDebugCallStack => Id -> CoreExpr -> CtsM MkStgRhs +coreToMkStgRhs bndr expr = do + let (args, body) = myCollectBinders expr + let args' = filterStgBinders args + extendVarEnvCts [ (a, LambdaBound) | a <- args' ] $ do + body' <- coreToStgExpr body + let mk_rhs = MkStgRhs + { rhs_args = args' + , rhs_expr = body' + , rhs_type = exprType body + , rhs_is_join = isJoinId bndr + } + pure mk_rhs -- --------------------------------------------------------------------------- -- A monad for the core-to-STG pass @@ -933,15 +791,6 @@ lookupBinding env v = case lookupVarEnv env v of Just xx -> xx Nothing -> assertPpr (isGlobalId v) (ppr v) ImportBound -getAllCAFsCC :: Module -> (CostCentre, CostCentreStack) -getAllCAFsCC this_mod = - let - span = mkGeneralSrcSpan (mkFastString "") -- XXX do better - all_cafs_cc = mkAllCafsCC this_mod span - all_cafs_ccs = mkSingletonCCS all_cafs_cc - in - (all_cafs_cc, all_cafs_ccs) - -- Misc. filterStgBinders :: [Var] -> [Var] ===================================== compiler/GHC/Driver/Config/Stg/Pipeline.hs ===================================== @@ -7,6 +7,7 @@ import GHC.Prelude import Control.Monad (guard) import GHC.Stg.Pipeline +import GHC.Stg.Utils import GHC.Driver.Config.Diagnostic import GHC.Driver.Config.Stg.Lift @@ -15,15 +16,19 @@ import GHC.Driver.DynFlags -- | Initialize STG pretty-printing options from DynFlags initStgPipelineOpts :: DynFlags -> Bool -> StgPipelineOpts -initStgPipelineOpts dflags for_bytecode = StgPipelineOpts - { stgPipeline_lint = do - guard $ gopt Opt_DoStgLinting dflags - Just $ initDiagOpts dflags - , stgPipeline_pprOpts = initStgPprOpts dflags - , stgPipeline_phases = getStgToDo for_bytecode dflags - , stgPlatform = targetPlatform dflags - , stgPipeline_forBytecode = for_bytecode - } +initStgPipelineOpts dflags for_bytecode = + let !platform = targetPlatform dflags + !ext_dyn_refs = gopt Opt_ExternalDynamicRefs dflags + in StgPipelineOpts + { stgPipeline_lint = do + guard $ gopt Opt_DoStgLinting dflags + Just $ initDiagOpts dflags + , stgPipeline_pprOpts = initStgPprOpts dflags + , stgPipeline_phases = getStgToDo for_bytecode dflags + , stgPlatform = platform + , stgPipeline_forBytecode = for_bytecode + , stgPipeline_allowTopLevelConApp = allowTopLevelConApp platform ext_dyn_refs + } -- | Which Stg-to-Stg passes to run. Depends on flags, ways etc. getStgToDo ===================================== compiler/GHC/Stg/Make.hs ===================================== @@ -0,0 +1,172 @@ +module GHC.Stg.Make + ( MkStgRhs (..) + , mkTopStgRhs + , mkStgRhs + , mkStgRhsCon_maybe + , mkTopStgRhsCon_maybe + ) +where + +import GHC.Prelude +import GHC.Unit.Module + +import GHC.Core.DataCon +import GHC.Core.Type (Type) + +import GHC.Stg.Syntax +import GHC.Stg.Utils (stripStgTicksTop) + +import GHC.Types.Id +import GHC.Types.Name +import GHC.Types.CostCentre +import GHC.Types.Demand ( isAtMostOnceDmd ) +import GHC.Types.Tickish + +-- Represents the RHS of a binding for use with mk(Top)StgRhs and +-- mk(Top)StgRhsCon_maybe. +data MkStgRhs = MkStgRhs + { rhs_args :: [Id] -- ^ Empty for thunks + , rhs_expr :: StgExpr -- ^ RHS expression + , rhs_type :: Type -- ^ RHS type (only used in the JS backend: layering violation) + , rhs_is_join :: !Bool -- ^ Is it a RHS for a join-point? + } + + +-- Generate a top-level RHS. Any new cost centres generated for CAFs will be +-- appended to `CollectedCCs` argument. +mkTopStgRhs :: (Module -> DataCon -> [StgArg] -> Bool) + -> Bool -> Module -> CollectedCCs + -> Id -> MkStgRhs -> (StgRhs, CollectedCCs) +mkTopStgRhs allow_toplevel_con_app opt_AutoSccsOnIndividualCafs this_mod ccs bndr mk_rhs@(MkStgRhs bndrs rhs typ _) + -- try to make a StgRhsCon first + | Just rhs_con <- mkTopStgRhsCon_maybe (allow_toplevel_con_app this_mod) mk_rhs + = ( rhs_con, ccs ) + + | not (null bndrs) + = -- The list of arguments is non-empty, so not CAF + ( StgRhsClosure noExtFieldSilent + dontCareCCS + ReEntrant + bndrs rhs typ + , ccs ) + + -- Otherwise it's a CAF, see Note [Cost-centre initialization plan]. + | opt_AutoSccsOnIndividualCafs + = ( StgRhsClosure noExtFieldSilent + caf_ccs + upd_flag [] rhs typ + , collectCC caf_cc caf_ccs ccs ) + + | otherwise + = ( StgRhsClosure noExtFieldSilent + all_cafs_ccs + upd_flag [] rhs typ + , ccs ) + + where + upd_flag | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry + | otherwise = Updatable + + -- CAF cost centres generated for -fcaf-all + caf_cc = mkAutoCC bndr modl + caf_ccs = mkSingletonCCS caf_cc + -- careful: the binder might be :Main.main, + -- which doesn't belong to module mod_name. + -- bug #249, tests prof001, prof002 + modl | Just m <- nameModule_maybe (idName bndr) = m + | otherwise = this_mod + + -- default CAF cost centre + (_, all_cafs_ccs) = getAllCAFsCC this_mod + +-- Generate a non-top-level RHS. Cost-centre is always currentCCS, +-- see Note [Cost-centre initialization plan]. +mkStgRhs :: Id -> MkStgRhs -> StgRhs +mkStgRhs bndr mk_rhs@(MkStgRhs bndrs rhs typ is_join) + -- try to make a StgRhsCon first + | Just rhs_con <- mkStgRhsCon_maybe mk_rhs + = rhs_con + + | otherwise + = StgRhsClosure noExtFieldSilent + currentCCS + upd_flag bndrs rhs typ + where + upd_flag | is_join = JumpedTo + | not (null bndrs) = ReEntrant + | isAtMostOnceDmd (idDemandInfo bndr) = SingleEntry + | otherwise = Updatable + + {- + SDM: disabled. Eval/Apply can't handle functions with arity zero very + well; and making these into simple non-updatable thunks breaks other + assumptions (namely that they will be entered only once). + + upd_flag | isPAP env rhs = ReEntrant + | otherwise = Updatable + +-- Detect thunks which will reduce immediately to PAPs, and make them +-- non-updatable. This has several advantages: +-- +-- - the non-updatable thunk behaves exactly like the PAP, +-- +-- - the thunk is more efficient to enter, because it is +-- specialised to the task. +-- +-- - we save one update frame, one stg_update_PAP, one update +-- and lots of PAP_enters. +-- +-- - in the case where the thunk is top-level, we save building +-- a black hole and furthermore the thunk isn't considered to +-- be a CAF any more, so it doesn't appear in any SRTs. +-- +-- We do it here, because the arity information is accurate, and we need +-- to do it before the SRT pass to save the SRT entries associated with +-- any top-level PAPs. + +isPAP env (StgApp f args) = listLengthCmp args arity == LT -- idArity f > length args + where + arity = stgArity f (lookupBinding env f) +isPAP env _ = False + +-} + +{- ToDo: + upd = if isOnceDem dem + then (if isNotTop toplev + then SingleEntry -- HA! Paydirt for "dem" + else + (if debugIsOn then trace "WARNING: SE CAFs unsupported, forcing UPD instead" else id) $ + Updatable) + else Updatable + -- For now we forbid SingleEntry CAFs; they tickle the + -- ASSERT in rts/Storage.c line 215 at newCAF() re mut_link, + -- and I don't understand why. There's only one SE_CAF (well, + -- only one that tickled a great gaping bug in an earlier attempt + -- at ClosureInfo.getEntryConvention) in the whole of nofib, + -- specifically Main.lvl6 in spectral/cryptarithm2. + -- So no great loss. KSW 2000-07. +-} + + +-- | Try to make a non top-level StgRhsCon if appropriate +mkStgRhsCon_maybe :: MkStgRhs -> Maybe StgRhs +mkStgRhsCon_maybe (MkStgRhs bndrs rhs typ is_join) + | [] <- bndrs + , not is_join + , (ticks, StgConApp con mn args _) <- stripStgTicksTop (not . tickishIsCode) rhs + = Just (StgRhsCon currentCCS con mn ticks args typ) + + | otherwise = Nothing + + +-- | Try to make a top-level StgRhsCon if appropriate +mkTopStgRhsCon_maybe :: (DataCon -> [StgArg] -> Bool) -> MkStgRhs -> Maybe StgRhs +mkTopStgRhsCon_maybe allow_static_con_app (MkStgRhs bndrs rhs typ is_join) + | [] <- bndrs + , not is_join -- shouldn't happen at top-level + , (ticks, StgConApp con mn args _) <- stripStgTicksTop (not . tickishIsCode) rhs + , allow_static_con_app con args + = Just (StgRhsCon dontCareCCS con mn ticks args typ) + + | otherwise = Nothing ===================================== compiler/GHC/Stg/Pipeline.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Stg.CSE ( stgCse ) import GHC.Stg.Lift ( StgLiftConfig, stgLiftLams ) import GHC.Unit.Module ( Module ) +import GHC.Core.DataCon (DataCon) + import GHC.Utils.Error import GHC.Types.Var import GHC.Types.Var.Set @@ -52,6 +54,12 @@ data StgPipelineOpts = StgPipelineOpts , stgPipeline_pprOpts :: !StgPprOpts , stgPlatform :: !Platform , stgPipeline_forBytecode :: !Bool + + , stgPipeline_allowTopLevelConApp :: Module -> DataCon -> [StgArg] -> Bool + -- ^ Is a top-level (static) StgConApp allowed or not. If not, use dynamic allocation. + -- + -- This is typically used to support dynamic linking on Windows and the + -- -fexternal-dynamic-refs flag. See GHC.Stg.Utils.allowTopLevelConApp. } newtype StgM a = StgM { _unStgM :: ReaderT Char IO a } @@ -136,7 +144,7 @@ stg2stg logger extra_vars opts this_mod binds StgUnarise -> do us <- getUniqueSupplyM liftIO (stg_linter False "Pre-unarise" binds) - let binds' = {-# SCC "StgUnarise" #-} unarise us binds + let binds' = {-# SCC "StgUnarise" #-} unarise us (stgPipeline_allowTopLevelConApp opts this_mod) binds liftIO (dump_when Opt_D_dump_stg_unarised "Unarised STG:" binds') liftIO (stg_linter True "Unarise" binds') return binds' ===================================== compiler/GHC/Stg/Stats.hs ===================================== @@ -46,6 +46,7 @@ data CounterType | ReEntrantBinds Bool{-ditto-} | SingleEntryBinds Bool{-ditto-} | UpdatableBinds Bool{-ditto-} + | JoinPointBinds Bool{-ditto-} deriving (Eq, Ord) type Count = Int @@ -94,6 +95,7 @@ showStgStats prog s (ReEntrantBinds _) = "ReEntrantBindsBinds_Nested " s (SingleEntryBinds _) = "SingleEntryBinds_Nested " s (UpdatableBinds _) = "UpdatableBinds_Nested " + s (JoinPointBinds _) = "JoinPointBinds_Nested " gatherStgStats :: [StgTopBinding] -> StatEnv gatherStgStats binds = combineSEs (map statTopBinding binds) @@ -132,6 +134,7 @@ statRhs top (_, StgRhsClosure _ _ u _ body _) ReEntrant -> ReEntrantBinds top Updatable -> UpdatableBinds top SingleEntry -> SingleEntryBinds top + JumpedTo -> JoinPointBinds top ) {- ===================================== compiler/GHC/Stg/Syntax.hs ===================================== @@ -54,7 +54,6 @@ module GHC.Stg.Syntax ( -- utils stgRhsArity, freeVarsOfRhs, - isDllConApp, stgArgType, stgArgRep, stgArgRep1, @@ -87,17 +86,14 @@ import GHC.Core.Ppr( {- instances -} ) import GHC.Types.ForeignCall ( ForeignCall ) import GHC.Types.Id -import GHC.Types.Name ( isDynLinkName ) import GHC.Types.Tickish ( StgTickish ) import GHC.Types.Var.Set import GHC.Types.Literal ( Literal, literalType ) import GHC.Types.RepType ( typePrimRep, typePrimRep1, typePrimRepU, typePrimRep_maybe ) -import GHC.Unit.Module ( Module ) import GHC.Utils.Outputable import GHC.Utils.Panic.Plain -import GHC.Platform import GHC.Builtin.PrimOps ( PrimOp, PrimCall ) import Data.ByteString ( ByteString ) @@ -138,51 +134,6 @@ data StgArg = StgVarArg Id | StgLitArg Literal --- | Does this constructor application refer to anything in a different --- *Windows* DLL? --- If so, we can't allocate it statically -isDllConApp - :: Platform - -> Bool -- is Opt_ExternalDynamicRefs enabled? - -> Module - -> DataCon - -> [StgArg] - -> Bool -isDllConApp platform ext_dyn_refs this_mod con args - | not ext_dyn_refs = False - | platformOS platform == OSMinGW32 - = isDynLinkName platform this_mod (dataConName con) || any is_dll_arg args - | otherwise = False - where - -- NB: typePrimRep1 is legit because any free variables won't have - -- unlifted type (there are no unlifted things at top level) - is_dll_arg :: StgArg -> Bool - is_dll_arg (StgVarArg v) = isAddrRep (typePrimRep1 (idType v)) - && isDynLinkName platform this_mod (idName v) - is_dll_arg _ = False - --- True of machine addresses; these are the things that don't work across DLLs. --- The key point here is that VoidRep comes out False, so that a top level --- nullary GADT constructor is False for isDllConApp --- --- data T a where --- T1 :: T Int --- --- gives --- --- T1 :: forall a. (a~Int) -> T a --- --- and hence the top-level binding --- --- $WT1 :: T Int --- $WT1 = T1 Int (Coercion (Refl Int)) --- --- The coercion argument here gets VoidRep -isAddrRep :: PrimOrVoidRep -> Bool -isAddrRep (NVRep AddrRep) = True -isAddrRep (NVRep (BoxedRep _)) = True -- FIXME: not true for JavaScript -isAddrRep _ = False - -- | Type of an @StgArg@ -- -- Very half baked because we have lost the type arguments. @@ -721,24 +672,35 @@ UpdateFlag This is also used in @LambdaFormInfo@ in the @ClosureInfo@ module. -A @ReEntrant@ closure may be entered multiple times, but should not be updated -or blackholed. An @Updatable@ closure should be updated after evaluation (and -may be blackholed during evaluation). A @SingleEntry@ closure will only be -entered once, and so need not be updated but may safely be blackholed. -} -data UpdateFlag = ReEntrant | Updatable | SingleEntry +data UpdateFlag + = ReEntrant + -- ^ A @ReEntrant@ closure may be entered multiple times, but should not + -- be updated or blackholed. + | Updatable + -- ^ An @Updatable@ closure should be updated after evaluation (and may be + -- blackholed during evaluation). + | SingleEntry + -- ^ A @SingleEntry@ closure will only be entered once, and so need not be + -- updated but may safely be blackholed. + | JumpedTo + -- ^ A @JumpedTo@ (join-point) closure is entered once or multiple times + -- but has no heap-allocated associated closure. + deriving (Show,Eq) instance Outputable UpdateFlag where ppr u = char $ case u of ReEntrant -> 'r' Updatable -> 'u' SingleEntry -> 's' + JumpedTo -> 'j' isUpdatable :: UpdateFlag -> Bool isUpdatable ReEntrant = False isUpdatable SingleEntry = False isUpdatable Updatable = True +isUpdatable JumpedTo = False {- ************************************************************************ ===================================== compiler/GHC/Stg/Unarise.hs ===================================== @@ -1,6 +1,7 @@ {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE TupleSections #-} +{-# LANGUAGE MultiWayIf #-} {- (c) The GRASP/AQUA Project, Glasgow University, 1992-2012 @@ -401,6 +402,7 @@ import GHC.Utils.Panic import GHC.Types.RepType import GHC.Stg.Syntax import GHC.Stg.Utils +import GHC.Stg.Make import GHC.Core.Type import GHC.Builtin.Types.Prim (intPrimTy) import GHC.Builtin.Types @@ -442,10 +444,14 @@ import Data.List (mapAccumL) -- INVARIANT: OutStgArgs in the range only have NvUnaryTypes -- (i.e. no unboxed tuples, sums or voids) -- -newtype UnariseEnv = UnariseEnv { ue_rho :: (VarEnv UnariseVal) } +data UnariseEnv = UnariseEnv + { ue_rho :: (VarEnv UnariseVal) + , ue_allow_static_conapp :: DataCon -> [StgArg] -> Bool + } -initUnariseEnv :: VarEnv UnariseVal -> UnariseEnv +initUnariseEnv :: VarEnv UnariseVal -> (DataCon -> [StgArg] -> Bool) -> UnariseEnv initUnariseEnv = UnariseEnv + data UnariseVal = MultiVal [OutStgArg] -- MultiVal to tuple. Can be empty list (void). | UnaryVal OutStgArg -- See Note [Renaming during unarisation]. @@ -477,27 +483,57 @@ lookupRho env v = lookupVarEnv (ue_rho env) v -------------------------------------------------------------------------------- -unarise :: UniqSupply -> [StgTopBinding] -> [StgTopBinding] -unarise us binds = initUs_ us (mapM (unariseTopBinding (initUnariseEnv emptyVarEnv)) binds) +unarise :: UniqSupply -> (DataCon -> [StgArg] -> Bool) -> [StgTopBinding] -> [StgTopBinding] +unarise us is_dll_con_app binds = initUs_ us (mapM (unariseTopBinding (initUnariseEnv emptyVarEnv is_dll_con_app)) binds) unariseTopBinding :: UnariseEnv -> StgTopBinding -> UniqSM StgTopBinding unariseTopBinding rho (StgTopLifted bind) - = StgTopLifted <$> unariseBinding rho bind + = StgTopLifted <$> unariseBinding rho True bind unariseTopBinding _ bind at StgTopStringLit{} = return bind -unariseBinding :: UnariseEnv -> StgBinding -> UniqSM StgBinding -unariseBinding rho (StgNonRec x rhs) - = StgNonRec x <$> unariseRhs rho rhs -unariseBinding rho (StgRec xrhss) - = StgRec <$> mapM (\(x, rhs) -> (x,) <$> unariseRhs rho rhs) xrhss +unariseBinding :: UnariseEnv -> Bool -> StgBinding -> UniqSM StgBinding +unariseBinding rho top_level (StgNonRec x rhs) + = StgNonRec x <$> unariseRhs rho top_level rhs +unariseBinding rho top_level (StgRec xrhss) + = StgRec <$> mapM (\(x, rhs) -> (x,) <$> unariseRhs rho top_level rhs) xrhss -unariseRhs :: UnariseEnv -> StgRhs -> UniqSM StgRhs -unariseRhs rho (StgRhsClosure ext ccs update_flag args expr typ) +unariseRhs :: UnariseEnv -> Bool -> StgRhs -> UniqSM StgRhs +unariseRhs rho top_level (StgRhsClosure ext ccs update_flag args expr typ) = do (rho', args1) <- unariseFunArgBinders rho args expr' <- unariseExpr rho' expr - return (StgRhsClosure ext ccs update_flag args1 expr' typ) - -unariseRhs rho (StgRhsCon ccs con mu ts args typ) + -- Unarisation can lead to a StgRhsClosure becoming a StgRhsCon. + -- Hence, we call `mk(Top)StgRhsCon_maybe` rather than just building + -- another `StgRhsClosure`. + -- + -- For example with unboxed sums (#25166): + -- + -- foo = \u [] case (# | _ | #) [(##)] of tag { __DEFAULT -> D [True tag] } + -- + -- ====> {unarisation} + -- + -- foo = D [True 2#] + -- + -- Transforming an appropriate StgRhsClosure into a StgRhsCon is + -- important as top-level StgRhsCon are statically allocated. + -- + let mk_rhs = MkStgRhs + { rhs_args = args1 + , rhs_expr = expr' + , rhs_type = typ + , rhs_is_join = update_flag == JumpedTo + } + if | top_level + , Just rhs_con <- mkTopStgRhsCon_maybe (ue_allow_static_conapp rho) mk_rhs + -> pure rhs_con + + | not top_level + , Just rhs_con <- mkStgRhsCon_maybe mk_rhs + -> pure rhs_con + + | otherwise + -> pure (StgRhsClosure ext ccs update_flag args1 expr' typ) + +unariseRhs rho _top (StgRhsCon ccs con mu ts args typ) = assert (not (isUnboxedTupleDataCon con || isUnboxedSumDataCon con)) return (StgRhsCon ccs con mu ts (unariseConArgs rho args) typ) @@ -576,10 +612,10 @@ unariseExpr rho (StgCase scrut bndr alt_ty alts) -- dead after unarise (checked in GHC.Stg.Lint) unariseExpr rho (StgLet ext bind e) - = StgLet ext <$> unariseBinding rho bind <*> unariseExpr rho e + = StgLet ext <$> unariseBinding rho False bind <*> unariseExpr rho e unariseExpr rho (StgLetNoEscape ext bind e) - = StgLetNoEscape ext <$> unariseBinding rho bind <*> unariseExpr rho e + = StgLetNoEscape ext <$> unariseBinding rho False bind <*> unariseExpr rho e unariseExpr rho (StgTick tick e) = StgTick tick <$> unariseExpr rho e ===================================== compiler/GHC/Stg/Utils.hs ===================================== @@ -9,9 +9,12 @@ module GHC.Stg.Utils , idArgs , mkUnarisedId, mkUnarisedIds + + , allowTopLevelConApp ) where import GHC.Prelude +import GHC.Platform import GHC.Types.Id import GHC.Core.Type @@ -22,6 +25,8 @@ import GHC.Types.Tickish import GHC.Types.Unique.Supply import GHC.Types.RepType +import GHC.Types.Name ( isDynLinkName ) +import GHC.Unit.Module ( Module ) import GHC.Stg.Syntax import GHC.Utils.Outputable @@ -122,3 +127,54 @@ stripStgTicksTopE :: (StgTickish -> Bool) -> GenStgExpr p -> GenStgExpr p stripStgTicksTopE p = go where go (StgTick t e) | p t = go e go other = other + +-- | Do we allow the given top-level (static) ConApp? +allowTopLevelConApp + :: Platform + -> Bool -- is Opt_ExternalDynamicRefs enabled? + -> Module + -> DataCon + -> [StgArg] + -> Bool +allowTopLevelConApp platform ext_dyn_refs this_mod con args + -- we're not using dynamic linking + | not ext_dyn_refs = True + -- if the target OS is Windows, we only allow top-level ConApps if they don't + -- reference external names (Windows DLLs have a problem with static cross-DLL + -- refs) + | platformOS platform == OSMinGW32 = not is_external_con_app + -- otherwise, allowed + -- Sylvain: shouldn't this be False when (ext_dyn_refs && is_external_con_app)? + | otherwise = True + where + is_external_con_app = isDynLinkName platform this_mod (dataConName con) || any is_dll_arg args + + -- NB: typePrimRep1 is legit because any free variables won't have + -- unlifted type (there are no unlifted things at top level) + is_dll_arg :: StgArg -> Bool + is_dll_arg (StgVarArg v) = isAddrRep (typePrimRep1 (idType v)) + && isDynLinkName platform this_mod (idName v) + is_dll_arg _ = False + +-- True of machine addresses; these are the things that don't work across DLLs. +-- The key point here is that VoidRep comes out False, so that a top level +-- nullary GADT constructor is True for allowTopLevelConApp +-- +-- data T a where +-- T1 :: T Int +-- +-- gives +-- +-- T1 :: forall a. (a~Int) -> T a +-- +-- and hence the top-level binding +-- +-- $WT1 :: T Int +-- $WT1 = T1 Int (Coercion (Refl Int)) +-- +-- The coercion argument here gets VoidRep +isAddrRep :: PrimOrVoidRep -> Bool +isAddrRep (NVRep AddrRep) = True +isAddrRep (NVRep (BoxedRep _)) = True -- FIXME: not true for JavaScript +isAddrRep _ = False + ===================================== compiler/GHC/StgToCmm/DataCon.hs ===================================== @@ -19,6 +19,7 @@ import GHC.Prelude import GHC.Platform +import GHC.Stg.Utils (allowTopLevelConApp) import GHC.Stg.Syntax import GHC.Core ( AltCon(..) ) @@ -48,7 +49,6 @@ import GHC.Utils.Panic import GHC.Utils.Misc import GHC.Utils.Monad (mapMaybeM) -import Control.Monad import Data.Char import GHC.StgToCmm.Config (stgToCmmPlatform) import GHC.StgToCmm.TagCheck (checkConArgsStatic, checkConArgsDyn) @@ -90,10 +90,8 @@ cgTopRhsCon cfg id con mn args gen_code = do { profile <- getProfile ; this_mod <- getModuleName - ; when (platformOS platform == OSMinGW32) $ - -- Windows DLLs have a problem with static cross-DLL refs. - massert (not (isDllConApp platform (stgToCmmExtDynRefs cfg) this_mod con (map fromNonVoid args))) - ; assert (args `lengthIs` countConRepArgs con ) return () + ; massert (allowTopLevelConApp platform (stgToCmmExtDynRefs cfg) this_mod con (map fromNonVoid args)) + ; massert (args `lengthIs` countConRepArgs con ) ; checkConArgsStatic (text "TagCheck failed - Top level con") con (map fromNonVoid args) -- LAY IT OUT ; let ===================================== compiler/GHC/Types/CostCentre.hs ===================================== @@ -4,6 +4,7 @@ module GHC.Types.CostCentre ( CostCentre(..), CcName, CCFlavour, mkCafFlavour, mkExprCCFlavour, mkDeclCCFlavour, mkHpcCCFlavour, mkLateCCFlavour, mkCallerCCFlavour, + getAllCAFsCC, pprCostCentre, CostCentreStack, @@ -393,3 +394,13 @@ instance Binary CostCentre where -- ok, because we only need the SrcSpan when declaring the -- CostCentre in the original module, it is not used by importing -- modules. + +getAllCAFsCC :: Module -> (CostCentre, CostCentreStack) +getAllCAFsCC this_mod = + let + span = mkGeneralSrcSpan (mkFastString "") -- XXX do better + all_cafs_cc = mkAllCafsCC this_mod span + all_cafs_ccs = mkSingletonCCS all_cafs_cc + in + (all_cafs_cc, all_cafs_ccs) + ===================================== compiler/ghc.cabal.in ===================================== @@ -702,6 +702,7 @@ Library GHC.Stg.InferTags.Rewrite GHC.Stg.InferTags.TagSig GHC.Stg.InferTags.Types + GHC.Stg.Make GHC.Stg.Pipeline GHC.Stg.Stats GHC.Stg.Subst ===================================== testsuite/tests/codeGen/should_compile/Makefile ===================================== @@ -77,3 +77,6 @@ T17648: -fcatch-nonexhaustive-cases T17648.hs -v0 -fforce-recomp '$(TEST_HC)' --show-iface T17648.hi | tr -d '\n\r' | \ grep -F 'f :: T GHC.Types.Int -> () [TagSig' >/dev/null + +T25166: + '$(TEST_HC)' -O2 -dno-typeable-binds -ddump-cmm T25166.hs | awk '/foo_closure/{flag=1}/}]/{flag=0}flag' ===================================== testsuite/tests/codeGen/should_compile/T25166.hs ===================================== @@ -0,0 +1,7 @@ +module Test where + +data A = A | B | C + +data D = D !Bool {-# UNPACK #-} !A + +foo = D True B ===================================== testsuite/tests/codeGen/should_compile/T25166.stdout ===================================== @@ -0,0 +1,6 @@ +[section ""data" . Test.foo_closure" { + Test.foo_closure: + const Test.D_con_info; + const GHC.Types.True_closure+2; + const 2; + const 3; ===================================== testsuite/tests/codeGen/should_compile/all.T ===================================== @@ -139,5 +139,7 @@ test('callee-no-local', [ ['-ddump-cmm-raw'] ) +test('T25166', [req_cmm], makefile_test, []) + # dump Core to ensure that d is defined as: d = D 10## RUBBISH(IntRep) test('T25177', normal, compile, ['-O2 -dno-typeable-binds -ddump-simpl -dsuppress-all -dsuppress-uniques -v0']) ===================================== utils/haddock/doc/requirements.txt ===================================== @@ -0,0 +1 @@ +sphinxcontrib-applehelp ==2.0.0 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7658b654218f4496d47d62551de0c54e8819438c...573f9833a7c56bf8fd8c6731069a7160541e5287 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7658b654218f4496d47d62551de0c54e8819438c...573f9833a7c56bf8fd8c6731069a7160541e5287 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 13:28:49 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sun, 08 Sep 2024 09:28:49 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 3 commits: haddock: Add missing requirements.txt for the online manual Message-ID: <66dda691d459_f46243eb32c30355@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - faf45b7b by Hécate Kleidukos at 2024-09-08T09:28:41-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 3 changed files: - compiler/GHC/CmmToAsm/AArch64/Instr.hs - utils/haddock/.readthedocs.yaml - + utils/haddock/doc/requirements.txt Changes: ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -30,6 +30,7 @@ import GHC.Utils.Panic import Data.Maybe (fromMaybe) import GHC.Stack +import GHC.Platform.Reg.Class -- | LR and FP (8 byte each) are the prologue of each stack frame stackFrameHeaderSize :: Int @@ -454,10 +455,22 @@ isMetaInstr instr mkRegRegMoveInstr :: Reg -> Reg -> Instr mkRegRegMoveInstr src dst = ANN (text "Reg->Reg Move: " <> ppr src <> text " -> " <> ppr dst) $ MOV (OpReg W64 dst) (OpReg W64 src) --- | Take the source and destination from this reg -> reg move instruction --- or Nothing if it's not one +-- | Take the source and destination registers from a move instruction of same +-- register class (`RegClass`). +-- +-- The idea is to identify moves that can be eliminated by the register +-- allocator: If the source register serves no special purpose, one could +-- continue using it; saving one move instruction. For this, the register kinds +-- (classes) must be the same (no conversion involved.) takeRegRegMoveInstr :: Instr -> Maybe (Reg,Reg) ---takeRegRegMoveInstr (MOV (OpReg fmt dst) (OpReg fmt' src)) | fmt == fmt' = Just (src, dst) +takeRegRegMoveInstr (MOV (OpReg _fmt dst) (OpReg _fmt' src)) + | classOfReg dst == classOfReg src = pure (src, dst) + where + classOfReg ::Reg -> RegClass + classOfReg reg + = case reg of + RegVirtual vr -> classOfVirtualReg vr + RegReal rr -> classOfRealReg rr takeRegRegMoveInstr _ = Nothing -- | Make an unconditional jump instruction. ===================================== utils/haddock/.readthedocs.yaml ===================================== @@ -2,7 +2,7 @@ version: 2 sphinx: - configuration: doc/conf.py + configuration: utils/haddock/doc/conf.py build: os: "ubuntu-22.04" @@ -11,4 +11,4 @@ build: python: install: - - requirements: doc/requirements.txt + - requirements: utils/haddock/doc/requirements.txt ===================================== utils/haddock/doc/requirements.txt ===================================== @@ -0,0 +1 @@ +sphinxcontrib-applehelp ==2.0.0 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5bba677e1bd93fc514b66283b87fd281fc620402...faf45b7b2153085e16fca4e9f39f61234a3a8e4b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5bba677e1bd93fc514b66283b87fd281fc620402...faf45b7b2153085e16fca4e9f39f61234a3a8e4b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 15:22:19 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 08 Sep 2024 11:22:19 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/supersven/aarch64-jump-tables Message-ID: <66ddc12bbfc8e_2c027e8acdc28069@gitlab.mail> Sven Tennie pushed new branch wip/supersven/aarch64-jump-tables at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/supersven/aarch64-jump-tables You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 15:25:20 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 08 Sep 2024 11:25:20 -0400 Subject: [Git][ghc/ghc][wip/supersven/aarch64-jump-tables] AArch64: Implement switch/jump tables (#19912) Message-ID: <66ddc1e018062_2c027e8aed030074@gitlab.mail> Sven Tennie pushed to branch wip/supersven/aarch64-jump-tables at Glasgow Haskell Compiler / GHC Commits: d73402bb by Sven Tennie at 2024-09-08T17:24:32+02:00 AArch64: Implement switch/jump tables (#19912) This improves the performance of Cmm switch statements (compared to a chain of if statements.) - - - - - 3 changed files: - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs Changes: ===================================== compiler/GHC/CmmToAsm/AArch64/CodeGen.hs ===================================== @@ -23,7 +23,7 @@ import GHC.Cmm.DebugBlock import GHC.CmmToAsm.Monad ( NatM, getNewRegNat , getPicBaseMaybeNat, getPlatform, getConfig - , getDebugBlock, getFileId + , getDebugBlock, getFileId, getNewLabelNat ) -- import GHC.CmmToAsm.Instr import GHC.CmmToAsm.PIC @@ -50,7 +50,7 @@ import GHC.Types.Unique.Supply import GHC.Data.OrdList import GHC.Utils.Outputable -import Control.Monad ( mapAndUnzipM, foldM ) +import Control.Monad ( mapAndUnzipM ) import GHC.Float import GHC.Types.Basic @@ -209,43 +209,79 @@ annExpr e instr {- debugIsOn -} = ANN (text . show $ e) instr -- ----------------------------------------------------------------------------- -- Generating a table-branch --- TODO jump tables would be a lot faster, but we'll use bare bones for now. --- this is usually done by sticking the jump table ids into an instruction --- and then have the @generateJumpTableForInstr@ callback produce the jump --- table as a static. +-- | Generate jump to jump table target -- --- See Ticket 19912 --- --- data SwitchTargets = --- SwitchTargets --- Bool -- Signed values --- (Integer, Integer) -- Range --- (Maybe Label) -- Default value --- (M.Map Integer Label) -- The branches --- --- Non Jumptable plan: --- xE <- expr +-- The index into the jump table is calulated by evaluating @expr at . The +-- corresponding table entry contains the relative address to jump to (relative +-- to the jump table's first entry / the table's own label). +genSwitch :: NCGConfig -> CmmExpr -> SwitchTargets -> NatM InstrBlock +genSwitch config expr targets = do + (reg, fmt1, e_code) <- getSomeReg indexExpr + let fmt = II64 + targetReg <- getNewRegNat fmt + lbl <- getNewLabelNat + dynRef <- cmmMakeDynamicReference config DataReference lbl + (tableReg, fmt2, t_code) <- getSomeReg dynRef + let code = + toOL + [ COMMENT (text "indexExpr" <+> (text . show) indexExpr), + COMMENT (text "dynRef" <+> (text . show) dynRef) + ] + `appOL` e_code + `appOL` t_code + `appOL` toOL + [ COMMENT (ftext "Jump table for switch"), + -- index to offset into the table (relative to tableReg) + annExpr expr (LSL (OpReg (formatToWidth fmt1) reg) (OpReg (formatToWidth fmt1) reg) (OpImm (ImmInt 3))), + -- calculate table entry address + ADD (OpReg W64 targetReg) (OpReg (formatToWidth fmt1) reg) (OpReg (formatToWidth fmt2) tableReg), + -- load table entry (relative offset from tableReg (first entry) to target label) + LDR II64 (OpReg W64 targetReg) (OpAddr (AddrRegImm targetReg (ImmInt 0))), + -- calculate absolute address of the target label + ADD (OpReg W64 targetReg) (OpReg W64 targetReg) (OpReg W64 tableReg), + -- prepare jump to target label + J_TBL ids (Just lbl) targetReg + ] + return code + where + -- See Note [Sub-word subtlety during jump-table indexing] in + -- GHC.CmmToAsm.X86.CodeGen for why we must first offset, then widen. + indexExpr0 = cmmOffset platform expr offset + -- We widen to a native-width register to sanitize the high bits + indexExpr = + CmmMachOp + (MO_UU_Conv expr_w (platformWordWidth platform)) + [indexExpr0] + expr_w = cmmExprWidth platform expr + (offset, ids) = switchTargetsToTable targets + platform = ncgPlatform config + +-- | Generate jump table data (if required) -- -genSwitch :: CmmExpr -> SwitchTargets -> NatM InstrBlock -genSwitch expr targets = do -- pprPanic "genSwitch" (ppr expr) - (reg, format, code) <- getSomeReg expr - let w = formatToWidth format - let mkbranch acc (key, bid) = do - (keyReg, _format, code) <- getSomeReg (CmmLit (CmmInt key w)) - return $ code `appOL` - toOL [ CMP (OpReg w reg) (OpReg w keyReg) - , BCOND EQ (TBlock bid) - ] `appOL` acc - def_code = case switchTargetsDefault targets of - Just bid -> unitOL (B (TBlock bid)) - Nothing -> nilOL - - switch_code <- foldM mkbranch nilOL (switchTargetsCases targets) - return $ code `appOL` switch_code `appOL` def_code - --- We don't do jump tables for now, see Ticket 19912 -generateJumpTableForInstr :: NCGConfig -> Instr - -> Maybe (NatCmmDecl RawCmmStatics Instr) +-- The idea is to emit one table entry per case. The entry is the relative +-- address of the block to jump to (relative to the table's first entry / +-- table's own label.) The calculation itself is done by the linker. +generateJumpTableForInstr :: + NCGConfig -> + Instr -> + Maybe (NatCmmDecl RawCmmStatics Instr) +generateJumpTableForInstr config (J_TBL ids (Just lbl) _) = + let jumpTable = + map jumpTableEntryRel ids + where + jumpTableEntryRel Nothing = + CmmStaticLit (CmmInt 0 (ncgWordWidth config)) + jumpTableEntryRel (Just blockid) = + CmmStaticLit + ( CmmLabelDiffOff + blockLabel + lbl + 0 + (ncgWordWidth config) + ) + where + blockLabel = blockLbl blockid + in Just (CmmData (Section ReadOnlyData lbl) (CmmStaticsRaw lbl jumpTable)) generateJumpTableForInstr _ _ = Nothing -- ----------------------------------------------------------------------------- @@ -266,6 +302,7 @@ stmtToInstrs :: CmmNode e x -- ^ Cmm Statement stmtToInstrs stmt = do -- traceM $ "-- -------------------------- stmtToInstrs -------------------------- --\n" -- ++ showSDocUnsafe (ppr stmt) + config <- getConfig platform <- getPlatform case stmt of CmmUnsafeForeignCall target result_regs args @@ -294,7 +331,7 @@ stmtToInstrs stmt = do CmmCondBranch arg true false _prediction -> genCondBranch true false arg - CmmSwitch arg ids -> genSwitch arg ids + CmmSwitch arg ids -> genSwitch config arg ids CmmCall { cml_target = arg } -> genJump arg ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -27,7 +27,7 @@ import GHC.Types.Unique.Supply import GHC.Utils.Panic -import Data.Maybe (fromMaybe) +import Data.Maybe (fromMaybe, catMaybes) import GHC.Stack @@ -118,6 +118,7 @@ regUsageOfInstr platform instr = case instr of ORR dst src1 src2 -> usage (regOp src1 ++ regOp src2, regOp dst) -- 4. Branch Instructions ---------------------------------------------------- J t -> usage (regTarget t, []) + J_TBL _ _ t -> usage ([t], []) B t -> usage (regTarget t, []) BCOND _ t -> usage (regTarget t, []) BL t ps -> usage (regTarget t ++ ps, callerSavedRegisters) @@ -264,10 +265,11 @@ patchRegsOfInstr instr env = case instr of ORR o1 o2 o3 -> ORR (patchOp o1) (patchOp o2) (patchOp o3) -- 4. Branch Instructions -------------------------------------------------- - J t -> J (patchTarget t) - B t -> B (patchTarget t) - BL t rs -> BL (patchTarget t) rs - BCOND c t -> BCOND c (patchTarget t) + J t -> J (patchTarget t) + J_TBL ids mbLbl t -> J_TBL ids mbLbl (env t) + B t -> B (patchTarget t) + BL t rs -> BL (patchTarget t) rs + BCOND c t -> BCOND c (patchTarget t) -- 5. Atomic Instructions -------------------------------------------------- -- 6. Conditional Instructions --------------------------------------------- @@ -319,6 +321,7 @@ isJumpishInstr instr = case instr of CBZ{} -> True CBNZ{} -> True J{} -> True + J_TBL{} -> True B{} -> True BL{} -> True BCOND{} -> True @@ -332,6 +335,7 @@ jumpDestsOfInstr (ANN _ i) = jumpDestsOfInstr i jumpDestsOfInstr (CBZ _ t) = [ id | TBlock id <- [t]] jumpDestsOfInstr (CBNZ _ t) = [ id | TBlock id <- [t]] jumpDestsOfInstr (J t) = [id | TBlock id <- [t]] +jumpDestsOfInstr (J_TBL ids _mbLbl _r) = catMaybes ids jumpDestsOfInstr (B t) = [id | TBlock id <- [t]] jumpDestsOfInstr (BL t _) = [ id | TBlock id <- [t]] jumpDestsOfInstr (BCOND _ t) = [ id | TBlock id <- [t]] @@ -340,6 +344,11 @@ jumpDestsOfInstr _ = [] canFallthroughTo :: Instr -> BlockId -> Bool canFallthroughTo (ANN _ i) bid = canFallthroughTo i bid canFallthroughTo (J (TBlock target)) bid = bid == target +canFallthroughTo (J_TBL targets _ _) bid = all isTargetBid targets + where + isTargetBid target = case target of + Nothing -> True + Just target -> target == bid canFallthroughTo (B (TBlock target)) bid = bid == target canFallthroughTo _ _ = False @@ -353,6 +362,7 @@ patchJumpInstr instr patchF CBZ r (TBlock bid) -> CBZ r (TBlock (patchF bid)) CBNZ r (TBlock bid) -> CBNZ r (TBlock (patchF bid)) J (TBlock bid) -> J (TBlock (patchF bid)) + J_TBL ids mbLbl r -> J_TBL (map (fmap patchF) ids) mbLbl r B (TBlock bid) -> B (TBlock (patchF bid)) BL (TBlock bid) ps -> BL (TBlock (patchF bid)) ps BCOND c (TBlock bid) -> BCOND c (TBlock (patchF bid)) @@ -516,6 +526,7 @@ allocMoreStack platform slots proc@(CmmProc info lbl live (ListGraph code)) = do insert_dealloc insn r = case insn of J _ -> dealloc ++ (insn : r) + J_TBL {} -> dealloc ++ (insn : r) ANN _ (J _) -> dealloc ++ (insn : r) _other | jumpDestsOfInstr insn /= [] -> patchJumpInstr insn retarget : r @@ -644,6 +655,7 @@ data Instr | CBNZ Operand Target -- if op /= 0, then branch. -- Branching. | J Target -- like B, but only generated from genJump. Used to distinguish genJumps from others. + | J_TBL [Maybe BlockId] (Maybe CLabel) Reg -- A jump instruction with data for switch/jump tables | B Target -- unconditional branching b/br. (To a blockid, label or register) | BL Target [Reg] -- branch and link (e.g. set x30 to next pc, and branch) | BCOND Cond Target -- branch with condition. b. @@ -730,6 +742,7 @@ instrCon i = CBZ{} -> "CBZ" CBNZ{} -> "CBNZ" J{} -> "J" + J_TBL {} -> "J_TBL" B{} -> "B" BL{} -> "BL" BCOND{} -> "BCOND" ===================================== compiler/GHC/CmmToAsm/AArch64/Ppr.hs ===================================== @@ -426,6 +426,7 @@ pprInstr platform instr = case instr of -- 4. Branch Instructions ---------------------------------------------------- J t -> pprInstr platform (B t) + J_TBL _ _ r -> pprInstr platform (B (TReg r)) B (TBlock bid) -> line $ text "\tb" <+> pprAsmLabel platform (mkLocalBlockLabel (getUnique bid)) B (TLabel lbl) -> line $ text "\tb" <+> pprAsmLabel platform lbl B (TReg r) -> line $ text "\tbr" <+> pprReg W64 r View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d73402bb639340540672f0a19b0087d67de976c4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d73402bb639340540672f0a19b0087d67de976c4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 16:34:36 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 08 Sep 2024 12:34:36 -0400 Subject: [Git][ghc/ghc][wip/supersven/aarch64-jump-tables] AArch64: Implement switch/jump tables (#19912) Message-ID: <66ddd21c7e3ce_2c027e43c3d031680@gitlab.mail> Sven Tennie pushed to branch wip/supersven/aarch64-jump-tables at Glasgow Haskell Compiler / GHC Commits: c6721691 by Sven Tennie at 2024-09-08T18:34:04+02:00 AArch64: Implement switch/jump tables (#19912) This improves the performance of Cmm switch statements (compared to a chain of if statements.) - - - - - 3 changed files: - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs Changes: ===================================== compiler/GHC/CmmToAsm/AArch64/CodeGen.hs ===================================== @@ -23,7 +23,7 @@ import GHC.Cmm.DebugBlock import GHC.CmmToAsm.Monad ( NatM, getNewRegNat , getPicBaseMaybeNat, getPlatform, getConfig - , getDebugBlock, getFileId + , getDebugBlock, getFileId, getNewLabelNat ) -- import GHC.CmmToAsm.Instr import GHC.CmmToAsm.PIC @@ -50,7 +50,7 @@ import GHC.Types.Unique.Supply import GHC.Data.OrdList import GHC.Utils.Outputable -import Control.Monad ( mapAndUnzipM, foldM ) +import Control.Monad ( mapAndUnzipM ) import GHC.Float import GHC.Types.Basic @@ -209,43 +209,79 @@ annExpr e instr {- debugIsOn -} = ANN (text . show $ e) instr -- ----------------------------------------------------------------------------- -- Generating a table-branch --- TODO jump tables would be a lot faster, but we'll use bare bones for now. --- this is usually done by sticking the jump table ids into an instruction --- and then have the @generateJumpTableForInstr@ callback produce the jump --- table as a static. +-- | Generate jump to jump table target -- --- See Ticket 19912 --- --- data SwitchTargets = --- SwitchTargets --- Bool -- Signed values --- (Integer, Integer) -- Range --- (Maybe Label) -- Default value --- (M.Map Integer Label) -- The branches --- --- Non Jumptable plan: --- xE <- expr +-- The index into the jump table is calulated by evaluating @expr at . The +-- corresponding table entry contains the relative address to jump to (relative +-- to the jump table's first entry / the table's own label). +genSwitch :: NCGConfig -> CmmExpr -> SwitchTargets -> NatM InstrBlock +genSwitch config expr targets = do + (reg, fmt1, e_code) <- getSomeReg indexExpr + let fmt = II64 + targetReg <- getNewRegNat fmt + lbl <- getNewLabelNat + dynRef <- cmmMakeDynamicReference config DataReference lbl + (tableReg, fmt2, t_code) <- getSomeReg dynRef + let code = + toOL + [ COMMENT (text "indexExpr" <+> (text . show) indexExpr), + COMMENT (text "dynRef" <+> (text . show) dynRef) + ] + `appOL` e_code + `appOL` t_code + `appOL` toOL + [ COMMENT (ftext "Jump table for switch"), + -- index to offset into the table (relative to tableReg) + annExpr expr (LSL (OpReg (formatToWidth fmt1) reg) (OpReg (formatToWidth fmt1) reg) (OpImm (ImmInt 3))), + -- calculate table entry address + ADD (OpReg W64 targetReg) (OpReg (formatToWidth fmt1) reg) (OpReg (formatToWidth fmt2) tableReg), + -- load table entry (relative offset from tableReg (first entry) to target label) + LDR II64 (OpReg W64 targetReg) (OpAddr (AddrRegImm targetReg (ImmInt 0))), + -- calculate absolute address of the target label + ADD (OpReg W64 targetReg) (OpReg W64 targetReg) (OpReg W64 tableReg), + -- prepare jump to target label + J_TBL ids (Just lbl) targetReg + ] + return code + where + -- See Note [Sub-word subtlety during jump-table indexing] in + -- GHC.CmmToAsm.X86.CodeGen for why we must first offset, then widen. + indexExpr0 = cmmOffset platform expr offset + -- We widen to a native-width register to sanitize the high bits + indexExpr = + CmmMachOp + (MO_UU_Conv expr_w (platformWordWidth platform)) + [indexExpr0] + expr_w = cmmExprWidth platform expr + (offset, ids) = switchTargetsToTable targets + platform = ncgPlatform config + +-- | Generate jump table data (if required) -- -genSwitch :: CmmExpr -> SwitchTargets -> NatM InstrBlock -genSwitch expr targets = do -- pprPanic "genSwitch" (ppr expr) - (reg, format, code) <- getSomeReg expr - let w = formatToWidth format - let mkbranch acc (key, bid) = do - (keyReg, _format, code) <- getSomeReg (CmmLit (CmmInt key w)) - return $ code `appOL` - toOL [ CMP (OpReg w reg) (OpReg w keyReg) - , BCOND EQ (TBlock bid) - ] `appOL` acc - def_code = case switchTargetsDefault targets of - Just bid -> unitOL (B (TBlock bid)) - Nothing -> nilOL - - switch_code <- foldM mkbranch nilOL (switchTargetsCases targets) - return $ code `appOL` switch_code `appOL` def_code - --- We don't do jump tables for now, see Ticket 19912 -generateJumpTableForInstr :: NCGConfig -> Instr - -> Maybe (NatCmmDecl RawCmmStatics Instr) +-- The idea is to emit one table entry per case. The entry is the relative +-- address of the block to jump to (relative to the table's first entry / +-- table's own label.) The calculation itself is done by the linker. +generateJumpTableForInstr :: + NCGConfig -> + Instr -> + Maybe (NatCmmDecl RawCmmStatics Instr) +generateJumpTableForInstr config (J_TBL ids (Just lbl) _) = + let jumpTable = + map jumpTableEntryRel ids + where + jumpTableEntryRel Nothing = + CmmStaticLit (CmmInt 0 (ncgWordWidth config)) + jumpTableEntryRel (Just blockid) = + CmmStaticLit + ( CmmLabelDiffOff + blockLabel + lbl + 0 + (ncgWordWidth config) + ) + where + blockLabel = blockLbl blockid + in Just (CmmData (Section ReadOnlyData lbl) (CmmStaticsRaw lbl jumpTable)) generateJumpTableForInstr _ _ = Nothing -- ----------------------------------------------------------------------------- @@ -266,6 +302,7 @@ stmtToInstrs :: CmmNode e x -- ^ Cmm Statement stmtToInstrs stmt = do -- traceM $ "-- -------------------------- stmtToInstrs -------------------------- --\n" -- ++ showSDocUnsafe (ppr stmt) + config <- getConfig platform <- getPlatform case stmt of CmmUnsafeForeignCall target result_regs args @@ -294,7 +331,7 @@ stmtToInstrs stmt = do CmmCondBranch arg true false _prediction -> genCondBranch true false arg - CmmSwitch arg ids -> genSwitch arg ids + CmmSwitch arg ids -> genSwitch config arg ids CmmCall { cml_target = arg } -> genJump arg @@ -339,12 +376,6 @@ getRegisterReg platform (CmmGlobal reg@(GlobalRegUse mid _)) -- ones which map to a real machine register on this -- platform. Hence if it's not mapped to a registers something -- went wrong earlier in the pipeline. --- | Convert a BlockId to some CmmStatic data --- TODO: Add JumpTable Logic, see Ticket 19912 --- jumpTableEntry :: NCGConfig -> Maybe BlockId -> CmmStatic --- jumpTableEntry config Nothing = CmmStaticLit (CmmInt 0 (ncgWordWidth config)) --- jumpTableEntry _ (Just blockid) = CmmStaticLit (CmmLabel blockLabel) --- where blockLabel = blockLbl blockid -- ----------------------------------------------------------------------------- -- General things for putting together code sequences ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -27,7 +27,7 @@ import GHC.Types.Unique.Supply import GHC.Utils.Panic -import Data.Maybe (fromMaybe) +import Data.Maybe (fromMaybe, catMaybes) import GHC.Stack @@ -118,6 +118,7 @@ regUsageOfInstr platform instr = case instr of ORR dst src1 src2 -> usage (regOp src1 ++ regOp src2, regOp dst) -- 4. Branch Instructions ---------------------------------------------------- J t -> usage (regTarget t, []) + J_TBL _ _ t -> usage ([t], []) B t -> usage (regTarget t, []) BCOND _ t -> usage (regTarget t, []) BL t ps -> usage (regTarget t ++ ps, callerSavedRegisters) @@ -264,10 +265,11 @@ patchRegsOfInstr instr env = case instr of ORR o1 o2 o3 -> ORR (patchOp o1) (patchOp o2) (patchOp o3) -- 4. Branch Instructions -------------------------------------------------- - J t -> J (patchTarget t) - B t -> B (patchTarget t) - BL t rs -> BL (patchTarget t) rs - BCOND c t -> BCOND c (patchTarget t) + J t -> J (patchTarget t) + J_TBL ids mbLbl t -> J_TBL ids mbLbl (env t) + B t -> B (patchTarget t) + BL t rs -> BL (patchTarget t) rs + BCOND c t -> BCOND c (patchTarget t) -- 5. Atomic Instructions -------------------------------------------------- -- 6. Conditional Instructions --------------------------------------------- @@ -319,6 +321,7 @@ isJumpishInstr instr = case instr of CBZ{} -> True CBNZ{} -> True J{} -> True + J_TBL{} -> True B{} -> True BL{} -> True BCOND{} -> True @@ -332,6 +335,7 @@ jumpDestsOfInstr (ANN _ i) = jumpDestsOfInstr i jumpDestsOfInstr (CBZ _ t) = [ id | TBlock id <- [t]] jumpDestsOfInstr (CBNZ _ t) = [ id | TBlock id <- [t]] jumpDestsOfInstr (J t) = [id | TBlock id <- [t]] +jumpDestsOfInstr (J_TBL ids _mbLbl _r) = catMaybes ids jumpDestsOfInstr (B t) = [id | TBlock id <- [t]] jumpDestsOfInstr (BL t _) = [ id | TBlock id <- [t]] jumpDestsOfInstr (BCOND _ t) = [ id | TBlock id <- [t]] @@ -340,6 +344,11 @@ jumpDestsOfInstr _ = [] canFallthroughTo :: Instr -> BlockId -> Bool canFallthroughTo (ANN _ i) bid = canFallthroughTo i bid canFallthroughTo (J (TBlock target)) bid = bid == target +canFallthroughTo (J_TBL targets _ _) bid = all isTargetBid targets + where + isTargetBid target = case target of + Nothing -> True + Just target -> target == bid canFallthroughTo (B (TBlock target)) bid = bid == target canFallthroughTo _ _ = False @@ -353,6 +362,7 @@ patchJumpInstr instr patchF CBZ r (TBlock bid) -> CBZ r (TBlock (patchF bid)) CBNZ r (TBlock bid) -> CBNZ r (TBlock (patchF bid)) J (TBlock bid) -> J (TBlock (patchF bid)) + J_TBL ids mbLbl r -> J_TBL (map (fmap patchF) ids) mbLbl r B (TBlock bid) -> B (TBlock (patchF bid)) BL (TBlock bid) ps -> BL (TBlock (patchF bid)) ps BCOND c (TBlock bid) -> BCOND c (TBlock (patchF bid)) @@ -516,6 +526,7 @@ allocMoreStack platform slots proc@(CmmProc info lbl live (ListGraph code)) = do insert_dealloc insn r = case insn of J _ -> dealloc ++ (insn : r) + J_TBL {} -> dealloc ++ (insn : r) ANN _ (J _) -> dealloc ++ (insn : r) _other | jumpDestsOfInstr insn /= [] -> patchJumpInstr insn retarget : r @@ -644,6 +655,7 @@ data Instr | CBNZ Operand Target -- if op /= 0, then branch. -- Branching. | J Target -- like B, but only generated from genJump. Used to distinguish genJumps from others. + | J_TBL [Maybe BlockId] (Maybe CLabel) Reg -- A jump instruction with data for switch/jump tables | B Target -- unconditional branching b/br. (To a blockid, label or register) | BL Target [Reg] -- branch and link (e.g. set x30 to next pc, and branch) | BCOND Cond Target -- branch with condition. b. @@ -730,6 +742,7 @@ instrCon i = CBZ{} -> "CBZ" CBNZ{} -> "CBNZ" J{} -> "J" + J_TBL {} -> "J_TBL" B{} -> "B" BL{} -> "BL" BCOND{} -> "BCOND" ===================================== compiler/GHC/CmmToAsm/AArch64/Ppr.hs ===================================== @@ -426,6 +426,7 @@ pprInstr platform instr = case instr of -- 4. Branch Instructions ---------------------------------------------------- J t -> pprInstr platform (B t) + J_TBL _ _ r -> pprInstr platform (B (TReg r)) B (TBlock bid) -> line $ text "\tb" <+> pprAsmLabel platform (mkLocalBlockLabel (getUnique bid)) B (TLabel lbl) -> line $ text "\tb" <+> pprAsmLabel platform lbl B (TReg r) -> line $ text "\tbr" <+> pprReg W64 r View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c67216918cd9c31391e3b30ba2ddbd50e0ab6958 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c67216918cd9c31391e3b30ba2ddbd50e0ab6958 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 17:19:49 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sun, 08 Sep 2024 13:19:49 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/supersven/strict-instr-data Message-ID: <66dddcb510137_2c027e5f864c357a4@gitlab.mail> Sven Tennie pushed new branch wip/supersven/strict-instr-data at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/supersven/strict-instr-data You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 18:19:12 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sun, 08 Sep 2024 14:19:12 -0400 Subject: [Git][ghc/ghc][master] AArch64: Implement takeRegRegMoveInstr Message-ID: <66ddeaa0ce2f8_2c027e8661b858222@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 1 changed file: - compiler/GHC/CmmToAsm/AArch64/Instr.hs Changes: ===================================== compiler/GHC/CmmToAsm/AArch64/Instr.hs ===================================== @@ -30,6 +30,7 @@ import GHC.Utils.Panic import Data.Maybe (fromMaybe) import GHC.Stack +import GHC.Platform.Reg.Class -- | LR and FP (8 byte each) are the prologue of each stack frame stackFrameHeaderSize :: Int @@ -454,10 +455,22 @@ isMetaInstr instr mkRegRegMoveInstr :: Reg -> Reg -> Instr mkRegRegMoveInstr src dst = ANN (text "Reg->Reg Move: " <> ppr src <> text " -> " <> ppr dst) $ MOV (OpReg W64 dst) (OpReg W64 src) --- | Take the source and destination from this reg -> reg move instruction --- or Nothing if it's not one +-- | Take the source and destination registers from a move instruction of same +-- register class (`RegClass`). +-- +-- The idea is to identify moves that can be eliminated by the register +-- allocator: If the source register serves no special purpose, one could +-- continue using it; saving one move instruction. For this, the register kinds +-- (classes) must be the same (no conversion involved.) takeRegRegMoveInstr :: Instr -> Maybe (Reg,Reg) ---takeRegRegMoveInstr (MOV (OpReg fmt dst) (OpReg fmt' src)) | fmt == fmt' = Just (src, dst) +takeRegRegMoveInstr (MOV (OpReg _fmt dst) (OpReg _fmt' src)) + | classOfReg dst == classOfReg src = pure (src, dst) + where + classOfReg ::Reg -> RegClass + classOfReg reg + = case reg of + RegVirtual vr -> classOfVirtualReg vr + RegReal rr -> classOfRealReg rr takeRegRegMoveInstr _ = Nothing -- | Make an unconditional jump instruction. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/573f9833a7c56bf8fd8c6731069a7160541e5287 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/573f9833a7c56bf8fd8c6731069a7160541e5287 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 8 18:19:47 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sun, 08 Sep 2024 14:19:47 -0400 Subject: [Git][ghc/ghc][master] haddock: Configuration fix for ReadTheDocs Message-ID: <66ddeac316df6_2c027e88ff9061328@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 1 changed file: - utils/haddock/.readthedocs.yaml Changes: ===================================== utils/haddock/.readthedocs.yaml ===================================== @@ -2,7 +2,7 @@ version: 2 sphinx: - configuration: doc/conf.py + configuration: utils/haddock/doc/conf.py build: os: "ubuntu-22.04" @@ -11,4 +11,4 @@ build: python: install: - - requirements: doc/requirements.txt + - requirements: utils/haddock/doc/requirements.txt View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/20b0de7d2cecca20f5d369058906a7bcf9b03662 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/20b0de7d2cecca20f5d369058906a7bcf9b03662 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 04:41:52 2024 From: gitlab at gitlab.haskell.org (Apoorv Ingle (@ani)) Date: Mon, 09 Sep 2024 00:41:52 -0400 Subject: [Git][ghc/ghc][wip/expansions-appdo] 85 commits: TTG HsCmdArrForm: use Fixity via extension point Message-ID: <66de7c901055b_36584c113e6a0886b2@gitlab.mail> Apoorv Ingle pushed to branch wip/expansions-appdo at Glasgow Haskell Compiler / GHC Commits: d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 5be55c1f by Apoorv Ingle at 2024-09-08T18:53:58-05:00 Make ApplicativeDo work with HsExpansions testcase added: T24406 Issues Fixed: #24406, #16135 Code Changes: - Remove `XStmtLR GhcTc` as `XStmtLR GhcRn` is now compiled to `HsExpr GhcTc` - The expanded statements are guided by `GHC.Hs.Expr.TcFunInfo` which is used to decide if the `XExpr GhcRn` is to be typechecked using `tcApp` or `tcExpr` Note [Expanding HsDo with XXExprGhcRn] explains the change in more detail - - - - - 40f4ea3f by Apoorv Ingle at 2024-09-08T23:41:06-05:00 simplify data structures. remove doTcApp and applicative stmt fail blocks do not refer stmts - - - - - 30 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/GenericOpt.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Reg/Linear/State.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Predicate.hs - compiler/GHC/Core/Type.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Config/Finder.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Errors/Types.hs - compiler/GHC/Driver/Flags.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c533d25fb9c0b4cc70668794f97a2a367f60fabc...40f4ea3f4f75cc30c11ea0961c04b5bbdc50ebe8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c533d25fb9c0b4cc70668794f97a2a367f60fabc...40f4ea3f4f75cc30c11ea0961c04b5bbdc50ebe8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 04:46:28 2024 From: gitlab at gitlab.haskell.org (Apoorv Ingle (@ani)) Date: Mon, 09 Sep 2024 00:46:28 -0400 Subject: [Git][ghc/ghc][wip/expansions-appdo] simplify data structures. remove doTcApp and applicative stmt fail blocks do not refer stmts Message-ID: <66de7da484332_262686952cc4646f@gitlab.mail> Apoorv Ingle pushed to branch wip/expansions-appdo at Glasgow Haskell Compiler / GHC Commits: 1f024c7e by Apoorv Ingle at 2024-09-08T23:46:08-05:00 simplify data structures. remove doTcApp and applicative stmt fail blocks do not refer stmts - - - - - 8 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Tc/Gen/App.hs - compiler/GHC/Tc/Gen/Do.hs - compiler/GHC/Tc/Gen/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Types/Origin.hs - testsuite/tests/ghci.debugger/scripts/break029.script Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -496,7 +496,6 @@ data HsThingRn = OrigExpr (HsExpr GhcRn) -- ^ The source, user wr | OrigStmt (ExprLStmt GhcRn) HsDoFlavour -- ^ which kind of do-block did this statement come from | OrigPat (LPat GhcRn) -- ^ The source, user written, pattern HsDoFlavour -- ^ which kind of do-block did this statement come from - (Maybe (ExprLStmt GhcRn)) -- ^ Optional statement binding this pattern isHsThingRnExpr, isHsThingRnStmt, isHsThingRnPat :: HsThingRn -> Bool isHsThingRnExpr (OrigExpr{}) = True @@ -511,9 +510,7 @@ isHsThingRnPat _ = False data XXExprGhcRn = ExpandedThingRn { xrn_orig :: HsThingRn -- The original source thing to be used for error messages , xrn_expanded :: HsExpr GhcRn -- The compiler generated expanded thing - , xrn_doTcApp :: Bool } -- A Hint to the type checker of how to proceed - -- True <=> use GHC.Tc.Gen.Expr.tcApp on xrn_expanded - -- False <=> use GHC.Tc.Gen.Expr.tcExpr on xrn_expanded + } | PopErrCtxt -- A hint for typechecker to pop {-# UNPACK #-} !(LHsExpr GhcRn) -- the top of the error context stack @@ -538,8 +535,7 @@ mkExpandedExpr -> HsExpr GhcRn -- ^ expanded expression -> HsExpr GhcRn -- ^ suitably wrapped 'XXExprGhcRn' mkExpandedExpr oExpr eExpr = XExpr (ExpandedThingRn { xrn_orig = OrigExpr oExpr - , xrn_expanded = eExpr - , xrn_doTcApp = False }) + , xrn_expanded = eExpr }) -- | Build an expression using the extension constructor `XExpr`, -- and the two components of the expansion: original do stmt and @@ -547,22 +543,18 @@ mkExpandedExpr oExpr eExpr = XExpr (ExpandedThingRn { xrn_orig = OrigExpr oExpr mkExpandedStmt :: ExprLStmt GhcRn -- ^ source statement -> HsDoFlavour -- ^ source statement do flavour - -> Bool -- ^ should this be type checked using tcApp? -> HsExpr GhcRn -- ^ expanded expression -> HsExpr GhcRn -- ^ suitably wrapped 'XXExprGhcRn' -mkExpandedStmt oStmt flav doTcApp eExpr = XExpr (ExpandedThingRn { xrn_orig = OrigStmt oStmt flav - , xrn_expanded = eExpr - , xrn_doTcApp = doTcApp}) +mkExpandedStmt oStmt flav eExpr = XExpr (ExpandedThingRn { xrn_orig = OrigStmt oStmt flav + , xrn_expanded = eExpr }) mkExpandedPatRn :: LPat GhcRn -- ^ source pattern -> HsDoFlavour -- ^ source statement do flavour - -> Maybe (ExprLStmt GhcRn) -- ^ pattern statement origin -> HsExpr GhcRn -- ^ expanded expression -> HsExpr GhcRn -- ^ suitably wrapped 'XXExprGhcRn' -mkExpandedPatRn oPat flav mb_stmt eExpr = XExpr (ExpandedThingRn { xrn_orig = OrigPat oPat flav mb_stmt - , xrn_expanded = eExpr - , xrn_doTcApp = False}) +mkExpandedPatRn oPat flav eExpr = XExpr (ExpandedThingRn { xrn_orig = OrigPat oPat flav + , xrn_expanded = eExpr }) -- | Build an expression using the extension constructor `XExpr`, -- and the two components of the expansion: original do stmt and @@ -572,14 +564,13 @@ mkExpandedStmtAt -> SrcSpanAnnA -- ^ Location for the expansion expression -> ExprLStmt GhcRn -- ^ source statement -> HsDoFlavour -- ^ the flavour of the statement - -> Bool -- ^ should type check with tcApp? -> HsExpr GhcRn -- ^ expanded expression -> LHsExpr GhcRn -- ^ suitably wrapped located 'XXExprGhcRn' -mkExpandedStmtAt addPop loc oStmt flav doTcApp eExpr +mkExpandedStmtAt addPop loc oStmt flav eExpr | addPop - = mkPopErrCtxtExprAt loc (L loc $ mkExpandedStmt oStmt flav doTcApp eExpr) + = mkPopErrCtxtExprAt loc (L loc $ mkExpandedStmt oStmt flav eExpr) | otherwise - = L loc $ mkExpandedStmt oStmt flav doTcApp eExpr + = L loc $ mkExpandedStmt oStmt flav eExpr data XXExprGhcTc = WrapExpr -- Type and evidence application and abstractions @@ -888,12 +879,12 @@ instance Outputable HsThingRn where = case thing of OrigExpr x -> ppr_builder ":" x OrigStmt x _ -> ppr_builder ":" x - OrigPat x _ mb_stmt -> ifPprDebug (braces (text "" <+> parens (ppr x) <+> parens (ppr mb_stmt))) (ppr x) + OrigPat x _ -> ifPprDebug (braces (text "" <+> parens (ppr x))) (ppr x) where ppr_builder prefix x = ifPprDebug (braces (text prefix <+> parens (ppr x))) (ppr x) instance Outputable XXExprGhcRn where - ppr (ExpandedThingRn o e _) = ifPprDebug (braces $ vcat [ppr o, text ";;" , ppr e]) (ppr o) - ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) + ppr (ExpandedThingRn o e) = ifPprDebug (braces $ vcat [ppr o, text ";;" , ppr e]) (ppr o) + ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) instance Outputable XXExprGhcTc where ppr (WrapExpr (HsWrap co_fn e)) @@ -933,7 +924,7 @@ ppr_infix_expr (XExpr x) = case ghcPass @p of ppr_infix_expr _ = Nothing ppr_infix_expr_rn :: XXExprGhcRn -> Maybe SDoc -ppr_infix_expr_rn (ExpandedThingRn thing _ _) = ppr_infix_hs_expansion thing +ppr_infix_expr_rn (ExpandedThingRn thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a ppr_infix_expr_tc :: XXExprGhcTc -> Maybe SDoc @@ -1047,7 +1038,7 @@ hsExprNeedsParens prec = go go_x_tc (HsBinTick _ _ (L _ e)) = hsExprNeedsParens prec e go_x_rn :: XXExprGhcRn -> Bool - go_x_rn (ExpandedThingRn thing _ _) = hsExpandedNeedsParens thing + go_x_rn (ExpandedThingRn thing _ ) = hsExpandedNeedsParens thing go_x_rn (PopErrCtxt (L _ a)) = hsExprNeedsParens prec a hsExpandedNeedsParens :: HsThingRn -> Bool @@ -1099,7 +1090,7 @@ isAtomicHsExpr (XExpr x) go_x_tc (HsBinTick {}) = False go_x_rn :: XXExprGhcRn -> Bool - go_x_rn (ExpandedThingRn thing _ _) = isAtomicExpandedThingRn thing + go_x_rn (ExpandedThingRn thing _) = isAtomicExpandedThingRn thing go_x_rn (PopErrCtxt (L _ a)) = isAtomicHsExpr a isAtomicExpandedThingRn :: HsThingRn -> Bool ===================================== compiler/GHC/HsToCore/Quote.hs ===================================== @@ -1684,7 +1684,6 @@ repE (HsProjection _ xs) = repProjection (fmap (field_label . unLoc . dfoLabel . repE (HsEmbTy _ t) = do t1 <- repLTy (hswc_body t) rep2 typeEName [unC t1] - repE (HsQual _ (L _ ctx) body) = do ctx' <- repLEs ctx body' <- repLE body @@ -1704,7 +1703,7 @@ repE (HsFunArr _ mult arg res) = do arg' <- repLE arg res' <- repLE res repApps fun [arg', res'] -repE e@(XExpr (ExpandedThingRn o x _)) +repE e@(XExpr (ExpandedThingRn o x)) | OrigExpr e <- o = do { rebindable_on <- lift $ xoptM LangExt.RebindableSyntax ; if rebindable_on -- See Note [Quotation and rebindable syntax] ===================================== compiler/GHC/Tc/Gen/App.hs ===================================== @@ -651,7 +651,7 @@ tcInstFun do_ql inst_final (tc_fun, fun_ctxt) fun_sigma rn_args where fun_orig = case fun_ctxt of VAExpansion (OrigStmt{}) _ _ -> DoOrigin - VAExpansion (OrigPat pat _ _) _ _ -> DoPatOrigin pat + VAExpansion (OrigPat pat _) _ _ -> DoPatOrigin pat VAExpansion (OrigExpr e) _ _ -> exprCtOrigin e VACall e _ _ -> exprCtOrigin e @@ -912,6 +912,8 @@ addArgCtxt ctxt (L arg_loc arg) thing_inside -> setSrcSpanA arg_loc $ addStmtCtxt stmt flav $ thing_inside + VAExpansion (OrigStmt (L _ (XStmtLR (ApplicativeStmt{}))) _) _ _ + -> thing_inside VAExpansion (OrigStmt (L loc stmt) flav) _ _ -> setSrcSpanA loc $ addStmtCtxt stmt flav $ @@ -1062,7 +1064,7 @@ expr_to_type earg = | otherwise = not_in_scope where occ = occName rdr not_in_scope = failWith $ mkTcRnNotInScope rdr NotInScope - go (L l (XExpr (ExpandedThingRn (OrigExpr orig) _ _))) = + go (L l (XExpr (ExpandedThingRn (OrigExpr orig) _))) = -- Use the original, user-written expression (before expansion). -- Example. Say we have vfun :: forall a -> blah -- and the call vfun (Maybe [1,2,3]) ===================================== compiler/GHC/Tc/Gen/Do.hs ===================================== @@ -79,7 +79,7 @@ expand_do_stmts addPop flav [stmt@(L loc (LastStmt _ (L body_loc body) _ ret_exp -- See `checkLastStmt` and `Syntax.Expr.StmtLR.LastStmt` | NoSyntaxExprRn <- ret_expr -- Last statement is just body if we are not in ListComp context. See Syntax.Expr.LastStmt - = return $ mkExpandedStmtAt addPop loc stmt flav False body + = return $ mkExpandedStmtAt addPop loc stmt flav body | SyntaxExprRn ret <- ret_expr -- @@ -87,7 +87,7 @@ expand_do_stmts addPop flav [stmt@(L loc (LastStmt _ (L body_loc body) _ ret_exp -- return e ~~> return e -- to make T18324 work = do let expansion = genHsApp ret (L body_loc body) - return $ mkExpandedStmtAt addPop loc stmt flav False expansion + return $ mkExpandedStmtAt addPop loc stmt flav expansion expand_do_stmts addPop doFlavour (stmt@(L loc (LetStmt _ bs)) : lstmts) = -- See Note [Expanding HsDo with XXExprGhcRn] Equation (3) below @@ -96,7 +96,7 @@ expand_do_stmts addPop doFlavour (stmt@(L loc (LetStmt _ bs)) : lstmts) = -- let x = e ; stmts ~~> let x = e in stmts' do expand_stmts <- expand_do_stmts True doFlavour lstmts let expansion = genHsLet bs expand_stmts - return $ mkExpandedStmtAt addPop loc stmt doFlavour False expansion + return $ mkExpandedStmtAt addPop loc stmt doFlavour expansion expand_do_stmts addPop doFlavour (stmt@(L loc (BindStmt xbsrn pat e)): lstmts) | SyntaxExprRn bind_op <- xbsrn_bindOp xbsrn @@ -108,11 +108,11 @@ expand_do_stmts addPop doFlavour (stmt@(L loc (BindStmt xbsrn pat e)): lstmts) -- ------------------------------------------------------- -- pat <- e ; stmts ~~> (>>=) e f = do expand_stmts <- expand_do_stmts True doFlavour lstmts - failable_expr <- mk_failable_expr False doFlavour Nothing pat expand_stmts fail_op + failable_expr <- mk_failable_expr False doFlavour pat expand_stmts fail_op let expansion = genHsExpApps bind_op -- (>>=) [ e , failable_expr ] - return $ mkExpandedStmtAt addPop loc stmt doFlavour True expansion + return $ mkExpandedStmtAt addPop loc stmt doFlavour expansion | otherwise = pprPanic "expand_do_stmts: The impossible happened, missing bind operator from renamer" (text "stmt" <+> ppr stmt) @@ -127,7 +127,7 @@ expand_do_stmts addPop doFlavour (stmt@(L loc (BodyStmt _ e (SyntaxExprRn then_o let expansion = genHsExpApps then_op -- (>>) [ e , expand_stmts_expr ] - return $ mkExpandedStmtAt addPop loc stmt doFlavour True expansion + return $ mkExpandedStmtAt addPop loc stmt doFlavour expansion expand_do_stmts _ doFlavour ((L loc (RecStmt { recS_stmts = L stmts_loc rec_stmts @@ -193,7 +193,7 @@ expand_do_stmts addPop doFlavour ((L _ (XStmtLR (ApplicativeStmt _ args mb_join) ; (pats_can_fail, rhss) <- unzip <$> mapM (do_arg . snd) args -- add blocks for failable patterns - ; body_with_fails <- foldrM match_args xexpr (zip pats_can_fail rhss) + ; body_with_fails <- foldrM match_args xexpr pats_can_fail -- builds (((body <$> e1) <*> e2) ...) ; let expand_ado_expr = foldl mk_apps body_with_fails (zip (map fst args) rhss) @@ -216,7 +216,7 @@ expand_do_stmts addPop doFlavour ((L _ (XStmtLR (ApplicativeStmt _ args mb_join) { xarg_app_arg_one = mb_fail_op , app_arg_pattern = pat , arg_expr = (L rhs_loc rhs) }) = - do let xx_expr = mkExpandedStmtAt addPop (noAnnSrcSpan generatedSrcSpan) stmt doFlavour False rhs + do let xx_expr = mkExpandedStmtAt addPop (noAnnSrcSpan generatedSrcSpan) stmt doFlavour rhs traceTc "do_arg" (text "OneArg" <+> vcat [ppr pat, ppr xx_expr]) return ((pat, mb_fail_op) , xx_expr) @@ -230,12 +230,8 @@ expand_do_stmts addPop doFlavour ((L _ (XStmtLR (ApplicativeStmt _ args mb_join) ; return ((pat, Nothing) , xx_expr) } - match_args :: ((LPat GhcRn, FailOperator GhcRn), LHsExpr GhcRn) -> LHsExpr GhcRn -> TcM (LHsExpr GhcRn) - match_args ((pat, fail_op), stmt_expr) body = mk_failable_expr addPop doFlavour mb_stmt pat body fail_op - where mb_stmt = case unLoc stmt_expr of - XExpr (ExpandedThingRn (OrigStmt s _) _ _) -> Just s - XExpr (PopErrCtxt (L _ (XExpr (ExpandedThingRn (OrigStmt s _) _ _)))) -> Just s - _ -> Nothing + match_args :: (LPat GhcRn, FailOperator GhcRn) -> LHsExpr GhcRn -> TcM (LHsExpr GhcRn) + match_args (pat, fail_op) body = mk_failable_expr addPop doFlavour pat body fail_op mk_apps :: LHsExpr GhcRn -> (SyntaxExprRn, LHsExpr GhcRn) -> LHsExpr GhcRn mk_apps l_expr (op, r_expr) = @@ -246,8 +242,8 @@ expand_do_stmts addPop doFlavour ((L _ (XStmtLR (ApplicativeStmt _ args mb_join) expand_do_stmts _ _ stmts = pprPanic "expand_do_stmts: impossible happened" $ (ppr stmts) -- checks the pattern `pat` for irrefutability which decides if we need to wrap it with a fail block -mk_failable_expr :: Bool -> HsDoFlavour -> Maybe (ExprLStmt GhcRn) -> LPat GhcRn -> LHsExpr GhcRn -> FailOperator GhcRn -> TcM (LHsExpr GhcRn) -mk_failable_expr addPop doFlav mb_stmt lpat@(L loc pat) expr fail_op = +mk_failable_expr :: Bool -> HsDoFlavour -> LPat GhcRn -> LHsExpr GhcRn -> FailOperator GhcRn -> TcM (LHsExpr GhcRn) +mk_failable_expr addPop doFlav lpat@(L loc pat) expr@(L exprloc _) fail_op = do { is_strict <- xoptM LangExt.Strict ; hscEnv <- getTopEnv ; rdrEnv <- getGlobalRdrEnv @@ -256,23 +252,21 @@ mk_failable_expr addPop doFlav mb_stmt lpat@(L loc pat) expr fail_op = ; traceTc "mk_failable_expr" (vcat [ text "pat:" <+> ppr pat , text "isIrrefutable:" <+> ppr irrf_pat ]) - + ; let xexpr | addPop = mkPopErrCtxtExprAt exprloc expr + | otherwise = expr ; if irrf_pat -- don't wrap with fail block if -- the pattern is irrefutable then case pat of - (WildPat{}) -> return $ genHsLamDoExp doFlav [L noSrcSpanA pat] expr - _ -> return $ case mb_stmt of - Nothing -> genHsLamDoExp doFlav [lpat] expr - Just s -> mkExpandedStmtAt addPop (noAnnSrcSpan generatedSrcSpan) s doFlav False - (unLoc $ (genHsLamDoExp doFlav [lpat] - $ wrapGenSpan (mkPopErrCtxtExpr expr))) - else L loc <$> mk_fail_block doFlav mb_stmt lpat expr fail_op + (WildPat{}) -> return $ genHsLamDoExp doFlav [L noSrcSpanA pat] xexpr + _ -> return $ genHsLamDoExp doFlav [lpat] xexpr + + else L loc <$> mk_fail_block doFlav lpat expr fail_op } -- makes the fail block with a given fail_op -mk_fail_block :: HsDoFlavour -> Maybe (ExprLStmt GhcRn) +mk_fail_block :: HsDoFlavour -> LPat GhcRn -> LHsExpr GhcRn -> FailOperator GhcRn -> TcM (HsExpr GhcRn) -mk_fail_block doFlav mb_stmt pat@(L ploc _) e (Just (SyntaxExprRn fail_op)) = +mk_fail_block doFlav pat@(L ploc _) e (Just (SyntaxExprRn fail_op)) = do dflags <- getDynFlags return $ HsLam noAnn LamCases $ mkMatchGroup (doExpansionOrigin doFlav) -- \ (wrapGenSpan [ genHsCaseAltDoExp doFlav pat e -- pat -> expr @@ -285,7 +279,7 @@ mk_fail_block doFlav mb_stmt pat@(L ploc _) e (Just (SyntaxExprRn fail_op)) = fail_op_expr :: DynFlags -> LPat GhcRn -> HsExpr GhcRn -> HsExpr GhcRn fail_op_expr dflags pat fail_op - = mkExpandedPatRn pat doFlav mb_stmt $ + = mkExpandedPatRn pat doFlav $ genHsApp fail_op (mk_fail_msg_expr dflags pat) mk_fail_msg_expr :: DynFlags -> LPat GhcRn -> LHsExpr GhcRn @@ -295,7 +289,7 @@ mk_fail_block doFlav mb_stmt pat@(L ploc _) e (Just (SyntaxExprRn fail_op)) = <+> text "at" <+> ppr (getLocA pat) -mk_fail_block _ _ _ _ _ = pprPanic "mk_fail_block: impossible happened" empty +mk_fail_block _ _ _ _ = pprPanic "mk_fail_block: impossible happened" empty {- Note [Expanding HsDo with XXExprGhcRn] ===================================== compiler/GHC/Tc/Gen/Expr.hs ===================================== @@ -717,15 +717,10 @@ tcXExpr (PopErrCtxt (L loc e)) res_ty setSrcSpanA loc $ tcExpr e res_ty -tcXExpr xe@(ExpandedThingRn o e' doTcApp) res_ty - | OrigPat (L loc _) flav (Just s) <- o -- testcase T16628 - = setSrcSpanA loc $ - addStmtCtxt (unLoc s) flav $ - tcApp (XExpr xe) res_ty - +tcXExpr xe@(ExpandedThingRn o e') res_ty | OrigStmt ls@(L loc s) flav <- o , HsLet x binds e <- e' - = do { (binds', wrapper, e') <- setSrcSpanA loc $ + = do { (binds', wrapper, e') <- setSrcSpanA loc $ addStmtCtxt s flav $ tcLocalBinds binds $ tcMonoExprNC e res_ty -- NB: Do not call tcMonoExpr here as it adds @@ -733,22 +728,14 @@ tcXExpr xe@(ExpandedThingRn o e' doTcApp) res_ty ; return $ mkExpandedStmtTc ls flav (HsLet x binds' (mkLHsWrap wrapper e')) } - | OrigStmt ls@(L loc _) flav <- o - , doTcApp + | OrigStmt s@(L loc LastStmt{}) flav <- o = setSrcSpanA loc $ - mkExpandedStmtTc ls flav <$> tcApp (XExpr xe) res_ty + addStmtCtxt (unLoc s) flav $ + mkExpandedStmtTc s flav <$> tcApp e' res_ty - -- There are currently 2 `do`-statements that require calling `tcExpr` and not `tcApp`: - -- `LastStmt`, `AppStmt` - -- The reason is that the expanded expression `e` is the last statement's body expression - -- (or the the argument expression of an applicative statement) - -- It is not an HsApp of a generated (>>) or (>>=) - -- This improves error messages e.g. tests: DoExpansion1, DoExpansion2, DoExpansion3, ado002 etc. - | OrigStmt ls@(L loc s) flav <- o - , not doTcApp + | OrigStmt ls@(L loc _) flav <- o = setSrcSpanA loc $ - addStmtCtxt s flav $ - mkExpandedStmtTc ls flav <$> tcExpr e' res_ty + mkExpandedStmtTc ls flav <$> tcApp (XExpr xe) res_ty tcXExpr xe res_ty = tcApp (XExpr xe) res_ty ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -300,7 +300,7 @@ splitHsApps e = go e (top_ctxt 0 e) [] top_ctxt n (HsPragE _ _ fun) = top_lctxt n fun top_ctxt n (HsAppType _ fun _) = top_lctxt (n+1) fun top_ctxt n (HsApp _ fun _) = top_lctxt (n+1) fun - top_ctxt n (XExpr (ExpandedThingRn (OrigExpr fun) _ _)) + top_ctxt n (XExpr (ExpandedThingRn (OrigExpr fun) _)) = VACall fun n noSrcSpan top_ctxt n other_fun = VACall other_fun n noSrcSpan @@ -325,7 +325,7 @@ splitHsApps e = go e (top_ctxt 0 e) [] HsQuasiQuote _ _ (L l _) -> set l ctxt -- l :: SrcAnn NoEpAnns -- See Note [Looking through ExpandedThingRn] - go (XExpr (ExpandedThingRn o e _)) ctxt args + go (XExpr (ExpandedThingRn o e)) ctxt args = go e (VAExpansion o (appCtxtLoc ctxt) (appCtxtLoc ctxt)) (EWrap (EExpand o) : args) @@ -567,8 +567,6 @@ addHeadCtxt fun_ctxt thing_inside do case fun_ctxt of VAExpansion (OrigExpr orig) _ _ -> addExprCtxt orig thing_inside - VAExpansion (OrigPat _ flav (Just (L loc stmt))) _ _ - -> setSrcSpanA loc $ addStmtCtxt stmt flav thing_inside _ -> thing_inside where fun_loc = appCtxtLoc fun_ctxt @@ -1268,7 +1266,7 @@ addExprCtxt e thing_inside = case e of HsUnboundVar {} -> thing_inside XExpr (PopErrCtxt (L _ e)) -> addExprCtxt e $ thing_inside - XExpr (ExpandedThingRn (OrigStmt stmt flav) _ _) -> addStmtCtxt (unLoc stmt) flav thing_inside + XExpr (ExpandedThingRn (OrigStmt stmt flav) _) -> addStmtCtxt (unLoc stmt) flav thing_inside _ -> addErrCtxt (exprCtxt e) thing_inside -- The HsUnboundVar special case addresses situations like -- f x = _ ===================================== compiler/GHC/Tc/Types/Origin.hs ===================================== @@ -755,9 +755,9 @@ exprCtOrigin (HsEmbTy {}) = Shouldn'tHappenOrigin "type expression" exprCtOrigin (HsForAll {}) = Shouldn'tHappenOrigin "forall telescope" -- See Note [Types in terms] exprCtOrigin (HsQual {}) = Shouldn'tHappenOrigin "constraint context" -- See Note [Types in terms] exprCtOrigin (HsFunArr {}) = Shouldn'tHappenOrigin "function arrow" -- See Note [Types in terms] -exprCtOrigin (XExpr (ExpandedThingRn thing _ _)) | OrigExpr a <- thing = exprCtOrigin a - | OrigStmt _ _ <- thing = DoOrigin - | OrigPat p _ _ <- thing = DoPatOrigin p +exprCtOrigin (XExpr (ExpandedThingRn thing _)) | OrigExpr a <- thing = exprCtOrigin a + | OrigStmt _ _ <- thing = DoOrigin + | OrigPat p _ <- thing = DoPatOrigin p exprCtOrigin (XExpr (PopErrCtxt {})) = Shouldn'tHappenOrigin "PopErrCtxt" -- | Extract a suitable CtOrigin from a MatchGroup ===================================== testsuite/tests/ghci.debugger/scripts/break029.script ===================================== @@ -1,5 +1,4 @@ :load break029.hs :step f 3 :step -:step y View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1f024c7ebc2988cdcf3867fed29251a36faf6d88 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1f024c7ebc2988cdcf3867fed29251a36faf6d88 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 06:45:50 2024 From: gitlab at gitlab.haskell.org (Sylvain Henry (@hsyl20)) Date: Mon, 09 Sep 2024 02:45:50 -0400 Subject: [Git][ghc/ghc][wip/libffi-dep-hadrian] 83 commits: Deriving-via one-shot strict state Monad instances Message-ID: <66de999e71f93_2626866ee3446514b@gitlab.mail> Sylvain Henry pushed to branch wip/libffi-dep-hadrian at Glasgow Haskell Compiler / GHC Commits: 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 56bb2734 by Matthew Pickering at 2024-09-09T08:36:57+02:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 - - - - - 5cddf22a by Sylvain Henry at 2024-09-09T08:45:25+02:00 Try fixing Wasm - - - - - 30 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/GenericOpt.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Reg/Linear/State.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Predicate.hs - compiler/GHC/Core/Type.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Config/Finder.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Errors/Types.hs - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/Main.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/97d33100091dd10b81b63bd278cf23ac9dc95f49...5cddf22a03bb381a645d3ca63fab07a309311ffe -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/97d33100091dd10b81b63bd278cf23ac9dc95f49...5cddf22a03bb381a645d3ca63fab07a309311ffe You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 06:57:12 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 09 Sep 2024 02:57:12 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 12 commits: haddock: Configuration fix for ReadTheDocs Message-ID: <66de9c48bfec2_2626866a7cf0682f3@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 1e19bd13 by Sven Tennie at 2024-09-09T02:56:53-04:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 81bb0113 by Sven Tennie at 2024-09-09T02:56:53-04:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - 87456694 by Sven Tennie at 2024-09-09T02:56:53-04:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - e9f863f9 by Sven Tennie at 2024-09-09T02:56:53-04:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - b0aabf62 by Sven Tennie at 2024-09-09T02:56:53-04:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 2d6832d4 by Sven Tennie at 2024-09-09T02:56:53-04:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 5bcb442a by Sven Tennie at 2024-09-09T02:56:53-04:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - ea993f0c by Sven Tennie at 2024-09-09T02:56:53-04:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - 0167c852 by Sven Tennie at 2024-09-09T02:56:53-04:00 RISCV64: Set codeowners of the NCG - - - - - 249caea7 by Sven Tennie at 2024-09-09T02:56:53-04:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. - - - - - d569f2d5 by Sylvain Henry at 2024-09-09T02:57:08-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 10 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/faf45b7b2153085e16fca4e9f39f61234a3a8e4b...d569f2d5a13dd82d6108842fe521688d6eff198b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/faf45b7b2153085e16fca4e9f39f61234a3a8e4b...d569f2d5a13dd82d6108842fe521688d6eff198b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 11:04:45 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 09 Sep 2024 07:04:45 -0400 Subject: [Git][ghc/ghc][wip/T25029] 22 commits: Haddock: Add no-compilation flag Message-ID: <66ded64dbc01b_8620e138d001074f0@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25029 at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 957a9cf0 by Simon Peyton Jones at 2024-09-09T10:59:57+01:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 0f79fc3f by Simon Peyton Jones at 2024-09-09T10:59:57+01:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 5d14c348 by Simon Peyton Jones at 2024-09-09T10:59:57+01:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/Core/InstEnv.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs - compiler/GHC/Stg/Unarise.hs - compiler/GHC/Stg/Utils.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9f06a23376822bcad98555800734d139fd2de226...5d14c348e928162b69c305023a62c16868843997 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9f06a23376822bcad98555800734d139fd2de226...5d14c348e928162b69c305023a62c16868843997 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 11:17:37 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 09 Sep 2024 07:17:37 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T24934 Message-ID: <66ded95144dc7_8620e4703d812032a@gitlab.mail> Simon Peyton Jones pushed new branch wip/T24934 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T24934 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 11:47:56 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 09 Sep 2024 07:47:56 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 2 commits: JS: fake support for native adjustors (#25159) Message-ID: <66dee06c25ef4_8620e8730ac1375ab@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 699a5041 by Sylvain Henry at 2024-09-09T07:47:47-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 38b0927b by Sylvain Henry at 2024-09-09T07:47:50-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - 3 changed files: - libraries/ghc-internal/jsbits/base.js - m4/ghc_adjustors_method.m4 - utils/ghc-toolchain/exe/Main.hs Changes: ===================================== libraries/ghc-internal/jsbits/base.js ===================================== @@ -367,23 +367,6 @@ function h$base_lstat(file, file_off, stat, stat_off, c) { h$unsupported(-1, c); } -function h$lstat(file, file_off, stat, stat_off) { - TRACE_IO("lstat") -#ifndef GHCJS_BROWSER - if(h$isNode()) { - try { - var fs = h$fs.lstatSync(h$decodeUtf8z(file, file_off)); - h$base_fillStat(fs, stat, stat_off); - return 0; - } catch(e) { - h$setErrno(e); - return -1; - } - } else -#endif - h$unsupported(-1); -} - function h$rmdir(file, file_off) { TRACE_IO("rmdir") #ifndef GHCJS_BROWSER ===================================== m4/ghc_adjustors_method.m4 ===================================== @@ -4,7 +4,7 @@ dnl Use libffi for adjustors? AC_DEFUN([GHC_ADJUSTORS_METHOD], [ case [$]{$1[Arch]} in - i386|x86_64) + i386|x86_64|javascript) # We have native adjustor support on these platforms HaveNativeAdjustor=yes ;; ===================================== utils/ghc-toolchain/exe/Main.hs ===================================== @@ -374,6 +374,7 @@ archHasNativeAdjustors :: Arch -> Bool archHasNativeAdjustors = \case ArchX86 -> True ArchX86_64 -> True + ArchJavaScript -> True _ -> False View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d569f2d5a13dd82d6108842fe521688d6eff198b...38b0927bd76f835abe3646450a2aa9116b5d5b9a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d569f2d5a13dd82d6108842fe521688d6eff198b...38b0927bd76f835abe3646450a2aa9116b5d5b9a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 12:29:35 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 09 Sep 2024 08:29:35 -0400 Subject: [Git][ghc/ghc][wip/T24817a] 55 commits: Extend -reexported-module flag to support module renaming Message-ID: <66deea2f9f166_89c3023e6f0257a8@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24817a at Glasgow Haskell Compiler / GHC Commits: ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 58843b90 by Simon Peyton Jones at 2024-09-09T11:18:36+01:00 Towards ZonkAnyTyCon fixing #24817 - - - - - f3ebc4eb by Simon Peyton Jones at 2024-09-09T11:18:36+01:00 Wibbles - - - - - adea172d by Simon Peyton Jones at 2024-09-09T11:18:36+01:00 Complete the patch ..still needs documentation - - - - - 9da177c4 by Simon Peyton Jones at 2024-09-09T13:28:03+01:00 Wibbles This accepts several test outputs that (sadly) mention `Any` in the error message. That's bad enough but now it's `ZonkAny`. But that just makes the infelicity worse; it does not introduce it. - - - - - 30 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Predicate.hs - compiler/GHC/Core/Type.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Config/Finder.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Errors/Types.hs - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/446045e785685c1ed46f131e62157ca05b3d0dfa...9da177c40a95f5035629047837a46d1d1b098d4f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/446045e785685c1ed46f131e62157ca05b3d0dfa...9da177c40a95f5035629047837a46d1d1b098d4f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 12:40:29 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 09 Sep 2024 08:40:29 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] 5 commits: Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Message-ID: <66deecbd438b5_89c3023e948304f4@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 6c1a78cb by Cheng Shao at 2024-09-09T14:40:23+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). Metric Decrease: MultiLayerModules T13701 - - - - - 30 changed files: - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs - compiler/GHC/Stg/Unarise.hs - compiler/GHC/Stg/Utils.hs - compiler/GHC/StgToCmm/DataCon.hs - compiler/GHC/Types/CostCentre.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - compiler/ghc.cabal.in - + testsuite/tests/bytecode/T25090/A.hs - + testsuite/tests/bytecode/T25090/B.hs - + testsuite/tests/bytecode/T25090/C.hs - + testsuite/tests/bytecode/T25090/C.hs-boot - + testsuite/tests/bytecode/T25090/D.hs - + testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/T25090-debug.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2088616d47d89831851c54d36917137ba949fcf9...6c1a78cb43b63e7d9a21a04fb0530baf59de10c3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2088616d47d89831851c54d36917137ba949fcf9...6c1a78cb43b63e7d9a21a04fb0530baf59de10c3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 14:30:19 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Mon, 09 Sep 2024 10:30:19 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/sgraf-resumptive-fork Message-ID: <66df067b7e319_28a43c18ec107012@gitlab.mail> Sebastian Graf pushed new branch wip/sgraf-resumptive-fork at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/sgraf-resumptive-fork You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 14:30:36 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Mon, 09 Sep 2024 10:30:36 -0400 Subject: [Git][ghc/ghc][wip/sgraf-resumptive-fork] Catch more parser errors gracefully Message-ID: <66df068ce1e3d_28a438b1341072c@gitlab.mail> Sebastian Graf pushed to branch wip/sgraf-resumptive-fork at Glasgow Haskell Compiler / GHC Commits: 9075666e by Sebastian Graf at 2024-09-09T16:30:28+02:00 Catch more parser errors gracefully - - - - - 4 changed files: - compiler/GHC/Cmm/Parser.y - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs Changes: ===================================== compiler/GHC/Cmm/Parser.y ===================================== @@ -1234,7 +1234,7 @@ isPtrGlobalRegUse (GlobalRegUse reg ty) go _ = False happyError :: PD a -happyError = PD $ \_ _ s -> unP srcParseFail s +happyError = PD $ \_ _ s -> unP srcParseAbort s -- ----------------------------------------------------------------------------- -- Statement-level macros ===================================== compiler/GHC/Parser.y ===================================== @@ -100,7 +100,7 @@ import Language.Haskell.Syntax.Basic (FieldLabelString(..)) import qualified Data.Semigroup as Semi } -%error { srcParseFail } { srcParseFail' } +%error { srcParseAbort } { srcParseReport } %error.expected %expect 0 -- shift/reduce conflicts @@ -2289,7 +2289,7 @@ infixtype :: { forall b. DisambTD b => PV (LocatedA b) } ftype :: { forall b. DisambTD b => PV (LocatedA b) } : atype { mkHsAppTyHeadPV $1 } - | tyop { failOpFewArgs (fst $1) } + | tyop { failOpFewArgs (fst $1) >> mkHsAppTyHeadPV (sL1a (fst $1) $ mkAnonWildCardTy) } -- TODO! | ftype tyarg { $1 >>= \ $1 -> mkHsAppTyPV $1 $2 } | ftype PREFIX_AT atype { $1 >>= \ $1 -> @@ -4159,7 +4159,7 @@ catchHsExpr = ecpFromExp (noLocA (HsPartial noExtField)) ----------------------------------------------------------------------------- happyError :: P a -happyError = srcParseFail +happyError = srcParseAbort getVARID (L _ (ITvarid x)) = x getCONID (L _ (ITconid x)) = x ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -61,7 +61,7 @@ module GHC.Parser.Lexer ( allocateComments, allocatePriorComments, allocateFinalComments, MonadP(..), getBit, getRealSrcLoc, getPState, - failMsgP, failLocMsgP, srcParseFail, srcParseFail', srcParseErr, + failMsgP, failLocMsgP, srcParseReport, srcParseAbort, srcParseErr, getPsErrorMessages, getPsMessages, popContext, pushModuleContext, setLastToken, setSrcLoc, activeContext, nextIsEOF, @@ -3264,11 +3264,11 @@ srcParseErr options buf len loc expected = mkPlainErrorMsgEnvelope loc (PsErrPar , ped_expected = expected } --- Report a parse failure, giving the span of the previous token as +-- Report a fatal parse failure, giving the span of the previous token as -- the location of the error. This is the entry point for errors -- detected during parsing. -srcParseFail :: P a -srcParseFail = +srcParseAbort :: P a +srcParseAbort = P $ \s at PState{ buffer = buf , options = o , last_len = len @@ -3300,12 +3300,16 @@ srcParseWarn options buf len loc expected = mkPlainWarningMsgEnvelope loc (PsErr , ped_expected = expected } -srcParseFail' :: Located Token -> [String] -> P a -> P a -srcParseFail' (L loc _last_tk) expected_toks resume = +-- Report a resumable parse failure, giving the span of the previous token as +-- the location of the error. This is the entry point for errors +-- detected during parsing. +srcParseReport :: Located Token -> [String] -> P a -> P a +srcParseReport (L loc _last_tk) expected_toks resume = do P $ \s at PState{ buffer = buf , options = o - , last_len = len } -> - unP (addWarning $ srcParseWarn o buf len loc expected_toks) s + , last_len = len + , last_loc = last_loc } -> + unP (addError $ srcParseErr o buf len (mkSrcSpanPs last_loc) []) s resume -- A lexical error is reported at a particular position in the source file, ===================================== compiler/GHC/Parser/PostProcess.hs ===================================== @@ -146,6 +146,7 @@ import GHC.Parser.Errors.Types import GHC.Utils.Lexeme ( okConOcc ) import GHC.Types.TyThing import GHC.Core.Type ( Specificity(..) ) +import GHC.Builtin.Names ( wildCardName ) import GHC.Builtin.Types( cTupleTyConName, tupleTyCon, tupleDataCon, nilDataConName, nilDataConKey, listTyConName, listTyConKey, sumDataCon, @@ -1077,12 +1078,14 @@ checkTyClHdr is_cls ty ; return (L a' (Unqual name), acc, fix , (reverse ops') ++ cps', cs) } - go cs l (HsTyVar _ _ ltc@(L _ tc)) acc ops cps fix - | isRdrTc tc = return (ltc, acc, fix, (reverse ops) ++ cps, cs Semi.<> comments l) - go cs l (HsOpTy _ _ t1 ltc@(L _ tc) t2) acc ops cps _fix - | isRdrTc tc = return (ltc, lhs:rhs:acc, Infix, (reverse ops) ++ cps, cs Semi.<> comments l) - where lhs = HsValArg noExtField t1 - rhs = HsValArg noExtField t2 + go cs l (HsTyVar _ _ ltc@(L _ tc)) acc ops cps fix = do + unless (isRdrTc tc) $ reportMalformed ty l + return (ltc, acc, fix, (reverse ops) ++ cps, cs Semi.<> comments l) + go cs l (HsOpTy _ _ t1 ltc@(L _ tc) t2) acc ops cps _fix = do + unless (isRdrTc tc) $ reportMalformed ty l + let lhs = HsValArg noExtField t1 + rhs = HsValArg noExtField t2 + return (ltc, lhs:rhs:acc, Infix, (reverse ops) ++ cps, cs Semi.<> comments l) go cs l (HsParTy _ ty) acc ops cps fix = goL (cs Semi.<> comments l) ty acc (o:ops) (c:cps) fix where (o,c) = mkParensEpAnn (realSrcSpan (locA l)) @@ -1096,9 +1099,16 @@ checkTyClHdr is_cls ty tup_name | is_cls = cTupleTyConName arity | otherwise = getName (tupleTyCon Boxed arity) -- See Note [Unit tuples] in GHC.Hs.Type (TODO: is this still relevant?) - go _ l _ _ _ _ _ - = addFatalError $ mkPlainErrorMsgEnvelope (locA l) $ - (PsErrMalformedTyOrClDecl ty) + go cs l this_ty acc ops cps fix = do + reportMalformed ty l + case this_ty of + HsWildCardTy{} -> -- This is actually the partial node case + return (L (l2l l) (nameRdrName wildCardName), acc, fix, (reverse ops) ++ cps, cs Semi.<> comments l) + _ -> + P PFailed + + reportMalformed ty l = + addError $ mkPlainErrorMsgEnvelope (locA l) (PsErrMalformedTyOrClDecl ty) -- Combine the annotations from the HsParTy and HsStarTy into a -- new one for the LocatedN RdrName @@ -3253,11 +3263,11 @@ failImportQualifiedTwice loc = warnStarIsType :: SrcSpan -> P () warnStarIsType span = addPsMessage span PsWarnStarIsType -failOpFewArgs :: MonadP m => LocatedN RdrName -> m a +failOpFewArgs :: MonadP m => LocatedN RdrName -> m () failOpFewArgs (L loc op) = do { star_is_type <- getBit StarIsTypeBit ; let is_star_type = if star_is_type then StarIsType else StarIsNotType - ; addFatalError $ mkPlainErrorMsgEnvelope (locA loc) $ + ; addError $ mkPlainErrorMsgEnvelope (locA loc) $ (PsErrOpFewArgs is_star_type op) } requireExplicitNamespaces :: MonadP m => SrcSpan -> m () View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9075666ef84e50f57d5686e5db0ea35f8fe29a89 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9075666ef84e50f57d5686e5db0ea35f8fe29a89 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 15:27:12 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 09 Sep 2024 11:27:12 -0400 Subject: [Git][ghc/ghc][wip/T24817a] Andd ZonkAny and document it Message-ID: <66df13d0cb8a1_28a4600ec8119547@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24817a at Glasgow Haskell Compiler / GHC Commits: 68b9f806 by Simon Peyton Jones at 2024-09-09T16:26:51+01:00 Andd ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 12 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Tc/Types.hs - compiler/GHC/Tc/Utils/Monad.hs - compiler/GHC/Tc/Zonk/Type.hs - testsuite/tests/perf/compiler/T11068.stdout - + testsuite/tests/pmcheck/should_compile/T24817.hs - testsuite/tests/pmcheck/should_compile/all.T - testsuite/tests/printer/T17697.stderr - testsuite/tests/profiling/should_run/staticcallstack002.stdout - testsuite/tests/simplCore/should_compile/T13156.stdout - testsuite/tests/typecheck/should_fail/T13292.stderr Changes: ===================================== compiler/GHC/Builtin/Names.hs ===================================== @@ -2007,6 +2007,12 @@ uWordTyConKey = mkPreludeTyConUnique 163 unsatisfiableClassNameKey :: Unique unsatisfiableClassNameKey = mkPreludeTyConUnique 170 +anyTyConKey :: Unique +anyTyConKey = mkPreludeTyConUnique 171 + +zonkAnyTyConKey :: Unique +zonkAnyTyConKey = mkPreludeTyConUnique 172 + -- Custom user type-errors errorMessageTypeErrorFamKey :: Unique errorMessageTypeErrorFamKey = mkPreludeTyConUnique 181 @@ -2020,9 +2026,6 @@ proxyPrimTyConKey = mkPreludeTyConUnique 184 specTyConKey :: Unique specTyConKey = mkPreludeTyConUnique 185 -anyTyConKey :: Unique -anyTyConKey = mkPreludeTyConUnique 186 - smallArrayPrimTyConKey = mkPreludeTyConUnique 187 smallMutableArrayPrimTyConKey = mkPreludeTyConUnique 188 ===================================== compiler/GHC/Builtin/Types.hs ===================================== @@ -91,7 +91,7 @@ module GHC.Builtin.Types ( cTupleSelId, cTupleSelIdName, -- * Any - anyTyCon, anyTy, anyTypeOfKind, + anyTyCon, anyTy, anyTypeOfKind, zonkAnyTyCon, -- * Recovery TyCon makeRecoveryTyCon, @@ -184,7 +184,7 @@ import GHC.Core.ConLike import GHC.Core.TyCon import GHC.Core.Class ( Class, mkClass ) import GHC.Core.Map.Type ( TypeMap, emptyTypeMap, extendTypeMap, lookupTypeMap ) -import qualified GHC.Core.TyCo.Rep as TyCoRep (Type(TyConApp)) +import qualified GHC.Core.TyCo.Rep as TyCoRep ( Type(TyConApp) ) import GHC.Types.TyThing import GHC.Types.SourceText @@ -309,6 +309,7 @@ wiredInTyCons = map (dataConTyCon . snd) boxingDataCons , soloTyCon , anyTyCon + , zonkAnyTyCon , boolTyCon , charTyCon , stringTyCon @@ -419,65 +420,102 @@ doubleDataConName = mkWiredInDataConName UserSyntax gHC_TYPES (fsLit "D#") {- Note [Any types] ~~~~~~~~~~~~~~~~ -The type constructor Any, +The type constructors `Any` and `ZonkAny` are closed type families declared thus: - type family Any :: k where { } + type family Any :: forall k. k where { } + type family ZonkAny :: forall k. Nat -> k where { } -It has these properties: +They are used when we want a type of a particular kind, but we don't really care +what that type is. The leading example is this: `ZonkAny` is used to instantiate +un-constrained type variables after type checking. For example, consider the +term (length [] :: Int), where - * Note that 'Any' is kind polymorphic since in some program we may - need to use Any to fill in a type variable of some kind other than * - (see #959 for examples). Its kind is thus `forall k. k``. + length :: forall a. [a] -> Int + [] :: forall a. [a] - * It is defined in module GHC.Types, and exported so that it is - available to users. For this reason it's treated like any other - wired-in type: - - has a fixed unique, anyTyConKey, - - lives in the global name cache +We must type-apply `length` and `[]`, but to what type? It doesn't matter! +The typechecker will end up with - * It is a *closed* type family, with no instances. This means that - if ty :: '(k1, k2) we add a given coercion - g :: ty ~ (Fst ty, Snd ty) - If Any was a *data* type, then we'd get inconsistency because 'ty' - could be (Any '(k1,k2)) and then we'd have an equality with Any on - one side and '(,) on the other. See also #9097 and #9636. + length @alpha ([] @alpha) - * When instantiated at a lifted type it is inhabited by at least one value, - namely bottom +where `alpha` is an un-constrained unification variable. The "zonking" process zaps +that unconstrained `alpha` to an arbitrary type (ZonkAny @Type 3), where the `3` is +arbitrary (see wrinkle (Any5) below). This is done in `GHC.Tc.Zonk.Type.commitFlexi`. +So we end up with - * You can safely coerce any /lifted/ type to Any and back with unsafeCoerce. - You can safely coerce any /unlifted/ type to Any and back with unsafeCoerceUnlifted. - You can coerce /any/ type to Any and back with unsafeCoerce#, but it's only safe when - the kinds of both the type and Any match. + length @(ZonkAny @Type 3) ([] @(ZonkAny @Type 3)) - For lifted/unlifted types unsafeCoerce[Unlifted] should be preferred over unsafeCoerce# - as they prevent accidentally coercing between types with kinds that don't match. +`Any` and `ZonkAny` differ only in the presence of the `Nat` argument; see +wrinkle (Any4). - See examples in ghc-prim:GHC.Types +Wrinkles: - * It does not claim to be a *data* type, and that's important for - the code generator, because the code gen may *enter* a data value - but never enters a function value. +(Any1) `Any` and `ZonkAny` are kind polymorphic since in some program we may + need to use `ZonkAny` to fill in a type variable of some kind other than * + (see #959 for examples). - * It is wired-in so we can easily refer to it where we don't have a name - environment (e.g. see Rules.matchRule for one example) +(Any2) They are /closed/ type families, with no instances. For example, suppose that + with alpha :: '(k1, k2) we add a given coercion + g :: alpha ~ (Fst alpha, Snd alpha) + and we zonked alpha = ZonkAny @(k1,k2) n. Then, if `ZonkAny` was a /data/ type, + we'd get inconsistency because we'd have a Given equality with `ZonkAny` on one + side and '(,) on the other. See also #9097 and #9636. -It's used to instantiate un-constrained type variables after type checking. For -example, 'length' has type + See #25244 for a suggestion that we instead use an /open/ type family for which + you cannot provide instances. Probably the difference is not very important. - length :: forall a. [a] -> Int +(Any3) They do not claim to be /data/ types, and that's important for + the code generator, because the code gen may /enter/ a data value + but never enters a function value. + +(Any4) `ZonkAny` takes a `Nat` argument so that we can readily make up /distinct/ + types (#24817). Consider -and the list datacon for the empty list has type + data SBool a where { STrue :: SBool True; SFalse :: SBool False } - [] :: forall a. [a] + foo :: forall a b. (SBool a, SBool b) -In order to compose these two terms as @length []@ a type -application is required, but there is no constraint on the -choice. In this situation GHC uses 'Any', + bar :: Bool + bar = case foo @alpha @beta of + (STrue, SFalse) -> True -- This branch is not inaccessible! + _ -> False -> length @(Any @Type) ([] @(Any @Type)) + Now, what are `alpha` and `beta`? If we zonk both of them to the same type + `Any @Type`, the pattern-match checker will (wrongly) report that the first + branch is inaccessible. So we zonk them to two /different/ types: + alpha := ZonkAny @Type 4 and beta := ZonkAny @Type k 5 + (The actual numbers are arbitrary; they just need to differ.) + + The unique-name generation comes from field `tcg_zany_n` of `TcGblEnv`; and + `GHC.Tc.Zonk.Type.commitFlexi` calls `GHC.Tc.Utils.Monad.newZonkAnyType` to + make up a fresh type. + +(Any5) `Any` and `ZonkAny` are wired-in so we can easily refer to it where we + don't have a name environment (e.g. see Rules.matchRule for one example) + +(Any6) `Any` is defined in library module ghc-prim:GHC.Types, and exported so that + it is available to users. For this reason it's treated like any other + wired-in type: + - has a fixed unique, anyTyConKey, + - lives in the global name cache + Currently `ZonkAny` is not available to users; but it could easily be. -Above, we print kinds explicitly, as if with -fprint-explicit-kinds. +(Any7) Properties of `Any`: + * When `Any` is instantiated at a lifted type it is inhabited by at least one value, + namely bottom. + + * You can safely coerce any /lifted/ type to `Any` and back with `unsafeCoerce`. + + * You can safely coerce any /unlifted/ type to `Any` and back with `unsafeCoerceUnlifted`. + + * You can coerce /any/ type to `Any` and back with `unsafeCoerce#`, but it's only safe when + the kinds of both the type and `Any` match. + + * For lifted/unlifted types `unsafeCoerce[Unlifted]` should be preferred over + `unsafeCoerce#` as they prevent accidentally coercing between types with kinds + that don't match. + + See examples in ghc-prim:GHC.Types The Any tycon used to be quite magic, but we have since been able to implement it merely with an empty kind polymorphic type family. See #10886 for a @@ -490,6 +528,7 @@ anyTyConName = mkWiredInTyConName UserSyntax gHC_TYPES (fsLit "Any") anyTyConKey anyTyCon anyTyCon :: TyCon +-- See Note [Any types] anyTyCon = mkFamilyTyCon anyTyConName binders res_kind Nothing (ClosedSynFamilyTyCon Nothing) Nothing @@ -504,6 +543,24 @@ anyTy = mkTyConTy anyTyCon anyTypeOfKind :: Kind -> Type anyTypeOfKind kind = mkTyConApp anyTyCon [kind] +zonkAnyTyConName :: Name +zonkAnyTyConName = + mkWiredInTyConName UserSyntax gHC_TYPES (fsLit "ZonkAny") zonkAnyTyConKey zonkAnyTyCon + +zonkAnyTyCon :: TyCon +-- ZonkAnyTyCon :: forall k. Nat -> k +-- See Note [Any types] +zonkAnyTyCon = mkFamilyTyCon zonkAnyTyConName + [ mkNamedTyConBinder Specified kv + , mkAnonTyConBinder nat_kv ] + (mkTyVarTy kv) + Nothing + (ClosedSynFamilyTyCon Nothing) + Nothing + NotInjective + where + [kv,nat_kv] = mkTemplateKindVars [liftedTypeKind, naturalTy] + -- | Make a fake, recovery 'TyCon' from an existing one. -- Used when recovering from errors in type declarations makeRecoveryTyCon :: TyCon -> TyCon ===================================== compiler/GHC/Tc/Types.hs ===================================== @@ -567,6 +567,10 @@ data TcGblEnv tcg_dfun_n :: TcRef OccSet, -- ^ Allows us to choose unique DFun names. + tcg_zany_n :: TcRef Integer, + -- ^ A source of unique identities for ZonkAny instances + -- See Note [Any types] in GHC.Builtin.Types, wrinkle (Any4) + tcg_merged :: [(Module, Fingerprint)], -- ^ The requirements we merged with; we always have to recompile -- if any of these changed. ===================================== compiler/GHC/Tc/Utils/Monad.hs ===================================== @@ -142,7 +142,7 @@ module GHC.Tc.Utils.Monad( getCCIndexM, getCCIndexTcM, -- * Zonking - liftZonkM, + liftZonkM, newZonkAnyType, -- * Complete matches localAndImportedCompleteMatches, getCompleteMatchesTcM, @@ -156,6 +156,7 @@ import GHC.Prelude import GHC.Builtin.Names +import GHC.Builtin.Types( zonkAnyTyCon ) import GHC.Tc.Errors.Types import GHC.Tc.Types -- Re-export all @@ -178,6 +179,7 @@ import GHC.Core.UsageEnv import GHC.Core.Multiplicity import GHC.Core.InstEnv import GHC.Core.FamInstEnv +import GHC.Core.Type( mkNumLitTy ) import GHC.Driver.Env import GHC.Driver.Session @@ -258,6 +260,7 @@ initTc hsc_env hsc_src keep_rn_syntax mod loc do_this infer_var <- newIORef True ; infer_reasons_var <- newIORef emptyMessages ; dfun_n_var <- newIORef emptyOccSet ; + zany_n_var <- newIORef 0 ; let { type_env_var = hsc_type_env_vars hsc_env }; dependent_files_var <- newIORef [] ; @@ -348,6 +351,7 @@ initTc hsc_env hsc_src keep_rn_syntax mod loc do_this tcg_patsyns = [], tcg_merged = [], tcg_dfun_n = dfun_n_var, + tcg_zany_n = zany_n_var, tcg_keep = keep_var, tcg_hdr_info = (Nothing,Nothing), tcg_hpc = False, @@ -1787,6 +1791,18 @@ chooseUniqueOccTc fn = ; writeTcRef dfun_n_var (extendOccSet set occ) ; return occ } +newZonkAnyType :: Kind -> TcM Type +-- Return a type (ZonkAny @k n), where n is fresh +-- Recall ZonkAny :: forall k. Natural -> k +-- See Note [Any types] in GHC.Builtin.Types, wrinkle (Any4) +newZonkAnyType kind + = do { env <- getGblEnv + ; let zany_n_var = tcg_zany_n env + ; i <- readTcRef zany_n_var + ; let !i2 = i+1 + ; writeTcRef zany_n_var i2 + ; return (mkTyConApp zonkAnyTyCon [kind, mkNumLitTy i]) } + getConstraintVar :: TcM (TcRef WantedConstraints) getConstraintVar = do { env <- getLclEnv; return (tcl_lie env) } ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -54,7 +54,7 @@ import GHC.Tc.Types.TcRef import GHC.Tc.TyCl.Build ( TcMethInfo, MethInfo ) import GHC.Tc.Utils.Env ( tcLookupGlobalOnly ) import GHC.Tc.Utils.TcType -import GHC.Tc.Utils.Monad ( setSrcSpanA, liftZonkM, traceTc, addErr ) +import GHC.Tc.Utils.Monad ( newZonkAnyType, setSrcSpanA, liftZonkM, traceTc, addErr ) import GHC.Tc.Types.Constraint import GHC.Tc.Types.Evidence import GHC.Tc.Errors.Types @@ -469,8 +469,9 @@ commitFlexi tv zonked_kind -> do { addErr $ TcRnZonkerMessage (ZonkerCannotDefaultConcrete origin) ; return (anyTypeOfKind zonked_kind) } | otherwise - -> do { traceTc "Defaulting flexi tyvar to Any:" (pprTyVar tv) - ; return (anyTypeOfKind zonked_kind) } + -> do { traceTc "Defaulting flexi tyvar to ZonkAny:" (pprTyVar tv) + -- See Note [Any types] in GHC.Builtin.Types, esp wrinkle (Any4) + ; newZonkAnyType zonked_kind } RuntimeUnkFlexi -> do { traceTc "Defaulting flexi tyvar to RuntimeUnk:" (pprTyVar tv) ===================================== testsuite/tests/perf/compiler/T11068.stdout ===================================== @@ -23,137 +23,137 @@ `cast` (GHC.Internal.Generics.N:M1[0] `cast` (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 ===================================== testsuite/tests/pmcheck/should_compile/T24817.hs ===================================== @@ -0,0 +1,15 @@ +{-# LANGUAGE GADTs, DataKinds #-} + +module T24817 where + +data SBool b where + STrue :: SBool True + SFalse :: SBool False + +foo :: forall a b. (SBool a, SBool b) +foo = error "urk" + +bar :: Bool +bar = case foo of + (STrue, SFalse) -> True + _ -> False ===================================== testsuite/tests/pmcheck/should_compile/all.T ===================================== @@ -171,3 +171,4 @@ test('DsIncompleteRecSel1', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel2', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel3', [collect_compiler_stats('bytes allocated', 10)], compile, ['-Wincomplete-record-selectors']) test('DoubleMatch', normal, compile, [overlapping_incomplete]) +test('T24817', normal, compile, [overlapping_incomplete]) ===================================== testsuite/tests/printer/T17697.stderr ===================================== @@ -1,7 +1,8 @@ - T17697.hs:6:5: warning: [GHC-88464] [-Wdeferred-out-of-scope-variables (in -Wdefault)] Variable not in scope: threadDelay :: t0 -> IO a0 T17697.hs:6:5: warning: [GHC-81995] [-Wunused-do-bind (in -Wall)] - A do-notation statement discarded a result of type ‘GHC.Types.Any’ + A do-notation statement discarded a result of type + ‘GHC.Types.ZonkAny 1’ Suggested fix: Suppress this warning by saying ‘_ <- threadDelay 1’ + ===================================== testsuite/tests/profiling/should_run/staticcallstack002.stdout ===================================== @@ -1,4 +1,4 @@ -Just (InfoProv {ipName = "sat_s237_info", ipDesc = THUNK, ipTyDesc = "Any", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "10:23-39"}) -Just (InfoProv {ipName = "sat_s23r_info", ipDesc = THUNK, ipTyDesc = "Any", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "11:23-42"}) -Just (InfoProv {ipName = "sat_s23L_info", ipDesc = THUNK, ipTyDesc = "Any", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "12:23-46"}) -Just (InfoProv {ipName = "sat_s245_info", ipDesc = THUNK, ipTyDesc = "Any", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "13:23-44"}) +Just (InfoProv {ipName = "sat_s1Rh_info", ipDesc = THUNK, ipTyDesc = "ZonkAny 0", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "10:23-39"}) +Just (InfoProv {ipName = "sat_s1RB_info", ipDesc = THUNK, ipTyDesc = "ZonkAny 1", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "11:23-42"}) +Just (InfoProv {ipName = "sat_s1RV_info", ipDesc = THUNK, ipTyDesc = "ZonkAny 2", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "12:23-46"}) +Just (InfoProv {ipName = "sat_s1Sf_info", ipDesc = THUNK, ipTyDesc = "ZonkAny 3", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "13:23-44"}) ===================================== testsuite/tests/simplCore/should_compile/T13156.stdout ===================================== @@ -1 +1,2 @@ - case r @GHC.Types.Any of { __DEFAULT -> r @a } + case r @(GHC.Types.ZonkAny 0) of { __DEFAULT -> + case r @(GHC.Types.ZonkAny 1) of { __DEFAULT -> r @a } ===================================== testsuite/tests/typecheck/should_fail/T13292.stderr ===================================== @@ -1,4 +1,3 @@ - T13292a.hs:4:12: warning: [GHC-39999] [-Wdeferred-type-errors (in -Wdefault)] • Ambiguous type variable ‘m0’ arising from a use of ‘return’ prevents the constraint ‘(Monad m0)’ from being solved. @@ -15,14 +14,15 @@ T13292a.hs:4:12: warning: [GHC-39999] [-Wdeferred-type-errors (in -Wdefault)] In an equation for ‘someFunc’: someFunc = return () T13292.hs:6:1: warning: [GHC-83865] [-Wdeferred-type-errors (in -Wdefault)] - • Couldn't match type ‘GHC.Types.Any’ with ‘IO’ + • Couldn't match type ‘GHC.Types.ZonkAny 0’ with ‘IO’ Expected: IO () - Actual: GHC.Types.Any () + Actual: GHC.Types.ZonkAny 0 () • When checking the type of the IO action ‘main’ T13292.hs:6:1: warning: [GHC-83865] [-Wdeferred-type-errors (in -Wdefault)] - • Couldn't match type ‘GHC.Types.Any’ with ‘IO’ + • Couldn't match type ‘GHC.Types.ZonkAny 0’ with ‘IO’ Expected: IO () - Actual: GHC.Types.Any () + Actual: GHC.Types.ZonkAny 0 () • In the expression: main When checking the type of the IO action ‘main’ + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/68b9f8063ff7f787d639ef530d3c1b9357333e08 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/68b9f8063ff7f787d639ef530d3c1b9357333e08 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 17:55:18 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Mon, 09 Sep 2024 13:55:18 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/az/epa-exactprint-sync Message-ID: <66df36866b4bd_33413a32f5dc101188@gitlab.mail> Alan Zimmerman pushed new branch wip/az/epa-exactprint-sync at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/az/epa-exactprint-sync You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 18:58:36 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 09 Sep 2024 14:58:36 -0400 Subject: [Git][ghc/ghc][master] JS: fake support for native adjustors (#25159) Message-ID: <66df455c83632_3b1dde2ace70787ed@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 2 changed files: - m4/ghc_adjustors_method.m4 - utils/ghc-toolchain/exe/Main.hs Changes: ===================================== m4/ghc_adjustors_method.m4 ===================================== @@ -4,7 +4,7 @@ dnl Use libffi for adjustors? AC_DEFUN([GHC_ADJUSTORS_METHOD], [ case [$]{$1[Arch]} in - i386|x86_64) + i386|x86_64|javascript) # We have native adjustor support on these platforms HaveNativeAdjustor=yes ;; ===================================== utils/ghc-toolchain/exe/Main.hs ===================================== @@ -374,6 +374,7 @@ archHasNativeAdjustors :: Arch -> Bool archHasNativeAdjustors = \case ArchX86 -> True ArchX86_64 -> True + ArchJavaScript -> True _ -> False View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/03055c71446c8f1265de3a6ccb5bc9e09827821e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/03055c71446c8f1265de3a6ccb5bc9e09827821e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 18:59:19 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 09 Sep 2024 14:59:19 -0400 Subject: [Git][ghc/ghc][master] JS: remove redundant h$lstat Message-ID: <66df4586d6f00_3b1dde2544f08184c@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - 1 changed file: - libraries/ghc-internal/jsbits/base.js Changes: ===================================== libraries/ghc-internal/jsbits/base.js ===================================== @@ -367,23 +367,6 @@ function h$base_lstat(file, file_off, stat, stat_off, c) { h$unsupported(-1, c); } -function h$lstat(file, file_off, stat, stat_off) { - TRACE_IO("lstat") -#ifndef GHCJS_BROWSER - if(h$isNode()) { - try { - var fs = h$fs.lstatSync(h$decodeUtf8z(file, file_off)); - h$base_fillStat(fs, stat, stat_off); - return 0; - } catch(e) { - h$setErrno(e); - return -1; - } - } else -#endif - h$unsupported(-1); -} - function h$rmdir(file, file_off) { TRACE_IO("rmdir") #ifndef GHCJS_BROWSER View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5bf0e6bc3e3024c289cb637a19c9660c848e831f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5bf0e6bc3e3024c289cb637a19c9660c848e831f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 19:19:40 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Mon, 09 Sep 2024 15:19:40 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] Try to fix CCallConv test Message-ID: <66df4a4c724c9_3b1dde717dfc857a@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 69230df5 by Sven Tennie at 2024-09-09T21:19:17+02:00 Try to fix CCallConv test - Make ccalls's results strict - Use better test config - - - - - 3 changed files: - testsuite/tests/codeGen/should_run/CCallConv.hs - testsuite/tests/codeGen/should_run/CCallConv_c.c - testsuite/tests/codeGen/should_run/all.T Changes: ===================================== testsuite/tests/codeGen/should_run/CCallConv.hs ===================================== @@ -3,6 +3,7 @@ {-# LANGUAGE MagicHash #-} {-# LANGUAGE UnboxedTuples #-} {-# LANGUAGE UnliftedFFITypes #-} +{-# LANGUAGE BangPatterns #-} -- | This test ensures that sub-word signed and unsigned parameters are correctly -- handed over to C functions. I.e. it asserts the calling-convention. @@ -92,18 +93,18 @@ main = -- twos-complement, which is the same as the max word value. let i8 :: Int8# = intToInt8# (-1#) w8 :: Word8# = wordToWord8# (255##) - res8 :: Int64# = fun8 i8 w8 i8 i8 i8 i8 i8 i8 w8 i8 + !res8 :: Int64# = fun8 i8 w8 i8 i8 i8 i8 i8 i8 w8 i8 expected_res8 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word8) + 8 * (-1) i16 :: Int16# = intToInt16# (-1#) w16 :: Word16# = wordToWord16# (65535##) - res16 :: Int64# = fun16 i16 w16 i16 i16 i16 i16 i16 i16 w16 i16 + !res16 :: Int64# = fun16 i16 w16 i16 i16 i16 i16 i16 i16 w16 i16 expected_res16 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word16) + 8 * (-1) i32 :: Int32# = intToInt32# (-1#) w32 :: Word32# = wordToWord32# (4294967295##) - res32 :: Int64# = fun32 i32 w32 i32 i32 i32 i32 i32 i32 w32 i32 + !res32 :: Int64# = fun32 i32 w32 i32 i32 i32 i32 i32 i32 w32 i32 expected_res32 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word32) + 8 * (-1) - resFloat :: Float = F# (funFloat 1.0# 1.1# 1.2# 1.3# 1.4# 1.5# 1.6# 1.7# 1.8# 1.9#) - resDouble :: Double = D# (funDouble 1.0## 1.1## 1.2## 1.3## 1.4## 1.5## 1.6## 1.7## 1.8## 1.9##) + !resFloat :: Float = F# (funFloat 1.0# 1.1# 1.2# 1.3# 1.4# 1.5# 1.6# 1.7# 1.8# 1.9#) + !resDouble :: Double = D# (funDouble 1.0## 1.1## 1.2## 1.3## 1.4## 1.5## 1.6## 1.7## 1.8## 1.9##) in do print $ "fun8 result:" ++ show (I64# res8) assertEqual expected_res8 (I64# res8) ===================================== testsuite/tests/codeGen/should_run/CCallConv_c.c ===================================== @@ -1,5 +1,5 @@ -#include "stdint.h" -#include "stdio.h" +#include +#include int64_t fun8(int8_t a0, uint8_t a1, int8_t a2, int8_t a3, int8_t a4, int8_t a5, int8_t a6, int8_t a7, int8_t s0, uint8_t s1) { ===================================== testsuite/tests/codeGen/should_run/all.T ===================================== @@ -251,7 +251,4 @@ test('T23034', req_c, compile_and_run, ['-O2 T23034_c.c']) test('T24700', normal, compile_and_run, ['-O']) test('T24893', normal, compile_and_run, ['-O']) -test('CCallConv', - [], - multi_compile_and_run, - ['CCallConv', [('CCallConv_c.c', '')], '']) +test('CCallConv', req_c, compile_and_run, ['CCallConv_c.c']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/69230df530a8035c2f1c682904b868d2c5b8a9af -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/69230df530a8035c2f1c682904b868d2c5b8a9af You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 19:30:32 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 09 Sep 2024 15:30:32 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 8 commits: JS: fake support for native adjustors (#25159) Message-ID: <66df4cd865688_3b1dde85778089933@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - dd9e89cb by Simon Peyton Jones at 2024-09-09T15:30:13-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - bb94b0a9 by Simon Peyton Jones at 2024-09-09T15:30:13-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 2a7d5a33 by Simon Peyton Jones at 2024-09-09T15:30:14-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - df6a6e5b by Andrzej Rybczak at 2024-09-09T15:30:19-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - 19d8909f by Simon Peyton Jones at 2024-09-09T15:30:20-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 0b9cdbab by Hécate Kleidukos at 2024-09-09T15:30:25-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 30 changed files: - compiler/GHC/Core/InstEnv.hs - compiler/GHC/Core/Opt/CallerCC/Types.hs - compiler/GHC/Core/Opt/WorkWrap/Utils.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Dict.hs - compiler/GHC/Tc/Solver/Equality.hs - compiler/GHC/Tc/Solver/Irred.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Solver/Solve.hs - compiler/GHC/Tc/TyCl/Instance.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Utils/Instantiate.hs - compiler/GHC/Tc/Utils/Unify.hs - libraries/ghc-internal/jsbits/base.js - m4/ghc_adjustors_method.m4 - testsuite/tests/indexed-types/should_compile/Simple14.hs - − testsuite/tests/indexed-types/should_compile/Simple14.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/38b0927bd76f835abe3646450a2aa9116b5d5b9a...0b9cdbab4c6882ac1da6bbaacb4c9249b6c0d27a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/38b0927bd76f835abe3646450a2aa9116b5d5b9a...0b9cdbab4c6882ac1da6bbaacb4c9249b6c0d27a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 19:38:21 2024 From: gitlab at gitlab.haskell.org (Bodigrim (@Bodigrim)) Date: Mon, 09 Sep 2024 15:38:21 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/deepseq-1.5.1.0 Message-ID: <66df4eadaebd0_3b1ddea83054997bf@gitlab.mail> Bodigrim pushed new branch wip/deepseq-1.5.1.0 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/deepseq-1.5.1.0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 21:07:09 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Mon, 09 Sep 2024 17:07:09 -0400 Subject: [Git][ghc/ghc][wip/az/epa-remove-anchor-alias] 24 commits: simplifier: Fix space leak during demand analysis Message-ID: <66df637dd1c89_3c4da63e7a4c358b9@gitlab.mail> Alan Zimmerman pushed to branch wip/az/epa-remove-anchor-alias at Glasgow Haskell Compiler / GHC Commits: 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - db092e30 by Alan Zimmerman at 2024-09-09T20:50:41+01:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Annotation.hs - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs - compiler/GHC/Stg/Unarise.hs - compiler/GHC/Stg/Utils.hs - compiler/GHC/StgToCmm/DataCon.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Utils/Env.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/CostCentre.hs - compiler/GHC/Types/Demand.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/992b25c3fb7288ba6604bfb3f76095bdc7af4fc0...db092e30408fa052c46f9b30355f13104ace36a1 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/992b25c3fb7288ba6604bfb3f76095bdc7af4fc0...db092e30408fa052c46f9b30355f13104ace36a1 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 21:48:47 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Mon, 09 Sep 2024 17:48:47 -0400 Subject: [Git][ghc/ghc][wip/az/epa-remove-anchor-alias] EPA: Remove Anchor = EpaLocation synonym Message-ID: <66df6d3fb8a67_3c4da665b4904414f@gitlab.mail> Alan Zimmerman pushed to branch wip/az/epa-remove-anchor-alias at Glasgow Haskell Compiler / GHC Commits: 9788163f by Alan Zimmerman at 2024-09-09T22:48:32+01:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 5 changed files: - compiler/GHC/Parser.y - compiler/GHC/Parser/Annotation.hs - compiler/GHC/Parser/PostProcess.hs - utils/check-exact/ExactPrint.hs - utils/check-exact/Utils.hs Changes: ===================================== compiler/GHC/Parser.y ===================================== @@ -4483,13 +4483,13 @@ gl = getLoc glA :: HasLoc a => a -> SrcSpan glA = getHasLoc -glR :: HasLoc a => a -> Anchor +glR :: HasLoc a => a -> EpaLocation glR !la = EpaSpan (getHasLoc la) -glEE :: (HasLoc a, HasLoc b) => a -> b -> Anchor +glEE :: (HasLoc a, HasLoc b) => a -> b -> EpaLocation glEE !x !y = spanAsAnchor $ comb2 x y -glRM :: Located a -> Maybe Anchor +glRM :: Located a -> Maybe EpaLocation glRM (L !l _) = Just $ spanAsAnchor l glAA :: HasLoc a => a -> EpaLocation @@ -4609,11 +4609,11 @@ hsDoAnn :: Located a -> LocatedAn t b -> AnnKeywordId -> AnnList hsDoAnn (L l _) (L ll _) kw = AnnList (Just $ spanAsAnchor (locA ll)) Nothing Nothing [AddEpAnn kw (srcSpan2e l)] [] -listAsAnchor :: [LocatedAn t a] -> Located b -> Anchor +listAsAnchor :: [LocatedAn t a] -> Located b -> EpaLocation listAsAnchor [] (L l _) = spanAsAnchor l listAsAnchor (h:_) s = spanAsAnchor (comb2 h s) -listAsAnchorM :: [LocatedAn t a] -> Maybe Anchor +listAsAnchorM :: [LocatedAn t a] -> Maybe EpaLocation listAsAnchorM [] = Nothing listAsAnchorM (L l _:_) = case locA l of ===================================== compiler/GHC/Parser/Annotation.hs ===================================== @@ -23,7 +23,7 @@ module GHC.Parser.Annotation ( TokenLocation(..), DeltaPos(..), deltaPos, getDeltaLine, - EpAnn(..), Anchor, + EpAnn(..), anchor, spanAsAnchor, realSpanAsAnchor, noSpanAnchor, @@ -528,7 +528,7 @@ instance Outputable AddEpAnn where -- new AST fragments out of old ones, and have them still printed out -- in a precise way. data EpAnn ann - = EpAnn { entry :: !Anchor + = EpAnn { entry :: !EpaLocation -- ^ Base location for the start of the syntactic element -- holding the annotations. , anns :: !ann -- ^ Annotations added by the Parser @@ -539,15 +539,6 @@ data EpAnn ann deriving (Data, Eq, Functor) -- See Note [XRec and Anno in the AST] --- | An 'Anchor' records the base location for the start of the --- syntactic element holding the annotations, and is used as the point --- of reference for calculating delta positions for contained --- annotations. --- It is also normally used as the reference point for the spacing of --- the element relative to its container. If the AST element is moved, --- that relationship is tracked using the 'EpaDelta' constructor instead. -type Anchor = EpaLocation -- Transitional - anchor :: (EpaLocation' a) -> RealSrcSpan anchor (EpaSpan (RealSrcSpan r _)) = r anchor _ = panic "anchor" @@ -676,7 +667,7 @@ data AnnListItem -- keywords such as 'where'. data AnnList = AnnList { - al_anchor :: Maybe Anchor, -- ^ start point of a list having layout + al_anchor :: Maybe EpaLocation, -- ^ start point of a list having layout al_open :: Maybe AddEpAnn, al_close :: Maybe AddEpAnn, al_rest :: [AddEpAnn], -- ^ context, such as 'where' keyword @@ -1143,7 +1134,7 @@ listLocation as = EpaSpan (go noSrcSpan as) go acc (L (EpAnn (EpaSpan s) _ _) _:rest) = go (combine acc s) rest go acc (_:rest) = go acc rest -widenAnchor :: Anchor -> [AddEpAnn] -> Anchor +widenAnchor :: EpaLocation -> [AddEpAnn] -> EpaLocation widenAnchor (EpaSpan (RealSrcSpan s mb)) as = EpaSpan (RealSrcSpan (widenRealSpan s as) (liftA2 combineBufSpans mb (bufSpanFromAnns as))) widenAnchor (EpaSpan us) _ = EpaSpan us @@ -1151,7 +1142,7 @@ widenAnchor a at EpaDelta{} as = case (realSpanFromAnns as) of Strict.Nothing -> a Strict.Just r -> EpaSpan (RealSrcSpan r Strict.Nothing) -widenAnchorS :: Anchor -> SrcSpan -> Anchor +widenAnchorS :: EpaLocation -> SrcSpan -> EpaLocation widenAnchorS (EpaSpan (RealSrcSpan s mbe)) (RealSrcSpan r mbr) = EpaSpan (RealSrcSpan (combineRealSrcSpans s r) (liftA2 combineBufSpans mbe mbr)) widenAnchorS (EpaSpan us) _ = EpaSpan us ===================================== compiler/GHC/Parser/PostProcess.hs ===================================== @@ -478,13 +478,13 @@ add_where an@(AddEpAnn _ (EpaSpan (RealSrcSpan rs _))) (EpAnn a (AnnList anc o c add_where (AddEpAnn _ _) _ _ = panic "add_where" -- EpaDelta should only be used for transformations -valid_anchor :: Anchor -> Bool +valid_anchor :: EpaLocation -> Bool valid_anchor (EpaSpan (RealSrcSpan r _)) = srcSpanStartLine r >= 0 valid_anchor _ = False -- If the decl list for where binds is empty, the anchor ends up -- invalid. In this case, use the parent one -patch_anchor :: RealSrcSpan -> Anchor -> Anchor +patch_anchor :: RealSrcSpan -> EpaLocation -> EpaLocation patch_anchor r EpaDelta{} = EpaSpan (RealSrcSpan r Strict.Nothing) patch_anchor r1 (EpaSpan (RealSrcSpan r0 mb)) = EpaSpan (RealSrcSpan r mb) where @@ -495,9 +495,9 @@ fixValbindsAnn :: EpAnn AnnList -> EpAnn AnnList fixValbindsAnn (EpAnn anchor (AnnList ma o c r t) cs) = (EpAnn (widenAnchor anchor (r ++ map trailingAnnToAddEpAnn t)) (AnnList ma o c r t) cs) --- | The 'Anchor' for a stmtlist is based on either the location or +-- | The anchor for a stmtlist is based on either the location or -- the first semicolon annotion. -stmtsAnchor :: Located (OrdList AddEpAnn,a) -> Maybe Anchor +stmtsAnchor :: Located (OrdList AddEpAnn,a) -> Maybe EpaLocation stmtsAnchor (L (RealSrcSpan l mb) ((ConsOL (AddEpAnn _ (EpaSpan (RealSrcSpan r rb))) _), _)) = Just $ widenAnchorS (EpaSpan (RealSrcSpan l mb)) (RealSrcSpan r rb) stmtsAnchor (L (RealSrcSpan l mb) _) = Just $ EpaSpan (RealSrcSpan l mb) ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -175,8 +175,8 @@ data EPState = EPState { uAnchorSpan :: !RealSrcSpan -- ^ in pre-changed AST -- reference frame, from -- Annotation - , uExtraDP :: !(Maybe Anchor) -- ^ Used to anchor a - -- list + , uExtraDP :: !(Maybe EpaLocation) -- ^ Used to anchor a + -- list , pAcceptSpan :: Bool -- ^ When we have processed an -- entry of EpaDelta, accept the -- next `EpaSpan` start as the @@ -214,21 +214,21 @@ class HasTrailing a where setTrailing :: a -> [TrailingAnn] -> a setAnchorEpa :: (HasTrailing an, NoAnn an) - => EpAnn an -> Anchor -> [TrailingAnn] -> EpAnnComments -> EpAnn an + => EpAnn an -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn an setAnchorEpa (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing an ts) cs -setAnchorHsModule :: HsModule GhcPs -> Anchor -> EpAnnComments -> HsModule GhcPs +setAnchorHsModule :: HsModule GhcPs -> EpaLocation -> EpAnnComments -> HsModule GhcPs setAnchorHsModule hsmod anc cs = hsmod { hsmodExt = (hsmodExt hsmod) {hsmodAnn = an'} } where anc' = anc an' = setAnchorEpa (hsmodAnn $ hsmodExt hsmod) anc' [] cs setAnchorAn :: (HasTrailing an, NoAnn an) - => LocatedAn an a -> Anchor -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a + => LocatedAn an a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a setAnchorAn (L (EpAnn _ an _) a) anc ts cs = (L (EpAnn anc (setTrailing an ts) cs) a) -- `debug` ("setAnchorAn: anc=" ++ showAst anc) -setAnchorEpaL :: EpAnn AnnList -> Anchor -> [TrailingAnn] -> EpAnnComments -> EpAnn AnnList +setAnchorEpaL :: EpAnn AnnList -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn AnnList setAnchorEpaL (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing (an {al_anchor = Nothing}) ts) cs -- --------------------------------------------------------------------- @@ -250,14 +250,14 @@ data CanUpdateAnchor = CanUpdateAnchor | NoCanUpdateAnchor deriving (Eq, Show) -data Entry = Entry Anchor [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor +data Entry = Entry EpaLocation [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor | NoEntryVal -- | For flagging whether to capture comments in an EpaDelta or not data CaptureComments = CaptureComments | NoCaptureComments -mkEntry :: Anchor -> [TrailingAnn] -> EpAnnComments -> Entry +mkEntry :: EpaLocation -> [TrailingAnn] -> EpAnnComments -> Entry mkEntry anc ts cs = Entry anc ts cs NoFlushComments CanUpdateAnchor instance (HasTrailing a) => HasEntry (EpAnn a) where @@ -642,7 +642,7 @@ withPpr a = do -- 'ppr'. class (Typeable a) => ExactPrint a where getAnnotationEntry :: a -> Entry - setAnnotationAnchor :: a -> Anchor -> [TrailingAnn] -> EpAnnComments -> a + setAnnotationAnchor :: a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> a exact :: (Monad m, Monoid w) => a -> EP w m a -- --------------------------------------------------------------------- @@ -4277,7 +4277,7 @@ instance ExactPrint (LocatedN RdrName) where locFromAdd :: AddEpAnn -> EpaLocation locFromAdd (AddEpAnn _ loc) = loc -printUnicode :: (Monad m, Monoid w) => Anchor -> RdrName -> EP w m Anchor +printUnicode :: (Monad m, Monoid w) => EpaLocation -> RdrName -> EP w m EpaLocation printUnicode anc n = do let str = case (showPprUnsafe n) of -- TODO: unicode support? @@ -4977,10 +4977,10 @@ setPosP l = do debugM $ "setPosP:" ++ show l modify (\s -> s {epPos = l}) -getExtraDP :: (Monad m, Monoid w) => EP w m (Maybe Anchor) +getExtraDP :: (Monad m, Monoid w) => EP w m (Maybe EpaLocation) getExtraDP = gets uExtraDP -setExtraDP :: (Monad m, Monoid w) => Maybe Anchor -> EP w m () +setExtraDP :: (Monad m, Monoid w) => Maybe EpaLocation -> EP w m () setExtraDP md = do debugM $ "setExtraDP:" ++ show md modify (\s -> s {uExtraDP = md}) ===================================== utils/check-exact/Utils.hs ===================================== @@ -255,7 +255,7 @@ tokComment t@(L lt c) = (GHC.EpaComment (EpaDocComment dc) pt) -> hsDocStringComments (noCommentsToEpaLocation lt) pt dc _ -> [mkComment (normaliseCommentText (ghcCommentText t)) lt (ac_prior_tok c)] -hsDocStringComments :: Anchor -> RealSrcSpan -> GHC.HsDocString -> [Comment] +hsDocStringComments :: EpaLocation -> RealSrcSpan -> GHC.HsDocString -> [Comment] hsDocStringComments _ pt (MultiLineDocString dec (x :| xs)) = let decStr = printDecorator dec @@ -342,7 +342,7 @@ isKWComment c = isJust (commentOrigin c) noKWComments :: [Comment] -> [Comment] noKWComments = filter (\c -> not (isKWComment c)) -sortAnchorLocated :: [GenLocated Anchor a] -> [GenLocated Anchor a] +sortAnchorLocated :: [GenLocated EpaLocation a] -> [GenLocated EpaLocation a] sortAnchorLocated = sortBy (compare `on` (anchor . getLoc)) -- | Calculates the distance from the start of a string to the end of @@ -401,12 +401,6 @@ setTrailingAnnLoc (AddDarrowUAnn _) ss = (AddDarrowUAnn ss) addEpAnnLoc :: AddEpAnn -> EpaLocation addEpAnnLoc (AddEpAnn _ l) = l --- --------------------------------------------------------------------- - --- TODO: get rid of this identity function -anchorToEpaLocation :: Anchor -> EpaLocation -anchorToEpaLocation a = a - -- --------------------------------------------------------------------- -- Horrible hack for dealing with some things still having a SrcSpan, -- not an Anchor. @@ -432,7 +426,7 @@ To be absolutely sure, we make the delta versions use -ve values. -} -hackSrcSpanToAnchor :: SrcSpan -> Anchor +hackSrcSpanToAnchor :: SrcSpan -> EpaLocation hackSrcSpanToAnchor (UnhelpfulSpan s) = error $ "hackSrcSpanToAnchor : UnhelpfulSpan:" ++ show s hackSrcSpanToAnchor ss@(RealSrcSpan r mb) = case mb of @@ -443,7 +437,7 @@ hackSrcSpanToAnchor ss@(RealSrcSpan r mb) else EpaSpan (RealSrcSpan r mb) _ -> EpaSpan (RealSrcSpan r mb) -hackAnchorToSrcSpan :: Anchor -> SrcSpan +hackAnchorToSrcSpan :: EpaLocation -> SrcSpan hackAnchorToSrcSpan (EpaSpan s) = s hackAnchorToSrcSpan _ = error $ "hackAnchorToSrcSpan" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9788163f832c5bc73939b9d8463f891369b4ec22 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9788163f832c5bc73939b9d8463f891369b4ec22 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 9 23:46:24 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 09 Sep 2024 19:46:24 -0400 Subject: [Git][ghc/ghc][wip/andreask/setNonBlockingMode] Allow unknown fd device types for setNonBlockingMode. Message-ID: <66df88d0d9d15_39d61393a148761d@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/setNonBlockingMode at Glasgow Haskell Compiler / GHC Commits: 883eb9d9 by Andreas Klebinger at 2024-09-10T01:27:22+02:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - 7 changed files: - libraries/base/changelog.md - libraries/ghc-internal/src/GHC/Internal/IO/FD.hs - libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/interface-stability/base-exports.stdout-ws-32 Changes: ===================================== libraries/base/changelog.md ===================================== @@ -16,6 +16,7 @@ * Add `firstA` and `secondA` to `Data.Bitraversable`. ([CLC proposal #172](https://github.com/haskell/core-libraries-committee/issues/172)) * Deprecate `GHC.TypeNats.Internal`, `GHC.TypeLits.Internal`, `GHC.ExecutionStack.Internal` ([CLC proposal #217](https://github.com/haskell/core-libraries-committee/issues/217)) * Define `Eq1`, `Ord1`, `Show1` and `Read1` instances for basic `Generic` representation types. ([CLC proposal #273](https://github.com/haskell/core-libraries-committee/issues/273)) + * `setNonBlockingMode` will no longer throw an exception when called on a FD associated with a unknown device type. ([CLC proposal #282](https://github.com/haskell/core-libraries-committee/issues/282)) * Add exception type metadata to default exception handler output. ([CLC proposal #231](https://github.com/haskell/core-libraries-committee/issues/231) and [CLC proposal #261](https://github.com/haskell/core-libraries-committee/issues/261)) ===================================== libraries/ghc-internal/src/GHC/Internal/IO/FD.hs ===================================== @@ -57,6 +57,7 @@ import GHC.Internal.Foreign.Storable import GHC.Internal.Foreign.C.Types import GHC.Internal.Foreign.C.Error import GHC.Internal.Foreign.Marshal.Utils +import GHC.Internal.Foreign.Marshal.Alloc (allocaBytes) import qualified GHC.Internal.System.Posix.Internals import GHC.Internal.System.Posix.Internals hiding (FD, setEcho, getEcho) @@ -466,10 +467,15 @@ setNonBlockingMode fd set = do -- O_NONBLOCK has no effect on regular files and block devices; -- utilities inspecting fdIsNonBlocking (such as readRawBufferPtr) -- should not be tricked to think otherwise. - is_nonblock <- if set then do - (fd_type, _, _) <- fdStat (fdFD fd) - pure $ fd_type /= RegularFile && fd_type /= RawDevice - else pure False + is_nonblock <- + if set + then do + allocaBytes sizeof_stat $ \ p_stat -> do + throwErrnoIfMinus1Retry_ "fileSize" $ + c_fstat (fdFD fd) p_stat + fd_type <- statGetType_maybe p_stat + pure $ fd_type /= Just RegularFile && fd_type /= Just RawDevice + else pure False setNonBlockingFD (fdFD fd) is_nonblock #if defined(mingw32_HOST_OS) return fd ===================================== libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs ===================================== @@ -140,17 +140,30 @@ fdStat fd = fdType :: FD -> IO IODeviceType fdType fd = do (ty,_,_) <- fdStat fd; return ty +-- | Return a known device type or throw an exception if the device +-- type is unknown. statGetType :: Ptr CStat -> IO IODeviceType statGetType p_stat = do + dev_ty_m <- statGetType_maybe p_stat + case dev_ty_m of + Nothing -> ioError ioe_unknownfiletype + Just dev_ty -> pure dev_ty + +-- | Unlike @statGetType@, @statGetType_maybe@ will not throw an exception +-- if the CStat refers to a unknown device type. +-- +-- @since base-4.21.0.0 +statGetType_maybe :: Ptr CStat -> IO (Maybe IODeviceType) +statGetType_maybe p_stat = do c_mode <- st_mode p_stat :: IO CMode case () of - _ | s_isdir c_mode -> return Directory + _ | s_isdir c_mode -> return $ Just Directory | s_isfifo c_mode || s_issock c_mode || s_ischr c_mode - -> return Stream - | s_isreg c_mode -> return RegularFile + -> return $ Just Stream + | s_isreg c_mode -> return $ Just RegularFile -- Q: map char devices to RawDevice too? - | s_isblk c_mode -> return RawDevice - | otherwise -> ioError ioe_unknownfiletype + | s_isblk c_mode -> return $ Just RawDevice + | otherwise -> return Nothing ioe_unknownfiletype :: IOException ioe_unknownfiletype = IOError Nothing UnsupportedOperation "fdType" ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -10549,6 +10549,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10596,6 +10597,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs ===================================== @@ -13590,6 +13590,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -13637,6 +13638,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -10821,6 +10821,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10865,6 +10866,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.CWString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-ws-32 ===================================== @@ -10549,6 +10549,7 @@ module System.Posix.Internals where fdFileSize :: FD -> GHC.Types.IO GHC.Num.Integer.Integer fdGetMode :: FD -> GHC.Types.IO GHC.Internal.IO.IOMode.IOMode fdStat :: FD -> GHC.Types.IO (GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) + fdStat_maybe :: FD -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType, GHC.Internal.System.Posix.Types.CDev, GHC.Internal.System.Posix.Types.CIno) fdType :: FD -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType fileType :: GHC.Internal.IO.FilePath -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType getEcho :: FD -> GHC.Types.IO GHC.Types.Bool @@ -10596,6 +10597,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/883eb9d9268601b94043a9bc1a5ca9c15ff43b43 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/883eb9d9268601b94043a9bc1a5ca9c15ff43b43 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 00:55:41 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 09 Sep 2024 20:55:41 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/andreask/base_c_fstat_refactor Message-ID: <66df990dc487e_2be1ce23e72c25682@gitlab.mail> Andreas Klebinger pushed new branch wip/andreask/base_c_fstat_refactor at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/andreask/base_c_fstat_refactor You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 04:41:10 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 00:41:10 -0400 Subject: [Git][ghc/ghc][master] 3 commits: Refactor only newSysLocalDs Message-ID: <66dfcde61ef4d_2fcb413039141211f@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - 30 changed files: - compiler/GHC/Core/InstEnv.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Dict.hs - compiler/GHC/Tc/Solver/Equality.hs - compiler/GHC/Tc/Solver/Irred.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Solver/Solve.hs - compiler/GHC/Tc/TyCl/Instance.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Utils/Instantiate.hs - compiler/GHC/Tc/Utils/Unify.hs - testsuite/tests/indexed-types/should_compile/Simple14.hs - − testsuite/tests/indexed-types/should_compile/Simple14.stderr - testsuite/tests/indexed-types/should_compile/all.T - + testsuite/tests/simplCore/should_compile/T25160.hs - + testsuite/tests/simplCore/should_compile/T25160.stderr - testsuite/tests/simplCore/should_compile/all.T The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5bf0e6bc3e3024c289cb637a19c9660c848e831f...663daf8dfaf878ee5463d56666b6d5e8670f33d8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5bf0e6bc3e3024c289cb637a19c9660c848e831f...663daf8dfaf878ee5463d56666b6d5e8670f33d8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 04:41:37 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 00:41:37 -0400 Subject: [Git][ghc/ghc][master] Don't name a binding pattern Message-ID: <66dfce015fa10_2fcb4139aef415222@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - 1 changed file: - compiler/GHC/Core/Opt/CallerCC/Types.hs Changes: ===================================== compiler/GHC/Core/Opt/CallerCC/Types.hs ===================================== @@ -61,14 +61,14 @@ occNameMatches pat = go pat . occNameString type Parser = P.ReadP parseNamePattern :: Parser NamePattern -parseNamePattern = pattern +parseNamePattern = namePattern where - pattern = star P.<++ wildcard P.<++ char P.<++ end - star = PChar '*' <$ P.string "\\*" <*> pattern + namePattern = star P.<++ wildcard P.<++ char P.<++ end + star = PChar '*' <$ P.string "\\*" <*> namePattern wildcard = do void $ P.char '*' - PWildcard <$> pattern - char = PChar <$> P.get <*> pattern + PWildcard <$> namePattern + char = PChar <$> P.get <*> namePattern end = PEnd <$ P.eof data CallerCcFilter @@ -110,4 +110,4 @@ parseCallerCcFilter' = c <- P.satisfy isUpper cs <- P.munch1 (\c -> isUpper c || isLower c || isDigit c || c == '_') rest <- optional $ P.char '.' >> fmap ('.':) moduleName - return $ c : (cs ++ fromMaybe "" rest) \ No newline at end of file + return $ c : (cs ++ fromMaybe "" rest) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e28cd0210ba5f380cf4da96c581043c19e629b44 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e28cd0210ba5f380cf4da96c581043c19e629b44 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 04:42:20 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 00:42:20 -0400 Subject: [Git][ghc/ghc][master] Do not use an error thunk for an absent dictionary Message-ID: <66dfce2cad2a0_2fcb415504ec185ab@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 1 changed file: - compiler/GHC/Core/Opt/WorkWrap/Utils.hs Changes: ===================================== compiler/GHC/Core/Opt/WorkWrap/Utils.hs ===================================== @@ -29,6 +29,7 @@ import GHC.Core.Subst import GHC.Core.Type import GHC.Core.Multiplicity import GHC.Core.Coercion +import GHC.Core.Predicate( isDictTy ) import GHC.Core.Reduction import GHC.Core.FamInstEnv import GHC.Core.TyCon @@ -1007,21 +1008,25 @@ unbox_one_arg opts arg_var -- same type as @id at . Otherwise, no suitable filler could be found. mkAbsentFiller :: WwOpts -> Id -> StrictnessMark -> Maybe CoreExpr mkAbsentFiller opts arg str - -- The lifted case: Bind 'absentError' for a nice panic message if we are - -- wrong (like we were in #11126). See (1) in Note [Absent fillers] + -- The lifted case: bind 'absentError'. See (AF1) in Note [Absent fillers] + -- We want to use this case if possible, because we get a nice runtime panic message + -- if we are wrong (like we were in #11126). Otherwise we fall through to the + -- less-desirable mkLitRubbish case. | mightBeLiftedType arg_ty - , not is_strict - , not (isMarkedStrict str) -- See (2) in Note [Absent fillers] + , not (isDictTy arg_ty) -- See (AF4) in Note [Absent fillers] + , not (isStrictDmd (idDemandInfo arg)) -- See (AF2) + , not (isMarkedStrict str) -- in Note [Absent fillers] = Just (mkAbsentErrorApp arg_ty msg) -- The default case for mono rep: Bind `RUBBISH[rr] arg_ty` - -- See Note [Absent fillers], the main part + -- See Note [Absent fillers] + -- (AF3): mkLitRubbish returns Nothing if the representation is not + -- monomorphic, in which case we can't make a filler | otherwise = mkLitRubbish arg_ty where - arg_ty = idType arg - is_strict = isStrictDmd (idDemandInfo arg) + arg_ty = idType arg msg = renderWithContext (defaultSDocContext { sdocSuppressUniques = True }) @@ -1184,7 +1189,7 @@ conjure filler values at any type (and any representation or levity!). Needless to say, there are some wrinkles: - 1. In case we have a absent, /lazy/, and /lifted/ arg, we use an error-thunk +(AF1) In case we have a absent, /lazy/, and /lifted/ arg, we use an error-thunk instead. If absence analysis was wrong (e.g., #11126) and the binding in fact is used, then we get a nice panic message instead of undefined runtime behavior (See Modes of failure from Note [Rubbish literals]). @@ -1192,7 +1197,7 @@ Needless to say, there are some wrinkles: Obviously, we can't use an error-thunk if the value is of unlifted rep (like 'Int#' or 'MutVar#'), because we'd immediately evaluate the panic. - 2. We also mustn't put an error-thunk (that fills in for an absent value of +(AF2) We also mustn't put an error-thunk (that fills in for an absent value of lifted rep) in a strict field, because #16970 establishes the invariant that strict fields are always evaluated, by possibly (re-)evaluating what is put in a strict field. That's the reason why 'zs' binds a rubbish literal instead @@ -1217,8 +1222,8 @@ Needless to say, there are some wrinkles: in place on top of threading through the marks from the constructor. It's a *really* cheap and easy check to make anyway. - 3. We can only emit a LitRubbish if the arg's type @arg_ty@ is mono-rep, e.g. - of the form @TYPE rep@ where @rep@ is not (and doesn't contain) a variable. +(AF3) We can only emit a LitRubbish if the arg's type `arg_ty` is mono-rep, e.g. + of the form `TYPE rep` where `rep` is not (and doesn't contain) a variable. Why? Because if we don't know its representation (e.g. size in memory, register class), we don't know what or how much rubbish to emit in codegen. 'mkLitRubbish' returns 'Nothing' in this case and we simply fall @@ -1228,8 +1233,30 @@ Needless to say, there are some wrinkles: have to be representation monomorphic. But in the future, we might allow levity polymorphism, e.g. a polymorphic levity variable in 'BoxedRep'. -While (1) and (2) are simply an optimisation in terms of compiler debugging -experience, (3) should be irrelevant in most programs, if not all. +(AF4) Consider (#24934) + f :: (a~b) => blah {-# INLINE f #-} + f d x = case eq_sel d of co -> body + In #24934 it turned out that `co` was unused; and we discarded the + entire case-scrutinisation via the `exprOkToDiscard` test in + `GHC.Core.Opt.Simplify.Iteration.rebuildCase`. So now `d` is absent. + But in the /unfolding/ for some reason we did not discard the `case`; + so when we inline `f` we end up evaluating that `d` argument. So we had + better not replace it with an error thunk! + + The root of it is this: `exprOkToDiscard` assumes that a dictionary is + non-bottom (Note [exprOkForSpeculation and type classes]); but then we replace + the (a~b) dictionary with an error thunk, breaking the invariant that every + dictionary is non-bottom. (If -XDictsStrict is on, the invariant is even + more important.) + + Simple solution: never use an error thunk for a dictionary; instead fall + through to mkRubbishLit. (The only downside is that we lose the compiler + debugging advantages of (AF1).) + + This is quite delicate. + +While (AF1) and (AF2) are simply an optimisation in terms of compiler debugging +experience, (AF3) should be irrelevant in most programs, if not all. Historical note: I did try the experiment of using an error thunk for unlifted things too, relying on the simplifier to drop it as dead code. But this is View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/b09571e2055c24daffa170e74fe8ef424285307e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/b09571e2055c24daffa170e74fe8ef424285307e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 04:42:58 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 00:42:58 -0400 Subject: [Git][ghc/ghc][master] haddock: Remove support for applehelp format in the Manual Message-ID: <66dfce5274719_2fcb41a145dc216fe@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 1 changed file: - utils/haddock/doc/requirements.txt Changes: ===================================== utils/haddock/doc/requirements.txt ===================================== @@ -1 +0,0 @@ -sphinxcontrib-applehelp ==2.0.0 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8bc9f5f6c18a6788fbd69d2879b6b355b2515b97 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8bc9f5f6c18a6788fbd69d2879b6b355b2515b97 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 08:26:00 2024 From: gitlab at gitlab.haskell.org (Sylvain Henry (@hsyl20)) Date: Tue, 10 Sep 2024 04:26:00 -0400 Subject: [Git][ghc/ghc][wip/libffi-dep-hadrian] 10 commits: JS: fake support for native adjustors (#25159) Message-ID: <66e00298533e8_221cfc840df014339@gitlab.mail> Sylvain Henry pushed to branch wip/libffi-dep-hadrian at Glasgow Haskell Compiler / GHC Commits: 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - e3ff1b67 by Matthew Pickering at 2024-09-10T10:23:40+02:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 - - - - - 86e78903 by Sylvain Henry at 2024-09-10T10:25:36+02:00 Try fixing Wasm - - - - - 30 changed files: - compiler/GHC/Core/InstEnv.hs - compiler/GHC/Core/Opt/CallerCC/Types.hs - compiler/GHC/Core/Opt/WorkWrap/Utils.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Dict.hs - compiler/GHC/Tc/Solver/Equality.hs - compiler/GHC/Tc/Solver/Irred.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Solver/Solve.hs - compiler/GHC/Tc/TyCl/Instance.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Utils/Instantiate.hs - compiler/GHC/Tc/Utils/Unify.hs - hadrian/src/Builder.hs - libraries/ghc-internal/jsbits/base.js - m4/ghc_adjustors_method.m4 - testsuite/tests/indexed-types/should_compile/Simple14.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5cddf22a03bb381a645d3ca63fab07a309311ffe...86e7890321477c72a20abc9c5c4d3dc39ee8bfea -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5cddf22a03bb381a645d3ca63fab07a309311ffe...86e7890321477c72a20abc9c5c4d3dc39ee8bfea You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 08:53:58 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Tue, 10 Sep 2024 04:53:58 -0400 Subject: [Git][ghc/ghc][wip/T24817a] 9 commits: JS: fake support for native adjustors (#25159) Message-ID: <66e009266d6d7_221cfcae06f027946@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24817a at Glasgow Haskell Compiler / GHC Commits: 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 24f5ac2d by Simon Peyton Jones at 2024-09-10T09:53:42+01:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 30 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Core/InstEnv.hs - compiler/GHC/Core/Opt/CallerCC/Types.hs - compiler/GHC/Core/Opt/WorkWrap/Utils.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Dict.hs - compiler/GHC/Tc/Solver/Equality.hs - compiler/GHC/Tc/Solver/Irred.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Solver/Solve.hs - compiler/GHC/Tc/TyCl/Instance.hs - compiler/GHC/Tc/Types.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Utils/Instantiate.hs - compiler/GHC/Tc/Utils/Monad.hs - compiler/GHC/Tc/Utils/Unify.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/68b9f8063ff7f787d639ef530d3c1b9357333e08...24f5ac2dbf562c7840ef99cd330ef4244b37f74e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/68b9f8063ff7f787d639ef530d3c1b9357333e08...24f5ac2dbf562c7840ef99cd330ef4244b37f74e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 09:46:42 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 05:46:42 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 9 commits: Refactor only newSysLocalDs Message-ID: <66e01581dc92c_1c6fe1c24e8895d5@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 49c0fa64 by doyougnu at 2024-09-10T05:46:25-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 696e8dc3 by Sven Tennie at 2024-09-10T05:46:26-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 24952620 by Sylvain Henry at 2024-09-10T05:46:29-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 30 changed files: - compiler/GHC/Core/InstEnv.hs - compiler/GHC/Core/Opt/CallerCC/Types.hs - compiler/GHC/Core/Opt/WorkWrap/Utils.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/StgToJS/Symbols.hs - compiler/GHC/StgToJS/Types.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Dict.hs - compiler/GHC/Tc/Solver/Equality.hs - compiler/GHC/Tc/Solver/Irred.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Solver/Solve.hs - compiler/GHC/Tc/TyCl/Instance.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Utils/Instantiate.hs - compiler/GHC/Tc/Utils/Unify.hs - libraries/ghc-heap/tests/stack_misc_closures_c.c - rts/Linker.c The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0b9cdbab4c6882ac1da6bbaacb4c9249b6c0d27a...24952620517344cc2def2c7ee0ee78e76b105980 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0b9cdbab4c6882ac1da6bbaacb4c9249b6c0d27a...24952620517344cc2def2c7ee0ee78e76b105980 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 10:05:36 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Tue, 10 Sep 2024 06:05:36 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] Try to fix CCallConv test Message-ID: <66e019f02076b_1c6fe12af120990fd@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 5ddc125e by Sven Tennie at 2024-09-10T12:04:23+02:00 Try to fix CCallConv test - Enforce order of thunks - Use better test config - Flush stdout - - - - - 4 changed files: - testsuite/tests/codeGen/should_run/CCallConv.hs - testsuite/tests/codeGen/should_run/CCallConv.stdout - testsuite/tests/codeGen/should_run/CCallConv_c.c - testsuite/tests/codeGen/should_run/all.T Changes: ===================================== testsuite/tests/codeGen/should_run/CCallConv.hs ===================================== @@ -15,6 +15,7 @@ module Main where import Data.Word import GHC.Exts import GHC.Int +import System.IO foreign import ccall "fun8" fun8 :: @@ -87,34 +88,42 @@ foreign import ccall "funDouble" Double# -- result main :: IO () -main = +main = do -- N.B. the values here aren't choosen by accident: -1 means all bits one in -- twos-complement, which is the same as the max word value. let i8 :: Int8# = intToInt8# (-1#) w8 :: Word8# = wordToWord8# (255##) res8 :: Int64# = fun8 i8 w8 i8 i8 i8 i8 i8 i8 w8 i8 expected_res8 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word8) + 8 * (-1) - i16 :: Int16# = intToInt16# (-1#) + print $ "fun8 result:" ++ show (I64# res8) + hFlush stdout + assertEqual expected_res8 (I64# res8) + + let i16 :: Int16# = intToInt16# (-1#) w16 :: Word16# = wordToWord16# (65535##) res16 :: Int64# = fun16 i16 w16 i16 i16 i16 i16 i16 i16 w16 i16 expected_res16 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word16) + 8 * (-1) - i32 :: Int32# = intToInt32# (-1#) + print $ "fun16 result:" ++ show (I64# res16) + hFlush stdout + assertEqual expected_res16 (I64# res16) + + let i32 :: Int32# = intToInt32# (-1#) w32 :: Word32# = wordToWord32# (4294967295##) res32 :: Int64# = fun32 i32 w32 i32 i32 i32 i32 i32 i32 w32 i32 expected_res32 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word32) + 8 * (-1) - resFloat :: Float = F# (funFloat 1.0# 1.1# 1.2# 1.3# 1.4# 1.5# 1.6# 1.7# 1.8# 1.9#) - resDouble :: Double = D# (funDouble 1.0## 1.1## 1.2## 1.3## 1.4## 1.5## 1.6## 1.7## 1.8## 1.9##) - in do - print $ "fun8 result:" ++ show (I64# res8) - assertEqual expected_res8 (I64# res8) - print $ "fun16 result:" ++ show (I64# res16) - assertEqual expected_res16 (I64# res16) - print $ "fun32 result:" ++ show (I64# res32) - assertEqual expected_res32 (I64# res32) - print $ "funFloat result:" ++ show resFloat - assertEqual (14.5 :: Float) resFloat - print $ "funDouble result:" ++ show resDouble - assertEqual (14.5 :: Double) resDouble + print $ "fun32 result:" ++ show (I64# res32) + hFlush stdout + assertEqual expected_res32 (I64# res32) + + let resFloat :: Float = F# (funFloat 1.0# 1.1# 1.2# 1.3# 1.4# 1.5# 1.6# 1.7# 1.8# 1.9#) + print $ "funFloat result:" ++ show resFloat + hFlush stdout + assertEqual (14.5 :: Float) resFloat + + let resDouble :: Double = D# (funDouble 1.0## 1.1## 1.2## 1.3## 1.4## 1.5## 1.6## 1.7## 1.8## 1.9##) + print $ "funDouble result:" ++ show resDouble + hFlush stdout + assertEqual (14.5 :: Double) resDouble assertEqual :: (Eq a, Show a) => a -> a -> IO () assertEqual a b = ===================================== testsuite/tests/codeGen/should_run/CCallConv.stdout ===================================== @@ -1,11 +1,6 @@ -"fun8 result:502" -"fun16 result:131062" -"fun32 result:8589934582" -"funFloat result:14.5" -"funDouble result:14.5" -fun32: +fun8: a0: 0xffffffff -1 -a1: 0xffffffff 4294967295 +a1: 0xff 255 a2: 0xffffffff -1 a3: 0xffffffff -1 a4: 0xffffffff -1 @@ -13,7 +8,8 @@ a5: 0xffffffff -1 a6: 0xffffffff -1 a7: 0xffffffff -1 s0: 0xffffffff -1 -s1: 0xffffffff 4294967295 +s1: 0xff 255 +"fun8 result:502" fun16: a0: 0xffffffff -1 a1: 0xffff 65535 @@ -25,9 +21,10 @@ a6: 0xffffffff -1 a7: 0xffffffff -1 s0: 0xffffffff -1 s1: 0xffff 65535 -fun8: +"fun16 result:131062" +fun32: a0: 0xffffffff -1 -a1: 0xff 255 +a1: 0xffffffff 4294967295 a2: 0xffffffff -1 a3: 0xffffffff -1 a4: 0xffffffff -1 @@ -35,7 +32,8 @@ a5: 0xffffffff -1 a6: 0xffffffff -1 a7: 0xffffffff -1 s0: 0xffffffff -1 -s1: 0xff 255 +s1: 0xffffffff 4294967295 +"fun32 result:8589934582" funFloat: a0: 1.000000 a1: 1.100000 @@ -47,6 +45,7 @@ a6: 1.600000 a7: 1.700000 s0: 1.800000 s1: 1.900000 +"funFloat result:14.5" funDouble: a0: 1.000000 a1: 1.100000 @@ -58,3 +57,4 @@ a6: 1.600000 a7: 1.700000 s0: 1.800000 s1: 1.900000 +"funDouble result:14.5" ===================================== testsuite/tests/codeGen/should_run/CCallConv_c.c ===================================== @@ -1,5 +1,5 @@ -#include "stdint.h" -#include "stdio.h" +#include +#include int64_t fun8(int8_t a0, uint8_t a1, int8_t a2, int8_t a3, int8_t a4, int8_t a5, int8_t a6, int8_t a7, int8_t s0, uint8_t s1) { @@ -15,6 +15,8 @@ int64_t fun8(int8_t a0, uint8_t a1, int8_t a2, int8_t a3, int8_t a4, int8_t a5, printf("s0: %#x %hhd\n", s0, s0); printf("s1: %#x %hhu\n", s1, s1); + fflush(stdout); + return a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + s1; } @@ -32,6 +34,8 @@ int64_t fun16(int16_t a0, uint16_t a1, int16_t a2, int16_t a3, int16_t a4, printf("s0: %#x %hd\n", s0, s0); printf("s1: %#x %hu\n", s1, s1); + fflush(stdout); + return a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + s1; } @@ -49,6 +53,8 @@ int64_t fun32(int32_t a0, uint32_t a1, int32_t a2, int32_t a3, int32_t a4, printf("s0: %#x %d\n", s0, s0); printf("s1: %#x %u\n", s1, s1); + fflush(stdout); + // Ensure the addition happens in long int (not just int) precission. // Otherwise, the result is truncated during the operation. int64_t force_int64_precission = 0; @@ -70,6 +76,8 @@ float funFloat(float a0, float a1, float a2, float a3, float a4, float a5, printf("s0: %f\n", s0); printf("s1: %f\n", s1); + fflush(stdout); + return a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + s1; } @@ -87,5 +95,7 @@ double funDouble(double a0, double a1, double a2, double a3, double a4, double a printf("s0: %f\n", s0); printf("s1: %f\n", s1); + fflush(stdout); + return a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + s1; } ===================================== testsuite/tests/codeGen/should_run/all.T ===================================== @@ -251,7 +251,4 @@ test('T23034', req_c, compile_and_run, ['-O2 T23034_c.c']) test('T24700', normal, compile_and_run, ['-O']) test('T24893', normal, compile_and_run, ['-O']) -test('CCallConv', - [], - multi_compile_and_run, - ['CCallConv', [('CCallConv_c.c', '')], '']) +test('CCallConv', req_c, compile_and_run, ['CCallConv_c.c']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5ddc125e968a6fb19c90430523312f7c781e5631 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5ddc125e968a6fb19c90430523312f7c781e5631 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 10:27:09 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Tue, 10 Sep 2024 06:27:09 -0400 Subject: [Git][ghc/ghc][wip/andreask/setNonBlockingMode] Allow unknown fd device types for setNonBlockingMode. Message-ID: <66e01efd785cc_1c6fe13e0b0c106517@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/setNonBlockingMode at Glasgow Haskell Compiler / GHC Commits: 90778487 by Andreas Klebinger at 2024-09-10T12:08:26+02:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - 7 changed files: - libraries/base/changelog.md - libraries/ghc-internal/src/GHC/Internal/IO/FD.hs - libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/interface-stability/base-exports.stdout-ws-32 Changes: ===================================== libraries/base/changelog.md ===================================== @@ -16,6 +16,7 @@ * Add `firstA` and `secondA` to `Data.Bitraversable`. ([CLC proposal #172](https://github.com/haskell/core-libraries-committee/issues/172)) * Deprecate `GHC.TypeNats.Internal`, `GHC.TypeLits.Internal`, `GHC.ExecutionStack.Internal` ([CLC proposal #217](https://github.com/haskell/core-libraries-committee/issues/217)) * Define `Eq1`, `Ord1`, `Show1` and `Read1` instances for basic `Generic` representation types. ([CLC proposal #273](https://github.com/haskell/core-libraries-committee/issues/273)) + * `setNonBlockingMode` will no longer throw an exception when called on a FD associated with a unknown device type. ([CLC proposal #282](https://github.com/haskell/core-libraries-committee/issues/282)) * Add exception type metadata to default exception handler output. ([CLC proposal #231](https://github.com/haskell/core-libraries-committee/issues/231) and [CLC proposal #261](https://github.com/haskell/core-libraries-committee/issues/261)) ===================================== libraries/ghc-internal/src/GHC/Internal/IO/FD.hs ===================================== @@ -57,6 +57,7 @@ import GHC.Internal.Foreign.Storable import GHC.Internal.Foreign.C.Types import GHC.Internal.Foreign.C.Error import GHC.Internal.Foreign.Marshal.Utils +import GHC.Internal.Foreign.Marshal.Alloc (allocaBytes) import qualified GHC.Internal.System.Posix.Internals import GHC.Internal.System.Posix.Internals hiding (FD, setEcho, getEcho) @@ -466,10 +467,15 @@ setNonBlockingMode fd set = do -- O_NONBLOCK has no effect on regular files and block devices; -- utilities inspecting fdIsNonBlocking (such as readRawBufferPtr) -- should not be tricked to think otherwise. - is_nonblock <- if set then do - (fd_type, _, _) <- fdStat (fdFD fd) - pure $ fd_type /= RegularFile && fd_type /= RawDevice - else pure False + is_nonblock <- + if set + then do + allocaBytes sizeof_stat $ \ p_stat -> do + throwErrnoIfMinus1Retry_ "fileSize" $ + c_fstat (fdFD fd) p_stat + fd_type <- statGetType_maybe p_stat + pure $ fd_type /= Just RegularFile && fd_type /= Just RawDevice + else pure False setNonBlockingFD (fdFD fd) is_nonblock #if defined(mingw32_HOST_OS) return fd ===================================== libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs ===================================== @@ -140,17 +140,30 @@ fdStat fd = fdType :: FD -> IO IODeviceType fdType fd = do (ty,_,_) <- fdStat fd; return ty +-- | Return a known device type or throw an exception if the device +-- type is unknown. statGetType :: Ptr CStat -> IO IODeviceType statGetType p_stat = do + dev_ty_m <- statGetType_maybe p_stat + case dev_ty_m of + Nothing -> ioError ioe_unknownfiletype + Just dev_ty -> pure dev_ty + +-- | Unlike @statGetType@, @statGetType_maybe@ will not throw an exception +-- if the CStat refers to a unknown device type. +-- +-- @since base-4.21.0.0 +statGetType_maybe :: Ptr CStat -> IO (Maybe IODeviceType) +statGetType_maybe p_stat = do c_mode <- st_mode p_stat :: IO CMode case () of - _ | s_isdir c_mode -> return Directory + _ | s_isdir c_mode -> return $ Just Directory | s_isfifo c_mode || s_issock c_mode || s_ischr c_mode - -> return Stream - | s_isreg c_mode -> return RegularFile + -> return $ Just Stream + | s_isreg c_mode -> return $ Just RegularFile -- Q: map char devices to RawDevice too? - | s_isblk c_mode -> return RawDevice - | otherwise -> ioError ioe_unknownfiletype + | s_isblk c_mode -> return $ Just RawDevice + | otherwise -> return Nothing ioe_unknownfiletype :: IOException ioe_unknownfiletype = IOError Nothing UnsupportedOperation "fdType" ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -10596,6 +10596,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs ===================================== @@ -13637,6 +13637,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -10865,6 +10865,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.CWString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-ws-32 ===================================== @@ -10596,6 +10596,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/90778487224c2db434772cc7dd3e649e629df076 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/90778487224c2db434772cc7dd3e649e629df076 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 10:52:40 2024 From: gitlab at gitlab.haskell.org (Sylvain Henry (@hsyl20)) Date: Tue, 10 Sep 2024 06:52:40 -0400 Subject: [Git][ghc/ghc][wip/libffi-dep-hadrian] hadrian: Make sure ffi headers are built before using a compiler Message-ID: <66e024f8a4c09_1c6fe157ec20108011@gitlab.mail> Sylvain Henry pushed to branch wip/libffi-dep-hadrian at Glasgow Haskell Compiler / GHC Commits: ba633b99 by Matthew Pickering at 2024-09-10T12:51:34+02:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 1 changed file: - hadrian/src/Builder.hs Changes: ===================================== hadrian/src/Builder.hs ===================================== @@ -237,16 +237,25 @@ instance H.Builder Builder where -- changes (#18001). _bootGhcVersion <- setting GhcVersion pure [] - Ghc {} -> do + Ghc _ st -> do root <- buildRoot unlitPath <- builderPath Unlit distro_mingw <- settingsFileSetting ToolchainSetting_DistroMinGW + libffi_adjustors <- useLibffiForAdjustors + use_system_ffi <- flag UseSystemFfi return $ [ unlitPath ] ++ [ root -/- mingwStamp | windowsHost, distro_mingw == "NO" ] -- proxy for the entire mingw toolchain that -- we have in inplace/mingw initially, and then at -- root -/- mingw. + -- ffi.h needed by the compiler when using libffi_adjustors (#24864) + -- It would be nicer to not duplicate this logic between here + -- and needRtsLibffiTargets and libffiHeaderFiles but this doesn't change + -- very often. + ++ [ root -/- buildDir (rtsContext st) -/- "include" -/- header + | header <- ["ffi.h", "ffitarget.h"] + , libffi_adjustors && not use_system_ffi ] Hsc2Hs stage -> (\p -> [p]) <$> templateHscPath stage Make dir -> return [dir -/- "Makefile"] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ba633b9916c455995734b290399d28271ab9c548 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ba633b9916c455995734b290399d28271ab9c548 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 12:06:35 2024 From: gitlab at gitlab.haskell.org (Sjoerd Visscher (@trac-sjoerd_visscher)) Date: Tue, 10 Sep 2024 08:06:35 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/finder-boot] finder: Add `IsBootInterface` to finder cache keys Message-ID: <66e0364b42384_3d42f13d50f4546f0@gitlab.mail> Sjoerd Visscher pushed to branch wip/torsten.schmits/finder-boot at Glasgow Haskell Compiler / GHC Commits: c3c3970b by Torsten Schmits at 2024-09-10T14:06:24+02:00 finder: Add `IsBootInterface` to finder cache keys - - - - - 12 changed files: - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Finder/Types.hs - compiler/GHC/Unit/Module/Env.hs - compiler/GHC/Unit/Types.hs - + testsuite/tests/driver/boot-target/A.hs - + testsuite/tests/driver/boot-target/A.hs-boot - + testsuite/tests/driver/boot-target/B.hs - + testsuite/tests/driver/boot-target/Makefile - + testsuite/tests/driver/boot-target/all.T Changes: ===================================== compiler/GHC/Driver/Backpack.hs ===================================== @@ -781,7 +781,7 @@ summariseRequirement pn mod_name = do let loc = srcLocSpan (mkSrcLoc (mkFastString (bkp_filename env)) 1 1) let fc = hsc_FC hsc_env - mod <- liftIO $ addHomeModuleToFinder fc home_unit mod_name location + mod <- liftIO $ addHomeModuleToFinder fc home_unit (notBoot mod_name) location extra_sig_imports <- liftIO $ findExtraSigImports hsc_env HsigFile mod_name @@ -893,7 +893,7 @@ hsModuleToModSummary home_keys pn hsc_src modname this_mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit modname location + addHomeModuleToFinder fc home_unit (GWIB modname (hscSourceToIsBoot hsc_src)) location let ms = ModSummary { ms_mod = this_mod, ms_hsc_src = hsc_src, ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2044,25 +2044,43 @@ summariseFile hsc_env' home_unit old_summaries src_fn mb_phase maybe_buf <- getPreprocessedImports hsc_env src_fn mb_phase maybe_buf let fopts = initFinderOpts (hsc_dflags hsc_env) - - -- Make a ModLocation for this file - let location = mkHomeModLocation fopts pi_mod_name (unsafeEncodeUtf src_fn) + src_path = unsafeEncodeUtf src_fn + + is_boot = case takeExtension src_fn of + ".hs-boot" -> IsBoot + ".lhs-boot" -> IsBoot + _ -> NotBoot + + (path_without_boot, hsc_src) + | isHaskellSigFilename src_fn = (src_path, HsigFile) + | IsBoot <- is_boot = (removeBootSuffix src_path, HsBootFile) + | otherwise = (src_path, HsSrcFile) + + -- Make a ModLocation for the Finder, who only has one entry for + -- each @ModuleName@, and therefore needs to use the locations for + -- the non-boot files. + location_without_boot = + mkHomeModLocation fopts pi_mod_name path_without_boot + + -- Make a ModLocation for this file, adding the @-boot@ suffix to + -- all paths if the original was a boot file. + location + | IsBoot <- is_boot + = addBootSuffixLocn location_without_boot + | otherwise + = location_without_boot -- Tell the Finder cache where it is, so that subsequent calls -- to findModule will find it, even if it's not on any search path mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit pi_mod_name location + addHomeModuleToFinder fc home_unit (GWIB pi_mod_name is_boot) location liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = NotBoot - , nms_hsc_src = - if isHaskellSigFilename src_fn - then HsigFile - else HsSrcFile + , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod , nms_preimps = preimps @@ -2090,9 +2108,10 @@ checkSummaryHash -- Also, only add to finder cache for non-boot modules as the finder cache -- makes sure to add a boot suffix for boot files. _ <- do - let fc = hsc_FC hsc_env + let fc = hsc_FC hsc_env + gwib = GWIB (ms_mod old_summary) (isBootSummary old_summary) case ms_hsc_src old_summary of - HsSrcFile -> addModuleToFinder fc (ms_mod old_summary) location + HsSrcFile -> addModuleToFinder fc gwib location _ -> return () hi_timestamp <- modificationTimeIfExists (ml_hi_file location) @@ -2230,7 +2249,6 @@ summariseModule hsc_env' home_unit old_summary_map is_boot (L _ wanted_mod) mb_p liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = is_boot , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod @@ -2243,7 +2261,6 @@ data MakeNewModSummary = MakeNewModSummary { nms_src_fn :: FilePath , nms_src_hash :: Fingerprint - , nms_is_boot :: IsBootInterface , nms_hsc_src :: HscSource , nms_location :: ModLocation , nms_mod :: Module ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -734,7 +734,7 @@ runHscPhase pipe_env hsc_env0 input_fn src_flavour = do mod <- do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit mod_name location + addHomeModuleToFinder fc home_unit (GWIB mod_name (hscSourceToIsBoot src_flavour)) location -- Make the ModSummary to hand to hscMain let ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -89,23 +89,23 @@ type BaseName = OsPath -- Basename of file initFinderCache :: IO FinderCache initFinderCache = do - mod_cache <- newIORef emptyInstalledModuleEnv + mod_cache <- newIORef emptyInstalledModuleWithIsBootEnv file_cache <- newIORef M.empty let flushFinderCaches :: UnitEnv -> IO () flushFinderCaches ue = do - atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleEnv is_ext fm, ()) + atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleWithIsBootEnv is_ext fm, ()) atomicModifyIORef' file_cache $ \_ -> (M.empty, ()) where - is_ext mod _ = not (isUnitEnvInstalledModule ue mod) + is_ext mod _ = not (isUnitEnvInstalledModule ue (gwib_mod mod)) - addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () addToFinderCache key val = - atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleEnv c key val, ()) + atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleWithIsBootEnv c key val, ()) - lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) lookupFinderCache key = do c <- readIORef mod_cache - return $! lookupInstalledModuleEnv c key + return $! lookupInstalledModuleWithIsBootEnv c key lookupFileCache :: FilePath -> IO Fingerprint lookupFileCache key = do @@ -255,7 +255,7 @@ orIfNotFound this or_this = do homeSearchCache :: FinderCache -> UnitId -> ModuleName -> IO InstalledFindResult -> IO InstalledFindResult homeSearchCache fc home_unit mod_name do_this = do let mod = mkModule home_unit mod_name - modLocationCache fc mod do_this + modLocationCache fc (notBoot mod) do_this findExposedPackageModule :: FinderCache -> FinderOpts -> UnitState -> ModuleName -> PkgQual -> IO FindResult findExposedPackageModule fc fopts units mod_name mb_pkg = @@ -312,7 +312,7 @@ findLookupResult fc fopts r = case r of , fr_unusables = [] , fr_suggestions = suggest' }) -modLocationCache :: FinderCache -> InstalledModule -> IO InstalledFindResult -> IO InstalledFindResult +modLocationCache :: FinderCache -> InstalledModuleWithIsBoot -> IO InstalledFindResult -> IO InstalledFindResult modLocationCache fc mod do_this = do m <- lookupFinderCache fc mod case m of @@ -322,17 +322,17 @@ modLocationCache fc mod do_this = do addToFinderCache fc mod result return result -addModuleToFinder :: FinderCache -> Module -> ModLocation -> IO () +addModuleToFinder :: FinderCache -> ModuleWithIsBoot -> ModLocation -> IO () addModuleToFinder fc mod loc = do - let imod = toUnitId <$> mod - addToFinderCache fc imod (InstalledFound loc imod) + let imod = fmap toUnitId <$> mod + addToFinderCache fc imod (InstalledFound loc (gwib_mod imod)) -- This returns a module because it's more convenient for users -addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleName -> ModLocation -> IO Module +addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleNameWithIsBoot -> ModLocation -> IO Module addHomeModuleToFinder fc home_unit mod_name loc = do - let mod = mkHomeInstalledModule home_unit mod_name - addToFinderCache fc mod (InstalledFound loc mod) - return (mkHomeModule home_unit mod_name) + let mod = mkHomeInstalledModule home_unit <$> mod_name + addToFinderCache fc mod (InstalledFound loc (gwib_mod mod)) + return (mkHomeModule home_unit (gwib_mod mod_name)) -- ----------------------------------------------------------------------------- -- The internal workers @@ -466,7 +466,7 @@ findPackageModule_ :: FinderCache -> FinderOpts -> InstalledModule -> UnitInfo - findPackageModule_ fc fopts mod pkg_conf = do massertPpr (moduleUnit mod == unitId pkg_conf) (ppr (moduleUnit mod) <+> ppr (unitId pkg_conf)) - modLocationCache fc mod $ + modLocationCache fc (notBoot mod) $ -- special case for GHC.Prim; we won't find it in the filesystem. if mod `installedModuleEq` gHC_PRIM ===================================== compiler/GHC/Unit/Finder/Types.hs ===================================== @@ -30,9 +30,9 @@ data FinderCache = FinderCache { flushFinderCaches :: UnitEnv -> IO () -- ^ remove all the home modules from the cache; package modules are -- assumed to not move around during a session; also flush the file hash -- cache. - , addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + , addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () -- ^ Add a found location to the cache for the module. - , lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + , lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) -- ^ Look for a location in the cache. , lookupFileCache :: FilePath -> IO Fingerprint -- ^ Look for the hash of a file in the cache. This should add it to the ===================================== compiler/GHC/Unit/Module/Env.hs ===================================== @@ -33,6 +33,17 @@ module GHC.Unit.Module.Env , mergeInstalledModuleEnv , plusInstalledModuleEnv , installedModuleEnvElts + + -- * InstalledModuleWithIsBootEnv + , InstalledModuleWithIsBootEnv + , emptyInstalledModuleWithIsBootEnv + , lookupInstalledModuleWithIsBootEnv + , extendInstalledModuleWithIsBootEnv + , filterInstalledModuleWithIsBootEnv + , delInstalledModuleWithIsBootEnv + , mergeInstalledModuleWithIsBootEnv + , plusInstalledModuleWithIsBootEnv + , installedModuleWithIsBootEnvElts ) where @@ -283,3 +294,56 @@ plusInstalledModuleEnv :: (elt -> elt -> elt) plusInstalledModuleEnv f (InstalledModuleEnv xm) (InstalledModuleEnv ym) = InstalledModuleEnv $ Map.unionWith f xm ym + + +-------------------------------------------------------------------- +-- InstalledModuleWithIsBootEnv +-------------------------------------------------------------------- + +-- | A map keyed off of 'InstalledModuleWithIsBoot' +newtype InstalledModuleWithIsBootEnv elt = InstalledModuleWithIsBootEnv (Map InstalledModuleWithIsBoot elt) + +instance Outputable elt => Outputable (InstalledModuleWithIsBootEnv elt) where + ppr (InstalledModuleWithIsBootEnv env) = ppr env + + +emptyInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a +emptyInstalledModuleWithIsBootEnv = InstalledModuleWithIsBootEnv Map.empty + +lookupInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> Maybe a +lookupInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = Map.lookup m e + +extendInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> a -> InstalledModuleWithIsBootEnv a +extendInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m x = InstalledModuleWithIsBootEnv (Map.insert m x e) + +filterInstalledModuleWithIsBootEnv :: (InstalledModuleWithIsBoot -> a -> Bool) -> InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBootEnv a +filterInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv e) = + InstalledModuleWithIsBootEnv (Map.filterWithKey f e) + +delInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> InstalledModuleWithIsBootEnv a +delInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = InstalledModuleWithIsBootEnv (Map.delete m e) + +installedModuleWithIsBootEnvElts :: InstalledModuleWithIsBootEnv a -> [(InstalledModuleWithIsBoot, a)] +installedModuleWithIsBootEnvElts (InstalledModuleWithIsBootEnv e) = Map.assocs e + +mergeInstalledModuleWithIsBootEnv + :: (elta -> eltb -> Maybe eltc) + -> (InstalledModuleWithIsBootEnv elta -> InstalledModuleWithIsBootEnv eltc) -- map X + -> (InstalledModuleWithIsBootEnv eltb -> InstalledModuleWithIsBootEnv eltc) -- map Y + -> InstalledModuleWithIsBootEnv elta + -> InstalledModuleWithIsBootEnv eltb + -> InstalledModuleWithIsBootEnv eltc +mergeInstalledModuleWithIsBootEnv f g h (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) + = InstalledModuleWithIsBootEnv $ Map.mergeWithKey + (\_ x y -> (x `f` y)) + (coerce g) + (coerce h) + xm ym + +plusInstalledModuleWithIsBootEnv :: (elt -> elt -> elt) + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt +plusInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) = + InstalledModuleWithIsBootEnv $ Map.unionWith f xm ym + ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -86,6 +86,8 @@ module GHC.Unit.Types , GenWithIsBoot (..) , ModuleNameWithIsBoot , ModuleWithIsBoot + , InstalledModuleWithIsBoot + , notBoot ) where @@ -723,6 +725,8 @@ type ModuleNameWithIsBoot = GenWithIsBoot ModuleName type ModuleWithIsBoot = GenWithIsBoot Module +type InstalledModuleWithIsBoot = GenWithIsBoot InstalledModule + instance Binary a => Binary (GenWithIsBoot a) where put_ bh (GWIB { gwib_mod, gwib_isBoot }) = do put_ bh gwib_mod @@ -736,3 +740,6 @@ instance Outputable a => Outputable (GenWithIsBoot a) where ppr (GWIB { gwib_mod, gwib_isBoot }) = hsep $ ppr gwib_mod : case gwib_isBoot of IsBoot -> [ text "{-# SOURCE #-}" ] NotBoot -> [] + +notBoot :: mod -> GenWithIsBoot mod +notBoot gwib_mod = GWIB {gwib_mod, gwib_isBoot = NotBoot} ===================================== testsuite/tests/driver/boot-target/A.hs ===================================== @@ -0,0 +1,5 @@ +module A where + +import B + +data A = A B ===================================== testsuite/tests/driver/boot-target/A.hs-boot ===================================== @@ -0,0 +1,3 @@ +module A where + +data A ===================================== testsuite/tests/driver/boot-target/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} A + +data B = B A ===================================== testsuite/tests/driver/boot-target/Makefile ===================================== @@ -0,0 +1,5 @@ +boot1: + $(TEST_HC) -c A.hs-boot B.hs + +boot2: + $(TEST_HC) A.hs-boot A.hs B.hs -v0 ===================================== testsuite/tests/driver/boot-target/all.T ===================================== @@ -0,0 +1,9 @@ +def test_boot(name): + return test(name, + [extra_files(['A.hs', 'A.hs-boot', 'B.hs']), + ], + makefile_test, + []) + +test_boot('boot1') +test_boot('boot2') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c3c3970b12429d1883dc9ddd9ddd53373fd8fd46 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c3c3970b12429d1883dc9ddd9ddd53373fd8fd46 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 13:09:34 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 10 Sep 2024 09:09:34 -0400 Subject: [Git][ghc/ghc][wip/romes/25110] 28 commits: Haddock: Add no-compilation flag Message-ID: <66e0450e95891_3d42f168f7e0582f4@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/25110 at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - af216f6d by Rodrigo Mesquita at 2024-09-10T14:09:25+01:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/Core/InstEnv.hs - compiler/GHC/Core/Opt/CallerCC/Types.hs - compiler/GHC/Core/Opt/WorkWrap/Utils.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Driver/Config/Stg/Pipeline.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - + compiler/GHC/Stg/Make.hs - compiler/GHC/Stg/Pipeline.hs - compiler/GHC/Stg/Stats.hs - compiler/GHC/Stg/Syntax.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/99984e36f346ff59968d9e33b64c0add2a412356...af216f6d369d7c7c152a8218ff4b9ed06c4a4491 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/99984e36f346ff59968d9e33b64c0add2a412356...af216f6d369d7c7c152a8218ff4b9ed06c4a4491 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 14:47:05 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 10:47:05 -0400 Subject: [Git][ghc/ghc][master] RTS linker: add support for hidden symbols (#25191) Message-ID: <66e05be96720a_f8848598f44791b8@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 11 changed files: - rts/Linker.c - rts/LinkerInternals.h - rts/linker/Elf.c - rts/linker/ElfTypes.h - rts/linker/PEi386.c - testsuite/tests/rts/linker/Makefile - + testsuite/tests/rts/linker/T25191.hs - + testsuite/tests/rts/linker/T25191.stdout - + testsuite/tests/rts/linker/T25191_foo1.c - + testsuite/tests/rts/linker/T25191_foo2.c - testsuite/tests/rts/linker/all.T Changes: ===================================== rts/Linker.c ===================================== @@ -232,11 +232,11 @@ static void ghciRemoveSymbolTable(StrHashTable *table, const SymbolName* key, static const char * symbolTypeString (SymType type) { - switch (type & ~SYM_TYPE_DUP_DISCARD) { + switch (type & ~(SYM_TYPE_DUP_DISCARD | SYM_TYPE_HIDDEN)) { case SYM_TYPE_CODE: return "code"; case SYM_TYPE_DATA: return "data"; case SYM_TYPE_INDIRECT_DATA: return "indirect-data"; - default: barf("symbolTypeString: unknown symbol type"); + default: barf("symbolTypeString: unknown symbol type (%d)", type); } } @@ -283,10 +283,19 @@ int ghciInsertSymbolTable( } else if (pinfo->type ^ type) { + if(pinfo->type & SYM_TYPE_HIDDEN) + { + /* The existing symbol is hidden, let's replace it */ + pinfo->value = data; + pinfo->owner = owner; + pinfo->strength = strength; + pinfo->type = type; + return 1; + } /* We were asked to discard the symbol on duplicates, do so quietly. */ - if (!(type & SYM_TYPE_DUP_DISCARD)) + if (!(type & (SYM_TYPE_DUP_DISCARD | SYM_TYPE_HIDDEN))) { - debugBelch("Symbol type mismatch.\n"); + debugBelch("Symbol type mismatch (existing %d, new %d).\n", pinfo->type, type); debugBelch("Symbol %s was defined by %" PATH_FMT " to be a %s symbol.\n", key, obj_name, symbolTypeString(type)); debugBelch(" yet was defined by %" PATH_FMT " to be a %s symbol.\n", ===================================== rts/LinkerInternals.h ===================================== @@ -64,6 +64,8 @@ typedef enum _SymType { SYM_TYPE_DUP_DISCARD = 1 << 3, /* the symbol is a symbol in a BFD import library however if a duplicate is found with a mismatching SymType then discard this one. */ + SYM_TYPE_HIDDEN = 1 << 4, /* the symbol is hidden and should not be exported */ + } SymType; ===================================== rts/linker/Elf.c ===================================== @@ -1073,6 +1073,9 @@ ocGetNames_ELF ( ObjectCode* oc ) } else { sym_type = SYM_TYPE_DATA; } + if(ELF_ST_VISIBILITY(symbol->elf_sym->st_other) == STV_HIDDEN) { + sym_type |= SYM_TYPE_HIDDEN; + } /* And the decision is ... */ ===================================== rts/linker/ElfTypes.h ===================================== @@ -33,6 +33,9 @@ #define Elf_Sym Elf64_Sym #define Elf_Rel Elf64_Rel #define Elf_Rela Elf64_Rela +#if !defined(ELF_ST_VISIBILITY) +#define ELF_ST_VISIBILITY ELF64_ST_VISIBILITY +#endif #if !defined(ELF_ST_TYPE) #define ELF_ST_TYPE ELF64_ST_TYPE #endif @@ -57,6 +60,9 @@ #define Elf_Sym Elf32_Sym #define Elf_Rel Elf32_Rel #define Elf_Rela Elf32_Rela +#if !defined(ELF_ST_VISIBILITY) +#define ELF_ST_VISIBILITY ELF32_ST_VISIBILITY +#endif /* ELF_ST_VISIBILITY */ #if !defined(ELF_ST_TYPE) #define ELF_ST_TYPE ELF32_ST_TYPE #endif /* ELF_ST_TYPE */ ===================================== rts/linker/PEi386.c ===================================== @@ -1891,6 +1891,9 @@ ocGetNames_PEi386 ( ObjectCode* oc ) sname[size-start]='\0'; stgFree(tmp); sname = strdup (sname); + if(secNumber == IMAGE_SYM_UNDEFINED) + type |= SYM_TYPE_HIDDEN; + if (!ghciInsertSymbolTable(oc->fileName, symhash, sname, addr, false, type, oc)) return false; @@ -1905,6 +1908,8 @@ ocGetNames_PEi386 ( ObjectCode* oc ) && (!section || (section && section->kind != SECTIONKIND_IMPORT))) { /* debugBelch("addSymbol %p `%s' Weak:%lld \n", addr, sname, isWeak); */ sname = strdup (sname); + if(secNumber == IMAGE_SYM_UNDEFINED) + type |= SYM_TYPE_HIDDEN; IF_DEBUG(linker_verbose, debugBelch("addSymbol %p `%s'\n", addr, sname)); ASSERT(i < (uint32_t)oc->n_symbols); oc->symbols[i].name = sname; @@ -1936,7 +1941,7 @@ static size_t makeSymbolExtra_PEi386( ObjectCode* oc, uint64_t index STG_UNUSED, size_t s, char* symbol STG_UNUSED, SymType type ) { SymbolExtra *extra; - switch(type & ~SYM_TYPE_DUP_DISCARD) { + switch(type & ~(SYM_TYPE_DUP_DISCARD | SYM_TYPE_HIDDEN)) { case SYM_TYPE_CODE: { // jmp *-14(%rip) extra = m32_alloc(oc->rx_m32, sizeof(SymbolExtra), 8); ===================================== testsuite/tests/rts/linker/Makefile ===================================== @@ -145,3 +145,10 @@ reloc-none: "$(TEST_HC)" load-object.c -o load-object -no-hs-main -debug "$(TEST_HC)" -c reloc-none.c -o reloc-none.o ./load-object reloc-none.o + +.PHONY: T25191 +T25191: + "$(TEST_HC)" -c T25191_foo1.c -o foo1.o -v0 + "$(TEST_HC)" -c T25191_foo2.c -o foo2.o -v0 + "$(TEST_HC)" T25191.hs -v0 + ./T25191 ===================================== testsuite/tests/rts/linker/T25191.hs ===================================== @@ -0,0 +1,29 @@ +{-# LANGUAGE ForeignFunctionInterface, CPP #-} +import Foreign.C.String +import Control.Monad +import System.FilePath +import Foreign.Ptr + +-- Type of paths is different on Windows +#if defined(mingw32_HOST_OS) +type PathString = CWString +withPathString = withCWString +#else +type PathString = CString +withPathString = withCString +#endif + +main = do + initLinker + r1 <- withPathString "foo1.o" loadObj + when (r1 /= 1) $ error "loadObj failed" + r2 <- withPathString "foo2.o" loadObj + when (r2 /= 1) $ error "loadObj failed" + r <- resolveObjs + when (r /= 1) $ error "resolveObj failed" + putStrLn "success" + +foreign import ccall "initLinker" initLinker :: IO () +foreign import ccall "addDLL" addDLL :: PathString -> IO CString +foreign import ccall "loadObj" loadObj :: PathString -> IO Int +foreign import ccall "resolveObjs" resolveObjs :: IO Int ===================================== testsuite/tests/rts/linker/T25191.stdout ===================================== @@ -0,0 +1 @@ +success ===================================== testsuite/tests/rts/linker/T25191_foo1.c ===================================== @@ -0,0 +1,10 @@ +#include + +void __attribute__ ((__visibility__ ("hidden"))) foo(void) { + printf("HIDDEN FOO\n"); +} + +void bar(void) { + printf("BAR\n"); + foo(); +} ===================================== testsuite/tests/rts/linker/T25191_foo2.c ===================================== @@ -0,0 +1,8 @@ +#include + +extern void bar(void); + +void foo(void) { + printf("VISIBLE FOO\n"); + bar(); +} ===================================== testsuite/tests/rts/linker/all.T ===================================== @@ -168,3 +168,11 @@ test('reloc-none', unless(opsys('linux'), skip), req_rts_linker], makefile_test, ['reloc-none']) + +test('T25191', + [req_rts_linker, + extra_files(['T25191_foo1.c','T25191_foo2.c']), + when(opsys('darwin'), expect_broken(25191)), # not supported in the MachO linker yet + when(opsys('mingw32'), expect_broken(25191)) # not supported in the PE linker yet + ], + makefile_test, ['T25191']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9ca155065b39968d784846902ec1e0bcbe60ee40 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9ca155065b39968d784846902ec1e0bcbe60ee40 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 14:47:35 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 10:47:35 -0400 Subject: [Git][ghc/ghc][master] Fix C warnings (#25237) Message-ID: <66e05c0715008_f884855924082477@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 2 changed files: - libraries/ghc-heap/tests/stack_misc_closures_c.c - testsuite/tests/rts/pause-resume/list_threads_and_misc_roots_c.c Changes: ===================================== libraries/ghc-heap/tests/stack_misc_closures_c.c ===================================== @@ -193,7 +193,7 @@ void create_any_bco_frame(Capability *cap, StgStack *stack, StgWord w) { StgWord bcoSizeWords = sizeofW(StgBCO) + sizeofW(StgLargeBitmap) + sizeofW(StgWord); StgBCO *bco = (StgBCO *)allocate(cap, bcoSizeWords); - SET_HDR(bco, &stg_BCO_info, CCS_MAIN); + SET_HDR(bco, (StgInfoTable*) &stg_BCO_info, CCS_MAIN); c->payload[0] = (StgClosure *)bco; bco->size = bcoSizeWords; ===================================== testsuite/tests/rts/pause-resume/list_threads_and_misc_roots_c.c ===================================== @@ -27,10 +27,10 @@ void checkGcRoots(void) rts_listThreads(&collectTSOsCallback, NULL); for (int i = 0; i < tsoCount; i++) { - StgTSO *tso = UNTAG_CLOSURE(tsos[i]); + StgClosure *tso = UNTAG_CLOSURE((StgClosure*) tsos[i]); if (get_itbl(tso)->type != TSO) { - fprintf(stderr, "tso returned a non-TSO type %zu at index %i\n", + fprintf(stderr, "tso returned a non-TSO type %u at index %i\n", tso->header.info->type, i); exit(1); @@ -44,7 +44,7 @@ void checkGcRoots(void) StgClosure *root = UNTAG_CLOSURE(miscRoots[i]); if (get_itbl(root)->type == TSO) { - fprintf(stderr, "rts_listThreads unexpectedly returned an TSO type at index %i (TSO=%zu)\n", i, TSO); + fprintf(stderr, "rts_listThreads unexpectedly returned an TSO type at index %i (TSO=%d)\n", i, TSO); exit(1); } } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3b2dc826c9d4e65546ae3083303fdd491e74bd5a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3b2dc826c9d4e65546ae3083303fdd491e74bd5a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 14:48:17 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 10:48:17 -0400 Subject: [Git][ghc/ghc][master] JS: fix codegen of static string data Message-ID: <66e05c313886_f884857533c8537f@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 2 changed files: - compiler/GHC/StgToJS/Symbols.hs - compiler/GHC/StgToJS/Types.hs Changes: ===================================== compiler/GHC/StgToJS/Symbols.hs ===================================== @@ -1212,3 +1212,6 @@ hdStlStr = fsLit "h$stl" hdStiStr :: FastString hdStiStr = fsLit "h$sti" + +hdStrStr :: FastString +hdStrStr = fsLit "h$str" ===================================== compiler/GHC/StgToJS/Types.hs ===================================== @@ -284,7 +284,7 @@ instance ToJExpr StaticLit where toJExpr (IntLit i) = toJExpr i toJExpr NullLit = null_ toJExpr (DoubleLit d) = toJExpr (unSaneDouble d) - toJExpr (StringLit t) = app hdStr [toJExpr t] + toJExpr (StringLit t) = app hdStrStr [toJExpr t] toJExpr (BinLit b) = app hdRawStr [toJExpr (map toInteger (BS.unpack b))] toJExpr (LabelLit _isFun lbl) = global lbl View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/057159947ae88f6bad703cc357913d8ca60384d0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/057159947ae88f6bad703cc357913d8ca60384d0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 14:49:00 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 10 Sep 2024 10:49:00 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 3 commits: UniqDSMT... Message-ID: <66e05c5c79040_f8848565ef085582@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 44c907b8 by Rodrigo Mesquita at 2024-09-09T11:41:38+01:00 UniqDSMT... - - - - - 2ed99824 by Rodrigo Mesquita at 2024-09-09T11:41:53+01:00 liftEff hoistEff - - - - - 90b8bdb0 by Rodrigo Mesquita at 2024-09-10T15:48:42+01:00 CgStream Revert "Very bad..." This reverts commit 1d6c826927df8abfd18649baa5d772bbbd3721df. Note tweaks Reapply "Very bad..." This reverts commit 09f3a4ec6c78024e235c8f9c2488398b719a519f. GODO - - - - - 15 changed files: - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Wasm.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/Data/Stream.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/GenerateCgIPEStub.hs - compiler/GHC/Driver/Hooks.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/StgToCmm.hs - compiler/GHC/StgToCmm/CgUtils.hs - compiler/GHC/StgToCmm/Types.hs - compiler/GHC/Types/Unique/DSM.hs - compiler/GHC/Types/Unique/Supply.hs Changes: ===================================== compiler/GHC/Cmm/Info.hs ===================================== @@ -36,9 +36,10 @@ import GHC.Prelude import GHC.Cmm import GHC.Cmm.Utils import GHC.Cmm.CLabel +import GHC.StgToCmm.CgUtils (CgStream) import GHC.Runtime.Heap.Layout import GHC.Data.Bitmap -import GHC.Data.Stream (Stream) +import GHC.Data.Stream (liftEff) import qualified GHC.Data.Stream as Stream import GHC.Cmm.Dataflow.Label @@ -54,7 +55,7 @@ import GHC.Utils.Outputable import GHC.Types.Unique.DSM import Data.ByteString (ByteString) -import Data.IORef +import qualified Control.Monad.Trans.State.Strict as T -- When we split at proc points, we need an empty info table. mkEmptyContInfoTable :: CLabel -> CmmInfoTable @@ -65,26 +66,23 @@ mkEmptyContInfoTable info_lbl , cit_srt = Nothing , cit_clo = Nothing } -cmmToRawCmm :: Logger -> Profile -> Stream IO CmmGroupSRTs a - -> IO (Stream IO RawCmmGroup a) +cmmToRawCmm :: Logger -> Profile -> CgStream CmmGroupSRTs a + -> IO (CgStream RawCmmGroup a) cmmToRawCmm logger profile cmms - = do { detUqSupply <- newIORef (initDUniqSupply 'i' 1) - ; let do_one :: [CmmDeclSRTs] -> IO [RawCmmDecl] + = do { let do_one :: [CmmDeclSRTs] -> UniqDSMT IO [RawCmmDecl] do_one cmm = do -- NB. strictness fixes a space leak. DO NOT REMOVE. withTimingSilent logger (text "Cmm -> Raw Cmm") (\x -> seqList x ()) $ do - -- We have to store the deterministic unique supply - -- to produce uniques across cmm decls. - nextUq <- readIORef detUqSupply - -- By using a local namespace 'i' here, we can have other - -- deterministic supplies starting from the same unique in - -- other parts of the Cmm backend -- See Note [Deterministic Uniques in the NCG] - let (a, us) = runUniqueDSM nextUq $ - concatMapM (mkInfoTable profile) cmm - writeIORef detUqSupply us + us <- UDSMT T.get + let (a, us') = runUniqueDSM us $ + concatMapM (mkInfoTable profile) cmm + UDSMT (T.put us') return a - ; return (Stream.mapM do_one cmms) + ; return $ do + -- Override the unique supply tag to 'i' + _ <- liftEff $ UDSMT (T.modify @IO (newTagDUniqSupply 'i')) + Stream.mapM do_one cmms } ===================================== compiler/GHC/Cmm/Pipeline.hs ===================================== @@ -44,12 +44,12 @@ cmmPipeline :: Logger -> CmmConfig -> ModuleSRTInfo -- Info about SRTs generated so far - -> DUniqSupply -> CmmGroup -- Input C-- with Procedures - -> IO (ModuleSRTInfo, DUniqSupply, CmmGroupSRTs) -- Output CPS transformed C-- + -> DUniqSupply + -> IO ((ModuleSRTInfo, CmmGroupSRTs), DUniqSupply) -- Output CPS transformed C-- -cmmPipeline logger cmm_config srtInfo dus0 prog = do - let forceRes (info, us, group) = info `seq` us `seq` foldr seq () group +cmmPipeline logger cmm_config srtInfo prog dus0 = do + let forceRes ((info, group), us) = info `seq` us `seq` foldr seq () group let platform = cmmPlatform cmm_config withTimingSilent logger (text "Cmm pipeline") forceRes $ do (dus1, prog') <- {-# SCC "tops" #-} mapAccumLM (cpsTop logger platform cmm_config) dus0 prog @@ -57,7 +57,7 @@ cmmPipeline logger cmm_config srtInfo dus0 prog = do (srtInfo, dus, cmms) <- {-# SCC "doSRTs" #-} doSRTs cmm_config srtInfo dus1 procs data_ dumpWith logger Opt_D_dump_cmm_cps "Post CPS Cmm" FormatCMM (pdoc platform cmms) - return (srtInfo, dus, cmms) + return ((srtInfo, cmms), dus) -- | The Cmm pipeline for a single 'CmmDecl'. Returns: -- ===================================== compiler/GHC/CmmToAsm.hs ===================================== @@ -116,7 +116,8 @@ import GHC.Utils.Constants (debugIsOn) import GHC.Data.FastString import GHC.Types.Unique.Set import GHC.Unit -import GHC.Data.Stream (Stream) +import GHC.StgToCmm.CgUtils (CgStream) +import GHC.Data.Stream (liftIO) import qualified GHC.Data.Stream as Stream import GHC.Settings @@ -129,14 +130,14 @@ import System.IO import System.Directory ( getCurrentDirectory ) -------------------- -nativeCodeGen :: forall a . Logger -> ToolSettings -> NCGConfig -> ModLocation -> Handle -> DUniqSupply - -> Stream IO RawCmmGroup a - -> IO a -nativeCodeGen logger ts config modLoc h us cmms +nativeCodeGen :: forall a . Logger -> ToolSettings -> NCGConfig -> ModLocation -> Handle + -> CgStream RawCmmGroup a + -> UniqDSMT IO a +nativeCodeGen logger ts config modLoc h cmms = let platform = ncgPlatform config nCG' :: ( OutputableP Platform statics, Outputable jumpDest, Instruction instr) - => NcgImpl statics instr jumpDest -> IO a - nCG' ncgImpl = nativeCodeGen' logger config modLoc ncgImpl h us cmms + => NcgImpl statics instr jumpDest -> UniqDSMT IO a + nCG' ncgImpl = nativeCodeGen' logger config modLoc ncgImpl h cmms in case platformArch platform of ArchX86 -> nCG' (X86.ncgX86 config) ArchX86_64 -> nCG' (X86.ncgX86_64 config) @@ -152,7 +153,7 @@ nativeCodeGen logger ts config modLoc h us cmms ArchLoongArch64->panic "nativeCodeGen: No NCG for LoongArch64" ArchUnknown -> panic "nativeCodeGen: No NCG for unknown arch" ArchJavaScript-> panic "nativeCodeGen: No NCG for JavaScript" - ArchWasm32 -> Wasm32.ncgWasm config logger platform ts us modLoc h cmms + ArchWasm32 -> Wasm32.ncgWasm config logger platform ts modLoc h cmms -- | Data accumulated during code generation. Mostly about statistics, -- but also collects debug data for DWARF generation. @@ -203,19 +204,17 @@ nativeCodeGen' :: (OutputableP Platform statics, Outputable jumpDest, Instructio -> ModLocation -> NcgImpl statics instr jumpDest -> Handle - -> DUniqSupply - -> Stream IO RawCmmGroup a - -> IO a -nativeCodeGen' logger config modLoc ncgImpl h us cmms + -> CgStream RawCmmGroup a + -> UniqDSMT IO a +nativeCodeGen' logger config modLoc ncgImpl h cmms = do -- BufHandle is a performance hack. We could hide it inside -- Pretty if it weren't for the fact that we do lots of little -- printDocs here (in order to do codegen in constant space). - bufh <- newBufHandle h + bufh <- liftIO $ newBufHandle h let ngs0 = NGS [] [] [] [] [] [] emptyUFM mapEmpty - (ngs, us', a) <- cmmNativeGenStream logger config modLoc ncgImpl bufh us - cmms ngs0 - _ <- finishNativeGen logger config modLoc bufh us' ngs + (ngs, a) <- cmmNativeGenStream logger config modLoc ncgImpl bufh cmms ngs0 + _ <- finishNativeGen logger config modLoc bufh ngs return a finishNativeGen :: Instruction instr @@ -223,20 +222,20 @@ finishNativeGen :: Instruction instr -> NCGConfig -> ModLocation -> BufHandle - -> DUniqSupply -> NativeGenAcc statics instr - -> IO DUniqSupply -finishNativeGen logger config modLoc bufh us ngs + -> UniqDSMT IO () +finishNativeGen logger config modLoc bufh ngs = withTimingSilent logger (text "NCG") (`seq` ()) $ do - -- Write debug data and finish - us' <- if not (ncgDwarfEnabled config) - then return us - else do - compPath <- getCurrentDirectory - let (dwarf_h, us') = dwarfGen compPath config modLoc us (ngs_debug ngs) - (dwarf_s, _) = dwarfGen compPath config modLoc us (ngs_debug ngs) - emitNativeCode logger config bufh dwarf_h dwarf_s - return us' + -- Write debug data and finish + if not (ncgDwarfEnabled config) + then return () + else withDUS $ \us -> do + compPath <- getCurrentDirectory + let (dwarf_h, us') = dwarfGen compPath config modLoc us (ngs_debug ngs) + (dwarf_s, _) = dwarfGen compPath config modLoc us (ngs_debug ngs) + emitNativeCode logger config bufh dwarf_h dwarf_s + return ((), us') + liftIO $ do -- dump global NCG stats for graph coloring allocator let stats = concat (ngs_colorStats ngs) @@ -272,7 +271,7 @@ finishNativeGen logger config modLoc bufh us ngs bPutHDoc bufh ctx $ makeImportsDoc config (concat (ngs_imports ngs)) bFlush bufh - return us' + return () where dump_stats = logDumpFile logger (mkDumpStyle alwaysQualify) Opt_D_dump_asm_stats "NCG stats" @@ -284,20 +283,18 @@ cmmNativeGenStream :: forall statics jumpDest instr a . (OutputableP Platform st -> ModLocation -> NcgImpl statics instr jumpDest -> BufHandle - -> DUniqSupply - -> Stream.Stream IO RawCmmGroup a + -> CgStream RawCmmGroup a -> NativeGenAcc statics instr - -> IO (NativeGenAcc statics instr, DUniqSupply, a) + -> UniqDSMT IO (NativeGenAcc statics instr, a) -cmmNativeGenStream logger config modLoc ncgImpl h us cmm_stream ngs - = loop us (Stream.runStream cmm_stream) ngs +cmmNativeGenStream logger config modLoc ncgImpl h cmm_stream ngs + = loop (Stream.runStream cmm_stream) ngs where ncglabel = text "NCG" - loop :: DUniqSupply - -> Stream.StreamS IO RawCmmGroup a - -> NativeGenAcc statics instr - -> IO (NativeGenAcc statics instr, DUniqSupply, a) - loop us s ngs = + loop :: Stream.StreamS (UniqDSMT IO) RawCmmGroup a + -> NativeGenAcc statics instr + -> UniqDSMT IO (NativeGenAcc statics instr, a) + loop s ngs = case s of Stream.Done a -> return (ngs { ngs_imports = reverse $ ngs_imports ngs @@ -305,35 +302,33 @@ cmmNativeGenStream logger config modLoc ncgImpl h us cmm_stream ngs , ngs_colorStats = reverse $ ngs_colorStats ngs , ngs_linearStats = reverse $ ngs_linearStats ngs }, - us, a) - Stream.Effect m -> m >>= \cmm_stream' -> loop us cmm_stream' ngs + Stream.Effect m -> m >>= \cmm_stream' -> loop cmm_stream' ngs Stream.Yield cmms cmm_stream' -> do - (us', ngs'') <- - withTimingSilent logger - ncglabel (\(a, b) -> a `seq` b `seq` ()) $ do + ngs'' <- + withTimingSilent logger ncglabel (`seq` ()) $ do -- Generate debug information let !ndbgs | ncgDwarfEnabled config = cmmDebugGen modLoc cmms | otherwise = [] dbgMap = debugToMap ndbgs -- Generate native code - (ngs',us') <- cmmNativeGens logger config ncgImpl h - dbgMap us cmms ngs 0 + ngs' <- withDUS $ cmmNativeGens logger config ncgImpl h + dbgMap cmms ngs 0 -- Link native code information into debug blocks -- See Note [What is this unwinding business?] in "GHC.Cmm.DebugBlock". let !ldbgs = cmmDebugLink (ngs_labels ngs') (ngs_unwinds ngs') ndbgs platform = ncgPlatform config - unless (null ldbgs) $ + unless (null ldbgs) $ liftIO $ putDumpFileMaybe logger Opt_D_dump_debug "Debug Infos" FormatText (vcat $ map (pdoc platform) ldbgs) -- Accumulate debug information for emission in finishNativeGen. let ngs'' = ngs' { ngs_debug = ngs_debug ngs' ++ ldbgs, ngs_labels = [] } - return (us', ngs'') + return ngs'' - loop us' cmm_stream' ngs'' + loop cmm_stream' ngs'' -- | Do native code generation on all these cmms. @@ -345,22 +340,22 @@ cmmNativeGens :: forall statics instr jumpDest. -> NcgImpl statics instr jumpDest -> BufHandle -> LabelMap DebugBlock - -> DUniqSupply -> [RawCmmDecl] -> NativeGenAcc statics instr -> Int + -> DUniqSupply -> IO (NativeGenAcc statics instr, DUniqSupply) cmmNativeGens logger config ncgImpl h dbgMap = go where - go :: DUniqSupply -> [RawCmmDecl] - -> NativeGenAcc statics instr -> Int + go :: [RawCmmDecl] + -> NativeGenAcc statics instr -> Int -> DUniqSupply -> IO (NativeGenAcc statics instr, DUniqSupply) - go us [] ngs !_ = + go [] ngs !_ us = return (ngs, us) - go us (cmm : cmms) ngs count = do + go (cmm : cmms) ngs count us = do let fileIds = ngs_dwarfFiles ngs (us', fileIds', native, imports, colorStats, linearStats, unwinds, mcfg) <- {-# SCC "cmmNativeGen" #-} @@ -409,7 +404,7 @@ cmmNativeGens logger config ncgImpl h dbgMap = go , ngs_dwarfFiles = fileIds' , ngs_unwinds = ngs_unwinds ngs `mapUnion` unwinds } - go us' cmms ngs' (count + 1) + go cmms ngs' (count + 1) us' -- see Note [pprNatCmmDeclS and pprNatCmmDeclH] in GHC.CmmToAsm.Monad ===================================== compiler/GHC/CmmToAsm/Wasm.hs ===================================== @@ -16,7 +16,8 @@ import GHC.CmmToAsm.Config import GHC.CmmToAsm.Wasm.Asm import GHC.CmmToAsm.Wasm.FromCmm import GHC.CmmToAsm.Wasm.Types -import GHC.Data.Stream (Stream, StreamS (..), runStream) +import GHC.StgToCmm.CgUtils (CgStream) +import GHC.Data.Stream (Stream, StreamS (..), runStream, liftIO) import GHC.Driver.DynFlags import GHC.Platform import GHC.Prelude @@ -32,13 +33,12 @@ ncgWasm :: Logger -> Platform -> ToolSettings -> - DUniqSupply -> ModLocation -> Handle -> - Stream IO RawCmmGroup a -> - IO a -ncgWasm ncg_config logger platform ts us loc h cmms = do - (r, s) <- streamCmmGroups ncg_config platform us cmms + CgStream RawCmmGroup a -> + UniqDSMT IO a +ncgWasm ncg_config logger platform ts loc h cmms = do + (r, s) <- streamCmmGroups ncg_config platform cmms outputWasm $ "# " <> string7 (fromJust $ ml_hs_file loc) <> "\n\n" outputWasm $ execWasmAsmM do_tail_call $ asmTellEverything TagI32 s pure r @@ -46,7 +46,7 @@ ncgWasm ncg_config logger platform ts us loc h cmms = do -- See Note [WasmTailCall] do_tail_call = doTailCall ts - outputWasm builder = do + outputWasm builder = liftIO $ do putDumpFileMaybe logger Opt_D_dump_asm @@ -58,14 +58,16 @@ ncgWasm ncg_config logger platform ts us loc h cmms = do streamCmmGroups :: NCGConfig -> Platform -> - DUniqSupply -> - Stream IO RawCmmGroup a -> - IO (a, WasmCodeGenState 'I32) -streamCmmGroups ncg_config platform us cmms = - go (initialWasmCodeGenState platform us) $ runStream cmms + CgStream RawCmmGroup a -> + UniqDSMT IO (a, WasmCodeGenState 'I32) +streamCmmGroups ncg_config platform cmms = withDUS $ \us -> do + (r,s) <- go (initialWasmCodeGenState platform us) $ runStream cmms + return ((r,s), wasmDUniqSupply s) where go s (Done r) = pure (r, s) - go s (Effect m) = m >>= go s + go s (Effect m) = do + (a, us') <- runUDSMT (wasmDUniqSupply s) m + go s{wasmDUniqSupply = us'} a go s (Yield decls k) = go (wasmExecM (onCmmGroup $ map opt decls) s) k where -- Run the generic cmm optimizations like other NCGs, followed ===================================== compiler/GHC/CmmToLlvm.hs ===================================== @@ -23,10 +23,11 @@ import GHC.CmmToLlvm.Regs import GHC.CmmToLlvm.Mangler import GHC.CmmToLlvm.Version -import GHC.StgToCmm.CgUtils ( fixStgRegisters ) +import GHC.StgToCmm.CgUtils ( fixStgRegisters, CgStream ) import GHC.Cmm import GHC.Cmm.Dataflow.Label +import GHC.Types.Unique.DSM import GHC.Utils.BufHandle import GHC.Driver.DynFlags import GHC.Platform ( platformArch, Arch(..) ) @@ -46,9 +47,11 @@ import System.IO -- | Top-level of the LLVM Code generator -- llvmCodeGen :: Logger -> LlvmCgConfig -> Handle - -> Stream.Stream IO RawCmmGroup a - -> IO a -llvmCodeGen logger cfg h cmm_stream + -> DUniqSupply -- ^ The deterministic uniq supply to run the CgStream. + -- See Note [Deterministic Uniques in the CG] + -> CgStream RawCmmGroup a + -> IO a +llvmCodeGen logger cfg h dus cmm_stream = withTiming logger (text "LLVM CodeGen") (const ()) $ do bufh <- newBufHandle h @@ -83,21 +86,24 @@ llvmCodeGen logger cfg h cmm_stream -- run code generation a <- runLlvm logger cfg llvm_ver bufh $ - llvmCodeGen' cfg cmm_stream + llvmCodeGen' cfg dus cmm_stream bFlush bufh return a -llvmCodeGen' :: LlvmCgConfig -> Stream.Stream IO RawCmmGroup a -> LlvmM a -llvmCodeGen' cfg cmm_stream +llvmCodeGen' :: LlvmCgConfig + -> DUniqSupply -- ^ The deterministic uniq supply to run the CgStream. + -- See Note [Deterministic Uniques in the CG] + -> CgStream RawCmmGroup a -> LlvmM a +llvmCodeGen' cfg dus cmm_stream = do -- Preamble renderLlvm (llvmHeader cfg) (llvmHeader cfg) ghcInternalFunctions cmmMetaLlvmPrelude -- Procedures - a <- Stream.consume cmm_stream liftIO llvmGroupLlvmGens + (a, _) <- runUDSMT dus $ Stream.consume cmm_stream (hoistUDSMT liftIO) (liftUDSMT . llvmGroupLlvmGens) -- Declare aliases for forward references decls <- generateExternDecls ===================================== compiler/GHC/Data/Stream.hs ===================================== @@ -9,7 +9,7 @@ -- | Monadic streams module GHC.Data.Stream ( - Stream(..), StreamS(..), runStream, yield, liftIO, + Stream(..), StreamS(..), runStream, yield, liftIO, liftEff, hoistEff, collect, consume, fromList, map, mapM, mapAccumL_ ) where @@ -140,3 +140,22 @@ mapAccumL_ f c str = Stream $ \f h -> go c f h (runStream str) go c f1 h1 (Yield a p) = Effect (f c a >>= (\(c', b) -> f1 b >>= \r' -> return $ Yield r' (go c' f1 h1 p))) go c f1 h1 (Effect m) = Effect (go c f1 h1 <$> m) + +-- | Lift an effect into the Stream +liftEff :: Monad m => m b -> Stream m a b +liftEff eff = Stream $ \_f g -> Effect (g <$> eff) + +-- | Hoist the underlying Stream effect +-- Note this is not very efficience since, just like 'mapAccumL_', it also needs +-- to traverse and rebuild the whole stream. +hoistEff :: forall m n a b. (Applicative m, Monad n) => (forall x. m x -> n x) -> Stream m a b -> Stream n a b +hoistEff h s = Stream $ \f g -> hs f g (runStream s :: StreamS m a b) where + hs :: (a -> n r') + -> (b -> StreamS n r' r) + -> StreamS m a b + -> StreamS n r' r + hs f g x = case x of + Done d -> g d + Yield a r -> Effect (f a >>= \r' -> return $ Yield r' (hs f g r)) + Effect e -> Effect (h (hs f g <$> e)) + ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -27,6 +27,8 @@ import GHC.Cmm.Lint ( cmmLint ) import GHC.Cmm import GHC.Cmm.CLabel +import GHC.StgToCmm.CgUtils (CgStream) + import GHC.Driver.DynFlags import GHC.Driver.Config.Finder ( initFinderOpts ) import GHC.Driver.Config.CmmToAsm ( initNCGConfig ) @@ -37,7 +39,7 @@ import GHC.Driver.Backend import GHC.Data.OsPath import qualified GHC.Data.ShortText as ST -import GHC.Data.Stream ( Stream ) +import GHC.Data.Stream ( liftIO ) import qualified GHC.Data.Stream as Stream import GHC.Utils.TmpFs @@ -86,19 +88,21 @@ codeOutput -> [(ForeignSrcLang, FilePath)] -- ^ additional files to be compiled with the C compiler -> Set UnitId -- ^ Dependencies - -> Stream IO RawCmmGroup a -- Compiled C-- + -> DUniqSupply -- ^ The deterministic unique supply to run the CgStream. + -- See Note [Deterministic Uniques in the CG] + -> CgStream RawCmmGroup a -- ^ Compiled C-- -> IO (FilePath, (Bool{-stub_h_exists-}, Maybe FilePath{-stub_c_exists-}), [(ForeignSrcLang, FilePath)]{-foreign_fps-}, a) -codeOutput logger tmpfs llvm_config dflags unit_state this_mod filenm location genForeignStubs foreign_fps pkg_deps +codeOutput logger tmpfs llvm_config dflags unit_state this_mod filenm location genForeignStubs foreign_fps pkg_deps dus0 cmm_stream = do { -- Lint each CmmGroup as it goes past ; let linted_cmm_stream = if gopt Opt_DoCmmLinting dflags - then Stream.mapM do_lint cmm_stream + then Stream.mapM (liftIO . do_lint) cmm_stream else cmm_stream do_lint cmm = withTimingSilent logger @@ -115,25 +119,26 @@ codeOutput logger tmpfs llvm_config dflags unit_state this_mod filenm location g ; return cmm } - ; let final_stream :: Stream IO RawCmmGroup (ForeignStubs, a) + ; let final_stream :: CgStream RawCmmGroup (ForeignStubs, a) final_stream = do { a <- linted_cmm_stream ; let stubs = genForeignStubs a ; emitInitializerDecls this_mod stubs ; return (stubs, a) } + ; let dus1 = newTagDUniqSupply 'n' dus0 ; (stubs, a) <- case backendCodeOutput (backend dflags) of - NcgCodeOutput -> outputAsm logger dflags this_mod location filenm + NcgCodeOutput -> outputAsm logger dflags this_mod location filenm dus1 final_stream - ViaCCodeOutput -> outputC logger dflags filenm final_stream pkg_deps - LlvmCodeOutput -> outputLlvm logger llvm_config dflags filenm final_stream + ViaCCodeOutput -> outputC logger dflags filenm dus1 final_stream pkg_deps + LlvmCodeOutput -> outputLlvm logger llvm_config dflags filenm dus1 final_stream JSCodeOutput -> outputJS logger llvm_config dflags filenm final_stream ; stubs_exist <- outputForeignStubs logger tmpfs dflags unit_state this_mod location stubs ; return (filenm, stubs_exist, foreign_fps, a) } -- | See Note [Initializers and finalizers in Cmm] in GHC.Cmm.InitFini for details. -emitInitializerDecls :: Module -> ForeignStubs -> Stream IO RawCmmGroup () +emitInitializerDecls :: Module -> ForeignStubs -> CgStream RawCmmGroup () emitInitializerDecls this_mod (ForeignStubs _ cstub) | initializers <- getInitializers cstub , not $ null initializers = @@ -161,15 +166,18 @@ doOutput filenm io_action = bracket (openFile filenm WriteMode) hClose io_action outputC :: Logger -> DynFlags -> FilePath - -> Stream IO RawCmmGroup a + -> DUniqSupply -- ^ The deterministic uniq supply to run the CgStream + -- See Note [Deterministic Uniques in the CG] + -> CgStream RawCmmGroup a -> Set UnitId -> IO a -outputC logger dflags filenm cmm_stream unit_deps = +outputC logger dflags filenm dus cmm_stream unit_deps = withTiming logger (text "C codegen") (\a -> seq a () {- FIXME -}) $ do let pkg_names = map unitIdString (Set.toAscList unit_deps) - doOutput filenm $ \ h -> do - hPutStr h ("/* GHC_PACKAGES " ++ unwords pkg_names ++ "\n*/\n") - hPutStr h "#include \"Stg.h\"\n" + doOutput filenm $ \ h -> fmap fst $ runUDSMT dus $ do + liftIO $ do + hPutStr h ("/* GHC_PACKAGES " ++ unwords pkg_names ++ "\n*/\n") + hPutStr h "#include \"Stg.h\"\n" let platform = targetPlatform dflags writeC cmm = do let doc = cmmToC platform cmm @@ -179,7 +187,7 @@ outputC logger dflags filenm cmm_stream unit_deps = doc let ctx = initSDocContext dflags PprCode printSDocLn ctx LeftMode h doc - Stream.consume cmm_stream id writeC + Stream.consume cmm_stream id (liftIO . writeC) {- ************************************************************************ @@ -194,15 +202,19 @@ outputAsm :: Logger -> Module -> ModLocation -> FilePath - -> Stream IO RawCmmGroup a + -> DUniqSupply -- ^ The deterministic uniq supply to run the CgStream + -- See Note [Deterministic Uniques in the CG] + -> CgStream RawCmmGroup a -> IO a -outputAsm logger dflags this_mod location filenm cmm_stream = do - let ncg_uniqs = initDUniqSupply 'n' 0 +outputAsm logger dflags this_mod location filenm dus cmm_stream = do + -- Update tag of uniques in Stream debugTraceMsg logger 4 (text "Outputing asm to" <+> text filenm) let ncg_config = initNCGConfig dflags this_mod {-# SCC "OutputAsm" #-} doOutput filenm $ \h -> {-# SCC "NativeCodeGen" #-} - nativeCodeGen logger (toolSettings dflags) ncg_config location h ncg_uniqs cmm_stream + fmap fst $ + runUDSMT dus $ + nativeCodeGen logger (toolSettings dflags) ncg_config location h cmm_stream {- ************************************************************************ @@ -212,12 +224,15 @@ outputAsm logger dflags this_mod location filenm cmm_stream = do ************************************************************************ -} -outputLlvm :: Logger -> LlvmConfigCache -> DynFlags -> FilePath -> Stream IO RawCmmGroup a -> IO a -outputLlvm logger llvm_config dflags filenm cmm_stream = do +outputLlvm :: Logger -> LlvmConfigCache -> DynFlags -> FilePath + -> DUniqSupply -- ^ The deterministic uniq supply to run the CgStream + -- See Note [Deterministic Uniques in the CG] + -> CgStream RawCmmGroup a -> IO a +outputLlvm logger llvm_config dflags filenm dus cmm_stream = do lcg_config <- initLlvmCgConfig logger llvm_config dflags {-# SCC "llvm_output" #-} doOutput filenm $ \f -> {-# SCC "llvm_CodeGen" #-} - llvmCodeGen logger lcg_config f cmm_stream + llvmCodeGen logger lcg_config f dus cmm_stream {- ************************************************************************ @@ -226,7 +241,7 @@ outputLlvm logger llvm_config dflags filenm cmm_stream = do * * ************************************************************************ -} -outputJS :: Logger -> LlvmConfigCache -> DynFlags -> FilePath -> Stream IO RawCmmGroup a -> IO a +outputJS :: Logger -> LlvmConfigCache -> DynFlags -> FilePath -> CgStream RawCmmGroup a -> IO a outputJS _ _ _ _ _ = pgmError $ "codeOutput: Hit JavaScript case. We should never reach here!" ++ "\nThe JS backend should shortcircuit to StgToJS after Stg." ++ "\nIf you reached this point then you've somehow made it to Cmm!" ===================================== compiler/GHC/Driver/GenerateCgIPEStub.hs ===================================== @@ -13,7 +13,7 @@ import GHC.Cmm.Dataflow.Block (blockSplit, blockToList) import GHC.Cmm.Dataflow.Label import GHC.Cmm.Info.Build (emptySRT) import GHC.Cmm.Pipeline (cmmPipeline) -import GHC.Data.Stream (Stream, liftIO) +import GHC.Data.Stream (liftIO, liftEff) import qualified GHC.Data.Stream as Stream import GHC.Driver.Env (hsc_dflags, hsc_logger) import GHC.Driver.Env.Types (HscEnv) @@ -28,6 +28,7 @@ import GHC.StgToCmm.Monad (getCmm, initC, runC, initFCodeState) import GHC.StgToCmm.Prof (initInfoTableProv) import GHC.StgToCmm.Types (CmmCgInfos (..), ModuleLFInfos) import GHC.StgToCmm.Utils +import GHC.StgToCmm.CgUtils (CgStream) import GHC.Types.IPE (InfoTableProvMap (provInfoTables), IpeSourceLocation) import GHC.Types.Name.Set (NonCaffySet) import GHC.Types.Tickish (GenTickish (SourceNote)) @@ -199,11 +200,10 @@ generateCgIPEStub , ModuleLFInfos , Map CmmInfoTable (Maybe IpeSourceLocation) , IPEStats - , DUniqSupply , DetUniqFM ) - -> Stream IO CmmGroupSRTs CmmCgInfos -generateCgIPEStub hsc_env this_mod denv (nonCaffySet, moduleLFInfos, infoTablesWithTickishes, initStats, dus, detRnEnv) = do + -> CgStream CmmGroupSRTs CmmCgInfos +generateCgIPEStub hsc_env this_mod denv (nonCaffySet, moduleLFInfos, infoTablesWithTickishes, initStats, detRnEnv) = do let dflags = hsc_dflags hsc_env platform = targetPlatform dflags logger = hsc_logger hsc_env @@ -217,7 +217,7 @@ generateCgIPEStub hsc_env this_mod denv (nonCaffySet, moduleLFInfos, infoTablesW (_detRnEnv', rn_ipeCmmGroup) = detRenameCmmGroup detRnEnv ipeCmmGroup - (_, _, ipeCmmGroupSRTs) <- liftIO $ cmmPipeline logger cmm_cfg (emptySRT this_mod) dus rn_ipeCmmGroup + (_, ipeCmmGroupSRTs) <- liftEff $ withDUS $ cmmPipeline logger cmm_cfg (emptySRT this_mod) rn_ipeCmmGroup Stream.yield ipeCmmGroupSRTs ipeStub <- ===================================== compiler/GHC/Driver/Hooks.hs ===================================== @@ -61,6 +61,7 @@ import GHC.Core.Type import GHC.Tc.Types import GHC.Stg.Syntax +import GHC.StgToCmm.CgUtils (CgStream) import GHC.StgToCmm.Types (ModuleLFInfos) import GHC.StgToCmm.Config import GHC.Cmm @@ -150,8 +151,8 @@ data Hooks = Hooks , createIservProcessHook :: !(Maybe (CreateProcess -> IO ProcessHandle)) , stgToCmmHook :: !(Maybe (StgToCmmConfig -> InfoTableProvMap -> [TyCon] -> CollectedCCs -> [CgStgTopBinding] -> HpcInfo -> Stream IO CmmGroup ModuleLFInfos)) - , cmmToRawCmmHook :: !(forall a . Maybe (DynFlags -> Maybe Module -> Stream IO CmmGroupSRTs a - -> IO (Stream IO RawCmmGroup a))) + , cmmToRawCmmHook :: !(forall a . Maybe (DynFlags -> Maybe Module -> CgStream CmmGroupSRTs a + -> IO (CgStream RawCmmGroup a))) } {-# DEPRECATED cmmToRawCmmHook "cmmToRawCmmHook is being deprecated. If you do use it in your project, please raise a GHC issue!" #-} ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -208,6 +208,7 @@ import GHC.Builtin.Names import qualified GHC.StgToCmm as StgToCmm ( codeGen ) import GHC.StgToCmm.Types (CmmCgInfos (..), ModuleLFInfos, LambdaFormInfo(..)) +import GHC.StgToCmm.CgUtils (CgStream) import GHC.Cmm import GHC.Cmm.Info.Build @@ -268,7 +269,6 @@ import GHC.Data.Bag import GHC.Data.OsPath (unsafeEncodeUtf) import GHC.Data.StringBuffer import qualified GHC.Data.Stream as Stream -import GHC.Data.Stream (Stream) import GHC.Data.Maybe import GHC.SysTools (initSysTools) @@ -301,6 +301,7 @@ import GHC.StgToCmm.Utils (IPEStats) import GHC.Types.Unique.FM import GHC.Types.Unique.DFM import GHC.Cmm.Config (CmmConfig) +import Data.Bifunctor {- ********************************************************************** %* * @@ -1999,7 +2000,7 @@ hscGenHardCode hsc_env cgguts location output_filename = do let dump a = do unless (null a) $ putDumpFileMaybe logger Opt_D_dump_cmm_raw "Raw Cmm" FormatCMM (pdoc platform a) return a - rawcmms1 = Stream.mapM dump rawcmms0 + rawcmms1 = Stream.mapM (liftIO . dump) rawcmms0 let foreign_stubs st = foreign_stubs0 `appendStubC` prof_init @@ -2008,7 +2009,7 @@ hscGenHardCode hsc_env cgguts location output_filename = do (output_filename, (_stub_h_exists, stub_c_exists), foreign_fps, cmm_cg_infos) <- {-# SCC "codeOutput" #-} codeOutput logger tmpfs llvm_config dflags (hsc_units hsc_env) this_mod output_filename location - foreign_stubs foreign_files dependencies rawcmms1 + foreign_stubs foreign_files dependencies (initDUniqSupply 'n' 0) rawcmms1 return ( output_filename, stub_c_exists, foreign_fps , Just stg_cg_infos, Just cmm_cg_infos) @@ -2137,9 +2138,9 @@ hscCompileCmmFile hsc_env original_filename filename output_filename = runHsc hs -- Re-ordering here causes breakage when booting with C backend because -- in C we must declare before use, but SRT algorithm is free to -- re-order [A, B] (B refers to A) when A is not CAFFY and return [B, A] - cmmgroup <- concat . snd <$> + ((_,dus1), cmmgroup) <- second concat <$> mapAccumLM (\(msrt0, dus0) cmm -> do - (msrt1, dus1, cmm') <- cmmPipeline logger cmm_config msrt0 dus0 [cmm] + ((msrt1, cmm'), dus1) <- cmmPipeline logger cmm_config msrt0 [cmm] dus0 return ((msrt1, dus1), cmm')) (emptySRT cmm_mod, initDUniqSupply 'u' 1) cmm unless (null cmmgroup) $ @@ -2148,12 +2149,12 @@ hscCompileCmmFile hsc_env original_filename filename output_filename = runHsc hs rawCmms0 <- case cmmToRawCmmHook hooks of Nothing -> cmmToRawCmm logger profile (Stream.yield cmmgroup) - Just h -> h dflags Nothing (Stream.yield cmmgroup) + Just h -> error "same as error below" $ h dflags Nothing (Stream.yield cmmgroup) let dump a = do unless (null a) $ putDumpFileMaybe logger Opt_D_dump_cmm_raw "Raw Cmm" FormatCMM (pdoc platform a) return a - rawCmms = Stream.mapM dump rawCmms0 + rawCmms = Stream.mapM (liftIO . dump) rawCmms0 let foreign_stubs _ | not $ null ipe_ents = @@ -2162,7 +2163,7 @@ hscCompileCmmFile hsc_env original_filename filename output_filename = runHsc hs | otherwise = NoStubs (_output_filename, (_stub_h_exists, stub_c_exists), _foreign_fps, _caf_infos) <- codeOutput logger tmpfs llvm_config dflags (hsc_units hsc_env) cmm_mod output_filename no_loc foreign_stubs [] S.empty - rawCmms + dus1 rawCmms return stub_c_exists where no_loc = OsPathModLocation @@ -2196,7 +2197,7 @@ doCodeGen :: HscEnv -> Module -> InfoTableProvMap -> [TyCon] -> CollectedCCs -> [CgStgTopBinding] -- ^ Bindings come already annotated with fvs -> HpcInfo - -> IO (Stream IO CmmGroupSRTs CmmCgInfos) + -> IO (CgStream CmmGroupSRTs CmmCgInfos) -- Note we produce a 'Stream' of CmmGroups, so that the -- backend can be run incrementally. Otherwise it generates all -- the C-- up front, which has a significant space cost. @@ -2214,9 +2215,9 @@ doCodeGen hsc_env this_mod denv data_tycons let stg_to_cmm dflags mod a b c d e = case stgToCmmHook hooks of Nothing -> StgToCmm.codeGen logger tmpfs (initStgToCmmConfig dflags mod) a b c d e - Just h -> (,emptyDetUFM) <$> h (initStgToCmmConfig dflags mod) a b c d e + Just h -> error "should we change the API or implement hoist?" $ (,emptyDetUFM) <$> h (initStgToCmmConfig dflags mod) a b c d e - let cmm_stream :: Stream IO CmmGroup (ModuleLFInfos, DetUniqFM) + let cmm_stream :: CgStream CmmGroup (ModuleLFInfos, DetUniqFM) -- See Note [Forcing of stg_binds] cmm_stream = stg_binds_w_fvs `seqList` {-# SCC "StgToCmm" #-} stg_to_cmm dflags this_mod denv data_tycons cost_centre_info stg_binds_w_fvs hpc_info @@ -2232,45 +2233,45 @@ doCodeGen hsc_env this_mod denv data_tycons "Cmm produced by codegen" FormatCMM (pdoc platform a) return a - ppr_stream1 = Stream.mapM dump1 cmm_stream + ppr_stream1 = Stream.mapM (liftIO . dump1) cmm_stream cmm_config = initCmmConfig dflags - pipeline_stream :: Stream IO CmmGroupSRTs CmmCgInfos + pipeline_stream :: CgStream CmmGroupSRTs CmmCgInfos pipeline_stream = do - ((mod_srt_info, ipes, ipe_stats, dus), (lf_infos, detRnEnv)) <- + ((mod_srt_info, ipes, ipe_stats), (lf_infos, detRnEnv)) <- {-# SCC "cmmPipeline" #-} - Stream.mapAccumL_ (pipeline_action logger cmm_config) (emptySRT this_mod, M.empty, mempty, initDUniqSupply 'u' 1) ppr_stream1 + Stream.mapAccumL_ (pipeline_action logger cmm_config) (emptySRT this_mod, M.empty, mempty) ppr_stream1 let nonCaffySet = srtMapNonCAFs (moduleSRTMap mod_srt_info) - cmmCgInfos <- generateCgIPEStub hsc_env this_mod denv (nonCaffySet, lf_infos, ipes, ipe_stats, dus, detRnEnv) + cmmCgInfos <- generateCgIPEStub hsc_env this_mod denv (nonCaffySet, lf_infos, ipes, ipe_stats, detRnEnv) return cmmCgInfos pipeline_action :: Logger -> CmmConfig - -> (ModuleSRTInfo, Map CmmInfoTable (Maybe IpeSourceLocation), IPEStats, DUniqSupply) + -> (ModuleSRTInfo, Map CmmInfoTable (Maybe IpeSourceLocation), IPEStats) -> CmmGroup - -> IO ((ModuleSRTInfo, Map CmmInfoTable (Maybe IpeSourceLocation), IPEStats, DUniqSupply), CmmGroupSRTs) - pipeline_action logger cmm_config (mod_srt_info, ipes, stats, dus) cmm_group = do - (mod_srt_info', dus', cmm_srts) <- cmmPipeline logger cmm_config mod_srt_info dus cmm_group + -> UniqDSMT IO ((ModuleSRTInfo, Map CmmInfoTable (Maybe IpeSourceLocation), IPEStats), CmmGroupSRTs) + pipeline_action logger cmm_config (mod_srt_info, ipes, stats) cmm_group = do + (mod_srt_info', cmm_srts) <- withDUS $ cmmPipeline logger cmm_config mod_srt_info cmm_group -- If -finfo-table-map is enabled, we precompute a map from info -- tables to source locations. See Note [Mapping Info Tables to Source -- Positions] in GHC.Stg.Debug. (ipes', stats') <- if (gopt Opt_InfoTableMap dflags) then - lookupEstimatedTicks hsc_env ipes stats cmm_srts + liftIO $ lookupEstimatedTicks hsc_env ipes stats cmm_srts else return (ipes, stats) - return ((mod_srt_info', ipes', stats', dus'), cmm_srts) + return ((mod_srt_info', ipes', stats'), cmm_srts) dump2 a = do unless (null a) $ putDumpFileMaybe logger Opt_D_dump_cmm "Output Cmm" FormatCMM (pdoc platform a) return a - return $ Stream.mapM dump2 pipeline_stream + return $ Stream.mapM (liftIO . dump2) pipeline_stream myCoreToStg :: Logger -> DynFlags -> [Var] -> Bool ===================================== compiler/GHC/StgToCmm.hs ===================================== @@ -27,6 +27,7 @@ import GHC.StgToCmm.Config import GHC.StgToCmm.Hpc import GHC.StgToCmm.Ticky import GHC.StgToCmm.Types (ModuleLFInfos) +import GHC.StgToCmm.CgUtils (CgStream) import GHC.Cmm import GHC.Cmm.Utils @@ -77,7 +78,7 @@ codeGen :: Logger -> CollectedCCs -- (Local/global) cost-centres needing declaring/registering. -> [CgStgTopBinding] -- Bindings to convert -> HpcInfo - -> Stream IO CmmGroup (ModuleLFInfos, DetUniqFM) + -> CgStream CmmGroup (ModuleLFInfos, DetUniqFM) -- See Note [Deterministic Uniques in the NCG] on why UniqDSMT -- Output as a stream, so codegen can -- be interleaved with output @@ -90,7 +91,7 @@ codeGen logger tmpfs cfg (InfoTableProvMap denv _ _) data_tycons ; cgref <- liftIO $ initC >>= \s -> newIORef s ; uniqRnRef <- liftIO $ newIORef emptyDetUFM ; let fstate = initFCodeState $ stgToCmmPlatform cfg - ; let cg :: FCode a -> Stream IO CmmGroup a + ; let cg :: FCode a -> CgStream CmmGroup a cg fcode = do (a, cmm) <- liftIO . withTimingSilent logger (text "STG -> Cmm") (`seq` ()) $ do st <- readIORef cgref ===================================== compiler/GHC/StgToCmm/CgUtils.hs ===================================== @@ -14,6 +14,9 @@ module GHC.StgToCmm.CgUtils ( get_Regtable_addr_from_offset, regTableOffset, get_GlobalReg_addr, + + -- * Streaming for CG + CgStream ) where import GHC.Prelude @@ -28,6 +31,21 @@ import GHC.Cmm.CLabel import GHC.Utils.Panic import GHC.Cmm.Dataflow.Label +import GHC.Data.Stream (Stream) +import GHC.Types.Unique.DSM (UniqDSMT) + +-- ----------------------------------------------------------------------------- +-- Streaming + +-- | The Stream instantiation used for code generation. +-- Note the underlying monad is @UniqDSMT IO@, where @UniqDSMT@ is a transformer +-- that propagates a deterministic unique supply (essentially an incrementing +-- counter) from which new uniques are deterministically created during the +-- code generation stages following StgToCmm. +-- See Note [Object determinism]. +type CgStream = Stream (UniqDSMT IO) + + -- ----------------------------------------------------------------------------- -- Information about global registers ===================================== compiler/GHC/StgToCmm/Types.hs ===================================== @@ -22,7 +22,6 @@ import GHC.Types.Name.Set import GHC.Utils.Outputable - {- Note [Conveying CAF-info and LFInfo between modules] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Types/Unique/DSM.hs ===================================== @@ -8,11 +8,15 @@ import Control.Monad.Fix import GHC.Types.Unique import qualified GHC.Utils.Monad.State.Strict as Strict import qualified GHC.Types.Unique.Supply as USM +import qualified Control.Monad.Trans.State.Strict as T +import Control.Monad.IO.Class {- -Note [Deterministic Uniques in the NCG] +Note [Deterministic Uniques in the CG] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Explain why should the Cmm stream be UniqDSMT IO starting from StgToCmm.codeGen + (TODO: Is there anything about locality that I need to add? better to avoid if possible. Need to double check.) See also Note [Object determinism] in GHC.StgToCmm @@ -31,6 +35,9 @@ See also Note [Object determinism] in GHC.StgToCmm -- pass. -- todo:check: UniqSM is only used before Cmm (grep for it), afterwards only UniqDSM is used. + +-Describe how we thread through the uniqds in the backend that is streamed by using UniqDSMT IO in the Stream. +That's the only sane way to thread it through all the Streaming -} newtype DUniqSupply = DUS Word64 -- supply uniques iteratively @@ -90,6 +97,45 @@ class Monad m => MonadGetUnique m where instance MonadGetUnique UniqDSM where getUniqueM = getUniqueDSM +-- non deterministic instance instance MonadGetUnique USM.UniqSM where getUniqueM = USM.getUniqueM +-------------------------------------------------------------------------------- +-- UniqDSMT +-------------------------------------------------------------------------------- + +-- | Transformer version of 'UniqDSM' to use when threading a deterministic +-- uniq supply over a Monad. Specifically, it is used in the `Stream` of Cmm +-- decls. +newtype UniqDSMT m result = UDSMT { unUDSMT :: T.StateT DUniqSupply m (result) } + deriving (Functor, Applicative, Monad, MonadIO) + +instance Monad m => MonadGetUnique (UniqDSMT m) where + getUniqueM = UDSMT $ do + us <- T.get + let (u, us') = takeUniqueFromDSupply us + T.put us' + return u + +-- | Like 'runUniqueDSM' but for 'UniqDSMT' +runUDSMT :: DUniqSupply -> UniqDSMT m a -> m (a, DUniqSupply) +runUDSMT dus (UDSMT st) = T.runStateT st dus + +-- | Lift an IO action that depends on, and threads through, a unique supply +-- into UniqDSMT IO. +withDUS :: (DUniqSupply -> IO (a, DUniqSupply)) -> UniqDSMT IO a +withDUS f = UDSMT $ do + us <- T.get + (a, us') <- liftIO (f us) + _ <- T.put us' + return a + +-- | Change the monad underyling an applied @UniqDSMT@, i.e. transform a +-- @UniqDSMT m@ into a @UniqDSMT n@ given @m ~> n at . +hoistUDSMT :: (forall x. m x -> n x) -> UniqDSMT m a -> UniqDSMT n a +hoistUDSMT nt (UDSMT s) = UDSMT $ T.mapStateT nt s + +-- | Lift a monadic action @m a@ into an @UniqDSMT m a@ +liftUDSMT :: Functor m => m a -> UniqDSMT m a +liftUDSMT m = UDSMT $ T.StateT $ \s -> (,s) <$> m ===================================== compiler/GHC/Types/Unique/Supply.hs ===================================== @@ -171,8 +171,7 @@ and hardcode the tag into the MonadUnique instance. On top of all the benefits of threading the tag this *also* has the benefit of avoiding the tag getting captured in thunks, or being passed around at runtime. It does however come at the cost of having to use a fixed tag for all -code run in this Monad. But remember, the tag is purely cosmetic: -See Note [Uniques and tags]. +code run in this Monad. The tag is mostly cosmetic: See Note [Uniques and tags]. NB: It's *not* an optimization to pass around the UniqSupply inside an IORef instead of the tag. While this would avoid frequent state updates @@ -203,9 +202,8 @@ data UniqSupply mkSplitUniqSupply :: Char -> IO UniqSupply -- ^ Create a unique supply out of thin air. --- The "tag" (Char) supplied is purely cosmetic, making it easier --- to figure out where a Unique was born. See --- Note [Uniques and tags]. +-- The "tag" (Char) supplied is mostly cosmetic, making it easier +-- to figure out where a Unique was born. See Note [Uniques and tags]. -- -- The payload part of the Uniques allocated from this UniqSupply are -- guaranteed distinct wrt all other supplies, regardless of their "tag". View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0e766d8eb501a9f37802d119777117e361e09649...90b8bdb0b3bd298c31e48694546bf6ca2d2720b3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0e766d8eb501a9f37802d119777117e361e09649...90b8bdb0b3bd298c31e48694546bf6ca2d2720b3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 14:51:05 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Tue, 10 Sep 2024 10:51:05 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/torsten.schmits/persistent-worker-test-2024-09-10 Message-ID: <66e05cd964ccd_f8848b895fc86398@gitlab.mail> Torsten Schmits pushed new branch wip/torsten.schmits/persistent-worker-test-2024-09-10 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/torsten.schmits/persistent-worker-test-2024-09-10 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 14:53:21 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Tue, 10 Sep 2024 10:53:21 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/andreask/ghc-primops Message-ID: <66e05d61e5272_f8848bc1b8c8658e@gitlab.mail> Andreas Klebinger pushed new branch wip/andreask/ghc-primops at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/andreask/ghc-primops You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 15:07:47 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Tue, 10 Sep 2024 11:07:47 -0400 Subject: [Git][ghc/ghc][wip/andreask/ghc-primops] ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps Message-ID: <66e060c3e0707_f8848598f449718@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/ghc-primops at Glasgow Haskell Compiler / GHC Commits: ac9b303d by Andreas Klebinger at 2024-09-10T16:48:53+02:00 ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps This will be the new place for functions that would have gone into GHC.Exts in the past but are not stable enough to do so now. Addresses #25242 - - - - - 2 changed files: - libraries/ghc-experimental/ghc-experimental.cabal.in - + libraries/ghc-experimental/src/GHC/PrimOps.hs Changes: ===================================== libraries/ghc-experimental/ghc-experimental.cabal.in ===================================== @@ -29,6 +29,7 @@ library exposed-modules: Data.Sum.Experimental Data.Tuple.Experimental + GHC.PrimOps GHC.Profiling.Eras GHC.TypeLits.Experimental GHC.TypeNats.Experimental ===================================== libraries/ghc-experimental/src/GHC/PrimOps.hs ===================================== @@ -0,0 +1,31 @@ +{-# LANGUAGE FlexibleInstances #-} +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE MultiParamTypeClasses #-} +{-# LANGUAGE NoImplicitPrelude #-} +{-# LANGUAGE TypeFamilies #-} +{-# LANGUAGE UnboxedTuples #-} +{-# LANGUAGE Unsafe #-} +{-# LANGUAGE DeriveDataTypeable #-} +{-# OPTIONS_HADDOCK print-explicit-runtime-reps #-} + +----------------------------------------------------------------------------- +-- | +-- Module : GHC.PrimOps +-- Copyright : Andreas Klebinger 2024 +-- License : see libraries/ghc-experimental/LICENSE +-- +-- Maintainer : ghc-devs at haskell.org +-- Stability : internal +-- Portability : non-portable (GHC Extensions) +-- +-- GHC Extensions: This is the Approved Way to get at GHC-specific extensions +-- without relying on the ghc-internal package. +----------------------------------------------------------------------------- + +module GHC.PrimOps + ( + module GHC.Internal.Exts, + ) where + +import GHC.Internal.Exts + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ac9b303dda899ca82b5671375aac1dfcc485a62f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ac9b303dda899ca82b5671375aac1dfcc485a62f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 15:49:39 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 11:49:39 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 7 commits: RTS linker: add support for hidden symbols (#25191) Message-ID: <66e06a9317b_1ae7801a1b70791cb@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - ad0974af by Hécate Kleidukos at 2024-09-10T11:49:26-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 65c07577 by Hécate Kleidukos at 2024-09-10T11:49:26-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 43f18f86 by Hécate Kleidukos at 2024-09-10T11:49:26-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - 1c0fe5eb by Simon Peyton Jones at 2024-09-10T11:49:27-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 30 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/StgToJS/Symbols.hs - compiler/GHC/StgToJS/Types.hs - compiler/GHC/Tc/Types.hs - compiler/GHC/Tc/Utils/Monad.hs - compiler/GHC/Tc/Zonk/Type.hs - libraries/ghc-heap/tests/stack_misc_closures_c.c - rts/Linker.c - rts/LinkerInternals.h - rts/linker/Elf.c - rts/linker/ElfTypes.h - rts/linker/PEi386.c - testsuite/tests/perf/compiler/T11068.stdout - testsuite/tests/pmcheck/should_compile/T12957.stderr - + testsuite/tests/pmcheck/should_compile/T24817.hs - testsuite/tests/pmcheck/should_compile/all.T - testsuite/tests/printer/T17697.stderr - testsuite/tests/profiling/should_run/staticcallstack002.stdout - testsuite/tests/rts/linker/Makefile - + testsuite/tests/rts/linker/T25191.hs - + testsuite/tests/rts/linker/T25191.stdout - + testsuite/tests/rts/linker/T25191_foo1.c - + testsuite/tests/rts/linker/T25191_foo2.c - testsuite/tests/rts/linker/all.T - testsuite/tests/rts/pause-resume/list_threads_and_misc_roots_c.c - testsuite/tests/simplCore/should_compile/T13156.stdout - testsuite/tests/typecheck/should_fail/T13292.stderr - + utils/haddock/haddock-api/compat/posix/Haddock/Compat.hs - + utils/haddock/haddock-api/compat/windows/Haddock/Compat.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/24952620517344cc2def2c7ee0ee78e76b105980...1c0fe5eb122f82e121e71fa5fbd4f6060b5646dc -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/24952620517344cc2def2c7ee0ee78e76b105980...1c0fe5eb122f82e121e71fa5fbd4f6060b5646dc You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 16:33:42 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Tue, 10 Sep 2024 12:33:42 -0400 Subject: [Git][ghc/ghc][wip/T24978] 75 commits: haddock: decrease margin on top of small headings Message-ID: <66e074e69711b_1ae7803aacb48504f@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24978 at Glasgow Haskell Compiler / GHC Commits: af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - fd7ec4b4 by Simon Peyton Jones at 2024-09-09T12:02:44+01:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to address As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv - - - - - 2076ab66 by Simon Peyton Jones at 2024-09-09T12:02:45+01:00 Rename GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 2e6ecfdf by Simon Peyton Jones at 2024-09-09T12:02:45+01:00 Reduce allocation - - - - - 8399e88f by Simon Peyton Jones at 2024-09-09T12:02:45+01:00 Accept CountDeps changes - - - - - a1f66dcb by Simon Peyton Jones at 2024-09-10T17:08:11+01:00 Wibble - - - - - 22 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/86e6e87bd5da49eb65b9d74c611fc407c33ee0a9...a1f66dcb6e8d17f2c89b2a82f6a822fd64817886 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/86e6e87bd5da49eb65b9d74c611fc407c33ee0a9...a1f66dcb6e8d17f2c89b2a82f6a822fd64817886 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 16:39:12 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Tue, 10 Sep 2024 12:39:12 -0400 Subject: [Git][ghc/ghc][wip/T24359] 330 commits: Migrate `Finder` component to `OsPath`, fixed #24616 Message-ID: <66e07630d81f1_1ae780563f10871fb@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24359 at Glasgow Haskell Compiler / GHC Commits: c8ece0df by Fendor at 2024-06-03T19:43:22-04:00 Migrate `Finder` component to `OsPath`, fixed #24616 For each module in a GHCi session, we keep alive one `ModLocation`. A `ModLocation` is fairly inefficiently packed, as `String`s are expensive in memory usage. While benchmarking the agda codebase, we concluded that we keep alive around 11MB of `FilePath`'s, solely retained by `ModLocation`. We provide a more densely packed encoding of `ModLocation`, by moving from `FilePath` to `OsPath`. Further, we migrate the full `Finder` component to `OsPath` to avoid unnecessary transformations. As the `Finder` component is well-encapsulated, this requires only a minimal amount of changes in other modules. We introduce pattern synonym for 'ModLocation' which maintains backwards compatibility and avoids breaking consumers of 'ModLocation'. - - - - - 0cff083a by Cheng Shao at 2024-06-03T19:43:58-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. - - - - - 18f63970 by Sebastian Graf at 2024-06-04T05:05:27-04:00 Parser: Remove unused `apats` rule - - - - - 38757c30 by David Knothe at 2024-06-04T05:05:27-04:00 Implement Or Patterns (#22596) This commit introduces a new language extension, `-XOrPatterns`, as described in GHC Proposal 522. An or-pattern `pat1; ...; patk` succeeds iff one of the patterns `pat1`, ..., `patk` succeed, in this order. See also the summary `Note [Implmentation of OrPatterns]`. Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> - - - - - 395412e8 by Cheng Shao at 2024-06-04T05:06:04-04:00 compiler/ghci/rts: remove stdcall support completely We have formally dropped i386 windows support (#18487) a long time ago. The stdcall foreign call convention is only used by i386 windows, and the legacy logic around it is a significant maintenance burden for future work that adds arm64 windows support (#24603). Therefore, this patch removes stdcall support completely from the compiler as well as the RTS (#24883): - stdcall is still recognized as a FFI calling convention in Haskell syntax. GHC will now unconditionally emit a warning (-Wunsupported-calling-conventions) and treat it as ccall. - Apart from minimum logic to support the parsing and warning logic, all other code paths related to stdcall has been completely stripped from the compiler. - ghci only supports FFI_DEFAULT_ABI and ccall convention from now on. - FFI foreign export adjustor code on all platforms no longer handles the stdcall case and only handles ccall from now on. - The Win32 specific parts of RTS no longer has special code paths for stdcall. This commit is the final nail on the coffin for i386 windows support. Further commits will perform more housecleaning to strip the legacy code paths and pave way for future arm64 windows support. - - - - - d1fe9ab6 by Cheng Shao at 2024-06-04T05:06:04-04:00 rts: remove legacy i386 windows code paths This commit removes some legacy i386 windows related code paths in the RTS, given this target is no longer supported. - - - - - a605e4b2 by Cheng Shao at 2024-06-04T05:06:04-04:00 autoconf: remove i386 windows related logic This commit removes legacy i386 windows logic in autoconf scripts. - - - - - 91e5ac5e by Cheng Shao at 2024-06-04T05:06:04-04:00 llvm-targets: remove i386 windows support This commit removes i386 windows from llvm-targets and the script to generate it. - - - - - 65fe75a4 by Cheng Shao at 2024-06-04T05:06:04-04:00 libraries/utils: remove stdcall related legacy logic This commit removes stdcall related legacy logic in libraries and utils. ccall should be used uniformly for all supported windows hosts from now on. - - - - - d2a83302 by Cheng Shao at 2024-06-04T05:06:04-04:00 testsuite: adapt the testsuite for stdcall removal This patch adjusts test cases to handle the stdcall removal: - Some stdcall usages are replaced with ccall since stdcall doesn't make sense anymore. - We also preserve some stdcall usages, and check in the expected warning messages to ensure GHC always warn about stdcall usages (-Wunsupported-calling-conventions) as expected. - Error code testsuite coverage is slightly improved, -Wunsupported-calling-conventions is now tested. - Obsolete code paths related to i386 windows are also removed. - - - - - cef8f47a by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: minor adjustments for stdcall removal This commit include minor adjustments of documentation related to stdcall removal. - - - - - 54332437 by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: mention i386 Windows removal in 9.12 changelog This commit mentions removal of i386 Windows support and stdcall related change in the 9.12 changelog. - - - - - 2aaea8a1 by Cheng Shao at 2024-06-04T05:06:40-04:00 hadrian: improve user settings documentation This patch adds minor improvements to hadrian user settings documentation: - Add missing `ghc.cpp.opts` case - Remove non-existent `cxx` case - Clarify `cc.c.opts` also works for C++, while `cc.deps.opts` doesn't - Add example of passing configure argument to autoconf packages - - - - - 71010381 by Alex Mason at 2024-06-04T12:09:07-04:00 Add AArch64 CLZ, CTZ, RBIT primop implementations. Adds support for emitting the clz and rbit instructions, which are used by GHC.Prim.clz*#, GHC.Prim.ctz*# and GHC.Prim.bitReverse*#. - - - - - 44e2abfb by Cheng Shao at 2024-06-04T12:09:43-04:00 hadrian: add +text_simdutf flavour transformer to allow building text with simdutf This patch adds a +text_simdutf flavour transformer to hadrian to allow downstream packagers and users that build from source to opt-in simdutf support for text, in order to benefit from SIMD speedup at run-time. It's still disabled by default for the time being. - - - - - 077cb2e1 by Cheng Shao at 2024-06-04T12:09:43-04:00 ci: enable +text_simdutf flavour transformer for wasm jobs This commit enables +text_simdutf flavour transformer for wasm jobs, so text is now built with simdutf support for wasm. - - - - - b23746ad by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Use TemplateHaskellQuotes in instance Lift ByteArray Resolves #24852 - - - - - 3fd25743 by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Mark addrToByteArray as NOINLINE This function should never be inlined in order to keep code size small. - - - - - 98ad1ea5 by Cheng Shao at 2024-06-04T22:51:26-04:00 compiler: remove unused CompilerInfo/LinkerInfo types This patch removes CompilerInfo/LinkerInfo types from the compiler since they aren't actually used anywhere. - - - - - 11795244 by Cheng Shao at 2024-06-05T06:33:17-04:00 rts: remove unused PowerPC/IA64 native adjustor code This commit removes unused PowerPC/IA64 native adjustor code which is never actually enabled by autoconf/hadrian. Fixes #24920. - - - - - 5132754b by Sylvain Henry at 2024-06-05T06:33:57-04:00 RTS: fix warnings with doing*Profiling (#24918) - - - - - accc8c33 by Cheng Shao at 2024-06-05T11:35:36-04:00 hadrian: don't depend on inplace/mingw when --enable-distro-toolchain on Windows - - - - - 6ffbd678 by Cheng Shao at 2024-06-05T11:35:37-04:00 autoconf: normalize paths of some build-time dependencies on Windows This commit applies path normalization via cygpath -m to some build-time dependencies on Windows. Without this logic, the /clang64/bin prefixed msys2-style paths cause the build to fail with --enable-distro-toolchain. - - - - - 075dc6d4 by Cheng Shao at 2024-06-05T11:36:12-04:00 hadrian: remove OSDarwin mention from speedHack This commit removes mentioning of OSDarwin from speedHack, since speedHack is purely for i386 and we no longer support i386 darwin (#24921). - - - - - 83235c4c by Cheng Shao at 2024-06-05T11:36:12-04:00 compiler: remove 32-bit darwin logic This commit removes all 32-bit darwin logic from the compiler, given we no longer support 32-bit apple systems (#24921). Also contains a bit more cleanup of obsolete i386 windows logic. - - - - - 1eb99bc3 by Cheng Shao at 2024-06-05T11:36:12-04:00 rts: remove 32-bit darwin/ios logic This commit removes 32-bit darwin/ios related logic from the rts, given we no longer support them (#24921). - - - - - 24f65892 by Cheng Shao at 2024-06-05T11:36:12-04:00 llvm-targets: remove 32-bit darwin/ios targets This commit removes 32-bit darwin/ios targets from llvm-targets given we no longer support them (#24921). - - - - - ccdbd689 by Cheng Shao at 2024-06-05T11:36:12-04:00 testsuite: remove 32-bit darwin logic This commit removes 32-bit darwin logic from the testsuite given it's no longer supported (#24921). Also contains more cleanup of obsolete i386 windows logic. - - - - - 11d661c4 by Cheng Shao at 2024-06-05T11:36:13-04:00 docs: mention 32-bit darwin/ios removal in 9.12 changelog This commit mentions removal of 32-bit darwin/ios support (#24921) in the 9.12 changelog. - - - - - 7c173310 by Georgi Lyubenov at 2024-06-05T15:17:22-04:00 Add firstA and secondA to Data.Bitraversable Please see https://github.com/haskell/core-libraries-committee/issues/172 for related discussion - - - - - 3b6f9fd1 by Ben Gamari at 2024-06-05T15:17:59-04:00 base: Fix name of changelog Fixes #24899. Also place it under `extra-doc-files` to better reflect its nature and avoid triggering unnecessary recompilation if it changes. - - - - - 1f4d2ef7 by Sebastian Graf at 2024-06-05T15:18:34-04:00 Announce Or-patterns in the release notes for GHC 9.12 (#22596) Leftover from !9229. - - - - - 8650338d by Jan Hrček at 2024-06-06T10:39:24-04:00 Improve haddocks of Language.Haskell.Syntax.Pat.Pat - - - - - 2eee65e1 by Cheng Shao at 2024-06-06T10:40:00-04:00 testsuite: bump T7653 timeout for wasm - - - - - 990fed60 by Sylvain Henry at 2024-06-07T14:45:23-04:00 StgToCmm: refactor opTranslate and friends - Change arguments order to avoid `\args -> ...` lambdas - Fix documentation - Rename StgToCmm options ("big" doesn't mean anything) - - - - - 1afad514 by Sylvain Henry at 2024-06-07T14:45:23-04:00 NCG x86: remove dead code (#5444) Since 6755d833af8c21bbad6585144b10e20ac4a0a1ab this code is dead. - - - - - 595c0894 by Cheng Shao at 2024-06-07T14:45:58-04:00 testsuite: skip objc-hi/objcxx-hi when cross compiling objc-hi/objcxx-hi should be skipped when cross compiling. The existing opsys('darwin') predicate only asserts the host system is darwin but tells us nothing about the target, hence the oversight. - - - - - edfe6140 by qqwy at 2024-06-08T11:23:54-04:00 Replace '?callStack' implicit param with HasCallStack in GHC.Internal.Exception.throw - - - - - 35a64220 by Cheng Shao at 2024-06-08T11:24:30-04:00 rts: cleanup inlining logic This patch removes pre-C11 legacy code paths related to INLINE_HEADER/STATIC_INLINE/EXTERN_INLINE macros, ensure EXTERN_INLINE is treated as static inline in most cases (fixes #24945), and also corrects the comments accordingly. - - - - - 9ea90ed2 by Andrew Lelechenko at 2024-06-08T11:25:06-04:00 CODEOWNERS: add @core-libraries to track base interface changes A low-tech tactical solution for #24919 - - - - - 580fef7b by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update CHANGELOG to reflect current version - - - - - 391ecff5 by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update prologue.txt to reflect package description - - - - - 3dca3b7d by Ben Gamari at 2024-06-09T01:27:57-04:00 compiler: Clarify comment regarding need for MOVABS The comment wasn't clear in stating that it was only applicable to immediate source and memory target operands. - - - - - 6bd850e8 by doyougnu at 2024-06-09T21:02:14-04:00 JS: establish single source of truth for symbols In pursuit of: #22736. This MR moves ad-hoc symbols used throughout the js backend into a single symbols file. Why? First, this cleans up the code by removing ad-hoc strings created on the fly and therefore makes the code more maintainable. Second, it makes it much easier to eventually type these identifiers. - - - - - f3017dd3 by Cheng Shao at 2024-06-09T21:02:49-04:00 rts: replace ad-hoc MYTASK_USE_TLV with proper CC_SUPPORTS_TLS This patch replaces the ad-hoc `MYTASK_USE_TLV` with the `CC_SUPPORTS_TLS` macro. If TLS support is detected by autoconf, then we should use that for managing `myTask` in the threaded RTS. - - - - - e17d7e8c by Ben Gamari at 2024-06-11T05:25:21-04:00 users-guide: Fix stylistic issues in 9.12 release notes - - - - - 8a8a982a by Hugo Peters at 2024-06-11T05:25:57-04:00 fix typo in the simplifier debug output: baling -> bailing - - - - - 16475bb8 by Hécate Moonlight at 2024-06-12T03:07:55-04:00 haddock: Correct the Makefile to take into account Darwin systems - - - - - a2f60da5 by Hécate Kleidukos at 2024-06-12T03:08:35-04:00 haddock: Remove obsolete links to github.com/haskell/haddock in the docs - - - - - de4395cd by qqwy at 2024-06-12T03:09:12-04:00 Add `__GLASGOW_HASKELL_ASSERTS_IGNORED__` as CPP macro name if `-fasserts-ignored is set. This allows users to create their own Control.Exception.assert-like functionality that does something other than raising an `AssertFailed` exception. Fixes #24967 - - - - - 0e9c4dee by Ryan Hendrickson at 2024-06-12T03:09:53-04:00 compiler: add hint to TcRnBadlyStaged message - - - - - 2747cd34 by Simon Peyton Jones at 2024-06-12T12:51:37-04:00 Fix a QuickLook bug This MR fixes the bug exposed by #24676. The problem was that quickLookArg was trying to avoid calling tcInstFun unnecessarily; but it was in fact necessary. But that in turn forced me into a significant refactoring, putting more fields into EValArgQL. Highlights: see Note [Quick Look overview] in GHC.Tc.Gen.App * Instantiation variables are now distinguishable from ordinary unification variables, by level number = QLInstVar. This is treated like "level infinity". See Note [The QLInstVar TcLevel] in GHC.Tc.Utils.TcType. * In `tcApp`, we don't track the instantiation variables in a set Delta any more; instead, we just tell them apart by their level number. * EValArgQL now much more clearly captures the "half-done" state of typechecking an argument, ready for later resumption. See Note [Quick Look at value arguments] in GHC.Tc.Gen.App * Elminated a bogus (never used) fast-path in GHC.Tc.Utils.Instantiate.instCallConstraints See Note [Possible fast path for equality constraints] Many other small refactorings. - - - - - 1b1523b1 by George Thomas at 2024-06-12T12:52:18-04:00 Fix non-compiling extensible record `HasField` example - - - - - 97b141a3 by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Fix hyperlinker source urls (#24907) This fixes a bug introduced by f56838c36235febb224107fa62334ebfe9941aba Links to external modules in the hyperlinker are uniformly generated using splicing the template given to us instead of attempting to construct the url in an ad-hoc manner. - - - - - 954f864c by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Add name anchor to external source urls from documentation page URLs for external source links from documentation pages were missing a splice location for the name. Fixes #24912 - - - - - b0b64177 by Simon Peyton Jones at 2024-06-12T12:53:31-04:00 Prioritise nominal equalities The main payload of this patch is * Prioritise nominal equalities in the constraint solver. This ameliorates the incompleteness of solving for representational constraints over newtypes: see #24887. See (EX2) in Note [Decomposing newtype equalities] in GHC.Tc.Solver.Equality In doing this patch I tripped over some other things that I refactored: * Move `isCoVarType` from `GHC.Core.Type` to `GHC.Core.Predicate` where it seems more at home. * Clarify the "rewrite role" of a constraint. I was very puzzled about what the role of, say `(Eq a)` might be, but see the new Note [The rewrite-role of a constraint]. In doing so I made predTypeEqRel crash when given a non-equality. Usually it expects an equality; but it was being mis-used for the above rewrite-role stuff. - - - - - cb7c1b83 by Liam Goodacre at 2024-06-12T12:54:09-04:00 compiler: missing-deriving-strategies suggested fix Extends the missing-deriving-strategies warning with a suggested fix that includes which deriving strategies were assumed. For info about the warning, see comments for `TcRnNoDerivStratSpecified`, `TcRnNoDerivingClauseStrategySpecified`, & `TcRnNoStandaloneDerivingStrategySpecified`. For info about the suggested fix, see `SuggestExplicitDerivingClauseStrategies` & `SuggestExplicitStandalanoDerivingStrategy`. docs: Rewords missing-deriving-strategies to mention the suggested fix. Resolves #24955 - - - - - 4e36d3a3 by Jan Hrček at 2024-06-12T12:54:48-04:00 Further haddocks improvements in Language.Haskell.Syntax.Pat.Pat - - - - - 558353f4 by Cheng Shao at 2024-06-12T12:55:24-04:00 rts: use page sized mblocks on wasm This patch changes mblock size to page size on wasm. It allows us to simplify our wasi-libc fork, makes it much easier to test third party libc allocators like emmalloc/mimalloc, as well as experimenting with threaded RTS in wasm. - - - - - b3cc5366 by Matthew Pickering at 2024-06-12T23:06:57-04:00 compiler: Make ghc-experimental not wired in If you need to wire in definitions, then place them in ghc-internal and reexport them from ghc-experimental. Ticket #24903 - - - - - 700eeab9 by Hécate Kleidukos at 2024-06-12T23:07:37-04:00 base: Use a more appropriate unicode arrow for the ByteArray diagram This commit rectifies the usage of a unicode arrow in favour of one that doesn't provoke mis-alignment. - - - - - cca7de25 by Matthew Pickering at 2024-06-12T23:08:14-04:00 ghcup-metadata: Fix debian version ranges This was caught by `ghcup-ci` failing and attempting to install a deb12 bindist on deb11. ``` configure: WARNING: m4/prep_target_file.m4: Expecting YES/NO but got in ArSupportsDashL_STAGE0. Defaulting to False. bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by bin/ghc-toolchain-bin) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) ``` Fixes #24974 - - - - - 7b23ce8b by Pierre Le Marre at 2024-06-13T15:35:04-04:00 ucd2haskell: remove Streamly dependency + misc - Remove dead code. - Remove `streamly` dependency. - Process files with `bytestring`. - Replace Unicode files parsers with the corresponding ones from the package `unicode-data-parser`. - Simplify cabal file and rename module - Regenerate `ghc-internal` Unicode files with new header - - - - - 4570319f by Jacco Krijnen at 2024-06-13T15:35:41-04:00 Document how to run haddocks tests (#24976) Also remove ghc 9.7 requirement - - - - - fb629e24 by amesgen at 2024-06-14T00:28:20-04:00 compiler: refactor lower_CmmExpr_Ptr - - - - - def46c8c by amesgen at 2024-06-14T00:28:20-04:00 compiler: handle CmmRegOff in lower_CmmExpr_Ptr - - - - - ce76bf78 by Simon Peyton Jones at 2024-06-14T00:28:56-04:00 Small documentation update in Quick Look - - - - - 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 2661f565 by Simon Peyton Jones at 2024-09-10T14:29:17+01:00 Just a start on specialising expressions Addresses #24359. Just a start, does not compile. - - - - - 78f847e7 by Simon Peyton Jones at 2024-09-10T17:15:59+01:00 More progress (Still does not compile.) - - - - - 4a267ade by Simon Peyton Jones at 2024-09-10T17:21:28+01:00 More progress - - - - - 2764a723 by Simon Peyton Jones at 2024-09-10T17:21:28+01:00 Wibble - - - - - 924fc9ea by Simon Peyton Jones at 2024-09-10T17:23:35+01:00 More progress - - - - - e3d891fc by Simon Peyton Jones at 2024-09-10T17:23:35+01:00 More progress - - - - - 5935d858 by Simon Peyton Jones at 2024-09-10T17:25:29+01:00 Finally runnable! - - - - - 94a2b741 by Simon Peyton Jones at 2024-09-10T17:25:29+01:00 Progress - - - - - 07002d7c by Simon Peyton Jones at 2024-09-10T17:26:01+01:00 Working I think - - - - - 278bb732 by Andrei Borzenkov at 2024-09-10T17:27:27+01:00 Fix derivations conflict in parser, disambiguate them in post-process - - - - - 59508c3b by Alan Zimmerman at 2024-09-10T17:27:28+01:00 Fix exact printing for RuleBndrs This puts the exact print annotations inside a TTG extension point in RuleBndrs. It also adds an exact print case for SpecSigE - - - - - 82d7c720 by Simon Peyton Jones at 2024-09-10T17:27:28+01:00 Wibble imports - - - - - 596bdeca by Simon Peyton Jones at 2024-09-10T17:27:28+01:00 Typo in comments - - - - - cd859a9b by Simon Peyton Jones at 2024-09-10T17:28:40+01:00 Wibbles - - - - - b8d2b9a1 by Simon Peyton Jones at 2024-09-10T17:28:40+01:00 Go via new route for simple SPECIALISE pragmas - - - - - 11afff8d by Simon Peyton Jones at 2024-09-10T17:29:25+01:00 Further work - - - - - 72d379d8 by Simon Peyton Jones at 2024-09-10T17:38:37+01:00 Wibble - - - - - 23 changed files: - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - CODEOWNERS - compiler/GHC.hs - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/PrimOps.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/Utils.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/ByteCode/Types.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/559e86ce0b41c8e1b0b669824ba3b02e0c9c632e...72d379d8edb4fd3e44ba3ed67d0ab578be7093c4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/559e86ce0b41c8e1b0b669824ba3b02e0c9c632e...72d379d8edb4fd3e44ba3ed67d0ab578be7093c4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 17:00:27 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Tue, 10 Sep 2024 13:00:27 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] Add test for C calling convention Message-ID: <66e07b2bcad98_17411a8db58189f@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: eab0e738 by Sven Tennie at 2024-09-10T18:58:12+02:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - 4 changed files: - + testsuite/tests/codeGen/should_run/CCallConv.hs - + testsuite/tests/codeGen/should_run/CCallConv.stdout - + testsuite/tests/codeGen/should_run/CCallConv_c.c - testsuite/tests/codeGen/should_run/all.T Changes: ===================================== testsuite/tests/codeGen/should_run/CCallConv.hs ===================================== @@ -0,0 +1,132 @@ +{-# LANGUAGE ForeignFunctionInterface #-} +{-# LANGUAGE GHCForeignImportPrim #-} +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE UnboxedTuples #-} +{-# LANGUAGE UnliftedFFITypes #-} + +-- | This test ensures that sub-word signed and unsigned parameters are correctly +-- handed over to C functions. I.e. it asserts the calling-convention. +-- +-- The number of parameters is currently shaped for the RISCV64 calling-convention. +-- You may need to add more parameters to the C functions in case there are more +-- registers reserved for parameters in your architecture. +module Main where + +import Data.Word +import GHC.Exts +import GHC.Int +import System.IO + +foreign import ccall "fun8" + fun8 :: + Int8# -> -- a0 + Word8# -> -- a1 + Int8# -> -- a2 + Int8# -> -- a3 + Int8# -> -- a4 + Int8# -> -- a5 + Int8# -> -- a6 + Int8# -> -- a7 + Word8# -> -- s0 + Int8# -> -- s1 + Int64# -- result + +foreign import ccall "fun16" + fun16 :: + Int16# -> -- a0 + Word16# -> -- a1 + Int16# -> -- a2 + Int16# -> -- a3 + Int16# -> -- a4 + Int16# -> -- a5 + Int16# -> -- a6 + Int16# -> -- a7 + Word16# -> -- s0 + Int16# -> -- s1 + Int64# -- result + +foreign import ccall "fun32" + fun32 :: + Int32# -> -- a0 + Word32# -> -- a1 + Int32# -> -- a2 + Int32# -> -- a3 + Int32# -> -- a4 + Int32# -> -- a5 + Int32# -> -- a6 + Int32# -> -- a7 + Word32# -> -- s0 + Int32# -> -- s1 + Int64# -- result + +foreign import ccall "funFloat" + funFloat :: + Float# -> -- a0 + Float# -> -- a1 + Float# -> -- a2 + Float# -> -- a3 + Float# -> -- a4 + Float# -> -- a5 + Float# -> -- a6 + Float# -> -- a7 + Float# -> -- s0 + Float# -> -- s1 + Float# -- result + +foreign import ccall "funDouble" + funDouble :: + Double# -> -- a0 + Double# -> -- a1 + Double# -> -- a2 + Double# -> -- a3 + Double# -> -- a4 + Double# -> -- a5 + Double# -> -- a6 + Double# -> -- a7 + Double# -> -- s0 + Double# -> -- s1 + Double# -- result + +main :: IO () +main = do + -- N.B. the values here aren't choosen by accident: -1 means all bits one in + -- twos-complement, which is the same as the max word value. + let i8 :: Int8# = intToInt8# (-1#) + w8 :: Word8# = wordToWord8# (255##) + res8 :: Int64# = fun8 i8 w8 i8 i8 i8 i8 i8 i8 w8 i8 + expected_res8 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word8) + 8 * (-1) + print $ "fun8 result:" ++ show (I64# res8) + hFlush stdout + assertEqual expected_res8 (I64# res8) + + let i16 :: Int16# = intToInt16# (-1#) + w16 :: Word16# = wordToWord16# (65535##) + res16 :: Int64# = fun16 i16 w16 i16 i16 i16 i16 i16 i16 w16 i16 + expected_res16 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word16) + 8 * (-1) + print $ "fun16 result:" ++ show (I64# res16) + hFlush stdout + assertEqual expected_res16 (I64# res16) + + let i32 :: Int32# = intToInt32# (-1#) + w32 :: Word32# = wordToWord32# (4294967295##) + res32 :: Int64# = fun32 i32 w32 i32 i32 i32 i32 i32 i32 w32 i32 + expected_res32 :: Int64 = 2 * (fromInteger . fromIntegral) (maxBound :: Word32) + 8 * (-1) + print $ "fun32 result:" ++ show (I64# res32) + hFlush stdout + assertEqual expected_res32 (I64# res32) + + let resFloat :: Float = F# (funFloat 1.0# 1.1# 1.2# 1.3# 1.4# 1.5# 1.6# 1.7# 1.8# 1.9#) + print $ "funFloat result:" ++ show resFloat + hFlush stdout + assertEqual (14.5 :: Float) resFloat + + let resDouble :: Double = D# (funDouble 1.0## 1.1## 1.2## 1.3## 1.4## 1.5## 1.6## 1.7## 1.8## 1.9##) + print $ "funDouble result:" ++ show resDouble + hFlush stdout + assertEqual (14.5 :: Double) resDouble + +assertEqual :: (Eq a, Show a) => a -> a -> IO () +assertEqual a b = + if a == b + then pure () + else error $ show a ++ " =/= " ++ show b ===================================== testsuite/tests/codeGen/should_run/CCallConv.stdout ===================================== @@ -0,0 +1,60 @@ +fun8: +a0: 0xffffffff -1 +a1: 0xff 255 +a2: 0xffffffff -1 +a3: 0xffffffff -1 +a4: 0xffffffff -1 +a5: 0xffffffff -1 +a6: 0xffffffff -1 +a7: 0xffffffff -1 +s0: 0xffffffff -1 +s1: 0xff 255 +"fun8 result:502" +fun16: +a0: 0xffffffff -1 +a1: 0xffff 65535 +a2: 0xffffffff -1 +a3: 0xffffffff -1 +a4: 0xffffffff -1 +a5: 0xffffffff -1 +a6: 0xffffffff -1 +a7: 0xffffffff -1 +s0: 0xffffffff -1 +s1: 0xffff 65535 +"fun16 result:131062" +fun32: +a0: 0xffffffff -1 +a1: 0xffffffff 4294967295 +a2: 0xffffffff -1 +a3: 0xffffffff -1 +a4: 0xffffffff -1 +a5: 0xffffffff -1 +a6: 0xffffffff -1 +a7: 0xffffffff -1 +s0: 0xffffffff -1 +s1: 0xffffffff 4294967295 +"fun32 result:8589934582" +funFloat: +a0: 1.000000 +a1: 1.100000 +a2: 1.200000 +a3: 1.300000 +a4: 1.400000 +a5: 1.500000 +a6: 1.600000 +a7: 1.700000 +s0: 1.800000 +s1: 1.900000 +"funFloat result:14.5" +funDouble: +a0: 1.000000 +a1: 1.100000 +a2: 1.200000 +a3: 1.300000 +a4: 1.400000 +a5: 1.500000 +a6: 1.600000 +a7: 1.700000 +s0: 1.800000 +s1: 1.900000 +"funDouble result:14.5" ===================================== testsuite/tests/codeGen/should_run/CCallConv_c.c ===================================== @@ -0,0 +1,101 @@ +#include +#include + +int64_t fun8(int8_t a0, uint8_t a1, int8_t a2, int8_t a3, int8_t a4, int8_t a5, + int8_t a6, int8_t a7, int8_t s0, uint8_t s1) { + printf("fun8:\n"); + printf("a0: %#x %hhd\n", a0, a0); + printf("a1: %#x %hhu\n", a1, a1); + printf("a2: %#x %hhd\n", a2, a2); + printf("a3: %#x %hhd\n", a3, a3); + printf("a4: %#x %hhd\n", a4, a4); + printf("a5: %#x %hhd\n", a5, a5); + printf("a6: %#x %hhd\n", a6, a6); + printf("a7: %#x %hhd\n", a7, a7); + printf("s0: %#x %hhd\n", s0, s0); + printf("s1: %#x %hhu\n", s1, s1); + + fflush(stdout); + + return a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + s1; +} + +int64_t fun16(int16_t a0, uint16_t a1, int16_t a2, int16_t a3, int16_t a4, + int16_t a5, int16_t a6, int16_t a7, int16_t s0, uint16_t s1) { + printf("fun16:\n"); + printf("a0: %#x %hd\n", a0, a0); + printf("a1: %#x %hu\n", a1, a1); + printf("a2: %#x %hd\n", a2, a2); + printf("a3: %#x %hd\n", a3, a3); + printf("a4: %#x %hd\n", a4, a4); + printf("a5: %#x %hd\n", a5, a5); + printf("a6: %#x %hd\n", a6, a6); + printf("a7: %#x %hd\n", a7, a7); + printf("s0: %#x %hd\n", s0, s0); + printf("s1: %#x %hu\n", s1, s1); + + fflush(stdout); + + return a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + s1; +} + +int64_t fun32(int32_t a0, uint32_t a1, int32_t a2, int32_t a3, int32_t a4, + int32_t a5, int32_t a6, int32_t a7, int32_t s0, uint32_t s1) { + printf("fun32:\n"); + printf("a0: %#x %d\n", a0, a0); + printf("a1: %#x %u\n", a1, a1); + printf("a2: %#x %d\n", a2, a2); + printf("a3: %#x %d\n", a3, a3); + printf("a4: %#x %d\n", a4, a4); + printf("a5: %#x %d\n", a5, a5); + printf("a6: %#x %d\n", a6, a6); + printf("a7: %#x %d\n", a7, a7); + printf("s0: %#x %d\n", s0, s0); + printf("s1: %#x %u\n", s1, s1); + + fflush(stdout); + + // Ensure the addition happens in long int (not just int) precission. + // Otherwise, the result is truncated during the operation. + int64_t force_int64_precission = 0; + return force_int64_precission + a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + + s1; +} + +float funFloat(float a0, float a1, float a2, float a3, float a4, float a5, + float a6, float a7, float s0, float s1) { + printf("funFloat:\n"); + printf("a0: %f\n", a0); + printf("a1: %f\n", a1); + printf("a2: %f\n", a2); + printf("a3: %f\n", a3); + printf("a4: %f\n", a4); + printf("a5: %f\n", a5); + printf("a6: %f\n", a6); + printf("a7: %f\n", a7); + printf("s0: %f\n", s0); + printf("s1: %f\n", s1); + + fflush(stdout); + + return a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + s1; +} + +double funDouble(double a0, double a1, double a2, double a3, double a4, double a5, + double a6, double a7, double s0, double s1) { + printf("funDouble:\n"); + printf("a0: %f\n", a0); + printf("a1: %f\n", a1); + printf("a2: %f\n", a2); + printf("a3: %f\n", a3); + printf("a4: %f\n", a4); + printf("a5: %f\n", a5); + printf("a6: %f\n", a6); + printf("a7: %f\n", a7); + printf("s0: %f\n", s0); + printf("s1: %f\n", s1); + + fflush(stdout); + + return a0 + a1 + a2 + a3 + a4 + a5 + a6 + a7 + s0 + s1; +} ===================================== testsuite/tests/codeGen/should_run/all.T ===================================== @@ -250,3 +250,5 @@ test('CtzClz0', normal, compile_and_run, ['']) test('T23034', req_c, compile_and_run, ['-O2 T23034_c.c']) test('T24700', normal, compile_and_run, ['-O']) test('T24893', normal, compile_and_run, ['-O']) + +test('CCallConv', [req_c, when(arch('wasm32'), fragile(25249))], compile_and_run, ['CCallConv_c.c']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/eab0e738eb8f4a69b0299455e3336a579f4a7e98 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/eab0e738eb8f4a69b0299455e3336a579f4a7e98 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 18:02:45 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Tue, 10 Sep 2024 14:02:45 -0400 Subject: [Git][ghc/ghc][wip/az/epa-exactprint-sync] 14 commits: AArch64: Implement takeRegRegMoveInstr Message-ID: <66e089c55734d_29b94c8ad2c28660@gitlab.mail> Alan Zimmerman pushed to branch wip/az/epa-exactprint-sync at Glasgow Haskell Compiler / GHC Commits: 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 5b61758c by Alan Zimmerman at 2024-09-10T19:02:26+01:00 EPA: Sync ghc-exactprint to GHC - - - - - 30 changed files: - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/Core/InstEnv.hs - compiler/GHC/Core/Opt/CallerCC/Types.hs - compiler/GHC/Core/Opt/WorkWrap/Utils.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/StgToJS/Symbols.hs - compiler/GHC/StgToJS/Types.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Dict.hs - compiler/GHC/Tc/Solver/Equality.hs - compiler/GHC/Tc/Solver/Irred.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Solver/Solve.hs - compiler/GHC/Tc/TyCl/Instance.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Utils/Instantiate.hs - compiler/GHC/Tc/Utils/Unify.hs - libraries/ghc-heap/tests/stack_misc_closures_c.c The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4f06b32165362a950bfd9982dbacd761add86c59...5b61758c50eaf6f9c8b50124bbabfc345fcf1a1b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4f06b32165362a950bfd9982dbacd761add86c59...5b61758c50eaf6f9c8b50124bbabfc345fcf1a1b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 18:57:44 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Tue, 10 Sep 2024 14:57:44 -0400 Subject: [Git][ghc/ghc][wip/az/epa-exactprint-sync] EPA: Sync ghc-exactprint to GHC Message-ID: <66e096a89c452_29b94c2fba983469f@gitlab.mail> Alan Zimmerman pushed to branch wip/az/epa-exactprint-sync at Glasgow Haskell Compiler / GHC Commits: 42c95600 by Alan Zimmerman at 2024-09-10T19:57:01+01:00 EPA: Sync ghc-exactprint to GHC - - - - - 6 changed files: - utils/check-exact/ExactPrint.hs - utils/check-exact/Main.hs - utils/check-exact/Parsers.hs - utils/check-exact/Transform.hs - utils/check-exact/Types.hs - utils/check-exact/Utils.hs Changes: ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -1,3 +1,4 @@ +{-# LANGUAGE BangPatterns #-} {-# LANGUAGE DeriveDataTypeable #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE FlexibleInstances #-} @@ -25,7 +26,7 @@ module ExactPrint , makeDeltaAst -- * Configuration - , EPOptions(epRigidity, epAstPrint, epTokenPrint, epWhitespacePrint, epUpdateAnchors) + , EPOptions(epTokenPrint, epWhitespacePrint) , stringOptions , epOptions , deltaOptions @@ -43,10 +44,11 @@ import GHC.Types.ForeignCall import GHC.Types.Name.Reader import GHC.Types.PkgQual import GHC.Types.SourceText +import GHC.Types.SrcLoc import GHC.Types.Var -import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Unit.Module.Warnings import GHC.Utils.Misc +import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Utils.Panic import Language.Haskell.Syntax.Basic (FieldLabelString(..)) @@ -77,8 +79,7 @@ import Types exactPrint :: ExactPrint ast => ast -> String exactPrint ast = snd $ runIdentity (runEP stringOptions (markAnnotated ast)) --- | The additional option to specify the rigidity and printing --- configuration. +-- | The additional option to specify the printing configuration. exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) => EPOptions m b -> ast @@ -86,9 +87,8 @@ exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) exactPrintWithOptions r ast = runEP r (markAnnotated ast) --- | Transform concrete annotations into relative annotations which --- are more useful when transforming an AST. This corresponds to the --- earlier 'relativiseApiAnns'. +-- | Transform concrete annotations into relative annotations. +-- This should be unnecessary from GHC 9.10 makeDeltaAst :: ExactPrint ast => ast -> ast makeDeltaAst ast = fst $ runIdentity (runEP deltaOptions (markAnnotated ast)) @@ -115,6 +115,7 @@ defaultEPState = EPState , dPriorEndPosition = (1,1) , uAnchorSpan = badRealSrcSpan , uExtraDP = Nothing + , uExtraDPReturn = Nothing , pAcceptSpan = False , epComments = [] , epCommentsApplied = [] @@ -128,39 +129,27 @@ defaultEPState = EPState -- | The R part of RWS. The environment. Updated via 'local' as we -- enter a new AST element, having a different anchor point. data EPOptions m a = EPOptions - { - epAstPrint :: forall ast . Data ast => GHC.Located ast -> a -> m a - , epTokenPrint :: String -> m a + { epTokenPrint :: String -> m a , epWhitespacePrint :: String -> m a - , epRigidity :: Rigidity - , epUpdateAnchors :: Bool } -- | Helper to create a 'EPOptions' -epOptions :: - (forall ast . Data ast => GHC.Located ast -> a -> m a) - -> (String -> m a) - -> (String -> m a) - -> Rigidity - -> Bool - -> EPOptions m a -epOptions astPrint tokenPrint wsPrint rigidity delta = EPOptions - { - epAstPrint = astPrint - , epWhitespacePrint = wsPrint +epOptions :: (String -> m a) + -> (String -> m a) + -> EPOptions m a +epOptions tokenPrint wsPrint = EPOptions + { epWhitespacePrint = wsPrint , epTokenPrint = tokenPrint - , epRigidity = rigidity - , epUpdateAnchors = delta } -- | Options which can be used to print as a normal String. stringOptions :: EPOptions Identity String -stringOptions = epOptions (\_ b -> return b) return return NormalLayout False +stringOptions = epOptions return return -- | Options which can be used to simply update the AST to be in delta -- form, without generating output deltaOptions :: EPOptions Identity () -deltaOptions = epOptions (\_ _ -> return ()) (\_ -> return ()) (\_ -> return ()) NormalLayout True +deltaOptions = epOptions (\_ -> return ()) (\_ -> return ()) data EPWriter a = EPWriter { output :: !a } @@ -177,6 +166,8 @@ data EPState = EPState -- Annotation , uExtraDP :: !(Maybe Anchor) -- ^ Used to anchor a -- list + , uExtraDPReturn :: !(Maybe DeltaPos) + -- ^ Used to return Delta version of uExtraDP , pAcceptSpan :: Bool -- ^ When we have processed an -- entry of EpaDelta, accept the -- next `EpaSpan` start as the @@ -213,7 +204,7 @@ class HasTrailing a where trailing :: a -> [TrailingAnn] setTrailing :: a -> [TrailingAnn] -> a -setAnchorEpa :: (HasTrailing an, NoAnn an) +setAnchorEpa :: (HasTrailing an) => EpAnn an -> Anchor -> [TrailingAnn] -> EpAnnComments -> EpAnn an setAnchorEpa (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing an ts) cs @@ -223,7 +214,7 @@ setAnchorHsModule hsmod anc cs = hsmod { hsmodExt = (hsmodExt hsmod) {hsmodAnn = anc' = anc an' = setAnchorEpa (hsmodAnn $ hsmodExt hsmod) anc' [] cs -setAnchorAn :: (HasTrailing an, NoAnn an) +setAnchorAn :: (HasTrailing an) => LocatedAn an a -> Anchor -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a setAnchorAn (L (EpAnn _ an _) a) anc ts cs = (L (EpAnn anc (setTrailing an ts) cs) a) -- `debug` ("setAnchorAn: anc=" ++ showAst anc) @@ -248,7 +239,7 @@ data FlushComments = FlushComments data CanUpdateAnchor = CanUpdateAnchor | CanUpdateAnchorOnly | NoCanUpdateAnchor - deriving (Eq, Show) + deriving (Eq, Show, Data) data Entry = Entry Anchor [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor | NoEntryVal @@ -402,7 +393,7 @@ enterAnn NoEntryVal a = do r <- exact a debugM $ "enterAnn:done:NO ANN:p =" ++ show (p, astId a) return r -enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do +enterAnn !(Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do acceptSpan <- getAcceptSpan setAcceptSpan False case anchor' of @@ -421,9 +412,11 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do _ -> return () case anchor' of EpaDelta _ _ dcs -> do - debugM $ "enterAnn:Printing comments:" ++ showGhc (priorComments cs) + debugM $ "enterAnn:Delta:Flushing comments" + flushComments [] + debugM $ "enterAnn:Delta:Printing prior comments:" ++ showGhc (priorComments cs) mapM_ printOneComment (concatMap tokComment $ priorComments cs) - debugM $ "enterAnn:Printing EpaDelta comments:" ++ showGhc dcs + debugM $ "enterAnn:Delta:Printing EpaDelta comments:" ++ showGhc dcs mapM_ printOneComment (concatMap tokComment dcs) _ -> do debugM $ "enterAnn:Adding comments:" ++ showGhc (priorComments cs) @@ -465,7 +458,7 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- The first part corresponds to the delta phase, so should only use -- delta phase variables ----------------------------------- -- Calculate offset required to get to the start of the SrcSPan - off <- getLayoutOffsetD + !off <- getLayoutOffsetD let spanStart = ss2pos curAnchor priorEndAfterComments <- getPriorEndD let edp' = adjustDeltaForOffset @@ -480,17 +473,18 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- --------------------------------------------- med <- getExtraDP setExtraDP Nothing - let edp = case med of - Nothing -> edp'' - Just (EpaDelta _ dp _) -> dp + let (edp, medr) = case med of + Nothing -> (edp'', Nothing) + Just (EpaDelta _ dp _) -> (dp, Nothing) -- Replace original with desired one. Allows all -- list entry values to be DP (1,0) - Just (EpaSpan (RealSrcSpan r _)) -> dp + Just (EpaSpan (RealSrcSpan r _)) -> (dp, Just dp) where dp = adjustDeltaForOffset off (ss2delta priorEndAfterComments r) Just (EpaSpan (UnhelpfulSpan r)) -> panic $ "enterAnn: UnhelpfulSpan:" ++ show r when (isJust med) $ debugM $ "enterAnn:(med,edp)=" ++ showAst (med,edp) + when (isJust medr) $ setExtraDPReturn medr -- --------------------------------------------- -- Preparation complete, perform the action when (priorEndAfterComments < spanStart) (do @@ -511,12 +505,15 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do debugM $ "enterAnn:exact a starting:" ++ show (showAst anchor') a' <- exact a debugM $ "enterAnn:exact a done:" ++ show (showAst anchor') + + -- Core recursive exactprint done, start end of Entry processing + when (flush == FlushComments) $ do - debugM $ "flushing comments in enterAnn:" ++ showAst cs + debugM $ "flushing comments in enterAnn:" ++ showAst (cs, getFollowingComments cs) flushComments (getFollowingComments cs) debugM $ "flushing comments in enterAnn done" - eof <- getEofPos + !eof <- getEofPos case eof of Nothing -> return () Just (pos, prior) -> do @@ -544,28 +541,50 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- Outside the anchor, mark any trailing postCs <- cua canUpdateAnchor takeAppliedCommentsPop - when (flush == NoFlushComments) $ do - when ((getFollowingComments cs) /= []) $ do - - -- debugM $ "enterAnn:in:(anchor') =" ++ show (eloc2str anchor') - debugM $ "starting trailing comments:" ++ showAst (getFollowingComments cs) - mapM_ printOneComment (concatMap tokComment $ getFollowingComments cs) - debugM $ "ending trailing comments" - trailing' <- markTrailing trailing_anns + following <- if (flush == NoFlushComments) + then do + let (before, after) = splitAfterTrailingAnns trailing_anns + (getFollowingComments cs) + addCommentsA before + return after + else return [] + !trailing' <- markTrailing trailing_anns + -- mapM_ printOneComment (concatMap tokComment $ following) + addCommentsA following -- Update original anchor, comments based on the printing process -- TODO:AZ: probably need to put something appropriate in instead of noSrcSpan - let newAchor = EpaDelta noSrcSpan edp [] + let newAnchor = EpaDelta noSrcSpan edp [] let r = case canUpdateAnchor of - CanUpdateAnchor -> setAnnotationAnchor a' newAchor trailing' (mkEpaComments (priorCs ++ postCs) []) - CanUpdateAnchorOnly -> setAnnotationAnchor a' newAchor [] emptyComments + CanUpdateAnchor -> setAnnotationAnchor a' newAnchor trailing' (mkEpaComments priorCs postCs) + CanUpdateAnchorOnly -> setAnnotationAnchor a' newAnchor [] emptyComments NoCanUpdateAnchor -> a' return r -- --------------------------------------------------------------------- +-- | Split the span following comments into ones that occur prior to +-- the last trailing ann, and ones after. +splitAfterTrailingAnns :: [TrailingAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitAfterTrailingAnns [] cs = ([], cs) +splitAfterTrailingAnns tas cs = (before, after) + where + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + (before, after) = case reverse (concatMap trailing_loc tas) of + [] -> ([],cs) + (s:_) -> (b,a) + where + s_pos = ss2pos s + (b,a) = break (\(L ll _) -> (ss2pos $ anchor ll) > s_pos) + cs + + +-- --------------------------------------------------------------------- + addCommentsA :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -addCommentsA csNew = addComments (concatMap tokComment csNew) +addCommentsA csNew = addComments False (concatMap tokComment csNew) {- TODO: When we addComments, some may have an anchor that is no longer @@ -583,24 +602,36 @@ By definition it is the current anchor, so work against that. And that also means that the first entry comment that has moved should not have a line offset. -} -addComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -addComments csNew = do - -- debugM $ "addComments:" ++ show csNew +addComments :: (Monad m, Monoid w) => Bool -> [Comment] -> EP w m () +addComments sortNeeded csNew = do + debugM $ "addComments:csNew" ++ show csNew cs <- getUnallocatedComments + debugM $ "addComments:cs" ++ show cs + -- We can only sort the comments if we are in the first phase, + -- where all comments have locations. If any have EpaDelta the + -- sort will fail, so we do not try. + if sortNeeded && all noDelta (csNew ++ cs) + then putUnallocatedComments (sort (cs ++ csNew)) + else putUnallocatedComments (cs ++ csNew) - putUnallocatedComments (sort (cs ++ csNew)) +noDelta :: Comment -> Bool +noDelta c = case commentLoc c of + EpaSpan _ -> True + _ -> False -- --------------------------------------------------------------------- -- | Just before we print out the EOF comments, flush the remaining -- ones in the state. flushComments :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -flushComments trailing_anns = do +flushComments !trailing_anns = do + debugM $ "flushComments entered: " ++ showAst trailing_anns addCommentsA trailing_anns + debugM $ "flushComments after addCommentsA" cs <- getUnallocatedComments - debugM $ "flushing comments starting" - -- AZ:TODO: is the sort still needed? - mapM_ printOneComment (sortComments cs) + debugM $ "flushComments: got cs" + debugM $ "flushing comments starting: cs" ++ showAst cs + mapM_ printOneComment cs putUnallocatedComments [] debugM $ "flushing comments done" @@ -612,7 +643,7 @@ annotationsToComments :: (Monad m, Monoid w) => a -> Lens a [AddEpAnn] -> [AnnKeywordId] -> EP w m a annotationsToComments a l kws = do let (newComments, newAnns) = go ([],[]) (view l a) - addComments newComments + addComments True newComments return (set l (reverse newAnns) a) where keywords = Set.fromList kws @@ -654,14 +685,11 @@ printSourceText (NoSourceText) txt = printStringAdvance txt >> return () printSourceText (SourceText txt) _ = printStringAdvance (unpackFS txt) >> return () printSourceTextAA :: (Monad m, Monoid w) => SourceText -> String -> EP w m () -printSourceTextAA (NoSourceText) txt = printStringAtAA noAnn txt >> return () -printSourceTextAA (SourceText txt) _ = printStringAtAA noAnn (unpackFS txt) >> return () +printSourceTextAA (NoSourceText) txt = printStringAdvanceA txt >> return () +printSourceTextAA (SourceText txt) _ = printStringAdvanceA (unpackFS txt) >> return () -- --------------------------------------------------------------------- -printStringAtSs :: (Monad m, Monoid w) => SrcSpan -> String -> EP w m () -printStringAtSs ss str = printStringAtRs (realSrcSpan ss) str >> return () - printStringAtRs :: (Monad m, Monoid w) => RealSrcSpan -> String -> EP w m EpaLocation printStringAtRs pa str = printStringAtRsC CaptureComments pa str @@ -676,7 +704,7 @@ printStringAtRsC capture pa str = do p' <- adjustDeltaForOffsetM p debugM $ "printStringAtRsC:(p,p')=" ++ show (p,p') printStringAtLsDelta p' str - setPriorEndASTD True pa + setPriorEndASTD pa cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -709,6 +737,9 @@ printStringAtMLocL (EpAnn anc an cs) l s = do printStringAtLsDelta (SameLine 1) str return (Just (EpaDelta noSrcSpan (SameLine 1) [])) +printStringAdvanceA :: (Monad m, Monoid w) => String -> EP w m () +printStringAdvanceA str = printStringAtAA (EpaDelta noSrcSpan (SameLine 0) []) str >> return () + printStringAtAA :: (Monad m, Monoid w) => EpaLocation -> String -> EP w m EpaLocation printStringAtAA el str = printStringAtAAC CaptureComments el str @@ -735,7 +766,7 @@ printStringAtAAC capture (EpaDelta ss d cs) s = do p2 <- getPosP pe2 <- getPriorEndD debugM $ "printStringAtAA:(pe1,pe2,p1,p2)=" ++ show (pe1,pe2,p1,p2) - setPriorEndASTPD True (pe1,pe2) + setPriorEndASTPD (pe1,pe2) cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -883,8 +914,7 @@ markAnnOpenP' :: (Monad m, Monoid w) => AnnPragma -> SourceText -> String -> EP markAnnOpenP' an NoSourceText txt = markEpAnnLMS0 an lapr_open AnnOpen (Just txt) markAnnOpenP' an (SourceText txt) _ = markEpAnnLMS0 an lapr_open AnnOpen (Just $ unpackFS txt) -markAnnOpen :: (Monad m, Monoid w) - => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] +markAnnOpen :: (Monad m, Monoid w) => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] markAnnOpen an NoSourceText txt = markEpAnnLMS'' an lidl AnnOpen (Just txt) markAnnOpen an (SourceText txt) _ = markEpAnnLMS'' an lidl AnnOpen (Just $ unpackFS txt) @@ -1589,7 +1619,7 @@ markTopLevelList ls = mapM (\a -> setLayoutTopLevelP $ markAnnotated a) ls instance (ExactPrint a) => ExactPrint (Located a) where getAnnotationEntry (L l _) = case l of UnhelpfulSpan _ -> NoEntryVal - _ -> Entry (hackSrcSpanToAnchor l) [] emptyComments NoFlushComments CanUpdateAnchorOnly + _ -> Entry (EpaSpan l) [] emptyComments NoFlushComments CanUpdateAnchorOnly setAnnotationAnchor (L l a) _anc _ts _cs = L l a @@ -1664,16 +1694,10 @@ instance ExactPrint (HsModule GhcPs) where _ -> return lo am_decls' <- markTrailing (am_decls $ anns an0) - imports' <- markTopLevelList imports - - case lo of - EpExplicitBraces _ _ -> return () - _ -> do - -- Get rid of the balance of the preceding comments before starting on the decls - flushComments [] - putUnallocatedComments [] - decls' <- markTopLevelList (filter removeDocDecl decls) + mid <- markAnnotated (HsModuleImpDecls (am_cs $ anns an0) imports decls) + let imports' = id_imps mid + let decls' = id_decls mid lo1 <- case lo0 of EpExplicitBraces open close -> do @@ -1688,15 +1712,32 @@ instance ExactPrint (HsModule GhcPs) where debugM $ "am_eof:" ++ showGhc (pos, prior) setEofPos (Just (pos, prior)) - let anf = an0 { anns = (anns an0) { am_decls = am_decls' }} + let anf = an0 { anns = (anns an0) { am_decls = am_decls', am_cs = [] }} debugM $ "HsModule, anf=" ++ showAst anf return (HsModule (XModulePs anf lo1 mdeprec' mbDoc') mmn' mexports' imports' decls') +-- --------------------------------------------------------------------- + +-- | This is used to ensure the comments are updated into the right +-- place for makeDeltaAst. +data HsModuleImpDecls + = HsModuleImpDecls { + id_cs :: [LEpaComment], + id_imps :: [LImportDecl GhcPs], + id_decls :: [LHsDecl GhcPs] + } deriving Data + +instance ExactPrint HsModuleImpDecls where + -- Use an UnhelpfulSpan for the anchor, we are only interested in the comments + getAnnotationEntry mid = mkEntry (EpaSpan (UnhelpfulSpan UnhelpfulNoLocationInfo)) [] (EpaComments (id_cs mid)) + setAnnotationAnchor mid _anc _ cs = mid { id_cs = priorComments cs ++ getFollowingComments cs } + `debug` ("HsModuleImpDecls.setAnnotationAnchor:cs=" ++ showAst cs) + exact (HsModuleImpDecls cs imports decls) = do + imports' <- markTopLevelList imports + decls' <- markTopLevelList (filter notDocDecl decls) + return (HsModuleImpDecls cs imports' decls') -removeDocDecl :: LHsDecl GhcPs -> Bool -removeDocDecl (L _ DocD{}) = False -removeDocDecl _ = True -- --------------------------------------------------------------------- @@ -1737,8 +1778,8 @@ instance ExactPrint InWarningCategory where exact (InWarningCategory tkIn source (L l wc)) = do tkIn' <- markEpToken tkIn - L _ (_,wc') <- markAnnotated (L l (source, wc)) - return (InWarningCategory tkIn' source (L l wc')) + L l' (_,wc') <- markAnnotated (L l (source, wc)) + return (InWarningCategory tkIn' source (L l' wc')) instance ExactPrint (SourceText, WarningCategory) where getAnnotationEntry _ = NoEntryVal @@ -1943,14 +1984,14 @@ exactDataFamInstDecl an top_lvl , feqn_pats = pats , feqn_fixity = fixity , feqn_rhs = defn })) = do - (an', an2', tycon', bndrs', _, _mc, defn') <- exactDataDefn an2 pp_hdr defn - -- See Note [an and an2 in exactDataFamInstDecl] + (an', an2', tycon', bndrs', pats', defn') <- exactDataDefn an2 pp_hdr defn + -- See Note [an and an2 in exactDataFamInstDecl] return (an', DataFamInstDecl ( FamEqn { feqn_ext = an2' , feqn_tycon = tycon' , feqn_bndrs = bndrs' - , feqn_pats = pats + , feqn_pats = pats' , feqn_fixity = fixity , feqn_rhs = defn' })) `debug` ("exactDataFamInstDecl: defn' derivs:" ++ showAst (dd_derivs defn')) @@ -2233,11 +2274,11 @@ instance ExactPrint (RoleAnnotDecl GhcPs) where an1 <- markEpAnnL an0 lidl AnnRole ltycon' <- markAnnotated ltycon let markRole (L l (Just r)) = do - (L _ r') <- markAnnotated (L l r) - return (L l (Just r')) + (L l' r') <- markAnnotated (L l r) + return (L l' (Just r')) markRole (L l Nothing) = do - printStringAtSs (locA l) "_" - return (L l Nothing) + e' <- printStringAtAA (entry l) "_" + return (L (l { entry = e'}) Nothing) roles' <- mapM markRole roles return (RoleAnnotDecl an1 ltycon' roles') @@ -2340,8 +2381,13 @@ instance (ExactPrint tm, ExactPrint ty, Outputable tm, Outputable ty) getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact a@(HsValArg _ tm) = markAnnotated tm >> return a - exact a@(HsTypeArg at ty) = markEpToken at >> markAnnotated ty >> return a + exact (HsValArg x tm) = do + tm' <- markAnnotated tm + return (HsValArg x tm') + exact (HsTypeArg at ty) = do + at' <- markEpToken at + ty' <- markAnnotated ty + return (HsTypeArg at' ty') exact x@(HsArgPar _sp) = withPpr x -- Does not appear in original source -- --------------------------------------------------------------------- @@ -2359,9 +2405,9 @@ instance ExactPrint (ClsInstDecl GhcPs) where (mbWarn', an0, mbOverlap', inst_ty') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lid AnnSemi - ds <- withSortKey sortKey - [(ClsAtdTag, prepareListAnnotationA ats), - (ClsAtdTag, prepareListAnnotationF an adts), + (sortKey', ds) <- withSortKey sortKey + [(ClsAtTag, prepareListAnnotationA ats), + (ClsAtdTag, prepareListAnnotationF adts), (ClsMethodTag, prepareListAnnotationA binds), (ClsSigTag, prepareListAnnotationA sigs) ] @@ -2371,7 +2417,7 @@ instance ExactPrint (ClsInstDecl GhcPs) where adts' = undynamic ds binds' = undynamic ds sigs' = undynamic ds - return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey) + return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey') , cid_poly_ty = inst_ty', cid_binds = binds' , cid_sigs = sigs', cid_tyfam_insts = ats' , cid_overlap_mode = mbOverlap' @@ -2452,15 +2498,29 @@ instance ExactPrint (HsBind GhcPs) where return (FunBind x fun_id' matches') exact (PatBind x pat q grhss) = do + q' <- markAnnotated q pat' <- markAnnotated pat grhss' <- markAnnotated grhss - return (PatBind x pat' q grhss') + return (PatBind x pat' q' grhss') exact (PatSynBind x bind) = do bind' <- markAnnotated bind return (PatSynBind x bind') exact x = error $ "HsBind: exact for " ++ showAst x +instance ExactPrint (HsMultAnn GhcPs) where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact (HsNoMultAnn x) = return (HsNoMultAnn x) + exact (HsPct1Ann tok) = do + tok' <- markEpToken tok + return (HsPct1Ann tok') + exact (HsMultAnn tok ty) = do + tok' <- markEpToken tok + ty' <- markAnnotated ty + return (HsMultAnn tok' ty') + -- --------------------------------------------------------------------- instance ExactPrint (PatSynBind GhcPs GhcPs) where @@ -2519,8 +2579,9 @@ instance ExactPrint (PatSynBind GhcPs GhcPs) where instance ExactPrint (RecordPatSynField GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact r@(RecordPatSynField { recordPatSynField = v }) = markAnnotated v - >> return r + exact (RecordPatSynField f v) = do + f' <- markAnnotated f + return (RecordPatSynField f' v) -- --------------------------------------------------------------------- @@ -2648,15 +2709,20 @@ instance ExactPrint (HsLocalBinds GhcPs) where (an1, valbinds') <- markAnnList an0 $ markAnnotatedWithLayout valbinds debugM $ "exact HsValBinds: an1=" ++ showAst an1 - return (HsValBinds an1 valbinds') + medr <- getExtraDPReturn + an2 <- case medr of + Nothing -> return an1 + Just dp -> do + setExtraDPReturn Nothing + return $ an1 { anns = (anns an1) { al_anchor = Just (EpaDelta noSrcSpan dp []) }} + return (HsValBinds an2 valbinds') exact (HsIPBinds an bs) = do - (as, ipb) <- markAnnList an (markEpAnnL' an lal_rest AnnWhere - >> markAnnotated bs - >>= \bs' -> return (HsIPBinds an bs'::HsLocalBinds GhcPs)) - case ipb of - HsIPBinds _ bs' -> return (HsIPBinds as bs'::HsLocalBinds GhcPs) - _ -> error "should not happen HsIPBinds" + (an2,bs') <- markAnnListA an $ \an0 -> do + an1 <- markEpAnnL' an0 lal_rest AnnWhere + bs' <- markAnnotated bs + return (an1, bs') + return (HsIPBinds an2 bs') exact b@(EmptyLocalBinds _) = return b @@ -2670,7 +2736,8 @@ instance ExactPrint (HsValBindsLR GhcPs GhcPs) where let binds' = concatMap decl2Bind decls sigs' = concatMap decl2Sig decls - return (ValBinds sortKey binds' sigs') + sortKey' = captureOrderBinds decls + return (ValBinds sortKey' binds' sigs') exact (XValBindsLR _) = panic "XValBindsLR" undynamic :: Typeable a => [Dynamic] -> [a] @@ -2682,7 +2749,9 @@ instance ExactPrint (HsIPBinds GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact b@(IPBinds _ binds) = setLayoutBoth $ markAnnotated binds >> return b + exact (IPBinds x binds) = setLayoutBoth $ do + binds' <- markAnnotated binds + return (IPBinds x binds') -- --------------------------------------------------------------------- @@ -2703,18 +2772,18 @@ instance ExactPrint HsIPName where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact i@(HsIPName fs) = printStringAdvance ("?" ++ (unpackFS fs)) >> return i + exact i@(HsIPName fs) = printStringAdvanceA ("?" ++ (unpackFS fs)) >> return i -- --------------------------------------------------------------------- -- Managing lists which have been separated, e.g. Sigs and Binds prepareListAnnotationF :: (Monad m, Monoid w) => - [AddEpAnn] -> [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] -prepareListAnnotationF an ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls + [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] +prepareListAnnotationF ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls where go (L l a) = do - d' <- markAnnotated (DataFamInstDeclWithContext an NotTopLevel a) - return (toDyn (L l (dc_d d'))) + (L l' d') <- markAnnotated (L l (DataFamInstDeclWithContext noAnn NotTopLevel a)) + return (toDyn (L l' (dc_d d'))) prepareListAnnotationA :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => [LocatedAn an a] -> [(RealSrcSpan,EP w m Dynamic)] @@ -2725,15 +2794,23 @@ prepareListAnnotationA ls = map (\b -> (realSrcSpan $ getLocA b,go b)) ls return (toDyn b') withSortKey :: (Monad m, Monoid w) - => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] -> EP w m [Dynamic] + => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] + -> EP w m (AnnSortKey DeclTag, [Dynamic]) withSortKey annSortKey xs = do debugM $ "withSortKey:annSortKey=" ++ showAst annSortKey - let ordered = case annSortKey of - NoAnnSortKey -> sortBy orderByFst $ concatMap snd xs - AnnSortKey _keys -> orderedDecls annSortKey (Map.fromList xs) - mapM snd ordered -orderByFst :: Ord a => (a, b1) -> (a, b2) -> Ordering -orderByFst (a,_) (b,_) = compare a b + let (sk, ordered) = case annSortKey of + NoAnnSortKey -> (annSortKey', map snd os) + where + doOne (tag, ds) = map (\d -> (tag, d)) ds + xsExpanded = concatMap doOne xs + os = sortBy orderByFst $ xsExpanded + annSortKey' = AnnSortKey (map fst os) + AnnSortKey _keys -> (annSortKey, orderedDecls annSortKey (Map.fromList xs)) + ordered' <- mapM snd ordered + return (sk, ordered') + +orderByFst :: Ord a => (t, (a,b1)) -> (t, (a, b2)) -> Ordering +orderByFst (_,(a,_)) (_,(b,_)) = compare a b -- --------------------------------------------------------------------- @@ -2761,15 +2838,16 @@ instance ExactPrint (Sig GhcPs) where (an0, vars',ty') <- exactVarSig an vars ty return (ClassOpSig an0 is_deflt vars' ty') - exact (FixSig (an,src) (FixitySig x names (Fixity v fdir))) = do + exact (FixSig (an,src) (FixitySig ns names (Fixity v fdir))) = do let fixstr = case fdir of InfixL -> "infixl" InfixR -> "infixr" InfixN -> "infix" an0 <- markEpAnnLMS'' an lidl AnnInfix (Just fixstr) an1 <- markEpAnnLMS'' an0 lidl AnnVal (Just (sourceTextToString src (show v))) + ns' <- markAnnotated ns names' <- markAnnotated names - return (FixSig (an1,src) (FixitySig x names' (Fixity v fdir))) + return (FixSig (an1,src) (FixitySig ns' names' (Fixity v fdir))) exact (InlineSig an ln inl) = do an0 <- markAnnOpen an (inl_src inl) "{-# INLINE" @@ -2809,7 +2887,7 @@ instance ExactPrint (Sig GhcPs) where exact (CompleteMatchSig (an,src) cs mty) = do an0 <- markAnnOpen an src "{-# COMPLETE" - cs' <- markAnnotated cs + cs' <- mapM markAnnotated cs (an1, mty') <- case mty of Nothing -> return (an0, mty) @@ -2822,6 +2900,20 @@ instance ExactPrint (Sig GhcPs) where -- --------------------------------------------------------------------- +instance ExactPrint NamespaceSpecifier where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact NoNamespaceSpecifier = return NoNamespaceSpecifier + exact (TypeNamespaceSpecifier typeTok) = do + typeTok' <- markEpToken typeTok + return (TypeNamespaceSpecifier typeTok') + exact (DataNamespaceSpecifier dataTok) = do + dataTok' <- markEpToken dataTok + return (DataNamespaceSpecifier dataTok') + +-- --------------------------------------------------------------------- + exactVarSig :: (Monad m, Monoid w, ExactPrint a) => AnnSig -> [LocatedN RdrName] -> a -> EP w m (AnnSig, [LocatedN RdrName], a) exactVarSig an vars ty = do @@ -2875,7 +2967,7 @@ instance ExactPrint (AnnDecl GhcPs) where n' <- markAnnotated n return (an1, TypeAnnProvenance n') ModuleAnnProvenance -> do - an1 <- markEpAnnL an lapr_rest AnnModule + an1 <- markEpAnnL an0 lapr_rest AnnModule return (an1, prov) e' <- markAnnotated e @@ -2950,21 +3042,21 @@ instance ExactPrint (HsExpr GhcPs) where then markAnnotated n else return n return (HsVar x n') - exact x@(HsUnboundVar an _) = do + exact (HsUnboundVar an n) = do case an of Just (EpAnnUnboundVar (ob,cb) l) -> do - printStringAtAA ob "`" >> return () - printStringAtAA l "_" >> return () - printStringAtAA cb "`" >> return () - return x + ob' <- printStringAtAA ob "`" + l' <- printStringAtAA l "_" + cb' <- printStringAtAA cb "`" + return (HsUnboundVar (Just (EpAnnUnboundVar (ob',cb') l')) n) _ -> do - printStringAtLsDelta (SameLine 0) "_" - return x + printStringAdvanceA "_" >> return () + return (HsUnboundVar an n) exact x@(HsOverLabel src l) = do - printStringAtLsDelta (SameLine 0) "#" + printStringAdvanceA "#" >> return () case src of - NoSourceText -> printStringAtLsDelta (SameLine 0) (unpackFS l) - SourceText txt -> printStringAtLsDelta (SameLine 0) (unpackFS txt) + NoSourceText -> printStringAdvanceA (unpackFS l) >> return () + SourceText txt -> printStringAdvanceA (unpackFS txt) >> return () return x exact x@(HsIPVar _ (HsIPName n)) @@ -3204,11 +3296,11 @@ instance ExactPrint (HsExpr GhcPs) where exact (HsTypedSplice an s) = do an0 <- markEpAnnL an lidl AnnDollarDollar - s' <- exact s + s' <- markAnnotated s return (HsTypedSplice an0 s') exact (HsUntypedSplice an s) = do - s' <- exact s + s' <- markAnnotated s return (HsUntypedSplice an s') exact (HsProc an p c) = do @@ -3274,12 +3366,15 @@ exactMdo an (Just module_name) kw = markEpAnnLMS'' an lal_rest kw (Just n) markMaybeDodgyStmts :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => AnnList -> LocatedAn an a -> EP w m (AnnList, LocatedAn an a) markMaybeDodgyStmts an stmts = - if isGoodSrcSpan (getLocA stmts) + if notDodgy stmts then do r <- markAnnotatedWithLayout stmts return (an, r) else return (an, stmts) +notDodgy :: GenLocated (EpAnn ann) a -> Bool +notDodgy (L (EpAnn anc _ _) _) = notDodgyE anc + notDodgyE :: EpaLocation -> Bool notDodgyE anc = case anc of @@ -3341,7 +3436,7 @@ instance ExactPrint (MatchGroup GhcPs (LocatedA (HsCmd GhcPs))) where setAnnotationAnchor a _ _ _ = a exact (MG x matches) = do -- TODO:AZ use SortKey, in MG ann. - matches' <- if isGoodSrcSpan (getLocA matches) + matches' <- if notDodgy matches then markAnnotated matches else return matches return (MG x matches') @@ -3661,6 +3756,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- There may be arbitrary parens around parts of the constructor -- that are infix. Turn these into comments so that they feed -- into the right place automatically + -- TODO: no longer sorting on insert. What now? an0 <- annotationsToComments an lidl [AnnOpenP,AnnCloseP] an1 <- markEpAnnL an0 lidl AnnType @@ -3674,7 +3770,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- TODO: add a workaround for https://gitlab.haskell.org/ghc/ghc/-/issues/20452 exact (DataDecl { tcdDExt = an, tcdLName = ltycon, tcdTyVars = tyvars , tcdFixity = fixity, tcdDataDefn = defn }) = do - (_, an', ltycon', tyvars', _, _mctxt', defn') <- + (_, an', ltycon', tyvars', _, defn') <- exactDataDefn an (exactVanillaDeclHead ltycon tyvars fixity) defn return (DataDecl { tcdDExt = an', tcdLName = ltycon', tcdTyVars = tyvars' , tcdFixity = fixity, tcdDataDefn = defn' }) @@ -3707,7 +3803,7 @@ instance ExactPrint (TyClDecl GhcPs) where (an0, fds', lclas', tyvars',context') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lidl AnnSemi - ds <- withSortKey sortKey + (sortKey', ds) <- withSortKey sortKey [(ClsSigTag, prepareListAnnotationA sigs), (ClsMethodTag, prepareListAnnotationA methods), (ClsAtTag, prepareListAnnotationA ats), @@ -3720,7 +3816,7 @@ instance ExactPrint (TyClDecl GhcPs) where methods' = undynamic ds ats' = undynamic ds at_defs' = undynamic ds - return (ClassDecl {tcdCExt = (an3, lo, sortKey), + return (ClassDecl {tcdCExt = (an3, lo, sortKey'), tcdCtxt = context', tcdLName = lclas', tcdTyVars = tyvars', tcdFixity = fixity, tcdFDs = fds', @@ -3845,7 +3941,7 @@ exactDataDefn -> HsDataDefn GhcPs -> EP w m ( [AddEpAnn] -- ^ from exactHdr , [AddEpAnn] -- ^ updated one passed in - , LocatedN RdrName, a, b, Maybe (LHsContext GhcPs), HsDataDefn GhcPs) + , LocatedN RdrName, a, b, HsDataDefn GhcPs) exactDataDefn an exactHdr (HsDataDefn { dd_ext = x, dd_ctxt = context , dd_cType = mb_ct @@ -3883,8 +3979,8 @@ exactDataDefn an exactHdr _ -> panic "exacprint NewTypeCon" an6 <- markEpAnnL an5 lidl AnnCloseC derivings' <- mapM markAnnotated derivings - return (anx, an6, ln', tvs', b, mctxt', - (HsDataDefn { dd_ext = x, dd_ctxt = context + return (anx, an6, ln', tvs', b, + (HsDataDefn { dd_ext = x, dd_ctxt = mctxt' , dd_cType = mb_ct' , dd_kindSig = mb_sig' , dd_cons = condecls'', dd_derivs = derivings' })) @@ -3941,22 +4037,23 @@ instance ExactPrint (InjectivityAnn GhcPs) where class Typeable flag => ExactPrintTVFlag flag where exactTVDelimiters :: (Monad m, Monoid w) - => [AddEpAnn] -> flag -> EP w m (HsTyVarBndr flag GhcPs) - -> EP w m ([AddEpAnn], (HsTyVarBndr flag GhcPs)) + => [AddEpAnn] -> flag + -> ([AddEpAnn] -> EP w m ([AddEpAnn], HsTyVarBndr flag GhcPs)) + -> EP w m ([AddEpAnn], flag, (HsTyVarBndr flag GhcPs)) instance ExactPrintTVFlag () where - exactTVDelimiters an _ thing_inside = do + exactTVDelimiters an flag thing_inside = do an0 <- markEpAnnAllL' an lid AnnOpenP - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid AnnCloseP - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid AnnCloseP + return (an2, flag, r) instance ExactPrintTVFlag Specificity where exactTVDelimiters an s thing_inside = do an0 <- markEpAnnAllL' an lid open - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid close - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid close + return (an2, s, r) where (open, close) = case s of SpecifiedSpec -> (AnnOpenP, AnnCloseP) @@ -3964,33 +4061,33 @@ instance ExactPrintTVFlag Specificity where instance ExactPrintTVFlag (HsBndrVis GhcPs) where exactTVDelimiters an0 bvis thing_inside = do - case bvis of - HsBndrRequired _ -> return () - HsBndrInvisible at -> markEpToken at >> return () + bvis' <- case bvis of + HsBndrRequired _ -> return bvis + HsBndrInvisible at -> HsBndrInvisible <$> markEpToken at an1 <- markEpAnnAllL' an0 lid AnnOpenP - r <- thing_inside - an2 <- markEpAnnAllL' an1 lid AnnCloseP - return (an2, r) + (an2, r) <- thing_inside an1 + an3 <- markEpAnnAllL' an2 lid AnnCloseP + return (an3, bvis', r) instance ExactPrintTVFlag flag => ExactPrint (HsTyVarBndr flag GhcPs) where getAnnotationEntry _ = NoEntryVal setAnnotationAnchor a _ _ _ = a exact (UserTyVar an flag n) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - return (UserTyVar an flag n') + return (ani, UserTyVar an flag n') case r of - (an', UserTyVar _ flag'' n'') -> return (UserTyVar an' flag'' n'') + (an', flag', UserTyVar _ _ n'') -> return (UserTyVar an' flag' n'') _ -> error "KindedTyVar should never happen here" exact (KindedTyVar an flag n k) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - an0 <- markEpAnnL an lidl AnnDcolon + an0 <- markEpAnnL ani lidl AnnDcolon k' <- markAnnotated k - return (KindedTyVar an0 flag n' k') + return (an0, KindedTyVar an0 flag n' k') case r of - (an',KindedTyVar _ flag'' n'' k'') -> return (KindedTyVar an' flag'' n'' k'') + (an',flag', KindedTyVar _ _ n'' k'') -> return (KindedTyVar an' flag' n'' k'') _ -> error "UserTyVar should never happen here" -- --------------------------------------------------------------------- @@ -4150,17 +4247,16 @@ instance ExactPrint (HsDerivingClause GhcPs) where , deriv_clause_strategy = dcs , deriv_clause_tys = dct }) = do an0 <- markEpAnnL an lidl AnnDeriving - exact_strat_before + dcs0 <- case dcs of + Just (L _ ViaStrategy{}) -> return dcs + _ -> mapM markAnnotated dcs dct' <- markAnnotated dct - exact_strat_after + dcs1 <- case dcs0 of + Just (L _ ViaStrategy{}) -> mapM markAnnotated dcs0 + _ -> return dcs0 return (HsDerivingClause { deriv_clause_ext = an0 - , deriv_clause_strategy = dcs + , deriv_clause_strategy = dcs1 , deriv_clause_tys = dct' }) - where - (exact_strat_before, exact_strat_after) = - case dcs of - Just v@(L _ ViaStrategy{}) -> (pure (), markAnnotated v >> pure ()) - _ -> (mapM_ markAnnotated dcs, pure ()) -- --------------------------------------------------------------------- @@ -4467,7 +4563,9 @@ instance ExactPrint (ConDeclField GhcPs) where instance ExactPrint (FieldOcc GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact f@(FieldOcc _ n) = markAnnotated n >> return f + exact (FieldOcc x n) = do + n' <- markAnnotated n + return (FieldOcc x n') -- --------------------------------------------------------------------- @@ -4535,7 +4633,7 @@ instance ExactPrint (LocatedL [LocatedA (IE GhcPs)]) where an0 <- markEpAnnL' an lal_rest AnnHiding p <- getPosP debugM $ "LocatedL [LIE:p=" ++ showPprUnsafe p - (an1, ies') <- markAnnList an0 (markAnnotated ies) + (an1, ies') <- markAnnList an0 (markAnnotated (filter notIEDoc ies)) return (L an1 ies') instance (ExactPrint (Match GhcPs (LocatedA body))) @@ -4985,6 +5083,14 @@ setExtraDP md = do debugM $ "setExtraDP:" ++ show md modify (\s -> s {uExtraDP = md}) +getExtraDPReturn :: (Monad m, Monoid w) => EP w m (Maybe DeltaPos) +getExtraDPReturn = gets uExtraDPReturn + +setExtraDPReturn :: (Monad m, Monoid w) => Maybe DeltaPos -> EP w m () +setExtraDPReturn md = do + debugM $ "setExtraDPReturn:" ++ show md + modify (\s -> s {uExtraDPReturn = md}) + getPriorEndD :: (Monad m, Monoid w) => EP w m Pos getPriorEndD = gets dPriorEndPosition @@ -5007,13 +5113,13 @@ setPriorEndNoLayoutD pe = do debugM $ "setPriorEndNoLayoutD:pe=" ++ show pe modify (\s -> s { dPriorEndPosition = pe }) -setPriorEndASTD :: (Monad m, Monoid w) => Bool -> RealSrcSpan -> EP w m () -setPriorEndASTD layout pe = setPriorEndASTPD layout (rs2range pe) +setPriorEndASTD :: (Monad m, Monoid w) => RealSrcSpan -> EP w m () +setPriorEndASTD pe = setPriorEndASTPD (rs2range pe) -setPriorEndASTPD :: (Monad m, Monoid w) => Bool -> (Pos,Pos) -> EP w m () -setPriorEndASTPD layout pe@(fm,to) = do +setPriorEndASTPD :: (Monad m, Monoid w) => (Pos,Pos) -> EP w m () +setPriorEndASTPD pe@(fm,to) = do debugM $ "setPriorEndASTD:pe=" ++ show pe - when layout $ setLayoutStartD (snd fm) + setLayoutStartD (snd fm) modify (\s -> s { dPriorEndPosition = to } ) setLayoutStartD :: (Monad m, Monoid w) => Int -> EP w m () @@ -5044,7 +5150,7 @@ getUnallocatedComments :: (Monad m, Monoid w) => EP w m [Comment] getUnallocatedComments = gets epComments putUnallocatedComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -putUnallocatedComments cs = modify (\s -> s { epComments = cs } ) +putUnallocatedComments !cs = modify (\s -> s { epComments = cs } ) -- | Push a fresh stack frame for the applied comments gatherer pushAppliedComments :: (Monad m, Monoid w) => EP w m () @@ -5054,7 +5160,7 @@ pushAppliedComments = modify (\s -> s { epCommentsApplied = []:(epCommentsApplie -- takeAppliedComments, and clear them, not popping the stack takeAppliedComments :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedComments = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5067,7 +5173,7 @@ takeAppliedComments = do -- takeAppliedComments, and clear them, popping the stack takeAppliedCommentsPop :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedCommentsPop = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5080,7 +5186,7 @@ takeAppliedCommentsPop = do -- when doing delta processing applyComment :: (Monad m, Monoid w) => Comment -> EP w m () applyComment c = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> modify (\s -> s { epCommentsApplied = [[c]] } ) (h:t) -> modify (\s -> s { epCommentsApplied = (c:h):t } ) ===================================== utils/check-exact/Main.hs ===================================== @@ -470,7 +470,7 @@ changeAddDecl1 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtStart m decl' + replaceTopLevelDecls m = return $ insertAtStart m decl' return p' -- --------------------------------------------------------------------- @@ -483,7 +483,7 @@ changeAddDecl2 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtEnd m decl' + replaceTopLevelDecls m = return $ insertAtEnd m decl' return p' -- --------------------------------------------------------------------- @@ -500,7 +500,7 @@ changeAddDecl3 libdir top = do l2' = setEntryDP l2 (DifferentLine 2 0) replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAt f m decl' + replaceTopLevelDecls m = return $ insertAt f m decl' return p' -- --------------------------------------------------------------------- @@ -571,8 +571,9 @@ changeLocalDecls2 libdir (L l p) = do changeWhereIn3a :: Changer changeWhereIn3a _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) - debugM $ unlines w + decls = balanceCommentsList decls0 + (_de0:_:de1:_d2:_) = decls + debugM $ "changeWhereIn3a:de1:" ++ showAst de1 let p2 = p { hsmodDecls = decls} return (L l p2) @@ -581,13 +582,12 @@ changeWhereIn3a _libdir (L l p) = do changeWhereIn3b :: Changer changeWhereIn3b _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) + decls = balanceCommentsList decls0 (de0:tdecls@(_:de1:d2:_)) = decls de0' = setEntryDP de0 (DifferentLine 2 0) de1' = setEntryDP de1 (DifferentLine 2 0) d2' = setEntryDP d2 (DifferentLine 2 0) decls' = d2':de1':de0':tdecls - debugM $ unlines w debugM $ "changeWhereIn3b:de1':" ++ showAst de1' let p2 = p { hsmodDecls = decls'} return (L l p2) @@ -598,37 +598,37 @@ addLocaLDecl1 :: Changer addLocaLDecl1 libdir top = do Right (L ld (ValD _ decl)) <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let decl' = setEntryDP (L ld decl) (DifferentLine 1 5) - doAddLocal = do - let lp = top - (de1:d2:d3:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - (de1',_) <- modifyValD (getLocA de1'') de1'' $ \_m d -> do - return ((wrapDecl decl' : d),Nothing) - replaceDecls lp [de1', d2', d3] - - (lp',_,w) <- runTransformT doAddLocal - debugM $ "addLocaLDecl1:" ++ intercalate "\n" w + doAddLocal :: ParsedSource + doAddLocal = replaceDecls lp [de1', d2', d3] + where + lp = top + (de1:d2:d3:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 + (de1',_) = modifyValD (getLocA de1'') de1'' $ \_m d -> ((wrapDecl decl' : d),Nothing) + + let lp' = doAddLocal return lp' -- --------------------------------------------------------------------- + addLocaLDecl2 :: Changer addLocaLDecl2 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 + doAddLocal = replaceDecls lp [parent',d2'] + where + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - newDecl' <- transferEntryDP' d newDecl - let d' = setEntryDP d (DifferentLine 1 0) - return ((newDecl':d':ds),Nothing) + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = transferEntryDP' d (makeDeltaAst newDecl) + d' = setEntryDP d (DifferentLine 1 0) + in ((newDecl':d':ds),Nothing) - replaceDecls lp [parent',d2'] - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -637,19 +637,18 @@ addLocaLDecl3 :: Changer addLocaLDecl3 libdir top = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - let lp = top - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - return (((d:ds) ++ [newDecl']),Nothing) + doAddLocal = replaceDecls (anchorEof lp) [parent',d2'] + where + lp = top + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - replaceDecls (anchorEof lp) [parent',d2'] + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = setEntryDP newDecl (DifferentLine 1 0) + in (((d:ds) ++ [newDecl']),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -659,40 +658,38 @@ addLocaLDecl4 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") Right newSig <- withDynFlags libdir (\df -> parseDecl df "sig" "nn :: Int") let - doAddLocal = do - (parent:ds) <- hsDecls lp + doAddLocal = replaceDecls (anchorEof lp) (parent':ds) + where + (parent:ds) = hsDecls (makeDeltaAst lp) - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - let newSig' = setEntryDP newSig (DifferentLine 1 4) + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 0) + newSig' = setEntryDP (makeDeltaAst newSig) (DifferentLine 1 5) - (parent',_) <- modifyValD (getLocA parent) parent $ \_m decls -> do - return ((decls++[newSig',newDecl']),Nothing) + (parent',_) = modifyValD (getLocA parent) parent $ \_m decls -> + ((decls++[newSig',newDecl']),Nothing) - replaceDecls (anchorEof lp) (parent':ds) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' - -- --------------------------------------------------------------------- addLocaLDecl5 :: Changer addLocaLDecl5 _libdir lp = do let - doAddLocal = do - decls <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList decls + doAddLocal = replaceDecls lp [s1,de1',d3'] + where + decls = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList decls - let d3' = setEntryDP d3 (DifferentLine 2 0) + d3' = setEntryDP d3 (DifferentLine 2 0) - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m _decls -> do - let d2' = setEntryDP d2 (DifferentLine 1 0) - return ([d2'],Nothing) - replaceDecls lp [s1,de1',d3'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m _decls -> + let + d2' = setEntryDP d2 (DifferentLine 1 0) + in ([d2'],Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -701,39 +698,36 @@ addLocaLDecl6 :: Changer addLocaLDecl6 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "x = 3") let - newDecl' = setEntryDP newDecl (DifferentLine 1 4) - doAddLocal = do - decls0 <- hsDecls lp - [de1'',d2] <- balanceCommentsList decls0 + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 5) + doAddLocal = replaceDecls lp [de1', d2] + where + decls0 = hsDecls lp + [de1'',d2] = balanceCommentsList decls0 - let de1 = captureMatchLineSpacing de1'' - let L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 - let [ma1,_ma2] = ms + de1 = captureMatchLineSpacing de1'' + L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 + [ma1,_ma2] = ms - (de1',_) <- modifyValD (getLocA ma1) de1 $ \_m decls -> do - return ((newDecl' : decls),Nothing) - replaceDecls lp [de1', d2] + (de1',_) = modifyValD (getLocA ma1) de1 $ \_m decls -> + ((newDecl' : decls),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- rmDecl1 :: Changer -rmDecl1 _libdir top = do - let doRmDecl = do - let lp = top - tlDecs0 <- hsDecls lp - tlDecs' <- balanceCommentsList tlDecs0 - let tlDecs = captureLineSpacing tlDecs' - let (de1:_s1:_d2:d3:ds) = tlDecs - let d3' = setEntryDP d3 (DifferentLine 2 0) - - replaceDecls lp (de1:d3':ds) - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" +rmDecl1 _libdir lp = do + let + doRmDecl = replaceDecls lp (de1:d3':ds) + where + tlDecs0 = hsDecls lp + tlDecs = balanceCommentsList tlDecs0 + (de1:_s1:_d2:d3:ds) = tlDecs + d3' = setEntryDP d3 (DifferentLine 2 0) + + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -745,13 +739,13 @@ rmDecl2 _libdir lp = do let go :: GHC.LHsExpr GhcPs -> Transform (GHC.LHsExpr GhcPs) go e@(GHC.L _ (GHC.HsLet{})) = do - decs0 <- hsDecls e - decs <- balanceCommentsList $ captureLineSpacing decs0 - e' <- replaceDecls e (init decs) + let decs0 = hsDecls e + let decs = balanceCommentsList $ captureLineSpacing decs0 + let e' = replaceDecls e (init decs) return e' go x = return x - everywhereM (mkM go) lp + everywhereM (mkM go) (makeDeltaAst lp) let (lp',_,_w) = runTransform doRmDecl debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" @@ -762,17 +756,15 @@ rmDecl2 _libdir lp = do rmDecl3 :: Changer rmDecl3 _libdir lp = do let - doRmDecl = do - [de1,d2] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1] -> do - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([],Just sd1') - - replaceDecls lp [de1',sd1,d2] + doRmDecl = replaceDecls lp [de1',sd1,d2] + where + [de1,d2] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a] -> + let + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([],Just sd1') - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -780,19 +772,15 @@ rmDecl3 _libdir lp = do rmDecl4 :: Changer rmDecl4 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1,sd2] -> do - sd2' <- transferEntryDP' sd1 sd2 - - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([sd2'],Just sd1') - - replaceDecls (anchorEof lp) [de1',sd1] - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + doRmDecl = replaceDecls (anchorEof lp) [de1',sd1] + where + [de1] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a,sd2] -> + let + sd2' = transferEntryDP' sd1a sd2 + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([sd2'],Just sd1') + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -805,10 +793,8 @@ rmDecl5 _libdir lp = do go :: HsExpr GhcPs -> Transform (HsExpr GhcPs) go (HsLet (tkLet, tkIn) lb expr) = do let decs = hsDeclsLocalBinds lb - let hdecs : _ = decs let dec = last decs - _ <- transferEntryDP hdecs dec - lb' <- replaceDeclsValbinds WithoutWhere lb [dec] + let lb' = replaceDeclsValbinds WithoutWhere lb [dec] return (HsLet (tkLet, tkIn) lb' expr) go x = return x @@ -823,73 +809,61 @@ rmDecl5 _libdir lp = do rmDecl6 :: Changer rmDecl6 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m subDecs -> do - let subDecs' = captureLineSpacing subDecs - let (ss1:_sd1:sd2:sds) = subDecs' - sd2' <- transferEntryDP' ss1 sd2 - - return (sd2':sds,Nothing) + doRmDecl = replaceDecls lp [de1'] + where + [de1] = hsDecls lp - replaceDecls lp [de1'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m subDecs -> + let + subDecs' = captureLineSpacing subDecs + (ss1:_sd1:sd2:sds) = subDecs' + sd2' = transferEntryDP' ss1 sd2 + in (sd2':sds,Nothing) - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmDecl7 :: Changer -rmDecl7 _libdir top = do +rmDecl7 _libdir lp = do let - doRmDecl = do - let lp = top - tlDecs <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList tlDecs - - d3' <- transferEntryDP' d2 d3 - - replaceDecls lp [s1,de1,d3'] + doRmDecl = replaceDecls lp [s1,de1,d3'] + where + tlDecs = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList tlDecs + d3' = transferEntryDP' d2 d3 - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig1 :: Changer rmTypeSig1 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let (s0:de1:d2) = tlDecs - s1 = captureTypeSigSpacing s0 - (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 - L ln n2' <- transferEntryDP n1 n2 - let s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) - replaceDecls lp (s1':de1:d2) - - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let doRmDecl = replaceDecls lp (s1':de1:d2) + where + tlDecs = hsDecls lp + (s0:de1:d2) = tlDecs + s1 = captureTypeSigSpacing s0 + (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 + L ln n2' = transferEntryDP n1 n2 + s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig2 :: Changer rmTypeSig2 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let [de1] = tlDecs - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m [s,d] -> do - d' <- transferEntryDP' s d - return $ ([d'],Nothing) - `debug` ("rmTypeSig2:(d,d')" ++ showAst (d,d')) - replaceDecls lp [de1'] + let doRmDecl = replaceDecls lp [de1'] + where + tlDecs = hsDecls lp + [de1] = tlDecs + (de1',_) = modifyValD (getLocA de1) de1 $ \_m [_s,d] -> ([d],Nothing) - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -958,13 +932,15 @@ addClassMethod libdir lp = do let decl' = setEntryDP decl (DifferentLine 1 3) let sig' = setEntryDP sig (DifferentLine 2 3) let doAddMethod = do - [cd] <- hsDecls lp - (f1:f2s:f2d:_) <- hsDecls cd - let f2s' = setEntryDP f2s (DifferentLine 2 3) - cd' <- replaceDecls cd [f1, sig', decl', f2s', f2d] - replaceDecls lp [cd'] - - (lp',_,w) <- runTransformT doAddMethod + let + [cd] = hsDecls lp + (f1:f2s:f2d:_) = hsDecls cd + f2s' = setEntryDP f2s (DifferentLine 2 3) + cd' = replaceDecls cd [f1, sig', decl', f2s', f2d] + lp' = replaceDecls lp [cd'] + return lp' + + let (lp',_,w) = runTransform doAddMethod debugM $ "addClassMethod:" ++ intercalate "\n" w return lp' ===================================== utils/check-exact/Parsers.hs ===================================== @@ -260,7 +260,7 @@ parseModuleEpAnnsWithCppInternal cppOptions dflags file = do GHC.PFailed pst -> Left (GHC.GhcPsMessage <$> GHC.getPsErrorMessages pst) GHC.POk _ pmod - -> Right $ (injectedComments, dflags', fixModuleTrailingComments pmod) + -> Right $ (injectedComments, dflags', fixModuleComments pmod) -- | Internal function. Exposed if you want to muck with DynFlags -- before parsing. Or after parsing. @@ -269,8 +269,10 @@ postParseTransform -> Either a (GHC.ParsedSource) postParseTransform parseRes = fmap mkAnns parseRes where - -- TODO:AZ perhaps inject the comments into the parsedsource here already - mkAnns (_cs, _, m) = fixModuleTrailingComments m + mkAnns (_cs, _, m) = fixModuleComments m + +fixModuleComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleComments p = fixModuleHeaderComments $ fixModuleTrailingComments p fixModuleTrailingComments :: GHC.ParsedSource -> GHC.ParsedSource fixModuleTrailingComments (GHC.L l p) = GHC.L l p' @@ -293,6 +295,47 @@ fixModuleTrailingComments (GHC.L l p) = GHC.L l p' in cs'' _ -> cs +-- Deal with https://gitlab.haskell.org/ghc/ghc/-/issues/23984 +-- The Lexer works bottom-up, so does not have module declaration info +-- when the first top decl processed +fixModuleHeaderComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleHeaderComments (GHC.L l p) = GHC.L l p' + where + moveComments :: GHC.EpaLocation -> GHC.LHsDecl GHC.GhcPs -> GHC.EpAnnComments + -> (GHC.LHsDecl GHC.GhcPs, GHC.EpAnnComments) + moveComments GHC.EpaDelta{} dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.UnhelpfulSpan _)) dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.RealSrcSpan r _)) (GHC.L (GHC.EpAnn anc an csd) a) cs = (dd,css) + where + -- Move any comments on the decl that occur prior to the location + pc = GHC.priorComments csd + fc = GHC.getFollowingComments csd + bf (GHC.L anch _) = GHC.anchor anch > r + (move,keep) = break bf pc + csd' = GHC.EpaCommentsBalanced keep fc + + dd = GHC.L (GHC.EpAnn anc an csd') a + css = cs <> GHC.EpaComments move + + (ds',an') = rebalance (GHC.hsmodDecls p, GHC.hsmodAnn $ GHC.hsmodExt p) + p' = p { GHC.hsmodExt = (GHC.hsmodExt p){ GHC.hsmodAnn = an' }, + GHC.hsmodDecls = ds' + } + + rebalance :: ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + -> ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + rebalance (ds, GHC.EpAnn a an cs) = (ds1, GHC.EpAnn a an cs') + where + (ds1,cs') = case break (\(GHC.AddEpAnn k _) -> k == GHC.AnnWhere) (GHC.am_main an) of + (_, (GHC.AddEpAnn _ whereLoc:_)) -> + case GHC.hsmodDecls p of + (d:ds0) -> (d':ds0, cs0) + where (d',cs0) = moveComments whereLoc d cs + ds0 -> (ds0,cs) + _ -> (ds,cs) + + + -- | Internal function. Initializes DynFlags value for parsing. -- -- Passes "-hide-all-packages" to the GHC API to prevent parsing of ===================================== utils/check-exact/Transform.hs ===================================== @@ -63,7 +63,7 @@ module Transform -- *** Low level operations used in 'HasDecls' , balanceComments , balanceCommentsList - , balanceCommentsList' + , balanceCommentsListA , anchorEof -- ** Managing lists, pure functions @@ -92,6 +92,7 @@ import qualified Control.Monad.Fail as Fail import GHC hiding (parseModule, parsedSource) import GHC.Data.FastString +import GHC.Types.SrcLoc import Data.Data import Data.Maybe @@ -154,6 +155,7 @@ logDataWithAnnsTr str ast = do -- |If we need to add new elements to the AST, they need their own -- 'SrcSpan' for this. +-- This should no longer be needed, we use an @EpaDelta@ location instead. uniqueSrcSpanT :: (Monad m) => TransformT m SrcSpan uniqueSrcSpanT = do col <- get @@ -171,15 +173,6 @@ srcSpanStartLine' _ = 0 -- --------------------------------------------------------------------- -captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag -captureOrderBinds ls = AnnSortKey $ map go ls - where - go (L _ (ValD _ _)) = BindTag - go (L _ (SigD _ _)) = SigDTag - go d = error $ "captureOrderBinds:" ++ showGhc d - --- --------------------------------------------------------------------- - captureMatchLineSpacing :: LHsDecl GhcPs -> LHsDecl GhcPs captureMatchLineSpacing (L l (ValD x (FunBind a b (MG c (L d ms ))))) = L l (ValD x (FunBind a b (MG c (L d ms')))) @@ -253,7 +246,7 @@ setEntryDPDecl d dp = setEntryDP d dp -- |Set the true entry 'DeltaPos' from the annotation for a given AST -- element. This is the 'DeltaPos' ignoring any comments. -setEntryDP :: NoAnn t => LocatedAn t a -> DeltaPos -> LocatedAn t a +setEntryDP :: LocatedAn t a -> DeltaPos -> LocatedAn t a setEntryDP (L (EpAnn (EpaSpan ss@(UnhelpfulSpan _)) an cs) a) dp = L (EpAnn (EpaDelta ss dp []) an cs) a setEntryDP (L (EpAnn (EpaSpan ss) an (EpaComments [])) a) dp @@ -293,7 +286,7 @@ setEntryDP (L (EpAnn (EpaSpan ss@(RealSrcSpan r _)) an cs) a) dp L (EpAnn (EpaDelta ss edp csd) an cs'') a where cs'' = setPriorComments cs [] - csd = L (EpaDelta ss dp NoComments) c:cs' + csd = L (EpaDelta ss dp NoComments) c:commentOrigDeltas cs' lc = last $ (L ca c:cs') delta = case getLoc lc of EpaSpan (RealSrcSpan rr _) -> ss2delta (ss2pos rr) r @@ -335,18 +328,15 @@ setEntryDPFromAnchor off (EpaSpan (RealSrcSpan anc _)) ll@(L la _) = setEntryDP -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -transferEntryDP :: (Monad m, NoAnn t2, Typeable t1, Typeable t2) - => LocatedAn t1 a -> LocatedAn t2 b -> TransformT m (LocatedAn t2 b) -transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = do - logTr $ "transferEntryDP': EpAnn,EpAnn" +transferEntryDP :: (Typeable t1, Typeable t2) + => LocatedAn t1 a -> LocatedAn t2 b -> (LocatedAn t2 b) +transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = -- Problem: if the original had preceding comments, blindly -- transferring the location is not correct case priorComments cs1 of - [] -> return (L (EpAnn anc1 (combine an1 an2) cs2) b) + [] -> (L (EpAnn anc1 (combine an1 an2) cs2) b) -- TODO: what happens if the receiving side already has comments? - (L anc _:_) -> do - logDataWithAnnsTr "transferEntryDP':priorComments anc=" anc - return (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) + (L _ _:_) -> (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) -- |If a and b are the same type return first arg, else return second @@ -356,10 +346,11 @@ combine x y = fromMaybe y (cast x) -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -- TODO: call transferEntryDP, and use pushDeclDP -transferEntryDP' :: (Monad m) => LHsDecl GhcPs -> LHsDecl GhcPs -> TransformT m (LHsDecl GhcPs) -transferEntryDP' la lb = do - (L l2 b) <- transferEntryDP la lb - return (L l2 (pushDeclDP b (SameLine 0))) +transferEntryDP' :: LHsDecl GhcPs -> LHsDecl GhcPs -> (LHsDecl GhcPs) +transferEntryDP' la lb = + let + (L l2 b) = transferEntryDP la lb + in (L l2 (pushDeclDP b (SameLine 0))) pushDeclDP :: HsDecl GhcPs -> DeltaPos -> HsDecl GhcPs @@ -375,13 +366,24 @@ pushDeclDP d _dp = d -- --------------------------------------------------------------------- -balanceCommentsList :: (Monad m) => [LHsDecl GhcPs] -> TransformT m [LHsDecl GhcPs] -balanceCommentsList [] = return [] -balanceCommentsList [x] = return [x] -balanceCommentsList (a:b:ls) = do - (a',b') <- balanceComments a b - r <- balanceCommentsList (b':ls) - return (a':r) +-- | If we compile in haddock mode, the haddock processing inserts +-- DocDecls to carry the Haddock Documentation. We ignore these in +-- exact printing, as all the comments are also available in their +-- normal location, and the haddock processing is lossy, in that it +-- does not preserve all haddock-like comments. When we balance +-- comments in a list, we migrate some to preceding or following +-- declarations in the list. We must make sure we do not move any to +-- these DocDecls, which are not printed. +balanceCommentsList :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList decls = balanceCommentsList' (filter notDocDecl decls) + +balanceCommentsList' :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList' [] = [] +balanceCommentsList' [x] = [x] +balanceCommentsList' (a:b:ls) = (a':r) + where + (a',b') = balanceComments a b + r = balanceCommentsList' (b':ls) -- |The GHC parser puts all comments appearing between the end of one AST -- item and the beginning of the next as 'annPriorComments' for the second one. @@ -389,28 +391,27 @@ balanceCommentsList (a:b:ls) = do -- from the second one to the 'annFollowingComments' of the first if they belong -- to it instead. This is typically required before deleting or duplicating -- either of the AST elements. -balanceComments :: (Monad m) - => LHsDecl GhcPs -> LHsDecl GhcPs - -> TransformT m (LHsDecl GhcPs, LHsDecl GhcPs) -balanceComments first second = do +balanceComments :: LHsDecl GhcPs -> LHsDecl GhcPs + -> (LHsDecl GhcPs, LHsDecl GhcPs) +balanceComments first second = case first of - (L l (ValD x fb@(FunBind{}))) -> do - (L l' fb',second') <- balanceCommentsFB (L l fb) second - return (L l' (ValD x fb'), second') - _ -> balanceComments' first second + (L l (ValD x fb@(FunBind{}))) -> + let + (L l' fb',second') = balanceCommentsFB (L l fb) second + in (L l' (ValD x fb'), second') + _ -> balanceCommentsA first second --- |Once 'balanceComments' has been called to move trailing comments to a +-- |Once 'balanceCommentsA has been called to move trailing comments to a -- 'FunBind', these need to be pushed down from the top level to the last -- 'Match' if that 'Match' needs to be manipulated. -balanceCommentsFB :: (Monad m) - => LHsBind GhcPs -> LocatedA b -> TransformT m (LHsBind GhcPs, LocatedA b) -balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do - debugM $ "balanceCommentsFB entered: " ++ showGhc (ss2range $ locA lf) +balanceCommentsFB :: LHsBind GhcPs -> LocatedA b -> (LHsBind GhcPs, LocatedA b) +balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second + = balanceCommentsA (packFunBind bind) second' -- There are comments on lf. We need to -- + Keep the prior ones here -- + move the interior ones to the first match, -- + move the trailing ones to the last match. - let + where (before,middle,after) = case entry lf of EpaSpan (RealSrcSpan ss _) -> let @@ -426,40 +427,29 @@ balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do getFollowingComments $ comments lf) lf' = setCommentsEpAnn lf (EpaComments before) - debugM $ "balanceCommentsFB (before, after): " ++ showAst (before, after) - debugM $ "balanceCommentsFB lf': " ++ showAst lf' - -- let matches' = case matches of - let matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] - matches' = case matches of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaComments middle )) m':ms') - _ -> error "balanceCommentsFB" - matches'' <- balanceCommentsList' matches' - let (m,ms) = case reverse matches'' of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m',ms') - -- (L (addCommentsToEpAnnS lm' (EpaCommentsBalanced [] after)) m',ms') - _ -> error "balanceCommentsFB4" - debugM $ "balanceCommentsFB: (m,ms):" ++ showAst (m,ms) - (m',second') <- balanceComments' m second - m'' <- balanceCommentsMatch m' - let (m''',lf'') = case ms of - [] -> moveLeadingComments m'' lf' - _ -> (m'',lf') - debugM $ "balanceCommentsFB: (lf'', m'''):" ++ showAst (lf'',m''') - debugM $ "balanceCommentsFB done" - let bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) - debugM $ "balanceCommentsFB returning:" ++ showAst bind - balanceComments' (packFunBind bind) second' -balanceCommentsFB f s = balanceComments' f s + matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] + matches' = case matches of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaComments middle )) m0:ms') + _ -> error "balanceCommentsFB" + matches'' = balanceCommentsListA matches' + (m,ms) = case reverse matches'' of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m0,ms') + _ -> error "balanceCommentsFB4" + (m',second') = balanceCommentsA m second + m'' = balanceCommentsMatch m' + (m''',lf'') = case ms of + [] -> moveLeadingComments m'' lf' + _ -> (m'',lf') + bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) +balanceCommentsFB f s = balanceCommentsA f s -- | Move comments on the same line as the end of the match into the -- GRHS, prior to the binds -balanceCommentsMatch :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do - logTr $ "balanceCommentsMatch: (logInfo)=" ++ showAst (logInfo) - return (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) +balanceCommentsMatch :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) + = (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) where simpleBreak (r,_) = r /= 0 an1 = l @@ -468,7 +458,7 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do (move',stay') = break simpleBreak (trailingCommentsDeltas (anchorFromLocatedA (L l ())) cs1f) move = map snd move' stay = map snd stay' - (l'', grhss', binds', logInfo) + (l'', grhss', binds', _logInfo) = case reverse grhss of [] -> (l, [], binds, (EpaComments [], noSrcSpanA)) (L lg (GRHS ag grs rhs):gs) -> @@ -491,26 +481,24 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do pushTrailingComments :: WithWhere -> EpAnnComments -> HsLocalBinds GhcPs -> (Bool, HsLocalBinds GhcPs) pushTrailingComments _ _cs b at EmptyLocalBinds{} = (False, b) pushTrailingComments _ _cs (HsIPBinds _ _) = error "TODO: pushTrailingComments:HsIPBinds" -pushTrailingComments w cs lb@(HsValBinds an _) - = (True, HsValBinds an' vb) +pushTrailingComments w cs lb@(HsValBinds an _) = (True, HsValBinds an' vb) where decls = hsDeclsLocalBinds lb (an', decls') = case reverse decls of [] -> (addCommentsToEpAnn an cs, decls) (L la d:ds) -> (an, L (addCommentsToEpAnn la cs) d:ds) - (vb,_ws2) = case runTransform (replaceDeclsValbinds w lb (reverse decls')) of - ((HsValBinds _ vb'), _, ws2') -> (vb', ws2') - _ -> (ValBinds NoAnnSortKey [] [], []) + vb = case replaceDeclsValbinds w lb (reverse decls') of + (HsValBinds _ vb') -> vb' + _ -> ValBinds NoAnnSortKey [] [] -balanceCommentsList' :: (Monad m) => [LocatedA a] -> TransformT m [LocatedA a] -balanceCommentsList' [] = return [] -balanceCommentsList' [x] = return [x] -balanceCommentsList' (a:b:ls) = do - logTr $ "balanceCommentsList' entered" - (a',b') <- balanceComments' a b - r <- balanceCommentsList' (b':ls) - return (a':r) +balanceCommentsListA :: [LocatedA a] -> [LocatedA a] +balanceCommentsListA [] = [] +balanceCommentsListA [x] = [x] +balanceCommentsListA (a:b:ls) = (a':r) + where + (a',b') = balanceCommentsA a b + r = balanceCommentsListA (b':ls) -- |Prior to moving an AST element, make sure any trailing comments belonging to -- it are attached to it, and not the following element. Of necessity this is a @@ -518,13 +506,8 @@ balanceCommentsList' (a:b:ls) = do -- with a passed-in decision function. -- The initial situation is that all comments for a given anchor appear as prior comments -- Many of these should in fact be following comments for the previous anchor -balanceComments' :: (Monad m) => LocatedA a -> LocatedA b -> TransformT m (LocatedA a, LocatedA b) -balanceComments' la1 la2 = do - debugM $ "balanceComments': (anc1)=" ++ showAst (anc1) - debugM $ "balanceComments': (cs1s)=" ++ showAst (cs1s) - debugM $ "balanceComments': (cs1stay,cs1move)=" ++ showAst (cs1stay,cs1move) - debugM $ "balanceComments': (an1',an2')=" ++ showAst (an1',an2') - return (la1', la2') +balanceCommentsA :: LocatedA a -> LocatedA b -> (LocatedA a, LocatedA b) +balanceCommentsA la1 la2 = (la1', la2') where simpleBreak n (r,_) = r > n L an1 f = la1 @@ -532,26 +515,31 @@ balanceComments' la1 la2 = do anc1 = comments an1 anc2 = comments an2 - cs1s = splitCommentsEnd (anchorFromLocatedA la1) anc1 - cs1p = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1s) - cs1f = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1s) + (p1,m1,f1) = splitComments (anchorFromLocatedA la1) anc1 + cs1p = priorCommentsDeltas (anchorFromLocatedA la1) p1 - cs2s = splitCommentsEnd (anchorFromLocatedA la2) anc2 - cs2p = priorCommentsDeltas (anchorFromLocatedA la2) (priorComments cs2s) - cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) (getFollowingComments cs2s) + -- Split cs1 following comments into those before any + -- TrailingAnn's on an1, and any after + cs1f = splitCommentsEnd (fullSpanFromLocatedA la1) $ EpaComments f1 + cs1fp = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1f) + cs1ff = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1f) - -- Split cs1f into those that belong on an1 and ones that must move to an2 - (cs1move,cs1stay) = break (simpleBreak 1) cs1f + -- Split cs1ff into those that belong on an1 and ones that must move to an2 + (cs1move,cs1stay) = break (simpleBreak 1) cs1ff + + (p2,m2,f2) = splitComments (anchorFromLocatedA la2) anc2 + cs2p = priorCommentsDeltas (anchorFromLocatedA la2) p2 + cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) f2 (stay'',move') = break (simpleBreak 1) cs2p -- Need to also check for comments more closely attached to la1, -- ie trailing on the same line (move'',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchorFromLocatedA la1) (map snd stay'')) - move = sortEpaComments $ map snd (cs1move ++ move'' ++ move') - stay = sortEpaComments $ map snd (cs1stay ++ stay') + move = sortEpaComments $ map snd (cs1fp ++ cs1move ++ move'' ++ move') + stay = sortEpaComments $ m2 ++ map snd (cs1stay ++ stay') - an1' = setCommentsEpAnn (getLoc la1) (EpaCommentsBalanced (map snd cs1p) move) - an2' = setCommentsEpAnn (getLoc la2) (EpaCommentsBalanced stay (map snd cs2f)) + an1' = setCommentsEpAnn (getLoc la1) (epaCommentsBalanced (m1 ++ map snd cs1p) move) + an2' = setCommentsEpAnn (getLoc la2) (epaCommentsBalanced stay (map snd cs2f)) la1' = L an1' f la2' = L an2' s @@ -569,10 +557,9 @@ trailingCommentsDeltas r (la@(L l _):las) (al,_) = ss2posEnd rs' (ll,_) = ss2pos (anchor loc) --- AZ:TODO: this is identical to commentsDeltas priorCommentsDeltas :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] -priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) +priorCommentsDeltas r cs = go r (sortEpaComments cs) where go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] go _ [] = [] @@ -588,6 +575,21 @@ priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) -- --------------------------------------------------------------------- +-- | Split comments into ones occurring before the end of the reference +-- span, and those after it. +splitComments :: RealSrcSpan -> EpAnnComments -> ([LEpaComment], [LEpaComment], [LEpaComment]) +splitComments p cs = (before, middle, after) + where + cmpe (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p + cmpe (L _ _) = True + + cmpb (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2pos p + cmpb (L _ _) = True + + (beforeEnd, after) = break cmpe ((priorComments cs) ++ (getFollowingComments cs)) + (before, middle) = break cmpb beforeEnd + + -- | Split comments into ones occurring before the end of the reference -- span, and those after it. splitCommentsEnd :: RealSrcSpan -> EpAnnComments -> EpAnnComments @@ -598,8 +600,8 @@ splitCommentsEnd p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsEnd p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsEnd p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -617,8 +619,8 @@ splitCommentsStart p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsStart p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsStart p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -638,8 +640,8 @@ moveLeadingComments (L la a) lb = (L la' a, lb') -- TODO: need to set an entry delta on lb' to zero, and move the -- original spacing to the first comment. - la' = setCommentsEpAnn la (EpaCommentsBalanced [] after) - lb' = addCommentsToEpAnn lb (EpaCommentsBalanced before []) + la' = setCommentsEpAnn la (epaCommentsBalanced [] after) + lb' = addCommentsToEpAnn lb (epaCommentsBalanced before []) -- | A GHC comment includes the span of the preceding (non-comment) -- token. Takes an original list of comments, and converts the @@ -662,17 +664,27 @@ addCommentOrigDeltasAnn (EpAnn e a cs) = EpAnn e a (addCommentOrigDeltas cs) anchorFromLocatedA :: LocatedA a -> RealSrcSpan anchorFromLocatedA (L (EpAnn anc _ _) _) = anchor anc +-- | Get the full span of interest for comments from a LocatedA. +-- This extends up to the last TrailingAnn +fullSpanFromLocatedA :: LocatedA a -> RealSrcSpan +fullSpanFromLocatedA (L (EpAnn anc (AnnListItem tas) _) _) = rr + where + r = anchor anc + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + rr = case reverse (concatMap trailing_loc tas) of + [] -> r + (s:_) -> combineRealSrcSpans r s + -- --------------------------------------------------------------------- -balanceSameLineComments :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do - logTr $ "balanceSameLineComments: (la)=" ++ showGhc (ss2range $ locA la) - logTr $ "balanceSameLineComments: [logInfo]=" ++ showAst logInfo - return (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) +balanceSameLineComments :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) + = (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) where simpleBreak n (r,_) = r > n - (la',grhss', logInfo) = case reverse grhss of + (la',grhss', _logInfo) = case reverse grhss of [] -> (la,grhss,[]) (L lg (GRHS ga gs rhs):grs) -> (la'',reverse $ (L lg (GRHS ga' gs rhs)):grs,[(gac,(csp,csf))]) where @@ -684,7 +696,7 @@ balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do (move',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchor anc) csf) move = map snd move' stay = map snd stay' - cs1 = EpaCommentsBalanced csp stay + cs1 = epaCommentsBalanced csp stay gac = epAnnComments ga gfc = getFollowingComments gac @@ -734,24 +746,21 @@ addComma (EpAnn anc (AnnListItem as) cs) -- | Insert a declaration into an AST element having sub-declarations -- (@HasDecls@) according to the given location function. insertAt :: (HasDecls ast) - => (LHsDecl GhcPs - -> [LHsDecl GhcPs] - -> [LHsDecl GhcPs]) - -> ast - -> LHsDecl GhcPs - -> Transform ast -insertAt f t decl = do - oldDecls <- hsDecls t - oldDeclsb <- balanceCommentsList oldDecls - let oldDecls' = oldDeclsb - replaceDecls t (f decl oldDecls') + => (LHsDecl GhcPs + -> [LHsDecl GhcPs] + -> [LHsDecl GhcPs]) + -> ast + -> LHsDecl GhcPs + -> ast +insertAt f t decl = replaceDecls t (f decl oldDecls') + where + oldDecls = hsDecls t + oldDeclsb = balanceCommentsList oldDecls + oldDecls' = oldDeclsb -- |Insert a declaration at the beginning or end of the subdecls of the given -- AST item -insertAtStart, insertAtEnd :: (HasDecls ast) - => ast - -> LHsDecl GhcPs - -> Transform ast +insertAtStart, insertAtEnd :: HasDecls ast => ast -> LHsDecl GhcPs -> ast insertAtEnd = insertAt (\x xs -> xs ++ [x]) @@ -766,11 +775,11 @@ insertAtStart = insertAt insertFirst -- |Insert a declaration at a specific location in the subdecls of the given -- AST item -insertAfter, insertBefore :: (HasDecls (LocatedA ast)) +insertAfter, insertBefore :: HasDecls (LocatedA ast) => LocatedA old -> LocatedA ast -> LHsDecl GhcPs - -> Transform (LocatedA ast) + -> LocatedA ast insertAfter (getLocA -> k) = insertAt findAfter where findAfter x xs = @@ -797,10 +806,10 @@ class (Data t) => HasDecls t where -- given syntax phrase. They are always returned in the wrapped 'HsDecl' -- form, even if orginating in local decls. This is safe, as annotations -- never attach to the wrapper, only to the wrapped item. - hsDecls :: (Monad m) => t -> TransformT m [LHsDecl GhcPs] + hsDecls :: t -> [LHsDecl GhcPs] -- | Replace the directly enclosed decl list by the given - -- decl list. Runs in the 'Transform' monad to be able to update list order + -- decl list. As part of replacing it will update list order -- annotations, and rebalance comments and other layout changes as needed. -- -- For example, a call on replaceDecls for a wrapped 'FunBind' having no @@ -818,96 +827,86 @@ class (Data t) => HasDecls t where -- where -- nn = 2 -- @ - replaceDecls :: (Monad m) => t -> [LHsDecl GhcPs] -> TransformT m t + replaceDecls :: t -> [LHsDecl GhcPs] -> t -- --------------------------------------------------------------------- instance HasDecls ParsedSource where - hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = return decls + hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = decls replaceDecls (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps _decls)) decls - = do - logTr "replaceDecls LHsModule" - return (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) + = (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsDecl GhcPs)) where - hsDecls (L _ (TyClD _ c at ClassDecl{})) = return $ hsDeclsClassDecl c - hsDecls decl = do - error $ "hsDecls:decl=" ++ showAst decl - replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = do - let decl' = replaceDeclsClassDecl dec decls - return (L l (TyClD e decl')) - replaceDecls decl _decls = do - error $ "replaceDecls:decl=" ++ showAst decl + hsDecls (L _ (TyClD _ c at ClassDecl{})) = hsDeclsClassDecl c + hsDecls decl = error $ "hsDecls:decl=" ++ showAst decl + replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = + let + decl' = replaceDeclsClassDecl dec decls + in (L l (TyClD e decl')) + replaceDecls decl _decls + = error $ "replaceDecls:decl=" ++ showAst decl -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Match GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = hsDeclsLocalBinds lb replaceDecls (L l (Match xm c p (GRHSs xr rhs binds))) [] - = do - logTr "replaceDecls LMatch empty decls" - binds'' <- replaceDeclsValbinds WithoutWhere binds [] - return (L l (Match xm c p (GRHSs xr rhs binds''))) + = let + binds'' = replaceDeclsValbinds WithoutWhere binds [] + in (L l (Match xm c p (GRHSs xr rhs binds''))) replaceDecls m@(L l (Match xm c p (GRHSs xr rhs binds))) newBinds - = do - logTr "replaceDecls LMatch nonempty decls" + = let -- Need to throw in a fresh where clause if the binds were empty, -- in the annotations. - (l', rhs') <- case binds of - EmptyLocalBinds{} -> do - logTr $ "replaceDecls LMatch empty binds" - - logDataWithAnnsTr "Match.replaceDecls:balancing comments:m" m - L l' m' <- balanceSameLineComments m - logDataWithAnnsTr "Match.replaceDecls:(m1')" (L l' m') - return (l', grhssGRHSs $ m_grhss m') - _ -> return (l, rhs) - binds'' <- replaceDeclsValbinds WithWhere binds newBinds - logDataWithAnnsTr "Match.replaceDecls:binds'" binds'' - return (L l' (Match xm c p (GRHSs xr rhs' binds''))) + (l', rhs') = case binds of + EmptyLocalBinds{} -> + let + L l0 m' = balanceSameLineComments m + in (l0, grhssGRHSs $ m_grhss m') + _ -> (l, rhs) + binds'' = replaceDeclsValbinds WithWhere binds newBinds + in (L l' (Match xm c p (GRHSs xr rhs' binds''))) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsExpr GhcPs)) where - hsDecls (L _ (HsLet _ decls _ex)) = return $ hsDeclsLocalBinds decls - hsDecls _ = return [] + hsDecls (L _ (HsLet _ decls _ex)) = hsDeclsLocalBinds decls + hsDecls _ = [] replaceDecls (L ll (HsLet (tkLet, tkIn) binds ex)) newDecls - = do - logTr "replaceDecls HsLet" - let lastAnc = realSrcSpan $ spanHsLocaLBinds binds + = let + lastAnc = realSrcSpan $ spanHsLocaLBinds binds -- TODO: may be an intervening comment, take account for lastAnc - let (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of - (EpTok l, EpTok i) -> - let - off = case l of - (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r - (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 - (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 - (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c - ex'' = setEntryDPFromAnchor off i ex - newDecls'' = case newDecls of - [] -> newDecls - (d:ds) -> setEntryDPDecl d (SameLine 0) : ds - in ( EpTok l - , EpTok (addEpaLocationDelta off lastAnc i) - , ex'' - , newDecls'') - (_,_) -> (tkLet, tkIn, ex, newDecls) - binds' <- replaceDeclsValbinds WithoutWhere binds newDecls' - return (L ll (HsLet (tkLet', tkIn') binds' ex')) + (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of + (EpTok l, EpTok i) -> + let + off = case l of + (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r + (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 + (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 + (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c + ex'' = setEntryDPFromAnchor off i ex + newDecls'' = case newDecls of + [] -> newDecls + (d:ds) -> setEntryDPDecl d (SameLine 0) : ds + in ( EpTok l + , EpTok (addEpaLocationDelta off lastAnc i) + , ex'' + , newDecls'') + (_,_) -> (tkLet, tkIn, ex, newDecls) + binds' = replaceDeclsValbinds WithoutWhere binds newDecls' + in (L ll (HsLet (tkLet', tkIn') binds' ex')) -- TODO: does this make sense? Especially as no hsDecls for HsPar replaceDecls (L l (HsPar x e)) newDecls - = do - logTr "replaceDecls HsPar" - e' <- replaceDecls e newDecls - return (L l (HsPar x e')) + = let + e' = replaceDecls e newDecls + in (L l (HsPar x e')) replaceDecls old _new = error $ "replaceDecls (LHsExpr GhcPs) undefined for:" ++ showGhc old -- --------------------------------------------------------------------- @@ -934,53 +933,51 @@ hsDeclsPatBind x = error $ "hsDeclsPatBind called for:" ++ showGhc x -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBindD' \/ 'replaceDeclsPatBindD' is -- idempotent. -replaceDeclsPatBindD :: (Monad m) => LHsDecl GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsDecl GhcPs) -replaceDeclsPatBindD (L l (ValD x d)) newDecls = do - (L _ d') <- replaceDeclsPatBind (L l d) newDecls - return (L l (ValD x d')) +replaceDeclsPatBindD :: LHsDecl GhcPs -> [LHsDecl GhcPs] -> (LHsDecl GhcPs) +replaceDeclsPatBindD (L l (ValD x d)) newDecls = + let + (L _ d') = replaceDeclsPatBind (L l d) newDecls + in (L l (ValD x d')) replaceDeclsPatBindD x _ = error $ "replaceDeclsPatBindD called for:" ++ showGhc x -- | Replace the immediate declarations for a 'PatBind'. This -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBind' \/ 'replaceDeclsPatBind' is -- idempotent. -replaceDeclsPatBind :: (Monad m) => LHsBind GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsBind GhcPs) +replaceDeclsPatBind :: LHsBind GhcPs -> [LHsDecl GhcPs] -> (LHsBind GhcPs) replaceDeclsPatBind (L l (PatBind x a p (GRHSs xr rhss binds))) newDecls - = do - logTr "replaceDecls PatBind" - binds'' <- replaceDeclsValbinds WithWhere binds newDecls - return (L l (PatBind x a p (GRHSs xr rhss binds''))) + = (L l (PatBind x a p (GRHSs xr rhss binds''))) + where + binds'' = replaceDeclsValbinds WithWhere binds newDecls replaceDeclsPatBind x _ = error $ "replaceDeclsPatBind called for:" ++ showGhc x -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Stmt GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (LetStmt _ lb)) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (LetStmt _ lb)) = hsDeclsLocalBinds lb hsDecls (L _ (LastStmt _ e _ _)) = hsDecls e hsDecls (L _ (BindStmt _ _pat e)) = hsDecls e hsDecls (L _ (BodyStmt _ e _ _)) = hsDecls e - hsDecls _ = return [] + hsDecls _ = [] replaceDecls (L l (LetStmt x lb)) newDecls - = do - lb'' <- replaceDeclsValbinds WithWhere lb newDecls - return (L l (LetStmt x lb'')) + = let + lb'' = replaceDeclsValbinds WithWhere lb newDecls + in (L l (LetStmt x lb'')) replaceDecls (L l (LastStmt x e d se)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (LastStmt x e' d se)) + = let + e' = replaceDecls e newDecls + in (L l (LastStmt x e' d se)) replaceDecls (L l (BindStmt x pat e)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BindStmt x pat e')) + = let + e' = replaceDecls e newDecls + in (L l (BindStmt x pat e')) replaceDecls (L l (BodyStmt x e a b)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BodyStmt x e' a b)) - replaceDecls x _newDecls = return x + = let + e' = replaceDecls e newDecls + in (L l (BodyStmt x e' a b)) + replaceDecls x _newDecls = x -- ===================================================================== -- end of HasDecls instances @@ -1062,61 +1059,55 @@ data WithWhere = WithWhere -- care, as this does not manage the declaration order, the -- ordering should be done by the calling function from the 'HsLocalBinds' -- context in the AST. -replaceDeclsValbinds :: (Monad m) - => WithWhere +replaceDeclsValbinds :: WithWhere -> HsLocalBinds GhcPs -> [LHsDecl GhcPs] - -> TransformT m (HsLocalBinds GhcPs) -replaceDeclsValbinds _ _ [] = do - return (EmptyLocalBinds NoExtField) + -> HsLocalBinds GhcPs +replaceDeclsValbinds _ _ [] = EmptyLocalBinds NoExtField replaceDeclsValbinds w b@(HsValBinds a _) new - = do - logTr "replaceDeclsValbinds" - let oldSpan = spanHsLocaLBinds b - an <- oldWhereAnnotation a w (realSrcSpan oldSpan) - let decs = concatMap decl2Bind new - let sigs = concatMap decl2Sig new - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) + = let + oldSpan = spanHsLocaLBinds b + an = oldWhereAnnotation a w (realSrcSpan oldSpan) + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) replaceDeclsValbinds _ (HsIPBinds {}) _new = error "undefined replaceDecls HsIPBinds" replaceDeclsValbinds w (EmptyLocalBinds _) new - = do - logTr "replaceDecls HsLocalBinds" - an <- newWhereAnnotation w - let newBinds = concatMap decl2Bind new - newSigs = concatMap decl2Sig new - let decs = newBinds - let sigs = newSigs - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) - -oldWhereAnnotation :: (Monad m) - => EpAnn AnnList -> WithWhere -> RealSrcSpan -> TransformT m (EpAnn AnnList) -oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = do - -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, change the AnnList anchor to have the correct DP too - let (AnnList ancl o c _r t) = an - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - (anc', ancl') <- do - case ww of - WithWhere -> return (anc, ancl) - WithoutWhere -> return (anc, ancl) - let an' = EpAnn anc' - (AnnList ancl' o c w t) - cs - return an' - -newWhereAnnotation :: (Monad m) => WithWhere -> TransformT m (EpAnn AnnList) -newWhereAnnotation ww = do - let anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] - let anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - let an = EpAnn anc - (AnnList (Just anc2) Nothing Nothing w []) - emptyComments - return an + = let + an = newWhereAnnotation w + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) + +oldWhereAnnotation :: EpAnn AnnList -> WithWhere -> RealSrcSpan -> (EpAnn AnnList) +oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = an' + -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, + -- change the AnnList anchor to have the correct DP too + where + (AnnList ancl o c _r t) = an + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + (anc', ancl') = + case ww of + WithWhere -> (anc, ancl) + WithoutWhere -> (anc, ancl) + an' = EpAnn anc' + (AnnList ancl' o c w t) + cs + +newWhereAnnotation :: WithWhere -> (EpAnn AnnList) +newWhereAnnotation ww = an + where + anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] + anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + an = EpAnn anc + (AnnList (Just anc2) Nothing Nothing w []) + emptyComments -- --------------------------------------------------------------------- @@ -1127,32 +1118,32 @@ type PMatch = LMatch GhcPs (LHsExpr GhcPs) -- declarations are extracted and returned after modification. For a -- 'FunBind' the supplied 'SrcSpan' is used to identify the specific -- 'Match' to be transformed, for when there are multiple of them. -modifyValD :: forall m t. (HasTransform m) - => SrcSpan +modifyValD :: forall t. + SrcSpan -> Decl - -> (PMatch -> [Decl] -> m ([Decl], Maybe t)) - -> m (Decl,Maybe t) + -> (PMatch -> [Decl] -> ([Decl], Maybe t)) + -> (Decl,Maybe t) modifyValD p pb@(L ss (ValD _ (PatBind {} ))) f = if (locA ss) == p - then do - let ds = hsDeclsPatBindD pb - (ds',r) <- f (error "modifyValD.PatBind should not touch Match") ds - pb' <- liftT $ replaceDeclsPatBindD pb ds' - return (pb',r) - else return (pb,Nothing) -modifyValD p decl f = do - (decl',r) <- runStateT (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing - return (packFunDecl decl',r) + then + let + ds = hsDeclsPatBindD pb + (ds',r) = f (error "modifyValD.PatBind should not touch Match") ds + pb' = replaceDeclsPatBindD pb ds' + in (pb',r) + else (pb,Nothing) +modifyValD p decl f = (packFunDecl decl', r) where - doModLocal :: PMatch -> StateT (Maybe t) m PMatch + (decl',r) = runState (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing + doModLocal :: PMatch -> State (Maybe t) PMatch doModLocal (match@(L ss _) :: PMatch) = do if (locA ss) == p then do - ds <- lift $ liftT $ hsDecls match - `debug` ("modifyValD: match=" ++ showAst match) - (ds',r) <- lift $ f match ds - put r - match' <- lift $ liftT $ replaceDecls match ds' + let + ds = hsDecls match + (ds',r0) = f match ds + put r0 + let match' = replaceDecls match ds' return match' else return match @@ -1172,6 +1163,6 @@ modifyDeclsT :: (HasDecls t,HasTransform m) => ([LHsDecl GhcPs] -> m [LHsDecl GhcPs]) -> t -> m t modifyDeclsT action t = do - decls <- liftT $ hsDecls t + let decls = hsDecls t decls' <- action decls - liftT $ replaceDecls t decls' + return $ replaceDecls t decls' ===================================== utils/check-exact/Types.hs ===================================== @@ -21,10 +21,6 @@ type Pos = (Int,Int) -- --------------------------------------------------------------------- -data Rigidity = NormalLayout | RigidLayout deriving (Eq, Ord, Show) - --- --------------------------------------------------------------------- - -- | A Haskell comment. The @AnnKeywordId@ is present if it has been converted -- from an @AnnKeywordId@ because the annotation must be interleaved into the -- stream and does not have a well-defined position ===================================== utils/check-exact/Utils.hs ===================================== @@ -20,9 +20,8 @@ module Utils where import Control.Monad (when) +import GHC.Utils.Monad.State.Strict import Data.Function -import Data.Maybe (isJust) -import Data.Ord (comparing) import GHC.Hs.Dump import Lookup @@ -36,9 +35,10 @@ import GHC.Driver.Ppr import GHC.Data.FastString import qualified GHC.Data.Strict as Strict import GHC.Base (NonEmpty(..)) +import GHC.Parser.Lexer (allocateComments) import Data.Data hiding ( Fixity ) -import Data.List (sortBy, elemIndex) +import Data.List (sortBy, partition) import qualified Data.Map.Strict as Map import Debug.Trace @@ -60,12 +60,32 @@ debug c s = if debugEnabledFlag debugM :: Monad m => String -> m () debugM s = when debugEnabledFlag $ traceM s --- --------------------------------------------------------------------- - warn :: c -> String -> c -- warn = flip trace warn c _ = c +-- --------------------------------------------------------------------- + +captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag +captureOrderBinds ls = AnnSortKey $ map go ls + where + go (L _ (ValD _ _)) = BindTag + go (L _ (SigD _ _)) = SigDTag + go d = error $ "captureOrderBinds:" ++ showGhc d + +-- --------------------------------------------------------------------- + +notDocDecl :: LHsDecl GhcPs -> Bool +notDocDecl (L _ DocD{}) = False +notDocDecl _ = True + +notIEDoc :: LIE GhcPs -> Bool +notIEDoc (L _ IEGroup {}) = False +notIEDoc (L _ IEDoc {}) = False +notIEDoc (L _ IEDocNamed {}) = False +notIEDoc _ = True + +-- --------------------------------------------------------------------- -- | A good delta has no negative values. isGoodDelta :: DeltaPos -> Bool isGoodDelta (SameLine co) = co >= 0 @@ -108,7 +128,6 @@ pos2delta (refl,refc) (l,c) = deltaPos lo co lo = l - refl co = if lo == 0 then c - refc else c - -- else c - 1 -- | Apply the delta to the current position, taking into account the -- current column offset if advancing to a new line @@ -200,23 +219,6 @@ origDelta pos pp = ss2delta (ss2posEnd pp) pos -- --------------------------------------------------------------------- --- |Given a list of items and a list of keys, returns a list of items --- ordered by their position in the list of keys. -orderByKey :: [(DeclTag,a)] -> [DeclTag] -> [(DeclTag,a)] -orderByKey keys order - -- AZ:TODO: if performance becomes a problem, consider a Map of the order - -- SrcSpan to an index, and do a lookup instead of elemIndex. - - -- Items not in the ordering are placed to the start - = sortBy (comparing (flip elemIndex order . fst)) keys - --- --------------------------------------------------------------------- - -isListComp :: HsDoFlavour -> Bool -isListComp = isDoComprehensionContext - --- --------------------------------------------------------------------- - needsWhere :: DataDefnCons (LConDecl (GhcPass p)) -> Bool needsWhere (NewTypeCon _) = True needsWhere (DataTypeCons _ []) = True @@ -225,21 +227,214 @@ needsWhere _ = False -- --------------------------------------------------------------------- +-- | Insert the comments at the appropriate places in the AST insertCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource -insertCppComments (L l p) cs = L l p' +-- insertCppComments p [] = p +insertCppComments (L l p) cs0 = insertRemainingCppComments (L l p2) remaining + where + (EpAnn anct ant cst) = hsmodAnn $ hsmodExt p + cs = sortEpaComments $ priorComments cst ++ getFollowingComments cst ++ cs0 + p0 = p { hsmodExt = (hsmodExt p) { hsmodAnn = EpAnn anct ant emptyComments }} + -- Comments embedded within spans + -- everywhereM is a bottom-up traversal + (p1, toplevel) = runState (everywhereM (mkM addCommentsListItem + `extM` addCommentsGrhs + `extM` addCommentsList) p0) cs + (p2, remaining) = insertTopLevelCppComments p1 toplevel + + addCommentsListItem :: EpAnn AnnListItem -> State [LEpaComment] (EpAnn AnnListItem) + addCommentsListItem = addComments + + addCommentsList :: EpAnn AnnList -> State [LEpaComment] (EpAnn AnnList) + addCommentsList = addComments + + addCommentsGrhs :: EpAnn GrhsAnn -> State [LEpaComment] (EpAnn GrhsAnn) + addCommentsGrhs = addComments + + addComments :: forall ann. EpAnn ann -> State [LEpaComment] (EpAnn ann) + addComments (EpAnn anc an ocs) = do + case anc of + EpaSpan (RealSrcSpan s _) -> do + unAllocated <- get + let + (rest, these) = GHC.Parser.Lexer.allocateComments s unAllocated + cs' = workInComments ocs these + put rest + return $ EpAnn anc an cs' + + _ -> return $ EpAnn anc an ocs + +workInComments :: EpAnnComments -> [LEpaComment] -> EpAnnComments +workInComments ocs [] = ocs +workInComments ocs new = cs' + where + pc = priorComments ocs + fc = getFollowingComments ocs + cs' = case fc of + [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ new + (L ac _:_) -> epaCommentsBalanced (sortEpaComments $ pc ++ cs_before) + (sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) + new + +insertTopLevelCppComments :: HsModule GhcPs -> [LEpaComment] -> (HsModule GhcPs, [LEpaComment]) +insertTopLevelCppComments (HsModule (XModulePs an lo mdeprec mbDoc) mmn mexports imports decls) cs + = (HsModule (XModulePs an4 lo mdeprec mbDoc) mmn mexports' imports' decls', cs3) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0,hc1,hc_cs)) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0i,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0i,hc0,hc1,hc_cs)) + where + -- Comments at the top level. + (an0, cs0) = + case mmn of + Nothing -> (an, cs) + Just _ -> + -- We have a module name. Capture all comments up to the `where` + let + (these, remaining) = splitOnWhere Before (am_main $ anns an) cs + (EpAnn a anno ocs) = an :: EpAnn AnnsModule + anm = EpAnn a anno (workInComments ocs these) + in + (anm, remaining) + (an1,cs0a) = case lo of + EpExplicitBraces (EpTok (EpaSpan (RealSrcSpan s _))) _close -> + let + (stay,cs0a') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0 + cs' = workInComments (comments an0) stay + in (an0 { comments = cs' }, cs0a') + _ -> (an0,cs0) + -- Deal with possible leading semis + (an2, cs0b) = case am_decls $ anns an1 of + (AddSemiAnn (EpaSpan (RealSrcSpan s _)):_) -> (an1 {comments = cs'}, cs0b') + where + (stay,cs0b') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0a + cs' = workInComments (comments an1) stay + _ -> (an1,cs0a) + + (mexports', an3, cs1) = + case mexports of + Nothing -> (Nothing, an2, cs0b) + Just (L l exports) -> (Just (L l exports'), an3', cse) + where + hc1' = workInComments (comments an2) csh' + an3' = an2 { comments = hc1' } + (csh', cs0b') = case al_open $ anns l of + Just (AddEpAnn _ (EpaSpan (RealSrcSpan s _))) ->(h, n) + where + (h,n) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + cs0b + + _ -> ([], cs0b) + (exports', cse) = allocPreceding exports cs0b' + (imports0, cs2) = allocPreceding imports cs1 + (imports', hc0i) = balanceFirstLocatedAComments imports0 + + (decls0, cs3) = allocPreceding decls cs2 + (decls', hc0d) = balanceFirstLocatedAComments decls0 + + -- Either hc0i or hc0d should have comments. Combine them + hc0 = hc0i ++ hc0d + + (hc1,hc_cs) = if null ( am_main $ anns an3) + then (hc0,[]) + else splitOnWhere After (am_main $ anns an3) hc0 + hc2 = workInComments (comments an3) hc1 + an4 = an3 { anns = (anns an3) {am_cs = hc_cs}, comments = hc2 } + + allocPreceding :: [LocatedA a] -> [LEpaComment] -> ([LocatedA a], [LEpaComment]) + allocPreceding [] cs' = ([], cs') + allocPreceding (L (EpAnn anc4 an5 cs4) a:xs) cs' = ((L (EpAnn anc4 an5 cs4') a:xs'), rest') + where + (rest, these) = + case anc4 of + EpaSpan (RealSrcSpan s _) -> + allocatePriorComments (ss2pos s) cs' + _ -> (cs', []) + cs4' = workInComments cs4 these + (xs',rest') = allocPreceding xs rest + +data SplitWhere = Before | After +splitOnWhere :: SplitWhere -> [AddEpAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitOnWhere _ [] csIn = (csIn,[]) +splitOnWhere w (AddEpAnn AnnWhere (EpaSpan (RealSrcSpan s _)):_) csIn = (hc, fc) + where + splitFunc Before anc_pos c_pos = c_pos < anc_pos + splitFunc After anc_pos c_pos = anc_pos < c_pos + (hc,fc) = break (\(L ll _) -> splitFunc w (ss2pos $ anchor ll) (ss2pos s)) csIn +splitOnWhere _ (AddEpAnn AnnWhere _:_) csIn = (csIn, []) +splitOnWhere f (_:as) csIn = splitOnWhere f as csIn + +balanceFirstLocatedAComments :: [LocatedA a] -> ([LocatedA a], [LEpaComment]) +balanceFirstLocatedAComments [] = ([],[]) +balanceFirstLocatedAComments ((L (EpAnn anc an csd) a):ds) = (L (EpAnn anc an csd0) a:ds, hc') + where + (csd0, hc') = case anc of + EpaSpan (RealSrcSpan s _) -> (csd', hc) + `debug` ("balanceFirstLocatedAComments: (csd,csd',attached,header)=" ++ showAst (csd,csd',attached,header)) + where + (priors, inners) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + (priorComments csd) + pcds = priorCommentsDeltas' s priors + (attached, header) = break (\(d,_c) -> d /= 1) pcds + csd' = setPriorComments csd (reverse (map snd attached) ++ inners) + hc = reverse (map snd header) + _ -> (csd, []) + + + +priorCommentsDeltas' :: RealSrcSpan -> [LEpaComment] + -> [(Int, LEpaComment)] +priorCommentsDeltas' r cs = go r (reverse cs) where - an' = case GHC.hsmodAnn $ GHC.hsmodExt p of - (EpAnn a an ocs) -> EpAnn a an cs' - where - pc = priorComments ocs - fc = getFollowingComments ocs - cs' = case fc of - [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ cs - (L ac _:_) -> EpaCommentsBalanced (sortEpaComments $ pc ++ cs_before) - (sortEpaComments $ fc ++ cs_after) - where - (cs_before,cs_after) = break (\(L ll _) -> (ss2pos $ anchor ll) < (ss2pos $ anchor ac) ) cs + go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] + go _ [] = [] + go _ (la@(L l@(EpaDelta _ dp _) _):las) = (deltaLine dp, la) : go (anchor l) las + go rs' (la@(L l _):las) = deltaComment rs' la : go (anchor l) las + + deltaComment :: RealSrcSpan -> LEpaComment -> (Int, LEpaComment) + deltaComment rs' (L loc c) = (abs(ll - al), L loc c) + where + (al,_) = ss2pos rs' + (ll,_) = ss2pos (anchor loc) + +allocatePriorComments + :: Pos + -> [LEpaComment] + -> ([LEpaComment], [LEpaComment]) +allocatePriorComments ss_loc comment_q = + let + cmp (L l _) = ss2pos (anchor l) <= ss_loc + (newAnns,after) = partition cmp comment_q + in + (after, newAnns) + +insertRemainingCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource +insertRemainingCppComments (L l p) cs = L l p' + -- `debug` ("insertRemainingCppComments: (cs,an')=" ++ showAst (cs,an')) + where + (EpAnn a an ocs) = GHC.hsmodAnn $ GHC.hsmodExt p + an' = EpAnn a an (addTrailingComments end_loc ocs cs) p' = p { GHC.hsmodExt = (GHC.hsmodExt p) { GHC.hsmodAnn = an' } } + end_loc = case GHC.hsmodLayout $ GHC.hsmodExt p of + EpExplicitBraces _open close -> case close of + EpTok (EpaSpan (RealSrcSpan s _)) -> ss2pos s + _ -> (1,1) + _ -> (1,1) + (new_before, new_after) = break (\(L ll _) -> (ss2pos $ anchor ll) > end_loc ) cs + + addTrailingComments end_loc' cur new = epaCommentsBalanced pc' fc' + where + pc = priorComments cur + fc = getFollowingComments cur + (pc', fc') = case reverse pc of + [] -> (sortEpaComments $ pc ++ new_before, sortEpaComments $ fc ++ new_after) + (L ac _:_) -> (sortEpaComments $ pc ++ cs_before, sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = if (ss2pos $ anchor ac) > end_loc' + then break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) new + else (new_before, new_after) -- --------------------------------------------------------------------- @@ -291,11 +486,16 @@ dedentDocChunkBy dedent (L (RealSrcSpan l mb) c) = L (RealSrcSpan l' mb) c dedentDocChunkBy _ x = x + +epaCommentsBalanced :: [LEpaComment] -> [LEpaComment] -> EpAnnComments +epaCommentsBalanced priorCs [] = EpaComments priorCs +epaCommentsBalanced priorCs postCs = EpaCommentsBalanced priorCs postCs + mkEpaComments :: [Comment] -> [Comment] -> EpAnnComments mkEpaComments priorCs [] = EpaComments (map comment2LEpaComment priorCs) mkEpaComments priorCs postCs - = EpaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) + = epaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) comment2LEpaComment :: Comment -> LEpaComment comment2LEpaComment (Comment s anc r _mk) = mkLEpaComment s anc r @@ -330,18 +530,11 @@ sortEpaComments cs = sortBy cmp cs mkKWComment :: AnnKeywordId -> NoCommentsLocation -> Comment mkKWComment kw (EpaSpan (RealSrcSpan ss mb)) = Comment (keywordToString kw) (EpaSpan (RealSrcSpan ss mb)) ss (Just kw) -mkKWComment kw (EpaSpan ss@(UnhelpfulSpan _)) - = Comment (keywordToString kw) (EpaDelta ss (SameLine 0) NoComments) placeholderRealSpan (Just kw) +mkKWComment kw (EpaSpan (UnhelpfulSpan _)) + = Comment (keywordToString kw) (EpaDelta noSrcSpan (SameLine 0) NoComments) placeholderRealSpan (Just kw) mkKWComment kw (EpaDelta ss dp cs) = Comment (keywordToString kw) (EpaDelta ss dp cs) placeholderRealSpan (Just kw) --- | Detects a comment which originates from a specific keyword. -isKWComment :: Comment -> Bool -isKWComment c = isJust (commentOrigin c) - -noKWComments :: [Comment] -> [Comment] -noKWComments = filter (\c -> not (isKWComment c)) - sortAnchorLocated :: [GenLocated Anchor a] -> [GenLocated Anchor a] sortAnchorLocated = sortBy (compare `on` (anchor . getLoc)) @@ -379,11 +572,6 @@ name2String = showPprUnsafe -- --------------------------------------------------------------------- -locatedAnAnchor :: LocatedAn a t -> RealSrcSpan -locatedAnAnchor (L (EpAnn a _ _) _) = anchor a - --- --------------------------------------------------------------------- - trailingAnnLoc :: TrailingAnn -> EpaLocation trailingAnnLoc (AddSemiAnn ss) = ss trailingAnnLoc (AddCommaAnn ss) = ss @@ -403,52 +591,6 @@ addEpAnnLoc (AddEpAnn _ l) = l -- --------------------------------------------------------------------- --- TODO: get rid of this identity function -anchorToEpaLocation :: Anchor -> EpaLocation -anchorToEpaLocation a = a - --- --------------------------------------------------------------------- --- Horrible hack for dealing with some things still having a SrcSpan, --- not an Anchor. - -{- -A SrcSpan is defined as - -data SrcSpan = - RealSrcSpan !RealSrcSpan !(Maybe BufSpan) -- See Note [Why Maybe BufPos] - | UnhelpfulSpan !UnhelpfulSpanReason - -data BufSpan = - BufSpan { bufSpanStart, bufSpanEnd :: {-# UNPACK #-} !BufPos } - deriving (Eq, Ord, Show) - -newtype BufPos = BufPos { bufPos :: Int } - - -We use the BufPos to encode a delta, using bufSpanStart for the line, -and bufSpanEnd for the col. - -To be absolutely sure, we make the delta versions use -ve values. - --} - -hackSrcSpanToAnchor :: SrcSpan -> Anchor -hackSrcSpanToAnchor (UnhelpfulSpan s) = error $ "hackSrcSpanToAnchor : UnhelpfulSpan:" ++ show s -hackSrcSpanToAnchor ss@(RealSrcSpan r mb) - = case mb of - (Strict.Just (BufSpan (BufPos s) (BufPos e))) -> - if s <= 0 && e <= 0 - then EpaDelta ss (deltaPos (-s) (-e)) [] - `debug` ("hackSrcSpanToAnchor: (r,s,e)=" ++ showAst (r,s,e) ) - else EpaSpan (RealSrcSpan r mb) - _ -> EpaSpan (RealSrcSpan r mb) - -hackAnchorToSrcSpan :: Anchor -> SrcSpan -hackAnchorToSrcSpan (EpaSpan s) = s -hackAnchorToSrcSpan _ = error $ "hackAnchorToSrcSpan" - --- --------------------------------------------------------------------- - type DeclsByTag a = Map.Map DeclTag [(RealSrcSpan, a)] orderedDecls View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/42c9560074c03c74e96c58f792c0d281d73e0eea -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/42c9560074c03c74e96c58f792c0d281d73e0eea You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 20:07:48 2024 From: gitlab at gitlab.haskell.org (Bodigrim (@Bodigrim)) Date: Tue, 10 Sep 2024 16:07:48 -0400 Subject: [Git][ghc/ghc][wip/deepseq-1.5.1.0] Bump submodule deepseq to 1.5.1.0 Message-ID: <66e0a7148c37e_29b94c75e720414f6@gitlab.mail> Bodigrim pushed to branch wip/deepseq-1.5.1.0 at Glasgow Haskell Compiler / GHC Commits: 680b4b89 by Andrew Lelechenko at 2024-09-10T21:07:38+01:00 Bump submodule deepseq to 1.5.1.0 - - - - - 2 changed files: - compiler/GHC/Unit/Types.hs - libraries/deepseq Changes: ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -101,9 +101,9 @@ import GHC.Utils.Fingerprint import GHC.Utils.Misc import GHC.Settings.Config (cProjectUnitId) -import Control.DeepSeq +import Control.DeepSeq (NFData(..)) import Data.Data -import Data.List (sortBy ) +import Data.List (sortBy) import Data.Function import Data.Bifunctor import qualified Data.ByteString as BS ===================================== libraries/deepseq ===================================== @@ -1 +1 @@ -Subproject commit 09aed1bf774f2f05c8b390539ce35adf5cd68c30 +Subproject commit 7ce6e2d3760b23336fd5f9a36f50df6571606947 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/680b4b894d59237cf136391c664b98dfe3b75e5a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/680b4b894d59237cf136391c664b98dfe3b75e5a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 20:59:36 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 10 Sep 2024 16:59:36 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 21 commits: Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Message-ID: <66e0b338e246e_29b94c9e4df0443b1@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - c6f4c81c by Matthew Pickering at 2024-09-10T15:50:05+01:00 Run on test-abi label - - - - - c41aaf44 by Rodrigo Mesquita at 2024-09-10T15:50:06+01:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - 03147966 by Rodrigo Mesquita at 2024-09-10T17:50:54+01:00 Deterministic Uniques in the NCG See Note [Deterministic Uniques in the NCG] UniqDSM det uniques + use in Cmm.Info Now for SRTs SRT generation using deterministic uniq supply Back LabelMap with deterministic UDFM TSAN uniq rename hard Revert "TSAN uniq rename hard" This reverts commit 7ca5ab3036c15f38c6d4cbcb616d415958c6bcda. improvements to uniqdsm UniqDSM ProcPoint CmmLayoutStack UniqDet 90% of cpsTop UniqDSM Major progress in using UniqDSM in CmmToAsm and Ncg backends Fix imports Un-back label map with udfm Revert "Un-back label map with udfm" This reverts commit f5d2e4257214a3f7b7d845651e6662c5babfd6a3. Make UDSM oneshot deriving via state Fix -fllvm hang - - - - - 5713ab47 by Rodrigo Mesquita at 2024-09-10T19:14:18+01:00 start note - - - - - b13ea17f by Rodrigo Mesquita at 2024-09-10T19:15:17+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 39071432 by Rodrigo Mesquita at 2024-09-10T19:15:18+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - 6d148b3e by Rodrigo Mesquita at 2024-09-10T19:15:18+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 39dac004 by Rodrigo Mesquita at 2024-09-10T19:15:19+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 3b1a97f5 by Rodrigo Mesquita at 2024-09-10T19:15:19+01:00 determinism: Rename CmmGroups in generateCgIPEStub Make sure to also deterministically rename the IPE Cmm (as per Note [Renaming uniques deterministically]) to guarantee deterministic objects when IPE information is requested. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Reg/Graph.hs - compiler/GHC/CmmToAsm/Reg/Graph/Spill.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/90b8bdb0b3bd298c31e48694546bf6ca2d2720b3...3b1a97f53b07fc0bd2642ec46a9ae0f9ee8185c1 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/90b8bdb0b3bd298c31e48694546bf6ca2d2720b3...3b1a97f53b07fc0bd2642ec46a9ae0f9ee8185c1 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 23:00:56 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Tue, 10 Sep 2024 19:00:56 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] 169 commits: compiler: Turn `FinderCache` into a record of operations so that GHC API clients can Message-ID: <66e0cfa8a4dcc_27f30327d82851784@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - ae64a31e by doyougnu at 2024-09-10T19:00:24-04:00 linker: add 0xDEADBEEF upstream from haskell.nix patches - - - - - 28 changed files: - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - compiler/GHC.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/GenericOpt.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Reg/Graph/TrivColorable.hs - compiler/GHC/CmmToAsm/Reg/Linear/State.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/Core/ConLike.hs - compiler/GHC/Core/DataCon.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/93a982ffe72c00e4b43caf6be6e69da80043b3ea...ae64a31e00d266934b1656ffc39353cc1f78bbe7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/93a982ffe72c00e4b43caf6be6e69da80043b3ea...ae64a31e00d266934b1656ffc39353cc1f78bbe7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 23:20:02 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 19:20:02 -0400 Subject: [Git][ghc/ghc][master] 3 commits: haddock: Re-organise cross-OS compatibility layer Message-ID: <66e0d42277f7d_27f303530174591fe@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - 8 changed files: - + utils/haddock/haddock-api/compat/posix/Haddock/Compat.hs - + utils/haddock/haddock-api/compat/windows/Haddock/Compat.hs - utils/haddock/haddock-api/haddock-api.cabal - utils/haddock/haddock-api/src/Haddock.hs - utils/haddock/haddock-api/src/Haddock/Interface.hs - utils/haddock/haddock-api/src/Haddock/Types.hs - utils/haddock/haddock-api/src/Haddock/Utils.hs - utils/haddock/haddock-test/src/Test/Haddock/Config.hs Changes: ===================================== utils/haddock/haddock-api/compat/posix/Haddock/Compat.hs ===================================== @@ -0,0 +1,14 @@ +module Haddock.Compat + ( getProcessID + , setEncoding + ) where + +import System.Posix.Types (ProcessID) +import qualified System.Posix.Process as Posix + +-- | Windows-only failsafe, not applicable on POSIX plateforms +setEncoding :: IO () +setEncoding = pure () + +getProcessID :: IO Int +getProcessID = fromIntegral @ProcessID @Int <$> Posix.getProcessID ===================================== utils/haddock/haddock-api/compat/windows/Haddock/Compat.hs ===================================== @@ -0,0 +1,19 @@ +module Haddock.Compat + ( getProcessID + , setEncoding + ) where + +import GHC.IO.Encoding.CodePage (mkLocaleEncoding) +import GHC.IO.Encoding.Failure (CodingFailureMode(TransliterateCodingFailure)) +import System.IO (hSetEncoding, stdout, stderr) +import System.Win32.Process (ProcessId) +import qualified System.Win32.Process as Windows + +-- | Avoid internal error: : hPutChar: invalid argument (invalid character)' non UTF-8 Windows +setEncoding :: IO () +setEncoding = do + hSetEncoding stdout $ mkLocaleEncoding TransliterateCodingFailure + hSetEncoding stderr $ mkLocaleEncoding TransliterateCodingFailure + +getProcessID :: IO Int +getProcessID = fromIntegral @ProcessId @Int <$> Windows.getCurrentProcessId ===================================== utils/haddock/haddock-api/haddock-api.cabal ===================================== @@ -14,7 +14,7 @@ copyright: (c) Simon Marlow, David Waern category: Documentation build-type: Simple -extra-source-files: +extra-doc-files: CHANGES.md data-dir: @@ -99,6 +99,16 @@ library , transformers hs-source-dirs: src + + if os(windows) + hs-source-dirs: compat/windows/ + build-depends: + Win32 + else + hs-source-dirs: compat/posix/ + build-depends: + unix + exposed-modules: Documentation.Haddock @@ -142,6 +152,7 @@ library Haddock.Options Haddock.GhcUtils Haddock.Convert + Haddock.Compat Paths_haddock_api autogen-modules: ===================================== utils/haddock/haddock-api/src/Haddock.hs ===================================== @@ -83,6 +83,7 @@ import Haddock.InterfaceFile import Haddock.Options import Haddock.Utils import Haddock.GhcUtils (modifySessionDynFlags, setOutputDir, getSupportedLanguagesAndExtensions) +import Haddock.Compat (getProcessID) -------------------------------------------------------------------------------- -- * Exception handling ===================================== utils/haddock/haddock-api/src/Haddock/Interface.hs ===================================== @@ -78,12 +78,6 @@ import GHC.Utils.Outputable (Outputable, (<+>), pprModuleName, text) import GHC.Utils.Error (withTiming) import GHC.Utils.Monad (mapMaybeM) -#if defined(mingw32_HOST_OS) -import System.IO -import GHC.IO.Encoding.CodePage (mkLocaleEncoding) -import GHC.IO.Encoding.Failure (CodingFailureMode(TransliterateCodingFailure)) -#endif - import Haddock.GhcUtils (moduleString, pretty) import Haddock.Interface.AttachInstances (attachInstances) import Haddock.Interface.Create (createInterface1, createInterface1') @@ -92,6 +86,7 @@ import Haddock.InterfaceFile (InterfaceFile, ifInstalledIfaces, ifLinkEnv) import Haddock.Options hiding (verbosity) import Haddock.Types import Haddock.Utils (Verbosity (..), normal, out, verbose) +import qualified Haddock.Compat as Compat -- | Create 'Interface's and a link environment by typechecking the list of -- modules using the GHC API and processing the resulting syntax trees. @@ -104,12 +99,7 @@ processModules -> Ghc ([Interface], LinkEnv) -- ^ Resulting list of interfaces and renaming -- environment processModules verbosity modules flags extIfaces = do -#if defined(mingw32_HOST_OS) - -- Avoid internal error: : hPutChar: invalid argument (invalid character)' non UTF-8 Windows - liftIO $ hSetEncoding stdout $ mkLocaleEncoding TransliterateCodingFailure - liftIO $ hSetEncoding stderr $ mkLocaleEncoding TransliterateCodingFailure -#endif - + liftIO Compat.setEncoding dflags <- getDynFlags -- Map from a module to a corresponding installed interface ===================================== utils/haddock/haddock-api/src/Haddock/Types.hs ===================================== @@ -605,13 +605,6 @@ instance DocHeader a -> a `deepseq` () DocTable a -> a `deepseq` () -#if !MIN_VERSION_ghc(8,0,2) --- These were added to GHC itself in 8.0.2 -instance NFData Name where rnf x = seq x () -instance NFData OccName where rnf x = seq x () -instance NFData ModuleName where rnf x = seq x () -#endif - instance NFData id => NFData (Header id) where rnf (Header a b) = a `deepseq` b `deepseq` () ===================================== utils/haddock/haddock-api/src/Haddock/Utils.hs ===================================== @@ -62,9 +62,6 @@ module Haddock.Utils , verbose , deafening , out - - -- * System tools - , getProcessID ) where import Control.Monad.Catch (MonadMask, bracket_) @@ -84,10 +81,6 @@ import qualified System.FilePath.Posix as HtmlPath import System.IO (IOMode (..), hPutStr, hSetEncoding, utf8, withFile) import System.IO.Unsafe (unsafePerformIO) -#ifndef mingw32_HOST_OS -import qualified System.Posix.Internals -#endif - import Documentation.Haddock.Doc (emptyMetaDoc) import Haddock.Types @@ -345,16 +338,3 @@ spanWith _ [] = ([], []) spanWith p xs@(a : as) | Just b <- p a = let (bs, cs) = spanWith p as in (b : bs, cs) | otherwise = ([], xs) - ------------------------------------------------------------------------------ - --- * System tools - ------------------------------------------------------------------------------ - -#ifdef mingw32_HOST_OS -foreign import ccall unsafe "_getpid" getProcessID :: IO Int -- relies on Int == Int32 on Windows -#else -getProcessID :: IO Int -getProcessID = fmap fromIntegral System.Posix.Internals.c_getpid -#endif ===================================== utils/haddock/haddock-test/src/Test/Haddock/Config.hs ===================================== @@ -257,11 +257,7 @@ baseDependencies ghcPath = do (comp, _, cfg) <- configure normal (Just ghcPath) Nothing defaultProgramDb -#if MIN_VERSION_Cabal(1,23,0) pkgIndex <- getInstalledPackages normal comp [GlobalPackageDB] cfg -#else - pkgIndex <- getInstalledPackages normal [GlobalPackageDB] cfg -#endif let pkgs = [ "array" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/057159947ae88f6bad703cc357913d8ca60384d0...370d1599ac2c68fc305cef4c29b4b4ecbf8e7969 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/057159947ae88f6bad703cc357913d8ca60384d0...370d1599ac2c68fc305cef4c29b4b4ecbf8e7969 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 23:20:48 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 19:20:48 -0400 Subject: [Git][ghc/ghc][master] Add ZonkAny and document it Message-ID: <66e0d450e9707_27f3036cd0686421@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 13 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Tc/Types.hs - compiler/GHC/Tc/Utils/Monad.hs - compiler/GHC/Tc/Zonk/Type.hs - testsuite/tests/perf/compiler/T11068.stdout - testsuite/tests/pmcheck/should_compile/T12957.stderr - + testsuite/tests/pmcheck/should_compile/T24817.hs - testsuite/tests/pmcheck/should_compile/all.T - testsuite/tests/printer/T17697.stderr - testsuite/tests/profiling/should_run/staticcallstack002.stdout - testsuite/tests/simplCore/should_compile/T13156.stdout - testsuite/tests/typecheck/should_fail/T13292.stderr Changes: ===================================== compiler/GHC/Builtin/Names.hs ===================================== @@ -2007,6 +2007,12 @@ uWordTyConKey = mkPreludeTyConUnique 163 unsatisfiableClassNameKey :: Unique unsatisfiableClassNameKey = mkPreludeTyConUnique 170 +anyTyConKey :: Unique +anyTyConKey = mkPreludeTyConUnique 171 + +zonkAnyTyConKey :: Unique +zonkAnyTyConKey = mkPreludeTyConUnique 172 + -- Custom user type-errors errorMessageTypeErrorFamKey :: Unique errorMessageTypeErrorFamKey = mkPreludeTyConUnique 181 @@ -2020,9 +2026,6 @@ proxyPrimTyConKey = mkPreludeTyConUnique 184 specTyConKey :: Unique specTyConKey = mkPreludeTyConUnique 185 -anyTyConKey :: Unique -anyTyConKey = mkPreludeTyConUnique 186 - smallArrayPrimTyConKey = mkPreludeTyConUnique 187 smallMutableArrayPrimTyConKey = mkPreludeTyConUnique 188 ===================================== compiler/GHC/Builtin/Types.hs ===================================== @@ -91,7 +91,7 @@ module GHC.Builtin.Types ( cTupleSelId, cTupleSelIdName, -- * Any - anyTyCon, anyTy, anyTypeOfKind, + anyTyCon, anyTy, anyTypeOfKind, zonkAnyTyCon, -- * Recovery TyCon makeRecoveryTyCon, @@ -184,7 +184,7 @@ import GHC.Core.ConLike import GHC.Core.TyCon import GHC.Core.Class ( Class, mkClass ) import GHC.Core.Map.Type ( TypeMap, emptyTypeMap, extendTypeMap, lookupTypeMap ) -import qualified GHC.Core.TyCo.Rep as TyCoRep (Type(TyConApp)) +import qualified GHC.Core.TyCo.Rep as TyCoRep ( Type(TyConApp) ) import GHC.Types.TyThing import GHC.Types.SourceText @@ -309,6 +309,7 @@ wiredInTyCons = map (dataConTyCon . snd) boxingDataCons , soloTyCon , anyTyCon + , zonkAnyTyCon , boolTyCon , charTyCon , stringTyCon @@ -419,65 +420,105 @@ doubleDataConName = mkWiredInDataConName UserSyntax gHC_TYPES (fsLit "D#") {- Note [Any types] ~~~~~~~~~~~~~~~~ -The type constructor Any, +The type constructors `Any` and `ZonkAny` are closed type families declared thus: - type family Any :: k where { } + type family Any :: forall k. k where { } + type family ZonkAny :: forall k. Nat -> k where { } -It has these properties: +They are used when we want a type of a particular kind, but we don't really care +what that type is. The leading example is this: `ZonkAny` is used to instantiate +un-constrained type variables after type checking. For example, consider the +term (length [] :: Int), where - * Note that 'Any' is kind polymorphic since in some program we may - need to use Any to fill in a type variable of some kind other than * - (see #959 for examples). Its kind is thus `forall k. k``. + length :: forall a. [a] -> Int + [] :: forall a. [a] - * It is defined in module GHC.Types, and exported so that it is - available to users. For this reason it's treated like any other - wired-in type: - - has a fixed unique, anyTyConKey, - - lives in the global name cache +We must type-apply `length` and `[]`, but to what type? It doesn't matter! +The typechecker will end up with - * It is a *closed* type family, with no instances. This means that - if ty :: '(k1, k2) we add a given coercion - g :: ty ~ (Fst ty, Snd ty) - If Any was a *data* type, then we'd get inconsistency because 'ty' - could be (Any '(k1,k2)) and then we'd have an equality with Any on - one side and '(,) on the other. See also #9097 and #9636. + length @alpha ([] @alpha) - * When instantiated at a lifted type it is inhabited by at least one value, - namely bottom +where `alpha` is an un-constrained unification variable. The "zonking" process zaps +that unconstrained `alpha` to an arbitrary type (ZonkAny @Type 3), where the `3` is +arbitrary (see wrinkle (Any5) below). This is done in `GHC.Tc.Zonk.Type.commitFlexi`. +So we end up with - * You can safely coerce any /lifted/ type to Any and back with unsafeCoerce. - You can safely coerce any /unlifted/ type to Any and back with unsafeCoerceUnlifted. - You can coerce /any/ type to Any and back with unsafeCoerce#, but it's only safe when - the kinds of both the type and Any match. + length @(ZonkAny @Type 3) ([] @(ZonkAny @Type 3)) - For lifted/unlifted types unsafeCoerce[Unlifted] should be preferred over unsafeCoerce# - as they prevent accidentally coercing between types with kinds that don't match. +`Any` and `ZonkAny` differ only in the presence of the `Nat` argument; see +wrinkle (Any4). - See examples in ghc-prim:GHC.Types +Wrinkles: - * It does not claim to be a *data* type, and that's important for - the code generator, because the code gen may *enter* a data value - but never enters a function value. +(Any1) `Any` and `ZonkAny` are kind polymorphic since in some program we may + need to use `ZonkAny` to fill in a type variable of some kind other than * + (see #959 for examples). - * It is wired-in so we can easily refer to it where we don't have a name - environment (e.g. see Rules.matchRule for one example) +(Any2) They are /closed/ type families, with no instances. For example, suppose that + with alpha :: '(k1, k2) we add a given coercion + g :: alpha ~ (Fst alpha, Snd alpha) + and we zonked alpha = ZonkAny @(k1,k2) n. Then, if `ZonkAny` was a /data/ type, + we'd get inconsistency because we'd have a Given equality with `ZonkAny` on one + side and '(,) on the other. See also #9097 and #9636. -It's used to instantiate un-constrained type variables after type checking. For -example, 'length' has type + See #25244 for a suggestion that we instead use an /open/ type family for which + you cannot provide instances. Probably the difference is not very important. - length :: forall a. [a] -> Int +(Any3) They do not claim to be /data/ types, and that's important for + the code generator, because the code gen may /enter/ a data value + but never enters a function value. + +(Any4) `ZonkAny` takes a `Nat` argument so that we can readily make up /distinct/ + types (#24817). Consider + + data SBool a where { STrue :: SBool True; SFalse :: SBool False } + + foo :: forall a b. (SBool a, SBool b) + + bar :: Bool + bar = case foo @alpha @beta of + (STrue, SFalse) -> True -- This branch is not inaccessible! + _ -> False + + Now, what are `alpha` and `beta`? If we zonk both of them to the same type + `Any @Type`, the pattern-match checker will (wrongly) report that the first + branch is inaccessible. So we zonk them to two /different/ types: + alpha := ZonkAny @Type 4 and beta := ZonkAny @Type k 5 + (The actual numbers are arbitrary; they just need to differ.) + + The unique-name generation comes from field `tcg_zany_n` of `TcGblEnv`; and + `GHC.Tc.Zonk.Type.commitFlexi` calls `GHC.Tc.Utils.Monad.newZonkAnyType` to + make up a fresh type. + + If this example seems unconvincing (e.g. in this case foo must be bottom) + see #24817 for larger but more compelling examples. -and the list datacon for the empty list has type +(Any5) `Any` and `ZonkAny` are wired-in so we can easily refer to it where we + don't have a name environment (e.g. see Rules.matchRule for one example) - [] :: forall a. [a] +(Any6) `Any` is defined in library module ghc-prim:GHC.Types, and exported so that + it is available to users. For this reason it's treated like any other + wired-in type: + - has a fixed unique, anyTyConKey, + - lives in the global name cache + Currently `ZonkAny` is not available to users; but it could easily be. + +(Any7) Properties of `Any`: + * When `Any` is instantiated at a lifted type it is inhabited by at least one value, + namely bottom. -In order to compose these two terms as @length []@ a type -application is required, but there is no constraint on the -choice. In this situation GHC uses 'Any', + * You can safely coerce any /lifted/ type to `Any` and back with `unsafeCoerce`. -> length @(Any @Type) ([] @(Any @Type)) + * You can safely coerce any /unlifted/ type to `Any` and back with `unsafeCoerceUnlifted`. -Above, we print kinds explicitly, as if with -fprint-explicit-kinds. + * You can coerce /any/ type to `Any` and back with `unsafeCoerce#`, but it's only safe when + the kinds of both the type and `Any` match. + + * For lifted/unlifted types `unsafeCoerce[Unlifted]` should be preferred over + `unsafeCoerce#` as they prevent accidentally coercing between types with kinds + that don't match. + + See examples in ghc-prim:GHC.Types The Any tycon used to be quite magic, but we have since been able to implement it merely with an empty kind polymorphic type family. See #10886 for a @@ -490,6 +531,7 @@ anyTyConName = mkWiredInTyConName UserSyntax gHC_TYPES (fsLit "Any") anyTyConKey anyTyCon anyTyCon :: TyCon +-- See Note [Any types] anyTyCon = mkFamilyTyCon anyTyConName binders res_kind Nothing (ClosedSynFamilyTyCon Nothing) Nothing @@ -504,6 +546,24 @@ anyTy = mkTyConTy anyTyCon anyTypeOfKind :: Kind -> Type anyTypeOfKind kind = mkTyConApp anyTyCon [kind] +zonkAnyTyConName :: Name +zonkAnyTyConName = + mkWiredInTyConName UserSyntax gHC_TYPES (fsLit "ZonkAny") zonkAnyTyConKey zonkAnyTyCon + +zonkAnyTyCon :: TyCon +-- ZonkAnyTyCon :: forall k. Nat -> k +-- See Note [Any types] +zonkAnyTyCon = mkFamilyTyCon zonkAnyTyConName + [ mkNamedTyConBinder Specified kv + , mkAnonTyConBinder nat_kv ] + (mkTyVarTy kv) + Nothing + (ClosedSynFamilyTyCon Nothing) + Nothing + NotInjective + where + [kv,nat_kv] = mkTemplateKindVars [liftedTypeKind, naturalTy] + -- | Make a fake, recovery 'TyCon' from an existing one. -- Used when recovering from errors in type declarations makeRecoveryTyCon :: TyCon -> TyCon ===================================== compiler/GHC/Tc/Types.hs ===================================== @@ -567,6 +567,10 @@ data TcGblEnv tcg_dfun_n :: TcRef OccSet, -- ^ Allows us to choose unique DFun names. + tcg_zany_n :: TcRef Integer, + -- ^ A source of unique identities for ZonkAny instances + -- See Note [Any types] in GHC.Builtin.Types, wrinkle (Any4) + tcg_merged :: [(Module, Fingerprint)], -- ^ The requirements we merged with; we always have to recompile -- if any of these changed. ===================================== compiler/GHC/Tc/Utils/Monad.hs ===================================== @@ -142,7 +142,7 @@ module GHC.Tc.Utils.Monad( getCCIndexM, getCCIndexTcM, -- * Zonking - liftZonkM, + liftZonkM, newZonkAnyType, -- * Complete matches localAndImportedCompleteMatches, getCompleteMatchesTcM, @@ -156,6 +156,7 @@ import GHC.Prelude import GHC.Builtin.Names +import GHC.Builtin.Types( zonkAnyTyCon ) import GHC.Tc.Errors.Types import GHC.Tc.Types -- Re-export all @@ -178,6 +179,7 @@ import GHC.Core.UsageEnv import GHC.Core.Multiplicity import GHC.Core.InstEnv import GHC.Core.FamInstEnv +import GHC.Core.Type( mkNumLitTy ) import GHC.Driver.Env import GHC.Driver.Session @@ -258,6 +260,7 @@ initTc hsc_env hsc_src keep_rn_syntax mod loc do_this infer_var <- newIORef True ; infer_reasons_var <- newIORef emptyMessages ; dfun_n_var <- newIORef emptyOccSet ; + zany_n_var <- newIORef 0 ; let { type_env_var = hsc_type_env_vars hsc_env }; dependent_files_var <- newIORef [] ; @@ -348,6 +351,7 @@ initTc hsc_env hsc_src keep_rn_syntax mod loc do_this tcg_patsyns = [], tcg_merged = [], tcg_dfun_n = dfun_n_var, + tcg_zany_n = zany_n_var, tcg_keep = keep_var, tcg_hdr_info = (Nothing,Nothing), tcg_hpc = False, @@ -1787,6 +1791,18 @@ chooseUniqueOccTc fn = ; writeTcRef dfun_n_var (extendOccSet set occ) ; return occ } +newZonkAnyType :: Kind -> TcM Type +-- Return a type (ZonkAny @k n), where n is fresh +-- Recall ZonkAny :: forall k. Natural -> k +-- See Note [Any types] in GHC.Builtin.Types, wrinkle (Any4) +newZonkAnyType kind + = do { env <- getGblEnv + ; let zany_n_var = tcg_zany_n env + ; i <- readTcRef zany_n_var + ; let !i2 = i+1 + ; writeTcRef zany_n_var i2 + ; return (mkTyConApp zonkAnyTyCon [kind, mkNumLitTy i]) } + getConstraintVar :: TcM (TcRef WantedConstraints) getConstraintVar = do { env <- getLclEnv; return (tcl_lie env) } ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -54,7 +54,7 @@ import GHC.Tc.Types.TcRef import GHC.Tc.TyCl.Build ( TcMethInfo, MethInfo ) import GHC.Tc.Utils.Env ( tcLookupGlobalOnly ) import GHC.Tc.Utils.TcType -import GHC.Tc.Utils.Monad ( setSrcSpanA, liftZonkM, traceTc, addErr ) +import GHC.Tc.Utils.Monad ( newZonkAnyType, setSrcSpanA, liftZonkM, traceTc, addErr ) import GHC.Tc.Types.Constraint import GHC.Tc.Types.Evidence import GHC.Tc.Errors.Types @@ -469,8 +469,9 @@ commitFlexi tv zonked_kind -> do { addErr $ TcRnZonkerMessage (ZonkerCannotDefaultConcrete origin) ; return (anyTypeOfKind zonked_kind) } | otherwise - -> do { traceTc "Defaulting flexi tyvar to Any:" (pprTyVar tv) - ; return (anyTypeOfKind zonked_kind) } + -> do { traceTc "Defaulting flexi tyvar to ZonkAny:" (pprTyVar tv) + -- See Note [Any types] in GHC.Builtin.Types, esp wrinkle (Any4) + ; newZonkAnyType zonked_kind } RuntimeUnkFlexi -> do { traceTc "Defaulting flexi tyvar to RuntimeUnk:" (pprTyVar tv) ===================================== testsuite/tests/perf/compiler/T11068.stdout ===================================== @@ -23,137 +23,137 @@ `cast` (GHC.Internal.Generics.N:M1[0] `cast` (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.L1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 - ((GHC.Internal.Generics.U1 @(*) @GHC.Types.Any) + ((GHC.Internal.Generics.U1 @(*) @(GHC.Types.ZonkAny 0)) `cast` (Sym (GHC.Internal.Generics.N:M1[0] = GHC.Internal.Generics.R1 = GHC.Internal.Generics.R1 ===================================== testsuite/tests/pmcheck/should_compile/T12957.stderr ===================================== @@ -1,9 +1,9 @@ - T12957.hs:4:5: warning: [GHC-62161] [-Wincomplete-patterns (in -Wextra)] Pattern match(es) are non-exhaustive In a case alternative: - Patterns of type ‘[GHC.Types.Any]’ not matched: [] + Patterns of type ‘[GHC.Types.ZonkAny 0]’ not matched: [] T12957.hs:4:16: warning: [GHC-53633] [-Woverlapping-patterns (in -Wdefault)] Pattern match is redundant In a case alternative: (_ : _) -> ... + ===================================== testsuite/tests/pmcheck/should_compile/T24817.hs ===================================== @@ -0,0 +1,15 @@ +{-# LANGUAGE GADTs, DataKinds #-} + +module T24817 where + +data SBool b where + STrue :: SBool True + SFalse :: SBool False + +foo :: forall a b. (SBool a, SBool b) +foo = error "urk" + +bar :: Bool +bar = case foo of + (STrue, SFalse) -> True + _ -> False ===================================== testsuite/tests/pmcheck/should_compile/all.T ===================================== @@ -171,3 +171,4 @@ test('DsIncompleteRecSel1', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel2', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel3', [collect_compiler_stats('bytes allocated', 10)], compile, ['-Wincomplete-record-selectors']) test('DoubleMatch', normal, compile, [overlapping_incomplete]) +test('T24817', normal, compile, [overlapping_incomplete]) ===================================== testsuite/tests/printer/T17697.stderr ===================================== @@ -1,7 +1,8 @@ - T17697.hs:6:5: warning: [GHC-88464] [-Wdeferred-out-of-scope-variables (in -Wdefault)] Variable not in scope: threadDelay :: t0 -> IO a0 T17697.hs:6:5: warning: [GHC-81995] [-Wunused-do-bind (in -Wall)] - A do-notation statement discarded a result of type ‘GHC.Types.Any’ + A do-notation statement discarded a result of type + ‘GHC.Types.ZonkAny 1’ Suggested fix: Suppress this warning by saying ‘_ <- threadDelay 1’ + ===================================== testsuite/tests/profiling/should_run/staticcallstack002.stdout ===================================== @@ -1,4 +1,4 @@ -Just (InfoProv {ipName = "sat_s237_info", ipDesc = THUNK, ipTyDesc = "Any", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "10:23-39"}) -Just (InfoProv {ipName = "sat_s23r_info", ipDesc = THUNK, ipTyDesc = "Any", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "11:23-42"}) -Just (InfoProv {ipName = "sat_s23L_info", ipDesc = THUNK, ipTyDesc = "Any", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "12:23-46"}) -Just (InfoProv {ipName = "sat_s245_info", ipDesc = THUNK, ipTyDesc = "Any", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "13:23-44"}) +Just (InfoProv {ipName = "sat_s1Rh_info", ipDesc = THUNK, ipTyDesc = "ZonkAny 0", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "10:23-39"}) +Just (InfoProv {ipName = "sat_s1RB_info", ipDesc = THUNK, ipTyDesc = "ZonkAny 1", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "11:23-42"}) +Just (InfoProv {ipName = "sat_s1RV_info", ipDesc = THUNK, ipTyDesc = "ZonkAny 2", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "12:23-46"}) +Just (InfoProv {ipName = "sat_s1Sf_info", ipDesc = THUNK, ipTyDesc = "ZonkAny 3", ipLabel = "main", ipUnitId = "main", ipMod = "Main", ipSrcFile = "staticcallstack002.hs", ipSrcSpan = "13:23-44"}) ===================================== testsuite/tests/simplCore/should_compile/T13156.stdout ===================================== @@ -1 +1,2 @@ - case r @GHC.Types.Any of { __DEFAULT -> r @a } + case r @(GHC.Types.ZonkAny 0) of { __DEFAULT -> + case r @(GHC.Types.ZonkAny 1) of { __DEFAULT -> r @a } ===================================== testsuite/tests/typecheck/should_fail/T13292.stderr ===================================== @@ -1,4 +1,3 @@ - T13292a.hs:4:12: warning: [GHC-39999] [-Wdeferred-type-errors (in -Wdefault)] • Ambiguous type variable ‘m0’ arising from a use of ‘return’ prevents the constraint ‘(Monad m0)’ from being solved. @@ -15,14 +14,15 @@ T13292a.hs:4:12: warning: [GHC-39999] [-Wdeferred-type-errors (in -Wdefault)] In an equation for ‘someFunc’: someFunc = return () T13292.hs:6:1: warning: [GHC-83865] [-Wdeferred-type-errors (in -Wdefault)] - • Couldn't match type ‘GHC.Types.Any’ with ‘IO’ + • Couldn't match type ‘GHC.Types.ZonkAny 0’ with ‘IO’ Expected: IO () - Actual: GHC.Types.Any () + Actual: GHC.Types.ZonkAny 0 () • When checking the type of the IO action ‘main’ T13292.hs:6:1: warning: [GHC-83865] [-Wdeferred-type-errors (in -Wdefault)] - • Couldn't match type ‘GHC.Types.Any’ with ‘IO’ + • Couldn't match type ‘GHC.Types.ZonkAny 0’ with ‘IO’ Expected: IO () - Actual: GHC.Types.Any () + Actual: GHC.Types.ZonkAny 0 () • In the expression: main When checking the type of the IO action ‘main’ + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/cfbff65a8bde902b4510cdaead847bf7a52b4018 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/cfbff65a8bde902b4510cdaead847bf7a52b4018 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 10 23:51:37 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 10 Sep 2024 19:51:37 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 7 commits: haddock: Re-organise cross-OS compatibility layer Message-ID: <66e0db89caa_27f303997578659cc@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - b5f3ee90 by Matthew Pickering at 2024-09-10T19:51:31-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 431b28e5 by Rodrigo Mesquita at 2024-09-10T19:51:32-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - d57344fc by Alan Zimmerman at 2024-09-10T19:51:32-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 30 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Annotation.hs - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Tc/Types.hs - compiler/GHC/Tc/Utils/Monad.hs - compiler/GHC/Tc/Zonk/Type.hs - hadrian/src/Builder.hs - libraries/base/src/GHC/Exts.hs - libraries/ghci/GHCi/CreateBCO.hs - libraries/ghci/ghci.cabal.in - + testsuite/tests/module/T21752.stderr - testsuite/tests/perf/compiler/T11068.stdout - testsuite/tests/pmcheck/should_compile/T12957.stderr - + testsuite/tests/pmcheck/should_compile/T24817.hs - testsuite/tests/pmcheck/should_compile/all.T - testsuite/tests/printer/T17697.stderr - testsuite/tests/profiling/should_run/staticcallstack002.stdout - testsuite/tests/simplCore/should_compile/T13156.stdout - testsuite/tests/typecheck/should_fail/T13292.stderr - utils/check-exact/ExactPrint.hs - utils/check-exact/Utils.hs - + utils/haddock/haddock-api/compat/posix/Haddock/Compat.hs - + utils/haddock/haddock-api/compat/windows/Haddock/Compat.hs - utils/haddock/haddock-api/haddock-api.cabal - utils/haddock/haddock-api/src/Haddock.hs - utils/haddock/haddock-api/src/Haddock/Interface.hs - utils/haddock/haddock-api/src/Haddock/Types.hs - utils/haddock/haddock-api/src/Haddock/Utils.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1c0fe5eb122f82e121e71fa5fbd4f6060b5646dc...d57344fc13bdbb23d73636a0a4635b060743d77c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1c0fe5eb122f82e121e71fa5fbd4f6060b5646dc...d57344fc13bdbb23d73636a0a4635b060743d77c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 06:42:12 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 11 Sep 2024 02:42:12 -0400 Subject: [Git][ghc/ghc][master] hadrian: Make sure ffi headers are built before using a compiler Message-ID: <66e13bc47e43b_1aa37057a9e080477@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 1 changed file: - hadrian/src/Builder.hs Changes: ===================================== hadrian/src/Builder.hs ===================================== @@ -237,16 +237,25 @@ instance H.Builder Builder where -- changes (#18001). _bootGhcVersion <- setting GhcVersion pure [] - Ghc {} -> do + Ghc _ st -> do root <- buildRoot unlitPath <- builderPath Unlit distro_mingw <- settingsFileSetting ToolchainSetting_DistroMinGW + libffi_adjustors <- useLibffiForAdjustors + use_system_ffi <- flag UseSystemFfi return $ [ unlitPath ] ++ [ root -/- mingwStamp | windowsHost, distro_mingw == "NO" ] -- proxy for the entire mingw toolchain that -- we have in inplace/mingw initially, and then at -- root -/- mingw. + -- ffi.h needed by the compiler when using libffi_adjustors (#24864) + -- It would be nicer to not duplicate this logic between here + -- and needRtsLibffiTargets and libffiHeaderFiles but this doesn't change + -- very often. + ++ [ root -/- buildDir (rtsContext st) -/- "include" -/- header + | header <- ["ffi.h", "ffitarget.h"] + , libffi_adjustors && not use_system_ffi ] Hsc2Hs stage -> (\p -> [p]) <$> templateHscPath stage Make dir -> return [dir -/- "Makefile"] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0167e472e7035d31591777983d8e35528aceefff -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0167e472e7035d31591777983d8e35528aceefff You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 06:42:37 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 11 Sep 2024 02:42:37 -0400 Subject: [Git][ghc/ghc][master] base: Deprecate BCO primops exports from GHC.Exts Message-ID: <66e13bdda512d_1aa3705722cc8368a@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - 4 changed files: - libraries/base/src/GHC/Exts.hs - libraries/ghci/GHCi/CreateBCO.hs - libraries/ghci/ghci.cabal.in - + testsuite/tests/module/T21752.stderr Changes: ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -23,6 +23,12 @@ module GHC.Exts -- ** Legacy interface for arrays of arrays module GHC.Internal.ArrayArray, -- * Primitive operations + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.BCO, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.mkApUpd0#, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.newBCO#, module GHC.Prim, module GHC.Prim.Ext, -- ** Running 'RealWorld' state thread @@ -114,7 +120,10 @@ import GHC.Internal.Exts import GHC.Internal.ArrayArray import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom#, - isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned# ) + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned#, + -- Deprecated + BCO, mkApUpd0#, newBCO# ) +import qualified GHC.Prim as Prim ( BCO, mkApUpd0#, newBCO# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/ghci/GHCi/CreateBCO.hs ===================================== @@ -6,6 +6,10 @@ {-# LANGUAGE UnboxedTuples #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE CPP #-} +{-# OPTIONS_GHC -Wno-warnings-deprecations #-} +-- TODO We want to import GHC.Internal.Base (BCO, mkApUpd0#, newBCO#) instead +-- of from GHC.Exts when we can require of the bootstrap compiler to have +-- ghc-internal. -- -- (c) The University of Glasgow 2002-2006 ===================================== libraries/ghci/ghci.cabal.in ===================================== @@ -86,9 +86,10 @@ library array == 0.5.*, base >= 4.8 && < 4.21, -- ghc-internal == @ProjectVersionForLib at .* - -- TODO: Use GHC.Internal.Desugar from ghc-internal instead of ignoring - -- the deprecation warning of GHC.Desugar when we require ghc-internal - -- of the bootstrap compiler + -- TODO: Use GHC.Internal.Desugar and GHC.Internal.Base from + -- ghc-internal instead of ignoring the deprecation warning in GHCi.TH + -- and GHCi.CreateBCO when we require ghc-internal of the bootstrap + -- compiler ghc-prim >= 0.5.0 && < 0.12, binary == 0.8.*, bytestring >= 0.10 && < 0.13, ===================================== testsuite/tests/module/T21752.stderr ===================================== @@ -0,0 +1,32 @@ +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘newBCO#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘newBCO#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘mkApUpd0#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘mkApUpd0#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of type constructor or class ‘BCO’ + (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of type constructor or class ‘BCO’ + (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0f696958db9e52b5b07d3d52658b4786280eb419 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0f696958db9e52b5b07d3d52658b4786280eb419 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 06:43:19 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 11 Sep 2024 02:43:19 -0400 Subject: [Git][ghc/ghc][master] EPA: Remove Anchor = EpaLocation synonym Message-ID: <66e13c0721efd_1aa37091788c871c6@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 5 changed files: - compiler/GHC/Parser.y - compiler/GHC/Parser/Annotation.hs - compiler/GHC/Parser/PostProcess.hs - utils/check-exact/ExactPrint.hs - utils/check-exact/Utils.hs Changes: ===================================== compiler/GHC/Parser.y ===================================== @@ -4483,13 +4483,13 @@ gl = getLoc glA :: HasLoc a => a -> SrcSpan glA = getHasLoc -glR :: HasLoc a => a -> Anchor +glR :: HasLoc a => a -> EpaLocation glR !la = EpaSpan (getHasLoc la) -glEE :: (HasLoc a, HasLoc b) => a -> b -> Anchor +glEE :: (HasLoc a, HasLoc b) => a -> b -> EpaLocation glEE !x !y = spanAsAnchor $ comb2 x y -glRM :: Located a -> Maybe Anchor +glRM :: Located a -> Maybe EpaLocation glRM (L !l _) = Just $ spanAsAnchor l glAA :: HasLoc a => a -> EpaLocation @@ -4609,11 +4609,11 @@ hsDoAnn :: Located a -> LocatedAn t b -> AnnKeywordId -> AnnList hsDoAnn (L l _) (L ll _) kw = AnnList (Just $ spanAsAnchor (locA ll)) Nothing Nothing [AddEpAnn kw (srcSpan2e l)] [] -listAsAnchor :: [LocatedAn t a] -> Located b -> Anchor +listAsAnchor :: [LocatedAn t a] -> Located b -> EpaLocation listAsAnchor [] (L l _) = spanAsAnchor l listAsAnchor (h:_) s = spanAsAnchor (comb2 h s) -listAsAnchorM :: [LocatedAn t a] -> Maybe Anchor +listAsAnchorM :: [LocatedAn t a] -> Maybe EpaLocation listAsAnchorM [] = Nothing listAsAnchorM (L l _:_) = case locA l of ===================================== compiler/GHC/Parser/Annotation.hs ===================================== @@ -23,7 +23,7 @@ module GHC.Parser.Annotation ( TokenLocation(..), DeltaPos(..), deltaPos, getDeltaLine, - EpAnn(..), Anchor, + EpAnn(..), anchor, spanAsAnchor, realSpanAsAnchor, noSpanAnchor, @@ -528,7 +528,7 @@ instance Outputable AddEpAnn where -- new AST fragments out of old ones, and have them still printed out -- in a precise way. data EpAnn ann - = EpAnn { entry :: !Anchor + = EpAnn { entry :: !EpaLocation -- ^ Base location for the start of the syntactic element -- holding the annotations. , anns :: !ann -- ^ Annotations added by the Parser @@ -539,15 +539,6 @@ data EpAnn ann deriving (Data, Eq, Functor) -- See Note [XRec and Anno in the AST] --- | An 'Anchor' records the base location for the start of the --- syntactic element holding the annotations, and is used as the point --- of reference for calculating delta positions for contained --- annotations. --- It is also normally used as the reference point for the spacing of --- the element relative to its container. If the AST element is moved, --- that relationship is tracked using the 'EpaDelta' constructor instead. -type Anchor = EpaLocation -- Transitional - anchor :: (EpaLocation' a) -> RealSrcSpan anchor (EpaSpan (RealSrcSpan r _)) = r anchor _ = panic "anchor" @@ -676,7 +667,7 @@ data AnnListItem -- keywords such as 'where'. data AnnList = AnnList { - al_anchor :: Maybe Anchor, -- ^ start point of a list having layout + al_anchor :: Maybe EpaLocation, -- ^ start point of a list having layout al_open :: Maybe AddEpAnn, al_close :: Maybe AddEpAnn, al_rest :: [AddEpAnn], -- ^ context, such as 'where' keyword @@ -1143,7 +1134,7 @@ listLocation as = EpaSpan (go noSrcSpan as) go acc (L (EpAnn (EpaSpan s) _ _) _:rest) = go (combine acc s) rest go acc (_:rest) = go acc rest -widenAnchor :: Anchor -> [AddEpAnn] -> Anchor +widenAnchor :: EpaLocation -> [AddEpAnn] -> EpaLocation widenAnchor (EpaSpan (RealSrcSpan s mb)) as = EpaSpan (RealSrcSpan (widenRealSpan s as) (liftA2 combineBufSpans mb (bufSpanFromAnns as))) widenAnchor (EpaSpan us) _ = EpaSpan us @@ -1151,7 +1142,7 @@ widenAnchor a at EpaDelta{} as = case (realSpanFromAnns as) of Strict.Nothing -> a Strict.Just r -> EpaSpan (RealSrcSpan r Strict.Nothing) -widenAnchorS :: Anchor -> SrcSpan -> Anchor +widenAnchorS :: EpaLocation -> SrcSpan -> EpaLocation widenAnchorS (EpaSpan (RealSrcSpan s mbe)) (RealSrcSpan r mbr) = EpaSpan (RealSrcSpan (combineRealSrcSpans s r) (liftA2 combineBufSpans mbe mbr)) widenAnchorS (EpaSpan us) _ = EpaSpan us ===================================== compiler/GHC/Parser/PostProcess.hs ===================================== @@ -478,13 +478,13 @@ add_where an@(AddEpAnn _ (EpaSpan (RealSrcSpan rs _))) (EpAnn a (AnnList anc o c add_where (AddEpAnn _ _) _ _ = panic "add_where" -- EpaDelta should only be used for transformations -valid_anchor :: Anchor -> Bool +valid_anchor :: EpaLocation -> Bool valid_anchor (EpaSpan (RealSrcSpan r _)) = srcSpanStartLine r >= 0 valid_anchor _ = False -- If the decl list for where binds is empty, the anchor ends up -- invalid. In this case, use the parent one -patch_anchor :: RealSrcSpan -> Anchor -> Anchor +patch_anchor :: RealSrcSpan -> EpaLocation -> EpaLocation patch_anchor r EpaDelta{} = EpaSpan (RealSrcSpan r Strict.Nothing) patch_anchor r1 (EpaSpan (RealSrcSpan r0 mb)) = EpaSpan (RealSrcSpan r mb) where @@ -495,9 +495,9 @@ fixValbindsAnn :: EpAnn AnnList -> EpAnn AnnList fixValbindsAnn (EpAnn anchor (AnnList ma o c r t) cs) = (EpAnn (widenAnchor anchor (r ++ map trailingAnnToAddEpAnn t)) (AnnList ma o c r t) cs) --- | The 'Anchor' for a stmtlist is based on either the location or +-- | The anchor for a stmtlist is based on either the location or -- the first semicolon annotion. -stmtsAnchor :: Located (OrdList AddEpAnn,a) -> Maybe Anchor +stmtsAnchor :: Located (OrdList AddEpAnn,a) -> Maybe EpaLocation stmtsAnchor (L (RealSrcSpan l mb) ((ConsOL (AddEpAnn _ (EpaSpan (RealSrcSpan r rb))) _), _)) = Just $ widenAnchorS (EpaSpan (RealSrcSpan l mb)) (RealSrcSpan r rb) stmtsAnchor (L (RealSrcSpan l mb) _) = Just $ EpaSpan (RealSrcSpan l mb) ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -175,8 +175,8 @@ data EPState = EPState { uAnchorSpan :: !RealSrcSpan -- ^ in pre-changed AST -- reference frame, from -- Annotation - , uExtraDP :: !(Maybe Anchor) -- ^ Used to anchor a - -- list + , uExtraDP :: !(Maybe EpaLocation) -- ^ Used to anchor a + -- list , pAcceptSpan :: Bool -- ^ When we have processed an -- entry of EpaDelta, accept the -- next `EpaSpan` start as the @@ -214,21 +214,21 @@ class HasTrailing a where setTrailing :: a -> [TrailingAnn] -> a setAnchorEpa :: (HasTrailing an, NoAnn an) - => EpAnn an -> Anchor -> [TrailingAnn] -> EpAnnComments -> EpAnn an + => EpAnn an -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn an setAnchorEpa (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing an ts) cs -setAnchorHsModule :: HsModule GhcPs -> Anchor -> EpAnnComments -> HsModule GhcPs +setAnchorHsModule :: HsModule GhcPs -> EpaLocation -> EpAnnComments -> HsModule GhcPs setAnchorHsModule hsmod anc cs = hsmod { hsmodExt = (hsmodExt hsmod) {hsmodAnn = an'} } where anc' = anc an' = setAnchorEpa (hsmodAnn $ hsmodExt hsmod) anc' [] cs setAnchorAn :: (HasTrailing an, NoAnn an) - => LocatedAn an a -> Anchor -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a + => LocatedAn an a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a setAnchorAn (L (EpAnn _ an _) a) anc ts cs = (L (EpAnn anc (setTrailing an ts) cs) a) -- `debug` ("setAnchorAn: anc=" ++ showAst anc) -setAnchorEpaL :: EpAnn AnnList -> Anchor -> [TrailingAnn] -> EpAnnComments -> EpAnn AnnList +setAnchorEpaL :: EpAnn AnnList -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn AnnList setAnchorEpaL (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing (an {al_anchor = Nothing}) ts) cs -- --------------------------------------------------------------------- @@ -250,14 +250,14 @@ data CanUpdateAnchor = CanUpdateAnchor | NoCanUpdateAnchor deriving (Eq, Show) -data Entry = Entry Anchor [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor +data Entry = Entry EpaLocation [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor | NoEntryVal -- | For flagging whether to capture comments in an EpaDelta or not data CaptureComments = CaptureComments | NoCaptureComments -mkEntry :: Anchor -> [TrailingAnn] -> EpAnnComments -> Entry +mkEntry :: EpaLocation -> [TrailingAnn] -> EpAnnComments -> Entry mkEntry anc ts cs = Entry anc ts cs NoFlushComments CanUpdateAnchor instance (HasTrailing a) => HasEntry (EpAnn a) where @@ -642,7 +642,7 @@ withPpr a = do -- 'ppr'. class (Typeable a) => ExactPrint a where getAnnotationEntry :: a -> Entry - setAnnotationAnchor :: a -> Anchor -> [TrailingAnn] -> EpAnnComments -> a + setAnnotationAnchor :: a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> a exact :: (Monad m, Monoid w) => a -> EP w m a -- --------------------------------------------------------------------- @@ -4277,7 +4277,7 @@ instance ExactPrint (LocatedN RdrName) where locFromAdd :: AddEpAnn -> EpaLocation locFromAdd (AddEpAnn _ loc) = loc -printUnicode :: (Monad m, Monoid w) => Anchor -> RdrName -> EP w m Anchor +printUnicode :: (Monad m, Monoid w) => EpaLocation -> RdrName -> EP w m EpaLocation printUnicode anc n = do let str = case (showPprUnsafe n) of -- TODO: unicode support? @@ -4977,10 +4977,10 @@ setPosP l = do debugM $ "setPosP:" ++ show l modify (\s -> s {epPos = l}) -getExtraDP :: (Monad m, Monoid w) => EP w m (Maybe Anchor) +getExtraDP :: (Monad m, Monoid w) => EP w m (Maybe EpaLocation) getExtraDP = gets uExtraDP -setExtraDP :: (Monad m, Monoid w) => Maybe Anchor -> EP w m () +setExtraDP :: (Monad m, Monoid w) => Maybe EpaLocation -> EP w m () setExtraDP md = do debugM $ "setExtraDP:" ++ show md modify (\s -> s {uExtraDP = md}) ===================================== utils/check-exact/Utils.hs ===================================== @@ -255,7 +255,7 @@ tokComment t@(L lt c) = (GHC.EpaComment (EpaDocComment dc) pt) -> hsDocStringComments (noCommentsToEpaLocation lt) pt dc _ -> [mkComment (normaliseCommentText (ghcCommentText t)) lt (ac_prior_tok c)] -hsDocStringComments :: Anchor -> RealSrcSpan -> GHC.HsDocString -> [Comment] +hsDocStringComments :: EpaLocation -> RealSrcSpan -> GHC.HsDocString -> [Comment] hsDocStringComments _ pt (MultiLineDocString dec (x :| xs)) = let decStr = printDecorator dec @@ -342,7 +342,7 @@ isKWComment c = isJust (commentOrigin c) noKWComments :: [Comment] -> [Comment] noKWComments = filter (\c -> not (isKWComment c)) -sortAnchorLocated :: [GenLocated Anchor a] -> [GenLocated Anchor a] +sortAnchorLocated :: [GenLocated EpaLocation a] -> [GenLocated EpaLocation a] sortAnchorLocated = sortBy (compare `on` (anchor . getLoc)) -- | Calculates the distance from the start of a string to the end of @@ -401,12 +401,6 @@ setTrailingAnnLoc (AddDarrowUAnn _) ss = (AddDarrowUAnn ss) addEpAnnLoc :: AddEpAnn -> EpaLocation addEpAnnLoc (AddEpAnn _ l) = l --- --------------------------------------------------------------------- - --- TODO: get rid of this identity function -anchorToEpaLocation :: Anchor -> EpaLocation -anchorToEpaLocation a = a - -- --------------------------------------------------------------------- -- Horrible hack for dealing with some things still having a SrcSpan, -- not an Anchor. @@ -432,7 +426,7 @@ To be absolutely sure, we make the delta versions use -ve values. -} -hackSrcSpanToAnchor :: SrcSpan -> Anchor +hackSrcSpanToAnchor :: SrcSpan -> EpaLocation hackSrcSpanToAnchor (UnhelpfulSpan s) = error $ "hackSrcSpanToAnchor : UnhelpfulSpan:" ++ show s hackSrcSpanToAnchor ss@(RealSrcSpan r mb) = case mb of @@ -443,7 +437,7 @@ hackSrcSpanToAnchor ss@(RealSrcSpan r mb) else EpaSpan (RealSrcSpan r mb) _ -> EpaSpan (RealSrcSpan r mb) -hackAnchorToSrcSpan :: Anchor -> SrcSpan +hackAnchorToSrcSpan :: EpaLocation -> SrcSpan hackAnchorToSrcSpan (EpaSpan s) = s hackAnchorToSrcSpan _ = error $ "hackAnchorToSrcSpan" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/cf0e7729f0cb3788ce123b9ec4a71e44f0f825d4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/cf0e7729f0cb3788ce123b9ec4a71e44f0f825d4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 06:52:30 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Wed, 11 Sep 2024 02:52:30 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25250 Message-ID: <66e13e2e637ea_1aa370b01f6c90832@gitlab.mail> Sebastian Graf pushed new branch wip/T25250 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25250 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 06:53:30 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Wed, 11 Sep 2024 02:53:30 -0400 Subject: [Git][ghc/ghc][wip/T25250] 19 commits: JS: fake support for native adjustors (#25159) Message-ID: <66e13e6a5e016_1aa370bc1c40926b7@gitlab.mail> Sebastian Graf pushed to branch wip/T25250 at Glasgow Haskell Compiler / GHC Commits: 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - d271048b by Sebastian Graf at 2024-09-11T08:53:22+02:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 30 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Core/InstEnv.hs - compiler/GHC/Core/Opt/CallerCC/Types.hs - compiler/GHC/Core/Opt/WorkWrap/Utils.hs - compiler/GHC/HsToCore/Arrows.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Foreign/C.hs - compiler/GHC/HsToCore/Foreign/Call.hs - compiler/GHC/HsToCore/Foreign/JavaScript.hs - compiler/GHC/HsToCore/Foreign/Wasm.hs - compiler/GHC/HsToCore/ListComp.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Annotation.hs - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/StgToJS/Symbols.hs - compiler/GHC/StgToJS/Types.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/Splice.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Dict.hs - compiler/GHC/Tc/Solver/Equality.hs - compiler/GHC/Tc/Solver/Irred.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Solver/Solve.hs - compiler/GHC/Tc/TyCl/Instance.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/497ba969844c42279d8c655eee56840be10f69dc...d271048b1d31853c5605fe0ec9aada550a1264d5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/497ba969844c42279d8c655eee56840be10f69dc...d271048b1d31853c5605fe0ec9aada550a1264d5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 13:50:22 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 11 Sep 2024 09:50:22 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 6 commits: determinism: Sampling uniques in the CG Message-ID: <66e1a01e1e78b_2cf0ea5d7564486a9@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 6388e22f by Rodrigo Mesquita at 2024-09-11T14:49:19+01:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - c8fced61 by Rodrigo Mesquita at 2024-09-11T14:50:02+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 119f0061 by Rodrigo Mesquita at 2024-09-11T14:50:02+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - f1b5d059 by Rodrigo Mesquita at 2024-09-11T14:50:02+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 254db8c3 by Rodrigo Mesquita at 2024-09-11T14:50:02+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - c4eea906 by Rodrigo Mesquita at 2024-09-11T14:50:02+01:00 determinism: Rename CmmGroups in generateCgIPEStub Make sure to also deterministically rename the IPE Cmm (as per Note [Renaming uniques deterministically]) to guarantee deterministic objects when IPE information is requested. - - - - - 30 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Reg/Graph.hs - compiler/GHC/CmmToAsm/Reg/Graph/Spill.hs - compiler/GHC/CmmToAsm/Reg/Linear.hs - compiler/GHC/CmmToAsm/Reg/Linear/Base.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3b1a97f53b07fc0bd2642ec46a9ae0f9ee8185c1...c4eea9067c5cce38f6875efd7f6a1ba2ef71b362 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3b1a97f53b07fc0bd2642ec46a9ae0f9ee8185c1...c4eea9067c5cce38f6875efd7f6a1ba2ef71b362 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 14:39:48 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 11 Sep 2024 10:39:48 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 6 commits: determinism: Sampling uniques in the CG Message-ID: <66e1abb4c700_d7cbf8aa349934a@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 1f5e4708 by Rodrigo Mesquita at 2024-09-11T15:39:38+01:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 05c01c99 by Rodrigo Mesquita at 2024-09-11T15:39:38+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - cb2465f3 by Rodrigo Mesquita at 2024-09-11T15:39:38+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - 5bee1ca6 by Rodrigo Mesquita at 2024-09-11T15:39:38+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - e51be21a by Rodrigo Mesquita at 2024-09-11T15:39:38+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 79688990 by Rodrigo Mesquita at 2024-09-11T15:39:39+01:00 determinism: Rename CmmGroups in generateCgIPEStub Make sure to also deterministically rename the IPE Cmm (as per Note [Renaming uniques deterministically]) to guarantee deterministic objects when IPE information is requested. - - - - - 30 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Reg/Graph.hs - compiler/GHC/CmmToAsm/Reg/Graph/Spill.hs - compiler/GHC/CmmToAsm/Reg/Linear.hs - compiler/GHC/CmmToAsm/Reg/Linear/Base.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c4eea9067c5cce38f6875efd7f6a1ba2ef71b362...79688990da1f23cc0602fa3a37677b255a3c897f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c4eea9067c5cce38f6875efd7f6a1ba2ef71b362...79688990da1f23cc0602fa3a37677b255a3c897f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 15:01:47 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Wed, 11 Sep 2024 11:01:47 -0400 Subject: [Git][ghc/ghc][wip/T24978] 2 commits: Fix windows test output Message-ID: <66e1b0dbacb9e_d7cbf27649c100225@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24978 at Glasgow Haskell Compiler / GHC Commits: ede5eaa4 by Simon Peyton Jones at 2024-09-11T10:00:34+01:00 Fix windows test output - - - - - e4bdc821 by Simon Peyton Jones at 2024-09-11T16:01:14+01:00 Cache AxiomRule in CoAxBranch - - - - - 18 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/TyCon.hs - compiler/GHC/Iface/Decl.hs - compiler/GHC/Iface/Rename.hs - compiler/GHC/Iface/Syntax.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Tc/Module.hs - compiler/GHC/Tc/TyCl.hs - compiler/GHC/Tc/TyCl/Build.hs - compiler/GHC/Tc/TyCl/Instance.hs - compiler/GHC/Tc/Utils/Backpack.hs - compiler/Language/Haskell/Syntax/Decls.hs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - utils/haddock/haddock-api/src/Haddock/Convert.hs Changes: ===================================== compiler/GHC/Builtin/Names.hs ===================================== @@ -2009,55 +2009,56 @@ unsatisfiableClassNameKey = mkPreludeTyConUnique 170 -- Custom user type-errors errorMessageTypeErrorFamKey :: Unique -errorMessageTypeErrorFamKey = mkPreludeTyConUnique 181 +errorMessageTypeErrorFamKey = mkPreludeTyConUnique 171 coercibleTyConKey :: Unique -coercibleTyConKey = mkPreludeTyConUnique 183 +coercibleTyConKey = mkPreludeTyConUnique 172 proxyPrimTyConKey :: Unique -proxyPrimTyConKey = mkPreludeTyConUnique 184 +proxyPrimTyConKey = mkPreludeTyConUnique 173 specTyConKey :: Unique -specTyConKey = mkPreludeTyConUnique 185 +specTyConKey = mkPreludeTyConUnique 174 -anyTyConKey :: Unique -anyTyConKey = mkPreludeTyConUnique 186 +anyTyConKey, anyAxKey :: Unique +anyTyConKey = mkPreludeTyConUnique 175 +anyAxKey = mkPreludeTyConUnique 176 -smallArrayPrimTyConKey = mkPreludeTyConUnique 187 -smallMutableArrayPrimTyConKey = mkPreludeTyConUnique 188 +smallArrayPrimTyConKey = mkPreludeTyConUnique 180 +smallMutableArrayPrimTyConKey = mkPreludeTyConUnique 181 staticPtrTyConKey :: Unique -staticPtrTyConKey = mkPreludeTyConUnique 189 +staticPtrTyConKey = mkPreludeTyConUnique 182 staticPtrInfoTyConKey :: Unique -staticPtrInfoTyConKey = mkPreludeTyConUnique 190 +staticPtrInfoTyConKey = mkPreludeTyConUnique 183 callStackTyConKey :: Unique -callStackTyConKey = mkPreludeTyConUnique 191 +callStackTyConKey = mkPreludeTyConUnique 184 -- Typeables typeRepTyConKey, someTypeRepTyConKey, someTypeRepDataConKey :: Unique -typeRepTyConKey = mkPreludeTyConUnique 192 -someTypeRepTyConKey = mkPreludeTyConUnique 193 -someTypeRepDataConKey = mkPreludeTyConUnique 194 +typeRepTyConKey = mkPreludeTyConUnique 185 +someTypeRepTyConKey = mkPreludeTyConUnique 186 +someTypeRepDataConKey = mkPreludeTyConUnique 187 typeSymbolAppendFamNameKey :: Unique -typeSymbolAppendFamNameKey = mkPreludeTyConUnique 195 +typeSymbolAppendFamNameKey = mkPreludeTyConUnique 188 -- Unsafe equality unsafeEqualityTyConKey :: Unique -unsafeEqualityTyConKey = mkPreludeTyConUnique 196 +unsafeEqualityTyConKey = mkPreludeTyConUnique 189 -- Linear types multiplicityTyConKey :: Unique -multiplicityTyConKey = mkPreludeTyConUnique 197 +multiplicityTyConKey = mkPreludeTyConUnique 190 unrestrictedFunTyConKey :: Unique -unrestrictedFunTyConKey = mkPreludeTyConUnique 198 +unrestrictedFunTyConKey = mkPreludeTyConUnique 191 multMulTyConKey :: Unique -multMulTyConKey = mkPreludeTyConUnique 199 +multMulTyConKey = mkPreludeTyConUnique 192 ---------------- Template Haskell ------------------- -- GHC.Builtin.Names.TH: USES TyConUniques 200-299 ===================================== compiler/GHC/Builtin/Types.hs ===================================== @@ -196,6 +196,7 @@ import GHC.Types.Name.Env ( lookupNameEnv_NF, mkNameEnv ) import GHC.Types.Basic import GHC.Types.ForeignCall import GHC.Types.Unique.Set +import GHC.Types.SrcLoc( wiredInSrcSpan ) import {-# SOURCE #-} GHC.Tc.Types.Origin ( FixedRuntimeRepOrigin(..), mkFRRUnboxedTuple, mkFRRUnboxedSum ) @@ -486,17 +487,21 @@ bit of history. anyTyConName :: Name -anyTyConName = - mkWiredInTyConName UserSyntax gHC_TYPES (fsLit "Any") anyTyConKey anyTyCon +anyTyConName = mkWiredInTyConName UserSyntax gHC_TYPES (fsLit "Any") anyTyConKey anyTyCon + +anyAxName :: Name +anyAxName = mkExternalName anyAxKey gHC_TYPES (mkTcOccFS (fsLit "R:Any")) wiredInSrcSpan anyTyCon :: TyCon -anyTyCon = mkFamilyTyCon anyTyConName binders res_kind Nothing - (ClosedSynFamilyTyCon Nothing) - Nothing - NotInjective +anyTyCon = tc where + tc = mkFamilyTyCon anyTyConName binders res_kind Nothing + (ClosedSynFamilyTyCon co_ax) + Nothing + NotInjective binders@[kv] = mkTemplateKindTyConBinders [liftedTypeKind] res_kind = mkTyVarTy (binderVar kv) + co_ax = mkEmptyCoAxiom anyAxName tc anyTy :: Type anyTy = mkTyConTy anyTyCon ===================================== compiler/GHC/Core/Coercion.hs ===================================== @@ -1051,8 +1051,12 @@ mkAxiomCo = AxiomCo -- to be used only with unbranched axioms mkUnbranchedAxInstCo :: Role -> CoAxiom Unbranched -> [Type] -> [Coercion] -> Coercion -mkUnbranchedAxInstCo role ax tys cos - = mkAxInstCo role (UnbranchedAxiom ax) tys cos +mkUnbranchedAxInstCo role ax tys cos + = mkAxInstCo role ax_rule tys cos + where + !ax_rule = cab_axr (coAxiomNthBranch ax 0) + -- This will be (UnbranchedAxiom ax), + -- but we avoid allocating it every time mkAxInstRHS :: CoAxiom br -> BranchIndex -> [Type] -> [Coercion] -> Type -- Instantiate the axiom with specified types, ===================================== compiler/GHC/Core/Coercion/Axiom.hs ===================================== @@ -19,6 +19,7 @@ module GHC.Core.Coercion.Axiom ( CoAxiom(..), CoAxBranch(..), + mkEmptyCoAxiom, toBranchedAxiom, toUnbranchedAxiom, coAxiomName, coAxiomArity, coAxiomBranches, coAxiomTyCon, isImplicitCoAxiom, coAxiomNumPats, @@ -26,7 +27,7 @@ module GHC.Core.Coercion.Axiom ( coAxiomSingleBranch, coAxBranchTyVars, coAxBranchCoVars, coAxBranchRoles, coAxBranchLHS, coAxBranchRHS, coAxBranchSpan, coAxBranchIncomps, - placeHolderIncomps, + placeHolderIncomps, pairWithAxiomRules, Role(..), fsFromRole, @@ -158,10 +159,13 @@ newtype Branches (br :: BranchFlag) type role Branches nominal manyBranches :: [CoAxBranch] -> Branches Branched -manyBranches brs = assert (snd bnds >= fst bnds ) +manyBranches brs = assert (snd bnds >= fst bnds - 1) MkBranches (listArray bnds brs) where bnds = (0, length brs - 1) + -- The list of branches can be empty; + -- e.g. type family F a where {} + -- Then bnds = (0,-1) and the array is empty unbranched :: CoAxBranch -> Branches Unbranched unbranched br = MkBranches (listArray (0, 0) [br]) @@ -279,9 +283,25 @@ data CoAxBranch , cab_incomps :: [CoAxBranch] -- ^ The previous incompatible branches -- See Note [Storing compatibility] + + , cab_axr :: CoAxiomRule + -- ^ The parent axiom of this branch + -- Cached here so that we allocate it once and for all, + -- rather than re-allocating it every time we pick this branch + -- See Note [Avoiding allocating lots of CoAxiomRules] } deriving Data.Data +mkEmptyCoAxiom :: Name -> TyCon -> CoAxiom Branched +-- An axiom with no branches +mkEmptyCoAxiom ax_name tycon + = CoAxiom { co_ax_unique = getUnique ax_name + , co_ax_name = ax_name + , co_ax_role = Nominal + , co_ax_tc = tycon + , co_ax_branches = manyBranches [] + , co_ax_implicit = False } + toBranchedAxiom :: CoAxiom br -> CoAxiom Branched toBranchedAxiom ax@(CoAxiom { co_ax_branches = branches }) = ax { co_ax_branches = toBranched branches } @@ -353,6 +373,11 @@ coAxBranchIncomps = cab_incomps placeHolderIncomps :: [CoAxBranch] placeHolderIncomps = panic "placeHolderIncomps" +pairWithAxiomRules :: CoAxiom Branched -> [a] -> [(a, CoAxiomRule)] +pairWithAxiomRules ax [branch] = [(branch, UnbranchedAxiom (toUnbranchedAxiom ax))] +pairWithAxiomRules ax branches = [(b, BranchedAxiom ax i) + | (b,i) <- branches `zip` [0..]] + {- Note [CoAxBranch type variables] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Core/FamInstEnv.hs ===================================== @@ -71,7 +71,7 @@ import qualified GHC.Data.List.Infinite as Inf import Control.Monad import Data.List( mapAccumL ) -import Data.Array( Array, assocs ) +import Data.Array as Arr( Array, elems ) {- ************************************************************************ @@ -710,16 +710,16 @@ are tidying (changing OccNames only), not freshening, in accordance with that Note. -} --- all axiom roles are Nominal, as this is only used with type families mkCoAxBranch :: [TyVar] -- original, possibly stale, tyvars -> [TyVar] -- Extra eta tyvars -> [CoVar] -- possibly stale covars -> [Type] -- LHS patterns -> Type -- RHS - -> [Role] + -> [Role] -- All axiom roles are Nominal, as this is only used with type families -> SrcSpan + -> CoAxiomRule -> CoAxBranch -mkCoAxBranch tvs eta_tvs cvs lhs rhs roles loc +mkCoAxBranch tvs eta_tvs cvs lhs rhs roles loc ax_rule = CoAxBranch { cab_tvs = tvs' , cab_eta_tvs = eta_tvs' , cab_cvs = cvs' @@ -727,7 +727,8 @@ mkCoAxBranch tvs eta_tvs cvs lhs rhs roles loc , cab_roles = roles , cab_rhs = tidyType env rhs , cab_loc = loc - , cab_incomps = placeHolderIncomps } + , cab_incomps = placeHolderIncomps + , cab_axr = ax_rule } where (env1, tvs') = tidyVarBndrs init_tidy_env tvs (env2, eta_tvs') = tidyVarBndrs env1 eta_tvs @@ -767,16 +768,17 @@ mkSingleCoAxiom :: Role -> Name -- Used for both type family (Nominal) and data family (Representational) -- axioms, hence passing in the Role mkSingleCoAxiom role ax_name tvs eta_tvs cvs fam_tc lhs_tys rhs_ty - = CoAxiom { co_ax_unique = nameUnique ax_name - , co_ax_name = ax_name - , co_ax_tc = fam_tc - , co_ax_role = role - , co_ax_implicit = False - , co_ax_branches = unbranched (branch { cab_incomps = [] }) } + = co_ax where + co_ax = CoAxiom { co_ax_unique = nameUnique ax_name + , co_ax_name = ax_name + , co_ax_tc = fam_tc + , co_ax_role = role + , co_ax_implicit = False + , co_ax_branches = unbranched (branch { cab_incomps = [] }) } branch = mkCoAxBranch tvs eta_tvs cvs lhs_tys rhs_ty (map (const Nominal) tvs) - (getSrcSpan ax_name) + (getSrcSpan ax_name) (UnbranchedAxiom co_ax) -- | Create a coercion constructor (axiom) suitable for the given -- newtype 'TyCon'. The 'Name' should be that of a new coercion @@ -784,16 +786,18 @@ mkSingleCoAxiom role ax_name tvs eta_tvs cvs fam_tc lhs_tys rhs_ty -- the type the appropriate right hand side of the @newtype@, with -- the free variables a subset of those 'TyVar's. mkNewTypeCoAxiom :: Name -> TyCon -> [TyVar] -> [Role] -> Type -> CoAxiom Unbranched -mkNewTypeCoAxiom name tycon tvs roles rhs_ty - = CoAxiom { co_ax_unique = nameUnique name - , co_ax_name = name - , co_ax_implicit = True -- See Note [Implicit axioms] in GHC.Core.TyCon - , co_ax_role = Representational - , co_ax_tc = tycon - , co_ax_branches = unbranched (branch { cab_incomps = [] }) } +mkNewTypeCoAxiom ax_name tycon tvs roles rhs_ty + = co_ax where + co_ax = CoAxiom { co_ax_unique = nameUnique ax_name + , co_ax_name = ax_name + , co_ax_implicit = True -- See Note [Implicit axioms] in GHC.Core.TyCon + , co_ax_role = Representational + , co_ax_tc = tycon + , co_ax_branches = unbranched (branch { cab_incomps = [] }) } + branch = mkCoAxBranch tvs [] [] (mkTyVarTys tvs) rhs_ty - roles (getSrcSpan name) + roles (getSrcSpan ax_name) (UnbranchedAxiom co_ax) {- ************************************************************************ @@ -1192,8 +1196,8 @@ reduceTyFamApp_maybe envs role tc tys in Just $ coercionRedn co | Just ax <- isClosedSynFamilyTyConWithAxiom_maybe tc - , Just (ind, inst_tys, inst_cos) <- chooseBranch ax tys - = let co = mkAxInstCo role (BranchedAxiom ax ind) inst_tys inst_cos + , Just (ax_rule, inst_tys, inst_cos) <- chooseBranch ax tys + = let co = mkAxInstCo role ax_rule inst_tys inst_cos in Just $ coercionRedn co | Just builtin_fam <- isBuiltInSynFamTyCon_maybe tc @@ -1206,30 +1210,30 @@ reduceTyFamApp_maybe envs role tc tys -- The axiom can be oversaturated. (Closed families only.) chooseBranch :: CoAxiom Branched -> [Type] - -> Maybe (BranchIndex, [Type], [Coercion]) -- found match, with args + -> Maybe (CoAxiomRule, [Type], [Coercion]) -- found match, with args chooseBranch axiom tys = do { let num_pats = coAxiomNumPats axiom (target_tys, extra_tys) = splitAt num_pats tys branches = coAxiomBranches axiom - ; (ind, inst_tys, inst_cos) + ; (ax_rule, inst_tys, inst_cos) <- findBranch (unMkBranches branches) target_tys - ; return ( ind, inst_tys `chkAppend` extra_tys, inst_cos ) } + ; return (ax_rule, inst_tys `chkAppend` extra_tys, inst_cos) } -- The axiom must *not* be oversaturated findBranch :: Array BranchIndex CoAxBranch -> [Type] - -> Maybe (BranchIndex, [Type], [Coercion]) + -> Maybe (CoAxiomRule, [Type], [Coercion]) -- coercions relate requested types to returned axiom LHS at role N findBranch branches target_tys - = foldr go Nothing (assocs branches) + = foldr go Nothing (Arr.elems branches) where - go :: (BranchIndex, CoAxBranch) - -> Maybe (BranchIndex, [Type], [Coercion]) - -> Maybe (BranchIndex, [Type], [Coercion]) - go (index, branch) other + go :: CoAxBranch + -> Maybe (CoAxiomRule, [Type], [Coercion]) + -> Maybe (CoAxiomRule, [Type], [Coercion]) + go branch other = let (CoAxBranch { cab_tvs = tpl_tvs, cab_cvs = tpl_cvs , cab_lhs = tpl_lhs - , cab_incomps = incomps }) = branch + , cab_incomps = incomps, cab_axr = ax_rule }) = branch in_scope = mkInScopeSet (unionVarSets $ map (tyCoVarsOfTypes . coAxBranchLHS) incomps) -- See Note [Flattening type-family applications when matching instances] @@ -1241,7 +1245,7 @@ findBranch branches target_tys -> -- matching worked & we're apart from all incompatible branches. -- success assert (all (isJust . lookupCoVar subst) tpl_cvs) $ - Just (index, substTyVars subst tpl_tvs, substCoVars subst tpl_cvs) + Just (ax_rule, substTyVars subst tpl_tvs, substCoVars subst tpl_cvs) -- failure. keep looking _ -> other ===================================== compiler/GHC/Core/TyCon.hs ===================================== @@ -1280,7 +1280,7 @@ data FamTyConFlav -- | A closed type synonym family e.g. -- @type family F x where { F Int = Bool }@ - | ClosedSynFamilyTyCon (Maybe (CoAxiom Branched)) + | ClosedSynFamilyTyCon (CoAxiom Branched) -- See Note [Closed type families] -- | A closed type synonym family declared in an hs-boot file with @@ -1293,8 +1293,7 @@ data FamTyConFlav instance Outputable FamTyConFlav where ppr (DataFamilyTyCon n) = text "data family" <+> ppr n ppr OpenSynFamilyTyCon = text "open type family" - ppr (ClosedSynFamilyTyCon Nothing) = text "closed type family" - ppr (ClosedSynFamilyTyCon (Just coax)) = text "closed type family" <+> ppr coax + ppr (ClosedSynFamilyTyCon coax) = text "closed type family" <+> ppr coax ppr AbstractClosedSynFamilyTyCon = text "abstract closed type family" ppr (BuiltInSynFamTyCon _) = text "built-in type family" @@ -2232,7 +2231,7 @@ isOpenTypeFamilyTyCon (TyCon { tyConDetails = details }) -- abstract or empty closed families. isClosedSynFamilyTyConWithAxiom_maybe :: TyCon -> Maybe (CoAxiom Branched) isClosedSynFamilyTyConWithAxiom_maybe (TyCon { tyConDetails = details }) - | FamilyTyCon {famTcFlav = ClosedSynFamilyTyCon mb} <- details = mb + | FamilyTyCon {famTcFlav = ClosedSynFamilyTyCon ax} <- details = Just ax | otherwise = Nothing isBuiltInSynFamTyCon_maybe :: TyCon -> Maybe BuiltInSynFamily ===================================== compiler/GHC/Iface/Decl.hs ===================================== @@ -201,9 +201,8 @@ tyConToIfaceDecl env tycon to_if_fam_flav AbstractClosedSynFamilyTyCon = IfaceAbstractClosedSynFamilyTyCon to_if_fam_flav (DataFamilyTyCon {}) = IfaceDataFamilyTyCon to_if_fam_flav (BuiltInSynFamTyCon {}) = IfaceBuiltInSynFamTyCon - to_if_fam_flav (ClosedSynFamilyTyCon Nothing) = IfaceClosedSynFamilyTyCon Nothing - to_if_fam_flav (ClosedSynFamilyTyCon (Just ax)) - = IfaceClosedSynFamilyTyCon (Just (axn, ibr)) + to_if_fam_flav (ClosedSynFamilyTyCon ax) + = IfaceClosedSynFamilyTyCon (axn, ibr) where defs = fromBranches $ coAxiomBranches ax lhss = map coAxBranchLHS defs ibr = map (coAxBranchToIfaceBranch tycon lhss) defs ===================================== compiler/GHC/Iface/Rename.hs ===================================== @@ -518,9 +518,9 @@ rnIfaceClassBody d at IfConcreteClass{} = do return d { ifClassCtxt = ctxt, ifATs = ats, ifSigs = sigs } rnIfaceFamTyConFlav :: Rename IfaceFamTyConFlav -rnIfaceFamTyConFlav (IfaceClosedSynFamilyTyCon (Just (n, axs))) - = IfaceClosedSynFamilyTyCon . Just <$> ((,) <$> rnIfaceNeverExported n - <*> mapM rnIfaceAxBranch axs) +rnIfaceFamTyConFlav (IfaceClosedSynFamilyTyCon (n, axs)) + = IfaceClosedSynFamilyTyCon <$> ((,) <$> rnIfaceNeverExported n + <*> mapM rnIfaceAxBranch axs) rnIfaceFamTyConFlav flav = pure flav rnIfaceAT :: Rename IfaceAT ===================================== compiler/GHC/Iface/Syntax.hs ===================================== @@ -237,9 +237,8 @@ data IfaceTyConParent data IfaceFamTyConFlav = IfaceDataFamilyTyCon -- Data family | IfaceOpenSynFamilyTyCon - | IfaceClosedSynFamilyTyCon (Maybe (IfExtName, [IfaceAxBranch])) + | IfaceClosedSynFamilyTyCon (IfExtName, [IfaceAxBranch]) -- ^ Name of associated axiom and branches for pretty printing purposes, - -- or 'Nothing' for an empty closed family without an axiom -- See Note [Pretty printing via Iface syntax] in "GHC.Types.TyThing.Ppr" | IfaceAbstractClosedSynFamilyTyCon | IfaceBuiltInSynFamTyCon -- for pretty printing purposes only @@ -1118,7 +1117,7 @@ pprIfaceDecl ss (IfaceFamily { ifName = tycon pp_rhs IfaceBuiltInSynFamTyCon = ppShowIface ss (text "built-in") - pp_branches (IfaceClosedSynFamilyTyCon (Just (ax, brs))) + pp_branches (IfaceClosedSynFamilyTyCon (ax, brs)) = vcat (unzipWith (pprAxBranch (pprPrefixIfDeclBndr (ss_how_much ss) @@ -1708,9 +1707,8 @@ freeNamesIfIdDetails IfDFunId = emptyNameSet freeNamesIfFamFlav :: IfaceFamTyConFlav -> NameSet freeNamesIfFamFlav IfaceOpenSynFamilyTyCon = emptyNameSet freeNamesIfFamFlav IfaceDataFamilyTyCon = emptyNameSet -freeNamesIfFamFlav (IfaceClosedSynFamilyTyCon (Just (ax, br))) +freeNamesIfFamFlav (IfaceClosedSynFamilyTyCon (ax, br)) = unitNameSet ax &&& fnList freeNamesIfAxBranch br -freeNamesIfFamFlav (IfaceClosedSynFamilyTyCon Nothing) = emptyNameSet freeNamesIfFamFlav IfaceAbstractClosedSynFamilyTyCon = emptyNameSet freeNamesIfFamFlav IfaceBuiltInSynFamTyCon = emptyNameSet ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -755,12 +755,13 @@ tc_iface_decl parent _ (IfaceFamily {ifName = tc_name, tc_fam_flav :: Name -> IfaceFamTyConFlav -> IfL FamTyConFlav tc_fam_flav tc_name IfaceDataFamilyTyCon - = do { tc_rep_name <- newTyConRepName tc_name - ; return (DataFamilyTyCon tc_rep_name) } - tc_fam_flav _ IfaceOpenSynFamilyTyCon= return OpenSynFamilyTyCon - tc_fam_flav _ (IfaceClosedSynFamilyTyCon mb_ax_name_branches) - = do { ax <- traverse (tcIfaceBranchedAxiom . fst) mb_ax_name_branches - ; return (ClosedSynFamilyTyCon ax) } + = do { tc_rep_name <- newTyConRepName tc_name + ; return (DataFamilyTyCon tc_rep_name) } + tc_fam_flav _ IfaceOpenSynFamilyTyCon + = return OpenSynFamilyTyCon + tc_fam_flav _ (IfaceClosedSynFamilyTyCon (ax_name, _)) + = do { ax <- tcIfaceBranchedAxiom ax_name + ; return (ClosedSynFamilyTyCon ax) } tc_fam_flav _ IfaceAbstractClosedSynFamilyTyCon = return AbstractClosedSynFamilyTyCon tc_fam_flav _ IfaceBuiltInSynFamTyCon @@ -852,20 +853,22 @@ tc_iface_decl _parent ignore_prags tc_iface_decl _ _ (IfaceAxiom { ifName = tc_name, ifTyCon = tc , ifAxBranches = branches, ifRole = role }) = do { tc_tycon <- tcIfaceTyCon tc - -- Must be done lazily, because axioms are forced when checking - -- for family instance consistency, and the RHS may mention - -- a hs-boot declared type constructor that is going to be - -- defined by this module. - -- e.g. type instance F Int = ToBeDefined - -- See #13803 - ; tc_branches <- forkM (text "Axiom branches" <+> ppr tc_name) - $ tc_ax_branches branches - ; let axiom = CoAxiom { co_ax_unique = nameUnique tc_name - , co_ax_name = tc_name - , co_ax_tc = tc_tycon - , co_ax_role = role - , co_ax_branches = manyBranches tc_branches - , co_ax_implicit = False } + ; axiom <- fixM $ \ (axiom :: CoAxiom Branched) -> + -- Might actually be unbranched + -- fixM: just knot-tying for cab_ax_rule + do { tc_branches <- forkM (text "Axiom branches" <+> ppr tc_name) $ + tc_ax_branches (pairWithAxiomRules axiom branches) + -- forkM: Must be done lazily, because axioms are forced when + -- checking for family instance consistency, and the RHS may + -- mention a hs-boot declared type constructor that is going + -- to be defined by this module (see #13803) + -- e.g. type instance F Int = ToBeDefined + ; return (CoAxiom { co_ax_unique = nameUnique tc_name + , co_ax_name = tc_name + , co_ax_tc = tc_tycon + , co_ax_role = role + , co_ax_branches = manyBranches tc_branches + , co_ax_implicit = False }) } ; return (ACoAxiom axiom) } tc_iface_decl _ _ (IfacePatSyn{ ifName = name @@ -1068,16 +1071,16 @@ tc_fd (tvs1, tvs2) = do { tvs1' <- mapM tcIfaceTyVar tvs1 ; tvs2' <- mapM tcIfaceTyVar tvs2 ; return (tvs1', tvs2') } -tc_ax_branches :: [IfaceAxBranch] -> IfL [CoAxBranch] +tc_ax_branches :: [(IfaceAxBranch, CoAxiomRule)] -> IfL [CoAxBranch] tc_ax_branches if_branches = foldlM tc_ax_branch [] if_branches -tc_ax_branch :: [CoAxBranch] -> IfaceAxBranch -> IfL [CoAxBranch] +tc_ax_branch :: [CoAxBranch] -> (IfaceAxBranch, CoAxiomRule) -> IfL [CoAxBranch] tc_ax_branch prev_branches (IfaceAxBranch { ifaxbTyVars = tv_bndrs , ifaxbEtaTyVars = eta_tv_bndrs , ifaxbCoVars = cv_bndrs , ifaxbLHS = lhs, ifaxbRHS = rhs - , ifaxbRoles = roles, ifaxbIncomps = incomps }) + , ifaxbRoles = roles, ifaxbIncomps = incomps }, ax_rule) = bindIfaceTyConBinders_AT (map (\b -> Bndr (IfaceTvBndr b) (NamedTCB Inferred)) tv_bndrs) $ \ tvs -> -- The _AT variant is needed here; see Note [CoAxBranch type variables] in GHC.Core.Coercion.Axiom @@ -1095,7 +1098,8 @@ tc_ax_branch prev_branches , cab_lhs = tc_lhs , cab_roles = roles , cab_rhs = tc_rhs - , cab_incomps = map (prev_branches `getNth`) incomps } + , cab_incomps = map (prev_branches `getNth`) incomps + , cab_axr = ax_rule } ; return (prev_branches ++ [br]) } tcIfaceDataCons :: Name -> TyCon -> [TyConBinder] -> IfaceConDecls -> IfL AlgTyConRhs ===================================== compiler/GHC/Tc/Module.hs ===================================== @@ -1422,13 +1422,10 @@ compatCon c1 c2 ; check (eqType (dataConWrapperType c1) (dataConWrapperType c2)) MismatchedDataConTypes } -eqClosedFamilyAx :: Maybe (CoAxiom br) -> Maybe (CoAxiom br1) +eqClosedFamilyAx :: CoAxiom br -> CoAxiom br1 -> BootErrsM BootTyConMismatch -eqClosedFamilyAx Nothing Nothing = checkSuccess -eqClosedFamilyAx Nothing (Just _) = bootErr $ TyConAxiomMismatch $ NE.singleton MismatchedLength -eqClosedFamilyAx (Just _) Nothing = bootErr $ TyConAxiomMismatch $ NE.singleton MismatchedLength -eqClosedFamilyAx (Just (CoAxiom { co_ax_branches = branches1 })) - (Just (CoAxiom { co_ax_branches = branches2 })) +eqClosedFamilyAx (CoAxiom { co_ax_branches = branches1 }) + (CoAxiom { co_ax_branches = branches2 }) = checkListBy eqClosedFamilyBranch branch_list1 branch_list2 TyConAxiomMismatch where ===================================== compiler/GHC/Tc/TyCl.hs ===================================== @@ -2978,28 +2978,31 @@ tcFamDecl1 parent (FamilyDecl { fdInfo = fam_info False {- this doesn't matter here -} ClosedTypeFamilyFlavour - ; (branches, axiom_validity_infos) <- - unzip <$> mapAndReportM (tcTyFamInstEqn tc_fam_tc NotAssociated) eqns - -- Do not attempt to drop equations dominated by earlier - -- ones here; in the case of mutual recursion with a data - -- type, we get a knot-tying failure. Instead we check - -- for this afterwards, in GHC.Tc.Validity.checkValidCoAxiom - -- Example: tc265 - -- Create a CoAxiom, with the correct src location. ; co_ax_name <- newFamInstAxiomName tc_lname [] - ; let mb_co_ax - | null eqns = Nothing -- mkBranchedCoAxiom fails on empty list - | otherwise = Just (mkBranchedCoAxiom co_ax_name fam_tc branches) + ; (fam_tc, _, ax_validity_infos) <- fixM $ \ ~(_, co_ax, _) -> + -- fixM: just knot-tying, to pass co_ax to pairWithAxiomRule + do { (branches, ax_validity_infos) <- + unzip <$> mapAndReportM (tcTyFamInstEqn tc_fam_tc NotAssociated) + (pairWithAxiomRules co_ax eqns) + -- Do not attempt to drop equations dominated by earlier + -- ones here; in the case of mutual recursion with a data + -- type, we get a knot-tying failure. Instead we check + -- for this afterwards, in GHC.Tc.Validity.checkValidCoAxiom + -- Example: tc265 + + ; let fam_tc = mkFamilyTyCon tc_name tc_bndrs res_kind (resultVariableName sig) + (ClosedSynFamilyTyCon co_ax) parent inj' + co_ax = mkBranchedCoAxiom co_ax_name fam_tc branches - fam_tc = mkFamilyTyCon tc_name tc_bndrs res_kind (resultVariableName sig) - (ClosedSynFamilyTyCon mb_co_ax) parent inj' + ; return ( fam_tc, co_ax, ax_validity_infos ) } -- We check for instance validity later, when doing validity -- checking for the tycon. Exception: checking equations -- overlap done by dropDominatedAxioms - ; return (fam_tc, axiom_validity_infos) } } + + ; return (fam_tc, ax_validity_infos) } } -- | Maybe return a list of Bools that say whether a type family was declared -- injective in the corresponding type arguments. Length of the list is equal to @@ -3206,7 +3209,8 @@ kcTyFamInstEqn tc_fam_tc } -------------------------- -tcTyFamInstEqn :: TcTyCon -> AssocInstInfo -> LTyFamInstEqn GhcRn +tcTyFamInstEqn :: TcTyCon -> AssocInstInfo + -> (LTyFamInstEqn GhcRn, KnotTied CoAxiomRule) -> TcM (KnotTied CoAxBranch, TyFamEqnValidityInfo) -- Needs to be here, not in GHC.Tc.TyCl.Instance, because closed families -- (typechecked here) have TyFamInstEqns @@ -3215,7 +3219,7 @@ tcTyFamInstEqn fam_tc mb_clsinfo (L loc (FamEqn { feqn_tycon = L _ eqn_tc_name , feqn_bndrs = outer_bndrs , feqn_pats = hs_pats - , feqn_rhs = hs_rhs_ty })) + , feqn_rhs = hs_rhs_ty }), ax_rule) = setSrcSpanA loc $ do { traceTc "tcTyFamInstEqn" $ vcat [ ppr loc, ppr fam_tc <+> ppr hs_pats @@ -3234,7 +3238,7 @@ tcTyFamInstEqn fam_tc mb_clsinfo ; let ax = mkCoAxBranch qtvs [] [] pats rhs_ty (map (const Nominal) qtvs) - (locA loc) + (locA loc) ax_rule vi = VI { vi_loc = locA loc , vi_qtvs = qtvs , vi_non_user_tvs = non_user_tvs @@ -4505,10 +4509,9 @@ checkValidTyCon tc | Just fam_flav <- famTyConFlav_maybe tc -> case fam_flav of - { ClosedSynFamilyTyCon (Just ax) + { ClosedSynFamilyTyCon ax -> tcAddClosedTypeFamilyDeclCtxt tc $ checkValidCoAxiom ax - ; ClosedSynFamilyTyCon Nothing -> return () ; AbstractClosedSynFamilyTyCon -> do { hsBoot <- tcIsHsBootOrSig ; checkTc hsBoot $ TcRnAbstractClosedTyFamDecl } ===================================== compiler/GHC/Tc/TyCl/Build.hs ===================================== @@ -50,8 +50,8 @@ mkNewTyConRhs :: Name -> TyCon -> DataCon -> TcRnIf m n AlgTyConRhs -- We pass the Name of the parent TyCon, as well as the TyCon itself, -- because the latter is part of a knot, whereas the former is not. mkNewTyConRhs tycon_name tycon con - = do { co_tycon_name <- newImplicitBinder tycon_name mkNewTyCoOcc - ; let nt_ax = mkNewTypeCoAxiom co_tycon_name tycon etad_tvs etad_roles etad_rhs + = do { co_ax_name <- newImplicitBinder tycon_name mkNewTyCoOcc + ; let nt_ax = mkNewTypeCoAxiom co_ax_name tycon etad_tvs etad_roles etad_rhs ; traceIf (text "mkNewTyConRhs" <+> ppr nt_ax) ; return (NewTyCon { data_con = con, nt_rhs = rhs_ty, ===================================== compiler/GHC/Tc/TyCl/Instance.hs ===================================== @@ -605,22 +605,25 @@ tcTyFamInstDecl mb_clsinfo (L loc decl@(TyFamInstDecl { tfid_eqn = eqn })) TcRnIllegalInstance $ IllegalFamilyInstance $ NotAnOpenFamilyTyCon fam_tc - -- (1) do the work of verifying the synonym group - -- For some reason we don't have a location for the equation - -- itself, so we make do with the location of family name - ; (co_ax_branch, co_ax_validity_info) - <- tcTyFamInstEqn fam_tc mb_clsinfo - (L (l2l $ getLoc fam_lname) eqn) - - -- (2) check for validity + ; (co_ax,co_ax_branch, co_ax_validity_info) <- fixM $ \ ~(co_ax, _, _) -> + -- fixM: just knot-tying + do { let eqn_w_loc = L (l2l $ getLoc fam_lname) eqn + -- eqn_w_loc: for some reason we don't have a location for the + -- equation itself, so we make do with the location of family name + + ; (co_ax_branch, co_ax_validity_info) <- tcTyFamInstEqn fam_tc mb_clsinfo + (eqn_w_loc, UnbranchedAxiom co_ax) + + ; rep_tc_name <- newFamInstAxiomName fam_lname [coAxBranchLHS co_ax_branch] + ; let co_ax = mkUnbranchedCoAxiom rep_tc_name fam_tc co_ax_branch + ; return (co_ax, co_ax_branch, co_ax_validity_info) } + + -- Check for validity ; checkConsistentFamInst mb_clsinfo fam_tc co_ax_branch ; checkTyFamEqnValidityInfo fam_tc co_ax_validity_info ; checkValidCoAxBranch fam_tc co_ax_branch - -- (3) construct coercion axiom - ; rep_tc_name <- newFamInstAxiomName fam_lname [coAxBranchLHS co_ax_branch] - ; let axiom = mkUnbranchedCoAxiom rep_tc_name fam_tc co_ax_branch - ; newFamInst SynFamilyInst axiom } + ; newFamInst SynFamilyInst co_ax } --------------------- ===================================== compiler/GHC/Tc/Utils/Backpack.hs ===================================== @@ -410,9 +410,8 @@ thinModIface avails iface = ifaceDeclNeverExportedRefs :: IfaceDecl -> [Name] ifaceDeclNeverExportedRefs d at IfaceFamily{} = case ifFamFlav d of - IfaceClosedSynFamilyTyCon (Just (n, _)) - -> [n] - _ -> [] + IfaceClosedSynFamilyTyCon (n, _) -> [n] + _ -> [] ifaceDeclNeverExportedRefs _ = [] ===================================== compiler/Language/Haskell/Syntax/Decls.hs ===================================== @@ -858,7 +858,9 @@ data InjectivityAnn pass data FamilyInfo pass = DataFamily + | OpenTypeFamily + -- | 'Nothing' if we're in an hs-boot file and the user -- said "type family Foo x where .." | ClosedTypeFamily (Maybe [LTyFamInstEqn pass]) @@ -869,11 +871,11 @@ familyInfoTyConFlavour -> TyConFlavour tc familyInfoTyConFlavour mb_parent_tycon info = case info of - DataFamily -> OpenFamilyFlavour IAmData mb_parent_tycon - OpenTypeFamily -> OpenFamilyFlavour IAmType mb_parent_tycon - ClosedTypeFamily _ -> assert (isNothing mb_parent_tycon) - -- See Note [Closed type family mb_parent_tycon] - ClosedTypeFamilyFlavour + DataFamily -> OpenFamilyFlavour IAmData mb_parent_tycon + OpenTypeFamily -> OpenFamilyFlavour IAmType mb_parent_tycon + ClosedTypeFamily {} -> assert (isNothing mb_parent_tycon) + -- See Note [Closed type family mb_parent_tycon] + ClosedTypeFamilyFlavour {- Note [Closed type family mb_parent_tycon] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -5424,8 +5424,8 @@ module GHC.Event.Windows.ManagedThreadPool where thrCallBack :: GHC.Internal.Event.Windows.ManagedThreadPool.WorkerJob, thrActiveThreads :: GHC.Internal.MVar.MVar GHC.Types.Int, thrMonitor :: GHC.Internal.MVar.MVar (), - thrThreadIds :: ! {-# UNPACK #-}(ghc-internal-0.1.0.0:GHC.Internal.Event.Array.N:Array[0] _P - ; GHC.Internal.IORef.N:IORef[0] _N)(ghc-internal-0.1.0.0:GHC.Internal.Event.Array.Array GHC.Internal.Conc.Sync.ThreadId)} + thrThreadIds :: ! {-# UNPACK #-}(ghc-internal-0.1.0.0:GHC.Internal.Event.Array.N:Array _P + ; GHC.Internal.IORef.N:IORef _N)(ghc-internal-0.1.0.0:GHC.Internal.Event.Array.Array GHC.Internal.Conc.Sync.ThreadId)} monitorThreadPool :: GHC.Internal.MVar.MVar () -> GHC.Types.IO () notifyRunning :: GHC.Internal.Maybe.Maybe ThreadPool -> GHC.Types.IO () notifyWaiting :: GHC.Internal.Maybe.Maybe ThreadPool -> GHC.Types.IO () ===================================== utils/haddock/haddock-api/src/Haddock/Convert.hs ===================================== @@ -291,14 +291,9 @@ synifyTyCon _prr _coax tc case flav of -- Type families OpenSynFamilyTyCon -> mkFamDecl OpenTypeFamily - ClosedSynFamilyTyCon mb - | Just (CoAxiom{co_ax_branches = branches}) <- mb -> - mkFamDecl $ - ClosedTypeFamily $ - Just $ - map (noLocA . synifyAxBranch tc) (fromBranches branches) - | otherwise -> - mkFamDecl $ ClosedTypeFamily $ Just [] + ClosedSynFamilyTyCon (CoAxiom{co_ax_branches = branches}) -> + mkFamDecl $ ClosedTypeFamily $ + Just $ map (noLocA . synifyAxBranch tc) (fromBranches branches) BuiltInSynFamTyCon{} -> mkFamDecl $ ClosedTypeFamily $ Just [] AbstractClosedSynFamilyTyCon{} -> View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a1f66dcb6e8d17f2c89b2a82f6a822fd64817886...e4bdc821f1e4863074a73cc032082dd2251a80f8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a1f66dcb6e8d17f2c89b2a82f6a822fd64817886...e4bdc821f1e4863074a73cc032082dd2251a80f8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 15:41:47 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Wed, 11 Sep 2024 11:41:47 -0400 Subject: [Git][ghc/ghc][wip/T24359] Fix build Message-ID: <66e1ba3b26836_d7cbf4bd700105094@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24359 at Glasgow Haskell Compiler / GHC Commits: c412451f by Simon Peyton Jones at 2024-09-11T16:41:22+01:00 Fix build - - - - - 8 changed files: - compiler/GHC/HsToCore.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Rename/Module.hs - compiler/GHC/Tc/Gen/Sig.hs - compiler/GHC/Types/Error/Codes.hs - testsuite/tests/typecheck/should_compile/tc186.hs Changes: ===================================== compiler/GHC/HsToCore.hs ===================================== @@ -293,7 +293,7 @@ deSugar hsc_env dsImpSpecs :: [LTcSpecPrag] -> DsM (OrdList (Id,CoreExpr), [CoreRule]) dsImpSpecs imp_specs - = do { spec_prs <- mapMaybeM (dsSpec Nothing) imp_specs + = do { spec_prs <- mapMaybeM (dsSpec Nothing . unLoc) imp_specs ; let (spec_binds, spec_rules) = unzip spec_prs ; return (concatOL spec_binds, spec_rules) } ===================================== compiler/GHC/HsToCore/Binds.hs ===================================== @@ -805,7 +805,7 @@ dsSpecs :: CoreExpr -- Its rhs -- See Note [Overview of SPECIALISE pragmas] in GHC.Tc.Gen.Sig dsSpecs _ IsDefaultMethod = return (nilOL, []) dsSpecs poly_rhs (SpecPrags sps) - = do { pairs <- mapMaybeM (dsSpec (Just poly_rhs)) sps + = do { pairs <- mapMaybeM (dsLSpec (Just poly_rhs)) sps ; let (spec_binds_s, rules) = unzip pairs ; return (concatOL spec_binds_s, rules) } @@ -829,7 +829,7 @@ dsSpec mb_poly_rhs (SpecPrag poly_id spec_co spec_inl) -- \spec_bndrs. [] spec_args -- perhaps with the body of the lambda wrapped in some WpLets -- E.g. /\a \(d:Eq a). let d2 = $df d in [] (Maybe a) d2 - = dsHsWrapper spec_app $ \core_app -> + = dsHsWrapperForRuleLHS spec_app $ \core_app -> finishSpecPrag mb_poly_rhs spec_bndrs (core_app (Var poly_id)) spec_bndrs (\_ poly_rhs -> core_app poly_rhs) @@ -840,7 +840,7 @@ dsSpec mb_poly_rhs (SpecPragE { spe_poly_id = poly_id , spe_id_bndrs = id_bndrs , spe_lhs_ev_bndrs = lhs_evs , spe_lhs_binds = lhs_binds - , spe_call = the_call + , spe_lhs_call = the_call , spe_rhs_ev_bndrs = rhs_evs , spe_rhs_binds = rhs_binds , spe_inl = inl }) @@ -869,7 +869,7 @@ failBecauseOfClassOp :: Id -> DsM (Maybe a) -- There is no point in trying to specialise a class op -- Moreover, classops don't (currently) have an inl_sat arity set -- (it would be Just 0) and that in turn makes makeCorePair bleat -failBecauseOfClassOp loc poly_id +failBecauseOfClassOp poly_id = do { diagnosticDs (DsUselessSpecialiseForClassMethodSelector poly_id) ; return Nothing } @@ -919,9 +919,7 @@ finishSpecPrag mb_poly_rhs ; tracePm "dsSpec" (vcat [ text "fun:" <+> ppr poly_id - , text "spec_co:" <+> ppr spec_co , text "spec_bndrs:" <+> ppr spec_bndrs - , text "ds_lhs:" <+> ppr ds_lhs , text "args:" <+> ppr rule_lhs_args ]) ; return (Just (unitOL (spec_id, spec_rhs), rule)) -- NB: do *not* use makeCorePair on (spec_id,spec_rhs), because @@ -1261,7 +1259,7 @@ drop_dicts drops dictionary bindings on the LHS where possible. Of course, the ($dfEqlist d) in the pattern makes it less likely to match, but there is no other way to get d:Eq a - NB 2: We do drop_dicts *before* simplOptEpxr, so that we expect all + NB 2: We do drop_dicts *before* simplOptExpr, so that we expect all the evidence bindings to be wrapped around the outside of the LHS. (After simplOptExpr they'll usually have been inlined.) dsHsWrapper does dependency analysis, so that civilised ones ===================================== compiler/GHC/Parser/Errors/Ppr.hs ===================================== @@ -695,7 +695,6 @@ instance Diagnostic PsMessage where PsErrInvalidPun {} -> ErrorWithoutFlag PsErrIllegalOrPat{} -> ErrorWithoutFlag PsErrTypeSyntaxInPat{} -> ErrorWithoutFlag - PsErrSpecEpxrMultipleTypeAscription{} -> ErrorWithoutFlag PsErrSpecExprMultipleTypeAscription{} -> ErrorWithoutFlag diagnosticHints = \case @@ -866,7 +865,6 @@ instance Diagnostic PsMessage where PsErrInvalidPun {} -> [suggestExtension LangExt.ListTuplePuns] PsErrIllegalOrPat{} -> [suggestExtension LangExt.OrPatterns] PsErrTypeSyntaxInPat{} -> noHints - PsErrSpecEpxrMultipleTypeAscription {} -> noHints PsErrSpecExprMultipleTypeAscription {} -> noHints diagnosticCode = constructorCode ===================================== compiler/GHC/Parser/Errors/Types.hs ===================================== @@ -491,7 +491,6 @@ data PsMessage -- T24159_pat_parse_error_6 | PsErrTypeSyntaxInPat !PsErrTypeSyntaxDetails - | PsErrSpecEpxrMultipleTypeAscription | PsErrSpecExprMultipleTypeAscription deriving Generic ===================================== compiler/GHC/Rename/Module.hs ===================================== @@ -24,8 +24,7 @@ import {-# SOURCE #-} GHC.Rename.Expr( rnLExpr ) import {-# SOURCE #-} GHC.Rename.Splice ( rnSpliceDecl, rnTopSpliceDecls ) import GHC.Hs -import GHC.Types.FieldLabel -import GHC.Types.Name.Reader + import GHC.Rename.HsType import GHC.Rename.Bind import GHC.Rename.Doc @@ -37,6 +36,7 @@ import GHC.Rename.Utils ( mapFvRn, bindLocalNames , addNoNestedForallsContextsErr, checkInferredVars ) import GHC.Rename.Unbound ( mkUnboundName, notInScopeErr, WhereLooking(WL_Global) ) import GHC.Rename.Names + import GHC.Tc.Errors.Types import GHC.Tc.Gen.Annotation ( annCtxt ) import GHC.Tc.Utils.Monad @@ -50,29 +50,28 @@ import GHC.Builtin.Names( applicativeClassName, pureAName, thenAName , monoidClassName, mappendName ) +import GHC.Types.FieldLabel +import GHC.Types.Name.Reader import GHC.Types.ForeignCall ( CCallTarget(..) ) import GHC.Types.Name import GHC.Types.Name.Set import GHC.Types.Name.Env -import GHC.Types.Basic ( TypeOrKind(..), RuleName ) +import GHC.Types.Basic ( Arity, TypeOrKind(..), RuleName ) import GHC.Types.GREInfo (ConLikeInfo (..), ConInfo, mkConInfo, conInfoFields) - - +import GHC.Types.Unique.Set import GHC.Types.SrcLoc as SrcLoc + import GHC.Driver.DynFlags -import GHC.Utils.Misc ( lengthExceeds, partitionWith ) -import GHC.Utils.Panic import GHC.Driver.Env ( HscEnv(..), hsc_home_unit) -import GHC.Types.Unique.Set +import GHC.Utils.Misc ( lengthExceeds, partitionWith ) +import GHC.Utils.Panic import GHC.Utils.Outputable import GHC.Data.FastString -import GHC.Data.Bag import GHC.Data.List.SetOps ( findDupsEq, removeDupsOn, equivClasses ) import GHC.Data.Graph.Directed ( SCC, flattenSCC, flattenSCCs, Node(..) , stronglyConnCompFromEdgedVerticesUniq ) - import GHC.Data.OrdList import qualified GHC.LanguageExtensions as LangExt import GHC.Core.DataCon ( isSrcStrict ) ===================================== compiler/GHC/Tc/Gen/Sig.hs ===================================== @@ -938,7 +938,7 @@ tcSpecPrag poly_id (SpecSigE nm bndrs spec_e inl) , spe_id_bndrs = id_bndrs , spe_lhs_ev_bndrs = lhs_evs , spe_lhs_binds = lhs_binds - , spe_call = spec_e' + , spe_lhs_call = spec_e' , spe_rhs_ev_bndrs = rhs_evs , spe_rhs_binds = rhs_binds , spe_inl = inl }] } @@ -1173,11 +1173,12 @@ tcRule (HsRule { rd_ext = ext , rd_bndrs = mkTcRuleBndrs bndrs (qtkvs ++ tpl_ids) , rd_lhs = mkHsDictLet lhs_binds lhs' , rd_rhs = mkHsDictLet rhs_binds rhs' } } - -mkTcRuleBndrs (RuleBndrs { rb_tyvs = tyvs }) vars - = RuleBndrs { rb_ext = noAnn - , rb_tyvs = tyvs -- preserved for ppr-ing - , rb_tmvs = map (noLocA . RuleBndr noAnn . noLocA) vars } + where + mkTcRuleBndrs (RuleBndrs { rb_tyvs = tyvs }) vars + = RuleBndrs { rb_ext = noAnn + , rb_tyvs = tyvs -- preserved for ppr-ing + , rb_tmvs = map (noLocA . RuleBndr noAnn . noLocA) vars } + mkTcRuleBndrs (XRuleBndrs {}) _ = panic "mkTCRuleBndrs" generateRuleConstraints :: SkolemInfo -> RuleBndrs GhcRn ===================================== compiler/GHC/Types/Error/Codes.hs ===================================== @@ -291,7 +291,6 @@ type family GhcDiagnosticCode c = n | n -> c where GhcDiagnosticCode "PsErrInvalidPun" = 52943 GhcDiagnosticCode "PsErrIllegalOrPat" = 29847 GhcDiagnosticCode "PsErrTypeSyntaxInPat" = 32181 - GhcDiagnosticCode "PsErrSpecEpxrMultipleTypeAscription" = 62037 GhcDiagnosticCode "PsErrSpecExprMultipleTypeAscription" = 62037 -- Driver diagnostic codes ===================================== testsuite/tests/typecheck/should_compile/tc186.hs ===================================== @@ -2,7 +2,7 @@ -- Killed 6.2.2 -- The trouble was that 1 was instantiated to a type (t::?) -- and the constraint (Foo (t::? -> s::*)) didn't match Foo (a::* -> b::*). --- Solution is to zap the expected type in TcEpxr.tc_expr(HsOverLit). +-- Solution is to zap the expected type in TcExpr.tc_expr(HsOverLit). module ShouldCompile where View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c412451fcf9960df43ecd24b1c8a384a77a1ac31 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c412451fcf9960df43ecd24b1c8a384a77a1ac31 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 16:46:41 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Wed, 11 Sep 2024 12:46:41 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] 30 commits: AArch64: Implement takeRegRegMoveInstr Message-ID: <66e1c9719f045_d7cbf84380c11287@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 59a75bb5 by Sven Tennie at 2024-09-11T16:46:37+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - fd507e82 by Sven Tennie at 2024-09-11T16:46:37+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - c98e7596 by Sven Tennie at 2024-09-11T16:46:37+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 2e8ff687 by Sven Tennie at 2024-09-11T16:46:37+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - 56fe71de by Sven Tennie at 2024-09-11T16:46:37+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 304563ae by Sven Tennie at 2024-09-11T16:46:37+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 4442dd29 by Sven Tennie at 2024-09-11T16:46:37+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 9d562390 by Sven Tennie at 2024-09-11T16:46:37+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - d49701ae by Sven Tennie at 2024-09-11T16:46:37+00:00 RISCV64: Set codeowners of the NCG - - - - - fddd2272 by Sven Tennie at 2024-09-11T16:46:37+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - 13 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/eab0e738eb8f4a69b0299455e3336a579f4a7e98...fddd2272d7ac4edc3083ec284a5652624475932c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/eab0e738eb8f4a69b0299455e3336a579f4a7e98...fddd2272d7ac4edc3083ec284a5652624475932c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 17:10:51 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Wed, 11 Sep 2024 13:10:51 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] 2 commits: wrk: add guards for deadbeef Message-ID: <66e1cf1bdbb41_d7cbf9dabc011402c@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 34807a34 by doyougnu at 2024-09-11T08:32:47-04:00 wrk: add guards for deadbeef - - - - - a0e8db32 by doyougnu at 2024-09-11T09:26:13-04:00 wrk: add rts option -link-unknown-symbol - - - - - 5 changed files: - libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc - rts/Linker.c - rts/RtsFlags.c - rts/include/rts/Flags.h - rts/linker/elf_got.c Changes: ===================================== libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc ===================================== @@ -162,6 +162,7 @@ data MiscFlags = MiscFlags , disableDelayedOsMemoryReturn :: Bool , internalCounters :: Bool , linkerAlwaysPic :: Bool + , linkUnknownSymbols :: Bool , linkerMemBase :: Word -- ^ address to ask the OS for memory for the linker, 0 ==> off , ioManager :: IoManagerFlag @@ -536,6 +537,8 @@ getMiscFlags = do (#{peek MISC_FLAGS, internalCounters} ptr :: IO CBool)) <*> (toBool <$> (#{peek MISC_FLAGS, linkerAlwaysPic} ptr :: IO CBool)) + <*> (toBool <$> + (#{peek MISC_FLAGS, linkUnknownSymbols} ptr :: IO CBool)) <*> #{peek MISC_FLAGS, linkerMemBase} ptr <*> (toEnum . fromIntegral <$> (#{peek MISC_FLAGS, ioManager} ptr :: IO Word32)) ===================================== rts/Linker.c ===================================== @@ -950,7 +950,13 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) "See top entry above.\n", lbl); IF_DEBUG(linker, printLoadedObjects()); fflush(stderr); - r = 0xDEADBEEF; + + // if -link-unknown-symbols is passed into the RTS we allow the linker + // to optimistically continue + if (RtsFlags.MiscFlags.linkUnknownSymbols){ + r = 0xDEADBEEF; + } + } if (!runPendingInitializers()) { ===================================== rts/RtsFlags.c ===================================== @@ -269,6 +269,7 @@ void initRtsFlagsDefaults(void) RtsFlags.MiscFlags.disableDelayedOsMemoryReturn = false; RtsFlags.MiscFlags.internalCounters = false; RtsFlags.MiscFlags.linkerAlwaysPic = DEFAULT_LINKER_ALWAYS_PIC; + RtsFlags.MiscFlags.linkUnknownSymbols = false; RtsFlags.MiscFlags.linkerMemBase = 0; RtsFlags.MiscFlags.ioManager = IO_MNGR_FLAG_AUTO; #if defined(THREADED_RTS) && defined(mingw32_HOST_OS) @@ -998,6 +999,11 @@ error = true; OPTION_UNSAFE; RtsFlags.MiscFlags.generate_dump_file = true; } + else if (strequal("link-unknown-symbols", + &rts_argv[arg][2])) { + OPTION_UNSAFE; + RtsFlags.MiscFlags.linkUnknownSymbols = true; + } else if (strequal("null-eventlog-writer", &rts_argv[arg][2])) { OPTION_UNSAFE; ===================================== rts/include/rts/Flags.h ===================================== @@ -267,6 +267,7 @@ typedef struct _MISC_FLAGS { there as well. */ bool internalCounters; /* See Note [Internal Counters Stats] */ bool linkerAlwaysPic; /* Assume the object code is always PIC */ + bool linkUnknownSymbols; /* Should the runtime linker optimistically continue */ StgWord linkerMemBase; /* address to ask the OS for memory * for the linker, NULL ==> off */ IO_MANAGER_FLAG ioManager; /* The I/O manager to use. */ ===================================== rts/linker/elf_got.c ===================================== @@ -99,8 +99,16 @@ fillGot(ObjectCode * oc) { } else { errorBelch("Failed to lookup symbol: %s\n", symbol->name); - // return EXIT_FAILURE; - symbol->addr = 0xDEADBEEF; + + // if -link-unknown-symbols is passed into the + // RTS we allow the linker to optimistically + // continue + if (RtsFlags.MiscFlags.linkUnknownSymbols) { + symbol->addr = 0xDEADBEEF; + } else { + return EXIT_FAILURE; + } + } } } else { View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ae64a31e00d266934b1656ffc39353cc1f78bbe7...a0e8db32f4e565b5336ec226b853fe3c68845779 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ae64a31e00d266934b1656ffc39353cc1f78bbe7...a0e8db32f4e565b5336ec226b853fe3c68845779 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 17:45:28 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Wed, 11 Sep 2024 13:45:28 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] wrk: correctly cast to the pointer Message-ID: <66e1d73883944_d7cbfbe346c114750@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 6b76535c by doyougnu at 2024-09-11T13:39:01-04:00 wrk: correctly cast to the pointer - - - - - 5 changed files: - rts/Linker.c - rts/linker/elf_got.c - + testsuite/tests/rts/T25240/FFI.hs - + testsuite/tests/rts/T25240/TH.hs - + testsuite/tests/rts/T25240/all.T Changes: ===================================== rts/Linker.c ===================================== @@ -954,7 +954,7 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) // if -link-unknown-symbols is passed into the RTS we allow the linker // to optimistically continue if (RtsFlags.MiscFlags.linkUnknownSymbols){ - r = 0xDEADBEEF; + r = (void*) 0xDEADBEEF; } } ===================================== rts/linker/elf_got.c ===================================== @@ -104,7 +104,7 @@ fillGot(ObjectCode * oc) { // RTS we allow the linker to optimistically // continue if (RtsFlags.MiscFlags.linkUnknownSymbols) { - symbol->addr = 0xDEADBEEF; + symbol->addr = (void*) 0xDEADBEEF; } else { return EXIT_FAILURE; } ===================================== testsuite/tests/rts/T25240/FFI.hs ===================================== @@ -0,0 +1,24 @@ +{-# LANGUAGE GHCForeignImportPrim #-} +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE UnliftedFFITypes #-} + +module T25240.FFI + ( D(..), D2(..), tabulate + ) where + +import GHC.Exts ( Double#, Double(D#) ) +import Language.Haskell.TH ( CodeQ ) + +data D = D Double +data D2 = D2 { x, y :: D } + +tabulate :: ( Word -> CodeQ D ) -> CodeQ ( D2 ) +tabulate f = [|| D2 $$( f 1 ) $$( f 2 ) ||] + + +-- Now an unrelated "Num D" instance. +instance Num D where + ( D ( D# x ) ) * ( D ( D# y ) ) = D ( D# ( func x y ) ) +foreign import prim "prim_func" + func :: Double# -> Double# -> Double# ===================================== testsuite/tests/rts/T25240/TH.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE TemplateHaskell #-} + +module TH where +import FFI + +th_bug :: D2 +th_bug = $$( tabulate $ \ _ -> [|| D 0 ||] ) ===================================== testsuite/tests/rts/T25240/all.T ===================================== @@ -0,0 +1,2 @@ +test('T25240', + [req_rts_linker, extra_files(['FFI.hs', 'TH.hs'])], compile, ['-flink-unknown-symbols']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6b76535c2067bdecec26830bc8077342dbd7776a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6b76535c2067bdecec26830bc8077342dbd7776a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 19:10:33 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Wed, 11 Sep 2024 15:10:33 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] wrk: add test, still failing though Message-ID: <66e1eb293303d_2201f2248f38374fc@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 04077f81 by doyougnu at 2024-09-11T15:10:01-04:00 wrk: add test, still failing though - - - - - 7 changed files: - rts/Linker.c - rts/linker/elf_got.c - − testsuite/tests/rts/T25240/all.T - testsuite/tests/rts/T25240/FFI.hs → testsuite/tests/rts/flags/T25240/FFI.hs - + testsuite/tests/rts/flags/T25240/T25240.hs - testsuite/tests/rts/T25240/TH.hs → testsuite/tests/rts/flags/T25240/TH.hs - + testsuite/tests/rts/flags/T25240/all.T Changes: ===================================== rts/Linker.c ===================================== @@ -951,7 +951,7 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) IF_DEBUG(linker, printLoadedObjects()); fflush(stderr); - // if -link-unknown-symbols is passed into the RTS we allow the linker + // if --link-unknown-symbols is passed into the RTS we allow the linker // to optimistically continue if (RtsFlags.MiscFlags.linkUnknownSymbols){ r = (void*) 0xDEADBEEF; ===================================== rts/linker/elf_got.c ===================================== @@ -100,13 +100,13 @@ fillGot(ObjectCode * oc) { errorBelch("Failed to lookup symbol: %s\n", symbol->name); - // if -link-unknown-symbols is passed into the + // if --link-unknown-symbols is passed into the // RTS we allow the linker to optimistically // continue if (RtsFlags.MiscFlags.linkUnknownSymbols) { symbol->addr = (void*) 0xDEADBEEF; } else { - return EXIT_FAILURE; + return EXIT_FAILURE; } } ===================================== testsuite/tests/rts/T25240/all.T deleted ===================================== @@ -1,2 +0,0 @@ -test('T25240', - [req_rts_linker, extra_files(['FFI.hs', 'TH.hs'])], compile, ['-flink-unknown-symbols']) ===================================== testsuite/tests/rts/T25240/FFI.hs → testsuite/tests/rts/flags/T25240/FFI.hs ===================================== @@ -3,7 +3,7 @@ {-# LANGUAGE TemplateHaskell #-} {-# LANGUAGE UnliftedFFITypes #-} -module T25240.FFI +module FFI ( D(..), D2(..), tabulate ) where @@ -11,7 +11,9 @@ import GHC.Exts ( Double#, Double(D#) ) import Language.Haskell.TH ( CodeQ ) data D = D Double + deriving Show data D2 = D2 { x, y :: D } + deriving Show tabulate :: ( Word -> CodeQ D ) -> CodeQ ( D2 ) tabulate f = [|| D2 $$( f 1 ) $$( f 2 ) ||] ===================================== testsuite/tests/rts/flags/T25240/T25240.hs ===================================== @@ -0,0 +1,5 @@ + +import TH + +main :: IO () +main = print th_bug ===================================== testsuite/tests/rts/T25240/TH.hs → testsuite/tests/rts/flags/T25240/TH.hs ===================================== ===================================== testsuite/tests/rts/flags/T25240/all.T ===================================== @@ -0,0 +1,2 @@ +test('T25240', [req_rts_linker, extra_files(['FFI.hs', 'TH.hs'])], + multimod_compile, ['T25240', '+RTS --link-unknown-symbols -RTS']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/04077f810f82218c85a5cb31a130b28abc1bfde0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/04077f810f82218c85a5cb31a130b28abc1bfde0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 20:52:37 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Wed, 11 Sep 2024 16:52:37 -0400 Subject: [Git][ghc/ghc][wip/az/epa-exactprint-sync] 8 commits: haddock: Re-organise cross-OS compatibility layer Message-ID: <66e2031534c80_1efe5e2e7b10473e5@gitlab.mail> Alan Zimmerman pushed to branch wip/az/epa-exactprint-sync at Glasgow Haskell Compiler / GHC Commits: 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 281da00c by Alan Zimmerman at 2024-09-11T21:33:44+01:00 EPA: Sync ghc-exactprint to GHC - - - - - 26 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Annotation.hs - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Tc/Types.hs - compiler/GHC/Tc/Utils/Monad.hs - compiler/GHC/Tc/Zonk/Type.hs - hadrian/src/Builder.hs - libraries/base/src/GHC/Exts.hs - libraries/ghci/GHCi/CreateBCO.hs - libraries/ghci/ghci.cabal.in - + testsuite/tests/module/T21752.stderr - testsuite/tests/perf/compiler/T11068.stdout - testsuite/tests/pmcheck/should_compile/T12957.stderr - + testsuite/tests/pmcheck/should_compile/T24817.hs - testsuite/tests/pmcheck/should_compile/all.T - testsuite/tests/printer/T17697.stderr - testsuite/tests/profiling/should_run/staticcallstack002.stdout - testsuite/tests/simplCore/should_compile/T13156.stdout - testsuite/tests/typecheck/should_fail/T13292.stderr - utils/check-exact/ExactPrint.hs - utils/check-exact/Main.hs - utils/check-exact/Parsers.hs - utils/check-exact/Transform.hs - utils/check-exact/Types.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/42c9560074c03c74e96c58f792c0d281d73e0eea...281da00ca3cbbb5de4cd5c5b147aaf88553edaef -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/42c9560074c03c74e96c58f792c0d281d73e0eea...281da00ca3cbbb5de4cd5c5b147aaf88553edaef You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 11 22:50:44 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 11 Sep 2024 18:50:44 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 5 commits: hadrian: Make sure ffi headers are built before using a compiler Message-ID: <66e21ec4972ab_1efe5e6b8eec6375c@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - bd15e649 by Andrew Lelechenko at 2024-09-11T18:50:31-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - 1930fd32 by Sebastian Graf at 2024-09-11T18:50:31-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 13 changed files: - compiler/GHC/Parser.y - compiler/GHC/Parser/Annotation.hs - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Unit/Types.hs - docs/users_guide/exts/or_patterns.rst - hadrian/src/Builder.hs - libraries/base/src/GHC/Exts.hs - libraries/deepseq - libraries/ghci/GHCi/CreateBCO.hs - libraries/ghci/ghci.cabal.in - + testsuite/tests/module/T21752.stderr - utils/check-exact/ExactPrint.hs - utils/check-exact/Utils.hs Changes: ===================================== compiler/GHC/Parser.y ===================================== @@ -4483,13 +4483,13 @@ gl = getLoc glA :: HasLoc a => a -> SrcSpan glA = getHasLoc -glR :: HasLoc a => a -> Anchor +glR :: HasLoc a => a -> EpaLocation glR !la = EpaSpan (getHasLoc la) -glEE :: (HasLoc a, HasLoc b) => a -> b -> Anchor +glEE :: (HasLoc a, HasLoc b) => a -> b -> EpaLocation glEE !x !y = spanAsAnchor $ comb2 x y -glRM :: Located a -> Maybe Anchor +glRM :: Located a -> Maybe EpaLocation glRM (L !l _) = Just $ spanAsAnchor l glAA :: HasLoc a => a -> EpaLocation @@ -4609,11 +4609,11 @@ hsDoAnn :: Located a -> LocatedAn t b -> AnnKeywordId -> AnnList hsDoAnn (L l _) (L ll _) kw = AnnList (Just $ spanAsAnchor (locA ll)) Nothing Nothing [AddEpAnn kw (srcSpan2e l)] [] -listAsAnchor :: [LocatedAn t a] -> Located b -> Anchor +listAsAnchor :: [LocatedAn t a] -> Located b -> EpaLocation listAsAnchor [] (L l _) = spanAsAnchor l listAsAnchor (h:_) s = spanAsAnchor (comb2 h s) -listAsAnchorM :: [LocatedAn t a] -> Maybe Anchor +listAsAnchorM :: [LocatedAn t a] -> Maybe EpaLocation listAsAnchorM [] = Nothing listAsAnchorM (L l _:_) = case locA l of ===================================== compiler/GHC/Parser/Annotation.hs ===================================== @@ -23,7 +23,7 @@ module GHC.Parser.Annotation ( TokenLocation(..), DeltaPos(..), deltaPos, getDeltaLine, - EpAnn(..), Anchor, + EpAnn(..), anchor, spanAsAnchor, realSpanAsAnchor, noSpanAnchor, @@ -528,7 +528,7 @@ instance Outputable AddEpAnn where -- new AST fragments out of old ones, and have them still printed out -- in a precise way. data EpAnn ann - = EpAnn { entry :: !Anchor + = EpAnn { entry :: !EpaLocation -- ^ Base location for the start of the syntactic element -- holding the annotations. , anns :: !ann -- ^ Annotations added by the Parser @@ -539,15 +539,6 @@ data EpAnn ann deriving (Data, Eq, Functor) -- See Note [XRec and Anno in the AST] --- | An 'Anchor' records the base location for the start of the --- syntactic element holding the annotations, and is used as the point --- of reference for calculating delta positions for contained --- annotations. --- It is also normally used as the reference point for the spacing of --- the element relative to its container. If the AST element is moved, --- that relationship is tracked using the 'EpaDelta' constructor instead. -type Anchor = EpaLocation -- Transitional - anchor :: (EpaLocation' a) -> RealSrcSpan anchor (EpaSpan (RealSrcSpan r _)) = r anchor _ = panic "anchor" @@ -676,7 +667,7 @@ data AnnListItem -- keywords such as 'where'. data AnnList = AnnList { - al_anchor :: Maybe Anchor, -- ^ start point of a list having layout + al_anchor :: Maybe EpaLocation, -- ^ start point of a list having layout al_open :: Maybe AddEpAnn, al_close :: Maybe AddEpAnn, al_rest :: [AddEpAnn], -- ^ context, such as 'where' keyword @@ -1143,7 +1134,7 @@ listLocation as = EpaSpan (go noSrcSpan as) go acc (L (EpAnn (EpaSpan s) _ _) _:rest) = go (combine acc s) rest go acc (_:rest) = go acc rest -widenAnchor :: Anchor -> [AddEpAnn] -> Anchor +widenAnchor :: EpaLocation -> [AddEpAnn] -> EpaLocation widenAnchor (EpaSpan (RealSrcSpan s mb)) as = EpaSpan (RealSrcSpan (widenRealSpan s as) (liftA2 combineBufSpans mb (bufSpanFromAnns as))) widenAnchor (EpaSpan us) _ = EpaSpan us @@ -1151,7 +1142,7 @@ widenAnchor a at EpaDelta{} as = case (realSpanFromAnns as) of Strict.Nothing -> a Strict.Just r -> EpaSpan (RealSrcSpan r Strict.Nothing) -widenAnchorS :: Anchor -> SrcSpan -> Anchor +widenAnchorS :: EpaLocation -> SrcSpan -> EpaLocation widenAnchorS (EpaSpan (RealSrcSpan s mbe)) (RealSrcSpan r mbr) = EpaSpan (RealSrcSpan (combineRealSrcSpans s r) (liftA2 combineBufSpans mbe mbr)) widenAnchorS (EpaSpan us) _ = EpaSpan us ===================================== compiler/GHC/Parser/PostProcess.hs ===================================== @@ -478,13 +478,13 @@ add_where an@(AddEpAnn _ (EpaSpan (RealSrcSpan rs _))) (EpAnn a (AnnList anc o c add_where (AddEpAnn _ _) _ _ = panic "add_where" -- EpaDelta should only be used for transformations -valid_anchor :: Anchor -> Bool +valid_anchor :: EpaLocation -> Bool valid_anchor (EpaSpan (RealSrcSpan r _)) = srcSpanStartLine r >= 0 valid_anchor _ = False -- If the decl list for where binds is empty, the anchor ends up -- invalid. In this case, use the parent one -patch_anchor :: RealSrcSpan -> Anchor -> Anchor +patch_anchor :: RealSrcSpan -> EpaLocation -> EpaLocation patch_anchor r EpaDelta{} = EpaSpan (RealSrcSpan r Strict.Nothing) patch_anchor r1 (EpaSpan (RealSrcSpan r0 mb)) = EpaSpan (RealSrcSpan r mb) where @@ -495,9 +495,9 @@ fixValbindsAnn :: EpAnn AnnList -> EpAnn AnnList fixValbindsAnn (EpAnn anchor (AnnList ma o c r t) cs) = (EpAnn (widenAnchor anchor (r ++ map trailingAnnToAddEpAnn t)) (AnnList ma o c r t) cs) --- | The 'Anchor' for a stmtlist is based on either the location or +-- | The anchor for a stmtlist is based on either the location or -- the first semicolon annotion. -stmtsAnchor :: Located (OrdList AddEpAnn,a) -> Maybe Anchor +stmtsAnchor :: Located (OrdList AddEpAnn,a) -> Maybe EpaLocation stmtsAnchor (L (RealSrcSpan l mb) ((ConsOL (AddEpAnn _ (EpaSpan (RealSrcSpan r rb))) _), _)) = Just $ widenAnchorS (EpaSpan (RealSrcSpan l mb)) (RealSrcSpan r rb) stmtsAnchor (L (RealSrcSpan l mb) _) = Just $ EpaSpan (RealSrcSpan l mb) ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -101,9 +101,9 @@ import GHC.Utils.Fingerprint import GHC.Utils.Misc import GHC.Settings.Config (cProjectUnitId) -import Control.DeepSeq +import Control.DeepSeq (NFData(..)) import Data.Data -import Data.List (sortBy ) +import Data.List (sortBy) import Data.Function import Data.Bifunctor import qualified Data.ByteString as BS ===================================== docs/users_guide/exts/or_patterns.rst ===================================== @@ -114,8 +114,8 @@ Or-patterns do not employ backtracking when given guarded right hand sides, i.e. when one alternative of the or-pattern matches, the others are not tried when the guard fails. The following code yields ``"no backtracking"``: :: - case (True, error "backtracking") of - ((True; _); (_; True)) | False -> error "inaccessible" + case error "backtracking" of + (_; True) | False -> error "inaccessible" _ -> error "no backtracking" (The exact syntax and semantics of or-patterns are found ===================================== hadrian/src/Builder.hs ===================================== @@ -237,16 +237,25 @@ instance H.Builder Builder where -- changes (#18001). _bootGhcVersion <- setting GhcVersion pure [] - Ghc {} -> do + Ghc _ st -> do root <- buildRoot unlitPath <- builderPath Unlit distro_mingw <- settingsFileSetting ToolchainSetting_DistroMinGW + libffi_adjustors <- useLibffiForAdjustors + use_system_ffi <- flag UseSystemFfi return $ [ unlitPath ] ++ [ root -/- mingwStamp | windowsHost, distro_mingw == "NO" ] -- proxy for the entire mingw toolchain that -- we have in inplace/mingw initially, and then at -- root -/- mingw. + -- ffi.h needed by the compiler when using libffi_adjustors (#24864) + -- It would be nicer to not duplicate this logic between here + -- and needRtsLibffiTargets and libffiHeaderFiles but this doesn't change + -- very often. + ++ [ root -/- buildDir (rtsContext st) -/- "include" -/- header + | header <- ["ffi.h", "ffitarget.h"] + , libffi_adjustors && not use_system_ffi ] Hsc2Hs stage -> (\p -> [p]) <$> templateHscPath stage Make dir -> return [dir -/- "Makefile"] ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -23,6 +23,12 @@ module GHC.Exts -- ** Legacy interface for arrays of arrays module GHC.Internal.ArrayArray, -- * Primitive operations + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.BCO, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.mkApUpd0#, + {-# DEPRECATED ["The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14", "These symbols should be imported from ghc-internal instead if needed."] #-} + Prim.newBCO#, module GHC.Prim, module GHC.Prim.Ext, -- ** Running 'RealWorld' state thread @@ -114,7 +120,10 @@ import GHC.Internal.Exts import GHC.Internal.ArrayArray import GHC.Prim hiding ( coerce, dataToTagSmall#, dataToTagLarge#, whereFrom#, - isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned# ) + isByteArrayWeaklyPinned#, isMutableByteArrayWeaklyPinned#, + -- Deprecated + BCO, mkApUpd0#, newBCO# ) +import qualified GHC.Prim as Prim ( BCO, mkApUpd0#, newBCO# ) -- Hide dataToTag# ops because they are expected to break for -- GHC-internal reasons in the near future, and shouldn't -- be exposed from base (not even GHC.Exts) ===================================== libraries/deepseq ===================================== @@ -1 +1 @@ -Subproject commit 09aed1bf774f2f05c8b390539ce35adf5cd68c30 +Subproject commit 7ce6e2d3760b23336fd5f9a36f50df6571606947 ===================================== libraries/ghci/GHCi/CreateBCO.hs ===================================== @@ -6,6 +6,10 @@ {-# LANGUAGE UnboxedTuples #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE CPP #-} +{-# OPTIONS_GHC -Wno-warnings-deprecations #-} +-- TODO We want to import GHC.Internal.Base (BCO, mkApUpd0#, newBCO#) instead +-- of from GHC.Exts when we can require of the bootstrap compiler to have +-- ghc-internal. -- -- (c) The University of Glasgow 2002-2006 ===================================== libraries/ghci/ghci.cabal.in ===================================== @@ -86,9 +86,10 @@ library array == 0.5.*, base >= 4.8 && < 4.21, -- ghc-internal == @ProjectVersionForLib at .* - -- TODO: Use GHC.Internal.Desugar from ghc-internal instead of ignoring - -- the deprecation warning of GHC.Desugar when we require ghc-internal - -- of the bootstrap compiler + -- TODO: Use GHC.Internal.Desugar and GHC.Internal.Base from + -- ghc-internal instead of ignoring the deprecation warning in GHCi.TH + -- and GHCi.CreateBCO when we require ghc-internal of the bootstrap + -- compiler ghc-prim >= 0.5.0 && < 0.12, binary == 0.8.*, bytestring >= 0.10 && < 0.13, ===================================== testsuite/tests/module/T21752.stderr ===================================== @@ -0,0 +1,32 @@ +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘newBCO#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘newBCO#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘mkApUpd0#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of ‘mkApUpd0#’ (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of type constructor or class ‘BCO’ + (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + +T21752A.hs:4:5: warning: [GHC-68441] [-Wdeprecations (in -Wextended-warnings)] + In the use of type constructor or class ‘BCO’ + (imported from GHC.Exts): + Deprecated: "The BCO, mkApUpd0#, and newBCO# re-exports from GHC.Exts have been deprecated and will be removed in 9.14 + These symbols should be imported from ghc-internal instead if needed." + ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -175,8 +175,8 @@ data EPState = EPState { uAnchorSpan :: !RealSrcSpan -- ^ in pre-changed AST -- reference frame, from -- Annotation - , uExtraDP :: !(Maybe Anchor) -- ^ Used to anchor a - -- list + , uExtraDP :: !(Maybe EpaLocation) -- ^ Used to anchor a + -- list , pAcceptSpan :: Bool -- ^ When we have processed an -- entry of EpaDelta, accept the -- next `EpaSpan` start as the @@ -214,21 +214,21 @@ class HasTrailing a where setTrailing :: a -> [TrailingAnn] -> a setAnchorEpa :: (HasTrailing an, NoAnn an) - => EpAnn an -> Anchor -> [TrailingAnn] -> EpAnnComments -> EpAnn an + => EpAnn an -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn an setAnchorEpa (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing an ts) cs -setAnchorHsModule :: HsModule GhcPs -> Anchor -> EpAnnComments -> HsModule GhcPs +setAnchorHsModule :: HsModule GhcPs -> EpaLocation -> EpAnnComments -> HsModule GhcPs setAnchorHsModule hsmod anc cs = hsmod { hsmodExt = (hsmodExt hsmod) {hsmodAnn = an'} } where anc' = anc an' = setAnchorEpa (hsmodAnn $ hsmodExt hsmod) anc' [] cs setAnchorAn :: (HasTrailing an, NoAnn an) - => LocatedAn an a -> Anchor -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a + => LocatedAn an a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a setAnchorAn (L (EpAnn _ an _) a) anc ts cs = (L (EpAnn anc (setTrailing an ts) cs) a) -- `debug` ("setAnchorAn: anc=" ++ showAst anc) -setAnchorEpaL :: EpAnn AnnList -> Anchor -> [TrailingAnn] -> EpAnnComments -> EpAnn AnnList +setAnchorEpaL :: EpAnn AnnList -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn AnnList setAnchorEpaL (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing (an {al_anchor = Nothing}) ts) cs -- --------------------------------------------------------------------- @@ -250,14 +250,14 @@ data CanUpdateAnchor = CanUpdateAnchor | NoCanUpdateAnchor deriving (Eq, Show) -data Entry = Entry Anchor [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor +data Entry = Entry EpaLocation [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor | NoEntryVal -- | For flagging whether to capture comments in an EpaDelta or not data CaptureComments = CaptureComments | NoCaptureComments -mkEntry :: Anchor -> [TrailingAnn] -> EpAnnComments -> Entry +mkEntry :: EpaLocation -> [TrailingAnn] -> EpAnnComments -> Entry mkEntry anc ts cs = Entry anc ts cs NoFlushComments CanUpdateAnchor instance (HasTrailing a) => HasEntry (EpAnn a) where @@ -642,7 +642,7 @@ withPpr a = do -- 'ppr'. class (Typeable a) => ExactPrint a where getAnnotationEntry :: a -> Entry - setAnnotationAnchor :: a -> Anchor -> [TrailingAnn] -> EpAnnComments -> a + setAnnotationAnchor :: a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> a exact :: (Monad m, Monoid w) => a -> EP w m a -- --------------------------------------------------------------------- @@ -4277,7 +4277,7 @@ instance ExactPrint (LocatedN RdrName) where locFromAdd :: AddEpAnn -> EpaLocation locFromAdd (AddEpAnn _ loc) = loc -printUnicode :: (Monad m, Monoid w) => Anchor -> RdrName -> EP w m Anchor +printUnicode :: (Monad m, Monoid w) => EpaLocation -> RdrName -> EP w m EpaLocation printUnicode anc n = do let str = case (showPprUnsafe n) of -- TODO: unicode support? @@ -4977,10 +4977,10 @@ setPosP l = do debugM $ "setPosP:" ++ show l modify (\s -> s {epPos = l}) -getExtraDP :: (Monad m, Monoid w) => EP w m (Maybe Anchor) +getExtraDP :: (Monad m, Monoid w) => EP w m (Maybe EpaLocation) getExtraDP = gets uExtraDP -setExtraDP :: (Monad m, Monoid w) => Maybe Anchor -> EP w m () +setExtraDP :: (Monad m, Monoid w) => Maybe EpaLocation -> EP w m () setExtraDP md = do debugM $ "setExtraDP:" ++ show md modify (\s -> s {uExtraDP = md}) ===================================== utils/check-exact/Utils.hs ===================================== @@ -255,7 +255,7 @@ tokComment t@(L lt c) = (GHC.EpaComment (EpaDocComment dc) pt) -> hsDocStringComments (noCommentsToEpaLocation lt) pt dc _ -> [mkComment (normaliseCommentText (ghcCommentText t)) lt (ac_prior_tok c)] -hsDocStringComments :: Anchor -> RealSrcSpan -> GHC.HsDocString -> [Comment] +hsDocStringComments :: EpaLocation -> RealSrcSpan -> GHC.HsDocString -> [Comment] hsDocStringComments _ pt (MultiLineDocString dec (x :| xs)) = let decStr = printDecorator dec @@ -342,7 +342,7 @@ isKWComment c = isJust (commentOrigin c) noKWComments :: [Comment] -> [Comment] noKWComments = filter (\c -> not (isKWComment c)) -sortAnchorLocated :: [GenLocated Anchor a] -> [GenLocated Anchor a] +sortAnchorLocated :: [GenLocated EpaLocation a] -> [GenLocated EpaLocation a] sortAnchorLocated = sortBy (compare `on` (anchor . getLoc)) -- | Calculates the distance from the start of a string to the end of @@ -401,12 +401,6 @@ setTrailingAnnLoc (AddDarrowUAnn _) ss = (AddDarrowUAnn ss) addEpAnnLoc :: AddEpAnn -> EpaLocation addEpAnnLoc (AddEpAnn _ l) = l --- --------------------------------------------------------------------- - --- TODO: get rid of this identity function -anchorToEpaLocation :: Anchor -> EpaLocation -anchorToEpaLocation a = a - -- --------------------------------------------------------------------- -- Horrible hack for dealing with some things still having a SrcSpan, -- not an Anchor. @@ -432,7 +426,7 @@ To be absolutely sure, we make the delta versions use -ve values. -} -hackSrcSpanToAnchor :: SrcSpan -> Anchor +hackSrcSpanToAnchor :: SrcSpan -> EpaLocation hackSrcSpanToAnchor (UnhelpfulSpan s) = error $ "hackSrcSpanToAnchor : UnhelpfulSpan:" ++ show s hackSrcSpanToAnchor ss@(RealSrcSpan r mb) = case mb of @@ -443,7 +437,7 @@ hackSrcSpanToAnchor ss@(RealSrcSpan r mb) else EpaSpan (RealSrcSpan r mb) _ -> EpaSpan (RealSrcSpan r mb) -hackAnchorToSrcSpan :: Anchor -> SrcSpan +hackAnchorToSrcSpan :: EpaLocation -> SrcSpan hackAnchorToSrcSpan (EpaSpan s) = s hackAnchorToSrcSpan _ = error $ "hackAnchorToSrcSpan" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d57344fc13bdbb23d73636a0a4635b060743d77c...1930fd32831d5a3fa87c40e3af34fdfab19aeec7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d57344fc13bdbb23d73636a0a4635b060743d77c...1930fd32831d5a3fa87c40e3af34fdfab19aeec7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 02:20:57 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 11 Sep 2024 22:20:57 -0400 Subject: [Git][ghc/ghc][master] Bump submodule deepseq to 1.5.1.0 Message-ID: <66e25009a383b_1efe5ef560e08597b@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - 2 changed files: - compiler/GHC/Unit/Types.hs - libraries/deepseq Changes: ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -101,9 +101,9 @@ import GHC.Utils.Fingerprint import GHC.Utils.Misc import GHC.Settings.Config (cProjectUnitId) -import Control.DeepSeq +import Control.DeepSeq (NFData(..)) import Data.Data -import Data.List (sortBy ) +import Data.List (sortBy) import Data.Function import Data.Bifunctor import qualified Data.ByteString as BS ===================================== libraries/deepseq ===================================== @@ -1 +1 @@ -Subproject commit 09aed1bf774f2f05c8b390539ce35adf5cd68c30 +Subproject commit 7ce6e2d3760b23336fd5f9a36f50df6571606947 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8e462f4d4bdf2a6c34c249e7be8084565600d300 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8e462f4d4bdf2a6c34c249e7be8084565600d300 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 02:21:36 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 11 Sep 2024 22:21:36 -0400 Subject: [Git][ghc/ghc][master] User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Message-ID: <66e250302e25c_1efe5ef9e9f890575@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1 changed file: - docs/users_guide/exts/or_patterns.rst Changes: ===================================== docs/users_guide/exts/or_patterns.rst ===================================== @@ -114,8 +114,8 @@ Or-patterns do not employ backtracking when given guarded right hand sides, i.e. when one alternative of the or-pattern matches, the others are not tried when the guard fails. The following code yields ``"no backtracking"``: :: - case (True, error "backtracking") of - ((True; _); (_; True)) | False -> error "inaccessible" + case error "backtracking" of + (_; True) | False -> error "inaccessible" _ -> error "no backtracking" (The exact syntax and semantics of or-patterns are found View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/aa4500aecca763be5ce27c4784473f00e2ce5aa3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/aa4500aecca763be5ce27c4784473f00e2ce5aa3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 10:39:40 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 12 Sep 2024 06:39:40 -0400 Subject: [Git][ghc/ghc][wip/supersven/ghc-master-riscv-ncg] 12 commits: Bump submodule deepseq to 1.5.1.0 Message-ID: <66e2c4ec83a52_33da7282d86844c@gitlab.mail> Sven Tennie pushed to branch wip/supersven/ghc-master-riscv-ncg at Glasgow Haskell Compiler / GHC Commits: 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - 10 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fddd2272d7ac4edc3083ec284a5652624475932c...8e6d58cfb0e701fcb0401456bd46c530655b6e8e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fddd2272d7ac4edc3083ec284a5652624475932c...8e6d58cfb0e701fcb0401456bd46c530655b6e8e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 14:22:59 2024 From: gitlab at gitlab.haskell.org (Sjoerd Visscher (@trac-sjoerd_visscher)) Date: Thu, 12 Sep 2024 10:22:59 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/finder-boot] finder: Add `IsBootInterface` to finder cache keys Message-ID: <66e2f943dc2c4_3272af2bfebc54297@gitlab.mail> Sjoerd Visscher pushed to branch wip/torsten.schmits/finder-boot at Glasgow Haskell Compiler / GHC Commits: ba3dc458 by Torsten Schmits at 2024-09-12T16:22:45+02:00 finder: Add `IsBootInterface` to finder cache keys - - - - - 12 changed files: - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Finder/Types.hs - compiler/GHC/Unit/Module/Env.hs - compiler/GHC/Unit/Types.hs - + testsuite/tests/driver/boot-target/A.hs - + testsuite/tests/driver/boot-target/A.hs-boot - + testsuite/tests/driver/boot-target/B.hs - + testsuite/tests/driver/boot-target/Makefile - + testsuite/tests/driver/boot-target/all.T Changes: ===================================== compiler/GHC/Driver/Backpack.hs ===================================== @@ -781,7 +781,7 @@ summariseRequirement pn mod_name = do let loc = srcLocSpan (mkSrcLoc (mkFastString (bkp_filename env)) 1 1) let fc = hsc_FC hsc_env - mod <- liftIO $ addHomeModuleToFinder fc home_unit mod_name location + mod <- liftIO $ addHomeModuleToFinder fc home_unit (notBoot mod_name) location extra_sig_imports <- liftIO $ findExtraSigImports hsc_env HsigFile mod_name @@ -893,7 +893,7 @@ hsModuleToModSummary home_keys pn hsc_src modname this_mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit modname location + addHomeModuleToFinder fc home_unit (GWIB modname (hscSourceToIsBoot hsc_src)) location let ms = ModSummary { ms_mod = this_mod, ms_hsc_src = hsc_src, ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2044,25 +2044,43 @@ summariseFile hsc_env' home_unit old_summaries src_fn mb_phase maybe_buf <- getPreprocessedImports hsc_env src_fn mb_phase maybe_buf let fopts = initFinderOpts (hsc_dflags hsc_env) - - -- Make a ModLocation for this file - let location = mkHomeModLocation fopts pi_mod_name (unsafeEncodeUtf src_fn) + src_path = unsafeEncodeUtf src_fn + + is_boot = case takeExtension src_fn of + ".hs-boot" -> IsBoot + ".lhs-boot" -> IsBoot + _ -> NotBoot + + (path_without_boot, hsc_src) + | isHaskellSigFilename src_fn = (src_path, HsigFile) + | IsBoot <- is_boot = (removeBootSuffix src_path, HsBootFile) + | otherwise = (src_path, HsSrcFile) + + -- Make a ModLocation for the Finder, who only has one entry for + -- each @ModuleName@, and therefore needs to use the locations for + -- the non-boot files. + location_without_boot = + mkHomeModLocation fopts pi_mod_name path_without_boot + + -- Make a ModLocation for this file, adding the @-boot@ suffix to + -- all paths if the original was a boot file. + location + | IsBoot <- is_boot + = addBootSuffixLocn location_without_boot + | otherwise + = location_without_boot -- Tell the Finder cache where it is, so that subsequent calls -- to findModule will find it, even if it's not on any search path mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit pi_mod_name location + addHomeModuleToFinder fc home_unit (GWIB pi_mod_name is_boot) location liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = NotBoot - , nms_hsc_src = - if isHaskellSigFilename src_fn - then HsigFile - else HsSrcFile + , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod , nms_preimps = preimps @@ -2090,9 +2108,10 @@ checkSummaryHash -- Also, only add to finder cache for non-boot modules as the finder cache -- makes sure to add a boot suffix for boot files. _ <- do - let fc = hsc_FC hsc_env + let fc = hsc_FC hsc_env + gwib = GWIB (ms_mod old_summary) (isBootSummary old_summary) case ms_hsc_src old_summary of - HsSrcFile -> addModuleToFinder fc (ms_mod old_summary) location + HsSrcFile -> addModuleToFinder fc gwib location _ -> return () hi_timestamp <- modificationTimeIfExists (ml_hi_file location) @@ -2230,7 +2249,6 @@ summariseModule hsc_env' home_unit old_summary_map is_boot (L _ wanted_mod) mb_p liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = is_boot , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod @@ -2243,7 +2261,6 @@ data MakeNewModSummary = MakeNewModSummary { nms_src_fn :: FilePath , nms_src_hash :: Fingerprint - , nms_is_boot :: IsBootInterface , nms_hsc_src :: HscSource , nms_location :: ModLocation , nms_mod :: Module ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -734,7 +734,7 @@ runHscPhase pipe_env hsc_env0 input_fn src_flavour = do mod <- do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit mod_name location + addHomeModuleToFinder fc home_unit (GWIB mod_name (hscSourceToIsBoot src_flavour)) location -- Make the ModSummary to hand to hscMain let ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -89,23 +89,23 @@ type BaseName = OsPath -- Basename of file initFinderCache :: IO FinderCache initFinderCache = do - mod_cache <- newIORef emptyInstalledModuleEnv + mod_cache <- newIORef emptyInstalledModuleWithIsBootEnv file_cache <- newIORef M.empty let flushFinderCaches :: UnitEnv -> IO () flushFinderCaches ue = do - atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleEnv is_ext fm, ()) + atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleWithIsBootEnv is_ext fm, ()) atomicModifyIORef' file_cache $ \_ -> (M.empty, ()) where - is_ext mod _ = not (isUnitEnvInstalledModule ue mod) + is_ext mod _ = not (isUnitEnvInstalledModule ue (gwib_mod mod)) - addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () addToFinderCache key val = - atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleEnv c key val, ()) + atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleWithIsBootEnv c key val, ()) - lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) lookupFinderCache key = do c <- readIORef mod_cache - return $! lookupInstalledModuleEnv c key + return $! lookupInstalledModuleWithIsBootEnv c key lookupFileCache :: FilePath -> IO Fingerprint lookupFileCache key = do @@ -255,7 +255,7 @@ orIfNotFound this or_this = do homeSearchCache :: FinderCache -> UnitId -> ModuleName -> IO InstalledFindResult -> IO InstalledFindResult homeSearchCache fc home_unit mod_name do_this = do let mod = mkModule home_unit mod_name - modLocationCache fc mod do_this + modLocationCache fc (notBoot mod) do_this findExposedPackageModule :: FinderCache -> FinderOpts -> UnitState -> ModuleName -> PkgQual -> IO FindResult findExposedPackageModule fc fopts units mod_name mb_pkg = @@ -312,7 +312,7 @@ findLookupResult fc fopts r = case r of , fr_unusables = [] , fr_suggestions = suggest' }) -modLocationCache :: FinderCache -> InstalledModule -> IO InstalledFindResult -> IO InstalledFindResult +modLocationCache :: FinderCache -> InstalledModuleWithIsBoot -> IO InstalledFindResult -> IO InstalledFindResult modLocationCache fc mod do_this = do m <- lookupFinderCache fc mod case m of @@ -322,17 +322,17 @@ modLocationCache fc mod do_this = do addToFinderCache fc mod result return result -addModuleToFinder :: FinderCache -> Module -> ModLocation -> IO () +addModuleToFinder :: FinderCache -> ModuleWithIsBoot -> ModLocation -> IO () addModuleToFinder fc mod loc = do - let imod = toUnitId <$> mod - addToFinderCache fc imod (InstalledFound loc imod) + let imod = fmap toUnitId <$> mod + addToFinderCache fc imod (InstalledFound loc (gwib_mod imod)) -- This returns a module because it's more convenient for users -addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleName -> ModLocation -> IO Module +addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleNameWithIsBoot -> ModLocation -> IO Module addHomeModuleToFinder fc home_unit mod_name loc = do - let mod = mkHomeInstalledModule home_unit mod_name - addToFinderCache fc mod (InstalledFound loc mod) - return (mkHomeModule home_unit mod_name) + let mod = mkHomeInstalledModule home_unit <$> mod_name + addToFinderCache fc mod (InstalledFound loc (gwib_mod mod)) + return (mkHomeModule home_unit (gwib_mod mod_name)) -- ----------------------------------------------------------------------------- -- The internal workers @@ -466,7 +466,7 @@ findPackageModule_ :: FinderCache -> FinderOpts -> InstalledModule -> UnitInfo - findPackageModule_ fc fopts mod pkg_conf = do massertPpr (moduleUnit mod == unitId pkg_conf) (ppr (moduleUnit mod) <+> ppr (unitId pkg_conf)) - modLocationCache fc mod $ + modLocationCache fc (notBoot mod) $ -- special case for GHC.Prim; we won't find it in the filesystem. if mod `installedModuleEq` gHC_PRIM ===================================== compiler/GHC/Unit/Finder/Types.hs ===================================== @@ -30,9 +30,9 @@ data FinderCache = FinderCache { flushFinderCaches :: UnitEnv -> IO () -- ^ remove all the home modules from the cache; package modules are -- assumed to not move around during a session; also flush the file hash -- cache. - , addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + , addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () -- ^ Add a found location to the cache for the module. - , lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + , lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) -- ^ Look for a location in the cache. , lookupFileCache :: FilePath -> IO Fingerprint -- ^ Look for the hash of a file in the cache. This should add it to the ===================================== compiler/GHC/Unit/Module/Env.hs ===================================== @@ -33,6 +33,17 @@ module GHC.Unit.Module.Env , mergeInstalledModuleEnv , plusInstalledModuleEnv , installedModuleEnvElts + + -- * InstalledModuleWithIsBootEnv + , InstalledModuleWithIsBootEnv + , emptyInstalledModuleWithIsBootEnv + , lookupInstalledModuleWithIsBootEnv + , extendInstalledModuleWithIsBootEnv + , filterInstalledModuleWithIsBootEnv + , delInstalledModuleWithIsBootEnv + , mergeInstalledModuleWithIsBootEnv + , plusInstalledModuleWithIsBootEnv + , installedModuleWithIsBootEnvElts ) where @@ -283,3 +294,56 @@ plusInstalledModuleEnv :: (elt -> elt -> elt) plusInstalledModuleEnv f (InstalledModuleEnv xm) (InstalledModuleEnv ym) = InstalledModuleEnv $ Map.unionWith f xm ym + + +-------------------------------------------------------------------- +-- InstalledModuleWithIsBootEnv +-------------------------------------------------------------------- + +-- | A map keyed off of 'InstalledModuleWithIsBoot' +newtype InstalledModuleWithIsBootEnv elt = InstalledModuleWithIsBootEnv (Map InstalledModuleWithIsBoot elt) + +instance Outputable elt => Outputable (InstalledModuleWithIsBootEnv elt) where + ppr (InstalledModuleWithIsBootEnv env) = ppr env + + +emptyInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a +emptyInstalledModuleWithIsBootEnv = InstalledModuleWithIsBootEnv Map.empty + +lookupInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> Maybe a +lookupInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = Map.lookup m e + +extendInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> a -> InstalledModuleWithIsBootEnv a +extendInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m x = InstalledModuleWithIsBootEnv (Map.insert m x e) + +filterInstalledModuleWithIsBootEnv :: (InstalledModuleWithIsBoot -> a -> Bool) -> InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBootEnv a +filterInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv e) = + InstalledModuleWithIsBootEnv (Map.filterWithKey f e) + +delInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> InstalledModuleWithIsBootEnv a +delInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = InstalledModuleWithIsBootEnv (Map.delete m e) + +installedModuleWithIsBootEnvElts :: InstalledModuleWithIsBootEnv a -> [(InstalledModuleWithIsBoot, a)] +installedModuleWithIsBootEnvElts (InstalledModuleWithIsBootEnv e) = Map.assocs e + +mergeInstalledModuleWithIsBootEnv + :: (elta -> eltb -> Maybe eltc) + -> (InstalledModuleWithIsBootEnv elta -> InstalledModuleWithIsBootEnv eltc) -- map X + -> (InstalledModuleWithIsBootEnv eltb -> InstalledModuleWithIsBootEnv eltc) -- map Y + -> InstalledModuleWithIsBootEnv elta + -> InstalledModuleWithIsBootEnv eltb + -> InstalledModuleWithIsBootEnv eltc +mergeInstalledModuleWithIsBootEnv f g h (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) + = InstalledModuleWithIsBootEnv $ Map.mergeWithKey + (\_ x y -> (x `f` y)) + (coerce g) + (coerce h) + xm ym + +plusInstalledModuleWithIsBootEnv :: (elt -> elt -> elt) + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt +plusInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) = + InstalledModuleWithIsBootEnv $ Map.unionWith f xm ym + ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -86,6 +86,8 @@ module GHC.Unit.Types , GenWithIsBoot (..) , ModuleNameWithIsBoot , ModuleWithIsBoot + , InstalledModuleWithIsBoot + , notBoot ) where @@ -723,6 +725,8 @@ type ModuleNameWithIsBoot = GenWithIsBoot ModuleName type ModuleWithIsBoot = GenWithIsBoot Module +type InstalledModuleWithIsBoot = GenWithIsBoot InstalledModule + instance Binary a => Binary (GenWithIsBoot a) where put_ bh (GWIB { gwib_mod, gwib_isBoot }) = do put_ bh gwib_mod @@ -736,3 +740,6 @@ instance Outputable a => Outputable (GenWithIsBoot a) where ppr (GWIB { gwib_mod, gwib_isBoot }) = hsep $ ppr gwib_mod : case gwib_isBoot of IsBoot -> [ text "{-# SOURCE #-}" ] NotBoot -> [] + +notBoot :: mod -> GenWithIsBoot mod +notBoot gwib_mod = GWIB {gwib_mod, gwib_isBoot = NotBoot} ===================================== testsuite/tests/driver/boot-target/A.hs ===================================== @@ -0,0 +1,5 @@ +module A where + +import B + +data A = A B ===================================== testsuite/tests/driver/boot-target/A.hs-boot ===================================== @@ -0,0 +1,3 @@ +module A where + +data A ===================================== testsuite/tests/driver/boot-target/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} A + +data B = B A ===================================== testsuite/tests/driver/boot-target/Makefile ===================================== @@ -0,0 +1,8 @@ +boot1: + $(TEST_HC) -c A.hs-boot B.hs + +boot2: + $(TEST_HC) A.hs-boot A.hs B.hs -v0 + +boot3: + $(TEST_HC) A.hs-boot B.hs -v0 \ No newline at end of file ===================================== testsuite/tests/driver/boot-target/all.T ===================================== @@ -0,0 +1,10 @@ +def test_boot(name): + return test(name, + [extra_files(['A.hs', 'A.hs-boot', 'B.hs']), + ], + makefile_test, + []) + +test_boot('boot1') +test_boot('boot2') +test_boot('boot3') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ba3dc458e236795e5f22d9489af2861cef403edf -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ba3dc458e236795e5f22d9489af2861cef403edf You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 14:27:35 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 12 Sep 2024 10:27:35 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 13 commits: Bump submodule deepseq to 1.5.1.0 Message-ID: <66e2fa572b18_3272af286d885616e@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - 02733138 by Alan Zimmerman at 2024-09-12T10:27:30-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 10 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1930fd32831d5a3fa87c40e3af34fdfab19aeec7...02733138052c08b8e5149f4ca8d21712d3d23f11 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1930fd32831d5a3fa87c40e3af34fdfab19aeec7...02733138052c08b8e5149f4ca8d21712d3d23f11 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 15:17:38 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 12 Sep 2024 11:17:38 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 18 commits: RTS linker: add support for hidden symbols (#25191) Message-ID: <66e3061294627_3272af6c96e8678e8@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 1558786d by Matthew Pickering at 2024-09-11T16:17:12+01:00 Run on test-abi label - - - - - 33251d4f by Rodrigo Mesquita at 2024-09-11T16:17:12+01:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - 5f339d8c by Rodrigo Mesquita at 2024-09-11T16:17:12+01:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - cbc5e74e by Rodrigo Mesquita at 2024-09-11T16:17:13+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - db464e7c by Rodrigo Mesquita at 2024-09-11T16:17:13+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - 7232863b by Rodrigo Mesquita at 2024-09-11T16:17:13+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 2d6620c7 by Rodrigo Mesquita at 2024-09-11T16:17:13+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 2cfccf9f by Rodrigo Mesquita at 2024-09-12T16:17:28+01:00 determinism: Rename CmmGroups in generateCgIPEStub Make sure to also deterministically rename the IPE Cmm (as per Note [Renaming uniques deterministically]) to guarantee deterministic objects when IPE information is requested. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/79688990da1f23cc0602fa3a37677b255a3c897f...2cfccf9f76faef7389f7d3318db7b762e94e2ebe -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/79688990da1f23cc0602fa3a37677b255a3c897f...2cfccf9f76faef7389f7d3318db7b762e94e2ebe You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 15:29:31 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 12 Sep 2024 11:29:31 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 6 commits: determinism: Sampling uniques in the CG Message-ID: <66e308dbaa0e8_3272af70176470740@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 0c1d7fc9 by Rodrigo Mesquita at 2024-09-12T16:29:18+01:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 79583f20 by Rodrigo Mesquita at 2024-09-12T16:29:18+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - ca7f0e13 by Rodrigo Mesquita at 2024-09-12T16:29:18+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - 8531a6e8 by Rodrigo Mesquita at 2024-09-12T16:29:18+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 5540655e by Rodrigo Mesquita at 2024-09-12T16:29:18+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - d7091d6d by Rodrigo Mesquita at 2024-09-12T16:29:18+01:00 determinism: Rename CmmGroups in generateCgIPEStub Make sure to also deterministically rename the IPE Cmm (as per Note [Renaming uniques deterministically]) to guarantee deterministic objects when IPE information is requested. - - - - - 30 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Reg/Graph.hs - compiler/GHC/CmmToAsm/Reg/Graph/Spill.hs - compiler/GHC/CmmToAsm/Reg/Linear.hs - compiler/GHC/CmmToAsm/Reg/Linear/Base.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2cfccf9f76faef7389f7d3318db7b762e94e2ebe...d7091d6dea51df10ab5de2d0b815192c5bb7577e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2cfccf9f76faef7389f7d3318db7b762e94e2ebe...d7091d6dea51df10ab5de2d0b815192c5bb7577e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 15:50:59 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Thu, 12 Sep 2024 11:50:59 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] wrk: done, just need a better name and good commit Message-ID: <66e30de3e51c7_eb108aad426041@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 501963a9 by doyougnu at 2024-09-12T11:50:24-04:00 wrk: done, just need a better name and good commit - - - - - 8 changed files: - rts/Linker.c - testsuite/tests/ghci/linking/Makefile - testsuite/tests/rts/flags/T25240/FFI.hs → testsuite/tests/ghci/linking/T25240/FFI.hs - + testsuite/tests/ghci/linking/T25240/Makefile - + testsuite/tests/ghci/linking/T25240/T25240.hs - testsuite/tests/rts/flags/T25240/T25240.hs → testsuite/tests/ghci/linking/T25240/T25240f.hs - testsuite/tests/rts/flags/T25240/TH.hs → testsuite/tests/ghci/linking/T25240/TH.hs - testsuite/tests/rts/flags/T25240/all.T → testsuite/tests/ghci/linking/T25240/all.T Changes: ===================================== rts/Linker.c ===================================== @@ -946,16 +946,17 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) // lookupDependentSymbol directly. SymbolAddr* r = lookupDependentSymbol(lbl, NULL, NULL); if (!r) { - errorBelch("^^ Could not load '%s', dependency unresolved. " - "See top entry above.\n", lbl); - IF_DEBUG(linker, printLoadedObjects()); - fflush(stderr); + if (!RtsFlags.MiscFlags.linkUnknownSymbols) { + errorBelch("^^ Could not load '%s', dependency unresolved. " + "See top entry above.\n", + lbl); + IF_DEBUG(linker, printLoadedObjects()); + fflush(stderr); + } // if --link-unknown-symbols is passed into the RTS we allow the linker // to optimistically continue - if (RtsFlags.MiscFlags.linkUnknownSymbols){ - r = (void*) 0xDEADBEEF; - } + r = (void*) 0xDEADBEEF; } ===================================== testsuite/tests/ghci/linking/Makefile ===================================== @@ -145,3 +145,7 @@ T15729: big-obj: '$(TEST_CC)' -c -Wa,-mbig-obj big-obj-c.c -o big-obj-c.o echo "main" | '$(TEST_HC)' $(TEST_HC_OPTS_INTERACTIVE) big-obj-c.o big-obj.hs + +.PHONY: T25240 +T25240: + echo "main" | "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) FFI.hs TH.hs +RTS --link-unknown-objects -RTS ===================================== testsuite/tests/rts/flags/T25240/FFI.hs → testsuite/tests/ghci/linking/T25240/FFI.hs ===================================== @@ -3,6 +3,8 @@ {-# LANGUAGE TemplateHaskell #-} {-# LANGUAGE UnliftedFFITypes #-} +{-# OPTIONS_GHC -fno-warn-missing-methods #-} + module FFI ( D(..), D2(..), tabulate ) where ===================================== testsuite/tests/ghci/linking/T25240/Makefile ===================================== @@ -0,0 +1,11 @@ +TOP=../../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +.PHONY: T25240 +T25240: + "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) FFI.hs TH.hs T25240.hs +RTS --link-unknown-symbols -RTS + +.PHONY: T25240f +T25240f: + "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) FFI.hs TH.hs T25240f.hs ===================================== testsuite/tests/ghci/linking/T25240/T25240.hs ===================================== @@ -0,0 +1,33 @@ + +{- +Here we have two modules, FFI and TH. + +In the FFI module, we define a data type D and a function to be used in Template +Haskell, tabulate. Completely unrelatedly, we also include a Num D instance +which makes use of a foreign import prim declaration, foreign import prim +"prim_func". In the TH module, we use tabulate and get a linking error: + +FFI.o: unknown symbol `prim_func' + + +This suggests we are trying to run some code at splice time involving the Num D +instance, even though Num D is not involved in any way. In particular, if we +change the use of prim_func to something like error "blah", then the error never +gets thrown, which shows we don't run that code at splice time. Note that it is +important to use a typeclass to trigger this bug. If we instead use the FFI +import in a top-level binding, there are no problems. It doesn't matter that the +class is Num or some other (possibly user-defined) class. + +The essential bug is that we have a symbol at link time which comes from dead +code and yet we try to resolve it anyway. The fix is to pass the flag +--link-unknown-symbols to the RTS which will assign the symbol a magic number +0xDEADBEEF and allow the linker to continue. See MR!13012 for the implementation +of that flag +-} + +module DeadBeef where + +import TH + +main :: IO () +main = print th_bug ===================================== testsuite/tests/rts/flags/T25240/T25240.hs → testsuite/tests/ghci/linking/T25240/T25240f.hs ===================================== @@ -1,4 +1,7 @@ +-- duplicate of T25240 to test the failure case +module DeadBeef where + import TH main :: IO () ===================================== testsuite/tests/rts/flags/T25240/TH.hs → testsuite/tests/ghci/linking/T25240/TH.hs ===================================== ===================================== testsuite/tests/rts/flags/T25240/all.T → testsuite/tests/ghci/linking/T25240/all.T ===================================== @@ -1,2 +1,5 @@ test('T25240', [req_rts_linker, extra_files(['FFI.hs', 'TH.hs'])], - multimod_compile, ['T25240', '+RTS --link-unknown-symbols -RTS']) + makefile_test, ['T25240']) + +test('T25240f', [req_rts_linker, expect_fail, extra_files(['FFI.hs', 'TH.hs'])], + makefile_test, ['T25240f']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/501963a972af8f4266ea15da89946e8c943ccef0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/501963a972af8f4266ea15da89946e8c943ccef0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 17:01:31 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Thu, 12 Sep 2024 13:01:31 -0400 Subject: [Git][ghc/ghc][wip/az/epa-exactprint-sync] 3 commits: Bump submodule deepseq to 1.5.1.0 Message-ID: <66e31e6ba9afe_eb10535e1c445c4@gitlab.mail> Alan Zimmerman pushed to branch wip/az/epa-exactprint-sync at Glasgow Haskell Compiler / GHC Commits: 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 3b2e27f3 by Alan Zimmerman at 2024-09-12T18:00:05+01:00 EPA: Sync ghc-exactprint to GHC - - - - - 9 changed files: - compiler/GHC/Unit/Types.hs - docs/users_guide/exts/or_patterns.rst - libraries/deepseq - utils/check-exact/ExactPrint.hs - utils/check-exact/Main.hs - utils/check-exact/Parsers.hs - utils/check-exact/Transform.hs - utils/check-exact/Types.hs - utils/check-exact/Utils.hs Changes: ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -101,9 +101,9 @@ import GHC.Utils.Fingerprint import GHC.Utils.Misc import GHC.Settings.Config (cProjectUnitId) -import Control.DeepSeq +import Control.DeepSeq (NFData(..)) import Data.Data -import Data.List (sortBy ) +import Data.List (sortBy) import Data.Function import Data.Bifunctor import qualified Data.ByteString as BS ===================================== docs/users_guide/exts/or_patterns.rst ===================================== @@ -114,8 +114,8 @@ Or-patterns do not employ backtracking when given guarded right hand sides, i.e. when one alternative of the or-pattern matches, the others are not tried when the guard fails. The following code yields ``"no backtracking"``: :: - case (True, error "backtracking") of - ((True; _); (_; True)) | False -> error "inaccessible" + case error "backtracking" of + (_; True) | False -> error "inaccessible" _ -> error "no backtracking" (The exact syntax and semantics of or-patterns are found ===================================== libraries/deepseq ===================================== @@ -1 +1 @@ -Subproject commit 09aed1bf774f2f05c8b390539ce35adf5cd68c30 +Subproject commit 7ce6e2d3760b23336fd5f9a36f50df6571606947 ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -1,3 +1,4 @@ +{-# LANGUAGE BangPatterns #-} {-# LANGUAGE DeriveDataTypeable #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE FlexibleInstances #-} @@ -25,7 +26,7 @@ module ExactPrint , makeDeltaAst -- * Configuration - , EPOptions(epRigidity, epAstPrint, epTokenPrint, epWhitespacePrint, epUpdateAnchors) + , EPOptions(epTokenPrint, epWhitespacePrint) , stringOptions , epOptions , deltaOptions @@ -43,10 +44,11 @@ import GHC.Types.ForeignCall import GHC.Types.Name.Reader import GHC.Types.PkgQual import GHC.Types.SourceText +import GHC.Types.SrcLoc import GHC.Types.Var -import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Unit.Module.Warnings import GHC.Utils.Misc +import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Utils.Panic import Language.Haskell.Syntax.Basic (FieldLabelString(..)) @@ -77,8 +79,7 @@ import Types exactPrint :: ExactPrint ast => ast -> String exactPrint ast = snd $ runIdentity (runEP stringOptions (markAnnotated ast)) --- | The additional option to specify the rigidity and printing --- configuration. +-- | The additional option to specify the printing configuration. exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) => EPOptions m b -> ast @@ -86,9 +87,8 @@ exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) exactPrintWithOptions r ast = runEP r (markAnnotated ast) --- | Transform concrete annotations into relative annotations which --- are more useful when transforming an AST. This corresponds to the --- earlier 'relativiseApiAnns'. +-- | Transform concrete annotations into relative annotations. +-- This should be unnecessary from GHC 9.10 makeDeltaAst :: ExactPrint ast => ast -> ast makeDeltaAst ast = fst $ runIdentity (runEP deltaOptions (markAnnotated ast)) @@ -115,6 +115,7 @@ defaultEPState = EPState , dPriorEndPosition = (1,1) , uAnchorSpan = badRealSrcSpan , uExtraDP = Nothing + , uExtraDPReturn = Nothing , pAcceptSpan = False , epComments = [] , epCommentsApplied = [] @@ -128,39 +129,27 @@ defaultEPState = EPState -- | The R part of RWS. The environment. Updated via 'local' as we -- enter a new AST element, having a different anchor point. data EPOptions m a = EPOptions - { - epAstPrint :: forall ast . Data ast => GHC.Located ast -> a -> m a - , epTokenPrint :: String -> m a + { epTokenPrint :: String -> m a , epWhitespacePrint :: String -> m a - , epRigidity :: Rigidity - , epUpdateAnchors :: Bool } -- | Helper to create a 'EPOptions' -epOptions :: - (forall ast . Data ast => GHC.Located ast -> a -> m a) - -> (String -> m a) - -> (String -> m a) - -> Rigidity - -> Bool - -> EPOptions m a -epOptions astPrint tokenPrint wsPrint rigidity delta = EPOptions - { - epAstPrint = astPrint - , epWhitespacePrint = wsPrint +epOptions :: (String -> m a) + -> (String -> m a) + -> EPOptions m a +epOptions tokenPrint wsPrint = EPOptions + { epWhitespacePrint = wsPrint , epTokenPrint = tokenPrint - , epRigidity = rigidity - , epUpdateAnchors = delta } -- | Options which can be used to print as a normal String. stringOptions :: EPOptions Identity String -stringOptions = epOptions (\_ b -> return b) return return NormalLayout False +stringOptions = epOptions return return -- | Options which can be used to simply update the AST to be in delta -- form, without generating output deltaOptions :: EPOptions Identity () -deltaOptions = epOptions (\_ _ -> return ()) (\_ -> return ()) (\_ -> return ()) NormalLayout True +deltaOptions = epOptions (\_ -> return ()) (\_ -> return ()) data EPWriter a = EPWriter { output :: !a } @@ -177,6 +166,8 @@ data EPState = EPState -- Annotation , uExtraDP :: !(Maybe EpaLocation) -- ^ Used to anchor a -- list + , uExtraDPReturn :: !(Maybe DeltaPos) + -- ^ Used to return Delta version of uExtraDP , pAcceptSpan :: Bool -- ^ When we have processed an -- entry of EpaDelta, accept the -- next `EpaSpan` start as the @@ -213,7 +204,7 @@ class HasTrailing a where trailing :: a -> [TrailingAnn] setTrailing :: a -> [TrailingAnn] -> a -setAnchorEpa :: (HasTrailing an, NoAnn an) +setAnchorEpa :: (HasTrailing an) => EpAnn an -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn an setAnchorEpa (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing an ts) cs @@ -223,7 +214,7 @@ setAnchorHsModule hsmod anc cs = hsmod { hsmodExt = (hsmodExt hsmod) {hsmodAnn = anc' = anc an' = setAnchorEpa (hsmodAnn $ hsmodExt hsmod) anc' [] cs -setAnchorAn :: (HasTrailing an, NoAnn an) +setAnchorAn :: (HasTrailing an) => LocatedAn an a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a setAnchorAn (L (EpAnn _ an _) a) anc ts cs = (L (EpAnn anc (setTrailing an ts) cs) a) -- `debug` ("setAnchorAn: anc=" ++ showAst anc) @@ -248,7 +239,7 @@ data FlushComments = FlushComments data CanUpdateAnchor = CanUpdateAnchor | CanUpdateAnchorOnly | NoCanUpdateAnchor - deriving (Eq, Show) + deriving (Eq, Show, Data) data Entry = Entry EpaLocation [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor | NoEntryVal @@ -402,7 +393,7 @@ enterAnn NoEntryVal a = do r <- exact a debugM $ "enterAnn:done:NO ANN:p =" ++ show (p, astId a) return r -enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do +enterAnn !(Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do acceptSpan <- getAcceptSpan setAcceptSpan False case anchor' of @@ -421,9 +412,11 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do _ -> return () case anchor' of EpaDelta _ _ dcs -> do - debugM $ "enterAnn:Printing comments:" ++ showGhc (priorComments cs) + debugM $ "enterAnn:Delta:Flushing comments" + flushComments [] + debugM $ "enterAnn:Delta:Printing prior comments:" ++ showGhc (priorComments cs) mapM_ printOneComment (concatMap tokComment $ priorComments cs) - debugM $ "enterAnn:Printing EpaDelta comments:" ++ showGhc dcs + debugM $ "enterAnn:Delta:Printing EpaDelta comments:" ++ showGhc dcs mapM_ printOneComment (concatMap tokComment dcs) _ -> do debugM $ "enterAnn:Adding comments:" ++ showGhc (priorComments cs) @@ -465,7 +458,7 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- The first part corresponds to the delta phase, so should only use -- delta phase variables ----------------------------------- -- Calculate offset required to get to the start of the SrcSPan - off <- getLayoutOffsetD + !off <- getLayoutOffsetD let spanStart = ss2pos curAnchor priorEndAfterComments <- getPriorEndD let edp' = adjustDeltaForOffset @@ -480,17 +473,18 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- --------------------------------------------- med <- getExtraDP setExtraDP Nothing - let edp = case med of - Nothing -> edp'' - Just (EpaDelta _ dp _) -> dp + let (edp, medr) = case med of + Nothing -> (edp'', Nothing) + Just (EpaDelta _ dp _) -> (dp, Nothing) -- Replace original with desired one. Allows all -- list entry values to be DP (1,0) - Just (EpaSpan (RealSrcSpan r _)) -> dp + Just (EpaSpan (RealSrcSpan r _)) -> (dp, Just dp) where dp = adjustDeltaForOffset off (ss2delta priorEndAfterComments r) Just (EpaSpan (UnhelpfulSpan r)) -> panic $ "enterAnn: UnhelpfulSpan:" ++ show r when (isJust med) $ debugM $ "enterAnn:(med,edp)=" ++ showAst (med,edp) + when (isJust medr) $ setExtraDPReturn medr -- --------------------------------------------- -- Preparation complete, perform the action when (priorEndAfterComments < spanStart) (do @@ -511,12 +505,15 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do debugM $ "enterAnn:exact a starting:" ++ show (showAst anchor') a' <- exact a debugM $ "enterAnn:exact a done:" ++ show (showAst anchor') + + -- Core recursive exactprint done, start end of Entry processing + when (flush == FlushComments) $ do - debugM $ "flushing comments in enterAnn:" ++ showAst cs + debugM $ "flushing comments in enterAnn:" ++ showAst (cs, getFollowingComments cs) flushComments (getFollowingComments cs) debugM $ "flushing comments in enterAnn done" - eof <- getEofPos + !eof <- getEofPos case eof of Nothing -> return () Just (pos, prior) -> do @@ -544,28 +541,50 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- Outside the anchor, mark any trailing postCs <- cua canUpdateAnchor takeAppliedCommentsPop - when (flush == NoFlushComments) $ do - when ((getFollowingComments cs) /= []) $ do - - -- debugM $ "enterAnn:in:(anchor') =" ++ show (eloc2str anchor') - debugM $ "starting trailing comments:" ++ showAst (getFollowingComments cs) - mapM_ printOneComment (concatMap tokComment $ getFollowingComments cs) - debugM $ "ending trailing comments" - trailing' <- markTrailing trailing_anns + following <- if (flush == NoFlushComments) + then do + let (before, after) = splitAfterTrailingAnns trailing_anns + (getFollowingComments cs) + addCommentsA before + return after + else return [] + !trailing' <- markTrailing trailing_anns + -- mapM_ printOneComment (concatMap tokComment $ following) + addCommentsA following -- Update original anchor, comments based on the printing process -- TODO:AZ: probably need to put something appropriate in instead of noSrcSpan - let newAchor = EpaDelta noSrcSpan edp [] + let newAnchor = EpaDelta noSrcSpan edp [] let r = case canUpdateAnchor of - CanUpdateAnchor -> setAnnotationAnchor a' newAchor trailing' (mkEpaComments (priorCs ++ postCs) []) - CanUpdateAnchorOnly -> setAnnotationAnchor a' newAchor [] emptyComments + CanUpdateAnchor -> setAnnotationAnchor a' newAnchor trailing' (mkEpaComments priorCs postCs) + CanUpdateAnchorOnly -> setAnnotationAnchor a' newAnchor [] emptyComments NoCanUpdateAnchor -> a' return r -- --------------------------------------------------------------------- +-- | Split the span following comments into ones that occur prior to +-- the last trailing ann, and ones after. +splitAfterTrailingAnns :: [TrailingAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitAfterTrailingAnns [] cs = ([], cs) +splitAfterTrailingAnns tas cs = (before, after) + where + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + (before, after) = case reverse (concatMap trailing_loc tas) of + [] -> ([],cs) + (s:_) -> (b,a) + where + s_pos = ss2pos s + (b,a) = break (\(L ll _) -> (ss2pos $ anchor ll) > s_pos) + cs + + +-- --------------------------------------------------------------------- + addCommentsA :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -addCommentsA csNew = addComments (concatMap tokComment csNew) +addCommentsA csNew = addComments False (concatMap tokComment csNew) {- TODO: When we addComments, some may have an anchor that is no longer @@ -583,24 +602,36 @@ By definition it is the current anchor, so work against that. And that also means that the first entry comment that has moved should not have a line offset. -} -addComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -addComments csNew = do - -- debugM $ "addComments:" ++ show csNew +addComments :: (Monad m, Monoid w) => Bool -> [Comment] -> EP w m () +addComments sortNeeded csNew = do + debugM $ "addComments:csNew" ++ show csNew cs <- getUnallocatedComments + debugM $ "addComments:cs" ++ show cs + -- We can only sort the comments if we are in the first phase, + -- where all comments have locations. If any have EpaDelta the + -- sort will fail, so we do not try. + if sortNeeded && all noDelta (csNew ++ cs) + then putUnallocatedComments (sort (cs ++ csNew)) + else putUnallocatedComments (cs ++ csNew) - putUnallocatedComments (sort (cs ++ csNew)) +noDelta :: Comment -> Bool +noDelta c = case commentLoc c of + EpaSpan _ -> True + _ -> False -- --------------------------------------------------------------------- -- | Just before we print out the EOF comments, flush the remaining -- ones in the state. flushComments :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -flushComments trailing_anns = do +flushComments !trailing_anns = do + debugM $ "flushComments entered: " ++ showAst trailing_anns addCommentsA trailing_anns + debugM $ "flushComments after addCommentsA" cs <- getUnallocatedComments - debugM $ "flushing comments starting" - -- AZ:TODO: is the sort still needed? - mapM_ printOneComment (sortComments cs) + debugM $ "flushComments: got cs" + debugM $ "flushing comments starting: cs" ++ showAst cs + mapM_ printOneComment cs putUnallocatedComments [] debugM $ "flushing comments done" @@ -612,7 +643,7 @@ annotationsToComments :: (Monad m, Monoid w) => a -> Lens a [AddEpAnn] -> [AnnKeywordId] -> EP w m a annotationsToComments a l kws = do let (newComments, newAnns) = go ([],[]) (view l a) - addComments newComments + addComments True newComments return (set l (reverse newAnns) a) where keywords = Set.fromList kws @@ -654,14 +685,11 @@ printSourceText (NoSourceText) txt = printStringAdvance txt >> return () printSourceText (SourceText txt) _ = printStringAdvance (unpackFS txt) >> return () printSourceTextAA :: (Monad m, Monoid w) => SourceText -> String -> EP w m () -printSourceTextAA (NoSourceText) txt = printStringAtAA noAnn txt >> return () -printSourceTextAA (SourceText txt) _ = printStringAtAA noAnn (unpackFS txt) >> return () +printSourceTextAA (NoSourceText) txt = printStringAdvanceA txt >> return () +printSourceTextAA (SourceText txt) _ = printStringAdvanceA (unpackFS txt) >> return () -- --------------------------------------------------------------------- -printStringAtSs :: (Monad m, Monoid w) => SrcSpan -> String -> EP w m () -printStringAtSs ss str = printStringAtRs (realSrcSpan ss) str >> return () - printStringAtRs :: (Monad m, Monoid w) => RealSrcSpan -> String -> EP w m EpaLocation printStringAtRs pa str = printStringAtRsC CaptureComments pa str @@ -676,7 +704,7 @@ printStringAtRsC capture pa str = do p' <- adjustDeltaForOffsetM p debugM $ "printStringAtRsC:(p,p')=" ++ show (p,p') printStringAtLsDelta p' str - setPriorEndASTD True pa + setPriorEndASTD pa cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -709,6 +737,9 @@ printStringAtMLocL (EpAnn anc an cs) l s = do printStringAtLsDelta (SameLine 1) str return (Just (EpaDelta noSrcSpan (SameLine 1) [])) +printStringAdvanceA :: (Monad m, Monoid w) => String -> EP w m () +printStringAdvanceA str = printStringAtAA (EpaDelta noSrcSpan (SameLine 0) []) str >> return () + printStringAtAA :: (Monad m, Monoid w) => EpaLocation -> String -> EP w m EpaLocation printStringAtAA el str = printStringAtAAC CaptureComments el str @@ -735,7 +766,7 @@ printStringAtAAC capture (EpaDelta ss d cs) s = do p2 <- getPosP pe2 <- getPriorEndD debugM $ "printStringAtAA:(pe1,pe2,p1,p2)=" ++ show (pe1,pe2,p1,p2) - setPriorEndASTPD True (pe1,pe2) + setPriorEndASTPD (pe1,pe2) cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -883,8 +914,7 @@ markAnnOpenP' :: (Monad m, Monoid w) => AnnPragma -> SourceText -> String -> EP markAnnOpenP' an NoSourceText txt = markEpAnnLMS0 an lapr_open AnnOpen (Just txt) markAnnOpenP' an (SourceText txt) _ = markEpAnnLMS0 an lapr_open AnnOpen (Just $ unpackFS txt) -markAnnOpen :: (Monad m, Monoid w) - => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] +markAnnOpen :: (Monad m, Monoid w) => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] markAnnOpen an NoSourceText txt = markEpAnnLMS'' an lidl AnnOpen (Just txt) markAnnOpen an (SourceText txt) _ = markEpAnnLMS'' an lidl AnnOpen (Just $ unpackFS txt) @@ -1589,7 +1619,7 @@ markTopLevelList ls = mapM (\a -> setLayoutTopLevelP $ markAnnotated a) ls instance (ExactPrint a) => ExactPrint (Located a) where getAnnotationEntry (L l _) = case l of UnhelpfulSpan _ -> NoEntryVal - _ -> Entry (hackSrcSpanToAnchor l) [] emptyComments NoFlushComments CanUpdateAnchorOnly + _ -> Entry (EpaSpan l) [] emptyComments NoFlushComments CanUpdateAnchorOnly setAnnotationAnchor (L l a) _anc _ts _cs = L l a @@ -1664,16 +1694,10 @@ instance ExactPrint (HsModule GhcPs) where _ -> return lo am_decls' <- markTrailing (am_decls $ anns an0) - imports' <- markTopLevelList imports - - case lo of - EpExplicitBraces _ _ -> return () - _ -> do - -- Get rid of the balance of the preceding comments before starting on the decls - flushComments [] - putUnallocatedComments [] - decls' <- markTopLevelList (filter removeDocDecl decls) + mid <- markAnnotated (HsModuleImpDecls (am_cs $ anns an0) imports decls) + let imports' = id_imps mid + let decls' = id_decls mid lo1 <- case lo0 of EpExplicitBraces open close -> do @@ -1688,15 +1712,32 @@ instance ExactPrint (HsModule GhcPs) where debugM $ "am_eof:" ++ showGhc (pos, prior) setEofPos (Just (pos, prior)) - let anf = an0 { anns = (anns an0) { am_decls = am_decls' }} + let anf = an0 { anns = (anns an0) { am_decls = am_decls', am_cs = [] }} debugM $ "HsModule, anf=" ++ showAst anf return (HsModule (XModulePs anf lo1 mdeprec' mbDoc') mmn' mexports' imports' decls') +-- --------------------------------------------------------------------- + +-- | This is used to ensure the comments are updated into the right +-- place for makeDeltaAst. +data HsModuleImpDecls + = HsModuleImpDecls { + id_cs :: [LEpaComment], + id_imps :: [LImportDecl GhcPs], + id_decls :: [LHsDecl GhcPs] + } deriving Data + +instance ExactPrint HsModuleImpDecls where + -- Use an UnhelpfulSpan for the anchor, we are only interested in the comments + getAnnotationEntry mid = mkEntry (EpaSpan (UnhelpfulSpan UnhelpfulNoLocationInfo)) [] (EpaComments (id_cs mid)) + setAnnotationAnchor mid _anc _ cs = mid { id_cs = priorComments cs ++ getFollowingComments cs } + `debug` ("HsModuleImpDecls.setAnnotationAnchor:cs=" ++ showAst cs) + exact (HsModuleImpDecls cs imports decls) = do + imports' <- markTopLevelList imports + decls' <- markTopLevelList (filter notDocDecl decls) + return (HsModuleImpDecls cs imports' decls') -removeDocDecl :: LHsDecl GhcPs -> Bool -removeDocDecl (L _ DocD{}) = False -removeDocDecl _ = True -- --------------------------------------------------------------------- @@ -1737,8 +1778,8 @@ instance ExactPrint InWarningCategory where exact (InWarningCategory tkIn source (L l wc)) = do tkIn' <- markEpToken tkIn - L _ (_,wc') <- markAnnotated (L l (source, wc)) - return (InWarningCategory tkIn' source (L l wc')) + L l' (_,wc') <- markAnnotated (L l (source, wc)) + return (InWarningCategory tkIn' source (L l' wc')) instance ExactPrint (SourceText, WarningCategory) where getAnnotationEntry _ = NoEntryVal @@ -1943,14 +1984,14 @@ exactDataFamInstDecl an top_lvl , feqn_pats = pats , feqn_fixity = fixity , feqn_rhs = defn })) = do - (an', an2', tycon', bndrs', _, _mc, defn') <- exactDataDefn an2 pp_hdr defn - -- See Note [an and an2 in exactDataFamInstDecl] + (an', an2', tycon', bndrs', pats', defn') <- exactDataDefn an2 pp_hdr defn + -- See Note [an and an2 in exactDataFamInstDecl] return (an', DataFamInstDecl ( FamEqn { feqn_ext = an2' , feqn_tycon = tycon' , feqn_bndrs = bndrs' - , feqn_pats = pats + , feqn_pats = pats' , feqn_fixity = fixity , feqn_rhs = defn' })) `debug` ("exactDataFamInstDecl: defn' derivs:" ++ showAst (dd_derivs defn')) @@ -2233,11 +2274,11 @@ instance ExactPrint (RoleAnnotDecl GhcPs) where an1 <- markEpAnnL an0 lidl AnnRole ltycon' <- markAnnotated ltycon let markRole (L l (Just r)) = do - (L _ r') <- markAnnotated (L l r) - return (L l (Just r')) + (L l' r') <- markAnnotated (L l r) + return (L l' (Just r')) markRole (L l Nothing) = do - printStringAtSs (locA l) "_" - return (L l Nothing) + e' <- printStringAtAA (entry l) "_" + return (L (l { entry = e'}) Nothing) roles' <- mapM markRole roles return (RoleAnnotDecl an1 ltycon' roles') @@ -2340,8 +2381,13 @@ instance (ExactPrint tm, ExactPrint ty, Outputable tm, Outputable ty) getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact a@(HsValArg _ tm) = markAnnotated tm >> return a - exact a@(HsTypeArg at ty) = markEpToken at >> markAnnotated ty >> return a + exact (HsValArg x tm) = do + tm' <- markAnnotated tm + return (HsValArg x tm') + exact (HsTypeArg at ty) = do + at' <- markEpToken at + ty' <- markAnnotated ty + return (HsTypeArg at' ty') exact x@(HsArgPar _sp) = withPpr x -- Does not appear in original source -- --------------------------------------------------------------------- @@ -2359,9 +2405,9 @@ instance ExactPrint (ClsInstDecl GhcPs) where (mbWarn', an0, mbOverlap', inst_ty') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lid AnnSemi - ds <- withSortKey sortKey - [(ClsAtdTag, prepareListAnnotationA ats), - (ClsAtdTag, prepareListAnnotationF an adts), + (sortKey', ds) <- withSortKey sortKey + [(ClsAtTag, prepareListAnnotationA ats), + (ClsAtdTag, prepareListAnnotationF adts), (ClsMethodTag, prepareListAnnotationA binds), (ClsSigTag, prepareListAnnotationA sigs) ] @@ -2371,7 +2417,7 @@ instance ExactPrint (ClsInstDecl GhcPs) where adts' = undynamic ds binds' = undynamic ds sigs' = undynamic ds - return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey) + return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey') , cid_poly_ty = inst_ty', cid_binds = binds' , cid_sigs = sigs', cid_tyfam_insts = ats' , cid_overlap_mode = mbOverlap' @@ -2452,15 +2498,29 @@ instance ExactPrint (HsBind GhcPs) where return (FunBind x fun_id' matches') exact (PatBind x pat q grhss) = do + q' <- markAnnotated q pat' <- markAnnotated pat grhss' <- markAnnotated grhss - return (PatBind x pat' q grhss') + return (PatBind x pat' q' grhss') exact (PatSynBind x bind) = do bind' <- markAnnotated bind return (PatSynBind x bind') exact x = error $ "HsBind: exact for " ++ showAst x +instance ExactPrint (HsMultAnn GhcPs) where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact (HsNoMultAnn x) = return (HsNoMultAnn x) + exact (HsPct1Ann tok) = do + tok' <- markEpToken tok + return (HsPct1Ann tok') + exact (HsMultAnn tok ty) = do + tok' <- markEpToken tok + ty' <- markAnnotated ty + return (HsMultAnn tok' ty') + -- --------------------------------------------------------------------- instance ExactPrint (PatSynBind GhcPs GhcPs) where @@ -2519,8 +2579,9 @@ instance ExactPrint (PatSynBind GhcPs GhcPs) where instance ExactPrint (RecordPatSynField GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact r@(RecordPatSynField { recordPatSynField = v }) = markAnnotated v - >> return r + exact (RecordPatSynField f v) = do + f' <- markAnnotated f + return (RecordPatSynField f' v) -- --------------------------------------------------------------------- @@ -2648,15 +2709,20 @@ instance ExactPrint (HsLocalBinds GhcPs) where (an1, valbinds') <- markAnnList an0 $ markAnnotatedWithLayout valbinds debugM $ "exact HsValBinds: an1=" ++ showAst an1 - return (HsValBinds an1 valbinds') + medr <- getExtraDPReturn + an2 <- case medr of + Nothing -> return an1 + Just dp -> do + setExtraDPReturn Nothing + return $ an1 { anns = (anns an1) { al_anchor = Just (EpaDelta noSrcSpan dp []) }} + return (HsValBinds an2 valbinds') exact (HsIPBinds an bs) = do - (as, ipb) <- markAnnList an (markEpAnnL' an lal_rest AnnWhere - >> markAnnotated bs - >>= \bs' -> return (HsIPBinds an bs'::HsLocalBinds GhcPs)) - case ipb of - HsIPBinds _ bs' -> return (HsIPBinds as bs'::HsLocalBinds GhcPs) - _ -> error "should not happen HsIPBinds" + (an2,bs') <- markAnnListA an $ \an0 -> do + an1 <- markEpAnnL' an0 lal_rest AnnWhere + bs' <- markAnnotated bs + return (an1, bs') + return (HsIPBinds an2 bs') exact b@(EmptyLocalBinds _) = return b @@ -2670,7 +2736,8 @@ instance ExactPrint (HsValBindsLR GhcPs GhcPs) where let binds' = concatMap decl2Bind decls sigs' = concatMap decl2Sig decls - return (ValBinds sortKey binds' sigs') + sortKey' = captureOrderBinds decls + return (ValBinds sortKey' binds' sigs') exact (XValBindsLR _) = panic "XValBindsLR" undynamic :: Typeable a => [Dynamic] -> [a] @@ -2682,7 +2749,9 @@ instance ExactPrint (HsIPBinds GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact b@(IPBinds _ binds) = setLayoutBoth $ markAnnotated binds >> return b + exact (IPBinds x binds) = setLayoutBoth $ do + binds' <- markAnnotated binds + return (IPBinds x binds') -- --------------------------------------------------------------------- @@ -2703,18 +2772,18 @@ instance ExactPrint HsIPName where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact i@(HsIPName fs) = printStringAdvance ("?" ++ (unpackFS fs)) >> return i + exact i@(HsIPName fs) = printStringAdvanceA ("?" ++ (unpackFS fs)) >> return i -- --------------------------------------------------------------------- -- Managing lists which have been separated, e.g. Sigs and Binds prepareListAnnotationF :: (Monad m, Monoid w) => - [AddEpAnn] -> [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] -prepareListAnnotationF an ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls + [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] +prepareListAnnotationF ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls where go (L l a) = do - d' <- markAnnotated (DataFamInstDeclWithContext an NotTopLevel a) - return (toDyn (L l (dc_d d'))) + (L l' d') <- markAnnotated (L l (DataFamInstDeclWithContext noAnn NotTopLevel a)) + return (toDyn (L l' (dc_d d'))) prepareListAnnotationA :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => [LocatedAn an a] -> [(RealSrcSpan,EP w m Dynamic)] @@ -2725,15 +2794,23 @@ prepareListAnnotationA ls = map (\b -> (realSrcSpan $ getLocA b,go b)) ls return (toDyn b') withSortKey :: (Monad m, Monoid w) - => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] -> EP w m [Dynamic] + => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] + -> EP w m (AnnSortKey DeclTag, [Dynamic]) withSortKey annSortKey xs = do debugM $ "withSortKey:annSortKey=" ++ showAst annSortKey - let ordered = case annSortKey of - NoAnnSortKey -> sortBy orderByFst $ concatMap snd xs - AnnSortKey _keys -> orderedDecls annSortKey (Map.fromList xs) - mapM snd ordered -orderByFst :: Ord a => (a, b1) -> (a, b2) -> Ordering -orderByFst (a,_) (b,_) = compare a b + let (sk, ordered) = case annSortKey of + NoAnnSortKey -> (annSortKey', map snd os) + where + doOne (tag, ds) = map (\d -> (tag, d)) ds + xsExpanded = concatMap doOne xs + os = sortBy orderByFst $ xsExpanded + annSortKey' = AnnSortKey (map fst os) + AnnSortKey _keys -> (annSortKey, orderedDecls annSortKey (Map.fromList xs)) + ordered' <- mapM snd ordered + return (sk, ordered') + +orderByFst :: Ord a => (t, (a,b1)) -> (t, (a, b2)) -> Ordering +orderByFst (_,(a,_)) (_,(b,_)) = compare a b -- --------------------------------------------------------------------- @@ -2761,15 +2838,16 @@ instance ExactPrint (Sig GhcPs) where (an0, vars',ty') <- exactVarSig an vars ty return (ClassOpSig an0 is_deflt vars' ty') - exact (FixSig (an,src) (FixitySig x names (Fixity v fdir))) = do + exact (FixSig (an,src) (FixitySig ns names (Fixity v fdir))) = do let fixstr = case fdir of InfixL -> "infixl" InfixR -> "infixr" InfixN -> "infix" an0 <- markEpAnnLMS'' an lidl AnnInfix (Just fixstr) an1 <- markEpAnnLMS'' an0 lidl AnnVal (Just (sourceTextToString src (show v))) + ns' <- markAnnotated ns names' <- markAnnotated names - return (FixSig (an1,src) (FixitySig x names' (Fixity v fdir))) + return (FixSig (an1,src) (FixitySig ns' names' (Fixity v fdir))) exact (InlineSig an ln inl) = do an0 <- markAnnOpen an (inl_src inl) "{-# INLINE" @@ -2809,7 +2887,7 @@ instance ExactPrint (Sig GhcPs) where exact (CompleteMatchSig (an,src) cs mty) = do an0 <- markAnnOpen an src "{-# COMPLETE" - cs' <- markAnnotated cs + cs' <- mapM markAnnotated cs (an1, mty') <- case mty of Nothing -> return (an0, mty) @@ -2822,6 +2900,20 @@ instance ExactPrint (Sig GhcPs) where -- --------------------------------------------------------------------- +instance ExactPrint NamespaceSpecifier where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact NoNamespaceSpecifier = return NoNamespaceSpecifier + exact (TypeNamespaceSpecifier typeTok) = do + typeTok' <- markEpToken typeTok + return (TypeNamespaceSpecifier typeTok') + exact (DataNamespaceSpecifier dataTok) = do + dataTok' <- markEpToken dataTok + return (DataNamespaceSpecifier dataTok') + +-- --------------------------------------------------------------------- + exactVarSig :: (Monad m, Monoid w, ExactPrint a) => AnnSig -> [LocatedN RdrName] -> a -> EP w m (AnnSig, [LocatedN RdrName], a) exactVarSig an vars ty = do @@ -2875,7 +2967,7 @@ instance ExactPrint (AnnDecl GhcPs) where n' <- markAnnotated n return (an1, TypeAnnProvenance n') ModuleAnnProvenance -> do - an1 <- markEpAnnL an lapr_rest AnnModule + an1 <- markEpAnnL an0 lapr_rest AnnModule return (an1, prov) e' <- markAnnotated e @@ -2950,21 +3042,21 @@ instance ExactPrint (HsExpr GhcPs) where then markAnnotated n else return n return (HsVar x n') - exact x@(HsUnboundVar an _) = do + exact (HsUnboundVar an n) = do case an of Just (EpAnnUnboundVar (ob,cb) l) -> do - printStringAtAA ob "`" >> return () - printStringAtAA l "_" >> return () - printStringAtAA cb "`" >> return () - return x + ob' <- printStringAtAA ob "`" + l' <- printStringAtAA l "_" + cb' <- printStringAtAA cb "`" + return (HsUnboundVar (Just (EpAnnUnboundVar (ob',cb') l')) n) _ -> do - printStringAtLsDelta (SameLine 0) "_" - return x + printStringAdvanceA "_" >> return () + return (HsUnboundVar an n) exact x@(HsOverLabel src l) = do - printStringAtLsDelta (SameLine 0) "#" + printStringAdvanceA "#" >> return () case src of - NoSourceText -> printStringAtLsDelta (SameLine 0) (unpackFS l) - SourceText txt -> printStringAtLsDelta (SameLine 0) (unpackFS txt) + NoSourceText -> printStringAdvanceA (unpackFS l) >> return () + SourceText txt -> printStringAdvanceA (unpackFS txt) >> return () return x exact x@(HsIPVar _ (HsIPName n)) @@ -3204,11 +3296,11 @@ instance ExactPrint (HsExpr GhcPs) where exact (HsTypedSplice an s) = do an0 <- markEpAnnL an lidl AnnDollarDollar - s' <- exact s + s' <- markAnnotated s return (HsTypedSplice an0 s') exact (HsUntypedSplice an s) = do - s' <- exact s + s' <- markAnnotated s return (HsUntypedSplice an s') exact (HsProc an p c) = do @@ -3274,12 +3366,15 @@ exactMdo an (Just module_name) kw = markEpAnnLMS'' an lal_rest kw (Just n) markMaybeDodgyStmts :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => AnnList -> LocatedAn an a -> EP w m (AnnList, LocatedAn an a) markMaybeDodgyStmts an stmts = - if isGoodSrcSpan (getLocA stmts) + if notDodgy stmts then do r <- markAnnotatedWithLayout stmts return (an, r) else return (an, stmts) +notDodgy :: GenLocated (EpAnn ann) a -> Bool +notDodgy (L (EpAnn anc _ _) _) = notDodgyE anc + notDodgyE :: EpaLocation -> Bool notDodgyE anc = case anc of @@ -3341,7 +3436,7 @@ instance ExactPrint (MatchGroup GhcPs (LocatedA (HsCmd GhcPs))) where setAnnotationAnchor a _ _ _ = a exact (MG x matches) = do -- TODO:AZ use SortKey, in MG ann. - matches' <- if isGoodSrcSpan (getLocA matches) + matches' <- if notDodgy matches then markAnnotated matches else return matches return (MG x matches') @@ -3661,6 +3756,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- There may be arbitrary parens around parts of the constructor -- that are infix. Turn these into comments so that they feed -- into the right place automatically + -- TODO: no longer sorting on insert. What now? an0 <- annotationsToComments an lidl [AnnOpenP,AnnCloseP] an1 <- markEpAnnL an0 lidl AnnType @@ -3674,7 +3770,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- TODO: add a workaround for https://gitlab.haskell.org/ghc/ghc/-/issues/20452 exact (DataDecl { tcdDExt = an, tcdLName = ltycon, tcdTyVars = tyvars , tcdFixity = fixity, tcdDataDefn = defn }) = do - (_, an', ltycon', tyvars', _, _mctxt', defn') <- + (_, an', ltycon', tyvars', _, defn') <- exactDataDefn an (exactVanillaDeclHead ltycon tyvars fixity) defn return (DataDecl { tcdDExt = an', tcdLName = ltycon', tcdTyVars = tyvars' , tcdFixity = fixity, tcdDataDefn = defn' }) @@ -3707,7 +3803,7 @@ instance ExactPrint (TyClDecl GhcPs) where (an0, fds', lclas', tyvars',context') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lidl AnnSemi - ds <- withSortKey sortKey + (sortKey', ds) <- withSortKey sortKey [(ClsSigTag, prepareListAnnotationA sigs), (ClsMethodTag, prepareListAnnotationA methods), (ClsAtTag, prepareListAnnotationA ats), @@ -3720,7 +3816,7 @@ instance ExactPrint (TyClDecl GhcPs) where methods' = undynamic ds ats' = undynamic ds at_defs' = undynamic ds - return (ClassDecl {tcdCExt = (an3, lo, sortKey), + return (ClassDecl {tcdCExt = (an3, lo, sortKey'), tcdCtxt = context', tcdLName = lclas', tcdTyVars = tyvars', tcdFixity = fixity, tcdFDs = fds', @@ -3845,7 +3941,7 @@ exactDataDefn -> HsDataDefn GhcPs -> EP w m ( [AddEpAnn] -- ^ from exactHdr , [AddEpAnn] -- ^ updated one passed in - , LocatedN RdrName, a, b, Maybe (LHsContext GhcPs), HsDataDefn GhcPs) + , LocatedN RdrName, a, b, HsDataDefn GhcPs) exactDataDefn an exactHdr (HsDataDefn { dd_ext = x, dd_ctxt = context , dd_cType = mb_ct @@ -3883,8 +3979,8 @@ exactDataDefn an exactHdr _ -> panic "exacprint NewTypeCon" an6 <- markEpAnnL an5 lidl AnnCloseC derivings' <- mapM markAnnotated derivings - return (anx, an6, ln', tvs', b, mctxt', - (HsDataDefn { dd_ext = x, dd_ctxt = context + return (anx, an6, ln', tvs', b, + (HsDataDefn { dd_ext = x, dd_ctxt = mctxt' , dd_cType = mb_ct' , dd_kindSig = mb_sig' , dd_cons = condecls'', dd_derivs = derivings' })) @@ -3941,22 +4037,23 @@ instance ExactPrint (InjectivityAnn GhcPs) where class Typeable flag => ExactPrintTVFlag flag where exactTVDelimiters :: (Monad m, Monoid w) - => [AddEpAnn] -> flag -> EP w m (HsTyVarBndr flag GhcPs) - -> EP w m ([AddEpAnn], (HsTyVarBndr flag GhcPs)) + => [AddEpAnn] -> flag + -> ([AddEpAnn] -> EP w m ([AddEpAnn], HsTyVarBndr flag GhcPs)) + -> EP w m ([AddEpAnn], flag, (HsTyVarBndr flag GhcPs)) instance ExactPrintTVFlag () where - exactTVDelimiters an _ thing_inside = do + exactTVDelimiters an flag thing_inside = do an0 <- markEpAnnAllL' an lid AnnOpenP - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid AnnCloseP - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid AnnCloseP + return (an2, flag, r) instance ExactPrintTVFlag Specificity where exactTVDelimiters an s thing_inside = do an0 <- markEpAnnAllL' an lid open - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid close - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid close + return (an2, s, r) where (open, close) = case s of SpecifiedSpec -> (AnnOpenP, AnnCloseP) @@ -3964,33 +4061,33 @@ instance ExactPrintTVFlag Specificity where instance ExactPrintTVFlag (HsBndrVis GhcPs) where exactTVDelimiters an0 bvis thing_inside = do - case bvis of - HsBndrRequired _ -> return () - HsBndrInvisible at -> markEpToken at >> return () + bvis' <- case bvis of + HsBndrRequired _ -> return bvis + HsBndrInvisible at -> HsBndrInvisible <$> markEpToken at an1 <- markEpAnnAllL' an0 lid AnnOpenP - r <- thing_inside - an2 <- markEpAnnAllL' an1 lid AnnCloseP - return (an2, r) + (an2, r) <- thing_inside an1 + an3 <- markEpAnnAllL' an2 lid AnnCloseP + return (an3, bvis', r) instance ExactPrintTVFlag flag => ExactPrint (HsTyVarBndr flag GhcPs) where getAnnotationEntry _ = NoEntryVal setAnnotationAnchor a _ _ _ = a exact (UserTyVar an flag n) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - return (UserTyVar an flag n') + return (ani, UserTyVar an flag n') case r of - (an', UserTyVar _ flag'' n'') -> return (UserTyVar an' flag'' n'') + (an', flag', UserTyVar _ _ n'') -> return (UserTyVar an' flag' n'') _ -> error "KindedTyVar should never happen here" exact (KindedTyVar an flag n k) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - an0 <- markEpAnnL an lidl AnnDcolon + an0 <- markEpAnnL ani lidl AnnDcolon k' <- markAnnotated k - return (KindedTyVar an0 flag n' k') + return (an0, KindedTyVar an0 flag n' k') case r of - (an',KindedTyVar _ flag'' n'' k'') -> return (KindedTyVar an' flag'' n'' k'') + (an',flag', KindedTyVar _ _ n'' k'') -> return (KindedTyVar an' flag' n'' k'') _ -> error "UserTyVar should never happen here" -- --------------------------------------------------------------------- @@ -4150,17 +4247,16 @@ instance ExactPrint (HsDerivingClause GhcPs) where , deriv_clause_strategy = dcs , deriv_clause_tys = dct }) = do an0 <- markEpAnnL an lidl AnnDeriving - exact_strat_before + dcs0 <- case dcs of + Just (L _ ViaStrategy{}) -> return dcs + _ -> mapM markAnnotated dcs dct' <- markAnnotated dct - exact_strat_after + dcs1 <- case dcs0 of + Just (L _ ViaStrategy{}) -> mapM markAnnotated dcs0 + _ -> return dcs0 return (HsDerivingClause { deriv_clause_ext = an0 - , deriv_clause_strategy = dcs + , deriv_clause_strategy = dcs1 , deriv_clause_tys = dct' }) - where - (exact_strat_before, exact_strat_after) = - case dcs of - Just v@(L _ ViaStrategy{}) -> (pure (), markAnnotated v >> pure ()) - _ -> (mapM_ markAnnotated dcs, pure ()) -- --------------------------------------------------------------------- @@ -4467,7 +4563,9 @@ instance ExactPrint (ConDeclField GhcPs) where instance ExactPrint (FieldOcc GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact f@(FieldOcc _ n) = markAnnotated n >> return f + exact (FieldOcc x n) = do + n' <- markAnnotated n + return (FieldOcc x n') -- --------------------------------------------------------------------- @@ -4535,7 +4633,7 @@ instance ExactPrint (LocatedL [LocatedA (IE GhcPs)]) where an0 <- markEpAnnL' an lal_rest AnnHiding p <- getPosP debugM $ "LocatedL [LIE:p=" ++ showPprUnsafe p - (an1, ies') <- markAnnList an0 (markAnnotated ies) + (an1, ies') <- markAnnList an0 (markAnnotated (filter notIEDoc ies)) return (L an1 ies') instance (ExactPrint (Match GhcPs (LocatedA body))) @@ -4985,6 +5083,14 @@ setExtraDP md = do debugM $ "setExtraDP:" ++ show md modify (\s -> s {uExtraDP = md}) +getExtraDPReturn :: (Monad m, Monoid w) => EP w m (Maybe DeltaPos) +getExtraDPReturn = gets uExtraDPReturn + +setExtraDPReturn :: (Monad m, Monoid w) => Maybe DeltaPos -> EP w m () +setExtraDPReturn md = do + debugM $ "setExtraDPReturn:" ++ show md + modify (\s -> s {uExtraDPReturn = md}) + getPriorEndD :: (Monad m, Monoid w) => EP w m Pos getPriorEndD = gets dPriorEndPosition @@ -5007,13 +5113,13 @@ setPriorEndNoLayoutD pe = do debugM $ "setPriorEndNoLayoutD:pe=" ++ show pe modify (\s -> s { dPriorEndPosition = pe }) -setPriorEndASTD :: (Monad m, Monoid w) => Bool -> RealSrcSpan -> EP w m () -setPriorEndASTD layout pe = setPriorEndASTPD layout (rs2range pe) +setPriorEndASTD :: (Monad m, Monoid w) => RealSrcSpan -> EP w m () +setPriorEndASTD pe = setPriorEndASTPD (rs2range pe) -setPriorEndASTPD :: (Monad m, Monoid w) => Bool -> (Pos,Pos) -> EP w m () -setPriorEndASTPD layout pe@(fm,to) = do +setPriorEndASTPD :: (Monad m, Monoid w) => (Pos,Pos) -> EP w m () +setPriorEndASTPD pe@(fm,to) = do debugM $ "setPriorEndASTD:pe=" ++ show pe - when layout $ setLayoutStartD (snd fm) + setLayoutStartD (snd fm) modify (\s -> s { dPriorEndPosition = to } ) setLayoutStartD :: (Monad m, Monoid w) => Int -> EP w m () @@ -5044,7 +5150,7 @@ getUnallocatedComments :: (Monad m, Monoid w) => EP w m [Comment] getUnallocatedComments = gets epComments putUnallocatedComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -putUnallocatedComments cs = modify (\s -> s { epComments = cs } ) +putUnallocatedComments !cs = modify (\s -> s { epComments = cs } ) -- | Push a fresh stack frame for the applied comments gatherer pushAppliedComments :: (Monad m, Monoid w) => EP w m () @@ -5054,7 +5160,7 @@ pushAppliedComments = modify (\s -> s { epCommentsApplied = []:(epCommentsApplie -- takeAppliedComments, and clear them, not popping the stack takeAppliedComments :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedComments = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5067,7 +5173,7 @@ takeAppliedComments = do -- takeAppliedComments, and clear them, popping the stack takeAppliedCommentsPop :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedCommentsPop = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5080,7 +5186,7 @@ takeAppliedCommentsPop = do -- when doing delta processing applyComment :: (Monad m, Monoid w) => Comment -> EP w m () applyComment c = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> modify (\s -> s { epCommentsApplied = [[c]] } ) (h:t) -> modify (\s -> s { epCommentsApplied = (c:h):t } ) ===================================== utils/check-exact/Main.hs ===================================== @@ -470,7 +470,7 @@ changeAddDecl1 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtStart m decl' + replaceTopLevelDecls m = return $ insertAtStart m decl' return p' -- --------------------------------------------------------------------- @@ -483,7 +483,7 @@ changeAddDecl2 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtEnd m decl' + replaceTopLevelDecls m = return $ insertAtEnd m decl' return p' -- --------------------------------------------------------------------- @@ -500,7 +500,7 @@ changeAddDecl3 libdir top = do l2' = setEntryDP l2 (DifferentLine 2 0) replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAt f m decl' + replaceTopLevelDecls m = return $ insertAt f m decl' return p' -- --------------------------------------------------------------------- @@ -571,8 +571,9 @@ changeLocalDecls2 libdir (L l p) = do changeWhereIn3a :: Changer changeWhereIn3a _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) - debugM $ unlines w + decls = balanceCommentsList decls0 + (_de0:_:de1:_d2:_) = decls + debugM $ "changeWhereIn3a:de1:" ++ showAst de1 let p2 = p { hsmodDecls = decls} return (L l p2) @@ -581,13 +582,12 @@ changeWhereIn3a _libdir (L l p) = do changeWhereIn3b :: Changer changeWhereIn3b _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) + decls = balanceCommentsList decls0 (de0:tdecls@(_:de1:d2:_)) = decls de0' = setEntryDP de0 (DifferentLine 2 0) de1' = setEntryDP de1 (DifferentLine 2 0) d2' = setEntryDP d2 (DifferentLine 2 0) decls' = d2':de1':de0':tdecls - debugM $ unlines w debugM $ "changeWhereIn3b:de1':" ++ showAst de1' let p2 = p { hsmodDecls = decls'} return (L l p2) @@ -598,37 +598,37 @@ addLocaLDecl1 :: Changer addLocaLDecl1 libdir top = do Right (L ld (ValD _ decl)) <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let decl' = setEntryDP (L ld decl) (DifferentLine 1 5) - doAddLocal = do - let lp = top - (de1:d2:d3:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - (de1',_) <- modifyValD (getLocA de1'') de1'' $ \_m d -> do - return ((wrapDecl decl' : d),Nothing) - replaceDecls lp [de1', d2', d3] - - (lp',_,w) <- runTransformT doAddLocal - debugM $ "addLocaLDecl1:" ++ intercalate "\n" w + doAddLocal :: ParsedSource + doAddLocal = replaceDecls lp [de1', d2', d3] + where + lp = top + (de1:d2:d3:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 + (de1',_) = modifyValD (getLocA de1'') de1'' $ \_m d -> ((wrapDecl decl' : d),Nothing) + + let lp' = doAddLocal return lp' -- --------------------------------------------------------------------- + addLocaLDecl2 :: Changer addLocaLDecl2 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 + doAddLocal = replaceDecls lp [parent',d2'] + where + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - newDecl' <- transferEntryDP' d newDecl - let d' = setEntryDP d (DifferentLine 1 0) - return ((newDecl':d':ds),Nothing) + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = transferEntryDP' d (makeDeltaAst newDecl) + d' = setEntryDP d (DifferentLine 1 0) + in ((newDecl':d':ds),Nothing) - replaceDecls lp [parent',d2'] - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -637,19 +637,18 @@ addLocaLDecl3 :: Changer addLocaLDecl3 libdir top = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - let lp = top - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - return (((d:ds) ++ [newDecl']),Nothing) + doAddLocal = replaceDecls (anchorEof lp) [parent',d2'] + where + lp = top + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - replaceDecls (anchorEof lp) [parent',d2'] + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = setEntryDP newDecl (DifferentLine 1 0) + in (((d:ds) ++ [newDecl']),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -659,40 +658,38 @@ addLocaLDecl4 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") Right newSig <- withDynFlags libdir (\df -> parseDecl df "sig" "nn :: Int") let - doAddLocal = do - (parent:ds) <- hsDecls lp + doAddLocal = replaceDecls (anchorEof lp) (parent':ds) + where + (parent:ds) = hsDecls (makeDeltaAst lp) - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - let newSig' = setEntryDP newSig (DifferentLine 1 4) + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 0) + newSig' = setEntryDP (makeDeltaAst newSig) (DifferentLine 1 5) - (parent',_) <- modifyValD (getLocA parent) parent $ \_m decls -> do - return ((decls++[newSig',newDecl']),Nothing) + (parent',_) = modifyValD (getLocA parent) parent $ \_m decls -> + ((decls++[newSig',newDecl']),Nothing) - replaceDecls (anchorEof lp) (parent':ds) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' - -- --------------------------------------------------------------------- addLocaLDecl5 :: Changer addLocaLDecl5 _libdir lp = do let - doAddLocal = do - decls <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList decls + doAddLocal = replaceDecls lp [s1,de1',d3'] + where + decls = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList decls - let d3' = setEntryDP d3 (DifferentLine 2 0) + d3' = setEntryDP d3 (DifferentLine 2 0) - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m _decls -> do - let d2' = setEntryDP d2 (DifferentLine 1 0) - return ([d2'],Nothing) - replaceDecls lp [s1,de1',d3'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m _decls -> + let + d2' = setEntryDP d2 (DifferentLine 1 0) + in ([d2'],Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -701,39 +698,36 @@ addLocaLDecl6 :: Changer addLocaLDecl6 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "x = 3") let - newDecl' = setEntryDP newDecl (DifferentLine 1 4) - doAddLocal = do - decls0 <- hsDecls lp - [de1'',d2] <- balanceCommentsList decls0 + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 5) + doAddLocal = replaceDecls lp [de1', d2] + where + decls0 = hsDecls lp + [de1'',d2] = balanceCommentsList decls0 - let de1 = captureMatchLineSpacing de1'' - let L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 - let [ma1,_ma2] = ms + de1 = captureMatchLineSpacing de1'' + L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 + [ma1,_ma2] = ms - (de1',_) <- modifyValD (getLocA ma1) de1 $ \_m decls -> do - return ((newDecl' : decls),Nothing) - replaceDecls lp [de1', d2] + (de1',_) = modifyValD (getLocA ma1) de1 $ \_m decls -> + ((newDecl' : decls),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- rmDecl1 :: Changer -rmDecl1 _libdir top = do - let doRmDecl = do - let lp = top - tlDecs0 <- hsDecls lp - tlDecs' <- balanceCommentsList tlDecs0 - let tlDecs = captureLineSpacing tlDecs' - let (de1:_s1:_d2:d3:ds) = tlDecs - let d3' = setEntryDP d3 (DifferentLine 2 0) - - replaceDecls lp (de1:d3':ds) - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" +rmDecl1 _libdir lp = do + let + doRmDecl = replaceDecls lp (de1:d3':ds) + where + tlDecs0 = hsDecls lp + tlDecs = balanceCommentsList tlDecs0 + (de1:_s1:_d2:d3:ds) = tlDecs + d3' = setEntryDP d3 (DifferentLine 2 0) + + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -745,13 +739,13 @@ rmDecl2 _libdir lp = do let go :: GHC.LHsExpr GhcPs -> Transform (GHC.LHsExpr GhcPs) go e@(GHC.L _ (GHC.HsLet{})) = do - decs0 <- hsDecls e - decs <- balanceCommentsList $ captureLineSpacing decs0 - e' <- replaceDecls e (init decs) + let decs0 = hsDecls e + let decs = balanceCommentsList $ captureLineSpacing decs0 + let e' = replaceDecls e (init decs) return e' go x = return x - everywhereM (mkM go) lp + everywhereM (mkM go) (makeDeltaAst lp) let (lp',_,_w) = runTransform doRmDecl debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" @@ -762,17 +756,15 @@ rmDecl2 _libdir lp = do rmDecl3 :: Changer rmDecl3 _libdir lp = do let - doRmDecl = do - [de1,d2] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1] -> do - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([],Just sd1') - - replaceDecls lp [de1',sd1,d2] + doRmDecl = replaceDecls lp [de1',sd1,d2] + where + [de1,d2] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a] -> + let + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([],Just sd1') - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -780,19 +772,15 @@ rmDecl3 _libdir lp = do rmDecl4 :: Changer rmDecl4 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1,sd2] -> do - sd2' <- transferEntryDP' sd1 sd2 - - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([sd2'],Just sd1') - - replaceDecls (anchorEof lp) [de1',sd1] - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + doRmDecl = replaceDecls (anchorEof lp) [de1',sd1] + where + [de1] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a,sd2] -> + let + sd2' = transferEntryDP' sd1a sd2 + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([sd2'],Just sd1') + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -805,10 +793,8 @@ rmDecl5 _libdir lp = do go :: HsExpr GhcPs -> Transform (HsExpr GhcPs) go (HsLet (tkLet, tkIn) lb expr) = do let decs = hsDeclsLocalBinds lb - let hdecs : _ = decs let dec = last decs - _ <- transferEntryDP hdecs dec - lb' <- replaceDeclsValbinds WithoutWhere lb [dec] + let lb' = replaceDeclsValbinds WithoutWhere lb [dec] return (HsLet (tkLet, tkIn) lb' expr) go x = return x @@ -823,73 +809,61 @@ rmDecl5 _libdir lp = do rmDecl6 :: Changer rmDecl6 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m subDecs -> do - let subDecs' = captureLineSpacing subDecs - let (ss1:_sd1:sd2:sds) = subDecs' - sd2' <- transferEntryDP' ss1 sd2 - - return (sd2':sds,Nothing) + doRmDecl = replaceDecls lp [de1'] + where + [de1] = hsDecls lp - replaceDecls lp [de1'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m subDecs -> + let + subDecs' = captureLineSpacing subDecs + (ss1:_sd1:sd2:sds) = subDecs' + sd2' = transferEntryDP' ss1 sd2 + in (sd2':sds,Nothing) - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmDecl7 :: Changer -rmDecl7 _libdir top = do +rmDecl7 _libdir lp = do let - doRmDecl = do - let lp = top - tlDecs <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList tlDecs - - d3' <- transferEntryDP' d2 d3 - - replaceDecls lp [s1,de1,d3'] + doRmDecl = replaceDecls lp [s1,de1,d3'] + where + tlDecs = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList tlDecs + d3' = transferEntryDP' d2 d3 - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig1 :: Changer rmTypeSig1 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let (s0:de1:d2) = tlDecs - s1 = captureTypeSigSpacing s0 - (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 - L ln n2' <- transferEntryDP n1 n2 - let s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) - replaceDecls lp (s1':de1:d2) - - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let doRmDecl = replaceDecls lp (s1':de1:d2) + where + tlDecs = hsDecls lp + (s0:de1:d2) = tlDecs + s1 = captureTypeSigSpacing s0 + (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 + L ln n2' = transferEntryDP n1 n2 + s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig2 :: Changer rmTypeSig2 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let [de1] = tlDecs - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m [s,d] -> do - d' <- transferEntryDP' s d - return $ ([d'],Nothing) - `debug` ("rmTypeSig2:(d,d')" ++ showAst (d,d')) - replaceDecls lp [de1'] + let doRmDecl = replaceDecls lp [de1'] + where + tlDecs = hsDecls lp + [de1] = tlDecs + (de1',_) = modifyValD (getLocA de1) de1 $ \_m [_s,d] -> ([d],Nothing) - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -958,13 +932,15 @@ addClassMethod libdir lp = do let decl' = setEntryDP decl (DifferentLine 1 3) let sig' = setEntryDP sig (DifferentLine 2 3) let doAddMethod = do - [cd] <- hsDecls lp - (f1:f2s:f2d:_) <- hsDecls cd - let f2s' = setEntryDP f2s (DifferentLine 2 3) - cd' <- replaceDecls cd [f1, sig', decl', f2s', f2d] - replaceDecls lp [cd'] - - (lp',_,w) <- runTransformT doAddMethod + let + [cd] = hsDecls lp + (f1:f2s:f2d:_) = hsDecls cd + f2s' = setEntryDP f2s (DifferentLine 2 3) + cd' = replaceDecls cd [f1, sig', decl', f2s', f2d] + lp' = replaceDecls lp [cd'] + return lp' + + let (lp',_,w) = runTransform doAddMethod debugM $ "addClassMethod:" ++ intercalate "\n" w return lp' ===================================== utils/check-exact/Parsers.hs ===================================== @@ -260,7 +260,7 @@ parseModuleEpAnnsWithCppInternal cppOptions dflags file = do GHC.PFailed pst -> Left (GHC.GhcPsMessage <$> GHC.getPsErrorMessages pst) GHC.POk _ pmod - -> Right $ (injectedComments, dflags', fixModuleTrailingComments pmod) + -> Right $ (injectedComments, dflags', fixModuleComments pmod) -- | Internal function. Exposed if you want to muck with DynFlags -- before parsing. Or after parsing. @@ -269,8 +269,10 @@ postParseTransform -> Either a (GHC.ParsedSource) postParseTransform parseRes = fmap mkAnns parseRes where - -- TODO:AZ perhaps inject the comments into the parsedsource here already - mkAnns (_cs, _, m) = fixModuleTrailingComments m + mkAnns (_cs, _, m) = fixModuleComments m + +fixModuleComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleComments p = fixModuleHeaderComments $ fixModuleTrailingComments p fixModuleTrailingComments :: GHC.ParsedSource -> GHC.ParsedSource fixModuleTrailingComments (GHC.L l p) = GHC.L l p' @@ -293,6 +295,47 @@ fixModuleTrailingComments (GHC.L l p) = GHC.L l p' in cs'' _ -> cs +-- Deal with https://gitlab.haskell.org/ghc/ghc/-/issues/23984 +-- The Lexer works bottom-up, so does not have module declaration info +-- when the first top decl processed +fixModuleHeaderComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleHeaderComments (GHC.L l p) = GHC.L l p' + where + moveComments :: GHC.EpaLocation -> GHC.LHsDecl GHC.GhcPs -> GHC.EpAnnComments + -> (GHC.LHsDecl GHC.GhcPs, GHC.EpAnnComments) + moveComments GHC.EpaDelta{} dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.UnhelpfulSpan _)) dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.RealSrcSpan r _)) (GHC.L (GHC.EpAnn anc an csd) a) cs = (dd,css) + where + -- Move any comments on the decl that occur prior to the location + pc = GHC.priorComments csd + fc = GHC.getFollowingComments csd + bf (GHC.L anch _) = GHC.anchor anch > r + (move,keep) = break bf pc + csd' = GHC.EpaCommentsBalanced keep fc + + dd = GHC.L (GHC.EpAnn anc an csd') a + css = cs <> GHC.EpaComments move + + (ds',an') = rebalance (GHC.hsmodDecls p, GHC.hsmodAnn $ GHC.hsmodExt p) + p' = p { GHC.hsmodExt = (GHC.hsmodExt p){ GHC.hsmodAnn = an' }, + GHC.hsmodDecls = ds' + } + + rebalance :: ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + -> ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + rebalance (ds, GHC.EpAnn a an cs) = (ds1, GHC.EpAnn a an cs') + where + (ds1,cs') = case break (\(GHC.AddEpAnn k _) -> k == GHC.AnnWhere) (GHC.am_main an) of + (_, (GHC.AddEpAnn _ whereLoc:_)) -> + case GHC.hsmodDecls p of + (d:ds0) -> (d':ds0, cs0) + where (d',cs0) = moveComments whereLoc d cs + ds0 -> (ds0,cs) + _ -> (ds,cs) + + + -- | Internal function. Initializes DynFlags value for parsing. -- -- Passes "-hide-all-packages" to the GHC API to prevent parsing of ===================================== utils/check-exact/Transform.hs ===================================== @@ -63,7 +63,7 @@ module Transform -- *** Low level operations used in 'HasDecls' , balanceComments , balanceCommentsList - , balanceCommentsList' + , balanceCommentsListA , anchorEof -- ** Managing lists, pure functions @@ -92,6 +92,7 @@ import qualified Control.Monad.Fail as Fail import GHC hiding (parseModule, parsedSource) import GHC.Data.FastString +import GHC.Types.SrcLoc import Data.Data import Data.Maybe @@ -154,6 +155,7 @@ logDataWithAnnsTr str ast = do -- |If we need to add new elements to the AST, they need their own -- 'SrcSpan' for this. +-- This should no longer be needed, we use an @EpaDelta@ location instead. uniqueSrcSpanT :: (Monad m) => TransformT m SrcSpan uniqueSrcSpanT = do col <- get @@ -171,15 +173,6 @@ srcSpanStartLine' _ = 0 -- --------------------------------------------------------------------- -captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag -captureOrderBinds ls = AnnSortKey $ map go ls - where - go (L _ (ValD _ _)) = BindTag - go (L _ (SigD _ _)) = SigDTag - go d = error $ "captureOrderBinds:" ++ showGhc d - --- --------------------------------------------------------------------- - captureMatchLineSpacing :: LHsDecl GhcPs -> LHsDecl GhcPs captureMatchLineSpacing (L l (ValD x (FunBind a b (MG c (L d ms ))))) = L l (ValD x (FunBind a b (MG c (L d ms')))) @@ -253,7 +246,7 @@ setEntryDPDecl d dp = setEntryDP d dp -- |Set the true entry 'DeltaPos' from the annotation for a given AST -- element. This is the 'DeltaPos' ignoring any comments. -setEntryDP :: NoAnn t => LocatedAn t a -> DeltaPos -> LocatedAn t a +setEntryDP :: LocatedAn t a -> DeltaPos -> LocatedAn t a setEntryDP (L (EpAnn (EpaSpan ss@(UnhelpfulSpan _)) an cs) a) dp = L (EpAnn (EpaDelta ss dp []) an cs) a setEntryDP (L (EpAnn (EpaSpan ss) an (EpaComments [])) a) dp @@ -293,7 +286,7 @@ setEntryDP (L (EpAnn (EpaSpan ss@(RealSrcSpan r _)) an cs) a) dp L (EpAnn (EpaDelta ss edp csd) an cs'') a where cs'' = setPriorComments cs [] - csd = L (EpaDelta ss dp NoComments) c:cs' + csd = L (EpaDelta ss dp NoComments) c:commentOrigDeltas cs' lc = last $ (L ca c:cs') delta = case getLoc lc of EpaSpan (RealSrcSpan rr _) -> ss2delta (ss2pos rr) r @@ -335,18 +328,15 @@ setEntryDPFromAnchor off (EpaSpan (RealSrcSpan anc _)) ll@(L la _) = setEntryDP -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -transferEntryDP :: (Monad m, NoAnn t2, Typeable t1, Typeable t2) - => LocatedAn t1 a -> LocatedAn t2 b -> TransformT m (LocatedAn t2 b) -transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = do - logTr $ "transferEntryDP': EpAnn,EpAnn" +transferEntryDP :: (Typeable t1, Typeable t2) + => LocatedAn t1 a -> LocatedAn t2 b -> (LocatedAn t2 b) +transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = -- Problem: if the original had preceding comments, blindly -- transferring the location is not correct case priorComments cs1 of - [] -> return (L (EpAnn anc1 (combine an1 an2) cs2) b) + [] -> (L (EpAnn anc1 (combine an1 an2) cs2) b) -- TODO: what happens if the receiving side already has comments? - (L anc _:_) -> do - logDataWithAnnsTr "transferEntryDP':priorComments anc=" anc - return (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) + (L _ _:_) -> (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) -- |If a and b are the same type return first arg, else return second @@ -356,10 +346,11 @@ combine x y = fromMaybe y (cast x) -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -- TODO: call transferEntryDP, and use pushDeclDP -transferEntryDP' :: (Monad m) => LHsDecl GhcPs -> LHsDecl GhcPs -> TransformT m (LHsDecl GhcPs) -transferEntryDP' la lb = do - (L l2 b) <- transferEntryDP la lb - return (L l2 (pushDeclDP b (SameLine 0))) +transferEntryDP' :: LHsDecl GhcPs -> LHsDecl GhcPs -> (LHsDecl GhcPs) +transferEntryDP' la lb = + let + (L l2 b) = transferEntryDP la lb + in (L l2 (pushDeclDP b (SameLine 0))) pushDeclDP :: HsDecl GhcPs -> DeltaPos -> HsDecl GhcPs @@ -375,13 +366,24 @@ pushDeclDP d _dp = d -- --------------------------------------------------------------------- -balanceCommentsList :: (Monad m) => [LHsDecl GhcPs] -> TransformT m [LHsDecl GhcPs] -balanceCommentsList [] = return [] -balanceCommentsList [x] = return [x] -balanceCommentsList (a:b:ls) = do - (a',b') <- balanceComments a b - r <- balanceCommentsList (b':ls) - return (a':r) +-- | If we compile in haddock mode, the haddock processing inserts +-- DocDecls to carry the Haddock Documentation. We ignore these in +-- exact printing, as all the comments are also available in their +-- normal location, and the haddock processing is lossy, in that it +-- does not preserve all haddock-like comments. When we balance +-- comments in a list, we migrate some to preceding or following +-- declarations in the list. We must make sure we do not move any to +-- these DocDecls, which are not printed. +balanceCommentsList :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList decls = balanceCommentsList' (filter notDocDecl decls) + +balanceCommentsList' :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList' [] = [] +balanceCommentsList' [x] = [x] +balanceCommentsList' (a:b:ls) = (a':r) + where + (a',b') = balanceComments a b + r = balanceCommentsList' (b':ls) -- |The GHC parser puts all comments appearing between the end of one AST -- item and the beginning of the next as 'annPriorComments' for the second one. @@ -389,28 +391,27 @@ balanceCommentsList (a:b:ls) = do -- from the second one to the 'annFollowingComments' of the first if they belong -- to it instead. This is typically required before deleting or duplicating -- either of the AST elements. -balanceComments :: (Monad m) - => LHsDecl GhcPs -> LHsDecl GhcPs - -> TransformT m (LHsDecl GhcPs, LHsDecl GhcPs) -balanceComments first second = do +balanceComments :: LHsDecl GhcPs -> LHsDecl GhcPs + -> (LHsDecl GhcPs, LHsDecl GhcPs) +balanceComments first second = case first of - (L l (ValD x fb@(FunBind{}))) -> do - (L l' fb',second') <- balanceCommentsFB (L l fb) second - return (L l' (ValD x fb'), second') - _ -> balanceComments' first second + (L l (ValD x fb@(FunBind{}))) -> + let + (L l' fb',second') = balanceCommentsFB (L l fb) second + in (L l' (ValD x fb'), second') + _ -> balanceCommentsA first second --- |Once 'balanceComments' has been called to move trailing comments to a +-- |Once 'balanceCommentsA has been called to move trailing comments to a -- 'FunBind', these need to be pushed down from the top level to the last -- 'Match' if that 'Match' needs to be manipulated. -balanceCommentsFB :: (Monad m) - => LHsBind GhcPs -> LocatedA b -> TransformT m (LHsBind GhcPs, LocatedA b) -balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do - debugM $ "balanceCommentsFB entered: " ++ showGhc (ss2range $ locA lf) +balanceCommentsFB :: LHsBind GhcPs -> LocatedA b -> (LHsBind GhcPs, LocatedA b) +balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second + = balanceCommentsA (packFunBind bind) second' -- There are comments on lf. We need to -- + Keep the prior ones here -- + move the interior ones to the first match, -- + move the trailing ones to the last match. - let + where (before,middle,after) = case entry lf of EpaSpan (RealSrcSpan ss _) -> let @@ -426,40 +427,29 @@ balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do getFollowingComments $ comments lf) lf' = setCommentsEpAnn lf (EpaComments before) - debugM $ "balanceCommentsFB (before, after): " ++ showAst (before, after) - debugM $ "balanceCommentsFB lf': " ++ showAst lf' - -- let matches' = case matches of - let matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] - matches' = case matches of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaComments middle )) m':ms') - _ -> error "balanceCommentsFB" - matches'' <- balanceCommentsList' matches' - let (m,ms) = case reverse matches'' of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m',ms') - -- (L (addCommentsToEpAnnS lm' (EpaCommentsBalanced [] after)) m',ms') - _ -> error "balanceCommentsFB4" - debugM $ "balanceCommentsFB: (m,ms):" ++ showAst (m,ms) - (m',second') <- balanceComments' m second - m'' <- balanceCommentsMatch m' - let (m''',lf'') = case ms of - [] -> moveLeadingComments m'' lf' - _ -> (m'',lf') - debugM $ "balanceCommentsFB: (lf'', m'''):" ++ showAst (lf'',m''') - debugM $ "balanceCommentsFB done" - let bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) - debugM $ "balanceCommentsFB returning:" ++ showAst bind - balanceComments' (packFunBind bind) second' -balanceCommentsFB f s = balanceComments' f s + matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] + matches' = case matches of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaComments middle )) m0:ms') + _ -> error "balanceCommentsFB" + matches'' = balanceCommentsListA matches' + (m,ms) = case reverse matches'' of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m0,ms') + _ -> error "balanceCommentsFB4" + (m',second') = balanceCommentsA m second + m'' = balanceCommentsMatch m' + (m''',lf'') = case ms of + [] -> moveLeadingComments m'' lf' + _ -> (m'',lf') + bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) +balanceCommentsFB f s = balanceCommentsA f s -- | Move comments on the same line as the end of the match into the -- GRHS, prior to the binds -balanceCommentsMatch :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do - logTr $ "balanceCommentsMatch: (logInfo)=" ++ showAst (logInfo) - return (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) +balanceCommentsMatch :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) + = (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) where simpleBreak (r,_) = r /= 0 an1 = l @@ -468,7 +458,7 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do (move',stay') = break simpleBreak (trailingCommentsDeltas (anchorFromLocatedA (L l ())) cs1f) move = map snd move' stay = map snd stay' - (l'', grhss', binds', logInfo) + (l'', grhss', binds', _logInfo) = case reverse grhss of [] -> (l, [], binds, (EpaComments [], noSrcSpanA)) (L lg (GRHS ag grs rhs):gs) -> @@ -491,26 +481,24 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do pushTrailingComments :: WithWhere -> EpAnnComments -> HsLocalBinds GhcPs -> (Bool, HsLocalBinds GhcPs) pushTrailingComments _ _cs b at EmptyLocalBinds{} = (False, b) pushTrailingComments _ _cs (HsIPBinds _ _) = error "TODO: pushTrailingComments:HsIPBinds" -pushTrailingComments w cs lb@(HsValBinds an _) - = (True, HsValBinds an' vb) +pushTrailingComments w cs lb@(HsValBinds an _) = (True, HsValBinds an' vb) where decls = hsDeclsLocalBinds lb (an', decls') = case reverse decls of [] -> (addCommentsToEpAnn an cs, decls) (L la d:ds) -> (an, L (addCommentsToEpAnn la cs) d:ds) - (vb,_ws2) = case runTransform (replaceDeclsValbinds w lb (reverse decls')) of - ((HsValBinds _ vb'), _, ws2') -> (vb', ws2') - _ -> (ValBinds NoAnnSortKey [] [], []) + vb = case replaceDeclsValbinds w lb (reverse decls') of + (HsValBinds _ vb') -> vb' + _ -> ValBinds NoAnnSortKey [] [] -balanceCommentsList' :: (Monad m) => [LocatedA a] -> TransformT m [LocatedA a] -balanceCommentsList' [] = return [] -balanceCommentsList' [x] = return [x] -balanceCommentsList' (a:b:ls) = do - logTr $ "balanceCommentsList' entered" - (a',b') <- balanceComments' a b - r <- balanceCommentsList' (b':ls) - return (a':r) +balanceCommentsListA :: [LocatedA a] -> [LocatedA a] +balanceCommentsListA [] = [] +balanceCommentsListA [x] = [x] +balanceCommentsListA (a:b:ls) = (a':r) + where + (a',b') = balanceCommentsA a b + r = balanceCommentsListA (b':ls) -- |Prior to moving an AST element, make sure any trailing comments belonging to -- it are attached to it, and not the following element. Of necessity this is a @@ -518,13 +506,8 @@ balanceCommentsList' (a:b:ls) = do -- with a passed-in decision function. -- The initial situation is that all comments for a given anchor appear as prior comments -- Many of these should in fact be following comments for the previous anchor -balanceComments' :: (Monad m) => LocatedA a -> LocatedA b -> TransformT m (LocatedA a, LocatedA b) -balanceComments' la1 la2 = do - debugM $ "balanceComments': (anc1)=" ++ showAst (anc1) - debugM $ "balanceComments': (cs1s)=" ++ showAst (cs1s) - debugM $ "balanceComments': (cs1stay,cs1move)=" ++ showAst (cs1stay,cs1move) - debugM $ "balanceComments': (an1',an2')=" ++ showAst (an1',an2') - return (la1', la2') +balanceCommentsA :: LocatedA a -> LocatedA b -> (LocatedA a, LocatedA b) +balanceCommentsA la1 la2 = (la1', la2') where simpleBreak n (r,_) = r > n L an1 f = la1 @@ -532,26 +515,31 @@ balanceComments' la1 la2 = do anc1 = comments an1 anc2 = comments an2 - cs1s = splitCommentsEnd (anchorFromLocatedA la1) anc1 - cs1p = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1s) - cs1f = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1s) + (p1,m1,f1) = splitComments (anchorFromLocatedA la1) anc1 + cs1p = priorCommentsDeltas (anchorFromLocatedA la1) p1 - cs2s = splitCommentsEnd (anchorFromLocatedA la2) anc2 - cs2p = priorCommentsDeltas (anchorFromLocatedA la2) (priorComments cs2s) - cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) (getFollowingComments cs2s) + -- Split cs1 following comments into those before any + -- TrailingAnn's on an1, and any after + cs1f = splitCommentsEnd (fullSpanFromLocatedA la1) $ EpaComments f1 + cs1fp = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1f) + cs1ff = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1f) - -- Split cs1f into those that belong on an1 and ones that must move to an2 - (cs1move,cs1stay) = break (simpleBreak 1) cs1f + -- Split cs1ff into those that belong on an1 and ones that must move to an2 + (cs1move,cs1stay) = break (simpleBreak 1) cs1ff + + (p2,m2,f2) = splitComments (anchorFromLocatedA la2) anc2 + cs2p = priorCommentsDeltas (anchorFromLocatedA la2) p2 + cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) f2 (stay'',move') = break (simpleBreak 1) cs2p -- Need to also check for comments more closely attached to la1, -- ie trailing on the same line (move'',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchorFromLocatedA la1) (map snd stay'')) - move = sortEpaComments $ map snd (cs1move ++ move'' ++ move') - stay = sortEpaComments $ map snd (cs1stay ++ stay') + move = sortEpaComments $ map snd (cs1fp ++ cs1move ++ move'' ++ move') + stay = sortEpaComments $ m2 ++ map snd (cs1stay ++ stay') - an1' = setCommentsEpAnn (getLoc la1) (EpaCommentsBalanced (map snd cs1p) move) - an2' = setCommentsEpAnn (getLoc la2) (EpaCommentsBalanced stay (map snd cs2f)) + an1' = setCommentsEpAnn (getLoc la1) (epaCommentsBalanced (m1 ++ map snd cs1p) move) + an2' = setCommentsEpAnn (getLoc la2) (epaCommentsBalanced stay (map snd cs2f)) la1' = L an1' f la2' = L an2' s @@ -569,10 +557,9 @@ trailingCommentsDeltas r (la@(L l _):las) (al,_) = ss2posEnd rs' (ll,_) = ss2pos (anchor loc) --- AZ:TODO: this is identical to commentsDeltas priorCommentsDeltas :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] -priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) +priorCommentsDeltas r cs = go r (sortEpaComments cs) where go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] go _ [] = [] @@ -588,6 +575,21 @@ priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) -- --------------------------------------------------------------------- +-- | Split comments into ones occurring before the end of the reference +-- span, and those after it. +splitComments :: RealSrcSpan -> EpAnnComments -> ([LEpaComment], [LEpaComment], [LEpaComment]) +splitComments p cs = (before, middle, after) + where + cmpe (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p + cmpe (L _ _) = True + + cmpb (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2pos p + cmpb (L _ _) = True + + (beforeEnd, after) = break cmpe ((priorComments cs) ++ (getFollowingComments cs)) + (before, middle) = break cmpb beforeEnd + + -- | Split comments into ones occurring before the end of the reference -- span, and those after it. splitCommentsEnd :: RealSrcSpan -> EpAnnComments -> EpAnnComments @@ -598,8 +600,8 @@ splitCommentsEnd p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsEnd p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsEnd p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -617,8 +619,8 @@ splitCommentsStart p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsStart p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsStart p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -638,8 +640,8 @@ moveLeadingComments (L la a) lb = (L la' a, lb') -- TODO: need to set an entry delta on lb' to zero, and move the -- original spacing to the first comment. - la' = setCommentsEpAnn la (EpaCommentsBalanced [] after) - lb' = addCommentsToEpAnn lb (EpaCommentsBalanced before []) + la' = setCommentsEpAnn la (epaCommentsBalanced [] after) + lb' = addCommentsToEpAnn lb (epaCommentsBalanced before []) -- | A GHC comment includes the span of the preceding (non-comment) -- token. Takes an original list of comments, and converts the @@ -662,17 +664,27 @@ addCommentOrigDeltasAnn (EpAnn e a cs) = EpAnn e a (addCommentOrigDeltas cs) anchorFromLocatedA :: LocatedA a -> RealSrcSpan anchorFromLocatedA (L (EpAnn anc _ _) _) = anchor anc +-- | Get the full span of interest for comments from a LocatedA. +-- This extends up to the last TrailingAnn +fullSpanFromLocatedA :: LocatedA a -> RealSrcSpan +fullSpanFromLocatedA (L (EpAnn anc (AnnListItem tas) _) _) = rr + where + r = anchor anc + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + rr = case reverse (concatMap trailing_loc tas) of + [] -> r + (s:_) -> combineRealSrcSpans r s + -- --------------------------------------------------------------------- -balanceSameLineComments :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do - logTr $ "balanceSameLineComments: (la)=" ++ showGhc (ss2range $ locA la) - logTr $ "balanceSameLineComments: [logInfo]=" ++ showAst logInfo - return (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) +balanceSameLineComments :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) + = (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) where simpleBreak n (r,_) = r > n - (la',grhss', logInfo) = case reverse grhss of + (la',grhss', _logInfo) = case reverse grhss of [] -> (la,grhss,[]) (L lg (GRHS ga gs rhs):grs) -> (la'',reverse $ (L lg (GRHS ga' gs rhs)):grs,[(gac,(csp,csf))]) where @@ -684,7 +696,7 @@ balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do (move',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchor anc) csf) move = map snd move' stay = map snd stay' - cs1 = EpaCommentsBalanced csp stay + cs1 = epaCommentsBalanced csp stay gac = epAnnComments ga gfc = getFollowingComments gac @@ -734,24 +746,21 @@ addComma (EpAnn anc (AnnListItem as) cs) -- | Insert a declaration into an AST element having sub-declarations -- (@HasDecls@) according to the given location function. insertAt :: (HasDecls ast) - => (LHsDecl GhcPs - -> [LHsDecl GhcPs] - -> [LHsDecl GhcPs]) - -> ast - -> LHsDecl GhcPs - -> Transform ast -insertAt f t decl = do - oldDecls <- hsDecls t - oldDeclsb <- balanceCommentsList oldDecls - let oldDecls' = oldDeclsb - replaceDecls t (f decl oldDecls') + => (LHsDecl GhcPs + -> [LHsDecl GhcPs] + -> [LHsDecl GhcPs]) + -> ast + -> LHsDecl GhcPs + -> ast +insertAt f t decl = replaceDecls t (f decl oldDecls') + where + oldDecls = hsDecls t + oldDeclsb = balanceCommentsList oldDecls + oldDecls' = oldDeclsb -- |Insert a declaration at the beginning or end of the subdecls of the given -- AST item -insertAtStart, insertAtEnd :: (HasDecls ast) - => ast - -> LHsDecl GhcPs - -> Transform ast +insertAtStart, insertAtEnd :: HasDecls ast => ast -> LHsDecl GhcPs -> ast insertAtEnd = insertAt (\x xs -> xs ++ [x]) @@ -766,11 +775,11 @@ insertAtStart = insertAt insertFirst -- |Insert a declaration at a specific location in the subdecls of the given -- AST item -insertAfter, insertBefore :: (HasDecls (LocatedA ast)) +insertAfter, insertBefore :: HasDecls (LocatedA ast) => LocatedA old -> LocatedA ast -> LHsDecl GhcPs - -> Transform (LocatedA ast) + -> LocatedA ast insertAfter (getLocA -> k) = insertAt findAfter where findAfter x xs = @@ -797,10 +806,10 @@ class (Data t) => HasDecls t where -- given syntax phrase. They are always returned in the wrapped 'HsDecl' -- form, even if orginating in local decls. This is safe, as annotations -- never attach to the wrapper, only to the wrapped item. - hsDecls :: (Monad m) => t -> TransformT m [LHsDecl GhcPs] + hsDecls :: t -> [LHsDecl GhcPs] -- | Replace the directly enclosed decl list by the given - -- decl list. Runs in the 'Transform' monad to be able to update list order + -- decl list. As part of replacing it will update list order -- annotations, and rebalance comments and other layout changes as needed. -- -- For example, a call on replaceDecls for a wrapped 'FunBind' having no @@ -818,96 +827,86 @@ class (Data t) => HasDecls t where -- where -- nn = 2 -- @ - replaceDecls :: (Monad m) => t -> [LHsDecl GhcPs] -> TransformT m t + replaceDecls :: t -> [LHsDecl GhcPs] -> t -- --------------------------------------------------------------------- instance HasDecls ParsedSource where - hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = return decls + hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = decls replaceDecls (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps _decls)) decls - = do - logTr "replaceDecls LHsModule" - return (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) + = (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsDecl GhcPs)) where - hsDecls (L _ (TyClD _ c at ClassDecl{})) = return $ hsDeclsClassDecl c - hsDecls decl = do - error $ "hsDecls:decl=" ++ showAst decl - replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = do - let decl' = replaceDeclsClassDecl dec decls - return (L l (TyClD e decl')) - replaceDecls decl _decls = do - error $ "replaceDecls:decl=" ++ showAst decl + hsDecls (L _ (TyClD _ c at ClassDecl{})) = hsDeclsClassDecl c + hsDecls decl = error $ "hsDecls:decl=" ++ showAst decl + replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = + let + decl' = replaceDeclsClassDecl dec decls + in (L l (TyClD e decl')) + replaceDecls decl _decls + = error $ "replaceDecls:decl=" ++ showAst decl -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Match GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = hsDeclsLocalBinds lb replaceDecls (L l (Match xm c p (GRHSs xr rhs binds))) [] - = do - logTr "replaceDecls LMatch empty decls" - binds'' <- replaceDeclsValbinds WithoutWhere binds [] - return (L l (Match xm c p (GRHSs xr rhs binds''))) + = let + binds'' = replaceDeclsValbinds WithoutWhere binds [] + in (L l (Match xm c p (GRHSs xr rhs binds''))) replaceDecls m@(L l (Match xm c p (GRHSs xr rhs binds))) newBinds - = do - logTr "replaceDecls LMatch nonempty decls" + = let -- Need to throw in a fresh where clause if the binds were empty, -- in the annotations. - (l', rhs') <- case binds of - EmptyLocalBinds{} -> do - logTr $ "replaceDecls LMatch empty binds" - - logDataWithAnnsTr "Match.replaceDecls:balancing comments:m" m - L l' m' <- balanceSameLineComments m - logDataWithAnnsTr "Match.replaceDecls:(m1')" (L l' m') - return (l', grhssGRHSs $ m_grhss m') - _ -> return (l, rhs) - binds'' <- replaceDeclsValbinds WithWhere binds newBinds - logDataWithAnnsTr "Match.replaceDecls:binds'" binds'' - return (L l' (Match xm c p (GRHSs xr rhs' binds''))) + (l', rhs') = case binds of + EmptyLocalBinds{} -> + let + L l0 m' = balanceSameLineComments m + in (l0, grhssGRHSs $ m_grhss m') + _ -> (l, rhs) + binds'' = replaceDeclsValbinds WithWhere binds newBinds + in (L l' (Match xm c p (GRHSs xr rhs' binds''))) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsExpr GhcPs)) where - hsDecls (L _ (HsLet _ decls _ex)) = return $ hsDeclsLocalBinds decls - hsDecls _ = return [] + hsDecls (L _ (HsLet _ decls _ex)) = hsDeclsLocalBinds decls + hsDecls _ = [] replaceDecls (L ll (HsLet (tkLet, tkIn) binds ex)) newDecls - = do - logTr "replaceDecls HsLet" - let lastAnc = realSrcSpan $ spanHsLocaLBinds binds + = let + lastAnc = realSrcSpan $ spanHsLocaLBinds binds -- TODO: may be an intervening comment, take account for lastAnc - let (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of - (EpTok l, EpTok i) -> - let - off = case l of - (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r - (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 - (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 - (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c - ex'' = setEntryDPFromAnchor off i ex - newDecls'' = case newDecls of - [] -> newDecls - (d:ds) -> setEntryDPDecl d (SameLine 0) : ds - in ( EpTok l - , EpTok (addEpaLocationDelta off lastAnc i) - , ex'' - , newDecls'') - (_,_) -> (tkLet, tkIn, ex, newDecls) - binds' <- replaceDeclsValbinds WithoutWhere binds newDecls' - return (L ll (HsLet (tkLet', tkIn') binds' ex')) + (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of + (EpTok l, EpTok i) -> + let + off = case l of + (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r + (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 + (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 + (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c + ex'' = setEntryDPFromAnchor off i ex + newDecls'' = case newDecls of + [] -> newDecls + (d:ds) -> setEntryDPDecl d (SameLine 0) : ds + in ( EpTok l + , EpTok (addEpaLocationDelta off lastAnc i) + , ex'' + , newDecls'') + (_,_) -> (tkLet, tkIn, ex, newDecls) + binds' = replaceDeclsValbinds WithoutWhere binds newDecls' + in (L ll (HsLet (tkLet', tkIn') binds' ex')) -- TODO: does this make sense? Especially as no hsDecls for HsPar replaceDecls (L l (HsPar x e)) newDecls - = do - logTr "replaceDecls HsPar" - e' <- replaceDecls e newDecls - return (L l (HsPar x e')) + = let + e' = replaceDecls e newDecls + in (L l (HsPar x e')) replaceDecls old _new = error $ "replaceDecls (LHsExpr GhcPs) undefined for:" ++ showGhc old -- --------------------------------------------------------------------- @@ -934,53 +933,51 @@ hsDeclsPatBind x = error $ "hsDeclsPatBind called for:" ++ showGhc x -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBindD' \/ 'replaceDeclsPatBindD' is -- idempotent. -replaceDeclsPatBindD :: (Monad m) => LHsDecl GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsDecl GhcPs) -replaceDeclsPatBindD (L l (ValD x d)) newDecls = do - (L _ d') <- replaceDeclsPatBind (L l d) newDecls - return (L l (ValD x d')) +replaceDeclsPatBindD :: LHsDecl GhcPs -> [LHsDecl GhcPs] -> (LHsDecl GhcPs) +replaceDeclsPatBindD (L l (ValD x d)) newDecls = + let + (L _ d') = replaceDeclsPatBind (L l d) newDecls + in (L l (ValD x d')) replaceDeclsPatBindD x _ = error $ "replaceDeclsPatBindD called for:" ++ showGhc x -- | Replace the immediate declarations for a 'PatBind'. This -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBind' \/ 'replaceDeclsPatBind' is -- idempotent. -replaceDeclsPatBind :: (Monad m) => LHsBind GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsBind GhcPs) +replaceDeclsPatBind :: LHsBind GhcPs -> [LHsDecl GhcPs] -> (LHsBind GhcPs) replaceDeclsPatBind (L l (PatBind x a p (GRHSs xr rhss binds))) newDecls - = do - logTr "replaceDecls PatBind" - binds'' <- replaceDeclsValbinds WithWhere binds newDecls - return (L l (PatBind x a p (GRHSs xr rhss binds''))) + = (L l (PatBind x a p (GRHSs xr rhss binds''))) + where + binds'' = replaceDeclsValbinds WithWhere binds newDecls replaceDeclsPatBind x _ = error $ "replaceDeclsPatBind called for:" ++ showGhc x -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Stmt GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (LetStmt _ lb)) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (LetStmt _ lb)) = hsDeclsLocalBinds lb hsDecls (L _ (LastStmt _ e _ _)) = hsDecls e hsDecls (L _ (BindStmt _ _pat e)) = hsDecls e hsDecls (L _ (BodyStmt _ e _ _)) = hsDecls e - hsDecls _ = return [] + hsDecls _ = [] replaceDecls (L l (LetStmt x lb)) newDecls - = do - lb'' <- replaceDeclsValbinds WithWhere lb newDecls - return (L l (LetStmt x lb'')) + = let + lb'' = replaceDeclsValbinds WithWhere lb newDecls + in (L l (LetStmt x lb'')) replaceDecls (L l (LastStmt x e d se)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (LastStmt x e' d se)) + = let + e' = replaceDecls e newDecls + in (L l (LastStmt x e' d se)) replaceDecls (L l (BindStmt x pat e)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BindStmt x pat e')) + = let + e' = replaceDecls e newDecls + in (L l (BindStmt x pat e')) replaceDecls (L l (BodyStmt x e a b)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BodyStmt x e' a b)) - replaceDecls x _newDecls = return x + = let + e' = replaceDecls e newDecls + in (L l (BodyStmt x e' a b)) + replaceDecls x _newDecls = x -- ===================================================================== -- end of HasDecls instances @@ -1062,61 +1059,55 @@ data WithWhere = WithWhere -- care, as this does not manage the declaration order, the -- ordering should be done by the calling function from the 'HsLocalBinds' -- context in the AST. -replaceDeclsValbinds :: (Monad m) - => WithWhere +replaceDeclsValbinds :: WithWhere -> HsLocalBinds GhcPs -> [LHsDecl GhcPs] - -> TransformT m (HsLocalBinds GhcPs) -replaceDeclsValbinds _ _ [] = do - return (EmptyLocalBinds NoExtField) + -> HsLocalBinds GhcPs +replaceDeclsValbinds _ _ [] = EmptyLocalBinds NoExtField replaceDeclsValbinds w b@(HsValBinds a _) new - = do - logTr "replaceDeclsValbinds" - let oldSpan = spanHsLocaLBinds b - an <- oldWhereAnnotation a w (realSrcSpan oldSpan) - let decs = concatMap decl2Bind new - let sigs = concatMap decl2Sig new - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) + = let + oldSpan = spanHsLocaLBinds b + an = oldWhereAnnotation a w (realSrcSpan oldSpan) + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) replaceDeclsValbinds _ (HsIPBinds {}) _new = error "undefined replaceDecls HsIPBinds" replaceDeclsValbinds w (EmptyLocalBinds _) new - = do - logTr "replaceDecls HsLocalBinds" - an <- newWhereAnnotation w - let newBinds = concatMap decl2Bind new - newSigs = concatMap decl2Sig new - let decs = newBinds - let sigs = newSigs - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) - -oldWhereAnnotation :: (Monad m) - => EpAnn AnnList -> WithWhere -> RealSrcSpan -> TransformT m (EpAnn AnnList) -oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = do - -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, change the AnnList anchor to have the correct DP too - let (AnnList ancl o c _r t) = an - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - (anc', ancl') <- do - case ww of - WithWhere -> return (anc, ancl) - WithoutWhere -> return (anc, ancl) - let an' = EpAnn anc' - (AnnList ancl' o c w t) - cs - return an' - -newWhereAnnotation :: (Monad m) => WithWhere -> TransformT m (EpAnn AnnList) -newWhereAnnotation ww = do - let anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] - let anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - let an = EpAnn anc - (AnnList (Just anc2) Nothing Nothing w []) - emptyComments - return an + = let + an = newWhereAnnotation w + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) + +oldWhereAnnotation :: EpAnn AnnList -> WithWhere -> RealSrcSpan -> (EpAnn AnnList) +oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = an' + -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, + -- change the AnnList anchor to have the correct DP too + where + (AnnList ancl o c _r t) = an + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + (anc', ancl') = + case ww of + WithWhere -> (anc, ancl) + WithoutWhere -> (anc, ancl) + an' = EpAnn anc' + (AnnList ancl' o c w t) + cs + +newWhereAnnotation :: WithWhere -> (EpAnn AnnList) +newWhereAnnotation ww = an + where + anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] + anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + an = EpAnn anc + (AnnList (Just anc2) Nothing Nothing w []) + emptyComments -- --------------------------------------------------------------------- @@ -1127,32 +1118,32 @@ type PMatch = LMatch GhcPs (LHsExpr GhcPs) -- declarations are extracted and returned after modification. For a -- 'FunBind' the supplied 'SrcSpan' is used to identify the specific -- 'Match' to be transformed, for when there are multiple of them. -modifyValD :: forall m t. (HasTransform m) - => SrcSpan +modifyValD :: forall t. + SrcSpan -> Decl - -> (PMatch -> [Decl] -> m ([Decl], Maybe t)) - -> m (Decl,Maybe t) + -> (PMatch -> [Decl] -> ([Decl], Maybe t)) + -> (Decl,Maybe t) modifyValD p pb@(L ss (ValD _ (PatBind {} ))) f = if (locA ss) == p - then do - let ds = hsDeclsPatBindD pb - (ds',r) <- f (error "modifyValD.PatBind should not touch Match") ds - pb' <- liftT $ replaceDeclsPatBindD pb ds' - return (pb',r) - else return (pb,Nothing) -modifyValD p decl f = do - (decl',r) <- runStateT (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing - return (packFunDecl decl',r) + then + let + ds = hsDeclsPatBindD pb + (ds',r) = f (error "modifyValD.PatBind should not touch Match") ds + pb' = replaceDeclsPatBindD pb ds' + in (pb',r) + else (pb,Nothing) +modifyValD p decl f = (packFunDecl decl', r) where - doModLocal :: PMatch -> StateT (Maybe t) m PMatch + (decl',r) = runState (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing + doModLocal :: PMatch -> State (Maybe t) PMatch doModLocal (match@(L ss _) :: PMatch) = do if (locA ss) == p then do - ds <- lift $ liftT $ hsDecls match - `debug` ("modifyValD: match=" ++ showAst match) - (ds',r) <- lift $ f match ds - put r - match' <- lift $ liftT $ replaceDecls match ds' + let + ds = hsDecls match + (ds',r0) = f match ds + put r0 + let match' = replaceDecls match ds' return match' else return match @@ -1172,6 +1163,6 @@ modifyDeclsT :: (HasDecls t,HasTransform m) => ([LHsDecl GhcPs] -> m [LHsDecl GhcPs]) -> t -> m t modifyDeclsT action t = do - decls <- liftT $ hsDecls t + let decls = hsDecls t decls' <- action decls - liftT $ replaceDecls t decls' + return $ replaceDecls t decls' ===================================== utils/check-exact/Types.hs ===================================== @@ -21,10 +21,6 @@ type Pos = (Int,Int) -- --------------------------------------------------------------------- -data Rigidity = NormalLayout | RigidLayout deriving (Eq, Ord, Show) - --- --------------------------------------------------------------------- - -- | A Haskell comment. The @AnnKeywordId@ is present if it has been converted -- from an @AnnKeywordId@ because the annotation must be interleaved into the -- stream and does not have a well-defined position ===================================== utils/check-exact/Utils.hs ===================================== @@ -20,9 +20,8 @@ module Utils where import Control.Monad (when) +import GHC.Utils.Monad.State.Strict import Data.Function -import Data.Maybe (isJust) -import Data.Ord (comparing) import GHC.Hs.Dump import Lookup @@ -36,9 +35,10 @@ import GHC.Driver.Ppr import GHC.Data.FastString import qualified GHC.Data.Strict as Strict import GHC.Base (NonEmpty(..)) +import GHC.Parser.Lexer (allocateComments) import Data.Data hiding ( Fixity ) -import Data.List (sortBy, elemIndex) +import Data.List (sortBy, partition) import qualified Data.Map.Strict as Map import Debug.Trace @@ -60,12 +60,32 @@ debug c s = if debugEnabledFlag debugM :: Monad m => String -> m () debugM s = when debugEnabledFlag $ traceM s --- --------------------------------------------------------------------- - warn :: c -> String -> c -- warn = flip trace warn c _ = c +-- --------------------------------------------------------------------- + +captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag +captureOrderBinds ls = AnnSortKey $ map go ls + where + go (L _ (ValD _ _)) = BindTag + go (L _ (SigD _ _)) = SigDTag + go d = error $ "captureOrderBinds:" ++ showGhc d + +-- --------------------------------------------------------------------- + +notDocDecl :: LHsDecl GhcPs -> Bool +notDocDecl (L _ DocD{}) = False +notDocDecl _ = True + +notIEDoc :: LIE GhcPs -> Bool +notIEDoc (L _ IEGroup {}) = False +notIEDoc (L _ IEDoc {}) = False +notIEDoc (L _ IEDocNamed {}) = False +notIEDoc _ = True + +-- --------------------------------------------------------------------- -- | A good delta has no negative values. isGoodDelta :: DeltaPos -> Bool isGoodDelta (SameLine co) = co >= 0 @@ -108,7 +128,6 @@ pos2delta (refl,refc) (l,c) = deltaPos lo co lo = l - refl co = if lo == 0 then c - refc else c - -- else c - 1 -- | Apply the delta to the current position, taking into account the -- current column offset if advancing to a new line @@ -200,23 +219,6 @@ origDelta pos pp = ss2delta (ss2posEnd pp) pos -- --------------------------------------------------------------------- --- |Given a list of items and a list of keys, returns a list of items --- ordered by their position in the list of keys. -orderByKey :: [(DeclTag,a)] -> [DeclTag] -> [(DeclTag,a)] -orderByKey keys order - -- AZ:TODO: if performance becomes a problem, consider a Map of the order - -- SrcSpan to an index, and do a lookup instead of elemIndex. - - -- Items not in the ordering are placed to the start - = sortBy (comparing (flip elemIndex order . fst)) keys - --- --------------------------------------------------------------------- - -isListComp :: HsDoFlavour -> Bool -isListComp = isDoComprehensionContext - --- --------------------------------------------------------------------- - needsWhere :: DataDefnCons (LConDecl (GhcPass p)) -> Bool needsWhere (NewTypeCon _) = True needsWhere (DataTypeCons _ []) = True @@ -225,21 +227,214 @@ needsWhere _ = False -- --------------------------------------------------------------------- +-- | Insert the comments at the appropriate places in the AST insertCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource -insertCppComments (L l p) cs = L l p' +-- insertCppComments p [] = p +insertCppComments (L l p) cs0 = insertRemainingCppComments (L l p2) remaining + where + (EpAnn anct ant cst) = hsmodAnn $ hsmodExt p + cs = sortEpaComments $ priorComments cst ++ getFollowingComments cst ++ cs0 + p0 = p { hsmodExt = (hsmodExt p) { hsmodAnn = EpAnn anct ant emptyComments }} + -- Comments embedded within spans + -- everywhereM is a bottom-up traversal + (p1, toplevel) = runState (everywhereM (mkM addCommentsListItem + `extM` addCommentsGrhs + `extM` addCommentsList) p0) cs + (p2, remaining) = insertTopLevelCppComments p1 toplevel + + addCommentsListItem :: EpAnn AnnListItem -> State [LEpaComment] (EpAnn AnnListItem) + addCommentsListItem = addComments + + addCommentsList :: EpAnn AnnList -> State [LEpaComment] (EpAnn AnnList) + addCommentsList = addComments + + addCommentsGrhs :: EpAnn GrhsAnn -> State [LEpaComment] (EpAnn GrhsAnn) + addCommentsGrhs = addComments + + addComments :: forall ann. EpAnn ann -> State [LEpaComment] (EpAnn ann) + addComments (EpAnn anc an ocs) = do + case anc of + EpaSpan (RealSrcSpan s _) -> do + unAllocated <- get + let + (rest, these) = GHC.Parser.Lexer.allocateComments s unAllocated + cs' = workInComments ocs these + put rest + return $ EpAnn anc an cs' + + _ -> return $ EpAnn anc an ocs + +workInComments :: EpAnnComments -> [LEpaComment] -> EpAnnComments +workInComments ocs [] = ocs +workInComments ocs new = cs' + where + pc = priorComments ocs + fc = getFollowingComments ocs + cs' = case fc of + [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ new + (L ac _:_) -> epaCommentsBalanced (sortEpaComments $ pc ++ cs_before) + (sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) + new + +insertTopLevelCppComments :: HsModule GhcPs -> [LEpaComment] -> (HsModule GhcPs, [LEpaComment]) +insertTopLevelCppComments (HsModule (XModulePs an lo mdeprec mbDoc) mmn mexports imports decls) cs + = (HsModule (XModulePs an4 lo mdeprec mbDoc) mmn mexports' imports' decls', cs3) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0,hc1,hc_cs)) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0i,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0i,hc0,hc1,hc_cs)) where - an' = case GHC.hsmodAnn $ GHC.hsmodExt p of - (EpAnn a an ocs) -> EpAnn a an cs' - where - pc = priorComments ocs - fc = getFollowingComments ocs - cs' = case fc of - [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ cs - (L ac _:_) -> EpaCommentsBalanced (sortEpaComments $ pc ++ cs_before) - (sortEpaComments $ fc ++ cs_after) - where - (cs_before,cs_after) = break (\(L ll _) -> (ss2pos $ anchor ll) < (ss2pos $ anchor ac) ) cs + -- Comments at the top level. + (an0, cs0) = + case mmn of + Nothing -> (an, cs) + Just _ -> + -- We have a module name. Capture all comments up to the `where` + let + (these, remaining) = splitOnWhere Before (am_main $ anns an) cs + (EpAnn a anno ocs) = an :: EpAnn AnnsModule + anm = EpAnn a anno (workInComments ocs these) + in + (anm, remaining) + (an1,cs0a) = case lo of + EpExplicitBraces (EpTok (EpaSpan (RealSrcSpan s _))) _close -> + let + (stay,cs0a') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0 + cs' = workInComments (comments an0) stay + in (an0 { comments = cs' }, cs0a') + _ -> (an0,cs0) + -- Deal with possible leading semis + (an2, cs0b) = case am_decls $ anns an1 of + (AddSemiAnn (EpaSpan (RealSrcSpan s _)):_) -> (an1 {comments = cs'}, cs0b') + where + (stay,cs0b') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0a + cs' = workInComments (comments an1) stay + _ -> (an1,cs0a) + + (mexports', an3, cs1) = + case mexports of + Nothing -> (Nothing, an2, cs0b) + Just (L l exports) -> (Just (L l exports'), an3', cse) + where + hc1' = workInComments (comments an2) csh' + an3' = an2 { comments = hc1' } + (csh', cs0b') = case al_open $ anns l of + Just (AddEpAnn _ (EpaSpan (RealSrcSpan s _))) ->(h, n) + where + (h,n) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + cs0b + + _ -> ([], cs0b) + (exports', cse) = allocPreceding exports cs0b' + (imports0, cs2) = allocPreceding imports cs1 + (imports', hc0i) = balanceFirstLocatedAComments imports0 + + (decls0, cs3) = allocPreceding decls cs2 + (decls', hc0d) = balanceFirstLocatedAComments decls0 + + -- Either hc0i or hc0d should have comments. Combine them + hc0 = hc0i ++ hc0d + + (hc1,hc_cs) = if null ( am_main $ anns an3) + then (hc0,[]) + else splitOnWhere After (am_main $ anns an3) hc0 + hc2 = workInComments (comments an3) hc1 + an4 = an3 { anns = (anns an3) {am_cs = hc_cs}, comments = hc2 } + + allocPreceding :: [LocatedA a] -> [LEpaComment] -> ([LocatedA a], [LEpaComment]) + allocPreceding [] cs' = ([], cs') + allocPreceding (L (EpAnn anc4 an5 cs4) a:xs) cs' = ((L (EpAnn anc4 an5 cs4') a:xs'), rest') + where + (rest, these) = + case anc4 of + EpaSpan (RealSrcSpan s _) -> + allocatePriorComments (ss2pos s) cs' + _ -> (cs', []) + cs4' = workInComments cs4 these + (xs',rest') = allocPreceding xs rest + +data SplitWhere = Before | After +splitOnWhere :: SplitWhere -> [AddEpAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitOnWhere _ [] csIn = (csIn,[]) +splitOnWhere w (AddEpAnn AnnWhere (EpaSpan (RealSrcSpan s _)):_) csIn = (hc, fc) + where + splitFunc Before anc_pos c_pos = c_pos < anc_pos + splitFunc After anc_pos c_pos = anc_pos < c_pos + (hc,fc) = break (\(L ll _) -> splitFunc w (ss2pos $ anchor ll) (ss2pos s)) csIn +splitOnWhere _ (AddEpAnn AnnWhere _:_) csIn = (csIn, []) +splitOnWhere f (_:as) csIn = splitOnWhere f as csIn + +balanceFirstLocatedAComments :: [LocatedA a] -> ([LocatedA a], [LEpaComment]) +balanceFirstLocatedAComments [] = ([],[]) +balanceFirstLocatedAComments ((L (EpAnn anc an csd) a):ds) = (L (EpAnn anc an csd0) a:ds, hc') + where + (csd0, hc') = case anc of + EpaSpan (RealSrcSpan s _) -> (csd', hc) + `debug` ("balanceFirstLocatedAComments: (csd,csd',attached,header)=" ++ showAst (csd,csd',attached,header)) + where + (priors, inners) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + (priorComments csd) + pcds = priorCommentsDeltas' s priors + (attached, header) = break (\(d,_c) -> d /= 1) pcds + csd' = setPriorComments csd (reverse (map snd attached) ++ inners) + hc = reverse (map snd header) + _ -> (csd, []) + + + +priorCommentsDeltas' :: RealSrcSpan -> [LEpaComment] + -> [(Int, LEpaComment)] +priorCommentsDeltas' r cs = go r (reverse cs) + where + go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] + go _ [] = [] + go _ (la@(L l@(EpaDelta _ dp _) _):las) = (deltaLine dp, la) : go (anchor l) las + go rs' (la@(L l _):las) = deltaComment rs' la : go (anchor l) las + + deltaComment :: RealSrcSpan -> LEpaComment -> (Int, LEpaComment) + deltaComment rs' (L loc c) = (abs(ll - al), L loc c) + where + (al,_) = ss2pos rs' + (ll,_) = ss2pos (anchor loc) + +allocatePriorComments + :: Pos + -> [LEpaComment] + -> ([LEpaComment], [LEpaComment]) +allocatePriorComments ss_loc comment_q = + let + cmp (L l _) = ss2pos (anchor l) <= ss_loc + (newAnns,after) = partition cmp comment_q + in + (after, newAnns) + +insertRemainingCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource +insertRemainingCppComments (L l p) cs = L l p' + -- `debug` ("insertRemainingCppComments: (cs,an')=" ++ showAst (cs,an')) + where + (EpAnn a an ocs) = GHC.hsmodAnn $ GHC.hsmodExt p + an' = EpAnn a an (addTrailingComments end_loc ocs cs) p' = p { GHC.hsmodExt = (GHC.hsmodExt p) { GHC.hsmodAnn = an' } } + end_loc = case GHC.hsmodLayout $ GHC.hsmodExt p of + EpExplicitBraces _open close -> case close of + EpTok (EpaSpan (RealSrcSpan s _)) -> ss2pos s + _ -> (1,1) + _ -> (1,1) + (new_before, new_after) = break (\(L ll _) -> (ss2pos $ anchor ll) > end_loc ) cs + + addTrailingComments end_loc' cur new = epaCommentsBalanced pc' fc' + where + pc = priorComments cur + fc = getFollowingComments cur + (pc', fc') = case reverse pc of + [] -> (sortEpaComments $ pc ++ new_before, sortEpaComments $ fc ++ new_after) + (L ac _:_) -> (sortEpaComments $ pc ++ cs_before, sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = if (ss2pos $ anchor ac) > end_loc' + then break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) new + else (new_before, new_after) -- --------------------------------------------------------------------- @@ -291,11 +486,16 @@ dedentDocChunkBy dedent (L (RealSrcSpan l mb) c) = L (RealSrcSpan l' mb) c dedentDocChunkBy _ x = x + +epaCommentsBalanced :: [LEpaComment] -> [LEpaComment] -> EpAnnComments +epaCommentsBalanced priorCs [] = EpaComments priorCs +epaCommentsBalanced priorCs postCs = EpaCommentsBalanced priorCs postCs + mkEpaComments :: [Comment] -> [Comment] -> EpAnnComments mkEpaComments priorCs [] = EpaComments (map comment2LEpaComment priorCs) mkEpaComments priorCs postCs - = EpaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) + = epaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) comment2LEpaComment :: Comment -> LEpaComment comment2LEpaComment (Comment s anc r _mk) = mkLEpaComment s anc r @@ -330,18 +530,11 @@ sortEpaComments cs = sortBy cmp cs mkKWComment :: AnnKeywordId -> NoCommentsLocation -> Comment mkKWComment kw (EpaSpan (RealSrcSpan ss mb)) = Comment (keywordToString kw) (EpaSpan (RealSrcSpan ss mb)) ss (Just kw) -mkKWComment kw (EpaSpan ss@(UnhelpfulSpan _)) - = Comment (keywordToString kw) (EpaDelta ss (SameLine 0) NoComments) placeholderRealSpan (Just kw) +mkKWComment kw (EpaSpan (UnhelpfulSpan _)) + = Comment (keywordToString kw) (EpaDelta noSrcSpan (SameLine 0) NoComments) placeholderRealSpan (Just kw) mkKWComment kw (EpaDelta ss dp cs) = Comment (keywordToString kw) (EpaDelta ss dp cs) placeholderRealSpan (Just kw) --- | Detects a comment which originates from a specific keyword. -isKWComment :: Comment -> Bool -isKWComment c = isJust (commentOrigin c) - -noKWComments :: [Comment] -> [Comment] -noKWComments = filter (\c -> not (isKWComment c)) - sortAnchorLocated :: [GenLocated EpaLocation a] -> [GenLocated EpaLocation a] sortAnchorLocated = sortBy (compare `on` (anchor . getLoc)) @@ -379,11 +572,6 @@ name2String = showPprUnsafe -- --------------------------------------------------------------------- -locatedAnAnchor :: LocatedAn a t -> RealSrcSpan -locatedAnAnchor (L (EpAnn a _ _) _) = anchor a - --- --------------------------------------------------------------------- - trailingAnnLoc :: TrailingAnn -> EpaLocation trailingAnnLoc (AddSemiAnn ss) = ss trailingAnnLoc (AddCommaAnn ss) = ss @@ -401,46 +589,6 @@ setTrailingAnnLoc (AddDarrowUAnn _) ss = (AddDarrowUAnn ss) addEpAnnLoc :: AddEpAnn -> EpaLocation addEpAnnLoc (AddEpAnn _ l) = l --- --------------------------------------------------------------------- --- Horrible hack for dealing with some things still having a SrcSpan, --- not an Anchor. - -{- -A SrcSpan is defined as - -data SrcSpan = - RealSrcSpan !RealSrcSpan !(Maybe BufSpan) -- See Note [Why Maybe BufPos] - | UnhelpfulSpan !UnhelpfulSpanReason - -data BufSpan = - BufSpan { bufSpanStart, bufSpanEnd :: {-# UNPACK #-} !BufPos } - deriving (Eq, Ord, Show) - -newtype BufPos = BufPos { bufPos :: Int } - - -We use the BufPos to encode a delta, using bufSpanStart for the line, -and bufSpanEnd for the col. - -To be absolutely sure, we make the delta versions use -ve values. - --} - -hackSrcSpanToAnchor :: SrcSpan -> EpaLocation -hackSrcSpanToAnchor (UnhelpfulSpan s) = error $ "hackSrcSpanToAnchor : UnhelpfulSpan:" ++ show s -hackSrcSpanToAnchor ss@(RealSrcSpan r mb) - = case mb of - (Strict.Just (BufSpan (BufPos s) (BufPos e))) -> - if s <= 0 && e <= 0 - then EpaDelta ss (deltaPos (-s) (-e)) [] - `debug` ("hackSrcSpanToAnchor: (r,s,e)=" ++ showAst (r,s,e) ) - else EpaSpan (RealSrcSpan r mb) - _ -> EpaSpan (RealSrcSpan r mb) - -hackAnchorToSrcSpan :: EpaLocation -> SrcSpan -hackAnchorToSrcSpan (EpaSpan s) = s -hackAnchorToSrcSpan _ = error $ "hackAnchorToSrcSpan" - -- --------------------------------------------------------------------- type DeclsByTag a = Map.Map DeclTag [(RealSrcSpan, a)] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/281da00ca3cbbb5de4cd5c5b147aaf88553edaef...3b2e27f3f102a79555a4cfe5395591b3abf648d4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/281da00ca3cbbb5de4cd5c5b147aaf88553edaef...3b2e27f3f102a79555a4cfe5395591b3abf648d4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 19:08:01 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 12 Sep 2024 15:08:01 -0400 Subject: [Git][ghc/ghc][master] 10 commits: RISCV64: Add Native Code Generator (NCG) Message-ID: <66e33c11a39c8_11232339bb74849e1@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - 10 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/aa4500aecca763be5ce27c4784473f00e2ce5aa3...8e6d58cfb0e701fcb0401456bd46c530655b6e8e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/aa4500aecca763be5ce27c4784473f00e2ce5aa3...8e6d58cfb0e701fcb0401456bd46c530655b6e8e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 19:44:32 2024 From: gitlab at gitlab.haskell.org (Bodigrim (@Bodigrim)) Date: Thu, 12 Sep 2024 15:44:32 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/array-0.5.8.0 Message-ID: <66e344a06aebd_1123236a42f895251@gitlab.mail> Bodigrim pushed new branch wip/array-0.5.8.0 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/array-0.5.8.0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 20:07:52 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Thu, 12 Sep 2024 16:07:52 -0400 Subject: [Git][ghc/ghc][wip/T25196] 60 commits: git: remove a.out and include it in .gitignore Message-ID: <66e34a183ab42_14558811a080308f4@gitlab.mail> Sebastian Graf pushed to branch wip/T25196 at Glasgow Haskell Compiler / GHC Commits: f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - 48d064ab by Sebastian Graf at 2024-09-12T20:07:48+00:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 23 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - − a.out - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9b96ccba2446abe56434a909fb4e50e2f22dbd94...48d064ab500b3334ae540e447c8c4e1f11f6bdaf -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9b96ccba2446abe56434a909fb4e50e2f22dbd94...48d064ab500b3334ae540e447c8c4e1f11f6bdaf You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 20:08:42 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Thu, 12 Sep 2024 16:08:42 -0400 Subject: [Git][ghc/ghc][wip/T25196] DmdAnal: Fast path for `multDmdType` (#25196) Message-ID: <66e34a4a48474_145588fd32c3111c@gitlab.mail> Sebastian Graf pushed to branch wip/T25196 at Glasgow Haskell Compiler / GHC Commits: 12a92f12 by Sebastian Graf at 2024-09-12T22:08:28+02:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 5 changed files: - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Types/Demand.hs - + testsuite/tests/dmdanal/should_compile/T25196.hs - + testsuite/tests/dmdanal/should_compile/T25196_aux.hs - testsuite/tests/dmdanal/should_compile/all.T Changes: ===================================== compiler/GHC/Core/Opt/DmdAnal.hs ===================================== @@ -426,7 +426,7 @@ dmdAnalStar env (n :* sd) e , n' <- anticipateANF e n -- See Note [Anticipating ANF in demand analysis] -- and Note [Analysing with absent demand] - = (discardArgDmds $ multDmdType n' dmd_ty, e') + = (multDmdEnv n' (discardArgDmds dmd_ty), e') -- Main Demand Analysis machinery dmdAnal, dmdAnal' :: AnalEnv ===================================== compiler/GHC/Types/Demand.hs ===================================== @@ -51,7 +51,7 @@ module GHC.Types.Demand ( -- * Demand environments DmdEnv(..), addVarDmdEnv, mkTermDmdEnv, nopDmdEnv, plusDmdEnv, plusDmdEnvs, - reuseEnv, + multDmdEnv, reuseEnv, -- * Demand types DmdType(..), dmdTypeDepth, @@ -1910,7 +1910,8 @@ splitDmdTy ty at DmdType{dt_args=dmd:args} = (dmd, ty{dt_args=args}) splitDmdTy ty at DmdType{dt_env=env} = (defaultArgDmd (de_div env), ty) multDmdType :: Card -> DmdType -> DmdType -multDmdType n (DmdType fv args) +multDmdType C_11 dmd_ty = dmd_ty -- a vital optimisation for T25196 +multDmdType n (DmdType fv args) = -- pprTrace "multDmdType" (ppr n $$ ppr fv $$ ppr (multDmdEnv n fv)) $ DmdType (multDmdEnv n fv) (strictMap (multDmd n) args) ===================================== testsuite/tests/dmdanal/should_compile/T25196.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25196 where + +import T25196_aux + +bar = $(gen 10000) ===================================== testsuite/tests/dmdanal/should_compile/T25196_aux.hs ===================================== @@ -0,0 +1,10 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25196_aux where + +import Language.Haskell.TH +import Language.Haskell.TH.Syntax +import Language.Haskell.TH.Lib + +gen :: Int -> Q Exp +gen n = lamE (replicate n (newName "x" >>= varP)) [| () |] ===================================== testsuite/tests/dmdanal/should_compile/all.T ===================================== @@ -98,3 +98,5 @@ test('T22388', [ grep_errmsg(r'^\S+\$w\S+') ], compile, ['-dsuppress-uniques -dd test('T22997', normal, compile, ['']) test('T23398', normal, compile, ['-dsuppress-uniques -ddump-simpl -dno-typeable-binds']) test('T24623', normal, compile, ['']) +test('T25196', [ req_th, collect_compiler_stats('bytes allocated', 10) ], + multimod_compile, ['T25196', '-v0']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/12a92f124820eea3e9835059600b980953d6fc76 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/12a92f124820eea3e9835059600b980953d6fc76 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 12 20:41:16 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 12 Sep 2024 16:41:16 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 2 commits: finder: Add `IsBootInterface` to finder cache keys Message-ID: <66e351ec3be77_1455883b029041137@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: bf856255 by Torsten Schmits at 2024-09-12T16:40:26-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - dda9c763 by Alan Zimmerman at 2024-09-12T16:40:27-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 18 changed files: - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Finder/Types.hs - compiler/GHC/Unit/Module/Env.hs - compiler/GHC/Unit/Types.hs - + testsuite/tests/driver/boot-target/A.hs - + testsuite/tests/driver/boot-target/A.hs-boot - + testsuite/tests/driver/boot-target/B.hs - + testsuite/tests/driver/boot-target/Makefile - + testsuite/tests/driver/boot-target/all.T - utils/check-exact/ExactPrint.hs - utils/check-exact/Main.hs - utils/check-exact/Parsers.hs - utils/check-exact/Transform.hs - utils/check-exact/Types.hs - utils/check-exact/Utils.hs Changes: ===================================== compiler/GHC/Driver/Backpack.hs ===================================== @@ -781,7 +781,7 @@ summariseRequirement pn mod_name = do let loc = srcLocSpan (mkSrcLoc (mkFastString (bkp_filename env)) 1 1) let fc = hsc_FC hsc_env - mod <- liftIO $ addHomeModuleToFinder fc home_unit mod_name location + mod <- liftIO $ addHomeModuleToFinder fc home_unit (notBoot mod_name) location extra_sig_imports <- liftIO $ findExtraSigImports hsc_env HsigFile mod_name @@ -893,7 +893,7 @@ hsModuleToModSummary home_keys pn hsc_src modname this_mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit modname location + addHomeModuleToFinder fc home_unit (GWIB modname (hscSourceToIsBoot hsc_src)) location let ms = ModSummary { ms_mod = this_mod, ms_hsc_src = hsc_src, ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2044,25 +2044,43 @@ summariseFile hsc_env' home_unit old_summaries src_fn mb_phase maybe_buf <- getPreprocessedImports hsc_env src_fn mb_phase maybe_buf let fopts = initFinderOpts (hsc_dflags hsc_env) - - -- Make a ModLocation for this file - let location = mkHomeModLocation fopts pi_mod_name (unsafeEncodeUtf src_fn) + src_path = unsafeEncodeUtf src_fn + + is_boot = case takeExtension src_fn of + ".hs-boot" -> IsBoot + ".lhs-boot" -> IsBoot + _ -> NotBoot + + (path_without_boot, hsc_src) + | isHaskellSigFilename src_fn = (src_path, HsigFile) + | IsBoot <- is_boot = (removeBootSuffix src_path, HsBootFile) + | otherwise = (src_path, HsSrcFile) + + -- Make a ModLocation for the Finder, who only has one entry for + -- each @ModuleName@, and therefore needs to use the locations for + -- the non-boot files. + location_without_boot = + mkHomeModLocation fopts pi_mod_name path_without_boot + + -- Make a ModLocation for this file, adding the @-boot@ suffix to + -- all paths if the original was a boot file. + location + | IsBoot <- is_boot + = addBootSuffixLocn location_without_boot + | otherwise + = location_without_boot -- Tell the Finder cache where it is, so that subsequent calls -- to findModule will find it, even if it's not on any search path mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit pi_mod_name location + addHomeModuleToFinder fc home_unit (GWIB pi_mod_name is_boot) location liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = NotBoot - , nms_hsc_src = - if isHaskellSigFilename src_fn - then HsigFile - else HsSrcFile + , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod , nms_preimps = preimps @@ -2090,9 +2108,10 @@ checkSummaryHash -- Also, only add to finder cache for non-boot modules as the finder cache -- makes sure to add a boot suffix for boot files. _ <- do - let fc = hsc_FC hsc_env + let fc = hsc_FC hsc_env + gwib = GWIB (ms_mod old_summary) (isBootSummary old_summary) case ms_hsc_src old_summary of - HsSrcFile -> addModuleToFinder fc (ms_mod old_summary) location + HsSrcFile -> addModuleToFinder fc gwib location _ -> return () hi_timestamp <- modificationTimeIfExists (ml_hi_file location) @@ -2230,7 +2249,6 @@ summariseModule hsc_env' home_unit old_summary_map is_boot (L _ wanted_mod) mb_p liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = is_boot , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod @@ -2243,7 +2261,6 @@ data MakeNewModSummary = MakeNewModSummary { nms_src_fn :: FilePath , nms_src_hash :: Fingerprint - , nms_is_boot :: IsBootInterface , nms_hsc_src :: HscSource , nms_location :: ModLocation , nms_mod :: Module ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -734,7 +734,7 @@ runHscPhase pipe_env hsc_env0 input_fn src_flavour = do mod <- do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit mod_name location + addHomeModuleToFinder fc home_unit (GWIB mod_name (hscSourceToIsBoot src_flavour)) location -- Make the ModSummary to hand to hscMain let ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -89,23 +89,23 @@ type BaseName = OsPath -- Basename of file initFinderCache :: IO FinderCache initFinderCache = do - mod_cache <- newIORef emptyInstalledModuleEnv + mod_cache <- newIORef emptyInstalledModuleWithIsBootEnv file_cache <- newIORef M.empty let flushFinderCaches :: UnitEnv -> IO () flushFinderCaches ue = do - atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleEnv is_ext fm, ()) + atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleWithIsBootEnv is_ext fm, ()) atomicModifyIORef' file_cache $ \_ -> (M.empty, ()) where - is_ext mod _ = not (isUnitEnvInstalledModule ue mod) + is_ext mod _ = not (isUnitEnvInstalledModule ue (gwib_mod mod)) - addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () addToFinderCache key val = - atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleEnv c key val, ()) + atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleWithIsBootEnv c key val, ()) - lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) lookupFinderCache key = do c <- readIORef mod_cache - return $! lookupInstalledModuleEnv c key + return $! lookupInstalledModuleWithIsBootEnv c key lookupFileCache :: FilePath -> IO Fingerprint lookupFileCache key = do @@ -255,7 +255,7 @@ orIfNotFound this or_this = do homeSearchCache :: FinderCache -> UnitId -> ModuleName -> IO InstalledFindResult -> IO InstalledFindResult homeSearchCache fc home_unit mod_name do_this = do let mod = mkModule home_unit mod_name - modLocationCache fc mod do_this + modLocationCache fc (notBoot mod) do_this findExposedPackageModule :: FinderCache -> FinderOpts -> UnitState -> ModuleName -> PkgQual -> IO FindResult findExposedPackageModule fc fopts units mod_name mb_pkg = @@ -312,7 +312,7 @@ findLookupResult fc fopts r = case r of , fr_unusables = [] , fr_suggestions = suggest' }) -modLocationCache :: FinderCache -> InstalledModule -> IO InstalledFindResult -> IO InstalledFindResult +modLocationCache :: FinderCache -> InstalledModuleWithIsBoot -> IO InstalledFindResult -> IO InstalledFindResult modLocationCache fc mod do_this = do m <- lookupFinderCache fc mod case m of @@ -322,17 +322,17 @@ modLocationCache fc mod do_this = do addToFinderCache fc mod result return result -addModuleToFinder :: FinderCache -> Module -> ModLocation -> IO () +addModuleToFinder :: FinderCache -> ModuleWithIsBoot -> ModLocation -> IO () addModuleToFinder fc mod loc = do - let imod = toUnitId <$> mod - addToFinderCache fc imod (InstalledFound loc imod) + let imod = fmap toUnitId <$> mod + addToFinderCache fc imod (InstalledFound loc (gwib_mod imod)) -- This returns a module because it's more convenient for users -addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleName -> ModLocation -> IO Module +addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleNameWithIsBoot -> ModLocation -> IO Module addHomeModuleToFinder fc home_unit mod_name loc = do - let mod = mkHomeInstalledModule home_unit mod_name - addToFinderCache fc mod (InstalledFound loc mod) - return (mkHomeModule home_unit mod_name) + let mod = mkHomeInstalledModule home_unit <$> mod_name + addToFinderCache fc mod (InstalledFound loc (gwib_mod mod)) + return (mkHomeModule home_unit (gwib_mod mod_name)) -- ----------------------------------------------------------------------------- -- The internal workers @@ -466,7 +466,7 @@ findPackageModule_ :: FinderCache -> FinderOpts -> InstalledModule -> UnitInfo - findPackageModule_ fc fopts mod pkg_conf = do massertPpr (moduleUnit mod == unitId pkg_conf) (ppr (moduleUnit mod) <+> ppr (unitId pkg_conf)) - modLocationCache fc mod $ + modLocationCache fc (notBoot mod) $ -- special case for GHC.Prim; we won't find it in the filesystem. if mod `installedModuleEq` gHC_PRIM ===================================== compiler/GHC/Unit/Finder/Types.hs ===================================== @@ -30,9 +30,9 @@ data FinderCache = FinderCache { flushFinderCaches :: UnitEnv -> IO () -- ^ remove all the home modules from the cache; package modules are -- assumed to not move around during a session; also flush the file hash -- cache. - , addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + , addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () -- ^ Add a found location to the cache for the module. - , lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + , lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) -- ^ Look for a location in the cache. , lookupFileCache :: FilePath -> IO Fingerprint -- ^ Look for the hash of a file in the cache. This should add it to the ===================================== compiler/GHC/Unit/Module/Env.hs ===================================== @@ -33,6 +33,17 @@ module GHC.Unit.Module.Env , mergeInstalledModuleEnv , plusInstalledModuleEnv , installedModuleEnvElts + + -- * InstalledModuleWithIsBootEnv + , InstalledModuleWithIsBootEnv + , emptyInstalledModuleWithIsBootEnv + , lookupInstalledModuleWithIsBootEnv + , extendInstalledModuleWithIsBootEnv + , filterInstalledModuleWithIsBootEnv + , delInstalledModuleWithIsBootEnv + , mergeInstalledModuleWithIsBootEnv + , plusInstalledModuleWithIsBootEnv + , installedModuleWithIsBootEnvElts ) where @@ -283,3 +294,56 @@ plusInstalledModuleEnv :: (elt -> elt -> elt) plusInstalledModuleEnv f (InstalledModuleEnv xm) (InstalledModuleEnv ym) = InstalledModuleEnv $ Map.unionWith f xm ym + + +-------------------------------------------------------------------- +-- InstalledModuleWithIsBootEnv +-------------------------------------------------------------------- + +-- | A map keyed off of 'InstalledModuleWithIsBoot' +newtype InstalledModuleWithIsBootEnv elt = InstalledModuleWithIsBootEnv (Map InstalledModuleWithIsBoot elt) + +instance Outputable elt => Outputable (InstalledModuleWithIsBootEnv elt) where + ppr (InstalledModuleWithIsBootEnv env) = ppr env + + +emptyInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a +emptyInstalledModuleWithIsBootEnv = InstalledModuleWithIsBootEnv Map.empty + +lookupInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> Maybe a +lookupInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = Map.lookup m e + +extendInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> a -> InstalledModuleWithIsBootEnv a +extendInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m x = InstalledModuleWithIsBootEnv (Map.insert m x e) + +filterInstalledModuleWithIsBootEnv :: (InstalledModuleWithIsBoot -> a -> Bool) -> InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBootEnv a +filterInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv e) = + InstalledModuleWithIsBootEnv (Map.filterWithKey f e) + +delInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> InstalledModuleWithIsBootEnv a +delInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = InstalledModuleWithIsBootEnv (Map.delete m e) + +installedModuleWithIsBootEnvElts :: InstalledModuleWithIsBootEnv a -> [(InstalledModuleWithIsBoot, a)] +installedModuleWithIsBootEnvElts (InstalledModuleWithIsBootEnv e) = Map.assocs e + +mergeInstalledModuleWithIsBootEnv + :: (elta -> eltb -> Maybe eltc) + -> (InstalledModuleWithIsBootEnv elta -> InstalledModuleWithIsBootEnv eltc) -- map X + -> (InstalledModuleWithIsBootEnv eltb -> InstalledModuleWithIsBootEnv eltc) -- map Y + -> InstalledModuleWithIsBootEnv elta + -> InstalledModuleWithIsBootEnv eltb + -> InstalledModuleWithIsBootEnv eltc +mergeInstalledModuleWithIsBootEnv f g h (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) + = InstalledModuleWithIsBootEnv $ Map.mergeWithKey + (\_ x y -> (x `f` y)) + (coerce g) + (coerce h) + xm ym + +plusInstalledModuleWithIsBootEnv :: (elt -> elt -> elt) + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt +plusInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) = + InstalledModuleWithIsBootEnv $ Map.unionWith f xm ym + ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -86,6 +86,8 @@ module GHC.Unit.Types , GenWithIsBoot (..) , ModuleNameWithIsBoot , ModuleWithIsBoot + , InstalledModuleWithIsBoot + , notBoot ) where @@ -723,6 +725,8 @@ type ModuleNameWithIsBoot = GenWithIsBoot ModuleName type ModuleWithIsBoot = GenWithIsBoot Module +type InstalledModuleWithIsBoot = GenWithIsBoot InstalledModule + instance Binary a => Binary (GenWithIsBoot a) where put_ bh (GWIB { gwib_mod, gwib_isBoot }) = do put_ bh gwib_mod @@ -736,3 +740,6 @@ instance Outputable a => Outputable (GenWithIsBoot a) where ppr (GWIB { gwib_mod, gwib_isBoot }) = hsep $ ppr gwib_mod : case gwib_isBoot of IsBoot -> [ text "{-# SOURCE #-}" ] NotBoot -> [] + +notBoot :: mod -> GenWithIsBoot mod +notBoot gwib_mod = GWIB {gwib_mod, gwib_isBoot = NotBoot} ===================================== testsuite/tests/driver/boot-target/A.hs ===================================== @@ -0,0 +1,5 @@ +module A where + +import B + +data A = A B ===================================== testsuite/tests/driver/boot-target/A.hs-boot ===================================== @@ -0,0 +1,3 @@ +module A where + +data A ===================================== testsuite/tests/driver/boot-target/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} A + +data B = B A ===================================== testsuite/tests/driver/boot-target/Makefile ===================================== @@ -0,0 +1,8 @@ +boot1: + $(TEST_HC) -c A.hs-boot B.hs + +boot2: + $(TEST_HC) A.hs-boot A.hs B.hs -v0 + +boot3: + $(TEST_HC) A.hs-boot B.hs -v0 \ No newline at end of file ===================================== testsuite/tests/driver/boot-target/all.T ===================================== @@ -0,0 +1,10 @@ +def test_boot(name): + return test(name, + [extra_files(['A.hs', 'A.hs-boot', 'B.hs']), + ], + makefile_test, + []) + +test_boot('boot1') +test_boot('boot2') +test_boot('boot3') ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -1,3 +1,4 @@ +{-# LANGUAGE BangPatterns #-} {-# LANGUAGE DeriveDataTypeable #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE FlexibleInstances #-} @@ -25,7 +26,7 @@ module ExactPrint , makeDeltaAst -- * Configuration - , EPOptions(epRigidity, epAstPrint, epTokenPrint, epWhitespacePrint, epUpdateAnchors) + , EPOptions(epTokenPrint, epWhitespacePrint) , stringOptions , epOptions , deltaOptions @@ -43,10 +44,11 @@ import GHC.Types.ForeignCall import GHC.Types.Name.Reader import GHC.Types.PkgQual import GHC.Types.SourceText +import GHC.Types.SrcLoc import GHC.Types.Var -import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Unit.Module.Warnings import GHC.Utils.Misc +import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Utils.Panic import Language.Haskell.Syntax.Basic (FieldLabelString(..)) @@ -77,8 +79,7 @@ import Types exactPrint :: ExactPrint ast => ast -> String exactPrint ast = snd $ runIdentity (runEP stringOptions (markAnnotated ast)) --- | The additional option to specify the rigidity and printing --- configuration. +-- | The additional option to specify the printing configuration. exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) => EPOptions m b -> ast @@ -86,9 +87,8 @@ exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) exactPrintWithOptions r ast = runEP r (markAnnotated ast) --- | Transform concrete annotations into relative annotations which --- are more useful when transforming an AST. This corresponds to the --- earlier 'relativiseApiAnns'. +-- | Transform concrete annotations into relative annotations. +-- This should be unnecessary from GHC 9.10 makeDeltaAst :: ExactPrint ast => ast -> ast makeDeltaAst ast = fst $ runIdentity (runEP deltaOptions (markAnnotated ast)) @@ -115,6 +115,7 @@ defaultEPState = EPState , dPriorEndPosition = (1,1) , uAnchorSpan = badRealSrcSpan , uExtraDP = Nothing + , uExtraDPReturn = Nothing , pAcceptSpan = False , epComments = [] , epCommentsApplied = [] @@ -128,39 +129,27 @@ defaultEPState = EPState -- | The R part of RWS. The environment. Updated via 'local' as we -- enter a new AST element, having a different anchor point. data EPOptions m a = EPOptions - { - epAstPrint :: forall ast . Data ast => GHC.Located ast -> a -> m a - , epTokenPrint :: String -> m a + { epTokenPrint :: String -> m a , epWhitespacePrint :: String -> m a - , epRigidity :: Rigidity - , epUpdateAnchors :: Bool } -- | Helper to create a 'EPOptions' -epOptions :: - (forall ast . Data ast => GHC.Located ast -> a -> m a) - -> (String -> m a) - -> (String -> m a) - -> Rigidity - -> Bool - -> EPOptions m a -epOptions astPrint tokenPrint wsPrint rigidity delta = EPOptions - { - epAstPrint = astPrint - , epWhitespacePrint = wsPrint +epOptions :: (String -> m a) + -> (String -> m a) + -> EPOptions m a +epOptions tokenPrint wsPrint = EPOptions + { epWhitespacePrint = wsPrint , epTokenPrint = tokenPrint - , epRigidity = rigidity - , epUpdateAnchors = delta } -- | Options which can be used to print as a normal String. stringOptions :: EPOptions Identity String -stringOptions = epOptions (\_ b -> return b) return return NormalLayout False +stringOptions = epOptions return return -- | Options which can be used to simply update the AST to be in delta -- form, without generating output deltaOptions :: EPOptions Identity () -deltaOptions = epOptions (\_ _ -> return ()) (\_ -> return ()) (\_ -> return ()) NormalLayout True +deltaOptions = epOptions (\_ -> return ()) (\_ -> return ()) data EPWriter a = EPWriter { output :: !a } @@ -177,6 +166,8 @@ data EPState = EPState -- Annotation , uExtraDP :: !(Maybe EpaLocation) -- ^ Used to anchor a -- list + , uExtraDPReturn :: !(Maybe DeltaPos) + -- ^ Used to return Delta version of uExtraDP , pAcceptSpan :: Bool -- ^ When we have processed an -- entry of EpaDelta, accept the -- next `EpaSpan` start as the @@ -213,7 +204,7 @@ class HasTrailing a where trailing :: a -> [TrailingAnn] setTrailing :: a -> [TrailingAnn] -> a -setAnchorEpa :: (HasTrailing an, NoAnn an) +setAnchorEpa :: (HasTrailing an) => EpAnn an -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn an setAnchorEpa (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing an ts) cs @@ -223,7 +214,7 @@ setAnchorHsModule hsmod anc cs = hsmod { hsmodExt = (hsmodExt hsmod) {hsmodAnn = anc' = anc an' = setAnchorEpa (hsmodAnn $ hsmodExt hsmod) anc' [] cs -setAnchorAn :: (HasTrailing an, NoAnn an) +setAnchorAn :: (HasTrailing an) => LocatedAn an a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a setAnchorAn (L (EpAnn _ an _) a) anc ts cs = (L (EpAnn anc (setTrailing an ts) cs) a) -- `debug` ("setAnchorAn: anc=" ++ showAst anc) @@ -248,7 +239,7 @@ data FlushComments = FlushComments data CanUpdateAnchor = CanUpdateAnchor | CanUpdateAnchorOnly | NoCanUpdateAnchor - deriving (Eq, Show) + deriving (Eq, Show, Data) data Entry = Entry EpaLocation [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor | NoEntryVal @@ -402,7 +393,7 @@ enterAnn NoEntryVal a = do r <- exact a debugM $ "enterAnn:done:NO ANN:p =" ++ show (p, astId a) return r -enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do +enterAnn !(Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do acceptSpan <- getAcceptSpan setAcceptSpan False case anchor' of @@ -421,9 +412,11 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do _ -> return () case anchor' of EpaDelta _ _ dcs -> do - debugM $ "enterAnn:Printing comments:" ++ showGhc (priorComments cs) + debugM $ "enterAnn:Delta:Flushing comments" + flushComments [] + debugM $ "enterAnn:Delta:Printing prior comments:" ++ showGhc (priorComments cs) mapM_ printOneComment (concatMap tokComment $ priorComments cs) - debugM $ "enterAnn:Printing EpaDelta comments:" ++ showGhc dcs + debugM $ "enterAnn:Delta:Printing EpaDelta comments:" ++ showGhc dcs mapM_ printOneComment (concatMap tokComment dcs) _ -> do debugM $ "enterAnn:Adding comments:" ++ showGhc (priorComments cs) @@ -465,7 +458,7 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- The first part corresponds to the delta phase, so should only use -- delta phase variables ----------------------------------- -- Calculate offset required to get to the start of the SrcSPan - off <- getLayoutOffsetD + !off <- getLayoutOffsetD let spanStart = ss2pos curAnchor priorEndAfterComments <- getPriorEndD let edp' = adjustDeltaForOffset @@ -480,17 +473,18 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- --------------------------------------------- med <- getExtraDP setExtraDP Nothing - let edp = case med of - Nothing -> edp'' - Just (EpaDelta _ dp _) -> dp + let (edp, medr) = case med of + Nothing -> (edp'', Nothing) + Just (EpaDelta _ dp _) -> (dp, Nothing) -- Replace original with desired one. Allows all -- list entry values to be DP (1,0) - Just (EpaSpan (RealSrcSpan r _)) -> dp + Just (EpaSpan (RealSrcSpan r _)) -> (dp, Just dp) where dp = adjustDeltaForOffset off (ss2delta priorEndAfterComments r) Just (EpaSpan (UnhelpfulSpan r)) -> panic $ "enterAnn: UnhelpfulSpan:" ++ show r when (isJust med) $ debugM $ "enterAnn:(med,edp)=" ++ showAst (med,edp) + when (isJust medr) $ setExtraDPReturn medr -- --------------------------------------------- -- Preparation complete, perform the action when (priorEndAfterComments < spanStart) (do @@ -511,12 +505,15 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do debugM $ "enterAnn:exact a starting:" ++ show (showAst anchor') a' <- exact a debugM $ "enterAnn:exact a done:" ++ show (showAst anchor') + + -- Core recursive exactprint done, start end of Entry processing + when (flush == FlushComments) $ do - debugM $ "flushing comments in enterAnn:" ++ showAst cs + debugM $ "flushing comments in enterAnn:" ++ showAst (cs, getFollowingComments cs) flushComments (getFollowingComments cs) debugM $ "flushing comments in enterAnn done" - eof <- getEofPos + !eof <- getEofPos case eof of Nothing -> return () Just (pos, prior) -> do @@ -544,28 +541,50 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- Outside the anchor, mark any trailing postCs <- cua canUpdateAnchor takeAppliedCommentsPop - when (flush == NoFlushComments) $ do - when ((getFollowingComments cs) /= []) $ do - - -- debugM $ "enterAnn:in:(anchor') =" ++ show (eloc2str anchor') - debugM $ "starting trailing comments:" ++ showAst (getFollowingComments cs) - mapM_ printOneComment (concatMap tokComment $ getFollowingComments cs) - debugM $ "ending trailing comments" - trailing' <- markTrailing trailing_anns + following <- if (flush == NoFlushComments) + then do + let (before, after) = splitAfterTrailingAnns trailing_anns + (getFollowingComments cs) + addCommentsA before + return after + else return [] + !trailing' <- markTrailing trailing_anns + -- mapM_ printOneComment (concatMap tokComment $ following) + addCommentsA following -- Update original anchor, comments based on the printing process -- TODO:AZ: probably need to put something appropriate in instead of noSrcSpan - let newAchor = EpaDelta noSrcSpan edp [] + let newAnchor = EpaDelta noSrcSpan edp [] let r = case canUpdateAnchor of - CanUpdateAnchor -> setAnnotationAnchor a' newAchor trailing' (mkEpaComments (priorCs ++ postCs) []) - CanUpdateAnchorOnly -> setAnnotationAnchor a' newAchor [] emptyComments + CanUpdateAnchor -> setAnnotationAnchor a' newAnchor trailing' (mkEpaComments priorCs postCs) + CanUpdateAnchorOnly -> setAnnotationAnchor a' newAnchor [] emptyComments NoCanUpdateAnchor -> a' return r -- --------------------------------------------------------------------- +-- | Split the span following comments into ones that occur prior to +-- the last trailing ann, and ones after. +splitAfterTrailingAnns :: [TrailingAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitAfterTrailingAnns [] cs = ([], cs) +splitAfterTrailingAnns tas cs = (before, after) + where + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + (before, after) = case reverse (concatMap trailing_loc tas) of + [] -> ([],cs) + (s:_) -> (b,a) + where + s_pos = ss2pos s + (b,a) = break (\(L ll _) -> (ss2pos $ anchor ll) > s_pos) + cs + + +-- --------------------------------------------------------------------- + addCommentsA :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -addCommentsA csNew = addComments (concatMap tokComment csNew) +addCommentsA csNew = addComments False (concatMap tokComment csNew) {- TODO: When we addComments, some may have an anchor that is no longer @@ -583,24 +602,36 @@ By definition it is the current anchor, so work against that. And that also means that the first entry comment that has moved should not have a line offset. -} -addComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -addComments csNew = do - -- debugM $ "addComments:" ++ show csNew +addComments :: (Monad m, Monoid w) => Bool -> [Comment] -> EP w m () +addComments sortNeeded csNew = do + debugM $ "addComments:csNew" ++ show csNew cs <- getUnallocatedComments + debugM $ "addComments:cs" ++ show cs + -- We can only sort the comments if we are in the first phase, + -- where all comments have locations. If any have EpaDelta the + -- sort will fail, so we do not try. + if sortNeeded && all noDelta (csNew ++ cs) + then putUnallocatedComments (sort (cs ++ csNew)) + else putUnallocatedComments (cs ++ csNew) - putUnallocatedComments (sort (cs ++ csNew)) +noDelta :: Comment -> Bool +noDelta c = case commentLoc c of + EpaSpan _ -> True + _ -> False -- --------------------------------------------------------------------- -- | Just before we print out the EOF comments, flush the remaining -- ones in the state. flushComments :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -flushComments trailing_anns = do +flushComments !trailing_anns = do + debugM $ "flushComments entered: " ++ showAst trailing_anns addCommentsA trailing_anns + debugM $ "flushComments after addCommentsA" cs <- getUnallocatedComments - debugM $ "flushing comments starting" - -- AZ:TODO: is the sort still needed? - mapM_ printOneComment (sortComments cs) + debugM $ "flushComments: got cs" + debugM $ "flushing comments starting: cs" ++ showAst cs + mapM_ printOneComment cs putUnallocatedComments [] debugM $ "flushing comments done" @@ -612,7 +643,7 @@ annotationsToComments :: (Monad m, Monoid w) => a -> Lens a [AddEpAnn] -> [AnnKeywordId] -> EP w m a annotationsToComments a l kws = do let (newComments, newAnns) = go ([],[]) (view l a) - addComments newComments + addComments True newComments return (set l (reverse newAnns) a) where keywords = Set.fromList kws @@ -654,14 +685,11 @@ printSourceText (NoSourceText) txt = printStringAdvance txt >> return () printSourceText (SourceText txt) _ = printStringAdvance (unpackFS txt) >> return () printSourceTextAA :: (Monad m, Monoid w) => SourceText -> String -> EP w m () -printSourceTextAA (NoSourceText) txt = printStringAtAA noAnn txt >> return () -printSourceTextAA (SourceText txt) _ = printStringAtAA noAnn (unpackFS txt) >> return () +printSourceTextAA (NoSourceText) txt = printStringAdvanceA txt >> return () +printSourceTextAA (SourceText txt) _ = printStringAdvanceA (unpackFS txt) >> return () -- --------------------------------------------------------------------- -printStringAtSs :: (Monad m, Monoid w) => SrcSpan -> String -> EP w m () -printStringAtSs ss str = printStringAtRs (realSrcSpan ss) str >> return () - printStringAtRs :: (Monad m, Monoid w) => RealSrcSpan -> String -> EP w m EpaLocation printStringAtRs pa str = printStringAtRsC CaptureComments pa str @@ -676,7 +704,7 @@ printStringAtRsC capture pa str = do p' <- adjustDeltaForOffsetM p debugM $ "printStringAtRsC:(p,p')=" ++ show (p,p') printStringAtLsDelta p' str - setPriorEndASTD True pa + setPriorEndASTD pa cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -709,6 +737,9 @@ printStringAtMLocL (EpAnn anc an cs) l s = do printStringAtLsDelta (SameLine 1) str return (Just (EpaDelta noSrcSpan (SameLine 1) [])) +printStringAdvanceA :: (Monad m, Monoid w) => String -> EP w m () +printStringAdvanceA str = printStringAtAA (EpaDelta noSrcSpan (SameLine 0) []) str >> return () + printStringAtAA :: (Monad m, Monoid w) => EpaLocation -> String -> EP w m EpaLocation printStringAtAA el str = printStringAtAAC CaptureComments el str @@ -735,7 +766,7 @@ printStringAtAAC capture (EpaDelta ss d cs) s = do p2 <- getPosP pe2 <- getPriorEndD debugM $ "printStringAtAA:(pe1,pe2,p1,p2)=" ++ show (pe1,pe2,p1,p2) - setPriorEndASTPD True (pe1,pe2) + setPriorEndASTPD (pe1,pe2) cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -883,8 +914,7 @@ markAnnOpenP' :: (Monad m, Monoid w) => AnnPragma -> SourceText -> String -> EP markAnnOpenP' an NoSourceText txt = markEpAnnLMS0 an lapr_open AnnOpen (Just txt) markAnnOpenP' an (SourceText txt) _ = markEpAnnLMS0 an lapr_open AnnOpen (Just $ unpackFS txt) -markAnnOpen :: (Monad m, Monoid w) - => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] +markAnnOpen :: (Monad m, Monoid w) => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] markAnnOpen an NoSourceText txt = markEpAnnLMS'' an lidl AnnOpen (Just txt) markAnnOpen an (SourceText txt) _ = markEpAnnLMS'' an lidl AnnOpen (Just $ unpackFS txt) @@ -1589,7 +1619,7 @@ markTopLevelList ls = mapM (\a -> setLayoutTopLevelP $ markAnnotated a) ls instance (ExactPrint a) => ExactPrint (Located a) where getAnnotationEntry (L l _) = case l of UnhelpfulSpan _ -> NoEntryVal - _ -> Entry (hackSrcSpanToAnchor l) [] emptyComments NoFlushComments CanUpdateAnchorOnly + _ -> Entry (EpaSpan l) [] emptyComments NoFlushComments CanUpdateAnchorOnly setAnnotationAnchor (L l a) _anc _ts _cs = L l a @@ -1664,16 +1694,10 @@ instance ExactPrint (HsModule GhcPs) where _ -> return lo am_decls' <- markTrailing (am_decls $ anns an0) - imports' <- markTopLevelList imports - - case lo of - EpExplicitBraces _ _ -> return () - _ -> do - -- Get rid of the balance of the preceding comments before starting on the decls - flushComments [] - putUnallocatedComments [] - decls' <- markTopLevelList (filter removeDocDecl decls) + mid <- markAnnotated (HsModuleImpDecls (am_cs $ anns an0) imports decls) + let imports' = id_imps mid + let decls' = id_decls mid lo1 <- case lo0 of EpExplicitBraces open close -> do @@ -1688,15 +1712,32 @@ instance ExactPrint (HsModule GhcPs) where debugM $ "am_eof:" ++ showGhc (pos, prior) setEofPos (Just (pos, prior)) - let anf = an0 { anns = (anns an0) { am_decls = am_decls' }} + let anf = an0 { anns = (anns an0) { am_decls = am_decls', am_cs = [] }} debugM $ "HsModule, anf=" ++ showAst anf return (HsModule (XModulePs anf lo1 mdeprec' mbDoc') mmn' mexports' imports' decls') +-- --------------------------------------------------------------------- + +-- | This is used to ensure the comments are updated into the right +-- place for makeDeltaAst. +data HsModuleImpDecls + = HsModuleImpDecls { + id_cs :: [LEpaComment], + id_imps :: [LImportDecl GhcPs], + id_decls :: [LHsDecl GhcPs] + } deriving Data + +instance ExactPrint HsModuleImpDecls where + -- Use an UnhelpfulSpan for the anchor, we are only interested in the comments + getAnnotationEntry mid = mkEntry (EpaSpan (UnhelpfulSpan UnhelpfulNoLocationInfo)) [] (EpaComments (id_cs mid)) + setAnnotationAnchor mid _anc _ cs = mid { id_cs = priorComments cs ++ getFollowingComments cs } + `debug` ("HsModuleImpDecls.setAnnotationAnchor:cs=" ++ showAst cs) + exact (HsModuleImpDecls cs imports decls) = do + imports' <- markTopLevelList imports + decls' <- markTopLevelList (filter notDocDecl decls) + return (HsModuleImpDecls cs imports' decls') -removeDocDecl :: LHsDecl GhcPs -> Bool -removeDocDecl (L _ DocD{}) = False -removeDocDecl _ = True -- --------------------------------------------------------------------- @@ -1737,8 +1778,8 @@ instance ExactPrint InWarningCategory where exact (InWarningCategory tkIn source (L l wc)) = do tkIn' <- markEpToken tkIn - L _ (_,wc') <- markAnnotated (L l (source, wc)) - return (InWarningCategory tkIn' source (L l wc')) + L l' (_,wc') <- markAnnotated (L l (source, wc)) + return (InWarningCategory tkIn' source (L l' wc')) instance ExactPrint (SourceText, WarningCategory) where getAnnotationEntry _ = NoEntryVal @@ -1943,14 +1984,14 @@ exactDataFamInstDecl an top_lvl , feqn_pats = pats , feqn_fixity = fixity , feqn_rhs = defn })) = do - (an', an2', tycon', bndrs', _, _mc, defn') <- exactDataDefn an2 pp_hdr defn - -- See Note [an and an2 in exactDataFamInstDecl] + (an', an2', tycon', bndrs', pats', defn') <- exactDataDefn an2 pp_hdr defn + -- See Note [an and an2 in exactDataFamInstDecl] return (an', DataFamInstDecl ( FamEqn { feqn_ext = an2' , feqn_tycon = tycon' , feqn_bndrs = bndrs' - , feqn_pats = pats + , feqn_pats = pats' , feqn_fixity = fixity , feqn_rhs = defn' })) `debug` ("exactDataFamInstDecl: defn' derivs:" ++ showAst (dd_derivs defn')) @@ -2233,11 +2274,11 @@ instance ExactPrint (RoleAnnotDecl GhcPs) where an1 <- markEpAnnL an0 lidl AnnRole ltycon' <- markAnnotated ltycon let markRole (L l (Just r)) = do - (L _ r') <- markAnnotated (L l r) - return (L l (Just r')) + (L l' r') <- markAnnotated (L l r) + return (L l' (Just r')) markRole (L l Nothing) = do - printStringAtSs (locA l) "_" - return (L l Nothing) + e' <- printStringAtAA (entry l) "_" + return (L (l { entry = e'}) Nothing) roles' <- mapM markRole roles return (RoleAnnotDecl an1 ltycon' roles') @@ -2340,8 +2381,13 @@ instance (ExactPrint tm, ExactPrint ty, Outputable tm, Outputable ty) getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact a@(HsValArg _ tm) = markAnnotated tm >> return a - exact a@(HsTypeArg at ty) = markEpToken at >> markAnnotated ty >> return a + exact (HsValArg x tm) = do + tm' <- markAnnotated tm + return (HsValArg x tm') + exact (HsTypeArg at ty) = do + at' <- markEpToken at + ty' <- markAnnotated ty + return (HsTypeArg at' ty') exact x@(HsArgPar _sp) = withPpr x -- Does not appear in original source -- --------------------------------------------------------------------- @@ -2359,9 +2405,9 @@ instance ExactPrint (ClsInstDecl GhcPs) where (mbWarn', an0, mbOverlap', inst_ty') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lid AnnSemi - ds <- withSortKey sortKey - [(ClsAtdTag, prepareListAnnotationA ats), - (ClsAtdTag, prepareListAnnotationF an adts), + (sortKey', ds) <- withSortKey sortKey + [(ClsAtTag, prepareListAnnotationA ats), + (ClsAtdTag, prepareListAnnotationF adts), (ClsMethodTag, prepareListAnnotationA binds), (ClsSigTag, prepareListAnnotationA sigs) ] @@ -2371,7 +2417,7 @@ instance ExactPrint (ClsInstDecl GhcPs) where adts' = undynamic ds binds' = undynamic ds sigs' = undynamic ds - return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey) + return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey') , cid_poly_ty = inst_ty', cid_binds = binds' , cid_sigs = sigs', cid_tyfam_insts = ats' , cid_overlap_mode = mbOverlap' @@ -2452,15 +2498,29 @@ instance ExactPrint (HsBind GhcPs) where return (FunBind x fun_id' matches') exact (PatBind x pat q grhss) = do + q' <- markAnnotated q pat' <- markAnnotated pat grhss' <- markAnnotated grhss - return (PatBind x pat' q grhss') + return (PatBind x pat' q' grhss') exact (PatSynBind x bind) = do bind' <- markAnnotated bind return (PatSynBind x bind') exact x = error $ "HsBind: exact for " ++ showAst x +instance ExactPrint (HsMultAnn GhcPs) where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact (HsNoMultAnn x) = return (HsNoMultAnn x) + exact (HsPct1Ann tok) = do + tok' <- markEpToken tok + return (HsPct1Ann tok') + exact (HsMultAnn tok ty) = do + tok' <- markEpToken tok + ty' <- markAnnotated ty + return (HsMultAnn tok' ty') + -- --------------------------------------------------------------------- instance ExactPrint (PatSynBind GhcPs GhcPs) where @@ -2519,8 +2579,9 @@ instance ExactPrint (PatSynBind GhcPs GhcPs) where instance ExactPrint (RecordPatSynField GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact r@(RecordPatSynField { recordPatSynField = v }) = markAnnotated v - >> return r + exact (RecordPatSynField f v) = do + f' <- markAnnotated f + return (RecordPatSynField f' v) -- --------------------------------------------------------------------- @@ -2648,15 +2709,20 @@ instance ExactPrint (HsLocalBinds GhcPs) where (an1, valbinds') <- markAnnList an0 $ markAnnotatedWithLayout valbinds debugM $ "exact HsValBinds: an1=" ++ showAst an1 - return (HsValBinds an1 valbinds') + medr <- getExtraDPReturn + an2 <- case medr of + Nothing -> return an1 + Just dp -> do + setExtraDPReturn Nothing + return $ an1 { anns = (anns an1) { al_anchor = Just (EpaDelta noSrcSpan dp []) }} + return (HsValBinds an2 valbinds') exact (HsIPBinds an bs) = do - (as, ipb) <- markAnnList an (markEpAnnL' an lal_rest AnnWhere - >> markAnnotated bs - >>= \bs' -> return (HsIPBinds an bs'::HsLocalBinds GhcPs)) - case ipb of - HsIPBinds _ bs' -> return (HsIPBinds as bs'::HsLocalBinds GhcPs) - _ -> error "should not happen HsIPBinds" + (an2,bs') <- markAnnListA an $ \an0 -> do + an1 <- markEpAnnL' an0 lal_rest AnnWhere + bs' <- markAnnotated bs + return (an1, bs') + return (HsIPBinds an2 bs') exact b@(EmptyLocalBinds _) = return b @@ -2670,7 +2736,8 @@ instance ExactPrint (HsValBindsLR GhcPs GhcPs) where let binds' = concatMap decl2Bind decls sigs' = concatMap decl2Sig decls - return (ValBinds sortKey binds' sigs') + sortKey' = captureOrderBinds decls + return (ValBinds sortKey' binds' sigs') exact (XValBindsLR _) = panic "XValBindsLR" undynamic :: Typeable a => [Dynamic] -> [a] @@ -2682,7 +2749,9 @@ instance ExactPrint (HsIPBinds GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact b@(IPBinds _ binds) = setLayoutBoth $ markAnnotated binds >> return b + exact (IPBinds x binds) = setLayoutBoth $ do + binds' <- markAnnotated binds + return (IPBinds x binds') -- --------------------------------------------------------------------- @@ -2703,18 +2772,18 @@ instance ExactPrint HsIPName where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact i@(HsIPName fs) = printStringAdvance ("?" ++ (unpackFS fs)) >> return i + exact i@(HsIPName fs) = printStringAdvanceA ("?" ++ (unpackFS fs)) >> return i -- --------------------------------------------------------------------- -- Managing lists which have been separated, e.g. Sigs and Binds prepareListAnnotationF :: (Monad m, Monoid w) => - [AddEpAnn] -> [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] -prepareListAnnotationF an ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls + [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] +prepareListAnnotationF ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls where go (L l a) = do - d' <- markAnnotated (DataFamInstDeclWithContext an NotTopLevel a) - return (toDyn (L l (dc_d d'))) + (L l' d') <- markAnnotated (L l (DataFamInstDeclWithContext noAnn NotTopLevel a)) + return (toDyn (L l' (dc_d d'))) prepareListAnnotationA :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => [LocatedAn an a] -> [(RealSrcSpan,EP w m Dynamic)] @@ -2725,15 +2794,23 @@ prepareListAnnotationA ls = map (\b -> (realSrcSpan $ getLocA b,go b)) ls return (toDyn b') withSortKey :: (Monad m, Monoid w) - => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] -> EP w m [Dynamic] + => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] + -> EP w m (AnnSortKey DeclTag, [Dynamic]) withSortKey annSortKey xs = do debugM $ "withSortKey:annSortKey=" ++ showAst annSortKey - let ordered = case annSortKey of - NoAnnSortKey -> sortBy orderByFst $ concatMap snd xs - AnnSortKey _keys -> orderedDecls annSortKey (Map.fromList xs) - mapM snd ordered -orderByFst :: Ord a => (a, b1) -> (a, b2) -> Ordering -orderByFst (a,_) (b,_) = compare a b + let (sk, ordered) = case annSortKey of + NoAnnSortKey -> (annSortKey', map snd os) + where + doOne (tag, ds) = map (\d -> (tag, d)) ds + xsExpanded = concatMap doOne xs + os = sortBy orderByFst $ xsExpanded + annSortKey' = AnnSortKey (map fst os) + AnnSortKey _keys -> (annSortKey, orderedDecls annSortKey (Map.fromList xs)) + ordered' <- mapM snd ordered + return (sk, ordered') + +orderByFst :: Ord a => (t, (a,b1)) -> (t, (a, b2)) -> Ordering +orderByFst (_,(a,_)) (_,(b,_)) = compare a b -- --------------------------------------------------------------------- @@ -2761,15 +2838,16 @@ instance ExactPrint (Sig GhcPs) where (an0, vars',ty') <- exactVarSig an vars ty return (ClassOpSig an0 is_deflt vars' ty') - exact (FixSig (an,src) (FixitySig x names (Fixity v fdir))) = do + exact (FixSig (an,src) (FixitySig ns names (Fixity v fdir))) = do let fixstr = case fdir of InfixL -> "infixl" InfixR -> "infixr" InfixN -> "infix" an0 <- markEpAnnLMS'' an lidl AnnInfix (Just fixstr) an1 <- markEpAnnLMS'' an0 lidl AnnVal (Just (sourceTextToString src (show v))) + ns' <- markAnnotated ns names' <- markAnnotated names - return (FixSig (an1,src) (FixitySig x names' (Fixity v fdir))) + return (FixSig (an1,src) (FixitySig ns' names' (Fixity v fdir))) exact (InlineSig an ln inl) = do an0 <- markAnnOpen an (inl_src inl) "{-# INLINE" @@ -2809,7 +2887,7 @@ instance ExactPrint (Sig GhcPs) where exact (CompleteMatchSig (an,src) cs mty) = do an0 <- markAnnOpen an src "{-# COMPLETE" - cs' <- markAnnotated cs + cs' <- mapM markAnnotated cs (an1, mty') <- case mty of Nothing -> return (an0, mty) @@ -2822,6 +2900,20 @@ instance ExactPrint (Sig GhcPs) where -- --------------------------------------------------------------------- +instance ExactPrint NamespaceSpecifier where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact NoNamespaceSpecifier = return NoNamespaceSpecifier + exact (TypeNamespaceSpecifier typeTok) = do + typeTok' <- markEpToken typeTok + return (TypeNamespaceSpecifier typeTok') + exact (DataNamespaceSpecifier dataTok) = do + dataTok' <- markEpToken dataTok + return (DataNamespaceSpecifier dataTok') + +-- --------------------------------------------------------------------- + exactVarSig :: (Monad m, Monoid w, ExactPrint a) => AnnSig -> [LocatedN RdrName] -> a -> EP w m (AnnSig, [LocatedN RdrName], a) exactVarSig an vars ty = do @@ -2875,7 +2967,7 @@ instance ExactPrint (AnnDecl GhcPs) where n' <- markAnnotated n return (an1, TypeAnnProvenance n') ModuleAnnProvenance -> do - an1 <- markEpAnnL an lapr_rest AnnModule + an1 <- markEpAnnL an0 lapr_rest AnnModule return (an1, prov) e' <- markAnnotated e @@ -2950,21 +3042,21 @@ instance ExactPrint (HsExpr GhcPs) where then markAnnotated n else return n return (HsVar x n') - exact x@(HsUnboundVar an _) = do + exact (HsUnboundVar an n) = do case an of Just (EpAnnUnboundVar (ob,cb) l) -> do - printStringAtAA ob "`" >> return () - printStringAtAA l "_" >> return () - printStringAtAA cb "`" >> return () - return x + ob' <- printStringAtAA ob "`" + l' <- printStringAtAA l "_" + cb' <- printStringAtAA cb "`" + return (HsUnboundVar (Just (EpAnnUnboundVar (ob',cb') l')) n) _ -> do - printStringAtLsDelta (SameLine 0) "_" - return x + printStringAdvanceA "_" >> return () + return (HsUnboundVar an n) exact x@(HsOverLabel src l) = do - printStringAtLsDelta (SameLine 0) "#" + printStringAdvanceA "#" >> return () case src of - NoSourceText -> printStringAtLsDelta (SameLine 0) (unpackFS l) - SourceText txt -> printStringAtLsDelta (SameLine 0) (unpackFS txt) + NoSourceText -> printStringAdvanceA (unpackFS l) >> return () + SourceText txt -> printStringAdvanceA (unpackFS txt) >> return () return x exact x@(HsIPVar _ (HsIPName n)) @@ -3204,11 +3296,11 @@ instance ExactPrint (HsExpr GhcPs) where exact (HsTypedSplice an s) = do an0 <- markEpAnnL an lidl AnnDollarDollar - s' <- exact s + s' <- markAnnotated s return (HsTypedSplice an0 s') exact (HsUntypedSplice an s) = do - s' <- exact s + s' <- markAnnotated s return (HsUntypedSplice an s') exact (HsProc an p c) = do @@ -3274,12 +3366,15 @@ exactMdo an (Just module_name) kw = markEpAnnLMS'' an lal_rest kw (Just n) markMaybeDodgyStmts :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => AnnList -> LocatedAn an a -> EP w m (AnnList, LocatedAn an a) markMaybeDodgyStmts an stmts = - if isGoodSrcSpan (getLocA stmts) + if notDodgy stmts then do r <- markAnnotatedWithLayout stmts return (an, r) else return (an, stmts) +notDodgy :: GenLocated (EpAnn ann) a -> Bool +notDodgy (L (EpAnn anc _ _) _) = notDodgyE anc + notDodgyE :: EpaLocation -> Bool notDodgyE anc = case anc of @@ -3341,7 +3436,7 @@ instance ExactPrint (MatchGroup GhcPs (LocatedA (HsCmd GhcPs))) where setAnnotationAnchor a _ _ _ = a exact (MG x matches) = do -- TODO:AZ use SortKey, in MG ann. - matches' <- if isGoodSrcSpan (getLocA matches) + matches' <- if notDodgy matches then markAnnotated matches else return matches return (MG x matches') @@ -3661,6 +3756,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- There may be arbitrary parens around parts of the constructor -- that are infix. Turn these into comments so that they feed -- into the right place automatically + -- TODO: no longer sorting on insert. What now? an0 <- annotationsToComments an lidl [AnnOpenP,AnnCloseP] an1 <- markEpAnnL an0 lidl AnnType @@ -3674,7 +3770,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- TODO: add a workaround for https://gitlab.haskell.org/ghc/ghc/-/issues/20452 exact (DataDecl { tcdDExt = an, tcdLName = ltycon, tcdTyVars = tyvars , tcdFixity = fixity, tcdDataDefn = defn }) = do - (_, an', ltycon', tyvars', _, _mctxt', defn') <- + (_, an', ltycon', tyvars', _, defn') <- exactDataDefn an (exactVanillaDeclHead ltycon tyvars fixity) defn return (DataDecl { tcdDExt = an', tcdLName = ltycon', tcdTyVars = tyvars' , tcdFixity = fixity, tcdDataDefn = defn' }) @@ -3707,7 +3803,7 @@ instance ExactPrint (TyClDecl GhcPs) where (an0, fds', lclas', tyvars',context') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lidl AnnSemi - ds <- withSortKey sortKey + (sortKey', ds) <- withSortKey sortKey [(ClsSigTag, prepareListAnnotationA sigs), (ClsMethodTag, prepareListAnnotationA methods), (ClsAtTag, prepareListAnnotationA ats), @@ -3720,7 +3816,7 @@ instance ExactPrint (TyClDecl GhcPs) where methods' = undynamic ds ats' = undynamic ds at_defs' = undynamic ds - return (ClassDecl {tcdCExt = (an3, lo, sortKey), + return (ClassDecl {tcdCExt = (an3, lo, sortKey'), tcdCtxt = context', tcdLName = lclas', tcdTyVars = tyvars', tcdFixity = fixity, tcdFDs = fds', @@ -3845,7 +3941,7 @@ exactDataDefn -> HsDataDefn GhcPs -> EP w m ( [AddEpAnn] -- ^ from exactHdr , [AddEpAnn] -- ^ updated one passed in - , LocatedN RdrName, a, b, Maybe (LHsContext GhcPs), HsDataDefn GhcPs) + , LocatedN RdrName, a, b, HsDataDefn GhcPs) exactDataDefn an exactHdr (HsDataDefn { dd_ext = x, dd_ctxt = context , dd_cType = mb_ct @@ -3883,8 +3979,8 @@ exactDataDefn an exactHdr _ -> panic "exacprint NewTypeCon" an6 <- markEpAnnL an5 lidl AnnCloseC derivings' <- mapM markAnnotated derivings - return (anx, an6, ln', tvs', b, mctxt', - (HsDataDefn { dd_ext = x, dd_ctxt = context + return (anx, an6, ln', tvs', b, + (HsDataDefn { dd_ext = x, dd_ctxt = mctxt' , dd_cType = mb_ct' , dd_kindSig = mb_sig' , dd_cons = condecls'', dd_derivs = derivings' })) @@ -3941,22 +4037,23 @@ instance ExactPrint (InjectivityAnn GhcPs) where class Typeable flag => ExactPrintTVFlag flag where exactTVDelimiters :: (Monad m, Monoid w) - => [AddEpAnn] -> flag -> EP w m (HsTyVarBndr flag GhcPs) - -> EP w m ([AddEpAnn], (HsTyVarBndr flag GhcPs)) + => [AddEpAnn] -> flag + -> ([AddEpAnn] -> EP w m ([AddEpAnn], HsTyVarBndr flag GhcPs)) + -> EP w m ([AddEpAnn], flag, (HsTyVarBndr flag GhcPs)) instance ExactPrintTVFlag () where - exactTVDelimiters an _ thing_inside = do + exactTVDelimiters an flag thing_inside = do an0 <- markEpAnnAllL' an lid AnnOpenP - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid AnnCloseP - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid AnnCloseP + return (an2, flag, r) instance ExactPrintTVFlag Specificity where exactTVDelimiters an s thing_inside = do an0 <- markEpAnnAllL' an lid open - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid close - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid close + return (an2, s, r) where (open, close) = case s of SpecifiedSpec -> (AnnOpenP, AnnCloseP) @@ -3964,33 +4061,33 @@ instance ExactPrintTVFlag Specificity where instance ExactPrintTVFlag (HsBndrVis GhcPs) where exactTVDelimiters an0 bvis thing_inside = do - case bvis of - HsBndrRequired _ -> return () - HsBndrInvisible at -> markEpToken at >> return () + bvis' <- case bvis of + HsBndrRequired _ -> return bvis + HsBndrInvisible at -> HsBndrInvisible <$> markEpToken at an1 <- markEpAnnAllL' an0 lid AnnOpenP - r <- thing_inside - an2 <- markEpAnnAllL' an1 lid AnnCloseP - return (an2, r) + (an2, r) <- thing_inside an1 + an3 <- markEpAnnAllL' an2 lid AnnCloseP + return (an3, bvis', r) instance ExactPrintTVFlag flag => ExactPrint (HsTyVarBndr flag GhcPs) where getAnnotationEntry _ = NoEntryVal setAnnotationAnchor a _ _ _ = a exact (UserTyVar an flag n) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - return (UserTyVar an flag n') + return (ani, UserTyVar an flag n') case r of - (an', UserTyVar _ flag'' n'') -> return (UserTyVar an' flag'' n'') + (an', flag', UserTyVar _ _ n'') -> return (UserTyVar an' flag' n'') _ -> error "KindedTyVar should never happen here" exact (KindedTyVar an flag n k) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - an0 <- markEpAnnL an lidl AnnDcolon + an0 <- markEpAnnL ani lidl AnnDcolon k' <- markAnnotated k - return (KindedTyVar an0 flag n' k') + return (an0, KindedTyVar an0 flag n' k') case r of - (an',KindedTyVar _ flag'' n'' k'') -> return (KindedTyVar an' flag'' n'' k'') + (an',flag', KindedTyVar _ _ n'' k'') -> return (KindedTyVar an' flag' n'' k'') _ -> error "UserTyVar should never happen here" -- --------------------------------------------------------------------- @@ -4150,17 +4247,16 @@ instance ExactPrint (HsDerivingClause GhcPs) where , deriv_clause_strategy = dcs , deriv_clause_tys = dct }) = do an0 <- markEpAnnL an lidl AnnDeriving - exact_strat_before + dcs0 <- case dcs of + Just (L _ ViaStrategy{}) -> return dcs + _ -> mapM markAnnotated dcs dct' <- markAnnotated dct - exact_strat_after + dcs1 <- case dcs0 of + Just (L _ ViaStrategy{}) -> mapM markAnnotated dcs0 + _ -> return dcs0 return (HsDerivingClause { deriv_clause_ext = an0 - , deriv_clause_strategy = dcs + , deriv_clause_strategy = dcs1 , deriv_clause_tys = dct' }) - where - (exact_strat_before, exact_strat_after) = - case dcs of - Just v@(L _ ViaStrategy{}) -> (pure (), markAnnotated v >> pure ()) - _ -> (mapM_ markAnnotated dcs, pure ()) -- --------------------------------------------------------------------- @@ -4467,7 +4563,9 @@ instance ExactPrint (ConDeclField GhcPs) where instance ExactPrint (FieldOcc GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact f@(FieldOcc _ n) = markAnnotated n >> return f + exact (FieldOcc x n) = do + n' <- markAnnotated n + return (FieldOcc x n') -- --------------------------------------------------------------------- @@ -4535,7 +4633,7 @@ instance ExactPrint (LocatedL [LocatedA (IE GhcPs)]) where an0 <- markEpAnnL' an lal_rest AnnHiding p <- getPosP debugM $ "LocatedL [LIE:p=" ++ showPprUnsafe p - (an1, ies') <- markAnnList an0 (markAnnotated ies) + (an1, ies') <- markAnnList an0 (markAnnotated (filter notIEDoc ies)) return (L an1 ies') instance (ExactPrint (Match GhcPs (LocatedA body))) @@ -4985,6 +5083,14 @@ setExtraDP md = do debugM $ "setExtraDP:" ++ show md modify (\s -> s {uExtraDP = md}) +getExtraDPReturn :: (Monad m, Monoid w) => EP w m (Maybe DeltaPos) +getExtraDPReturn = gets uExtraDPReturn + +setExtraDPReturn :: (Monad m, Monoid w) => Maybe DeltaPos -> EP w m () +setExtraDPReturn md = do + debugM $ "setExtraDPReturn:" ++ show md + modify (\s -> s {uExtraDPReturn = md}) + getPriorEndD :: (Monad m, Monoid w) => EP w m Pos getPriorEndD = gets dPriorEndPosition @@ -5007,13 +5113,13 @@ setPriorEndNoLayoutD pe = do debugM $ "setPriorEndNoLayoutD:pe=" ++ show pe modify (\s -> s { dPriorEndPosition = pe }) -setPriorEndASTD :: (Monad m, Monoid w) => Bool -> RealSrcSpan -> EP w m () -setPriorEndASTD layout pe = setPriorEndASTPD layout (rs2range pe) +setPriorEndASTD :: (Monad m, Monoid w) => RealSrcSpan -> EP w m () +setPriorEndASTD pe = setPriorEndASTPD (rs2range pe) -setPriorEndASTPD :: (Monad m, Monoid w) => Bool -> (Pos,Pos) -> EP w m () -setPriorEndASTPD layout pe@(fm,to) = do +setPriorEndASTPD :: (Monad m, Monoid w) => (Pos,Pos) -> EP w m () +setPriorEndASTPD pe@(fm,to) = do debugM $ "setPriorEndASTD:pe=" ++ show pe - when layout $ setLayoutStartD (snd fm) + setLayoutStartD (snd fm) modify (\s -> s { dPriorEndPosition = to } ) setLayoutStartD :: (Monad m, Monoid w) => Int -> EP w m () @@ -5044,7 +5150,7 @@ getUnallocatedComments :: (Monad m, Monoid w) => EP w m [Comment] getUnallocatedComments = gets epComments putUnallocatedComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -putUnallocatedComments cs = modify (\s -> s { epComments = cs } ) +putUnallocatedComments !cs = modify (\s -> s { epComments = cs } ) -- | Push a fresh stack frame for the applied comments gatherer pushAppliedComments :: (Monad m, Monoid w) => EP w m () @@ -5054,7 +5160,7 @@ pushAppliedComments = modify (\s -> s { epCommentsApplied = []:(epCommentsApplie -- takeAppliedComments, and clear them, not popping the stack takeAppliedComments :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedComments = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5067,7 +5173,7 @@ takeAppliedComments = do -- takeAppliedComments, and clear them, popping the stack takeAppliedCommentsPop :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedCommentsPop = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5080,7 +5186,7 @@ takeAppliedCommentsPop = do -- when doing delta processing applyComment :: (Monad m, Monoid w) => Comment -> EP w m () applyComment c = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> modify (\s -> s { epCommentsApplied = [[c]] } ) (h:t) -> modify (\s -> s { epCommentsApplied = (c:h):t } ) ===================================== utils/check-exact/Main.hs ===================================== @@ -470,7 +470,7 @@ changeAddDecl1 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtStart m decl' + replaceTopLevelDecls m = return $ insertAtStart m decl' return p' -- --------------------------------------------------------------------- @@ -483,7 +483,7 @@ changeAddDecl2 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtEnd m decl' + replaceTopLevelDecls m = return $ insertAtEnd m decl' return p' -- --------------------------------------------------------------------- @@ -500,7 +500,7 @@ changeAddDecl3 libdir top = do l2' = setEntryDP l2 (DifferentLine 2 0) replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAt f m decl' + replaceTopLevelDecls m = return $ insertAt f m decl' return p' -- --------------------------------------------------------------------- @@ -571,8 +571,9 @@ changeLocalDecls2 libdir (L l p) = do changeWhereIn3a :: Changer changeWhereIn3a _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) - debugM $ unlines w + decls = balanceCommentsList decls0 + (_de0:_:de1:_d2:_) = decls + debugM $ "changeWhereIn3a:de1:" ++ showAst de1 let p2 = p { hsmodDecls = decls} return (L l p2) @@ -581,13 +582,12 @@ changeWhereIn3a _libdir (L l p) = do changeWhereIn3b :: Changer changeWhereIn3b _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) + decls = balanceCommentsList decls0 (de0:tdecls@(_:de1:d2:_)) = decls de0' = setEntryDP de0 (DifferentLine 2 0) de1' = setEntryDP de1 (DifferentLine 2 0) d2' = setEntryDP d2 (DifferentLine 2 0) decls' = d2':de1':de0':tdecls - debugM $ unlines w debugM $ "changeWhereIn3b:de1':" ++ showAst de1' let p2 = p { hsmodDecls = decls'} return (L l p2) @@ -598,37 +598,37 @@ addLocaLDecl1 :: Changer addLocaLDecl1 libdir top = do Right (L ld (ValD _ decl)) <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let decl' = setEntryDP (L ld decl) (DifferentLine 1 5) - doAddLocal = do - let lp = top - (de1:d2:d3:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - (de1',_) <- modifyValD (getLocA de1'') de1'' $ \_m d -> do - return ((wrapDecl decl' : d),Nothing) - replaceDecls lp [de1', d2', d3] - - (lp',_,w) <- runTransformT doAddLocal - debugM $ "addLocaLDecl1:" ++ intercalate "\n" w + doAddLocal :: ParsedSource + doAddLocal = replaceDecls lp [de1', d2', d3] + where + lp = top + (de1:d2:d3:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 + (de1',_) = modifyValD (getLocA de1'') de1'' $ \_m d -> ((wrapDecl decl' : d),Nothing) + + let lp' = doAddLocal return lp' -- --------------------------------------------------------------------- + addLocaLDecl2 :: Changer addLocaLDecl2 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 + doAddLocal = replaceDecls lp [parent',d2'] + where + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - newDecl' <- transferEntryDP' d newDecl - let d' = setEntryDP d (DifferentLine 1 0) - return ((newDecl':d':ds),Nothing) + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = transferEntryDP' d (makeDeltaAst newDecl) + d' = setEntryDP d (DifferentLine 1 0) + in ((newDecl':d':ds),Nothing) - replaceDecls lp [parent',d2'] - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -637,19 +637,18 @@ addLocaLDecl3 :: Changer addLocaLDecl3 libdir top = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - let lp = top - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - return (((d:ds) ++ [newDecl']),Nothing) + doAddLocal = replaceDecls (anchorEof lp) [parent',d2'] + where + lp = top + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - replaceDecls (anchorEof lp) [parent',d2'] + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = setEntryDP newDecl (DifferentLine 1 0) + in (((d:ds) ++ [newDecl']),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -659,40 +658,38 @@ addLocaLDecl4 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") Right newSig <- withDynFlags libdir (\df -> parseDecl df "sig" "nn :: Int") let - doAddLocal = do - (parent:ds) <- hsDecls lp + doAddLocal = replaceDecls (anchorEof lp) (parent':ds) + where + (parent:ds) = hsDecls (makeDeltaAst lp) - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - let newSig' = setEntryDP newSig (DifferentLine 1 4) + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 0) + newSig' = setEntryDP (makeDeltaAst newSig) (DifferentLine 1 5) - (parent',_) <- modifyValD (getLocA parent) parent $ \_m decls -> do - return ((decls++[newSig',newDecl']),Nothing) + (parent',_) = modifyValD (getLocA parent) parent $ \_m decls -> + ((decls++[newSig',newDecl']),Nothing) - replaceDecls (anchorEof lp) (parent':ds) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' - -- --------------------------------------------------------------------- addLocaLDecl5 :: Changer addLocaLDecl5 _libdir lp = do let - doAddLocal = do - decls <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList decls + doAddLocal = replaceDecls lp [s1,de1',d3'] + where + decls = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList decls - let d3' = setEntryDP d3 (DifferentLine 2 0) + d3' = setEntryDP d3 (DifferentLine 2 0) - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m _decls -> do - let d2' = setEntryDP d2 (DifferentLine 1 0) - return ([d2'],Nothing) - replaceDecls lp [s1,de1',d3'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m _decls -> + let + d2' = setEntryDP d2 (DifferentLine 1 0) + in ([d2'],Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -701,39 +698,36 @@ addLocaLDecl6 :: Changer addLocaLDecl6 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "x = 3") let - newDecl' = setEntryDP newDecl (DifferentLine 1 4) - doAddLocal = do - decls0 <- hsDecls lp - [de1'',d2] <- balanceCommentsList decls0 + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 5) + doAddLocal = replaceDecls lp [de1', d2] + where + decls0 = hsDecls lp + [de1'',d2] = balanceCommentsList decls0 - let de1 = captureMatchLineSpacing de1'' - let L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 - let [ma1,_ma2] = ms + de1 = captureMatchLineSpacing de1'' + L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 + [ma1,_ma2] = ms - (de1',_) <- modifyValD (getLocA ma1) de1 $ \_m decls -> do - return ((newDecl' : decls),Nothing) - replaceDecls lp [de1', d2] + (de1',_) = modifyValD (getLocA ma1) de1 $ \_m decls -> + ((newDecl' : decls),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- rmDecl1 :: Changer -rmDecl1 _libdir top = do - let doRmDecl = do - let lp = top - tlDecs0 <- hsDecls lp - tlDecs' <- balanceCommentsList tlDecs0 - let tlDecs = captureLineSpacing tlDecs' - let (de1:_s1:_d2:d3:ds) = tlDecs - let d3' = setEntryDP d3 (DifferentLine 2 0) - - replaceDecls lp (de1:d3':ds) - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" +rmDecl1 _libdir lp = do + let + doRmDecl = replaceDecls lp (de1:d3':ds) + where + tlDecs0 = hsDecls lp + tlDecs = balanceCommentsList tlDecs0 + (de1:_s1:_d2:d3:ds) = tlDecs + d3' = setEntryDP d3 (DifferentLine 2 0) + + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -745,13 +739,13 @@ rmDecl2 _libdir lp = do let go :: GHC.LHsExpr GhcPs -> Transform (GHC.LHsExpr GhcPs) go e@(GHC.L _ (GHC.HsLet{})) = do - decs0 <- hsDecls e - decs <- balanceCommentsList $ captureLineSpacing decs0 - e' <- replaceDecls e (init decs) + let decs0 = hsDecls e + let decs = balanceCommentsList $ captureLineSpacing decs0 + let e' = replaceDecls e (init decs) return e' go x = return x - everywhereM (mkM go) lp + everywhereM (mkM go) (makeDeltaAst lp) let (lp',_,_w) = runTransform doRmDecl debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" @@ -762,17 +756,15 @@ rmDecl2 _libdir lp = do rmDecl3 :: Changer rmDecl3 _libdir lp = do let - doRmDecl = do - [de1,d2] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1] -> do - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([],Just sd1') - - replaceDecls lp [de1',sd1,d2] + doRmDecl = replaceDecls lp [de1',sd1,d2] + where + [de1,d2] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a] -> + let + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([],Just sd1') - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -780,19 +772,15 @@ rmDecl3 _libdir lp = do rmDecl4 :: Changer rmDecl4 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1,sd2] -> do - sd2' <- transferEntryDP' sd1 sd2 - - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([sd2'],Just sd1') - - replaceDecls (anchorEof lp) [de1',sd1] - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + doRmDecl = replaceDecls (anchorEof lp) [de1',sd1] + where + [de1] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a,sd2] -> + let + sd2' = transferEntryDP' sd1a sd2 + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([sd2'],Just sd1') + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -805,10 +793,8 @@ rmDecl5 _libdir lp = do go :: HsExpr GhcPs -> Transform (HsExpr GhcPs) go (HsLet (tkLet, tkIn) lb expr) = do let decs = hsDeclsLocalBinds lb - let hdecs : _ = decs let dec = last decs - _ <- transferEntryDP hdecs dec - lb' <- replaceDeclsValbinds WithoutWhere lb [dec] + let lb' = replaceDeclsValbinds WithoutWhere lb [dec] return (HsLet (tkLet, tkIn) lb' expr) go x = return x @@ -823,73 +809,61 @@ rmDecl5 _libdir lp = do rmDecl6 :: Changer rmDecl6 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m subDecs -> do - let subDecs' = captureLineSpacing subDecs - let (ss1:_sd1:sd2:sds) = subDecs' - sd2' <- transferEntryDP' ss1 sd2 - - return (sd2':sds,Nothing) + doRmDecl = replaceDecls lp [de1'] + where + [de1] = hsDecls lp - replaceDecls lp [de1'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m subDecs -> + let + subDecs' = captureLineSpacing subDecs + (ss1:_sd1:sd2:sds) = subDecs' + sd2' = transferEntryDP' ss1 sd2 + in (sd2':sds,Nothing) - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmDecl7 :: Changer -rmDecl7 _libdir top = do +rmDecl7 _libdir lp = do let - doRmDecl = do - let lp = top - tlDecs <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList tlDecs - - d3' <- transferEntryDP' d2 d3 - - replaceDecls lp [s1,de1,d3'] + doRmDecl = replaceDecls lp [s1,de1,d3'] + where + tlDecs = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList tlDecs + d3' = transferEntryDP' d2 d3 - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig1 :: Changer rmTypeSig1 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let (s0:de1:d2) = tlDecs - s1 = captureTypeSigSpacing s0 - (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 - L ln n2' <- transferEntryDP n1 n2 - let s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) - replaceDecls lp (s1':de1:d2) - - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let doRmDecl = replaceDecls lp (s1':de1:d2) + where + tlDecs = hsDecls lp + (s0:de1:d2) = tlDecs + s1 = captureTypeSigSpacing s0 + (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 + L ln n2' = transferEntryDP n1 n2 + s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig2 :: Changer rmTypeSig2 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let [de1] = tlDecs - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m [s,d] -> do - d' <- transferEntryDP' s d - return $ ([d'],Nothing) - `debug` ("rmTypeSig2:(d,d')" ++ showAst (d,d')) - replaceDecls lp [de1'] + let doRmDecl = replaceDecls lp [de1'] + where + tlDecs = hsDecls lp + [de1] = tlDecs + (de1',_) = modifyValD (getLocA de1) de1 $ \_m [_s,d] -> ([d],Nothing) - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -958,13 +932,15 @@ addClassMethod libdir lp = do let decl' = setEntryDP decl (DifferentLine 1 3) let sig' = setEntryDP sig (DifferentLine 2 3) let doAddMethod = do - [cd] <- hsDecls lp - (f1:f2s:f2d:_) <- hsDecls cd - let f2s' = setEntryDP f2s (DifferentLine 2 3) - cd' <- replaceDecls cd [f1, sig', decl', f2s', f2d] - replaceDecls lp [cd'] - - (lp',_,w) <- runTransformT doAddMethod + let + [cd] = hsDecls lp + (f1:f2s:f2d:_) = hsDecls cd + f2s' = setEntryDP f2s (DifferentLine 2 3) + cd' = replaceDecls cd [f1, sig', decl', f2s', f2d] + lp' = replaceDecls lp [cd'] + return lp' + + let (lp',_,w) = runTransform doAddMethod debugM $ "addClassMethod:" ++ intercalate "\n" w return lp' ===================================== utils/check-exact/Parsers.hs ===================================== @@ -260,7 +260,7 @@ parseModuleEpAnnsWithCppInternal cppOptions dflags file = do GHC.PFailed pst -> Left (GHC.GhcPsMessage <$> GHC.getPsErrorMessages pst) GHC.POk _ pmod - -> Right $ (injectedComments, dflags', fixModuleTrailingComments pmod) + -> Right $ (injectedComments, dflags', fixModuleComments pmod) -- | Internal function. Exposed if you want to muck with DynFlags -- before parsing. Or after parsing. @@ -269,8 +269,10 @@ postParseTransform -> Either a (GHC.ParsedSource) postParseTransform parseRes = fmap mkAnns parseRes where - -- TODO:AZ perhaps inject the comments into the parsedsource here already - mkAnns (_cs, _, m) = fixModuleTrailingComments m + mkAnns (_cs, _, m) = fixModuleComments m + +fixModuleComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleComments p = fixModuleHeaderComments $ fixModuleTrailingComments p fixModuleTrailingComments :: GHC.ParsedSource -> GHC.ParsedSource fixModuleTrailingComments (GHC.L l p) = GHC.L l p' @@ -293,6 +295,47 @@ fixModuleTrailingComments (GHC.L l p) = GHC.L l p' in cs'' _ -> cs +-- Deal with https://gitlab.haskell.org/ghc/ghc/-/issues/23984 +-- The Lexer works bottom-up, so does not have module declaration info +-- when the first top decl processed +fixModuleHeaderComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleHeaderComments (GHC.L l p) = GHC.L l p' + where + moveComments :: GHC.EpaLocation -> GHC.LHsDecl GHC.GhcPs -> GHC.EpAnnComments + -> (GHC.LHsDecl GHC.GhcPs, GHC.EpAnnComments) + moveComments GHC.EpaDelta{} dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.UnhelpfulSpan _)) dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.RealSrcSpan r _)) (GHC.L (GHC.EpAnn anc an csd) a) cs = (dd,css) + where + -- Move any comments on the decl that occur prior to the location + pc = GHC.priorComments csd + fc = GHC.getFollowingComments csd + bf (GHC.L anch _) = GHC.anchor anch > r + (move,keep) = break bf pc + csd' = GHC.EpaCommentsBalanced keep fc + + dd = GHC.L (GHC.EpAnn anc an csd') a + css = cs <> GHC.EpaComments move + + (ds',an') = rebalance (GHC.hsmodDecls p, GHC.hsmodAnn $ GHC.hsmodExt p) + p' = p { GHC.hsmodExt = (GHC.hsmodExt p){ GHC.hsmodAnn = an' }, + GHC.hsmodDecls = ds' + } + + rebalance :: ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + -> ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + rebalance (ds, GHC.EpAnn a an cs) = (ds1, GHC.EpAnn a an cs') + where + (ds1,cs') = case break (\(GHC.AddEpAnn k _) -> k == GHC.AnnWhere) (GHC.am_main an) of + (_, (GHC.AddEpAnn _ whereLoc:_)) -> + case GHC.hsmodDecls p of + (d:ds0) -> (d':ds0, cs0) + where (d',cs0) = moveComments whereLoc d cs + ds0 -> (ds0,cs) + _ -> (ds,cs) + + + -- | Internal function. Initializes DynFlags value for parsing. -- -- Passes "-hide-all-packages" to the GHC API to prevent parsing of ===================================== utils/check-exact/Transform.hs ===================================== @@ -63,7 +63,7 @@ module Transform -- *** Low level operations used in 'HasDecls' , balanceComments , balanceCommentsList - , balanceCommentsList' + , balanceCommentsListA , anchorEof -- ** Managing lists, pure functions @@ -92,6 +92,7 @@ import qualified Control.Monad.Fail as Fail import GHC hiding (parseModule, parsedSource) import GHC.Data.FastString +import GHC.Types.SrcLoc import Data.Data import Data.Maybe @@ -154,6 +155,7 @@ logDataWithAnnsTr str ast = do -- |If we need to add new elements to the AST, they need their own -- 'SrcSpan' for this. +-- This should no longer be needed, we use an @EpaDelta@ location instead. uniqueSrcSpanT :: (Monad m) => TransformT m SrcSpan uniqueSrcSpanT = do col <- get @@ -171,15 +173,6 @@ srcSpanStartLine' _ = 0 -- --------------------------------------------------------------------- -captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag -captureOrderBinds ls = AnnSortKey $ map go ls - where - go (L _ (ValD _ _)) = BindTag - go (L _ (SigD _ _)) = SigDTag - go d = error $ "captureOrderBinds:" ++ showGhc d - --- --------------------------------------------------------------------- - captureMatchLineSpacing :: LHsDecl GhcPs -> LHsDecl GhcPs captureMatchLineSpacing (L l (ValD x (FunBind a b (MG c (L d ms ))))) = L l (ValD x (FunBind a b (MG c (L d ms')))) @@ -253,7 +246,7 @@ setEntryDPDecl d dp = setEntryDP d dp -- |Set the true entry 'DeltaPos' from the annotation for a given AST -- element. This is the 'DeltaPos' ignoring any comments. -setEntryDP :: NoAnn t => LocatedAn t a -> DeltaPos -> LocatedAn t a +setEntryDP :: LocatedAn t a -> DeltaPos -> LocatedAn t a setEntryDP (L (EpAnn (EpaSpan ss@(UnhelpfulSpan _)) an cs) a) dp = L (EpAnn (EpaDelta ss dp []) an cs) a setEntryDP (L (EpAnn (EpaSpan ss) an (EpaComments [])) a) dp @@ -293,7 +286,7 @@ setEntryDP (L (EpAnn (EpaSpan ss@(RealSrcSpan r _)) an cs) a) dp L (EpAnn (EpaDelta ss edp csd) an cs'') a where cs'' = setPriorComments cs [] - csd = L (EpaDelta ss dp NoComments) c:cs' + csd = L (EpaDelta ss dp NoComments) c:commentOrigDeltas cs' lc = last $ (L ca c:cs') delta = case getLoc lc of EpaSpan (RealSrcSpan rr _) -> ss2delta (ss2pos rr) r @@ -335,18 +328,15 @@ setEntryDPFromAnchor off (EpaSpan (RealSrcSpan anc _)) ll@(L la _) = setEntryDP -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -transferEntryDP :: (Monad m, NoAnn t2, Typeable t1, Typeable t2) - => LocatedAn t1 a -> LocatedAn t2 b -> TransformT m (LocatedAn t2 b) -transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = do - logTr $ "transferEntryDP': EpAnn,EpAnn" +transferEntryDP :: (Typeable t1, Typeable t2) + => LocatedAn t1 a -> LocatedAn t2 b -> (LocatedAn t2 b) +transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = -- Problem: if the original had preceding comments, blindly -- transferring the location is not correct case priorComments cs1 of - [] -> return (L (EpAnn anc1 (combine an1 an2) cs2) b) + [] -> (L (EpAnn anc1 (combine an1 an2) cs2) b) -- TODO: what happens if the receiving side already has comments? - (L anc _:_) -> do - logDataWithAnnsTr "transferEntryDP':priorComments anc=" anc - return (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) + (L _ _:_) -> (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) -- |If a and b are the same type return first arg, else return second @@ -356,10 +346,11 @@ combine x y = fromMaybe y (cast x) -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -- TODO: call transferEntryDP, and use pushDeclDP -transferEntryDP' :: (Monad m) => LHsDecl GhcPs -> LHsDecl GhcPs -> TransformT m (LHsDecl GhcPs) -transferEntryDP' la lb = do - (L l2 b) <- transferEntryDP la lb - return (L l2 (pushDeclDP b (SameLine 0))) +transferEntryDP' :: LHsDecl GhcPs -> LHsDecl GhcPs -> (LHsDecl GhcPs) +transferEntryDP' la lb = + let + (L l2 b) = transferEntryDP la lb + in (L l2 (pushDeclDP b (SameLine 0))) pushDeclDP :: HsDecl GhcPs -> DeltaPos -> HsDecl GhcPs @@ -375,13 +366,24 @@ pushDeclDP d _dp = d -- --------------------------------------------------------------------- -balanceCommentsList :: (Monad m) => [LHsDecl GhcPs] -> TransformT m [LHsDecl GhcPs] -balanceCommentsList [] = return [] -balanceCommentsList [x] = return [x] -balanceCommentsList (a:b:ls) = do - (a',b') <- balanceComments a b - r <- balanceCommentsList (b':ls) - return (a':r) +-- | If we compile in haddock mode, the haddock processing inserts +-- DocDecls to carry the Haddock Documentation. We ignore these in +-- exact printing, as all the comments are also available in their +-- normal location, and the haddock processing is lossy, in that it +-- does not preserve all haddock-like comments. When we balance +-- comments in a list, we migrate some to preceding or following +-- declarations in the list. We must make sure we do not move any to +-- these DocDecls, which are not printed. +balanceCommentsList :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList decls = balanceCommentsList' (filter notDocDecl decls) + +balanceCommentsList' :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList' [] = [] +balanceCommentsList' [x] = [x] +balanceCommentsList' (a:b:ls) = (a':r) + where + (a',b') = balanceComments a b + r = balanceCommentsList' (b':ls) -- |The GHC parser puts all comments appearing between the end of one AST -- item and the beginning of the next as 'annPriorComments' for the second one. @@ -389,28 +391,27 @@ balanceCommentsList (a:b:ls) = do -- from the second one to the 'annFollowingComments' of the first if they belong -- to it instead. This is typically required before deleting or duplicating -- either of the AST elements. -balanceComments :: (Monad m) - => LHsDecl GhcPs -> LHsDecl GhcPs - -> TransformT m (LHsDecl GhcPs, LHsDecl GhcPs) -balanceComments first second = do +balanceComments :: LHsDecl GhcPs -> LHsDecl GhcPs + -> (LHsDecl GhcPs, LHsDecl GhcPs) +balanceComments first second = case first of - (L l (ValD x fb@(FunBind{}))) -> do - (L l' fb',second') <- balanceCommentsFB (L l fb) second - return (L l' (ValD x fb'), second') - _ -> balanceComments' first second + (L l (ValD x fb@(FunBind{}))) -> + let + (L l' fb',second') = balanceCommentsFB (L l fb) second + in (L l' (ValD x fb'), second') + _ -> balanceCommentsA first second --- |Once 'balanceComments' has been called to move trailing comments to a +-- |Once 'balanceCommentsA has been called to move trailing comments to a -- 'FunBind', these need to be pushed down from the top level to the last -- 'Match' if that 'Match' needs to be manipulated. -balanceCommentsFB :: (Monad m) - => LHsBind GhcPs -> LocatedA b -> TransformT m (LHsBind GhcPs, LocatedA b) -balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do - debugM $ "balanceCommentsFB entered: " ++ showGhc (ss2range $ locA lf) +balanceCommentsFB :: LHsBind GhcPs -> LocatedA b -> (LHsBind GhcPs, LocatedA b) +balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second + = balanceCommentsA (packFunBind bind) second' -- There are comments on lf. We need to -- + Keep the prior ones here -- + move the interior ones to the first match, -- + move the trailing ones to the last match. - let + where (before,middle,after) = case entry lf of EpaSpan (RealSrcSpan ss _) -> let @@ -426,40 +427,29 @@ balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do getFollowingComments $ comments lf) lf' = setCommentsEpAnn lf (EpaComments before) - debugM $ "balanceCommentsFB (before, after): " ++ showAst (before, after) - debugM $ "balanceCommentsFB lf': " ++ showAst lf' - -- let matches' = case matches of - let matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] - matches' = case matches of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaComments middle )) m':ms') - _ -> error "balanceCommentsFB" - matches'' <- balanceCommentsList' matches' - let (m,ms) = case reverse matches'' of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m',ms') - -- (L (addCommentsToEpAnnS lm' (EpaCommentsBalanced [] after)) m',ms') - _ -> error "balanceCommentsFB4" - debugM $ "balanceCommentsFB: (m,ms):" ++ showAst (m,ms) - (m',second') <- balanceComments' m second - m'' <- balanceCommentsMatch m' - let (m''',lf'') = case ms of - [] -> moveLeadingComments m'' lf' - _ -> (m'',lf') - debugM $ "balanceCommentsFB: (lf'', m'''):" ++ showAst (lf'',m''') - debugM $ "balanceCommentsFB done" - let bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) - debugM $ "balanceCommentsFB returning:" ++ showAst bind - balanceComments' (packFunBind bind) second' -balanceCommentsFB f s = balanceComments' f s + matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] + matches' = case matches of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaComments middle )) m0:ms') + _ -> error "balanceCommentsFB" + matches'' = balanceCommentsListA matches' + (m,ms) = case reverse matches'' of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m0,ms') + _ -> error "balanceCommentsFB4" + (m',second') = balanceCommentsA m second + m'' = balanceCommentsMatch m' + (m''',lf'') = case ms of + [] -> moveLeadingComments m'' lf' + _ -> (m'',lf') + bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) +balanceCommentsFB f s = balanceCommentsA f s -- | Move comments on the same line as the end of the match into the -- GRHS, prior to the binds -balanceCommentsMatch :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do - logTr $ "balanceCommentsMatch: (logInfo)=" ++ showAst (logInfo) - return (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) +balanceCommentsMatch :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) + = (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) where simpleBreak (r,_) = r /= 0 an1 = l @@ -468,7 +458,7 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do (move',stay') = break simpleBreak (trailingCommentsDeltas (anchorFromLocatedA (L l ())) cs1f) move = map snd move' stay = map snd stay' - (l'', grhss', binds', logInfo) + (l'', grhss', binds', _logInfo) = case reverse grhss of [] -> (l, [], binds, (EpaComments [], noSrcSpanA)) (L lg (GRHS ag grs rhs):gs) -> @@ -491,26 +481,24 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do pushTrailingComments :: WithWhere -> EpAnnComments -> HsLocalBinds GhcPs -> (Bool, HsLocalBinds GhcPs) pushTrailingComments _ _cs b at EmptyLocalBinds{} = (False, b) pushTrailingComments _ _cs (HsIPBinds _ _) = error "TODO: pushTrailingComments:HsIPBinds" -pushTrailingComments w cs lb@(HsValBinds an _) - = (True, HsValBinds an' vb) +pushTrailingComments w cs lb@(HsValBinds an _) = (True, HsValBinds an' vb) where decls = hsDeclsLocalBinds lb (an', decls') = case reverse decls of [] -> (addCommentsToEpAnn an cs, decls) (L la d:ds) -> (an, L (addCommentsToEpAnn la cs) d:ds) - (vb,_ws2) = case runTransform (replaceDeclsValbinds w lb (reverse decls')) of - ((HsValBinds _ vb'), _, ws2') -> (vb', ws2') - _ -> (ValBinds NoAnnSortKey [] [], []) + vb = case replaceDeclsValbinds w lb (reverse decls') of + (HsValBinds _ vb') -> vb' + _ -> ValBinds NoAnnSortKey [] [] -balanceCommentsList' :: (Monad m) => [LocatedA a] -> TransformT m [LocatedA a] -balanceCommentsList' [] = return [] -balanceCommentsList' [x] = return [x] -balanceCommentsList' (a:b:ls) = do - logTr $ "balanceCommentsList' entered" - (a',b') <- balanceComments' a b - r <- balanceCommentsList' (b':ls) - return (a':r) +balanceCommentsListA :: [LocatedA a] -> [LocatedA a] +balanceCommentsListA [] = [] +balanceCommentsListA [x] = [x] +balanceCommentsListA (a:b:ls) = (a':r) + where + (a',b') = balanceCommentsA a b + r = balanceCommentsListA (b':ls) -- |Prior to moving an AST element, make sure any trailing comments belonging to -- it are attached to it, and not the following element. Of necessity this is a @@ -518,13 +506,8 @@ balanceCommentsList' (a:b:ls) = do -- with a passed-in decision function. -- The initial situation is that all comments for a given anchor appear as prior comments -- Many of these should in fact be following comments for the previous anchor -balanceComments' :: (Monad m) => LocatedA a -> LocatedA b -> TransformT m (LocatedA a, LocatedA b) -balanceComments' la1 la2 = do - debugM $ "balanceComments': (anc1)=" ++ showAst (anc1) - debugM $ "balanceComments': (cs1s)=" ++ showAst (cs1s) - debugM $ "balanceComments': (cs1stay,cs1move)=" ++ showAst (cs1stay,cs1move) - debugM $ "balanceComments': (an1',an2')=" ++ showAst (an1',an2') - return (la1', la2') +balanceCommentsA :: LocatedA a -> LocatedA b -> (LocatedA a, LocatedA b) +balanceCommentsA la1 la2 = (la1', la2') where simpleBreak n (r,_) = r > n L an1 f = la1 @@ -532,26 +515,31 @@ balanceComments' la1 la2 = do anc1 = comments an1 anc2 = comments an2 - cs1s = splitCommentsEnd (anchorFromLocatedA la1) anc1 - cs1p = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1s) - cs1f = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1s) + (p1,m1,f1) = splitComments (anchorFromLocatedA la1) anc1 + cs1p = priorCommentsDeltas (anchorFromLocatedA la1) p1 - cs2s = splitCommentsEnd (anchorFromLocatedA la2) anc2 - cs2p = priorCommentsDeltas (anchorFromLocatedA la2) (priorComments cs2s) - cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) (getFollowingComments cs2s) + -- Split cs1 following comments into those before any + -- TrailingAnn's on an1, and any after + cs1f = splitCommentsEnd (fullSpanFromLocatedA la1) $ EpaComments f1 + cs1fp = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1f) + cs1ff = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1f) - -- Split cs1f into those that belong on an1 and ones that must move to an2 - (cs1move,cs1stay) = break (simpleBreak 1) cs1f + -- Split cs1ff into those that belong on an1 and ones that must move to an2 + (cs1move,cs1stay) = break (simpleBreak 1) cs1ff + + (p2,m2,f2) = splitComments (anchorFromLocatedA la2) anc2 + cs2p = priorCommentsDeltas (anchorFromLocatedA la2) p2 + cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) f2 (stay'',move') = break (simpleBreak 1) cs2p -- Need to also check for comments more closely attached to la1, -- ie trailing on the same line (move'',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchorFromLocatedA la1) (map snd stay'')) - move = sortEpaComments $ map snd (cs1move ++ move'' ++ move') - stay = sortEpaComments $ map snd (cs1stay ++ stay') + move = sortEpaComments $ map snd (cs1fp ++ cs1move ++ move'' ++ move') + stay = sortEpaComments $ m2 ++ map snd (cs1stay ++ stay') - an1' = setCommentsEpAnn (getLoc la1) (EpaCommentsBalanced (map snd cs1p) move) - an2' = setCommentsEpAnn (getLoc la2) (EpaCommentsBalanced stay (map snd cs2f)) + an1' = setCommentsEpAnn (getLoc la1) (epaCommentsBalanced (m1 ++ map snd cs1p) move) + an2' = setCommentsEpAnn (getLoc la2) (epaCommentsBalanced stay (map snd cs2f)) la1' = L an1' f la2' = L an2' s @@ -569,10 +557,9 @@ trailingCommentsDeltas r (la@(L l _):las) (al,_) = ss2posEnd rs' (ll,_) = ss2pos (anchor loc) --- AZ:TODO: this is identical to commentsDeltas priorCommentsDeltas :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] -priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) +priorCommentsDeltas r cs = go r (sortEpaComments cs) where go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] go _ [] = [] @@ -588,6 +575,21 @@ priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) -- --------------------------------------------------------------------- +-- | Split comments into ones occurring before the end of the reference +-- span, and those after it. +splitComments :: RealSrcSpan -> EpAnnComments -> ([LEpaComment], [LEpaComment], [LEpaComment]) +splitComments p cs = (before, middle, after) + where + cmpe (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p + cmpe (L _ _) = True + + cmpb (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2pos p + cmpb (L _ _) = True + + (beforeEnd, after) = break cmpe ((priorComments cs) ++ (getFollowingComments cs)) + (before, middle) = break cmpb beforeEnd + + -- | Split comments into ones occurring before the end of the reference -- span, and those after it. splitCommentsEnd :: RealSrcSpan -> EpAnnComments -> EpAnnComments @@ -598,8 +600,8 @@ splitCommentsEnd p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsEnd p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsEnd p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -617,8 +619,8 @@ splitCommentsStart p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsStart p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsStart p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -638,8 +640,8 @@ moveLeadingComments (L la a) lb = (L la' a, lb') -- TODO: need to set an entry delta on lb' to zero, and move the -- original spacing to the first comment. - la' = setCommentsEpAnn la (EpaCommentsBalanced [] after) - lb' = addCommentsToEpAnn lb (EpaCommentsBalanced before []) + la' = setCommentsEpAnn la (epaCommentsBalanced [] after) + lb' = addCommentsToEpAnn lb (epaCommentsBalanced before []) -- | A GHC comment includes the span of the preceding (non-comment) -- token. Takes an original list of comments, and converts the @@ -662,17 +664,27 @@ addCommentOrigDeltasAnn (EpAnn e a cs) = EpAnn e a (addCommentOrigDeltas cs) anchorFromLocatedA :: LocatedA a -> RealSrcSpan anchorFromLocatedA (L (EpAnn anc _ _) _) = anchor anc +-- | Get the full span of interest for comments from a LocatedA. +-- This extends up to the last TrailingAnn +fullSpanFromLocatedA :: LocatedA a -> RealSrcSpan +fullSpanFromLocatedA (L (EpAnn anc (AnnListItem tas) _) _) = rr + where + r = anchor anc + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + rr = case reverse (concatMap trailing_loc tas) of + [] -> r + (s:_) -> combineRealSrcSpans r s + -- --------------------------------------------------------------------- -balanceSameLineComments :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do - logTr $ "balanceSameLineComments: (la)=" ++ showGhc (ss2range $ locA la) - logTr $ "balanceSameLineComments: [logInfo]=" ++ showAst logInfo - return (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) +balanceSameLineComments :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) + = (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) where simpleBreak n (r,_) = r > n - (la',grhss', logInfo) = case reverse grhss of + (la',grhss', _logInfo) = case reverse grhss of [] -> (la,grhss,[]) (L lg (GRHS ga gs rhs):grs) -> (la'',reverse $ (L lg (GRHS ga' gs rhs)):grs,[(gac,(csp,csf))]) where @@ -684,7 +696,7 @@ balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do (move',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchor anc) csf) move = map snd move' stay = map snd stay' - cs1 = EpaCommentsBalanced csp stay + cs1 = epaCommentsBalanced csp stay gac = epAnnComments ga gfc = getFollowingComments gac @@ -734,24 +746,21 @@ addComma (EpAnn anc (AnnListItem as) cs) -- | Insert a declaration into an AST element having sub-declarations -- (@HasDecls@) according to the given location function. insertAt :: (HasDecls ast) - => (LHsDecl GhcPs - -> [LHsDecl GhcPs] - -> [LHsDecl GhcPs]) - -> ast - -> LHsDecl GhcPs - -> Transform ast -insertAt f t decl = do - oldDecls <- hsDecls t - oldDeclsb <- balanceCommentsList oldDecls - let oldDecls' = oldDeclsb - replaceDecls t (f decl oldDecls') + => (LHsDecl GhcPs + -> [LHsDecl GhcPs] + -> [LHsDecl GhcPs]) + -> ast + -> LHsDecl GhcPs + -> ast +insertAt f t decl = replaceDecls t (f decl oldDecls') + where + oldDecls = hsDecls t + oldDeclsb = balanceCommentsList oldDecls + oldDecls' = oldDeclsb -- |Insert a declaration at the beginning or end of the subdecls of the given -- AST item -insertAtStart, insertAtEnd :: (HasDecls ast) - => ast - -> LHsDecl GhcPs - -> Transform ast +insertAtStart, insertAtEnd :: HasDecls ast => ast -> LHsDecl GhcPs -> ast insertAtEnd = insertAt (\x xs -> xs ++ [x]) @@ -766,11 +775,11 @@ insertAtStart = insertAt insertFirst -- |Insert a declaration at a specific location in the subdecls of the given -- AST item -insertAfter, insertBefore :: (HasDecls (LocatedA ast)) +insertAfter, insertBefore :: HasDecls (LocatedA ast) => LocatedA old -> LocatedA ast -> LHsDecl GhcPs - -> Transform (LocatedA ast) + -> LocatedA ast insertAfter (getLocA -> k) = insertAt findAfter where findAfter x xs = @@ -797,10 +806,10 @@ class (Data t) => HasDecls t where -- given syntax phrase. They are always returned in the wrapped 'HsDecl' -- form, even if orginating in local decls. This is safe, as annotations -- never attach to the wrapper, only to the wrapped item. - hsDecls :: (Monad m) => t -> TransformT m [LHsDecl GhcPs] + hsDecls :: t -> [LHsDecl GhcPs] -- | Replace the directly enclosed decl list by the given - -- decl list. Runs in the 'Transform' monad to be able to update list order + -- decl list. As part of replacing it will update list order -- annotations, and rebalance comments and other layout changes as needed. -- -- For example, a call on replaceDecls for a wrapped 'FunBind' having no @@ -818,96 +827,86 @@ class (Data t) => HasDecls t where -- where -- nn = 2 -- @ - replaceDecls :: (Monad m) => t -> [LHsDecl GhcPs] -> TransformT m t + replaceDecls :: t -> [LHsDecl GhcPs] -> t -- --------------------------------------------------------------------- instance HasDecls ParsedSource where - hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = return decls + hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = decls replaceDecls (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps _decls)) decls - = do - logTr "replaceDecls LHsModule" - return (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) + = (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsDecl GhcPs)) where - hsDecls (L _ (TyClD _ c at ClassDecl{})) = return $ hsDeclsClassDecl c - hsDecls decl = do - error $ "hsDecls:decl=" ++ showAst decl - replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = do - let decl' = replaceDeclsClassDecl dec decls - return (L l (TyClD e decl')) - replaceDecls decl _decls = do - error $ "replaceDecls:decl=" ++ showAst decl + hsDecls (L _ (TyClD _ c at ClassDecl{})) = hsDeclsClassDecl c + hsDecls decl = error $ "hsDecls:decl=" ++ showAst decl + replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = + let + decl' = replaceDeclsClassDecl dec decls + in (L l (TyClD e decl')) + replaceDecls decl _decls + = error $ "replaceDecls:decl=" ++ showAst decl -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Match GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = hsDeclsLocalBinds lb replaceDecls (L l (Match xm c p (GRHSs xr rhs binds))) [] - = do - logTr "replaceDecls LMatch empty decls" - binds'' <- replaceDeclsValbinds WithoutWhere binds [] - return (L l (Match xm c p (GRHSs xr rhs binds''))) + = let + binds'' = replaceDeclsValbinds WithoutWhere binds [] + in (L l (Match xm c p (GRHSs xr rhs binds''))) replaceDecls m@(L l (Match xm c p (GRHSs xr rhs binds))) newBinds - = do - logTr "replaceDecls LMatch nonempty decls" + = let -- Need to throw in a fresh where clause if the binds were empty, -- in the annotations. - (l', rhs') <- case binds of - EmptyLocalBinds{} -> do - logTr $ "replaceDecls LMatch empty binds" - - logDataWithAnnsTr "Match.replaceDecls:balancing comments:m" m - L l' m' <- balanceSameLineComments m - logDataWithAnnsTr "Match.replaceDecls:(m1')" (L l' m') - return (l', grhssGRHSs $ m_grhss m') - _ -> return (l, rhs) - binds'' <- replaceDeclsValbinds WithWhere binds newBinds - logDataWithAnnsTr "Match.replaceDecls:binds'" binds'' - return (L l' (Match xm c p (GRHSs xr rhs' binds''))) + (l', rhs') = case binds of + EmptyLocalBinds{} -> + let + L l0 m' = balanceSameLineComments m + in (l0, grhssGRHSs $ m_grhss m') + _ -> (l, rhs) + binds'' = replaceDeclsValbinds WithWhere binds newBinds + in (L l' (Match xm c p (GRHSs xr rhs' binds''))) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsExpr GhcPs)) where - hsDecls (L _ (HsLet _ decls _ex)) = return $ hsDeclsLocalBinds decls - hsDecls _ = return [] + hsDecls (L _ (HsLet _ decls _ex)) = hsDeclsLocalBinds decls + hsDecls _ = [] replaceDecls (L ll (HsLet (tkLet, tkIn) binds ex)) newDecls - = do - logTr "replaceDecls HsLet" - let lastAnc = realSrcSpan $ spanHsLocaLBinds binds + = let + lastAnc = realSrcSpan $ spanHsLocaLBinds binds -- TODO: may be an intervening comment, take account for lastAnc - let (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of - (EpTok l, EpTok i) -> - let - off = case l of - (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r - (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 - (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 - (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c - ex'' = setEntryDPFromAnchor off i ex - newDecls'' = case newDecls of - [] -> newDecls - (d:ds) -> setEntryDPDecl d (SameLine 0) : ds - in ( EpTok l - , EpTok (addEpaLocationDelta off lastAnc i) - , ex'' - , newDecls'') - (_,_) -> (tkLet, tkIn, ex, newDecls) - binds' <- replaceDeclsValbinds WithoutWhere binds newDecls' - return (L ll (HsLet (tkLet', tkIn') binds' ex')) + (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of + (EpTok l, EpTok i) -> + let + off = case l of + (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r + (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 + (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 + (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c + ex'' = setEntryDPFromAnchor off i ex + newDecls'' = case newDecls of + [] -> newDecls + (d:ds) -> setEntryDPDecl d (SameLine 0) : ds + in ( EpTok l + , EpTok (addEpaLocationDelta off lastAnc i) + , ex'' + , newDecls'') + (_,_) -> (tkLet, tkIn, ex, newDecls) + binds' = replaceDeclsValbinds WithoutWhere binds newDecls' + in (L ll (HsLet (tkLet', tkIn') binds' ex')) -- TODO: does this make sense? Especially as no hsDecls for HsPar replaceDecls (L l (HsPar x e)) newDecls - = do - logTr "replaceDecls HsPar" - e' <- replaceDecls e newDecls - return (L l (HsPar x e')) + = let + e' = replaceDecls e newDecls + in (L l (HsPar x e')) replaceDecls old _new = error $ "replaceDecls (LHsExpr GhcPs) undefined for:" ++ showGhc old -- --------------------------------------------------------------------- @@ -934,53 +933,51 @@ hsDeclsPatBind x = error $ "hsDeclsPatBind called for:" ++ showGhc x -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBindD' \/ 'replaceDeclsPatBindD' is -- idempotent. -replaceDeclsPatBindD :: (Monad m) => LHsDecl GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsDecl GhcPs) -replaceDeclsPatBindD (L l (ValD x d)) newDecls = do - (L _ d') <- replaceDeclsPatBind (L l d) newDecls - return (L l (ValD x d')) +replaceDeclsPatBindD :: LHsDecl GhcPs -> [LHsDecl GhcPs] -> (LHsDecl GhcPs) +replaceDeclsPatBindD (L l (ValD x d)) newDecls = + let + (L _ d') = replaceDeclsPatBind (L l d) newDecls + in (L l (ValD x d')) replaceDeclsPatBindD x _ = error $ "replaceDeclsPatBindD called for:" ++ showGhc x -- | Replace the immediate declarations for a 'PatBind'. This -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBind' \/ 'replaceDeclsPatBind' is -- idempotent. -replaceDeclsPatBind :: (Monad m) => LHsBind GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsBind GhcPs) +replaceDeclsPatBind :: LHsBind GhcPs -> [LHsDecl GhcPs] -> (LHsBind GhcPs) replaceDeclsPatBind (L l (PatBind x a p (GRHSs xr rhss binds))) newDecls - = do - logTr "replaceDecls PatBind" - binds'' <- replaceDeclsValbinds WithWhere binds newDecls - return (L l (PatBind x a p (GRHSs xr rhss binds''))) + = (L l (PatBind x a p (GRHSs xr rhss binds''))) + where + binds'' = replaceDeclsValbinds WithWhere binds newDecls replaceDeclsPatBind x _ = error $ "replaceDeclsPatBind called for:" ++ showGhc x -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Stmt GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (LetStmt _ lb)) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (LetStmt _ lb)) = hsDeclsLocalBinds lb hsDecls (L _ (LastStmt _ e _ _)) = hsDecls e hsDecls (L _ (BindStmt _ _pat e)) = hsDecls e hsDecls (L _ (BodyStmt _ e _ _)) = hsDecls e - hsDecls _ = return [] + hsDecls _ = [] replaceDecls (L l (LetStmt x lb)) newDecls - = do - lb'' <- replaceDeclsValbinds WithWhere lb newDecls - return (L l (LetStmt x lb'')) + = let + lb'' = replaceDeclsValbinds WithWhere lb newDecls + in (L l (LetStmt x lb'')) replaceDecls (L l (LastStmt x e d se)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (LastStmt x e' d se)) + = let + e' = replaceDecls e newDecls + in (L l (LastStmt x e' d se)) replaceDecls (L l (BindStmt x pat e)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BindStmt x pat e')) + = let + e' = replaceDecls e newDecls + in (L l (BindStmt x pat e')) replaceDecls (L l (BodyStmt x e a b)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BodyStmt x e' a b)) - replaceDecls x _newDecls = return x + = let + e' = replaceDecls e newDecls + in (L l (BodyStmt x e' a b)) + replaceDecls x _newDecls = x -- ===================================================================== -- end of HasDecls instances @@ -1062,61 +1059,55 @@ data WithWhere = WithWhere -- care, as this does not manage the declaration order, the -- ordering should be done by the calling function from the 'HsLocalBinds' -- context in the AST. -replaceDeclsValbinds :: (Monad m) - => WithWhere +replaceDeclsValbinds :: WithWhere -> HsLocalBinds GhcPs -> [LHsDecl GhcPs] - -> TransformT m (HsLocalBinds GhcPs) -replaceDeclsValbinds _ _ [] = do - return (EmptyLocalBinds NoExtField) + -> HsLocalBinds GhcPs +replaceDeclsValbinds _ _ [] = EmptyLocalBinds NoExtField replaceDeclsValbinds w b@(HsValBinds a _) new - = do - logTr "replaceDeclsValbinds" - let oldSpan = spanHsLocaLBinds b - an <- oldWhereAnnotation a w (realSrcSpan oldSpan) - let decs = concatMap decl2Bind new - let sigs = concatMap decl2Sig new - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) + = let + oldSpan = spanHsLocaLBinds b + an = oldWhereAnnotation a w (realSrcSpan oldSpan) + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) replaceDeclsValbinds _ (HsIPBinds {}) _new = error "undefined replaceDecls HsIPBinds" replaceDeclsValbinds w (EmptyLocalBinds _) new - = do - logTr "replaceDecls HsLocalBinds" - an <- newWhereAnnotation w - let newBinds = concatMap decl2Bind new - newSigs = concatMap decl2Sig new - let decs = newBinds - let sigs = newSigs - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) - -oldWhereAnnotation :: (Monad m) - => EpAnn AnnList -> WithWhere -> RealSrcSpan -> TransformT m (EpAnn AnnList) -oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = do - -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, change the AnnList anchor to have the correct DP too - let (AnnList ancl o c _r t) = an - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - (anc', ancl') <- do - case ww of - WithWhere -> return (anc, ancl) - WithoutWhere -> return (anc, ancl) - let an' = EpAnn anc' - (AnnList ancl' o c w t) - cs - return an' - -newWhereAnnotation :: (Monad m) => WithWhere -> TransformT m (EpAnn AnnList) -newWhereAnnotation ww = do - let anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] - let anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - let an = EpAnn anc - (AnnList (Just anc2) Nothing Nothing w []) - emptyComments - return an + = let + an = newWhereAnnotation w + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) + +oldWhereAnnotation :: EpAnn AnnList -> WithWhere -> RealSrcSpan -> (EpAnn AnnList) +oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = an' + -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, + -- change the AnnList anchor to have the correct DP too + where + (AnnList ancl o c _r t) = an + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + (anc', ancl') = + case ww of + WithWhere -> (anc, ancl) + WithoutWhere -> (anc, ancl) + an' = EpAnn anc' + (AnnList ancl' o c w t) + cs + +newWhereAnnotation :: WithWhere -> (EpAnn AnnList) +newWhereAnnotation ww = an + where + anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] + anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + an = EpAnn anc + (AnnList (Just anc2) Nothing Nothing w []) + emptyComments -- --------------------------------------------------------------------- @@ -1127,32 +1118,32 @@ type PMatch = LMatch GhcPs (LHsExpr GhcPs) -- declarations are extracted and returned after modification. For a -- 'FunBind' the supplied 'SrcSpan' is used to identify the specific -- 'Match' to be transformed, for when there are multiple of them. -modifyValD :: forall m t. (HasTransform m) - => SrcSpan +modifyValD :: forall t. + SrcSpan -> Decl - -> (PMatch -> [Decl] -> m ([Decl], Maybe t)) - -> m (Decl,Maybe t) + -> (PMatch -> [Decl] -> ([Decl], Maybe t)) + -> (Decl,Maybe t) modifyValD p pb@(L ss (ValD _ (PatBind {} ))) f = if (locA ss) == p - then do - let ds = hsDeclsPatBindD pb - (ds',r) <- f (error "modifyValD.PatBind should not touch Match") ds - pb' <- liftT $ replaceDeclsPatBindD pb ds' - return (pb',r) - else return (pb,Nothing) -modifyValD p decl f = do - (decl',r) <- runStateT (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing - return (packFunDecl decl',r) + then + let + ds = hsDeclsPatBindD pb + (ds',r) = f (error "modifyValD.PatBind should not touch Match") ds + pb' = replaceDeclsPatBindD pb ds' + in (pb',r) + else (pb,Nothing) +modifyValD p decl f = (packFunDecl decl', r) where - doModLocal :: PMatch -> StateT (Maybe t) m PMatch + (decl',r) = runState (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing + doModLocal :: PMatch -> State (Maybe t) PMatch doModLocal (match@(L ss _) :: PMatch) = do if (locA ss) == p then do - ds <- lift $ liftT $ hsDecls match - `debug` ("modifyValD: match=" ++ showAst match) - (ds',r) <- lift $ f match ds - put r - match' <- lift $ liftT $ replaceDecls match ds' + let + ds = hsDecls match + (ds',r0) = f match ds + put r0 + let match' = replaceDecls match ds' return match' else return match @@ -1172,6 +1163,6 @@ modifyDeclsT :: (HasDecls t,HasTransform m) => ([LHsDecl GhcPs] -> m [LHsDecl GhcPs]) -> t -> m t modifyDeclsT action t = do - decls <- liftT $ hsDecls t + let decls = hsDecls t decls' <- action decls - liftT $ replaceDecls t decls' + return $ replaceDecls t decls' ===================================== utils/check-exact/Types.hs ===================================== @@ -21,10 +21,6 @@ type Pos = (Int,Int) -- --------------------------------------------------------------------- -data Rigidity = NormalLayout | RigidLayout deriving (Eq, Ord, Show) - --- --------------------------------------------------------------------- - -- | A Haskell comment. The @AnnKeywordId@ is present if it has been converted -- from an @AnnKeywordId@ because the annotation must be interleaved into the -- stream and does not have a well-defined position ===================================== utils/check-exact/Utils.hs ===================================== @@ -20,9 +20,8 @@ module Utils where import Control.Monad (when) +import GHC.Utils.Monad.State.Strict import Data.Function -import Data.Maybe (isJust) -import Data.Ord (comparing) import GHC.Hs.Dump import Lookup @@ -36,9 +35,10 @@ import GHC.Driver.Ppr import GHC.Data.FastString import qualified GHC.Data.Strict as Strict import GHC.Base (NonEmpty(..)) +import GHC.Parser.Lexer (allocateComments) import Data.Data hiding ( Fixity ) -import Data.List (sortBy, elemIndex) +import Data.List (sortBy, partition) import qualified Data.Map.Strict as Map import Debug.Trace @@ -60,12 +60,32 @@ debug c s = if debugEnabledFlag debugM :: Monad m => String -> m () debugM s = when debugEnabledFlag $ traceM s --- --------------------------------------------------------------------- - warn :: c -> String -> c -- warn = flip trace warn c _ = c +-- --------------------------------------------------------------------- + +captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag +captureOrderBinds ls = AnnSortKey $ map go ls + where + go (L _ (ValD _ _)) = BindTag + go (L _ (SigD _ _)) = SigDTag + go d = error $ "captureOrderBinds:" ++ showGhc d + +-- --------------------------------------------------------------------- + +notDocDecl :: LHsDecl GhcPs -> Bool +notDocDecl (L _ DocD{}) = False +notDocDecl _ = True + +notIEDoc :: LIE GhcPs -> Bool +notIEDoc (L _ IEGroup {}) = False +notIEDoc (L _ IEDoc {}) = False +notIEDoc (L _ IEDocNamed {}) = False +notIEDoc _ = True + +-- --------------------------------------------------------------------- -- | A good delta has no negative values. isGoodDelta :: DeltaPos -> Bool isGoodDelta (SameLine co) = co >= 0 @@ -108,7 +128,6 @@ pos2delta (refl,refc) (l,c) = deltaPos lo co lo = l - refl co = if lo == 0 then c - refc else c - -- else c - 1 -- | Apply the delta to the current position, taking into account the -- current column offset if advancing to a new line @@ -200,23 +219,6 @@ origDelta pos pp = ss2delta (ss2posEnd pp) pos -- --------------------------------------------------------------------- --- |Given a list of items and a list of keys, returns a list of items --- ordered by their position in the list of keys. -orderByKey :: [(DeclTag,a)] -> [DeclTag] -> [(DeclTag,a)] -orderByKey keys order - -- AZ:TODO: if performance becomes a problem, consider a Map of the order - -- SrcSpan to an index, and do a lookup instead of elemIndex. - - -- Items not in the ordering are placed to the start - = sortBy (comparing (flip elemIndex order . fst)) keys - --- --------------------------------------------------------------------- - -isListComp :: HsDoFlavour -> Bool -isListComp = isDoComprehensionContext - --- --------------------------------------------------------------------- - needsWhere :: DataDefnCons (LConDecl (GhcPass p)) -> Bool needsWhere (NewTypeCon _) = True needsWhere (DataTypeCons _ []) = True @@ -225,21 +227,214 @@ needsWhere _ = False -- --------------------------------------------------------------------- +-- | Insert the comments at the appropriate places in the AST insertCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource -insertCppComments (L l p) cs = L l p' +-- insertCppComments p [] = p +insertCppComments (L l p) cs0 = insertRemainingCppComments (L l p2) remaining + where + (EpAnn anct ant cst) = hsmodAnn $ hsmodExt p + cs = sortEpaComments $ priorComments cst ++ getFollowingComments cst ++ cs0 + p0 = p { hsmodExt = (hsmodExt p) { hsmodAnn = EpAnn anct ant emptyComments }} + -- Comments embedded within spans + -- everywhereM is a bottom-up traversal + (p1, toplevel) = runState (everywhereM (mkM addCommentsListItem + `extM` addCommentsGrhs + `extM` addCommentsList) p0) cs + (p2, remaining) = insertTopLevelCppComments p1 toplevel + + addCommentsListItem :: EpAnn AnnListItem -> State [LEpaComment] (EpAnn AnnListItem) + addCommentsListItem = addComments + + addCommentsList :: EpAnn AnnList -> State [LEpaComment] (EpAnn AnnList) + addCommentsList = addComments + + addCommentsGrhs :: EpAnn GrhsAnn -> State [LEpaComment] (EpAnn GrhsAnn) + addCommentsGrhs = addComments + + addComments :: forall ann. EpAnn ann -> State [LEpaComment] (EpAnn ann) + addComments (EpAnn anc an ocs) = do + case anc of + EpaSpan (RealSrcSpan s _) -> do + unAllocated <- get + let + (rest, these) = GHC.Parser.Lexer.allocateComments s unAllocated + cs' = workInComments ocs these + put rest + return $ EpAnn anc an cs' + + _ -> return $ EpAnn anc an ocs + +workInComments :: EpAnnComments -> [LEpaComment] -> EpAnnComments +workInComments ocs [] = ocs +workInComments ocs new = cs' + where + pc = priorComments ocs + fc = getFollowingComments ocs + cs' = case fc of + [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ new + (L ac _:_) -> epaCommentsBalanced (sortEpaComments $ pc ++ cs_before) + (sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) + new + +insertTopLevelCppComments :: HsModule GhcPs -> [LEpaComment] -> (HsModule GhcPs, [LEpaComment]) +insertTopLevelCppComments (HsModule (XModulePs an lo mdeprec mbDoc) mmn mexports imports decls) cs + = (HsModule (XModulePs an4 lo mdeprec mbDoc) mmn mexports' imports' decls', cs3) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0,hc1,hc_cs)) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0i,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0i,hc0,hc1,hc_cs)) where - an' = case GHC.hsmodAnn $ GHC.hsmodExt p of - (EpAnn a an ocs) -> EpAnn a an cs' - where - pc = priorComments ocs - fc = getFollowingComments ocs - cs' = case fc of - [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ cs - (L ac _:_) -> EpaCommentsBalanced (sortEpaComments $ pc ++ cs_before) - (sortEpaComments $ fc ++ cs_after) - where - (cs_before,cs_after) = break (\(L ll _) -> (ss2pos $ anchor ll) < (ss2pos $ anchor ac) ) cs + -- Comments at the top level. + (an0, cs0) = + case mmn of + Nothing -> (an, cs) + Just _ -> + -- We have a module name. Capture all comments up to the `where` + let + (these, remaining) = splitOnWhere Before (am_main $ anns an) cs + (EpAnn a anno ocs) = an :: EpAnn AnnsModule + anm = EpAnn a anno (workInComments ocs these) + in + (anm, remaining) + (an1,cs0a) = case lo of + EpExplicitBraces (EpTok (EpaSpan (RealSrcSpan s _))) _close -> + let + (stay,cs0a') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0 + cs' = workInComments (comments an0) stay + in (an0 { comments = cs' }, cs0a') + _ -> (an0,cs0) + -- Deal with possible leading semis + (an2, cs0b) = case am_decls $ anns an1 of + (AddSemiAnn (EpaSpan (RealSrcSpan s _)):_) -> (an1 {comments = cs'}, cs0b') + where + (stay,cs0b') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0a + cs' = workInComments (comments an1) stay + _ -> (an1,cs0a) + + (mexports', an3, cs1) = + case mexports of + Nothing -> (Nothing, an2, cs0b) + Just (L l exports) -> (Just (L l exports'), an3', cse) + where + hc1' = workInComments (comments an2) csh' + an3' = an2 { comments = hc1' } + (csh', cs0b') = case al_open $ anns l of + Just (AddEpAnn _ (EpaSpan (RealSrcSpan s _))) ->(h, n) + where + (h,n) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + cs0b + + _ -> ([], cs0b) + (exports', cse) = allocPreceding exports cs0b' + (imports0, cs2) = allocPreceding imports cs1 + (imports', hc0i) = balanceFirstLocatedAComments imports0 + + (decls0, cs3) = allocPreceding decls cs2 + (decls', hc0d) = balanceFirstLocatedAComments decls0 + + -- Either hc0i or hc0d should have comments. Combine them + hc0 = hc0i ++ hc0d + + (hc1,hc_cs) = if null ( am_main $ anns an3) + then (hc0,[]) + else splitOnWhere After (am_main $ anns an3) hc0 + hc2 = workInComments (comments an3) hc1 + an4 = an3 { anns = (anns an3) {am_cs = hc_cs}, comments = hc2 } + + allocPreceding :: [LocatedA a] -> [LEpaComment] -> ([LocatedA a], [LEpaComment]) + allocPreceding [] cs' = ([], cs') + allocPreceding (L (EpAnn anc4 an5 cs4) a:xs) cs' = ((L (EpAnn anc4 an5 cs4') a:xs'), rest') + where + (rest, these) = + case anc4 of + EpaSpan (RealSrcSpan s _) -> + allocatePriorComments (ss2pos s) cs' + _ -> (cs', []) + cs4' = workInComments cs4 these + (xs',rest') = allocPreceding xs rest + +data SplitWhere = Before | After +splitOnWhere :: SplitWhere -> [AddEpAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitOnWhere _ [] csIn = (csIn,[]) +splitOnWhere w (AddEpAnn AnnWhere (EpaSpan (RealSrcSpan s _)):_) csIn = (hc, fc) + where + splitFunc Before anc_pos c_pos = c_pos < anc_pos + splitFunc After anc_pos c_pos = anc_pos < c_pos + (hc,fc) = break (\(L ll _) -> splitFunc w (ss2pos $ anchor ll) (ss2pos s)) csIn +splitOnWhere _ (AddEpAnn AnnWhere _:_) csIn = (csIn, []) +splitOnWhere f (_:as) csIn = splitOnWhere f as csIn + +balanceFirstLocatedAComments :: [LocatedA a] -> ([LocatedA a], [LEpaComment]) +balanceFirstLocatedAComments [] = ([],[]) +balanceFirstLocatedAComments ((L (EpAnn anc an csd) a):ds) = (L (EpAnn anc an csd0) a:ds, hc') + where + (csd0, hc') = case anc of + EpaSpan (RealSrcSpan s _) -> (csd', hc) + `debug` ("balanceFirstLocatedAComments: (csd,csd',attached,header)=" ++ showAst (csd,csd',attached,header)) + where + (priors, inners) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + (priorComments csd) + pcds = priorCommentsDeltas' s priors + (attached, header) = break (\(d,_c) -> d /= 1) pcds + csd' = setPriorComments csd (reverse (map snd attached) ++ inners) + hc = reverse (map snd header) + _ -> (csd, []) + + + +priorCommentsDeltas' :: RealSrcSpan -> [LEpaComment] + -> [(Int, LEpaComment)] +priorCommentsDeltas' r cs = go r (reverse cs) + where + go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] + go _ [] = [] + go _ (la@(L l@(EpaDelta _ dp _) _):las) = (deltaLine dp, la) : go (anchor l) las + go rs' (la@(L l _):las) = deltaComment rs' la : go (anchor l) las + + deltaComment :: RealSrcSpan -> LEpaComment -> (Int, LEpaComment) + deltaComment rs' (L loc c) = (abs(ll - al), L loc c) + where + (al,_) = ss2pos rs' + (ll,_) = ss2pos (anchor loc) + +allocatePriorComments + :: Pos + -> [LEpaComment] + -> ([LEpaComment], [LEpaComment]) +allocatePriorComments ss_loc comment_q = + let + cmp (L l _) = ss2pos (anchor l) <= ss_loc + (newAnns,after) = partition cmp comment_q + in + (after, newAnns) + +insertRemainingCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource +insertRemainingCppComments (L l p) cs = L l p' + -- `debug` ("insertRemainingCppComments: (cs,an')=" ++ showAst (cs,an')) + where + (EpAnn a an ocs) = GHC.hsmodAnn $ GHC.hsmodExt p + an' = EpAnn a an (addTrailingComments end_loc ocs cs) p' = p { GHC.hsmodExt = (GHC.hsmodExt p) { GHC.hsmodAnn = an' } } + end_loc = case GHC.hsmodLayout $ GHC.hsmodExt p of + EpExplicitBraces _open close -> case close of + EpTok (EpaSpan (RealSrcSpan s _)) -> ss2pos s + _ -> (1,1) + _ -> (1,1) + (new_before, new_after) = break (\(L ll _) -> (ss2pos $ anchor ll) > end_loc ) cs + + addTrailingComments end_loc' cur new = epaCommentsBalanced pc' fc' + where + pc = priorComments cur + fc = getFollowingComments cur + (pc', fc') = case reverse pc of + [] -> (sortEpaComments $ pc ++ new_before, sortEpaComments $ fc ++ new_after) + (L ac _:_) -> (sortEpaComments $ pc ++ cs_before, sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = if (ss2pos $ anchor ac) > end_loc' + then break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) new + else (new_before, new_after) -- --------------------------------------------------------------------- @@ -291,11 +486,16 @@ dedentDocChunkBy dedent (L (RealSrcSpan l mb) c) = L (RealSrcSpan l' mb) c dedentDocChunkBy _ x = x + +epaCommentsBalanced :: [LEpaComment] -> [LEpaComment] -> EpAnnComments +epaCommentsBalanced priorCs [] = EpaComments priorCs +epaCommentsBalanced priorCs postCs = EpaCommentsBalanced priorCs postCs + mkEpaComments :: [Comment] -> [Comment] -> EpAnnComments mkEpaComments priorCs [] = EpaComments (map comment2LEpaComment priorCs) mkEpaComments priorCs postCs - = EpaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) + = epaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) comment2LEpaComment :: Comment -> LEpaComment comment2LEpaComment (Comment s anc r _mk) = mkLEpaComment s anc r @@ -330,18 +530,11 @@ sortEpaComments cs = sortBy cmp cs mkKWComment :: AnnKeywordId -> NoCommentsLocation -> Comment mkKWComment kw (EpaSpan (RealSrcSpan ss mb)) = Comment (keywordToString kw) (EpaSpan (RealSrcSpan ss mb)) ss (Just kw) -mkKWComment kw (EpaSpan ss@(UnhelpfulSpan _)) - = Comment (keywordToString kw) (EpaDelta ss (SameLine 0) NoComments) placeholderRealSpan (Just kw) +mkKWComment kw (EpaSpan (UnhelpfulSpan _)) + = Comment (keywordToString kw) (EpaDelta noSrcSpan (SameLine 0) NoComments) placeholderRealSpan (Just kw) mkKWComment kw (EpaDelta ss dp cs) = Comment (keywordToString kw) (EpaDelta ss dp cs) placeholderRealSpan (Just kw) --- | Detects a comment which originates from a specific keyword. -isKWComment :: Comment -> Bool -isKWComment c = isJust (commentOrigin c) - -noKWComments :: [Comment] -> [Comment] -noKWComments = filter (\c -> not (isKWComment c)) - sortAnchorLocated :: [GenLocated EpaLocation a] -> [GenLocated EpaLocation a] sortAnchorLocated = sortBy (compare `on` (anchor . getLoc)) @@ -379,11 +572,6 @@ name2String = showPprUnsafe -- --------------------------------------------------------------------- -locatedAnAnchor :: LocatedAn a t -> RealSrcSpan -locatedAnAnchor (L (EpAnn a _ _) _) = anchor a - --- --------------------------------------------------------------------- - trailingAnnLoc :: TrailingAnn -> EpaLocation trailingAnnLoc (AddSemiAnn ss) = ss trailingAnnLoc (AddCommaAnn ss) = ss @@ -401,46 +589,6 @@ setTrailingAnnLoc (AddDarrowUAnn _) ss = (AddDarrowUAnn ss) addEpAnnLoc :: AddEpAnn -> EpaLocation addEpAnnLoc (AddEpAnn _ l) = l --- --------------------------------------------------------------------- --- Horrible hack for dealing with some things still having a SrcSpan, --- not an Anchor. - -{- -A SrcSpan is defined as - -data SrcSpan = - RealSrcSpan !RealSrcSpan !(Maybe BufSpan) -- See Note [Why Maybe BufPos] - | UnhelpfulSpan !UnhelpfulSpanReason - -data BufSpan = - BufSpan { bufSpanStart, bufSpanEnd :: {-# UNPACK #-} !BufPos } - deriving (Eq, Ord, Show) - -newtype BufPos = BufPos { bufPos :: Int } - - -We use the BufPos to encode a delta, using bufSpanStart for the line, -and bufSpanEnd for the col. - -To be absolutely sure, we make the delta versions use -ve values. - --} - -hackSrcSpanToAnchor :: SrcSpan -> EpaLocation -hackSrcSpanToAnchor (UnhelpfulSpan s) = error $ "hackSrcSpanToAnchor : UnhelpfulSpan:" ++ show s -hackSrcSpanToAnchor ss@(RealSrcSpan r mb) - = case mb of - (Strict.Just (BufSpan (BufPos s) (BufPos e))) -> - if s <= 0 && e <= 0 - then EpaDelta ss (deltaPos (-s) (-e)) [] - `debug` ("hackSrcSpanToAnchor: (r,s,e)=" ++ showAst (r,s,e) ) - else EpaSpan (RealSrcSpan r mb) - _ -> EpaSpan (RealSrcSpan r mb) - -hackAnchorToSrcSpan :: EpaLocation -> SrcSpan -hackAnchorToSrcSpan (EpaSpan s) = s -hackAnchorToSrcSpan _ = error $ "hackAnchorToSrcSpan" - -- --------------------------------------------------------------------- type DeclsByTag a = Map.Map DeclTag [(RealSrcSpan, a)] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/02733138052c08b8e5149f4ca8d21712d3d23f11...dda9c763e92ae2c27a103b226e501d2f708e902b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/02733138052c08b8e5149f4ca8d21712d3d23f11...dda9c763e92ae2c27a103b226e501d2f708e902b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 00:17:55 2024 From: gitlab at gitlab.haskell.org (Adriaan Leijnse (@aidylns)) Date: Thu, 12 Sep 2024 20:17:55 -0400 Subject: [Git][ghc/ghc][wip/aidylns/ttg-remove-hsunboundvar-via-hshole] 50 commits: Haddock: Add no-compilation flag Message-ID: <66e384b3680cd_38fde459cbf880031@gitlab.mail> Adriaan Leijnse pushed to branch wip/aidylns/ttg-remove-hsunboundvar-via-hshole at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - 144251c2 by Adriaan Leijnse at 2024-09-13T02:11:18+02:00 TTG: Remove HsUnboundVar, replace with HsHole, HsUnboundVarRn/Tc This commit removes HsUnboundVar from the Language AST, which was used to parse Holes to. Instead it introduces a new HsHole AST constructor for this purpose. The renaming and type checking phases keep their original HsUnboundVar implementation using HsUnboundVarRn and HsUnboundVarTc constructors (HsHole is turned into HsUnboundVarRn during renaming). Also, the note explaining Holes is rewritten to reflect the current state of the code. - - - - - 19 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3bbd76d6d2a946f8b5ec276e9930dbee30dc1988...144251c2e74c7d2fc333dcd9b7a9fdea5e299e50 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3bbd76d6d2a946f8b5ec276e9930dbee30dc1988...144251c2e74c7d2fc333dcd9b7a9fdea5e299e50 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 01:51:00 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 12 Sep 2024 21:51:00 -0400 Subject: [Git][ghc/ghc][master] finder: Add `IsBootInterface` to finder cache keys Message-ID: <66e39a84b0ac6_333c9c46f1e077427@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - 12 changed files: - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Finder/Types.hs - compiler/GHC/Unit/Module/Env.hs - compiler/GHC/Unit/Types.hs - + testsuite/tests/driver/boot-target/A.hs - + testsuite/tests/driver/boot-target/A.hs-boot - + testsuite/tests/driver/boot-target/B.hs - + testsuite/tests/driver/boot-target/Makefile - + testsuite/tests/driver/boot-target/all.T Changes: ===================================== compiler/GHC/Driver/Backpack.hs ===================================== @@ -781,7 +781,7 @@ summariseRequirement pn mod_name = do let loc = srcLocSpan (mkSrcLoc (mkFastString (bkp_filename env)) 1 1) let fc = hsc_FC hsc_env - mod <- liftIO $ addHomeModuleToFinder fc home_unit mod_name location + mod <- liftIO $ addHomeModuleToFinder fc home_unit (notBoot mod_name) location extra_sig_imports <- liftIO $ findExtraSigImports hsc_env HsigFile mod_name @@ -893,7 +893,7 @@ hsModuleToModSummary home_keys pn hsc_src modname this_mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit modname location + addHomeModuleToFinder fc home_unit (GWIB modname (hscSourceToIsBoot hsc_src)) location let ms = ModSummary { ms_mod = this_mod, ms_hsc_src = hsc_src, ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2044,25 +2044,43 @@ summariseFile hsc_env' home_unit old_summaries src_fn mb_phase maybe_buf <- getPreprocessedImports hsc_env src_fn mb_phase maybe_buf let fopts = initFinderOpts (hsc_dflags hsc_env) - - -- Make a ModLocation for this file - let location = mkHomeModLocation fopts pi_mod_name (unsafeEncodeUtf src_fn) + src_path = unsafeEncodeUtf src_fn + + is_boot = case takeExtension src_fn of + ".hs-boot" -> IsBoot + ".lhs-boot" -> IsBoot + _ -> NotBoot + + (path_without_boot, hsc_src) + | isHaskellSigFilename src_fn = (src_path, HsigFile) + | IsBoot <- is_boot = (removeBootSuffix src_path, HsBootFile) + | otherwise = (src_path, HsSrcFile) + + -- Make a ModLocation for the Finder, who only has one entry for + -- each @ModuleName@, and therefore needs to use the locations for + -- the non-boot files. + location_without_boot = + mkHomeModLocation fopts pi_mod_name path_without_boot + + -- Make a ModLocation for this file, adding the @-boot@ suffix to + -- all paths if the original was a boot file. + location + | IsBoot <- is_boot + = addBootSuffixLocn location_without_boot + | otherwise + = location_without_boot -- Tell the Finder cache where it is, so that subsequent calls -- to findModule will find it, even if it's not on any search path mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit pi_mod_name location + addHomeModuleToFinder fc home_unit (GWIB pi_mod_name is_boot) location liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = NotBoot - , nms_hsc_src = - if isHaskellSigFilename src_fn - then HsigFile - else HsSrcFile + , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod , nms_preimps = preimps @@ -2090,9 +2108,10 @@ checkSummaryHash -- Also, only add to finder cache for non-boot modules as the finder cache -- makes sure to add a boot suffix for boot files. _ <- do - let fc = hsc_FC hsc_env + let fc = hsc_FC hsc_env + gwib = GWIB (ms_mod old_summary) (isBootSummary old_summary) case ms_hsc_src old_summary of - HsSrcFile -> addModuleToFinder fc (ms_mod old_summary) location + HsSrcFile -> addModuleToFinder fc gwib location _ -> return () hi_timestamp <- modificationTimeIfExists (ml_hi_file location) @@ -2230,7 +2249,6 @@ summariseModule hsc_env' home_unit old_summary_map is_boot (L _ wanted_mod) mb_p liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = is_boot , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod @@ -2243,7 +2261,6 @@ data MakeNewModSummary = MakeNewModSummary { nms_src_fn :: FilePath , nms_src_hash :: Fingerprint - , nms_is_boot :: IsBootInterface , nms_hsc_src :: HscSource , nms_location :: ModLocation , nms_mod :: Module ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -734,7 +734,7 @@ runHscPhase pipe_env hsc_env0 input_fn src_flavour = do mod <- do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit mod_name location + addHomeModuleToFinder fc home_unit (GWIB mod_name (hscSourceToIsBoot src_flavour)) location -- Make the ModSummary to hand to hscMain let ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -89,23 +89,23 @@ type BaseName = OsPath -- Basename of file initFinderCache :: IO FinderCache initFinderCache = do - mod_cache <- newIORef emptyInstalledModuleEnv + mod_cache <- newIORef emptyInstalledModuleWithIsBootEnv file_cache <- newIORef M.empty let flushFinderCaches :: UnitEnv -> IO () flushFinderCaches ue = do - atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleEnv is_ext fm, ()) + atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleWithIsBootEnv is_ext fm, ()) atomicModifyIORef' file_cache $ \_ -> (M.empty, ()) where - is_ext mod _ = not (isUnitEnvInstalledModule ue mod) + is_ext mod _ = not (isUnitEnvInstalledModule ue (gwib_mod mod)) - addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () addToFinderCache key val = - atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleEnv c key val, ()) + atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleWithIsBootEnv c key val, ()) - lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) lookupFinderCache key = do c <- readIORef mod_cache - return $! lookupInstalledModuleEnv c key + return $! lookupInstalledModuleWithIsBootEnv c key lookupFileCache :: FilePath -> IO Fingerprint lookupFileCache key = do @@ -255,7 +255,7 @@ orIfNotFound this or_this = do homeSearchCache :: FinderCache -> UnitId -> ModuleName -> IO InstalledFindResult -> IO InstalledFindResult homeSearchCache fc home_unit mod_name do_this = do let mod = mkModule home_unit mod_name - modLocationCache fc mod do_this + modLocationCache fc (notBoot mod) do_this findExposedPackageModule :: FinderCache -> FinderOpts -> UnitState -> ModuleName -> PkgQual -> IO FindResult findExposedPackageModule fc fopts units mod_name mb_pkg = @@ -312,7 +312,7 @@ findLookupResult fc fopts r = case r of , fr_unusables = [] , fr_suggestions = suggest' }) -modLocationCache :: FinderCache -> InstalledModule -> IO InstalledFindResult -> IO InstalledFindResult +modLocationCache :: FinderCache -> InstalledModuleWithIsBoot -> IO InstalledFindResult -> IO InstalledFindResult modLocationCache fc mod do_this = do m <- lookupFinderCache fc mod case m of @@ -322,17 +322,17 @@ modLocationCache fc mod do_this = do addToFinderCache fc mod result return result -addModuleToFinder :: FinderCache -> Module -> ModLocation -> IO () +addModuleToFinder :: FinderCache -> ModuleWithIsBoot -> ModLocation -> IO () addModuleToFinder fc mod loc = do - let imod = toUnitId <$> mod - addToFinderCache fc imod (InstalledFound loc imod) + let imod = fmap toUnitId <$> mod + addToFinderCache fc imod (InstalledFound loc (gwib_mod imod)) -- This returns a module because it's more convenient for users -addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleName -> ModLocation -> IO Module +addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleNameWithIsBoot -> ModLocation -> IO Module addHomeModuleToFinder fc home_unit mod_name loc = do - let mod = mkHomeInstalledModule home_unit mod_name - addToFinderCache fc mod (InstalledFound loc mod) - return (mkHomeModule home_unit mod_name) + let mod = mkHomeInstalledModule home_unit <$> mod_name + addToFinderCache fc mod (InstalledFound loc (gwib_mod mod)) + return (mkHomeModule home_unit (gwib_mod mod_name)) -- ----------------------------------------------------------------------------- -- The internal workers @@ -466,7 +466,7 @@ findPackageModule_ :: FinderCache -> FinderOpts -> InstalledModule -> UnitInfo - findPackageModule_ fc fopts mod pkg_conf = do massertPpr (moduleUnit mod == unitId pkg_conf) (ppr (moduleUnit mod) <+> ppr (unitId pkg_conf)) - modLocationCache fc mod $ + modLocationCache fc (notBoot mod) $ -- special case for GHC.Prim; we won't find it in the filesystem. if mod `installedModuleEq` gHC_PRIM ===================================== compiler/GHC/Unit/Finder/Types.hs ===================================== @@ -30,9 +30,9 @@ data FinderCache = FinderCache { flushFinderCaches :: UnitEnv -> IO () -- ^ remove all the home modules from the cache; package modules are -- assumed to not move around during a session; also flush the file hash -- cache. - , addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + , addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () -- ^ Add a found location to the cache for the module. - , lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + , lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) -- ^ Look for a location in the cache. , lookupFileCache :: FilePath -> IO Fingerprint -- ^ Look for the hash of a file in the cache. This should add it to the ===================================== compiler/GHC/Unit/Module/Env.hs ===================================== @@ -33,6 +33,17 @@ module GHC.Unit.Module.Env , mergeInstalledModuleEnv , plusInstalledModuleEnv , installedModuleEnvElts + + -- * InstalledModuleWithIsBootEnv + , InstalledModuleWithIsBootEnv + , emptyInstalledModuleWithIsBootEnv + , lookupInstalledModuleWithIsBootEnv + , extendInstalledModuleWithIsBootEnv + , filterInstalledModuleWithIsBootEnv + , delInstalledModuleWithIsBootEnv + , mergeInstalledModuleWithIsBootEnv + , plusInstalledModuleWithIsBootEnv + , installedModuleWithIsBootEnvElts ) where @@ -283,3 +294,56 @@ plusInstalledModuleEnv :: (elt -> elt -> elt) plusInstalledModuleEnv f (InstalledModuleEnv xm) (InstalledModuleEnv ym) = InstalledModuleEnv $ Map.unionWith f xm ym + + +-------------------------------------------------------------------- +-- InstalledModuleWithIsBootEnv +-------------------------------------------------------------------- + +-- | A map keyed off of 'InstalledModuleWithIsBoot' +newtype InstalledModuleWithIsBootEnv elt = InstalledModuleWithIsBootEnv (Map InstalledModuleWithIsBoot elt) + +instance Outputable elt => Outputable (InstalledModuleWithIsBootEnv elt) where + ppr (InstalledModuleWithIsBootEnv env) = ppr env + + +emptyInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a +emptyInstalledModuleWithIsBootEnv = InstalledModuleWithIsBootEnv Map.empty + +lookupInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> Maybe a +lookupInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = Map.lookup m e + +extendInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> a -> InstalledModuleWithIsBootEnv a +extendInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m x = InstalledModuleWithIsBootEnv (Map.insert m x e) + +filterInstalledModuleWithIsBootEnv :: (InstalledModuleWithIsBoot -> a -> Bool) -> InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBootEnv a +filterInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv e) = + InstalledModuleWithIsBootEnv (Map.filterWithKey f e) + +delInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> InstalledModuleWithIsBootEnv a +delInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = InstalledModuleWithIsBootEnv (Map.delete m e) + +installedModuleWithIsBootEnvElts :: InstalledModuleWithIsBootEnv a -> [(InstalledModuleWithIsBoot, a)] +installedModuleWithIsBootEnvElts (InstalledModuleWithIsBootEnv e) = Map.assocs e + +mergeInstalledModuleWithIsBootEnv + :: (elta -> eltb -> Maybe eltc) + -> (InstalledModuleWithIsBootEnv elta -> InstalledModuleWithIsBootEnv eltc) -- map X + -> (InstalledModuleWithIsBootEnv eltb -> InstalledModuleWithIsBootEnv eltc) -- map Y + -> InstalledModuleWithIsBootEnv elta + -> InstalledModuleWithIsBootEnv eltb + -> InstalledModuleWithIsBootEnv eltc +mergeInstalledModuleWithIsBootEnv f g h (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) + = InstalledModuleWithIsBootEnv $ Map.mergeWithKey + (\_ x y -> (x `f` y)) + (coerce g) + (coerce h) + xm ym + +plusInstalledModuleWithIsBootEnv :: (elt -> elt -> elt) + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt +plusInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) = + InstalledModuleWithIsBootEnv $ Map.unionWith f xm ym + ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -86,6 +86,8 @@ module GHC.Unit.Types , GenWithIsBoot (..) , ModuleNameWithIsBoot , ModuleWithIsBoot + , InstalledModuleWithIsBoot + , notBoot ) where @@ -723,6 +725,8 @@ type ModuleNameWithIsBoot = GenWithIsBoot ModuleName type ModuleWithIsBoot = GenWithIsBoot Module +type InstalledModuleWithIsBoot = GenWithIsBoot InstalledModule + instance Binary a => Binary (GenWithIsBoot a) where put_ bh (GWIB { gwib_mod, gwib_isBoot }) = do put_ bh gwib_mod @@ -736,3 +740,6 @@ instance Outputable a => Outputable (GenWithIsBoot a) where ppr (GWIB { gwib_mod, gwib_isBoot }) = hsep $ ppr gwib_mod : case gwib_isBoot of IsBoot -> [ text "{-# SOURCE #-}" ] NotBoot -> [] + +notBoot :: mod -> GenWithIsBoot mod +notBoot gwib_mod = GWIB {gwib_mod, gwib_isBoot = NotBoot} ===================================== testsuite/tests/driver/boot-target/A.hs ===================================== @@ -0,0 +1,5 @@ +module A where + +import B + +data A = A B ===================================== testsuite/tests/driver/boot-target/A.hs-boot ===================================== @@ -0,0 +1,3 @@ +module A where + +data A ===================================== testsuite/tests/driver/boot-target/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} A + +data B = B A ===================================== testsuite/tests/driver/boot-target/Makefile ===================================== @@ -0,0 +1,8 @@ +boot1: + $(TEST_HC) -c A.hs-boot B.hs + +boot2: + $(TEST_HC) A.hs-boot A.hs B.hs -v0 + +boot3: + $(TEST_HC) A.hs-boot B.hs -v0 \ No newline at end of file ===================================== testsuite/tests/driver/boot-target/all.T ===================================== @@ -0,0 +1,10 @@ +def test_boot(name): + return test(name, + [extra_files(['A.hs', 'A.hs-boot', 'B.hs']), + ], + makefile_test, + []) + +test_boot('boot1') +test_boot('boot2') +test_boot('boot3') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fff55592a7b9c9487c043d055f2d0d77fa549f4e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fff55592a7b9c9487c043d055f2d0d77fa549f4e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 01:52:05 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 12 Sep 2024 21:52:05 -0400 Subject: [Git][ghc/ghc][master] EPA: Sync ghc-exactprint to GHC Message-ID: <66e39ac4e6a25_333c9c40088080971@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 6 changed files: - utils/check-exact/ExactPrint.hs - utils/check-exact/Main.hs - utils/check-exact/Parsers.hs - utils/check-exact/Transform.hs - utils/check-exact/Types.hs - utils/check-exact/Utils.hs Changes: ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -1,3 +1,4 @@ +{-# LANGUAGE BangPatterns #-} {-# LANGUAGE DeriveDataTypeable #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE FlexibleInstances #-} @@ -25,7 +26,7 @@ module ExactPrint , makeDeltaAst -- * Configuration - , EPOptions(epRigidity, epAstPrint, epTokenPrint, epWhitespacePrint, epUpdateAnchors) + , EPOptions(epTokenPrint, epWhitespacePrint) , stringOptions , epOptions , deltaOptions @@ -43,10 +44,11 @@ import GHC.Types.ForeignCall import GHC.Types.Name.Reader import GHC.Types.PkgQual import GHC.Types.SourceText +import GHC.Types.SrcLoc import GHC.Types.Var -import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Unit.Module.Warnings import GHC.Utils.Misc +import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Utils.Panic import Language.Haskell.Syntax.Basic (FieldLabelString(..)) @@ -77,8 +79,7 @@ import Types exactPrint :: ExactPrint ast => ast -> String exactPrint ast = snd $ runIdentity (runEP stringOptions (markAnnotated ast)) --- | The additional option to specify the rigidity and printing --- configuration. +-- | The additional option to specify the printing configuration. exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) => EPOptions m b -> ast @@ -86,9 +87,8 @@ exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) exactPrintWithOptions r ast = runEP r (markAnnotated ast) --- | Transform concrete annotations into relative annotations which --- are more useful when transforming an AST. This corresponds to the --- earlier 'relativiseApiAnns'. +-- | Transform concrete annotations into relative annotations. +-- This should be unnecessary from GHC 9.10 makeDeltaAst :: ExactPrint ast => ast -> ast makeDeltaAst ast = fst $ runIdentity (runEP deltaOptions (markAnnotated ast)) @@ -115,6 +115,7 @@ defaultEPState = EPState , dPriorEndPosition = (1,1) , uAnchorSpan = badRealSrcSpan , uExtraDP = Nothing + , uExtraDPReturn = Nothing , pAcceptSpan = False , epComments = [] , epCommentsApplied = [] @@ -128,39 +129,27 @@ defaultEPState = EPState -- | The R part of RWS. The environment. Updated via 'local' as we -- enter a new AST element, having a different anchor point. data EPOptions m a = EPOptions - { - epAstPrint :: forall ast . Data ast => GHC.Located ast -> a -> m a - , epTokenPrint :: String -> m a + { epTokenPrint :: String -> m a , epWhitespacePrint :: String -> m a - , epRigidity :: Rigidity - , epUpdateAnchors :: Bool } -- | Helper to create a 'EPOptions' -epOptions :: - (forall ast . Data ast => GHC.Located ast -> a -> m a) - -> (String -> m a) - -> (String -> m a) - -> Rigidity - -> Bool - -> EPOptions m a -epOptions astPrint tokenPrint wsPrint rigidity delta = EPOptions - { - epAstPrint = astPrint - , epWhitespacePrint = wsPrint +epOptions :: (String -> m a) + -> (String -> m a) + -> EPOptions m a +epOptions tokenPrint wsPrint = EPOptions + { epWhitespacePrint = wsPrint , epTokenPrint = tokenPrint - , epRigidity = rigidity - , epUpdateAnchors = delta } -- | Options which can be used to print as a normal String. stringOptions :: EPOptions Identity String -stringOptions = epOptions (\_ b -> return b) return return NormalLayout False +stringOptions = epOptions return return -- | Options which can be used to simply update the AST to be in delta -- form, without generating output deltaOptions :: EPOptions Identity () -deltaOptions = epOptions (\_ _ -> return ()) (\_ -> return ()) (\_ -> return ()) NormalLayout True +deltaOptions = epOptions (\_ -> return ()) (\_ -> return ()) data EPWriter a = EPWriter { output :: !a } @@ -177,6 +166,8 @@ data EPState = EPState -- Annotation , uExtraDP :: !(Maybe EpaLocation) -- ^ Used to anchor a -- list + , uExtraDPReturn :: !(Maybe DeltaPos) + -- ^ Used to return Delta version of uExtraDP , pAcceptSpan :: Bool -- ^ When we have processed an -- entry of EpaDelta, accept the -- next `EpaSpan` start as the @@ -213,7 +204,7 @@ class HasTrailing a where trailing :: a -> [TrailingAnn] setTrailing :: a -> [TrailingAnn] -> a -setAnchorEpa :: (HasTrailing an, NoAnn an) +setAnchorEpa :: (HasTrailing an) => EpAnn an -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn an setAnchorEpa (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing an ts) cs @@ -223,7 +214,7 @@ setAnchorHsModule hsmod anc cs = hsmod { hsmodExt = (hsmodExt hsmod) {hsmodAnn = anc' = anc an' = setAnchorEpa (hsmodAnn $ hsmodExt hsmod) anc' [] cs -setAnchorAn :: (HasTrailing an, NoAnn an) +setAnchorAn :: (HasTrailing an) => LocatedAn an a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a setAnchorAn (L (EpAnn _ an _) a) anc ts cs = (L (EpAnn anc (setTrailing an ts) cs) a) -- `debug` ("setAnchorAn: anc=" ++ showAst anc) @@ -248,7 +239,7 @@ data FlushComments = FlushComments data CanUpdateAnchor = CanUpdateAnchor | CanUpdateAnchorOnly | NoCanUpdateAnchor - deriving (Eq, Show) + deriving (Eq, Show, Data) data Entry = Entry EpaLocation [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor | NoEntryVal @@ -402,7 +393,7 @@ enterAnn NoEntryVal a = do r <- exact a debugM $ "enterAnn:done:NO ANN:p =" ++ show (p, astId a) return r -enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do +enterAnn !(Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do acceptSpan <- getAcceptSpan setAcceptSpan False case anchor' of @@ -421,9 +412,11 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do _ -> return () case anchor' of EpaDelta _ _ dcs -> do - debugM $ "enterAnn:Printing comments:" ++ showGhc (priorComments cs) + debugM $ "enterAnn:Delta:Flushing comments" + flushComments [] + debugM $ "enterAnn:Delta:Printing prior comments:" ++ showGhc (priorComments cs) mapM_ printOneComment (concatMap tokComment $ priorComments cs) - debugM $ "enterAnn:Printing EpaDelta comments:" ++ showGhc dcs + debugM $ "enterAnn:Delta:Printing EpaDelta comments:" ++ showGhc dcs mapM_ printOneComment (concatMap tokComment dcs) _ -> do debugM $ "enterAnn:Adding comments:" ++ showGhc (priorComments cs) @@ -465,7 +458,7 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- The first part corresponds to the delta phase, so should only use -- delta phase variables ----------------------------------- -- Calculate offset required to get to the start of the SrcSPan - off <- getLayoutOffsetD + !off <- getLayoutOffsetD let spanStart = ss2pos curAnchor priorEndAfterComments <- getPriorEndD let edp' = adjustDeltaForOffset @@ -480,17 +473,18 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- --------------------------------------------- med <- getExtraDP setExtraDP Nothing - let edp = case med of - Nothing -> edp'' - Just (EpaDelta _ dp _) -> dp + let (edp, medr) = case med of + Nothing -> (edp'', Nothing) + Just (EpaDelta _ dp _) -> (dp, Nothing) -- Replace original with desired one. Allows all -- list entry values to be DP (1,0) - Just (EpaSpan (RealSrcSpan r _)) -> dp + Just (EpaSpan (RealSrcSpan r _)) -> (dp, Just dp) where dp = adjustDeltaForOffset off (ss2delta priorEndAfterComments r) Just (EpaSpan (UnhelpfulSpan r)) -> panic $ "enterAnn: UnhelpfulSpan:" ++ show r when (isJust med) $ debugM $ "enterAnn:(med,edp)=" ++ showAst (med,edp) + when (isJust medr) $ setExtraDPReturn medr -- --------------------------------------------- -- Preparation complete, perform the action when (priorEndAfterComments < spanStart) (do @@ -511,12 +505,15 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do debugM $ "enterAnn:exact a starting:" ++ show (showAst anchor') a' <- exact a debugM $ "enterAnn:exact a done:" ++ show (showAst anchor') + + -- Core recursive exactprint done, start end of Entry processing + when (flush == FlushComments) $ do - debugM $ "flushing comments in enterAnn:" ++ showAst cs + debugM $ "flushing comments in enterAnn:" ++ showAst (cs, getFollowingComments cs) flushComments (getFollowingComments cs) debugM $ "flushing comments in enterAnn done" - eof <- getEofPos + !eof <- getEofPos case eof of Nothing -> return () Just (pos, prior) -> do @@ -544,28 +541,50 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- Outside the anchor, mark any trailing postCs <- cua canUpdateAnchor takeAppliedCommentsPop - when (flush == NoFlushComments) $ do - when ((getFollowingComments cs) /= []) $ do - - -- debugM $ "enterAnn:in:(anchor') =" ++ show (eloc2str anchor') - debugM $ "starting trailing comments:" ++ showAst (getFollowingComments cs) - mapM_ printOneComment (concatMap tokComment $ getFollowingComments cs) - debugM $ "ending trailing comments" - trailing' <- markTrailing trailing_anns + following <- if (flush == NoFlushComments) + then do + let (before, after) = splitAfterTrailingAnns trailing_anns + (getFollowingComments cs) + addCommentsA before + return after + else return [] + !trailing' <- markTrailing trailing_anns + -- mapM_ printOneComment (concatMap tokComment $ following) + addCommentsA following -- Update original anchor, comments based on the printing process -- TODO:AZ: probably need to put something appropriate in instead of noSrcSpan - let newAchor = EpaDelta noSrcSpan edp [] + let newAnchor = EpaDelta noSrcSpan edp [] let r = case canUpdateAnchor of - CanUpdateAnchor -> setAnnotationAnchor a' newAchor trailing' (mkEpaComments (priorCs ++ postCs) []) - CanUpdateAnchorOnly -> setAnnotationAnchor a' newAchor [] emptyComments + CanUpdateAnchor -> setAnnotationAnchor a' newAnchor trailing' (mkEpaComments priorCs postCs) + CanUpdateAnchorOnly -> setAnnotationAnchor a' newAnchor [] emptyComments NoCanUpdateAnchor -> a' return r -- --------------------------------------------------------------------- +-- | Split the span following comments into ones that occur prior to +-- the last trailing ann, and ones after. +splitAfterTrailingAnns :: [TrailingAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitAfterTrailingAnns [] cs = ([], cs) +splitAfterTrailingAnns tas cs = (before, after) + where + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + (before, after) = case reverse (concatMap trailing_loc tas) of + [] -> ([],cs) + (s:_) -> (b,a) + where + s_pos = ss2pos s + (b,a) = break (\(L ll _) -> (ss2pos $ anchor ll) > s_pos) + cs + + +-- --------------------------------------------------------------------- + addCommentsA :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -addCommentsA csNew = addComments (concatMap tokComment csNew) +addCommentsA csNew = addComments False (concatMap tokComment csNew) {- TODO: When we addComments, some may have an anchor that is no longer @@ -583,24 +602,36 @@ By definition it is the current anchor, so work against that. And that also means that the first entry comment that has moved should not have a line offset. -} -addComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -addComments csNew = do - -- debugM $ "addComments:" ++ show csNew +addComments :: (Monad m, Monoid w) => Bool -> [Comment] -> EP w m () +addComments sortNeeded csNew = do + debugM $ "addComments:csNew" ++ show csNew cs <- getUnallocatedComments + debugM $ "addComments:cs" ++ show cs + -- We can only sort the comments if we are in the first phase, + -- where all comments have locations. If any have EpaDelta the + -- sort will fail, so we do not try. + if sortNeeded && all noDelta (csNew ++ cs) + then putUnallocatedComments (sort (cs ++ csNew)) + else putUnallocatedComments (cs ++ csNew) - putUnallocatedComments (sort (cs ++ csNew)) +noDelta :: Comment -> Bool +noDelta c = case commentLoc c of + EpaSpan _ -> True + _ -> False -- --------------------------------------------------------------------- -- | Just before we print out the EOF comments, flush the remaining -- ones in the state. flushComments :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -flushComments trailing_anns = do +flushComments !trailing_anns = do + debugM $ "flushComments entered: " ++ showAst trailing_anns addCommentsA trailing_anns + debugM $ "flushComments after addCommentsA" cs <- getUnallocatedComments - debugM $ "flushing comments starting" - -- AZ:TODO: is the sort still needed? - mapM_ printOneComment (sortComments cs) + debugM $ "flushComments: got cs" + debugM $ "flushing comments starting: cs" ++ showAst cs + mapM_ printOneComment cs putUnallocatedComments [] debugM $ "flushing comments done" @@ -612,7 +643,7 @@ annotationsToComments :: (Monad m, Monoid w) => a -> Lens a [AddEpAnn] -> [AnnKeywordId] -> EP w m a annotationsToComments a l kws = do let (newComments, newAnns) = go ([],[]) (view l a) - addComments newComments + addComments True newComments return (set l (reverse newAnns) a) where keywords = Set.fromList kws @@ -654,14 +685,11 @@ printSourceText (NoSourceText) txt = printStringAdvance txt >> return () printSourceText (SourceText txt) _ = printStringAdvance (unpackFS txt) >> return () printSourceTextAA :: (Monad m, Monoid w) => SourceText -> String -> EP w m () -printSourceTextAA (NoSourceText) txt = printStringAtAA noAnn txt >> return () -printSourceTextAA (SourceText txt) _ = printStringAtAA noAnn (unpackFS txt) >> return () +printSourceTextAA (NoSourceText) txt = printStringAdvanceA txt >> return () +printSourceTextAA (SourceText txt) _ = printStringAdvanceA (unpackFS txt) >> return () -- --------------------------------------------------------------------- -printStringAtSs :: (Monad m, Monoid w) => SrcSpan -> String -> EP w m () -printStringAtSs ss str = printStringAtRs (realSrcSpan ss) str >> return () - printStringAtRs :: (Monad m, Monoid w) => RealSrcSpan -> String -> EP w m EpaLocation printStringAtRs pa str = printStringAtRsC CaptureComments pa str @@ -676,7 +704,7 @@ printStringAtRsC capture pa str = do p' <- adjustDeltaForOffsetM p debugM $ "printStringAtRsC:(p,p')=" ++ show (p,p') printStringAtLsDelta p' str - setPriorEndASTD True pa + setPriorEndASTD pa cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -709,6 +737,9 @@ printStringAtMLocL (EpAnn anc an cs) l s = do printStringAtLsDelta (SameLine 1) str return (Just (EpaDelta noSrcSpan (SameLine 1) [])) +printStringAdvanceA :: (Monad m, Monoid w) => String -> EP w m () +printStringAdvanceA str = printStringAtAA (EpaDelta noSrcSpan (SameLine 0) []) str >> return () + printStringAtAA :: (Monad m, Monoid w) => EpaLocation -> String -> EP w m EpaLocation printStringAtAA el str = printStringAtAAC CaptureComments el str @@ -735,7 +766,7 @@ printStringAtAAC capture (EpaDelta ss d cs) s = do p2 <- getPosP pe2 <- getPriorEndD debugM $ "printStringAtAA:(pe1,pe2,p1,p2)=" ++ show (pe1,pe2,p1,p2) - setPriorEndASTPD True (pe1,pe2) + setPriorEndASTPD (pe1,pe2) cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -883,8 +914,7 @@ markAnnOpenP' :: (Monad m, Monoid w) => AnnPragma -> SourceText -> String -> EP markAnnOpenP' an NoSourceText txt = markEpAnnLMS0 an lapr_open AnnOpen (Just txt) markAnnOpenP' an (SourceText txt) _ = markEpAnnLMS0 an lapr_open AnnOpen (Just $ unpackFS txt) -markAnnOpen :: (Monad m, Monoid w) - => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] +markAnnOpen :: (Monad m, Monoid w) => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] markAnnOpen an NoSourceText txt = markEpAnnLMS'' an lidl AnnOpen (Just txt) markAnnOpen an (SourceText txt) _ = markEpAnnLMS'' an lidl AnnOpen (Just $ unpackFS txt) @@ -1589,7 +1619,7 @@ markTopLevelList ls = mapM (\a -> setLayoutTopLevelP $ markAnnotated a) ls instance (ExactPrint a) => ExactPrint (Located a) where getAnnotationEntry (L l _) = case l of UnhelpfulSpan _ -> NoEntryVal - _ -> Entry (hackSrcSpanToAnchor l) [] emptyComments NoFlushComments CanUpdateAnchorOnly + _ -> Entry (EpaSpan l) [] emptyComments NoFlushComments CanUpdateAnchorOnly setAnnotationAnchor (L l a) _anc _ts _cs = L l a @@ -1664,16 +1694,10 @@ instance ExactPrint (HsModule GhcPs) where _ -> return lo am_decls' <- markTrailing (am_decls $ anns an0) - imports' <- markTopLevelList imports - - case lo of - EpExplicitBraces _ _ -> return () - _ -> do - -- Get rid of the balance of the preceding comments before starting on the decls - flushComments [] - putUnallocatedComments [] - decls' <- markTopLevelList (filter removeDocDecl decls) + mid <- markAnnotated (HsModuleImpDecls (am_cs $ anns an0) imports decls) + let imports' = id_imps mid + let decls' = id_decls mid lo1 <- case lo0 of EpExplicitBraces open close -> do @@ -1688,15 +1712,32 @@ instance ExactPrint (HsModule GhcPs) where debugM $ "am_eof:" ++ showGhc (pos, prior) setEofPos (Just (pos, prior)) - let anf = an0 { anns = (anns an0) { am_decls = am_decls' }} + let anf = an0 { anns = (anns an0) { am_decls = am_decls', am_cs = [] }} debugM $ "HsModule, anf=" ++ showAst anf return (HsModule (XModulePs anf lo1 mdeprec' mbDoc') mmn' mexports' imports' decls') +-- --------------------------------------------------------------------- + +-- | This is used to ensure the comments are updated into the right +-- place for makeDeltaAst. +data HsModuleImpDecls + = HsModuleImpDecls { + id_cs :: [LEpaComment], + id_imps :: [LImportDecl GhcPs], + id_decls :: [LHsDecl GhcPs] + } deriving Data + +instance ExactPrint HsModuleImpDecls where + -- Use an UnhelpfulSpan for the anchor, we are only interested in the comments + getAnnotationEntry mid = mkEntry (EpaSpan (UnhelpfulSpan UnhelpfulNoLocationInfo)) [] (EpaComments (id_cs mid)) + setAnnotationAnchor mid _anc _ cs = mid { id_cs = priorComments cs ++ getFollowingComments cs } + `debug` ("HsModuleImpDecls.setAnnotationAnchor:cs=" ++ showAst cs) + exact (HsModuleImpDecls cs imports decls) = do + imports' <- markTopLevelList imports + decls' <- markTopLevelList (filter notDocDecl decls) + return (HsModuleImpDecls cs imports' decls') -removeDocDecl :: LHsDecl GhcPs -> Bool -removeDocDecl (L _ DocD{}) = False -removeDocDecl _ = True -- --------------------------------------------------------------------- @@ -1737,8 +1778,8 @@ instance ExactPrint InWarningCategory where exact (InWarningCategory tkIn source (L l wc)) = do tkIn' <- markEpToken tkIn - L _ (_,wc') <- markAnnotated (L l (source, wc)) - return (InWarningCategory tkIn' source (L l wc')) + L l' (_,wc') <- markAnnotated (L l (source, wc)) + return (InWarningCategory tkIn' source (L l' wc')) instance ExactPrint (SourceText, WarningCategory) where getAnnotationEntry _ = NoEntryVal @@ -1943,14 +1984,14 @@ exactDataFamInstDecl an top_lvl , feqn_pats = pats , feqn_fixity = fixity , feqn_rhs = defn })) = do - (an', an2', tycon', bndrs', _, _mc, defn') <- exactDataDefn an2 pp_hdr defn - -- See Note [an and an2 in exactDataFamInstDecl] + (an', an2', tycon', bndrs', pats', defn') <- exactDataDefn an2 pp_hdr defn + -- See Note [an and an2 in exactDataFamInstDecl] return (an', DataFamInstDecl ( FamEqn { feqn_ext = an2' , feqn_tycon = tycon' , feqn_bndrs = bndrs' - , feqn_pats = pats + , feqn_pats = pats' , feqn_fixity = fixity , feqn_rhs = defn' })) `debug` ("exactDataFamInstDecl: defn' derivs:" ++ showAst (dd_derivs defn')) @@ -2233,11 +2274,11 @@ instance ExactPrint (RoleAnnotDecl GhcPs) where an1 <- markEpAnnL an0 lidl AnnRole ltycon' <- markAnnotated ltycon let markRole (L l (Just r)) = do - (L _ r') <- markAnnotated (L l r) - return (L l (Just r')) + (L l' r') <- markAnnotated (L l r) + return (L l' (Just r')) markRole (L l Nothing) = do - printStringAtSs (locA l) "_" - return (L l Nothing) + e' <- printStringAtAA (entry l) "_" + return (L (l { entry = e'}) Nothing) roles' <- mapM markRole roles return (RoleAnnotDecl an1 ltycon' roles') @@ -2340,8 +2381,13 @@ instance (ExactPrint tm, ExactPrint ty, Outputable tm, Outputable ty) getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact a@(HsValArg _ tm) = markAnnotated tm >> return a - exact a@(HsTypeArg at ty) = markEpToken at >> markAnnotated ty >> return a + exact (HsValArg x tm) = do + tm' <- markAnnotated tm + return (HsValArg x tm') + exact (HsTypeArg at ty) = do + at' <- markEpToken at + ty' <- markAnnotated ty + return (HsTypeArg at' ty') exact x@(HsArgPar _sp) = withPpr x -- Does not appear in original source -- --------------------------------------------------------------------- @@ -2359,9 +2405,9 @@ instance ExactPrint (ClsInstDecl GhcPs) where (mbWarn', an0, mbOverlap', inst_ty') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lid AnnSemi - ds <- withSortKey sortKey - [(ClsAtdTag, prepareListAnnotationA ats), - (ClsAtdTag, prepareListAnnotationF an adts), + (sortKey', ds) <- withSortKey sortKey + [(ClsAtTag, prepareListAnnotationA ats), + (ClsAtdTag, prepareListAnnotationF adts), (ClsMethodTag, prepareListAnnotationA binds), (ClsSigTag, prepareListAnnotationA sigs) ] @@ -2371,7 +2417,7 @@ instance ExactPrint (ClsInstDecl GhcPs) where adts' = undynamic ds binds' = undynamic ds sigs' = undynamic ds - return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey) + return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey') , cid_poly_ty = inst_ty', cid_binds = binds' , cid_sigs = sigs', cid_tyfam_insts = ats' , cid_overlap_mode = mbOverlap' @@ -2452,15 +2498,29 @@ instance ExactPrint (HsBind GhcPs) where return (FunBind x fun_id' matches') exact (PatBind x pat q grhss) = do + q' <- markAnnotated q pat' <- markAnnotated pat grhss' <- markAnnotated grhss - return (PatBind x pat' q grhss') + return (PatBind x pat' q' grhss') exact (PatSynBind x bind) = do bind' <- markAnnotated bind return (PatSynBind x bind') exact x = error $ "HsBind: exact for " ++ showAst x +instance ExactPrint (HsMultAnn GhcPs) where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact (HsNoMultAnn x) = return (HsNoMultAnn x) + exact (HsPct1Ann tok) = do + tok' <- markEpToken tok + return (HsPct1Ann tok') + exact (HsMultAnn tok ty) = do + tok' <- markEpToken tok + ty' <- markAnnotated ty + return (HsMultAnn tok' ty') + -- --------------------------------------------------------------------- instance ExactPrint (PatSynBind GhcPs GhcPs) where @@ -2519,8 +2579,9 @@ instance ExactPrint (PatSynBind GhcPs GhcPs) where instance ExactPrint (RecordPatSynField GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact r@(RecordPatSynField { recordPatSynField = v }) = markAnnotated v - >> return r + exact (RecordPatSynField f v) = do + f' <- markAnnotated f + return (RecordPatSynField f' v) -- --------------------------------------------------------------------- @@ -2648,15 +2709,20 @@ instance ExactPrint (HsLocalBinds GhcPs) where (an1, valbinds') <- markAnnList an0 $ markAnnotatedWithLayout valbinds debugM $ "exact HsValBinds: an1=" ++ showAst an1 - return (HsValBinds an1 valbinds') + medr <- getExtraDPReturn + an2 <- case medr of + Nothing -> return an1 + Just dp -> do + setExtraDPReturn Nothing + return $ an1 { anns = (anns an1) { al_anchor = Just (EpaDelta noSrcSpan dp []) }} + return (HsValBinds an2 valbinds') exact (HsIPBinds an bs) = do - (as, ipb) <- markAnnList an (markEpAnnL' an lal_rest AnnWhere - >> markAnnotated bs - >>= \bs' -> return (HsIPBinds an bs'::HsLocalBinds GhcPs)) - case ipb of - HsIPBinds _ bs' -> return (HsIPBinds as bs'::HsLocalBinds GhcPs) - _ -> error "should not happen HsIPBinds" + (an2,bs') <- markAnnListA an $ \an0 -> do + an1 <- markEpAnnL' an0 lal_rest AnnWhere + bs' <- markAnnotated bs + return (an1, bs') + return (HsIPBinds an2 bs') exact b@(EmptyLocalBinds _) = return b @@ -2670,7 +2736,8 @@ instance ExactPrint (HsValBindsLR GhcPs GhcPs) where let binds' = concatMap decl2Bind decls sigs' = concatMap decl2Sig decls - return (ValBinds sortKey binds' sigs') + sortKey' = captureOrderBinds decls + return (ValBinds sortKey' binds' sigs') exact (XValBindsLR _) = panic "XValBindsLR" undynamic :: Typeable a => [Dynamic] -> [a] @@ -2682,7 +2749,9 @@ instance ExactPrint (HsIPBinds GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact b@(IPBinds _ binds) = setLayoutBoth $ markAnnotated binds >> return b + exact (IPBinds x binds) = setLayoutBoth $ do + binds' <- markAnnotated binds + return (IPBinds x binds') -- --------------------------------------------------------------------- @@ -2703,18 +2772,18 @@ instance ExactPrint HsIPName where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact i@(HsIPName fs) = printStringAdvance ("?" ++ (unpackFS fs)) >> return i + exact i@(HsIPName fs) = printStringAdvanceA ("?" ++ (unpackFS fs)) >> return i -- --------------------------------------------------------------------- -- Managing lists which have been separated, e.g. Sigs and Binds prepareListAnnotationF :: (Monad m, Monoid w) => - [AddEpAnn] -> [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] -prepareListAnnotationF an ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls + [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] +prepareListAnnotationF ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls where go (L l a) = do - d' <- markAnnotated (DataFamInstDeclWithContext an NotTopLevel a) - return (toDyn (L l (dc_d d'))) + (L l' d') <- markAnnotated (L l (DataFamInstDeclWithContext noAnn NotTopLevel a)) + return (toDyn (L l' (dc_d d'))) prepareListAnnotationA :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => [LocatedAn an a] -> [(RealSrcSpan,EP w m Dynamic)] @@ -2725,15 +2794,23 @@ prepareListAnnotationA ls = map (\b -> (realSrcSpan $ getLocA b,go b)) ls return (toDyn b') withSortKey :: (Monad m, Monoid w) - => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] -> EP w m [Dynamic] + => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] + -> EP w m (AnnSortKey DeclTag, [Dynamic]) withSortKey annSortKey xs = do debugM $ "withSortKey:annSortKey=" ++ showAst annSortKey - let ordered = case annSortKey of - NoAnnSortKey -> sortBy orderByFst $ concatMap snd xs - AnnSortKey _keys -> orderedDecls annSortKey (Map.fromList xs) - mapM snd ordered -orderByFst :: Ord a => (a, b1) -> (a, b2) -> Ordering -orderByFst (a,_) (b,_) = compare a b + let (sk, ordered) = case annSortKey of + NoAnnSortKey -> (annSortKey', map snd os) + where + doOne (tag, ds) = map (\d -> (tag, d)) ds + xsExpanded = concatMap doOne xs + os = sortBy orderByFst $ xsExpanded + annSortKey' = AnnSortKey (map fst os) + AnnSortKey _keys -> (annSortKey, orderedDecls annSortKey (Map.fromList xs)) + ordered' <- mapM snd ordered + return (sk, ordered') + +orderByFst :: Ord a => (t, (a,b1)) -> (t, (a, b2)) -> Ordering +orderByFst (_,(a,_)) (_,(b,_)) = compare a b -- --------------------------------------------------------------------- @@ -2761,15 +2838,16 @@ instance ExactPrint (Sig GhcPs) where (an0, vars',ty') <- exactVarSig an vars ty return (ClassOpSig an0 is_deflt vars' ty') - exact (FixSig (an,src) (FixitySig x names (Fixity v fdir))) = do + exact (FixSig (an,src) (FixitySig ns names (Fixity v fdir))) = do let fixstr = case fdir of InfixL -> "infixl" InfixR -> "infixr" InfixN -> "infix" an0 <- markEpAnnLMS'' an lidl AnnInfix (Just fixstr) an1 <- markEpAnnLMS'' an0 lidl AnnVal (Just (sourceTextToString src (show v))) + ns' <- markAnnotated ns names' <- markAnnotated names - return (FixSig (an1,src) (FixitySig x names' (Fixity v fdir))) + return (FixSig (an1,src) (FixitySig ns' names' (Fixity v fdir))) exact (InlineSig an ln inl) = do an0 <- markAnnOpen an (inl_src inl) "{-# INLINE" @@ -2809,7 +2887,7 @@ instance ExactPrint (Sig GhcPs) where exact (CompleteMatchSig (an,src) cs mty) = do an0 <- markAnnOpen an src "{-# COMPLETE" - cs' <- markAnnotated cs + cs' <- mapM markAnnotated cs (an1, mty') <- case mty of Nothing -> return (an0, mty) @@ -2822,6 +2900,20 @@ instance ExactPrint (Sig GhcPs) where -- --------------------------------------------------------------------- +instance ExactPrint NamespaceSpecifier where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact NoNamespaceSpecifier = return NoNamespaceSpecifier + exact (TypeNamespaceSpecifier typeTok) = do + typeTok' <- markEpToken typeTok + return (TypeNamespaceSpecifier typeTok') + exact (DataNamespaceSpecifier dataTok) = do + dataTok' <- markEpToken dataTok + return (DataNamespaceSpecifier dataTok') + +-- --------------------------------------------------------------------- + exactVarSig :: (Monad m, Monoid w, ExactPrint a) => AnnSig -> [LocatedN RdrName] -> a -> EP w m (AnnSig, [LocatedN RdrName], a) exactVarSig an vars ty = do @@ -2875,7 +2967,7 @@ instance ExactPrint (AnnDecl GhcPs) where n' <- markAnnotated n return (an1, TypeAnnProvenance n') ModuleAnnProvenance -> do - an1 <- markEpAnnL an lapr_rest AnnModule + an1 <- markEpAnnL an0 lapr_rest AnnModule return (an1, prov) e' <- markAnnotated e @@ -2950,21 +3042,21 @@ instance ExactPrint (HsExpr GhcPs) where then markAnnotated n else return n return (HsVar x n') - exact x@(HsUnboundVar an _) = do + exact (HsUnboundVar an n) = do case an of Just (EpAnnUnboundVar (ob,cb) l) -> do - printStringAtAA ob "`" >> return () - printStringAtAA l "_" >> return () - printStringAtAA cb "`" >> return () - return x + ob' <- printStringAtAA ob "`" + l' <- printStringAtAA l "_" + cb' <- printStringAtAA cb "`" + return (HsUnboundVar (Just (EpAnnUnboundVar (ob',cb') l')) n) _ -> do - printStringAtLsDelta (SameLine 0) "_" - return x + printStringAdvanceA "_" >> return () + return (HsUnboundVar an n) exact x@(HsOverLabel src l) = do - printStringAtLsDelta (SameLine 0) "#" + printStringAdvanceA "#" >> return () case src of - NoSourceText -> printStringAtLsDelta (SameLine 0) (unpackFS l) - SourceText txt -> printStringAtLsDelta (SameLine 0) (unpackFS txt) + NoSourceText -> printStringAdvanceA (unpackFS l) >> return () + SourceText txt -> printStringAdvanceA (unpackFS txt) >> return () return x exact x@(HsIPVar _ (HsIPName n)) @@ -3204,11 +3296,11 @@ instance ExactPrint (HsExpr GhcPs) where exact (HsTypedSplice an s) = do an0 <- markEpAnnL an lidl AnnDollarDollar - s' <- exact s + s' <- markAnnotated s return (HsTypedSplice an0 s') exact (HsUntypedSplice an s) = do - s' <- exact s + s' <- markAnnotated s return (HsUntypedSplice an s') exact (HsProc an p c) = do @@ -3274,12 +3366,15 @@ exactMdo an (Just module_name) kw = markEpAnnLMS'' an lal_rest kw (Just n) markMaybeDodgyStmts :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => AnnList -> LocatedAn an a -> EP w m (AnnList, LocatedAn an a) markMaybeDodgyStmts an stmts = - if isGoodSrcSpan (getLocA stmts) + if notDodgy stmts then do r <- markAnnotatedWithLayout stmts return (an, r) else return (an, stmts) +notDodgy :: GenLocated (EpAnn ann) a -> Bool +notDodgy (L (EpAnn anc _ _) _) = notDodgyE anc + notDodgyE :: EpaLocation -> Bool notDodgyE anc = case anc of @@ -3341,7 +3436,7 @@ instance ExactPrint (MatchGroup GhcPs (LocatedA (HsCmd GhcPs))) where setAnnotationAnchor a _ _ _ = a exact (MG x matches) = do -- TODO:AZ use SortKey, in MG ann. - matches' <- if isGoodSrcSpan (getLocA matches) + matches' <- if notDodgy matches then markAnnotated matches else return matches return (MG x matches') @@ -3661,6 +3756,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- There may be arbitrary parens around parts of the constructor -- that are infix. Turn these into comments so that they feed -- into the right place automatically + -- TODO: no longer sorting on insert. What now? an0 <- annotationsToComments an lidl [AnnOpenP,AnnCloseP] an1 <- markEpAnnL an0 lidl AnnType @@ -3674,7 +3770,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- TODO: add a workaround for https://gitlab.haskell.org/ghc/ghc/-/issues/20452 exact (DataDecl { tcdDExt = an, tcdLName = ltycon, tcdTyVars = tyvars , tcdFixity = fixity, tcdDataDefn = defn }) = do - (_, an', ltycon', tyvars', _, _mctxt', defn') <- + (_, an', ltycon', tyvars', _, defn') <- exactDataDefn an (exactVanillaDeclHead ltycon tyvars fixity) defn return (DataDecl { tcdDExt = an', tcdLName = ltycon', tcdTyVars = tyvars' , tcdFixity = fixity, tcdDataDefn = defn' }) @@ -3707,7 +3803,7 @@ instance ExactPrint (TyClDecl GhcPs) where (an0, fds', lclas', tyvars',context') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lidl AnnSemi - ds <- withSortKey sortKey + (sortKey', ds) <- withSortKey sortKey [(ClsSigTag, prepareListAnnotationA sigs), (ClsMethodTag, prepareListAnnotationA methods), (ClsAtTag, prepareListAnnotationA ats), @@ -3720,7 +3816,7 @@ instance ExactPrint (TyClDecl GhcPs) where methods' = undynamic ds ats' = undynamic ds at_defs' = undynamic ds - return (ClassDecl {tcdCExt = (an3, lo, sortKey), + return (ClassDecl {tcdCExt = (an3, lo, sortKey'), tcdCtxt = context', tcdLName = lclas', tcdTyVars = tyvars', tcdFixity = fixity, tcdFDs = fds', @@ -3845,7 +3941,7 @@ exactDataDefn -> HsDataDefn GhcPs -> EP w m ( [AddEpAnn] -- ^ from exactHdr , [AddEpAnn] -- ^ updated one passed in - , LocatedN RdrName, a, b, Maybe (LHsContext GhcPs), HsDataDefn GhcPs) + , LocatedN RdrName, a, b, HsDataDefn GhcPs) exactDataDefn an exactHdr (HsDataDefn { dd_ext = x, dd_ctxt = context , dd_cType = mb_ct @@ -3883,8 +3979,8 @@ exactDataDefn an exactHdr _ -> panic "exacprint NewTypeCon" an6 <- markEpAnnL an5 lidl AnnCloseC derivings' <- mapM markAnnotated derivings - return (anx, an6, ln', tvs', b, mctxt', - (HsDataDefn { dd_ext = x, dd_ctxt = context + return (anx, an6, ln', tvs', b, + (HsDataDefn { dd_ext = x, dd_ctxt = mctxt' , dd_cType = mb_ct' , dd_kindSig = mb_sig' , dd_cons = condecls'', dd_derivs = derivings' })) @@ -3941,22 +4037,23 @@ instance ExactPrint (InjectivityAnn GhcPs) where class Typeable flag => ExactPrintTVFlag flag where exactTVDelimiters :: (Monad m, Monoid w) - => [AddEpAnn] -> flag -> EP w m (HsTyVarBndr flag GhcPs) - -> EP w m ([AddEpAnn], (HsTyVarBndr flag GhcPs)) + => [AddEpAnn] -> flag + -> ([AddEpAnn] -> EP w m ([AddEpAnn], HsTyVarBndr flag GhcPs)) + -> EP w m ([AddEpAnn], flag, (HsTyVarBndr flag GhcPs)) instance ExactPrintTVFlag () where - exactTVDelimiters an _ thing_inside = do + exactTVDelimiters an flag thing_inside = do an0 <- markEpAnnAllL' an lid AnnOpenP - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid AnnCloseP - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid AnnCloseP + return (an2, flag, r) instance ExactPrintTVFlag Specificity where exactTVDelimiters an s thing_inside = do an0 <- markEpAnnAllL' an lid open - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid close - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid close + return (an2, s, r) where (open, close) = case s of SpecifiedSpec -> (AnnOpenP, AnnCloseP) @@ -3964,33 +4061,33 @@ instance ExactPrintTVFlag Specificity where instance ExactPrintTVFlag (HsBndrVis GhcPs) where exactTVDelimiters an0 bvis thing_inside = do - case bvis of - HsBndrRequired _ -> return () - HsBndrInvisible at -> markEpToken at >> return () + bvis' <- case bvis of + HsBndrRequired _ -> return bvis + HsBndrInvisible at -> HsBndrInvisible <$> markEpToken at an1 <- markEpAnnAllL' an0 lid AnnOpenP - r <- thing_inside - an2 <- markEpAnnAllL' an1 lid AnnCloseP - return (an2, r) + (an2, r) <- thing_inside an1 + an3 <- markEpAnnAllL' an2 lid AnnCloseP + return (an3, bvis', r) instance ExactPrintTVFlag flag => ExactPrint (HsTyVarBndr flag GhcPs) where getAnnotationEntry _ = NoEntryVal setAnnotationAnchor a _ _ _ = a exact (UserTyVar an flag n) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - return (UserTyVar an flag n') + return (ani, UserTyVar an flag n') case r of - (an', UserTyVar _ flag'' n'') -> return (UserTyVar an' flag'' n'') + (an', flag', UserTyVar _ _ n'') -> return (UserTyVar an' flag' n'') _ -> error "KindedTyVar should never happen here" exact (KindedTyVar an flag n k) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - an0 <- markEpAnnL an lidl AnnDcolon + an0 <- markEpAnnL ani lidl AnnDcolon k' <- markAnnotated k - return (KindedTyVar an0 flag n' k') + return (an0, KindedTyVar an0 flag n' k') case r of - (an',KindedTyVar _ flag'' n'' k'') -> return (KindedTyVar an' flag'' n'' k'') + (an',flag', KindedTyVar _ _ n'' k'') -> return (KindedTyVar an' flag' n'' k'') _ -> error "UserTyVar should never happen here" -- --------------------------------------------------------------------- @@ -4150,17 +4247,16 @@ instance ExactPrint (HsDerivingClause GhcPs) where , deriv_clause_strategy = dcs , deriv_clause_tys = dct }) = do an0 <- markEpAnnL an lidl AnnDeriving - exact_strat_before + dcs0 <- case dcs of + Just (L _ ViaStrategy{}) -> return dcs + _ -> mapM markAnnotated dcs dct' <- markAnnotated dct - exact_strat_after + dcs1 <- case dcs0 of + Just (L _ ViaStrategy{}) -> mapM markAnnotated dcs0 + _ -> return dcs0 return (HsDerivingClause { deriv_clause_ext = an0 - , deriv_clause_strategy = dcs + , deriv_clause_strategy = dcs1 , deriv_clause_tys = dct' }) - where - (exact_strat_before, exact_strat_after) = - case dcs of - Just v@(L _ ViaStrategy{}) -> (pure (), markAnnotated v >> pure ()) - _ -> (mapM_ markAnnotated dcs, pure ()) -- --------------------------------------------------------------------- @@ -4467,7 +4563,9 @@ instance ExactPrint (ConDeclField GhcPs) where instance ExactPrint (FieldOcc GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact f@(FieldOcc _ n) = markAnnotated n >> return f + exact (FieldOcc x n) = do + n' <- markAnnotated n + return (FieldOcc x n') -- --------------------------------------------------------------------- @@ -4535,7 +4633,7 @@ instance ExactPrint (LocatedL [LocatedA (IE GhcPs)]) where an0 <- markEpAnnL' an lal_rest AnnHiding p <- getPosP debugM $ "LocatedL [LIE:p=" ++ showPprUnsafe p - (an1, ies') <- markAnnList an0 (markAnnotated ies) + (an1, ies') <- markAnnList an0 (markAnnotated (filter notIEDoc ies)) return (L an1 ies') instance (ExactPrint (Match GhcPs (LocatedA body))) @@ -4985,6 +5083,14 @@ setExtraDP md = do debugM $ "setExtraDP:" ++ show md modify (\s -> s {uExtraDP = md}) +getExtraDPReturn :: (Monad m, Monoid w) => EP w m (Maybe DeltaPos) +getExtraDPReturn = gets uExtraDPReturn + +setExtraDPReturn :: (Monad m, Monoid w) => Maybe DeltaPos -> EP w m () +setExtraDPReturn md = do + debugM $ "setExtraDPReturn:" ++ show md + modify (\s -> s {uExtraDPReturn = md}) + getPriorEndD :: (Monad m, Monoid w) => EP w m Pos getPriorEndD = gets dPriorEndPosition @@ -5007,13 +5113,13 @@ setPriorEndNoLayoutD pe = do debugM $ "setPriorEndNoLayoutD:pe=" ++ show pe modify (\s -> s { dPriorEndPosition = pe }) -setPriorEndASTD :: (Monad m, Monoid w) => Bool -> RealSrcSpan -> EP w m () -setPriorEndASTD layout pe = setPriorEndASTPD layout (rs2range pe) +setPriorEndASTD :: (Monad m, Monoid w) => RealSrcSpan -> EP w m () +setPriorEndASTD pe = setPriorEndASTPD (rs2range pe) -setPriorEndASTPD :: (Monad m, Monoid w) => Bool -> (Pos,Pos) -> EP w m () -setPriorEndASTPD layout pe@(fm,to) = do +setPriorEndASTPD :: (Monad m, Monoid w) => (Pos,Pos) -> EP w m () +setPriorEndASTPD pe@(fm,to) = do debugM $ "setPriorEndASTD:pe=" ++ show pe - when layout $ setLayoutStartD (snd fm) + setLayoutStartD (snd fm) modify (\s -> s { dPriorEndPosition = to } ) setLayoutStartD :: (Monad m, Monoid w) => Int -> EP w m () @@ -5044,7 +5150,7 @@ getUnallocatedComments :: (Monad m, Monoid w) => EP w m [Comment] getUnallocatedComments = gets epComments putUnallocatedComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -putUnallocatedComments cs = modify (\s -> s { epComments = cs } ) +putUnallocatedComments !cs = modify (\s -> s { epComments = cs } ) -- | Push a fresh stack frame for the applied comments gatherer pushAppliedComments :: (Monad m, Monoid w) => EP w m () @@ -5054,7 +5160,7 @@ pushAppliedComments = modify (\s -> s { epCommentsApplied = []:(epCommentsApplie -- takeAppliedComments, and clear them, not popping the stack takeAppliedComments :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedComments = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5067,7 +5173,7 @@ takeAppliedComments = do -- takeAppliedComments, and clear them, popping the stack takeAppliedCommentsPop :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedCommentsPop = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5080,7 +5186,7 @@ takeAppliedCommentsPop = do -- when doing delta processing applyComment :: (Monad m, Monoid w) => Comment -> EP w m () applyComment c = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> modify (\s -> s { epCommentsApplied = [[c]] } ) (h:t) -> modify (\s -> s { epCommentsApplied = (c:h):t } ) ===================================== utils/check-exact/Main.hs ===================================== @@ -470,7 +470,7 @@ changeAddDecl1 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtStart m decl' + replaceTopLevelDecls m = return $ insertAtStart m decl' return p' -- --------------------------------------------------------------------- @@ -483,7 +483,7 @@ changeAddDecl2 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtEnd m decl' + replaceTopLevelDecls m = return $ insertAtEnd m decl' return p' -- --------------------------------------------------------------------- @@ -500,7 +500,7 @@ changeAddDecl3 libdir top = do l2' = setEntryDP l2 (DifferentLine 2 0) replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAt f m decl' + replaceTopLevelDecls m = return $ insertAt f m decl' return p' -- --------------------------------------------------------------------- @@ -571,8 +571,9 @@ changeLocalDecls2 libdir (L l p) = do changeWhereIn3a :: Changer changeWhereIn3a _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) - debugM $ unlines w + decls = balanceCommentsList decls0 + (_de0:_:de1:_d2:_) = decls + debugM $ "changeWhereIn3a:de1:" ++ showAst de1 let p2 = p { hsmodDecls = decls} return (L l p2) @@ -581,13 +582,12 @@ changeWhereIn3a _libdir (L l p) = do changeWhereIn3b :: Changer changeWhereIn3b _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) + decls = balanceCommentsList decls0 (de0:tdecls@(_:de1:d2:_)) = decls de0' = setEntryDP de0 (DifferentLine 2 0) de1' = setEntryDP de1 (DifferentLine 2 0) d2' = setEntryDP d2 (DifferentLine 2 0) decls' = d2':de1':de0':tdecls - debugM $ unlines w debugM $ "changeWhereIn3b:de1':" ++ showAst de1' let p2 = p { hsmodDecls = decls'} return (L l p2) @@ -598,37 +598,37 @@ addLocaLDecl1 :: Changer addLocaLDecl1 libdir top = do Right (L ld (ValD _ decl)) <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let decl' = setEntryDP (L ld decl) (DifferentLine 1 5) - doAddLocal = do - let lp = top - (de1:d2:d3:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - (de1',_) <- modifyValD (getLocA de1'') de1'' $ \_m d -> do - return ((wrapDecl decl' : d),Nothing) - replaceDecls lp [de1', d2', d3] - - (lp',_,w) <- runTransformT doAddLocal - debugM $ "addLocaLDecl1:" ++ intercalate "\n" w + doAddLocal :: ParsedSource + doAddLocal = replaceDecls lp [de1', d2', d3] + where + lp = top + (de1:d2:d3:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 + (de1',_) = modifyValD (getLocA de1'') de1'' $ \_m d -> ((wrapDecl decl' : d),Nothing) + + let lp' = doAddLocal return lp' -- --------------------------------------------------------------------- + addLocaLDecl2 :: Changer addLocaLDecl2 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 + doAddLocal = replaceDecls lp [parent',d2'] + where + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - newDecl' <- transferEntryDP' d newDecl - let d' = setEntryDP d (DifferentLine 1 0) - return ((newDecl':d':ds),Nothing) + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = transferEntryDP' d (makeDeltaAst newDecl) + d' = setEntryDP d (DifferentLine 1 0) + in ((newDecl':d':ds),Nothing) - replaceDecls lp [parent',d2'] - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -637,19 +637,18 @@ addLocaLDecl3 :: Changer addLocaLDecl3 libdir top = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - let lp = top - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - return (((d:ds) ++ [newDecl']),Nothing) + doAddLocal = replaceDecls (anchorEof lp) [parent',d2'] + where + lp = top + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - replaceDecls (anchorEof lp) [parent',d2'] + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = setEntryDP newDecl (DifferentLine 1 0) + in (((d:ds) ++ [newDecl']),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -659,40 +658,38 @@ addLocaLDecl4 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") Right newSig <- withDynFlags libdir (\df -> parseDecl df "sig" "nn :: Int") let - doAddLocal = do - (parent:ds) <- hsDecls lp + doAddLocal = replaceDecls (anchorEof lp) (parent':ds) + where + (parent:ds) = hsDecls (makeDeltaAst lp) - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - let newSig' = setEntryDP newSig (DifferentLine 1 4) + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 0) + newSig' = setEntryDP (makeDeltaAst newSig) (DifferentLine 1 5) - (parent',_) <- modifyValD (getLocA parent) parent $ \_m decls -> do - return ((decls++[newSig',newDecl']),Nothing) + (parent',_) = modifyValD (getLocA parent) parent $ \_m decls -> + ((decls++[newSig',newDecl']),Nothing) - replaceDecls (anchorEof lp) (parent':ds) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' - -- --------------------------------------------------------------------- addLocaLDecl5 :: Changer addLocaLDecl5 _libdir lp = do let - doAddLocal = do - decls <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList decls + doAddLocal = replaceDecls lp [s1,de1',d3'] + where + decls = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList decls - let d3' = setEntryDP d3 (DifferentLine 2 0) + d3' = setEntryDP d3 (DifferentLine 2 0) - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m _decls -> do - let d2' = setEntryDP d2 (DifferentLine 1 0) - return ([d2'],Nothing) - replaceDecls lp [s1,de1',d3'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m _decls -> + let + d2' = setEntryDP d2 (DifferentLine 1 0) + in ([d2'],Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -701,39 +698,36 @@ addLocaLDecl6 :: Changer addLocaLDecl6 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "x = 3") let - newDecl' = setEntryDP newDecl (DifferentLine 1 4) - doAddLocal = do - decls0 <- hsDecls lp - [de1'',d2] <- balanceCommentsList decls0 + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 5) + doAddLocal = replaceDecls lp [de1', d2] + where + decls0 = hsDecls lp + [de1'',d2] = balanceCommentsList decls0 - let de1 = captureMatchLineSpacing de1'' - let L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 - let [ma1,_ma2] = ms + de1 = captureMatchLineSpacing de1'' + L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 + [ma1,_ma2] = ms - (de1',_) <- modifyValD (getLocA ma1) de1 $ \_m decls -> do - return ((newDecl' : decls),Nothing) - replaceDecls lp [de1', d2] + (de1',_) = modifyValD (getLocA ma1) de1 $ \_m decls -> + ((newDecl' : decls),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- rmDecl1 :: Changer -rmDecl1 _libdir top = do - let doRmDecl = do - let lp = top - tlDecs0 <- hsDecls lp - tlDecs' <- balanceCommentsList tlDecs0 - let tlDecs = captureLineSpacing tlDecs' - let (de1:_s1:_d2:d3:ds) = tlDecs - let d3' = setEntryDP d3 (DifferentLine 2 0) - - replaceDecls lp (de1:d3':ds) - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" +rmDecl1 _libdir lp = do + let + doRmDecl = replaceDecls lp (de1:d3':ds) + where + tlDecs0 = hsDecls lp + tlDecs = balanceCommentsList tlDecs0 + (de1:_s1:_d2:d3:ds) = tlDecs + d3' = setEntryDP d3 (DifferentLine 2 0) + + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -745,13 +739,13 @@ rmDecl2 _libdir lp = do let go :: GHC.LHsExpr GhcPs -> Transform (GHC.LHsExpr GhcPs) go e@(GHC.L _ (GHC.HsLet{})) = do - decs0 <- hsDecls e - decs <- balanceCommentsList $ captureLineSpacing decs0 - e' <- replaceDecls e (init decs) + let decs0 = hsDecls e + let decs = balanceCommentsList $ captureLineSpacing decs0 + let e' = replaceDecls e (init decs) return e' go x = return x - everywhereM (mkM go) lp + everywhereM (mkM go) (makeDeltaAst lp) let (lp',_,_w) = runTransform doRmDecl debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" @@ -762,17 +756,15 @@ rmDecl2 _libdir lp = do rmDecl3 :: Changer rmDecl3 _libdir lp = do let - doRmDecl = do - [de1,d2] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1] -> do - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([],Just sd1') - - replaceDecls lp [de1',sd1,d2] + doRmDecl = replaceDecls lp [de1',sd1,d2] + where + [de1,d2] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a] -> + let + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([],Just sd1') - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -780,19 +772,15 @@ rmDecl3 _libdir lp = do rmDecl4 :: Changer rmDecl4 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1,sd2] -> do - sd2' <- transferEntryDP' sd1 sd2 - - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([sd2'],Just sd1') - - replaceDecls (anchorEof lp) [de1',sd1] - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + doRmDecl = replaceDecls (anchorEof lp) [de1',sd1] + where + [de1] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a,sd2] -> + let + sd2' = transferEntryDP' sd1a sd2 + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([sd2'],Just sd1') + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -805,10 +793,8 @@ rmDecl5 _libdir lp = do go :: HsExpr GhcPs -> Transform (HsExpr GhcPs) go (HsLet (tkLet, tkIn) lb expr) = do let decs = hsDeclsLocalBinds lb - let hdecs : _ = decs let dec = last decs - _ <- transferEntryDP hdecs dec - lb' <- replaceDeclsValbinds WithoutWhere lb [dec] + let lb' = replaceDeclsValbinds WithoutWhere lb [dec] return (HsLet (tkLet, tkIn) lb' expr) go x = return x @@ -823,73 +809,61 @@ rmDecl5 _libdir lp = do rmDecl6 :: Changer rmDecl6 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m subDecs -> do - let subDecs' = captureLineSpacing subDecs - let (ss1:_sd1:sd2:sds) = subDecs' - sd2' <- transferEntryDP' ss1 sd2 - - return (sd2':sds,Nothing) + doRmDecl = replaceDecls lp [de1'] + where + [de1] = hsDecls lp - replaceDecls lp [de1'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m subDecs -> + let + subDecs' = captureLineSpacing subDecs + (ss1:_sd1:sd2:sds) = subDecs' + sd2' = transferEntryDP' ss1 sd2 + in (sd2':sds,Nothing) - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmDecl7 :: Changer -rmDecl7 _libdir top = do +rmDecl7 _libdir lp = do let - doRmDecl = do - let lp = top - tlDecs <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList tlDecs - - d3' <- transferEntryDP' d2 d3 - - replaceDecls lp [s1,de1,d3'] + doRmDecl = replaceDecls lp [s1,de1,d3'] + where + tlDecs = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList tlDecs + d3' = transferEntryDP' d2 d3 - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig1 :: Changer rmTypeSig1 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let (s0:de1:d2) = tlDecs - s1 = captureTypeSigSpacing s0 - (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 - L ln n2' <- transferEntryDP n1 n2 - let s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) - replaceDecls lp (s1':de1:d2) - - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let doRmDecl = replaceDecls lp (s1':de1:d2) + where + tlDecs = hsDecls lp + (s0:de1:d2) = tlDecs + s1 = captureTypeSigSpacing s0 + (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 + L ln n2' = transferEntryDP n1 n2 + s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig2 :: Changer rmTypeSig2 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let [de1] = tlDecs - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m [s,d] -> do - d' <- transferEntryDP' s d - return $ ([d'],Nothing) - `debug` ("rmTypeSig2:(d,d')" ++ showAst (d,d')) - replaceDecls lp [de1'] + let doRmDecl = replaceDecls lp [de1'] + where + tlDecs = hsDecls lp + [de1] = tlDecs + (de1',_) = modifyValD (getLocA de1) de1 $ \_m [_s,d] -> ([d],Nothing) - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -958,13 +932,15 @@ addClassMethod libdir lp = do let decl' = setEntryDP decl (DifferentLine 1 3) let sig' = setEntryDP sig (DifferentLine 2 3) let doAddMethod = do - [cd] <- hsDecls lp - (f1:f2s:f2d:_) <- hsDecls cd - let f2s' = setEntryDP f2s (DifferentLine 2 3) - cd' <- replaceDecls cd [f1, sig', decl', f2s', f2d] - replaceDecls lp [cd'] - - (lp',_,w) <- runTransformT doAddMethod + let + [cd] = hsDecls lp + (f1:f2s:f2d:_) = hsDecls cd + f2s' = setEntryDP f2s (DifferentLine 2 3) + cd' = replaceDecls cd [f1, sig', decl', f2s', f2d] + lp' = replaceDecls lp [cd'] + return lp' + + let (lp',_,w) = runTransform doAddMethod debugM $ "addClassMethod:" ++ intercalate "\n" w return lp' ===================================== utils/check-exact/Parsers.hs ===================================== @@ -260,7 +260,7 @@ parseModuleEpAnnsWithCppInternal cppOptions dflags file = do GHC.PFailed pst -> Left (GHC.GhcPsMessage <$> GHC.getPsErrorMessages pst) GHC.POk _ pmod - -> Right $ (injectedComments, dflags', fixModuleTrailingComments pmod) + -> Right $ (injectedComments, dflags', fixModuleComments pmod) -- | Internal function. Exposed if you want to muck with DynFlags -- before parsing. Or after parsing. @@ -269,8 +269,10 @@ postParseTransform -> Either a (GHC.ParsedSource) postParseTransform parseRes = fmap mkAnns parseRes where - -- TODO:AZ perhaps inject the comments into the parsedsource here already - mkAnns (_cs, _, m) = fixModuleTrailingComments m + mkAnns (_cs, _, m) = fixModuleComments m + +fixModuleComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleComments p = fixModuleHeaderComments $ fixModuleTrailingComments p fixModuleTrailingComments :: GHC.ParsedSource -> GHC.ParsedSource fixModuleTrailingComments (GHC.L l p) = GHC.L l p' @@ -293,6 +295,47 @@ fixModuleTrailingComments (GHC.L l p) = GHC.L l p' in cs'' _ -> cs +-- Deal with https://gitlab.haskell.org/ghc/ghc/-/issues/23984 +-- The Lexer works bottom-up, so does not have module declaration info +-- when the first top decl processed +fixModuleHeaderComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleHeaderComments (GHC.L l p) = GHC.L l p' + where + moveComments :: GHC.EpaLocation -> GHC.LHsDecl GHC.GhcPs -> GHC.EpAnnComments + -> (GHC.LHsDecl GHC.GhcPs, GHC.EpAnnComments) + moveComments GHC.EpaDelta{} dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.UnhelpfulSpan _)) dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.RealSrcSpan r _)) (GHC.L (GHC.EpAnn anc an csd) a) cs = (dd,css) + where + -- Move any comments on the decl that occur prior to the location + pc = GHC.priorComments csd + fc = GHC.getFollowingComments csd + bf (GHC.L anch _) = GHC.anchor anch > r + (move,keep) = break bf pc + csd' = GHC.EpaCommentsBalanced keep fc + + dd = GHC.L (GHC.EpAnn anc an csd') a + css = cs <> GHC.EpaComments move + + (ds',an') = rebalance (GHC.hsmodDecls p, GHC.hsmodAnn $ GHC.hsmodExt p) + p' = p { GHC.hsmodExt = (GHC.hsmodExt p){ GHC.hsmodAnn = an' }, + GHC.hsmodDecls = ds' + } + + rebalance :: ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + -> ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + rebalance (ds, GHC.EpAnn a an cs) = (ds1, GHC.EpAnn a an cs') + where + (ds1,cs') = case break (\(GHC.AddEpAnn k _) -> k == GHC.AnnWhere) (GHC.am_main an) of + (_, (GHC.AddEpAnn _ whereLoc:_)) -> + case GHC.hsmodDecls p of + (d:ds0) -> (d':ds0, cs0) + where (d',cs0) = moveComments whereLoc d cs + ds0 -> (ds0,cs) + _ -> (ds,cs) + + + -- | Internal function. Initializes DynFlags value for parsing. -- -- Passes "-hide-all-packages" to the GHC API to prevent parsing of ===================================== utils/check-exact/Transform.hs ===================================== @@ -63,7 +63,7 @@ module Transform -- *** Low level operations used in 'HasDecls' , balanceComments , balanceCommentsList - , balanceCommentsList' + , balanceCommentsListA , anchorEof -- ** Managing lists, pure functions @@ -92,6 +92,7 @@ import qualified Control.Monad.Fail as Fail import GHC hiding (parseModule, parsedSource) import GHC.Data.FastString +import GHC.Types.SrcLoc import Data.Data import Data.Maybe @@ -154,6 +155,7 @@ logDataWithAnnsTr str ast = do -- |If we need to add new elements to the AST, they need their own -- 'SrcSpan' for this. +-- This should no longer be needed, we use an @EpaDelta@ location instead. uniqueSrcSpanT :: (Monad m) => TransformT m SrcSpan uniqueSrcSpanT = do col <- get @@ -171,15 +173,6 @@ srcSpanStartLine' _ = 0 -- --------------------------------------------------------------------- -captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag -captureOrderBinds ls = AnnSortKey $ map go ls - where - go (L _ (ValD _ _)) = BindTag - go (L _ (SigD _ _)) = SigDTag - go d = error $ "captureOrderBinds:" ++ showGhc d - --- --------------------------------------------------------------------- - captureMatchLineSpacing :: LHsDecl GhcPs -> LHsDecl GhcPs captureMatchLineSpacing (L l (ValD x (FunBind a b (MG c (L d ms ))))) = L l (ValD x (FunBind a b (MG c (L d ms')))) @@ -253,7 +246,7 @@ setEntryDPDecl d dp = setEntryDP d dp -- |Set the true entry 'DeltaPos' from the annotation for a given AST -- element. This is the 'DeltaPos' ignoring any comments. -setEntryDP :: NoAnn t => LocatedAn t a -> DeltaPos -> LocatedAn t a +setEntryDP :: LocatedAn t a -> DeltaPos -> LocatedAn t a setEntryDP (L (EpAnn (EpaSpan ss@(UnhelpfulSpan _)) an cs) a) dp = L (EpAnn (EpaDelta ss dp []) an cs) a setEntryDP (L (EpAnn (EpaSpan ss) an (EpaComments [])) a) dp @@ -293,7 +286,7 @@ setEntryDP (L (EpAnn (EpaSpan ss@(RealSrcSpan r _)) an cs) a) dp L (EpAnn (EpaDelta ss edp csd) an cs'') a where cs'' = setPriorComments cs [] - csd = L (EpaDelta ss dp NoComments) c:cs' + csd = L (EpaDelta ss dp NoComments) c:commentOrigDeltas cs' lc = last $ (L ca c:cs') delta = case getLoc lc of EpaSpan (RealSrcSpan rr _) -> ss2delta (ss2pos rr) r @@ -335,18 +328,15 @@ setEntryDPFromAnchor off (EpaSpan (RealSrcSpan anc _)) ll@(L la _) = setEntryDP -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -transferEntryDP :: (Monad m, NoAnn t2, Typeable t1, Typeable t2) - => LocatedAn t1 a -> LocatedAn t2 b -> TransformT m (LocatedAn t2 b) -transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = do - logTr $ "transferEntryDP': EpAnn,EpAnn" +transferEntryDP :: (Typeable t1, Typeable t2) + => LocatedAn t1 a -> LocatedAn t2 b -> (LocatedAn t2 b) +transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = -- Problem: if the original had preceding comments, blindly -- transferring the location is not correct case priorComments cs1 of - [] -> return (L (EpAnn anc1 (combine an1 an2) cs2) b) + [] -> (L (EpAnn anc1 (combine an1 an2) cs2) b) -- TODO: what happens if the receiving side already has comments? - (L anc _:_) -> do - logDataWithAnnsTr "transferEntryDP':priorComments anc=" anc - return (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) + (L _ _:_) -> (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) -- |If a and b are the same type return first arg, else return second @@ -356,10 +346,11 @@ combine x y = fromMaybe y (cast x) -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -- TODO: call transferEntryDP, and use pushDeclDP -transferEntryDP' :: (Monad m) => LHsDecl GhcPs -> LHsDecl GhcPs -> TransformT m (LHsDecl GhcPs) -transferEntryDP' la lb = do - (L l2 b) <- transferEntryDP la lb - return (L l2 (pushDeclDP b (SameLine 0))) +transferEntryDP' :: LHsDecl GhcPs -> LHsDecl GhcPs -> (LHsDecl GhcPs) +transferEntryDP' la lb = + let + (L l2 b) = transferEntryDP la lb + in (L l2 (pushDeclDP b (SameLine 0))) pushDeclDP :: HsDecl GhcPs -> DeltaPos -> HsDecl GhcPs @@ -375,13 +366,24 @@ pushDeclDP d _dp = d -- --------------------------------------------------------------------- -balanceCommentsList :: (Monad m) => [LHsDecl GhcPs] -> TransformT m [LHsDecl GhcPs] -balanceCommentsList [] = return [] -balanceCommentsList [x] = return [x] -balanceCommentsList (a:b:ls) = do - (a',b') <- balanceComments a b - r <- balanceCommentsList (b':ls) - return (a':r) +-- | If we compile in haddock mode, the haddock processing inserts +-- DocDecls to carry the Haddock Documentation. We ignore these in +-- exact printing, as all the comments are also available in their +-- normal location, and the haddock processing is lossy, in that it +-- does not preserve all haddock-like comments. When we balance +-- comments in a list, we migrate some to preceding or following +-- declarations in the list. We must make sure we do not move any to +-- these DocDecls, which are not printed. +balanceCommentsList :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList decls = balanceCommentsList' (filter notDocDecl decls) + +balanceCommentsList' :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList' [] = [] +balanceCommentsList' [x] = [x] +balanceCommentsList' (a:b:ls) = (a':r) + where + (a',b') = balanceComments a b + r = balanceCommentsList' (b':ls) -- |The GHC parser puts all comments appearing between the end of one AST -- item and the beginning of the next as 'annPriorComments' for the second one. @@ -389,28 +391,27 @@ balanceCommentsList (a:b:ls) = do -- from the second one to the 'annFollowingComments' of the first if they belong -- to it instead. This is typically required before deleting or duplicating -- either of the AST elements. -balanceComments :: (Monad m) - => LHsDecl GhcPs -> LHsDecl GhcPs - -> TransformT m (LHsDecl GhcPs, LHsDecl GhcPs) -balanceComments first second = do +balanceComments :: LHsDecl GhcPs -> LHsDecl GhcPs + -> (LHsDecl GhcPs, LHsDecl GhcPs) +balanceComments first second = case first of - (L l (ValD x fb@(FunBind{}))) -> do - (L l' fb',second') <- balanceCommentsFB (L l fb) second - return (L l' (ValD x fb'), second') - _ -> balanceComments' first second + (L l (ValD x fb@(FunBind{}))) -> + let + (L l' fb',second') = balanceCommentsFB (L l fb) second + in (L l' (ValD x fb'), second') + _ -> balanceCommentsA first second --- |Once 'balanceComments' has been called to move trailing comments to a +-- |Once 'balanceCommentsA has been called to move trailing comments to a -- 'FunBind', these need to be pushed down from the top level to the last -- 'Match' if that 'Match' needs to be manipulated. -balanceCommentsFB :: (Monad m) - => LHsBind GhcPs -> LocatedA b -> TransformT m (LHsBind GhcPs, LocatedA b) -balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do - debugM $ "balanceCommentsFB entered: " ++ showGhc (ss2range $ locA lf) +balanceCommentsFB :: LHsBind GhcPs -> LocatedA b -> (LHsBind GhcPs, LocatedA b) +balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second + = balanceCommentsA (packFunBind bind) second' -- There are comments on lf. We need to -- + Keep the prior ones here -- + move the interior ones to the first match, -- + move the trailing ones to the last match. - let + where (before,middle,after) = case entry lf of EpaSpan (RealSrcSpan ss _) -> let @@ -426,40 +427,29 @@ balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do getFollowingComments $ comments lf) lf' = setCommentsEpAnn lf (EpaComments before) - debugM $ "balanceCommentsFB (before, after): " ++ showAst (before, after) - debugM $ "balanceCommentsFB lf': " ++ showAst lf' - -- let matches' = case matches of - let matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] - matches' = case matches of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaComments middle )) m':ms') - _ -> error "balanceCommentsFB" - matches'' <- balanceCommentsList' matches' - let (m,ms) = case reverse matches'' of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m',ms') - -- (L (addCommentsToEpAnnS lm' (EpaCommentsBalanced [] after)) m',ms') - _ -> error "balanceCommentsFB4" - debugM $ "balanceCommentsFB: (m,ms):" ++ showAst (m,ms) - (m',second') <- balanceComments' m second - m'' <- balanceCommentsMatch m' - let (m''',lf'') = case ms of - [] -> moveLeadingComments m'' lf' - _ -> (m'',lf') - debugM $ "balanceCommentsFB: (lf'', m'''):" ++ showAst (lf'',m''') - debugM $ "balanceCommentsFB done" - let bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) - debugM $ "balanceCommentsFB returning:" ++ showAst bind - balanceComments' (packFunBind bind) second' -balanceCommentsFB f s = balanceComments' f s + matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] + matches' = case matches of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaComments middle )) m0:ms') + _ -> error "balanceCommentsFB" + matches'' = balanceCommentsListA matches' + (m,ms) = case reverse matches'' of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m0,ms') + _ -> error "balanceCommentsFB4" + (m',second') = balanceCommentsA m second + m'' = balanceCommentsMatch m' + (m''',lf'') = case ms of + [] -> moveLeadingComments m'' lf' + _ -> (m'',lf') + bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) +balanceCommentsFB f s = balanceCommentsA f s -- | Move comments on the same line as the end of the match into the -- GRHS, prior to the binds -balanceCommentsMatch :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do - logTr $ "balanceCommentsMatch: (logInfo)=" ++ showAst (logInfo) - return (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) +balanceCommentsMatch :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) + = (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) where simpleBreak (r,_) = r /= 0 an1 = l @@ -468,7 +458,7 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do (move',stay') = break simpleBreak (trailingCommentsDeltas (anchorFromLocatedA (L l ())) cs1f) move = map snd move' stay = map snd stay' - (l'', grhss', binds', logInfo) + (l'', grhss', binds', _logInfo) = case reverse grhss of [] -> (l, [], binds, (EpaComments [], noSrcSpanA)) (L lg (GRHS ag grs rhs):gs) -> @@ -491,26 +481,24 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do pushTrailingComments :: WithWhere -> EpAnnComments -> HsLocalBinds GhcPs -> (Bool, HsLocalBinds GhcPs) pushTrailingComments _ _cs b at EmptyLocalBinds{} = (False, b) pushTrailingComments _ _cs (HsIPBinds _ _) = error "TODO: pushTrailingComments:HsIPBinds" -pushTrailingComments w cs lb@(HsValBinds an _) - = (True, HsValBinds an' vb) +pushTrailingComments w cs lb@(HsValBinds an _) = (True, HsValBinds an' vb) where decls = hsDeclsLocalBinds lb (an', decls') = case reverse decls of [] -> (addCommentsToEpAnn an cs, decls) (L la d:ds) -> (an, L (addCommentsToEpAnn la cs) d:ds) - (vb,_ws2) = case runTransform (replaceDeclsValbinds w lb (reverse decls')) of - ((HsValBinds _ vb'), _, ws2') -> (vb', ws2') - _ -> (ValBinds NoAnnSortKey [] [], []) + vb = case replaceDeclsValbinds w lb (reverse decls') of + (HsValBinds _ vb') -> vb' + _ -> ValBinds NoAnnSortKey [] [] -balanceCommentsList' :: (Monad m) => [LocatedA a] -> TransformT m [LocatedA a] -balanceCommentsList' [] = return [] -balanceCommentsList' [x] = return [x] -balanceCommentsList' (a:b:ls) = do - logTr $ "balanceCommentsList' entered" - (a',b') <- balanceComments' a b - r <- balanceCommentsList' (b':ls) - return (a':r) +balanceCommentsListA :: [LocatedA a] -> [LocatedA a] +balanceCommentsListA [] = [] +balanceCommentsListA [x] = [x] +balanceCommentsListA (a:b:ls) = (a':r) + where + (a',b') = balanceCommentsA a b + r = balanceCommentsListA (b':ls) -- |Prior to moving an AST element, make sure any trailing comments belonging to -- it are attached to it, and not the following element. Of necessity this is a @@ -518,13 +506,8 @@ balanceCommentsList' (a:b:ls) = do -- with a passed-in decision function. -- The initial situation is that all comments for a given anchor appear as prior comments -- Many of these should in fact be following comments for the previous anchor -balanceComments' :: (Monad m) => LocatedA a -> LocatedA b -> TransformT m (LocatedA a, LocatedA b) -balanceComments' la1 la2 = do - debugM $ "balanceComments': (anc1)=" ++ showAst (anc1) - debugM $ "balanceComments': (cs1s)=" ++ showAst (cs1s) - debugM $ "balanceComments': (cs1stay,cs1move)=" ++ showAst (cs1stay,cs1move) - debugM $ "balanceComments': (an1',an2')=" ++ showAst (an1',an2') - return (la1', la2') +balanceCommentsA :: LocatedA a -> LocatedA b -> (LocatedA a, LocatedA b) +balanceCommentsA la1 la2 = (la1', la2') where simpleBreak n (r,_) = r > n L an1 f = la1 @@ -532,26 +515,31 @@ balanceComments' la1 la2 = do anc1 = comments an1 anc2 = comments an2 - cs1s = splitCommentsEnd (anchorFromLocatedA la1) anc1 - cs1p = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1s) - cs1f = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1s) + (p1,m1,f1) = splitComments (anchorFromLocatedA la1) anc1 + cs1p = priorCommentsDeltas (anchorFromLocatedA la1) p1 - cs2s = splitCommentsEnd (anchorFromLocatedA la2) anc2 - cs2p = priorCommentsDeltas (anchorFromLocatedA la2) (priorComments cs2s) - cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) (getFollowingComments cs2s) + -- Split cs1 following comments into those before any + -- TrailingAnn's on an1, and any after + cs1f = splitCommentsEnd (fullSpanFromLocatedA la1) $ EpaComments f1 + cs1fp = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1f) + cs1ff = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1f) - -- Split cs1f into those that belong on an1 and ones that must move to an2 - (cs1move,cs1stay) = break (simpleBreak 1) cs1f + -- Split cs1ff into those that belong on an1 and ones that must move to an2 + (cs1move,cs1stay) = break (simpleBreak 1) cs1ff + + (p2,m2,f2) = splitComments (anchorFromLocatedA la2) anc2 + cs2p = priorCommentsDeltas (anchorFromLocatedA la2) p2 + cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) f2 (stay'',move') = break (simpleBreak 1) cs2p -- Need to also check for comments more closely attached to la1, -- ie trailing on the same line (move'',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchorFromLocatedA la1) (map snd stay'')) - move = sortEpaComments $ map snd (cs1move ++ move'' ++ move') - stay = sortEpaComments $ map snd (cs1stay ++ stay') + move = sortEpaComments $ map snd (cs1fp ++ cs1move ++ move'' ++ move') + stay = sortEpaComments $ m2 ++ map snd (cs1stay ++ stay') - an1' = setCommentsEpAnn (getLoc la1) (EpaCommentsBalanced (map snd cs1p) move) - an2' = setCommentsEpAnn (getLoc la2) (EpaCommentsBalanced stay (map snd cs2f)) + an1' = setCommentsEpAnn (getLoc la1) (epaCommentsBalanced (m1 ++ map snd cs1p) move) + an2' = setCommentsEpAnn (getLoc la2) (epaCommentsBalanced stay (map snd cs2f)) la1' = L an1' f la2' = L an2' s @@ -569,10 +557,9 @@ trailingCommentsDeltas r (la@(L l _):las) (al,_) = ss2posEnd rs' (ll,_) = ss2pos (anchor loc) --- AZ:TODO: this is identical to commentsDeltas priorCommentsDeltas :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] -priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) +priorCommentsDeltas r cs = go r (sortEpaComments cs) where go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] go _ [] = [] @@ -588,6 +575,21 @@ priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) -- --------------------------------------------------------------------- +-- | Split comments into ones occurring before the end of the reference +-- span, and those after it. +splitComments :: RealSrcSpan -> EpAnnComments -> ([LEpaComment], [LEpaComment], [LEpaComment]) +splitComments p cs = (before, middle, after) + where + cmpe (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p + cmpe (L _ _) = True + + cmpb (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2pos p + cmpb (L _ _) = True + + (beforeEnd, after) = break cmpe ((priorComments cs) ++ (getFollowingComments cs)) + (before, middle) = break cmpb beforeEnd + + -- | Split comments into ones occurring before the end of the reference -- span, and those after it. splitCommentsEnd :: RealSrcSpan -> EpAnnComments -> EpAnnComments @@ -598,8 +600,8 @@ splitCommentsEnd p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsEnd p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsEnd p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -617,8 +619,8 @@ splitCommentsStart p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsStart p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsStart p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -638,8 +640,8 @@ moveLeadingComments (L la a) lb = (L la' a, lb') -- TODO: need to set an entry delta on lb' to zero, and move the -- original spacing to the first comment. - la' = setCommentsEpAnn la (EpaCommentsBalanced [] after) - lb' = addCommentsToEpAnn lb (EpaCommentsBalanced before []) + la' = setCommentsEpAnn la (epaCommentsBalanced [] after) + lb' = addCommentsToEpAnn lb (epaCommentsBalanced before []) -- | A GHC comment includes the span of the preceding (non-comment) -- token. Takes an original list of comments, and converts the @@ -662,17 +664,27 @@ addCommentOrigDeltasAnn (EpAnn e a cs) = EpAnn e a (addCommentOrigDeltas cs) anchorFromLocatedA :: LocatedA a -> RealSrcSpan anchorFromLocatedA (L (EpAnn anc _ _) _) = anchor anc +-- | Get the full span of interest for comments from a LocatedA. +-- This extends up to the last TrailingAnn +fullSpanFromLocatedA :: LocatedA a -> RealSrcSpan +fullSpanFromLocatedA (L (EpAnn anc (AnnListItem tas) _) _) = rr + where + r = anchor anc + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + rr = case reverse (concatMap trailing_loc tas) of + [] -> r + (s:_) -> combineRealSrcSpans r s + -- --------------------------------------------------------------------- -balanceSameLineComments :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do - logTr $ "balanceSameLineComments: (la)=" ++ showGhc (ss2range $ locA la) - logTr $ "balanceSameLineComments: [logInfo]=" ++ showAst logInfo - return (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) +balanceSameLineComments :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) + = (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) where simpleBreak n (r,_) = r > n - (la',grhss', logInfo) = case reverse grhss of + (la',grhss', _logInfo) = case reverse grhss of [] -> (la,grhss,[]) (L lg (GRHS ga gs rhs):grs) -> (la'',reverse $ (L lg (GRHS ga' gs rhs)):grs,[(gac,(csp,csf))]) where @@ -684,7 +696,7 @@ balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do (move',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchor anc) csf) move = map snd move' stay = map snd stay' - cs1 = EpaCommentsBalanced csp stay + cs1 = epaCommentsBalanced csp stay gac = epAnnComments ga gfc = getFollowingComments gac @@ -734,24 +746,21 @@ addComma (EpAnn anc (AnnListItem as) cs) -- | Insert a declaration into an AST element having sub-declarations -- (@HasDecls@) according to the given location function. insertAt :: (HasDecls ast) - => (LHsDecl GhcPs - -> [LHsDecl GhcPs] - -> [LHsDecl GhcPs]) - -> ast - -> LHsDecl GhcPs - -> Transform ast -insertAt f t decl = do - oldDecls <- hsDecls t - oldDeclsb <- balanceCommentsList oldDecls - let oldDecls' = oldDeclsb - replaceDecls t (f decl oldDecls') + => (LHsDecl GhcPs + -> [LHsDecl GhcPs] + -> [LHsDecl GhcPs]) + -> ast + -> LHsDecl GhcPs + -> ast +insertAt f t decl = replaceDecls t (f decl oldDecls') + where + oldDecls = hsDecls t + oldDeclsb = balanceCommentsList oldDecls + oldDecls' = oldDeclsb -- |Insert a declaration at the beginning or end of the subdecls of the given -- AST item -insertAtStart, insertAtEnd :: (HasDecls ast) - => ast - -> LHsDecl GhcPs - -> Transform ast +insertAtStart, insertAtEnd :: HasDecls ast => ast -> LHsDecl GhcPs -> ast insertAtEnd = insertAt (\x xs -> xs ++ [x]) @@ -766,11 +775,11 @@ insertAtStart = insertAt insertFirst -- |Insert a declaration at a specific location in the subdecls of the given -- AST item -insertAfter, insertBefore :: (HasDecls (LocatedA ast)) +insertAfter, insertBefore :: HasDecls (LocatedA ast) => LocatedA old -> LocatedA ast -> LHsDecl GhcPs - -> Transform (LocatedA ast) + -> LocatedA ast insertAfter (getLocA -> k) = insertAt findAfter where findAfter x xs = @@ -797,10 +806,10 @@ class (Data t) => HasDecls t where -- given syntax phrase. They are always returned in the wrapped 'HsDecl' -- form, even if orginating in local decls. This is safe, as annotations -- never attach to the wrapper, only to the wrapped item. - hsDecls :: (Monad m) => t -> TransformT m [LHsDecl GhcPs] + hsDecls :: t -> [LHsDecl GhcPs] -- | Replace the directly enclosed decl list by the given - -- decl list. Runs in the 'Transform' monad to be able to update list order + -- decl list. As part of replacing it will update list order -- annotations, and rebalance comments and other layout changes as needed. -- -- For example, a call on replaceDecls for a wrapped 'FunBind' having no @@ -818,96 +827,86 @@ class (Data t) => HasDecls t where -- where -- nn = 2 -- @ - replaceDecls :: (Monad m) => t -> [LHsDecl GhcPs] -> TransformT m t + replaceDecls :: t -> [LHsDecl GhcPs] -> t -- --------------------------------------------------------------------- instance HasDecls ParsedSource where - hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = return decls + hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = decls replaceDecls (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps _decls)) decls - = do - logTr "replaceDecls LHsModule" - return (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) + = (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsDecl GhcPs)) where - hsDecls (L _ (TyClD _ c at ClassDecl{})) = return $ hsDeclsClassDecl c - hsDecls decl = do - error $ "hsDecls:decl=" ++ showAst decl - replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = do - let decl' = replaceDeclsClassDecl dec decls - return (L l (TyClD e decl')) - replaceDecls decl _decls = do - error $ "replaceDecls:decl=" ++ showAst decl + hsDecls (L _ (TyClD _ c at ClassDecl{})) = hsDeclsClassDecl c + hsDecls decl = error $ "hsDecls:decl=" ++ showAst decl + replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = + let + decl' = replaceDeclsClassDecl dec decls + in (L l (TyClD e decl')) + replaceDecls decl _decls + = error $ "replaceDecls:decl=" ++ showAst decl -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Match GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = hsDeclsLocalBinds lb replaceDecls (L l (Match xm c p (GRHSs xr rhs binds))) [] - = do - logTr "replaceDecls LMatch empty decls" - binds'' <- replaceDeclsValbinds WithoutWhere binds [] - return (L l (Match xm c p (GRHSs xr rhs binds''))) + = let + binds'' = replaceDeclsValbinds WithoutWhere binds [] + in (L l (Match xm c p (GRHSs xr rhs binds''))) replaceDecls m@(L l (Match xm c p (GRHSs xr rhs binds))) newBinds - = do - logTr "replaceDecls LMatch nonempty decls" + = let -- Need to throw in a fresh where clause if the binds were empty, -- in the annotations. - (l', rhs') <- case binds of - EmptyLocalBinds{} -> do - logTr $ "replaceDecls LMatch empty binds" - - logDataWithAnnsTr "Match.replaceDecls:balancing comments:m" m - L l' m' <- balanceSameLineComments m - logDataWithAnnsTr "Match.replaceDecls:(m1')" (L l' m') - return (l', grhssGRHSs $ m_grhss m') - _ -> return (l, rhs) - binds'' <- replaceDeclsValbinds WithWhere binds newBinds - logDataWithAnnsTr "Match.replaceDecls:binds'" binds'' - return (L l' (Match xm c p (GRHSs xr rhs' binds''))) + (l', rhs') = case binds of + EmptyLocalBinds{} -> + let + L l0 m' = balanceSameLineComments m + in (l0, grhssGRHSs $ m_grhss m') + _ -> (l, rhs) + binds'' = replaceDeclsValbinds WithWhere binds newBinds + in (L l' (Match xm c p (GRHSs xr rhs' binds''))) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsExpr GhcPs)) where - hsDecls (L _ (HsLet _ decls _ex)) = return $ hsDeclsLocalBinds decls - hsDecls _ = return [] + hsDecls (L _ (HsLet _ decls _ex)) = hsDeclsLocalBinds decls + hsDecls _ = [] replaceDecls (L ll (HsLet (tkLet, tkIn) binds ex)) newDecls - = do - logTr "replaceDecls HsLet" - let lastAnc = realSrcSpan $ spanHsLocaLBinds binds + = let + lastAnc = realSrcSpan $ spanHsLocaLBinds binds -- TODO: may be an intervening comment, take account for lastAnc - let (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of - (EpTok l, EpTok i) -> - let - off = case l of - (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r - (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 - (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 - (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c - ex'' = setEntryDPFromAnchor off i ex - newDecls'' = case newDecls of - [] -> newDecls - (d:ds) -> setEntryDPDecl d (SameLine 0) : ds - in ( EpTok l - , EpTok (addEpaLocationDelta off lastAnc i) - , ex'' - , newDecls'') - (_,_) -> (tkLet, tkIn, ex, newDecls) - binds' <- replaceDeclsValbinds WithoutWhere binds newDecls' - return (L ll (HsLet (tkLet', tkIn') binds' ex')) + (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of + (EpTok l, EpTok i) -> + let + off = case l of + (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r + (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 + (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 + (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c + ex'' = setEntryDPFromAnchor off i ex + newDecls'' = case newDecls of + [] -> newDecls + (d:ds) -> setEntryDPDecl d (SameLine 0) : ds + in ( EpTok l + , EpTok (addEpaLocationDelta off lastAnc i) + , ex'' + , newDecls'') + (_,_) -> (tkLet, tkIn, ex, newDecls) + binds' = replaceDeclsValbinds WithoutWhere binds newDecls' + in (L ll (HsLet (tkLet', tkIn') binds' ex')) -- TODO: does this make sense? Especially as no hsDecls for HsPar replaceDecls (L l (HsPar x e)) newDecls - = do - logTr "replaceDecls HsPar" - e' <- replaceDecls e newDecls - return (L l (HsPar x e')) + = let + e' = replaceDecls e newDecls + in (L l (HsPar x e')) replaceDecls old _new = error $ "replaceDecls (LHsExpr GhcPs) undefined for:" ++ showGhc old -- --------------------------------------------------------------------- @@ -934,53 +933,51 @@ hsDeclsPatBind x = error $ "hsDeclsPatBind called for:" ++ showGhc x -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBindD' \/ 'replaceDeclsPatBindD' is -- idempotent. -replaceDeclsPatBindD :: (Monad m) => LHsDecl GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsDecl GhcPs) -replaceDeclsPatBindD (L l (ValD x d)) newDecls = do - (L _ d') <- replaceDeclsPatBind (L l d) newDecls - return (L l (ValD x d')) +replaceDeclsPatBindD :: LHsDecl GhcPs -> [LHsDecl GhcPs] -> (LHsDecl GhcPs) +replaceDeclsPatBindD (L l (ValD x d)) newDecls = + let + (L _ d') = replaceDeclsPatBind (L l d) newDecls + in (L l (ValD x d')) replaceDeclsPatBindD x _ = error $ "replaceDeclsPatBindD called for:" ++ showGhc x -- | Replace the immediate declarations for a 'PatBind'. This -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBind' \/ 'replaceDeclsPatBind' is -- idempotent. -replaceDeclsPatBind :: (Monad m) => LHsBind GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsBind GhcPs) +replaceDeclsPatBind :: LHsBind GhcPs -> [LHsDecl GhcPs] -> (LHsBind GhcPs) replaceDeclsPatBind (L l (PatBind x a p (GRHSs xr rhss binds))) newDecls - = do - logTr "replaceDecls PatBind" - binds'' <- replaceDeclsValbinds WithWhere binds newDecls - return (L l (PatBind x a p (GRHSs xr rhss binds''))) + = (L l (PatBind x a p (GRHSs xr rhss binds''))) + where + binds'' = replaceDeclsValbinds WithWhere binds newDecls replaceDeclsPatBind x _ = error $ "replaceDeclsPatBind called for:" ++ showGhc x -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Stmt GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (LetStmt _ lb)) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (LetStmt _ lb)) = hsDeclsLocalBinds lb hsDecls (L _ (LastStmt _ e _ _)) = hsDecls e hsDecls (L _ (BindStmt _ _pat e)) = hsDecls e hsDecls (L _ (BodyStmt _ e _ _)) = hsDecls e - hsDecls _ = return [] + hsDecls _ = [] replaceDecls (L l (LetStmt x lb)) newDecls - = do - lb'' <- replaceDeclsValbinds WithWhere lb newDecls - return (L l (LetStmt x lb'')) + = let + lb'' = replaceDeclsValbinds WithWhere lb newDecls + in (L l (LetStmt x lb'')) replaceDecls (L l (LastStmt x e d se)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (LastStmt x e' d se)) + = let + e' = replaceDecls e newDecls + in (L l (LastStmt x e' d se)) replaceDecls (L l (BindStmt x pat e)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BindStmt x pat e')) + = let + e' = replaceDecls e newDecls + in (L l (BindStmt x pat e')) replaceDecls (L l (BodyStmt x e a b)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BodyStmt x e' a b)) - replaceDecls x _newDecls = return x + = let + e' = replaceDecls e newDecls + in (L l (BodyStmt x e' a b)) + replaceDecls x _newDecls = x -- ===================================================================== -- end of HasDecls instances @@ -1062,61 +1059,55 @@ data WithWhere = WithWhere -- care, as this does not manage the declaration order, the -- ordering should be done by the calling function from the 'HsLocalBinds' -- context in the AST. -replaceDeclsValbinds :: (Monad m) - => WithWhere +replaceDeclsValbinds :: WithWhere -> HsLocalBinds GhcPs -> [LHsDecl GhcPs] - -> TransformT m (HsLocalBinds GhcPs) -replaceDeclsValbinds _ _ [] = do - return (EmptyLocalBinds NoExtField) + -> HsLocalBinds GhcPs +replaceDeclsValbinds _ _ [] = EmptyLocalBinds NoExtField replaceDeclsValbinds w b@(HsValBinds a _) new - = do - logTr "replaceDeclsValbinds" - let oldSpan = spanHsLocaLBinds b - an <- oldWhereAnnotation a w (realSrcSpan oldSpan) - let decs = concatMap decl2Bind new - let sigs = concatMap decl2Sig new - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) + = let + oldSpan = spanHsLocaLBinds b + an = oldWhereAnnotation a w (realSrcSpan oldSpan) + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) replaceDeclsValbinds _ (HsIPBinds {}) _new = error "undefined replaceDecls HsIPBinds" replaceDeclsValbinds w (EmptyLocalBinds _) new - = do - logTr "replaceDecls HsLocalBinds" - an <- newWhereAnnotation w - let newBinds = concatMap decl2Bind new - newSigs = concatMap decl2Sig new - let decs = newBinds - let sigs = newSigs - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) - -oldWhereAnnotation :: (Monad m) - => EpAnn AnnList -> WithWhere -> RealSrcSpan -> TransformT m (EpAnn AnnList) -oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = do - -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, change the AnnList anchor to have the correct DP too - let (AnnList ancl o c _r t) = an - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - (anc', ancl') <- do - case ww of - WithWhere -> return (anc, ancl) - WithoutWhere -> return (anc, ancl) - let an' = EpAnn anc' - (AnnList ancl' o c w t) - cs - return an' - -newWhereAnnotation :: (Monad m) => WithWhere -> TransformT m (EpAnn AnnList) -newWhereAnnotation ww = do - let anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] - let anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - let an = EpAnn anc - (AnnList (Just anc2) Nothing Nothing w []) - emptyComments - return an + = let + an = newWhereAnnotation w + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) + +oldWhereAnnotation :: EpAnn AnnList -> WithWhere -> RealSrcSpan -> (EpAnn AnnList) +oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = an' + -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, + -- change the AnnList anchor to have the correct DP too + where + (AnnList ancl o c _r t) = an + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + (anc', ancl') = + case ww of + WithWhere -> (anc, ancl) + WithoutWhere -> (anc, ancl) + an' = EpAnn anc' + (AnnList ancl' o c w t) + cs + +newWhereAnnotation :: WithWhere -> (EpAnn AnnList) +newWhereAnnotation ww = an + where + anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] + anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + an = EpAnn anc + (AnnList (Just anc2) Nothing Nothing w []) + emptyComments -- --------------------------------------------------------------------- @@ -1127,32 +1118,32 @@ type PMatch = LMatch GhcPs (LHsExpr GhcPs) -- declarations are extracted and returned after modification. For a -- 'FunBind' the supplied 'SrcSpan' is used to identify the specific -- 'Match' to be transformed, for when there are multiple of them. -modifyValD :: forall m t. (HasTransform m) - => SrcSpan +modifyValD :: forall t. + SrcSpan -> Decl - -> (PMatch -> [Decl] -> m ([Decl], Maybe t)) - -> m (Decl,Maybe t) + -> (PMatch -> [Decl] -> ([Decl], Maybe t)) + -> (Decl,Maybe t) modifyValD p pb@(L ss (ValD _ (PatBind {} ))) f = if (locA ss) == p - then do - let ds = hsDeclsPatBindD pb - (ds',r) <- f (error "modifyValD.PatBind should not touch Match") ds - pb' <- liftT $ replaceDeclsPatBindD pb ds' - return (pb',r) - else return (pb,Nothing) -modifyValD p decl f = do - (decl',r) <- runStateT (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing - return (packFunDecl decl',r) + then + let + ds = hsDeclsPatBindD pb + (ds',r) = f (error "modifyValD.PatBind should not touch Match") ds + pb' = replaceDeclsPatBindD pb ds' + in (pb',r) + else (pb,Nothing) +modifyValD p decl f = (packFunDecl decl', r) where - doModLocal :: PMatch -> StateT (Maybe t) m PMatch + (decl',r) = runState (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing + doModLocal :: PMatch -> State (Maybe t) PMatch doModLocal (match@(L ss _) :: PMatch) = do if (locA ss) == p then do - ds <- lift $ liftT $ hsDecls match - `debug` ("modifyValD: match=" ++ showAst match) - (ds',r) <- lift $ f match ds - put r - match' <- lift $ liftT $ replaceDecls match ds' + let + ds = hsDecls match + (ds',r0) = f match ds + put r0 + let match' = replaceDecls match ds' return match' else return match @@ -1172,6 +1163,6 @@ modifyDeclsT :: (HasDecls t,HasTransform m) => ([LHsDecl GhcPs] -> m [LHsDecl GhcPs]) -> t -> m t modifyDeclsT action t = do - decls <- liftT $ hsDecls t + let decls = hsDecls t decls' <- action decls - liftT $ replaceDecls t decls' + return $ replaceDecls t decls' ===================================== utils/check-exact/Types.hs ===================================== @@ -21,10 +21,6 @@ type Pos = (Int,Int) -- --------------------------------------------------------------------- -data Rigidity = NormalLayout | RigidLayout deriving (Eq, Ord, Show) - --- --------------------------------------------------------------------- - -- | A Haskell comment. The @AnnKeywordId@ is present if it has been converted -- from an @AnnKeywordId@ because the annotation must be interleaved into the -- stream and does not have a well-defined position ===================================== utils/check-exact/Utils.hs ===================================== @@ -20,9 +20,8 @@ module Utils where import Control.Monad (when) +import GHC.Utils.Monad.State.Strict import Data.Function -import Data.Maybe (isJust) -import Data.Ord (comparing) import GHC.Hs.Dump import Lookup @@ -36,9 +35,10 @@ import GHC.Driver.Ppr import GHC.Data.FastString import qualified GHC.Data.Strict as Strict import GHC.Base (NonEmpty(..)) +import GHC.Parser.Lexer (allocateComments) import Data.Data hiding ( Fixity ) -import Data.List (sortBy, elemIndex) +import Data.List (sortBy, partition) import qualified Data.Map.Strict as Map import Debug.Trace @@ -60,12 +60,32 @@ debug c s = if debugEnabledFlag debugM :: Monad m => String -> m () debugM s = when debugEnabledFlag $ traceM s --- --------------------------------------------------------------------- - warn :: c -> String -> c -- warn = flip trace warn c _ = c +-- --------------------------------------------------------------------- + +captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag +captureOrderBinds ls = AnnSortKey $ map go ls + where + go (L _ (ValD _ _)) = BindTag + go (L _ (SigD _ _)) = SigDTag + go d = error $ "captureOrderBinds:" ++ showGhc d + +-- --------------------------------------------------------------------- + +notDocDecl :: LHsDecl GhcPs -> Bool +notDocDecl (L _ DocD{}) = False +notDocDecl _ = True + +notIEDoc :: LIE GhcPs -> Bool +notIEDoc (L _ IEGroup {}) = False +notIEDoc (L _ IEDoc {}) = False +notIEDoc (L _ IEDocNamed {}) = False +notIEDoc _ = True + +-- --------------------------------------------------------------------- -- | A good delta has no negative values. isGoodDelta :: DeltaPos -> Bool isGoodDelta (SameLine co) = co >= 0 @@ -108,7 +128,6 @@ pos2delta (refl,refc) (l,c) = deltaPos lo co lo = l - refl co = if lo == 0 then c - refc else c - -- else c - 1 -- | Apply the delta to the current position, taking into account the -- current column offset if advancing to a new line @@ -200,23 +219,6 @@ origDelta pos pp = ss2delta (ss2posEnd pp) pos -- --------------------------------------------------------------------- --- |Given a list of items and a list of keys, returns a list of items --- ordered by their position in the list of keys. -orderByKey :: [(DeclTag,a)] -> [DeclTag] -> [(DeclTag,a)] -orderByKey keys order - -- AZ:TODO: if performance becomes a problem, consider a Map of the order - -- SrcSpan to an index, and do a lookup instead of elemIndex. - - -- Items not in the ordering are placed to the start - = sortBy (comparing (flip elemIndex order . fst)) keys - --- --------------------------------------------------------------------- - -isListComp :: HsDoFlavour -> Bool -isListComp = isDoComprehensionContext - --- --------------------------------------------------------------------- - needsWhere :: DataDefnCons (LConDecl (GhcPass p)) -> Bool needsWhere (NewTypeCon _) = True needsWhere (DataTypeCons _ []) = True @@ -225,21 +227,214 @@ needsWhere _ = False -- --------------------------------------------------------------------- +-- | Insert the comments at the appropriate places in the AST insertCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource -insertCppComments (L l p) cs = L l p' +-- insertCppComments p [] = p +insertCppComments (L l p) cs0 = insertRemainingCppComments (L l p2) remaining + where + (EpAnn anct ant cst) = hsmodAnn $ hsmodExt p + cs = sortEpaComments $ priorComments cst ++ getFollowingComments cst ++ cs0 + p0 = p { hsmodExt = (hsmodExt p) { hsmodAnn = EpAnn anct ant emptyComments }} + -- Comments embedded within spans + -- everywhereM is a bottom-up traversal + (p1, toplevel) = runState (everywhereM (mkM addCommentsListItem + `extM` addCommentsGrhs + `extM` addCommentsList) p0) cs + (p2, remaining) = insertTopLevelCppComments p1 toplevel + + addCommentsListItem :: EpAnn AnnListItem -> State [LEpaComment] (EpAnn AnnListItem) + addCommentsListItem = addComments + + addCommentsList :: EpAnn AnnList -> State [LEpaComment] (EpAnn AnnList) + addCommentsList = addComments + + addCommentsGrhs :: EpAnn GrhsAnn -> State [LEpaComment] (EpAnn GrhsAnn) + addCommentsGrhs = addComments + + addComments :: forall ann. EpAnn ann -> State [LEpaComment] (EpAnn ann) + addComments (EpAnn anc an ocs) = do + case anc of + EpaSpan (RealSrcSpan s _) -> do + unAllocated <- get + let + (rest, these) = GHC.Parser.Lexer.allocateComments s unAllocated + cs' = workInComments ocs these + put rest + return $ EpAnn anc an cs' + + _ -> return $ EpAnn anc an ocs + +workInComments :: EpAnnComments -> [LEpaComment] -> EpAnnComments +workInComments ocs [] = ocs +workInComments ocs new = cs' + where + pc = priorComments ocs + fc = getFollowingComments ocs + cs' = case fc of + [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ new + (L ac _:_) -> epaCommentsBalanced (sortEpaComments $ pc ++ cs_before) + (sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) + new + +insertTopLevelCppComments :: HsModule GhcPs -> [LEpaComment] -> (HsModule GhcPs, [LEpaComment]) +insertTopLevelCppComments (HsModule (XModulePs an lo mdeprec mbDoc) mmn mexports imports decls) cs + = (HsModule (XModulePs an4 lo mdeprec mbDoc) mmn mexports' imports' decls', cs3) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0,hc1,hc_cs)) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0i,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0i,hc0,hc1,hc_cs)) where - an' = case GHC.hsmodAnn $ GHC.hsmodExt p of - (EpAnn a an ocs) -> EpAnn a an cs' - where - pc = priorComments ocs - fc = getFollowingComments ocs - cs' = case fc of - [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ cs - (L ac _:_) -> EpaCommentsBalanced (sortEpaComments $ pc ++ cs_before) - (sortEpaComments $ fc ++ cs_after) - where - (cs_before,cs_after) = break (\(L ll _) -> (ss2pos $ anchor ll) < (ss2pos $ anchor ac) ) cs + -- Comments at the top level. + (an0, cs0) = + case mmn of + Nothing -> (an, cs) + Just _ -> + -- We have a module name. Capture all comments up to the `where` + let + (these, remaining) = splitOnWhere Before (am_main $ anns an) cs + (EpAnn a anno ocs) = an :: EpAnn AnnsModule + anm = EpAnn a anno (workInComments ocs these) + in + (anm, remaining) + (an1,cs0a) = case lo of + EpExplicitBraces (EpTok (EpaSpan (RealSrcSpan s _))) _close -> + let + (stay,cs0a') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0 + cs' = workInComments (comments an0) stay + in (an0 { comments = cs' }, cs0a') + _ -> (an0,cs0) + -- Deal with possible leading semis + (an2, cs0b) = case am_decls $ anns an1 of + (AddSemiAnn (EpaSpan (RealSrcSpan s _)):_) -> (an1 {comments = cs'}, cs0b') + where + (stay,cs0b') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0a + cs' = workInComments (comments an1) stay + _ -> (an1,cs0a) + + (mexports', an3, cs1) = + case mexports of + Nothing -> (Nothing, an2, cs0b) + Just (L l exports) -> (Just (L l exports'), an3', cse) + where + hc1' = workInComments (comments an2) csh' + an3' = an2 { comments = hc1' } + (csh', cs0b') = case al_open $ anns l of + Just (AddEpAnn _ (EpaSpan (RealSrcSpan s _))) ->(h, n) + where + (h,n) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + cs0b + + _ -> ([], cs0b) + (exports', cse) = allocPreceding exports cs0b' + (imports0, cs2) = allocPreceding imports cs1 + (imports', hc0i) = balanceFirstLocatedAComments imports0 + + (decls0, cs3) = allocPreceding decls cs2 + (decls', hc0d) = balanceFirstLocatedAComments decls0 + + -- Either hc0i or hc0d should have comments. Combine them + hc0 = hc0i ++ hc0d + + (hc1,hc_cs) = if null ( am_main $ anns an3) + then (hc0,[]) + else splitOnWhere After (am_main $ anns an3) hc0 + hc2 = workInComments (comments an3) hc1 + an4 = an3 { anns = (anns an3) {am_cs = hc_cs}, comments = hc2 } + + allocPreceding :: [LocatedA a] -> [LEpaComment] -> ([LocatedA a], [LEpaComment]) + allocPreceding [] cs' = ([], cs') + allocPreceding (L (EpAnn anc4 an5 cs4) a:xs) cs' = ((L (EpAnn anc4 an5 cs4') a:xs'), rest') + where + (rest, these) = + case anc4 of + EpaSpan (RealSrcSpan s _) -> + allocatePriorComments (ss2pos s) cs' + _ -> (cs', []) + cs4' = workInComments cs4 these + (xs',rest') = allocPreceding xs rest + +data SplitWhere = Before | After +splitOnWhere :: SplitWhere -> [AddEpAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitOnWhere _ [] csIn = (csIn,[]) +splitOnWhere w (AddEpAnn AnnWhere (EpaSpan (RealSrcSpan s _)):_) csIn = (hc, fc) + where + splitFunc Before anc_pos c_pos = c_pos < anc_pos + splitFunc After anc_pos c_pos = anc_pos < c_pos + (hc,fc) = break (\(L ll _) -> splitFunc w (ss2pos $ anchor ll) (ss2pos s)) csIn +splitOnWhere _ (AddEpAnn AnnWhere _:_) csIn = (csIn, []) +splitOnWhere f (_:as) csIn = splitOnWhere f as csIn + +balanceFirstLocatedAComments :: [LocatedA a] -> ([LocatedA a], [LEpaComment]) +balanceFirstLocatedAComments [] = ([],[]) +balanceFirstLocatedAComments ((L (EpAnn anc an csd) a):ds) = (L (EpAnn anc an csd0) a:ds, hc') + where + (csd0, hc') = case anc of + EpaSpan (RealSrcSpan s _) -> (csd', hc) + `debug` ("balanceFirstLocatedAComments: (csd,csd',attached,header)=" ++ showAst (csd,csd',attached,header)) + where + (priors, inners) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + (priorComments csd) + pcds = priorCommentsDeltas' s priors + (attached, header) = break (\(d,_c) -> d /= 1) pcds + csd' = setPriorComments csd (reverse (map snd attached) ++ inners) + hc = reverse (map snd header) + _ -> (csd, []) + + + +priorCommentsDeltas' :: RealSrcSpan -> [LEpaComment] + -> [(Int, LEpaComment)] +priorCommentsDeltas' r cs = go r (reverse cs) + where + go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] + go _ [] = [] + go _ (la@(L l@(EpaDelta _ dp _) _):las) = (deltaLine dp, la) : go (anchor l) las + go rs' (la@(L l _):las) = deltaComment rs' la : go (anchor l) las + + deltaComment :: RealSrcSpan -> LEpaComment -> (Int, LEpaComment) + deltaComment rs' (L loc c) = (abs(ll - al), L loc c) + where + (al,_) = ss2pos rs' + (ll,_) = ss2pos (anchor loc) + +allocatePriorComments + :: Pos + -> [LEpaComment] + -> ([LEpaComment], [LEpaComment]) +allocatePriorComments ss_loc comment_q = + let + cmp (L l _) = ss2pos (anchor l) <= ss_loc + (newAnns,after) = partition cmp comment_q + in + (after, newAnns) + +insertRemainingCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource +insertRemainingCppComments (L l p) cs = L l p' + -- `debug` ("insertRemainingCppComments: (cs,an')=" ++ showAst (cs,an')) + where + (EpAnn a an ocs) = GHC.hsmodAnn $ GHC.hsmodExt p + an' = EpAnn a an (addTrailingComments end_loc ocs cs) p' = p { GHC.hsmodExt = (GHC.hsmodExt p) { GHC.hsmodAnn = an' } } + end_loc = case GHC.hsmodLayout $ GHC.hsmodExt p of + EpExplicitBraces _open close -> case close of + EpTok (EpaSpan (RealSrcSpan s _)) -> ss2pos s + _ -> (1,1) + _ -> (1,1) + (new_before, new_after) = break (\(L ll _) -> (ss2pos $ anchor ll) > end_loc ) cs + + addTrailingComments end_loc' cur new = epaCommentsBalanced pc' fc' + where + pc = priorComments cur + fc = getFollowingComments cur + (pc', fc') = case reverse pc of + [] -> (sortEpaComments $ pc ++ new_before, sortEpaComments $ fc ++ new_after) + (L ac _:_) -> (sortEpaComments $ pc ++ cs_before, sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = if (ss2pos $ anchor ac) > end_loc' + then break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) new + else (new_before, new_after) -- --------------------------------------------------------------------- @@ -291,11 +486,16 @@ dedentDocChunkBy dedent (L (RealSrcSpan l mb) c) = L (RealSrcSpan l' mb) c dedentDocChunkBy _ x = x + +epaCommentsBalanced :: [LEpaComment] -> [LEpaComment] -> EpAnnComments +epaCommentsBalanced priorCs [] = EpaComments priorCs +epaCommentsBalanced priorCs postCs = EpaCommentsBalanced priorCs postCs + mkEpaComments :: [Comment] -> [Comment] -> EpAnnComments mkEpaComments priorCs [] = EpaComments (map comment2LEpaComment priorCs) mkEpaComments priorCs postCs - = EpaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) + = epaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) comment2LEpaComment :: Comment -> LEpaComment comment2LEpaComment (Comment s anc r _mk) = mkLEpaComment s anc r @@ -330,18 +530,11 @@ sortEpaComments cs = sortBy cmp cs mkKWComment :: AnnKeywordId -> NoCommentsLocation -> Comment mkKWComment kw (EpaSpan (RealSrcSpan ss mb)) = Comment (keywordToString kw) (EpaSpan (RealSrcSpan ss mb)) ss (Just kw) -mkKWComment kw (EpaSpan ss@(UnhelpfulSpan _)) - = Comment (keywordToString kw) (EpaDelta ss (SameLine 0) NoComments) placeholderRealSpan (Just kw) +mkKWComment kw (EpaSpan (UnhelpfulSpan _)) + = Comment (keywordToString kw) (EpaDelta noSrcSpan (SameLine 0) NoComments) placeholderRealSpan (Just kw) mkKWComment kw (EpaDelta ss dp cs) = Comment (keywordToString kw) (EpaDelta ss dp cs) placeholderRealSpan (Just kw) --- | Detects a comment which originates from a specific keyword. -isKWComment :: Comment -> Bool -isKWComment c = isJust (commentOrigin c) - -noKWComments :: [Comment] -> [Comment] -noKWComments = filter (\c -> not (isKWComment c)) - sortAnchorLocated :: [GenLocated EpaLocation a] -> [GenLocated EpaLocation a] sortAnchorLocated = sortBy (compare `on` (anchor . getLoc)) @@ -379,11 +572,6 @@ name2String = showPprUnsafe -- --------------------------------------------------------------------- -locatedAnAnchor :: LocatedAn a t -> RealSrcSpan -locatedAnAnchor (L (EpAnn a _ _) _) = anchor a - --- --------------------------------------------------------------------- - trailingAnnLoc :: TrailingAnn -> EpaLocation trailingAnnLoc (AddSemiAnn ss) = ss trailingAnnLoc (AddCommaAnn ss) = ss @@ -401,46 +589,6 @@ setTrailingAnnLoc (AddDarrowUAnn _) ss = (AddDarrowUAnn ss) addEpAnnLoc :: AddEpAnn -> EpaLocation addEpAnnLoc (AddEpAnn _ l) = l --- --------------------------------------------------------------------- --- Horrible hack for dealing with some things still having a SrcSpan, --- not an Anchor. - -{- -A SrcSpan is defined as - -data SrcSpan = - RealSrcSpan !RealSrcSpan !(Maybe BufSpan) -- See Note [Why Maybe BufPos] - | UnhelpfulSpan !UnhelpfulSpanReason - -data BufSpan = - BufSpan { bufSpanStart, bufSpanEnd :: {-# UNPACK #-} !BufPos } - deriving (Eq, Ord, Show) - -newtype BufPos = BufPos { bufPos :: Int } - - -We use the BufPos to encode a delta, using bufSpanStart for the line, -and bufSpanEnd for the col. - -To be absolutely sure, we make the delta versions use -ve values. - --} - -hackSrcSpanToAnchor :: SrcSpan -> EpaLocation -hackSrcSpanToAnchor (UnhelpfulSpan s) = error $ "hackSrcSpanToAnchor : UnhelpfulSpan:" ++ show s -hackSrcSpanToAnchor ss@(RealSrcSpan r mb) - = case mb of - (Strict.Just (BufSpan (BufPos s) (BufPos e))) -> - if s <= 0 && e <= 0 - then EpaDelta ss (deltaPos (-s) (-e)) [] - `debug` ("hackSrcSpanToAnchor: (r,s,e)=" ++ showAst (r,s,e) ) - else EpaSpan (RealSrcSpan r mb) - _ -> EpaSpan (RealSrcSpan r mb) - -hackAnchorToSrcSpan :: EpaLocation -> SrcSpan -hackAnchorToSrcSpan (EpaSpan s) = s -hackAnchorToSrcSpan _ = error $ "hackAnchorToSrcSpan" - -- --------------------------------------------------------------------- type DeclsByTag a = Map.Map DeclTag [(RealSrcSpan, a)] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/cdf530df10ae9453fffb60f453e5dabc2df31c2c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/cdf530df10ae9453fffb60f453e5dabc2df31c2c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 02:22:50 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 12 Sep 2024 22:22:50 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 4 commits: finder: Add `IsBootInterface` to finder cache keys Message-ID: <66e3a1fae7e21_333c9c853ef085423@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 893d3869 by Sebastian Graf at 2024-09-12T22:21:58-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - b49daf5e by Andrew Lelechenko at 2024-09-12T22:21:59-04:00 Bump submodule array to 0.5.8.0 - - - - - 24 changed files: - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Types/Demand.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Finder/Types.hs - compiler/GHC/Unit/Module/Env.hs - compiler/GHC/Unit/Types.hs - libraries/array - + testsuite/tests/dmdanal/should_compile/T25196.hs - + testsuite/tests/dmdanal/should_compile/T25196_aux.hs - testsuite/tests/dmdanal/should_compile/all.T - + testsuite/tests/driver/boot-target/A.hs - + testsuite/tests/driver/boot-target/A.hs-boot - + testsuite/tests/driver/boot-target/B.hs - + testsuite/tests/driver/boot-target/Makefile - + testsuite/tests/driver/boot-target/all.T - utils/check-exact/ExactPrint.hs - utils/check-exact/Main.hs - utils/check-exact/Parsers.hs - utils/check-exact/Transform.hs - utils/check-exact/Types.hs - utils/check-exact/Utils.hs Changes: ===================================== compiler/GHC/Core/Opt/DmdAnal.hs ===================================== @@ -426,7 +426,7 @@ dmdAnalStar env (n :* sd) e , n' <- anticipateANF e n -- See Note [Anticipating ANF in demand analysis] -- and Note [Analysing with absent demand] - = (discardArgDmds $ multDmdType n' dmd_ty, e') + = (multDmdEnv n' (discardArgDmds dmd_ty), e') -- Main Demand Analysis machinery dmdAnal, dmdAnal' :: AnalEnv ===================================== compiler/GHC/Driver/Backpack.hs ===================================== @@ -781,7 +781,7 @@ summariseRequirement pn mod_name = do let loc = srcLocSpan (mkSrcLoc (mkFastString (bkp_filename env)) 1 1) let fc = hsc_FC hsc_env - mod <- liftIO $ addHomeModuleToFinder fc home_unit mod_name location + mod <- liftIO $ addHomeModuleToFinder fc home_unit (notBoot mod_name) location extra_sig_imports <- liftIO $ findExtraSigImports hsc_env HsigFile mod_name @@ -893,7 +893,7 @@ hsModuleToModSummary home_keys pn hsc_src modname this_mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit modname location + addHomeModuleToFinder fc home_unit (GWIB modname (hscSourceToIsBoot hsc_src)) location let ms = ModSummary { ms_mod = this_mod, ms_hsc_src = hsc_src, ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -2044,25 +2044,43 @@ summariseFile hsc_env' home_unit old_summaries src_fn mb_phase maybe_buf <- getPreprocessedImports hsc_env src_fn mb_phase maybe_buf let fopts = initFinderOpts (hsc_dflags hsc_env) - - -- Make a ModLocation for this file - let location = mkHomeModLocation fopts pi_mod_name (unsafeEncodeUtf src_fn) + src_path = unsafeEncodeUtf src_fn + + is_boot = case takeExtension src_fn of + ".hs-boot" -> IsBoot + ".lhs-boot" -> IsBoot + _ -> NotBoot + + (path_without_boot, hsc_src) + | isHaskellSigFilename src_fn = (src_path, HsigFile) + | IsBoot <- is_boot = (removeBootSuffix src_path, HsBootFile) + | otherwise = (src_path, HsSrcFile) + + -- Make a ModLocation for the Finder, who only has one entry for + -- each @ModuleName@, and therefore needs to use the locations for + -- the non-boot files. + location_without_boot = + mkHomeModLocation fopts pi_mod_name path_without_boot + + -- Make a ModLocation for this file, adding the @-boot@ suffix to + -- all paths if the original was a boot file. + location + | IsBoot <- is_boot + = addBootSuffixLocn location_without_boot + | otherwise + = location_without_boot -- Tell the Finder cache where it is, so that subsequent calls -- to findModule will find it, even if it's not on any search path mod <- liftIO $ do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit pi_mod_name location + addHomeModuleToFinder fc home_unit (GWIB pi_mod_name is_boot) location liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = NotBoot - , nms_hsc_src = - if isHaskellSigFilename src_fn - then HsigFile - else HsSrcFile + , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod , nms_preimps = preimps @@ -2090,9 +2108,10 @@ checkSummaryHash -- Also, only add to finder cache for non-boot modules as the finder cache -- makes sure to add a boot suffix for boot files. _ <- do - let fc = hsc_FC hsc_env + let fc = hsc_FC hsc_env + gwib = GWIB (ms_mod old_summary) (isBootSummary old_summary) case ms_hsc_src old_summary of - HsSrcFile -> addModuleToFinder fc (ms_mod old_summary) location + HsSrcFile -> addModuleToFinder fc gwib location _ -> return () hi_timestamp <- modificationTimeIfExists (ml_hi_file location) @@ -2230,7 +2249,6 @@ summariseModule hsc_env' home_unit old_summary_map is_boot (L _ wanted_mod) mb_p liftIO $ makeNewModSummary hsc_env $ MakeNewModSummary { nms_src_fn = src_fn , nms_src_hash = src_hash - , nms_is_boot = is_boot , nms_hsc_src = hsc_src , nms_location = location , nms_mod = mod @@ -2243,7 +2261,6 @@ data MakeNewModSummary = MakeNewModSummary { nms_src_fn :: FilePath , nms_src_hash :: Fingerprint - , nms_is_boot :: IsBootInterface , nms_hsc_src :: HscSource , nms_location :: ModLocation , nms_mod :: Module ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -734,7 +734,7 @@ runHscPhase pipe_env hsc_env0 input_fn src_flavour = do mod <- do let home_unit = hsc_home_unit hsc_env let fc = hsc_FC hsc_env - addHomeModuleToFinder fc home_unit mod_name location + addHomeModuleToFinder fc home_unit (GWIB mod_name (hscSourceToIsBoot src_flavour)) location -- Make the ModSummary to hand to hscMain let ===================================== compiler/GHC/Types/Demand.hs ===================================== @@ -51,7 +51,7 @@ module GHC.Types.Demand ( -- * Demand environments DmdEnv(..), addVarDmdEnv, mkTermDmdEnv, nopDmdEnv, plusDmdEnv, plusDmdEnvs, - reuseEnv, + multDmdEnv, reuseEnv, -- * Demand types DmdType(..), dmdTypeDepth, @@ -1910,7 +1910,8 @@ splitDmdTy ty at DmdType{dt_args=dmd:args} = (dmd, ty{dt_args=args}) splitDmdTy ty at DmdType{dt_env=env} = (defaultArgDmd (de_div env), ty) multDmdType :: Card -> DmdType -> DmdType -multDmdType n (DmdType fv args) +multDmdType C_11 dmd_ty = dmd_ty -- a vital optimisation for T25196 +multDmdType n (DmdType fv args) = -- pprTrace "multDmdType" (ppr n $$ ppr fv $$ ppr (multDmdEnv n fv)) $ DmdType (multDmdEnv n fv) (strictMap (multDmd n) args) ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -89,23 +89,23 @@ type BaseName = OsPath -- Basename of file initFinderCache :: IO FinderCache initFinderCache = do - mod_cache <- newIORef emptyInstalledModuleEnv + mod_cache <- newIORef emptyInstalledModuleWithIsBootEnv file_cache <- newIORef M.empty let flushFinderCaches :: UnitEnv -> IO () flushFinderCaches ue = do - atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleEnv is_ext fm, ()) + atomicModifyIORef' mod_cache $ \fm -> (filterInstalledModuleWithIsBootEnv is_ext fm, ()) atomicModifyIORef' file_cache $ \_ -> (M.empty, ()) where - is_ext mod _ = not (isUnitEnvInstalledModule ue mod) + is_ext mod _ = not (isUnitEnvInstalledModule ue (gwib_mod mod)) - addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () addToFinderCache key val = - atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleEnv c key val, ()) + atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleWithIsBootEnv c key val, ()) - lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) lookupFinderCache key = do c <- readIORef mod_cache - return $! lookupInstalledModuleEnv c key + return $! lookupInstalledModuleWithIsBootEnv c key lookupFileCache :: FilePath -> IO Fingerprint lookupFileCache key = do @@ -255,7 +255,7 @@ orIfNotFound this or_this = do homeSearchCache :: FinderCache -> UnitId -> ModuleName -> IO InstalledFindResult -> IO InstalledFindResult homeSearchCache fc home_unit mod_name do_this = do let mod = mkModule home_unit mod_name - modLocationCache fc mod do_this + modLocationCache fc (notBoot mod) do_this findExposedPackageModule :: FinderCache -> FinderOpts -> UnitState -> ModuleName -> PkgQual -> IO FindResult findExposedPackageModule fc fopts units mod_name mb_pkg = @@ -312,7 +312,7 @@ findLookupResult fc fopts r = case r of , fr_unusables = [] , fr_suggestions = suggest' }) -modLocationCache :: FinderCache -> InstalledModule -> IO InstalledFindResult -> IO InstalledFindResult +modLocationCache :: FinderCache -> InstalledModuleWithIsBoot -> IO InstalledFindResult -> IO InstalledFindResult modLocationCache fc mod do_this = do m <- lookupFinderCache fc mod case m of @@ -322,17 +322,17 @@ modLocationCache fc mod do_this = do addToFinderCache fc mod result return result -addModuleToFinder :: FinderCache -> Module -> ModLocation -> IO () +addModuleToFinder :: FinderCache -> ModuleWithIsBoot -> ModLocation -> IO () addModuleToFinder fc mod loc = do - let imod = toUnitId <$> mod - addToFinderCache fc imod (InstalledFound loc imod) + let imod = fmap toUnitId <$> mod + addToFinderCache fc imod (InstalledFound loc (gwib_mod imod)) -- This returns a module because it's more convenient for users -addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleName -> ModLocation -> IO Module +addHomeModuleToFinder :: FinderCache -> HomeUnit -> ModuleNameWithIsBoot -> ModLocation -> IO Module addHomeModuleToFinder fc home_unit mod_name loc = do - let mod = mkHomeInstalledModule home_unit mod_name - addToFinderCache fc mod (InstalledFound loc mod) - return (mkHomeModule home_unit mod_name) + let mod = mkHomeInstalledModule home_unit <$> mod_name + addToFinderCache fc mod (InstalledFound loc (gwib_mod mod)) + return (mkHomeModule home_unit (gwib_mod mod_name)) -- ----------------------------------------------------------------------------- -- The internal workers @@ -466,7 +466,7 @@ findPackageModule_ :: FinderCache -> FinderOpts -> InstalledModule -> UnitInfo - findPackageModule_ fc fopts mod pkg_conf = do massertPpr (moduleUnit mod == unitId pkg_conf) (ppr (moduleUnit mod) <+> ppr (unitId pkg_conf)) - modLocationCache fc mod $ + modLocationCache fc (notBoot mod) $ -- special case for GHC.Prim; we won't find it in the filesystem. if mod `installedModuleEq` gHC_PRIM ===================================== compiler/GHC/Unit/Finder/Types.hs ===================================== @@ -30,9 +30,9 @@ data FinderCache = FinderCache { flushFinderCaches :: UnitEnv -> IO () -- ^ remove all the home modules from the cache; package modules are -- assumed to not move around during a session; also flush the file hash -- cache. - , addToFinderCache :: InstalledModule -> InstalledFindResult -> IO () + , addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () -- ^ Add a found location to the cache for the module. - , lookupFinderCache :: InstalledModule -> IO (Maybe InstalledFindResult) + , lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) -- ^ Look for a location in the cache. , lookupFileCache :: FilePath -> IO Fingerprint -- ^ Look for the hash of a file in the cache. This should add it to the ===================================== compiler/GHC/Unit/Module/Env.hs ===================================== @@ -33,6 +33,17 @@ module GHC.Unit.Module.Env , mergeInstalledModuleEnv , plusInstalledModuleEnv , installedModuleEnvElts + + -- * InstalledModuleWithIsBootEnv + , InstalledModuleWithIsBootEnv + , emptyInstalledModuleWithIsBootEnv + , lookupInstalledModuleWithIsBootEnv + , extendInstalledModuleWithIsBootEnv + , filterInstalledModuleWithIsBootEnv + , delInstalledModuleWithIsBootEnv + , mergeInstalledModuleWithIsBootEnv + , plusInstalledModuleWithIsBootEnv + , installedModuleWithIsBootEnvElts ) where @@ -283,3 +294,56 @@ plusInstalledModuleEnv :: (elt -> elt -> elt) plusInstalledModuleEnv f (InstalledModuleEnv xm) (InstalledModuleEnv ym) = InstalledModuleEnv $ Map.unionWith f xm ym + + +-------------------------------------------------------------------- +-- InstalledModuleWithIsBootEnv +-------------------------------------------------------------------- + +-- | A map keyed off of 'InstalledModuleWithIsBoot' +newtype InstalledModuleWithIsBootEnv elt = InstalledModuleWithIsBootEnv (Map InstalledModuleWithIsBoot elt) + +instance Outputable elt => Outputable (InstalledModuleWithIsBootEnv elt) where + ppr (InstalledModuleWithIsBootEnv env) = ppr env + + +emptyInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a +emptyInstalledModuleWithIsBootEnv = InstalledModuleWithIsBootEnv Map.empty + +lookupInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> Maybe a +lookupInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = Map.lookup m e + +extendInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> a -> InstalledModuleWithIsBootEnv a +extendInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m x = InstalledModuleWithIsBootEnv (Map.insert m x e) + +filterInstalledModuleWithIsBootEnv :: (InstalledModuleWithIsBoot -> a -> Bool) -> InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBootEnv a +filterInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv e) = + InstalledModuleWithIsBootEnv (Map.filterWithKey f e) + +delInstalledModuleWithIsBootEnv :: InstalledModuleWithIsBootEnv a -> InstalledModuleWithIsBoot -> InstalledModuleWithIsBootEnv a +delInstalledModuleWithIsBootEnv (InstalledModuleWithIsBootEnv e) m = InstalledModuleWithIsBootEnv (Map.delete m e) + +installedModuleWithIsBootEnvElts :: InstalledModuleWithIsBootEnv a -> [(InstalledModuleWithIsBoot, a)] +installedModuleWithIsBootEnvElts (InstalledModuleWithIsBootEnv e) = Map.assocs e + +mergeInstalledModuleWithIsBootEnv + :: (elta -> eltb -> Maybe eltc) + -> (InstalledModuleWithIsBootEnv elta -> InstalledModuleWithIsBootEnv eltc) -- map X + -> (InstalledModuleWithIsBootEnv eltb -> InstalledModuleWithIsBootEnv eltc) -- map Y + -> InstalledModuleWithIsBootEnv elta + -> InstalledModuleWithIsBootEnv eltb + -> InstalledModuleWithIsBootEnv eltc +mergeInstalledModuleWithIsBootEnv f g h (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) + = InstalledModuleWithIsBootEnv $ Map.mergeWithKey + (\_ x y -> (x `f` y)) + (coerce g) + (coerce h) + xm ym + +plusInstalledModuleWithIsBootEnv :: (elt -> elt -> elt) + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt + -> InstalledModuleWithIsBootEnv elt +plusInstalledModuleWithIsBootEnv f (InstalledModuleWithIsBootEnv xm) (InstalledModuleWithIsBootEnv ym) = + InstalledModuleWithIsBootEnv $ Map.unionWith f xm ym + ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -86,6 +86,8 @@ module GHC.Unit.Types , GenWithIsBoot (..) , ModuleNameWithIsBoot , ModuleWithIsBoot + , InstalledModuleWithIsBoot + , notBoot ) where @@ -723,6 +725,8 @@ type ModuleNameWithIsBoot = GenWithIsBoot ModuleName type ModuleWithIsBoot = GenWithIsBoot Module +type InstalledModuleWithIsBoot = GenWithIsBoot InstalledModule + instance Binary a => Binary (GenWithIsBoot a) where put_ bh (GWIB { gwib_mod, gwib_isBoot }) = do put_ bh gwib_mod @@ -736,3 +740,6 @@ instance Outputable a => Outputable (GenWithIsBoot a) where ppr (GWIB { gwib_mod, gwib_isBoot }) = hsep $ ppr gwib_mod : case gwib_isBoot of IsBoot -> [ text "{-# SOURCE #-}" ] NotBoot -> [] + +notBoot :: mod -> GenWithIsBoot mod +notBoot gwib_mod = GWIB {gwib_mod, gwib_isBoot = NotBoot} ===================================== libraries/array ===================================== @@ -1 +1 @@ -Subproject commit ba5e9dcf1370190239395b8361b1c92ea9fc7632 +Subproject commit c9cb2c1e8762aa83b6e77af82c87a55e03e990e4 ===================================== testsuite/tests/dmdanal/should_compile/T25196.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25196 where + +import T25196_aux + +bar = $(gen 10000) ===================================== testsuite/tests/dmdanal/should_compile/T25196_aux.hs ===================================== @@ -0,0 +1,10 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25196_aux where + +import Language.Haskell.TH +import Language.Haskell.TH.Syntax +import Language.Haskell.TH.Lib + +gen :: Int -> Q Exp +gen n = lamE (replicate n (newName "x" >>= varP)) [| () |] ===================================== testsuite/tests/dmdanal/should_compile/all.T ===================================== @@ -98,3 +98,5 @@ test('T22388', [ grep_errmsg(r'^\S+\$w\S+') ], compile, ['-dsuppress-uniques -dd test('T22997', normal, compile, ['']) test('T23398', normal, compile, ['-dsuppress-uniques -ddump-simpl -dno-typeable-binds']) test('T24623', normal, compile, ['']) +test('T25196', [ req_th, collect_compiler_stats('bytes allocated', 10) ], + multimod_compile, ['T25196', '-v0']) ===================================== testsuite/tests/driver/boot-target/A.hs ===================================== @@ -0,0 +1,5 @@ +module A where + +import B + +data A = A B ===================================== testsuite/tests/driver/boot-target/A.hs-boot ===================================== @@ -0,0 +1,3 @@ +module A where + +data A ===================================== testsuite/tests/driver/boot-target/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} A + +data B = B A ===================================== testsuite/tests/driver/boot-target/Makefile ===================================== @@ -0,0 +1,8 @@ +boot1: + $(TEST_HC) -c A.hs-boot B.hs + +boot2: + $(TEST_HC) A.hs-boot A.hs B.hs -v0 + +boot3: + $(TEST_HC) A.hs-boot B.hs -v0 \ No newline at end of file ===================================== testsuite/tests/driver/boot-target/all.T ===================================== @@ -0,0 +1,10 @@ +def test_boot(name): + return test(name, + [extra_files(['A.hs', 'A.hs-boot', 'B.hs']), + ], + makefile_test, + []) + +test_boot('boot1') +test_boot('boot2') +test_boot('boot3') ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -1,3 +1,4 @@ +{-# LANGUAGE BangPatterns #-} {-# LANGUAGE DeriveDataTypeable #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE FlexibleInstances #-} @@ -25,7 +26,7 @@ module ExactPrint , makeDeltaAst -- * Configuration - , EPOptions(epRigidity, epAstPrint, epTokenPrint, epWhitespacePrint, epUpdateAnchors) + , EPOptions(epTokenPrint, epWhitespacePrint) , stringOptions , epOptions , deltaOptions @@ -43,10 +44,11 @@ import GHC.Types.ForeignCall import GHC.Types.Name.Reader import GHC.Types.PkgQual import GHC.Types.SourceText +import GHC.Types.SrcLoc import GHC.Types.Var -import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Unit.Module.Warnings import GHC.Utils.Misc +import GHC.Utils.Outputable hiding ( (<>) ) import GHC.Utils.Panic import Language.Haskell.Syntax.Basic (FieldLabelString(..)) @@ -77,8 +79,7 @@ import Types exactPrint :: ExactPrint ast => ast -> String exactPrint ast = snd $ runIdentity (runEP stringOptions (markAnnotated ast)) --- | The additional option to specify the rigidity and printing --- configuration. +-- | The additional option to specify the printing configuration. exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) => EPOptions m b -> ast @@ -86,9 +87,8 @@ exactPrintWithOptions :: (ExactPrint ast, Monoid b, Monad m) exactPrintWithOptions r ast = runEP r (markAnnotated ast) --- | Transform concrete annotations into relative annotations which --- are more useful when transforming an AST. This corresponds to the --- earlier 'relativiseApiAnns'. +-- | Transform concrete annotations into relative annotations. +-- This should be unnecessary from GHC 9.10 makeDeltaAst :: ExactPrint ast => ast -> ast makeDeltaAst ast = fst $ runIdentity (runEP deltaOptions (markAnnotated ast)) @@ -115,6 +115,7 @@ defaultEPState = EPState , dPriorEndPosition = (1,1) , uAnchorSpan = badRealSrcSpan , uExtraDP = Nothing + , uExtraDPReturn = Nothing , pAcceptSpan = False , epComments = [] , epCommentsApplied = [] @@ -128,39 +129,27 @@ defaultEPState = EPState -- | The R part of RWS. The environment. Updated via 'local' as we -- enter a new AST element, having a different anchor point. data EPOptions m a = EPOptions - { - epAstPrint :: forall ast . Data ast => GHC.Located ast -> a -> m a - , epTokenPrint :: String -> m a + { epTokenPrint :: String -> m a , epWhitespacePrint :: String -> m a - , epRigidity :: Rigidity - , epUpdateAnchors :: Bool } -- | Helper to create a 'EPOptions' -epOptions :: - (forall ast . Data ast => GHC.Located ast -> a -> m a) - -> (String -> m a) - -> (String -> m a) - -> Rigidity - -> Bool - -> EPOptions m a -epOptions astPrint tokenPrint wsPrint rigidity delta = EPOptions - { - epAstPrint = astPrint - , epWhitespacePrint = wsPrint +epOptions :: (String -> m a) + -> (String -> m a) + -> EPOptions m a +epOptions tokenPrint wsPrint = EPOptions + { epWhitespacePrint = wsPrint , epTokenPrint = tokenPrint - , epRigidity = rigidity - , epUpdateAnchors = delta } -- | Options which can be used to print as a normal String. stringOptions :: EPOptions Identity String -stringOptions = epOptions (\_ b -> return b) return return NormalLayout False +stringOptions = epOptions return return -- | Options which can be used to simply update the AST to be in delta -- form, without generating output deltaOptions :: EPOptions Identity () -deltaOptions = epOptions (\_ _ -> return ()) (\_ -> return ()) (\_ -> return ()) NormalLayout True +deltaOptions = epOptions (\_ -> return ()) (\_ -> return ()) data EPWriter a = EPWriter { output :: !a } @@ -177,6 +166,8 @@ data EPState = EPState -- Annotation , uExtraDP :: !(Maybe EpaLocation) -- ^ Used to anchor a -- list + , uExtraDPReturn :: !(Maybe DeltaPos) + -- ^ Used to return Delta version of uExtraDP , pAcceptSpan :: Bool -- ^ When we have processed an -- entry of EpaDelta, accept the -- next `EpaSpan` start as the @@ -213,7 +204,7 @@ class HasTrailing a where trailing :: a -> [TrailingAnn] setTrailing :: a -> [TrailingAnn] -> a -setAnchorEpa :: (HasTrailing an, NoAnn an) +setAnchorEpa :: (HasTrailing an) => EpAnn an -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> EpAnn an setAnchorEpa (EpAnn _ an _) anc ts cs = EpAnn anc (setTrailing an ts) cs @@ -223,7 +214,7 @@ setAnchorHsModule hsmod anc cs = hsmod { hsmodExt = (hsmodExt hsmod) {hsmodAnn = anc' = anc an' = setAnchorEpa (hsmodAnn $ hsmodExt hsmod) anc' [] cs -setAnchorAn :: (HasTrailing an, NoAnn an) +setAnchorAn :: (HasTrailing an) => LocatedAn an a -> EpaLocation -> [TrailingAnn] -> EpAnnComments -> LocatedAn an a setAnchorAn (L (EpAnn _ an _) a) anc ts cs = (L (EpAnn anc (setTrailing an ts) cs) a) -- `debug` ("setAnchorAn: anc=" ++ showAst anc) @@ -248,7 +239,7 @@ data FlushComments = FlushComments data CanUpdateAnchor = CanUpdateAnchor | CanUpdateAnchorOnly | NoCanUpdateAnchor - deriving (Eq, Show) + deriving (Eq, Show, Data) data Entry = Entry EpaLocation [TrailingAnn] EpAnnComments FlushComments CanUpdateAnchor | NoEntryVal @@ -402,7 +393,7 @@ enterAnn NoEntryVal a = do r <- exact a debugM $ "enterAnn:done:NO ANN:p =" ++ show (p, astId a) return r -enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do +enterAnn !(Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do acceptSpan <- getAcceptSpan setAcceptSpan False case anchor' of @@ -421,9 +412,11 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do _ -> return () case anchor' of EpaDelta _ _ dcs -> do - debugM $ "enterAnn:Printing comments:" ++ showGhc (priorComments cs) + debugM $ "enterAnn:Delta:Flushing comments" + flushComments [] + debugM $ "enterAnn:Delta:Printing prior comments:" ++ showGhc (priorComments cs) mapM_ printOneComment (concatMap tokComment $ priorComments cs) - debugM $ "enterAnn:Printing EpaDelta comments:" ++ showGhc dcs + debugM $ "enterAnn:Delta:Printing EpaDelta comments:" ++ showGhc dcs mapM_ printOneComment (concatMap tokComment dcs) _ -> do debugM $ "enterAnn:Adding comments:" ++ showGhc (priorComments cs) @@ -465,7 +458,7 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- The first part corresponds to the delta phase, so should only use -- delta phase variables ----------------------------------- -- Calculate offset required to get to the start of the SrcSPan - off <- getLayoutOffsetD + !off <- getLayoutOffsetD let spanStart = ss2pos curAnchor priorEndAfterComments <- getPriorEndD let edp' = adjustDeltaForOffset @@ -480,17 +473,18 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- --------------------------------------------- med <- getExtraDP setExtraDP Nothing - let edp = case med of - Nothing -> edp'' - Just (EpaDelta _ dp _) -> dp + let (edp, medr) = case med of + Nothing -> (edp'', Nothing) + Just (EpaDelta _ dp _) -> (dp, Nothing) -- Replace original with desired one. Allows all -- list entry values to be DP (1,0) - Just (EpaSpan (RealSrcSpan r _)) -> dp + Just (EpaSpan (RealSrcSpan r _)) -> (dp, Just dp) where dp = adjustDeltaForOffset off (ss2delta priorEndAfterComments r) Just (EpaSpan (UnhelpfulSpan r)) -> panic $ "enterAnn: UnhelpfulSpan:" ++ show r when (isJust med) $ debugM $ "enterAnn:(med,edp)=" ++ showAst (med,edp) + when (isJust medr) $ setExtraDPReturn medr -- --------------------------------------------- -- Preparation complete, perform the action when (priorEndAfterComments < spanStart) (do @@ -511,12 +505,15 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do debugM $ "enterAnn:exact a starting:" ++ show (showAst anchor') a' <- exact a debugM $ "enterAnn:exact a done:" ++ show (showAst anchor') + + -- Core recursive exactprint done, start end of Entry processing + when (flush == FlushComments) $ do - debugM $ "flushing comments in enterAnn:" ++ showAst cs + debugM $ "flushing comments in enterAnn:" ++ showAst (cs, getFollowingComments cs) flushComments (getFollowingComments cs) debugM $ "flushing comments in enterAnn done" - eof <- getEofPos + !eof <- getEofPos case eof of Nothing -> return () Just (pos, prior) -> do @@ -544,28 +541,50 @@ enterAnn (Entry anchor' trailing_anns cs flush canUpdateAnchor) a = do -- Outside the anchor, mark any trailing postCs <- cua canUpdateAnchor takeAppliedCommentsPop - when (flush == NoFlushComments) $ do - when ((getFollowingComments cs) /= []) $ do - - -- debugM $ "enterAnn:in:(anchor') =" ++ show (eloc2str anchor') - debugM $ "starting trailing comments:" ++ showAst (getFollowingComments cs) - mapM_ printOneComment (concatMap tokComment $ getFollowingComments cs) - debugM $ "ending trailing comments" - trailing' <- markTrailing trailing_anns + following <- if (flush == NoFlushComments) + then do + let (before, after) = splitAfterTrailingAnns trailing_anns + (getFollowingComments cs) + addCommentsA before + return after + else return [] + !trailing' <- markTrailing trailing_anns + -- mapM_ printOneComment (concatMap tokComment $ following) + addCommentsA following -- Update original anchor, comments based on the printing process -- TODO:AZ: probably need to put something appropriate in instead of noSrcSpan - let newAchor = EpaDelta noSrcSpan edp [] + let newAnchor = EpaDelta noSrcSpan edp [] let r = case canUpdateAnchor of - CanUpdateAnchor -> setAnnotationAnchor a' newAchor trailing' (mkEpaComments (priorCs ++ postCs) []) - CanUpdateAnchorOnly -> setAnnotationAnchor a' newAchor [] emptyComments + CanUpdateAnchor -> setAnnotationAnchor a' newAnchor trailing' (mkEpaComments priorCs postCs) + CanUpdateAnchorOnly -> setAnnotationAnchor a' newAnchor [] emptyComments NoCanUpdateAnchor -> a' return r -- --------------------------------------------------------------------- +-- | Split the span following comments into ones that occur prior to +-- the last trailing ann, and ones after. +splitAfterTrailingAnns :: [TrailingAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitAfterTrailingAnns [] cs = ([], cs) +splitAfterTrailingAnns tas cs = (before, after) + where + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + (before, after) = case reverse (concatMap trailing_loc tas) of + [] -> ([],cs) + (s:_) -> (b,a) + where + s_pos = ss2pos s + (b,a) = break (\(L ll _) -> (ss2pos $ anchor ll) > s_pos) + cs + + +-- --------------------------------------------------------------------- + addCommentsA :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -addCommentsA csNew = addComments (concatMap tokComment csNew) +addCommentsA csNew = addComments False (concatMap tokComment csNew) {- TODO: When we addComments, some may have an anchor that is no longer @@ -583,24 +602,36 @@ By definition it is the current anchor, so work against that. And that also means that the first entry comment that has moved should not have a line offset. -} -addComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -addComments csNew = do - -- debugM $ "addComments:" ++ show csNew +addComments :: (Monad m, Monoid w) => Bool -> [Comment] -> EP w m () +addComments sortNeeded csNew = do + debugM $ "addComments:csNew" ++ show csNew cs <- getUnallocatedComments + debugM $ "addComments:cs" ++ show cs + -- We can only sort the comments if we are in the first phase, + -- where all comments have locations. If any have EpaDelta the + -- sort will fail, so we do not try. + if sortNeeded && all noDelta (csNew ++ cs) + then putUnallocatedComments (sort (cs ++ csNew)) + else putUnallocatedComments (cs ++ csNew) - putUnallocatedComments (sort (cs ++ csNew)) +noDelta :: Comment -> Bool +noDelta c = case commentLoc c of + EpaSpan _ -> True + _ -> False -- --------------------------------------------------------------------- -- | Just before we print out the EOF comments, flush the remaining -- ones in the state. flushComments :: (Monad m, Monoid w) => [LEpaComment] -> EP w m () -flushComments trailing_anns = do +flushComments !trailing_anns = do + debugM $ "flushComments entered: " ++ showAst trailing_anns addCommentsA trailing_anns + debugM $ "flushComments after addCommentsA" cs <- getUnallocatedComments - debugM $ "flushing comments starting" - -- AZ:TODO: is the sort still needed? - mapM_ printOneComment (sortComments cs) + debugM $ "flushComments: got cs" + debugM $ "flushing comments starting: cs" ++ showAst cs + mapM_ printOneComment cs putUnallocatedComments [] debugM $ "flushing comments done" @@ -612,7 +643,7 @@ annotationsToComments :: (Monad m, Monoid w) => a -> Lens a [AddEpAnn] -> [AnnKeywordId] -> EP w m a annotationsToComments a l kws = do let (newComments, newAnns) = go ([],[]) (view l a) - addComments newComments + addComments True newComments return (set l (reverse newAnns) a) where keywords = Set.fromList kws @@ -654,14 +685,11 @@ printSourceText (NoSourceText) txt = printStringAdvance txt >> return () printSourceText (SourceText txt) _ = printStringAdvance (unpackFS txt) >> return () printSourceTextAA :: (Monad m, Monoid w) => SourceText -> String -> EP w m () -printSourceTextAA (NoSourceText) txt = printStringAtAA noAnn txt >> return () -printSourceTextAA (SourceText txt) _ = printStringAtAA noAnn (unpackFS txt) >> return () +printSourceTextAA (NoSourceText) txt = printStringAdvanceA txt >> return () +printSourceTextAA (SourceText txt) _ = printStringAdvanceA (unpackFS txt) >> return () -- --------------------------------------------------------------------- -printStringAtSs :: (Monad m, Monoid w) => SrcSpan -> String -> EP w m () -printStringAtSs ss str = printStringAtRs (realSrcSpan ss) str >> return () - printStringAtRs :: (Monad m, Monoid w) => RealSrcSpan -> String -> EP w m EpaLocation printStringAtRs pa str = printStringAtRsC CaptureComments pa str @@ -676,7 +704,7 @@ printStringAtRsC capture pa str = do p' <- adjustDeltaForOffsetM p debugM $ "printStringAtRsC:(p,p')=" ++ show (p,p') printStringAtLsDelta p' str - setPriorEndASTD True pa + setPriorEndASTD pa cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -709,6 +737,9 @@ printStringAtMLocL (EpAnn anc an cs) l s = do printStringAtLsDelta (SameLine 1) str return (Just (EpaDelta noSrcSpan (SameLine 1) [])) +printStringAdvanceA :: (Monad m, Monoid w) => String -> EP w m () +printStringAdvanceA str = printStringAtAA (EpaDelta noSrcSpan (SameLine 0) []) str >> return () + printStringAtAA :: (Monad m, Monoid w) => EpaLocation -> String -> EP w m EpaLocation printStringAtAA el str = printStringAtAAC CaptureComments el str @@ -735,7 +766,7 @@ printStringAtAAC capture (EpaDelta ss d cs) s = do p2 <- getPosP pe2 <- getPriorEndD debugM $ "printStringAtAA:(pe1,pe2,p1,p2)=" ++ show (pe1,pe2,p1,p2) - setPriorEndASTPD True (pe1,pe2) + setPriorEndASTPD (pe1,pe2) cs' <- case capture of CaptureComments -> takeAppliedComments NoCaptureComments -> return [] @@ -883,8 +914,7 @@ markAnnOpenP' :: (Monad m, Monoid w) => AnnPragma -> SourceText -> String -> EP markAnnOpenP' an NoSourceText txt = markEpAnnLMS0 an lapr_open AnnOpen (Just txt) markAnnOpenP' an (SourceText txt) _ = markEpAnnLMS0 an lapr_open AnnOpen (Just $ unpackFS txt) -markAnnOpen :: (Monad m, Monoid w) - => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] +markAnnOpen :: (Monad m, Monoid w) => [AddEpAnn] -> SourceText -> String -> EP w m [AddEpAnn] markAnnOpen an NoSourceText txt = markEpAnnLMS'' an lidl AnnOpen (Just txt) markAnnOpen an (SourceText txt) _ = markEpAnnLMS'' an lidl AnnOpen (Just $ unpackFS txt) @@ -1589,7 +1619,7 @@ markTopLevelList ls = mapM (\a -> setLayoutTopLevelP $ markAnnotated a) ls instance (ExactPrint a) => ExactPrint (Located a) where getAnnotationEntry (L l _) = case l of UnhelpfulSpan _ -> NoEntryVal - _ -> Entry (hackSrcSpanToAnchor l) [] emptyComments NoFlushComments CanUpdateAnchorOnly + _ -> Entry (EpaSpan l) [] emptyComments NoFlushComments CanUpdateAnchorOnly setAnnotationAnchor (L l a) _anc _ts _cs = L l a @@ -1664,16 +1694,10 @@ instance ExactPrint (HsModule GhcPs) where _ -> return lo am_decls' <- markTrailing (am_decls $ anns an0) - imports' <- markTopLevelList imports - - case lo of - EpExplicitBraces _ _ -> return () - _ -> do - -- Get rid of the balance of the preceding comments before starting on the decls - flushComments [] - putUnallocatedComments [] - decls' <- markTopLevelList (filter removeDocDecl decls) + mid <- markAnnotated (HsModuleImpDecls (am_cs $ anns an0) imports decls) + let imports' = id_imps mid + let decls' = id_decls mid lo1 <- case lo0 of EpExplicitBraces open close -> do @@ -1688,15 +1712,32 @@ instance ExactPrint (HsModule GhcPs) where debugM $ "am_eof:" ++ showGhc (pos, prior) setEofPos (Just (pos, prior)) - let anf = an0 { anns = (anns an0) { am_decls = am_decls' }} + let anf = an0 { anns = (anns an0) { am_decls = am_decls', am_cs = [] }} debugM $ "HsModule, anf=" ++ showAst anf return (HsModule (XModulePs anf lo1 mdeprec' mbDoc') mmn' mexports' imports' decls') +-- --------------------------------------------------------------------- + +-- | This is used to ensure the comments are updated into the right +-- place for makeDeltaAst. +data HsModuleImpDecls + = HsModuleImpDecls { + id_cs :: [LEpaComment], + id_imps :: [LImportDecl GhcPs], + id_decls :: [LHsDecl GhcPs] + } deriving Data + +instance ExactPrint HsModuleImpDecls where + -- Use an UnhelpfulSpan for the anchor, we are only interested in the comments + getAnnotationEntry mid = mkEntry (EpaSpan (UnhelpfulSpan UnhelpfulNoLocationInfo)) [] (EpaComments (id_cs mid)) + setAnnotationAnchor mid _anc _ cs = mid { id_cs = priorComments cs ++ getFollowingComments cs } + `debug` ("HsModuleImpDecls.setAnnotationAnchor:cs=" ++ showAst cs) + exact (HsModuleImpDecls cs imports decls) = do + imports' <- markTopLevelList imports + decls' <- markTopLevelList (filter notDocDecl decls) + return (HsModuleImpDecls cs imports' decls') -removeDocDecl :: LHsDecl GhcPs -> Bool -removeDocDecl (L _ DocD{}) = False -removeDocDecl _ = True -- --------------------------------------------------------------------- @@ -1737,8 +1778,8 @@ instance ExactPrint InWarningCategory where exact (InWarningCategory tkIn source (L l wc)) = do tkIn' <- markEpToken tkIn - L _ (_,wc') <- markAnnotated (L l (source, wc)) - return (InWarningCategory tkIn' source (L l wc')) + L l' (_,wc') <- markAnnotated (L l (source, wc)) + return (InWarningCategory tkIn' source (L l' wc')) instance ExactPrint (SourceText, WarningCategory) where getAnnotationEntry _ = NoEntryVal @@ -1943,14 +1984,14 @@ exactDataFamInstDecl an top_lvl , feqn_pats = pats , feqn_fixity = fixity , feqn_rhs = defn })) = do - (an', an2', tycon', bndrs', _, _mc, defn') <- exactDataDefn an2 pp_hdr defn - -- See Note [an and an2 in exactDataFamInstDecl] + (an', an2', tycon', bndrs', pats', defn') <- exactDataDefn an2 pp_hdr defn + -- See Note [an and an2 in exactDataFamInstDecl] return (an', DataFamInstDecl ( FamEqn { feqn_ext = an2' , feqn_tycon = tycon' , feqn_bndrs = bndrs' - , feqn_pats = pats + , feqn_pats = pats' , feqn_fixity = fixity , feqn_rhs = defn' })) `debug` ("exactDataFamInstDecl: defn' derivs:" ++ showAst (dd_derivs defn')) @@ -2233,11 +2274,11 @@ instance ExactPrint (RoleAnnotDecl GhcPs) where an1 <- markEpAnnL an0 lidl AnnRole ltycon' <- markAnnotated ltycon let markRole (L l (Just r)) = do - (L _ r') <- markAnnotated (L l r) - return (L l (Just r')) + (L l' r') <- markAnnotated (L l r) + return (L l' (Just r')) markRole (L l Nothing) = do - printStringAtSs (locA l) "_" - return (L l Nothing) + e' <- printStringAtAA (entry l) "_" + return (L (l { entry = e'}) Nothing) roles' <- mapM markRole roles return (RoleAnnotDecl an1 ltycon' roles') @@ -2340,8 +2381,13 @@ instance (ExactPrint tm, ExactPrint ty, Outputable tm, Outputable ty) getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact a@(HsValArg _ tm) = markAnnotated tm >> return a - exact a@(HsTypeArg at ty) = markEpToken at >> markAnnotated ty >> return a + exact (HsValArg x tm) = do + tm' <- markAnnotated tm + return (HsValArg x tm') + exact (HsTypeArg at ty) = do + at' <- markEpToken at + ty' <- markAnnotated ty + return (HsTypeArg at' ty') exact x@(HsArgPar _sp) = withPpr x -- Does not appear in original source -- --------------------------------------------------------------------- @@ -2359,9 +2405,9 @@ instance ExactPrint (ClsInstDecl GhcPs) where (mbWarn', an0, mbOverlap', inst_ty') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lid AnnSemi - ds <- withSortKey sortKey - [(ClsAtdTag, prepareListAnnotationA ats), - (ClsAtdTag, prepareListAnnotationF an adts), + (sortKey', ds) <- withSortKey sortKey + [(ClsAtTag, prepareListAnnotationA ats), + (ClsAtdTag, prepareListAnnotationF adts), (ClsMethodTag, prepareListAnnotationA binds), (ClsSigTag, prepareListAnnotationA sigs) ] @@ -2371,7 +2417,7 @@ instance ExactPrint (ClsInstDecl GhcPs) where adts' = undynamic ds binds' = undynamic ds sigs' = undynamic ds - return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey) + return (ClsInstDecl { cid_ext = (mbWarn', an3, sortKey') , cid_poly_ty = inst_ty', cid_binds = binds' , cid_sigs = sigs', cid_tyfam_insts = ats' , cid_overlap_mode = mbOverlap' @@ -2452,15 +2498,29 @@ instance ExactPrint (HsBind GhcPs) where return (FunBind x fun_id' matches') exact (PatBind x pat q grhss) = do + q' <- markAnnotated q pat' <- markAnnotated pat grhss' <- markAnnotated grhss - return (PatBind x pat' q grhss') + return (PatBind x pat' q' grhss') exact (PatSynBind x bind) = do bind' <- markAnnotated bind return (PatSynBind x bind') exact x = error $ "HsBind: exact for " ++ showAst x +instance ExactPrint (HsMultAnn GhcPs) where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact (HsNoMultAnn x) = return (HsNoMultAnn x) + exact (HsPct1Ann tok) = do + tok' <- markEpToken tok + return (HsPct1Ann tok') + exact (HsMultAnn tok ty) = do + tok' <- markEpToken tok + ty' <- markAnnotated ty + return (HsMultAnn tok' ty') + -- --------------------------------------------------------------------- instance ExactPrint (PatSynBind GhcPs GhcPs) where @@ -2519,8 +2579,9 @@ instance ExactPrint (PatSynBind GhcPs GhcPs) where instance ExactPrint (RecordPatSynField GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact r@(RecordPatSynField { recordPatSynField = v }) = markAnnotated v - >> return r + exact (RecordPatSynField f v) = do + f' <- markAnnotated f + return (RecordPatSynField f' v) -- --------------------------------------------------------------------- @@ -2648,15 +2709,20 @@ instance ExactPrint (HsLocalBinds GhcPs) where (an1, valbinds') <- markAnnList an0 $ markAnnotatedWithLayout valbinds debugM $ "exact HsValBinds: an1=" ++ showAst an1 - return (HsValBinds an1 valbinds') + medr <- getExtraDPReturn + an2 <- case medr of + Nothing -> return an1 + Just dp -> do + setExtraDPReturn Nothing + return $ an1 { anns = (anns an1) { al_anchor = Just (EpaDelta noSrcSpan dp []) }} + return (HsValBinds an2 valbinds') exact (HsIPBinds an bs) = do - (as, ipb) <- markAnnList an (markEpAnnL' an lal_rest AnnWhere - >> markAnnotated bs - >>= \bs' -> return (HsIPBinds an bs'::HsLocalBinds GhcPs)) - case ipb of - HsIPBinds _ bs' -> return (HsIPBinds as bs'::HsLocalBinds GhcPs) - _ -> error "should not happen HsIPBinds" + (an2,bs') <- markAnnListA an $ \an0 -> do + an1 <- markEpAnnL' an0 lal_rest AnnWhere + bs' <- markAnnotated bs + return (an1, bs') + return (HsIPBinds an2 bs') exact b@(EmptyLocalBinds _) = return b @@ -2670,7 +2736,8 @@ instance ExactPrint (HsValBindsLR GhcPs GhcPs) where let binds' = concatMap decl2Bind decls sigs' = concatMap decl2Sig decls - return (ValBinds sortKey binds' sigs') + sortKey' = captureOrderBinds decls + return (ValBinds sortKey' binds' sigs') exact (XValBindsLR _) = panic "XValBindsLR" undynamic :: Typeable a => [Dynamic] -> [a] @@ -2682,7 +2749,9 @@ instance ExactPrint (HsIPBinds GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact b@(IPBinds _ binds) = setLayoutBoth $ markAnnotated binds >> return b + exact (IPBinds x binds) = setLayoutBoth $ do + binds' <- markAnnotated binds + return (IPBinds x binds') -- --------------------------------------------------------------------- @@ -2703,18 +2772,18 @@ instance ExactPrint HsIPName where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact i@(HsIPName fs) = printStringAdvance ("?" ++ (unpackFS fs)) >> return i + exact i@(HsIPName fs) = printStringAdvanceA ("?" ++ (unpackFS fs)) >> return i -- --------------------------------------------------------------------- -- Managing lists which have been separated, e.g. Sigs and Binds prepareListAnnotationF :: (Monad m, Monoid w) => - [AddEpAnn] -> [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] -prepareListAnnotationF an ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls + [LDataFamInstDecl GhcPs] -> [(RealSrcSpan,EP w m Dynamic)] +prepareListAnnotationF ls = map (\b -> (realSrcSpan $ getLocA b, go b)) ls where go (L l a) = do - d' <- markAnnotated (DataFamInstDeclWithContext an NotTopLevel a) - return (toDyn (L l (dc_d d'))) + (L l' d') <- markAnnotated (L l (DataFamInstDeclWithContext noAnn NotTopLevel a)) + return (toDyn (L l' (dc_d d'))) prepareListAnnotationA :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => [LocatedAn an a] -> [(RealSrcSpan,EP w m Dynamic)] @@ -2725,15 +2794,23 @@ prepareListAnnotationA ls = map (\b -> (realSrcSpan $ getLocA b,go b)) ls return (toDyn b') withSortKey :: (Monad m, Monoid w) - => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] -> EP w m [Dynamic] + => AnnSortKey DeclTag -> [(DeclTag, [(RealSrcSpan, EP w m Dynamic)])] + -> EP w m (AnnSortKey DeclTag, [Dynamic]) withSortKey annSortKey xs = do debugM $ "withSortKey:annSortKey=" ++ showAst annSortKey - let ordered = case annSortKey of - NoAnnSortKey -> sortBy orderByFst $ concatMap snd xs - AnnSortKey _keys -> orderedDecls annSortKey (Map.fromList xs) - mapM snd ordered -orderByFst :: Ord a => (a, b1) -> (a, b2) -> Ordering -orderByFst (a,_) (b,_) = compare a b + let (sk, ordered) = case annSortKey of + NoAnnSortKey -> (annSortKey', map snd os) + where + doOne (tag, ds) = map (\d -> (tag, d)) ds + xsExpanded = concatMap doOne xs + os = sortBy orderByFst $ xsExpanded + annSortKey' = AnnSortKey (map fst os) + AnnSortKey _keys -> (annSortKey, orderedDecls annSortKey (Map.fromList xs)) + ordered' <- mapM snd ordered + return (sk, ordered') + +orderByFst :: Ord a => (t, (a,b1)) -> (t, (a, b2)) -> Ordering +orderByFst (_,(a,_)) (_,(b,_)) = compare a b -- --------------------------------------------------------------------- @@ -2761,15 +2838,16 @@ instance ExactPrint (Sig GhcPs) where (an0, vars',ty') <- exactVarSig an vars ty return (ClassOpSig an0 is_deflt vars' ty') - exact (FixSig (an,src) (FixitySig x names (Fixity v fdir))) = do + exact (FixSig (an,src) (FixitySig ns names (Fixity v fdir))) = do let fixstr = case fdir of InfixL -> "infixl" InfixR -> "infixr" InfixN -> "infix" an0 <- markEpAnnLMS'' an lidl AnnInfix (Just fixstr) an1 <- markEpAnnLMS'' an0 lidl AnnVal (Just (sourceTextToString src (show v))) + ns' <- markAnnotated ns names' <- markAnnotated names - return (FixSig (an1,src) (FixitySig x names' (Fixity v fdir))) + return (FixSig (an1,src) (FixitySig ns' names' (Fixity v fdir))) exact (InlineSig an ln inl) = do an0 <- markAnnOpen an (inl_src inl) "{-# INLINE" @@ -2809,7 +2887,7 @@ instance ExactPrint (Sig GhcPs) where exact (CompleteMatchSig (an,src) cs mty) = do an0 <- markAnnOpen an src "{-# COMPLETE" - cs' <- markAnnotated cs + cs' <- mapM markAnnotated cs (an1, mty') <- case mty of Nothing -> return (an0, mty) @@ -2822,6 +2900,20 @@ instance ExactPrint (Sig GhcPs) where -- --------------------------------------------------------------------- +instance ExactPrint NamespaceSpecifier where + getAnnotationEntry _ = NoEntryVal + setAnnotationAnchor a _ _ _ = a + + exact NoNamespaceSpecifier = return NoNamespaceSpecifier + exact (TypeNamespaceSpecifier typeTok) = do + typeTok' <- markEpToken typeTok + return (TypeNamespaceSpecifier typeTok') + exact (DataNamespaceSpecifier dataTok) = do + dataTok' <- markEpToken dataTok + return (DataNamespaceSpecifier dataTok') + +-- --------------------------------------------------------------------- + exactVarSig :: (Monad m, Monoid w, ExactPrint a) => AnnSig -> [LocatedN RdrName] -> a -> EP w m (AnnSig, [LocatedN RdrName], a) exactVarSig an vars ty = do @@ -2875,7 +2967,7 @@ instance ExactPrint (AnnDecl GhcPs) where n' <- markAnnotated n return (an1, TypeAnnProvenance n') ModuleAnnProvenance -> do - an1 <- markEpAnnL an lapr_rest AnnModule + an1 <- markEpAnnL an0 lapr_rest AnnModule return (an1, prov) e' <- markAnnotated e @@ -2950,21 +3042,21 @@ instance ExactPrint (HsExpr GhcPs) where then markAnnotated n else return n return (HsVar x n') - exact x@(HsUnboundVar an _) = do + exact (HsUnboundVar an n) = do case an of Just (EpAnnUnboundVar (ob,cb) l) -> do - printStringAtAA ob "`" >> return () - printStringAtAA l "_" >> return () - printStringAtAA cb "`" >> return () - return x + ob' <- printStringAtAA ob "`" + l' <- printStringAtAA l "_" + cb' <- printStringAtAA cb "`" + return (HsUnboundVar (Just (EpAnnUnboundVar (ob',cb') l')) n) _ -> do - printStringAtLsDelta (SameLine 0) "_" - return x + printStringAdvanceA "_" >> return () + return (HsUnboundVar an n) exact x@(HsOverLabel src l) = do - printStringAtLsDelta (SameLine 0) "#" + printStringAdvanceA "#" >> return () case src of - NoSourceText -> printStringAtLsDelta (SameLine 0) (unpackFS l) - SourceText txt -> printStringAtLsDelta (SameLine 0) (unpackFS txt) + NoSourceText -> printStringAdvanceA (unpackFS l) >> return () + SourceText txt -> printStringAdvanceA (unpackFS txt) >> return () return x exact x@(HsIPVar _ (HsIPName n)) @@ -3204,11 +3296,11 @@ instance ExactPrint (HsExpr GhcPs) where exact (HsTypedSplice an s) = do an0 <- markEpAnnL an lidl AnnDollarDollar - s' <- exact s + s' <- markAnnotated s return (HsTypedSplice an0 s') exact (HsUntypedSplice an s) = do - s' <- exact s + s' <- markAnnotated s return (HsUntypedSplice an s') exact (HsProc an p c) = do @@ -3274,12 +3366,15 @@ exactMdo an (Just module_name) kw = markEpAnnLMS'' an lal_rest kw (Just n) markMaybeDodgyStmts :: (Monad m, Monoid w, ExactPrint (LocatedAn an a)) => AnnList -> LocatedAn an a -> EP w m (AnnList, LocatedAn an a) markMaybeDodgyStmts an stmts = - if isGoodSrcSpan (getLocA stmts) + if notDodgy stmts then do r <- markAnnotatedWithLayout stmts return (an, r) else return (an, stmts) +notDodgy :: GenLocated (EpAnn ann) a -> Bool +notDodgy (L (EpAnn anc _ _) _) = notDodgyE anc + notDodgyE :: EpaLocation -> Bool notDodgyE anc = case anc of @@ -3341,7 +3436,7 @@ instance ExactPrint (MatchGroup GhcPs (LocatedA (HsCmd GhcPs))) where setAnnotationAnchor a _ _ _ = a exact (MG x matches) = do -- TODO:AZ use SortKey, in MG ann. - matches' <- if isGoodSrcSpan (getLocA matches) + matches' <- if notDodgy matches then markAnnotated matches else return matches return (MG x matches') @@ -3661,6 +3756,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- There may be arbitrary parens around parts of the constructor -- that are infix. Turn these into comments so that they feed -- into the right place automatically + -- TODO: no longer sorting on insert. What now? an0 <- annotationsToComments an lidl [AnnOpenP,AnnCloseP] an1 <- markEpAnnL an0 lidl AnnType @@ -3674,7 +3770,7 @@ instance ExactPrint (TyClDecl GhcPs) where -- TODO: add a workaround for https://gitlab.haskell.org/ghc/ghc/-/issues/20452 exact (DataDecl { tcdDExt = an, tcdLName = ltycon, tcdTyVars = tyvars , tcdFixity = fixity, tcdDataDefn = defn }) = do - (_, an', ltycon', tyvars', _, _mctxt', defn') <- + (_, an', ltycon', tyvars', _, defn') <- exactDataDefn an (exactVanillaDeclHead ltycon tyvars fixity) defn return (DataDecl { tcdDExt = an', tcdLName = ltycon', tcdTyVars = tyvars' , tcdFixity = fixity, tcdDataDefn = defn' }) @@ -3707,7 +3803,7 @@ instance ExactPrint (TyClDecl GhcPs) where (an0, fds', lclas', tyvars',context') <- top_matter an1 <- markEpAnnL an0 lidl AnnOpenC an2 <- markEpAnnAllL' an1 lidl AnnSemi - ds <- withSortKey sortKey + (sortKey', ds) <- withSortKey sortKey [(ClsSigTag, prepareListAnnotationA sigs), (ClsMethodTag, prepareListAnnotationA methods), (ClsAtTag, prepareListAnnotationA ats), @@ -3720,7 +3816,7 @@ instance ExactPrint (TyClDecl GhcPs) where methods' = undynamic ds ats' = undynamic ds at_defs' = undynamic ds - return (ClassDecl {tcdCExt = (an3, lo, sortKey), + return (ClassDecl {tcdCExt = (an3, lo, sortKey'), tcdCtxt = context', tcdLName = lclas', tcdTyVars = tyvars', tcdFixity = fixity, tcdFDs = fds', @@ -3845,7 +3941,7 @@ exactDataDefn -> HsDataDefn GhcPs -> EP w m ( [AddEpAnn] -- ^ from exactHdr , [AddEpAnn] -- ^ updated one passed in - , LocatedN RdrName, a, b, Maybe (LHsContext GhcPs), HsDataDefn GhcPs) + , LocatedN RdrName, a, b, HsDataDefn GhcPs) exactDataDefn an exactHdr (HsDataDefn { dd_ext = x, dd_ctxt = context , dd_cType = mb_ct @@ -3883,8 +3979,8 @@ exactDataDefn an exactHdr _ -> panic "exacprint NewTypeCon" an6 <- markEpAnnL an5 lidl AnnCloseC derivings' <- mapM markAnnotated derivings - return (anx, an6, ln', tvs', b, mctxt', - (HsDataDefn { dd_ext = x, dd_ctxt = context + return (anx, an6, ln', tvs', b, + (HsDataDefn { dd_ext = x, dd_ctxt = mctxt' , dd_cType = mb_ct' , dd_kindSig = mb_sig' , dd_cons = condecls'', dd_derivs = derivings' })) @@ -3941,22 +4037,23 @@ instance ExactPrint (InjectivityAnn GhcPs) where class Typeable flag => ExactPrintTVFlag flag where exactTVDelimiters :: (Monad m, Monoid w) - => [AddEpAnn] -> flag -> EP w m (HsTyVarBndr flag GhcPs) - -> EP w m ([AddEpAnn], (HsTyVarBndr flag GhcPs)) + => [AddEpAnn] -> flag + -> ([AddEpAnn] -> EP w m ([AddEpAnn], HsTyVarBndr flag GhcPs)) + -> EP w m ([AddEpAnn], flag, (HsTyVarBndr flag GhcPs)) instance ExactPrintTVFlag () where - exactTVDelimiters an _ thing_inside = do + exactTVDelimiters an flag thing_inside = do an0 <- markEpAnnAllL' an lid AnnOpenP - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid AnnCloseP - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid AnnCloseP + return (an2, flag, r) instance ExactPrintTVFlag Specificity where exactTVDelimiters an s thing_inside = do an0 <- markEpAnnAllL' an lid open - r <- thing_inside - an1 <- markEpAnnAllL' an0 lid close - return (an1, r) + (an1, r) <- thing_inside an0 + an2 <- markEpAnnAllL' an1 lid close + return (an2, s, r) where (open, close) = case s of SpecifiedSpec -> (AnnOpenP, AnnCloseP) @@ -3964,33 +4061,33 @@ instance ExactPrintTVFlag Specificity where instance ExactPrintTVFlag (HsBndrVis GhcPs) where exactTVDelimiters an0 bvis thing_inside = do - case bvis of - HsBndrRequired _ -> return () - HsBndrInvisible at -> markEpToken at >> return () + bvis' <- case bvis of + HsBndrRequired _ -> return bvis + HsBndrInvisible at -> HsBndrInvisible <$> markEpToken at an1 <- markEpAnnAllL' an0 lid AnnOpenP - r <- thing_inside - an2 <- markEpAnnAllL' an1 lid AnnCloseP - return (an2, r) + (an2, r) <- thing_inside an1 + an3 <- markEpAnnAllL' an2 lid AnnCloseP + return (an3, bvis', r) instance ExactPrintTVFlag flag => ExactPrint (HsTyVarBndr flag GhcPs) where getAnnotationEntry _ = NoEntryVal setAnnotationAnchor a _ _ _ = a exact (UserTyVar an flag n) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - return (UserTyVar an flag n') + return (ani, UserTyVar an flag n') case r of - (an', UserTyVar _ flag'' n'') -> return (UserTyVar an' flag'' n'') + (an', flag', UserTyVar _ _ n'') -> return (UserTyVar an' flag' n'') _ -> error "KindedTyVar should never happen here" exact (KindedTyVar an flag n k) = do - r <- exactTVDelimiters an flag $ do + r <- exactTVDelimiters an flag $ \ani -> do n' <- markAnnotated n - an0 <- markEpAnnL an lidl AnnDcolon + an0 <- markEpAnnL ani lidl AnnDcolon k' <- markAnnotated k - return (KindedTyVar an0 flag n' k') + return (an0, KindedTyVar an0 flag n' k') case r of - (an',KindedTyVar _ flag'' n'' k'') -> return (KindedTyVar an' flag'' n'' k'') + (an',flag', KindedTyVar _ _ n'' k'') -> return (KindedTyVar an' flag' n'' k'') _ -> error "UserTyVar should never happen here" -- --------------------------------------------------------------------- @@ -4150,17 +4247,16 @@ instance ExactPrint (HsDerivingClause GhcPs) where , deriv_clause_strategy = dcs , deriv_clause_tys = dct }) = do an0 <- markEpAnnL an lidl AnnDeriving - exact_strat_before + dcs0 <- case dcs of + Just (L _ ViaStrategy{}) -> return dcs + _ -> mapM markAnnotated dcs dct' <- markAnnotated dct - exact_strat_after + dcs1 <- case dcs0 of + Just (L _ ViaStrategy{}) -> mapM markAnnotated dcs0 + _ -> return dcs0 return (HsDerivingClause { deriv_clause_ext = an0 - , deriv_clause_strategy = dcs + , deriv_clause_strategy = dcs1 , deriv_clause_tys = dct' }) - where - (exact_strat_before, exact_strat_after) = - case dcs of - Just v@(L _ ViaStrategy{}) -> (pure (), markAnnotated v >> pure ()) - _ -> (mapM_ markAnnotated dcs, pure ()) -- --------------------------------------------------------------------- @@ -4467,7 +4563,9 @@ instance ExactPrint (ConDeclField GhcPs) where instance ExactPrint (FieldOcc GhcPs) where getAnnotationEntry = const NoEntryVal setAnnotationAnchor a _ _ _ = a - exact f@(FieldOcc _ n) = markAnnotated n >> return f + exact (FieldOcc x n) = do + n' <- markAnnotated n + return (FieldOcc x n') -- --------------------------------------------------------------------- @@ -4535,7 +4633,7 @@ instance ExactPrint (LocatedL [LocatedA (IE GhcPs)]) where an0 <- markEpAnnL' an lal_rest AnnHiding p <- getPosP debugM $ "LocatedL [LIE:p=" ++ showPprUnsafe p - (an1, ies') <- markAnnList an0 (markAnnotated ies) + (an1, ies') <- markAnnList an0 (markAnnotated (filter notIEDoc ies)) return (L an1 ies') instance (ExactPrint (Match GhcPs (LocatedA body))) @@ -4985,6 +5083,14 @@ setExtraDP md = do debugM $ "setExtraDP:" ++ show md modify (\s -> s {uExtraDP = md}) +getExtraDPReturn :: (Monad m, Monoid w) => EP w m (Maybe DeltaPos) +getExtraDPReturn = gets uExtraDPReturn + +setExtraDPReturn :: (Monad m, Monoid w) => Maybe DeltaPos -> EP w m () +setExtraDPReturn md = do + debugM $ "setExtraDPReturn:" ++ show md + modify (\s -> s {uExtraDPReturn = md}) + getPriorEndD :: (Monad m, Monoid w) => EP w m Pos getPriorEndD = gets dPriorEndPosition @@ -5007,13 +5113,13 @@ setPriorEndNoLayoutD pe = do debugM $ "setPriorEndNoLayoutD:pe=" ++ show pe modify (\s -> s { dPriorEndPosition = pe }) -setPriorEndASTD :: (Monad m, Monoid w) => Bool -> RealSrcSpan -> EP w m () -setPriorEndASTD layout pe = setPriorEndASTPD layout (rs2range pe) +setPriorEndASTD :: (Monad m, Monoid w) => RealSrcSpan -> EP w m () +setPriorEndASTD pe = setPriorEndASTPD (rs2range pe) -setPriorEndASTPD :: (Monad m, Monoid w) => Bool -> (Pos,Pos) -> EP w m () -setPriorEndASTPD layout pe@(fm,to) = do +setPriorEndASTPD :: (Monad m, Monoid w) => (Pos,Pos) -> EP w m () +setPriorEndASTPD pe@(fm,to) = do debugM $ "setPriorEndASTD:pe=" ++ show pe - when layout $ setLayoutStartD (snd fm) + setLayoutStartD (snd fm) modify (\s -> s { dPriorEndPosition = to } ) setLayoutStartD :: (Monad m, Monoid w) => Int -> EP w m () @@ -5044,7 +5150,7 @@ getUnallocatedComments :: (Monad m, Monoid w) => EP w m [Comment] getUnallocatedComments = gets epComments putUnallocatedComments :: (Monad m, Monoid w) => [Comment] -> EP w m () -putUnallocatedComments cs = modify (\s -> s { epComments = cs } ) +putUnallocatedComments !cs = modify (\s -> s { epComments = cs } ) -- | Push a fresh stack frame for the applied comments gatherer pushAppliedComments :: (Monad m, Monoid w) => EP w m () @@ -5054,7 +5160,7 @@ pushAppliedComments = modify (\s -> s { epCommentsApplied = []:(epCommentsApplie -- takeAppliedComments, and clear them, not popping the stack takeAppliedComments :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedComments = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5067,7 +5173,7 @@ takeAppliedComments = do -- takeAppliedComments, and clear them, popping the stack takeAppliedCommentsPop :: (Monad m, Monoid w) => EP w m [Comment] takeAppliedCommentsPop = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> do modify (\s -> s { epCommentsApplied = [] }) @@ -5080,7 +5186,7 @@ takeAppliedCommentsPop = do -- when doing delta processing applyComment :: (Monad m, Monoid w) => Comment -> EP w m () applyComment c = do - ccs <- gets epCommentsApplied + !ccs <- gets epCommentsApplied case ccs of [] -> modify (\s -> s { epCommentsApplied = [[c]] } ) (h:t) -> modify (\s -> s { epCommentsApplied = (c:h):t } ) ===================================== utils/check-exact/Main.hs ===================================== @@ -470,7 +470,7 @@ changeAddDecl1 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtStart m decl' + replaceTopLevelDecls m = return $ insertAtStart m decl' return p' -- --------------------------------------------------------------------- @@ -483,7 +483,7 @@ changeAddDecl2 libdir top = do let (p',_,_) = runTransform doAddDecl doAddDecl = everywhereM (mkM replaceTopLevelDecls) top replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAtEnd m decl' + replaceTopLevelDecls m = return $ insertAtEnd m decl' return p' -- --------------------------------------------------------------------- @@ -500,7 +500,7 @@ changeAddDecl3 libdir top = do l2' = setEntryDP l2 (DifferentLine 2 0) replaceTopLevelDecls :: ParsedSource -> Transform ParsedSource - replaceTopLevelDecls m = insertAt f m decl' + replaceTopLevelDecls m = return $ insertAt f m decl' return p' -- --------------------------------------------------------------------- @@ -571,8 +571,9 @@ changeLocalDecls2 libdir (L l p) = do changeWhereIn3a :: Changer changeWhereIn3a _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) - debugM $ unlines w + decls = balanceCommentsList decls0 + (_de0:_:de1:_d2:_) = decls + debugM $ "changeWhereIn3a:de1:" ++ showAst de1 let p2 = p { hsmodDecls = decls} return (L l p2) @@ -581,13 +582,12 @@ changeWhereIn3a _libdir (L l p) = do changeWhereIn3b :: Changer changeWhereIn3b _libdir (L l p) = do let decls0 = hsmodDecls p - (decls,_,w) = runTransform (balanceCommentsList decls0) + decls = balanceCommentsList decls0 (de0:tdecls@(_:de1:d2:_)) = decls de0' = setEntryDP de0 (DifferentLine 2 0) de1' = setEntryDP de1 (DifferentLine 2 0) d2' = setEntryDP d2 (DifferentLine 2 0) decls' = d2':de1':de0':tdecls - debugM $ unlines w debugM $ "changeWhereIn3b:de1':" ++ showAst de1' let p2 = p { hsmodDecls = decls'} return (L l p2) @@ -598,37 +598,37 @@ addLocaLDecl1 :: Changer addLocaLDecl1 libdir top = do Right (L ld (ValD _ decl)) <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let decl' = setEntryDP (L ld decl) (DifferentLine 1 5) - doAddLocal = do - let lp = top - (de1:d2:d3:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - (de1',_) <- modifyValD (getLocA de1'') de1'' $ \_m d -> do - return ((wrapDecl decl' : d),Nothing) - replaceDecls lp [de1', d2', d3] - - (lp',_,w) <- runTransformT doAddLocal - debugM $ "addLocaLDecl1:" ++ intercalate "\n" w + doAddLocal :: ParsedSource + doAddLocal = replaceDecls lp [de1', d2', d3] + where + lp = top + (de1:d2:d3:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 + (de1',_) = modifyValD (getLocA de1'') de1'' $ \_m d -> ((wrapDecl decl' : d),Nothing) + + let lp' = doAddLocal return lp' -- --------------------------------------------------------------------- + addLocaLDecl2 :: Changer addLocaLDecl2 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 + doAddLocal = replaceDecls lp [parent',d2'] + where + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - newDecl' <- transferEntryDP' d newDecl - let d' = setEntryDP d (DifferentLine 1 0) - return ((newDecl':d':ds),Nothing) + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = transferEntryDP' d (makeDeltaAst newDecl) + d' = setEntryDP d (DifferentLine 1 0) + in ((newDecl':d':ds),Nothing) - replaceDecls lp [parent',d2'] - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -637,19 +637,18 @@ addLocaLDecl3 :: Changer addLocaLDecl3 libdir top = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") let - doAddLocal = do - let lp = top - (de1:d2:_) <- hsDecls lp - (de1'',d2') <- balanceComments de1 d2 - - (parent',_) <- modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> do - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - return (((d:ds) ++ [newDecl']),Nothing) + doAddLocal = replaceDecls (anchorEof lp) [parent',d2'] + where + lp = top + (de1:d2:_) = hsDecls lp + (de1'',d2') = balanceComments de1 d2 - replaceDecls (anchorEof lp) [parent',d2'] + (parent',_) = modifyValD (getLocA de1) de1'' $ \_m (d:ds) -> + let + newDecl' = setEntryDP newDecl (DifferentLine 1 0) + in (((d:ds) ++ [newDecl']),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -659,40 +658,38 @@ addLocaLDecl4 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "nn = 2") Right newSig <- withDynFlags libdir (\df -> parseDecl df "sig" "nn :: Int") let - doAddLocal = do - (parent:ds) <- hsDecls lp + doAddLocal = replaceDecls (anchorEof lp) (parent':ds) + where + (parent:ds) = hsDecls (makeDeltaAst lp) - let newDecl' = setEntryDP newDecl (DifferentLine 1 0) - let newSig' = setEntryDP newSig (DifferentLine 1 4) + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 0) + newSig' = setEntryDP (makeDeltaAst newSig) (DifferentLine 1 5) - (parent',_) <- modifyValD (getLocA parent) parent $ \_m decls -> do - return ((decls++[newSig',newDecl']),Nothing) + (parent',_) = modifyValD (getLocA parent) parent $ \_m decls -> + ((decls++[newSig',newDecl']),Nothing) - replaceDecls (anchorEof lp) (parent':ds) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' - -- --------------------------------------------------------------------- addLocaLDecl5 :: Changer addLocaLDecl5 _libdir lp = do let - doAddLocal = do - decls <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList decls + doAddLocal = replaceDecls lp [s1,de1',d3'] + where + decls = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList decls - let d3' = setEntryDP d3 (DifferentLine 2 0) + d3' = setEntryDP d3 (DifferentLine 2 0) - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m _decls -> do - let d2' = setEntryDP d2 (DifferentLine 1 0) - return ([d2'],Nothing) - replaceDecls lp [s1,de1',d3'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m _decls -> + let + d2' = setEntryDP d2 (DifferentLine 1 0) + in ([d2'],Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- @@ -701,39 +698,36 @@ addLocaLDecl6 :: Changer addLocaLDecl6 libdir lp = do Right newDecl <- withDynFlags libdir (\df -> parseDecl df "decl" "x = 3") let - newDecl' = setEntryDP newDecl (DifferentLine 1 4) - doAddLocal = do - decls0 <- hsDecls lp - [de1'',d2] <- balanceCommentsList decls0 + newDecl' = setEntryDP (makeDeltaAst newDecl) (DifferentLine 1 5) + doAddLocal = replaceDecls lp [de1', d2] + where + decls0 = hsDecls lp + [de1'',d2] = balanceCommentsList decls0 - let de1 = captureMatchLineSpacing de1'' - let L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 - let [ma1,_ma2] = ms + de1 = captureMatchLineSpacing de1'' + L _ (ValD _ (FunBind _ _ (MG _ (L _ ms)))) = de1 + [ma1,_ma2] = ms - (de1',_) <- modifyValD (getLocA ma1) de1 $ \_m decls -> do - return ((newDecl' : decls),Nothing) - replaceDecls lp [de1', d2] + (de1',_) = modifyValD (getLocA ma1) de1 $ \_m decls -> + ((newDecl' : decls),Nothing) - (lp',_,_w) <- runTransformT doAddLocal - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doAddLocal return lp' -- --------------------------------------------------------------------- rmDecl1 :: Changer -rmDecl1 _libdir top = do - let doRmDecl = do - let lp = top - tlDecs0 <- hsDecls lp - tlDecs' <- balanceCommentsList tlDecs0 - let tlDecs = captureLineSpacing tlDecs' - let (de1:_s1:_d2:d3:ds) = tlDecs - let d3' = setEntryDP d3 (DifferentLine 2 0) - - replaceDecls lp (de1:d3':ds) - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" +rmDecl1 _libdir lp = do + let + doRmDecl = replaceDecls lp (de1:d3':ds) + where + tlDecs0 = hsDecls lp + tlDecs = balanceCommentsList tlDecs0 + (de1:_s1:_d2:d3:ds) = tlDecs + d3' = setEntryDP d3 (DifferentLine 2 0) + + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -745,13 +739,13 @@ rmDecl2 _libdir lp = do let go :: GHC.LHsExpr GhcPs -> Transform (GHC.LHsExpr GhcPs) go e@(GHC.L _ (GHC.HsLet{})) = do - decs0 <- hsDecls e - decs <- balanceCommentsList $ captureLineSpacing decs0 - e' <- replaceDecls e (init decs) + let decs0 = hsDecls e + let decs = balanceCommentsList $ captureLineSpacing decs0 + let e' = replaceDecls e (init decs) return e' go x = return x - everywhereM (mkM go) lp + everywhereM (mkM go) (makeDeltaAst lp) let (lp',_,_w) = runTransform doRmDecl debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" @@ -762,17 +756,15 @@ rmDecl2 _libdir lp = do rmDecl3 :: Changer rmDecl3 _libdir lp = do let - doRmDecl = do - [de1,d2] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1] -> do - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([],Just sd1') - - replaceDecls lp [de1',sd1,d2] + doRmDecl = replaceDecls lp [de1',sd1,d2] + where + [de1,d2] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a] -> + let + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([],Just sd1') - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -780,19 +772,15 @@ rmDecl3 _libdir lp = do rmDecl4 :: Changer rmDecl4 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',Just sd1) <- modifyValD (getLocA de1) de1 $ \_m [sd1,sd2] -> do - sd2' <- transferEntryDP' sd1 sd2 - - let sd1' = setEntryDP sd1 (DifferentLine 2 0) - return ([sd2'],Just sd1') - - replaceDecls (anchorEof lp) [de1',sd1] - - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + doRmDecl = replaceDecls (anchorEof lp) [de1',sd1] + where + [de1] = hsDecls lp + (de1',Just sd1) = modifyValD (getLocA de1) de1 $ \_m [sd1a,sd2] -> + let + sd2' = transferEntryDP' sd1a sd2 + sd1' = setEntryDP sd1a (DifferentLine 2 0) + in ([sd2'],Just sd1') + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -805,10 +793,8 @@ rmDecl5 _libdir lp = do go :: HsExpr GhcPs -> Transform (HsExpr GhcPs) go (HsLet (tkLet, tkIn) lb expr) = do let decs = hsDeclsLocalBinds lb - let hdecs : _ = decs let dec = last decs - _ <- transferEntryDP hdecs dec - lb' <- replaceDeclsValbinds WithoutWhere lb [dec] + let lb' = replaceDeclsValbinds WithoutWhere lb [dec] return (HsLet (tkLet, tkIn) lb' expr) go x = return x @@ -823,73 +809,61 @@ rmDecl5 _libdir lp = do rmDecl6 :: Changer rmDecl6 _libdir lp = do let - doRmDecl = do - [de1] <- hsDecls lp - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m subDecs -> do - let subDecs' = captureLineSpacing subDecs - let (ss1:_sd1:sd2:sds) = subDecs' - sd2' <- transferEntryDP' ss1 sd2 - - return (sd2':sds,Nothing) + doRmDecl = replaceDecls lp [de1'] + where + [de1] = hsDecls lp - replaceDecls lp [de1'] + (de1',_) = modifyValD (getLocA de1) de1 $ \_m subDecs -> + let + subDecs' = captureLineSpacing subDecs + (ss1:_sd1:sd2:sds) = subDecs' + sd2' = transferEntryDP' ss1 sd2 + in (sd2':sds,Nothing) - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmDecl7 :: Changer -rmDecl7 _libdir top = do +rmDecl7 _libdir lp = do let - doRmDecl = do - let lp = top - tlDecs <- hsDecls lp - [s1,de1,d2,d3] <- balanceCommentsList tlDecs - - d3' <- transferEntryDP' d2 d3 - - replaceDecls lp [s1,de1,d3'] + doRmDecl = replaceDecls lp [s1,de1,d3'] + where + tlDecs = hsDecls lp + [s1,de1,d2,d3] = balanceCommentsList tlDecs + d3' = transferEntryDP' d2 d3 - (lp',_,_w) <- runTransformT doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig1 :: Changer rmTypeSig1 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let (s0:de1:d2) = tlDecs - s1 = captureTypeSigSpacing s0 - (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 - L ln n2' <- transferEntryDP n1 n2 - let s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) - replaceDecls lp (s1':de1:d2) - - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let doRmDecl = replaceDecls lp (s1':de1:d2) + where + tlDecs = hsDecls lp + (s0:de1:d2) = tlDecs + s1 = captureTypeSigSpacing s0 + (L l (SigD x1 (TypeSig x2 [n1,n2] typ))) = s1 + L ln n2' = transferEntryDP n1 n2 + s1' = (L l (SigD x1 (TypeSig x2 [L (noTrailingN ln) n2'] typ))) + + lp' = doRmDecl return lp' -- --------------------------------------------------------------------- rmTypeSig2 :: Changer rmTypeSig2 _libdir lp = do - let doRmDecl = do - tlDecs <- hsDecls lp - let [de1] = tlDecs - - (de1',_) <- modifyValD (getLocA de1) de1 $ \_m [s,d] -> do - d' <- transferEntryDP' s d - return $ ([d'],Nothing) - `debug` ("rmTypeSig2:(d,d')" ++ showAst (d,d')) - replaceDecls lp [de1'] + let doRmDecl = replaceDecls lp [de1'] + where + tlDecs = hsDecls lp + [de1] = tlDecs + (de1',_) = modifyValD (getLocA de1) de1 $ \_m [_s,d] -> ([d],Nothing) - let (lp',_,_w) = runTransform doRmDecl - debugM $ "log:[\n" ++ intercalate "\n" _w ++ "]log end\n" + let lp' = doRmDecl return lp' -- --------------------------------------------------------------------- @@ -958,13 +932,15 @@ addClassMethod libdir lp = do let decl' = setEntryDP decl (DifferentLine 1 3) let sig' = setEntryDP sig (DifferentLine 2 3) let doAddMethod = do - [cd] <- hsDecls lp - (f1:f2s:f2d:_) <- hsDecls cd - let f2s' = setEntryDP f2s (DifferentLine 2 3) - cd' <- replaceDecls cd [f1, sig', decl', f2s', f2d] - replaceDecls lp [cd'] - - (lp',_,w) <- runTransformT doAddMethod + let + [cd] = hsDecls lp + (f1:f2s:f2d:_) = hsDecls cd + f2s' = setEntryDP f2s (DifferentLine 2 3) + cd' = replaceDecls cd [f1, sig', decl', f2s', f2d] + lp' = replaceDecls lp [cd'] + return lp' + + let (lp',_,w) = runTransform doAddMethod debugM $ "addClassMethod:" ++ intercalate "\n" w return lp' ===================================== utils/check-exact/Parsers.hs ===================================== @@ -260,7 +260,7 @@ parseModuleEpAnnsWithCppInternal cppOptions dflags file = do GHC.PFailed pst -> Left (GHC.GhcPsMessage <$> GHC.getPsErrorMessages pst) GHC.POk _ pmod - -> Right $ (injectedComments, dflags', fixModuleTrailingComments pmod) + -> Right $ (injectedComments, dflags', fixModuleComments pmod) -- | Internal function. Exposed if you want to muck with DynFlags -- before parsing. Or after parsing. @@ -269,8 +269,10 @@ postParseTransform -> Either a (GHC.ParsedSource) postParseTransform parseRes = fmap mkAnns parseRes where - -- TODO:AZ perhaps inject the comments into the parsedsource here already - mkAnns (_cs, _, m) = fixModuleTrailingComments m + mkAnns (_cs, _, m) = fixModuleComments m + +fixModuleComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleComments p = fixModuleHeaderComments $ fixModuleTrailingComments p fixModuleTrailingComments :: GHC.ParsedSource -> GHC.ParsedSource fixModuleTrailingComments (GHC.L l p) = GHC.L l p' @@ -293,6 +295,47 @@ fixModuleTrailingComments (GHC.L l p) = GHC.L l p' in cs'' _ -> cs +-- Deal with https://gitlab.haskell.org/ghc/ghc/-/issues/23984 +-- The Lexer works bottom-up, so does not have module declaration info +-- when the first top decl processed +fixModuleHeaderComments :: GHC.ParsedSource -> GHC.ParsedSource +fixModuleHeaderComments (GHC.L l p) = GHC.L l p' + where + moveComments :: GHC.EpaLocation -> GHC.LHsDecl GHC.GhcPs -> GHC.EpAnnComments + -> (GHC.LHsDecl GHC.GhcPs, GHC.EpAnnComments) + moveComments GHC.EpaDelta{} dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.UnhelpfulSpan _)) dd cs = (dd,cs) + moveComments (GHC.EpaSpan (GHC.RealSrcSpan r _)) (GHC.L (GHC.EpAnn anc an csd) a) cs = (dd,css) + where + -- Move any comments on the decl that occur prior to the location + pc = GHC.priorComments csd + fc = GHC.getFollowingComments csd + bf (GHC.L anch _) = GHC.anchor anch > r + (move,keep) = break bf pc + csd' = GHC.EpaCommentsBalanced keep fc + + dd = GHC.L (GHC.EpAnn anc an csd') a + css = cs <> GHC.EpaComments move + + (ds',an') = rebalance (GHC.hsmodDecls p, GHC.hsmodAnn $ GHC.hsmodExt p) + p' = p { GHC.hsmodExt = (GHC.hsmodExt p){ GHC.hsmodAnn = an' }, + GHC.hsmodDecls = ds' + } + + rebalance :: ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + -> ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) + rebalance (ds, GHC.EpAnn a an cs) = (ds1, GHC.EpAnn a an cs') + where + (ds1,cs') = case break (\(GHC.AddEpAnn k _) -> k == GHC.AnnWhere) (GHC.am_main an) of + (_, (GHC.AddEpAnn _ whereLoc:_)) -> + case GHC.hsmodDecls p of + (d:ds0) -> (d':ds0, cs0) + where (d',cs0) = moveComments whereLoc d cs + ds0 -> (ds0,cs) + _ -> (ds,cs) + + + -- | Internal function. Initializes DynFlags value for parsing. -- -- Passes "-hide-all-packages" to the GHC API to prevent parsing of ===================================== utils/check-exact/Transform.hs ===================================== @@ -63,7 +63,7 @@ module Transform -- *** Low level operations used in 'HasDecls' , balanceComments , balanceCommentsList - , balanceCommentsList' + , balanceCommentsListA , anchorEof -- ** Managing lists, pure functions @@ -92,6 +92,7 @@ import qualified Control.Monad.Fail as Fail import GHC hiding (parseModule, parsedSource) import GHC.Data.FastString +import GHC.Types.SrcLoc import Data.Data import Data.Maybe @@ -154,6 +155,7 @@ logDataWithAnnsTr str ast = do -- |If we need to add new elements to the AST, they need their own -- 'SrcSpan' for this. +-- This should no longer be needed, we use an @EpaDelta@ location instead. uniqueSrcSpanT :: (Monad m) => TransformT m SrcSpan uniqueSrcSpanT = do col <- get @@ -171,15 +173,6 @@ srcSpanStartLine' _ = 0 -- --------------------------------------------------------------------- -captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag -captureOrderBinds ls = AnnSortKey $ map go ls - where - go (L _ (ValD _ _)) = BindTag - go (L _ (SigD _ _)) = SigDTag - go d = error $ "captureOrderBinds:" ++ showGhc d - --- --------------------------------------------------------------------- - captureMatchLineSpacing :: LHsDecl GhcPs -> LHsDecl GhcPs captureMatchLineSpacing (L l (ValD x (FunBind a b (MG c (L d ms ))))) = L l (ValD x (FunBind a b (MG c (L d ms')))) @@ -253,7 +246,7 @@ setEntryDPDecl d dp = setEntryDP d dp -- |Set the true entry 'DeltaPos' from the annotation for a given AST -- element. This is the 'DeltaPos' ignoring any comments. -setEntryDP :: NoAnn t => LocatedAn t a -> DeltaPos -> LocatedAn t a +setEntryDP :: LocatedAn t a -> DeltaPos -> LocatedAn t a setEntryDP (L (EpAnn (EpaSpan ss@(UnhelpfulSpan _)) an cs) a) dp = L (EpAnn (EpaDelta ss dp []) an cs) a setEntryDP (L (EpAnn (EpaSpan ss) an (EpaComments [])) a) dp @@ -293,7 +286,7 @@ setEntryDP (L (EpAnn (EpaSpan ss@(RealSrcSpan r _)) an cs) a) dp L (EpAnn (EpaDelta ss edp csd) an cs'') a where cs'' = setPriorComments cs [] - csd = L (EpaDelta ss dp NoComments) c:cs' + csd = L (EpaDelta ss dp NoComments) c:commentOrigDeltas cs' lc = last $ (L ca c:cs') delta = case getLoc lc of EpaSpan (RealSrcSpan rr _) -> ss2delta (ss2pos rr) r @@ -335,18 +328,15 @@ setEntryDPFromAnchor off (EpaSpan (RealSrcSpan anc _)) ll@(L la _) = setEntryDP -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -transferEntryDP :: (Monad m, NoAnn t2, Typeable t1, Typeable t2) - => LocatedAn t1 a -> LocatedAn t2 b -> TransformT m (LocatedAn t2 b) -transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = do - logTr $ "transferEntryDP': EpAnn,EpAnn" +transferEntryDP :: (Typeable t1, Typeable t2) + => LocatedAn t1 a -> LocatedAn t2 b -> (LocatedAn t2 b) +transferEntryDP (L (EpAnn anc1 an1 cs1) _) (L (EpAnn _anc2 an2 cs2) b) = -- Problem: if the original had preceding comments, blindly -- transferring the location is not correct case priorComments cs1 of - [] -> return (L (EpAnn anc1 (combine an1 an2) cs2) b) + [] -> (L (EpAnn anc1 (combine an1 an2) cs2) b) -- TODO: what happens if the receiving side already has comments? - (L anc _:_) -> do - logDataWithAnnsTr "transferEntryDP':priorComments anc=" anc - return (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) + (L _ _:_) -> (L (EpAnn anc1 (combine an1 an2) (cs1 <> cs2)) b) -- |If a and b are the same type return first arg, else return second @@ -356,10 +346,11 @@ combine x y = fromMaybe y (cast x) -- |Take the annEntryDelta associated with the first item and associate it with the second. -- Also transfer any comments occurring before it. -- TODO: call transferEntryDP, and use pushDeclDP -transferEntryDP' :: (Monad m) => LHsDecl GhcPs -> LHsDecl GhcPs -> TransformT m (LHsDecl GhcPs) -transferEntryDP' la lb = do - (L l2 b) <- transferEntryDP la lb - return (L l2 (pushDeclDP b (SameLine 0))) +transferEntryDP' :: LHsDecl GhcPs -> LHsDecl GhcPs -> (LHsDecl GhcPs) +transferEntryDP' la lb = + let + (L l2 b) = transferEntryDP la lb + in (L l2 (pushDeclDP b (SameLine 0))) pushDeclDP :: HsDecl GhcPs -> DeltaPos -> HsDecl GhcPs @@ -375,13 +366,24 @@ pushDeclDP d _dp = d -- --------------------------------------------------------------------- -balanceCommentsList :: (Monad m) => [LHsDecl GhcPs] -> TransformT m [LHsDecl GhcPs] -balanceCommentsList [] = return [] -balanceCommentsList [x] = return [x] -balanceCommentsList (a:b:ls) = do - (a',b') <- balanceComments a b - r <- balanceCommentsList (b':ls) - return (a':r) +-- | If we compile in haddock mode, the haddock processing inserts +-- DocDecls to carry the Haddock Documentation. We ignore these in +-- exact printing, as all the comments are also available in their +-- normal location, and the haddock processing is lossy, in that it +-- does not preserve all haddock-like comments. When we balance +-- comments in a list, we migrate some to preceding or following +-- declarations in the list. We must make sure we do not move any to +-- these DocDecls, which are not printed. +balanceCommentsList :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList decls = balanceCommentsList' (filter notDocDecl decls) + +balanceCommentsList' :: [LHsDecl GhcPs] -> [LHsDecl GhcPs] +balanceCommentsList' [] = [] +balanceCommentsList' [x] = [x] +balanceCommentsList' (a:b:ls) = (a':r) + where + (a',b') = balanceComments a b + r = balanceCommentsList' (b':ls) -- |The GHC parser puts all comments appearing between the end of one AST -- item and the beginning of the next as 'annPriorComments' for the second one. @@ -389,28 +391,27 @@ balanceCommentsList (a:b:ls) = do -- from the second one to the 'annFollowingComments' of the first if they belong -- to it instead. This is typically required before deleting or duplicating -- either of the AST elements. -balanceComments :: (Monad m) - => LHsDecl GhcPs -> LHsDecl GhcPs - -> TransformT m (LHsDecl GhcPs, LHsDecl GhcPs) -balanceComments first second = do +balanceComments :: LHsDecl GhcPs -> LHsDecl GhcPs + -> (LHsDecl GhcPs, LHsDecl GhcPs) +balanceComments first second = case first of - (L l (ValD x fb@(FunBind{}))) -> do - (L l' fb',second') <- balanceCommentsFB (L l fb) second - return (L l' (ValD x fb'), second') - _ -> balanceComments' first second + (L l (ValD x fb@(FunBind{}))) -> + let + (L l' fb',second') = balanceCommentsFB (L l fb) second + in (L l' (ValD x fb'), second') + _ -> balanceCommentsA first second --- |Once 'balanceComments' has been called to move trailing comments to a +-- |Once 'balanceCommentsA has been called to move trailing comments to a -- 'FunBind', these need to be pushed down from the top level to the last -- 'Match' if that 'Match' needs to be manipulated. -balanceCommentsFB :: (Monad m) - => LHsBind GhcPs -> LocatedA b -> TransformT m (LHsBind GhcPs, LocatedA b) -balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do - debugM $ "balanceCommentsFB entered: " ++ showGhc (ss2range $ locA lf) +balanceCommentsFB :: LHsBind GhcPs -> LocatedA b -> (LHsBind GhcPs, LocatedA b) +balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second + = balanceCommentsA (packFunBind bind) second' -- There are comments on lf. We need to -- + Keep the prior ones here -- + move the interior ones to the first match, -- + move the trailing ones to the last match. - let + where (before,middle,after) = case entry lf of EpaSpan (RealSrcSpan ss _) -> let @@ -426,40 +427,29 @@ balanceCommentsFB (L lf (FunBind x n (MG o (L lm matches)))) second = do getFollowingComments $ comments lf) lf' = setCommentsEpAnn lf (EpaComments before) - debugM $ "balanceCommentsFB (before, after): " ++ showAst (before, after) - debugM $ "balanceCommentsFB lf': " ++ showAst lf' - -- let matches' = case matches of - let matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] - matches' = case matches of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaComments middle )) m':ms') - _ -> error "balanceCommentsFB" - matches'' <- balanceCommentsList' matches' - let (m,ms) = case reverse matches'' of - (L lm' m':ms') -> - (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m',ms') - -- (L (addCommentsToEpAnnS lm' (EpaCommentsBalanced [] after)) m',ms') - _ -> error "balanceCommentsFB4" - debugM $ "balanceCommentsFB: (m,ms):" ++ showAst (m,ms) - (m',second') <- balanceComments' m second - m'' <- balanceCommentsMatch m' - let (m''',lf'') = case ms of - [] -> moveLeadingComments m'' lf' - _ -> (m'',lf') - debugM $ "balanceCommentsFB: (lf'', m'''):" ++ showAst (lf'',m''') - debugM $ "balanceCommentsFB done" - let bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) - debugM $ "balanceCommentsFB returning:" ++ showAst bind - balanceComments' (packFunBind bind) second' -balanceCommentsFB f s = balanceComments' f s + matches' :: [LocatedA (Match GhcPs (LHsExpr GhcPs))] + matches' = case matches of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaComments middle )) m0:ms') + _ -> error "balanceCommentsFB" + matches'' = balanceCommentsListA matches' + (m,ms) = case reverse matches'' of + (L lm' m0:ms') -> + (L (addCommentsToEpAnn lm' (EpaCommentsBalanced [] after)) m0,ms') + _ -> error "balanceCommentsFB4" + (m',second') = balanceCommentsA m second + m'' = balanceCommentsMatch m' + (m''',lf'') = case ms of + [] -> moveLeadingComments m'' lf' + _ -> (m'',lf') + bind = L lf'' (FunBind x n (MG o (L lm (reverse (m''':ms))))) +balanceCommentsFB f s = balanceCommentsA f s -- | Move comments on the same line as the end of the match into the -- GRHS, prior to the binds -balanceCommentsMatch :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do - logTr $ "balanceCommentsMatch: (logInfo)=" ++ showAst (logInfo) - return (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) +balanceCommentsMatch :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) + = (L l'' (Match am mctxt pats (GRHSs xg grhss' binds'))) where simpleBreak (r,_) = r /= 0 an1 = l @@ -468,7 +458,7 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do (move',stay') = break simpleBreak (trailingCommentsDeltas (anchorFromLocatedA (L l ())) cs1f) move = map snd move' stay = map snd stay' - (l'', grhss', binds', logInfo) + (l'', grhss', binds', _logInfo) = case reverse grhss of [] -> (l, [], binds, (EpaComments [], noSrcSpanA)) (L lg (GRHS ag grs rhs):gs) -> @@ -491,26 +481,24 @@ balanceCommentsMatch (L l (Match am mctxt pats (GRHSs xg grhss binds))) = do pushTrailingComments :: WithWhere -> EpAnnComments -> HsLocalBinds GhcPs -> (Bool, HsLocalBinds GhcPs) pushTrailingComments _ _cs b at EmptyLocalBinds{} = (False, b) pushTrailingComments _ _cs (HsIPBinds _ _) = error "TODO: pushTrailingComments:HsIPBinds" -pushTrailingComments w cs lb@(HsValBinds an _) - = (True, HsValBinds an' vb) +pushTrailingComments w cs lb@(HsValBinds an _) = (True, HsValBinds an' vb) where decls = hsDeclsLocalBinds lb (an', decls') = case reverse decls of [] -> (addCommentsToEpAnn an cs, decls) (L la d:ds) -> (an, L (addCommentsToEpAnn la cs) d:ds) - (vb,_ws2) = case runTransform (replaceDeclsValbinds w lb (reverse decls')) of - ((HsValBinds _ vb'), _, ws2') -> (vb', ws2') - _ -> (ValBinds NoAnnSortKey [] [], []) + vb = case replaceDeclsValbinds w lb (reverse decls') of + (HsValBinds _ vb') -> vb' + _ -> ValBinds NoAnnSortKey [] [] -balanceCommentsList' :: (Monad m) => [LocatedA a] -> TransformT m [LocatedA a] -balanceCommentsList' [] = return [] -balanceCommentsList' [x] = return [x] -balanceCommentsList' (a:b:ls) = do - logTr $ "balanceCommentsList' entered" - (a',b') <- balanceComments' a b - r <- balanceCommentsList' (b':ls) - return (a':r) +balanceCommentsListA :: [LocatedA a] -> [LocatedA a] +balanceCommentsListA [] = [] +balanceCommentsListA [x] = [x] +balanceCommentsListA (a:b:ls) = (a':r) + where + (a',b') = balanceCommentsA a b + r = balanceCommentsListA (b':ls) -- |Prior to moving an AST element, make sure any trailing comments belonging to -- it are attached to it, and not the following element. Of necessity this is a @@ -518,13 +506,8 @@ balanceCommentsList' (a:b:ls) = do -- with a passed-in decision function. -- The initial situation is that all comments for a given anchor appear as prior comments -- Many of these should in fact be following comments for the previous anchor -balanceComments' :: (Monad m) => LocatedA a -> LocatedA b -> TransformT m (LocatedA a, LocatedA b) -balanceComments' la1 la2 = do - debugM $ "balanceComments': (anc1)=" ++ showAst (anc1) - debugM $ "balanceComments': (cs1s)=" ++ showAst (cs1s) - debugM $ "balanceComments': (cs1stay,cs1move)=" ++ showAst (cs1stay,cs1move) - debugM $ "balanceComments': (an1',an2')=" ++ showAst (an1',an2') - return (la1', la2') +balanceCommentsA :: LocatedA a -> LocatedA b -> (LocatedA a, LocatedA b) +balanceCommentsA la1 la2 = (la1', la2') where simpleBreak n (r,_) = r > n L an1 f = la1 @@ -532,26 +515,31 @@ balanceComments' la1 la2 = do anc1 = comments an1 anc2 = comments an2 - cs1s = splitCommentsEnd (anchorFromLocatedA la1) anc1 - cs1p = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1s) - cs1f = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1s) + (p1,m1,f1) = splitComments (anchorFromLocatedA la1) anc1 + cs1p = priorCommentsDeltas (anchorFromLocatedA la1) p1 - cs2s = splitCommentsEnd (anchorFromLocatedA la2) anc2 - cs2p = priorCommentsDeltas (anchorFromLocatedA la2) (priorComments cs2s) - cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) (getFollowingComments cs2s) + -- Split cs1 following comments into those before any + -- TrailingAnn's on an1, and any after + cs1f = splitCommentsEnd (fullSpanFromLocatedA la1) $ EpaComments f1 + cs1fp = priorCommentsDeltas (anchorFromLocatedA la1) (priorComments cs1f) + cs1ff = trailingCommentsDeltas (anchorFromLocatedA la1) (getFollowingComments cs1f) - -- Split cs1f into those that belong on an1 and ones that must move to an2 - (cs1move,cs1stay) = break (simpleBreak 1) cs1f + -- Split cs1ff into those that belong on an1 and ones that must move to an2 + (cs1move,cs1stay) = break (simpleBreak 1) cs1ff + + (p2,m2,f2) = splitComments (anchorFromLocatedA la2) anc2 + cs2p = priorCommentsDeltas (anchorFromLocatedA la2) p2 + cs2f = trailingCommentsDeltas (anchorFromLocatedA la2) f2 (stay'',move') = break (simpleBreak 1) cs2p -- Need to also check for comments more closely attached to la1, -- ie trailing on the same line (move'',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchorFromLocatedA la1) (map snd stay'')) - move = sortEpaComments $ map snd (cs1move ++ move'' ++ move') - stay = sortEpaComments $ map snd (cs1stay ++ stay') + move = sortEpaComments $ map snd (cs1fp ++ cs1move ++ move'' ++ move') + stay = sortEpaComments $ m2 ++ map snd (cs1stay ++ stay') - an1' = setCommentsEpAnn (getLoc la1) (EpaCommentsBalanced (map snd cs1p) move) - an2' = setCommentsEpAnn (getLoc la2) (EpaCommentsBalanced stay (map snd cs2f)) + an1' = setCommentsEpAnn (getLoc la1) (epaCommentsBalanced (m1 ++ map snd cs1p) move) + an2' = setCommentsEpAnn (getLoc la2) (epaCommentsBalanced stay (map snd cs2f)) la1' = L an1' f la2' = L an2' s @@ -569,10 +557,9 @@ trailingCommentsDeltas r (la@(L l _):las) (al,_) = ss2posEnd rs' (ll,_) = ss2pos (anchor loc) --- AZ:TODO: this is identical to commentsDeltas priorCommentsDeltas :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] -priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) +priorCommentsDeltas r cs = go r (sortEpaComments cs) where go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] go _ [] = [] @@ -588,6 +575,21 @@ priorCommentsDeltas r cs = go r (reverse $ sortEpaComments cs) -- --------------------------------------------------------------------- +-- | Split comments into ones occurring before the end of the reference +-- span, and those after it. +splitComments :: RealSrcSpan -> EpAnnComments -> ([LEpaComment], [LEpaComment], [LEpaComment]) +splitComments p cs = (before, middle, after) + where + cmpe (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p + cmpe (L _ _) = True + + cmpb (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2pos p + cmpb (L _ _) = True + + (beforeEnd, after) = break cmpe ((priorComments cs) ++ (getFollowingComments cs)) + (before, middle) = break cmpb beforeEnd + + -- | Split comments into ones occurring before the end of the reference -- span, and those after it. splitCommentsEnd :: RealSrcSpan -> EpAnnComments -> EpAnnComments @@ -598,8 +600,8 @@ splitCommentsEnd p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsEnd p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsEnd p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -617,8 +619,8 @@ splitCommentsStart p (EpaComments cs) = cs' (before, after) = break cmp cs cs' = case after of [] -> EpaComments cs - _ -> EpaCommentsBalanced before after -splitCommentsStart p (EpaCommentsBalanced cs ts) = EpaCommentsBalanced cs' ts' + _ -> epaCommentsBalanced before after +splitCommentsStart p (EpaCommentsBalanced cs ts) = epaCommentsBalanced cs' ts' where cmp (L (EpaSpan (RealSrcSpan l _)) _) = ss2pos l > ss2posEnd p cmp (L _ _) = True @@ -638,8 +640,8 @@ moveLeadingComments (L la a) lb = (L la' a, lb') -- TODO: need to set an entry delta on lb' to zero, and move the -- original spacing to the first comment. - la' = setCommentsEpAnn la (EpaCommentsBalanced [] after) - lb' = addCommentsToEpAnn lb (EpaCommentsBalanced before []) + la' = setCommentsEpAnn la (epaCommentsBalanced [] after) + lb' = addCommentsToEpAnn lb (epaCommentsBalanced before []) -- | A GHC comment includes the span of the preceding (non-comment) -- token. Takes an original list of comments, and converts the @@ -662,17 +664,27 @@ addCommentOrigDeltasAnn (EpAnn e a cs) = EpAnn e a (addCommentOrigDeltas cs) anchorFromLocatedA :: LocatedA a -> RealSrcSpan anchorFromLocatedA (L (EpAnn anc _ _) _) = anchor anc +-- | Get the full span of interest for comments from a LocatedA. +-- This extends up to the last TrailingAnn +fullSpanFromLocatedA :: LocatedA a -> RealSrcSpan +fullSpanFromLocatedA (L (EpAnn anc (AnnListItem tas) _) _) = rr + where + r = anchor anc + trailing_loc ta = case ta_location ta of + EpaSpan (RealSrcSpan s _) -> [s] + _ -> [] + rr = case reverse (concatMap trailing_loc tas) of + [] -> r + (s:_) -> combineRealSrcSpans r s + -- --------------------------------------------------------------------- -balanceSameLineComments :: (Monad m) - => LMatch GhcPs (LHsExpr GhcPs) -> TransformT m (LMatch GhcPs (LHsExpr GhcPs)) -balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do - logTr $ "balanceSameLineComments: (la)=" ++ showGhc (ss2range $ locA la) - logTr $ "balanceSameLineComments: [logInfo]=" ++ showAst logInfo - return (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) +balanceSameLineComments :: LMatch GhcPs (LHsExpr GhcPs) -> (LMatch GhcPs (LHsExpr GhcPs)) +balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) + = (L la' (Match anm mctxt pats (GRHSs x grhss' lb))) where simpleBreak n (r,_) = r > n - (la',grhss', logInfo) = case reverse grhss of + (la',grhss', _logInfo) = case reverse grhss of [] -> (la,grhss,[]) (L lg (GRHS ga gs rhs):grs) -> (la'',reverse $ (L lg (GRHS ga' gs rhs)):grs,[(gac,(csp,csf))]) where @@ -684,7 +696,7 @@ balanceSameLineComments (L la (Match anm mctxt pats (GRHSs x grhss lb))) = do (move',stay') = break (simpleBreak 0) (trailingCommentsDeltas (anchor anc) csf) move = map snd move' stay = map snd stay' - cs1 = EpaCommentsBalanced csp stay + cs1 = epaCommentsBalanced csp stay gac = epAnnComments ga gfc = getFollowingComments gac @@ -734,24 +746,21 @@ addComma (EpAnn anc (AnnListItem as) cs) -- | Insert a declaration into an AST element having sub-declarations -- (@HasDecls@) according to the given location function. insertAt :: (HasDecls ast) - => (LHsDecl GhcPs - -> [LHsDecl GhcPs] - -> [LHsDecl GhcPs]) - -> ast - -> LHsDecl GhcPs - -> Transform ast -insertAt f t decl = do - oldDecls <- hsDecls t - oldDeclsb <- balanceCommentsList oldDecls - let oldDecls' = oldDeclsb - replaceDecls t (f decl oldDecls') + => (LHsDecl GhcPs + -> [LHsDecl GhcPs] + -> [LHsDecl GhcPs]) + -> ast + -> LHsDecl GhcPs + -> ast +insertAt f t decl = replaceDecls t (f decl oldDecls') + where + oldDecls = hsDecls t + oldDeclsb = balanceCommentsList oldDecls + oldDecls' = oldDeclsb -- |Insert a declaration at the beginning or end of the subdecls of the given -- AST item -insertAtStart, insertAtEnd :: (HasDecls ast) - => ast - -> LHsDecl GhcPs - -> Transform ast +insertAtStart, insertAtEnd :: HasDecls ast => ast -> LHsDecl GhcPs -> ast insertAtEnd = insertAt (\x xs -> xs ++ [x]) @@ -766,11 +775,11 @@ insertAtStart = insertAt insertFirst -- |Insert a declaration at a specific location in the subdecls of the given -- AST item -insertAfter, insertBefore :: (HasDecls (LocatedA ast)) +insertAfter, insertBefore :: HasDecls (LocatedA ast) => LocatedA old -> LocatedA ast -> LHsDecl GhcPs - -> Transform (LocatedA ast) + -> LocatedA ast insertAfter (getLocA -> k) = insertAt findAfter where findAfter x xs = @@ -797,10 +806,10 @@ class (Data t) => HasDecls t where -- given syntax phrase. They are always returned in the wrapped 'HsDecl' -- form, even if orginating in local decls. This is safe, as annotations -- never attach to the wrapper, only to the wrapped item. - hsDecls :: (Monad m) => t -> TransformT m [LHsDecl GhcPs] + hsDecls :: t -> [LHsDecl GhcPs] -- | Replace the directly enclosed decl list by the given - -- decl list. Runs in the 'Transform' monad to be able to update list order + -- decl list. As part of replacing it will update list order -- annotations, and rebalance comments and other layout changes as needed. -- -- For example, a call on replaceDecls for a wrapped 'FunBind' having no @@ -818,96 +827,86 @@ class (Data t) => HasDecls t where -- where -- nn = 2 -- @ - replaceDecls :: (Monad m) => t -> [LHsDecl GhcPs] -> TransformT m t + replaceDecls :: t -> [LHsDecl GhcPs] -> t -- --------------------------------------------------------------------- instance HasDecls ParsedSource where - hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = return decls + hsDecls (L _ (HsModule (XModulePs _ _lo _ _) _mn _exps _imps decls)) = decls replaceDecls (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps _decls)) decls - = do - logTr "replaceDecls LHsModule" - return (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) + = (L l (HsModule (XModulePs a lo deps haddocks) mname exps imps decls)) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsDecl GhcPs)) where - hsDecls (L _ (TyClD _ c at ClassDecl{})) = return $ hsDeclsClassDecl c - hsDecls decl = do - error $ "hsDecls:decl=" ++ showAst decl - replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = do - let decl' = replaceDeclsClassDecl dec decls - return (L l (TyClD e decl')) - replaceDecls decl _decls = do - error $ "replaceDecls:decl=" ++ showAst decl + hsDecls (L _ (TyClD _ c at ClassDecl{})) = hsDeclsClassDecl c + hsDecls decl = error $ "hsDecls:decl=" ++ showAst decl + replaceDecls (L l (TyClD e dec at ClassDecl{})) decls = + let + decl' = replaceDeclsClassDecl dec decls + in (L l (TyClD e decl')) + replaceDecls decl _decls + = error $ "replaceDecls:decl=" ++ showAst decl -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Match GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (Match _ _ _ (GRHSs _ _ lb))) = hsDeclsLocalBinds lb replaceDecls (L l (Match xm c p (GRHSs xr rhs binds))) [] - = do - logTr "replaceDecls LMatch empty decls" - binds'' <- replaceDeclsValbinds WithoutWhere binds [] - return (L l (Match xm c p (GRHSs xr rhs binds''))) + = let + binds'' = replaceDeclsValbinds WithoutWhere binds [] + in (L l (Match xm c p (GRHSs xr rhs binds''))) replaceDecls m@(L l (Match xm c p (GRHSs xr rhs binds))) newBinds - = do - logTr "replaceDecls LMatch nonempty decls" + = let -- Need to throw in a fresh where clause if the binds were empty, -- in the annotations. - (l', rhs') <- case binds of - EmptyLocalBinds{} -> do - logTr $ "replaceDecls LMatch empty binds" - - logDataWithAnnsTr "Match.replaceDecls:balancing comments:m" m - L l' m' <- balanceSameLineComments m - logDataWithAnnsTr "Match.replaceDecls:(m1')" (L l' m') - return (l', grhssGRHSs $ m_grhss m') - _ -> return (l, rhs) - binds'' <- replaceDeclsValbinds WithWhere binds newBinds - logDataWithAnnsTr "Match.replaceDecls:binds'" binds'' - return (L l' (Match xm c p (GRHSs xr rhs' binds''))) + (l', rhs') = case binds of + EmptyLocalBinds{} -> + let + L l0 m' = balanceSameLineComments m + in (l0, grhssGRHSs $ m_grhss m') + _ -> (l, rhs) + binds'' = replaceDeclsValbinds WithWhere binds newBinds + in (L l' (Match xm c p (GRHSs xr rhs' binds''))) -- --------------------------------------------------------------------- instance HasDecls (LocatedA (HsExpr GhcPs)) where - hsDecls (L _ (HsLet _ decls _ex)) = return $ hsDeclsLocalBinds decls - hsDecls _ = return [] + hsDecls (L _ (HsLet _ decls _ex)) = hsDeclsLocalBinds decls + hsDecls _ = [] replaceDecls (L ll (HsLet (tkLet, tkIn) binds ex)) newDecls - = do - logTr "replaceDecls HsLet" - let lastAnc = realSrcSpan $ spanHsLocaLBinds binds + = let + lastAnc = realSrcSpan $ spanHsLocaLBinds binds -- TODO: may be an intervening comment, take account for lastAnc - let (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of - (EpTok l, EpTok i) -> - let - off = case l of - (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r - (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 - (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 - (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c - ex'' = setEntryDPFromAnchor off i ex - newDecls'' = case newDecls of - [] -> newDecls - (d:ds) -> setEntryDPDecl d (SameLine 0) : ds - in ( EpTok l - , EpTok (addEpaLocationDelta off lastAnc i) - , ex'' - , newDecls'') - (_,_) -> (tkLet, tkIn, ex, newDecls) - binds' <- replaceDeclsValbinds WithoutWhere binds newDecls' - return (L ll (HsLet (tkLet', tkIn') binds' ex')) + (tkLet', tkIn', ex',newDecls') = case (tkLet, tkIn) of + (EpTok l, EpTok i) -> + let + off = case l of + (EpaSpan (RealSrcSpan r _)) -> LayoutStartCol $ snd $ ss2pos r + (EpaSpan (UnhelpfulSpan _)) -> LayoutStartCol 0 + (EpaDelta _ (SameLine _) _) -> LayoutStartCol 0 + (EpaDelta _ (DifferentLine _ c) _) -> LayoutStartCol c + ex'' = setEntryDPFromAnchor off i ex + newDecls'' = case newDecls of + [] -> newDecls + (d:ds) -> setEntryDPDecl d (SameLine 0) : ds + in ( EpTok l + , EpTok (addEpaLocationDelta off lastAnc i) + , ex'' + , newDecls'') + (_,_) -> (tkLet, tkIn, ex, newDecls) + binds' = replaceDeclsValbinds WithoutWhere binds newDecls' + in (L ll (HsLet (tkLet', tkIn') binds' ex')) -- TODO: does this make sense? Especially as no hsDecls for HsPar replaceDecls (L l (HsPar x e)) newDecls - = do - logTr "replaceDecls HsPar" - e' <- replaceDecls e newDecls - return (L l (HsPar x e')) + = let + e' = replaceDecls e newDecls + in (L l (HsPar x e')) replaceDecls old _new = error $ "replaceDecls (LHsExpr GhcPs) undefined for:" ++ showGhc old -- --------------------------------------------------------------------- @@ -934,53 +933,51 @@ hsDeclsPatBind x = error $ "hsDeclsPatBind called for:" ++ showGhc x -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBindD' \/ 'replaceDeclsPatBindD' is -- idempotent. -replaceDeclsPatBindD :: (Monad m) => LHsDecl GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsDecl GhcPs) -replaceDeclsPatBindD (L l (ValD x d)) newDecls = do - (L _ d') <- replaceDeclsPatBind (L l d) newDecls - return (L l (ValD x d')) +replaceDeclsPatBindD :: LHsDecl GhcPs -> [LHsDecl GhcPs] -> (LHsDecl GhcPs) +replaceDeclsPatBindD (L l (ValD x d)) newDecls = + let + (L _ d') = replaceDeclsPatBind (L l d) newDecls + in (L l (ValD x d')) replaceDeclsPatBindD x _ = error $ "replaceDeclsPatBindD called for:" ++ showGhc x -- | Replace the immediate declarations for a 'PatBind'. This -- cannot be a member of 'HasDecls' because a 'FunBind' is not idempotent -- for 'hsDecls' \/ 'replaceDecls'. 'hsDeclsPatBind' \/ 'replaceDeclsPatBind' is -- idempotent. -replaceDeclsPatBind :: (Monad m) => LHsBind GhcPs -> [LHsDecl GhcPs] - -> TransformT m (LHsBind GhcPs) +replaceDeclsPatBind :: LHsBind GhcPs -> [LHsDecl GhcPs] -> (LHsBind GhcPs) replaceDeclsPatBind (L l (PatBind x a p (GRHSs xr rhss binds))) newDecls - = do - logTr "replaceDecls PatBind" - binds'' <- replaceDeclsValbinds WithWhere binds newDecls - return (L l (PatBind x a p (GRHSs xr rhss binds''))) + = (L l (PatBind x a p (GRHSs xr rhss binds''))) + where + binds'' = replaceDeclsValbinds WithWhere binds newDecls replaceDeclsPatBind x _ = error $ "replaceDeclsPatBind called for:" ++ showGhc x -- --------------------------------------------------------------------- instance HasDecls (LocatedA (Stmt GhcPs (LocatedA (HsExpr GhcPs)))) where - hsDecls (L _ (LetStmt _ lb)) = return $ hsDeclsLocalBinds lb + hsDecls (L _ (LetStmt _ lb)) = hsDeclsLocalBinds lb hsDecls (L _ (LastStmt _ e _ _)) = hsDecls e hsDecls (L _ (BindStmt _ _pat e)) = hsDecls e hsDecls (L _ (BodyStmt _ e _ _)) = hsDecls e - hsDecls _ = return [] + hsDecls _ = [] replaceDecls (L l (LetStmt x lb)) newDecls - = do - lb'' <- replaceDeclsValbinds WithWhere lb newDecls - return (L l (LetStmt x lb'')) + = let + lb'' = replaceDeclsValbinds WithWhere lb newDecls + in (L l (LetStmt x lb'')) replaceDecls (L l (LastStmt x e d se)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (LastStmt x e' d se)) + = let + e' = replaceDecls e newDecls + in (L l (LastStmt x e' d se)) replaceDecls (L l (BindStmt x pat e)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BindStmt x pat e')) + = let + e' = replaceDecls e newDecls + in (L l (BindStmt x pat e')) replaceDecls (L l (BodyStmt x e a b)) newDecls - = do - e' <- replaceDecls e newDecls - return (L l (BodyStmt x e' a b)) - replaceDecls x _newDecls = return x + = let + e' = replaceDecls e newDecls + in (L l (BodyStmt x e' a b)) + replaceDecls x _newDecls = x -- ===================================================================== -- end of HasDecls instances @@ -1062,61 +1059,55 @@ data WithWhere = WithWhere -- care, as this does not manage the declaration order, the -- ordering should be done by the calling function from the 'HsLocalBinds' -- context in the AST. -replaceDeclsValbinds :: (Monad m) - => WithWhere +replaceDeclsValbinds :: WithWhere -> HsLocalBinds GhcPs -> [LHsDecl GhcPs] - -> TransformT m (HsLocalBinds GhcPs) -replaceDeclsValbinds _ _ [] = do - return (EmptyLocalBinds NoExtField) + -> HsLocalBinds GhcPs +replaceDeclsValbinds _ _ [] = EmptyLocalBinds NoExtField replaceDeclsValbinds w b@(HsValBinds a _) new - = do - logTr "replaceDeclsValbinds" - let oldSpan = spanHsLocaLBinds b - an <- oldWhereAnnotation a w (realSrcSpan oldSpan) - let decs = concatMap decl2Bind new - let sigs = concatMap decl2Sig new - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) + = let + oldSpan = spanHsLocaLBinds b + an = oldWhereAnnotation a w (realSrcSpan oldSpan) + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) replaceDeclsValbinds _ (HsIPBinds {}) _new = error "undefined replaceDecls HsIPBinds" replaceDeclsValbinds w (EmptyLocalBinds _) new - = do - logTr "replaceDecls HsLocalBinds" - an <- newWhereAnnotation w - let newBinds = concatMap decl2Bind new - newSigs = concatMap decl2Sig new - let decs = newBinds - let sigs = newSigs - let sortKey = captureOrderBinds new - return (HsValBinds an (ValBinds sortKey decs sigs)) - -oldWhereAnnotation :: (Monad m) - => EpAnn AnnList -> WithWhere -> RealSrcSpan -> TransformT m (EpAnn AnnList) -oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = do - -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, change the AnnList anchor to have the correct DP too - let (AnnList ancl o c _r t) = an - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - (anc', ancl') <- do - case ww of - WithWhere -> return (anc, ancl) - WithoutWhere -> return (anc, ancl) - let an' = EpAnn anc' - (AnnList ancl' o c w t) - cs - return an' - -newWhereAnnotation :: (Monad m) => WithWhere -> TransformT m (EpAnn AnnList) -newWhereAnnotation ww = do - let anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] - let anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] - let w = case ww of - WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] - WithoutWhere -> [] - let an = EpAnn anc - (AnnList (Just anc2) Nothing Nothing w []) - emptyComments - return an + = let + an = newWhereAnnotation w + decs = concatMap decl2Bind new + sigs = concatMap decl2Sig new + sortKey = captureOrderBinds new + in (HsValBinds an (ValBinds sortKey decs sigs)) + +oldWhereAnnotation :: EpAnn AnnList -> WithWhere -> RealSrcSpan -> (EpAnn AnnList) +oldWhereAnnotation (EpAnn anc an cs) ww _oldSpan = an' + -- TODO: when we set DP (0,0) for the HsValBinds EpEpaLocation, + -- change the AnnList anchor to have the correct DP too + where + (AnnList ancl o c _r t) = an + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + (anc', ancl') = + case ww of + WithWhere -> (anc, ancl) + WithoutWhere -> (anc, ancl) + an' = EpAnn anc' + (AnnList ancl' o c w t) + cs + +newWhereAnnotation :: WithWhere -> (EpAnn AnnList) +newWhereAnnotation ww = an + where + anc = EpaDelta noSrcSpan (DifferentLine 1 3) [] + anc2 = EpaDelta noSrcSpan (DifferentLine 1 5) [] + w = case ww of + WithWhere -> [AddEpAnn AnnWhere (EpaDelta noSrcSpan (SameLine 0) [])] + WithoutWhere -> [] + an = EpAnn anc + (AnnList (Just anc2) Nothing Nothing w []) + emptyComments -- --------------------------------------------------------------------- @@ -1127,32 +1118,32 @@ type PMatch = LMatch GhcPs (LHsExpr GhcPs) -- declarations are extracted and returned after modification. For a -- 'FunBind' the supplied 'SrcSpan' is used to identify the specific -- 'Match' to be transformed, for when there are multiple of them. -modifyValD :: forall m t. (HasTransform m) - => SrcSpan +modifyValD :: forall t. + SrcSpan -> Decl - -> (PMatch -> [Decl] -> m ([Decl], Maybe t)) - -> m (Decl,Maybe t) + -> (PMatch -> [Decl] -> ([Decl], Maybe t)) + -> (Decl,Maybe t) modifyValD p pb@(L ss (ValD _ (PatBind {} ))) f = if (locA ss) == p - then do - let ds = hsDeclsPatBindD pb - (ds',r) <- f (error "modifyValD.PatBind should not touch Match") ds - pb' <- liftT $ replaceDeclsPatBindD pb ds' - return (pb',r) - else return (pb,Nothing) -modifyValD p decl f = do - (decl',r) <- runStateT (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing - return (packFunDecl decl',r) + then + let + ds = hsDeclsPatBindD pb + (ds',r) = f (error "modifyValD.PatBind should not touch Match") ds + pb' = replaceDeclsPatBindD pb ds' + in (pb',r) + else (pb,Nothing) +modifyValD p decl f = (packFunDecl decl', r) where - doModLocal :: PMatch -> StateT (Maybe t) m PMatch + (decl',r) = runState (everywhereM (mkM doModLocal) (unpackFunDecl decl)) Nothing + doModLocal :: PMatch -> State (Maybe t) PMatch doModLocal (match@(L ss _) :: PMatch) = do if (locA ss) == p then do - ds <- lift $ liftT $ hsDecls match - `debug` ("modifyValD: match=" ++ showAst match) - (ds',r) <- lift $ f match ds - put r - match' <- lift $ liftT $ replaceDecls match ds' + let + ds = hsDecls match + (ds',r0) = f match ds + put r0 + let match' = replaceDecls match ds' return match' else return match @@ -1172,6 +1163,6 @@ modifyDeclsT :: (HasDecls t,HasTransform m) => ([LHsDecl GhcPs] -> m [LHsDecl GhcPs]) -> t -> m t modifyDeclsT action t = do - decls <- liftT $ hsDecls t + let decls = hsDecls t decls' <- action decls - liftT $ replaceDecls t decls' + return $ replaceDecls t decls' ===================================== utils/check-exact/Types.hs ===================================== @@ -21,10 +21,6 @@ type Pos = (Int,Int) -- --------------------------------------------------------------------- -data Rigidity = NormalLayout | RigidLayout deriving (Eq, Ord, Show) - --- --------------------------------------------------------------------- - -- | A Haskell comment. The @AnnKeywordId@ is present if it has been converted -- from an @AnnKeywordId@ because the annotation must be interleaved into the -- stream and does not have a well-defined position ===================================== utils/check-exact/Utils.hs ===================================== @@ -20,9 +20,8 @@ module Utils where import Control.Monad (when) +import GHC.Utils.Monad.State.Strict import Data.Function -import Data.Maybe (isJust) -import Data.Ord (comparing) import GHC.Hs.Dump import Lookup @@ -36,9 +35,10 @@ import GHC.Driver.Ppr import GHC.Data.FastString import qualified GHC.Data.Strict as Strict import GHC.Base (NonEmpty(..)) +import GHC.Parser.Lexer (allocateComments) import Data.Data hiding ( Fixity ) -import Data.List (sortBy, elemIndex) +import Data.List (sortBy, partition) import qualified Data.Map.Strict as Map import Debug.Trace @@ -60,12 +60,32 @@ debug c s = if debugEnabledFlag debugM :: Monad m => String -> m () debugM s = when debugEnabledFlag $ traceM s --- --------------------------------------------------------------------- - warn :: c -> String -> c -- warn = flip trace warn c _ = c +-- --------------------------------------------------------------------- + +captureOrderBinds :: [LHsDecl GhcPs] -> AnnSortKey BindTag +captureOrderBinds ls = AnnSortKey $ map go ls + where + go (L _ (ValD _ _)) = BindTag + go (L _ (SigD _ _)) = SigDTag + go d = error $ "captureOrderBinds:" ++ showGhc d + +-- --------------------------------------------------------------------- + +notDocDecl :: LHsDecl GhcPs -> Bool +notDocDecl (L _ DocD{}) = False +notDocDecl _ = True + +notIEDoc :: LIE GhcPs -> Bool +notIEDoc (L _ IEGroup {}) = False +notIEDoc (L _ IEDoc {}) = False +notIEDoc (L _ IEDocNamed {}) = False +notIEDoc _ = True + +-- --------------------------------------------------------------------- -- | A good delta has no negative values. isGoodDelta :: DeltaPos -> Bool isGoodDelta (SameLine co) = co >= 0 @@ -108,7 +128,6 @@ pos2delta (refl,refc) (l,c) = deltaPos lo co lo = l - refl co = if lo == 0 then c - refc else c - -- else c - 1 -- | Apply the delta to the current position, taking into account the -- current column offset if advancing to a new line @@ -200,23 +219,6 @@ origDelta pos pp = ss2delta (ss2posEnd pp) pos -- --------------------------------------------------------------------- --- |Given a list of items and a list of keys, returns a list of items --- ordered by their position in the list of keys. -orderByKey :: [(DeclTag,a)] -> [DeclTag] -> [(DeclTag,a)] -orderByKey keys order - -- AZ:TODO: if performance becomes a problem, consider a Map of the order - -- SrcSpan to an index, and do a lookup instead of elemIndex. - - -- Items not in the ordering are placed to the start - = sortBy (comparing (flip elemIndex order . fst)) keys - --- --------------------------------------------------------------------- - -isListComp :: HsDoFlavour -> Bool -isListComp = isDoComprehensionContext - --- --------------------------------------------------------------------- - needsWhere :: DataDefnCons (LConDecl (GhcPass p)) -> Bool needsWhere (NewTypeCon _) = True needsWhere (DataTypeCons _ []) = True @@ -225,21 +227,214 @@ needsWhere _ = False -- --------------------------------------------------------------------- +-- | Insert the comments at the appropriate places in the AST insertCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource -insertCppComments (L l p) cs = L l p' +-- insertCppComments p [] = p +insertCppComments (L l p) cs0 = insertRemainingCppComments (L l p2) remaining + where + (EpAnn anct ant cst) = hsmodAnn $ hsmodExt p + cs = sortEpaComments $ priorComments cst ++ getFollowingComments cst ++ cs0 + p0 = p { hsmodExt = (hsmodExt p) { hsmodAnn = EpAnn anct ant emptyComments }} + -- Comments embedded within spans + -- everywhereM is a bottom-up traversal + (p1, toplevel) = runState (everywhereM (mkM addCommentsListItem + `extM` addCommentsGrhs + `extM` addCommentsList) p0) cs + (p2, remaining) = insertTopLevelCppComments p1 toplevel + + addCommentsListItem :: EpAnn AnnListItem -> State [LEpaComment] (EpAnn AnnListItem) + addCommentsListItem = addComments + + addCommentsList :: EpAnn AnnList -> State [LEpaComment] (EpAnn AnnList) + addCommentsList = addComments + + addCommentsGrhs :: EpAnn GrhsAnn -> State [LEpaComment] (EpAnn GrhsAnn) + addCommentsGrhs = addComments + + addComments :: forall ann. EpAnn ann -> State [LEpaComment] (EpAnn ann) + addComments (EpAnn anc an ocs) = do + case anc of + EpaSpan (RealSrcSpan s _) -> do + unAllocated <- get + let + (rest, these) = GHC.Parser.Lexer.allocateComments s unAllocated + cs' = workInComments ocs these + put rest + return $ EpAnn anc an cs' + + _ -> return $ EpAnn anc an ocs + +workInComments :: EpAnnComments -> [LEpaComment] -> EpAnnComments +workInComments ocs [] = ocs +workInComments ocs new = cs' + where + pc = priorComments ocs + fc = getFollowingComments ocs + cs' = case fc of + [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ new + (L ac _:_) -> epaCommentsBalanced (sortEpaComments $ pc ++ cs_before) + (sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) + new + +insertTopLevelCppComments :: HsModule GhcPs -> [LEpaComment] -> (HsModule GhcPs, [LEpaComment]) +insertTopLevelCppComments (HsModule (XModulePs an lo mdeprec mbDoc) mmn mexports imports decls) cs + = (HsModule (XModulePs an4 lo mdeprec mbDoc) mmn mexports' imports' decls', cs3) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0,hc1,hc_cs)) + -- `debug` ("insertTopLevelCppComments: (cs2,cs3,hc0i,hc0,hc1,hc_cs)" ++ showAst (cs2,cs3,hc0i,hc0,hc1,hc_cs)) where - an' = case GHC.hsmodAnn $ GHC.hsmodExt p of - (EpAnn a an ocs) -> EpAnn a an cs' - where - pc = priorComments ocs - fc = getFollowingComments ocs - cs' = case fc of - [] -> EpaComments $ sortEpaComments $ pc ++ fc ++ cs - (L ac _:_) -> EpaCommentsBalanced (sortEpaComments $ pc ++ cs_before) - (sortEpaComments $ fc ++ cs_after) - where - (cs_before,cs_after) = break (\(L ll _) -> (ss2pos $ anchor ll) < (ss2pos $ anchor ac) ) cs + -- Comments at the top level. + (an0, cs0) = + case mmn of + Nothing -> (an, cs) + Just _ -> + -- We have a module name. Capture all comments up to the `where` + let + (these, remaining) = splitOnWhere Before (am_main $ anns an) cs + (EpAnn a anno ocs) = an :: EpAnn AnnsModule + anm = EpAnn a anno (workInComments ocs these) + in + (anm, remaining) + (an1,cs0a) = case lo of + EpExplicitBraces (EpTok (EpaSpan (RealSrcSpan s _))) _close -> + let + (stay,cs0a') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0 + cs' = workInComments (comments an0) stay + in (an0 { comments = cs' }, cs0a') + _ -> (an0,cs0) + -- Deal with possible leading semis + (an2, cs0b) = case am_decls $ anns an1 of + (AddSemiAnn (EpaSpan (RealSrcSpan s _)):_) -> (an1 {comments = cs'}, cs0b') + where + (stay,cs0b') = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ s)) cs0a + cs' = workInComments (comments an1) stay + _ -> (an1,cs0a) + + (mexports', an3, cs1) = + case mexports of + Nothing -> (Nothing, an2, cs0b) + Just (L l exports) -> (Just (L l exports'), an3', cse) + where + hc1' = workInComments (comments an2) csh' + an3' = an2 { comments = hc1' } + (csh', cs0b') = case al_open $ anns l of + Just (AddEpAnn _ (EpaSpan (RealSrcSpan s _))) ->(h, n) + where + (h,n) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + cs0b + + _ -> ([], cs0b) + (exports', cse) = allocPreceding exports cs0b' + (imports0, cs2) = allocPreceding imports cs1 + (imports', hc0i) = balanceFirstLocatedAComments imports0 + + (decls0, cs3) = allocPreceding decls cs2 + (decls', hc0d) = balanceFirstLocatedAComments decls0 + + -- Either hc0i or hc0d should have comments. Combine them + hc0 = hc0i ++ hc0d + + (hc1,hc_cs) = if null ( am_main $ anns an3) + then (hc0,[]) + else splitOnWhere After (am_main $ anns an3) hc0 + hc2 = workInComments (comments an3) hc1 + an4 = an3 { anns = (anns an3) {am_cs = hc_cs}, comments = hc2 } + + allocPreceding :: [LocatedA a] -> [LEpaComment] -> ([LocatedA a], [LEpaComment]) + allocPreceding [] cs' = ([], cs') + allocPreceding (L (EpAnn anc4 an5 cs4) a:xs) cs' = ((L (EpAnn anc4 an5 cs4') a:xs'), rest') + where + (rest, these) = + case anc4 of + EpaSpan (RealSrcSpan s _) -> + allocatePriorComments (ss2pos s) cs' + _ -> (cs', []) + cs4' = workInComments cs4 these + (xs',rest') = allocPreceding xs rest + +data SplitWhere = Before | After +splitOnWhere :: SplitWhere -> [AddEpAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitOnWhere _ [] csIn = (csIn,[]) +splitOnWhere w (AddEpAnn AnnWhere (EpaSpan (RealSrcSpan s _)):_) csIn = (hc, fc) + where + splitFunc Before anc_pos c_pos = c_pos < anc_pos + splitFunc After anc_pos c_pos = anc_pos < c_pos + (hc,fc) = break (\(L ll _) -> splitFunc w (ss2pos $ anchor ll) (ss2pos s)) csIn +splitOnWhere _ (AddEpAnn AnnWhere _:_) csIn = (csIn, []) +splitOnWhere f (_:as) csIn = splitOnWhere f as csIn + +balanceFirstLocatedAComments :: [LocatedA a] -> ([LocatedA a], [LEpaComment]) +balanceFirstLocatedAComments [] = ([],[]) +balanceFirstLocatedAComments ((L (EpAnn anc an csd) a):ds) = (L (EpAnn anc an csd0) a:ds, hc') + where + (csd0, hc') = case anc of + EpaSpan (RealSrcSpan s _) -> (csd', hc) + `debug` ("balanceFirstLocatedAComments: (csd,csd',attached,header)=" ++ showAst (csd,csd',attached,header)) + where + (priors, inners) = break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos s) ) + (priorComments csd) + pcds = priorCommentsDeltas' s priors + (attached, header) = break (\(d,_c) -> d /= 1) pcds + csd' = setPriorComments csd (reverse (map snd attached) ++ inners) + hc = reverse (map snd header) + _ -> (csd, []) + + + +priorCommentsDeltas' :: RealSrcSpan -> [LEpaComment] + -> [(Int, LEpaComment)] +priorCommentsDeltas' r cs = go r (reverse cs) + where + go :: RealSrcSpan -> [LEpaComment] -> [(Int, LEpaComment)] + go _ [] = [] + go _ (la@(L l@(EpaDelta _ dp _) _):las) = (deltaLine dp, la) : go (anchor l) las + go rs' (la@(L l _):las) = deltaComment rs' la : go (anchor l) las + + deltaComment :: RealSrcSpan -> LEpaComment -> (Int, LEpaComment) + deltaComment rs' (L loc c) = (abs(ll - al), L loc c) + where + (al,_) = ss2pos rs' + (ll,_) = ss2pos (anchor loc) + +allocatePriorComments + :: Pos + -> [LEpaComment] + -> ([LEpaComment], [LEpaComment]) +allocatePriorComments ss_loc comment_q = + let + cmp (L l _) = ss2pos (anchor l) <= ss_loc + (newAnns,after) = partition cmp comment_q + in + (after, newAnns) + +insertRemainingCppComments :: ParsedSource -> [LEpaComment] -> ParsedSource +insertRemainingCppComments (L l p) cs = L l p' + -- `debug` ("insertRemainingCppComments: (cs,an')=" ++ showAst (cs,an')) + where + (EpAnn a an ocs) = GHC.hsmodAnn $ GHC.hsmodExt p + an' = EpAnn a an (addTrailingComments end_loc ocs cs) p' = p { GHC.hsmodExt = (GHC.hsmodExt p) { GHC.hsmodAnn = an' } } + end_loc = case GHC.hsmodLayout $ GHC.hsmodExt p of + EpExplicitBraces _open close -> case close of + EpTok (EpaSpan (RealSrcSpan s _)) -> ss2pos s + _ -> (1,1) + _ -> (1,1) + (new_before, new_after) = break (\(L ll _) -> (ss2pos $ anchor ll) > end_loc ) cs + + addTrailingComments end_loc' cur new = epaCommentsBalanced pc' fc' + where + pc = priorComments cur + fc = getFollowingComments cur + (pc', fc') = case reverse pc of + [] -> (sortEpaComments $ pc ++ new_before, sortEpaComments $ fc ++ new_after) + (L ac _:_) -> (sortEpaComments $ pc ++ cs_before, sortEpaComments $ fc ++ cs_after) + where + (cs_before,cs_after) + = if (ss2pos $ anchor ac) > end_loc' + then break (\(L ll _) -> (ss2pos $ anchor ll) > (ss2pos $ anchor ac) ) new + else (new_before, new_after) -- --------------------------------------------------------------------- @@ -291,11 +486,16 @@ dedentDocChunkBy dedent (L (RealSrcSpan l mb) c) = L (RealSrcSpan l' mb) c dedentDocChunkBy _ x = x + +epaCommentsBalanced :: [LEpaComment] -> [LEpaComment] -> EpAnnComments +epaCommentsBalanced priorCs [] = EpaComments priorCs +epaCommentsBalanced priorCs postCs = EpaCommentsBalanced priorCs postCs + mkEpaComments :: [Comment] -> [Comment] -> EpAnnComments mkEpaComments priorCs [] = EpaComments (map comment2LEpaComment priorCs) mkEpaComments priorCs postCs - = EpaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) + = epaCommentsBalanced (map comment2LEpaComment priorCs) (map comment2LEpaComment postCs) comment2LEpaComment :: Comment -> LEpaComment comment2LEpaComment (Comment s anc r _mk) = mkLEpaComment s anc r @@ -330,18 +530,11 @@ sortEpaComments cs = sortBy cmp cs mkKWComment :: AnnKeywordId -> NoCommentsLocation -> Comment mkKWComment kw (EpaSpan (RealSrcSpan ss mb)) = Comment (keywordToString kw) (EpaSpan (RealSrcSpan ss mb)) ss (Just kw) -mkKWComment kw (EpaSpan ss@(UnhelpfulSpan _)) - = Comment (keywordToString kw) (EpaDelta ss (SameLine 0) NoComments) placeholderRealSpan (Just kw) +mkKWComment kw (EpaSpan (UnhelpfulSpan _)) + = Comment (keywordToString kw) (EpaDelta noSrcSpan (SameLine 0) NoComments) placeholderRealSpan (Just kw) mkKWComment kw (EpaDelta ss dp cs) = Comment (keywordToString kw) (EpaDelta ss dp cs) placeholderRealSpan (Just kw) --- | Detects a comment which originates from a specific keyword. -isKWComment :: Comment -> Bool -isKWComment c = isJust (commentOrigin c) - -noKWComments :: [Comment] -> [Comment] -noKWComments = filter (\c -> not (isKWComment c)) - sortAnchorLocated :: [GenLocated EpaLocation a] -> [GenLocated EpaLocation a] sortAnchorLocated = sortBy (compare `on` (anchor . getLoc)) @@ -379,11 +572,6 @@ name2String = showPprUnsafe -- --------------------------------------------------------------------- -locatedAnAnchor :: LocatedAn a t -> RealSrcSpan -locatedAnAnchor (L (EpAnn a _ _) _) = anchor a - --- --------------------------------------------------------------------- - trailingAnnLoc :: TrailingAnn -> EpaLocation trailingAnnLoc (AddSemiAnn ss) = ss trailingAnnLoc (AddCommaAnn ss) = ss @@ -401,46 +589,6 @@ setTrailingAnnLoc (AddDarrowUAnn _) ss = (AddDarrowUAnn ss) addEpAnnLoc :: AddEpAnn -> EpaLocation addEpAnnLoc (AddEpAnn _ l) = l --- --------------------------------------------------------------------- --- Horrible hack for dealing with some things still having a SrcSpan, --- not an Anchor. - -{- -A SrcSpan is defined as - -data SrcSpan = - RealSrcSpan !RealSrcSpan !(Maybe BufSpan) -- See Note [Why Maybe BufPos] - | UnhelpfulSpan !UnhelpfulSpanReason - -data BufSpan = - BufSpan { bufSpanStart, bufSpanEnd :: {-# UNPACK #-} !BufPos } - deriving (Eq, Ord, Show) - -newtype BufPos = BufPos { bufPos :: Int } - - -We use the BufPos to encode a delta, using bufSpanStart for the line, -and bufSpanEnd for the col. - -To be absolutely sure, we make the delta versions use -ve values. - --} - -hackSrcSpanToAnchor :: SrcSpan -> EpaLocation -hackSrcSpanToAnchor (UnhelpfulSpan s) = error $ "hackSrcSpanToAnchor : UnhelpfulSpan:" ++ show s -hackSrcSpanToAnchor ss@(RealSrcSpan r mb) - = case mb of - (Strict.Just (BufSpan (BufPos s) (BufPos e))) -> - if s <= 0 && e <= 0 - then EpaDelta ss (deltaPos (-s) (-e)) [] - `debug` ("hackSrcSpanToAnchor: (r,s,e)=" ++ showAst (r,s,e) ) - else EpaSpan (RealSrcSpan r mb) - _ -> EpaSpan (RealSrcSpan r mb) - -hackAnchorToSrcSpan :: EpaLocation -> SrcSpan -hackAnchorToSrcSpan (EpaSpan s) = s -hackAnchorToSrcSpan _ = error $ "hackAnchorToSrcSpan" - -- --------------------------------------------------------------------- type DeclsByTag a = Map.Map DeclTag [(RealSrcSpan, a)] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/dda9c763e92ae2c27a103b226e501d2f708e902b...b49daf5e49670ad097aa3a17caacac8f52355ad9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/dda9c763e92ae2c27a103b226e501d2f708e902b...b49daf5e49670ad097aa3a17caacac8f52355ad9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 07:50:31 2024 From: gitlab at gitlab.haskell.org (Adriaan Leijnse (@aidylns)) Date: Fri, 13 Sep 2024 03:50:31 -0400 Subject: [Git][ghc/ghc][wip/aidylns/ttg-remove-hsunboundvar-via-hshole] 3 commits: finder: Add `IsBootInterface` to finder cache keys Message-ID: <66e3eec747ece_30e1055c834817614@gitlab.mail> Adriaan Leijnse pushed to branch wip/aidylns/ttg-remove-hsunboundvar-via-hshole at Glasgow Haskell Compiler / GHC Commits: fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 74939e45 by Adriaan Leijnse at 2024-09-13T09:50:11+02:00 TTG: Remove HsUnboundVar, replace with HsHole, HsUnboundVarRn/Tc This commit removes HsUnboundVar from the Language AST, which was used to parse Holes to. Instead it introduces a new HsHole AST constructor for this purpose. The renaming and type checking phases keep their original HsUnboundVar implementation using HsUnboundVarRn and HsUnboundVarTc constructors (HsHole is turned into HsUnboundVarRn during renaming). Also, the note explaining Holes is rewritten to reflect the current state of the code. - - - - - 30 changed files: - compiler/GHC/Driver/Backpack.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Rename/Module.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/App.hs - compiler/GHC/Tc/Gen/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Types/Constraint.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Types/Origin.hs - compiler/GHC/Tc/Zonk/Type.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Finder/Types.hs - compiler/GHC/Unit/Module/Env.hs - compiler/GHC/Unit/Types.hs - compiler/Language/Haskell/Syntax/Expr.hs - compiler/Language/Haskell/Syntax/Extension.hs - + testsuite/tests/driver/boot-target/A.hs - + testsuite/tests/driver/boot-target/A.hs-boot The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/144251c2e74c7d2fc333dcd9b7a9fdea5e299e50...74939e45e91c8e769821eaf1464969eb7039c48b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/144251c2e74c7d2fc333dcd9b7a9fdea5e299e50...74939e45e91c8e769821eaf1464969eb7039c48b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 11:52:36 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 13 Sep 2024 07:52:36 -0400 Subject: [Git][ghc/ghc][master] DmdAnal: Fast path for `multDmdType` (#25196) Message-ID: <66e4278413234_3de40473a154102169@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 5 changed files: - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Types/Demand.hs - + testsuite/tests/dmdanal/should_compile/T25196.hs - + testsuite/tests/dmdanal/should_compile/T25196_aux.hs - testsuite/tests/dmdanal/should_compile/all.T Changes: ===================================== compiler/GHC/Core/Opt/DmdAnal.hs ===================================== @@ -426,7 +426,7 @@ dmdAnalStar env (n :* sd) e , n' <- anticipateANF e n -- See Note [Anticipating ANF in demand analysis] -- and Note [Analysing with absent demand] - = (discardArgDmds $ multDmdType n' dmd_ty, e') + = (multDmdEnv n' (discardArgDmds dmd_ty), e') -- Main Demand Analysis machinery dmdAnal, dmdAnal' :: AnalEnv ===================================== compiler/GHC/Types/Demand.hs ===================================== @@ -51,7 +51,7 @@ module GHC.Types.Demand ( -- * Demand environments DmdEnv(..), addVarDmdEnv, mkTermDmdEnv, nopDmdEnv, plusDmdEnv, plusDmdEnvs, - reuseEnv, + multDmdEnv, reuseEnv, -- * Demand types DmdType(..), dmdTypeDepth, @@ -1910,7 +1910,8 @@ splitDmdTy ty at DmdType{dt_args=dmd:args} = (dmd, ty{dt_args=args}) splitDmdTy ty at DmdType{dt_env=env} = (defaultArgDmd (de_div env), ty) multDmdType :: Card -> DmdType -> DmdType -multDmdType n (DmdType fv args) +multDmdType C_11 dmd_ty = dmd_ty -- a vital optimisation for T25196 +multDmdType n (DmdType fv args) = -- pprTrace "multDmdType" (ppr n $$ ppr fv $$ ppr (multDmdEnv n fv)) $ DmdType (multDmdEnv n fv) (strictMap (multDmd n) args) ===================================== testsuite/tests/dmdanal/should_compile/T25196.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25196 where + +import T25196_aux + +bar = $(gen 10000) ===================================== testsuite/tests/dmdanal/should_compile/T25196_aux.hs ===================================== @@ -0,0 +1,10 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25196_aux where + +import Language.Haskell.TH +import Language.Haskell.TH.Syntax +import Language.Haskell.TH.Lib + +gen :: Int -> Q Exp +gen n = lamE (replicate n (newName "x" >>= varP)) [| () |] ===================================== testsuite/tests/dmdanal/should_compile/all.T ===================================== @@ -98,3 +98,5 @@ test('T22388', [ grep_errmsg(r'^\S+\$w\S+') ], compile, ['-dsuppress-uniques -dd test('T22997', normal, compile, ['']) test('T23398', normal, compile, ['-dsuppress-uniques -ddump-simpl -dno-typeable-binds']) test('T24623', normal, compile, ['']) +test('T25196', [ req_th, collect_compiler_stats('bytes allocated', 10) ], + multimod_compile, ['T25196', '-v0']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1374349ba392c6354214dc2a1ceddbdcf4d65e53 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1374349ba392c6354214dc2a1ceddbdcf4d65e53 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 11:53:07 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 13 Sep 2024 07:53:07 -0400 Subject: [Git][ghc/ghc][master] Bump submodule array to 0.5.8.0 Message-ID: <66e427a39a4ca_3de40473a15410521c@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 1 changed file: - libraries/array Changes: ===================================== libraries/array ===================================== @@ -1 +1 @@ -Subproject commit ba5e9dcf1370190239395b8361b1c92ea9fc7632 +Subproject commit c9cb2c1e8762aa83b6e77af82c87a55e03e990e4 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/80769bc9f56541601796366485283a697c52a18b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/80769bc9f56541601796366485283a697c52a18b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 14:49:00 2024 From: gitlab at gitlab.haskell.org (Adriaan Leijnse (@aidylns)) Date: Fri, 13 Sep 2024 10:49:00 -0400 Subject: [Git][ghc/ghc][wip/aidylns/ttg-remove-hsunboundvar-via-hshole] TTG: Remove HsUnboundVar, replace with HsHole, HsUnboundVarRn/Tc Message-ID: <66e450dcf0a9a_1d9c0713c770343ec@gitlab.mail> Adriaan Leijnse pushed to branch wip/aidylns/ttg-remove-hsunboundvar-via-hshole at Glasgow Haskell Compiler / GHC Commits: 7ed954d9 by Adriaan Leijnse at 2024-09-13T16:48:38+02:00 TTG: Remove HsUnboundVar, replace with HsHole, HsUnboundVarRn/Tc This commit removes HsUnboundVar from the Language AST, which was used to parse Holes to. Instead it introduces a new HsHole AST constructor for this purpose. The renaming and type checking phases keep their original HsUnboundVar implementation using HsUnboundVarRn and HsUnboundVarTc constructors (HsHole is turned into HsUnboundVarRn during renaming). Also, the note explaining Holes is rewritten to reflect the current state of the code. - - - - - 25 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Rename/Module.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/App.hs - compiler/GHC/Tc/Gen/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Types/Constraint.hs - compiler/GHC/Tc/Types/Evidence.hs - compiler/GHC/Tc/Types/Origin.hs - compiler/GHC/Tc/Zonk/Type.hs - compiler/Language/Haskell/Syntax/Expr.hs - compiler/Language/Haskell/Syntax/Extension.hs - testsuite/tests/perf/compiler/hard_hole_fits.hs - testsuite/tests/perf/compiler/hard_hole_fits.stderr - testsuite/tests/plugins/T20803-plugin/FixErrorsPlugin.hs - utils/check-exact/ExactPrint.hs Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -208,9 +208,10 @@ data EpAnnHsCase = EpAnnHsCase instance NoAnn EpAnnHsCase where noAnn = EpAnnHsCase noAnn noAnn noAnn -data EpAnnUnboundVar = EpAnnUnboundVar - { hsUnboundBackquotes :: (EpaLocation, EpaLocation) - , hsUnboundHole :: EpaLocation +-- | ExactPrint annotation for holes. +data EpAnnHole = EpAnnHole + { hsHoleBackquotes :: (EpaLocation, EpaLocation) + , hsHoleHole :: EpaLocation } deriving Data -- Record selectors at parse time are HsVar; they convert to HsRecSel @@ -229,14 +230,6 @@ type instance XOverLabel GhcTc = DataConCantHappen type instance XVar (GhcPass _) = NoExtField -type instance XUnboundVar GhcPs = Maybe EpAnnUnboundVar -type instance XUnboundVar GhcRn = NoExtField -type instance XUnboundVar GhcTc = HoleExprRef - -- We really don't need the whole HoleExprRef; just the IORef EvTerm - -- would be enough. But then deriving a Data instance becomes impossible. - -- Much, much easier just to define HoleExprRef with a Data instance and - -- store the whole structure. - type instance XIPVar GhcPs = NoExtField type instance XIPVar GhcRn = NoExtField type instance XIPVar GhcTc = DataConCantHappen @@ -382,6 +375,11 @@ type instance XEmbTy GhcTc = DataConCantHappen -- A free-standing HsEmbTy is an error. -- Valid usages are immediately desugared into Type. +type instance XHole GhcPs = Maybe EpAnnHole +-- After renaming holes are handled using HsUnboundVarRn/HsUnboundVarTc: +type instance XHole GhcRn = DataConCantHappen +type instance XHole GhcTc = DataConCantHappen + type instance XForAll GhcPs = NoExtField type instance XForAll GhcRn = NoExtField type instance XForAll GhcTc = DataConCantHappen @@ -514,6 +512,10 @@ data XXExprGhcRn -- Does not presist post renaming phase -- See Part 3. of Note [Expanding HsDo with XXExprGhcRn] -- in `GHC.Tc.Gen.Do` + | HsUnboundVarRn RdrName -- ^ Unbound variable. + -- Turned from HsVar to HsUnboundVarRn by the + -- renamer, when it finds an out-of-scope + -- variable. -- | Wrap a located expression with a `PopErrCtxt` @@ -594,6 +596,19 @@ data XXExprGhcTc Int -- module-local tick number for True Int -- module-local tick number for False (LHsExpr GhcTc) -- sub-expression + | HsUnboundVarTc HoleExprRef RdrName + -- ^ Unbound variable; also used for "holes" + -- (_ or _x). + -- In the case of expression holes, the HoleExprRef is where the + -- erroring expression will be written after solving. + -- See Note [Holes in expressions] in GHC.Tc.Types.Constraint. + -- + -- We really don't need the whole HoleExprRef; just + -- the IORef EvTerm would be enough. But then + -- deriving a Data instance becomes impossible. + -- Much, much easier just to define HoleExprRef + -- with a Data instance and store the whole + -- structure. -- | Build a 'XXExprGhcRn' out of an extension constructor, -- and the two components of the expansion: original and @@ -654,7 +669,10 @@ ppr_lexpr e = ppr_expr (unLoc e) ppr_expr :: forall p. (OutputableBndrId p) => HsExpr (GhcPass p) -> SDoc ppr_expr (HsVar _ (L _ v)) = pprPrefixOcc v -ppr_expr (HsUnboundVar _ uv) = pprPrefixOcc uv +ppr_expr (HsHole x) = case ghcPass @p of + GhcPs -> pprPrefixOcc (mkUnqual varName (fsLit "_")) + GhcRn -> dataConCantHappen x + GhcTc -> dataConCantHappen x ppr_expr (HsRecSel _ f) = pprPrefixOcc f ppr_expr (HsIPVar _ v) = ppr v ppr_expr (HsOverLabel s l) = case ghcPass @p of @@ -879,6 +897,7 @@ instance Outputable HsThingRn where instance Outputable XXExprGhcRn where ppr (ExpandedThingRn o e) = ifPprDebug (braces $ vcat [ppr o, ppr e]) (ppr o) ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) + ppr (HsUnboundVarRn uv) = pprPrefixOcc uv instance Outputable XXExprGhcTc where ppr (WrapExpr (HsWrap co_fn e)) @@ -907,11 +926,15 @@ instance Outputable XXExprGhcTc where ppr tickIdFalse, text ">(", ppr exp, text ")"] + ppr (HsUnboundVarTc _ uv) = pprPrefixOcc uv ppr_infix_expr :: forall p. (OutputableBndrId p) => HsExpr (GhcPass p) -> Maybe SDoc ppr_infix_expr (HsVar _ (L _ v)) = Just (pprInfixOcc v) ppr_infix_expr (HsRecSel _ f) = Just (pprInfixOcc f) -ppr_infix_expr (HsUnboundVar _ occ) = Just (pprInfixOcc occ) +ppr_infix_expr (HsHole x) = case ghcPass @p of + GhcPs -> Just (pprInfixOcc (mkUnqual varName (fsLit "_"))) + GhcRn -> dataConCantHappen x + GhcTc -> dataConCantHappen x ppr_infix_expr (XExpr x) = case ghcPass @p of GhcRn -> ppr_infix_expr_rn x GhcTc -> ppr_infix_expr_tc x @@ -920,6 +943,7 @@ ppr_infix_expr _ = Nothing ppr_infix_expr_rn :: XXExprGhcRn -> Maybe SDoc ppr_infix_expr_rn (ExpandedThingRn thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a +ppr_infix_expr_rn (HsUnboundVarRn occ) = Just (pprInfixOcc occ) ppr_infix_expr_tc :: XXExprGhcTc -> Maybe SDoc ppr_infix_expr_tc (WrapExpr (HsWrap _ e)) = ppr_infix_expr e @@ -927,6 +951,7 @@ ppr_infix_expr_tc (ExpandedThingTc thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_tc (ConLikeTc {}) = Nothing ppr_infix_expr_tc (HsTick {}) = Nothing ppr_infix_expr_tc (HsBinTick {}) = Nothing +ppr_infix_expr_tc (HsUnboundVarTc _ occ) = Just (pprInfixOcc occ) ppr_infix_hs_expansion :: HsThingRn -> Maybe SDoc ppr_infix_hs_expansion (OrigExpr e) = ppr_infix_expr e @@ -974,7 +999,6 @@ hsExprNeedsParens prec = go where go :: HsExpr (GhcPass p) -> Bool go (HsVar{}) = False - go (HsUnboundVar{}) = False go (HsIPVar{}) = False go (HsOverLabel{}) = False go (HsLit _ l) = hsLitNeedsParens prec l @@ -1017,6 +1041,7 @@ hsExprNeedsParens prec = go go (HsProjection{}) = True go (HsGetField{}) = False go (HsEmbTy{}) = prec > topPrec + go (HsHole{}) = False go (HsForAll{}) = prec >= funPrec go (HsQual{}) = prec >= funPrec go (HsFunArr{}) = prec >= funPrec @@ -1030,10 +1055,12 @@ hsExprNeedsParens prec = go go_x_tc (ConLikeTc {}) = False go_x_tc (HsTick _ (L _ e)) = hsExprNeedsParens prec e go_x_tc (HsBinTick _ _ (L _ e)) = hsExprNeedsParens prec e + go_x_tc (HsUnboundVarTc {}) = False go_x_rn :: XXExprGhcRn -> Bool go_x_rn (ExpandedThingRn thing _) = hsExpandedNeedsParens thing go_x_rn (PopErrCtxt (L _ a)) = hsExprNeedsParens prec a + go_x_rn (HsUnboundVarRn _) = False hsExpandedNeedsParens :: HsThingRn -> Bool hsExpandedNeedsParens (OrigExpr e) = hsExprNeedsParens prec e @@ -1070,8 +1097,8 @@ isAtomicHsExpr (HsLit {}) = True isAtomicHsExpr (HsOverLit {}) = True isAtomicHsExpr (HsIPVar {}) = True isAtomicHsExpr (HsOverLabel {}) = True -isAtomicHsExpr (HsUnboundVar {}) = True isAtomicHsExpr (HsRecSel{}) = True +isAtomicHsExpr (HsHole{}) = True isAtomicHsExpr (XExpr x) | GhcTc <- ghcPass @p = go_x_tc x | GhcRn <- ghcPass @p = go_x_rn x @@ -1082,10 +1109,12 @@ isAtomicHsExpr (XExpr x) go_x_tc (ConLikeTc {}) = True go_x_tc (HsTick {}) = False go_x_tc (HsBinTick {}) = False + go_x_tc (HsUnboundVarTc {}) = True go_x_rn :: XXExprGhcRn -> Bool go_x_rn (ExpandedThingRn thing _) = isAtomicExpandedThingRn thing go_x_rn (PopErrCtxt (L _ a)) = isAtomicHsExpr a + go_x_rn (HsUnboundVarRn {}) = True isAtomicExpandedThingRn :: HsThingRn -> Bool isAtomicExpandedThingRn (OrigExpr e) = isAtomicHsExpr e ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -102,7 +102,6 @@ lhsExprType (L _ e) = hsExprType e -- | Compute the 'Type' of an @'HsExpr' 'GhcTc'@ in a pure fashion. hsExprType :: HsExpr GhcTc -> Type hsExprType (HsVar _ (L _ id)) = idType id -hsExprType (HsUnboundVar (HER _ ty _) _) = ty hsExprType (HsRecSel _ (FieldOcc id _)) = idType id hsExprType (HsOverLabel v _) = dataConCantHappen v hsExprType (HsIPVar v _) = dataConCantHappen v @@ -146,6 +145,7 @@ hsExprType (HsProc _ _ lcmd_top) = lhsCmdTopType lcmd_top hsExprType (HsStatic (_, ty) _s) = ty hsExprType (HsPragE _ _ e) = lhsExprType e hsExprType (HsEmbTy x _) = dataConCantHappen x +hsExprType (HsHole x) = dataConCantHappen x hsExprType (HsQual x _ _) = dataConCantHappen x hsExprType (HsForAll x _ _) = dataConCantHappen x hsExprType (HsFunArr x _ _ _) = dataConCantHappen x @@ -154,6 +154,7 @@ hsExprType (XExpr (ExpandedThingTc _ e)) = hsExprType e hsExprType (XExpr (ConLikeTc con _ _)) = conLikeType con hsExprType (XExpr (HsTick _ e)) = lhsExprType e hsExprType (XExpr (HsBinTick _ _ e)) = lhsExprType e +hsExprType (XExpr (HsUnboundVarTc (HER _ ty _) _)) = ty arithSeqInfoType :: ArithSeqInfo GhcTc -> Type arithSeqInfoType asi = mkListTy $ case asi of ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -292,8 +292,8 @@ dsExpr (HsRecSel _ (FieldOcc id _)) return $ take maxConstructors cons_wo_field -dsExpr (HsUnboundVar (HER ref _ _) _) = dsEvTerm =<< readMutVar ref - -- See Note [Holes] in GHC.Tc.Types.Constraint +dsExpr (HsHole x) = dataConCantHappen x + -- Holes are handled by HsUnboundVarTc. dsExpr (HsPar _ e) = dsLExpr e dsExpr (ExprWithTySig _ e _) = dsLExpr e @@ -336,6 +336,9 @@ dsExpr e@(XExpr ext_expr_tc) do { assert (exprType e2 `eqType` boolTy) mkBinaryTickBox ixT ixF e2 } + HsUnboundVarTc (HER ref _ _) _ -> dsEvTerm =<< readMutVar ref + -- See Note [Holes in expressions] in GHC.Tc.Types.Constraint. + -- Strip ticks due to #21701, need to be invariant about warnings we produce whether -- this is enabled or not. ===================================== compiler/GHC/HsToCore/Quote.hs ===================================== @@ -1674,9 +1674,6 @@ repE (HsTypedSplice n _) = rep_splice n repE (HsUntypedSplice (HsUntypedSpliceNested n) _) = rep_splice n repE e@(HsUntypedSplice (HsUntypedSpliceTop _ _) _) = pprPanic "repE: top level splice" (ppr e) repE (HsStatic _ e) = repLE e >>= rep2 staticEName . (:[]) . unC -repE (HsUnboundVar _ uv) = do - name <- repRdrName uv - repUnboundVar name repE (HsGetField _ e (L _ (DotFieldOcc _ (L _ (FieldLabelString f))))) = do e1 <- repLE e repGetField e1 f @@ -1713,10 +1710,14 @@ repE e@(XExpr (ExpandedThingRn o x)) = notHandled (ThExpressionForm e) repE (XExpr (PopErrCtxt (L _ e))) = repE e +repE (XExpr (HsUnboundVarRn uv)) = do + name <- repRdrName uv + repUnboundVar name repE e@(HsPragE _ (HsPragSCC {}) _) = notHandled (ThCostCentres e) repE e@(HsTypedBracket{}) = notHandled (ThExpressionForm e) repE e@(HsUntypedBracket{}) = notHandled (ThExpressionForm e) repE e@(HsProc{}) = notHandled (ThExpressionForm e) +repE e@(HsHole{}) = notHandled (ThExpressionForm e) repFunArr :: HsArrowOf (LocatedA (HsExpr GhcRn)) GhcRn -> MetaM (Core (M TH.Exp)) repFunArr HsUnrestrictedArrow{} = repConName unrestrictedFunTyConName ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -475,7 +475,6 @@ addBinTickLHsExpr boxLabel e@(L pos e0) addTickHsExpr :: HsExpr GhcTc -> TM (HsExpr GhcTc) addTickHsExpr e@(HsVar _ (L _ id)) = do freeVar id; return e -addTickHsExpr e@(HsUnboundVar {}) = return e addTickHsExpr e@(HsRecSel _ (FieldOcc id _)) = do freeVar id; return e addTickHsExpr e@(HsIPVar {}) = return e @@ -483,6 +482,7 @@ addTickHsExpr e@(HsOverLit {}) = return e addTickHsExpr e@(HsOverLabel{}) = return e addTickHsExpr e@(HsLit {}) = return e addTickHsExpr e@(HsEmbTy {}) = return e +addTickHsExpr e@(HsHole {}) = return e addTickHsExpr e@(HsQual {}) = return e addTickHsExpr e@(HsForAll {}) = return e addTickHsExpr e@(HsFunArr {}) = return e @@ -603,6 +603,7 @@ addTickHsExpr (XExpr (HsTick t e)) = liftM (XExpr . HsTick t) (addTickLHsExprNever e) addTickHsExpr (XExpr (HsBinTick t0 t1 e)) = liftM (XExpr . HsBinTick t0 t1) (addTickLHsExprNever e) +addTickHsExpr e@(XExpr (HsUnboundVarTc {})) = return e addTickHsExpr (HsDo srcloc cxt (L l stmts)) = do { (stmts', _) <- addTickLStmts' forQual stmts (return ()) ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1176,7 +1176,6 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where [ toHie $ C Use (L mspan var) -- Patch up var location since typechecker removes it ] - HsUnboundVar _ _ -> [] -- there is an unbound name here, but that causes trouble HsRecSel _ fld -> [ toHie $ RFC RecFieldOcc Nothing (L mspan fld) ] @@ -1325,6 +1324,7 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where ] HsGetField {} -> [] HsProjection {} -> [] + HsHole _ -> [] -- there is a hole here, but that causes trouble XExpr x | HieTc <- hiePass @p -> case x of @@ -1341,6 +1341,7 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where HsBinTick _ _ expr -> [ toHie expr ] + HsUnboundVarTc _ _ -> [] -- there is an unbound name here, but that causes trouble | otherwise -> [] -- NOTE: no longer have the location ===================================== compiler/GHC/Parser.y ===================================== @@ -3896,7 +3896,7 @@ qopm :: { forall b. DisambInfixOp b => PV (LocatedN b) } -- used in section | hole_op { mkHsInfixHolePV $1 } hole_op :: { LocatedN (HsExpr GhcPs) } -- used in sections -hole_op : '`' '_' '`' { sLLa $1 $> (hsHoleExpr (Just $ EpAnnUnboundVar (glAA $1, glAA $3) (glAA $2))) } +hole_op : '`' '_' '`' { sLLa $1 $> (HsHole (Just $ EpAnnHole (glAA $1, glAA $3) (glAA $2))) } qvarop :: { LocatedN RdrName } : qvarsym { $1 } ===================================== compiler/GHC/Parser/PostProcess.hs ===================================== @@ -109,7 +109,6 @@ module GHC.Parser.PostProcess ( withArrowParsingMode, withArrowParsingMode', setTelescopeBndrsNameSpace, PatBuilder, - hsHoleExpr, -- Type/datacon ambiguity resolution DisambTD(..), @@ -1899,11 +1898,11 @@ instance DisambECP (HsExpr GhcPs) where type Body (HsExpr GhcPs) = HsExpr ecpFromCmd' (L l c) = do addError $ mkPlainErrorMsgEnvelope (locA l) $ PsErrArrowCmdInExpr c - return (L l (hsHoleExpr noAnn)) + return (L l (HsHole noAnn)) ecpFromExp' = return ecpFromPat' p@(L l _) = do addError $ mkPlainErrorMsgEnvelope (locA l) $ PsErrOrPatInExpr p - return (L l (hsHoleExpr noAnn)) + return (L l (HsHole noAnn)) mkHsProjUpdatePV l fields arg isPun anns = do !cs <- getCommentsFor l return $ mkRdrProjUpdate (EpAnn (spanAsAnchor l) noAnn cs) fields arg isPun anns @@ -1952,7 +1951,7 @@ instance DisambECP (HsExpr GhcPs) where mkHsOverLitPV (L (EpAnn l an csIn) a) = do !cs <- getCommentsFor (locA l) return $ L (EpAnn l an (cs Semi.<> csIn)) (HsOverLit NoExtField a) - mkHsWildCardPV l = return $ L (noAnnSrcSpan l) (hsHoleExpr noAnn) + mkHsWildCardPV l = return $ L (noAnnSrcSpan l) (HsHole noAnn) mkHsTySigPV l@(EpAnn anc an csIn) a sig anns = do !cs <- getCommentsFor (locA l) return $ L (EpAnn anc an (csIn Semi.<> cs)) (ExprWithTySig anns a (hsTypeToHsSigWcType sig)) @@ -1973,11 +1972,11 @@ instance DisambECP (HsExpr GhcPs) where !cs <- getCommentsFor l return $ L (EpAnn (spanAsAnchor l) noAnn cs) (SectionR noExtField op e) mkHsAsPatPV l v _ e = addError (mkPlainErrorMsgEnvelope l $ PsErrTypeAppWithoutSpace (unLoc v) e) - >> return (L (noAnnSrcSpan l) (hsHoleExpr noAnn)) + >> return (L (noAnnSrcSpan l) (HsHole noAnn)) mkHsLazyPatPV l e _ = addError (mkPlainErrorMsgEnvelope l $ PsErrLazyPatWithoutSpace e) - >> return (L (noAnnSrcSpan l) (hsHoleExpr noAnn)) + >> return (L (noAnnSrcSpan l) (HsHole noAnn)) mkHsBangPatPV l e _ = addError (mkPlainErrorMsgEnvelope l $ PsErrBangPatWithoutSpace e) - >> return (L (noAnnSrcSpan l) (hsHoleExpr noAnn)) + >> return (L (noAnnSrcSpan l) (HsHole noAnn)) mkSumOrTuplePV = mkSumOrTupleExpr mkHsEmbTyPV l toktype ty = return $ L (noAnnSrcSpan l) $ @@ -2004,9 +2003,6 @@ instance DisambECP (HsExpr GhcPs) where (PsErrUnallowedPragma prag) rejectPragmaPV _ = return () -hsHoleExpr :: Maybe EpAnnUnboundVar -> HsExpr GhcPs -hsHoleExpr anns = HsUnboundVar anns (mkRdrUnqual (mkVarOccFS (fsLit "_"))) - instance DisambECP (PatBuilder GhcPs) where type Body (PatBuilder GhcPs) = PatBuilder ecpFromCmd' (L l c) = addFatalError $ mkPlainErrorMsgEnvelope (locA l) $ PsErrArrowCmdInPat c ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -320,7 +320,7 @@ rnUnboundVar v = do deferOutofScopeVariables <- goptM Opt_DeferOutOfScopeVariables -- See Note [Reporting unbound names] for difference between qualified and unqualified names. unless (isUnqual v || deferOutofScopeVariables) (reportUnboundName v >> return ()) - return (HsUnboundVar noExtField v, emptyFVs) + return (XExpr (HsUnboundVarRn v), emptyFVs) rnExpr (HsVar _ (L l v)) = do { dflags <- getDynFlags @@ -355,8 +355,8 @@ rnExpr (HsVar _ (L l v)) rnExpr (HsIPVar x v) = return (HsIPVar x v, emptyFVs) -rnExpr (HsUnboundVar _ v) - = return (HsUnboundVar noExtField v, emptyFVs) +rnExpr (HsHole _) + = return (XExpr (HsUnboundVarRn (mkUnqual varName (fsLit "_"))), emptyFVs) -- HsOverLabel: see Note [Handling overloaded and rebindable constructs] rnExpr (HsOverLabel src v) @@ -862,8 +862,8 @@ See #18151. Note [Reporting unbound names] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Faced with an out-of-scope `RdrName` there are two courses of action -A. Report an error immediately (and return a HsUnboundVar). This will halt GHC after the renamer is complete -B. Return a HsUnboundVar without reporting an error. That will allow the typechecker to run, which in turn +A. Report an error immediately (and return a HsUnboundVarRn). This will halt GHC after the renamer is complete +B. Return a HsUnboundVarRn without reporting an error. That will allow the typechecker to run, which in turn can give a better error message, notably giving the type of the variable via the "typed holes" mechanism. When `-fdefer-out-of-scope-variables` is on we follow plan B. ===================================== compiler/GHC/Rename/HsType.hs ===================================== @@ -1431,10 +1431,10 @@ data NegationHandling = ReassociateNegation | KeepNegationIntact ---------------------------- get_op :: LHsExpr GhcRn -> OpName --- An unbound name could be either HsVar or HsUnboundVar +-- An unbound name could be either HsVar or (XExpr (HsUnboundVarRn uv)) -- See GHC.Rename.Expr.rnUnboundVar get_op (L _ (HsVar _ n)) = NormalOp (unLoc n) -get_op (L _ (HsUnboundVar _ uv)) = UnboundOp uv +get_op (L _ (XExpr (HsUnboundVarRn uv))) = UnboundOp uv get_op (L _ (HsRecSel _ fld)) = RecFldOp fld get_op other = pprPanic "get_op" (ppr other) ===================================== compiler/GHC/Rename/Module.hs ===================================== @@ -1299,7 +1299,7 @@ badRuleLhsErr name lhs bad_e = TcRnIllegalRuleLhs errReason name lhs bad_e where errReason = case bad_e of - HsUnboundVar _ uv -> + XExpr (HsUnboundVarRn uv) -> UnboundVariable uv $ notInScopeErr WL_Global uv _ -> IllegalExpression ===================================== compiler/GHC/Tc/Errors.hs ===================================== @@ -1550,7 +1550,7 @@ maybeAddDeferredBindings ctxt hole report = do when (deferringAnyBindings ctxt) $ do err_tm <- mkErrorTerm ctxt (hole_loc hole) ref_ty report -- NB: ref_ty, not hole_ty. hole_ty might be rewritten. - -- See Note [Holes] in GHC.Tc.Types.Constraint + -- See Note [Holes in expressions] in GHC.Tc.Types.Constraint writeMutVar ref err_tm _ -> pure () ===================================== compiler/GHC/Tc/Gen/App.hs ===================================== @@ -220,7 +220,7 @@ A "head" has three special cases (for which we can infer a polytype using tcInferAppHead_maybe); otherwise is just any old expression (for which we can infer a rho-type (via tcInfer). -There is no special treatment for HsUnboundVar, HsOverLit etc, because +There is no special treatment for HsUnboundVarTc, HsOverLit etc, because we can't get a polytype from them. Left and right sections (e.g. (x +) and (+ x)) are not yet supported. @@ -674,7 +674,7 @@ tcInstFun do_ql inst_final (tc_fun, fun_ctxt) fun_sigma rn_args fun_is_out_of_scope -- See Note [VTA for out-of-scope functions] = case tc_fun of - HsUnboundVar {} -> True + XExpr (HsUnboundVarTc {}) -> True _ -> False inst_fun :: [HsExprArg 'TcpRn] -> ForAllTyFlag -> Bool @@ -1050,7 +1050,7 @@ expr_to_type earg = = do { t <- go (L l e) ; let splice_result' = HsUntypedSpliceTop finalizers t ; return (L l (HsSpliceTy splice_result' splice)) } - go (L l (HsUnboundVar _ rdr)) + go (L l (XExpr (HsUnboundVarRn rdr))) | isUnderscore occ = return (L l (HsWildCardTy noExtField)) | startsWithUnderscore occ = -- See Note [Wildcards in the T2T translation] @@ -1122,7 +1122,7 @@ This conversion is in the TcM monad because vfun [x | x <- xs] Can't convert list comprehension to a type vfun (\x -> x) Can't convert a lambda to a type * It needs to check for LangExt.NamedWildCards to generate an appropriate - error message for HsUnboundVar. + error message for HsUnboundVarRn. vfun _a Not in scope: ‘_a’ (NamedWildCards disabled) vfun _a Illegal named wildcard in a required type argument: ‘_a’ @@ -1473,7 +1473,7 @@ Note [VTA for out-of-scope functions] Suppose 'wurble' is not in scope, and we have (wurble @Int @Bool True 'x') -Then the renamer will make (HsUnboundVar "wurble") for 'wurble', +Then the renamer will make (HsUnboundVarRn "wurble") for 'wurble', and the typechecker will typecheck it with tcUnboundId, giving it a type 'alpha', and emitting a deferred Hole constraint, to be reported later. @@ -1488,7 +1488,7 @@ tcUnboundId. It later reports 'wurble' as out of scope, and tries to give its type. Fortunately in tcInstFun we still have access to the function, so we -can check if it is a HsUnboundVar. We use this info to simply skip +can check if it is a HsUnboundVarTc. We use this info to simply skip over any visible type arguments. We'll /already/ have emitted a Hole constraint; failing preserves that constraint. ===================================== compiler/GHC/Tc/Gen/Expr.hs ===================================== @@ -304,17 +304,6 @@ tcExpr e@(HsOverLit _ lit) res_ty Just lit' -> return (HsOverLit noExtField lit') Nothing -> tcApp e res_ty } --- Typecheck an occurrence of an unbound Id --- --- Some of these started life as a true expression hole "_". --- Others might simply be variables that accidentally have no binding site -tcExpr (HsUnboundVar _ occ) res_ty - = do { ty <- expTypeToType res_ty -- Allow Int# etc (#12531) - ; her <- emitNewExprHole occ ty - ; tcEmitBindingUsage bottomUE -- Holes fit any usage environment - -- (#18491) - ; return (HsUnboundVar her occ) } - tcExpr e@(HsLit x lit) res_ty = do { let lit_ty = hsLitType lit ; tcWrapResult e (HsLit x (convertLit lit)) lit_ty res_ty } @@ -700,6 +689,7 @@ tcExpr (HsUntypedSplice splice _) res_ty tcExpr (HsOverLabel {}) ty = pprPanic "tcExpr:HsOverLabel" (ppr ty) tcExpr (SectionL {}) ty = pprPanic "tcExpr:SectionL" (ppr ty) tcExpr (SectionR {}) ty = pprPanic "tcExpr:SectionR" (ppr ty) +tcExpr (HsHole {}) ty = pprPanic "tcExpr:HsHole" (ppr ty) {- @@ -739,6 +729,17 @@ tcXExpr xe@(ExpandedThingRn o e') res_ty = setSrcSpanA loc $ mkExpandedStmtTc ls <$> tcApp (XExpr xe) res_ty +-- Typecheck an occurrence of an unbound Id +-- +-- Some of these started life as a true expression hole "_". +-- Others might simply be variables that accidentally have no binding site. +tcXExpr (HsUnboundVarRn occ) res_ty + = do { ty <- expTypeToType res_ty -- Allow Int# etc (#12531) + ; her <- emitNewExprHole occ ty + ; tcEmitBindingUsage bottomUE -- Holes fit any usage environment + -- (#18491) + ; return (XExpr (HsUnboundVarTc her occ)) } + tcXExpr xe res_ty = tcApp (XExpr xe) res_ty {- ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -1281,9 +1281,9 @@ addStmtCtxt stmt thing_inside addExprCtxt :: HsExpr GhcRn -> TcRn a -> TcRn a addExprCtxt e thing_inside = case e of - HsUnboundVar {} -> thing_inside + XExpr (HsUnboundVarRn {}) -> thing_inside _ -> addErrCtxt (exprCtxt e) thing_inside - -- The HsUnboundVar special case addresses situations like + -- The HsUnboundVarRn special case addresses situations like -- f x = _ -- when we don't want to say "In the expression: _", -- because it is mentioned in the error message itself ===================================== compiler/GHC/Tc/Types/Constraint.hs ===================================== @@ -370,7 +370,7 @@ data DelayedError = DE_Hole Hole -- ^ A hole (in a type or in a term). -- - -- See Note [Holes]. + -- See Note [Holes in expressions]. | DE_NotConcrete NotConcreteError -- ^ A type could not be ensured to be concrete. -- @@ -383,7 +383,7 @@ instance Outputable DelayedError where -- | A hole stores the information needed to report diagnostics -- about holes in terms (unbound identifiers or underscores) or -- in types (also called wildcards, as used in partial type --- signatures). See Note [Holes]. +-- signatures). See Note [Holes in expressions] for holes in terms. data Hole = Hole { hole_sort :: HoleSort -- ^ What flavour of hole is this? , hole_occ :: RdrName -- ^ The name of this hole @@ -673,43 +673,78 @@ of the rhs. This is necessary because these constraints are used for substitutio during solving. If the kinds differed, then the substitution would take a well-kinded type to an ill-kinded one. -Note [Holes] -~~~~~~~~~~~~ -This Note explains how GHC tracks *holes*. - -A hole represents one of two conditions: - - A missing bit of an expression. Example: foo x = x + _ - - A missing bit of a type. Example: bar :: Int -> _ - -What these have in common is that both cause GHC to emit a diagnostic to the -user describing the bit that is left out. - -When a hole is encountered, a new entry of type Hole is added to the ambient -WantedConstraints. The type (hole_ty) of the hole is then simplified during -solving (with respect to any Givens in surrounding implications). It is -reported with all the other errors in GHC.Tc.Errors. - -For expression holes, the user has the option of deferring errors until runtime -with -fdefer-type-errors. In this case, the hole actually has evidence: this -evidence is an erroring expression that prints an error and crashes at runtime. -The ExprHole variant of holes stores an IORef EvTerm that will contain this evidence; -during constraint generation, this IORef was stored in the HsUnboundVar extension -field by the type checker. The desugarer simply dereferences to get the CoreExpr. - -Prior to fixing #17812, we used to invent an Id to hold the erroring -expression, and then bind it during type-checking. But this does not support -representation-polymorphic out-of-scope identifiers. See -typecheck/should_compile/T17812. We thus use the mutable-CoreExpr approach -described above. - -You might think that the type in the HoleExprRef is the same as the type of the -hole. However, because the hole type (hole_ty) is rewritten with respect to -givens, this might not be the case. That is, the hole_ty is always (~) to the -type of the HoleExprRef, but they might not be `eqType`. We need the type of the generated -evidence to match what is expected in the context of the hole, and so we must -store these types separately. - -Type-level holes have no evidence at all. +Note [Holes in expressions] +~~~~~~~~~~~~~~~~~~~~~~~~~~~ +This Note explains how GHC tracks "holes" in expressions. It does not +deal with holes in types, nor with partial type signatures. + +A hole represents a missing bit of an expression. Example: + foo x = x && _ +GHC then emits a diagnostic, describing the bit that is left out: + Foo.hs:5:14: error: [GHC-88464] + • Found hole: _ :: Bool + • In the second argument of ‘(&&)’, namely ‘_’ + In the expression: x && _ + +GHC uses the same mechanism is used to give diagnostics for out-of-scope +variables: + foo x = x && y +gives diagnostic + Foo.hs:5:14: error: [GHC-88464] + Variable not in scope: y :: Bool + +Here is how holes are represented in expressions: + +* If the user wrote "_": + Parser HsHole + Renamer XExpr (HsUnboundVarRn "_") + Typechecker XExpr (HsUnboundVarTc ref "_") + +* If the user wrote "x", where `x` is not in scope + Parser HsVar "x" + Renamer XExpr (HsUnboundVarRn "x") + Typechecker XExpr (HsUnboundVarTc ref "x") + +In both cases (ref::HoleExprRef) contains + - The type of the hole. + - A ref-cell that is filled in (by the typechecker) with an + error thunk. With -fdefer-type errors we use this as the + value of the hole. + - A Unique (see Note [Uniques and tags]). + +Typechecking holes + +* When the typechecker encounters a `HsUnboundVarRn`, it returns + a `HsUnboundVarTc`, but also emits a `DelayedError` into the `WantedConstraints`. + +* This DelayedError later triggers the error reporting, and the filling-in of + the error thunk, in GHC.Tc.Errors. + +* The user has the option of deferring errors until runtime with + `-fdefer-type-errors`. In this case, the hole carries evidence in its + `HoleExprRef`. This evidence is an erroring expression that prints an error + and crashes at runtime. + +Desugaring holes + +* During desugaring, the `(HsExpr (HsUnboundVarTc ref "x"))` is desugared by + reading the ref-cell to find the error thunk evidence term, put there by the + constraint solver. + +Wrinkles: + +* Prior to fixing #17812, we used to invent an Id to hold the erroring + expression, and then bind it during type-checking. But this does not support + representation-polymorphic out-of-scope identifiers. See + typecheck/should_compile/T17812. We thus use the mutable-CoreExpr approach + described above. + +* You might think that the type in the HoleExprRef is the same as the type of the + hole. However, because the hole type (hole_ty) is rewritten with respect to + givens, this might not be the case. That is, the hole_ty is always (~) to the + type of the HoleExprRef, but they might not be `eqType`. We need the type of the generated + evidence to match what is expected in the context of the hole, and so we must + store these types separately. -} mkNonCanonical :: CtEvidence -> Ct ===================================== compiler/GHC/Tc/Types/Evidence.hs ===================================== @@ -594,7 +594,7 @@ data EvCallStack -} -- | Where to store evidence for expression holes --- See Note [Holes] in GHC.Tc.Types.Constraint +-- See Note [Holes in expressions] in GHC.Tc.Types.Constraint data HoleExprRef = HER (IORef EvTerm) -- ^ where to write the erroring expression TcType -- ^ expected type of that expression Unique -- ^ for debug output only ===================================== compiler/GHC/Tc/Types/Origin.hs ===================================== @@ -717,7 +717,6 @@ lexprCtOrigin (L _ e) = exprCtOrigin e exprCtOrigin :: HsExpr GhcRn -> CtOrigin exprCtOrigin (HsVar _ (L _ name)) = OccurrenceOf name exprCtOrigin (HsGetField _ _ (L _ f)) = HasFieldOrigin (field_label $ unLoc $ dfoLabel f) -exprCtOrigin (HsUnboundVar {}) = Shouldn'tHappenOrigin "unbound variable" exprCtOrigin (HsRecSel _ f) = OccurrenceOfRecSel (unLoc $ foLabel f) exprCtOrigin (HsOverLabel _ l) = OverLabelOrigin l exprCtOrigin (ExplicitList {}) = ListOrigin @@ -752,12 +751,14 @@ exprCtOrigin (HsUntypedSplice {}) = Shouldn'tHappenOrigin "TH untyped splice" exprCtOrigin (HsProc {}) = Shouldn'tHappenOrigin "proc" exprCtOrigin (HsStatic {}) = Shouldn'tHappenOrigin "static expression" exprCtOrigin (HsEmbTy {}) = Shouldn'tHappenOrigin "type expression" +exprCtOrigin (HsHole {}) = Shouldn'tHappenOrigin "hole expression" exprCtOrigin (HsForAll {}) = Shouldn'tHappenOrigin "forall telescope" -- See Note [Types in terms] exprCtOrigin (HsQual {}) = Shouldn'tHappenOrigin "constraint context" -- See Note [Types in terms] exprCtOrigin (HsFunArr {}) = Shouldn'tHappenOrigin "function arrow" -- See Note [Types in terms] exprCtOrigin (XExpr (ExpandedThingRn thing _)) | OrigExpr a <- thing = exprCtOrigin a | OrigStmt _ <- thing = DoOrigin | OrigPat p <- thing = DoPatOrigin p +exprCtOrigin (XExpr (HsUnboundVarRn {})) = Shouldn'tHappenOrigin "unbound variable" exprCtOrigin (XExpr (PopErrCtxt {})) = Shouldn'tHappenOrigin "PopErrCtxt" -- | Extract a suitable CtOrigin from a MatchGroup ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -925,9 +925,9 @@ zonkExpr (HsVar x (L l id)) do { id' <- zonkIdOcc id ; return (HsVar x (L l id')) } -zonkExpr (HsUnboundVar her occ) +zonkExpr (XExpr (HsUnboundVarTc her occ)) = do her' <- zonk_her her - return (HsUnboundVar her' occ) + return (XExpr (HsUnboundVarTc her' occ)) where zonk_her :: HoleExprRef -> ZonkTcM HoleExprRef zonk_her (HER ref ty u) @@ -1079,6 +1079,7 @@ zonkExpr (HsEmbTy x _) = dataConCantHappen x zonkExpr (HsQual x _ _) = dataConCantHappen x zonkExpr (HsForAll x _ _) = dataConCantHappen x zonkExpr (HsFunArr x _ _ _) = dataConCantHappen x +zonkExpr (HsHole x) = dataConCantHappen x zonkExpr (XExpr (WrapExpr (HsWrap co_fn expr))) = runZonkBndrT (zonkCoFn co_fn) $ \ new_co_fn -> ===================================== compiler/Language/Haskell/Syntax/Expr.hs ===================================== @@ -41,7 +41,6 @@ import Data.Bool import Data.Eq import Data.Maybe import Data.List.NonEmpty ( NonEmpty ) -import GHC.Types.Name.Reader {- Note [RecordDotSyntax field updates] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -338,19 +337,6 @@ data HsExpr p = HsVar (XVar p) (LIdP p) -- ^ Variable -- See Note [Located RdrNames] - - | HsUnboundVar (XUnboundVar p) - RdrName -- ^ Unbound variable; also used for "holes" - -- (_ or _x). - -- Turned from HsVar to HsUnboundVar by the - -- renamer, when it finds an out-of-scope - -- variable or hole. - -- The (XUnboundVar p) field becomes an HoleExprRef - -- after typechecking; this is where the - -- erroring expression will be written after - -- solving. See Note [Holes] in GHC.Tc.Types.Constraint. - - | HsRecSel (XRecSel p) (FieldOcc p) -- ^ Variable pointing to record selector -- See Note [Non-overloaded record field selectors] and @@ -404,7 +390,7 @@ data HsExpr p -- NB Bracketed ops such as (+) come out as Vars. -- NB Sadly, we need an expr for the operator in an OpApp/Section since - -- the renamer may turn a HsVar into HsRecSel or HsUnboundVar + -- the renamer may turn a HsVar into HsRecSel or HsUnboundVarRn | OpApp (XOpApp p) (LHsExpr p) -- left operand @@ -651,6 +637,10 @@ data HsExpr p | HsEmbTy (XEmbTy p) (LHsWcType (NoGhcTc p)) + -- | Holes in expressions. + -- See Note [Holes in expressions] in GHC.Tc.Types.Constraint. + | HsHole (XHole p) + -- | Forall-types @forall tvs. t@ and @forall tvs -> t at . -- Used with @RequiredTypeArguments@, e.g. @fn (forall a. Proxy a)@. -- See Note [Types in terms] ===================================== compiler/Language/Haskell/Syntax/Extension.hs ===================================== @@ -448,6 +448,7 @@ type family XTick x type family XBinTick x type family XPragE x type family XEmbTy x +type family XHole x type family XForAll x type family XQual x type family XFunArr x ===================================== testsuite/tests/perf/compiler/hard_hole_fits.hs ===================================== @@ -12,7 +12,7 @@ import GHC (GhcPs) testMe :: HsExpr GhcPs -> Int testMe (HsVar a b) = _ -testMe (HsUnboundVar xuv uv) = _ +testMe (HsHole a) = _ testMe (HsOverLabel xol m_ip) = _ testMe (HsIPVar xv hin) = _ testMe (HsOverLit xole hol) = _ ===================================== testsuite/tests/perf/compiler/hard_hole_fits.stderr ===================================== @@ -17,14 +17,12 @@ hard_hole_fits.hs:14:22: warning: [GHC-88464] [-Wtyped-holes (in -Wdefault)] (imported from ‘Prelude’ at hard_hole_fits.hs:8:8-20 (and originally defined in ‘GHC.Internal.Enum’)) -hard_hole_fits.hs:15:32: warning: [GHC-88464] [-Wtyped-holes (in -Wdefault)] - • Found hole: _ :: Int - • In an equation for ‘testMe’: testMe (HsUnboundVar xuv uv) = _ - • Relevant bindings include - uv :: GHC.Types.Name.Reader.RdrName - (bound at hard_hole_fits.hs:15:26) - xuv :: Language.Haskell.Syntax.Extension.XUnboundVar GhcPs - (bound at hard_hole_fits.hs:15:22) +hard_hole_fits.hs:15:21: warning: [GHC-88464] [-Wtyped-holes (in -Wdefault)] + Found hole: _ :: Int + In an equation for ‘testMe’: testMe (HsHole a) = _ + Relevant bindings include + a :: Language.Haskell.Syntax.Extension.XHole GhcPs + (bound at hard_hole_fits.hs:15:16) testMe :: HsExpr GhcPs -> Int (bound at hard_hole_fits.hs:14:1) Valid hole fits include maxBound :: forall a. Bounded a => a ===================================== testsuite/tests/plugins/T20803-plugin/FixErrorsPlugin.hs ===================================== @@ -20,15 +20,15 @@ import Data.Maybe plugin :: Plugin plugin = defaultPlugin {parsedResultAction = parsedAction} --- Replace every hole (and other unbound vars) with the given expression +-- Replace every hole with the given expression replaceHoles :: forall a . Data a => HsExpr GhcPs -> a -> a replaceHoles new = gmapT \case (d :: d) -> replaceHoles new d `fromMaybe` tryHole where tryHole :: Maybe d tryHole = eqT @d @(HsExpr GhcPs) >>= \case - Eq.Refl | HsUnboundVar _ _ <- d -> Just new - _ -> Nothing + Eq.Refl | HsHole _ <- d -> Just new + _ -> Nothing parsedAction :: [CommandLineOption] -> ModSummary -> ParsedResult -> Hsc ParsedResult ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -329,7 +329,7 @@ instance HasTrailing AnnExplicitSum where trailing _ = [] setTrailing a _ = a -instance HasTrailing (Maybe EpAnnUnboundVar) where +instance HasTrailing (Maybe EpAnnHole) where trailing _ = [] setTrailing a _ = a @@ -3042,16 +3042,16 @@ instance ExactPrint (HsExpr GhcPs) where then markAnnotated n else return n return (HsVar x n') - exact (HsUnboundVar an n) = do + exact x@(HsHole an) = do case an of - Just (EpAnnUnboundVar (ob,cb) l) -> do - ob' <- printStringAtAA ob "`" + Just (EpAnnHole (ob,cb) l) -> do + ob' <- printStringAtAA ob "`" l' <- printStringAtAA l "_" cb' <- printStringAtAA cb "`" - return (HsUnboundVar (Just (EpAnnUnboundVar (ob',cb') l')) n) - _ -> do - printStringAdvanceA "_" >> return () - return (HsUnboundVar an n) + return (HsHole (Just (EpAnnHole (ob',cb') l'))) + Nothing -> do + printStringAdvanceA "_" + return x exact x@(HsOverLabel src l) = do printStringAdvanceA "#" >> return () case src of View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7ed954d99ef1a3f3798630bb46d08b404567e9d1 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7ed954d99ef1a3f3798630bb46d08b404567e9d1 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 16:06:13 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Fri, 13 Sep 2024 12:06:13 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 5 commits: determinism: Cmm unique renaming pass Message-ID: <66e462f5d49fc_34e4c11be7d42785@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: bec9f340 by Rodrigo Mesquita at 2024-09-13T16:53:09+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - a96081f1 by Rodrigo Mesquita at 2024-09-13T16:53:09+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - 45de906a by Rodrigo Mesquita at 2024-09-13T16:53:09+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 52c39590 by Rodrigo Mesquita at 2024-09-13T16:53:10+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - bea3b26c by Rodrigo Mesquita at 2024-09-13T17:05:46+01:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 30 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm/Reg/Liveness.hs - compiler/GHC/Driver/Config/StgToCmm.hs - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/GenerateCgIPEStub.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Session.hs - compiler/GHC/Stg/Debug.hs - compiler/GHC/StgToCmm.hs - compiler/GHC/StgToCmm/CgUtils.hs - compiler/GHC/StgToCmm/Config.hs - compiler/GHC/StgToCmm/InfoTableProv.hs - compiler/GHC/StgToCmm/Monad.hs - compiler/GHC/StgToCmm/Prof.hs - compiler/GHC/StgToCmm/Utils.hs - compiler/GHC/Types/IPE.hs - compiler/GHC/Types/Name.hs - compiler/GHC/Types/Unique.hs - compiler/GHC/Types/Unique/DFM.hs - compiler/GHC/Utils/Outputable.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d7091d6dea51df10ab5de2d0b815192c5bb7577e...bea3b26cfbc860d9de2afa337c1c07fc3d13364e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d7091d6dea51df10ab5de2d0b815192c5bb7577e...bea3b26cfbc860d9de2afa337c1c07fc3d13364e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 16:07:26 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Fri, 13 Sep 2024 12:07:26 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] 2 commits: wrk: safer c code Message-ID: <66e4633ecd232_34e4c110edac37f@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 368ec171 by doyougnu at 2024-09-13T12:02:16-04:00 wrk: safer c code previously my patch was accidentally leaking the deadbeef addr - - - - - 6b17092f by doyougnu at 2024-09-13T12:06:02-04:00 wrk: update base-exports to accomodate new flag - - - - - 2 changed files: - rts/Linker.c - testsuite/tests/interface-stability/base-exports.stdout Changes: ===================================== rts/Linker.c ===================================== @@ -952,12 +952,11 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) lbl); IF_DEBUG(linker, printLoadedObjects()); fflush(stderr); + } else { + // if --link-unknown-symbols is passed into the RTS we allow the linker + // to optimistically continue + r = (void*) 0xDEADBEEF; } - - // if --link-unknown-symbols is passed into the RTS we allow the linker - // to optimistically continue - r = (void*) 0xDEADBEEF; - } if (!runPendingInitializers()) { ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -234,7 +234,7 @@ module Control.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -303,7 +303,7 @@ module Control.Exception where module Control.Exception.Annotation where -- Safety: None type ExceptionAnnotation :: * -> Constraint - class ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where + class ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where displayExceptionAnnotation :: a -> GHC.Internal.Base.String default displayExceptionAnnotation :: GHC.Internal.Show.Show a => a -> GHC.Internal.Base.String {-# MINIMAL #-} @@ -345,7 +345,7 @@ module Control.Exception.Base where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -850,11 +850,11 @@ module Data.Data where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b constrFields :: Constr -> [GHC.Internal.Base.String] constrFixity :: Constr -> Fixity @@ -897,7 +897,7 @@ module Data.Data where showConstr :: Constr -> GHC.Internal.Base.String showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -921,14 +921,14 @@ module Data.Dynamic where -- Safety: Safe type Dynamic :: * data Dynamic where - Dynamic :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic + Dynamic :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} dynApp :: Dynamic -> Dynamic -> Dynamic dynApply :: Dynamic -> Dynamic -> GHC.Internal.Maybe.Maybe Dynamic - dynTypeRep :: Dynamic -> ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + dynTypeRep :: Dynamic -> ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep fromDyn :: forall a. Typeable a => Dynamic -> a -> a fromDynamic :: forall a. Typeable a => Dynamic -> GHC.Internal.Maybe.Maybe a toDyn :: forall a. Typeable a => a -> Dynamic @@ -1815,11 +1815,11 @@ module Data.Typeable where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b decT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Data.Either.Either ((a :~: b) -> GHC.Internal.Base.Void) (a :~: b) eqT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Maybe.Maybe (a :~: b) @@ -1834,7 +1834,7 @@ module Data.Typeable where rnfTypeRep :: TypeRep -> () showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -5255,7 +5255,7 @@ module GHC.Event where data EventManager = ... type FdKey :: * data FdKey - = ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] + = ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] ; GHC.Internal.Foreign.C.Types.N:CInt[0])GHC.Internal.System.Posix.Types.Fd, ...} type IOCallback :: * @@ -5288,9 +5288,9 @@ module GHC.Event.TimeOut where type TimeoutEdit :: * type TimeoutEdit = TimeoutQueue -> TimeoutQueue type TimeoutKey :: * - newtype TimeoutKey = TK ghc-internal-9.1001.0:GHC.Internal.Event.Unique.Unique + newtype TimeoutKey = TK ghc-internal-9.1100.0:GHC.Internal.Event.Unique.Unique type TimeoutQueue :: * - type TimeoutQueue = ghc-internal-9.1001.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback + type TimeoutQueue = ghc-internal-9.1100.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback module GHC.Exception where -- Safety: Safe @@ -5302,7 +5302,7 @@ module GHC.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -5331,7 +5331,7 @@ module GHC.Exception.Type where type ArithException :: * data ArithException = Overflow | Underflow | LossOfPrecision | DivideByZero | Denormal | RatioZeroDenominator type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -7884,8 +7884,8 @@ module GHC.IO.Handle where hTryLock :: Handle -> LockMode -> GHC.Types.IO GHC.Types.Bool hWaitForInput :: Handle -> GHC.Types.Int -> GHC.Types.IO GHC.Types.Bool isEOF :: GHC.Types.IO GHC.Types.Bool - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle nativeNewline :: Newline nativeNewlineMode :: NewlineMode noNewlineTranslation :: NewlineMode @@ -7935,11 +7935,11 @@ module GHC.IO.Handle.Internals where ioe_notReadable :: forall a. GHC.Types.IO a ioe_notWritable :: forall a. GHC.Types.IO a ioe_semiclosedHandle :: forall a. GHC.Types.IO a - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle openTextEncoding :: forall a. GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.HandleType -> (forall es ds. GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextEncoder es) -> GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextDecoder ds) -> GHC.Types.IO a) -> GHC.Types.IO a readTextDevice :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer readTextDeviceNonBlocking :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer @@ -8003,7 +8003,7 @@ module GHC.IO.Handle.Types where type Handle__ :: * data Handle__ = forall dev enc_state dec_state. - (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => + (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => Handle__ {haDevice :: !dev, haType :: HandleType, haByteBuffer :: ! {-# UNPACK #-}(GHC.Internal.IORef.N:IORef[0] _N)(GHC.Internal.IORef.IORef (GHC.Internal.IO.Buffer.Buffer GHC.Internal.Word.Word8)), @@ -9106,7 +9106,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkUnknownSymbols :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * @@ -9417,7 +9417,7 @@ module GHC.StaticPtr where -- Safety: None type IsStatic :: (* -> *) -> Constraint class IsStatic p where - fromStaticPtr :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a + fromStaticPtr :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a {-# MINIMAL fromStaticPtr #-} type StaticKey :: * type StaticKey = GHC.Internal.Fingerprint.Type.Fingerprint @@ -10813,8 +10813,8 @@ module Type.Reflection where data (:~~:) a b where HRefl :: forall {k1} (a :: k1). (:~~:) a a pattern App :: forall k2 (t :: k2). () => forall k1 (a :: k1 -> k2) (b :: k1). (t ~ a b) => TypeRep a -> TypeRep b -> TypeRep t - pattern Con :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a - pattern Con' :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a + pattern Con :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a + pattern Con' :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a pattern Fun :: forall k (fun :: k). () => forall (r1 :: GHC.Types.RuntimeRep) (r2 :: GHC.Types.RuntimeRep) (arg :: TYPE r1) (res :: TYPE r2). (k ~ *, fun ~~ (arg -> res)) => TypeRep arg -> TypeRep res -> TypeRep fun type Module :: * data Module = ... @@ -10831,7 +10831,7 @@ module Type.Reflection where type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} decTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Data.Either.Either ((a :~~: b) -> GHC.Internal.Base.Void) (a :~~: b) eqTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Maybe.Maybe (a :~~: b) moduleName :: Module -> GHC.Internal.Base.String @@ -10866,9 +10866,9 @@ module Type.Reflection.Unsafe where data TypeRep a where ... mkTrApp :: forall k1 k2 (a :: k1 -> k2) (b :: k1). TypeRep a -> TypeRep b -> TypeRep (a b) - mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a + mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a mkTyCon :: GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Types.Int -> KindRep -> TyCon - someTypeRepFingerprint :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint + someTypeRepFingerprint :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConKindArgs :: TyCon -> GHC.Types.Int tyConKindRep :: TyCon -> KindRep @@ -11202,9 +11202,9 @@ instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.I instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Monoid (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Monoid (f p), GHC.Internal.Base.Monoid (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Monoid (f (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Monoid (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Monoid (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11260,9 +11260,9 @@ instance forall a. GHC.Internal.Base.Semigroup (Data.Semigroup.Last a) -- Define instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Semigroup (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Semigroup (f p), GHC.Internal.Base.Semigroup (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Semigroup (f (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11392,20 +11392,20 @@ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Last -- Defined instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Max -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Min -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.IO.Class.MonadIO GHC.Types.IO -- Defined in ‘GHC.Internal.Control.Monad.IO.Class’ -instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ -instance forall (m :: * -> *) a. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ +instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ +instance forall (m :: * -> *) a. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Functor.ZipList.ZipList a) -- Defined in ‘GHC.Internal.Functor.ZipList’ instance GHC.Internal.Data.Data.Data Data.Array.Byte.ByteArray -- Defined in ‘Data.Array.Byte’ -instance forall s. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ +instance forall s. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Complex.Complex a) -- Defined in ‘Data.Complex’ -instance forall i j (a :: i) (b :: j). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall i j (a :: i) (b :: j). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.All -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.Any -- Defined in ‘GHC.Internal.Data.Data’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b, GHC.Internal.Ix.Ix a) => GHC.Internal.Data.Data.Data (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.Associativity -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Bool -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Char -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k a (b :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall k a (b :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.DecidedStrictness -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Double -- Defined in ‘GHC.Internal.Data.Data’ @@ -11453,10 +11453,10 @@ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word32 -- Defined in ‘G instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word64 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word8 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Foreign.Ptr.WordPtr -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ -instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ +instance forall k (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ +instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b) => GHC.Internal.Data.Data.Data (Data.Semigroup.Arg a b) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.First a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.Last a) -- Defined in ‘Data.Semigroup’ @@ -11522,7 +11522,7 @@ instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SCha instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SSymbol -- Defined in ‘GHC.Internal.TypeLits’ instance forall k (a :: k). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k (a :: k1). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ -instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1). GHC.Internal.Data.Type.Equality.TestEquality f => GHC.Internal.Data.Type.Equality.TestEquality (Data.Functor.Compose.Compose f g) -- Defined in ‘Data.Functor.Compose’ instance forall a k (b :: k). GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ instance forall a. GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Bits.And a) -- Defined in ‘GHC.Internal.Data.Bits’ @@ -11719,7 +11719,7 @@ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Ba instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.RecUpdError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.TypeError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ -instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ instance [safe] GHC.Internal.Exception.Type.Exception System.Timeout.Timeout -- Defined in ‘System.Timeout’ instance forall a k (b :: k). GHC.Internal.Float.Floating a => GHC.Internal.Float.Floating (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ @@ -12390,8 +12390,8 @@ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataRep -- Defined in ‘ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataType -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Internal.Show.Show (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance GHC.Internal.Show.Show GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ instance forall a b. (GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). Data.Fixed.HasResolution a => GHC.Internal.Show.Show (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ @@ -12462,13 +12462,13 @@ instance GHC.Internal.Show.Show GHC.Internal.Foreign.C.Types.CWchar -- Defined i instance forall a. GHC.Internal.Show.Show (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Foreign.C.ConstPtr’ instance forall a b. (GHC.Internal.Ix.Ix a, GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Internal.Show.Show GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Internal.Show.Show GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ instance GHC.Internal.Show.Show GHC.Types.Double -- Defined in ‘GHC.Internal.Float’ instance GHC.Internal.Show.Show GHC.Types.Float -- Defined in ‘GHC.Internal.Float’ @@ -12502,7 +12502,7 @@ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Handle -- Defined i instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.HandleType -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Newline -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.NewlineMode -- Defined in ‘GHC.Internal.IO.Handle.Types’ -instance [safe] GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.HandlePosn -- Defined in ‘GHC.Internal.IO.Handle’ instance GHC.Internal.Show.Show GHC.Internal.IO.IOMode.IOMode -- Defined in ‘GHC.Internal.IO.IOMode’ instance GHC.Internal.Show.Show GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ @@ -12599,8 +12599,8 @@ instance GHC.Classes.Eq GHC.Internal.Data.Data.ConstrRep -- Defined in ‘GHC.In instance GHC.Classes.Eq GHC.Internal.Data.Data.DataRep -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Classes.Eq GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Classes.Eq (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Eq a, GHC.Classes.Eq b) => GHC.Classes.Eq (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Eq (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Eq (f (g a)) => GHC.Classes.Eq (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ @@ -12676,13 +12676,13 @@ instance forall a. GHC.Classes.Eq (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) - instance forall i e. (GHC.Internal.Ix.Ix i, GHC.Classes.Eq e) => GHC.Classes.Eq (GHC.Internal.Arr.Array i e) -- Defined in ‘GHC.Internal.Arr’ instance forall s i e. GHC.Classes.Eq (GHC.Internal.Arr.STArray s i e) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Classes.Eq GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ instance GHC.Classes.Eq GHC.Internal.Event.TimeOut.TimeoutKey -- Defined in ‘GHC.Internal.Event.TimeOut’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Classes.Eq GHC.Internal.Stack.Types.SrcLoc -- Defined in ‘GHC.Internal.Stack.Types’ instance GHC.Classes.Eq GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ instance GHC.Classes.Eq GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ @@ -12778,8 +12778,8 @@ instance GHC.Classes.Ord GHC.Internal.Unicode.GeneralCategory -- Defined in ‘G instance forall k (a :: k) (b :: k). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k2 (a :: k1) (b :: k2). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k (s :: k). GHC.Classes.Ord (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Ord ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Ord ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Ord a, GHC.Classes.Ord b) => GHC.Classes.Ord (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Ord (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Ord (f (g a)) => GHC.Classes.Ord (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/501963a972af8f4266ea15da89946e8c943ccef0...6b17092ff82644a47d7f41ccf649f77c4f8ed8ff -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/501963a972af8f4266ea15da89946e8c943ccef0...6b17092ff82644a47d7f41ccf649f77c4f8ed8ff You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 21:59:00 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Fri, 13 Sep 2024 17:59:00 -0400 Subject: [Git][ghc/ghc][wip/T23479] 56 commits: rts: fix checkClosure error message Message-ID: <66e4b5a463f94_1aa025642c2410389@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 363eb215 by Serge S. Gulin at 2024-09-14T00:56:28+03:00 JS: Re-add optimization for literal strings in genApp (fixes 23479 (muted temporary)) Based on https://gitlab.haskell.org/ghc/ghc/-/merge_requests/10588/ - - - - - 5acd4edb by Serge S. Gulin at 2024-09-14T00:56:32+03:00 Use name defined at `GHC.Builtin.Names` - - - - - edcee7f0 by Serge S. Gulin at 2024-09-14T00:56:32+03:00 Apply 1 suggestion(s) to 1 file(s) Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 37f73337 by Serge S. Gulin at 2024-09-14T00:56:32+03:00 Attempt to take 805 for id - - - - - a4e51ca0 by Serge S. Gulin at 2024-09-14T00:56:32+03:00 Attempt to add to basicKnownKeyNames Co-authored-by: Andrei Borzenkov <root at sandwitch.dev> Co-authored-by: Danil Berestov <goosedb at yandex.ru> - - - - - 13bd2d8a by Serge S. Gulin at 2024-09-14T00:56:32+03:00 Naive attempt to add `StgLitArg (LitString bs)` - - - - - f875a0a4 by Serge S. Gulin at 2024-09-14T00:56:32+03:00 WIP add logging - - - - - 18 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f601a3efabfcdea2c92daa05c8692c1f4831eefb...f875a0a4c5b1c5b6ddc70a2ff01bd82b920821e6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f601a3efabfcdea2c92daa05c8692c1f4831eefb...f875a0a4c5b1c5b6ddc70a2ff01bd82b920821e6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 13 22:23:26 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Fri, 13 Sep 2024 18:23:26 -0400 Subject: [Git][ghc/ghc][wip/T23479] WIP add logging Message-ID: <66e4bb5e85647_1aa0257ec2f01086ae@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: 1a145d43 by Serge S. Gulin at 2024-09-14T01:23:19+03:00 WIP add logging - - - - - 1 changed file: - compiler/GHC/StgToJS/Sinker.hs Changes: ===================================== compiler/GHC/StgToJS/Sinker.hs ===================================== @@ -43,9 +43,11 @@ sinkPgm m pgm ++ ((DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) stringLits stringLits) ) where - selectLifted :: CgStgTopBinding -> Either CgStgBinding (Id, ByteString) + -- selectLifted :: CgStgTopBinding -> Either CgStgBinding (Id, ByteString) + selectLifted :: CgStgTopBinding -> Either CgStgBinding CgStgTopBinding selectLifted (StgTopLifted b) = Left b - selectLifted (StgTopStringLit i b) = Right (i, b) + -- selectLifted (StgTopStringLit i b) = Right (i, b) + selectLifted x = Right x selectLiftedInCase :: [CgStgBinding] -> (Id, ByteString) -> Either (Id, ByteString) (Id, ByteString) selectLiftedInCase bindings (i, _) = undefined View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1a145d43264aa9c933fa4bbc1913c0aac28ea174 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1a145d43264aa9c933fa4bbc1913c0aac28ea174 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 01:17:10 2024 From: gitlab at gitlab.haskell.org (Matthew Craven (@clyring)) Date: Fri, 13 Sep 2024 21:17:10 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T23847 Message-ID: <66e4e41641289_1aa025f5a7d012662e@gitlab.mail> Matthew Craven pushed new branch wip/T23847 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T23847 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 10:50:27 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sat, 14 Sep 2024 06:50:27 -0400 Subject: [Git][ghc/ghc][wip/supersven/aarch64-jump-tables] 37 commits: AArch64: Implement takeRegRegMoveInstr Message-ID: <66e56a73d103d_310f1f1fa8339a@gitlab.mail> Sven Tennie pushed to branch wip/supersven/aarch64-jump-tables at Glasgow Haskell Compiler / GHC Commits: 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 3b037c58 by Sven Tennie at 2024-09-14T10:50:18+00:00 AArch64: Implement switch/jump tables (#19912) This improves the performance of Cmm switch statements (compared to a chain of if statements.) - - - - - 15 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c67216918cd9c31391e3b30ba2ddbd50e0ab6958...3b037c583aa46b04722f715e9f1a6f564e59c075 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c67216918cd9c31391e3b30ba2ddbd50e0ab6958...3b037c583aa46b04722f715e9f1a6f564e59c075 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 15:50:22 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sat, 14 Sep 2024 11:50:22 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/supersven/riscv64-cross-ci Message-ID: <66e5b0be52fa2_310fd055fc50332@gitlab.mail> Sven Tennie pushed new branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/supersven/riscv64-cross-ci You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 15:56:47 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sat, 14 Sep 2024 11:56:47 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Patch job to see some action Message-ID: <66e5b23f9ba6c_310fec8d3054253@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 4d6f3701 by Sven Tennie at 2024-09-14T17:56:30+02:00 Patch job to see some action - - - - - 1 changed file: - .gitlab/jobs.yaml Changes: ===================================== .gitlab/jobs.yaml ===================================== @@ -1891,7 +1891,7 @@ ], "rules": [ { - "if": "(\"true\" == \"true\") && ($RELEASE_JOB != \"yes\") && ($NIGHTLY)", + "if": "(\"true\" == \"true\") && ($RELEASE_JOB != \"yes\") && ($NIGHTLY == null)", "when": "on_success" } ], View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4d6f370153ea19b0d69936396ded1f875557f74e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4d6f370153ea19b0d69936396ded1f875557f74e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 16:38:34 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sat, 14 Sep 2024 12:38:34 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Use latest supersven image Message-ID: <66e5bc0a7acaf_395eee24f158516d8@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: ba87f1a0 by Sven Tennie at 2024-09-14T18:38:07+02:00 Use latest supersven image The image is WIP, too. - - - - - 1 changed file: - .gitlab/jobs.yaml Changes: ===================================== .gitlab/jobs.yaml ===================================== @@ -1882,7 +1882,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12-riscv64-qemu:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12-riscv64-qemu:latest", "needs": [ { "artifacts": false, @@ -5313,7 +5313,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12-riscv64-qemu:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12-riscv64-qemu:latest", "needs": [ { "artifacts": false, View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ba87f1a04e2540c12d8030df5c22d36480370d48 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ba87f1a04e2540c12d8030df5c22d36480370d48 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 17:26:41 2024 From: gitlab at gitlab.haskell.org (Krzysztof Gogolewski (@monoidal)) Date: Sat, 14 Sep 2024 13:26:41 -0400 Subject: [Git][ghc/ghc][wip/T25243] Only allow (a => b) :: Constraint rather than CONSTRAINT rep Message-ID: <66e5c751cd07a_395eee4d83fc66362@gitlab.mail> Krzysztof Gogolewski pushed to branch wip/T25243 at Glasgow Haskell Compiler / GHC Commits: a5617c32 by Krzysztof Gogolewski at 2024-09-14T19:22:55+02:00 Only allow (a => b) :: Constraint rather than CONSTRAINT rep Fixes #25243 - - - - - 4 changed files: - compiler/GHC/Tc/Gen/HsType.hs - + testsuite/tests/quantified-constraints/T25243.hs - + testsuite/tests/quantified-constraints/T25243.stderr - testsuite/tests/quantified-constraints/all.T Changes: ===================================== compiler/GHC/Tc/Gen/HsType.hs ===================================== @@ -1158,10 +1158,13 @@ tcHsType mode (HsQualTy { hst_ctxt = ctxt, hst_body = rn_ty }) exp_kind | null (unLoc ctxt) = tcLHsType mode rn_ty exp_kind -- See Note [Body kind of a HsQualTy] - | Check kind <- exp_kind, isConstraintLikeKind kind + | Check kind <- exp_kind -- Checking mode + , isConstraintLikeKind kind -- CONSTRAINT rep = do { ctxt' <- tc_hs_context mode ctxt ; ty' <- tc_check_lhs_type mode rn_ty constraintKind - ; return (tcMkDFunPhiTy ctxt' ty') } + ; let res_ty = tcMkDFunPhiTy ctxt' ty' + ; checkExpKind (unLoc rn_ty) res_ty + constraintKind exp_kind } | otherwise = do { ctxt' <- tc_hs_context mode ctxt @@ -2121,11 +2124,14 @@ should be '*' we risk getting TWO error messages, one saying that Eq the left of the outer (=>). How do we figure out the right body kind? Well, it's a bit of a -kludge: I just look at the expected kind. If it's Constraint, we -must be in this instance situation context. It's a kludge because it -wouldn't work if any unification was involved to compute that result -kind -- but it isn't. (The true way might be to use the 'mode' -parameter, but that seemed like a sledgehammer to crack a nut.) +kludge: I just look at the expected kind. If we are in checking mode +(`exp_kind` = `Check k`), and the pushed-in kind `k` is `Constraint`, then +we check that the body type has kind `Constraint` too. +This is a kludge because it wouldn't work if any unification was +involved to compute that result kind -- but it isn't. +Actually, we only check whether `k` is a `CONSTRAINT rep`, but in that +case enforce that `rep` is a LiftedRep. This gives a better error message +in T25243. Note [Inferring tuple kinds] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== testsuite/tests/quantified-constraints/T25243.hs ===================================== @@ -0,0 +1,8 @@ +{-# LANGUAGE DataKinds, QuantifiedConstraints, UndecidableInstances #-} +module T25243 where + +import GHC.Exts +import Data.Kind + +type T :: Constraint -> Constraint -> CONSTRAINT IntRep +type T a b = a => b ===================================== testsuite/tests/quantified-constraints/T25243.stderr ===================================== @@ -0,0 +1,5 @@ +T25243.hs:8:14: error: [GHC-83865] + • Expected an IntRep constraint, but ‘b’ is a lifted constraint + • In the type ‘a => b’ + In the type declaration for ‘T’ + ===================================== testsuite/tests/quantified-constraints/all.T ===================================== @@ -45,3 +45,4 @@ test('T23143', normal, compile, ['']) test('T23333', normal, compile, ['']) test('T23323', normal, compile, ['']) test('T22238', normal, compile, ['']) +test('T25243', normal, compile_fail, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a5617c32a99127692984ceb23a91e9d3bbb25a83 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a5617c32a99127692984ceb23a91e9d3bbb25a83 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 20:52:00 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Sat, 14 Sep 2024 16:52:00 -0400 Subject: [Git][ghc/ghc][wip/T23479] 2 commits: Add STG debug from JS Sinker Message-ID: <66e5f7705ab46_3ba008fc1fc3758f@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: c878e3d4 by Serge S. Gulin at 2024-09-14T23:51:08+03:00 Add STG debug from JS Sinker - - - - - a547d899 by Serge S. Gulin at 2024-09-14T23:51:44+03:00 Add eager Sinker's strings unfloater - - - - - 4 changed files: - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/Session.hs - compiler/GHC/StgToJS/CodeGen.hs - compiler/GHC/StgToJS/Sinker.hs Changes: ===================================== compiler/GHC/Driver/Flags.hs ===================================== @@ -470,6 +470,7 @@ data DumpFlag | Opt_D_dump_stg_cg -- ^ STG (after stg2stg) | Opt_D_dump_stg_tags -- ^ Result of tag inference analysis. | Opt_D_dump_stg_final -- ^ Final STG (before cmm gen) + | Opt_D_dump_stg_from_js_sinker -- ^ STG after JS sinker | Opt_D_dump_call_arity | Opt_D_dump_exitify | Opt_D_dump_dmdanal ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -1515,6 +1515,8 @@ dynamic_flags_deps = [ "Use `-ddump-stg-from-core` or `-ddump-stg-final` instead" , make_ord_flag defGhcFlag "ddump-stg-tags" (setDumpFlag Opt_D_dump_stg_tags) + , make_ord_flag defGhcFlag "ddump-stg-from-js-sinker" + (setDumpFlag Opt_D_dump_stg_tags) , make_ord_flag defGhcFlag "ddump-call-arity" (setDumpFlag Opt_D_dump_call_arity) , make_ord_flag defGhcFlag "ddump-exitify" ===================================== compiler/GHC/StgToJS/CodeGen.hs ===================================== @@ -11,7 +11,7 @@ where import GHC.Prelude -import GHC.Driver.Flags (DumpFlag (Opt_D_dump_js)) +import GHC.Driver.Flags (DumpFlag (Opt_D_dump_js, Opt_D_dump_stg_from_js_sinker)) import GHC.JS.Ppr import GHC.JS.JStg.Syntax @@ -81,7 +81,8 @@ stgToJS logger config stg_binds0 this_mod spt_entries foreign_stubs cccs output_ -- TODO: avoid top level lifting in core-2-core when the JS backend is -- enabled instead of undoing it here - -- TODO: add dump pass for optimized STG ast for JS + putDumpFileMaybe logger Opt_D_dump_stg_from_js_sinker "STG Optimized JS Sinker:" FormatSTG + (pprGenStgTopBindings (StgPprOpts False) stg_binds) (deps,lus) <- runG config this_mod unfloated_binds $ do ifProfilingM $ initCostCentres cccs ===================================== compiler/GHC/StgToJS/Sinker.hs ===================================== @@ -19,8 +19,8 @@ import GHC.Utils.Misc (partitionWith) import GHC.StgToJS.Utils import Data.Char -import Data.List (partition) import Data.Maybe +import Data.Bifunctor (Bifunctor (..)) import GHC.Utils.Outputable (showSDocUnsafe, showPprUnsafe) import qualified GHC.Utils.Trace as DT import Data.ByteString (ByteString) @@ -39,40 +39,21 @@ sinkPgm :: Module -> (UniqFM Id CgStgExpr, [CgStgTopBinding]) sinkPgm m pgm = ( (DT.trace . (++) "" . showPprUnsafe) sunk sunk - , ((DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (map StgTopLifted pgm'') (map StgTopLifted pgm'')) - ++ ((DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) stringLits stringLits) + , traceDoc "" (map StgTopLifted pgm''') (map StgTopLifted pgm''') + ++ traceDoc "" stringLits stringLits ) where - -- selectLifted :: CgStgTopBinding -> Either CgStgBinding (Id, ByteString) - selectLifted :: CgStgTopBinding -> Either CgStgBinding CgStgTopBinding + traceDoc head = DT.trace . (++) head . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts + + selectLifted :: CgStgTopBinding -> Either CgStgBinding (Id, ByteString) selectLifted (StgTopLifted b) = Left b - -- selectLifted (StgTopStringLit i b) = Right (i, b) - selectLifted x = Right x + selectLifted (StgTopStringLit i b) = Right (i, b) - selectLiftedInCase :: [CgStgBinding] -> (Id, ByteString) -> Either (Id, ByteString) (Id, ByteString) - selectLiftedInCase bindings (i, _) = undefined - where - isUsedInCase :: CgStgBinding -> (Id, ByteString) -> Either (Id, ByteString) (Id, ByteString) - -- Recursive expressions are not a target - isUsedInCase (StgRec _) x = Left x - -- Constructors should be left as is - isUsedInCase (StgNonRec _ (StgRhsCon {})) x = Left x - -- If function has no arguments it is not a target - isUsedInCase (StgNonRec i' (StgRhsClosure _ _ _ [] _ _)) x = Left x - -- If function has more arguments than 1 it is not a target - isUsedInCase (StgNonRec i' (StgRhsClosure _ _ _ (_ : (_ : _)) _ _)) x = Left x - -- If function has single argument probably it uses a string - isUsedInCase (StgNonRec i' (StgRhsClosure _ _ _ [arg1] _ _)) x - | arg1 == i = undefined - | otherwise = Left x - - selectLiftedToUnlift :: [CgStgTopBinding] -> ([CgStgTopBinding], [CgStgTopBinding]) - selectLiftedToUnlift = undefined - where - xxx = undefined + (pgm', stringLits') = partitionWith selectLifted pgm + stringLits = uncurry StgTopStringLit <$> stringLits' - (pgm', stringLits) = partitionWith selectLifted pgm - (sunk, pgm'') = sinkPgm' m $ (DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (map StgTopLifted pgm') pgm' + (pgm'', _usedStringLitNames) = unfloatStringLits (listToUFM $ first idName <$> stringLits') pgm' + (sunk, pgm''') = sinkPgm' m $ (DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (map StgTopLifted pgm'') pgm'' sinkPgm' :: Module @@ -211,3 +192,112 @@ topSortDecls _m binds = rest ++ nr' nr' | (not . null) [()| CyclicSCC _ <- stronglyConnCompG g] = error "topSortDecls: unexpected cycle" | otherwise = map node_payload (topologicalSortG g) + +-- | We are doing attempts to unfloat string literals back to +-- the call site. Further special JS optimizations +-- can generate more performant operations over them. +unfloatStringLits :: UniqFM Name ByteString -> [CgStgBinding] -> ([CgStgBinding], UniqSet Name) +unfloatStringLits stringLits allBindings = (binderWithoutChanges ++ binderWithUnfloatedStringLit, actuallyUsedStringLitNames) + where + (binderWithoutChanges, binderWithUnfloatedStringLitPairs) = partitionWith substituteStringLit allBindings + + binderWithUnfloatedStringLit = fst <$> binderWithUnfloatedStringLitPairs + actuallyUsedStringLitNames = unionManyUniqSets (snd <$> binderWithUnfloatedStringLitPairs) + + substituteStringLit :: CgStgBinding -> Either CgStgBinding (CgStgBinding, UniqSet Name) + substituteStringLit x@(StgRec bnds) + | isEmptyUniqSet names = Left x + | otherwise = Right (StgRec bnds', names) + where + (bnds', names) = extractNames id $ do + (i, rhs) <- bnds + pure $ case processStgRhs rhs of + Nothing -> Left (i, rhs) + Just (rhs', names) -> Right ((i, rhs'), names) + substituteStringLit x@(StgNonRec binder rhs) + = maybe (Left x) + (\(body', names) -> Right (StgNonRec binder body', names)) + (processStgRhs rhs) + + processStgRhs :: CgStgRhs -> Maybe (CgStgRhs, UniqSet Name) + processStgRhs (StgRhsCon ccs dataCon mu ticks args typ) + | isEmptyUniqSet names = Nothing + | otherwise = Just (StgRhsCon ccs dataCon mu ticks unified typ, names) + where + (unified, names) = substituteArgWithNames args + processStgRhs (StgRhsClosure fvs ccs upd bndrs body typ) + = (\(body', names) -> (StgRhsClosure fvs ccs upd bndrs body' typ, names)) <$> + processStgExpr body + + -- Recursive expressions + processStgExpr :: CgStgExpr -> Maybe (CgStgExpr, UniqSet Name) + processStgExpr (StgLit _) = Nothing + processStgExpr (StgTick _ _) = Nothing + processStgExpr (StgLet n b e) = + case (substituteStringLit b, processStgExpr e) of + (Left _, Nothing) -> Nothing + (Right (b', names), Nothing) -> Just (StgLet n b' e, names) + (Left _, Just (e', names)) -> Just (StgLet n b e', names) + (Right (b', names), Just (e', names')) -> Just (StgLet n b' e', names `unionUniqSets` names') + processStgExpr (StgLetNoEscape n b e) = + case (substituteStringLit b, processStgExpr e) of + (Left _, Nothing) -> Nothing + (Right (b', names), Nothing) -> Just (StgLetNoEscape n b' e, names) + (Left _, Just (e', names)) -> Just (StgLetNoEscape n b e', names) + (Right (b', names), Just (e', names')) -> Just (StgLetNoEscape n b' e', names `unionUniqSets` names') + -- We should keep the order: See Note [Case expression invariants] + processStgExpr (StgCase e bndr alt_type alts) = + case (isEmptyUniqSet names, processStgExpr e) of + (True, Nothing) -> Nothing + (True, Just (e', names')) -> Just (StgCase e' bndr alt_type alts, names') + (False, Nothing) -> Just (StgCase e bndr alt_type unified, names) + (False, Just (e', names')) -> Just (StgCase e' bndr alt_type unified, names `unionUniqSets` names') + where + (unified, names) = extractNames splitAlts alts + + splitAlts :: CgStgAlt -> Either CgStgAlt (CgStgAlt, UniqSet Name) + splitAlts alt@(GenStgAlt con bndrs rhs) = + case processStgExpr rhs of + Nothing -> Left alt + Just (alt', names) -> Right (GenStgAlt con bndrs alt', names) + + -- No args + processStgExpr (StgApp _ []) = Nothing + processStgExpr (StgConApp _ _ [] _) = Nothing + processStgExpr (StgOpApp _ [] _) = Nothing + + -- Main targets. Preserving the order of args is important + processStgExpr (StgApp fn args@(_:_)) + | isEmptyUniqSet names = Nothing + | otherwise = Just (StgApp fn unified, names) + where + (unified, names) = substituteArgWithNames args + processStgExpr (StgConApp dc n args@(_:_) tys) + | isEmptyUniqSet names = Nothing + | otherwise = Just (StgConApp dc n unified tys, names) + where + (unified, names) = substituteArgWithNames args + processStgExpr (StgOpApp op args@(_:_) tys) + | isEmptyUniqSet names = Nothing + | otherwise = Just (StgOpApp op unified tys, names) + where + (unified, names) = substituteArgWithNames args + + substituteArg :: StgArg -> Either StgArg (StgArg, Name) + substituteArg a@(StgLitArg _) = Left a + substituteArg a@(StgVarArg i) = + let name = idName i + in case lookupUFM stringLits name of + Nothing -> Left a + Just b -> Right (StgLitArg $ LitString b, name) + + substituteArgWithNames = extractNames (second (second unitUniqSet) . substituteArg) + + extractNames :: (a -> Either x (x, UniqSet Name)) -> [a] -> ([x], UniqSet Name) + extractNames splitter target = + let + splitted = splitter <$> target + combined = either (, emptyUniqSet) id <$> splitted + unified = fst <$> combined + names = unionManyUniqSets (snd <$> combined) + in (unified, names) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1a145d43264aa9c933fa4bbc1913c0aac28ea174...a547d8990b3ac40ba4a1a187bd7413aa36327d46 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1a145d43264aa9c933fa4bbc1913c0aac28ea174...a547d8990b3ac40ba4a1a187bd7413aa36327d46 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 21:45:45 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Sat, 14 Sep 2024 17:45:45 -0400 Subject: [Git][ghc/ghc][wip/T23479] Add eager Sinker's strings unfloater Message-ID: <66e6040984777_3e21498ab10671a6@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: 2b4abb88 by Serge S. Gulin at 2024-09-15T00:45:29+03:00 Add eager Sinker's strings unfloater - - - - - 1 changed file: - compiler/GHC/StgToJS/Sinker.hs Changes: ===================================== compiler/GHC/StgToJS/Sinker.hs ===================================== @@ -21,6 +21,7 @@ import GHC.StgToJS.Utils import Data.Char import Data.List (partition) import Data.Maybe +import Data.Bifunctor (Bifunctor (..)) import GHC.Utils.Outputable (showSDocUnsafe, showPprUnsafe) import qualified GHC.Utils.Trace as DT import Data.ByteString (ByteString) @@ -39,40 +40,21 @@ sinkPgm :: Module -> (UniqFM Id CgStgExpr, [CgStgTopBinding]) sinkPgm m pgm = ( (DT.trace . (++) "" . showPprUnsafe) sunk sunk - , ((DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (map StgTopLifted pgm'') (map StgTopLifted pgm'')) - ++ ((DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) stringLits stringLits) + , traceDoc "" (map StgTopLifted pgm''') (map StgTopLifted pgm''') + ++ traceDoc "" stringLits stringLits ) where - -- selectLifted :: CgStgTopBinding -> Either CgStgBinding (Id, ByteString) - selectLifted :: CgStgTopBinding -> Either CgStgBinding CgStgTopBinding + traceDoc head = DT.trace . (++) head . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts + + selectLifted :: CgStgTopBinding -> Either CgStgBinding (Id, ByteString) selectLifted (StgTopLifted b) = Left b - -- selectLifted (StgTopStringLit i b) = Right (i, b) - selectLifted x = Right x + selectLifted (StgTopStringLit i b) = Right (i, b) - selectLiftedInCase :: [CgStgBinding] -> (Id, ByteString) -> Either (Id, ByteString) (Id, ByteString) - selectLiftedInCase bindings (i, _) = undefined - where - isUsedInCase :: CgStgBinding -> (Id, ByteString) -> Either (Id, ByteString) (Id, ByteString) - -- Recursive expressions are not a target - isUsedInCase (StgRec _) x = Left x - -- Constructors should be left as is - isUsedInCase (StgNonRec _ (StgRhsCon {})) x = Left x - -- If function has no arguments it is not a target - isUsedInCase (StgNonRec i' (StgRhsClosure _ _ _ [] _ _)) x = Left x - -- If function has more arguments than 1 it is not a target - isUsedInCase (StgNonRec i' (StgRhsClosure _ _ _ (_ : (_ : _)) _ _)) x = Left x - -- If function has single argument probably it uses a string - isUsedInCase (StgNonRec i' (StgRhsClosure _ _ _ [arg1] _ _)) x - | arg1 == i = undefined - | otherwise = Left x - - selectLiftedToUnlift :: [CgStgTopBinding] -> ([CgStgTopBinding], [CgStgTopBinding]) - selectLiftedToUnlift = undefined - where - xxx = undefined + (pgm', stringLits') = partitionWith selectLifted pgm + stringLits = uncurry StgTopStringLit <$> stringLits' - (pgm', stringLits) = partitionWith selectLifted pgm - (sunk, pgm'') = sinkPgm' m $ (DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (map StgTopLifted pgm') pgm' + (pgm'', _usedStringLitNames) = unfloatStringLits (listToUFM $ first idName <$> stringLits') pgm' + (sunk, pgm''') = sinkPgm' m $ (DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (map StgTopLifted pgm'') pgm'' sinkPgm' :: Module @@ -211,3 +193,112 @@ topSortDecls _m binds = rest ++ nr' nr' | (not . null) [()| CyclicSCC _ <- stronglyConnCompG g] = error "topSortDecls: unexpected cycle" | otherwise = map node_payload (topologicalSortG g) + +-- | We are doing attempts to unfloat string literals back to +-- the call site. Further special JS optimizations +-- can generate more performant operations over them. +unfloatStringLits :: UniqFM Name ByteString -> [CgStgBinding] -> ([CgStgBinding], UniqSet Name) +unfloatStringLits stringLits allBindings = (binderWithoutChanges ++ binderWithUnfloatedStringLit, actuallyUsedStringLitNames) + where + (binderWithoutChanges, binderWithUnfloatedStringLitPairs) = partitionWith substituteStringLit allBindings + + binderWithUnfloatedStringLit = fst <$> binderWithUnfloatedStringLitPairs + actuallyUsedStringLitNames = unionManyUniqSets (snd <$> binderWithUnfloatedStringLitPairs) + + substituteStringLit :: CgStgBinding -> Either CgStgBinding (CgStgBinding, UniqSet Name) + substituteStringLit x@(StgRec bnds) + | isEmptyUniqSet names = Left x + | otherwise = Right (StgRec bnds', names) + where + (bnds', names) = extractNames id $ do + (i, rhs) <- bnds + pure $ case processStgRhs rhs of + Nothing -> Left (i, rhs) + Just (rhs', names) -> Right ((i, rhs'), names) + substituteStringLit x@(StgNonRec binder rhs) + = maybe (Left x) + (\(body', names) -> Right (StgNonRec binder body', names)) + (processStgRhs rhs) + + processStgRhs :: CgStgRhs -> Maybe (CgStgRhs, UniqSet Name) + processStgRhs (StgRhsCon ccs dataCon mu ticks args typ) + | isEmptyUniqSet names = Nothing + | otherwise = Just (StgRhsCon ccs dataCon mu ticks unified typ, names) + where + (unified, names) = substituteArgWithNames args + processStgRhs (StgRhsClosure fvs ccs upd bndrs body typ) + = (\(body', names) -> (StgRhsClosure fvs ccs upd bndrs body' typ, names)) <$> + processStgExpr body + + -- Recursive expressions + processStgExpr :: CgStgExpr -> Maybe (CgStgExpr, UniqSet Name) + processStgExpr (StgLit _) = Nothing + processStgExpr (StgTick _ _) = Nothing + processStgExpr (StgLet n b e) = + case (substituteStringLit b, processStgExpr e) of + (Left _, Nothing) -> Nothing + (Right (b', names), Nothing) -> Just (StgLet n b' e, names) + (Left _, Just (e', names)) -> Just (StgLet n b e', names) + (Right (b', names), Just (e', names')) -> Just (StgLet n b' e', names `unionUniqSets` names') + processStgExpr (StgLetNoEscape n b e) = + case (substituteStringLit b, processStgExpr e) of + (Left _, Nothing) -> Nothing + (Right (b', names), Nothing) -> Just (StgLetNoEscape n b' e, names) + (Left _, Just (e', names)) -> Just (StgLetNoEscape n b e', names) + (Right (b', names), Just (e', names')) -> Just (StgLetNoEscape n b' e', names `unionUniqSets` names') + -- We should keep the order: See Note [Case expression invariants] + processStgExpr (StgCase e bndr alt_type alts) = + case (isEmptyUniqSet names, processStgExpr e) of + (True, Nothing) -> Nothing + (True, Just (e', names')) -> Just (StgCase e' bndr alt_type alts, names') + (False, Nothing) -> Just (StgCase e bndr alt_type unified, names) + (False, Just (e', names')) -> Just (StgCase e' bndr alt_type unified, names `unionUniqSets` names') + where + (unified, names) = extractNames splitAlts alts + + splitAlts :: CgStgAlt -> Either CgStgAlt (CgStgAlt, UniqSet Name) + splitAlts alt@(GenStgAlt con bndrs rhs) = + case processStgExpr rhs of + Nothing -> Left alt + Just (alt', names) -> Right (GenStgAlt con bndrs alt', names) + + -- No args + processStgExpr (StgApp _ []) = Nothing + processStgExpr (StgConApp _ _ [] _) = Nothing + processStgExpr (StgOpApp _ [] _) = Nothing + + -- Main targets. Preserving the order of args is important + processStgExpr (StgApp fn args@(_:_)) + | isEmptyUniqSet names = Nothing + | otherwise = Just (StgApp fn unified, names) + where + (unified, names) = substituteArgWithNames args + processStgExpr (StgConApp dc n args@(_:_) tys) + | isEmptyUniqSet names = Nothing + | otherwise = Just (StgConApp dc n unified tys, names) + where + (unified, names) = substituteArgWithNames args + processStgExpr (StgOpApp op args@(_:_) tys) + | isEmptyUniqSet names = Nothing + | otherwise = Just (StgOpApp op unified tys, names) + where + (unified, names) = substituteArgWithNames args + + substituteArg :: StgArg -> Either StgArg (StgArg, Name) + substituteArg a@(StgLitArg _) = Left a + substituteArg a@(StgVarArg i) = + let name = idName i + in case lookupUFM stringLits name of + Nothing -> Left a + Just b -> Right (StgLitArg $ LitString b, name) + + substituteArgWithNames = extractNames (second (second unitUniqSet) . substituteArg) + + extractNames :: (a -> Either x (x, UniqSet Name)) -> [a] -> ([x], UniqSet Name) + extractNames splitter target = + let + splitted = splitter <$> target + combined = either (, emptyUniqSet) id <$> splitted + unified = fst <$> combined + names = unionManyUniqSets (snd <$> combined) + in (unified, names) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2b4abb8812cd4301aeb5da3623d3c90d3e2c6fe8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2b4abb8812cd4301aeb5da3623d3c90d3e2c6fe8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 14 22:15:38 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Sat, 14 Sep 2024 18:15:38 -0400 Subject: [Git][ghc/ghc][wip/T23479] Add STG debug from JS Sinker Message-ID: <66e60b0a63d48_3e214921c4746974c@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: ea4e6427 by Serge S. Gulin at 2024-09-15T01:15:30+03:00 Add STG debug from JS Sinker - - - - - 1 changed file: - compiler/GHC/Driver/Session.hs Changes: ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -1516,7 +1516,7 @@ dynamic_flags_deps = [ , make_ord_flag defGhcFlag "ddump-stg-tags" (setDumpFlag Opt_D_dump_stg_tags) , make_ord_flag defGhcFlag "ddump-stg-from-js-sinker" - (setDumpFlag Opt_D_dump_stg_tags) + (setDumpFlag Opt_D_dump_stg_from_js_sinker) , make_ord_flag defGhcFlag "ddump-call-arity" (setDumpFlag Opt_D_dump_call_arity) , make_ord_flag defGhcFlag "ddump-exitify" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ea4e6427efb57849fe6ef8faa6cd861bee15369f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ea4e6427efb57849fe6ef8faa6cd861bee15369f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 15 13:30:42 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Sun, 15 Sep 2024 09:30:42 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/ttg-booleanformula Message-ID: <66e6e1827c44_1869a6cd5834262dc@gitlab.mail> Hassan Al-Awwadi pushed new branch wip/ttg-booleanformula at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/ttg-booleanformula You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 09:22:49 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 16 Sep 2024 05:22:49 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 4 commits: DmdAnal: Fast path for `multDmdType` (#25196) Message-ID: <66e7f8e8ebea6_3892c31a753547591f@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 56d2523f by Sylvain Henry at 2024-09-16T05:22:31-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - cf40cfff by Samuel Thibault at 2024-09-16T05:22:37-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - 20 changed files: - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Types/Demand.hs - docs/users_guide/ghci.rst - libraries/array - libraries/ghc-internal/src/GHC/Internal/System/Environment/ExecutablePath.hsc - rts/Linker.c - rts/LinkerInternals.h - rts/RtsSymbols.c - rts/RtsSymbols.h - rts/rts.cabal - + testsuite/tests/dmdanal/should_compile/T25196.hs - + testsuite/tests/dmdanal/should_compile/T25196_aux.hs - testsuite/tests/dmdanal/should_compile/all.T - testsuite/tests/ghci/linking/Makefile - + testsuite/tests/ghci/linking/T25155.hs - + testsuite/tests/ghci/linking/T25155.stdout - + testsuite/tests/ghci/linking/T25155_TH.hs - + testsuite/tests/ghci/linking/T25155_iserv.hs - + testsuite/tests/ghci/linking/T25155_iserv_main.c - testsuite/tests/ghci/linking/all.T Changes: ===================================== compiler/GHC/Core/Opt/DmdAnal.hs ===================================== @@ -426,7 +426,7 @@ dmdAnalStar env (n :* sd) e , n' <- anticipateANF e n -- See Note [Anticipating ANF in demand analysis] -- and Note [Analysing with absent demand] - = (discardArgDmds $ multDmdType n' dmd_ty, e') + = (multDmdEnv n' (discardArgDmds dmd_ty), e') -- Main Demand Analysis machinery dmdAnal, dmdAnal' :: AnalEnv ===================================== compiler/GHC/Types/Demand.hs ===================================== @@ -51,7 +51,7 @@ module GHC.Types.Demand ( -- * Demand environments DmdEnv(..), addVarDmdEnv, mkTermDmdEnv, nopDmdEnv, plusDmdEnv, plusDmdEnvs, - reuseEnv, + multDmdEnv, reuseEnv, -- * Demand types DmdType(..), dmdTypeDepth, @@ -1910,7 +1910,8 @@ splitDmdTy ty at DmdType{dt_args=dmd:args} = (dmd, ty{dt_args=args}) splitDmdTy ty at DmdType{dt_env=env} = (defaultArgDmd (de_div env), ty) multDmdType :: Card -> DmdType -> DmdType -multDmdType n (DmdType fv args) +multDmdType C_11 dmd_ty = dmd_ty -- a vital optimisation for T25196 +multDmdType n (DmdType fv args) = -- pprTrace "multDmdType" (ppr n $$ ppr fv $$ ppr (multDmdEnv n fv)) $ DmdType (multDmdEnv n fv) (strictMap (multDmd n) args) ===================================== docs/users_guide/ghci.rst ===================================== @@ -3487,10 +3487,48 @@ dynamically-linked) from GHC itself. So for example: This feature is experimental in GHC 8.0.x, but it may become the default in future releases. +Building an external interpreter +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +The source code for the external interpreter program is in `utils/iserv`. It is +very simple because most of the heavy lifting code is from the `ghci` library. + +It is sometimes desirable to customize the external interpreter program. For +example, it is possible to add symbols to the RTS linker used by the external +interpreter. This is done simply at link time by linking an additional `.o` that +defines a `rtsExtraSyms` function returning the extra symbols. Doing it this way +avoids the need to recompile the RTS with symbols added to its built-in list. +A typical C file would look like this: + +.. code:: C + + #include + + #define CODE_SYM(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), \ + (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_CODE }, + #define DATA_SYM(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), \ + (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_DATA }, + + RtsSymbolVal my_iserv_syms[] = { + CODE_SYM(malloc) + CODE_SYM(getauxval) + CODE_SYM(posix_spawn_file_actions_init) + ... + { 0, 0, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ + }; + + RtsSymbolVal* rtsExtraSyms() { + return my_iserv_syms; + } + +For more information, read the Note [Extra RTS symbols] in the RTS. + + + .. _external-interpreter-proxy: Running the interpreter on a different host -------------------------------------------- +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ When using the flag :ghc-flag:`-fexternal-interpreter` GHC will spawn and communicate with the separate process using pipes. There @@ -3520,6 +3558,7 @@ will be executed on the target. As such packages like ``git-embed``, ``file-embed`` and others might not behave as expected if the target and host do not share the same filesystem. + .. _building-ghci-libraries: Building GHCi libraries ===================================== libraries/array ===================================== @@ -1 +1 @@ -Subproject commit ba5e9dcf1370190239395b8361b1c92ea9fc7632 +Subproject commit c9cb2c1e8762aa83b6e77af82c87a55e03e990e4 ===================================== libraries/ghc-internal/src/GHC/Internal/System/Environment/ExecutablePath.hsc ===================================== @@ -200,9 +200,9 @@ executablePath = Just (fmap Just getExecutablePath `catch` f) | otherwise = throw e -------------------------------------------------------------------------------- --- Linux / Solaris +-- Linux / Solaris / Hurd -#elif defined(linux_HOST_OS) || defined(solaris2_HOST_OS) +#elif defined(linux_HOST_OS) || defined(solaris2_HOST_OS) || defined(gnu_HOST_OS) foreign import ccall unsafe "readlink" c_readlink :: CString -> CString -> CSize -> IO CInt @@ -219,7 +219,7 @@ readSymbolicLink file = c_readlink s buf 4096 peekFilePathLen (buf,fromIntegral len) -# if defined(linux_HOST_OS) +# if defined(linux_HOST_OS) || defined(gnu_HOST_OS) getExecutablePath = readSymbolicLink $ "/proc/self/exe" executablePath = Just (check <$> getExecutablePath) where ===================================== rts/Linker.c ===================================== @@ -472,14 +472,35 @@ initLinker_ (int retain_cafs) symhash = allocStrHashTable(); /* populate the symbol table with stuff from the RTS */ + IF_DEBUG(linker, debugBelch("populating linker symbol table with built-in RTS symbols\n")); for (const RtsSymbolVal *sym = rtsSyms; sym->lbl != NULL; sym++) { + IF_DEBUG(linker, debugBelch("initLinker: inserting rts symbol %s, %p\n", sym->lbl, sym->addr)); if (! ghciInsertSymbolTable(WSTR("(GHCi built-in symbols)"), symhash, sym->lbl, sym->addr, sym->strength, sym->type, NULL)) { barf("ghciInsertSymbolTable failed"); } - IF_DEBUG(linker, debugBelch("initLinker: inserting rts symbol %s, %p\n", sym->lbl, sym->addr)); } + IF_DEBUG(linker, debugBelch("done with built-in RTS symbols\n")); + + /* Add extra symbols. rtsExtraSyms() is a weakly defined symbol in the rts, + * that can be overrided by linking in an object with a corresponding + * definition later. This is useful to build an external-interpreter or some + * other process with extra symbols (mostly libc, or similar). + * See Note [Extra RTS symbols] + */ + IF_DEBUG(linker, debugBelch("populating linker symbol table with extra RTS symbols\n")); + if(rtsExtraSyms && rtsExtraSyms() != NULL) { + for(RtsSymbolVal *sym = rtsExtraSyms(); sym->lbl != NULL; sym++) { + IF_DEBUG(linker, debugBelch("initLinker: inserting extra rts symbol %s, %p\n", sym->lbl, sym->addr)); + if (! ghciInsertSymbolTable(WSTR("(GHCi built-in symbols)"), + symhash, sym->lbl, sym->addr, + sym->strength, sym->type, NULL)) { + barf("ghciInsertSymbolTable failed"); + } + } + } + IF_DEBUG(linker, debugBelch("done with extra RTS symbols\n")); // Redirect newCAF to newRetainedCAF if retain_cafs is true. if (! ghciInsertSymbolTable(WSTR("(GHCi built-in symbols)"), symhash, ===================================== rts/LinkerInternals.h ===================================== @@ -9,6 +9,7 @@ #pragma once #include "Rts.h" +#include "RtsSymbols.h" #include "Hash.h" #include "linker/M32Alloc.h" @@ -33,8 +34,6 @@ void printLoadedObjects(void); # define OBJFORMAT_WASM32 #endif -typedef void SymbolAddr; -typedef char SymbolName; typedef struct _ObjectCode ObjectCode; typedef struct _Section Section; @@ -53,22 +52,6 @@ typedef struct _Section Section; * to recompile their program as position-independent. */ -/* What kind of thing a symbol identifies. We need to know this to determine how - * to process overflowing relocations. See Note [Processing overflowed relocations]. - * This is bitfield however only the option SYM_TYPE_DUP_DISCARD can be combined - * with the other values. */ -typedef enum _SymType { - SYM_TYPE_CODE = 1 << 0, /* the symbol is a function and can be relocated via a jump island */ - SYM_TYPE_DATA = 1 << 1, /* the symbol is data */ - SYM_TYPE_INDIRECT_DATA = 1 << 2, /* see Note [_iob_func symbol] */ - SYM_TYPE_DUP_DISCARD = 1 << 3, /* the symbol is a symbol in a BFD import library - however if a duplicate is found with a mismatching - SymType then discard this one. */ - SYM_TYPE_HIDDEN = 1 << 4, /* the symbol is hidden and should not be exported */ - -} SymType; - - #if defined(OBJFORMAT_ELF) # include "linker/ElfTypes.h" #elif defined(OBJFORMAT_PEi386) @@ -425,12 +408,6 @@ typedef void (*init_t) (int argc, char **argv, char **env); /* Type of a finalizer */ typedef void (*fini_t) (void); -typedef enum _SymStrength { - STRENGTH_NORMAL, - STRENGTH_WEAK, - STRENGTH_STRONG, -} SymStrength; - /* SymbolInfo tracks a symbol's address, the object code from which it originated, and whether or not it's weak. ===================================== rts/RtsSymbols.c ===================================== @@ -1090,3 +1090,104 @@ RtsSymbolVal rtsSyms[] = { SymI_HasDataProto(nonmoving_write_barrier_enabled) { 0, 0, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ }; + + +// Note [Extra RTS symbols] +// ~~~~~~~~~~~~~~~~~~~~~~~~ +// How does the RTS linker know the location of the exposed RTS functions in the +// current Haskell process? Especially if it is statically linked, there is no +// dynamic table containing a symbol->address mapping. What we do is that we +// compile the following static "rtsSyms" array: +// +// RtsSymbolVar rtsSyms[] = { +// { "rts_foo", &rts_foo, ...} +// ... +// }; +// +// This array associates a predefined set of symbol names (e.g. "rts_foo") to +// their address (e.g. &rts_foo) as determined by the linker (static or +// dynamic). +// +// Note that also use this mechanism for libraries other than the RTS, e.g. +// libc. +// +// Why do we need this mapping? +// ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +// Suppose we have a Haskell program statically linked against the "libfoo" +// library and using the "foo" symbol. If we try to dynamically link some code +// using "foo" with the RTS linker, we will load another instance of "libfoo" in +// memory and use the "foo" symbol in it for the dynamically loaded code. +// However the statically linked code still uses the statically linked instance +// of libfoo and its foo symbol. +// +// This isn't a problem as long a we use side-effect free functions. However +// both instances of a "foo" function using global variables may use different +// global variables! +// +// This was a real issue with programs linked against the ghc library and +// also loading it dynamically: the GHC lib uses global variables (e.g. for the +// unique counter), and those weren't shared... This wasn't solved by adding +// GHC's global variables to the rtsSyms list but it could have been (instead a +// generic mechanism to explicit manage Haskell global variables through the RTS +// was added and used). However for other libraries, especially foreign ones, +// the issue remains! +// +// Which symbols should go into the list? +// ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +// Ideally all problematic symbols (wrt to the double loading issue mentioned +// above) should go into the list. However: +// - we can't easily predict all the symbols that will be problematic/needed +// - symbols in the list prevent some dead code elimination! If symbol "foo" +// is in the list, even if we never load any code using "foo", we'll carry the +// code for "foo" and its dependencies. +// +// The second point is particularly problematic if some symbols are only needed +// to run Template Haskell splices at compilation time: we need them in the RTS +// used by the interpreter but not in the RTS of the target program! +// +// Extra symbols +// ~~~~~~~~~~~~~ +// Since #25155 we also support specifying a second list of "extra" symbols +// which is empty by default. By overriding the weak `rtsExtraSyms` function, +// build tools can extend the list of symbols linked into the RTS. This feature +// coupled with the external-interpreter feature allows specific external +// interpreter programs to be built easily with extra symbols supported. +// +// As a concrete example, one can build an `iserv-proxy-interpreter` program +// with the following additional Cabal stanza (see [1]): +// +// if os(linux) && arch(aarch64) +// c-sources: cbits/symbols.aarch64-musl.c +// +// Where the C file contains (see [2] and [3]): +// +// #include +// +// #define CODE_SYM(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_CODE }, +// #define DATA_SYM(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_DATA }, +// +// RtsSymbolVal my_iserv_syms[] = { +// CODE_SYM(malloc) +// CODE_SYM(getauxval) +// CODE_SYM(posix_spawn_file_actions_init) +// ... +// { 0, 0, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ +// }; +// +// RtsSymbolVal* rtsExtraSyms() { +// return my_iserv_syms; +// } +// +// [1] https://github.com/stable-haskell/iserv-proxy/blob/2ed34002247213fc435d0062350b91bab920626e/iserv-proxy.cabal#L110-L111 +// [2] https://github.com/stable-haskell/iserv-proxy/blob/2ed34002247213fc435d0062350b91bab920626e/cbits/symbols.aarch64-musl.c +// [3] https://github.com/stable-haskell/iserv-proxy/blob/2ed34002247213fc435d0062350b91bab920626e/cbits/symbols.aarch64-musl.h +// + +static RtsSymbolVal default_extra_syms[] = { + { 0, 0, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ +}; + +/* Default empty extra RTS symbols. See Note[Extra RTS symbols] */ +RtsSymbolVal* __attribute__((weak)) rtsExtraSyms(void) { + return default_extra_syms; +} ===================================== rts/RtsSymbols.h ===================================== @@ -9,8 +9,6 @@ #pragma once #include "ghcautoconf.h" -#include "LinkerInternals.h" -#include #if defined(LEADING_UNDERSCORE) #define MAYBE_LEADING_UNDERSCORE_STR(s) ("_" s) @@ -18,6 +16,29 @@ #define MAYBE_LEADING_UNDERSCORE_STR(s) (s) #endif +typedef void SymbolAddr; +typedef char SymbolName; + +/* What kind of thing a symbol identifies. We need to know this to determine how + * to process overflowing relocations. See Note [Processing overflowed relocations]. + * This is bitfield however only the option SYM_TYPE_DUP_DISCARD can be combined + * with the other values. */ +typedef enum _SymType { + SYM_TYPE_CODE = 1 << 0, /* the symbol is a function and can be relocated via a jump island */ + SYM_TYPE_DATA = 1 << 1, /* the symbol is data */ + SYM_TYPE_INDIRECT_DATA = 1 << 2, /* see Note [_iob_func symbol] */ + SYM_TYPE_DUP_DISCARD = 1 << 3, /* the symbol is a symbol in a BFD import library + however if a duplicate is found with a mismatching + SymType then discard this one. */ + SYM_TYPE_HIDDEN = 1 << 4, /* the symbol is hidden and should not be exported */ +} SymType; + +typedef enum _SymStrength { + STRENGTH_NORMAL, + STRENGTH_WEAK, + STRENGTH_STRONG, +} SymStrength; + typedef struct _RtsSymbolVal { const SymbolName* lbl; SymbolAddr* addr; @@ -27,6 +48,8 @@ typedef struct _RtsSymbolVal { extern RtsSymbolVal rtsSyms[]; +extern RtsSymbolVal* __attribute__((weak)) rtsExtraSyms(void); + /* See Note [_iob_func symbol]. */ #if defined(mingw32_HOST_OS) extern const void* __rts_iob_func; ===================================== rts/rts.cabal ===================================== @@ -250,7 +250,7 @@ library include-dirs: include includes: Rts.h autogen-includes: ghcautoconf.h ghcplatform.h - install-includes: Cmm.h HsFFI.h MachDeps.h Rts.h RtsAPI.h Stg.h + install-includes: Cmm.h HsFFI.h MachDeps.h Rts.h RtsAPI.h RtsSymbols.h Stg.h ghcautoconf.h ghcconfig.h ghcplatform.h ghcversion.h -- ^ from include DerivedConstants.h ===================================== testsuite/tests/dmdanal/should_compile/T25196.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25196 where + +import T25196_aux + +bar = $(gen 10000) ===================================== testsuite/tests/dmdanal/should_compile/T25196_aux.hs ===================================== @@ -0,0 +1,10 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25196_aux where + +import Language.Haskell.TH +import Language.Haskell.TH.Syntax +import Language.Haskell.TH.Lib + +gen :: Int -> Q Exp +gen n = lamE (replicate n (newName "x" >>= varP)) [| () |] ===================================== testsuite/tests/dmdanal/should_compile/all.T ===================================== @@ -98,3 +98,5 @@ test('T22388', [ grep_errmsg(r'^\S+\$w\S+') ], compile, ['-dsuppress-uniques -dd test('T22997', normal, compile, ['']) test('T23398', normal, compile, ['-dsuppress-uniques -ddump-simpl -dno-typeable-binds']) test('T24623', normal, compile, ['']) +test('T25196', [ req_th, collect_compiler_stats('bytes allocated', 10) ], + multimod_compile, ['T25196', '-v0']) ===================================== testsuite/tests/ghci/linking/Makefile ===================================== @@ -145,3 +145,8 @@ T15729: big-obj: '$(TEST_CC)' -c -Wa,-mbig-obj big-obj-c.c -o big-obj-c.o echo "main" | '$(TEST_HC)' $(TEST_HC_OPTS_INTERACTIVE) big-obj-c.o big-obj.hs + +.PHONY: T25155 +T25155: + '$(TEST_HC)' T25155_iserv_main.c T25155_iserv.hs -package ghci -no-hs-main -v0 -o T25155_iserv + '$(TEST_HC)' -fexternal-interpreter -pgmi ./T25155_iserv -v0 T25155.hs ===================================== testsuite/tests/ghci/linking/T25155.hs ===================================== @@ -0,0 +1,8 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25155 where + +import Language.Haskell.TH (runIO) +import T25155_TH + +$(runIO (foobar 7) >> pure []) ===================================== testsuite/tests/ghci/linking/T25155.stdout ===================================== @@ -0,0 +1 @@ +Called foobar with 7 ===================================== testsuite/tests/ghci/linking/T25155_TH.hs ===================================== @@ -0,0 +1,3 @@ +module T25155_TH (foobar) where + +foreign import ccall "foobar" foobar :: Int -> IO Int ===================================== testsuite/tests/ghci/linking/T25155_iserv.hs ===================================== @@ -0,0 +1,6 @@ +module Main (main) where + +import GHCi.Server (defaultServer) + +main :: IO () +main = defaultServer ===================================== testsuite/tests/ghci/linking/T25155_iserv_main.c ===================================== @@ -0,0 +1,41 @@ +#include +#include +#include +#include +#include + +int main (int argc, char *argv[]) +{ + RtsConfig conf = defaultRtsConfig; + + // We never know what symbols GHC will look up in the future, so + // we must retain CAFs for running interpreted code. + conf.keep_cafs = 1; + + conf.rts_opts_enabled = RtsOptsAll; + extern StgClosure ZCMain_main_closure; + hs_main(argc, argv, &ZCMain_main_closure, conf); +} + +// Inject "foobar" in the linker symbol table. +// +// The target object isn't compiled against any object defining foobar, yet we +// can use the FFI call to foobar in a TH splice. + +#define SymI_HasProto(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), \ + (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_CODE }, + + +void foobar(int x) { + printf("Called foobar with %d\n", x); +} + +RtsSymbolVal extra_syms[] = { + SymI_HasProto(foobar) + { NULL, NULL, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ + }; + +/* Extend the list of built-in symbols */ +RtsSymbolVal* rtsExtraSyms() { + return extra_syms; +} ===================================== testsuite/tests/ghci/linking/all.T ===================================== @@ -76,3 +76,4 @@ test('big-obj', [extra_files(['big-obj-c.c', 'big-obj.hs']), unless(doing_ghci, skip), unless(opsys('mingw32'), skip)], makefile_test, ['big-obj']) +test('T25155', [req_c, req_th, req_interp, extra_files(['T25155_iserv_main.c', 'T25155_iserv.hs', 'T25155_TH.hs'])], makefile_test, []) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b49daf5e49670ad097aa3a17caacac8f52355ad9...cf40cfff06017ad24ef4dc3bafbd37606a5b5c28 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b49daf5e49670ad097aa3a17caacac8f52355ad9...cf40cfff06017ad24ef4dc3bafbd37606a5b5c28 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 09:54:22 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 16 Sep 2024 05:54:22 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] Refactor FieldOcc vs AmbiguousFieldOcc with TTG Message-ID: <66e8004ed9de1_eeb08152f84500fc@gitlab.mail> Rodrigo Mesquita pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 5e11f34d by Jade at 2024-09-16T10:53:51+01:00 Refactor FieldOcc vs AmbiguousFieldOcc with TTG Improves the design of FieldOcc vs AmbiguousFieldOcc, and removes a dependency on `RdrName` from the Language.Haskell.* namespace (#21592). The design: * The FieldOcc constructor of FieldOcc always refers to an unambiguous field occurrence. * During renaming, a FieldOcc may be ambiguous and only be resolvable during Typechecking * Therefore, we extend (with TTG) `FieldOcc GhcRn` with a constructor `AmbiguousFieldOcc` that constructs a definitely ambiguous `FieldOcc`. * During typechecking, all ambiguous field occurrences must be resolved, so the `AmbiguousFieldOcc` constructor no longer exists See Note [Lifecycle of a FieldOcc] Co-authored-by: Rodrigo Mesquita <rodrigo.m.mesquita at gmail.com> - - - - - 20 changed files: - compiler/GHC/Hs/Binds.hs - compiler/GHC/Hs/Extension.hs - compiler/GHC/Hs/Instances.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Docs.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Fixity.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Tc/Types/Origin.hs - compiler/GHC/Tc/Zonk/Type.hs - compiler/GHC/ThToHs.hs - compiler/Language/Haskell/Syntax/Extension.hs - compiler/Language/Haskell/Syntax/Pat.hs - compiler/Language/Haskell/Syntax/Type.hs Changes: ===================================== compiler/GHC/Hs/Binds.hs ===================================== @@ -49,7 +49,6 @@ import GHC.Types.SrcLoc as SrcLoc import GHC.Types.Var import GHC.Data.Bag import GHC.Data.BooleanFormula (LBooleanFormula) -import GHC.Types.Name.Reader import GHC.Types.Name import GHC.Utils.Outputable @@ -653,7 +652,7 @@ pprTicks pp_no_debug pp_when_debug then pp_when_debug else pp_no_debug -instance Outputable (XRec a RdrName) => Outputable (RecordPatSynField a) where +instance Outputable (XRec pass (IdP pass)) => Outputable (RecordPatSynField pass) where ppr (RecordPatSynField { recordPatSynField = v }) = ppr v ===================================== compiler/GHC/Hs/Extension.hs ===================================== @@ -240,4 +240,4 @@ instance Outputable NoExtField where ppr _ = text "NoExtField" instance Outputable DataConCantHappen where - ppr = dataConCantHappen \ No newline at end of file + ppr = dataConCantHappen ===================================== compiler/GHC/Hs/Instances.hs ===================================== @@ -544,12 +544,7 @@ deriving instance Data (ConDeclField GhcTc) deriving instance Data (FieldOcc GhcPs) deriving instance Data (FieldOcc GhcRn) deriving instance Data (FieldOcc GhcTc) - --- deriving instance DataId p => Data (AmbiguousFieldOcc p) -deriving instance Data (AmbiguousFieldOcc GhcPs) -deriving instance Data (AmbiguousFieldOcc GhcRn) -deriving instance Data (AmbiguousFieldOcc GhcTc) - +deriving instance Data AmbiguousFieldOcc -- deriving instance (DataId name) => Data (ImportDecl name) deriving instance Data (ImportDecl GhcPs) ===================================== compiler/GHC/Hs/Pat.hs ===================================== @@ -37,7 +37,6 @@ module GHC.Hs.Pat ( HsRecUpdField, LHsRecUpdField, RecFieldsDotDot(..), hsRecFields, hsRecFieldSel, hsRecFieldId, hsRecFieldsArgs, - hsRecUpdFieldId, hsRecUpdFieldOcc, hsRecUpdFieldRdr, mkPrefixConPat, mkCharLitPat, mkNilPat, @@ -335,16 +334,6 @@ data ConPatTc hsRecFieldId :: HsRecField GhcTc arg -> Id hsRecFieldId = hsRecFieldSel -hsRecUpdFieldRdr :: HsRecUpdField (GhcPass p) q -> Located RdrName -hsRecUpdFieldRdr = fmap ambiguousFieldOccRdrName . reLoc . hfbLHS - -hsRecUpdFieldId :: HsFieldBind (LAmbiguousFieldOcc GhcTc) arg -> Located Id -hsRecUpdFieldId = fmap foExt . reLoc . hsRecUpdFieldOcc - -hsRecUpdFieldOcc :: HsFieldBind (LAmbiguousFieldOcc GhcTc) arg -> LFieldOcc GhcTc -hsRecUpdFieldOcc = fmap unambiguousFieldOcc . hfbLHS - - {- ************************************************************************ * * ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -102,7 +102,7 @@ lhsExprType (L _ e) = hsExprType e hsExprType :: HsExpr GhcTc -> Type hsExprType (HsVar _ (L _ id)) = idType id hsExprType (HsUnboundVar (HER _ ty _) _) = ty -hsExprType (HsRecSel _ (FieldOcc id _)) = idType id +hsExprType (HsRecSel _ (FieldOcc _ id)) = idType (unLoc id) hsExprType (HsOverLabel v _ _) = dataConCantHappen v hsExprType (HsIPVar v _) = dataConCantHappen v hsExprType (HsOverLit _ lit) = overLitType lit ===================================== compiler/GHC/Hs/Type.hs ===================================== @@ -58,10 +58,9 @@ module GHC.Hs.Type ( HsConDetails(..), noTypeArgs, FieldOcc(..), LFieldOcc, mkFieldOcc, - AmbiguousFieldOcc(..), LAmbiguousFieldOcc, mkAmbiguousFieldOcc, - ambiguousFieldOccRdrName, ambiguousFieldOccLRdrName, - selectorAmbiguousFieldOcc, - unambiguousFieldOcc, ambiguousFieldOcc, + fieldOccRdrName, fieldOccLRdrName, + AmbiguousFieldOcc(..), + ambiguousFieldOccRdrName, OpName(..), @@ -108,7 +107,6 @@ import GHC.Hs.Extension import GHC.Parser.Annotation import GHC.Types.Fixity ( LexicalFixity(..) ) -import GHC.Types.Id ( Id ) import GHC.Types.SourceText import GHC.Types.Name import GHC.Types.Name.Reader ( RdrName ) @@ -1040,59 +1038,94 @@ also forbids them in types involved with `deriving`: FieldOcc * * ************************************************************************ --} -type instance XCFieldOcc GhcPs = NoExtField -type instance XCFieldOcc GhcRn = Name -type instance XCFieldOcc GhcTc = Id +Note [Lifecycle of a FieldOcc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +A field occurrence (FieldOcc) has a slightly special lifecycle because field +occurrences may be deemed /ambiguous/ during renaming. Ambiguous field +occurrences can only be resolved during typechecking (since we do type-directed +disambiguation). To accommodate the fact that we may be unable to produce a +`Name` for a `FieldOcc` during renaming, `FieldOcc` is extended with +`AmbiguousFieldOcc` during renaming. Here's the life cycle: -type instance XXFieldOcc (GhcPass _) = DataConCantHappen +* The `FieldOcc` constructor of the `FieldOcc` type always refers to an + unambiguous field occurrence. We parse field occurrences into `FieldOcc`. -mkFieldOcc :: LocatedN RdrName -> FieldOcc GhcPs -mkFieldOcc rdr = FieldOcc noExtField rdr +* During renaming, a FieldOcc may be ambiguous and only be resolvable + during typechecking. To construct an /ambiguous/ `FieldOcc GhcRn`, we use the + extension point for constructors, which is instanced with `AmbiguousFieldOcc`: + (XFieldOcc . Ambiguous) :: FieldOcc GhcRn -type instance XUnambiguous GhcPs = NoExtField -type instance XUnambiguous GhcRn = Name -type instance XUnambiguous GhcTc = Id +* During typechecking, all ambiguous field occurrences must be resolved. We + statically guarantee this by making it impossible to construct a `FieldOcc + GhcTc` with an ambiguous name: -type instance XAmbiguous GhcPs = NoExtField -type instance XAmbiguous GhcRn = NoExtField -type instance XAmbiguous GhcTc = Id + type instance XXFieldOcc GhcTc = DataConCantHappen -type instance XXAmbiguousFieldOcc (GhcPass _) = DataConCantHappen +Note that throughout this lifecycle, we preserve the `RdrName` the user +originally wrote in the extension point during renaming and typechecking. +-} -instance Outputable (AmbiguousFieldOcc (GhcPass p)) where - ppr = ppr . ambiguousFieldOccRdrName +-- | Ambiguous Field Occurrence +-- +-- Represents an *occurrence* of a field that is definiely +-- ambiguous after the renamer, with the ambiguity resolved by the +-- typechecker. We always store the 'RdrName' that the user +-- originally wrote, and store the selector function after the typechecker (for +-- ambiguous occurrences). +-- +-- Unambiguous field occurrences should be stored in the proper FieldOcc datacon of FieldOcc. +-- +-- See Note [HsRecField and HsRecUpdField] in "GHC.Hs.Pat". +-- See Note [Located RdrNames] in "GHC.Hs.Expr". +newtype AmbiguousFieldOcc + = Ambiguous (LocatedN RdrName) -instance OutputableBndr (AmbiguousFieldOcc (GhcPass p)) where - pprInfixOcc = pprInfixOcc . ambiguousFieldOccRdrName - pprPrefixOcc = pprPrefixOcc . ambiguousFieldOccRdrName +type instance XCFieldOcc GhcPs = NoExtField -- RdrName is stored in the proper IdP field +type instance XCFieldOcc GhcRn = RdrName +type instance XCFieldOcc GhcTc = RdrName -instance OutputableBndr (Located (AmbiguousFieldOcc (GhcPass p))) where - pprInfixOcc = pprInfixOcc . unLoc - pprPrefixOcc = pprPrefixOcc . unLoc +type instance XXFieldOcc GhcPs = DataConCantHappen +type instance XXFieldOcc GhcRn = AmbiguousFieldOcc +type instance XXFieldOcc GhcTc = DataConCantHappen -mkAmbiguousFieldOcc :: LocatedN RdrName -> AmbiguousFieldOcc GhcPs -mkAmbiguousFieldOcc rdr = Unambiguous noExtField rdr +-------------------------------------------------------------------------------- -ambiguousFieldOccRdrName :: AmbiguousFieldOcc (GhcPass p) -> RdrName -ambiguousFieldOccRdrName = unLoc . ambiguousFieldOccLRdrName +mkFieldOcc :: LocatedN RdrName -> FieldOcc GhcPs +mkFieldOcc rdr = FieldOcc noExtField rdr -ambiguousFieldOccLRdrName :: AmbiguousFieldOcc (GhcPass p) -> LocatedN RdrName -ambiguousFieldOccLRdrName (Unambiguous _ rdr) = rdr -ambiguousFieldOccLRdrName (Ambiguous _ rdr) = rdr +fieldOccRdrName :: forall p. IsPass p => FieldOcc (GhcPass p) -> RdrName +fieldOccRdrName fo = case ghcPass @p of + GhcPs -> unLoc $ foLabel fo + GhcRn -> foExt fo + GhcTc -> foExt fo + +fieldOccLRdrName :: forall p. IsPass p => FieldOcc (GhcPass p) -> LocatedN RdrName +fieldOccLRdrName fo = case ghcPass @p of + GhcPs -> foLabel fo + GhcRn -> case fo of + FieldOcc rdr sel -> + let (L l _) = sel + in L l rdr + XFieldOcc (Ambiguous l) -> l + GhcTc -> + let (L l _) = foLabel fo + in L l (foExt fo) + +instance Outputable AmbiguousFieldOcc where + ppr = ppr . ambiguousFieldOccRdrName -selectorAmbiguousFieldOcc :: AmbiguousFieldOcc GhcTc -> Id -selectorAmbiguousFieldOcc (Unambiguous sel _) = sel -selectorAmbiguousFieldOcc (Ambiguous sel _) = sel +instance OutputableBndr AmbiguousFieldOcc where + pprInfixOcc = pprInfixOcc . ambiguousFieldOccRdrName + pprPrefixOcc = pprPrefixOcc . ambiguousFieldOccRdrName -unambiguousFieldOcc :: AmbiguousFieldOcc GhcTc -> FieldOcc GhcTc -unambiguousFieldOcc (Unambiguous rdr sel) = FieldOcc rdr sel -unambiguousFieldOcc (Ambiguous rdr sel) = FieldOcc rdr sel +instance OutputableBndr (Located AmbiguousFieldOcc) where + pprInfixOcc = pprInfixOcc . unLoc + pprPrefixOcc = pprPrefixOcc . unLoc -ambiguousFieldOcc :: FieldOcc GhcTc -> AmbiguousFieldOcc GhcTc -ambiguousFieldOcc (FieldOcc sel rdr) = Unambiguous sel rdr +ambiguousFieldOccRdrName :: AmbiguousFieldOcc -> RdrName +ambiguousFieldOccRdrName (Ambiguous rdr) = unLoc rdr {- ************************************************************************ @@ -1210,14 +1243,14 @@ instance (Outputable tyarg, Outputable arg, Outputable rec) ppr (RecCon rec) = text "RecCon:" <+> ppr rec ppr (InfixCon l r) = text "InfixCon:" <+> ppr [l, r] -instance Outputable (XRec pass RdrName) => Outputable (FieldOcc pass) where +instance Outputable (XRec pass (IdP pass)) => Outputable (FieldOcc pass) where ppr = ppr . foLabel -instance (UnXRec pass, OutputableBndr (XRec pass RdrName)) => OutputableBndr (FieldOcc pass) where - pprInfixOcc = pprInfixOcc . unXRec @pass . foLabel - pprPrefixOcc = pprPrefixOcc . unXRec @pass . foLabel +instance (OutputableBndrId pass) => OutputableBndr (FieldOcc (GhcPass pass)) where + pprInfixOcc = pprInfixOcc . unXRec @(GhcPass pass) . foLabel + pprPrefixOcc = pprPrefixOcc . unXRec @(GhcPass pass) . foLabel -instance (UnXRec pass, OutputableBndr (XRec pass RdrName)) => OutputableBndr (GenLocated SrcSpan (FieldOcc pass)) where +instance (OutputableBndrId pass) => OutputableBndr (GenLocated SrcSpan (FieldOcc (GhcPass pass))) where pprInfixOcc = pprInfixOcc . unLoc pprPrefixOcc = pprPrefixOcc . unLoc @@ -1282,7 +1315,7 @@ pprLHsContextAlways (L _ ctxt) [L _ ty] -> ppr_mono_ty ty <+> darrow _ -> parens (interpp'SP ctxt) <+> darrow -pprConDeclFields :: OutputableBndrId p +pprConDeclFields :: forall p. OutputableBndrId p => [LConDeclField (GhcPass p)] -> SDoc pprConDeclFields fields = braces (sep (punctuate comma (map ppr_fld fields))) where @@ -1290,7 +1323,7 @@ pprConDeclFields fields = braces (sep (punctuate comma (map ppr_fld fields))) cd_fld_doc = doc })) = pprMaybeWithDoc doc (ppr_names ns <+> dcolon <+> ppr ty) - ppr_names :: [LFieldOcc (GhcPass p)] -> SDoc + ppr_names :: forall p. OutputableBndrId p => [LFieldOcc (GhcPass p)] -> SDoc ppr_names [n] = pprPrefixOcc n ppr_names ns = sep (punctuate comma (map pprPrefixOcc ns)) @@ -1531,4 +1564,4 @@ type instance Anno HsIPName = EpAnnCO type instance Anno (ConDeclField (GhcPass p)) = SrcSpanAnnA type instance Anno (FieldOcc (GhcPass p)) = SrcSpanAnnA -type instance Anno (AmbiguousFieldOcc (GhcPass p)) = SrcSpanAnnA +type instance Anno AmbiguousFieldOcc = SrcSpanAnnA ===================================== compiler/GHC/Hs/Utils.hs ===================================== @@ -1435,7 +1435,7 @@ hsTyClForeignBinders tycl_decls foreign_decls (foldMap (foldMap hsLInstDeclBinders . group_instds) tycl_decls)) where getSelectorNames :: ([LocatedA Name], [LFieldOcc GhcRn]) -> [Name] - getSelectorNames (ns, fs) = map unLoc ns ++ map (foExt . unLoc) fs + getSelectorNames (ns, fs) = map unLoc ns ++ map (unLoc . foLabel . unLoc) fs ------------------- @@ -1677,7 +1677,7 @@ emptyFieldIndices = , fieldIndices = Map.empty , newInt = 0 } -insertField :: LFieldOcc (GhcPass p) -> FieldIndices p -> (Located Int, FieldIndices p) +insertField :: IsPass p => LFieldOcc (GhcPass p) -> FieldIndices p -> (Located Int, FieldIndices p) insertField new_fld fi@(FieldIndices flds idxs new_idx) | Just i <- Map.lookup rdr idxs = (L loc i, fi) @@ -1688,7 +1688,7 @@ insertField new_fld fi@(FieldIndices flds idxs new_idx) (new_idx + 1)) where loc = getLocA new_fld - rdr = unLoc . foLabel . unLoc $ new_fld + rdr = fieldOccRdrName . unLoc $ new_fld {- @@ -1864,5 +1864,5 @@ rec_field_expl_impl rec_flds (RecFieldsDotDot { .. }) where (explicit_binds, implicit_binds) = splitAt unRecFieldsDotDot rec_flds implicit_field_binders (L _ (HsFieldBind { hfbLHS = L _ fld, hfbRHS = rhs })) = ImplicitFieldBinders - { implFlBndr_field = foExt fld + { implFlBndr_field = unLoc $ foLabel fld , implFlBndr_binders = collectPatBinders CollNoDictBinders rhs } ===================================== compiler/GHC/HsToCore/Docs.hs ===================================== @@ -394,7 +394,7 @@ subordinates env instMap decl = case decl of , maybeToList $ fmap unLoc $ con_doc c , conArgDocs c) | c <- toList cons, cname <- getConNames c ] - fields = [ (foExt n, maybeToList $ fmap unLoc doc, IM.empty) + fields = [ (unLoc $ foLabel n, maybeToList $ fmap unLoc doc, IM.empty) | Just flds <- toList $ fmap getRecConArgs_maybe cons , (L _ (ConDeclField _ ns _ doc)) <- (unLoc flds) , (L _ n) <- ns ] ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -477,7 +477,7 @@ addBinTickLHsExpr boxLabel e@(L pos e0) addTickHsExpr :: HsExpr GhcTc -> TM (HsExpr GhcTc) addTickHsExpr e@(HsVar _ (L _ id)) = do freeVar id; return e addTickHsExpr e@(HsUnboundVar {}) = return e -addTickHsExpr e@(HsRecSel _ (FieldOcc id _)) = do freeVar id; return e +addTickHsExpr e@(HsRecSel _ (FieldOcc _ id)) = do freeVar (unLoc id); return e addTickHsExpr e@(HsIPVar {}) = return e addTickHsExpr e@(HsOverLit {}) = return e ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -800,7 +800,7 @@ class ( HiePass (NoGhcTcPass p) , Data (Stmt (GhcPass p) (LocatedA (HsCmd (GhcPass p)))) , Data (HsExpr (GhcPass p)) , Data (HsCmd (GhcPass p)) - , Data (AmbiguousFieldOcc (GhcPass p)) + , Data AmbiguousFieldOcc , Data (HsCmdTop (GhcPass p)) , Data (GRHS (GhcPass p) (LocatedA (HsCmd (GhcPass p)))) , Data (HsUntypedSplice (GhcPass p)) @@ -1461,23 +1461,19 @@ instance ( ToHie (RFContext label) ] instance HiePass p => ToHie (RFContext (LocatedA (FieldOcc (GhcPass p)))) where - toHie (RFC c rhs (L nspan f)) = concatM $ case f of - FieldOcc fld _ -> - case hiePass @p of - HieRn -> [toHie $ C (RecField c rhs) (L (locA nspan) fld)] - HieTc -> [toHie $ C (RecField c rhs) (L (locA nspan) fld)] - -instance HiePass p => ToHie (RFContext (LocatedA (AmbiguousFieldOcc (GhcPass p)))) where - - toHie (RFC c rhs (L nspan afo)) = concatM $ case afo of - Unambiguous fld _ -> - case hiePass @p of - HieRn -> [toHie $ C (RecField c rhs) $ L (locA nspan) fld] - HieTc -> [toHie $ C (RecField c rhs) $ L (locA nspan) fld] - Ambiguous fld _ -> - case hiePass @p of - HieRn -> [] - HieTc -> [ toHie $ C (RecField c rhs) (L (locA nspan) fld) ] + toHie (RFC c rhs (L nspan f)) = concatM $ + case hiePass @p of + HieRn -> + case f of + FieldOcc _ fld -> + [toHie $ C (RecField c rhs) (L (locA nspan) $ unLoc fld)] + XFieldOcc (Ambiguous (L nspan fld)) + -> [ toHie $ C (RecField c rhs) (L (locA nspan) fld) ] + HieTc -> + case f of + FieldOcc _ fld -> + [toHie $ C (RecField c rhs) (L (locA nspan) $ unLoc fld)] + instance HiePass p => ToHie (RScoped (ApplicativeArg (GhcPass p))) where toHie (RS sc (ApplicativeArgOne _ pat expr _)) = concatM ===================================== compiler/GHC/Parser/PostProcess.hs ===================================== @@ -2665,7 +2665,7 @@ mkRdrRecordUpd overloaded_on exp@(L loc _) fbinds anns = do True -> do let qualifiedFields = [ L l lbl | L _ (HsFieldBind _ (L l lbl) _ _) <- fs' - , isQual . ambiguousFieldOccRdrName $ lbl + , isQual . fieldOccRdrName $ lbl ] case qualifiedFields of qf:_ -> addFatalError $ mkPlainErrorMsgEnvelope (getLocA qf) $ @@ -2711,7 +2711,7 @@ mk_rec_fields fs (Just s) = HsRecFields { rec_flds = fs mk_rec_upd_field :: HsRecField GhcPs (LHsExpr GhcPs) -> HsRecUpdField GhcPs GhcPs mk_rec_upd_field (HsFieldBind noAnn (L loc (FieldOcc _ rdr)) arg pun) - = HsFieldBind noAnn (L loc (Unambiguous noExtField rdr)) arg pun + = HsFieldBind noAnn (L loc (FieldOcc noExtField rdr)) arg pun mkInlinePragma :: SourceText -> (InlineSpec, RuleMatchInfo) -> Maybe Activation -> InlinePragma ===================================== compiler/GHC/Rename/Env.hs ===================================== @@ -1508,8 +1508,8 @@ lookupGlobalOccRn_overloaded rdr_name = addNameClashErrRn rdr_name gres return (Just gre) } -getFieldUpdLbl :: LHsRecUpdField (GhcPass p) q -> LocatedN RdrName -getFieldUpdLbl = ambiguousFieldOccLRdrName . unLoc . hfbLHS . unLoc +getFieldUpdLbl :: IsPass p => LHsRecUpdField (GhcPass p) q -> LocatedN RdrName +getFieldUpdLbl = fieldOccLRdrName . unLoc . hfbLHS . unLoc -- | Returns all possible collections of field labels for the given -- record update. @@ -1603,10 +1603,10 @@ lookupRecUpdFields flds * * **********************************************************************-} -getUpdFieldLbls :: forall p q. UnXRec (GhcPass p) +getUpdFieldLbls :: forall p q. IsPass p => [LHsRecUpdField (GhcPass p) q] -> [RdrName] getUpdFieldLbls - = map $ ambiguousFieldOccRdrName + = map $ fieldOccRdrName . unXRec @(GhcPass p) . hfbLHS . unXRec @(GhcPass p) ===================================== compiler/GHC/Rename/Fixity.hs ===================================== @@ -202,5 +202,6 @@ lookupFixityRn_help name lookupTyFixityRn :: LocatedN Name -> RnM Fixity lookupTyFixityRn = lookupFixityRn . unLoc -lookupFieldFixityRn :: FieldOcc GhcRn -> RnM Fixity -lookupFieldFixityRn (FieldOcc n _) = lookupFixityRn n +lookupFieldFixityRn :: FieldOcc GhcRn -> RnM (Maybe Fixity) +lookupFieldFixityRn (FieldOcc _ n) = Just <$> lookupFixityRn (unLoc n) +lookupFieldFixityRn _ = pure Nothing ===================================== compiler/GHC/Rename/HsType.hs ===================================== @@ -1302,7 +1302,7 @@ rnField fl_env env (L l (ConDeclField _ names ty haddock_doc)) lookupField :: FastStringEnv FieldLabel -> FieldOcc GhcPs -> FieldOcc GhcRn lookupField fl_env (FieldOcc _ (L lr rdr)) = - FieldOcc sel (L lr $ mkRdrUnqual $ occName sel) + FieldOcc rdr (L lr sel) where lbl = occNameFS $ rdrNameOcc rdr sel = flSelector @@ -1598,7 +1598,12 @@ lookupFixityOp :: OpName -> RnM Fixity lookupFixityOp (NormalOp n) = lookupFixityRn n lookupFixityOp NegateOp = lookupFixityRn negateName lookupFixityOp (UnboundOp u) = lookupFixityRn (mkUnboundName (occName u)) -lookupFixityOp (RecFldOp f) = lookupFieldFixityRn f +lookupFixityOp (RecFldOp f) = + -- We could get rid of this panic by parametrising FieldOcc with (k :: + -- MaybeAmbiguous), say 'PossiblyAmbiguous | 'Unambiguous, and using + -- 'PossiblyAmbiguous only for record field updates... but seems overkill + fromMaybe (panic "lookupFixityOp: RecFldOp should not be ambiguous!") <$> + lookupFieldFixityRn f -- Precedence-related error messages ===================================== compiler/GHC/Tc/Types/Origin.hs ===================================== @@ -717,7 +717,7 @@ exprCtOrigin :: HsExpr GhcRn -> CtOrigin exprCtOrigin (HsVar _ (L _ name)) = OccurrenceOf name exprCtOrigin (HsGetField _ _ (L _ f)) = HasFieldOrigin (field_label $ unLoc $ dfoLabel f) exprCtOrigin (HsUnboundVar {}) = Shouldn'tHappenOrigin "unbound variable" -exprCtOrigin (HsRecSel _ f) = OccurrenceOfRecSel (unLoc $ foLabel f) +exprCtOrigin (HsRecSel _ f) = OccurrenceOfRecSel (foExt f) exprCtOrigin (HsOverLabel _ _ l) = OverLabelOrigin l exprCtOrigin (ExplicitList {}) = ListOrigin exprCtOrigin (HsIPVar _ ip) = IPOccOrigin ip ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -607,8 +607,8 @@ zonkTopBndrs :: [TcId] -> TcM [Id] zonkTopBndrs ids = initZonkEnv DefaultFlexi $ zonkIdBndrs ids zonkFieldOcc :: FieldOcc GhcTc -> ZonkTcM (FieldOcc GhcTc) -zonkFieldOcc (FieldOcc sel lbl) - = fmap ((flip FieldOcc) lbl) $ zonkIdBndr sel +zonkFieldOcc (FieldOcc lbl (L l sel)) + = FieldOcc lbl . L l <$> zonkIdBndr sel zonkEvBndrsX :: [EvVar] -> ZonkBndrTcM [EvVar] zonkEvBndrsX = traverse zonkEvBndrX @@ -934,9 +934,9 @@ zonkExpr (HsUnboundVar her occ) ty' <- zonkTcTypeToTypeX ty return (HER ref ty' u) -zonkExpr (HsRecSel _ (FieldOcc v occ)) +zonkExpr (HsRecSel _ (FieldOcc occ (L l v))) = do { v' <- zonkIdOcc v - ; return (HsRecSel noExtField (FieldOcc v' occ)) } + ; return (HsRecSel noExtField (FieldOcc occ (L l v'))) } zonkExpr (HsIPVar x _) = dataConCantHappen x ===================================== compiler/GHC/ThToHs.hs ===================================== @@ -1147,7 +1147,7 @@ cvtl e = wrapLA (cvt e) ; return $ mkRdrRecordCon c' (HsRecFields flds' Nothing) noAnn } cvt (RecUpdE e flds) = do { e' <- cvtl e ; flds' - <- mapM (cvtFld (wrapParLA mkAmbiguousFieldOcc)) + <- mapM (cvtFld (wrapParLA mkFieldOcc)) flds ; return $ RecordUpd noAnn e' $ RegularRecUpdFields ===================================== compiler/Language/Haskell/Syntax/Extension.hs ===================================== @@ -460,13 +460,6 @@ type family XXDotFieldOcc x type family XSCC x type family XXPragE x - --- ------------------------------------- --- AmbiguousFieldOcc type families -type family XUnambiguous x -type family XAmbiguous x -type family XXAmbiguousFieldOcc x - -- ------------------------------------- -- HsTupArg type families type family XPresent x ===================================== compiler/Language/Haskell/Syntax/Pat.hs ===================================== @@ -323,7 +323,7 @@ type LHsRecUpdField p q = XRec p (HsRecUpdField p q) type HsRecField p arg = HsFieldBind (LFieldOcc p) arg -- | Haskell Record Update Field -type HsRecUpdField p q = HsFieldBind (LAmbiguousFieldOcc p) (LHsExpr q) +type HsRecUpdField p q = HsFieldBind (LFieldOcc p) (LHsExpr q) -- | Haskell Field Binding -- @@ -392,11 +392,11 @@ data HsFieldBind lhs rhs = HsFieldBind { -- -- See also Note [Disambiguating record updates] in GHC.Rename.Pat. -hsRecFields :: forall p arg.UnXRec p => HsRecFields p arg -> [XCFieldOcc p] +hsRecFields :: forall p arg.UnXRec p => HsRecFields p arg -> [IdP p] hsRecFields rbinds = Data.List.map (hsRecFieldSel . unXRec @p) (rec_flds rbinds) hsRecFieldsArgs :: forall p arg. UnXRec p => HsRecFields p arg -> [arg] hsRecFieldsArgs rbinds = Data.List.map (hfbRHS . unXRec @p) (rec_flds rbinds) -hsRecFieldSel :: forall p arg. UnXRec p => HsRecField p arg -> XCFieldOcc p -hsRecFieldSel = foExt . unXRec @p . hfbLHS +hsRecFieldSel :: forall p arg. UnXRec p => HsRecField p arg -> IdP p +hsRecFieldSel = unXRec @p . foLabel . unXRec @p . hfbLHS ===================================== compiler/Language/Haskell/Syntax/Type.hs ===================================== @@ -51,7 +51,6 @@ module Language.Haskell.Syntax.Type ( HsConDetails(..), noTypeArgs, conDetailsArity, FieldOcc(..), LFieldOcc, - AmbiguousFieldOcc(..), LAmbiguousFieldOcc, mapHsOuterImplicit, hsQTvExplicit, @@ -63,7 +62,6 @@ import {-# SOURCE #-} Language.Haskell.Syntax.Expr ( HsUntypedSplice ) import Language.Haskell.Syntax.Extension -import GHC.Types.Name.Reader ( RdrName ) import GHC.Core.DataCon( HsSrcBang(..) ) import GHC.Core.Type (Specificity) import GHC.Types.Basic (Arity) @@ -1286,38 +1284,20 @@ type LFieldOcc pass = XRec pass (FieldOcc pass) -- We store both the 'RdrName' the user originally wrote, and after -- the renamer we use the extension field to store the selector -- function. +-- +-- See Note [Lifecycle of a FieldOcc] data FieldOcc pass = FieldOcc { foExt :: XCFieldOcc pass - , foLabel :: XRec pass RdrName -- See Note [Located RdrNames] in Language.Haskell.Syntax.Expr + , foLabel :: LIdP pass } | XFieldOcc !(XXFieldOcc pass) deriving instance ( - Eq (XRec pass RdrName) + Eq (LIdP pass) , Eq (XCFieldOcc pass) , Eq (XXFieldOcc pass) ) => Eq (FieldOcc pass) --- | Located Ambiguous Field Occurrence -type LAmbiguousFieldOcc pass = XRec pass (AmbiguousFieldOcc pass) - --- | Ambiguous Field Occurrence --- --- Represents an *occurrence* of a field that is potentially --- ambiguous after the renamer, with the ambiguity resolved by the --- typechecker. We always store the 'RdrName' that the user --- originally wrote, and store the selector function after the renamer --- (for unambiguous occurrences) or the typechecker (for ambiguous --- occurrences). --- --- See Note [HsRecField and HsRecUpdField] in "GHC.Hs.Pat". --- See Note [Located RdrNames] in "GHC.Hs.Expr". -data AmbiguousFieldOcc pass - = Unambiguous (XUnambiguous pass) (XRec pass RdrName) - | Ambiguous (XAmbiguous pass) (XRec pass RdrName) - | XAmbiguousFieldOcc !(XXAmbiguousFieldOcc pass) - - {- ************************************************************************ * * View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5e11f34d6f5a089e5c4659bdf20e715750e08120 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5e11f34d6f5a089e5c4659bdf20e715750e08120 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 12:22:25 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 16 Sep 2024 08:22:25 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] TEMPORARY(DROP-ME): Try CI on object-determinism by default Message-ID: <66e823012a869_eeb086dfa6064586@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: bdec61b8 by Rodrigo Mesquita at 2024-09-16T13:22:18+01:00 TEMPORARY(DROP-ME): Try CI on object-determinism by default - - - - - 2 changed files: - compiler/GHC/Driver/Config/StgToCmm.hs - compiler/GHC/Driver/Main.hs Changes: ===================================== compiler/GHC/Driver/Config/StgToCmm.hs ===================================== @@ -53,7 +53,7 @@ initStgToCmmConfig dflags mod = StgToCmmConfig , stgToCmmExtDynRefs = gopt Opt_ExternalDynamicRefs dflags , stgToCmmDoBoundsCheck = gopt Opt_DoBoundsChecking dflags , stgToCmmDoTagCheck = gopt Opt_DoTagInferenceChecks dflags - , stgToCmmObjectDeterminism = gopt Opt_ObjectDeterminism dflags + , stgToCmmObjectDeterminism = not (gopt Opt_ObjectDeterminism dflags) -- backend flags: ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -2250,7 +2250,7 @@ doCodeGen hsc_env this_mod denv data_tycons -- Cmm was renamed. Update all the symbols by renaming them with -- the renaming map in that case. (_drn, rn_denv) - | gopt Opt_ObjectDeterminism dflags = detRenameIPEMap detRnEnv denv + | not (gopt Opt_ObjectDeterminism dflags) = detRenameIPEMap detRnEnv denv | otherwise = (detRnEnv, denv) cmmCgInfos <- generateCgIPEStub hsc_env this_mod rn_denv (nonCaffySet, lf_infos, ipes, ipe_stats) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/bdec61b8e5d7f83ce8ae722a83563366b2287b6f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/bdec61b8e5d7f83ce8ae722a83563366b2287b6f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 13:17:51 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Mon, 16 Sep 2024 09:17:51 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] 3 commits: wrk: remove typeclass in T25240 Message-ID: <66e82fff23c96_eeb08ae40fc87373@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 5d683e0f by doyougnu at 2024-09-16T08:36:06-04:00 wrk: remove typeclass in T25240 - - - - - 5101d05e by doyougnu at 2024-09-16T08:43:08-04:00 wrk: link-unknown-symbols -> optimistic-linking - - - - - 44eeff48 by doyougnu at 2024-09-16T09:17:28-04:00 wrk: add warnings - - - - - 9 changed files: - libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc - rts/Linker.c - rts/RtsFlags.c - rts/include/rts/Flags.h - rts/linker/elf_got.c - testsuite/tests/ghci/linking/T25240/FFI.hs - testsuite/tests/ghci/linking/T25240/Makefile - testsuite/tests/ghci/linking/T25240/T25240.hs - testsuite/tests/interface-stability/base-exports.stdout Changes: ===================================== libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc ===================================== @@ -162,7 +162,7 @@ data MiscFlags = MiscFlags , disableDelayedOsMemoryReturn :: Bool , internalCounters :: Bool , linkerAlwaysPic :: Bool - , linkUnknownSymbols :: Bool + , linkerOptimistic :: Bool , linkerMemBase :: Word -- ^ address to ask the OS for memory for the linker, 0 ==> off , ioManager :: IoManagerFlag @@ -538,7 +538,7 @@ getMiscFlags = do <*> (toBool <$> (#{peek MISC_FLAGS, linkerAlwaysPic} ptr :: IO CBool)) <*> (toBool <$> - (#{peek MISC_FLAGS, linkUnknownSymbols} ptr :: IO CBool)) + (#{peek MISC_FLAGS, linkerOptimistic} ptr :: IO CBool)) <*> #{peek MISC_FLAGS, linkerMemBase} ptr <*> (toEnum . fromIntegral <$> (#{peek MISC_FLAGS, ioManager} ptr :: IO Word32)) ===================================== rts/Linker.c ===================================== @@ -946,15 +946,18 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) // lookupDependentSymbol directly. SymbolAddr* r = lookupDependentSymbol(lbl, NULL, NULL); if (!r) { - if (!RtsFlags.MiscFlags.linkUnknownSymbols) { + if (!RtsFlags.MiscFlags.linkerOptimistic) { errorBelch("^^ Could not load '%s', dependency unresolved. " - "See top entry above.\n", + "See top entry above. You might consider using --optimistic-linking\n", lbl); IF_DEBUG(linker, printLoadedObjects()); fflush(stderr); } else { - // if --link-unknown-symbols is passed into the RTS we allow the linker + // if --optimistic-linking is passed into the RTS we allow the linker // to optimistically continue + errorBelch("^^ Could not load '%s', dependency unresolved, " + "optimistically continuing; assigning '%s' to 0xDEADBEEF.\n", + lbl, lbl); r = (void*) 0xDEADBEEF; } } ===================================== rts/RtsFlags.c ===================================== @@ -269,7 +269,7 @@ void initRtsFlagsDefaults(void) RtsFlags.MiscFlags.disableDelayedOsMemoryReturn = false; RtsFlags.MiscFlags.internalCounters = false; RtsFlags.MiscFlags.linkerAlwaysPic = DEFAULT_LINKER_ALWAYS_PIC; - RtsFlags.MiscFlags.linkUnknownSymbols = false; + RtsFlags.MiscFlags.linkerOptimistic = false; RtsFlags.MiscFlags.linkerMemBase = 0; RtsFlags.MiscFlags.ioManager = IO_MNGR_FLAG_AUTO; #if defined(THREADED_RTS) && defined(mingw32_HOST_OS) @@ -999,10 +999,10 @@ error = true; OPTION_UNSAFE; RtsFlags.MiscFlags.generate_dump_file = true; } - else if (strequal("link-unknown-symbols", + else if (strequal("optimistic-linking", &rts_argv[arg][2])) { OPTION_UNSAFE; - RtsFlags.MiscFlags.linkUnknownSymbols = true; + RtsFlags.MiscFlags.linkerOptimistic = true; } else if (strequal("null-eventlog-writer", &rts_argv[arg][2])) { ===================================== rts/include/rts/Flags.h ===================================== @@ -267,7 +267,7 @@ typedef struct _MISC_FLAGS { there as well. */ bool internalCounters; /* See Note [Internal Counters Stats] */ bool linkerAlwaysPic; /* Assume the object code is always PIC */ - bool linkUnknownSymbols; /* Should the runtime linker optimistically continue */ + bool linkerOptimistic; /* Should the runtime linker optimistically continue */ StgWord linkerMemBase; /* address to ask the OS for memory * for the linker, NULL ==> off */ IO_MANAGER_FLAG ioManager; /* The I/O manager to use. */ ===================================== rts/linker/elf_got.c ===================================== @@ -97,13 +97,16 @@ fillGot(ObjectCode * oc) { if(0 == strncmp(symbol->name,"_GLOBAL_OFFSET_TABLE_",21)) { symbol->addr = oc->info->got_start; } else { - errorBelch("Failed to lookup symbol: %s\n", + errorBelch("Failed to lookup symbol: %s," + " you might consider using --optimistic-linking\n", symbol->name); - // if --link-unknown-symbols is passed into the + // if --optimistic-linking is passed into the // RTS we allow the linker to optimistically // continue - if (RtsFlags.MiscFlags.linkUnknownSymbols) { + if (RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("Warning: optimistically continuing by assigning %s to OxDEADBEEF.\n", + symbol->name); symbol->addr = (void*) 0xDEADBEEF; } else { return EXIT_FAILURE; ===================================== testsuite/tests/ghci/linking/T25240/FFI.hs ===================================== @@ -6,10 +6,10 @@ {-# OPTIONS_GHC -fno-warn-missing-methods #-} module FFI - ( D(..), D2(..), tabulate + ( D(..), D2(..), tabulate, func ) where -import GHC.Exts ( Double#, Double(D#) ) +import GHC.Exts ( Double#, Double ) import Language.Haskell.TH ( CodeQ ) data D = D Double @@ -17,12 +17,9 @@ data D = D Double data D2 = D2 { x, y :: D } deriving Show -tabulate :: ( Word -> CodeQ D ) -> CodeQ ( D2 ) +tabulate :: ( Word -> CodeQ D ) -> CodeQ D2 tabulate f = [|| D2 $$( f 1 ) $$( f 2 ) ||] --- Now an unrelated "Num D" instance. -instance Num D where - ( D ( D# x ) ) * ( D ( D# y ) ) = D ( D# ( func x y ) ) foreign import prim "prim_func" func :: Double# -> Double# -> Double# ===================================== testsuite/tests/ghci/linking/T25240/Makefile ===================================== @@ -4,7 +4,7 @@ include $(TOP)/mk/test.mk .PHONY: T25240 T25240: - "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) FFI.hs TH.hs T25240.hs +RTS --link-unknown-symbols -RTS + "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) FFI.hs TH.hs T25240.hs +RTS --optimistic-linking -RTS .PHONY: T25240f T25240f: ===================================== testsuite/tests/ghci/linking/T25240/T25240.hs ===================================== @@ -3,24 +3,21 @@ Here we have two modules, FFI and TH. In the FFI module, we define a data type D and a function to be used in Template -Haskell, tabulate. Completely unrelatedly, we also include a Num D instance -which makes use of a foreign import prim declaration, foreign import prim -"prim_func". In the TH module, we use tabulate and get a linking error: +Haskell, tabulate. Completely unrelatedly, we also include foreign import prim +declaration, foreign import prim "prim_func". In the TH module, we use tabulate +and get a linking error: FFI.o: unknown symbol `prim_func' -This suggests we are trying to run some code at splice time involving the Num D -instance, even though Num D is not involved in any way. In particular, if we -change the use of prim_func to something like error "blah", then the error never -gets thrown, which shows we don't run that code at splice time. Note that it is -important to use a typeclass to trigger this bug. If we instead use the FFI -import in a top-level binding, there are no problems. It doesn't matter that the -class is Num or some other (possibly user-defined) class. +This suggests we are trying to run some code at splice time involving prim_func, +even prim_func is not involved in any way. In particular, if we change the use +of prim_func to something like error "blah", then the error never gets thrown, +which shows we don't run that code at splice time. The essential bug is that we have a symbol at link time which comes from dead code and yet we try to resolve it anyway. The fix is to pass the flag ---link-unknown-symbols to the RTS which will assign the symbol a magic number +--optimistic-linking to the RTS which will assign the symbol a magic number 0xDEADBEEF and allow the linker to continue. See MR!13012 for the implementation of that flag -} ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -9106,7 +9106,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkUnknownSymbols :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6b17092ff82644a47d7f41ccf649f77c4f8ed8ff...44eeff48eaaf43ebaa6e42fe2677a90ec38c8d0a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6b17092ff82644a47d7f41ccf649f77c4f8ed8ff...44eeff48eaaf43ebaa6e42fe2677a90ec38c8d0a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 13:46:28 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Mon, 16 Sep 2024 09:46:28 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] wrk: add user_guide and release_notes sections Message-ID: <66e836b429559_12f581e4d2c108d0@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 50d18acf by doyougnu at 2024-09-16T09:45:56-04:00 wrk: add user_guide and release_notes sections - - - - - 2 changed files: - docs/users_guide/9.12.1-notes.rst - docs/users_guide/runtime_control.rst Changes: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -147,6 +147,10 @@ Runtime system For this reason, a warning is emitted if a .tix file is read in implicitly. You can silence this warning by explicitly passing ``--read-tix-file=yes``. Details can be found in `GHC proposal 612 `__. +- Add new runtime flag :rts-flag:`--optimistic-linking` which instructs the + runtime linker to continue in the presence of unknown symbols. By default this + flag is not passed, preserving previous behavior. + ``base`` library ~~~~~~~~~~~~~~~~ ===================================== docs/users_guide/runtime_control.rst ===================================== @@ -397,6 +397,12 @@ Miscellaneous RTS options heap larger than 1T. ``-xr`` is a no-op if GHC is configured with ``--disable-large-address-space`` or if the platform is 32-bit. +.. rts-flag:: --optimistic-linking + + If given, instruct the runtime linker to try to continue linking in the + presence of an unresolved symbol by assigning that symbol the magic number + ``OxDEADBEEF``. + .. _rts-options-gc: RTS options to control the garbage collector View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/50d18acf21475f89d4ea30b8f76eaec81c50d01c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/50d18acf21475f89d4ea30b8f76eaec81c50d01c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 14:19:23 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Mon, 16 Sep 2024 10:19:23 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] wrk: repair base-export test on windows Message-ID: <66e83e6b9808a_12f5813a7d5c1426c@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 8dcb2f43 by doyougnu at 2024-09-16T10:18:56-04:00 wrk: repair base-export test on windows - - - - - 1 changed file: - testsuite/tests/interface-stability/base-exports.stdout-mingw32 Changes: ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -9330,7 +9330,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8dcb2f43939ab684555e98e3e188aa00a8c08fa9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8dcb2f43939ab684555e98e3e188aa00a8c08fa9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 14:33:29 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 16 Sep 2024 10:33:29 -0400 Subject: [Git][ghc/ghc][master] Linker: add support for extra built-in symbols (#25155) Message-ID: <66e841b9266d1_12f5815663c818352@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 13 changed files: - docs/users_guide/ghci.rst - rts/Linker.c - rts/LinkerInternals.h - rts/RtsSymbols.c - rts/RtsSymbols.h - rts/rts.cabal - testsuite/tests/ghci/linking/Makefile - + testsuite/tests/ghci/linking/T25155.hs - + testsuite/tests/ghci/linking/T25155.stdout - + testsuite/tests/ghci/linking/T25155_TH.hs - + testsuite/tests/ghci/linking/T25155_iserv.hs - + testsuite/tests/ghci/linking/T25155_iserv_main.c - testsuite/tests/ghci/linking/all.T Changes: ===================================== docs/users_guide/ghci.rst ===================================== @@ -3487,10 +3487,48 @@ dynamically-linked) from GHC itself. So for example: This feature is experimental in GHC 8.0.x, but it may become the default in future releases. +Building an external interpreter +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +The source code for the external interpreter program is in `utils/iserv`. It is +very simple because most of the heavy lifting code is from the `ghci` library. + +It is sometimes desirable to customize the external interpreter program. For +example, it is possible to add symbols to the RTS linker used by the external +interpreter. This is done simply at link time by linking an additional `.o` that +defines a `rtsExtraSyms` function returning the extra symbols. Doing it this way +avoids the need to recompile the RTS with symbols added to its built-in list. +A typical C file would look like this: + +.. code:: C + + #include + + #define CODE_SYM(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), \ + (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_CODE }, + #define DATA_SYM(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), \ + (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_DATA }, + + RtsSymbolVal my_iserv_syms[] = { + CODE_SYM(malloc) + CODE_SYM(getauxval) + CODE_SYM(posix_spawn_file_actions_init) + ... + { 0, 0, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ + }; + + RtsSymbolVal* rtsExtraSyms() { + return my_iserv_syms; + } + +For more information, read the Note [Extra RTS symbols] in the RTS. + + + .. _external-interpreter-proxy: Running the interpreter on a different host -------------------------------------------- +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ When using the flag :ghc-flag:`-fexternal-interpreter` GHC will spawn and communicate with the separate process using pipes. There @@ -3520,6 +3558,7 @@ will be executed on the target. As such packages like ``git-embed``, ``file-embed`` and others might not behave as expected if the target and host do not share the same filesystem. + .. _building-ghci-libraries: Building GHCi libraries ===================================== rts/Linker.c ===================================== @@ -472,14 +472,35 @@ initLinker_ (int retain_cafs) symhash = allocStrHashTable(); /* populate the symbol table with stuff from the RTS */ + IF_DEBUG(linker, debugBelch("populating linker symbol table with built-in RTS symbols\n")); for (const RtsSymbolVal *sym = rtsSyms; sym->lbl != NULL; sym++) { + IF_DEBUG(linker, debugBelch("initLinker: inserting rts symbol %s, %p\n", sym->lbl, sym->addr)); if (! ghciInsertSymbolTable(WSTR("(GHCi built-in symbols)"), symhash, sym->lbl, sym->addr, sym->strength, sym->type, NULL)) { barf("ghciInsertSymbolTable failed"); } - IF_DEBUG(linker, debugBelch("initLinker: inserting rts symbol %s, %p\n", sym->lbl, sym->addr)); } + IF_DEBUG(linker, debugBelch("done with built-in RTS symbols\n")); + + /* Add extra symbols. rtsExtraSyms() is a weakly defined symbol in the rts, + * that can be overrided by linking in an object with a corresponding + * definition later. This is useful to build an external-interpreter or some + * other process with extra symbols (mostly libc, or similar). + * See Note [Extra RTS symbols] + */ + IF_DEBUG(linker, debugBelch("populating linker symbol table with extra RTS symbols\n")); + if(rtsExtraSyms && rtsExtraSyms() != NULL) { + for(RtsSymbolVal *sym = rtsExtraSyms(); sym->lbl != NULL; sym++) { + IF_DEBUG(linker, debugBelch("initLinker: inserting extra rts symbol %s, %p\n", sym->lbl, sym->addr)); + if (! ghciInsertSymbolTable(WSTR("(GHCi built-in symbols)"), + symhash, sym->lbl, sym->addr, + sym->strength, sym->type, NULL)) { + barf("ghciInsertSymbolTable failed"); + } + } + } + IF_DEBUG(linker, debugBelch("done with extra RTS symbols\n")); // Redirect newCAF to newRetainedCAF if retain_cafs is true. if (! ghciInsertSymbolTable(WSTR("(GHCi built-in symbols)"), symhash, ===================================== rts/LinkerInternals.h ===================================== @@ -9,6 +9,7 @@ #pragma once #include "Rts.h" +#include "RtsSymbols.h" #include "Hash.h" #include "linker/M32Alloc.h" @@ -33,8 +34,6 @@ void printLoadedObjects(void); # define OBJFORMAT_WASM32 #endif -typedef void SymbolAddr; -typedef char SymbolName; typedef struct _ObjectCode ObjectCode; typedef struct _Section Section; @@ -53,22 +52,6 @@ typedef struct _Section Section; * to recompile their program as position-independent. */ -/* What kind of thing a symbol identifies. We need to know this to determine how - * to process overflowing relocations. See Note [Processing overflowed relocations]. - * This is bitfield however only the option SYM_TYPE_DUP_DISCARD can be combined - * with the other values. */ -typedef enum _SymType { - SYM_TYPE_CODE = 1 << 0, /* the symbol is a function and can be relocated via a jump island */ - SYM_TYPE_DATA = 1 << 1, /* the symbol is data */ - SYM_TYPE_INDIRECT_DATA = 1 << 2, /* see Note [_iob_func symbol] */ - SYM_TYPE_DUP_DISCARD = 1 << 3, /* the symbol is a symbol in a BFD import library - however if a duplicate is found with a mismatching - SymType then discard this one. */ - SYM_TYPE_HIDDEN = 1 << 4, /* the symbol is hidden and should not be exported */ - -} SymType; - - #if defined(OBJFORMAT_ELF) # include "linker/ElfTypes.h" #elif defined(OBJFORMAT_PEi386) @@ -425,12 +408,6 @@ typedef void (*init_t) (int argc, char **argv, char **env); /* Type of a finalizer */ typedef void (*fini_t) (void); -typedef enum _SymStrength { - STRENGTH_NORMAL, - STRENGTH_WEAK, - STRENGTH_STRONG, -} SymStrength; - /* SymbolInfo tracks a symbol's address, the object code from which it originated, and whether or not it's weak. ===================================== rts/RtsSymbols.c ===================================== @@ -1090,3 +1090,104 @@ RtsSymbolVal rtsSyms[] = { SymI_HasDataProto(nonmoving_write_barrier_enabled) { 0, 0, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ }; + + +// Note [Extra RTS symbols] +// ~~~~~~~~~~~~~~~~~~~~~~~~ +// How does the RTS linker know the location of the exposed RTS functions in the +// current Haskell process? Especially if it is statically linked, there is no +// dynamic table containing a symbol->address mapping. What we do is that we +// compile the following static "rtsSyms" array: +// +// RtsSymbolVar rtsSyms[] = { +// { "rts_foo", &rts_foo, ...} +// ... +// }; +// +// This array associates a predefined set of symbol names (e.g. "rts_foo") to +// their address (e.g. &rts_foo) as determined by the linker (static or +// dynamic). +// +// Note that also use this mechanism for libraries other than the RTS, e.g. +// libc. +// +// Why do we need this mapping? +// ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +// Suppose we have a Haskell program statically linked against the "libfoo" +// library and using the "foo" symbol. If we try to dynamically link some code +// using "foo" with the RTS linker, we will load another instance of "libfoo" in +// memory and use the "foo" symbol in it for the dynamically loaded code. +// However the statically linked code still uses the statically linked instance +// of libfoo and its foo symbol. +// +// This isn't a problem as long a we use side-effect free functions. However +// both instances of a "foo" function using global variables may use different +// global variables! +// +// This was a real issue with programs linked against the ghc library and +// also loading it dynamically: the GHC lib uses global variables (e.g. for the +// unique counter), and those weren't shared... This wasn't solved by adding +// GHC's global variables to the rtsSyms list but it could have been (instead a +// generic mechanism to explicit manage Haskell global variables through the RTS +// was added and used). However for other libraries, especially foreign ones, +// the issue remains! +// +// Which symbols should go into the list? +// ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +// Ideally all problematic symbols (wrt to the double loading issue mentioned +// above) should go into the list. However: +// - we can't easily predict all the symbols that will be problematic/needed +// - symbols in the list prevent some dead code elimination! If symbol "foo" +// is in the list, even if we never load any code using "foo", we'll carry the +// code for "foo" and its dependencies. +// +// The second point is particularly problematic if some symbols are only needed +// to run Template Haskell splices at compilation time: we need them in the RTS +// used by the interpreter but not in the RTS of the target program! +// +// Extra symbols +// ~~~~~~~~~~~~~ +// Since #25155 we also support specifying a second list of "extra" symbols +// which is empty by default. By overriding the weak `rtsExtraSyms` function, +// build tools can extend the list of symbols linked into the RTS. This feature +// coupled with the external-interpreter feature allows specific external +// interpreter programs to be built easily with extra symbols supported. +// +// As a concrete example, one can build an `iserv-proxy-interpreter` program +// with the following additional Cabal stanza (see [1]): +// +// if os(linux) && arch(aarch64) +// c-sources: cbits/symbols.aarch64-musl.c +// +// Where the C file contains (see [2] and [3]): +// +// #include +// +// #define CODE_SYM(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_CODE }, +// #define DATA_SYM(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_DATA }, +// +// RtsSymbolVal my_iserv_syms[] = { +// CODE_SYM(malloc) +// CODE_SYM(getauxval) +// CODE_SYM(posix_spawn_file_actions_init) +// ... +// { 0, 0, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ +// }; +// +// RtsSymbolVal* rtsExtraSyms() { +// return my_iserv_syms; +// } +// +// [1] https://github.com/stable-haskell/iserv-proxy/blob/2ed34002247213fc435d0062350b91bab920626e/iserv-proxy.cabal#L110-L111 +// [2] https://github.com/stable-haskell/iserv-proxy/blob/2ed34002247213fc435d0062350b91bab920626e/cbits/symbols.aarch64-musl.c +// [3] https://github.com/stable-haskell/iserv-proxy/blob/2ed34002247213fc435d0062350b91bab920626e/cbits/symbols.aarch64-musl.h +// + +static RtsSymbolVal default_extra_syms[] = { + { 0, 0, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ +}; + +/* Default empty extra RTS symbols. See Note[Extra RTS symbols] */ +RtsSymbolVal* __attribute__((weak)) rtsExtraSyms(void) { + return default_extra_syms; +} ===================================== rts/RtsSymbols.h ===================================== @@ -9,8 +9,6 @@ #pragma once #include "ghcautoconf.h" -#include "LinkerInternals.h" -#include #if defined(LEADING_UNDERSCORE) #define MAYBE_LEADING_UNDERSCORE_STR(s) ("_" s) @@ -18,6 +16,29 @@ #define MAYBE_LEADING_UNDERSCORE_STR(s) (s) #endif +typedef void SymbolAddr; +typedef char SymbolName; + +/* What kind of thing a symbol identifies. We need to know this to determine how + * to process overflowing relocations. See Note [Processing overflowed relocations]. + * This is bitfield however only the option SYM_TYPE_DUP_DISCARD can be combined + * with the other values. */ +typedef enum _SymType { + SYM_TYPE_CODE = 1 << 0, /* the symbol is a function and can be relocated via a jump island */ + SYM_TYPE_DATA = 1 << 1, /* the symbol is data */ + SYM_TYPE_INDIRECT_DATA = 1 << 2, /* see Note [_iob_func symbol] */ + SYM_TYPE_DUP_DISCARD = 1 << 3, /* the symbol is a symbol in a BFD import library + however if a duplicate is found with a mismatching + SymType then discard this one. */ + SYM_TYPE_HIDDEN = 1 << 4, /* the symbol is hidden and should not be exported */ +} SymType; + +typedef enum _SymStrength { + STRENGTH_NORMAL, + STRENGTH_WEAK, + STRENGTH_STRONG, +} SymStrength; + typedef struct _RtsSymbolVal { const SymbolName* lbl; SymbolAddr* addr; @@ -27,6 +48,8 @@ typedef struct _RtsSymbolVal { extern RtsSymbolVal rtsSyms[]; +extern RtsSymbolVal* __attribute__((weak)) rtsExtraSyms(void); + /* See Note [_iob_func symbol]. */ #if defined(mingw32_HOST_OS) extern const void* __rts_iob_func; ===================================== rts/rts.cabal ===================================== @@ -250,7 +250,7 @@ library include-dirs: include includes: Rts.h autogen-includes: ghcautoconf.h ghcplatform.h - install-includes: Cmm.h HsFFI.h MachDeps.h Rts.h RtsAPI.h Stg.h + install-includes: Cmm.h HsFFI.h MachDeps.h Rts.h RtsAPI.h RtsSymbols.h Stg.h ghcautoconf.h ghcconfig.h ghcplatform.h ghcversion.h -- ^ from include DerivedConstants.h ===================================== testsuite/tests/ghci/linking/Makefile ===================================== @@ -145,3 +145,8 @@ T15729: big-obj: '$(TEST_CC)' -c -Wa,-mbig-obj big-obj-c.c -o big-obj-c.o echo "main" | '$(TEST_HC)' $(TEST_HC_OPTS_INTERACTIVE) big-obj-c.o big-obj.hs + +.PHONY: T25155 +T25155: + '$(TEST_HC)' T25155_iserv_main.c T25155_iserv.hs -package ghci -no-hs-main -v0 -o T25155_iserv + '$(TEST_HC)' -fexternal-interpreter -pgmi ./T25155_iserv -v0 T25155.hs ===================================== testsuite/tests/ghci/linking/T25155.hs ===================================== @@ -0,0 +1,8 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25155 where + +import Language.Haskell.TH (runIO) +import T25155_TH + +$(runIO (foobar 7) >> pure []) ===================================== testsuite/tests/ghci/linking/T25155.stdout ===================================== @@ -0,0 +1 @@ +Called foobar with 7 ===================================== testsuite/tests/ghci/linking/T25155_TH.hs ===================================== @@ -0,0 +1,3 @@ +module T25155_TH (foobar) where + +foreign import ccall "foobar" foobar :: Int -> IO Int ===================================== testsuite/tests/ghci/linking/T25155_iserv.hs ===================================== @@ -0,0 +1,6 @@ +module Main (main) where + +import GHCi.Server (defaultServer) + +main :: IO () +main = defaultServer ===================================== testsuite/tests/ghci/linking/T25155_iserv_main.c ===================================== @@ -0,0 +1,41 @@ +#include +#include +#include +#include +#include + +int main (int argc, char *argv[]) +{ + RtsConfig conf = defaultRtsConfig; + + // We never know what symbols GHC will look up in the future, so + // we must retain CAFs for running interpreted code. + conf.keep_cafs = 1; + + conf.rts_opts_enabled = RtsOptsAll; + extern StgClosure ZCMain_main_closure; + hs_main(argc, argv, &ZCMain_main_closure, conf); +} + +// Inject "foobar" in the linker symbol table. +// +// The target object isn't compiled against any object defining foobar, yet we +// can use the FFI call to foobar in a TH splice. + +#define SymI_HasProto(vvv) { MAYBE_LEADING_UNDERSCORE_STR(#vvv), \ + (void*)(&(vvv)), STRENGTH_NORMAL, SYM_TYPE_CODE }, + + +void foobar(int x) { + printf("Called foobar with %d\n", x); +} + +RtsSymbolVal extra_syms[] = { + SymI_HasProto(foobar) + { NULL, NULL, STRENGTH_NORMAL, SYM_TYPE_CODE } /* sentinel */ + }; + +/* Extend the list of built-in symbols */ +RtsSymbolVal* rtsExtraSyms() { + return extra_syms; +} ===================================== testsuite/tests/ghci/linking/all.T ===================================== @@ -76,3 +76,4 @@ test('big-obj', [extra_files(['big-obj-c.c', 'big-obj.hs']), unless(doing_ghci, skip), unless(opsys('mingw32'), skip)], makefile_test, ['big-obj']) +test('T25155', [req_c, req_th, req_interp, extra_files(['T25155_iserv_main.c', 'T25155_iserv.hs', 'T25155_TH.hs'])], makefile_test, []) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/49ac3fb8cabc0ed584f3261296feb656d32a9ae7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/49ac3fb8cabc0ed584f3261296feb656d32a9ae7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 14:34:04 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 16 Sep 2024 10:34:04 -0400 Subject: [Git][ghc/ghc][master] GNU/Hurd: Add getExecutablePath support Message-ID: <66e841dc496f1_12f58154f434214ef@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - 1 changed file: - libraries/ghc-internal/src/GHC/Internal/System/Environment/ExecutablePath.hsc Changes: ===================================== libraries/ghc-internal/src/GHC/Internal/System/Environment/ExecutablePath.hsc ===================================== @@ -200,9 +200,9 @@ executablePath = Just (fmap Just getExecutablePath `catch` f) | otherwise = throw e -------------------------------------------------------------------------------- --- Linux / Solaris +-- Linux / Solaris / Hurd -#elif defined(linux_HOST_OS) || defined(solaris2_HOST_OS) +#elif defined(linux_HOST_OS) || defined(solaris2_HOST_OS) || defined(gnu_HOST_OS) foreign import ccall unsafe "readlink" c_readlink :: CString -> CString -> CSize -> IO CInt @@ -219,7 +219,7 @@ readSymbolicLink file = c_readlink s buf 4096 peekFilePathLen (buf,fromIntegral len) -# if defined(linux_HOST_OS) +# if defined(linux_HOST_OS) || defined(gnu_HOST_OS) getExecutablePath = readSymbolicLink $ "/proc/self/exe" executablePath = Just (check <$> getExecutablePath) where View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3939a8bf93e27d8151aa1d92bf3ce10bbbc96a72 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3939a8bf93e27d8151aa1d92bf3ce10bbbc96a72 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 15:09:40 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Mon, 16 Sep 2024 11:09:40 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25257 Message-ID: <66e84a346830b_12f581a29e50311e@gitlab.mail> Sebastian Graf pushed new branch wip/T25257 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25257 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 17:42:33 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Mon, 16 Sep 2024 13:42:33 -0400 Subject: [Git][ghc/ghc][wip/T25257] 2 commits: HsExpr: Inline `HsWrap` into `WrapExpr` Message-ID: <66e86e09d1b3c_1c43e28ab7413ee@gitlab.mail> Sebastian Graf pushed to branch wip/T25257 at Glasgow Haskell Compiler / GHC Commits: dbcac232 by Sebastian Graf at 2024-09-16T19:40:42+02:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 077028f3 by Sebastian Graf at 2024-09-16T19:42:22+02:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - 13 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Zonk/Type.hs - + testsuite/tests/pmcheck/should_compile/T25257.hs - testsuite/tests/pmcheck/should_compile/all.T Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -569,7 +569,7 @@ mkExpandedStmtPopAt loc oStmt eExpr = mkPopErrCtxtExprAt loc $ mkExpandedStmtAt data XXExprGhcTc = WrapExpr -- Type and evidence application and abstractions - {-# UNPACK #-} !(HsWrap HsExpr) + HsWrapper (HsExpr GhcTc) | ExpandedThingTc -- See Note [Rebindable syntax and XXExprGhcRn] -- See Note [Expanding HsDo with XXExprGhcRn] in `GHC.Tc.Gen.Do` @@ -881,7 +881,7 @@ instance Outputable XXExprGhcRn where ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) instance Outputable XXExprGhcTc where - ppr (WrapExpr (HsWrap co_fn e)) + ppr (WrapExpr co_fn e) = pprHsWrapper co_fn (\_parens -> pprExpr e) ppr (ExpandedThingTc o e) @@ -922,7 +922,7 @@ ppr_infix_expr_rn (ExpandedThingRn thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a ppr_infix_expr_tc :: XXExprGhcTc -> Maybe SDoc -ppr_infix_expr_tc (WrapExpr (HsWrap _ e)) = ppr_infix_expr e +ppr_infix_expr_tc (WrapExpr _ e) = ppr_infix_expr e ppr_infix_expr_tc (ExpandedThingTc thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_tc (ConLikeTc {}) = Nothing ppr_infix_expr_tc (HsTick {}) = Nothing @@ -1025,7 +1025,7 @@ hsExprNeedsParens prec = go GhcRn -> go_x_rn x go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = hsExprNeedsParens prec e + go_x_tc (WrapExpr _ e) = hsExprNeedsParens prec e go_x_tc (ExpandedThingTc thing _) = hsExpandedNeedsParens thing go_x_tc (ConLikeTc {}) = False go_x_tc (HsTick _ (L _ e)) = hsExprNeedsParens prec e @@ -1077,7 +1077,7 @@ isAtomicHsExpr (XExpr x) | GhcRn <- ghcPass @p = go_x_rn x where go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = isAtomicHsExpr e + go_x_tc (WrapExpr _ e) = isAtomicHsExpr e go_x_tc (ExpandedThingTc thing _) = isAtomicExpandedThingRn thing go_x_tc (ConLikeTc {}) = True go_x_tc (HsTick {}) = False ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -149,7 +149,7 @@ hsExprType (HsEmbTy x _) = dataConCantHappen x hsExprType (HsQual x _ _) = dataConCantHappen x hsExprType (HsForAll x _ _) = dataConCantHappen x hsExprType (HsFunArr x _ _ _) = dataConCantHappen x -hsExprType (XExpr (WrapExpr (HsWrap wrap e))) = hsWrapperType wrap $ hsExprType e +hsExprType (XExpr (WrapExpr wrap e)) = hsWrapperType wrap $ hsExprType e hsExprType (XExpr (ExpandedThingTc _ e)) = hsExprType e hsExprType (XExpr (ConLikeTc con _ _)) = conLikeType con hsExprType (XExpr (HsTick _ e)) = lhsExprType e ===================================== compiler/GHC/Hs/Utils.hs ===================================== @@ -780,7 +780,7 @@ mkLHsWrap co_fn (L loc e) = L loc (mkHsWrap co_fn e) mkHsWrap :: HsWrapper -> HsExpr GhcTc -> HsExpr GhcTc mkHsWrap co_fn e | isIdHsWrapper co_fn = e -mkHsWrap co_fn e = XExpr (WrapExpr $ HsWrap co_fn e) +mkHsWrap co_fn e = XExpr (WrapExpr co_fn e) mkHsWrapCo :: TcCoercionN -- A Nominal coercion a ~N b -> HsExpr GhcTc -> HsExpr GhcTc ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -970,8 +970,8 @@ warnUnusedBindValue fun arg@(L loc _) arg_ty fish_var :: LHsExpr GhcTc -> Maybe (SrcSpan , Id) fish_var (L l (HsVar _ id)) = return (locA l, unLoc id) fish_var (L _ (HsAppType _ e _)) = fish_var e - fish_var (L l (XExpr (WrapExpr (HsWrap _ e)))) = do (l, e') <- fish_var (L l e) - return (l, e') + fish_var (L l (XExpr (WrapExpr _ e))) = do (l, e') <- fish_var (L l e) + return (l, e') fish_var (L l (XExpr (ExpandedThingTc _ e))) = fish_var (L l e) fish_var _ = Nothing @@ -1019,7 +1019,7 @@ dsHsWrapped orig_hs_expr where go wrap (HsPar _ (L _ hs_e)) = go wrap hs_e - go wrap1 (XExpr (WrapExpr (HsWrap wrap2 hs_e))) + go wrap1 (XExpr (WrapExpr wrap2 hs_e)) = go (wrap1 <.> wrap2) hs_e go wrap (HsAppType ty (L _ hs_e) _) = go (wrap <.> WpTyApp ty) hs_e ===================================== compiler/GHC/HsToCore/Match.hs ===================================== @@ -1175,7 +1175,7 @@ viewLExprEq (e1,_) (e2,_) = lexp e1 e2 exp e (HsPar _ (L _ e')) = exp e e' -- because the expressions do not necessarily have the same type, -- we have to compare the wrappers - exp (XExpr (WrapExpr (HsWrap h e))) (XExpr (WrapExpr (HsWrap h' e'))) = + exp (XExpr (WrapExpr h e)) (XExpr (WrapExpr h' e')) = wrap h h' && exp e e' exp (XExpr (ExpandedThingTc o x)) (XExpr (ExpandedThingTc o' x')) | isHsThingRnExpr o ===================================== compiler/GHC/HsToCore/Match/Literal.hs ===================================== @@ -447,7 +447,7 @@ getLHsIntegralLit (L _ e) = go e go (XExpr (HsBinTick _ _ e)) = getLHsIntegralLit e -- The literal might be wrapped in a case with -XOverloadedLists - go (XExpr (WrapExpr (HsWrap _ e))) = go e + go (XExpr (WrapExpr _ e)) = go e go _ = Nothing -- | If 'Integral', extract the value and type of the overloaded literal. ===================================== compiler/GHC/HsToCore/Pmc/Desugar.hs ===================================== @@ -27,7 +27,7 @@ import GHC.Types.Id import GHC.Core.ConLike import GHC.Types.Name import GHC.Builtin.Types -import GHC.Builtin.Names (rationalTyConName) +import GHC.Builtin.Names (rationalTyConName, toListName) import GHC.Types.SrcLoc import GHC.Utils.Outputable import GHC.Utils.Panic @@ -44,7 +44,6 @@ import GHC.Core.TyCo.Rep import GHC.Core.TyCo.Compare( eqType ) import GHC.Core.Type import GHC.Data.Maybe -import qualified GHC.LanguageExtensions as LangExt import GHC.Types.SourceText (FractionalLit(..)) import Control.Monad (zipWithM, replicateM) import Data.List (elemIndex) @@ -125,21 +124,22 @@ desugarPat x pat = case pat of XPat ext -> case ext of ExpansionPat orig expansion -> do - dflags <- getDynFlags case orig of -- We add special logic for overloaded list patterns. When: -- - a ViewPat is the expansion of a ListPat, - -- - RebindableSyntax is off, -- - the type of the pattern is the built-in list type, -- then we assume that the view function, 'toList', is the identity. -- This improves pattern-match overload checks, as this will allow -- the pattern match checker to directly inspect the inner pattern. -- See #14547, and Note [Desugaring overloaded list patterns] (Wrinkle). ListPat {} - | ViewPat arg_ty _lexpr pat <- expansion - , not (xopt LangExt.RebindableSyntax dflags) + | ViewPat arg_ty lrhs pat <- expansion , Just tc <- tyConAppTyCon_maybe arg_ty , tc == listTyCon + -- `pat` looks like `toList @[a] -> [p1,...,pn]`. + -- Now take care of -XRebindableSyntax: + , L _ (XExpr (WrapExpr _ (HsVar _ (L _ to_list)))) <- lrhs + , idName to_list == toListName -> desugarLPat x pat _ -> desugarPat x expansion ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -588,8 +588,8 @@ addTickHsExpr (HsProc x pat cmdtop) = liftM2 (HsProc x) (addTickLPat pat) (traverse (addTickHsCmdTop) cmdtop) -addTickHsExpr (XExpr (WrapExpr (HsWrap w e))) = - liftM (XExpr . WrapExpr . HsWrap w) $ +addTickHsExpr (XExpr (WrapExpr w e)) = + liftM (XExpr . WrapExpr w) $ (addTickHsExpr e) -- Explicitly no tick on inside addTickHsExpr (XExpr (ExpandedThingTc o e)) = addTickHsExpanded o e ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1328,7 +1328,7 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where XExpr x | HieTc <- hiePass @p -> case x of - WrapExpr (HsWrap w a) + WrapExpr w a -> [ toHie $ L mspan a , toHie (L mspan w) ] ExpandedThingTc _ e ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -333,7 +333,7 @@ We expect to add to this list as we deal with more patterns via the expansion mechanism. Note [Desugaring overloaded list patterns] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If OverloadedLists is enabled, we desugar a list pattern to a view pattern: [p1, p2, p3] @@ -350,11 +350,12 @@ See Note [Invertible view patterns] in GHC.Tc.TyCl.PatSyn for details. == Wrinkle == This is all fine, except in one very specific case: - - when RebindableSyntax is off, - - and the type being matched on is already a list type. - -In this case, it is undesirable to desugar an overloaded list pattern into -a view pattern. To illustrate, consider the following program: +When the type being matched on is already a list type, so that the +pattern looks like + toList @[ty] dict -> pat +then we know for certain that `toList` is an identity function, so we can +behave exactly as if the pattern was just `pat`. This is important when +we have `OverloadedLists`. For example (#14547, #25257) > {-# LANGUAGE OverloadedLists #-} > @@ -375,6 +376,8 @@ as it isn't able to look through view patterns. We can see that this is silly: as we are matching on a list, `toList` doesn't actually do anything. So we ignore it, and desugar the pattern to an explicit list pattern, instead of a view pattern. +(NB: Because of -XRebindableSyntax we have to check that the `toList` we see is +actually resolved to `GHC.Exts.toList`.) Note however that this is not necessarily sound, because it is possible to have a list `l` such that `toList l` is not the same as `l`. ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -1080,10 +1080,10 @@ zonkExpr (HsQual x _ _) = dataConCantHappen x zonkExpr (HsForAll x _ _) = dataConCantHappen x zonkExpr (HsFunArr x _ _ _) = dataConCantHappen x -zonkExpr (XExpr (WrapExpr (HsWrap co_fn expr))) +zonkExpr (XExpr (WrapExpr co_fn expr)) = runZonkBndrT (zonkCoFn co_fn) $ \ new_co_fn -> do new_expr <- zonkExpr expr - return (XExpr (WrapExpr (HsWrap new_co_fn new_expr))) + return (XExpr (WrapExpr new_co_fn new_expr)) zonkExpr (XExpr (ExpandedThingTc thing e)) = do e' <- zonkExpr e ===================================== testsuite/tests/pmcheck/should_compile/T25257.hs ===================================== @@ -0,0 +1,11 @@ +{-# LANGUAGE OverloadedLists, RebindableSyntax #-} +{-# OPTIONS -Wincomplete-patterns #-} + +module T25257 where + +import Prelude (Bool(..), IO, print) +import GHC.Exts (IsList(fromListN, toList)) + +null :: [a] -> Bool +null [] = True +null (_:_) = False ===================================== testsuite/tests/pmcheck/should_compile/all.T ===================================== @@ -172,3 +172,4 @@ test('DsIncompleteRecSel2', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel3', [collect_compiler_stats('bytes allocated', 10)], compile, ['-Wincomplete-record-selectors']) test('DoubleMatch', normal, compile, [overlapping_incomplete]) test('T24817', normal, compile, [overlapping_incomplete]) +test('T25257', normal, compile, [overlapping_incomplete]) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4b8dbe2815e02c7532ce749e1f028f0d468f2bce...077028f3f316f7526f7babbc52f804bf42a454eb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4b8dbe2815e02c7532ce749e1f028f0d468f2bce...077028f3f316f7526f7babbc52f804bf42a454eb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 19:22:52 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Mon, 16 Sep 2024 15:22:52 -0400 Subject: [Git][ghc/ghc][wip/T25257] 2 commits: HsExpr: Inline `HsWrap` into `WrapExpr` Message-ID: <66e8858cb35c1_1c43e25d0624159c9@gitlab.mail> Sebastian Graf pushed to branch wip/T25257 at Glasgow Haskell Compiler / GHC Commits: 76074ae0 by Sebastian Graf at 2024-09-16T21:22:19+02:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7377be17 by Sebastian Graf at 2024-09-16T21:22:19+02:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - 14 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Zonk/Type.hs - ghc/GHCi/UI/Info.hs - + testsuite/tests/pmcheck/should_compile/T25257.hs - testsuite/tests/pmcheck/should_compile/all.T Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -569,7 +569,7 @@ mkExpandedStmtPopAt loc oStmt eExpr = mkPopErrCtxtExprAt loc $ mkExpandedStmtAt data XXExprGhcTc = WrapExpr -- Type and evidence application and abstractions - {-# UNPACK #-} !(HsWrap HsExpr) + HsWrapper (HsExpr GhcTc) | ExpandedThingTc -- See Note [Rebindable syntax and XXExprGhcRn] -- See Note [Expanding HsDo with XXExprGhcRn] in `GHC.Tc.Gen.Do` @@ -881,7 +881,7 @@ instance Outputable XXExprGhcRn where ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) instance Outputable XXExprGhcTc where - ppr (WrapExpr (HsWrap co_fn e)) + ppr (WrapExpr co_fn e) = pprHsWrapper co_fn (\_parens -> pprExpr e) ppr (ExpandedThingTc o e) @@ -922,7 +922,7 @@ ppr_infix_expr_rn (ExpandedThingRn thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a ppr_infix_expr_tc :: XXExprGhcTc -> Maybe SDoc -ppr_infix_expr_tc (WrapExpr (HsWrap _ e)) = ppr_infix_expr e +ppr_infix_expr_tc (WrapExpr _ e) = ppr_infix_expr e ppr_infix_expr_tc (ExpandedThingTc thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_tc (ConLikeTc {}) = Nothing ppr_infix_expr_tc (HsTick {}) = Nothing @@ -1025,7 +1025,7 @@ hsExprNeedsParens prec = go GhcRn -> go_x_rn x go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = hsExprNeedsParens prec e + go_x_tc (WrapExpr _ e) = hsExprNeedsParens prec e go_x_tc (ExpandedThingTc thing _) = hsExpandedNeedsParens thing go_x_tc (ConLikeTc {}) = False go_x_tc (HsTick _ (L _ e)) = hsExprNeedsParens prec e @@ -1077,7 +1077,7 @@ isAtomicHsExpr (XExpr x) | GhcRn <- ghcPass @p = go_x_rn x where go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = isAtomicHsExpr e + go_x_tc (WrapExpr _ e) = isAtomicHsExpr e go_x_tc (ExpandedThingTc thing _) = isAtomicExpandedThingRn thing go_x_tc (ConLikeTc {}) = True go_x_tc (HsTick {}) = False ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -149,7 +149,7 @@ hsExprType (HsEmbTy x _) = dataConCantHappen x hsExprType (HsQual x _ _) = dataConCantHappen x hsExprType (HsForAll x _ _) = dataConCantHappen x hsExprType (HsFunArr x _ _ _) = dataConCantHappen x -hsExprType (XExpr (WrapExpr (HsWrap wrap e))) = hsWrapperType wrap $ hsExprType e +hsExprType (XExpr (WrapExpr wrap e)) = hsWrapperType wrap $ hsExprType e hsExprType (XExpr (ExpandedThingTc _ e)) = hsExprType e hsExprType (XExpr (ConLikeTc con _ _)) = conLikeType con hsExprType (XExpr (HsTick _ e)) = lhsExprType e ===================================== compiler/GHC/Hs/Utils.hs ===================================== @@ -780,7 +780,7 @@ mkLHsWrap co_fn (L loc e) = L loc (mkHsWrap co_fn e) mkHsWrap :: HsWrapper -> HsExpr GhcTc -> HsExpr GhcTc mkHsWrap co_fn e | isIdHsWrapper co_fn = e -mkHsWrap co_fn e = XExpr (WrapExpr $ HsWrap co_fn e) +mkHsWrap co_fn e = XExpr (WrapExpr co_fn e) mkHsWrapCo :: TcCoercionN -- A Nominal coercion a ~N b -> HsExpr GhcTc -> HsExpr GhcTc ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -970,8 +970,8 @@ warnUnusedBindValue fun arg@(L loc _) arg_ty fish_var :: LHsExpr GhcTc -> Maybe (SrcSpan , Id) fish_var (L l (HsVar _ id)) = return (locA l, unLoc id) fish_var (L _ (HsAppType _ e _)) = fish_var e - fish_var (L l (XExpr (WrapExpr (HsWrap _ e)))) = do (l, e') <- fish_var (L l e) - return (l, e') + fish_var (L l (XExpr (WrapExpr _ e))) = do (l, e') <- fish_var (L l e) + return (l, e') fish_var (L l (XExpr (ExpandedThingTc _ e))) = fish_var (L l e) fish_var _ = Nothing @@ -1019,7 +1019,7 @@ dsHsWrapped orig_hs_expr where go wrap (HsPar _ (L _ hs_e)) = go wrap hs_e - go wrap1 (XExpr (WrapExpr (HsWrap wrap2 hs_e))) + go wrap1 (XExpr (WrapExpr wrap2 hs_e)) = go (wrap1 <.> wrap2) hs_e go wrap (HsAppType ty (L _ hs_e) _) = go (wrap <.> WpTyApp ty) hs_e ===================================== compiler/GHC/HsToCore/Match.hs ===================================== @@ -1175,7 +1175,7 @@ viewLExprEq (e1,_) (e2,_) = lexp e1 e2 exp e (HsPar _ (L _ e')) = exp e e' -- because the expressions do not necessarily have the same type, -- we have to compare the wrappers - exp (XExpr (WrapExpr (HsWrap h e))) (XExpr (WrapExpr (HsWrap h' e'))) = + exp (XExpr (WrapExpr h e)) (XExpr (WrapExpr h' e')) = wrap h h' && exp e e' exp (XExpr (ExpandedThingTc o x)) (XExpr (ExpandedThingTc o' x')) | isHsThingRnExpr o ===================================== compiler/GHC/HsToCore/Match/Literal.hs ===================================== @@ -447,7 +447,7 @@ getLHsIntegralLit (L _ e) = go e go (XExpr (HsBinTick _ _ e)) = getLHsIntegralLit e -- The literal might be wrapped in a case with -XOverloadedLists - go (XExpr (WrapExpr (HsWrap _ e))) = go e + go (XExpr (WrapExpr _ e)) = go e go _ = Nothing -- | If 'Integral', extract the value and type of the overloaded literal. ===================================== compiler/GHC/HsToCore/Pmc/Desugar.hs ===================================== @@ -27,7 +27,7 @@ import GHC.Types.Id import GHC.Core.ConLike import GHC.Types.Name import GHC.Builtin.Types -import GHC.Builtin.Names (rationalTyConName) +import GHC.Builtin.Names (rationalTyConName, toListName) import GHC.Types.SrcLoc import GHC.Utils.Outputable import GHC.Utils.Panic @@ -44,7 +44,6 @@ import GHC.Core.TyCo.Rep import GHC.Core.TyCo.Compare( eqType ) import GHC.Core.Type import GHC.Data.Maybe -import qualified GHC.LanguageExtensions as LangExt import GHC.Types.SourceText (FractionalLit(..)) import Control.Monad (zipWithM, replicateM) import Data.List (elemIndex) @@ -125,21 +124,22 @@ desugarPat x pat = case pat of XPat ext -> case ext of ExpansionPat orig expansion -> do - dflags <- getDynFlags case orig of -- We add special logic for overloaded list patterns. When: -- - a ViewPat is the expansion of a ListPat, - -- - RebindableSyntax is off, -- - the type of the pattern is the built-in list type, -- then we assume that the view function, 'toList', is the identity. -- This improves pattern-match overload checks, as this will allow -- the pattern match checker to directly inspect the inner pattern. -- See #14547, and Note [Desugaring overloaded list patterns] (Wrinkle). ListPat {} - | ViewPat arg_ty _lexpr pat <- expansion - , not (xopt LangExt.RebindableSyntax dflags) + | ViewPat arg_ty lrhs pat <- expansion , Just tc <- tyConAppTyCon_maybe arg_ty , tc == listTyCon + -- `pat` looks like `toList @[a] -> [p1,...,pn]`. + -- Now take care of -XRebindableSyntax: + , L _ (XExpr (WrapExpr _ (HsVar _ (L _ to_list)))) <- lrhs + , idName to_list == toListName -> desugarLPat x pat _ -> desugarPat x expansion ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -588,8 +588,8 @@ addTickHsExpr (HsProc x pat cmdtop) = liftM2 (HsProc x) (addTickLPat pat) (traverse (addTickHsCmdTop) cmdtop) -addTickHsExpr (XExpr (WrapExpr (HsWrap w e))) = - liftM (XExpr . WrapExpr . HsWrap w) $ +addTickHsExpr (XExpr (WrapExpr w e)) = + liftM (XExpr . WrapExpr w) $ (addTickHsExpr e) -- Explicitly no tick on inside addTickHsExpr (XExpr (ExpandedThingTc o e)) = addTickHsExpanded o e ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1328,7 +1328,7 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where XExpr x | HieTc <- hiePass @p -> case x of - WrapExpr (HsWrap w a) + WrapExpr w a -> [ toHie $ L mspan a , toHie (L mspan w) ] ExpandedThingTc _ e ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -333,7 +333,7 @@ We expect to add to this list as we deal with more patterns via the expansion mechanism. Note [Desugaring overloaded list patterns] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If OverloadedLists is enabled, we desugar a list pattern to a view pattern: [p1, p2, p3] @@ -350,11 +350,12 @@ See Note [Invertible view patterns] in GHC.Tc.TyCl.PatSyn for details. == Wrinkle == This is all fine, except in one very specific case: - - when RebindableSyntax is off, - - and the type being matched on is already a list type. - -In this case, it is undesirable to desugar an overloaded list pattern into -a view pattern. To illustrate, consider the following program: +When the type being matched on is already a list type, so that the +pattern looks like + toList @[ty] dict -> pat +then we know for certain that `toList` is an identity function, so we can +behave exactly as if the pattern was just `pat`. This is important when +we have `OverloadedLists`. For example (#14547, #25257) > {-# LANGUAGE OverloadedLists #-} > @@ -375,6 +376,8 @@ as it isn't able to look through view patterns. We can see that this is silly: as we are matching on a list, `toList` doesn't actually do anything. So we ignore it, and desugar the pattern to an explicit list pattern, instead of a view pattern. +(NB: Because of -XRebindableSyntax we have to check that the `toList` we see is +actually resolved to `GHC.Exts.toList`.) Note however that this is not necessarily sound, because it is possible to have a list `l` such that `toList l` is not the same as `l`. ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -1080,10 +1080,10 @@ zonkExpr (HsQual x _ _) = dataConCantHappen x zonkExpr (HsForAll x _ _) = dataConCantHappen x zonkExpr (HsFunArr x _ _ _) = dataConCantHappen x -zonkExpr (XExpr (WrapExpr (HsWrap co_fn expr))) +zonkExpr (XExpr (WrapExpr co_fn expr)) = runZonkBndrT (zonkCoFn co_fn) $ \ new_co_fn -> do new_expr <- zonkExpr expr - return (XExpr (WrapExpr (HsWrap new_co_fn new_expr))) + return (XExpr (WrapExpr new_co_fn new_expr)) zonkExpr (XExpr (ExpandedThingTc thing e)) = do e' <- zonkExpr e ===================================== ghc/GHCi/UI/Info.hs ===================================== @@ -355,8 +355,8 @@ processAllTypeCheckedModule tcm mid | HsVar _ (L _ i) <- unwrapVar (unLoc e) = Just i | otherwise = Nothing - unwrapVar (XExpr (WrapExpr (HsWrap _ var))) = var - unwrapVar e' = e' + unwrapVar (XExpr (WrapExpr _ var)) = var + unwrapVar e' = e' -- | Extract 'Id', 'SrcSpan', and 'Type' for 'LPats's getTypeLPat :: LPat GhcTc -> Maybe (Maybe Id,SrcSpan,Type) ===================================== testsuite/tests/pmcheck/should_compile/T25257.hs ===================================== @@ -0,0 +1,11 @@ +{-# LANGUAGE OverloadedLists, RebindableSyntax #-} +{-# OPTIONS -Wincomplete-patterns #-} + +module T25257 where + +import Prelude (Bool(..), IO, print) +import GHC.Exts (IsList(fromListN, toList)) + +null :: [a] -> Bool +null [] = True +null (_:_) = False ===================================== testsuite/tests/pmcheck/should_compile/all.T ===================================== @@ -172,3 +172,4 @@ test('DsIncompleteRecSel2', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel3', [collect_compiler_stats('bytes allocated', 10)], compile, ['-Wincomplete-record-selectors']) test('DoubleMatch', normal, compile, [overlapping_incomplete]) test('T24817', normal, compile, [overlapping_incomplete]) +test('T25257', normal, compile, [overlapping_incomplete]) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/077028f3f316f7526f7babbc52f804bf42a454eb...7377be17d5d936ec2babaf48afdea81b27e33680 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/077028f3f316f7526f7babbc52f804bf42a454eb...7377be17d5d936ec2babaf48afdea81b27e33680 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 16 21:56:14 2024 From: gitlab at gitlab.haskell.org (Krzysztof Gogolewski (@monoidal)) Date: Mon, 16 Sep 2024 17:56:14 -0400 Subject: [Git][ghc/ghc][wip/T25243] Only allow (a => b) :: Constraint rather than CONSTRAINT rep Message-ID: <66e8a97e616ed_2ceaf5aecf4134d5@gitlab.mail> Krzysztof Gogolewski pushed to branch wip/T25243 at Glasgow Haskell Compiler / GHC Commits: 1e41f636 by Krzysztof Gogolewski at 2024-09-16T23:55:57+02:00 Only allow (a => b) :: Constraint rather than CONSTRAINT rep Fixes #25243 - - - - - 4 changed files: - compiler/GHC/Tc/Gen/HsType.hs - + testsuite/tests/quantified-constraints/T25243.hs - + testsuite/tests/quantified-constraints/T25243.stderr - testsuite/tests/quantified-constraints/all.T Changes: ===================================== compiler/GHC/Tc/Gen/HsType.hs ===================================== @@ -1158,10 +1158,13 @@ tcHsType mode (HsQualTy { hst_ctxt = ctxt, hst_body = rn_ty }) exp_kind | null (unLoc ctxt) = tcLHsType mode rn_ty exp_kind -- See Note [Body kind of a HsQualTy] - | Check kind <- exp_kind, isConstraintLikeKind kind + | Check kind <- exp_kind -- Checking mode + , isConstraintLikeKind kind -- CONSTRAINT rep = do { ctxt' <- tc_hs_context mode ctxt ; ty' <- tc_check_lhs_type mode rn_ty constraintKind - ; return (tcMkDFunPhiTy ctxt' ty') } + ; let res_ty = tcMkDFunPhiTy ctxt' ty' + ; checkExpKind (unLoc rn_ty) res_ty + constraintKind exp_kind } | otherwise = do { ctxt' <- tc_hs_context mode ctxt @@ -2110,7 +2113,7 @@ However, consider instance Eq a => Eq [a] where ... or f :: (Eq a => Eq [a]) => blah -Here both body-kind of the HsQualTy is Constraint rather than *. +Here both body-kind and result kind of the HsQualTy is Constraint rather than *. Rather crudely we tell the difference by looking at exp_kind. It's very convenient to typecheck instance types like any other HsSigType. @@ -2121,11 +2124,14 @@ should be '*' we risk getting TWO error messages, one saying that Eq the left of the outer (=>). How do we figure out the right body kind? Well, it's a bit of a -kludge: I just look at the expected kind. If it's Constraint, we -must be in this instance situation context. It's a kludge because it -wouldn't work if any unification was involved to compute that result -kind -- but it isn't. (The true way might be to use the 'mode' -parameter, but that seemed like a sledgehammer to crack a nut.) +kludge: I just look at the expected kind. If we are in checking mode +(`exp_kind` = `Check k`), and the pushed-in kind `k` is `Constraint`, then +we check that the body type has kind `Constraint` too. +This is a kludge because it wouldn't work if any unification was +involved to compute that result kind -- but it isn't. +Actually, we only check whether `k` is a `CONSTRAINT rep`, but in that +case enforce that `rep` is a LiftedRep. This gives a better error message +in T25243. Note [Inferring tuple kinds] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== testsuite/tests/quantified-constraints/T25243.hs ===================================== @@ -0,0 +1,8 @@ +{-# LANGUAGE DataKinds, QuantifiedConstraints, UndecidableInstances #-} +module T25243 where + +import GHC.Exts +import Data.Kind + +type T :: Constraint -> Constraint -> CONSTRAINT IntRep +type T a b = a => b ===================================== testsuite/tests/quantified-constraints/T25243.stderr ===================================== @@ -0,0 +1,5 @@ +T25243.hs:8:14: error: [GHC-83865] + • Expected an IntRep constraint, but ‘b’ is a lifted constraint + • In the type ‘a => b’ + In the type declaration for ‘T’ + ===================================== testsuite/tests/quantified-constraints/all.T ===================================== @@ -45,3 +45,4 @@ test('T23143', normal, compile, ['']) test('T23333', normal, compile, ['']) test('T23323', normal, compile, ['']) test('T22238', normal, compile, ['']) +test('T25243', normal, compile_fail, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1e41f6367a01a3068390cfbc8baa1702d9138b76 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1e41f6367a01a3068390cfbc8baa1702d9138b76 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 07:19:16 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Tue, 17 Sep 2024 03:19:16 -0400 Subject: [Git][ghc/ghc][wip/T25257] 2 commits: HsExpr: Inline `HsWrap` into `WrapExpr` Message-ID: <66e92d7423558_2ceaf51412070430a8@gitlab.mail> Sebastian Graf pushed to branch wip/T25257 at Glasgow Haskell Compiler / GHC Commits: 0948e83b by Sebastian Graf at 2024-09-17T09:19:05+02:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 60dca249 by Sebastian Graf at 2024-09-17T09:19:05+02:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - 17 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Zonk/Type.hs - ghc/GHCi/UI/Info.hs - testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr - testsuite/tests/plugins/simple-plugin/Simple/RemovePlugin.hs - + testsuite/tests/pmcheck/should_compile/T25257.hs - testsuite/tests/pmcheck/should_compile/all.T Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -569,7 +569,7 @@ mkExpandedStmtPopAt loc oStmt eExpr = mkPopErrCtxtExprAt loc $ mkExpandedStmtAt data XXExprGhcTc = WrapExpr -- Type and evidence application and abstractions - {-# UNPACK #-} !(HsWrap HsExpr) + HsWrapper (HsExpr GhcTc) | ExpandedThingTc -- See Note [Rebindable syntax and XXExprGhcRn] -- See Note [Expanding HsDo with XXExprGhcRn] in `GHC.Tc.Gen.Do` @@ -881,7 +881,7 @@ instance Outputable XXExprGhcRn where ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) instance Outputable XXExprGhcTc where - ppr (WrapExpr (HsWrap co_fn e)) + ppr (WrapExpr co_fn e) = pprHsWrapper co_fn (\_parens -> pprExpr e) ppr (ExpandedThingTc o e) @@ -922,7 +922,7 @@ ppr_infix_expr_rn (ExpandedThingRn thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a ppr_infix_expr_tc :: XXExprGhcTc -> Maybe SDoc -ppr_infix_expr_tc (WrapExpr (HsWrap _ e)) = ppr_infix_expr e +ppr_infix_expr_tc (WrapExpr _ e) = ppr_infix_expr e ppr_infix_expr_tc (ExpandedThingTc thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_tc (ConLikeTc {}) = Nothing ppr_infix_expr_tc (HsTick {}) = Nothing @@ -1025,7 +1025,7 @@ hsExprNeedsParens prec = go GhcRn -> go_x_rn x go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = hsExprNeedsParens prec e + go_x_tc (WrapExpr _ e) = hsExprNeedsParens prec e go_x_tc (ExpandedThingTc thing _) = hsExpandedNeedsParens thing go_x_tc (ConLikeTc {}) = False go_x_tc (HsTick _ (L _ e)) = hsExprNeedsParens prec e @@ -1077,7 +1077,7 @@ isAtomicHsExpr (XExpr x) | GhcRn <- ghcPass @p = go_x_rn x where go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = isAtomicHsExpr e + go_x_tc (WrapExpr _ e) = isAtomicHsExpr e go_x_tc (ExpandedThingTc thing _) = isAtomicExpandedThingRn thing go_x_tc (ConLikeTc {}) = True go_x_tc (HsTick {}) = False ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -149,7 +149,7 @@ hsExprType (HsEmbTy x _) = dataConCantHappen x hsExprType (HsQual x _ _) = dataConCantHappen x hsExprType (HsForAll x _ _) = dataConCantHappen x hsExprType (HsFunArr x _ _ _) = dataConCantHappen x -hsExprType (XExpr (WrapExpr (HsWrap wrap e))) = hsWrapperType wrap $ hsExprType e +hsExprType (XExpr (WrapExpr wrap e)) = hsWrapperType wrap $ hsExprType e hsExprType (XExpr (ExpandedThingTc _ e)) = hsExprType e hsExprType (XExpr (ConLikeTc con _ _)) = conLikeType con hsExprType (XExpr (HsTick _ e)) = lhsExprType e ===================================== compiler/GHC/Hs/Utils.hs ===================================== @@ -780,7 +780,7 @@ mkLHsWrap co_fn (L loc e) = L loc (mkHsWrap co_fn e) mkHsWrap :: HsWrapper -> HsExpr GhcTc -> HsExpr GhcTc mkHsWrap co_fn e | isIdHsWrapper co_fn = e -mkHsWrap co_fn e = XExpr (WrapExpr $ HsWrap co_fn e) +mkHsWrap co_fn e = XExpr (WrapExpr co_fn e) mkHsWrapCo :: TcCoercionN -- A Nominal coercion a ~N b -> HsExpr GhcTc -> HsExpr GhcTc ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -970,8 +970,8 @@ warnUnusedBindValue fun arg@(L loc _) arg_ty fish_var :: LHsExpr GhcTc -> Maybe (SrcSpan , Id) fish_var (L l (HsVar _ id)) = return (locA l, unLoc id) fish_var (L _ (HsAppType _ e _)) = fish_var e - fish_var (L l (XExpr (WrapExpr (HsWrap _ e)))) = do (l, e') <- fish_var (L l e) - return (l, e') + fish_var (L l (XExpr (WrapExpr _ e))) = do (l, e') <- fish_var (L l e) + return (l, e') fish_var (L l (XExpr (ExpandedThingTc _ e))) = fish_var (L l e) fish_var _ = Nothing @@ -1019,7 +1019,7 @@ dsHsWrapped orig_hs_expr where go wrap (HsPar _ (L _ hs_e)) = go wrap hs_e - go wrap1 (XExpr (WrapExpr (HsWrap wrap2 hs_e))) + go wrap1 (XExpr (WrapExpr wrap2 hs_e)) = go (wrap1 <.> wrap2) hs_e go wrap (HsAppType ty (L _ hs_e) _) = go (wrap <.> WpTyApp ty) hs_e ===================================== compiler/GHC/HsToCore/Match.hs ===================================== @@ -1175,7 +1175,7 @@ viewLExprEq (e1,_) (e2,_) = lexp e1 e2 exp e (HsPar _ (L _ e')) = exp e e' -- because the expressions do not necessarily have the same type, -- we have to compare the wrappers - exp (XExpr (WrapExpr (HsWrap h e))) (XExpr (WrapExpr (HsWrap h' e'))) = + exp (XExpr (WrapExpr h e)) (XExpr (WrapExpr h' e')) = wrap h h' && exp e e' exp (XExpr (ExpandedThingTc o x)) (XExpr (ExpandedThingTc o' x')) | isHsThingRnExpr o ===================================== compiler/GHC/HsToCore/Match/Literal.hs ===================================== @@ -447,7 +447,7 @@ getLHsIntegralLit (L _ e) = go e go (XExpr (HsBinTick _ _ e)) = getLHsIntegralLit e -- The literal might be wrapped in a case with -XOverloadedLists - go (XExpr (WrapExpr (HsWrap _ e))) = go e + go (XExpr (WrapExpr _ e)) = go e go _ = Nothing -- | If 'Integral', extract the value and type of the overloaded literal. ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -189,6 +189,7 @@ pmcMatches origin ctxt vars matches = {-# SCC "pmcMatches" #-} do Just matches -> do matches <- {-# SCC "desugarMatches" #-} noCheckDs $ desugarMatches vars matches + tracePm "desugared matches" (ppr matches) result <- {-# SCC "checkMatchGroup" #-} unCA (checkMatchGroup matches) missing tracePm "}: " (ppr (cr_uncov result)) ===================================== compiler/GHC/HsToCore/Pmc/Desugar.hs ===================================== @@ -27,7 +27,7 @@ import GHC.Types.Id import GHC.Core.ConLike import GHC.Types.Name import GHC.Builtin.Types -import GHC.Builtin.Names (rationalTyConName) +import GHC.Builtin.Names (rationalTyConName, toListName) import GHC.Types.SrcLoc import GHC.Utils.Outputable import GHC.Utils.Panic @@ -44,7 +44,6 @@ import GHC.Core.TyCo.Rep import GHC.Core.TyCo.Compare( eqType ) import GHC.Core.Type import GHC.Data.Maybe -import qualified GHC.LanguageExtensions as LangExt import GHC.Types.SourceText (FractionalLit(..)) import Control.Monad (zipWithM, replicateM) import Data.List (elemIndex) @@ -125,21 +124,24 @@ desugarPat x pat = case pat of XPat ext -> case ext of ExpansionPat orig expansion -> do - dflags <- getDynFlags case orig of -- We add special logic for overloaded list patterns. When: -- - a ViewPat is the expansion of a ListPat, - -- - RebindableSyntax is off, -- - the type of the pattern is the built-in list type, -- then we assume that the view function, 'toList', is the identity. -- This improves pattern-match overload checks, as this will allow -- the pattern match checker to directly inspect the inner pattern. -- See #14547, and Note [Desugaring overloaded list patterns] (Wrinkle). ListPat {} - | ViewPat arg_ty _lexpr pat <- expansion - , not (xopt LangExt.RebindableSyntax dflags) + | ViewPat arg_ty lrhs pat <- expansion , Just tc <- tyConAppTyCon_maybe arg_ty , tc == listTyCon + -- `pat` looks like `coerce toList -> [p1,...,pn]`. + -- Now take care of -XRebindableSyntax: + , let is_to_list (HsVar _ (L _ to_list)) = idName to_list == toListName + is_to_list (XExpr (WrapExpr _ e)) = is_to_list e + is_to_list _ = False + , is_to_list (unLoc lrhs) -> desugarLPat x pat _ -> desugarPat x expansion ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -588,8 +588,8 @@ addTickHsExpr (HsProc x pat cmdtop) = liftM2 (HsProc x) (addTickLPat pat) (traverse (addTickHsCmdTop) cmdtop) -addTickHsExpr (XExpr (WrapExpr (HsWrap w e))) = - liftM (XExpr . WrapExpr . HsWrap w) $ +addTickHsExpr (XExpr (WrapExpr w e)) = + liftM (XExpr . WrapExpr w) $ (addTickHsExpr e) -- Explicitly no tick on inside addTickHsExpr (XExpr (ExpandedThingTc o e)) = addTickHsExpanded o e ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1328,7 +1328,7 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where XExpr x | HieTc <- hiePass @p -> case x of - WrapExpr (HsWrap w a) + WrapExpr w a -> [ toHie $ L mspan a , toHie (L mspan w) ] ExpandedThingTc _ e ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -333,7 +333,7 @@ We expect to add to this list as we deal with more patterns via the expansion mechanism. Note [Desugaring overloaded list patterns] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If OverloadedLists is enabled, we desugar a list pattern to a view pattern: [p1, p2, p3] @@ -350,11 +350,12 @@ See Note [Invertible view patterns] in GHC.Tc.TyCl.PatSyn for details. == Wrinkle == This is all fine, except in one very specific case: - - when RebindableSyntax is off, - - and the type being matched on is already a list type. - -In this case, it is undesirable to desugar an overloaded list pattern into -a view pattern. To illustrate, consider the following program: +When the type being matched on is already a list type, so that the +pattern looks like + toList @[ty] dict -> pat +then we know for certain that `toList` is an identity function, so we can +behave exactly as if the pattern was just `pat`. This is important when +we have `OverloadedLists`. For example (#14547, #25257) > {-# LANGUAGE OverloadedLists #-} > @@ -375,6 +376,8 @@ as it isn't able to look through view patterns. We can see that this is silly: as we are matching on a list, `toList` doesn't actually do anything. So we ignore it, and desugar the pattern to an explicit list pattern, instead of a view pattern. +(NB: Because of -XRebindableSyntax we have to check that the `toList` we see is +actually resolved to `GHC.Exts.toList`.) Note however that this is not necessarily sound, because it is possible to have a list `l` such that `toList l` is not the same as `l`. ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -1080,10 +1080,10 @@ zonkExpr (HsQual x _ _) = dataConCantHappen x zonkExpr (HsForAll x _ _) = dataConCantHappen x zonkExpr (HsFunArr x _ _ _) = dataConCantHappen x -zonkExpr (XExpr (WrapExpr (HsWrap co_fn expr))) +zonkExpr (XExpr (WrapExpr co_fn expr)) = runZonkBndrT (zonkCoFn co_fn) $ \ new_co_fn -> do new_expr <- zonkExpr expr - return (XExpr (WrapExpr (HsWrap new_co_fn new_expr))) + return (XExpr (WrapExpr new_co_fn new_expr)) zonkExpr (XExpr (ExpandedThingTc thing e)) = do e' <- zonkExpr e ===================================== ghc/GHCi/UI/Info.hs ===================================== @@ -355,8 +355,8 @@ processAllTypeCheckedModule tcm mid | HsVar _ (L _ i) <- unwrapVar (unLoc e) = Just i | otherwise = Nothing - unwrapVar (XExpr (WrapExpr (HsWrap _ var))) = var - unwrapVar e' = e' + unwrapVar (XExpr (WrapExpr _ var)) = var + unwrapVar e' = e' -- | Extract 'Id', 'SrcSpan', and 'Type' for 'LPats's getTypeLPat :: LPat GhcTc -> Maybe (Maybe Id,SrcSpan,Type) ===================================== testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr ===================================== @@ -1512,16 +1512,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1575,16 +1574,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1638,16 +1636,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1674,16 +1671,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))))))))))))))))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))))))))))))))))))) ,(L (EpAnn (EpaSpan { }) @@ -1750,16 +1746,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))))))) ,(L (EpAnn (EpaSpan { }) ===================================== testsuite/tests/plugins/simple-plugin/Simple/RemovePlugin.hs ===================================== @@ -53,9 +53,9 @@ typecheckPlugin [name, "typecheck"] _ tc typecheckPlugin _ _ tc = return tc metaPlugin' :: [CommandLineOption] -> LHsExpr GhcTc -> TcM (LHsExpr GhcTc) -metaPlugin' [name, "meta"] (L l (XExpr (WrapExpr (HsWrap w (HsPar _ (L _ (HsApp noExt (L _ (HsVar _ (L _ id))) e))))))) +metaPlugin' [name, "meta"] (L l (XExpr (WrapExpr w (HsPar _ (L _ (HsApp noExt (L _ (HsVar _ (L _ id))) e)))))) | occNameString (getOccName id) == name - = return (L l (XExpr (WrapExpr (HsWrap w (unLoc e))))) + = return (L l (XExpr (WrapExpr w (unLoc e)))) -- The test should always match this first case. If the desugaring changes -- again in the future then the panic is more useful than the previous -- inscrutable failure. ===================================== testsuite/tests/pmcheck/should_compile/T25257.hs ===================================== @@ -0,0 +1,11 @@ +{-# LANGUAGE OverloadedLists, RebindableSyntax #-} +{-# OPTIONS -Wincomplete-patterns #-} + +module T25257 where + +import Prelude (Bool(..), IO, print) +import GHC.Exts (IsList(fromListN, toList)) + +null :: [a] -> Bool +null [] = True +null (_:_) = False ===================================== testsuite/tests/pmcheck/should_compile/all.T ===================================== @@ -172,3 +172,4 @@ test('DsIncompleteRecSel2', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel3', [collect_compiler_stats('bytes allocated', 10)], compile, ['-Wincomplete-record-selectors']) test('DoubleMatch', normal, compile, [overlapping_incomplete]) test('T24817', normal, compile, [overlapping_incomplete]) +test('T25257', normal, compile, [overlapping_incomplete]) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7377be17d5d936ec2babaf48afdea81b27e33680...60dca24936a65a217f01c61fda54c37fd0346ab9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7377be17d5d936ec2babaf48afdea81b27e33680...60dca24936a65a217f01c61fda54c37fd0346ab9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 09:05:18 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 17 Sep 2024 05:05:18 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] fixup! determinism: Sampling uniques in the CG Message-ID: <66e9464e98f09_5453e23d610694ca@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 53d54e2b by Rodrigo Mesquita at 2024-09-17T10:04:56+01:00 fixup! determinism: Sampling uniques in the CG - - - - - 2 changed files: - compiler/GHC/Driver/GenerateCgIPEStub.hs - compiler/GHC/Types/Unique/DSM.hs Changes: ===================================== compiler/GHC/Driver/GenerateCgIPEStub.hs ===================================== @@ -37,7 +37,6 @@ import GHC.Unit.Module (moduleNameString) import qualified GHC.Utils.Logger as Logger import GHC.Utils.Outputable (ppr) import GHC.Types.Unique.DSM -import qualified Control.Monad.Trans.State.Strict as T {- Note [Stacktraces from Info Table Provenance Entries (IPE based stack unwinding)] @@ -216,15 +215,15 @@ generateCgIPEStub hsc_env this_mod denv (nonCaffySet, moduleLFInfos, infoTablesW -- of the code! Instead, make sure all labels generated for IPE related code -- sources uniques from the DUniqSupply gotten from CgStream (see its use in -- initInfoTableProv/emitIpeBufferListNode). - dus <- liftEff $ UDSMT $ T.get + (mIpeStub, ipeCmmGroup) <- liftEff $ UDSMT $ \dus -> do - -- Yield Cmm for Info Table Provenance Entries (IPEs) - let denv' = denv {provInfoTables = Map.mapKeys cit_lbl infoTablesWithTickishes} - (((mIpeStub, dus'), ipeCmmGroup), _) = - runC (initStgToCmmConfig dflags this_mod) fstate cgState $ - getCmm (initInfoTableProv initStats (Map.keys infoTablesWithTickishes) denv' dus) + -- Yield Cmm for Info Table Provenance Entries (IPEs) + let denv' = denv {provInfoTables = Map.mapKeys cit_lbl infoTablesWithTickishes} + (((mIpeStub, dus'), ipeCmmGroup), _) = + runC (initStgToCmmConfig dflags this_mod) fstate cgState $ + getCmm (initInfoTableProv initStats (Map.keys infoTablesWithTickishes) denv' dus) - liftEff $ UDSMT $ T.put dus' + return ((mIpeStub, ipeCmmGroup), dus') (_, ipeCmmGroupSRTs) <- liftEff $ withDUS $ cmmPipeline logger cmm_cfg (emptySRT this_mod) (removeDeterm ipeCmmGroup) Stream.yield ipeCmmGroupSRTs ===================================== compiler/GHC/Types/Unique/DSM.hs ===================================== @@ -8,7 +8,6 @@ import Control.Monad.Fix import GHC.Types.Unique import qualified GHC.Utils.Monad.State.Strict as Strict import qualified GHC.Types.Unique.Supply as USM -import qualified Control.Monad.Trans.State.Strict as T import Control.Monad.IO.Class {- @@ -129,48 +128,68 @@ instance MonadGetUnique USM.UniqSM where -- | Transformer version of 'UniqDSM' to use when threading a deterministic -- uniq supply over a Monad. Specifically, it is used in the `Stream` of Cmm -- decls. -newtype UniqDSMT m result = UDSMT { unUDSMT :: T.StateT DUniqSupply m (result) } - deriving (Functor, Applicative, Monad, MonadIO) +newtype UniqDSMT m result = UDSMT' { unUDSMT :: DUniqSupply -> m (result, DUniqSupply) } + deriving (Functor) + +-- Similar to GHC.Utils.Monad.State.Strict, using Note [The one-shot state monad trick] + +pattern UDSMT :: (DUniqSupply -> m (result, DUniqSupply)) -> UniqDSMT m result +pattern UDSMT m <- UDSMT' m + where + UDSMT m = UDSMT' (oneShot $ \s -> m s) +{-# COMPLETE UDSMT #-} + +instance Monad m => Applicative (UniqDSMT m) where + pure x = UDSMT $ \s -> pure (x, s) + UDSMT f <*> UDSMT x = UDSMT $ \s0 -> do + (f', s1) <- f s0 + (x', s2) <- x s1 + pure (f' x', s2) + +instance Monad m => Monad (UniqDSMT m) where + UDSMT x >>= f = UDSMT $ \s0 -> do + (x', s1) <- x s0 + case f x' of UDSMT y -> y s1 + +instance MonadIO m => MonadIO (UniqDSMT m) where + liftIO x = UDSMT $ \s -> (,s) <$> liftIO x instance Monad m => MonadGetUnique (UniqDSMT m) where - getUniqueM = UDSMT $ do - us <- T.get + getUniqueM = UDSMT $ \us -> do let (u, us') = takeUniqueFromDSupply us - T.put us' - return u + return (u, us') -- | Set the tag of the running @UniqDSMT@ supply to the given tag and run an action with it. -- All uniques produced in the given action will use this tag, until the tag is changed -- again. setTagUDSMT :: Monad m => Char {-^ Tag -} -> UniqDSMT m a -> UniqDSMT m a -setTagUDSMT tag act = do - origtag <- UDSMT (T.gets getTagDUniqSupply) - UDSMT (T.modify (newTagDUniqSupply tag)) - a <- act - UDSMT (T.modify (newTagDUniqSupply origtag)) -- restore original tag - return a +setTagUDSMT tag (UDSMT act) = UDSMT $ \us -> do + let origtag = getTagDUniqSupply us + new_us = newTagDUniqSupply tag us + (a, us') <- act new_us + let us'_origtag = newTagDUniqSupply origtag us' + -- restore original tag + return (a, us'_origtag) -- | Like 'runUniqueDSM' but for 'UniqDSMT' runUDSMT :: DUniqSupply -> UniqDSMT m a -> m (a, DUniqSupply) -runUDSMT dus (UDSMT st) = T.runStateT st dus +runUDSMT dus (UDSMT st) = st dus -- | Lift an IO action that depends on, and threads through, a unique supply -- into UniqDSMT IO. withDUS :: (DUniqSupply -> IO (a, DUniqSupply)) -> UniqDSMT IO a -withDUS f = UDSMT $ do - us <- T.get +withDUS f = UDSMT $ \us -> do (a, us') <- liftIO (f us) - _ <- T.put us' - return a + return (a, us') -- | Change the monad underyling an applied @UniqDSMT@, i.e. transform a -- @UniqDSMT m@ into a @UniqDSMT n@ given @m ~> n at . hoistUDSMT :: (forall x. m x -> n x) -> UniqDSMT m a -> UniqDSMT n a -hoistUDSMT nt (UDSMT s) = UDSMT $ T.mapStateT nt s +hoistUDSMT nt (UDSMT m) = UDSMT $ \s -> nt (m s) -- | Lift a monadic action @m a@ into an @UniqDSMT m a@ liftUDSMT :: Functor m => m a -> UniqDSMT m a -liftUDSMT m = UDSMT $ T.StateT $ \s -> (,s) <$> m +liftUDSMT m = UDSMT $ \s -> (,s) <$> m -------------------------------------------------------------------------------- -- MonadUniqDSM @@ -184,8 +203,4 @@ instance MonadUniqDSM UniqDSM where liftUniqDSM = id instance Monad m => MonadUniqDSM (UniqDSMT m) where - liftUniqDSM act = do - us <- UDSMT T.get - let (a, us') = runUniqueDSM us act - UDSMT (T.put us') - return a + liftUniqDSM act = UDSMT $ \us -> pure $ runUniqueDSM us act View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/53d54e2b2fe19d93b3ac92690cf6f6afb5054996 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/53d54e2b2fe19d93b3ac92690cf6f6afb5054996 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 10:13:10 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 17 Sep 2024 06:13:10 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 5 commits: Linker: add support for extra built-in symbols (#25155) Message-ID: <66e95636c174d_5453e5b3b3c99681@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - 20ef0ac7 by Sylvain Henry at 2024-09-17T06:13:05-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - f4db085f by Sebastian Graf at 2024-09-17T06:13:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 6d4a2a04 by Sebastian Graf at 2024-09-17T06:13:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - 30 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Zonk/Type.hs - docs/users_guide/ghci.rst - ghc/GHCi/UI/Info.hs - libraries/ghc-internal/src/GHC/Internal/System/Environment/ExecutablePath.hsc - rts/Linker.c - rts/LinkerInternals.h - rts/RtsSymbols.c - rts/RtsSymbols.h - rts/rts.cabal - testsuite/tests/ghci/linking/Makefile - + testsuite/tests/ghci/linking/T25155.hs - + testsuite/tests/ghci/linking/T25155.stdout - + testsuite/tests/ghci/linking/T25155_TH.hs - + testsuite/tests/ghci/linking/T25155_iserv.hs - + testsuite/tests/ghci/linking/T25155_iserv_main.c - testsuite/tests/ghci/linking/all.T - testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr - testsuite/tests/plugins/simple-plugin/Simple/RemovePlugin.hs - + testsuite/tests/pmcheck/should_compile/T25257.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/cf40cfff06017ad24ef4dc3bafbd37606a5b5c28...6d4a2a041727147c9b6cedf089124151cf32d7ed -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/cf40cfff06017ad24ef4dc3bafbd37606a5b5c28...6d4a2a041727147c9b6cedf089124151cf32d7ed You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 12:25:30 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Tue, 17 Sep 2024 08:25:30 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 62 commits: rts: fix checkClosure error message Message-ID: <66e9753a13c18_9896c9063c7467@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 185320a3 by sheaf at 2024-09-17T14:24:31+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - b28f75a1 by sheaf at 2024-09-17T14:24:34+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - df6d94f7 by sheaf at 2024-09-17T14:24:34+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 91458a88 by sheaf at 2024-09-17T14:24:35+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 9b74fe13 by sheaf at 2024-09-17T14:24:35+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - bcc6b168 by sheaf at 2024-09-17T14:24:35+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 7e48a328 by sheaf at 2024-09-17T14:24:36+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 8573fa7d by sheaf at 2024-09-17T14:24:36+02:00 Add test for C calls & SIMD vectors - - - - - 548a26b1 by sheaf at 2024-09-17T14:24:36+02:00 Add test for #25169 - - - - - 480c07dc by sheaf at 2024-09-17T14:24:36+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - ac8481e8 by sheaf at 2024-09-17T14:25:15+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - acf56839 by sheaf at 2024-09-17T14:25:15+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/086ca50e7d242fcc8102df437637496b1dfb4136...acf568390a4296b44032a5e6264be016ac6aef5a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/086ca50e7d242fcc8102df437637496b1dfb4136...acf568390a4296b44032a5e6264be016ac6aef5a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 12:30:36 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 17 Sep 2024 08:30:36 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 26 commits: Bump submodule deepseq to 1.5.1.0 Message-ID: <66e9766c89171_9896c90268875ee@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - 505764ba by Matthew Pickering at 2024-09-17T13:19:52+01:00 ci: Run abi-test on test-abi label - - - - - 57860775 by Rodrigo Mesquita at 2024-09-17T13:22:30+01:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - 0a42f397 by Rodrigo Mesquita at 2024-09-17T13:24:10+01:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 4325ffc1 by Rodrigo Mesquita at 2024-09-17T13:24:10+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 6457d108 by Rodrigo Mesquita at 2024-09-17T13:24:10+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - 9ae19dd2 by Rodrigo Mesquita at 2024-09-17T13:24:10+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 33412bb6 by Rodrigo Mesquita at 2024-09-17T13:24:10+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 57c3dd9b by Rodrigo Mesquita at 2024-09-17T13:24:11+01:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/Monad.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/53d54e2b2fe19d93b3ac92690cf6f6afb5054996...57c3dd9b9aefa01a096c41849ff95d7babf950e2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/53d54e2b2fe19d93b3ac92690cf6f6afb5054996...57c3dd9b9aefa01a096c41849ff95d7babf950e2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 13:34:23 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Tue, 17 Sep 2024 09:34:23 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] linker: add --optimistic-linking flag Message-ID: <66e9855f18f41_23a2121622c1058f6@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 939bd90b by doyougnu at 2024-09-17T09:26:02-04:00 linker: add --optimistic-linking flag This patch adds: - the --optimistic-linking flag which binds unknown symbols in the runtime linker to 0xDEADBEEF instead of exiting with failure - The test T25240 which tests these flags using dead code in the FFI system. - closes #25240 This patch is part of the upstreaming haskell.nix patches project. - - - - - 16 changed files: - docs/users_guide/9.12.1-notes.rst - docs/users_guide/runtime_control.rst - libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc - rts/Linker.c - rts/RtsFlags.c - rts/include/rts/Flags.h - rts/linker/elf_got.c - testsuite/tests/ghci/linking/Makefile - + testsuite/tests/ghci/linking/T25240/Makefile - + testsuite/tests/ghci/linking/T25240/T25240.hs - + testsuite/tests/ghci/linking/T25240/T25240.stderr - + testsuite/tests/ghci/linking/T25240/T25240a.hs - + testsuite/tests/ghci/linking/T25240/all.T - testsuite/tests/ghci/should_run/T18064.stderr - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-mingw32 Changes: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -147,6 +147,10 @@ Runtime system For this reason, a warning is emitted if a .tix file is read in implicitly. You can silence this warning by explicitly passing ``--read-tix-file=yes``. Details can be found in `GHC proposal 612 `__. +- Add new runtime flag :rts-flag:`--optimistic-linking` which instructs the + runtime linker to continue in the presence of unknown symbols. By default this + flag is not passed, preserving previous behavior. + ``base`` library ~~~~~~~~~~~~~~~~ ===================================== docs/users_guide/runtime_control.rst ===================================== @@ -397,6 +397,12 @@ Miscellaneous RTS options heap larger than 1T. ``-xr`` is a no-op if GHC is configured with ``--disable-large-address-space`` or if the platform is 32-bit. +.. rts-flag:: --optimistic-linking + + If given, instruct the runtime linker to try to continue linking in the + presence of an unresolved symbol by assigning that symbol the magic number + ``OxDEADBEEF``. + .. _rts-options-gc: RTS options to control the garbage collector ===================================== libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc ===================================== @@ -162,6 +162,7 @@ data MiscFlags = MiscFlags , disableDelayedOsMemoryReturn :: Bool , internalCounters :: Bool , linkerAlwaysPic :: Bool + , linkerOptimistic :: Bool , linkerMemBase :: Word -- ^ address to ask the OS for memory for the linker, 0 ==> off , ioManager :: IoManagerFlag @@ -536,6 +537,8 @@ getMiscFlags = do (#{peek MISC_FLAGS, internalCounters} ptr :: IO CBool)) <*> (toBool <$> (#{peek MISC_FLAGS, linkerAlwaysPic} ptr :: IO CBool)) + <*> (toBool <$> + (#{peek MISC_FLAGS, linkerOptimistic} ptr :: IO CBool)) <*> #{peek MISC_FLAGS, linkerMemBase} ptr <*> (toEnum . fromIntegral <$> (#{peek MISC_FLAGS, ioManager} ptr :: IO Word32)) ===================================== rts/Linker.c ===================================== @@ -946,10 +946,20 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) // lookupDependentSymbol directly. SymbolAddr* r = lookupDependentSymbol(lbl, NULL, NULL); if (!r) { - errorBelch("^^ Could not load '%s', dependency unresolved. " - "See top entry above.\n", lbl); - IF_DEBUG(linker, printLoadedObjects()); - fflush(stderr); + if (!RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("^^ Could not load '%s', dependency unresolved. " + "See top entry above. You might consider using --optimistic-linking\n", + lbl); + IF_DEBUG(linker, printLoadedObjects()); + fflush(stderr); + } else { + // if --optimistic-linking is passed into the RTS we allow the linker + // to optimistically continue + errorBelch("^^ Could not load '%s', dependency unresolved, " + "optimistically continuing; assigning '%s' to 0xDEADBEEF.\n", + lbl, lbl); + r = (void*) 0xDEADBEEF; + } } if (!runPendingInitializers()) { ===================================== rts/RtsFlags.c ===================================== @@ -269,6 +269,7 @@ void initRtsFlagsDefaults(void) RtsFlags.MiscFlags.disableDelayedOsMemoryReturn = false; RtsFlags.MiscFlags.internalCounters = false; RtsFlags.MiscFlags.linkerAlwaysPic = DEFAULT_LINKER_ALWAYS_PIC; + RtsFlags.MiscFlags.linkerOptimistic = false; RtsFlags.MiscFlags.linkerMemBase = 0; RtsFlags.MiscFlags.ioManager = IO_MNGR_FLAG_AUTO; #if defined(THREADED_RTS) && defined(mingw32_HOST_OS) @@ -998,6 +999,11 @@ error = true; OPTION_UNSAFE; RtsFlags.MiscFlags.generate_dump_file = true; } + else if (strequal("optimistic-linking", + &rts_argv[arg][2])) { + OPTION_UNSAFE; + RtsFlags.MiscFlags.linkerOptimistic = true; + } else if (strequal("null-eventlog-writer", &rts_argv[arg][2])) { OPTION_UNSAFE; ===================================== rts/include/rts/Flags.h ===================================== @@ -267,6 +267,7 @@ typedef struct _MISC_FLAGS { there as well. */ bool internalCounters; /* See Note [Internal Counters Stats] */ bool linkerAlwaysPic; /* Assume the object code is always PIC */ + bool linkerOptimistic; /* Should the runtime linker optimistically continue */ StgWord linkerMemBase; /* address to ask the OS for memory * for the linker, NULL ==> off */ IO_MANAGER_FLAG ioManager; /* The I/O manager to use. */ ===================================== rts/linker/elf_got.c ===================================== @@ -97,9 +97,21 @@ fillGot(ObjectCode * oc) { if(0 == strncmp(symbol->name,"_GLOBAL_OFFSET_TABLE_",21)) { symbol->addr = oc->info->got_start; } else { - errorBelch("Failed to lookup symbol: %s\n", + errorBelch("Failed to lookup symbol: %s," + " you might consider using --optimistic-linking\n", symbol->name); - return EXIT_FAILURE; + + // if --optimistic-linking is passed into the + // RTS we allow the linker to optimistically + // continue + if (RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("Warning: optimistically continuing by assigning %s to OxDEADBEEF.\n", + symbol->name); + symbol->addr = (void*) 0xDEADBEEF; + } else { + return EXIT_FAILURE; + } + } } } else { ===================================== testsuite/tests/ghci/linking/Makefile ===================================== @@ -145,3 +145,7 @@ T15729: big-obj: '$(TEST_CC)' -c -Wa,-mbig-obj big-obj-c.c -o big-obj-c.o echo "main" | '$(TEST_HC)' $(TEST_HC_OPTS_INTERACTIVE) big-obj-c.o big-obj.hs + +.PHONY: T25240 +T25240: + echo "main" | "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) FFI.hs TH.hs +RTS --link-unknown-objects -RTS ===================================== testsuite/tests/ghci/linking/T25240/Makefile ===================================== @@ -0,0 +1,7 @@ +TOP=../../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +.PHONY: T25240 +T25240: + "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) T25240a.hs T25240.hs +RTS --optimistic-linking -RTS ===================================== testsuite/tests/ghci/linking/T25240/T25240.hs ===================================== @@ -0,0 +1,30 @@ +{-# LANGUAGE TemplateHaskell #-} + +{- +Here we have two modules, FFI and TH. + +In the FFI module, we define a data type D and a function to be used in Template +Haskell, tabulate. Completely unrelatedly, we also include foreign import prim +declaration, foreign import prim "prim_func". In the TH module, we use tabulate +and get a linking error: + +FFI.o: unknown symbol `prim_func' + + +This suggests we are trying to run some code at splice time involving prim_func, +even prim_func is not involved in any way. In particular, if we change the use +of prim_func to something like error "blah", then the error never gets thrown, +which shows we don't run that code at splice time. + +The essential bug is that we have a symbol at link time which comes from dead +code and yet we try to resolve it anyway. The fix is to pass the flag +--optimistic-linking to the RTS which will assign the symbol a magic number +0xDEADBEEF and allow the linker to continue. See MR!13012 for the implementation +of that flag +-} + +module T25240 where + +import T25240a + +$(pure foo) ===================================== testsuite/tests/ghci/linking/T25240/T25240.stderr ===================================== @@ -0,0 +1,2 @@ +ghc: ^^ Could not load 'func', dependency unresolved, optimistically continuing; assigning 'func' to 0xDEADBEEF. + ===================================== testsuite/tests/ghci/linking/T25240/T25240a.hs ===================================== @@ -0,0 +1,13 @@ +{-# LANGUAGE TemplateHaskell #-} + + +module T25240a + ( foo, func + ) where + + +foo :: [a] +foo = [] + +foreign import ccall "func" + func :: Int -> Int ===================================== testsuite/tests/ghci/linking/T25240/all.T ===================================== @@ -0,0 +1,2 @@ +test('T25240', [req_rts_linker, extra_files(['T25240a.hs'])], + makefile_test, ['T25240']) ===================================== testsuite/tests/ghci/should_run/T18064.stderr ===================================== @@ -1,2 +1,2 @@ -: ^^ Could not load 'blah', dependency unresolved. See top entry above. +: ^^ Could not load 'blah', dependency unresolved. See top entry above. You might consider using --optimistic-linking ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -234,7 +234,7 @@ module Control.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -303,7 +303,7 @@ module Control.Exception where module Control.Exception.Annotation where -- Safety: None type ExceptionAnnotation :: * -> Constraint - class ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where + class ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where displayExceptionAnnotation :: a -> GHC.Internal.Base.String default displayExceptionAnnotation :: GHC.Internal.Show.Show a => a -> GHC.Internal.Base.String {-# MINIMAL #-} @@ -345,7 +345,7 @@ module Control.Exception.Base where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -850,11 +850,11 @@ module Data.Data where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b constrFields :: Constr -> [GHC.Internal.Base.String] constrFixity :: Constr -> Fixity @@ -897,7 +897,7 @@ module Data.Data where showConstr :: Constr -> GHC.Internal.Base.String showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -921,14 +921,14 @@ module Data.Dynamic where -- Safety: Safe type Dynamic :: * data Dynamic where - Dynamic :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic + Dynamic :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} dynApp :: Dynamic -> Dynamic -> Dynamic dynApply :: Dynamic -> Dynamic -> GHC.Internal.Maybe.Maybe Dynamic - dynTypeRep :: Dynamic -> ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + dynTypeRep :: Dynamic -> ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep fromDyn :: forall a. Typeable a => Dynamic -> a -> a fromDynamic :: forall a. Typeable a => Dynamic -> GHC.Internal.Maybe.Maybe a toDyn :: forall a. Typeable a => a -> Dynamic @@ -1815,11 +1815,11 @@ module Data.Typeable where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b decT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Data.Either.Either ((a :~: b) -> GHC.Internal.Base.Void) (a :~: b) eqT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Maybe.Maybe (a :~: b) @@ -1834,7 +1834,7 @@ module Data.Typeable where rnfTypeRep :: TypeRep -> () showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -5255,7 +5255,7 @@ module GHC.Event where data EventManager = ... type FdKey :: * data FdKey - = ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] + = ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] ; GHC.Internal.Foreign.C.Types.N:CInt[0])GHC.Internal.System.Posix.Types.Fd, ...} type IOCallback :: * @@ -5288,9 +5288,9 @@ module GHC.Event.TimeOut where type TimeoutEdit :: * type TimeoutEdit = TimeoutQueue -> TimeoutQueue type TimeoutKey :: * - newtype TimeoutKey = TK ghc-internal-9.1001.0:GHC.Internal.Event.Unique.Unique + newtype TimeoutKey = TK ghc-internal-9.1100.0:GHC.Internal.Event.Unique.Unique type TimeoutQueue :: * - type TimeoutQueue = ghc-internal-9.1001.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback + type TimeoutQueue = ghc-internal-9.1100.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback module GHC.Exception where -- Safety: Safe @@ -5302,7 +5302,7 @@ module GHC.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -5331,7 +5331,7 @@ module GHC.Exception.Type where type ArithException :: * data ArithException = Overflow | Underflow | LossOfPrecision | DivideByZero | Denormal | RatioZeroDenominator type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -7884,8 +7884,8 @@ module GHC.IO.Handle where hTryLock :: Handle -> LockMode -> GHC.Types.IO GHC.Types.Bool hWaitForInput :: Handle -> GHC.Types.Int -> GHC.Types.IO GHC.Types.Bool isEOF :: GHC.Types.IO GHC.Types.Bool - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle nativeNewline :: Newline nativeNewlineMode :: NewlineMode noNewlineTranslation :: NewlineMode @@ -7935,11 +7935,11 @@ module GHC.IO.Handle.Internals where ioe_notReadable :: forall a. GHC.Types.IO a ioe_notWritable :: forall a. GHC.Types.IO a ioe_semiclosedHandle :: forall a. GHC.Types.IO a - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle openTextEncoding :: forall a. GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.HandleType -> (forall es ds. GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextEncoder es) -> GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextDecoder ds) -> GHC.Types.IO a) -> GHC.Types.IO a readTextDevice :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer readTextDeviceNonBlocking :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer @@ -8003,7 +8003,7 @@ module GHC.IO.Handle.Types where type Handle__ :: * data Handle__ = forall dev enc_state dec_state. - (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => + (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => Handle__ {haDevice :: !dev, haType :: HandleType, haByteBuffer :: ! {-# UNPACK #-}(GHC.Internal.IORef.N:IORef[0] _N)(GHC.Internal.IORef.IORef (GHC.Internal.IO.Buffer.Buffer GHC.Internal.Word.Word8)), @@ -9106,7 +9106,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * @@ -9417,7 +9417,7 @@ module GHC.StaticPtr where -- Safety: None type IsStatic :: (* -> *) -> Constraint class IsStatic p where - fromStaticPtr :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a + fromStaticPtr :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a {-# MINIMAL fromStaticPtr #-} type StaticKey :: * type StaticKey = GHC.Internal.Fingerprint.Type.Fingerprint @@ -10813,8 +10813,8 @@ module Type.Reflection where data (:~~:) a b where HRefl :: forall {k1} (a :: k1). (:~~:) a a pattern App :: forall k2 (t :: k2). () => forall k1 (a :: k1 -> k2) (b :: k1). (t ~ a b) => TypeRep a -> TypeRep b -> TypeRep t - pattern Con :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a - pattern Con' :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a + pattern Con :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a + pattern Con' :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a pattern Fun :: forall k (fun :: k). () => forall (r1 :: GHC.Types.RuntimeRep) (r2 :: GHC.Types.RuntimeRep) (arg :: TYPE r1) (res :: TYPE r2). (k ~ *, fun ~~ (arg -> res)) => TypeRep arg -> TypeRep res -> TypeRep fun type Module :: * data Module = ... @@ -10831,7 +10831,7 @@ module Type.Reflection where type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} decTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Data.Either.Either ((a :~~: b) -> GHC.Internal.Base.Void) (a :~~: b) eqTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Maybe.Maybe (a :~~: b) moduleName :: Module -> GHC.Internal.Base.String @@ -10866,9 +10866,9 @@ module Type.Reflection.Unsafe where data TypeRep a where ... mkTrApp :: forall k1 k2 (a :: k1 -> k2) (b :: k1). TypeRep a -> TypeRep b -> TypeRep (a b) - mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a + mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a mkTyCon :: GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Types.Int -> KindRep -> TyCon - someTypeRepFingerprint :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint + someTypeRepFingerprint :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConKindArgs :: TyCon -> GHC.Types.Int tyConKindRep :: TyCon -> KindRep @@ -11202,9 +11202,9 @@ instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.I instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Monoid (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Monoid (f p), GHC.Internal.Base.Monoid (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Monoid (f (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Monoid (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Monoid (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11260,9 +11260,9 @@ instance forall a. GHC.Internal.Base.Semigroup (Data.Semigroup.Last a) -- Define instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Semigroup (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Semigroup (f p), GHC.Internal.Base.Semigroup (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Semigroup (f (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11392,20 +11392,20 @@ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Last -- Defined instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Max -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Min -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.IO.Class.MonadIO GHC.Types.IO -- Defined in ‘GHC.Internal.Control.Monad.IO.Class’ -instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ -instance forall (m :: * -> *) a. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ +instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ +instance forall (m :: * -> *) a. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Functor.ZipList.ZipList a) -- Defined in ‘GHC.Internal.Functor.ZipList’ instance GHC.Internal.Data.Data.Data Data.Array.Byte.ByteArray -- Defined in ‘Data.Array.Byte’ -instance forall s. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ +instance forall s. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Complex.Complex a) -- Defined in ‘Data.Complex’ -instance forall i j (a :: i) (b :: j). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall i j (a :: i) (b :: j). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.All -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.Any -- Defined in ‘GHC.Internal.Data.Data’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b, GHC.Internal.Ix.Ix a) => GHC.Internal.Data.Data.Data (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.Associativity -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Bool -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Char -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k a (b :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall k a (b :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.DecidedStrictness -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Double -- Defined in ‘GHC.Internal.Data.Data’ @@ -11453,10 +11453,10 @@ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word32 -- Defined in ‘G instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word64 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word8 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Foreign.Ptr.WordPtr -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ -instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ +instance forall k (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ +instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b) => GHC.Internal.Data.Data.Data (Data.Semigroup.Arg a b) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.First a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.Last a) -- Defined in ‘Data.Semigroup’ @@ -11522,7 +11522,7 @@ instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SCha instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SSymbol -- Defined in ‘GHC.Internal.TypeLits’ instance forall k (a :: k). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k (a :: k1). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ -instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1). GHC.Internal.Data.Type.Equality.TestEquality f => GHC.Internal.Data.Type.Equality.TestEquality (Data.Functor.Compose.Compose f g) -- Defined in ‘Data.Functor.Compose’ instance forall a k (b :: k). GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ instance forall a. GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Bits.And a) -- Defined in ‘GHC.Internal.Data.Bits’ @@ -11719,7 +11719,7 @@ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Ba instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.RecUpdError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.TypeError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ -instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ instance [safe] GHC.Internal.Exception.Type.Exception System.Timeout.Timeout -- Defined in ‘System.Timeout’ instance forall a k (b :: k). GHC.Internal.Float.Floating a => GHC.Internal.Float.Floating (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ @@ -12390,8 +12390,8 @@ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataRep -- Defined in ‘ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataType -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Internal.Show.Show (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance GHC.Internal.Show.Show GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ instance forall a b. (GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). Data.Fixed.HasResolution a => GHC.Internal.Show.Show (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ @@ -12462,13 +12462,13 @@ instance GHC.Internal.Show.Show GHC.Internal.Foreign.C.Types.CWchar -- Defined i instance forall a. GHC.Internal.Show.Show (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Foreign.C.ConstPtr’ instance forall a b. (GHC.Internal.Ix.Ix a, GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Internal.Show.Show GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Internal.Show.Show GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ instance GHC.Internal.Show.Show GHC.Types.Double -- Defined in ‘GHC.Internal.Float’ instance GHC.Internal.Show.Show GHC.Types.Float -- Defined in ‘GHC.Internal.Float’ @@ -12502,7 +12502,7 @@ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Handle -- Defined i instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.HandleType -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Newline -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.NewlineMode -- Defined in ‘GHC.Internal.IO.Handle.Types’ -instance [safe] GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.HandlePosn -- Defined in ‘GHC.Internal.IO.Handle’ instance GHC.Internal.Show.Show GHC.Internal.IO.IOMode.IOMode -- Defined in ‘GHC.Internal.IO.IOMode’ instance GHC.Internal.Show.Show GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ @@ -12599,8 +12599,8 @@ instance GHC.Classes.Eq GHC.Internal.Data.Data.ConstrRep -- Defined in ‘GHC.In instance GHC.Classes.Eq GHC.Internal.Data.Data.DataRep -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Classes.Eq GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Classes.Eq (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Eq a, GHC.Classes.Eq b) => GHC.Classes.Eq (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Eq (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Eq (f (g a)) => GHC.Classes.Eq (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ @@ -12676,13 +12676,13 @@ instance forall a. GHC.Classes.Eq (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) - instance forall i e. (GHC.Internal.Ix.Ix i, GHC.Classes.Eq e) => GHC.Classes.Eq (GHC.Internal.Arr.Array i e) -- Defined in ‘GHC.Internal.Arr’ instance forall s i e. GHC.Classes.Eq (GHC.Internal.Arr.STArray s i e) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Classes.Eq GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ instance GHC.Classes.Eq GHC.Internal.Event.TimeOut.TimeoutKey -- Defined in ‘GHC.Internal.Event.TimeOut’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Classes.Eq GHC.Internal.Stack.Types.SrcLoc -- Defined in ‘GHC.Internal.Stack.Types’ instance GHC.Classes.Eq GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ instance GHC.Classes.Eq GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ @@ -12778,8 +12778,8 @@ instance GHC.Classes.Ord GHC.Internal.Unicode.GeneralCategory -- Defined in ‘G instance forall k (a :: k) (b :: k). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k2 (a :: k1) (b :: k2). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k (s :: k). GHC.Classes.Ord (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Ord ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Ord ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Ord a, GHC.Classes.Ord b) => GHC.Classes.Ord (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Ord (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Ord (f (g a)) => GHC.Classes.Ord (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -9330,7 +9330,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/939bd90b2e778aad77c30c97e2939d1d171ab851 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/939bd90b2e778aad77c30c97e2939d1d171ab851 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 13:36:29 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Tue, 17 Sep 2024 09:36:29 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] linker: add --optimistic-linking flag Message-ID: <66e985dd60ea8_23a21360c6810827@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: a44c0248 by doyougnu at 2024-09-17T09:36:07-04:00 linker: add --optimistic-linking flag This patch adds: - the --optimistic-linking flag which binds unknown symbols in the runtime linker to 0xDEADBEEF instead of exiting with failure - The test T25240 which tests these flags using dead code in the FFI system. - closes #25240 This patch is part of the upstreaming haskell.nix patches project. - - - - - 16 changed files: - docs/users_guide/9.12.1-notes.rst - docs/users_guide/runtime_control.rst - libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc - rts/Linker.c - rts/RtsFlags.c - rts/include/rts/Flags.h - rts/linker/elf_got.c - testsuite/tests/ghci/linking/Makefile - + testsuite/tests/ghci/linking/T25240/Makefile - + testsuite/tests/ghci/linking/T25240/T25240.hs - + testsuite/tests/ghci/linking/T25240/T25240.stderr - + testsuite/tests/ghci/linking/T25240/T25240a.hs - + testsuite/tests/ghci/linking/T25240/all.T - testsuite/tests/ghci/should_run/T18064.stderr - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-mingw32 Changes: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -147,6 +147,10 @@ Runtime system For this reason, a warning is emitted if a .tix file is read in implicitly. You can silence this warning by explicitly passing ``--read-tix-file=yes``. Details can be found in `GHC proposal 612 `__. +- Add new runtime flag :rts-flag:`--optimistic-linking` which instructs the + runtime linker to continue in the presence of unknown symbols. By default this + flag is not passed, preserving previous behavior. + ``base`` library ~~~~~~~~~~~~~~~~ ===================================== docs/users_guide/runtime_control.rst ===================================== @@ -397,6 +397,12 @@ Miscellaneous RTS options heap larger than 1T. ``-xr`` is a no-op if GHC is configured with ``--disable-large-address-space`` or if the platform is 32-bit. +.. rts-flag:: --optimistic-linking + + If given, instruct the runtime linker to try to continue linking in the + presence of an unresolved symbol by assigning that symbol the magic number + ``OxDEADBEEF``. + .. _rts-options-gc: RTS options to control the garbage collector ===================================== libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc ===================================== @@ -162,6 +162,7 @@ data MiscFlags = MiscFlags , disableDelayedOsMemoryReturn :: Bool , internalCounters :: Bool , linkerAlwaysPic :: Bool + , linkerOptimistic :: Bool , linkerMemBase :: Word -- ^ address to ask the OS for memory for the linker, 0 ==> off , ioManager :: IoManagerFlag @@ -536,6 +537,8 @@ getMiscFlags = do (#{peek MISC_FLAGS, internalCounters} ptr :: IO CBool)) <*> (toBool <$> (#{peek MISC_FLAGS, linkerAlwaysPic} ptr :: IO CBool)) + <*> (toBool <$> + (#{peek MISC_FLAGS, linkerOptimistic} ptr :: IO CBool)) <*> #{peek MISC_FLAGS, linkerMemBase} ptr <*> (toEnum . fromIntegral <$> (#{peek MISC_FLAGS, ioManager} ptr :: IO Word32)) ===================================== rts/Linker.c ===================================== @@ -946,10 +946,20 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) // lookupDependentSymbol directly. SymbolAddr* r = lookupDependentSymbol(lbl, NULL, NULL); if (!r) { - errorBelch("^^ Could not load '%s', dependency unresolved. " - "See top entry above.\n", lbl); - IF_DEBUG(linker, printLoadedObjects()); - fflush(stderr); + if (!RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("^^ Could not load '%s', dependency unresolved. " + "See top entry above. You might consider using --optimistic-linking\n", + lbl); + IF_DEBUG(linker, printLoadedObjects()); + fflush(stderr); + } else { + // if --optimistic-linking is passed into the RTS we allow the linker + // to optimistically continue + errorBelch("^^ Could not load '%s', dependency unresolved, " + "optimistically continuing; assigning '%s' to 0xDEADBEEF.\n", + lbl, lbl); + r = (void*) 0xDEADBEEF; + } } if (!runPendingInitializers()) { ===================================== rts/RtsFlags.c ===================================== @@ -269,6 +269,7 @@ void initRtsFlagsDefaults(void) RtsFlags.MiscFlags.disableDelayedOsMemoryReturn = false; RtsFlags.MiscFlags.internalCounters = false; RtsFlags.MiscFlags.linkerAlwaysPic = DEFAULT_LINKER_ALWAYS_PIC; + RtsFlags.MiscFlags.linkerOptimistic = false; RtsFlags.MiscFlags.linkerMemBase = 0; RtsFlags.MiscFlags.ioManager = IO_MNGR_FLAG_AUTO; #if defined(THREADED_RTS) && defined(mingw32_HOST_OS) @@ -998,6 +999,11 @@ error = true; OPTION_UNSAFE; RtsFlags.MiscFlags.generate_dump_file = true; } + else if (strequal("optimistic-linking", + &rts_argv[arg][2])) { + OPTION_UNSAFE; + RtsFlags.MiscFlags.linkerOptimistic = true; + } else if (strequal("null-eventlog-writer", &rts_argv[arg][2])) { OPTION_UNSAFE; ===================================== rts/include/rts/Flags.h ===================================== @@ -267,6 +267,7 @@ typedef struct _MISC_FLAGS { there as well. */ bool internalCounters; /* See Note [Internal Counters Stats] */ bool linkerAlwaysPic; /* Assume the object code is always PIC */ + bool linkerOptimistic; /* Should the runtime linker optimistically continue */ StgWord linkerMemBase; /* address to ask the OS for memory * for the linker, NULL ==> off */ IO_MANAGER_FLAG ioManager; /* The I/O manager to use. */ ===================================== rts/linker/elf_got.c ===================================== @@ -97,9 +97,21 @@ fillGot(ObjectCode * oc) { if(0 == strncmp(symbol->name,"_GLOBAL_OFFSET_TABLE_",21)) { symbol->addr = oc->info->got_start; } else { - errorBelch("Failed to lookup symbol: %s\n", + errorBelch("Failed to lookup symbol: %s," + " you might consider using --optimistic-linking\n", symbol->name); - return EXIT_FAILURE; + + // if --optimistic-linking is passed into the + // RTS we allow the linker to optimistically + // continue + if (RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("Warning: optimistically continuing by assigning %s to OxDEADBEEF.\n", + symbol->name); + symbol->addr = (void*) 0xDEADBEEF; + } else { + return EXIT_FAILURE; + } + } } } else { ===================================== testsuite/tests/ghci/linking/Makefile ===================================== @@ -145,3 +145,7 @@ T15729: big-obj: '$(TEST_CC)' -c -Wa,-mbig-obj big-obj-c.c -o big-obj-c.o echo "main" | '$(TEST_HC)' $(TEST_HC_OPTS_INTERACTIVE) big-obj-c.o big-obj.hs + +.PHONY: T25240 +T25240: + echo "main" | "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) FFI.hs TH.hs +RTS --link-unknown-objects -RTS ===================================== testsuite/tests/ghci/linking/T25240/Makefile ===================================== @@ -0,0 +1,7 @@ +TOP=../../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +.PHONY: T25240 +T25240: + "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) T25240a.hs T25240.hs +RTS --optimistic-linking -RTS ===================================== testsuite/tests/ghci/linking/T25240/T25240.hs ===================================== @@ -0,0 +1,17 @@ +{-# LANGUAGE TemplateHaskell #-} + +{- + +When the RTS linker loads the T25240a module to run the pure foo splice, it +tries to resolve the func symbol even if this function isn't required to run the +splice code, i.e., its dead code. This test checks that by passing the +--optimistic-linking flag the RTS linker continues to link even in the presence +of unknown symbols. + +-} + +module T25240 where + +import T25240a + +$(pure foo) ===================================== testsuite/tests/ghci/linking/T25240/T25240.stderr ===================================== @@ -0,0 +1,2 @@ +ghc: ^^ Could not load 'func', dependency unresolved, optimistically continuing; assigning 'func' to 0xDEADBEEF. + ===================================== testsuite/tests/ghci/linking/T25240/T25240a.hs ===================================== @@ -0,0 +1,13 @@ +{-# LANGUAGE TemplateHaskell #-} + + +module T25240a + ( foo, func + ) where + + +foo :: [a] +foo = [] + +foreign import ccall "func" + func :: Int -> Int ===================================== testsuite/tests/ghci/linking/T25240/all.T ===================================== @@ -0,0 +1,2 @@ +test('T25240', [req_rts_linker, extra_files(['T25240a.hs'])], + makefile_test, ['T25240']) ===================================== testsuite/tests/ghci/should_run/T18064.stderr ===================================== @@ -1,2 +1,2 @@ -: ^^ Could not load 'blah', dependency unresolved. See top entry above. +: ^^ Could not load 'blah', dependency unresolved. See top entry above. You might consider using --optimistic-linking ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -234,7 +234,7 @@ module Control.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -303,7 +303,7 @@ module Control.Exception where module Control.Exception.Annotation where -- Safety: None type ExceptionAnnotation :: * -> Constraint - class ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where + class ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where displayExceptionAnnotation :: a -> GHC.Internal.Base.String default displayExceptionAnnotation :: GHC.Internal.Show.Show a => a -> GHC.Internal.Base.String {-# MINIMAL #-} @@ -345,7 +345,7 @@ module Control.Exception.Base where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -850,11 +850,11 @@ module Data.Data where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b constrFields :: Constr -> [GHC.Internal.Base.String] constrFixity :: Constr -> Fixity @@ -897,7 +897,7 @@ module Data.Data where showConstr :: Constr -> GHC.Internal.Base.String showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -921,14 +921,14 @@ module Data.Dynamic where -- Safety: Safe type Dynamic :: * data Dynamic where - Dynamic :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic + Dynamic :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} dynApp :: Dynamic -> Dynamic -> Dynamic dynApply :: Dynamic -> Dynamic -> GHC.Internal.Maybe.Maybe Dynamic - dynTypeRep :: Dynamic -> ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + dynTypeRep :: Dynamic -> ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep fromDyn :: forall a. Typeable a => Dynamic -> a -> a fromDynamic :: forall a. Typeable a => Dynamic -> GHC.Internal.Maybe.Maybe a toDyn :: forall a. Typeable a => a -> Dynamic @@ -1815,11 +1815,11 @@ module Data.Typeable where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b decT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Data.Either.Either ((a :~: b) -> GHC.Internal.Base.Void) (a :~: b) eqT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Maybe.Maybe (a :~: b) @@ -1834,7 +1834,7 @@ module Data.Typeable where rnfTypeRep :: TypeRep -> () showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -5255,7 +5255,7 @@ module GHC.Event where data EventManager = ... type FdKey :: * data FdKey - = ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] + = ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] ; GHC.Internal.Foreign.C.Types.N:CInt[0])GHC.Internal.System.Posix.Types.Fd, ...} type IOCallback :: * @@ -5288,9 +5288,9 @@ module GHC.Event.TimeOut where type TimeoutEdit :: * type TimeoutEdit = TimeoutQueue -> TimeoutQueue type TimeoutKey :: * - newtype TimeoutKey = TK ghc-internal-9.1001.0:GHC.Internal.Event.Unique.Unique + newtype TimeoutKey = TK ghc-internal-9.1100.0:GHC.Internal.Event.Unique.Unique type TimeoutQueue :: * - type TimeoutQueue = ghc-internal-9.1001.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback + type TimeoutQueue = ghc-internal-9.1100.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback module GHC.Exception where -- Safety: Safe @@ -5302,7 +5302,7 @@ module GHC.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -5331,7 +5331,7 @@ module GHC.Exception.Type where type ArithException :: * data ArithException = Overflow | Underflow | LossOfPrecision | DivideByZero | Denormal | RatioZeroDenominator type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -7884,8 +7884,8 @@ module GHC.IO.Handle where hTryLock :: Handle -> LockMode -> GHC.Types.IO GHC.Types.Bool hWaitForInput :: Handle -> GHC.Types.Int -> GHC.Types.IO GHC.Types.Bool isEOF :: GHC.Types.IO GHC.Types.Bool - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle nativeNewline :: Newline nativeNewlineMode :: NewlineMode noNewlineTranslation :: NewlineMode @@ -7935,11 +7935,11 @@ module GHC.IO.Handle.Internals where ioe_notReadable :: forall a. GHC.Types.IO a ioe_notWritable :: forall a. GHC.Types.IO a ioe_semiclosedHandle :: forall a. GHC.Types.IO a - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle openTextEncoding :: forall a. GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.HandleType -> (forall es ds. GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextEncoder es) -> GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextDecoder ds) -> GHC.Types.IO a) -> GHC.Types.IO a readTextDevice :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer readTextDeviceNonBlocking :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer @@ -8003,7 +8003,7 @@ module GHC.IO.Handle.Types where type Handle__ :: * data Handle__ = forall dev enc_state dec_state. - (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => + (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => Handle__ {haDevice :: !dev, haType :: HandleType, haByteBuffer :: ! {-# UNPACK #-}(GHC.Internal.IORef.N:IORef[0] _N)(GHC.Internal.IORef.IORef (GHC.Internal.IO.Buffer.Buffer GHC.Internal.Word.Word8)), @@ -9106,7 +9106,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * @@ -9417,7 +9417,7 @@ module GHC.StaticPtr where -- Safety: None type IsStatic :: (* -> *) -> Constraint class IsStatic p where - fromStaticPtr :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a + fromStaticPtr :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a {-# MINIMAL fromStaticPtr #-} type StaticKey :: * type StaticKey = GHC.Internal.Fingerprint.Type.Fingerprint @@ -10813,8 +10813,8 @@ module Type.Reflection where data (:~~:) a b where HRefl :: forall {k1} (a :: k1). (:~~:) a a pattern App :: forall k2 (t :: k2). () => forall k1 (a :: k1 -> k2) (b :: k1). (t ~ a b) => TypeRep a -> TypeRep b -> TypeRep t - pattern Con :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a - pattern Con' :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a + pattern Con :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a + pattern Con' :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a pattern Fun :: forall k (fun :: k). () => forall (r1 :: GHC.Types.RuntimeRep) (r2 :: GHC.Types.RuntimeRep) (arg :: TYPE r1) (res :: TYPE r2). (k ~ *, fun ~~ (arg -> res)) => TypeRep arg -> TypeRep res -> TypeRep fun type Module :: * data Module = ... @@ -10831,7 +10831,7 @@ module Type.Reflection where type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} decTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Data.Either.Either ((a :~~: b) -> GHC.Internal.Base.Void) (a :~~: b) eqTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Maybe.Maybe (a :~~: b) moduleName :: Module -> GHC.Internal.Base.String @@ -10866,9 +10866,9 @@ module Type.Reflection.Unsafe where data TypeRep a where ... mkTrApp :: forall k1 k2 (a :: k1 -> k2) (b :: k1). TypeRep a -> TypeRep b -> TypeRep (a b) - mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a + mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a mkTyCon :: GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Types.Int -> KindRep -> TyCon - someTypeRepFingerprint :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint + someTypeRepFingerprint :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConKindArgs :: TyCon -> GHC.Types.Int tyConKindRep :: TyCon -> KindRep @@ -11202,9 +11202,9 @@ instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.I instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Monoid (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Monoid (f p), GHC.Internal.Base.Monoid (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Monoid (f (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Monoid (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Monoid (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11260,9 +11260,9 @@ instance forall a. GHC.Internal.Base.Semigroup (Data.Semigroup.Last a) -- Define instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Semigroup (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Semigroup (f p), GHC.Internal.Base.Semigroup (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Semigroup (f (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11392,20 +11392,20 @@ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Last -- Defined instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Max -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Min -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.IO.Class.MonadIO GHC.Types.IO -- Defined in ‘GHC.Internal.Control.Monad.IO.Class’ -instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ -instance forall (m :: * -> *) a. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ +instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ +instance forall (m :: * -> *) a. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Functor.ZipList.ZipList a) -- Defined in ‘GHC.Internal.Functor.ZipList’ instance GHC.Internal.Data.Data.Data Data.Array.Byte.ByteArray -- Defined in ‘Data.Array.Byte’ -instance forall s. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ +instance forall s. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Complex.Complex a) -- Defined in ‘Data.Complex’ -instance forall i j (a :: i) (b :: j). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall i j (a :: i) (b :: j). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.All -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.Any -- Defined in ‘GHC.Internal.Data.Data’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b, GHC.Internal.Ix.Ix a) => GHC.Internal.Data.Data.Data (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.Associativity -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Bool -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Char -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k a (b :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall k a (b :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.DecidedStrictness -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Double -- Defined in ‘GHC.Internal.Data.Data’ @@ -11453,10 +11453,10 @@ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word32 -- Defined in ‘G instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word64 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word8 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Foreign.Ptr.WordPtr -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ -instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ +instance forall k (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ +instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b) => GHC.Internal.Data.Data.Data (Data.Semigroup.Arg a b) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.First a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.Last a) -- Defined in ‘Data.Semigroup’ @@ -11522,7 +11522,7 @@ instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SCha instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SSymbol -- Defined in ‘GHC.Internal.TypeLits’ instance forall k (a :: k). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k (a :: k1). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ -instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1). GHC.Internal.Data.Type.Equality.TestEquality f => GHC.Internal.Data.Type.Equality.TestEquality (Data.Functor.Compose.Compose f g) -- Defined in ‘Data.Functor.Compose’ instance forall a k (b :: k). GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ instance forall a. GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Bits.And a) -- Defined in ‘GHC.Internal.Data.Bits’ @@ -11719,7 +11719,7 @@ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Ba instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.RecUpdError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.TypeError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ -instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ instance [safe] GHC.Internal.Exception.Type.Exception System.Timeout.Timeout -- Defined in ‘System.Timeout’ instance forall a k (b :: k). GHC.Internal.Float.Floating a => GHC.Internal.Float.Floating (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ @@ -12390,8 +12390,8 @@ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataRep -- Defined in ‘ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataType -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Internal.Show.Show (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance GHC.Internal.Show.Show GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ instance forall a b. (GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). Data.Fixed.HasResolution a => GHC.Internal.Show.Show (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ @@ -12462,13 +12462,13 @@ instance GHC.Internal.Show.Show GHC.Internal.Foreign.C.Types.CWchar -- Defined i instance forall a. GHC.Internal.Show.Show (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Foreign.C.ConstPtr’ instance forall a b. (GHC.Internal.Ix.Ix a, GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Internal.Show.Show GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Internal.Show.Show GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ instance GHC.Internal.Show.Show GHC.Types.Double -- Defined in ‘GHC.Internal.Float’ instance GHC.Internal.Show.Show GHC.Types.Float -- Defined in ‘GHC.Internal.Float’ @@ -12502,7 +12502,7 @@ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Handle -- Defined i instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.HandleType -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Newline -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.NewlineMode -- Defined in ‘GHC.Internal.IO.Handle.Types’ -instance [safe] GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.HandlePosn -- Defined in ‘GHC.Internal.IO.Handle’ instance GHC.Internal.Show.Show GHC.Internal.IO.IOMode.IOMode -- Defined in ‘GHC.Internal.IO.IOMode’ instance GHC.Internal.Show.Show GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ @@ -12599,8 +12599,8 @@ instance GHC.Classes.Eq GHC.Internal.Data.Data.ConstrRep -- Defined in ‘GHC.In instance GHC.Classes.Eq GHC.Internal.Data.Data.DataRep -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Classes.Eq GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Classes.Eq (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Eq a, GHC.Classes.Eq b) => GHC.Classes.Eq (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Eq (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Eq (f (g a)) => GHC.Classes.Eq (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ @@ -12676,13 +12676,13 @@ instance forall a. GHC.Classes.Eq (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) - instance forall i e. (GHC.Internal.Ix.Ix i, GHC.Classes.Eq e) => GHC.Classes.Eq (GHC.Internal.Arr.Array i e) -- Defined in ‘GHC.Internal.Arr’ instance forall s i e. GHC.Classes.Eq (GHC.Internal.Arr.STArray s i e) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Classes.Eq GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ instance GHC.Classes.Eq GHC.Internal.Event.TimeOut.TimeoutKey -- Defined in ‘GHC.Internal.Event.TimeOut’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Classes.Eq GHC.Internal.Stack.Types.SrcLoc -- Defined in ‘GHC.Internal.Stack.Types’ instance GHC.Classes.Eq GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ instance GHC.Classes.Eq GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ @@ -12778,8 +12778,8 @@ instance GHC.Classes.Ord GHC.Internal.Unicode.GeneralCategory -- Defined in ‘G instance forall k (a :: k) (b :: k). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k2 (a :: k1) (b :: k2). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k (s :: k). GHC.Classes.Ord (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Ord ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Ord ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Ord a, GHC.Classes.Ord b) => GHC.Classes.Ord (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Ord (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Ord (f (g a)) => GHC.Classes.Ord (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -9330,7 +9330,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a44c0248a0d4ec01e85e5b0fe7435553e2b08e27 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a44c0248a0d4ec01e85e5b0fe7435553e2b08e27 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 13:39:07 2024 From: gitlab at gitlab.haskell.org (John Ericson (@Ericson2314)) Date: Tue, 17 Sep 2024 09:39:07 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/rm-ghc-internal-pack Message-ID: <66e9867b7ad22_23a21360c54112948@gitlab.mail> John Ericson pushed new branch wip/rm-ghc-internal-pack at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/rm-ghc-internal-pack You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 13:39:25 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Tue, 17 Sep 2024 09:39:25 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] linker: add --optimistic-linking flag Message-ID: <66e9868d63354_23a21500dfc113466@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 6f3894f7 by doyougnu at 2024-09-17T09:39:02-04:00 linker: add --optimistic-linking flag This patch adds: - the --optimistic-linking flag which binds unknown symbols in the runtime linker to 0xDEADBEEF instead of exiting with failure - The test T25240 which tests these flags using dead code in the FFI system. - closes #25240 This patch is part of the upstreaming haskell.nix patches project. - - - - - 15 changed files: - docs/users_guide/9.12.1-notes.rst - docs/users_guide/runtime_control.rst - libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc - rts/Linker.c - rts/RtsFlags.c - rts/include/rts/Flags.h - rts/linker/elf_got.c - + testsuite/tests/ghci/linking/T25240/Makefile - + testsuite/tests/ghci/linking/T25240/T25240.hs - + testsuite/tests/ghci/linking/T25240/T25240.stderr - + testsuite/tests/ghci/linking/T25240/T25240a.hs - + testsuite/tests/ghci/linking/T25240/all.T - testsuite/tests/ghci/should_run/T18064.stderr - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-mingw32 Changes: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -147,6 +147,10 @@ Runtime system For this reason, a warning is emitted if a .tix file is read in implicitly. You can silence this warning by explicitly passing ``--read-tix-file=yes``. Details can be found in `GHC proposal 612 `__. +- Add new runtime flag :rts-flag:`--optimistic-linking` which instructs the + runtime linker to continue in the presence of unknown symbols. By default this + flag is not passed, preserving previous behavior. + ``base`` library ~~~~~~~~~~~~~~~~ ===================================== docs/users_guide/runtime_control.rst ===================================== @@ -397,6 +397,12 @@ Miscellaneous RTS options heap larger than 1T. ``-xr`` is a no-op if GHC is configured with ``--disable-large-address-space`` or if the platform is 32-bit. +.. rts-flag:: --optimistic-linking + + If given, instruct the runtime linker to try to continue linking in the + presence of an unresolved symbol by assigning that symbol the magic number + ``OxDEADBEEF``. + .. _rts-options-gc: RTS options to control the garbage collector ===================================== libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc ===================================== @@ -162,6 +162,7 @@ data MiscFlags = MiscFlags , disableDelayedOsMemoryReturn :: Bool , internalCounters :: Bool , linkerAlwaysPic :: Bool + , linkerOptimistic :: Bool , linkerMemBase :: Word -- ^ address to ask the OS for memory for the linker, 0 ==> off , ioManager :: IoManagerFlag @@ -536,6 +537,8 @@ getMiscFlags = do (#{peek MISC_FLAGS, internalCounters} ptr :: IO CBool)) <*> (toBool <$> (#{peek MISC_FLAGS, linkerAlwaysPic} ptr :: IO CBool)) + <*> (toBool <$> + (#{peek MISC_FLAGS, linkerOptimistic} ptr :: IO CBool)) <*> #{peek MISC_FLAGS, linkerMemBase} ptr <*> (toEnum . fromIntegral <$> (#{peek MISC_FLAGS, ioManager} ptr :: IO Word32)) ===================================== rts/Linker.c ===================================== @@ -946,10 +946,20 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) // lookupDependentSymbol directly. SymbolAddr* r = lookupDependentSymbol(lbl, NULL, NULL); if (!r) { - errorBelch("^^ Could not load '%s', dependency unresolved. " - "See top entry above.\n", lbl); - IF_DEBUG(linker, printLoadedObjects()); - fflush(stderr); + if (!RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("^^ Could not load '%s', dependency unresolved. " + "See top entry above. You might consider using --optimistic-linking\n", + lbl); + IF_DEBUG(linker, printLoadedObjects()); + fflush(stderr); + } else { + // if --optimistic-linking is passed into the RTS we allow the linker + // to optimistically continue + errorBelch("^^ Could not load '%s', dependency unresolved, " + "optimistically continuing; assigning '%s' to 0xDEADBEEF.\n", + lbl, lbl); + r = (void*) 0xDEADBEEF; + } } if (!runPendingInitializers()) { ===================================== rts/RtsFlags.c ===================================== @@ -269,6 +269,7 @@ void initRtsFlagsDefaults(void) RtsFlags.MiscFlags.disableDelayedOsMemoryReturn = false; RtsFlags.MiscFlags.internalCounters = false; RtsFlags.MiscFlags.linkerAlwaysPic = DEFAULT_LINKER_ALWAYS_PIC; + RtsFlags.MiscFlags.linkerOptimistic = false; RtsFlags.MiscFlags.linkerMemBase = 0; RtsFlags.MiscFlags.ioManager = IO_MNGR_FLAG_AUTO; #if defined(THREADED_RTS) && defined(mingw32_HOST_OS) @@ -998,6 +999,11 @@ error = true; OPTION_UNSAFE; RtsFlags.MiscFlags.generate_dump_file = true; } + else if (strequal("optimistic-linking", + &rts_argv[arg][2])) { + OPTION_UNSAFE; + RtsFlags.MiscFlags.linkerOptimistic = true; + } else if (strequal("null-eventlog-writer", &rts_argv[arg][2])) { OPTION_UNSAFE; ===================================== rts/include/rts/Flags.h ===================================== @@ -267,6 +267,7 @@ typedef struct _MISC_FLAGS { there as well. */ bool internalCounters; /* See Note [Internal Counters Stats] */ bool linkerAlwaysPic; /* Assume the object code is always PIC */ + bool linkerOptimistic; /* Should the runtime linker optimistically continue */ StgWord linkerMemBase; /* address to ask the OS for memory * for the linker, NULL ==> off */ IO_MANAGER_FLAG ioManager; /* The I/O manager to use. */ ===================================== rts/linker/elf_got.c ===================================== @@ -97,9 +97,21 @@ fillGot(ObjectCode * oc) { if(0 == strncmp(symbol->name,"_GLOBAL_OFFSET_TABLE_",21)) { symbol->addr = oc->info->got_start; } else { - errorBelch("Failed to lookup symbol: %s\n", + errorBelch("Failed to lookup symbol: %s," + " you might consider using --optimistic-linking\n", symbol->name); - return EXIT_FAILURE; + + // if --optimistic-linking is passed into the + // RTS we allow the linker to optimistically + // continue + if (RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("Warning: optimistically continuing by assigning %s to OxDEADBEEF.\n", + symbol->name); + symbol->addr = (void*) 0xDEADBEEF; + } else { + return EXIT_FAILURE; + } + } } } else { ===================================== testsuite/tests/ghci/linking/T25240/Makefile ===================================== @@ -0,0 +1,7 @@ +TOP=../../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +.PHONY: T25240 +T25240: + "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) T25240a.hs T25240.hs +RTS --optimistic-linking -RTS ===================================== testsuite/tests/ghci/linking/T25240/T25240.hs ===================================== @@ -0,0 +1,17 @@ +{-# LANGUAGE TemplateHaskell #-} + +{- + +When the RTS linker loads the T25240a module to run the pure foo splice, it +tries to resolve the func symbol even if this function isn't required to run the +splice code, i.e., its dead code. This test checks that by passing the +--optimistic-linking flag the RTS linker continues to link even in the presence +of unknown symbols. + +-} + +module T25240 where + +import T25240a + +$(pure foo) ===================================== testsuite/tests/ghci/linking/T25240/T25240.stderr ===================================== @@ -0,0 +1,2 @@ +ghc: ^^ Could not load 'func', dependency unresolved, optimistically continuing; assigning 'func' to 0xDEADBEEF. + ===================================== testsuite/tests/ghci/linking/T25240/T25240a.hs ===================================== @@ -0,0 +1,13 @@ +{-# LANGUAGE TemplateHaskell #-} + + +module T25240a + ( foo, func + ) where + + +foo :: [a] +foo = [] + +foreign import ccall "func" + func :: Int -> Int ===================================== testsuite/tests/ghci/linking/T25240/all.T ===================================== @@ -0,0 +1,2 @@ +test('T25240', [req_rts_linker, extra_files(['T25240a.hs'])], + makefile_test, ['T25240']) ===================================== testsuite/tests/ghci/should_run/T18064.stderr ===================================== @@ -1,2 +1,2 @@ -: ^^ Could not load 'blah', dependency unresolved. See top entry above. +: ^^ Could not load 'blah', dependency unresolved. See top entry above. You might consider using --optimistic-linking ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -234,7 +234,7 @@ module Control.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -303,7 +303,7 @@ module Control.Exception where module Control.Exception.Annotation where -- Safety: None type ExceptionAnnotation :: * -> Constraint - class ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where + class ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where displayExceptionAnnotation :: a -> GHC.Internal.Base.String default displayExceptionAnnotation :: GHC.Internal.Show.Show a => a -> GHC.Internal.Base.String {-# MINIMAL #-} @@ -345,7 +345,7 @@ module Control.Exception.Base where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -850,11 +850,11 @@ module Data.Data where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b constrFields :: Constr -> [GHC.Internal.Base.String] constrFixity :: Constr -> Fixity @@ -897,7 +897,7 @@ module Data.Data where showConstr :: Constr -> GHC.Internal.Base.String showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -921,14 +921,14 @@ module Data.Dynamic where -- Safety: Safe type Dynamic :: * data Dynamic where - Dynamic :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic + Dynamic :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} dynApp :: Dynamic -> Dynamic -> Dynamic dynApply :: Dynamic -> Dynamic -> GHC.Internal.Maybe.Maybe Dynamic - dynTypeRep :: Dynamic -> ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + dynTypeRep :: Dynamic -> ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep fromDyn :: forall a. Typeable a => Dynamic -> a -> a fromDynamic :: forall a. Typeable a => Dynamic -> GHC.Internal.Maybe.Maybe a toDyn :: forall a. Typeable a => a -> Dynamic @@ -1815,11 +1815,11 @@ module Data.Typeable where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b decT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Data.Either.Either ((a :~: b) -> GHC.Internal.Base.Void) (a :~: b) eqT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Maybe.Maybe (a :~: b) @@ -1834,7 +1834,7 @@ module Data.Typeable where rnfTypeRep :: TypeRep -> () showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -5255,7 +5255,7 @@ module GHC.Event where data EventManager = ... type FdKey :: * data FdKey - = ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] + = ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] ; GHC.Internal.Foreign.C.Types.N:CInt[0])GHC.Internal.System.Posix.Types.Fd, ...} type IOCallback :: * @@ -5288,9 +5288,9 @@ module GHC.Event.TimeOut where type TimeoutEdit :: * type TimeoutEdit = TimeoutQueue -> TimeoutQueue type TimeoutKey :: * - newtype TimeoutKey = TK ghc-internal-9.1001.0:GHC.Internal.Event.Unique.Unique + newtype TimeoutKey = TK ghc-internal-9.1100.0:GHC.Internal.Event.Unique.Unique type TimeoutQueue :: * - type TimeoutQueue = ghc-internal-9.1001.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback + type TimeoutQueue = ghc-internal-9.1100.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback module GHC.Exception where -- Safety: Safe @@ -5302,7 +5302,7 @@ module GHC.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -5331,7 +5331,7 @@ module GHC.Exception.Type where type ArithException :: * data ArithException = Overflow | Underflow | LossOfPrecision | DivideByZero | Denormal | RatioZeroDenominator type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -7884,8 +7884,8 @@ module GHC.IO.Handle where hTryLock :: Handle -> LockMode -> GHC.Types.IO GHC.Types.Bool hWaitForInput :: Handle -> GHC.Types.Int -> GHC.Types.IO GHC.Types.Bool isEOF :: GHC.Types.IO GHC.Types.Bool - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle nativeNewline :: Newline nativeNewlineMode :: NewlineMode noNewlineTranslation :: NewlineMode @@ -7935,11 +7935,11 @@ module GHC.IO.Handle.Internals where ioe_notReadable :: forall a. GHC.Types.IO a ioe_notWritable :: forall a. GHC.Types.IO a ioe_semiclosedHandle :: forall a. GHC.Types.IO a - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle openTextEncoding :: forall a. GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.HandleType -> (forall es ds. GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextEncoder es) -> GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextDecoder ds) -> GHC.Types.IO a) -> GHC.Types.IO a readTextDevice :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer readTextDeviceNonBlocking :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer @@ -8003,7 +8003,7 @@ module GHC.IO.Handle.Types where type Handle__ :: * data Handle__ = forall dev enc_state dec_state. - (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => + (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => Handle__ {haDevice :: !dev, haType :: HandleType, haByteBuffer :: ! {-# UNPACK #-}(GHC.Internal.IORef.N:IORef[0] _N)(GHC.Internal.IORef.IORef (GHC.Internal.IO.Buffer.Buffer GHC.Internal.Word.Word8)), @@ -9106,7 +9106,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * @@ -9417,7 +9417,7 @@ module GHC.StaticPtr where -- Safety: None type IsStatic :: (* -> *) -> Constraint class IsStatic p where - fromStaticPtr :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a + fromStaticPtr :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a {-# MINIMAL fromStaticPtr #-} type StaticKey :: * type StaticKey = GHC.Internal.Fingerprint.Type.Fingerprint @@ -10813,8 +10813,8 @@ module Type.Reflection where data (:~~:) a b where HRefl :: forall {k1} (a :: k1). (:~~:) a a pattern App :: forall k2 (t :: k2). () => forall k1 (a :: k1 -> k2) (b :: k1). (t ~ a b) => TypeRep a -> TypeRep b -> TypeRep t - pattern Con :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a - pattern Con' :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a + pattern Con :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a + pattern Con' :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a pattern Fun :: forall k (fun :: k). () => forall (r1 :: GHC.Types.RuntimeRep) (r2 :: GHC.Types.RuntimeRep) (arg :: TYPE r1) (res :: TYPE r2). (k ~ *, fun ~~ (arg -> res)) => TypeRep arg -> TypeRep res -> TypeRep fun type Module :: * data Module = ... @@ -10831,7 +10831,7 @@ module Type.Reflection where type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} decTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Data.Either.Either ((a :~~: b) -> GHC.Internal.Base.Void) (a :~~: b) eqTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Maybe.Maybe (a :~~: b) moduleName :: Module -> GHC.Internal.Base.String @@ -10866,9 +10866,9 @@ module Type.Reflection.Unsafe where data TypeRep a where ... mkTrApp :: forall k1 k2 (a :: k1 -> k2) (b :: k1). TypeRep a -> TypeRep b -> TypeRep (a b) - mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a + mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a mkTyCon :: GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Types.Int -> KindRep -> TyCon - someTypeRepFingerprint :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint + someTypeRepFingerprint :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConKindArgs :: TyCon -> GHC.Types.Int tyConKindRep :: TyCon -> KindRep @@ -11202,9 +11202,9 @@ instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.I instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Monoid (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Monoid (f p), GHC.Internal.Base.Monoid (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Monoid (f (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Monoid (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Monoid (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11260,9 +11260,9 @@ instance forall a. GHC.Internal.Base.Semigroup (Data.Semigroup.Last a) -- Define instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Semigroup (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Semigroup (f p), GHC.Internal.Base.Semigroup (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Semigroup (f (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11392,20 +11392,20 @@ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Last -- Defined instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Max -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Min -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.IO.Class.MonadIO GHC.Types.IO -- Defined in ‘GHC.Internal.Control.Monad.IO.Class’ -instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ -instance forall (m :: * -> *) a. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ +instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ +instance forall (m :: * -> *) a. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Functor.ZipList.ZipList a) -- Defined in ‘GHC.Internal.Functor.ZipList’ instance GHC.Internal.Data.Data.Data Data.Array.Byte.ByteArray -- Defined in ‘Data.Array.Byte’ -instance forall s. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ +instance forall s. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Complex.Complex a) -- Defined in ‘Data.Complex’ -instance forall i j (a :: i) (b :: j). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall i j (a :: i) (b :: j). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.All -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.Any -- Defined in ‘GHC.Internal.Data.Data’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b, GHC.Internal.Ix.Ix a) => GHC.Internal.Data.Data.Data (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.Associativity -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Bool -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Char -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k a (b :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall k a (b :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.DecidedStrictness -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Double -- Defined in ‘GHC.Internal.Data.Data’ @@ -11453,10 +11453,10 @@ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word32 -- Defined in ‘G instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word64 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word8 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Foreign.Ptr.WordPtr -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ -instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ +instance forall k (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ +instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b) => GHC.Internal.Data.Data.Data (Data.Semigroup.Arg a b) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.First a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.Last a) -- Defined in ‘Data.Semigroup’ @@ -11522,7 +11522,7 @@ instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SCha instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SSymbol -- Defined in ‘GHC.Internal.TypeLits’ instance forall k (a :: k). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k (a :: k1). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ -instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1). GHC.Internal.Data.Type.Equality.TestEquality f => GHC.Internal.Data.Type.Equality.TestEquality (Data.Functor.Compose.Compose f g) -- Defined in ‘Data.Functor.Compose’ instance forall a k (b :: k). GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ instance forall a. GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Bits.And a) -- Defined in ‘GHC.Internal.Data.Bits’ @@ -11719,7 +11719,7 @@ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Ba instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.RecUpdError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.TypeError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ -instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ instance [safe] GHC.Internal.Exception.Type.Exception System.Timeout.Timeout -- Defined in ‘System.Timeout’ instance forall a k (b :: k). GHC.Internal.Float.Floating a => GHC.Internal.Float.Floating (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ @@ -12390,8 +12390,8 @@ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataRep -- Defined in ‘ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataType -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Internal.Show.Show (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance GHC.Internal.Show.Show GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ instance forall a b. (GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). Data.Fixed.HasResolution a => GHC.Internal.Show.Show (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ @@ -12462,13 +12462,13 @@ instance GHC.Internal.Show.Show GHC.Internal.Foreign.C.Types.CWchar -- Defined i instance forall a. GHC.Internal.Show.Show (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Foreign.C.ConstPtr’ instance forall a b. (GHC.Internal.Ix.Ix a, GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Internal.Show.Show GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Internal.Show.Show GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ instance GHC.Internal.Show.Show GHC.Types.Double -- Defined in ‘GHC.Internal.Float’ instance GHC.Internal.Show.Show GHC.Types.Float -- Defined in ‘GHC.Internal.Float’ @@ -12502,7 +12502,7 @@ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Handle -- Defined i instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.HandleType -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Newline -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.NewlineMode -- Defined in ‘GHC.Internal.IO.Handle.Types’ -instance [safe] GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.HandlePosn -- Defined in ‘GHC.Internal.IO.Handle’ instance GHC.Internal.Show.Show GHC.Internal.IO.IOMode.IOMode -- Defined in ‘GHC.Internal.IO.IOMode’ instance GHC.Internal.Show.Show GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ @@ -12599,8 +12599,8 @@ instance GHC.Classes.Eq GHC.Internal.Data.Data.ConstrRep -- Defined in ‘GHC.In instance GHC.Classes.Eq GHC.Internal.Data.Data.DataRep -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Classes.Eq GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Classes.Eq (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Eq a, GHC.Classes.Eq b) => GHC.Classes.Eq (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Eq (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Eq (f (g a)) => GHC.Classes.Eq (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ @@ -12676,13 +12676,13 @@ instance forall a. GHC.Classes.Eq (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) - instance forall i e. (GHC.Internal.Ix.Ix i, GHC.Classes.Eq e) => GHC.Classes.Eq (GHC.Internal.Arr.Array i e) -- Defined in ‘GHC.Internal.Arr’ instance forall s i e. GHC.Classes.Eq (GHC.Internal.Arr.STArray s i e) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Classes.Eq GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ instance GHC.Classes.Eq GHC.Internal.Event.TimeOut.TimeoutKey -- Defined in ‘GHC.Internal.Event.TimeOut’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Classes.Eq GHC.Internal.Stack.Types.SrcLoc -- Defined in ‘GHC.Internal.Stack.Types’ instance GHC.Classes.Eq GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ instance GHC.Classes.Eq GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ @@ -12778,8 +12778,8 @@ instance GHC.Classes.Ord GHC.Internal.Unicode.GeneralCategory -- Defined in ‘G instance forall k (a :: k) (b :: k). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k2 (a :: k1) (b :: k2). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k (s :: k). GHC.Classes.Ord (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Ord ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Ord ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Ord a, GHC.Classes.Ord b) => GHC.Classes.Ord (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Ord (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Ord (f (g a)) => GHC.Classes.Ord (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -9330,7 +9330,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6f3894f7dd16b2b81544385ae91943175b57a5e7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6f3894f7dd16b2b81544385ae91943175b57a5e7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 13:40:34 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Tue, 17 Sep 2024 09:40:34 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] 26 commits: haddock: Re-organise cross-OS compatibility layer Message-ID: <66e986d2af693_23a21500dfc117273@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - de779f7a by doyougnu at 2024-09-17T09:39:45-04:00 linker: add --optimistic-linking flag This patch adds: - the --optimistic-linking flag which binds unknown symbols in the runtime linker to 0xDEADBEEF instead of exiting with failure - The test T25240 which tests these flags using dead code in the FFI system. - closes #25240 This patch is part of the upstreaming haskell.nix patches project. - - - - - 12 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6f3894f7dd16b2b81544385ae91943175b57a5e7...de779f7a3f81b08a1b06fbb37b4286fbec053100 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6f3894f7dd16b2b81544385ae91943175b57a5e7...de779f7a3f81b08a1b06fbb37b4286fbec053100 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 15:03:56 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 17 Sep 2024 11:03:56 -0400 Subject: [Git][ghc/ghc][master] RTS: expose closure_sizeW_ (#25252) Message-ID: <66e99a5c18d65_152688488fc810112@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 5 changed files: - rts/RtsSymbols.c - + testsuite/tests/th/T25252.hs - + testsuite/tests/th/T25252B.hs - + testsuite/tests/th/T25252_c.c - testsuite/tests/th/all.T Changes: ===================================== rts/RtsSymbols.c ===================================== @@ -954,6 +954,7 @@ extern char **environ; SymI_HasDataProto(stg_castDoubleToWord64zh) \ SymI_HasDataProto(stg_castWord32ToFloatzh) \ SymI_HasDataProto(stg_castFloatToWord32zh) \ + SymI_HasProto(closure_sizeW_) \ RTS_USER_SIGNALS_SYMBOLS \ RTS_INTCHAR_SYMBOLS ===================================== testsuite/tests/th/T25252.hs ===================================== @@ -0,0 +1,11 @@ +{-# LANGUAGE TemplateHaskell #-} + +module Main where + +import Data.Word +import Foreign.StablePtr +import Control.Monad.IO.Class +import T25252B + +main :: IO () +main = pure $(liftIO foo >> [| () |]) ===================================== testsuite/tests/th/T25252B.hs ===================================== @@ -0,0 +1,15 @@ +module T25252B where + +import Data.Word +import Foreign.StablePtr + +foreign import ccall "hs_custom_closureSize" closureSize :: StablePtr a -> Word64 + +foo :: IO () +foo = do + let + x :: [Int] + x = cycle [10,20] -- segfaults without "cycle"... + sp <- newStablePtr x + print (closureSize sp /= 0) + ===================================== testsuite/tests/th/T25252_c.c ===================================== @@ -0,0 +1,8 @@ +#include + + uint64_t hs_custom_closureSize(StgStablePtr const sp) { + StgPtr const root = deRefStablePtr(sp); + ASSERT(LOOKS_LIKE_CLOSURE_PTR(root)); + + return closure_sizeW((StgClosure*)root); +} ===================================== testsuite/tests/th/all.T ===================================== @@ -624,3 +624,9 @@ test('T24572c', normal, compile_fail, ['']) test('T24572d', normal, compile, ['']) test('T25209', normal, compile, ['-v0 -ddump-splices -dsuppress-uniques']) test('TH_MultilineStrings', normal, compile_and_run, ['']) +test('T25252', + [extra_files(['T25252B.hs', 'T25252_c.c']), + when(arch('i386'), expect_broken_for(25260,['ext-interp'])), + req_th, + req_c], + compile_and_run, ['-fPIC T25252_c.c']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d3b19851626411eeea35954ff745dfd7f663ac5c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d3b19851626411eeea35954ff745dfd7f663ac5c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 15:04:44 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 17 Sep 2024 11:04:44 -0400 Subject: [Git][ghc/ghc][master] 2 commits: HsExpr: Inline `HsWrap` into `WrapExpr` Message-ID: <66e99a8cb7c3b_152688488ff01077af@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - 17 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Zonk/Type.hs - ghc/GHCi/UI/Info.hs - testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr - testsuite/tests/plugins/simple-plugin/Simple/RemovePlugin.hs - + testsuite/tests/pmcheck/should_compile/T25257.hs - testsuite/tests/pmcheck/should_compile/all.T Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -569,7 +569,7 @@ mkExpandedStmtPopAt loc oStmt eExpr = mkPopErrCtxtExprAt loc $ mkExpandedStmtAt data XXExprGhcTc = WrapExpr -- Type and evidence application and abstractions - {-# UNPACK #-} !(HsWrap HsExpr) + HsWrapper (HsExpr GhcTc) | ExpandedThingTc -- See Note [Rebindable syntax and XXExprGhcRn] -- See Note [Expanding HsDo with XXExprGhcRn] in `GHC.Tc.Gen.Do` @@ -881,7 +881,7 @@ instance Outputable XXExprGhcRn where ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) instance Outputable XXExprGhcTc where - ppr (WrapExpr (HsWrap co_fn e)) + ppr (WrapExpr co_fn e) = pprHsWrapper co_fn (\_parens -> pprExpr e) ppr (ExpandedThingTc o e) @@ -922,7 +922,7 @@ ppr_infix_expr_rn (ExpandedThingRn thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a ppr_infix_expr_tc :: XXExprGhcTc -> Maybe SDoc -ppr_infix_expr_tc (WrapExpr (HsWrap _ e)) = ppr_infix_expr e +ppr_infix_expr_tc (WrapExpr _ e) = ppr_infix_expr e ppr_infix_expr_tc (ExpandedThingTc thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_tc (ConLikeTc {}) = Nothing ppr_infix_expr_tc (HsTick {}) = Nothing @@ -1025,7 +1025,7 @@ hsExprNeedsParens prec = go GhcRn -> go_x_rn x go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = hsExprNeedsParens prec e + go_x_tc (WrapExpr _ e) = hsExprNeedsParens prec e go_x_tc (ExpandedThingTc thing _) = hsExpandedNeedsParens thing go_x_tc (ConLikeTc {}) = False go_x_tc (HsTick _ (L _ e)) = hsExprNeedsParens prec e @@ -1077,7 +1077,7 @@ isAtomicHsExpr (XExpr x) | GhcRn <- ghcPass @p = go_x_rn x where go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = isAtomicHsExpr e + go_x_tc (WrapExpr _ e) = isAtomicHsExpr e go_x_tc (ExpandedThingTc thing _) = isAtomicExpandedThingRn thing go_x_tc (ConLikeTc {}) = True go_x_tc (HsTick {}) = False ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -149,7 +149,7 @@ hsExprType (HsEmbTy x _) = dataConCantHappen x hsExprType (HsQual x _ _) = dataConCantHappen x hsExprType (HsForAll x _ _) = dataConCantHappen x hsExprType (HsFunArr x _ _ _) = dataConCantHappen x -hsExprType (XExpr (WrapExpr (HsWrap wrap e))) = hsWrapperType wrap $ hsExprType e +hsExprType (XExpr (WrapExpr wrap e)) = hsWrapperType wrap $ hsExprType e hsExprType (XExpr (ExpandedThingTc _ e)) = hsExprType e hsExprType (XExpr (ConLikeTc con _ _)) = conLikeType con hsExprType (XExpr (HsTick _ e)) = lhsExprType e ===================================== compiler/GHC/Hs/Utils.hs ===================================== @@ -780,7 +780,7 @@ mkLHsWrap co_fn (L loc e) = L loc (mkHsWrap co_fn e) mkHsWrap :: HsWrapper -> HsExpr GhcTc -> HsExpr GhcTc mkHsWrap co_fn e | isIdHsWrapper co_fn = e -mkHsWrap co_fn e = XExpr (WrapExpr $ HsWrap co_fn e) +mkHsWrap co_fn e = XExpr (WrapExpr co_fn e) mkHsWrapCo :: TcCoercionN -- A Nominal coercion a ~N b -> HsExpr GhcTc -> HsExpr GhcTc ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -970,8 +970,8 @@ warnUnusedBindValue fun arg@(L loc _) arg_ty fish_var :: LHsExpr GhcTc -> Maybe (SrcSpan , Id) fish_var (L l (HsVar _ id)) = return (locA l, unLoc id) fish_var (L _ (HsAppType _ e _)) = fish_var e - fish_var (L l (XExpr (WrapExpr (HsWrap _ e)))) = do (l, e') <- fish_var (L l e) - return (l, e') + fish_var (L l (XExpr (WrapExpr _ e))) = do (l, e') <- fish_var (L l e) + return (l, e') fish_var (L l (XExpr (ExpandedThingTc _ e))) = fish_var (L l e) fish_var _ = Nothing @@ -1019,7 +1019,7 @@ dsHsWrapped orig_hs_expr where go wrap (HsPar _ (L _ hs_e)) = go wrap hs_e - go wrap1 (XExpr (WrapExpr (HsWrap wrap2 hs_e))) + go wrap1 (XExpr (WrapExpr wrap2 hs_e)) = go (wrap1 <.> wrap2) hs_e go wrap (HsAppType ty (L _ hs_e) _) = go (wrap <.> WpTyApp ty) hs_e ===================================== compiler/GHC/HsToCore/Match.hs ===================================== @@ -1175,7 +1175,7 @@ viewLExprEq (e1,_) (e2,_) = lexp e1 e2 exp e (HsPar _ (L _ e')) = exp e e' -- because the expressions do not necessarily have the same type, -- we have to compare the wrappers - exp (XExpr (WrapExpr (HsWrap h e))) (XExpr (WrapExpr (HsWrap h' e'))) = + exp (XExpr (WrapExpr h e)) (XExpr (WrapExpr h' e')) = wrap h h' && exp e e' exp (XExpr (ExpandedThingTc o x)) (XExpr (ExpandedThingTc o' x')) | isHsThingRnExpr o ===================================== compiler/GHC/HsToCore/Match/Literal.hs ===================================== @@ -447,7 +447,7 @@ getLHsIntegralLit (L _ e) = go e go (XExpr (HsBinTick _ _ e)) = getLHsIntegralLit e -- The literal might be wrapped in a case with -XOverloadedLists - go (XExpr (WrapExpr (HsWrap _ e))) = go e + go (XExpr (WrapExpr _ e)) = go e go _ = Nothing -- | If 'Integral', extract the value and type of the overloaded literal. ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -189,6 +189,7 @@ pmcMatches origin ctxt vars matches = {-# SCC "pmcMatches" #-} do Just matches -> do matches <- {-# SCC "desugarMatches" #-} noCheckDs $ desugarMatches vars matches + tracePm "desugared matches" (ppr matches) result <- {-# SCC "checkMatchGroup" #-} unCA (checkMatchGroup matches) missing tracePm "}: " (ppr (cr_uncov result)) ===================================== compiler/GHC/HsToCore/Pmc/Desugar.hs ===================================== @@ -27,7 +27,7 @@ import GHC.Types.Id import GHC.Core.ConLike import GHC.Types.Name import GHC.Builtin.Types -import GHC.Builtin.Names (rationalTyConName) +import GHC.Builtin.Names (rationalTyConName, toListName) import GHC.Types.SrcLoc import GHC.Utils.Outputable import GHC.Utils.Panic @@ -44,7 +44,6 @@ import GHC.Core.TyCo.Rep import GHC.Core.TyCo.Compare( eqType ) import GHC.Core.Type import GHC.Data.Maybe -import qualified GHC.LanguageExtensions as LangExt import GHC.Types.SourceText (FractionalLit(..)) import Control.Monad (zipWithM, replicateM) import Data.List (elemIndex) @@ -125,21 +124,24 @@ desugarPat x pat = case pat of XPat ext -> case ext of ExpansionPat orig expansion -> do - dflags <- getDynFlags case orig of -- We add special logic for overloaded list patterns. When: -- - a ViewPat is the expansion of a ListPat, - -- - RebindableSyntax is off, -- - the type of the pattern is the built-in list type, -- then we assume that the view function, 'toList', is the identity. -- This improves pattern-match overload checks, as this will allow -- the pattern match checker to directly inspect the inner pattern. -- See #14547, and Note [Desugaring overloaded list patterns] (Wrinkle). ListPat {} - | ViewPat arg_ty _lexpr pat <- expansion - , not (xopt LangExt.RebindableSyntax dflags) + | ViewPat arg_ty lrhs pat <- expansion , Just tc <- tyConAppTyCon_maybe arg_ty , tc == listTyCon + -- `pat` looks like `coerce toList -> [p1,...,pn]`. + -- Now take care of -XRebindableSyntax: + , let is_to_list (HsVar _ (L _ to_list)) = idName to_list == toListName + is_to_list (XExpr (WrapExpr _ e)) = is_to_list e + is_to_list _ = False + , is_to_list (unLoc lrhs) -> desugarLPat x pat _ -> desugarPat x expansion ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -588,8 +588,8 @@ addTickHsExpr (HsProc x pat cmdtop) = liftM2 (HsProc x) (addTickLPat pat) (traverse (addTickHsCmdTop) cmdtop) -addTickHsExpr (XExpr (WrapExpr (HsWrap w e))) = - liftM (XExpr . WrapExpr . HsWrap w) $ +addTickHsExpr (XExpr (WrapExpr w e)) = + liftM (XExpr . WrapExpr w) $ (addTickHsExpr e) -- Explicitly no tick on inside addTickHsExpr (XExpr (ExpandedThingTc o e)) = addTickHsExpanded o e ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1328,7 +1328,7 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where XExpr x | HieTc <- hiePass @p -> case x of - WrapExpr (HsWrap w a) + WrapExpr w a -> [ toHie $ L mspan a , toHie (L mspan w) ] ExpandedThingTc _ e ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -333,7 +333,7 @@ We expect to add to this list as we deal with more patterns via the expansion mechanism. Note [Desugaring overloaded list patterns] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If OverloadedLists is enabled, we desugar a list pattern to a view pattern: [p1, p2, p3] @@ -350,11 +350,12 @@ See Note [Invertible view patterns] in GHC.Tc.TyCl.PatSyn for details. == Wrinkle == This is all fine, except in one very specific case: - - when RebindableSyntax is off, - - and the type being matched on is already a list type. - -In this case, it is undesirable to desugar an overloaded list pattern into -a view pattern. To illustrate, consider the following program: +When the type being matched on is already a list type, so that the +pattern looks like + toList @[ty] dict -> pat +then we know for certain that `toList` is an identity function, so we can +behave exactly as if the pattern was just `pat`. This is important when +we have `OverloadedLists`. For example (#14547, #25257) > {-# LANGUAGE OverloadedLists #-} > @@ -375,6 +376,8 @@ as it isn't able to look through view patterns. We can see that this is silly: as we are matching on a list, `toList` doesn't actually do anything. So we ignore it, and desugar the pattern to an explicit list pattern, instead of a view pattern. +(NB: Because of -XRebindableSyntax we have to check that the `toList` we see is +actually resolved to `GHC.Exts.toList`.) Note however that this is not necessarily sound, because it is possible to have a list `l` such that `toList l` is not the same as `l`. ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -1080,10 +1080,10 @@ zonkExpr (HsQual x _ _) = dataConCantHappen x zonkExpr (HsForAll x _ _) = dataConCantHappen x zonkExpr (HsFunArr x _ _ _) = dataConCantHappen x -zonkExpr (XExpr (WrapExpr (HsWrap co_fn expr))) +zonkExpr (XExpr (WrapExpr co_fn expr)) = runZonkBndrT (zonkCoFn co_fn) $ \ new_co_fn -> do new_expr <- zonkExpr expr - return (XExpr (WrapExpr (HsWrap new_co_fn new_expr))) + return (XExpr (WrapExpr new_co_fn new_expr)) zonkExpr (XExpr (ExpandedThingTc thing e)) = do e' <- zonkExpr e ===================================== ghc/GHCi/UI/Info.hs ===================================== @@ -355,8 +355,8 @@ processAllTypeCheckedModule tcm mid | HsVar _ (L _ i) <- unwrapVar (unLoc e) = Just i | otherwise = Nothing - unwrapVar (XExpr (WrapExpr (HsWrap _ var))) = var - unwrapVar e' = e' + unwrapVar (XExpr (WrapExpr _ var)) = var + unwrapVar e' = e' -- | Extract 'Id', 'SrcSpan', and 'Type' for 'LPats's getTypeLPat :: LPat GhcTc -> Maybe (Maybe Id,SrcSpan,Type) ===================================== testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr ===================================== @@ -1512,16 +1512,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1575,16 +1574,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1638,16 +1636,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1674,16 +1671,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))))))))))))))))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))))))))))))))))))) ,(L (EpAnn (EpaSpan { }) @@ -1750,16 +1746,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))))))) ,(L (EpAnn (EpaSpan { }) ===================================== testsuite/tests/plugins/simple-plugin/Simple/RemovePlugin.hs ===================================== @@ -53,9 +53,9 @@ typecheckPlugin [name, "typecheck"] _ tc typecheckPlugin _ _ tc = return tc metaPlugin' :: [CommandLineOption] -> LHsExpr GhcTc -> TcM (LHsExpr GhcTc) -metaPlugin' [name, "meta"] (L l (XExpr (WrapExpr (HsWrap w (HsPar _ (L _ (HsApp noExt (L _ (HsVar _ (L _ id))) e))))))) +metaPlugin' [name, "meta"] (L l (XExpr (WrapExpr w (HsPar _ (L _ (HsApp noExt (L _ (HsVar _ (L _ id))) e)))))) | occNameString (getOccName id) == name - = return (L l (XExpr (WrapExpr (HsWrap w (unLoc e))))) + = return (L l (XExpr (WrapExpr w (unLoc e)))) -- The test should always match this first case. If the desugaring changes -- again in the future then the panic is more useful than the previous -- inscrutable failure. ===================================== testsuite/tests/pmcheck/should_compile/T25257.hs ===================================== @@ -0,0 +1,11 @@ +{-# LANGUAGE OverloadedLists, RebindableSyntax #-} +{-# OPTIONS -Wincomplete-patterns #-} + +module T25257 where + +import Prelude (Bool(..), IO, print) +import GHC.Exts (IsList(fromListN, toList)) + +null :: [a] -> Bool +null [] = True +null (_:_) = False ===================================== testsuite/tests/pmcheck/should_compile/all.T ===================================== @@ -172,3 +172,4 @@ test('DsIncompleteRecSel2', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel3', [collect_compiler_stats('bytes allocated', 10)], compile, ['-Wincomplete-record-selectors']) test('DoubleMatch', normal, compile, [overlapping_incomplete]) test('T24817', normal, compile, [overlapping_incomplete]) +test('T25257', normal, compile, [overlapping_incomplete]) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d3b19851626411eeea35954ff745dfd7f663ac5c...7fd9e5e29ab54eb406880077463e8552e2ddd39a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d3b19851626411eeea35954ff745dfd7f663ac5c...7fd9e5e29ab54eb406880077463e8552e2ddd39a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 15:35:39 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 17 Sep 2024 11:35:39 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 5 commits: RTS: expose closure_sizeW_ (#25252) Message-ID: <66e9a1cb2180a_1526889ac9dc1131d7@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - 0fef9baa by Ben Gamari at 2024-09-17T11:35:21-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 3e00c410 by Sylvain Henry at 2024-09-17T11:35:25-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 26 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Zonk/Type.hs - compiler/GHC/Types/Var.hs - ghc/GHCi/UI/Info.hs - m4/fptools_set_haskell_platform_vars.m4 - rts/RtsSymbols.c - + testsuite/tests/ffi/should_compile/T25255.hs - testsuite/tests/ffi/should_compile/all.T - testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr - testsuite/tests/plugins/simple-plugin/Simple/RemovePlugin.hs - + testsuite/tests/pmcheck/should_compile/T25257.hs - testsuite/tests/pmcheck/should_compile/all.T - + testsuite/tests/th/T25252.hs - + testsuite/tests/th/T25252B.hs - + testsuite/tests/th/T25252_c.c - testsuite/tests/th/all.T Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -569,7 +569,7 @@ mkExpandedStmtPopAt loc oStmt eExpr = mkPopErrCtxtExprAt loc $ mkExpandedStmtAt data XXExprGhcTc = WrapExpr -- Type and evidence application and abstractions - {-# UNPACK #-} !(HsWrap HsExpr) + HsWrapper (HsExpr GhcTc) | ExpandedThingTc -- See Note [Rebindable syntax and XXExprGhcRn] -- See Note [Expanding HsDo with XXExprGhcRn] in `GHC.Tc.Gen.Do` @@ -881,7 +881,7 @@ instance Outputable XXExprGhcRn where ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) instance Outputable XXExprGhcTc where - ppr (WrapExpr (HsWrap co_fn e)) + ppr (WrapExpr co_fn e) = pprHsWrapper co_fn (\_parens -> pprExpr e) ppr (ExpandedThingTc o e) @@ -922,7 +922,7 @@ ppr_infix_expr_rn (ExpandedThingRn thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a ppr_infix_expr_tc :: XXExprGhcTc -> Maybe SDoc -ppr_infix_expr_tc (WrapExpr (HsWrap _ e)) = ppr_infix_expr e +ppr_infix_expr_tc (WrapExpr _ e) = ppr_infix_expr e ppr_infix_expr_tc (ExpandedThingTc thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_tc (ConLikeTc {}) = Nothing ppr_infix_expr_tc (HsTick {}) = Nothing @@ -1025,7 +1025,7 @@ hsExprNeedsParens prec = go GhcRn -> go_x_rn x go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = hsExprNeedsParens prec e + go_x_tc (WrapExpr _ e) = hsExprNeedsParens prec e go_x_tc (ExpandedThingTc thing _) = hsExpandedNeedsParens thing go_x_tc (ConLikeTc {}) = False go_x_tc (HsTick _ (L _ e)) = hsExprNeedsParens prec e @@ -1077,7 +1077,7 @@ isAtomicHsExpr (XExpr x) | GhcRn <- ghcPass @p = go_x_rn x where go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr (HsWrap _ e)) = isAtomicHsExpr e + go_x_tc (WrapExpr _ e) = isAtomicHsExpr e go_x_tc (ExpandedThingTc thing _) = isAtomicExpandedThingRn thing go_x_tc (ConLikeTc {}) = True go_x_tc (HsTick {}) = False ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -149,7 +149,7 @@ hsExprType (HsEmbTy x _) = dataConCantHappen x hsExprType (HsQual x _ _) = dataConCantHappen x hsExprType (HsForAll x _ _) = dataConCantHappen x hsExprType (HsFunArr x _ _ _) = dataConCantHappen x -hsExprType (XExpr (WrapExpr (HsWrap wrap e))) = hsWrapperType wrap $ hsExprType e +hsExprType (XExpr (WrapExpr wrap e)) = hsWrapperType wrap $ hsExprType e hsExprType (XExpr (ExpandedThingTc _ e)) = hsExprType e hsExprType (XExpr (ConLikeTc con _ _)) = conLikeType con hsExprType (XExpr (HsTick _ e)) = lhsExprType e ===================================== compiler/GHC/Hs/Utils.hs ===================================== @@ -780,7 +780,7 @@ mkLHsWrap co_fn (L loc e) = L loc (mkHsWrap co_fn e) mkHsWrap :: HsWrapper -> HsExpr GhcTc -> HsExpr GhcTc mkHsWrap co_fn e | isIdHsWrapper co_fn = e -mkHsWrap co_fn e = XExpr (WrapExpr $ HsWrap co_fn e) +mkHsWrap co_fn e = XExpr (WrapExpr co_fn e) mkHsWrapCo :: TcCoercionN -- A Nominal coercion a ~N b -> HsExpr GhcTc -> HsExpr GhcTc ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -970,8 +970,8 @@ warnUnusedBindValue fun arg@(L loc _) arg_ty fish_var :: LHsExpr GhcTc -> Maybe (SrcSpan , Id) fish_var (L l (HsVar _ id)) = return (locA l, unLoc id) fish_var (L _ (HsAppType _ e _)) = fish_var e - fish_var (L l (XExpr (WrapExpr (HsWrap _ e)))) = do (l, e') <- fish_var (L l e) - return (l, e') + fish_var (L l (XExpr (WrapExpr _ e))) = do (l, e') <- fish_var (L l e) + return (l, e') fish_var (L l (XExpr (ExpandedThingTc _ e))) = fish_var (L l e) fish_var _ = Nothing @@ -1019,7 +1019,7 @@ dsHsWrapped orig_hs_expr where go wrap (HsPar _ (L _ hs_e)) = go wrap hs_e - go wrap1 (XExpr (WrapExpr (HsWrap wrap2 hs_e))) + go wrap1 (XExpr (WrapExpr wrap2 hs_e)) = go (wrap1 <.> wrap2) hs_e go wrap (HsAppType ty (L _ hs_e) _) = go (wrap <.> WpTyApp ty) hs_e ===================================== compiler/GHC/HsToCore/Match.hs ===================================== @@ -1175,7 +1175,7 @@ viewLExprEq (e1,_) (e2,_) = lexp e1 e2 exp e (HsPar _ (L _ e')) = exp e e' -- because the expressions do not necessarily have the same type, -- we have to compare the wrappers - exp (XExpr (WrapExpr (HsWrap h e))) (XExpr (WrapExpr (HsWrap h' e'))) = + exp (XExpr (WrapExpr h e)) (XExpr (WrapExpr h' e')) = wrap h h' && exp e e' exp (XExpr (ExpandedThingTc o x)) (XExpr (ExpandedThingTc o' x')) | isHsThingRnExpr o ===================================== compiler/GHC/HsToCore/Match/Literal.hs ===================================== @@ -447,7 +447,7 @@ getLHsIntegralLit (L _ e) = go e go (XExpr (HsBinTick _ _ e)) = getLHsIntegralLit e -- The literal might be wrapped in a case with -XOverloadedLists - go (XExpr (WrapExpr (HsWrap _ e))) = go e + go (XExpr (WrapExpr _ e)) = go e go _ = Nothing -- | If 'Integral', extract the value and type of the overloaded literal. ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -189,6 +189,7 @@ pmcMatches origin ctxt vars matches = {-# SCC "pmcMatches" #-} do Just matches -> do matches <- {-# SCC "desugarMatches" #-} noCheckDs $ desugarMatches vars matches + tracePm "desugared matches" (ppr matches) result <- {-# SCC "checkMatchGroup" #-} unCA (checkMatchGroup matches) missing tracePm "}: " (ppr (cr_uncov result)) ===================================== compiler/GHC/HsToCore/Pmc/Desugar.hs ===================================== @@ -27,7 +27,7 @@ import GHC.Types.Id import GHC.Core.ConLike import GHC.Types.Name import GHC.Builtin.Types -import GHC.Builtin.Names (rationalTyConName) +import GHC.Builtin.Names (rationalTyConName, toListName) import GHC.Types.SrcLoc import GHC.Utils.Outputable import GHC.Utils.Panic @@ -44,7 +44,6 @@ import GHC.Core.TyCo.Rep import GHC.Core.TyCo.Compare( eqType ) import GHC.Core.Type import GHC.Data.Maybe -import qualified GHC.LanguageExtensions as LangExt import GHC.Types.SourceText (FractionalLit(..)) import Control.Monad (zipWithM, replicateM) import Data.List (elemIndex) @@ -125,21 +124,24 @@ desugarPat x pat = case pat of XPat ext -> case ext of ExpansionPat orig expansion -> do - dflags <- getDynFlags case orig of -- We add special logic for overloaded list patterns. When: -- - a ViewPat is the expansion of a ListPat, - -- - RebindableSyntax is off, -- - the type of the pattern is the built-in list type, -- then we assume that the view function, 'toList', is the identity. -- This improves pattern-match overload checks, as this will allow -- the pattern match checker to directly inspect the inner pattern. -- See #14547, and Note [Desugaring overloaded list patterns] (Wrinkle). ListPat {} - | ViewPat arg_ty _lexpr pat <- expansion - , not (xopt LangExt.RebindableSyntax dflags) + | ViewPat arg_ty lrhs pat <- expansion , Just tc <- tyConAppTyCon_maybe arg_ty , tc == listTyCon + -- `pat` looks like `coerce toList -> [p1,...,pn]`. + -- Now take care of -XRebindableSyntax: + , let is_to_list (HsVar _ (L _ to_list)) = idName to_list == toListName + is_to_list (XExpr (WrapExpr _ e)) = is_to_list e + is_to_list _ = False + , is_to_list (unLoc lrhs) -> desugarLPat x pat _ -> desugarPat x expansion ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -588,8 +588,8 @@ addTickHsExpr (HsProc x pat cmdtop) = liftM2 (HsProc x) (addTickLPat pat) (traverse (addTickHsCmdTop) cmdtop) -addTickHsExpr (XExpr (WrapExpr (HsWrap w e))) = - liftM (XExpr . WrapExpr . HsWrap w) $ +addTickHsExpr (XExpr (WrapExpr w e)) = + liftM (XExpr . WrapExpr w) $ (addTickHsExpr e) -- Explicitly no tick on inside addTickHsExpr (XExpr (ExpandedThingTc o e)) = addTickHsExpanded o e ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1328,7 +1328,7 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where XExpr x | HieTc <- hiePass @p -> case x of - WrapExpr (HsWrap w a) + WrapExpr w a -> [ toHie $ L mspan a , toHie (L mspan w) ] ExpandedThingTc _ e ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -333,7 +333,7 @@ We expect to add to this list as we deal with more patterns via the expansion mechanism. Note [Desugaring overloaded list patterns] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If OverloadedLists is enabled, we desugar a list pattern to a view pattern: [p1, p2, p3] @@ -350,11 +350,12 @@ See Note [Invertible view patterns] in GHC.Tc.TyCl.PatSyn for details. == Wrinkle == This is all fine, except in one very specific case: - - when RebindableSyntax is off, - - and the type being matched on is already a list type. - -In this case, it is undesirable to desugar an overloaded list pattern into -a view pattern. To illustrate, consider the following program: +When the type being matched on is already a list type, so that the +pattern looks like + toList @[ty] dict -> pat +then we know for certain that `toList` is an identity function, so we can +behave exactly as if the pattern was just `pat`. This is important when +we have `OverloadedLists`. For example (#14547, #25257) > {-# LANGUAGE OverloadedLists #-} > @@ -375,6 +376,8 @@ as it isn't able to look through view patterns. We can see that this is silly: as we are matching on a list, `toList` doesn't actually do anything. So we ignore it, and desugar the pattern to an explicit list pattern, instead of a view pattern. +(NB: Because of -XRebindableSyntax we have to check that the `toList` we see is +actually resolved to `GHC.Exts.toList`.) Note however that this is not necessarily sound, because it is possible to have a list `l` such that `toList l` is not the same as `l`. ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -1080,10 +1080,10 @@ zonkExpr (HsQual x _ _) = dataConCantHappen x zonkExpr (HsForAll x _ _) = dataConCantHappen x zonkExpr (HsFunArr x _ _ _) = dataConCantHappen x -zonkExpr (XExpr (WrapExpr (HsWrap co_fn expr))) +zonkExpr (XExpr (WrapExpr co_fn expr)) = runZonkBndrT (zonkCoFn co_fn) $ \ new_co_fn -> do new_expr <- zonkExpr expr - return (XExpr (WrapExpr (HsWrap new_co_fn new_expr))) + return (XExpr (WrapExpr new_co_fn new_expr)) zonkExpr (XExpr (ExpandedThingTc thing e)) = do e' <- zonkExpr e ===================================== compiler/GHC/Types/Var.hs ===================================== @@ -343,9 +343,12 @@ arbitrary value which will (and must!) be ignored. -} instance Outputable Var where - ppr var = sdocOption sdocSuppressVarKinds $ \supp_var_kinds -> + ppr var = docWithStyle ppr_code ppr_normal + where + -- don't display debug info with Code style (#25255) + ppr_code = ppr (varName var) + ppr_normal sty = sdocOption sdocSuppressVarKinds $ \supp_var_kinds -> getPprDebug $ \debug -> - getPprStyle $ \sty -> let ppr_var = case var of (TyVar {}) ===================================== ghc/GHCi/UI/Info.hs ===================================== @@ -355,8 +355,8 @@ processAllTypeCheckedModule tcm mid | HsVar _ (L _ i) <- unwrapVar (unLoc e) = Just i | otherwise = Nothing - unwrapVar (XExpr (WrapExpr (HsWrap _ var))) = var - unwrapVar e' = e' + unwrapVar (XExpr (WrapExpr _ var)) = var + unwrapVar e' = e' -- | Extract 'Id', 'SrcSpan', and 'Type' for 'LPats's getTypeLPat :: LPat GhcTc -> Maybe (Maybe Id,SrcSpan,Type) ===================================== m4/fptools_set_haskell_platform_vars.m4 ===================================== @@ -154,12 +154,14 @@ AC_DEFUN([GHC_SUBSECTIONS_VIA_SYMBOLS], dnl See Note [autoconf assembler checks and -flto] AC_LINK_IFELSE( [AC_LANG_PROGRAM([], [__asm__ (".subsections_via_symbols");])], - [AC_MSG_RESULT(yes) + [ if test x"$TargetArch" = xaarch64; then dnl subsections via symbols is busted on arm64 TargetHasSubsectionsViaSymbols=NO + AC_MSG_RESULT([no, subsections-via-symbols are broken on AArch64/Darwin (GHC 24962)]) else TargetHasSubsectionsViaSymbols=YES + AC_MSG_RESULT(yes) fi ], [TargetHasSubsectionsViaSymbols=NO ===================================== rts/RtsSymbols.c ===================================== @@ -954,6 +954,7 @@ extern char **environ; SymI_HasDataProto(stg_castDoubleToWord64zh) \ SymI_HasDataProto(stg_castWord32ToFloatzh) \ SymI_HasDataProto(stg_castFloatToWord32zh) \ + SymI_HasProto(closure_sizeW_) \ RTS_USER_SIGNALS_SYMBOLS \ RTS_INTCHAR_SYMBOLS ===================================== testsuite/tests/ffi/should_compile/T25255.hs ===================================== @@ -0,0 +1,6 @@ +module T25255 where + +foreign export ccall foo :: Int -> Int + +foo :: Int -> Int +foo x = x + 10 ===================================== testsuite/tests/ffi/should_compile/all.T ===================================== @@ -42,3 +42,4 @@ test('T22043', normal, compile, ['']) test('T22774', unless(js_arch() or arch('wasm32'), expect_fail), compile, ['']) test('T24034', normal, compile, ['']) +test('T25255', normal, compile, ['-dppr-debug']) ===================================== testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr ===================================== @@ -1512,16 +1512,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1575,16 +1574,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1638,16 +1636,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))) (L (EpAnn (EpaSpan { }) @@ -1674,16 +1671,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))))))))))))))))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))))))))))))))))))) ,(L (EpAnn (EpaSpan { }) @@ -1750,16 +1746,15 @@ [])) (XExpr (WrapExpr - (HsWrap - (WpTyApp - (TyConApp - ({abstract:TyCon}) - [])) - (XExpr - (ConLikeTc - ({abstract:ConLike}) - [] - [])))))))))) + (WpTyApp + (TyConApp + ({abstract:TyCon}) + [])) + (XExpr + (ConLikeTc + ({abstract:ConLike}) + [] + []))))))))) ,(L (EpAnn (EpaSpan { }) ===================================== testsuite/tests/plugins/simple-plugin/Simple/RemovePlugin.hs ===================================== @@ -53,9 +53,9 @@ typecheckPlugin [name, "typecheck"] _ tc typecheckPlugin _ _ tc = return tc metaPlugin' :: [CommandLineOption] -> LHsExpr GhcTc -> TcM (LHsExpr GhcTc) -metaPlugin' [name, "meta"] (L l (XExpr (WrapExpr (HsWrap w (HsPar _ (L _ (HsApp noExt (L _ (HsVar _ (L _ id))) e))))))) +metaPlugin' [name, "meta"] (L l (XExpr (WrapExpr w (HsPar _ (L _ (HsApp noExt (L _ (HsVar _ (L _ id))) e)))))) | occNameString (getOccName id) == name - = return (L l (XExpr (WrapExpr (HsWrap w (unLoc e))))) + = return (L l (XExpr (WrapExpr w (unLoc e)))) -- The test should always match this first case. If the desugaring changes -- again in the future then the panic is more useful than the previous -- inscrutable failure. ===================================== testsuite/tests/pmcheck/should_compile/T25257.hs ===================================== @@ -0,0 +1,11 @@ +{-# LANGUAGE OverloadedLists, RebindableSyntax #-} +{-# OPTIONS -Wincomplete-patterns #-} + +module T25257 where + +import Prelude (Bool(..), IO, print) +import GHC.Exts (IsList(fromListN, toList)) + +null :: [a] -> Bool +null [] = True +null (_:_) = False ===================================== testsuite/tests/pmcheck/should_compile/all.T ===================================== @@ -172,3 +172,4 @@ test('DsIncompleteRecSel2', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel3', [collect_compiler_stats('bytes allocated', 10)], compile, ['-Wincomplete-record-selectors']) test('DoubleMatch', normal, compile, [overlapping_incomplete]) test('T24817', normal, compile, [overlapping_incomplete]) +test('T25257', normal, compile, [overlapping_incomplete]) ===================================== testsuite/tests/th/T25252.hs ===================================== @@ -0,0 +1,11 @@ +{-# LANGUAGE TemplateHaskell #-} + +module Main where + +import Data.Word +import Foreign.StablePtr +import Control.Monad.IO.Class +import T25252B + +main :: IO () +main = pure $(liftIO foo >> [| () |]) ===================================== testsuite/tests/th/T25252B.hs ===================================== @@ -0,0 +1,15 @@ +module T25252B where + +import Data.Word +import Foreign.StablePtr + +foreign import ccall "hs_custom_closureSize" closureSize :: StablePtr a -> Word64 + +foo :: IO () +foo = do + let + x :: [Int] + x = cycle [10,20] -- segfaults without "cycle"... + sp <- newStablePtr x + print (closureSize sp /= 0) + ===================================== testsuite/tests/th/T25252_c.c ===================================== @@ -0,0 +1,8 @@ +#include + + uint64_t hs_custom_closureSize(StgStablePtr const sp) { + StgPtr const root = deRefStablePtr(sp); + ASSERT(LOOKS_LIKE_CLOSURE_PTR(root)); + + return closure_sizeW((StgClosure*)root); +} ===================================== testsuite/tests/th/all.T ===================================== @@ -624,3 +624,9 @@ test('T24572c', normal, compile_fail, ['']) test('T24572d', normal, compile, ['']) test('T25209', normal, compile, ['-v0 -ddump-splices -dsuppress-uniques']) test('TH_MultilineStrings', normal, compile_and_run, ['']) +test('T25252', + [extra_files(['T25252B.hs', 'T25252_c.c']), + when(arch('i386'), expect_broken_for(25260,['ext-interp'])), + req_th, + req_c], + compile_and_run, ['-fPIC T25252_c.c']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6d4a2a041727147c9b6cedf089124151cf32d7ed...3e00c410f68105e7cb92457527e721118c4d3d49 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6d4a2a041727147c9b6cedf089124151cf32d7ed...3e00c410f68105e7cb92457527e721118c4d3d49 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 16:05:17 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Tue, 17 Sep 2024 12:05:17 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/supersven/ghc-toolchain-fuse-ld-bfd-flag Message-ID: <66e9a8bd18bae_2d66c798530620e8@gitlab.mail> Sven Tennie pushed new branch wip/supersven/ghc-toolchain-fuse-ld-bfd-flag at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/supersven/ghc-toolchain-fuse-ld-bfd-flag You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 20:45:33 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 17 Sep 2024 16:45:33 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 3 commits: configure: Correctly report when subsections-via-symbols is disabled Message-ID: <66e9ea6d64045_2b6f2ebc5c07042@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: e436ce2d by Ben Gamari at 2024-09-17T16:44:52-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 8905a6c0 by Mario Blažević at 2024-09-17T16:44:57-04:00 Modified the default export implementation to match the amended spec - - - - - c6e6749b by Sylvain Henry at 2024-09-17T16:45:09-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9 changed files: - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/Types/Var.hs - m4/fptools_set_haskell_platform_vars.m4 - − testsuite/tests/default/DefaultImport04.stdout - testsuite/tests/default/DefaultImport04.hs → testsuite/tests/default/DefaultImportFail07.hs - + testsuite/tests/default/DefaultImportFail07.stderr - testsuite/tests/default/all.T - + testsuite/tests/ffi/should_compile/T25255.hs - testsuite/tests/ffi/should_compile/all.T Changes: ===================================== compiler/GHC/Tc/Gen/Export.hs ===================================== @@ -42,7 +42,7 @@ import GHC.Types.SrcLoc as SrcLoc import GHC.Types.Name import GHC.Types.Name.Env import GHC.Types.Name.Set -import GHC.Types.DefaultEnv (ClassDefaults (cd_class, cd_module), DefaultEnv, +import GHC.Types.DefaultEnv (ClassDefaults (cd_class), DefaultEnv, emptyDefaultEnv, filterDefaultEnv, isEmptyDefaultEnv) import GHC.Types.Avail import GHC.Types.SourceFile @@ -192,7 +192,6 @@ rnExports explicit_mod exports ; let dflags = hsc_dflags hsc_env TcGblEnv { tcg_mod = this_mod , tcg_rdr_env = rdr_env - , tcg_default = defaults , tcg_imports = imports , tcg_warns = warns , tcg_src = hsc_src } = tcg_env @@ -237,15 +236,7 @@ rnExports explicit_mod exports Nothing -> Nothing Just _ -> map drop_defaults <$> rn_exports , tcg_default_exports = case exports of - Nothing -> - if xopt LangExt.NamedDefaults dflags then - -- NamedDefaults on: implicitly export the defaults declared in this module. - -- Test case: default/DefaultImport04.hs - filterDefaultEnv ((Just this_mod ==) . cd_module) defaults - else - -- NamedDefaults off: do not export any defaults (fixes #25206). - -- Test case: default/T25206.hs - emptyDefaultEnv + Nothing -> emptyDefaultEnv _ -> foldMap (foldMap sndOf3) rn_exports , tcg_dus = tcg_dus tcg_env `plusDU` usesOnly final_ns @@ -265,18 +256,17 @@ type DontWarnExportNames = NameEnv (NE.NonEmpty SrcSpan) ~~~~~~~~~~~~~~~~~~~~~~~~~ Named default declarations (see Note [Named default declarations] in GHC.Tc.Gen.Default) can be exported. A named default declaration is -exported when - -* there is no export list, and we export all locally-declared defaults - -* or it is specified in the export list, using the `default` keyword - and the class name. For example: +exported only when it's specified in the export list, using the `default` +keyword and the class name. For example: module TextWrap (Text, default IsString) where import Data.String (IsString) import Data.Text (Text) default IsString (Text, String) +A module with no explicit export list does not export any default +declarations, and neither does the re-export of a whole imported module. + The export item `default IsString` is parsed into the `IE` item IEThingAbs ext (L loc (IEDefault ext "IsString")) doc ===================================== compiler/GHC/Types/Var.hs ===================================== @@ -343,9 +343,12 @@ arbitrary value which will (and must!) be ignored. -} instance Outputable Var where - ppr var = sdocOption sdocSuppressVarKinds $ \supp_var_kinds -> + ppr var = docWithStyle ppr_code ppr_normal + where + -- don't display debug info with Code style (#25255) + ppr_code = ppr (varName var) + ppr_normal sty = sdocOption sdocSuppressVarKinds $ \supp_var_kinds -> getPprDebug $ \debug -> - getPprStyle $ \sty -> let ppr_var = case var of (TyVar {}) ===================================== m4/fptools_set_haskell_platform_vars.m4 ===================================== @@ -154,12 +154,14 @@ AC_DEFUN([GHC_SUBSECTIONS_VIA_SYMBOLS], dnl See Note [autoconf assembler checks and -flto] AC_LINK_IFELSE( [AC_LANG_PROGRAM([], [__asm__ (".subsections_via_symbols");])], - [AC_MSG_RESULT(yes) + [ if test x"$TargetArch" = xaarch64; then dnl subsections via symbols is busted on arm64 TargetHasSubsectionsViaSymbols=NO + AC_MSG_RESULT([no, subsections-via-symbols are broken on AArch64/Darwin (GHC 24962)]) else TargetHasSubsectionsViaSymbols=YES + AC_MSG_RESULT(yes) fi ], [TargetHasSubsectionsViaSymbols=NO ===================================== testsuite/tests/default/DefaultImport04.stdout deleted ===================================== @@ -1 +0,0 @@ -Product {getProduct = 1} ===================================== testsuite/tests/default/DefaultImport04.hs → testsuite/tests/default/DefaultImportFail07.hs ===================================== ===================================== testsuite/tests/default/DefaultImportFail07.stderr ===================================== @@ -0,0 +1,32 @@ +[1 of 5] Compiling ExportShowSum ( ExportShowSum.hs, ExportShowSum.o ) +[2 of 5] Compiling ExportImplicitMonoidProduct ( ExportImplicitMonoidProduct.hs, ExportImplicitMonoidProduct.o ) +[3 of 5] Compiling ReExportShowSumModule ( ReExportShowSumModule.hs, ReExportShowSumModule.o ) +[4 of 5] Compiling Main ( DefaultImportFail07.hs, DefaultImportFail07.o ) +DefaultImportFail07.hs:6:8: error: [GHC-39999] + • Ambiguous type variable ‘a0’ arising from a use of ‘print’ + prevents the constraint ‘(Show a0)’ from being solved. + Probable fix: use a type annotation to specify what ‘a0’ should be. + Potentially matching instances: + instance Show Ordering -- Defined in ‘GHC.Internal.Show’ + instance Show Integer -- Defined in ‘GHC.Internal.Show’ + ...plus 25 others + ...plus 19 instances involving out-of-scope types + (use -fprint-potential-instances to see them all) + • In the expression: print mempty + In an equation for ‘main’: main = print mempty + +DefaultImportFail07.hs:6:14: error: [GHC-39999] + • Ambiguous type variable ‘a0’ arising from a use of ‘mempty’ + prevents the constraint ‘(Monoid a0)’ from being solved. + Probable fix: use a type annotation to specify what ‘a0’ should be. + Potentially matching instances: + instance Monoid a => Monoid (IO a) + -- Defined in ‘GHC.Internal.Base’ + instance Monoid Ordering -- Defined in ‘GHC.Internal.Base’ + ...plus 9 others + ...plus 7 instances involving out-of-scope types + (use -fprint-potential-instances to see them all) + • In the first argument of ‘print’, namely ‘mempty’ + In the expression: print mempty + In an equation for ‘main’: main = print mempty + ===================================== testsuite/tests/default/all.T ===================================== @@ -11,8 +11,6 @@ test('default11', normal, compile_and_run, ['']) test('DefaultImport01', [extra_files(['ExportMonoidSum.hs'])], multimod_compile_and_run, ['DefaultImport01', '']) test('DefaultImport02', [extra_files(['ExportMonoidProduct.hs', 'ExportMonoidSum.hs'])], multimod_compile_and_run, ['DefaultImport02', '']) test('DefaultImport03', [extra_files(['ExportMonoidSum.hs', 'ReExportMonoidSum.hs'])], multimod_compile_and_run, ['DefaultImport03', '']) -test('DefaultImport04', [extra_files(['ExportImplicitMonoidProduct.hs', 'ExportShowSum.hs', 'ReExportShowSumModule.hs'])], - multimod_compile_and_run, ['DefaultImport04', '']) test('DefaultImport05', [extra_files(['ExportBitsInt.hs'])], multimod_compile_and_run, ['DefaultImport05', '']) test('DefaultImport07', [extra_files(['ExportMonoidProduct.hs', 'ExportMonoidSum.hs'])], multimod_compile, ['DefaultImport07', '-Wtype-defaults']) test('DefaultImport08', [extra_files(['ExportMonoidProduct.hs'])], multimod_compile, ['DefaultImport08', '-Wtype-defaults']) @@ -22,6 +20,8 @@ test('DefaultImportFail03', [extra_files(['ExportMonoidSum.hs', 'UnExportMonoidS test('DefaultImportFail04', [extra_files(['ExportShowSum.hs', 'ReExportShowSumModule.hs'])], multimod_compile_fail, ['DefaultImportFail04', '']) test('DefaultImportFail05', [extra_files(['ExportMonoidProduct.hs', 'ExportShowSum.hs'])], multimod_compile_fail, ['DefaultImportFail05', '']) test('DefaultImportFail06', [extra_files(['ExportBitsInt.hs'])], multimod_compile_fail, ['DefaultImportFail06', '']) +test('DefaultImportFail07', [extra_files(['ExportImplicitMonoidProduct.hs', 'ExportShowSum.hs', 'ReExportShowSumModule.hs'])], + multimod_compile_fail, ['DefaultImportFail07', '']) test('default-fail01', normal, compile_fail, ['']) test('default-fail02', normal, compile_fail, ['']) test('default-fail03', normal, compile_fail, ['']) ===================================== testsuite/tests/ffi/should_compile/T25255.hs ===================================== @@ -0,0 +1,6 @@ +module T25255 where + +foreign export ccall foo :: Int -> Int + +foo :: Int -> Int +foo x = x + 10 ===================================== testsuite/tests/ffi/should_compile/all.T ===================================== @@ -42,3 +42,4 @@ test('T22043', normal, compile, ['']) test('T22774', unless(js_arch() or arch('wasm32'), expect_fail), compile, ['']) test('T24034', normal, compile, ['']) +test('T25255', normal, compile, ['-dppr-debug']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3e00c410f68105e7cb92457527e721118c4d3d49...c6e6749b2b6b984dba7a707f4a1c15970763fc9e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3e00c410f68105e7cb92457527e721118c4d3d49...c6e6749b2b6b984dba7a707f4a1c15970763fc9e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 17 21:13:42 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Tue, 17 Sep 2024 17:13:42 -0400 Subject: [Git][ghc/ghc][wip/T23479] Add limitations to unfloat string lits Message-ID: <66e9f10685b43_2b6f2e311a007839c@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: 251595d8 by Serge S. Gulin at 2024-09-18T00:13:30+03:00 Add limitations to unfloat string lits - - - - - 5 changed files: - compiler/GHC/StgToJS/CodeGen.hs - + compiler/GHC/StgToJS/Sinker/Collect.hs - + compiler/GHC/StgToJS/Sinker/Sinker.hs - compiler/GHC/StgToJS/Sinker.hs → compiler/GHC/StgToJS/Sinker/StringsUnfloat.hs - compiler/ghc.cabal.in Changes: ===================================== compiler/GHC/StgToJS/CodeGen.hs ===================================== @@ -21,7 +21,7 @@ import GHC.JS.Transform import GHC.JS.Optimizer import GHC.StgToJS.Arg -import GHC.StgToJS.Sinker +import GHC.StgToJS.Sinker.Sinker import GHC.StgToJS.Types import qualified GHC.StgToJS.Object as Object import GHC.StgToJS.Utils ===================================== compiler/GHC/StgToJS/Sinker/Collect.hs ===================================== @@ -0,0 +1,72 @@ +{-# LANGUAGE LambdaCase #-} + +module GHC.StgToJS.Sinker.Collect + ( collectArgsTop + , collectArgs + , selectUsedOnce + ) + where + +import GHC.Prelude +import GHC.Types.Unique.Set +import GHC.Stg.Syntax +import GHC.Types.Id +import GHC.Types.Unique + +-- | fold over all id in StgArg used at the top level in an StgRhsCon +collectArgsTop :: CgStgBinding -> [Id] +collectArgsTop = \case + StgNonRec _b r -> collectArgsTopRhs r + StgRec bs -> concatMap (collectArgsTopRhs . snd) bs + where + collectArgsTopRhs :: CgStgRhs -> [Id] + collectArgsTopRhs = \case + StgRhsCon _ccs _dc _mu _ticks args _typ -> concatMap collectArgsA args + StgRhsClosure {} -> [] + +-- | fold over all Id in StgArg in the AST +collectArgs :: CgStgBinding -> [Id] +collectArgs = \case + StgNonRec _b r -> collectArgsR r + StgRec bs -> concatMap (collectArgsR . snd) bs + where + collectArgsR :: CgStgRhs -> [Id] + collectArgsR = \case + StgRhsClosure _x0 _x1 _x2 _x3 e _typ -> collectArgsE e + StgRhsCon _ccs _con _mu _ticks args _typ -> concatMap collectArgsA args + + collectArgsAlt :: CgStgAlt -> [Id] + collectArgsAlt alt = collectArgsE (alt_rhs alt) + + collectArgsE :: CgStgExpr -> [Id] + collectArgsE = \case + StgApp x args + -> x : concatMap collectArgsA args + StgConApp _con _mn args _ts + -> concatMap collectArgsA args + StgOpApp _x args _t + -> concatMap collectArgsA args + StgCase e _b _a alts + -> collectArgsE e ++ concatMap collectArgsAlt alts + StgLet _x b e + -> collectArgs b ++ collectArgsE e + StgLetNoEscape _x b e + -> collectArgs b ++ collectArgsE e + StgTick _i e + -> collectArgsE e + StgLit _ + -> [] + +collectArgsA :: StgArg -> [Id] +collectArgsA = \case + StgVarArg i -> [i] + StgLitArg _ -> [] + +selectUsedOnce :: (Foldable t, Uniquable a) => t a -> UniqSet a +selectUsedOnce = fst . foldr g (emptyUniqSet, emptyUniqSet) + where + g i t@(once, mult) + | i `elementOfUniqSet` mult = t + | i `elementOfUniqSet` once + = (delOneFromUniqSet once i, addOneToUniqSet mult i) + | otherwise = (addOneToUniqSet once i, mult) ===================================== compiler/GHC/StgToJS/Sinker/Sinker.hs ===================================== @@ -0,0 +1,158 @@ +{-# LANGUAGE FlexibleContexts #-} +{-# LANGUAGE RankNTypes #-} +{-# LANGUAGE LambdaCase #-} + +module GHC.StgToJS.Sinker.Sinker (sinkPgm) where + +import GHC.Prelude +import GHC.Types.Unique.Set +import GHC.Types.Unique.FM +import GHC.Types.Var.Set +import GHC.Stg.Syntax +import GHC.Types.Id +import GHC.Types.Name +import GHC.Unit.Module +import GHC.Types.Literal +import GHC.Data.Graph.Directed +import GHC.StgToJS.Sinker.Collect +import GHC.StgToJS.Sinker.StringsUnfloat + +import GHC.Utils.Misc (partitionWith) +import GHC.StgToJS.Utils + +import Data.Char +import Data.List (partition) +import Data.Maybe +import Data.Bifunctor (Bifunctor (..)) +import GHC.Utils.Outputable (showSDocUnsafe, showPprUnsafe) +import qualified GHC.Utils.Trace as DT +import Data.ByteString (ByteString) + + +-- | Unfloat some top-level unexported things +-- +-- GHC floats constants to the top level. This is fine in native code, but with JS +-- they occupy some global variable name. We can unfloat some unexported things: +-- +-- - global constructors, as long as they're referenced only once by another global +-- constructor and are not in a recursive binding group +-- - literals (small literals may also be sunk if they are used more than once) +sinkPgm :: Module + -> [CgStgTopBinding] + -> (UniqFM Id CgStgExpr, [CgStgTopBinding]) +sinkPgm m pgm + = ( tracePpr "" id sunk + , traceDoc "" id (map StgTopLifted pgm''') + ++ traceDoc "" id stringLits + ) + where + selectLifted :: CgStgTopBinding -> Either CgStgBinding (Id, ByteString) + selectLifted (StgTopLifted b) = Left b + selectLifted (StgTopStringLit i b) = Right (i, b) + + (pgm', allStringLits) = partitionWith selectLifted pgm + usedOnceIds = selectUsedOnce $ concatMap collectArgs pgm' + + stringLitsUFM = listToUFM $ (\(i, b) -> (idName i, (i, b))) <$> allStringLits + (pgm'', usedStringLitNames) = + unfloatStringLits + (idName `mapUniqSet` usedOnceIds) + (snd `mapUFM` stringLitsUFM) + pgm' + + stringLitsWithoutUnfolded = + nonDetEltsUFM + $ delListFromUFM stringLitsUFM + $ nonDetEltsUniqSet usedStringLitNames + + stringLits = uncurry StgTopStringLit <$> stringLitsWithoutUnfolded + + (sunk, pgm''') = sinkPgm' m usedOnceIds $ traceDoc "" (map StgTopLifted) pgm'' + +sinkPgm' + :: Module + -- ^ the module, since we treat definitions from the current module + -- differently + -> IdSet + -- ^ the set of used once ids + -> [CgStgBinding] + -- ^ the bindings + -> (UniqFM Id CgStgExpr, [CgStgBinding]) + -- ^ a map with sunken replacements for nodes, for where the replacement + -- does not fit in the 'StgBinding' AST and the new bindings +sinkPgm' m usedOnceIds pgm = + let usedOnce = tracePpr "" id (collectTopLevelUsedOnce usedOnceIds pgm) + sinkables = listToUFM $ + (tracePpr "" id (concatMap alwaysSinkable pgm)) ++ + filter ((`elementOfUniqSet` usedOnce) . fst) (concatMap (onceSinkable m) pgm) + isSunkBind (StgNonRec b _e) | elemUFM b sinkables = True + isSunkBind _ = False + in (sinkables, filter (not . isSunkBind) $ topSortDecls m pgm) + +-- | always sinkable, values that may be duplicated in the generated code (e.g. +-- small literals) +alwaysSinkable :: CgStgBinding -> [(Id, CgStgExpr)] +alwaysSinkable (StgRec {}) = [] +alwaysSinkable (StgNonRec b rhs) = case rhs of + StgRhsClosure _ _ _ _ e@(StgLit l) _ + | isSmallSinkableLit l + , isLocal (tracePpr "" id b) + -> [(b,e)] + StgRhsCon _ccs dc cnum _ticks as@[StgLitArg l] _typ + | isSmallSinkableLit l + , isLocal (tracePpr "" id b) + , isUnboxableCon dc + -> [(b,StgConApp dc cnum as [])] + _ -> [] + +isSmallSinkableLit :: Literal -> Bool +isSmallSinkableLit (LitChar c) = ord c < 100000 +isSmallSinkableLit (LitNumber _ i) = abs i < 100000 +isSmallSinkableLit _ = False + + +-- | once sinkable: may be sunk, but duplication is not ok +onceSinkable :: Module -> CgStgBinding -> [(Id, CgStgExpr)] +onceSinkable _m (StgNonRec b rhs) + | Just e <- getSinkable rhs + , isLocal b = [(b,e)] + where + getSinkable = \case + StgRhsCon _ccs dc cnum _ticks args _typ -> Just (StgConApp dc cnum args []) + StgRhsClosure _ _ _ _ e@(StgLit{}) _typ -> Just e + _ -> Nothing +onceSinkable _ _ = [] + +-- | collect all idents used only once in an argument at the top level +-- and never anywhere else +collectTopLevelUsedOnce :: IdSet -> [CgStgBinding] -> IdSet +collectTopLevelUsedOnce usedOnceIds binds = intersectUniqSets usedOnceIds (selectUsedOnce top_args) + where + top_args = concatMap collectArgsTop binds + +isLocal :: Id -> Bool +isLocal i = isNothing (nameModule_maybe . idName $ i) && not (isExportedId i) + +-- | since we have sequential initialization, topsort the non-recursive +-- constructor bindings +topSortDecls :: Module -> [CgStgBinding] -> [CgStgBinding] +topSortDecls _m binds = rest ++ nr' + where + (nr, rest) = partition isNonRec binds + isNonRec StgNonRec{} = True + isNonRec _ = False + vs = map getV nr + keys = mkUniqSet (map node_key vs) + getV e@(StgNonRec b _) = DigraphNode e b [] + getV _ = error "topSortDecls: getV, unexpected binding" + collectDeps (StgNonRec b (StgRhsCon _cc _dc _cnum _ticks args _typ)) = + [ (i, b) | StgVarArg i <- args, i `elementOfUniqSet` keys ] + collectDeps _ = [] + g = graphFromVerticesAndAdjacency vs (concatMap collectDeps nr) + nr' | (not . null) [()| CyclicSCC _ <- stronglyConnCompG g] + = error "topSortDecls: unexpected cycle" + | otherwise = map node_payload (topologicalSortG g) + + +traceDoc head m v = (DT.trace . (++) head . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (m v) v +tracePpr head m v = (DT.trace . (++) head . showPprUnsafe) (m v) v ===================================== compiler/GHC/StgToJS/Sinker.hs → compiler/GHC/StgToJS/Sinker/StringsUnfloat.hs ===================================== @@ -1,204 +1,36 @@ -{-# LANGUAGE FlexibleContexts #-} -{-# LANGUAGE RankNTypes #-} {-# LANGUAGE LambdaCase #-} -module GHC.StgToJS.Sinker (sinkPgm) where +module GHC.StgToJS.Sinker.StringsUnfloat + ( unfloatStringLits + ) + where import GHC.Prelude import GHC.Types.Unique.Set import GHC.Types.Unique.FM -import GHC.Types.Var.Set import GHC.Stg.Syntax import GHC.Types.Id import GHC.Types.Name -import GHC.Unit.Module import GHC.Types.Literal -import GHC.Data.Graph.Directed - import GHC.Utils.Misc (partitionWith) -import GHC.StgToJS.Utils -import Data.Char -import Data.List (partition) -import Data.Maybe -import Data.Bifunctor (Bifunctor (..)) -import GHC.Utils.Outputable (showSDocUnsafe, showPprUnsafe) -import qualified GHC.Utils.Trace as DT +import Data.ByteString qualified as BS import Data.ByteString (ByteString) +import Data.Bifunctor (Bifunctor (..)) - --- | Unfloat some top-level unexported things --- --- GHC floats constants to the top level. This is fine in native code, but with JS --- they occupy some global variable name. We can unfloat some unexported things: --- --- - global constructors, as long as they're referenced only once by another global --- constructor and are not in a recursive binding group --- - literals (small literals may also be sunk if they are used more than once) -sinkPgm :: Module - -> [CgStgTopBinding] - -> (UniqFM Id CgStgExpr, [CgStgTopBinding]) -sinkPgm m pgm - = ( (DT.trace . (++) "" . showPprUnsafe) sunk sunk - , traceDoc "" (map StgTopLifted pgm''') (map StgTopLifted pgm''') - ++ traceDoc "" stringLits stringLits - ) - where - traceDoc head = DT.trace . (++) head . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts - - selectLifted :: CgStgTopBinding -> Either CgStgBinding (Id, ByteString) - selectLifted (StgTopLifted b) = Left b - selectLifted (StgTopStringLit i b) = Right (i, b) - - (pgm', stringLits') = partitionWith selectLifted pgm - stringLits = uncurry StgTopStringLit <$> stringLits' - - (pgm'', _usedStringLitNames) = unfloatStringLits (listToUFM $ first idName <$> stringLits') pgm' - (sunk, pgm''') = sinkPgm' m $ (DT.trace . (++) "" . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (map StgTopLifted pgm'') pgm'' - -sinkPgm' - :: Module - -- ^ the module, since we treat definitions from the current module - -- differently +unfloatStringLits + :: UniqSet Name + -> UniqFM Name ByteString -> [CgStgBinding] - -- ^ the bindings - -> (UniqFM Id CgStgExpr, [CgStgBinding]) - -- ^ a map with sunken replacements for nodes, for where the replacement - -- does not fit in the 'StgBinding' AST and the new bindings -sinkPgm' m pgm = - let usedOnce = collectUsedOnce pgm - sinkables = listToUFM $ - concatMap alwaysSinkable pgm ++ - filter ((`elementOfUniqSet` usedOnce) . fst) (concatMap (onceSinkable m) pgm) - isSunkBind (StgNonRec b _e) | elemUFM b sinkables = True - isSunkBind _ = False - in (sinkables, filter (not . isSunkBind) $ topSortDecls m pgm) - --- | always sinkable, values that may be duplicated in the generated code (e.g. --- small literals) -alwaysSinkable :: CgStgBinding -> [(Id, CgStgExpr)] -alwaysSinkable (StgRec {}) = [] -alwaysSinkable (StgNonRec b rhs) = case rhs of - StgRhsClosure _ _ _ _ e@(StgLit l) _ - | isSmallSinkableLit l - , isLocal b - -> [(b,e)] - StgRhsCon _ccs dc cnum _ticks as@[StgLitArg l] _typ - | isSmallSinkableLit l - , isLocal b - , isUnboxableCon dc - -> [(b,StgConApp dc cnum as [])] - _ -> [] - -isSmallSinkableLit :: Literal -> Bool -isSmallSinkableLit (LitChar c) = ord c < 100000 -isSmallSinkableLit (LitNumber _ i) = abs i < 100000 -isSmallSinkableLit _ = False - - --- | once sinkable: may be sunk, but duplication is not ok -onceSinkable :: Module -> CgStgBinding -> [(Id, CgStgExpr)] -onceSinkable _m (StgNonRec b rhs) - | Just e <- getSinkable rhs - , isLocal b = [(b,e)] - where - getSinkable = \case - StgRhsCon _ccs dc cnum _ticks args _typ -> Just (StgConApp dc cnum args []) - StgRhsClosure _ _ _ _ e@(StgLit{}) _typ -> Just e - _ -> Nothing -onceSinkable _ _ = [] - --- | collect all idents used only once in an argument at the top level --- and never anywhere else -collectUsedOnce :: [CgStgBinding] -> IdSet -collectUsedOnce binds = intersectUniqSets (usedOnce args) (usedOnce top_args) - where - top_args = concatMap collectArgsTop binds - args = concatMap collectArgs binds - usedOnce = fst . foldr g (emptyUniqSet, emptyUniqSet) - g i t@(once, mult) - | i `elementOfUniqSet` mult = t - | i `elementOfUniqSet` once - = (delOneFromUniqSet once i, addOneToUniqSet mult i) - | otherwise = (addOneToUniqSet once i, mult) - --- | fold over all id in StgArg used at the top level in an StgRhsCon -collectArgsTop :: CgStgBinding -> [Id] -collectArgsTop = \case - StgNonRec _b r -> collectArgsTopRhs r - StgRec bs -> concatMap (collectArgsTopRhs . snd) bs - -collectArgsTopRhs :: CgStgRhs -> [Id] -collectArgsTopRhs = \case - StgRhsCon _ccs _dc _mu _ticks args _typ -> concatMap collectArgsA args - StgRhsClosure {} -> [] - --- | fold over all Id in StgArg in the AST -collectArgs :: CgStgBinding -> [Id] -collectArgs = \case - StgNonRec _b r -> collectArgsR r - StgRec bs -> concatMap (collectArgsR . snd) bs - -collectArgsR :: CgStgRhs -> [Id] -collectArgsR = \case - StgRhsClosure _x0 _x1 _x2 _x3 e _typ -> collectArgsE e - StgRhsCon _ccs _con _mu _ticks args _typ -> concatMap collectArgsA args - -collectArgsAlt :: CgStgAlt -> [Id] -collectArgsAlt alt = collectArgsE (alt_rhs alt) - -collectArgsE :: CgStgExpr -> [Id] -collectArgsE = \case - StgApp x args - -> x : concatMap collectArgsA args - StgConApp _con _mn args _ts - -> concatMap collectArgsA args - StgOpApp _x args _t - -> concatMap collectArgsA args - StgCase e _b _a alts - -> collectArgsE e ++ concatMap collectArgsAlt alts - StgLet _x b e - -> collectArgs b ++ collectArgsE e - StgLetNoEscape _x b e - -> collectArgs b ++ collectArgsE e - StgTick _i e - -> collectArgsE e - StgLit _ - -> [] - -collectArgsA :: StgArg -> [Id] -collectArgsA = \case - StgVarArg i -> [i] - StgLitArg _ -> [] - -isLocal :: Id -> Bool -isLocal i = isNothing (nameModule_maybe . idName $ i) && not (isExportedId i) - --- | since we have sequential initialization, topsort the non-recursive --- constructor bindings -topSortDecls :: Module -> [CgStgBinding] -> [CgStgBinding] -topSortDecls _m binds = rest ++ nr' - where - (nr, rest) = partition isNonRec binds - isNonRec StgNonRec{} = True - isNonRec _ = False - vs = map getV nr - keys = mkUniqSet (map node_key vs) - getV e@(StgNonRec b _) = DigraphNode e b [] - getV _ = error "topSortDecls: getV, unexpected binding" - collectDeps (StgNonRec b (StgRhsCon _cc _dc _cnum _ticks args _typ)) = - [ (i, b) | StgVarArg i <- args, i `elementOfUniqSet` keys ] - collectDeps _ = [] - g = graphFromVerticesAndAdjacency vs (concatMap collectDeps nr) - nr' | (not . null) [()| CyclicSCC _ <- stronglyConnCompG g] - = error "topSortDecls: unexpected cycle" - | otherwise = map node_payload (topologicalSortG g) + -> ([CgStgBinding], UniqSet Name) +unfloatStringLits usedOnceStringLits stringLits = + unfloatStringLits' (selectStringLitsForUnfloat usedOnceStringLits stringLits) -- | We are doing attempts to unfloat string literals back to -- the call site. Further special JS optimizations -- can generate more performant operations over them. -unfloatStringLits :: UniqFM Name ByteString -> [CgStgBinding] -> ([CgStgBinding], UniqSet Name) -unfloatStringLits stringLits allBindings = (binderWithoutChanges ++ binderWithUnfloatedStringLit, actuallyUsedStringLitNames) +unfloatStringLits' :: UniqFM Name ByteString -> [CgStgBinding] -> ([CgStgBinding], UniqSet Name) +unfloatStringLits' stringLits allBindings = (binderWithoutChanges ++ binderWithUnfloatedStringLit, actuallyUsedStringLitNames) where (binderWithoutChanges, binderWithUnfloatedStringLitPairs) = partitionWith substituteStringLit allBindings @@ -302,3 +134,16 @@ unfloatStringLits stringLits allBindings = (binderWithoutChanges ++ binderWithUn unified = fst <$> combined names = unionManyUniqSets (snd <$> combined) in (unified, names) + +selectStringLitsForUnfloat :: UniqSet Name -> UniqFM Name ByteString -> UniqFM Name ByteString +selectStringLitsForUnfloat usedOnceStringLits stringLits = alwaysUnfloat `plusUFM` usedOnceUnfloat + where + alwaysUnfloat = alwaysUnfloatStringLits stringLits + usedOnceUnfloat = selectUsedOnceStringLits usedOnceStringLits stringLits + + alwaysUnfloatStringLits :: UniqFM Name ByteString -> UniqFM Name ByteString + alwaysUnfloatStringLits = filterUFM $ \b -> BS.length b < 3 + + selectUsedOnceStringLits :: UniqSet Name -> UniqFM Name ByteString -> UniqFM Name ByteString + selectUsedOnceStringLits usedOnceStringLits stringLits = + stringLits `intersectUFM` getUniqSet usedOnceStringLits ===================================== compiler/ghc.cabal.in ===================================== @@ -761,7 +761,9 @@ Library GHC.StgToJS.Regs GHC.StgToJS.Rts.Types GHC.StgToJS.Rts.Rts - GHC.StgToJS.Sinker + GHC.StgToJS.Sinker.Collect + GHC.StgToJS.Sinker.StringsUnfloat + GHC.StgToJS.Sinker.Sinker GHC.StgToJS.Stack GHC.StgToJS.StaticPtr GHC.StgToJS.Symbols View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/251595d8a2cf80a9655f4e3e9d1a49e5434f7656 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/251595d8a2cf80a9655f4e3e9d1a49e5434f7656 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 05:07:36 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Wed, 18 Sep 2024 01:07:36 -0400 Subject: [Git][ghc/ghc][wip/strings] 3 commits: Inline alexScanUser Message-ID: <66ea6018a62d9_2b6f2e163a1cc1056e4@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC Commits: fb6c36bd by Brandon Chinn at 2024-09-17T21:17:20-07:00 Inline alexScanUser - - - - - c3780088 by Brandon Chinn at 2024-09-17T21:17:23-07:00 Use dlist to reduce allocations - - - - - 4607c624 by Brandon Chinn at 2024-09-17T21:17:23-07:00 Duplicate processChars for specialization - - - - - 2 changed files: - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs Changes: ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -3551,6 +3551,11 @@ topNoLayoutContainsCommas [] = False topNoLayoutContainsCommas (ALRLayout _ _ : ls) = topNoLayoutContainsCommas ls topNoLayoutContainsCommas (ALRNoLayout b _ : _) = b +-- If the generated alexScan/alexScanUser functions are called multiple times +-- in this file, alexScanUser gets broken out into a separate function and +-- increases memory usage. Make sure GHC inlines this function and optimizes it. +{-# INLINE alexScanUser #-} + lexToken :: P (PsLocated Token) lexToken = do inp@(AI loc1 buf) <- getInput ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -18,11 +18,9 @@ import GHC.Prelude hiding (getChar) import Control.Arrow ((>>>)) import Control.Monad (when) import Data.Char (chr, ord) -import qualified Data.Foldable as Seq (toList) import qualified Data.Foldable1 as Foldable1 import qualified Data.List.NonEmpty as NonEmpty import Data.Maybe (listToMaybe, mapMaybe) -import qualified Data.Sequence as Seq import GHC.Data.StringBuffer (StringBuffer) import qualified GHC.Data.StringBuffer as StringBuffer import GHC.Parser.CharClass ( @@ -40,7 +38,7 @@ type BufPos = Int data StringLexError = StringLexError LexErr BufPos lexString :: Int -> StringBuffer -> Either StringLexError String -lexString = lexStringWith processChars +lexString = lexStringWith processChars processChars where processChars :: HasChar c => [c] -> Either (c, LexErr) [c] processChars = @@ -68,19 +66,24 @@ So what we'll do is do two passes. The first pass is optimistic; just convert to a plain String and process it. If this results in an error, we do a second pass, this time where each character is annotated with its position. Now, the error has all the information it needs. + +Ideally, lexStringWith would take a single (forall c. HasChar c => ...) function, +but to help the specializer, we pass it in twice to concretize it for the two +types we actually use. -} -- | See Note [Lexing strings] lexStringWith :: - (forall c. HasChar c => [c] -> Either (c, LexErr) [c]) + ([Char] -> Either (Char, LexErr) [Char]) + -> ([CharPos] -> Either (CharPos, LexErr) [CharPos]) -> Int -> StringBuffer -> Either StringLexError String -lexStringWith processChars len buf = +lexStringWith processChars processCharsPos len buf = case processChars $ bufferChars buf len of Right s -> Right s Left _ -> - case processChars $ bufferLocatedChars buf len of + case processCharsPos $ bufferLocatedChars buf len of Right _ -> panic "expected lex error on second pass" Left ((_, pos), e) -> Left $ StringLexError e pos @@ -103,7 +106,9 @@ pattern Char c <- (getChar -> c) bufferChars :: StringBuffer -> Int -> [Char] bufferChars = StringBuffer.lexemeToString -bufferLocatedChars :: StringBuffer -> Int -> [(Char, BufPos)] +type CharPos = (Char, BufPos) + +bufferLocatedChars :: StringBuffer -> Int -> [CharPos] bufferLocatedChars initialBuf len = go initialBuf where go buf @@ -133,17 +138,16 @@ collapseGaps = go [] -> panic "gap unexpectedly ended" resolveEscapes :: HasChar c => [c] -> Either (c, LexErr) [c] -resolveEscapes = go Seq.empty +resolveEscapes = go dlistEmpty where - -- FIXME.bchinn: see if dlist/reverselist improves performance go !acc = \case - [] -> pure $ Seq.toList acc + [] -> pure $ dlistToList acc Char '\\' : Char '&' : cs -> go acc cs backslash@(Char '\\') : cs -> case resolveEscapeChar cs of - Right (esc, cs') -> go (acc Seq.|> setChar esc backslash) cs' + Right (esc, cs') -> go (acc `dlistSnoc` setChar esc backslash) cs' Left (c, e) -> Left (c, e) - c : cs -> go (acc Seq.|> c) cs + c : cs -> go (acc `dlistSnoc` c) cs -- ----------------------------------------------------------------------------- -- Escape characters @@ -252,7 +256,7 @@ isSingleSmartQuote = \case -- and rejoining lines, and instead manually find newline characters, -- for performance. lexMultilineString :: Int -> StringBuffer -> Either StringLexError String -lexMultilineString = lexStringWith processChars +lexMultilineString = lexStringWith processChars processChars where processChars :: HasChar c => [c] -> Either (c, LexErr) [c] processChars = @@ -373,3 +377,17 @@ It's more precisely defined with the following algorithm: * Lines with only whitespace characters 3. Calculate the longest prefix of whitespace shared by all lines in the remaining list -} + +-- ----------------------------------------------------------------------------- +-- DList + +newtype DList a = DList ([a] -> [a]) + +dlistEmpty :: DList a +dlistEmpty = DList id + +dlistToList :: DList a -> [a] +dlistToList (DList f) = f [] + +dlistSnoc :: DList a -> a -> DList a +dlistSnoc (DList f) x = DList (f . (x :)) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d70e41ae439b6a6f2967b43d977b8af6a8febe63...4607c624e4cf4cf27ba9a9c443b2cd203a43b883 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d70e41ae439b6a6f2967b43d977b8af6a8febe63...4607c624e4cf4cf27ba9a9c443b2cd203a43b883 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 08:20:22 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 18 Sep 2024 04:20:22 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] Apply 3 suggestion(s) to 2 file(s) Message-ID: <66ea8d46242f4_1cc73e66362c2671@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: e88731dc by Rodrigo Mesquita at 2024-09-18T08:20:17+00:00 Apply 3 suggestion(s) to 2 file(s) Co-authored-by: Ben Gamari <ben at well-typed.com> - - - - - 2 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/UniqueRenamer.hs Changes: ===================================== compiler/GHC/Cmm.hs ===================================== @@ -378,10 +378,10 @@ CmmGroup uses LabelMap in two places: * In CmmProc for info tables * In CmmGraph for the blocks of the graph -LabelMap is not a deterministic strucutre, so traversing a LabelMap can process +LabelMap is not a deterministic structure, so traversing a LabelMap can process elements in different order (depending on the given uniques). -Therefore before we do the renaming we need to use a deterministic strucutre, one +Therefore before we do the renaming we need to use a deterministic structure, one which we can traverse in a guaranteed order. A list does the job perfectly. Once the renaming happens it is converted back into a LabelMap, which is now deterministic ===================================== compiler/GHC/Cmm/UniqueRenamer.hs ===================================== @@ -124,8 +124,8 @@ detRenameCmmGroup dufm group = swap (runState (mapM detRenameCmmDecl group) dufm regs' <- mapM detRenameGlobalReg regs lbl' <- detRenameCLabel lbl -- Rename the info table last! This is necessary for determinism, the - -- info table contents not always appear in the same order accross - -- runs. As long as all uniques have already been renamed in a deterministic + -- info table contents not always appear in the same order across runs due to the existence of a LabelMap in `CmmTopInfo`. + -- As long as all uniques have already been renamed in a deterministic -- order, renaming the info table uniques will only lookup the -- corresponding deterministic ones rather than creating any. h' <- detRenameCmmTop h View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e88731dc98c5f9271f92956e1eb0d60bb86760a7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e88731dc98c5f9271f92956e1eb0d60bb86760a7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 09:50:45 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Wed, 18 Sep 2024 05:50:45 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25272 Message-ID: <66eaa2754c513_1cc73ec54748472f8@gitlab.mail> Teo Camarasu pushed new branch wip/T25272 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25272 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:18:45 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Wed, 18 Sep 2024 07:18:45 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/25232 Message-ID: <66eab715ac1e9_3d77602bd220607fe@gitlab.mail> Zubin pushed new branch wip/25232 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/25232 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:31:57 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Wed, 18 Sep 2024 07:31:57 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] wrk: update T11223 golden test output Message-ID: <66eaba2d8ffba_3d776046297c664db@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: 5364cd98 by doyougnu at 2024-09-18T07:27:40-04:00 wrk: update T11223 golden test output - - - - - 6 changed files: - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-32-mingw32 - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-64-mingw32 - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-32-mingw32 - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-64-mingw32 Changes: ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-32-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-64-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-32-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-64-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5364cd98aa1dd9338305dc1320e6f421621671d8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5364cd98aa1dd9338305dc1320e6f421621671d8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:35:21 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Wed, 18 Sep 2024 07:35:21 -0400 Subject: [Git][ghc/ghc][wip/T25272] 6 commits: Linker: add support for extra built-in symbols (#25155) Message-ID: <66eabaf9a3985_3d77604222b4666de@gitlab.mail> Teo Camarasu pushed to branch wip/T25272 at Glasgow Haskell Compiler / GHC Commits: 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - eab798f8 by Teo Camarasu at 2024-09-18T12:35:05+01:00 Fix Lift Bytes and Lift ByteArray when using RebindableSyntax Previously these Lift instances could not be used with a custom fromInteger introduced by RebindableSyntax. We implement the same workaround as: https://github.com/haskell/text/pull/534 to fix this. Resolves #25272 - - - - - 30 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Zonk/Type.hs - docs/users_guide/ghci.rst - ghc/GHCi/UI/Info.hs - libraries/base/src/Data/Array/Byte.hs - libraries/ghc-internal/src/GHC/Internal/System/Environment/ExecutablePath.hsc - libraries/ghc-internal/src/GHC/Internal/TH/Lift.hs - rts/Linker.c - rts/LinkerInternals.h - rts/RtsSymbols.c - rts/RtsSymbols.h - rts/rts.cabal - testsuite/tests/ghci/linking/Makefile - + testsuite/tests/ghci/linking/T25155.hs - + testsuite/tests/ghci/linking/T25155.stdout - + testsuite/tests/ghci/linking/T25155_TH.hs - + testsuite/tests/ghci/linking/T25155_iserv.hs - + testsuite/tests/ghci/linking/T25155_iserv_main.c - testsuite/tests/ghci/linking/all.T - testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ea5f3088d74c801407322e642ecda60c905faae2...eab798f83f52ae5503787ba0080073c094b39847 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ea5f3088d74c801407322e642ecda60c905faae2...eab798f83f52ae5503787ba0080073c094b39847 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:52:49 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Wed, 18 Sep 2024 07:52:49 -0400 Subject: [Git][ghc/ghc][wip/25232] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66eabf11ed72f_3d77606bca9c7865a@gitlab.mail> Zubin pushed to branch wip/25232 at Glasgow Haskell Compiler / GHC Commits: 7516b16c by Zubin Duggal at 2024-09-18T17:22:40+05:30 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 4 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,17 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +957,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [extra_ways(['nonmoving', 'nonmoving_prof', 'nonmoving_thr_prof'])], compile_and_run, ['-threaded -with-rtsopts "-N2"']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7516b16c3c051ae46799686be317de0be482281b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7516b16c3c051ae46799686be317de0be482281b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:53:41 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Wed, 18 Sep 2024 07:53:41 -0400 Subject: [Git][ghc/ghc][wip/25232] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66eabf4582806_3d77606d709078946@gitlab.mail> Zubin pushed to branch wip/25232 at Glasgow Haskell Compiler / GHC Commits: ce854703 by Zubin Duggal at 2024-09-18T17:23:32+05:30 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 4 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,18 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + // Not doing this lead to #25232 and #23958 + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +958,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [extra_ways(['nonmoving', 'nonmoving_prof', 'nonmoving_thr_prof'])], compile_and_run, ['-threaded -with-rtsopts "-N2"']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ce8547035120d12d0ddee4cf7817496da8007dcc -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ce8547035120d12d0ddee4cf7817496da8007dcc You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:55:36 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Wed, 18 Sep 2024 07:55:36 -0400 Subject: [Git][ghc/ghc][wip/25232] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66eabfb819237_3d77606d6c088186b@gitlab.mail> Zubin pushed to branch wip/25232 at Glasgow Haskell Compiler / GHC Commits: 5ee34950 by Zubin Duggal at 2024-09-18T17:25:25+05:30 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 5 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - + testsuite/tests/rts/T25232.hs - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,18 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + // Not doing this lead to #25232 and #23958 + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +958,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): ===================================== testsuite/tests/rts/T25232.hs ===================================== @@ -0,0 +1,14 @@ +module Main where + +import Control.Monad +import Control.Monad.Trans.State.Strict +import Data.IORef + +main :: IO () +main = do + ref <- newIORef () + replicateM_ 1000 $ withIORef ref $ runStateT (pure ()) + +withIORef :: IORef a -> (a -> IO (b, a)) -> IO b +withIORef ref f = + readIORef ref >>= f >>= \(b, a) -> writeIORef ref a >> pure b ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [extra_ways(['nonmoving', 'nonmoving_prof', 'nonmoving_thr_prof'])], compile_and_run, ['-threaded -with-rtsopts "-N2"']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5ee34950c6449249d184d5f6c2de03b57f0e94a4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5ee34950c6449249d184d5f6c2de03b57f0e94a4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:56:13 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 18 Sep 2024 07:56:13 -0400 Subject: [Git][ghc/ghc][master] configure: Correctly report when subsections-via-symbols is disabled Message-ID: <66eabfdd3e0e_3d776092a88885260@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 1 changed file: - m4/fptools_set_haskell_platform_vars.m4 Changes: ===================================== m4/fptools_set_haskell_platform_vars.m4 ===================================== @@ -154,12 +154,14 @@ AC_DEFUN([GHC_SUBSECTIONS_VIA_SYMBOLS], dnl See Note [autoconf assembler checks and -flto] AC_LINK_IFELSE( [AC_LANG_PROGRAM([], [__asm__ (".subsections_via_symbols");])], - [AC_MSG_RESULT(yes) + [ if test x"$TargetArch" = xaarch64; then dnl subsections via symbols is busted on arm64 TargetHasSubsectionsViaSymbols=NO + AC_MSG_RESULT([no, subsections-via-symbols are broken on AArch64/Darwin (GHC 24962)]) else TargetHasSubsectionsViaSymbols=YES + AC_MSG_RESULT(yes) fi ], [TargetHasSubsectionsViaSymbols=NO View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e4169ba910cefbbf9e8fe5770c3f4b43f24bb509 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e4169ba910cefbbf9e8fe5770c3f4b43f24bb509 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:56:38 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 18 Sep 2024 07:56:38 -0400 Subject: [Git][ghc/ghc][master] Modified the default export implementation to match the amended spec Message-ID: <66eabff658cef_3d77606dfefc88553@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 5 changed files: - compiler/GHC/Tc/Gen/Export.hs - − testsuite/tests/default/DefaultImport04.stdout - testsuite/tests/default/DefaultImport04.hs → testsuite/tests/default/DefaultImportFail07.hs - + testsuite/tests/default/DefaultImportFail07.stderr - testsuite/tests/default/all.T Changes: ===================================== compiler/GHC/Tc/Gen/Export.hs ===================================== @@ -42,7 +42,7 @@ import GHC.Types.SrcLoc as SrcLoc import GHC.Types.Name import GHC.Types.Name.Env import GHC.Types.Name.Set -import GHC.Types.DefaultEnv (ClassDefaults (cd_class, cd_module), DefaultEnv, +import GHC.Types.DefaultEnv (ClassDefaults (cd_class), DefaultEnv, emptyDefaultEnv, filterDefaultEnv, isEmptyDefaultEnv) import GHC.Types.Avail import GHC.Types.SourceFile @@ -192,7 +192,6 @@ rnExports explicit_mod exports ; let dflags = hsc_dflags hsc_env TcGblEnv { tcg_mod = this_mod , tcg_rdr_env = rdr_env - , tcg_default = defaults , tcg_imports = imports , tcg_warns = warns , tcg_src = hsc_src } = tcg_env @@ -237,15 +236,7 @@ rnExports explicit_mod exports Nothing -> Nothing Just _ -> map drop_defaults <$> rn_exports , tcg_default_exports = case exports of - Nothing -> - if xopt LangExt.NamedDefaults dflags then - -- NamedDefaults on: implicitly export the defaults declared in this module. - -- Test case: default/DefaultImport04.hs - filterDefaultEnv ((Just this_mod ==) . cd_module) defaults - else - -- NamedDefaults off: do not export any defaults (fixes #25206). - -- Test case: default/T25206.hs - emptyDefaultEnv + Nothing -> emptyDefaultEnv _ -> foldMap (foldMap sndOf3) rn_exports , tcg_dus = tcg_dus tcg_env `plusDU` usesOnly final_ns @@ -265,18 +256,17 @@ type DontWarnExportNames = NameEnv (NE.NonEmpty SrcSpan) ~~~~~~~~~~~~~~~~~~~~~~~~~ Named default declarations (see Note [Named default declarations] in GHC.Tc.Gen.Default) can be exported. A named default declaration is -exported when - -* there is no export list, and we export all locally-declared defaults - -* or it is specified in the export list, using the `default` keyword - and the class name. For example: +exported only when it's specified in the export list, using the `default` +keyword and the class name. For example: module TextWrap (Text, default IsString) where import Data.String (IsString) import Data.Text (Text) default IsString (Text, String) +A module with no explicit export list does not export any default +declarations, and neither does the re-export of a whole imported module. + The export item `default IsString` is parsed into the `IE` item IEThingAbs ext (L loc (IEDefault ext "IsString")) doc ===================================== testsuite/tests/default/DefaultImport04.stdout deleted ===================================== @@ -1 +0,0 @@ -Product {getProduct = 1} ===================================== testsuite/tests/default/DefaultImport04.hs → testsuite/tests/default/DefaultImportFail07.hs ===================================== ===================================== testsuite/tests/default/DefaultImportFail07.stderr ===================================== @@ -0,0 +1,32 @@ +[1 of 5] Compiling ExportShowSum ( ExportShowSum.hs, ExportShowSum.o ) +[2 of 5] Compiling ExportImplicitMonoidProduct ( ExportImplicitMonoidProduct.hs, ExportImplicitMonoidProduct.o ) +[3 of 5] Compiling ReExportShowSumModule ( ReExportShowSumModule.hs, ReExportShowSumModule.o ) +[4 of 5] Compiling Main ( DefaultImportFail07.hs, DefaultImportFail07.o ) +DefaultImportFail07.hs:6:8: error: [GHC-39999] + • Ambiguous type variable ‘a0’ arising from a use of ‘print’ + prevents the constraint ‘(Show a0)’ from being solved. + Probable fix: use a type annotation to specify what ‘a0’ should be. + Potentially matching instances: + instance Show Ordering -- Defined in ‘GHC.Internal.Show’ + instance Show Integer -- Defined in ‘GHC.Internal.Show’ + ...plus 25 others + ...plus 19 instances involving out-of-scope types + (use -fprint-potential-instances to see them all) + • In the expression: print mempty + In an equation for ‘main’: main = print mempty + +DefaultImportFail07.hs:6:14: error: [GHC-39999] + • Ambiguous type variable ‘a0’ arising from a use of ‘mempty’ + prevents the constraint ‘(Monoid a0)’ from being solved. + Probable fix: use a type annotation to specify what ‘a0’ should be. + Potentially matching instances: + instance Monoid a => Monoid (IO a) + -- Defined in ‘GHC.Internal.Base’ + instance Monoid Ordering -- Defined in ‘GHC.Internal.Base’ + ...plus 9 others + ...plus 7 instances involving out-of-scope types + (use -fprint-potential-instances to see them all) + • In the first argument of ‘print’, namely ‘mempty’ + In the expression: print mempty + In an equation for ‘main’: main = print mempty + ===================================== testsuite/tests/default/all.T ===================================== @@ -11,8 +11,6 @@ test('default11', normal, compile_and_run, ['']) test('DefaultImport01', [extra_files(['ExportMonoidSum.hs'])], multimod_compile_and_run, ['DefaultImport01', '']) test('DefaultImport02', [extra_files(['ExportMonoidProduct.hs', 'ExportMonoidSum.hs'])], multimod_compile_and_run, ['DefaultImport02', '']) test('DefaultImport03', [extra_files(['ExportMonoidSum.hs', 'ReExportMonoidSum.hs'])], multimod_compile_and_run, ['DefaultImport03', '']) -test('DefaultImport04', [extra_files(['ExportImplicitMonoidProduct.hs', 'ExportShowSum.hs', 'ReExportShowSumModule.hs'])], - multimod_compile_and_run, ['DefaultImport04', '']) test('DefaultImport05', [extra_files(['ExportBitsInt.hs'])], multimod_compile_and_run, ['DefaultImport05', '']) test('DefaultImport07', [extra_files(['ExportMonoidProduct.hs', 'ExportMonoidSum.hs'])], multimod_compile, ['DefaultImport07', '-Wtype-defaults']) test('DefaultImport08', [extra_files(['ExportMonoidProduct.hs'])], multimod_compile, ['DefaultImport08', '-Wtype-defaults']) @@ -22,6 +20,8 @@ test('DefaultImportFail03', [extra_files(['ExportMonoidSum.hs', 'UnExportMonoidS test('DefaultImportFail04', [extra_files(['ExportShowSum.hs', 'ReExportShowSumModule.hs'])], multimod_compile_fail, ['DefaultImportFail04', '']) test('DefaultImportFail05', [extra_files(['ExportMonoidProduct.hs', 'ExportShowSum.hs'])], multimod_compile_fail, ['DefaultImportFail05', '']) test('DefaultImportFail06', [extra_files(['ExportBitsInt.hs'])], multimod_compile_fail, ['DefaultImportFail06', '']) +test('DefaultImportFail07', [extra_files(['ExportImplicitMonoidProduct.hs', 'ExportShowSum.hs', 'ReExportShowSumModule.hs'])], + multimod_compile_fail, ['DefaultImportFail07', '']) test('default-fail01', normal, compile_fail, ['']) test('default-fail02', normal, compile_fail, ['']) test('default-fail03', normal, compile_fail, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9d20a787490a3b758da89d60eac63c70578f5ad5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9d20a787490a3b758da89d60eac63c70578f5ad5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 11:57:23 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 18 Sep 2024 07:57:23 -0400 Subject: [Git][ghc/ghc][master] FFI: don't ppr Id/Var symbols with debug info (#25255) Message-ID: <66eac0233cb1e_3d77607118e4929bf@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 3 changed files: - compiler/GHC/Types/Var.hs - + testsuite/tests/ffi/should_compile/T25255.hs - testsuite/tests/ffi/should_compile/all.T Changes: ===================================== compiler/GHC/Types/Var.hs ===================================== @@ -343,9 +343,12 @@ arbitrary value which will (and must!) be ignored. -} instance Outputable Var where - ppr var = sdocOption sdocSuppressVarKinds $ \supp_var_kinds -> + ppr var = docWithStyle ppr_code ppr_normal + where + -- don't display debug info with Code style (#25255) + ppr_code = ppr (varName var) + ppr_normal sty = sdocOption sdocSuppressVarKinds $ \supp_var_kinds -> getPprDebug $ \debug -> - getPprStyle $ \sty -> let ppr_var = case var of (TyVar {}) ===================================== testsuite/tests/ffi/should_compile/T25255.hs ===================================== @@ -0,0 +1,6 @@ +module T25255 where + +foreign export ccall foo :: Int -> Int + +foo :: Int -> Int +foo x = x + 10 ===================================== testsuite/tests/ffi/should_compile/all.T ===================================== @@ -42,3 +42,4 @@ test('T22043', normal, compile, ['']) test('T22774', unless(js_arch() or arch('wasm32'), expect_fail), compile, ['']) test('T24034', normal, compile, ['']) +test('T25255', normal, compile, ['-dppr-debug']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/35eb4f428ab72b712ea78d6ef86b956e321c3bb2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/35eb4f428ab72b712ea78d6ef86b956e321c3bb2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 12:17:14 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Wed, 18 Sep 2024 08:17:14 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/persistent-worker-test-2024-09-10] compiler: implement --show-iface-abi-hash major mode Message-ID: <66eac4cab9288_1344ccc3474487cb@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/persistent-worker-test-2024-09-10 at Glasgow Haskell Compiler / GHC Commits: 4ab7031d by Cheng Shao at 2024-09-16T20:54:44+02:00 compiler: implement --show-iface-abi-hash major mode - - - - - 3 changed files: - compiler/GHC/Iface/Load.hs - docs/users_guide/expected-undocumented-flags.txt - ghc/Main.hs Changes: ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -33,7 +33,7 @@ module GHC.Iface.Load ( WhereFrom(..), pprModIfaceSimple, - ifaceStats, pprModIface, showIface, + ifaceStats, pprModIface, showIface, showIfaceAbiHash, module Iface_Errors -- avoids boot files in Ppr modules ) where @@ -1090,6 +1090,19 @@ showIface logger dflags unit_state name_cache filename = do $ withPprStyle (mkDumpStyle name_ppr_ctx) $ pprModIface unit_state iface +-- | Read binary interface, and print ABI hash. Unlike the +-- @--abi-hash@ major mode, the output is the ABI hash deserialized +-- from the interface directly. +showIfaceAbiHash :: Logger -> DynFlags -> NameCache -> FilePath -> IO () +showIfaceAbiHash logger dflags name_cache filename = do + let profile = targetProfile dflags + + -- skip the hi way check and silence warnings + iface <- readBinIface profile name_cache IgnoreHiWay QuietBinIFace filename + + logMsg logger MCDump noSrcSpan + $ ppr $ mi_mod_hash $ mi_final_exts iface + -- | Show a ModIface but don't display details; suitable for ModIfaces stored in -- the EPT. pprModIfaceSimple :: UnitState -> ModIface -> SDoc ===================================== docs/users_guide/expected-undocumented-flags.txt ===================================== @@ -1,6 +1,7 @@ -#include --abi-hash --backpack +--show-iface-abi-hash --show-packages -Onot -Walternative-layout-rule-transitional ===================================== ghc/Main.hs ===================================== @@ -291,6 +291,10 @@ main' postLoadMode units dflags0 args flagWarnings = do (hsc_units hsc_env) (hsc_NC hsc_env) f + ShowInterfaceAbiHash f -> liftIO $ showIfaceAbiHash logger + (hsc_dflags hsc_env) + (hsc_NC hsc_env) + f DoMake -> doMake units srcs DoMkDependHS -> doMkDependHS (map fst srcs) StopBefore p -> liftIO (oneShot hsc_env p srcs) @@ -481,6 +485,8 @@ isShowGhciUsageMode _ = False data PostLoadMode = ShowInterface FilePath -- ghc --show-iface + | ShowInterfaceAbiHash FilePath + -- ghc --show-iface-abi-hash | DoMkDependHS -- ghc -M | StopBefore StopPhase -- ghc -E | -C | -S -- StopBefore StopLn is the default @@ -505,6 +511,9 @@ showUnitsMode = mkPostLoadMode ShowPackages showInterfaceMode :: FilePath -> Mode showInterfaceMode fp = mkPostLoadMode (ShowInterface fp) +showInterfaceAbiHashMode :: FilePath -> Mode +showInterfaceAbiHashMode fp = mkPostLoadMode (ShowInterfaceAbiHash fp) + stopBeforeMode :: StopPhase -> Mode stopBeforeMode phase = mkPostLoadMode (StopBefore phase) @@ -639,6 +648,9 @@ mode_flags = [ defFlag "-show-iface" (HasArg (\f -> setMode (showInterfaceMode f) "--show-iface")) + , defFlag "-show-iface-abi-hash" (HasArg (\f -> setMode (showInterfaceAbiHashMode f) + "--show-iface-abi-hash")) + ------- primary modes ------------------------------------------------ , defFlag "c" (PassFlag (\f -> do setMode (stopBeforeMode NoStop) f addFlag "-no-link" f)) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4ab7031d16db4e94bd9eb195455d85c923b9177c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4ab7031d16db4e94bd9eb195455d85c923b9177c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 13:08:21 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Wed, 18 Sep 2024 09:08:21 -0400 Subject: [Git][ghc/ghc][wip/T25272] Fix Lift Bytes and Lift ByteArray when using RebindableSyntax Message-ID: <66ead0c567335_1344cc3e0da0557f3@gitlab.mail> Teo Camarasu pushed to branch wip/T25272 at Glasgow Haskell Compiler / GHC Commits: ecdaedac by Teo Camarasu at 2024-09-18T14:08:12+01:00 Fix Lift Bytes and Lift ByteArray when using RebindableSyntax Previously these Lift instances could not be used with a custom fromInteger introduced by RebindableSyntax. We implement the same workaround as: https://github.com/haskell/text/pull/534 to fix this. Resolves #25272 - - - - - 6 changed files: - docs/users_guide/9.12.1-notes.rst - libraries/base/src/Data/Array/Byte.hs - libraries/ghc-internal/src/GHC/Internal/TH/Lift.hs - + testsuite/tests/th/T25272.hs - testsuite/tests/th/TH_Lift.stderr - testsuite/tests/th/all.T Changes: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -159,6 +159,8 @@ Runtime system - The `deprecation process of GHC.Pack ` has come its term. The module has now been removed from ``base``. +- Fix `Lift ByteArray` instance when using a user overriden `fromInteger` function through `RebindableSyntax`. + ``ghc-prim`` library ~~~~~~~~~~~~~~~~~~~~ ===================================== libraries/base/src/Data/Array/Byte.hs ===================================== @@ -203,7 +203,7 @@ instance Show ByteArray where instance Lift ByteArray where liftTyped = unsafeCodeCoerce . lift lift (ByteArray b) = - [| addrToByteArray $(lift len) + [| addrToByteArray $(pure . LitE . IntPrimL $ fromIntegral len) $(pure . LitE . BytesPrimL $ Bytes ptr 0 (fromIntegral len)) |] where @@ -221,8 +221,8 @@ instance Lift ByteArray where ptr = ForeignPtr (byteArrayContents# pb) (PlainPtr (unsafeCoerce# pb)) {-# NOINLINE addrToByteArray #-} -addrToByteArray :: Int -> Addr# -> ByteArray -addrToByteArray (I# len) addr = runST $ ST $ +addrToByteArray :: Int# -> Addr# -> ByteArray +addrToByteArray len addr = runST $ ST $ \s -> case newByteArray# len s of (# s', mb #) -> case copyAddrToByteArray# addr mb 0# len s' of s'' -> case unsafeFreezeByteArray# mb s'' of ===================================== libraries/ghc-internal/src/GHC/Internal/TH/Lift.hs ===================================== @@ -350,8 +350,9 @@ instance Lift Bytes where lift bytes at Bytes{} = -- See Note [Why FinalPtr] [| Bytes { bytesPtr = ForeignPtr $(Lib.litE (BytesPrimL bytes)) FinalPtr - , bytesOffset = 0 - , bytesSize = $(lift (bytesSize bytes)) + -- we cannot use Lift Word here because of #25267 and #25272 + , bytesOffset = W# 0## + , bytesSize = W# $(Lib.litE . WordPrimL . fromIntegral $ bytesSize bytes) } |] -- | @since template-haskell-2.22.1.0 ===================================== testsuite/tests/th/T25272.hs ===================================== @@ -0,0 +1,26 @@ +{-# LANGUAGE RebindableSyntax #-} +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE UnboxedTuples #-} +module T25272 where +import Prelude (Integer, String, ($), (>>=), mempty) +import qualified Prelude +import GHC.Exts (ByteArray#) +import GHC.IO (IO(..)) +import Language.Haskell.TH.Syntax (lift, Bytes(..), runIO) +import Foreign.Ptr (nullPtr) +import Foreign.ForeignPtr (newForeignPtr_) +import Data.Array.Byte + +fromInteger :: Integer -> String +fromInteger _ = "yikes" + +x :: Bytes +x = $(do + let fromInteger = Prelude.fromInteger + b <- runIO $ newForeignPtr_ nullPtr + lift $ Bytes b 0 0 + ) + +y :: ByteArray +y = $(lift (mempty :: ByteArray)) ===================================== testsuite/tests/th/TH_Lift.stderr ===================================== @@ -182,7 +182,7 @@ TH_Lift.hs:(93,10)-(99,13): Splicing expression Bytes {bytesPtr = GHC.Internal.ForeignPtr.ForeignPtr "Hello"# GHC.Internal.ForeignPtr.FinalPtr, - bytesOffset = 0, bytesSize = 5} + bytesOffset = GHC.Types.W# 0##, bytesSize = GHC.Types.W# 5##} TH_Lift.hs:90:10-59: Splicing expression examineCode [|| 3 + 4 ||] `bindCode` liftTyped ======> ===================================== testsuite/tests/th/all.T ===================================== @@ -630,3 +630,4 @@ test('T25252', req_th, req_c], compile_and_run, ['-fPIC T25252_c.c']) +test('T25272', normal, compile, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ecdaedac2e240741ac36aff277422b0c9edb4653 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ecdaedac2e240741ac36aff277422b0c9edb4653 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 13:10:59 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Wed, 18 Sep 2024 09:10:59 -0400 Subject: [Git][ghc/ghc][wip/T25272] Fix Lift Bytes and Lift ByteArray when using RebindableSyntax Message-ID: <66ead16382c92_1344cc3cb75c582b7@gitlab.mail> Teo Camarasu pushed to branch wip/T25272 at Glasgow Haskell Compiler / GHC Commits: 821ec44c by Teo Camarasu at 2024-09-18T14:10:50+01:00 Fix Lift Bytes and Lift ByteArray when using RebindableSyntax Previously these Lift instances could not be used with a custom fromInteger introduced by RebindableSyntax. We implement the same workaround as: https://github.com/haskell/text/pull/534 to fix this. Resolves #25272 - - - - - 6 changed files: - docs/users_guide/9.12.1-notes.rst - libraries/base/src/Data/Array/Byte.hs - libraries/ghc-internal/src/GHC/Internal/TH/Lift.hs - + testsuite/tests/th/T25272.hs - testsuite/tests/th/TH_Lift.stderr - testsuite/tests/th/all.T Changes: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -159,6 +159,8 @@ Runtime system - The `deprecation process of GHC.Pack ` has come its term. The module has now been removed from ``base``. +- Fix `Lift ByteArray` instance when using a user overriden `fromInteger` function through `RebindableSyntax`. + ``ghc-prim`` library ~~~~~~~~~~~~~~~~~~~~ ===================================== libraries/base/src/Data/Array/Byte.hs ===================================== @@ -203,7 +203,8 @@ instance Show ByteArray where instance Lift ByteArray where liftTyped = unsafeCodeCoerce . lift lift (ByteArray b) = - [| addrToByteArray $(lift len) + -- we cannot use Lift Int here because of #25267 and #25272 + [| addrToByteArray $(pure . LitE . IntPrimL $ fromIntegral len) $(pure . LitE . BytesPrimL $ Bytes ptr 0 (fromIntegral len)) |] where @@ -221,8 +222,8 @@ instance Lift ByteArray where ptr = ForeignPtr (byteArrayContents# pb) (PlainPtr (unsafeCoerce# pb)) {-# NOINLINE addrToByteArray #-} -addrToByteArray :: Int -> Addr# -> ByteArray -addrToByteArray (I# len) addr = runST $ ST $ +addrToByteArray :: Int# -> Addr# -> ByteArray +addrToByteArray len addr = runST $ ST $ \s -> case newByteArray# len s of (# s', mb #) -> case copyAddrToByteArray# addr mb 0# len s' of s'' -> case unsafeFreezeByteArray# mb s'' of ===================================== libraries/ghc-internal/src/GHC/Internal/TH/Lift.hs ===================================== @@ -350,8 +350,9 @@ instance Lift Bytes where lift bytes at Bytes{} = -- See Note [Why FinalPtr] [| Bytes { bytesPtr = ForeignPtr $(Lib.litE (BytesPrimL bytes)) FinalPtr - , bytesOffset = 0 - , bytesSize = $(lift (bytesSize bytes)) + -- we cannot use Lift Word here because of #25267 and #25272 + , bytesOffset = W# 0## + , bytesSize = W# $(Lib.litE . WordPrimL . fromIntegral $ bytesSize bytes) } |] -- | @since template-haskell-2.22.1.0 ===================================== testsuite/tests/th/T25272.hs ===================================== @@ -0,0 +1,26 @@ +{-# LANGUAGE RebindableSyntax #-} +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE UnboxedTuples #-} +module T25272 where +import Prelude (Integer, String, ($), (>>=), mempty) +import qualified Prelude +import GHC.Exts (ByteArray#) +import GHC.IO (IO(..)) +import Language.Haskell.TH.Syntax (lift, Bytes(..), runIO) +import Foreign.Ptr (nullPtr) +import Foreign.ForeignPtr (newForeignPtr_) +import Data.Array.Byte + +fromInteger :: Integer -> String +fromInteger _ = "yikes" + +x :: Bytes +x = $(do + let fromInteger = Prelude.fromInteger + b <- runIO $ newForeignPtr_ nullPtr + lift $ Bytes b 0 0 + ) + +y :: ByteArray +y = $(lift (mempty :: ByteArray)) ===================================== testsuite/tests/th/TH_Lift.stderr ===================================== @@ -182,7 +182,7 @@ TH_Lift.hs:(93,10)-(99,13): Splicing expression Bytes {bytesPtr = GHC.Internal.ForeignPtr.ForeignPtr "Hello"# GHC.Internal.ForeignPtr.FinalPtr, - bytesOffset = 0, bytesSize = 5} + bytesOffset = GHC.Types.W# 0##, bytesSize = GHC.Types.W# 5##} TH_Lift.hs:90:10-59: Splicing expression examineCode [|| 3 + 4 ||] `bindCode` liftTyped ======> ===================================== testsuite/tests/th/all.T ===================================== @@ -630,3 +630,4 @@ test('T25252', req_th, req_c], compile_and_run, ['-fPIC T25252_c.c']) +test('T25272', normal, compile, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/821ec44cd09dfad7474cf72277b538f520166d57 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/821ec44cd09dfad7474cf72277b538f520166d57 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 15:33:13 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 18 Sep 2024 11:33:13 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 14 commits: RTS: expose closure_sizeW_ (#25252) Message-ID: <66eaf2b934e87_25dfcb6b8a00332c@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - c5ce2b8e by Matthew Pickering at 2024-09-18T16:28:36+01:00 ci: Run abi-test on test-abi label - - - - - db862ba8 by Rodrigo Mesquita at 2024-09-18T16:28:37+01:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - e8881e42 by Rodrigo Mesquita at 2024-09-18T16:28:37+01:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 1eb56672 by Rodrigo Mesquita at 2024-09-18T16:29:05+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 0a2ae5fb by Rodrigo Mesquita at 2024-09-18T16:29:05+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - b66dd8fd by Rodrigo Mesquita at 2024-09-18T16:29:05+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - fe39a7a8 by Rodrigo Mesquita at 2024-09-18T16:29:05+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 8a9ad234 by Rodrigo Mesquita at 2024-09-18T16:29:06+01:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/RV64/CodeGen.hs - compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e88731dc98c5f9271f92956e1eb0d60bb86760a7...8a9ad23479585405f88b80eaca7be4b654980078 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e88731dc98c5f9271f92956e1eb0d60bb86760a7...8a9ad23479585405f88b80eaca7be4b654980078 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 16:04:16 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Wed, 18 Sep 2024 12:04:16 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 12 commits: The X86 SIMD patch. Message-ID: <66eafa0081129_25dfcb9e5048462b2@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: a3d7ae36 by sheaf at 2024-09-18T17:38:19+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 0c3f2c3c by sheaf at 2024-09-18T17:38:19+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 2223f050 by sheaf at 2024-09-18T17:38:20+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 62da41e3 by sheaf at 2024-09-18T17:38:20+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - ab81d5a0 by sheaf at 2024-09-18T18:04:03+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 84ccea6a by sheaf at 2024-09-18T18:04:03+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 4af6befe by sheaf at 2024-09-18T18:04:03+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - c55bcb25 by sheaf at 2024-09-18T18:04:04+02:00 Add test for C calls & SIMD vectors - - - - - fcb330ad by sheaf at 2024-09-18T18:04:04+02:00 Add test for #25169 - - - - - 635e00a6 by sheaf at 2024-09-18T18:04:04+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - d03a091f by sheaf at 2024-09-18T18:04:04+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f36849c9 by sheaf at 2024-09-18T18:04:04+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/acf568390a4296b44032a5e6264be016ac6aef5a...f36849c917a6bc0d4ed96b1703e60d1dc4b10447 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/acf568390a4296b44032a5e6264be016ac6aef5a...f36849c917a6bc0d4ed96b1703e60d1dc4b10447 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 16:13:54 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Wed, 18 Sep 2024 12:13:54 -0400 Subject: [Git][ghc/ghc][wip/ttg-booleanformula] moved as much as I can towards the correct places... Message-ID: <66eafc4244140_3ac274bedd49181e@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/ttg-booleanformula at Glasgow Haskell Compiler / GHC Commits: 2ad76802 by Hassan Al-Awwadi at 2024-09-18T18:13:19+02:00 moved as much as I can towards the correct places... - - - - - 3 changed files: - compiler/GHC/Data/BooleanFormula.hs - compiler/GHC/Hs/Decls.hs - compiler/Language/Haskell/Syntax/BooleanFormula.hs Changes: ===================================== compiler/GHC/Data/BooleanFormula.hs ===================================== @@ -1,10 +1,5 @@ -{-# LANGUAGE DeriveDataTypeable #-} - -{-# LANGUAGE TypeFamilies #-} -{-# LANGUAGE UndecidableInstances #-} -{-# OPTIONS_GHC -Wno-unrecognised-pragmas #-} -{-# LANGUAGE AllowAmbiguousTypes #-} {-# OPTIONS_GHC -Wno-orphans #-} +{-# LANGUAGE TypeFamilies #-} -------------------------------------------------------------------------------- -- | Boolean formulas without quantifiers and without negation. @@ -21,31 +16,39 @@ module GHC.Data.BooleanFormula ( pprBooleanFormula, pprBooleanFormulaNice ) where -import Language.Haskell.Syntax.BooleanFormula - -import GHC.Prelude hiding ( init, last ) - import Data.List ( intersperse ) import Data.List.NonEmpty ( NonEmpty (..), init, last ) -import Data.Data -import GHC.Utils.Outputable -import GHC.Parser.Annotation ( SrcSpanAnnL ) +import GHC.Prelude hiding ( init, last ) import GHC.Types.Unique import GHC.Types.Unique.Set -import Language.Haskell.Syntax.Extension (Anno) import GHC.Types.SrcLoc (GenLocated(L), unLoc) +import GHC.Utils.Outputable +import GHC.Parser.Annotation ( SrcSpanAnnL ) import GHC.Hs.Extension (GhcPass) +import Language.Haskell.Syntax.Extension (Anno) +import Language.Haskell.Syntax.BooleanFormula + + + ---------------------------------------------------------------------- -- Boolean formula type and smart constructors ---------------------------------------------------------------------- +-- Actually should this be moved to GHC.Hs.Decls? +-- That's where most of the anno instances are, anyway. +-- But I don't know how strictly that is just an implementation detail +-- I'm allowed to ignore? type instance Anno (BooleanFormula (GhcPass p) a) = SrcSpanAnnL -deriving instance (Eq a, Eq (LBooleanFormula p a)) => Eq (BooleanFormula p a) -deriving instance (Data a, Data p, Data (LBooleanFormula p a)) => Data (BooleanFormula p a) - +-- In a perfect world I could do something like +-- instance (Functor (LBooleanFormula p) => Functor (BooleanFormula p)). +-- But type synonyms need to be fully applied, and I could not figure out the +-- correct way to hack my way around wrapping and unwrapping and ~ to make it work +-- less cleanly. My last hope for using mapXRec also doe not work, because it has +-- the Anno p a ~ Anno p b constraint which seems impossible to get in the class +-- constraint header. instance Functor (BooleanFormula (GhcPass p)) where fmap :: (a -> b) -> BooleanFormula (GhcPass p) a -> BooleanFormula (GhcPass p) b fmap f (Var a ) = Var (f a) @@ -53,13 +56,7 @@ instance Functor (BooleanFormula (GhcPass p)) where fmap f (Or bfs) = Or $ fmap (fmap (fmap f)) bfs fmap f (Parens bf ) = Parens $ fmap (fmap f) bf -instance Foldable (BooleanFormula (GhcPass p)) where - foldMap :: Monoid m => (a -> m) -> BooleanFormula (GhcPass p) a -> m - foldMap f (Var a ) = f a - foldMap f (And bfs) = foldMap (foldMap f . unLoc) bfs - foldMap f (Or bfs) = foldMap (foldMap f . unLoc) bfs - foldMap f (Parens bf ) = foldMap f $ unLoc bf - +-- See comment above Functor instance. instance Traversable (BooleanFormula (GhcPass p)) where sequenceA :: Applicative f => BooleanFormula (GhcPass p) (f a) -> f (BooleanFormula (GhcPass p) a) sequenceA (Var a) = Var <$> a @@ -87,7 +84,6 @@ lbfSwitchPass :: forall p p' l a lbfSwitchPass (L loc bf) = L loc (bfSwitchPass bf) - {- Note [Simplification of BooleanFormulas] ~~~~~~~~~~~~~~~~~~~~~~ @@ -134,7 +130,7 @@ isTrue :: BooleanFormula (GhcPass p) a -> Bool isTrue (And []) = True isTrue _ = False -eval :: forall p a. (a -> Bool) -> BooleanFormula (GhcPass p) a -> Bool +eval :: (a -> Bool) -> BooleanFormula (GhcPass p) a -> Bool eval f (Var x) = f x eval f (And xs) = all (eval f . unLoc) xs eval f (Or xs) = any (eval f . unLoc) xs @@ -142,7 +138,7 @@ eval f (Parens x) = eval f (unLoc x) -- Simplify a boolean formula. -- The argument function should give the truth of the atoms, or Nothing if undecided. -simplify :: forall p a. Eq a => (a -> Maybe Bool) -> BooleanFormula (GhcPass p) a -> BooleanFormula (GhcPass p) a +simplify :: Eq a => (a -> Maybe Bool) -> BooleanFormula (GhcPass p) a -> BooleanFormula (GhcPass p) a simplify f (Var a) = case f a of Nothing -> Var a Just b -> mkBool b @@ -153,7 +149,7 @@ simplify f (Parens x) = simplify f (unLoc x) -- Test if a boolean formula is satisfied when the given values are assigned to the atoms -- if it is, returns Nothing -- if it is not, return (Just remainder) -isUnsatisfied :: forall p a. Eq a => (a -> Bool) -> BooleanFormula (GhcPass p) a -> Maybe (BooleanFormula (GhcPass p) a) +isUnsatisfied :: Eq a => (a -> Bool) -> BooleanFormula (GhcPass p) a -> Maybe (BooleanFormula (GhcPass p) a) isUnsatisfied f bf | isTrue bf' = Nothing | otherwise = Just bf' @@ -166,14 +162,14 @@ isUnsatisfied f bf -- eval f x == False <==> isFalse (simplify (Just . f) x) -- If the boolean formula holds, does that mean that the given atom is always true? -impliesAtom :: forall p a. Eq a => BooleanFormula (GhcPass p) a -> a -> Bool +impliesAtom :: Eq a => BooleanFormula (GhcPass p) a -> a -> Bool Var x `impliesAtom` y = x == y And xs `impliesAtom` y = any (\x -> unLoc x `impliesAtom` y) xs -- we have all of xs, so one of them implying y is enough Or xs `impliesAtom` y = all (\x -> unLoc x `impliesAtom` y) xs Parens x `impliesAtom` y = unLoc x `impliesAtom` y -implies :: forall p a. Uniquable a => BooleanFormula (GhcPass p) a -> BooleanFormula (GhcPass p) a -> Bool +implies :: Uniquable a => BooleanFormula (GhcPass p) a -> BooleanFormula (GhcPass p) a -> Bool implies e1 e2 = go (Clause emptyUniqSet [e1]) (Clause emptyUniqSet [e2]) where go :: Uniquable a => Clause (GhcPass p) a -> Clause (GhcPass p) a -> Bool @@ -210,8 +206,7 @@ memberClauseAtoms x c = x `elementOfUniqSet` clauseAtoms c -- Pretty print a BooleanFormula, -- using the arguments as pretty printers for Var, And and Or respectively -pprBooleanFormula' :: forall p a. () - => (Rational -> a -> SDoc) +pprBooleanFormula' :: (Rational -> a -> SDoc) -> (Rational -> [SDoc] -> SDoc) -> (Rational -> [SDoc] -> SDoc) -> Rational -> BooleanFormula (GhcPass p) a -> SDoc @@ -245,8 +240,7 @@ pprBooleanFormulaNice = pprBooleanFormula' pprVar pprAnd pprOr 0 instance (OutputableBndr a) => Outputable (BooleanFormula (GhcPass p) a) where ppr = pprBooleanFormulaNormal -pprBooleanFormulaNormal :: forall p a. (OutputableBndr a) - => BooleanFormula (GhcPass p) a -> SDoc +pprBooleanFormulaNormal :: (OutputableBndr a) => BooleanFormula (GhcPass p) a -> SDoc pprBooleanFormulaNormal = go where go (Var x) = pprPrefixOcc x ===================================== compiler/GHC/Hs/Decls.hs ===================================== @@ -103,6 +103,8 @@ module GHC.Hs.Decls ( import GHC.Prelude import Language.Haskell.Syntax.Decls +import Language.Haskell.Syntax.Extension +import Language.Haskell.Syntax.BooleanFormula (BooleanFormula) import {-# SOURCE #-} GHC.Hs.Expr ( pprExpr, pprUntypedSplice ) -- Because Expr imports Decls via HsBracket @@ -112,7 +114,7 @@ import GHC.Hs.Type import GHC.Hs.Doc import GHC.Types.Basic import GHC.Core.Coercion -import Language.Haskell.Syntax.Extension + import GHC.Hs.Extension import GHC.Parser.Annotation import GHC.Types.Name @@ -1395,3 +1397,4 @@ type instance Anno (Maybe Role) = EpAnnCO type instance Anno CCallConv = EpaLocation type instance Anno Safety = EpaLocation type instance Anno CExportSpec = EpaLocation +type instance Anno (BooleanFormula (GhcPass p) a) = SrcSpanAnnL ===================================== compiler/Language/Haskell/Syntax/BooleanFormula.hs ===================================== @@ -1,11 +1,14 @@ -{-# LANGUAGE RoleAnnotations #-} {-# LANGUAGE AllowAmbiguousTypes #-} +{-# LANGUAGE UndecidableInstances #-} +{-# LANGUAGE QuantifiedConstraints #-} + module Language.Haskell.Syntax.BooleanFormula( BooleanFormula(..), LBooleanFormula, mkVar, mkFalse, mkTrue, mkBool, mkAnd, mkOr) where import Prelude hiding ( init, last ) import Data.List ( nub ) +import Data.Data (Data) import Language.Haskell.Syntax.Extension (XRec, UnXRec (..)) @@ -14,16 +17,16 @@ type LBooleanFormula p a = XRec p (BooleanFormula p a) --type role BooleanFormula phantom nominal data BooleanFormula p a = Var a | And [LBooleanFormula p a] | Or [LBooleanFormula p a] | Parens (LBooleanFormula p a) - deriving () --, Functor, Foldable, Traversable) --- | Monadic version of concatMap -concatMapM :: (Monad m, Traversable f) => (a -> m [b]) -> f a -> m [b] -concatMapM f xs = fmap concat (mapM f xs) --- A dupe from GHC.Utils.Monad, until I figure out what to actually do with it. -{-# INLINE concatMapM #-} --- It's better to inline to inline this than to specialise --- concatMapM :: (Monad m) => (a -> m [b]) -> [a] -> m [b] --- Inlining cuts compiler allocation by around 1% +-- instances +deriving instance (Eq a, Eq (LBooleanFormula p a)) => Eq (BooleanFormula p a) +deriving instance (Data a, Data p, Data (LBooleanFormula p a)) => Data (BooleanFormula p a) +instance UnXRec p => Foldable (BooleanFormula p) where + foldMap :: Monoid m => (a -> m) -> BooleanFormula p a -> m + foldMap f (Var a ) = f a + foldMap f (And bfs) = foldMap (foldMap f . unXRec @p) bfs + foldMap f (Or bfs) = foldMap (foldMap f . unXRec @p) bfs + foldMap f (Parens bf ) = foldMap f $ unXRec @p bf mkVar :: a -> BooleanFormula p a @@ -40,7 +43,7 @@ mkBool True = mkTrue -- Make a conjunction, and try to simplify mkAnd :: forall p a. (UnXRec p, Eq a, Eq (LBooleanFormula p a)) => [LBooleanFormula p a] -> BooleanFormula p a -mkAnd = maybe mkFalse (mkAnd' . nub) . concatMapM fromAnd +mkAnd = maybe mkFalse (mkAnd' . nub . concat) . mapM fromAnd where -- See Note [Simplification of BooleanFormulas] fromAnd :: LBooleanFormula p a -> Maybe [LBooleanFormula p a] @@ -55,7 +58,7 @@ mkAnd = maybe mkFalse (mkAnd' . nub) . concatMapM fromAnd mkAnd' xs = And xs mkOr :: forall p a. (UnXRec p, Eq a, Eq (LBooleanFormula p a)) => [LBooleanFormula p a] -> BooleanFormula p a -mkOr = maybe mkTrue (mkOr' . nub) . concatMapM fromOr +mkOr = maybe mkTrue (mkOr' . nub . concat) . mapM fromOr where -- See Note [Simplification of BooleanFormulas] fromOr bf = case unXRec @p bf of View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2ad768025845f7a772047290c9c7a781bc5e6ab5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2ad768025845f7a772047290c9c7a781bc5e6ab5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 16:30:02 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Wed, 18 Sep 2024 12:30:02 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/persistent-worker-test-2024-09-10] Link bytecode from interface-stored core bindings in oneshot mode Message-ID: <66eb000a55f66_3ac27424d4ac920fd@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/persistent-worker-test-2024-09-10 at Glasgow Haskell Compiler / GHC Commits: 0c6cdd92 by Cheng Shao at 2024-09-18T16:15:35+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). Metric Decrease: MultiLayerModules T13701 - - - - - 20 changed files: - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/A.hs - + testsuite/tests/bytecode/T25090/B.hs - + testsuite/tests/bytecode/T25090/C.hs - + testsuite/tests/bytecode/T25090/C.hs-boot - + testsuite/tests/bytecode/T25090/D.hs - + testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/T25090-debug.stderr - + testsuite/tests/bytecode/T25090/T25090.stdout - + testsuite/tests/bytecode/T25090/all.T Changes: ===================================== compiler/GHC/CoreToIface.hs ===================================== @@ -602,7 +602,7 @@ toIfaceTopBind b = in (top_bndr, rhs') -- The sharing behaviour is currently disabled due to #22807, and relies on - -- finished #220056 to be re-enabled. + -- finished #20056 to be re-enabled. disabledDueTo22807 = True already_has_unfolding b = not disabledDueTo22807 @@ -774,8 +774,8 @@ outside of the hs-boot loop. Note [Interface File with Core: Sharing RHSs] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -IMPORTANT: This optimisation is currently disabled due to #22027, it can be - re-enabled once #220056 is implemented. +IMPORTANT: This optimisation is currently disabled due to #22807, it can be + re-enabled once #22056 is implemented. In order to avoid duplicating definitions for bindings which already have unfoldings we do some minor headstands to avoid serialising the RHS of a definition if it has ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -260,7 +260,6 @@ outputForeignStubs Maybe FilePath) -- C file created outputForeignStubs logger tmpfs dflags unit_state mod location stubs = do - let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location stub_c <- newTempName logger tmpfs (tmpDir dflags) TFL_CurrentModule "c" case stubs of @@ -276,8 +275,6 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs stub_h_output_d = pprCode h_code stub_h_output_w = showSDoc dflags stub_h_output_d - createDirectoryIfMissing True (takeDirectory stub_h) - putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export header file" FormatC @@ -299,9 +296,20 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs | platformMisc_libFFI $ platformMisc dflags = "#include \"rts/ghc_ffi.h\"\n" | otherwise = "" - stub_h_file_exists - <- outputForeignStubs_help stub_h stub_h_output_w - ("#include \n" ++ cplusplus_hdr) cplusplus_ftr + stub_h_file_exists <- + if null stub_h_output_w + then pure False + else do + -- The header path is computed from the module source path, which + -- does not exist when loading interface core bindings for Template + -- Haskell. + -- The header is only generated for foreign exports. + -- Since those aren't supported for TH with bytecode, we can skip + -- this here for now. + let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location + createDirectoryIfMissing True (takeDirectory stub_h) + outputForeignStubs_help stub_h stub_h_output_w + ("#include \n" ++ cplusplus_hdr) cplusplus_ftr putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export stubs" FormatC stub_c_output_d ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,6 +49,7 @@ module GHC.Driver.Main , HscBackendAction (..), HscRecompStatus (..) , initModDetails , initWholeCoreBindings + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -105,6 +106,7 @@ module GHC.Driver.Main , showModuleIndex , hscAddSptEntries , writeInterfaceOnlyMode + , loadByteCode ) where import GHC.Prelude @@ -274,7 +276,8 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.List ( nub, isPrefixOf, partition ) +import Data.Functor ((<&>)) +import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad import Data.IORef @@ -970,19 +973,23 @@ loadByteCode iface mod_sum = do (mi_foreign iface) return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) _ -> return $ outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- +add_iface_to_hpt :: ModIface -> ModDetails -> HscEnv -> HscEnv +add_iface_to_hpt iface details = + hscUpdateHPT $ \ hpt -> + addToHpt hpt (moduleName (mi_module iface)) + (HomeModInfo iface details emptyHomeModInfoLinkable) -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] initModDetails :: HscEnv -> ModIface -> IO ModDetails initModDetails hsc_env iface = fixIO $ \details' -> do - let act hpt = addToHpt hpt (moduleName $ mi_module iface) - (HomeModInfo iface details' emptyHomeModInfoLinkable) - let !hsc_env' = hscUpdateHPT act hsc_env + let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful @@ -1010,8 +1017,52 @@ compile_for_interpreter hsc_env use = adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Return an 'IO' that hydrates Core bindings and compiles them to bytecode if +-- the interface contains any, using the supplied type env for typechecking. +-- +-- Unlike 'initWholeCoreBindings', this does not use lazy IO. +-- Instead, the 'IO' is only evaluated (in @get_link_deps@) when it is clear +-- that it will be used immediately (because we're linking TH with +-- @-fprefer-byte-code@ in oneshot mode), and the result is cached in +-- 'LoaderState'. +-- +-- 'initWholeCoreBindings' needs the laziness because it is used to populate +-- 'HomeModInfo', which is done preemptively, in anticipation of downstream +-- modules using the bytecode for TH in make mode, which might never happen. +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + compile <$> iface_core_bindings iface location + where + compile decls = do + (bcos, fos) <- compileWholeCoreBindings hsc_env type_env decls + linkable $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time (mi_module iface) parts + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects. +-- them with a lazy IO thunk that compiles them to bytecode and foreign objects, +-- using the supplied environment for type checking. -- -- The laziness is necessary because this value is stored purely in a -- 'HomeModLinkable' in the home package table, rather than some dedicated @@ -1025,29 +1076,71 @@ compile_for_interpreter hsc_env use = -- -- This is sound because generateByteCode just depends on things already loaded -- in the interface file. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env mod_iface details (Linkable utc_time this_mod uls) = +initWholeCoreBindings :: + HscEnv -> + ModIface -> + ModDetails -> + Linkable -> + IO Linkable +initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = Linkable utc_time this_mod <$> mapM go uls where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef (md_types details) - let act hpt = addToHpt hpt (moduleName $ mi_module mod_iface) - (HomeModInfo mod_iface details emptyHomeModInfoLinkable) - kv = knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - hsc_env' = hscUpdateHPT act hsc_env { hsc_type_env_vars = kv } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") hsc_env' $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons (md_types details)) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + go = \case + CoreBindings wcb -> do + ~(bco, fos) <- unsafeInterleaveIO $ + compileWholeCoreBindings hsc_env' type_env wcb + pure (LazyBCOs bco fos) + l -> pure l + + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written or were unavailable due to boot import +-- cycles, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +compileWholeCoreBindings :: + HscEnv -> + TypeEnv -> + WholeCoreBindings -> + IO (CompiledByteCode, [FilePath]) +compileWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + gen_bytecode core_binds stubs foreign_files + where + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + generateByteCode hsc_env cgi_guts wcb_mod_location + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -0,0 +1,15 @@ +module GHC.Driver.Main where + +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) + +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1310,8 +1310,10 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- MP: This is a bit janky, because before you add the entries you have to extend the HPT with the module -- you just compiled. Another option, would be delay adding anything until after upsweep has finished, but I -- am unsure if this is sound (wrt running TH splices for example). - -- This function only does anything if the linkable produced is a BCO, which only happens with the - -- bytecode backend, no need to guard against the backend type additionally. + -- This function only does anything if the linkable produced is a BCO, which + -- used to only happen with the bytecode backend, but with + -- @-fprefer-byte-code@, @HomeModInfo@ has bytecode even when generating + -- object code, see #25230. addSptEntries (hscUpdateHPT (\hpt -> addToHpt hpt (ms_mod_name summary) hmi) hsc_env) (homeModInfoByteCode hmi) @@ -3012,7 +3014,7 @@ which can be checked easily using ghc-debug. a reference to the entire HscEnv, if we are not careful the HscEnv will contain the HomePackageTable at the time the interface was loaded and it will never be released. - Where? dontLeakTheHPT in GHC.Iface.Load + Where? dontLeakTheHUG in GHC.Iface.Load 2. No KnotVars are live at the end of upsweep (#20491) Why? KnotVars contains an old stale reference to the TypeEnv for modules ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -505,12 +506,14 @@ loadInterface doc_str mod from || mod == gHC_PRIM) (text "Attempting to load home package interface into the EPS" $$ ppr hug $$ doc_str $$ ppr mod $$ ppr (moduleUnitId mod)) ; ignore_prags <- goptM Opt_IgnoreInterfacePragmas + ; prefer_bytecode <- goptM Opt_UseBytecodeRatherThanObjects ; new_eps_decls <- tcIfaceDecls ignore_prags (mi_decls iface) ; new_eps_insts <- mapM tcIfaceInst (mi_insts iface) ; new_eps_fam_insts <- mapM tcIfaceFamInst (mi_fam_insts iface) ; new_eps_rules <- tcIfaceRules ignore_prags (mi_rules iface) ; new_eps_anns <- tcIfaceAnnotations (mi_anns iface) ; new_eps_complete_matches <- tcIfaceCompleteMatches (mi_complete_matches iface) + ; purged_hsc_env <- getTopEnv ; let final_iface = iface & set_mi_decls (panic "No mi_decls in PIT") @@ -518,13 +521,27 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + | prefer_bytecode + , Just action <- loadIfaceByteCode purged_hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have 'extra_decls' + -- so 'get_link_deps' knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,6 +553,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -569,7 +587,7 @@ loadInterface doc_str mod from {- Note [Loading your own hi-boot file] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Generally speaking, when compiling module M, we should not -load M.hi boot into the EPS. After all, we are very shortly +load M.hi-boot into the EPS. After all, we are very shortly going to have full information about M. Moreover, see Note [Do not update EPS with your own hi-boot] in GHC.Iface.Recomp. @@ -698,7 +716,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -845,7 +863,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -875,7 +893,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -900,7 +918,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -905,11 +905,11 @@ tcTopIfaceBindings :: IORef TypeEnv -> [IfaceBindingX IfaceMaybeRhs IfaceTopBndr -> IfL [CoreBind] tcTopIfaceBindings ty_var ver_decls = do - int <- mapM tcTopBinders ver_decls + int <- mapM tcTopBinders ver_decls let all_ids :: [Id] = concatMap toList int liftIO $ modifyIORef ty_var (flip extendTypeEnvList (map AnId all_ids)) - extendIfaceIdEnv all_ids $ mapM (tc_iface_bindings) int + extendIfaceIdEnv all_ids $ mapM tc_iface_bindings int tcTopBinders :: IfaceBindingX a IfaceTopBndrInfo -> IfL (IfaceBindingX a Id) tcTopBinders = traverse mk_top_id ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -60,16 +60,16 @@ import System.Directory data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -275,21 +275,21 @@ get_link_deps opts pls maybe_normal_osuf span mods = do case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod = do { - -- ...and then find the linkable for it - mb_lnk <- findObjectLinkableMaybe mod loc ; - case mb_lnk of { - Nothing -> no_obj mod ; - Just lnk -> adjust_linkable lnk - }} + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,18 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -68,6 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -139,6 +142,12 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), + eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules eps_fam_inst_env :: !PackageFamInstEnv,-- ^ The total 'FamInstEnv' accumulated ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -28,12 +28,12 @@ import System.FilePath (takeExtension) {- Note [Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A interface file can optionally contain the definitions of all core bindings, this is enabled by the flag `-fwrite-if-simplified-core`. This provides everything needed in addition to the normal ModIface and ModDetails -to restart compilation after typechecking to generate bytecode. The `fi_bindings` field +to restart compilation after typechecking to generate bytecode. The `wcb_bindings` field is stored in the normal interface file and the other fields populated whilst loading the interface file. @@ -62,8 +62,50 @@ after whatever simplification the user requested has been performed. So the simp of the interface file agree with the optimisation level as reported by the interface file. +The lifecycle differs beyond laziness depending on the provenance of a module. +In all cases, the main consumer for interface bytecode is 'get_link_deps', which +traverses a splice's or GHCi expression's dependencies and collects the needed +build artifacts, which can be objects or bytecode, depending on the build +settings. + +1. In make mode, all eligible modules are part of the dependency graph. + Their interfaces are loaded unconditionally and in dependency order by the + compilation manager, and each module's bytecode is prepared before its + dependents are compiled, in one of two ways: + + - If the interface file for a module is missing or out of sync with its + source, it is recompiled and bytecode is generated directly and + immediately, not involving 'WholeCoreBindings' (in 'runHscBackendPhase'). + + - If the interface file is up to date, no compilation is performed, and a + lazy thunk generating bytecode from interface Core bindings is created in + 'compileOne'', which will only be compiled if a downstream module contains + a splice that depends on it, as described above. + + In both cases, the bytecode 'Linkable' is stored in a 'HomeModLinkable' in + the Home Unit Graph, lazy or not. + +2. In oneshot mode, which compiles individual modules without a shared home unit + graph, a previously compiled module is not reprocessed as described for make + mode above. + When 'get_link_deps' encounters a dependency on a local module, it requests + its bytecode from the External Package State, who loads the interface + on-demand. + + Since the EPS stores interfaces for all package dependencies in addition to + local modules in oneshot mode, it has a substantial memory footprint. + We try to curtail that by extracting important data into specialized fields + in the EPS, and retaining only a few fields of 'ModIface' by overwriting the + others with bottom values. + + In order to avoid keeping around all of the interface's components needed for + compiling bytecode, we instead store an IO action in 'eps_iface_bytecode'. + When 'get_link_deps' evaluates this action, the result is not retained in the + EPS, but stored in 'LoaderState', where it may eventually get evicted to free + up the memory. + Note [Size of Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ How much overhead does `-fwrite-if-simplified-core` add to a typical interface file? As an experiment I compiled the `Cabal` library and `ghc` library (Aug 22) with ===================================== testsuite/tests/bytecode/T25090/A.hs ===================================== @@ -0,0 +1,7 @@ +{-# language TemplateHaskell #-} +module Main where + +import D + +main :: IO () +main = putStrLn (show ($splc :: Int)) ===================================== testsuite/tests/bytecode/T25090/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} C (C) + +data B = B C ===================================== testsuite/tests/bytecode/T25090/C.hs ===================================== @@ -0,0 +1,8 @@ +module C where + +import B + +data C = C Int + +b :: B +b = B (C 2024) ===================================== testsuite/tests/bytecode/T25090/C.hs-boot ===================================== @@ -0,0 +1,3 @@ +module C where + +data C ===================================== testsuite/tests/bytecode/T25090/D.hs ===================================== @@ -0,0 +1,12 @@ +module D where + +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +import B +import C + +splc :: ExpQ +splc = + lift @_ @Int num + where + B (C num) = b ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -0,0 +1,21 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +# Verify that the object files aren't linked by clobbering them. +T25090a: + $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot + $(TEST_HC) -c -fbyte-code-and-object-code B.hs + $(TEST_HC) -c -fbyte-code-and-object-code C.hs + echo 'corrupt' > B.o + echo 'corrupt' > C.o + echo 'corrupt' > C.o-boot + $(TEST_HC) -c -fbyte-code-and-object-code D.hs + echo 'corrupt' > D.o + $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A.o -o exe + ./exe + +T25090b: + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 + ./exe ===================================== testsuite/tests/bytecode/T25090/T25090-debug.stderr ===================================== @@ -0,0 +1,6 @@ +WARNING: + loadInterface + C + Call stack: + CallStack (from HasCallStack): + warnPprTrace, called at compiler/GHC/Iface/Load.hs:: in :GHC.Iface.Load ===================================== testsuite/tests/bytecode/T25090/T25090.stdout ===================================== @@ -0,0 +1 @@ +2024 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -0,0 +1,18 @@ +# This test compiles the boot file separately from its source file, which causes +# a debug assertion warning. +# Since this appears to be intentional according to the Note [Loading your own hi-boot file], +# the warning is added to the expected stderr for debugged builds. +def test_T25090(name): + assert_warn_spec = {'stderr': 'T25090-debug.stderr'} + extra_specs = assert_warn_spec if name == 'T25090a' and compiler_debugged() else {} + return test(name, + [extra_files(['A.hs', 'B.hs', 'C.hs-boot', 'C.hs', 'D.hs']), + req_th, + js_skip, + use_specs(dict(stdout = 'T25090.stdout', **extra_specs)), + ], + makefile_test, + []) + +test_T25090('T25090a') +test_T25090('T25090b') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0c6cdd92173c9404a4e181d3d9d322a5e31e171a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0c6cdd92173c9404a4e181d3d9d322a5e31e171a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 16:38:06 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Wed, 18 Sep 2024 12:38:06 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/persistent-worker-test-2024-09-10] be less eager to skip creating stub dirs Message-ID: <66eb01eecd52b_3ac274327580922c6@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/persistent-worker-test-2024-09-10 at Glasgow Haskell Compiler / GHC Commits: de4e6137 by Torsten Schmits at 2024-09-18T18:37:57+02:00 be less eager to skip creating stub dirs - - - - - 1 changed file: - compiler/GHC/Driver/CodeOutput.hs Changes: ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -296,10 +296,9 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs | platformMisc_libFFI $ platformMisc dflags = "#include \"rts/ghc_ffi.h\"\n" | otherwise = "" - stub_h_file_exists <- - if null stub_h_output_w - then pure False - else do + stub_h_file_exists <- case ml_hs_file_ospath location of + Nothing -> pure False + Just _ -> do -- The header path is computed from the module source path, which -- does not exist when loading interface core bindings for Template -- Haskell. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/de4e6137133d3fd12d88eb7bf527e18d43ba3c45 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/de4e6137133d3fd12d88eb7bf527e18d43ba3c45 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 17:10:22 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 18 Sep 2024 13:10:22 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 5 commits: determinism: Cmm unique renaming pass Message-ID: <66eb097e7e66a_3ac2744582749589@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: f0641005 by Rodrigo Mesquita at 2024-09-18T18:09:53+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - b88486d4 by Rodrigo Mesquita at 2024-09-18T18:09:54+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. - - - - - db8e1fd6 by Rodrigo Mesquita at 2024-09-18T18:09:54+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 3e2939a0 by Rodrigo Mesquita at 2024-09-18T18:09:54+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 4b1fc47d by Rodrigo Mesquita at 2024-09-18T18:09:54+01:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 30 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm/Reg/Liveness.hs - compiler/GHC/Driver/Config/StgToCmm.hs - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/GenerateCgIPEStub.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Driver/Session.hs - compiler/GHC/Stg/Debug.hs - compiler/GHC/StgToCmm.hs - compiler/GHC/StgToCmm/CgUtils.hs - compiler/GHC/StgToCmm/Config.hs - compiler/GHC/StgToCmm/InfoTableProv.hs - compiler/GHC/StgToCmm/Monad.hs - compiler/GHC/StgToCmm/Prof.hs - compiler/GHC/StgToCmm/Utils.hs - compiler/GHC/Types/IPE.hs - compiler/GHC/Types/Name.hs - compiler/GHC/Types/Unique.hs - compiler/GHC/Types/Unique/DFM.hs - compiler/GHC/Utils/Outputable.hs - compiler/ghc.cabal.in The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8a9ad23479585405f88b80eaca7be4b654980078...4b1fc47dfc8ec3abbbba36bc6dfba2ff1824cf69 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8a9ad23479585405f88b80eaca7be4b654980078...4b1fc47dfc8ec3abbbba36bc6dfba2ff1824cf69 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 18:16:48 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 18 Sep 2024 14:16:48 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 4 commits: determinism: DCmmGroup vs CmmGroup Message-ID: <66eb19106a482_3ac27481345412465b@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 0a5c2dac by Rodrigo Mesquita at 2024-09-18T19:16:08+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 475d1f2e by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 5af98c82 by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - ed43ad42 by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 23 changed files: - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ThreadSanitizer.hs - compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm/Reg/Liveness.hs - compiler/GHC/Driver/GenerateCgIPEStub.hs - compiler/GHC/Driver/Main.hs - compiler/GHC/Stg/Debug.hs - compiler/GHC/StgToCmm.hs - compiler/GHC/StgToCmm/CgUtils.hs - compiler/GHC/StgToCmm/InfoTableProv.hs - compiler/GHC/StgToCmm/Monad.hs - compiler/GHC/StgToCmm/Prof.hs - compiler/GHC/StgToCmm/Utils.hs - compiler/GHC/Types/IPE.hs - compiler/GHC/Types/Name.hs - compiler/GHC/Types/Unique/DFM.hs - compiler/GHC/Utils/Outputable.hs - testsuite/tests/regalloc/regalloc_unit_tests.hs - testsuite/tests/wasm/should_run/control-flow/LoadCmmGroup.hs Changes: ===================================== compiler/GHC/Cmm.hs ===================================== @@ -12,22 +12,28 @@ module GHC.Cmm ( -- * Cmm top-level datatypes + DCmmGroup, CmmProgram, CmmGroup, CmmGroupSRTs, RawCmmGroup, GenCmmGroup, - CmmDecl, CmmDeclSRTs, GenCmmDecl(..), - CmmDataDecl, cmmDataDeclCmmDecl, - CmmGraph, GenCmmGraph(..), + CmmDecl, DCmmDecl, CmmDeclSRTs, GenCmmDecl(..), + CmmDataDecl, cmmDataDeclCmmDecl, DCmmGraph, + CmmGraph, GenCmmGraph, GenGenCmmGraph(..), toBlockMap, revPostorder, toBlockList, CmmBlock, RawCmmDecl, Section(..), SectionType(..), GenCmmStatics(..), type CmmStatics, type RawCmmStatics, CmmStatic(..), SectionProtection(..), sectionProtection, + DWrap(..), unDeterm, removeDeterm, removeDetermDecl, removeDetermGraph, + -- ** Blocks containing lists GenBasicBlock(..), blockId, ListGraph(..), pprBBlock, -- * Info Tables - CmmTopInfo(..), CmmStackInfo(..), CmmInfoTable(..), topInfoTable, + GenCmmTopInfo(..) + , DCmmTopInfo + , CmmTopInfo + , CmmStackInfo(..), CmmInfoTable(..), topInfoTable, topInfoTableD, ClosureTypeInfo(..), ProfilingInfo(..), ConstrDescription, @@ -74,6 +80,8 @@ import qualified Data.ByteString as BS type CmmProgram = [CmmGroup] type GenCmmGroup d h g = [GenCmmDecl d h g] +-- | Cmm group after STG generation +type DCmmGroup = GenCmmGroup CmmStatics DCmmTopInfo DCmmGraph -- | Cmm group before SRT generation type CmmGroup = GenCmmGroup CmmStatics CmmTopInfo CmmGraph -- | Cmm group with SRTs @@ -117,6 +125,7 @@ instance (OutputableP Platform d, OutputableP Platform info, OutputableP Platfor => OutputableP Platform (GenCmmDecl d info i) where pdoc = pprTop +type DCmmDecl = GenCmmDecl CmmStatics DCmmTopInfo DCmmGraph type CmmDecl = GenCmmDecl CmmStatics CmmTopInfo CmmGraph type CmmDeclSRTs = GenCmmDecl RawCmmStatics CmmTopInfo CmmGraph type CmmDataDecl = GenCmmDataDecl CmmStatics @@ -139,7 +148,11 @@ type RawCmmDecl ----------------------------------------------------------------------------- type CmmGraph = GenCmmGraph CmmNode -data GenCmmGraph n = CmmGraph { g_entry :: BlockId, g_graph :: Graph n C C } +type DCmmGraph = GenGenCmmGraph DWrap CmmNode + +type GenCmmGraph n = GenGenCmmGraph LabelMap n + +data GenGenCmmGraph s n = CmmGraph { g_entry :: BlockId, g_graph :: Graph' s Block n C C } type CmmBlock = Block CmmNode C C instance OutputableP Platform CmmGraph where @@ -171,8 +184,16 @@ toBlockList g = mapElems $ toBlockMap g -- | CmmTopInfo is attached to each CmmDecl (see defn of CmmGroup), and contains -- the extra info (beyond the executable code) that belongs to that CmmDecl. -data CmmTopInfo = TopInfo { info_tbls :: LabelMap CmmInfoTable - , stack_info :: CmmStackInfo } +data GenCmmTopInfo f = TopInfo { info_tbls :: f CmmInfoTable + , stack_info :: CmmStackInfo } + +newtype DWrap a = DWrap [(BlockId, a)] + +unDeterm :: DWrap a -> [(BlockId, a)] +unDeterm (DWrap f) = f + +type DCmmTopInfo = GenCmmTopInfo DWrap +type CmmTopInfo = GenCmmTopInfo LabelMap instance OutputableP Platform CmmTopInfo where pdoc = pprTopInfo @@ -182,7 +203,12 @@ pprTopInfo platform (TopInfo {info_tbls=info_tbl, stack_info=stack_info}) = vcat [text "info_tbls: " <> pdoc platform info_tbl, text "stack_info: " <> ppr stack_info] -topInfoTable :: GenCmmDecl a CmmTopInfo (GenCmmGraph n) -> Maybe CmmInfoTable +topInfoTableD :: GenCmmDecl a DCmmTopInfo (GenGenCmmGraph s n) -> Maybe CmmInfoTable +topInfoTableD (CmmProc infos _ _ g) = case (info_tbls infos) of + DWrap xs -> lookup (g_entry g) xs +topInfoTableD _ = Nothing + +topInfoTable :: GenCmmDecl a CmmTopInfo (GenGenCmmGraph s n) -> Maybe CmmInfoTable topInfoTable (CmmProc infos _ _ g) = mapLookup (g_entry g) (info_tbls infos) topInfoTable _ = Nothing @@ -237,6 +263,7 @@ data ProfilingInfo = NoProfilingInfo | ProfilingInfo ByteString ByteString -- closure_type, closure_desc deriving (Eq, Ord) + ----------------------------------------------------------------------------- -- Static Data ----------------------------------------------------------------------------- @@ -328,6 +355,61 @@ instance OutputableP Platform (GenCmmStatics a) where type CmmStatics = GenCmmStatics 'False type RawCmmStatics = GenCmmStatics 'True +{- +----------------------------------------------------------------------------- +-- Deterministic Cmm / Info Tables +----------------------------------------------------------------------------- + +Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Consulting Note [Object determinism] one will learn that in order to produce +deterministic objects just after cmm is produced we perform a renaming pass which +provides fresh uniques for all unique-able things in the input Cmm. + +After this point, we use a deterministic unique supply (an incrementing counter) +so any resulting labels which make their way into object code have a deterministic name. + +A key assumption to this process is that the input is deterministic modulo the uniques +and the order that bindings appear in the definitions is the same. + +CmmGroup uses LabelMap in two places: + +* In CmmProc for info tables +* In CmmGraph for the blocks of the graph + +LabelMap is not a deterministic structure, so traversing a LabelMap can process +elements in different order (depending on the given uniques). + +Therefore before we do the renaming we need to use a deterministic structure, one +which we can traverse in a guaranteed order. A list does the job perfectly. + +Once the renaming happens it is converted back into a LabelMap, which is now deterministic +due to the uniques being generated and assigned in a deterministic manner. + +We prefer using the renamed LabelMap rather than the list in the rest of the +code generation because it is much more efficient than lists for the needs of +the code generator. +-} + +-- Converting out of deterministic Cmm + +removeDeterm :: DCmmGroup -> CmmGroup +removeDeterm = map removeDetermDecl + +removeDetermDecl :: DCmmDecl -> CmmDecl +removeDetermDecl (CmmProc h e r g) = CmmProc (removeDetermTop h) e r (removeDetermGraph g) +removeDetermDecl (CmmData a b) = CmmData a b + +removeDetermTop :: DCmmTopInfo -> CmmTopInfo +removeDetermTop (TopInfo a b) = TopInfo (mapFromList $ unDeterm a) b + +removeDetermGraph :: DCmmGraph -> CmmGraph +removeDetermGraph (CmmGraph x y) = + let y' = case y of + GMany a (DWrap b) c -> GMany a (mapFromList b) c + in CmmGraph x y' + -- ----------------------------------------------------------------------------- -- Basic blocks consisting of lists ===================================== compiler/GHC/Cmm/Dataflow/Graph.hs ===================================== @@ -26,10 +26,10 @@ import GHC.Cmm.Dataflow.Block import Data.Kind -- | A (possibly empty) collection of closed/closed blocks -type Body n = LabelMap (Block n C C) +type Body s n = Body' s Block n -- | @Body@ abstracted over @block@ -type Body' block (n :: Extensibility -> Extensibility -> Type) = LabelMap (block n C C) +type Body' s block (n :: Extensibility -> Extensibility -> Type) = s (block n C C) ------------------------------- -- | Gives access to the anchor points for @@ -46,13 +46,13 @@ instance NonLocal n => NonLocal (Block n) where successors (BlockCC _ _ n) = successors n -emptyBody :: Body' block n +emptyBody :: Body' LabelMap block n emptyBody = mapEmpty -bodyList :: Body' block n -> [(Label,block n C C)] +bodyList :: Body' LabelMap block n -> [(Label,block n C C)] bodyList body = mapToList body -bodyToBlockList :: Body n -> [Block n C C] +bodyToBlockList :: Body LabelMap n -> [Block n C C] bodyToBlockList body = mapElems body addBlock @@ -72,18 +72,18 @@ addBlock block body = mapAlter add lbl body -- O/C, C/O, C/C). A graph open at the entry has a single, -- distinguished, anonymous entry point; if a graph is closed at the -- entry, its entry point(s) are supplied by a context. -type Graph = Graph' Block +type Graph = Graph' LabelMap Block -- | @Graph'@ is abstracted over the block type, so that we can build -- graphs of annotated blocks for example (Compiler.Hoopl.Dataflow -- needs this). -data Graph' block (n :: Extensibility -> Extensibility -> Type) e x where - GNil :: Graph' block n O O - GUnit :: block n O O -> Graph' block n O O +data Graph' s block (n :: Extensibility -> Extensibility -> Type) e x where + GNil :: Graph' s block n O O + GUnit :: block n O O -> Graph' s block n O O GMany :: MaybeO e (block n O C) - -> Body' block n + -> Body' s block n -> MaybeO x (block n C O) - -> Graph' block n e x + -> Graph' s block n e x -- ----------------------------------------------------------------------------- @@ -91,26 +91,27 @@ data Graph' block (n :: Extensibility -> Extensibility -> Type) e x where -- | Maps over all nodes in a graph. mapGraph :: (forall e x. n e x -> n' e x) -> Graph n e x -> Graph n' e x -mapGraph f = mapGraphBlocks (mapBlock f) +mapGraph f = mapGraphBlocks mapMap (mapBlock f) -- | Function 'mapGraphBlocks' enables a change of representation of blocks, -- nodes, or both. It lifts a polymorphic block transform into a polymorphic -- graph transform. When the block representation stabilizes, a similar -- function should be provided for blocks. -mapGraphBlocks :: forall block n block' n' e x . - (forall e x . block n e x -> block' n' e x) - -> (Graph' block n e x -> Graph' block' n' e x) +mapGraphBlocks :: forall s block n block' n' e x . + (forall a b . (a -> b) -> s a -> s b) + -> (forall e x . block n e x -> block' n' e x) + -> (Graph' s block n e x -> Graph' s block' n' e x) -mapGraphBlocks f = map - where map :: Graph' block n e x -> Graph' block' n' e x +mapGraphBlocks f g = map + where map :: Graph' s block n e x -> Graph' s block' n' e x map GNil = GNil - map (GUnit b) = GUnit (f b) - map (GMany e b x) = GMany (fmap f e) (mapMap f b) (fmap f x) + map (GUnit b) = GUnit (g b) + map (GMany e b x) = GMany (fmap g e) (f g b) (fmap g x) -- ----------------------------------------------------------------------------- -- Extracting Labels from graphs -labelsDefined :: forall block n e x . NonLocal (block n) => Graph' block n e x +labelsDefined :: forall block n e x . NonLocal (block n) => Graph' LabelMap block n e x -> LabelSet labelsDefined GNil = setEmpty labelsDefined (GUnit{}) = setEmpty ===================================== compiler/GHC/Cmm/Graph.hs ===================================== @@ -73,12 +73,12 @@ data CgStmt | CgLast (CmmNode O C) | CgFork BlockId CmmAGraph CmmTickScope -flattenCmmAGraph :: BlockId -> CmmAGraphScoped -> CmmGraph +flattenCmmAGraph :: BlockId -> CmmAGraphScoped -> DCmmGraph flattenCmmAGraph id (stmts_t, tscope) = CmmGraph { g_entry = id, g_graph = GMany NothingO body NothingO } where - body = foldr addBlock emptyBody $ flatten id stmts_t tscope [] + body = DWrap [(entryLabel b, b) | b <- flatten id stmts_t tscope [] ] -- -- flatten: given an entry label and a CmmAGraph, make a list of blocks. @@ -169,13 +169,13 @@ outOfLine :: BlockId -> CmmAGraphScoped -> CmmAGraph outOfLine l (c,s) = unitOL (CgFork l c s) -- | allocate a fresh label for the entry point -lgraphOfAGraph :: CmmAGraphScoped -> UniqDSM CmmGraph +lgraphOfAGraph :: CmmAGraphScoped -> UniqDSM DCmmGraph lgraphOfAGraph g = do u <- getUniqueDSM return (labelAGraph (mkBlockId u) g) -- | use the given BlockId as the label of the entry point -labelAGraph :: BlockId -> CmmAGraphScoped -> CmmGraph +labelAGraph :: BlockId -> CmmAGraphScoped -> DCmmGraph labelAGraph lbl ag = flattenCmmAGraph lbl ag ---------- No-ops ===================================== compiler/GHC/Cmm/LayoutStack.hs ===================================== @@ -282,12 +282,18 @@ layout cfg procpoints liveness entry entry_args final_stackmaps final_sp_high bl where (updfr, cont_info) = collectContInfo blocks - init_stackmap = mapSingleton entry StackMap{ sm_sp = entry_args - , sm_args = entry_args - , sm_ret_off = updfr - , sm_regs = emptyUFM - } - + init_stackmap = mapSingleton entry + StackMap{ sm_sp = entry_args + , sm_args = entry_args + , sm_ret_off = updfr + , sm_regs = emptyUFM + } + + go :: [Block CmmNode C C] + -> LabelMap StackMap + -> StackLoc + -> [CmmBlock] + -> UniqDSM (LabelMap StackMap, StackLoc, [CmmBlock]) go [] acc_stackmaps acc_hwm acc_blocks = return (acc_stackmaps, acc_hwm, acc_blocks) @@ -1180,7 +1186,7 @@ lowerSafeForeignCall profile block copyout <*> mkLast jump, tscp) - case toBlockList graph' of + case toBlockList (removeDetermGraph graph') of [one] -> let (_, middle', last) = blockSplit one in return (blockJoin entry (middle `blockAppend` middle') last) _ -> panic "lowerSafeForeignCall0" ===================================== compiler/GHC/Cmm/Parser.y ===================================== @@ -278,6 +278,7 @@ import GHC.Parser.Lexer import GHC.Parser.Errors.Types import GHC.Parser.Errors.Ppr +import GHC.Types.Unique.DSM import GHC.Types.CostCentre import GHC.Types.ForeignCall import GHC.Unit.Module @@ -1575,7 +1576,7 @@ parseCmmFile :: CmmParserConfig -> Module -> HomeUnit -> FilePath - -> IO (Messages PsMessage, Messages PsMessage, Maybe (CmmGroup, [InfoProvEnt])) + -> IO (Messages PsMessage, Messages PsMessage, Maybe (DCmmGroup, [InfoProvEnt])) parseCmmFile cmmpConfig this_mod home_unit filename = do buf <- hGetStringBuffer filename let @@ -1595,11 +1596,17 @@ parseCmmFile cmmpConfig this_mod home_unit filename = do ((), cmm) <- getCmm $ unEC code "global" (initEnv (pdProfile pdConfig)) [] >> return () -- See Note [Mapping Info Tables to Source Positions] (IPE Maps) let used_info - | do_ipe = map (cmmInfoTableToInfoProvEnt this_mod) (mapMaybe topInfoTable cmm) + | do_ipe = map (cmmInfoTableToInfoProvEnt this_mod) (mapMaybe topInfoTableD cmm) | otherwise = [] where do_ipe = stgToCmmInfoTableMap $ cmmpStgToCmmConfig cmmpConfig - ((), cmm2) <- getCmm $ emitIpeBufferListNode this_mod used_info + -- We need to pass a deterministic unique supply to generate IPE + -- symbols deterministically. The symbols created by + -- emitIpeBufferListNode must all be local to the object (see + -- comment on its definition). If the symbols weren't local, using a + -- counter starting from zero for every Cmm file would cause + -- conflicts when compiling more than one Cmm file together. + (_, cmm2) <- getCmm $ emitIpeBufferListNode this_mod used_info (initDUniqSupply 'P' 0) return (cmm ++ cmm2, used_info) (cmm, _) = runC (cmmpStgToCmmConfig cmmpConfig) fstate st fcode (warnings,errors) = getPsMessages pst ===================================== compiler/GHC/Cmm/ThreadSanitizer.hs ===================================== @@ -19,6 +19,7 @@ import GHC.Types.Basic import GHC.Types.ForeignCall import GHC.Types.Unique import GHC.Types.Unique.Supply +import GHC.Cmm.Dataflow.Label import Data.Maybe (fromMaybe) @@ -29,7 +30,7 @@ data Env = Env { platform :: Platform annotateTSAN :: Platform -> CmmGraph -> UniqSM CmmGraph annotateTSAN platform graph = do env <- Env platform <$> getUniqueSupplyM - return $ modifyGraph (mapGraphBlocks (annotateBlock env)) graph + return $ modifyGraph (mapGraphBlocks mapMap (annotateBlock env)) graph mapBlockList :: (forall e' x'. n e' x' -> Block n e' x') -> Block n e x -> Block n e x ===================================== compiler/GHC/Cmm/UniqueRenamer.hs ===================================== @@ -1,6 +1,7 @@ {-# LANGUAGE LambdaCase, RecordWildCards, MagicHash, UnboxedTuples, PatternSynonyms, ExplicitNamespaces #-} module GHC.Cmm.UniqueRenamer ( detRenameCmmGroup + , detRenameIPEMap , MonadGetUnique(..) -- Careful! Not for general use! @@ -10,7 +11,7 @@ module GHC.Cmm.UniqueRenamer ) where -import Prelude +import GHC.Prelude import GHC.Utils.Monad.State.Strict import Data.Tuple (swap) import GHC.Word @@ -22,12 +23,13 @@ import GHC.Cmm.Dataflow.Label import GHC.Cmm.Switch import GHC.Types.Unique import GHC.Types.Unique.FM +import GHC.Types.Unique.DFM import GHC.Utils.Outputable as Outputable import GHC.Types.Id import GHC.Types.Unique.DSM import GHC.Types.Name hiding (varName) import GHC.Types.Var - +import GHC.Types.IPE {- Note [Renaming uniques deterministically] @@ -104,6 +106,12 @@ detRenameId i | isExternalName (varName i) = return i | otherwise = setIdUnique i <$> renameDetUniq (getUnique i) +-- | Similar to `detRenameId`, but for `Name`. +detRenameName :: Name -> DetRnM Name +detRenameName n + | isExternalName n = return n + | otherwise = setNameUnique n <$> renameDetUniq (getUnique n) + detRenameCmmGroup :: DetUniqFM -> DCmmGroup -> (DetUniqFM, CmmGroup) detRenameCmmGroup dufm group = swap (runState (mapM detRenameCmmDecl group) dufm) where @@ -251,3 +259,19 @@ detRenameCmmGroup dufm group = swap (runState (mapM detRenameCmmDecl group) dufm detRenameMaybe f (Just x) = Just <$> f x detRenamePair f g (a, b) = (,) <$> f a <*> g b + +detRenameIPEMap :: DetUniqFM -> InfoTableProvMap -> (DetUniqFM, InfoTableProvMap) +detRenameIPEMap dufm InfoTableProvMap{ provDC, provClosure, provInfoTables } = + (dufm2, InfoTableProvMap { provDC, provClosure = cm, provInfoTables }) + where + (cm, dufm2) = runState (detRenameClosureMap provClosure) dufm + + detRenameClosureMap :: ClosureMap -> DetRnM ClosureMap + detRenameClosureMap m = + -- `eltsUDFM` preserves the deterministic order, but it doesn't matter + -- since we will rename all uniques deterministically, thus the + -- reconstructed map will necessarily be deterministic too. + listToUDFM <$> mapM (\(n,r) -> do + n' <- detRenameName n + return (n', (n', r)) + ) (eltsUDFM m) ===================================== compiler/GHC/CmmToAsm/Reg/Liveness.hs ===================================== @@ -14,7 +14,7 @@ module GHC.CmmToAsm.Reg.Liveness ( RegSet, RegMap, emptyRegMap, - BlockMap, mapEmpty, + BlockMap, LiveCmmDecl, InstrSR (..), LiveInstr (..), @@ -260,7 +260,7 @@ instance OutputableP Platform LiveInfo where = (pdoc env mb_static) $$ text "# entryIds = " <> ppr entryIds $$ text "# liveVRegsOnEntry = " <> ppr liveVRegsOnEntry - $$ text "# liveSlotsOnEntry = " <> text (show liveSlotsOnEntry) + $$ text "# liveSlotsOnEntry = " <> ppr liveSlotsOnEntry ===================================== compiler/GHC/Driver/GenerateCgIPEStub.hs ===================================== @@ -20,7 +20,7 @@ import GHC.Driver.Env.Types (HscEnv) import GHC.Driver.Flags (GeneralFlag (..), DumpFlag(Opt_D_ipe_stats)) import GHC.Driver.DynFlags (gopt, targetPlatform) import GHC.Driver.Config.StgToCmm -import GHC.Driver.Config.Cmm +import GHC.Driver.Config.Cmm ( initCmmConfig ) import GHC.Prelude import GHC.Runtime.Heap.Layout (isStackRep) import GHC.Settings (platformTablesNextToCode) @@ -195,6 +195,7 @@ generateCgIPEStub :: HscEnv -> Module -> InfoTableProvMap + -- ^ If the CmmInfoTables map refer Cmm symbols which were deterministically renamed, the info table provenance map must also be accordingly renamed. -> ( NonCaffySet , ModuleLFInfos , Map CmmInfoTable (Maybe IpeSourceLocation) @@ -209,11 +210,22 @@ generateCgIPEStub hsc_env this_mod denv (nonCaffySet, moduleLFInfos, infoTablesW cmm_cfg = initCmmConfig dflags cgState <- liftIO initC - -- Yield Cmm for Info Table Provenance Entries (IPEs) - let denv' = denv {provInfoTables = Map.mapKeys cit_lbl infoTablesWithTickishes} - ((mIpeStub, ipeCmmGroup), _) = runC (initStgToCmmConfig dflags this_mod) fstate cgState $ getCmm (initInfoTableProv initStats (Map.keys infoTablesWithTickishes) denv') + -- NB: For determinism, don't use DetUniqFM to rename the IPE Cmm because + -- detRenameCmm isn't idempotent and this Cmm references symbols in the rest + -- of the code! Instead, make sure all labels generated for IPE related code + -- sources uniques from the DUniqSupply gotten from CgStream (see its use in + -- initInfoTableProv/emitIpeBufferListNode). + (mIpeStub, ipeCmmGroup) <- liftEff $ UDSMT $ \dus -> do - (_, ipeCmmGroupSRTs) <- liftEff $ withDUS $ cmmPipeline logger cmm_cfg (emptySRT this_mod) ipeCmmGroup + -- Yield Cmm for Info Table Provenance Entries (IPEs) + let denv' = denv {provInfoTables = Map.mapKeys cit_lbl infoTablesWithTickishes} + (((mIpeStub, dus'), ipeCmmGroup), _) = + runC (initStgToCmmConfig dflags this_mod) fstate cgState $ + getCmm (initInfoTableProv initStats (Map.keys infoTablesWithTickishes) denv' dus) + + return ((mIpeStub, ipeCmmGroup), dus') + + (_, ipeCmmGroupSRTs) <- liftEff $ withDUS $ cmmPipeline logger cmm_cfg (emptySRT this_mod) (removeDeterm ipeCmmGroup) Stream.yield ipeCmmGroupSRTs ipeStub <- ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -215,6 +215,7 @@ import GHC.Cmm.Info.Build import GHC.Cmm.Pipeline import GHC.Cmm.Info import GHC.Cmm.Parser +import GHC.Cmm.UniqueRenamer import GHC.Unit import GHC.Unit.Env @@ -299,7 +300,6 @@ import GHC.Stg.InferTags.TagSig (seqTagSig) import GHC.StgToCmm.Utils (IPEStats) import GHC.Types.Unique.FM import GHC.Types.Unique.DFM -import GHC.Types.Unique.DSM import GHC.Cmm.Config (CmmConfig) import Data.Bifunctor @@ -2121,12 +2121,14 @@ hscCompileCmmFile hsc_env original_filename filename output_filename = runHsc hs mod_name = mkModuleName $ "Cmm$" ++ original_filename cmm_mod = mkHomeModule home_unit mod_name cmmpConfig = initCmmParserConfig dflags - (cmm, ipe_ents) <- ioMsgMaybe + (dcmm, ipe_ents) <- ioMsgMaybe $ do (warns,errs,cmm) <- withTiming logger (text "ParseCmm"<+>brackets (text filename)) (\_ -> ()) $ parseCmmFile cmmpConfig cmm_mod home_unit filename let msgs = warns `unionMessages` errs return (GhcPsMessage <$> msgs, cmm) + -- Probably need to rename cmm here + let cmm = removeDeterm dcmm liftIO $ do putDumpFileMaybe logger Opt_D_dump_cmm_verbose_by_proc "Parsed Cmm" FormatCMM (pdoc platform cmm) @@ -2211,11 +2213,11 @@ doCodeGen hsc_env this_mod denv data_tycons putDumpFileMaybe logger Opt_D_dump_stg_final "Final STG:" FormatSTG (pprGenStgTopBindings stg_ppr_opts stg_binds_w_fvs) - let stg_to_cmm dflags mod = case stgToCmmHook hooks of - Nothing -> StgToCmm.codeGen logger tmpfs (initStgToCmmConfig dflags mod) - Just h -> h (initStgToCmmConfig dflags mod) + let stg_to_cmm dflags mod a b c d e = case stgToCmmHook hooks of + Nothing -> StgToCmm.codeGen logger tmpfs (initStgToCmmConfig dflags mod) a b c d e + Just h -> (,emptyDetUFM) <$> h (initStgToCmmConfig dflags mod) a b c d e - let cmm_stream :: CgStream CmmGroup ModuleLFInfos + let cmm_stream :: CgStream CmmGroup (ModuleLFInfos, DetUniqFM) -- See Note [Forcing of stg_binds] cmm_stream = stg_binds_w_fvs `seqList` {-# SCC "StgToCmm" #-} stg_to_cmm dflags this_mod denv data_tycons cost_centre_info stg_binds_w_fvs hpc_info @@ -2237,11 +2239,21 @@ doCodeGen hsc_env this_mod denv data_tycons pipeline_stream :: CgStream CmmGroupSRTs CmmCgInfos pipeline_stream = do - ((mod_srt_info, ipes, ipe_stats), lf_infos) <- + ((mod_srt_info, ipes, ipe_stats), (lf_infos, detRnEnv)) <- {-# SCC "cmmPipeline" #-} Stream.mapAccumL_ (pipeline_action logger cmm_config) (emptySRT this_mod, M.empty, mempty) ppr_stream1 + let nonCaffySet = srtMapNonCAFs (moduleSRTMap mod_srt_info) - cmmCgInfos <- generateCgIPEStub hsc_env this_mod denv (nonCaffySet, lf_infos, ipes, ipe_stats) + + -- denv::InfoTableProvMap refers to symbols that no longer exist + -- if -fobject-determinism is on, since it was created before the + -- Cmm was renamed. Update all the symbols by renaming them with + -- the renaming map in that case. + (_drn, rn_denv) + | gopt Opt_ObjectDeterminism dflags = detRenameIPEMap detRnEnv denv + | otherwise = (detRnEnv, denv) + + cmmCgInfos <- generateCgIPEStub hsc_env this_mod rn_denv (nonCaffySet, lf_infos, ipes, ipe_stats) return cmmCgInfos pipeline_action ===================================== compiler/GHC/Stg/Debug.hs ===================================== @@ -11,6 +11,7 @@ import GHC.Prelude import GHC.Stg.Syntax +import GHC.Types.Unique.DFM import GHC.Types.Id import GHC.Types.Tickish import GHC.Core.DataCon @@ -23,7 +24,6 @@ import Control.Monad (when) import Control.Monad.Trans.Reader import GHC.Utils.Monad.State.Strict import Control.Monad.Trans.Class -import GHC.Types.Unique.Map import GHC.Types.SrcLoc import Control.Applicative import qualified Data.List.NonEmpty as NE @@ -153,7 +153,7 @@ recordStgIdPosition id best_span ss = do --Useful for debugging why a certain Id gets given a certain span --pprTraceM "recordStgIdPosition" (ppr id $$ ppr cc $$ ppr best_span $$ ppr ss) let mbspan = (\(SpanWithLabel rss d) -> (rss, d)) <$> (best_span <|> cc <|> ss) - lift $ modify (\env -> env { provClosure = addToUniqMap (provClosure env) (idName id) (idType id, mbspan) }) + lift $ modify (\env -> env { provClosure = addToUDFM (provClosure env) (idName id) (idName id, (idType id, mbspan)) }) numberDataCon :: DataCon -> [StgTickish] -> M ConstructorNumber -- Unboxed tuples and sums do not allocate so they @@ -166,13 +166,13 @@ numberDataCon dc ts = do env <- lift get mcc <- asks rSpan let !mbest_span = (\(SpanWithLabel rss l) -> (rss, l)) <$> (selectTick ts <|> mcc) - let !dcMap' = alterUniqMap (maybe (Just ((0, mbest_span) :| [] )) - (\xs@((k, _):|_) -> Just $! ((k + 1, mbest_span) `NE.cons` xs))) (provDC env) dc + let !dcMap' = alterUDFM (maybe (Just (dc, (0, mbest_span) :| [] )) + (\(_dc, xs@((k, _):|_)) -> Just $! (dc, (k + 1, mbest_span) `NE.cons` xs))) (provDC env) dc lift $ put (env { provDC = dcMap' }) - let r = lookupUniqMap dcMap' dc + let r = lookupUDFM dcMap' dc return $ case r of Nothing -> NoNumber - Just res -> Numbered (fst (NE.head res)) + Just (_, res) -> Numbered (fst (NE.head res)) selectTick :: [StgTickish] -> Maybe SpanWithLabel selectTick [] = Nothing ===================================== compiler/GHC/StgToCmm.hs ===================================== @@ -44,6 +44,7 @@ import GHC.Types.Id.Info import GHC.Types.RepType import GHC.Types.Basic import GHC.Types.Var.Set ( isEmptyDVarSet ) +import GHC.Types.Unique.DFM import GHC.Types.Unique.FM import GHC.Types.Name.Env @@ -61,7 +62,6 @@ import GHC.Utils.TmpFs import GHC.Data.Stream import GHC.Data.OrdList -import GHC.Types.Unique.Map import Control.Monad (when,void, forM_) import GHC.Utils.Misc @@ -82,7 +82,7 @@ codeGen :: Logger -- Output as a stream, so codegen can -- be interleaved with output -codeGen logger tmpfs cfg (InfoTableProvMap (UniqMap denv) _ _) data_tycons +codeGen logger tmpfs cfg (InfoTableProvMap denv _ _) data_tycons cost_centre_info stg_binds hpc_info = do { -- cg: run the code generator, and yield the resulting CmmGroup -- Using an IORef to store the state is a bit crude, but otherwise @@ -137,7 +137,7 @@ codeGen logger tmpfs cfg (InfoTableProvMap (UniqMap denv) _ _) data_tycons -- Emit special info tables for everything used in this module -- This will only do something if `-fdistinct-info-tables` is turned on. - ; mapM_ (\(dc, ns) -> forM_ ns $ \(k, _ss) -> cg (cgDataCon (UsageSite (stgToCmmThisModule cfg) k) dc)) (nonDetEltsUFM denv) + ; mapM_ (\(dc, ns) -> forM_ ns $ \(k, _ss) -> cg (cgDataCon (UsageSite (stgToCmmThisModule cfg) k) dc)) (eltsUDFM denv) ; final_state <- liftIO (readIORef cgref) ; let cg_id_infos = cgs_binds final_state @@ -158,7 +158,7 @@ codeGen logger tmpfs cfg (InfoTableProvMap (UniqMap denv) _ _) data_tycons ; rn_mapping <- liftIO (readIORef uniqRnRef) ; liftIO $ debugTraceMsg logger 3 (text "DetRnM mapping:" <+> ppr rn_mapping) - ; return generatedInfo + ; return (generatedInfo, rn_mapping) } {- ===================================== compiler/GHC/StgToCmm/CgUtils.hs ===================================== @@ -29,6 +29,7 @@ import GHC.Cmm.Dataflow.Graph import GHC.Cmm.Utils import GHC.Cmm.CLabel import GHC.Utils.Panic +import GHC.Cmm.Dataflow.Label import GHC.Data.Stream (Stream) import GHC.Types.Unique.DSM (UniqDSMT) @@ -150,7 +151,7 @@ fixStgRegisters :: Platform -> RawCmmDecl -> RawCmmDecl fixStgRegisters _ top@(CmmData _ _) = top fixStgRegisters platform (CmmProc info lbl live graph) = - let graph' = modifyGraph (mapGraphBlocks (fixStgRegBlock platform)) graph + let graph' = modifyGraph (mapGraphBlocks mapMap (fixStgRegBlock platform)) graph in CmmProc info lbl live graph' fixStgRegBlock :: Platform -> Block CmmNode e x -> Block CmmNode e x ===================================== compiler/GHC/StgToCmm/InfoTableProv.hs ===================================== @@ -13,6 +13,7 @@ import GHC.IO (unsafePerformIO) import GHC.Prelude import GHC.Platform import GHC.Types.SrcLoc (pprUserRealSpan, srcSpanFile) +import GHC.Types.Unique.DSM import GHC.Unit.Module import GHC.Utils.Outputable import GHC.Data.FastString (fastStringToShortText, unpackFS, LexicalFastString(..)) @@ -70,16 +71,26 @@ construction. emitIpeBufferListNode :: Module -> [InfoProvEnt] - -> FCode () -emitIpeBufferListNode _ [] = return () -emitIpeBufferListNode this_mod ents = do + -> DUniqSupply -- ^ Symbols created source uniques deterministically + -- All uniques must be created from this supply. + -- NB: If you are creating a new symbol within this function, + -- make sure it is local only (as in not `externallyVisibleCLabel`). + -- If you need it to be global, reconsider the comment on the + -- call of emitIpeBufferListNode in Cmm.Parser. + -> FCode DUniqSupply +emitIpeBufferListNode _ [] dus = return dus +emitIpeBufferListNode this_mod ents dus0 = do cfg <- getStgToCmmConfig - tables_lbl <- mkStringLitLabel <$> newUnique - strings_lbl <- mkStringLitLabel <$> newUnique - entries_lbl <- mkStringLitLabel <$> newUnique + let (u1, dus1) = takeUniqueFromDSupply dus0 + (u2, dus2) = takeUniqueFromDSupply dus1 + (u3, dus3) = takeUniqueFromDSupply dus2 - let ctx = stgToCmmContext cfg + tables_lbl = mkStringLitLabel u1 + strings_lbl = mkStringLitLabel u2 + entries_lbl = mkStringLitLabel u3 + + ctx = stgToCmmContext cfg platform = stgToCmmPlatform cfg int n = mkIntCLit platform n @@ -166,6 +177,8 @@ emitIpeBufferListNode this_mod ents = do (Section Data ipe_buffer_lbl) (CmmStaticsRaw ipe_buffer_lbl ipe_buffer_node) + return dus3 + -- | Emit the fields of an IpeBufferEntry struct for each entry in a given list. toIpeBufferEntries :: ByteOrder -- ^ Byte order to write the data in ===================================== compiler/GHC/StgToCmm/Monad.hs ===================================== @@ -76,7 +76,6 @@ import GHC.StgToCmm.Sequel import GHC.Cmm.Graph as CmmGraph import GHC.Cmm.BlockId import GHC.Cmm.CLabel -import GHC.Cmm.Dataflow.Label import GHC.Runtime.Heap.Layout import GHC.Unit import GHC.Types.Id @@ -285,7 +284,7 @@ data CgState = MkCgState { cgs_stmts :: CmmAGraph, -- Current procedure - cgs_tops :: OrdList CmmDecl, + cgs_tops :: OrdList DCmmDecl, -- Other procedures and data blocks in this compilation unit -- Both are ordered only so that we can -- reduce forward references, when it's easy to do so @@ -744,7 +743,7 @@ emit ag = do { state <- getState ; setState $ state { cgs_stmts = cgs_stmts state CmmGraph.<*> ag } } -emitDecl :: CmmDecl -> FCode () +emitDecl :: DCmmDecl -> FCode () emitDecl decl = do { state <- getState ; setState $ state { cgs_tops = cgs_tops state `snocOL` decl } } @@ -787,16 +786,16 @@ emitProc :: Maybe CmmInfoTable -> CLabel -> [GlobalReg] -> CmmAGraphScoped emitProc mb_info lbl live blocks offset do_layout = do { l <- newBlockId ; let - blks :: CmmGraph + blks :: DCmmGraph blks = labelAGraph l blocks - infos | Just info <- mb_info = mapSingleton (g_entry blks) info - | otherwise = mapEmpty + infos | Just info <- mb_info = [((g_entry blks), info)] + | otherwise = [] sinfo = StackInfo { arg_space = offset , do_layout = do_layout } - tinfo = TopInfo { info_tbls = infos + tinfo = TopInfo { info_tbls = DWrap infos , stack_info=sinfo} proc_block = CmmProc tinfo lbl live blks @@ -804,7 +803,7 @@ emitProc mb_info lbl live blocks offset do_layout ; state <- getState ; setState $ state { cgs_tops = cgs_tops state `snocOL` proc_block } } -getCmm :: FCode a -> FCode (a, CmmGroup) +getCmm :: FCode a -> FCode (a, DCmmGroup) -- Get all the CmmTops (there should be no stmts) -- Return a single Cmm which may be split from other Cmms by -- object splitting (at a later stage) @@ -880,7 +879,7 @@ mkCmmCall f results actuals updfr_off -- ---------------------------------------------------------------------------- -- turn CmmAGraph into CmmGraph, for making a new proc. -aGraphToGraph :: CmmAGraphScoped -> FCode CmmGraph +aGraphToGraph :: CmmAGraphScoped -> FCode DCmmGraph aGraphToGraph stmts = do { l <- newBlockId ; return (labelAGraph l stmts) } ===================================== compiler/GHC/StgToCmm/Prof.hs ===================================== @@ -43,6 +43,7 @@ import GHC.Cmm import GHC.Cmm.Utils import GHC.Cmm.CLabel +import GHC.Types.Unique.DSM import GHC.Types.CostCentre import GHC.Types.IPE import GHC.Types.ForeignStubs @@ -279,8 +280,8 @@ sizeof_ccs_words platform -- Note that the stats passed to this function will (rather, should) only ever -- contain stats for skipped STACK info tables accumulated in -- 'generateCgIPEStub'. -initInfoTableProv :: IPEStats -> [CmmInfoTable] -> InfoTableProvMap -> FCode (Maybe (IPEStats, CStub)) -initInfoTableProv stats infos itmap +initInfoTableProv :: IPEStats -> [CmmInfoTable] -> InfoTableProvMap -> DUniqSupply -> FCode (Maybe (IPEStats, CStub), DUniqSupply) +initInfoTableProv stats infos itmap dus = do cfg <- getStgToCmmConfig let (stats', ents) = convertInfoProvMap cfg this_mod itmap stats infos @@ -288,13 +289,13 @@ initInfoTableProv stats infos itmap platform = stgToCmmPlatform cfg this_mod = stgToCmmThisModule cfg case ents of - [] -> return Nothing + [] -> return (Nothing, dus) _ -> do -- Emit IPE buffer - emitIpeBufferListNode this_mod ents + dus' <- emitIpeBufferListNode this_mod ents dus -- Create the C stub which initialises the IPE map - return (Just (stats', ipInitCode info_table platform this_mod)) + return (Just (stats', ipInitCode info_table platform this_mod), dus') -- --------------------------------------------------------------------------- -- Set the current cost centre stack ===================================== compiler/GHC/StgToCmm/Utils.hs ===================================== @@ -86,11 +86,10 @@ import GHC.Types.IPE import qualified Data.Map as M import Data.List (sortBy) import Data.Ord -import GHC.Types.Unique.Map import Data.Maybe import qualified Data.List.NonEmpty as NE import GHC.Core.DataCon -import GHC.Types.Unique.FM +import GHC.Types.Unique.DFM import GHC.Data.Maybe import Control.Monad import qualified Data.Map.Strict as Map @@ -673,7 +672,7 @@ pprIPEStats (IPEStats{..}) = -- for stack info tables skipped during 'generateCgIPEStub'. As the fold -- progresses, counts of tables per closure type will be accumulated. convertInfoProvMap :: StgToCmmConfig -> Module -> InfoTableProvMap -> IPEStats -> [CmmInfoTable] -> (IPEStats, [InfoProvEnt]) -convertInfoProvMap cfg this_mod (InfoTableProvMap (UniqMap dcenv) denv infoTableToSourceLocationMap) initStats cmits = +convertInfoProvMap cfg this_mod (InfoTableProvMap dcenv denv infoTableToSourceLocationMap) initStats cmits = foldl' convertInfoProvMap' (initStats, []) cmits where convertInfoProvMap' :: (IPEStats, [InfoProvEnt]) -> CmmInfoTable -> (IPEStats, [InfoProvEnt]) @@ -686,7 +685,7 @@ convertInfoProvMap cfg this_mod (InfoTableProvMap (UniqMap dcenv) denv infoTable tyString = renderWithContext defaultSDocContext . ppr lookupClosureMap :: Maybe (IPEStats, InfoProvEnt) - lookupClosureMap = case hasHaskellName cl >>= lookupUniqMap denv of + lookupClosureMap = case hasHaskellName cl >>= fmap snd . lookupUDFM denv of Just (ty, mbspan) -> Just (closureIpeStats cn, (InfoProvEnt cl cn (tyString ty) this_mod mbspan)) Nothing -> Nothing @@ -694,7 +693,7 @@ convertInfoProvMap cfg this_mod (InfoTableProvMap (UniqMap dcenv) denv infoTable lookupDataConMap = (closureIpeStats cn,) <$> do UsageSite _ n <- hasIdLabelInfo cl >>= getConInfoTableLocation -- This is a bit grimy, relies on the DataCon and Name having the same Unique, which they do - (dc, ns) <- hasHaskellName cl >>= lookupUFM_Directly dcenv . getUnique + (dc, ns) <- hasHaskellName cl >>= lookupUDFM_Directly dcenv . getUnique -- Lookup is linear but lists will be small (< 100) return $ (InfoProvEnt cl cn (tyString (dataConTyCon dc)) this_mod (join $ lookup n (NE.toList ns))) ===================================== compiler/GHC/Types/IPE.hs ===================================== @@ -13,7 +13,7 @@ import GHC.Data.FastString import GHC.Types.SrcLoc import GHC.Core.DataCon -import GHC.Types.Unique.Map +import GHC.Types.Unique.DFM import GHC.Core.Type import Data.List.NonEmpty import GHC.Cmm.CLabel (CLabel) @@ -25,8 +25,8 @@ type IpeSourceLocation = (RealSrcSpan, LexicalFastString) -- | A map from a 'Name' to the best approximate source position that -- name arose from. -type ClosureMap = UniqMap Name -- The binding - (Type, Maybe IpeSourceLocation) +type ClosureMap = UniqDFM Name -- The binding + (Name, (Type, Maybe IpeSourceLocation)) -- The best approximate source position. -- (rendered type, source position, source note -- label) @@ -38,7 +38,7 @@ type ClosureMap = UniqMap Name -- The binding -- the constructor was used at, if possible and a string which names -- the source location. This is the same information as is the payload -- for the 'GHC.Core.SourceNote' constructor. -type DCMap = UniqMap DataCon (NonEmpty (Int, Maybe IpeSourceLocation)) +type DCMap = UniqDFM DataCon (DataCon, NonEmpty (Int, Maybe IpeSourceLocation)) type InfoTableToSourceLocationMap = Map.Map CLabel (Maybe IpeSourceLocation) @@ -49,4 +49,4 @@ data InfoTableProvMap = InfoTableProvMap } emptyInfoTableProvMap :: InfoTableProvMap -emptyInfoTableProvMap = InfoTableProvMap emptyUniqMap emptyUniqMap Map.empty +emptyInfoTableProvMap = InfoTableProvMap emptyUDFM emptyUDFM Map.empty ===================================== compiler/GHC/Types/Name.hs ===================================== @@ -716,9 +716,9 @@ pprName name@(Name {n_sort = sort, n_uniq = uniq, n_occ = occ}) {-# SPECIALISE pprName :: Name -> SDoc #-} {-# SPECIALISE pprName :: Name -> HLine #-} -- see Note [SPECIALIZE to HDoc] in GHC.Utils.Outputable --- | Print fully qualified name (with unit-id, module and unique) +-- | Print fully qualified name (with unit-id and module, but no unique) pprFullName :: Module -> Name -> SDoc -pprFullName this_mod Name{n_sort = sort, n_uniq = uniq, n_occ = occ} = +pprFullName this_mod Name{n_sort = sort, n_occ = occ} = let mod = case sort of WiredIn m _ _ -> m External m -> m @@ -727,8 +727,6 @@ pprFullName this_mod Name{n_sort = sort, n_uniq = uniq, n_occ = occ} = in ftext (unitIdFS (moduleUnitId mod)) <> colon <> ftext (moduleNameFS $ moduleName mod) <> dot <> ftext (occNameFS occ) - <> char '_' <> pprUniqueAlways uniq - -- | Print a ticky ticky styled name -- ===================================== compiler/GHC/Types/Unique/DFM.hs ===================================== @@ -43,10 +43,10 @@ module GHC.Types.Unique.DFM ( mapMaybeUDFM, mapMUDFM, plusUDFM, - plusUDFM_C, + plusUDFM_C, plusUDFM_CK, lookupUDFM, lookupUDFM_Directly, elemUDFM, - foldUDFM, + foldUDFM, foldWithKeyUDFM, eltsUDFM, filterUDFM, filterUDFM_Directly, isNullUDFM, @@ -56,6 +56,7 @@ module GHC.Types.Unique.DFM ( equalKeysUDFM, minusUDFM, listToUDFM, listToUDFM_Directly, + listToUDFM_C_Directly, udfmMinusUFM, ufmMinusUDFM, partitionUDFM, udfmRestrictKeys, @@ -224,6 +225,12 @@ addListToUDFM_Directly_C addListToUDFM_Directly_C f = foldl' (\m (k, v) -> addToUDFM_C_Directly f m k v) {-# INLINEABLE addListToUDFM_Directly_C #-} +-- | Like 'addListToUDFM_Directly_C' but also passes the unique key to the combine function +addListToUDFM_Directly_CK + :: (Unique -> elt -> elt -> elt) -> UniqDFM key elt -> [(Unique,elt)] -> UniqDFM key elt +addListToUDFM_Directly_CK f = foldl' (\m (k, v) -> addToUDFM_C_Directly (f k) m k v) +{-# INLINEABLE addListToUDFM_Directly_CK #-} + delFromUDFM :: Uniquable key => UniqDFM key elt -> key -> UniqDFM key elt delFromUDFM (UDFM m i) k = UDFM (M.delete (getKey $ getUnique k) m) i @@ -234,6 +241,15 @@ plusUDFM_C f udfml@(UDFM _ i) udfmr@(UDFM _ j) | i > j = insertUDFMIntoLeft_C f udfml udfmr | otherwise = insertUDFMIntoLeft_C f udfmr udfml +-- | Like 'plusUDFM_C' but the combine function also receives the unique key +plusUDFM_CK :: (Unique -> elt -> elt -> elt) -> UniqDFM key elt -> UniqDFM key elt -> UniqDFM key elt +plusUDFM_CK f udfml@(UDFM _ i) udfmr@(UDFM _ j) + -- we will use the upper bound on the tag as a proxy for the set size, + -- to insert the smaller one into the bigger one + | i > j = insertUDFMIntoLeft_CK f udfml udfmr + | otherwise = insertUDFMIntoLeft_CK f udfmr udfml + + -- Note [Overflow on plusUDFM] -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~ -- There are multiple ways of implementing plusUDFM. @@ -282,6 +298,12 @@ insertUDFMIntoLeft_C insertUDFMIntoLeft_C f udfml udfmr = addListToUDFM_Directly_C f udfml $ udfmToList udfmr +-- | Like 'insertUDFMIntoLeft_C', but the merge function also receives the unique key +insertUDFMIntoLeft_CK + :: (Unique -> elt -> elt -> elt) -> UniqDFM key elt -> UniqDFM key elt -> UniqDFM key elt +insertUDFMIntoLeft_CK f udfml udfmr = + addListToUDFM_Directly_CK f udfml $ udfmToList udfmr + lookupUDFM :: Uniquable key => UniqDFM key elt -> key -> Maybe elt lookupUDFM (UDFM m _i) k = taggedFst `fmap` M.lookup (getKey $ getUnique k) m @@ -298,6 +320,12 @@ foldUDFM :: (elt -> a -> a) -> a -> UniqDFM key elt -> a -- This INLINE prevents a regression in !10568 foldUDFM k z m = foldr k z (eltsUDFM m) +-- | Like 'foldUDFM' but the function also receives a key +foldWithKeyUDFM :: (Unique -> elt -> a -> a) -> a -> UniqDFM key elt -> a +{-# INLINE foldWithKeyUDFM #-} +-- This INLINE was copied from foldUDFM +foldWithKeyUDFM k z m = foldr (uncurry k) z (udfmToList m) + -- | Performs a nondeterministic strict fold over the UniqDFM. -- It's O(n), same as the corresponding function on `UniqFM`. -- If you use this please provide a justification why it doesn't introduce @@ -397,6 +425,9 @@ listToUDFM = foldl' (\m (k, v) -> addToUDFM m k v) emptyUDFM listToUDFM_Directly :: [(Unique, elt)] -> UniqDFM key elt listToUDFM_Directly = foldl' (\m (u, v) -> addToUDFM_Directly m u v) emptyUDFM +listToUDFM_C_Directly :: (elt -> elt -> elt) -> [(Unique, elt)] -> UniqDFM key elt +listToUDFM_C_Directly f = foldl' (\m (u, v) -> addToUDFM_C_Directly f m u v) emptyUDFM + -- | Apply a function to a particular element adjustUDFM :: Uniquable key => (elt -> elt) -> UniqDFM key elt -> key -> UniqDFM key elt adjustUDFM f (UDFM m i) k = UDFM (M.adjust (fmap f) (getKey $ getUnique k) m) i ===================================== compiler/GHC/Utils/Outputable.hs ===================================== @@ -142,6 +142,7 @@ import Data.Int import qualified Data.IntMap as IM import Data.Set (Set) import qualified Data.Set as Set +import qualified Data.IntSet as IntSet import qualified GHC.Data.Word64Set as Word64Set import Data.String import Data.Word @@ -991,6 +992,9 @@ instance (Outputable a) => Outputable (Set a) where instance Outputable Word64Set.Word64Set where ppr s = braces (pprWithCommas ppr (Word64Set.toList s)) +instance Outputable IntSet.IntSet where + ppr s = braces (pprWithCommas ppr (IntSet.toList s)) + instance (Outputable a, Outputable b) => Outputable (a, b) where ppr (x,y) = parens (sep [ppr x <> comma, ppr y]) @@ -1200,6 +1204,9 @@ instance OutputableP env a => OutputableP env [a] where instance OutputableP env a => OutputableP env (Maybe a) where pdoc env xs = ppr (fmap (pdoc env) xs) +instance OutputableP env () where + pdoc _ _ = ppr () + instance (OutputableP env a, OutputableP env b) => OutputableP env (a, b) where pdoc env (a,b) = ppr (pdoc env a, pdoc env b) ===================================== testsuite/tests/regalloc/regalloc_unit_tests.hs ===================================== @@ -137,7 +137,9 @@ compileCmmForRegAllocStats logger home_unit dflags cmmFile ncgImplF us = do -- parse the cmm file and output any warnings or errors let fake_mod = mkHomeModule home_unit (mkModuleName "fake") cmmpConfig = initCmmParserConfig dflags - (warnings, errors, parsedCmm) <- parseCmmFile cmmpConfig fake_mod home_unit cmmFile + (warnings, errors, dparsedCmm) <- parseCmmFile cmmpConfig fake_mod home_unit cmmFile + + let parsedCmm = removeDeterm (fst (fromJust dparsedCmm)) -- print parser errors or warnings let !diag_opts = initDiagOpts dflags ===================================== testsuite/tests/wasm/should_run/control-flow/LoadCmmGroup.hs ===================================== @@ -122,7 +122,7 @@ slurpCmm hsc_env filename = runHsc hsc_env $ do $ parseCmmFile cmmpConfig cmm_mod home_unit filename let msgs = warns `unionMessages` errs return (GhcPsMessage <$> msgs, cmm) - return cmm + return (removeDeterm cmm) collectAll :: Monad m => Stream m a b -> m ([a], b) collectAll = gobble . runStream View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4b1fc47dfc8ec3abbbba36bc6dfba2ff1824cf69...ed43ad4297715de55a9f7e3a9be4017a2f55f1a2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4b1fc47dfc8ec3abbbba36bc6dfba2ff1824cf69...ed43ad4297715de55a9f7e3a9be4017a2f55f1a2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 18 21:27:55 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Wed, 18 Sep 2024 17:27:55 -0400 Subject: [Git][ghc/ghc][wip/T24807] 4404 commits: [haddock @ 2002-04-04 16:23:43 by simonmar] Message-ID: <66eb45db93951_3ac2741213eb81340c4@gitlab.mail> Ben Gamari pushed to branch wip/T24807 at Glasgow Haskell Compiler / GHC Commits: 2b39cd94 by Simon Marlow at 2002-04-04T16:23:43+00:00 [haddock @ 2002-04-04 16:23:43 by simonmar] This is Haddock, my stab at a Haskell documentation tool. It's not quite ready for release yet, but I'm putting it in the repository so others can take a look. It uses a locally modified version of the hssource parser, extended with support for GHC extensions and documentation annotations. - - - - - 99ede94f by Simon Marlow at 2002-04-04T16:24:10+00:00 [haddock @ 2002-04-04 16:24:10 by simonmar] forgot one file - - - - - 8363294c by Simon Marlow at 2002-04-05T13:58:15+00:00 [haddock @ 2002-04-05 13:58:15 by simonmar] Remap names in the exported declarations to be "closer" to the current module. eg. if an exported declaration mentions a type 'T' which is imported from module A then re-exported from the current module, then links from the type or indeed the documentation will point to the current module rather than module A. This is to support better hiding: module A won't be referred to in the generated output. - - - - - 1570cbc1 by Simon Marlow at 2002-04-05T13:58:23+00:00 [haddock @ 2002-04-05 13:58:23 by simonmar] update the TODO list - - - - - 3a62f96b by Simon Marlow at 2002-04-05T14:11:51+00:00 [haddock @ 2002-04-05 14:11:51 by simonmar] Fix the anchor for a class declaration - - - - - c5d9a471 by Simon Marlow at 2002-04-05T14:18:41+00:00 [haddock @ 2002-04-05 14:18:41 by simonmar] remove underlines on visited links - - - - - 97280525 by Simon Marlow at 2002-04-05T16:11:47+00:00 [haddock @ 2002-04-05 16:11:47 by simonmar] - Update to generate more correct HTML. - Use our own non-overloaded table combinators, as the overloaded versions were giving me a headache. The improved type safety caught several errors in the HTML generation. - - - - - 9acd3a4d by Simon Marlow at 2002-04-05T16:32:19+00:00 [haddock @ 2002-04-05 16:32:19 by simonmar] Add width property to the title, and add TD.children for the module contents page. - - - - - ec9a0847 by Simon Marlow at 2002-04-08T16:39:56+00:00 [haddock @ 2002-04-08 16:39:56 by simonmar] Fix a problem with exports of the form T(..). - - - - - e4627dc8 by Simon Marlow at 2002-04-08T16:41:38+00:00 [haddock @ 2002-04-08 16:41:37 by simonmar] - Add our own versions of Html & BlockTable for the time being. - Add support for generating an index to the HTML backend - - - - - 2d73fd75 by Simon Marlow at 2002-04-09T11:23:24+00:00 [haddock @ 2002-04-09 11:23:24 by simonmar] Add '-- /' as a synonym for '-- |', for compatibility with IDoc. - - - - - 3675464e by Simon Marlow at 2002-04-09T11:33:55+00:00 [haddock @ 2002-04-09 11:33:54 by simonmar] - add the <...> syntax for marking up URLs in documentation - Make the output for data & class declarations more compact when there aren't any documentation annotations on the individual methods or constructors respectively. - - - - - 5077f5b1 by Simon Marlow at 2002-04-09T11:36:04+00:00 [haddock @ 2002-04-09 11:36:04 by simonmar] Update the TODO list - - - - - 9e83c54d by Simon Marlow at 2002-04-10T10:50:06+00:00 [haddock @ 2002-04-10 10:50:06 by simonmar] Use explicit 'px' suffix on pixel sizes; IE seems to prefer them - - - - - 052de51c by Simon Marlow at 2002-04-10T13:23:13+00:00 [haddock @ 2002-04-10 13:23:13 by simonmar] Lex URLs as a single token to avoid having to escape special characters inside the URL string. - - - - - 47187edb by Simon Marlow at 2002-04-10T13:23:55+00:00 [haddock @ 2002-04-10 13:23:55 by simonmar] Not sure why I made the constructor name for a record declaration into a TyCls name, but change it back into a Var name anyhow. - - - - - 3dc6aa81 by Simon Marlow at 2002-04-10T13:26:10+00:00 [haddock @ 2002-04-10 13:26:09 by simonmar] Lots of changes, including: - add index support to the HTML backend - clean up the renamer, put it into a monad - propogate unresolved names to the top level and report them in a nicer way - various bugfixes - - - - - c2a70a72 by Simon Marlow at 2002-04-10T13:32:39+00:00 [haddock @ 2002-04-10 13:32:39 by simonmar] Skeleton documentation - - - - - 50c98d17 by Simon Marlow at 2002-04-10T13:37:23+00:00 [haddock @ 2002-04-10 13:37:23 by simonmar] Update the TODO list, separate into pre-1.0 and post-1.0 items - - - - - f3778be6 by Simon Marlow at 2002-04-10T14:30:58+00:00 [haddock @ 2002-04-10 14:30:58 by simonmar] Add an introduction - - - - - cfbaf9f7 by Simon Marlow at 2002-04-10T14:59:51+00:00 [haddock @ 2002-04-10 14:59:51 by simonmar] Sort the module tree - - - - - 76bd7b34 by Simon Marlow at 2002-04-10T15:50:11+00:00 [haddock @ 2002-04-10 15:50:10 by simonmar] Generate a little table of contents at the top of the module doc (only if the module actually contains some section headings, though). - - - - - bb8560a1 by Simon Marlow at 2002-04-10T16:10:26+00:00 [haddock @ 2002-04-10 16:10:26 by simonmar] Now we understand (or at least don't barf on) type signatures in patterns such as you might find when scoped type variables are in use. - - - - - 86c2a026 by Simon Marlow at 2002-04-10T16:10:49+00:00 [haddock @ 2002-04-10 16:10:49 by simonmar] more updates - - - - - 1c052b0e by Simon Marlow at 2002-04-10T16:28:05+00:00 [haddock @ 2002-04-10 16:28:05 by simonmar] Parse errors in doc strings are now reported as warnings rather that causing the whole thing to fall over. It still needs cleaning up (the warning is emitted with trace) but this will do for the time being. - - - - - ace03e8f by Simon Marlow at 2002-04-10T16:38:03+00:00 [haddock @ 2002-04-10 16:38:03 by simonmar] update again - - - - - 69006c3e by Simon Marlow at 2002-04-11T13:38:02+00:00 [haddock @ 2002-04-11 13:38:02 by simonmar] mention Opera - - - - - fe9b10f8 by Simon Marlow at 2002-04-11T13:40:31+00:00 [haddock @ 2002-04-11 13:40:30 by simonmar] - copy haddock.css into the same place as the generated HTML - new option: --css <file> specifies the style sheet to use - new option: -o <dir> specifies the directory in which to generate the output. - because Haddock now needs to know where to find its default stylesheet, we have to have a wrapper script and do the haddock-inplace thing (Makefile code copied largely from fptools/happy). - - - - - 106adbbe by Simon Marlow at 2002-04-24T15:12:41+00:00 [haddock @ 2002-04-24 15:12:41 by simonmar] Stop slurping comment lines when we see a row of dashes longer than length 2: these are useful as separators. - - - - - 995d3f9e by Simon Marlow at 2002-04-24T15:14:12+00:00 [haddock @ 2002-04-24 15:14:11 by simonmar] Grok the kind of module headers we use in fptools/libraries, and pass the "portability", "stability", and "maintainer" strings through into the generated HTML. If the module header doesn't match the pattern, then we don't include the info in the HTML. - - - - - e14da136 by Simon Marlow at 2002-04-24T15:16:57+00:00 [haddock @ 2002-04-24 15:16:57 by simonmar] Done module headers now. - - - - - 2ca8dfd4 by Simon Marlow at 2002-04-24T15:57:48+00:00 [haddock @ 2002-04-24 15:57:47 by simonmar] Handle gcons in export lists (a common extension). - - - - - 044cea81 by Simon Marlow at 2002-04-25T14:20:12+00:00 [haddock @ 2002-04-25 14:20:12 by simonmar] Add the little lambda icon - - - - - 63955027 by Simon Marlow at 2002-04-25T14:40:05+00:00 [haddock @ 2002-04-25 14:40:05 by simonmar] - Add support for named chunks of documentation which can be referenced from the export list. - Copy the icon from $libdir to the destination in HTML mode. - - - - - 36e3f913 by Simon Marlow at 2002-04-25T16:48:36+00:00 [haddock @ 2002-04-25 16:48:36 by simonmar] More keyboard bashing - - - - - 7ae18dd0 by Simon Marlow at 2002-04-26T08:43:33+00:00 [haddock @ 2002-04-26 08:43:33 by simonmar] Package util reqd. to compile with 4.08.2 - - - - - bbd5fbab by Simon Marlow at 2002-04-26T10:13:00+00:00 [haddock @ 2002-04-26 10:13:00 by simonmar] Include $(GHC_HAPPY_OPTS) when compiling HsParser - - - - - 31c53d79 by Simon Marlow at 2002-04-26T11:18:57+00:00 [haddock @ 2002-04-26 11:18:56 by simonmar] - support for fundeps (partially contributed by Brett Letner - thanks Brett). - make it build with GHC 4.08.2 - - - - - c415ce76 by Simon Marlow at 2002-04-26T13:15:02+00:00 [haddock @ 2002-04-26 13:15:02 by simonmar] Move the explicit formatting of the little table for the stability/portability/maintainer info from the HTML into the CSS, and remove the explicit table size (just right-align it). - - - - - 520ee21a by Simon Marlow at 2002-04-26T16:01:44+00:00 [haddock @ 2002-04-26 16:01:44 by simonmar] Yet more keyboard bashing - this is pretty much complete now. - - - - - 2ae37179 by Simon Marlow at 2002-04-26T16:02:14+00:00 [haddock @ 2002-04-26 16:02:14 by simonmar] Add a couple of things I forgot about - - - - - b7211e04 by Simon Marlow at 2002-04-29T15:28:12+00:00 [haddock @ 2002-04-29 15:28:12 by simonmar] bugfix for declBinders on a NewTypeDecl - - - - - 640c154a by Simon Marlow at 2002-04-29T15:28:54+00:00 [haddock @ 2002-04-29 15:28:54 by simonmar] Allow '-- |' style annotations on constructors and record fields. - - - - - 393f258a by Simon Marlow at 2002-04-29T15:37:32+00:00 [haddock @ 2002-04-29 15:37:32 by simonmar] syntax fix - - - - - 8a2c2549 by Simon Marlow at 2002-04-29T15:37:48+00:00 [haddock @ 2002-04-29 15:37:48 by simonmar] Add an example - - - - - db88f8a2 by Simon Marlow at 2002-04-29T15:55:46+00:00 [haddock @ 2002-04-29 15:55:46 by simonmar] remove a trace - - - - - 2b0248e0 by Simon Marlow at 2002-04-29T15:56:19+00:00 [haddock @ 2002-04-29 15:56:19 by simonmar] Fix for 'make install' - - - - - 120453a0 by Simon Marlow at 2002-04-29T15:56:39+00:00 [haddock @ 2002-04-29 15:56:39 by simonmar] Install the auxilliary bits - - - - - 950e6dbb by Simon Marlow at 2002-04-29T15:57:30+00:00 [haddock @ 2002-04-29 15:57:30 by simonmar] Add BinDist bits - - - - - 154b9d71 by Simon Marlow at 2002-05-01T11:02:52+00:00 [haddock @ 2002-05-01 11:02:52 by simonmar] update - - - - - ba6c39fa by Simon Marlow at 2002-05-01T11:03:26+00:00 [haddock @ 2002-05-01 11:03:26 by simonmar] Add another item - - - - - bacb5e33 by Simon Marlow at 2002-05-03T08:50:00+00:00 [haddock @ 2002-05-03 08:50:00 by simonmar] Fix some typos. - - - - - 54c87895 by Sven Panne at 2002-05-05T19:40:51+00:00 [haddock @ 2002-05-05 19:40:51 by panne] As a temporary hack/workaround for a bug in GHC's simplifier, don't pass Happy the -c option for generating the parsers in this subdir. Furthermore, disable -O for HaddocParse, too. - - - - - e6c08703 by Simon Marlow at 2002-05-06T09:51:10+00:00 [haddock @ 2002-05-06 09:51:10 by simonmar] Add RPM spec file (thanks to Tom Moertel <tom-rpms at moertel.com>) - - - - - 7b8fa8e7 by Simon Marlow at 2002-05-06T12:29:26+00:00 [haddock @ 2002-05-06 12:29:26 by simonmar] Add missing type signature (a different workaround for the bug in GHC's simplifier). - - - - - cd0e300d by Simon Marlow at 2002-05-06T12:30:09+00:00 [haddock @ 2002-05-06 12:30:09 by simonmar] Remove workaround for simplifier bug in previous revision. - - - - - 687e68fa by Simon Marlow at 2002-05-06T12:32:32+00:00 [haddock @ 2002-05-06 12:32:32 by simonmar] Allow empty data declarations (another GHC extension). - - - - - 8f29f696 by Simon Marlow at 2002-05-06T12:49:21+00:00 [haddock @ 2002-05-06 12:49:21 by simonmar] Fix silly bug in named documentation block lookup. - - - - - 8e0059af by Simon Marlow at 2002-05-06T13:02:42+00:00 [haddock @ 2002-05-06 13:02:42 by simonmar] Add another named chunk with a different name - - - - - 68f8a896 by Simon Marlow at 2002-05-06T13:32:32+00:00 [haddock @ 2002-05-06 13:32:32 by simonmar] Be more lenient about extra paragraph breaks - - - - - 65fc31db by Simon Marlow at 2002-05-07T15:36:36+00:00 [haddock @ 2002-05-07 15:36:36 by simonmar] DocEmpty is a right and left-unit of DocAppend (remove it in the smart constructor). - - - - - adc81078 by Simon Marlow at 2002-05-07T15:37:15+00:00 [haddock @ 2002-05-07 15:37:15 by simonmar] Allow code blocks to be denoted with bird-tracks in addition to [...]. - - - - - 1283a3c1 by Simon Marlow at 2002-05-08T11:21:56+00:00 [haddock @ 2002-05-08 11:21:56 by simonmar] Add a facility for specifying options that affect Haddock's treatment of the module. Options are given at the top of the module in a comma-separated list, beginning with '-- #'. eg. -- # prune, hide, ignore-exports Options currently available, with their meanings: prune: ignore declarations which have no documentation annotations ignore-exports: act as if the export list were not specified (i.e. export everything local to the module). hide: do not include this module in the generated documentation, but propagate any exported definitions to modules which re-export them. There's a slight change in the semantics for re-exporting a full module by giving 'module M' in the export list: if module M does not have the 'hide' option, then the documentation will now just contain a reference to module M rather than the full inlined contents of that module. These features, and some other changes in the pipeline, are the result of discussions between myself and Manuel Chakravarty <chak at cse.unsw.edu.au> (author of IDoc) yesterday. Also: some cleanups, use a Writer monad to collect error messages in some places instead of just printing them with trace. - - - - - a2239cf5 by Simon Marlow at 2002-05-08T11:22:30+00:00 [haddock @ 2002-05-08 11:22:30 by simonmar] Update to test new features. - - - - - 6add955f by Simon Marlow at 2002-05-08T13:37:25+00:00 [haddock @ 2002-05-08 13:37:25 by simonmar] Change the markup for typewriter-font from [...] to @... at . The reasoning is that the '@' symbol is much less likely to be needed than square brackets, and we don't want to have to escape square brackets in code fragments. This will be mildly painful in the short term, but it's better to get the change out of the way as early as possible. - - - - - cda06447 by Simon Marlow at 2002-05-08T13:39:56+00:00 [haddock @ 2002-05-08 13:39:56 by simonmar] Allow nested-style comments to be used as documentation annotations too. eg. {-| ... -} is equivalent to -- | ... An extra space can also be left after the comment opener: {- | ... -}. The only version that isn't allowed is {-# ... -}, because this syntax overlaps with Haskell pragmas; use {- # ... -} instead. - - - - - db23f65e by Simon Marlow at 2002-05-08T14:48:41+00:00 [haddock @ 2002-05-08 14:48:39 by simonmar] Add support for existential quantifiers on constructors. - - - - - adce3794 by Simon Marlow at 2002-05-08T15:43:25+00:00 [haddock @ 2002-05-08 15:43:25 by simonmar] update - - - - - 62a1f436 by Simon Marlow at 2002-05-08T15:44:10+00:00 [haddock @ 2002-05-08 15:44:10 by simonmar] Update to version 0.2 - - - - - f6a24ba3 by Simon Marlow at 2002-05-09T08:48:29+00:00 [haddock @ 2002-05-09 08:48:29 by simonmar] typo - - - - - 9f9522a4 by Simon Marlow at 2002-05-09T10:33:14+00:00 [haddock @ 2002-05-09 10:33:14 by simonmar] oops, left out '/' from the special characters in the last change. - - - - - 14abcb39 by Simon Marlow at 2002-05-09T10:34:44+00:00 [haddock @ 2002-05-09 10:34:44 by simonmar] Fix buglet - - - - - b8d878be by Simon Marlow at 2002-05-09T10:35:00+00:00 [haddock @ 2002-05-09 10:35:00 by simonmar] Give a more useful instance of Show for Module. - - - - - f7bfd626 by Simon Marlow at 2002-05-09T10:37:07+00:00 [haddock @ 2002-05-09 10:37:07 by simonmar] The last commit to Main.lhs broke the delicate balance of laziness which was being used to avoid computing the dependency graph of modules. So I finally bit the bullet and did a proper topological sort of the module graph, which turned out to be easy (stealing the Digraph module from GHC - this really ought to be in the libraries somewhere). - - - - - b481c1d0 by Simon Marlow at 2002-05-09T10:37:25+00:00 [haddock @ 2002-05-09 10:37:25 by simonmar] another item done - - - - - 032e2b42 by Simon Marlow at 2002-05-09T10:44:15+00:00 [haddock @ 2002-05-09 10:44:15 by simonmar] Don't consider a module re-export as having documentation, for the purposes of deciding whether we need a Synopsis section or not. - - - - - 5fb45e92 by Simon Marlow at 2002-05-09T11:10:55+00:00 [haddock @ 2002-05-09 11:10:55 by simonmar] Add a special case for list types in ppHsAType - - - - - 1937e428 by Simon Marlow at 2002-05-09T12:43:06+00:00 [haddock @ 2002-05-09 12:43:06 by simonmar] Type synonyms can accept a ctype on the RHS, to match GHC. - - - - - 0f16ce56 by Simon Marlow at 2002-05-09T12:45:19+00:00 [haddock @ 2002-05-09 12:45:19 by simonmar] Add 'stdcall' keyword - - - - - 29b0d7d2 by Simon Marlow at 2002-05-09T13:35:45+00:00 [haddock @ 2002-05-09 13:35:45 by simonmar] Add System Requirements section - - - - - bf14dddd by Simon Marlow at 2002-05-09T13:36:11+00:00 [haddock @ 2002-05-09 13:36:11 by simonmar] Test existential types, amongst other things - - - - - 502f8f6f by Simon Marlow at 2002-05-09T13:37:35+00:00 [haddock @ 2002-05-09 13:37:35 by simonmar] Print the module name in a doc-string parse error - - - - - ca1f8d49 by Simon Marlow at 2002-05-09T13:38:04+00:00 [haddock @ 2002-05-09 13:38:04 by simonmar] Add dependency - - - - - 8d3d91ff by Simon Marlow at 2002-05-09T15:37:57+00:00 [haddock @ 2002-05-09 15:37:57 by simonmar] Add the changelog/release notes - - - - - f3960959 by Simon Marlow at 2002-05-09T15:47:47+00:00 [haddock @ 2002-05-09 15:47:47 by simonmar] mention the backquote-style of markup - - - - - 089fb6e6 by Simon Marlow at 2002-05-09T15:59:45+00:00 [haddock @ 2002-05-09 15:59:45 by simonmar] update - - - - - bdd3be0b by Simon Marlow at 2002-05-09T15:59:56+00:00 [haddock @ 2002-05-09 15:59:56 by simonmar] Document changes since 0.1 - - - - - 00fc4af8 by Simon Marlow at 2002-05-10T08:22:48+00:00 [haddock @ 2002-05-10 08:22:48 by simonmar] oops, update to version 0.2 - - - - - a8a79041 by Simon Marlow at 2002-05-10T16:05:08+00:00 [haddock @ 2002-05-10 16:05:08 by simonmar] Only include a mini-contents if there are 2 or more sections - - - - - 06653319 by Simon Marlow at 2002-05-13T09:13:12+00:00 [haddock @ 2002-05-13 09:13:12 by simonmar] fix typos - - - - - 1402b19b by Simon Marlow at 2002-05-13T10:14:22+00:00 [haddock @ 2002-05-13 10:14:22 by simonmar] Allow backquote as the right-hand quote as well as the left-hand quote, as suggested by Dean Herrington. Clean up the grammar a litte. - - - - - dcd5320d by Simon Marlow at 2002-05-13T10:44:10+00:00 [haddock @ 2002-05-13 10:44:10 by simonmar] a couple more things, prioritise a bit - - - - - a90130c4 by Simon Marlow at 2002-05-13T15:19:03+00:00 [haddock @ 2002-05-13 15:19:03 by simonmar] Cope with datatypes which have documentation on the constructor but not the type itself, and records which have documentation on the fields but not the constructor. (Thanks to Ross Paterson for pointing out the bugs). - - - - - a774d432 by Simon Marlow at 2002-05-13T15:20:54+00:00 [haddock @ 2002-05-13 15:20:54 by simonmar] Fix one of the record examples - - - - - 2d1d5218 by Simon Marlow at 2002-05-15T12:44:35+00:00 [haddock @ 2002-05-15 12:44:35 by simonmar] Preserve the newline before a bird-track, but only within a paragraph. - - - - - 1554c09a by Simon Marlow at 2002-05-15T13:03:02+00:00 [haddock @ 2002-05-15 13:03:01 by simonmar] Reworking of the internals to support documenting function arguments (the Most Wanted new feature by the punters). The old method of keeping parsed documentation in a Name -> Doc mapping wasn't going to cut it for anntations on type components, where there's no name to attach the documentation to, so I've moved to storing all the documentation in the abstract syntax. Previously some of the documentation was left in the abstract syntax by the parser, but was later extracted into the mapping. In order to avoid having to parameterise the abstract syntax over the type of documentation stored in it, we have to parse the documentation at the same time as we parse the Haskell source (well, I suppose we could store 'Either String Doc' in the HsSyn, but that's clunky). One upshot is that documentation is now parsed eagerly, and documentation parse errors are fatal (but have better line numbers in the error message). The new story simplifies matters for the code that processes the source modules, because we don't have to maintain the extra Name->Doc mapping, and it should improve efficiency a little too. New features: - Function arguments and return values can now have doc annotations. - If you refer to a qualified name in a doc string, eg. 'IO.putStr', then Haddock will emit a hyperlink even if the identifier is not in scope, so you don't have to make sure everything referred to from the documentation is imported. - several bugs & minor infelicities fixed. - - - - - 57344dc3 by Simon Marlow at 2002-05-15T13:03:19+00:00 [haddock @ 2002-05-15 13:03:19 by simonmar] Bump to version 0.3 - - - - - b2791812 by Simon Marlow at 2002-05-15T13:03:41+00:00 [haddock @ 2002-05-15 13:03:41 by simonmar] update - - - - - fead183e by Simon Marlow at 2002-05-15T13:10:15+00:00 [haddock @ 2002-05-15 13:10:15 by simonmar] Rename Foo.hs to Test.hs, and add a Makefile - - - - - b0b1f89f by Simon Marlow at 2002-05-15T13:16:07+00:00 [haddock @ 2002-05-15 13:16:07 by simonmar] - Remove the note about function argument docs not being implemented - Note that qualified identifiers can be used to point to entities that aren't in scope. - - - - - 5665f31a by Simon Marlow at 2002-05-15T13:28:46+00:00 [haddock @ 2002-05-15 13:28:46 by simonmar] Patch to add support for GHC-style primitive strings ".."#, from Ross Paterson. - - - - - 0564505d by Simon Marlow at 2002-05-17T10:51:57+00:00 [haddock @ 2002-05-17 10:51:57 by simonmar] Fix bugs in qualified name handling (A.B.f was returned as B.f) - - - - - 10e7311c by Simon Marlow at 2002-05-21T10:24:52+00:00 [haddock @ 2002-05-21 10:24:52 by simonmar] - Use an alternate tabular layout for datatypes, which is more compact - Fix some problems with the function argument documentation - - - - - 2f91c2a6 by Simon Marlow at 2002-05-21T10:27:40+00:00 [haddock @ 2002-05-21 10:27:40 by simonmar] add a few more test cases - - - - - 01c2ddd2 by Simon Marlow at 2002-05-21T10:28:33+00:00 [haddock @ 2002-05-21 10:28:33 by simonmar] Rearrange a bit, and add support for tabular datatype rendering - - - - - a4e4c5f8 by Simon Marlow at 2002-05-27T09:03:52+00:00 [haddock @ 2002-05-27 09:03:51 by simonmar] Lots of changes: - instances of a class are listed with the class, and instances involving a datatype are listed with that type. Derived instances aren't included at the moment: the calculation to find the instance head for a derived instance is non-trivial. - some formatting changes; use rows with specified height rather than cellspacing in some places. - various fixes (source file links were wrong, amongst others) - - - - - 48722e68 by Simon Marlow at 2002-05-27T12:30:38+00:00 [haddock @ 2002-05-27 12:30:37 by simonmar] - Put function arguments *before* the doc for the function, as suggested by Sven Panne. This looks nicer when the function documentation is long. - Switch to using bold for binders at the definition site, and use underline for keywords. This makes the binder stand out more. - - - - - 657204d2 by Simon Marlow at 2002-05-27T13:19:49+00:00 [haddock @ 2002-05-27 13:19:49 by simonmar] Fix bug: we weren't renaming HsDocCommentNamed in renameDecl - - - - - 592aae66 by Simon Marlow at 2002-05-27T14:10:27+00:00 [haddock @ 2002-05-27 14:10:27 by simonmar] Fix some bugs in the rendering of qualified type signatures. - - - - - 69c8f763 by Simon Marlow at 2002-05-27T14:36:45+00:00 [haddock @ 2002-05-27 14:36:45 by simonmar] warning message tweak - - - - - 16e64e21 by Simon Marlow at 2002-05-27T14:53:53+00:00 [haddock @ 2002-05-27 14:53:53 by simonmar] hyperlinked identifiers should be in <tt> - - - - - 8d5e4783 by Simon Marlow at 2002-05-27T15:56:45+00:00 [haddock @ 2002-05-27 15:56:45 by simonmar] Do something sensible for modules which don't export anything (except instances). - - - - - 9d3ef811 by Simon Marlow at 2002-05-28T10:12:50+00:00 [haddock @ 2002-05-28 10:12:50 by simonmar] Rename the module documentation properly (bug reported by Sven Panne). - - - - - ef03a1cc by Simon Marlow at 2002-05-28T10:13:04+00:00 [haddock @ 2002-05-28 10:13:04 by simonmar] Add some more test cases - - - - - 92baa0e8 by Simon Marlow at 2002-05-28T11:17:55+00:00 [haddock @ 2002-05-28 11:17:55 by simonmar] If an identifier doesn't lex, then just replace it by a DocString. - - - - - a3156213 by Simon Marlow at 2002-05-28T16:16:19+00:00 [haddock @ 2002-05-28 16:16:19 by simonmar] Only link to names in the current module which are actually listed in the documentation. A name may be exported but not present in the documentation if it is exported as part of a 'module M' export specifier. - - - - - 31acf941 by Simon Marlow at 2002-05-28T16:17:11+00:00 [haddock @ 2002-05-28 16:17:11 by simonmar] update - - - - - 7e474ebf by Sigbjorn Finne at 2002-05-28T22:42:08+00:00 [haddock @ 2002-05-28 22:42:08 by sof] Handle lone occurrences of '/', e.g., -- | This/that. [did this in the lexer rather than in the parser, as I couldn't see a way not to introduce an S/R conflict that way.] - - - - - 093f7e53 by Simon Marlow at 2002-05-29T09:09:49+00:00 [haddock @ 2002-05-29 09:09:49 by simonmar] Back out previous change until we can find a better way to do this. - - - - - 9234389c by Simon Marlow at 2002-05-29T13:19:06+00:00 [haddock @ 2002-05-29 13:19:06 by simonmar] Make the markup syntax a little more friendly: - single quotes are now interpreted literally unless they surround a valid Haskell identifier. So for example now there's no need to escape a single quote used as an apostrophe. - text to the right of a bird track is now literal (if you want marked-up text in a code block, use @...@). - - - - - b3333526 by Simon Marlow at 2002-05-29T13:38:51+00:00 [haddock @ 2002-05-29 13:38:51 by simonmar] Document recent changes to markup syntax - - - - - f93641d6 by Simon Marlow at 2002-05-29T15:27:18+00:00 [haddock @ 2002-05-29 15:27:18 by simonmar] Include the instances in abstract data types too - - - - - 613f21e3 by Simon Marlow at 2002-06-03T13:05:58+00:00 [haddock @ 2002-06-03 13:05:57 by simonmar] Allow exporting of individual class methods and record selectors. For these we have to invent the correct type signature, which we do in the simplest possible way (i.e. no context reduction nonsense in the class case). - - - - - 14b36807 by Simon Marlow at 2002-06-03T13:20:00+00:00 [haddock @ 2002-06-03 13:20:00 by simonmar] Fix linking to qualified names again (thanks to Sven Panne for pointing out the bug). - - - - - 95b10eac by Simon Marlow at 2002-06-03T13:46:48+00:00 [haddock @ 2002-06-03 13:46:48 by simonmar] Fix for exporting record selectors from a newtype declaration - - - - - 272f932e by Simon Marlow at 2002-06-03T13:56:38+00:00 [haddock @ 2002-06-03 13:56:38 by simonmar] update to version 0.3 - - - - - 1c0a3bed by Simon Marlow at 2002-06-03T14:05:07+00:00 [haddock @ 2002-06-03 14:05:07 by simonmar] Add changes in version 0.3 - - - - - 145b4626 by Simon Marlow at 2002-06-03T14:12:38+00:00 [haddock @ 2002-06-03 14:12:38 by simonmar] Render class names as proper binders - - - - - 052106b3 by Simon Marlow at 2002-06-03T14:15:10+00:00 [haddock @ 2002-06-03 14:15:10 by simonmar] update, and separate into bugs, features, and cosmetic items. - - - - - 854f4914 by Simon Marlow at 2002-06-03T14:16:13+00:00 [haddock @ 2002-06-03 14:16:13 by simonmar] More test cases - - - - - 466922c8 by Simon Marlow at 2002-06-03T14:16:56+00:00 [haddock @ 2002-06-03 14:16:56 by simonmar] Example from the paper - - - - - 9962a045 by Simon Marlow at 2002-06-03T14:17:49+00:00 [haddock @ 2002-06-03 14:17:49 by simonmar] A debugging version of the style-sheet, which gives some tables coloured backgrounds so we can see what's going on. - - - - - f16b79db by Simon Marlow at 2002-06-03T14:19:46+00:00 [haddock @ 2002-06-03 14:19:46 by simonmar] typo - - - - - 620db27b by Simon Marlow at 2002-06-03T14:48:32+00:00 [haddock @ 2002-06-03 14:48:32 by simonmar] oops, fix markup bugs - - - - - 53fd105c by Simon Marlow at 2002-06-05T09:05:07+00:00 [haddock @ 2002-06-05 09:05:07 by simonmar] Keep foreign imports when there is no export list (bug reported by Sven Panne). - - - - - 6d98989c by Simon Marlow at 2002-06-05T09:12:02+00:00 [haddock @ 2002-06-05 09:12:02 by simonmar] Identifiers in single quotes can be symbol names too (bug reported by Hal Daume). - - - - - 001811e5 by Sven Panne at 2002-06-08T14:03:36+00:00 [haddock @ 2002-06-08 14:03:36 by panne] Tiny workaround for the fact that Haddock currently ignores HsImportSpecs: Let the local_orig_env take precedence. This is no real solution at all, but improves things sometimes, e.g. in my GLUT documentation. :-) - - - - - 504d19c9 by Simon Marlow at 2002-06-11T09:23:25+00:00 [haddock @ 2002-06-11 09:23:25 by simonmar] portability nit - - - - - e13b5af4 by Simon Marlow at 2002-06-20T12:38:07+00:00 [haddock @ 2002-06-20 12:38:07 by simonmar] Empty declaration fixes. - - - - - f467a9b6 by Simon Marlow at 2002-06-20T12:39:02+00:00 [haddock @ 2002-06-20 12:39:01 by simonmar] Add support for a "prologue" - a description for the whole library, placed on the contents page before the module list. - - - - - b8dbfe20 by Simon Marlow at 2002-06-21T12:43:06+00:00 [haddock @ 2002-06-21 12:43:06 by simonmar] When we have a single code block paragraph, don't place it in <pre>..</pre>, just use <tt>..</tt> to avoid generating extra vertical white space in some browsers. - - - - - 4831dbbd by Simon Marlow at 2002-06-21T15:50:42+00:00 [haddock @ 2002-06-21 15:50:42 by simonmar] Add support for reading and writing interface files(!) This turned out to be quite easy, and necessary to get decent hyperlinks between the documentation for separate packages in the libraries. The functionality isn't quite complete yet: for a given package of modules, you'd like to say "the HTML for these modules lives in directory <dir>" (currently they are assumed to be all in the same place). Two new flags: --dump-interface=FILE dump an interface file in FILE --read-interface=FILE read interface from FILE an interface file describes *all* the modules being processed. Only the exported names are kept in the interface: if you re-export a name from a module in another interface the signature won't be copied. This is a compromise to keep the size of the interfaces sensible. Also, I added another useful option: --no-implicit-prelude avoids trying to import the Prelude. Previously this was the default, but now importing the Prelude from elsewhere makes sense if you also read in an interface containing the Prelude module, so Haddock imports the Prelude implicitly according to the Haskell spec. - - - - - d3640a19 by Sven Panne at 2002-06-23T14:54:00+00:00 [haddock @ 2002-06-23 14:54:00 by panne] Make it compile with newer GHCs - - - - - 780c506b by Sven Panne at 2002-06-23T15:44:31+00:00 [haddock @ 2002-06-23 15:44:31 by panne] Cleaned up build root handling and added more docs - - - - - 45290d2e by Simon Marlow at 2002-06-24T14:37:43+00:00 [haddock @ 2002-06-24 14:37:42 by simonmar] When reading an interface, allow a file path offset to be specified which represents the path to the HTML files for the modules specified by that interface. The path may be either relative (to the location of the HTML for this package), or absolute. The syntax is --read-interface=PATH,FILE where PATH is the path to the HTML, and FILE is the filename containing the interface. - - - - - 4e2b9ae6 by Simon Marlow at 2002-07-03T16:01:08+00:00 [haddock @ 2002-07-03 16:01:07 by simonmar] Handle import specs properly, include 'hiding'. Haddock now has a complete implementation of the Haskell module system (more or less; I won't claim it's 100% correct). - - - - - 9a9aa1a8 by Simon Marlow at 2002-07-03T16:18:16+00:00 [haddock @ 2002-07-03 16:18:16 by simonmar] Update - - - - - 560c3026 by Simon Marlow at 2002-07-04T14:56:10+00:00 [haddock @ 2002-07-04 14:56:10 by simonmar] Clean up the code that constructs the exported declarations, and fix a couple of bugs along the way. Now if you import a class hiding one of the methods, then re-export the class, the version in the documentation will correctly have the appropriate method removed. - - - - - 2c26e77d by Simon Marlow at 2002-07-04T15:26:13+00:00 [haddock @ 2002-07-04 15:26:13 by simonmar] More bugfixes to the export handling - - - - - 03e0710d by Simon Marlow at 2002-07-09T10:12:10+00:00 [haddock @ 2002-07-09 10:12:10 by simonmar] Don't require that the list type comes from "Prelude" for it to be treated as special syntax (sometimes it comes from Data.List or maybe even GHC.Base). - - - - - 44f3891a by Simon Marlow at 2002-07-09T10:12:51+00:00 [haddock @ 2002-07-09 10:12:51 by simonmar] commented-out debugging code - - - - - 97280873 by Krasimir Angelov at 2002-07-09T16:33:33+00:00 [haddock @ 2002-07-09 16:33:31 by krasimir] 'Microsoft HTML Help' support - - - - - 3dc04655 by Simon Marlow at 2002-07-10T09:40:56+00:00 [haddock @ 2002-07-10 09:40:56 by simonmar] Fix for rendering of the (->) type constructor, from Ross Paterson. - - - - - c9f149c6 by Simon Marlow at 2002-07-10T10:26:11+00:00 [haddock @ 2002-07-10 10:26:11 by simonmar] Tweaks to the MS Help support: the extra files are now only generated if you ask for them (--ms-help). - - - - - e8acc1e6 by Simon Marlow at 2002-07-10T10:57:10+00:00 [haddock @ 2002-07-10 10:57:10 by simonmar] Document all the new options since 0.3 - - - - - 8bb85544 by Simon Marlow at 2002-07-10T10:58:31+00:00 [haddock @ 2002-07-10 10:58:31 by simonmar] Sort the options a bit - - - - - abc0dd59 by Simon Marlow at 2002-07-15T09:19:38+00:00 [haddock @ 2002-07-15 09:19:38 by simonmar] Fix a bug in mkExportItems when processing a module without an explicit export list. We were placing one copy of a declaration for each binder in the declaration, which for a data type would mean one copy of the whole declaration per constructor or record selector. - - - - - dde65bb9 by Simon Marlow at 2002-07-15T09:54:16+00:00 [haddock @ 2002-07-15 09:54:16 by simonmar] merge rev. 1.35 - - - - - bd7eb8c4 by Simon Marlow at 2002-07-15T10:14:31+00:00 [haddock @ 2002-07-15 10:14:30 by simonmar] Be a bit more liberal in the kind of commenting styles we allow, as suggested by Malcolm Wallace. Mostly this consists of allowing doc comments either side of a separator token. In an export list, a section heading is now allowed before the comma, as well as after it. eg. module M where ( T(..) -- * a section heading , f -- * another section heading , g ) In record fields, doc comments are allowed anywhere (previously a doc-next was allowed only after the comma, and a doc-before was allowed only before the comma). eg. data R = C { -- | describes 'f' f :: Int -- | describes 'g' , g :: Int } - - - - - 8f6dfe34 by Simon Marlow at 2002-07-15T10:21:56+00:00 [haddock @ 2002-07-15 10:21:56 by simonmar] Mention alternative commenting styles. - - - - - fc515bb7 by Simon Marlow at 2002-07-15T16:16:50+00:00 [haddock @ 2002-07-15 16:16:50 by simonmar] Allow multiple sections/subsections before and after a comma in the export list. Also at the same time I made the syntax a little stricter (multiple commas now aren't allowed between export specs). - - - - - 80a97e74 by Simon Marlow at 2002-07-19T09:13:10+00:00 [haddock @ 2002-07-19 09:13:10 by simonmar] Allow special id's ([], (), etc.) to be used in an import declaration. - - - - - a69d7378 by Simon Marlow at 2002-07-19T09:59:02+00:00 [haddock @ 2002-07-19 09:59:02 by simonmar] Allow special id's ([], (), etc.) to be used in an import declarations. - - - - - d205fa60 by Simon Marlow at 2002-07-19T10:00:16+00:00 [haddock @ 2002-07-19 10:00:16 by simonmar] Relax the restrictions which require doc comments to be followed by semi colons - in some cases this isn't necessary. Now you can write module M where { -- | some doc class C where {} } without needing to put a semicolon before the class declaration. - - - - - e9301e14 by Simon Marlow at 2002-07-23T08:24:09+00:00 [haddock @ 2002-07-23 08:24:09 by simonmar] A new TODO list item - - - - - e5d77586 by Simon Marlow at 2002-07-23T08:40:56+00:00 [haddock @ 2002-07-23 08:40:56 by simonmar] - update the acknowledgements - remove the paragraph that described how to use explicit layout with doc comments; it isn't relevant any more. - - - - - 78a94137 by Simon Marlow at 2002-07-23T08:43:02+00:00 [haddock @ 2002-07-23 08:43:02 by simonmar] more tests - - - - - 5c320927 by Simon Marlow at 2002-07-23T08:43:26+00:00 [haddock @ 2002-07-23 08:43:26 by simonmar] Updates for version 0.4 - - - - - 488e99ae by Simon Marlow at 2002-07-23T09:10:46+00:00 [haddock @ 2002-07-23 09:10:46 by simonmar] Fix the %changelog (rpm complained that it wasn't in the right order) - - - - - a77bb373 by Simon Marlow at 2002-07-23T09:12:38+00:00 [haddock @ 2002-07-23 09:12:38 by simonmar] Another item for the TODO list - - - - - f1ec1813 by Simon Marlow at 2002-07-23T10:18:46+00:00 [haddock @ 2002-07-23 10:18:46 by simonmar] Add a version banner when invoked with -v - - - - - 1d44cadf by Simon Marlow at 2002-07-24T09:28:19+00:00 [haddock @ 2002-07-24 09:28:19 by simonmar] Remove ^Ms - - - - - 4d8d5e94 by Simon Marlow at 2002-07-24T09:42:18+00:00 [haddock @ 2002-07-24 09:42:17 by simonmar] Patches to quieten ghc -Wall, from those nice folks at Galois. - - - - - d6edc43e by Simon Marlow at 2002-07-25T14:37:29+00:00 [haddock @ 2002-07-25 14:37:28 by simonmar] Patch to allow simple hyperlinking to an arbitrary location in another module's documentation, from Volker Stolz. Now in a doc comment: #foo# creates <a name="foo"></a> And you can use the form "M\#foo" to hyperlink to the label 'foo' in module 'M'. Note that the backslash is necessary for now. - - - - - b34d18fa by Simon Marlow at 2002-08-02T09:08:22+00:00 [haddock @ 2002-08-02 09:08:22 by simonmar] The <TT> and <PRE> environments seem to use a font that is a little too small in IE. Compensate. (suggestion from Daan Leijen). - - - - - 8106b086 by Simon Marlow at 2002-08-02T09:25:23+00:00 [haddock @ 2002-08-02 09:25:20 by simonmar] Remove <P>..</P> from around list items, to reduce excess whitespace between the items of bulleted and ordered lists. (Suggestion from Daan Leijen). - - - - - c1acff8f by Simon Marlow at 2002-08-05T09:03:49+00:00 [haddock @ 2002-08-05 09:03:49 by simonmar] update - - - - - f968661c by Simon Marlow at 2002-11-11T09:32:57+00:00 [haddock @ 2002-11-11 09:32:57 by simonmar] Fix cut-n-pasto - - - - - 12d02619 by Simon Marlow at 2002-11-13T09:49:46+00:00 [haddock @ 2002-11-13 09:49:46 by simonmar] Small bugfix in the --read-interface option parsing from Brett Letner. - - - - - 30e32d5e by Ross Paterson at 2003-01-16T15:07:57+00:00 [haddock @ 2003-01-16 15:07:57 by ross] Adjust for the new exception libraries (as well as the old ones). - - - - - 871f65df by Sven Panne at 2003-02-20T21:31:40+00:00 [haddock @ 2003-02-20 21:31:40 by panne] * Add varsyms and consyms to index * Exclude empty entries from index - - - - - bc42cc87 by Sven Panne at 2003-02-24T21:26:29+00:00 [haddock @ 2003-02-24 21:26:29 by panne] Don't convert a "newtype" to a single-constructor "data" for non-abstractly exported types, they are quite different regarding strictness/pattern matching. Now a "data" without any constructors is only emitted for an abstractly exported type, regardless if it is actually a "newtype" or a "data". - - - - - 0c2a1d99 by Sven Panne at 2003-03-08T19:02:38+00:00 [haddock @ 2003-03-08 19:02:38 by panne] Fixed some broken/redirected/canonicalized links found by a very picky link checker. - - - - - 25459269 by Sven Panne at 2003-03-09T21:13:43+00:00 [haddock @ 2003-03-09 21:13:43 by panne] Don't append a fragment to non-defining index entries, only documents with a defining occurrence have a name anchor. - - - - - 6be4db86 by Sven Panne at 2003-03-10T21:34:25+00:00 [haddock @ 2003-03-10 21:34:24 by panne] Escape fragments. This fixes e.g. links to operators. - - - - - eb12972c by Ross Paterson at 2003-04-25T10:50:06+00:00 [haddock @ 2003-04-25 10:50:05 by ross] An 80% solution to generating derived instances. A complete solution would duplicate the instance inference logic, but if a type variable occurs as a constructor argument, then we can just propagate the derived class to the variable. But we know nothing of the constraints on any type variables that occur elsewhere. For example, the declarations data Either a b = Left a | Right b deriving (Eq, Ord) data Ptr a = Ptr Addr# deriving (Eq, Ord) newtype IORef a = IORef (STRef RealWorld a) deriving Eq yield the instances (Eq a, Eq b) => Eq (Either a b) (Ord a, Ord b) => Ord (Either a b) Eq (Ptr a) Ord (Ptr a) (??? a) => Eq (IORef a) The last example shows the limits of this local analysis. Note that a type variable may be in both categories: then we know a constraint, but there may be more, or a stronger constraint, e.g. data Tree a = Node a [Tree a] deriving Eq yields (Eq a, ??? a) => Eq (Tree a) - - - - - de886f78 by Simon Marlow at 2003-04-25T11:17:55+00:00 [haddock @ 2003-04-25 11:17:55 by simonmar] Some updates, including moving the derived instance item down to the bottom of the list now that Ross has contributed some code that does the job for common cases. - - - - - 1b52cffd by Simon Marlow at 2003-04-30T14:02:32+00:00 [haddock @ 2003-04-30 14:02:32 by simonmar] When installing on Windows, run cygpath over $(HADDOCKLIB) so that haddock (a mingw program, built by GHC) can understand it. You still need to be in a cygwin environment to run Haddock, because of the shell script wrapper. - - - - - d4f638de by Simon Marlow at 2003-05-06T10:04:47+00:00 [haddock @ 2003-05-06 10:04:47 by simonmar] Catch another case of a paragraph containing just a DocMonospaced that should turn into a DocCodeBlock. - - - - - 4162b2b9 by Simon Marlow at 2003-05-06T10:11:44+00:00 [haddock @ 2003-05-06 10:11:44 by simonmar] Add some more code-block tests. - - - - - 4f5802c8 by Simon Marlow at 2003-05-06T10:14:52+00:00 [haddock @ 2003-05-06 10:14:52 by simonmar] Don't turn a single DocCodeBlock into a DocMonospaced, because that tends to remove the line breaks in the code. - - - - - ef8c45f7 by Simon Marlow at 2003-05-21T15:07:21+00:00 [haddock @ 2003-05-21 15:07:21 by simonmar] Only omit the module contents when there are no section headings at all. - - - - - bcee1e75 by Sigbjorn Finne at 2003-05-30T16:50:45+00:00 [haddock @ 2003-05-30 16:50:45 by sof] cygpath: for now, steer clear of --mixed - - - - - 30567af3 by Sigbjorn Finne at 2003-05-30T17:59:28+00:00 [haddock @ 2003-05-30 17:59:28 by sof] oops, drop test defn from prev commit - - - - - b0856e7d by Simon Marlow at 2003-06-03T09:55:26+00:00 [haddock @ 2003-06-03 09:55:26 by simonmar] Two small fixes to make the output valid HTML 4.01 (transitional). Thanks to Malcolm Wallace for pointing out the problems. - - - - - 70e137ea by Simon Marlow at 2003-07-28T13:30:35+00:00 [haddock @ 2003-07-28 13:30:35 by simonmar] Add tests for a couple of bugs. - - - - - 122bd578 by Simon Marlow at 2003-07-28T13:31:25+00:00 [haddock @ 2003-07-28 13:31:25 by simonmar] Add documentation for anchors. - - - - - 0bd27cb2 by Simon Marlow at 2003-07-28T13:31:46+00:00 [haddock @ 2003-07-28 13:31:46 by simonmar] Update - - - - - 08052d42 by Simon Marlow at 2003-07-28T13:32:12+00:00 [haddock @ 2003-07-28 13:32:12 by simonmar] layout tweak. - - - - - 13942749 by Simon Marlow at 2003-07-28T13:33:03+00:00 [haddock @ 2003-07-28 13:33:03 by simonmar] Differentiate links to types/classes from links to variables/constructors with a prefix ("t:" and "v:" respectively). - - - - - d7f493b9 by Simon Marlow at 2003-07-28T13:35:17+00:00 [haddock @ 2003-07-28 13:35:16 by simonmar] When a module A exports another module's contents via 'module B', then modules which import entities from B re-exported by A should link to B.foo rather than A.foo. See examples/Bug2.hs. - - - - - d94cf705 by Simon Marlow at 2003-07-28T13:36:14+00:00 [haddock @ 2003-07-28 13:36:14 by simonmar] Update to version 0.5 - - - - - dbb776cd by Sven Panne at 2003-07-28T14:02:43+00:00 [haddock @ 2003-07-28 14:02:43 by panne] * Updated to version 0.5 * Automagically generate configure if it is not there - - - - - 6cfeee53 by Simon Marlow at 2003-07-28T14:32:43+00:00 [haddock @ 2003-07-28 14:32:42 by simonmar] Update to avoid using hslibs with GHC >= 5.04 - - - - - a1ce838f by Simon Marlow at 2003-07-28T14:33:37+00:00 [haddock @ 2003-07-28 14:33:37 by simonmar] Update for 0.5 - - - - - c0fe6493 by Simon Marlow at 2003-07-28T14:53:22+00:00 [haddock @ 2003-07-28 14:53:22 by simonmar] Markup fix - - - - - 6ea31596 by Sven Panne at 2003-07-28T16:40:45+00:00 [haddock @ 2003-07-28 16:40:45 by panne] Make it compile with GHC >= 6.01 - - - - - afcd30fc by Simon Marlow at 2003-07-30T15:04:52+00:00 [haddock @ 2003-07-30 15:04:52 by simonmar] Pay attention to import specs when building the the import env, as well as the orig env. This may fix some wrong links in documentation when import specs are being used. - - - - - 17c3137f by Simon Marlow at 2003-07-30T16:05:41+00:00 [haddock @ 2003-07-30 16:05:40 by simonmar] Rename instances based on the import_env for the module in which they are to be displayed. This should give, in many cases, better links for the types and classes mentioned in the instance head. This involves keeping around the import_env in the iface until the end, because instances are not collected up until all the modules have been processed. Fortunately it doesn't seem to affect performance much. Instance heads are now attached to ExportDecls, rather than the HTML backend passing around a separate mapping for instances. This is a cleanup. - - - - - 3d3b5c87 by Sven Panne at 2003-08-04T10:18:24+00:00 [haddock @ 2003-08-04 10:18:24 by panne] Don't print parentheses around one-element contexts - - - - - 9e3f3f2d by Simon Marlow at 2003-08-04T12:59:47+00:00 [haddock @ 2003-08-04 12:59:47 by simonmar] A couple of TODOs. - - - - - e9d8085c by Simon Marlow at 2003-08-05T14:10:31+00:00 [haddock @ 2003-08-05 14:10:31 by simonmar] I'm not sure why, but it seems that the index entries for non-defining occurrences of entities did not have an anchor - the link just pointed to the module. This fixes it. - - - - - ff5c7d6d by Simon Marlow at 2003-08-15T14:42:59+00:00 [haddock @ 2003-08-15 14:42:59 by simonmar] Convert the lexer to Alex, and fix a bug in the process. - - - - - 1aa077bf by Simon Marlow at 2003-08-15T15:00:18+00:00 [haddock @ 2003-08-15 15:00:18 by simonmar] Update - - - - - d3de1e38 by Simon Marlow at 2003-08-15T15:01:03+00:00 [haddock @ 2003-08-15 15:01:03 by simonmar] wibbles - - - - - b40ece3b by Simon Marlow at 2003-08-18T10:04:47+00:00 [haddock @ 2003-08-18 10:04:47 by simonmar] Lex the 'mdo' keyword as 'do'. - - - - - 8f9a1146 by Simon Marlow at 2003-08-18T11:48:24+00:00 [haddock @ 2003-08-18 11:48:24 by simonmar] Two bugs from Sven. - - - - - ea54ebc0 by Simon Marlow at 2003-08-18T11:48:46+00:00 [haddock @ 2003-08-18 11:48:46 by simonmar] Fixes to the new lexer. - - - - - d5f6a4b5 by Simon Marlow at 2003-08-19T09:09:03+00:00 [haddock @ 2003-08-19 09:09:03 by simonmar] Further wibbles to the syntax. - - - - - 6bbdadb7 by Sven Panne at 2003-08-26T18:45:35+00:00 [haddock @ 2003-08-26 18:45:35 by panne] Use autoreconf instead of autoconf - - - - - 32e889cb by Sven Panne at 2003-08-26T19:01:19+00:00 [haddock @ 2003-08-26 19:01:18 by panne] Made option handling a bit more consistent with other tools, in particular: Every program in fptools should output * version info on stdout and terminate successfully when -V or --version * usage info on stdout and terminate successfully when -? or --help * usage info on stderr and terminate unsuccessfully when an unknown option is given. - - - - - 5d156a91 by Sven Panne at 2003-08-26T19:20:55+00:00 [haddock @ 2003-08-26 19:20:55 by panne] Make it *very* clear that we terminate when given a -V/--version flag - - - - - e6577265 by Sven Panne at 2003-08-27T07:50:03+00:00 [haddock @ 2003-08-27 07:50:02 by panne] * Made -D a short option for --dump-interface. * Made -m a short option for --ms-help. * Made -n a short option for --no-implicit-prelude. * Made -c a short option for --css. * Removed DocBook options from executable (they didn't do anything), but mark them as reserved in the docs. Note that the short option for DocBook output is now -S (from SGML) instead of -d. The latter is now a short option for --debug. * The order of the Options in the documentation now matches the order printed by Haddock itself. Note: Although changing the names of options is often a bad idea, I'd really like to make the options for the programs in fptools more consistent and compatible to the ones used in common GNU programs. - - - - - d303ff98 by Simon Marlow at 2003-09-10T08:23:48+00:00 [haddock @ 2003-09-10 08:23:48 by simonmar] Add doc subdir. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 9a70e46a by Simon Marlow at 2003-09-10T08:24:32+00:00 [haddock @ 2003-09-10 08:24:32 by simonmar] Install these files in $(datadir), not $(libdir), since they're architecture independent. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - bbb87e7a by Simon Marlow at 2003-09-10T08:25:31+00:00 [haddock @ 2003-09-10 08:25:31 by simonmar] Haddock's supplementary HTML bits now live in $(datadir), not $(libdir). Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 3587c24b by Simon Marlow at 2003-09-22T10:34:38+00:00 [haddock @ 2003-09-22 10:34:38 by simonmar] Allow installing of docs. - - - - - d510b517 by Sven Panne at 2003-10-11T08:10:44+00:00 [haddock @ 2003-10-11 08:10:44 by panne] Include architecture-independent files in file list - - - - - 187d7618 by Sigbjorn Finne at 2003-10-20T17:19:24+00:00 [haddock @ 2003-10-20 17:19:22 by sof] support for i-parameters + zip comprehensions - - - - - b6c7a273 by Simon Marlow at 2003-11-03T14:24:24+00:00 [haddock @ 2003-11-03 14:24:24 by simonmar] Update TODO file. - - - - - 58513e33 by Simon Marlow at 2003-11-05T11:22:04+00:00 [haddock @ 2003-11-05 11:22:04 by simonmar] Remove the last of the uses of 'trace' to emit warnings, and tidy up a couple of places where duplicate warnings were being emitted. - - - - - 33a78846 by Simon Marlow at 2003-11-05T11:30:53+00:00 [haddock @ 2003-11-05 11:30:52 by simonmar] - Suppress warnings about unknown imported modules by default. - Add a -v/--verbose flag to re-enable these warnings. The general idea is to suppress the "Warning: unknown module: Prelude" warnings which most Haddock users will see every time, and which aren't terribly useful. - - - - - a969de7f by Simon Marlow at 2003-11-05T12:30:28+00:00 [haddock @ 2003-11-05 12:30:28 by simonmar] - Remove the emboldening of index entries for defining locations. This isn't useful, and breaks abstractions. - If an entity is re-exported by a module but the module doesn't include documentation for that entity (perhaps because it is re-exported by 'module M'), then don't attempt to hyperlink to the documentation from the index. Instead, just list that module in the index, to indicate that the entity is exported from there. - - - - - f14ea82a by Simon Marlow at 2003-11-05T15:15:59+00:00 [haddock @ 2003-11-05 15:15:59 by simonmar] Index overhaul: - no more separate type/class and variable/function indices - the index now makes a distinction between different entities with the same name. One example is a type constructor with the same name as a data constructor, but another example is simply a function with the same name exported by two different modules. For example, the index entry for 'catch' now looks like this: catch 1 (Function) Control.Exception 2 (Function) GHC.Exception, Prelude, System.IO, System.IO.Error making it clear that there are two different 'catch'es, but one of them is exported by several modules. - Each index page now has the index contents (A B C ...) at the top. Please let me know if you really hate any of this. - - - - - 01a25ca6 by Simon Marlow at 2003-11-05T15:16:38+00:00 [haddock @ 2003-11-05 15:16:38 by simonmar] Update - - - - - 1a7ccb86 by Simon Marlow at 2003-11-05T17:16:05+00:00 [haddock @ 2003-11-05 17:16:04 by simonmar] Support for generating a single unified index for several packages. --use-index=URL turns off normal index generation, causes Index links to point to URL. --gen-index generates an combined index from the specified interfaces. Currently doesn't work exactly right, because the interfaces don't contain the iface_reexported info. I'll need to fix that up. - - - - - a2bca16d by Simon Marlow at 2003-11-06T10:44:52+00:00 [haddock @ 2003-11-06 10:44:52 by simonmar] Include iface_reexported in the .haddock file. This unfortunately bloats the file (40% for base). If this gets to be a problem we can always apply the dictionary trick that GHC uses for squashing .hi files. - - - - - 0a09c293 by Simon Marlow at 2003-11-06T12:39:47+00:00 [haddock @ 2003-11-06 12:39:46 by simonmar] - Add definition lists, marked up like this: -- | This is a definition list: -- -- [@foo@] The description of @foo at . -- -- [@bar@] The description of @bar at . Cunningly, the [] characters are not treated specially unless a [ is found at the beginning of a paragraph, in which case the ] becomes special in the following text. - Add --use-contents and --gen-contents, along the lines of --use-index and --gen-index added yesterday. Now we can generate a combined index and contents for the whole of the hierarchical libraries, and in theory the index/contents on the system could be updated as new packages are added. - - - - - fe1b3460 by Simon Marlow at 2003-11-06T14:47:36+00:00 [haddock @ 2003-11-06 14:47:36 by simonmar] Remove the 'Parent' button - it is of dubious use, and often points into thin air. - - - - - db6d762f by Simon Marlow at 2003-11-06T16:48:14+00:00 [haddock @ 2003-11-06 16:48:11 by simonmar] - Include the OptHide setting in the interface, so we don't include hidden modules in the combined index/contents. - Add a -k/--package flag to set the package name for the current set of modules. The package name for each module is now shown in the right-hand column of the contents, in a combined contents page. - - - - - 7d71718b by Simon Marlow at 2003-11-06T16:50:28+00:00 [haddock @ 2003-11-06 16:50:28 by simonmar] Add -k/--package docs - - - - - ef43949d by Simon Marlow at 2003-11-06T16:51:23+00:00 [haddock @ 2003-11-06 16:51:23 by simonmar] Bump to 0.6 - - - - - 1c419e06 by Simon Marlow at 2003-11-06T16:51:50+00:00 [haddock @ 2003-11-06 16:51:50 by simonmar] update - - - - - 69422327 by Simon Marlow at 2003-11-10T14:41:06+00:00 [haddock @ 2003-11-10 14:41:05 by simonmar] Re-exporting names from a different package is problematic, because we don't have access to the full documentation for the entity. Currently Haddock just ignores entities with no documentation, but this results in bogus-looking empty documentation for many of the modules in the haskell98 package. So: - the documentation will now just list the name, as a link pointing to the location of the actual documentation. - now we don't attempt to link to these re-exported entities if they are referred to by the current module. Additionally: - If there is no documentation in the current module, include just the Synopsis section (rather than just the documentation section, as it was before). This just looks nicer and was on the TODO list. - - - - - 3c3fc433 by Simon Marlow at 2003-11-10T14:51:59+00:00 [haddock @ 2003-11-10 14:51:59 by simonmar] Fix for getReExports: take into account names which are not visible because they are re-exported from a different package. - - - - - 31c8437b by Simon Marlow at 2003-11-10T15:10:53+00:00 [haddock @ 2003-11-10 15:10:53 by simonmar] Version 0.6 changes - - - - - a7c2430b by Simon Marlow at 2003-11-10T15:15:58+00:00 [haddock @ 2003-11-10 15:15:58 by simonmar] getReExports: one error case that isn't - - - - - 00cc459c by Simon Marlow at 2003-11-10T16:15:19+00:00 [haddock @ 2003-11-10 16:15:18 by simonmar] copyright update - - - - - ca62408d by Simon Marlow at 2003-11-11T09:57:25+00:00 [haddock @ 2003-11-11 09:57:25 by simonmar] Version 0.6 - - - - - 3acbf818 by Simon Marlow at 2003-11-11T12:10:44+00:00 [haddock @ 2003-11-11 12:10:44 by simonmar] Go back to producing just the documentation section, rather than just the synopsis section, for a module with no documentation annotations. One reason is that the synopsis section tries to link each entity to its documentation on the same page. Also, the doc section anchors each entity, and it lists instances which the synopsis doesn't. - - - - - 6c90abc2 by Simon Marlow at 2003-11-12T10:03:39+00:00 [haddock @ 2003-11-12 10:03:39 by simonmar] 2002 -> 2003 - - - - - 090bbc4c by Simon Marlow at 2003-11-28T12:08:00+00:00 [haddock @ 2003-11-28 12:08:00 by simonmar] update - - - - - 8096a832 by Simon Marlow at 2003-11-28T12:09:58+00:00 [haddock @ 2003-11-28 12:09:58 by simonmar] Fix some of the problems with Haddock generating pages that are too wide. Now we only specify 'nowrap' when it is necessary to avoid a code box getting squashed up by the text to the right of it. - - - - - 35294929 by Sven Panne at 2003-12-29T17:16:31+00:00 [haddock @ 2003-12-29 17:16:31 by panne] Updated my email address - - - - - cdb697bf by Simon Marlow at 2004-01-08T10:14:24+00:00 [haddock @ 2004-01-08 10:14:24 by simonmar] Add instructions for using GHC to pre-process source for feeding to Haddock. - - - - - 8dfc491f by Simon Marlow at 2004-01-09T12:45:46+00:00 [haddock @ 2004-01-09 12:45:46 by simonmar] Add -optP-P to example ghc command line. - - - - - ac41b820 by Simon Marlow at 2004-02-03T11:02:03+00:00 [haddock @ 2004-02-03 11:02:03 by simonmar] Fix bug in index generation - - - - - f4e7edcb by Simon Marlow at 2004-02-10T11:51:16+00:00 [haddock @ 2004-02-10 11:51:16 by simonmar] Don't throw away whitespace at the beginning of a line (experimental fix). - - - - - 68e212d2 by Simon Marlow at 2004-02-10T12:10:08+00:00 [haddock @ 2004-02-10 12:10:08 by simonmar] Fix for previous commit: I now realise why the whitespace was stripped from the beginning of the line. Work around it. - - - - - e7d7f2df by Sven Panne at 2004-02-10T18:38:45+00:00 [haddock @ 2004-02-10 18:38:45 by panne] Make Haddock link with the latest relocated monad transformer package - - - - - 992d4225 by Simon Marlow at 2004-02-16T10:21:35+00:00 [haddock @ 2004-02-16 10:21:35 by simonmar] Add a TODO - - - - - 1ac55326 by Simon Marlow at 2004-03-12T11:33:39+00:00 [haddock @ 2004-03-12 11:33:39 by simonmar] Add an item. - - - - - 0478e903 by Simon Marlow at 2004-03-15T12:24:05+00:00 [haddock @ 2004-03-15 12:24:05 by simonmar] Add an item. - - - - - 6f26d21a by Simon Marlow at 2004-03-18T14:21:29+00:00 [haddock @ 2004-03-18 14:21:29 by simonmar] Fix URL - - - - - 19b6bb99 by Simon Marlow at 2004-03-22T14:09:03+00:00 [haddock @ 2004-03-22 14:09:03 by simonmar] getReExports was bogus: we should really look in the import_env to find the documentation for an entity which we are re-exporting without documentation. Suggested by: Ross Paterson (patch modified by me). - - - - - 5c756031 by Simon Marlow at 2004-03-24T09:42:11+00:00 [haddock @ 2004-03-24 09:42:10 by simonmar] hiding bug from Ross Paterson (fixed in rev 1.59 of Main.hs) - - - - - 1b692e6c by Simon Marlow at 2004-03-24T10:10:50+00:00 [haddock @ 2004-03-24 10:10:50 by simonmar] mkExportItems fix & simplification: we should be looking at the actual exported names (calculated earlier) to figure out which subordinates of a declaration are exported. This means that if you export a record, and name its fields separately in the export list, the fields will still be visible in the documentation for the constructor. - - - - - 90e5e294 by Simon Marlow at 2004-03-24T10:12:08+00:00 [haddock @ 2004-03-24 10:12:08 by simonmar] Make restrictCons take into account record field names too (removing a ToDo). - - - - - 2600efa4 by Simon Marlow at 2004-03-24T10:16:17+00:00 [haddock @ 2004-03-24 10:16:17 by simonmar] Record export tests. - - - - - 6a8575c7 by Simon Marlow at 2004-03-25T09:35:14+00:00 [haddock @ 2004-03-25 09:35:14 by simonmar] restrictTo: fix for restricting a newtype with a record field. - - - - - dcf55a8d by Simon Marlow at 2004-03-25T10:01:42+00:00 [haddock @ 2004-03-25 10:01:42 by simonmar] Fix duplicate instance bug - - - - - f49aa758 by Simon Marlow at 2004-03-25T10:02:41+00:00 [haddock @ 2004-03-25 10:02:41 by simonmar] Duplicate instance bug. - - - - - 7b87344c by Simon Marlow at 2004-03-25T10:29:56+00:00 [haddock @ 2004-03-25 10:29:56 by simonmar] If a name is imported from two places, one hidden and one not, choose the unhidden one to link to. Also, when there's only a hidden module to link to, don't try linking to it. - - - - - 40f44d7b by Simon Marlow at 2004-03-25T15:17:24+00:00 [haddock @ 2004-03-25 15:17:23 by simonmar] Add support for collaspible parts of the page, with a +/- button and a bit of JavaScript. Make the instances collapsible, and collapse them by default. This makes documentation with long lists of instances (eg. the Prelude) much easier to read. Maybe we should give other documentation sections the same treatment. - - - - - 9b64dc0f by Simon Marlow at 2004-03-25T15:20:55+00:00 [haddock @ 2004-03-25 15:20:55 by simonmar] Update - - - - - c2fff7f2 by Simon Marlow at 2004-03-25T15:45:10+00:00 [haddock @ 2004-03-25 15:45:10 by simonmar] Eliminate some unnecessary spaces in the HTML rendering - - - - - b7948ff0 by Simon Marlow at 2004-03-25T16:00:37+00:00 [haddock @ 2004-03-25 16:00:36 by simonmar] Remove all that indentation in the generated HTML to keep the file sizes down. - - - - - da2bb4ca by Sven Panne at 2004-03-27T09:57:58+00:00 [haddock @ 2004-03-27 09:57:57 by panne] Added the new-born haddock.js to the build process and the documentation. - - - - - b99e6f8c by Sven Panne at 2004-03-27T10:32:20+00:00 [haddock @ 2004-03-27 10:32:20 by panne] "type" is a required attribute of the "script" element - - - - - 562b185a by Sven Panne at 2004-03-27T12:52:34+00:00 [haddock @ 2004-03-27 12:52:34 by panne] Add a doctype for the contents page, too. - - - - - f6a99c2d by Simon Marlow at 2004-04-14T10:03:25+00:00 [haddock @ 2004-04-14 10:03:25 by simonmar] fix for single-line comment syntax - - - - - de366303 by Simon Marlow at 2004-04-20T13:08:04+00:00 [haddock @ 2004-04-20 13:08:04 by simonmar] Allow a 'type' declaration to include documentation comments. These will be ignored by Haddock, but at least one user (Johannes Waldmann) finds this feature useful, and it's easy to add. - - - - - fd78f51e by Simon Marlow at 2004-05-07T15:14:56+00:00 [haddock @ 2004-05-07 15:14:56 by simonmar] - update copyright - add version to abstract - - - - - 59f53e32 by Sven Panne at 2004-05-09T14:39:53+00:00 [haddock @ 2004-05-09 14:39:53 by panne] Fix the fix for single-line comment syntax, ------------------------------------------- is now a valid comment line again. - - - - - 8b18f2fe by Simon Marlow at 2004-05-10T10:11:51+00:00 [haddock @ 2004-05-10 10:11:51 by simonmar] Update - - - - - 225a491d by Ross Paterson at 2004-05-19T13:10:23+00:00 [haddock @ 2004-05-19 13:10:23 by ross] Make the handling of "deriving" slightly smarter, by ignoring data constructor arguments that are identical to the lhs. Now handles things like data Tree a = Leaf a | Branch (Tree a) (Tree a) deriving ... - - - - - 37588686 by Mike Thomas at 2004-05-21T06:38:14+00:00 [haddock @ 2004-05-21 06:38:14 by mthomas] Windows exe extensions (bin remains for Unix). - - - - - cf2b9152 by Simon Marlow at 2004-05-25T09:34:54+00:00 [haddock @ 2004-05-25 09:34:54 by simonmar] Add some TODO items - - - - - 4d29cdfc by Simon Marlow at 2004-05-25T10:41:46+00:00 [haddock @ 2004-05-25 10:41:46 by simonmar] Complain if -h is used with --gen-index or --gen-contents, because it'll overwrite the new index/contents. - - - - - 2e0771e0 by Mike Thomas at 2004-05-28T20:17:55+00:00 [haddock @ 2004-05-28 20:17:55 by mthomas] Windows: search for templates in executable directory. Unix: Haddock tries cwd first rather than error if no -l arg. - - - - - 8d10bde1 by Sven Panne at 2004-06-05T16:53:34+00:00 [haddock @ 2004-06-05 16:53:34 by panne] Misc. rpm spec file cleanup, including: * make BuildRoot handling more consistent * added default file attributes * consistent defines and tags - - - - - 59974349 by Sven Panne at 2004-06-05T18:01:00+00:00 [haddock @ 2004-06-05 18:01:00 by panne] More rpm spec file cleanup, including: * added some BuildRequires * changed packager to me, so people can complain at the right place :-] * consistently refer to haskell.org instead of www.haskell.org - - - - - b94d4903 by Simon Marlow at 2004-07-01T11:08:58+00:00 [haddock @ 2004-07-01 11:08:57 by simonmar] Update to the +/- buttons: use a resized image rather than a <button>. Still seeing some strange effects in Konqueror, so might need to use a fixed-size image instead. - - - - - d5278f67 by Sven Panne at 2004-07-04T15:15:55+00:00 [haddock @ 2004-07-04 15:15:55 by panne] Install pictures for +/- pictures, too (JPEG is a strange format for graphics like this, I would have expected GIF or PNG here.) Things look fine with Konqueror and Netscape on Linux now, the only downside is that the cursor doesn't change when positioned above the "button". - - - - - 46dec6c5 by Sven Panne at 2004-07-13T17:59:28+00:00 [haddock @ 2004-07-13 17:59:28 by panne] A quote is a valid part of a Haskell identifier, but it would interfere with an ECMA script string delimiter, so escape it there. - - - - - 1d7bc432 by Simon Marlow at 2004-07-22T08:54:06+00:00 [haddock @ 2004-07-22 08:54:06 by simonmar] Add single quote to $ident, so you can say eg. 'foldl'' to refer to foldl' (the longest match rule is our friend). Bug reported by Adrian Hey <ahey at iee.org> - - - - - f183618b by Krasimir Angelov at 2004-07-27T22:59:35+00:00 [haddock @ 2004-07-27 22:58:23 by krasimir] Add basic support for Microsoft HTML Help 2.0 - - - - - d515d0c2 by Krasimir Angelov at 2004-07-27T23:02:36+00:00 [haddock @ 2004-07-27 23:02:36 by krasimir] escape names in the index - - - - - a5f1be23 by Krasimir Angelov at 2004-07-27T23:05:21+00:00 [haddock @ 2004-07-27 23:05:21 by krasimir] Add jsFile, plusFile and minusFile to the file list - - - - - c4fb4881 by Krasimir Angelov at 2004-07-28T22:12:10+00:00 [haddock @ 2004-07-28 22:12:09 by krasimir] bugfix. Move contentsHtmlFile, indexHtmlFile and subIndexHtmlFile functions to HaddockUtil.hs module to make them accessible from HaddockHH2.hs - - - - - 64d30b1d by Krasimir Angelov at 2004-07-30T22:15:47+00:00 [haddock @ 2004-07-30 22:15:45 by krasimir] more stuffs - support for separated compilation of packages - the contents page now uses DHTML TreeView - fixed copyFile bug - - - - - 133c8c5c by Krasimir Angelov at 2004-07-31T12:04:38+00:00 [haddock @ 2004-07-31 12:04:37 by krasimir] make the DHtmlTree in contents page more portable. The +/- buttons are replaced with new images which looks more beatiful. - - - - - 79040963 by Krasimir Angelov at 2004-07-31T13:10:20+00:00 [haddock @ 2004-07-31 13:10:20 by krasimir] Make DHtmlTree compatible with Mozila browser - - - - - 1a55dc90 by Krasimir Angelov at 2004-07-31T14:52:55+00:00 [haddock @ 2004-07-31 14:52:55 by krasimir] fix - - - - - 85ce0237 by Krasimir Angelov at 2004-07-31T14:53:28+00:00 [haddock @ 2004-07-31 14:53:28 by krasimir] HtmlHelp 1.x - - - - - 3c0c53ba by Krasimir Angelov at 2004-07-31T20:35:21+00:00 [haddock @ 2004-07-31 20:35:21 by krasimir] Added support for DevHelp - - - - - d42b5af1 by Krasimir Angelov at 2004-07-31T21:17:51+00:00 [haddock @ 2004-07-31 21:17:51 by krasimir] Document new features in HtmlHelp - - - - - 790fe21e by Krasimir Angelov at 2004-08-01T15:14:02+00:00 [haddock @ 2004-08-01 15:14:02 by krasimir] add missing imports - - - - - fd7cc6bc by Krasimir Angelov at 2004-08-01T19:52:08+00:00 [haddock @ 2004-08-01 19:52:06 by krasimir] fix some bugs. Now I have got the entire libraries documentation in HtmlHelp 2.0 format. - - - - - 94ad7ac8 by Krasimir Angelov at 2004-08-01T19:53:50+00:00 [haddock @ 2004-08-01 19:53:50 by krasimir] I forgot to add the new +/- images - - - - - f0c65388 by Krasimir Angelov at 2004-08-02T16:25:53+00:00 [haddock @ 2004-08-02 16:25:53 by krasimir] Add root node to the table of contents. All modules in tree are not children of the root - - - - - f50bd85d by Sven Panne at 2004-08-02T18:17:46+00:00 [haddock @ 2004-08-02 18:17:46 by panne] Mainly DocBook fixes - - - - - 09527ce3 by Sven Panne at 2004-08-02T20:02:29+00:00 [haddock @ 2004-08-02 20:02:29 by panne] Fixed -o/--odir handling. Generating the output, especially the directory handling, is getting a bit convoluted nowadays... - - - - - c8fbacfa by Sven Panne at 2004-08-02T20:31:13+00:00 [haddock @ 2004-08-02 20:31:13 by panne] Warning police - - - - - 37830bff by Sven Panne at 2004-08-02T20:32:29+00:00 [haddock @ 2004-08-02 20:32:28 by panne] Nuked dead code - - - - - 13847171 by Sven Panne at 2004-08-02T21:12:27+00:00 [haddock @ 2004-08-02 21:12:25 by panne] Use pathJoin instead of low-level list-based manipulation for FilePaths - - - - - c711d61e by Sven Panne at 2004-08-02T21:16:02+00:00 [haddock @ 2004-08-02 21:16:02 by panne] Removed WinDoze CRs - - - - - b1f7dc88 by Sven Panne at 2004-08-03T19:35:59+00:00 [haddock @ 2004-08-03 19:35:59 by panne] Fixed spelling of "http-equiv" attribute - - - - - dd5f394e by Sven Panne at 2004-08-03T19:44:03+00:00 [haddock @ 2004-08-03 19:44:03 by panne] Pacify W3C validator: * Added document encoding (currently UTF-8, not sure if this is completely correct) * Fixed syntax of `id' attributes * Added necessary `alt' attribute for +/- images Small layout improvement: * Added space after +/- images (still not perfect, but better than before) - - - - - 919c47c6 by Sigbjorn Finne at 2004-08-03T19:45:11+00:00 [haddock @ 2004-08-03 19:45:11 by sof] make it compile with <= ghc-6.1 - - - - - 4d6f01d8 by Sigbjorn Finne at 2004-08-03T19:45:30+00:00 [haddock @ 2004-08-03 19:45:30 by sof] ffi wibble - - - - - 4770643a by Sven Panne at 2004-08-03T20:47:46+00:00 [haddock @ 2004-08-03 20:47:46 by panne] Fixed CSS for button style. Note that only "0" is a valid measure without a unit! - - - - - 14aaf2e5 by Sven Panne at 2004-08-03T21:07:59+00:00 [haddock @ 2004-08-03 21:07:58 by panne] Improved spacing of dynamic module tree - - - - - 97c3579a by Simon Marlow at 2004-08-09T11:03:04+00:00 [haddock @ 2004-08-09 11:03:04 by simonmar] Add FormatVersion Patch submitted by: George Russell <ger at informatik.uni-bremen.de> - - - - - af7f8c03 by Simon Marlow at 2004-08-09T11:55:07+00:00 [haddock @ 2004-08-09 11:55:05 by simonmar] Add support for a short description for each module, which is included in the contents. The short description should be given in a "Description: " field of the header. Included in this patch are changes that make the format of the header a little more flexible. From the comments: -- all fields in the header are optional and have the form -- -- [spaces1][field name][spaces] ":" -- [text]"\n" ([spaces2][space][text]"\n" | [spaces]"\n")* -- where each [spaces2] should have [spaces1] as a prefix. -- -- Thus for the key "Description", -- -- > Description : this is a -- > rather long -- > -- > description -- > -- > The module comment starts here -- -- the value will be "this is a .. description" and the rest will begin -- at "The module comment". The header fields must be in the following order: Module, Description, Copyright, License, Maintainer, Stability, Portability. Patches submitted by: George Russell <ger at informatik.uni-bremen.de>, with a few small changes be me, mostly to merge with other recent changes. ToDo: document the module header. - - - - - 7b865ad3 by Simon Marlow at 2004-08-10T14:09:57+00:00 [haddock @ 2004-08-10 14:09:57 by simonmar] Fixes for DevHelp/HtmlHelp following introduction of short module description. - - - - - 814766cd by Simon Marlow at 2004-08-10T14:33:46+00:00 [haddock @ 2004-08-10 14:33:45 by simonmar] Fixes to installation under Windows. - - - - - 39cf9ede by Simon Marlow at 2004-08-12T12:08:23+00:00 [haddock @ 2004-08-12 12:08:23 by simonmar] Avoid using string-gap tricks. - - - - - b6d78551 by Simon Marlow at 2004-08-13T10:53:21+00:00 [haddock @ 2004-08-13 10:53:21 by simonmar] Update - - - - - eaae7417 by Simon Marlow at 2004-08-13T10:53:50+00:00 [haddock @ 2004-08-13 10:53:50 by simonmar] Test for primes in quoted links - - - - - 68c34f06 by Sven Panne at 2004-08-16T19:59:38+00:00 [haddock @ 2004-08-16 19:59:36 by panne] XMLification - - - - - 7f45a6f9 by Sven Panne at 2004-08-18T16:42:54+00:00 [haddock @ 2004-08-18 16:42:54 by panne] Re-added indices + minor fixes - - - - - 8a5dd97c by Sigbjorn Finne at 2004-08-25T17:15:42+00:00 [haddock @ 2004-08-25 17:15:42 by sof] backquote HADDOCK_VERSION defn for <= ghc-6.0.x; believe this is only needed under mingw - - - - - 4b1b42ea by Sven Panne at 2004-08-26T20:08:50+00:00 [haddock @ 2004-08-26 20:08:49 by panne] SGML is dead, long live DocBook XML! Note: The BuildRequires tags in the spec files are still incomplete and the documentation about the DocBook tools needs to be updated, too. Stay tuned... - - - - - 8d52cedb by Sven Panne at 2004-08-26T21:03:19+00:00 [haddock @ 2004-08-26 21:03:19 by panne] Updated BuildRequires tags. Alas, there seems to be no real standard here, so your mileage may vary... At least the current specs should work on SuSE Linux. - - - - - e6982912 by Sigbjorn Finne at 2004-08-30T15:44:59+00:00 [haddock @ 2004-08-30 15:44:59 by sof] escape HADDOCK_VERSION double quotes on all platforms when compiling with <=6.0.x - - - - - b3fbc867 by Simon Marlow at 2004-08-31T13:09:42+00:00 [haddock @ 2004-08-31 13:09:42 by simonmar] Avoid GHC/shell versionitis and create Version.hs - - - - - c359e16a by Sven Panne at 2004-09-05T19:12:33+00:00 [haddock @ 2004-09-05 19:12:32 by panne] * HTML documentation for "foo.xml" goes into directory "foo" again, not "foo-html". This is nicer and consistent with the behaviour for building the docs from SGML. * Disabled building PostScript documentation in the spec files for now, there are some strange issues with the FO->PS conversion for some files which have to be clarified first. - - - - - c68b1eba by Sven Panne at 2004-09-24T07:04:38+00:00 [haddock @ 2004-09-24 07:04:38 by panne] Switched the default state for instances and the module hierarchy to non-collapsed. This can be reversed when we finally use cookies from JavaScript to have a more persistent state. Previously going back and forth in the documentation was simply too annoying because everything was collapsed again and therefore the documentation was not easily navigatable. - - - - - dfb32615 by Simon Marlow at 2004-09-30T08:21:29+00:00 [haddock @ 2004-09-30 08:21:29 by simonmar] Add a feature request - - - - - 45ff783c by Sven Panne at 2004-10-23T19:54:00+00:00 [haddock @ 2004-10-23 19:54:00 by panne] Improved the Cygwin/MinGW chaos a little bit. There is still confusion about host platform vs. target platform... - - - - - 5f644714 by Krasimir Angelov at 2004-10-28T16:01:51+00:00 [haddock @ 2004-10-28 16:01:51 by krasimir] update for ghc-6.3+ - - - - - 92d9753e by Sven Panne at 2004-11-01T16:39:01+00:00 [haddock @ 2004-11-01 16:39:01 by panne] Revert previous commit: It's Network.URI which should be changed, not Haddock. - - - - - 05f70f6e by Simon Marlow at 2005-01-04T16:15:51+00:00 [haddock @ 2005-01-04 16:15:51 by simonmar] parser fix: allow qualified specialids. - - - - - 47870837 by Simon Marlow at 2005-01-04T16:16:54+00:00 [haddock @ 2005-01-04 16:16:54 by simonmar] Add a test - - - - - ff11fc2c by Ross Paterson at 2005-01-10T19:18:22+00:00 [haddock @ 2005-01-10 19:18:22 by ross] Render non-ASCII characters using numeric character references, to simplify charset issues. There's a META tag saying the charset is UTF-8, but GHC outputs characters as raw bytes. Ideally we need an encoding on the input side too, primarily in comments, because source files containing non-ASCII characters aren't portable between locales. - - - - - eba2fc4e by Simon Marlow at 2005-01-11T10:44:37+00:00 [haddock @ 2005-01-11 10:44:37 by simonmar] Remove string gap - - - - - b899a381 by Ross Paterson at 2005-01-13T11:41:33+00:00 [haddock @ 2005-01-13 11:41:33 by ross] recognize SGML-style numeric character references &#ddd; or &#xhhhh; and translate them into Chars. - - - - - 106e3cf0 by Ross Paterson at 2005-01-13T14:43:41+00:00 [haddock @ 2005-01-13 14:43:41 by ross] also allow uppercase X in hexadecimal character references (like SGML) - - - - - e8f54f25 by Ross Paterson at 2005-01-13T14:44:24+00:00 [haddock @ 2005-01-13 14:44:24 by ross] Describe numeric character references. - - - - - 914ccdce by Sven Panne at 2005-01-15T18:44:48+00:00 [haddock @ 2005-01-15 18:44:45 by panne] Make Haddock compile again after the recent base package changed. The Map/Set legacy hell has been factored out, so that all modules can simply use the new non-deprecated interfaces. Probably a lot of things can be improved by a little bit of Map/Set/List algebra, this can be done later if needed. Small note: Currently the list of instances in HTML code is reversed. This will hopefully be fixed later. - - - - - 6ab20e84 by Sven Panne at 2005-01-16T12:18:26+00:00 [haddock @ 2005-01-16 12:18:26 by panne] Trim imports - - - - - efb81da9 by Sven Panne at 2005-01-16T12:58:08+00:00 [haddock @ 2005-01-16 12:58:03 by panne] Correctly handle the new order of arguments for the combining function given to fromListWith. - - - - - e27b5834 by Sven Panne at 2005-01-16T14:14:41+00:00 [haddock @ 2005-01-16 14:14:39 by panne] Data.Map.unions is left-biased. - - - - - dae3cc3e by Sven Panne at 2005-01-16T14:22:44+00:00 [haddock @ 2005-01-16 14:22:44 by panne] Added the last missing "flip" to get identical HTML output as previous versions. - - - - - 951d8408 by Sven Panne at 2005-01-16T14:37:10+00:00 [haddock @ 2005-01-16 14:37:10 by panne] Refactored Text.PrettyPrint legacy hell into a separate module. - - - - - f1c4b892 by Sven Panne at 2005-01-16T15:41:25+00:00 [haddock @ 2005-01-16 15:41:21 by panne] Cleaned up imports and dropped support for GHC < 5.03, it never worked, anyway. - - - - - 60824c6e by Simon Marlow at 2005-01-18T10:02:48+00:00 [haddock @ 2005-01-18 10:02:48 by simonmar] Add a TODO - - - - - a8c82f23 by Krasimir Angelov at 2005-01-28T23:19:39+00:00 [haddock @ 2005-01-28 23:19:39 by krasimir] import Foreign/Foreign.C are required for Windows - - - - - d8450a23 by Simon Marlow at 2005-02-02T16:23:04+00:00 [haddock @ 2005-02-02 16:23:00 by simonmar] Revamp the linking strategy in Haddock. Now name resolution is done in two phases: - first resolve everything to original names, like a Haskell compiler would. - then, figure out the "home" location for every entity, and point all the links to there. The home location is the lowest non-hidden module in the import hierarchy that documents the entity. If there are multiple candidates, one is chosen at random. Also: - Haddock should not generate any HTML with dangling links any more. Unlinked references are just rendered as plain text. - Error reporting is better: if we can't find a link destination for an entity reference, we now emit a warning. - - - - - 1cce71d0 by Simon Marlow at 2005-02-03T13:42:19+00:00 [haddock @ 2005-02-03 13:42:19 by simonmar] - add --ignore-all-exports flag, which behaves as if every module has the ignore-exports attribute (requested by Chris Ryder). - add --hide option to hide a module on the command line. - add --use-package option to get Haddock info for a package from ghc-pkg (largely untested). - remove reexports from the .haddock file, they aren't used any more. - - - - - 767123ef by Ross Paterson at 2005-02-03T16:17:37+00:00 [haddock @ 2005-02-03 16:17:37 by ross] fix typo for < 6.3 - - - - - 0c680c04 by Simon Marlow at 2005-02-04T12:03:31+00:00 [haddock @ 2005-02-04 12:03:31 by simonmar] Fix bug in renameExportItems that meant links in instances weren't being renamed properly. - - - - - ff7abe5f by Simon Marlow at 2005-02-04T12:15:53+00:00 [haddock @ 2005-02-04 12:15:52 by simonmar] Add attribute #not-home, to indicate that the current module should not be considered to be a home module for the each entity it exports, unless there is no other module that exports the entity. - - - - - fc2cfd27 by Simon Marlow at 2005-02-04T12:40:02+00:00 [haddock @ 2005-02-04 12:40:02 by simonmar] Update the documentation w.r.t. home modules and the not-home attribute. - - - - - 26b8ddf7 by Ross Paterson at 2005-02-04T13:36:06+00:00 [haddock @ 2005-02-04 13:36:05 by ross] sort lists of instances by - arity of the type constructors (so higher-kinded instances come first) - name of the class - argument types - - - - - 26bfb19c by Simon Marlow at 2005-02-23T15:57:12+00:00 [haddock @ 2005-02-23 15:57:12 by simonmar] Fix documentation regarding the module attributes. - - - - - 9c3afd02 by Simon Marlow at 2005-02-28T16:18:17+00:00 [haddock @ 2005-02-28 16:18:17 by simonmar] version 0.7 - - - - - a95fd63f by Simon Marlow at 2005-02-28T16:22:08+00:00 [haddock @ 2005-02-28 16:22:08 by simonmar] Attempt to fix the layout of the package names in the contents. Having tried just about everything, the only thing I can get to work reliably is to make the package names line up on a fixed offset from the left margin. This obviously isn't ideal, so anyone else that would like to have a go at improving it is welcome. One option is to remove the +/- buttons from the contents list and go back to a plain table. The contents page now uses CSS for layout rather than tables. It seems that most browsers have different interpretations of CSS layout, so only the simplest things lead to consistent results. - - - - - 905d42f7 by Simon Marlow at 2005-03-01T17:16:42+00:00 [haddock @ 2005-03-01 17:16:40 by simonmar] Another attempt at lining up the package names on the contents page. Now, they line up with Konqueror, and almost line up with Firefox & IE (different layout in each case). - - - - - a0e1d178 by Wolfgang Thaller at 2005-03-09T08:28:39+00:00 [haddock @ 2005-03-09 08:28:39 by wolfgang] Hack haddock's lexer to accept the output from Apple's broken version of cpp (Apple's cpp leaves #pragma set_debug_pwd directives in it's output). - - - - - 9e1eb784 by Simon Marlow at 2005-04-22T14:27:15+00:00 [haddock @ 2005-04-22 14:27:15 by simonmar] Add a TODO item - - - - - 23281f78 by Ross Paterson at 2005-05-18T12:41:59+00:00 [haddock @ 2005-05-18 12:41:59 by ross] fix 3 bugs in --use-package, and document it. - - - - - 00074a68 by Sven Panne at 2005-05-21T12:35:29+00:00 [haddock @ 2005-05-21 12:35:29 by panne] Warning/versionitis police - - - - - 341fa822 by Simon Marlow at 2005-06-15T15:43:21+00:00 [haddock @ 2005-06-15 15:43:21 by simonmar] Allow "licence" as an alternate spelling of "license" - - - - - 3b953f8b by Simon Marlow at 2005-06-16T08:14:12+00:00 [haddock @ 2005-06-16 08:14:12 by simonmar] wibble - - - - - abfd9826 by Simon Marlow at 2005-06-27T14:46:40+00:00 [haddock @ 2005-06-27 14:46:40 by simonmar] name hierarchical HTML files as A-B-C.html instead of A.B.C.html. The old way confused Apache because the extensions are sometimes interpreted as having special meanings. - - - - - a01eea00 by Simon Marlow at 2005-08-04T13:59:40+00:00 [haddock @ 2005-08-04 13:59:40 by simonmar] 0.7 changes - - - - - 170ef87e by Simon Marlow at 2005-08-04T15:08:03+00:00 [haddock @ 2005-08-04 15:08:03 by simonmar] spec file from Jens Peterson - - - - - 7621fde4 by Simon Marlow at 2005-08-04T15:59:30+00:00 [haddock @ 2005-08-04 15:59:30 by simonmar] replace mingw tests with $(Windows) - - - - - a20739bb by Sven Panne at 2005-08-05T07:01:12+00:00 [haddock @ 2005-08-05 07:01:12 by panne] Reverted to previous version (but with bumped version number), the last commit broke RPM building on SuSE systems due to differently named dependencies. As a clarification: All .spec files in the repository have to work at least on SuSE, because that's the system I'm using. And as "Mr. Building Police", I reserve me the right to keep them that way... >:-) It might very well be the case that we need different .spec files for different platforms, so packagers which are unhappy with the current .spec files should contact me, stating the actual problems. - - - - - 4afb15cf by Simon Marlow at 2005-10-05T10:51:45+00:00 [haddock @ 2005-10-05 10:51:45 by simonmar] Add a bug - - - - - 60f69f82 by Simon Marlow at 2005-10-05T12:52:03+00:00 [haddock @ 2005-10-05 12:52:03 by simonmar] Document new behaviour of -s option - - - - - f7e520ca by Simon Marlow at 2005-10-10T15:02:55+00:00 [haddock @ 2005-10-10 15:02:55 by simonmar] extractRecSel: ignore non-record constructors (fixes a crash when using datatypes with a mixture of record and non-record style constructors). - - - - - b2edbedb by Simon Marlow at 2005-10-14T09:44:21+00:00 Start CHANGES for 0.8 - - - - - 21c7ac8d by Simon Marlow at 2005-10-14T23:11:19+00:00 First cut of Cabal build system - - - - - 766cecdd by Simon Marlow at 2005-10-29T08:14:43+00:00 Add configure script and Makefile for the docs Add a separate configure script and build system for building the documentation. The configure and Makefile code is stolen from fptools. This is left as a separate build system so that the main Cabal setup doesn't require a Unix build environment or DocBook XML tools. - - - - - aa36c783 by Duncan Coutts at 2006-01-17T19:29:55+00:00 Add a --wiki=URL flag to add a per-module link to a correspondng wiki page. So each html page gets an extra link (placed next to the source code and contents links) to a corresponding wiki page. The idea is to let readers contribute their own notes, examples etc to the documentation. Also slightly tidy up the code for the --source option. - - - - - e06e2da2 by Simon Marlow at 2006-01-18T09:28:15+00:00 TODO: documnet --wiki - - - - - 17adfda9 by Duncan Coutts at 2006-01-19T20:17:59+00:00 Add an optional wiki link for each top level exported name. In each module, for each "top level" exported entity we add a hyper link to a corresponding wiki page. The link url gets the name of the exported entity as a '#'-style anchor, so if there is an anchor in the page with that name then the users browser should jump directly to it. By "top level" we mean functions, classes, class members and data types (data, type, newtype), but not data constructors, class instances or data type class membership. The link is added at the right of the page and in a small font. Hopefully this is the right balance of visibility/distraction. We also include a link to the wiki base url in the contents and index pages. - - - - - f52324bb by Duncan Coutts at 2006-01-19T20:28:27+00:00 Rewrite pathJoin to only add a path separator when necessary. When the path ends in a file seperator there is no need to add another. Now using "--wiki=http://blah.com/foo/" should do the right thing. (Code snippet adapted from Isaac's FilePath package.) - - - - - 43bb89fa by Duncan Coutts at 2006-01-21T17:15:27+00:00 Teach haddock about line pragmas and add accurate source code links Teach haddock about C and Haskell style line pragmas. Extend the lexer/parser's source location tracking to include the file name as well as line/column. This way each AST item that is tagged with a SrcLoc gets the original file name too. Use this original file name to add source links to each exported item, in the same visual style as the wiki links. Note that the per-export source links are to the defining module rather than whichever module haddock pretends it is exported from. This is what we want for source code links. The source code link URL can also contain the name of the export so one could implement jumping to the actual location of the function in the file if it were linked to an html version of the source rather than just plain text. The name can be selected with the %N wild card. So for linking to the raw source code one might use: --source=http://darcs/haskell.org/foo/%F Or for linking to html syntax highlighted code: --source=http://darcs/haskell.org/foo/%M.html#%N - - - - - edd9f229 by Duncan Coutts at 2006-01-22T00:02:00+00:00 Extend URL variable expansion syntax and add source links to the contents page Like the wiki link on the contents and index page, add a source code link too. Extend the wiki & source URL variable expansion syntax. The original syntax was: %F for the source file name (the .hs version only, not the .lhs or .hs.pp one) %M for the module name (with '.' replaced by '/') The new syntax is: %F or %{FILE} for the original source file name %M or %{MODULE} for the module name (no replacements) %N or %{NAME} for the function/type export name %K or %{KIND} for a type/value flag "t" or "v" with these extensions: %{MODULE/./c} to replace the '.' module seperator with any other char c %{VAR|some text with the % char in it} which means if the VAR is not in use in this URL context then "" else replace the given text with the '%' char replaced by the string value of the VAR. This extension allows us to construct URLs wit optional parts, since the module/file name is not available for the URL in the contents/index pages and the value/type name is not available for the URL at the top level of each module. - - - - - eb3c6ada by Duncan Coutts at 2006-01-23T13:42:34+00:00 Remove the complex substitutions and add more command line flags instead. Instead of incomprehensable URL substitutions like ${MODULE/./-|?m=%} we now use three seperate command line flags for the top level, per-module and per-entity source and wiki links. They are: --source-base, --source-module, --source-entity --comments-base, --comments-module, --comments-entity We leave -s, --source as an alias for --source-module which is how that option behaved previously. The long forms of the substitutions are still available, ${FILE} ${MODULE} etc and the only non-trivial substitution is ${MODULE/./c} to replace the '.' characters in the module name with any other character c. eg ${MODULE/./-} Seperating the source and wiki url flags has the added bonus that they can be turned on or off individually. So users can have per-module links for example without having to also have per-entity links.` - - - - - a2f0f2af by Duncan Coutts at 2006-01-23T13:54:52+00:00 Make the --help output fit in 80 columns. This is a purely cosmetic patch, feel free to ignore it. The only trickery going on is that we don't display the deprecated -s, --source flags in the help message, but we do still accept them. - - - - - 2d3a4b0c by Duncan Coutts at 2006-01-23T14:12:16+00:00 Add documentation for the new --source-* and --comments-* command line options - - - - - 1a82a297 by Simon Marlow at 2006-01-23T17:03:27+00:00 fix markup - - - - - 100d464a by Duncan Coutts at 2006-01-23T18:31:13+00:00 remove a couple TODO items that have been done The --wiki, or rather the --comment-* options are now documented. There is probably no need to have haddock invoke unlit or cpp itself since it can now pick up the line pragmas to get the source locations right. Tools like Cabal will arrange for preprocessors to be run so there is less of a need for tools like haddock to do it themselves. - - - - - 3162fa91 by Simon Marlow at 2006-01-24T14:21:56+00:00 add a test I had lying around - - - - - 98947063 by Simon Marlow at 2006-01-31T13:52:54+00:00 add scabal-version field - - - - - c41876e6 by Neil Mitchell at 2006-02-26T17:48:21+00:00 Add Hoogle output option - - - - - f86fb9c0 by Simon Marlow at 2006-03-08T09:15:20+00:00 add haskell.vim Contributed by Brad Bowman <bsb at bereft.net>, thanks! - - - - - 35d3c511 by benjamin.franksen at 2006-03-03T22:39:54+00:00 fixed libdir (/html was missing) - - - - - 4d08fd7d by Simon Marlow at 2006-03-10T11:13:31+00:00 add PatternGuards extension - - - - - 3f095e70 by Simon Marlow at 2006-03-13T11:40:42+00:00 bug fixes from Brad Bowman - - - - - 8610849d by Sven Panne at 2006-03-19T17:02:56+00:00 Fixed Cabal/RPM build - - - - - 34a994d6 by sven.panne at 2006-04-20T12:39:23+00:00 Avoid pattern guards Due to the use of pattern guards in Haddock, GHC was called with -fglasgow-exts. This in turn enables bang patterns, too, which broke the Haddock build. Removing some unnecessary pattern guards seemed to be the better way of fixing this instead of using a pragma to disable pattern guards. - - - - - bb523f51 by Ross Paterson at 2006-04-24T09:03:25+00:00 extend 'deriving' heuristic a little If an argument of a data constructor has a type variable head, it is irreducible and the same type class can be copied into the constraint. (Formerly we just did this for type variable arguments.) - - - - - dab9fe7a by Simon Marlow at 2006-04-26T10:02:31+00:00 record an idea - - - - - 748b7078 by Simon Marlow at 2006-05-08T08:28:53+00:00 add section about deriving - - - - - 11252ea1 by Simon Marlow at 2006-05-24T15:43:10+00:00 replace a fatal error in lexChar with a parseError - - - - - 382c9411 by Simon Marlow at 2006-05-24T15:45:47+00:00 add a bug - - - - - b79272f5 by Simon Marlow at 2006-05-24T15:46:29+00:00 add a bug report - - - - - 912edf65 by David Waern at 2006-07-10T19:09:23+00:00 Initial modifications -- doesn't compile - - - - - a3c7ba99 by David Waern at 2006-07-11T00:54:19+00:00 More porting work -- doesn't compile - - - - - 0a173d19 by David Waern at 2006-07-11T11:30:03+00:00 Make the repos temporarily compile and illustrate a problem - - - - - bad316de by David Waern at 2006-07-11T15:43:47+00:00 Progress on the porting process - - - - - bbf12d02 by David Waern at 2006-07-11T23:07:44+00:00 More progress on the porting -- first pass starting to shape up - - - - - de580ba2 by David Waern at 2006-07-20T17:48:30+00:00 More progress -- still on phase1 - - - - - 75a917a2 by David Waern at 2006-07-23T18:22:43+00:00 More work on pass1 -- mostly done - - - - - 6697b3f7 by David Waern at 2006-07-23T22:17:40+00:00 More work, started working on the renaming phase -- this code will need a cleanup soon :) - - - - - 82a5bcbb by David Waern at 2006-07-29T16:16:43+00:00 Add instances, build renaming environment, start on the renamer - - - - - c3f8f4f1 by David Waern at 2006-07-29T21:37:48+00:00 Complete the renamer - - - - - 7e00d464 by David Waern at 2006-07-30T21:01:57+00:00 Start porting the Html renderer - - - - - f04ce121 by David Waern at 2006-08-09T20:04:56+00:00 More Html rendering progress - - - - - 20c21b53 by David Waern at 2006-08-10T17:37:47+00:00 More progress - - - - - d7097e0d by David Waern at 2006-08-11T20:31:51+00:00 Cleanup - - - - - a7351e86 by David Waern at 2006-08-12T11:44:47+00:00 Render H98 Data declarations - - - - - 3fb2208e by David Waern at 2006-08-12T17:15:34+00:00 Perfect rendering of Test.hs - - - - - 454fd062 by David Waern at 2006-08-13T21:57:08+00:00 Misc fixes and interface load/save - - - - - 7ef7e7be by David Waern at 2006-08-14T00:56:07+00:00 Some refactoring - - - - - a7d3efef by David Waern at 2006-08-19T20:07:55+00:00 Adapt to latest GHC - - - - - 5fc3c0d7 by David Waern at 2006-08-20T21:28:11+00:00 Move interface read/write to its own module + some cleanup - - - - - 037e011c by David Waern at 2006-08-20T21:38:24+00:00 Small cleanup - - - - - da3a1023 by David Waern at 2006-09-03T16:05:22+00:00 Change mode to BatchCompile to avoid GHC API bug - - - - - 3cc9be3b by David Waern at 2006-09-03T16:06:59+00:00 Starting work on GADT rendering - - - - - 94506037 by David Waern at 2006-09-03T20:02:48+00:00 Compensate for change of export list order in GHC - - - - - c2cec4eb by David Waern at 2006-09-04T20:53:01+00:00 Rename a function - - - - - 9a9735ba by David Waern at 2006-09-05T15:51:21+00:00 Change version number to 2.0 - - - - - 3758a714 by David Waern at 2006-09-05T15:51:49+00:00 Align comment properly - - - - - 68478d9e by David Waern at 2006-09-15T18:03:00+00:00 Remove interface reading/writing code and use the GHC api for creating package environments instead - - - - - d2eedd95 by David Waern at 2006-09-15T18:05:29+00:00 Change the executable name to haddock-ghc-nolib - - - - - fcfbcf66 by David Waern at 2006-09-15T18:05:45+00:00 Small source code cleanup - - - - - d08eb017 by David Waern at 2006-09-15T18:06:21+00:00 Remove handling of --package flag - - - - - b8a4cf53 by David Waern at 2006-09-15T18:07:16+00:00 Remove commented-out code - - - - - bef0a684 by David Waern at 2006-09-15T18:37:57+00:00 Don't warn about missing links to () - - - - - e7d25fd7 by David Waern at 2006-09-15T19:50:49+00:00 Remove Interface and Binary2 modules - - - - - 9894f2a1 by David Waern at 2006-09-15T19:53:43+00:00 Remove debug printing from HaddockHtml - - - - - a0e7455d by David Waern at 2006-09-16T00:16:29+00:00 Comments only - - - - - d5b26fa7 by David Waern at 2006-09-16T00:16:57+00:00 Refactor PackageData creation code and start on building the doc env propery (unfinished) - - - - - 06aaa779 by David Waern at 2006-09-16T00:19:25+00:00 Better comments in Main.hs - - - - - 1a52d1b4 by David Waern at 2006-09-18T22:17:11+00:00 Comments and spacing change - - - - - e5a97767 by David Waern at 2006-09-21T17:02:45+00:00 Remove unnecessary fmapM import in Main - - - - - 9d0f9d3a by David Waern at 2006-09-22T18:07:07+00:00 Make import list in HaddockHtml prettier - - - - - 3452f662 by David Waern at 2006-09-22T18:08:47+00:00 Refactor context rendering - - - - - 12d0a6d0 by David Waern at 2006-09-22T18:09:52+00:00 Do proper HsType rendering (inser parentheses correctly) - - - - - 2c20c2f9 by David Waern at 2006-09-22T18:10:45+00:00 Fix a bug in Main.toHsType - - - - - c5396443 by David Waern at 2006-09-22T18:11:16+00:00 Skip external package modules sort for now - - - - - 3fb95547 by David Waern at 2006-09-22T20:35:40+00:00 Take away trailin "2" on all previously clashing type names - - - - - 2174755f by David Waern at 2006-09-22T20:51:43+00:00 Remove unused imports in Main - - - - - 1e9f7a39 by David Waern at 2006-09-22T20:52:11+00:00 Fix a comment in Main - - - - - 32d9e028 by David Waern at 2006-10-05T16:40:11+00:00 Merge with changes to ghc HEAD - - - - - 3058c8f5 by David Waern at 2006-10-05T16:41:02+00:00 Comment fixes - - - - - b9c217ec by David Waern at 2006-10-05T16:49:59+00:00 Filter out more builtin type constructors from warning messages - - - - - 67e7d252 by David Waern at 2006-10-05T19:38:22+00:00 Refactoring -- better structured pass1 - - - - - cd21c0c1 by David Waern at 2006-10-05T19:44:42+00:00 Remove read/dump interface flags - - - - - 313f9e69 by David Waern at 2006-10-05T19:49:26+00:00 Remove unused pretty printing - - - - - 480f09d1 by David Waern at 2006-12-28T13:22:24+00:00 Update to build with latest GHC HEAD - - - - - 63dccfcb by David Waern at 2007-01-05T01:38:45+00:00 Fixed a bug so that --ghc-flag works correctly - - - - - 3117dadc by David Waern at 2006-12-29T18:53:39+00:00 Automatically get the GHC lib dir - - - - - 9dc84a5c by David Waern at 2006-12-29T19:58:53+00:00 Comments - - - - - 0b0237cc by David Waern at 2007-01-05T16:48:30+00:00 Collect docs based on SrcLoc, syncing with removal of DeclEntity from GHC - - - - - a962c256 by David Waern at 2007-01-05T17:02:47+00:00 Add tabs in haddock.cabal - - - - - 0ca30c97 by David Waern at 2007-01-05T17:04:11+00:00 Add GHCUtils.hs - - - - - c0ab9abe by David Waern at 2007-01-10T11:43:08+00:00 Change package name to haddock-ghc, version 0.1 - - - - - 38e18b27 by David Waern at 2007-01-12T12:03:52+00:00 No binder name for foreign exports - - - - - d18587ab by David Waern at 2007-01-12T12:08:15+00:00 Temp record - - - - - ba6251a0 by David Waern at 2007-01-12T18:27:55+00:00 Remove read/dump-interface (again) - - - - - f4ba2b39 by David Waern at 2007-01-12T18:31:36+00:00 Remove DocOption, use the GHC type - - - - - 511be8bd by David Waern at 2007-01-12T18:32:41+00:00 Use exceptions instead of Either when loading package info - - - - - 0f2144d8 by David Waern at 2007-01-12T18:33:23+00:00 Small type change - - - - - 77507eb7 by David Waern at 2007-01-12T18:33:59+00:00 Remove interface file read/write - - - - - 0ea1e14f by David Waern at 2007-01-17T21:40:26+00:00 Add trace_ppr to GHCUtils - - - - - 3878b493 by David Waern at 2007-01-17T21:40:53+00:00 Sort external package modules and build a doc env - - - - - 8dc323fc by David Waern at 2007-01-17T21:42:41+00:00 Remove comment - - - - - f4c5b097 by David Waern at 2007-01-18T23:22:18+00:00 Add haddock-ghc.cabal and remove ghc option pragma in source file - - - - - da242b2c by David Waern at 2007-01-18T23:22:46+00:00 Remove some tabs - - - - - 288ed096 by David Waern at 2007-01-18T23:39:28+00:00 Moved the defaultErrorHandler to scope only over sortAndCheckModules for now - - - - - 4dd150fe by David Waern at 2007-02-03T21:23:56+00:00 Let restrictCons handle infix constructors - - - - - 97893442 by David Waern at 2007-02-04T16:26:00+00:00 Render infix data constructors - - - - - da89db72 by David Waern at 2007-02-04T16:26:33+00:00 CHange project name to Haddock-GHC - - - - - e93d48af by David Waern at 2007-02-04T16:59:08+00:00 Render infix type constructors properly - - - - - 357bc99b by David Waern at 2007-02-04T17:37:08+00:00 Insert spaces around infix function names - - - - - ab6cfc49 by David Waern at 2007-02-04T17:59:54+00:00 Do not list entities without documentation - - - - - 04249c7e by David Waern at 2007-02-04T19:16:25+00:00 Add GADT support (quite untested) - - - - - 2c223f8d by David Waern at 2007-02-04T19:25:10+00:00 Add package file write/save again! - - - - - b07ed218 by David Waern at 2007-02-04T19:33:02+00:00 Comment out minf_iface based stuff - - - - - 953d1fa7 by David Waern at 2007-02-05T00:12:23+00:00 Solve conflicts - - - - - 593247fc by David Waern at 2007-02-06T19:48:48+00:00 Remove -package flag, GHC's can be used instead - - - - - f658ded2 by David Waern at 2007-02-06T20:50:44+00:00 Start for support of ATs - - - - - 97f9e913 by David Waern at 2007-02-06T20:52:27+00:00 Wibble - - - - - 2ce8e4cf by David Waern at 2007-02-16T12:09:49+00:00 Add the DocOptions change - - - - - dee4a9b5 by David Waern at 2007-03-06T01:24:48+00:00 Wibble - - - - - 7cb99d18 by David Waern at 2007-03-06T01:24:58+00:00 Change version to 2.0 and executable name to haddock - - - - - c5aa02bc by David Waern at 2007-03-08T15:59:49+00:00 Go back to -B flag - - - - - 3a349201 by David Waern at 2007-03-09T13:31:59+00:00 Better exception handling and parsing of GHC flags - - - - - 05a69b71 by David Waern at 2007-03-09T17:45:44+00:00 Remove commented-out DocEntity printing - - - - - 755032cb by davve at dtek.chalmers.se at 2007-03-23T23:30:20+00:00 Remove a file that shouldn't be here - - - - - a7077e5f by davve at dtek.chalmers.se at 2007-03-24T03:58:48+00:00 Remove an import - - - - - 6f55aa8b by davve at dtek.chalmers.se at 2007-03-25T00:46:48+00:00 Start work on Haddock API - - - - - f0199480 by davve at dtek.chalmers.se at 2007-03-25T00:56:36+00:00 Prettify some comments - - - - - f952f9d1 by davve at dtek.chalmers.se at 2007-03-25T00:56:53+00:00 Remove ppr in HaddockTypes - - - - - bc594904 by davve at dtek.chalmers.se at 2007-03-25T00:57:53+00:00 Remove commented out doc env inference - - - - - 11ebf08d by davve at dtek.chalmers.se at 2007-03-25T01:23:25+00:00 De-flatten the namespace - - - - - f696b4bc by davve at dtek.chalmers.se at 2007-03-25T03:21:48+00:00 Add missing stuff to API - - - - - 9a2a04c3 by davve at dtek.chalmers.se at 2007-03-25T03:22:02+00:00 Wibble - - - - - 7d04a6d5 by davve at dtek.chalmers.se at 2007-03-25T03:22:08+00:00 Avoid a GHC bug with parseStaticFlags [] - - - - - 4d2820ba by davve at dtek.chalmers.se at 2007-03-26T04:57:01+00:00 Add fall-through case to mkExportItem - - - - - 6ebc8950 by Stefan O'Rear at 2007-03-26T04:14:53+00:00 Add shebang line to Setup.lhs - - - - - 80966ec5 by davve at dtek.chalmers.se at 2007-03-26T05:24:26+00:00 Fix stupid compile error - - - - - 1ea1385d by davve at dtek.chalmers.se at 2007-04-05T17:19:56+00:00 Do save/read of interface files properly - - - - - 0e4f6541 by David Waern at 2007-04-10T21:08:36+00:00 Add version to ghc dependency - - - - - b0499b63 by David Waern at 2007-04-10T21:37:08+00:00 Change package name to haddock - - - - - 9d50d27e by David Waern at 2007-04-24T00:22:14+00:00 Use filepath package instead of FilePath - - - - - 87c7fcdf by David Waern at 2007-07-10T21:03:04+00:00 Add new package dependencies - - - - - 4768709c by David Waern at 2007-07-11T20:37:11+00:00 Follow changes to record constructor representation - - - - - b9a02fee by Simon Marlow at 2007-05-30T14:00:48+00:00 update to compile with the latest GHC & Cabal - - - - - c0ebdc01 by David Waern at 2007-07-11T21:35:45+00:00 Fix conflicts - - - - - 97f7afd4 by David Waern at 2007-07-11T21:52:38+00:00 Follow changes to the GHC API - - - - - a5b7b58f by David Waern at 2007-07-12T20:36:48+00:00 Call parseStaticFlags before newSession - - - - - f7f50dbc by David Waern at 2007-08-01T21:52:58+00:00 Better indentation in haddock.cabal - - - - - d84e52ad by David Waern at 2007-08-02T00:08:18+00:00 Wibble - - - - - a23f494a by David Waern at 2007-08-02T00:08:24+00:00 Be better at trying to load all module dependencies (debugging) - - - - - ee917f13 by David Waern at 2007-08-03T18:48:08+00:00 Load all targets explicitly (checkModule doesn't chase dependencies anymore) - - - - - 5182d631 by David Waern at 2007-08-16T16:48:55+00:00 Finalize support for links to other packages - - - - - dfd1e3da by David Waern at 2007-08-16T16:51:11+00:00 Fix haddock comment errors in Haddock.Types - - - - - 50c0d83e by David Waern at 2007-08-16T16:51:37+00:00 Remove a debug import - - - - - d84b7c2b by David Waern at 2007-08-16T17:06:30+00:00 Rename PackageData to HaddockPackage - - - - - 3b52cb9f by David Waern at 2007-08-16T17:09:42+00:00 Simplify some comments - - - - - 66fa68d9 by David Waern at 2007-08-16T17:11:38+00:00 Comment the HaddockPackage definition - - - - - 8674c761 by David Waern at 2007-08-16T17:25:54+00:00 Improve code layout in Main - - - - - 571a3a0b by David Waern at 2007-08-16T17:32:13+00:00 Remove explict module imports in Main - - - - - d31b3cb0 by David Waern at 2007-08-16T17:36:23+00:00 Correct comments - - - - - 7f8a9f2b by David Waern at 2007-08-16T17:39:50+00:00 Fix layout problems in Haddock.Types - - - - - 9f421d7f by David Waern at 2007-08-17T11:16:48+00:00 Move options out of Main into Haddock.Options - - - - - 80042b63 by David Waern at 2007-08-17T11:26:59+00:00 Small comment/layout fixes - - - - - b141b982 by David Waern at 2007-08-17T11:28:28+00:00 Change project name from Haddock-GHC to Haddock - - - - - dbeb4a81 by David Waern at 2007-08-17T11:41:05+00:00 Add top module comment to all files - - - - - ce99cc9e by David Waern at 2007-08-17T14:53:04+00:00 Factor out typechecking phase into Haddock.Typecheck - - - - - 6bf75d9e by David Waern at 2007-08-17T16:55:35+00:00 Factor out package code to Haddock.Packages - - - - - b396db37 by David Waern at 2007-08-29T22:40:23+00:00 Major refactoring - - - - - 3d4f95ee by David Waern at 2007-08-29T23:26:24+00:00 Rename HaddockModule to Interface and a few more refactorings - - - - - c55326db by David Waern at 2007-08-29T23:48:03+00:00 Some comment cleanup - - - - - 9a84fc46 by David Waern at 2007-08-29T23:49:29+00:00 Add some modules that I forgot to add earlier - - - - - 4536dce2 by David Waern at 2007-08-29T23:55:24+00:00 Wibble - - - - - 9b7f0206 by David Waern at 2007-08-30T16:03:29+00:00 Wibble - - - - - c52c050a by David Waern at 2007-08-30T16:30:37+00:00 Rename HaddockModule to Interface - - - - - eae2995f by David Waern at 2007-08-30T16:42:59+00:00 Simplify createInterfaces - - - - - 53f99caa by David Waern at 2007-09-29T00:04:31+00:00 Add build-type: Simple to the cabal file - - - - - 0d3103a8 by David Waern at 2007-09-29T00:04:58+00:00 Add containers and array dependency - - - - - 6acf5f30 by David Waern at 2007-09-29T00:13:36+00:00 Prettify the cabal file - - - - - 87c1e378 by David Waern at 2007-09-29T13:16:39+00:00 FIX: consym data headers with more than two variables - - - - - b67fc16a by David Waern at 2007-09-29T14:01:32+00:00 FIX: prefix types used as operators should be quoted - - - - - a8f925bc by David Waern at 2007-09-29T14:02:26+00:00 Use isSymOcc from OccName instead of isConSym - - - - - fc330701 by David Waern at 2007-09-29T14:15:37+00:00 Use isLexConSym/isLexVarSym from OccName - - - - - e4f3dbad by David Waern at 2007-09-29T15:01:08+00:00 FIX: do not quote varsym type operators - - - - - 402207d2 by David Waern at 2007-09-29T15:01:50+00:00 Wibble - - - - - f9d89ef0 by David Waern at 2007-09-29T15:17:40+00:00 Take care when pp tyvars - add parens on syms - - - - - 849e2a77 by David Waern at 2007-10-01T21:56:39+00:00 Go back to using a ModuleMap instead of LookupMod - fixes a bug - - - - - 549dbac6 by David Waern at 2007-10-02T01:05:19+00:00 Improve parsing of doc options - - - - - a36021b8 by David Waern at 2007-10-02T23:05:00+00:00 FIX: double arrows in constructor contexts - - - - - d03bf347 by David Waern at 2007-10-09T16:14:05+00:00 Add a simple test suite - - - - - c252c140 by David Waern at 2007-10-17T16:02:28+00:00 Add --optghc=.. style flag passing to GHC - - - - - cce6c1b3 by David Waern at 2007-10-18T22:03:20+00:00 Add support for --read-interface again - - - - - 33d059c0 by David Waern at 2007-10-18T22:30:18+00:00 Refactoring -- get rid of Haddock.Packages - - - - - f9ed0a4c by David Waern at 2007-10-18T22:34:36+00:00 Name changes - - - - - 8a1c816f by David Waern at 2007-10-20T14:24:23+00:00 Add --ghc-version option - - - - - 4925aaa1 by David Waern at 2007-10-21T14:34:26+00:00 Add some Outputable utils - - - - - 69e7e47f by David Waern at 2007-10-21T14:35:49+00:00 FIX: Ord for OrdName was not comparing modules - - - - - 5a4ae535 by David Waern at 2007-10-21T21:18:48+00:00 Wibble - - - - - 03d48e20 by David Waern at 2007-10-24T15:52:56+00:00 Remove Main from "other modules" - - - - - c66f6d82 by David Waern at 2007-10-24T16:37:18+00:00 Make it possible to run haddock on itself - - - - - 21d156d8 by David Waern at 2007-10-25T14:02:14+00:00 Don't set boot modules as targets - - - - - f8bcf91c by David Waern at 2007-10-31T22:11:17+00:00 Add optimisation flags - - - - - 7ac758f2 by David Waern at 2007-11-04T09:48:28+00:00 Go back to loading only targets (seems to work now) - - - - - 4862aae1 by David Waern at 2007-11-05T22:24:57+00:00 Do full compilation of modules -- temporary fix for GHC API problem - - - - - 697e1517 by David Waern at 2007-11-05T22:25:50+00:00 Don't warn about not being able to link to wired/system/builtin-names - - - - - 892186da by David Waern at 2007-11-06T00:49:21+00:00 Filter out instances with TyCons that are not exported - - - - - 9548314c by David Waern at 2007-11-06T09:37:14+00:00 Wibble - - - - - 5cafd627 by David Waern at 2007-11-08T01:43:07+00:00 Filter out all non-vanilla type sigs - - - - - 04621830 by David Waern at 2007-11-08T01:45:13+00:00 Synch loading of names from .haddock files with GHC's name cache - - - - - 88d37f77 by David Waern at 2007-11-08T01:46:21+00:00 Remove commented-out code - - - - - 6409c911 by David Waern at 2007-11-08T01:56:00+00:00 Small bugfix and cleanup in getDeclFromTyCls - - - - - af59d9c2 by David Waern at 2007-11-08T02:08:44+00:00 Remove OrdName stuff - - - - - 3a615e2e by David Waern at 2007-11-08T02:13:41+00:00 Update runtests.hs following changes to haddock - - - - - 01f3314e by David Waern at 2007-11-08T02:33:01+00:00 Complain if we can't link to wired-in names - - - - - fcafb5d1 by David Waern at 2007-11-09T02:40:16+00:00 Don't exit when there are no file arguments - - - - - 194bc332 by David Waern at 2007-11-09T02:55:37+00:00 Wibble - - - - - dbe4cb55 by David Waern at 2007-11-09T02:56:14+00:00 Wibble - - - - - 82869fda by David Waern at 2007-11-10T17:01:43+00:00 Introduce InstalledInterface structure and add more stuff to the .haddock files We introduce InstalledInterface capturing the part of Interface that is stored in the interface files. We change the ppHtmlContents and ppHtmllIndex to take this structure instead of a partial Interface. We add stuff like the doc map and exported names to the .haddock file (via InstalledInterface). - - - - - d6bb57bf by David Waern at 2007-11-10T17:19:48+00:00 FIX: contents and index should include external package modules when --gen-contents/--gen-index - - - - - e8814716 by David Waern at 2007-11-11T00:29:27+00:00 Remove lDocLinkName and its use in Html backend - - - - - 6f9bd702 by David Waern at 2007-11-11T00:50:57+00:00 Do some refactoring in the html backend This also merges an old patch by Augustsson: Wed Jul 12 19:54:36 CEST 2006 lennart.augustsson at credit-suisse.com * Print type definitions like signatures if given arrows. - - - - - 09d0ce24 by Malcolm.Wallace at 2006-07-20T13:13:57+00:00 mention HsColour in the docs, next to option flags for linking to source code - - - - - 24da6c34 by Malcolm.Wallace at 2006-07-20T13:14:50+00:00 change doc references to CVS to give darcs repository location instead - - - - - 74d52cd6 by David Waern at 2007-11-11T00:55:33+00:00 Update copyright - - - - - fcaa3b4f by Duncan Coutts at 2006-09-08T13:41:00+00:00 Eliminate dep on network by doing a little cut'n'paste haddock depending on the network causes a circular dependency at least if you want to build the network lib with haddock docs. - - - - - 10cc9bda by David Waern at 2007-11-11T02:09:41+00:00 Fix conflicts - - - - - 4e3acd39 by David Waern at 2007-11-11T02:21:19+00:00 Manual merge of a patch from Duncan Coutts that removes the dependency on mtl - - - - - fa9070da by Neil Mitchell at 2006-09-29T15:52:03+00:00 Do not generate an empty table if there are no exports, this fixes a <table></table> tag being generated, which is not valid HTML 4.01 - - - - - d7431c85 by David Waern at 2007-11-11T02:28:50+00:00 Fix conflicts - - - - - f87e8f98 by Simon Marlow at 2006-10-10T11:37:16+00:00 changes for 0.8 - - - - - db929565 by Simon Marlow at 2006-10-10T12:07:12+00:00 fix the name of the source file - - - - - 8220aa4b by Simon Marlow at 2006-10-11T14:17:37+00:00 Rename haddock.js to haddock-util.js haddock.js will be run automatically by Windows when you type 'haddock' if it is found on the PATH, so rename to avoid confusion. Spotted by Adrian Hey. - - - - - 6bccdaa1 by sven.panne at 2006-10-12T15:28:23+00:00 Cabal's sdist does not generate "-src.tar.gz" files, but ".tar.gz" ones - - - - - d3f3fc19 by Simon Marlow at 2006-12-06T16:05:07+00:00 add todo item for --maintainer - - - - - 2da7e269 by Simon Marlow at 2006-12-15T15:52:00+00:00 TODO: do something better about re-exported symbols from another package - - - - - 42d85549 by David Waern at 2007-11-11T02:30:59+00:00 Fix conflicts - - - - - 5e7ef6e5 by Neil Mitchell at 2007-01-11T15:41:15+00:00 Never do spliting index files into many - - - - - f3d4aebe by Neil Mitchell at 2007-01-11T17:07:09+00:00 Add searching on the index page - - - - - bad3ab66 by Neil Mitchell at 2007-01-11T18:17:46+00:00 Delete dead code, now there is only one index page - - - - - cd09eedb by Neil Mitchell at 2007-01-11T18:21:19+00:00 Delete more stuff that is no longer required - - - - - e2806646 by David Waern at 2007-11-11T02:41:53+00:00 Fix conflicts - - - - - a872a823 by Neil Mitchell at 2007-01-11T18:51:43+00:00 Make the index be in case-insensitive alphabetic order - - - - - 8bddd9d7 by Neil Mitchell at 2007-02-06T17:49:12+00:00 Do not create empty tables for data declarations which don't have any constructors, instances or comments. Gets better HTML 4.01 compliance - - - - - 036b8120 by David Waern at 2007-11-11T02:56:58+00:00 Fix conflicts - - - - - f50c1639 by Conal Elliott at 2007-02-14T21:54:00+00:00 added substitution %{FILE///c} - - - - - 402e166a by David Waern at 2007-11-11T03:35:46+00:00 Manual merge of old patch: Sat Apr 21 04:36:43 CEST 2007 Roberto Zunino <zunrob at users.sf.net> * URL expansion for %%, %L, %{LINE} - - - - - 2f264fbd by David Waern at 2007-11-11T03:40:33+00:00 Manual merge of an old patch: Thu Apr 19 20:23:40 CEST 2007 Wolfgang Jeltsch <g9ks157k at acme.softbase.org> * bug fix When Haddock was invoked with the --ignore-all-exports flag but the ignore-exports module attribute wasn't used, hyperlinks weren't created for non-exported names. This fix might not be as clean as one would wish (since --ignore-all-exports now results in ignore_all_exports = True *and* an additional OptIgnoreExports option for every module) but at least the bug seems to be resolved now. - - - - - 7d7ae106 by sven.panne at 2007-09-02T12:18:02+00:00 Install LICENSE in the correct place - - - - - 66eaa924 by David Waern at 2007-11-11T19:02:46+00:00 Fix a bug that made haddock loop - - - - - 4ed47b58 by David Waern at 2007-11-11T19:03:09+00:00 Rename java-script file (this wasn't merge correctly) - - - - - d569534a by David Waern at 2007-11-11T19:06:44+00:00 Don't require -B <ghc-libdir> when no argument files Change readInterfaceFile to take a Maybe Session, to avoid having to pass -B <ghc-libdir> to Haddock when there're no source files to process. This is nice when computing contents/index for external packages. - - - - - 373368bc by Neil Mitchell at 2007-01-11T18:22:44+00:00 Change from tabs to spaces in the ppHtmlIndex function - - - - - 6b063a77 by Neil Mitchell at 2007-01-12T12:17:46+00:00 Rewrite much of the index searching code, previously was too slow to execute on the base library with IE, the new version guarantees less than O(log n) operations be performed, where n is the number in the list (before was always O(n)) - - - - - bfad00b7 by David Waern at 2007-11-11T23:33:53+00:00 Fix conflicts - - - - - cd2dcc09 by Neil Mitchell at 2007-01-12T12:25:01+00:00 Make the max number of results 75 instead of 50, to allow map searching in the base library to work - - - - - 3ae74764 by Neil Mitchell at 2007-01-12T12:58:17+00:00 Make the search box in a form so that enter does the default search - - - - - 142103e5 by David Waern at 2007-11-12T00:03:18+00:00 Merge patch from the old branch: Fri Aug 31 13:21:45 CEST 2007 Duncan Coutts <duncan at haskell.org> * Add category: Development to .cabal file Otherwise it appears on the hackage website in the "Unclassified" category. - - - - - 22ec2ddb by David Waern at 2007-11-25T01:55:29+00:00 A a list of small improvements to the TODO file - - - - - eb0129f4 by Wolfgang Jeltsch at 2007-12-03T23:47:55+00:00 addition of type equality support (at least for HTML generation) - - - - - 816a7e22 by David Waern at 2007-12-08T15:46:26+00:00 Handle class operators correctly when rendering predicates - - - - - 68baaad2 by David Waern at 2007-12-08T16:15:54+00:00 Code layout changes - - - - - 09b77fb4 by David Waern at 2007-12-08T16:16:03+00:00 Handle infix operators correctly in the Type -> HsType translation - - - - - 31c36da2 by David Waern at 2007-12-08T16:24:27+00:00 Add ppLParendTypes/ppLParendType - - - - - b17cc818 by David Waern at 2007-12-08T16:26:12+00:00 Use ppParendType when printing types args in predicates - - - - - ffd1f2cf by David Waern at 2007-12-08T16:45:06+00:00 Fix rendering of instance heads to handle infix operators This is also a refactoring to share this code for rendering predicates. - - - - - ff886d45 by David Waern at 2007-12-08T17:27:46+00:00 Fix rendering of class operators - - - - - e2fcbb9e by David Waern at 2007-12-08T17:59:28+00:00 Fix a bug (use ppTyName instead of ppName to print names in type apps) - - - - - 79a1056e by David Waern at 2007-12-08T21:25:18+00:00 Update tests - - - - - 867741ac by David Waern at 2007-12-08T21:25:49+00:00 Give a diff on test failure - - - - - 7e5eb274 by David Waern at 2008-01-05T14:33:45+00:00 Add DrIFT commands - - - - - 3656454d by David Waern at 2008-01-05T20:26:00+00:00 Add "cabal-version: >= 1.2" to the cabal file - - - - - 77974efc by Simon Marlow at 2007-12-20T09:52:44+00:00 add an item - - - - - f6ac1708 by Simon Marlow at 2007-12-06T14:00:10+00:00 Source links must point to the original module, not the referring module - - - - - eda1d5c9 by David Waern at 2008-01-06T14:40:52+00:00 Manual merge of a patch to the 0.8 branch Thu Dec 6 15:00:10 CET 2007 Simon Marlow <simonmar at microsoft.com> * Source links must point to the original module, not the referring module - - - - - 378f4085 by David Waern at 2008-01-06T16:03:45+00:00 Change stability from stable to experimental - - - - - 8bdafe44 by David Waern at 2008-01-06T16:14:22+00:00 Add haskell.vim (it had been removed somehow) - - - - - ea34d02e by David Waern at 2008-01-06T16:36:57+00:00 Change version to 2.0.0.0 - - - - - 34631ac0 by David Waern at 2008-01-06T16:44:57+00:00 Add missing modules to the cabal file - - - - - 9e142935 by David Waern at 2008-01-06T17:25:42+00:00 Depend on ghc >= 6.8.2 && < 6.9 - - - - - 59f9eeaa by Simon Marlow at 2007-12-20T10:43:04+00:00 add build scripts - - - - - 1c29ae30 by Simon Marlow at 2007-12-20T10:47:07+00:00 update version number - - - - - fe16a3e4 by Simon Marlow at 2007-12-20T10:48:03+00:00 update version - - - - - f688530f by Simon Marlow at 2007-12-20T10:48:29+00:00 doc updates - - - - - ce71b611 by David Waern at 2008-01-07T13:46:32+00:00 Change version in docs and spec - - - - - 03ab8d6f by David Waern at 2008-01-07T13:47:38+00:00 Manually merge over changes to CHANGES for 0.9 - - - - - 39f1b042 by David Waern at 2008-01-07T15:17:41+00:00 Remove the -use-package flag, we don't support it anyway - - - - - 7274a544 by David Waern at 2008-01-07T15:33:05+00:00 Update CHANGES for 2.0.0.0 - - - - - 96594f5d by David Waern at 2008-01-07T15:46:49+00:00 Wibble - - - - - f4c5a4c4 by David Waern at 2008-01-07T15:55:36+00:00 Change url to repo in documentation - - - - - 8a4c77f0 by David Waern at 2008-01-07T16:00:54+00:00 Update CHANGES - - - - - cb3a9288 by David Waern at 2008-01-07T16:02:55+00:00 Documentation fix - - - - - d8e45539 by David Waern at 2008-01-07T16:12:00+00:00 Update docs to say that Haddock accets .lhs files and module names - - - - - 4b5ce824 by David Waern at 2008-01-07T16:12:25+00:00 Document -B option - - - - - 47274262 by David Waern at 2008-01-07T16:23:07+00:00 Update CHANGES - - - - - 7ff314a9 by David Waern at 2008-01-07T16:23:20+00:00 Remove --use-package, --package & --no-implicit.. flags from docs - - - - - 6c3819c0 by David Waern at 2008-01-07T16:23:52+00:00 Remove --no-implicit-prelide flag - - - - - 1b14ae40 by David Waern at 2008-01-07T16:32:26+00:00 Update the "Using literate or pre-processed source" section - - - - - 0117f620 by David Waern at 2008-01-07T16:41:55+00:00 Document the --optghc flag - - - - - 087ab1cf by David Waern at 2008-01-07T16:42:10+00:00 Remove the documenation section on derived instances The problem mentioned there doesn't exist in Haddock 2.0.0.0 - - - - - 7253951e by David Waern at 2008-01-07T16:48:40+00:00 Document OPTIONS_HADDOCK - - - - - 3b6bdcf6 by David Waern at 2008-01-07T16:56:54+00:00 Wibble - - - - - 3025adf9 by David Waern at 2008-01-07T17:08:14+00:00 Wibble - - - - - 5f30f1a0 by David Waern at 2008-01-07T17:15:44+00:00 Change synopsis field to description - - - - - 1673f54b by David Waern at 2008-01-07T17:18:21+00:00 Change my email address in the cabal file - - - - - 55aa9808 by David Waern at 2008-01-07T18:18:02+00:00 Add documentation for readInterfaceFile - - - - - eaea417f by David Waern at 2008-01-07T18:21:30+00:00 Export necessary stuff from Distribution.Haddock - - - - - 7ea18759 by David Waern at 2008-01-07T18:31:49+00:00 Remove dep on Cabal - - - - - 7b79c74e by David Waern at 2008-01-07T18:33:49+00:00 Remove dep on process - - - - - ce3054e6 by David Waern at 2008-01-16T23:01:21+00:00 Add feature-requsts from Henning Thielemann to TODO - - - - - 0c08f1ec by David Waern at 2008-01-16T23:03:02+00:00 Record a bug in TODO - - - - - b04605f3 by David Waern at 2008-01-23T16:59:06+00:00 Add a bug reported by Ross to TODO - - - - - 5b17c030 by David Waern at 2008-01-23T18:05:53+00:00 A a bug report to TODO - - - - - 1c993b0d by David Waern at 2008-01-25T16:30:25+00:00 Accept test output - - - - - c22fc0d0 by David Waern at 2008-01-25T16:34:49+00:00 Accept test output - - - - - 4b795811 by David Waern at 2008-01-25T16:38:37+00:00 Change Hidden.hs (test) to use OPTIONS_HADDOCK - - - - - c124dbd9 by David Waern at 2008-01-25T16:39:23+00:00 Accept test output - - - - - ec6f6eea by David Waern at 2008-01-25T16:42:08+00:00 Add Hidden.html.ref to tests - - - - - 1dc9610c by David Waern at 2008-02-02T20:50:51+00:00 Add a comment about UNPACK bug in TODO - - - - - 2d3f7081 by David Waern at 2008-02-09T22:33:24+00:00 Change the representation of DocNames Ross Paterson reported a bug where links would point to the defining module instead of the "best" module for an identifier (e.g Int pointing to GHC.Base instead of Data.Int). This patch fixes this problem by refactoring the way renamed names are represented. Instead of representing them by: > data DocName = Link Name | NoLink Name they are now represented as such: > data DocName = Documented Name Module | Undocumented Name and the the link-env looks like this: > type LinkEnv = Map Name Module There are several reasons for this. First of all, the bug was caused by changing the module part of Names during the renaming process, without changing the Unique field. This caused names to be overwritten during the loading of .haddock files (which caches names using the NameCache of the GHC session). So we might create new Uniques during renaming to fix this (but I'm not sure that would be problem-free). Instead, we just keep the Name and add the Module where the name is best documented, since it can be useful to keep the original Name around (for e.g. source-code location info and for users of the Haddock API). Also, the names Link/NoLink don't really make sense, since wether to use links or not is entirely up to the users of DocName. In the process of following this change into H.Backends.Html I removed the assumption that binder names are Undocumented (which was just an unnecessary assumption, the OccName is the only thing needed to render these). This will probably make it possible to get rid of the renamer and replace it with a traversal from SYB or Uniplate. Since DocName has changed, InterfaceFile has changed so this patch also increments the file-format version. No backwards-compatibility is implemented. - - - - - 0f28c921 by David Waern at 2008-02-09T23:00:36+00:00 H.GHC.Utils: remove unused imports/exports - - - - - 0c44cad5 by David Waern at 2008-02-10T00:28:13+00:00 H.GHC.Utils: add some functions that were removed by mistake - - - - - e3452f49 by David Waern at 2008-02-10T00:28:48+00:00 Fix some trivial warnings in H.InterfaceFile - - - - - a6d74644 by David Waern at 2008-02-10T00:48:06+00:00 Update the version message to fit in small terminals - - - - - 76c9cd3e by David Waern at 2008-02-10T14:47:39+00:00 Remove bugs from TODO that don't apply anymore since the port - - - - - 5e10e090 by David Waern at 2008-02-10T15:22:47+00:00 Remove bugs from TODO that weren't actual bugs - - - - - fef70878 by David Waern at 2008-02-10T15:23:44+00:00 Remove yet another item from TODO that was not an actual bug - - - - - e1af47b8 by David Waern at 2008-02-11T10:25:57+00:00 Bump the version number to 2.1.0 Since the exported datatype DocName has changed, we need to bump the major version number. Let's also drop the fourth version component, it's not that useful. - - - - - e3be7825 by David Waern at 2008-04-11T14:29:04+00:00 Add a bug to TODO - - - - - cb6574be by David Waern at 2008-04-11T16:00:45+00:00 Use the in-place haddock when running tests - - - - - c6d7af0d by David Waern at 2008-04-11T16:09:16+00:00 Turn off GHC warnings when running tests - - - - - 7f61b546 by David Waern at 2008-04-11T17:24:00+00:00 Add a flag for turning off all warnings - - - - - 883b8422 by David Waern at 2008-04-12T14:02:18+00:00 Fix printing of data binders - - - - - 2a0db8fc by David Waern at 2008-04-12T18:52:46+00:00 Fix missing parenthesis in constructor args bug - - - - - 1b3ac3f9 by David Waern at 2008-04-12T18:57:23+00:00 Simplify test suite and add tests I move all tests into one single directory to simplify things, and add a test for the last bug that was fixed. - - - - - 8f178376 by David Waern at 2008-04-12T19:00:15+00:00 Add a script for copying test output to "expected" output - - - - - 193e3a03 by David Waern at 2008-04-12T19:16:37+00:00 Remove two fixed bugs from TODO - - - - - ddc9130c by David Waern at 2008-04-12T19:37:06+00:00 Update test README - - - - - 956069c0 by David Waern at 2008-05-01T12:16:14+00:00 Update version number in spec and docs - - - - - 5478621c by David Waern at 2008-05-01T12:28:12+00:00 Remove claim of backwards compatibility from docs for readInterfaceFile - - - - - 4a16dea9 by David Waern at 2008-05-01T12:33:04+00:00 Update CHANGES - - - - - 804216fb by David Waern at 2008-05-01T12:43:16+00:00 Add a synopsis - - - - - fd0c84d5 by David Waern at 2008-05-01T12:44:44+00:00 Add Haddock.DocName to the cabal file - - - - - 9f4a7439 by David Waern at 2008-05-01T12:45:53+00:00 Remove -fglasgow-exts and -fasm - - - - - aee7c145 by David Waern at 2008-05-01T12:54:01+00:00 Add LANGUAGE pragmas to source files - - - - - 9a58428b by David Waern at 2008-05-01T12:54:19+00:00 Add extensions to cabal file - - - - - 494f1bee by David Waern at 2008-05-01T13:12:09+00:00 Export DocName in the API - - - - - c938196b by David Waern at 2008-05-01T13:12:19+00:00 Add hide options to some source files - - - - - 236e86af by Neil Mitchell at 2008-06-07T20:45:10+00:00 Rewrite the --hoogle flag support - - - - - 6d910950 by Neil Mitchell at 2008-06-14T10:56:50+00:00 Simplify the newtype/data outputting in Hoogle, as haddock does it automatically - - - - - f87a95a8 by Neil Mitchell at 2008-06-14T12:10:18+00:00 Add initial structure for outputting documentation as well, but does not yet output anything - - - - - 7c3bce54 by Neil Mitchell at 2008-06-14T12:27:07+00:00 Remove <document comment> from the Hoogle output - - - - - 9504a325 by Neil Mitchell at 2008-06-16T06:33:21+00:00 Default to "main" if there is no package, otherwise will clobber hoogle's hoogle info - - - - - 4a794a79 by Neil Mitchell at 2008-06-16T06:53:29+00:00 Change packageName to packageStr, as it better reflects the information stored in it - - - - - 7abc9baf by Neil Mitchell at 2008-06-16T07:09:49+00:00 Add modulePkgInfo to Haddock.GHC.Utils, which gives back package name and version info - - - - - 8ca11514 by Neil Mitchell at 2008-06-16T07:13:48+00:00 Change Hoogle to take the package name and package version separately - - - - - a6da452d by Neil Mitchell at 2008-06-18T11:29:46+00:00 In Hoogle do not list things that are not local to this module - - - - - 974b76b7 by David Waern at 2008-06-19T18:40:13+00:00 Be more consistent with GHC API naming in H.GHC.Utils - - - - - 2facb4eb by David Waern at 2008-06-19T19:03:03+00:00 Update test output - - - - - c501de72 by David Waern at 2008-06-26T20:26:49+00:00 Use ghc-paths to get the lib dir The path can still be overridden using the -B flag. It's not longer required to pass the lib dir to the program that runs the test suite. - - - - - ac4c6836 by David Waern at 2008-06-26T20:33:08+00:00 Update CHANGES - - - - - 9d21c60a by David Waern at 2008-06-26T20:34:53+00:00 Update README - - - - - 741448f0 by David Waern at 2008-06-26T21:12:57+00:00 Improve wording in the help message - - - - - b1b42b11 by David Waern at 2008-06-30T10:16:17+00:00 Rename ForeignType - - - - - 6d6c2b34 by David Waern at 2008-06-30T10:25:09+00:00 Rename TyFamily - - - - - 8d1125ed by David Waern at 2008-06-30T10:37:21+00:00 Rename type patterns - - - - - 7610a4cb by David Waern at 2008-06-30T10:45:07+00:00 Rename associated types - - - - - 8eeba14c by David Waern at 2008-06-30T10:47:41+00:00 Remove the TODO file now that we have a trac - - - - - 1af5b25b by David Waern at 2008-07-02T18:19:28+00:00 Render type family declarations (untested) - - - - - ceb99797 by David Waern at 2008-07-02T18:24:06+00:00 Remove redundant check for summary when rendering data types - - - - - b36a58e0 by David Waern at 2008-07-02T22:01:38+00:00 More support for type families and associated types Now we just need to render the instances - - - - - 78784879 by David Waern at 2008-07-07T22:13:58+00:00 Remove filtering of instances We were filtering out all instances for types with unknown names. This was probably an attempt to filter out instances for internal types. I am removing the filtering for the moment, and will try to fix this properly later. - - - - - 3e758dad by David Waern at 2008-06-30T18:50:30+00:00 Run haddock in-place during testing - - - - - d9dab0ce by David Waern at 2008-07-08T21:04:32+00:00 Remove index.html and doc-index.html from output, they should not be versioned - - - - - 3e6c4681 by David Waern at 2008-07-08T21:06:42+00:00 Update test output following change to instance filtering - - - - - e34a3f14 by David Waern at 2008-07-12T16:48:28+00:00 Stop using the map from exported names to declarations During creation of the interface, we were using two maps: one from exported names to declarations, and one from all defined names in the module to declarations. The first contained subordinate names while the second one didn't. The first map was never used to look up names not defined in the associated module, so if we add subordinate names to the second map, we could use it everywhere. That's that this patch does. This simplifies code because we don't have to pass around two maps everywhere. We now store the map from locally defined things in the interface structure instead of the one from exported names. - - - - - 2e1d2766 by David Waern at 2008-07-12T16:55:21+00:00 Get the all locally defined names from GHC API We previously had some code to compute all locally defined names in a module including subordinate names. We don't need it since we can get the names from modInfoTyThings in the GHC API. - - - - - bf637994 by David Waern at 2008-07-13T13:09:16+00:00 Refactoring in H.Interface.Create We were creating a doc map, a declaration map and a list of entities separately by going through the HsGroup. These structures were all used to build the interface of a module. Instead of doing this, we can start by creating a list of declarations from the HsGroup, then collect the docs directly from this list (instead of using the list of entities), creating a documentation map. We no longer need the Entity data type, and we can store a single map from names to declarations and docs in the interface, instead of the declaration map and the doc map. This way, there is only one place where we filter out the declarations that we don't want, and we can remove a lot of code. Another advantage of this is that we can create the exports directly out of the list of declarations when we export the full module contents. (Previously we did a look up for each name to find the declarations). This is faster and removes another point where we depend on names to identify exported declarations, which is good because it eliminates problems with instances (which don't have names). - - - - - 547e410e by David Waern at 2008-07-13T13:34:51+00:00 Remove FastString import and FSLIT macro in H.I.Create -- they were unused - - - - - 693759d1 by David Waern at 2008-07-13T13:36:23+00:00 Remove unused import from H.I.Create - - - - - cde6e7fb by David Waern at 2008-07-13T13:51:54+00:00 Small touches - - - - - 96de8f1d by David Waern at 2008-07-20T11:21:46+00:00 Preparation for rendering instances as separate declarations We want to be able to render instances as separate declarations. So we remove the Name argument of ExportDecl, since instances are nameless. This patch also contains the first steps needed to gather type family instances and display them in the backend, but the implementation is far from complete. Because of this, we don't actually show the instances yet. - - - - - b0f824fb by David Waern at 2008-07-20T15:53:08+00:00 Follow changes to ExportDecl in Hoogle - - - - - 1192eff3 by Neil Mitchell at 2008-06-26T00:28:10+00:00 Change how the Hoogle backend outputs classes, adding the context in - - - - - 7a0d1464 by Neil Mitchell at 2008-06-26T00:28:46+00:00 Remove the indent utility function from Hoogle backend - - - - - 3361241b by Neil Mitchell at 2008-06-26T09:45:09+00:00 Add support for Hoogle writing ForeignImport/ForeignExport properly - - - - - 795ad3bf by Neil Mitchell at 2008-06-26T12:15:25+00:00 Flesh out the Hoogle code to render documentation - - - - - 23277995 by Neil Mitchell at 2008-06-26T14:56:41+00:00 Fix a bug in the Hoogle backend, unordered lists were being written out <ul>...</u> - - - - - db739b27 by Neil Mitchell at 2008-06-26T15:09:54+00:00 Remove any white space around a <li> element - - - - - f2e6bb8c by Neil Mitchell at 2008-07-10T15:30:47+00:00 Remove the TODO in the Hoogle HTML generation, was already done - - - - - 693ec9a3 by Neil Mitchell at 2008-07-10T15:53:00+00:00 Put brackets round operators in more places in the Hoogle output - - - - - 842313aa by Neil Mitchell at 2008-07-10T16:01:25+00:00 Print type signatures with brackets around the name - - - - - cf93deb0 by David Waern at 2008-07-20T17:04:22+00:00 Bump version number to 2.2.0 - - - - - 30e6a8d1 by David Waern at 2008-07-20T17:04:41+00:00 Resolve conflicts in H.B.Hoogle - - - - - 1f0071c9 by David Waern at 2008-07-23T23:05:01+00:00 Add "all" command to runtests.hs that runs all tests despite failures - - - - - f2723023 by David Waern at 2008-07-23T23:08:39+00:00 Update tests/README - - - - - c0304a11 by David Waern at 2008-07-23T23:21:15+00:00 Be compatible with GHC 6.8.3 The cabal file is converted to use the "new" syntax with explicit Library and Executable sections. We define the __GHC_PATCHLEVEL__ symbol using a conditinal cpp-options field in the cabal file. (Ideally, Cabal would define the symbol for us, like it does for __GLASGOW_HASKELL__). We use these symbols to #ifdef around a small difference between 6.8.2 and 6.8.3. Previously, we only supported GHC 6.8.2 officially but the dependencies field said "ghc <= 6.9". This was just for convenience when testing against the (then compatible) HEAD version of GHC, and was left in the release by mistake. Now, we support both GHC 6.8.2 and 6.8.3 and the dependencies field correctly reflects this. - - - - - 88a5fe71 by David Waern at 2008-07-23T23:54:16+00:00 Depend on the currently available ghc-paths versions only - - - - - 8738d97b by David Waern at 2008-07-24T10:50:44+00:00 FIX haskell/haddock#44: Propagate parenthesis level when printing documented types - - - - - 05339119 by David Waern at 2008-07-24T16:06:18+00:00 Drop unnecessary parenthesis in types, put in by the user We were putting in parenthesis were the user did. Let's remove this since it just clutters up the types. The types are readable anyway since we print parens around infix operators and do not rely on fixity levels. When doing this I discovered that we were relying on user parenthesis when printin types like (a `O` b) c. This patchs fixes this problem so that parenthesis are always inserted around an infix op application in case it is applied to further arguments, or if it's an arguments to a type constructor. Tests are updated. - - - - - b3a99828 by David Waern at 2008-07-24T10:19:43+00:00 Print parenthesis around non-atomic banged types Fixes half of haskell/haddock#44 - - - - - ab5238e0 by David Waern at 2008-07-24T22:07:49+00:00 Add a reference file for the TypeFamilies test - - - - - 1941cc11 by David Waern at 2008-07-25T17:15:53+00:00 Simplify definition of pretty and trace_ppr - - - - - e3bfa33c by David Waern at 2008-07-25T17:18:27+00:00 Warning messages Output a warning when filtering out data/type instances and associated types in instances. We don't show these in the documentation yet, and we need to let the user know. - - - - - 9b85fc89 by David Waern at 2008-07-25T17:45:40+00:00 Doc: Mention Hoogle in the Introduction - - - - - afb2dd60 by David Waern at 2008-07-25T17:49:00+00:00 Doc: update -B description - - - - - 584c0c91 by David Waern at 2008-07-25T18:11:38+00:00 Doc: describe -w flag - - - - - 77619c24 by David Waern at 2008-07-28T12:29:07+00:00 Remove TODO from cabal file - - - - - 96717d5f by David Waern at 2008-07-28T12:29:27+00:00 Support type equality predicates - - - - - c2fd2330 by David Waern at 2008-07-29T19:45:14+00:00 Move unL from H.B.Hoogle to H.GHC.Utils I like Neil's shorter unL better than unLoc from the GHC API. - - - - - c4c3bf6a by David Waern at 2008-07-29T19:47:36+00:00 Do not export ATs when not in list of subitems - - - - - bf9a7b85 by David Waern at 2008-08-03T11:42:59+00:00 Filter out ForeignExports - - - - - df59fcb0 by David Waern at 2008-08-03T14:02:51+00:00 Filter out more declarations The previous refactorings in H.I.Create introduced a few bugs. Filtering of some types of declarations that we don't handle was removed. This patch fixes this. - - - - - 2f8a958b by David Waern at 2008-08-03T15:24:07+00:00 Move reL to H.GHC.Utils so we can use it everywhere - - - - - 8ec15efd by David Waern at 2008-08-03T15:25:00+00:00 Use isVanillaLSig from GHC API instead of home brewn function - - - - - 300f93a2 by David Waern at 2008-08-03T15:25:27+00:00 Filter out separately exported ATs This is a quick and dirty hack to get rid of separately exported ATs. We haven't decided how to handle them yet. No warning message is given. - - - - - 8776d1ec by David Waern at 2008-08-03T16:21:21+00:00 Filter out more declarations and keep only vanilla type sigs in classes - - - - - ea07eada by David Waern at 2008-08-03T16:48:00+00:00 Fix layout - - - - - dd5e8199 by David Waern at 2008-08-03T16:50:52+00:00 Move some utility functions from H.I.Create to H.GHC.Utils - - - - - 4a1dbd72 by David Waern at 2008-08-03T17:39:55+00:00 Do not filter out doc declarations - - - - - 0bc8dca4 by David Waern at 2008-08-03T17:47:26+00:00 Filter out separately exported ATs (take two) - - - - - af970fe8 by David Waern at 2008-08-03T22:39:17+00:00 Update CHANGES - - - - - 5436ad24 by David Waern at 2008-08-03T22:40:20+00:00 Bump version number to 2.2.1 - - - - - d66de448 by David Waern at 2008-08-05T19:00:32+00:00 Remove version restriction on ghc-paths - - - - - 534b1364 by David Waern at 2008-08-05T19:04:35+00:00 Bump version to 2.2.2 and update CHANGES - - - - - 549188ff by David Waern at 2008-08-05T19:16:49+00:00 Fix CHANGES - - - - - 0d156bb4 by Luke Plant at 2008-08-11T15:20:59+00:00 invoking haddock clarification and help - - - - - 748295cc by David Waern at 2008-08-11T18:56:37+00:00 Doc: say that the --hoogle option is functional - - - - - 43301db4 by David Waern at 2008-08-05T19:26:08+00:00 Change ghc version dependency to >= 6.8.2 - - - - - 3e5a53b6 by David Waern at 2008-08-10T22:42:05+00:00 Make H.GHC.Utils build with GHC HEAD - - - - - 7568ace0 by David Waern at 2008-08-11T19:41:54+00:00 Import Control.OldException instead of C.Exception when using ghc >= 6.9 We should really test for base version instead, but I don't currently know which version to test for. - - - - - b71ae991 by David Waern at 2008-08-12T22:40:39+00:00 Make our .haddock file version number depend on the GHC version We need to do this, since our .haddock format can potentially change whenever GHC's version changes (even when only the patchlevel changes). - - - - - 6307ce3f by David Waern at 2008-08-12T22:49:57+00:00 Remove matching on NoteTy in AttachInstances, it has been removed - - - - - 2dbcfd5f by David Waern at 2008-08-12T23:02:02+00:00 Comment out H.GHC.loadPackages - it is unused and doesn't build with ghc >= 6.9 - - - - - c74db5c2 by David Waern at 2008-08-12T23:03:58+00:00 Hide <.> from GHC import in Hoogle only for ghc <= 6.8.3 - - - - - 69a44ebb by David Waern at 2008-08-12T23:11:12+00:00 Follow changes to parseDynamic/StaticFlags - - - - - 5881f3f0 by David Waern at 2008-08-13T21:43:58+00:00 Add __GHC_PATCHLEVEL__ symbol also when building the library - - - - - 8574dc11 by David Waern at 2008-08-13T21:44:17+00:00 Follow move of package string functions from PackageConfig to Module - - - - - c9baa77f by David Waern at 2008-08-13T21:45:29+00:00 Follow extensible exceptions changes - - - - - 9092de15 by David Waern at 2008-08-13T21:46:20+00:00 Update test following Haddock version change - - - - - ebe569a4 by David Waern at 2008-08-13T21:46:54+00:00 Follow changes to parseDynamic- parseStaticFlags in GHC - - - - - b8a5ffd3 by David Waern at 2008-08-13T21:47:36+00:00 Follow changes to Binary in GHC 6.9 - - - - - edfda1cc by David Waern at 2008-08-13T21:50:17+00:00 Change ghc version dependency to >= 6.8.2 && <= 6.9 - - - - - d59be1cf by Neil Mitchell at 2008-08-12T16:02:53+00:00 Output all items, even if they are not defined in this module - ensures map comes from Prelude, not just GHC.Base - - - - - dda93b9f by Neil Mitchell at 2008-08-12T21:37:32+00:00 Add support for type synonyms to Hoogle, was accidentally missing before (woops!) - - - - - b6ee795c by Neil Mitchell at 2008-08-13T14:03:24+00:00 Generalise Hoogle.doc and add a docWith - - - - - 415e1bb2 by Neil Mitchell at 2008-08-13T14:03:46+00:00 Make Hoogle add documentation to a package - - - - - 790a1202 by Neil Mitchell at 2008-08-18T12:52:43+00:00 Use the same method to put out signatures as class methods in the Hoogle backend - - - - - ded37eba by Neil Mitchell at 2008-08-18T12:53:04+00:00 Remove Explicit top-level forall's when pretty-printing signatures - - - - - 6468c722 by Neil Mitchell at 2008-08-20T07:59:13+00:00 Simplify the code by removing not-to-important use of <.> in the Hoogle back end - - - - - 788c3a8b by Neil Mitchell at 2008-08-21T18:20:24+00:00 In the hoogle back end, markup definition lists using <i>, not <b> - - - - - 77d4b000 by Ian Lynagh at 2008-08-14T10:49:14+00:00 Add a Makefile for GHC's build system. Still won't work yet, but we're closer - - - - - 920440d7 by Ian Lynagh at 2008-08-27T18:06:46+00:00 Add haddock.wrapper - - - - - bcda925f by Ian Lynagh at 2008-08-27T18:07:02+00:00 Add a manual Cabal flag to control the ghc-paths dependency - - - - - 04d194e2 by Ian Lynagh at 2008-08-27T20:41:27+00:00 Update extensions in Cabal file Use ScopedTypeVariables instead of PatternSignatures - - - - - 12480043 by Ian Lynagh at 2008-08-27T20:41:55+00:00 Increase the upper bound on the GHC version number - - - - - b1f809a5 by Ian Lynagh at 2008-08-27T21:32:22+00:00 Fix some warnings - - - - - aea0453d by Ian Lynagh at 2008-08-28T14:22:29+00:00 Fixes for using haddock in a GHC build tree - - - - - ad23bf86 by Ian Lynagh at 2008-08-28T21:14:27+00:00 Don't use Cabal wrappers on Windows - - - - - 35858e4c by Ian Lynagh at 2008-08-29T00:07:42+00:00 Fix in-tree haddock on Windows - - - - - c2642066 by Ian Lynagh at 2008-09-03T22:35:53+00:00 follow library changes - - - - - 2eb55d50 by Ian Lynagh at 2008-09-07T18:52:51+00:00 bindist fixes - - - - - 3daa5b59 by Ian Lynagh at 2008-09-10T16:58:18+00:00 We need to tell haddock that its datasubdir is . or it can't find package.conf - - - - - 388fd8c2 by Ian Lynagh at 2008-09-10T19:47:44+00:00 Fix haddock inplace on Windows - - - - - 70a641c1 by Ian Lynagh at 2008-09-10T22:15:44+00:00 Fix installed haddock on Windows - - - - - 83c1e997 by Neil Mitchell at 2008-09-11T10:48:55+00:00 Import GHC.Paths if not IN_GHC_TREE, seems to match the use of GHC.Paths functions much better - - - - - b452519b by Ian Lynagh at 2008-09-12T12:58:24+00:00 Add a LANGUAGE ForeignFunctionInterface pragma - - - - - afbd592c by Ian Lynagh at 2008-09-12T12:59:13+00:00 Wibble imports - - - - - 547ac4ad by Ian Lynagh at 2008-09-14T15:34:22+00:00 Add a "#!/bin/sh" to haddock.wrapper - - - - - f207a807 by Ian Lynagh at 2008-09-15T10:02:32+00:00 Use "exec" when calling haddock in the wrapper - - - - - 2ee68509 by Thomas Schilling at 2008-09-15T09:09:16+00:00 Port Haddock.Interface to new GHC API. This required one bigger change: 'readInterfaceFile' used to take an optional 'Session' argument. This was used to optionally update the name cache of an existing GHC session. This does not work with the new GHC API, because an active session requires the function to return a 'GhcMonad' action, but this is not possible if no session is provided. The solution is to use an argument of functions for reading and updating the name cache and to make the function work for any monad that embeds IO, so it's result type can adapt to the calling context. While refactoring, I tried to make the code a little more self-documenting, mostly turning comments into function names. - - - - - 3bb96431 by Thomas Schilling at 2008-09-15T09:09:37+00:00 Reflect GHC API changes. - - - - - 2e60f714 by Thomas Schilling at 2008-09-15T09:10:37+00:00 Port Haddock.GHC.Typecheck to new GHC API. - - - - - 9cfd4cff by Thomas Schilling at 2008-09-15T09:11:00+00:00 Port Haddock.GHC to new GHC API. - - - - - caffa003 by Thomas Schilling at 2008-09-15T09:11:25+00:00 Port Main to new GHC API. - - - - - 069a4608 by Ian Lynagh at 2008-09-21T11:19:00+00:00 Fix paths used on Windows frmo a GHC tree: There is no whare directory - - - - - 7ceee1f7 by Ian Lynagh at 2008-09-21T12:20:16+00:00 Fix the in-tree haddock on Windows - - - - - 0d486514 by Ian Lynagh at 2008-09-23T18:06:58+00:00 Increase the GHC upper bound from 6.11 to 6.13 - - - - - f092c414 by Neil Mitchell at 2008-09-11T14:56:07+00:00 Do not wrap __ in brackets - - - - - 036bdd13 by Ian Lynagh at 2008-09-28T01:42:35+00:00 Fix building haddock when GhcProfiled=YES - - - - - 01434a89 by David Waern at 2008-09-24T20:24:21+00:00 Add PatternSignatures LANGUAGE pragma to Main and Utils - - - - - 1671a750 by David Waern at 2008-10-02T22:57:25+00:00 For source links, get original module from declaration name instead of environment. Getting it from the environment must have been a remnant from the times when we were using unqualified names (versions 0.x). - - - - - a25dde99 by David Waern at 2008-10-02T22:59:57+00:00 Remove ifaceEnv from Interface - it's no longer used - - - - - 610993da by David Waern at 2008-10-02T23:04:58+00:00 Write a comment about source links for type instance declarations - - - - - 5a96b5d5 by Thomas Schilling at 2008-10-03T10:45:08+00:00 Follow GHC API change of parseModule. - - - - - 5a943ae5 by Ian Lynagh at 2008-10-03T15:56:58+00:00 TAG 2008-10-03 - - - - - 76cdd6ae by Thomas Schilling at 2008-10-08T12:29:50+00:00 Only load modules once when typechecking with GHC. This still doesn't fix the memory leak since the typechecked source is retained and then processed separately. To fix the leak, modules must be processed directly after typechecking. - - - - - 7074d251 by David Waern at 2008-10-09T23:53:54+00:00 Interleave typechecking with interface creation At the same time, we fix a bug where the list of interfaces were processed in the wrong order, when building the links and renaming the interfaces. - - - - - 4b9b2b2d by David Waern at 2008-10-09T23:54:49+00:00 Add some strictness annotations in Interface We add some strictness annotations to the fields of Interface, so that less GHC data is hold on to during processing. - - - - - 22035628 by David Waern at 2008-10-10T20:02:31+00:00 Remove typecheckFiles and MonadUtils import from H.GHC.Typeccheck - - - - - be637ad3 by David Waern at 2008-10-10T20:33:38+00:00 Make Haddock build with GHC 6.8.2 - - - - - 523b3404 by David Waern at 2008-10-10T21:08:09+00:00 Fix documentation for createInterfaces - - - - - e1556702 by David Waern at 2008-10-10T21:26:19+00:00 Hide H.Utils in library - - - - - a8e751c3 by David Waern at 2008-10-10T21:34:59+00:00 Add back .haddock file versioning based on GHC version It was accidentally removed in the patch for GHC 6.8.2 compatibility - - - - - 06fb3c01 by David Waern at 2008-10-10T21:47:15+00:00 Bump version number to 2.3.0 - - - - - ff087fce by David Waern at 2008-10-10T22:35:49+00:00 Add support for DocPic The support for DocPic was merged into the GHC source long ago, but the support in Haddock was forgotten. Thanks Peter Gavin for submitting this fix! - - - - - 3af85bf6 by David Waern at 2008-10-10T23:34:05+00:00 Update tests - - - - - 0966873c by Simon Marlow at 2008-10-10T14:43:04+00:00 no need for handleErrMsg now, we don't throw any ErrMsgs - - - - - f1870de3 by Clemens Fruhwirth at 2008-10-10T13:29:36+00:00 Compile with wrapper but remove it for dist-install - - - - - 7b440dc2 by David Waern at 2008-10-11T14:02:25+00:00 Remove interface from LinksInfo It was there to know the documentation home module when creating a wiki link, but we already know this since we have the DocName. - - - - - e5729e6a by David Waern at 2008-10-15T20:49:18+00:00 Wibble - - - - - b2a8e01a by David Waern at 2008-10-15T21:03:36+00:00 Use type synonyms for declarations and docs in H.I.Create - - - - - be71a15b by David Waern at 2008-10-15T21:12:17+00:00 Comment out unused type family stuff completely - - - - - 91aaf075 by David Waern at 2008-10-15T21:49:04+00:00 Wibble - - - - - 42ba4eb4 by David Waern at 2008-10-15T21:53:53+00:00 Move convenient type synonym to H.Types - - - - - db11b723 by David Waern at 2008-10-15T22:14:07+00:00 Add DeclInfo to H.Types - - - - - 193552b6 by David Waern at 2008-10-15T22:15:01+00:00 Add subordinates with docs to the declaration map The only place in the code where we want the subordinates for a declaration is right after having looked up the declaration in the map. And since we include subordinates in the map, we might as well take the opportunity to store those subordinates that belong to a particular declaration together with that declaration. We also store the documentation for each subordinate. - - - - - 31e6eebc by David Waern at 2008-10-16T17:18:47+00:00 Wibble - - - - - 0dcbd79f by David Waern at 2008-10-16T20:58:42+00:00 Fix haskell/haddock#61 We were not getting docs for re-exported class methods. This was because we were looking up the docs in a map made from the declarations in the current module being rendered. Obviously, re-exported class methods come from another module. Class methods and ATs were the only thing we were looking up using the doc map, everything else we found in the ExporItems. So now I've put subordinate docs in the ExportItem's directly, to make things a bit more consistent. To do this, I added subordinates to the the declarations in the declaration map. This was easy since we were computing subordinates anyway, to store stand-alone in the map. I added a new type synonym 'DeclInfo', which is what we call what is now stored in the map. This little refactoring removes duplicate code to retrieve subordinates and documentation from the HsGroup. - - - - - de47f20a by David Waern at 2008-10-16T22:06:35+00:00 Document function and improve its layout - - - - - e74e625a by Thomas Schilling at 2008-10-20T11:12:57+00:00 Force interface more aggressively. For running Haddock on GHC this reduces memory usage by about 50 MB on a 32 bit system. A heap profile shows total memory usage peak at about 100 MB, but actual usage is at around 300 MB even with compacting GC (+RTS -c). - - - - - b63ac9a1 by David Waern at 2008-10-20T20:25:50+00:00 Make renamer consistent Instead of explicitly making some binders Undocumented, treat all names the same way (that is, try to find a Documented name). - - - - - f6de0bb0 by Ian Lynagh at 2008-09-19T00:54:43+00:00 TAG GHC 6.10 fork - - - - - 74599cd0 by David Waern at 2008-10-20T21:13:24+00:00 Do not save hidden modules in the .haddock file We were saving interfaces of all processed modules including those hidden using {-# OPTIONS_HADDOCK hide #-} in the .haddock file. This caused broken links when generating the index for the libraries that come with GHC. This patch excludes modules with hidden documentation when writing .haddock files. It should fix the above problem. - - - - - 7b6742e9 by David Waern at 2008-10-21T19:54:52+00:00 Do not save hidden modules in the .haddock file (also for ghc >= 6.9) When writing the first patch, I forgot to do the fix in both branches of an #if macro. - - - - - b99b1951 by David Waern at 2008-10-22T20:04:18+00:00 Remove subordinate map and its usage It is not needed now that we store subordinate names in the DeclInfo map. - - - - - da97cddc by David Waern at 2008-10-22T20:11:46+00:00 Tidy up code in H.I.Create a little Remove commented out half-done type instance support, and remove DeclWithDoc synonym. - - - - - 6afa76f3 by David Waern at 2008-10-22T21:17:29+00:00 Fix warnings in H.GHC.Utils - - - - - 171ea1e8 by David Waern at 2008-10-22T21:35:04+00:00 Fix warnings in H.Utils - - - - - c8cb3b91 by David Waern at 2008-10-22T21:36:49+00:00 Wibble - - - - - 767fa06a by David Waern at 2008-10-27T19:59:04+00:00 Make named doc comments into ExportDoc instead of ExportDecl Fixes a crash when processing modules without export lists containing named docs. - - - - - e638bbc6 by David Waern at 2008-11-02T22:21:10+00:00 Add HCAR entry - - - - - 92b4ffcf by David Waern at 2008-11-02T22:44:19+00:00 Update CHANGES - - - - - 84d4da6e by David Waern at 2008-11-03T11:25:04+00:00 Add failing test for template haskell crash - - - - - 2a9cd2b1 by David Waern at 2008-11-04T21:13:44+00:00 Add tests/TH.hs - - - - - 8a59348e by David Waern at 2008-11-04T21:30:26+00:00 TAG 2.3.0 - - - - - 54f70d31 by Thomas Schilling at 2008-10-24T17:04:08+00:00 Enable framed view of the HTML documentation. This patch introduces: - A page that displays the documentation in a framed view. The left side will show a full module index. Clicking a module name will show it in the right frame. If Javascript is enabled, the left side is split again to show the modules at the top and a very short synopsis for the module currently displayed on the right. - Code to generate the mini-synopsis for each module and the mini module index ("index-frames.html"). - CSS rules for the mini-synopsis. - A very small amount of javascript to update the mini-synopsis (but only if inside a frame.) Some perhaps controversial things: - Sharing code was very difficult, so there is a small amount of code duplication. - The amount of generated pages has been doubled, since every module now also gets a mini-synopsis. The overhead should not be too much, but I haven't checked. Alternatively, the mini-synopsis could also be generated using Javascript if we properly annotate the actual synopsis. - - - - - 5d7ea5a6 by David Waern at 2008-11-04T23:20:17+00:00 Follow change to ExportDecl in frames code - - - - - 60e16308 by David Waern at 2008-11-04T23:35:26+00:00 Update CHANGES - - - - - d63fd26d by David Waern at 2008-11-04T23:37:43+00:00 Bump version number - - - - - c1660c39 by David Waern at 2008-11-04T23:44:46+00:00 Update CHANGES - - - - - 995ab384 by David Waern at 2008-11-04T23:55:21+00:00 Remove .ref files from tests/output/ - - - - - 1abbbe75 by David Waern at 2008-11-04T23:57:41+00:00 Output version info before running tests - - - - - 649b182f by David Waern at 2008-11-05T22:45:37+00:00 Add ANNOUNCE message - - - - - c36ae0bb by David Waern at 2008-11-05T23:15:35+00:00 Update ANNOUNCE - - - - - 9c4f3d40 by David Waern at 2008-11-05T23:18:30+00:00 Wibble - - - - - 5aac87ce by David Waern at 2008-11-06T21:07:48+00:00 Depend on base 4.* when using GHC >= 6.9, otherwise 3.* - - - - - b9796a74 by David Waern at 2008-11-06T21:13:40+00:00 Bump version to 2.4.1 and update CHANGES - - - - - d4b26baa by David Waern at 2008-11-06T21:26:33+00:00 Depend on base 4.0.* instead of 4.* - - - - - 2cb0903c by David Waern at 2008-11-06T21:46:53+00:00 Fix warnings in H.B.HH and H.B.HH2 - - - - - e568e89a by David Waern at 2008-11-06T21:47:12+00:00 Fix warnings in Haddock.ModuleTree - - - - - 9dc14fbd by David Waern at 2008-11-06T21:47:52+00:00 Fix warnings in Haddock.Version - - - - - 02ac197c by David Waern at 2008-11-06T21:51:31+00:00 Fix warnings in H.InterfaceFile and H.Options - - - - - 63e7439a by David Waern at 2008-11-06T21:59:45+00:00 Fix warnings in H.GHC.Typecheck - - - - - 4bca5b68 by David Waern at 2008-11-08T13:43:42+00:00 Set HscTarget to HscNothing instead of HscAsm There used to be a bug in the GHC API that prevented us from setting this value. - - - - - 07357aec by David Waern at 2008-11-09T22:27:00+00:00 Re-export NameCache and friends from Distribution.Haddock - - - - - ea554b5a by David Waern at 2008-11-09T23:14:10+00:00 Add Haddock.GHC.Utils to other-modules in library - - - - - 74aecfd7 by David Waern at 2008-11-10T01:18:57+00:00 Export DocName in the library - - - - - 241a58b3 by David Waern at 2008-11-10T01:19:18+00:00 Document the functions in H.DocName - - - - - edc2ef1b by David Waern at 2008-11-10T01:20:52+00:00 Export H.DocName in the library - - - - - 4f588d55 by David Waern at 2008-11-10T01:29:14+00:00 Make DocName an instance of NamedThing - - - - - b4647244 by David Waern at 2008-11-15T22:58:18+00:00 Reflect version bump in test suite - - - - - 4bee8ce2 by David Waern at 2008-11-15T22:58:45+00:00 Update tests For unknown reasons, test output for Bug1 and Test has changed for the better. - - - - - 1690e2f9 by David Waern at 2008-11-15T22:59:33+00:00 Store hidden modules in .haddock files We store documentation for an entity in the 'InstalledInterface' of the definition site module, and never in the same structure for a module which re-exports the entity. So when a client of the Haddock library wants to look up some documentation, he/she might need to access a hidden module. But we currently don't store hidden modules in the .haddock files. So we add the hidden modules and the Haddock options to the .haddock files. The options will be used to filter the module list to obtain the visible modules only, which is necessary for generating the contents and index for installed packages. - - - - - 8add6435 by David Waern at 2008-11-16T14:35:50+00:00 Bump major version number due to .haddock file format change - - - - - 48bfcf82 by David Waern at 2008-11-23T14:32:52+00:00 Update tests to account for version number bump - - - - - 0bbd1738 by David Waern at 2008-11-23T14:33:31+00:00 HADDOCK_DATA_DIR changed to haddock_datadir - - - - - 5088b78c by David Waern at 2008-11-23T17:13:21+00:00 FIX haskell/haddock#45: generate two anchors for each name We generate two anchor tags for each name, one where we don't escape the name and one where we URI-encode it. This is for compatibility between IE and Opera. Test output is updated. - - - - - 5ee5ca3b by Neil Mitchell at 2008-11-27T14:38:11+00:00 Drop HsDocTy annotations, they mess up pretty printing and also have a bracketing bug (#2584) - - - - - 51c014e9 by Roman Cheplyaka at 2008-11-27T22:27:36+00:00 Allow referring to a specific section within a module in a module link Fixes haskell/haddock#65 - - - - - 4094bdc5 by David Waern at 2008-11-28T21:13:33+00:00 Update tests following anchor change - - - - - f89552dd by Thomas Schilling at 2008-11-29T16:16:20+00:00 Haddock really shouldn't try to overwrite files. - - - - - 98127499 by David Waern at 2008-12-07T14:09:15+00:00 Solve conflict - - - - - 319356c5 by David Waern at 2008-10-22T21:16:55+00:00 Add -Wall -Werror to ghc-options - - - - - 3c4968c9 by David Waern at 2008-11-04T23:38:56+00:00 TAG 2.4.0 - - - - - 4b21e003 by David Waern at 2008-11-06T21:14:04+00:00 TAG 2.4.1 - - - - - 8e0cad5c by David Waern at 2008-12-07T14:12:54+00:00 Remove -Werror - - - - - 299d6deb by David Waern at 2008-12-07T14:25:18+00:00 Remove -Wall, we'll focus on warnings after 6.10.2 is out - - - - - 5f4216b6 by David Waern at 2008-12-07T20:58:05+00:00 Resolve conflict properly - - - - - 67d774e7 by Neil Mitchell at 2008-12-15T11:44:26+00:00 Make forall's in constructors explicit, i.e. data Foo = Foo {foo :: Eq a => a} - - - - - 61851792 by Neil Mitchell at 2008-12-18T15:39:39+00:00 Try and find a better package name than "main" for Hoogle, goes wrong when working on an executable rather than a library - - - - - 2fab8554 by David Waern at 2008-12-08T23:19:48+00:00 Make visible names from ExportItems Instead of a complicated calculation of visible names out of GHC's export items, we can get them straight out of the already calculated ExportItems. The ExportItems should represent exactly those items that are visible in an interface. If store all the exported sub-names in ExportDecl instead of only those with documentation, the calculation becomes very simple. So we do this change as well (should perhaps have been a separate patch). This should fix the problem with names from ghc-prim not appearing in the link environment. - - - - - 7caadd8c by Ian Lynagh at 2008-12-11T17:01:04+00:00 Wrap the GHC usage with defaultCleanupHandler This fixes a bug where haddock leaves /tmp/ghc* directories uncleaned. - - - - - 7c9fc9a5 by David Waern at 2009-01-02T21:38:27+00:00 Show re-exported names from external packages again This fixes GHC ticket 2746. In order to also link to the exported subordinate names of a declaration, we need to re-introduce the sub map in the .haddock files. - - - - - 119e4e05 by David Waern at 2009-01-06T23:34:17+00:00 Do not process boot modules We should of course not try to produce documentation for boot modules! The reason this has worked in the past is that the output of "real" modules overwrites the output of boot modules later in the process. However, this causes a subtle link environment problem. So let's get rid of this stupid behaviour. We avoid processing boot modules, but we continue to typecheck them. - - - - - c285b9d2 by David Waern at 2009-01-08T18:03:36+00:00 Export modules also when coming from external packages This seems to have regressed since a refactoring that was part of the 2.3.0 release. - - - - - 24031c17 by David Waern at 2009-01-10T15:26:26+00:00 Change version to 2.4.2 - no need to go to 2.5.0 - - - - - 864d1c3f by David Waern at 2009-01-10T15:35:20+00:00 Update tests to account for version number change - - - - - 524ba886 by David Waern at 2009-01-10T18:29:17+00:00 Add test for Template Haskell splicing - - - - - 05e6e003 by David Waern at 2009-01-10T19:35:42+00:00 Fix Trac haskell/haddock#68: Turn on compilation via C for Template Haskell packages We can't use HscNothing if we need to run code coming from modules inside the processed package during typechecking, which is the case for some packages using Template Haskell. This could be improved, to e.g. use HscInterpreted and HscNothing where possible, instead of using HscC for all modules in the package. - - - - - 2b2bafa1 by David Waern at 2009-01-10T20:22:25+00:00 Only use needsTemplateHaskell when compiling with GHC 6.10.2 or above - - - - - bedc3a93 by Ian Lynagh at 2009-01-11T14:58:41+00:00 Fix the location of INPLACE_PKG_CONF; fixes the build Spotted by Conal Elliott - - - - - 943107c8 by David Waern at 2009-01-20T19:27:39+00:00 Document H.I.Create.collectDocs better - - - - - c6252e37 by David Waern at 2009-01-20T19:29:51+00:00 Fix Trac haskell/haddock#59: TH-generated declarations disappearing This patch was contributed by Joachim Breitner (nomeata). - - - - - 3568a6af by David Waern at 2009-01-21T21:41:48+00:00 Do not indicate that a constructor argument is unboxed We only show the strictness annotation for an unboxed constructor argument. The fact that it is unboxed is an implementation detail and should not be part of the module interface. - - - - - 562a4523 by David Waern at 2009-01-22T18:53:49+00:00 Fix Trac haskell/haddock#50: do not attach docs to pragmas or other kinds of non-declarations We now filter out everything that is not a proper Haskell declaration before collecting the docs and attaching them to declarations. - - - - - 6fdf21c2 by David Waern at 2009-01-22T19:48:09+00:00 Add test for quasi quotation. No reference output yet. - - - - - dc4100fd by David Waern at 2009-01-22T19:57:47+00:00 Improve quasi-quotation test and add reference output - - - - - 908b74bb by David Waern at 2009-01-23T23:22:03+00:00 Filter out separately exported associated types in a smarter way - - - - - f6b42ecb by David Waern at 2009-01-24T16:54:39+00:00 Correct spelling mistake in error message - - - - - 24e4245d by David Waern at 2009-01-24T17:48:03+00:00 Correct comment - - - - - b5e8462f by David Waern at 2009-02-07T13:22:29+00:00 Do not show a subordinate at the top level if its parent is also exported See note in the source code for more info. - - - - - 4b09de57 by David Waern at 2009-02-07T13:53:53+00:00 Update test following change to top level subordinates - - - - - 76379896 by David Waern at 2009-02-07T13:58:04+00:00 Remove html files in the tests/output/ directory which have been accidentally added - - - - - 1a6d8b10 by Joachim Breitner at 2009-02-20T10:29:43+00:00 Typo in comment - - - - - fec367d0 by David Waern at 2009-02-24T20:21:17+00:00 Fix small bug The rule is to prefer type constructors to other things when an identifier in a doc string can refer to multiple things. This stopped working with newer GHC versions (due to a tiny change in the GHC renamer). We implement this rule in the HTML backend for now, instead of fixing it in GHC, since we will move renaming of doc strings to Haddock in the future anyway. - - - - - 9b4172eb by David Waern at 2009-02-25T20:04:38+00:00 Fix bad error handling with newer GHCs When support for GHC 6.10 was added, an error handler was installed only around the typechecking phase. This had the effect that errors thrown during dependency chasing were caught in the top-level exception handler and not printed with enough detail. With this patch we wrap the error handler around all our usage of the Ghc monad. - - - - - de2df363 by Simon Peyton Jones at 2009-02-02T16:47:42+00:00 Hide funTyConName, now exported by TypeRep - - - - - 4d40a29f by Ian Lynagh at 2009-02-12T18:57:49+00:00 Don't build the library when building in the GHC tree - - - - - 1cd0abe4 by Ian Lynagh at 2009-02-13T13:58:53+00:00 Add a ghc.mk - - - - - 3d814eeb by Ian Lynagh at 2009-02-13T18:50:28+00:00 do .depend generation for haddock with the stage1 compiler This is a bit of a hack. We mkdepend with stage1 as if .depend depends on the stage2 compiler then make goes wrong: haddock's .depend gets included, which means that make won't reload until it's built, but we can't build it without the stage2 compiler. We therefore build the stage2 compiler before its .depend file is available, and so compilation fails. - - - - - b55036a4 by Ian Lynagh at 2009-02-25T01:38:13+00:00 Give haddock a wrapper on unix in the new GHC build system - - - - - 9eabfe68 by Ian Lynagh at 2009-02-25T19:21:32+00:00 Create inplace/lib/html in the new GHC build system - - - - - 93af30c7 by Ian Lynagh at 2008-11-07T19:18:23+00:00 TAG GHC 6.10.1 release - - - - - 06e6e34a by Thomas Schilling at 2009-02-24T18:11:00+00:00 Define __GHC_PATCHLEVEL__ for recent version of GHC (stable). - - - - - 680e6ed8 by Thomas Schilling at 2009-02-24T18:12:26+00:00 'needsTemplateHaskell' is not defined in current stable GHC. - - - - - 6c5619df by David Waern at 2009-02-25T22:15:23+00:00 Hide fynTyConName only for recent GHC versions - - - - - 6b2344f1 by Ian Lynagh at 2009-02-26T00:49:56+00:00 Add the module to one of haddocks warnings - - - - - e5d11c70 by David Waern at 2009-02-27T21:37:20+00:00 Bug fix We tried to filter out subordinates that were already exported through their parent. This didn't work properly since we were in some cases looking at the grand-parent and not the parent. We now properly compute all the parent-child relations of a declaration, and use this information to get the parent of a subordinate. We also didn't consider record fields with multiple parents. This is now handled correctly. We don't currently support separately exported associated types. But when we do, they should be handled correctly by this process too. Also slightly improved the warning message that we give when filtering out subordinates. - - - - - 10a79a60 by David Waern at 2009-02-27T22:08:08+00:00 Fix error message conflict The module name is already written in the beginning of the message, as seems to be the convention in Haddock. Perhaps not so clear, but we should change it everywhere in that case. Leaving it as it is for now. - - - - - c5055c7f by David Waern at 2009-02-27T22:15:17+00:00 Shorten warning message - - - - - a72fed3a by David Waern at 2009-02-28T00:53:55+00:00 Do not show package name in warning message - - - - - a5daccb2 by Ian Lynagh at 2009-03-01T14:59:35+00:00 Install haddock in the new GHC build system - - - - - dfdb025c by Ian Lynagh at 2009-03-07T23:56:29+00:00 Relax base dependency to < 4.2, not < 4.1 - - - - - 5769c8b4 by David Waern at 2009-03-21T14:58:52+00:00 Bump .haddock file version number (due to change of format) - - - - - f1b8f67b by David Waern at 2009-03-21T14:59:26+00:00 Define __GHC_PATCHLEVEL__=1 when using ghc-6.10.1 - - - - - 23f78831 by David Waern at 2009-03-21T16:40:52+00:00 Update CHANGES - - - - - 7d2735e9 by David Waern at 2009-03-21T16:50:33+00:00 Update ANNOUNCE - - - - - 0771e00a by David Waern at 2009-03-21T16:54:40+00:00 Update ANNOUNCE, again - - - - - 81a6942a by David Waern at 2009-03-21T17:50:06+00:00 Don't be too verbose in CHANGES - - - - - 29861dcf by David Waern at 2009-03-21T18:03:31+00:00 TAG 2.4.2 - - - - - a585f285 by David Waern at 2009-03-21T19:20:29+00:00 Require Cabal >= 1.2.3 - - - - - 7c611662 by David Waern at 2009-03-21T19:21:48+00:00 TAG 2.4.2 with cabal-version >= 1.2.3 - - - - - 23b7deff by Simon Marlow at 2009-03-20T15:43:42+00:00 new GHC build system: use shell-wrappers macro - - - - - 25f8afe7 by Ian Lynagh at 2009-03-21T19:13:53+00:00 Fix (with a hack?) haddock in teh new build system - - - - - 6a29a37e by David Waern at 2009-03-24T22:10:15+00:00 Remove unnecessary LANGUAGE pragma - - - - - 954da57d by David Waern at 2009-03-24T22:21:23+00:00 Fix warnings in H.B.DevHelp - - - - - 1619f1df by David Waern at 2009-03-26T23:20:44+00:00 -Wall police in H.B.Html - - - - - b211e13b by Simon Marlow at 2009-03-24T13:00:56+00:00 install Haddock's html stuff - - - - - 78e0b107 by David Waern at 2008-12-07T19:58:53+00:00 Add verbosity flag and utils, remove "verbose" flag - - - - - 913dae06 by David Waern at 2008-12-07T20:01:05+00:00 Add some basic "verbose" mode logging in H.Interface - - - - - 1cbff3bf by David Waern at 2009-03-27T00:07:26+00:00 Fix conflicts - - - - - 22f82032 by David Waern at 2009-03-27T21:15:11+00:00 Remove H.GHC.Typecheck - - - - - 81557804 by David Waern at 2009-03-27T21:19:22+00:00 Remove docNameOrig and use getName everywhere instead - - - - - d8267213 by David Waern at 2009-03-27T21:21:46+00:00 Use docNameOcc instead of nameOccName . getName - - - - - 5d55deab by David Waern at 2009-03-27T21:33:04+00:00 Remove H.DocName and put DocName in H.Types - - - - - 8ba72611 by David Waern at 2009-03-27T22:06:26+00:00 Document DocName - - - - - 605f8ca5 by David Waern at 2009-03-27T22:45:21+00:00 -Wall police - - - - - e4da93ae by David Waern at 2009-03-27T23:12:53+00:00 -Wall police in H.B.Hoogle - - - - - bb255519 by David Waern at 2009-03-27T23:41:28+00:00 Define Foldable and Traversable instances for Located - - - - - f1195cfe by David Waern at 2009-03-27T23:51:34+00:00 Wibble - - - - - 23818d7c by David Waern at 2009-03-28T00:03:55+00:00 -Wall police in H.I.Rename - - - - - 0f050d67 by David Waern at 2009-03-28T00:15:15+00:00 -Wall police in H.I.AttachInstances - - - - - 0f3fe038 by David Waern at 2009-03-28T21:09:41+00:00 Wibble - - - - - 275d4865 by David Waern at 2009-03-28T21:27:06+00:00 Layout fix - - - - - 54ff0ef8 by David Waern at 2009-03-28T21:59:07+00:00 -Wall police in H.I.Create - - - - - 7f58b117 by David Waern at 2009-03-28T22:10:19+00:00 -Wall police in H.Interface - - - - - f0c03b44 by David Waern at 2009-03-28T22:22:59+00:00 -Wall police in Main - - - - - 29da355c by David Waern at 2009-03-28T22:23:39+00:00 Turn on -Wall -Werror - - - - - 446d3060 by David Waern at 2009-04-01T20:40:30+00:00 hlint police - - - - - 3867c9fc by David Waern at 2009-04-01T20:48:42+00:00 hlint police - - - - - bd1f1600 by David Waern at 2009-04-01T20:58:02+00:00 hlint police - - - - - e0e90866 by David Waern at 2009-04-05T12:42:53+00:00 Move H.GHC.Utils to H.GhcUtils - - - - - 9cbd426b by David Waern at 2009-04-05T12:57:21+00:00 Remove Haddock.GHC and move its (small) contents to Main - - - - - b5c2cbfd by David Waern at 2009-04-05T13:07:04+00:00 Fix whitespace and stylistic issues in Main - - - - - 3c04aa56 by porges at 2008-12-07T08:22:19+00:00 add unicode output - - - - - 607918da by David Waern at 2009-04-26T15:09:43+00:00 Resolve conflict - - - - - 4bec6b6b by Simon Marlow at 2009-05-13T10:00:31+00:00 fix markup - - - - - 436ad6f4 by Simon Marlow at 2009-03-23T11:54:45+00:00 clean up - - - - - bdcd1398 by Simon Marlow at 2009-03-24T10:36:45+00:00 new GHC build system: add $(exeext) - - - - - 9c0972f3 by Simon Marlow at 2009-03-24T11:04:31+00:00 update for new GHC build system layout - - - - - d0f3f83a by Ian Lynagh at 2009-03-29T15:31:43+00:00 GHC new build system fixes - - - - - 5a8245c2 by Ian Lynagh at 2009-04-04T20:44:23+00:00 Tweak new build system - - - - - 9c6f2d7b by Simon Marlow at 2009-05-13T10:01:27+00:00 add build instructions for GHC - - - - - 66d07c76 by Ian Lynagh at 2009-05-31T00:37:53+00:00 Quote program paths in ghc.mk - - - - - bb7de2cd by Ian Lynagh at 2009-06-03T22:57:55+00:00 Use a bang pattern on an unlifted binding - - - - - 3ad283fc by Ian Lynagh at 2009-06-13T16:17:50+00:00 Include haddock in GHC bindists - - - - - ac447ff4 by David Waern at 2009-06-24T21:07:50+00:00 Delete Haddock.Exception and move contents to Haddock.Types Only a few lines of code that mainly declares a type - why not just put it in Haddock.Types. - - - - - 4464fb9b by David Waern at 2009-06-24T22:23:23+00:00 Add Haddock module headers Add a proper Haddock module header to each module, with a more finegrained copyright. If you feel mis-accreditted, please correct any copyright notice! The maintainer field is set to haddock at projects.haskell.org. Next step is to add a brief description to each module. - - - - - 5f4c95dd by David Waern at 2009-06-24T22:39:44+00:00 Fix spelling error - - - - - 6d074cdb by David Waern at 2009-06-25T21:53:56+00:00 Document Interface and InstalledInterface better - - - - - d0cbd183 by David Waern at 2009-06-27T12:46:46+00:00 Remove misplaced whitespace in H.I.Rename - - - - - fa381c49 by David Waern at 2009-06-27T13:26:03+00:00 Fix haskell/haddock#104 - create output directory if missing - - - - - 91fb77ae by Ian Lynagh at 2009-06-25T15:59:50+00:00 TAG 2009-06-25 - - - - - 0d853f40 by Simon Peyton Jones at 2009-07-02T15:35:22+00:00 Follow extra field in ConDecl - - - - - b201735d by Ian Lynagh at 2009-07-05T16:50:35+00:00 Update Makefile for the new GHC build system - - - - - df6c0092 by Ian Lynagh at 2009-07-05T17:01:13+00:00 Resolve conflicts - - - - - 1066870a by Ian Lynagh at 2009-07-05T17:01:48+00:00 Remove the -Wwarn hack in the GHC build system - - - - - 7e856076 by Ian Lynagh at 2009-07-05T17:17:59+00:00 Fix warnings - - - - - 5d4cd958 by Ian Lynagh at 2009-07-05T19:35:40+00:00 Bump version number Cabal needs to distinguish between haddocks having a --verbose and --verbosity flag - - - - - 6ee07c99 by David Waern at 2009-07-06T20:14:57+00:00 Wibble - - - - - 2308b66f by David Waern at 2009-07-06T20:24:20+00:00 Clearer printing of versions by runtests.hs - - - - - d4b5d9ab by David Waern at 2009-07-06T21:22:42+00:00 Fix (invisible) bug introduced by unicode patch - - - - - 2caca8d8 by David Waern at 2009-07-06T21:44:10+00:00 Use HscAsm instead of HscC when using TH - - - - - 18f3b755 by David Waern at 2009-07-06T22:10:22+00:00 Update HCAR entry (by Janis) - - - - - a72ac9db by David Waern at 2009-07-06T23:01:35+00:00 Follow HsRecTy change with an #if __GLASGOW_HASKEL__ >= 611 - - - - - 549135d2 by David Waern at 2009-07-06T23:11:41+00:00 Remove unused functions from Haddock.Utils - - - - - b450134a by Isaac Dupree at 2009-07-11T14:59:00+00:00 revert to split-index for large indices - remove the search-box, because browsers have search-for-text abilities anyway. - pick 150 items in index as the arbitrary time at which to split it - notice the bug that identifiers starting with non-ASCII characters won't be listed in split-index, but don't bother to fix it yet (see ticket haskell/haddock#116, http://trac.haskell.org/haddock/ticket/116 ) - - - - - 78a5661e by Isaac Dupree at 2009-07-20T15:37:18+00:00 Implement GADT records in HTML backend - - - - - 4e163555 by Isaac Dupree at 2009-07-21T22:03:25+00:00 add test for GADT records - - - - - 79aa4d6e by David Waern at 2009-07-23T20:40:37+00:00 Update test suite following version bump - - - - - 5932c011 by David Waern at 2009-08-02T10:25:39+00:00 Fix documentation bug - - - - - a6970fca by David Waern at 2009-08-12T23:08:53+00:00 Remove support for ghc 6.8.* from .cabal file - - - - - c1695902 by Ian Lynagh at 2009-07-07T13:35:45+00:00 Fix unused import warnings - - - - - fb6df7f9 by Ian Lynagh at 2009-07-16T00:20:31+00:00 Use cProjectVersion directly rather than going through compilerInfo Fixes the build after changes in GHC - - - - - 548cdd66 by Simon Marlow at 2009-07-28T14:27:04+00:00 follow changes in GHC's ForeignType - - - - - 9395aaa0 by David Waern at 2009-08-13T22:17:33+00:00 Switch from PatternSignatures to ScopedTypeVariables in Main - - - - - eebf39bd by David Waern at 2009-08-14T17:14:28+00:00 Version .haddock files made with GHC 6.10.3/4 correclty - - - - - 58f3e735 by David Waern at 2009-08-14T17:19:37+00:00 Support GHC 6.10.* and 6.11.* only - - - - - 5f63cecc by David Waern at 2009-08-14T22:03:20+00:00 Do not version .haddock file based on GHC patchlevel version We require that the instances of Binary that we use from GHC will not change between patchlevel versions. - - - - - d519de9f by David Waern at 2009-08-14T23:50:00+00:00 Update CHANGES - - - - - 35dccf5c by David Waern at 2009-08-14T23:51:38+00:00 Update version number everywhere - - - - - 6d363fea by David Waern at 2009-08-15T09:46:49+00:00 Update ANNOUNCE - - - - - c7ee6bc2 by David Waern at 2009-08-15T09:47:13+00:00 Remove -Werror Forgot that Hackage doesn't like it. - - - - - a125c12b by David Waern at 2009-08-15T09:49:50+00:00 Require Cabal >= 1.6 - - - - - adb2f560 by Isaac Dupree at 2009-08-12T03:47:14+00:00 Cross-Package Documentation version 4 - - - - - 3d6dc04d by David Waern at 2009-08-15T23:42:57+00:00 Put all the IN_GHC_TREE stuff inside getGhcLibDir - - - - - 56624097 by David Waern at 2009-08-15T23:52:03+00:00 Add --print-ghc-libdir - - - - - f15d3ccb by David Waern at 2009-08-16T00:37:52+00:00 Read base.haddock when running tests We can now test cross-package docs. - - - - - 283f0fb9 by David Waern at 2009-08-16T00:50:59+00:00 Update test output - we now have more links - - - - - 673d1004 by David Waern at 2009-08-16T01:26:08+00:00 Read process.haddock when running tests - - - - - 0d127f82 by David Waern at 2009-08-16T01:43:04+00:00 Add a test for cross-package documentation - - - - - f94db967 by Ian Lynagh at 2009-08-16T18:42:44+00:00 Follow GHC build system changes - - - - - 5151278a by Isaac Dupree at 2009-08-16T19:58:05+00:00 make cross-package list types look nicer - - - - - c41e8228 by Isaac Dupree at 2009-08-18T01:47:47+00:00 Haddock.Convert: export more functions This lets us remove some code in Haddock.Interface.AttachInstances - - - - - 2e5fa398 by Isaac Dupree at 2009-08-18T02:11:05+00:00 switch AttachInstances to use synify code It changed an instance from showing ((,) a b) to (a, b) because my synify code is more sophisticated; I hope the latter is a good thing rather than a bad thing aesthetically, here. But this definitely reduces code duplication! - - - - - b8b07123 by Isaac Dupree at 2009-08-18T02:23:31+00:00 Find instances using GHC, which is more complete. In particular, it works cross-package. An intermediate patch also moved the instance-finding into createInterface, but that move turned out not to be necessary, so if we want to do that, it'd go in a separate patch. (Is that possible? Or will we need GHC to have loaded all the modules first, before we can go searching for the instances (e.g. if the modules are recursive or something)?) - - - - - 6959b451 by Isaac Dupree at 2009-08-17T00:37:18+00:00 fix preprocessor conditional sense - - - - - 942823af by Isaac Dupree at 2009-08-16T22:46:48+00:00 remove ghc 6.8 conditionals from Haddock.Interface - - - - - 4b3ad888 by Isaac Dupree at 2009-08-18T20:24:38+00:00 Fix GHC 6.11 build in Haddock.Convert - - - - - 0a89c5ab by Isaac Dupree at 2009-08-23T00:08:58+00:00 hacks to make it compile without fnArgDocsn - - - - - 7b3bed43 by Isaac Dupree at 2009-08-23T03:01:28+00:00 less big-Map-based proper extraction of constructor subdocs - - - - - b21c279a by Isaac Dupree at 2009-08-23T03:02:06+00:00 Html: remove unnecessary+troublesome GHC. qualifications - - - - - 96c97115 by Isaac Dupree at 2009-08-23T03:08:03+00:00 Move doc parsing/lexing into Haddock for ghc>=6.11 - - - - - e1cec02d by Isaac Dupree at 2009-08-23T05:08:14+00:00 get rid of unused DocMap parameter in Html - - - - - 66960c59 by Isaac Dupree at 2009-08-23T05:54:20+00:00 fix horrible named-docs-disappearing bug :-) - - - - - a9d7eff3 by Isaac Dupree at 2009-08-23T06:26:36+00:00 re-implement function-argument docs ..on top of the lexParseRn work. This patch doesn't change the InstalledInterface format, and thus, it does not work cross-package, but that will be easy to add subsequently. - - - - - 8bf6852c by Isaac Dupree at 2009-08-23T07:26:05+00:00 cross-package fnArgDocs. WARNING: changes .haddock binary format While breaking the format, I took the opportunity to unrename the DocMap that's saved to disk, because there's really no reason that we want to know what *another* package's favorite place to link a Name to was. (Is that true? Or might we want to know, someday?) Also, I added instance Binary Map in InterfaceFile. It makes the code a little simpler without changing anything of substance. Also it lets us add another Map hidden inside another Map (fnArgsDocs in instDocMap) without having really-convoluted serialization code. Instances are neat! I don't understand why this change to InterfaceFile seemed to subtly break binary compatibility all by itself, but no matter, I'll just roll it into the greater format-changing patch. Done! - - - - - 30115a64 by Isaac Dupree at 2009-08-23T18:22:47+00:00 Improve behavior for unfindable .haddock - - - - - aa364bda by Isaac Dupree at 2009-08-23T18:28:16+00:00 add comment for FnArgsDoc type - - - - - 49b23a99 by Isaac Dupree at 2009-08-23T21:52:48+00:00 bugfix: restore fnArgDocs for type-synonyms - - - - - f65f9467 by Isaac Dupree at 2009-08-23T22:06:55+00:00 Backends.Hoogle: eliminate warnings - - - - - a292d216 by Isaac Dupree at 2009-08-23T22:10:24+00:00 Haddock.Convert: eliminate warnings - - - - - 5546cd20 by Isaac Dupree at 2009-08-23T22:12:31+00:00 Haddock.Interface.Rename: eliminate warnings - - - - - 0a9798b6 by Isaac Dupree at 2009-08-23T22:18:47+00:00 Main.hs: remove ghc<6.9 conditionals - - - - - e8f9867f by Isaac Dupree at 2009-08-23T22:27:46+00:00 Main.hs: eliminate warnings (except for OldException) - - - - - 61c64247 by Isaac Dupree at 2009-08-23T22:41:01+00:00 move get*LibDir code in Main.hs, to +consistent code, -duplication - - - - - 948f1e69 by Isaac Dupree at 2009-08-23T23:14:26+00:00 Main.hs: OldException->Exception: which eliminates warnings - - - - - 3d5d5e03 by Isaac Dupree at 2009-08-23T23:20:11+00:00 GhcUtils: ghc >= 6.10 - - - - - 2771d657 by Isaac Dupree at 2009-08-23T23:21:55+00:00 InterfaceFile: ghc >= 6.10 - - - - - d9f2b9d1 by Isaac Dupree at 2009-08-23T23:22:58+00:00 Types: ghc >= 6.10 - - - - - ca39210e by Isaac Dupree at 2009-08-23T23:23:26+00:00 ModuleTree: ghc >= 6.10 - - - - - 883c4e59 by Isaac Dupree at 2009-08-23T23:24:04+00:00 Backends.DevHelp: ghc >= 6.10 - - - - - 04667df5 by Isaac Dupree at 2009-08-23T23:24:37+00:00 Backends.Html: ghc >= 6.10 - - - - - a9f7f25f by Isaac Dupree at 2009-08-23T23:25:24+00:00 Utils: ghc >= 6.10 - - - - - b7105022 by Isaac Dupree at 2009-08-23T23:37:47+00:00 eliminate haskell98 dependency, following GHC's example It turns out I/we already had, and it was only a matter of deleting it from the cabal file. - - - - - 292e0911 by Isaac Dupree at 2009-08-24T01:22:44+00:00 refactor out subordinatesWithNoDocs dep of inferenced-decls fix - - - - - c2ed46a2 by Isaac Dupree at 2009-08-24T01:24:03+00:00 Eradicate wrong runtime warning for type-inferenced exported-functions see the long comment in the patch for why I did it this way :-) - - - - - 4ac0b57c by David Waern at 2009-09-04T22:56:20+00:00 Clean up tyThingToHsSynSig a little Factor out noLoc and use the case construct. Also rename the function to tyThingToLHsDecl, since it doesn't just create type signatures. - - - - - 28ab9201 by David Waern at 2009-09-04T22:58:50+00:00 Wibble - - - - - 0d9fe6d0 by David Waern at 2009-09-06T18:39:30+00:00 Add more copyright owners to H.I.AttachInstances - - - - - 122441b1 by David Waern at 2009-09-06T18:44:12+00:00 Style police - - - - - 1fa79463 by David Waern at 2009-09-06T18:57:45+00:00 Move toHsInstHead to Haddock.Convert and call it synifyInstHead - - - - - 0d42a8aa by David Waern at 2009-09-06T21:11:38+00:00 Use colordiff to display test results if available - - - - - ea9d8e03 by Simon Marlow at 2009-08-24T08:46:14+00:00 Follow changes in GHC's interface file format Word32 instead of Int for FastString and Name offsets - - - - - 537e051e by Simon Marlow at 2009-07-29T14:16:53+00:00 define unpackPackageId (it was removed from GHC) - - - - - 50c63aa7 by David Waern at 2009-09-09T23:18:03+00:00 Remove commented-out code - - - - - 511631fe by David Waern at 2009-09-09T23:19:05+00:00 Correct copyright in H.I.ParseModuleHeader - - - - - 898ec768 by David Waern at 2009-09-11T11:22:29+00:00 Use Map.fromList/toList intead of fromAscList/toAscList when serializing Maps This fixes the missing docs problem. The Eq and Ord instances for Name uses the unique number in Name. This number is created at deserialization time by GHC's magic Binary instance for Name, and it is random. Thus, fromAscList can't be used at deserialization time, even though toAscList was used at serialization time. - - - - - 37bec0d5 by Simon Peyton Jones at 2009-09-11T08:28:04+00:00 Track change in HsType - - - - - eb3a97c3 by Ian Lynagh at 2009-09-11T16:07:09+00:00 Allow building with base 4.2 - - - - - bb4205ed by Ian Lynagh at 2009-09-22T13:50:02+00:00 Loosen the GHC dependency - - - - - 5c75deb2 by Ian Lynagh at 2009-09-22T14:08:39+00:00 Fix building with GHC >= 6.12 - - - - - fb131481 by David Waern at 2009-09-11T11:24:48+00:00 Update runtests.hs to work with GHC 6.11 - - - - - ac3a419d by David Waern at 2009-09-11T11:25:14+00:00 Update CrossPackageDocs test - - - - - ec65c3c6 by David Waern at 2009-09-11T11:25:40+00:00 Add reference output for CrossPackageDocs - - - - - 520c2758 by Ian Lynagh at 2009-10-25T17:26:40+00:00 Fix installation in the GHC build system - - - - - 28b3d7df by Ian Lynagh at 2009-11-05T15:57:27+00:00 GHC build system: Make *nix installation work in paths containing spaces - - - - - 5c9bb541 by David Waern at 2009-11-14T11:56:39+00:00 Track change in HsType for the right compiler version - - - - - 905097ce by David Waern at 2009-11-14T12:10:47+00:00 hlint police - - - - - 04920630 by Ian Lynagh at 2009-11-20T13:46:30+00:00 Use defaultObjectTarget rather than HscAsm This fixes haddock when we don't have a native code generator - - - - - 966eb079 by David Waern at 2009-11-15T12:32:21+00:00 Remove commented-out code - - - - - 37f00fc4 by David Waern at 2009-11-22T13:58:48+00:00 Make runtests.hs strip links before diffing Generates easier to read diffs when tests fail. The content of the links is not important anyway since it is not taken into account by the tests. - - - - - 3a9bb8ef by David Waern at 2009-11-22T14:05:06+00:00 Follow findProgramOnPath signature change in runtests.hs - - - - - b26b9e5a by David Waern at 2009-11-22T14:08:40+00:00 Follow removal of GHC.MVar from base in CrossPackageDocs - - - - - f4d90ae4 by David Waern at 2009-11-22T14:48:47+00:00 Make copy.hs strip link contents before copying No more updating of reference files when URLs in links changes. - - - - - 4c9c420d by David Waern at 2009-11-22T15:26:41+00:00 Update test reference output * More links (Int, Float etc) * Stripped link contents - - - - - a62b80e3 by David Waern at 2009-11-23T23:19:39+00:00 Update CrossPackageDocs reference output - Remove GHC.MVar import (removed from base) - Strip link contents - - - - - 43491394 by David Waern at 2009-11-23T23:20:00+00:00 Update test reference files with comments on instances - - - - - 0d370a0b by David Waern at 2009-11-23T23:25:16+00:00 Bump version number - - - - - 2293113e by David Waern at 2009-11-24T20:55:49+00:00 Comments on instances Implementing this was a little trickier than I thought, since we need to match up instances from the renamed syntax with instances represented by InstEnv.Instance. This is due to the current design of Haddock, which matches comments with declarations from the renamed syntax, while getting the list of instances of a class/family directly using the GHC API. - Works for class instances only (Haddock has no support for type family instances yet) - The comments are rendered to the right of the instance head in the HTML output - No change to the .haddock file format - Works for normal user-written instances only. No comments are added on derived or TH-generated instances - - - - - bf586f29 by David Waern at 2009-11-27T22:05:15+00:00 Whitespace police - - - - - b8f03afa by David Waern at 2009-11-27T22:11:46+00:00 Remove bad whitespace and commented-out pieces - - - - - 90b8ee90 by David Waern at 2009-11-27T22:15:04+00:00 Whitespace police - - - - - b5ede900 by David Waern at 2009-11-27T22:15:50+00:00 Whitespace police - - - - - e3fddbfe by David Waern at 2009-11-28T13:37:59+00:00 Remove Name from DocInstance It's not used. - - - - - 9502786c by David Waern at 2009-11-28T13:56:54+00:00 Require at least GHC 6.12 While regression testing Haddock, I found a bug that happens with GHC 6.10.3, but not with GHC 6.12-rc2 (haven't tried 6.10.4). I don't have time to track it down. I think we should just always require the latest major GHC version. The time spent on making Haddock work with older versions is too high compared to the time spent on bugfixing, refactoring and features. - - - - - 8fa688d8 by David Waern at 2009-11-28T15:05:03+00:00 Remove cruft due to compatibility with older GHCs - - - - - 46fbbe9d by David Waern at 2009-11-28T15:07:50+00:00 Add a documentation header to Haddock.Convert - - - - - c3d2cc4a by David Waern at 2009-11-28T15:10:14+00:00 Remove unused H.Utils.FastMutInt2 - - - - - 490aba80 by David Waern at 2009-11-28T15:36:36+00:00 Rename Distribution.Haddock into Documentation.Haddock - - - - - 33ee2397 by David Waern at 2009-11-28T15:36:47+00:00 Fix error message - - - - - a5a3b950 by David Waern at 2009-11-28T16:58:39+00:00 Add a test flag that brings in QuickCheck - - - - - fa049e13 by David Waern at 2009-11-28T19:32:18+00:00 Say that we want quickcheck 2 - - - - - f32b0d9b by David Waern at 2009-11-28T19:32:40+00:00 Add an Arbitrary instance for HsDoc - - - - - da9a8bd7 by David Waern at 2009-11-28T20:15:30+00:00 Rename HsDoc back into Doc - - - - - edb60101 by David Waern at 2009-11-28T22:16:16+00:00 Move H.Interface.Parse/Lex to H.Parse/Lex These are not just used to build Interfaces. - - - - - 0656a9b8 by David Waern at 2009-11-28T23:12:14+00:00 Update version number in test suite - - - - - 5e8c6f4a by David Waern at 2009-12-21T14:12:41+00:00 Improve doc of DocName - - - - - 7868e551 by Ian Lynagh at 2009-09-22T10:43:03+00:00 TAG GHC 6.12-branch created - - - - - 0452a3ea by Ian Lynagh at 2009-12-15T12:46:07+00:00 TAG GHC 6.12.1 release - - - - - 65e9be62 by David Waern at 2009-12-21T16:58:58+00:00 Update CHANGES - - - - - 145cee32 by David Waern at 2009-12-21T16:59:09+00:00 TAG 2.6.0 - - - - - 3c552008 by David Waern at 2009-12-22T17:11:14+00:00 Update ANNOUNCE - - - - - 931f9db4 by David Waern at 2010-01-22T19:57:17+00:00 Convert haddock.vim to use unix newlines - - - - - 4e56588f by David Waern at 2010-01-22T22:11:17+00:00 Remove unnecessary (and inexplicable) uses of nub - - - - - 744bb4d1 by David Waern at 2010-01-22T22:12:14+00:00 Follow move of parser and lexer - - - - - e34bab14 by David Waern at 2010-01-22T22:49:13+00:00 Use findProgramLocation instead of findProgramOnPath in runtests.hs - - - - - 8d39891b by Isaac Dupree at 2010-01-14T18:53:18+00:00 fix html arg-doc off-by-one and silliness - - - - - 9401f2e9 by David Waern at 2010-01-22T22:57:03+00:00 Create a test for function argument docs - - - - - 507a82d7 by David Waern at 2010-01-22T23:24:47+00:00 Put parenthesis around type signature arguments of function type - - - - - 8a305c28 by David Waern at 2010-01-23T17:26:59+00:00 Add reference file for the FunArgs test - - - - - 1309d5e1 by David Waern at 2010-01-24T16:05:08+00:00 Improve FunArg test and update Test.html.ref - - - - - 2990f055 by Yitzchak Gale at 2010-02-14T16:03:46+00:00 Do not generate illegal character in HTML ID attribute. - - - - - c5bcab7a by David Waern at 2010-02-22T22:10:30+00:00 Fix Haddock markup error in comment - - - - - c6416a73 by David Waern at 2010-02-24T22:55:08+00:00 Large additions to the Haddock API Also improved and added more doc comments. - - - - - 57d289d7 by David Waern at 2010-02-24T22:58:02+00:00 Remove unused ifaceLocals - - - - - 80528d93 by David Waern at 2010-02-25T21:05:09+00:00 Add HaddockModInfo to the API - - - - - 82806848 by David Waern at 2010-02-25T21:05:27+00:00 Wibble - - - - - 744cad4c by David Waern at 2010-02-25T23:30:59+00:00 Make it possible to run a single test - - - - - 6a806e4c by David Waern at 2010-03-14T14:19:39+00:00 Bump version number - - - - - a5a8e4a7 by David Waern at 2010-03-14T14:36:35+00:00 Update ANNOUNCE - - - - - 6f05435e by Simon Hengel at 2010-03-15T20:52:42+00:00 Add missing dependencies for 'library' in haddock.cabal - - - - - faefe2bd by David Waern at 2010-03-15T22:29:37+00:00 Solve conflicts - - - - - 9808ad52 by David Waern at 2010-03-15T22:51:21+00:00 Bump version number - - - - - eb0bf60b by David Waern at 2010-03-15T22:52:32+00:00 Update CHANGES - - - - - f95cd891 by David Waern at 2010-03-15T23:01:06+00:00 Add Paths_haddock to other-modules of library - - - - - 65997b0a by David Waern at 2010-03-15T23:14:59+00:00 Update CHANGES - - - - - 7e251731 by David Waern at 2010-03-15T23:15:30+00:00 Bump version number - - - - - c9cd0ddc by David Waern at 2010-03-16T00:28:34+00:00 Fix warning - - - - - 1cac2d93 by Simon Peyton Jones at 2010-01-04T15:22:16+00:00 Fix imports for new location of splitKindFunTys - - - - - 474f26f6 by Simon Peyton Jones at 2010-02-10T14:36:06+00:00 Update Haddock for quasiquotes - - - - - 0dcc06c0 by Simon Peyton Jones at 2010-02-10T10:59:45+00:00 Track changes in HsTyVarBndr - - - - - 2d84733a by Simon Peyton Jones at 2010-02-10T14:52:44+00:00 Track HsSyn chnages - - - - - 9e3adb8b by Ian Lynagh at 2010-02-20T17:09:42+00:00 Resolve conflicts - - - - - a3e72ff8 by Simon Peyton Jones at 2010-03-04T13:05:16+00:00 Track change in HsUtils; and use a nicer function not an internal one - - - - - 27994854 by David Waern at 2010-03-18T22:22:27+00:00 Fix build with GHC 6.12.1 - - - - - 11f6e488 by David Waern at 2010-03-18T22:24:09+00:00 Bump version in test reference files - - - - - 0ef2f11b by David Waern at 2010-03-20T00:56:30+00:00 Fix library part of cabal file when in ghc tree - - - - - 3f6146ff by Mark Lentczner at 2010-03-20T22:30:11+00:00 First, experimental XHTML rendering switch to using the xhtml package copied Html.hs to Xhtml.hs and split into sub-modules under Haddock/Backends/Xhtml and detabify moved footer into div, got ready for iface change headers converted to semantic markup contents in semantic markup summary as semantic markup description in semantic markup, info block in header fixed factored out rendering so during debug it can be readable (see renderToString) - - - - - b8ab329b by Mark Lentczner at 2010-03-20T22:54:01+00:00 apply changes to Html.hs to Xhtml/*.hs incorporate changes that were made between the time Html.hs was copied and split into Xhtml.hs and Xhtml/*.hs includes patchs after "Wibble" (!) through "Fix build with GHC 6.12.1" - - - - - 73df2433 by Ian Lynagh at 2010-03-20T21:56:37+00:00 Follow LazyUniqFM->UniqFM in GHC - - - - - db4f602b by David Waern at 2010-03-29T22:00:01+00:00 Fix build with GHC 6.12 - - - - - d8dca088 by Simon Hengel at 2010-04-02T16:39:55+00:00 Add missing dependencies to cabal file - - - - - e2adc437 by Simon Hengel at 2010-04-02T14:08:40+00:00 Add markup support for interactive examples - - - - - e882ac05 by Simon Hengel at 2010-04-02T14:11:53+00:00 Add tests for interactive examples - - - - - 5a07a6d3 by David Waern at 2010-04-07T17:05:20+00:00 Propagate source positions from Lex.x to Parse.y - - - - - 6493b46f by David Waern at 2010-04-07T21:48:57+00:00 Let runtests.hs die when haddock has not been built - - - - - 5e34423e by David Waern at 2010-04-07T22:01:13+00:00 Make runtests.hs slightly more readable - - - - - 321d59b3 by David Waern at 2010-04-07T22:13:27+00:00 Fix haskell/haddock#75 Add colons to the $ident character set. - - - - - 37b08b8d by David Waern at 2010-04-08T00:32:52+00:00 Fix haskell/haddock#118 Avoid being too greedy when lexing URL markup (<..>), in order to allow multiple URLs on the same line. Do the same thing with <<..>> and #..#. - - - - - df8feac9 by David Waern at 2010-04-08T00:57:33+00:00 Make it easier to add new package deps to test suite This is a hack - we should use Cabal to get the package details instead. - - - - - 1ca6f84b by David Waern at 2010-04-08T01:03:06+00:00 Add ghc-prim to test suite deps - - - - - 27371e3a by Simon Hengel at 2010-04-08T19:26:34+00:00 Let parsing fails on paragraphs that are immediately followed by an example This is more consistent with the way we treat code blocks. - - - - - 83096e4a by David Waern at 2010-04-08T21:20:00+00:00 Improve function name - - - - - 439983ce by David Waern at 2010-04-10T10:46:14+00:00 Fix haskell/haddock#112 No link was generated for 'Addr#' in a doc comment. The reason was simply that the identifier didn't parse. We were using parseIdentifier from the GHC API, with a parser state built from 'defaultDynFlags'. If we pass the dynflags of the module instead, the right options are turned on on while parsing the identifer (in this case -XMagicHash), and the parse succeeds. - - - - - 5c0d35d7 by David Waern at 2010-04-10T10:54:06+00:00 Rename startGhc into withGhc - - - - - dca081fa by Simon Hengel at 2010-04-12T19:09:16+00:00 Add documentation for interactive examples - - - - - c7f26bfa by David Waern at 2010-04-13T00:51:51+00:00 Slight fix to the documentation of examples - - - - - 06eb7c4c by David Waern at 2010-04-13T00:57:05+00:00 Rename Interactive Examples into Examples (and simplify explanation) - - - - - 264830cb by David Waern at 2010-05-10T20:07:27+00:00 Update CHANGES with info about 2.6.1 - - - - - 8e5d4514 by Simon Hengel at 2010-04-18T18:16:54+00:00 Add unit tests for parser - - - - - 68297f40 by David Waern at 2010-05-10T21:53:37+00:00 Improve testsuite README - - - - - f04eb6e4 by David Waern at 2010-05-11T19:14:31+00:00 Re-organise the testsuite structure - - - - - a360f710 by David Waern at 2010-05-11T19:18:03+00:00 Shorten function name - - - - - 1d5dd359 by David Waern at 2010-05-11T21:40:02+00:00 Update runtests.hs following testsuite re-organisation - - - - - ffebe217 by David Waern at 2010-05-11T21:40:10+00:00 Update runtests.hs to use base-4.2.0.1 - - - - - 635de402 by David Waern at 2010-05-11T21:41:11+00:00 Update runparsetests.hs following testsuite reorganisation - - - - - 72137910 by Ian Lynagh at 2010-05-06T20:43:06+00:00 Fix build - - - - - 1a80b76e by Ian Lynagh at 2010-05-06T22:25:29+00:00 Remove redundant import - - - - - 1031a80c by Simon Peyton Jones at 2010-05-07T13:21:09+00:00 Minor wibbles to HsBang stuff - - - - - dd8e7fe5 by Ian Lynagh at 2010-05-08T15:22:00+00:00 GHC build system: Follow "rm" variable changes - - - - - 7f5e6748 by David Waern at 2010-05-13T11:53:02+00:00 Fix build with GHC 6.12.2 - - - - - 7953d4d8 by David Waern at 2010-05-13T18:45:01+00:00 Fixes to comments only - - - - - 8ae8eb64 by David Waern at 2010-05-13T18:57:26+00:00 ModuleMap -> IfaceMap - - - - - 1c3eadc6 by David Waern at 2010-05-13T19:03:13+00:00 Fix whitespace style issues - - - - - e96783c0 by David Waern at 2010-05-13T19:08:53+00:00 Fix comment - - - - - c998a78b by David Waern at 2010-05-13T19:39:00+00:00 Position the module header the same way everywhere Silly, but nice with some consistency :-) - - - - - b48a714e by David Waern at 2010-05-13T19:41:32+00:00 Position of module header, this time in the HTML backends - - - - - f9bfb12e by David Waern at 2010-05-13T19:43:05+00:00 Two newlines between declarations in Main - - - - - 071d44c7 by David Waern at 2010-05-13T19:44:21+00:00 Newlines in Convert - - - - - 036346db by David Waern at 2010-05-13T19:46:47+00:00 Fix a few stylistic issues in H.InterfaceFile - - - - - f0b8379e by David Waern at 2010-05-13T19:47:53+00:00 Add newlines to H.ModuleTree - - - - - 27409f8e by David Waern at 2010-05-13T19:51:10+00:00 Fix stylistic issues in H.Utils - - - - - 24774a11 by David Waern at 2010-05-13T20:00:43+00:00 Structure H.Types better - - - - - 7b6f5e40 by David Waern at 2010-05-13T20:01:04+00:00 Remove bad Arbitrary instance - - - - - fac9f1f6 by David Waern at 2010-05-13T20:05:50+00:00 Get rid of H.Utils.pathJoin and use System.FilePath.joinPath instead - - - - - fe6d00c4 by David Waern at 2010-05-13T20:51:55+00:00 Export a couple of more types from the API - - - - - b2e33a5f by David Waern at 2010-05-13T21:27:51+00:00 Improve doc comment for Interface - - - - - c585f2ce by David Waern at 2010-05-13T21:30:14+00:00 Improve documentation of Haddock.Interface - - - - - e6791db2 by David Waern at 2010-05-13T22:07:35+00:00 Remove meaningless comments - - - - - 7801b390 by David Waern at 2010-05-14T17:53:33+00:00 Remove unused modules - - - - - f813e937 by David Waern at 2010-05-14T17:55:17+00:00 Re-direct compilation output to a temporary directory Also add a flag --no-tmp-comp-dir that can be used to get the old behaviour of writing compilation files to GHC's output directory (default "."). - - - - - e56737ec by David Waern at 2010-05-14T18:06:11+00:00 Wibble - - - - - e40b0447 by David Waern at 2010-05-14T19:01:52+00:00 Move flag evaluation code from Main to Haddock.Options Determining the value of "singular" flags (by e.g. taking the last occurrence of the flag) and other flag evaluation should done in Haddock.Options which is the module that is supposed to define the command line interface. This makes Main a bit easier on the eyes as well. - - - - - 27091f57 by David Waern at 2010-05-14T19:05:10+00:00 Wibble - - - - - c658cf61 by David Waern at 2010-05-14T19:06:49+00:00 Re-order things in Haddock.Options a bit - - - - - 8cfdd342 by David Waern at 2010-05-14T19:20:29+00:00 De-tabify Haddock.Options and fix other whitespace issues - - - - - 0df16b62 by David Waern at 2010-05-14T19:25:07+00:00 Improve comments - - - - - 80b38e2b by David Waern at 2010-05-14T19:26:42+00:00 Whitespace police - - - - - fe580255 by David Waern at 2010-05-14T19:31:23+00:00 Wibbles to comments - - - - - a2b43fad by David Waern at 2010-05-14T20:24:32+00:00 Move some more flag functions to Haddock.Options - - - - - 3f895547 by David Waern at 2010-05-14T20:37:12+00:00 Make renderStep a top-level function in Main - - - - - 5cdca11d by David Waern at 2010-05-14T20:39:27+00:00 Spelling in comment - - - - - ad98d14c by David Waern at 2010-05-14T20:40:26+00:00 Comment fixes - - - - - 0bb9218f by David Waern at 2010-05-14T20:49:01+00:00 Whitespace police - - - - - 0f0a533f by David Waern at 2010-05-15T16:42:29+00:00 Improve description of --dump-interface - - - - - 5b2833ac by David Waern at 2010-05-15T17:16:53+00:00 Document --no-tmp-comp-dir - - - - - 8160b170 by David Waern at 2010-05-15T17:18:59+00:00 Wibble - - - - - 570dbe33 by David Waern at 2010-05-18T21:15:38+00:00 HLint police - - - - - 204e425f by David Waern at 2010-05-18T21:16:30+00:00 HLint police - - - - - 6db657ac by David Waern at 2010-05-18T21:16:37+00:00 Wibble - - - - - b942ccd7 by Simon Marlow at 2010-06-02T08:27:30+00:00 Interrupted disappeared in GHC 6.13 (GHC ticket haskell/haddock#4100) - - - - - 3b94a819 by Simon Marlow at 2010-06-02T08:45:08+00:00 Allow base-4.3 - - - - - c5a1fb7c by Simon Marlow at 2010-06-02T09:03:04+00:00 Fix compilation with GHC 6.13 - - - - - 6181296c by David Waern at 2010-06-08T21:09:05+00:00 Display name of prologue file when parsing it fails - - - - - 7cbc6f60 by Ian Lynagh at 2010-06-13T16:20:25+00:00 Remove redundant imports - - - - - 980c804b by Simon Marlow at 2010-06-22T08:41:50+00:00 isLocalAndTypeInferenced: fix for local module names overlapping package modules - - - - - d74d4a12 by Simon Marlow at 2010-06-23T12:03:27+00:00 Unresolved identifiers in Doc get replaced with DocMonospaced rather than plain strings - - - - - d8546783 by Simon Marlow at 2010-06-30T12:45:17+00:00 LaTeX backend (new options: --latex, --latex-style=<style>) - - - - - 437afa9e by David Waern at 2010-07-01T12:02:44+00:00 Fix a few stylistic whitespace issues in LaTeX backend - - - - - 85bc1fae by David Waern at 2010-07-01T15:42:45+00:00 Make runtest.hs work with GHC 6.12.3 (we should really stop hard coding this) - - - - - 7d2eb86f by David Waern at 2010-07-01T15:43:33+00:00 Update test following Simon's patch to render unresolved names in monospaced font - - - - - 08fcbcd2 by David Waern at 2010-07-01T16:12:18+00:00 Warning police - - - - - d04a8d7a by David Waern at 2010-07-04T14:53:39+00:00 Fix a bug in attachInstances We didn't look for instance docs in all the interfaces of the package. This had the effect of instance docs not always showing up under a declaration. I took the opportunity to clean up the code in H.I.AttachInstances a bit as well. More cleanup is needed, however. - - - - - d10344eb by Simon Hengel at 2010-07-10T09:19:04+00:00 Add missing dependencies to cabal file - - - - - 24090531 by Mark Lentczner at 2010-03-21T04:51:16+00:00 add exports to Xhtml modules - - - - - 84f9a333 by Mark Lentczner at 2010-04-03T19:14:22+00:00 clean up Doc formatting code - add CSS for lists - renderToString now uses showHtml since prettyHtml messes up <pre> sections - - - - - bebccf52 by Mark Lentczner at 2010-04-04T04:51:08+00:00 tweak list css - - - - - 0c2aeb5e by Mark Lentczner at 2010-04-04T06:24:14+00:00 all decls now generate Html not HtmlTable - ppDecl return Html, and so now do all of the functions it calls - added some internal tables to some decls, which is wrong, and will have to be fixed - decl "Box" functions became "Elem" functions to make clear they aren't in a table anymore (see Layout.hs) - docBox went away, as only used in one place (and its days are numbered) - cleaned up logic in a number of places, removed dead code - added maybeDocToHtml which simplified a number of places in the code - - - - - dbf73e6e by Mark Lentczner at 2010-04-05T05:02:43+00:00 clean up processExport and place a div around each decl - - - - - e25b7e9f by Mark Lentczner at 2010-04-10T21:23:21+00:00 data decls are now a sequence of paragraphs, not a table - - - - - 89ee0294 by Mark Lentczner at 2010-04-10T21:29:16+00:00 removed commented out code that can't be maintained - - - - - d466f536 by Mark Lentczner at 2010-04-12T04:56:27+00:00 removed declWithDoc and cleaned up data decls in summary - - - - - ed755832 by Mark Lentczner at 2010-04-12T05:07:53+00:00 merge in markupExample changes - - - - - c36f51fd by Mark Lentczner at 2010-04-25T04:56:37+00:00 made record fields be an unordList, not a table - - - - - ed3a28d6 by Mark Lentczner at 2010-04-25T05:23:28+00:00 fixed surround of instance and constructor tables - - - - - 0e35bbc4 by Mark Lentczner at 2010-04-25T05:36:59+00:00 fix class member boxes in summary - - - - - 5041749b by Mark Lentczner at 2010-04-25T05:38:35+00:00 remove unused bodyBox - - - - - e91724db by Mark Lentczner at 2010-04-25T06:26:10+00:00 fixed javascript quoting/escpaing issue - - - - - f4abbb73 by Mark Lentczner at 2010-05-03T23:04:31+00:00 adjust css for current markup - - - - - e75fec4c by Mark Lentczner at 2010-05-04T06:14:34+00:00 added assoicated types and methods back into class decls - - - - - 84169323 by Mark Lentczner at 2010-05-24T13:13:42+00:00 merge in changes from the big-whitespace cleanup - - - - - 3c1c872e by Mark Lentczner at 2010-06-11T21:03:58+00:00 adjust synopsis and bottom bar spacing - - - - - 3c1f9ef7 by Mark Lentczner at 2010-06-11T21:14:44+00:00 fix missing space in "module" lines in synoposis - - - - - 9a137e6d by Mark Lentczner at 2010-06-11T21:34:08+00:00 changed tt elements to code elements - - - - - 50f71ef1 by Mark Lentczner at 2010-06-11T23:27:46+00:00 factored out ppInstances - - - - - 3b9a9de5 by Mark Lentczner at 2010-06-17T17:36:01+00:00 push single constructors (newtype) onto line with decl - - - - - e0f8f2ec by Mark Lentczner at 2010-06-17T22:20:56+00:00 remove <++> connector - - - - - 56c075dd by Mark Lentczner at 2010-07-13T05:26:21+00:00 change to new page structure - - - - - 04be6ca7 by Mark Lentczner at 2010-07-14T04:21:55+00:00 constructors and args as dl lists, built in Layout.hs - - - - - 65aeafc2 by Mark Lentczner at 2010-07-14T05:38:32+00:00 better interface to subDecls - - - - - 72032189 by Mark Lentczner at 2010-07-14T07:04:10+00:00 made subDecl tables looks just so - - - - - b782eca2 by Mark Lentczner at 2010-07-14T16:00:54+00:00 convert args to SubDecl format - - - - - cc75e98f by Mark Lentczner at 2010-07-14T16:28:53+00:00 convert instances to SubDecl - - - - - 34e2aa5a by Mark Lentczner at 2010-07-14T21:07:32+00:00 removing old table cruft from Layout.hs - - - - - d5810d95 by Mark Lentczner at 2010-07-14T21:54:58+00:00 methods and associated types in new layout scheme - - - - - 65ef9579 by Mark Lentczner at 2010-07-14T23:43:42+00:00 clean up synopsis lists - - - - - e523318f by Mark Lentczner at 2010-07-15T05:02:26+00:00 clean up of anchors - - - - - 1215dfc5 by Mark Lentczner at 2010-07-15T23:53:01+00:00 added two new themes and rough css switcher - - - - - 7f0fd36f by Mark Lentczner at 2010-07-16T04:57:38+00:00 fixed package catpion, added style menu - - - - - 0dd4999c by Mark Lentczner at 2010-07-16T20:12:39+00:00 new output for mini_ pages - - - - - 64b2810b by Mark Lentczner at 2010-07-16T20:58:41+00:00 reformat index-frames - - - - - 3173f555 by Mark Lentczner at 2010-07-16T22:41:53+00:00 convert index to new markup - - - - - b0a4b7c9 by Mark Lentczner at 2010-07-17T04:07:22+00:00 convert index.html to new markup, adjust module markup - - - - - 8261ae1e by Mark Lentczner at 2010-07-17T05:07:29+00:00 classing styling of ancillary pages - - - - - 2a4fb025 by Mark Lentczner at 2010-07-17T05:11:45+00:00 clean up Layout.hs: no more vanillaTable - - - - - 87eec685 by Mark Lentczner at 2010-07-17T05:35:16+00:00 clean up Util.hs - - - - - d304e9b0 by Mark Lentczner at 2010-07-17T05:38:50+00:00 qualify import of XHtml as XHtml - - - - - 7dc05807 by Mark Lentczner at 2010-07-17T06:17:53+00:00 factored out head element generation - - - - - 9cdaec9e by Mark Lentczner at 2010-07-17T06:44:54+00:00 refactored out main page body generation - - - - - 8a51019e by Mark Lentczner at 2010-07-17T06:48:20+00:00 moved footer into only place that used it - - - - - efa479da by Mark Lentczner at 2010-07-17T18:48:30+00:00 styling auxillary pages for tibbe and snappy themes - - - - - 81de5509 by Mark Lentczner at 2010-07-18T04:41:38+00:00 fixed alphabet on index page, and styling of it and packages in module lists - - - - - 20718c1a by Mark Lentczner at 2010-07-18T05:34:29+00:00 cleaned up div functions in Layout.hs - - - - - 60d50453 by Mark Lentczner at 2010-07-18T05:48:39+00:00 added content div to main pages - - - - - ed16561c by Mark Lentczner at 2010-07-18T06:12:22+00:00 add .doc class to documentation blocks - - - - - f5c781b0 by Mark Lentczner at 2010-07-19T05:20:53+00:00 refactoring of anchor ID and fragment handling - - - - - a69a93bf by Mark Lentczner at 2010-07-19T05:35:55+00:00 remove an explicit bold tag - replace with .def class - - - - - d76c7225 by Mark Lentczner at 2010-07-19T06:56:15+00:00 rename Haddock.Backends.Xhtml.Util to Utils - - - - - 5a58c0da by David Waern at 2010-07-21T13:30:54+00:00 Remove trailing whitespace in Haddock.Backends.Xhtml - - - - - 0652aa17 by David Waern at 2010-07-21T13:33:21+00:00 Align a few comments - - - - - 785776c3 by David Waern at 2010-07-21T13:39:04+00:00 Remove trailing whitespace in H.B.X.Decl - - - - - 71a30710 by David Waern at 2010-07-21T13:44:27+00:00 Remove more trailing whitespace - - - - - 38750394 by David Waern at 2010-07-21T13:50:43+00:00 Style police - - - - - 3023d940 by David Waern at 2010-07-21T14:01:22+00:00 Style police in H.B.X.Decl - - - - - df16e9e6 by David Waern at 2010-07-21T14:14:45+00:00 Style police in H.B.X.DocMarkup - - - - - 6020e321 by David Waern at 2010-07-21T14:17:32+00:00 More style police - - - - - 86ad8bf5 by David Waern at 2010-07-21T14:21:02+00:00 Style police in H.B.Xhtml - - - - - aea27d03 by David Waern at 2010-07-21T14:42:03+00:00 Fix warnings in LaTeX backend - - - - - 2aff34a9 by David Waern at 2010-07-21T14:50:46+00:00 Style police in LaTeX backend (mainly more newlines) - - - - - e517162d by David Waern at 2010-07-21T15:05:47+00:00 Doc sections in Main - - - - - b971aa0c by David Waern at 2010-07-21T15:06:17+00:00 Trailing whitespace in Documentation.Haddock - - - - - f11628fb by David Waern at 2010-07-21T15:07:06+00:00 Trailing whitespace in Haddock.Convert - - - - - cbaf284c by David Waern at 2010-07-21T15:08:11+00:00 Style police in Haddock.GhcUtils - - - - - 71feb77b by David Waern at 2010-07-21T15:09:06+00:00 Style police in Haddock.InterfaceFile - - - - - 0a9c80e6 by David Waern at 2010-07-21T15:11:33+00:00 Whitespace police - - - - - 6168376c by David Waern at 2010-07-21T15:16:35+00:00 Style police in Haddock.Utils - - - - - 9fe4dd90 by David Waern at 2010-07-21T15:19:31+00:00 Add -fwarn-tabs - - - - - a000d752 by Mark Lentczner at 2010-07-20T17:25:52+00:00 move CSS Theme functions into Themes.hs - - - - - b52b440f by Mark Lentczner at 2010-07-20T17:29:35+00:00 add Thomas Schilling's theme - - - - - e43fa7e8 by Mark Lentczner at 2010-07-21T04:49:34+00:00 correct icon used with Snappy theme - - - - - ba5092d3 by Mark Lentczner at 2010-07-21T04:56:47+00:00 apply Tibbe's updates to his theme - - - - - 7804eef6 by Mark Lentczner at 2010-07-21T05:15:49+00:00 space between "Style" and the downward triangle - - - - - 7131d4c6 by Mark Lentczner at 2010-07-21T17:43:35+00:00 merge with David's source cleanups - - - - - ee65f1cb by David Waern at 2010-07-22T16:50:46+00:00 Fix a bug where we allowed --hoogle, --latex, etc without input files - - - - - e413ff7a by David Waern at 2010-07-22T17:21:58+00:00 Improve function name - - - - - a0fd14f3 by Simon Marlow at 2010-06-30T15:34:32+00:00 fix warnings - - - - - 31f73d2a by David Waern at 2010-07-22T19:29:41+00:00 Solve conflicts - - - - - d563b4a5 by Simon Marlow at 2010-06-30T15:34:37+00:00 fix warning - - - - - 412b6469 by David Waern at 2010-07-22T19:31:28+00:00 Solve conflict - - - - - 35174b94 by Ian Lynagh at 2010-07-06T17:27:16+00:00 Follow mkPState argument order change - - - - - b5c3585c by Simon Marlow at 2010-07-14T08:49:21+00:00 common up code for instance rendering - - - - - d8009560 by Simon Marlow at 2010-07-14T12:37:11+00:00 fix warnings - - - - - a6d88695 by David Waern at 2010-07-24T15:33:33+00:00 Fix build with ghc < 6.13 - - - - - 94cf9de1 by David Waern at 2010-07-24T15:34:37+00:00 Remove conflict left-over - - - - - 313b15c0 by Mark Lentczner at 2010-07-21T22:09:04+00:00 reorganization of nhaddock.css with tibbe - - - - - 9defed80 by Mark Lentczner at 2010-07-21T22:42:14+00:00 further cleanup of nhaddock.css, float TOC, support aux. pages - - - - - 6d944c1b by Mark Lentczner at 2010-07-22T06:22:23+00:00 remove old HTML backend - - - - - b3e8cba5 by Mark Lentczner at 2010-07-22T06:43:32+00:00 remove --html-help support - it was old, out-of-date, and mostly missing - - - - - d2654a08 by Mark Lentczner at 2010-07-22T21:45:34+00:00 tweaks to nhaddock.css - - - - - f73b285c by Mark Lentczner at 2010-07-23T06:19:35+00:00 command like processing for theme selection The bulk of the change is threadnig the selected theme set through functions in Xhtml.hs so that the selected themes can be used when generating the page output. There isn't much going on in most of these changes, just passing it along. The real work is all done in Themes.hs. - - - - - 8bddc90d by Mark Lentczner at 2010-07-23T06:58:31+00:00 drop --themes support, add named theme support decided that --themes was silly - no one would do that, just use multiple --theme arguments made --theme a synonym for --css and -c made those arguments, if no file is found, look up the argument as the name of a built in theme all of this let's haddock be invoked with "--theme=classic" for example. - - - - - 20cafd4f by Mark Lentczner at 2010-07-23T17:44:29+00:00 rename --default-themes to --built-in-themes - - - - - 0fe41307 by Mark Lentczner at 2010-07-23T18:33:02+00:00 tweaks to theme for info table, headings, and tables - - - - - cba4fee0 by Mark Lentczner at 2010-07-23T19:13:59+00:00 tweaks for dl layout, though still not used - - - - - 463fa294 by Mark Lentczner at 2010-07-23T21:07:19+00:00 tweak look of mini pages, keywords, and preblocks - - - - - 5472fc02 by Mark Lentczner at 2010-07-24T05:36:15+00:00 slide out Synopsis drawer - - - - - 9d5d5de5 by Mark Lentczner at 2010-07-24T06:02:42+00:00 extend package header and footer to edges of page - - - - - a47c91a2 by Mark Lentczner at 2010-07-24T06:28:44+00:00 fields are def lists, tweak css for style menu, mini pages, arguments - - - - - ca20f23b by Mark Lentczner at 2010-07-24T16:55:22+00:00 excisting last vestiges of the --xhtml flag - - - - - 71fb012e by Mark Lentczner at 2010-07-25T18:47:49+00:00 change how collapsing sections are done make whole .caption be the target improve javascript for class toggling have plus/minus images come from .css, not img tags - - - - - c168c8d3 by Mark Lentczner at 2010-07-26T00:32:05+00:00 reorganize files in the html lib data dir - - - - - 93324301 by Mark Lentczner at 2010-07-26T01:27:42+00:00 cleaned up Themes.hs - - - - - ad3b5dd4 by Mark Lentczner at 2010-07-26T02:39:15+00:00 make module list use new collapsers - - - - - 1df9bfc6 by Mark Lentczner at 2010-07-27T19:09:25+00:00 remove Tibbe theme - - - - - 8b9b01b3 by Mark Lentczner at 2010-07-27T20:04:03+00:00 move themes into html dir with .theme and .std-theme extensions - - - - - a7beb965 by Mark Lentczner at 2010-07-27T21:06:34+00:00 give a class to empty dd elements so they can be hidden - - - - - a258c117 by Mark Lentczner at 2010-07-27T21:23:58+00:00 remove custom version of copyFile in Xhtml.hs - - - - - b70dba6e by Mark Lentczner at 2010-07-27T22:12:45+00:00 apply margin changes to pre and headings as per group decision, and small cleanups - - - - - e6f722a2 by Mark Lentczner at 2010-07-28T00:03:12+00:00 make info block and package bar links be floatable by placing them first in the dom tree - - - - - c8278867 by Mark Lentczner at 2010-07-28T19:01:18+00:00 styling source links on declarations - - - - - 88fdc399 by Mark Lentczner at 2010-07-29T01:12:46+00:00 styling tweaks don't generate an empty li for absent style menu in links area update css for Classic and Snappy to handle: dl lists links in package header and in declarations floating of links and info block in package and module headers - - - - - 8a75b213 by Ian Lynagh at 2010-07-30T20:21:46+00:00 Fix build in GHC tree - - - - - ce8e18b3 by Simon Hengel at 2010-08-03T18:37:26+00:00 Adapt paths to data files in cabal file - - - - - 9701a455 by Simon Hengel at 2010-08-07T13:20:27+00:00 Add missing dependency to cabal file - - - - - 01b838d1 by Mark Lentczner at 2010-07-30T20:19:40+00:00 improved synopsis drawer: on click, not hover - - - - - 7b6f3e59 by Mark Lentczner at 2010-07-30T23:38:55+00:00 put the synopsis back in the other themes - - - - - 7b2904c9 by Mark Lentczner at 2010-08-11T11:11:26+00:00 close arrows on expanded synopsis drawer - - - - - ea19e177 by Mark Lentczner at 2010-08-12T21:16:45+00:00 width and font changes removed the max width restrictions on the page as a whole and the synopsis made the main font size smaller (nominally 14pt) and then tweaked most font sizes (relative) to be more consistent - - - - - 5ced00c0 by Mark Lentczner at 2010-08-13T15:09:55+00:00 implemented YUI's CSS font approach - - - - - 2799c548 by Mark Lentczner at 2010-08-13T15:11:59+00:00 adjusted margin to 2em, 1 wasn't enough - - - - - 58f06893 by Mark Lentczner at 2010-08-13T15:48:44+00:00 removed underlining on hover for named anchors headings in interface lost thier a element, no need, just put id on heading css for a elements now only applies to those with href attribute - - - - - 7aced4c4 by Mark Lentczner at 2010-08-13T15:50:22+00:00 more space between elements - - - - - 5a3c1cce by Mark Lentczner at 2010-08-13T16:43:43+00:00 adjusted font sizes of auxilary pages per new scheme - - - - - 487539ef by Mark Lentczner at 2010-08-13T21:43:41+00:00 add Frames button and clean up frames.html - - - - - c1a140b6 by Mark Lentczner at 2010-08-13T22:17:48+00:00 move frames button to js - - - - - b0bdb68e by Mark Lentczner at 2010-08-14T03:44:46+00:00 build style menu in javascript moved to javascript, so as to not polute the content with the style menu removed menu building code in Themes.hs removed onclick in Utils.hs changed text of button in header from "Source code" to "Source" more consistent with links in rest of page - - - - - 43ab7120 by Mark Lentczner at 2010-08-16T15:15:37+00:00 font size and margin tweaks - - - - - c0b68652 by Mark Lentczner at 2010-08-17T18:19:52+00:00 clean up collapser logics javascript code for collapasble sections cleaned up rewrote class utilities in javascript to be more robust refactored utilities for generating collapsable sections made toc be same color as synopsis module list has needed clear attribute in CSS - - - - - 5d573427 by Mark Lentczner at 2010-08-17T23:06:02+00:00 don't collapse entries in module list when clicking on links - - - - - 8c307c4a by Mark Lentczner at 2010-08-17T23:21:43+00:00 add missing data file to .cabal - - - - - 414bcfcf by Mark Lentczner at 2010-08-17T23:28:47+00:00 remove synopsis when in frames - - - - - ba0fa98a by Mark Lentczner at 2010-08-18T16:16:11+00:00 layout tweeks - mini page font size, toc color, etc. - - - - - 63c1bed1 by Mark Lentczner at 2010-08-18T19:50:02+00:00 margin fiddling - - - - - c311c094 by Mark Lentczner at 2010-08-20T01:37:55+00:00 better synopsis handling logic - no flashing - - - - - f1fe5fa8 by Mark Lentczner at 2010-08-20T01:41:06+00:00 fix small layout issues mini frames should have same size top heading give info block dts some padding so they don't collide in some browsers - - - - - 0de84d77 by Mark Lentczner at 2010-08-20T02:13:09+00:00 made style changing and cookies storage robust - - - - - 1ef064f9 by Thomas Schilling at 2010-08-04T13:12:22+00:00 Make synopsis frame behave properly in Firefox. In Firefox, pressing the back button first reverted the synopsis frame, and only clicking the back button a second time would update the main frame. - - - - - dd1c9a94 by Mark Lentczner at 2010-08-21T01:46:19+00:00 remove Snappy theme - - - - - 2353a90d by Mark Lentczner at 2010-08-25T05:16:19+00:00 fix occasional v.scroll bars on pre blocks (I think) - - - - - 459b8bf1 by Simon Hengel at 2010-08-08T10:12:45+00:00 Add createInterfaces' (a more high-level alternative to createInterfaces) to Haddock API - - - - - b1b68675 by David Waern at 2010-08-26T20:31:58+00:00 Follow recent API additions with some refactorings Simon Hegel's patch prompted me to do some refactorings in Main, Haddock.Documentation and Haddock.Interface. - - - - - 264d4d67 by David Waern at 2010-08-26T21:40:59+00:00 Get rid of GhcModule and related cruft We can get everything we need directly from TypecheckedModule. - - - - - 0feacec2 by Mark Lentczner at 2010-08-26T23:44:13+00:00 fixed CSS for ordered lists and def lists in doc blocks - - - - - 2997e0c2 by Mark Lentczner at 2010-08-26T23:45:03+00:00 support both kinds of enumerated lists in doc markup The documentation for Haddock says enumerated lists can use either of (1) first item 2. second item The second form wasn't actually supported - - - - - 5d4ddeec by Mark Lentczner at 2010-08-27T21:29:48+00:00 fix broken header link margins - - - - - 614456ba by Mark Lentczner at 2010-08-27T22:16:19+00:00 fix table of contents CSS - - - - - 03f329a2 by David Waern at 2010-08-28T16:36:09+00:00 Update tests following switch to the Xhtml backend - - - - - ca689fa2 by Mark Lentczner at 2010-08-28T18:25:16+00:00 fix def lists - - - - - 18e1d3d2 by Mark Lentczner at 2010-08-28T18:26:18+00:00 push footer to bottom of window - - - - - b0ab8d82 by David Waern at 2010-08-28T22:04:32+00:00 Whitespace police - - - - - 2d217977 by David Waern at 2010-08-29T12:44:45+00:00 Remove Snappy data files - - - - - 01e27d5f by David Waern at 2010-08-29T13:03:28+00:00 Add source entity path to --read-interface You can now use this flag like this: --read-interface=<html path>,<source entity path>,<.haddock file> By "source entity path" I mean the same thing that is specified with the --source-entity flag. The purpose of this is to be able to specify the source entity path per package, to allow source links to work in the presence of cross-package documentation. When given two arguments or less the --read-interface flag behaves as before. - - - - - 20bf4aaa by David Waern at 2010-08-29T13:11:03+00:00 Naming wibbles - - - - - ad22463f by Mark Lentczner at 2010-08-29T15:14:54+00:00 make portability block be a table - solves layout issues - - - - - 97bd1ae6 by Mark Lentczner at 2010-08-29T15:17:42+00:00 update golden test for Test due to portability box change - - - - - d37e139e by Mark Lentczner at 2010-08-29T17:07:17+00:00 move TOC and Info blocks down 0.5em to improve layout issue w/Test.hs - - - - - acf52501 by David Waern at 2010-08-29T17:32:36+00:00 Allow building with ghc < 6.16 - - - - - 1cb34ed8 by Ian Lynagh at 2010-07-24T23:18:49+00:00 Flatten the dynflags before parsing - - - - - b36845b4 by Ian Lynagh at 2010-07-24T23:26:49+00:00 Follow flattenLanguageFlags -> flattenExtensionFlags rename - - - - - 7f7fcc7e by David Waern at 2010-08-29T17:46:23+00:00 Use flattenExtensionFlags with ghc >= 6.13 only - - - - - 13cf9411 by Ian Lynagh at 2010-08-01T18:09:54+00:00 Make the main haddock script versioned, and make plain "haddock" a symlink - - - - - 495cbff2 by Ian Lynagh at 2010-08-18T18:57:24+00:00 Fix installation in the GHC build system Data-files are now in subdirectories, so we need to handle that - - - - - 88ebab0a by Ian Lynagh at 2010-08-18T19:43:53+00:00 GHC build system: Add all the data files to BINDIST_EXTRAS - - - - - 65837172 by David Waern at 2010-08-29T20:12:34+00:00 Update Test - - - - - 094bbaa2 by David Waern at 2010-08-29T20:55:14+00:00 Revert update to Test - - - - - a881cfb3 by David Waern at 2010-08-31T18:24:15+00:00 Bump version number - - - - - 1fc8a3eb by David Waern at 2010-08-31T22:32:27+00:00 Update ANNOUNCE - - - - - ee1df9d0 by David Waern at 2010-08-31T22:33:11+00:00 Update CHANGES - - - - - 394cc854 by David Waern at 2010-08-31T22:33:23+00:00 Update interface file versioning to work with ghc 6.14/15 - - - - - 7d03b79b by David Waern at 2010-08-31T22:36:00+00:00 Update test output following version change - - - - - a48d82d1 by Mark Lentczner at 2010-09-01T04:29:35+00:00 sort options in doc to match --help output removed --html-help option, as it is no longer supported - - - - - 06561aeb by Mark Lentczner at 2010-09-01T05:29:32+00:00 update options documentation rewrote doc for --html added doc for --theme and --built-in-themes added --use-contents and --gen-contents - - - - - 57dea832 by Mark Lentczner at 2010-09-01T05:31:27+00:00 slight wording change about Frames mode - - - - - fa1f6da3 by David Waern at 2010-09-01T10:57:44+00:00 Update doc configure script to find docbook stylesheets on arch linux - - - - - addff770 by David Waern at 2010-09-01T11:02:29+00:00 Wibble - - - - - 8399006d by David Waern at 2010-09-01T11:19:21+00:00 Replace ghci> with >>> in example syntax - - - - - 35074cf8 by David Waern at 2010-09-01T19:03:27+00:00 Improve docs for --no-tmp-comp-dir - - - - - 0f8f8cfd by David Waern at 2010-09-02T11:22:27+00:00 Add a list of contributors to the user guide Break out everyone thanked in the `Acknowledgements` chapter into a separate contributor list and add everyone from `darcs show authors`. We consider everyone who is thanked to be a contributor as a conservative estimation :-) I have added some more contributors that I know about, who were not in the darcs history, but others may be missing. So please add anyone that you think is missing from the list. - - - - - 42ccf099 by David Waern at 2010-09-02T11:29:22+00:00 Update copyright years in license - - - - - 0d560479 by David Waern at 2010-09-02T11:38:52+00:00 Update release instructions - - - - - 72ab7796 by David Waern at 2010-09-02T19:27:08+00:00 Add a note to ANNOUNCE - - - - - bf9d9c5d by David Waern at 2010-09-02T19:27:48+00:00 H.Utils needs FFI on Win+MinGW - - - - - 048ae44a by Mark Lentczner at 2010-09-04T23:19:47+00:00 make TOC group header identifiers validate - - - - - 8c6faf36 by Simon Michael at 2010-09-22T07:12:34+00:00 add hints for cleaner darcs show authors output - - - - - 9909bd17 by Simon Michael at 2010-09-22T17:58:06+00:00 print haddock coverage info on stdout when generating docs A module's haddockable items are its exports and the module itself. The output is lightly formatted so you can align the :'s and sort for readability. - - - - - 6da72171 by David Waern at 2010-10-03T21:31:24+00:00 Style wibble - - - - - 2f8d8e4d by Tobias Brandt at 2010-08-27T07:01:21+00:00 adding the option to fully qualify identifiers - - - - - 833be6c6 by Tobias Brandt at 2010-08-27T15:50:28+00:00 adding support for local and relative name qualification - - - - - df15c4e9 by Tobias Brandt at 2010-08-27T15:56:37+00:00 corrected qualification help message - - - - - 449e9ce1 by David Waern at 2010-10-16T17:34:30+00:00 Solve conflicts - - - - - 3469bda5 by David Waern at 2010-10-16T18:42:40+00:00 Use "qual" as an abbreviation for qualification instead of "quali" for consistency - - - - - 97c2d728 by David Waern at 2010-10-16T18:47:07+00:00 Style police - - - - - ce14fbea by David Waern at 2010-10-16T21:15:25+00:00 Style police - - - - - fdf29e9d by David Waern at 2010-10-17T00:30:44+00:00 Add a pointer to the style guide - - - - - 8e6b44e8 by rrnewton at 2010-10-24T03:19:28+00:00 Change to index pages: include an 'All' option even when subdividing A-Z. - - - - - 755b131c by David Waern at 2010-11-14T19:39:36+00:00 Bump version - - - - - d0345a04 by David Waern at 2010-11-14T19:41:59+00:00 TAG 2.8.1 - - - - - f6221508 by Simon Peyton Jones at 2010-09-13T09:53:00+00:00 Adapt to minor changes in internal GHC functions - - - - - 1290713d by Ian Lynagh at 2010-09-15T10:37:18+00:00 Remove duplicate Outputable instance for Data.Map.Map - - - - - 87f69eef by Ian Lynagh at 2010-09-21T15:01:10+00:00 Bump GHC dep upper bound - - - - - af36e087 by Ian Lynagh at 2010-09-21T15:12:02+00:00 Fix up __GLASGOW_HASKELL__ tests - - - - - ad67716c by Ian Lynagh at 2010-09-21T20:31:35+00:00 Don't build haddock is HADDOCK_DOCS is NO - - - - - 63b3f1f5 by Ian Lynagh at 2010-09-21T21:39:51+00:00 Fixes for when HADDOCK_DOCS=NO - - - - - e92bfa42 by Ian Lynagh at 2010-09-29T21:15:38+00:00 Fix URL creation on Windows: Use / not \ in URLs. Fixes haskell/haddock#4353 - - - - - 66c55e05 by Ian Lynagh at 2010-09-30T17:03:34+00:00 Tidy up haddock symlink installation In particular, it now doesn't get created if we aren't installing haddock. - - - - - 549b5556 by Ian Lynagh at 2010-10-23T21:17:14+00:00 Follow extension-flattening change in GHC - - - - - d7c2f72b by David Waern at 2010-11-14T20:17:55+00:00 Bump version to 2.8.2 - - - - - 6989a3a9 by David Waern at 2010-11-14T20:26:01+00:00 Solve conflict - - - - - 055c6910 by Ian Lynagh at 2010-09-22T15:36:20+00:00 Bump GHC dep - - - - - c96c0763 by Simon Marlow at 2010-10-27T11:09:44+00:00 follow changes in the GHC API - - - - - 45907129 by David Waern at 2010-11-07T14:00:58+00:00 Update the HCAR entry - - - - - 61940b95 by David Waern at 2010-11-07T14:07:34+00:00 Make the HCAR entry smaller - - - - - aa590b7d by David Waern at 2010-11-14T21:30:59+00:00 Update HCAR entry with November 2010 version - - - - - 587f9847 by David Waern at 2010-11-14T23:48:17+00:00 Require ghc >= 7.0 - - - - - ff5c647c by David Waern at 2010-11-14T23:49:09+00:00 TAG 2.8.2 - - - - - 937fcb4f by David Waern at 2010-11-14T23:49:45+00:00 Solve conflict - - - - - 8e5d0c1a by David Waern at 2010-11-15T21:09:50+00:00 Remove code for ghc < 7 - - - - - 3d47b70a by David Waern at 2010-11-15T21:11:06+00:00 Fix bad merge - - - - - 7f4a0d8a by David Waern at 2010-11-15T21:13:57+00:00 Remove more ghc < 7 code - - - - - 9ee34b50 by David Waern at 2010-11-15T21:31:25+00:00 Match all AsyncExceptions in exception handler - - - - - 42849c70 by David Waern at 2010-11-15T21:35:31+00:00 Just say "internal error" instead of "internal Haddock or GHC error" - - - - - c88c809b by David Waern at 2010-11-15T21:44:19+00:00 Remove docNameOcc under the motto "don't name compositions" - - - - - b798fc7c by David Waern at 2010-11-15T23:27:13+00:00 Wibble - - - - - 2228197e by David Waern at 2010-11-15T23:28:24+00:00 Rename the HCAR entry file - - - - - 8a3f9090 by David Waern at 2010-11-16T00:05:29+00:00 Remove Haskell 2010 extensions from .cabal file - - - - - c7a0c597 by David Waern at 2010-11-16T00:10:28+00:00 Style wibbles - - - - - cde707a5 by David Waern at 2010-11-16T00:12:00+00:00 Remove LANGUAGE ForeignFunctionInterface pragmas - - - - - 1dbda8ed by David Waern at 2010-11-16T00:17:21+00:00 Make a little more use of DoAndIfThenElse - - - - - 4c45ff6e by David Waern at 2010-11-16T00:59:41+00:00 hlint police - - - - - d2feaf09 by David Waern at 2010-11-16T01:14:15+00:00 hlint police - - - - - 99876e97 by David Waern at 2010-11-20T19:06:00+00:00 Haddock documentation updates - - - - - 65ce6987 by David Waern at 2010-11-20T19:42:51+00:00 Follow the style guide closer in Haddock.Types and improve docs - - - - - 28ca304a by tob.brandt at 2010-11-20T17:04:40+00:00 add full qualification for undocumented names - - - - - d61341e3 by David Waern at 2010-11-20T20:04:15+00:00 Re-structure qualification code a little - - - - - 0057e4d6 by David Waern at 2010-11-20T20:07:55+00:00 Re-order functions - - - - - d7279afd by David Waern at 2010-11-21T03:39:54+00:00 Add BangPatterns to alex and happy source files - - - - - 629fe60e by tob.brandt at 2010-11-23T23:35:11+00:00 documentation for qualification - - - - - 37031cee by David Waern at 2010-11-23T21:06:44+00:00 Update CHANGES - don't mention 2.8.2, we won't release it - - - - - f2489e19 by David Waern at 2010-12-01T21:57:11+00:00 Update deps of runtests.hs to work with ghc 7.0.1 - - - - - d3657e9a by David Waern at 2010-12-01T22:04:57+00:00 Make tests compile with ghc 7.0.1 - - - - - a2f09d9b by David Waern at 2010-12-01T22:06:59+00:00 Update tests following version bump - - - - - 50883ebb by David Waern at 2010-12-06T14:09:18+00:00 Update tests following recent changes - - - - - fc2fadeb by David Waern at 2010-12-06T14:17:29+00:00 Add a flag --pretty-html for rendering indented html with newlines - - - - - 30832ef2 by David Waern at 2010-12-06T14:17:35+00:00 Use --pretty-html when running the test suite. Makes it easier to compare output - - - - - a0b81b31 by David Waern at 2010-12-06T14:18:27+00:00 Wibble - - - - - 3aaa23fe by David Waern at 2010-12-06T14:19:29+00:00 Haddockify ppHtml comments - - - - - 24bb24f0 by David Waern at 2010-12-06T14:23:15+00:00 Remove --debug. It was't used, and --verbosity should take its place - - - - - 6bc076e5 by David Waern at 2010-12-06T14:25:37+00:00 Rename golden-tests into html-tests. "golden tests" sounds strange - - - - - 53301e55 by David Waern at 2010-12-06T14:26:26+00:00 QUALI -> QUAL in the description --qual for consistency - - - - - 98b6affb by David Waern at 2010-12-06T21:54:02+00:00 Bump version - - - - - 371bf1b3 by David Waern at 2010-12-06T22:08:55+00:00 Update tests following version bump - - - - - 25be762d by David Waern at 2010-12-06T22:21:03+00:00 Update CHANGES - - - - - 7c7dac71 by David Waern at 2010-12-06T22:33:43+00:00 Update ANNOUNCE - - - - - 30d7a5f2 by Simon Peyton Jones at 2010-11-15T08:38:38+00:00 Alex generates BangPatterns, so make Lex.x accept them (It'd be better for Alex to generate this pragma.) - - - - - 605e8018 by Simon Marlow at 2010-11-17T11:37:24+00:00 Add {-# LANGUAGE BangPatterns #-} to mollify GHC - - - - - a46607ba by David Waern at 2010-12-07T14:08:10+00:00 Solve conflicts - - - - - b28cda66 by David Waern at 2010-12-09T20:41:35+00:00 Docs: Mention that \ is a special character in markup - - - - - a435bfdd by Ian Lynagh at 2010-11-17T14:01:19+00:00 TAG GHC 7.0.1 release - - - - - 5a15a05a by David Waern at 2010-12-11T17:51:19+00:00 Fix indentation problem - - - - - 4232289a by Lennart Kolmodin at 2010-12-17T18:32:03+00:00 Revise haddock.cabal given that we now require ghc-7 default-language should be Haskell2010, slight new semantics for extensions. Rewrite into clearer dependencies of base and Cabal. - - - - - a36302dc by David Waern at 2010-12-19T17:12:37+00:00 Update CHANGES - - - - - 7c8b85b3 by David Waern at 2010-12-19T17:14:24+00:00 Bump version - - - - - cff22813 by Ian Lynagh at 2011-01-05T18:24:27+00:00 Write hoogle output in utf8; fixes GHC build on Windows - - - - - c7e762ea by David Waern at 2011-01-22T00:00:35+00:00 Put title outside doc div when HTML:fying title+prologue Avoids indenting the title, and makes more sense since the title is not a doc string anyway. - - - - - 5f639054 by David Waern at 2011-01-22T16:09:44+00:00 Fix spelling error - contributed by Marco Silva - - - - - c11dce78 by Ian Lynagh at 2011-01-07T02:33:11+00:00 Follow GHC build system changes - - - - - 101cfaf5 by David Waern at 2011-01-08T14:06:44+00:00 Bump version - - - - - af62348b by David Waern at 2011-01-08T14:07:07+00:00 TAG 2.9.2 - - - - - 4d1f6461 by Ian Lynagh at 2011-01-07T23:06:57+00:00 Name the haddock script haddock-ghc-7.0.2 instead of haddock-7.0.2; haskell/haddock#4882 "7.0.2" looked like a haddock version number before - - - - - 8ee4d5d3 by Simon Peyton Jones at 2011-01-10T17:31:12+00:00 Update Haddock to reflect change in hs_tyclds field of HsGroup - - - - - 06f3e3db by Ian Lynagh at 2011-03-03T15:02:37+00:00 TAG GHC 7.0.2 release - - - - - 7de0667d by David Waern at 2011-03-10T22:47:13+00:00 Update CHANGES - - - - - 33a9f1c8 by David Waern at 2011-03-10T22:47:31+00:00 Fix build with ghc 7.0.1 - - - - - 4616f861 by David Waern at 2011-03-10T22:47:50+00:00 TAG 2.9.2-actual - - - - - 0dab5e3c by Simon Hengel at 2011-04-08T15:53:01+00:00 Set shell script for unit tests back to work - - - - - 85c54dee by Simon Hengel at 2011-04-08T16:01:24+00:00 Set unit tests back to work Here "ghci>" was still used instead of ">>>". - - - - - 1cea9b78 by Simon Hengel at 2011-04-08T16:25:36+00:00 Update runtests.hs for GHC 7.0.2 - - - - - 8e5b3bbb by Simon Hengel at 2011-04-08T16:28:49+00:00 Update Haddock version in *.html.ref - - - - - 2545e955 by Simon Hengel at 2011-04-08T17:09:28+00:00 Add support for blank lines in the result of examples Result lines that only contain the string "<BLANKLINE>" are treated as a blank line. - - - - - adf64d2e by Simon Hengel at 2011-04-08T17:36:50+00:00 Add documentation for "support for blank lines in the result of examples" - - - - - c51352ca by David Waern at 2011-05-21T23:57:56+00:00 Improve a haddock comment - - - - - 7419cf2c by David Waern at 2011-05-22T15:41:52+00:00 Use cabal's test suite support to run the test suite This gives up proper dependency tracking of the test script. - - - - - 7770070c by David Waern at 2011-05-22T01:45:44+00:00 We don't need to send DocOptions nor a flag to mkExportItems - - - - - 9d95b7b6 by David Waern at 2011-05-22T21:39:03+00:00 Fix a bug - - - - - 1f93699b by David Waern at 2011-05-22T21:40:21+00:00 Break out fullContentsOf, give it a better name and some documentation The documentation describes how we want this function to eventually behave, once we have fixed a few problems with the current implementation. - - - - - 9a86432f by David Waern at 2011-05-22T21:53:52+00:00 Fix some stylistic issues in mkExportItems - - - - - c271ff0c by David Waern at 2011-05-22T22:09:11+00:00 Indentation - - - - - 93e602b1 by David Waern at 2011-06-10T01:35:31+00:00 Add git commits since switchover: darcs format (followed by a conflict resolution): commit 6f92cdd12d1354dfbd80f8323ca333bea700896a Merge: f420cc4 28df3a1 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Thu May 19 17:54:34 2011 +0100 Merge remote branch 'origin/master' into ghc-generics commit 28df3a119f770fdfe85c687dd73d5f6712b8e7d0 Author: Max Bolingbroke <batterseapower at hotmail.com> Date: Sat May 14 22:37:02 2011 +0100 Unicode fix for getExecDir on Windows commit 89813e729be8bce26765b95419a171a7826f6d70 Merge: 6df3a04 797ab27 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 9 11:55:17 2011 +0100 Merge branch 'ghc-new-co' commit 6df3a040da3dbddee67c6e30a892f87e6b164383 Author: Ian Lynagh <igloo at earth.li> Date: Sun May 8 17:05:50 2011 +0100 Follow changes in SDoc commit f420cc48b9259f0b1afd2438b12f9a2bde57053d Author: Jose Pedro Magalhaes <jpm at cs.uu.nl> Date: Wed May 4 17:31:52 2011 +0200 Adapt haddock to the removal of HsNumTy and TypePat. commit 797ab27bdccf39c73ccad374fea265f124cb52ea Merge: 1d81436 5a91450 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:05:03 2011 +0100 Merge remote branch 'origin/master' into ghc-new-co commit 1d8143659a81cf9611668348e33fd0775c7ab1d2 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:03:46 2011 +0100 Wibbles for ghc-new-co branch commit 5a91450e2ea5a93c70bd3904b022445c9cc82488 Author: Ian Lynagh <igloo at earth.li> Date: Fri Apr 22 00:51:56 2011 +0100 Follow defaultDynFlags change in GHC - - - - - 498da5ae by David Waern at 2011-06-11T00:33:33+00:00 * Merge in git patch from Michal Terepeta >From 6fc71d067738ef4b7de159327bb6dc3d0596be29 Mon Sep 17 00:00:00 2001 From: Michal Terepeta <michal.terepeta at gmail.com> Date: Sat, 14 May 2011 19:18:22 +0200 Subject: [PATCH] Follow the change of TypeSig in GHC. This follows the change in GHC to make TypeSig take a list of names (instead of just one); GHC ticket haskell/haddock#1595. This should also improve the Haddock output in case the user writes a type signature that refers to many names: -- | Some comment.. foo, bar :: ... will now generate the expected output with one signature for both names. - - - - - 094607fe by Ian Lynagh at 2011-06-17T19:10:29+01:00 Fix build - - - - - 8fa35740 by Ian Lynagh at 2011-06-26T21:06:40+01:00 Bump GHC dep to allow 7.2 - - - - - e4d2ca3c by Ian Lynagh at 2011-07-07T23:06:28+01:00 Relax base dep - - - - - b948fde9 by Ian Lynagh at 2011-07-28T16:39:45+01:00 GHC build system: Don't install the datafiles twice - - - - - f82f6d70 by Simon Marlow at 2011-08-11T12:08:15+01:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - b341cc12 by Max Bolingbroke at 2011-08-22T20:25:27+01:00 Fix compilation with no-pred-ty GHC - - - - - 30494581 by Max Bolingbroke at 2011-08-23T10:20:54+01:00 Remaining fixes for PredTy removal - - - - - 0b197138 by Max Bolingbroke at 2011-08-26T08:27:45+01:00 Rename factKind to constraintKind - - - - - a379bec5 by Max Bolingbroke at 2011-09-04T12:54:47+01:00 Deal with change to IParam handling in GHC - - - - - f94e421b by Max Bolingbroke at 2011-09-06T17:34:31+01:00 Adapt Haddock for the ConstraintKind extension changes - - - - - 8821e5cc by Max Bolingbroke at 2011-09-09T08:24:59+01:00 Ignore associated type defaults (just as we ignore default methods) - - - - - 31a0afd4 by Max Bolingbroke at 2011-09-09T09:06:00+01:00 Merge branch 'no-pred-ty' of ssh://darcs.haskell.org/srv/darcs/haddock into no-pred-ty - - - - - dd3b530a by Max Bolingbroke at 2011-09-09T14:10:25+01:00 Merge branch 'no-pred-ty' Conflicts: src/Haddock/Convert.hs - - - - - 5f25ec96 by Max Bolingbroke at 2011-09-09T14:10:40+01:00 Replace FactTuple with ConstraintTuple - - - - - cd30b9cc by David Waern at 2011-09-26T02:17:55+02:00 Bump to version 2.9.3 - - - - - 4fbfd397 by Max Bolingbroke at 2011-09-27T14:55:21+01:00 Follow changes to BinIface Name serialization - - - - - 92257d90 by David Waern at 2011-09-30T23:45:07+02:00 Fix problem with test files not added to distribution tarball - - - - - 00255bda by David Waern at 2011-09-30T23:48:24+02:00 Merge branch 'development' - - - - - 5421264f by David Waern at 2011-10-01T01:25:39+02:00 Merge in darcs patch from Simon Meier: Wed Jun 1 19:41:16 CEST 2011 iridcode at gmail.com * prettier haddock coverage info The new coverage info rendering uses less horizontal space. This reduces the number of unnecessary line-wrappings. Moreover, the most important information, how much has been documented already, is now put up front. Hopefully, this makes it more likely that a library author is bothered by the low coverage of his modules and fixes that issue ;-) - - - - - 07d318ef by David Waern at 2011-10-01T01:34:10+02:00 Use printException instead of deprecated printExceptionAndWarnings - - - - - 40d52ee4 by David Waern at 2011-10-01T01:41:13+02:00 Merge in darcs pach: Mon Apr 11 18:09:54 JST 2011 Liyang HU <haddock at liyang.hu> * Remember collapsed sections in index.html / haddock-util.js - - - - - 279d6dd4 by David Waern at 2011-10-01T01:55:45+02:00 Merge in darcs patch: Joachim Breitner <mail at joachim-breitner.de>**20110619201645 Ignore-this: f6c51228205b0902ad5bfad5040b989a As reported on http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=578301, generating the global index takes much too long if type-level (with lots of auto-generated types) is installed. The patch avoids a quadratic runtime in the subfunction getIfaceIndex of ppHtmlIndex by using a temporary set. Runtime improvement observed here from 25.36s to 2.86s. - - - - - d1612383 by David Waern at 2011-10-01T01:56:48+02:00 Merge branch 'development' - - - - - 347520c1 by David Waern at 2011-10-01T01:56:54+02:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9a0c95e8 by David Waern at 2011-10-01T02:19:10+02:00 Improve .cabal file - - - - - 6967dc64 by Ian Lynagh at 2011-10-01T01:34:06+01:00 Follow changes to ForeignImport/ForeignExport in GHC - - - - - 565cb26b by Simon Marlow at 2011-10-04T00:15:04+02:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - 8b74f512 by David Waern at 2011-10-04T00:18:17+02:00 Requre ghc >= 7.2 - - - - - 271d360c by David Waern at 2011-10-04T00:22:50+02:00 Bump version to 2.9.4 - - - - - 37f3edb0 by David Waern at 2011-10-06T02:30:21+02:00 Add alex and happy to build-tools. - - - - - 7ac2bb6e by David Terei at 2011-10-12T14:02:55-07:00 Add safe haskell indication to haddock output - - - - - 42c91a47 by David Terei at 2011-10-12T14:06:03-07:00 Fix CSS issue with info table not being contained in module header - - - - - 0eddab6c by David Terei at 2011-10-12T14:06:58-07:00 Add safe haskell indication to haddock output - - - - - 3df058eb by David Terei at 2011-10-12T14:07:07-07:00 Fix CSS issue with info table not being contained in module header - - - - - a40a6c3f by David Waern at 2011-10-22T11:29:06+02:00 Bump .haddock file version since the format has changed recently - - - - - 8a6254be by David Waern at 2011-10-22T11:30:42+02:00 Merge branch 'development' - - - - - 642e3e02 by David Waern at 2011-10-23T21:23:39+02:00 Sort import list - - - - - 36371cf8 by David Waern at 2011-10-23T22:48:18+02:00 Remove NEW_GHC_LAYOUT conditional. - - - - - 5604b499 by David Waern at 2011-10-27T00:15:03+02:00 Add --print-ghc-path. - - - - - 463499fa by David Waern at 2011-10-27T00:16:22+02:00 Make testsuite able to find its dependencies automatically. - - - - - a3506172 by Ryan Newton at 2011-11-05T05:59:58-04:00 Improved declNames internal error. Added a case to handle DocD. - - - - - 001b8baf by David Waern at 2011-11-05T20:37:29+01:00 Rename copy.hs -> accept.hs. - - - - - 55d808d3 by David Waern at 2011-11-05T23:30:02+01:00 Fix build. - - - - - deb5c3be by David Waern at 2011-11-06T00:01:47+01:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9b663554 by David Waern at 2011-11-06T00:03:45+01:00 Merge https://github.com/rrnewton/haddock - - - - - 1abb0ff6 by David Waern at 2011-11-06T01:20:37+01:00 Use getDeclMainBinder instead of declNames. - - - - - 4b005c01 by David Waern at 2011-11-06T19:09:53+01:00 Fix build. - - - - - c2c51bc7 by Ian Lynagh at 2011-11-06T23:01:33+00:00 Remove -DNEW_GHC_LAYOUT in ghc.mk - - - - - f847d703 by Jose Pedro Magalhaes at 2011-11-11T09:07:39+00:00 New kind-polymorphic core This big patch implements a kind-polymorphic core for GHC. The current implementation focuses on making sure that all kind-monomorphic programs still work in the new core; it is not yet guaranteed that kind-polymorphic programs (using the new -XPolyKinds flag) will work. For more information, see http://haskell.org/haskellwiki/GHC/Kinds - - - - - 7d7c3b09 by Jose Pedro Magalhaes at 2011-11-16T21:42:22+01:00 Follow changes to tuple sorts in master - - - - - 8430e03e by Simon Peyton Jones at 2011-11-17T10:20:27+00:00 Remove redundant imports - - - - - d1b06832 by Ian Lynagh at 2011-11-19T01:33:21+00:00 Follow GHC build system change to the way we call rm - - - - - 9e2230ed by David Waern at 2011-11-24T15:00:24+01:00 Fix a bug in test runner and get rid of regex-compat dependency. - - - - - 52039b21 by David Waern at 2011-11-24T23:55:36+01:00 Avoid haskell98 dependency in test - - - - - 92e1220d by David Waern at 2011-11-25T00:03:33+01:00 Avoid depency on regex-compat also in accept.hs. - - - - - ddac6b6f by David Waern at 2011-11-25T02:13:38+01:00 Accept test output. - - - - - 5a720455 by David Waern at 2011-11-25T02:16:20+01:00 Some more changes to test scripts. - - - - - 170a9004 by David Waern at 2011-11-25T02:30:41+01:00 Add flag --interface-version. - - - - - d225576c by David Waern at 2011-11-25T02:39:26+01:00 Remove #ifs for older compiler versions. - - - - - f0d0a4f5 by David Waern at 2011-11-26T04:20:12+01:00 Give preference to type over data constructors for doc comment links at renaming time. Previously this was done in the backends. Also, warn when a doc comment refers to something that is in scope but which we don't have the .haddock file for. These changes mean we can make DocIdentifier [a] into DocIdentifier a. - - - - - eef0e776 by David Waern at 2011-11-26T17:01:06+01:00 Allow doc comments to link to out-of-scope things (#78). (A bug that should have been fixed long ago.) - - - - - 565ad529 by David Waern at 2011-11-26T19:56:21+01:00 Update tests. - - - - - fb3ce7b9 by David Waern at 2011-11-26T21:44:28+01:00 Cleanup. - - - - - d0328126 by David Waern at 2011-11-26T22:10:28+01:00 Fix module reference bug. - - - - - c03765f8 by David Waern at 2011-12-03T05:20:20+01:00 Slightly better behaviour on top-levels without type signatures. - Docs don't get attached to the next top-level with signature by mistake. - If there's an export list and the top-level is part of it, its doc comment shows up in the documentation. - - - - - 48461d31 by David Waern at 2011-12-03T05:38:10+01:00 Add a test for Unicode doc comments. - - - - - 549c4b4e by David Waern at 2011-12-03T19:07:55+01:00 Cleanup. - - - - - 7bfecf91 by David Waern at 2011-12-03T20:13:08+01:00 More cleanup. - - - - - 14fab722 by Ian Lynagh at 2011-12-12T21:21:35+00:00 Update dependencies and binaryInterfaceVersion - - - - - 469e6568 by Ian Lynagh at 2011-12-18T12:56:16+00:00 Fix (untested) building from source tarball without alex/happy haddock's .cabal file was declaring that it needed alex and happy to build, but in the GHC source tarballs it doesn't. - - - - - 895c9a8c by David Waern at 2011-12-27T12:57:43+01:00 Go back to having a doc, sub and decl map instead of one big decl map. This setup makes more sense since when we add value bindings to the processed declarations (for type inference), we will have multiple declarations which should share documentation. Also, we already have a separate doc map for instances which we can now merge into the main doc map. Another benefit is that we don't need the DeclInfo type any longer. - - - - - 736767d9 by David Waern at 2011-12-27T13:33:41+01:00 Merge ../../../haddock Conflicts: src/Haddock/InterfaceFile.hs - - - - - 20016f79 by David Waern at 2011-12-27T13:57:23+01:00 Bump version. - - - - - 31f276fb by David Waern at 2011-12-27T13:57:32+01:00 Merge ../ghc/utils/haddock - - - - - 95b367cd by David Waern at 2011-12-27T14:57:29+01:00 Update tests following version bump. - - - - - fa3c94cd by David Waern at 2011-12-27T14:57:51+01:00 Get rid of quite unnecessary use of different lists. - - - - - 9c4d3c54 by David Waern at 2011-12-27T15:26:42+01:00 Cleanup. - - - - - 2caf9f90 by David Waern at 2011-12-27T16:18:05+01:00 Wibbles. - - - - - 3757d09b by David Waern at 2011-12-27T20:50:26+01:00 Complete support for inferring types for top-level bindings. - - - - - 53418734 by David Waern at 2011-12-28T15:02:13+01:00 Minor fixes and cleanup. - - - - - 0c9d0385 by Ian Lynagh at 2012-01-03T18:31:29+00:00 Follow rename of Instance to ClsInst in GHC - - - - - c9bc969a by Simon Hengel at 2012-01-12T21:28:14+01:00 Make sure that generated xhtml is valid (close haskell/haddock#186) Thanks to Phyx. - - - - - 836a0b9a by David Waern at 2012-02-01T02:30:05+01:00 Fix bug introduced in my recent refactoring. - - - - - c7d733eb by David Waern at 2012-02-01T02:30:26+01:00 Cleanup mkMaps and avoid quadratic behaviour. - - - - - da3cda8f by David Waern at 2012-02-01T02:56:56+01:00 Require ghc >= 7.4. - - - - - 83a3287e by David Waern at 2012-02-01T02:57:36+01:00 Update CHANGES. - - - - - 93408f0b by Simon Hengel at 2012-02-04T00:48:04+01:00 Add reference renderings - - - - - 49d00d2c by Simon Hengel at 2012-02-04T00:48:25+01:00 Set unit tests for parser back to work - - - - - eb450980 by Simon Hengel at 2012-02-04T00:49:07+01:00 Add .gitignore - - - - - a841602c by Simon Hengel at 2012-02-04T00:49:16+01:00 Add .ghci file - - - - - 8861199d by Simon Hengel at 2012-02-04T00:49:29+01:00 tests/html-tests/copy.hs: Use mapM_ instead of mapM So we do net get a list of () on stdout when running with runhaskell. - - - - - b477d9b5 by Simon Hengel at 2012-02-04T00:49:46+01:00 Remove index files from golden tests - - - - - 9dbda34e by Simon Hengel at 2012-02-04T00:49:57+01:00 Add /tests/html-tests/tests/*index*.ref to .gitignore - - - - - a9434817 by Simon Hengel at 2012-02-04T00:50:04+01:00 Add DocWarning to Doc The Xhtml backend has special markup for that, Hoogle and LaTeX reuse what we have for DocEmphasis. - - - - - de2fb6fa by Simon Hengel at 2012-02-04T00:50:13+01:00 Add support for module warnings - - - - - 0640920e by Simon Hengel at 2012-02-04T00:50:21+01:00 Add tests for module warnings - - - - - 30ce0d77 by Simon Hengel at 2012-02-04T00:50:29+01:00 Add support for warnings - - - - - bb367960 by Simon Hengel at 2012-02-04T00:50:37+01:00 Add tests for warnings - - - - - 6af1dc2d by Simon Hengel at 2012-02-04T00:50:50+01:00 Expand type signatures in export list (fixes haskell/haddock#192) - - - - - a06cbf25 by Simon Hengel at 2012-02-04T00:51:04+01:00 Expand type signatures for modules without explicit export list - - - - - 57dda796 by Simon Hengel at 2012-02-04T00:51:15+01:00 Remove obsolete TODO - - - - - 270c3253 by David Waern at 2012-02-04T00:51:24+01:00 Fix issues in support for warnings. * Match against local names only. * Simplify (it's OK to map over the warnings). - - - - - 683634bd by David Waern at 2012-02-04T00:55:11+01:00 Some cleanup and make sure we filter warnings through exports. - - - - - 210cb4ca by David Waern at 2012-02-04T03:01:30+01:00 Merge branch 'fix-for-186' of https://github.com/sol/haddock into ghc-7.4 - - - - - e8db9031 by David Waern at 2012-02-04T03:07:51+01:00 Style police. - - - - - 261f9462 by David Waern at 2012-02-04T03:20:16+01:00 Update tests. - - - - - 823cfc7c by David Waern at 2012-02-04T03:21:12+01:00 Use mapM_ in accept.hs as well. - - - - - 873dd619 by David Waern at 2012-02-04T03:21:33+01:00 Remove copy.hs - use accept.hs instead. - - - - - 0e31a14a by David Waern at 2012-02-04T03:47:33+01:00 Use <> instead of mappend. - - - - - 2ff7544f by David Waern at 2012-02-04T03:48:55+01:00 Remove code for older ghc versions. - - - - - dacf2786 by David Waern at 2012-02-04T15:52:51+01:00 Clean up some code from last SoC project. - - - - - 00cbb117 by David Waern at 2012-02-04T21:43:49+01:00 Mostly hlint-inspired cleanup. - - - - - 7dc86cc2 by Simon Peyton Jones at 2012-02-06T09:14:41+00:00 Track changes in HsDecls - - - - - f91f82fe by Ian Lynagh at 2012-02-16T13:40:11+00:00 Follow changes in GHC caused by the CAPI CTYPE pragma - - - - - a0ea6b0b by Ian Lynagh at 2012-02-22T02:26:12+00:00 Follow changes in GHC - - - - - b23b07d1 by Simon Peyton Jones at 2012-03-02T16:36:41+00:00 Follow changes in data representation from the big PolyKinds commit - - - - - 43406022 by Simon Hengel at 2012-03-05T11:18:34+01:00 Save/restore global state for static flags when running GHC actions This is necessary if we want to run createInterfaces (from Documentation.Haddock) multiple times in the same process. - - - - - 9fba16fe by Paolo Capriotti at 2012-03-06T10:57:33+00:00 Update .gitignore. - - - - - a9325044 by Simon Peyton Jones at 2012-03-14T17:35:42+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - fd48065a by Iavor Diatchki at 2012-03-15T22:43:35-07:00 Add support for type-level literals. - - - - - 2e8206dd by Simon Peyton Jones at 2012-03-16T14:18:22+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - 93e13319 by Simon Peyton Jones at 2012-03-17T01:04:05+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock Conflicts: src/Haddock/Convert.hs - - - - - d253fa71 by Iavor Diatchki at 2012-03-19T20:12:18-07:00 Merge remote-tracking branch 'origin/master' into type-nats - - - - - fc40acc8 by Iavor Diatchki at 2012-03-19T20:31:27-07:00 Add a missing case for type literals. - - - - - fd2ad699 by Iavor Diatchki at 2012-03-24T13:28:29-07:00 Rename variable to avoid shadowing warning. - - - - - 9369dd3c by Simon Peyton Jones at 2012-03-26T09:14:23+01:00 Follow refactoring of TyClDecl/HsTyDefn - - - - - 38825ca5 by Simon Peyton Jones at 2012-03-26T09:14:37+01:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - 4324ac0f by David Waern at 2012-04-01T01:51:19+02:00 Disable unicode test. - - - - - 3165b750 by David Waern at 2012-04-01T01:51:34+02:00 Take reader environment directly from TypecheckedSource. - - - - - 213b644c by David Waern at 2012-04-01T01:55:20+02:00 Cleanup. - - - - - 3118b4ba by David Waern at 2012-04-01T02:16:15+02:00 Don't filter out unexported names from the four maps - fixes a regression. - - - - - d6524e17 by David Waern at 2012-04-01T02:40:34+02:00 Fix crash when using --qual. Naughty GHC API! - - - - - ea3c43d8 by Henning Thielemann at 2012-04-01T13:03:07+02:00 add QualOption type for distinction between qualification argument given by the user and the actual qualification for a concrete module - - - - - 5422ff05 by Henning Thielemann at 2012-04-01T16:25:02+02:00 emit an error message when the --qual option is used incorrectly - - - - - 026e3404 by David Waern at 2012-04-01T18:10:30+02:00 Don't crash on unicode strings in doc comments. - - - - - ce006632 by David Waern at 2012-04-01T20:13:35+02:00 Add test for --ignore-all-exports flag/ignore-exports pragma. - - - - - 6e4dd33c by David Waern at 2012-04-01T20:21:03+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.4 - - - - - 734ae124 by Henning Thielemann at 2012-04-01T20:22:10+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - 622f9ba5 by David Waern at 2012-04-01T21:26:13+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 55ce17cb by Henning Thielemann at 2012-04-01T22:03:25+02:00 'abbreviate' qualification style - basic support Currently we ignore the package a module is imported from. This means that a module import would shadow another one with the same module name from a different package. - - - - - c85314ef by David Waern at 2012-04-01T22:05:12+02:00 Check qualification option before processing modules. - - - - - ae4b626c by Henning Thielemann at 2012-04-02T00:19:36+02:00 abbreviated qualification: use Packages.lookupModuleInAllPackages for finding the package that a module belongs to - - - - - 60bdbcf5 by Henning Thielemann at 2012-04-02T00:25:31+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - df44301d by Henning Thielemann at 2012-04-02T00:29:05+02:00 qualification style 'abbreviated' -> 'aliased' - - - - - f4192a64 by David Waern at 2012-04-02T01:05:47+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 7ba09067 by David Terei at 2012-04-04T15:08:21-07:00 Fix reporting of modules safe haskell mode (#5989) - - - - - d0cc33d0 by David Terei at 2012-04-06T15:50:41+01:00 Fix reporting of modules safe haskell mode (#5989) - - - - - 6e3434c5 by Simon Peyton Jones at 2012-04-20T18:37:46+01:00 Track changes in HsSyn - - - - - 22014ed0 by Simon Peyton Jones at 2012-05-11T22:45:15+01:00 Follow changes to LHsTyVarBndrs - - - - - d9a07b24 by David Waern at 2012-05-15T01:46:35+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - a6c4ebc6 by David Waern at 2012-05-16T02:18:32+02:00 Update CHANGES. - - - - - 8e181d29 by David Waern at 2012-05-16T02:27:56+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - e358210d by David Waern at 2012-05-16T02:35:33+02:00 Mention the new aliased --qual mode in CHANGES. - - - - - efd36a28 by David Waern at 2012-05-16T21:33:13+02:00 Bump version number. - - - - - d6b3af14 by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for deprecated record field - - - - - 927f800e by Simon Hengel at 2012-05-17T19:08:20+02:00 Use >>= instead of fmap and join - - - - - 048b41d5 by Simon Hengel at 2012-05-17T19:08:20+02:00 newtype-wrap Doc nodes for things that may have warnings attached - - - - - e3a89fc3 by Simon Hengel at 2012-05-17T19:08:20+02:00 Attach warnings to `Documentation` type - - - - - 5d4cc43d by Simon Hengel at 2012-05-17T19:08:20+02:00 Simplify lookupWarning - - - - - cf8ae69d by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for haskell/haddock#205 - - - - - cb409b19 by Simon Peyton Jones at 2012-05-25T08:30:11+01:00 Follow changes in LHsTyVarBndrs - - - - - 2d5f4179 by Simon Hengel at 2012-05-26T19:21:29+02:00 Add Applicative instance for (GenRnM a) - - - - - e4373060 by Simon Hengel at 2012-05-26T19:21:33+02:00 Use a map for warnings, as suggested by @waern - - - - - 597a68c7 by Simon Hengel at 2012-05-27T08:48:24+02:00 Add an optional label to URLs - - - - - ef1ac7fe by Simon Hengel at 2012-05-27T08:48:24+02:00 Add support for hyperlink labels to parser - - - - - 41f2adce by Simon Hengel at 2012-05-27T08:48:24+02:00 Add golden test for hyperlinks - - - - - 83d5e764 by Simon Hengel at 2012-05-27T08:50:02+02:00 Use LANGUAGE pragmas instead of default-extensions in cabal file - - - - - ddb755e5 by Simon Hengel at 2012-05-27T08:50:02+02:00 Fix typo in comment - - - - - 110676b4 by Simon Hengel at 2012-05-27T08:50:02+02:00 Add a type signature for a where-binding - - - - - 7d9ba2a0 by Ian Lynagh at 2012-06-12T14:38:01+01:00 Follow changes in GHC - - - - - 47c704f2 by Ian Lynagh at 2012-06-12T18:52:16+01:00 Follow changes in GHC - - - - - e1efe1ab by Simon Peyton Jones at 2012-06-13T17:25:29+01:00 Follow changes for the implementation of implicit parameters - - - - - 69abc81c by Ian Lynagh at 2012-06-19T22:52:58+01:00 Follow changes in base - - - - - 9d074a21 by Paolo Capriotti at 2012-06-22T18:26:47+01:00 Use right docMap to get decl documentation. - - - - - e3292ef6 by Ian Lynagh at 2012-07-15T01:31:19+01:00 Follow changes in GHC - - - - - ceae56b0 by Ian Lynagh at 2012-07-16T21:22:48+01:00 Fix haddock following some GHC changes Passing _|_ as the Settings for defaultDynFlags no longer works well enough - - - - - 9df72735 by Paolo Capriotti at 2012-07-19T16:49:32+01:00 Forward port changes from stable. - - - - - 572f5fcf by Ian Lynagh at 2012-07-19T20:38:26+01:00 Merge branch 'master' of darcs.haskell.org:/srv/darcs//haddock - - - - - 9195aca4 by Paolo Capriotti at 2012-07-20T10:27:28+01:00 Update dependencies. - - - - - 33db3923 by Ian Lynagh at 2012-07-20T17:54:43+01:00 Build with GHC 7.7 - - - - - 925a2cea by David Waern at 2012-07-23T16:50:40+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.6 Conflicts: src/Haddock/InterfaceFile.hs - - - - - d710ef97 by David Waern at 2012-07-23T16:52:07+02:00 Bump version number. - - - - - eb0c2f83 by David Waern at 2012-07-23T16:57:58+02:00 Update CHANGES. - - - - - b3f56943 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Hide "internal" instances This fixes haskell/haddock#37 (http://trac.haskell.org/haddock/ticket/37) Precisely, we show an instance iff its class and all the types are exported by non-hidden modules. - - - - - a70aa412 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Tests for hiding instances (#37) - - - - - c0f4aa58 by Simon Hengel at 2012-07-27T13:00:13+03:00 Add an other test for hiding instances (#37) - - - - - a7ed6268 by Ian Lynagh at 2012-08-07T14:48:13+01:00 Follow changes in GHC - - - - - 0ab30d38 by Ian Lynagh at 2012-08-13T22:12:27+01:00 Improve haddock memory usage - - - - - 0eaa4e30 by Ian Lynagh at 2012-08-13T23:58:46+01:00 Improve haddock memory usage - - - - - 659d26cf by Ian Lynagh at 2012-08-14T13:16:48+01:00 Remove some temporary pragmas I accidentally recorded - - - - - d97fceb6 by Simon Hengel at 2012-08-25T13:19:34+02:00 Add missing dependency to library - - - - - 4c910697 by Simon Hengel at 2012-08-28T07:39:14+02:00 Move .ghci to project root - - - - - fc3c601a by Simon Hengel at 2012-08-28T07:39:14+02:00 accept.hs: Ignore some files - - - - - 1af9b984 by Simon Hengel at 2012-08-28T07:40:04+02:00 Update reference renderings (bump version) - - - - - 980dc253 by Simon Hengel at 2012-08-28T07:40:32+02:00 Update reference renderings (remove links for ()) - - - - - 33651dbf by Simon Hengel at 2012-08-28T07:41:50+02:00 Update documentation of `runInteractiveProcess` in reference rendering - - - - - 7ab25078 by David Waern at 2012-09-07T10:38:50+02:00 Merge branch 'hiddenInstances2' of http://github.com/feuerbach/haddock into ghc-7.6 - - - - - c3de3a4b by David Waern at 2012-09-07T14:29:27+02:00 Follow changes in GHC. - - - - - 298c43ac by David Waern at 2012-09-07T14:59:24+02:00 Update CHANGES. - - - - - e797993a by David Waern at 2012-09-07T15:21:30+02:00 Update ANNOUNCE. - - - - - d0b44790 by David Waern at 2012-09-07T15:22:43+02:00 Merge branch 'hidden-instances' into ghc-7.6 - - - - - 41a4adc8 by Simon Hengel at 2012-09-08T12:08:37+02:00 Update doc/README - - - - - 71ad1040 by Simon Hengel at 2012-09-08T12:17:17+02:00 Add documentation for URL labels - - - - - 9bb41afd by Simon Peyton Jones at 2012-09-20T18:14:26+01:00 Follow data type changes in the tc-untouchables branch Relating entirely to SynTyConRhs - - - - - b8139bfa by Simon Hengel at 2012-09-21T14:24:16+02:00 Disable Unicode test for now - - - - - a5fafdd7 by Simon Hengel at 2012-09-21T14:35:45+02:00 Update TypeOperators test for GHC 7.6.1 Type operators can't be used as type variables anymore! - - - - - 6ccf0025 by Simon Hengel at 2012-09-21T16:02:24+02:00 Remove (Monad (Either e)) instance from ref. rendering of CrossPackageDocs I do not really understand why the behavior changed, so I'll open a ticket, so that we can further investigate. - - - - - b5c6c138 by Ian Lynagh at 2012-09-27T02:00:57+01:00 Follow changes in GHC build system - - - - - b98eded0 by David Waern at 2012-09-27T15:37:02+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 76cc2051 by David Waern at 2012-09-27T15:48:19+02:00 Update hidden instances tests. - - - - - aeaa1c59 by David Waern at 2012-09-28T10:21:32+02:00 Make API buildable with GHC 7.6. - - - - - d76be1b0 by Simon Peyton Jones at 2012-09-28T15:57:05+01:00 Merge remote-tracking branch 'origin/master' into tc-untouchables - - - - - a1922af8 by David Waern at 2012-09-28T19:50:20+02:00 Fix spurious superclass constraints bug. - - - - - bc41bdbb by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove old examples - - - - - bed7d3dd by Simon Hengel at 2012-10-01T11:30:51+02:00 Adapt parsetests for GHC 7.6.1 - - - - - dcdb22bb by Simon Hengel at 2012-10-01T11:30:51+02:00 Add test-suite section for parsetests to cabal file + get rid of HUnit dependency - - - - - 1e5263c9 by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove test flag from cabal file This was not really used. - - - - - 4beee98b by David Waern at 2012-09-28T23:42:28+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 11dd2256 by Ian Lynagh at 2012-10-03T16:17:35+01:00 Follow change in GHC build system - - - - - fbd77962 by Simon Hengel at 2012-10-03T18:49:40+02:00 Remove redundant dependency from cabal file - - - - - 09218989 by Simon Hengel at 2012-10-04T16:03:05+02:00 Fix typo - - - - - 93a2d5f9 by Simon Hengel at 2012-10-04T16:11:41+02:00 Remove trailing whitespace from cabal file - - - - - c8b46cd3 by Simon Hengel at 2012-10-04T16:12:17+02:00 Export Haddock's main entry point from library - - - - - b411e77b by Simon Hengel at 2012-10-04T16:29:46+02:00 Depend on library for executable The main motivation for this is to increase build speed. In GHC's source tree the library is not build, but all modules are now required for the executable, so that GHC's validate will now detect build failures for the library. - - - - - f8f0979f by Simon Hengel at 2012-10-05T00:32:57+02:00 Set executable flag for Setup.lhs - - - - - dd045998 by Simon Hengel at 2012-10-07T16:44:06+02:00 Extend rather than set environment when running HTML tests On some platforms (e.g. ppc64) GHC requires gcc in the path. - - - - - 7b39c3ae by Simon Hengel at 2012-10-07T17:05:45+02:00 cross-package test: re-export IsString instead of Monad There is a monad instance for Q, which is not available on platforms that do not have GHCi support. This caused CrossPackageDocs to fail on those platforms. Re-exporting IsString should test the same thing, but it works on all platforms. - - - - - 0700c605 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Fix some warnings - - - - - f78eca79 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Make -Wall proof - - - - - 6beec041 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Use listToMaybe/fromMaybe instead of safeHead/maybe - - - - - 44b8ce86 by Ian Lynagh at 2012-10-08T21:59:46+01:00 Follow changes in GHC - - - - - 6da5f702 by Simon Hengel at 2012-10-09T11:16:19+02:00 Update .ghci - - - - - 9ac1a1b9 by Kazu Yamamoto at 2012-10-09T12:45:31+02:00 Add markup support for properties - - - - - 1944cb42 by Simon Hengel at 2012-10-09T12:45:31+02:00 Simplify lexing/parsing of properties In contrast to what we do for examples, we do not really need to capture the "prompt" here. - - - - - bffd8e62 by Simon Hengel at 2012-10-09T13:40:14+02:00 Add HTML test for properties - - - - - 2fe9c5cb by Simon Hengel at 2012-10-09T13:40:21+02:00 Add unit tests for properties - - - - - 874e361b by Simon Hengel at 2012-10-09T13:40:33+02:00 Bump interface version - - - - - 2506cc37 by Simon Hengel at 2012-10-09T15:15:04+02:00 Fix parser bug - - - - - 743d2b7d by Simon Hengel at 2012-10-09T15:31:06+02:00 Allow to load interface files with compatible versions - - - - - 981a1660 by Simon Hengel at 2012-10-10T10:32:05+02:00 Export more types from Documentation.Haddock (fixes haskell/haddock#216) - - - - - dff7dc76 by Simon Hengel at 2012-10-10T11:15:19+02:00 Update ANNOUNCE and CHANGES - - - - - edd2bb01 by Simon Hengel at 2012-10-10T11:22:50+02:00 Bump version - - - - - 5039163b by Simon Hengel at 2012-10-10T13:56:04+02:00 Fix typo in documentation - - - - - e4ce34da by Simon Hengel at 2012-10-10T14:28:35+02:00 Add documentation for properties - - - - - 9555ebca by Simon Hengel at 2012-10-11T10:49:04+02:00 Remove redundant if-defs, more source documentation - - - - - 87aa67e1 by Simon Hengel at 2012-10-11T12:32:51+02:00 Adapt cabal file - - - - - c44c1dee by Simon Hengel at 2012-10-11T12:41:58+02:00 Require ghc 7.6 - - - - - 8383bc34 by Simon Hengel at 2012-10-11T12:50:24+02:00 Bump version - - - - - 1030eb38 by Simon Hengel at 2012-10-11T12:55:44+02:00 Update ANNOUNCE and CHANGES - - - - - 74955088 by Simon Hengel at 2012-10-12T09:49:31+02:00 Improve note about `binaryInterfaceVersion` (thanks David) - - - - - ee30f6b7 by Simon Hengel at 2012-10-13T13:40:59+02:00 Update version in html tests, rpm spec file, and user manual - - - - - f2861f18 by Simon Hengel at 2012-10-13T14:40:33+02:00 Remove unused MonadFix constraint - - - - - dfdf1a74 by Simon Hengel at 2012-10-13T15:15:38+02:00 Minor code simplification - - - - - 4ecd1e70 by Simon Hengel at 2012-10-13T15:33:43+02:00 Increase code locality - - - - - f7df5cc9 by Simon Hengel at 2012-10-13T16:03:12+02:00 Minor code simplification - - - - - e737eb6e by Simon Hengel at 2012-10-13T19:03:04+02:00 Handle HsExplicitListTy in renameer (fixes haskell/haddock#213) - - - - - c2dc8f17 by Simon Hengel at 2012-10-13T20:46:31+02:00 Better error messages - - - - - 14d48b4c by Simon Hengel at 2012-10-14T00:21:07+02:00 Simplify RnM type - - - - - 6c2cc547 by Simon Hengel at 2012-10-14T00:23:35+02:00 Simplify lookupRn - - - - - bc77ce85 by Simon Hengel at 2012-10-14T01:51:32+02:00 Organize unite tests hierarchically - - - - - 2306d117 by Simon Hengel at 2012-10-14T10:34:58+02:00 Handle more cases in renameType - - - - - 8a864203 by Simon Hengel at 2012-10-14T11:47:59+02:00 Add mini_HiddenInstances.html.ref and mini_HiddenInstancesB.html.ref - - - - - 3a978eca by Simon Hengel at 2012-10-14T11:49:28+02:00 Add /tests/html-tests/output/ to .gitignore - - - - - db18888a by Simon Hengel at 2012-10-14T13:38:21+02:00 Allow haddock markup in deprecation messages - - - - - e7cfee9f by Simon Hengel at 2012-10-14T14:00:23+02:00 If parsing of deprecation message fails, include it verbatim - - - - - 242a85be by Simon Hengel at 2012-10-14T14:13:24+02:00 Add description for PruneWithWarning test - - - - - 43d33df1 by Simon Hengel at 2012-10-14T15:40:53+02:00 Minor formatting change - - - - - 22768c44 by Simon Hengel at 2012-10-14T16:03:43+02:00 Properly handle deprecation messages for re-exported things (fixes haskell/haddock#220) - - - - - cb4b9111 by Simon Hengel at 2012-10-14T17:30:28+02:00 Add build artifacts for documentation to .gitignore - - - - - 854cd8de by Simon Hengel at 2012-10-14T23:34:51+02:00 unit-tests: Improve readability Add IsString instance for (Doc RdrName) + use <> instead of DocAppend. - - - - - c4446d54 by Simon Hengel at 2012-10-14T23:37:21+02:00 unit-tests: Minor refactoring Rename parse to parseParas. - - - - - 04f2703c by Simon Hengel at 2012-10-15T00:36:42+02:00 Fix typo - - - - - 3d109e44 by Simon Hengel at 2012-10-15T10:30:07+02:00 Add description for DeprecatedReExport test - - - - - 84f0985c by Simon Hengel at 2012-10-15T14:54:19+02:00 Move resources to /resources directory - - - - - a5de7ca6 by Simon Hengel at 2012-10-15T15:46:18+02:00 Move HTML tests to directory /html-test/ - - - - - e21f727d by Simon Hengel at 2012-10-15T19:32:42+02:00 Move HTML reference renderings to /html-test/ref/ - - - - - 3a3c6c75 by Simon Hengel at 2012-10-15T19:32:42+02:00 Copy css, images, etc. on accept - - - - - 40ead6dc by Simon Hengel at 2012-10-15T19:32:42+02:00 Move unit tests to /test directory - - - - - 99a28231 by Simon Hengel at 2012-10-15T19:32:42+02:00 Fix Setup.lhs /usr/bin/runhaskell is not installed on all systems. - - - - - 95faf45e by Simon Hengel at 2012-10-15T19:32:42+02:00 Make test management scripts more robust * They are now independent from the current directory, and hence can be called from everywhere * On UNIX/Linux they can now be run as scripts - - - - - 027aaa2d by Simon Hengel at 2012-10-15T19:53:40+02:00 Add 'dev' flag to cabal file, that builds without -O2 That way --disable-optimization can be used, which decreases build time considerably. - - - - - e0266ede by Simon Hengel at 2012-10-15T20:03:43+02:00 Add test case for "spurious superclass constraints bug" - - - - - 52a2aa92 by Simon Hengel at 2012-10-15T20:28:55+02:00 Adapt accept.lhs, so that it ignores more index files - - - - - 53530781 by Simon Hengel at 2012-10-15T20:49:39+02:00 Rename html-test/runtests.lhs to html-test/run.lhs - - - - - 84518797 by Simon Hengel at 2012-10-15T20:49:39+02:00 Move source files for HTML tests to html-test/src - - - - - a911dc6c by Simon Hengel at 2012-10-15T20:49:39+02:00 Adapt output directory for HTML tests - - - - - d3c15857 by Ian Lynagh at 2012-10-16T16:54:43+01:00 Follow dopt->gopt rename - - - - - 956665a5 by Simon Hengel at 2012-10-18T08:42:48+02:00 Update html-test/README - - - - - 903b1029 by Simon Hengel at 2012-10-18T08:50:26+02:00 Use markdown for html-test/README - - - - - 150b4d63 by Ian Lynagh at 2012-10-18T16:36:00+01:00 Follow changes in GHC: 'flags' has been renamed 'generalFlags' - - - - - 41e04ff9 by Simon Hengel at 2012-11-28T09:54:35+01:00 Export missing types from Documentation.Haddock - - - - - 9be59237 by Ian Lynagh at 2012-11-30T23:20:47+00:00 Update dependencies - - - - - e06842f5 by Simon Hengel at 2012-12-07T20:58:05+01:00 Bump version - - - - - e3dbede0 by Simon Hengel at 2012-12-07T20:58:05+01:00 Add missing test files to cabal file (fixes haskell/haddock#230) - - - - - ee0dcca7 by Simon Hengel at 2012-12-07T20:58:05+01:00 Update CHANGES - - - - - 51601bdb by Simon Peyton Jones at 2012-12-19T17:28:35+00:00 Track changes in UNPACK pragma stuff - - - - - f2573bc1 by Richard Eisenberg at 2012-12-21T20:56:25-05:00 Implement overlapping type family instances. An ordered, overlapping type family instance is introduced by 'type instance where', followed by equations. See the new section in the user manual (7.7.2.2) for details. The canonical example is Boolean equality at the type level: type family Equals (a :: k) (b :: k) :: Bool type instance where Equals a a = True Equals a b = False A branched family instance, such as this one, checks its equations in order and applies only the first the matches. As explained in the note [Instance checking within groups] in FamInstEnv.lhs, we must be careful not to simplify, say, (Equals Int b) to False, because b might later unify with Int. This commit includes all of the commits on the overlapping-tyfams branch. SPJ requested that I combine all my commits over the past several months into one monolithic commit. The following GHC repos are affected: ghc, testsuite, utils/haddock, libraries/template-haskell, and libraries/dph. Here are some details for the interested: - The definition of CoAxiom has been moved from TyCon.lhs to a new file CoAxiom.lhs. I made this decision because of the number of definitions necessary to support BranchList. - BranchList is a GADT whose type tracks whether it is a singleton list or not-necessarily-a-singleton-list. The reason I introduced this type is to increase static checking of places where GHC code assumes that a FamInst or CoAxiom is indeed a singleton. This assumption takes place roughly 10 times throughout the code. I was worried that a future change to GHC would invalidate the assumption, and GHC might subtly fail to do the right thing. By explicitly labeling CoAxioms and FamInsts as being Unbranched (singleton) or Branched (not-necessarily-singleton), we make this assumption explicit and checkable. Furthermore, to enforce the accuracy of this label, the list of branches of a CoAxiom or FamInst is stored using a BranchList, whose constructors constrain its type index appropriately. I think that the decision to use BranchList is probably the most controversial decision I made from a code design point of view. Although I provide conversions to/from ordinary lists, it is more efficient to use the brList... functions provided in CoAxiom than always to convert. The use of these functions does not wander far from the core CoAxiom/FamInst logic. BranchLists are motivated and explained in the note [Branched axioms] in CoAxiom.lhs. - The CoAxiom type has changed significantly. You can see the new type in CoAxiom.lhs. It uses a CoAxBranch type to track branches of the CoAxiom. Correspondingly various functions producing and consuming CoAxioms had to change, including the binary layout of interface files. - To get branched axioms to work correctly, it is important to have a notion of type "apartness": two types are apart if they cannot unify, and no substitution of variables can ever get them to unify, even after type family simplification. (This is different than the normal failure to unify because of the type family bit.) This notion in encoded in tcApartTys, in Unify.lhs. Because apartness is finer-grained than unification, the tcUnifyTys now calls tcApartTys. - CoreLinting axioms has been updated, both to reflect the new form of CoAxiom and to enforce the apartness rules of branch application. The formalization of the new rules is in docs/core-spec/core-spec.pdf. - The FamInst type (in types/FamInstEnv.lhs) has changed significantly, paralleling the changes to CoAxiom. Of course, this forced minor changes in many files. - There are several new Notes in FamInstEnv.lhs, including one discussing confluent overlap and why we're not doing it. - lookupFamInstEnv, lookupFamInstEnvConflicts, and lookup_fam_inst_env' (the function that actually does the work) have all been more-or-less completely rewritten. There is a Note [lookup_fam_inst_env' implementation] describing the implementation. One of the changes that affects other files is to change the type of matches from a pair of (FamInst, [Type]) to a new datatype (which now includes the index of the matching branch). This seemed a better design. - The TySynInstD constructor in Template Haskell was updated to use the new datatype TySynEqn. I also bumped the TH version number, requiring changes to DPH cabal files. (That's why the DPH repo has an overlapping-tyfams branch.) - As SPJ requested, I refactored some of the code in HsDecls: * splitting up TyDecl into SynDecl and DataDecl, correspondingly changing HsTyDefn to HsDataDefn (with only one constructor) * splitting FamInstD into TyFamInstD and DataFamInstD and splitting FamInstDecl into DataFamInstDecl and TyFamInstDecl * making the ClsInstD take a ClsInstDecl, for parallelism with InstDecl's other constructors * changing constructor TyFamily into FamDecl * creating a FamilyDecl type that stores the details for a family declaration; this is useful because FamilyDecls can appear in classes but other decls cannot * restricting the associated types and associated type defaults for a * class to be the new, more restrictive types * splitting cid_fam_insts into cid_tyfam_insts and cid_datafam_insts, according to the new types * perhaps one or two more that I'm overlooking None of these changes has far-reaching implications. - The user manual, section 7.7.2.2, is updated to describe the new type family instances. - - - - - f788d0fb by Simon Peyton Jones at 2012-12-23T15:49:58+00:00 Track changes in HsBang - - - - - ca460a0c by Simon Peyton Jones at 2012-12-23T15:50:28+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - f078fea6 by Simon Peyton Jones at 2013-01-02T08:33:13+00:00 Use InstEnv.instanceSig rather than instanceHead (name change) - - - - - 88e41305 by Simon Peyton Jones at 2013-01-14T17:10:27+00:00 Track change to HsBang type - - - - - e1ad4e19 by Kazu Yamamoto at 2013-02-01T11:59:24+09:00 Merge branch 'ghc-7.6' into ghc-7.6-merge-2 Conflicts: haddock.cabal src/Haddock/Interface/AttachInstances.hs src/Haddock/Interface/Create.hs src/Haddock/Interface/LexParseRn.hs src/Haddock/InterfaceFile.hs src/Haddock/Types.hs Only GHC HEAD can compile this. GHC 7.6.x cannot compile this. Some test fail. - - - - - 62bec012 by Kazu Yamamoto at 2013-02-06T11:12:28+09:00 Using tcSplitSigmaTy in instanceHead' (FIXME is resolved.) - - - - - 013fd2e4 by Kazu Yamamoto at 2013-02-06T17:56:21+09:00 Refactoring instanceHead'. - - - - - 3148ce0e by Kazu Yamamoto at 2013-02-07T17:45:10+09:00 Using new syntax in html-test/src/GADTRecords.hs. - - - - - 626dabe7 by Gabor Greif at 2013-02-15T22:42:01+01:00 Typo - - - - - 1eb667ae by Ian Lynagh at 2013-02-16T17:02:07+00:00 Follow changes in base - - - - - 3ef8253a by Ian Lynagh at 2013-03-01T23:23:57+00:00 Follow changes in GHC's build system - - - - - 1a265a3c by Ian Lynagh at 2013-03-03T23:12:07+00:00 Follow changes in GHC build system - - - - - 69941c79 by Max Bolingbroke at 2013-03-10T09:38:28-07:00 Use Alex 3's Unicode support to properly lex source files as UTF-8 Signed-off-by: David Waern <david.waern at gmail.com> - - - - - ea687dad by Simon Peyton Jones at 2013-03-15T14:16:10+00:00 Adapt to tcRnGetInfo returning family instances too This API change was part of the fix to Trac haskell/haddock#4175. But it offers new information to Haddock: the type-family instances, as well as the class instances, of this type. This patch just drops the new information on the floor, but there's an open opportunity to use it in the information that Haddock displays. - - - - - 971a30b0 by Andreas Voellmy at 2013-05-19T20:47:39+01:00 Fix for haskell/haddock#7879. Changed copy of utils/haddock/html/resources/html to use "cp -RL" rather than "cp -R". This allows users to run validate in a build tree, where the build tree was setup using lndir with a relative path to the source directory. - - - - - 31fb7694 by Ian Lynagh at 2013-05-19T20:47:49+01:00 Use "cp -L" when making $(INPLACE_LIB)/latex too - - - - - e9952233 by Simon Hengel at 2013-06-01T18:06:50+02:00 Add -itest to .ghci - - - - - b06873b3 by Mateusz Kowalczyk at 2013-06-01T18:06:50+02:00 Workaround for a failing build with --enable-tests. - - - - - e7858d16 by Simon Hengel at 2013-06-01T19:29:28+02:00 Fix broken test - - - - - 0690acb1 by Richard Eisenberg at 2013-06-21T14:08:25+01:00 Updates to reflect changes in HsDecls to support closed type families. - - - - - 7fd347ec by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 53ed81b6 by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 931c4f4f by Richard Eisenberg at 2013-07-24T13:15:59+01:00 Remove (error "synifyKind") to use WithinType, to allow haddock to process base. - - - - - 55a9c804 by Richard Eisenberg at 2013-08-02T15:54:55+01:00 Changes to reflect changes in GHC's type HsTyVarBndr - - - - - b6e9226c by Mathieu Boespflug at 2013-08-04T10:39:43-07:00 Output Copright and License keys in Xhtml backend. This information is as relevant in the documentation as it is in the source files themselves. Signed-off-by: David Waern <david.waern at gmail.com> - - - - - 4c66028a by David Waern at 2013-08-04T15:27:36-07:00 Bump interface file version. - - - - - 67340163 by David Waern at 2013-08-09T16:12:51-07:00 Update tests. - - - - - 2087569b by Mateusz Kowalczyk at 2013-08-25T09:24:13+02:00 Add spec tests. This adds tests for all elements we can create during regular parsing. This also adds tests for text with unicode in it. - - - - - 97f36a11 by Mateusz Kowalczyk at 2013-08-27T06:59:12+01:00 Fix ticket haskell/haddock#247. I do the same thing that the XHTML backend does: give these no special treatment and just act as if they are regular functions. - - - - - 60681b4f by Mateusz Kowalczyk at 2013-08-27T21:22:48+02:00 LaTeX tests setup - - - - - fa4c27b2 by Mateusz Kowalczyk at 2013-09-02T23:21:43+01:00 Fixes haskell/haddock#253 - - - - - 1a202490 by Mateusz Kowalczyk at 2013-09-03T01:12:50+01:00 Use Hspec instead of nanospec This is motivated by the fact that Haddock tests are not ran by the GHC's ‘validate’ script so we're pretty liberal on dependencies in that area. Full Hspec gives us some nice features such as Quickcheck integration. - - - - - 8cde3b20 by David Luposchainsky at 2013-09-08T07:27:28-05:00 Fix AMP warnings Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - d10661f2 by Herbert Valerio Riedel at 2013-09-11T15:15:01+02:00 Update Git repo URL in `.cabal` file - - - - - 16a44eb5 by Richard Eisenberg at 2013-09-17T09:34:26-04:00 Revision to reflect new role annotation syntax in GHC. - - - - - 4b9833b9 by Herbert Valerio Riedel at 2013-09-18T10:15:28+02:00 Add missing `traverse` method for `GenLocated` As `Traversable` needs at least one of `traverse` or `sequenceA` to be overridden. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - b71fed5d by Simon Hengel at 2013-09-18T22:43:34+02:00 Add test helper - - - - - 4fc1ea86 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#231 - - - - - 435872f6 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#256 We inject -dynamic-too into flags before we run all our actions in the GHC monad. - - - - - b8b24abb by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Add new field to DynFlags - - - - - 49558795 by Simon Hengel at 2013-09-18T22:43:35+02:00 Fallback to ./resources when Cabal data is not found (so that themes are found during development) - - - - - bf79d05c by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Fixes haskell/haddock#5 - - - - - e1baebc2 by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Print missing documentation. Fixes haskell/haddock#258. - - - - - 02ea74de by Austin Seipp at 2013-10-09T10:52:22-05:00 Don't consider StaticFlags when parsing arguments. Instead, discard any static flags before parsing the command line using GHC's DynFlags parser. See http://ghc.haskell.org/trac/ghc/ticket/8276 Based off a patch from Simon Hengel. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 704fd5bb by Simon Hengel at 2013-11-09T00:15:13+01:00 Update HTML tests - - - - - f9fed49e by Simon Hengel at 2013-11-10T18:43:58+01:00 Bump version - - - - - 97ae1999 by Simon Peyton Jones at 2013-11-25T17:25:14+00:00 Track changes in HsSpliceTy data constructor - - - - - 59ad8268 by Simon Peyton Jones at 2014-01-10T18:17:43+00:00 Adapt to small change in Pretty's exports - - - - - 8b12e6aa by Simon Hengel at 2014-01-12T14:48:35-06:00 Some code simplification by using traverse - - - - - fc5ea9a2 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix warnings in test helper - - - - - 6dbb3ba5 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Add ByteString version of Attoparsec - - - - - 968d7774 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 One pass parser and tests. We remove the HTML test as it is no longer necessary. We cover the test case in spec tests and other HTML tests but keeping this around fails: this is because the new parser has different semantics there. In fact, I suspect the original behaviour was a bug that wasn't caught/fixed but simply included as-is during the testing. - - - - - 37a07c9c by Simon Hengel at 2014-01-12T14:48:35-06:00 Rename Haddock.ParseSpec to Haddock.ParserSpec - - - - - f0f68fe9 by Simon Hengel at 2014-01-12T14:48:35-06:00 Don't append newline to parseString input We also check that we have parsed everything with endOfInput. - - - - - 95d60093 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix totality, unicode, examples, paragraph parsing Also simplify specs and parsers while we're at it. Some parsers were made more generic. This commit is a part of GHC pre-merge squash, email fuuzetsu at fuuzetsu.co.uk if you need the full commit history. - - - - - 7d99108c by Simon Hengel at 2014-01-12T14:48:35-06:00 Update acceptance tests - - - - - d1b59640 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Support for bold. Conflicts: src/Haddock/Backends/Hoogle.hs src/Haddock/Interface/Rename.hs src/Haddock/Parser.hs - - - - - 4b412b39 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Allow for headings inside function documentation. LaTeX will treat the h3-h6 headings the same as we'd have to hack the style file heavily otherwise and it would make the headings tiny anyway. Hoogle upstream said they will put in the functionality on their end. Conflicts: src/Haddock/Interface/Rename.hs src/Haddock/Types.hs test/Haddock/ParserSpec.hs - - - - - fdcca428 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Per-module extension flags and language listing. Any extensions that are not enabled by a used language (Haskell2010 &c) will be shown. Furthermore, any implicitly enabled are also going to be shown. While we could eliminate this either by using the GHC API or a dirty hack, I opted not to: if a user doesn't want the implied flags to show, they are recommended to use enable extensions more carefully or individually. Perhaps this will encourage users to not enable the most powerful flags needlessly. Enabled with show-extensions. Conflicts: src/Haddock/InterfaceFile.hs - - - - - 368942a2 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Bump interface version There were some breaking changes over the last few patches so we bump the interface version. This causes a big problem with testing: 1. To generate cross package docs, we first need to generate docs for the package used. 2. To generate package docs with new interface version, we need to use Haddock which has the version bumped. 3. To get Haddock with the version bump, we first need to test cross package docs 4. GOTO 1 So the problem is the chicken and the egg problem. It seems that the only solution would be to generate some interface files on the fly but it is non-trivial. To run this test, you'll have to: * build Haddock without the test (make sure everything else passes) * rebuild the packages used in the test with your shiny new binary making sure they are visible to Haddock * remove the ‘_hidden’ suffix and re-run the tests Note: because the packages currently used for this test are those provided by GHC, it's probably non-trivial to just re-build them. Preferably something less tedious to rebuild should be used and something that is not subject to change. - - - - - 124ae7a9 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow for nesting of paragraphs under lists. The nesting rules are similar to Markdown's with the exception that we can not simply indent the first line of a hard wrapped indented paragraph and have it treated as if it was fully indented. The reason is differences in markup as some of our constructs care about whitespace while others just swallow everything up so it's just a lot easier to not bother with it rather than making arbitrary rules. Note that we now drop trailing for string entities inside of lists. They weren't needed and it makes the output look uniform whether we use a single or double newline between list elements. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - c7913535 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow escaping in URLs and pictures. Some tests were moved under parseString as they weren't about paragraph level markup. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - 32326680 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update documentation. - - - - - fbef6406 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update maintainer - - - - - b40e82f4 by Mateusz Kowalczyk at 2014-01-13T02:39:25-06:00 Fixes haskell/haddock#271 Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - f4eafbf8 by Gergő Érdi at 2014-01-19T15:35:16-06:00 Support for -XPatternSynonyms Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - a8939591 by Austin Seipp at 2014-01-29T08:09:04-06:00 Update CPP check for __GLASGOW_HASKELL__ Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 30d7e9d5 by Gergő Érdi at 2014-01-31T00:15:01+08:00 <+>: Don't insert a space when concatenating empty nodes - - - - - a25ccd4d by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Fix @ code blocks In cases where we had some horizontal space before the closing ‘@’, the parser would not accept the block as a code block and we'd get ugly output. - - - - - 0f67305a by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Update tests This updates tests due to Haddock Trac haskell/haddock#271 fix and due to removal of TypeHoles as an extension from GHC. - - - - - 157322a7 by Gergő Érdi at 2014-01-31T01:03:17+08:00 Handle infix vs prefix names correctly everywhere, by explicitly specifying the context The basic idea is that "a" and "+" are either pretty-printed as "a" and "(+)" or "`a`" and "+" - - - - - aa6d9685 by Mateusz Kowalczyk at 2014-01-30T17:21:50+00:00 Correct whitespace in ‘hidden’ test for <+> change - - - - - 121872f0 by Mateusz Kowalczyk at 2014-02-09T17:59:12+00:00 Document module header. Fixes Haddock Trac haskell/haddock#270. - - - - - e3253746 by Mateusz Kowalczyk at 2014-02-10T21:37:48+00:00 Insert a space between module link and description Fixes Haddock Trac haskell/haddock#277. - - - - - 771d2384 by Mateusz Kowalczyk at 2014-02-10T23:27:21+00:00 Ensure a space between type signature and ‘Source’ This is briefly related to Haddock Trac haskell/haddock#249 and employs effectively the suggested fix _but_ it doesn't actually fix the reported issue. This commit simply makes copying the full line a bit less of a pain. - - - - - 8cda9eff by nand at 2014-02-11T15:48:30+00:00 Add support for type/data families This adds support for type/data families with their respective instances, as well as closed type families and associated type/data families. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 3f22c510 by nand at 2014-02-11T15:53:50+00:00 Improve display of poly-kinded type operators This now displays them as (==) k a b c ... to mirror GHC's behavior, instead of the old (k == a) b c ... which was just wrong. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - effb2d6b by nand at 2014-02-11T15:56:50+00:00 Add test case for PatternSynonyms This just tests various stuff including poly-kinded patterns and operator patterns to make sure the rendering isn't broken. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - b38faf0d by Niklas Haas at 2014-02-13T21:53:32+00:00 Get rid of re-implementation of sortBy I have no idea what this was doing lying around here, and due to the usage of tuples it's actually slower, too. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - ac1e0413 by Mateusz Kowalczyk at 2014-02-13T23:57:16+00:00 Only warn about missing docs when docs are missing This fixes the ‘Missing documentation for…’ message for modules with 100% coverage. - - - - - cae2e36a by Niklas Haas at 2014-02-15T21:56:18+00:00 Add test case for inter-module type/data family instances These should show up in every place where the class is visible, and indeed they do right now. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 8bea5c3a by Mateusz Kowalczyk at 2014-02-19T05:11:34+00:00 Use a bespoke data type to indicate fixity This deals with what I imagine was an ancient TODO and makes it much clearer what the argument actually does rather than having the user chase down the comment. - - - - - 5b52d57c by Niklas Haas at 2014-02-22T21:31:03+01:00 Strip a single leading space from bird tracks (#201) This makes bird tracks in the form > foo > bar > bat parse as if they had been written as >foo >bar >bat ie. without the leading whitespace in front of every line. Ideally we also want to look into how leading whitespace affects code blocks written using the @ @ syntax, which are currently unaffected by this patch. - - - - - 5a1315a5 by Simon Hengel at 2014-02-22T21:55:35+01:00 Turn a source code comment into specs - - - - - 784cfe58 by Mateusz Kowalczyk at 2014-02-23T05:02:22+00:00 Update test case for lifted GADT type rendering The parsing of these seems to have been fixed by GHC folk and it now renders differently. IMHO it now renders in a better way so I'm updating the test to reflect this. - - - - - c3c88c2f by Mateusz Kowalczyk at 2014-02-23T06:37:14+00:00 Don't shadow ‘strip’. -Wall complains - - - - - 293031d8 by Niklas Haas at 2014-02-23T15:21:52+01:00 Make ImplicitParams render correctly (#260) This introduces a new precedence level for single contexts (because implicit param contexts always need parens around them, but other types of contexts don't necessarily, even when alone) - - - - - 4200842d by Niklas Haas at 2014-02-23T15:37:13+01:00 Lower precedence of equality constraints This drops them to the new precedence pREC_CTX, which makes single eqaulity constraints show up as (a ~ b) => ty, in line with GHC's rendering. Additional tests added to make sure other type operators render as intended. Current behavior matches GHC - - - - - b59e3227 by Niklas Haas at 2014-02-23T16:11:22+01:00 Add RankNTypes test case to ImplicitParams.hs This test actually tests what haskell/haddock#260 originally reported - I omitted the RankNTypes scenario from the original fix because I realized it's not relevant to the underlying issue and indeed, this renders as intended now. Still good to have more tests. - - - - - c373dbf7 by Mateusz Kowalczyk at 2014-02-24T06:09:54+00:00 Fix rendering of Contents when links are present Fixes Haddock Trac haskell/haddock#267. - - - - - 9ecb0e56 by Mateusz Kowalczyk at 2014-02-24T06:26:50+00:00 Fix wording in the docs - - - - - 4f4dcd8e by Mateusz Kowalczyk at 2014-02-27T03:00:33+00:00 Change rendering of duplicate record field docs See Haddock Trac haskell/haddock#195. We now change this behaviour to only rendering the documentation attached to the first instance of a duplicate field. Perhaps we could improve this by rendering the first instance that has documentation attached to it but for now, we'll stick with this. - - - - - ad8aa609 by Niklas Haas at 2014-03-08T09:43:26+01:00 Render fixity information Affects functions, type synonyms, type families, class names, data type names, constructors, data families, associated TFs/DFs, type synonyms, pattern synonyms and everything else I could think of. - - - - - 6a39c917 by Niklas Haas at 2014-03-09T07:43:39+01:00 Reorder topDeclElem to move the source/wiki links to the top They appear in the same position due to the float: right attribute but now they're always at the top of the box instead of at the bottom. - - - - - 2d34b3b4 by Niklas Haas at 2014-03-09T07:53:46+01:00 Use optLast instead of listToMaybe for sourceUrls/wikiUrls This lets you override them using eg. cabal haddock --haddock-options, which can come in handy if you want to use a different layout or URL for your source code links than cabal-install generates. - - - - - 0eff4624 by Niklas Haas at 2014-03-09T07:53:46+01:00 Differentiate between TH splices (line-links) and regular names This adds a new type of source code link, to a specific line rather than a specific declaration/name - this is used to link to the location of a TH splice that defines a certain name. Rather hefty changes throughout and still one unresolved issue (the line URLs aren't parsed from the third form of --read-interface which means they're currently restricted to same-interface links). Not sure if this issue is really worth all the hassle, especially since we could just use line links in general. This commit also contains some cleanup/clarification of the types in Haddock.Backends.Xhtml.Decl and shortens some overlong lines in the process. Notably, the Bool parameter was replaced by a Unicode type synonym to help clarify its presence in type signatures. - - - - - 66d6f77b by Niklas Haas at 2014-03-09T20:02:43+01:00 Group similar fixities together Identical fixities declared for the same line should now render using syntax like: infix 4 <, >=, >, <= - - - - - 6587f9f5 by Mateusz Kowalczyk at 2014-03-10T04:24:18+00:00 Update changelog - - - - - 7387ddad by Niklas Haas at 2014-03-11T10:26:04+01:00 Include fixity information in the Interface file This resolves fixity information not appearing across package borders. The binary file version has been increased accordingly. - - - - - ab46ef44 by Niklas Haas at 2014-03-11T10:26:04+01:00 Update changelog - - - - - 565cab6f by Niklas Haas at 2014-03-11T10:26:04+01:00 Update appearance of fixity annotations This moves them in-line with their corresponding lines, similar to a presentation envision by @hvr and described in #ghc. Redundant operator names are also omitted when no ambiguity is present. - - - - - 5d7afd67 by Niklas Haas at 2014-03-11T10:26:05+01:00 Filter family instances of hidden types Currently, this check does not extend to hidden right hand sides, although it probably should hide them in that case. - - - - - ec291b0c by Niklas Haas at 2014-03-11T10:26:05+01:00 Add documentation for --source-entity-line - - - - - 0922e581 by Niklas Haas at 2014-03-11T10:37:32+01:00 Revert "Reorder topDeclElem to move the source/wiki links to the top" This reverts commit 843c42c4179526a2ad3526e4c7d38cbf4d50001d. This change is no longer needed with the new rendering style, and it messes with copy/pasting lines. - - - - - 30618e8b by Mateusz Kowalczyk at 2014-03-11T09:41:07+00:00 Bump version to 2.15.0 - - - - - adf3f1bb by Mateusz Kowalczyk at 2014-03-11T09:41:09+00:00 Fix up some whitespace - - - - - 8905f57d by Niklas Haas at 2014-03-13T19:18:06+00:00 Hide RHS of TFs with non-exported right hand sides Not sure what to do about data families yet, since technically it would not make a lot of sense to display constructors that cannot be used by the user. - - - - - 5c44d5c2 by Niklas Haas at 2014-03-13T19:18:08+00:00 Add UnicodeSyntax alternatives for * and -> I could not find a cleaner way to do this other than checking for string equality with the given built-in types. But seeing as it's actually equivalent to string rewriting in GHC's implementation of UnicodeSyntax, it's probably fitting. - - - - - b04a63e6 by Niklas Haas at 2014-03-13T19:18:10+00:00 Display minimal complete definitions for type classes This corresponds to the new {-# MINIMAL #-} pragma present in GHC 7.8+. I also cleaned up some of the places in which ExportDecl is used to make adding fields easier in the future. Lots of test cases have been updated since they now render with minimality information. - - - - - a4a20b16 by Niklas Haas at 2014-03-13T19:18:12+00:00 Strip links from recently added html tests These were accidentally left there when the tests were originally added - - - - - d624f315 by Mateusz Kowalczyk at 2014-03-13T19:19:31+00:00 Update changelog - - - - - d27a21ac by Mateusz Kowalczyk at 2014-03-13T21:19:07+00:00 Always read in prologue files as UTF8 (#286). - - - - - 54b2fd78 by Mateusz Kowalczyk at 2014-03-13T21:28:09+00:00 Style only - - - - - fa4fe650 by Simon Hengel at 2014-03-15T09:04:18+01:00 Add Fuuzetsu maintainers field in cabal file - - - - - f83484b7 by Niklas Haas at 2014-03-15T18:20:24+00:00 Hide minimal definition for only-method classes Previously this was not covered by the All xs check since here it is not actually an All, rather a single Var n. This also adds the previously missing html-test/src/Minimal.hs. - - - - - 0099d276 by Niklas Haas at 2014-03-15T18:20:26+00:00 Fix issue haskell/haddock#281 This is a regression from the data family instances change. Data instances are now distinguished from regular lists by usage of the new class "inst", and the style has been updated to only apply to those. I've also updated the appropriate test case to test this a bit better, including GADT instances with GADT-style records. - - - - - 1f9687bd by Mateusz Kowalczyk at 2014-03-21T17:48:37+00:00 Please cabal sdist - - - - - 75542693 by Mateusz Kowalczyk at 2014-03-22T16:36:16+00:00 Drop needless --split-objs which slows us down. Involves tiny cleanup of all the dynflag bindings. Fixes haskell/haddock#292. - - - - - 31214dc3 by Herbert Valerio Riedel at 2014-03-23T18:01:01+01:00 Fix a few typos Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 0b73e638 by Mateusz Kowalczyk at 2014-03-31T05:34:36+01:00 Print kind signatures on GADTs - - - - - 2bab42f3 by Mateusz Kowalczyk at 2014-03-31T16:53:25+01:00 Add default for new PlatformConstraints field - - - - - 42647c5f by Mateusz Kowalczyk at 2014-03-31T18:29:04+01:00 Drop leading whitespace in @-style blocks. Fixes haskell/haddock#201. - - - - - 98208294 by Niklas Haas at 2014-03-31T20:09:58+02:00 Crash when exporting record selectors of data family instances This fixes bug haskell/haddock#294. This also fixes a related but never-before-mentioned bug about the display of GADT record selectors with non-polymorphic type signatures. Note: Associated data type constructors fail to show up if nothing is exported that they could be attached to. Exporting any of the data types in the instance head, or the class + data family itself, causes them to show up, but in the absence of either of these, exporting just the associated data type with the constructor itself will result in it being hidden. The only scenario I can come up that would involve this kind of situation involved OverlappingInstances, and even then it can be mitigated by just exporting the class itself, so I'm not going to solve it since the logic would most likely be very complicated. - - - - - 3832d171 by Mateusz Kowalczyk at 2014-04-01T19:07:33+01:00 Make CHANGES consistent with what's now in 2.14.2 - - - - - c386ae89 by Mateusz Kowalczyk at 2014-04-01T19:18:36+01:00 Actually bundle extra spec tests in sdist - - - - - bd57a6d3 by Mateusz Kowalczyk at 2014-04-03T21:13:48+01:00 Update test cases for GHC bug haskell/haddock#8945, Haddock haskell/haddock#188 The order of signature groups has been corrected upstream. Here we add a test case and update some existing test-cases to reflect this change. We remove grouped signature in test cases that we can (Minimal, BugDeprecated &c) so that the test is as self-contained as possible. - - - - - 708b88b1 by Mateusz Kowalczyk at 2014-04-03T21:16:07+01:00 Enforce strict GHC version in cabal file This stops people with 7.6.3 trying to install 2.15.x which clearly won't work. Unfortunately we shipped 2.14.x without realising this. - - - - - 60334f7c by Mateusz Kowalczyk at 2014-04-03T21:19:24+01:00 Initialise some new PlatformConstants fields - - - - - ea77f668 by Mateusz Kowalczyk at 2014-04-11T16:52:23+01:00 We don't actually want unicode here - - - - - 0b651cae by Mateusz Kowalczyk at 2014-04-11T18:13:30+01:00 Parse identifiers with ^ and ⋆ in them. Fixes haskell/haddock#298. - - - - - e8ad0f5f by Mateusz Kowalczyk at 2014-04-11T18:47:41+01:00 Ignore version string during HTML tests. - - - - - de489089 by Mateusz Kowalczyk at 2014-04-11T18:59:30+01:00 Update CHANGES to follow 2.14.3 - - - - - beb464a9 by Gergő Érdi at 2014-04-13T16:31:10+08:00 remove Origin flag from LHsBindsLR - - - - - cb16f07c by Herbert Valerio Riedel at 2014-04-21T17:16:50+02:00 Replace local `die` by new `System.Exit.die` Starting with GHC 7.10, System.Exit exports the new `die` which is essentially the same as Haddock.Util.die, so this commit changes Haddock.Util.die to be a simple re-export of System.Exit.die. See also https://ghc.haskell.org/trac/ghc/ticket/9016 for more details. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 9b9b23c7 by Mateusz Kowalczyk at 2014-05-03T15:40:11+02:00 Disambiguate ‘die’ in test runners. - - - - - 5d28a2b8 by Mateusz Kowalczyk at 2014-05-05T09:19:49+02:00 Prepare modules for parser split. We have to generalise the Doc (now DocH) slightly to remove the dependency on GHC-supplied type. - - - - - d3967ff3 by Mateusz Kowalczyk at 2014-05-05T11:00:41+02:00 Move parser + parser tests out to own package. We move some types out that are necessary as well and then re-export and specialise them in the core Haddock. Reason for moving out spec tests is that if we're working on the parser, we can simply work on that and we can ignore the rest of Haddock. The downside is that it's a little inconvenient if at the end of the day we want to see that everything passes. - - - - - 522a448d by Mateusz Kowalczyk at 2014-05-05T11:14:47+02:00 Move out Show and Eq instances to Types They are much more useful to the users here. - - - - - 11a6f0f2 by Mateusz Kowalczyk at 2014-05-06T13:50:31+02:00 Remove no longer necessary parser error handling. We can now drop some Maybe tests and even lets us strip an error handling monad away in a few places. - - - - - 6992c924 by Mateusz Kowalczyk at 2014-05-14T02:23:55+02:00 Please the GHC build-system. As I can not figure out how to do this properly, if we're in GHC tree, we treat the library as being the same package. If we're not in the tree, we require that the library be installed separately. - - - - - 7a8ad763 by Mateusz Kowalczyk at 2014-05-14T14:50:25+02:00 Update issue tracker URL - - - - - f616c521 by Mateusz Kowalczyk at 2014-05-14T14:53:32+02:00 Update issue tracker URL for haddock-library - - - - - 66580ded by Gergő Érdi at 2014-05-25T14:24:16+08:00 Accomodate change in PatSyn representation - - - - - 0e43b988 by Mateusz Kowalczyk at 2014-05-29T03:15:29+02:00 Revert "Accomodate change in PatSyn representation" This reverts commit 57aa591362d7c8ba21285fccd6a958629a422091. I am reverting this because I pushed it to master when it was meant to stay on a wip-branch. Sorry Gergo and everyone who had trouble due to this. - - - - - e10d7ec8 by Mateusz Kowalczyk at 2014-05-29T03:24:11+02:00 Revert "Revert "Accomodate change in PatSyn representation"" This reverts commit e110e6e70e40eed06c06676fd2e62578da01d295. Apparently as per GHC commit ac2796e6ddbd54c5762c53e2fcf29f20ea162fd5 this was actually intended. Embarrasing for me. - - - - - 5861aca9 by Mateusz Kowalczyk at 2014-06-05T19:49:27+02:00 Clear up highlighting of identifiers with ‘'’s. - - - - - d7cc420f by Simon Peyton Jones at 2014-06-06T12:41:09+01:00 Follow change in patSynSig - - - - - 938b4fd8 by Mateusz Kowalczyk at 2014-06-12T07:24:29+02:00 Slightly update the readme. Style-sheets are no longer a recent thing, dead links, old maintainers, different formats. - - - - - c7799dea by Mateusz Kowalczyk at 2014-06-18T00:05:56+02:00 Update cabal files Update repository urls, use subdir property for haddock-library and use a separate versioning scheme for haddock-library in preparation for release. - - - - - a2750b6a by Simon Hengel at 2014-06-18T11:01:18+08:00 Compatibility with older versions of base and bytestring - - - - - 009b4b03 by Simon Hengel at 2014-06-18T11:14:01+08:00 Enable travis-ci for haddock-library - - - - - 9b5862eb by Simon Hengel at 2014-06-18T11:14:01+08:00 haddock-library: Do not depend on haddock-library in test suite I think you either add src to hs-source-dirs or the library to build-depends. But doing both does not make sense (AFAICT). - - - - - fb1f3279 by Simon Hengel at 2014-06-18T11:49:05+08:00 haddock-library: Use -Wall for specs - - - - - 649340e1 by Mateusz Kowalczyk at 2014-06-18T06:58:54+02:00 Use Travis with multiple GHC versions When using HEAD, we build haddock-library directly from repository as a dependency (and thanks to --enable-tests, the tests get ran anyway). In all other cases, we manually run the tests on haddock-library only and don't test the main project. - - - - - d7eeeec2 by Mateusz Kowalczyk at 2014-06-18T07:49:04+02:00 Comment improvements + few words in cabal file - - - - - 0f8db914 by Simon Hengel at 2014-06-18T13:52:23+08:00 Use doctest to check examples in documentation - - - - - 2888a8dc by Simon Hengel at 2014-06-18T14:16:48+08:00 Remove doctest dependency (so that we can use haddock-library with doctest) - - - - - 626d5e85 by Mateusz Kowalczyk at 2014-06-18T08:41:25+02:00 Travis tweaks - - - - - 41d4f9cc by Mateusz Kowalczyk at 2014-06-18T08:58:43+02:00 Don't actually forget to install specified GHC. - - - - - c6aa512a by John MacFarlane at 2014-06-18T10:43:57-07:00 Removed reliance on LambdaCase (which breaks build with ghc 7.4). - - - - - b9b93b6f by John MacFarlane at 2014-06-18T10:54:56-07:00 Fixed haddock warnings. - - - - - a41b0ab5 by Mateusz Kowalczyk at 2014-06-19T01:20:10+02:00 Update Travis, bump version - - - - - 864bf62a by Mateusz Kowalczyk at 2014-06-25T10:36:54+02:00 Fix anchors. Closes haskell/haddock#308. - - - - - 53df91bb by Mateusz Kowalczyk at 2014-06-25T15:04:49+02:00 Drop DocParagraph from front of headers I can not remember why they were wrapped in paragraphs to begin with and it seems unnecessary now that I test it. Closes haskell/haddock#307. - - - - - 29b5f2fa by Mateusz Kowalczyk at 2014-06-25T15:17:20+02:00 Don't mangle append order for nested lists. The benefit of this is that the ‘top-level’ element of such lists is properly wrapped in <p> tags so any CSS working with these will be applied properly. It also just makes more sense. Pointed out at jgm/pandoc#1346. - - - - - 05cb6e9c by Mateusz Kowalczyk at 2014-06-25T15:19:45+02:00 Bump haddock-library to 1.1.0 for release - - - - - 70feab15 by Iavor Diatchki at 2014-07-01T03:37:07-07:00 Propagate overloading-mode for instance declarations in haddock (#9242) - - - - - d4ca34a7 by Simon Peyton Jones at 2014-07-14T16:23:15+01:00 Adapt to new definition of HsDecls.TyFamEqn This is a knock-on from the refactoring from Trac haskell/haddock#9063. I'll push the corresponding changes to GHC shortly. - - - - - f91e2276 by Edward Z. Yang at 2014-07-21T08:14:19-07:00 Track GHC PackageId to PackageKey renaming. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - b010f9ef by Edward Z. Yang at 2014-07-25T16:28:46-07:00 Track changes for module reexports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - 8b85f9f9 by Mateusz Kowalczyk at 2014-07-28T13:25:43+02:00 Catch mid-line URLs. Fixes haskell/haddock#314. - - - - - 4c613a78 by Edward Z. Yang at 2014-08-05T03:11:00-07:00 Track type signature change of lookupModuleInAllPackages Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - e80b051c by Edward Z. Yang at 2014-08-05T17:34:26+01:00 If GhcProfiled, also build Haddock profiled. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f9cccd29 by Edward Z. Yang at 2014-08-07T14:23:35+01:00 Ignore TAGS files. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 00b3af52 by Mateusz Kowalczyk at 2014-08-08T04:58:19+02:00 Update to attoparsec-0.12.1.1 There seems to be memory and speed improvement. - - - - - 5457dc71 by Mateusz Kowalczyk at 2014-08-08T18:24:02+02:00 Fix forgotten src - - - - - 3520cb04 by Mateusz Kowalczyk at 2014-08-14T20:19:07+01:00 Bump down the version for master to 2.14.4 - - - - - dc98c21b by Mateusz Kowalczyk at 2014-08-14T20:23:27+01:00 Revert "Track type signature change of lookupModuleInAllPackages" This reverts commit d59fec2c9551b5662a3507c0011e32a09a9c118f. - - - - - 3f2038c0 by Mateusz Kowalczyk at 2014-08-14T20:23:31+01:00 Revert "Track changes for module reexports." This reverts commit b99b57c0df072d12b67816b45eca2a03cb1da96d. - - - - - 56d4e49e by Mateusz Kowalczyk at 2014-08-14T20:23:42+01:00 Revert "Track GHC PackageId to PackageKey renaming." This reverts commit 8ac42d3327473939c013551750425cac191ff0fd. - - - - - 726ea3cb by Mateusz Kowalczyk at 2014-08-14T20:23:47+01:00 Revert "Adapt to new definition of HsDecls.TyFamEqn" This reverts commit cb96b4f1ed0462b4a394b9fda6612c3bea9886bd. - - - - - 61a88ff0 by Mateusz Kowalczyk at 2014-08-14T20:23:52+01:00 Revert "Propagate overloading-mode for instance declarations in haddock (#9242)" This reverts commit 8d20ca8d5a9bee73252ff2035ec45f9c03d0820c. - - - - - a32ba674 by Mateusz Kowalczyk at 2014-08-14T20:26:03+01:00 Revert "Disambiguate ‘die’ in test runners." This reverts commit dba02d6df32534aac5d257f2d28596238d248942. - - - - - f335820f by Mateusz Kowalczyk at 2014-08-14T20:26:09+01:00 Revert "Replace local `die` by new `System.Exit.die`" This reverts commit 08aa509ebac58bfb202ea79c7c41291ec280a1c5. - - - - - 107078e4 by Mateusz Kowalczyk at 2014-08-14T20:27:34+01:00 Merge branch 'reverts' This reverts any changes that were made to have Haddock compile with 7.9. When 7.10 release comes, we can simply re-apply all the patches and any patches that occur on ghc-head branch from now on. This allows us to build master with 7.8.3 - - - - - b44b3871 by Mateusz Kowalczyk at 2014-08-15T02:47:40+01:00 Fix haskell/haddock#313 by doing some list munging. I get rid of the Monoid instance because we weren't satisfying the laws. Convenience of having <> didn't outweigh the shock-factor of having it behave badly. - - - - - e1a62cde by Mateusz Kowalczyk at 2014-08-15T02:52:56+01:00 Stop testing haskell/haddock#188. Because the change is in GHC 7.9 and we now work against 7.8.3, this test no longer makes sense. We revert it until 7.10 becomes the standard version. If anything, there should be a test for this in GHC itself. - - - - - 54e8286d by Mateusz Kowalczyk at 2014-08-15T05:31:57+01:00 Add haskell/haddock#313 to CHANGES - - - - - 9df7ad5d by Simon Hengel at 2014-08-20T11:25:32+08:00 Fix warning - - - - - ee2574d6 by Simon Hengel at 2014-08-20T12:07:01+08:00 Fix travis builds - - - - - 384cf2e6 by Simon Hengel at 2014-08-20T12:14:31+08:00 Require GHC 7.8.3 - - - - - d4779863 by Simon Hengel at 2014-08-22T12:14:16+08:00 Move Haddock API to a separate package - - - - - 80f3e0e1 by Simon Hengel at 2014-08-22T14:57:38+08:00 Bump version to 2.15.0 and add version constraints - - - - - 309a94ce by Simon Hengel at 2014-08-22T15:18:06+08:00 Add deprecated compatibility module - - - - - 4d1e4e3f by Luite Stegeman at 2014-08-22T20:46:45+02:00 export things to allow customizing how the Ghc session is run - - - - - 47884591 by Luite Stegeman at 2014-08-22T20:46:51+02:00 ghc 7.8.2 compatibility - - - - - 5ea94e2c by Luite Stegeman at 2014-08-22T22:08:58+02:00 install dependencies for haddock-api on travis - - - - - 9fb845b2 by Mateusz Kowalczyk at 2014-08-23T10:09:34+01:00 Move sources under haddock-api/src - - - - - 85817dc4 by Mateusz Kowalczyk at 2014-08-23T10:10:48+01:00 Remove compat stuff - - - - - 151c6169 by Niklas Haas at 2014-08-24T08:14:10+02:00 Fix extra whitespace on signatures and update all test cases This was long overdue, now running ./accept.lhs on a clean test from master will not generate a bunch of changes. - - - - - d320e0d2 by Niklas Haas at 2014-08-24T08:14:35+02:00 Omit unnecessary foralls and fix haskell/haddock#315 This also fixes haskell/haddock#86. - - - - - bdafe108 by Mateusz Kowalczyk at 2014-08-24T15:06:46+01:00 Update CHANGES - - - - - fafa6d6e by Mateusz Kowalczyk at 2014-08-24T15:14:23+01:00 Delete few unused/irrelevant/badly-place files. - - - - - 3634923d by Duncan Coutts at 2014-08-27T13:49:31+01:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. - - - - - 539a7e70 by Herbert Valerio Riedel at 2014-08-31T11:36:32+02:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 9e3a0e5b by Mateusz Kowalczyk at 2014-08-31T12:54:43+01:00 Bump version in doc - - - - - 4a177525 by Mateusz Kowalczyk at 2014-08-31T13:01:23+01:00 Bump haddock-library version - - - - - f99c1384 by Mateusz Kowalczyk at 2014-08-31T13:05:25+01:00 Remove references to deleted files - - - - - 5e51a247 by Mateusz Kowalczyk at 2014-08-31T14:18:44+01:00 Make the doc parser not complain - - - - - 2cedb49a by Mateusz Kowalczyk at 2014-09-03T03:33:15+01:00 CONTRIBUTING file for issues - - - - - 88027143 by Mateusz Kowalczyk at 2014-09-04T00:46:59+01:00 Mention --print-missing-docs - - - - - 42f6754f by Alan Zimmerman at 2014-09-05T18:13:24-05:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e712719e by Austin Seipp at 2014-09-09T01:03:27-05:00 Fix import of 'empty' due to AMP. Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 71c29755 by Herbert Valerio Riedel at 2014-09-09T17:35:20+02:00 Bump `base` constraint for AMP - - - - - 0bf9f3ed by Mateusz Kowalczyk at 2014-09-12T19:18:32+01:00 Delete stale ANNOUNCE - - - - - cac89ee6 by Krzysztof Gogolewski at 2014-09-14T17:17:09+02:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 4d683426 by Edward Z. Yang at 2014-09-18T13:38:11-07:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 80697fd5 by Herbert Valerio Riedel at 2014-09-19T00:07:52+02:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. - - - - - c015eb70 by Herbert Valerio Riedel at 2014-09-19T00:10:36+02:00 Revert "Followup changes to addition of -fwarn-context-quantification" This reverts commit 4023817d7c0e46db012ba2eea28022626841ca9b temporarily as the respective feature hasn't landed in GHC HEAD yet, but this commit blocks later commits from being referenced in GHC HEAD. - - - - - 38ded784 by Edward Z. Yang at 2014-09-18T15:32:15-07:00 Revert "Revert "Followup changes to addition of -fwarn-context-quantification"" This reverts commit db14fd8ab4fab43694139bc203808b814eafb2dc. It's in HEAD now. - - - - - f55d59c9 by Herbert Valerio Riedel at 2014-09-26T19:18:28+02:00 Revert "Fix import of 'empty' due to AMP." This reverts commit 0cc5bc85e9fca92ab712b68a2ba2c0dd9d3d79f4 since it turns out we don't need to re-export `empty` from Control.Monad after all. - - - - - 467050f1 by David Feuer at 2014-10-09T20:07:36-04:00 Fix improper lazy IO use Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. - - - - - cc47b699 by Edward Z. Yang at 2014-10-09T21:38:19-07:00 Fix use-after-close lazy IO bug Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. Signed-off-by: David Feuer <David.Feuer at gmail.com> Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 87babcbe by Austin Seipp at 2014-10-20T20:05:27-05:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - ab259516 by Austin Seipp at 2014-10-20T20:07:01-05:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - b918093c by Mateusz Kowalczyk at 2014-10-29T03:59:39+00:00 Experimental support for collapsable headers Closes haskell/haddock#335 - - - - - 849db129 by Mateusz Kowalczyk at 2014-10-29T10:07:26+01:00 Experimental support for collapsable headers (cherry picked from commit e2ed3b9d8dfab09f1b1861dbc8e74f08e137ebcc) - - - - - a4cc4789 by Herbert Valerio Riedel at 2014-10-31T11:08:26+01:00 Collapse user-defined section by default (re haskell/haddock#335) - - - - - 9da1b33e by Yuras Shumovich at 2014-10-31T16:11:04-05:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - c625aefc by Austin Seipp at 2014-10-31T19:34:10-05:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - c7738e5e by Simon Hengel at 2014-11-02T07:25:30+08:00 Remove -fobject-code from .ghci (this slows down reloads on modifications) - - - - - d4a86e95 by Simon Hengel at 2014-11-03T09:26:11+08:00 Get rid of StandaloneDeriving - - - - - a974e311 by Simon Hengel at 2014-11-03T09:26:11+08:00 Derive more instances - - - - - 8aa0c4d7 by Simon Hengel at 2014-11-03T09:27:08+08:00 Remove unused language extensions - - - - - 3052d46a by Simon Hengel at 2014-11-03T09:30:46+08:00 Minor refactoring - - - - - 4281d3cb by Simon Hengel at 2014-11-03T09:30:46+08:00 parser: Try to parse definition lists right before text paragraphs - - - - - 8ba12bf9 by Simon Hengel at 2014-11-03T09:34:19+08:00 Add support for markdown links (closes haskell/haddock#336) - - - - - a2f8d747 by Simon Hengel at 2014-11-03T09:34:19+08:00 Allow markdown links at the beginning of a paragraph - - - - - 53b11207 by Simon Hengel at 2014-11-03T09:34:20+08:00 Update documentation - - - - - 652267c6 by Simon Hengel at 2014-11-03T09:34:20+08:00 Add support for markdown images - - - - - 9d667502 by Simon Hengel at 2014-11-03T09:34:20+08:00 Allow an optional colon after the closing bracket of definition lists This is to disambiguate them from markdown links and will be require with a future release. - - - - - 8167fc32 by Mateusz Kowalczyk at 2014-11-04T01:16:51+00:00 whitespace only - - - - - 3da62981 by Mateusz Kowalczyk at 2014-11-04T01:17:31+00:00 Fix re-exports of built-in type families Fixes haskell/haddock#310 - - - - - edc76b34 by Mateusz Kowalczyk at 2014-11-04T02:54:28+00:00 Turn some uses of error into recoverable warnings This should at the very least not abort when something weird happens. It does feel like we should have a type that carries these errors until the end however as the user might not see them unless they are printed at the end. - - - - - 0a137400 by Mateusz Kowalczyk at 2014-11-04T04:09:44+00:00 Fix warnings - - - - - d068fc21 by Mateusz Kowalczyk at 2014-11-04T21:04:07+00:00 Fix parsing of identifiers written in infix way - - - - - 1a9f2f3d by Simon Hengel at 2014-11-08T11:32:42+08:00 Minor code simplification - - - - - 6475e9b1 by Simon Hengel at 2014-11-08T17:28:33+08:00 newtype-wrap parser monad - - - - - dc1ea105 by Herbert Valerio Riedel at 2014-11-15T11:55:43+01:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - fbb1aca4 by Simon Hengel at 2014-11-16T08:51:38+08:00 State intention rather than implementation details in Haddock comment - - - - - 97851ab2 by Simon Hengel at 2014-11-16T10:20:19+08:00 (wip) Add support for @since (closes haskell/haddock#26) - - - - - 34bcd18e by Gergő Érdi at 2014-11-20T22:35:38+08:00 Update Haddock to new pattern synonym type signature syntax - - - - - 304b7dc3 by Jan Stolarek at 2014-11-20T17:48:43+01:00 Follow changes from haskell/haddock#9812 - - - - - 920f9b03 by Richard Eisenberg at 2014-11-20T16:52:50-05:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 0bfe4e78 by Alan Zimmerman at 2014-11-21T11:23:09-06:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 356ed45a by Thomas Winant at 2014-11-28T16:11:22-06:00 Support for PartialTypeSignatures - - - - - 5dc8f3b1 by Gergő Érdi at 2014-11-29T15:39:09+08:00 For pattern synonyms, render "pattern" as a keyword - - - - - fe704480 by Mateusz Kowalczyk at 2014-12-09T03:38:32+00:00 List new module in cabal file - - - - - b9ad5a29 by Mateusz Kowalczyk at 2014-12-10T00:58:24+00:00 Allow the parser to spit out meta-info Currently we only use it only for ‘since’ annotations but with these patches it should be fairly simple to add new attributes if we wish to. Closes haskell/haddock#26. It seems to work fine but due to 7.10 rush I don't have the chance to do more exhaustive testing right now. The way the meta is output (emphasis at the end of the whole comment) is fairly arbitrary and subject to bikeshedding. Note that this makes test for Bug310 fail due to interface version bump: it can't find the docs for base with this interface version so it fails. There is not much we can do to help this because it tests for ’built-in’ identifier, not something we can provide ourselves. - - - - - 765af0e3 by Mateusz Kowalczyk at 2014-12-10T01:17:19+00:00 Update doctest parts of comments - - - - - 8670272b by jpmoresmau at 2014-12-10T01:35:31+00:00 header could contain several lines Closes haskell/haddock#348 - - - - - 4f9ae4f3 by Mateusz Kowalczyk at 2014-12-12T06:22:31+00:00 Revert "Merge branch 'reverts'" This reverts commit 5c93cc347773c7634321edd5f808d5b55b46301f, reversing changes made to 5b81a9e53894d2ae591ca0c6c96199632d39eb06. Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - e974ac94 by Duncan Coutts at 2014-12-12T06:26:11+00:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. Conflicts: haddock-api/src/Haddock.hs - - - - - 2f3a2365 by Herbert Valerio Riedel at 2014-12-12T06:26:51+00:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 1dbd6390 by Alan Zimmerman at 2014-12-12T06:32:07+00:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - bb6ff1f4 by Mateusz Kowalczyk at 2014-12-12T06:35:07+00:00 Bump ‘base’ constraint Follows the similar commit made on ghc-head branch - - - - - 466fe4ab by Krzysztof Gogolewski at 2014-12-12T06:37:42+00:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 97e080c9 by Edward Z. Yang at 2014-12-12T06:39:35+00:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: haddock-api/src/Haddock/ModuleTree.hs - - - - - 20b2af56 by Herbert Valerio Riedel at 2014-12-12T06:42:50+00:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. Conflicts: haddock-library/src/Documentation/Haddock/Parser.hs - - - - - b3ad269d by Austin Seipp at 2014-12-12T06:44:14+00:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 072df0dd by Austin Seipp at 2014-12-12T06:45:01+00:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - dbb9294a by Herbert Valerio Riedel at 2014-12-12T06:46:17+00:00 Collapse user-defined section by default (re haskell/haddock#335) Conflicts: haddock-api/src/Haddock/Backends/Xhtml/DocMarkup.hs - - - - - f23ab545 by Yuras Shumovich at 2014-12-12T06:46:41+00:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - 753a4b67 by Austin Seipp at 2014-12-12T06:46:51+00:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8954e8f5 by Herbert Valerio Riedel at 2014-12-12T06:50:53+00:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - d2b06d61 by Gergő Érdi at 2014-12-12T07:07:30+00:00 Update Haddock to new pattern synonym type signature syntax Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 1ff02426 by Jan Stolarek at 2014-12-12T07:13:24+00:00 Follow changes from haskell/haddock#9812 Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - 06ad7600 by Richard Eisenberg at 2014-12-12T07:13:43+00:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 8fd2aa8b by Alan Zimmerman at 2014-12-12T07:22:25+00:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Backends/LaTeX.hs haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 95c3db98 by Thomas Winant at 2014-12-12T07:35:49+00:00 Support for PartialTypeSignatures Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs haddock-api/src/Haddock/Interface/Create.hs - - - - - 45494428 by Gergő Érdi at 2014-12-12T07:36:18+00:00 For pattern synonyms, render "pattern" as a keyword - - - - - a237e3eb by Mateusz Kowalczyk at 2014-12-12T12:27:13+00:00 Various fixups and bumps for next release - - - - - 22918bcd by Herbert Valerio Riedel at 2014-12-14T10:11:47+01:00 Remove redundant wild-card pattern match (this would otherwise cause a build-failure with `-Werror`) - - - - - 1d6ce947 by Herbert Valerio Riedel at 2014-12-14T10:17:06+01:00 Treat GHC 7.10 the same as GHC 7.9 ...since the current GHC 7.9 is going to become GHC 7.10 real-soon-now anyway - - - - - f434ea89 by Herbert Valerio Riedel at 2014-12-14T18:26:50+01:00 Fixup ghc.mk (follow-up to 1739375eb23342) This makes the GHC build-system aware of the data-files to be copied into the bindist (as haddock.cabal doesn't list those anymore) - - - - - 6fb839eb by Mateusz Kowalczyk at 2014-12-17T09:28:59+00:00 Only keep one Version instead of blindly appending - - - - - 40645489 by Mateusz Kowalczyk at 2014-12-18T07:09:44+00:00 Fix dependency version - - - - - 8b3b927b by Mateusz Kowalczyk at 2014-12-18T07:14:23+00:00 Print missing docs by default Adds --no-print-missing-docs - - - - - 59666694 by Mateusz Kowalczyk at 2014-12-18T07:21:37+00:00 update changelog - - - - - aa6d168e by Mateusz Kowalczyk at 2014-12-18T07:30:58+00:00 Update docs for @since - - - - - 2d7043ee by Luite Stegeman at 2014-12-19T18:29:35-06:00 hide projectVersion from DynFlags since it clashes with Haddock.Version.projectVersion - - - - - aaa70fc0 by Luite Stegeman at 2014-12-22T15:58:43+01:00 Add missing import for standalone haddock-api package - - - - - 9ce01269 by Herbert Valerio Riedel at 2014-12-22T17:48:45+01:00 Reset ghc-head with master's tree (this is an overwriting git merge of master into ghc-head) - - - - - fcd6fec1 by Herbert Valerio Riedel at 2014-12-22T17:51:52+01:00 Bump versions for ghc-7.11 - - - - - 525ec900 by Mateusz Kowalczyk at 2014-12-23T13:36:24+00:00 travis-ci: test with HEAD - - - - - cbf494b5 by Simon Peyton Jones at 2014-12-23T15:22:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 50e01c99 by Mateusz Kowalczyk at 2014-12-29T15:28:47+00:00 Make travis use 7.10.x - - - - - 475e60b0 by Njagi Mwaniki at 2014-12-29T15:30:44+00:00 Turn the README into GitHub Markdown format. Closes haskell/haddock#354 - - - - - 8cacf48e by Luite Stegeman at 2015-01-05T16:25:37+01:00 bump haddock-api ghc dependency to allow release candidate and first release - - - - - 6ed6cf1f by Simon Peyton Jones at 2015-01-06T16:37:47+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 8b484f33 by Simon Peyton Jones at 2015-01-08T15:50:22+00:00 Track naming change in DataCon - - - - - 23c5c0b5 by Alan Zimmerman at 2015-01-16T10:15:11-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e7a5532c by JP Moresmau at 2015-01-22T17:19:03+00:00 Ignore warnings, install Cabal 1.22 - - - - - 86942c84 by jpmoresmau at 2015-01-22T17:19:04+00:00 solve dataDir ambiguity - - - - - 5ceb743e by jpmoresmau at 2015-01-22T19:17:32+00:00 support GHC 7.10: no Safe-Inferred, Foldable instance - - - - - 6a3b3fb5 by Mateusz Kowalczyk at 2015-01-22T19:32:10+00:00 Update test files Test: a correct behaviour for fields comma-separating values. I'm surprised we had no bug open for this. Maybe it affects how haskell/haddock#301 renders now but I doubt. Operators: Seems GHC is giving us a new order for operators, something must have changed on their side again. cc @haasn , this makes the fixity to the side not match the order on the LHS which is a bit unpleasant. Maybe the fixity can be made to match the GHC order? Bug335: We expand examples by default now. Bug310: Now inferred safe. - - - - - 708f8b2f by jpmoresmau at 2015-01-22T19:36:59+00:00 Links to source location of class instance definitions - - - - - 5cf8a6da by Vincent Berthoux at 2015-01-22T19:59:58+00:00 Filter '\r' from comments due to Windows problems. On Windows this was causing newline to be rendered twice in code blocks. Closes haskell/haddock#359, fixes haskell/haddock#356. - - - - - 1749e6f0 by Mateusz Kowalczyk at 2015-01-22T20:31:27+00:00 Changelog only - - - - - c8145f90 by Mateusz Kowalczyk at 2015-01-22T23:34:05+00:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. - - - - - 14248254 by Mateusz Kowalczyk at 2015-01-22T23:43:18+00:00 Sort out some module import warnings - - - - - d8a38989 by Simon Peyton Jones at 2015-01-23T07:10:16-06:00 Track naming change in DataCon (cherry picked from commit 04cf63d0195837ed52075ed7d2676e71831e8a0b) - - - - - d3ac6ae4 by Alan Zimmerman at 2015-01-23T07:17:19-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit d61bbc75890e4eb0ad508b9c2a27b91f691213e6) - - - - - 4c1ffeb0 by Simon Peyton Jones at 2015-02-10T12:10:33+00:00 Track changes in HsSyn for quasi-quotes - - - - - 775d20f7 by Mateusz Kowalczyk at 2015-03-15T08:11:48+01:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. (cherry picked from commit 8e06728afb0784128ab2df0be7a5d7a191d30ff4) - - - - - f9245e72 by Phil Ruffwind at 2015-03-16T04:32:01-04:00 Prevent Synopsis from using up too much horizontal space When long type signatures occur in the Synopsis, the element is stretched beyond the width of the window. Scrollbars don't appear, so it's impossible to read anything when this happens. - - - - - cd8fa415 by Mateusz Kowalczyk at 2015-03-17T21:59:39+00:00 Update changelog Closes haskell/haddock#151 due to 71170fc77962f10d7d001e3b8bc8b92bfeda99bc - - - - - b5248b47 by Ben Gamari at 2015-03-25T17:12:17+00:00 Make the error encountered when a package can't be found more user-friendly Closes haskell/haddock#369 - - - - - b756b772 by Mateusz Kowalczyk at 2015-03-26T16:31:40+00:00 Remove now redundant imports - - - - - 5ea5e8dd by Mateusz Kowalczyk at 2015-03-26T16:45:52+00:00 Update test to account for \r filtering - - - - - 6539bfb3 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Test for anchor defaulting I delete the old tests because it turns out that: * test runner would never put them in scope of each other even with imports so just one would suffice * test runner actually needed some hacking to keep links so in the end we would end up with no anchors making them useless - - - - - 1a01d950 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Clearly default to variables in out of scope case - - - - - 7943abe8 by Mateusz Kowalczyk at 2015-03-27T01:14:11+00:00 Fix Hoogle display of constructors Fixes haskell/haddock#361 - - - - - 6d6e587e by Mateusz Kowalczyk at 2015-03-27T01:45:18+00:00 Fully qualify names in Hoogle instances output Closes haskell/haddock#263 - - - - - 52dac365 by Mateusz Kowalczyk at 2015-03-27T01:55:01+00:00 Update changelog - - - - - ca5af9a8 by Mateusz Kowalczyk at 2015-03-27T02:43:55+00:00 Output method documentation in Hoogle backend One thing of note is that we no longer preserve grouping of methods and print each method on its own line. We could preserve it if no documentation is present for any methods in the group if someone asks for it though. Fixes haskell/haddock#259 - - - - - a33f0c10 by Mateusz Kowalczyk at 2015-03-27T03:04:21+00:00 Don't print instance safety information in Hoogle Fixes haskell/haddock#168 - - - - - df6c935a by Mateusz Kowalczyk at 2015-03-28T00:11:47+00:00 Post-release version bumps and changelog - - - - - dde8f7c0 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Loosen bounds on haddock-* - - - - - de93bf89 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Expand response files in arguments Closes haskell/haddock#285 - - - - - 1f0b0856 by Zejun Wu at 2015-04-26T16:35:35-07:00 Do not insert anchor for section headings in contents box - - - - - 860439d7 by Simon Peyton Jones at 2015-05-01T09:36:47+01:00 Track change in API of TyCon - - - - - a32f3e5f by Adam Gundry at 2015-05-04T15:32:59+01:00 Track API changes to support empty closed type familes - - - - - 77e98bee by Ben Gamari at 2015-05-06T20:17:08+01:00 Ignore doc/haddock.{ps,pdf} - - - - - 663d0204 by Murray Campbell at 2015-05-11T04:47:37-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8bb0dcf5 by Murray Campbell at 2015-05-11T06:35:06-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit 2380f07c430c525b205ce2eae6dab23c8388d899) - - - - - bad900ea by Adam Bergmark at 2015-05-11T15:29:39+01:00 haddock-library: require GHC >= 7.4 `Data.Monoid.<>` was added in base-4.5/GHC-7.4 Closes haskell/haddock#394 Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - daceff85 by Simon Peyton Jones at 2015-05-13T12:04:21+01:00 Track the new location of setRdrNameSpace - - - - - 1937d1c4 by Alan Zimmerman at 2015-05-25T21:27:15+02:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - ee0fb6c2 by Łukasz Hanuszczak at 2015-05-27T11:51:31+02:00 Create simple method for indentation parsing. - - - - - 7d6fcad5 by Łukasz Hanuszczak at 2015-05-27T21:36:13+02:00 Make nested lists count indentation according to first item. - - - - - d6819398 by Łukasz Hanuszczak at 2015-05-27T22:46:13+02:00 Add simple test case for arbitrary-depth list nesting. - - - - - 2929c54d by Łukasz Hanuszczak at 2015-06-03T02:11:31+02:00 Add arbitrary-indent spec test for parser. - - - - - 9a0a9bb0 by Mateusz Kowalczyk at 2015-06-03T05:25:29+01:00 Update docs with info on new list nesting rule Fixes haskell/haddock#278 through commits from PR haskell/haddock#401 - - - - - 12efc92c by Mateusz Kowalczyk at 2015-06-03T05:29:26+01:00 Update some meta data at the top of the docs - - - - - 765ee49f by Bartosz Nitka at 2015-06-07T08:40:59-07:00 Add some Hacking docs for getting started - - - - - 19aaf851 by Bartosz Nitka at 2015-06-07T08:44:30-07:00 Fix markdown - - - - - 2a90cb70 by Mateusz Kowalczyk at 2015-06-08T15:08:36+01:00 Refine hacking instructions slightly - - - - - 0894da6e by Thomas Winant at 2015-06-08T23:47:28-05:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 10a9bb76 by Emanuel Borsboom at 2015-06-12T02:46:23+01:00 Build executable with '-threaded' (fixes haskell/haddock#399) - - - - - 7696b94f by Mateusz Kowalczyk at 2015-06-12T02:59:19+01:00 Update changelog for -threaded Closes haskell/haddock#400 - - - - - d3c118ec by Bartosz Nitka at 2015-06-12T03:00:58+01:00 Fix haddock: internal error: spliceURL UnhelpfulSpan (#207) Inferred type signatures don't have SrcSpans, so let's use the one from the declaration. I've tested this manually on the test-case from haskell/haddock#207, but I got stuck at trying to run the test-suite. - - - - - b67e843b by Mateusz Kowalczyk at 2015-06-12T03:01:50+01:00 Changelog for haskell/haddock#207 Fixes haskell/haddock#207, closes haskell/haddock#402 - - - - - 841d785e by jpmoresmau at 2015-06-12T16:03:16+01:00 Attach to instance location the name that has the same location file Fixes haskell/haddock#383 - - - - - 98791cae by Mateusz Kowalczyk at 2015-06-12T16:08:27+01:00 Update changelog Closes haskell/haddock#398 - - - - - 7c0b5a87 by Phil Ruffwind at 2015-06-12T13:07:25-04:00 Fix alignment of Source links in instance table in Firefox Due to a Firefox bug [1], a combination of 'whitespace: nowrap' on the parent element with 'float: right' on the inner element can cause the floated element to be displaced downwards for no apparent reason. To work around this, the left side is wrapped in its own <span> and set to 'float: left'. As a precautionary measure to prevent the parent element from collapsing entirely, we also add the classic "clearfix" hack. The latter is not strictly needed but it helps prevent bugs if the layout is altered again in the future. Fixes haskell/haddock#384. Remark: line 159 of src/Haddock/Backends/Xhtml/Layout.hs was indented to prevent confusion over the operator precedence of (<+>) vs (<<). [1]: https://bugzilla.mozilla.org/show_bug.cgi?id=488725 - - - - - cfe86e73 by Mateusz Kowalczyk at 2015-06-14T10:49:01+01:00 Update tests for the CSS changes - - - - - 2d4983c1 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding for Haskell source parser module. - - - - - 29548785 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement function for tagging parsed chunks with source spans. - - - - - 6a5e4074 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement simple string chunking based on HsColour library. - - - - - 6e52291f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create basic token classification method. - - - - - da971a27 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Adapt source span tagging to work with current whitespace handling. - - - - - 4feb5a22 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add record accessors to exports of hyperlinker parser module. - - - - - a8cc4e39 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Make parser module export all types and associated accessors. - - - - - fb8d468f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple HTML renderer for parsed source file. - - - - - 80747822 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for specifying the CSS file path in HTML source renderer. - - - - - 994dc1f5 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix identifier recognition in Haskell source parser. - - - - - b1bd0430 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix comment recognition in Haskell source parser. - - - - - 11db85ae by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for recognizing compiler pragmas in source parser. - - - - - 736c7bd3 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding of module for associating tokens with AST names. - - - - - 7e149bc2 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement utility method for extracting variable identifiers from AST. - - - - - 32eb640a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple mechanism for associating tokens with AST names. - - - - - d4eba5bc by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add dummy support for hyperlinking named tokens. - - - - - 2b76141f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matcher bug causing wrong items being hyperlinked. - - - - - 2d48002e by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Constrain elements exported by hyperlinker modules. - - - - - 9715eec6 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for type token recognition. - - - - - 8fa401cb by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for binding token recognition. - - - - - d062400b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement go-to-definition mechanism for local bindings. - - - - - f4dc229b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement module export- and import-list item hyperlinking. - - - - - c9a46d58 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matching to allow parenthesized operators hyperlinking. - - - - - 03aad95a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix weird hyperlinking of parenthesized operators. - - - - - b4694a7d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for type declaration anchors. - - - - - 7358d2d2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for top-level function declaration anchors. - - - - - dfc24b24 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix external anchors to contain HTML file extension. - - - - - a045926c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Refactor the way AST names are handled within detailed tokens. - - - - - c76049b4 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement hyperlinking of imported module names. - - - - - 2d2a1572 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix parsing of single line comments with broken up newlines. - - - - - 11afdcf2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix bug with improper newline handling. - - - - - 8137f104 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issues with escaped newlines in comments. - - - - - 34759b19 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for parsing C preprocessor macros. - - - - - 09f0f847 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for parser module of source hyperlinker. - - - - - 709a8389 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for AST module of source hyperlinker. - - - - - 4df5c227 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add command line option for generating hyperlinked source. - - - - - 7a755ea2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Extend module interface with rich source token stream field. - - - - - 494f4ab1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement source tokenization during interface creation process. - - - - - 5f21c953 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create hyperlinker module and plug it into the Haddock pipeline. - - - - - 0cc8a216 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for providing custom CSS files for hyperlinked source. - - - - - a32bbdc1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for fancy highlighting upon hovering over identifier. - - - - - d16d642a by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make source hyperlinker generate output in apropriate directory. - - - - - ae12953d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create module with hyperlinker utility functions. - - - - - 6d4952c5 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make external hyperlinks point to locations specified by source URLs. - - - - - 8417555d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Rewrite source generation to fixed links and directory structure. - - - - - ce9cec01 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add basic support for cross-package hyperlink generation. - - - - - 7eaf025c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Disable generating hyperlinks for module references. - - - - - a50bf92e by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make Haddock generate source for all interfaces (also hidden ones). - - - - - f5ae2838 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Prevent source parser from throwing exception when lexing fails. - - - - - db9ffbe0 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement workaround for Chrome highlighting issues. - - - - - 0b6b453b by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate correct anchors for data constructors. - - - - - c86d38bc by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate anchors for record field declarations. - - - - - 063abf7f by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issue with hyperlink highlight styling in Chrome browser. - - - - - 880fc611 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking constructor names in patters. - - - - - c9e89b95 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record patterns. - - - - - 17a11996 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record expressions. - - - - - 0eef932d by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Make hyperlinker respect pretty-printer flag and add documentation. - - - - - f87c1776 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Unexpose hyperlinker modules in Cabal configuration. - - - - - 4c9e2b06 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Setup HSpec framework for Haddock API package. - - - - - 4b20cb30 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add basic tests related to comment parsing. - - - - - 6842e919 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add tests related to parsing basic language constructs. - - - - - 87bffb35 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add simple tests for do-notation parsing. - - - - - e7af1841 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add very simple QuickCheck properties for source parser spec. - - - - - c84efcf1 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Create simple test runner for hyperlinker tests. - - - - - 76b90447 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for basic identifier hyperlinking. - - - - - 0fbf4df6 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for operator hyperlinking. - - - - - 731aa039 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for constructor hyperlinking. - - - - - 995a78a2 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for record expressions and patterns hyperlinking. - - - - - 3566875a by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for literal syntax highlighting. - - - - - 68469a35 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add hyperlinker test runner to .cabal and .gitignore files. - - - - - aa946c93 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Adapt hyperlinker test runner to have the same interface as HTML one. - - - - - ce34da16 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Fix hyperlinker test runner file paths and add pretty-printing option. - - - - - 0d7dd65e by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add reference files for hyperlinker test cases. - - - - - efa4a1e0 by Łukasz Hanuszczak at 2015-07-01T00:47:32+02:00 Make hyperlinker test runner strip local links from generated source. - - - - - 3e96e584 by Łukasz Hanuszczak at 2015-07-01T01:14:59+02:00 Create simple script for accepting hyperlinker test case references. - - - - - 526fe610 by Łukasz Hanuszczak at 2015-07-01T01:16:41+02:00 Re-accept hyperlinker test cases with local references stripped out. - - - - - 892e2cb3 by Łukasz Hanuszczak at 2015-07-01T01:22:09+02:00 Fix bug with diffing wrong files in hyperlinker test runner. - - - - - 9ff46039 by Łukasz Hanuszczak at 2015-07-01T18:04:46+02:00 Remove unused dependencies in Haddock API spec configuration. - - - - - 47969c07 by Łukasz Hanuszczak at 2015-07-01T18:32:19+02:00 Add support for hyperlinking synonyms in patterns. - - - - - a73449e0 by Łukasz Hanuszczak at 2015-07-01T18:33:44+02:00 Create test case for hyperlinking @-patterns. - - - - - c2077ed8 by Łukasz Hanuszczak at 2015-07-01T19:06:04+02:00 Add support for hyperlinking universally quantified type variables. - - - - - 68017342 by Łukasz Hanuszczak at 2015-07-01T19:28:32+02:00 Create hyperlinker test case with quantified type variables. - - - - - 51c01a78 by Łukasz Hanuszczak at 2015-07-01T19:34:22+02:00 Add scoped type variables test for polymorphism test case. - - - - - 13181ae2 by Łukasz Hanuszczak at 2015-07-01T19:56:27+02:00 Add record wildcards test for records hyperlinking test case. - - - - - 991b81dd by Łukasz Hanuszczak at 2015-07-01T21:01:42+02:00 Document some functions in XHTML utlity module. - - - - - 98c8dfe5 by Łukasz Hanuszczak at 2015-07-01T22:25:21+02:00 Make hyperlinker render qualified names as one entity. - - - - - 75e13b9b by Łukasz Hanuszczak at 2015-07-01T22:27:38+02:00 Add qualified name test for identifiers hyperlinking test case. - - - - - de1e143f by Łukasz Hanuszczak at 2015-07-02T12:32:59+02:00 Fix crash happening when hyperlinking type family declarations. - - - - - 7a8fb175 by Łukasz Hanuszczak at 2015-07-02T12:47:03+02:00 Add support for anchoring data family constructor declarations. - - - - - 3b404e49 by Łukasz Hanuszczak at 2015-07-02T13:31:05+02:00 Improve support for hyperlinking type families. - - - - - 59eb7143 by Łukasz Hanuszczak at 2015-07-02T13:33:34+02:00 Add hyperlinker test case for checking type and type family declarations. - - - - - d1cda0c0 by Łukasz Hanuszczak at 2015-07-02T13:41:38+02:00 Fix issue with operators being recognized as preprocessor directives. - - - - - da206c9d by Łukasz Hanuszczak at 2015-07-02T17:18:12+02:00 Fix broken tests for parsing and hyperlinking hash operators. - - - - - 53750d1b by Łukasz Hanuszczak at 2015-07-02T18:53:28+02:00 Add support for anchoring signatures in type class declarations. - - - - - 1fa5bb10 by Łukasz Hanuszczak at 2015-07-02T19:04:47+02:00 Make hyperlinker generate anchors only to top-level value bindings. - - - - - a542305c by Łukasz Hanuszczak at 2015-07-02T19:05:58+02:00 Create hyperlinker test case for type classes. - - - - - b0dd4581 by Łukasz Hanuszczak at 2015-07-04T16:28:26+02:00 Update docs with information about source hyperlinking. - - - - - 9795302a by Łukasz Hanuszczak at 2015-07-04T16:52:15+02:00 Update docs on using `--read-interface` option. - - - - - 9acdc002 by Łukasz Hanuszczak at 2015-07-04T17:15:26+02:00 Remove potentially dangerous record access in hyperlinker AST module. - - - - - fb3ab7be by Łukasz Hanuszczak at 2015-07-04T17:40:10+02:00 Make Haddock generate warnings about potential misuse of hyperlinker. - - - - - a324c504 by Łukasz Hanuszczak at 2015-07-04T17:43:22+02:00 Fix incorrect specification of source style option in doc file. - - - - - 3f01a8e4 by Łukasz Hanuszczak at 2015-07-05T17:06:36+02:00 Refactor source path mapping to use modules as indices. - - - - - ac70f5b1 by Łukasz Hanuszczak at 2015-07-05T17:47:34+02:00 Fix bug where not all module interfaces were added to source mapping. - - - - - f5e57da9 by Łukasz Hanuszczak at 2015-07-06T16:39:57+02:00 Extract main hyperlinker types to separate module. - - - - - 43974905 by Łukasz Hanuszczak at 2015-07-06T16:52:13+02:00 Move source paths types to hyperlinker types module. - - - - - 3e236055 by Łukasz Hanuszczak at 2015-07-06T17:06:19+02:00 Add support for hyperlinking modules in import lists. - - - - - 58233d9f by Łukasz Hanuszczak at 2015-07-06T17:26:49+02:00 Add short documentation for hyperlinker source map type. - - - - - 14da016d by Łukasz Hanuszczak at 2015-07-06T18:07:20+02:00 Fix bug with module name being hyperlinked to `Prelude`. - - - - - 8f79db52 by Łukasz Hanuszczak at 2015-07-06T18:23:47+02:00 Fix problem with spec build in Haddock API configuration. - - - - - e7cc056c by Adam Sandberg Eriksson at 2015-07-07T23:22:21+01:00 StrictData: print correct strictness marks - - - - - e8253ca8 by Mateusz Kowalczyk at 2015-07-07T23:58:28+01:00 Update changelog - - - - - 0aba676b by Mateusz Kowalczyk at 2015-07-07T23:58:33+01:00 Relax upper bound on GHC a bit - - - - - 7a595381 by Mateusz Kowalczyk at 2015-07-07T23:58:52+01:00 Delete trailing whitespace - - - - - 50976d5e by Adam Sandberg Eriksson at 2015-07-08T15:03:04+02:00 StrictData: changes in HsBang type - - - - - 83b045fa by Mateusz Kowalczyk at 2015-07-11T14:35:18+01:00 Fix expansion icon for user-collapsible sections Closes haskell/haddock#412 - - - - - b2a3b0d1 by Mateusz Kowalczyk at 2015-07-22T22:03:21+01:00 Make some version changes after 2.16.1 release - - - - - a8294423 by Ben Gamari at 2015-07-27T13:16:07+02:00 Merge pull request haskell/haddock#422 from adamse/adamse-D1033 Merge for GHC D1033 - - - - - c0173f17 by randen at 2015-07-30T14:49:08-07:00 Break the response file by line termination rather than spaces, since spaces may be within the parameters. This simple approach avoids having the need for any quoting and/or escaping (although a newline char will not be possible in a parameter and has no escape mechanism to allow it). - - - - - 47c0ca14 by Alan Zimmerman at 2015-07-31T10:41:52+02:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - 45a9d770 by Mateusz Kowalczyk at 2015-07-31T09:47:43+01:00 Update changelog - - - - - 347a20a3 by Phil Ruffwind at 2015-08-02T23:15:26+01:00 Avoid JavaScript error during page load in non-frame mode In non-frame mode, parent.window.synopsis refers to the synopsis div rather than the nonexistent frame. Unfortunately, the script wrongly assumes that if it exists it must be a frame, leading to an error where it tries to access the nonexistent attribute 'replace' of an undefined value (synopsis.location). Closes haskell/haddock#406 - - - - - 54ebd519 by Phil Ruffwind at 2015-08-02T23:27:10+01:00 Link to the definitions to themselves Currently, the definitions already have an anchor tag that allows URLs with fragment identifiers to locate them, but it is rather inconvenient to obtain such a URL (so-called "permalink") as it would require finding the a link to the corresponding item in the Synopsis or elsewhere. This commit adds hyperlinks to the definitions themselves, allowing users to obtain links to them easily. To preserve the original aesthetics of the definitions, we alter the color of the link so as to be identical to what it was, except it now has a hover effect indicating that it is clickable. Additionally, the anchor now uses the 'id' attribute instead of the (obsolete) 'name' attribute. Closes haskell/haddock#407 - - - - - 02cc8bb7 by Phil Ruffwind at 2015-08-02T23:28:02+01:00 Fix typo in Haddock.Backends.Xhtml.Layout: divSynposis -> divSynopsis Closes haskell/haddock#408 - - - - - 2eb0a458 by Phil Ruffwind at 2015-08-02T23:30:07+01:00 Fix record field alignment when name is too long Change <dl> to <ul> and use display:table rather than floats to layout the record fields. This avoids bug haskell/haddock#301 that occurs whenever the field name gets too long. Slight aesthetic change: the entire cell of the field's source code is now shaded gray rather than just the area where text exists. Fixes haskell/haddock#301. Closes haskell/haddock#421 - - - - - 7abb3402 by Łukasz Hanuszczak at 2015-08-02T23:32:14+01:00 Add some utility definitions for generating line anchors. - - - - - e0b1d79b by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Make hyperlinked source renderer generate line anchors. - - - - - 24dd4c9f by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Re-accept test cases after adding line anchors for each of them. - - - - - 0372cfcb by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Override source line flags when source hyperlinker is enabled. - - - - - a81bcd07 by Mateusz Kowalczyk at 2015-08-02T23:58:25+01:00 Update tests to follow HTML changes - - - - - d2d7426f by Łukasz Hanuszczak at 2015-08-06T20:54:59+02:00 Fix quote syntax for promoted types. - - - - - 668cf029 by Łukasz Hanuszczak at 2015-08-06T21:12:00+02:00 Apply promoted type quoting to type-level consing. - - - - - 89f8e7c6 by Łukasz Hanuszczak at 2015-08-06T21:17:10+02:00 Extend advanced types test case with other examples. - - - - - 86494bca by Łukasz Hanuszczak at 2015-08-06T21:22:06+02:00 Rename advanced types test case and accept new output. - - - - - dbb7c7c0 by Adam Sandberg Eriksson at 2015-08-09T23:01:05+02:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 2a7704fa by Ben Gamari at 2015-08-10T13:18:05+02:00 Merge pull request haskell/haddock#433 from adamse/split-hsbang HsBang is split into HsSrcBang and HsImplBang - - - - - 891954bc by Thomas Miedema at 2015-08-15T14:51:18+02:00 Follow changes in GHC build system - - - - - b55d32ab by Mateusz Kowalczyk at 2015-08-21T18:06:09+01:00 Make Travis use 7.10.2 - - - - - 97348b51 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Move SYB utilities to standalone module. - - - - - 748ec081 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement `everywhere` transformation in SYB module. - - - - - 011cc543 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement generic transformation constructor. - - - - - b9510db2 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Create simple utility module for type specialization. - - - - - 43229fa6 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Make type of type specialization function more general. - - - - - fd844e90 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Add basic HTML test case for checking instance specialization. - - - - - 6ea0ad04 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Make HTML class instance printer take optional signature argument. - - - - - 65aa41b6 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Refactor instance head type to record instead of a meaningless tuple. - - - - - 3fc3bede by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add expandable method section for each class instance declaration. - - - - - 99ceb107 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Move dummy post-family instances for `DocName` to `Types` module. - - - - - e98f4708 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create convenience functions for type specialization module. - - - - - b947552f by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Hook type specialization logic with HTML pretty-printer. - - - - - dcaa8030 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create stub functions for sugaring specialized types. - - - - - fa84bc65 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement list syntax sugaring logic for specialized types. - - - - - e8b05b07 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement tuple syntax sugaring logic for specialized types. - - - - - 68a2e5bc by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Get rid of code duplication in type specialization module. - - - - - 4721c336 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create scaffolding of a framework for renaming specialized types. - - - - - 271b488d by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in missing cases in specialized type renaming function. - - - - - bfa5f2a4 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Remove code duplication in specialized type renamer. - - - - - ea6bd0e8 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Change state of the type renaming monad. - - - - - 77c5496e by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement simple mechanism for generating new type names. - - - - - 91bfb48b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in stub behaviour with actual environment renaming. - - - - - d244517b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix logic behind binder type renaming. - - - - - f3c5e360 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add SYB-like utility function for performing stateful queries. - - - - - eb3f9154 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create function for retrieving free variables from given type. - - - - - a94561d3 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix compilation error caused by incorrect type signature. - - - - - 8bb707cf by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Move `SetName` class definition to types module. - - - - - 5800b13b by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Hook type renamer with instance method HTML pretty-printer. - - - - - 6a480164 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some test cases for type renamer. - - - - - 839842f7 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized signatures refer to original signature declaration. - - - - - 4880f7c9 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized methods be nicely formatted again. - - - - - ab5a6a2e by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Attach source locations to the specialized class methods. - - - - - 43f8a559 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Extend instances test case to also test multi-name type signatures. - - - - - 59bc751c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix tab-based indentation in instances test case. - - - - - c2126815 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Improve placement of instance methods expander button. - - - - - 0a32e287 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add new data type declaration to instance specialization test case. - - - - - 5281af1f by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make type renamer first try single-letter names as alternatives. - - - - - 7d509475 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix type renamer bug with incorrect names being generated. - - - - - 0f35bf7c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some documentation and refactor type specialization module. - - - - - da1d0803 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix another bug where type renamer was generating incorrect names. - - - - - cd39b5cb by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Refactor type renamer to rebinding and pure renaming phases. - - - - - 850251f4 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix unwitting compilation bug. - - - - - e5e9fc01 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Integrate instance specification type into class instance definition. - - - - - 825b0ea0 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer neccessary instance specification type. - - - - - cdba44eb by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix declaration converter to use more appropriate mode for methods. - - - - - bc45c309 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug with types not being specialized at all. - - - - - 5d8e5d89 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug where instance expander was opening wrong section. - - - - - 6001ee41 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix another type renamer bug where not all names were rebound. - - - - - 5f58ce2a by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix yet another renamer bug where some names were not unique. - - - - - 8265e521 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split instance subsection layout method to top-level declarations. - - - - - e5e66298 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Rearrange layout of instance methods in generated documentation. - - - - - a50b4eea by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer used layout method. - - - - - 2ff36ec2 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach section title to the instance methods block. - - - - - 7ac15300 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Add basic tests for associated types in instances test case. - - - - - db0ea2f9 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach associated types information to instance header. - - - - - 71cad4d5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details section contain associated types information. - - - - - deee2809 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Improve look of rendered associated families in instance details. - - - - - 839d13a5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Introduce alternative type for family declarations. - - - - - d397f03f by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details record use new type for family declarations. - - - - - 2b23fe97 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split printer of type family header to separate functions. - - - - - c3498cdc by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Implement HTML renderer for pseudo-family declarations. - - - - - c12bbb04 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Apply type specializer to associated type family declarations. - - - - - 2fd69ff2 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Create helper method for specializing type signatures. - - - - - 475826e7 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Refactor specializer module to be independent from XHTML backend. - - - - - f00b431c by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add some documentation for instance head specializer. - - - - - a9fef2dc by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix bug with missing space in documentation for associated types. - - - - - 50e29056 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with incorrect instance details sections being expanded. - - - - - e6dfdd03 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by adding instance details section. - - - - - 75565b2a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make section identifier of instance details more GHC-independent. - - - - - add0c23e by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Re-accept tests after applying deterministic section identifiers. - - - - - 878f2534 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make identifier generation also architecture-independent. - - - - - 48be69f8 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with instance expander hijacking type hyperlink click. - - - - - 47830c1f by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Get rid of dreadful hashing function for generating identifiers. - - - - - 956cd5af by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Move `InstOrigin` type declaration to more appropriate module. - - - - - bf672ed3 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by changes related to instance expander. - - - - - 8f2a949a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add examples with type operators to the instances test case. - - - - - 64600a84 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add basic support for sugaring infix type operators. - - - - - 747d71b8 by Łukasz Hanuszczak at 2015-08-21T18:22:34+01:00 Add support for sugaring built-in function syntax. - - - - - d4696ffb by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Remove default methods from Hoogle class output. - - - - - bf0e09d7 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add fixity declarations in Hoogle backend output. - - - - - 90e91a51 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix bug with incorrect fixities being generated in Hoogle backend. - - - - - 48f11d35 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve class type family declarations output in Hoogle backend. - - - - - 661e8e8f by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add missing default family equations in Hoogle output. - - - - - e2d64103 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve formatting of class details output in Hoogle backend. - - - - - 490fc377 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix weird-looking Hoogle output for familyless classes. - - - - - ea115b64 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create script file for new HTML test runner. - - - - - 609913d3 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Set default behaviour if no arguments given. - - - - - dc115f67 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for providing optional arguments for test runner. - - - - - d93ec867 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve output of test runner error messages. - - - - - 0be9fe12 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for executing Haddock process in test runner. - - - - - 4e4d00d9 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add GHC path to test runner configuration. - - - - - d67a2086 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make GHC path a test runner command-line argument. - - - - - c810079a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Extend test runner configuration with Haddock arguments. - - - - - fee18845 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor test runner and create stub functions. - - - - - ff7c161f by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make test runner actually run Haddock executable. - - - - - 391f73e6 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with test runner not producing any output files. - - - - - 81a74e2d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Setup skeleton of framework for running tests. - - - - - f8a79ec4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with modules not being found in global search mode. - - - - - 7e700b4d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make Haddock standard output redirection be more configurable. - - - - - 53b4c17a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Incorporate old, ugly functions for comparing output files. - - - - - 8277c8aa by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor architecture of test runner output checking functions. - - - - - 587bb414 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement actual diffing mechanism. - - - - - 9ed2b5e4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve code style to match popular guidelines. - - - - - 14bffaf8 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make it possible to choose alternative diff tool. - - - - - 5cdfb005 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create stub methods for processing test output as XML documents. - - - - - 7ef8e12e by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement link-stripping logic as simple SYB transformation. - - - - - 8a1fcd4f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Incorporate link stripping to output diffing mechanism. - - - - - 37dba2bc by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement footer-stripping logic. - - - - - 9cd52120 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Add missing dependencies in Cabal configuration file. - - - - - e0f83c6e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix issue with output being printed in incorrect order. - - - - - 0a94fbb0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to run tests without generating diff. - - - - - 76a58c6f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor HTML test suite boilerplate to external package. - - - - - af41e6b0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utilities for storing directory configuration. - - - - - d8f0698f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move IO-dependent config of HTML test suite to test package. - - - - - 17369fa0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Enable all compiler warnings in Haddock test package configuration. - - - - - 9d03b47a by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move Haddock runner of HTML test suite to Haddock test package. - - - - - 4b3483c5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 03754194 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience wrappers to simplify in test entry points. - - - - - 27476ab7 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adjust module visibility and items they export. - - - - - c40002ba by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Remove no longer useful test option. - - - - - 55ab2541 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Change extension of test files used for diffing. - - - - - 136bf4e4 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor and simplify XHTML helper module of test package. - - - - - 69f7e3df by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix typo in link stripper of HTML test suite runner. - - - - - 0c3c1c6b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience script for running specific HTML tests. - - - - - 489e1b05 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement utility functions for conditional link stripping. - - - - - 0f985dc3 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt `hypsrc-test` module to work with new testing framework. - - - - - 927406f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement output accepting mechanism in test package. - - - - - 8545715e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utility function for recursive obtaining directory contents. - - - - - cb70381f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 019599b5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix path handling in test runner. - - - - - 399b985b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to specify ignored files for test output. - - - - - 41b3d93d by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt HTML test runner to use new ignoring functionality. - - - - - e2091c8b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix bug with not all test output files being checked. - - - - - b22134f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Specify ignored files for hyperlinker source test runner. - - - - - 3301dfa1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Copy test runner script for hyperlinked source case. - - - - - d39a6dfa by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner invoking Haddock in incorrect mode. - - - - - f32c8ff3 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix path handling in test module loader. - - - - - 10f94ee9 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make test runner ignore test packages with no modules. - - - - - 5dc4239c by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create test runner entry points for LaTeX test suite. - - - - - 58d1f7cf by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with unnecessary checking old test output. - - - - - c7ce76e1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Re-implement test acceptance functionality. - - - - - 13bbabe8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix warning about no longer needed definition. - - - - - 958a99b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Adapt Cabal configuration to execute LaTeX suite with new runner. - - - - - 550ff663 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Setup test suite for Hoogle backend. - - - - - 3aa969c4 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make Hoogle backend create output directory if needed. - - - - - eb085b02 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Add appropriate .gitignore entry and configure Hoogle test suite. - - - - - a50bf915 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner failing when run on multiple test packages. - - - - - bf5368b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create simple test cases for Hoogle backend. - - - - - 6121ba4b by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create helper function for conversion between XML and XHTML. - - - - - cb516061 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Refactor existing code to use XHTML printer instead of XML one. - - - - - e2de8c82 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Improve portability of test runner scripts. - - - - - 9563e774 by Łukasz Hanuszczak at 2015-08-22T23:43:16+02:00 Remove redundant import statement. - - - - - 55353df1 by Łukasz Hanuszczak at 2015-08-24T23:09:20+02:00 Fix bug with accepting to non-existing directory. - - - - - 00a334ca by Łukasz Hanuszczak at 2015-08-24T23:09:47+02:00 Accept output for Hoogle and LaTeX backends. - - - - - 29191d8b by Łukasz Hanuszczak at 2015-08-24T23:14:18+02:00 Get rid of obsolete testing utilities. - - - - - bbb25db3 by Łukasz Hanuszczak at 2015-08-24T23:18:50+02:00 Update sandbox setup guide to work with Haddock test package. - - - - - cfd45248 by Łukasz Hanuszczak at 2015-08-24T23:51:30+02:00 Make Travis aware of Haddock test package. - - - - - 74185b7a by Łukasz Hanuszczak at 2015-08-25T17:41:59+02:00 Fix test suite failure when used with Stack. - - - - - 18769697 by Łukasz Hanuszczak at 2015-08-25T18:02:09+02:00 Add sample Stack setup to the hacking guide. - - - - - 22715eeb by Łukasz Hanuszczak at 2015-08-25T18:04:47+02:00 Fix Markdown formatting of README file. - - - - - b49ec386 by Łukasz Hanuszczak at 2015-08-25T18:13:36+02:00 Setup Haddock executable path in Travis configuration. - - - - - 5d29eb03 by Eric Seidel at 2015-08-30T09:55:58-07:00 account for changes to ipClass - - - - - f111740a by Ben Gamari at 2015-09-02T13:20:37+02:00 Merge pull request haskell/haddock#443 from bgamari/ghc-head account for changes to ipClass - - - - - a2654bf6 by Jan Stolarek at 2015-09-03T01:32:57+02:00 Follow changes from haskell/haddock#6018 - - - - - 2678bafe by Richard Eisenberg at 2015-09-21T12:00:47-04:00 React to refactoring CoAxiom branch lists. - - - - - ebc56e24 by Edward Z. Yang at 2015-09-21T11:53:46-07:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4a8c4198 by Tamar Christina at 2015-09-27T13:59:08+02:00 Create Process: removed PhaseFailed - - - - - 7e99b790 by Oleg Grenrus at 2015-09-27T20:52:10+03:00 Generate docs for orphan instances - - - - - 32e932e2 by Oleg Grenrus at 2015-09-28T07:21:11+03:00 Have source links for orphan instances - - - - - c2eb9f4f by Oleg Grenrus at 2015-09-28T07:24:58+03:00 Print orphan instances header only if required - - - - - ff96f978 by Oleg Grenrus at 2015-09-28T07:40:54+03:00 Add orphan instances link to contents box - - - - - d72490a6 by Oleg Grenrus at 2015-09-28T16:37:44+03:00 Fix orphan instance collapsing - - - - - 25d3dfe5 by Ben Gamari at 2015-10-03T12:38:09+02:00 Merge pull request haskell/haddock#448 from Mistuke/fix-silent-death-of-runInteractive Remove PhaseFailed - - - - - 1e45e43b by Edward Z. Yang at 2015-10-11T13:10:10-07:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b1370ac1 by Adam Gundry at 2015-10-16T16:26:42+01:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 60bef421 by Simon Peyton Jones at 2015-10-26T12:52:36+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 4c1898ca by Simon Peyton Jones at 2015-10-27T14:24:56+00:00 Track change to PatSyn.patSynSig - - - - - 25108e85 by Simon Peyton Jones at 2015-10-27T17:34:18+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - 693643ac by Ben Gamari at 2015-10-28T14:33:06+01:00 Account for Typeable changes The treatment of type families changed. - - - - - cd7c2221 by Simon Peyton Jones at 2015-10-30T13:03:51+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - 712032cb by Herbert Valerio Riedel at 2015-10-31T11:01:45+01:00 Relax upper bound on `base` to allow base-4.9 - - - - - 0bfa0475 by Simon Peyton Jones at 2015-10-31T19:08:13+00:00 More adaption to wildcard-refactor - - - - - 0a3c0cb7 by Simon Peyton Jones at 2015-10-31T22:14:43+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - c4fd4ec9 by Alan Zimmerman at 2015-11-01T11:16:34+01:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 42cdd882 by Matthew Pickering at 2015-11-06T20:02:16+00:00 Change for IEThingWith - - - - - f368b7be by Ben Gamari at 2015-11-11T11:35:51+01:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - e32965b8 by Simon Peyton Jones at 2015-11-13T12:18:17+00:00 Merge with origin/head - - - - - ebcf795a by Edward Z. Yang at 2015-11-13T21:56:27-08:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4e23989f by Simon Peyton Jones at 2015-11-18T11:32:54+00:00 Wibbles to Haddock - - - - - 2289cd4a by Simon Peyton Jones at 2015-11-20T23:12:49+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor - - - - - 695975a6 by Alan Zimmerman at 2015-11-21T21:16:12+02:00 Update to match GHC wip/T11019 - - - - - bbba21e7 by Simon Peyton Jones at 2015-11-23T13:54:31+00:00 merge with origin/ghc-head - - - - - 3d664258 by Simon Peyton Jones at 2015-11-23T17:17:18+00:00 Wibble - - - - - e64cf586 by Herbert Valerio Riedel at 2015-12-05T00:29:55+01:00 Canonicalise Monad instances - - - - - a2de15a7 by Alan Zimmerman at 2015-12-05T17:33:52+02:00 Matching changes for haskell/haddock#11028 - - - - - cc29a3e4 by Alan Zimmerman at 2015-12-05T19:45:33+02:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 95dd15d1 by Richard Eisenberg at 2015-12-11T17:33:39-06:00 Update for type=kinds - - - - - cb5fd9ed by Herbert Valerio Riedel at 2015-12-14T15:07:30+00:00 Bump versions for ghc-7.11 - - - - - 4f286d96 by Simon Peyton Jones at 2015-12-14T15:10:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 13ea2733 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 098df8b8 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track changes in HsSyn for quasi-quotes - - - - - 716a64de by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track change in API of TyCon - - - - - 77a66bca by Adam Gundry at 2015-12-14T15:10:58+00:00 Track API changes to support empty closed type familes - - - - - f2808305 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track the new location of setRdrNameSpace - - - - - ba8b08a4 by Alan Zimmerman at 2015-12-14T15:10:59+00:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - a4ded87e by Thomas Winant at 2015-12-14T15:14:05+00:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 25c78107 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: print correct strictness marks - - - - - 6cbc41c4 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: changes in HsBang type - - - - - ad46821a by Alan Zimmerman at 2015-12-14T15:14:06+00:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - abc0ae5b by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 3308d06c by Thomas Miedema at 2015-12-14T15:14:07+00:00 Follow changes in GHC build system - - - - - 6c763deb by Eric Seidel at 2015-12-14T15:14:07+00:00 account for changes to ipClass - - - - - ae5b4eac by Jan Stolarek at 2015-12-14T15:17:00+00:00 Follow changes from haskell/haddock#6018 - - - - - ffbc40e0 by Richard Eisenberg at 2015-12-14T15:17:02+00:00 React to refactoring CoAxiom branch lists. - - - - - d1f531e9 by Edward Z. Yang at 2015-12-14T15:17:02+00:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 79f73754 by Tamar Christina at 2015-12-14T15:17:02+00:00 Create Process: removed PhaseFailed - - - - - 3d37bebb by Edward Z. Yang at 2015-12-14T15:20:46+00:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 5f8a9e44 by Adam Gundry at 2015-12-14T15:20:48+00:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 79dda70f by Simon Peyton Jones at 2015-12-14T15:26:02+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 959930fb by Simon Peyton Jones at 2015-12-14T15:37:50+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - e18a8df5 by Simon Peyton Jones at 2015-12-14T15:37:52+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - aa35ab52 by Simon Peyton Jones at 2015-12-14T15:40:18+00:00 More adaption to wildcard-refactor - - - - - 8ceef94b by Simon Peyton Jones at 2015-12-14T15:46:04+00:00 Track change to PatSyn.patSynSig - - - - - cd81e83d by Ben Gamari at 2015-12-14T15:46:06+00:00 Account for Typeable changes The treatment of type families changed. - - - - - 63c9117c by Herbert Valerio Riedel at 2015-12-14T15:46:34+00:00 Relax upper bound on `base` to allow base-4.9 - - - - - a484c613 by Alan Zimmerman at 2015-12-14T15:47:46+00:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 2c26fa51 by Matthew Pickering at 2015-12-14T15:47:47+00:00 Change for IEThingWith - - - - - 593baa0f by Ben Gamari at 2015-12-14T15:49:21+00:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - b6b5ca78 by Edward Z. Yang at 2015-12-14T15:49:54+00:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b5b0e072 by Alan Zimmerman at 2015-12-14T15:54:20+00:00 Update to match GHC wip/T11019 - - - - - 14ddeb68 by Simon Peyton Jones at 2015-12-14T15:54:22+00:00 Wibble - - - - - 10a90ad8 by Herbert Valerio Riedel at 2015-12-14T15:54:22+00:00 Canonicalise Monad instances - - - - - ed68ac50 by Alan Zimmerman at 2015-12-14T15:55:48+00:00 Matching changes for haskell/haddock#11028 - - - - - 3f7e5a2d by Alan Zimmerman at 2015-12-14T15:55:49+00:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 6543a73f by Richard Eisenberg at 2015-12-14T15:59:55+00:00 Update for type=kinds - - - - - 193a5c48 by Matthew Pickering at 2015-12-14T18:17:00+00:00 Changes to compile with 8.0 - - - - - add669ec by Matthew Pickering at 2015-12-14T18:47:12+00:00 Warnings - - - - - 223f3fb4 by Ben Gamari at 2015-12-15T23:45:05+01:00 Update for D1200 - - - - - d058388f by Ben Gamari at 2015-12-16T05:40:17-05:00 Types: Add Outputable[Bndr] DocName instances - - - - - 62ecd7fb by Ben Gamari at 2015-12-16T09:23:09-05:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ddbc187a by Ben Gamari at 2015-12-16T17:54:55+01:00 Update for D1200 - - - - - cec83b52 by Ben Gamari at 2015-12-16T17:54:55+01:00 Types: Add Outputable[Bndr] DocName instances - - - - - d12ecc98 by Ben Gamari at 2015-12-16T17:54:55+01:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ada1616f by Ben Gamari at 2015-12-16T17:54:58+01:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a4f0383d by Ben Gamari at 2015-12-16T23:32:38+01:00 Fix Hyperlinker GHC.con_names is now GHC.getConNames - - - - - a10e6849 by Ben Gamari at 2015-12-20T00:54:11+01:00 Merge remote-tracking branch 'mrhania/testing-framework-improvements' into ghc-head - - - - - f078b4fd by Ben Gamari at 2015-12-20T00:59:51+01:00 test: Compatibility with Cabal 1.23 - - - - - 88a511a9 by Ben Gamari at 2015-12-20T01:14:35+01:00 Merge remote-tracking branch 'phadej/orphans' into ghc-head - - - - - 4e250f36 by Ben Gamari at 2015-12-20T01:14:52+01:00 Add html-test for orphan instances output - - - - - 87fffbad by Alan Zimmerman at 2015-12-20T09:50:42+02:00 Update for GHC trac#11258 Adding locations to RdrName in FieldOcc and AmbiguousFieldOcc - - - - - 6b7e51c9 by idontgetoutmuch at 2015-12-20T21:01:47+00:00 Merge pull request haskell/haddock#1 from haskell/ghc-head Ghc head - - - - - 229c1fb5 by Dominic Steinitz at 2015-12-21T07:19:16+00:00 Handle inline math with mathjax. - - - - - 57902d66 by Dominic Steinitz at 2015-12-21T08:07:11+00:00 Fix the documentation for haddock itself. Change notation and add support for inline math. Allow newlines in display math. Add a command line option for the mathjax url (you might want to use a locally installed version). Rebase tests because of extra url and version change. Respond to (some of the) comments. Fix warnings in InterfaceFile.hs - - - - - 0e69f236 by Herbert Valerio Riedel at 2015-12-21T18:30:43+01:00 Fix-up left-over assumptions of GHC 7.12 into GHC 8.0 - - - - - c67f8444 by Simon Peyton Jones at 2015-12-22T16:26:56+00:00 Follow removal of NamedWildCard from HsType - - - - - da40327a by Ben Gamari at 2015-12-23T14:15:28+01:00 html-test/Operators: Clear up ambiguous types For reasons that aren't entirely clear a class with ambiguous types was accepted by GHC <8.0. I've added a functional dependency to clear up this ambiguity. - - - - - 541b7fa4 by Ben Gamari at 2015-12-23T14:18:51+01:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 0febc947 by Ben Gamari at 2015-12-24T00:30:20+01:00 hoogle-test/AssocTypes: Allow AmbiguousTypes GHC 8.0 complains otherwise - - - - - 25810841 by Ben Gamari at 2015-12-24T00:33:18+01:00 OrphanInstances: Accept test output - - - - - 841987f3 by Ben Gamari at 2015-12-25T11:03:11+01:00 Merge remote-tracking branch 'idontgetoutmuch/ghc-head' into ghc-head - - - - - 358391f0 by Ben Gamari at 2015-12-26T10:44:50+01:00 Add missing import - - - - - a8896885 by Ben Gamari at 2015-12-26T10:45:27+01:00 travis: Use Travis containers - - - - - 85e82134 by Herbert Valerio Riedel at 2015-12-30T17:25:39+01:00 tweak version bounds for GHC-8.1 - - - - - 672a5f75 by randen at 2016-01-01T23:45:25-08:00 The Haddock part for fully gcc-like response files " driver/Main.hs * Moved the response file handling into ResponseFile.hs, updating import section as appropriate. * driver/ResponseFile.hs * New file. In anticipation that maybe some day this could be provided by another library, and to make it possible to unit test, this functionality is pulled out of the Main.hs module, and expanded to support the style/format of response files which gcc uses. * The specification for the format of response files which gcc generates and consumes, seems to be best derived from the gcc code itself (libiberty/argv.c), so that is what has been done here. * This is intended to fix haskell/haddock#379 * driver-test/Main.hs * New file for testing code in the driver source tree * driver-test/ResponseFileSpec.hs * Tests, adapted/adopted from the same gcc code where the escaping/unescaping is from, in the hspec style of unit tests * haddock.cabal * Add the driver-test test-suite. Introduces a new library dependency (upon hspec) for the haddock driver target in the haddock.cabal file, but practically, this should not be a problem as the haddock-api tests already depend on hspec. - - - - - 498781df by Ben Gamari at 2016-01-06T13:41:04+01:00 Version bumps and changelog - - - - - 8451e46a by Ben Gamari at 2016-01-06T13:47:17+01:00 Merge remote-tracking branch 'randen/bug468' - - - - - fb2d9181 by Ben Gamari at 2016-01-06T08:14:42-05:00 Add ResponseFile to OtherModules - - - - - 2cb2d2e3 by Ben Gamari at 2016-01-06T14:35:00+01:00 Merge branch 'master' into ghc-head - - - - - 913477d4 by Eric Seidel at 2016-01-11T14:57:57-08:00 deal with un-wiring of IP class - - - - - c557a4b3 by Alan Zimmerman at 2016-01-15T11:14:35+02:00 Update to match wip/T11430 in GHC - - - - - 3e135093 by Alan Zimmerman at 2016-01-16T18:21:59+01:00 Update to match wip/T11430 in GHC - - - - - c48ef2f9 by Ben Gamari at 2016-01-18T09:50:06+01:00 Merge remote-tracking branch 'gridaphobe/ghc-head' into ghc-head - - - - - 9138a1b0 by Eric Seidel at 2016-01-18T12:50:15+01:00 deal with un-wiring of IP class (cherry picked from commit 17388b0f0029d969d79353be7737eb01c7b8dc5f) - - - - - b48c172e by Joachim Breitner at 2016-01-19T00:11:38+01:00 Make sure --mathjax affects all written HTML files This fixes haskell/haddock#475. - - - - - af61fe63 by Ryan Scott at 2016-02-07T23:25:57+01:00 Render */# instead of TYPE 'Lifted/TYPE 'Unlifted (fixes haskell/haddock#473) - - - - - b6458693 by Ben Gamari at 2016-02-07T23:29:27+01:00 Merge pull request haskell/haddock#477 from haskell/issue-475 Make sure --mathjax affects all written HTML files - - - - - adcc0071 by Ben Gamari at 2016-02-07T23:34:52+01:00 Merge branch 'master' into ghc-head - - - - - d0404e61 by Ben Gamari at 2016-02-08T12:46:49+01:00 doc: Switch to Sphinx - - - - - acb153b3 by Ben Gamari at 2016-02-08T12:46:56+01:00 Document --use-unicode flag - - - - - c20bdf1d by Ben Gamari at 2016-02-08T13:41:24+01:00 Fix GHC and haddock-library dependency bounds - - - - - 8d946801 by Ben Gamari at 2016-02-08T14:54:56+01:00 testsuite: Rework handling of output sanitization Previously un-cleaned artifacts were kept as reference output, making it difficult to tell what has changed and causing spurious changes in the version control history. Here we rework this, cleaning the output during acceptance. To accomplish this it was necessary to move to strict I/O to ensure the reference handle was closed before accept attempts to open the reference file. - - - - - c465705d by Ben Gamari at 2016-02-08T15:36:05+01:00 test: Compare on dump For reasons I don't understand the Xml representations differ despite their textual representations being identical. - - - - - 1ec0227a by Ben Gamari at 2016-02-08T15:36:05+01:00 html-test: Accept test output - - - - - eefbd63a by Ben Gamari at 2016-02-08T15:36:08+01:00 hypsrc-test: Accept test output And fix impredicative Polymorphism testcase. - - - - - d1df4372 by Ben Gamari at 2016-02-08T15:40:44+01:00 Merge branch 'fix-up-testsuite' - - - - - 206a3859 by Phil Ruffwind at 2016-02-08T17:51:21+01:00 Move the permalinks to "#" on the right side Since pull request haskell/haddock#407, the identifiers have been permalinked to themselves, but this makes it difficult to copy the identifier by double-clicking. To work around this usability problem, the permalinks are now placed on the far right adjacent to "Source", indicated by "#". Also, 'namedAnchor' now uses 'id' instead of 'name' (which is obsolete). - - - - - 6c89fa03 by Phil Ruffwind at 2016-02-08T17:54:44+01:00 Update tests for previous commit - - - - - effaa832 by Ben Gamari at 2016-02-08T17:56:17+01:00 Merge branch 'anchors-redux' - - - - - 9a2bec90 by Ben Gamari at 2016-02-08T17:58:40+01:00 Use -fprint-unicode-syntax when --use-unicode is enabled This allows GHC to render `*` as its Unicode representation, among other things. - - - - - 28ecac5b by Ben Gamari at 2016-02-11T18:53:03+01:00 Merge pull request haskell/haddock#480 from bgamari/sphinx Move documentation to ReStructuredText - - - - - 222e5920 by Ryan Scott at 2016-02-11T15:42:42-05:00 Collapse type/data family instances by default - - - - - a80ac03b by Ryan Scott at 2016-02-11T20:17:09-05:00 Ensure expanded family instances render correctly - - - - - 7f985231 by Ben Gamari at 2016-02-12T10:04:22+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - d4eda086 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Various cleanups - - - - - 79bee48d by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show kind signatures for type family variables Addresses GHC haskell/haddock#11588. - - - - - b2981d98 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show 'where ...' after closed type family Seems like we should ideally show the actual equations as well but that seems like it would be a fair amount of work - - - - - cfc0e621 by Ben Gamari at 2016-02-18T22:48:12+01:00 Merge pull request haskell/haddock#483 from bgamari/T11588 Fix GHC haskell/haddock#11588 This fixes GHC haskell/haddock#11588: * Show where ... after closed type families * Show kind signatures on type family type variables - - - - - 256e8a0d by Ben Gamari at 2016-02-18T23:15:39+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 32402036 by Richard Eisenberg at 2016-02-24T13:21:44-05:00 Follow-on changes to support RuntimeRep - - - - - 2b1c572d by Matthew Pickering at 2016-03-04T21:04:02+00:00 Remove unused functions - - - - - eb906f50 by Richard Eisenberg at 2016-03-13T21:17:20+01:00 Follow-on changes to support RuntimeRep (cherry picked from commit ab954263a793d8ced734459d6194a5d89214b66c) - - - - - 8c34ef34 by Richard Eisenberg at 2016-03-14T23:47:23-04:00 Changes due to fix for GHC#11648. - - - - - 0e022014 by Richard Eisenberg at 2016-03-15T14:06:45+01:00 Changes due to fix for GHC#11648. (cherry picked from commit bb994de1ab0c76d1aaf1e39c54158db2526d31f1) - - - - - ed3f78ab by Rik Steenkamp at 2016-04-02T22:20:36+01:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 - - - - - d3210042 by Rik Steenkamp at 2016-04-04T15:43:32+02:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 (cherry picked from commit 3ddcbd6b8e6884bd95028381176eb33bee6896fb) - - - - - 236eec90 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 692ee7e0 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. - - - - - 79619f57 by Ben Gamari at 2016-04-10T23:46:22+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. (cherry picked from commit c6d6a18d85e5e2d9bb5904e6919e8a8d7e31c4c5) - - - - - 3358ccb4 by Ben Gamari at 2016-04-10T23:47:27+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 264949b1 by Ben Gamari at 2016-04-16T17:50:23+02:00 Merge pull request haskell/haddock#482 from RyanGlScott/ghc-head Collapse type/data family instances by default - - - - - 478c483a by Ben Gamari at 2016-04-16T17:51:09+02:00 Merge pull request haskell/haddock#489 from mpickering/unused-functions Remove some unused functions - - - - - c94e55f0 by Ryan Scott at 2016-04-16T17:57:54+02:00 Collapse type/data family instances by default (cherry picked from commit 2da130a8db8f995c119b544fad807533236cf088) - - - - - 31e633d3 by Ryan Scott at 2016-04-16T17:58:06+02:00 Ensure expanded family instances render correctly (cherry picked from commit 1338b5d7c32939de6bbc31af0049477e4f847103) - - - - - 03e4d197 by Matthew Pickering at 2016-04-16T17:58:21+02:00 Remove unused functions (cherry picked from commit b89d1c2456bdb2d4208d94ded56155f7088a37d0) - - - - - ed4116f6 by Ben Gamari at 2016-04-20T10:46:57+02:00 ghc: Install files for needed --hyperlinked-source - - - - - 0be999c4 by Ben Gamari at 2016-04-20T11:37:54+02:00 ghc: Install files for needed --hyperlinked-source (cherry picked from commit 5c82c9fc2d21ddaae4a2470f1c375426968f19c6) - - - - - 4d17544c by Simon Peyton Jones at 2016-04-20T12:42:28+01:00 Track change to HsGroup This relates to a big GHC patch for Trac haskell/haddock#11348 - - - - - 1700a50d by Ben Gamari at 2016-05-01T13:19:27+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. - - - - - 0b7c8125 by Ben Gamari at 2016-05-01T13:21:43+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. (cherry picked from commit 609018dd09c4ffe27f9248b2d8b50f6196cd42b9) - - - - - af115ce0 by Ryan Scott at 2016-05-04T22:15:50-04:00 Render Haddocks for derived instances Currently, one can document top-level instance declarations, but derived instances (both those in `deriving` clauses and standalone `deriving` instances) do not enjoy the same privilege. This makes the necessary changes to the Haddock API to enable rendering Haddock comments for derived instances. This is part of a fix for Trac haskell/haddock#11768. - - - - - 76fa1edc by Ben Gamari at 2016-05-10T18:13:25+02:00 haddock-test: A bit of refactoring for debuggability - - - - - 7d4c4b20 by Ben Gamari at 2016-05-10T18:13:25+02:00 Create: Mark a comment as TODO - - - - - 2a6d0c90 by Ben Gamari at 2016-05-10T18:13:25+02:00 html-test: Update reference output - - - - - bd60913d by Ben Gamari at 2016-05-10T18:13:25+02:00 hypsrc-test: Fix reference file path in cabal file It appears the haddock insists on prefixing --hyperlinked-sourcer output with directory which the source appeared in. - - - - - c1548057 by Ben Gamari at 2016-05-10T18:22:12+02:00 doc: Update extra-source-files in Cabal file - - - - - 41d5bae3 by Ben Gamari at 2016-05-10T18:29:21+02:00 Bump versions - - - - - ca75b779 by Ben Gamari at 2016-05-11T16:03:44+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 4e3cfd62 by Ben Gamari at 2016-05-11T16:06:45+02:00 Merge remote-tracking branch 'RyanGlScott/ghc-head' into ghc-head - - - - - a2379970 by Ben Gamari at 2016-05-11T23:15:11+02:00 doc: Add clean targets - - - - - f275212e by Ben Gamari at 2016-05-11T23:15:14+02:00 doc: Add html as an all-target for ghc Otherwise the html documentation won't be installed for binary-dist. - - - - - 388fc0af by Ben Gamari at 2016-05-12T09:49:12+02:00 Update CHANGES - - - - - bad81ad5 by Ben Gamari at 2016-05-12T09:49:38+02:00 Version bump - - - - - c01688a7 by Ben Gamari at 2016-05-12T10:04:58+02:00 Revert "Version bump" This bump was a bit premature. This reverts commit 7b238d9c5be9b07aa2d10df323b5c7b8d1634dc8. - - - - - 7ed05724 by Ben Gamari at 2016-05-12T10:05:33+02:00 doc: Fix GHC clean rule Apparently GHC's build system doesn't permit wildcards in clean paths. - - - - - 5d9611f4 by Ben Gamari at 2016-05-12T17:43:50+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 653566b2 by Ben Gamari at 2016-05-14T09:57:31+02:00 Version bump to 2.17.2 - - - - - b355c439 by Ben Gamari at 2016-05-14T09:57:51+02:00 doc: Use `$(MAKE)` instead of `make` This is necessary to ensure we use gmake. - - - - - 8a18537d by Ben Gamari at 2016-05-14T10:15:45+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - b3290ef1 by Sebastian Meric de Bellefon at 2016-05-14T11:29:47-04:00 Fix haskell/haddock#303. Hide footer when printing The "Produced by Haddock" footer was overlapping the page's body when printing. This patch hides the footer with a css media rule. - - - - - b4a76f89 by Sebastian Meric de Bellefon at 2016-05-15T02:12:46-04:00 Fix haskell/haddock#280. Parsing of module header The initial newlines were counted as indentation spaces, thus disturbing the parsing of next lines - - - - - ba797c9e by Ben Gamari at 2016-05-16T14:53:46+02:00 doc: Vendorize alabaster Sphinx theme Alabaster is now the default sphinx theme and is a significant improvement over the previous default that it's worthproviding it when unavailable (e.g. Sphinx <1.3). - - - - - c9283e44 by Ben Gamari at 2016-05-16T14:55:17+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 1c9ea198 by Sebastian Méric de Bellefon at 2016-05-16T12:30:40-04:00 Merge pull request haskell/haddock#502 from Helkafen/master Fix haskell/haddock#303. Hide footer when printing - - - - - 33631016 by Ben Gamari at 2016-05-16T19:56:11+02:00 Revert "doc: Vendorize alabaster Sphinx theme" This ended up causes far too many issues to be worthwhile. We'll just have to live with inconsistent haddock documentation. This reverts commit cec21957001143794e71bcd9420283df18e7de40. - - - - - 93317d26 by Ben Gamari at 2016-05-16T19:56:11+02:00 cabal: Fix README path - - - - - c8695b22 by Ben Gamari at 2016-05-16T19:58:51+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 0b50eaaa by Ben Gamari at 2016-05-16T21:02:08+02:00 doc: Use whichever theme sphinx deems appropriate - - - - - 857c1c9c by Ben Gamari at 2016-05-16T21:07:08+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 15fc5637 by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Remove redundant imports - - - - - 132ddc6a by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Better debug output For tracking down haskell/haddock#505 - - - - - 2252a149 by Ben Gamari at 2016-05-22T12:43:59+02:00 Don't consider default class ops when looking for decls When we are looking for an operation within a class we don't care about `default`-type declarations. This was the cause of haskell/haddock#505. - - - - - 4886b2ec by Oleg Grenrus at 2016-05-24T16:19:48+03:00 UnfelpfulSpan line number omitted Kind of resolves https://github.com/haskell/haddock/issues/508 - - - - - a4befd36 by Oleg Grenrus at 2016-05-24T16:53:35+03:00 Change Hyperlinked lexer to know about DataKinds ticks - - - - - f45cb52e by David Feuer at 2016-05-24T18:48:53-04:00 Make parser state a newtype Previously, it was `data` wrapping a `Maybe`, which seems a bit silly. Obviously, this can be changed back if anyone wants to add more fields some day. - - - - - 05013dd7 by Sebastian Meric de Bellefon at 2016-05-24T22:03:55-04:00 remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) Frames are a bit broken, ignored by Hackage, and considered obsolete in general. This patch disables frames generation. The mini_*.html files are still used in the synopsis. - - - - - b8163a88 by Ben Gamari at 2016-05-25T14:44:15+02:00 Merge pull request haskell/haddock#507 from bgamari/T505 Fix haskell/haddock#505 - - - - - ea1b30c6 by Sebastian Meric de Bellefon at 2016-05-25T14:17:00-04:00 Update CHANGES - - - - - eddfc258 by Sebastian Méric de Bellefon at 2016-05-25T15:17:40-04:00 Merge pull request haskell/haddock#514 from Helkafen/frames remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) - - - - - 0e506818 by Alex Biehl at 2016-05-26T12:43:09+02:00 Remove misplaced haddock comment - - - - - a07d28c0 by Ben Gamari at 2016-05-27T11:34:59+02:00 Merge pull request haskell/haddock#515 from alexbiehl/master Remove misplaced haddock comment - - - - - 9001d267 by Ben Gamari at 2016-05-27T11:35:46+02:00 Merge pull request haskell/haddock#513 from treeowl/newtype-since Make parser state a newtype - - - - - 74e1a018 by Sebastian Méric de Bellefon at 2016-05-28T17:28:15-04:00 Merge pull request haskell/haddock#504 from Helkafen/issue-280 Fix haskell/haddock#280. Parsing of module header - - - - - 37557f4f by Alan Zimmerman at 2016-05-29T23:36:50+02:00 Matching changes for haskell/haddock#12105 - - - - - 7d09e5d6 by Sebastian Meric de Bellefon at 2016-06-03T18:07:48-04:00 Version bumps (2.17.3, 1.4.2) - - - - - 85b4bc15 by Sebastian Méric de Bellefon at 2016-06-06T18:35:13-04:00 Merge pull request haskell/haddock#521 from Helkafen/master Version bumps (2.17.3, 1.4.2) - - - - - e95f0dee by Sebastian Meric de Bellefon at 2016-06-06T19:11:35-04:00 publish haddock-test library - - - - - 4de40586 by Sebastian Méric de Bellefon at 2016-06-06T20:26:30-04:00 Merge pull request haskell/haddock#512 from phadej/oleg-fixes Fixes for haskell/haddock#508 and haskell/haddock#510 - - - - - ddfd0789 by Dominic Steinitz at 2016-06-09T09:27:28+01:00 Documentation for LaTeX markup. - - - - - 697a503a by Dominic Steinitz at 2016-06-09T09:33:59+01:00 Fix spelling mistake. - - - - - 246f6fff by Dominic Steinitz at 2016-06-09T09:37:15+01:00 Camel case MathJax. - - - - - 4684bd23 by Dominic Steinitz at 2016-06-09T09:44:53+01:00 Fix math typo and add link. - - - - - f20c037c by Simon Peyton Jones at 2016-06-13T18:26:03+01:00 Follow changes to LHsSigWcType - - - - - 0c58996d by Simon Peyton Jones at 2016-06-15T12:56:01+01:00 Follow GHC re-adding FunTy - - - - - 401b5ca7 by Sebastian Méric de Bellefon at 2016-06-15T12:16:47-04:00 Merge pull request haskell/haddock#525 from idontgetoutmuch/master Documentation for LaTeX markup. - - - - - 92d263b7 by Sebastian Méric de Bellefon at 2016-06-15T12:17:29-04:00 Merge pull request haskell/haddock#522 from Helkafen/master publish haddock-test library - - - - - 0953a2ca by Sebastian Meric de Bellefon at 2016-06-16T00:46:46-04:00 Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 65453e14 by Ben Gamari at 2016-06-16T11:16:32+02:00 ocean: Ensure that synopsis fully covers other content Previously MathJax content was being rendered on top of the synopsis due to ambiguous z-ordering. Here we explicitly give the synopsis block a higher z-index to ensure it is rendered on top. Fixes haskell/haddock#531. - - - - - 68e411a1 by Sebastian Méric de Bellefon at 2016-06-16T23:34:39-04:00 Merge pull request haskell/haddock#534 from bgamari/T531 ocean: Ensure that synopsis fully covers other content - - - - - fad6491b by Sebastian Méric de Bellefon at 2016-06-18T23:57:20-04:00 Merge pull request haskell/haddock#533 from Helkafen/master Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 6108e21b by Sebastian Meric de Bellefon at 2016-06-22T23:08:28-04:00 do not create empty src directory Fix haskell/haddock#536. - - - - - 1ef23823 by Sebastian Méric de Bellefon at 2016-06-24T00:04:48-04:00 Merge pull request haskell/haddock#537 from Helkafen/master do not create empty src directory - - - - - 966baa96 by Omari Norman at 2016-06-29T21:59:34-04:00 Add $ as a special character If this character is not escaped, documentation built with Haddock 2.17.2 will fail. This was not an issue with 2.16 series, which causes builds to fail and there is nothing in the docs or error message giving a clue about why builds that used to succeed now don't. - - - - - 324adb60 by Ben Gamari at 2016-07-01T12:18:51+02:00 GhcUtils: Changes for multi-pattern signatures - - - - - d7571675 by Ömer Sinan Ağacan at 2016-07-21T13:30:47+02:00 Add support for unboxed sums - - - - - 29d0907b by Simon Marlow at 2016-07-22T13:55:48+01:00 Disable NFData instances for GHC types when GHC >= 8.2 - - - - - 702d95f3 by Simon Marlow at 2016-08-02T15:57:30+02:00 Disable NFData instances for GHC types when GHC >= 8.0.2 (cherry picked from commit a3309e797c42dae9bccdeb17ce52fcababbaff8a) - - - - - f4fa79c3 by Ben Gamari at 2016-08-07T13:51:18+02:00 ghc.mk: Don't attempt to install html/frames.html The frames business has been removed. - - - - - 9cd63daf by Ben Gamari at 2016-08-07T13:51:40+02:00 Haddock.Types: More precise version guard This allows haddock to be built with GHC 8.0.2 pre-releases. - - - - - f3d7e03f by Mateusz Kowalczyk at 2016-08-29T20:47:45+01:00 Merge pull request haskell/haddock#538 from massysett/master Add $ as a special character - - - - - 16dbf7fd by Bartosz Nitka at 2016-09-20T19:44:04+01:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 - - - - - 7c31c1ff by Bartosz Nitka at 2016-09-27T17:32:22-04:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 (cherry picked from commit 073d899a8f94ddec698f617a38d3420160a7fd0b) - - - - - 33a90dce by Ryan Scott at 2016-09-30T20:53:41-04:00 Haddock changes for T10598 See https://ghc.haskell.org/trac/ghc/ticket/10598 - - - - - 1f32f7cb by Ben Gamari at 2016-10-13T20:01:26-04:00 Update for refactoring of NameCache - - - - - 1678ff2e by Ben Gamari at 2016-11-15T17:42:48-05:00 Bump upper bound on base - - - - - 9262a7c5 by Alan Zimmerman at 2016-12-07T21:14:28+02:00 Match changes in GHC wip/T3384 branch - - - - - ac0eaf1a by Ben Gamari at 2016-12-09T09:48:41-05:00 haddock-api: Don't use stdcall calling convention on 64-bit Windows See GHC haskell/haddock#12890. - - - - - 04afe4f7 by Alan Zimmerman at 2016-12-12T20:07:21+02:00 Matching changes for GHC wip/T12942 - - - - - e1d1701d by Ben Gamari at 2016-12-13T16:50:41-05:00 Bump base upper bound - - - - - 3d3eacd1 by Alan Zimmerman at 2017-01-10T16:59:38+02:00 HsIParamTy now has a Located name - - - - - 7dbceefd by Kyrill Briantsev at 2017-01-12T13:23:50+03:00 Prevent GHC API from doing optimization passes. - - - - - d48d1e33 by Richard Eisenberg at 2017-01-19T08:41:41-05:00 Upstream changes re levity polymorphism - - - - - 40c25ed6 by Alan Zimmerman at 2017-01-26T15:16:18+02:00 Changes to match haskell/haddock#13163 in GHC - - - - - 504f586d by Ben Gamari at 2017-02-02T17:19:37-05:00 Kill remaining static flags - - - - - 49147ea0 by Justus Adam at 2017-03-02T15:33:34+01:00 Adding MDoc to exports of Documentation.Haddock - - - - - 1cfba9b4 by Justus Adam at 2017-03-09T11:41:44+01:00 Also exposing toInstalledIface - - - - - 53f0c0dd by Ben Gamari at 2017-03-09T13:10:08-05:00 Bump for GHC 8.3 - - - - - c7902d2e by Ben Gamari at 2017-03-09T23:46:02-05:00 Bump for GHC 8.2 - - - - - 4f3a74f8 by Ben Gamari at 2017-03-10T10:21:55-05:00 Merge branch 'ghc-head' - - - - - e273b72f by Richard Eisenberg at 2017-03-14T13:34:04-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - 6ec3d436 by Richard Eisenberg at 2017-03-14T15:15:52-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - eee3cda1 by Ben Gamari at 2017-03-15T15:19:59-04:00 Adapt to EnumSet - - - - - 017cf58e by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 736d6773 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 475f84a0 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 13240b53 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - cd16d529 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 3bea97ae by Edward Z. Yang at 2017-03-15T22:50:46-07:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b2b051ce by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 0f082795 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 20ef63c9 by Edward Z. Yang at 2017-03-22T13:48:12-07:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 45692dcb by Edward Z. Yang at 2017-03-22T14:11:25-07:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4eae8caf by Ben Gamari at 2017-03-23T09:25:33-04:00 Merge commit '240bc38b94ed2d0af27333b23392d03eeb615e82' into HEAD - - - - - 0bbe03f5 by Ben Gamari at 2017-03-23T09:27:28-04:00 haddock-api: Bump bound on GHC - - - - - 65f3ac9d by Alex Biehl at 2017-03-23T17:36:11+01:00 Merge pull request haskell/haddock#581 from JustusAdam/master Adding more exports to Documentation.Haddock - - - - - 37d49a47 by Alex Biehl at 2017-03-23T17:39:14+01:00 Merge pull request haskell/haddock#568 from awson/ghc-head Prevent GHC API from doing optimization passes. - - - - - 1ed047e4 by Brian Huffman at 2017-03-23T17:45:58+01:00 Print any user-supplied kind signatures on type parameters. This applies to type parameters on data, newtype, type, and class declarations, and also to forall-bound type vars in type signatures. - - - - - 1b78ca5c by Brian Huffman at 2017-03-23T17:45:58+01:00 Update test suite to expect kind annotations on type parameters. - - - - - a856b162 by Alex Biehl at 2017-03-23T17:49:32+01:00 Include travis build indication badge - - - - - 8e2e2c56 by Ben Gamari at 2017-03-23T17:20:08-04:00 haddock-api: Bump bound on GHC - - - - - 4d2d9995 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 26d6c150b31bc4580ab17cfd07b6e7f9afe10737) - - - - - a650e20f by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 930cfbe58e2e87f5a4d431d89a3c204934e6e858) - - - - - caa282c2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 2067a2d0afa9cef381d26fb7140b67c62f433fc0) - - - - - 49684884 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 0671abfe7e8ceae2269467a30b77ed9d9656e2cc) - - - - - 4dcfeb1a by Edward Z. Yang at 2017-03-23T17:20:08-04:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 3d77b373dd5807d5d956719dd7c849a11534fa6a) - - - - - 74dd19d2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 94610e9b446324f4231fa6ad4c6ac51e4eba8c0e) - - - - - a9b19a23 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 6cc832dfb1de6088a4abcaae62b25a7e944d55c3) - - - - - d3631064 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit bf3c4d72a0fda38561376eac7eda216158783267) - - - - - ef2148fc by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 07b88c5d4e79b87a319fbb08f8ea01dbb41063c1) - - - - - 2f29518b by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 4eb765ca4205c79539d60b7afa9b7e261a4a49fe) - - - - - 37de047d by Phil Ruffwind at 2017-04-03T11:57:14+02:00 Update MathJax URL MathJax is shutting down their CDN: https://www.mathjax.org/cdn-shutting-down/ They recommend migrating to cdnjs. - - - - - e9d24ba8 by David C. Turner at 2017-04-03T14:58:01+02:00 Add highlight for :target to ocean.css - - - - - 4819a202 by Alex Biehl at 2017-04-11T19:36:48+02:00 Allow base-4.10 for haddock-test - - - - - 44cec69c by Alex Biehl at 2017-04-11T19:39:22+02:00 cabal.project for haddock-api, haddock-library and haddock-test - - - - - 935d0f6a by Alex Biehl at 2017-04-11T19:46:29+02:00 Move dist scripts to scripts/ - - - - - 128e150c by Alex Biehl at 2017-04-11T20:34:46+02:00 Add haddock to cabal.project - - - - - cc8e08ea by Alex Biehl at 2017-04-11T20:35:08+02:00 Read files for hyperlinker eagerly This also exposes Documentation.Haddock.Utf8 - - - - - 152dda78 by Alex Biehl at 2017-04-11T20:37:06+02:00 Explicit import list ofr Control.DeepSeq in Haddock.Interface.Create - - - - - 501b33c4 by Kyrill Briantsev at 2017-04-11T21:01:42+02:00 Prevent GHC API from doing optimization passes. - - - - - c9f3f5ff by Alexander Biehl at 2017-04-12T16:36:53+02:00 Add @alexbiehl as maintaner - - - - - 76f214cc by Alex Biehl at 2017-04-13T07:27:18+02:00 Disable doctest with ghc-8.3 Currently doctest doesn't support ghc-head - - - - - 46b4f5fc by Edward Z. Yang at 2017-04-22T20:38:26-07:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f0555235 by Alex Biehl at 2017-04-25T10:08:48+02:00 Travis: Use ghc-8.2.1 on master - - - - - 966ea348 by Alex Biehl at 2017-04-25T10:32:01+02:00 Travis: Verbose cabal output cf. https://travis-ci.org/haskell/haddock/jobs/225512194#L377 - - - - - 36972bcd by Alex Biehl at 2017-04-25T10:40:43+02:00 Use travis_retry for cabal invocations - - - - - b3a09d2c by Alex Biehl at 2017-04-25T17:02:20+02:00 Use new MathJax URL in html-test 18ed871afb82560d5433b2f53e31b4db9353a74e switched to a new MathJax URL but didn't update the tests. - - - - - ae331e5f by Alexander Biehl at 2017-04-25T17:02:20+02:00 Expand signatures for class declarations - - - - - e573c65a by Alexander Biehl at 2017-04-25T17:02:20+02:00 Hoogle: Correctly print classes with associated data types - - - - - 3fc6be9b by Edward Z. Yang at 2017-04-25T17:02:20+02:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit a0c4790e15a2d3fab8d830eee8fcd639fe6d39c9) - - - - - 6725c060 by Herbert Valerio Riedel at 2017-04-25T17:02:20+02:00 `html-test --accept` deltas to reference samples - - - - - 7d444d61 by Alex Biehl at 2017-04-26T07:13:50+02:00 Remove anything related to obsolete frames mode - - - - - b888972c by Alex Biehl at 2017-04-26T07:49:10+02:00 Cherry-picked remaining commits from haddock-2.17.4-release (#603) * Release haddock/haddock-api 2.17.4 and haddock-library 1.4.3 * Set version bounds for haddock-library NB: This allows GHC 8.2.1's base * Set version bounds for haddock & haddock-api The version bounds support GHC 8.2 * Merge (temporary) v2.17.3 branch into v2.17 This allows us to delete the v2.17.3 branch * Fixup changelog * Pin down haddock-api to a single version as otherwise `haddock`'s package version has no proper meaning * fix source-repo spec for haddock-api - - - - - 4161099b by Alex Biehl at 2017-04-26T11:11:20+02:00 Update changelog to reflect news in HEAD - - - - - eed72cb8 by Alex Biehl at 2017-04-26T11:11:20+02:00 Markdownify changelog - - - - - 5815cea1 by Alex Biehl at 2017-04-26T11:32:33+02:00 Bump to 2.18.0 (#605) - - - - - a551d558 by Alex Biehl at 2017-04-29T22:00:25+02:00 Update attoparsec-0.12.1.1 to attoparsec-0.13.1.0 - - - - - ea164a8d by Sergey Vinokurov at 2017-04-29T22:42:36+02:00 Improve error message - - - - - 2e10122f by Alex Biehl at 2017-04-30T10:07:46+02:00 Correctly remember collapsed sections (#608) Now the "collapsed" cookie stores which sections have changed state instead of which are collapsed. - - - - - f9b24d99 by Alex Biehl at 2017-05-01T17:40:36+02:00 Lazily decode docMap and argMap (#610) These are only used in case of a doc reexport so most of the time decoding these is wasted work. - - - - - 2372af62 by Alex Biehl at 2017-05-01T21:59:23+02:00 Fix Binary instance for InstalledInterface (#611) (#610) introduced lazy decoding for docs from InstalledInterface but forgot to remove the original calls to get and put_ - - - - - 6c633c13 by Nathan Collins at 2017-05-11T11:47:55+02:00 Improve documenation of Haddock markup (#614) * Improve documentation of Haddock markup. - document that Haddock supports inferring types top-level functions with without type signatures, but also explain why using this feature is discouraged. Looks like this feature has been around since version 2.0.0.0 in 2008! - rework the "Module description" section: - move the general discussion of field formatting to the section intro and add examples illustrating the prose for multiline fields. - mention that newlines are preserved in some multiline fields, but not in others (I also noticed that commas in the `Copyright` field are not preserved; I'll look into this bug later). - add a subsection for the module description fields documentation, and put the field keywords in code formatting (double back ticks) instead of double quotes, to be consistent with the typesetting of keywords in other parts of the documentation. - mention that "Named chunks" are not supported in the long-form "Module description" documentation. - fix formatting of keywords in the "Module attributes" section. Perhaps these errors were left over from an automatic translation to ReST from some other format as part of the transition to using Sphinx for Haddock documentation? Also, add a missing reference here; it just said "See ?"! - update footnote about special treatment for re-exporting partially imported modules not being implemented. In my tests it's not implemented at all -- I tried re-exporting both `import B hiding (f)` and `import B (a, b)` style partial imports, and in both cases got the same result as with full imports `import B`: I only get a module reference. * Rework the `Controlling the documentation structure` section. My main goal was to better explain how to use Haddock without an export list, since that's my most common use case, but I hope I improved the section overall: - remove the incomplete `Omitting the export list` section and fold it into the other sections. In particular, summarize the differences between using and not using an export list -- i.e. control over what and in what order is documented -- in the section lead. - add "realistic" examples that use the structure markup, both with and without an export list. I wanted a realistic example here to capture how it can be useful to explain the relationship between a group of functions in a section, in addition to documenting their individual APIs. - make it clear that you can associate documentation chunks with documentation sections when you aren't using an export list, and that doing it in the most obvious way -- i.e. with `-- |`, as you can in the export list -- doesn't work without an export list. It took me a while to figure this out the first time, since the docs didn't explain it at all before. - add a "no export list" example to the section header section. - add more cross references. * Add examples of gotchas for markup in `@...@`. I'm not sure this will help anyone, since I think most people first learn about `@...@` by reading other people's Haddocks, but I've documented the mistakes which I've made and then gotten confused by. * Use consistent Capitalization of Titles. Some titles were in usual title caps, and others only had the first word capitalized. I chose making them all use title caps because that seems to make the cross references look better. - - - - - d4734f45 by Ben Gamari at 2017-05-12T20:36:08+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 93883f37 by Alex Biehl at 2017-05-12T21:02:33+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 5b8f179c by Alex Biehl at 2017-05-13T12:48:10+02:00 Consequently use inClass and notInClass in haddock-library (#617) These allow attoparsec to do some clever lookup optimization - - - - - 77984b82 by Doug Wilson at 2017-05-27T17:37:38+02:00 Don't enable compilation for template haskell (#624) This is no longer necessary after ghc commit 53c78be0aab76a3107c4dacbb1d177afacdd37fa - - - - - 5a3de2b4 by Doug Wilson at 2017-05-27T19:54:53+02:00 Improve Syb code (#621) Specialize.hs and Ast.hs are modified to have their Syb code not recurse into Name or Id in HsSyn types. Specialize.hs is refactored to have fewer calls to Syb functions. Syb.hs has some foldl calls replaced with foldl' calls. There is still a lot of performance on the floor of Ast.hs. The RenamedSource is traversed many times, and lookupBySpan is very inefficient. everywhereBut and lookupBySpan dominate the runtime whenever --hyperlinked-source is passed. - - - - - 3d35a949 by Alex Biehl at 2017-05-30T19:01:37+02:00 Clear fixme comment (#625) - - - - - 2a44bd0c by Alex Biehl at 2017-05-30T19:02:12+02:00 Make haddock-library and haddock-api warning free (#626) - - - - - bd1a0e42 by Alex Biehl at 2017-06-01T10:40:33+02:00 Include `driver-test/*.hs` sdist (#630) This lead to haskell/haddock#629. - - - - - 184a3ab6 by Doug Wilson at 2017-06-03T12:02:08+02:00 Disable pattern match warnings (#628) This disables the pattern match checker which can be very expensive in some cases. The disabled warnings include: * Opt_WarnIncompletePatterns * Opt_WarnIncompleteUniPatterns * Opt_WarnIncompletePatternsRecUpd * Opt_WarnOverlappingPatterns - - - - - 0cf68004 by Alex Biehl at 2017-06-03T20:37:28+02:00 Allow user defined signatures for pattern synonyms (#631) - - - - - 7f51a58a by Alex Biehl at 2017-06-04T11:56:38+02:00 Use NameSet for isExported check (#632) - - - - - d8f044a9 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Match new AST as per GHC wip/new-tree-one-param See https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow - - - - - da1254e3 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Rename extension index tags - - - - - 538c7514 by Christiaan Baaij at 2017-06-09T08:26:43+02:00 Haddock support for bundled pattern synonyms (#627) * Haddock support for bundled pattern synonyms * Add fixities to bundled pattern synonyms * Add bundled pattern synonyms to the synopsis * Store bundled pattern fixities in expItemFixities * Add test for bundled pattern synonyms * Stop threading fixities * Include bundled pattern synonyms for re-exported data types Sadly, fixity information isn't found for re-exported data types * Support for pattern synonyms * Modify tests after haskell/haddock#631 * Test some reexport variations * Also lookup bundled pattern synonyms from `InstalledInterface`s * Check isExported for bundled pattern synonyms * Pattern synonym is exported check * Always look for pattern synonyms in the current module Another overlooked cornercase * Account for types named twice in export lists Also introduce a fast function for nubbing on a `Name` and use it throughout the code base. * correct fixities for reexported pattern synonyms * Fuse concatMap and map * Remove obsolete import * Add pattern synonyms to visible exports * Fix test * Remove corner case - - - - - a050bffd by Doug Wilson at 2017-06-21T09:27:33+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#636) There is some performance improvement. GHC compiler: | version | bytes allocated | cpu_seconds --------------------------------- | before | 56057108648 | 41.0 | after | 51592019560 | 35.1 base: | version | bytes allocated | cpu_seconds --------------------------------- | before | 25174011784 | 14.6 | after | 23712637272 | 13.1 Cabal: | version | bytes allocated | cpu_seconds --------------------------------- | before | 18754966920 | 12.6 | after | 18198208864 | 11.6 - - - - - 5d06b871 by Doug Wilson at 2017-06-22T20:23:29+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#639) * Use new function getNameToInstancesIndex instead of tcRnGetInfo There is some significant performance improvement in the ghc testsuite. haddock.base: -23.3% haddock.Cabal: -16.7% haddock.compiler: -19.8% * Remove unused imports - - - - - b11bb73a by Alex Biehl at 2017-06-23T14:44:41+02:00 Lookup fixities for reexports without subordinates (#642) So we agree that reexported declarations which do not have subordinates (for example top-level functions) shouldn't have gotten fixities reexported according to the current logic. I wondered why for example Prelude.($) which is obviously reexported from GHC.Base has fixities attached (c.f. http://hackage.haskell.org/package/base-4.9.1.0/docs/Prelude.html#v:-36-). The reason is this: In mkMaps we lookup all the subordinates of top-level declarations, of course top-level functions don't have subordinates so for them the resulting list is empty. In haskell/haddock#644 I established the invariant that there won't be any empty lists in the subordinate map. Without the patch from haskell/haddock#642 top-level functions now started to fail reexporting their fixities. - - - - - d2a6dad6 by Alex Biehl at 2017-06-23T18:30:45+02:00 Don't include names with empty subordinates in maps (#644) These are unecessary anyway and just blow up interface size - - - - - 69c2aac4 by Alex Biehl at 2017-06-29T19:54:49+02:00 Make per-argument docs for class methods work again (#648) * Make per-argument docs for class methods work again * Test case - - - - - c9448d54 by Bartosz Nitka at 2017-07-02T12:12:01+02:00 Fix haddock: internal error: links: UnhelpfulSpan (#561) * Fix haddock: internal error: links: UnhelpfulSpan This fixes haskell/haddock#554 for me. I believe this is another fall out of `wildcard-refactor`, like haskell/haddock#549. * Comment to clarify why we take the methods name location - - - - - d4f29eb7 by Alex Biehl at 2017-07-03T19:43:04+02:00 Document record fields when DuplicateRecordFields is enabled (#649) - - - - - 9d6e3423 by Yuji Yamamoto at 2017-07-03T22:37:58+02:00 Fix test failures on Windows (#564) * Ignore .stack-work * Fix for windows: use nul instead of /dev/null * Fix for windows: canonicalize line separator * Also normalize osx line endings - - - - - 7d81e8b3 by Yuji Yamamoto at 2017-07-04T16:13:12+02:00 Avoid errors on non UTF-8 Windows (#566) * Avoid errors on non UTF-8 Windows Problem ==== haddock exits with errors like below: `(1)` ``` haddock: internal error: <stderr>: hPutChar: invalid argument (invalid character) ``` `(2)` ``` haddock: internal error: Language\Haskell\HsColour\Anchors.hs: hGetContents: invalid argument (invalid byte sequence) ``` `(1)` is caused by printing [the "bullet" character](http://www.fileformat.info/info/unicode/char/2022/index.htm) onto stderr. For example, this warning contains it: ``` Language\Haskell\HsColour\ANSI.hs:62:10: warning: [-Wmissing-methods] • No explicit implementation for ‘toEnum’ • In the instance declaration for ‘Enum Highlight’ ``` `(2)` is caused when the input file of `readFile` contains some Unicode characters. In the case above, '⇒' is the cause. Environment ---- OS: Windows 10 haddock: 2.17.3 GHC: 8.0.1 Solution ==== Add `hSetEncoding handle utf8` to avoid the errors. Note ==== - I found the detailed causes by these changes for debugging: - https://github.com/haskell/haddock/commit/8f29edb6b02691c1cf4c479f6c6f3f922b35a55b - https://github.com/haskell/haddock/commit/1dd23bf2065a1e1f2c14d0f4abd847c906b4ecb4 - These errors happen even after executing `chcp 65001` on the console. According to the debug code, `hGetEncoding stderr` returns `CP932` regardless of the console encoding. * Avoid 'internal error: <stderr>: hPutChar: invalid argument (invalid character)' non UTF-8 Windows Better solution for 59411754a6db41d17820733c076e6a72bcdbd82b's (1) - - - - - eded67d2 by Alex Biehl at 2017-07-07T19:17:15+02:00 Remove redudant import warning (#651) - - - - - 05114757 by Alex Biehl at 2017-07-08T00:33:12+02:00 Avoid missing home module warning (#652) * Avoid missing home module warning * Update haddock-library.cabal - - - - - e9cfc902 by Bryn Edwards at 2017-07-17T07:51:20+02:00 Fix haskell/haddock#249 (#655) - - - - - eb02792b by Herbert Valerio Riedel at 2017-07-20T09:09:15+02:00 Fix compilation of lib:haddock-library w/ GHC < 8 - - - - - 9200bfbc by Alex Biehl at 2017-07-20T09:20:38+02:00 Prepare 2.18.1 release (#657) - - - - - 46ddd22c by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Tweak haddock-api.cabal for pending release - - - - - 85e33d29 by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Avoid trivial use of LambdaCase otherwise we can't test w/ e.g. GHC 7.4.2 - - - - - 3afb4bfe by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Refactor .cabal to use sub-lib for vendored lib A practical benefit is that we can control the build-depends and also avoid some recompilation between library and test-suite. - - - - - e56a552e by Herbert Valerio Riedel at 2017-07-20T10:17:48+02:00 haddock-api: add changelog pointing to haddock's changelog This addresses https://github.com/haskell/haddock/issues/638#issuecomment-309283297 - - - - - 2222ff0d by Herbert Valerio Riedel at 2017-07-20T10:19:56+02:00 Drop obsolete/misleading `stability: experimental` This .cabal property has long been considered obsolete - - - - - 9b882905 by Alex Biehl at 2017-07-20T11:25:54+02:00 Beef up haddock description (#658) * Beef up haddock description * Handle empty lines - - - - - bb60e95c by Herbert Valerio Riedel at 2017-07-20T12:08:53+02:00 Import @aisamanra's Haddock cheatsheet from https://github.com/aisamanra/haddock-cheatsheet - - - - - 0761e456 by Herbert Valerio Riedel at 2017-07-20T12:12:55+02:00 Add cheatsheet to haddock.cabal - - - - - 2ece0f0f by Herbert Valerio Riedel at 2017-07-20T12:18:38+02:00 Mention new-build in README - - - - - 947b7865 by Herbert Valerio Riedel at 2017-07-20T12:32:16+02:00 Update README Also improves markup and removes/fixes redundant/obsolete parts [skip ci] - - - - - 785e09ad by Alex Biehl at 2017-07-27T07:28:57+02:00 Bump haddock to 2.18.2, haddock-library to 1.4.5 - - - - - e3ff1ca3 by Alex Biehl at 2017-07-31T20:15:32+02:00 Move `DocMarkup` from haddock-api to haddock-library (#659) * Move `DocMarkup` from haddock-api to haddock-library * Move more markup related functions * Markup module * CHANGELOG - - - - - cda7c20c by Alex Biehl at 2017-07-31T20:35:49+02:00 Fixup haddock - - - - - 583b6812 by Alex Biehl at 2017-07-31T21:20:45+02:00 Changelog for haddock-library - - - - - bac6a0eb by Alex Biehl at 2017-07-31T21:50:24+02:00 Prepare haddock-library-1.4.5 release - - - - - 58ce6877 by Moritz Drexl at 2017-08-05T16:44:40+02:00 Fix renaming after instance signature specializing (#660) * rework rename * Add regression test for Bug 613 * update tests * update changelog - - - - - b8137ec8 by Tim Baumann at 2017-08-06T11:33:38+02:00 Fix: Generate pattern signatures for constructors exported as patterns (#663) * Fix pretty-printing of pattern signatures Pattern synonyms can have up to two contexts, both having a different semantic meaning: The first holds the constraints required to perform the matching, the second contains the constraints provided by a successful pattern match. When the first context is empty but the second is not it is necessary to render the first, empty context. * Generate pattern synonym signatures for ctors exported as patterns This fixes haskell/haddock#653. * Simplify extractPatternSyn It is not necessary to generate the simplest type signature since it will be simplified when pretty-printed. * Add changelog entries for PR haskell/haddock#663 * Fix extractPatternSyn error message - - - - - d037086b by Alex Biehl at 2017-08-06T12:43:25+02:00 Bump haddock-library - - - - - 99d7e792 by Alex Biehl at 2017-08-06T12:44:07+02:00 Bump haddock-library in haddock-api - - - - - 94802a5b by Alex Biehl at 2017-08-06T13:18:02+02:00 Provide --show-interface option to dump interfaces (#645) * WIP: Provide --show-interface option to dump interfaces Like ghcs own --show-iface this flag dumps a binary interface file to stdout in a human (and machine) readable fashion. Currently it uses json as output format. * Fill all the jsonNull stubs * Rework Bifunctor instance of DocH, update changelog and documentation * replace changelog, bring DocMarkupH doc back * Update CHANGES.md * Update CHANGES.md * Move Control.Arrow up It would result in unused import if the Bifunctor instance is not generated. - - - - - c662e476 by Ryan Scott at 2017-08-14T21:00:21-04:00 Adapt to haskell/haddock#14060 - - - - - b891eb73 by Alex Biehl at 2017-08-16T08:24:48+02:00 Bifoldable and Bitraversable for DocH and MetaDoc - - - - - 021bb56c by Alex Biehl at 2017-08-16T09:06:40+02:00 Refactoring: Make doc renaming monadic This allows us to later throw warnings if can't find an identifier - - - - - 39fbf022 by Alex Biehl at 2017-08-19T20:35:27+02:00 Hyperlinker: Avoid linear lookup in enrichToken (#669) * Make Span strict in Position * Hyperlinker: Use a proper map to enrich tokens - - - - - e13baedd by Alex Biehl at 2017-08-21T20:05:42+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 27dd6e87 by Alex Biehl at 2017-08-21T22:06:35+02:00 Drop Avails from export list - - - - - 86b247e2 by Alex Biehl at 2017-08-22T08:44:22+02:00 Bump ghc version for haddock-api tests - - - - - d4607ca0 by Alex Biehl at 2017-08-22T08:45:17+02:00 Revert "Drop Avails from export list" This reverts commit a850ba86d88a4fb9c0bd175453a2580e544e3def. - - - - - c9c54c30 by Alex Biehl at 2017-08-22T09:26:01+02:00 IntefaceFile version - - - - - a85b7c02 by Ben Gamari at 2017-08-22T09:29:52-04:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 34e976f5 by Ben Gamari at 2017-08-22T17:40:06+02:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 577abf06 by Ryan Scott at 2017-08-23T14:47:29-04:00 Update for haskell/haddock#14131 - - - - - da68fc55 by Florian Eggenhofer at 2017-08-27T18:21:56+02:00 Generate an index for package content search (#662) Generate an index for package content search - - - - - 39e62302 by Alex Biehl at 2017-08-27T18:50:16+02:00 Content search for haddock html doc - - - - - 91fd6fb2 by Alex Biehl at 2017-08-28T18:39:58+02:00 Fix tests for content search - - - - - b4a3798a by Alex Biehl at 2017-08-28T18:44:08+02:00 Add search button to #page-menu - - - - - 25a7ca65 by Alex Biehl at 2017-08-28T18:47:43+02:00 Load javascript below the fold - - - - - 8d323c1a by Alex Biehl at 2017-08-28T18:49:22+02:00 Accept tests - - - - - c5dac557 by Alex Biehl at 2017-08-28T19:14:55+02:00 Content search css - - - - - 89a5af57 by Paolo Veronelli at 2017-08-29T07:42:13+02:00 Removed `nowrap` for interface method sigs (#674) with nowrap the interfaces method sigs would expand at libitum - - - - - a505f6f7 by Alex Biehl at 2017-08-29T08:05:33+02:00 Include subordinates in content index - - - - - 4bb698c4 by Alexander Biehl at 2017-08-29T11:40:19+02:00 QuickNav: Make docbase configurable - - - - - c783bf44 by Alexander Biehl at 2017-08-29T11:48:36+02:00 QuickNav: Also use baseUrl for doc-index.json request - - - - - 47017510 by Alex Biehl at 2017-08-29T17:56:47+02:00 Fix test fallout (again) - - - - - 924fc318 by Alex Biehl at 2017-08-30T09:24:56+02:00 Write meta.json when generating html output (#676) - - - - - 717dea52 by Alex Biehl at 2017-09-01T09:20:34+02:00 Use relative URL when no docBaseUrl given - - - - - e5d85f3b by Alex Biehl at 2017-09-01T09:35:19+02:00 Add missing js files to data-files (#677) - - - - - 95b9231a by Alex Biehl at 2017-09-01T11:01:36+02:00 Rename "Search" tab to "Quick Jump" - - - - - da0ead0b by Alex Biehl at 2017-09-01T13:03:49+02:00 Make trigger link configurable (#678) QuickNav: Configurable show/hide trigger - - - - - de7da594 by Ben Gamari at 2017-09-05T06:49:55-04:00 Account for "Remember the AvailInfo for each IE" As of GHC commit f609374a55bdcf3b79f3a299104767aae2ffbf21 GHC retains the AvailInfo associated with each IE. @alexbiehl has a patch making proper use of this change, but this is just to keep things building. - - - - - b05cd3b3 by Ben Gamari at 2017-09-14T07:55:07-04:00 Bump upper bound on base - - - - - 79db899e by Herbert Valerio Riedel at 2017-09-21T23:27:52+02:00 Make compatible with Prelude.<> export in GHC 8.4/base-4.11 - - - - - 3405dd52 by Tim Baumann at 2017-09-23T22:02:01+02:00 Add compile step that bundles and compresses JS files (#684) * Add compile step that bundles and compresses JS files Also, manage dependencies on third-party JS libraries using NPM. * Compile JS from TypeScript * Enable 'noImplicitAny' in TypeScript * QuickJump: use JSX syntax * Generate source maps from TypeScript for easier debugging * TypeScript: more accurate type * Separate quick jump css file from ocean theme - - - - - df0b5742 by Alex Biehl at 2017-09-29T21:15:40+02:00 Bump base for haddock-library and haddock-test - - - - - 62b12ea0 by Merijn Verstraaten at 2017-10-04T16:03:13+02:00 Inhibit output of coverage information for hidden modules. (#687) * Inhibit output of coverage information for hidden modules. * Add changelog entry. - - - - - 8daf8bc1 by Alexander Biehl at 2017-10-05T11:27:05+02:00 Don't use subMap in attachInstances - - - - - ad75114e by Alexander Biehl at 2017-10-05T11:27:58+02:00 Revert "Don't use subMap in attachInstances" This reverts commit 3adf5bcb1a6c5326ab33dc77b4aa229a91d91ce9. - - - - - 7d4aa02f by Alex Biehl at 2017-10-08T15:32:28+02:00 Precise Haddock: Use Avails for export resolution (#688) * Use Avails for export resolution * Support reexported modules * Factor out availExportItem * Use avails for fullModuleExports * Don't use subMap in attachInstances * lookupDocs without subMap * Completely remove subMap * Only calculate unqualified modules when explicit export list is given * Refactor * Refine comment * return * Fix * Refactoring * Split avail if declaration is not exported itself * Move avail splitting - - - - - b9b4faa8 by Alex Biehl at 2017-10-08T19:38:21+02:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 43325295 by Alex Biehl at 2017-10-08T20:18:46+02:00 Fix merge fallout - - - - - c6423cc0 by Alex Biehl at 2017-10-08T20:36:12+02:00 Copy QuickJump files over - - - - - 1db587c3 by Tim Baumann at 2017-10-09T18:33:09+02:00 Use <details> element for collapsibles (#690) * Remove unnecessary call to 'collapseSection' The call is unnecessary since there is no corresponding toggle for hiding the section of orphan instances. * Use <details> for collapsibles This makes them work even when JS is disabled. Closes haskell/haddock#560. - - - - - 1b54c64b by Tim Baumann at 2017-10-10T09:50:59+02:00 Quick Jump: Show error when loading 'doc-index.json' failed (#691) - - - - - 910f716d by Veronika Romashkina at 2017-10-24T07:36:20+02:00 Fix tiny typo in docs (#693) - - - - - b21de7e5 by Ryan Scott at 2017-10-24T13:07:15+02:00 Overhaul Haddock's rendering of kind signatures (#681) * Overhaul Haddock's rendering of kind signatures * Strip off kind signatures when specializing As an added bonus, this lets us remove an ugly hack specifically for `(->)`. Yay! * Update due to 0390e4a0f61e37bd1dcc24a36d499e92f2561b67 * @alexbiehl's suggestions * Import injectiveVarsOfBinder from GHC - - - - - 6704405c by Ryan Scott at 2017-10-28T07:10:27+02:00 Fix Haddock rendering of kind-indexed data family instances (#694) - - - - - 470f6b9c by Alex Biehl at 2017-10-30T08:45:51+01:00 Add QuickJump version to meta.json (#696) - - - - - b89eccdf by Alex Biehl at 2017-10-30T10:15:49+01:00 Put Quickjump behind --quickjump flag (#697) - - - - - 3095fb58 by Alex Biehl at 2017-10-30T19:09:06+01:00 Add build command to package.json - - - - - f223fda9 by Alex Biehl at 2017-10-30T19:10:39+01:00 Decrease threshold for fuzzy matching - - - - - 80245dda by Edward Z. Yang at 2017-10-31T20:35:05+01:00 Supported reexported-modules via --reexport flag. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 7e389742 by Alex Biehl at 2017-10-31T20:37:56+01:00 Correct missing title in changelog - - - - - 1a2a1c03 by Alex Biehl at 2017-10-31T20:59:07+01:00 Copy quickjump.css for nicer error messages - - - - - db234bb9 by Alex Biehl at 2017-10-31T21:31:18+01:00 Reexported modules: Report warnings if argument cannot be parsed or ... module cannot be found - - - - - eea8a205 by Carlo Hamalainen at 2017-10-31T21:43:14+01:00 More general type for nameCacheFromGhc. (#539) - - - - - 580eb42a by Alex Biehl at 2017-10-31T21:46:52+01:00 Remote tab - - - - - 0e599498 by Alex Biehl at 2017-10-31T21:48:55+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 7b8539bb by Alex Biehl at 2017-10-31T22:28:34+01:00 fullModuleContents traverses exports in declaration order - - - - - 0c91fbf2 by Alex Biehl at 2017-10-31T22:32:31+01:00 Remove excessive use of list comprehensions - - - - - f7356e02 by Alex Biehl at 2017-11-01T19:11:03+01:00 Make better use of AvailInfo - - - - - f3e512d5 by Alex Biehl at 2017-11-02T12:16:22+01:00 Always return documentation for exported subordinates ... event if they have no documentation (e.g. noDocForDecl) By using the information in the AvailInfo we don't need additional export checks. - - - - - 7cf58898 by Alan Zimmerman at 2017-11-07T08:28:03+02:00 Match changes for Trees that Grow in GHC - - - - - e5105a41 by Alan Zimmerman at 2017-11-08T17:21:58+02:00 Match Trees That Grow - - - - - 55178266 by Alan Zimmerman at 2017-11-11T22:20:31+02:00 Match Trees that Grow in GHC for HsExpr - - - - - 2082ab02 by Ryan Scott at 2017-11-14T15:27:03+01:00 Actually render infix type operators as infix (#703) * Actually render infix type operators as infix * Account for things like `(f :*: g) p`, too - - - - - c52ab7d0 by Alan Zimmerman at 2017-11-14T23:14:26+02:00 Clean up use of PlaceHolder, to match TTG - - - - - 81cc9851 by Moritz Angermann at 2017-11-20T07:52:49+01:00 Declare use of `Paths_haddock` module in other-modules (#705) This was detected by `-Wmissing-home-modules` - - - - - f9d27598 by Moritz Angermann at 2017-11-20T12:47:34+01:00 Drop Paths_haddock from ghc.mk (#707) With haskell/haddock#705 and haskell/haddock#706, the custom addition should not be necessary any more. # Conflicts: # ghc.mk - - - - - f34818dc by Moritz Angermann at 2017-11-20T12:47:59+01:00 Add autogen-modules (#706) > Packages using 'cabal-version: >= 1.25' and the autogenerated module Paths_* must include it also on the 'autogen-modules' field besides 'exposed-modules' and 'other-modules'. This specifies that the module does not come with the package and is generated on setup. Modules built with a custom Setup.hs script also go here to ensure that commands like sdist don't fail. # Conflicts: # haddock.cabal - - - - - bb43a0aa by Ben Gamari at 2017-11-21T15:50:12-05:00 Revert "Clean up use of PlaceHolder, to match TTG" This reverts commit 134a7bb054ea730b13c8629a76232d73e3ace049. - - - - - af9ebb2b by Ben Gamari at 2017-11-21T15:50:14-05:00 Revert "Match Trees that Grow in GHC for HsExpr" This reverts commit 9f054dc365379c66668de6719840918190ae6e44. - - - - - 5d35c3af by Ben Gamari at 2017-11-21T15:50:15-05:00 Revert "Match Trees That Grow" This reverts commit 73a26af844ac50b8bec39de11d64452a6286b00c. - - - - - 99a8e43b by Ben Gamari at 2017-11-21T16:36:06-05:00 Revert "Match changes for Trees that Grow in GHC" This reverts commit 01eeeb048acd2dd05ff6471ae148a97cf0720547. - - - - - c4d650c2 by Ben Gamari at 2017-12-04T15:06:07-05:00 Bump GHC version - - - - - 027b2274 by Ben Gamari at 2017-12-04T17:06:31-05:00 Bump GHC bound to 8.4.* - - - - - 58eaf755 by Alex Biehl at 2017-12-06T15:44:24+01:00 Update changelog - - - - - d68f5584 by Simon Peyton Jones at 2017-12-07T14:39:56+00:00 Track changes to follow Trac haskell/haddock#14529 This tracks the refactoring of HsDecl.ConDecl. - - - - - dc519d6b by Alec Theriault at 2018-01-06T08:20:43-08:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - 8285118c by Alec Theriault at 2018-01-13T12:12:37+01:00 Constructor and pattern synonym argument docs (#709) * Support Haddocks on constructor arguments This is in conjunction with https://phabricator.haskell.org/D4094. Adds support for rendering Haddock's on (non-record) constructor arguments, both for regular and GADT constructors. * Support haddocks on pattern synonym arguments It appears that GHC already parsed these - we just weren't using them. In the process of doing this, I tried to deduplicate some code around handling patterns. * Update the markup guide Add some information about the new support for commenting constructor arguments, and mention pattern synonyms and GADT-style constructors. * Overhaul LaTeX support for data/pattern decls This includes at least * fixing several bugs that resulted in invalid LaTeX * fixing GADT data declaration headers * overhaul handling of record fields * overhaul handling of GADT constructors * overhaul handling of bundled patterns * add support for constructor argument docs * Support GADT record constructors This means changes what existing HTML docs look like. As for LaTeX, looks like GADT records were never even supported. Now they are. * Clean up code/comments Made code/comments consistent between the LaTeX and XHTML backend when possible. * Update changelog * Patch post-rebase regressions * Another post-rebase change We want return values to be documentable on record GADT constructors. - - - - - ca4fabb4 by Alec Theriault at 2018-01-15T17:12:18-08:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 4c472fea by Ryan Scott at 2018-01-19T10:44:02+01:00 Fix haskell/haddock#732 (#733) - - - - - bff14dbd by Alex Biehl at 2018-01-19T15:33:30+01:00 extractDecl: Extract associated types correctly (#736) - - - - - a2a94a73 by Alex Biehl at 2018-01-19T15:34:40+01:00 extractDecl: Extract associated types correctly (#736) - - - - - 26df93dc by Alex Biehl at 2018-01-20T10:18:22+01:00 haddock-api: bump ghc to ^>= 8.4 - - - - - f65aeb1d by Alex Biehl at 2018-01-20T19:18:20+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - 0e721b97 by Alex Biehl at 2018-01-20T19:20:19+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - cb6234f6 by Ben Gamari at 2018-01-26T13:40:55-05:00 Merge remote-tracking branch 'harpocrates/fix/missing-orphan-instances' into ghc-head - - - - - 0fc28554 by Alec Theriault at 2018-02-01T14:58:18+01:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - b9123772 by Alec Theriault at 2018-02-01T14:58:18+01:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 0c12e274 by Ryan Scott at 2018-02-01T14:58:18+01:00 Fix haskell/haddock#548 by rendering datatype kinds more carefully (#702) - - - - - 8876d20b by Alec Theriault at 2018-02-01T14:58:18+01:00 Use the GHC lexer for the Hyperlinker backend (#714) * Start changing to use GHC lexer * better cpp * Change SrcSpan to RealSrcSpan * Remove error * Try to stop too many open files * wip * wip * Revert "wip" This reverts commit b605510a195f26315e3d8ca90e6d95a6737553e1. Conflicts: haddock-api/haddock-api.cabal haddock-api/src/Haddock/Interface.hs * Remove pointless 'caching' * Use dlist rather than lists when finding vars * Use a map rather than list * Delete bogus comment * Rebase followup Things now run using the GHC lexer. There are still - stray debug statements - unnecessary changes w.r.t. master * Cleaned up differences w.r.t. current Haddock HEAD Things are looking good. quasiquotes in particular look beautiful: the TH ones (with Haskell source inside) colour/link their contents too! Haven't yet begun to check for possible performance problems. * Support CPP and top-level pragmas The support for these is hackier - but no more hacky than the existing support. * Tests pass, CPP is better recognized The tests were in some cases altered: I consider the new output to be more correct than the old one.... * Fix shrinking of source without tabs in test * Replace 'Position'/'Span' with GHC counterparts Replaces 'Position' -> 'GHC.RealSrcLoc' and 'Span' -> 'GHC.RealSrcSpan'. * Nits * Forgot entry in .cabal * Update changelog - - - - - 95c6a771 by Alec Theriault at 2018-02-01T14:58:18+01:00 Clickable anchors for headings (#716) See haskell/haddock#579. This just adds an <a> tag around the heading, pointing to the heading itself. - - - - - 21463d28 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump: Matches on function names weight more than matches in ... module names. - - - - - 8023af39 by Alex Biehl at 2018-02-01T14:58:18+01:00 Treat escaped \] better in definition lists (#717) This fixes haskell/haddock#546. - - - - - e4866dc1 by Alex Biehl at 2018-02-01T14:58:18+01:00 Remove scanner, takeWhile1_ already takes care of escaping - - - - - 9bcaa49d by Alex Biehl at 2018-02-01T14:58:18+01:00 Take until line feed - - - - - 01d2af93 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Add simple framework for running parser fixtures (#668) * Add simple framework for running parser fixtures * Compatible with tree-diff-0.0.0.1 * Use parseParas to parse fixtures This allows to test all syntactic constructs available in haddock markup. - - - - - 31128417 by Alec Theriault at 2018-02-01T14:58:18+01:00 Patch flaky parser test (#720) * Patch flaky parser test This test was a great idea, but it doesn't port over too well to using the GHC lexer. GHC rewrites its input a bit - nothing surprising, but we need to guard against those cases for the test. * Change instance head * Change use site - - - - - 9704f214 by Herbert Valerio Riedel at 2018-02-01T14:58:18+01:00 Include secondary LICENSE file in source dist - - - - - 51f25074 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Grid Tables (#718) * Add table examples * Add table types and adopt simple parser Simple parser is done by Giovanni Cappellotto (@potomak) in https://github.com/haskell/haddock/pull/577 It seems to support single fine full tables, so far from full RST-grid tables, but it's good start. Table type support row- and colspans, but obviously parser is lacking. Still TODO: - Latex backend. Should we use multirow package https://ctan.org/pkg/multirow?lang=en? - Hoogle backend: ? * Implement grid-tables * Refactor table parser * Add two ill-examples * Update CHANGES.md * Basic documentation for tables * Fix documentation example - - - - - 670d6200 by Alex Biehl at 2018-02-01T14:58:18+01:00 Add grid table example to cheatsheet (pdf and svg need to be regenerated thought) - - - - - 4262dec9 by Alec Theriault at 2018-02-01T14:58:18+01:00 Fix infinite loop when specializing instance heads (#723) * Fix infinite loop when specializing instance heads The bug can only be triggered from TH, hence why it went un-noticed for so long. * Add test for haskell/haddock#679 and haskell/haddock#710 - - - - - 67ecd803 by Alec Theriault at 2018-02-01T14:58:18+01:00 Filter RTS arguments from 'ghc-options' arguments (#725) This fixes haskell/haddock#666. - - - - - 7db26992 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump Scrollable overlay - - - - - da9ff634 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Hyperlinker: Adjust parser to new PFailed constructor - - - - - 7b7cf8cb by Alexander Biehl at 2018-02-01T14:58:18+01:00 Specialize: Add missing IdP annotations - - - - - 78cd7231 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Convert: Correct pass type - - - - - a2d0f590 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Warning free compilation - - - - - cd861cf3 by Alexander Biehl at 2018-02-01T14:58:18+01:00 hadock-2.19.0 / haddock-api-2.19.0 / haddock-library-1.5.0 - - - - - c6651b72 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Adjust changelogs - - - - - 1e93da0b by Alexander Biehl at 2018-02-01T14:58:18+01:00 haddock-library: Info about breaking changes - - - - - f9b11db8 by Alec Theriault at 2018-02-02T12:36:02+01:00 Properly color pragma contents in hyperlinker The hyperlinker backend now classifies the content of pragmas as 'TkPragma'. That means that in something like '{-# INLINE foo #-}', 'foo' still gets classified as a pragma token. - - - - - c40b0043 by Alec Theriault at 2018-02-02T12:36:02+01:00 Support the new 'ITcolumn_prag' token - - - - - 4a2a4d39 by Alex Biehl at 2018-02-03T12:11:55+01:00 QuickJump: Mitigate encoding problems on Windows - - - - - bb34503a by Alex Biehl at 2018-02-04T18:39:31+01:00 Use withBinaryFile - - - - - 637605bf by Herbert Valerio Riedel at 2018-02-05T09:48:32+01:00 Try GHC 8.4.1 for Travis CI job - - - - - 7abb67e4 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 try harder to build w/ GHC 8.4.1 - - - - - 8255cc98 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 Add `SPDX-License-Identifier` as alised for "license" module header tokens C.f. SPDX 2.1 - Appendix V https://spdx.org/spdx-specification-21-web-version#h.twlc0ztnng3b The tag should appear on its own line in the source file, generally as part of a comment. SPDX-License-Identifier: <SPDX License Expression> Cherry-picked from haskell/haddock#743 - - - - - 267cd23d by Herbert Valerio Riedel at 2018-02-05T10:24:34+01:00 Make test-suite SMP compatible - - - - - 95d4bf40 by Alec Theriault at 2018-02-05T22:01:04+01:00 Hyperlink pattern synonyms and 'module' imports (#744) Links to pattern synonyms are now generated, as well as links from modules in import lists. Fixes haskell/haddock#731. - - - - - 67838dcd by Alec Theriault at 2018-02-06T08:23:36+01:00 Don't warn about missing '~' (#746) This manually filters out '~' from the list of things to warn about. It truly makes no sense to warn on this since '~' has nothing it could link to - it is magical. This fixes haskell/haddock#532. - - - - - ab6c3f9f by Alec Theriault at 2018-02-06T08:24:47+01:00 Don't barf on 'HsSpliceTy' (#745) This handles 'HsSpliceTy's by replacing them with what they expand to. IIUC everything that is happening, 'renameHsSpliceTy' should not be able to fail for the inputs we feed it from GHC. This fixes haskell/haddock#574. - - - - - 92bf95ad by Alex Biehl at 2018-02-06T08:28:23+01:00 Rename: renameHsSpliceTy ttg - - - - - 3130b1e1 by Alex Biehl at 2018-02-06T09:02:14+01:00 Expand SigDs - - - - - c72adae5 by Alex Biehl at 2018-02-06T09:20:51+01:00 fullModuleContents: support named docs - - - - - de2e4dbf by Alex Biehl at 2018-02-06T13:56:17+01:00 Hyperlinker: Also link pattern synonym arguments - - - - - b7c98237 by Alex Biehl at 2018-02-09T18:44:23+01:00 Expand SigD in a better place In https://github.com/haskell/haddock/issues/287 we found that haddock-2.19.0 would miss documentation on class methods with multiples names. This patch uses expandSigDecls in a more sensible place. - - - - - 8f598b27 by Alec Theriault at 2018-02-11T12:29:56+01:00 Add module tooltips to linked identifiers (#753) No more clicking to figure out whether your bytestring is strict or lazy! - - - - - d812e65d by Alec Theriault at 2018-02-11T12:31:44+01:00 Add 'show' option to complement 'hide' (#752) * Add 'show' option to complement 'hide' The behaviour is for flags passed in the command line to override flags in file headers. In the command line, later flags override earlier ones. Fixes haskell/haddock#751 and haskell/haddock#266. * Add a '--show-all' option - - - - - 6676cecb by Alex Biehl at 2018-02-18T11:07:15-05:00 QuickJump: Mitigate encoding problems on Windows (cherry picked from commit 86292c54bfee2343aee84559ec01f1fc68f52231) - - - - - e753dd88 by Alex Biehl at 2018-02-18T17:59:54+01:00 Use withBinaryFile - - - - - 724dc881 by Tamar Christina at 2018-02-19T05:34:49+01:00 Haddock: support splitted include paths. (#689) - - - - - 9b6d6f50 by Alex Biehl at 2018-02-19T05:57:02+01:00 Teach the HTML backend how to render methods with multiple names - - - - - a74aa754 by Alexander Biehl at 2018-02-19T10:04:34+01:00 Hoogle/Latex: Remove use of partial function - - - - - 66d8bb0e by Alec Theriault at 2018-02-25T16:04:01+01:00 Fix file handle leak (#763) (#764) Brought back some mistakenly deleted code for handling encoding and eager reading of files from e0ada1743cb722d2f82498a95b201f3ffb303137. - - - - - bb92d03d by Alex Biehl at 2018-03-02T14:21:23+01:00 Enable running test suite with stock haddock and ghc using ``` $ cabal new-run -- html-test --haddock-path=$(which haddock) --ghc-path=$(which ghc) ``` - - - - - dddb3cb2 by Alex Biehl at 2018-03-02T15:43:21+01:00 Make testsuite work with haddock-1.19.0 release (#766) - - - - - f38636ed by Alec Theriault at 2018-03-02T15:48:36+01:00 Support unicode operators, proper modules Unicode operators are a pretty big thing in Haskell, so supporting linking them seems like it outweighs the cost of the extra machinery to force Attoparsec to look for unicode. Fixes haskell/haddock#458. - - - - - 09d89f7c by Alec Theriault at 2018-03-02T15:48:43+01:00 Remove bang pattern - - - - - d150a687 by Alex Biehl at 2018-03-02T15:48:48+01:00 fix test - - - - - d6fd71a5 by Alex Biehl at 2018-03-02T16:22:38+01:00 haddock-test: Be more explicit which packages to pass We now pass `-hide-all-packages` to haddock when invoking the testsuite. This ensures we don't accidentally pick up any dependencies up through ghc.env files. - - - - - 0932c78c by Alex Biehl at 2018-03-02T17:50:38+01:00 Revert "fix test" This reverts commit 1ac2f9569242f6cb074ba6e577285a4c33ae1197. - - - - - 52516029 by Alex Biehl at 2018-03-02T18:16:50+01:00 Fix Bug548 for real - - - - - 89df9eb5 by Alex Biehl at 2018-03-05T18:28:19+01:00 Hyperlinker: Links for TyOps, class methods and associated types - - - - - d019a4cb by Ryan Scott at 2018-03-06T13:43:56-05:00 Updates for haskell/haddock#13324 - - - - - 6d5a42ce by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump haddock-2.19.0.1, haddock-api-2.19.0.1, haddock-library-1.5.0.1 - - - - - c0e6f380 by Alex Biehl at 2018-03-10T18:25:57+01:00 Update changelogs for haddock-2.19.0.1 and haddock-library-1.5.0.1 - - - - - 500da489 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Update to QC 2.11 - - - - - ce8362e9 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Restore backward-compat with base-4.5 through base-4.8 - - - - - baae4435 by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump lower bound for haddock-library - - - - - 10b7a73e by Alex Biehl at 2018-03-10T18:25:57+01:00 Haddock: Straighten out base bound - - - - - a6096f7b by Alex Biehl at 2018-03-13T08:45:06+01:00 extractDecl: Extract constructor patterns from data family instances (#776) * extractDecl: Allow extraction of data family instance constructors * extractDecl: extract data family instance constructors - - - - - ba4a0744 by Simon Jakobi at 2018-03-14T08:26:42+01:00 Readme: Update GHC version (#778) - - - - - 8de157d4 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for definition lists - - - - - 425b46f9 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for links - - - - - d53945d8 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for inline links - - - - - f1dc7c99 by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Slightly unmangle output - - - - - 0879d31c by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Prevent stdout buffering - - - - - 1f9e5f1b by Simon Jakobi at 2018-03-14T20:39:29+01:00 haddock-library.cabal: Clean up GHC options - - - - - 066b891a by Simon Jakobi at 2018-03-14T20:39:29+01:00 Make a proper definition for the <link> parser - - - - - 573d6ba7 by Alec Theriault at 2018-03-21T09:16:57+01:00 Show where instances are defined (#748) * Indicate source module of instances Above instance, we now also display a link to the module where the instance was defined. This is sometimes helpful in figuring out what to import. * Source module for type/data families too * Remove parens * Accept tests - - - - - 99b5d28b by Alex Biehl at 2018-03-21T09:20:36+01:00 Prepare changelog for next release - - - - - 482d3a93 by Alex Biehl at 2018-03-23T15:57:36+01:00 Useful cost centres, timers and allocation counters (#785) * Add some useful cost-centres for profiling * Add withTiming for each haddock phase Invoking haddock with `--optghc=-ddump-timings` now shows the amount of time spent and the number of allocated bytes for each phase. - - - - - 773b41bb by Alec Theriault at 2018-03-27T08:35:59+02:00 @since includes package name (#749) * Metadoc stores a package name This means that '@since' annotations can be package aware. * Get the package name the right way This should extract the package name for `@since` annotations the right way. I had to move `modulePackageInfo` around to do this and, in the process, I took the liberty to update it. Since it appears that finding the package name is something that can fail, I added a warning for this case. * Silence warnings * Hide package for local 'since' annotations As discussed, this is still the usual case (and we should avoid being noisy for it). Although this commit is large, it is basically only about threading a 'Maybe Package' from 'Haddock.render' all the way to 'Haddock.Backends.Xhtml.DocMarkup.renderMeta'. * Bump binary interface version * Add a '--since-qual' option This controls when to qualify since annotations with the package they come from. The default is always, but I've left an 'external' variant where only those annotations coming from outside of the current package are qualified. * Make ParserSpec work * Make Fixtures work * Use package name even if package version is not available The @since stuff needs only the package name passed in, so it makes sense to not be forced to pass in a version too. - - - - - e42c57bc by Alex Biehl at 2018-03-27T08:42:50+02:00 haddock-2.19.1, haddock-api-2.19.1, haddock-library-1.6.0 - - - - - 8373a529 by Alex Biehl at 2018-03-28T10:17:11+02:00 Bump haddock and haddock-api to 2.20.0 - - - - - 5038eddd by Jack Henahan at 2018-04-03T13:28:12+02:00 Clear search string on hide for haskell/haddock#781 (#789) - - - - - 920ca1eb by Alex Biehl at 2018-04-03T16:35:50+02:00 Travis: Build with ghc-8.4.2 (#793) - - - - - a232f0eb by Alan Zimmerman at 2018-04-07T14:14:32+02:00 Match changes in GHC for D4199 Removing HasSourceText and SourceTextX classes. - - - - - ab85060b by Alan Zimmerman at 2018-04-09T21:20:24+02:00 Match GHC changes for TTG - - - - - 739302b6 by Alan Zimmerman at 2018-04-13T13:31:44+02:00 Match GHC for TTG implemented on HsBinds, D4581 - - - - - 2f56d3cb by Ryan Scott at 2018-04-19T11:42:58-04:00 Bump upper bound on base to < 4.13 See https://ghc.haskell.org/trac/ghc/ticket/15018. - - - - - a49df92a by Alex Biehl at 2018-04-20T07:31:44+02:00 Don't treat fixity signatures like declarations - - - - - d02c103b by Ryan Scott at 2018-04-24T11:20:11-04:00 Add regression test for haskell/haddock#413 Fixes haskell/haddock#413. - - - - - c7577f52 by Ryan Scott at 2018-04-24T13:51:06-07:00 Improve the Hoogle backend's treatment of type families (#808) Fixes parts 1 and 2 of haskell/haddock#806. - - - - - d88f85b1 by Alec Theriault at 2018-04-25T11:24:07-07:00 Replace 'attoparsec' with 'parsec' (#799) * Remove attoparsec with parsec and start fixing failed parses * Make tests pass * Fix encoding issues The Haddock parser no longer needs to worry about bytestrings. All the internal parsing work in haddock-library happens over 'Text'. * Remove attoparsec vendor * Fix stuff broken in 'attoparsec' -> 'parsec' * hyperlinks * codeblocks * examples Pretty much all issues are due to attoparsec's backtracking failure behaviour vs. parsec's non-backtracking failure behaviour. * Fix small TODOs * Missing quote + Haddocks * Better handle spaces before/after paragraphs * Address review comments - - - - - fc25e2fe by Alan Zimmerman at 2018-04-27T15:36:53+02:00 Match changes in GHC for TTG - - - - - 06175f91 by Herbert Valerio Riedel at 2018-05-01T18:11:09+02:00 Merge branch 'ghc-head' with 'ghc-8.4' - - - - - 879caaa8 by Alec Theriault at 2018-05-07T18:53:15-07:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 3e0120cb by Simon Jakobi at 2018-05-07T19:00:18-07:00 Add docs for some DocH constructors (#814) - - - - - 0a32c6db by Alec Theriault at 2018-05-08T02:15:45-07:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 8816e783 by Simon Jakobi at 2018-05-08T10:48:11-07:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - ad60366f by Ryan Scott at 2018-05-10T11:19:47-04:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - 03b7cc3b by Ryan Scott at 2018-05-10T11:24:38-04:00 Wibbles - - - - - b03dd563 by Chaitanya Koparkar at 2018-05-10T11:44:58-04:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 9f298a40 by Ben Gamari at 2018-05-13T17:36:04-04:00 Account for refactoring of LitString - - - - - ea3dabe7 by Ryan Scott at 2018-05-16T09:21:43-04:00 Merge pull request haskell/haddock#826 from haskell/T825 Remove Hoogle backend hack that butchers infix datatype names - - - - - 0d234f7c by Alec Theriault at 2018-05-23T11:29:05+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 15fc9712 by Simon Jakobi at 2018-05-31T04:17:47+02:00 Adjust to new HsDocString internals - - - - - 6f1e19a8 by Ben Gamari at 2018-06-02T16:18:58-04:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 0d0355d9 by Ryan Scott at 2018-06-04T21:26:59-04:00 DerivingVia changes - - - - - 0d93475a by Simon Jakobi at 2018-06-05T19:47:05+02:00 Bump a few dependency bounds (#845) - - - - - 5cbef804 by Alec Theriault at 2018-06-05T19:47:16+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 9094c56f by Alec Theriault at 2018-06-05T22:53:25+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 70188719 by Simon Jakobi at 2018-06-08T22:20:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - 495cd1fc by Chaitanya Koparkar at 2018-06-13T23:01:34+02:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 81088732 by Ben Gamari at 2018-06-13T23:01:34+02:00 Account for refactoring of LitString - - - - - 7baf6587 by Simon Jakobi at 2018-06-13T23:05:08+02:00 Adjust to new HsDocString internals - - - - - bb61464d by Ben Gamari at 2018-06-13T23:05:22+02:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 5d8cb87f by Ryan Scott at 2018-06-13T23:39:30+02:00 DerivingVia changes - - - - - 73d373a3 by Alec Theriault at 2018-06-13T23:39:30+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 4865e254 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Remove `ITtildehsh` token - - - - - b867db54 by Alec Theriault at 2018-06-13T23:39:30+02:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 9598e392 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Add docs for some DocH constructors (#814) - - - - - 8a59035b by Alec Theriault at 2018-06-13T23:39:30+02:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 29350fc8 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - 2590bbd9 by Ryan Scott at 2018-06-13T23:39:30+02:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - a9939fdc by Ryan Scott at 2018-06-13T23:39:30+02:00 Wibbles - - - - - a22f7df4 by Alec Theriault at 2018-06-13T23:39:30+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 8741015d by Simon Jakobi at 2018-06-13T23:39:30+02:00 Bump a few dependency bounds (#845) - - - - - 4791e1cc by Alec Theriault at 2018-06-13T23:39:30+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 311d3216 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - d0577817 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Complete FixitySig and FamilyDecl pattern matches - - - - - 055b3aa7 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Fix redundant import warnings - - - - - f9ce19b1 by Simon Jakobi at 2018-06-13T23:49:52+02:00 html-test: Accept output - - - - - 04604ea7 by Simon Jakobi at 2018-06-13T23:54:37+02:00 Bump bounds on Cabal - - - - - 0713b692 by Simon Jakobi at 2018-06-14T00:00:12+02:00 Merge branch 'ghc-head' into ghc-head-update-3 - - - - - c6a56bfd by Simon Jakobi at 2018-06-14T02:33:27+02:00 Bump ghc bound for haddock-api spec test-suite - - - - - 119d04b2 by Simon Jakobi at 2018-06-14T12:37:48+02:00 Travis: `--allow-newer` for all packages - - - - - 0e876e2c by Alex Biehl at 2018-06-14T15:28:52+02:00 Merge pull request haskell/haddock#857 from sjakobi/ghc-head-update-3 Update ghc-head - - - - - 5be46454 by Alec Theriault at 2018-06-14T21:42:45+02:00 Improved handling of interfaces in 'haddock-test' (#851) This should now work with an inplace GHC where (for instance) HTML directories may not be properly recorded in the package DB. - - - - - 96ab1387 by Vladislav Zavialov at 2018-06-14T17:06:21-04:00 Handle -XStarIsType - - - - - e518f8c4 by Ben Gamari at 2018-06-14T17:48:00-04:00 Revert unintentional reversion of fix of haskell/haddock#548 - - - - - 01b9f96d by Alan Zimmerman at 2018-06-19T11:52:22+02:00 Match changes in GHC for haskell/haddock#14259 - - - - - 7f8c8298 by Ben Gamari at 2018-06-19T18:14:27-04:00 Bump GHC version to 8.6 - - - - - 11c6b5d2 by Ryan Scott at 2018-06-19T23:17:31-04:00 Remove HsEqTy and XEqTy - - - - - b33347c2 by Herbert Valerio Riedel at 2018-06-20T23:14:52+02:00 Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. - - - - - f0d2460e by Herbert Valerio Riedel at 2018-06-20T23:28:46+02:00 Update Travis CI job - - - - - ef239223 by Herbert Valerio Riedel at 2018-06-20T23:32:41+02:00 Drop GHC HEAD from CI and update GHC to 8.4.3 It's a waste of resource to even try to build this branch w/ ghc-head; so let's not do that... - - - - - 41c4a9fa by Ben Gamari at 2018-06-20T18:26:20-04:00 Bump GHC version to 8.7 - - - - - 8be593dc by Herbert Valerio Riedel at 2018-06-21T22:32:15+02:00 Update CI job to use GHC 8.7.* - - - - - b91d334a by Simon Jakobi at 2018-06-30T13:41:38+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section - - - - - f707d848 by Alec Theriault at 2018-07-05T10:43:35-04:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. - - - - - a6d2b8dc by Alec Theriault at 2018-07-06T10:06:32-04:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case - - - - - 13819f71 by Alan Zimmerman at 2018-07-15T19:33:51+02:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 - - - - - c346aa78 by Simon Jakobi at 2018-07-19T12:29:32+02:00 haddock-library: Bump bounds for containers - - - - - 722e733c by Simon Jakobi at 2018-07-19T13:36:45+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] - - - - - f0bd83fd by Alec Theriault at 2018-07-19T14:39:57+02:00 Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. - - - - - 394053a8 by Simon Jakobi at 2018-07-19T14:58:07+02:00 haddock-library: Bump bounds for containers - - - - - 1bda11a2 by Alec Theriault at 2018-07-20T09:04:03+02:00 Add HEAD.hackage overlay (#887) * Add HEAD.hackage overlay * Add HCPKG variable - - - - - c7b4ab45 by Alec Theriault at 2018-07-20T12:01:16+02:00 Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test - - - - - c05d32ad by Alec Theriault at 2018-07-20T12:01:49+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output - - - - - 24b39ee4 by Alec Theriault at 2018-07-20T12:02:16+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. - - - - - cb9d2099 by Simon Jakobi at 2018-07-20T13:39:29+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) - - - - - 133f24f5 by Alec Theriault at 2018-07-20T13:39:29+02:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) - - - - - 11024149 by Alec Theriault at 2018-07-20T13:39:29+02:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) - - - - - de0c139e by Simon Jakobi at 2018-07-20T13:39:29+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) - - - - - 6435e952 by Alec Theriault at 2018-07-20T13:39:29+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) - - - - - 1461af39 by Alec Theriault at 2018-07-20T13:39:29+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) - - - - - 69d3bde1 by Alec Theriault at 2018-07-20T13:49:47+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) - - - - - 6a5c73c7 by Alec Theriault at 2018-07-20T13:50:00+02:00 Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests - - - - - 92ca94c6 by Alec Theriault at 2018-07-20T13:55:36+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) - - - - - 981bc7fa by Simon Jakobi at 2018-07-20T15:06:06+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers - - - - - 27e7c0c5 by Simon Jakobi at 2018-07-20T15:09:05+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 49e1a415 by Simon Jakobi at 2018-07-20T16:02:02+02:00 Update the ghc-8.6 branch (#889) * Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. * README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) * Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) * Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) * tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) * Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. * Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test * Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) * Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) * Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) * Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests * Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 5ca14bed by Simon Jakobi at 2018-07-20T16:05:47+02:00 Revert "Revert "Bump GHC version to 8.6"" That commit didn't belong onto the ghc-8.6 branch. This reverts commit acbaef3b9daf1d2dea10017964bf886e77a8e967. - - - - - 2dd600dd by Simon Jakobi at 2018-07-20T16:18:21+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. - - - - - fceb2422 by Simon Jakobi at 2018-07-20T16:18:21+02:00 outOfScope: Recommend qualifying the identifier - - - - - acea5d23 by Simon Jakobi at 2018-07-20T16:19:35+02:00 outOfScope: Recommend qualifying the identifier (cherry picked from commit 73707ed58d879cc04cb644c5dab88c39ca1465b7) - - - - - 1a83ca55 by Simon Jakobi at 2018-07-20T16:19:35+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. (cherry picked from commit d504a2864a4e1982e142cf88c023e7caeea3b76f) - - - - - 48374451 by Masahiro Sakai at 2018-07-20T17:06:42+02:00 Add # as a special character (#884) '#' has special meaning used for anchors and can be escaped using backslash. Therefore it would be nice to be listed as special characters. - - - - - 5e1a5275 by Alec Theriault at 2018-07-20T23:37:24+02:00 Let `haddock-test` bypass interface version check (#890) This means `haddock-test` might * crash during deserialization * deserialize incorrectly Still - it means things _might_ work where they were previously sure not to. - - - - - 27286754 by Yuji Yamamoto at 2018-07-23T08:16:01+02:00 Avoid "invalid argument (invalid character)" on non-unicode Windows (#892) Steps to reproduce and the error message ==== ``` > stack haddock basement ... snip ... Warning: 'A' is out of scope. Warning: 'haddock: internal error: <stdout>: commitBuffer: invalid argument (invalid character) ``` Environment ==== OS: Windows 10 ver. 1709 haddock: [HEAD of ghc-8.4 when I reproduce the error](https://github.com/haskell/haddock/commit/532b209d127e4cecdbf7e9e3dcf4f653a5605b5a). (I had to use this version to avoid another probrem already fixed in HEAD) GHC: 8.4.3 stack: Version 1.7.1, Git revision 681c800873816c022739ca7ed14755e85a579565 (5807 commits) x86_64 hpack-0.28.2 Related pull request ==== https://github.com/haskell/haddock/pull/566 - - - - - 6729d361 by Alec Theriault at 2018-07-23T13:52:56-07:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). - - - - - 8cf4e6b5 by Ryan Scott at 2018-07-27T11:28:03-04:00 eqTyCon_RDR now lives in TysWiredIn After GHC commit http://git.haskell.org/ghc.git/commit/f265008fb6f70830e7e92ce563f6d83833cef071 - - - - - 1ad251a6 by Alan Zimmerman at 2018-07-30T13:28:09-04:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 (cherry picked from commit e3926b50ab8a7269fd6904b06e881745f08bc5d6) - - - - - 8aea2492 by Richard Eisenberg at 2018-08-02T10:54:17-04:00 Update against new HsImplicitBndrs - - - - - e42cada9 by Alec Theriault at 2018-08-04T17:51:30+02:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 0e852512 by Alex Biehl at 2018-08-06T13:04:02+02:00 Make --package-version optional for --hoogle generation (#899) * Make --package-version optional for --hoogle generation * Import mkVersion * It's makeVersion not mkVersion - - - - - d2abd684 by Noel Bourke at 2018-08-21T09:34:18+02:00 Remove unnecessary backslashes from docs (#908) On https://haskell-haddock.readthedocs.io/en/latest/markup.html#special-characters the backslash and backtick special characters showed up with an extra backslash before them – I think the escaping is not (or no longer) needed for those characters in rst. - - - - - 7a578a9e by Matthew Pickering at 2018-08-21T09:34:50+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - aa3d4db3 by Matthew Pickering at 2018-08-21T09:37:34+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - ede91744 by Alec Theriault at 2018-08-21T09:42:52+02:00 Better test output when Haddock crashes on a test (#902) In particular: we report the tests that crashed seperately from the tests that produced incorrect output. In order for tests to pass (and exit 0), they must not crash and must produce the right output. - - - - - 4a872b84 by Guillaume Bouchard at 2018-08-21T09:45:57+02:00 Fix a typo (#878) - - - - - 4dbf7595 by Ben Sklaroff at 2018-08-21T12:04:09-04:00 Add ITcomment_line_prag token to Hyperlinker Parser This token is necessary for parsing #line pragmas inside nested comments. Reviewers: bgamari Reviewed By: bgamari Differential Revision: https://phabricator.haskell.org/D4935 - - - - - 9170b2a9 by Ben Gamari at 2018-08-21T17:55:15-04:00 Merge pull request haskell/haddock#893 from harpocrates/get-name-to-instances Accumulate explicitly which modules to load for 'attachInstances' - - - - - d57b57cc by Ben Gamari at 2018-08-21T17:59:13-04:00 Merge branch 'ghc-head' of github.com:haskell/haddock into ghc-head - - - - - 14601ca2 by Alec Theriault at 2018-08-21T19:09:37-04:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). (cherry picked from commit 5c7c596c51d69b92164e9ba920157b36ce2b2ec1) - - - - - 438c645e by Matthew Pickering at 2018-08-21T19:12:39-04:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 (cherry picked from commit e6aa8fb47b9477cc5ef5e46097524fe83e080f6d) - - - - - a80c5161 by Alec Theriault at 2018-08-21T22:06:40-07:00 Better rendering of unboxed sums/tuples * adds space after/before the '#' marks * properly reify 'HsSumTy' in 'synifyType' - - - - - 88456cc1 by Alec Theriault at 2018-08-21T22:06:40-07:00 Handle promoted tuples in 'synifyType' When we have a fully applied promoted tuple, we can expand it out properly. - - - - - fd1c1094 by Alec Theriault at 2018-08-21T22:19:34-07:00 Accept test cases - - - - - 6e80d9e0 by Alec Theriault at 2018-08-21T22:24:03-07:00 Merge pull request haskell/haddock#914 from harpocrates/feature/unboxed-stuff Better rendering of unboxed sums, unboxed tuples, promoted tuples. - - - - - 181a23f1 by Ben Gamari at 2018-08-23T15:53:48-04:00 Merge remote-tracking branch 'origin/ghc-8.6' into ghc-8.6 - - - - - 3a18c1d8 by Alec Theriault at 2018-08-27T14:15:25-07:00 Properly synify promoted list types We reconstruct promoted list literals whenever possible. That means that 'synifyType' produces '[Int, Bool, ()] instead of (Int ': (() ': (Bool ': ([] :: [Type])))) - - - - - b4794946 by Alec Theriault at 2018-09-03T07:19:55-07:00 Only look at visible types when synifying a 'HsListTy' The other types are still looked at when considering whether to make a kind signature or not. - - - - - a231fce2 by Alec Theriault at 2018-09-03T07:38:10-07:00 Merge pull request haskell/haddock#922 from harpocrates/promoted-lists Properly synify promoted list types - - - - - 0fdf044e by Ningning Xie at 2018-09-15T10:25:58-04:00 Update according to GHC Core changes - - - - - 7379b115 by Ningning Xie at 2018-09-15T15:40:18-04:00 update dataFullSig to work with Co Quantification This should have been in the previous patch, but wasn't. - - - - - cf84a046 by Alec Theriault at 2018-09-17T20:12:18-07:00 Fix/add to various docs * Add documentation for a bunch of previously undocumented options (fixes haskell/haddock#870) * Extend the documentation of `--hoogle` considerably (see haskell/haddock#807) * Describe how to add docs to `deriving` clauses (fixes haskell/haddock#912) * Fix inaccurate docs about hyperlinking infix identifiers (fixes haskell/haddock#780) - - - - - ae017935 by Alec Theriault at 2018-09-22T08:32:16-07:00 Update Travis - - - - - d95ae753 by Alec Theriault at 2018-09-22T09:34:10-07:00 Accept failing tests Also silence orphan warnings. - - - - - f3e67024 by Alec Theriault at 2018-09-22T09:41:23-07:00 Bump haddock-api-2.21.0, haddock-library-1.7.0 * Update CHANGELOGS * Update new versions in Cabal files * Purge references to ghc-8.4/master branches in README - - - - - 3f136d4a by Alec Theriault at 2018-09-22T10:53:31-07:00 Turn haddock-library into a minor release Fix some version bounds in haddock-library too. - - - - - b9def006 by Alec Theriault at 2018-09-22T13:07:35-07:00 keep cabal.project file - - - - - 4909aca7 by Alec Theriault at 2018-10-16T09:36:30-07:00 Build on 7.4 and 7.8 - - - - - 99d20a28 by Herbert Valerio Riedel at 2018-10-16T18:45:52+02:00 Minor tweak to package description - - - - - a8059618 by Herbert Valerio Riedel at 2018-10-16T18:47:24+02:00 Merge pull request haskell/haddock#945 haddock-api 2.21.0 and haddock-library 1.6.1 release - - - - - 2d9bdfc1 by Alec Theriault at 2018-10-16T10:54:21-07:00 Bump haddock-library to 1.7.0 The 1.6.1 release should've been a major bump, since types in the `Documentation.Haddock.Parser.Monad` module changed. This version makes that module internal (as it morally should be). - - - - - ed340cef by Alec Theriault at 2018-10-16T14:59:13-07:00 Merge branch 'ghc-8.4' into ghc-8.6 - - - - - 2821a8df by Alec Theriault at 2018-10-16T15:14:48-07:00 Merge branch 'ghc-8.6' into ghc-head - - - - - a722dc84 by Alec Theriault at 2018-10-16T16:28:55-07:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 63377496 by Alec Theriault at 2018-10-16T16:39:07-07:00 Update changelog - - - - - 099a0110 by Alec Theriault at 2018-10-16T16:49:28-07:00 Merge pull request haskell/haddock#942 from harpocrates/update-docs Fix & add to documentation - - - - - 0927416f by Alec Theriault at 2018-10-16T16:50:14-07:00 Set UTF-8 encoding before writing files (#934) This should fix haskell/haddock#929, as well as guard against future problems of this sort in other places. Basically replaces 'writeFile' (which selects the users default locale) with 'writeUtf8File' (which always uses utf8). - - - - - 83b7b017 by Alec Theriault at 2018-10-16T17:42:05-07:00 Output pattern synonyms in Hoogle backend (#947) * Output pattern synonyms in Hoogle backend We were previously weren't outputting _any_ pattern synonyms, bundled or not. Now, we output both. Fixes haskell/haddock#946. * Update changelog - - - - - 81e5033d by Alec Theriault at 2018-10-16T18:04:40-07:00 Release `haddock{,-api}-2.22.0` This version will accompany ghc-8.6.2 - - - - - 9661744e by Alex Biehl at 2018-10-18T08:14:32-07:00 Add NewOcean theme And make it the default theme. - - - - - 7ae6d722 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Improve appearance and readability These changes include: - use latest Haskell's logo colors - decrease #content width to improve readability - use nicer font - improve sizes and distances - - - - - 37f8703d by NunoAlexandre at 2018-10-18T08:14:32-07:00 Include custom font in the html head - - - - - 1d5e1d79 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Update html test reference files - - - - - 53b7651f by NunoAlexandre at 2018-10-18T08:14:32-07:00 Make it responsive - It makes small screens taking more space than larger ones - fixes a few issues present in small screens currently - make it look good across different screen sizes. - - - - - 6aa1aeb1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make the style consistent with hackage Several things are addressed here: - better responsive behaviour on the header - better space usage - consistent colors overall - other nit PR comments - - - - - 3a250c5c by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Place the package name before the menu links This supports the expected responsive menu design, where the package name appears above the menu links. - - - - - cae699b3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update html-test reference files The package name element in the package-header is now a div instead of a paragraph, and it is now above the menu ul.links instead of below. - - - - - 2ec7fd2d by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve synopsis style and code - Use CSS3 instead of loading pictures to show "+" and "-" symbols - Drop redundant code - - - - - 0c874c01 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Decrease space between code blocks There was too much space between code blocks as pointed out by reviewers. - - - - - 85568ce2 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Add an initial-scale property to all haddock pages This solves an issue reported about the content looking incredibly small on mobile devices. - - - - - c1538926 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Address several PR comments - Darken text color like done for hackage - Move synopsis to left side - Make table of contents stick to the left on wide screens - Wrap links to avoid page overflow - Improve expand/collapse buttons - Fix issue with content size on mobile devices - Fix issue with font-size on landscape mode - Increase width of the content - Change colors of table of contents and synopsis - Etc - - - - - e6639e5f by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make synopsis scrollable on wide screens When the synopsis is longer than the screen, you can’t see its end and you can't scroll down either, making the content unreachable. - - - - - 1f0591ff by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve information density - Reduce font size - Improve space between and within code blocks - Improve alignments - Improve spacing within sub-blocks - - - - - bf083097 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Minor adjustments Bring in some adjustments made to hackage: - link colors - page header show everything when package title is too long - - - - - 10375fc7 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix responsive triggers overlap issue The min and max width triggers have the same values, which caused the style resolution to take an intersection of both style declarations when the screen resolution had the size of the limts (say 1280px), causing an odd behaviour and look. - - - - - 95ff2f95 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix issue with menu alignment on firefox Reported and described here: https://github.com/haskell/haddock/pull/721#issuecomment-374668869 - - - - - dc86587e by Alex Biehl at 2018-10-18T08:14:32-07:00 Changelog entry for NewOcean - - - - - 27195e47 by Herbert Valerio Riedel at 2018-10-18T08:14:32-07:00 html-test --accept - - - - - 83f4f9c0 by Alex Biehl at 2018-10-18T08:14:32-07:00 Avoid name shadowing - - - - - 231487f1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font to PT Sans Also migrate some general text related changes from hackage. - - - - - 313db81a by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Use 'flex' to fix header alignment - - - - - 5087367b by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Misc of tweaks - Update link colors to hackage scheme - Tune spacing between content elements - Update footer style - Fix and improve code blocks identation - - - - - b08020df by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font in Xhtml.hs to PT Sans - - - - - 78ce06e3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve code blocks styling - Fix and improve spacing - Improve colors and borders - - - - - 81262d20 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make package-header caption backward-compatible The current html generator of this branch wraps the package-header caption as a div, which does not work (without style adjustments) with the old themes. Changing it from div to span does the trick, without needing to adjust the old stylesheets. - - - - - dc4475cb by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update test-suite reference html pages - - - - - 393d35d8 by Alec Theriault at 2018-10-18T08:25:36-07:00 Accept tests - - - - - a94484ba by Alec Theriault at 2018-10-21T10:29:29-07:00 Fix CHANGELOG - - - - - 8797eca3 by Alec Theriault at 2018-10-21T10:36:19-07:00 Update 'data-files' to include NewOcean stuff - - - - - 1ae51e4a by Simon Jakobi at 2018-10-23T11:29:14+02:00 Fix typo in a warning - - - - - 009ad8e8 by Alec Theriault at 2018-10-24T12:47:47-07:00 Update JS dependencies This was done via `npm audit fix`. I think this fixes haskell/haddock#903 along with some more serious vulnerabilities that nobody seems to have noticed. - - - - - 051994db by Alec Theriault at 2018-10-24T17:31:09-07:00 Resurrect the style-switcher This fixes haskell/haddock#810. Looks like things were broken during the quickjump refactor of the JS. For the (git) record: I do not think the style switcher is a good idea. I'm fixing it for the same reason @mzero added it; as an answer to "rumblings from some that they didn't want their pixels changed on bit" - - - - - 2a1d620f by Alec Theriault at 2018-10-24T17:38:07-07:00 Fix copy-pasta error in data-files - - - - - ed5bfb7f by Alec Theriault at 2018-10-24T20:42:14-07:00 Fix the synopsis button Here's these changes are supposed to do: * put the synopsis back on the right side * properly have it on the edge of the screen on wide screens * adjust the background of the synopsis to match the button (otherwise the grey blends in with what is underneath) * get rid of the dotted purple line * the synopsis contents are now scrollable even when in wide screens (this has been a long-standing bug) - - - - - 883fd74b by Alec Theriault at 2018-10-25T20:16:46-07:00 Avoid more conflicts in generated ids (#954) This fixes haskell/haddock#953 by passing more names into the generated ids. - - - - - ea54e331 by Alec Theriault at 2018-10-25T21:07:12-07:00 Don't hide bullets in method docs I think thst CSS was meant only to deal with fields and the effect on bullets was accidental. Fixes haskell/haddock#926. - - - - - 9a14ef4a by Alec Theriault at 2018-10-25T22:02:07-07:00 Indent more things + slightly smaller font - - - - - b9f17e29 by Alec Theriault at 2018-10-25T22:10:01-07:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 096a3cfa by Alec Theriault at 2018-10-25T22:24:38-07:00 Accept HTML output - - - - - 2669517d by Alec Theriault at 2018-10-26T09:02:35-07:00 User manual + stuff for building GHC docs - - - - - 46b27687 by Alec Theriault at 2018-10-26T09:10:59-07:00 Make 'Contents' in NewOcean scrollable This only happens if the contents block on the left is so big that it doesn't fit (vertically) on the page. If that happens, we want it to be scrollable. - - - - - 3443dd94 by Alec Theriault at 2018-10-26T09:36:46-07:00 Revert "Make 'Contents' in NewOcean scrollable" This reverts commit f909ffd8353d6463fd5dd184998a32aa98d5c922. I missed the fact this also forces the 'Contents' to always go down to the bottom of the page. - - - - - ed081424 by Alec Theriault at 2018-10-26T14:22:23-07:00 Avoid some partiality AFAICT this wasn't causing any crashes, but that's mostly because we happen not to be forcing `pkgStr` when it would diverge. We come dangerously close to doing that in `ppHtmlIndex`. Fixes haskell/haddock#569. - - - - - 6a5bec41 by Alec Theriault at 2018-10-27T10:05:04-07:00 Fix documentation in `haddock-api` (#957) * Fix misplaced Haddocks in Haddock itself Haddock should be able to generate documentation for 'haddock-api' again. * Make CI check that documentation can be built. * Add back a doc that is OK - - - - - 5100450a by Matthew Yacavone at 2018-10-27T14:51:38-04:00 More explicit foralls (GHC Proposal 0007) - - - - - 8771a6b0 by Alec Theriault at 2018-11-05T13:58:11-08:00 Only run MathJax on entities with "mathjax" class (#960) Correspondingly, we wrap all inline/diplay math in <span class="mathjax"> ... the math .... </span> This fixes haskell/haddock#959. - - - - - bd7ff5c5 by Alec Theriault at 2018-11-05T15:54:22-08:00 Deduplicate some work in 'AttachInstances' Perf only change: * avoid needlessly union-ing maps * avoid synify-ing instances twice Took this opportunity to add some docs too - - - - - cf99fd8f by Alec Theriault at 2018-11-05T15:54:22-08:00 Specialize some SYB functions Perf only change: * Add a 'SPECIALIZE' pragma to help GHC optimize a 'Data a =>' constraint * Manually specialize the needlessly general type of 'specializeTyVarBndrs' - - - - - 4f91c473 by Alec Theriault at 2018-11-05T15:54:22-08:00 Improve perf of renaming Perf only change: * don't look up type variable names (they're never in the environment) * use a difference list for accumulating missing names * more efficient 'Functor'/'Applicative' instances for 'RnM' - - - - - 4bbab0d4 by Alec Theriault at 2018-11-05T15:54:22-08:00 Faster 'Text' driven parser combinators Perf only change: * use 'getParserState'/'setParserState' to make 'Text'-optimized parser combinators * minimize uses of 'Data.Text.{pack,unpack,cons,snoc}' - - - - - fa430c02 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support hyperlink labels with inline markup The parser for pictures hasn't been properly adjusted yet. - - - - - c1431035 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support (and flatten) inline markup in image links Inline markup is supported in image links but, as per the [commonmark recommendation][0], it is stripped back to a plain text representation. [0]: https://spec.commonmark.org/0.28/#example-547 - - - - - d4ee1ba5 by Alec Theriault at 2018-11-06T12:03:24-08:00 Accept test case - - - - - 8088aeb1 by Alec Theriault at 2018-11-06T12:03:24-08:00 Fix/add to haddock-library test suite - - - - - e78f644d by Alec Theriault at 2018-11-06T13:26:31-08:00 Bump version bounds - - - - - 644335eb by Alec Theriault at 2018-11-06T13:53:30-08:00 Merge pull request haskell/haddock#875 from harpocrates/feature/markup-in-hyperlinks Inline markup in markdown-style links and images - - - - - e173ed0d by Alec Theriault at 2018-11-07T12:37:18-08:00 Fix issues around plus/minus * swap the minimize unicode to something more intuitive * use new unicode expander/collapser for instance lists * address some alignment issues in the "index" page - - - - - b2d92df7 by Alec Theriault at 2018-11-07T13:41:57-08:00 Allow "Contents" summary to scroll in a fixed div In the unfortunate event that the "Contents" summary doesn't fit vertically (like in the "Prelude"), it will be scrollable. - - - - - ca704c23 by Alec Theriault at 2018-11-07T13:45:15-08:00 Accept HTML output changes - - - - - 82c0ec6d by Alec Theriault at 2018-11-07T18:12:54-08:00 overflow-y 'scroll' -> 'auto' - - - - - 571d7657 by Alec Theriault at 2018-11-08T19:44:12-08:00 Clicking on "Contents" navigates to top of page - - - - - 8065a012 by Alec Theriault at 2018-11-08T19:44:17-08:00 Space out functions more Also, functions and data decls now have the same space before and after them. - - - - - cc650ede by Alec Theriault at 2018-11-09T08:13:35-08:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 65f8c17f by Alec Theriault at 2018-11-10T14:04:06-08:00 Update changelog - - - - - 20473847 by Alec Theriault at 2018-11-10T14:21:40-08:00 Replace oplus/ominus expander/collapser icons with triangles - - - - - 16592957 by Alec Theriault at 2018-11-10T14:35:10-08:00 Merge pull request haskell/haddock#949 from haskell/wip/new-ocean Introduce NewOcean theme. - - - - - 357cefe1 by Alec Theriault at 2018-11-10T16:02:13-08:00 Merge branch 'ghc-8.6' into ghc-head - - - - - de612267 by Alec Theriault at 2018-11-11T20:01:21-08:00 Rename 'NewOcean' theme to 'Linuwial' - - - - - 954b5baa by Alec Theriault at 2018-11-12T08:33:18-08:00 Add blockquote styling Matches b71da1feabf33efbbc517ac376bb690b5a604c2f from hackage-server. Fixes haskell/haddock#967. - - - - - d32c0b0b by Fangyi Zhou at 2018-11-12T10:24:13-08:00 Fix some broken links (#15733) Summary: For links in subpackages as well. https://phabricator.haskell.org/D5257 Test Plan: Manually verify links Reviewers: mpickering, bgamari, osa1 Reviewed By: osa1 GHC Trac Issues: haskell/haddock#15733 Differential Revision: https://phabricator.haskell.org/D5262 - - - - - 41098b1f by Alp Mestanogullari at 2018-11-15T22:40:09+01:00 Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change It got introduced in ghc/ghc at ae2c9b40f5b6bf272251d1f4107c60003f541b62. - - - - - c5c1c7e0 by Alec Theriault at 2018-11-15T13:48:13-08:00 Merge pull request haskell/haddock#970 from alpmestan/alp/fix-promotionflag Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change - - - - - 6473d3a4 by Shayan-Najd at 2018-11-23T01:38:49+01:00 [TTG: Handling Source Locations] Foundation and Pat Trac Issues haskell/haddock#15495 This patch removes the ping-pong style from HsPat (only, for now), using the plan laid out at https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow/HandlingSourceLocations (solution A). - the class `HasSrcSpan`, and its functions (e.g., `cL` and `dL`), are introduced - some instances of `HasSrcSpan` are introduced - some constructors `L` are replaced with `cL` - some patterns `L` are replaced with `dL->L` view pattern - some type annotation are necessarily updated (e.g., `Pat p` --> `Pat (GhcPass p)`) - - - - - 7a088dfe by Alec Theriault at 2018-11-26T11:11:28-08:00 More uniform handling of `forall`'s in HTML/LaTeX * don't forget to print explicit `forall`'s when there are arg docs * when printing an explicit `forall`, print all tyvars Fixes haskell/haddock#973 - - - - - d735e570 by Alec Theriault at 2018-12-12T08:42:09-08:00 Fix warnings, accept output * remove redundant imports (only brought to light due to recent work for improving redundant import detection) * fix a bug that was casuing exports to appear in reverse order * fix something in haddock-library that prevented compilation on old GHC's - - - - - a3852f8a by Zejun Wu at 2018-12-14T09:37:47-05:00 Output better debug infromation on internal error in extractDecl This will make investigation of haskell/haddock#979 easier - - - - - 2eccb5b9 by Alec Theriault at 2018-12-17T09:25:10-05:00 Refactor names + unused functions (#982) This commit should not introduce any change in functionality! * consistently use `getOccString` to convert `Name`s to strings * compare names directly when possible (instead of comparing strings) * get rid of unused utility functions - - - - - e82e4df8 by Alec Theriault at 2018-12-20T16:16:30-05:00 Load plugins when compiling each module (#983) * WIP: Load (typechecker) plugins from language pragmas * Revert "Load plugins when starting a GHC session (#905)" This reverts commit 72d82e52f2a6225686d9668790ac33c1d1743193. * Simplify plugin initialization code - - - - - 96e86f38 by Alec Theriault at 2018-12-23T10:23:20-05:00 Properly synify and render promoted type variables (#985) * Synify and render properly promoted type variables Fixes haskell/haddock#923. * Accept output - - - - - 23343345 by Alec Theriault at 2018-12-27T16:39:38-05:00 Remove `haddock-test`'s dep. on `syb` (#987) The functionality is easily inlined into one short function: `gmapEverywhere`. This doesn't warrant pulling in another package. - - - - - d0734f21 by Alec Theriault at 2018-12-27T16:39:52-05:00 Address deprecation warnings in `haddock-test` (#988) Fixes haskell/haddock#885. - - - - - 4d9f144e by mynguyen at 2018-12-30T23:42:26-05:00 Visible kind application haddock update - - - - - ffe0e9ed by Alec Theriault at 2019-01-07T13:55:22-08:00 Print kinded tyvars in constructors for Hoogle (#993) Fixes haskell/haddock#992 - - - - - 2e18b55d by Alec Theriault at 2019-01-10T16:42:45-08:00 Accept new output `GHC.Maybe` -> `Data.Maybe` (#996) Since 53874834b779ad0dfbcde6650069c37926da1b79 in GHC, "GHC.Maybe" is marked as `not-home`. That changes around some test output. - - - - - 055da666 by Gabor Greif at 2019-01-22T14:41:51+01:00 Lone typofix - - - - - 01bb71c9 by Alec Theriault at 2019-01-23T11:46:46-08:00 Keep forall on H98 existential data constructors (#1003) The information about whether or not there is a source-level `forall` is already available on a `ConDecl` (as `con_forall`), so we should use it instead of always assuming `False`! Fixes haskell/haddock#1002. - - - - - f9b9bc0e by Ryan Scott at 2019-01-27T09:28:12-08:00 Fix haskell/haddock#1004 with a pinch of dropForAlls - - - - - 5cfcdd0a by Alec Theriault at 2019-01-28T16:49:57-08:00 Loosen 'QuickCheck' and 'hspec' bounds It looks like the new versions don't cause any breakage and loosening the bounds helps deps fit in one stack resolver. - - - - - 3545d3dd by Alec Theriault at 2019-01-31T01:37:25-08:00 Use `.hie` files for the Hyperlinker backend (#977) # Summary This is a large architectural change to the Hyperlinker. * extract link (and now also type) information from `.hie` instead of doing ad-hoc SYB traversals of the `RenamedSource`. Also adds a superb type-on-hover feature (#715). * re-engineer the lexer to avoid needless string conversions. By going directly through GHC's `P` monad and taking bytestring slices, we avoid a ton of allocation and have better handling of position pragmas and CPP. In terms of performance, the Haddock side of things has gotten _much_ more efficient. Unfortunately, much of this is cancelled out by the increased GHC workload for generating `.hie` files. For the full set of boot libs (including `ghc`-the-library) * the sum of total time went down by 9-10% overall * the sum of total allocations went down by 6-7% # Motivation Haddock is moving towards working entirely over `.hi` and `.hie` files. This change means we no longer need the `RenamedSource` from `TypecheckedModule` (something which is _not_ in `.hi` files). # Details Along the way a bunch of things were fixed: * Cross package (and other) links are now more reliable (#496) * The lexer tries to recover from errors on every line (instead of at CPP boundaries) * `LINE`/`COLUMN` pragmas are taken into account * filter out zero length tokens before rendering * avoid recomputing the `ModuleName`-based `SrcMap` * remove the last use of `Documentation.Haddock.Utf8` (see haskell/haddock#998) * restructure temporary folder logic for `.hi`/`.hie` model - - - - - 2ded3359 by Herbert Valerio Riedel at 2019-02-02T12:06:12+01:00 Update/modernise haddock-library.cabal file - - - - - 62b93451 by Herbert Valerio Riedel at 2019-02-02T12:19:31+01:00 Tentatively declare support for unreleased base-4.13/ghc-8.8 - - - - - 6041e767 by Herbert Valerio Riedel at 2019-02-02T16:04:32+01:00 Normalise LICENSE text w/ cabal's BSD2 template Also, correct the `.cabal` files to advertise `BSD2` instead of the incorrect `BSD3` license. - - - - - 0b459d7f by Alec Theriault at 2019-02-02T18:06:12-08:00 CI: fetch GHC from validate artifact Should help make CI be less broken - - - - - 6b5c07cf by Alec Theriault at 2019-02-02T18:06:12-08:00 Fix some Hyperlinker test suite fallout * Amend `ParserSpec` to match new Hyperlinker API - pass in compiler info - strip out null tokens * Make `hypsrc-test` pass reliably - strip out `local-*` ids - strip out `line-*` ids from the `ClangCppBug` test - re-accept output - - - - - ded34791 by Nathan Collins at 2019-02-02T18:31:23-08:00 Update README instructions for Stack No need to `stack install` Haddock to test it. Indeed, `stack install` changes the `haddock` on user's `PATH` if `~/.local/bin` is on user's `PATH` which may not be desirable when hacking on Haddock. - - - - - 723298c9 by Alec Theriault at 2019-02-03T09:11:05-08:00 Remove `Documentation.Haddock.Utf8` The circumstances under which this module appeared are completely gone. The Hyperlinker backend no longer needs this module (it uses the more efficient `Encoding` module from `ghc`). Why no deprecation? Because this module really shouldn't exist! - It isn't used in `haddock-library`/`haddock-api` anymore - It was copy pasted directly from `utf8-string` - Folks seeking a boot-lib only solution can use `ghc`'s `Encoding` - - - - - 51050006 by Alec Theriault at 2019-02-03T22:58:58-08:00 Miscellaneous improvements to `Convert` (#1020) Now that Haddock is moving towards working entirely over `.hi` and `.hie` files, all declarations and types are going to be synthesized via the `Convert` module. In preparation for this change, here are a bunch of fixes to this module: * Add kind annotations to type variables in `forall`'s whose kind is not `Type`, unless the kind can be inferred from some later use of the variable. See `implicitForAll` and `noKindTyVars` in particular if you wish to dive into this. * Properly detect `HsQualTy` in `synifyType`. This is done by following suit with what GHC's `toIfaceTypeX` does and checking the first argument of `FunTy{} :: Type` to see if it classified as a given/wanted in the typechecker (see `isPredTy`). * Beef up the logic around figuring out when an explicit `forall` is needed. This includes: observing if any of the type variables will need kind signatures, if the inferred type variable order _without_ a forall will still match the one GHC claims, and some other small things. * Add some (not yet used) functionality for default levity polymorphic type signatures. This functionality similar to `fprint-explicit-runtime-reps`. Couple other smaller fixes only worth mentioning: * Show the family result signature only when it isn't `Type` * Fix rendering of implicit parameters in the LaTeX and Hoogle backends * Better handling of the return kind of polykinded H98 data declarations * Class decls produced by `tyThingToLHsDecl` now contain associated type defaults and default method signatures when appropriate * Filter out more `forall`'s in pattern synonyms - - - - - 841980c4 by Oleg Grenrus at 2019-02-04T08:44:25-08:00 Make a fixture of weird parsing of lists (#997) The second example is interesting. If there's a list directly after the header, and that list has deeper structure, the parser is confused: It finds two lists: - One with the first nested element, - everything after it I'm not trying to fix this, as I'm not even sure this is a bug, and not a feature. - - - - - 7315c0c8 by Ryan Scott at 2019-02-04T12:17:56-08:00 Fix haskell/haddock#1015 with dataConUserTyVars (#1022) The central trick in this patch is to use `dataConUserTyVars` instead of `univ_tvs ++ ex_tvs`, which displays the foralls in a GADT constructor in a way that's more faithful to how the user originally wrote it. Fixes haskell/haddock#1015. - - - - - ee0b49a3 by Ryan Scott at 2019-02-04T15:25:17-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. - - - - - 1c850dc8 by Alan Zimmerman at 2019-02-05T21:54:18+02:00 Matching changes in GHC for haskell/haddock#16236 - - - - - ab03c38e by Simon Marlow at 2019-02-06T08:07:33+00:00 Merge pull request haskell/haddock#1014 from hvr/pr/bsd2-normalise Normalise LICENSE text w/ cabal's BSD2 template - - - - - 5a92ccae by Alec Theriault at 2019-02-10T06:21:55-05:00 Merge remote-tracking branch 'gitlab/wip/T16236-2' into ghc-head - - - - - c0485a1d by Alec Theriault at 2019-02-10T03:32:52-08:00 Removes `haddock-test`s dependency on `xml`/`xhtml` (#1027) This means that `html-test`, `latex-test`, `hoogle-test`, and `hypsrc-test` now only depend on GHC boot libs. So we should now be able to build and run these as part of GHC's testsuite. \o/ The reference output has changed very slightly, in three ways: * we don't convert quotes back into `&quot;` as the `xml` lib did * we don't add extra `&nbsp;` as the `xml` lib did * we now remove the entire footer `div` (instead of just emptying it) - - - - - 65a448e3 by Alec Theriault at 2019-02-11T12:27:41-05:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - 360ca937 by Alec Theriault at 2019-02-13T11:36:11-05:00 Clean up logic for guessing `-B` and `--lib` (#1026) Haddock built with the `in-ghc-tree` flag tries harder to find the GHC lib folder and its own resources. This should make it possible to use `in-ghc-tree`-built Haddock without having to specify the `-B` and `--lib` options (just how you can use in-tree GHC without always specifying the `-B` option). The logic to do this relies on `getExecutablePath`, so we only get this auto-detection on platforms where this function works. - - - - - d583e364 by Alec Theriault at 2019-02-16T10:41:22-05:00 Fix tests broken by GHC Changes in 19626218566ea709b5f6f287d3c296b0c4021de2 affected some of the hyperlinker output. Accepted the new output (hovering over a `..` now shows you what that wildcard binds). Also fixed some stray deprecation warnings. - - - - - da0c42cc by Vladislav Zavialov at 2019-02-17T11:39:19+03:00 Parser changes to match !380 - - - - - ab96bed7 by Ryan Scott at 2019-02-18T04:44:08-05:00 Bump ghc version to 8.9 - - - - - 44b7c714 by Alec Theriault at 2019-02-22T05:49:43-08:00 Match GHC changes for T16185 `FunTy` now has an `AnonArgFlag` that indicates whether the arrow is a `t1 => t2` or `t1 -> t2`. This commit shouldn't change any functionality in Haddock. - - - - - 2ee653b1 by Alec Theriault at 2019-02-24T18:53:33-08:00 Update .travis.yml Points to the new GHC CI artifact. - - - - - 90939d71 by Alec Theriault at 2019-02-25T00:42:41-08:00 Support value/type namespaces on identifier links Identifier links can be prefixed with a 'v' or 't' to indicate the value or type namespace of the desired identifier. For example: -- | Some link to a value: v'Data.Functor.Identity' -- -- Some link to a type: t'Data.Functor.Identity' The default is still the type (with a warning about the ambiguity) - - - - - d6ed496c by Alec Theriault at 2019-02-25T00:42:46-08:00 Better identifier parsing * '(<|>)' and '`elem`' now get parsed and rendered properly as links * 'DbModule'/'DbUnitId' now properly get split apart into two links * tuple names now get parsed properly * some more small niceties... The identifier parsing code is more precise and more efficient (although to be fair: it is also longer and in its own module). On the rendering side, we need to pipe through information about backticks/parens/neither all the way through from renaming to the backends. In terms of impact: a total of 35 modules in the entirety of the bootlib + ghc lib docs change. The only "regression" is things like '\0'. These should be changed to @\\0@ (the path by which this previously worked seems accidental). - - - - - 3c3b404c by Alec Theriault at 2019-02-25T22:12:11-08:00 Fix standalone deriving docs Docs on standalone deriving decls for classes with associated types should be associated with the class instance, not the associated type instance. Fixes haskell/haddock#1033 - - - - - d51ef69e by Alec Theriault at 2019-02-26T19:14:59-08:00 Fix bogus identifier defaulting This avoids a situation in which an identifier would get defaulted to a completely different identifier. Prior to this commit, the 'Bug1035' test case would hyperlink 'Foo' into 'Bar'! Fixes haskell/haddock#1035. - - - - - 88cbbdc7 by Ryan Scott at 2019-02-27T10:14:03-05:00 Visible dependent quantification (#16326) changes - - - - - 0dcf6cee by Xia Li-yao at 2019-02-27T21:53:27-05:00 Menu item controlling which instances are expanded/collapsed (#1007) Adds a menu item (like "Quick Jump") for options related to displaying instances. This provides functionality for: * expanding/collapsing all instances on the currently opened page * controlling whether instances are expanded/collapsed by default * controlling whether the state of instances should be "remembered" This new functionality is implemented in Typescript in `details-helper`. The built-in-themes style switcher also got a revamp so that all three of QuickJump, the style switcher, and instance preferences now have the same style and implementation structure. See also: https://mail.haskell.org/pipermail/haskell-cafe/2019-January/130495.html Fixes haskell/haddock#698. Co-authored-by: Lysxia <lysxia at gmail.com> Co-authored-by: Nathan Collins <conathan at galois.com> - - - - - 3828c0fb by Alec Theriault at 2019-02-28T12:42:49-05:00 `--show-interface` should output to stdout. (#1040) Fixes haskell/haddock#864. - - - - - a50f4cda by gbaz at 2019-03-01T07:43:16-08:00 Increase contrast of Linuwal theme (#1037) This is to address the concern that, on less nice and older screens, some of the shades of grey blend in too easily with the white background. * darken the font slightly * darken slightly the grey behind type signatures and such * add a border and round the corners on code blocks * knock the font down by one point - - - - - ab4d41de by Alec Theriault at 2019-03-03T09:23:26-08:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 12f509eb by Ben Gamari at 2019-03-04T22:13:20-05:00 Remove reference to Opt_SplitObjs flag Split-objects has been removed. - - - - - 5b3e4c9a by Ryan Scott at 2019-03-06T19:16:24-05:00 Update html-test output to reflect haskell/haddock#16391 changes - - - - - fc228af1 by Alec Theriault at 2019-03-09T08:29:23-08:00 Match changes for "Stop inferring over-polymorphic kinds" The `hsq_ext` field of `HsQTvs` is now just the implicit variables (instead of also including information about which of these variables are dependent). This commit shouldn't change any functionality in Haddock. - - - - - 6ac109eb by Alec Theriault at 2019-03-09T11:22:55-08:00 Add .hi, .dyn_hi, etc files to .gitignore Fixes haskell/haddock#1030. - - - - - b55f0c05 by Alec Theriault at 2019-03-09T11:22:55-08:00 Better support for default methods in classes * default methods now get rendered differently * default associated types get rendered * fix a forgotten `s/TypeSig/ClassOpSig/` refactor in LaTeX backend * LaTeX backend now renders default method signatures NB: there is still no way to document default class members and the NB: LaTeX backend still crashes on associated types - - - - - 10aea0cf by Alec Theriault at 2019-03-09T11:22:55-08:00 Avoid multi-line `emph` in LaTeX backend `markupWarning` often processes inputs which span across paragraphs. Unfortunately, LaTeX's `emph` is not made to handle this (and will crash). Fixes haskell/haddock#936. - - - - - d22dc2c9 by Alec Theriault at 2019-03-09T11:22:55-08:00 Many LaTeX backend fixes After this commit, we can run with `--latex` on all boot libraries without crashing (although the generated LaTeX still fails to compile in a handful of larger packages like `ghc` and `base`). * Add newlines after all block elements in LaTeX. This is important to prevent the final output from being more an more indented. See the `latext-test/src/Example` test case for a sample of this. * Support associated types in class declarations (but not yet defaults) * Several small issues for producing compiling LaTeX; - avoid empy `\haddockbeginargs` lists (ex: `type family Any`) - properly escape identifiers depending on context (ex: `Int#`) - add `vbox` around `itemize`/`enumerate` (so they can be in tables) * Several spacing fixes: - limit the width of `Pretty`-arranged monospaced code - cut out extra space characters in export lists - only escape spaces if there are _multiple_ spaces - allow type signatures to be multiline (even without docs) * Remove uninteresting and repetitive `main.tex`/`haddock.sty` files from `latex-test` test reference output. Fixes haskell/haddock#935, haskell/haddock#929 (LaTeX docs for `text` build & compile) Fixes haskell/haddock#727, haskell/haddock#930 (I think both are really about type families...) - - - - - 0e6cee00 by Alec Theriault at 2019-03-29T12:11:56-07:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - ce05434d by Alan Zimmerman at 2019-03-29T12:12:11-07:00 Matching changes in GHC for haskell/haddock#16236 (cherry picked from commit 3ee6526d4ae7bf4deb7cd1caf24b3d7355573576) - - - - - d85766b2 by Ben Gamari at 2019-03-29T12:14:04-07:00 Bump GHC to 8.8 - - - - - 5a82cbaf by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Redo ParseModuleHeader - - - - - b9033348 by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - bb55c8f4 by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove outdated `.ghci` files and `scripts` The `.ghci` files are actively annoying when trying to `cabal v2-repl`. As for the `scripts`, the distribution workflow is completely different. - - - - - 5ee244dc by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove obsolete arcanist files + STYLE Now that GHC is hosted on Gitlab, the arcanist files don't make sense anymore. The STYLE file contains nothing more than a dead link too. - - - - - d07c1928 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Redo ParseModuleHeader - - - - - 492762d2 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - af2ac773 by Ryan Scott at 2019-05-14T17:22:13-04:00 Changes for haskell/haddock#16110/#16356 - - - - - 6820ed0d by Alec Theriault at 2019-05-17T08:51:27-07:00 Unbreak haskell/haddock#1004 test case `fail` is no longer part of `Monad`. - - - - - 6bf7be98 by Alec Theriault at 2019-05-17T08:51:27-07:00 Fix haskell/haddock#1063 with better parenthesization logic for contexts The only other change in html/hoogle/hyperlinker output for the boot libraries that this caused is a fix to some Hoogle output for implicit params. ``` $ diff -r _build/docs/ old_docs diff -r _build/docs/html/libraries/base/base.txt old_docs/html/libraries/base/base.txt 13296c13296 < assertError :: (?callStack :: CallStack) => Bool -> a -> a --- > assertError :: ?callStack :: CallStack => Bool -> a -> a ``` - - - - - b5716b61 by Ryan Scott at 2019-05-22T17:24:32-04:00 Match changes with haskell/haddock#14332 - - - - - c115abf6 by Alec Theriault at 2019-05-26T16:01:58-04:00 Remove Haddock's dependency on `Cabal` At this point, Haddock depended on Cabal-the-library solely for a verbosity parser (which misleadingly accepts all sorts of verbosity options that Haddock never uses). Now, the only dependency on Cabal is for `haddock-test` (which uses Cabal to locate the Haddock interface files of a couple boot libraries). - - - - - e5b2d4a3 by Alec Theriault at 2019-05-26T16:16:25-04:00 Regression test: promoted lists in associated types When possible, associated types with promoted lists should use the promoted list literal syntax (instead of repeated applications of ': and '[]). This was fixed in 2122de5473fd5b434af690ff9ccb1a2e58491f8c. Closes haskell/haddock#466, - - - - - cc5ad5d3 by Alec Theriault at 2019-05-26T17:55:54-04:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 4b3301a6 by Alec Theriault at 2019-05-26T17:57:52-04:00 Release haddock-2.23, haddock-library-1.8.0 Tentatively adjust bounds and changelogs for the release to be bundled with GHC 8.8.1. - - - - - 69c7cfce by Matthew Pickering at 2019-05-30T10:54:27+01:00 Update hyperlinker tests for new types in .hie files - - - - - 29b7e738 by Zubin Duggal at 2019-05-30T10:57:51+01:00 update for new way to store hiefile headers - - - - - aeca5d5f by Zubin Duggal at 2019-06-04T18:57:42-04:00 update for new way to store hiefile headers - - - - - ba2ca518 by Ben Gamari at 2019-06-07T23:11:14+00:00 Update test output for introduction of Safe-Inferred - - - - - 3a975a6c by Ryan Scott at 2019-07-03T12:06:27-04:00 Changes for haskell/haddock#15247 - - - - - 0df46555 by Zubin Duggal at 2019-07-22T10:52:50+01:00 Fix haddockHypsrcTest - - - - - 2688686b by Sylvain Henry at 2019-09-12T23:19:39+02:00 Fix for GHC module renaming - - - - - 9ec0f3fc by Alec Theriault at 2019-09-20T03:21:00-04:00 Fix Travis CI, loosen .cabal bounds (#1089) Tentatively for the 2.23 release: * updated Travis CI to work again * tweaked bounds in the `.cabal` files * adjusted `extra-source-files` to properly identify test files - - - - - ca559beb by Matthías Páll Gissurarson at 2019-09-28T12:14:40-04:00 Small change in to facilitate extended typed-holes (#1090) This change has no functional effect on haddock itself, it just changes one pattern to use `_ (` rather than `_(`, so that we may use `_(` as a token for extended typed-holes later. - - - - - 02e28976 by Vladislav Zavialov at 2019-09-28T12:17:45-04:00 Remove spaces around @-patterns (#1093) This is needed to compile `haddock` when [GHC Proposal haskell/haddock#229](https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0229-whitespace-bang-patterns.rst) is implemented. - - - - - 83cbbf55 by Alexis King at 2019-09-30T21:12:42-04:00 Fix the ignore-exports option (#1082) The `ignore-exports` option has been broken since haskell/haddock#688, as mentioned in https://github.com/haskell/haddock/pull/766#issue-172505043. This PR fixes it. - - - - - e127e0ab by Ben Gamari at 2019-10-06T15:12:06-04:00 Fix a few haddock issues - - - - - 3a0f5c89 by Zubin Duggal at 2019-10-07T17:56:13-04:00 Fix crash when there are no srcspans in the file due to CPP - - - - - 339c5ff8 by Alec Theriault at 2019-10-07T17:56:13-04:00 Prefer un-hyperlinked sources to no sources It is possible to fail to extract an HIE ast. This is however not a reason to produce _no_ output - we should still make a colorized HTML page. - - - - - d47ef478 by Alec Theriault at 2019-10-07T17:56:13-04:00 Add a regression test for haskell/haddock#1091 Previously, this input would crash Haddock. - - - - - ed7c8b0f by Alec Theriault at 2019-10-07T20:56:48-04:00 Add Hyperlinker test cases for TH-related stuff Hopefully this will guard against regressions around quasiquotes, TH quotes, and TH splices. - - - - - d00436ab by Andreas Klebinger at 2019-10-21T15:53:03+02:00 Refactor for withTiming changes. - - - - - 4230e712 by Ben Gamari at 2019-10-22T09:36:37-04:00 Merge pull request haskell/haddock#1101 from AndreasPK/withTimingRefactor Refactor for withTiming changes. - - - - - d155c5f4 by Ryan Scott at 2019-10-23T10:37:17-04:00 Reify oversaturated data family instances correctly (#1103) This fixes haskell/haddock#1103 by adapting the corresponding patch for GHC (see https://gitlab.haskell.org/ghc/ghc/issues/17296 and https://gitlab.haskell.org/ghc/ghc/merge_requests/1877). - - - - - 331a5adf by Sebastian Graf at 2019-10-25T17:14:40+02:00 Refactor for OutputableBndrId changes - - - - - 48a490e0 by Ben Gamari at 2019-10-27T10:16:16-04:00 Merge pull request haskell/haddock#1105 from sgraf812/wip/flexible-outputable Refactor for OutputableBndrId changes - - - - - f62a7dfc by Sebastian Graf at 2019-11-01T11:54:16+00:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - d9b242ed by Ryan Scott at 2019-11-03T13:20:03-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. (cherry picked from commit cfd682c5fd03b099a3d78c44f9279faf56a0ac70) - - - - - dfd42406 by Sebastian Graf at 2019-11-04T07:02:14-05:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - 0b15be7c by Ben Gamari at 2019-11-09T13:21:33-05:00 Import isRuntimeRepVar from Type rather than TyCoRep isRuntimeRepVar is not longer exported from TyCoRep due to ghc#17441. - - - - - 091f7283 by Ben Gamari at 2019-11-10T12:47:06-05:00 Bump to GHC 8.10 - - - - - e88c71f2 by Ben Gamari at 2019-11-14T00:22:24-05:00 Merge pull request haskell/haddock#1110 from haskell/wip/T17441 Import isRuntimeRepVar from Type rather than TyCoRep - - - - - 4e0bbc17 by Ben Gamari at 2019-11-14T00:22:45-05:00 Version bumps for GHC 8.11 - - - - - 0e85ceb4 by Ben Gamari at 2019-11-15T11:59:45-05:00 Bump to GHC 8.10 - - - - - 00d6d68b by Ben Gamari at 2019-11-16T18:35:58-05:00 Bump ghc version to 8.11 - - - - - dde1fc3f by Ben Gamari at 2019-11-16T20:40:37-05:00 Drop support for base 4.13 - - - - - f52e331d by Vladislav Zavialov at 2019-11-24T13:02:28+03:00 Update Hyperlinker.Parser.classify to use ITdollar - - - - - 1ad96198 by Vladislav Zavialov at 2019-11-28T16:12:33+03:00 Remove HasSrcSpan (#17494) - - - - - 651afd70 by Herbert Valerio Riedel at 2019-12-08T12:08:16+01:00 Document error-prone conditional definition of instances This can easily trip up people if one isn't aware of it. Usually it's better to avoid this kind of conditionality especially for typeclasses for which there's an compat-package as conditional instances like these tend to fragment the ecosystem into those packages that go the extra mile to provide backward compat via those compat-packages and those that fail to do so. - - - - - b521af56 by Herbert Valerio Riedel at 2019-12-08T12:09:54+01:00 Fix build-failure regression for base < 4.7 The `$>` operator definition is available only since base-4.7 which unfortunately wasn't caught before release to Hackage (but has been fixed up by a metadata-revision) This commit introduces a `CompatPrelude` module which allows to reduce the amount of CPP by ousting it to a central location, i.e. the new `CompatPrelude` module. This pattern also tends to reduce the tricks needed to silence unused import warnings. Addresses haskell/haddock#1119 - - - - - 556c375d by Sylvain Henry at 2020-01-02T19:01:55+01:00 Fix after Iface modules renaming - - - - - bd6c53e5 by Sylvain Henry at 2020-01-07T00:48:48+01:00 hsyl20-modules-renamer - - - - - fb23713b by Ryan Scott at 2020-01-08T07:41:13-05:00 Changes for GHC#17608 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2372 - - - - - 4a4dd382 by Ryan Scott at 2020-01-25T08:08:26-05:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - e782a44d by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename PackageConfig into UnitInfo - - - - - ba3c9f05 by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename lookupPackage - - - - - ab37f9b3 by Ben Gamari at 2020-01-29T13:00:44-05:00 Merge pull request haskell/haddock#1125 from haskell/wip/T17566-take-two Changes for GHC#17566 - - - - - 3ebd5ae0 by Ryan Scott at 2020-01-31T05:56:50-05:00 Merge branch 'wip-hsyl20-package-refactor' into ghc-head - - - - - 602a747e by Richard Eisenberg at 2020-02-04T09:05:43+00:00 Echo GHC's removal of PlaceHolder module This goes with GHC's !2083. - - - - - ccfe5679 by Sylvain Henry at 2020-02-10T10:13:56+01:00 Module hierarchy: runtime (cf haskell/haddock#13009) - - - - - 554914ce by Cale Gibbard at 2020-02-10T16:10:39-05:00 Fix build of haddock in stage1 We have to use the correct version of the GHC API, but the version of the compiler itself doesn't matter. - - - - - 5b6fa2a7 by John Ericson at 2020-02-10T16:18:07-05:00 Noramlize `tested-with` fields in cabal files - - - - - e6eb3ebe by Vladislav Zavialov at 2020-02-16T13:25:26+03:00 No MonadFail/Alternative for P - - - - - 90e181f7 by Ben Gamari at 2020-02-18T14:13:47-05:00 Merge pull request haskell/haddock#1129 from obsidiansystems/wip/fix-stage1-build Fix build of haddock in stage1 - - - - - 93b64636 by Sylvain Henry at 2020-02-19T11:20:27+01:00 Modules: Driver (#13009) - - - - - da4f6c7b by Vladislav Zavialov at 2020-02-22T15:33:02+03:00 Use RealSrcSpan in InstMap - - - - - 479b1b50 by Ben Gamari at 2020-02-23T10:28:13-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 55ecacf0 by Sylvain Henry at 2020-02-25T15:18:27+01:00 Modules: Core (#13009) - - - - - 60867b3b by Vladislav Zavialov at 2020-02-28T15:53:52+03:00 Ignore the BufLoc/BufSpan added in GHC's !2516 - - - - - 1e5506d3 by Sylvain Henry at 2020-03-02T12:32:43+01:00 Modules: Core (#13009) - - - - - 6fb53177 by Richard Eisenberg at 2020-03-09T14:49:40+00:00 Changes in GHC's !1913. - - - - - 30b792ea by Ben Gamari at 2020-03-16T12:45:02-04:00 Merge pull request haskell/haddock#1130 from hsyl20/wip/hsyl20-modules-core2 Modules: Core (#13009) - - - - - cd761ffa by Sylvain Henry at 2020-03-18T15:24:00+01:00 Modules: Types - - - - - b6646486 by Ben Gamari at 2020-03-18T14:42:43-04:00 Merge pull request haskell/haddock#1133 from hsyl20/wip/hsyl20/modules/types Modules: Types - - - - - 9325d734 by Kleidukos at 2020-03-19T12:38:31-04:00 Replace the 'caption' class so that the collapsible sections are shown - - - - - 5e2bb555 by Kleidukos at 2020-03-19T12:38:31-04:00 Force ghc-8.8.3 - - - - - c6fcd0aa by Kleidukos at 2020-03-19T12:38:31-04:00 Update test fixtures - - - - - 5c849cb1 by Sylvain Henry at 2020-03-20T09:34:39+01:00 Modules: Types - - - - - 7f439155 by Alec Theriault at 2020-03-20T20:17:01-04:00 Merge branch 'ghc-8.8' into ghc-8.10 - - - - - b7904e5c by Alina Banerjee at 2020-03-20T20:24:17-04:00 Update parsing to strip whitespace from table cells (#1074) * Update parsing to strip leading & trailing whitespace from table cells * Update fixture data to disallow whitespaces at both ends in table cells * Add test case for whitespaces stripped from both ends of table cells * Update table reference test data for html tests - - - - - b9d60a59 by Alec Theriault at 2020-03-22T11:46:42-04:00 Clean up warnings * unused imports * imports of `Data.List` without import lists * missing `CompatPrelude` file in `.cabal` - - - - - 0c317dbe by Alec Theriault at 2020-03-22T18:46:54-04:00 Fix NPM security warnings This was done by calling `npm audit fix`. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. - - - - - 6e306242 by Alec Theriault at 2020-03-22T20:10:52-04:00 Tentative 2.24 release Adjusted changelogs and versions in `.cabal` files in preparation for the upcoming release bundled with GHC 8.10. - - - - - 1bfb4645 by Ben Gamari at 2020-03-23T16:40:54-04:00 Merge commit '3c2944c037263b426c4fe60a3424c27b852ea71c' into HEAD More changes from the GHC types module refactoring. - - - - - be8c6f3d by Alec Theriault at 2020-03-26T20:10:53-04:00 Update `.travis.yml` to work with GHC 8.10.1 * Regenerated the Travis file with `haskell-ci` * Beef up `.cabal` files with more `tested-with` information - - - - - b025a9c6 by Alec Theriault at 2020-03-26T20:10:53-04:00 Update README Removed some out of date links/info, added some more useful links. * badge to Hackage * update old trac link * `ghc-head` => `ghc-8.10` * `cabal new-*` is now `cabal v2-*` and it should Just Work * `--test-option='--accept'` is the way to accept testsuite output - - - - - 564d889a by Alec Theriault at 2020-03-27T20:34:33-04:00 Fix crash in `haddock-library` on unicode space Our quickcheck tests for `haddock-library` stumbled across an edge case input that was causing Haddock to crash: it was a unicode space character. The root cause of the crash is that we were implicitly assuming that if a space character was not " \t\f\v\r", it would have to be "\n". We fix this by instead defining horizontal space as: any space character that is not '\n'. Fixes haskell/haddock#1142 - - - - - 2d360ba1 by Alec Theriault at 2020-03-27T21:57:32-04:00 Disallow qualified uses of reserved identifiers This a GHC bug (https://gitlab.haskell.org/ghc/ghc/issues/14109) too, but it is a relatively easy fix in Haddock. Note that the fix must live in `haddock-api` instead of `haddock-library` because we can only really decide if an identifier is a reserved one by asking the GHC lexer. Fixes haskell/haddock#952 - - - - - 47ae22ed by Alec Theriault at 2020-03-28T13:36:25-04:00 Remove unused `Haddock.Utils` functions * removed functions in `Haddock.Utils` that were not used anywhere (or exported from the `haddock-api` package) * moved GHC-specific utils from `Haddock.Utils` to `Haddock.GhcUtils` - - - - - c0291245 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use TTG empty extensions to remove some `error`'s None of these error cases should ever have been reachable, so this is just a matter of leveraging the type system to assert this. * Use the `NoExtCon` and `noExtCon` to handle case matches for no extension constructors, instead of throwing an `error`. * Use the extension field of `HsSpliceTy` to ensure that this variant of `HsType` cannot exist in an `HsType DocNameI`. - - - - - 0aff8dc4 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use `unLoc`/`noLoc` from GHC instead of `unL`/`reL` * `unL` is already defined by GHC as `unLoc` * `reL` is already defined by GHC as `noLoc` (in a safer way too!) * Condense `setOutputDir` and add a about exporting from GHC Fixes haskell/haddock#978 - - - - - bf6f2fb7 by Alec Theriault at 2020-03-28T13:36:25-04:00 Cleanup up GHC flags in `.cabal` files * enable more useful warning flags in `haddock-api`, handle the new warnings generated * remove `-fwarn-tabs` (now we'd use `-Wtabs`, but this has been in `-Wall` for a while now) - - - - - c576fbf1 by Alec Theriault at 2020-03-28T13:36:25-04:00 `haddock-library` document header level Document the fact the header level is going to always be between 1 and 6 inclusive. Along the way, I also optimized the parsing code a bit. - - - - - 71bce0ee by Alec Theriault at 2020-03-28T14:26:27-04:00 Disallow links in section headers This is quite straightforward to implement, since we already had a function `docToHtmlNoAnchors` (which we used to generate the link in the sidebar "Contents"). This breaks test `Bug387`, but that test case has aged badly: we now automatically generate anchors for all headings, so manually adding an anchor in a section makes no sense. Nested anchors are, as pointed out in haskell/haddock#1054, disallowed by the HTML standard. Fixes haskell/haddock#1054 - - - - - b461b0ed by Sylvain Henry at 2020-03-30T10:34:23+02:00 Modules: type checker - - - - - cd8cd1ee by Ben Gamari at 2020-03-31T12:45:02-04:00 Merge pull request haskell/haddock#1152 from hsyl20/wip/hsyl20/modules/tc Module renaming - - - - - 5e8f8ea7 by Felix Yan at 2020-04-01T17:58:06-07:00 Allow QuickCheck 2.14 Builds fine and all tests pass. - - - - - dc6b1633 by Sylvain Henry at 2020-04-05T16:43:44+02:00 Module renaming: amend previous patch - - - - - eee2f4ae by Ryan Scott at 2020-04-05T09:04:43-07:00 Fix haskell/haddock#1050 by filtering out invisible AppTy arguments This makes the `synifyType` case for `AppTy` more intelligent by taking into consideration the visibilities of each `AppTy` argument and filtering out any invisible arguments, as they aren't intended to be displayed in the source code. (See haskell/haddock#1050 for an example of what can happen if you fail to filter these out.) Along the way, I noticed that a special `synifyType` case for `AppTy t1 (CoercionTy {})` could be consolidated with the case below it, so I took the opportunity to tidy this up. - - - - - 23eb99e8 by Ben Gamari at 2020-04-07T11:19:58-04:00 Merge pull request haskell/haddock#1154 from hsyl20/wip/hsyl20/modules/tc Module renaming: amend previous patch - - - - - 072d994d by Ryan Scott at 2020-04-07T19:32:47-04:00 Make NoExtCon fields strict These changes are a part of a fix for [GHC#17992](https://gitlab.haskell.org/ghc/ghc/issues/17992). - - - - - d8ebf6c8 by Ignat Insarov at 2020-04-09T21:15:01-04:00 Recode Doc to Json. (#1159) * Recode Doc to Json. * More descriptive field labels. - - - - - 52df4b4e by Sylvain Henry at 2020-04-10T12:39:18+02:00 Module renaming - - - - - d9ab8ec8 by Cale Gibbard at 2020-04-14T11:43:34-04:00 Add instance of XCollectPat for DocNameI - - - - - 323d221d by Cale Gibbard at 2020-04-14T11:43:34-04:00 Rename XCollectPat -> CollectPass - - - - - 2df80867 by Alec Theriault at 2020-04-15T07:30:51-07:00 Prune docstrings that are never rendered When first creating a Haddock interface, trim `ifaceDocMap` and `ifaceArgMap` to not include docstrings that can never appear in the final output. Besides checking with GHC which names are exported, we also need to keep all the docs attached to instance declarations (it is much tougher to detect when an instance is fully private). This change means: * slightly smaller interface files (7% reduction on boot libs) * slightly less work to do processing docstrings that aren't used * no warnings in Haddock's output about private docstrings (see haskell/haddock#1070) I've tested manually that this does not affect any of the boot library generated docs (the only change in output was some small re-ordering in a handful of instance lists). This should mean no docstrings have been incorrectly dropped. - - - - - f49c90cc by Alec Theriault at 2020-04-15T07:30:51-07:00 Don't warn about missing links in miminal sigs When renaming the Haddock interface, never emit warnings when renaming a minimal signature. Also added some documention around `renameInterface`. Minimal signatures intentionally include references to potentially un-exported methods (see the discussion in haskell/haddock#330), so it is expected that they will not always have a link destination. On the principle that warnings should always be resolvable, this shouldn't produce a warning. See haskell/haddock#1070. - - - - - a9eda64d by Ben Gamari at 2020-04-17T09:27:35-04:00 Merge pull request haskell/haddock#1160 from hsyl20/wip/hsyl20/modules/systools Module renaming - - - - - f40d7879 by Cale Gibbard at 2020-04-20T11:30:38-04:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ttg-con-pat - - - - - a50e7753 by Ben Gamari at 2020-04-20T11:36:10-04:00 Merge pull request haskell/haddock#1165 from obsidiansystems/wip/ttg-con-pat Trees that Grow refactor (GHC !2553) - - - - - 6a24795c by Alec Theriault at 2020-04-21T08:06:45-07:00 Fallback to `hiDecl` when `extractDecl` fails Sometimes, the declaration being exported is a subdecl (for instance, a record accessor getting exported at the top-level). For these cases, Haddock has to find a way to produce some synthetic sensible top-level declaration. This is done with `extractDecl`. As is shown by haskell/haddock#1067, this is sometimes impossible to do just at a syntactic level (for instance when the subdecl is re-exported). In these cases, the only sensible thing to do is to try to reify a declaration based on a GHC `TyThing` via `hiDecl`. - - - - - eee1a8b7 by Sylvain Henry at 2020-04-24T15:46:05+02:00 Module structure - - - - - 50b9259c by Iñaki at 2020-04-25T18:38:11-04:00 Add support for custom section anchors (#1179) This allows to have stable anchors for groups, even if the set of groups in the documentation is altered. The syntax for setting the anchor of a group is -- * Group name #desiredAnchor# Which will produce an html anchor of the form '#g:desiredAnchor' Co-authored-by: Iñaki García Etxebarria <git at inaki.blueleaf.cc> - - - - - 4003c97a by Ben Gamari at 2020-04-26T09:35:15-04:00 Merge pull request haskell/haddock#1166 from hsyl20/wip/hsyl20/modules/utils Module structure - - - - - 5206ab60 by Sylvain Henry at 2020-04-27T16:47:39+02:00 Renamed UnitInfo fields - - - - - c32c333b by Sylvain Henry at 2020-04-27T17:32:58+02:00 UnitId has been renamed into Unit - - - - - 3e87db64 by Sylvain Henry at 2020-04-27T17:36:00+02:00 Fix for GHC.Unit.* modules - - - - - ae3323a7 by Ben Gamari at 2020-04-29T12:36:37-04:00 Merge pull request haskell/haddock#1183 from hsyl20/wip/hsyl20/unitid Refactoring of Unit code - - - - - b105564a by Artem Pelenitsyn at 2020-05-03T08:14:10+01:00 add dependency on exceptions because GHC.Exception was boiled down (ghc haskell/haddock#18075) - - - - - 9857eff3 by Zubin Duggal at 2020-05-04T18:48:25+01:00 Atomic update of NameCache in readHieFile - - - - - 86bbb226 by Sylvain Henry at 2020-05-14T16:36:27+02:00 Fix after Config module renaming - - - - - a4bbdbc2 by Gert-Jan Bottu at 2020-05-15T22:09:44+02:00 Explicit Specificity Support for Haddock - - - - - 46199daf by Ben Gamari at 2020-05-19T09:59:56-04:00 Merge pull request haskell/haddock#1192 from hsyl20/hsyl20/modules-config Fix after Config module renaming - - - - - f9a9d2ba by Gert-Jan Bottu at 2020-05-20T16:48:38-04:00 Explicit Specificity Support for Haddock - - - - - 55c5b7ea by Ben Gamari at 2020-05-21T00:32:02-04:00 Merge commit 'a8d7e66da4dcc3b242103271875261604be42d6e' into ghc-head - - - - - a566557f by Cale Gibbard at 2020-05-21T16:02:06-04:00 isBootSummary now produces a result of type IsBootInterface - - - - - ea52f905 by Zubin Duggal at 2020-05-24T17:55:48+01:00 update for hiefile-typeclass-info - - - - - 49ba7a67 by Willem Van Onsem at 2020-05-25T12:23:01-04:00 Use floor over round to calculate the percentage (#1195) If we compile documentation where only a small fraction is undocumented, it is misleading to see 100% coverage - 99% is more intuitive. Fixes haskell/haddock#1194 - - - - - c025ebf1 by Ben Gamari at 2020-05-29T14:32:42-04:00 Merge pull request haskell/haddock#1185 from obsidiansystems/boot-disambig isBootSummary now produces a result of type IsBootInterface - - - - - 74ab9415 by Ben Gamari at 2020-05-29T20:23:39-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - b40be944 by Ben Gamari at 2020-06-03T17:02:31-04:00 testsuite: Update expected output for simplified subsumption - - - - - 624be71c by Ryan Scott at 2020-06-05T12:43:23-04:00 Changes for GHC#18191 See https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3337. - - - - - fbd8f7ce by Sylvain Henry at 2020-06-08T15:31:47+02:00 Fix after unit refactoring - - - - - 743fda4d by Ben Gamari at 2020-06-09T12:09:58-04:00 Merge pull request haskell/haddock#1202 from hsyl20/wip/hsyl20/unitid-ii Fix after unit refactoring - - - - - d07a06a9 by Ryan Scott at 2020-06-13T07:16:55-04:00 Use HsForAllTelescope (GHC#18235) - - - - - 389bb60d by Ben Gamari at 2020-06-13T15:30:52-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - 7a377f5f by Ben Gamari at 2020-06-17T14:53:16-04:00 Merge pull request haskell/haddock#1199 from bgamari/wip/ghc-8.12 haddock: Bounds bumps for GHC 8.12 - - - - - 9fd9e586 by Krzysztof Gogolewski at 2020-06-17T16:09:07-04:00 Adapt Haddock to LinearTypes See ghc/ghc!852. - - - - - 46fe7636 by Ben Gamari at 2020-06-18T14:20:02-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 35a3c9e2 by Zubin Duggal at 2020-06-21T21:19:18+05:30 Use functions exported from HsToCore - - - - - 8abe3928 by Ben Gamari at 2020-06-24T13:53:39-04:00 Merge pull request haskell/haddock#1204 from wz1000/wip/haddock-hstocore Use functions exported from GHC.HsToCore.Docs - - - - - 22f2c937 by Matthías Páll Gissurarson at 2020-06-26T19:07:03+02:00 Adapt Haddock for QualifiedDo - - - - - 3f6208d7 by Vladislav Zavialov at 2020-06-28T14:28:16+03:00 Handle LexicalNegation's ITprefixminus - - - - - 03a19f41 by Sylvain Henry at 2020-07-02T09:37:38+02:00 Rename hsctarget into backend - - - - - ea17ff23 by Andreas Klebinger at 2020-07-02T17:44:18+02:00 Update for UniqFM changes. - - - - - 9872f2f3 by Ben Gamari at 2020-07-09T10:39:19-04:00 Merge pull request haskell/haddock#1209 from AndreasPK/wip/typed_uniqfm Update for UniqFM changes. - - - - - 68f7b668 by Krzysztof Gogolewski at 2020-07-12T18:16:57+02:00 Sync with GHC removing {-# CORE #-} pragma See ghc ticket haskell/haddock#18048 - - - - - eb372681 by Sylvain Henry at 2020-07-20T11:41:30+02:00 Rename hscTarget into backend - - - - - fb7f78bf by Ben Gamari at 2020-07-21T12:15:25-04:00 Merge pull request haskell/haddock#1214 from hsyl20/wip/hsyl20/hadrian/ncg Rename hscTarget into backend - - - - - 1e8f5b56 by Ben Gamari at 2020-07-23T09:11:50-04:00 Merge commit '904dce0cafe0a241dd3ef355775db47fc12f434d' into ghc-head - - - - - d8fd1775 by Zubin Duggal at 2020-07-23T18:46:40+05:30 Update for modular ping pong - - - - - 8416f872 by Ben Gamari at 2020-07-23T09:35:03-04:00 Merge pull request haskell/haddock#1200 from wz1000/wip/wz1000-modular-ping-pong Modular ping pong - - - - - a24a8577 by Ben Gamari at 2020-07-28T15:23:36-04:00 Bump GHC version to 9.0 - - - - - 6a51c9dd by Sylvain Henry at 2020-08-05T18:47:05+02:00 Fix after Outputable refactoring - - - - - c05e1c99 by Ben Gamari at 2020-08-10T14:41:41-04:00 Merge pull request haskell/haddock#1223 from hsyl20/wip/hsyl20/dynflags/exception Fix after Outputable refactoring - - - - - d964f15b by Sylvain Henry at 2020-08-12T11:58:49+02:00 Fix after HomeUnit - - - - - 8e6d5b23 by Ben Gamari at 2020-08-12T14:25:30-04:00 Merge pull request haskell/haddock#1225 from hsyl20/wip/hsyl20/plugins/homeunit Fix after HomeUnit - - - - - 8c7880fe by Sylvain Henry at 2020-08-17T14:13:29+02:00 Remove Ord FastString instance - - - - - 8ea410db by Alex Biehl at 2020-08-19T10:56:32+02:00 Another round of `npm audit fix` (#1228) This should shut down the warnings on Github. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. Last NPM dependency audit happend in d576b2327e2bc117f912fe0a9d595e9ae62614e0 Co-authored-by: Alex Biehl <alex.biehl at target.com> - - - - - 7af6e2a8 by Ben Gamari at 2020-08-31T13:59:34-04:00 Merge pull request haskell/haddock#1226 from hsyl20/wip/hsyl20/fs_ord Remove Ord FastString instance - - - - - ffbc8702 by Alan Zimmerman at 2020-09-07T21:47:41+01:00 Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - a93f1268 by Alan Zimmerman at 2020-09-07T23:11:38+01:00 Merge pull request haskell/haddock#1232 from haskell/wip/T18639-remove-generated-pragma, Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - 1f605d50 by Ben Gamari at 2020-09-14T18:30:01-04:00 Bump GHC version to 9.1 - - - - - 6599df62 by Vladislav Zavialov at 2020-09-18T14:05:15+03:00 Bump base upper bound to 4.16 - - - - - a01b3c43 by Ben Gamari at 2020-09-22T15:41:48-04:00 Update hypsrc-test for QuickLook This appears to be a spurious change. - - - - - e9cc6cac by Vladislav Zavialov at 2020-09-26T21:00:12+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 30e3ca7c by Sylvain Henry at 2020-09-29T11:18:32-04:00 Update for parser (#1234) - - - - - b172f3e3 by Vladislav Zavialov at 2020-09-30T01:01:30+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 0b9c08d3 by Sylvain Henry at 2020-09-30T11:02:33+02:00 Adapt to GHC parser changes - - - - - b9540b7a by Sylvain Henry at 2020-10-12T09:13:38-04:00 Don't pass the HomeUnitId (#1239) - - - - - 34762e80 by HaskellMouse at 2020-10-13T12:58:04+03:00 Changed tests due to unification of `Nat` and `Natural` in the follwing merge request: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3583 - - - - - 256f86b6 by Vladislav Zavialov at 2020-10-15T10:48:03+03:00 Add whitespace in: map ($ v) - - - - - 4a3f711b by Alan Zimmerman at 2020-10-19T08:57:27+01:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled - - - - - 072cdd21 by Alan Zimmerman at 2020-10-21T14:48:28-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 9e09a445 by Alan Zimmerman at 2020-10-21T23:53:34-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 636d7de3 by Sylvain Henry at 2020-10-26T14:31:54-04:00 GHC.Driver.Types refactoring (#1242) - - - - - a597f000 by Ryan Scott at 2020-10-29T04:18:05-04:00 Adapt to the removal of Hs{Boxed,Constraint}Tuple See ghc/ghc!4097 and GHC#18723. - - - - - b96660fb by Ryan Scott at 2020-10-30T04:53:05-04:00 Adapt to HsConDecl{H98,GADT}Details split Needed for GHC#18844. - - - - - c287d82c by Ryan Scott at 2020-10-30T19:35:59-04:00 Adapt to HsOuterTyVarBndrs These changes accompany ghc/ghc!4107, which aims to be a fix for haskell/haddock#16762. - - - - - a34c31a1 by Ryan Scott at 2020-11-13T13:38:34-05:00 Adapt to splitPiTysInvisible being renamed to splitInvisPiTys This is a part of !4434, a fix for GHC#18939. - - - - - 66ea459d by Sylvain Henry at 2020-11-16T10:59:30+01:00 Fix after Plugins moved into HscEnv - - - - - 508556d8 by Ben Gamari at 2020-11-18T15:47:40-05:00 Merge pull request haskell/haddock#1253 from hsyl20/wip/hsyl20/plugins/hscenv Fix after Plugins moved into HscEnv - - - - - 620fec1a by Andreas Klebinger at 2020-11-24T20:51:59+01:00 Update for changes in GHC's Pretty - - - - - 01cc13ab by Richard Eisenberg at 2020-11-25T23:18:35-05:00 Avoid GHC#18932. - - - - - 8d29ba21 by Cale Gibbard at 2020-11-25T23:18:35-05:00 Add type arguments to PrefixCon - - - - - 414d5f87 by Sylvain Henry at 2020-11-30T17:06:04+01:00 DynFlags's unit fields moved to HscEnv - - - - - e356668c by Ben Gamari at 2020-11-30T11:11:37-05:00 Merge pull request haskell/haddock#1258 from hsyl20/wip/hsyl20/hscenv/unitstate Unit fields moved from DynFlags to HscEnv - - - - - 7cf552f1 by Ben Gamari at 2020-12-03T10:31:27-05:00 Merge pull request haskell/haddock#1257 from AndreasPK/wip/andreask/opt_dumps Update for changes in GHC's Pretty - - - - - fc0871c3 by Veronika Romashkina at 2020-12-08T16:35:33+01:00 Fix docs links from Darcs to GitHub in intro (#1262) - - - - - 7059e808 by Veronika Romashkina at 2020-12-08T16:36:16+01:00 Use gender neutral word in docs (#1260) - - - - - 1b16e5ee by Maximilian Tagher at 2020-12-08T16:40:03+01:00 Allow scrolling search results (#1235) Closes https://github.com/haskell/haddock/issues/1231 - - - - - 8a118c01 by dependabot[bot] at 2020-12-08T16:40:25+01:00 Bump bl from 1.2.2 to 1.2.3 in /haddock-api/resources/html (#1255) Bumps [bl](https://github.com/rvagg/bl) from 1.2.2 to 1.2.3. - [Release notes](https://github.com/rvagg/bl/releases) - [Commits](https://github.com/rvagg/bl/compare/v1.2.2...v1.2.3) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - c89ff587 by Xia Li-yao at 2020-12-08T16:42:17+01:00 Allow more characters in anchor following module reference (#1220) - - - - - 14af7d64 by Xia Li-yao at 2020-12-08T16:43:05+01:00 Add dangling changes from branches ghc-8.6 and ghc-8.8 (#1243) * Fix multiple typos and inconsistencies in doc/markup.rst Note: I noticed some overlap with haskell/haddock#1112 from @wygulmage and haskell/haddock#1081 from @parsonsmatt after creating these proposed changes - mea culpa for not looking at the open PRs sooner. * Fix haskell/haddock#1113 If no Signatures, no section of index.html * Change the formatting of missing link destinations The current formatting of the missing link destination does not really help user to understand the reasons of the missing link. To address this, I've changed the formatting in two ways: - the missing link symbol name is now fully qualified. This way you immediately know which haskell module cannot be linked. It is then easier to understand why this module does not have documentation (hidden module or broken documentation). - one line per missing link, that's more readable now that symbol name can be longer due to qualification. For example, before haddock was listing missing symbol such as: ``` could not find link destinations for: Word8 Word16 mapMaybe ``` Now it is listed as: ``` could not find link destinations for: - Data.Word.Word8 - Data.Word.Word16 - Data.Maybe.mapMaybe ``` * Add `--ignore-link-symbol` command line argument This argument can be used multiples time. A missing link to a symbol listed by `--ignore-link-symbol` won't trigger "missing link" warning. * Forbid spaces in anchors (#1148) * Improve error messages with context information (#1060) Co-authored-by: Matt Audesse <matt at mattaudesse.com> Co-authored-by: Mike Pilgrem <mpilgrem at users.noreply.github.com> Co-authored-by: Guillaume Bouchard <guillaume.bouchard at tweag.io> Co-authored-by: Pepe Iborra <pepeiborra at gmail.com> - - - - - 89e3af13 by tomjaguarpaw at 2020-12-08T18:00:04+01:00 Enable two warnings (#1245) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - c3320f8d by Willem Van Onsem at 2020-12-08T18:26:55+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 685df308 by Alex Biehl at 2020-12-08T20:06:26+01:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - be3ec3c0 by Alex Biehl at 2020-12-08T20:06:26+01:00 Import intercalate - - - - - 32c33912 by Matthías Páll Gissurarson at 2020-12-08T21:15:30+01:00 Adapt Haddock for QualifiedDo - - - - - 31696088 by Alex Biehl at 2020-12-08T22:06:02+01:00 Fix haddock-library tests - - - - - fbc0998a by Alex Biehl at 2020-12-08T23:08:23+01:00 Move to GitHub CI (#1266) * Initial version of ci.yml This is a straight copy from Dmitrii Kovanikov's blog post at https://kodimensional.dev/github-actions. Will adapt to haddock in successive commits. * Delete .travis.yml * Modify to only test on ghc-8.10.{1,2} * Use actions/setup-haskell at v1.1.4 * Relax QuickCheck bound on haddock-api * Remove stack matrix for now * Nail down to ghc-8.10 branch for now * Pin index state to 2020-12-08T20:13:44Z for now * Disable macOS and Windows tests for now for speed up - - - - - 5b946b9a by tomjaguarpaw at 2020-12-10T19:01:41+01:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - bc5a408f by dependabot[bot] at 2020-12-10T19:02:16+01:00 Bump ini from 1.3.5 to 1.3.7 in /haddock-api/resources/html (#1269) Bumps [ini](https://github.com/isaacs/ini) from 1.3.5 to 1.3.7. - [Release notes](https://github.com/isaacs/ini/releases) - [Commits](https://github.com/isaacs/ini/compare/v1.3.5...v1.3.7) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - d02995f1 by Andrew Martin at 2020-12-14T16:48:40-05:00 Update for boxed rep - - - - - a381aeff by Ben Gamari at 2020-12-15T15:13:30-05:00 Revert "Enable two warnings (#1245) (#1268)" As this does not build on GHC `master`. This reverts commit 7936692badfe38f23ae95b51fb7bd7c2ff7e9bce. - - - - - a63c0a9e by Ben Gamari at 2020-12-15T15:17:59-05:00 Revert "Update for boxed rep" This reverts commit 4ffb30d8b637ccebecc81ce610f0af451ac8088d. - - - - - 53bfbb29 by Ben Gamari at 2020-12-15T15:37:24-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - bae76a30 by Ben Gamari at 2020-12-16T02:44:42+00:00 Update output for nullary TyConApp optimisation (ghc/ghc!2952) - - - - - 4b733b57 by Krzysztof Gogolewski at 2020-12-16T20:03:14+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. - - - - - ee463bd3 by Ryan Scott at 2020-12-16T16:55:23-05:00 Adapt to HsCoreTy (formerly NewHsTypeX) becoming a type synonym Needed for !4417, the fix for GHC#15706 and GHC#18914. - - - - - ed0b02f8 by tomjaguarpaw at 2020-12-19T10:17:19+00:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - d80bf8f5 by Sylvain Henry at 2020-12-21T10:09:25+01:00 Fix after binder collect changes - - - - - bf4c9d32 by Adam Gundry at 2020-12-23T21:35:01+00:00 Adapt to changes to GlobalRdrElt and AvailInfo Needed for ghc/ghc!4467 - - - - - 37736c4c by John Ericson at 2020-12-28T12:27:02-05:00 Support a new ghc --make node type for parallel backpack upsweep - - - - - 717bdeac by Vladislav Zavialov at 2020-12-29T10:50:02+03:00 Inline and fix getGADTConTypeG The getGADTConTypeG used HsRecTy, which is at odds with GHC issue haskell/haddock#18782. I noticed that getGADTConTypeG was only used in the Hoogle backend. Interestingly, when handling H98 constructors, Hoogle converts RecCon to PrefixCon (see Haddock.Backends.Hoogle.ppCtor). So I changed getGADTConTypeG to handle RecConGADT in the same manner as PrefixConGADT, and after this simplification moved it into the 'where' clause of ppCtor, to the only place where it is used. The practical effect of this change is as follows. Consider this example: data TestH98 = T98 { bar::Int } data TestGADT where TG :: { foo :: Int } -> TestGADT Before this patch, haddock --hoogle used to produce: T98 :: Int -> TestH98 [TG] :: {foo :: Int} -> TestGADT Notice how the record syntax was discarded in T98 but not TG. With this patch, we always produce signatures without record syntax: T98 :: Int -> TestH98 [TG] :: Int -> TestGADT I suspect this might also be a bugfix, as currently Hoogle doesn't seem to render GADT record constructors properly. - - - - - cb1b8c56 by Andreas Abel at 2020-12-30T21:12:37+01:00 Build instructions: haddock-library and -api first! - - - - - b947f6ad by Ben Gamari at 2020-12-31T13:04:19-05:00 Merge pull request haskell/haddock#1281 from obsidiansystems/wip/backpack-j Changes to support -j with backpack - - - - - 120e1cfd by Hécate Moonlight at 2021-01-04T19:54:58+01:00 Merge pull request haskell/haddock#1282 from andreasabel/master Build instructions: haddock-library and -api first! - - - - - fd45e41a by Ben Gamari at 2021-01-05T16:14:31-05:00 Merge remote-tracking branch 'origin/ghc-8.10' into ghc-9.0 - - - - - b471bdec by Ben Gamari at 2021-01-05T16:23:02-05:00 Merge commit '1e56f63c3197e7ca1c1e506e083c2bad25d08793' into ghc-9.0 - - - - - 81cdbc41 by Alex Biehl at 2021-01-09T12:14:41+01:00 Prepare Haddock for being a GHC Plugin - - - - - b646d952 by Alex Biehl at 2021-01-09T12:14:41+01:00 Make Haddock a GHC Plugin - - - - - cc044674 by Alex Biehl at 2021-01-09T12:14:41+01:00 Add -j[n] CLI param to Haddock executable It translates to `--ghcopt=-j[n]` - - - - - 84a04073 by Alex Biehl at 2021-01-09T12:14:41+01:00 Abstract Monad for interface creation I found that when running as a plugin the lookupName function (which runs in Ghc monad) does not work correctly from the typeCheckResultAction hook. Instead, we abstracted the monad used when creating interfaces, so that access to GHC session specific parts is explicit and so that the TcM can provide their (correct) implementation of lookupName. - - - - - 5be2c4f7 by Alex Biehl at 2021-01-09T12:14:41+01:00 Accept tests - - - - - 8cefee9d by Alex Biehl at 2021-01-09T16:10:47+01:00 Add missing dependency for mtl - - - - - 3681f919 by Ben Gamari at 2021-01-13T18:39:25-05:00 Merge remote-tracking branch 'origin/ghc-9.0' into ghc-head - - - - - 33c6b152 by Hécate Moonlight at 2021-01-14T16:04:20+01:00 Merge pull request haskell/haddock#1273 from hsyl20/wip/hsyl20/arrows Fix after binder collect changes - - - - - 70d13e8e by Joachim Breitner at 2021-01-22T19:03:45+01:00 Make haddock more robust to changes to the `Language` data type With the introduction of GHC2021, the `Languages` data type in GHC will grow. In preparation of that (and to avoid changing haddock with each new language), this change makes the code handle extensions to that data type gracefully. (cherry picked from commit c341dd7c9c3fc5ebc83a2d577c5a726f3eb152a5) - - - - - 7d6dd57a by John Ericson at 2021-01-22T22:02:02+00:00 Add `NoGhcTc` instance now that it's not closed - - - - - e5fdaf0a by Alan Zimmerman at 2021-01-23T22:57:44+00:00 Merge pull request haskell/haddock#1293 from obsidiansystems/wip/fix-18936 Add `NoGhcTc` instance now that it's not closed - - - - - 989a1e05 by Oleg Grenrus at 2021-01-24T16:11:46+03:00 Add import list to Data.List - - - - - 368e144a by Ben Gamari at 2021-01-28T22:15:48+01:00 Adapt to "Make PatSyn immutable" - - - - - abe66c21 by Alfredo Di Napoli at 2021-02-01T08:05:35+01:00 Rename pprLogErrMsg to new name - - - - - e600e75c by Hécate Moonlight at 2021-02-05T14:53:00+01:00 Move CI to ghc-9.0 - - - - - dd492961 by Vladislav Zavialov at 2021-02-05T14:53:00+01:00 Update cabal.project and README build instructions - - - - - 31bd292a by Hécate Moonlight at 2021-02-05T15:03:56+01:00 Merge pull request haskell/haddock#1296 from Kleidukos/ghc-9.0 Merge the late additions to ghc-8.10 into ghc-9.0 - - - - - 6388989e by Vladislav Zavialov at 2021-02-05T17:41:57+03:00 Cleanup: fix build warnings - - - - - f99407ef by Daniel Rogozin at 2021-02-05T18:11:48+03:00 type level characters support for haddock (required for haskell/haddock#11342) - - - - - d8c6b26f by Hécate Moonlight at 2021-02-05T17:44:50+01:00 Add a CONTRIBUTING.md file - - - - - 6a01ad98 by Hécate Moonlight at 2021-02-05T17:58:16+01:00 Merge pull request haskell/haddock#1312 from Kleidukos/proper-branch-etiquette Add a CONTRIBUTING.md file - - - - - 955eecc4 by Vladislav Zavialov at 2021-02-05T20:29:00+03:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into ghc-head - - - - - 47b3d6ab by Hécate Moonlight at 2021-02-05T19:09:38+01:00 Amend the CONTRIBUTING.md file - - - - - 23de6137 by Hécate Moonlight at 2021-02-05T19:16:49+01:00 Merge pull request haskell/haddock#1313 from Kleidukos/amend-contributing Amend the CONTRIBUTING.md file - - - - - 69026b59 by Krzysztof Gogolewski at 2021-02-05T23:05:56+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. (cherry picked from commit b4b4d896d2d68d6c48e7db7bfe95c185ca0709cb) - - - - - ea026b78 by Oleg Grenrus at 2021-02-06T17:14:45+01:00 Add import list to Data.List - - - - - 5204326f by Hécate Moonlight at 2021-02-06T17:15:44+01:00 Merge pull request haskell/haddock#1316 from Kleidukos/explicit-imports-to-data-list Add import list to Data.List - - - - - 1f4d2136 by Ben Gamari at 2021-02-06T11:53:31-05:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ghc-head-merge - - - - - 13f0d09a by Ben Gamari at 2021-02-06T11:53:45-05:00 Fix partial record selector warning - - - - - 5c115f7e by Ben Gamari at 2021-02-06T11:55:52-05:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into wip/ghc-head-merge - - - - - b6fd8b75 by Ben Gamari at 2021-02-06T12:01:31-05:00 Merge commit '41964cb2fd54b5a10f8c0f28147015b7d5ad2c02' into wip/ghc-head-merge - - - - - a967194c by Ben Gamari at 2021-02-06T18:30:35-05:00 Merge branch 'wip/ghc-head-merge' into ghc-head - - - - - 1f4c3a91 by MorrowM at 2021-02-07T01:52:33+02:00 Fix search div not scrolling - - - - - 684b1287 by Iñaki García Etxebarria at 2021-02-07T16:13:04+01:00 Add support for labeled module references Support a markdown-style way of annotating module references. For instance -- | [label]("Module.Name#anchor") will create a link that points to the same place as the module reference "Module.Name#anchor" but the text displayed on the link will be "label". - - - - - bdb55a5d by Hécate Moonlight at 2021-02-07T16:18:10+01:00 Merge pull request haskell/haddock#1319 from alexbiehl/alex/compat Backward compat: Add support for labeled module references - - - - - 6ca70991 by Hécate Moonlight at 2021-02-07T16:21:29+01:00 Merge pull request haskell/haddock#1314 from tweag/show-linear-backport Backport haskell/haddock#1238 (linear types) to ghc-9.0 - - - - - d9d73298 by Alex Biehl at 2021-02-07T17:46:25+01:00 Remove dubious parseModLink Instead construct the ModLink value directly when parsing. - - - - - 33b4d020 by Hécate Moonlight at 2021-02-07T17:52:05+01:00 Merge pull request haskell/haddock#1320 from haskell/alex/fix Remove dubious parseModLink - - - - - 54211316 by Hécate Moonlight at 2021-02-07T18:12:07+01:00 Merge pull request haskell/haddock#1318 from MorrowM/ghc-9.0 Fix search div not scrolling - - - - - 19db679e by alexbiehl-gc at 2021-02-07T18:14:46+01:00 Merge pull request haskell/haddock#1317 from bgamari/wip/ghc-head-merge Merge ghc-8.10 into ghc-head - - - - - 6bc1e9e4 by Willem Van Onsem at 2021-02-07T18:25:30+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - c8537cf8 by alexbiehl-gc at 2021-02-07T18:30:40+01:00 Merge pull request haskell/haddock#1322 from haskell/alex/forward-port simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 2d47ae4e by alexbiehl-gc at 2021-02-07T18:39:59+01:00 Merge branch 'ghc-head' into ghc-9.0 - - - - - 849e4733 by Hécate Moonlight at 2021-02-07T18:43:19+01:00 Merge pull request haskell/haddock#1321 from Kleidukos/ghc-9.0 Merge ghc-9.0 into ghc-head - - - - - ee6095d7 by Sylvain Henry at 2021-02-08T11:36:38+01:00 Update for Logger - - - - - 4ad688c9 by Alex Biehl at 2021-02-08T18:11:24+01:00 Merge pull request haskell/haddock#1310 from hsyl20/wip/hsyl20/logger2 Logger refactoring - - - - - 922a9e0e by Ben Gamari at 2021-02-08T12:54:33-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - 991649d2 by Sylvain Henry at 2021-02-09T10:55:17+01:00 Fix to build with HEAD - - - - - a8348dc2 by Hécate Moonlight at 2021-02-09T10:58:51+01:00 Merge pull request haskell/haddock#1327 from hsyl20/wip/hsyl20/logger2 Fix to build with HEAD - - - - - 0abdbca6 by Fendor at 2021-02-09T20:06:15+01:00 Add UnitId to Target record - - - - - d5790a0e by Alex Biehl at 2021-02-11T10:32:32+01:00 Stable sort for (data/newtype) instances - - - - - 8e6036f5 by Alex Biehl at 2021-02-11T10:32:32+01:00 Also make TyLit deterministic - - - - - f76d2945 by Hécate Moonlight at 2021-02-11T11:00:31+01:00 Merge pull request haskell/haddock#1329 from hsyl20/hsyl20/stabe_iface Stable sort for instances - - - - - 5e0469ea by Oleg Grenrus at 2021-02-14T15:28:15+02:00 Add import list to Data.List in Haddock.Interface.Create - - - - - fa57cd24 by Hécate Moonlight at 2021-02-14T17:19:27+01:00 Merge pull request haskell/haddock#1331 from phadej/more-explicit-data-list-imports Add import list to Data.List in Haddock.Interface.Create - - - - - f0cd629c by Hécate Moonlight at 2021-02-21T00:22:01+01:00 Merge pull request haskell/haddock#1311 from fendor/wip/add-targetUnitId-to-target Add UnitId to Target record - - - - - 674ef723 by Joachim Breitner at 2021-02-22T10:39:18+01:00 html-test: Always set language from ghc-9.2 on, the “default” langauge of GHC is expected to change more wildly. To prepare for that (and unblock https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4853), this sets the language for all the test files to `Haskell2010`. This should insolate this test suite against changes to the default. Cherry-picked from https://github.com/haskell/haddock/pull/1341 - - - - - f072d623 by Hécate Moonlight at 2021-02-22T10:56:51+01:00 Merge pull request haskell/haddock#1342 from nomeata/joachim/haskell2010-in-tests-ghc-head html-test: Always set language - - - - - caebbfca by Hécate Moonlight at 2021-02-22T11:53:07+01:00 Clean-up of Interface and Interface.Create's imports and pragmata - - - - - f6caa89d by Hécate Moonlight at 2021-02-22T11:54:58+01:00 Merge pull request haskell/haddock#1345 from Kleidukos/head/fix-interface-imports [ghc-head] Clean-up of Interface and Interface.Create's imports and pragmata - - - - - 7395c9cb by Hécate Moonlight at 2021-02-22T18:44:57+01:00 Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 6e9fb5d5 by Hécate Moonlight at 2021-02-22T18:45:28+01:00 Merge pull request haskell/haddock#1348 from Kleidukos/head/explicit-imports-interface Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 9198b118 by Alan Zimmerman at 2021-02-22T20:04:24+00:00 Context becomes a Maybe in the GHC AST This prevents noLoc's appearing in the ParsedSource. Match the change in GHC. - - - - - 0af20f64 by Hécate Moonlight at 2021-02-23T12:36:12+01:00 Fix the call-site of guessTarget in Interface.hs Explicit the imports from GHC.HsToCore.Docs - - - - - b7886885 by Hécate Moonlight at 2021-02-23T12:37:54+01:00 Merge pull request haskell/haddock#1349 from Kleidukos/fix-interface-guesstarget-call Fix the call-site of guessTarget in Interface.hs - - - - - 9cf041ba by Sylvain Henry at 2021-02-24T11:08:20+01:00 Fix haddockHypsrcTest output in ghc-head - - - - - b194182a by Hécate Moonlight at 2021-02-24T11:12:36+01:00 Merge pull request haskell/haddock#1351 from hsyl20/wip/hsyl20/fix-head Fix haddockHypsrcTest output in ghc-head - - - - - 3ce8b375 by Shayne Fletcher at 2021-03-06T09:55:03-05:00 Add ITproj to parser - - - - - d2abf762 by Ben Gamari at 2021-03-06T19:26:49-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a0f6047d by Andrew Martin at 2021-03-07T11:25:23-05:00 Update for boxed rep - - - - - 6f63c99e by Ben Gamari at 2021-03-10T13:20:21-05:00 Update for "FastString: Use FastMutInt instead of IORef Int" - - - - - e13f01df by Luke Lau at 2021-03-10T15:38:40-05:00 Implement template-haskell's putDoc This catches up to GHC using the new extractTHDocs function, which returns documentation added via the putDoc function (provided it was compiled with Opt_Haddock). Since it's already a map from names -> docs, there's no need to do traversal etc. It also matches the change from the argument map being made an IntMap rather than a Map Int - - - - - 89263d94 by Alan Zimmerman at 2021-03-15T17:15:26+00:00 Match changes in GHC AST for in-tree API Annotations As landed via https://gitlab.haskell.org/ghc/ghc/-/merge_requests/2418 - - - - - 28db1934 by Alan Zimmerman at 2021-03-15T20:40:09+00:00 Change some type family test results. It is not clear to me whether the original was printing incorrectly (since we did not have the TopLevel flag before now), or if this behaviour is expected. For the time being I am assuming the former. - - - - - 7c11c989 by Sylvain Henry at 2021-03-22T10:05:19+01:00 Fix after NameCache changes - - - - - addbde15 by Sylvain Henry at 2021-03-22T10:05:19+01:00 NameCache doesn't store a UniqSupply anymore - - - - - 15ec6cec by Ben Gamari at 2021-03-22T17:53:44-04:00 Bump GHC version to 9.2 - - - - - dbd6aa63 by Hécate Moonlight at 2021-03-24T14:28:36+01:00 Merge pull request haskell/haddock#1365 from hsyl20/wip/hsyl20/iface1 NameCache refactoring - - - - - 2d32da7e by Oleg Grenrus at 2021-03-27T01:12:00+02:00 Specialization of Data.List - - - - - 32b84fa6 by Fendor at 2021-03-27T10:50:17+01:00 Add UnitId to Target record This way we always know to which home-unit a given target belongs to. So far, there only exists a single home-unit at a time, but it enables having multiple home-units at the same time. - - - - - 54bf9f0e by Hécate Moonlight at 2021-03-28T14:08:35+02:00 Merge pull request haskell/haddock#1368 from fendor/target-unit-id-revert Add UnitId to Target record - - - - - 7dea168a by Alan Zimmerman at 2021-03-29T08:45:52+01:00 EPA : Rename ApiAnn to EpAnn - - - - - 72967f65 by Alfredo Di Napoli at 2021-03-29T09:47:01+02:00 pprError changed name in GHC - - - - - 4bc61035 by Alan Zimmerman at 2021-03-29T16:16:27-04:00 EPA : Rename ApiAnn to EpAnn - - - - - 108d031d by Ben Gamari at 2021-03-29T18:49:36-04:00 Merge commit '36418c4f70d7d2b179a77925b3ad5caedb08c9b5' into HEAD - - - - - 1444f700 by Ben Gamari at 2021-03-31T09:18:39-04:00 Merge pull request haskell/haddock#1370 from adinapoli/wip/adinapoli-diag-reason-severity Rename pprError to mkParserErr - - - - - d3087b79 by Ben Gamari at 2021-03-31T11:34:17-04:00 Merge commit 'd8d8024ad6796549a8d3b5512dabf3288d14e30f' into ghc-head - - - - - 170b79e9 by Ben Gamari at 2021-03-31T12:24:56-04:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - db0d6bae by Ben Gamari at 2021-04-10T09:34:35-04:00 Bump GHC version to 9.3 - - - - - a9f2c421 by Alan Zimmerman at 2021-04-19T18:26:46-04:00 Update for EPA changes in GHC (cherry picked from commit cafb48118f7c111020663776845897e225607b41) - - - - - 1ee4b7c7 by Sylvain Henry at 2021-05-11T10:00:06+02:00 Removal of HsVersions.h (#1388) * Update for EPA changes in GHC * Account for HsVersions.h removal Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - 79e819e9 by Hécate Moonlight at 2021-05-11T10:14:47+02:00 Revert "Removal of HsVersions.h (#1388)" This reverts commit 72118896464f94d81f10c52f5d9261efcacc57a6. - - - - - 3dbd3f8b by Alan Zimmerman at 2021-05-11T10:15:17+02:00 Update for EPA changes in GHC - - - - - 2ce80c17 by Sylvain Henry at 2021-05-11T10:15:19+02:00 Account for HsVersions.h removal - - - - - 00e4c918 by Christiaan Baaij at 2021-05-13T08:21:56+02:00 Add Haddock support for the OPAQUE pragma (#1380) - - - - - 8f9049b2 by Hécate Moonlight at 2021-05-13T08:40:22+02:00 fixup! Use GHC 9.2 in CI runner - - - - - 27ddec38 by Alan Zimmerman at 2021-05-13T22:51:20+01:00 EPA: match changes from GHC T19834 - - - - - f8a1d714 by Felix Yan at 2021-05-14T17:10:04+02:00 Allow hspec 2.8 (#1389) All tests are passing. - - - - - df44453b by Divam Narula at 2021-05-20T15:42:42+02:00 Update ref, the variables got renamed. (#1391) This is due to ghc/ghc!5555 which caused a change in ModDetails in case of NoBackend. Now the initModDetails is used to recreate the ModDetails from interface and in-memory ModDetails is not used. - - - - - e46bfc87 by Alan Zimmerman at 2021-05-20T19:05:09+01:00 Remove Maybe from HsQualTy Match changes in GHC for haskell/haddock#19845 - - - - - 79bd7b62 by Shayne Fletcher at 2021-05-22T08:20:39+10:00 FieldOcc: rename extFieldOcc to foExt - - - - - 6ed68c74 by Ben Gamari at 2021-05-21T22:29:30-04:00 Merge commit '3b6a8774bdb543dad59b2618458b07feab8a55e9' into ghc-head - - - - - f9a02d34 by Alfredo Di Napoli at 2021-05-24T13:53:00+02:00 New Parser diagnostics interface - - - - - 392807d0 by Ben Gamari at 2021-05-24T09:57:40-04:00 Merge pull request haskell/haddock#1394 from adinapoli/wip/adinapoli-align-ps-messages Align Haddock to use the new Parser diagnostics interface - - - - - 33023cd8 by Ben Gamari at 2021-05-24T11:19:16-04:00 Revert "Add Haddock support for the OPAQUE pragma (#1380)" This reverts commit a1337c599ef7720b0482a25c55f11794112496dc. The GHC patch associated with this change is not yet ready to be merged. - - - - - 8c005af7 by Ben Simms at 2021-05-28T07:56:20+02:00 CI configuration for ghc-head (#1395) - - - - - 1e947612 by Hécate Moonlight at 2021-05-28T12:27:35+02:00 Use GHC 9.2 in CI runner (#1378) - - - - - e6fa10ab by CGenie at 2021-05-31T09:02:13+02:00 Add page about common errors (#1396) * Update index.rst Common errors page * Create common-errors.rst * Update common-errors.rst * Use GHC 9.2 in CI runner (#1378) * [haddock-api] remove .hspec-failures Co-authored-by: Hécate Moonlight <Kleidukos at users.noreply.github.com> - - - - - abc72a8d by Sylvain Henry at 2021-06-01T10:02:06+02:00 Adapt Haddock to Logger and Parser changes (#1399) - - - - - 91373656 by Zubin Duggal at 2021-06-01T20:45:10+02:00 Update haddockHypSrc tests since we now compute slighly more type info (#1397) - - - - - ed712822 by Marcin Szamotulski at 2021-06-02T08:54:33+02:00 Added myself to contributors - - - - - 49fdbcb7 by Marcin Szamotulski at 2021-06-02T08:57:24+02:00 Document multi component support - - - - - 9ddc8d7d by Hécate Moonlight at 2021-06-02T09:35:55+02:00 Merge pull request haskell/haddock#1379 from coot/coot/document-multi-component-support Document multi component support - - - - - 585b5c5e by Ben Simms at 2021-06-02T19:46:54+02:00 Update CONTRIBUTING.md (#1402) - - - - - 1df4a605 by Ben Simms at 2021-06-02T19:47:14+02:00 Update CONTRIBUTING.md (#1403) - - - - - 58ea43d2 by sheaf at 2021-06-02T22:09:06+02:00 Update Haddock Bug873 to account for renaming - - - - - c5d0ab23 by Vladislav Zavialov at 2021-06-10T13:35:42+03:00 HsToken in FunTy, RecConGADT - - - - - 1ae2f40c by Hécate Moonlight at 2021-06-11T11:19:09+02:00 Update the CI badges - - - - - 6fdc4de2 by Sylvain Henry at 2021-06-28T19:21:17+02:00 Fix mkParserOpts (#1411) - - - - - 18201670 by Alfredo Di Napoli at 2021-07-05T07:55:12+02:00 Rename getErrorMessages Lexer import This commit renames the Lexer import in `Hyperlinker.Parser` from `getErrorMessages` to `getPsErrorMessages` to eliminate the ambiguity with the `getErrorMessages` function defined in `GHC.Types.Error`. - - - - - 23173ca3 by Ben Gamari at 2021-07-07T11:31:44-04:00 Merge pull request haskell/haddock#1413 from adinapoli/wip/adinapoli-issue-19920 Rename getErrorMessages Lexer import - - - - - b3dc4ed8 by Alan Zimmerman at 2021-07-28T22:30:59+01:00 EPA: match changes from GHC T19834 (cherry picked from commit 2fec1b44e0ee7e263286709aa528b4ecb99ac6c2) - - - - - 5f177278 by Ben Gamari at 2021-08-06T01:17:37-04:00 Merge commit '2a966c8ca37' into HEAD - - - - - cdd81d08 by Marcin Szamotulski at 2021-08-08T17:19:06+02:00 coot/multiple packages (ghc-9.2) (#1418) - - - - - be0d71f1 by Marcin Szamotulski at 2021-08-16T08:46:03+02:00 coot/multiple package (ghc-head) (#1419) * FromJSON class Aeson style FromJSON class with Parsec based json parser. * doc-index.json file for multiple packages When creating haddock summary page for multiple packages render doc-index.json file using contents of all found 'doc-index.json' files. * Render doc-index.json When rendering html, render doc-index.json file independently of maybe_index_url option. doc-index.json file is useful now even if maybe_index_url is not `Nothing`. * base url option New `Flag_BaseURL` which configures from where static files are loaded (--base-url). If given and not equal "." static files are not coppied, as this indicates that they are not read from the the directory where we'd copy them. The default value is ".". - - - - - 3b09dbdf by Hécate Moonlight at 2021-10-07T23:26:03+02:00 Update GHC 9.2 to latest pre-release in CI - - - - - 7ac55417 by Zubin Duggal at 2021-10-11T12:10:19+02:00 Enable Haddock tests in GHC windows CI (#1428) * testsuite: strip windows line endings for haddock * hyperlinker: Work around double escaping (#19236) * deterministic SCC - - - - - 1cb81f25 by Andrew Lelechenko at 2021-10-12T15:23:19+02:00 haddock-library does not depend on bytestring or transformers (#1426) - - - - - a890b9aa by sheaf at 2021-10-15T22:19:42+02:00 update haddockHypsrcTest for GHC MR !6705 (#1430) - - - - - 42a55c6c by Sylvain Henry at 2021-10-15T22:20:10+02:00 Fix after PkgQual refactoring (#1429) - - - - - 91659238 by Alan Zimmerman at 2021-10-28T18:57:10+01:00 Update for changes in GHC for branch wip/az/no-srcspan-anno-instances - - - - - acf23e60 by Vladislav Zavialov at 2021-11-05T02:09:47+03:00 Do not use forall as an identifier See GHC ticket haskell/haddock#20609 - - - - - c565db0e by Krzysztof Gogolewski at 2021-11-27T02:42:35+01:00 Update after NoExtCon -> DataConCantHappen rename - - - - - b5f55590 by Artem Pelenitsyn at 2021-11-27T11:14:17+01:00 fix CI for 9.2 (#1436) - - - - - 25cd621e by Matthew Pickering at 2021-12-02T11:46:54+00:00 Update html-test for Data.List revert - - - - - 1d5ff85f by malteneuss at 2021-12-15T07:56:55+01:00 Add hint about inline link issue (#1444) - - - - - 791fde81 by Sylvain Henry at 2021-12-16T09:29:51+01:00 Bump ghc-head (#1445) * Update after NoExtCon -> DataConCantHappen rename * Update html-test for Data.List revert * Fix for new Plugins datatype Co-authored-by: Krzysztof Gogolewski <krzysztof.gogolewski at tweag.io> Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 44236317 by Sylvain Henry at 2021-12-17T09:39:00+01:00 Fix for new Plugins datatype - - - - - 80ada0fa by Hécate Moonlight at 2021-12-17T17:28:48+01:00 Remove ghc-head workflow (#1446) Contributions of GHC glue code are now done on the GHC gitlab, not in the GitHub repo anymore. - - - - - 49e171cd by Matthew Pickering at 2021-12-28T09:47:09+00:00 Remove use of ExtendedModSummary - - - - - 0e91b5ea by askeblad at 2022-01-04T09:18:35+01:00 update URLs - - - - - 9f13c212 by Hécate Moonlight at 2022-02-25T10:19:46+01:00 Fix solver for GHC 9.2 - - - - - 386751a1 by Meng Weng Wong at 2022-02-25T19:19:11+01:00 IDoc link has bitrotted; replaced with web.archive.org cache. (#1454) - - - - - d877cbe6 by Hécate Moonlight at 2022-02-25T19:21:58+01:00 Fix haddock user guide (#1456) - - - - - cc47f036 by Andrew Lelechenko at 2022-03-04T17:29:36+01:00 Allow text-2.0 in haddock-library (#1459) - - - - - 7b3685a3 by malteneuss at 2022-03-07T19:27:24+01:00 Add multi-line style hint to style section (#1460) - - - - - c51088b8 by John Ericson at 2022-03-11T16:46:26+01:00 Fix CollectPass instance to match TTG refactor Companion to GHC !7614 (https://gitlab.haskell.org/ghc/ghc/-/merge_requests/7614) - - - - - b882195b by Vladislav Zavialov at 2022-03-14T20:32:30+01:00 Link to (~) - - - - - 877349b8 by Christiaan Baaij at 2022-03-16T09:20:43+01:00 Add Haddock support for the OPAQUE pragma - - - - - 0ea22721 by askeblad at 2022-03-16T09:44:27+01:00 typos (#1464) - - - - - a6d13da1 by Matthew Pickering at 2022-03-22T13:41:17+00:00 Minimum changes needed for compilation with hi-haddock With hi-haddock, of course there is a much large refactoring of haddock which could be achieved but that is left for a future patch which can implemented at any time independently of GHC. - - - - - e7ac9129 by Matthew Pickering at 2022-03-22T21:17:50+00:00 Update test output - - - - - 6d916214 by Matthew Pickering at 2022-03-24T15:06:26+00:00 Merge branch 'wip/opaque_pragma' into 'ghc-head' Add Haddock support for the OPAQUE pragma See merge request ghc/haddock!2 - - - - - 42208183 by Steve Hart at 2022-03-25T20:43:50+01:00 Fix CI (#1467) * CI: Reinstall GHC with docs CI tests were failing because the GHC preinstalled to the CI environment does not include documentation, which is required for running the Haddock tests. This commit causes the CI workflow to reinstall GHC with docs so that tests can succeed. - - - - - 9676fd79 by Steve Hart at 2022-03-25T21:33:34+01:00 Make links in Synopsis functional again (#1458) Commit e41c1cbe9f0476997eac7b4a3f17cbc6b2262faf added a call to e.preventDefault() when handling click events that reach a toggle element. This prevents the browser from following hyperlinks within the Synopsis section when they are clicked by a user. This commit restores functioning hyperlinks within the Synopsis section by removing the call to e.preventDefault(), as it does not appear to be necessary, and removing it increases the flexibility of the details-helper code. - - - - - d1edd637 by sheaf at 2022-04-01T12:02:02+02:00 Keep track of promotion ticks in HsOpTy Keeping track of promotion ticks in HsOpTy allows us to properly pretty-print promoted constructors such as lists. - - - - - 9dcb2dfc by Jakob Brünker at 2022-04-01T15:46:22+00:00 Add support for \cases See merge request ghc/ghc!7873 - - - - - b0412ee5 by askeblad at 2022-04-06T17:47:57+02:00 spelling errors (#1471) - - - - - 6b18829b by Vladislav Zavialov at 2022-04-06T18:53:58+02:00 Rename [] to List - - - - - 2d046691 by Vladislav Zavialov at 2022-04-07T20:25:54+03:00 HsToken ConDeclGADT con_dcolon - - - - - 90b43da4 by Steve Hart at 2022-04-12T13:29:46+02:00 Parse Markdown links at beginning of line within a paragraph (#1470) * Catch Markdown links at beginning of line within paragraph Per Issue haskell/haddock#774, Markdown links were being parsed as ordinary text when they occurred at the beginning of a line other than the first line of the paragraph. This occurred because the parser was not interpreting a left square bracket as a special character that could delimit special markup. A space character was considered a special character, so, if a space occurred at the beginning of the new line, then the parser would interpret the space by itself and then continue parsing, thereby catching the Markdown link. '\n' was not treated as a special character, so the parser did not catch a Markdown link that may have followed. Note that this will allow for Markdown links that are not surrounded by spaces. For example, the following text includes a Markdown link that will be parsed: Hello, world[label](url) This is consistent with how the parser handles other types of markup. * Remove obsolete documentation hint Commit 6b9aeafddf20efc65d3725c16e3fc43a20aac343 should eliminate the need for the workaround suggested in the documentation. - - - - - 5b08312d by Hécate Moonlight at 2022-04-12T13:36:38+02:00 Force ghc-9.2 in the cabal.project - - - - - 0d0ea349 by dependabot[bot] at 2022-04-12T13:57:41+02:00 Bump path-parse from 1.0.5 to 1.0.7 in /haddock-api/resources/html (#1469) Bumps [path-parse](https://github.com/jbgutierrez/path-parse) from 1.0.5 to 1.0.7. - [Release notes](https://github.com/jbgutierrez/path-parse/releases) - [Commits](https://github.com/jbgutierrez/path-parse/commits/v1.0.7) --- updated-dependencies: - dependency-name: path-parse dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 2b9fc65e by dependabot[bot] at 2022-04-12T13:57:54+02:00 Bump copy-props from 2.0.4 to 2.0.5 in /haddock-api/resources/html (#1468) Bumps [copy-props](https://github.com/gulpjs/copy-props) from 2.0.4 to 2.0.5. - [Release notes](https://github.com/gulpjs/copy-props/releases) - [Changelog](https://github.com/gulpjs/copy-props/blob/master/CHANGELOG.md) - [Commits](https://github.com/gulpjs/copy-props/compare/2.0.4...2.0.5) --- updated-dependencies: - dependency-name: copy-props dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - ea98a6fb by Ben Gamari at 2022-04-23T22:54:37-04:00 Update for GHC 9.4 - - - - - 9e11864f by Ben Gamari at 2022-04-25T16:24:31-04:00 Merge remote-tracking branch 'upstream/ghc-9.2' into ghc-head - - - - - f83cc506 by Ben Gamari at 2022-04-25T17:00:25-04:00 Bump ghc version to 9.5 - - - - - e01c2e7d by Ben Gamari at 2022-04-28T16:19:04-04:00 Revert "Bump ghc-head (#1445)" This reverts commit b29a78ef6926101338f62e84f456dac8659dc9d2. This should not have been merged. - - - - - a2b5ee8c by Ben Gamari at 2022-04-28T16:19:24-04:00 Merge commit '2627a86c' into ghc-head - - - - - 0c6fe4f9 by Ben Gamari at 2022-04-29T10:05:54-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-9.4 - - - - - b6e5cb0a by Ben Gamari at 2022-04-29T11:46:06-04:00 Revert "HsToken ConDeclGADT con_dcolon" This reverts commit 24208496649a02d5f87373052c430ea4a97842c5. - - - - - 15a62888 by Ben Gamari at 2022-04-29T15:12:55-04:00 Bump base upper bound - - - - - 165b9031 by Ben Gamari at 2022-04-29T23:58:38-04:00 Update test output - - - - - e0c3e5da by Phil de Joux at 2022-05-02T14:46:38+02:00 Add hlint action .hlint.yaml with ignores & CPP. (#1475) - - - - - ead1158d by Raphael Das Gupta at 2022-05-02T14:46:48+02:00 fix grammar in docs: "can the" → "can be" (#1477) - - - - - cff97944 by Ben Gamari at 2022-05-02T18:38:56-04:00 Allow base-4.17 - - - - - e4ecb201 by Phil de Joux at 2022-05-03T13:14:55+02:00 Remove unused imports that GHC warned about. (#1480) - - - - - 222890b1 by Phil de Joux at 2022-05-03T13:15:46+02:00 Follow hlint suggestion to remove redundant bang. (#1479) - - - - - 058b671f by Phil de Joux at 2022-05-03T13:34:04+02:00 Follow hlint, remove language pragmas in libs. (#1478) - - - - - 0a645049 by Ben Simms at 2022-05-03T14:19:24+02:00 Keep track of ordered list indexes and render them (#1407) * Keep track of ordered list indexes and render them * Rename some identifiers to clarify - - - - - f0433304 by Norman Ramsey at 2022-05-04T15:13:34-04:00 update for changes in GHC API - - - - - 3740cf71 by Emily Martins at 2022-05-06T18:23:48+02:00 Add link to the readthedocs in cabal description to show on hackage. (cherry picked from commit 52e2d40d47295c02d3181aac0c53028e730f1e3b) - - - - - 5d754f1e by Hécate Moonlight at 2022-05-06T18:44:57+02:00 remove Bug873 - - - - - 968fc267 by Hécate Moonlight at 2022-05-06T18:48:28+02:00 Ignore "Use second" HLint suggestion. It increases laziness. - - - - - 02d14e97 by Jade Lovelace at 2022-05-07T17:42:08+02:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` - - - - - b22b87ed by Artem Pelenitsyn at 2022-05-08T16:19:47+02:00 fix parsing trailing quotes in backticked identifiers (#1408) (#1483) - - - - - 80ae107b by Alex Biehl at 2022-05-08T16:37:16+02:00 Fix "Defined by not used" error (cherry picked from commit 6e02a620a26c3a44f98675dd1b93b08070c36c0a) - - - - - 4c838e84 by Hécate Moonlight at 2022-05-08T16:37:16+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - fc9827b4 by Hécate Moonlight at 2022-05-08T16:40:40+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - b153b555 by Xia Li-yao at 2022-05-20T17:52:42+02:00 Hide synopsis from search when hidden (#1486) Fix haskell/haddock#1451 - - - - - f3e38b85 by Marcin Szamotulski at 2022-05-21T23:32:31+02:00 Allow to hide interfaces when rendering multiple components (#1487) This is useful when one wishes to `--gen-contents` when rendering multiple components, but one does not want to render all modules. This is in particular useful when adding base package. - - - - - f942863b by Marcin Szamotulski at 2022-05-24T08:29:59+02:00 Check if doc-index.json exists before reading it (#1488) - - - - - 31e92982 by Marcin Szamotulski at 2022-05-25T16:22:13+02:00 Version bump 2.26.1 (#1489) * Version bump 2.26.1 We extended format accepted by `--read-interface` option, which requires updating the minor version. * Update documentation of --read-interface option - - - - - 7cc873e0 by sheaf at 2022-05-25T16:42:31+02:00 Updated HaddockHypsrcTest output for record update changes (MR !7981) - - - - - cd196942 by Marcin Szamotulski at 2022-05-25T20:28:47+02:00 Use visibility to decide which interfaces are included in quickjump (#1490) This is also consistent with how html index is build. See haskell/cabal#7669 for rationale behind this decision. - - - - - 00c713c5 by Hécate Moonlight at 2022-05-26T17:09:15+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 2f3039f1 by Hécate Moonlight at 2022-05-26T17:10:59+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 63a5650c by romes at 2022-05-31T12:43:22+01:00 TTG: Match new GHC AST - - - - - dd7d1617 by romes at 2022-06-02T16:11:00+01:00 Update for IE changes in !8228 - - - - - c23aaab7 by cydparser at 2022-06-06T08:48:14+02:00 Fix and improve CI (#1495) * Pin GHC version before creating the freeze file * Use newest action versions * Improve caching * Avoid unnecessarily reinstalling GHC * Use GHC 9.2.2 for CI Co-authored-by: Cyd Wise <cwise at tripshot.com> - - - - - c156fa77 by Hécate Moonlight at 2022-06-06T11:59:35+02:00 Add Mergify configuration (#1496) - - - - - 2dba4188 by Hécate Moonlight at 2022-06-06T16:12:50+02:00 Bump haddock's version in cabal file to 2.26.1 (#1497) - - - - - d7d4b8b9 by Marcin Szamotulski at 2022-06-07T06:09:40+00:00 Render module tree per package in the content page (#1492) * Render module tree per package in the content page When rendering content page for multiple packages it is useful to split the module tree per package. Package names in this patch are inferred from haddock's interface file names. * Write PackageInfo into interface file To keep interface file format backward compatible, instead of using `Binary` instance for `InterfaceFile` we introduce functions to serialise and deserialise, which depends on the interface file version. - - - - - 77765665 by Mike Pilgrem at 2022-06-12T21:57:19+01:00 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - b0e079b0 by mergify[bot] at 2022-06-13T11:49:37+00:00 Merge pull request haskell/haddock#1108 from mpilgrem/fix783 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - 6c0292b1 by Hécate Moonlight at 2022-06-21T17:21:08+02:00 Update the contribution guide - - - - - e413b9fa by dependabot[bot] at 2022-06-21T23:38:19+02:00 Bump shell-quote from 1.6.1 to 1.7.3 in /haddock-api/resources/html (#1500) Bumps [shell-quote](https://github.com/substack/node-shell-quote) from 1.6.1 to 1.7.3. - [Release notes](https://github.com/substack/node-shell-quote/releases) - [Changelog](https://github.com/substack/node-shell-quote/blob/master/CHANGELOG.md) - [Commits](https://github.com/substack/node-shell-quote/compare/1.6.1...1.7.3) --- updated-dependencies: - dependency-name: shell-quote dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 29d0ef70 by romes at 2022-07-06T11:29:39+02:00 TTG: AST Updates for !8308 - - - - - 1bae7c87 by Alan Zimmerman at 2022-07-06T22:50:43+01:00 Match GHC changes for T21805 This brings in a newtype for FieldLabelString - - - - - 6fe8b988 by Phil de Joux at 2022-07-16T20:54:26+00:00 Bump hlint version to 3.4.1, the version with counts. (#1503) Redo the counts available with the --default option. - - - - - 48fb43af by Phil de Joux at 2022-07-19T09:32:55+02:00 Follow hlint suggestion: unused LANGUAGE pragma. (#1504) * Follow hlint suggestion: unused LANGUAGE pragma. * Ignore within modules to pass linting and pass tests. - - - - - c1cf1fa7 by Phil de Joux at 2022-07-24T13:45:59+02:00 Follow hlint suggestion: redundant $. (#1505) * Follow hlint suggestion: redundant $. * Remove $ and surplus blank lines in Operators. - - - - - 74777eb2 by Jade Lovelace at 2022-07-29T11:02:41+01:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` (cherry picked from commit ab53ccf089ea703b767581ac14be0f6c78a7678a) - - - - - faa4cfcf by Hécate Moonlight at 2022-07-29T20:31:20+02:00 Merge pull request haskell/haddock#1516 from duog/9-4-backport-fix-hyperlinks Backport 9-4: Fix hyperlinks to external items and modules (#1482) - - - - - 5d2450f3 by Ben Gamari at 2022-08-05T17:41:15-04:00 Merge remote-tracking branch 'origin/ghc-9.4' - - - - - 63954f73 by Ben Gamari at 2022-08-05T19:08:36-04:00 Clean up build and testsuite for GHC 9.4 - - - - - d4568cb8 by Hécate Moonlight at 2022-08-05T19:10:49-04:00 Bump the versions - - - - - 505583a4 by Ben Gamari at 2022-08-06T13:58:27-04:00 Merge pull request haskell/haddock#1518 from bgamari/wip/ghc-9.4-merge Merge GHC 9.4 into `main` - - - - - 5706f6a4 by Ben Gamari at 2022-08-06T22:57:21-04:00 html-test: Testsuite changes for GHC 9.4.1 - - - - - 5f2a45a2 by Ben Gamari at 2022-08-15T14:33:05-04:00 doc: Fix a few minor ReST issues Sphinx was complaining about too-short title underlines. - - - - - 220e6410 by Ben Gamari at 2022-08-15T14:41:24-04:00 Merge branch 'main' into ghc-head - - - - - fbeb1b02 by Ben Gamari at 2022-08-15T14:45:16-04:00 Updates for GHC 9.5 - - - - - eee562eb by Vladislav Zavialov at 2022-08-15T14:46:13-04:00 HsToken ConDeclGADT con_dcolon - - - - - c5f073db by Ben Gamari at 2022-08-15T16:55:35-04:00 Updates for GHC 9.5 - - - - - 3f7ab242 by Vladislav Zavialov at 2022-08-15T16:55:35-04:00 HsToken ConDeclGADT con_dcolon - - - - - a18e473d by Ben Gamari at 2022-08-16T08:35:19-04:00 Merge branch 'wip/ghc-head-bump' into ghc-head - - - - - af0ff3a4 by M Farkas-Dyck at 2022-09-15T21:16:05+00:00 Disuse `mapLoc`. - - - - - a748fc38 by Matthew Farkas-Dyck at 2022-09-17T10:44:18+00:00 Scrub partiality about `NewOrData`. - - - - - 2758fb6c by John Ericson at 2022-09-18T03:27:37+02:00 Test output changed because of change to `base` Spooky, but I guess that is intended? - - - - - a7eec128 by Torsten Schmits at 2022-09-21T11:06:55+02:00 update tests for the move of tuples to GHC.Tuple.Prim - - - - - 461e7b9d by Ross Paterson at 2022-09-24T22:01:25+00:00 match implementation of GHC proposal haskell/haddock#106 (Define Kinds Without Promotion) - - - - - f7fd77ef by sheaf at 2022-10-17T14:53:01+02:00 Update Haddock for GHC MR !8563 (configuration of diagnostics) - - - - - 3d3e85ab by Vladislav Zavialov at 2022-10-22T23:04:06+03:00 Class layout info - - - - - cbde4cb0 by Simon Peyton Jones at 2022-10-25T23:19:18+01:00 Adapt to Constraint-vs-Type See haskell/haddock#21623 and !8750 - - - - - 7108ba96 by Tom Smeding at 2022-11-01T22:33:23+01:00 Remove outdated footnote about module re-exports The footnote is invalid with GHC 9.2.4 (and possibly earlier): the described behaviour in the main text works fine. - - - - - 206c6bc7 by Hécate Moonlight at 2022-11-01T23:00:46+01:00 Merge pull request haskell/haddock#1534 from tomsmeding/patch-1 - - - - - a57b4c4b by Andrew Lelechenko at 2022-11-21T00:39:52+00:00 Support mtl-2.3 - - - - - e9d62453 by Simon Peyton Jones at 2022-11-25T13:49:12+01:00 Track small API change in TyCon.hs - - - - - eb1c73f7 by Ben Gamari at 2022-12-07T08:46:21-05:00 Update for GhC 9.6 - - - - - 063268dd by Ben Gamari at 2022-12-07T11:26:32-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 4ca722fe by Ben Gamari at 2022-12-08T14:43:26-05:00 Bump bounds to accomodate base-4.18 - - - - - 340b7511 by Vladislav Zavialov at 2022-12-10T12:31:28+00:00 HsToken in HsAppKindTy - - - - - 946226ec by Ben Gamari at 2022-12-13T20:12:56-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - fd8faa66 by Ben Gamari at 2022-12-22T13:44:28-05:00 Bump GHC version to 9.7 - - - - - 2958aa9c by Ben Gamari at 2022-12-22T14:49:16-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 9e0fefd8 by Andrei Borzenkov at 2023-01-30T14:02:04+04:00 Rename () to Unit, Rename (,,...,,) to Tuple<n> - - - - - eb3968b5 by Ben Gamari at 2023-03-10T02:32:43-05:00 Bump versions for ghc-9.6 release - - - - - 4aeead36 by Adam Gundry at 2023-03-23T13:53:47+01:00 Adapt to warning categories changes - - - - - 642d8d60 by sheaf at 2023-03-29T13:35:56+02:00 Adapt to record field refactor This commit adapts to the changes in GHC MR !8686, which overhauls the treatment of record fields in the renamer, adding separate record field namespaces and entirely removing the GreName datatype. - - - - - ac8d4333 by doyougnu at 2023-03-29T11:11:44-04:00 Update UniqMap API - - - - - 7866fc86 by Ben Orchard at 2023-04-20T11:29:33+02:00 update classify with new tokens - - - - - ffcdd683 by Finley McIlwaine at 2023-04-24T09:36:18-06:00 Remove index-state - - - - - 05b70982 by Finley McIlwaine at 2023-04-26T08:16:31-06:00 `renameInterface` space leak fixes - Change logic for accumulation of names for which link warnings will be generated - Change handling of `--ignore-link-symbol` to allow qualified and unqualified names. Added to CHANGES.md - Some formatting changes and comments here and there - - - - - e5697d7c by Finley McIlwaine at 2023-04-27T18:46:36-06:00 Messy things - ghc-debug dependency and instrumentation - cabal.project custom with-compiler - hie.yaml files - traces and such - - - - - 0b8ef80b by Finley McIlwaine at 2023-05-02T18:08:52-06:00 Stop retaining GRE closures GRE closures should never be necessary to Haddock, so we never want to keep them on the heap. Despite that, they are retained by a lot of the data structures that Haddock makes use of. - Attempt to fix that situation by adding strictness to various computations and pruning the `ifaceInstances` field of `Interface` to a much thinner data type. - Removes the `ifaceFamInstances` field, as it was never used. - Move some of the attach instances types (e.g. `SimpleType`) to the types module - - - - - 8bda991b by Finley McIlwaine at 2023-05-08T16:07:51-06:00 Memory usage fixes - Refactor `ifaceDeclMap` to drastically reduce memory footprint. We no longer store all declarations associated with a given name, since we only cared to determine if the only declaration associated with a name was a value declaration. Change the `DeclMap` type to better reflect this. - Drop pre-renaming export items after the renaming step. Since the Hoogle backend used the pre-renamed export items, this isn't trivial. We now generate Hoogle output for exported declarations during the renaming step (if Hoogle output /should/ be generated), and store that with the renamed export item. - Slightly refactor Hoogle backend to handle the above change and allow for early generation of Hoogle output. - Remove the `ifaceRnDocMap` and `ifaceRnArgMap` fields of the `Interface` type, as they were never used. - Remove some unnecessary strictness - Remove a lot of dead code from `Syb` module - - - - - 1611ac0c by Finley McIlwaine at 2023-05-09T11:51:57-06:00 Unify ErrMsgM and IfM - Delete ErrMsgM, stop accumulating warnings in a writer - Make IfM a state monad, print warnings directly to stdout, move IfM type into types module - Drop ErrMsg = String synonym - Unset IORefs from plugin after they are read, preventing unnecessary retention of interfaces - - - - - 42d696ab by Finley McIlwaine at 2023-05-11T15:52:07-06:00 Thunk leak fixes The strictness introduced in this commit was motivated by observing thunk leaks in the eventlog2html output. - Refactor attach instances list comprehension to avoid large intermediate thunks - Refactor some HTML backend list comprehensions to avoid large intermediate thunks - Avoid thunks accumulating in documentation types or documentation parser - A lot of orphan NFData instances to allow us to force documentation values - - - - - 68561cf6 by Finley McIlwaine at 2023-05-11T17:02:10-06:00 Remove GHC debug dep - - - - - 10519e3d by Finley McIlwaine at 2023-05-15T12:40:48-06:00 Force HIE file path Removes a potential retainer of `ModSummary`s - - - - - 1e4a6ec6 by Finley McIlwaine at 2023-05-15T14:20:34-06:00 Re-add index-state, with-compiler, delete hie.yamls - - - - - a2363fe9 by Hécate Moonlight at 2023-05-15T22:45:16+02:00 Merge pull request haskell/haddock#1594 from FinleyMcIlwaine/finley/ghc-9.6-mem-fixes Reduce memory usage - - - - - e8a78383 by Finley McIlwaine at 2023-05-17T12:19:16-06:00 Merge branch ghc-9.6 into ghc-head - - - - - 22e25581 by Finley McIlwaine at 2023-05-17T12:20:23-06:00 Merge branch 'ghc-head' of gitlab.haskell.org:ghc/haddock into ghc-head - - - - - 41bbf0df by Bartłomiej Cieślar at 2023-05-24T08:57:58+02:00 changes to the WarningTxt cases Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - c686ba9b by Hécate Moonlight at 2023-06-01T14:03:02-06:00 Port the remains of Hi-Haddock - - - - - 9d8a85fd by Hécate Moonlight at 2023-06-01T14:03:06-06:00 Stdout for tests - - - - - 36331d07 by Finley McIlwaine at 2023-06-01T14:06:02-06:00 Formatting, organize imports - - - - - a06059b1 by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix empty context confusion in Convert module - - - - - 379346ae by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix associated type families in Hoogle output - - - - - fc6ea7ed by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix test refs Accept several changes in Hoogle tests: Pretty printing logic no longer prints the `(Proxy (Proxy (...))` chain in Bug806 with parentheses. Since this test was only meant to verify that line breaks do not occur, accept the change. `tyThingToLHsDecl` is called for class and data declarations, which ends up "synifying" the type variables and giving unlifted types kind signatures. As a result, type variables of kind `Type -> Type` are now printed with kind signatures in Hoogle output. This could be changed by manually drop kind signatures from class variables in the Hoogle backend if the behavior is deemed unacceptable. Sometimes subordinate declarations are exported separate from their parent declarations (e.g. record selectors). In this case, a type signature is cobbled together for the export item in `extractDecl`. Since this type signature is very manually constructed, it may lack kind signatures of decls constructed from `tyThingToLHsDecl`. An example of this is the `type-sigs` Hoogle test. Change `*` to `Type` in Hoogle test refs. I don't think this will break Hoogle behavior, since it appears to not consider type signatures in search. I have not fully verified this. - - - - - e14b7e58 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix LaTeX backend test refs Changes to GHC pretty printing code have resulted in some differences to Haddock's LaTeX output. - Type variables are printed explicitly quantified in the LinearTypes test - Wildcard types in type family equations are now printed numbered, e.g. _1 _2, in the TypeFamilies3 test - Combined signatures in DefaultSignatures test are now documented as separate signatures - - - - - 41b5b296 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and test source updates - Stop using kind `*` in html test sources - Add TypeOperators where necessary to avoid warnings and future errors - Rename some test modules to match their module names - - - - - c640e2a2 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix missing deprecation warnings on record fields `lookupOccEnv` was used to resolve `OccNames` with warnings attached, but it doesn't look in the record field namespace. Thus, if a record field had a warning attached, it would not resolve and the warning map would not include it. This commit fixes by using `lookupOccEnv_WithFields` instead. - - - - - fad0c462 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and some comments - - - - - 751fd023 by Finley McIlwaine at 2023-06-01T14:11:41-06:00 Accept HTML test diffs All diffs now boil down to the expected differences resulting from declarations being reified from TyThings in hi-haddock. Surface syntax now has much less control over the syntax used in the documentation. - - - - - d835c845 by Finley McIlwaine at 2023-06-01T14:11:45-06:00 Adapt to new load' type - - - - - dcf776c4 by Finley McIlwaine at 2023-06-01T14:13:13-06:00 Update mkWarningMap and moduleWarning - - - - - 8e8432fd by Finley McIlwaine at 2023-06-01T14:28:54-06:00 Revert load' changes - - - - - aeb2982c by Finley McIlwaine at 2023-06-01T14:40:24-06:00 Accept change to Instances test in html-test Link to Data.Tuple instead of GHC.Tuple.Prim - - - - - 8adfdbac by Finley McIlwaine at 2023-06-01T15:53:17-06:00 Reset ghc dep to ^>= 9.6 - - - - - 2b1ce93d by Finley McIlwaine at 2023-06-06T07:50:04-06:00 Update CHANGES.md, user guide, recomp avoidance * Add --trace-args flag for tracing arguments received to standard output * Avoid recompiling due to changes in optimization flags * Update users guide and changes.md - - - - - f3da6676 by Finley McIlwaine at 2023-06-06T14:12:56-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - bf36c467 by Matthew Pickering at 2023-06-07T10:16:09+01:00 Revert back to e16e20d592a6f5d9ed1af17b77fafd6495242345 Neither of these MRs are ready to land yet which causes issues with other MRs which are ready to land and need haddock changes. - - - - - 421510a9 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 atSign has no unicode variant Prior to this change, atSign was defined as follows: atSign unicode = text (if unicode then "@" else "@") Yes, this is the same symbol '\64' and not your font playing tricks on you. Now we define: atSign = char '@' Both the LaTeX and the Xhtml backend are updated accordingly. - - - - - 3785c276 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 LaTeX: fix printing of type variable bindings Consider this type signature: kindOf :: forall {k} (a :: k). Proxy a -> Proxy k Prior to this fix, the LaTeX backend rendered it like this: kindOf :: forall k a. Proxy a -> Proxy k Now we preserve explicit specificity and kind annotations. - - - - - 0febf3a8 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 Add support for invisible binders in type declarations - - - - - 13e33bb3 by Finley McIlwaine at 2023-06-08T07:51:59-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - 3e5340ce by Finley McIlwaine at 2023-06-08T07:54:27-06:00 Add note about stubdir to recompilation docs - - - - - db7e84dc by Finley at 2023-06-08T08:11:03-06:00 Merge pull request haskell/haddock#1597 from haskell/finley/hi-haddock-9.6 hi-haddock for ghc 9.6 - - - - - 4e085d17 by Finley McIlwaine at 2023-06-14T13:41:06-06:00 Replace SYB traversals - - - - - 7b39aec5 by Finley McIlwaine at 2023-06-14T14:20:17-06:00 Test ref accepts, remove unused HaddockClsInst - - - - - df9c2090 by Finley McIlwaine at 2023-06-15T08:02:51-06:00 Use batchMsg for progress reporting during load With hi-haddock as is, there is an awkward silence during the load operation. This commit makes haddock use the default `batchMsg` Messager for progress reporting, and makes the default GHC verbosity level 1, so the user can see what GHC is doing. - - - - - f23679a8 by Hécate Moonlight at 2023-06-15T20:31:53+02:00 Merge pull request haskell/haddock#1600 from haskell/finley/hi-haddock-optim - - - - - a7982192 by Finley McIlwaine at 2023-06-15T15:02:16-06:00 hi-haddock squashed - - - - - c34f0c8d by Finley McIlwaine at 2023-06-15T16:22:03-06:00 Merge remote-tracking branch 'origin/ghc-9.6' into finley/hi-haddock-squashed - - - - - 40452797 by Bartłomiej Cieślar at 2023-06-16T12:26:04+02:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - e58673bf by Ben Gamari at 2023-06-16T09:33:35-04:00 Bump GHC version to 9.8 - - - - - 74bdf972 by Ben Gamari at 2023-06-16T09:36:18-04:00 Merge commit 'fcaaad06770a26d35d4aafd65772dedadf17669c' into ghc-head - - - - - 418ee3dc by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Remove NFData SourceText, IfaceWarnings updates The NFData SourceText instance is now available in GHC Handle mod_iface mi_warns now being IfaceWarnings - - - - - 62f31380 by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Accept Instances.hs test output Due to ghc!10469. - - - - - a8f2fc0e by Ben Gamari at 2023-06-20T15:48:08-04:00 Test fixes for "Fix associated data family doc structure items" Associated data families were being given their own export DocStructureItems, which resulted in them being documented separately from their classes in haddocks. This commit fixes it. - - - - - cb1ac33e by Bartłomiej Cieślar at 2023-06-21T12:56:02-04:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - 9933e10b by Ben Gamari at 2023-06-21T12:56:02-04:00 Bump GHC version to 9.8 - - - - - fe8c18b6 by Ben Gamari at 2023-06-21T15:36:29-04:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - c61a0d5b by Ben Gamari at 2023-06-21T16:10:51-04:00 Bump GHC version to 9.9 - - - - - 0c2a756e by sheaf at 2023-07-07T13:45:12+02:00 Avoid incomplete record update in Haddock Hoogle This commit avoids running into an incomplete record update warning in the Hoogle backend of Haddock. This was only noticed now, because incomplete record updates were broken in GHC 9.6. Now that they are fixed, we have to avoid running into them! - - - - - f9b952a7 by Ben Gamari at 2023-07-21T11:58:05-04:00 Bump base bound to <4.20 For GHC 9.8. - - - - - 1b27e151 by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Check for puns (see ghc#23368) - - - - - 457341fd by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Remove fake exports for (~), List, and Tuple<n> The old reasoning no longer applies, nowadays those names can be mentioned in export lists. - - - - - bf3dcddf by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Fix pretty-printing of Solo and MkSolo - - - - - 495b2241 by Matthew Pickering at 2023-09-01T13:02:07+02:00 Fix issue with duplicate reexported definitions (T23616) When a class method was reexported, it's default methods were also showing up in the generated html page. The simplest and most non-invasive fix is to not look for the default method if we are just exporting the class method.. because the backends are just showing default methods when the whole class is exported. In general it would be worthwhile to rewrite this bit of code I think as the logic and what gets included is split over `lookupDocs` and `availExportDecl` it would be clearer to combine the two. The result of lookupDocs is always just passed to availExportDecl so it seems simpler and more obvious to just write the function directly. - - - - - 6551824d by Finley McIlwaine at 2023-09-05T13:06:57-07:00 Remove fake export of `FUN` from Prelude This prevents `data FUN` from being shown at the top of the Prelude docs. Fixes \#23920 on GHC. - - - - - 9ab5a448 by Alan Zimmerman at 2023-09-08T18:26:53+01:00 Match changes in wip/az/T23885-unicode-funtycon - - - - - 4d08364e by Alan Zimmerman at 2023-10-31T19:46:45+00:00 EPA: match changes in GHC - EPA: Comments in AnchorOperation - EPA: Remove EpaEofComment - - - - - e7da0d25 by Alan Zimmerman at 2023-11-05T11:20:31+00:00 EPA: match changes in GHC, l2l cleanup - - - - - 4ceac14d by Alan Zimmerman at 2023-11-11T15:16:41+00:00 EPA: Replace Anchor with EpaLocation Match GHC - - - - - 94fb8d47 by Alan Zimmerman at 2023-11-29T18:10:26+00:00 Match GHC, No comments in EpaDelta for comments - - - - - 32d208e1 by Vladislav Zavialov at 2023-12-12T20:41:36+03:00 EPA: Match changes to LHsToken removal - - - - - eebdd316 by Apoorv Ingle at 2024-01-23T13:49:12+00:00 Changes for haskell/haddock#18324 - - - - - ae856a82 by Matthew Pickering at 2024-02-05T12:22:39+00:00 ghc-internals fallout - - - - - f8429266 by Jade at 2024-02-08T14:56:50+01:00 Adjust test for ghc MR !10993 - - - - - 6d1e2386 by Alan Zimmerman at 2024-02-13T22:00:28+03:00 EPA: Match changes to HsParTy and HsFunTy - - - - - 9c588f19 by Fendor at 2024-02-14T11:05:36+01:00 Adapt to GHC giving better Name's for linking - - - - - 778e1db3 by Andrei Borzenkov at 2024-02-16T16:12:07+03:00 Namespace specifiers for fixity signatures - - - - - 826c5b47 by Torsten Schmits at 2024-02-21T13:17:05+01:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 2cff14d5 by Ben Gamari at 2024-02-22T09:35:56-05:00 Bump bounds - - - - - f49376b3 by Ben Gamari at 2024-02-22T09:35:56-05:00 Allow `@since` annotations in export lists Here we extend Haddock to admit `@since` annotations in export lists. These can be attached to most export list items (although not subordinate lists). These annotations supercede the declaration's `@since` annotation in produced Haddocks. - - - - - b5aa93df by Ben Gamari at 2024-02-22T12:09:06-05:00 Allow package-qualified @since declarations - - - - - 8f5957f2 by Ben Gamari at 2024-02-22T13:55:19-05:00 Documentation changes from ghc-internal restructuring Previously many declarations (e.g. `Int`) were declared to have a "home" in `Prelude`. However, now Haddock instead chooses to put these in more specific homes (e.g. `Data.Int`). Given that the "home" decision is driven by heuristics and in general these changes seem quite reasonable I am accepting them: * `Int` moved from `Prelude` to `Data.Int` * `(~)` moved from `Prelude` to `Data.Type.Equality` * `Type` moved from `GHC.Types` to `Data.Kind` * `Maybe` moved from `Prelude` to `Data.Maybe` * `Bool` moved from `Prelude` to `Data.Bool` * `Ordering` moved from `Prelude` to `Data.Ord` As well, more identifiers are now hyperlinked; it's not immediately clear *why*, but it is an improvement nevertheless. - - - - - ec33fec3 by Ben Gamari at 2024-02-22T20:36:24-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 30cfd251 by Torsten Schmits at 2024-02-24T13:00:42-05:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 732db81d by Ben Gamari at 2024-02-24T19:12:18-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 86bf7010 by Ben Gamari at 2024-02-27T19:28:10-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 4b6e76b5 by Patrick at 2024-03-07T22:09:30+08:00 fix haskell/haddock#24493, with module name introduced in hieAst The accompanies haddoc PR with GHC PR https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12153 Two things have been done: 1. Link is introduced to every `X` in `module X where`, since we introduce the module name to HieAst, 2. `{-# LINE 4 "hypsrc-test/src/PositionPragmas.hs" #-}` is added before the `module PositionPragmas where` in ` hypsrc-test/ref/src/PositionPragmas.html `.It ensures only a single hieAst for file `hypsrc-test/src/PositionPragmas.hs` is generated. - - - - - 635abccc by Ben Gamari at 2024-03-08T17:09:06-05:00 Bump ghc version to 9.10 - - - - - 5b934048 by Ben Gamari at 2024-03-08T18:50:12-05:00 Bump base upper bound - - - - - b30d134e by Ben Gamari at 2024-03-08T18:50:44-05:00 Testsuite output update - - - - - 9bdf3586 by Ben Gamari at 2024-03-09T21:37:44-05:00 Merge branch 'ghc-9.10' into ghc-head - - - - - cec76981 by Ben Gamari at 2024-03-09T21:54:00-05:00 Bump GHC version to 9.11 - - - - - 4c59feb7 by Ben Gamari at 2024-03-09T22:15:01-05:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - bee54c24 by Krzysztof Gogolewski at 2024-04-19T11:13:00+02:00 Update quantification order following GHC haskell/haddock#23764 - - - - - 2814eb89 by Ben Gamari at 2024-04-19T18:57:05+02:00 hypsrc-test: Fix output of PositionPragmas.html - - - - - a27c6a49 by Fendor at 2024-04-22T10:13:03+02:00 Adapt to UserData split - - - - - 1efc5a7a by Fendor at 2024-04-22T10:13:03+02:00 Adapt to BinHandle split - - - - - c62239b7 by Sylvain Henry at 2024-04-29T10:35:00+02:00 Fix tests for T22229 - - - - - b2682534 by Rodrigo Mesquita at 2024-05-10T01:47:51-04:00 Document NcgImpl methods Fixes #19914 - - - - - 4d3acbcf by Zejun Wu at 2024-05-10T01:48:28-04:00 Make renamer to be more flexible with parens in the LHS of the rules We used to reject LHS like `(f a) b` in RULES and requires it to be written as `f a b`. It will be handy to allow both as the expression may be more readable with extra parens in some cases when infix operator is involved. Espceially when TemplateHaskell is used, extra parens may be added out of user's control and result in "valid" rules being rejected and there are not always ways to workaround it. Fixes #24621 - - - - - ab840ce6 by Ben Gamari at 2024-05-10T01:49:04-04:00 IPE: Eliminate dependency on Read Instead of encoding the closure type as decimal string we now simply represent it as an integer, eliminating the need for `Read` in `GHC.Internal.InfoProv.Types.peekInfoProv`. Closes #24504. ------------------------- Metric Decrease: T24602_perf_size size_hello_artifact ------------------------- - - - - - a9979f55 by Cheng Shao at 2024-05-10T01:49:43-04:00 testsuite: fix testwsdeque with recent clang This patch fixes compilation of testwsdeque.c with recent versions of clang, which will fail with the error below: ``` testwsdeque.c:95:33: error: warning: format specifies type 'long' but the argument has type 'void *' [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~~ ^ testwsdeque.c:95:39: error: warning: format specifies type 'int' but the argument has type 'StgWord' (aka 'unsigned long') [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~ ^~~ | %lu testwsdeque.c:133:42: error: error: incompatible function pointer types passing 'void (void *)' to parameter of type 'OSThreadProc *' (aka 'void *(*)(void *)') [-Wincompatible-function-pointer-types] 133 | createOSThread(&ids[n], "thief", thief, (void*)(StgWord)n); | ^~~~~ /workspace/ghc/_build/stage1/lib/../lib/x86_64-linux-ghc-9.11.20240502/rts-1.0.2/include/rts/OSThreads.h:193:51: error: note: passing argument to parameter 'startProc' here 193 | OSThreadProc *startProc, void *param); | ^ 2 warnings and 1 error generated. ``` - - - - - c2b33fc9 by Rodrigo Mesquita at 2024-05-10T01:50:20-04:00 Rename pre-processor invocation args Small clean up. Uses proper names for the various groups of arguments that make up the pre-processor invocation. - - - - - 2b1af08b by Cheng Shao at 2024-05-10T01:50:55-04:00 ghc-heap: fix typo in ghc-heap cbits - - - - - fc2d6de1 by Jade at 2024-05-10T21:07:16-04:00 Improve performance of Data.List.sort(By) This patch improves the algorithm to sort lists in base. It does so using two strategies: 1) Use a four-way-merge instead of the 'default' two-way-merge. This is able to save comparisons and allocations. 2) Use `(>) a b` over `compare a b == GT` and allow inlining and specialization. This mainly benefits types with a fast (>). Note that this *may* break instances with a *malformed* Ord instance where `a > b` is *not* equal to `compare a b == GT`. CLC proposal: https://github.com/haskell/core-libraries-committee/issues/236 Fixes #24280 ------------------------- Metric Decrease: MultiLayerModulesTH_Make T10421 T13719 T15164 T18698a T18698b T1969 T9872a T9961 T18730 WWRec T12425 T15703 ------------------------- - - - - - 1012e8aa by Matthew Pickering at 2024-05-10T21:07:52-04:00 Revert "ghcup-metadata: Drop output_name field" This reverts commit ecbf22a6ac397a791204590f94c0afa82e29e79f. This breaks the ghcup metadata generation on the nightly jobs. - - - - - daff1e30 by Jannis at 2024-05-12T13:38:35-04:00 Division by constants optimization - - - - - 413217ba by Andreas Klebinger at 2024-05-12T13:39:11-04:00 Tidy: Add flag to expose unfoldings if they take dictionary arguments. Add the flag `-fexpose-overloaded-unfoldings` to be able to control this behaviour. For ghc's boot libraries file size grew by less than 1% when it was enabled. However I refrained from enabling it by default for now. I've also added a section on specialization more broadly to the users guide. ------------------------- Metric Decrease: MultiLayerModulesTH_OneShot Metric Increase: T12425 T13386 hard_hole_fits ------------------------- - - - - - c5d89412 by Zubin Duggal at 2024-05-13T22:19:53-04:00 Don't store a GlobalRdrEnv in `mi_globals` for GHCi. GHCi only needs the `mi_globals` field for modules imported with :module +*SomeModule. It uses this field to make the top level environment in `SomeModule` available to the repl. By default, only the first target in the command line parameters is "star" loaded into GHCi. Other modules have to be manually "star" loaded into the repl. Storing the top level GlobalRdrEnv for each module is very wasteful, especially given that we will most likely never need most of these environments. Instead we store only the information needed to reconstruct the top level environment in a module, which is the `IfaceTopEnv` data structure, consisting of all import statements as well as all top level symbols defined in the module (not taking export lists into account) When a particular module is "star-loaded" into GHCi (as the first commandline target, or via an explicit `:module +*SomeModule`, we reconstruct the top level environment on demand using the `IfaceTopEnv`. - - - - - d65bf4a2 by Fendor at 2024-05-13T22:20:30-04:00 Add perf regression test for `-fwrite-if-simplified-core` - - - - - 2c0f8ddb by Andrei Borzenkov at 2024-05-13T22:21:07-04:00 Improve pattern to type pattern transformation (23739) `pat_to_type_pat` function now can handle more patterns: - TuplePat - ListPat - LitPat - NPat - ConPat Allowing these new constructors in type patterns significantly increases possible shapes of type patterns without `type` keyword. This patch also changes how lookups in `lookupOccRnConstr` are performed, because we need to fall back into types when we didn't find a constructor on data level to perform `ConPat` to type transformation properly. - - - - - be514bb4 by Cheng Shao at 2024-05-13T22:21:43-04:00 hadrian: fix hadrian building with ghc-9.10.1 - - - - - ad38e954 by Cheng Shao at 2024-05-13T22:21:43-04:00 linters: fix lint-whitespace compilation with ghc-9.10.1 - - - - - a593f284 by Andreas Klebinger at 2024-05-15T07:32:10-04:00 Expand the `inline` rule to look through casts/ticks. Fixes #24808 - - - - - b1e0c313 by Cheng Shao at 2024-05-15T07:32:46-04:00 testsuite: bump PartialDownSweep timeout to 5x on wasm32 - - - - - b2227487 by Fendor at 2024-05-15T17:14:06-04:00 Add Eq and Ord instance to `IfaceType` We add an `Ord` instance so that we can store `IfaceType` in a `Data.Map` container. This is required to deduplicate `IfaceType` while writing `.hi` files to disk. Deduplication has many beneficial consequences to both file size and memory usage, as the deduplication enables implicit sharing of values. See issue #24540 for more motivation. The `Ord` instance would be unnecessary if we used a `TrieMap` instead of `Data.Map` for the deduplication process. While in theory this is clerarly the better option, experiments on the agda code base showed that a `TrieMap` implementation has worse run-time performance characteristics. To the change itself, we mostly derive `Eq` and `Ord`. This requires us to change occurrences of `FastString` with `LexicalFastString`, since `FastString` has no `Ord` instance. We change the definition of `IfLclName` to a newtype of `LexicalFastString`, to make such changes in the future easier. Bump haddock submodule for IfLclName changes - - - - - d368f9a6 by Fendor at 2024-05-15T17:14:06-04:00 Move out LiteralMap to avoid cyclic module dependencies - - - - - 2fcc09fd by Fendor at 2024-05-15T17:14:06-04:00 Add deduplication table for `IfaceType` The type `IfaceType` is a highly redundant, tree-like data structure. While benchmarking, we realised that the high redundancy of `IfaceType` causes high memory consumption in GHCi sessions when byte code is embedded into the `.hi` file via `-fwrite-if-simplified-core` or `-fbyte-code-and-object-code`. Loading such `.hi` files from disk introduces many duplicates of memory expensive values in `IfaceType`, such as `IfaceTyCon`, `IfaceTyConApp`, `IA_Arg` and many more. We improve the memory behaviour of GHCi by adding an additional deduplication table for `IfaceType` to the serialisation of `ModIface`, similar to how we deduplicate `Name`s and `FastString`s. When reading the interface file back, the table allows us to automatically share identical values of `IfaceType`. To provide some numbers, we evaluated this patch on the agda code base. We loaded the full library from the `.hi` files, which contained the embedded core expressions (`-fwrite-if-simplified-core`). Before this patch: * Load time: 11.7 s, 2.5 GB maximum residency. After this patch: * Load time: 7.3 s, 1.7 GB maximum residency. This deduplication has the beneficial side effect to additionally reduce the size of the on-disk interface files tremendously. For example, on agda, we reduce the size of `.hi` files (with `-fwrite-if-simplified-core`): * Before: 101 MB on disk * Now: 24 MB on disk This has even a beneficial side effect on the cabal store. We reduce the size of the store on disk: * Before: 341 MB on disk * Now: 310 MB on disk Note, none of the dependencies have been compiled with `-fwrite-if-simplified-core`, but `IfaceType` occurs in multiple locations in a `ModIface`. We also add IfaceType deduplication table to .hie serialisation and refactor .hie file serialisation to use the same infrastrucutre as `putWithTables`. Bump haddock submodule to accomodate for changes to the deduplication table layout and binary interface. - - - - - 36aa7cf1 by Fendor at 2024-05-15T17:14:06-04:00 Add run-time configurability of `.hi` file compression Introduce the flag `-fwrite-if-compression=<n>` which allows to configure the compression level of writing .hi files. The motivation is that some deduplication operations are too expensive for the average use case. Hence, we introduce multiple compression levels with variable impact on performance, but still reduce the memory residency and `.hi` file size on disk considerably. We introduce three compression levels: * `1`: `Normal` mode. This is the least amount of compression. It deduplicates only `Name` and `FastString`s, and is naturally the fastest compression mode. * `2`: `Safe` mode. It has a noticeable impact on .hi file size and is marginally slower than `Normal` mode. In general, it should be safe to always use `Safe` mode. * `3`: `Full` deduplication mode. Deduplicate as much as we can, resulting in minimal .hi files, but at the cost of additional compilation time. Reading .hi files doesn't need to know the initial compression level, and can always deserialise a `ModIface`, as we write out a byte that indicates the next value has been deduplicated. This allows users to experiment with different compression levels for packages, without recompilation of dependencies. Note, the deduplication also has an additional side effect of reduced memory consumption to implicit sharing of deduplicated elements. See https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for example where that matters. ------------------------- Metric Decrease: MultiLayerModulesDefsGhciWithCore T16875 T21839c T24471 hard_hole_fits libdir ------------------------- - - - - - 1e63a6fb by Matthew Pickering at 2024-05-15T17:14:07-04:00 Introduce regression tests for `.hi` file sizes Add regression tests to track how `-fwrite-if-compression` levels affect the size of `.hi` files. - - - - - 639d742b by M Farkas-Dyck at 2024-05-15T17:14:49-04:00 TTG: ApplicativeStatement exist only in Rn and Tc Co-Authored-By: romes <rodrigo.m.mesquita at gmail.com> - - - - - aa7b336b by Jade at 2024-05-15T23:06:17-04:00 Documentation: Improve documentation for symbols exported from System.IO - - - - - c561de8f by Jade at 2024-05-15T23:06:54-04:00 Improve suggestions for language extensions - When suggesting Language extensions, also suggest Extensions which imply them - Suggest ExplicitForAll and GADTSyntax instead of more specific extensions - Rephrase suggestion to include the term 'Extension' - Also moves some flag specific definitions out of Session.hs into Flags.hs (#24478) Fixes: #24477 Fixes: #24448 Fixes: #10893 - - - - - 4c7ae2a1 by Andreas Klebinger at 2024-05-15T23:07:30-04:00 Testsuite: Check if llvm assembler is available for have_llvm - - - - - bc672166 by Torsten Schmits at 2024-05-15T23:08:06-04:00 refactor quadratic search in warnMissingHomeModules - - - - - 7875e8cb by Torsten Schmits at 2024-05-15T23:08:06-04:00 add test that runs MakeDepend on thousands of modules - - - - - b84b91f5 by Adam Gundry at 2024-05-16T15:32:06-04:00 Representation-polymorphic HasField (fixes #22156) This generalises the HasField class to support representation polymorphism, so that instead of type HasField :: forall {k} . k -> Type -> Type -> Constraint we have type HasField :: forall {k} {r_rep} {a_rep} . k -> TYPE r_rep -> TYPE a_rep -> Constraint - - - - - 05285090 by Matthew Pickering at 2024-05-16T15:32:43-04:00 Bump os-string submodule to 2.0.2.2 Closes #24786 - - - - - 886ab43a by Cheng Shao at 2024-05-17T01:34:50-04:00 rts: do not prefetch mark_closure bdescr in non-moving gc when ASSERTS_ENABLED This commit fixes a small an oversight in !12148: the prefetch logic in non-moving GC may trap in debug RTS because it calls Bdescr() for mark_closure which may be a static one. It's fine in non-debug RTS because even invalid bdescr addresses are prefetched, they will not cause segfaults, so this commit implements the most straightforward fix: don't prefetch mark_closure bdescr when assertions are enabled. - - - - - b38dcf39 by Teo Camarasu at 2024-05-17T01:34:50-04:00 rts: Allocate non-moving segments with megablocks Non-moving segments are 8 blocks long and need to be aligned. Previously we serviced allocations by grabbing 15 blocks, finding an aligned 8 block group in it and returning the rest. This proved to lead to high levels of fragmentation as a de-allocating a segment caused an 8 block gap to form, and this could not be reused for allocation. This patch introduces a segment allocator based around using entire megablocks to service segment allocations in bulk. When there are no free segments, we grab an entire megablock and fill it with aligned segments. As the megablock is free, we can easily guarantee alignment. Any unused segments are placed on a free list. It only makes sense to free segments in bulk when all of the segments in a megablock are freeable. After sweeping, we grab the free list, sort it, and find all groups of segments where they cover the megablock and free them. This introduces a period of time when free segments are not available to the mutator, but the risk that this would lead to excessive allocation is low. Right after sweep, we should have an abundance of partially full segments, and this pruning step is relatively quick. In implementing this we drop the logic that kept NONMOVING_MAX_FREE segments on the free list. We also introduce an eventlog event to log the amount of pruned/retained free segments. See Note [Segment allocation strategy] Resolves #24150 ------------------------- Metric Decrease: T13253 T19695 ------------------------- - - - - - 710665bd by Cheng Shao at 2024-05-17T01:35:30-04:00 rts: fix I/O manager compilation errors for win32 target This patch fixes I/O manager compilation errors for win32 target discovered when cross-compiling to win32 using recent clang: ``` rts/win32/ThrIOManager.c:117:7: error: error: call to undeclared function 'is_io_mng_native_p'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 117 | if (is_io_mng_native_p ()) { | ^ | 117 | if (is_io_mng_native_p ()) { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/fs.c:143:28: error: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] 143 | int setErrNoFromWin32Error () { | ^ | void | 143 | int setErrNoFromWin32Error () { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:227:9: error: error: call to undeclared function 'interruptIOManagerEvent'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 227 | interruptIOManagerEvent (); | ^ | 227 | interruptIOManagerEvent (); | ^ rts/win32/ConsoleHandler.c:227:9: error: note: did you mean 'getIOManagerEvent'? | 227 | interruptIOManagerEvent (); | ^ rts/include/rts/IOInterface.h:27:10: error: note: 'getIOManagerEvent' declared here 27 | void * getIOManagerEvent (void); | ^ | 27 | void * getIOManagerEvent (void); | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:196:9: error: error: call to undeclared function 'setThreadLabel'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/win32/ConsoleHandler.c:196:9: error: note: did you mean 'postThreadLabel'? | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/eventlog/EventLog.h:118:6: error: note: 'postThreadLabel' declared here 118 | void postThreadLabel(Capability *cap, | ^ | 118 | void postThreadLabel(Capability *cap, | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) ``` - - - - - 28b9cee0 by Rodrigo Mesquita at 2024-05-17T01:36:05-04:00 configure: Check C99-compat for Cmm preprocessor Fixes #24815 - - - - - 8927e0c3 by Andreas Klebinger at 2024-05-17T01:36:41-04:00 Ensure `tcHasFixedRuntimeRep (# #)` returns True. - - - - - 04179044 by doyougnu at 2024-05-17T09:00:32-04:00 testsuite: make find_so regex less general Closes #24759 Background. In MR !12372 we began tracking shared object files and directories sizes for dependencies. However, this broke release builds because release builds alter the filenames swapping "in-place" for a hash. This was not considered in the MR and thus broke release pipelines. Furthermore, the rts_so test was found to be wildly varying and was therefore disabled in !12561. This commit fixes both of these issues: - fix the rts_so test by making the regex less general, now the rts_so test and all other foo.so tests must match "libHS<some-lib>-<version>-<hash|'in-place>-<ghc>". This prevents the rts_so test from accidentally matching different rts variants such as rts_threaded, which was the cause of the wild swings after !12372. - add logic to match either a hash or the string in-place. This should make the find_so function build agnostic. - - - - - 0962b50d by Andreas Klebinger at 2024-05-17T09:01:08-04:00 TagAnalysis: Treat all bottom ids as tagged during analysis. Ticket #24806 showed that we also need to treat dead end thunks as tagged during the analysis. - - - - - 7eb9f184 by Ben Gamari at 2024-05-17T11:23:37-04:00 Remove haddock submodule In preparation for merge into the GHC, as proposed in #23178. - - - - - 47b14dcc by Fendor at 2024-05-17T11:28:17-04:00 Adapt to `IfLclName` newtype changes (cherry picked from commit a711607e29b925f3d69e27c5fde4ba655c711ff1) - - - - - 6cc6681d by Fendor at 2024-05-17T11:28:17-04:00 Add IfaceType deduplication table to interface file serialisation Although we do not really need it in the interface file serialisation, as the deserialisation uses `getWithUserData`, we need to mirror the structure `getWithUserData` expects. Thus, we write essentially an empty `IfaceType` table at the end of the file, as the interface file doesn't reference `IfaceType`. (cherry picked from commit c9bc29c6a708483d2abc3d8ec9262510ce87ca61) - - - - - b9721206 by Ben Gamari at 2024-05-17T11:30:22-04:00 ghc-tags.yaml: Initial commit - - - - - 074e7d8f by Ben Gamari at 2024-05-17T11:31:29-04:00 fourmolu: Add configuration - - - - - 151b1736 by Ben Gamari at 2024-05-17T11:32:52-04:00 Makefile: Rework for use by haddock developers Previously the Makefile was present only for GHC's old make-based build system. Now since the make-based build system is gone we can use it for more useful ends. - - - - - a7dcf13b by Ben Gamari at 2024-05-17T11:36:14-04:00 Reformat with fourmolu Using previously-added configuration and `fourmolu -i .` Note that we exclude the test-cases (`./{hoogle,html-hypsrc,latex}-test`) as they are sensitive to formatting. - - - - - 0ea6017b by Ben Gamari at 2024-05-17T11:40:04-04:00 Add 'utils/haddock/' from commit 'a7dcf13bfbb97b20e75cc8ce650e2bb628db4660' git-subtree-dir: utils/haddock git-subtree-mainline: 7eb9f1849b1c72a1c61dee88462b4244550406f3 git-subtree-split: a7dcf13bfbb97b20e75cc8ce650e2bb628db4660 - - - - - aba1d304 by Hécate Moonlight at 2024-05-17T11:40:48-04:00 Add exceptions to the dangling notes list - - - - - 527bfbfb by Hécate Moonlight at 2024-05-17T11:40:52-04:00 Add haddock to the whitespace lint ignore list - - - - - 43274677 by Ben Gamari at 2024-05-17T11:41:20-04:00 git-blame-ignore-revs: Ignore haddock reformatting - - - - - 0e679e37 by Fendor at 2024-05-18T00:27:24-04:00 Pass cpp options to the CC builder in hadrian - - - - - bb40244e by Sylvain Henry at 2024-05-18T00:28:06-04:00 JS: fix allocation constant (fix #24746) - - - - - 646d30ab by Jade at 2024-05-18T19:23:31+02:00 Add highlighting for inline-code snippets in haddock - - - - - 64459a3e by Hécate Moonlight at 2024-05-19T08:42:27-04:00 haddock: Add a .readthedocs.yml file for online documentation - - - - - 7d3d9bbf by Serge S. Gulin at 2024-05-19T18:47:05+00:00 Unicode: General Category size test (related #24789) Added trivial size performance test which involves unicode general category usage via `read`. The `read` itself uses general category to detect spaces. The purpose for this test is to measure outcome of applying improvements at General Category representation in code discussed at #24789. - - - - - 8e04efcf by Alan Zimmerman at 2024-05-19T21:29:34-04:00 EPA: Remove redundant code Remove unused epAnnAnns function various cases for showAstData that no longer exist - - - - - 071d7a1e by Rodrigo Mesquita at 2024-05-20T10:55:16-04:00 Improve docs on closed type families in hs-boots Fixes #24776 - - - - - d9e2c119 by Torsten Schmits at 2024-05-20T10:55:52-04:00 Use default deviation for large-project test This new performance test has the purpose of detecting regressions in complexity in relation to the number of modules in a project, so 1% deviation is way too small to avoid false positives. - - - - - 20b0136a by Ben Gamari at 2024-05-22T00:31:39-04:00 ghcup-metadata: Various fixes from 9.10.1 Use Debian 12/x86-64, Debian 10/aarch64, and Debian 11/aarch64 bindists where possible. - - - - - 6838a7c3 by Sylvain Henry at 2024-05-22T00:32:23-04:00 Reverse arguments to stgCallocBytes (fix #24828) - - - - - f50f46c3 by Fendor at 2024-05-22T00:32:59-04:00 Add log messages for Iface serialisation compression level Fix the label of the number of 'IfaceType' entries in the log message. Add log message for the compression level that is used to serialise a an interface file. Adds `Outputable` instance for 'CompressionIFace'. - - - - - 3bad5d55 by Hécate Moonlight at 2024-05-22T00:33:40-04:00 base: Update doctests outputs ghc-internal: Update doctests outputs - - - - - 9317c6fb by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix the testsuites of the haddock-library - Apply all the metadata revisions from Hackage to the cabal file. - Fix the `ParserSpec.hs` file in the `spec` testsuite of haddock-library. - Make `CHANGES.md` an extra-doc-file instead of an extra-source-file. - - - - - 54073b02 by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix parser of @since pragma The testsuite contained tests for annotations of the form `@since foo-bar-0.5.0`, but the parser was written incorrectly. - - - - - ede6ede3 by Matthew Pickering at 2024-05-22T00:34:57-04:00 Fix nightly pages job It seems likely broken by 9f99126a which moved `index.html` from the root folder into `docs/` folder. Fixes #24840 - - - - - b7bcf729 by Cheng Shao at 2024-05-22T00:35:32-04:00 autoconf: remove unused context diff check This patch removes redundant autoconf check for the context diff program given it isn't actually been used anywhere, especially since make removal. - - - - - ea2fe66e by Hécate Moonlight at 2024-05-22T00:36:13-04:00 haddock: Rework the contributing guide - - - - - 0f302a94 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 haddock: Add module relationships diagrams of haddock-api and haddock-library - - - - - d1a9f34f by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add instructions - - - - - b880ee80 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add SVG outputs - - - - - 6d7e6ad8 by Ben Gamari at 2024-05-22T13:40:05-04:00 rts: Fix size of StgOrigThunkInfo frames Previously the entry code of the `stg_orig_thunk` frame failed to account for the size of the profiling header as it hard-coded the frame size. Fix this. Fixes #24809. - - - - - c645fe40 by Fendor at 2024-05-22T13:40:05-04:00 Add regression test T24809 for stg_orig_thunk_info_frame size - - - - - 4181aa40 by Andreas Klebinger at 2024-05-22T13:40:42-04:00 bindists: Check for existence of share folder before trying to copy it. This folder isn't distributed in windows bindists A lack of doing so resulted us copying loads of files twice. - - - - - d216510e by Matthew Pickering at 2024-05-22T13:40:42-04:00 Remove ad-hoc installation of mingw toolchain in relocatable bindists This reverts 616ac30026e8dd7d2ebb98d92dde071eedf5d951 The choice about whether to install mingw is taken in the installation makefile. This is also broken on non-windows systems. The actual issue was the EnableDistroToolchain variable wasn't declared in mk/config.mk and therefore the check to install mingw was failing. - - - - - 7b4c1998 by Cheng Shao at 2024-05-22T21:52:52-04:00 testsuite: fix T17920 for wasm backend T17920 was marked as fragile on wasm before; it can be trivially fixed by avoiding calling variadic printf() in cmm. - - - - - c739383b by Cheng Shao at 2024-05-22T21:53:29-04:00 testsuite: bump T22744 timeout to 5x - - - - - c4c6d714 by Cheng Shao at 2024-05-22T21:54:06-04:00 testsuite: don't attempt to detect host cpu features when testing cross ghc The testsuite driver CPU feature detection logic only detects host CPU and only makes sense when we are not testing a cross GHC. - - - - - 3d9e4ce6 by Simon Peyton Jones at 2024-05-22T21:54:43-04:00 Better skolemisation As #24810 showed, it is (a little) better to skolemise en-bloc, so that Note [Let-bound skolems] fires more often. See Note [Skolemisation en bloc] in GHC.Tc.Utils.Instantiate. - - - - - a3cd3a1d by Ryan Scott at 2024-05-22T21:55:19-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. - - - - - 4bb2a7cc by Hécate Moonlight at 2024-05-22T21:55:59-04:00 [base] Document the memory overhead of ByteArray Add a diagram that shows the constituent parts of a ByteArray and their memory overhead. - - - - - 8b2a016a by Hécate Moonlight at 2024-05-22T21:56:38-04:00 Haddock: Add MR template for Haddock - - - - - ead75532 by Peter Trommler at 2024-05-23T02:28:05-04:00 PPC: Support ELF v2 on powerpc64 big-endian Detect ELF v2 on PowerPC 64-bit systems. Check for `_CALL_ELF` preprocessor macro. Fixes #21191 - - - - - 9d4c10f2 by Hécate Kleidukos at 2024-05-23T02:28:44-04:00 gitlab: Add @Kleidukos to CODEOWNERS for utils/haddock - - - - - 28e64170 by Preetham Gujjula at 2024-05-23T07:20:48-04:00 haddock: Add cabal-fmt to tools for `make style` - - - - - 00126a89 by Andrei Borzenkov at 2024-05-23T07:21:24-04:00 haddock: fix verbosity option parsing - - - - - a3e0b68b by Ryan Hendrickson at 2024-05-23T15:52:03-04:00 base: specify tie-breaking behavior of min, max, and related list/Foldable functions - - - - - bdcc0f37 by doyougnu at 2024-05-24T07:51:18-04:00 cmm: add word <-> double/float bitcast - closes: #25331 This is the last step in the project plan described in #25331. This commit: - adds bitcast operands for x86_64, LLVM, aarch64 - For PPC and i386 we resort to using the cmm implementations - renames conversion MachOps from Conv to Round|Truncate - - - - - f0d257f7 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor Some functions in StgToByteCode were filtering out void arguments. However, StgToByteCode is called after unarisation: the void arguments should have been removed earlier. Instead of filtering out, we assert that the args are non-void. - - - - - 03137fd2 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor `layoutNativeCall` was always called with a `primRepCmmType platform` callback. Hence we can put it inside of `layoutNativeCall` rather than repeat it. - - - - - 27c430f3 by David Binder at 2024-05-24T07:52:38-04:00 haddock: Remove compatibility shims for GHC < 8.4 from haddock-library - - - - - 8dd8a076 by Cheng Shao at 2024-05-24T07:53:14-04:00 compiler: avoid saving foreign call target to local when there are no caller-save GlobalRegs This patch makes the STG->Cmm backend avoid saving foreign call target to local when there are no caller-save GlobalRegs. Since 321941a8ebe25192cdeece723e1058f2f47809ea, when we lower a foreign call, we unconditionally save the foreign call target to a temporary local first, then rely on cmmSink to clean it up later, which only happens with -fcmm-sink (implied by -O) and not in unoptimized code. And this is troublesome for the wasm backend NCG, which needs to infer a foreign call target symbol's type signature from the Cmm call site. Previously, the NCG has been emitting incorrect type signatures for unoptimized code, which happens to work with `wasm-ld` most of the time, but this is never future-proof against upstream toolchain updates, and it causes horrible breakages when LTO objects are included in linker input. Hence this patch. - - - - - 986df1ab by Cheng Shao at 2024-05-24T07:53:14-04:00 testsuite: add callee-no-local regression test - - - - - 52d62e2a by Sylvain Henry at 2024-05-24T07:53:57-04:00 Fix HasCallStack leftovers from !12514 / #24726 - - - - - c5e00c35 by crumbtoo at 2024-05-24T07:54:38-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 - - - - - bd323b0e by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Ensure that CHANGELOG is included in extra-source-files This was missed in the `ghc-internal` split. Closes #24831. - - - - - 1bfd32e8 by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Fix changelog reference to setBacktraceMechanismState (cherry picked from commit b63f7ba01fdfd98a01d2f0dec8d9262b3e595c5d) - - - - - 43e8e4f3 by Sylvain Henry at 2024-05-24T12:16:43-04:00 Float/double unboxed literal support for HexFloatLiterals (fix #22155) - - - - - 4a7f4713 by Fendor at 2024-05-24T12:17:19-04:00 Improve test labels for binary interface file size tests Test labels for binary interface file sizes are hard to read and overly verbose at the same time. Extend the name for the metric title, but shorten it in the actual comparison table. - - - - - 14e554cf by Zubin Duggal at 2024-05-24T12:17:55-04:00 Revert "Fix haskell/haddock#783 Don't show button if --quickjump not present" This reverts commit 7776566531e72c415f66dd3b13da9041c52076aa. - - - - - f56838c3 by Ben Gamari at 2024-05-24T12:17:55-04:00 Fix default hyperlinked sources pattern Previously this didn't include the `%M` token which manifested as broken links to the hyperlinked sources of reexports of declarations defined in other packages. Fixes haddock#1628. (cherry picked from commit 1432bcc943d41736eca491ecec4eb9a6304dab36) - - - - - 42efa62c by Ben Gamari at 2024-05-24T12:17:55-04:00 Make DocPaths a proper data type (cherry picked from commit 7f3a5c4da0023ae47b4c376c9b1ea2d706c94d8c) - - - - - 53d9ceb3 by Ben Gamari at 2024-05-24T12:17:55-04:00 haddock: Bump version to 2.30 (cherry picked from commit 994989ed3d535177e57b778629726aeabe8c7602) - - - - - e4db1112 by Zubin Duggal at 2024-05-24T12:17:55-04:00 haddock-api: allow base 4.20 and ghc 9.11 - - - - - e294f7a2 by PHO at 2024-05-24T12:17:55-04:00 Add a flag "threaded" for building haddock with the threaded RTS GHC isn't guaranteed to have a threaded RTS. There should be a way to build it with the vanilla one. (cherry picked from commit 75a94e010fb5b0236c670d22b04f5472397dc15d) - - - - - 51165bc9 by Andreas Klebinger at 2024-05-25T10:58:03-04:00 Update ticky counter event docs. Add the info about the info table address and json fields. Fixes #23200 - - - - - 98597ad5 by Sylvain Henry at 2024-05-25T10:58:45-04:00 Export extractPromotedList (#24866) This can be useful in plugins. - - - - - 228dcae6 by Teo Camarasu at 2024-05-28T13:12:24+00:00 template-haskell: Move wired-ins to ghc-internal Thus we make `template-haskell` reinstallable and keep it as the public API for Template Haskell. All of the wired-in identifiers are moved to `ghc-internal`. This necessitates also moving much of `ghc-boot-th` into `ghc-internal`. These modules are then re-exported from `ghc-boot-th` and `template-haskell`. To avoid a dependency on `template-haskell` from `lib:ghc`, we instead depend on the TH ASTs via `ghc-boot-th`. As `template-haskell` no longer has special status, we can drop the logic adding an implicit dependency on `template-haskell` when using TH. We can also drop the `template-haskell-next` package, which was previously used when bootstrapping. When bootstrapping, we need to vendor the TH AST modules from `ghc-internal` into `ghc-boot-th`. This is controlled by the `bootstrap` cabal flag as before. See Note [Bootstrapping Template Haskell]. We split out a GHC.Internal.TH.Lift module resolving #24752. This module is only built when not bootstrapping. Resolves #24703 ------------------------- Metric Increase: ghc_boot_th_dir ghc_boot_th_so ------------------------- - - - - - 62dded28 by Teo Camarasu at 2024-05-28T13:12:24+00:00 testsuite: mark tests broken by #24886 Now that `template-haskell` is no longer wired-in. These tests are triggering #24886, and so need to be marked broken. - - - - - 3ca72ad9 by Cheng Shao at 2024-05-30T02:57:06-04:00 rts: fix missing function prototypes in ClosureMacros.h - - - - - e0029e3d by Andreas Klebinger at 2024-05-30T02:57:43-04:00 UnliftedFFITypes: Allow `(# #)` as argument when it's the only argument. This allows representing functions like: int foo(void); to be imported like this: foreign import ccall "a_number_c" c_number :: (# #) -> Int64# Which can be useful when the imported function isn't implicitly stateful. - - - - - d0401335 by Matthew Pickering at 2024-05-30T02:58:19-04:00 ci: Update ci-images commit for fedora38 image The fedora38 nightly job has been failing for quite a while because `diff` was no longer installed. The ci-images bump explicitly installs `diffutils` into these images so hopefully they now pass again. - - - - - 3c97c74a by Jan Hrček at 2024-05-30T02:58:58-04:00 Update exactprint docs - - - - - 77760cd7 by Jan Hrček at 2024-05-30T02:58:58-04:00 Incorporate review feedback - - - - - 87591368 by Jan Hrček at 2024-05-30T02:58:58-04:00 Remove no longer relevant reference to comments - - - - - 05f4f142 by Jan Hrček at 2024-05-30T02:58:59-04:00 Replace outdated code example - - - - - 45a4a5f3 by Andreas Klebinger at 2024-05-30T02:59:34-04:00 Reword error resulting from missing -XBangPatterns. It can be the result of either a bang pattern or strict binding, so now we say so instead of claiming it must be a bang pattern. Fixes #21032 - - - - - e17f2df9 by Cheng Shao at 2024-05-30T03:00:10-04:00 testsuite: bump MultiLayerModulesDefsGhciReload timeout to 10x - - - - - 7a660042 by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - c77a48af by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 05c4fafb by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 5ff83bfc by Sylvain Henry at 2024-05-30T14:43:10-04:00 JS: remove useless h$CLOCK_REALTIME (#23202) - - - - - 95ef2d58 by Matthew Pickering at 2024-05-30T14:43:47-04:00 ghcup-metadata: Fix metadata generation There were some syntax errors in the generation script which were preventing it from running. I have tested this with: ``` nix shell --extra-experimental-features nix-command -f .gitlab/rel_eng -c ghcup-metadata --metadata ghcup-0.0.7.yaml --date="2024-05-27" --pipeline-id=95534 --version=9.11.20240525 ``` which completed successfully. - - - - - 1bc66ee4 by Jakob Bruenker at 2024-05-30T14:44:22-04:00 Add diagrams to Arrows documentation This adds diagrams to the documentation of Arrows, similar to the ones found on https://www.haskell.org/arrows/. It does not add diagrams for ArrowChoice for the time being, mainly because it's not clear to me how to visually distinguish them from the ones for Arrow. Ideally, you might want to do something like highlight the arrows belonging to the same tuple or same Either in common colors, but that's not really possible with unicode. - - - - - d10a1c65 by Matthew Craven at 2024-05-30T23:35:48-04:00 Make UnsafeSNat et al. into pattern synonyms ...so that they do not cause coerce to bypass the nominal role on the corresponding singleton types when they are imported. See Note [Preventing unsafe coercions for singleton types] and the discussion at #23478. This also introduces unsafeWithSNatCo (and analogues for Char and Symbol) so that users can still access the dangerous coercions that importing the real constructors would allow, but only in a very localized way. - - - - - 0958937e by Cheng Shao at 2024-05-30T23:36:25-04:00 hadrian: build C/C++ with split sections when enabled When split sections is enabled, ensure -fsplit-sections is passed to GHC as well when invoking GHC to compile C/C++; and pass -ffunction-sections -fdata-sections to gcc/clang when compiling C/C++ with the hadrian Cc builder. Fixes #23381. - - - - - 02b1f91e by Cheng Shao at 2024-05-30T23:36:25-04:00 driver: build C/C++ with -ffunction-sections -fdata-sections when split sections is enabled When -fsplit-sections is passed to GHC, pass -ffunction-sections -fdata-sections to gcc/clang when building C/C++. Previously, -fsplit-sections was only respected by the NCG/LLVM backends, but not the unregisterised backend; the GHC driver did not pass -fdata-sections and -ffunction-sections to the C compiler, which resulted in excessive executable sizes. Fixes #23381. ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - fd47e2e3 by Cheng Shao at 2024-05-30T23:37:00-04:00 testsuite: mark process005 as fragile on JS - - - - - 34a04ea1 by Matthew Pickering at 2024-05-31T06:08:36-04:00 Add -Wderiving-typeable to -Wall Deriving `Typeable` does nothing, and it hasn't done for a long while. There has also been a warning for a long while which warns you about uselessly deriving it but it wasn't enabled in -Wall. Fixes #24784 - - - - - 75fa7b0b by Matthew Pickering at 2024-05-31T06:08:36-04:00 docs: Fix formatting of changelog entries - - - - - 303c4b33 by Preetham Gujjula at 2024-05-31T06:09:21-04:00 docs: Fix link to injective type families paper Closes #24863 - - - - - df97e9a6 by Ben Gamari at 2024-05-31T06:09:57-04:00 ghc-internal: Fix package description The previous description was inherited from `base` and was inappropriate for `ghc-internal`. Also fix the maintainer and bug reporting fields. Closes #24906. - - - - - bf0737c0 by Cheng Shao at 2024-05-31T06:10:33-04:00 compiler: remove ArchWasm32 special case in cmmDoCmmSwitchPlans This patch removes special consideration for ArchWasm32 in cmmDoCmmSwitchPlans, which means the compiler will now disable cmmImplementSwitchPlans for wasm unreg backend, just like unreg backend of other targets. We enabled it in the past to workaround some compile-time panic in older versions of LLVM, but those panics are no longer present, hence no need to keep this workaround. - - - - - 7eda4bd2 by Cheng Shao at 2024-05-31T15:52:04-04:00 utils: add hie.yaml config file for ghc-config Add hie.yaml to ghc-config project directory so it can be edited using HLS. - - - - - 1e5752f6 by Cheng Shao at 2024-05-31T15:52:05-04:00 hadrian: handle findExecutable "" gracefully hadrian may invoke findExecutable "" at run-time due to a certain program is not found by configure script. Which is fine and findExecutable is supposed to return Nothing in this case. However, on Windows there's a directory bug that throws an exception (see https://github.com/haskell/directory/issues/180), so we might as well use a wrapper for findExecutable and handle exceptions gracefully. - - - - - 4eb5ad09 by Cheng Shao at 2024-05-31T15:52:05-04:00 configure: do not set LLC/OPT/LLVMAS fallback values when FIND_LLVM_PROG fails When configure fails to find LLC/OPT/LLVMAS within supported version range, it used to set "llc"/"opt"/"clang" as fallback values. This behavior is particularly troublesome when the user has llc/opt/clang with other versions in their PATH and run the testsuite, since hadrian will incorrectly assume have_llvm=True and pass that to the testsuite driver, resulting in annoying optllvm test failures (#23186). If configure determines llc/opt/clang wouldn't work, then we shouldn't pretend it'll work at all, and the bindist configure will invoke FIND_LLVM_PROG check again at install time anyway. - - - - - 5f1afdf7 by Sylvain Henry at 2024-05-31T15:52:52-04:00 Introduce UniqueSet and use it to replace 'UniqSet Unique' 'UniqSet Unique' represents a set of uniques as a 'Map Unique Unique', which is wasting space (associated key/value are always the same). Fix #23572 and #23605 - - - - - e0aa42b9 by crumbtoo at 2024-05-31T15:53:33-04:00 Improve template-haskell haddocks Closes #15822 - - - - - ae170155 by Olivier Benz at 2024-06-01T09:35:17-04:00 Bump max LLVM version to 19 (not inclusive) - - - - - 92aa65ea by Matthew Pickering at 2024-06-01T09:35:17-04:00 ci: Update CI images to test LLVM 18 The debian12 image in this commit has llvm 18 installed. - - - - - adb1fe42 by Serge S. Gulin at 2024-06-01T09:35:53-04:00 Unicode: make ucd2haskell build-able again ucd2haskell tool used streamly library which version in cabal was out of date. It is updated to the latest version at hackage with deprecated parts rewritten. Also following fixes were applied to existing code in suppose that from its last run the code structure was changed and now it was required to be up to date with actual folder structures: 1. Ghc module path environment got a suffix with `src`. 2. Generated code got 2.1 `GHC.Internal` prefix for `Data.*`. 2.2 `GHC.Unicode.Internal` swapped on `GHC.Internal.Unicode` according to actual structure. - - - - - ad56fd84 by Jade at 2024-06-01T09:36:29-04:00 Replace 'NB' with 'Note' in error messages - - - - - 6346c669 by Cheng Shao at 2024-06-01T09:37:04-04:00 compiler: fix -ddump-cmm-raw when compiling .cmm This patch fixes missing -ddump-cmm-raw output when compiling .cmm, which is useful for debugging cmm related codegen issues. - - - - - 1c834ad4 by Ryan Scott at 2024-06-01T09:37:40-04:00 Print namespace specifiers in FixitySig's Outputable instance For whatever reason, the `Outputable` instance for `FixitySig` simply did not print out namespace specifiers, leading to the confusing `-ddump-splices` output seen in #24911. This patch corrects this oversight. Fixes #24911. - - - - - cf49fb5f by Sylvain Henry at 2024-06-01T09:38:19-04:00 Configure: display C++ compiler path - - - - - f9c1ae12 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable PIC for in-tree GMP on wasm32 This patch disables PIC for in-tree GMP on wasm32 target. Enabling PIC unconditionally adds undesired code size and runtime overhead for wasm32. - - - - - 1a32f828 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable in-tree gmp fft code path for wasm32 This patch disables in-tree GMP FFT code paths for wasm32 target in order to give up some performance of multiplying very large operands in exchange for reduced code size. - - - - - 06277d56 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: build in-tree GMP with malloc-notreentrant on wasm32 This patch makes hadrian build in-tree GMP with the --enable-alloca=malloc-notreentrant configure option. We will only need malloc-reentrant when we have threaded RTS and SMP support on wasm32, which will take some time to happen, before which we should use malloc-notreentrant to avoid undesired runtime overhead. - - - - - 9f614270 by ARATA Mizuki at 2024-06-02T14:02:35-04:00 Set package include paths when assembling .S files Fixes #24839. Co-authored-by: Sylvain Henry <hsyl20 at gmail.com> - - - - - 4998a6ed by Alex Mason at 2024-06-03T02:09:29-04:00 Improve performance of genericWordQuotRem2Op (#22966) Implements the algorithm from compiler-rt's udiv128by64to64default. This rewrite results in a roughly 24x improvement in runtime on AArch64 (and likely any other arch that uses it). - - - - - ae50a8eb by Cheng Shao at 2024-06-03T02:10:05-04:00 testsuite: mark T7773 as fragile on wasm - - - - - c8ece0df by Fendor at 2024-06-03T19:43:22-04:00 Migrate `Finder` component to `OsPath`, fixed #24616 For each module in a GHCi session, we keep alive one `ModLocation`. A `ModLocation` is fairly inefficiently packed, as `String`s are expensive in memory usage. While benchmarking the agda codebase, we concluded that we keep alive around 11MB of `FilePath`'s, solely retained by `ModLocation`. We provide a more densely packed encoding of `ModLocation`, by moving from `FilePath` to `OsPath`. Further, we migrate the full `Finder` component to `OsPath` to avoid unnecessary transformations. As the `Finder` component is well-encapsulated, this requires only a minimal amount of changes in other modules. We introduce pattern synonym for 'ModLocation' which maintains backwards compatibility and avoids breaking consumers of 'ModLocation'. - - - - - 0cff083a by Cheng Shao at 2024-06-03T19:43:58-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. - - - - - 18f63970 by Sebastian Graf at 2024-06-04T05:05:27-04:00 Parser: Remove unused `apats` rule - - - - - 38757c30 by David Knothe at 2024-06-04T05:05:27-04:00 Implement Or Patterns (#22596) This commit introduces a new language extension, `-XOrPatterns`, as described in GHC Proposal 522. An or-pattern `pat1; ...; patk` succeeds iff one of the patterns `pat1`, ..., `patk` succeed, in this order. See also the summary `Note [Implmentation of OrPatterns]`. Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> - - - - - 395412e8 by Cheng Shao at 2024-06-04T05:06:04-04:00 compiler/ghci/rts: remove stdcall support completely We have formally dropped i386 windows support (#18487) a long time ago. The stdcall foreign call convention is only used by i386 windows, and the legacy logic around it is a significant maintenance burden for future work that adds arm64 windows support (#24603). Therefore, this patch removes stdcall support completely from the compiler as well as the RTS (#24883): - stdcall is still recognized as a FFI calling convention in Haskell syntax. GHC will now unconditionally emit a warning (-Wunsupported-calling-conventions) and treat it as ccall. - Apart from minimum logic to support the parsing and warning logic, all other code paths related to stdcall has been completely stripped from the compiler. - ghci only supports FFI_DEFAULT_ABI and ccall convention from now on. - FFI foreign export adjustor code on all platforms no longer handles the stdcall case and only handles ccall from now on. - The Win32 specific parts of RTS no longer has special code paths for stdcall. This commit is the final nail on the coffin for i386 windows support. Further commits will perform more housecleaning to strip the legacy code paths and pave way for future arm64 windows support. - - - - - d1fe9ab6 by Cheng Shao at 2024-06-04T05:06:04-04:00 rts: remove legacy i386 windows code paths This commit removes some legacy i386 windows related code paths in the RTS, given this target is no longer supported. - - - - - a605e4b2 by Cheng Shao at 2024-06-04T05:06:04-04:00 autoconf: remove i386 windows related logic This commit removes legacy i386 windows logic in autoconf scripts. - - - - - 91e5ac5e by Cheng Shao at 2024-06-04T05:06:04-04:00 llvm-targets: remove i386 windows support This commit removes i386 windows from llvm-targets and the script to generate it. - - - - - 65fe75a4 by Cheng Shao at 2024-06-04T05:06:04-04:00 libraries/utils: remove stdcall related legacy logic This commit removes stdcall related legacy logic in libraries and utils. ccall should be used uniformly for all supported windows hosts from now on. - - - - - d2a83302 by Cheng Shao at 2024-06-04T05:06:04-04:00 testsuite: adapt the testsuite for stdcall removal This patch adjusts test cases to handle the stdcall removal: - Some stdcall usages are replaced with ccall since stdcall doesn't make sense anymore. - We also preserve some stdcall usages, and check in the expected warning messages to ensure GHC always warn about stdcall usages (-Wunsupported-calling-conventions) as expected. - Error code testsuite coverage is slightly improved, -Wunsupported-calling-conventions is now tested. - Obsolete code paths related to i386 windows are also removed. - - - - - cef8f47a by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: minor adjustments for stdcall removal This commit include minor adjustments of documentation related to stdcall removal. - - - - - 54332437 by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: mention i386 Windows removal in 9.12 changelog This commit mentions removal of i386 Windows support and stdcall related change in the 9.12 changelog. - - - - - 2aaea8a1 by Cheng Shao at 2024-06-04T05:06:40-04:00 hadrian: improve user settings documentation This patch adds minor improvements to hadrian user settings documentation: - Add missing `ghc.cpp.opts` case - Remove non-existent `cxx` case - Clarify `cc.c.opts` also works for C++, while `cc.deps.opts` doesn't - Add example of passing configure argument to autoconf packages - - - - - 71010381 by Alex Mason at 2024-06-04T12:09:07-04:00 Add AArch64 CLZ, CTZ, RBIT primop implementations. Adds support for emitting the clz and rbit instructions, which are used by GHC.Prim.clz*#, GHC.Prim.ctz*# and GHC.Prim.bitReverse*#. - - - - - 44e2abfb by Cheng Shao at 2024-06-04T12:09:43-04:00 hadrian: add +text_simdutf flavour transformer to allow building text with simdutf This patch adds a +text_simdutf flavour transformer to hadrian to allow downstream packagers and users that build from source to opt-in simdutf support for text, in order to benefit from SIMD speedup at run-time. It's still disabled by default for the time being. - - - - - 077cb2e1 by Cheng Shao at 2024-06-04T12:09:43-04:00 ci: enable +text_simdutf flavour transformer for wasm jobs This commit enables +text_simdutf flavour transformer for wasm jobs, so text is now built with simdutf support for wasm. - - - - - b23746ad by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Use TemplateHaskellQuotes in instance Lift ByteArray Resolves #24852 - - - - - 3fd25743 by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Mark addrToByteArray as NOINLINE This function should never be inlined in order to keep code size small. - - - - - 98ad1ea5 by Cheng Shao at 2024-06-04T22:51:26-04:00 compiler: remove unused CompilerInfo/LinkerInfo types This patch removes CompilerInfo/LinkerInfo types from the compiler since they aren't actually used anywhere. - - - - - 11795244 by Cheng Shao at 2024-06-05T06:33:17-04:00 rts: remove unused PowerPC/IA64 native adjustor code This commit removes unused PowerPC/IA64 native adjustor code which is never actually enabled by autoconf/hadrian. Fixes #24920. - - - - - 5132754b by Sylvain Henry at 2024-06-05T06:33:57-04:00 RTS: fix warnings with doing*Profiling (#24918) - - - - - accc8c33 by Cheng Shao at 2024-06-05T11:35:36-04:00 hadrian: don't depend on inplace/mingw when --enable-distro-toolchain on Windows - - - - - 6ffbd678 by Cheng Shao at 2024-06-05T11:35:37-04:00 autoconf: normalize paths of some build-time dependencies on Windows This commit applies path normalization via cygpath -m to some build-time dependencies on Windows. Without this logic, the /clang64/bin prefixed msys2-style paths cause the build to fail with --enable-distro-toolchain. - - - - - 075dc6d4 by Cheng Shao at 2024-06-05T11:36:12-04:00 hadrian: remove OSDarwin mention from speedHack This commit removes mentioning of OSDarwin from speedHack, since speedHack is purely for i386 and we no longer support i386 darwin (#24921). - - - - - 83235c4c by Cheng Shao at 2024-06-05T11:36:12-04:00 compiler: remove 32-bit darwin logic This commit removes all 32-bit darwin logic from the compiler, given we no longer support 32-bit apple systems (#24921). Also contains a bit more cleanup of obsolete i386 windows logic. - - - - - 1eb99bc3 by Cheng Shao at 2024-06-05T11:36:12-04:00 rts: remove 32-bit darwin/ios logic This commit removes 32-bit darwin/ios related logic from the rts, given we no longer support them (#24921). - - - - - 24f65892 by Cheng Shao at 2024-06-05T11:36:12-04:00 llvm-targets: remove 32-bit darwin/ios targets This commit removes 32-bit darwin/ios targets from llvm-targets given we no longer support them (#24921). - - - - - ccdbd689 by Cheng Shao at 2024-06-05T11:36:12-04:00 testsuite: remove 32-bit darwin logic This commit removes 32-bit darwin logic from the testsuite given it's no longer supported (#24921). Also contains more cleanup of obsolete i386 windows logic. - - - - - 11d661c4 by Cheng Shao at 2024-06-05T11:36:13-04:00 docs: mention 32-bit darwin/ios removal in 9.12 changelog This commit mentions removal of 32-bit darwin/ios support (#24921) in the 9.12 changelog. - - - - - 7c173310 by Georgi Lyubenov at 2024-06-05T15:17:22-04:00 Add firstA and secondA to Data.Bitraversable Please see https://github.com/haskell/core-libraries-committee/issues/172 for related discussion - - - - - 3b6f9fd1 by Ben Gamari at 2024-06-05T15:17:59-04:00 base: Fix name of changelog Fixes #24899. Also place it under `extra-doc-files` to better reflect its nature and avoid triggering unnecessary recompilation if it changes. - - - - - 1f4d2ef7 by Sebastian Graf at 2024-06-05T15:18:34-04:00 Announce Or-patterns in the release notes for GHC 9.12 (#22596) Leftover from !9229. - - - - - 8650338d by Jan Hrček at 2024-06-06T10:39:24-04:00 Improve haddocks of Language.Haskell.Syntax.Pat.Pat - - - - - 2eee65e1 by Cheng Shao at 2024-06-06T10:40:00-04:00 testsuite: bump T7653 timeout for wasm - - - - - 990fed60 by Sylvain Henry at 2024-06-07T14:45:23-04:00 StgToCmm: refactor opTranslate and friends - Change arguments order to avoid `\args -> ...` lambdas - Fix documentation - Rename StgToCmm options ("big" doesn't mean anything) - - - - - 1afad514 by Sylvain Henry at 2024-06-07T14:45:23-04:00 NCG x86: remove dead code (#5444) Since 6755d833af8c21bbad6585144b10e20ac4a0a1ab this code is dead. - - - - - 595c0894 by Cheng Shao at 2024-06-07T14:45:58-04:00 testsuite: skip objc-hi/objcxx-hi when cross compiling objc-hi/objcxx-hi should be skipped when cross compiling. The existing opsys('darwin') predicate only asserts the host system is darwin but tells us nothing about the target, hence the oversight. - - - - - edfe6140 by qqwy at 2024-06-08T11:23:54-04:00 Replace '?callStack' implicit param with HasCallStack in GHC.Internal.Exception.throw - - - - - 35a64220 by Cheng Shao at 2024-06-08T11:24:30-04:00 rts: cleanup inlining logic This patch removes pre-C11 legacy code paths related to INLINE_HEADER/STATIC_INLINE/EXTERN_INLINE macros, ensure EXTERN_INLINE is treated as static inline in most cases (fixes #24945), and also corrects the comments accordingly. - - - - - 9ea90ed2 by Andrew Lelechenko at 2024-06-08T11:25:06-04:00 CODEOWNERS: add @core-libraries to track base interface changes A low-tech tactical solution for #24919 - - - - - 580fef7b by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update CHANGELOG to reflect current version - - - - - 391ecff5 by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update prologue.txt to reflect package description - - - - - 3dca3b7d by Ben Gamari at 2024-06-09T01:27:57-04:00 compiler: Clarify comment regarding need for MOVABS The comment wasn't clear in stating that it was only applicable to immediate source and memory target operands. - - - - - 6bd850e8 by doyougnu at 2024-06-09T21:02:14-04:00 JS: establish single source of truth for symbols In pursuit of: #22736. This MR moves ad-hoc symbols used throughout the js backend into a single symbols file. Why? First, this cleans up the code by removing ad-hoc strings created on the fly and therefore makes the code more maintainable. Second, it makes it much easier to eventually type these identifiers. - - - - - f3017dd3 by Cheng Shao at 2024-06-09T21:02:49-04:00 rts: replace ad-hoc MYTASK_USE_TLV with proper CC_SUPPORTS_TLS This patch replaces the ad-hoc `MYTASK_USE_TLV` with the `CC_SUPPORTS_TLS` macro. If TLS support is detected by autoconf, then we should use that for managing `myTask` in the threaded RTS. - - - - - e17d7e8c by Ben Gamari at 2024-06-11T05:25:21-04:00 users-guide: Fix stylistic issues in 9.12 release notes - - - - - 8a8a982a by Hugo Peters at 2024-06-11T05:25:57-04:00 fix typo in the simplifier debug output: baling -> bailing - - - - - 16475bb8 by Hécate Moonlight at 2024-06-12T03:07:55-04:00 haddock: Correct the Makefile to take into account Darwin systems - - - - - a2f60da5 by Hécate Kleidukos at 2024-06-12T03:08:35-04:00 haddock: Remove obsolete links to github.com/haskell/haddock in the docs - - - - - de4395cd by qqwy at 2024-06-12T03:09:12-04:00 Add `__GLASGOW_HASKELL_ASSERTS_IGNORED__` as CPP macro name if `-fasserts-ignored is set. This allows users to create their own Control.Exception.assert-like functionality that does something other than raising an `AssertFailed` exception. Fixes #24967 - - - - - 0e9c4dee by Ryan Hendrickson at 2024-06-12T03:09:53-04:00 compiler: add hint to TcRnBadlyStaged message - - - - - 2747cd34 by Simon Peyton Jones at 2024-06-12T12:51:37-04:00 Fix a QuickLook bug This MR fixes the bug exposed by #24676. The problem was that quickLookArg was trying to avoid calling tcInstFun unnecessarily; but it was in fact necessary. But that in turn forced me into a significant refactoring, putting more fields into EValArgQL. Highlights: see Note [Quick Look overview] in GHC.Tc.Gen.App * Instantiation variables are now distinguishable from ordinary unification variables, by level number = QLInstVar. This is treated like "level infinity". See Note [The QLInstVar TcLevel] in GHC.Tc.Utils.TcType. * In `tcApp`, we don't track the instantiation variables in a set Delta any more; instead, we just tell them apart by their level number. * EValArgQL now much more clearly captures the "half-done" state of typechecking an argument, ready for later resumption. See Note [Quick Look at value arguments] in GHC.Tc.Gen.App * Elminated a bogus (never used) fast-path in GHC.Tc.Utils.Instantiate.instCallConstraints See Note [Possible fast path for equality constraints] Many other small refactorings. - - - - - 1b1523b1 by George Thomas at 2024-06-12T12:52:18-04:00 Fix non-compiling extensible record `HasField` example - - - - - 97b141a3 by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Fix hyperlinker source urls (#24907) This fixes a bug introduced by f56838c36235febb224107fa62334ebfe9941aba Links to external modules in the hyperlinker are uniformly generated using splicing the template given to us instead of attempting to construct the url in an ad-hoc manner. - - - - - 954f864c by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Add name anchor to external source urls from documentation page URLs for external source links from documentation pages were missing a splice location for the name. Fixes #24912 - - - - - b0b64177 by Simon Peyton Jones at 2024-06-12T12:53:31-04:00 Prioritise nominal equalities The main payload of this patch is * Prioritise nominal equalities in the constraint solver. This ameliorates the incompleteness of solving for representational constraints over newtypes: see #24887. See (EX2) in Note [Decomposing newtype equalities] in GHC.Tc.Solver.Equality In doing this patch I tripped over some other things that I refactored: * Move `isCoVarType` from `GHC.Core.Type` to `GHC.Core.Predicate` where it seems more at home. * Clarify the "rewrite role" of a constraint. I was very puzzled about what the role of, say `(Eq a)` might be, but see the new Note [The rewrite-role of a constraint]. In doing so I made predTypeEqRel crash when given a non-equality. Usually it expects an equality; but it was being mis-used for the above rewrite-role stuff. - - - - - cb7c1b83 by Liam Goodacre at 2024-06-12T12:54:09-04:00 compiler: missing-deriving-strategies suggested fix Extends the missing-deriving-strategies warning with a suggested fix that includes which deriving strategies were assumed. For info about the warning, see comments for `TcRnNoDerivStratSpecified`, `TcRnNoDerivingClauseStrategySpecified`, & `TcRnNoStandaloneDerivingStrategySpecified`. For info about the suggested fix, see `SuggestExplicitDerivingClauseStrategies` & `SuggestExplicitStandalanoDerivingStrategy`. docs: Rewords missing-deriving-strategies to mention the suggested fix. Resolves #24955 - - - - - 4e36d3a3 by Jan Hrček at 2024-06-12T12:54:48-04:00 Further haddocks improvements in Language.Haskell.Syntax.Pat.Pat - - - - - 558353f4 by Cheng Shao at 2024-06-12T12:55:24-04:00 rts: use page sized mblocks on wasm This patch changes mblock size to page size on wasm. It allows us to simplify our wasi-libc fork, makes it much easier to test third party libc allocators like emmalloc/mimalloc, as well as experimenting with threaded RTS in wasm. - - - - - b3cc5366 by Matthew Pickering at 2024-06-12T23:06:57-04:00 compiler: Make ghc-experimental not wired in If you need to wire in definitions, then place them in ghc-internal and reexport them from ghc-experimental. Ticket #24903 - - - - - 700eeab9 by Hécate Kleidukos at 2024-06-12T23:07:37-04:00 base: Use a more appropriate unicode arrow for the ByteArray diagram This commit rectifies the usage of a unicode arrow in favour of one that doesn't provoke mis-alignment. - - - - - cca7de25 by Matthew Pickering at 2024-06-12T23:08:14-04:00 ghcup-metadata: Fix debian version ranges This was caught by `ghcup-ci` failing and attempting to install a deb12 bindist on deb11. ``` configure: WARNING: m4/prep_target_file.m4: Expecting YES/NO but got in ArSupportsDashL_STAGE0. Defaulting to False. bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by bin/ghc-toolchain-bin) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) ``` Fixes #24974 - - - - - 7b23ce8b by Pierre Le Marre at 2024-06-13T15:35:04-04:00 ucd2haskell: remove Streamly dependency + misc - Remove dead code. - Remove `streamly` dependency. - Process files with `bytestring`. - Replace Unicode files parsers with the corresponding ones from the package `unicode-data-parser`. - Simplify cabal file and rename module - Regenerate `ghc-internal` Unicode files with new header - - - - - 4570319f by Jacco Krijnen at 2024-06-13T15:35:41-04:00 Document how to run haddocks tests (#24976) Also remove ghc 9.7 requirement - - - - - fb629e24 by amesgen at 2024-06-14T00:28:20-04:00 compiler: refactor lower_CmmExpr_Ptr - - - - - def46c8c by amesgen at 2024-06-14T00:28:20-04:00 compiler: handle CmmRegOff in lower_CmmExpr_Ptr - - - - - ce76bf78 by Simon Peyton Jones at 2024-06-14T00:28:56-04:00 Small documentation update in Quick Look - - - - - 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - 7e56649e by Ben Gamari at 2024-09-18T17:26:57-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes #24807. - - - - - 20 changed files: - + .git-blame-ignore-revs - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - + .gitlab/merge_request_templates/Haddock.md - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC.hs - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/PrimOps.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a7237b11ebbc327e123d43b6e0e304b16cc746ed...7e56649e56d0ba75b7bf88a73f71d7e11f2217ac -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a7237b11ebbc327e123d43b6e0e304b16cc746ed...7e56649e56d0ba75b7bf88a73f71d7e11f2217ac You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 02:13:10 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Wed, 18 Sep 2024 22:13:10 -0400 Subject: [Git][ghc/ghc][wip/strings] Simplify multiline string lexing Message-ID: <66eb88b68be86_3ac2741f285041494d6@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC Commits: da077a6d by Brandon Chinn at 2024-09-18T19:12:58-07:00 Simplify multiline string lexing - - - - - 1 changed file: - compiler/GHC/Parser/Lexer.x Changes: ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -692,17 +692,11 @@ $unigraphic / { isSmartQuote } { smart_quote_error } \" @stringchar* $unigraphic / { isSmartQuote } { smart_quote_error } } - { - ([\ $tab] | @gap)* { tok_string_multi_content } -} - { - @stringchar* { tok_string_multi_content } - $nl { tok_string_multi_content } + -- Parse as much of the multiline string as possible, except for quotes + @stringchar* ($nl ([\ $tab] | @gap)* @stringchar*)* { tok_string_multi_content } -- Allow bare quotes if it's not a triple quote - -- N.B. We need to explicitly check for \n in the right context because - -- the character set [^...] doesn't include newlines - (\" | \"\") / (\n | [^\"]) { tok_string_multi_content } + (\" | \"\") / ([\n .] # \") { tok_string_multi_content } } <0> { @@ -2277,8 +2271,6 @@ tok_string_multi startSpan startBuf _len _buf2 = do case alexScan i0 string_multi_content of AlexToken i1 len _ | Just i2 <- lexDelim i1 -> pure (i1, i2) - | -- did we just lex a newline? - Just ('\n', _) <- alexGetChar' i0 -> goBOL i1 | -- is the next token a tab character? -- need this explicitly because there's a global rule matching $tab Just ('\t', _) <- alexGetChar' i1 -> setInput i1 >> lexError LexError @@ -2288,12 +2280,6 @@ tok_string_multi startSpan startBuf _len _buf2 = do AlexSkip i1 _ -> goContent i1 _ -> lexError LexError - goBOL i0 = - case alexScan i0 string_multi_bol of - AlexToken i1 _ _ -> goContent i1 - AlexSkip i1 _ -> goBOL i1 - _ -> lexError LexError - lexDelim = let go 0 i = Just i go n i = View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/da077a6d96cd08c140c0ff2d3b81348fd5be07e1 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/da077a6d96cd08c140c0ff2d3b81348fd5be07e1 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 07:13:13 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Thu, 19 Sep 2024 03:13:13 -0400 Subject: [Git][ghc/ghc][wip/T23479] 2 commits: Fix build Message-ID: <66ebcf098e634_3d7b3ccc7f4590e7@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: a0f3672e by Serge S. Gulin at 2024-09-19T10:11:13+03:00 Fix build - - - - - 91765d8e by Serge S. Gulin at 2024-09-19T10:12:46+03:00 String lits were removed too early, need do it at Linker instead - - - - - 1 changed file: - compiler/GHC/StgToJS/Sinker/Sinker.hs Changes: ===================================== compiler/GHC/StgToJS/Sinker/Sinker.hs ===================================== @@ -23,11 +23,11 @@ import GHC.StgToJS.Utils import Data.Char import Data.List (partition) import Data.Maybe -import Data.Bifunctor (Bifunctor (..)) -import GHC.Utils.Outputable (showSDocUnsafe, showPprUnsafe) -import qualified GHC.Utils.Trace as DT import Data.ByteString (ByteString) +import GHC.Utils.Outputable (showSDocUnsafe, showPprUnsafe) +import qualified GHC.Utils.Trace as DT_GHC + -- | Unfloat some top-level unexported things -- @@ -54,18 +54,13 @@ sinkPgm m pgm usedOnceIds = selectUsedOnce $ concatMap collectArgs pgm' stringLitsUFM = listToUFM $ (\(i, b) -> (idName i, (i, b))) <$> allStringLits - (pgm'', usedStringLitNames) = + (pgm'', _actuallyUnfloatedStringLitNames) = unfloatStringLits (idName `mapUniqSet` usedOnceIds) (snd `mapUFM` stringLitsUFM) pgm' - stringLitsWithoutUnfolded = - nonDetEltsUFM - $ delListFromUFM stringLitsUFM - $ nonDetEltsUniqSet usedStringLitNames - - stringLits = uncurry StgTopStringLit <$> stringLitsWithoutUnfolded + stringLits = uncurry StgTopStringLit <$> allStringLits (sunk, pgm''') = sinkPgm' m usedOnceIds $ traceDoc "" (map StgTopLifted) pgm'' @@ -154,5 +149,5 @@ topSortDecls _m binds = rest ++ nr' | otherwise = map node_payload (topologicalSortG g) -traceDoc head m v = (DT.trace . (++) head . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (m v) v -tracePpr head m v = (DT.trace . (++) head . showPprUnsafe) (m v) v +traceDoc head m v = (DT_GHC.trace . (++) head . showSDocUnsafe . pprGenStgTopBindings shortStgPprOpts) (m v) v +tracePpr head m v = (DT_GHC.pprTraceIt . (++) head . showPprUnsafe) (m v) v View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/251595d8a2cf80a9655f4e3e9d1a49e5434f7656...91765d8ecc525c1bf17e0be6b008d6b9af184501 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/251595d8a2cf80a9655f4e3e9d1a49e5434f7656...91765d8ecc525c1bf17e0be6b008d6b9af184501 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 08:43:25 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Thu, 19 Sep 2024 04:43:25 -0400 Subject: [Git][ghc/ghc][wip/25232] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66ebe42d60313_2ab0e810b4cc99417@gitlab.mail> Zubin pushed to branch wip/25232 at Glasgow Haskell Compiler / GHC Commits: 382c79e8 by Zubin Duggal at 2024-09-19T14:13:15+05:30 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 5 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - + testsuite/tests/rts/T25232.hs - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,18 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + // Not doing this lead to #25232 and #23958 + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +958,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): ===================================== testsuite/tests/rts/T25232.hs ===================================== @@ -0,0 +1,14 @@ +module Main where + +import Control.Monad +import Control.Monad.Trans.State.Strict +import Data.IORef + +main :: IO () +main = do + ref <- newIORef () + replicateM_ 1000 $ withIORef ref $ runStateT (pure ()) + +withIORef :: IORef a -> (a -> IO (b, a)) -> IO b +withIORef ref f = + readIORef ref >>= f >>= \(b, a) -> writeIORef ref a >> pure b ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [req_profiling, extra_ways(['nonmoving', 'nonmoving_prof', 'nonmoving_thr_prof'])], compile_and_run, []) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/382c79e8abb6be01ab69aee91291032ae5497185 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/382c79e8abb6be01ab69aee91291032ae5497185 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 10:00:23 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 19 Sep 2024 06:00:23 -0400 Subject: [Git][ghc/ghc][wip/T24807] base: Propagate `error` CallStack to thrown exception Message-ID: <66ebf63736580_2ab0e854c770115975@gitlab.mail> Rodrigo Mesquita pushed to branch wip/T24807 at Glasgow Haskell Compiler / GHC Commits: 5963b8c2 by Ben Gamari at 2024-09-19T11:00:10+01:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes #24807. - - - - - 30 changed files: - libraries/base/tests/T19288.stderr - + libraries/base/tests/T24807.hs - + libraries/base/tests/T24807.stderr - libraries/base/tests/all.T - libraries/ghc-internal/src/GHC/Internal/Exception.hs - testsuite/tests/arityanal/should_run/T21694a.stderr - testsuite/tests/arityanal/should_run/T24296.stderr - testsuite/tests/codeGen/should_run/T16846.stderr - testsuite/tests/codeGen/should_run/T5626.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/codeGen/should_run/cgrun045.stderr - testsuite/tests/codeGen/should_run/cgrun051.stderr - testsuite/tests/codeGen/should_run/cgrun059.stderr - testsuite/tests/concurrent/should_run/T13330.stderr - testsuite/tests/concurrent/should_run/conc021.stderr - testsuite/tests/deSugar/should_run/Or5.stderr - testsuite/tests/deSugar/should_run/T11193.stderr - testsuite/tests/deSugar/should_run/T11572.stderr - testsuite/tests/deSugar/should_run/T11601.stderr - testsuite/tests/dmdanal/should_run/T12368.stderr - testsuite/tests/dmdanal/should_run/T12368a.stderr - testsuite/tests/dmdanal/should_run/T23208.stderr - testsuite/tests/dmdanal/should_run/strun002.stderr - testsuite/tests/driver/multipleHomeUnits/multipleHomeUnits_callstack.stderr - testsuite/tests/ffi/should_run/ffi008.stderr - testsuite/tests/hpc/simple/tixs/T10529a.stderr - testsuite/tests/hpc/simple/tixs/T10529b.stderr - testsuite/tests/hpc/simple/tixs/T10529c.stderr - testsuite/tests/rebindable/RebindableFailA.stderr - testsuite/tests/safeHaskell/safeLanguage/SafeLang09.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5963b8c2de54d397f1c860e011951109f9c95163 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5963b8c2de54d397f1c860e011951109f9c95163 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 10:08:33 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 19 Sep 2024 06:08:33 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 14 commits: configure: Correctly report when subsections-via-symbols is disabled Message-ID: <66ebf820ec1e6_2ab0e8544e08122328@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - c5ce2b8e by Matthew Pickering at 2024-09-18T16:28:36+01:00 ci: Run abi-test on test-abi label - - - - - db862ba8 by Rodrigo Mesquita at 2024-09-18T16:28:37+01:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - e8881e42 by Rodrigo Mesquita at 2024-09-18T16:28:37+01:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - f0641005 by Rodrigo Mesquita at 2024-09-18T18:09:53+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 0a5c2dac by Rodrigo Mesquita at 2024-09-18T19:16:08+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 475d1f2e by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 5af98c82 by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - ed43ad42 by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 98472267 by Sylvain Henry at 2024-09-19T06:08:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - 3ad18ccf by Sylvain Henry at 2024-09-19T06:08:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - e1bfbb7a by Sven Tennie at 2024-09-19T06:08:25-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/RV64/CodeGen.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c6e6749b2b6b984dba7a707f4a1c15970763fc9e...e1bfbb7a68249d2174c84280215e4cc4e562435f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c6e6749b2b6b984dba7a707f4a1c15970763fc9e...e1bfbb7a68249d2174c84280215e4cc4e562435f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:13:32 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:13:32 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] 12 commits: Linker: add support for extra built-in symbols (#25155) Message-ID: <66ec075cd2ed2_530bb3e7a4c10157c@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 083d2872 by Sven Tennie at 2024-09-19T13:13:16+02:00 WIP: RISCV64 cross-compile testing - - - - - 73becc23 by Sven Tennie at 2024-09-19T13:13:16+02:00 Patch job to see some action - - - - - 59567e6c by Sven Tennie at 2024-09-19T13:13:16+02:00 Use latest supersven image The image is WIP, too. - - - - - f370062a by Sven Tennie at 2024-09-19T13:13:16+02:00 Next round - - - - - 1 changed file: - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ba87f1a04e2540c12d8030df5c22d36480370d48...f370062af6af5633ea90fd113fc3070bbbf0a5d7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ba87f1a04e2540c12d8030df5c22d36480370d48...f370062af6af5633ea90fd113fc3070bbbf0a5d7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:16:26 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:16:26 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec080a633b4_530bb42d3d0102224@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 95ae5787 by Sven Tennie at 2024-09-19T13:16:17+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/95ae5787c60fd2acd9ef3198329e048aba4e9f9b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/95ae5787c60fd2acd9ef3198329e048aba4e9f9b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:18:00 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:18:00 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec0868e3e18_530bb43b070102967@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 86ceb2ab by Sven Tennie at 2024-09-19T13:17:51+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/86ceb2aba26ed9a525205567ff6022979398b1d6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/86ceb2aba26ed9a525205567ff6022979398b1d6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:19:25 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:19:25 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec08bde5b4b_530bb478c681034e4@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 79c6f886 by Sven Tennie at 2024-09-19T13:19:18+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/79c6f886695749275d06d64693e3d66ab2980841 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/79c6f886695749275d06d64693e3d66ab2980841 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:25:20 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:25:20 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec0a206df6b_530bb62a8cc10411e@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 56416813 by Sven Tennie at 2024-09-19T13:25:04+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/56416813e6ba50bf68c5678e6bd6bcaacb33fd71 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/56416813e6ba50bf68c5678e6bd6bcaacb33fd71 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:26:22 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:26:22 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec0a5e7b577_530bb42d27c1046f6@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 5809e7e7 by Sven Tennie at 2024-09-19T13:26:14+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5809e7e732619b03d3a4f71ddae294eb70bf36e6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5809e7e732619b03d3a4f71ddae294eb70bf36e6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:29:45 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:29:45 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec0b2931461_530bb6bac601053d@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: e7079a7f by Sven Tennie at 2024-09-19T13:29:34+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e7079a7f81e61ccfc3f069d26b8731c2357f1290 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e7079a7f81e61ccfc3f069d26b8731c2357f1290 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:30:45 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:30:45 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec0b65504a4_530bb62a8cc10588c@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 3c74af07 by Sven Tennie at 2024-09-19T13:30:36+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3c74af071eff52261796cc2d6d3b1f6904bf1c03 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3c74af071eff52261796cc2d6d3b1f6904bf1c03 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:32:43 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:32:43 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec0bdba721d_530bb48df00106519@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 3fd33039 by Sven Tennie at 2024-09-19T13:32:35+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3fd33039033d324ae585aef7e5055dc51b2d20f3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3fd33039033d324ae585aef7e5055dc51b2d20f3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:46:51 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:46:51 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec0f2bcbe71_530bb8a1b78109070@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: e6ead74c by Sven Tennie at 2024-09-19T13:46:43+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e6ead74caf439f1bdfd1bf9ef90938b735b7c73d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e6ead74caf439f1bdfd1bf9ef90938b735b7c73d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:48:24 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:48:24 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec0f88a0306_530bb8916d810972e@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: a23bb8a0 by Sven Tennie at 2024-09-19T13:48:15+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a23bb8a0506f853621130584a71002f140c2ec7b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a23bb8a0506f853621130584a71002f140c2ec7b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 11:51:04 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Thu, 19 Sep 2024 07:51:04 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Next round Message-ID: <66ec1028a140e_530bb8897f8110314@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 6e7cec14 by Sven Tennie at 2024-09-19T13:50:54+02:00 Next round - - - - - 2 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6e7cec1451041aa254a7003285f48d82a829155d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6e7cec1451041aa254a7003285f48d82a829155d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 12:09:13 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Thu, 19 Sep 2024 08:09:13 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] linker: add --optimistic-linking flag Message-ID: <66ec146990b01_530bbad4bac1133b@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: fc7431f8 by doyougnu at 2024-09-19T08:08:48-04:00 linker: add --optimistic-linking flag This patch adds: - the --optimistic-linking flag which binds unknown symbols in the runtime linker to 0xDEADBEEF instead of exiting with failure - The test T25240 which tests these flags using dead code in the FFI system. - closes #25240 This patch is part of the upstreaming haskell.nix patches project. - - - - - 21 changed files: - docs/users_guide/9.12.1-notes.rst - docs/users_guide/runtime_control.rst - libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc - rts/Linker.c - rts/RtsFlags.c - rts/include/rts/Flags.h - rts/linker/elf_got.c - + testsuite/tests/ghci/linking/T25240/Makefile - + testsuite/tests/ghci/linking/T25240/T25240.hs - + testsuite/tests/ghci/linking/T25240/T25240.stderr - + testsuite/tests/ghci/linking/T25240/T25240a.hs - + testsuite/tests/ghci/linking/T25240/all.T - testsuite/tests/ghci/should_run/T18064.stderr - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-32-mingw32 - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-64-mingw32 - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-32-mingw32 - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-64-mingw32 Changes: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -147,6 +147,10 @@ Runtime system For this reason, a warning is emitted if a .tix file is read in implicitly. You can silence this warning by explicitly passing ``--read-tix-file=yes``. Details can be found in `GHC proposal 612 `__. +- Add new runtime flag :rts-flag:`--optimistic-linking` which instructs the + runtime linker to continue in the presence of unknown symbols. By default this + flag is not passed, preserving previous behavior. + ``base`` library ~~~~~~~~~~~~~~~~ ===================================== docs/users_guide/runtime_control.rst ===================================== @@ -397,6 +397,11 @@ Miscellaneous RTS options heap larger than 1T. ``-xr`` is a no-op if GHC is configured with ``--disable-large-address-space`` or if the platform is 32-bit. +.. rts-flag:: --optimistic-linking + + If given, instruct the runtime linker to try to continue linking in the + presence of an unresolved symbol. + .. _rts-options-gc: RTS options to control the garbage collector ===================================== libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc ===================================== @@ -162,6 +162,7 @@ data MiscFlags = MiscFlags , disableDelayedOsMemoryReturn :: Bool , internalCounters :: Bool , linkerAlwaysPic :: Bool + , linkerOptimistic :: Bool , linkerMemBase :: Word -- ^ address to ask the OS for memory for the linker, 0 ==> off , ioManager :: IoManagerFlag @@ -536,6 +537,8 @@ getMiscFlags = do (#{peek MISC_FLAGS, internalCounters} ptr :: IO CBool)) <*> (toBool <$> (#{peek MISC_FLAGS, linkerAlwaysPic} ptr :: IO CBool)) + <*> (toBool <$> + (#{peek MISC_FLAGS, linkerOptimistic} ptr :: IO CBool)) <*> #{peek MISC_FLAGS, linkerMemBase} ptr <*> (toEnum . fromIntegral <$> (#{peek MISC_FLAGS, ioManager} ptr :: IO Word32)) ===================================== rts/Linker.c ===================================== @@ -967,10 +967,20 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) // lookupDependentSymbol directly. SymbolAddr* r = lookupDependentSymbol(lbl, NULL, NULL); if (!r) { - errorBelch("^^ Could not load '%s', dependency unresolved. " - "See top entry above.\n", lbl); - IF_DEBUG(linker, printLoadedObjects()); - fflush(stderr); + if (!RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("^^ Could not load '%s', dependency unresolved. " + "See top entry above. You might consider using --optimistic-linking\n", + lbl); + IF_DEBUG(linker, printLoadedObjects()); + fflush(stderr); + } else { + // if --optimistic-linking is passed into the RTS we allow the linker + // to optimistically continue + errorBelch("^^ Could not load '%s', dependency unresolved, " + "optimistically continuing\n", + lbl); + r = (void*) 0xDEADBEEF; + } } if (!runPendingInitializers()) { ===================================== rts/RtsFlags.c ===================================== @@ -269,6 +269,7 @@ void initRtsFlagsDefaults(void) RtsFlags.MiscFlags.disableDelayedOsMemoryReturn = false; RtsFlags.MiscFlags.internalCounters = false; RtsFlags.MiscFlags.linkerAlwaysPic = DEFAULT_LINKER_ALWAYS_PIC; + RtsFlags.MiscFlags.linkerOptimistic = false; RtsFlags.MiscFlags.linkerMemBase = 0; RtsFlags.MiscFlags.ioManager = IO_MNGR_FLAG_AUTO; #if defined(THREADED_RTS) && defined(mingw32_HOST_OS) @@ -998,6 +999,11 @@ error = true; OPTION_UNSAFE; RtsFlags.MiscFlags.generate_dump_file = true; } + else if (strequal("optimistic-linking", + &rts_argv[arg][2])) { + OPTION_UNSAFE; + RtsFlags.MiscFlags.linkerOptimistic = true; + } else if (strequal("null-eventlog-writer", &rts_argv[arg][2])) { OPTION_UNSAFE; ===================================== rts/include/rts/Flags.h ===================================== @@ -267,6 +267,7 @@ typedef struct _MISC_FLAGS { there as well. */ bool internalCounters; /* See Note [Internal Counters Stats] */ bool linkerAlwaysPic; /* Assume the object code is always PIC */ + bool linkerOptimistic; /* Should the runtime linker optimistically continue */ StgWord linkerMemBase; /* address to ask the OS for memory * for the linker, NULL ==> off */ IO_MANAGER_FLAG ioManager; /* The I/O manager to use. */ ===================================== rts/linker/elf_got.c ===================================== @@ -97,9 +97,22 @@ fillGot(ObjectCode * oc) { if(0 == strncmp(symbol->name,"_GLOBAL_OFFSET_TABLE_",21)) { symbol->addr = oc->info->got_start; } else { - errorBelch("Failed to lookup symbol: %s\n", + errorBelch("Failed to lookup symbol: %s," + " you might consider using --optimistic-linking\n", symbol->name); - return EXIT_FAILURE; + + // if --optimistic-linking is passed into the + // RTS we allow the linker to optimistically + // continue + if (RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("Failed to lookup symbol: %s," + " optimistically continuing.\n", + symbol->name); + symbol->addr = (void*) 0xDEADBEEF; + } else { + return EXIT_FAILURE; + } + } } } else { ===================================== testsuite/tests/ghci/linking/T25240/Makefile ===================================== @@ -0,0 +1,7 @@ +TOP=../../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +.PHONY: T25240 +T25240: + "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) T25240a.hs T25240.hs +RTS --optimistic-linking -RTS ===================================== testsuite/tests/ghci/linking/T25240/T25240.hs ===================================== @@ -0,0 +1,17 @@ +{-# LANGUAGE TemplateHaskell #-} + +{- + +When the RTS linker loads the T25240a module to run the pure foo splice, it +tries to resolve the func symbol even if this function isn't required to run the +splice code, i.e., its dead code. This test checks that by passing the +--optimistic-linking flag the RTS linker continues to link even in the presence +of unknown symbols. + +-} + +module T25240 where + +import T25240a + +$(pure foo) ===================================== testsuite/tests/ghci/linking/T25240/T25240.stderr ===================================== @@ -0,0 +1,2 @@ +ghc: ^^ Could not load 'func', dependency unresolved, optimistically continuing + ===================================== testsuite/tests/ghci/linking/T25240/T25240a.hs ===================================== @@ -0,0 +1,13 @@ +{-# LANGUAGE TemplateHaskell #-} + + +module T25240a + ( foo, func + ) where + + +foo :: [a] +foo = [] + +foreign import ccall "func" + func :: Int -> Int ===================================== testsuite/tests/ghci/linking/T25240/all.T ===================================== @@ -0,0 +1,3 @@ +# skip on darwin because the leading underscores will make the test fail +test('T25240', [when(leading_underscore(),skip), req_rts_linker, extra_files(['T25240a.hs'])], + makefile_test, ['T25240']) ===================================== testsuite/tests/ghci/should_run/T18064.stderr ===================================== @@ -1,2 +1,2 @@ -: ^^ Could not load 'blah', dependency unresolved. See top entry above. +: ^^ Could not load 'blah', dependency unresolved. See top entry above. You might consider using --optimistic-linking ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -234,7 +234,7 @@ module Control.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -303,7 +303,7 @@ module Control.Exception where module Control.Exception.Annotation where -- Safety: None type ExceptionAnnotation :: * -> Constraint - class ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where + class ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where displayExceptionAnnotation :: a -> GHC.Internal.Base.String default displayExceptionAnnotation :: GHC.Internal.Show.Show a => a -> GHC.Internal.Base.String {-# MINIMAL #-} @@ -345,7 +345,7 @@ module Control.Exception.Base where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -850,11 +850,11 @@ module Data.Data where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b constrFields :: Constr -> [GHC.Internal.Base.String] constrFixity :: Constr -> Fixity @@ -897,7 +897,7 @@ module Data.Data where showConstr :: Constr -> GHC.Internal.Base.String showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -921,14 +921,14 @@ module Data.Dynamic where -- Safety: Safe type Dynamic :: * data Dynamic where - Dynamic :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic + Dynamic :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} dynApp :: Dynamic -> Dynamic -> Dynamic dynApply :: Dynamic -> Dynamic -> GHC.Internal.Maybe.Maybe Dynamic - dynTypeRep :: Dynamic -> ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + dynTypeRep :: Dynamic -> ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep fromDyn :: forall a. Typeable a => Dynamic -> a -> a fromDynamic :: forall a. Typeable a => Dynamic -> GHC.Internal.Maybe.Maybe a toDyn :: forall a. Typeable a => a -> Dynamic @@ -1815,11 +1815,11 @@ module Data.Typeable where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b decT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Data.Either.Either ((a :~: b) -> GHC.Internal.Base.Void) (a :~: b) eqT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Maybe.Maybe (a :~: b) @@ -1834,7 +1834,7 @@ module Data.Typeable where rnfTypeRep :: TypeRep -> () showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -5255,7 +5255,7 @@ module GHC.Event where data EventManager = ... type FdKey :: * data FdKey - = ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] + = ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] ; GHC.Internal.Foreign.C.Types.N:CInt[0])GHC.Internal.System.Posix.Types.Fd, ...} type IOCallback :: * @@ -5288,9 +5288,9 @@ module GHC.Event.TimeOut where type TimeoutEdit :: * type TimeoutEdit = TimeoutQueue -> TimeoutQueue type TimeoutKey :: * - newtype TimeoutKey = TK ghc-internal-9.1001.0:GHC.Internal.Event.Unique.Unique + newtype TimeoutKey = TK ghc-internal-9.1100.0:GHC.Internal.Event.Unique.Unique type TimeoutQueue :: * - type TimeoutQueue = ghc-internal-9.1001.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback + type TimeoutQueue = ghc-internal-9.1100.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback module GHC.Exception where -- Safety: Safe @@ -5302,7 +5302,7 @@ module GHC.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -5331,7 +5331,7 @@ module GHC.Exception.Type where type ArithException :: * data ArithException = Overflow | Underflow | LossOfPrecision | DivideByZero | Denormal | RatioZeroDenominator type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -7884,8 +7884,8 @@ module GHC.IO.Handle where hTryLock :: Handle -> LockMode -> GHC.Types.IO GHC.Types.Bool hWaitForInput :: Handle -> GHC.Types.Int -> GHC.Types.IO GHC.Types.Bool isEOF :: GHC.Types.IO GHC.Types.Bool - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle nativeNewline :: Newline nativeNewlineMode :: NewlineMode noNewlineTranslation :: NewlineMode @@ -7935,11 +7935,11 @@ module GHC.IO.Handle.Internals where ioe_notReadable :: forall a. GHC.Types.IO a ioe_notWritable :: forall a. GHC.Types.IO a ioe_semiclosedHandle :: forall a. GHC.Types.IO a - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle openTextEncoding :: forall a. GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.HandleType -> (forall es ds. GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextEncoder es) -> GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextDecoder ds) -> GHC.Types.IO a) -> GHC.Types.IO a readTextDevice :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer readTextDeviceNonBlocking :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer @@ -8003,7 +8003,7 @@ module GHC.IO.Handle.Types where type Handle__ :: * data Handle__ = forall dev enc_state dec_state. - (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => + (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => Handle__ {haDevice :: !dev, haType :: HandleType, haByteBuffer :: ! {-# UNPACK #-}(GHC.Internal.IORef.N:IORef[0] _N)(GHC.Internal.IORef.IORef (GHC.Internal.IO.Buffer.Buffer GHC.Internal.Word.Word8)), @@ -9106,7 +9106,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * @@ -9417,7 +9417,7 @@ module GHC.StaticPtr where -- Safety: None type IsStatic :: (* -> *) -> Constraint class IsStatic p where - fromStaticPtr :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a + fromStaticPtr :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a {-# MINIMAL fromStaticPtr #-} type StaticKey :: * type StaticKey = GHC.Internal.Fingerprint.Type.Fingerprint @@ -10813,8 +10813,8 @@ module Type.Reflection where data (:~~:) a b where HRefl :: forall {k1} (a :: k1). (:~~:) a a pattern App :: forall k2 (t :: k2). () => forall k1 (a :: k1 -> k2) (b :: k1). (t ~ a b) => TypeRep a -> TypeRep b -> TypeRep t - pattern Con :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a - pattern Con' :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a + pattern Con :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a + pattern Con' :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a pattern Fun :: forall k (fun :: k). () => forall (r1 :: GHC.Types.RuntimeRep) (r2 :: GHC.Types.RuntimeRep) (arg :: TYPE r1) (res :: TYPE r2). (k ~ *, fun ~~ (arg -> res)) => TypeRep arg -> TypeRep res -> TypeRep fun type Module :: * data Module = ... @@ -10831,7 +10831,7 @@ module Type.Reflection where type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} decTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Data.Either.Either ((a :~~: b) -> GHC.Internal.Base.Void) (a :~~: b) eqTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Maybe.Maybe (a :~~: b) moduleName :: Module -> GHC.Internal.Base.String @@ -10866,9 +10866,9 @@ module Type.Reflection.Unsafe where data TypeRep a where ... mkTrApp :: forall k1 k2 (a :: k1 -> k2) (b :: k1). TypeRep a -> TypeRep b -> TypeRep (a b) - mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a + mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a mkTyCon :: GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Types.Int -> KindRep -> TyCon - someTypeRepFingerprint :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint + someTypeRepFingerprint :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConKindArgs :: TyCon -> GHC.Types.Int tyConKindRep :: TyCon -> KindRep @@ -11202,9 +11202,9 @@ instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.I instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Monoid (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Monoid (f p), GHC.Internal.Base.Monoid (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Monoid (f (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Monoid (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Monoid (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11260,9 +11260,9 @@ instance forall a. GHC.Internal.Base.Semigroup (Data.Semigroup.Last a) -- Define instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Semigroup (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Semigroup (f p), GHC.Internal.Base.Semigroup (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Semigroup (f (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11392,20 +11392,20 @@ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Last -- Defined instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Max -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Min -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.IO.Class.MonadIO GHC.Types.IO -- Defined in ‘GHC.Internal.Control.Monad.IO.Class’ -instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ -instance forall (m :: * -> *) a. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ +instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ +instance forall (m :: * -> *) a. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Functor.ZipList.ZipList a) -- Defined in ‘GHC.Internal.Functor.ZipList’ instance GHC.Internal.Data.Data.Data Data.Array.Byte.ByteArray -- Defined in ‘Data.Array.Byte’ -instance forall s. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ +instance forall s. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Complex.Complex a) -- Defined in ‘Data.Complex’ -instance forall i j (a :: i) (b :: j). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall i j (a :: i) (b :: j). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.All -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.Any -- Defined in ‘GHC.Internal.Data.Data’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b, GHC.Internal.Ix.Ix a) => GHC.Internal.Data.Data.Data (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.Associativity -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Bool -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Char -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k a (b :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall k a (b :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.DecidedStrictness -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Double -- Defined in ‘GHC.Internal.Data.Data’ @@ -11453,10 +11453,10 @@ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word32 -- Defined in ‘G instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word64 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word8 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Foreign.Ptr.WordPtr -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ -instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ +instance forall k (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ +instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b) => GHC.Internal.Data.Data.Data (Data.Semigroup.Arg a b) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.First a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.Last a) -- Defined in ‘Data.Semigroup’ @@ -11522,7 +11522,7 @@ instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SCha instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SSymbol -- Defined in ‘GHC.Internal.TypeLits’ instance forall k (a :: k). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k (a :: k1). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ -instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1). GHC.Internal.Data.Type.Equality.TestEquality f => GHC.Internal.Data.Type.Equality.TestEquality (Data.Functor.Compose.Compose f g) -- Defined in ‘Data.Functor.Compose’ instance forall a k (b :: k). GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ instance forall a. GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Bits.And a) -- Defined in ‘GHC.Internal.Data.Bits’ @@ -11719,7 +11719,7 @@ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Ba instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.RecUpdError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.TypeError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ -instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ instance [safe] GHC.Internal.Exception.Type.Exception System.Timeout.Timeout -- Defined in ‘System.Timeout’ instance forall a k (b :: k). GHC.Internal.Float.Floating a => GHC.Internal.Float.Floating (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ @@ -12390,8 +12390,8 @@ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataRep -- Defined in ‘ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataType -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Internal.Show.Show (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance GHC.Internal.Show.Show GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ instance forall a b. (GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). Data.Fixed.HasResolution a => GHC.Internal.Show.Show (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ @@ -12462,13 +12462,13 @@ instance GHC.Internal.Show.Show GHC.Internal.Foreign.C.Types.CWchar -- Defined i instance forall a. GHC.Internal.Show.Show (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Foreign.C.ConstPtr’ instance forall a b. (GHC.Internal.Ix.Ix a, GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Internal.Show.Show GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Internal.Show.Show GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ instance GHC.Internal.Show.Show GHC.Types.Double -- Defined in ‘GHC.Internal.Float’ instance GHC.Internal.Show.Show GHC.Types.Float -- Defined in ‘GHC.Internal.Float’ @@ -12502,7 +12502,7 @@ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Handle -- Defined i instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.HandleType -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Newline -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.NewlineMode -- Defined in ‘GHC.Internal.IO.Handle.Types’ -instance [safe] GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.HandlePosn -- Defined in ‘GHC.Internal.IO.Handle’ instance GHC.Internal.Show.Show GHC.Internal.IO.IOMode.IOMode -- Defined in ‘GHC.Internal.IO.IOMode’ instance GHC.Internal.Show.Show GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ @@ -12599,8 +12599,8 @@ instance GHC.Classes.Eq GHC.Internal.Data.Data.ConstrRep -- Defined in ‘GHC.In instance GHC.Classes.Eq GHC.Internal.Data.Data.DataRep -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Classes.Eq GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Classes.Eq (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Eq a, GHC.Classes.Eq b) => GHC.Classes.Eq (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Eq (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Eq (f (g a)) => GHC.Classes.Eq (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ @@ -12676,13 +12676,13 @@ instance forall a. GHC.Classes.Eq (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) - instance forall i e. (GHC.Internal.Ix.Ix i, GHC.Classes.Eq e) => GHC.Classes.Eq (GHC.Internal.Arr.Array i e) -- Defined in ‘GHC.Internal.Arr’ instance forall s i e. GHC.Classes.Eq (GHC.Internal.Arr.STArray s i e) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Classes.Eq GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ instance GHC.Classes.Eq GHC.Internal.Event.TimeOut.TimeoutKey -- Defined in ‘GHC.Internal.Event.TimeOut’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Classes.Eq GHC.Internal.Stack.Types.SrcLoc -- Defined in ‘GHC.Internal.Stack.Types’ instance GHC.Classes.Eq GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ instance GHC.Classes.Eq GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ @@ -12778,8 +12778,8 @@ instance GHC.Classes.Ord GHC.Internal.Unicode.GeneralCategory -- Defined in ‘G instance forall k (a :: k) (b :: k). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k2 (a :: k1) (b :: k2). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k (s :: k). GHC.Classes.Ord (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Ord ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Ord ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Ord a, GHC.Classes.Ord b) => GHC.Classes.Ord (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Ord (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Ord (f (g a)) => GHC.Classes.Ord (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -9330,7 +9330,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-32-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-64-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-32-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-64-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fc7431f8fec425da58df35ac7a76880a63b960d4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fc7431f8fec425da58df35ac7a76880a63b960d4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 12:33:46 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Thu, 19 Sep 2024 08:33:46 -0400 Subject: [Git][ghc/ghc][wip/24148] 111 commits: testsuite: extend size performance tests with gzip (fixes #25046) Message-ID: <66ec1a2a5778c_530bbe467401173eb@gitlab.mail> Zubin pushed to branch wip/24148 at Glasgow Haskell Compiler / GHC Commits: eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 0913faec by Zubin Duggal at 2024-09-19T12:33:39+00:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - 24 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/18d0e5606ad23b6deee17b0ee63a176286da592a...0913faec333d318bc13f6b3d7db2c2c15040db33 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/18d0e5606ad23b6deee17b0ee63a176286da592a...0913faec333d318bc13f6b3d7db2c2c15040db33 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 14:26:57 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Thu, 19 Sep 2024 10:26:57 -0400 Subject: [Git][ghc/ghc][wip/andreask/setNonBlockingMode] 62 commits: Haddock: Add no-compilation flag Message-ID: <66ec34b1a0e0_2f752549aa20681f0@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/setNonBlockingMode at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 63daed6c by Andreas Klebinger at 2024-09-19T16:08:26+02:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - 19 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/90778487224c2db434772cc7dd3e649e629df076...63daed6caa0f857c2c1752eb34d26efa37a75088 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/90778487224c2db434772cc7dd3e649e629df076...63daed6caa0f857c2c1752eb34d26efa37a75088 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 16:13:05 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Thu, 19 Sep 2024 12:13:05 -0400 Subject: [Git][ghc/ghc][wip/strings] 2 commits: Allow performance regression in MultilineStringsPerf Message-ID: <66ec4d913b16d_2f7525823f0c7836a@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC Commits: 4050b0f1 by Brandon Chinn at 2024-09-19T08:38:48-07:00 Allow performance regression in MultilineStringsPerf Metric Increase: MultilineStringsPerf - - - - - 2f0d9afe by Brandon Chinn at 2024-09-19T08:39:29-07:00 Allow performance regression in MultilineStringsPerf Metric Increase: MultilineStringsPerf - - - - - 0 changed files: Changes: View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/da077a6d96cd08c140c0ff2d3b81348fd5be07e1...2f0d9afebd33d4ccc326276a667668f56814dddb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/da077a6d96cd08c140c0ff2d3b81348fd5be07e1...2f0d9afebd33d4ccc326276a667668f56814dddb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 16:13:26 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Thu, 19 Sep 2024 12:13:26 -0400 Subject: [Git][ghc/ghc][wip/strings] Deleted 1 commit: Allow performance regression in MultilineStringsPerf Message-ID: <66ec4da6d14b_2f75258472187919d@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC WARNING: The push did not contain any new commits, but force pushed to delete the commits and changes below. Deleted commits: 2f0d9afe by Brandon Chinn at 2024-09-19T08:39:29-07:00 Allow performance regression in MultilineStringsPerf Metric Increase: MultilineStringsPerf - - - - - 0 changed files: Changes: View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2f0d9afebd33d4ccc326276a667668f56814dddb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2f0d9afebd33d4ccc326276a667668f56814dddb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 16:15:35 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Thu, 19 Sep 2024 12:15:35 -0400 Subject: [Git][ghc/ghc][wip/strings] Replace manual string lexing (#25158) Message-ID: <66ec4e27138de_2f75258aa96c809d7@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC Commits: 7378f606 by Brandon Chinn at 2024-09-19T09:14:51-07:00 Replace manual string lexing (#25158) - - - - - 26 changed files: - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - testsuite/driver/testlib.py - testsuite/tests/ghci/prog013/prog013.stderr - testsuite/tests/ghci/scripts/ghci022.stderr - testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr - testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr - testsuite/tests/parser/should_fail/T21843c.stderr - testsuite/tests/parser/should_fail/T21843e.stderr - testsuite/tests/parser/should_fail/T21843f.stderr - testsuite/tests/parser/should_fail/T3751.stderr - testsuite/tests/parser/should_fail/T5425.stderr - testsuite/tests/parser/should_fail/readFail002.stderr - testsuite/tests/parser/should_fail/readFail004.stderr - testsuite/tests/parser/should_fail/readFail005.stderr - testsuite/tests/parser/should_fail/readFail033.stderr - testsuite/tests/parser/unicode/utf8_010.stderr - testsuite/tests/parser/unicode/utf8_011.stderr - testsuite/tests/parser/unicode/utf8_020.stderr - testsuite/tests/parser/unicode/utf8_021.stderr - testsuite/tests/parser/unicode/utf8_022.stderr - utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs Changes: ===================================== compiler/GHC/Parser.y ===================================== @@ -728,7 +728,7 @@ are the most common patterns, rewritten as regular expressions for clarity: CHAR { L _ (ITchar _ _) } STRING { L _ (ITstring _ _) } - MULTILINESTRING { L _ (ITmultilinestring _ _) } + STRING_MULTI { L _ (ITstringMulti _ _) } INTEGER { L _ (ITinteger _) } RATIONAL { L _ (ITrational _) } @@ -2357,8 +2357,8 @@ atype :: { LHsType GhcPs } (getCHAR $1) } | STRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGs $1) (getSTRING $1) } - | MULTILINESTRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getMULTILINESTRINGs $1) - (getMULTILINESTRING $1) } + | STRING_MULTI { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGMULTIs $1) + (getSTRINGMULTI $1) } | '_' { sL1a $1 $ mkAnonWildCardTy } -- Type variables are never exported, so `M.tyvar` will be rejected by the renamer. -- We let it pass the parser because the renamer can generate a better error message. @@ -4047,8 +4047,8 @@ literal :: { Located (HsLit GhcPs) } : CHAR { sL1 $1 $ HsChar (getCHARs $1) $ getCHAR $1 } | STRING { sL1 $1 $ HsString (getSTRINGs $1) $ getSTRING $1 } - | MULTILINESTRING { sL1 $1 $ HsMultilineString (getMULTILINESTRINGs $1) - $ getMULTILINESTRING $1 } + | STRING_MULTI { sL1 $1 $ HsMultilineString (getSTRINGMULTIs $1) + $ getSTRINGMULTI $1 } | PRIMINTEGER { sL1 $1 $ HsIntPrim (getPRIMINTEGERs $1) $ getPRIMINTEGER $1 } | PRIMWORD { sL1 $1 $ HsWordPrim (getPRIMWORDs $1) @@ -4154,7 +4154,7 @@ getIPDUPVARID (L _ (ITdupipvarid x)) = x getLABELVARID (L _ (ITlabelvarid _ x)) = x getCHAR (L _ (ITchar _ x)) = x getSTRING (L _ (ITstring _ x)) = x -getMULTILINESTRING (L _ (ITmultilinestring _ x)) = x +getSTRINGMULTI (L _ (ITstringMulti _ x)) = x getINTEGER (L _ (ITinteger x)) = x getRATIONAL (L _ (ITrational x)) = x getPRIMCHAR (L _ (ITprimchar _ x)) = x @@ -4180,7 +4180,7 @@ getVOCURLY (L (RealSrcSpan l _) ITvocurly) = srcSpanStartCol l getINTEGERs (L _ (ITinteger (IL src _ _))) = src getCHARs (L _ (ITchar src _)) = src getSTRINGs (L _ (ITstring src _)) = src -getMULTILINESTRINGs (L _ (ITmultilinestring src _)) = src +getSTRINGMULTIs (L _ (ITstringMulti src _)) = src getPRIMCHARs (L _ (ITprimchar src _)) = src getPRIMSTRINGs (L _ (ITprimstring src _)) = src getPRIMINTEGERs (L _ (ITprimint src _)) = src ===================================== compiler/GHC/Parser/CharClass.hs ===================================== @@ -36,14 +36,14 @@ cDigit = 64 {-# INLINABLE is_ctype #-} is_ctype :: Word8 -> Char -> Bool -is_ctype mask c = (charType c .&. mask) /= 0 +is_ctype mask c = c <= '\127' && (charType c .&. mask) /= 0 is_ident, is_symbol, is_any, is_space, is_lower, is_upper, is_digit, is_alphanum :: Char -> Bool is_ident = is_ctype cIdent is_symbol = is_ctype cSymbol is_any = is_ctype cAny -is_space = \c -> c <= '\x7f' && is_ctype cSpace c -- is_space only works for <= '\x7f' (#3751, #5425) +is_space = is_ctype cSpace is_lower = is_ctype cLower is_upper = is_ctype cUpper is_digit = is_ctype cDigit ===================================== compiler/GHC/Parser/Errors/Ppr.hs ===================================== @@ -157,8 +157,6 @@ instance Diagnostic PsMessage where LexUnknownPragma -> text "unknown pragma" LexErrorInPragma -> text "lexical error in pragma" LexNumEscapeRange -> text "numeric escape sequence out of range" - LexStringCharLit -> text "lexical error in string/character literal" - LexStringCharLitEOF -> text "unexpected end-of-file in string/character literal" LexUnterminatedComment -> text "unterminated `{-'" LexUnterminatedOptions -> text "unterminated OPTIONS pragma" LexUnterminatedQQ -> text "unterminated quasiquotation" ===================================== compiler/GHC/Parser/Errors/Types.hs ===================================== @@ -591,8 +591,6 @@ data LexErr | LexUnknownPragma -- ^ Unknown pragma | LexErrorInPragma -- ^ Lexical error in pragma | LexNumEscapeRange -- ^ Numeric escape sequence out of range - | LexStringCharLit -- ^ Lexical error in string/character literal - | LexStringCharLitEOF -- ^ Unexpected end-of-file in string/character literal | LexUnterminatedComment -- ^ Unterminated `{-' | LexUnterminatedOptions -- ^ Unterminated OPTIONS pragma | LexUnterminatedQQ -- ^ Unterminated quasiquotation ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -167,6 +167,7 @@ $idchar = [$small $large $digit $uniidchar \'] $unigraphic = \x06 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $graphic = [$small $large $symbol $digit $idchar $special $unigraphic \"\'] +$charesc = [a b f n r t v \\ \" \' \&] $binit = 0-1 $octit = 0-7 @@ -213,6 +214,20 @@ $docsym = [\| \^ \* \$] @floating_point = @numspc @decimal \. @decimal @exponent? | @numspc @decimal @exponent @hex_floating_point = @numspc @hexadecimal \. @hexadecimal @bin_exponent? | @numspc @hexadecimal @bin_exponent + at gap = \\ $whitechar+ \\ + at cntrl = $asclarge | \@ | \[ | \\ | \] | \^ | \_ + at ascii = \^ @cntrl | "NUL" | "SOH" | "STX" | "ETX" | "EOT" | "ENQ" | "ACK" + | "BEL" | "BS" | "HT" | "LF" | "VT" | "FF" | "CR" | "SO" | "SI" | "DLE" + | "DC1" | "DC2" | "DC3" | "DC4" | "NAK" | "SYN" | "ETB" | "CAN" + | "EM" | "SUB" | "ESC" | "FS" | "GS" | "RS" | "US" | "SP" | "DEL" +-- N.B. ideally, we would do `@escape # \\ \&` instead of duplicating in @escapechar, +-- which is what the Haskell Report says, but this isn't valid Alex syntax, as only +-- character sets can be subtracted, not strings + at escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) + at escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) + at stringchar = ($graphic # [\\ \"]) | " " | @escape | @gap + at char = ($graphic # [\\ \']) | " " | @escapechar + -- normal signed numerical literals can only be explicitly negative, -- not explicitly positive (contrast @exponent) @negative = \- @@ -460,7 +475,7 @@ $unigraphic / { isSmartQuote } { smart_quote_error } <0> { "#" $idchar+ / { ifExtension OverloadedLabelsBit } { skip_one_varid_src ITlabelvarid } - "#" \" / { ifExtension OverloadedLabelsBit } { lex_quoted_label } + "#" \" @stringchar* \" / { ifExtension OverloadedLabelsBit } { tok_quoted_label } } <0> { @@ -660,14 +675,38 @@ $unigraphic / { isSmartQuote } { smart_quote_error } } --- Strings and chars are lexed by hand-written code. The reason is --- that even if we recognise the string or char here in the regex --- lexer, we would still have to parse the string afterward in order --- to convert it to a String. <0> { - \' { lex_char_tok } - \"\"\" / { ifExtension MultilineStringsBit} { lex_string_tok StringTypeMulti } - \" { lex_string_tok StringTypeSingle } + \"\"\" / { ifExtension MultilineStringsBit } { tok_string_multi } + \" @stringchar* \" { tok_string } + \" @stringchar* \" \# / { ifExtension MagicHashBit } { tok_string } + \' @char \' { tok_char } + \' @char \' \# / { ifExtension MagicHashBit } { tok_char } + + -- Check for smart quotes and throw better errors than a plain lexical error (#21843) + \' \\ $unigraphic / { isSmartQuote } { smart_quote_error } + \" @stringchar* \\ $unigraphic / { isSmartQuote } { smart_quote_error } + -- See Note [Bare smart quote error] + -- The valid string rule will take precedence because it'll match more + -- characters than this rule, so this rule will only fire if the string + -- could not be lexed correctly + \" @stringchar* $unigraphic / { isSmartQuote } { smart_quote_error } +} + + { + -- Parse as much of the multiline string as possible, except for quotes + @stringchar* ($nl ([\ $tab] | @gap)* @stringchar*)* { tok_string_multi_content } + -- Allow bare quotes if it's not a triple quote + (\" | \"\") / ([\n .] # \") { tok_string_multi_content } +} + +<0> { + \'\' { token ITtyQuote } + + -- The normal character match takes precedence over this because it matches + -- more characters. However, if that pattern didn't match, then this quote + -- could be a quoted identifier, like 'x. Here, just return ITsimpleQuote, + -- as the parser will lex the varid separately. + \' / ($graphic # \\ | " ") { token ITsimpleQuote } } -- Note [Whitespace-sensitive operator parsing] @@ -953,7 +992,7 @@ data Token | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" | ITstring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" - | ITmultilinestring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringMulti SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" | ITinteger IntegralLit -- Note [Literal source text] in "GHC.Types.SourceText" | ITrational FractionalLit @@ -2181,156 +2220,128 @@ lex_string_prag_comment mkTok span _buf _len _buf2 -- ----------------------------------------------------------------------------- -- Strings & Chars --- This stuff is horrible. I hates it. - -lex_string_tok :: LexStringType -> Action -lex_string_tok strType span buf _len _buf2 = do - s <- lex_string strType - - i <- getInput - case strType of - StringTypeSingle -> - lex_magic_hash i >>= \case - Just i' -> do - when (any (> '\xFF') s) $ do - pState <- getPState - let msg = PsErrPrimStringInvalidChar - let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg - addError err - - setInput i' - let (psSpan, src) = getStringLoc (buf, locStart) i' - pure $ L psSpan (ITprimstring src (unsafeMkByteString s)) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITstring src (mkFastString s)) - StringTypeMulti -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITmultilinestring src (mkFastString s)) +tok_string :: Action +tok_string span buf len _buf2 = do + s <- lex_chars ("\"", "\"") span buf (if endsInHash then len - 1 else len) + + if endsInHash + then do + when (any (> '\xFF') s) $ do + pState <- getPState + let msg = PsErrPrimStringInvalidChar + let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg + addError err + pure $ L span (ITprimstring src (unsafeMkByteString s)) + else + pure $ L span (ITstring src (mkFastString s)) where - locStart = psSpanStart span + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + +-- | Ideally, we would define this completely with Alex syntax, like normal strings. +-- Instead, this is defined as a hybrid solution by manually invoking lex states, which +-- we're doing for two reasons: +-- 1. The multiline string should all be one lexical token, not multiple +-- 2. We need to allow bare quotes, which can't be done with one regex +tok_string_multi :: Action +tok_string_multi startSpan startBuf _len _buf2 = do + -- advance to the end of the multiline string + let startLoc = psSpanStart startSpan + let i@(AI _ contentStartBuf) = + case lexDelim $ AI startLoc startBuf of + Just i -> i + Nothing -> panic "tok_string_multi did not start with a delimiter" + (AI _ contentEndBuf, i'@(AI endLoc endBuf)) <- goContent i + + -- build the values pertaining to the entire multiline string, including delimiters + let span = mkPsSpan startLoc endLoc + let len = byteDiff startBuf endBuf + let src = SourceText $ lexemeToFastString startBuf len + + -- load the content of the multiline string + let contentLen = byteDiff contentStartBuf contentEndBuf + s <- + either (throwStringLexError (AI startLoc startBuf)) pure $ + lexMultilineString contentLen contentStartBuf + + setInput i' + pure $ L span $ ITstringMulti src (mkFastString s) + where + goContent i0 = + case alexScan i0 string_multi_content of + AlexToken i1 len _ + | Just i2 <- lexDelim i1 -> pure (i1, i2) + | -- is the next token a tab character? + -- need this explicitly because there's a global rule matching $tab + Just ('\t', _) <- alexGetChar' i1 -> setInput i1 >> lexError LexError + | isEOF i1 -> checkSmartQuotes >> lexError LexError + | len == 0 -> panic $ "parsing multiline string got into infinite loop at: " ++ show i0 + | otherwise -> goContent i1 + AlexSkip i1 _ -> goContent i1 + _ -> lexError LexError + + lexDelim = + let go 0 i = Just i + go n i = + case alexGetChar' i of + Just ('"', i') -> go (n - 1) i' + _ -> Nothing + in go (3 :: Int) + + -- See Note [Bare smart quote error] + checkSmartQuotes = do + let findSmartQuote i0@(AI loc _) = + case alexGetChar' i0 of + Just ('\\', i1) | Just (_, i2) <- alexGetChar' i1 -> findSmartQuote i2 + Just (c, i1) + | isDoubleSmartQuote c -> Just (c, loc) + | otherwise -> findSmartQuote i1 + _ -> Nothing + case findSmartQuote (AI (psSpanStart startSpan) startBuf) of + Just (c, loc) -> throwSmartQuoteError c loc + Nothing -> pure () + +-- | Dummy action that should never be called. Should only be used in lex states +-- that are manually lexed in tok_string_multi. +tok_string_multi_content :: Action +tok_string_multi_content = panic "tok_string_multi_content unexpectedly invoked" + +lex_chars :: (String, String) -> PsSpan -> StringBuffer -> Int -> P String +lex_chars (startDelim, endDelim) span buf len = + either (throwStringLexError i0) pure $ + lexString contentLen contentBuf + where + i0@(AI _ contentBuf) = advanceInputBytes (length startDelim) $ AI (psSpanStart span) buf + -- assumes delimiters are ASCII, with 1 byte per Char + contentLen = len - length startDelim - length endDelim -lex_quoted_label :: Action -lex_quoted_label span buf _len _buf2 = do - s <- lex_string StringTypeSingle - (AI end bufEnd) <- getInput - let - token = ITlabelvarid (SourceText src) (mkFastString s) - src = lexemeToFastString (stepOn buf) (cur bufEnd - cur buf - 1) - start = psSpanStart span - - return $ L (mkPsSpan start end) token - - -lex_string :: LexStringType -> P String -lex_string strType = do - start <- getInput - (str, next) <- either fromStringLexError pure $ lexString strType alexGetChar' start - setInput next - pure str - - -lex_char_tok :: Action --- Here we are basically parsing character literals, such as 'x' or '\n' --- but we additionally spot 'x and ''T, returning ITsimpleQuote and --- ITtyQuote respectively, but WITHOUT CONSUMING the x or T part --- (the parser does that). --- So we have to do two characters of lookahead: when we see 'x we need to --- see if there's a trailing quote -lex_char_tok span buf _len _buf2 = do -- We've seen ' - i1 <- getInput -- Look ahead to first character - let loc = psSpanStart span - case alexGetChar' i1 of - Nothing -> lit_error i1 - - Just ('\'', i2@(AI end2 _)) -> do -- We've seen '' - setInput i2 - return (L (mkPsSpan loc end2) ITtyQuote) - - Just ('\\', i2@(AI end2 _)) -> do -- We've seen 'backslash - (lit_ch, i3) <- - either fromStringLexError pure $ - resolveEscapeCharacter alexGetChar' i2 - case alexGetChar' i3 of - Just ('\'', i4) -> do - setInput i4 - finish_char_tok buf loc lit_ch - Just (mc, _) | isSingleSmartQuote mc -> add_smart_quote_error mc end2 - _ -> lit_error i3 - - Just (c, i2@(AI end2 _)) - | not (isAnyChar c) -> lit_error i1 - | otherwise -> - - -- We've seen 'x, where x is a valid character - -- (i.e. not newline etc) but not a quote or backslash - case alexGetChar' i2 of -- Look ahead one more character - Just ('\'', i3) -> do -- We've seen 'x' - setInput i3 - finish_char_tok buf loc c - Just (c, _) | isSingleSmartQuote c -> add_smart_quote_error c end2 - _other -> do -- We've seen 'x not followed by quote - -- (including the possibility of EOF) - -- Just parse the quote only - let (AI end _) = i1 - return (L (mkPsSpan loc end) ITsimpleQuote) - --- We've already seen the closing quote --- Just need to check for trailing # -finish_char_tok :: StringBuffer -> PsLoc -> Char -> P (PsLocated Token) -finish_char_tok buf loc ch = do - i <- getInput - lex_magic_hash i >>= \case - Just i' -> do - setInput i' - -- Include the trailing # in SourceText - let (psSpan, src) = getStringLoc (buf, loc) i' - pure $ L psSpan (ITprimchar src ch) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, loc) i - pure $ L psSpan (ITchar src ch) +throwStringLexError :: AlexInput -> StringLexError -> P a +throwStringLexError i (StringLexError e pos) = setInput (advanceInputTo pos i) >> lexError e --- | Get the span and source text for a string from the given start to the given end. -getStringLoc :: (StringBuffer, PsLoc) -> AlexInput -> (PsSpan, SourceText) -getStringLoc (bufStart, locStart) (AI locEnd bufEnd) = (psSpan, SourceText src) +tok_quoted_label :: Action +tok_quoted_label span buf len _buf2 = do + s <- lex_chars ("#\"", "\"") span buf len + pure $ L span (ITlabelvarid src (mkFastString s)) where - psSpan = mkPsSpan locStart locEnd - src = lexemeToFastString bufStart (cur bufEnd - cur bufStart) - - --- Return Just if we found the magic hash, with the next input. -lex_magic_hash :: AlexInput -> P (Maybe AlexInput) -lex_magic_hash i = do - magicHash <- getBit MagicHashBit - if magicHash - then - case alexGetChar' i of - Just ('#', i') -> pure (Just i') - _other -> pure Nothing - else pure Nothing - -fromStringLexError :: StringLexError AlexInput -> P a -fromStringLexError = \case - UnexpectedEOF i squote -> checkSQuote squote >> throw i LexStringCharLitEOF - BadCharInitialLex i squote -> checkSQuote squote >> throw i LexStringCharLit - EscapeBadChar i -> throw i LexStringCharLit - EscapeUnexpectedEOF i -> throw i LexStringCharLitEOF - EscapeNumRangeError i -> throw i LexNumEscapeRange - EscapeSmartQuoteError c (AI loc _) -> add_smart_quote_error c loc + -- skip leading '#' + src = SourceText . mkFastString . drop 1 $ lexemeToString buf len + + +tok_char :: Action +tok_char span buf len _buf2 = do + c <- lex_chars ("'", "'") span buf (if endsInHash then len - 1 else len) >>= \case + [c] -> pure c + s -> panic $ "tok_char expected exactly one character, got: " ++ show s + pure . L span $ + if endsInHash + then ITprimchar src c + else ITchar src c where - throw i e = setInput i >> lexError e - checkSQuote = \case - NoSmartQuote -> pure () - SmartQuote c (AI loc _) -> add_nonfatal_smart_quote_error c loc - --- before calling lit_error, ensure that the current input is pointing to --- the position of the error in the buffer. This is so that we can report --- a correct location to the user, but also so we can detect UTF-8 decoding --- errors if they occur. -lit_error :: AlexInput -> P a -lit_error i = do setInput i; lexError LexStringCharLit + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + -- ----------------------------------------------------------------------------- -- QuasiQuote @@ -2389,32 +2400,28 @@ quasiquote_error start = do isSmartQuote :: AlexAccPred ExtsBitmap isSmartQuote _ _ _ (AI _ buf) = let c = prevChar buf ' ' in isSingleSmartQuote c || isDoubleSmartQuote c -smart_quote_error_message :: Char -> PsLoc -> MsgEnvelope PsMessage -smart_quote_error_message c loc = - let (correct_char, correct_char_name) = - if isSingleSmartQuote c then ('\'', "Single Quote") else ('"', "Quotation Mark") - err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ - PsErrUnicodeCharLooksLike c correct_char correct_char_name in - err - +throwSmartQuoteError :: Char -> PsLoc -> P a +throwSmartQuoteError c loc = addFatalError err + where + err = + mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ + PsErrUnicodeCharLooksLike c correct_char correct_char_name + (correct_char, correct_char_name) = + if isSingleSmartQuote c + then ('\'', "Single Quote") + else ('"', "Quotation Mark") + +-- | Throw a smart quote error, where the smart quote was the last character lexed smart_quote_error :: Action -smart_quote_error span buf _len _buf2 = do - let c = currentChar buf - addFatalError (smart_quote_error_message c (psSpanStart span)) - -add_smart_quote_error :: Char -> PsLoc -> P a -add_smart_quote_error c loc = addFatalError (smart_quote_error_message c loc) - -add_nonfatal_smart_quote_error :: Char -> PsLoc -> P () -add_nonfatal_smart_quote_error c loc = addError (smart_quote_error_message c loc) +smart_quote_error span _ _ buf2 = do + let c = prevChar buf2 (panic "smart_quote_error unexpectedly called on beginning of input") + throwSmartQuoteError c (psSpanStart span) -advance_to_smart_quote_character :: P () -advance_to_smart_quote_character = do - i <- getInput - case alexGetChar' i of - Just (c, _) | isDoubleSmartQuote c -> return () - Just (_, i2) -> do setInput i2; advance_to_smart_quote_character - Nothing -> return () -- should never get here +-- Note [Bare smart quote error] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- A smart quote inside of a string is allowed, but if a complete valid string +-- couldn't be lexed, we want to see if there's a smart quote that the user +-- thought ended the string, but in fact didn't. -- ----------------------------------------------------------------------------- -- Warnings @@ -2652,7 +2659,7 @@ getLastLocIncludingComments = P $ \s@(PState { prev_loc = prev_loc }) -> POk s p getLastLoc :: P PsSpan getLastLoc = P $ \s@(PState { last_loc = last_loc }) -> POk s last_loc -data AlexInput = AI !PsLoc !StringBuffer +data AlexInput = AI !PsLoc !StringBuffer deriving (Show) {- Note [Unicode in Alex] @@ -2763,6 +2770,19 @@ alexGetChar' (AI loc s) where (c,s') = nextChar s loc' = advancePsLoc loc c +-- | Advance the given input N bytes. +advanceInputBytes :: Int -> AlexInput -> AlexInput +advanceInputBytes n i0@(AI _ buf0) = advanceInputTo (cur buf0 + n) i0 + +-- | Advance the given input to the given position. +advanceInputTo :: Int -> AlexInput -> AlexInput +advanceInputTo pos = go + where + go i@(AI _ buf) + | cur buf >= pos = i + | Just (_, i') <- alexGetChar' i = go i' + | otherwise = i -- reached the end, just return the last input + getInput :: P AlexInput getInput = P $ \s at PState{ loc=l, buffer=b } -> POk s (AI l b) @@ -2770,9 +2790,10 @@ setInput :: AlexInput -> P () setInput (AI l b) = P $ \s -> POk s{ loc=l, buffer=b } () nextIsEOF :: P Bool -nextIsEOF = do - AI _ s <- getInput - return $ atEnd s +nextIsEOF = isEOF <$> getInput + +isEOF :: AlexInput -> Bool +isEOF (AI _ buf) = atEnd buf pushLexState :: Int -> P () pushLexState ls = P $ \s at PState{ lex_state=l } -> POk s{lex_state=ls:l} () @@ -3516,6 +3537,11 @@ topNoLayoutContainsCommas [] = False topNoLayoutContainsCommas (ALRLayout _ _ : ls) = topNoLayoutContainsCommas ls topNoLayoutContainsCommas (ALRNoLayout b _ : _) = b +-- If the generated alexScan/alexScanUser functions are called multiple times +-- in this file, alexScanUser gets broken out into a separate function and +-- increases memory usage. Make sure GHC inlines this function and optimizes it. +{-# INLINE alexScanUser #-} + lexToken :: P (PsLocated Token) lexToken = do inp@(AI loc1 buf) <- getInput ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -1,284 +1,199 @@ {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE PatternSynonyms #-} +{-# LANGUAGE TupleSections #-} +{-# LANGUAGE ViewPatterns #-} module GHC.Parser.String ( StringLexError (..), - ContainsSmartQuote (..), - LexStringType (..), lexString, + lexMultilineString, -- * Unicode smart quote helpers isDoubleSmartQuote, isSingleSmartQuote, - - -- * Other helpers - isAnyChar, - resolveEscapeCharacter, ) where -import GHC.Prelude +import GHC.Prelude hiding (getChar) import Control.Arrow ((>>>)) -import Control.Monad (guard, unless, when) -import Data.Char (chr, isPrint, ord) -import Data.List (unfoldr) +import Control.Monad (when) +import Data.Char (chr, ord) +import qualified Data.Foldable1 as Foldable1 +import qualified Data.List.NonEmpty as NonEmpty import Data.Maybe (listToMaybe, mapMaybe) +import GHC.Data.StringBuffer (StringBuffer) +import qualified GHC.Data.StringBuffer as StringBuffer import GHC.Parser.CharClass ( hexDigit, - is_any, is_decdigit, is_hexdigit, is_octdigit, is_space, octDecDigit, ) +import GHC.Parser.Errors.Types (LexErr (..)) import GHC.Utils.Panic (panic) -data LexStringType = StringTypeSingle | StringTypeMulti +type BufPos = Int +data StringLexError = StringLexError LexErr BufPos --- | State to accumulate while iterating through string literal. --- --- Fields are strict here to avoid space leak when iterating through large string literal --- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 -data LexStringState loc = LexStringState - { stringAcc :: !String - -- ^ The string seen so far, reversed - , multilineCommonWsPrefix :: !Int - -- ^ The common prefix for multiline strings. See Note [Multiline string literals] - , initialLoc :: !loc - -- ^ The location of the beginning of the string literal - } - --- | Get the character at the given location, with the location --- of the next character. Returns Nothing if at the end of the --- input. -type GetChar loc = loc -> Maybe (Char, loc) - -lexString :: LexStringType -> GetChar loc -> loc -> Either (StringLexError loc) (String, loc) -lexString strType getChar initialLoc = go initialState initialLoc +lexString :: Int -> StringBuffer -> Either StringLexError String +lexString = lexStringWith processChars processChars where - initialState = - LexStringState - { stringAcc = "" - , multilineCommonWsPrefix = - case strType of - StringTypeMulti -> maxBound - _ -> 0 - , initialLoc = initialLoc - } - - -- 's' is strict here to avoid space leak when iterating through large string literal - -- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 - go !s loc0 = - case getChar loc0 of - -- found closing delimiter - Just ('"', _) | Just loc1 <- checkDelimiter strType getChar loc0 -> do - let postprocess = - case strType of - StringTypeSingle -> id - StringTypeMulti -> postprocessMultiline (multilineCommonWsPrefix s) - Right (postprocess . reverse $ stringAcc s, loc1) - - -- found backslash - Just (c0@'\\', loc1) -> do - case getChar loc1 of - -- found '\&' character, which should be elided - Just ('&', loc2) -> go s loc2 - -- found start of a string gap - Just (c1, loc2) | is_space c1 -> collapseStringGap getChar s loc2 >>= go s - -- some other escape character - Just (c1, loc2) -> - case strType of - StringTypeSingle -> do - (c', loc') <- resolveEscapeCharacter getChar loc1 - go (addChar c' s) loc' - StringTypeMulti -> do - -- keep escape characters unresolved until after post-processing, - -- to distinguish between a user-newline and the user writing "\n". - -- but still process the characters here, to find any errors - _ <- resolveEscapeCharacter getChar loc1 - go (addChar c1 . addChar c0 $ s) loc2 - -- backslash at end of input - Nothing -> Left $ BadCharInitialLex loc1 (hasSQuote getChar s) - - -- found newline character in multiline string - Just (c0@'\n', loc1) | StringTypeMulti <- strType -> - uncurry go $ parseLeadingWS getChar (addChar c0 s) loc1 - - -- found some other character - Just (c0, loc1) | isAnyChar c0 -> go (addChar c0 s) loc1 - - -- found some unknown character - Just (_, _) -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - - -- reached EOF before finding end of string - Nothing -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) -{-# INLINE lexString #-} - -checkDelimiter :: LexStringType -> GetChar loc -> loc -> Maybe loc -checkDelimiter strType getChar loc0 = - case strType of - StringTypeSingle -> do - ('"', loc1) <- getChar loc0 - Just loc1 - StringTypeMulti -> do - ('"', loc1) <- getChar loc0 - ('"', loc2) <- getChar loc1 - ('"', loc3) <- getChar loc2 - Just loc3 -{-# INLINE checkDelimiter #-} - --- | A helper for adding the given character to the lexed string. -addChar :: Char -> LexStringState loc -> LexStringState loc -addChar c s = s{stringAcc = c : stringAcc s} -{-# INLINE addChar #-} - --- | Return whether the string we've parsed so far contains any smart quotes. -hasSQuote :: GetChar loc -> LexStringState loc -> ContainsSmartQuote loc -hasSQuote getChar s - | any isDoubleSmartQuote (stringAcc s) - , (c, loc) : _ <- filter (isDoubleSmartQuote . fst) allChars = - SmartQuote c loc - | otherwise = - NoSmartQuote + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps + >>> resolveEscapes + +-- ----------------------------------------------------------------------------- +-- Lexing interface + +{- +Note [Lexing strings] +~~~~~~~~~~~~~~~~~~~~~ + +After verifying if a string is lexically valid with Alex, we still need to do +some post processing of the string, namely: +1. Collapse string gaps +2. Resolve escape characters + +The problem: 'lexemeToString' is more performant than manually reading +characters from the StringBuffer. However, that completely erases the position +of each character, which we need in order to report the correct position for +error messages (e.g. when resolving escape characters). + +So what we'll do is do two passes. The first pass is optimistic; just convert +to a plain String and process it. If this results in an error, we do a second +pass, this time where each character is annotated with its position. Now, the +error has all the information it needs. + +Ideally, lexStringWith would take a single (forall c. HasChar c => ...) function, +but to help the specializer, we pass it in twice to concretize it for the two +types we actually use. +-} + +-- | See Note [Lexing strings] +lexStringWith :: + ([Char] -> Either (Char, LexErr) [Char]) + -> ([CharPos] -> Either (CharPos, LexErr) [CharPos]) + -> Int + -> StringBuffer + -> Either StringLexError String +lexStringWith processChars processCharsPos len buf = + case processChars $ bufferChars buf len of + Right s -> Right s + Left _ -> + case processCharsPos $ bufferLocatedChars buf len of + Right _ -> panic "expected lex error on second pass" + Left ((_, pos), e) -> Left $ StringLexError e pos + +class HasChar c where + getChar :: c -> Char + setChar :: Char -> c -> c + +instance HasChar Char where + getChar = id + setChar = const + +instance HasChar (Char, x) where + getChar = fst + setChar c (_, x) = (c, x) + +pattern Char :: HasChar c => Char -> c +pattern Char c <- (getChar -> c) +{-# COMPLETE Char #-} + +bufferChars :: StringBuffer -> Int -> [Char] +bufferChars = StringBuffer.lexemeToString + +type CharPos = (Char, BufPos) + +bufferLocatedChars :: StringBuffer -> Int -> [CharPos] +bufferLocatedChars initialBuf len = go initialBuf where - allChars = unfoldr getCharWithLoc (initialLoc s) - getCharWithLoc loc = - case getChar loc of - Just (c, loc') -> Just ((c, loc), loc') - Nothing -> Nothing -{-# INLINE hasSQuote #-} - --- | After parsing a backslash and a space character, consume the rest of --- the string gap and return the next location. -collapseStringGap :: GetChar loc -> LexStringState loc -> loc -> Either (StringLexError loc) loc -collapseStringGap getChar s = go + go buf + | atEnd buf = [] + | otherwise = + let (c, buf') = StringBuffer.nextChar buf + in (c, StringBuffer.cur buf) : go buf' + + atEnd buf = StringBuffer.byteDiff initialBuf buf >= len + +-- ----------------------------------------------------------------------------- +-- Lexing phases + +collapseGaps :: HasChar c => [c] -> [c] +collapseGaps = go where - go loc0 = - case getChar loc0 of - Just ('\\', loc1) -> pure loc1 - Just (c0, loc1) | is_space c0 -> go loc1 - Just _ -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - Nothing -> Left $ UnexpectedEOF loc0 (hasSQuote getChar s) -{-# INLINE collapseStringGap #-} + go = \case + c1@(Char '\\') : c2@(Char c) : cs + | is_space c -> go $ dropGap cs + | otherwise -> c1 : c2 : go cs + c : cs -> c : go cs + [] -> [] --- | See Note [Multiline string literals] -parseLeadingWS :: GetChar loc -> LexStringState loc -> loc -> (LexStringState loc, loc) -parseLeadingWS getChar = go 0 + dropGap = \case + Char '\\' : cs -> cs + _ : cs -> dropGap cs + [] -> panic "gap unexpectedly ended" + +resolveEscapes :: HasChar c => [c] -> Either (c, LexErr) [c] +resolveEscapes = go dlistEmpty where - go !col s loc = - case getChar loc of - Just (c@' ', loc') -> go (col + 1) (addChar c s) loc' - -- expand tabs - Just ('\t', loc') -> - let fill = 8 - (col `mod` 8) - s' = applyN fill (addChar ' ') s - in go (col + fill) s' loc' - -- if we see a newline or string delimiter, then this line only contained whitespace, so - -- don't include it in the common whitespace prefix - Just ('\n', _) -> (s, loc) - Just ('"', _) | Just _ <- checkDelimiter StringTypeMulti getChar loc -> (s, loc) - -- found some other character, so we're done parsing leading whitespace - _ -> - let s' = s{multilineCommonWsPrefix = min col (multilineCommonWsPrefix s)} - in (s', loc) - - applyN :: Int -> (a -> a) -> a -> a - applyN n f x0 = iterate f x0 !! n -{-# INLINE parseLeadingWS #-} - -data StringLexError loc - = UnexpectedEOF !loc !(ContainsSmartQuote loc) - -- ^ Unexpectedly hit EOF when lexing string - | BadCharInitialLex !loc !(ContainsSmartQuote loc) - -- ^ Found invalid character when initially lexing string - | EscapeBadChar !loc - -- ^ Found invalid character when parsing an escaped character - | EscapeUnexpectedEOF !loc - -- ^ Unexpectedly hit EOF when parsing an escaped character - | EscapeNumRangeError !loc - -- ^ Escaped number exceeds range - | EscapeSmartQuoteError !Char !loc - -- ^ Found escaped smart unicode chars as `\’` or `\”` - deriving (Show) - --- | When initially lexing the string, we want to track if we've --- seen a smart quote, to show a helpful "you might be accidentally --- using a smart quote" error. -data ContainsSmartQuote loc - = NoSmartQuote - | SmartQuote !Char !loc - deriving (Show) + go !acc = \case + [] -> pure $ dlistToList acc + Char '\\' : Char '&' : cs -> go acc cs + backslash@(Char '\\') : cs -> + case resolveEscapeChar cs of + Right (esc, cs') -> go (acc `dlistSnoc` setChar esc backslash) cs' + Left (c, e) -> Left (c, e) + c : cs -> go (acc `dlistSnoc` c) cs -- ----------------------------------------------------------------------------- -- Escape characters --- | After finding a backslash, parse the rest of the escape character, starting --- at the given location. -resolveEscapeCharacter :: GetChar loc -> loc -> Either (StringLexError loc) (Char, loc) -resolveEscapeCharacter getChar loc0 = do - (c0, loc1) <- expectChar loc0 - case c0 of - 'a' -> pure ('\a', loc1) - 'b' -> pure ('\b', loc1) - 'f' -> pure ('\f', loc1) - 'n' -> pure ('\n', loc1) - 'r' -> pure ('\r', loc1) - 't' -> pure ('\t', loc1) - 'v' -> pure ('\v', loc1) - '\\' -> pure ('\\', loc1) - '"' -> pure ('\"', loc1) - '\'' -> pure ('\'', loc1) - -- escape codes - 'x' -> expectNum is_hexdigit 16 hexDigit loc1 - 'o' -> expectNum is_octdigit 8 octDecDigit loc1 - _ | is_decdigit c0 -> expectNum is_decdigit 10 octDecDigit loc0 - -- control characters (e.g. '\^M') - '^' -> do - (c1, loc2) <- expectChar loc1 - unless ('@' <= c1 && c1 <= '_') $ Left $ EscapeBadChar loc1 - pure (chr $ ord c1 - ord '@', loc2) - -- long form escapes (e.g. '\NUL') - _ | Just (c1, loc2) <- parseLongEscape getChar c0 loc1 -> pure (c1, loc2) - -- check unicode smart quotes (#21843) - _ | isDoubleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - _ | isSingleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - -- unknown escape - _ -> Left $ EscapeBadChar loc0 +-- | Resolve a escape character, after having just lexed a backslash. +-- Assumes escape character is valid. +resolveEscapeChar :: HasChar c => [c] -> Either (c, LexErr) (Char, [c]) +resolveEscapeChar = \case + Char 'a' : cs -> pure ('\a', cs) + Char 'b' : cs -> pure ('\b', cs) + Char 'f' : cs -> pure ('\f', cs) + Char 'n' : cs -> pure ('\n', cs) + Char 'r' : cs -> pure ('\r', cs) + Char 't' : cs -> pure ('\t', cs) + Char 'v' : cs -> pure ('\v', cs) + Char '\\' : cs -> pure ('\\', cs) + Char '"' : cs -> pure ('\"', cs) + Char '\'' : cs -> pure ('\'', cs) + -- escape codes + Char 'x' : cs -> parseNum is_hexdigit 16 hexDigit cs + Char 'o' : cs -> parseNum is_octdigit 8 octDecDigit cs + cs@(Char c : _) | is_decdigit c -> parseNum is_decdigit 10 octDecDigit cs + -- control characters (e.g. '\^M') + Char '^' : Char c : cs -> pure (chr $ ord c - ord '@', cs) + -- long form escapes (e.g. '\NUL') + cs | Just (esc, cs') <- parseLongEscape cs -> pure (esc, cs') + -- shouldn't happen + Char c : _ -> panic $ "found unexpected escape character: " ++ show c + [] -> panic "escape character unexpectedly ended" where - expectChar loc = - case getChar loc of - Just x -> pure x - Nothing -> Left $ EscapeUnexpectedEOF loc - - expectNum isDigit base toDigit loc1 = do - (c1, loc2) <- expectChar loc1 - unless (isDigit c1) $ Left $ EscapeBadChar loc1 - let parseNum x loc = - case getChar loc of - Just (c, loc') | isDigit c -> do - let x' = x * base + toDigit c - when (x' > 0x10ffff) $ Left $ EscapeNumRangeError loc - parseNum x' loc' - _ -> - pure (chr x, loc) - parseNum (toDigit c1) loc2 -{-# INLINE resolveEscapeCharacter #-} - -parseLongEscape :: GetChar loc -> Char -> loc -> Maybe (Char, loc) -parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCodes + parseNum isDigit base toDigit = + let go x = \case + ch@(Char c) : cs | isDigit c -> do + let x' = x * base + toDigit c + when (x' > 0x10ffff) $ Left (ch, LexNumEscapeRange) + go x' cs + cs -> pure (chr x, cs) + in go 0 + +parseLongEscape :: HasChar c => [c] -> Maybe (Char, [c]) +parseLongEscape cs = listToMaybe (mapMaybe tryParse longEscapeCodes) where - tryParse (prefix, c) = do - p0 : p <- pure prefix - guard (p0 == c0) -- see if the first character matches - loc <- parsePrefix loc1 p -- see if the rest of the prefix matches - pure (c, loc) - - parsePrefix loc = \case - [] -> pure loc - p : ps -> do - (c, loc') <- getChar loc - guard (p == c) - parsePrefix loc' ps + tryParse (code, esc) = + case splitAt (length code) cs of + (pre, cs') | map getChar pre == code -> Just (esc, cs') + _ -> Nothing longEscapeCodes = [ ("NUL", '\NUL') @@ -289,14 +204,14 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("ENQ", '\ENQ') , ("ACK", '\ACK') , ("BEL", '\BEL') - , ("BS", '\BS') - , ("HT", '\HT') - , ("LF", '\LF') - , ("VT", '\VT') - , ("FF", '\FF') - , ("CR", '\CR') - , ("SO", '\SO') - , ("SI", '\SI') + , ("BS" , '\BS' ) + , ("HT" , '\HT' ) + , ("LF" , '\LF' ) + , ("VT" , '\VT' ) + , ("FF" , '\FF' ) + , ("CR" , '\CR' ) + , ("SO" , '\SO' ) + , ("SI" , '\SI' ) , ("DLE", '\DLE') , ("DC1", '\DC1') , ("DC2", '\DC2') @@ -306,17 +221,16 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("SYN", '\SYN') , ("ETB", '\ETB') , ("CAN", '\CAN') - , ("EM", '\EM') + , ("EM" , '\EM' ) , ("SUB", '\SUB') , ("ESC", '\ESC') - , ("FS", '\FS') - , ("GS", '\GS') - , ("RS", '\RS') - , ("US", '\US') - , ("SP", '\SP') + , ("FS" , '\FS' ) + , ("GS" , '\GS' ) + , ("RS" , '\RS' ) + , ("US" , '\US' ) + , ("SP" , '\SP' ) , ("DEL", '\DEL') ] -{-# INLINE parseLongEscape #-} -- ----------------------------------------------------------------------------- -- Unicode Smart Quote detection (#21843) @@ -333,6 +247,98 @@ isSingleSmartQuote = \case '’' -> True _ -> False +-- ----------------------------------------------------------------------------- +-- Multiline strings + +-- | See Note [Multiline string literals] +-- +-- Assumes string is lexically valid. Skips the steps about splitting +-- and rejoining lines, and instead manually find newline characters, +-- for performance. +lexMultilineString :: Int -> StringBuffer -> Either StringLexError String +lexMultilineString = lexStringWith processChars processChars + where + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps -- Step 1 + >>> expandLeadingTabs -- Step 3 + >>> rmCommonWhitespacePrefix -- Step 4 + >>> collapseOnlyWsLines -- Step 5 + >>> rmFirstNewline -- Step 7a + >>> rmLastNewline -- Step 7b + >>> resolveEscapes -- Step 8 + + -- expands all tabs, since the lexer will verify that tabs can only appear + -- as leading indentation + expandLeadingTabs :: HasChar c => [c] -> [c] + expandLeadingTabs = + let go !col = \case + c@(Char '\t') : cs -> + let fill = 8 - (col `mod` 8) + in replicate fill (setChar ' ' c) ++ go (col + fill) cs + c : cs -> c : go (if getChar c == '\n' then 0 else col + 1) cs + [] -> [] + in go 0 + + rmCommonWhitespacePrefix :: HasChar c => [c] -> [c] + rmCommonWhitespacePrefix cs0 = + let commonWSPrefix = getCommonWsPrefix (map getChar cs0) + go = \case + c@(Char '\n') : cs -> c : go (dropLine commonWSPrefix cs) + c : cs -> c : go cs + [] -> [] + -- drop x characters from the string, or up to a newline, whichever + -- comes first + dropLine !x = \case + cs | x <= 0 -> cs + cs@(Char '\n' : _) -> cs + _ : cs -> dropLine (x - 1) cs + [] -> [] + in go cs0 + + collapseOnlyWsLines :: HasChar c => [c] -> [c] + collapseOnlyWsLines = + let go = \case + c@(Char '\n') : cs | Just cs' <- checkAllWs cs -> c : go cs' + c : cs -> c : go cs + [] -> [] + checkAllWs = \case + -- got all the way to a newline or the end of the string, return + cs@(Char '\n' : _) -> Just cs + cs@[] -> Just cs + -- found whitespace, continue + Char c : cs | is_space c -> checkAllWs cs + -- anything else, stop + _ -> Nothing + in go + + rmFirstNewline :: HasChar c => [c] -> [c] + rmFirstNewline = \case + Char '\n' : cs -> cs + cs -> cs + + rmLastNewline :: HasChar c => [c] -> [c] + rmLastNewline = + let go = \case + [] -> [] + [Char '\n'] -> [] + c : cs -> c : go cs + in go + +-- | See step 4 in Note [Multiline string literals] +-- +-- Assumes tabs have already been expanded. +getCommonWsPrefix :: String -> Int +getCommonWsPrefix s = + case NonEmpty.nonEmpty includedLines of + Nothing -> 0 + Just ls -> Foldable1.minimum $ NonEmpty.map (length . takeWhile is_space) ls + where + includedLines = + filter (not . all is_space) -- ignore whitespace-only lines + . drop 1 -- ignore first line in calculation + $ lines s + {- Note [Multiline string literals] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -351,23 +357,13 @@ The canonical steps for post processing a multiline string are: 2. Split the string by newlines 3. Convert leading tabs into spaces * In each line, any tabs preceding non-whitespace characters are replaced with spaces up to the next tab stop -4. Remove common whitespace prefix in every line (see below) +4. Remove common whitespace prefix in every line except the first (see below) 5. If a line contains only whitespace, remove all of the whitespace 6. Join the string back with `\n` delimiters -7. If the first character of the string is a newline, remove it +7a. If the first character of the string is a newline, remove it +7b. If the last character of the string is a newline, remove it 8. Interpret escaped characters -However, for performance reasons, we do as much of this in one pass as possible: -1. As we lex the string, do the following steps as they appear: - a. Collapse string gaps - b. Keep track of the common whitespace prefix so far - c. Validate escaped characters -2. At the very end, post process the lexed string: - a. Remove the common whitespace prefix from every line - b. Remove all whitespace from all-whitespace lines - c. Remove initial newline character - d. Resolve escaped characters - The common whitespace prefix can be informally defined as "The longest prefix of whitespace shared by all lines in the string, excluding the first line and any whitespace-only lines". @@ -382,75 +378,16 @@ It's more precisely defined with the following algorithm: 3. Calculate the longest prefix of whitespace shared by all lines in the remaining list -} --- | See Note [Multiline string literals] -postprocessMultiline :: Int -> String -> String -postprocessMultiline commonWSPrefix = - rmCommonWhitespacePrefix - >>> collapseOnlyWsLines - >>> rmFirstNewline - >>> rmLastNewline - >>> resolveEscapeChars - where - rmCommonWhitespacePrefix = - let go = \case - '\n' : s -> '\n' : go (dropLine commonWSPrefix s) - c : s -> c : go s - [] -> [] - -- drop x characters from the string, or up to a newline, whichever - -- comes first - dropLine !x = \case - s | x <= 0 -> s - s@('\n' : _) -> s - _ : s -> dropLine (x - 1) s - [] -> [] - in go - - collapseOnlyWsLines = - let go = \case - '\n' : s | Just s' <- checkAllWs s -> '\n' : go s' - c : s -> c : go s - [] -> [] - checkAllWs = \case - -- got all the way to a newline or the end of the string, return - s@('\n' : _) -> Just s - s@[] -> Just s - -- found whitespace, continue - c : s | is_space c -> checkAllWs s - -- anything else, stop - _ -> Nothing - in go - - rmFirstNewline = \case - '\n' : s -> s - s -> s +-- ----------------------------------------------------------------------------- +-- DList - rmLastNewline = - let go = \case - [] -> [] - ['\n'] -> [] - c : cs -> c : go cs - in go +newtype DList a = DList ([a] -> [a]) - -- resolve escape characters, deferred from lexString. guaranteed - -- to not throw any errors, since we already checked them in lexString - resolveEscapeChars = \case - [] -> [] - '\\' : s -> - -- concretizing 'loc' to String: - -- resolveEscapeCharacter :: (String -> Maybe (Char, String)) -> String -> Either _ (Char, String) - case resolveEscapeCharacter uncons s of - Left e -> panic $ "resolving escape characters in multiline string unexpectedly found errors: " ++ show e - Right (c, s') -> c : resolveEscapeChars s' - c : s -> c : resolveEscapeChars s - - uncons = \case - c : cs -> Just (c, cs) - [] -> Nothing +dlistEmpty :: DList a +dlistEmpty = DList id --- ----------------------------------------------------------------------------- --- Helpers +dlistToList :: DList a -> [a] +dlistToList (DList f) = f [] -isAnyChar :: Char -> Bool -isAnyChar c - | c > '\x7f' = isPrint c - | otherwise = is_any c +dlistSnoc :: DList a -> a -> DList a +dlistSnoc (DList f) x = DList (f . (x :)) ===================================== testsuite/driver/testlib.py ===================================== @@ -2760,7 +2760,7 @@ def normalise_errmsg(s: str) -> str: # normalise slashes to minimise Windows/Unix filename differences, # but don't normalize backslashes in chars - s = re.sub(r"(?!')\\", '/', s) + s = re.sub(r"(?:9:1: error: [GHC-58481] parse error on input ‘+’ -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' + +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/ghci/scripts/ghci022.stderr ===================================== @@ -1,3 +1,2 @@ +ghci022.hs:2:5: error: [GHC-21231] lexical error at character '\n' -ghci022.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr ===================================== @@ -1,3 +1,3 @@ - MultilineStringsInnerTab.hs:8:5: error: [GHC-21231] - lexical error in string/character literal at character '\t' + lexical error at character '\t' + ===================================== testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr ===================================== @@ -1,6 +1,2 @@ - MultilineStringsSmartQuotes.hs:9:3: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -MultilineStringsSmartQuotes.hs:10:1: [GHC-21231] - lexical error in string/character literal at end of input ===================================== testsuite/tests/parser/should_fail/T21843c.stderr ===================================== @@ -1,6 +1,2 @@ - -T21843c.hs:3:19: [GHC-31623] +T21843c.hs:3:13: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -T21843c.hs:3:20: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/T21843e.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843e.hs:3:15: [GHC-31623] +T21843e.hs:3:13: error: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not + ===================================== testsuite/tests/parser/should_fail/T21843f.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843f.hs:3:13: [GHC-31623] +T21843f.hs:3:11: error: [GHC-31623] Unicode character '‘' ('\8216') looks like ''' (Single Quote), but it is not + ===================================== testsuite/tests/parser/should_fail/T3751.stderr ===================================== @@ -1,3 +1,2 @@ +T3751.hs:3:5: error: [GHC-21231] lexical error at character '\167' -T3751.hs:3:7: error: [GHC-21231] - lexical error in string/character literal at character '\167' ===================================== testsuite/tests/parser/should_fail/T5425.stderr ===================================== @@ -1,3 +1,2 @@ +T5425.hs:4:1: error: [GHC-21231] lexical error at character '\955' -T5425.hs:5:2: error: [GHC-21231] - lexical error in string/character literal at character '\955' ===================================== testsuite/tests/parser/should_fail/readFail002.stderr ===================================== @@ -1,3 +1,3 @@ +readFail002.hs:5:5: error: [GHC-21231] + lexical error at character '\n' -readFail002.hs:5:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/readFail004.stderr ===================================== @@ -1,3 +1,3 @@ +readFail004.hs:17:16: error: [GHC-21231] + lexical error at character '.' -readFail004.hs:19:1: error: [GHC-21231] - lexical error in string/character literal at character '.' ===================================== testsuite/tests/parser/should_fail/readFail005.stderr ===================================== @@ -1,3 +1,2 @@ - -readFail005.hs:4:7: error: [GHC-21231] - lexical error in string/character literal at character '&' +readFail005.hs:4:5: [GHC-21231] + lexical error at character '&' ===================================== testsuite/tests/parser/should_fail/readFail033.stderr ===================================== @@ -1,3 +1,3 @@ +readFail033.hs:2:5: error: [GHC-21231] + lexical error at character '\t' -readFail033.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\t' ===================================== testsuite/tests/parser/unicode/utf8_010.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_010.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_011.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_011.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_020.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_020.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_021.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_021.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_022.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_022.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs ===================================== @@ -320,7 +320,7 @@ classify tok = ITlabelvarid{} -> TkUnknown ITchar{} -> TkChar ITstring{} -> TkString - ITmultilinestring{} -> TkString + ITstringMulti{} -> TkString ITinteger{} -> TkNumber ITrational{} -> TkNumber ITprimchar{} -> TkChar View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7378f6062ce0e8cb1408382966cbbba872b7d138 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7378f6062ce0e8cb1408382966cbbba872b7d138 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 16:35:27 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Thu, 19 Sep 2024 12:35:27 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25251 Message-ID: <66ec52cfe637f_2f7525bc3af481523@gitlab.mail> Simon Peyton Jones pushed new branch wip/T25251 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25251 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 19:48:52 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Thu, 19 Sep 2024 15:48:52 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] 241 commits: Add hack for #24623 Message-ID: <66ec80249634c_1863ac56c0d47686a@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 8628c415 by Jade at 2024-09-19T21:20:27+02:00 Refactor FieldOcc vs AmbiguousFieldOcc with TTG Improves the design of FieldOcc vs AmbiguousFieldOcc, and removes a dependency on `RdrName` from the Language.Haskell.* namespace (#21592). The design: * The FieldOcc constructor of FieldOcc always refers to an unambiguous field occurrence. * During renaming, a FieldOcc may be ambiguous and only be resolvable during Typechecking * Therefore, we extend (with TTG) `FieldOcc GhcRn` with a constructor `AmbiguousFieldOcc` that constructs a definitely ambiguous `FieldOcc`. * During typechecking, all ambiguous field occurrences must be resolved, so the `AmbiguousFieldOcc` constructor no longer exists See Note [Lifecycle of a FieldOcc] Co-authored-by: Rodrigo Mesquita <rodrigo.m.mesquita at gmail.com> - - - - - 769089f4 by Hassan Al-Awwadi at 2024-09-19T21:41:42+02:00 Wrangled until everything compiled. There are two issues: - FieldOcc used to only have one constructor and always be unambigious, this is no longer the case. Calls to foLabel are thus partial now. Don't know how much we care about this, since the partial calls are mostly inside functions that used to operate on the operate on the unambigious FieldOcc - Lots of functions that take in a FieldOcc, or a HsExpr (with the HsRecSel constructor) now have a new case. It was not always clear to me what the correct implementation was for these. I have filled them in as far as I could and left one undefined... sidenote: I actually can't get it to compile, since there is some error for me in "libraries/ghc-internal/src/GHC/Internal/IO/Handle". I am unable to comprehend it and think it might have something to do with the windows/linux stuff so hopefully github can make it compile :pray: - - - - - a61ae7fc by Hassan Al-Awwadi at 2024-09-19T21:44:30+02:00 silly rebase conflict, ignore it since it'll be gone in the squash anyway :relieved: - - - - - 27 changed files: - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - compiler/GHC.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/PrimOps.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/Utils.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/DebugBlock.hs - compiler/GHC/Cmm/GenericOpt.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/CmmToAsm.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5e11f34d6f5a089e5c4659bdf20e715750e08120...a61ae7fc01046beb596a1fe3157ae0c8d989b381 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5e11f34d6f5a089e5c4659bdf20e715750e08120...a61ae7fc01046beb596a1fe3157ae0c8d989b381 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 20:12:11 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Thu, 19 Sep 2024 16:12:11 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] Wrangled until everything compiled. Message-ID: <66ec859b47a90_1863ac729bc48068b@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: fd237dfe by Hassan Al-Awwadi at 2024-09-19T22:11:38+02:00 Wrangled until everything compiled. There are two issues: - FieldOcc used to only have one constructor and always be unambigious, this is no longer the case. Calls to foLabel are thus partial now. Don't know how much we care about this, since the partial calls are mostly inside functions that used to operate on the operate on the unambigious FieldOcc - Lots of functions that take in a FieldOcc, or a HsExpr (with the HsRecSel constructor) now have a new case. It was not always clear to me what the correct implementation was for these. I have filled them in as far as I could and left one undefined... - - - - - 19 changed files: - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Bind.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Gen/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/Tc/TyCl/Utils.hs - utils/haddock/haddock-api/src/Haddock/Backends/Hoogle.hs - utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs - utils/haddock/haddock-api/src/Haddock/Convert.hs - utils/haddock/haddock-api/src/Haddock/GhcUtils.hs - utils/haddock/haddock-api/src/Haddock/Interface/Create.hs - utils/haddock/haddock-api/src/Haddock/Interface/Rename.hs - utils/haddock/haddock-api/src/Haddock/Types.hs Changes: ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -104,7 +104,7 @@ hsExprType :: HsExpr GhcTc -> Type hsExprType (HsVar _ (L _ id)) = idType id hsExprType (HsUnboundVar (HER _ ty _) _) = ty hsExprType (HsRecSel _ (FieldOcc _ id)) = idType (unLoc id) -hsExprType (HsOverLabel v _ _) = dataConCantHappen v +hsExprType (HsOverLabel v _) = dataConCantHappen v hsExprType (HsIPVar v _) = dataConCantHappen v hsExprType (HsOverLit _ lit) = overLitType lit hsExprType (HsLit _ lit) = hsLitType lit ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -273,7 +273,7 @@ application. For example: See the `HsApp` case for where it is filtered out -} -dsExpr (HsRecSel _ (FieldOcc id _)) +dsExpr (HsRecSel _ (FieldOcc _ (L _ id))) = do { let name = getName id RecSelId {sel_cons = (_, cons_wo_field)} = idDetails id ===================================== compiler/GHC/HsToCore/Quote.hs ===================================== @@ -284,7 +284,7 @@ repTopDs group@(HsGroup { hs_valds = valds , hs_docs = docs }) = do { let { bndrs = hsScopedTvBinders valds ++ hsGroupBinders group - ++ map foExt (hsPatSynSelectors valds) + ++ map (unLoc . foLabel) (hsPatSynSelectors valds) ; instds = tyclds >>= group_instds } ; ss <- mkGenSyms bndrs ; @@ -1538,7 +1538,8 @@ repE (HsVar _ (L _ x)) = repE (HsIPVar _ n) = rep_implicit_param_name n >>= repImplicitParamVar repE (HsOverLabel _ s) = repOverLabel s -repE (HsRecSel _ (FieldOcc x _)) = repE (HsVar noExtField (noLocA x)) +repE (HsRecSel _ (FieldOcc _ (L _ x))) = repE (HsVar noExtField (noLocA x)) +repE r@(HsRecSel _ (XFieldOcc _)) = notHandled (ThAmbiguousRecordSelectors r) -- Remember, we're desugaring renamer output here, so -- HsOverlit can definitely occur @@ -1816,10 +1817,11 @@ repUpdFields = repListM fieldExpTyConName rep_fld where rep_fld :: LHsRecUpdField GhcRn GhcRn -> MetaM (Core (M TH.FieldExp)) rep_fld (L l fld) = case unLoc (hfbLHS fld) of - Unambiguous sel_name _ -> do { fn <- lookupLOcc (L l sel_name) - ; e <- repLE (hfbRHS fld) - ; repFieldExp fn e } - Ambiguous{} -> notHandled (ThAmbiguousRecordUpdates fld) + FieldOcc _ (L _ sel_name) -> do { fn <- lookupLOcc (L l sel_name) + ; e <- repLE (hfbRHS fld) + ; repFieldExp fn e + } + (XFieldOcc _) -> notHandled (ThAmbiguousRecordUpdates fld) @@ -2022,7 +2024,7 @@ rep_bind (L loc (PatSynBind _ (PSB { psb_id = syn mkGenArgSyms (InfixCon arg1 arg2) = mkGenSyms [unLoc arg1, unLoc arg2] mkGenArgSyms (RecCon fields) = do { let pats = map (unLoc . recordPatSynPatVar) fields - sels = map (foExt . recordPatSynField) fields + sels = map (unLoc . foLabel . recordPatSynField) fields ; ss <- mkGenSyms sels ; return $ replaceNames (zip sels pats) ss } @@ -2054,7 +2056,7 @@ repPatSynArgs (InfixCon arg1 arg2) ; arg2' <- lookupLOcc arg2 ; repInfixPatSynArgs arg1' arg2' } repPatSynArgs (RecCon fields) - = do { sels' <- repList nameTyConName (lookupOcc . foExt) sels + = do { sels' <- repList nameTyConName (lookupOcc . unLoc . foLabel) sels ; repRecordPatSynArgs sels' } where sels = map recordPatSynField fields @@ -2877,7 +2879,7 @@ repRecConArgs ips = do rep_ip (L _ ip) = mapM (rep_one_ip (cd_fld_type ip)) (cd_fld_names ip) rep_one_ip :: LBangType GhcRn -> LFieldOcc GhcRn -> MetaM (Core (M TH.VarBangType)) - rep_one_ip t n = do { MkC v <- lookupOcc (foExt $ unLoc n) + rep_one_ip t n = do { MkC v <- lookupOcc (unLoc . foLabel $ unLoc n) ; MkC ty <- repBangTy t ; rep2 varBangTypeName [v,ty] } ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1483,8 +1483,7 @@ instance HiePass p => ToHie (RFContext (LocatedA (FieldOcc (GhcPass p)))) where case f of FieldOcc _ fld -> [toHie $ C (RecField c rhs) (L (locA nspan) $ unLoc fld)] - XFieldOcc (Ambiguous (L nspan fld)) - -> [ toHie $ C (RecField c rhs) (L (locA nspan) fld) ] + XFieldOcc (Ambiguous _) -> [] HieTc -> case f of FieldOcc _ fld -> @@ -2075,9 +2074,8 @@ instance ToHie (LocatedA (TyFamInstDecl GhcRn)) where toHie (L sp (TyFamInstDecl _ d)) = toHie $ TS (ResolvedScopes [mkScope sp]) d instance HiePass p => ToHie (Context (FieldOcc (GhcPass p))) where - toHie (C c (FieldOcc n (L l _))) = case hiePass @p of - HieTc -> toHie (C c (L l n)) - HieRn -> toHie (C c (L l n)) + toHie (C c (FieldOcc _ l)) = toHie (C c l) + toHie (C _ (XFieldOcc _)) = concatM [] instance HiePass p => ToHie (PatSynFieldContext (RecordPatSynField (GhcPass p))) where toHie (PSC sp (RecordPatSynField a b)) = concatM $ ===================================== compiler/GHC/Rename/Bind.hs ===================================== @@ -824,7 +824,9 @@ rnPatSynBind sig_fn bind@(PSB { psb_id = L l name , psb_ext = fvs' } selector_names = case details' of RecCon names -> - map (foExt . recordPatSynField) names + -- I don't actually know if its fine or not + -- that foLabel is partial. + map (unLoc . foLabel . recordPatSynField) names _ -> [] ; fvs' `seq` -- See Note [Free-variable space leak] ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -339,7 +339,7 @@ rnExpr (HsVar _ (L l v)) ; this_mod <- getModule ; when (nameIsLocalOrFrom this_mod sel_name) $ checkThLocalName sel_name - ; return (HsRecSel noExtField (FieldOcc sel_name (L l v) ), unitFV sel_name) + ; return (HsRecSel noExtField (FieldOcc v (L l sel_name)), unitFV sel_name) } | nm == nilDataConName -- Treat [] as an ExplicitList, so that @@ -417,7 +417,7 @@ rnExpr (OpApp _ e1 op e2) -- should prevent bad things happening. ; fixity <- case op' of L _ (HsVar _ (L _ n)) -> lookupFixityRn n - L _ (HsRecSel _ f) -> lookupFieldFixityRn f + L _ (HsRecSel _ f) -> fromJust <$> lookupFieldFixityRn f _ -> return (Fixity minPrecedence InfixL) -- c.f. lookupFixity for unbound ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -257,6 +257,42 @@ newPatName (LetMk is_top fix_env) rdr_name NotTopLevel -> newLocalBndrRn rdr_name TopLevel -> newTopSrcBinder rdr_name ; bindLocalNames [name] $ + -- Do *not* use bindLocalNameFV here; + -- Do *not* use bindLocalNameFV here; + -- Do *not* use bindLocalNameFV here; + -- Do *not* use bindLocalNameFV here; + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- For the TopLevel case + -- For the TopLevel case + -- For the TopLevel case + -- For the TopLevel case + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + + -- Do *not* use bindLocalNameFV here; + + -- Do *not* use bindLocalNameFV here; + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- For the TopLevel case + -- For the TopLevel case + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + + -- Do *not* use bindLocalNameFV here; + -- Do *not* use bindLocalNameFV here; + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- For the TopLevel case + -- For the TopLevel case + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + -- Do *not* use bindLocalNameFV here; -- see Note [View pattern usage] -- For the TopLevel case @@ -877,7 +913,7 @@ rnHsRecFields ctxt mk_arg (HsRecFields { rec_flds = flds, rec_dotdot = dotdot }) ; return $ L l $ HsFieldBind { hfbAnn = noAnn - , hfbLHS = L loc (FieldOcc sel (L ll arg_rdr)) + , hfbLHS = L loc (FieldOcc arg_rdr (L ll sel)) , hfbRHS = arg' , hfbPun = pun } } @@ -897,7 +933,7 @@ rnHsRecFields ctxt mk_arg (HsRecFields { rec_flds = flds, rec_dotdot = dotdot }) ; (rdr_env, lcl_env) <- getRdrEnvs ; conInfo <- lookupConstructorInfo con ; when (conFieldInfo conInfo == ConHasPositionalArgs) (addErr (TcRnIllegalWildcardsInConstructor con)) - ; let present_flds = mkOccSet $ map rdrNameOcc (getFieldLbls flds) + ; let present_flds = mkOccSet $ map nameOccName (getFieldLbls flds) -- For constructor uses (but not patterns) -- the arg should be in scope locally; @@ -923,7 +959,7 @@ rnHsRecFields ctxt mk_arg (HsRecFields { rec_flds = flds, rec_dotdot = dotdot }) ; return [ L (noAnnSrcSpan loc) (HsFieldBind { hfbAnn = noAnn , hfbLHS - = L (noAnnSrcSpan loc) (FieldOcc sel (L (noAnnSrcSpan loc) arg_rdr)) + = L (noAnnSrcSpan loc) (FieldOcc arg_rdr (L (noAnnSrcSpan loc) sel)) , hfbRHS = L locn (mk_arg loc arg_rdr) , hfbPun = False }) | fl <- dot_dot_fields @@ -1013,10 +1049,10 @@ rnHsRecUpdFields flds -> RnM ([LHsRecUpdField GhcRn GhcRn], FreeVars) rn_flds _ _ [] = return ([], emptyFVs) rn_flds pun_ok mb_unambig_lbls - ((L l (HsFieldBind { hfbLHS = L loc f + ((L l (HsFieldBind { hfbLHS = L loc (FieldOcc _ f) , hfbRHS = arg , hfbPun = pun })):flds) - = do { let lbl = ambiguousFieldOccRdrName f + = do { let lbl = unLoc f ; (arg' :: LHsExpr GhcPs) <- if pun then do { setSrcSpanA loc $ checkErr pun_ok (TcRnIllegalFieldPunning (L (locA loc) lbl)) @@ -1025,12 +1061,12 @@ rnHsRecUpdFields flds ; return (L (l2l loc) (HsVar noExtField (L (l2l loc) arg_rdr))) } else return arg ; (arg'', fvs) <- rnLExpr arg' - ; let lbl' :: AmbiguousFieldOcc GhcRn + ; let lbl' :: FieldOcc GhcRn lbl' = case mb_unambig_lbls of { Just (fl:_) -> let sel_name = flSelector fl - in Unambiguous sel_name (L (l2l loc) lbl) - ; _ -> Ambiguous noExtField (L (l2l loc) lbl) } + in FieldOcc lbl (L (l2l loc) sel_name) + ; _ -> XFieldOcc (Ambiguous (L (l2l loc) lbl)) } fld' :: LHsRecUpdField GhcRn GhcRn fld' = L l (HsFieldBind { hfbAnn = noAnn , hfbLHS = L loc lbl' @@ -1042,9 +1078,10 @@ rnHsRecUpdFields flds getFieldIds :: [LHsRecField GhcRn arg] -> [Name] getFieldIds flds = map (hsRecFieldSel . unLoc) flds -getFieldLbls :: forall p arg . UnXRec p => [LHsRecField p arg] -> [RdrName] -getFieldLbls flds - = map (unXRec @p . foLabel . unXRec @p . hfbLHS . unXRec @p) flds +-- The call to foLabel might be partial now. Don't know enough about +-- the rest of the function chain to say if this is an issue. +getFieldLbls :: forall p arg . UnXRec p => [LHsRecField p arg] -> [IdP p] +getFieldLbls = map (unXRec @p . foLabel . unXRec @p . hfbLHS . unXRec @p) needFlagDotDot :: HsRecFieldContext -> TcRnMessage needFlagDotDot = TcRnIllegalWildcardsInRecord . toRecordFieldPart ===================================== compiler/GHC/Tc/Gen/Expr.hs ===================================== @@ -1301,7 +1301,7 @@ expandRecordUpd record_expr possible_parents rbnds res_ty ; (cons, rbinds) <- disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty ; let upd_flds = map (unLoc . hfbLHS . unLoc) rbinds - sel_ids = map selectorAmbiguousFieldOcc upd_flds + sel_ids = map (unLoc . foLabel) upd_flds upd_fld_names = map idName sel_ids relevant_cons = nonDetEltsUniqSet cons relevant_con = head relevant_cons @@ -1584,7 +1584,7 @@ disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty -> TcM (LHsRecUpdField GhcTc GhcRn) lookupField fld_gre (L l upd) = do { let L loc af = hfbLHS upd - lbl = ambiguousFieldOccRdrName af + lbl = fieldOccRdrName af mb_gre = pickGREs lbl [fld_gre] -- NB: this GRE can be 'Nothing' when in GHCi. -- See test T10439. @@ -1596,7 +1596,7 @@ disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty ; sel <- tcLookupId (greName fld_gre) ; return $ L l HsFieldBind { hfbAnn = hfbAnn upd - , hfbLHS = L (l2l loc) $ Unambiguous sel (L (l2l loc) lbl) + , hfbLHS = L (l2l loc) $ FieldOcc lbl (L (l2l loc) sel) , hfbRHS = hfbRHS upd , hfbPun = hfbPun upd } } @@ -1669,11 +1669,11 @@ fieldCtxt field_name tcRecordField :: ConLike -> Assoc Name Type -> LFieldOcc GhcRn -> LHsExpr GhcRn -> TcM (Maybe (LFieldOcc GhcTc, LHsExpr GhcTc)) -tcRecordField con_like flds_w_tys (L loc (FieldOcc sel_name lbl)) rhs +tcRecordField con_like flds_w_tys (L loc (FieldOcc rdr (L l sel_name))) rhs | Just field_ty <- assocMaybe flds_w_tys sel_name = addErrCtxt (fieldCtxt field_lbl) $ do { rhs' <- tcCheckPolyExprNC rhs field_ty - ; hasFixedRuntimeRep_syntactic (FRRRecordCon (unLoc lbl) (unLoc rhs')) + ; hasFixedRuntimeRep_syntactic (FRRRecordCon rdr (unLoc rhs')) field_ty ; let field_id = mkUserLocal (nameOccName sel_name) (nameUnique sel_name) @@ -1682,12 +1682,13 @@ tcRecordField con_like flds_w_tys (L loc (FieldOcc sel_name lbl)) rhs -- (so we can find it easily) -- but is a LocalId with the appropriate type of the RHS -- (so the expansion knows the type of local binder to make) - ; return (Just (L loc (FieldOcc field_id lbl), rhs')) } + ; return (Just (L loc (FieldOcc rdr (L l field_id)), rhs')) } | otherwise = do { addErrTc (badFieldConErr (getName con_like) field_lbl) ; return Nothing } where - field_lbl = FieldLabelString $ occNameFS $ rdrNameOcc (unLoc lbl) + field_lbl = FieldLabelString $ occNameFS $ rdrNameOcc rdr +tcRecordField _ _ (L _ (XFieldOcc (Ambiguous _))) _ = pure Nothing checkMissingFields :: ConLike -> HsRecordBinds GhcRn -> [Scaled TcType] -> TcM () ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -567,7 +567,7 @@ tcInferAppHead_maybe :: HsExpr GhcRn tcInferAppHead_maybe fun = case fun of HsVar _ (L _ nm) -> Just <$> tcInferId nm - HsRecSel _ f -> Just <$> tcInferRecSelId f + HsRecSel _ f -> tcInferRecSelId f ExprWithTySig _ e hs_ty -> Just <$> tcExprWithSig e hs_ty HsOverLit _ lit -> Just <$> tcInferOverLit lit _ -> return Nothing @@ -596,21 +596,20 @@ addHeadCtxt fun_ctxt thing_inside ********************************************************************* -} tcInferRecSelId :: FieldOcc GhcRn - -> TcM (HsExpr GhcTc, TcSigmaType) -tcInferRecSelId (FieldOcc sel_name lbl) - = do { sel_id <- tc_rec_sel_id - ; let expr = HsRecSel noExtField (FieldOcc sel_id lbl) - ; return (expr, idType sel_id) + -> TcM (Maybe (HsExpr GhcTc, TcSigmaType)) +tcInferRecSelId (FieldOcc sel_name (L l n)) + = do { sel_id <- tc_rec_sel_id + ; let expr = HsRecSel noExtField (FieldOcc sel_name (L l sel_id)) + ; return $ Just(expr, idType sel_id) } where occ :: OccName - occ = rdrNameOcc (unLoc lbl) - + occ = nameOccName n tc_rec_sel_id :: TcM TcId -- Like tc_infer_id, but returns an Id not a HsExpr, -- so we can wrap it back up into a HsRecSel tc_rec_sel_id - = do { thing <- tcLookup sel_name + = do { thing <- tcLookup n ; case thing of ATcId { tct_id = id } -> do { check_naughty occ id -- See Note [Local record selectors] @@ -625,6 +624,7 @@ tcInferRecSelId (FieldOcc sel_name lbl) -- hence no checkTh stuff here _ -> failWithTc $ TcRnExpectedValueId thing } +tcInferRecSelId (XFieldOcc _) = pure Nothing ------------------------ ===================================== compiler/GHC/Tc/Gen/Pat.hs ===================================== @@ -1645,15 +1645,22 @@ tcConValArgs con_like arg_tys penv con_args thing_inside = case con_args of where tc_field :: Checker (LHsRecField GhcRn (LPat GhcRn)) (LHsRecField GhcTc (LPat GhcTc)) - tc_field penv - (L l (HsFieldBind ann (L loc (FieldOcc sel (L lr rdr))) pat pun)) - thing_inside + tc_field penv + (L l (HsFieldBind ann (L loc (FieldOcc rdr (L lr sel))) pat pun)) + thing_inside = do { sel' <- tcLookupId sel ; pat_ty <- setSrcSpanA loc $ find_field_ty sel (occNameFS $ rdrNameOcc rdr) ; (pat', res) <- tcConArg penv (pat, pat_ty) thing_inside - ; return (L l (HsFieldBind ann (L loc (FieldOcc sel' (L lr rdr))) pat' + ; return (L l (HsFieldBind ann (L loc (FieldOcc rdr (L lr sel'))) pat' pun), res) } + tc_field _ + (L _ (HsFieldBind _ (L _ (XFieldOcc (Ambiguous (L _ _)))) _ _)) + _ + -- I don't like leaving things undefined, and I don't like leaving + -- leaving pattern not matched. I think I should just throw in this + -- case, but I don't know exactly how to do that. + = undefined -- See Note [Omitted record fields and linearity] check_omitted_fields_multiplicity :: TcM MultiplicityCheckCoercions check_omitted_fields_multiplicity = do @@ -1682,7 +1689,7 @@ tcConValArgs con_like arg_tys penv con_args thing_inside = case con_args of (bound_field_tys, omitted_field_tys) = partition is_bound all_field_tys is_bound :: (Maybe FieldLabel, Scaled TcType) -> Bool - is_bound (Just fl, _) = elem (flSelector fl) (map (\(L _ (HsFieldBind _ (L _ (FieldOcc sel _ )) _ _)) -> sel) rpats) + is_bound (Just fl, _) = elem (flSelector fl) (map (\(L _ (HsFieldBind _ (L _ (FieldOcc _ sel )) _ _)) -> unLoc sel) rpats) is_bound _ = False all_field_tys :: [(Maybe FieldLabel, Scaled TcType)] ===================================== compiler/GHC/Tc/TyCl/Utils.hs ===================================== @@ -946,8 +946,7 @@ mkOneRecordSelector all_cons idDetails fl has_sel rec_field = noLocA (HsFieldBind { hfbAnn = noAnn , hfbLHS - = L locc (FieldOcc sel_name - (L locn $ mkRdrUnqual (nameOccName sel_name))) + = L locc (FieldOcc (mkRdrUnqual $ nameOccName sel_name) (L locn sel_name)) , hfbRHS = L loc' (VarPat noExtField (L locn field_var)) , hfbPun = False }) ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hoogle.hs ===================================== @@ -300,7 +300,7 @@ ppCtor sDocContext dat subdocs con at ConDeclH98{con_args = con_args'} = f (RecCon (L _ recs)) = f (PrefixCon [] $ map (hsLinear . cd_fld_type . unLoc) recs) ++ concat - [ (concatMap (lookupCon sDocContext subdocs . noLocA . foExt . unLoc) (cd_fld_names r)) + [ (concatMap (lookupCon sDocContext subdocs . noLocA . unLoc . foLabel . unLoc) (cd_fld_names r)) ++ [out sDocContext (map (foExt . unLoc) $ cd_fld_names r) `typeSig` [resType, cd_fld_type r]] | r <- map unLoc recs ] ===================================== utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs ===================================== @@ -1027,7 +1027,7 @@ ppSideBySideConstr subdocs unicode leader (L _ con) = ppSideBySideField :: [(DocName, DocForDecl DocName)] -> Bool -> ConDeclField DocNameI -> LaTeX ppSideBySideField subdocs unicode (ConDeclField _ names ltype _) = decltt - ( cat (punctuate comma (map (ppBinder . rdrNameOcc . unLoc . foLabel . unLoc) names)) + ( cat (punctuate comma (map (ppBinder . rdrNameOcc . foExt . unLoc) names)) <+> dcolon unicode <+> ppLType unicode ltype ) @@ -1035,7 +1035,7 @@ ppSideBySideField subdocs unicode (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc name) subdocs >>= fmap _doc . combineDocumentation . fst + mbDoc = lookup (unLoc . foLabel . unLoc $ name) subdocs >>= fmap _doc . combineDocumentation . fst name = case Maybe.listToMaybe names of Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs ===================================== @@ -1536,7 +1536,7 @@ ppSideBySideField subdocs unicode qual (ConDeclField _ names ltype _) = comma [ ppBinder False (rdrNameOcc field) | L _ name <- names - , let field = (unLoc . foLabel) name + , let field = (foExt) name ] ) <+> dcolon unicode @@ -1547,14 +1547,14 @@ ppSideBySideField subdocs unicode qual (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc declName) subdocs >>= combineDocumentation . fst + mbDoc = lookup (unLoc . foLabel $ unLoc declName) subdocs >>= combineDocumentation . fst declName = case Maybe.listToMaybe names of Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" Just hd -> hd ppShortField :: Bool -> Unicode -> Qualification -> ConDeclField DocNameI -> Html ppShortField summary unicode qual (ConDeclField _ names ltype _) = - hsep (punctuate comma (map ((ppBinder summary) . rdrNameOcc . unLoc . foLabel . unLoc) names)) + hsep (punctuate comma (map ((ppBinder summary) . rdrNameOcc . foExt . unLoc) names)) <+> dcolon unicode <+> ppLType unicode qual HideEmptyContexts ltype ===================================== utils/haddock/haddock-api/src/Haddock/Convert.hs ===================================== @@ -496,7 +496,7 @@ synifyDataCon use_gadt_syntax dc = noLocA $ ConDeclField noAnn - [noLocA $ FieldOcc (flSelector fl) (noLocA $ mkVarUnqual $ field_label $ flLabel fl)] + [noLocA $ FieldOcc (mkVarUnqual $ field_label $ flLabel fl) (noLocA (flSelector fl))] synTy Nothing ===================================== utils/haddock/haddock-api/src/Haddock/GhcUtils.hs ===================================== @@ -374,7 +374,7 @@ restrictCons names decls = [L p d | L p (Just d) <- fmap keep <$> decls] field_avail :: LConDeclField GhcRn -> Bool field_avail (L _ (ConDeclField _ fs _ _)) = - all (\f -> foExt (unLoc f) `elem` names) fs + all (\f -> (unLoc . foLabel . unLoc $ f) `elem` names) fs field_types flds = [hsUnrestricted t | L _ (ConDeclField _ _ t _) <- flds] keep _ = Nothing @@ -563,7 +563,7 @@ instance Parent (ConDecl GhcRn) where children con = case getRecConArgs_maybe con of Nothing -> [] - Just flds -> map (foExt . unLoc) $ concatMap (cd_fld_names . unLoc) (unLoc flds) + Just flds -> map (unLoc . foLabel . unLoc) $ concatMap (cd_fld_names . unLoc) (unLoc flds) instance Parent (TyClDecl GhcRn) where children d ===================================== utils/haddock/haddock-api/src/Haddock/Interface/Create.hs ===================================== @@ -927,7 +927,7 @@ extractDecl prr dflags sDocContext name decl Just rec <- toList $ getRecConArgs_maybe . unLoc <$> dd_cons (feqn_rhs d) , ConDeclField{cd_fld_names = ns} <- map unLoc (unLoc rec) , L _ n <- ns - , foExt n == name + , unLoc (foLabel n) == name ] in case matches of [d0] -> extractDecl prr dflags sDocContext name (noLocA . InstD noExtField $ DataFamInstD noExtField d0) @@ -996,7 +996,7 @@ extractRecSel nm t tvs (L _ con : rest) = where matching_fields :: [LConDeclField GhcRn] -> [(SrcSpan, LConDeclField GhcRn)] matching_fields flds = - [ (locA l, f) | f@(L _ (ConDeclField _ ns _ _)) <- flds, L l n <- ns, foExt n == nm + [ (locA l, f) | f@(L _ (ConDeclField _ ns _ _)) <- flds, L l n <- ns, unLoc (foLabel n) == nm ] data_ty -- ResTyGADT _ ty <- con_res con = ty ===================================== utils/haddock/haddock-api/src/Haddock/Interface/Rename.hs ===================================== @@ -745,9 +745,9 @@ renameConDeclFieldField (L l (ConDeclField _ names t doc)) = do return $ L (locA l) (ConDeclField noExtField names' t' doc') renameLFieldOcc :: LFieldOcc GhcRn -> RnM (LFieldOcc DocNameI) -renameLFieldOcc (L l (FieldOcc sel lbl)) = do +renameLFieldOcc (L l (FieldOcc rdr (L n sel))) = do sel' <- renameName sel - return $ L l (FieldOcc sel' lbl) + return $ L l (FieldOcc rdr (L n sel')) renameSig :: Sig GhcRn -> RnM (Sig DocNameI) renameSig sig = case sig of ===================================== utils/haddock/haddock-api/src/Haddock/Types.hs ===================================== @@ -885,8 +885,8 @@ type instance XUserTyVar DocNameI = NoExtField type instance XKindedTyVar DocNameI = NoExtField type instance XXTyVarBndr DocNameI = DataConCantHappen -type instance XCFieldOcc DocNameI = DocName -type instance XXFieldOcc DocNameI = NoExtField +type instance XCFieldOcc DocNameI = RdrName +type instance XXFieldOcc DocNameI = DataConCantHappen type instance XFixitySig DocNameI = NoExtField type instance XFixSig DocNameI = NoExtField View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fd237dfed07d4d9975c1cd226d2bfd73e04bc482 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fd237dfed07d4d9975c1cd226d2bfd73e04bc482 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 23:12:54 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Thu, 19 Sep 2024 19:12:54 -0400 Subject: [Git][ghc/ghc][wip/T25066] 166 commits: AArch64: Simplify stmtToInstrs type Message-ID: <66ecaff699f67_2d2c8f7dad48958b9@gitlab.mail> Ben Gamari pushed to branch wip/T25066 at Glasgow Haskell Compiler / GHC Commits: ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - 9b7850a8 by Ben Gamari at 2024-09-19T15:47:26-04:00 base: Add test for #25066 - - - - - 76b91a10 by Ben Gamari at 2024-09-19T19:09:28-04:00 base: Fix #25066 As noted in #25066, the exception backtrace proposal introduced a rather subtle performance regression due to simplification producing Core which the demand analyser concludes may diverge with a precise exception. The nature of the problem is more completely described in the new Note [Hiding precise exception signature in throw]. The (rather hacky) solution we use here hides the problematic optimisation through judicious use of `noinline`. Ultimately however we will want a more principled solution (e.g. #23847). - - - - - 28 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC.hs - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/GenericOpt.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/35251a17de9b44b30d4dd39e657033b02bc5aedc...76b91a10f371f150c72c53f8ec17a56c416e0a67 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/35251a17de9b44b30d4dd39e657033b02bc5aedc...76b91a10f371f150c72c53f8ec17a56c416e0a67 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 23:15:53 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Thu, 19 Sep 2024 19:15:53 -0400 Subject: [Git][ghc/ghc][wip/T25066] 2 commits: base: Add test for #25066 Message-ID: <66ecb0a93d763_2d2c8f7dad341013af@gitlab.mail> Ben Gamari pushed to branch wip/T25066 at Glasgow Haskell Compiler / GHC Commits: 262f8dca by Ben Gamari at 2024-09-19T19:15:36-04:00 base: Add test for #25066 - - - - - 63b1776b by Ben Gamari at 2024-09-19T19:15:36-04:00 base: Fix #25066 As noted in #25066, the exception backtrace proposal introduced a rather subtle performance regression due to simplification producing Core which the demand analyser concludes may diverge with a precise exception. The nature of the problem is more completely described in the new Note [Hiding precise exception signature in throw]. The (rather hacky) solution we use here hides the problematic optimisation through judicious use of `noinline`. Ultimately however we will want a more principled solution (e.g. #23847). - - - - - 4 changed files: - + libraries/base/tests/T25066.hs - + libraries/base/tests/T25066.stderr - libraries/base/tests/all.T - libraries/ghc-internal/src/GHC/Internal/Exception.hs Changes: ===================================== libraries/base/tests/T25066.hs ===================================== @@ -0,0 +1,15 @@ +-- | Check that the demand signature of 'throw' doesn't suggest that it will +-- throw a precise exception. Specifically, `g` should have a `b` divergence +-- type in its demand signature. + +module T25066 (g) where + +import Control.Exception + +data MyException = MyException + deriving (Show) + +instance Exception MyException + +g :: a +g = throw MyException ===================================== libraries/base/tests/T25066.stderr ===================================== @@ -0,0 +1,20 @@ + +==================== Demand signatures ==================== +T25066.$fExceptionMyException: +T25066.$fShowMyException: +T25066.$tc'MyException: +T25066.$tcMyException: +T25066.$trModule: +T25066.g: b + + + +==================== Demand signatures ==================== +T25066.$fExceptionMyException: +T25066.$fShowMyException: +T25066.$tc'MyException: +T25066.$tcMyException: +T25066.$trModule: +T25066.g: b + + ===================================== libraries/base/tests/all.T ===================================== @@ -324,3 +324,4 @@ test('T23697', , when(opsys('darwin'), skip) # permission denied ], makefile_test, ['T23697']) test('stimesEndo', normal, compile_and_run, ['']) +test('T25066', [only_ways(['optasm']), grep_errmsg('T25066.g')], compile, ['-ddump-dmd-signatures']) ===================================== libraries/ghc-internal/src/GHC/Internal/Exception.hs ===================================== @@ -81,9 +81,58 @@ import GHC.Internal.Exception.Type throw :: forall (r :: RuntimeRep). forall (a :: TYPE r). forall e. (HasCallStack, Exception e) => e -> a throw e = - let !se = unsafePerformIO (toExceptionWithBacktrace e) + -- See Note [Hiding precise exception signature in throw] + let !se = noinline (unsafePerformIO (toExceptionWithBacktrace e)) in raise# se +-- Note [Hiding precise exception signature in throw] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- In 'throw' we use `unsafePerformIO . toExceptionWithBacktrace' to collect +-- the backtraces which will be attached as the exception's 'ExceptionContext'. +-- We must ensure that this is evaluated immediately in throw since `toExceptionWithBacktrace` +-- must capture the execution state at the moment that the exception is thrown. +-- Unfortunately, unless we take particular care this can lead to a catastrophic +-- regression in 'throw's demand signature which will infect all callers (#25066) +-- +-- Specifically, GHC's demand analysis has an approximate heuristic for tracking +-- whether divergent functions diverge with precise or imprecise exceptions (namely +-- the 'ExnOrDiv' and 'Diverges' constructors of 'GHC.Types.Demand.Divergence', respectively). This is because we can take considerably more liberties in optimising around +-- functions which are known not to diverge via precise exception (see +-- Note [Precise exceptions and strictness analysis]). For this reason, it is +-- important that 'throw' have a 'Diverges' divergence type. +-- +-- Unfortunately, this is broken if we allow `unsafePerformIO` to inline. Specifically, +-- if we allow this inlining we will end up with Core of the form: +-- +-- throw = \e -> +-- case runRW# (\s -> ... toExceptionWithBacktrace e s ...) of +-- se -> raise# se +-- +-- so far this is fine; the demand analyzer's divergence heuristic +-- will give 'throw' the expected 'Diverges' divergence. +-- +-- However, the simplifier will subsequently notice that `raise#` can be fruitfully +-- floated into the body of the `runRW#`: +-- +-- throw = \e -> +-- runRW# (\s -> case toExceptionWithBacktrace e s of +-- (# s', se #) -> raise# se) +-- +-- This is problematic as one of the demand analyser's heuristics +-- examines the `case` scrutinees, looking for those that result in a RealWorld# +-- token (see Note [Which scrutinees may throw precise exceptions], test (1)). +-- The `case toExceptionWithBacktrace e of ...` case fails this check, causing the +-- heuristic to conclude that `throw` may indeed diverge with a precise exception. +-- This resulted in the significant performance regression noted in #25066. +-- +-- To avoid this, we use `noinline` to ensure that `unsafePerformIO` does not unfold, +-- meaning that the `raise#` cannot be floated under the `toExceptionWithBacktrace` +-- case analysis. +-- +-- Ultimately this is a bit of a horrible hack; the right solution would be to have +-- primops which allow more precise guidance of the demand analyser's heuristic +-- (e.g. #23847). + -- | @since base-4.20.0.0 toExceptionWithBacktrace :: (HasCallStack, Exception e) => e -> IO SomeException View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/76b91a10f371f150c72c53f8ec17a56c416e0a67...63b1776b68b3284368e784ee8493efeb1d079d29 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/76b91a10f371f150c72c53f8ec17a56c416e0a67...63b1776b68b3284368e784ee8493efeb1d079d29 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 23:19:31 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Thu, 19 Sep 2024 19:19:31 -0400 Subject: [Git][ghc/ghc][wip/T25066] base: Fix #25066 Message-ID: <66ecb183ef68b_2d2c8f7d2328101798@gitlab.mail> Ben Gamari pushed to branch wip/T25066 at Glasgow Haskell Compiler / GHC Commits: 440d9c60 by Ben Gamari at 2024-09-19T19:19:24-04:00 base: Fix #25066 As noted in #25066, the exception backtrace proposal introduced a rather subtle performance regression due to simplification producing Core which the demand analyser concludes may diverge with a precise exception. The nature of the problem is more completely described in the new Note [Hiding precise exception signature in throw]. The (rather hacky) solution we use here hides the problematic optimisation through judicious use of `noinline`. Ultimately however we will want a more principled solution (e.g. #23847). - - - - - 2 changed files: - libraries/base/tests/T25066.stderr - libraries/ghc-internal/src/GHC/Internal/Exception.hs Changes: ===================================== libraries/base/tests/T25066.stderr ===================================== @@ -5,7 +5,7 @@ T25066.$fShowMyException: T25066.$tc'MyException: T25066.$tcMyException: T25066.$trModule: -T25066.g: x +T25066.g: b @@ -15,6 +15,6 @@ T25066.$fShowMyException: T25066.$tc'MyException: T25066.$tcMyException: T25066.$trModule: -T25066.g: x +T25066.g: b ===================================== libraries/ghc-internal/src/GHC/Internal/Exception.hs ===================================== @@ -81,9 +81,61 @@ import GHC.Internal.Exception.Type throw :: forall (r :: RuntimeRep). forall (a :: TYPE r). forall e. (HasCallStack, Exception e) => e -> a throw e = - let !se = unsafePerformIO (toExceptionWithBacktrace e) + -- See Note [Hiding precise exception signature in throw] + let !se = noinline (unsafePerformIO (toExceptionWithBacktrace e)) in raise# se +-- Note [Hiding precise exception signature in throw] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- In 'throw' we use `unsafePerformIO . toExceptionWithBacktrace' to collect +-- the backtraces which will be attached as the exception's 'ExceptionContext'. +-- We must ensure that this is evaluated immediately in throw since +-- `toExceptionWithBacktrace` must capture the execution state at the moment +-- that the exception is thrown. Unfortunately, unless we take particular +-- care this can lead to a catastrophic regression in 'throw's demand signature +-- which will infect all callers (#25066) +-- +-- Specifically, GHC's demand analysis has an approximate heuristic for tracking +-- whether divergent functions diverge with precise or imprecise exceptions (namely +-- the 'ExnOrDiv' and 'Diverges' constructors of 'GHC.Types.Demand.Divergence', +-- respectively). This is because we can take considerably more liberties in +-- optimising around functions which are known not to diverge via precise +-- exception (see Note [Precise exceptions and strictness analysis]). +-- For this reason, it is important that 'throw' have a 'Diverges' divergence +-- type. +-- +-- Unfortunately, this is broken if we allow `unsafePerformIO` to inline. Specifically, +-- if we allow this inlining we will end up with Core of the form: +-- +-- throw = \e -> +-- case runRW# (\s -> ... toExceptionWithBacktrace e s ...) of +-- se -> raise# se +-- +-- so far this is fine; the demand analyzer's divergence heuristic +-- will give 'throw' the expected 'Diverges' divergence. +-- +-- However, the simplifier will subsequently notice that `raise#` can be fruitfully +-- floated into the body of the `runRW#`: +-- +-- throw = \e -> +-- runRW# (\s -> case toExceptionWithBacktrace e s of +-- (# s', se #) -> raise# se) +-- +-- This is problematic as one of the demand analyser's heuristics +-- examines the `case` scrutinees, looking for those that result in a RealWorld# +-- token (see Note [Which scrutinees may throw precise exceptions], test (1)). +-- The `case toExceptionWithBacktrace e of ...` case fails this check, causing the +-- heuristic to conclude that `throw` may indeed diverge with a precise exception. +-- This resulted in the significant performance regression noted in #25066. +-- +-- To avoid this, we use `noinline` to ensure that `unsafePerformIO` does not unfold, +-- meaning that the `raise#` cannot be floated under the `toExceptionWithBacktrace` +-- case analysis. +-- +-- Ultimately this is a bit of a horrible hack; the right solution would be to have +-- primops which allow more precise guidance of the demand analyser's heuristic +-- (e.g. #23847). + -- | @since base-4.20.0.0 toExceptionWithBacktrace :: (HasCallStack, Exception e) => e -> IO SomeException View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/440d9c6077b56858d4f3637d6c9480bbb7d2ff3d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/440d9c6077b56858d4f3637d6c9480bbb7d2ff3d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 19 23:24:33 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Thu, 19 Sep 2024 19:24:33 -0400 Subject: [Git][ghc/ghc][wip/T25066] base: Fix #25066 Message-ID: <66ecb2b146b41_2d2c8f7d60401038a4@gitlab.mail> Ben Gamari pushed to branch wip/T25066 at Glasgow Haskell Compiler / GHC Commits: 8945337e by Ben Gamari at 2024-09-19T19:24:21-04:00 base: Fix #25066 As noted in #25066, the exception backtrace proposal introduced a rather subtle performance regression due to simplification producing Core which the demand analyser concludes may diverge with a precise exception. The nature of the problem is more completely described in the new Note [Hiding precise exception signature in throw]. The (rather hacky) solution we use here hides the problematic optimisation through judicious use of `noinline`. Ultimately however we will want a more principled solution (e.g. #23847). - - - - - 2 changed files: - libraries/base/tests/T25066.stderr - libraries/ghc-internal/src/GHC/Internal/Exception.hs Changes: ===================================== libraries/base/tests/T25066.stderr ===================================== @@ -5,7 +5,7 @@ T25066.$fShowMyException: T25066.$tc'MyException: T25066.$tcMyException: T25066.$trModule: -T25066.g: x +T25066.g: b @@ -15,6 +15,6 @@ T25066.$fShowMyException: T25066.$tc'MyException: T25066.$tcMyException: T25066.$trModule: -T25066.g: x +T25066.g: b ===================================== libraries/ghc-internal/src/GHC/Internal/Exception.hs ===================================== @@ -81,9 +81,61 @@ import GHC.Internal.Exception.Type throw :: forall (r :: RuntimeRep). forall (a :: TYPE r). forall e. (HasCallStack, Exception e) => e -> a throw e = - let !se = unsafePerformIO (toExceptionWithBacktrace e) + -- See Note [Hiding precise exception signature in throw] + let !se = noinline (unsafePerformIO (toExceptionWithBacktrace e)) in raise# se +-- Note [Hiding precise exception signature in throw] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- In 'throw' we use `unsafePerformIO . toExceptionWithBacktrace' to collect +-- the backtraces which will be attached as the exception's 'ExceptionContext'. +-- We must ensure that this is evaluated immediately in throw since +-- `toExceptionWithBacktrace` must capture the execution state at the moment +-- that the exception is thrown. Unfortunately, unless we take particular +-- care this can lead to a catastrophic regression in 'throw's demand signature +-- which will infect all callers (#25066) +-- +-- Specifically, GHC's demand analysis has an approximate heuristic for tracking +-- whether divergent functions diverge with precise or imprecise exceptions (namely +-- the 'ExnOrDiv' and 'Diverges' constructors of 'GHC.Types.Demand.Divergence', +-- respectively). This is because we can take considerably more liberties in +-- optimising around functions which are known not to diverge via precise +-- exception (see Note [Precise exceptions and strictness analysis]). +-- For this reason, it is important that 'throw' have a 'Diverges' divergence +-- type. +-- +-- Unfortunately, this is broken if we allow `unsafePerformIO` to inline. Specifically, +-- if we allow this inlining we will end up with Core of the form: +-- +-- throw = \e -> +-- case runRW# (\s -> ... toExceptionWithBacktrace e s ...) of +-- se -> raise# se +-- +-- so far this is fine; the demand analyzer's divergence heuristic +-- will give 'throw' the expected 'Diverges' divergence. +-- +-- However, the simplifier will subsequently notice that `raise#` can be fruitfully +-- floated into the body of the `runRW#`: +-- +-- throw = \e -> +-- runRW# (\s -> case toExceptionWithBacktrace e s of +-- (# s', se #) -> raise# se) +-- +-- This is problematic as one of the demand analyser's heuristics +-- examines `case` scrutinees, looking for those that result in a `RealWorld#` +-- token (see Note [Which scrutinees may throw precise exceptions], test (1)). +-- The `case toExceptionWithBacktrace e of ...` here fails this check, causing the +-- heuristic to conclude that `throw` may indeed diverge with a precise exception. +-- This resulted in the significant performance regression noted in #25066. +-- +-- To avoid this, we use `noinline` to ensure that `unsafePerformIO` does not unfold, +-- meaning that the `raise#` cannot be floated under the `toExceptionWithBacktrace` +-- case analysis. +-- +-- Ultimately this is a bit of a horrible hack; the right solution would be to have +-- primops which allow more precise guidance of the demand analyser's heuristic +-- (e.g. #23847). + -- | @since base-4.20.0.0 toExceptionWithBacktrace :: (HasCallStack, Exception e) => e -> IO SomeException View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8945337e0d4c400040c3257cfc4a70889c7b05a5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8945337e0d4c400040c3257cfc4a70889c7b05a5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 02:00:38 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Thu, 19 Sep 2024 22:00:38 -0400 Subject: [Git][ghc/ghc][wip/strings] Replace manual string lexing (#25158) Message-ID: <66ecd7469396b_2d2c8f129cfb0119895@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC Commits: 05589606 by Brandon Chinn at 2024-09-19T19:00:03-07:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 26 changed files: - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - testsuite/driver/testlib.py - testsuite/tests/ghci/prog013/prog013.stderr - testsuite/tests/ghci/scripts/ghci022.stderr - testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr - testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr - testsuite/tests/parser/should_fail/T21843c.stderr - testsuite/tests/parser/should_fail/T21843e.stderr - testsuite/tests/parser/should_fail/T21843f.stderr - testsuite/tests/parser/should_fail/T3751.stderr - testsuite/tests/parser/should_fail/T5425.stderr - testsuite/tests/parser/should_fail/readFail002.stderr - testsuite/tests/parser/should_fail/readFail004.stderr - testsuite/tests/parser/should_fail/readFail005.stderr - testsuite/tests/parser/should_fail/readFail033.stderr - testsuite/tests/parser/unicode/utf8_010.stderr - testsuite/tests/parser/unicode/utf8_011.stderr - testsuite/tests/parser/unicode/utf8_020.stderr - testsuite/tests/parser/unicode/utf8_021.stderr - testsuite/tests/parser/unicode/utf8_022.stderr - utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs Changes: ===================================== compiler/GHC/Parser.y ===================================== @@ -728,7 +728,7 @@ are the most common patterns, rewritten as regular expressions for clarity: CHAR { L _ (ITchar _ _) } STRING { L _ (ITstring _ _) } - MULTILINESTRING { L _ (ITmultilinestring _ _) } + STRING_MULTI { L _ (ITstringMulti _ _) } INTEGER { L _ (ITinteger _) } RATIONAL { L _ (ITrational _) } @@ -2357,8 +2357,8 @@ atype :: { LHsType GhcPs } (getCHAR $1) } | STRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGs $1) (getSTRING $1) } - | MULTILINESTRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getMULTILINESTRINGs $1) - (getMULTILINESTRING $1) } + | STRING_MULTI { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGMULTIs $1) + (getSTRINGMULTI $1) } | '_' { sL1a $1 $ mkAnonWildCardTy } -- Type variables are never exported, so `M.tyvar` will be rejected by the renamer. -- We let it pass the parser because the renamer can generate a better error message. @@ -4047,8 +4047,8 @@ literal :: { Located (HsLit GhcPs) } : CHAR { sL1 $1 $ HsChar (getCHARs $1) $ getCHAR $1 } | STRING { sL1 $1 $ HsString (getSTRINGs $1) $ getSTRING $1 } - | MULTILINESTRING { sL1 $1 $ HsMultilineString (getMULTILINESTRINGs $1) - $ getMULTILINESTRING $1 } + | STRING_MULTI { sL1 $1 $ HsMultilineString (getSTRINGMULTIs $1) + $ getSTRINGMULTI $1 } | PRIMINTEGER { sL1 $1 $ HsIntPrim (getPRIMINTEGERs $1) $ getPRIMINTEGER $1 } | PRIMWORD { sL1 $1 $ HsWordPrim (getPRIMWORDs $1) @@ -4154,7 +4154,7 @@ getIPDUPVARID (L _ (ITdupipvarid x)) = x getLABELVARID (L _ (ITlabelvarid _ x)) = x getCHAR (L _ (ITchar _ x)) = x getSTRING (L _ (ITstring _ x)) = x -getMULTILINESTRING (L _ (ITmultilinestring _ x)) = x +getSTRINGMULTI (L _ (ITstringMulti _ x)) = x getINTEGER (L _ (ITinteger x)) = x getRATIONAL (L _ (ITrational x)) = x getPRIMCHAR (L _ (ITprimchar _ x)) = x @@ -4180,7 +4180,7 @@ getVOCURLY (L (RealSrcSpan l _) ITvocurly) = srcSpanStartCol l getINTEGERs (L _ (ITinteger (IL src _ _))) = src getCHARs (L _ (ITchar src _)) = src getSTRINGs (L _ (ITstring src _)) = src -getMULTILINESTRINGs (L _ (ITmultilinestring src _)) = src +getSTRINGMULTIs (L _ (ITstringMulti src _)) = src getPRIMCHARs (L _ (ITprimchar src _)) = src getPRIMSTRINGs (L _ (ITprimstring src _)) = src getPRIMINTEGERs (L _ (ITprimint src _)) = src ===================================== compiler/GHC/Parser/CharClass.hs ===================================== @@ -36,14 +36,14 @@ cDigit = 64 {-# INLINABLE is_ctype #-} is_ctype :: Word8 -> Char -> Bool -is_ctype mask c = (charType c .&. mask) /= 0 +is_ctype mask c = c <= '\127' && (charType c .&. mask) /= 0 is_ident, is_symbol, is_any, is_space, is_lower, is_upper, is_digit, is_alphanum :: Char -> Bool is_ident = is_ctype cIdent is_symbol = is_ctype cSymbol is_any = is_ctype cAny -is_space = \c -> c <= '\x7f' && is_ctype cSpace c -- is_space only works for <= '\x7f' (#3751, #5425) +is_space = is_ctype cSpace is_lower = is_ctype cLower is_upper = is_ctype cUpper is_digit = is_ctype cDigit ===================================== compiler/GHC/Parser/Errors/Ppr.hs ===================================== @@ -157,8 +157,6 @@ instance Diagnostic PsMessage where LexUnknownPragma -> text "unknown pragma" LexErrorInPragma -> text "lexical error in pragma" LexNumEscapeRange -> text "numeric escape sequence out of range" - LexStringCharLit -> text "lexical error in string/character literal" - LexStringCharLitEOF -> text "unexpected end-of-file in string/character literal" LexUnterminatedComment -> text "unterminated `{-'" LexUnterminatedOptions -> text "unterminated OPTIONS pragma" LexUnterminatedQQ -> text "unterminated quasiquotation" ===================================== compiler/GHC/Parser/Errors/Types.hs ===================================== @@ -591,8 +591,6 @@ data LexErr | LexUnknownPragma -- ^ Unknown pragma | LexErrorInPragma -- ^ Lexical error in pragma | LexNumEscapeRange -- ^ Numeric escape sequence out of range - | LexStringCharLit -- ^ Lexical error in string/character literal - | LexStringCharLitEOF -- ^ Unexpected end-of-file in string/character literal | LexUnterminatedComment -- ^ Unterminated `{-' | LexUnterminatedOptions -- ^ Unterminated OPTIONS pragma | LexUnterminatedQQ -- ^ Unterminated quasiquotation ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -167,6 +167,7 @@ $idchar = [$small $large $digit $uniidchar \'] $unigraphic = \x06 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $graphic = [$small $large $symbol $digit $idchar $special $unigraphic \"\'] +$charesc = [a b f n r t v \\ \" \' \&] $binit = 0-1 $octit = 0-7 @@ -213,6 +214,20 @@ $docsym = [\| \^ \* \$] @floating_point = @numspc @decimal \. @decimal @exponent? | @numspc @decimal @exponent @hex_floating_point = @numspc @hexadecimal \. @hexadecimal @bin_exponent? | @numspc @hexadecimal @bin_exponent + at gap = \\ $whitechar+ \\ + at cntrl = $asclarge | \@ | \[ | \\ | \] | \^ | \_ + at ascii = \^ @cntrl | "NUL" | "SOH" | "STX" | "ETX" | "EOT" | "ENQ" | "ACK" + | "BEL" | "BS" | "HT" | "LF" | "VT" | "FF" | "CR" | "SO" | "SI" | "DLE" + | "DC1" | "DC2" | "DC3" | "DC4" | "NAK" | "SYN" | "ETB" | "CAN" + | "EM" | "SUB" | "ESC" | "FS" | "GS" | "RS" | "US" | "SP" | "DEL" +-- N.B. ideally, we would do `@escape # \\ \&` instead of duplicating in @escapechar, +-- which is what the Haskell Report says, but this isn't valid Alex syntax, as only +-- character sets can be subtracted, not strings + at escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) + at escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) + at stringchar = ($graphic # [\\ \"]) | " " | @escape | @gap + at char = ($graphic # [\\ \']) | " " | @escapechar + -- normal signed numerical literals can only be explicitly negative, -- not explicitly positive (contrast @exponent) @negative = \- @@ -460,7 +475,7 @@ $unigraphic / { isSmartQuote } { smart_quote_error } <0> { "#" $idchar+ / { ifExtension OverloadedLabelsBit } { skip_one_varid_src ITlabelvarid } - "#" \" / { ifExtension OverloadedLabelsBit } { lex_quoted_label } + "#" \" @stringchar* \" / { ifExtension OverloadedLabelsBit } { tok_quoted_label } } <0> { @@ -660,14 +675,38 @@ $unigraphic / { isSmartQuote } { smart_quote_error } } --- Strings and chars are lexed by hand-written code. The reason is --- that even if we recognise the string or char here in the regex --- lexer, we would still have to parse the string afterward in order --- to convert it to a String. <0> { - \' { lex_char_tok } - \"\"\" / { ifExtension MultilineStringsBit} { lex_string_tok StringTypeMulti } - \" { lex_string_tok StringTypeSingle } + \"\"\" / { ifExtension MultilineStringsBit } { tok_string_multi } + \" @stringchar* \" { tok_string } + \" @stringchar* \" \# / { ifExtension MagicHashBit } { tok_string } + \' @char \' { tok_char } + \' @char \' \# / { ifExtension MagicHashBit } { tok_char } + + -- Check for smart quotes and throw better errors than a plain lexical error (#21843) + \' \\ $unigraphic / { isSmartQuote } { smart_quote_error } + \" @stringchar* \\ $unigraphic / { isSmartQuote } { smart_quote_error } + -- See Note [Bare smart quote error] + -- The valid string rule will take precedence because it'll match more + -- characters than this rule, so this rule will only fire if the string + -- could not be lexed correctly + \" @stringchar* $unigraphic / { isSmartQuote } { smart_quote_error } +} + + { + -- Parse as much of the multiline string as possible, except for quotes + @stringchar* ($nl ([\ $tab] | @gap)* @stringchar*)* { tok_string_multi_content } + -- Allow bare quotes if it's not a triple quote + (\" | \"\") / ([\n .] # \") { tok_string_multi_content } +} + +<0> { + \'\' { token ITtyQuote } + + -- The normal character match takes precedence over this because it matches + -- more characters. However, if that pattern didn't match, then this quote + -- could be a quoted identifier, like 'x. Here, just return ITsimpleQuote, + -- as the parser will lex the varid separately. + \' / ($graphic # \\ | " ") { token ITsimpleQuote } } -- Note [Whitespace-sensitive operator parsing] @@ -953,7 +992,7 @@ data Token | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" | ITstring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" - | ITmultilinestring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringMulti SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" | ITinteger IntegralLit -- Note [Literal source text] in "GHC.Types.SourceText" | ITrational FractionalLit @@ -2181,156 +2220,128 @@ lex_string_prag_comment mkTok span _buf _len _buf2 -- ----------------------------------------------------------------------------- -- Strings & Chars --- This stuff is horrible. I hates it. - -lex_string_tok :: LexStringType -> Action -lex_string_tok strType span buf _len _buf2 = do - s <- lex_string strType - - i <- getInput - case strType of - StringTypeSingle -> - lex_magic_hash i >>= \case - Just i' -> do - when (any (> '\xFF') s) $ do - pState <- getPState - let msg = PsErrPrimStringInvalidChar - let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg - addError err - - setInput i' - let (psSpan, src) = getStringLoc (buf, locStart) i' - pure $ L psSpan (ITprimstring src (unsafeMkByteString s)) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITstring src (mkFastString s)) - StringTypeMulti -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITmultilinestring src (mkFastString s)) +tok_string :: Action +tok_string span buf len _buf2 = do + s <- lex_chars ("\"", "\"") span buf (if endsInHash then len - 1 else len) + + if endsInHash + then do + when (any (> '\xFF') s) $ do + pState <- getPState + let msg = PsErrPrimStringInvalidChar + let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg + addError err + pure $ L span (ITprimstring src (unsafeMkByteString s)) + else + pure $ L span (ITstring src (mkFastString s)) where - locStart = psSpanStart span + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + +-- | Ideally, we would define this completely with Alex syntax, like normal strings. +-- Instead, this is defined as a hybrid solution by manually invoking lex states, which +-- we're doing for two reasons: +-- 1. The multiline string should all be one lexical token, not multiple +-- 2. We need to allow bare quotes, which can't be done with one regex +tok_string_multi :: Action +tok_string_multi startSpan startBuf _len _buf2 = do + -- advance to the end of the multiline string + let startLoc = psSpanStart startSpan + let i@(AI _ contentStartBuf) = + case lexDelim $ AI startLoc startBuf of + Just i -> i + Nothing -> panic "tok_string_multi did not start with a delimiter" + (AI _ contentEndBuf, i'@(AI endLoc endBuf)) <- goContent i + + -- build the values pertaining to the entire multiline string, including delimiters + let span = mkPsSpan startLoc endLoc + let len = byteDiff startBuf endBuf + let src = SourceText $ lexemeToFastString startBuf len + + -- load the content of the multiline string + let contentLen = byteDiff contentStartBuf contentEndBuf + s <- + either (throwStringLexError (AI startLoc startBuf)) pure $ + lexMultilineString contentLen contentStartBuf + + setInput i' + pure $ L span $ ITstringMulti src (mkFastString s) + where + goContent i0 = + case alexScan i0 string_multi_content of + AlexToken i1 len _ + | Just i2 <- lexDelim i1 -> pure (i1, i2) + | -- is the next token a tab character? + -- need this explicitly because there's a global rule matching $tab + Just ('\t', _) <- alexGetChar' i1 -> setInput i1 >> lexError LexError + | isEOF i1 -> checkSmartQuotes >> lexError LexError + | len == 0 -> panic $ "parsing multiline string got into infinite loop at: " ++ show i0 + | otherwise -> goContent i1 + AlexSkip i1 _ -> goContent i1 + _ -> lexError LexError + + lexDelim = + let go 0 i = Just i + go n i = + case alexGetChar' i of + Just ('"', i') -> go (n - 1) i' + _ -> Nothing + in go (3 :: Int) + + -- See Note [Bare smart quote error] + checkSmartQuotes = do + let findSmartQuote i0@(AI loc _) = + case alexGetChar' i0 of + Just ('\\', i1) | Just (_, i2) <- alexGetChar' i1 -> findSmartQuote i2 + Just (c, i1) + | isDoubleSmartQuote c -> Just (c, loc) + | otherwise -> findSmartQuote i1 + _ -> Nothing + case findSmartQuote (AI (psSpanStart startSpan) startBuf) of + Just (c, loc) -> throwSmartQuoteError c loc + Nothing -> pure () + +-- | Dummy action that should never be called. Should only be used in lex states +-- that are manually lexed in tok_string_multi. +tok_string_multi_content :: Action +tok_string_multi_content = panic "tok_string_multi_content unexpectedly invoked" + +lex_chars :: (String, String) -> PsSpan -> StringBuffer -> Int -> P String +lex_chars (startDelim, endDelim) span buf len = + either (throwStringLexError i0) pure $ + lexString contentLen contentBuf + where + i0@(AI _ contentBuf) = advanceInputBytes (length startDelim) $ AI (psSpanStart span) buf + -- assumes delimiters are ASCII, with 1 byte per Char + contentLen = len - length startDelim - length endDelim -lex_quoted_label :: Action -lex_quoted_label span buf _len _buf2 = do - s <- lex_string StringTypeSingle - (AI end bufEnd) <- getInput - let - token = ITlabelvarid (SourceText src) (mkFastString s) - src = lexemeToFastString (stepOn buf) (cur bufEnd - cur buf - 1) - start = psSpanStart span - - return $ L (mkPsSpan start end) token - - -lex_string :: LexStringType -> P String -lex_string strType = do - start <- getInput - (str, next) <- either fromStringLexError pure $ lexString strType alexGetChar' start - setInput next - pure str - - -lex_char_tok :: Action --- Here we are basically parsing character literals, such as 'x' or '\n' --- but we additionally spot 'x and ''T, returning ITsimpleQuote and --- ITtyQuote respectively, but WITHOUT CONSUMING the x or T part --- (the parser does that). --- So we have to do two characters of lookahead: when we see 'x we need to --- see if there's a trailing quote -lex_char_tok span buf _len _buf2 = do -- We've seen ' - i1 <- getInput -- Look ahead to first character - let loc = psSpanStart span - case alexGetChar' i1 of - Nothing -> lit_error i1 - - Just ('\'', i2@(AI end2 _)) -> do -- We've seen '' - setInput i2 - return (L (mkPsSpan loc end2) ITtyQuote) - - Just ('\\', i2@(AI end2 _)) -> do -- We've seen 'backslash - (lit_ch, i3) <- - either fromStringLexError pure $ - resolveEscapeCharacter alexGetChar' i2 - case alexGetChar' i3 of - Just ('\'', i4) -> do - setInput i4 - finish_char_tok buf loc lit_ch - Just (mc, _) | isSingleSmartQuote mc -> add_smart_quote_error mc end2 - _ -> lit_error i3 - - Just (c, i2@(AI end2 _)) - | not (isAnyChar c) -> lit_error i1 - | otherwise -> - - -- We've seen 'x, where x is a valid character - -- (i.e. not newline etc) but not a quote or backslash - case alexGetChar' i2 of -- Look ahead one more character - Just ('\'', i3) -> do -- We've seen 'x' - setInput i3 - finish_char_tok buf loc c - Just (c, _) | isSingleSmartQuote c -> add_smart_quote_error c end2 - _other -> do -- We've seen 'x not followed by quote - -- (including the possibility of EOF) - -- Just parse the quote only - let (AI end _) = i1 - return (L (mkPsSpan loc end) ITsimpleQuote) - --- We've already seen the closing quote --- Just need to check for trailing # -finish_char_tok :: StringBuffer -> PsLoc -> Char -> P (PsLocated Token) -finish_char_tok buf loc ch = do - i <- getInput - lex_magic_hash i >>= \case - Just i' -> do - setInput i' - -- Include the trailing # in SourceText - let (psSpan, src) = getStringLoc (buf, loc) i' - pure $ L psSpan (ITprimchar src ch) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, loc) i - pure $ L psSpan (ITchar src ch) +throwStringLexError :: AlexInput -> StringLexError -> P a +throwStringLexError i (StringLexError e pos) = setInput (advanceInputTo pos i) >> lexError e --- | Get the span and source text for a string from the given start to the given end. -getStringLoc :: (StringBuffer, PsLoc) -> AlexInput -> (PsSpan, SourceText) -getStringLoc (bufStart, locStart) (AI locEnd bufEnd) = (psSpan, SourceText src) +tok_quoted_label :: Action +tok_quoted_label span buf len _buf2 = do + s <- lex_chars ("#\"", "\"") span buf len + pure $ L span (ITlabelvarid src (mkFastString s)) where - psSpan = mkPsSpan locStart locEnd - src = lexemeToFastString bufStart (cur bufEnd - cur bufStart) - - --- Return Just if we found the magic hash, with the next input. -lex_magic_hash :: AlexInput -> P (Maybe AlexInput) -lex_magic_hash i = do - magicHash <- getBit MagicHashBit - if magicHash - then - case alexGetChar' i of - Just ('#', i') -> pure (Just i') - _other -> pure Nothing - else pure Nothing - -fromStringLexError :: StringLexError AlexInput -> P a -fromStringLexError = \case - UnexpectedEOF i squote -> checkSQuote squote >> throw i LexStringCharLitEOF - BadCharInitialLex i squote -> checkSQuote squote >> throw i LexStringCharLit - EscapeBadChar i -> throw i LexStringCharLit - EscapeUnexpectedEOF i -> throw i LexStringCharLitEOF - EscapeNumRangeError i -> throw i LexNumEscapeRange - EscapeSmartQuoteError c (AI loc _) -> add_smart_quote_error c loc + -- skip leading '#' + src = SourceText . mkFastString . drop 1 $ lexemeToString buf len + + +tok_char :: Action +tok_char span buf len _buf2 = do + c <- lex_chars ("'", "'") span buf (if endsInHash then len - 1 else len) >>= \case + [c] -> pure c + s -> panic $ "tok_char expected exactly one character, got: " ++ show s + pure . L span $ + if endsInHash + then ITprimchar src c + else ITchar src c where - throw i e = setInput i >> lexError e - checkSQuote = \case - NoSmartQuote -> pure () - SmartQuote c (AI loc _) -> add_nonfatal_smart_quote_error c loc - --- before calling lit_error, ensure that the current input is pointing to --- the position of the error in the buffer. This is so that we can report --- a correct location to the user, but also so we can detect UTF-8 decoding --- errors if they occur. -lit_error :: AlexInput -> P a -lit_error i = do setInput i; lexError LexStringCharLit + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + -- ----------------------------------------------------------------------------- -- QuasiQuote @@ -2389,32 +2400,28 @@ quasiquote_error start = do isSmartQuote :: AlexAccPred ExtsBitmap isSmartQuote _ _ _ (AI _ buf) = let c = prevChar buf ' ' in isSingleSmartQuote c || isDoubleSmartQuote c -smart_quote_error_message :: Char -> PsLoc -> MsgEnvelope PsMessage -smart_quote_error_message c loc = - let (correct_char, correct_char_name) = - if isSingleSmartQuote c then ('\'', "Single Quote") else ('"', "Quotation Mark") - err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ - PsErrUnicodeCharLooksLike c correct_char correct_char_name in - err - +throwSmartQuoteError :: Char -> PsLoc -> P a +throwSmartQuoteError c loc = addFatalError err + where + err = + mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ + PsErrUnicodeCharLooksLike c correct_char correct_char_name + (correct_char, correct_char_name) = + if isSingleSmartQuote c + then ('\'', "Single Quote") + else ('"', "Quotation Mark") + +-- | Throw a smart quote error, where the smart quote was the last character lexed smart_quote_error :: Action -smart_quote_error span buf _len _buf2 = do - let c = currentChar buf - addFatalError (smart_quote_error_message c (psSpanStart span)) - -add_smart_quote_error :: Char -> PsLoc -> P a -add_smart_quote_error c loc = addFatalError (smart_quote_error_message c loc) - -add_nonfatal_smart_quote_error :: Char -> PsLoc -> P () -add_nonfatal_smart_quote_error c loc = addError (smart_quote_error_message c loc) +smart_quote_error span _ _ buf2 = do + let c = prevChar buf2 (panic "smart_quote_error unexpectedly called on beginning of input") + throwSmartQuoteError c (psSpanStart span) -advance_to_smart_quote_character :: P () -advance_to_smart_quote_character = do - i <- getInput - case alexGetChar' i of - Just (c, _) | isDoubleSmartQuote c -> return () - Just (_, i2) -> do setInput i2; advance_to_smart_quote_character - Nothing -> return () -- should never get here +-- Note [Bare smart quote error] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- A smart quote inside of a string is allowed, but if a complete valid string +-- couldn't be lexed, we want to see if there's a smart quote that the user +-- thought ended the string, but in fact didn't. -- ----------------------------------------------------------------------------- -- Warnings @@ -2652,7 +2659,7 @@ getLastLocIncludingComments = P $ \s@(PState { prev_loc = prev_loc }) -> POk s p getLastLoc :: P PsSpan getLastLoc = P $ \s@(PState { last_loc = last_loc }) -> POk s last_loc -data AlexInput = AI !PsLoc !StringBuffer +data AlexInput = AI !PsLoc !StringBuffer deriving (Show) {- Note [Unicode in Alex] @@ -2763,6 +2770,19 @@ alexGetChar' (AI loc s) where (c,s') = nextChar s loc' = advancePsLoc loc c +-- | Advance the given input N bytes. +advanceInputBytes :: Int -> AlexInput -> AlexInput +advanceInputBytes n i0@(AI _ buf0) = advanceInputTo (cur buf0 + n) i0 + +-- | Advance the given input to the given position. +advanceInputTo :: Int -> AlexInput -> AlexInput +advanceInputTo pos = go + where + go i@(AI _ buf) + | cur buf >= pos = i + | Just (_, i') <- alexGetChar' i = go i' + | otherwise = i -- reached the end, just return the last input + getInput :: P AlexInput getInput = P $ \s at PState{ loc=l, buffer=b } -> POk s (AI l b) @@ -2770,9 +2790,10 @@ setInput :: AlexInput -> P () setInput (AI l b) = P $ \s -> POk s{ loc=l, buffer=b } () nextIsEOF :: P Bool -nextIsEOF = do - AI _ s <- getInput - return $ atEnd s +nextIsEOF = isEOF <$> getInput + +isEOF :: AlexInput -> Bool +isEOF (AI _ buf) = atEnd buf pushLexState :: Int -> P () pushLexState ls = P $ \s at PState{ lex_state=l } -> POk s{lex_state=ls:l} () @@ -3516,6 +3537,11 @@ topNoLayoutContainsCommas [] = False topNoLayoutContainsCommas (ALRLayout _ _ : ls) = topNoLayoutContainsCommas ls topNoLayoutContainsCommas (ALRNoLayout b _ : _) = b +-- If the generated alexScan/alexScanUser functions are called multiple times +-- in this file, alexScanUser gets broken out into a separate function and +-- increases memory usage. Make sure GHC inlines this function and optimizes it. +{-# INLINE alexScanUser #-} + lexToken :: P (PsLocated Token) lexToken = do inp@(AI loc1 buf) <- getInput ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -1,284 +1,199 @@ {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE PatternSynonyms #-} +{-# LANGUAGE TupleSections #-} +{-# LANGUAGE ViewPatterns #-} module GHC.Parser.String ( StringLexError (..), - ContainsSmartQuote (..), - LexStringType (..), lexString, + lexMultilineString, -- * Unicode smart quote helpers isDoubleSmartQuote, isSingleSmartQuote, - - -- * Other helpers - isAnyChar, - resolveEscapeCharacter, ) where -import GHC.Prelude +import GHC.Prelude hiding (getChar) import Control.Arrow ((>>>)) -import Control.Monad (guard, unless, when) -import Data.Char (chr, isPrint, ord) -import Data.List (unfoldr) +import Control.Monad (when) +import Data.Char (chr, ord) +import qualified Data.Foldable1 as Foldable1 +import qualified Data.List.NonEmpty as NonEmpty import Data.Maybe (listToMaybe, mapMaybe) +import GHC.Data.StringBuffer (StringBuffer) +import qualified GHC.Data.StringBuffer as StringBuffer import GHC.Parser.CharClass ( hexDigit, - is_any, is_decdigit, is_hexdigit, is_octdigit, is_space, octDecDigit, ) +import GHC.Parser.Errors.Types (LexErr (..)) import GHC.Utils.Panic (panic) -data LexStringType = StringTypeSingle | StringTypeMulti +type BufPos = Int +data StringLexError = StringLexError LexErr BufPos --- | State to accumulate while iterating through string literal. --- --- Fields are strict here to avoid space leak when iterating through large string literal --- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 -data LexStringState loc = LexStringState - { stringAcc :: !String - -- ^ The string seen so far, reversed - , multilineCommonWsPrefix :: !Int - -- ^ The common prefix for multiline strings. See Note [Multiline string literals] - , initialLoc :: !loc - -- ^ The location of the beginning of the string literal - } - --- | Get the character at the given location, with the location --- of the next character. Returns Nothing if at the end of the --- input. -type GetChar loc = loc -> Maybe (Char, loc) - -lexString :: LexStringType -> GetChar loc -> loc -> Either (StringLexError loc) (String, loc) -lexString strType getChar initialLoc = go initialState initialLoc +lexString :: Int -> StringBuffer -> Either StringLexError String +lexString = lexStringWith processChars processChars where - initialState = - LexStringState - { stringAcc = "" - , multilineCommonWsPrefix = - case strType of - StringTypeMulti -> maxBound - _ -> 0 - , initialLoc = initialLoc - } - - -- 's' is strict here to avoid space leak when iterating through large string literal - -- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 - go !s loc0 = - case getChar loc0 of - -- found closing delimiter - Just ('"', _) | Just loc1 <- checkDelimiter strType getChar loc0 -> do - let postprocess = - case strType of - StringTypeSingle -> id - StringTypeMulti -> postprocessMultiline (multilineCommonWsPrefix s) - Right (postprocess . reverse $ stringAcc s, loc1) - - -- found backslash - Just (c0@'\\', loc1) -> do - case getChar loc1 of - -- found '\&' character, which should be elided - Just ('&', loc2) -> go s loc2 - -- found start of a string gap - Just (c1, loc2) | is_space c1 -> collapseStringGap getChar s loc2 >>= go s - -- some other escape character - Just (c1, loc2) -> - case strType of - StringTypeSingle -> do - (c', loc') <- resolveEscapeCharacter getChar loc1 - go (addChar c' s) loc' - StringTypeMulti -> do - -- keep escape characters unresolved until after post-processing, - -- to distinguish between a user-newline and the user writing "\n". - -- but still process the characters here, to find any errors - _ <- resolveEscapeCharacter getChar loc1 - go (addChar c1 . addChar c0 $ s) loc2 - -- backslash at end of input - Nothing -> Left $ BadCharInitialLex loc1 (hasSQuote getChar s) - - -- found newline character in multiline string - Just (c0@'\n', loc1) | StringTypeMulti <- strType -> - uncurry go $ parseLeadingWS getChar (addChar c0 s) loc1 - - -- found some other character - Just (c0, loc1) | isAnyChar c0 -> go (addChar c0 s) loc1 - - -- found some unknown character - Just (_, _) -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - - -- reached EOF before finding end of string - Nothing -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) -{-# INLINE lexString #-} - -checkDelimiter :: LexStringType -> GetChar loc -> loc -> Maybe loc -checkDelimiter strType getChar loc0 = - case strType of - StringTypeSingle -> do - ('"', loc1) <- getChar loc0 - Just loc1 - StringTypeMulti -> do - ('"', loc1) <- getChar loc0 - ('"', loc2) <- getChar loc1 - ('"', loc3) <- getChar loc2 - Just loc3 -{-# INLINE checkDelimiter #-} - --- | A helper for adding the given character to the lexed string. -addChar :: Char -> LexStringState loc -> LexStringState loc -addChar c s = s{stringAcc = c : stringAcc s} -{-# INLINE addChar #-} - --- | Return whether the string we've parsed so far contains any smart quotes. -hasSQuote :: GetChar loc -> LexStringState loc -> ContainsSmartQuote loc -hasSQuote getChar s - | any isDoubleSmartQuote (stringAcc s) - , (c, loc) : _ <- filter (isDoubleSmartQuote . fst) allChars = - SmartQuote c loc - | otherwise = - NoSmartQuote + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps + >>> resolveEscapes + +-- ----------------------------------------------------------------------------- +-- Lexing interface + +{- +Note [Lexing strings] +~~~~~~~~~~~~~~~~~~~~~ + +After verifying if a string is lexically valid with Alex, we still need to do +some post processing of the string, namely: +1. Collapse string gaps +2. Resolve escape characters + +The problem: 'lexemeToString' is more performant than manually reading +characters from the StringBuffer. However, that completely erases the position +of each character, which we need in order to report the correct position for +error messages (e.g. when resolving escape characters). + +So what we'll do is do two passes. The first pass is optimistic; just convert +to a plain String and process it. If this results in an error, we do a second +pass, this time where each character is annotated with its position. Now, the +error has all the information it needs. + +Ideally, lexStringWith would take a single (forall c. HasChar c => ...) function, +but to help the specializer, we pass it in twice to concretize it for the two +types we actually use. +-} + +-- | See Note [Lexing strings] +lexStringWith :: + ([Char] -> Either (Char, LexErr) [Char]) + -> ([CharPos] -> Either (CharPos, LexErr) [CharPos]) + -> Int + -> StringBuffer + -> Either StringLexError String +lexStringWith processChars processCharsPos len buf = + case processChars $ bufferChars buf len of + Right s -> Right s + Left _ -> + case processCharsPos $ bufferLocatedChars buf len of + Right _ -> panic "expected lex error on second pass" + Left ((_, pos), e) -> Left $ StringLexError e pos + +class HasChar c where + getChar :: c -> Char + setChar :: Char -> c -> c + +instance HasChar Char where + getChar = id + setChar = const + +instance HasChar (Char, x) where + getChar = fst + setChar c (_, x) = (c, x) + +pattern Char :: HasChar c => Char -> c +pattern Char c <- (getChar -> c) +{-# COMPLETE Char #-} + +bufferChars :: StringBuffer -> Int -> [Char] +bufferChars = StringBuffer.lexemeToString + +type CharPos = (Char, BufPos) + +bufferLocatedChars :: StringBuffer -> Int -> [CharPos] +bufferLocatedChars initialBuf len = go initialBuf where - allChars = unfoldr getCharWithLoc (initialLoc s) - getCharWithLoc loc = - case getChar loc of - Just (c, loc') -> Just ((c, loc), loc') - Nothing -> Nothing -{-# INLINE hasSQuote #-} - --- | After parsing a backslash and a space character, consume the rest of --- the string gap and return the next location. -collapseStringGap :: GetChar loc -> LexStringState loc -> loc -> Either (StringLexError loc) loc -collapseStringGap getChar s = go + go buf + | atEnd buf = [] + | otherwise = + let (c, buf') = StringBuffer.nextChar buf + in (c, StringBuffer.cur buf) : go buf' + + atEnd buf = StringBuffer.byteDiff initialBuf buf >= len + +-- ----------------------------------------------------------------------------- +-- Lexing phases + +collapseGaps :: HasChar c => [c] -> [c] +collapseGaps = go where - go loc0 = - case getChar loc0 of - Just ('\\', loc1) -> pure loc1 - Just (c0, loc1) | is_space c0 -> go loc1 - Just _ -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - Nothing -> Left $ UnexpectedEOF loc0 (hasSQuote getChar s) -{-# INLINE collapseStringGap #-} + go = \case + c1@(Char '\\') : c2@(Char c) : cs + | is_space c -> go $ dropGap cs + | otherwise -> c1 : c2 : go cs + c : cs -> c : go cs + [] -> [] --- | See Note [Multiline string literals] -parseLeadingWS :: GetChar loc -> LexStringState loc -> loc -> (LexStringState loc, loc) -parseLeadingWS getChar = go 0 + dropGap = \case + Char '\\' : cs -> cs + _ : cs -> dropGap cs + [] -> panic "gap unexpectedly ended" + +resolveEscapes :: HasChar c => [c] -> Either (c, LexErr) [c] +resolveEscapes = go dlistEmpty where - go !col s loc = - case getChar loc of - Just (c@' ', loc') -> go (col + 1) (addChar c s) loc' - -- expand tabs - Just ('\t', loc') -> - let fill = 8 - (col `mod` 8) - s' = applyN fill (addChar ' ') s - in go (col + fill) s' loc' - -- if we see a newline or string delimiter, then this line only contained whitespace, so - -- don't include it in the common whitespace prefix - Just ('\n', _) -> (s, loc) - Just ('"', _) | Just _ <- checkDelimiter StringTypeMulti getChar loc -> (s, loc) - -- found some other character, so we're done parsing leading whitespace - _ -> - let s' = s{multilineCommonWsPrefix = min col (multilineCommonWsPrefix s)} - in (s', loc) - - applyN :: Int -> (a -> a) -> a -> a - applyN n f x0 = iterate f x0 !! n -{-# INLINE parseLeadingWS #-} - -data StringLexError loc - = UnexpectedEOF !loc !(ContainsSmartQuote loc) - -- ^ Unexpectedly hit EOF when lexing string - | BadCharInitialLex !loc !(ContainsSmartQuote loc) - -- ^ Found invalid character when initially lexing string - | EscapeBadChar !loc - -- ^ Found invalid character when parsing an escaped character - | EscapeUnexpectedEOF !loc - -- ^ Unexpectedly hit EOF when parsing an escaped character - | EscapeNumRangeError !loc - -- ^ Escaped number exceeds range - | EscapeSmartQuoteError !Char !loc - -- ^ Found escaped smart unicode chars as `\’` or `\”` - deriving (Show) - --- | When initially lexing the string, we want to track if we've --- seen a smart quote, to show a helpful "you might be accidentally --- using a smart quote" error. -data ContainsSmartQuote loc - = NoSmartQuote - | SmartQuote !Char !loc - deriving (Show) + go !acc = \case + [] -> pure $ dlistToList acc + Char '\\' : Char '&' : cs -> go acc cs + backslash@(Char '\\') : cs -> + case resolveEscapeChar cs of + Right (esc, cs') -> go (acc `dlistSnoc` setChar esc backslash) cs' + Left (c, e) -> Left (c, e) + c : cs -> go (acc `dlistSnoc` c) cs -- ----------------------------------------------------------------------------- -- Escape characters --- | After finding a backslash, parse the rest of the escape character, starting --- at the given location. -resolveEscapeCharacter :: GetChar loc -> loc -> Either (StringLexError loc) (Char, loc) -resolveEscapeCharacter getChar loc0 = do - (c0, loc1) <- expectChar loc0 - case c0 of - 'a' -> pure ('\a', loc1) - 'b' -> pure ('\b', loc1) - 'f' -> pure ('\f', loc1) - 'n' -> pure ('\n', loc1) - 'r' -> pure ('\r', loc1) - 't' -> pure ('\t', loc1) - 'v' -> pure ('\v', loc1) - '\\' -> pure ('\\', loc1) - '"' -> pure ('\"', loc1) - '\'' -> pure ('\'', loc1) - -- escape codes - 'x' -> expectNum is_hexdigit 16 hexDigit loc1 - 'o' -> expectNum is_octdigit 8 octDecDigit loc1 - _ | is_decdigit c0 -> expectNum is_decdigit 10 octDecDigit loc0 - -- control characters (e.g. '\^M') - '^' -> do - (c1, loc2) <- expectChar loc1 - unless ('@' <= c1 && c1 <= '_') $ Left $ EscapeBadChar loc1 - pure (chr $ ord c1 - ord '@', loc2) - -- long form escapes (e.g. '\NUL') - _ | Just (c1, loc2) <- parseLongEscape getChar c0 loc1 -> pure (c1, loc2) - -- check unicode smart quotes (#21843) - _ | isDoubleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - _ | isSingleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - -- unknown escape - _ -> Left $ EscapeBadChar loc0 +-- | Resolve a escape character, after having just lexed a backslash. +-- Assumes escape character is valid. +resolveEscapeChar :: HasChar c => [c] -> Either (c, LexErr) (Char, [c]) +resolveEscapeChar = \case + Char 'a' : cs -> pure ('\a', cs) + Char 'b' : cs -> pure ('\b', cs) + Char 'f' : cs -> pure ('\f', cs) + Char 'n' : cs -> pure ('\n', cs) + Char 'r' : cs -> pure ('\r', cs) + Char 't' : cs -> pure ('\t', cs) + Char 'v' : cs -> pure ('\v', cs) + Char '\\' : cs -> pure ('\\', cs) + Char '"' : cs -> pure ('\"', cs) + Char '\'' : cs -> pure ('\'', cs) + -- escape codes + Char 'x' : cs -> parseNum is_hexdigit 16 hexDigit cs + Char 'o' : cs -> parseNum is_octdigit 8 octDecDigit cs + cs@(Char c : _) | is_decdigit c -> parseNum is_decdigit 10 octDecDigit cs + -- control characters (e.g. '\^M') + Char '^' : Char c : cs -> pure (chr $ ord c - ord '@', cs) + -- long form escapes (e.g. '\NUL') + cs | Just (esc, cs') <- parseLongEscape cs -> pure (esc, cs') + -- shouldn't happen + Char c : _ -> panic $ "found unexpected escape character: " ++ show c + [] -> panic "escape character unexpectedly ended" where - expectChar loc = - case getChar loc of - Just x -> pure x - Nothing -> Left $ EscapeUnexpectedEOF loc - - expectNum isDigit base toDigit loc1 = do - (c1, loc2) <- expectChar loc1 - unless (isDigit c1) $ Left $ EscapeBadChar loc1 - let parseNum x loc = - case getChar loc of - Just (c, loc') | isDigit c -> do - let x' = x * base + toDigit c - when (x' > 0x10ffff) $ Left $ EscapeNumRangeError loc - parseNum x' loc' - _ -> - pure (chr x, loc) - parseNum (toDigit c1) loc2 -{-# INLINE resolveEscapeCharacter #-} - -parseLongEscape :: GetChar loc -> Char -> loc -> Maybe (Char, loc) -parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCodes + parseNum isDigit base toDigit = + let go x = \case + ch@(Char c) : cs | isDigit c -> do + let x' = x * base + toDigit c + when (x' > 0x10ffff) $ Left (ch, LexNumEscapeRange) + go x' cs + cs -> pure (chr x, cs) + in go 0 + +parseLongEscape :: HasChar c => [c] -> Maybe (Char, [c]) +parseLongEscape cs = listToMaybe (mapMaybe tryParse longEscapeCodes) where - tryParse (prefix, c) = do - p0 : p <- pure prefix - guard (p0 == c0) -- see if the first character matches - loc <- parsePrefix loc1 p -- see if the rest of the prefix matches - pure (c, loc) - - parsePrefix loc = \case - [] -> pure loc - p : ps -> do - (c, loc') <- getChar loc - guard (p == c) - parsePrefix loc' ps + tryParse (code, esc) = + case splitAt (length code) cs of + (pre, cs') | map getChar pre == code -> Just (esc, cs') + _ -> Nothing longEscapeCodes = [ ("NUL", '\NUL') @@ -289,14 +204,14 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("ENQ", '\ENQ') , ("ACK", '\ACK') , ("BEL", '\BEL') - , ("BS", '\BS') - , ("HT", '\HT') - , ("LF", '\LF') - , ("VT", '\VT') - , ("FF", '\FF') - , ("CR", '\CR') - , ("SO", '\SO') - , ("SI", '\SI') + , ("BS" , '\BS' ) + , ("HT" , '\HT' ) + , ("LF" , '\LF' ) + , ("VT" , '\VT' ) + , ("FF" , '\FF' ) + , ("CR" , '\CR' ) + , ("SO" , '\SO' ) + , ("SI" , '\SI' ) , ("DLE", '\DLE') , ("DC1", '\DC1') , ("DC2", '\DC2') @@ -306,17 +221,16 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("SYN", '\SYN') , ("ETB", '\ETB') , ("CAN", '\CAN') - , ("EM", '\EM') + , ("EM" , '\EM' ) , ("SUB", '\SUB') , ("ESC", '\ESC') - , ("FS", '\FS') - , ("GS", '\GS') - , ("RS", '\RS') - , ("US", '\US') - , ("SP", '\SP') + , ("FS" , '\FS' ) + , ("GS" , '\GS' ) + , ("RS" , '\RS' ) + , ("US" , '\US' ) + , ("SP" , '\SP' ) , ("DEL", '\DEL') ] -{-# INLINE parseLongEscape #-} -- ----------------------------------------------------------------------------- -- Unicode Smart Quote detection (#21843) @@ -333,6 +247,98 @@ isSingleSmartQuote = \case '’' -> True _ -> False +-- ----------------------------------------------------------------------------- +-- Multiline strings + +-- | See Note [Multiline string literals] +-- +-- Assumes string is lexically valid. Skips the steps about splitting +-- and rejoining lines, and instead manually find newline characters, +-- for performance. +lexMultilineString :: Int -> StringBuffer -> Either StringLexError String +lexMultilineString = lexStringWith processChars processChars + where + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps -- Step 1 + >>> expandLeadingTabs -- Step 3 + >>> rmCommonWhitespacePrefix -- Step 4 + >>> collapseOnlyWsLines -- Step 5 + >>> rmFirstNewline -- Step 7a + >>> rmLastNewline -- Step 7b + >>> resolveEscapes -- Step 8 + + -- expands all tabs, since the lexer will verify that tabs can only appear + -- as leading indentation + expandLeadingTabs :: HasChar c => [c] -> [c] + expandLeadingTabs = + let go !col = \case + c@(Char '\t') : cs -> + let fill = 8 - (col `mod` 8) + in replicate fill (setChar ' ' c) ++ go (col + fill) cs + c : cs -> c : go (if getChar c == '\n' then 0 else col + 1) cs + [] -> [] + in go 0 + + rmCommonWhitespacePrefix :: HasChar c => [c] -> [c] + rmCommonWhitespacePrefix cs0 = + let commonWSPrefix = getCommonWsPrefix (map getChar cs0) + go = \case + c@(Char '\n') : cs -> c : go (dropLine commonWSPrefix cs) + c : cs -> c : go cs + [] -> [] + -- drop x characters from the string, or up to a newline, whichever + -- comes first + dropLine !x = \case + cs | x <= 0 -> cs + cs@(Char '\n' : _) -> cs + _ : cs -> dropLine (x - 1) cs + [] -> [] + in go cs0 + + collapseOnlyWsLines :: HasChar c => [c] -> [c] + collapseOnlyWsLines = + let go = \case + c@(Char '\n') : cs | Just cs' <- checkAllWs cs -> c : go cs' + c : cs -> c : go cs + [] -> [] + checkAllWs = \case + -- got all the way to a newline or the end of the string, return + cs@(Char '\n' : _) -> Just cs + cs@[] -> Just cs + -- found whitespace, continue + Char c : cs | is_space c -> checkAllWs cs + -- anything else, stop + _ -> Nothing + in go + + rmFirstNewline :: HasChar c => [c] -> [c] + rmFirstNewline = \case + Char '\n' : cs -> cs + cs -> cs + + rmLastNewline :: HasChar c => [c] -> [c] + rmLastNewline = + let go = \case + [] -> [] + [Char '\n'] -> [] + c : cs -> c : go cs + in go + +-- | See step 4 in Note [Multiline string literals] +-- +-- Assumes tabs have already been expanded. +getCommonWsPrefix :: String -> Int +getCommonWsPrefix s = + case NonEmpty.nonEmpty includedLines of + Nothing -> 0 + Just ls -> Foldable1.minimum $ NonEmpty.map (length . takeWhile is_space) ls + where + includedLines = + filter (not . all is_space) -- ignore whitespace-only lines + . drop 1 -- ignore first line in calculation + $ lines s + {- Note [Multiline string literals] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -351,23 +357,13 @@ The canonical steps for post processing a multiline string are: 2. Split the string by newlines 3. Convert leading tabs into spaces * In each line, any tabs preceding non-whitespace characters are replaced with spaces up to the next tab stop -4. Remove common whitespace prefix in every line (see below) +4. Remove common whitespace prefix in every line except the first (see below) 5. If a line contains only whitespace, remove all of the whitespace 6. Join the string back with `\n` delimiters -7. If the first character of the string is a newline, remove it +7a. If the first character of the string is a newline, remove it +7b. If the last character of the string is a newline, remove it 8. Interpret escaped characters -However, for performance reasons, we do as much of this in one pass as possible: -1. As we lex the string, do the following steps as they appear: - a. Collapse string gaps - b. Keep track of the common whitespace prefix so far - c. Validate escaped characters -2. At the very end, post process the lexed string: - a. Remove the common whitespace prefix from every line - b. Remove all whitespace from all-whitespace lines - c. Remove initial newline character - d. Resolve escaped characters - The common whitespace prefix can be informally defined as "The longest prefix of whitespace shared by all lines in the string, excluding the first line and any whitespace-only lines". @@ -382,75 +378,16 @@ It's more precisely defined with the following algorithm: 3. Calculate the longest prefix of whitespace shared by all lines in the remaining list -} --- | See Note [Multiline string literals] -postprocessMultiline :: Int -> String -> String -postprocessMultiline commonWSPrefix = - rmCommonWhitespacePrefix - >>> collapseOnlyWsLines - >>> rmFirstNewline - >>> rmLastNewline - >>> resolveEscapeChars - where - rmCommonWhitespacePrefix = - let go = \case - '\n' : s -> '\n' : go (dropLine commonWSPrefix s) - c : s -> c : go s - [] -> [] - -- drop x characters from the string, or up to a newline, whichever - -- comes first - dropLine !x = \case - s | x <= 0 -> s - s@('\n' : _) -> s - _ : s -> dropLine (x - 1) s - [] -> [] - in go - - collapseOnlyWsLines = - let go = \case - '\n' : s | Just s' <- checkAllWs s -> '\n' : go s' - c : s -> c : go s - [] -> [] - checkAllWs = \case - -- got all the way to a newline or the end of the string, return - s@('\n' : _) -> Just s - s@[] -> Just s - -- found whitespace, continue - c : s | is_space c -> checkAllWs s - -- anything else, stop - _ -> Nothing - in go - - rmFirstNewline = \case - '\n' : s -> s - s -> s +-- ----------------------------------------------------------------------------- +-- DList - rmLastNewline = - let go = \case - [] -> [] - ['\n'] -> [] - c : cs -> c : go cs - in go +newtype DList a = DList ([a] -> [a]) - -- resolve escape characters, deferred from lexString. guaranteed - -- to not throw any errors, since we already checked them in lexString - resolveEscapeChars = \case - [] -> [] - '\\' : s -> - -- concretizing 'loc' to String: - -- resolveEscapeCharacter :: (String -> Maybe (Char, String)) -> String -> Either _ (Char, String) - case resolveEscapeCharacter uncons s of - Left e -> panic $ "resolving escape characters in multiline string unexpectedly found errors: " ++ show e - Right (c, s') -> c : resolveEscapeChars s' - c : s -> c : resolveEscapeChars s - - uncons = \case - c : cs -> Just (c, cs) - [] -> Nothing +dlistEmpty :: DList a +dlistEmpty = DList id --- ----------------------------------------------------------------------------- --- Helpers +dlistToList :: DList a -> [a] +dlistToList (DList f) = f [] -isAnyChar :: Char -> Bool -isAnyChar c - | c > '\x7f' = isPrint c - | otherwise = is_any c +dlistSnoc :: DList a -> a -> DList a +dlistSnoc (DList f) x = DList (f . (x :)) ===================================== testsuite/driver/testlib.py ===================================== @@ -2760,7 +2760,7 @@ def normalise_errmsg(s: str) -> str: # normalise slashes to minimise Windows/Unix filename differences, # but don't normalize backslashes in chars - s = re.sub(r"(?!')\\", '/', s) + s = re.sub(r"(?:9:1: error: [GHC-58481] parse error on input ‘+’ -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' + +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/ghci/scripts/ghci022.stderr ===================================== @@ -1,3 +1,2 @@ +ghci022.hs:2:5: error: [GHC-21231] lexical error at character '\n' -ghci022.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr ===================================== @@ -1,3 +1,3 @@ - MultilineStringsInnerTab.hs:8:5: error: [GHC-21231] - lexical error in string/character literal at character '\t' + lexical error at character '\t' + ===================================== testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr ===================================== @@ -1,6 +1,2 @@ - MultilineStringsSmartQuotes.hs:9:3: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -MultilineStringsSmartQuotes.hs:10:1: [GHC-21231] - lexical error in string/character literal at end of input ===================================== testsuite/tests/parser/should_fail/T21843c.stderr ===================================== @@ -1,6 +1,2 @@ - -T21843c.hs:3:19: [GHC-31623] +T21843c.hs:3:13: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -T21843c.hs:3:20: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/T21843e.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843e.hs:3:15: [GHC-31623] +T21843e.hs:3:13: error: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not + ===================================== testsuite/tests/parser/should_fail/T21843f.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843f.hs:3:13: [GHC-31623] +T21843f.hs:3:11: error: [GHC-31623] Unicode character '‘' ('\8216') looks like ''' (Single Quote), but it is not + ===================================== testsuite/tests/parser/should_fail/T3751.stderr ===================================== @@ -1,3 +1,2 @@ +T3751.hs:3:5: error: [GHC-21231] lexical error at character '\167' -T3751.hs:3:7: error: [GHC-21231] - lexical error in string/character literal at character '\167' ===================================== testsuite/tests/parser/should_fail/T5425.stderr ===================================== @@ -1,3 +1,2 @@ +T5425.hs:4:1: error: [GHC-21231] lexical error at character '\955' -T5425.hs:5:2: error: [GHC-21231] - lexical error in string/character literal at character '\955' ===================================== testsuite/tests/parser/should_fail/readFail002.stderr ===================================== @@ -1,3 +1,3 @@ +readFail002.hs:5:5: error: [GHC-21231] + lexical error at character '\n' -readFail002.hs:5:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/readFail004.stderr ===================================== @@ -1,3 +1,3 @@ +readFail004.hs:17:16: error: [GHC-21231] + lexical error at character '.' -readFail004.hs:19:1: error: [GHC-21231] - lexical error in string/character literal at character '.' ===================================== testsuite/tests/parser/should_fail/readFail005.stderr ===================================== @@ -1,3 +1,2 @@ - -readFail005.hs:4:7: error: [GHC-21231] - lexical error in string/character literal at character '&' +readFail005.hs:4:5: [GHC-21231] + lexical error at character '&' ===================================== testsuite/tests/parser/should_fail/readFail033.stderr ===================================== @@ -1,3 +1,3 @@ +readFail033.hs:2:5: error: [GHC-21231] + lexical error at character '\t' -readFail033.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\t' ===================================== testsuite/tests/parser/unicode/utf8_010.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_010.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_011.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_011.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_020.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_020.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_021.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_021.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_022.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_022.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs ===================================== @@ -320,7 +320,7 @@ classify tok = ITlabelvarid{} -> TkUnknown ITchar{} -> TkChar ITstring{} -> TkString - ITmultilinestring{} -> TkString + ITstringMulti{} -> TkString ITinteger{} -> TkNumber ITrational{} -> TkNumber ITprimchar{} -> TkChar View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/05589606db8c91f10bb96b69f3cb458c8d558f8b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/05589606db8c91f10bb96b69f3cb458c8d558f8b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 05:41:32 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 20 Sep 2024 01:41:32 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] ghc-toolchain: Set -fuse-ld even for ld.bfd Message-ID: <66ed0b0c51553_32250c1b551c1037ed@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 88bd9a1f by Sven Tennie at 2024-09-20T07:39:16+02:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - 1 changed file: - utils/ghc-toolchain/src/GHC/Toolchain/Tools/Link.hs Changes: ===================================== utils/ghc-toolchain/src/GHC/Toolchain/Tools/Link.hs ===================================== @@ -91,7 +91,7 @@ findLinkFlags enableOverride cc ccLink -- executable exists before trying cc. do _ <- findProgram (linker ++ " linker") emptyProgOpt ["ld."++linker] prog <$ checkLinkWorks cc prog - | linker <- ["lld", "gold"] + | linker <- ["lld", "gold", "bfd"] , let prog = over _prgFlags (++["-fuse-ld="++linker]) ccLink ] <|> (ccLink <$ checkLinkWorks cc ccLink) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/88bd9a1f06b6fe3efb1a0a1888e8e8ca82782388 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/88bd9a1f06b6fe3efb1a0a1888e8e8ca82782388 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 09:45:17 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 20 Sep 2024 05:45:17 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Cross-tests for RISCV64 Message-ID: <66ed442d53415_224d8113f68c942f6@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: e51d5654 by Sven Tennie at 2024-09-20T11:44:53+02:00 Cross-tests for RISCV64 - - - - - 1 changed file: - .gitlab/ci.sh Changes: ===================================== .gitlab/ci.sh ===================================== @@ -614,6 +614,13 @@ function test_hadrian() { "runtest.opts+=${RUNTEST_ARGS:-}" \ "runtest.opts+=--unexpected-output-dir=$TOP/unexpected-test-output" \ || fail "hadrian main testsuite targetting $CROSS_TARGET" + elif [[ -n "${CROSS_TARGET:-}" ]] && [[ "${CROSS_TARGET:-}" == *"riscv64"* ]]; then + run_hadrian \ + test \ + --summary-junit=./junit.xml \ + "runtest.opts+=${RUNTEST_ARGS:-}" \ + "runtest.opts+=--unexpected-output-dir=$TOP/unexpected-test-output" \ + || fail "hadrian main testsuite targetting $CROSS_TARGET" elif [ -n "${CROSS_TARGET:-}" ]; then local instdir="$TOP/_build/install" local test_compiler="$instdir/bin/${cross_prefix}ghc$exe" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e51d565429f0522bc9991e677d101e1cf9396d5c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e51d565429f0522bc9991e677d101e1cf9396d5c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 10:27:38 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Fri, 20 Sep 2024 06:27:38 -0400 Subject: [Git][ghc/ghc][wip/T25251] 7 commits: RTS: expose closure_sizeW_ (#25252) Message-ID: <66ed4e1a2b339_224d81401d20969fe@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25251 at Glasgow Haskell Compiler / GHC Commits: d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - b29e34b7 by Simon Peyton Jones at 2024-09-20T11:22:46+01:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 30 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Utils/Unify.hs - compiler/GHC/Tc/Zonk/Type.hs - compiler/GHC/Types/Var.hs - ghc/GHCi/UI/Info.hs - m4/fptools_set_haskell_platform_vars.m4 - rts/RtsSymbols.c - − testsuite/tests/default/DefaultImport04.stdout - testsuite/tests/default/DefaultImport04.hs → testsuite/tests/default/DefaultImportFail07.hs - + testsuite/tests/default/DefaultImportFail07.stderr - testsuite/tests/default/all.T - + testsuite/tests/ffi/should_compile/T25255.hs - testsuite/tests/ffi/should_compile/all.T - testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr - testsuite/tests/plugins/simple-plugin/Simple/RemovePlugin.hs - + testsuite/tests/pmcheck/should_compile/T25257.hs - testsuite/tests/pmcheck/should_compile/all.T The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/577f2c3ccfe65ecc18020330460bd072f13160d6...b29e34b7755c5c10b47b1d7915708cce0cb95c8c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/577f2c3ccfe65ecc18020330460bd072f13160d6...b29e34b7755c5c10b47b1d7915708cce0cb95c8c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 10:34:05 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Fri, 20 Sep 2024 06:34:05 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] Wrangled until everything compiled. Message-ID: <66ed4f9db3985_224d81461360103352@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 45ea0e59 by Hassan Al-Awwadi at 2024-09-20T12:33:28+02:00 Wrangled until everything compiled. There are two issues: - FieldOcc used to only have one constructor and always be unambigious, this is no longer the case. Calls to foLabel are thus partial now. Don't know how much we care about this, since the partial calls are mostly inside functions that used to operate on the operate on the unambigious FieldOcc - Lots of functions that take in a FieldOcc, or a HsExpr (with the HsRecSel constructor) now have a new case. It was not always clear to me what the correct implementation was for these. I have filled them in as far as I could and left one undefined... - - - - - 19 changed files: - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Bind.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Gen/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/Tc/TyCl/Utils.hs - utils/haddock/haddock-api/src/Haddock/Backends/Hoogle.hs - utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs - utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs - utils/haddock/haddock-api/src/Haddock/Convert.hs - utils/haddock/haddock-api/src/Haddock/GhcUtils.hs - utils/haddock/haddock-api/src/Haddock/Interface/Create.hs - utils/haddock/haddock-api/src/Haddock/Interface/Rename.hs - utils/haddock/haddock-api/src/Haddock/Types.hs Changes: ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -104,7 +104,7 @@ hsExprType :: HsExpr GhcTc -> Type hsExprType (HsVar _ (L _ id)) = idType id hsExprType (HsUnboundVar (HER _ ty _) _) = ty hsExprType (HsRecSel _ (FieldOcc _ id)) = idType (unLoc id) -hsExprType (HsOverLabel v _ _) = dataConCantHappen v +hsExprType (HsOverLabel v _) = dataConCantHappen v hsExprType (HsIPVar v _) = dataConCantHappen v hsExprType (HsOverLit _ lit) = overLitType lit hsExprType (HsLit _ lit) = hsLitType lit ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -273,7 +273,7 @@ application. For example: See the `HsApp` case for where it is filtered out -} -dsExpr (HsRecSel _ (FieldOcc id _)) +dsExpr (HsRecSel _ (FieldOcc _ (L _ id))) = do { let name = getName id RecSelId {sel_cons = (_, cons_wo_field)} = idDetails id ===================================== compiler/GHC/HsToCore/Quote.hs ===================================== @@ -284,7 +284,7 @@ repTopDs group@(HsGroup { hs_valds = valds , hs_docs = docs }) = do { let { bndrs = hsScopedTvBinders valds ++ hsGroupBinders group - ++ map foExt (hsPatSynSelectors valds) + ++ map (unLoc . foLabel) (hsPatSynSelectors valds) ; instds = tyclds >>= group_instds } ; ss <- mkGenSyms bndrs ; @@ -1538,7 +1538,8 @@ repE (HsVar _ (L _ x)) = repE (HsIPVar _ n) = rep_implicit_param_name n >>= repImplicitParamVar repE (HsOverLabel _ s) = repOverLabel s -repE (HsRecSel _ (FieldOcc x _)) = repE (HsVar noExtField (noLocA x)) +repE (HsRecSel _ (FieldOcc _ (L _ x))) = repE (HsVar noExtField (noLocA x)) +repE r@(HsRecSel _ (XFieldOcc _)) = notHandled (ThAmbiguousRecordSelectors r) -- Remember, we're desugaring renamer output here, so -- HsOverlit can definitely occur @@ -1816,10 +1817,11 @@ repUpdFields = repListM fieldExpTyConName rep_fld where rep_fld :: LHsRecUpdField GhcRn GhcRn -> MetaM (Core (M TH.FieldExp)) rep_fld (L l fld) = case unLoc (hfbLHS fld) of - Unambiguous sel_name _ -> do { fn <- lookupLOcc (L l sel_name) - ; e <- repLE (hfbRHS fld) - ; repFieldExp fn e } - Ambiguous{} -> notHandled (ThAmbiguousRecordUpdates fld) + FieldOcc _ (L _ sel_name) -> do { fn <- lookupLOcc (L l sel_name) + ; e <- repLE (hfbRHS fld) + ; repFieldExp fn e + } + (XFieldOcc _) -> notHandled (ThAmbiguousRecordUpdates fld) @@ -2022,7 +2024,7 @@ rep_bind (L loc (PatSynBind _ (PSB { psb_id = syn mkGenArgSyms (InfixCon arg1 arg2) = mkGenSyms [unLoc arg1, unLoc arg2] mkGenArgSyms (RecCon fields) = do { let pats = map (unLoc . recordPatSynPatVar) fields - sels = map (foExt . recordPatSynField) fields + sels = map (unLoc . foLabel . recordPatSynField) fields ; ss <- mkGenSyms sels ; return $ replaceNames (zip sels pats) ss } @@ -2054,7 +2056,7 @@ repPatSynArgs (InfixCon arg1 arg2) ; arg2' <- lookupLOcc arg2 ; repInfixPatSynArgs arg1' arg2' } repPatSynArgs (RecCon fields) - = do { sels' <- repList nameTyConName (lookupOcc . foExt) sels + = do { sels' <- repList nameTyConName (lookupOcc . unLoc . foLabel) sels ; repRecordPatSynArgs sels' } where sels = map recordPatSynField fields @@ -2877,7 +2879,7 @@ repRecConArgs ips = do rep_ip (L _ ip) = mapM (rep_one_ip (cd_fld_type ip)) (cd_fld_names ip) rep_one_ip :: LBangType GhcRn -> LFieldOcc GhcRn -> MetaM (Core (M TH.VarBangType)) - rep_one_ip t n = do { MkC v <- lookupOcc (foExt $ unLoc n) + rep_one_ip t n = do { MkC v <- lookupOcc (unLoc . foLabel $ unLoc n) ; MkC ty <- repBangTy t ; rep2 varBangTypeName [v,ty] } ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1483,8 +1483,7 @@ instance HiePass p => ToHie (RFContext (LocatedA (FieldOcc (GhcPass p)))) where case f of FieldOcc _ fld -> [toHie $ C (RecField c rhs) (L (locA nspan) $ unLoc fld)] - XFieldOcc (Ambiguous (L nspan fld)) - -> [ toHie $ C (RecField c rhs) (L (locA nspan) fld) ] + XFieldOcc (Ambiguous _) -> [] HieTc -> case f of FieldOcc _ fld -> @@ -2075,9 +2074,8 @@ instance ToHie (LocatedA (TyFamInstDecl GhcRn)) where toHie (L sp (TyFamInstDecl _ d)) = toHie $ TS (ResolvedScopes [mkScope sp]) d instance HiePass p => ToHie (Context (FieldOcc (GhcPass p))) where - toHie (C c (FieldOcc n (L l _))) = case hiePass @p of - HieTc -> toHie (C c (L l n)) - HieRn -> toHie (C c (L l n)) + toHie (C c (FieldOcc _ l)) = toHie (C c l) + toHie (C _ (XFieldOcc _)) = concatM [] instance HiePass p => ToHie (PatSynFieldContext (RecordPatSynField (GhcPass p))) where toHie (PSC sp (RecordPatSynField a b)) = concatM $ ===================================== compiler/GHC/Rename/Bind.hs ===================================== @@ -824,7 +824,9 @@ rnPatSynBind sig_fn bind@(PSB { psb_id = L l name , psb_ext = fvs' } selector_names = case details' of RecCon names -> - map (foExt . recordPatSynField) names + -- I don't actually know if its fine or not + -- that foLabel is partial. + map (unLoc . foLabel . recordPatSynField) names _ -> [] ; fvs' `seq` -- See Note [Free-variable space leak] ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -339,7 +339,7 @@ rnExpr (HsVar _ (L l v)) ; this_mod <- getModule ; when (nameIsLocalOrFrom this_mod sel_name) $ checkThLocalName sel_name - ; return (HsRecSel noExtField (FieldOcc sel_name (L l v) ), unitFV sel_name) + ; return (HsRecSel noExtField (FieldOcc v (L l sel_name)), unitFV sel_name) } | nm == nilDataConName -- Treat [] as an ExplicitList, so that @@ -417,7 +417,7 @@ rnExpr (OpApp _ e1 op e2) -- should prevent bad things happening. ; fixity <- case op' of L _ (HsVar _ (L _ n)) -> lookupFixityRn n - L _ (HsRecSel _ f) -> lookupFieldFixityRn f + L _ (HsRecSel _ f) -> fromJust <$> lookupFieldFixityRn f _ -> return (Fixity minPrecedence InfixL) -- c.f. lookupFixity for unbound ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -10,6 +10,8 @@ {-# LANGUAGE MultiWayIf #-} {-# LANGUAGE DerivingStrategies #-} {-# LANGUAGE GeneralizedNewtypeDeriving #-} +{-# OPTIONS_GHC -Wno-unrecognised-pragmas #-} +{-# HLINT ignore "Use camelCase" #-} {- (c) The GRASP/AQUA Project, Glasgow University, 1992-1998 @@ -257,6 +259,42 @@ newPatName (LetMk is_top fix_env) rdr_name NotTopLevel -> newLocalBndrRn rdr_name TopLevel -> newTopSrcBinder rdr_name ; bindLocalNames [name] $ + -- Do *not* use bindLocalNameFV here; + -- Do *not* use bindLocalNameFV here; + -- Do *not* use bindLocalNameFV here; + -- Do *not* use bindLocalNameFV here; + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- For the TopLevel case + -- For the TopLevel case + -- For the TopLevel case + -- For the TopLevel case + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + + -- Do *not* use bindLocalNameFV here; + + -- Do *not* use bindLocalNameFV here; + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- For the TopLevel case + -- For the TopLevel case + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External nam + + -- Do *not* use bindLocalNameFV here; + -- Do *not* use bindLocalNameFV here; + -- see Note [View pattern usage] + -- see Note [View pattern usage] + -- For the TopLevel case + -- For the TopLevel case + -- see Note [bindLocalNames for an External name] + -- see Note [bindLocalNames for an External name] + -- Do *not* use bindLocalNameFV here; -- see Note [View pattern usage] -- For the TopLevel case @@ -877,7 +915,7 @@ rnHsRecFields ctxt mk_arg (HsRecFields { rec_flds = flds, rec_dotdot = dotdot }) ; return $ L l $ HsFieldBind { hfbAnn = noAnn - , hfbLHS = L loc (FieldOcc sel (L ll arg_rdr)) + , hfbLHS = L loc (FieldOcc arg_rdr (L ll sel)) , hfbRHS = arg' , hfbPun = pun } } @@ -897,7 +935,7 @@ rnHsRecFields ctxt mk_arg (HsRecFields { rec_flds = flds, rec_dotdot = dotdot }) ; (rdr_env, lcl_env) <- getRdrEnvs ; conInfo <- lookupConstructorInfo con ; when (conFieldInfo conInfo == ConHasPositionalArgs) (addErr (TcRnIllegalWildcardsInConstructor con)) - ; let present_flds = mkOccSet $ map rdrNameOcc (getFieldLbls flds) + ; let present_flds = mkOccSet $ map nameOccName (getFieldLbls flds) -- For constructor uses (but not patterns) -- the arg should be in scope locally; @@ -923,7 +961,7 @@ rnHsRecFields ctxt mk_arg (HsRecFields { rec_flds = flds, rec_dotdot = dotdot }) ; return [ L (noAnnSrcSpan loc) (HsFieldBind { hfbAnn = noAnn , hfbLHS - = L (noAnnSrcSpan loc) (FieldOcc sel (L (noAnnSrcSpan loc) arg_rdr)) + = L (noAnnSrcSpan loc) (FieldOcc arg_rdr (L (noAnnSrcSpan loc) sel)) , hfbRHS = L locn (mk_arg loc arg_rdr) , hfbPun = False }) | fl <- dot_dot_fields @@ -1013,10 +1051,10 @@ rnHsRecUpdFields flds -> RnM ([LHsRecUpdField GhcRn GhcRn], FreeVars) rn_flds _ _ [] = return ([], emptyFVs) rn_flds pun_ok mb_unambig_lbls - ((L l (HsFieldBind { hfbLHS = L loc f + ((L l (HsFieldBind { hfbLHS = L loc (FieldOcc _ f) , hfbRHS = arg , hfbPun = pun })):flds) - = do { let lbl = ambiguousFieldOccRdrName f + = do { let lbl = unLoc f ; (arg' :: LHsExpr GhcPs) <- if pun then do { setSrcSpanA loc $ checkErr pun_ok (TcRnIllegalFieldPunning (L (locA loc) lbl)) @@ -1025,12 +1063,12 @@ rnHsRecUpdFields flds ; return (L (l2l loc) (HsVar noExtField (L (l2l loc) arg_rdr))) } else return arg ; (arg'', fvs) <- rnLExpr arg' - ; let lbl' :: AmbiguousFieldOcc GhcRn + ; let lbl' :: FieldOcc GhcRn lbl' = case mb_unambig_lbls of { Just (fl:_) -> let sel_name = flSelector fl - in Unambiguous sel_name (L (l2l loc) lbl) - ; _ -> Ambiguous noExtField (L (l2l loc) lbl) } + in FieldOcc lbl (L (l2l loc) sel_name) + ; _ -> XFieldOcc (Ambiguous (L (l2l loc) lbl)) } fld' :: LHsRecUpdField GhcRn GhcRn fld' = L l (HsFieldBind { hfbAnn = noAnn , hfbLHS = L loc lbl' @@ -1042,9 +1080,10 @@ rnHsRecUpdFields flds getFieldIds :: [LHsRecField GhcRn arg] -> [Name] getFieldIds flds = map (hsRecFieldSel . unLoc) flds -getFieldLbls :: forall p arg . UnXRec p => [LHsRecField p arg] -> [RdrName] -getFieldLbls flds - = map (unXRec @p . foLabel . unXRec @p . hfbLHS . unXRec @p) flds +-- The call to foLabel might be partial now. Don't know enough about +-- the rest of the function chain to say if this is an issue. +getFieldLbls :: forall p arg . UnXRec p => [LHsRecField p arg] -> [IdP p] +getFieldLbls = map (unXRec @p . foLabel . unXRec @p . hfbLHS . unXRec @p) needFlagDotDot :: HsRecFieldContext -> TcRnMessage needFlagDotDot = TcRnIllegalWildcardsInRecord . toRecordFieldPart ===================================== compiler/GHC/Tc/Gen/Expr.hs ===================================== @@ -1301,7 +1301,7 @@ expandRecordUpd record_expr possible_parents rbnds res_ty ; (cons, rbinds) <- disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty ; let upd_flds = map (unLoc . hfbLHS . unLoc) rbinds - sel_ids = map selectorAmbiguousFieldOcc upd_flds + sel_ids = map (unLoc . foLabel) upd_flds upd_fld_names = map idName sel_ids relevant_cons = nonDetEltsUniqSet cons relevant_con = head relevant_cons @@ -1584,7 +1584,7 @@ disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty -> TcM (LHsRecUpdField GhcTc GhcRn) lookupField fld_gre (L l upd) = do { let L loc af = hfbLHS upd - lbl = ambiguousFieldOccRdrName af + lbl = fieldOccRdrName af mb_gre = pickGREs lbl [fld_gre] -- NB: this GRE can be 'Nothing' when in GHCi. -- See test T10439. @@ -1596,7 +1596,7 @@ disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty ; sel <- tcLookupId (greName fld_gre) ; return $ L l HsFieldBind { hfbAnn = hfbAnn upd - , hfbLHS = L (l2l loc) $ Unambiguous sel (L (l2l loc) lbl) + , hfbLHS = L (l2l loc) $ FieldOcc lbl (L (l2l loc) sel) , hfbRHS = hfbRHS upd , hfbPun = hfbPun upd } } @@ -1669,11 +1669,11 @@ fieldCtxt field_name tcRecordField :: ConLike -> Assoc Name Type -> LFieldOcc GhcRn -> LHsExpr GhcRn -> TcM (Maybe (LFieldOcc GhcTc, LHsExpr GhcTc)) -tcRecordField con_like flds_w_tys (L loc (FieldOcc sel_name lbl)) rhs +tcRecordField con_like flds_w_tys (L loc (FieldOcc rdr (L l sel_name))) rhs | Just field_ty <- assocMaybe flds_w_tys sel_name = addErrCtxt (fieldCtxt field_lbl) $ do { rhs' <- tcCheckPolyExprNC rhs field_ty - ; hasFixedRuntimeRep_syntactic (FRRRecordCon (unLoc lbl) (unLoc rhs')) + ; hasFixedRuntimeRep_syntactic (FRRRecordCon rdr (unLoc rhs')) field_ty ; let field_id = mkUserLocal (nameOccName sel_name) (nameUnique sel_name) @@ -1682,12 +1682,13 @@ tcRecordField con_like flds_w_tys (L loc (FieldOcc sel_name lbl)) rhs -- (so we can find it easily) -- but is a LocalId with the appropriate type of the RHS -- (so the expansion knows the type of local binder to make) - ; return (Just (L loc (FieldOcc field_id lbl), rhs')) } + ; return (Just (L loc (FieldOcc rdr (L l field_id)), rhs')) } | otherwise = do { addErrTc (badFieldConErr (getName con_like) field_lbl) ; return Nothing } where - field_lbl = FieldLabelString $ occNameFS $ rdrNameOcc (unLoc lbl) + field_lbl = FieldLabelString $ occNameFS $ rdrNameOcc rdr +tcRecordField _ _ (L _ (XFieldOcc (Ambiguous _))) _ = pure Nothing checkMissingFields :: ConLike -> HsRecordBinds GhcRn -> [Scaled TcType] -> TcM () ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -567,7 +567,7 @@ tcInferAppHead_maybe :: HsExpr GhcRn tcInferAppHead_maybe fun = case fun of HsVar _ (L _ nm) -> Just <$> tcInferId nm - HsRecSel _ f -> Just <$> tcInferRecSelId f + HsRecSel _ f -> tcInferRecSelId f ExprWithTySig _ e hs_ty -> Just <$> tcExprWithSig e hs_ty HsOverLit _ lit -> Just <$> tcInferOverLit lit _ -> return Nothing @@ -596,21 +596,20 @@ addHeadCtxt fun_ctxt thing_inside ********************************************************************* -} tcInferRecSelId :: FieldOcc GhcRn - -> TcM (HsExpr GhcTc, TcSigmaType) -tcInferRecSelId (FieldOcc sel_name lbl) - = do { sel_id <- tc_rec_sel_id - ; let expr = HsRecSel noExtField (FieldOcc sel_id lbl) - ; return (expr, idType sel_id) + -> TcM (Maybe (HsExpr GhcTc, TcSigmaType)) +tcInferRecSelId (FieldOcc sel_name (L l n)) + = do { sel_id <- tc_rec_sel_id + ; let expr = HsRecSel noExtField (FieldOcc sel_name (L l sel_id)) + ; return $ Just(expr, idType sel_id) } where occ :: OccName - occ = rdrNameOcc (unLoc lbl) - + occ = nameOccName n tc_rec_sel_id :: TcM TcId -- Like tc_infer_id, but returns an Id not a HsExpr, -- so we can wrap it back up into a HsRecSel tc_rec_sel_id - = do { thing <- tcLookup sel_name + = do { thing <- tcLookup n ; case thing of ATcId { tct_id = id } -> do { check_naughty occ id -- See Note [Local record selectors] @@ -625,6 +624,7 @@ tcInferRecSelId (FieldOcc sel_name lbl) -- hence no checkTh stuff here _ -> failWithTc $ TcRnExpectedValueId thing } +tcInferRecSelId (XFieldOcc _) = pure Nothing ------------------------ ===================================== compiler/GHC/Tc/Gen/Pat.hs ===================================== @@ -1645,15 +1645,23 @@ tcConValArgs con_like arg_tys penv con_args thing_inside = case con_args of where tc_field :: Checker (LHsRecField GhcRn (LPat GhcRn)) (LHsRecField GhcTc (LPat GhcTc)) - tc_field penv - (L l (HsFieldBind ann (L loc (FieldOcc sel (L lr rdr))) pat pun)) - thing_inside + tc_field penv + (L l (HsFieldBind ann (L loc (FieldOcc rdr (L lr sel))) pat pun)) + thing_inside = do { sel' <- tcLookupId sel ; pat_ty <- setSrcSpanA loc $ find_field_ty sel (occNameFS $ rdrNameOcc rdr) ; (pat', res) <- tcConArg penv (pat, pat_ty) thing_inside - ; return (L l (HsFieldBind ann (L loc (FieldOcc sel' (L lr rdr))) pat' + ; return (L l (HsFieldBind ann (L loc (FieldOcc rdr (L lr sel'))) pat' pun), res) } + tc_field _ + (L _ (HsFieldBind _ (L _ (XFieldOcc (Ambiguous (L _ _)))) _ _)) + _ + -- I don't like leaving things undefined, and I don't like leaving + -- leaving pattern not matched. I think I should just throw in this + -- case, but I don't know exactly how to do that. + = undefined + -- See Note [Omitted record fields and linearity] check_omitted_fields_multiplicity :: TcM MultiplicityCheckCoercions check_omitted_fields_multiplicity = do @@ -1682,7 +1690,7 @@ tcConValArgs con_like arg_tys penv con_args thing_inside = case con_args of (bound_field_tys, omitted_field_tys) = partition is_bound all_field_tys is_bound :: (Maybe FieldLabel, Scaled TcType) -> Bool - is_bound (Just fl, _) = elem (flSelector fl) (map (\(L _ (HsFieldBind _ (L _ (FieldOcc sel _ )) _ _)) -> sel) rpats) + is_bound (Just fl, _) = elem (flSelector fl) (map (\(L _ (HsFieldBind _ (L _ (FieldOcc _ sel )) _ _)) -> unLoc sel) rpats) is_bound _ = False all_field_tys :: [(Maybe FieldLabel, Scaled TcType)] ===================================== compiler/GHC/Tc/TyCl/Utils.hs ===================================== @@ -946,8 +946,7 @@ mkOneRecordSelector all_cons idDetails fl has_sel rec_field = noLocA (HsFieldBind { hfbAnn = noAnn , hfbLHS - = L locc (FieldOcc sel_name - (L locn $ mkRdrUnqual (nameOccName sel_name))) + = L locc (FieldOcc (mkRdrUnqual $ nameOccName sel_name) (L locn sel_name)) , hfbRHS = L loc' (VarPat noExtField (L locn field_var)) , hfbPun = False }) ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hoogle.hs ===================================== @@ -300,7 +300,7 @@ ppCtor sDocContext dat subdocs con at ConDeclH98{con_args = con_args'} = f (RecCon (L _ recs)) = f (PrefixCon [] $ map (hsLinear . cd_fld_type . unLoc) recs) ++ concat - [ (concatMap (lookupCon sDocContext subdocs . noLocA . foExt . unLoc) (cd_fld_names r)) + [ (concatMap (lookupCon sDocContext subdocs . noLocA . unLoc . foLabel . unLoc) (cd_fld_names r)) ++ [out sDocContext (map (foExt . unLoc) $ cd_fld_names r) `typeSig` [resType, cd_fld_type r]] | r <- map unLoc recs ] ===================================== utils/haddock/haddock-api/src/Haddock/Backends/LaTeX.hs ===================================== @@ -1027,7 +1027,7 @@ ppSideBySideConstr subdocs unicode leader (L _ con) = ppSideBySideField :: [(DocName, DocForDecl DocName)] -> Bool -> ConDeclField DocNameI -> LaTeX ppSideBySideField subdocs unicode (ConDeclField _ names ltype _) = decltt - ( cat (punctuate comma (map (ppBinder . rdrNameOcc . unLoc . foLabel . unLoc) names)) + ( cat (punctuate comma (map (ppBinder . rdrNameOcc . foExt . unLoc) names)) <+> dcolon unicode <+> ppLType unicode ltype ) @@ -1035,7 +1035,7 @@ ppSideBySideField subdocs unicode (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc name) subdocs >>= fmap _doc . combineDocumentation . fst + mbDoc = lookup (unLoc . foLabel . unLoc $ name) subdocs >>= fmap _doc . combineDocumentation . fst name = case Maybe.listToMaybe names of Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Xhtml/Decl.hs ===================================== @@ -1536,7 +1536,7 @@ ppSideBySideField subdocs unicode qual (ConDeclField _ names ltype _) = comma [ ppBinder False (rdrNameOcc field) | L _ name <- names - , let field = (unLoc . foLabel) name + , let field = (foExt) name ] ) <+> dcolon unicode @@ -1547,14 +1547,14 @@ ppSideBySideField subdocs unicode qual (ConDeclField _ names ltype _) = where -- don't use cd_fld_doc for same reason we don't use con_doc above -- Where there is more than one name, they all have the same documentation - mbDoc = lookup (foExt $ unLoc declName) subdocs >>= combineDocumentation . fst + mbDoc = lookup (unLoc . foLabel $ unLoc declName) subdocs >>= combineDocumentation . fst declName = case Maybe.listToMaybe names of Nothing -> error "No names. An invariant was broken. Please report this to the Haddock project" Just hd -> hd ppShortField :: Bool -> Unicode -> Qualification -> ConDeclField DocNameI -> Html ppShortField summary unicode qual (ConDeclField _ names ltype _) = - hsep (punctuate comma (map ((ppBinder summary) . rdrNameOcc . unLoc . foLabel . unLoc) names)) + hsep (punctuate comma (map ((ppBinder summary) . rdrNameOcc . foExt . unLoc) names)) <+> dcolon unicode <+> ppLType unicode qual HideEmptyContexts ltype ===================================== utils/haddock/haddock-api/src/Haddock/Convert.hs ===================================== @@ -496,7 +496,7 @@ synifyDataCon use_gadt_syntax dc = noLocA $ ConDeclField noAnn - [noLocA $ FieldOcc (flSelector fl) (noLocA $ mkVarUnqual $ field_label $ flLabel fl)] + [noLocA $ FieldOcc (mkVarUnqual $ field_label $ flLabel fl) (noLocA (flSelector fl))] synTy Nothing ===================================== utils/haddock/haddock-api/src/Haddock/GhcUtils.hs ===================================== @@ -374,7 +374,7 @@ restrictCons names decls = [L p d | L p (Just d) <- fmap keep <$> decls] field_avail :: LConDeclField GhcRn -> Bool field_avail (L _ (ConDeclField _ fs _ _)) = - all (\f -> foExt (unLoc f) `elem` names) fs + all (\f -> (unLoc . foLabel . unLoc $ f) `elem` names) fs field_types flds = [hsUnrestricted t | L _ (ConDeclField _ _ t _) <- flds] keep _ = Nothing @@ -563,7 +563,7 @@ instance Parent (ConDecl GhcRn) where children con = case getRecConArgs_maybe con of Nothing -> [] - Just flds -> map (foExt . unLoc) $ concatMap (cd_fld_names . unLoc) (unLoc flds) + Just flds -> map (unLoc . foLabel . unLoc) $ concatMap (cd_fld_names . unLoc) (unLoc flds) instance Parent (TyClDecl GhcRn) where children d ===================================== utils/haddock/haddock-api/src/Haddock/Interface/Create.hs ===================================== @@ -927,7 +927,7 @@ extractDecl prr dflags sDocContext name decl Just rec <- toList $ getRecConArgs_maybe . unLoc <$> dd_cons (feqn_rhs d) , ConDeclField{cd_fld_names = ns} <- map unLoc (unLoc rec) , L _ n <- ns - , foExt n == name + , unLoc (foLabel n) == name ] in case matches of [d0] -> extractDecl prr dflags sDocContext name (noLocA . InstD noExtField $ DataFamInstD noExtField d0) @@ -996,7 +996,7 @@ extractRecSel nm t tvs (L _ con : rest) = where matching_fields :: [LConDeclField GhcRn] -> [(SrcSpan, LConDeclField GhcRn)] matching_fields flds = - [ (locA l, f) | f@(L _ (ConDeclField _ ns _ _)) <- flds, L l n <- ns, foExt n == nm + [ (locA l, f) | f@(L _ (ConDeclField _ ns _ _)) <- flds, L l n <- ns, unLoc (foLabel n) == nm ] data_ty -- ResTyGADT _ ty <- con_res con = ty ===================================== utils/haddock/haddock-api/src/Haddock/Interface/Rename.hs ===================================== @@ -745,9 +745,9 @@ renameConDeclFieldField (L l (ConDeclField _ names t doc)) = do return $ L (locA l) (ConDeclField noExtField names' t' doc') renameLFieldOcc :: LFieldOcc GhcRn -> RnM (LFieldOcc DocNameI) -renameLFieldOcc (L l (FieldOcc sel lbl)) = do +renameLFieldOcc (L l (FieldOcc rdr (L n sel))) = do sel' <- renameName sel - return $ L l (FieldOcc sel' lbl) + return $ L l (FieldOcc rdr (L n sel')) renameSig :: Sig GhcRn -> RnM (Sig DocNameI) renameSig sig = case sig of ===================================== utils/haddock/haddock-api/src/Haddock/Types.hs ===================================== @@ -885,8 +885,8 @@ type instance XUserTyVar DocNameI = NoExtField type instance XKindedTyVar DocNameI = NoExtField type instance XXTyVarBndr DocNameI = DataConCantHappen -type instance XCFieldOcc DocNameI = DocName -type instance XXFieldOcc DocNameI = NoExtField +type instance XCFieldOcc DocNameI = RdrName +type instance XXFieldOcc DocNameI = DataConCantHappen type instance XFixitySig DocNameI = NoExtField type instance XFixSig DocNameI = NoExtField View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/45ea0e5939e14ecbf978375816de20b10142f092 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/45ea0e5939e14ecbf978375816de20b10142f092 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 10:36:26 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Fri, 20 Sep 2024 06:36:26 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] 44 commits: JS: fake support for native adjustors (#25159) Message-ID: <66ed502a510e6_224d81574c481038b0@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 275d7390 by Jade at 2024-09-20T12:35:43+02:00 Refactor FieldOcc vs AmbiguousFieldOcc with TTG Improves the design of FieldOcc vs AmbiguousFieldOcc, and removes a dependency on `RdrName` from the Language.Haskell.* namespace (#21592). The design: * The FieldOcc constructor of FieldOcc always refers to an unambiguous field occurrence. * During renaming, a FieldOcc may be ambiguous and only be resolvable during Typechecking * Therefore, we extend (with TTG) `FieldOcc GhcRn` with a constructor `AmbiguousFieldOcc` that constructs a definitely ambiguous `FieldOcc`. * During typechecking, all ambiguous field occurrences must be resolved, so the `AmbiguousFieldOcc` constructor no longer exists See Note [Lifecycle of a FieldOcc] Co-authored-by: Rodrigo Mesquita <rodrigo.m.mesquita at gmail.com> - - - - - 5c6b288e by Hassan Al-Awwadi at 2024-09-20T12:35:43+02:00 Wrangled until everything compiled. There are two issues: - FieldOcc used to only have one constructor and always be unambigious, this is no longer the case. Calls to foLabel are thus partial now. Don't know how much we care about this, since the partial calls are mostly inside functions that used to operate on the operate on the unambigious FieldOcc - Lots of functions that take in a FieldOcc, or a HsExpr (with the HsRecSel constructor) now have a new case. It was not always clear to me what the correct implementation was for these. I have filled them in as far as I could and left one undefined... - - - - - 12 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/45ea0e5939e14ecbf978375816de20b10142f092...5c6b288e454daa1bbbaa8d078607506520a698ec -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/45ea0e5939e14ecbf978375816de20b10142f092...5c6b288e454daa1bbbaa8d078607506520a698ec You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 11:52:07 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 20 Sep 2024 07:52:07 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Test options Message-ID: <66ed61e72f2ef_36d9d2bcafc7932@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 35c34b15 by Sven Tennie at 2024-09-20T13:51:46+02:00 Test options - - - - - 1 changed file: - .gitlab/jobs.yaml Changes: ===================================== .gitlab/jobs.yaml ===================================== @@ -62,8 +62,10 @@ "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate", - "XZ_OPT": "-9" - } + "XZ_OPT": "-9", + "HADRIAN_ARGS": "--docs=none", + "EXTRA_HC_OPTS": "-fexternal-interpreter' + } }, "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate": { "after_script": [ @@ -126,7 +128,9 @@ "CROSS_TARGET": "riscv64-linux-gnu", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", - "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate" + "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate", + "HADRIAN_ARGS": "--docs=none", + "EXTRA_HC_OPTS": "-fexternal-interpreter' } } } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/35c34b15fa9fd80f31d8f0dd6c4b7b1f9031c24d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/35c34b15fa9fd80f31d8f0dd6c4b7b1f9031c24d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 11:59:30 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 20 Sep 2024 07:59:30 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Test options Message-ID: <66ed63a2a79ca_36d9d2bcbb08042b@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: e8e9d35c by Sven Tennie at 2024-09-20T13:59:21+02:00 Test options - - - - - 1 changed file: - .gitlab/jobs.yaml Changes: ===================================== .gitlab/jobs.yaml ===================================== @@ -62,7 +62,9 @@ "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate", - "XZ_OPT": "-9" + "XZ_OPT": "-9", + "HADRIAN_ARGS": "--docs=none", + "EXTRA_HC_OPTS": "-fexternal-interpreter" } }, "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate": { @@ -126,7 +128,9 @@ "CROSS_TARGET": "riscv64-linux-gnu", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", - "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate" + "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate", + "HADRIAN_ARGS": "--docs=none", + "EXTRA_HC_OPTS": "-fexternal-interpreter' } } } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e8e9d35cf99fae2f3c9bca1725ce3b3f3247a983 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e8e9d35cf99fae2f3c9bca1725ce3b3f3247a983 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 12:01:58 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 20 Sep 2024 08:01:58 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Test options Message-ID: <66ed6436c86eb_36d9d21c20788093a@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 00b915ba by Sven Tennie at 2024-09-20T14:01:50+02:00 Test options - - - - - 1 changed file: - .gitlab/jobs.yaml Changes: ===================================== .gitlab/jobs.yaml ===================================== @@ -62,7 +62,9 @@ "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate", - "XZ_OPT": "-9" + "XZ_OPT": "-9", + "HADRIAN_ARGS": "--docs=none", + "EXTRA_HC_OPTS": "-fexternal-interpreter" } }, "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate": { @@ -126,7 +128,9 @@ "CROSS_TARGET": "riscv64-linux-gnu", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", - "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate" + "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate", + "HADRIAN_ARGS": "--docs=none", + "EXTRA_HC_OPTS": "-fexternal-interpreter" } } } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/00b915bac2594afd648c93a9b3c41aee4b32b5bc -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/00b915bac2594afd648c93a9b3c41aee4b32b5bc You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 12:24:26 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 20 Sep 2024 08:24:26 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 17 commits: base: Propagate `error` CallStack to thrown exception Message-ID: <66ed697a60c63_36d9d2269990829cc@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 163b63ed by Ben Gamari at 2024-09-20T08:23:57-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes #24807. - - - - - 98e79c68 by Matthew Pickering at 2024-09-20T08:23:57-04:00 ci: Run abi-test on test-abi label - - - - - 67a8356a by Rodrigo Mesquita at 2024-09-20T08:23:57-04:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - 38185b4f by Rodrigo Mesquita at 2024-09-20T08:23:57-04:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 6ca5b05d by Rodrigo Mesquita at 2024-09-20T08:23:57-04:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - be5c4aa3 by Rodrigo Mesquita at 2024-09-20T08:23:57-04:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 1f54463a by Rodrigo Mesquita at 2024-09-20T08:23:57-04:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 8a2c862a by Rodrigo Mesquita at 2024-09-20T08:23:57-04:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 577c7dee by Rodrigo Mesquita at 2024-09-20T08:23:57-04:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7c0c11bc by doyougnu at 2024-09-20T08:23:58-04:00 linker: add --optimistic-linking flag This patch adds: - the --optimistic-linking flag which binds unknown symbols in the runtime linker to 0xDEADBEEF instead of exiting with failure - The test T25240 which tests these flags using dead code in the FFI system. - closes #25240 This patch is part of the upstreaming haskell.nix patches project. - - - - - 4125dfd0 by Ryan Hendrickson at 2024-09-20T08:23:59-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - d45db125 by Zubin Duggal at 2024-09-20T08:24:00-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 5924515d by Andreas Klebinger at 2024-09-20T08:24:01-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - bf31f6ad by Sylvain Henry at 2024-09-20T08:24:13-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - 2d7b4db4 by Sylvain Henry at 2024-09-20T08:24:13-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - bfc246a6 by Sven Tennie at 2024-09-20T08:24:14-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - 60db6c25 by Sylvain Henry at 2024-09-20T08:24:17-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/RV64/CodeGen.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e1bfbb7a68249d2174c84280215e4cc4e562435f...60db6c2581d1e80d04b2a56576d54677f36db069 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e1bfbb7a68249d2174c84280215e4cc4e562435f...60db6c2581d1e80d04b2a56576d54677f36db069 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 12:50:32 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Fri, 20 Sep 2024 08:50:32 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 8 commits: Add Broadcast MachOps Message-ID: <66ed6f98a1340_36d9d2537f50995a3@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: b825f971 by sheaf at 2024-09-20T14:50:18+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - dc42d915 by sheaf at 2024-09-20T14:50:18+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 148635c8 by sheaf at 2024-09-20T14:50:18+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - ed7a48f3 by sheaf at 2024-09-20T14:50:18+02:00 Add test for C calls & SIMD vectors - - - - - 9480fb17 by sheaf at 2024-09-20T14:50:18+02:00 Add test for #25169 - - - - - 020e35e9 by sheaf at 2024-09-20T14:50:19+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 1845b415 by sheaf at 2024-09-20T14:50:19+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 57defd65 by sheaf at 2024-09-20T14:50:19+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Reg.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/Wasm/Asm.hs - compiler/GHC/CmmToAsm/Wasm/FromCmm.hs - compiler/GHC/CmmToAsm/Wasm/Types.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToAsm/X86/Instr.hs - compiler/GHC/CmmToAsm/X86/Ppr.hs - compiler/GHC/CmmToC.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs - compiler/GHC/JS/Make.hs - compiler/GHC/Llvm/Types.hs - compiler/GHC/StgToByteCode.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f36849c917a6bc0d4ed96b1703e60d1dc4b10447...57defd651d44ee65be01c7508166a50f4a2c8de7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f36849c917a6bc0d4ed96b1703e60d1dc4b10447...57defd651d44ee65be01c7508166a50f4a2c8de7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 12:59:08 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Fri, 20 Sep 2024 08:59:08 -0400 Subject: [Git][ghc/ghc][wip/T25107] Demand: Combine examples into Note (#25107) Message-ID: <66ed719c2f8d3_36d9d2535d7c1048af@gitlab.mail> Sebastian Graf pushed to branch wip/T25107 at Glasgow Haskell Compiler / GHC Commits: 497d41b8 by Sebastian Graf at 2024-09-20T14:59:01+02:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 2 changed files: - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Types/Demand.hs Changes: ===================================== compiler/GHC/Core/Opt/DmdAnal.hs ===================================== @@ -1970,10 +1970,10 @@ W/W-transformed /caller of/ 'f' would immediately rebox any unboxed arguments that is applied to the wrapper of 'f'. When the wrapper is inlined, that kind of reboxing does not happen. -But now we have functions with OPAQUE pragmas, which by definition (See Note -[OPAQUE pragma]) do not get W/W-transformed. So in order to avoid reboxing -workers of any W/W-transformed /callers of/ 'f' we need to strip all boxity -information from 'f' in the demand analysis. This will inform the +But now we have functions with OPAQUE pragmas, which by definition +(See Note [OPAQUE pragma]) do not get W/W-transformed. So in order to avoid +reboxing workers of any W/W-transformed /callers of/ 'f' we need to strip all +boxity information from 'f' in the demand analysis. This will inform the W/W-transformation code that boxed arguments of 'f' must definitely be passed along in boxed form and as such dissuade the creation of reboxing workers. -} ===================================== compiler/GHC/Types/Demand.hs ===================================== @@ -606,22 +606,8 @@ multCard (Card a) (Card b) -- * How many times a variable is evaluated, via a 'Card'inality, and -- * How deep its value was evaluated in turn, via a 'SubDemand'. -- --- Examples (using Note [Demand notation]): --- --- * 'seq' puts demand `1A` on its first argument: It evaluates the argument --- strictly (`1`), but not any deeper (`A`). --- * 'fst' puts demand `1P(1L,A)` on its argument: It evaluates the argument --- pair strictly and the first component strictly, but no nested info --- beyond that (`L`). Its second argument is not used at all. --- * '$' puts demand `1C(1,L)` on its first argument: It calls (`C`) the --- argument function with one argument, exactly once (`1`). No info --- on how the result of that call is evaluated (`L`). --- * 'maybe' puts demand `MC(M,L)` on its second argument: It evaluates --- the argument function at most once ((M)aybe) and calls it once when --- it is evaluated. --- * `fst p + fst p` puts demand `SP(SL,A)` on `p`: It's `1P(1L,A)` --- multiplied by two, so we get `S` (used at least once, possibly multiple --- times). +-- See also Note [Demand notation] +-- and Note [Demand examples]. -- -- This data type is quite similar to `'Scaled' 'SubDemand'`, but it's scaled -- by 'Card', which is an /interval/ on 'Multiplicity', the upper bound of @@ -2656,12 +2642,8 @@ So, L can denote a 'Card', polymorphic 'SubDemand' or polymorphic 'Demand', but it's always clear from context which "overload" is meant. It's like return-type inference of e.g. 'read'. -Examples are in the haddock for 'Demand'. Here are some more: - SA Strict, but does not look at subcomponents (`seq`) - SP(L,L) Strict boxed pair, components lazy - S!P(L,L) Strict unboxed pair, components lazy - LP(SA,SA) Lazy pair, but if it is evaluated will evaluated its components - LC(1C(L)) Lazy, but if called will apply the result exactly once +An example of the demand syntax is 1!P(1!L,A), the demand of fst's argument. +See Note [Demand examples] for more examples and their semantics. This is the syntax for demand signatures: @@ -2679,7 +2661,39 @@ This is the syntax for demand signatures: (omitted if empty) (omitted if no information) - +Note [Demand examples] +~~~~~~~~~~~~~~~~~~~~~~ +Here are some examples of the demand notation, specified in Note [Demand notation], +in action. In each case we give the demand on the variable `x`. + +Demand on x Example Explanation + 1!A seq x y Evaluates `x` exactly once (`1`), but not + any deeper (`A`), and discards the box (`!`). + S!A seq x (seq x y) Twice the previous demand; hence eval'd + more than once (`S` for strict). + 1!P(1!L,A) fst x Evaluates pair `x` exactly once, first + component exactly once. No info that (`L`). + Second component is absent. Discards boxes (`!`). + 1P(1L,A) opq_fst x Like fst, but all boxes are retained. + SP(1!L,A) opq_seq x (fst x) Two evals of x but exactly one of its first component. + Box of x retained, but box of first component discarded. + 1!C(1,L) x $ 3 Evals x exactly once ( 1 ) and calls it + exactly once ( C(1,_) ). No info on how the + result is evaluated ( L ). + MC(M,L) maybe y x Evals x at most once ( 1 ) and calls it at + most once ( C(1,_) ). No info on how the + result is evaluated ( L ). + LP(SL,A) map (+ fst x) Evals x lazily and multiple times ( L ), + but when it is evaluated, the first + component is evaluated (strictly) as well. + +In the examples above, `opq_fst` is an opaque wrapper around `fst`, i.e. + + opq_fst = fst + {-# OPAQUE opq_fst #-} + +Similarly for `seq`. The effect of an OPAQUE pragma is that it discards any +boxity flags in the demand signature, as described in Note [OPAQUE pragma]. -} -- | See Note [Demand notation] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/497d41b8e41963578af98ff50ed2ce0d2f448af4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/497d41b8e41963578af98ff50ed2ce0d2f448af4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 13:47:07 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Fri, 20 Sep 2024 09:47:07 -0400 Subject: [Git][ghc/ghc][wip/T24824] 367 commits: utils: add hie.yaml config file for ghc-config Message-ID: <66ed7cdbedd79_36d9d2b934581144d4@gitlab.mail> Sebastian Graf pushed to branch wip/T24824 at Glasgow Haskell Compiler / GHC Commits: 7eda4bd2 by Cheng Shao at 2024-05-31T15:52:04-04:00 utils: add hie.yaml config file for ghc-config Add hie.yaml to ghc-config project directory so it can be edited using HLS. - - - - - 1e5752f6 by Cheng Shao at 2024-05-31T15:52:05-04:00 hadrian: handle findExecutable "" gracefully hadrian may invoke findExecutable "" at run-time due to a certain program is not found by configure script. Which is fine and findExecutable is supposed to return Nothing in this case. However, on Windows there's a directory bug that throws an exception (see https://github.com/haskell/directory/issues/180), so we might as well use a wrapper for findExecutable and handle exceptions gracefully. - - - - - 4eb5ad09 by Cheng Shao at 2024-05-31T15:52:05-04:00 configure: do not set LLC/OPT/LLVMAS fallback values when FIND_LLVM_PROG fails When configure fails to find LLC/OPT/LLVMAS within supported version range, it used to set "llc"/"opt"/"clang" as fallback values. This behavior is particularly troublesome when the user has llc/opt/clang with other versions in their PATH and run the testsuite, since hadrian will incorrectly assume have_llvm=True and pass that to the testsuite driver, resulting in annoying optllvm test failures (#23186). If configure determines llc/opt/clang wouldn't work, then we shouldn't pretend it'll work at all, and the bindist configure will invoke FIND_LLVM_PROG check again at install time anyway. - - - - - 5f1afdf7 by Sylvain Henry at 2024-05-31T15:52:52-04:00 Introduce UniqueSet and use it to replace 'UniqSet Unique' 'UniqSet Unique' represents a set of uniques as a 'Map Unique Unique', which is wasting space (associated key/value are always the same). Fix #23572 and #23605 - - - - - e0aa42b9 by crumbtoo at 2024-05-31T15:53:33-04:00 Improve template-haskell haddocks Closes #15822 - - - - - ae170155 by Olivier Benz at 2024-06-01T09:35:17-04:00 Bump max LLVM version to 19 (not inclusive) - - - - - 92aa65ea by Matthew Pickering at 2024-06-01T09:35:17-04:00 ci: Update CI images to test LLVM 18 The debian12 image in this commit has llvm 18 installed. - - - - - adb1fe42 by Serge S. Gulin at 2024-06-01T09:35:53-04:00 Unicode: make ucd2haskell build-able again ucd2haskell tool used streamly library which version in cabal was out of date. It is updated to the latest version at hackage with deprecated parts rewritten. Also following fixes were applied to existing code in suppose that from its last run the code structure was changed and now it was required to be up to date with actual folder structures: 1. Ghc module path environment got a suffix with `src`. 2. Generated code got 2.1 `GHC.Internal` prefix for `Data.*`. 2.2 `GHC.Unicode.Internal` swapped on `GHC.Internal.Unicode` according to actual structure. - - - - - ad56fd84 by Jade at 2024-06-01T09:36:29-04:00 Replace 'NB' with 'Note' in error messages - - - - - 6346c669 by Cheng Shao at 2024-06-01T09:37:04-04:00 compiler: fix -ddump-cmm-raw when compiling .cmm This patch fixes missing -ddump-cmm-raw output when compiling .cmm, which is useful for debugging cmm related codegen issues. - - - - - 1c834ad4 by Ryan Scott at 2024-06-01T09:37:40-04:00 Print namespace specifiers in FixitySig's Outputable instance For whatever reason, the `Outputable` instance for `FixitySig` simply did not print out namespace specifiers, leading to the confusing `-ddump-splices` output seen in #24911. This patch corrects this oversight. Fixes #24911. - - - - - cf49fb5f by Sylvain Henry at 2024-06-01T09:38:19-04:00 Configure: display C++ compiler path - - - - - f9c1ae12 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable PIC for in-tree GMP on wasm32 This patch disables PIC for in-tree GMP on wasm32 target. Enabling PIC unconditionally adds undesired code size and runtime overhead for wasm32. - - - - - 1a32f828 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable in-tree gmp fft code path for wasm32 This patch disables in-tree GMP FFT code paths for wasm32 target in order to give up some performance of multiplying very large operands in exchange for reduced code size. - - - - - 06277d56 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: build in-tree GMP with malloc-notreentrant on wasm32 This patch makes hadrian build in-tree GMP with the --enable-alloca=malloc-notreentrant configure option. We will only need malloc-reentrant when we have threaded RTS and SMP support on wasm32, which will take some time to happen, before which we should use malloc-notreentrant to avoid undesired runtime overhead. - - - - - 9f614270 by ARATA Mizuki at 2024-06-02T14:02:35-04:00 Set package include paths when assembling .S files Fixes #24839. Co-authored-by: Sylvain Henry <hsyl20 at gmail.com> - - - - - 4998a6ed by Alex Mason at 2024-06-03T02:09:29-04:00 Improve performance of genericWordQuotRem2Op (#22966) Implements the algorithm from compiler-rt's udiv128by64to64default. This rewrite results in a roughly 24x improvement in runtime on AArch64 (and likely any other arch that uses it). - - - - - ae50a8eb by Cheng Shao at 2024-06-03T02:10:05-04:00 testsuite: mark T7773 as fragile on wasm - - - - - c8ece0df by Fendor at 2024-06-03T19:43:22-04:00 Migrate `Finder` component to `OsPath`, fixed #24616 For each module in a GHCi session, we keep alive one `ModLocation`. A `ModLocation` is fairly inefficiently packed, as `String`s are expensive in memory usage. While benchmarking the agda codebase, we concluded that we keep alive around 11MB of `FilePath`'s, solely retained by `ModLocation`. We provide a more densely packed encoding of `ModLocation`, by moving from `FilePath` to `OsPath`. Further, we migrate the full `Finder` component to `OsPath` to avoid unnecessary transformations. As the `Finder` component is well-encapsulated, this requires only a minimal amount of changes in other modules. We introduce pattern synonym for 'ModLocation' which maintains backwards compatibility and avoids breaking consumers of 'ModLocation'. - - - - - 0cff083a by Cheng Shao at 2024-06-03T19:43:58-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. - - - - - 18f63970 by Sebastian Graf at 2024-06-04T05:05:27-04:00 Parser: Remove unused `apats` rule - - - - - 38757c30 by David Knothe at 2024-06-04T05:05:27-04:00 Implement Or Patterns (#22596) This commit introduces a new language extension, `-XOrPatterns`, as described in GHC Proposal 522. An or-pattern `pat1; ...; patk` succeeds iff one of the patterns `pat1`, ..., `patk` succeed, in this order. See also the summary `Note [Implmentation of OrPatterns]`. Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> - - - - - 395412e8 by Cheng Shao at 2024-06-04T05:06:04-04:00 compiler/ghci/rts: remove stdcall support completely We have formally dropped i386 windows support (#18487) a long time ago. The stdcall foreign call convention is only used by i386 windows, and the legacy logic around it is a significant maintenance burden for future work that adds arm64 windows support (#24603). Therefore, this patch removes stdcall support completely from the compiler as well as the RTS (#24883): - stdcall is still recognized as a FFI calling convention in Haskell syntax. GHC will now unconditionally emit a warning (-Wunsupported-calling-conventions) and treat it as ccall. - Apart from minimum logic to support the parsing and warning logic, all other code paths related to stdcall has been completely stripped from the compiler. - ghci only supports FFI_DEFAULT_ABI and ccall convention from now on. - FFI foreign export adjustor code on all platforms no longer handles the stdcall case and only handles ccall from now on. - The Win32 specific parts of RTS no longer has special code paths for stdcall. This commit is the final nail on the coffin for i386 windows support. Further commits will perform more housecleaning to strip the legacy code paths and pave way for future arm64 windows support. - - - - - d1fe9ab6 by Cheng Shao at 2024-06-04T05:06:04-04:00 rts: remove legacy i386 windows code paths This commit removes some legacy i386 windows related code paths in the RTS, given this target is no longer supported. - - - - - a605e4b2 by Cheng Shao at 2024-06-04T05:06:04-04:00 autoconf: remove i386 windows related logic This commit removes legacy i386 windows logic in autoconf scripts. - - - - - 91e5ac5e by Cheng Shao at 2024-06-04T05:06:04-04:00 llvm-targets: remove i386 windows support This commit removes i386 windows from llvm-targets and the script to generate it. - - - - - 65fe75a4 by Cheng Shao at 2024-06-04T05:06:04-04:00 libraries/utils: remove stdcall related legacy logic This commit removes stdcall related legacy logic in libraries and utils. ccall should be used uniformly for all supported windows hosts from now on. - - - - - d2a83302 by Cheng Shao at 2024-06-04T05:06:04-04:00 testsuite: adapt the testsuite for stdcall removal This patch adjusts test cases to handle the stdcall removal: - Some stdcall usages are replaced with ccall since stdcall doesn't make sense anymore. - We also preserve some stdcall usages, and check in the expected warning messages to ensure GHC always warn about stdcall usages (-Wunsupported-calling-conventions) as expected. - Error code testsuite coverage is slightly improved, -Wunsupported-calling-conventions is now tested. - Obsolete code paths related to i386 windows are also removed. - - - - - cef8f47a by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: minor adjustments for stdcall removal This commit include minor adjustments of documentation related to stdcall removal. - - - - - 54332437 by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: mention i386 Windows removal in 9.12 changelog This commit mentions removal of i386 Windows support and stdcall related change in the 9.12 changelog. - - - - - 2aaea8a1 by Cheng Shao at 2024-06-04T05:06:40-04:00 hadrian: improve user settings documentation This patch adds minor improvements to hadrian user settings documentation: - Add missing `ghc.cpp.opts` case - Remove non-existent `cxx` case - Clarify `cc.c.opts` also works for C++, while `cc.deps.opts` doesn't - Add example of passing configure argument to autoconf packages - - - - - 71010381 by Alex Mason at 2024-06-04T12:09:07-04:00 Add AArch64 CLZ, CTZ, RBIT primop implementations. Adds support for emitting the clz and rbit instructions, which are used by GHC.Prim.clz*#, GHC.Prim.ctz*# and GHC.Prim.bitReverse*#. - - - - - 44e2abfb by Cheng Shao at 2024-06-04T12:09:43-04:00 hadrian: add +text_simdutf flavour transformer to allow building text with simdutf This patch adds a +text_simdutf flavour transformer to hadrian to allow downstream packagers and users that build from source to opt-in simdutf support for text, in order to benefit from SIMD speedup at run-time. It's still disabled by default for the time being. - - - - - 077cb2e1 by Cheng Shao at 2024-06-04T12:09:43-04:00 ci: enable +text_simdutf flavour transformer for wasm jobs This commit enables +text_simdutf flavour transformer for wasm jobs, so text is now built with simdutf support for wasm. - - - - - b23746ad by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Use TemplateHaskellQuotes in instance Lift ByteArray Resolves #24852 - - - - - 3fd25743 by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Mark addrToByteArray as NOINLINE This function should never be inlined in order to keep code size small. - - - - - 98ad1ea5 by Cheng Shao at 2024-06-04T22:51:26-04:00 compiler: remove unused CompilerInfo/LinkerInfo types This patch removes CompilerInfo/LinkerInfo types from the compiler since they aren't actually used anywhere. - - - - - 11795244 by Cheng Shao at 2024-06-05T06:33:17-04:00 rts: remove unused PowerPC/IA64 native adjustor code This commit removes unused PowerPC/IA64 native adjustor code which is never actually enabled by autoconf/hadrian. Fixes #24920. - - - - - 5132754b by Sylvain Henry at 2024-06-05T06:33:57-04:00 RTS: fix warnings with doing*Profiling (#24918) - - - - - accc8c33 by Cheng Shao at 2024-06-05T11:35:36-04:00 hadrian: don't depend on inplace/mingw when --enable-distro-toolchain on Windows - - - - - 6ffbd678 by Cheng Shao at 2024-06-05T11:35:37-04:00 autoconf: normalize paths of some build-time dependencies on Windows This commit applies path normalization via cygpath -m to some build-time dependencies on Windows. Without this logic, the /clang64/bin prefixed msys2-style paths cause the build to fail with --enable-distro-toolchain. - - - - - 075dc6d4 by Cheng Shao at 2024-06-05T11:36:12-04:00 hadrian: remove OSDarwin mention from speedHack This commit removes mentioning of OSDarwin from speedHack, since speedHack is purely for i386 and we no longer support i386 darwin (#24921). - - - - - 83235c4c by Cheng Shao at 2024-06-05T11:36:12-04:00 compiler: remove 32-bit darwin logic This commit removes all 32-bit darwin logic from the compiler, given we no longer support 32-bit apple systems (#24921). Also contains a bit more cleanup of obsolete i386 windows logic. - - - - - 1eb99bc3 by Cheng Shao at 2024-06-05T11:36:12-04:00 rts: remove 32-bit darwin/ios logic This commit removes 32-bit darwin/ios related logic from the rts, given we no longer support them (#24921). - - - - - 24f65892 by Cheng Shao at 2024-06-05T11:36:12-04:00 llvm-targets: remove 32-bit darwin/ios targets This commit removes 32-bit darwin/ios targets from llvm-targets given we no longer support them (#24921). - - - - - ccdbd689 by Cheng Shao at 2024-06-05T11:36:12-04:00 testsuite: remove 32-bit darwin logic This commit removes 32-bit darwin logic from the testsuite given it's no longer supported (#24921). Also contains more cleanup of obsolete i386 windows logic. - - - - - 11d661c4 by Cheng Shao at 2024-06-05T11:36:13-04:00 docs: mention 32-bit darwin/ios removal in 9.12 changelog This commit mentions removal of 32-bit darwin/ios support (#24921) in the 9.12 changelog. - - - - - 7c173310 by Georgi Lyubenov at 2024-06-05T15:17:22-04:00 Add firstA and secondA to Data.Bitraversable Please see https://github.com/haskell/core-libraries-committee/issues/172 for related discussion - - - - - 3b6f9fd1 by Ben Gamari at 2024-06-05T15:17:59-04:00 base: Fix name of changelog Fixes #24899. Also place it under `extra-doc-files` to better reflect its nature and avoid triggering unnecessary recompilation if it changes. - - - - - 1f4d2ef7 by Sebastian Graf at 2024-06-05T15:18:34-04:00 Announce Or-patterns in the release notes for GHC 9.12 (#22596) Leftover from !9229. - - - - - 8650338d by Jan Hrček at 2024-06-06T10:39:24-04:00 Improve haddocks of Language.Haskell.Syntax.Pat.Pat - - - - - 2eee65e1 by Cheng Shao at 2024-06-06T10:40:00-04:00 testsuite: bump T7653 timeout for wasm - - - - - 990fed60 by Sylvain Henry at 2024-06-07T14:45:23-04:00 StgToCmm: refactor opTranslate and friends - Change arguments order to avoid `\args -> ...` lambdas - Fix documentation - Rename StgToCmm options ("big" doesn't mean anything) - - - - - 1afad514 by Sylvain Henry at 2024-06-07T14:45:23-04:00 NCG x86: remove dead code (#5444) Since 6755d833af8c21bbad6585144b10e20ac4a0a1ab this code is dead. - - - - - 595c0894 by Cheng Shao at 2024-06-07T14:45:58-04:00 testsuite: skip objc-hi/objcxx-hi when cross compiling objc-hi/objcxx-hi should be skipped when cross compiling. The existing opsys('darwin') predicate only asserts the host system is darwin but tells us nothing about the target, hence the oversight. - - - - - edfe6140 by qqwy at 2024-06-08T11:23:54-04:00 Replace '?callStack' implicit param with HasCallStack in GHC.Internal.Exception.throw - - - - - 35a64220 by Cheng Shao at 2024-06-08T11:24:30-04:00 rts: cleanup inlining logic This patch removes pre-C11 legacy code paths related to INLINE_HEADER/STATIC_INLINE/EXTERN_INLINE macros, ensure EXTERN_INLINE is treated as static inline in most cases (fixes #24945), and also corrects the comments accordingly. - - - - - 9ea90ed2 by Andrew Lelechenko at 2024-06-08T11:25:06-04:00 CODEOWNERS: add @core-libraries to track base interface changes A low-tech tactical solution for #24919 - - - - - 580fef7b by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update CHANGELOG to reflect current version - - - - - 391ecff5 by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update prologue.txt to reflect package description - - - - - 3dca3b7d by Ben Gamari at 2024-06-09T01:27:57-04:00 compiler: Clarify comment regarding need for MOVABS The comment wasn't clear in stating that it was only applicable to immediate source and memory target operands. - - - - - 6bd850e8 by doyougnu at 2024-06-09T21:02:14-04:00 JS: establish single source of truth for symbols In pursuit of: #22736. This MR moves ad-hoc symbols used throughout the js backend into a single symbols file. Why? First, this cleans up the code by removing ad-hoc strings created on the fly and therefore makes the code more maintainable. Second, it makes it much easier to eventually type these identifiers. - - - - - f3017dd3 by Cheng Shao at 2024-06-09T21:02:49-04:00 rts: replace ad-hoc MYTASK_USE_TLV with proper CC_SUPPORTS_TLS This patch replaces the ad-hoc `MYTASK_USE_TLV` with the `CC_SUPPORTS_TLS` macro. If TLS support is detected by autoconf, then we should use that for managing `myTask` in the threaded RTS. - - - - - e17d7e8c by Ben Gamari at 2024-06-11T05:25:21-04:00 users-guide: Fix stylistic issues in 9.12 release notes - - - - - 8a8a982a by Hugo Peters at 2024-06-11T05:25:57-04:00 fix typo in the simplifier debug output: baling -> bailing - - - - - 16475bb8 by Hécate Moonlight at 2024-06-12T03:07:55-04:00 haddock: Correct the Makefile to take into account Darwin systems - - - - - a2f60da5 by Hécate Kleidukos at 2024-06-12T03:08:35-04:00 haddock: Remove obsolete links to github.com/haskell/haddock in the docs - - - - - de4395cd by qqwy at 2024-06-12T03:09:12-04:00 Add `__GLASGOW_HASKELL_ASSERTS_IGNORED__` as CPP macro name if `-fasserts-ignored is set. This allows users to create their own Control.Exception.assert-like functionality that does something other than raising an `AssertFailed` exception. Fixes #24967 - - - - - 0e9c4dee by Ryan Hendrickson at 2024-06-12T03:09:53-04:00 compiler: add hint to TcRnBadlyStaged message - - - - - 2747cd34 by Simon Peyton Jones at 2024-06-12T12:51:37-04:00 Fix a QuickLook bug This MR fixes the bug exposed by #24676. The problem was that quickLookArg was trying to avoid calling tcInstFun unnecessarily; but it was in fact necessary. But that in turn forced me into a significant refactoring, putting more fields into EValArgQL. Highlights: see Note [Quick Look overview] in GHC.Tc.Gen.App * Instantiation variables are now distinguishable from ordinary unification variables, by level number = QLInstVar. This is treated like "level infinity". See Note [The QLInstVar TcLevel] in GHC.Tc.Utils.TcType. * In `tcApp`, we don't track the instantiation variables in a set Delta any more; instead, we just tell them apart by their level number. * EValArgQL now much more clearly captures the "half-done" state of typechecking an argument, ready for later resumption. See Note [Quick Look at value arguments] in GHC.Tc.Gen.App * Elminated a bogus (never used) fast-path in GHC.Tc.Utils.Instantiate.instCallConstraints See Note [Possible fast path for equality constraints] Many other small refactorings. - - - - - 1b1523b1 by George Thomas at 2024-06-12T12:52:18-04:00 Fix non-compiling extensible record `HasField` example - - - - - 97b141a3 by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Fix hyperlinker source urls (#24907) This fixes a bug introduced by f56838c36235febb224107fa62334ebfe9941aba Links to external modules in the hyperlinker are uniformly generated using splicing the template given to us instead of attempting to construct the url in an ad-hoc manner. - - - - - 954f864c by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Add name anchor to external source urls from documentation page URLs for external source links from documentation pages were missing a splice location for the name. Fixes #24912 - - - - - b0b64177 by Simon Peyton Jones at 2024-06-12T12:53:31-04:00 Prioritise nominal equalities The main payload of this patch is * Prioritise nominal equalities in the constraint solver. This ameliorates the incompleteness of solving for representational constraints over newtypes: see #24887. See (EX2) in Note [Decomposing newtype equalities] in GHC.Tc.Solver.Equality In doing this patch I tripped over some other things that I refactored: * Move `isCoVarType` from `GHC.Core.Type` to `GHC.Core.Predicate` where it seems more at home. * Clarify the "rewrite role" of a constraint. I was very puzzled about what the role of, say `(Eq a)` might be, but see the new Note [The rewrite-role of a constraint]. In doing so I made predTypeEqRel crash when given a non-equality. Usually it expects an equality; but it was being mis-used for the above rewrite-role stuff. - - - - - cb7c1b83 by Liam Goodacre at 2024-06-12T12:54:09-04:00 compiler: missing-deriving-strategies suggested fix Extends the missing-deriving-strategies warning with a suggested fix that includes which deriving strategies were assumed. For info about the warning, see comments for `TcRnNoDerivStratSpecified`, `TcRnNoDerivingClauseStrategySpecified`, & `TcRnNoStandaloneDerivingStrategySpecified`. For info about the suggested fix, see `SuggestExplicitDerivingClauseStrategies` & `SuggestExplicitStandalanoDerivingStrategy`. docs: Rewords missing-deriving-strategies to mention the suggested fix. Resolves #24955 - - - - - 4e36d3a3 by Jan Hrček at 2024-06-12T12:54:48-04:00 Further haddocks improvements in Language.Haskell.Syntax.Pat.Pat - - - - - 558353f4 by Cheng Shao at 2024-06-12T12:55:24-04:00 rts: use page sized mblocks on wasm This patch changes mblock size to page size on wasm. It allows us to simplify our wasi-libc fork, makes it much easier to test third party libc allocators like emmalloc/mimalloc, as well as experimenting with threaded RTS in wasm. - - - - - b3cc5366 by Matthew Pickering at 2024-06-12T23:06:57-04:00 compiler: Make ghc-experimental not wired in If you need to wire in definitions, then place them in ghc-internal and reexport them from ghc-experimental. Ticket #24903 - - - - - 700eeab9 by Hécate Kleidukos at 2024-06-12T23:07:37-04:00 base: Use a more appropriate unicode arrow for the ByteArray diagram This commit rectifies the usage of a unicode arrow in favour of one that doesn't provoke mis-alignment. - - - - - cca7de25 by Matthew Pickering at 2024-06-12T23:08:14-04:00 ghcup-metadata: Fix debian version ranges This was caught by `ghcup-ci` failing and attempting to install a deb12 bindist on deb11. ``` configure: WARNING: m4/prep_target_file.m4: Expecting YES/NO but got in ArSupportsDashL_STAGE0. Defaulting to False. bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by bin/ghc-toolchain-bin) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) ``` Fixes #24974 - - - - - 7b23ce8b by Pierre Le Marre at 2024-06-13T15:35:04-04:00 ucd2haskell: remove Streamly dependency + misc - Remove dead code. - Remove `streamly` dependency. - Process files with `bytestring`. - Replace Unicode files parsers with the corresponding ones from the package `unicode-data-parser`. - Simplify cabal file and rename module - Regenerate `ghc-internal` Unicode files with new header - - - - - 4570319f by Jacco Krijnen at 2024-06-13T15:35:41-04:00 Document how to run haddocks tests (#24976) Also remove ghc 9.7 requirement - - - - - fb629e24 by amesgen at 2024-06-14T00:28:20-04:00 compiler: refactor lower_CmmExpr_Ptr - - - - - def46c8c by amesgen at 2024-06-14T00:28:20-04:00 compiler: handle CmmRegOff in lower_CmmExpr_Ptr - - - - - ce76bf78 by Simon Peyton Jones at 2024-06-14T00:28:56-04:00 Small documentation update in Quick Look - - - - - 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - bfcb306f by Sebastian Graf at 2024-09-20T15:46:53+02:00 Pmc: Improve implementation of -Wincomplete-record-selectors (#24824, #24891) We now incorporate the result type of unsaturated record selector applications as well as consider long-distance information in getField applications. See the updated Note [Detecting incomplete record selectors]. Fixes #24824 and #24891. - - - - - 024addc0 by Sebastian Graf at 2024-09-20T15:46:53+02:00 Pmc: Improve warning messages of -Wincomplete-record-selectors ... as suggested by Adam Gundry in !12685. - - - - - 19 changed files: - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC.hs - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/PrimOps.hs - compiler/GHC/Builtin/Types.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f92cc9a869a4f25c72ea4c03c2c06efee169a236...024addc0420e57cfc882e6af0a4af2ec7409c413 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f92cc9a869a4f25c72ea4c03c2c06efee169a236...024addc0420e57cfc882e6af0a4af2ec7409c413 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 14:56:45 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Fri, 20 Sep 2024 10:56:45 -0400 Subject: [Git][ghc/ghc][wip/andreask/ghc-primops] 35 commits: RTS linker: add support for hidden symbols (#25191) Message-ID: <66ed8d2d6d324_36d9d2fecec8144547@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/ghc-primops at Glasgow Haskell Compiler / GHC Commits: 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 79b07e5c by Andreas Klebinger at 2024-09-20T16:37:22+02:00 ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps This will be the new place for functions that would have gone into GHC.Exts in the past but are not stable enough to do so now. Addresses #25242 - - - - - 12 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ac9b303dda899ca82b5671375aac1dfcc485a62f...79b07e5cc766f5445366c12aba2c4fb236c15181 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ac9b303dda899ca82b5671375aac1dfcc485a62f...79b07e5cc766f5445366c12aba2c4fb236c15181 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 14:59:27 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Fri, 20 Sep 2024 10:59:27 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 14 commits: The X86 SIMD patch. Message-ID: <66ed8dcfe64a3_36d9d2fef59c144955@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: b1681978 by sheaf at 2024-09-20T16:58:37+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - d9821a7e by sheaf at 2024-09-20T16:58:40+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 5e4e1f64 by sheaf at 2024-09-20T16:58:40+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 95787b4d by sheaf at 2024-09-20T16:58:41+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 1be7e9fc by sheaf at 2024-09-20T16:58:41+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 9cacd56a by sheaf at 2024-09-20T16:58:41+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - fdff26e9 by sheaf at 2024-09-20T16:58:41+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 8b6e545a by sheaf at 2024-09-20T16:58:41+02:00 Add test for C calls & SIMD vectors - - - - - 8035cbeb by sheaf at 2024-09-20T16:58:42+02:00 Add test for #25169 - - - - - b788180d by sheaf at 2024-09-20T16:58:42+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - b17d19ea by sheaf at 2024-09-20T16:59:12+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 3a08a57e by sheaf at 2024-09-20T16:59:13+02:00 X86 genCCall64: simply loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 57bd478d by sheaf at 2024-09-20T16:59:13+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - a34a4dc2 by sheaf at 2024-09-20T16:59:13+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/57defd651d44ee65be01c7508166a50f4a2c8de7...a34a4dc2269300b31c06173dad5d64f55317ecb4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/57defd651d44ee65be01c7508166a50f4a2c8de7...a34a4dc2269300b31c06173dad5d64f55317ecb4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 15:08:29 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Fri, 20 Sep 2024 11:08:29 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 3 commits: X86 CodeGen: refactor getRegister CmmLit Message-ID: <66ed8fedd6671_36d9d213019d81463fd@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: df5b64c4 by sheaf at 2024-09-20T17:08:06+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - cf56ab12 by sheaf at 2024-09-20T17:08:06+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - c27747f2 by sheaf at 2024-09-20T17:08:06+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 9 changed files: - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToAsm/X86/Instr.hs - compiler/GHC/CmmToAsm/X86/Ppr.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs Changes: ===================================== compiler/GHC/CmmToAsm/Format.hs ===================================== @@ -28,6 +28,7 @@ module GHC.CmmToAsm.Format ( scalarWidth, formatInBytes, isFloatScalarFormat, + isFloatOrFloatVecFormat, floatScalarFormat, scalarFormatFormat, VirtualRegWithFormat(..), @@ -134,6 +135,11 @@ isFloatScalarFormat = \case FmtDouble -> True _ -> False +isFloatOrFloatVecFormat :: Format -> Bool +isFloatOrFloatVecFormat = \case + VecFormat _ sFmt -> isFloatScalarFormat sFmt + fmt -> isFloatFormat fmt + floatScalarFormat :: Width -> ScalarFormat floatScalarFormat W32 = FmtFloat floatScalarFormat W64 = FmtDouble ===================================== compiler/GHC/CmmToAsm/X86/CodeGen.hs ===================================== @@ -913,20 +913,6 @@ getRegister' _ is32Bit (CmmMachOp (MO_UU_Conv W64 W16) [x]) ro <- getNewRegNat II16 return $ Fixed II16 ro (code `appOL` toOL [ MOVZxL II16 (OpReg rlo) (OpReg ro) ]) -getRegister' _ _ (CmmLit lit@(CmmFloat f w)) = - float_const_sse2 where - float_const_sse2 - | f == 0.0 = do - -- TODO: this mishandles negative zero floating point literals. - let - format = floatFormat w - code dst = unitOL (XOR format (OpReg dst) (OpReg dst)) - -- I don't know why there are xorpd, xorps, and pxor instructions. - -- They all appear to do the same thing --SDM - return (Any format code) - - | otherwise = getFloatLitRegister lit - -- catch simple cases of zero- or sign-extended load getRegister' _ _ (CmmMachOp (MO_UU_Conv W8 W32) [CmmLoad addr _ _]) = do code <- intLoadCode (MOVZxL II8) addr @@ -1932,7 +1918,7 @@ getRegister' platform is32Bit load@(CmmLoad mem ty _) | isFloatType ty = do Amode addr mem_code <- getAmode mem - loadFloatAmode width addr mem_code + loadAmode (floatFormat width) addr mem_code | is32Bit && not (isWord64 ty) = do @@ -1960,20 +1946,6 @@ getRegister' platform is32Bit load@(CmmLoad mem ty _) format = cmmTypeFormat ty width = typeWidth ty -getRegister' _ is32Bit (CmmLit (CmmInt 0 width)) - = let - format = intFormat width - - -- x86_64: 32-bit xor is one byte shorter, and zero-extends to 64 bits - format1 = if is32Bit then format - else case format of - II64 -> II32 - _ -> format - code dst - = unitOL (XOR format1 (OpReg dst) (OpReg dst)) - in - return (Any format code) - -- Handle symbol references with LEA and %rip-relative addressing. -- See Note [%rip-relative addressing on x86-64]. getRegister' platform is32Bit (CmmLit lit) @@ -1990,80 +1962,102 @@ getRegister' platform is32Bit (CmmLit lit) is_label (CmmLabelDiffOff {}) = True is_label _ = False - -- optimisation for loading small literals on x86_64: take advantage - -- of the automatic zero-extension from 32 to 64 bits, because the 32-bit - -- instruction forms are shorter. -getRegister' platform is32Bit (CmmLit lit) - | not is32Bit, isWord64 (cmmLitType platform lit), not (isBigLit lit) - = let - imm = litToImm lit - code dst = unitOL (MOV II32 (OpImm imm) (OpReg dst)) - in - return (Any II64 code) - where - isBigLit (CmmInt i _) = i < 0 || i > 0xffffffff - isBigLit _ = False +getRegister' platform is32Bit (CmmLit lit) = do + avx <- avxEnabled + + -- NB: it is important that the code produced here (to load a literal into + -- a register) doesn't clobber any registers other than the destination + -- register; the code for generating C calls relies on this property. + -- + -- In particular, we have: + -- + -- > loadIntoRegMightClobberOtherReg (CmmLit _) = False + -- + -- which means that we assume that loading a literal into a register + -- will not clobber any other registers. + + -- TODO: this function mishandles floating-point negative zero, + -- because -0.0 == 0.0 returns True and because we represent CmmFloat as + -- Rational, which can't properly represent negative zero. + + if + -- Zero: use XOR. + | isZeroLit lit + -> let code dst + | isIntFormat fmt + = let fmt' + | is32Bit + = fmt + | otherwise + -- x86_64: 32-bit xor is one byte shorter, + -- and zero-extends to 64 bits + = case fmt of + II64 -> II32 + _ -> fmt + in unitOL (XOR fmt' (OpReg dst) (OpReg dst)) + | avx + = if float_or_floatvec + then unitOL (VXOR fmt (OpReg dst) dst dst) + else unitOL (VPXOR fmt dst dst dst) + | otherwise + = if float_or_floatvec + then unitOL (XOR fmt (OpReg dst) (OpReg dst)) + else unitOL (PXOR fmt (OpReg dst) dst) + in return $ Any fmt code + + -- Constant vector: use broadcast. + | VecFormat l sFmt <- fmt + , CmmVec (f:fs) <- lit + , all (== f) fs + -> do let w = scalarWidth sFmt + broadcast = if isFloatScalarFormat sFmt + then MO_VF_Broadcast l w + else MO_V_Broadcast l w + valCode <- getAnyReg (CmmMachOp broadcast [CmmLit f]) + return $ Any fmt valCode + + -- Optimisation for loading small literals on x86_64: take advantage + -- of the automatic zero-extension from 32 to 64 bits, because the 32-bit + -- instruction forms are shorter. + | not is32Bit, isWord64 cmmTy, not (isBigLit lit) + -> let + imm = litToImm lit + code dst = unitOL (MOV II32 (OpImm imm) (OpReg dst)) + in + return (Any II64 code) + + -- Scalar integer: use an immediate. + | isIntFormat fmt + -> let imm = litToImm lit + code dst = unitOL (MOV fmt (OpImm imm) (OpReg dst)) + in return (Any fmt code) + + -- General case: load literal from data address. + | otherwise + -> do let w = formatToWidth fmt + Amode addr addr_code <- memConstant (mkAlignment $ widthInBytes w) lit + loadAmode fmt addr addr_code + + where + cmmTy = cmmLitType platform lit + fmt = cmmTypeFormat cmmTy + float_or_floatvec = isFloatOrFloatVecFormat fmt + isZeroLit (CmmInt i _) = i == 0 + isZeroLit (CmmFloat f _) = f == 0 -- TODO: mishandles negative zero + isZeroLit (CmmVec fs) = all isZeroLit fs + isZeroLit _ = False + + isBigLit (CmmInt i _) = i < 0 || i > 0xffffffff + isBigLit _ = False -- note1: not the same as (not.is32BitLit), because that checks for -- signed literals that fit in 32 bits, but we want unsigned -- literals here. -- note2: all labels are small, because we're assuming the -- small memory model. See Note [%rip-relative addressing on x86-64]. -getRegister' platform _ (CmmLit lit) = do - avx <- avxEnabled - case fmt of - VecFormat l sFmt - | CmmVec fs <- lit - , all is_zero fs - -> let code dst - | avx - = if isFloatScalarFormat sFmt - then unitOL (VXOR fmt (OpReg dst) dst dst) - else unitOL (VPXOR fmt dst dst dst) - | otherwise - = unitOL (XOR fmt (OpReg dst) (OpReg dst)) - in return (Any fmt code) - | CmmVec (f:fs) <- lit - , all (== f) fs - -- TODO: mishandles negative zero (because -0.0 == 0.0 returns True), and because we - -- represent CmmFloat as Rational which can't properly represent negative zero. - -> do let w = scalarWidth sFmt - broadcast = if isFloatScalarFormat sFmt - then MO_VF_Broadcast l w - else MO_V_Broadcast l w - valCode <- getAnyReg (CmmMachOp broadcast [CmmLit f]) - return $ Any fmt valCode - - | otherwise - -> do - let w = formatToWidth fmt - config <- getConfig - Amode addr addr_code <- memConstant (mkAlignment $ widthInBytes w) lit - let code dst = addr_code `snocOL` - movInstr config fmt (OpAddr addr) (OpReg dst) - return (Any fmt code) - where - is_zero (CmmInt i _) = i == 0 - is_zero (CmmFloat f _) = f == 0 -- TODO: mishandles negative zero - is_zero _ = False - - _ -> let imm = litToImm lit - code dst = unitOL (MOV fmt (OpImm imm) (OpReg dst)) - in return (Any fmt code) - where - cmmTy = cmmLitType platform lit - fmt = cmmTypeFormat cmmTy - getRegister' platform _ slot@(CmmStackSlot {}) = pprPanic "getRegister(x86) CmmStackSlot" (pdoc platform slot) -getFloatLitRegister :: CmmLit -> NatM Register -getFloatLitRegister lit = do - let w :: Width - w = case lit of { CmmInt _ w -> w; CmmFloat _ w -> w; _ -> panic "getFloatLitRegister" (ppr lit) } - Amode addr code <- memConstant (mkAlignment $ widthInBytes w) lit - loadFloatAmode w addr code - intLoadCode :: (Operand -> Operand -> Instr) -> CmmExpr -> NatM (Reg -> InstrBlock) intLoadCode instr mem = do @@ -2392,15 +2386,12 @@ memConstant align lit = do `consOL` addr_code return (Amode addr code) - -loadFloatAmode :: Width -> AddrMode -> InstrBlock -> NatM Register -loadFloatAmode w addr addr_code = do - let format = floatFormat w - code dst = addr_code `snocOL` - MOV format (OpAddr addr) (OpReg dst) - - return (Any format code) - +-- | Load the value at the given address into any register. +loadAmode :: Format -> AddrMode -> InstrBlock -> NatM Register +loadAmode fmt addr addr_code = do + config <- getConfig + let load dst = movInstr config fmt (OpAddr addr) (OpReg dst) + return $ Any fmt (\ dst -> addr_code `snocOL` load dst) -- if we want a floating-point literal as an operand, we can -- use it directly from memory. However, if the literal is @@ -3100,10 +3091,8 @@ genSimplePrim _ op dst args = do platform <- ncgPlatform <$> getConfig pprPanic "genSimplePrim: unhandled primop" (ppr (pprCallishMachOp op, dst, fmap (pdoc platform) args)) -{- -Note [Evaluate C-call arguments before placing in destination registers] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - +{- Note [Evaluate C-call arguments before placing in destination registers] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ When producing code for C calls we must take care when placing arguments in their final registers. Specifically, we must ensure that temporary register usage due to evaluation of one argument does not clobber a register in which we @@ -3154,15 +3143,11 @@ genForeignCall{32,64}. -- | See Note [Evaluate C-call arguments before placing in destination registers] evalArgs :: BlockId -> [CmmActual] -> NatM (InstrBlock, [CmmActual]) evalArgs bid actuals - | any mightContainMachOp actuals = do + | any loadIntoRegMightClobberOtherReg actuals = do regs_blks <- mapM evalArg actuals return (concatOL $ map fst regs_blks, map snd regs_blks) | otherwise = return (nilOL, actuals) where - mightContainMachOp (CmmReg _) = False - mightContainMachOp (CmmRegOff _ _) = False - mightContainMachOp (CmmLit _) = False - mightContainMachOp _ = True evalArg :: CmmActual -> NatM (InstrBlock, CmmExpr) evalArg actual = do @@ -3176,6 +3161,16 @@ evalArgs bid actuals newLocalReg :: CmmType -> NatM LocalReg newLocalReg ty = LocalReg <$> getUniqueM <*> pure ty +-- | Might the code to put this expression into a register +-- clobber any other registers? +loadIntoRegMightClobberOtherReg :: CmmExpr -> Bool +loadIntoRegMightClobberOtherReg (CmmReg _) = False +loadIntoRegMightClobberOtherReg (CmmRegOff _ _) = False +loadIntoRegMightClobberOtherReg (CmmLit _) = False + -- NB: this last 'False' is slightly risky, because the code for loading + -- a literal into a register is not entirely trivial. +loadIntoRegMightClobberOtherReg _ = True + -- Note [DIV/IDIV for bytes] -- ~~~~~~~~~~~~~~~~~~~~~~~~~ -- IDIV reminder: @@ -3437,7 +3432,6 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do { stackArgs = proper_stack_args , stackDataArgs = stack_data_args , usedRegs = arg_regs_used - , computeArgsCode = compute_args_code , assignArgsCode = assign_args_code } <- loadArgs config prom_args @@ -3536,7 +3530,6 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do return (align_call_code `appOL` push_code `appOL` - compute_args_code `appOL` assign_args_code `appOL` load_data_refs `appOL` shadow_space_code `appOL` @@ -3557,16 +3550,14 @@ data LoadArgs , stackDataArgs :: [CmmExpr] -- | Which registers are we using for argument passing? , usedRegs :: [RegWithFormat] - -- | The code to compute arguments into (possibly temporary) registers. - , computeArgsCode :: InstrBlock -- | The code to assign arguments to registers used for argument passing. , assignArgsCode :: InstrBlock } instance Semigroup LoadArgs where - LoadArgs a1 d1 r1 i1 j1 <> LoadArgs a2 d2 r2 i2 j2 - = LoadArgs (a1 ++ a2) (d1 ++ d2) (r1 ++ r2) (i1 S.<> i2) (j1 S.<> j2) + LoadArgs a1 d1 r1 j1 <> LoadArgs a2 d2 r2 j2 + = LoadArgs (a1 ++ a2) (d1 ++ d2) (r1 ++ r2) (j1 S.<> j2) instance Monoid LoadArgs where - mempty = LoadArgs [] [] [] nilOL nilOL + mempty = LoadArgs [] [] [] nilOL -- | An argument passed on the stack, either directly or by reference. -- @@ -3720,7 +3711,6 @@ loadArgsSysV config (arg:rest) = do LoadArgs { stackArgs = map RawStackArg (arg:rest) , stackDataArgs = [] - , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3740,12 +3730,11 @@ loadArgsSysV config (arg:rest) = do this_arg <- case mbReg of Just reg -> do - (compute_code, assign_code) <- lift $ loadArgIntoReg config arg rest reg + assign_code <- lift $ loadArgIntoReg arg reg return $ LoadArgs { stackArgs = [] -- passed in register , stackDataArgs = [] - , computeArgsCode = compute_code , assignArgsCode = assign_code , usedRegs = [RegWithFormat reg arg_fmt] } @@ -3755,7 +3744,6 @@ loadArgsSysV config (arg:rest) = do LoadArgs { stackArgs = [RawStackArg arg] , stackDataArgs = [] - , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3807,7 +3795,6 @@ loadArgsWin config (arg:rest) = do LoadArgs { stackArgs = stk_args , stackDataArgs = data_args - , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3823,8 +3810,7 @@ loadArgsWin config (arg:rest) = do -- Pass the reference in a register, -- and the argument data on the stack. { stackArgs = [RawStackArgRef (InReg ireg) (argSize platform arg)] - , stackDataArgs = [arg] - , computeArgsCode = nilOL -- we don't yet know where the data will reside, + , stackDataArgs = [arg] -- we don't yet know where the data will reside, , assignArgsCode = nilOL -- so we defer computing the reference and storing it -- in the register until later , usedRegs = [RegWithFormat ireg II64] @@ -3835,7 +3821,7 @@ loadArgsWin config (arg:rest) = do = freg | otherwise = ireg - (compute_code, assign_code) <- loadArgIntoReg config arg rest arg_reg + assign_code <- loadArgIntoReg arg arg_reg -- Recall that, for varargs, we must pass floating-point -- arguments in both fp and integer registers. let (assign_code', regs') @@ -3848,42 +3834,23 @@ loadArgsWin config (arg:rest) = do LoadArgs { stackArgs = [] -- passed in register , stackDataArgs = [] - , computeArgsCode = compute_code , assignArgsCode = assign_code' , usedRegs = regs' } - --- | Return two pieces of code: --- --- - code to compute a the given 'CmmExpr' into some (possibly temporary) register --- - code to assign the resulting value to the specified register +-- | Load an argument into a register. -- --- Using two separate pieces of code handles clobbering issues reported --- in e.g. #11792, #12614. -loadArgIntoReg :: NCGConfig -> CmmExpr -> [CmmExpr] -> Reg -> NatM (InstrBlock, InstrBlock) -loadArgIntoReg config arg rest reg - -- "operand" args can be directly assigned into the register - | isOperand platform arg - = do arg_code <- getAnyReg arg - return (nilOL, arg_code reg) - -- The last non-operand arg can be directly assigned after its - -- computation without going into a temporary register - | all (isOperand platform) rest - = do arg_code <- getAnyReg arg - return (arg_code reg, nilOL) - -- Other args need to be computed beforehand to avoid clobbering - -- previously assigned registers used to pass parameters (see - -- #11792, #12614). They are assigned into temporary registers - -- and get assigned to proper call ABI registers after they all - -- have been computed. - | otherwise - = do arg_code <- getAnyReg arg - tmp <- getNewRegNat arg_fmt - return (arg_code tmp, unitOL $ mkRegRegMoveInstr config arg_fmt tmp reg) - where - platform = ncgPlatform config - arg_fmt = cmmTypeFormat $ cmmExprType platform arg +-- Assumes that the expression does not contain any MachOps, +-- as per Note [Evaluate C-call arguments before placing in destination registers]. +loadArgIntoReg :: CmmExpr -> Reg -> NatM InstrBlock +loadArgIntoReg arg reg = do + when (debugIsOn && loadIntoRegMightClobberOtherReg arg) $ do + platform <- getPlatform + massertPpr False $ + vcat [ text "loadArgIntoReg: arg might contain MachOp" + , text "arg:" <+> pdoc platform arg ] + arg_code <- getAnyReg arg + return $ arg_code reg -- ----------------------------------------------------------------------------- -- Pushing arguments onto the stack for 64-bit C calls. ===================================== compiler/GHC/CmmToAsm/X86/Instr.hs ===================================== @@ -306,6 +306,7 @@ data Instr | VMOVDQU Format Operand Operand -- logic operations + | PXOR Format Operand Reg | VPXOR Format Reg Reg Reg -- Arithmetic @@ -492,6 +493,12 @@ regUsageOfInstr platform instr MOVDQU fmt src dst -> mkRU (use_R fmt src []) (use_R fmt dst []) VMOVDQU fmt src dst -> mkRU (use_R fmt src []) (use_R fmt dst []) + PXOR fmt (OpReg src) dst + | src == dst + -> mkRU [] [mk fmt dst] + | otherwise + -> mkRU [mk fmt src, mk fmt dst] [mk fmt dst] + VPXOR fmt s1 s2 dst | s1 == s2, s1 == dst -> mkRU [] [mk fmt dst] @@ -733,6 +740,7 @@ patchRegsOfInstr platform instr env MOVDQU fmt src dst -> MOVDQU fmt (patchOp src) (patchOp dst) VMOVDQU fmt src dst -> VMOVDQU fmt (patchOp src) (patchOp dst) + PXOR fmt src dst -> PXOR fmt (patchOp src) (env dst) VPXOR fmt s1 s2 dst -> VPXOR fmt (env s1) (env s2) (env dst) VADD fmt s1 s2 dst -> VADD fmt (patchOp s1) (env s2) (env dst) ===================================== compiler/GHC/CmmToAsm/X86/Ppr.hs ===================================== @@ -1016,6 +1016,8 @@ pprInstr platform i = case i of VecFormat 64 FmtInt8 -> text "vmovdqu32" -- require the additional AVX512BW extension _ -> text "vmovdqu" + PXOR format src dst + -> pprPXor (text "pxor") format src dst VPXOR format s1 s2 dst -> pprXor (text "vpxor") format s1 s2 dst VEXTRACT format offset from to @@ -1320,6 +1322,15 @@ pprInstr platform i = case i of pprReg platform format reg3 ] + pprPXor :: Line doc -> Format -> Operand -> Reg -> doc + pprPXor name format src dst + = line $ hcat [ + pprGenMnemonic name format, + pprOperand platform format src, + comma, + pprReg platform format dst + ] + pprVxor :: Format -> Operand -> Reg -> Reg -> doc pprVxor fmt src1 src2 dst = line $ hcat [ ===================================== compiler/GHC/CmmToLlvm.hs ===================================== @@ -139,7 +139,7 @@ llvmGroupLlvmGens cmm = do Nothing -> l Just (CmmStaticsRaw info_lbl _) -> info_lbl lml <- strCLabel_llvm l' - funInsert lml =<< llvmFunTy (map globalRegUse_reg live) + funInsert lml =<< llvmFunTy live return Nothing cdata <- fmap catMaybes $ mapM split cmm ===================================== compiler/GHC/CmmToLlvm/Base.hs ===================================== @@ -12,7 +12,7 @@ module GHC.CmmToLlvm.Base ( LlvmCmmDecl, LlvmBasicBlock, - LiveGlobalRegs, + LiveGlobalRegs, LiveGlobalRegUses, LlvmUnresData, LlvmData, UnresLabel, UnresStatic, LlvmM, @@ -29,6 +29,8 @@ module GHC.CmmToLlvm.Base ( llvmFunSig, llvmFunArgs, llvmStdFunAttrs, llvmFunAlign, llvmInfAlign, llvmPtrBits, tysToParams, llvmFunSection, padLiveArgs, isFPR, + lookupRegUse, + strCLabel_llvm, getGlobalPtr, generateExternDecls, @@ -58,9 +60,11 @@ import GHC.Types.Unique.Set import GHC.Types.Unique.Supply import GHC.Utils.Logger -import Data.Maybe (fromJust) import Control.Monad.Trans.State (StateT (..)) -import Data.List (isPrefixOf) +import Control.Applicative (Alternative((<|>))) +import Data.Maybe (fromJust, mapMaybe) + +import Data.List (find, isPrefixOf) import qualified Data.List.NonEmpty as NE import Data.Ord (comparing) @@ -73,6 +77,7 @@ type LlvmBasicBlock = GenBasicBlock LlvmStatement -- | Global registers live on proc entry type LiveGlobalRegs = [GlobalReg] +type LiveGlobalRegUses = [GlobalRegUse] -- | Unresolved code. -- Of the form: (data label, data type, unresolved data) @@ -116,16 +121,16 @@ llvmGhcCC platform | otherwise = CC_Ghc -- | Llvm Function type for Cmm function -llvmFunTy :: LiveGlobalRegs -> LlvmM LlvmType +llvmFunTy :: LiveGlobalRegUses -> LlvmM LlvmType llvmFunTy live = return . LMFunction =<< llvmFunSig' live (fsLit "a") ExternallyVisible -- | Llvm Function signature -llvmFunSig :: LiveGlobalRegs -> CLabel -> LlvmLinkageType -> LlvmM LlvmFunctionDecl +llvmFunSig :: LiveGlobalRegUses -> CLabel -> LlvmLinkageType -> LlvmM LlvmFunctionDecl llvmFunSig live lbl link = do lbl' <- strCLabel_llvm lbl llvmFunSig' live lbl' link -llvmFunSig' :: LiveGlobalRegs -> LMString -> LlvmLinkageType -> LlvmM LlvmFunctionDecl +llvmFunSig' :: LiveGlobalRegUses -> LMString -> LlvmLinkageType -> LlvmM LlvmFunctionDecl llvmFunSig' live lbl link = do let toParams x | isPointer x = (x, [NoAlias, NoCapture]) | otherwise = (x, []) @@ -149,16 +154,25 @@ llvmFunSection opts lbl | otherwise = Nothing -- | A Function's arguments -llvmFunArgs :: Platform -> LiveGlobalRegs -> [LlvmVar] +llvmFunArgs :: Platform -> LiveGlobalRegUses -> [LlvmVar] llvmFunArgs platform live = - map (lmGlobalRegArg platform) (filter isPassed allRegs) + map (lmGlobalRegArg platform) (mapMaybe isPassed allRegs) where allRegs = activeStgRegs platform paddingRegs = padLiveArgs platform live - isLive r = r `elem` alwaysLive - || r `elem` live - || r `elem` paddingRegs - isPassed r = not (isFPR r) || isLive r - + isLive :: GlobalReg -> Maybe GlobalRegUse + isLive r = + lookupRegUse r (alwaysLive platform) + <|> + lookupRegUse r live + <|> + lookupRegUse r paddingRegs + isPassed r = + if not (isFPR r) + then Just $ GlobalRegUse r (globalRegSpillType platform r) + else isLive r + +lookupRegUse :: GlobalReg -> [GlobalRegUse] -> Maybe GlobalRegUse +lookupRegUse r = find ((== r) . globalRegUse_reg) isFPR :: GlobalReg -> Bool isFPR (FloatReg _) = True @@ -179,7 +193,7 @@ isFPR _ = False -- Invariant: Cmm FPR regs with number "n" maps to real registers with number -- "n" If the calling convention uses registers in a different order or if the -- invariant doesn't hold, this code probably won't be correct. -padLiveArgs :: Platform -> LiveGlobalRegs -> LiveGlobalRegs +padLiveArgs :: Platform -> LiveGlobalRegUses -> LiveGlobalRegUses padLiveArgs platform live = if platformUnregisterised platform then [] -- not using GHC's register convention for platform. @@ -188,7 +202,7 @@ padLiveArgs platform live = ---------------------------------- -- handle floating-point registers (FPR) - fprLive = filter isFPR live -- real live FPR registers + fprLive = filter (isFPR . globalRegUse_reg) live -- real live FPR registers -- we group live registers sharing the same classes, i.e. that use the same -- set of real registers to be passed. E.g. FloatReg, DoubleReg and XmmReg @@ -196,39 +210,44 @@ padLiveArgs platform live = -- classes = NE.groupBy sharesClass fprLive sharesClass a b = globalRegsOverlap platform (norm a) (norm b) -- check if mapped to overlapping registers - norm x = fpr_ctor x 1 -- get the first register of the family + norm x = globalRegUse_reg (fpr_ctor x 1) -- get the first register of the family -- For each class, we just have to fill missing registers numbers. We use -- the constructor of the greatest register to build padding registers. -- -- E.g. sortedRs = [ F2, XMM4, D5] -- output = [D1, D3] + padded :: [GlobalRegUse] padded = concatMap padClass classes + + padClass :: NE.NonEmpty GlobalRegUse -> [GlobalRegUse] padClass rs = go (NE.toList sortedRs) 1 where - sortedRs = NE.sortBy (comparing fpr_num) rs + sortedRs = NE.sortBy (comparing (fpr_num . globalRegUse_reg)) rs maxr = NE.last sortedRs ctor = fpr_ctor maxr go [] _ = [] - go (c1:c2:_) _ -- detect bogus case (see #17920) + go (GlobalRegUse c1 _: GlobalRegUse c2 _:_) _ -- detect bogus case (see #17920) | fpr_num c1 == fpr_num c2 , Just real <- globalRegMaybe platform c1 = sorryDoc "LLVM code generator" $ text "Found two different Cmm registers (" <> ppr c1 <> text "," <> ppr c2 <> text ") both alive AND mapped to the same real register: " <> ppr real <> text ". This isn't currently supported by the LLVM backend." - go (c:cs) f - | fpr_num c == f = go cs f -- already covered by a real register - | otherwise = ctor f : go (c:cs) (f + 1) -- add padding register - - fpr_ctor :: GlobalReg -> Int -> GlobalReg - fpr_ctor (FloatReg _) = FloatReg - fpr_ctor (DoubleReg _) = DoubleReg - fpr_ctor (XmmReg _) = XmmReg - fpr_ctor (YmmReg _) = YmmReg - fpr_ctor (ZmmReg _) = ZmmReg - fpr_ctor _ = error "fpr_ctor expected only FPR regs" + go (cu@(GlobalRegUse c _):cs) f + | fpr_num c == f = go cs f -- already covered by a real register + | otherwise = ctor f : go (cu:cs) (f + 1) -- add padding register + + fpr_ctor :: GlobalRegUse -> Int -> GlobalRegUse + fpr_ctor (GlobalRegUse r fmt) i = + case r of + FloatReg _ -> GlobalRegUse (FloatReg i) fmt + DoubleReg _ -> GlobalRegUse (DoubleReg i) fmt + XmmReg _ -> GlobalRegUse (XmmReg i) fmt + YmmReg _ -> GlobalRegUse (YmmReg i) fmt + ZmmReg _ -> GlobalRegUse (ZmmReg i) fmt + _ -> error "fpr_ctor expected only FPR regs" fpr_num :: GlobalReg -> Int fpr_num (FloatReg i) = i ===================================== compiler/GHC/CmmToLlvm/CodeGen.hs ===================================== @@ -37,13 +37,14 @@ import GHC.Utils.Outputable import qualified GHC.Utils.Panic as Panic import GHC.Utils.Misc +import Control.Applicative (Alternative((<|>))) import Control.Monad.Trans.Class import Control.Monad.Trans.Writer import Control.Monad import qualified Data.Semigroup as Semigroup import Data.List ( nub ) -import Data.Maybe ( catMaybes ) +import Data.Maybe ( catMaybes, isJust ) type Atomic = Maybe MemoryOrdering type LlvmStatements = OrdList LlvmStatement @@ -57,7 +58,7 @@ genLlvmProc :: RawCmmDecl -> LlvmM [LlvmCmmDecl] genLlvmProc (CmmProc infos lbl live graph) = do let blocks = toBlockListEntryFirstFalseFallthrough graph - (lmblocks, lmdata) <- basicBlocksCodeGen (map globalRegUse_reg live) blocks + (lmblocks, lmdata) <- basicBlocksCodeGen live blocks let info = mapLookup (g_entry graph) infos proc = CmmProc info lbl live (ListGraph lmblocks) return (proc:lmdata) @@ -76,7 +77,7 @@ newtype UnreachableBlockId = UnreachableBlockId BlockId -- | Generate code for a list of blocks that make up a complete -- procedure. The first block in the list is expected to be the entry -- point. -basicBlocksCodeGen :: LiveGlobalRegs -> [CmmBlock] +basicBlocksCodeGen :: LiveGlobalRegUses -> [CmmBlock] -> LlvmM ([LlvmBasicBlock], [LlvmCmmDecl]) basicBlocksCodeGen _ [] = panic "no entry block!" basicBlocksCodeGen live cmmBlocks @@ -152,7 +153,7 @@ stmtToInstrs ubid stmt = case stmt of -- Tail call CmmCall { cml_target = arg, - cml_args_regs = live } -> genJump arg $ map globalRegUse_reg live + cml_args_regs = live } -> genJump arg live _ -> panic "Llvm.CodeGen.stmtToInstrs" @@ -1050,7 +1051,7 @@ cmmPrimOpFunctions mop = do -- | Tail function calls -genJump :: CmmExpr -> [GlobalReg] -> LlvmM StmtData +genJump :: CmmExpr -> LiveGlobalRegUses -> LlvmM StmtData -- Call to known function genJump (CmmLit (CmmLabel lbl)) live = do @@ -2056,14 +2057,13 @@ getCmmReg (CmmLocal (LocalReg un _)) -- have been assigned a value at some point, triggering -- "funPrologue" to allocate it on the stack. -getCmmReg (CmmGlobal g) - = do let r = globalRegUse_reg g - onStack <- checkStackReg r +getCmmReg (CmmGlobal ru@(GlobalRegUse r _)) + = do onStack <- checkStackReg r platform <- getPlatform if onStack - then return (lmGlobalRegVar platform r) + then return (lmGlobalRegVar platform ru) else pprPanic "getCmmReg: Cmm register " $ - ppr g <> text " not stack-allocated!" + ppr r <> text " not stack-allocated!" -- | Return the value of a given register, as well as its type. Might -- need to be load from stack. @@ -2074,7 +2074,7 @@ getCmmRegVal reg = onStack <- checkStackReg (globalRegUse_reg g) platform <- getPlatform if onStack then loadFromStack else do - let r = lmGlobalRegArg platform (globalRegUse_reg g) + let r = lmGlobalRegArg platform g return (r, getVarType r, nilOL) _ -> loadFromStack where loadFromStack = do @@ -2187,8 +2187,9 @@ convertMemoryOrdering MemOrderSeqCst = SyncSeqCst -- question is never written. Therefore we skip it where we can to -- save a few lines in the output and hopefully speed compilation up a -- bit. -funPrologue :: LiveGlobalRegs -> [CmmBlock] -> LlvmM StmtData +funPrologue :: LiveGlobalRegUses -> [CmmBlock] -> LlvmM StmtData funPrologue live cmmBlocks = do + platform <- getPlatform let getAssignedRegs :: CmmNode O O -> [CmmReg] getAssignedRegs (CmmAssign reg _) = [reg] @@ -2196,7 +2197,8 @@ funPrologue live cmmBlocks = do getAssignedRegs _ = [] getRegsBlock (_, body, _) = concatMap getAssignedRegs $ blockToList body assignedRegs = nub $ concatMap (getRegsBlock . blockSplit) cmmBlocks - isLive r = r `elem` alwaysLive || r `elem` live + mbLive r = + lookupRegUse r (alwaysLive platform) <|> lookupRegUse r live platform <- getPlatform stmtss <- forM assignedRegs $ \reg -> @@ -2205,12 +2207,12 @@ funPrologue live cmmBlocks = do let (newv, stmts) = allocReg reg varInsert un (pLower $ getVarType newv) return stmts - CmmGlobal (GlobalRegUse r _) -> do - let reg = lmGlobalRegVar platform r - arg = lmGlobalRegArg platform r + CmmGlobal ru@(GlobalRegUse r _) -> do + let reg = lmGlobalRegVar platform ru + arg = lmGlobalRegArg platform ru ty = (pLower . getVarType) reg trash = LMLitVar $ LMUndefLit ty - rval = if isLive r then arg else trash + rval = if isJust (mbLive r) then arg else trash alloc = Assignment reg $ Alloca (pLower $ getVarType reg) 1 markStackReg r return $ toOL [alloc, Store rval reg Nothing []] @@ -2222,7 +2224,7 @@ funPrologue live cmmBlocks = do -- | Function epilogue. Load STG variables to use as argument for call. -- STG Liveness optimisation done here. -funEpilogue :: LiveGlobalRegs -> LlvmM ([LlvmVar], LlvmStatements) +funEpilogue :: LiveGlobalRegUses -> LlvmM ([LlvmVar], LlvmStatements) funEpilogue live = do platform <- getPlatform @@ -2248,12 +2250,16 @@ funEpilogue live = do let allRegs = activeStgRegs platform loads <- forM allRegs $ \r -> if -- load live registers - | r `elem` alwaysLive -> loadExpr (GlobalRegUse r (globalRegSpillType platform r)) - | r `elem` live -> loadExpr (GlobalRegUse r (globalRegSpillType platform r)) + | Just ru <- lookupRegUse r (alwaysLive platform) + -> loadExpr ru + | Just ru <- lookupRegUse r live + -> loadExpr ru -- load all non Floating-Point Registers - | not (isFPR r) -> loadUndef r + | not (isFPR r) + -> loadUndef (GlobalRegUse r (globalRegSpillType platform r)) -- load padding Floating-Point Registers - | r `elem` paddingRegs -> loadUndef r + | Just ru <- lookupRegUse r paddingRegs + -> loadUndef ru | otherwise -> return (Nothing, nilOL) let (vars, stmts) = unzip loads @@ -2263,7 +2269,7 @@ funEpilogue live = do -- -- This is for Haskell functions, function type is assumed, so doesn't work -- with foreign functions. -getHsFunc :: LiveGlobalRegs -> CLabel -> LlvmM ExprData +getHsFunc :: LiveGlobalRegUses -> CLabel -> LlvmM ExprData getHsFunc live lbl = do fty <- llvmFunTy live name <- strCLabel_llvm lbl ===================================== compiler/GHC/CmmToLlvm/Ppr.hs ===================================== @@ -49,9 +49,8 @@ pprLlvmCmmDecl (CmmData _ lmdata) = do return ( vcat $ map (pprLlvmData opts) lmdata , vcat $ map (pprLlvmData opts) lmdata) -pprLlvmCmmDecl (CmmProc mb_info entry_lbl liveWithUses (ListGraph blks)) - = do let live = map globalRegUse_reg liveWithUses - lbl = case mb_info of +pprLlvmCmmDecl (CmmProc mb_info entry_lbl live (ListGraph blks)) + = do let lbl = case mb_info of Nothing -> entry_lbl Just (CmmStaticsRaw info_lbl _) -> info_lbl link = if externallyVisibleCLabel lbl ===================================== compiler/GHC/CmmToLlvm/Regs.hs ===================================== @@ -14,25 +14,27 @@ import GHC.Prelude import GHC.Llvm import GHC.Cmm.Expr +import GHC.CmmToAsm.Format import GHC.Platform import GHC.Data.FastString import GHC.Utils.Panic ( panic ) import GHC.Types.Unique + -- | Get the LlvmVar function variable storing the real register -lmGlobalRegVar :: Platform -> GlobalReg -> LlvmVar +lmGlobalRegVar :: Platform -> GlobalRegUse -> LlvmVar lmGlobalRegVar platform = pVarLift . lmGlobalReg platform "_Var" -- | Get the LlvmVar function argument storing the real register -lmGlobalRegArg :: Platform -> GlobalReg -> LlvmVar +lmGlobalRegArg :: Platform -> GlobalRegUse -> LlvmVar lmGlobalRegArg platform = lmGlobalReg platform "_Arg" {- Need to make sure the names here can't conflict with the unique generated names. Uniques generated names containing only base62 chars. So using say the '_' char guarantees this. -} -lmGlobalReg :: Platform -> String -> GlobalReg -> LlvmVar -lmGlobalReg platform suf reg +lmGlobalReg :: Platform -> String -> GlobalRegUse -> LlvmVar +lmGlobalReg platform suf (GlobalRegUse reg ty) = case reg of BaseReg -> ptrGlobal $ "Base" ++ suf Sp -> ptrGlobal $ "Sp" ++ suf @@ -88,13 +90,26 @@ lmGlobalReg platform suf reg ptrGlobal name = LMNLocalVar (fsLit name) (llvmWordPtr platform) floatGlobal name = LMNLocalVar (fsLit name) LMFloat doubleGlobal name = LMNLocalVar (fsLit name) LMDouble - xmmGlobal name = LMNLocalVar (fsLit name) (LMVector 4 (LMInt 32)) - ymmGlobal name = LMNLocalVar (fsLit name) (LMVector 8 (LMInt 32)) - zmmGlobal name = LMNLocalVar (fsLit name) (LMVector 16 (LMInt 32)) + fmt = cmmTypeFormat ty + xmmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + ymmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + zmmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + +formatLlvmType :: Format -> LlvmType +formatLlvmType II8 = LMInt 8 +formatLlvmType II16 = LMInt 16 +formatLlvmType II32 = LMInt 32 +formatLlvmType II64 = LMInt 64 +formatLlvmType FF32 = LMFloat +formatLlvmType FF64 = LMDouble +formatLlvmType (VecFormat l sFmt) = LMVector l (formatLlvmType $ scalarFormatFormat sFmt) -- | A list of STG Registers that should always be considered alive -alwaysLive :: [GlobalReg] -alwaysLive = [BaseReg, Sp, Hp, SpLim, HpLim, node] +alwaysLive :: Platform -> [GlobalRegUse] +alwaysLive platform = + [ GlobalRegUse r (globalRegSpillType platform r) + | r <- [BaseReg, Sp, Hp, SpLim, HpLim, node] + ] -- | STG Type Based Alias Analysis hierarchy stgTBAA :: [(Unique, LMString, Maybe Unique)] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a34a4dc2269300b31c06173dad5d64f55317ecb4...c27747f25c7209555ae96573f4525839297b7703 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a34a4dc2269300b31c06173dad5d64f55317ecb4...c27747f25c7209555ae96573f4525839297b7703 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 15:20:48 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Fri, 20 Sep 2024 11:20:48 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 4 commits: Fix C calls with SIMD vectors Message-ID: <66ed92d075c0a_36d9d213019741498a3@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: be70f8c0 by sheaf at 2024-09-20T17:20:18+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 970a1b81 by sheaf at 2024-09-20T17:20:18+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - d61d6f99 by sheaf at 2024-09-20T17:20:18+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - ad8c5c64 by sheaf at 2024-09-20T17:20:19+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 12 changed files: - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToAsm/X86/Instr.hs - compiler/GHC/CmmToAsm/X86/Ppr.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs - testsuite/tests/simd/should_run/simd013.hs - testsuite/tests/simd/should_run/simd013.stdout - testsuite/tests/simd/should_run/simd013C.c Changes: ===================================== compiler/GHC/CmmToAsm/Format.hs ===================================== @@ -28,6 +28,7 @@ module GHC.CmmToAsm.Format ( scalarWidth, formatInBytes, isFloatScalarFormat, + isFloatOrFloatVecFormat, floatScalarFormat, scalarFormatFormat, VirtualRegWithFormat(..), @@ -134,6 +135,11 @@ isFloatScalarFormat = \case FmtDouble -> True _ -> False +isFloatOrFloatVecFormat :: Format -> Bool +isFloatOrFloatVecFormat = \case + VecFormat _ sFmt -> isFloatScalarFormat sFmt + fmt -> isFloatFormat fmt + floatScalarFormat :: Width -> ScalarFormat floatScalarFormat W32 = FmtFloat floatScalarFormat W64 = FmtDouble ===================================== compiler/GHC/CmmToAsm/X86/CodeGen.hs ===================================== @@ -1,6 +1,9 @@ +{-# LANGUAGE DuplicateRecordFields #-} +{-# LANGUAGE GADTs #-} {-# LANGUAGE LambdaCase #-} {-# LANGUAGE MultiWayIf #-} -{-# LANGUAGE GADTs #-} +{-# LANGUAGE ParallelListComp #-} +{-# LANGUAGE TupleSections #-} {-# LANGUAGE NondecreasingIndentation #-} {-# OPTIONS_GHC -Wno-incomplete-uni-patterns #-} @@ -73,22 +76,29 @@ import GHC.Types.Tickish ( GenTickish(..) ) import GHC.Types.SrcLoc ( srcSpanFile, srcSpanStartLine, srcSpanStartCol ) -- The rest: +import GHC.Data.Maybe ( expectJust ) import GHC.Types.ForeignCall ( CCallConv(..) ) import GHC.Data.OrdList import GHC.Utils.Outputable import GHC.Utils.Constants (debugIsOn) +import GHC.Utils.Monad ( foldMapM ) import GHC.Utils.Panic import GHC.Data.FastString import GHC.Utils.Misc import GHC.Types.Unique.Supply ( getUniqueM ) +import qualified Data.Semigroup as S + import Control.Monad +import Control.Monad.Trans.State.Strict + ( StateT, evalStateT, get, put ) +import Control.Monad.Trans.Class (lift) import Data.Foldable (fold) import Data.Int import Data.Maybe import Data.Word -import qualified Data.Map as M +import qualified Data.Map as Map is32BitPlatform :: NatM Bool is32BitPlatform = do @@ -227,7 +237,7 @@ addSpUnwindings instr@(DELTA d) = do let platform = ncgPlatform config if ncgDwarfUnwindings config then do lbl <- mkAsmTempLabel <$> getUniqueM - let unwind = M.singleton MachSp (Just $ UwReg (GlobalRegUse MachSp (bWord platform)) $ negate d) + let unwind = Map.singleton MachSp (Just $ UwReg (GlobalRegUse MachSp (bWord platform)) $ negate d) return $ toOL [ instr, UNWIND lbl unwind ] else return (unitOL instr) addSpUnwindings instr = return $ unitOL instr @@ -328,10 +338,10 @@ stmtToInstrs bid stmt = do CmmUnwind regs -> do let to_unwind_entry :: (GlobalReg, Maybe CmmExpr) -> UnwindTable - to_unwind_entry (reg, expr) = M.singleton reg (fmap (toUnwindExpr platform) expr) + to_unwind_entry (reg, expr) = Map.singleton reg (fmap (toUnwindExpr platform) expr) case foldMap to_unwind_entry regs of - tbl | M.null tbl -> return nilOL - | otherwise -> do + tbl | Map.null tbl -> return nilOL + | otherwise -> do lbl <- mkAsmTempLabel <$> getUniqueM return $ unitOL $ UNWIND lbl tbl @@ -836,12 +846,12 @@ iselExpr64ParallelBin op e1 e2 = do -- targetted for any particular type like Int8, Int32 etc data VectorArithInstns = VA_Add | VA_Sub | VA_Mul | VA_Div | VA_Min | VA_Max -getRegister :: CmmExpr -> NatM Register +getRegister :: HasDebugCallStack => CmmExpr -> NatM Register getRegister e = do platform <- getPlatform is32Bit <- is32BitPlatform getRegister' platform is32Bit e -getRegister' :: Platform -> Bool -> CmmExpr -> NatM Register +getRegister' :: HasDebugCallStack => Platform -> Bool -> CmmExpr -> NatM Register getRegister' platform is32Bit (CmmReg reg) = case reg of @@ -903,20 +913,6 @@ getRegister' _ is32Bit (CmmMachOp (MO_UU_Conv W64 W16) [x]) ro <- getNewRegNat II16 return $ Fixed II16 ro (code `appOL` toOL [ MOVZxL II16 (OpReg rlo) (OpReg ro) ]) -getRegister' _ _ (CmmLit lit@(CmmFloat f w)) = - float_const_sse2 where - float_const_sse2 - | f == 0.0 = do - -- TODO: this mishandles negative zero floating point literals. - let - format = floatFormat w - code dst = unitOL (XOR format (OpReg dst) (OpReg dst)) - -- I don't know why there are xorpd, xorps, and pxor instructions. - -- They all appear to do the same thing --SDM - return (Any format code) - - | otherwise = getFloatLitRegister lit - -- catch simple cases of zero- or sign-extended load getRegister' _ _ (CmmMachOp (MO_UU_Conv W8 W32) [CmmLoad addr _ _]) = do code <- intLoadCode (MOVZxL II8) addr @@ -1922,7 +1918,7 @@ getRegister' platform is32Bit load@(CmmLoad mem ty _) | isFloatType ty = do Amode addr mem_code <- getAmode mem - loadFloatAmode width addr mem_code + loadAmode (floatFormat width) addr mem_code | is32Bit && not (isWord64 ty) = do @@ -1950,20 +1946,6 @@ getRegister' platform is32Bit load@(CmmLoad mem ty _) format = cmmTypeFormat ty width = typeWidth ty -getRegister' _ is32Bit (CmmLit (CmmInt 0 width)) - = let - format = intFormat width - - -- x86_64: 32-bit xor is one byte shorter, and zero-extends to 64 bits - format1 = if is32Bit then format - else case format of - II64 -> II32 - _ -> format - code dst - = unitOL (XOR format1 (OpReg dst) (OpReg dst)) - in - return (Any format code) - -- Handle symbol references with LEA and %rip-relative addressing. -- See Note [%rip-relative addressing on x86-64]. getRegister' platform is32Bit (CmmLit lit) @@ -1980,80 +1962,102 @@ getRegister' platform is32Bit (CmmLit lit) is_label (CmmLabelDiffOff {}) = True is_label _ = False - -- optimisation for loading small literals on x86_64: take advantage - -- of the automatic zero-extension from 32 to 64 bits, because the 32-bit - -- instruction forms are shorter. -getRegister' platform is32Bit (CmmLit lit) - | not is32Bit, isWord64 (cmmLitType platform lit), not (isBigLit lit) - = let - imm = litToImm lit - code dst = unitOL (MOV II32 (OpImm imm) (OpReg dst)) - in - return (Any II64 code) - where - isBigLit (CmmInt i _) = i < 0 || i > 0xffffffff - isBigLit _ = False +getRegister' platform is32Bit (CmmLit lit) = do + avx <- avxEnabled + + -- NB: it is important that the code produced here (to load a literal into + -- a register) doesn't clobber any registers other than the destination + -- register; the code for generating C calls relies on this property. + -- + -- In particular, we have: + -- + -- > loadIntoRegMightClobberOtherReg (CmmLit _) = False + -- + -- which means that we assume that loading a literal into a register + -- will not clobber any other registers. + + -- TODO: this function mishandles floating-point negative zero, + -- because -0.0 == 0.0 returns True and because we represent CmmFloat as + -- Rational, which can't properly represent negative zero. + + if + -- Zero: use XOR. + | isZeroLit lit + -> let code dst + | isIntFormat fmt + = let fmt' + | is32Bit + = fmt + | otherwise + -- x86_64: 32-bit xor is one byte shorter, + -- and zero-extends to 64 bits + = case fmt of + II64 -> II32 + _ -> fmt + in unitOL (XOR fmt' (OpReg dst) (OpReg dst)) + | avx + = if float_or_floatvec + then unitOL (VXOR fmt (OpReg dst) dst dst) + else unitOL (VPXOR fmt dst dst dst) + | otherwise + = if float_or_floatvec + then unitOL (XOR fmt (OpReg dst) (OpReg dst)) + else unitOL (PXOR fmt (OpReg dst) dst) + in return $ Any fmt code + + -- Constant vector: use broadcast. + | VecFormat l sFmt <- fmt + , CmmVec (f:fs) <- lit + , all (== f) fs + -> do let w = scalarWidth sFmt + broadcast = if isFloatScalarFormat sFmt + then MO_VF_Broadcast l w + else MO_V_Broadcast l w + valCode <- getAnyReg (CmmMachOp broadcast [CmmLit f]) + return $ Any fmt valCode + + -- Optimisation for loading small literals on x86_64: take advantage + -- of the automatic zero-extension from 32 to 64 bits, because the 32-bit + -- instruction forms are shorter. + | not is32Bit, isWord64 cmmTy, not (isBigLit lit) + -> let + imm = litToImm lit + code dst = unitOL (MOV II32 (OpImm imm) (OpReg dst)) + in + return (Any II64 code) + + -- Scalar integer: use an immediate. + | isIntFormat fmt + -> let imm = litToImm lit + code dst = unitOL (MOV fmt (OpImm imm) (OpReg dst)) + in return (Any fmt code) + + -- General case: load literal from data address. + | otherwise + -> do let w = formatToWidth fmt + Amode addr addr_code <- memConstant (mkAlignment $ widthInBytes w) lit + loadAmode fmt addr addr_code + + where + cmmTy = cmmLitType platform lit + fmt = cmmTypeFormat cmmTy + float_or_floatvec = isFloatOrFloatVecFormat fmt + isZeroLit (CmmInt i _) = i == 0 + isZeroLit (CmmFloat f _) = f == 0 -- TODO: mishandles negative zero + isZeroLit (CmmVec fs) = all isZeroLit fs + isZeroLit _ = False + + isBigLit (CmmInt i _) = i < 0 || i > 0xffffffff + isBigLit _ = False -- note1: not the same as (not.is32BitLit), because that checks for -- signed literals that fit in 32 bits, but we want unsigned -- literals here. -- note2: all labels are small, because we're assuming the -- small memory model. See Note [%rip-relative addressing on x86-64]. -getRegister' platform _ (CmmLit lit) = do - avx <- avxEnabled - case fmt of - VecFormat l sFmt - | CmmVec fs <- lit - , all is_zero fs - -> let code dst - | avx - = if isFloatScalarFormat sFmt - then unitOL (VXOR fmt (OpReg dst) dst dst) - else unitOL (VPXOR fmt dst dst dst) - | otherwise - = unitOL (XOR fmt (OpReg dst) (OpReg dst)) - in return (Any fmt code) - | CmmVec (f:fs) <- lit - , all (== f) fs - -- TODO: mishandles negative zero (because -0.0 == 0.0 returns True), and because we - -- represent CmmFloat as Rational which can't properly represent negative zero. - -> do let w = scalarWidth sFmt - broadcast = if isFloatScalarFormat sFmt - then MO_VF_Broadcast l w - else MO_V_Broadcast l w - valCode <- getAnyReg (CmmMachOp broadcast [CmmLit f]) - return $ Any fmt valCode - - | otherwise - -> do - let w = formatToWidth fmt - config <- getConfig - Amode addr addr_code <- memConstant (mkAlignment $ widthInBytes w) lit - let code dst = addr_code `snocOL` - movInstr config fmt (OpAddr addr) (OpReg dst) - return (Any fmt code) - where - is_zero (CmmInt i _) = i == 0 - is_zero (CmmFloat f _) = f == 0 -- TODO: mishandles negative zero - is_zero _ = False - - _ -> let imm = litToImm lit - code dst = unitOL (MOV fmt (OpImm imm) (OpReg dst)) - in return (Any fmt code) - where - cmmTy = cmmLitType platform lit - fmt = cmmTypeFormat cmmTy - getRegister' platform _ slot@(CmmStackSlot {}) = pprPanic "getRegister(x86) CmmStackSlot" (pdoc platform slot) -getFloatLitRegister :: CmmLit -> NatM Register -getFloatLitRegister lit = do - let w :: Width - w = case lit of { CmmInt _ w -> w; CmmFloat _ w -> w; _ -> panic "getFloatLitRegister" (ppr lit) } - Amode addr code <- memConstant (mkAlignment $ widthInBytes w) lit - loadFloatAmode w addr code - intLoadCode :: (Operand -> Operand -> Instr) -> CmmExpr -> NatM (Reg -> InstrBlock) intLoadCode instr mem = do @@ -2264,7 +2268,7 @@ getNonClobberedOperand (CmmLit lit) = return (OpAddr addr, code) else do platform <- getPlatform - if is32BitLit platform lit && not (isFloatType (cmmLitType platform lit)) + if is32BitLit platform lit && isIntFormat (cmmTypeFormat (cmmLitType platform lit)) then return (OpImm (litToImm lit), nilOL) else getNonClobberedOperand_generic (CmmLit lit) @@ -2321,13 +2325,13 @@ getOperand (CmmLit lit) = do else do platform <- getPlatform - if is32BitLit platform lit && not (isFloatType (cmmLitType platform lit)) + if is32BitLit platform lit && (isIntFormat $ cmmTypeFormat (cmmLitType platform lit)) then return (OpImm (litToImm lit), nilOL) else getOperand_generic (CmmLit lit) getOperand (CmmLoad mem ty _) = do is32Bit <- is32BitPlatform - if not (isFloatType ty) && (if is32Bit then not (isWord64 ty) else True) + if isIntFormat (cmmTypeFormat ty) && (if is32Bit then not (isWord64 ty) else True) then do Amode src mem_code <- getAmode mem return (OpAddr src, mem_code) @@ -2358,7 +2362,7 @@ addAlignmentCheck align reg = where check :: Format -> Reg -> InstrBlock check fmt reg = - assert (not $ isFloatFormat fmt) $ + assert (isIntFormat fmt) $ toOL [ TEST fmt (OpImm $ ImmInt $ align-1) (OpReg reg) , JXX_GBL NE $ ImmCLbl mkBadAlignmentLabel ] @@ -2382,15 +2386,12 @@ memConstant align lit = do `consOL` addr_code return (Amode addr code) - -loadFloatAmode :: Width -> AddrMode -> InstrBlock -> NatM Register -loadFloatAmode w addr addr_code = do - let format = floatFormat w - code dst = addr_code `snocOL` - MOV format (OpAddr addr) (OpReg dst) - - return (Any format code) - +-- | Load the value at the given address into any register. +loadAmode :: Format -> AddrMode -> InstrBlock -> NatM Register +loadAmode fmt addr addr_code = do + config <- getConfig + let load dst = movInstr config fmt (OpAddr addr) (OpReg dst) + return $ Any fmt (\ dst -> addr_code `snocOL` load dst) -- if we want a floating-point literal as an operand, we can -- use it directly from memory. However, if the literal is @@ -2403,7 +2404,7 @@ isSuitableFloatingPointLit _ = False getRegOrMem :: CmmExpr -> NatM (Operand, InstrBlock) getRegOrMem e@(CmmLoad mem ty _) = do is32Bit <- is32BitPlatform - if not (isFloatType ty) && (if is32Bit then not (isWord64 ty) else True) + if isIntFormat (cmmTypeFormat ty) && (if is32Bit then not (isWord64 ty) else True) then do Amode src mem_code <- getAmode mem return (OpAddr src, mem_code) @@ -3090,10 +3091,8 @@ genSimplePrim _ op dst args = do platform <- ncgPlatform <$> getConfig pprPanic "genSimplePrim: unhandled primop" (ppr (pprCallishMachOp op, dst, fmap (pdoc platform) args)) -{- -Note [Evaluate C-call arguments before placing in destination registers] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - +{- Note [Evaluate C-call arguments before placing in destination registers] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ When producing code for C calls we must take care when placing arguments in their final registers. Specifically, we must ensure that temporary register usage due to evaluation of one argument does not clobber a register in which we @@ -3144,15 +3143,11 @@ genForeignCall{32,64}. -- | See Note [Evaluate C-call arguments before placing in destination registers] evalArgs :: BlockId -> [CmmActual] -> NatM (InstrBlock, [CmmActual]) evalArgs bid actuals - | any mightContainMachOp actuals = do + | any loadIntoRegMightClobberOtherReg actuals = do regs_blks <- mapM evalArg actuals return (concatOL $ map fst regs_blks, map snd regs_blks) | otherwise = return (nilOL, actuals) where - mightContainMachOp (CmmReg _) = False - mightContainMachOp (CmmRegOff _ _) = False - mightContainMachOp (CmmLit _) = False - mightContainMachOp _ = True evalArg :: CmmActual -> NatM (InstrBlock, CmmExpr) evalArg actual = do @@ -3166,6 +3161,16 @@ evalArgs bid actuals newLocalReg :: CmmType -> NatM LocalReg newLocalReg ty = LocalReg <$> getUniqueM <*> pure ty +-- | Might the code to put this expression into a register +-- clobber any other registers? +loadIntoRegMightClobberOtherReg :: CmmExpr -> Bool +loadIntoRegMightClobberOtherReg (CmmReg _) = False +loadIntoRegMightClobberOtherReg (CmmRegOff _ _) = False +loadIntoRegMightClobberOtherReg (CmmLit _) = False + -- NB: this last 'False' is slightly risky, because the code for loading + -- a literal into a register is not entirely trivial. +loadIntoRegMightClobberOtherReg _ = True + -- Note [DIV/IDIV for bytes] -- ~~~~~~~~~~~~~~~~~~~~~~~~~ -- IDIV reminder: @@ -3246,6 +3251,15 @@ genCCall bid addr conv dest_regs args = do else genCCall64 addr conv dest_regs args' return (instrs0 `appOL` instrs1) +maybePromoteCArg :: Platform -> Width -> (CmmExpr, ForeignHint) -> CmmExpr +maybePromoteCArg platform wto (arg, hint) + | wfrom < wto = case hint of + SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] + _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] + | otherwise = arg + where + wfrom = cmmExprWidth platform arg + genCCall32 :: CmmExpr -- ^ address of the function to call -> ForeignConvention -- ^ calling convention -> [CmmFormal] -- ^ where to put the result @@ -3278,7 +3292,7 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do DELTA (delta-8)] ) - | isFloatType arg_ty = do + | isFloatType arg_ty || isVecType arg_ty = do (reg, code) <- getSomeReg arg delta <- getDeltaNat setDeltaNat (delta-size) @@ -3288,11 +3302,10 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do let addr = AddrBaseIndex (EABaseReg esp) EAIndexNone (ImmInt 0) - format = floatFormat (typeWidth arg_ty) + format = cmmTypeFormat arg_ty in - -- assume SSE2 - MOV format (OpReg reg) (OpAddr addr) + movInstr config format (OpReg reg) (OpAddr addr) ] ) @@ -3316,7 +3329,7 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do let -- Align stack to 16n for calls, assuming a starting stack -- alignment of 16n - word_size on procedure entry. Which we - -- maintiain. See Note [Stack Alignment on X86] in rts/StgCRun.c. + -- maintain. See Note [Stack Alignment on X86] in rts/StgCRun.c. sizes = map (arg_size_bytes . cmmExprType platform) (reverse args) raw_arg_size = sum sizes + platformWordSizeInBytes platform arg_pad_size = (roundTo 16 $ raw_arg_size) - raw_arg_size @@ -3361,6 +3374,8 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do -- assign the results, if necessary assign_code [] = nilOL assign_code [dest] + | isVecType ty + = unitOL (mkRegRegMoveInstr config (cmmTypeFormat ty) xmm0 r_dest) | isFloatType ty = -- we assume SSE2 let tmp_amode = AddrBaseIndex (EABaseReg esp) @@ -3402,189 +3417,70 @@ genCCall64 :: CmmExpr -- ^ address of function to call genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do config <- getConfig let platform = ncgPlatform config - -- load up the register arguments - let args_hints = zip args (argHints ++ repeat NoHint) - let prom_args = map (maybePromoteCArg platform W32) args_hints - - let load_args :: [CmmExpr] - -> [RegWithFormat] -- int regs avail for args - -> [RegWithFormat] -- FP regs avail for args - -> InstrBlock -- code computing args - -> InstrBlock -- code assigning args to ABI regs - -> NatM ([CmmExpr],[RegWithFormat],[RegWithFormat],InstrBlock,InstrBlock) - -- no more regs to use - load_args args [] [] code acode = - return (args, [], [], code, acode) - - -- no more args to push - load_args [] aregs fregs code acode = - return ([], aregs, fregs, code, acode) - - load_args (arg : rest) aregs fregs code acode - | isFloatType arg_rep = case fregs of - [] -> push_this_arg - (RegWithFormat r _fmt:rs) -> do - (code',acode') <- reg_this_arg r - load_args rest aregs rs code' acode' - | otherwise = case aregs of - [] -> push_this_arg - (RegWithFormat r _fmt:rs) -> do - (code',acode') <- reg_this_arg r - load_args rest rs fregs code' acode' - where + args_hints = zip args (argHints ++ repeat NoHint) + prom_args = map (maybePromoteCArg platform W32) args_hints + word_size = platformWordSizeInBytes platform + wordFmt = archWordFormat (target32Bit platform) - -- put arg into the list of stack pushed args - push_this_arg = do - (args',ars,frs,code',acode') - <- load_args rest aregs fregs code acode - return (arg:args', ars, frs, code', acode') - - -- pass the arg into the given register - reg_this_arg r - -- "operand" args can be directly assigned into r - | isOperand platform arg = do - arg_code <- getAnyReg arg - return (code, (acode `appOL` arg_code r)) - -- The last non-operand arg can be directly assigned after its - -- computation without going into a temporary register - | all (isOperand platform) rest = do - arg_code <- getAnyReg arg - return (code `appOL` arg_code r,acode) - - -- other args need to be computed beforehand to avoid clobbering - -- previously assigned registers used to pass parameters (see - -- #11792, #12614). They are assigned into temporary registers - -- and get assigned to proper call ABI registers after they all - -- have been computed. - | otherwise = do - arg_code <- getAnyReg arg - tmp <- getNewRegNat arg_fmt - let - code' = code `appOL` arg_code tmp - acode' = acode `snocOL` mkRegRegMoveInstr config arg_fmt tmp r - return (code',acode') - - arg_rep = cmmExprType platform arg - arg_fmt = cmmTypeFormat arg_rep - - load_args_win :: [CmmExpr] - -> [RegWithFormat] -- used int regs - -> [RegWithFormat] -- used FP regs - -> [(Reg, Reg)] -- (int, FP) regs avail for args - -> InstrBlock - -> NatM ([CmmExpr],[RegWithFormat],[RegWithFormat],InstrBlock,InstrBlock) - load_args_win args usedInt usedFP [] code - = return (args, usedInt, usedFP, code, nilOL) - -- no more regs to use - load_args_win [] usedInt usedFP _ code - = return ([], usedInt, usedFP, code, nilOL) - -- no more args to push - load_args_win (arg : rest) usedInt usedFP - ((ireg, freg) : regs) code - | isFloatType arg_rep = do - arg_code <- getAnyReg arg - load_args_win rest (RegWithFormat ireg II64: usedInt) (RegWithFormat freg FF64 : usedFP) regs - (code `appOL` - arg_code freg `snocOL` - -- If we are calling a varargs function - -- then we need to define ireg as well - -- as freg - MOVD FF64 (OpReg freg) (OpReg ireg)) - | otherwise = do - arg_code <- getAnyReg arg - load_args_win rest (RegWithFormat ireg II64: usedInt) usedFP regs - (code `appOL` arg_code ireg) - where - arg_rep = cmmExprType platform arg - - arg_size = 8 -- always, at the mo - - push_args [] code = return code - push_args (arg:rest) code - | isFloatType arg_rep = do - (arg_reg, arg_code) <- getSomeReg arg - delta <- getDeltaNat - setDeltaNat (delta-arg_size) - let fmt = floatFormat width - code' = code `appOL` arg_code `appOL` toOL [ - SUB (intFormat (wordWidth platform)) (OpImm (ImmInt arg_size)) (OpReg rsp), - DELTA (delta-arg_size), - MOV fmt (OpReg arg_reg) (OpAddr (spRel platform 0))] - push_args rest code' - - | otherwise = do - -- Arguments can be smaller than 64-bit, but we still use @PUSH - -- II64@ - the usual calling conventions expect integers to be - -- 8-byte aligned. - massert (width <= W64) - (arg_op, arg_code) <- getOperand arg - delta <- getDeltaNat - setDeltaNat (delta-arg_size) - let code' = code `appOL` arg_code `appOL` toOL [ - PUSH II64 arg_op, - DELTA (delta-arg_size)] - push_args rest code' - where - arg_rep = cmmExprType platform arg - width = typeWidth arg_rep - - leaveStackSpace n = do - delta <- getDeltaNat - setDeltaNat (delta - n * arg_size) - return $ toOL [ - SUB II64 (OpImm (ImmInt (n * platformWordSizeInBytes platform))) (OpReg rsp), - DELTA (delta - n * arg_size)] - - (stack_args, int_regs_used, fp_regs_used, load_args_code, assign_args_code) - <- - if platformOS platform == OSMinGW32 - then load_args_win prom_args [] [] (allArgRegs platform) nilOL - else do - let intArgRegs = map (\r -> RegWithFormat r II64) $ allIntArgRegs platform - fpArgRegs = map (\r -> RegWithFormat r FF64) $ allFPArgRegs platform - (stack_args, aregs, fregs, load_args_code, assign_args_code) - <- load_args prom_args intArgRegs fpArgRegs nilOL nilOL - let used_regs rs as = dropTail (length rs) as - fregs_used = used_regs fregs fpArgRegs - aregs_used = used_regs aregs intArgRegs - return (stack_args, aregs_used, fregs_used, load_args_code - , assign_args_code) + -- Compute the code for loading arguments into registers, + -- returning the leftover arguments that will need to be passed on the stack. + -- + -- NB: the code for loading references to data into registers is computed + -- later (in 'pushArgs'), because we don't yet know where the data will be + -- placed (due to alignment requirements). + LoadArgs + { stackArgs = proper_stack_args + , stackDataArgs = stack_data_args + , usedRegs = arg_regs_used + , assignArgsCode = assign_args_code + } + <- loadArgs config prom_args let - wordFmt = archWordFormat (target32Bit platform) - arg_regs_used = int_regs_used ++ fp_regs_used - arg_regs = [RegWithFormat eax wordFmt] ++ arg_regs_used - -- for annotating the call instruction with - sse_regs = length fp_regs_used - arg_stack_slots = if platformOS platform == OSMinGW32 - then length stack_args + length (allArgRegs platform) - else length stack_args - tot_arg_size = arg_size * arg_stack_slots + -- Pad all arguments and data passed on stack to align them properly. + (stk_args_with_padding, args_aligned_16) = + padStackArgs platform (proper_stack_args, stack_data_args) -- Align stack to 16n for calls, assuming a starting stack -- alignment of 16n - word_size on procedure entry. Which we -- maintain. See Note [Stack Alignment on X86] in rts/StgCRun.c - let word_size = platformWordSizeInBytes platform - (real_size, adjust_rsp) <- - if (tot_arg_size + word_size) `rem` 16 == 0 - then return (tot_arg_size, nilOL) - else do -- we need to adjust... - delta <- getDeltaNat - setDeltaNat (delta - word_size) - return (tot_arg_size + word_size, toOL [ - SUB II64 (OpImm (ImmInt word_size)) (OpReg rsp), - DELTA (delta - word_size) ]) - - -- push the stack args, right to left - push_code <- push_args (reverse stack_args) nilOL - -- On Win64, we also have to leave stack space for the arguments - -- that we are passing in registers - lss_code <- if platformOS platform == OSMinGW32 - then leaveStackSpace (length (allArgRegs platform)) - else return nilOL + need_realign_call = args_aligned_16 + align_call_code <- + if need_realign_call + then addStackPadding word_size + else return nilOL + + -- Compute the code that pushes data to the stack, and also + -- the code that loads references to that data into registers, + -- when the data is passed by reference in a register. + (load_data_refs, push_code) <- + pushArgs config proper_stack_args stk_args_with_padding + + -- On Windows, leave stack space for the arguments that we are passing + -- in registers (the so-called shadow space). + let shadow_space = + if platformOS platform == OSMinGW32 + then 8 * length (allArgRegs platform) + -- NB: the shadow store is always 8 * 4 = 32 bytes large, + -- i.e. the cumulative size of rcx, rdx, r8, r9 (see 'allArgRegs'). + else 0 + shadow_space_code <- addStackPadding shadow_space + + let total_args_size + = shadow_space + + sum (map (stackArgSpace platform) stk_args_with_padding) + real_size = + total_args_size + if need_realign_call then word_size else 0 + + -- End of argument passing. + -- + -- Next step: emit the appropriate call instruction. delta <- getDeltaNat + let arg_regs = [RegWithFormat eax wordFmt] ++ arg_regs_used + -- for annotating the call instruction with + -- deal with static vs dynamic call targets (callinsns,_cconv) <- case addr of CmmLit (CmmLabel lbl) -> @@ -3594,7 +3490,7 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do return (dyn_c `snocOL` CALL (Right dyn_r) arg_regs, conv) let - -- The x86_64 ABI requires us to set %al to the number of SSE2 + -- The System V AMD64 ABI requires us to set %al to the number of SSE2 -- registers that contain arguments, if the called routine -- is a varargs function. We don't know whether it's a -- varargs function or not, so we have to assume it is. @@ -3602,7 +3498,12 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do -- It's not safe to omit this assignment, even if the number -- of SSE2 regs in use is zero. If %al is larger than 8 -- on entry to a varargs function, seg faults ensue. - assign_eax n = unitOL (MOV II32 (OpImm (ImmInt n)) (OpReg eax)) + nb_sse_regs_used = count (isFloatFormat . regWithFormat_format) arg_regs_used + assign_eax_sse_regs + | platformOS platform == OSMinGW32 + = nilOL + | otherwise + = unitOL (MOV II32 (OpImm (ImmInt nb_sse_regs_used)) (OpReg eax)) let call = callinsns `appOL` toOL ( @@ -3619,7 +3520,7 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do assign_code [] = nilOL assign_code [dest] = unitOL $ - movInstr config fmt (OpReg reg) (OpReg r_dest) + mkRegRegMoveInstr config fmt reg r_dest where reg = if isIntFormat fmt then rax else xmm0 fmt = cmmTypeFormat rep @@ -3627,24 +3528,581 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do r_dest = getRegisterReg platform (CmmLocal dest) assign_code _many = panic "genForeignCall.assign_code many" - return (adjust_rsp `appOL` + return (align_call_code `appOL` push_code `appOL` - load_args_code `appOL` assign_args_code `appOL` - lss_code `appOL` - assign_eax sse_regs `appOL` + load_data_refs `appOL` + shadow_space_code `appOL` + assign_eax_sse_regs `appOL` call `appOL` assign_code dest_regs) +-- ----------------------------------------------------------------------------- +-- Loading arguments into registers for 64-bit C calls. + +-- | Information needed to know how to pass arguments in a C call, +-- and in particular how to load them into registers. +data LoadArgs + = LoadArgs + -- | Arguments that should be passed on the stack + { stackArgs :: [RawStackArg] + -- | Additional values to store onto the stack. + , stackDataArgs :: [CmmExpr] + -- | Which registers are we using for argument passing? + , usedRegs :: [RegWithFormat] + -- | The code to assign arguments to registers used for argument passing. + , assignArgsCode :: InstrBlock + } +instance Semigroup LoadArgs where + LoadArgs a1 d1 r1 j1 <> LoadArgs a2 d2 r2 j2 + = LoadArgs (a1 ++ a2) (d1 ++ d2) (r1 ++ r2) (j1 S.<> j2) +instance Monoid LoadArgs where + mempty = LoadArgs [] [] [] nilOL + +-- | An argument passed on the stack, either directly or by reference. +-- +-- The padding information hasn't yet been computed (see 'StackArg'). +data RawStackArg + -- | Pass the argument on the stack directly. + = RawStackArg { stackArgExpr :: CmmExpr } + -- | Pass the argument by reference. + | RawStackArgRef + { stackRef :: StackRef + -- ^ is the reference passed in a register, or on the stack? + , stackRefArgSize :: Int + -- ^ the size of the data pointed to + } + deriving ( Show ) + +-- | An argument passed on the stack, either directly or by reference, +-- with additional padding information. +data StackArg + -- | Pass the argument on the stack directly. + = StackArg + { stackArgExpr :: CmmExpr + , stackArgPadding :: Int + -- ^ padding required (in bytes) + } + -- | Pass the argument by reference. + | StackArgRef + { stackRef :: StackRef + -- ^ where the reference is passed + , stackRefArgSize :: Int + -- ^ the size of the data pointed to + , stackRefArgPadding :: Int + -- ^ padding of the data pointed to + -- (the reference itself never requires padding) + } + deriving ( Show ) + +-- | Where is a reference to data on the stack passed? +data StackRef + -- | In a register. + = InReg Reg + -- | On the stack. + | OnStack + deriving ( Eq, Ord, Show ) + +newtype Padding = Padding { paddingBytes :: Int } + deriving ( Show, Eq, Ord ) + +-- | How much space does this 'StackArg' take up on the stack? +-- +-- Only counts the "reference" part for references, not the data it points to. +stackArgSpace :: Platform -> StackArg -> Int +stackArgSpace platform = \case + StackArg arg padding -> + argSize platform arg + padding + StackArgRef { stackRef = ref } -> + case ref of + InReg {} -> 0 + OnStack {} -> 8 + +-- | Pad arguments, assuming we start aligned to a 16-byte boundary. +-- +-- Returns padded arguments, together with whether we end up aligned +-- to a 16-byte boundary. +padStackArgs :: Platform + -> ([RawStackArg], [CmmExpr]) + -> ([StackArg], Bool) +padStackArgs platform (args0, data_args0) = + let + -- Pad the direct args + (args, align_16_mid) = pad_args True args0 + + -- Pad the data section + (data_args, align_16_end) = pad_args align_16_mid (map RawStackArg data_args0) + + -- Now figure out where the data is placed relative to the direct arguments, + -- in order to resolve references. + resolve_args :: Int -> [((RawStackArg, Padding), Padding)] -> [StackArg] + resolve_args _ [] = [] + resolve_args i (((stk_arg, Padding pad), Padding arg_pad) : rest) = + let (this_arg, i') = + case stk_arg of + RawStackArg arg -> (StackArg arg pad, i) + RawStackArgRef ref size -> + let arg = + StackArgRef + { stackRef = ref + , stackRefArgSize = size + , stackRefArgPadding = arg_pad } + in (arg, i+1) + in this_arg : resolve_args i' rest + + args_data_pads :: [((RawStackArg, Padding), Padding)] + args_data_pads = zipWith (\ a ( _, p ) -> ( a, p )) args data_args + + in + ( resolve_args 0 args_data_pads ++ + [ case data_arg of + RawStackArg arg -> StackArg arg pad + RawStackArgRef {} -> panic "padStackArgs: reference in data section" + | (data_arg, Padding pad) <- data_args + ] + , align_16_end ) + + where + pad_args :: Bool -> [RawStackArg] -> ([(RawStackArg, Padding)], Bool) + pad_args aligned_16 [] = ([], aligned_16) + pad_args aligned_16 (arg:args) + | needed_alignment > 16 + -- We don't know if the stack is aligned to 8 (mod 32) or 24 (mod 32). + -- This makes aligning the stack to a 32 or 64 byte boundary more + -- complicated, in particular with DELTA. + = sorry $ unlines + [ "X86_86 C call: unsupported argument." + , " Alignment requirement: " ++ show needed_alignment ++ " bytes." + , if platformOS platform == OSMinGW32 + then " The X86_64 NCG does not (yet) support Windows C calls with 256/512 bit vectors." + else " The X86_64 NCG cannot (yet) pass 256/512 bit vectors on the stack for C calls." + , " Please use the LLVM backend (-fllvm)." ] + | otherwise + = let ( rest, final_align_16 ) = pad_args next_aligned_16 args + in ( (arg, Padding padding) : rest, final_align_16 ) -maybePromoteCArg :: Platform -> Width -> (CmmExpr, ForeignHint) -> CmmExpr -maybePromoteCArg platform wto (arg, hint) - | wfrom < wto = case hint of - SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] - _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] - | otherwise = arg - where - wfrom = cmmExprWidth platform arg + where + needed_alignment = case arg of + RawStackArg arg -> argSize platform arg + RawStackArgRef {} -> platformWordSizeInBytes platform + padding + | needed_alignment < 16 || aligned_16 + = 0 + | otherwise + = 8 + next_aligned_16 = not ( aligned_16 && needed_alignment < 16 ) + +-- | Load arguments into available registers. +loadArgs :: NCGConfig -> [CmmExpr] -> NatM LoadArgs +loadArgs config args + | platformOS platform == OSMinGW32 + = evalStateT (loadArgsWin config args) (allArgRegs platform) + | otherwise + = evalStateT (loadArgsSysV config args) (allIntArgRegs platform + ,allFPArgRegs platform) + where + platform = ncgPlatform config + +-- | Load arguments into available registers (System V AMD64 ABI). +loadArgsSysV :: NCGConfig + -> [CmmExpr] + -> StateT ([Reg], [Reg]) NatM LoadArgs +loadArgsSysV _ [] = return mempty +loadArgsSysV config (arg:rest) = do + (iregs, fregs) <- get + -- No available registers: pass everything on the stack (shortcut). + if null iregs && null fregs + then return $ + LoadArgs + { stackArgs = map RawStackArg (arg:rest) + , stackDataArgs = [] + , assignArgsCode = nilOL + , usedRegs = [] + } + else do + mbReg <- + if + | isIntFormat arg_fmt + , ireg:iregs' <- iregs + -> do put (iregs', fregs) + return $ Just ireg + | isFloatFormat arg_fmt || isVecFormat arg_fmt + , freg:fregs' <- fregs + -> do put (iregs, fregs') + return $ Just freg + | otherwise + -> return Nothing + this_arg <- + case mbReg of + Just reg -> do + assign_code <- lift $ loadArgIntoReg arg reg + return $ + LoadArgs + { stackArgs = [] -- passed in register + , stackDataArgs = [] + , assignArgsCode = assign_code + , usedRegs = [RegWithFormat reg arg_fmt] + } + Nothing -> do + return $ + -- No available register for this argument: pass it on the stack. + LoadArgs + { stackArgs = [RawStackArg arg] + , stackDataArgs = [] + , assignArgsCode = nilOL + , usedRegs = [] + } + others <- loadArgsSysV config rest + return $ this_arg S.<> others + + where + platform = ncgPlatform config + arg_fmt = cmmTypeFormat (cmmExprType platform arg) + +-- | Compute all things that will need to be pushed to the stack. +-- +-- On Windows, an argument passed by reference will require two pieces of data: +-- +-- - the reference (returned in the first position) +-- - the actual data (returned in the second position) +computeWinPushArgs :: Platform -> [CmmExpr] -> ([RawStackArg], [CmmExpr]) +computeWinPushArgs platform = go + where + go :: [CmmExpr] -> ([RawStackArg], [CmmExpr]) + go [] = ([], []) + go (arg:args) = + let + arg_size = argSize platform arg + (this_arg, add_this_arg) + | arg_size > 8 + = ( RawStackArgRef OnStack arg_size, (arg :) ) + | otherwise + = ( RawStackArg arg, id ) + (stk_args, stk_data) = go args + in + (this_arg:stk_args, add_this_arg stk_data) + +-- | Load arguments into available registers (Windows C X64 calling convention). +loadArgsWin :: NCGConfig -> [CmmExpr] -> StateT [(Reg,Reg)] NatM LoadArgs +loadArgsWin _ [] = return mempty +loadArgsWin config (arg:rest) = do + regs <- get + case regs of + reg:regs' -> do + put regs' + this_arg <- lift $ load_arg_win reg + rest <- loadArgsWin config rest + return $ this_arg S.<> rest + [] -> do + -- No more registers available: pass all (remaining) arguments on the stack. + let (stk_args, data_args) = computeWinPushArgs platform (arg:rest) + return $ + LoadArgs + { stackArgs = stk_args + , stackDataArgs = data_args + , assignArgsCode = nilOL + , usedRegs = [] + } + where + platform = ncgPlatform config + arg_fmt = cmmTypeFormat $ cmmExprType platform arg + load_arg_win (ireg, freg) + | isVecFormat arg_fmt + -- Vectors are passed by reference. + -- See Note [The Windows X64 C calling convention]. + = do return $ + LoadArgs + -- Pass the reference in a register, + -- and the argument data on the stack. + { stackArgs = [RawStackArgRef (InReg ireg) (argSize platform arg)] + , stackDataArgs = [arg] -- we don't yet know where the data will reside, + , assignArgsCode = nilOL -- so we defer computing the reference and storing it + -- in the register until later + , usedRegs = [RegWithFormat ireg II64] + } + | otherwise + = do let arg_reg + | isFloatFormat arg_fmt + = freg + | otherwise + = ireg + assign_code <- loadArgIntoReg arg arg_reg + -- Recall that, for varargs, we must pass floating-point + -- arguments in both fp and integer registers. + let (assign_code', regs') + | isFloatFormat arg_fmt = + ( assign_code `snocOL` MOVD FF64 (OpReg freg) (OpReg ireg), + [ RegWithFormat freg FF64 + , RegWithFormat ireg II64 ]) + | otherwise = (assign_code, [RegWithFormat ireg II64]) + return $ + LoadArgs + { stackArgs = [] -- passed in register + , stackDataArgs = [] + , assignArgsCode = assign_code' + , usedRegs = regs' + } + +-- | Load an argument into a register. +-- +-- Assumes that the expression does not contain any MachOps, +-- as per Note [Evaluate C-call arguments before placing in destination registers]. +loadArgIntoReg :: CmmExpr -> Reg -> NatM InstrBlock +loadArgIntoReg arg reg = do + when (debugIsOn && loadIntoRegMightClobberOtherReg arg) $ do + platform <- getPlatform + massertPpr False $ + vcat [ text "loadArgIntoReg: arg might contain MachOp" + , text "arg:" <+> pdoc platform arg ] + arg_code <- getAnyReg arg + return $ arg_code reg + +-- ----------------------------------------------------------------------------- +-- Pushing arguments onto the stack for 64-bit C calls. + +-- | The size of an argument (in bytes). +-- +-- Never smaller than the platform word width. +argSize :: Platform -> CmmExpr -> Int +argSize platform arg = + max (platformWordSizeInBytes platform) $ + widthInBytes (typeWidth $ cmmExprType platform arg) + +-- | Add the given amount of padding on the stack. +addStackPadding :: Int -- ^ padding (in bytes) + -> NatM InstrBlock +addStackPadding pad_bytes + | pad_bytes == 0 + = return nilOL + | otherwise + = do delta <- getDeltaNat + setDeltaNat (delta - pad_bytes) + return $ + toOL [ SUB II64 (OpImm (ImmInt pad_bytes)) (OpReg rsp) + , DELTA (delta - pad_bytes) + ] + +-- | Push one argument directly to the stack (by value). +-- +-- Assumes the current stack pointer fulfills any necessary alignment requirements. +pushArgByValue :: NCGConfig -> CmmExpr -> NatM InstrBlock +pushArgByValue config arg = do + (arg_reg, arg_code) <- getSomeReg arg + delta <- getDeltaNat + setDeltaNat (delta-arg_size) + let fmt = cmmTypeFormat arg_rep + return $ arg_code `appOL` toOL + [ SUB (intFormat (wordWidth platform)) (OpImm (ImmInt arg_size)) (OpReg rsp) + , DELTA (delta-arg_size) + -- NB: it's important to only move the actual size, + -- e.g. for W8 width we must only move 8 bits and not 64. + -- Suppose for example the argument is (CmmLoad addr W8 aln); + -- then we must make sure not to try to read more than 8 bits from 'addr'. + , movInstr config fmt (OpReg arg_reg) (OpAddr (spRel platform 0)) ] + where + platform = ncgPlatform config + arg_size = argSize platform arg + arg_rep = cmmExprType platform arg + +-- | Load an argument into a register or push it to the stack. +loadOrPushArg :: NCGConfig -> (StackArg, Maybe Int) -> NatM (InstrBlock, InstrBlock) +loadOrPushArg config (stk_arg, mb_off) = + case stk_arg of + StackArg arg pad -> do + push_code <- pushArgByValue config arg + pad_code <- addStackPadding pad + return (nilOL, push_code `appOL` pad_code) + StackArgRef { stackRef = ref } -> + case ref of + -- Pass the reference in a register + InReg ireg -> + return (unitOL $ LEA II64 (OpAddr (spRel platform off)) (OpReg ireg), nilOL) + -- Pass the reference on the stack + OnStack {} -> do + tmp <- getNewRegNat II64 + delta <- getDeltaNat + setDeltaNat (delta-arg_ref_size) + let push_code = toOL + [ SUB (intFormat (wordWidth platform)) (OpImm (ImmInt arg_ref_size)) (OpReg rsp) + , DELTA (delta-arg_ref_size) + , LEA II64 (OpAddr (spRel platform off)) (OpReg tmp) + , MOV II64 (OpReg tmp) (OpAddr (spRel platform 0)) ] + return (nilOL, push_code) + where off = expectJust "push_arg_win offset" mb_off + where + arg_ref_size = 8 -- passing a reference to the argument + platform = ncgPlatform config + +-- | Push arguments to the stack, right to left. +-- +-- On Windows, some arguments may need to be passed by reference, +-- which requires separately passing the data and the reference. +-- See Note [The Windows X64 C calling convention]. +pushArgs :: NCGConfig + -> [RawStackArg] + -- ^ arguments proper (i.e. don't include the data for arguments passed by reference) + -> [StackArg] + -- ^ arguments we are passing on the stack + -> NatM (InstrBlock, InstrBlock) +pushArgs config proper_args all_stk_args + = do { let + vec_offs :: [Maybe Int] + vec_offs + | platformOS platform == OSMinGW32 + = go stack_arg_size all_stk_args + | otherwise + = repeat Nothing + + --------------------- + -- Windows-only code + + -- Size of the arguments we are passing on the stack, counting only + -- the reference part for arguments passed by reference. + stack_arg_size = 8 * count not_in_reg proper_args + not_in_reg (RawStackArg {}) = True + not_in_reg (RawStackArgRef { stackRef = ref }) = + case ref of + InReg {} -> False + OnStack {} -> True + + -- Check an offset is valid (8-byte aligned), for assertions. + ok off = off `rem` 8 == 0 + + -- Tricky code: compute the stack offset to the vector data + -- for this argument. + -- + -- If you're confused, Note [The Windows X64 C calling convention] + -- contains a helpful diagram. + go :: Int -> [StackArg] -> [Maybe Int] + go _ [] = [] + go off (stk_arg:args) = + assertPpr (ok off) (text "unaligned offset:" <+> ppr off) $ + case stk_arg of + StackArg {} -> + -- Only account for the stack pointer movement. + let off' = off - stackArgSpace platform stk_arg + in Nothing : go off' args + StackArgRef + { stackRefArgSize = data_size + , stackRefArgPadding = data_pad } -> + assertPpr (ok data_size) (text "unaligned data size:" <+> ppr data_size) $ + assertPpr (ok data_pad) (text "unaligned data padding:" <+> ppr data_pad) $ + let off' = off + -- Next piece of data is after the data for this reference + + data_size + data_pad + -- ... and account for the stack pointer movement. + - stackArgSpace platform stk_arg + in Just (data_pad + off) : go off' args + + -- end of Windows-only code + ---------------------------- + + -- Push the stack arguments (right to left), + -- including both the reference and the data for arguments passed by reference. + ; (load_regs, push_args) <- foldMapM (loadOrPushArg config) (reverse $ zip all_stk_args vec_offs) + ; return (load_regs, push_args) } + where + platform = ncgPlatform config + +{- Note [The Windows X64 C calling convention] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Here are a few facts about the Windows X64 C calling convention that +are important: + + - any argument larger than 8 bytes must be passed by reference, + and arguments smaller than 8 bytes are padded to 8 bytes. + + - the first four arguments are passed in registers: + - floating-point scalar arguments are passed in %xmm0, %xmm1, %xmm2, %xmm3 + - other arguments are passed in %rcx, %rdx, %r8, %r9 + (this includes vector arguments, passed by reference) + + For variadic functions, it is additionally expected that floating point + scalar arguments are copied to the corresponding integer register, e.g. + the data in xmm2 should also be copied to r8. + + There is no requirement about setting %al like there is for the + System V AMD64 ABI. + + - subsequent arguments are passed on the stack. + +There are also alignment requirements: + + - the data for vectors must be aligned to the size of the vector, + e.g. a 32 byte vector must be aligned on a 32 byte boundary, + + - the call instruction must be aligned to 16 bytes. + (This differs from the System V AMD64 ABI, which mandates that the call + instruction must be aligned to 32 bytes if there are any 32 byte vectors + passed on the stack.) + +This motivates our handling of vector values. Suppose we have a function call +with many arguments, several of them being vectors. We proceed as follows: + + - Add some padding, if necessary, to ensure the stack, when executing the call + instruction, is 16-byte aligned. Whether this padding is necessary depends + on what happens next. (Recall also that we start off at 8 (mod 16) alignment, + as per Note [Stack Alignment on X86] in rts/StgCRun.c) + - Push all the vectors to the stack first, adding padding after each one + if necessary. + - Then push the arguments: + - for non-vectors, proceed as usual, + - for vectors, push the address of the vector data we pushed above. + - Then assign the registers: + - for non-vectors, proceed as usual, + - for vectors, store the address in a general-purpose register, as opposed + to storing the data in an xmm register. + +For a concrete example, suppose we have a call of the form: + + f x1 x2 x3 x4 x5 x6 x7 + +in which: + + - x2, x3, x5 and x7 are 16 byte vectors + - the other arguments are all 8 byte wide + +Now, x1, x2, x3, x4 will get passed in registers, except that we pass +x2 and x3 by reference, because they are vectors. We proceed as follows: + + - push the vectors to the stack: x7, x5, x3, x2 (in that order) + - push the stack arguments in order: addr(x7), x6, addr(x5) + - load the remaining arguments into registers: x4, addr(x3), addr(x2), x1 + +The tricky part is to get the right offsets for the addresses of the vector +data. The following visualisation will hopefully clear things up: + + ┌──┐ + │▓▓│ ─── padding to align the call instruction + ╭─╴ ╞══╡ (ensures Sp, below, is 16-byte aligned) + │ │ │ + │ x7 ───╴ │ │ + │ ├──┤ + │ │ │ + │ x5 ───╴ │ │ + │ ├──┤ + vector data ────┤ │ │ +(individually padded) │ x3 ───╴ │ │ + │ ├──┤ + │ │ │ + │ x2 ───╴ │ │ + │ ├┄┄┤ + │ │▓▓│ ─── padding to align x2 to 16 bytes + ╭─╴ ╰─╴ ╞══╡ + │ addr(x7) ───╴ │ │ ╭─ from here: x7 is +64 + │ ├──┤ ╾──╯ = 64 (position of x5) + stack ───┤ x6 ───╴ │ │ + 16 (size of x5) + 0 (padding of x7) + arguments │ ├──┤ - 2 * 8 (x7 is 2 arguments higher than x5) + │ addr(x5) ───╴ │ │ + ╰─╴ ╭─╴ ╞══╡ ╾─── from here: + │ │ │ - x2 is +32 = 24 (stack_arg_size) + 8 (padding of x2) + shadow ───┤ │ │ - x3 is +48 = 32 (position of x2) + 16 (size of x2) + 0 (padding of x3) + space │ │ │ - x5 is +64 = 48 (position of x3) + 16 (size of x3) + 0 (padding of x5) + │ │ │ + ╰─╴ └──┘ ╾─── Sp + +This is all tested in the simd013 test. +-} -- ----------------------------------------------------------------------------- -- Generating a table-branch ===================================== compiler/GHC/CmmToAsm/X86/Instr.hs ===================================== @@ -306,6 +306,7 @@ data Instr | VMOVDQU Format Operand Operand -- logic operations + | PXOR Format Operand Reg | VPXOR Format Reg Reg Reg -- Arithmetic @@ -492,6 +493,12 @@ regUsageOfInstr platform instr MOVDQU fmt src dst -> mkRU (use_R fmt src []) (use_R fmt dst []) VMOVDQU fmt src dst -> mkRU (use_R fmt src []) (use_R fmt dst []) + PXOR fmt (OpReg src) dst + | src == dst + -> mkRU [] [mk fmt dst] + | otherwise + -> mkRU [mk fmt src, mk fmt dst] [mk fmt dst] + VPXOR fmt s1 s2 dst | s1 == s2, s1 == dst -> mkRU [] [mk fmt dst] @@ -733,6 +740,7 @@ patchRegsOfInstr platform instr env MOVDQU fmt src dst -> MOVDQU fmt (patchOp src) (patchOp dst) VMOVDQU fmt src dst -> VMOVDQU fmt (patchOp src) (patchOp dst) + PXOR fmt src dst -> PXOR fmt (patchOp src) (env dst) VPXOR fmt s1 s2 dst -> VPXOR fmt (env s1) (env s2) (env dst) VADD fmt s1 s2 dst -> VADD fmt (patchOp s1) (env s2) (env dst) @@ -886,7 +894,7 @@ mkLoadInstr config (RegWithFormat reg fmt) delta slot = -- | A move instruction for moving the entire contents of an operand -- at the given 'Format'. -movInstr :: NCGConfig -> Format -> (Operand -> Operand -> Instr) +movInstr :: HasDebugCallStack => NCGConfig -> Format -> (Operand -> Operand -> Instr) movInstr config fmt = case fmt of VecFormat _ sFmt -> @@ -908,16 +916,37 @@ movInstr config fmt = w -> sorry $ "Unhandled SIMD vector width: " ++ show w ++ " bits" _ -> MOV fmt where + plat = ncgPlatform config avx = ncgAvxEnabled config avx2 = ncgAvx2Enabled config avx512f = ncgAvx512fEnabled config avx_move sFmt = if isFloatScalarFormat sFmt - then VMOVU fmt + then \ op1 op2 -> + if + | OpReg r1 <- op1 + , OpReg r2 <- op2 + , targetClassOfReg plat r1 /= targetClassOfReg plat r2 + -> pprPanic "movInstr: VMOVU between incompatible registers" + ( vcat [ text "fmt:" <+> ppr fmt + , text "r1:" <+> ppr r1 + , text "r2:" <+> ppr r2 ] ) + | otherwise + -> VMOVU fmt op1 op2 else VMOVDQU fmt sse_move sFmt = if isFloatScalarFormat sFmt - then MOVU fmt + then \ op1 op2 -> + if + | OpReg r1 <- op1 + , OpReg r2 <- op2 + , targetClassOfReg plat r1 /= targetClassOfReg plat r2 + -> pprPanic "movInstr: MOVU between incompatible registers" + ( vcat [ text "fmt:" <+> ppr fmt + , text "r1:" <+> ppr r1 + , text "r2:" <+> ppr r2 ] ) + | otherwise + -> MOVU fmt op1 op2 else MOVDQU fmt -- NB: we are using {V}MOVU and not {V}MOVA, because we have no guarantees -- about the stack being sufficiently aligned (even for even numbered stack slots). @@ -982,12 +1011,7 @@ mkRegRegMoveInstr -> Reg -> Instr mkRegRegMoveInstr config fmt src dst = - assertPpr (targetClassOfReg platform src == targetClassOfReg platform dst) - (vcat [ text "mkRegRegMoveInstr: incompatible register classes" - , text "fmt:" <+> ppr fmt - , text "src:" <+> ppr src - , text "dst:" <+> ppr dst ]) $ - movInstr config fmt' (OpReg src) (OpReg dst) + movInstr config fmt' (OpReg src) (OpReg dst) -- Move the platform word size, at a minimum where platform = ncgPlatform config ===================================== compiler/GHC/CmmToAsm/X86/Ppr.hs ===================================== @@ -1016,6 +1016,8 @@ pprInstr platform i = case i of VecFormat 64 FmtInt8 -> text "vmovdqu32" -- require the additional AVX512BW extension _ -> text "vmovdqu" + PXOR format src dst + -> pprPXor (text "pxor") format src dst VPXOR format s1 s2 dst -> pprXor (text "vpxor") format s1 s2 dst VEXTRACT format offset from to @@ -1320,6 +1322,15 @@ pprInstr platform i = case i of pprReg platform format reg3 ] + pprPXor :: Line doc -> Format -> Operand -> Reg -> doc + pprPXor name format src dst + = line $ hcat [ + pprGenMnemonic name format, + pprOperand platform format src, + comma, + pprReg platform format dst + ] + pprVxor :: Format -> Operand -> Reg -> Reg -> doc pprVxor fmt src1 src2 dst = line $ hcat [ ===================================== compiler/GHC/CmmToLlvm.hs ===================================== @@ -139,7 +139,7 @@ llvmGroupLlvmGens cmm = do Nothing -> l Just (CmmStaticsRaw info_lbl _) -> info_lbl lml <- strCLabel_llvm l' - funInsert lml =<< llvmFunTy (map globalRegUse_reg live) + funInsert lml =<< llvmFunTy live return Nothing cdata <- fmap catMaybes $ mapM split cmm ===================================== compiler/GHC/CmmToLlvm/Base.hs ===================================== @@ -12,7 +12,7 @@ module GHC.CmmToLlvm.Base ( LlvmCmmDecl, LlvmBasicBlock, - LiveGlobalRegs, + LiveGlobalRegs, LiveGlobalRegUses, LlvmUnresData, LlvmData, UnresLabel, UnresStatic, LlvmM, @@ -29,6 +29,8 @@ module GHC.CmmToLlvm.Base ( llvmFunSig, llvmFunArgs, llvmStdFunAttrs, llvmFunAlign, llvmInfAlign, llvmPtrBits, tysToParams, llvmFunSection, padLiveArgs, isFPR, + lookupRegUse, + strCLabel_llvm, getGlobalPtr, generateExternDecls, @@ -58,9 +60,11 @@ import GHC.Types.Unique.Set import GHC.Types.Unique.Supply import GHC.Utils.Logger -import Data.Maybe (fromJust) import Control.Monad.Trans.State (StateT (..)) -import Data.List (isPrefixOf) +import Control.Applicative (Alternative((<|>))) +import Data.Maybe (fromJust, mapMaybe) + +import Data.List (find, isPrefixOf) import qualified Data.List.NonEmpty as NE import Data.Ord (comparing) @@ -73,6 +77,7 @@ type LlvmBasicBlock = GenBasicBlock LlvmStatement -- | Global registers live on proc entry type LiveGlobalRegs = [GlobalReg] +type LiveGlobalRegUses = [GlobalRegUse] -- | Unresolved code. -- Of the form: (data label, data type, unresolved data) @@ -116,16 +121,16 @@ llvmGhcCC platform | otherwise = CC_Ghc -- | Llvm Function type for Cmm function -llvmFunTy :: LiveGlobalRegs -> LlvmM LlvmType +llvmFunTy :: LiveGlobalRegUses -> LlvmM LlvmType llvmFunTy live = return . LMFunction =<< llvmFunSig' live (fsLit "a") ExternallyVisible -- | Llvm Function signature -llvmFunSig :: LiveGlobalRegs -> CLabel -> LlvmLinkageType -> LlvmM LlvmFunctionDecl +llvmFunSig :: LiveGlobalRegUses -> CLabel -> LlvmLinkageType -> LlvmM LlvmFunctionDecl llvmFunSig live lbl link = do lbl' <- strCLabel_llvm lbl llvmFunSig' live lbl' link -llvmFunSig' :: LiveGlobalRegs -> LMString -> LlvmLinkageType -> LlvmM LlvmFunctionDecl +llvmFunSig' :: LiveGlobalRegUses -> LMString -> LlvmLinkageType -> LlvmM LlvmFunctionDecl llvmFunSig' live lbl link = do let toParams x | isPointer x = (x, [NoAlias, NoCapture]) | otherwise = (x, []) @@ -149,16 +154,25 @@ llvmFunSection opts lbl | otherwise = Nothing -- | A Function's arguments -llvmFunArgs :: Platform -> LiveGlobalRegs -> [LlvmVar] +llvmFunArgs :: Platform -> LiveGlobalRegUses -> [LlvmVar] llvmFunArgs platform live = - map (lmGlobalRegArg platform) (filter isPassed allRegs) + map (lmGlobalRegArg platform) (mapMaybe isPassed allRegs) where allRegs = activeStgRegs platform paddingRegs = padLiveArgs platform live - isLive r = r `elem` alwaysLive - || r `elem` live - || r `elem` paddingRegs - isPassed r = not (isFPR r) || isLive r - + isLive :: GlobalReg -> Maybe GlobalRegUse + isLive r = + lookupRegUse r (alwaysLive platform) + <|> + lookupRegUse r live + <|> + lookupRegUse r paddingRegs + isPassed r = + if not (isFPR r) + then Just $ GlobalRegUse r (globalRegSpillType platform r) + else isLive r + +lookupRegUse :: GlobalReg -> [GlobalRegUse] -> Maybe GlobalRegUse +lookupRegUse r = find ((== r) . globalRegUse_reg) isFPR :: GlobalReg -> Bool isFPR (FloatReg _) = True @@ -179,7 +193,7 @@ isFPR _ = False -- Invariant: Cmm FPR regs with number "n" maps to real registers with number -- "n" If the calling convention uses registers in a different order or if the -- invariant doesn't hold, this code probably won't be correct. -padLiveArgs :: Platform -> LiveGlobalRegs -> LiveGlobalRegs +padLiveArgs :: Platform -> LiveGlobalRegUses -> LiveGlobalRegUses padLiveArgs platform live = if platformUnregisterised platform then [] -- not using GHC's register convention for platform. @@ -188,7 +202,7 @@ padLiveArgs platform live = ---------------------------------- -- handle floating-point registers (FPR) - fprLive = filter isFPR live -- real live FPR registers + fprLive = filter (isFPR . globalRegUse_reg) live -- real live FPR registers -- we group live registers sharing the same classes, i.e. that use the same -- set of real registers to be passed. E.g. FloatReg, DoubleReg and XmmReg @@ -196,39 +210,44 @@ padLiveArgs platform live = -- classes = NE.groupBy sharesClass fprLive sharesClass a b = globalRegsOverlap platform (norm a) (norm b) -- check if mapped to overlapping registers - norm x = fpr_ctor x 1 -- get the first register of the family + norm x = globalRegUse_reg (fpr_ctor x 1) -- get the first register of the family -- For each class, we just have to fill missing registers numbers. We use -- the constructor of the greatest register to build padding registers. -- -- E.g. sortedRs = [ F2, XMM4, D5] -- output = [D1, D3] + padded :: [GlobalRegUse] padded = concatMap padClass classes + + padClass :: NE.NonEmpty GlobalRegUse -> [GlobalRegUse] padClass rs = go (NE.toList sortedRs) 1 where - sortedRs = NE.sortBy (comparing fpr_num) rs + sortedRs = NE.sortBy (comparing (fpr_num . globalRegUse_reg)) rs maxr = NE.last sortedRs ctor = fpr_ctor maxr go [] _ = [] - go (c1:c2:_) _ -- detect bogus case (see #17920) + go (GlobalRegUse c1 _: GlobalRegUse c2 _:_) _ -- detect bogus case (see #17920) | fpr_num c1 == fpr_num c2 , Just real <- globalRegMaybe platform c1 = sorryDoc "LLVM code generator" $ text "Found two different Cmm registers (" <> ppr c1 <> text "," <> ppr c2 <> text ") both alive AND mapped to the same real register: " <> ppr real <> text ". This isn't currently supported by the LLVM backend." - go (c:cs) f - | fpr_num c == f = go cs f -- already covered by a real register - | otherwise = ctor f : go (c:cs) (f + 1) -- add padding register - - fpr_ctor :: GlobalReg -> Int -> GlobalReg - fpr_ctor (FloatReg _) = FloatReg - fpr_ctor (DoubleReg _) = DoubleReg - fpr_ctor (XmmReg _) = XmmReg - fpr_ctor (YmmReg _) = YmmReg - fpr_ctor (ZmmReg _) = ZmmReg - fpr_ctor _ = error "fpr_ctor expected only FPR regs" + go (cu@(GlobalRegUse c _):cs) f + | fpr_num c == f = go cs f -- already covered by a real register + | otherwise = ctor f : go (cu:cs) (f + 1) -- add padding register + + fpr_ctor :: GlobalRegUse -> Int -> GlobalRegUse + fpr_ctor (GlobalRegUse r fmt) i = + case r of + FloatReg _ -> GlobalRegUse (FloatReg i) fmt + DoubleReg _ -> GlobalRegUse (DoubleReg i) fmt + XmmReg _ -> GlobalRegUse (XmmReg i) fmt + YmmReg _ -> GlobalRegUse (YmmReg i) fmt + ZmmReg _ -> GlobalRegUse (ZmmReg i) fmt + _ -> error "fpr_ctor expected only FPR regs" fpr_num :: GlobalReg -> Int fpr_num (FloatReg i) = i ===================================== compiler/GHC/CmmToLlvm/CodeGen.hs ===================================== @@ -37,13 +37,14 @@ import GHC.Utils.Outputable import qualified GHC.Utils.Panic as Panic import GHC.Utils.Misc +import Control.Applicative (Alternative((<|>))) import Control.Monad.Trans.Class import Control.Monad.Trans.Writer import Control.Monad import qualified Data.Semigroup as Semigroup import Data.List ( nub ) -import Data.Maybe ( catMaybes ) +import Data.Maybe ( catMaybes, isJust ) type Atomic = Maybe MemoryOrdering type LlvmStatements = OrdList LlvmStatement @@ -57,7 +58,7 @@ genLlvmProc :: RawCmmDecl -> LlvmM [LlvmCmmDecl] genLlvmProc (CmmProc infos lbl live graph) = do let blocks = toBlockListEntryFirstFalseFallthrough graph - (lmblocks, lmdata) <- basicBlocksCodeGen (map globalRegUse_reg live) blocks + (lmblocks, lmdata) <- basicBlocksCodeGen live blocks let info = mapLookup (g_entry graph) infos proc = CmmProc info lbl live (ListGraph lmblocks) return (proc:lmdata) @@ -76,7 +77,7 @@ newtype UnreachableBlockId = UnreachableBlockId BlockId -- | Generate code for a list of blocks that make up a complete -- procedure. The first block in the list is expected to be the entry -- point. -basicBlocksCodeGen :: LiveGlobalRegs -> [CmmBlock] +basicBlocksCodeGen :: LiveGlobalRegUses -> [CmmBlock] -> LlvmM ([LlvmBasicBlock], [LlvmCmmDecl]) basicBlocksCodeGen _ [] = panic "no entry block!" basicBlocksCodeGen live cmmBlocks @@ -152,7 +153,7 @@ stmtToInstrs ubid stmt = case stmt of -- Tail call CmmCall { cml_target = arg, - cml_args_regs = live } -> genJump arg $ map globalRegUse_reg live + cml_args_regs = live } -> genJump arg live _ -> panic "Llvm.CodeGen.stmtToInstrs" @@ -1050,7 +1051,7 @@ cmmPrimOpFunctions mop = do -- | Tail function calls -genJump :: CmmExpr -> [GlobalReg] -> LlvmM StmtData +genJump :: CmmExpr -> LiveGlobalRegUses -> LlvmM StmtData -- Call to known function genJump (CmmLit (CmmLabel lbl)) live = do @@ -2056,14 +2057,13 @@ getCmmReg (CmmLocal (LocalReg un _)) -- have been assigned a value at some point, triggering -- "funPrologue" to allocate it on the stack. -getCmmReg (CmmGlobal g) - = do let r = globalRegUse_reg g - onStack <- checkStackReg r +getCmmReg (CmmGlobal ru@(GlobalRegUse r _)) + = do onStack <- checkStackReg r platform <- getPlatform if onStack - then return (lmGlobalRegVar platform r) + then return (lmGlobalRegVar platform ru) else pprPanic "getCmmReg: Cmm register " $ - ppr g <> text " not stack-allocated!" + ppr r <> text " not stack-allocated!" -- | Return the value of a given register, as well as its type. Might -- need to be load from stack. @@ -2074,7 +2074,7 @@ getCmmRegVal reg = onStack <- checkStackReg (globalRegUse_reg g) platform <- getPlatform if onStack then loadFromStack else do - let r = lmGlobalRegArg platform (globalRegUse_reg g) + let r = lmGlobalRegArg platform g return (r, getVarType r, nilOL) _ -> loadFromStack where loadFromStack = do @@ -2187,8 +2187,9 @@ convertMemoryOrdering MemOrderSeqCst = SyncSeqCst -- question is never written. Therefore we skip it where we can to -- save a few lines in the output and hopefully speed compilation up a -- bit. -funPrologue :: LiveGlobalRegs -> [CmmBlock] -> LlvmM StmtData +funPrologue :: LiveGlobalRegUses -> [CmmBlock] -> LlvmM StmtData funPrologue live cmmBlocks = do + platform <- getPlatform let getAssignedRegs :: CmmNode O O -> [CmmReg] getAssignedRegs (CmmAssign reg _) = [reg] @@ -2196,7 +2197,8 @@ funPrologue live cmmBlocks = do getAssignedRegs _ = [] getRegsBlock (_, body, _) = concatMap getAssignedRegs $ blockToList body assignedRegs = nub $ concatMap (getRegsBlock . blockSplit) cmmBlocks - isLive r = r `elem` alwaysLive || r `elem` live + mbLive r = + lookupRegUse r (alwaysLive platform) <|> lookupRegUse r live platform <- getPlatform stmtss <- forM assignedRegs $ \reg -> @@ -2205,12 +2207,12 @@ funPrologue live cmmBlocks = do let (newv, stmts) = allocReg reg varInsert un (pLower $ getVarType newv) return stmts - CmmGlobal (GlobalRegUse r _) -> do - let reg = lmGlobalRegVar platform r - arg = lmGlobalRegArg platform r + CmmGlobal ru@(GlobalRegUse r _) -> do + let reg = lmGlobalRegVar platform ru + arg = lmGlobalRegArg platform ru ty = (pLower . getVarType) reg trash = LMLitVar $ LMUndefLit ty - rval = if isLive r then arg else trash + rval = if isJust (mbLive r) then arg else trash alloc = Assignment reg $ Alloca (pLower $ getVarType reg) 1 markStackReg r return $ toOL [alloc, Store rval reg Nothing []] @@ -2222,7 +2224,7 @@ funPrologue live cmmBlocks = do -- | Function epilogue. Load STG variables to use as argument for call. -- STG Liveness optimisation done here. -funEpilogue :: LiveGlobalRegs -> LlvmM ([LlvmVar], LlvmStatements) +funEpilogue :: LiveGlobalRegUses -> LlvmM ([LlvmVar], LlvmStatements) funEpilogue live = do platform <- getPlatform @@ -2248,12 +2250,16 @@ funEpilogue live = do let allRegs = activeStgRegs platform loads <- forM allRegs $ \r -> if -- load live registers - | r `elem` alwaysLive -> loadExpr (GlobalRegUse r (globalRegSpillType platform r)) - | r `elem` live -> loadExpr (GlobalRegUse r (globalRegSpillType platform r)) + | Just ru <- lookupRegUse r (alwaysLive platform) + -> loadExpr ru + | Just ru <- lookupRegUse r live + -> loadExpr ru -- load all non Floating-Point Registers - | not (isFPR r) -> loadUndef r + | not (isFPR r) + -> loadUndef (GlobalRegUse r (globalRegSpillType platform r)) -- load padding Floating-Point Registers - | r `elem` paddingRegs -> loadUndef r + | Just ru <- lookupRegUse r paddingRegs + -> loadUndef ru | otherwise -> return (Nothing, nilOL) let (vars, stmts) = unzip loads @@ -2263,7 +2269,7 @@ funEpilogue live = do -- -- This is for Haskell functions, function type is assumed, so doesn't work -- with foreign functions. -getHsFunc :: LiveGlobalRegs -> CLabel -> LlvmM ExprData +getHsFunc :: LiveGlobalRegUses -> CLabel -> LlvmM ExprData getHsFunc live lbl = do fty <- llvmFunTy live name <- strCLabel_llvm lbl ===================================== compiler/GHC/CmmToLlvm/Ppr.hs ===================================== @@ -49,9 +49,8 @@ pprLlvmCmmDecl (CmmData _ lmdata) = do return ( vcat $ map (pprLlvmData opts) lmdata , vcat $ map (pprLlvmData opts) lmdata) -pprLlvmCmmDecl (CmmProc mb_info entry_lbl liveWithUses (ListGraph blks)) - = do let live = map globalRegUse_reg liveWithUses - lbl = case mb_info of +pprLlvmCmmDecl (CmmProc mb_info entry_lbl live (ListGraph blks)) + = do let lbl = case mb_info of Nothing -> entry_lbl Just (CmmStaticsRaw info_lbl _) -> info_lbl link = if externallyVisibleCLabel lbl ===================================== compiler/GHC/CmmToLlvm/Regs.hs ===================================== @@ -14,25 +14,27 @@ import GHC.Prelude import GHC.Llvm import GHC.Cmm.Expr +import GHC.CmmToAsm.Format import GHC.Platform import GHC.Data.FastString import GHC.Utils.Panic ( panic ) import GHC.Types.Unique + -- | Get the LlvmVar function variable storing the real register -lmGlobalRegVar :: Platform -> GlobalReg -> LlvmVar +lmGlobalRegVar :: Platform -> GlobalRegUse -> LlvmVar lmGlobalRegVar platform = pVarLift . lmGlobalReg platform "_Var" -- | Get the LlvmVar function argument storing the real register -lmGlobalRegArg :: Platform -> GlobalReg -> LlvmVar +lmGlobalRegArg :: Platform -> GlobalRegUse -> LlvmVar lmGlobalRegArg platform = lmGlobalReg platform "_Arg" {- Need to make sure the names here can't conflict with the unique generated names. Uniques generated names containing only base62 chars. So using say the '_' char guarantees this. -} -lmGlobalReg :: Platform -> String -> GlobalReg -> LlvmVar -lmGlobalReg platform suf reg +lmGlobalReg :: Platform -> String -> GlobalRegUse -> LlvmVar +lmGlobalReg platform suf (GlobalRegUse reg ty) = case reg of BaseReg -> ptrGlobal $ "Base" ++ suf Sp -> ptrGlobal $ "Sp" ++ suf @@ -88,13 +90,26 @@ lmGlobalReg platform suf reg ptrGlobal name = LMNLocalVar (fsLit name) (llvmWordPtr platform) floatGlobal name = LMNLocalVar (fsLit name) LMFloat doubleGlobal name = LMNLocalVar (fsLit name) LMDouble - xmmGlobal name = LMNLocalVar (fsLit name) (LMVector 4 (LMInt 32)) - ymmGlobal name = LMNLocalVar (fsLit name) (LMVector 8 (LMInt 32)) - zmmGlobal name = LMNLocalVar (fsLit name) (LMVector 16 (LMInt 32)) + fmt = cmmTypeFormat ty + xmmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + ymmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + zmmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + +formatLlvmType :: Format -> LlvmType +formatLlvmType II8 = LMInt 8 +formatLlvmType II16 = LMInt 16 +formatLlvmType II32 = LMInt 32 +formatLlvmType II64 = LMInt 64 +formatLlvmType FF32 = LMFloat +formatLlvmType FF64 = LMDouble +formatLlvmType (VecFormat l sFmt) = LMVector l (formatLlvmType $ scalarFormatFormat sFmt) -- | A list of STG Registers that should always be considered alive -alwaysLive :: [GlobalReg] -alwaysLive = [BaseReg, Sp, Hp, SpLim, HpLim, node] +alwaysLive :: Platform -> [GlobalRegUse] +alwaysLive platform = + [ GlobalRegUse r (globalRegSpillType platform r) + | r <- [BaseReg, Sp, Hp, SpLim, HpLim, node] + ] -- | STG Type Based Alias Analysis hierarchy stgTBAA :: [(Unique, LMString, Maybe Unique)] ===================================== testsuite/tests/simd/should_run/simd013.hs ===================================== @@ -8,25 +8,26 @@ module Main where import GHC.Exts import GHC.Prim -foreign import ccall "sub" +foreign import ccall unsafe "sub" sub :: DoubleX2# -> DoubleX2# -> DoubleX2# -foreign import ccall "add6" - add6 :: DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# +foreign import ccall unsafe "add7" + add7 :: DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# main :: IO () main = do let x1, x2 :: DoubleX2# x1 = packDoubleX2# (# 9.9##, 99.99## #) x2 = packDoubleX2# (# 1.1##, 11.11## #) - y1, y2, y3, y4, y5, y6 :: DoubleX2# - !y1 = packDoubleX2# (# 1.5##, 2.5## #) - !y2 = packDoubleX2# (# 10.25##, 20.25## #) - !y3 = packDoubleX2# (# 100.125##, 200.125## #) - !y4 = packDoubleX2# (# 1000.0625##, 2000.0625## #) - !y5 = packDoubleX2# (# 10000.03125##, 20000.03125## #) - !y6 = packDoubleX2# (# 100000.015625##, 200000.015625## #) + y1, y2, y3, y4, y5, y6, y7 :: DoubleX2# + !y1 = packDoubleX2# (# 1.5##, 2.5## #) + !y2 = packDoubleX2# (# 10.25##, 20.25## #) + !y3 = packDoubleX2# (# 100.125##, 200.125## #) + !y4 = packDoubleX2# (# 1000.0625##, 2000.0625## #) + !y5 = packDoubleX2# (# 10000.03125##, 20000.03125## #) + !y6 = packDoubleX2# (# 100000.015625##, 200000.015625## #) + !y7 = packDoubleX2# (# 1000000.0078125##, 2000000.0078125## #) !(# a, b #) = unpackDoubleX2# ( sub x1 x2 ) - !(# c, d #) = unpackDoubleX2# ( add6 y1 y2 y3 y4 y5 y6 ) + !(# c, d #) = unpackDoubleX2# ( add7 y1 y2 y3 y4 y5 y6 y7 ) print ( D# a, D# b ) print ( D# c, D# d ) ===================================== testsuite/tests/simd/should_run/simd013.stdout ===================================== @@ -1,2 +1,2 @@ (8.8,88.88) -(111111.984375,222222.984375) +(1111111.9921875,2222222.9921875) ===================================== testsuite/tests/simd/should_run/simd013C.c ===================================== @@ -1,12 +1,12 @@ -#include +#include __m128d sub(__m128d x, __m128d y) { return _mm_sub_pd(x,y); } -__m128d add6(__m128d x1, __m128d x2, __m128d x3, __m128d x4, __m128d x5, __m128d x6) +__m128d add7(__m128d x1, __m128d x2, __m128d x3, __m128d x4, __m128d x5, __m128d x6, __m128d x7) { - return _mm_add_pd(x1,_mm_add_pd(x2,_mm_add_pd(x3,_mm_add_pd(x4,_mm_add_pd(x5,x6))))); + return _mm_add_pd(x1,_mm_add_pd(x2,_mm_add_pd(x3,_mm_add_pd(x4,_mm_add_pd(x5,_mm_add_pd(x6, x7)))))); } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c27747f25c7209555ae96573f4525839297b7703...ad8c5c6410e818a4677894e0efcb14d7aa45a5e6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c27747f25c7209555ae96573f4525839297b7703...ad8c5c6410e818a4677894e0efcb14d7aa45a5e6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 15:26:27 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Fri, 20 Sep 2024 11:26:27 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/romes/exceptions-layout Message-ID: <66ed9423deccb_36d9d214e8170151078@gitlab.mail> Rodrigo Mesquita pushed new branch wip/romes/exceptions-layout at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/romes/exceptions-layout You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 15:43:46 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Fri, 20 Sep 2024 11:43:46 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 14 commits: The X86 SIMD patch. Message-ID: <66ed9832d3808_19acd718fb64753e4@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 1f052d49 by sheaf at 2024-09-20T17:42:52+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - bdcc4a82 by sheaf at 2024-09-20T17:42:57+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 73adb56a by sheaf at 2024-09-20T17:42:57+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 25dd47f4 by sheaf at 2024-09-20T17:42:57+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 1539cd58 by sheaf at 2024-09-20T17:42:57+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 98a35470 by sheaf at 2024-09-20T17:42:58+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 5ac32479 by sheaf at 2024-09-20T17:42:58+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 63ab9b97 by sheaf at 2024-09-20T17:42:58+02:00 Add test for C calls & SIMD vectors - - - - - 92447156 by sheaf at 2024-09-20T17:42:58+02:00 Add test for #25169 - - - - - 4bf648bc by sheaf at 2024-09-20T17:42:58+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 7665f74a by sheaf at 2024-09-20T17:43:32+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 8863864e by sheaf at 2024-09-20T17:43:35+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - 3cc298b9 by sheaf at 2024-09-20T17:43:35+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 98cf481c by sheaf at 2024-09-20T17:43:35+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ad8c5c6410e818a4677894e0efcb14d7aa45a5e6...98cf481c5c8a49ea14dd75c50a8c77d4f865b091 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ad8c5c6410e818a4677894e0efcb14d7aa45a5e6...98cf481c5c8a49ea14dd75c50a8c77d4f865b091 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 15:59:51 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Fri, 20 Sep 2024 11:59:51 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] timeouts Message-ID: <66ed9bf75923c_19acd71444e88356d@gitlab.mail> Sven Tennie pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 8279e766 by Sven Tennie at 2024-09-20T17:58:36+02:00 timeouts - - - - - 1 changed file: - .gitlab/jobs.yaml Changes: ===================================== .gitlab/jobs.yaml ===================================== @@ -60,7 +60,7 @@ "CROSS_EMULATOR": "qemu-riscv64 -L /usr/riscv64-linux-gnu", "CROSS_TARGET": "riscv64-linux-gnu", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", - "RUNTEST_ARGS": "", + "RUNTEST_ARGS": "--config='config.timeout=2000'", "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate", "XZ_OPT": "-9", "HADRIAN_ARGS": "--docs=none", @@ -127,7 +127,7 @@ "CROSS_EMULATOR": "qemu-riscv64 -L /usr/riscv64-linux-gnu", "CROSS_TARGET": "riscv64-linux-gnu", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", - "RUNTEST_ARGS": "", + "RUNTEST_ARGS": "--config='config.timeout=2000'", "TEST_ENV": "x86_64-linux-deb12-cross_riscv64-linux-gnu-validate", "HADRIAN_ARGS": "--docs=none", "EXTRA_HC_OPTS": "-fexternal-interpreter" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8279e7667fbb6133128f3eb74461608424533c08 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8279e7667fbb6133128f3eb74461608424533c08 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 20 22:25:54 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 20 Sep 2024 18:25:54 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 10 commits: base: Propagate `error` CallStack to thrown exception Message-ID: <66edf67222f0e_a272d3d7cc89964a@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 4cc9a07e by Ben Gamari at 2024-09-20T18:25:23-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes #24807. - - - - - cba79db1 by doyougnu at 2024-09-20T18:25:24-04:00 linker: add --optimistic-linking flag This patch adds: - the --optimistic-linking flag which binds unknown symbols in the runtime linker to 0xDEADBEEF instead of exiting with failure - The test T25240 which tests these flags using dead code in the FFI system. - closes #25240 This patch is part of the upstreaming haskell.nix patches project. - - - - - 4138f79e by Sebastian Graf at 2024-09-20T18:25:25-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 7a6cad41 by sheaf at 2024-09-20T18:25:26-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 4f7dd10a by sheaf at 2024-09-20T18:25:26-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - d75c58ff by Ryan Hendrickson at 2024-09-20T18:25:28-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - a8dcc787 by Zubin Duggal at 2024-09-20T18:25:28-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - c83d4179 by Andreas Klebinger at 2024-09-20T18:25:29-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c1238ab5 by Sven Tennie at 2024-09-20T18:25:30-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - ac820b38 by Sylvain Henry at 2024-09-20T18:25:42-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - 30 changed files: - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/extended_literals.rst - docs/users_guide/runtime_control.rst - libraries/base/changelog.md - libraries/base/tests/T19288.stderr - + libraries/base/tests/T24807.hs - + libraries/base/tests/T24807.stderr - libraries/base/tests/all.T - libraries/ghc-internal/src/GHC/Internal/Exception.hs - libraries/ghc-internal/src/GHC/Internal/IO/FD.hs - libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc - libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs - llvm-targets - m4/ghc_llvm_target.m4 - rts/Linker.c The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/60db6c2581d1e80d04b2a56576d54677f36db069...ac820b38f632fe88aa835aed61930ce8edf598c2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/60db6c2581d1e80d04b2a56576d54677f36db069...ac820b38f632fe88aa835aed61930ce8edf598c2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 04:43:49 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Sat, 21 Sep 2024 00:43:49 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/interpolated-strings Message-ID: <66ee4f05668d5_16c149bda3449071a@gitlab.mail> Brandon Chinn pushed new branch wip/interpolated-strings at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/interpolated-strings You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 04:48:18 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Sat, 21 Sep 2024 00:48:18 -0400 Subject: [Git][ghc/ghc] Deleted branch wip/multiline-strings-th Message-ID: <66ee501290d03_16c149beaeb0909d8@gitlab.mail> Brandon Chinn deleted branch wip/multiline-strings-th at Glasgow Haskell Compiler / GHC -- You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 04:48:46 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Sat, 21 Sep 2024 00:48:46 -0400 Subject: [Git][ghc/ghc] Deleted branch wip/multiline-strings Message-ID: <66ee502e9398d_16c149beaec491169@gitlab.mail> Brandon Chinn deleted branch wip/multiline-strings at Glasgow Haskell Compiler / GHC -- You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 08:47:19 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 04:47:19 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 11 commits: base: Propagate `error` CallStack to thrown exception Message-ID: <66ee8817abd3_1cf4ac585afc1156b3@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: e7b54e36 by Ben Gamari at 2024-09-21T04:46:50-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes #24807. - - - - - e20c202e by Sebastian Graf at 2024-09-21T04:46:51-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - f42b26dc by sheaf at 2024-09-21T04:46:52-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 799282f6 by sheaf at 2024-09-21T04:46:52-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - 51ab91bb by Ryan Hendrickson at 2024-09-21T04:46:53-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - 9fbd08f6 by Zubin Duggal at 2024-09-21T04:46:54-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - a80f4234 by Andreas Klebinger at 2024-09-21T04:46:55-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - 892fa884 by Hécate Kleidukos at 2024-09-21T04:46:59-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 2f28bf0e by Zubin Duggal at 2024-09-21T04:46:59-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 1f66f3b9 by Sven Tennie at 2024-09-21T04:47:00-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - 2404d801 by Sylvain Henry at 2024-09-21T04:47:03-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - 25 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - docs/users_guide/exts/extended_literals.rst - hadrian/bootstrap/bootstrap.py - hadrian/bootstrap/generate_bootstrap_plans - hadrian/bootstrap/plan-9_10_1.json - − hadrian/bootstrap/plan-9_4_1.json - − hadrian/bootstrap/plan-9_4_2.json The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ac820b38f632fe88aa835aed61930ce8edf598c2...2404d8013374ed12efc2a1a26ce1840529f614da -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ac820b38f632fe88aa835aed61930ce8edf598c2...2404d8013374ed12efc2a1a26ce1840529f614da You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 10:12:46 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sat, 21 Sep 2024 06:12:46 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] X86 genCCall: promote arg before calling evalArgs Message-ID: <66ee9c1ed6041_2fdb172f1c14223a4@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 13c949c5 by sheaf at 2024-09-21T12:12:37+02:00 X86 genCCall: promote arg before calling evalArgs - - - - - 1 changed file: - compiler/GHC/CmmToAsm/X86/CodeGen.hs Changes: ===================================== compiler/GHC/CmmToAsm/X86/CodeGen.hs ===================================== @@ -3243,9 +3243,12 @@ genCCall -> [CmmFormal] -> [CmmActual] -> NatM InstrBlock -genCCall bid addr conv dest_regs args = do +genCCall bid addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do + platform <- getPlatform + let args_hints = zip args (argHints ++ repeat NoHint) + prom_args = map (maybePromoteCArg platform W32) args_hints is32Bit <- is32BitPlatform - (instrs0, args') <- evalArgs bid args + (instrs0, args') <- evalArgs bid prom_args instrs1 <- if is32Bit then genCCall32 addr conv dest_regs args' else genCCall64 addr conv dest_regs args' @@ -3265,11 +3268,9 @@ genCCall32 :: CmmExpr -- ^ address of the function to call -> [CmmFormal] -- ^ where to put the result -> [CmmActual] -- ^ arguments (of mixed type) -> NatM InstrBlock -genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do +genCCall32 addr _conv dest_regs args = do config <- getConfig let platform = ncgPlatform config - args_hints = zip args (argHints ++ repeat NoHint) - prom_args = map (maybePromoteCArg platform W32) args_hints -- If the size is smaller than the word, we widen things (see maybePromoteCArg) arg_size_bytes :: CmmType -> Int @@ -3339,7 +3340,7 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do delta0 <- getDeltaNat setDeltaNat (delta0 - arg_pad_size) - push_codes <- mapM push_arg (reverse prom_args) + push_codes <- mapM push_arg (reverse args) delta <- getDeltaNat massert (delta == delta0 - tot_arg_size) @@ -3414,11 +3415,9 @@ genCCall64 :: CmmExpr -- ^ address of function to call -> [CmmFormal] -- ^ where to put the result -> [CmmActual] -- ^ arguments (of mixed type) -> NatM InstrBlock -genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do +genCCall64 addr conv dest_regs args = do config <- getConfig let platform = ncgPlatform config - args_hints = zip args (argHints ++ repeat NoHint) - prom_args = map (maybePromoteCArg platform W32) args_hints word_size = platformWordSizeInBytes platform wordFmt = archWordFormat (target32Bit platform) @@ -3434,7 +3433,7 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do , usedRegs = arg_regs_used , assignArgsCode = assign_args_code } - <- loadArgs config prom_args + <- loadArgs config args let View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/13c949c5140994c7499551f105d8d7c85315470f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/13c949c5140994c7499551f105d8d7c85315470f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 10:40:36 2024 From: gitlab at gitlab.haskell.org (doyougnu (@doyougnu)) Date: Sat, 21 Sep 2024 06:40:36 -0400 Subject: [Git][ghc/ghc][wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef] linker: add --optimistic-linking flag Message-ID: <66eea2a444f43_3b6a25bc55c5389@gitlab.mail> doyougnu pushed to branch wip/haskell-nix-patches/musl64/ghc-9.6-missing-symbols-deadbeef at Glasgow Haskell Compiler / GHC Commits: bc903950 by doyougnu at 2024-09-21T06:40:04-04:00 linker: add --optimistic-linking flag This patch adds: - the --optimistic-linking flag which binds unknown symbols in the runtime linker to 0xDEADBEEF instead of exiting with failure - The test T25240 which tests these flags using dead code in the FFI system. - closes #25240 This patch is part of the upstreaming haskell.nix patches project. - - - - - 23 changed files: - docs/users_guide/9.12.1-notes.rst - docs/users_guide/runtime_control.rst - libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc - rts/Linker.c - rts/RtsFlags.c - rts/include/rts/Flags.h - rts/linker/elf_got.c - + testsuite/tests/ghci/linking/T25240/Makefile - + testsuite/tests/ghci/linking/T25240/T25240.hs - + testsuite/tests/ghci/linking/T25240/T25240.stderr - + testsuite/tests/ghci/linking/T25240/T25240a.hs - + testsuite/tests/ghci/linking/T25240/all.T - testsuite/tests/ghci/should_run/T18064.stderr - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/interface-stability/base-exports.stdout-ws-32 - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-32-mingw32 - testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-64-mingw32 - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-32-mingw32 - testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-64-mingw32 Changes: ===================================== docs/users_guide/9.12.1-notes.rst ===================================== @@ -147,6 +147,10 @@ Runtime system For this reason, a warning is emitted if a .tix file is read in implicitly. You can silence this warning by explicitly passing ``--read-tix-file=yes``. Details can be found in `GHC proposal 612 `__. +- Add new runtime flag :rts-flag:`--optimistic-linking` which instructs the + runtime linker to continue in the presence of unknown symbols. By default this + flag is not passed, preserving previous behavior. + ``base`` library ~~~~~~~~~~~~~~~~ ===================================== docs/users_guide/runtime_control.rst ===================================== @@ -397,6 +397,11 @@ Miscellaneous RTS options heap larger than 1T. ``-xr`` is a no-op if GHC is configured with ``--disable-large-address-space`` or if the platform is 32-bit. +.. rts-flag:: --optimistic-linking + + If given, instruct the runtime linker to try to continue linking in the + presence of an unresolved symbol. + .. _rts-options-gc: RTS options to control the garbage collector ===================================== libraries/ghc-internal/src/GHC/Internal/RTS/Flags.hsc ===================================== @@ -162,6 +162,7 @@ data MiscFlags = MiscFlags , disableDelayedOsMemoryReturn :: Bool , internalCounters :: Bool , linkerAlwaysPic :: Bool + , linkerOptimistic :: Bool , linkerMemBase :: Word -- ^ address to ask the OS for memory for the linker, 0 ==> off , ioManager :: IoManagerFlag @@ -536,6 +537,8 @@ getMiscFlags = do (#{peek MISC_FLAGS, internalCounters} ptr :: IO CBool)) <*> (toBool <$> (#{peek MISC_FLAGS, linkerAlwaysPic} ptr :: IO CBool)) + <*> (toBool <$> + (#{peek MISC_FLAGS, linkerOptimistic} ptr :: IO CBool)) <*> #{peek MISC_FLAGS, linkerMemBase} ptr <*> (toEnum . fromIntegral <$> (#{peek MISC_FLAGS, ioManager} ptr :: IO Word32)) ===================================== rts/Linker.c ===================================== @@ -967,10 +967,20 @@ SymbolAddr* lookupSymbol( SymbolName* lbl ) // lookupDependentSymbol directly. SymbolAddr* r = lookupDependentSymbol(lbl, NULL, NULL); if (!r) { - errorBelch("^^ Could not load '%s', dependency unresolved. " - "See top entry above.\n", lbl); - IF_DEBUG(linker, printLoadedObjects()); - fflush(stderr); + if (!RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("^^ Could not load '%s', dependency unresolved. " + "See top entry above. You might consider using --optimistic-linking\n", + lbl); + IF_DEBUG(linker, printLoadedObjects()); + fflush(stderr); + } else { + // if --optimistic-linking is passed into the RTS we allow the linker + // to optimistically continue + errorBelch("^^ Could not load '%s', dependency unresolved, " + "optimistically continuing\n", + lbl); + r = (void*) 0xDEADBEEF; + } } if (!runPendingInitializers()) { ===================================== rts/RtsFlags.c ===================================== @@ -269,6 +269,7 @@ void initRtsFlagsDefaults(void) RtsFlags.MiscFlags.disableDelayedOsMemoryReturn = false; RtsFlags.MiscFlags.internalCounters = false; RtsFlags.MiscFlags.linkerAlwaysPic = DEFAULT_LINKER_ALWAYS_PIC; + RtsFlags.MiscFlags.linkerOptimistic = false; RtsFlags.MiscFlags.linkerMemBase = 0; RtsFlags.MiscFlags.ioManager = IO_MNGR_FLAG_AUTO; #if defined(THREADED_RTS) && defined(mingw32_HOST_OS) @@ -998,6 +999,11 @@ error = true; OPTION_UNSAFE; RtsFlags.MiscFlags.generate_dump_file = true; } + else if (strequal("optimistic-linking", + &rts_argv[arg][2])) { + OPTION_UNSAFE; + RtsFlags.MiscFlags.linkerOptimistic = true; + } else if (strequal("null-eventlog-writer", &rts_argv[arg][2])) { OPTION_UNSAFE; ===================================== rts/include/rts/Flags.h ===================================== @@ -267,6 +267,7 @@ typedef struct _MISC_FLAGS { there as well. */ bool internalCounters; /* See Note [Internal Counters Stats] */ bool linkerAlwaysPic; /* Assume the object code is always PIC */ + bool linkerOptimistic; /* Should the runtime linker optimistically continue */ StgWord linkerMemBase; /* address to ask the OS for memory * for the linker, NULL ==> off */ IO_MANAGER_FLAG ioManager; /* The I/O manager to use. */ ===================================== rts/linker/elf_got.c ===================================== @@ -97,9 +97,22 @@ fillGot(ObjectCode * oc) { if(0 == strncmp(symbol->name,"_GLOBAL_OFFSET_TABLE_",21)) { symbol->addr = oc->info->got_start; } else { - errorBelch("Failed to lookup symbol: %s\n", + errorBelch("Failed to lookup symbol: %s," + " you might consider using --optimistic-linking\n", symbol->name); - return EXIT_FAILURE; + + // if --optimistic-linking is passed into the + // RTS we allow the linker to optimistically + // continue + if (RtsFlags.MiscFlags.linkerOptimistic) { + errorBelch("Failed to lookup symbol: %s," + " optimistically continuing.\n", + symbol->name); + symbol->addr = (void*) 0xDEADBEEF; + } else { + return EXIT_FAILURE; + } + } } } else { ===================================== testsuite/tests/ghci/linking/T25240/Makefile ===================================== @@ -0,0 +1,7 @@ +TOP=../../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +.PHONY: T25240 +T25240: + "$(TEST_HC)" $(TEST_HC_OPTS_INTERACTIVE) T25240a.hs T25240.hs +RTS --optimistic-linking -RTS ===================================== testsuite/tests/ghci/linking/T25240/T25240.hs ===================================== @@ -0,0 +1,17 @@ +{-# LANGUAGE TemplateHaskell #-} + +{- + +When the RTS linker loads the T25240a module to run the pure foo splice, it +tries to resolve the func symbol even if this function isn't required to run the +splice code, i.e., its dead code. This test checks that by passing the +--optimistic-linking flag the RTS linker continues to link even in the presence +of unknown symbols. + +-} + +module T25240 where + +import T25240a + +$(pure foo) ===================================== testsuite/tests/ghci/linking/T25240/T25240.stderr ===================================== @@ -0,0 +1,2 @@ +ghc: ^^ Could not load 'func', dependency unresolved, optimistically continuing + ===================================== testsuite/tests/ghci/linking/T25240/T25240a.hs ===================================== @@ -0,0 +1,13 @@ +{-# LANGUAGE TemplateHaskell #-} + + +module T25240a + ( foo, func + ) where + + +foo :: [a] +foo = [] + +foreign import ccall "func" + func :: Int -> Int ===================================== testsuite/tests/ghci/linking/T25240/all.T ===================================== @@ -0,0 +1,3 @@ +# skip on darwin because the leading underscores will make the test fail +test('T25240', [when(leading_underscore(),skip), req_rts_linker, extra_files(['T25240a.hs'])], + makefile_test, ['T25240']) ===================================== testsuite/tests/ghci/should_run/T18064.stderr ===================================== @@ -1,2 +1,2 @@ -: ^^ Could not load 'blah', dependency unresolved. See top entry above. +: ^^ Could not load 'blah', dependency unresolved. See top entry above. You might consider using --optimistic-linking ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -234,7 +234,7 @@ module Control.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -303,7 +303,7 @@ module Control.Exception where module Control.Exception.Annotation where -- Safety: None type ExceptionAnnotation :: * -> Constraint - class ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where + class ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where displayExceptionAnnotation :: a -> GHC.Internal.Base.String default displayExceptionAnnotation :: GHC.Internal.Show.Show a => a -> GHC.Internal.Base.String {-# MINIMAL #-} @@ -345,7 +345,7 @@ module Control.Exception.Base where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -850,11 +850,11 @@ module Data.Data where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b constrFields :: Constr -> [GHC.Internal.Base.String] constrFixity :: Constr -> Fixity @@ -897,7 +897,7 @@ module Data.Data where showConstr :: Constr -> GHC.Internal.Base.String showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -921,14 +921,14 @@ module Data.Dynamic where -- Safety: Safe type Dynamic :: * data Dynamic where - Dynamic :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic + Dynamic :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} dynApp :: Dynamic -> Dynamic -> Dynamic dynApply :: Dynamic -> Dynamic -> GHC.Internal.Maybe.Maybe Dynamic - dynTypeRep :: Dynamic -> ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + dynTypeRep :: Dynamic -> ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep fromDyn :: forall a. Typeable a => Dynamic -> a -> a fromDynamic :: forall a. Typeable a => Dynamic -> GHC.Internal.Maybe.Maybe a toDyn :: forall a. Typeable a => a -> Dynamic @@ -1815,11 +1815,11 @@ module Data.Typeable where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b decT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Data.Either.Either ((a :~: b) -> GHC.Internal.Base.Void) (a :~: b) eqT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Maybe.Maybe (a :~: b) @@ -1834,7 +1834,7 @@ module Data.Typeable where rnfTypeRep :: TypeRep -> () showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -5255,7 +5255,7 @@ module GHC.Event where data EventManager = ... type FdKey :: * data FdKey - = ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] + = ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] ; GHC.Internal.Foreign.C.Types.N:CInt[0])GHC.Internal.System.Posix.Types.Fd, ...} type IOCallback :: * @@ -5288,9 +5288,9 @@ module GHC.Event.TimeOut where type TimeoutEdit :: * type TimeoutEdit = TimeoutQueue -> TimeoutQueue type TimeoutKey :: * - newtype TimeoutKey = TK ghc-internal-9.1001.0:GHC.Internal.Event.Unique.Unique + newtype TimeoutKey = TK ghc-internal-9.1100.0:GHC.Internal.Event.Unique.Unique type TimeoutQueue :: * - type TimeoutQueue = ghc-internal-9.1001.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback + type TimeoutQueue = ghc-internal-9.1100.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback module GHC.Exception where -- Safety: Safe @@ -5302,7 +5302,7 @@ module GHC.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -5331,7 +5331,7 @@ module GHC.Exception.Type where type ArithException :: * data ArithException = Overflow | Underflow | LossOfPrecision | DivideByZero | Denormal | RatioZeroDenominator type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -7884,8 +7884,8 @@ module GHC.IO.Handle where hTryLock :: Handle -> LockMode -> GHC.Types.IO GHC.Types.Bool hWaitForInput :: Handle -> GHC.Types.Int -> GHC.Types.IO GHC.Types.Bool isEOF :: GHC.Types.IO GHC.Types.Bool - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle nativeNewline :: Newline nativeNewlineMode :: NewlineMode noNewlineTranslation :: NewlineMode @@ -7935,11 +7935,11 @@ module GHC.IO.Handle.Internals where ioe_notReadable :: forall a. GHC.Types.IO a ioe_notWritable :: forall a. GHC.Types.IO a ioe_semiclosedHandle :: forall a. GHC.Types.IO a - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle openTextEncoding :: forall a. GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.HandleType -> (forall es ds. GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextEncoder es) -> GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextDecoder ds) -> GHC.Types.IO a) -> GHC.Types.IO a readTextDevice :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer readTextDeviceNonBlocking :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer @@ -8003,7 +8003,7 @@ module GHC.IO.Handle.Types where type Handle__ :: * data Handle__ = forall dev enc_state dec_state. - (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => + (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => Handle__ {haDevice :: !dev, haType :: HandleType, haByteBuffer :: ! {-# UNPACK #-}(GHC.Internal.IORef.N:IORef[0] _N)(GHC.Internal.IORef.IORef (GHC.Internal.IO.Buffer.Buffer GHC.Internal.Word.Word8)), @@ -9106,7 +9106,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * @@ -9417,7 +9417,7 @@ module GHC.StaticPtr where -- Safety: None type IsStatic :: (* -> *) -> Constraint class IsStatic p where - fromStaticPtr :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a + fromStaticPtr :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a {-# MINIMAL fromStaticPtr #-} type StaticKey :: * type StaticKey = GHC.Internal.Fingerprint.Type.Fingerprint @@ -10813,8 +10813,8 @@ module Type.Reflection where data (:~~:) a b where HRefl :: forall {k1} (a :: k1). (:~~:) a a pattern App :: forall k2 (t :: k2). () => forall k1 (a :: k1 -> k2) (b :: k1). (t ~ a b) => TypeRep a -> TypeRep b -> TypeRep t - pattern Con :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a - pattern Con' :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a + pattern Con :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a + pattern Con' :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a pattern Fun :: forall k (fun :: k). () => forall (r1 :: GHC.Types.RuntimeRep) (r2 :: GHC.Types.RuntimeRep) (arg :: TYPE r1) (res :: TYPE r2). (k ~ *, fun ~~ (arg -> res)) => TypeRep arg -> TypeRep res -> TypeRep fun type Module :: * data Module = ... @@ -10831,7 +10831,7 @@ module Type.Reflection where type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} decTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Data.Either.Either ((a :~~: b) -> GHC.Internal.Base.Void) (a :~~: b) eqTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Maybe.Maybe (a :~~: b) moduleName :: Module -> GHC.Internal.Base.String @@ -10866,9 +10866,9 @@ module Type.Reflection.Unsafe where data TypeRep a where ... mkTrApp :: forall k1 k2 (a :: k1 -> k2) (b :: k1). TypeRep a -> TypeRep b -> TypeRep (a b) - mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a + mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a mkTyCon :: GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Types.Int -> KindRep -> TyCon - someTypeRepFingerprint :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint + someTypeRepFingerprint :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConKindArgs :: TyCon -> GHC.Types.Int tyConKindRep :: TyCon -> KindRep @@ -11202,9 +11202,9 @@ instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.I instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Monoid (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Monoid (f p), GHC.Internal.Base.Monoid (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Monoid (f (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Monoid (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Monoid (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11260,9 +11260,9 @@ instance forall a. GHC.Internal.Base.Semigroup (Data.Semigroup.Last a) -- Define instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Semigroup (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Semigroup (f p), GHC.Internal.Base.Semigroup (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Semigroup (f (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11392,20 +11392,20 @@ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Last -- Defined instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Max -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Min -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.IO.Class.MonadIO GHC.Types.IO -- Defined in ‘GHC.Internal.Control.Monad.IO.Class’ -instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ -instance forall (m :: * -> *) a. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ +instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ +instance forall (m :: * -> *) a. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Functor.ZipList.ZipList a) -- Defined in ‘GHC.Internal.Functor.ZipList’ instance GHC.Internal.Data.Data.Data Data.Array.Byte.ByteArray -- Defined in ‘Data.Array.Byte’ -instance forall s. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ +instance forall s. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Complex.Complex a) -- Defined in ‘Data.Complex’ -instance forall i j (a :: i) (b :: j). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall i j (a :: i) (b :: j). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.All -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.Any -- Defined in ‘GHC.Internal.Data.Data’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b, GHC.Internal.Ix.Ix a) => GHC.Internal.Data.Data.Data (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.Associativity -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Bool -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Char -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k a (b :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall k a (b :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.DecidedStrictness -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Double -- Defined in ‘GHC.Internal.Data.Data’ @@ -11453,10 +11453,10 @@ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word32 -- Defined in ‘G instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word64 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word8 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Foreign.Ptr.WordPtr -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ -instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ +instance forall k (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ +instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b) => GHC.Internal.Data.Data.Data (Data.Semigroup.Arg a b) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.First a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.Last a) -- Defined in ‘Data.Semigroup’ @@ -11522,7 +11522,7 @@ instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SCha instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SSymbol -- Defined in ‘GHC.Internal.TypeLits’ instance forall k (a :: k). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k (a :: k1). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ -instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1). GHC.Internal.Data.Type.Equality.TestEquality f => GHC.Internal.Data.Type.Equality.TestEquality (Data.Functor.Compose.Compose f g) -- Defined in ‘Data.Functor.Compose’ instance forall a k (b :: k). GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ instance forall a. GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Bits.And a) -- Defined in ‘GHC.Internal.Data.Bits’ @@ -11719,7 +11719,7 @@ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Ba instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.RecUpdError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.TypeError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ -instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ instance [safe] GHC.Internal.Exception.Type.Exception System.Timeout.Timeout -- Defined in ‘System.Timeout’ instance forall a k (b :: k). GHC.Internal.Float.Floating a => GHC.Internal.Float.Floating (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ @@ -12390,8 +12390,8 @@ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataRep -- Defined in ‘ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataType -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Internal.Show.Show (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance GHC.Internal.Show.Show GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ instance forall a b. (GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). Data.Fixed.HasResolution a => GHC.Internal.Show.Show (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ @@ -12462,13 +12462,13 @@ instance GHC.Internal.Show.Show GHC.Internal.Foreign.C.Types.CWchar -- Defined i instance forall a. GHC.Internal.Show.Show (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Foreign.C.ConstPtr’ instance forall a b. (GHC.Internal.Ix.Ix a, GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Internal.Show.Show GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Internal.Show.Show GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ instance GHC.Internal.Show.Show GHC.Types.Double -- Defined in ‘GHC.Internal.Float’ instance GHC.Internal.Show.Show GHC.Types.Float -- Defined in ‘GHC.Internal.Float’ @@ -12502,7 +12502,7 @@ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Handle -- Defined i instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.HandleType -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Newline -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.NewlineMode -- Defined in ‘GHC.Internal.IO.Handle.Types’ -instance [safe] GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.HandlePosn -- Defined in ‘GHC.Internal.IO.Handle’ instance GHC.Internal.Show.Show GHC.Internal.IO.IOMode.IOMode -- Defined in ‘GHC.Internal.IO.IOMode’ instance GHC.Internal.Show.Show GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ @@ -12599,8 +12599,8 @@ instance GHC.Classes.Eq GHC.Internal.Data.Data.ConstrRep -- Defined in ‘GHC.In instance GHC.Classes.Eq GHC.Internal.Data.Data.DataRep -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Classes.Eq GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Classes.Eq (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Eq a, GHC.Classes.Eq b) => GHC.Classes.Eq (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Eq (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Eq (f (g a)) => GHC.Classes.Eq (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ @@ -12676,13 +12676,13 @@ instance forall a. GHC.Classes.Eq (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) - instance forall i e. (GHC.Internal.Ix.Ix i, GHC.Classes.Eq e) => GHC.Classes.Eq (GHC.Internal.Arr.Array i e) -- Defined in ‘GHC.Internal.Arr’ instance forall s i e. GHC.Classes.Eq (GHC.Internal.Arr.STArray s i e) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Classes.Eq GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ instance GHC.Classes.Eq GHC.Internal.Event.TimeOut.TimeoutKey -- Defined in ‘GHC.Internal.Event.TimeOut’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Classes.Eq GHC.Internal.Stack.Types.SrcLoc -- Defined in ‘GHC.Internal.Stack.Types’ instance GHC.Classes.Eq GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ instance GHC.Classes.Eq GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ @@ -12778,8 +12778,8 @@ instance GHC.Classes.Ord GHC.Internal.Unicode.GeneralCategory -- Defined in ‘G instance forall k (a :: k) (b :: k). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k2 (a :: k1) (b :: k2). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k (s :: k). GHC.Classes.Ord (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Ord ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Ord ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Ord a, GHC.Classes.Ord b) => GHC.Classes.Ord (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Ord (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Ord (f (g a)) => GHC.Classes.Ord (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ ===================================== testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs ===================================== @@ -12148,7 +12148,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -9330,7 +9330,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * ===================================== testsuite/tests/interface-stability/base-exports.stdout-ws-32 ===================================== @@ -9106,7 +9106,7 @@ module GHC.RTS.Flags where type IoSubSystem :: * data IoSubSystem = IoPOSIX | IoNative type MiscFlags :: * - data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} + data MiscFlags = MiscFlags {tickInterval :: RtsTime, installSignalHandlers :: GHC.Types.Bool, installSEHHandlers :: GHC.Types.Bool, generateCrashDumpFile :: GHC.Types.Bool, generateStackTrace :: GHC.Types.Bool, machineReadable :: GHC.Types.Bool, disableDelayedOsMemoryReturn :: GHC.Types.Bool, internalCounters :: GHC.Types.Bool, linkerAlwaysPic :: GHC.Types.Bool, linkerOptimistic :: GHC.Types.Bool, linkerMemBase :: GHC.Types.Word, ioManager :: IoManagerFlag, numIoWorkerThreads :: GHC.Internal.Word.Word32} type ParFlags :: * data ParFlags = ParFlags {nCapabilities :: GHC.Internal.Word.Word32, migrate :: GHC.Types.Bool, maxLocalSparks :: GHC.Internal.Word.Word32, parGcEnabled :: GHC.Types.Bool, parGcGen :: GHC.Internal.Word.Word32, parGcLoadBalancingEnabled :: GHC.Types.Bool, parGcLoadBalancingGen :: GHC.Internal.Word.Word32, parGcNoSyncWithIdle :: GHC.Internal.Word.Word32, parGcThreads :: GHC.Internal.Word.Word32, setAffinity :: GHC.Types.Bool} type ProfFlags :: * ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-32-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_link_order_a_b_2_fail.stderr-ws-64-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-32-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load '_c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label ===================================== testsuite/tests/rts/linker/T11223/T11223_simple_duplicate_lib.stderr-ws-64-mingw32 ===================================== @@ -9,7 +9,7 @@ This could be caused by: * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. -ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. +ghc-stage2.exe: ^^ Could not load 'c', dependency unresolved. See top entry above. You might consider using --optimistic-linking GHC.ByteCode.Linker: can't find label View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/bc9039508debd5a9398a9e8f77d1df8a924c5f05 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/bc9039508debd5a9398a9e8f77d1df8a924c5f05 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 10:48:10 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sat, 21 Sep 2024 06:48:10 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 2 commits: Revert "X86 genCCall64: simplify loadArg code" Message-ID: <66eea46a1d6fd_3b6a2524799494af@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 2a02aea3 by sheaf at 2024-09-21T12:25:11+02:00 Revert "X86 genCCall64: simplify loadArg code" This reverts commit 3cc298b9343218d49507c2f669c9303e58e115b5. - - - - - 7f962063 by sheaf at 2024-09-21T12:47:58+02:00 WIP pushArgByValue - - - - - 1 changed file: - compiler/GHC/CmmToAsm/X86/CodeGen.hs Changes: ===================================== compiler/GHC/CmmToAsm/X86/CodeGen.hs ===================================== @@ -1965,17 +1965,6 @@ getRegister' platform is32Bit (CmmLit lit) getRegister' platform is32Bit (CmmLit lit) = do avx <- avxEnabled - -- NB: it is important that the code produced here (to load a literal into - -- a register) doesn't clobber any registers other than the destination - -- register; the code for generating C calls relies on this property. - -- - -- In particular, we have: - -- - -- > loadIntoRegMightClobberOtherReg (CmmLit _) = False - -- - -- which means that we assume that loading a literal into a register - -- will not clobber any other registers. - -- TODO: this function mishandles floating-point negative zero, -- because -0.0 == 0.0 returns True and because we represent CmmFloat as -- Rational, which can't properly represent negative zero. @@ -3091,8 +3080,10 @@ genSimplePrim _ op dst args = do platform <- ncgPlatform <$> getConfig pprPanic "genSimplePrim: unhandled primop" (ppr (pprCallishMachOp op, dst, fmap (pdoc platform) args)) -{- Note [Evaluate C-call arguments before placing in destination registers] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +{- +Note [Evaluate C-call arguments before placing in destination registers] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + When producing code for C calls we must take care when placing arguments in their final registers. Specifically, we must ensure that temporary register usage due to evaluation of one argument does not clobber a register in which we @@ -3143,11 +3134,15 @@ genForeignCall{32,64}. -- | See Note [Evaluate C-call arguments before placing in destination registers] evalArgs :: BlockId -> [CmmActual] -> NatM (InstrBlock, [CmmActual]) evalArgs bid actuals - | any loadIntoRegMightClobberOtherReg actuals = do + | any mightContainMachOp actuals = do regs_blks <- mapM evalArg actuals return (concatOL $ map fst regs_blks, map snd regs_blks) | otherwise = return (nilOL, actuals) where + mightContainMachOp (CmmReg _) = False + mightContainMachOp (CmmRegOff _ _) = False + mightContainMachOp (CmmLit _) = False + mightContainMachOp _ = True evalArg :: CmmActual -> NatM (InstrBlock, CmmExpr) evalArg actual = do @@ -3161,16 +3156,6 @@ evalArgs bid actuals newLocalReg :: CmmType -> NatM LocalReg newLocalReg ty = LocalReg <$> getUniqueM <*> pure ty --- | Might the code to put this expression into a register --- clobber any other registers? -loadIntoRegMightClobberOtherReg :: CmmExpr -> Bool -loadIntoRegMightClobberOtherReg (CmmReg _) = False -loadIntoRegMightClobberOtherReg (CmmRegOff _ _) = False -loadIntoRegMightClobberOtherReg (CmmLit _) = False - -- NB: this last 'False' is slightly risky, because the code for loading - -- a literal into a register is not entirely trivial. -loadIntoRegMightClobberOtherReg _ = True - -- Note [DIV/IDIV for bytes] -- ~~~~~~~~~~~~~~~~~~~~~~~~~ -- IDIV reminder: @@ -3245,9 +3230,10 @@ genCCall -> NatM InstrBlock genCCall bid addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do platform <- getPlatform - let args_hints = zip args (argHints ++ repeat NoHint) - prom_args = map (maybePromoteCArg platform W32) args_hints is32Bit <- is32BitPlatform + let args_hints = zip args (argHints ++ repeat NoHint) + prom_width = if is32Bit then W32 else W64 + prom_args = map (maybePromoteCArg platform prom_width) args_hints (instrs0, args') <- evalArgs bid prom_args instrs1 <- if is32Bit then genCCall32 addr conv dest_regs args' @@ -3256,12 +3242,17 @@ genCCall bid addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do maybePromoteCArg :: Platform -> Width -> (CmmExpr, ForeignHint) -> CmmExpr maybePromoteCArg platform wto (arg, hint) - | wfrom < wto = case hint of - SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] - _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] + | wfrom < wto = + if isFloatType ty + then CmmMachOp (MO_FF_Conv wfrom wto) [arg] + else + case hint of + SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] + _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] | otherwise = arg where - wfrom = cmmExprWidth platform arg + ty = cmmExprType platform arg + wfrom = typeWidth ty genCCall32 :: CmmExpr -- ^ address of the function to call -> ForeignConvention -- ^ calling convention @@ -3431,6 +3422,7 @@ genCCall64 addr conv dest_regs args = do { stackArgs = proper_stack_args , stackDataArgs = stack_data_args , usedRegs = arg_regs_used + , computeArgsCode = compute_args_code , assignArgsCode = assign_args_code } <- loadArgs config args @@ -3529,6 +3521,7 @@ genCCall64 addr conv dest_regs args = do return (align_call_code `appOL` push_code `appOL` + compute_args_code `appOL` assign_args_code `appOL` load_data_refs `appOL` shadow_space_code `appOL` @@ -3549,14 +3542,16 @@ data LoadArgs , stackDataArgs :: [CmmExpr] -- | Which registers are we using for argument passing? , usedRegs :: [RegWithFormat] + -- | The code to compute arguments into (possibly temporary) registers. + , computeArgsCode :: InstrBlock -- | The code to assign arguments to registers used for argument passing. , assignArgsCode :: InstrBlock } instance Semigroup LoadArgs where - LoadArgs a1 d1 r1 j1 <> LoadArgs a2 d2 r2 j2 - = LoadArgs (a1 ++ a2) (d1 ++ d2) (r1 ++ r2) (j1 S.<> j2) + LoadArgs a1 d1 r1 i1 j1 <> LoadArgs a2 d2 r2 i2 j2 + = LoadArgs (a1 ++ a2) (d1 ++ d2) (r1 ++ r2) (i1 S.<> i2) (j1 S.<> j2) instance Monoid LoadArgs where - mempty = LoadArgs [] [] [] nilOL + mempty = LoadArgs [] [] [] nilOL nilOL -- | An argument passed on the stack, either directly or by reference. -- @@ -3715,6 +3710,7 @@ loadArgsSysV config (arg:rest) = do LoadArgs { stackArgs = map RawStackArg (arg:rest) , stackDataArgs = [] + , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3734,11 +3730,12 @@ loadArgsSysV config (arg:rest) = do this_arg <- case mbReg of Just reg -> do - assign_code <- lift $ loadArgIntoReg arg reg + (compute_code, assign_code) <- lift $ loadArgIntoReg config arg rest reg return $ LoadArgs { stackArgs = [] -- passed in register , stackDataArgs = [] + , computeArgsCode = compute_code , assignArgsCode = assign_code , usedRegs = [RegWithFormat reg arg_fmt] } @@ -3748,6 +3745,7 @@ loadArgsSysV config (arg:rest) = do LoadArgs { stackArgs = [RawStackArg arg] , stackDataArgs = [] + , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3799,6 +3797,7 @@ loadArgsWin config (arg:rest) = do LoadArgs { stackArgs = stk_args , stackDataArgs = data_args + , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3814,7 +3813,8 @@ loadArgsWin config (arg:rest) = do -- Pass the reference in a register, -- and the argument data on the stack. { stackArgs = [RawStackArgRef (InReg ireg) (argSize platform arg)] - , stackDataArgs = [arg] -- we don't yet know where the data will reside, + , stackDataArgs = [arg] + , computeArgsCode = nilOL -- we don't yet know where the data will reside, , assignArgsCode = nilOL -- so we defer computing the reference and storing it -- in the register until later , usedRegs = [RegWithFormat ireg II64] @@ -3825,7 +3825,7 @@ loadArgsWin config (arg:rest) = do = freg | otherwise = ireg - assign_code <- loadArgIntoReg arg arg_reg + (compute_code, assign_code) <- loadArgIntoReg config arg rest arg_reg -- Recall that, for varargs, we must pass floating-point -- arguments in both fp and integer registers. let (assign_code', regs') @@ -3838,23 +3838,42 @@ loadArgsWin config (arg:rest) = do LoadArgs { stackArgs = [] -- passed in register , stackDataArgs = [] + , computeArgsCode = compute_code , assignArgsCode = assign_code' , usedRegs = regs' } --- | Load an argument into a register. + +-- | Return two pieces of code: -- --- Assumes that the expression does not contain any MachOps, --- as per Note [Evaluate C-call arguments before placing in destination registers]. -loadArgIntoReg :: CmmExpr -> Reg -> NatM InstrBlock -loadArgIntoReg arg reg = do - when (debugIsOn && loadIntoRegMightClobberOtherReg arg) $ do - platform <- getPlatform - massertPpr False $ - vcat [ text "loadArgIntoReg: arg might contain MachOp" - , text "arg:" <+> pdoc platform arg ] - arg_code <- getAnyReg arg - return $ arg_code reg +-- - code to compute a the given 'CmmExpr' into some (possibly temporary) register +-- - code to assign the resulting value to the specified register +-- +-- Using two separate pieces of code handles clobbering issues reported +-- in e.g. #11792, #12614. +loadArgIntoReg :: NCGConfig -> CmmExpr -> [CmmExpr] -> Reg -> NatM (InstrBlock, InstrBlock) +loadArgIntoReg config arg rest reg + -- "operand" args can be directly assigned into the register + | isOperand platform arg + = do arg_code <- getAnyReg arg + return (nilOL, arg_code reg) + -- The last non-operand arg can be directly assigned after its + -- computation without going into a temporary register + | all (isOperand platform) rest + = do arg_code <- getAnyReg arg + return (arg_code reg, nilOL) + -- Other args need to be computed beforehand to avoid clobbering + -- previously assigned registers used to pass parameters (see + -- #11792, #12614). They are assigned into temporary registers + -- and get assigned to proper call ABI registers after they all + -- have been computed. + | otherwise + = do arg_code <- getAnyReg arg + tmp <- getNewRegNat arg_fmt + return (arg_code tmp, unitOL $ mkRegRegMoveInstr config arg_fmt tmp reg) + where + platform = ncgPlatform config + arg_fmt = cmmTypeFormat $ cmmExprType platform arg -- ----------------------------------------------------------------------------- -- Pushing arguments onto the stack for 64-bit C calls. @@ -3885,23 +3904,39 @@ addStackPadding pad_bytes -- -- Assumes the current stack pointer fulfills any necessary alignment requirements. pushArgByValue :: NCGConfig -> CmmExpr -> NatM InstrBlock -pushArgByValue config arg = do - (arg_reg, arg_code) <- getSomeReg arg - delta <- getDeltaNat - setDeltaNat (delta-arg_size) - let fmt = cmmTypeFormat arg_rep - return $ arg_code `appOL` toOL - [ SUB (intFormat (wordWidth platform)) (OpImm (ImmInt arg_size)) (OpReg rsp) - , DELTA (delta-arg_size) - -- NB: it's important to only move the actual size, - -- e.g. for W8 width we must only move 8 bits and not 64. - -- Suppose for example the argument is (CmmLoad addr W8 aln); - -- then we must make sure not to try to read more than 8 bits from 'addr'. - , movInstr config fmt (OpReg arg_reg) (OpAddr (spRel platform 0)) ] +pushArgByValue config arg + | isIntFormat fmt + = do + -- Arguments have been promoted to 64-bits wide, so we can use @PUSH II64 at . + -- (Recall that both the Windows and System V 64-bit C calling conventions + -- expect all arguments to be 8-byte aligned.) + -- + -- Note that promotion is crucial. Otherwise, if we tried to push an + -- argument such as @CmmLoad addr W32 aln@, we could end up reading + -- unmapped memory and segfaulting. + (arg_op, arg_code) <- getOperand arg + delta <- getDeltaNat + setDeltaNat (delta-arg_size) + return $ + arg_code `appOL` toOL + [ PUSH II64 arg_op + , DELTA (delta-arg_size) ] + + | otherwise + = do + (arg_reg, arg_code) <- getSomeReg arg + delta <- getDeltaNat + setDeltaNat (delta-arg_size) + return $ arg_code `appOL` toOL + [ SUB (intFormat (wordWidth platform)) (OpImm (ImmInt arg_size)) (OpReg rsp) + , DELTA (delta-arg_size) + , movInstr config fmt (OpReg arg_reg) (OpAddr (spRel platform 0)) ] + where platform = ncgPlatform config arg_size = argSize platform arg arg_rep = cmmExprType platform arg + fmt = cmmTypeFormat arg_rep -- | Load an argument into a register or push it to the stack. loadOrPushArg :: NCGConfig -> (StackArg, Maybe Int) -> NatM (InstrBlock, InstrBlock) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/13c949c5140994c7499551f105d8d7c85315470f...7f962063eeb6676ae559d88146b1a2d7f4950596 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/13c949c5140994c7499551f105d8d7c85315470f...7f962063eeb6676ae559d88146b1a2d7f4950596 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 13:55:51 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sat, 21 Sep 2024 09:55:51 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 15 commits: The X86 SIMD patch. Message-ID: <66eed06798de2_fc9803fc1545456a@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: c7fe2bc4 by sheaf at 2024-09-21T15:55:38+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 02a6faa2 by sheaf at 2024-09-21T15:55:38+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - ce07649e by sheaf at 2024-09-21T15:55:38+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - dae615f7 by sheaf at 2024-09-21T15:55:38+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - bdd36d98 by sheaf at 2024-09-21T15:55:39+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 7403392b by sheaf at 2024-09-21T15:55:39+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - a09e8166 by sheaf at 2024-09-21T15:55:39+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - d311b8aa by sheaf at 2024-09-21T15:55:39+02:00 Add test for C calls & SIMD vectors - - - - - d6751ae8 by sheaf at 2024-09-21T15:55:39+02:00 Add test for #25169 - - - - - 3313c0e2 by sheaf at 2024-09-21T15:55:40+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 0f3d4096 by sheaf at 2024-09-21T15:55:40+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 534eaa78 by sheaf at 2024-09-21T15:55:40+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - d9a03bd6 by sheaf at 2024-09-21T15:55:40+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - da79961d by sheaf at 2024-09-21T15:55:40+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2aa95340 by sheaf at 2024-09-21T15:55:40+02:00 X86 genCCall: promote arg before calling evalArgs - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7f962063eeb6676ae559d88146b1a2d7f4950596...2aa953402a32625791d0e55b3e34d4f53a2c129d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7f962063eeb6676ae559d88146b1a2d7f4950596...2aa953402a32625791d0e55b3e34d4f53a2c129d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 13:58:46 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sat, 21 Sep 2024 09:58:46 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] X86 genCCall: promote arg before calling evalArgs Message-ID: <66eed11610e37_fc980582dc0579d5@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 0a288d07 by sheaf at 2024-09-21T15:56:57+02:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 1 changed file: - compiler/GHC/CmmToAsm/X86/CodeGen.hs Changes: ===================================== compiler/GHC/CmmToAsm/X86/CodeGen.hs ===================================== @@ -3243,33 +3243,39 @@ genCCall -> [CmmFormal] -> [CmmActual] -> NatM InstrBlock -genCCall bid addr conv dest_regs args = do +genCCall bid addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do + platform <- getPlatform is32Bit <- is32BitPlatform - (instrs0, args') <- evalArgs bid args + let args_hints = zip args (argHints ++ repeat NoHint) + prom_args = map (maybePromoteCArgToW32 platform) args_hints + (instrs0, args') <- evalArgs bid prom_args instrs1 <- if is32Bit then genCCall32 addr conv dest_regs args' else genCCall64 addr conv dest_regs args' return (instrs0 `appOL` instrs1) -maybePromoteCArg :: Platform -> Width -> (CmmExpr, ForeignHint) -> CmmExpr -maybePromoteCArg platform wto (arg, hint) - | wfrom < wto = case hint of - SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] - _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] +maybePromoteCArgToW32 :: Platform -> (CmmExpr, ForeignHint) -> CmmExpr +maybePromoteCArgToW32 platform (arg, hint) + | wfrom < wto = + -- As wto=W32, we only need to handle integer conversions, + -- never Float -> Double. + case hint of + SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] + _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] | otherwise = arg where - wfrom = cmmExprWidth platform arg + ty = cmmExprType platform arg + wfrom = typeWidth ty + wto = W32 genCCall32 :: CmmExpr -- ^ address of the function to call -> ForeignConvention -- ^ calling convention -> [CmmFormal] -- ^ where to put the result -> [CmmActual] -- ^ arguments (of mixed type) -> NatM InstrBlock -genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do +genCCall32 addr _conv dest_regs args = do config <- getConfig let platform = ncgPlatform config - args_hints = zip args (argHints ++ repeat NoHint) - prom_args = map (maybePromoteCArg platform W32) args_hints -- If the size is smaller than the word, we widen things (see maybePromoteCArg) arg_size_bytes :: CmmType -> Int @@ -3339,7 +3345,7 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do delta0 <- getDeltaNat setDeltaNat (delta0 - arg_pad_size) - push_codes <- mapM push_arg (reverse prom_args) + push_codes <- mapM push_arg (reverse args) delta <- getDeltaNat massert (delta == delta0 - tot_arg_size) @@ -3414,11 +3420,9 @@ genCCall64 :: CmmExpr -- ^ address of function to call -> [CmmFormal] -- ^ where to put the result -> [CmmActual] -- ^ arguments (of mixed type) -> NatM InstrBlock -genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do +genCCall64 addr conv dest_regs args = do config <- getConfig let platform = ncgPlatform config - args_hints = zip args (argHints ++ repeat NoHint) - prom_args = map (maybePromoteCArg platform W32) args_hints word_size = platformWordSizeInBytes platform wordFmt = archWordFormat (target32Bit platform) @@ -3434,7 +3438,7 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do , usedRegs = arg_regs_used , assignArgsCode = assign_args_code } - <- loadArgs config prom_args + <- loadArgs config args let View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0a288d07bb5458c5c0d3e17a5fcc63f8059782a2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0a288d07bb5458c5c0d3e17a5fcc63f8059782a2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 13:59:58 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sat, 21 Sep 2024 09:59:58 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 3 commits: X86 genCCall: promote arg before calling evalArgs Message-ID: <66eed15ddad03_fc9803fc1b8591b6@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 13bd1182 by sheaf at 2024-09-21T15:59:34+02:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - e9685fea by sheaf at 2024-09-21T15:59:34+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - a7b364be by sheaf at 2024-09-21T15:59:34+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 6 changed files: - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs Changes: ===================================== compiler/GHC/CmmToAsm/X86/CodeGen.hs ===================================== @@ -1965,6 +1965,17 @@ getRegister' platform is32Bit (CmmLit lit) getRegister' platform is32Bit (CmmLit lit) = do avx <- avxEnabled + -- NB: it is important that the code produced here (to load a literal into + -- a register) doesn't clobber any registers other than the destination + -- register; the code for generating C calls relies on this property. + -- + -- In particular, we have: + -- + -- > loadIntoRegMightClobberOtherReg (CmmLit _) = False + -- + -- which means that we assume that loading a literal into a register + -- will not clobber any other registers. + -- TODO: this function mishandles floating-point negative zero, -- because -0.0 == 0.0 returns True and because we represent CmmFloat as -- Rational, which can't properly represent negative zero. @@ -3080,10 +3091,8 @@ genSimplePrim _ op dst args = do platform <- ncgPlatform <$> getConfig pprPanic "genSimplePrim: unhandled primop" (ppr (pprCallishMachOp op, dst, fmap (pdoc platform) args)) -{- -Note [Evaluate C-call arguments before placing in destination registers] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - +{- Note [Evaluate C-call arguments before placing in destination registers] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ When producing code for C calls we must take care when placing arguments in their final registers. Specifically, we must ensure that temporary register usage due to evaluation of one argument does not clobber a register in which we @@ -3134,15 +3143,11 @@ genForeignCall{32,64}. -- | See Note [Evaluate C-call arguments before placing in destination registers] evalArgs :: BlockId -> [CmmActual] -> NatM (InstrBlock, [CmmActual]) evalArgs bid actuals - | any mightContainMachOp actuals = do + | any loadIntoRegMightClobberOtherReg actuals = do regs_blks <- mapM evalArg actuals return (concatOL $ map fst regs_blks, map snd regs_blks) | otherwise = return (nilOL, actuals) where - mightContainMachOp (CmmReg _) = False - mightContainMachOp (CmmRegOff _ _) = False - mightContainMachOp (CmmLit _) = False - mightContainMachOp _ = True evalArg :: CmmActual -> NatM (InstrBlock, CmmExpr) evalArg actual = do @@ -3156,6 +3161,16 @@ evalArgs bid actuals newLocalReg :: CmmType -> NatM LocalReg newLocalReg ty = LocalReg <$> getUniqueM <*> pure ty +-- | Might the code to put this expression into a register +-- clobber any other registers? +loadIntoRegMightClobberOtherReg :: CmmExpr -> Bool +loadIntoRegMightClobberOtherReg (CmmReg _) = False +loadIntoRegMightClobberOtherReg (CmmRegOff _ _) = False +loadIntoRegMightClobberOtherReg (CmmLit _) = False + -- NB: this last 'False' is slightly risky, because the code for loading + -- a literal into a register is not entirely trivial. +loadIntoRegMightClobberOtherReg _ = True + -- Note [DIV/IDIV for bytes] -- ~~~~~~~~~~~~~~~~~~~~~~~~~ -- IDIV reminder: @@ -3228,33 +3243,39 @@ genCCall -> [CmmFormal] -> [CmmActual] -> NatM InstrBlock -genCCall bid addr conv dest_regs args = do +genCCall bid addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do + platform <- getPlatform is32Bit <- is32BitPlatform - (instrs0, args') <- evalArgs bid args + let args_hints = zip args (argHints ++ repeat NoHint) + prom_args = map (maybePromoteCArgToW32 platform) args_hints + (instrs0, args') <- evalArgs bid prom_args instrs1 <- if is32Bit then genCCall32 addr conv dest_regs args' else genCCall64 addr conv dest_regs args' return (instrs0 `appOL` instrs1) -maybePromoteCArg :: Platform -> Width -> (CmmExpr, ForeignHint) -> CmmExpr -maybePromoteCArg platform wto (arg, hint) - | wfrom < wto = case hint of - SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] - _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] +maybePromoteCArgToW32 :: Platform -> (CmmExpr, ForeignHint) -> CmmExpr +maybePromoteCArgToW32 platform (arg, hint) + | wfrom < wto = + -- As wto=W32, we only need to handle integer conversions, + -- never Float -> Double. + case hint of + SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] + _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] | otherwise = arg where - wfrom = cmmExprWidth platform arg + ty = cmmExprType platform arg + wfrom = typeWidth ty + wto = W32 genCCall32 :: CmmExpr -- ^ address of the function to call -> ForeignConvention -- ^ calling convention -> [CmmFormal] -- ^ where to put the result -> [CmmActual] -- ^ arguments (of mixed type) -> NatM InstrBlock -genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do +genCCall32 addr _conv dest_regs args = do config <- getConfig let platform = ncgPlatform config - args_hints = zip args (argHints ++ repeat NoHint) - prom_args = map (maybePromoteCArg platform W32) args_hints -- If the size is smaller than the word, we widen things (see maybePromoteCArg) arg_size_bytes :: CmmType -> Int @@ -3324,7 +3345,7 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do delta0 <- getDeltaNat setDeltaNat (delta0 - arg_pad_size) - push_codes <- mapM push_arg (reverse prom_args) + push_codes <- mapM push_arg (reverse args) delta <- getDeltaNat massert (delta == delta0 - tot_arg_size) @@ -3399,11 +3420,9 @@ genCCall64 :: CmmExpr -- ^ address of function to call -> [CmmFormal] -- ^ where to put the result -> [CmmActual] -- ^ arguments (of mixed type) -> NatM InstrBlock -genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do +genCCall64 addr conv dest_regs args = do config <- getConfig let platform = ncgPlatform config - args_hints = zip args (argHints ++ repeat NoHint) - prom_args = map (maybePromoteCArg platform W32) args_hints word_size = platformWordSizeInBytes platform wordFmt = archWordFormat (target32Bit platform) @@ -3417,10 +3436,9 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do { stackArgs = proper_stack_args , stackDataArgs = stack_data_args , usedRegs = arg_regs_used - , computeArgsCode = compute_args_code , assignArgsCode = assign_args_code } - <- loadArgs config prom_args + <- loadArgs config args let @@ -3516,7 +3534,6 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do return (align_call_code `appOL` push_code `appOL` - compute_args_code `appOL` assign_args_code `appOL` load_data_refs `appOL` shadow_space_code `appOL` @@ -3537,16 +3554,14 @@ data LoadArgs , stackDataArgs :: [CmmExpr] -- | Which registers are we using for argument passing? , usedRegs :: [RegWithFormat] - -- | The code to compute arguments into (possibly temporary) registers. - , computeArgsCode :: InstrBlock -- | The code to assign arguments to registers used for argument passing. , assignArgsCode :: InstrBlock } instance Semigroup LoadArgs where - LoadArgs a1 d1 r1 i1 j1 <> LoadArgs a2 d2 r2 i2 j2 - = LoadArgs (a1 ++ a2) (d1 ++ d2) (r1 ++ r2) (i1 S.<> i2) (j1 S.<> j2) + LoadArgs a1 d1 r1 j1 <> LoadArgs a2 d2 r2 j2 + = LoadArgs (a1 ++ a2) (d1 ++ d2) (r1 ++ r2) (j1 S.<> j2) instance Monoid LoadArgs where - mempty = LoadArgs [] [] [] nilOL nilOL + mempty = LoadArgs [] [] [] nilOL -- | An argument passed on the stack, either directly or by reference. -- @@ -3703,7 +3718,6 @@ loadArgsSysV config (arg:rest) = do LoadArgs { stackArgs = map RawStackArg (arg:rest) , stackDataArgs = [] - , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3723,12 +3737,11 @@ loadArgsSysV config (arg:rest) = do this_arg <- case mbReg of Just reg -> do - (compute_code, assign_code) <- lift $ loadArgIntoReg config arg rest reg + assign_code <- lift $ loadArgIntoReg arg reg return $ LoadArgs { stackArgs = [] -- passed in register , stackDataArgs = [] - , computeArgsCode = compute_code , assignArgsCode = assign_code , usedRegs = [RegWithFormat reg arg_fmt] } @@ -3738,7 +3751,6 @@ loadArgsSysV config (arg:rest) = do LoadArgs { stackArgs = [RawStackArg arg] , stackDataArgs = [] - , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3790,7 +3802,6 @@ loadArgsWin config (arg:rest) = do LoadArgs { stackArgs = stk_args , stackDataArgs = data_args - , computeArgsCode = nilOL , assignArgsCode = nilOL , usedRegs = [] } @@ -3806,8 +3817,7 @@ loadArgsWin config (arg:rest) = do -- Pass the reference in a register, -- and the argument data on the stack. { stackArgs = [RawStackArgRef (InReg ireg) (argSize platform arg)] - , stackDataArgs = [arg] - , computeArgsCode = nilOL -- we don't yet know where the data will reside, + , stackDataArgs = [arg] -- we don't yet know where the data will reside, , assignArgsCode = nilOL -- so we defer computing the reference and storing it -- in the register until later , usedRegs = [RegWithFormat ireg II64] @@ -3818,7 +3828,7 @@ loadArgsWin config (arg:rest) = do = freg | otherwise = ireg - (compute_code, assign_code) <- loadArgIntoReg config arg rest arg_reg + assign_code <- loadArgIntoReg arg arg_reg -- Recall that, for varargs, we must pass floating-point -- arguments in both fp and integer registers. let (assign_code', regs') @@ -3831,42 +3841,23 @@ loadArgsWin config (arg:rest) = do LoadArgs { stackArgs = [] -- passed in register , stackDataArgs = [] - , computeArgsCode = compute_code , assignArgsCode = assign_code' , usedRegs = regs' } - --- | Return two pieces of code: --- --- - code to compute a the given 'CmmExpr' into some (possibly temporary) register --- - code to assign the resulting value to the specified register +-- | Load an argument into a register. -- --- Using two separate pieces of code handles clobbering issues reported --- in e.g. #11792, #12614. -loadArgIntoReg :: NCGConfig -> CmmExpr -> [CmmExpr] -> Reg -> NatM (InstrBlock, InstrBlock) -loadArgIntoReg config arg rest reg - -- "operand" args can be directly assigned into the register - | isOperand platform arg - = do arg_code <- getAnyReg arg - return (nilOL, arg_code reg) - -- The last non-operand arg can be directly assigned after its - -- computation without going into a temporary register - | all (isOperand platform) rest - = do arg_code <- getAnyReg arg - return (arg_code reg, nilOL) - -- Other args need to be computed beforehand to avoid clobbering - -- previously assigned registers used to pass parameters (see - -- #11792, #12614). They are assigned into temporary registers - -- and get assigned to proper call ABI registers after they all - -- have been computed. - | otherwise - = do arg_code <- getAnyReg arg - tmp <- getNewRegNat arg_fmt - return (arg_code tmp, unitOL $ mkRegRegMoveInstr config arg_fmt tmp reg) - where - platform = ncgPlatform config - arg_fmt = cmmTypeFormat $ cmmExprType platform arg +-- Assumes that the expression does not contain any MachOps, +-- as per Note [Evaluate C-call arguments before placing in destination registers]. +loadArgIntoReg :: CmmExpr -> Reg -> NatM InstrBlock +loadArgIntoReg arg reg = do + when (debugIsOn && loadIntoRegMightClobberOtherReg arg) $ do + platform <- getPlatform + massertPpr False $ + vcat [ text "loadArgIntoReg: arg might contain MachOp" + , text "arg:" <+> pdoc platform arg ] + arg_code <- getAnyReg arg + return $ arg_code reg -- ----------------------------------------------------------------------------- -- Pushing arguments onto the stack for 64-bit C calls. ===================================== compiler/GHC/CmmToLlvm.hs ===================================== @@ -139,7 +139,7 @@ llvmGroupLlvmGens cmm = do Nothing -> l Just (CmmStaticsRaw info_lbl _) -> info_lbl lml <- strCLabel_llvm l' - funInsert lml =<< llvmFunTy (map globalRegUse_reg live) + funInsert lml =<< llvmFunTy live return Nothing cdata <- fmap catMaybes $ mapM split cmm ===================================== compiler/GHC/CmmToLlvm/Base.hs ===================================== @@ -12,7 +12,7 @@ module GHC.CmmToLlvm.Base ( LlvmCmmDecl, LlvmBasicBlock, - LiveGlobalRegs, + LiveGlobalRegs, LiveGlobalRegUses, LlvmUnresData, LlvmData, UnresLabel, UnresStatic, LlvmM, @@ -29,6 +29,8 @@ module GHC.CmmToLlvm.Base ( llvmFunSig, llvmFunArgs, llvmStdFunAttrs, llvmFunAlign, llvmInfAlign, llvmPtrBits, tysToParams, llvmFunSection, padLiveArgs, isFPR, + lookupRegUse, + strCLabel_llvm, getGlobalPtr, generateExternDecls, @@ -58,9 +60,11 @@ import GHC.Types.Unique.Set import GHC.Types.Unique.Supply import GHC.Utils.Logger -import Data.Maybe (fromJust) import Control.Monad.Trans.State (StateT (..)) -import Data.List (isPrefixOf) +import Control.Applicative (Alternative((<|>))) +import Data.Maybe (fromJust, mapMaybe) + +import Data.List (find, isPrefixOf) import qualified Data.List.NonEmpty as NE import Data.Ord (comparing) @@ -73,6 +77,7 @@ type LlvmBasicBlock = GenBasicBlock LlvmStatement -- | Global registers live on proc entry type LiveGlobalRegs = [GlobalReg] +type LiveGlobalRegUses = [GlobalRegUse] -- | Unresolved code. -- Of the form: (data label, data type, unresolved data) @@ -116,16 +121,16 @@ llvmGhcCC platform | otherwise = CC_Ghc -- | Llvm Function type for Cmm function -llvmFunTy :: LiveGlobalRegs -> LlvmM LlvmType +llvmFunTy :: LiveGlobalRegUses -> LlvmM LlvmType llvmFunTy live = return . LMFunction =<< llvmFunSig' live (fsLit "a") ExternallyVisible -- | Llvm Function signature -llvmFunSig :: LiveGlobalRegs -> CLabel -> LlvmLinkageType -> LlvmM LlvmFunctionDecl +llvmFunSig :: LiveGlobalRegUses -> CLabel -> LlvmLinkageType -> LlvmM LlvmFunctionDecl llvmFunSig live lbl link = do lbl' <- strCLabel_llvm lbl llvmFunSig' live lbl' link -llvmFunSig' :: LiveGlobalRegs -> LMString -> LlvmLinkageType -> LlvmM LlvmFunctionDecl +llvmFunSig' :: LiveGlobalRegUses -> LMString -> LlvmLinkageType -> LlvmM LlvmFunctionDecl llvmFunSig' live lbl link = do let toParams x | isPointer x = (x, [NoAlias, NoCapture]) | otherwise = (x, []) @@ -149,16 +154,25 @@ llvmFunSection opts lbl | otherwise = Nothing -- | A Function's arguments -llvmFunArgs :: Platform -> LiveGlobalRegs -> [LlvmVar] +llvmFunArgs :: Platform -> LiveGlobalRegUses -> [LlvmVar] llvmFunArgs platform live = - map (lmGlobalRegArg platform) (filter isPassed allRegs) + map (lmGlobalRegArg platform) (mapMaybe isPassed allRegs) where allRegs = activeStgRegs platform paddingRegs = padLiveArgs platform live - isLive r = r `elem` alwaysLive - || r `elem` live - || r `elem` paddingRegs - isPassed r = not (isFPR r) || isLive r - + isLive :: GlobalReg -> Maybe GlobalRegUse + isLive r = + lookupRegUse r (alwaysLive platform) + <|> + lookupRegUse r live + <|> + lookupRegUse r paddingRegs + isPassed r = + if not (isFPR r) + then Just $ GlobalRegUse r (globalRegSpillType platform r) + else isLive r + +lookupRegUse :: GlobalReg -> [GlobalRegUse] -> Maybe GlobalRegUse +lookupRegUse r = find ((== r) . globalRegUse_reg) isFPR :: GlobalReg -> Bool isFPR (FloatReg _) = True @@ -179,7 +193,7 @@ isFPR _ = False -- Invariant: Cmm FPR regs with number "n" maps to real registers with number -- "n" If the calling convention uses registers in a different order or if the -- invariant doesn't hold, this code probably won't be correct. -padLiveArgs :: Platform -> LiveGlobalRegs -> LiveGlobalRegs +padLiveArgs :: Platform -> LiveGlobalRegUses -> LiveGlobalRegUses padLiveArgs platform live = if platformUnregisterised platform then [] -- not using GHC's register convention for platform. @@ -188,7 +202,7 @@ padLiveArgs platform live = ---------------------------------- -- handle floating-point registers (FPR) - fprLive = filter isFPR live -- real live FPR registers + fprLive = filter (isFPR . globalRegUse_reg) live -- real live FPR registers -- we group live registers sharing the same classes, i.e. that use the same -- set of real registers to be passed. E.g. FloatReg, DoubleReg and XmmReg @@ -196,39 +210,44 @@ padLiveArgs platform live = -- classes = NE.groupBy sharesClass fprLive sharesClass a b = globalRegsOverlap platform (norm a) (norm b) -- check if mapped to overlapping registers - norm x = fpr_ctor x 1 -- get the first register of the family + norm x = globalRegUse_reg (fpr_ctor x 1) -- get the first register of the family -- For each class, we just have to fill missing registers numbers. We use -- the constructor of the greatest register to build padding registers. -- -- E.g. sortedRs = [ F2, XMM4, D5] -- output = [D1, D3] + padded :: [GlobalRegUse] padded = concatMap padClass classes + + padClass :: NE.NonEmpty GlobalRegUse -> [GlobalRegUse] padClass rs = go (NE.toList sortedRs) 1 where - sortedRs = NE.sortBy (comparing fpr_num) rs + sortedRs = NE.sortBy (comparing (fpr_num . globalRegUse_reg)) rs maxr = NE.last sortedRs ctor = fpr_ctor maxr go [] _ = [] - go (c1:c2:_) _ -- detect bogus case (see #17920) + go (GlobalRegUse c1 _: GlobalRegUse c2 _:_) _ -- detect bogus case (see #17920) | fpr_num c1 == fpr_num c2 , Just real <- globalRegMaybe platform c1 = sorryDoc "LLVM code generator" $ text "Found two different Cmm registers (" <> ppr c1 <> text "," <> ppr c2 <> text ") both alive AND mapped to the same real register: " <> ppr real <> text ". This isn't currently supported by the LLVM backend." - go (c:cs) f - | fpr_num c == f = go cs f -- already covered by a real register - | otherwise = ctor f : go (c:cs) (f + 1) -- add padding register - - fpr_ctor :: GlobalReg -> Int -> GlobalReg - fpr_ctor (FloatReg _) = FloatReg - fpr_ctor (DoubleReg _) = DoubleReg - fpr_ctor (XmmReg _) = XmmReg - fpr_ctor (YmmReg _) = YmmReg - fpr_ctor (ZmmReg _) = ZmmReg - fpr_ctor _ = error "fpr_ctor expected only FPR regs" + go (cu@(GlobalRegUse c _):cs) f + | fpr_num c == f = go cs f -- already covered by a real register + | otherwise = ctor f : go (cu:cs) (f + 1) -- add padding register + + fpr_ctor :: GlobalRegUse -> Int -> GlobalRegUse + fpr_ctor (GlobalRegUse r fmt) i = + case r of + FloatReg _ -> GlobalRegUse (FloatReg i) fmt + DoubleReg _ -> GlobalRegUse (DoubleReg i) fmt + XmmReg _ -> GlobalRegUse (XmmReg i) fmt + YmmReg _ -> GlobalRegUse (YmmReg i) fmt + ZmmReg _ -> GlobalRegUse (ZmmReg i) fmt + _ -> error "fpr_ctor expected only FPR regs" fpr_num :: GlobalReg -> Int fpr_num (FloatReg i) = i ===================================== compiler/GHC/CmmToLlvm/CodeGen.hs ===================================== @@ -37,13 +37,14 @@ import GHC.Utils.Outputable import qualified GHC.Utils.Panic as Panic import GHC.Utils.Misc +import Control.Applicative (Alternative((<|>))) import Control.Monad.Trans.Class import Control.Monad.Trans.Writer import Control.Monad import qualified Data.Semigroup as Semigroup import Data.List ( nub ) -import Data.Maybe ( catMaybes ) +import Data.Maybe ( catMaybes, isJust ) type Atomic = Maybe MemoryOrdering type LlvmStatements = OrdList LlvmStatement @@ -57,7 +58,7 @@ genLlvmProc :: RawCmmDecl -> LlvmM [LlvmCmmDecl] genLlvmProc (CmmProc infos lbl live graph) = do let blocks = toBlockListEntryFirstFalseFallthrough graph - (lmblocks, lmdata) <- basicBlocksCodeGen (map globalRegUse_reg live) blocks + (lmblocks, lmdata) <- basicBlocksCodeGen live blocks let info = mapLookup (g_entry graph) infos proc = CmmProc info lbl live (ListGraph lmblocks) return (proc:lmdata) @@ -76,7 +77,7 @@ newtype UnreachableBlockId = UnreachableBlockId BlockId -- | Generate code for a list of blocks that make up a complete -- procedure. The first block in the list is expected to be the entry -- point. -basicBlocksCodeGen :: LiveGlobalRegs -> [CmmBlock] +basicBlocksCodeGen :: LiveGlobalRegUses -> [CmmBlock] -> LlvmM ([LlvmBasicBlock], [LlvmCmmDecl]) basicBlocksCodeGen _ [] = panic "no entry block!" basicBlocksCodeGen live cmmBlocks @@ -152,7 +153,7 @@ stmtToInstrs ubid stmt = case stmt of -- Tail call CmmCall { cml_target = arg, - cml_args_regs = live } -> genJump arg $ map globalRegUse_reg live + cml_args_regs = live } -> genJump arg live _ -> panic "Llvm.CodeGen.stmtToInstrs" @@ -1050,7 +1051,7 @@ cmmPrimOpFunctions mop = do -- | Tail function calls -genJump :: CmmExpr -> [GlobalReg] -> LlvmM StmtData +genJump :: CmmExpr -> LiveGlobalRegUses -> LlvmM StmtData -- Call to known function genJump (CmmLit (CmmLabel lbl)) live = do @@ -2056,14 +2057,13 @@ getCmmReg (CmmLocal (LocalReg un _)) -- have been assigned a value at some point, triggering -- "funPrologue" to allocate it on the stack. -getCmmReg (CmmGlobal g) - = do let r = globalRegUse_reg g - onStack <- checkStackReg r +getCmmReg (CmmGlobal ru@(GlobalRegUse r _)) + = do onStack <- checkStackReg r platform <- getPlatform if onStack - then return (lmGlobalRegVar platform r) + then return (lmGlobalRegVar platform ru) else pprPanic "getCmmReg: Cmm register " $ - ppr g <> text " not stack-allocated!" + ppr r <> text " not stack-allocated!" -- | Return the value of a given register, as well as its type. Might -- need to be load from stack. @@ -2074,7 +2074,7 @@ getCmmRegVal reg = onStack <- checkStackReg (globalRegUse_reg g) platform <- getPlatform if onStack then loadFromStack else do - let r = lmGlobalRegArg platform (globalRegUse_reg g) + let r = lmGlobalRegArg platform g return (r, getVarType r, nilOL) _ -> loadFromStack where loadFromStack = do @@ -2187,8 +2187,9 @@ convertMemoryOrdering MemOrderSeqCst = SyncSeqCst -- question is never written. Therefore we skip it where we can to -- save a few lines in the output and hopefully speed compilation up a -- bit. -funPrologue :: LiveGlobalRegs -> [CmmBlock] -> LlvmM StmtData +funPrologue :: LiveGlobalRegUses -> [CmmBlock] -> LlvmM StmtData funPrologue live cmmBlocks = do + platform <- getPlatform let getAssignedRegs :: CmmNode O O -> [CmmReg] getAssignedRegs (CmmAssign reg _) = [reg] @@ -2196,7 +2197,8 @@ funPrologue live cmmBlocks = do getAssignedRegs _ = [] getRegsBlock (_, body, _) = concatMap getAssignedRegs $ blockToList body assignedRegs = nub $ concatMap (getRegsBlock . blockSplit) cmmBlocks - isLive r = r `elem` alwaysLive || r `elem` live + mbLive r = + lookupRegUse r (alwaysLive platform) <|> lookupRegUse r live platform <- getPlatform stmtss <- forM assignedRegs $ \reg -> @@ -2205,12 +2207,12 @@ funPrologue live cmmBlocks = do let (newv, stmts) = allocReg reg varInsert un (pLower $ getVarType newv) return stmts - CmmGlobal (GlobalRegUse r _) -> do - let reg = lmGlobalRegVar platform r - arg = lmGlobalRegArg platform r + CmmGlobal ru@(GlobalRegUse r _) -> do + let reg = lmGlobalRegVar platform ru + arg = lmGlobalRegArg platform ru ty = (pLower . getVarType) reg trash = LMLitVar $ LMUndefLit ty - rval = if isLive r then arg else trash + rval = if isJust (mbLive r) then arg else trash alloc = Assignment reg $ Alloca (pLower $ getVarType reg) 1 markStackReg r return $ toOL [alloc, Store rval reg Nothing []] @@ -2222,7 +2224,7 @@ funPrologue live cmmBlocks = do -- | Function epilogue. Load STG variables to use as argument for call. -- STG Liveness optimisation done here. -funEpilogue :: LiveGlobalRegs -> LlvmM ([LlvmVar], LlvmStatements) +funEpilogue :: LiveGlobalRegUses -> LlvmM ([LlvmVar], LlvmStatements) funEpilogue live = do platform <- getPlatform @@ -2248,12 +2250,16 @@ funEpilogue live = do let allRegs = activeStgRegs platform loads <- forM allRegs $ \r -> if -- load live registers - | r `elem` alwaysLive -> loadExpr (GlobalRegUse r (globalRegSpillType platform r)) - | r `elem` live -> loadExpr (GlobalRegUse r (globalRegSpillType platform r)) + | Just ru <- lookupRegUse r (alwaysLive platform) + -> loadExpr ru + | Just ru <- lookupRegUse r live + -> loadExpr ru -- load all non Floating-Point Registers - | not (isFPR r) -> loadUndef r + | not (isFPR r) + -> loadUndef (GlobalRegUse r (globalRegSpillType platform r)) -- load padding Floating-Point Registers - | r `elem` paddingRegs -> loadUndef r + | Just ru <- lookupRegUse r paddingRegs + -> loadUndef ru | otherwise -> return (Nothing, nilOL) let (vars, stmts) = unzip loads @@ -2263,7 +2269,7 @@ funEpilogue live = do -- -- This is for Haskell functions, function type is assumed, so doesn't work -- with foreign functions. -getHsFunc :: LiveGlobalRegs -> CLabel -> LlvmM ExprData +getHsFunc :: LiveGlobalRegUses -> CLabel -> LlvmM ExprData getHsFunc live lbl = do fty <- llvmFunTy live name <- strCLabel_llvm lbl ===================================== compiler/GHC/CmmToLlvm/Ppr.hs ===================================== @@ -49,9 +49,8 @@ pprLlvmCmmDecl (CmmData _ lmdata) = do return ( vcat $ map (pprLlvmData opts) lmdata , vcat $ map (pprLlvmData opts) lmdata) -pprLlvmCmmDecl (CmmProc mb_info entry_lbl liveWithUses (ListGraph blks)) - = do let live = map globalRegUse_reg liveWithUses - lbl = case mb_info of +pprLlvmCmmDecl (CmmProc mb_info entry_lbl live (ListGraph blks)) + = do let lbl = case mb_info of Nothing -> entry_lbl Just (CmmStaticsRaw info_lbl _) -> info_lbl link = if externallyVisibleCLabel lbl ===================================== compiler/GHC/CmmToLlvm/Regs.hs ===================================== @@ -14,25 +14,27 @@ import GHC.Prelude import GHC.Llvm import GHC.Cmm.Expr +import GHC.CmmToAsm.Format import GHC.Platform import GHC.Data.FastString import GHC.Utils.Panic ( panic ) import GHC.Types.Unique + -- | Get the LlvmVar function variable storing the real register -lmGlobalRegVar :: Platform -> GlobalReg -> LlvmVar +lmGlobalRegVar :: Platform -> GlobalRegUse -> LlvmVar lmGlobalRegVar platform = pVarLift . lmGlobalReg platform "_Var" -- | Get the LlvmVar function argument storing the real register -lmGlobalRegArg :: Platform -> GlobalReg -> LlvmVar +lmGlobalRegArg :: Platform -> GlobalRegUse -> LlvmVar lmGlobalRegArg platform = lmGlobalReg platform "_Arg" {- Need to make sure the names here can't conflict with the unique generated names. Uniques generated names containing only base62 chars. So using say the '_' char guarantees this. -} -lmGlobalReg :: Platform -> String -> GlobalReg -> LlvmVar -lmGlobalReg platform suf reg +lmGlobalReg :: Platform -> String -> GlobalRegUse -> LlvmVar +lmGlobalReg platform suf (GlobalRegUse reg ty) = case reg of BaseReg -> ptrGlobal $ "Base" ++ suf Sp -> ptrGlobal $ "Sp" ++ suf @@ -88,13 +90,26 @@ lmGlobalReg platform suf reg ptrGlobal name = LMNLocalVar (fsLit name) (llvmWordPtr platform) floatGlobal name = LMNLocalVar (fsLit name) LMFloat doubleGlobal name = LMNLocalVar (fsLit name) LMDouble - xmmGlobal name = LMNLocalVar (fsLit name) (LMVector 4 (LMInt 32)) - ymmGlobal name = LMNLocalVar (fsLit name) (LMVector 8 (LMInt 32)) - zmmGlobal name = LMNLocalVar (fsLit name) (LMVector 16 (LMInt 32)) + fmt = cmmTypeFormat ty + xmmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + ymmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + zmmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + +formatLlvmType :: Format -> LlvmType +formatLlvmType II8 = LMInt 8 +formatLlvmType II16 = LMInt 16 +formatLlvmType II32 = LMInt 32 +formatLlvmType II64 = LMInt 64 +formatLlvmType FF32 = LMFloat +formatLlvmType FF64 = LMDouble +formatLlvmType (VecFormat l sFmt) = LMVector l (formatLlvmType $ scalarFormatFormat sFmt) -- | A list of STG Registers that should always be considered alive -alwaysLive :: [GlobalReg] -alwaysLive = [BaseReg, Sp, Hp, SpLim, HpLim, node] +alwaysLive :: Platform -> [GlobalRegUse] +alwaysLive platform = + [ GlobalRegUse r (globalRegSpillType platform r) + | r <- [BaseReg, Sp, Hp, SpLim, HpLim, node] + ] -- | STG Type Based Alias Analysis hierarchy stgTBAA :: [(Unique, LMString, Maybe Unique)] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0a288d07bb5458c5c0d3e17a5fcc63f8059782a2...a7b364be1d319a7f78246060eafb3d483bd94beb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0a288d07bb5458c5c0d3e17a5fcc63f8059782a2...a7b364be1d319a7f78246060eafb3d483bd94beb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 14:36:39 2024 From: gitlab at gitlab.haskell.org (Sven Tennie (@supersven)) Date: Sat, 21 Sep 2024 10:36:39 -0400 Subject: [Git][ghc/ghc][wip/supersven/add-LLVMAS-target] Debug Message-ID: <66eed9f74491a_fc9809de84c60961@gitlab.mail> Sven Tennie pushed to branch wip/supersven/add-LLVMAS-target at Glasgow Haskell Compiler / GHC Commits: fb82db15 by Sven Tennie at 2024-09-21T16:36:25+02:00 Debug - - - - - 3 changed files: - configure.ac - m4/fp_cc_supports_target.m4 - m4/fp_prog_llvm_as_args.m4 Changes: ===================================== configure.ac ===================================== @@ -545,9 +545,6 @@ FIND_LLVM_PROG([LLVMAS], [clang], [$LlvmMinVersion], [$LlvmMaxVersion]) LlvmAsCmd="$LLVMAS" AC_SUBST([LlvmAsCmd]) -dnl ** Which LLVM assembler arguments to use (e.g. `--target=...`)? -FP_PROG_LLVM_AS_ARGS() - dnl -------------------------------------------------------------- dnl End of configure script option section dnl -------------------------------------------------------------- @@ -655,6 +652,9 @@ FP_CC_SUPPORTS_TARGET([$CC_STAGE0], [CONF_CC_OPTS_STAGE0], [CONF_CXX_OPTS_STAGE0 FP_CC_SUPPORTS_TARGET([$CC], [CONF_CC_OPTS_STAGE1], [CONF_CXX_OPTS_STAGE1]) FP_CC_SUPPORTS_TARGET([$CC], [CONF_CC_OPTS_STAGE2], [CONF_CXX_OPTS_STAGE2]) +dnl ** Which LLVM assembler arguments to use (e.g. `--target=...`)? +FP_PROG_LLVM_AS_ARGS() + FP_PROG_CC_LINKER_TARGET([$CC_STAGE0], [CONF_CC_OPTS_STAGE0], [CONF_GCC_LINKER_OPTS_STAGE0]) FP_PROG_CC_LINKER_TARGET([$CC], [CONF_CC_OPTS_STAGE1], [CONF_GCC_LINKER_OPTS_STAGE1]) FP_PROG_CC_LINKER_TARGET([$CC], [CONF_CC_OPTS_STAGE2], [CONF_GCC_LINKER_OPTS_STAGE2]) ===================================== m4/fp_cc_supports_target.m4 ===================================== @@ -13,7 +13,7 @@ AC_DEFUN([FP_CC_SUPPORTS_TARGET], [ AC_REQUIRE([GHC_LLVM_TARGET_SET_VAR]) - AC_MSG_CHECKING([whether $1 supports --target]) + AC_MSG_CHECKING([whether $1 supports --target $LlvmTarget]) echo 'int main() { return 0; }' > conftest.c if test "$target_cpu" = "javascript" ; then @@ -32,6 +32,8 @@ AC_DEFUN([FP_CC_SUPPORTS_TARGET], if test $CONF_CC_SUPPORTS_TARGET = YES ; then $2="--target=$LlvmTarget $$2" $3="--target=$LlvmTarget $$3" + AC_MSG_NOTICE([Result 2: $2]) + AC_MSG_NOTICE([Result 3: $3]) fi ]) ===================================== m4/fp_prog_llvm_as_args.m4 ===================================== @@ -40,6 +40,9 @@ else # to build intermediate GHC stages (with host target). This ressembles old # behaviour and is added for backwards compatibility. LlvmAsArgsHost="" + + # TODO: Delete this debug output + $LLVMAS -v $LlvmAsArgsHost test.ll fi AC_MSG_CHECKING([whether \$LLVMAS supports target flags passed by GHC when compiling]) @@ -48,6 +51,9 @@ if $LLVMAS $LlvmAsArgsTarget test.ll > /dev/null 2>&1 ; then else AC_MSG_RESULT([no]) AC_MSG_ERROR([\$LLVMAS ($LLVMAS) does not support target flags: $LlvmAsArgsTarget]) + + # TODO: Delete this debug output + $LLVMAS -v $LlvmAsArgsTarget test.ll fi AC_SUBST(LlvmAsArgsHost) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fb82db15cd872ee993974b332a7bab2dd64d44f2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fb82db15cd872ee993974b332a7bab2dd64d44f2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 15:28:32 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Sat, 21 Sep 2024 11:28:32 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/az/epa-annsmodule-eptokens Message-ID: <66eee6204a43d_fc980e126c07697f@gitlab.mail> Alan Zimmerman pushed new branch wip/az/epa-annsmodule-eptokens at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/az/epa-annsmodule-eptokens You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 15:38:00 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 11:38:00 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 10 commits: Demand: Combine examples into Note (#25107) Message-ID: <66eee8586ac88_fc980e126c0824db@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 1eff88d7 by Sebastian Graf at 2024-09-21T11:37:38-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 70e0f4fe by sheaf at 2024-09-21T11:37:39-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 67017608 by sheaf at 2024-09-21T11:37:39-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - b5dea7b9 by Ryan Hendrickson at 2024-09-21T11:37:41-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - 1411ecd2 by Zubin Duggal at 2024-09-21T11:37:42-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 5224343b by Andreas Klebinger at 2024-09-21T11:37:42-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - 9f0d68ff by Hécate Kleidukos at 2024-09-21T11:37:44-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - abf985c8 by Zubin Duggal at 2024-09-21T11:37:44-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 6597b7b3 by Sven Tennie at 2024-09-21T11:37:45-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - e7fedcf4 by Sylvain Henry at 2024-09-21T11:37:48-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - 25 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - docs/users_guide/exts/extended_literals.rst - hadrian/bootstrap/bootstrap.py - hadrian/bootstrap/generate_bootstrap_plans - hadrian/bootstrap/plan-9_10_1.json - − hadrian/bootstrap/plan-9_4_1.json - − hadrian/bootstrap/plan-9_4_2.json The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2404d8013374ed12efc2a1a26ce1840529f614da...e7fedcf4d7356dcfad2af2df1cbad68e1f6aa397 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2404d8013374ed12efc2a1a26ce1840529f614da...e7fedcf4d7356dcfad2af2df1cbad68e1f6aa397 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 18:50:40 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Sat, 21 Sep 2024 14:50:40 -0400 Subject: [Git][ghc/ghc][wip/strings] 84 commits: compiler: Fix pretty printing of ticked prefix constructors (#24237) Message-ID: <66ef158077785_356d5346afc810209c@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC Commits: 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - b7f2b438 by Brandon Chinn at 2024-09-21T11:50:25-07:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 21 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/05589606db8c91f10bb96b69f3cb458c8d558f8b...b7f2b4383c217efc4e305f5f113ce7b67aec141f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/05589606db8c91f10bb96b69f3cb458c8d558f8b...b7f2b4383c217efc4e305f5f113ce7b67aec141f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 18:51:07 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Sat, 21 Sep 2024 14:51:07 -0400 Subject: [Git][ghc/ghc][wip/interpolated-strings] 87 commits: compiler: Fix pretty printing of ticked prefix constructors (#24237) Message-ID: <66ef159b41af8_356d534df198102671@gitlab.mail> Brandon Chinn pushed to branch wip/interpolated-strings at Glasgow Haskell Compiler / GHC Commits: 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - b7f2b438 by Brandon Chinn at 2024-09-21T11:50:25-07:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - ea3c515d by Brandon Chinn at 2024-09-21T11:50:55-07:00 Merge branch 'wip/strings' into HEAD - - - - - 05bedfcf by Brandon Chinn at 2024-09-21T11:50:55-07:00 Unify ITstring + ITstringMulti - - - - - 4fabf783 by Brandon Chinn at 2024-09-21T11:50:55-07:00 Implement interpolated strings - - - - - 21 changed files: - .gitignore - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9e72c94bd95d0295096574b4701b0cdc0972c508...4fabf783334f6fb26ea31aadc44dfee5e7e4ade0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9e72c94bd95d0295096574b4701b0cdc0972c508...4fabf783334f6fb26ea31aadc44dfee5e7e4ade0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 19:18:31 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Sat, 21 Sep 2024 15:18:31 -0400 Subject: [Git][ghc/ghc][wip/interpolated-strings] Implement interpolated strings Message-ID: <66ef1c07eb47c_23cb27bc764826f3@gitlab.mail> Brandon Chinn pushed to branch wip/interpolated-strings at Glasgow Haskell Compiler / GHC Commits: e0692156 by Brandon Chinn at 2024-09-21T12:18:23-07:00 Implement interpolated strings - - - - - 8 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - compiler/Language/Haskell/Syntax/Expr.hs - compiler/Language/Haskell/Syntax/Extension.hs - libraries/array - libraries/deepseq - libraries/directory Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -242,6 +242,9 @@ type instance XIPVar GhcRn = NoExtField type instance XIPVar GhcTc = DataConCantHappen type instance XOverLitE (GhcPass _) = NoExtField type instance XLitE (GhcPass _) = NoExtField +type instance XInterString (GhcPass _) = NoExtField +type instance XInterStringRaw (GhcPass _) = NoExtField +type instance XInterStringExp (GhcPass _) = NoExtField type instance XLam (GhcPass _) = [AddEpAnn] type instance XApp (GhcPass _) = NoExtField ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -167,7 +167,7 @@ $idchar = [$small $large $digit $uniidchar \'] $unigraphic = \x06 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $graphic = [$small $large $symbol $digit $idchar $special $unigraphic \"\'] -$charesc = [a b f n r t v \\ \" \' \&] +$charesc = [a b f n r t v \\ \" \' \& \$] $binit = 0-1 $octit = 0-7 @@ -223,10 +223,11 @@ $docsym = [\| \^ \* \$] -- N.B. ideally, we would do `@escape # \\ \&` instead of duplicating in @escapechar, -- which is what the Haskell Report says, but this isn't valid Alex syntax, as only -- character sets can be subtracted, not strings - at escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) - at escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) - at stringchar = ($graphic # [\\ \"]) | " " | @escape | @gap - at char = ($graphic # [\\ \']) | " " | @escapechar + at escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) + at escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) + at stringchar = ($graphic # [\\ \"]) | " " | @escape | @gap + at char = ($graphic # [\\ \']) | " " | @escapechar + at stringinterchar = ($graphic # [\\ \" \$]) | " " | @escape | @gap -- normal signed numerical literals can only be explicitly negative, -- not explicitly positive (contrast @exponent) @@ -699,6 +700,25 @@ $unigraphic / { isSmartQuote } { smart_quote_error } (\" | \"\") / ([\n .] # \") { tok_string_multi_content } } +-- See Note [Interpolated strings] +<0> { + s \" { \span _ _ _ -> pushLexState string_inter_content >> pure (L span (ITstringInterBegin src StringTypeSingle)) } + -- TODO: interpolated multiline strings +} + + { + @stringinterchar* { tok_string_inter_raw } + \$ \{ { \span _ _ _ -> pushLexState string_inter >> pure (L span (ITstringInterExpOpen src)) } + \" { \span _ _ _ -> popLexState >> pure (L span (ITstringInterEnd src StringTypeSingle)) } + + -- TODO: check for smart quotes +} + +-- TODO: add string_inter state to all <0> states that can be in an interpolated string + { + \} { \span _ _ _ -> popLexState >> pure (L span (ITstringInterExpClose src)) } +} + <0> { \'\' { token ITtyQuote } @@ -993,6 +1013,13 @@ data Token | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" | ITstring SourceText FastString StringType -- Note [Literal source text] in "GHC.Types.SourceText" + -- See Note [Interpolated strings] + | ITstringInterBegin SourceText StringType -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringInterRaw SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringInterExpOpen SourceText -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringInterExpClose SourceText -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringInterEnd SourceText StringType -- Note [Literal source text] in "GHC.Types.SourceText" + | ITinteger IntegralLit -- Note [Literal source text] in "GHC.Types.SourceText" | ITrational FractionalLit ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -378,6 +378,47 @@ It's more precisely defined with the following algorithm: 3. Calculate the longest prefix of whitespace shared by all lines in the remaining list -} +{- +Note [Interpolated strings] +~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Interpolated string syntax was accepted in this proposal: +https://github.com/ghc-proposals/ghc-proposals/pull/570 + +Interpolated strings are syntax sugar for + +Interpolated strings are implemented in the following manner: + +1. Lexer takes the string as input: + + s"Hello ${Text.toUpper name}!" + + and outputs the following tokens: + + [ ITstringInterBegin src StringTypeSingle + , ITstringInterRaw src "Hello " + , ITstringInterExpOpen src + , ITqvarid ("Text.toUpper", "name") + , ITvarid "name" + , ITstringInterExpClose src + , ITstringInterRaw src "!" + , ITstringInterEnd src StringTypeSingle + ] + +2. The parser will then parse the tokens into the following HsExpr: + + HsInterString ext + [ HsInterRaw ext "Hello " + , HsInterExp ext $ + HsApp ext + (HsVar ext 'Text.toUpper) + (HsVar ext 'name) + , HsInterRaw ext "!" + ] + +3. This will then be desugared into +-} + -- ----------------------------------------------------------------------------- -- DList ===================================== compiler/Language/Haskell/Syntax/Expr.hs ===================================== @@ -350,7 +350,6 @@ data HsExpr p -- erroring expression will be written after -- solving. See Note [Holes] in GHC.Tc.Types.Constraint. - | HsRecSel (XRecSel p) (FieldOcc p) -- ^ Variable pointing to record selector -- See Note [Non-overloaded record field selectors] and @@ -361,12 +360,18 @@ data HsExpr p | HsIPVar (XIPVar p) HsIPName -- ^ Implicit parameter (not in use after typechecking) + | HsOverLit (XOverLitE p) (HsOverLit p) -- ^ Overloaded literals | HsLit (XLitE p) (HsLit p) -- ^ Simple (non-overloaded) literals + | -- | See Note [Interpolated strings] + HsInterString + (XInterString p) + [Either FastString (HsExpr p)] + -- | Lambda, Lambda-case, and Lambda-cases -- -- - 'GHC.Parser.Annotation.AnnKeywordId' : 'GHC.Parser.Annotation.AnnLam', ===================================== compiler/Language/Haskell/Syntax/Extension.hs ===================================== @@ -413,6 +413,9 @@ type family XOverLabel x type family XIPVar x type family XOverLitE x type family XLitE x +type family XInterString x +type family XInterStringRaw x +type family XInterStringExp x type family XLam x type family XLamCase x type family XApp x ===================================== libraries/array ===================================== @@ -1 +1 @@ -Subproject commit c9cb2c1e8762aa83b6e77af82c87a55e03e990e4 +Subproject commit ba5e9dcf1370190239395b8361b1c92ea9fc7632 ===================================== libraries/deepseq ===================================== @@ -1 +1 @@ -Subproject commit 7ce6e2d3760b23336fd5f9a36f50df6571606947 +Subproject commit 09aed1bf774f2f05c8b390539ce35adf5cd68c30 ===================================== libraries/directory ===================================== @@ -1 +1 @@ -Subproject commit 6045b93c4ef7a713c8f3d6837ca69f8e96b12bf1 +Subproject commit a97a8a8f30d652f972192122fd5f459a147c13e5 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e069215631ede6dc7c4ee35fab30dbe049ee856d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e069215631ede6dc7c4ee35fab30dbe049ee856d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 19:39:51 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Sat, 21 Sep 2024 15:39:51 -0400 Subject: [Git][ghc/ghc][wip/25232] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66ef2107534ad_23cb271c57788845d@gitlab.mail> Zubin pushed to branch wip/25232 at Glasgow Haskell Compiler / GHC Commits: bfe8ae39 by Zubin Duggal at 2024-09-22T01:09:41+05:30 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 5 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - + testsuite/tests/rts/T25232.hs - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,18 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + // Not doing this lead to #25232 and #23958 + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +958,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): ===================================== testsuite/tests/rts/T25232.hs ===================================== @@ -0,0 +1,14 @@ +module Main where + +import Control.Monad +import Control.Monad.Trans.State.Strict +import Data.IORef + +main :: IO () +main = do + ref <- newIORef () + replicateM_ 1000 $ withIORef ref $ runStateT (pure ()) + +withIORef :: IORef a -> (a -> IO (b, a)) -> IO b +withIORef ref f = + readIORef ref >>= f >>= \(b, a) -> writeIORef ref a >> pure b ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [req_profiling, extra_ways(['nonmoving', 'nonmoving_prof', 'nonmoving_thr_prof'])], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/bfe8ae393d0d1bc056bdc9f5dfb31d79ec805562 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/bfe8ae393d0d1bc056bdc9f5dfb31d79ec805562 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:48:29 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 17:48:29 -0400 Subject: [Git][ghc/ghc][master] Demand: Combine examples into Note (#25107) Message-ID: <66ef3f2dc09d2_23cb27664cd49801e@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 2 changed files: - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Types/Demand.hs Changes: ===================================== compiler/GHC/Core/Opt/DmdAnal.hs ===================================== @@ -1970,10 +1970,10 @@ W/W-transformed /caller of/ 'f' would immediately rebox any unboxed arguments that is applied to the wrapper of 'f'. When the wrapper is inlined, that kind of reboxing does not happen. -But now we have functions with OPAQUE pragmas, which by definition (See Note -[OPAQUE pragma]) do not get W/W-transformed. So in order to avoid reboxing -workers of any W/W-transformed /callers of/ 'f' we need to strip all boxity -information from 'f' in the demand analysis. This will inform the +But now we have functions with OPAQUE pragmas, which by definition +(See Note [OPAQUE pragma]) do not get W/W-transformed. So in order to avoid +reboxing workers of any W/W-transformed /callers of/ 'f' we need to strip all +boxity information from 'f' in the demand analysis. This will inform the W/W-transformation code that boxed arguments of 'f' must definitely be passed along in boxed form and as such dissuade the creation of reboxing workers. -} ===================================== compiler/GHC/Types/Demand.hs ===================================== @@ -606,22 +606,8 @@ multCard (Card a) (Card b) -- * How many times a variable is evaluated, via a 'Card'inality, and -- * How deep its value was evaluated in turn, via a 'SubDemand'. -- --- Examples (using Note [Demand notation]): --- --- * 'seq' puts demand `1A` on its first argument: It evaluates the argument --- strictly (`1`), but not any deeper (`A`). --- * 'fst' puts demand `1P(1L,A)` on its argument: It evaluates the argument --- pair strictly and the first component strictly, but no nested info --- beyond that (`L`). Its second argument is not used at all. --- * '$' puts demand `1C(1,L)` on its first argument: It calls (`C`) the --- argument function with one argument, exactly once (`1`). No info --- on how the result of that call is evaluated (`L`). --- * 'maybe' puts demand `MC(M,L)` on its second argument: It evaluates --- the argument function at most once ((M)aybe) and calls it once when --- it is evaluated. --- * `fst p + fst p` puts demand `SP(SL,A)` on `p`: It's `1P(1L,A)` --- multiplied by two, so we get `S` (used at least once, possibly multiple --- times). +-- See also Note [Demand notation] +-- and Note [Demand examples]. -- -- This data type is quite similar to `'Scaled' 'SubDemand'`, but it's scaled -- by 'Card', which is an /interval/ on 'Multiplicity', the upper bound of @@ -2657,12 +2643,8 @@ So, L can denote a 'Card', polymorphic 'SubDemand' or polymorphic 'Demand', but it's always clear from context which "overload" is meant. It's like return-type inference of e.g. 'read'. -Examples are in the haddock for 'Demand'. Here are some more: - SA Strict, but does not look at subcomponents (`seq`) - SP(L,L) Strict boxed pair, components lazy - S!P(L,L) Strict unboxed pair, components lazy - LP(SA,SA) Lazy pair, but if it is evaluated will evaluated its components - LC(1C(L)) Lazy, but if called will apply the result exactly once +An example of the demand syntax is 1!P(1!L,A), the demand of fst's argument. +See Note [Demand examples] for more examples and their semantics. This is the syntax for demand signatures: @@ -2680,7 +2662,39 @@ This is the syntax for demand signatures: (omitted if empty) (omitted if no information) - +Note [Demand examples] +~~~~~~~~~~~~~~~~~~~~~~ +Here are some examples of the demand notation, specified in Note [Demand notation], +in action. In each case we give the demand on the variable `x`. + +Demand on x Example Explanation + 1!A seq x y Evaluates `x` exactly once (`1`), but not + any deeper (`A`), and discards the box (`!`). + S!A seq x (seq x y) Twice the previous demand; hence eval'd + more than once (`S` for strict). + 1!P(1!L,A) fst x Evaluates pair `x` exactly once, first + component exactly once. No info that (`L`). + Second component is absent. Discards boxes (`!`). + 1P(1L,A) opq_fst x Like fst, but all boxes are retained. + SP(1!L,A) opq_seq x (fst x) Two evals of x but exactly one of its first component. + Box of x retained, but box of first component discarded. + 1!C(1,L) x $ 3 Evals x exactly once ( 1 ) and calls it + exactly once ( C(1,_) ). No info on how the + result is evaluated ( L ). + MC(M,L) maybe y x Evals x at most once ( 1 ) and calls it at + most once ( C(1,_) ). No info on how the + result is evaluated ( L ). + LP(SL,A) map (+ fst x) Evals x lazily and multiple times ( L ), + but when it is evaluated, the first + component is evaluated (strictly) as well. + +In the examples above, `opq_fst` is an opaque wrapper around `fst`, i.e. + + opq_fst = fst + {-# OPAQUE opq_fst #-} + +Similarly for `seq`. The effect of an OPAQUE pragma is that it discards any +boxity flags in the demand signature, as described in Note [OPAQUE pragma]. -} -- | See Note [Demand notation] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9e96dad809667c777ebb86983a6e8b372d605e17 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9e96dad809667c777ebb86983a6e8b372d605e17 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:49:15 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 17:49:15 -0400 Subject: [Git][ghc/ghc][master] 2 commits: Use x86_64-unknown-windows-gnu target for LLVM on Windows Message-ID: <66ef3f5b3dd85_23cb2766c290103938@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - 7 changed files: - compiler/GHC/Driver/Pipeline/Execute.hs - llvm-targets - m4/ghc_llvm_target.m4 - + testsuite/tests/llvm/should_run/T22487.hs - + testsuite/tests/llvm/should_run/T22487.stdout - + testsuite/tests/llvm/should_run/all.T - utils/llvm-targets/gen-data-layout.sh Changes: ===================================== compiler/GHC/Driver/Pipeline/Execute.hs ===================================== @@ -969,13 +969,18 @@ llvmOptions llvm_config dflags = ++ [("", "-target-abi=" ++ abi) | not (null abi) ] where target = platformMisc_llvmTarget $ platformMisc dflags + target_os = platformOS (targetPlatform dflags) Just (LlvmTarget _ mcpu mattr) = lookup target (llvmTargets llvm_config) -- Relocation models - rmodel | gopt Opt_PIC dflags = "pic" - | positionIndependent dflags = "pic" - | ways dflags `hasWay` WayDyn = "dynamic-no-pic" - | otherwise = "static" + rmodel | gopt Opt_PIC dflags + || positionIndependent dflags + || target_os == OSMinGW32 -- #22487: use PIC on (64-bit) Windows + = "pic" + | ways dflags `hasWay` WayDyn + = "dynamic-no-pic" + | otherwise + = "static" platform = targetPlatform dflags arch = platformArch platform ===================================== llvm-targets ===================================== @@ -1,4 +1,4 @@ -[("x86_64-unknown-windows", ("e-m:w-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", "x86-64", "")) +[("x86_64-unknown-windows-gnu", ("e-m:w-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128", "x86-64", "")) ,("arm-unknown-linux-gnueabi", ("e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64", "arm7tdmi", "+strict-align")) ,("arm-unknown-linux-gnueabihf", ("e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64", "arm1176jzf-s", "+strict-align")) ,("arm-unknown-linux-musleabihf", ("e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64", "arm1176jzf-s", "+strict-align")) ===================================== m4/ghc_llvm_target.m4 ===================================== @@ -14,7 +14,7 @@ AC_DEFUN([GHC_LLVM_TARGET], [ ;; *-mingw32|*-mingw64|*-msys) llvm_target_vendor="unknown" - llvm_target_os="windows" + llvm_target_os="windows-gnu" ;; # apple is a bit about their naming scheme for # aarch64; and clang on macOS doesn't know that ===================================== testsuite/tests/llvm/should_run/T22487.hs ===================================== @@ -0,0 +1,8 @@ + +module Main where + +add :: Double -> Double -> Double +add x y = x + y +{-# NOINLINE add #-} +main = do putStrLn "Hello world!" + print (add 1.0 2.0) ===================================== testsuite/tests/llvm/should_run/T22487.stdout ===================================== @@ -0,0 +1,2 @@ +Hello world! +3.0 ===================================== testsuite/tests/llvm/should_run/all.T ===================================== @@ -0,0 +1,15 @@ + +def f( name, opts ): + opts.only_ways = ['optllvm', 'llvm', 'debugllvm'] + +setTestOpts(f) + +# Apples LLVM Toolchain knows about a `vortex` cpu (and possibly others), that +# the stock LLVM toolchain doesn't know about and will warn about. Let's not +# have test fail just because of processor name differences due to different +# LLVM Toolchains. GHC tries to pass what apple expects (on darwin), but can +# be used with the stock LLVM toolchain as well. +def ignore_llvm_and_vortex( msg ): + return re.sub(r".* is not a recognized processor for this target.*\n",r"",msg) + +test('T22487', [normal, normalise_errmsg_fun(ignore_llvm_and_vortex)], compile_and_run, ['']) ===================================== utils/llvm-targets/gen-data-layout.sh ===================================== @@ -26,7 +26,7 @@ TARGETS=( ######################### # Windows - "x86_64-unknown-windows" + "x86_64-unknown-windows-gnu" ######################### # Linux View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9e96dad809667c777ebb86983a6e8b372d605e17...992a7624817cdab4b566387f2e877f420f7b27d4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9e96dad809667c777ebb86983a6e8b372d605e17...992a7624817cdab4b566387f2e877f420f7b27d4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:50:02 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 17:50:02 -0400 Subject: [Git][ghc/ghc][master] compiler: Use type abstractions when deriving Message-ID: <66ef3f8a34de2_23cb2797fb08107629@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - 15 changed files: - compiler/GHC/Hs/Pat.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - testsuite/tests/deriving/should_compile/T14578.stderr - testsuite/tests/deriving/should_compile/T14579.stderr - testsuite/tests/deriving/should_compile/all.T - + testsuite/tests/deriving/should_compile/deriving-inferred-ty-arg.hs - + testsuite/tests/deriving/should_compile/deriving-inferred-ty-arg.stderr - testsuite/tests/ghc-e/should_run/ghc-e005.stderr - utils/check-exact/ExactPrint.hs Changes: ===================================== compiler/GHC/Hs/Pat.hs ===================================== @@ -192,8 +192,11 @@ type instance XHsRecFields GhcTc = MultiplicityCheckCoercions type instance XHsFieldBind _ = [AddEpAnn] -type instance XInvisPat GhcPs = EpToken "@" -type instance XInvisPat GhcRn = NoExtField +-- The specificity of an invisible pattern from the parser is always +-- SpecifiedSpec. The specificity field supports code generated when deriving +-- newtype or via; see Note [Inferred invisible patterns]. +type instance XInvisPat GhcPs = (EpToken "@", Specificity) +type instance XInvisPat GhcRn = Specificity type instance XInvisPat GhcTc = Type @@ -474,7 +477,17 @@ pprPat (ConPat { pat_con = con , cpt_binds = binds } = ext pprPat (EmbTyPat _ tp) = text "type" <+> ppr tp -pprPat (InvisPat _ tp) = char '@' <> ppr tp +pprPat (InvisPat x tp) = char '@' <> delimit (ppr tp) + where + delimit + | inferred = braces + | needs_parens = parens + | otherwise = id + inferred = case ghcPass @p of + GhcPs -> snd x == InferredSpec + GhcRn -> x == InferredSpec + GhcTc -> False + needs_parens = hsTypeNeedsParens appPrec $ unLoc $ hstp_body tp pprPat (XPat ext) = case ghcPass @p of GhcRn -> case ext of ===================================== compiler/GHC/Parser.y ===================================== @@ -3513,7 +3513,7 @@ bindpat : exp {% -- See Note [Parser-Validator Details] in GHC.Parse argpat :: { LPat GhcPs } argpat : apat { $1 } - | PREFIX_AT atype { sLLa $1 $> (InvisPat (epTok $1) (mkHsTyPat $2)) } + | PREFIX_AT atype { sLLa $1 $> (InvisPat (epTok $1, SpecifiedSpec) (mkHsTyPat $2)) } argpats :: { [LPat GhcPs] } : argpat argpats { $1 : $2 } ===================================== compiler/GHC/Parser/PostProcess.hs ===================================== @@ -1522,7 +1522,7 @@ isFunLhs e = go e [] [] [] reassociate _other = Nothing go (L l (PatBuilderAppType (L lp pat) tok ty_pat@(HsTP _ (L (EpAnn anc ann cs) _)))) es ops cps = go (L lp' pat) (L (EpAnn anc' ann cs) (ArgPatBuilderArgPat invis_pat) : es) ops cps - where invis_pat = InvisPat tok ty_pat + where invis_pat = InvisPat (tok, SpecifiedSpec) ty_pat anc' = case tok of NoEpTok -> anc EpTok l -> widenAnchor anc [AddEpAnn AnnAnyclass l] ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -507,11 +507,11 @@ rnBindPat name_maker pat = runCps (rnLPatAndThen name_maker pat) rnLArgPatAndThen :: NameMaker -> LocatedA (Pat GhcPs) -> CpsRn (LocatedA (Pat GhcRn)) rnLArgPatAndThen mk = wrapSrcSpanCps rnArgPatAndThen where - rnArgPatAndThen (InvisPat _ tp) = do + rnArgPatAndThen (InvisPat (_, spec) tp) = do liftCps $ unlessXOptM LangExt.TypeAbstractions $ addErr (TcRnIllegalInvisibleTypePattern tp) tp' <- rnHsTyPat HsTypePatCtx tp - pure (InvisPat noExtField tp') + pure (InvisPat spec tp') rnArgPatAndThen p = rnPatAndThen mk p -- ----------- Entry point 3: rnLPatAndThen ------------------- @@ -677,12 +677,12 @@ rnPatAndThen mk (SplicePat _ splice) rnPatAndThen _ (EmbTyPat _ tp) = do { tp' <- rnHsTyPat HsTypePatCtx tp ; return (EmbTyPat noExtField tp') } -rnPatAndThen _ (InvisPat _ tp) +rnPatAndThen _ (InvisPat (_, spec) tp) = do { liftCps $ addErr (TcRnMisplacedInvisPat tp) -- Invisible patterns are handled in `rnLArgPatAndThen` -- so unconditionally emit error here ; tp' <- rnHsTyPat HsTypePatCtx tp - ; return (InvisPat noExtField tp') + ; return (InvisPat spec tp') } -------------------- ===================================== compiler/GHC/Tc/Deriv.hs ===================================== @@ -281,8 +281,9 @@ renameDeriv inst_infos bagBinds setXOptM LangExt.KindSignatures $ -- Derived decls (for newtype-deriving) can use ScopedTypeVariables & -- KindSignatures + setXOptM LangExt.TypeAbstractions $ setXOptM LangExt.TypeApplications $ - -- GND/DerivingVia uses TypeApplications in generated code + -- GND/DerivingVia uses TypeAbstractions & TypeApplications in generated code -- (See Note [Newtype-deriving instances] in GHC.Tc.Deriv.Generate) unsetXOptM LangExt.RebindableSyntax $ -- See Note [Avoid RebindableSyntax when deriving] @@ -1971,12 +1972,9 @@ genInstBinds spec@(DS { ds_tvs = tyvars, ds_mechanism = mechanism extensions | isDerivSpecNewtype mechanism || isDerivSpecVia mechanism = [ - -- Both these flags are needed for higher-rank uses of coerce... - LangExt.ImpredicativeTypes, LangExt.RankNTypes - -- ...and this flag is needed to support the instance signatures - -- that bring type variables into scope. + -- Both these flags are needed for higher-rank uses of coerce -- See Note [Newtype-deriving instances] in GHC.Tc.Deriv.Generate - , LangExt.InstanceSigs + LangExt.ImpredicativeTypes, LangExt.RankNTypes -- Skip unboxed tuples checking for derived instances when imported -- in a different module, see #20524 , LangExt.UnboxedTuples @@ -2010,8 +2008,8 @@ genInstBinds spec@(DS { ds_tvs = tyvars, ds_mechanism = mechanism -> gen_newtype_or_via via_ty gen_newtype_or_via ty = do - let (binds, sigs) = gen_Newtype_binds loc clas tyvars inst_tys ty - return (binds, sigs, emptyBag, []) + let binds = gen_Newtype_binds loc clas tyvars inst_tys ty + return (binds, [], emptyBag, []) -- | Generate the associated type family instances for a derived instance. genFamInsts :: DerivSpec theta -> TcM [FamInst] ===================================== compiler/GHC/Tc/Deriv/Generate.hs ===================================== @@ -1697,17 +1697,21 @@ coercing from. So from, say, newtype T x = MkT instance C a => C a (T x) where - op :: forall c. a -> [T x] -> c -> Int - op = coerce @(a -> [] -> c -> Int) - @(a -> [T x] -> c -> Int) - op + op @c = coerce @(a -> [] -> c -> Int) + @(a -> [T x] -> c -> Int) + op -In addition to the type applications, we also have an explicit -type signature on the entire RHS. This brings the method-bound variable -`c` into scope over the two type applications. +In addition to the type applications, we also use a type abstraction to bring +the method-bound variable `c` into scope over the two type applications. See Note [GND and QuantifiedConstraints] for more information on why this is important. +(In the surface syntax, only specified type variables can be used in type +abstractions. Since a method signature could contain both specified and +inferred type variables, we need an internal-only way to represent the inferred +case. We handle this by smuggling a Specificity field in XInvisPat. See +Note [Inferred invisible patterns].) + Giving 'coerce' two explicitly-visible type arguments grants us finer control over how it should be instantiated. Recall @@ -1720,7 +1724,6 @@ a polytype. E.g. class C a where op :: a -> forall b. b -> b newtype T x = MkT instance C => C (T x) where - op :: T x -> forall b. b -> b op = coerce @( -> forall b. b -> b) @(T x -> forall b. b -> b) op @@ -1734,6 +1737,74 @@ However, to allow VTA with polytypes we must switch on -XImpredicativeTypes locally in GHC.Tc.Deriv.genInst. See #8503 for more discussion. +Note [Inferred invisible patterns] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Consider the following: + + class R a where + r :: forall b. Proxy b -> a + +When newtype-deriving an instance of `R`, following +Note [GND and QuantifiedConstraints], we might generate the following code: + + instance R => R where + r = \ @b -> coerce @(Proxy b -> ) + @(Proxy b -> ) + r + +The code being generated is an HsSyn AST, except for the arguments to coerce, +which are XHsTypes carrying Core types. As Core types, they must be fully +elaborated, so we actually want something more like the following: + + instance R => R where + r = \ @b -> coerce @(Proxy @{k} b -> ) + @(Proxy @{k} b -> ) + r + +where the `k` corresponds to the `k` in the elaborated type of `r`: + + class R (a :: Type) where + r :: forall {k :: Type} (b :: k). Proxy @{k} b -> a + +However, `k` is not bound in the definition of `r` in the derived instance, and +binding it requires a way to create an inferred (because `k` is inferred in the +signature of `r`) invisible pattern. + +So we actually generate the following for `R`: + + instance R => R where + r = \ @{k :: Type} -> \ @(b :: k) -> + coerce @(Proxy @{k} b -> ) + @(Proxy @{k} b -> ) + r + +The `\ @{k :: Type} ->` (note the braces!) is the big lambda that binds `k`, and +represents an inferred invisible pattern. Inferred invisible patterns aren't +allowed in the surface syntax of Haskell, for the reason that the order in +which inferred foralls are added to a signature is not specified, so it is +ambiguous which pattern would bind to which forall. But when deriving an +instance, the patterns are being created after the type of the method has been +elaborated, so an order for the inferred foralls has already been determined. +This makes inferred invisible patterns safe for internal use. + +(You might wonder if you could bring `k` into scope via the pattern signature +in `\ @(b :: k)`, but that does not work in general; e.g. if +`r :: Proxy Any -> a`; see `C5` in test `deriving-inferred-ty-arg`.) + +The implementation is straightforward: we have a Specificity field in +XInvisPat, which is always SpecifiedSpec when coming from the parser or +Template Haskell, but takes the specificity of the corresponding forall from +the method type during instance deriving. When type checking an invisible +pattern, we allow inferred patterns to bind inferred foralls just like we allow +specified patterns to bind specified foralls. + +More discussion of this scenario and some rejected alternatives at +https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13190 + +See also https://github.com/ghc-proposals/ghc-proposals/pull/675, which +was triggered by this ticket, and explores source-language syntax in this +space. + Note [Newtype-deriving trickiness] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Consider (#12768): @@ -1805,16 +1876,15 @@ See Note [Instances in no-evidence implications] in GHC.Tc.Solver.Equality. But this isn't the death knell for combining QuantifiedConstraints with GND. On the contrary, if we generate GND bindings in a slightly different way, then we can avoid this situation altogether. Instead of applying `coerce` to two -polymorphic types, we instead let an instance signature do the polymorphic -instantiation, and omit the `forall`s in the type applications. -More concretely, we generate the following code instead: +polymorphic types, we instead use a type abstraction to bind the type +variables, and omit the `forall`s in the type applications. More concretely, we +generate the following code instead: instance (C m, forall p q. Coercible p q => Coercible (m p) (m q)) => C (T m) where - join :: forall a. T m (T m a) -> T m a - join = coerce @( m (m a) -> m a) - @(T m (T m a) -> T m a) - join + join @a = coerce @( m (m a) -> m a) + @(T m (T m a) -> T m a) + join Now the visible type arguments are both monotypes, so we don't need any of this funny quantified constraint instantiation business. While this particular @@ -1823,120 +1893,25 @@ ImpredicativeTypes to typecheck GND-generated code for class methods with higher-rank types. See Note [Newtype-deriving instances]. You might think that that second @(T m (T m a) -> T m a) argument is redundant -in the presence of the instance signature, but in fact leaving it off will -break this example (from the T15290d test case): +with the type information provided by the class, but in fact leaving it off +will break the following example (from the T12616 test case): + + type m ~> n = forall a. m a -> n a + data StateT s m a = ... + newtype OtherStateT s m a = OtherStateT (StateT s m a) - class C a where - c :: Int -> forall b. b -> a + class MonadTrans t where + lift :: (Monad m) => m ~> t m - instance C Int + instance MonadTrans (StateT s) - instance C Age where - c :: Int -> forall b. b -> Age - c = coerce @(Int -> forall b. b -> Int) - c + instance MonadTrans (OtherStateT s) where + lift @m = coerce @(m ~> StateT s m) + lift That is because we still need to instantiate the second argument of coerce with a polytype, and we can only do that with VTA or QuickLook. -Be aware that the use of an instance signature doesn't /solve/ this -problem; it just makes it less likely to occur. For example, if a class has -a truly higher-rank type like so: - - class CProblem m where - op :: (forall b. ... (m b) ...) -> Int - -Then the same situation will arise again. But at least it won't arise for the -common case of methods with ordinary, prenex-quantified types. - ------ --- Wrinkle: Use HsOuterExplicit ------ - -One minor complication with the plan above is that we need to ensure that the -type variables from a method's instance signature properly scope over the body -of the method. For example, recall: - - instance (C m, forall p q. Coercible p q => Coercible (m p) (m q)) => - C (T m) where - join :: forall a. T m (T m a) -> T m a - join = coerce @( m (m a) -> m a) - @(T m (T m a) -> T m a) - join - -In the example above, it is imperative that the `a` in the instance signature -for `join` scope over the body of `join` by way of ScopedTypeVariables. -This might sound obvious, but note that in gen_Newtype_binds, which is -responsible for generating the code above, the type in `join`'s instance -signature is given as a Core type, whereas gen_Newtype_binds will eventually -produce HsBinds (i.e., source Haskell) that is renamed and typechecked. We -must ensure that `a` is in scope over the body of `join` during renaming -or else the generated code will be rejected. - -In short, we need to convert the instance signature from a Core type to an -HsType (i.e., a source Haskell type). Two possible options are: - -1. Convert the Core type entirely to an HsType (i.e., a source Haskell type). -2. Embed the entire Core type using HsCoreTy. - -Neither option is quite satisfactory: - -1. Converting a Core type to an HsType in full generality is surprisingly - complicated. Previous versions of GHCs did this, but it was the source of - numerous bugs (see #14579 and #16518, for instance). -2. While HsCoreTy is much less complicated that option (1), it's not quite - what we want. In order for `a` to be in scope over the body of `join` during - renaming, the `forall` must be contained in an HsOuterExplicit. - (See Note [Lexically scoped type variables] in GHC.Hs.Type.) HsCoreTy - bypasses HsOuterExplicit, so this won't work either. - -As a compromise, we adopt a combination of the two options above: - -* Split apart the top-level ForAllTys in the instance signature's Core type, -* Convert the top-level ForAllTys to an HsOuterExplicit, and -* Embed the remainder of the Core type in an HsCoreTy. - -This retains most of the simplicity of option (2) while still ensuring that -the type variables are correctly scoped. - -Note that splitting apart top-level ForAllTys will expand any type synonyms -in the Core type itself. This ends up being important to fix a corner case -observed in #18914. Consider this example: - - type T f = forall a. f a - - class C f where - m :: T f - - newtype N f a = MkN (f a) - deriving C - -What code should `deriving C` generate? It will have roughly the following -shape: - - instance C f => C (N f) where - m :: T (N f) - m = coerce @(...) (...) (m @f) - -At a minimum, we must instantiate `coerce` with `@(T f)` and `@(T (N f))`, but -with the `forall`s removed in order to make them monotypes. However, the -`forall` is hidden underneath the `T` type synonym, so we must first expand `T` -before we can strip of the `forall`. Expanding `T`, we get -`coerce @(forall a. f a) @(forall a. N f a)`, and after omitting the `forall`s, -we get `coerce @(f a) @(N f a)`. - -We can't stop there, however, or else we would end up with this code: - - instance C f => C (N f) where - m :: T (N f) - m = coerce @(f a) @(N f a) (m @f) - -Notice that the type variable `a` is completely unbound. In order to make sure -that `a` is in scope, we must /also/ expand the `T` in `m :: T (N f)` to get -`m :: forall a. N f a`. Fortunately, we will do just that in the plan outlined -above, since when we split off the top-level ForAllTys in the instance -signature, we must first expand the T type synonym. - Note [GND and ambiguity] ~~~~~~~~~~~~~~~~~~~~~~~~ We make an effort to make the code generated through GND be robust w.r.t. @@ -1950,7 +1925,6 @@ ambiguous type variables. As one example, consider the following example A naïve attempt and generating a C T instance would be: instance C T where - f :: String f = coerce @String @String f This isn't going to typecheck, however, since GHC doesn't know what to @@ -1959,7 +1933,6 @@ instantiate the type variable `a` with in the call to `f` in the method body. ambiguity here, we explicitly instantiate `a` like so: instance C T where - f :: String f = coerce @String @String (f @()) All better now. @@ -1972,22 +1945,19 @@ gen_Newtype_binds :: SrcSpan -- newtype itself) -> [Type] -- instance head parameters (incl. newtype) -> Type -- the representation type - -> (LHsBinds GhcPs, [LSig GhcPs]) + -> LHsBinds GhcPs -- See Note [Newtype-deriving instances] gen_Newtype_binds loc' cls inst_tvs inst_tys rhs_ty - = (binds, sigs) + = map mk_bind (classMethods cls) where - (binds, sigs) = mapAndUnzip mk_bind_and_sig (classMethods cls) - -- Same as inst_tys, but with the last argument type replaced by the -- representation type. underlying_inst_tys :: [Type] underlying_inst_tys = changeLast inst_tys rhs_ty locn = noAnnSrcSpan loc' - loca = noAnnSrcSpan loc' - -- For each class method, generate its derived binding and instance - -- signature. Using the first example from + -- For each class method, generate its derived binding. Using the first + -- example from -- Note [Newtype-deriving instances]: -- -- class C a b where @@ -1999,43 +1969,30 @@ gen_Newtype_binds loc' cls inst_tvs inst_tys rhs_ty -- -- instance C a => C a (T x) where -- - mk_bind_and_sig :: Id -> (LHsBind GhcPs, LSig GhcPs) - mk_bind_and_sig meth_id - = ( -- The derived binding, e.g., - -- - -- op = coerce @(a -> [] -> c -> Int) - -- @(a -> [T x] -> c -> Int) - -- op - mkRdrFunBind loc_meth_RDR [mkSimpleMatch - (mkPrefixFunRhs loc_meth_RDR) - (noLocA []) rhs_expr] - , -- The derived instance signature, e.g., - -- - -- op :: forall c. a -> [T x] -> c -> Int - -- - -- Make sure that `forall c` is in an HsOuterExplicit so that it - -- scopes over the body of `op`. See "Wrinkle: Use HsOuterExplicit" in - -- Note [GND and QuantifiedConstraints]. - L loca $ ClassOpSig noAnn False [loc_meth_RDR] - $ L loca $ mkHsExplicitSigType noAnn - (map mk_hs_tvb to_tvbs) - (nlHsCoreTy to_rho) - ) + mk_bind :: Id -> LHsBind GhcPs + mk_bind meth_id + = -- The derived binding, e.g., + -- + -- op @c = coerce @(a -> [] -> c -> Int) + -- @(a -> [T x] -> c -> Int) + -- op + mkRdrFunBind loc_meth_RDR [mkSimpleMatch + (mkPrefixFunRhs loc_meth_RDR) + (noLocA (map mk_ty_pat to_tvbs)) rhs_expr] + where Pair from_ty to_ty = mkCoerceClassMethEqn cls inst_tvs inst_tys rhs_ty meth_id (_, _, from_tau) = tcSplitSigmaTy from_ty (to_tvbs, to_rho) = tcSplitForAllInvisTVBinders to_ty (_, to_tau) = tcSplitPhiTy to_rho - -- The use of tcSplitForAllInvisTVBinders above expands type synonyms, - -- which is important to ensure correct type variable scoping. - -- See "Wrinkle: Use HsOuterExplicit" in - -- Note [GND and QuantifiedConstraints]. - - mk_hs_tvb :: VarBndr TyVar flag -> LHsTyVarBndr flag GhcPs - mk_hs_tvb (Bndr tv flag) = noLocA $ KindedTyVar noAnn - flag - (noLocA (getRdrName tv)) - (nlHsCoreTy (tyVarKind tv)) + -- The `to_tvbs` bind variables that are mentioned in `to_rho` and + -- hence in `to_tau`. So we bring `to_tvbs` into scope via the + -- `mkSimpleMatch` above, so that their use in `to_tau` in `rhs_expr` + -- is well-scoped. + + mk_ty_pat :: VarBndr TyVar Specificity -> LPat GhcPs + mk_ty_pat (Bndr tv spec) = noLocA $ InvisPat (noAnn, spec) $ mkHsTyPat $ + nlHsTyVar NotPromoted $ getRdrName tv meth_RDR = getRdrName meth_id loc_meth_RDR = L locn meth_RDR ===================================== compiler/GHC/Tc/Gen/Pat.hs ===================================== @@ -175,8 +175,11 @@ tcMatchPats match_ctxt pats pat_tys thing_inside -- E.g. f :: forall a. Bool -> a -> blah -- f @b True x = rhs1 -- b is bound to skolem a -- f @c False y = rhs2 -- c is bound to skolem a - | L _ (InvisPat _ tp) <- pat - , isSpecifiedForAllTyFlag vis + -- Also handles invisible (Inferred) case originating from type + -- class deriving; see Note [Inferred invisible patterns] + | L _ (InvisPat pat_spec tp) <- pat + , Invisible spec <- vis + , pat_spec == spec = do { (_p, (ps, res)) <- tc_ty_pat tp tv $ loop pats pat_tys ; return (ps, res) } ===================================== compiler/GHC/ThToHs.hs ===================================== @@ -1520,10 +1520,13 @@ cvtp (ViewP e p) = do { e' <- cvtl e; p' <- cvtPat p cvtp (TypeP t) = do { t' <- cvtType t ; return $ EmbTyPat noAnn (mkHsTyPat t') } cvtp (InvisP t) = do { t' <- parenthesizeHsType appPrec <$> cvtType t - ; pure (InvisPat noAnn (mkHsTyPat t'))} + ; pure (InvisPat noAnnSpecified (mkHsTyPat t'))} cvtp (OrP ps) = do { ps' <- cvtPats ps ; pure (OrPat noExtField ps')} +noAnnSpecified :: XInvisPat GhcPs +noAnnSpecified = (noAnn, Hs.SpecifiedSpec) + cvtPatFld :: (TH.Name, TH.Pat) -> CvtM (LHsRecField GhcPs (LPat GhcPs)) cvtPatFld (s,p) = do { L ls s' <- vNameN s ===================================== testsuite/tests/deriving/should_compile/T14578.stderr ===================================== @@ -4,13 +4,6 @@ Derived class instances: instance (GHC.Internal.Base.Applicative f, GHC.Internal.Base.Applicative g, GHC.Internal.Base.Semigroup a) => GHC.Internal.Base.Semigroup (T14578.Wat f g a) where - (GHC.Internal.Base.<>) :: - T14578.Wat f g a -> T14578.Wat f g a -> T14578.Wat f g a - GHC.Internal.Base.sconcat :: - GHC.Internal.Base.NonEmpty (T14578.Wat f g a) -> T14578.Wat f g a - GHC.Internal.Base.stimes :: - forall (b :: *). GHC.Internal.Real.Integral b => - b -> T14578.Wat f g a -> T14578.Wat f g a (GHC.Internal.Base.<>) = GHC.Prim.coerce @(T14578.App (Data.Functor.Compose.Compose f g) a @@ -28,7 +21,7 @@ Derived class instances: -> T14578.Wat f g a) (GHC.Internal.Base.sconcat @(T14578.App (Data.Functor.Compose.Compose f g) a)) - GHC.Internal.Base.stimes + GHC.Internal.Base.stimes @b = GHC.Prim.coerce @(b -> T14578.App (Data.Functor.Compose.Compose f g) a @@ -39,56 +32,38 @@ Derived class instances: instance GHC.Internal.Base.Functor f => GHC.Internal.Base.Functor (T14578.App f) where - GHC.Internal.Base.fmap :: - forall (a :: *) (b :: *). (a -> b) - -> T14578.App f a -> T14578.App f b - (GHC.Internal.Base.<$) :: - forall (a :: *) (b :: *). a -> T14578.App f b -> T14578.App f a - GHC.Internal.Base.fmap + GHC.Internal.Base.fmap @a @b = GHC.Prim.coerce @((a -> b) -> f a -> f b) @((a -> b) -> T14578.App f a -> T14578.App f b) (GHC.Internal.Base.fmap @f) - (GHC.Internal.Base.<$) + (GHC.Internal.Base.<$) @a @b = GHC.Prim.coerce @(a -> f b -> f a) @(a -> T14578.App f b -> T14578.App f a) ((GHC.Internal.Base.<$) @f) instance GHC.Internal.Base.Applicative f => GHC.Internal.Base.Applicative (T14578.App f) where - GHC.Internal.Base.pure :: forall (a :: *). a -> T14578.App f a - (GHC.Internal.Base.<*>) :: - forall (a :: *) (b :: *). T14578.App f (a -> b) - -> T14578.App f a -> T14578.App f b - GHC.Internal.Base.liftA2 :: - forall (a :: *) (b :: *) (c :: *). (a -> b -> c) - -> T14578.App f a -> T14578.App f b -> T14578.App f c - (GHC.Internal.Base.*>) :: - forall (a :: *) (b :: *). T14578.App f a - -> T14578.App f b -> T14578.App f b - (GHC.Internal.Base.<*) :: - forall (a :: *) (b :: *). T14578.App f a - -> T14578.App f b -> T14578.App f a - GHC.Internal.Base.pure + GHC.Internal.Base.pure @a = GHC.Prim.coerce @(a -> f a) @(a -> T14578.App f a) (GHC.Internal.Base.pure @f) - (GHC.Internal.Base.<*>) + (GHC.Internal.Base.<*>) @a @b = GHC.Prim.coerce @(f (a -> b) -> f a -> f b) @(T14578.App f (a -> b) -> T14578.App f a -> T14578.App f b) ((GHC.Internal.Base.<*>) @f) - GHC.Internal.Base.liftA2 + GHC.Internal.Base.liftA2 @a @b @c = GHC.Prim.coerce @((a -> b -> c) -> f a -> f b -> f c) @((a -> b -> c) -> T14578.App f a -> T14578.App f b -> T14578.App f c) (GHC.Internal.Base.liftA2 @f) - (GHC.Internal.Base.*>) + (GHC.Internal.Base.*>) @a @b = GHC.Prim.coerce @(f a -> f b -> f b) @(T14578.App f a -> T14578.App f b -> T14578.App f b) ((GHC.Internal.Base.*>) @f) - (GHC.Internal.Base.<*) + (GHC.Internal.Base.<*) @a @b = GHC.Prim.coerce @(f a -> f b -> f a) @(T14578.App f a -> T14578.App f b -> T14578.App f a) ===================================== testsuite/tests/deriving/should_compile/T14579.stderr ===================================== @@ -2,10 +2,6 @@ ==================== Derived instances ==================== Derived class instances: instance GHC.Classes.Eq a => GHC.Classes.Eq (T14579.Glurp a) where - (GHC.Classes.==) :: - T14579.Glurp a -> T14579.Glurp a -> GHC.Types.Bool - (GHC.Classes./=) :: - T14579.Glurp a -> T14579.Glurp a -> GHC.Types.Bool (GHC.Classes.==) = GHC.Prim.coerce @(T14579.Wat 'GHC.Internal.Data.Proxy.Proxy @@ -22,8 +18,6 @@ Derived class instances: instance forall a (x :: GHC.Internal.Data.Proxy.Proxy a). GHC.Classes.Eq a => GHC.Classes.Eq (T14579.Wat x) where - (GHC.Classes.==) :: T14579.Wat x -> T14579.Wat x -> GHC.Types.Bool - (GHC.Classes./=) :: T14579.Wat x -> T14579.Wat x -> GHC.Types.Bool (GHC.Classes.==) = GHC.Prim.coerce @(GHC.Internal.Maybe.Maybe a ===================================== testsuite/tests/deriving/should_compile/all.T ===================================== @@ -150,3 +150,4 @@ test('T15798c', normal, compile, ['']) test('T24955a', normal, compile, ['']) test('T24955b', normal, compile, ['']) test('T24955c', normal, compile, ['']) +test('deriving-inferred-ty-arg', normal, compile, ['-ddump-deriv -dsuppress-uniques']) ===================================== testsuite/tests/deriving/should_compile/deriving-inferred-ty-arg.hs ===================================== @@ -0,0 +1,27 @@ +{-# LANGUAGE DataKinds, TypeFamilies #-} +module DerivingInferredTyArg where + +import Data.Functor.Const +import Data.Proxy + +class C a where + m :: forall {k} (b :: k). Const a b + +class C2 a where + m2 :: forall {k} {b :: k}. Const a b + +class C3 a where + m3 :: forall {k} (b :: k) {p :: Proxy b}. Const a p + +data VisProxy k (a :: k) = VisProxy + +class C4 a where + m4 :: forall {k} (b :: k) {p :: VisProxy k b}. Const a p + +type family Any :: k + +class C5 a where + m5 :: Proxy Any -> a + +newtype T a = MkT a + deriving (C, C2, C3, C4, C5) ===================================== testsuite/tests/deriving/should_compile/deriving-inferred-ty-arg.stderr ===================================== @@ -0,0 +1,52 @@ + +==================== Derived instances ==================== +Derived class instances: + instance DerivingInferredTyArg.C a => + DerivingInferredTyArg.C (DerivingInferredTyArg.T a) where + DerivingInferredTyArg.m @{k} @b + = GHC.Prim.coerce + @(GHC.Internal.Data.Functor.Const.Const a b) + @(GHC.Internal.Data.Functor.Const.Const + (DerivingInferredTyArg.T a) b) + (DerivingInferredTyArg.m @a) + + instance DerivingInferredTyArg.C2 a => + DerivingInferredTyArg.C2 (DerivingInferredTyArg.T a) where + DerivingInferredTyArg.m2 @{k} @{b} + = GHC.Prim.coerce + @(GHC.Internal.Data.Functor.Const.Const a b) + @(GHC.Internal.Data.Functor.Const.Const + (DerivingInferredTyArg.T a) b) + (DerivingInferredTyArg.m2 @a) + + instance DerivingInferredTyArg.C3 a => + DerivingInferredTyArg.C3 (DerivingInferredTyArg.T a) where + DerivingInferredTyArg.m3 @{k} @b @{p} + = GHC.Prim.coerce + @(GHC.Internal.Data.Functor.Const.Const a p) + @(GHC.Internal.Data.Functor.Const.Const + (DerivingInferredTyArg.T a) p) + (DerivingInferredTyArg.m3 @a) + + instance DerivingInferredTyArg.C4 a => + DerivingInferredTyArg.C4 (DerivingInferredTyArg.T a) where + DerivingInferredTyArg.m4 @{k} @b @{p} + = GHC.Prim.coerce + @(GHC.Internal.Data.Functor.Const.Const a p) + @(GHC.Internal.Data.Functor.Const.Const + (DerivingInferredTyArg.T a) p) + (DerivingInferredTyArg.m4 @a) + + instance DerivingInferredTyArg.C5 a => + DerivingInferredTyArg.C5 (DerivingInferredTyArg.T a) where + DerivingInferredTyArg.m5 @{k} + = GHC.Prim.coerce + @(GHC.Internal.Data.Proxy.Proxy DerivingInferredTyArg.Any -> a) + @(GHC.Internal.Data.Proxy.Proxy DerivingInferredTyArg.Any + -> DerivingInferredTyArg.T a) + (DerivingInferredTyArg.m5 @a) + + +Derived type family instances: + + ===================================== testsuite/tests/ghc-e/should_run/ghc-e005.stderr ===================================== @@ -9,23 +9,18 @@ Module: GHC.Internal.Exception Type: ErrorCall HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO - throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at compiler/GHC/Driver/Monad.hs:167:54 in ghc-9.9-inplace:GHC.Driver.Monad - a type signature in an instance, called at compiler/GHC/Driver/Monad.hs:167:54 in ghc-9.9-inplace:GHC.Driver.Monad - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at ghc/GHCi/UI/Monad.hs:288:15 in ghc-bin-9.9.20240223-inplace:GHCi.UI.Monad - a type signature in an instance, called at ghc/GHCi/UI/Monad.hs:288:15 in ghc-bin-9.9.20240223-inplace:GHCi.UI.Monad - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch - throwM, called at libraries/haskeline/System/Console/Haskeline/InputT.hs:53:39 in haskeline-0.8.2.1-inplace:System.Console.Haskeline.InputT - a type signature in an instance, called at libraries/haskeline/System/Console/Haskeline/InputT.hs:53:39 in haskeline-0.8.2.1-inplace:System.Console.Haskeline.InputT - throwM, called at ghc/GHCi/UI/Monad.hs:215:52 in ghc-bin-9.9.20240223-inplace:GHCi.UI.Monad - - + collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch + throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + throwM, called at compiler/GHC/Driver/Monad.hs:167:54 in ghc-9.11-inplace:GHC.Driver.Monad + throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + throwM, called at ghc/GHCi/UI/Monad.hs:288:15 in ghc-bin-9.11.20240823-inplace:GHCi.UI.Monad + throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + throwM, called at libraries/haskeline/System/Console/Haskeline/InputT.hs:53:39 in haskeline-0.8.2.1-inplace:System.Console.Haskeline.InputT + throwM, called at ghc/GHCi/UI/Monad.hs:215:52 in ghc-bin-9.11.20240823-inplace:GHCi.UI.Monad ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -4877,10 +4877,10 @@ instance ExactPrint (Pat GhcPs) where tp' <- markAnnotated tp return (EmbTyPat toktype' tp') - exact (InvisPat tokat tp) = do + exact (InvisPat (tokat, spec) tp) = do tokat' <- markEpToken tokat tp' <- markAnnotated tp - pure (InvisPat tokat' tp') + pure (InvisPat (tokat', spec) tp') -- --------------------------------------------------------------------- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c50d29be95247b92f23386d4aa61d8bb9bfc767e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c50d29be95247b92f23386d4aa61d8bb9bfc767e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:51:06 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 17:51:06 -0400 Subject: [Git][ghc/ghc][master] driver: Ensure we run driverPlugin for staticPlugins (#25217) Message-ID: <66ef3fca15fd2_23cb27b7c26c114376@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 6 changed files: - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Runtime/Loader.hs - + testsuite/tests/plugins/T25217.hs - + testsuite/tests/plugins/T25217.stdout - testsuite/tests/plugins/all.T - testsuite/tests/plugins/static-plugins.hs Changes: ===================================== compiler/GHC/Driver/Plugins.hs ===================================== @@ -244,6 +244,8 @@ data ExternalPlugin = ExternalPlugin data StaticPlugin = StaticPlugin { spPlugin :: PluginWithArgs -- ^ the actual plugin together with its commandline arguments + , spInitialised :: Bool + -- ^ has this plugin been initialised (i.e. driverPlugin has been run) } lpModuleName :: LoadedPlugin -> ModuleName ===================================== compiler/GHC/Runtime/Loader.hs ===================================== @@ -121,14 +121,15 @@ initializePlugins hsc_env , external_plugins <- externalPlugins (hsc_plugins hsc_env) , check_external_plugins external_plugins (externalPluginSpecs dflags) - -- FIXME: we should check static plugins too + -- ensure we have initialised static plugins + , all spInitialised (staticPlugins (hsc_plugins hsc_env)) = return hsc_env -- no change, no need to reload plugins | otherwise = do (loaded_plugins, links, pkgs) <- loadPlugins hsc_env external_plugins <- loadExternalPlugins (externalPluginSpecs dflags) - let plugins' = (hsc_plugins hsc_env) { staticPlugins = staticPlugins (hsc_plugins hsc_env) + let plugins' = (hsc_plugins hsc_env) { staticPlugins = map (\sp -> sp{ spInitialised = True }) $ staticPlugins (hsc_plugins hsc_env) , externalPlugins = external_plugins , loadedPlugins = loaded_plugins , loadedPluginDeps = (links, pkgs) ===================================== testsuite/tests/plugins/T25217.hs ===================================== @@ -0,0 +1,42 @@ +module Main where + +import Control.Monad.IO.Class +import GHC +import GHC.Driver.Monad +import GHC.Plugins +import System.Environment + +main = do + libdir:args <- getArgs + defaultErrorHandler defaultFatalMessager defaultFlushOut $ do + runGhc (Just libdir) $ do + dflags <- getSessionDynFlags + -- we need to LinkInMemory otherwise `setTarget [] >> load LoadAllTargets` + -- below will fail. + setSessionDynFlags dflags { ghcLink = LinkInMemory} + + liftIO $ putStrLn "loading with driver plugin" + loadWithPlugins [StaticPlugin (PluginWithArgs plugin []) False] + + + where + loadWithPlugins the_plugins = do + -- first unload (like GHCi :load does) + GHC.setTargets [] + _ <- GHC.load LoadAllTargets + + target <- guessTarget "static-plugins-module.hs" Nothing Nothing + setTargets [target] + + modifySession $ \hsc_env -> + let old_plugins = hsc_plugins hsc_env + in hsc_env { hsc_plugins = old_plugins { staticPlugins = the_plugins } } + + dflags <- getSessionDynFlags + setSessionDynFlags dflags { outputFile_ = Nothing } + load LoadAllTargets + liftIO (putStrLn "loading done") + + +plugin = + defaultPlugin { driverPlugin = \_ env -> liftIO (putStrLn "driver plugin ran") >> pure env } ===================================== testsuite/tests/plugins/T25217.stdout ===================================== @@ -0,0 +1,3 @@ +loading with driver plugin +driver plugin ran +loading done ===================================== testsuite/tests/plugins/all.T ===================================== @@ -200,6 +200,15 @@ test('static-plugins', compile_and_run, ['-package ghc -isimple-plugin/ -j1']) +test('T25217', + [extra_files(['static-plugins-module.hs']), + unless(config.have_RTS_linker, skip), + expect_broken_for(16803, prof_ways), + extra_run_opts('"' + config.libdir + '"'), + ], + compile_and_run, + ['-package ghc -j1']) + test('T15858', [extra_files(['plugin-recomp/', 'plugin-recomp-test.hs']), # ===================================== testsuite/tests/plugins/static-plugins.hs ===================================== @@ -35,23 +35,23 @@ main = do -- Start with a pure plugin, this should trigger recomp. liftIO $ putStrLn "==pure.0" - loadWithPlugins [StaticPlugin $ PluginWithArgs plugin0_pure []] + loadWithPlugins [StaticPlugin (PluginWithArgs plugin0_pure []) False] -- The same (or a different) pure plugin shouldn't trigger recomp. liftIO $ putStrLn "==pure.1" - loadWithPlugins [StaticPlugin $ PluginWithArgs plugin0_pure []] + loadWithPlugins [StaticPlugin (PluginWithArgs plugin0_pure []) False] -- Next try with a fingerprint plugin, should trigger recomp. liftIO $ putStrLn "==fp0.0" - loadWithPlugins [StaticPlugin $ PluginWithArgs plugin_fp0 []] + loadWithPlugins [StaticPlugin (PluginWithArgs plugin_fp0 []) False] -- With the same fingerprint plugin, should not trigger recomp. liftIO $ putStrLn "==fp0.1" - loadWithPlugins [StaticPlugin $ PluginWithArgs plugin_fp0 []] + loadWithPlugins [StaticPlugin (PluginWithArgs plugin_fp0 []) False] -- Change the plugin fingerprint, should trigger recomp. liftIO $ putStrLn "==fp1" - loadWithPlugins [StaticPlugin $ PluginWithArgs plugin_fp1 []] + loadWithPlugins [StaticPlugin (PluginWithArgs plugin_fp1 []) False] -- TODO: this currently doesn't work, patch pending -- -- Even though the plugin is now pure we should still recomp since we View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f04fd0ae9b6283229f8891c8e7859b9f4cc3785a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f04fd0ae9b6283229f8891c8e7859b9f4cc3785a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:52:13 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 17:52:13 -0400 Subject: [Git][ghc/ghc][master] Allow unknown fd device types for setNonBlockingMode. Message-ID: <66ef400d8f2c0_23cb2797fb081207f2@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - 7 changed files: - libraries/base/changelog.md - libraries/ghc-internal/src/GHC/Internal/IO/FD.hs - libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/interface-stability/base-exports.stdout-ws-32 Changes: ===================================== libraries/base/changelog.md ===================================== @@ -16,6 +16,7 @@ * Add `firstA` and `secondA` to `Data.Bitraversable`. ([CLC proposal #172](https://github.com/haskell/core-libraries-committee/issues/172)) * Deprecate `GHC.TypeNats.Internal`, `GHC.TypeLits.Internal`, `GHC.ExecutionStack.Internal` ([CLC proposal #217](https://github.com/haskell/core-libraries-committee/issues/217)) * Define `Eq1`, `Ord1`, `Show1` and `Read1` instances for basic `Generic` representation types. ([CLC proposal #273](https://github.com/haskell/core-libraries-committee/issues/273)) + * `setNonBlockingMode` will no longer throw an exception when called on a FD associated with a unknown device type. ([CLC proposal #282](https://github.com/haskell/core-libraries-committee/issues/282)) * Add exception type metadata to default exception handler output. ([CLC proposal #231](https://github.com/haskell/core-libraries-committee/issues/231) and [CLC proposal #261](https://github.com/haskell/core-libraries-committee/issues/261)) ===================================== libraries/ghc-internal/src/GHC/Internal/IO/FD.hs ===================================== @@ -57,6 +57,7 @@ import GHC.Internal.Foreign.Storable import GHC.Internal.Foreign.C.Types import GHC.Internal.Foreign.C.Error import GHC.Internal.Foreign.Marshal.Utils +import GHC.Internal.Foreign.Marshal.Alloc (allocaBytes) import qualified GHC.Internal.System.Posix.Internals import GHC.Internal.System.Posix.Internals hiding (FD, setEcho, getEcho) @@ -466,10 +467,15 @@ setNonBlockingMode fd set = do -- O_NONBLOCK has no effect on regular files and block devices; -- utilities inspecting fdIsNonBlocking (such as readRawBufferPtr) -- should not be tricked to think otherwise. - is_nonblock <- if set then do - (fd_type, _, _) <- fdStat (fdFD fd) - pure $ fd_type /= RegularFile && fd_type /= RawDevice - else pure False + is_nonblock <- + if set + then do + allocaBytes sizeof_stat $ \ p_stat -> do + throwErrnoIfMinus1Retry_ "fileSize" $ + c_fstat (fdFD fd) p_stat + fd_type <- statGetType_maybe p_stat + pure $ fd_type /= Just RegularFile && fd_type /= Just RawDevice + else pure False setNonBlockingFD (fdFD fd) is_nonblock #if defined(mingw32_HOST_OS) return fd ===================================== libraries/ghc-internal/src/GHC/Internal/System/Posix/Internals.hs ===================================== @@ -140,17 +140,30 @@ fdStat fd = fdType :: FD -> IO IODeviceType fdType fd = do (ty,_,_) <- fdStat fd; return ty +-- | Return a known device type or throw an exception if the device +-- type is unknown. statGetType :: Ptr CStat -> IO IODeviceType statGetType p_stat = do + dev_ty_m <- statGetType_maybe p_stat + case dev_ty_m of + Nothing -> ioError ioe_unknownfiletype + Just dev_ty -> pure dev_ty + +-- | Unlike @statGetType@, @statGetType_maybe@ will not throw an exception +-- if the CStat refers to a unknown device type. +-- +-- @since base-4.20.1.0 +statGetType_maybe :: Ptr CStat -> IO (Maybe IODeviceType) +statGetType_maybe p_stat = do c_mode <- st_mode p_stat :: IO CMode case () of - _ | s_isdir c_mode -> return Directory + _ | s_isdir c_mode -> return $ Just Directory | s_isfifo c_mode || s_issock c_mode || s_ischr c_mode - -> return Stream - | s_isreg c_mode -> return RegularFile + -> return $ Just Stream + | s_isreg c_mode -> return $ Just RegularFile -- Q: map char devices to RawDevice too? - | s_isblk c_mode -> return RawDevice - | otherwise -> ioError ioe_unknownfiletype + | s_isblk c_mode -> return $ Just RawDevice + | otherwise -> return Nothing ioe_unknownfiletype :: IOException ioe_unknownfiletype = IOError Nothing UnsupportedOperation "fdType" ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -10596,6 +10596,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs ===================================== @@ -13637,6 +13637,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -10865,6 +10865,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Int.Int64 statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.CWString -> GHC.Types.IO a) -> GHC.Types.IO a ===================================== testsuite/tests/interface-stability/base-exports.stdout-ws-32 ===================================== @@ -10596,6 +10596,7 @@ module System.Posix.Internals where st_mtime :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.Foreign.C.Types.CTime st_size :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.System.Posix.Types.COff statGetType :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO GHC.Internal.IO.Device.IODeviceType + statGetType_maybe :: GHC.Internal.Ptr.Ptr CStat -> GHC.Types.IO (GHC.Internal.Maybe.Maybe GHC.Internal.IO.Device.IODeviceType) tcSetAttr :: forall a. FD -> (GHC.Internal.Ptr.Ptr CTermios -> GHC.Types.IO a) -> GHC.Types.IO a throwInternalNulError :: forall a. GHC.Internal.IO.FilePath -> GHC.Types.IO a withFilePath :: forall a. GHC.Internal.IO.FilePath -> (GHC.Internal.Foreign.C.String.Encoding.CString -> GHC.Types.IO a) -> GHC.Types.IO a View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/620becd72ec18ca08c1ed86759d65a0e614fe43c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/620becd72ec18ca08c1ed86759d65a0e614fe43c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:53:08 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 17:53:08 -0400 Subject: [Git][ghc/ghc][master] 2 commits: Use Hackage version of Cabal 3.14.0.0 for Hadrian. Message-ID: <66ef4044602f4_23cb277ebf3012816b@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 10 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - hadrian/bootstrap/bootstrap.py - hadrian/bootstrap/generate_bootstrap_plans - hadrian/bootstrap/plan-9_10_1.json - − hadrian/bootstrap/plan-9_4_1.json - − hadrian/bootstrap/plan-9_4_2.json - − hadrian/bootstrap/plan-9_4_3.json The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/620becd72ec18ca08c1ed86759d65a0e614fe43c...6c83fd7fa7fcb18c88616c6b30b1c67acc9273b7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/620becd72ec18ca08c1ed86759d65a0e614fe43c...6c83fd7fa7fcb18c88616c6b30b1c67acc9273b7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:53:23 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 17:53:23 -0400 Subject: [Git][ghc/ghc][master] ghc-toolchain: Set -fuse-ld even for ld.bfd Message-ID: <66ef405350194_23cb2766c29012831b@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - 1 changed file: - utils/ghc-toolchain/src/GHC/Toolchain/Tools/Link.hs Changes: ===================================== utils/ghc-toolchain/src/GHC/Toolchain/Tools/Link.hs ===================================== @@ -91,7 +91,7 @@ findLinkFlags enableOverride cc ccLink -- executable exists before trying cc. do _ <- findProgram (linker ++ " linker") emptyProgOpt ["ld."++linker] prog <$ checkLinkWorks cc prog - | linker <- ["lld", "gold"] + | linker <- ["lld", "gold", "bfd"] , let prog = over _prgFlags (++["-fuse-ld="++linker]) ccLink ] <|> (ccLink <$ checkLinkWorks cc ccLink) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9586998d09d6be7ca4478b344c7db04ecd512c03 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9586998d09d6be7ca4478b344c7db04ecd512c03 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:53:53 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 17:53:53 -0400 Subject: [Git][ghc/ghc][master] Parser: be more careful when lexing extended literals (#25258) Message-ID: <66ef4071e2398_23cb27671538128567@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - 13 changed files: - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Parser/Lexer.x - docs/users_guide/exts/extended_literals.rst - + testsuite/tests/parser/should_compile/T25258.hs - + testsuite/tests/parser/should_compile/T25258.stderr - testsuite/tests/parser/should_compile/all.T - + testsuite/tests/parser/should_fail/T25258a.hs - + testsuite/tests/parser/should_fail/T25258a.stderr - + testsuite/tests/parser/should_fail/T25258b.hs - + testsuite/tests/parser/should_fail/T25258b.stderr - + testsuite/tests/parser/should_fail/T25258c.hs - + testsuite/tests/parser/should_fail/T25258c.stderr - testsuite/tests/parser/should_fail/all.T Changes: ===================================== compiler/GHC/Data/StringBuffer.hs ===================================== @@ -9,6 +9,7 @@ Buffers for scanning string input stored in external arrays. {-# LANGUAGE CPP #-} {-# LANGUAGE MagicHash #-} {-# LANGUAGE UnboxedTuples #-} +{-# LANGUAGE LambdaCase #-} {-# OPTIONS_GHC -O2 #-} -- We always optimise this, otherwise performance of a non-optimised @@ -47,6 +48,7 @@ module GHC.Data.StringBuffer -- * Parsing integers parseUnsignedInteger, + findHashOffset, -- * Checking for bi-directional format characters containsBidirectionalFormatChar, @@ -417,3 +419,15 @@ parseUnsignedInteger (StringBuffer buf _ cur) len radix char_to_int '_' -> go (i + 1) x -- skip "_" (#14473) char -> go (i + 1) (x * radix + toInteger (char_to_int char)) in go 0 0 + +-- | Find the offset of the '#' character in the StringBuffer. +-- +-- Make sure that it contains one before calling this function! +findHashOffset :: StringBuffer -> Int +findHashOffset (StringBuffer buf _ cur) + = inlinePerformIO $ withForeignPtr buf $ \ptr -> do + let + go p = peek p >>= \case + (0x23 :: Word8) -> pure $! ((p `minusPtr` ptr) - cur) + _ -> go (p `plusPtr` 1) + go (ptr `plusPtr` cur) ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -562,6 +562,18 @@ $unigraphic / { isSmartQuote } { smart_quote_error } @octallit \# \# / { ifExtension MagicHashBit } { tok_primword 2 4 octal } @hexadecimallit \# \# / { ifExtension MagicHashBit } { tok_primword 2 4 hexadecimal } + @decimal \# $idchar+ / { ifExtension ExtendedLiteralsBit } { tok_prim_num_ext positive 0 decimal } + @binarylit \# $idchar+ / { ifExtension ExtendedLiteralsBit `alexAndPred` + ifExtension BinaryLiteralsBit } { tok_prim_num_ext positive 2 binary } + @octallit \# $idchar+ / { ifExtension ExtendedLiteralsBit } { tok_prim_num_ext positive 2 octal } + @hexadecimallit \# $idchar+ / { ifExtension ExtendedLiteralsBit } { tok_prim_num_ext positive 2 hexadecimal } + @negative @decimal \# $idchar+ / { negHashLitPred ExtendedLiteralsBit } { tok_prim_num_ext negative 1 decimal } + @negative @binarylit \# $idchar+ / { negHashLitPred ExtendedLiteralsBit `alexAndPred` + ifExtension BinaryLiteralsBit } { tok_prim_num_ext negative 3 binary } + @negative @octallit \# $idchar+ / { negHashLitPred ExtendedLiteralsBit } { tok_prim_num_ext negative 3 octal } + @negative @hexadecimallit \# $idchar+ / { negHashLitPred ExtendedLiteralsBit } { tok_prim_num_ext negative 3 hexadecimal } + + -- Unboxed floats and doubles (:: Float#, :: Double#) -- prim_{float,double} work with signed literals @floating_point \# / { ifExtension MagicHashBit } { tok_frac 1 tok_primfloat } @@ -573,91 +585,6 @@ $unigraphic / { isSmartQuote } { smart_quote_error } @negative 0[xX] @numspc @hex_floating_point \# / { ifExtension HexFloatLiteralsBit `alexAndPred` negHashLitPred MagicHashBit } { tok_frac 1 tok_prim_hex_float } @negative 0[xX] @numspc @hex_floating_point \# \# / { ifExtension HexFloatLiteralsBit `alexAndPred` negHashLitPred MagicHashBit } { tok_frac 2 tok_prim_hex_double } - @decimal \#"Int8" / { ifExtension ExtendedLiteralsBit } { tok_primint8 positive 0 decimal } - @binarylit \#"Int8" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint8 positive 2 binary } - @octallit \#"Int8" / { ifExtension ExtendedLiteralsBit } { tok_primint8 positive 2 octal } - @hexadecimallit \#"Int8" / { ifExtension ExtendedLiteralsBit } { tok_primint8 positive 2 hexadecimal } - @negative @decimal \#"Int8" / { negHashLitPred ExtendedLiteralsBit } { tok_primint8 negative 1 decimal } - @negative @binarylit \#"Int8" / { negHashLitPred ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint8 negative 3 binary } - @negative @octallit \#"Int8" / { negHashLitPred ExtendedLiteralsBit } { tok_primint8 negative 3 octal } - @negative @hexadecimallit \#"Int8" / { negHashLitPred ExtendedLiteralsBit } { tok_primint8 negative 3 hexadecimal } - - @decimal \#"Int16" / { ifExtension ExtendedLiteralsBit } { tok_primint16 positive 0 decimal } - @binarylit \#"Int16" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint16 positive 2 binary } - @octallit \#"Int16" / { ifExtension ExtendedLiteralsBit } { tok_primint16 positive 2 octal } - @hexadecimallit \#"Int16" / { ifExtension ExtendedLiteralsBit } { tok_primint16 positive 2 hexadecimal } - @negative @decimal \#"Int16" / { negHashLitPred ExtendedLiteralsBit} { tok_primint16 negative 1 decimal } - @negative @binarylit \#"Int16" / { negHashLitPred ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint16 negative 3 binary } - @negative @octallit \#"Int16" / { negHashLitPred ExtendedLiteralsBit} { tok_primint16 negative 3 octal } - @negative @hexadecimallit \#"Int16" / { negHashLitPred ExtendedLiteralsBit} { tok_primint16 negative 3 hexadecimal } - - @decimal \#"Int32" / { ifExtension ExtendedLiteralsBit } { tok_primint32 positive 0 decimal } - @binarylit \#"Int32" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint32 positive 2 binary } - @octallit \#"Int32" / { ifExtension ExtendedLiteralsBit } { tok_primint32 positive 2 octal } - @hexadecimallit \#"Int32" / { ifExtension ExtendedLiteralsBit } { tok_primint32 positive 2 hexadecimal } - @negative @decimal \#"Int32" / { negHashLitPred ExtendedLiteralsBit } { tok_primint32 negative 1 decimal } - @negative @binarylit \#"Int32" / { negHashLitPred ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint32 negative 3 binary } - @negative @octallit \#"Int32" / { negHashLitPred ExtendedLiteralsBit} { tok_primint32 negative 3 octal } - @negative @hexadecimallit \#"Int32" / { negHashLitPred ExtendedLiteralsBit} { tok_primint32 negative 3 hexadecimal } - - @decimal \#"Int64" / { ifExtension ExtendedLiteralsBit } { tok_primint64 positive 0 decimal } - @binarylit \#"Int64" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint64 positive 2 binary } - @octallit \#"Int64" / { ifExtension ExtendedLiteralsBit } { tok_primint64 positive 2 octal } - @hexadecimallit \#"Int64" / { ifExtension ExtendedLiteralsBit } { tok_primint64 positive 2 hexadecimal } - @negative @decimal \#"Int64" / { negHashLitPred ExtendedLiteralsBit } { tok_primint64 negative 1 decimal } - @negative @binarylit \#"Int64" / { negHashLitPred ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint64 negative 3 binary } - @negative @octallit \#"Int64" / { negHashLitPred ExtendedLiteralsBit } { tok_primint64 negative 3 octal } - @negative @hexadecimallit \#"Int64" / { negHashLitPred ExtendedLiteralsBit } { tok_primint64 negative 3 hexadecimal } - - @decimal \#"Int" / { ifExtension ExtendedLiteralsBit } { tok_primint positive 0 4 decimal } - @binarylit \#"Int" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint positive 2 6 binary } - @octallit \#"Int" / { ifExtension ExtendedLiteralsBit } { tok_primint positive 2 6 octal } - @hexadecimallit \#"Int" / { ifExtension ExtendedLiteralsBit } { tok_primint positive 2 6 hexadecimal } - @negative @decimal \#"Int" / { negHashLitPred ExtendedLiteralsBit } { tok_primint negative 1 5 decimal } - @negative @binarylit \#"Int" / { negHashLitPred ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primint negative 3 7 binary } - @negative @octallit \#"Int" / { negHashLitPred ExtendedLiteralsBit } { tok_primint negative 3 7 octal } - @negative @hexadecimallit \#"Int" / { negHashLitPred ExtendedLiteralsBit } { tok_primint negative 3 7 hexadecimal } - - @decimal \#"Word8" / { ifExtension ExtendedLiteralsBit } { tok_primword8 0 decimal } - @binarylit \#"Word8" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primword8 2 binary } - @octallit \#"Word8" / { ifExtension ExtendedLiteralsBit } { tok_primword8 2 octal } - @hexadecimallit \#"Word8" / { ifExtension ExtendedLiteralsBit } { tok_primword8 2 hexadecimal } - - @decimal \#"Word16" / { ifExtension ExtendedLiteralsBit } { tok_primword16 0 decimal } - @binarylit \#"Word16" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primword16 2 binary } - @octallit \#"Word16" / { ifExtension ExtendedLiteralsBit } { tok_primword16 2 octal } - @hexadecimallit \#"Word16" / { ifExtension ExtendedLiteralsBit } { tok_primword16 2 hexadecimal } - - @decimal \#"Word32" / { ifExtension ExtendedLiteralsBit } { tok_primword32 0 decimal } - @binarylit \#"Word32" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primword32 2 binary } - @octallit \#"Word32" / { ifExtension ExtendedLiteralsBit } { tok_primword32 2 octal } - @hexadecimallit \#"Word32" / { ifExtension ExtendedLiteralsBit } { tok_primword32 2 hexadecimal } - - @decimal \#"Word64" / { ifExtension ExtendedLiteralsBit } { tok_primword64 0 decimal } - @binarylit \#"Word64" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primword64 2 binary } - @octallit \#"Word64" / { ifExtension ExtendedLiteralsBit } { tok_primword64 2 octal } - @hexadecimallit \#"Word64" / { ifExtension ExtendedLiteralsBit } { tok_primword64 2 hexadecimal } - - @decimal \#"Word" / { ifExtension ExtendedLiteralsBit } { tok_primword 0 5 decimal } - @binarylit \#"Word" / { ifExtension ExtendedLiteralsBit `alexAndPred` - ifExtension BinaryLiteralsBit } { tok_primword 2 7 binary } - @octallit \#"Word" / { ifExtension ExtendedLiteralsBit } { tok_primword 2 7 octal } - @hexadecimallit \#"Word" / { ifExtension ExtendedLiteralsBit } { tok_primword 2 7 hexadecimal } - } -- Strings and chars are lexed by hand-written code. The reason is @@ -1905,21 +1832,56 @@ sym con span buf len _buf2 = !fs = lexemeToFastString buf len -- Variations on the integral numeric literal. -tok_integral :: (SourceText -> Integer -> Token) - -> (Integer -> Integer) - -> Int -> Int - -> (Integer, (Char -> Int)) - -> Action -tok_integral itint transint transbuf translen (radix,char_to_int) span buf len _buf2 = do +tok_integral + :: (SourceText -> Integer -> Token) -- ^ token constructor + -> (Integer -> Integer) -- ^ value transformation (e.g. negate) + -> Int -- ^ Offset of the unsigned value (e.g. 1 when we parsed "-", 2 for "0x", etc.) + -> Int -- ^ Number of non-numeric characters parsed (e.g. 6 in "-12#Int8") + -> (Integer, (Char -> Int)) -- ^ (radix, char_to_int parsing function) + -> Action +tok_integral mk_token transval offset translen (radix,char_to_int) span buf len _buf2 = do numericUnderscores <- getBit NumericUnderscoresBit -- #14473 let src = lexemeToFastString buf len when ((not numericUnderscores) && ('_' `elem` unpackFS src)) $ do pState <- getPState let msg = PsErrNumUnderscores NumUnderscore_Integral addError $ mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg - return $ L span $ itint (SourceText src) - $! transint $ parseUnsignedInteger - (offsetBytes transbuf buf) (subtract translen len) radix char_to_int + return $ L span $ mk_token (SourceText src) + $! transval $ parseUnsignedInteger + (offsetBytes offset buf) (subtract translen len) radix char_to_int + +-- | Helper to parse ExtendedLiterals (e.g. -0x10#Word32) +-- +-- This function finds the offset of the "#" character and checks that the +-- suffix is valid. Then it calls tok_integral with the appropriate suffix +-- length taken into account. +tok_prim_num_ext + :: (Integer -> Integer) -- ^ value transformation (e.g. negate) + -> Int -- ^ Offset of the unsigned value (e.g. 1 when we parsed "-", 2 for "0x", etc.) + -> (Integer, (Char -> Int)) -- ^ (radix, char_to_int parsing function) + -> Action +tok_prim_num_ext transval offset (radix,char_to_int) span buf len buf2 = do + let !suffix_offset = findHashOffset buf + 1 + let !suffix_len = len - suffix_offset + let !suffix = lexemeToFastString (offsetBytes suffix_offset buf) suffix_len + + mk_token <- if + | suffix == fsLit "Word" -> pure ITprimword + | suffix == fsLit "Word8" -> pure ITprimword8 + | suffix == fsLit "Word16" -> pure ITprimword16 + | suffix == fsLit "Word32" -> pure ITprimword32 + | suffix == fsLit "Word64" -> pure ITprimword64 + | suffix == fsLit "Int" -> pure ITprimint + | suffix == fsLit "Int8" -> pure ITprimint8 + | suffix == fsLit "Int16" -> pure ITprimint16 + | suffix == fsLit "Int32" -> pure ITprimint32 + | suffix == fsLit "Int64" -> pure ITprimint64 + | otherwise -> srcParseFail + + let !translen = suffix_len+offset+1 + tok_integral mk_token transval offset translen (radix,char_to_int) span buf len buf2 + + tok_num :: (Integer -> Integer) -> Int -> Int @@ -1941,48 +1903,16 @@ tok_primint = tok_integral ITprimint tok_primword :: Int -> Int -> (Integer, (Char->Int)) -> Action tok_primword = tok_integral ITprimword positive + positive, negative :: (Integer -> Integer) positive = id negative = negate -decimal, octal, hexadecimal :: (Integer, Char -> Int) -decimal = (10,octDecDigit) -binary = (2,octDecDigit) -octal = (8,octDecDigit) -hexadecimal = (16,hexDigit) --- | Helper for defining @IntX@ primitive literal parsers (specifically for --- the ExtendedLiterals extension, such as @123#Int8@). -tok_primintX :: (SourceText -> Integer -> Token) - -> Int - -> (Integer -> Integer) - -> Int - -> (Integer, (Char->Int)) -> Action -tok_primintX itint addlen transint transbuf = - tok_integral itint transint transbuf (transbuf+addlen) - -tok_primint8, tok_primint16, tok_primint32, tok_primint64 - :: (Integer -> Integer) - -> Int -> (Integer, (Char->Int)) -> Action -tok_primint8 = tok_primintX ITprimint8 5 -tok_primint16 = tok_primintX ITprimint16 6 -tok_primint32 = tok_primintX ITprimint32 6 -tok_primint64 = tok_primintX ITprimint64 6 - --- | Helper for defining @WordX@ primitive literal parsers (specifically for --- the ExtendedLiterals extension, such as @234#Word8@). -tok_primwordX :: (SourceText -> Integer -> Token) - -> Int - -> Int - -> (Integer, (Char->Int)) -> Action -tok_primwordX itint addlen transbuf = - tok_integral itint positive transbuf (transbuf+addlen) - -tok_primword8, tok_primword16, tok_primword32, tok_primword64 - :: Int -> (Integer, (Char->Int)) -> Action -tok_primword8 = tok_primwordX ITprimword8 6 -tok_primword16 = tok_primwordX ITprimword16 7 -tok_primword32 = tok_primwordX ITprimword32 7 -tok_primword64 = tok_primwordX ITprimword64 7 +binary, octal, decimal, hexadecimal :: (Integer, Char -> Int) +binary = (2,octDecDigit) +octal = (8,octDecDigit) +decimal = (10,octDecDigit) +hexadecimal = (16,hexDigit) -- readSignificandExponentPair can understand negative rationals, exponents, everything. tok_frac :: Int -> (String -> Token) -> Action ===================================== docs/users_guide/exts/extended_literals.rst ===================================== @@ -35,11 +35,10 @@ The primitive numeric types allowed are: - ``Word64#`` - ``Word#`` -All types permit any nonnegative Haskell integer lexeme, e.g. ``70``, ``0x2A``, -``0o1276``, ``0b1010`` (with :extension:`BinaryLiterals`). The signed ``Int`` -types also permit negative integer lexemes. Defining a literal with a value that -can't fit in its requested type will emit an overflow warning by default, the -same as boxed numeric literals. +All types permit any positive and negative Haskell integer lexeme. Defining a +literal with a value that can't fit in its requested type will emit an overflow +warning by default, the same as boxed numeric literals (see +:ghc-flag:`-Woverflowed-literals`). As with :extension:`MagicHash`, this extension does not bring anything into scope, nor change any semantics. The syntax only applies to numeric literals. ===================================== testsuite/tests/parser/should_compile/T25258.hs ===================================== @@ -0,0 +1,18 @@ +{-# LANGUAGE MagicHash, ExtendedLiterals, RequiredTypeArguments #-} + +module T25258 where + +import GHC.Exts +import GHC.Word + +f :: Int# -> forall t -> () +f _ _ = () + +x4 :: forall t -> () +x4 = f 0x08#Int + +x5 :: () +x5 = f 0x08# Int + +x6 :: Word8 +x6 = W8# (-10#Word8) -- we now allow negative unsigned extended literals (with a warning) ===================================== testsuite/tests/parser/should_compile/T25258.stderr ===================================== @@ -0,0 +1,3 @@ +T25258.hs:18:11: warning: [GHC-97441] [-Woverflowed-literals (in -Wdefault)] + Literal -10 is out of the Word8# range 0..255 + ===================================== testsuite/tests/parser/should_compile/all.T ===================================== @@ -204,3 +204,4 @@ test('ListTuplePunsFamilies', [expect_broken(23135), extra_files(['ListTuplePuns test('T22155', normal, compile, ['-dsuppress-uniques -ddump-simpl -dsuppress-all -dno-typeable-binds']) test('T25132', normal, compile, ['']) +test('T25258', normal, compile, ['']) ===================================== testsuite/tests/parser/should_fail/T25258a.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE MagicHash, ExtendedLiterals #-} + +module T25258a where + +import GHC.Exts + +x1 = 0x08#Misc -- invalid uppercase identifier ===================================== testsuite/tests/parser/should_fail/T25258a.stderr ===================================== @@ -0,0 +1,2 @@ +T25258a.hs:7:6: error: [GHC-58481] parse error on input ‘0x08#Misc’ + ===================================== testsuite/tests/parser/should_fail/T25258b.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE MagicHash, ExtendedLiterals #-} + +module T25258b where + +import GHC.Exts + +x2 = 0x80#Int3 ===================================== testsuite/tests/parser/should_fail/T25258b.stderr ===================================== @@ -0,0 +1,2 @@ +T25258b.hs:7:6: error: [GHC-58481] parse error on input ‘0x80#Int3’ + ===================================== testsuite/tests/parser/should_fail/T25258c.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE MagicHash, ExtendedLiterals #-} + +module T25258c where + +import GHC.Exts + +x0 = 0x08#misc -- invalid lowercase identifier ===================================== testsuite/tests/parser/should_fail/T25258c.stderr ===================================== @@ -0,0 +1,2 @@ +T25258c.hs:7:6: error: [GHC-58481] parse error on input ‘0x08#misc’ + ===================================== testsuite/tests/parser/should_fail/all.T ===================================== @@ -234,3 +234,6 @@ test('OrPatInExprErr', normal, compile_fail, ['']) test('MultilineStringsError', [normalise_whitespace_fun(lambda s: s)], compile_fail, ['']) test('MultilineStringsSmartQuotes', normal, compile_fail, ['']) test('MultilineStringsInnerTab', normal, compile_fail, ['']) +test('T25258a', normal, compile_fail, ['']) +test('T25258b', normal, compile_fail, ['']) +test('T25258c', normal, compile_fail, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d7016e0d41769f777c713158817bc0c43fb2a33f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d7016e0d41769f777c713158817bc0c43fb2a33f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 21:55:23 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Sat, 21 Sep 2024 17:55:23 -0400 Subject: [Git][ghc/ghc][wip/T24978] 43 commits: JS: fake support for native adjustors (#25159) Message-ID: <66ef40cb11843_23cb2715d57b8128745@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24978 at Glasgow Haskell Compiler / GHC Commits: 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - c4c9b21b by Simon Peyton Jones at 2024-09-21T22:53:52+01:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to address As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 10 changed files: - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e4bdc821f1e4863074a73cc032082dd2251a80f8...c4c9b21b26ddaa832a71d168ba8e07326bbc0989 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e4bdc821f1e4863074a73cc032082dd2251a80f8...c4c9b21b26ddaa832a71d168ba8e07326bbc0989 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 22:04:39 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Sat, 21 Sep 2024 18:04:39 -0400 Subject: [Git][ghc/ghc][wip/T24978] Add Given injectivity for built-in type families Message-ID: <66ef42f76096f_23cb2711278381295b@gitlab.mail> Simon Peyton Jones pushed to branch wip/T24978 at Glasgow Haskell Compiler / GHC Commits: f7054f5c by Simon Peyton Jones at 2024-09-21T23:03:54+01:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 10 changed files: - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f7054f5ceff24877a89982e5131cdaa9f510e3e5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f7054f5ceff24877a89982e5131cdaa9f510e3e5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 22:54:12 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 21 Sep 2024 18:54:12 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 12 commits: Demand: Combine examples into Note (#25107) Message-ID: <66ef4e943b066_1137ec2021c864663@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - d8071598 by Zubin Duggal at 2024-09-21T18:53:55-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - a938f57c by Alan Zimmerman at 2024-09-21T18:53:55-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 26 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - docs/users_guide/exts/extended_literals.rst - hadrian/bootstrap/bootstrap.py - hadrian/bootstrap/generate_bootstrap_plans - hadrian/bootstrap/plan-9_10_1.json - − hadrian/bootstrap/plan-9_4_1.json - − hadrian/bootstrap/plan-9_4_2.json The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e7fedcf4d7356dcfad2af2df1cbad68e1f6aa397...a938f57cf11a116b5fe7847e75c909ee9a16e0f3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e7fedcf4d7356dcfad2af2df1cbad68e1f6aa397...a938f57cf11a116b5fe7847e75c909ee9a16e0f3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 21 23:23:34 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Sat, 21 Sep 2024 19:23:34 -0400 Subject: [Git][ghc/ghc][wip/andreask/ghc-primops] 11 commits: Demand: Combine examples into Note (#25107) Message-ID: <66ef55764b28b_1137ec3f578c68395@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/ghc-primops at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - 3851cd4c by Andreas Klebinger at 2024-09-22T01:05:06+02:00 ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps This will be the new place for functions that would have gone into GHC.Exts in the past but are not stable enough to do so now. Addresses #25242 - - - - - 25 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - docs/users_guide/exts/extended_literals.rst - hadrian/bootstrap/bootstrap.py - hadrian/bootstrap/generate_bootstrap_plans - hadrian/bootstrap/plan-9_10_1.json - − hadrian/bootstrap/plan-9_4_1.json - − hadrian/bootstrap/plan-9_4_2.json The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/79b07e5cc766f5445366c12aba2c4fb236c15181...3851cd4c65c803440911015b80b7319f214e5f5b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/79b07e5cc766f5445366c12aba2c4fb236c15181...3851cd4c65c803440911015b80b7319f214e5f5b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 00:59:56 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Sat, 21 Sep 2024 20:59:56 -0400 Subject: [Git][ghc/ghc][wip/25232] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66ef6c0c4eaf2_1d20eb2c0a10961b5@gitlab.mail> Zubin pushed to branch wip/25232 at Glasgow Haskell Compiler / GHC Commits: e6d3e5dc by Zubin Duggal at 2024-09-22T06:29:46+05:30 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 5 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - + testsuite/tests/rts/T25232.hs - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,18 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + // Not doing this lead to #25232 and #23958 + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +958,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): ===================================== testsuite/tests/rts/T25232.hs ===================================== @@ -0,0 +1,14 @@ +module Main where + +import Control.Monad +import Control.Monad.Trans.State.Strict +import Data.IORef + +main :: IO () +main = do + ref <- newIORef () + replicateM_ 1000 $ withIORef ref $ runStateT (pure ()) + +withIORef :: IORef a -> (a -> IO (b, a)) -> IO b +withIORef ref f = + readIORef ref >>= f >>= \(b, a) -> writeIORef ref a >> pure b ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [only_ways(['normal','nonmoving','nonmoving_prof','nonmoving_thr_prof']), extra_ways(['nonmoving', 'nonmoving_prof', 'nonmoving_thr_prof'])], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e6d3e5dc2201e7adb922d1b5a20acea6ef5385e7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e6d3e5dc2201e7adb922d1b5a20acea6ef5385e7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 02:29:22 2024 From: gitlab at gitlab.haskell.org (Ryan Scott (@RyanGlScott)) Date: Sat, 21 Sep 2024 22:29:22 -0400 Subject: [Git][ghc/ghc][wip/T25148] 111 commits: haddock: Build haddock-api and haddock-library using hadrian Message-ID: <66ef8102b1aa7_1d20eb6f85d8108746@gitlab.mail> Ryan Scott pushed to branch wip/T25148 at Glasgow Haskell Compiler / GHC Commits: 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - 153f1cfe by Ryan Scott at 2024-09-21T19:55:25-04:00 Resolve ambiguous method-bound type variables in vanilla defaults When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from method type signatures)`. Fixes #14266. Fixes #25148. - - - - - 14ed5de9 by Ryan Scott at 2024-09-21T22:28:34-04:00 Also apply the same trick to GND, review suggestions - - - - - 24 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2542cb36cc8eff109a9d1a99215db724e65c2669...14ed5de97fcc5f1a761377e38176a1c3ca0c01e8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2542cb36cc8eff109a9d1a99215db724e65c2669...14ed5de97fcc5f1a761377e38176a1c3ca0c01e8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 03:22:25 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Sat, 21 Sep 2024 23:22:25 -0400 Subject: [Git][ghc/ghc][wip/strings] Replace manual string lexing (#25158) Message-ID: <66ef8d71ed700_1d20eb97e500114155@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC Commits: e7c83fe4 by Brandon Chinn at 2024-09-21T20:22:05-07:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 29 changed files: - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - testsuite/driver/testlib.py - testsuite/tests/ghci/prog013/prog013.stderr - testsuite/tests/ghci/scripts/ghci022.stderr - testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr - testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr - testsuite/tests/parser/should_fail/T21843c.stderr - testsuite/tests/parser/should_fail/T21843e.stderr - testsuite/tests/parser/should_fail/T21843f.stderr - testsuite/tests/parser/should_fail/T3751.stderr - testsuite/tests/parser/should_fail/T5425.stderr - testsuite/tests/parser/should_fail/readFail002.stderr - testsuite/tests/parser/should_fail/readFail004.stderr - testsuite/tests/parser/should_fail/readFail005.stderr - testsuite/tests/parser/should_fail/readFail033.stderr - testsuite/tests/parser/unicode/all.T - + testsuite/tests/parser/unicode/lex_unicode_ids.hs - + testsuite/tests/parser/unicode/lex_unispace.hs - testsuite/tests/parser/unicode/utf8_010.stderr - testsuite/tests/parser/unicode/utf8_011.stderr - testsuite/tests/parser/unicode/utf8_020.stderr - testsuite/tests/parser/unicode/utf8_021.stderr - testsuite/tests/parser/unicode/utf8_022.stderr - utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs Changes: ===================================== compiler/GHC/Parser.y ===================================== @@ -728,7 +728,7 @@ are the most common patterns, rewritten as regular expressions for clarity: CHAR { L _ (ITchar _ _) } STRING { L _ (ITstring _ _) } - MULTILINESTRING { L _ (ITmultilinestring _ _) } + STRING_MULTI { L _ (ITstringMulti _ _) } INTEGER { L _ (ITinteger _) } RATIONAL { L _ (ITrational _) } @@ -2357,8 +2357,8 @@ atype :: { LHsType GhcPs } (getCHAR $1) } | STRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGs $1) (getSTRING $1) } - | MULTILINESTRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getMULTILINESTRINGs $1) - (getMULTILINESTRING $1) } + | STRING_MULTI { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGMULTIs $1) + (getSTRINGMULTI $1) } | '_' { sL1a $1 $ mkAnonWildCardTy } -- Type variables are never exported, so `M.tyvar` will be rejected by the renamer. -- We let it pass the parser because the renamer can generate a better error message. @@ -4047,8 +4047,8 @@ literal :: { Located (HsLit GhcPs) } : CHAR { sL1 $1 $ HsChar (getCHARs $1) $ getCHAR $1 } | STRING { sL1 $1 $ HsString (getSTRINGs $1) $ getSTRING $1 } - | MULTILINESTRING { sL1 $1 $ HsMultilineString (getMULTILINESTRINGs $1) - $ getMULTILINESTRING $1 } + | STRING_MULTI { sL1 $1 $ HsMultilineString (getSTRINGMULTIs $1) + $ getSTRINGMULTI $1 } | PRIMINTEGER { sL1 $1 $ HsIntPrim (getPRIMINTEGERs $1) $ getPRIMINTEGER $1 } | PRIMWORD { sL1 $1 $ HsWordPrim (getPRIMWORDs $1) @@ -4154,7 +4154,7 @@ getIPDUPVARID (L _ (ITdupipvarid x)) = x getLABELVARID (L _ (ITlabelvarid _ x)) = x getCHAR (L _ (ITchar _ x)) = x getSTRING (L _ (ITstring _ x)) = x -getMULTILINESTRING (L _ (ITmultilinestring _ x)) = x +getSTRINGMULTI (L _ (ITstringMulti _ x)) = x getINTEGER (L _ (ITinteger x)) = x getRATIONAL (L _ (ITrational x)) = x getPRIMCHAR (L _ (ITprimchar _ x)) = x @@ -4180,7 +4180,7 @@ getVOCURLY (L (RealSrcSpan l _) ITvocurly) = srcSpanStartCol l getINTEGERs (L _ (ITinteger (IL src _ _))) = src getCHARs (L _ (ITchar src _)) = src getSTRINGs (L _ (ITstring src _)) = src -getMULTILINESTRINGs (L _ (ITmultilinestring src _)) = src +getSTRINGMULTIs (L _ (ITstringMulti src _)) = src getPRIMCHARs (L _ (ITprimchar src _)) = src getPRIMSTRINGs (L _ (ITprimstring src _)) = src getPRIMINTEGERs (L _ (ITprimint src _)) = src ===================================== compiler/GHC/Parser/CharClass.hs ===================================== @@ -36,14 +36,14 @@ cDigit = 64 {-# INLINABLE is_ctype #-} is_ctype :: Word8 -> Char -> Bool -is_ctype mask c = (charType c .&. mask) /= 0 +is_ctype mask c = c <= '\127' && (charType c .&. mask) /= 0 is_ident, is_symbol, is_any, is_space, is_lower, is_upper, is_digit, is_alphanum :: Char -> Bool is_ident = is_ctype cIdent is_symbol = is_ctype cSymbol is_any = is_ctype cAny -is_space = \c -> c <= '\x7f' && is_ctype cSpace c -- is_space only works for <= '\x7f' (#3751, #5425) +is_space = is_ctype cSpace is_lower = is_ctype cLower is_upper = is_ctype cUpper is_digit = is_ctype cDigit ===================================== compiler/GHC/Parser/Errors/Ppr.hs ===================================== @@ -157,8 +157,6 @@ instance Diagnostic PsMessage where LexUnknownPragma -> text "unknown pragma" LexErrorInPragma -> text "lexical error in pragma" LexNumEscapeRange -> text "numeric escape sequence out of range" - LexStringCharLit -> text "lexical error in string/character literal" - LexStringCharLitEOF -> text "unexpected end-of-file in string/character literal" LexUnterminatedComment -> text "unterminated `{-'" LexUnterminatedOptions -> text "unterminated OPTIONS pragma" LexUnterminatedQQ -> text "unterminated quasiquotation" ===================================== compiler/GHC/Parser/Errors/Types.hs ===================================== @@ -591,8 +591,6 @@ data LexErr | LexUnknownPragma -- ^ Unknown pragma | LexErrorInPragma -- ^ Lexical error in pragma | LexNumEscapeRange -- ^ Numeric escape sequence out of range - | LexStringCharLit -- ^ Lexical error in string/character literal - | LexStringCharLitEOF -- ^ Unexpected end-of-file in string/character literal | LexUnterminatedComment -- ^ Unterminated `{-' | LexUnterminatedOptions -- ^ Unterminated OPTIONS pragma | LexUnterminatedQQ -- ^ Unterminated quasiquotation ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -140,7 +140,8 @@ import GHC.Parser.String -- Any changes here should likely be reflected there. $unispace = \x05 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $nl = [\n\r\f] -$whitechar = [$nl\v\ $unispace] +$space = [\ $unispace] +$whitechar = [$nl \v $space] $white_no_nl = $whitechar # \n -- TODO #8424 $tab = \t @@ -167,6 +168,7 @@ $idchar = [$small $large $digit $uniidchar \'] $unigraphic = \x06 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $graphic = [$small $large $symbol $digit $idchar $special $unigraphic \"\'] +$charesc = [a b f n r t v \\ \" \' \&] $binit = 0-1 $octit = 0-7 @@ -213,6 +215,20 @@ $docsym = [\| \^ \* \$] @floating_point = @numspc @decimal \. @decimal @exponent? | @numspc @decimal @exponent @hex_floating_point = @numspc @hexadecimal \. @hexadecimal @bin_exponent? | @numspc @hexadecimal @bin_exponent + at gap = \\ $whitechar+ \\ + at cntrl = $asclarge | \@ | \[ | \\ | \] | \^ | \_ + at ascii = \^ @cntrl | "NUL" | "SOH" | "STX" | "ETX" | "EOT" | "ENQ" | "ACK" + | "BEL" | "BS" | "HT" | "LF" | "VT" | "FF" | "CR" | "SO" | "SI" | "DLE" + | "DC1" | "DC2" | "DC3" | "DC4" | "NAK" | "SYN" | "ETB" | "CAN" + | "EM" | "SUB" | "ESC" | "FS" | "GS" | "RS" | "US" | "SP" | "DEL" +-- N.B. ideally, we would do `@escape # \\ \&` instead of duplicating in @escapechar, +-- which is what the Haskell Report says, but this isn't valid Alex syntax, as only +-- character sets can be subtracted, not strings + at escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) + at escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) + at stringchar = ($graphic # [\\ \"]) | $space | @escape | @gap + at char = ($graphic # [\\ \']) | $space | @escapechar + -- normal signed numerical literals can only be explicitly negative, -- not explicitly positive (contrast @exponent) @negative = \- @@ -460,7 +476,7 @@ $unigraphic / { isSmartQuote } { smart_quote_error } <0> { "#" $idchar+ / { ifExtension OverloadedLabelsBit } { skip_one_varid_src ITlabelvarid } - "#" \" / { ifExtension OverloadedLabelsBit } { lex_quoted_label } + "#" \" @stringchar* \" / { ifExtension OverloadedLabelsBit } { tok_quoted_label } } <0> { @@ -660,14 +676,38 @@ $unigraphic / { isSmartQuote } { smart_quote_error } } --- Strings and chars are lexed by hand-written code. The reason is --- that even if we recognise the string or char here in the regex --- lexer, we would still have to parse the string afterward in order --- to convert it to a String. <0> { - \' { lex_char_tok } - \"\"\" / { ifExtension MultilineStringsBit} { lex_string_tok StringTypeMulti } - \" { lex_string_tok StringTypeSingle } + \"\"\" / { ifExtension MultilineStringsBit } { tok_string_multi } + \" @stringchar* \" { tok_string } + \" @stringchar* \" \# / { ifExtension MagicHashBit } { tok_string } + \' @char \' { tok_char } + \' @char \' \# / { ifExtension MagicHashBit } { tok_char } + + -- Check for smart quotes and throw better errors than a plain lexical error (#21843) + \' \\ $unigraphic / { isSmartQuote } { smart_quote_error } + \" @stringchar* \\ $unigraphic / { isSmartQuote } { smart_quote_error } + -- See Note [Bare smart quote error] + -- The valid string rule will take precedence because it'll match more + -- characters than this rule, so this rule will only fire if the string + -- could not be lexed correctly + \" @stringchar* $unigraphic / { isSmartQuote } { smart_quote_error } +} + + { + -- Parse as much of the multiline string as possible, except for quotes + @stringchar* ($nl ([\ $tab] | @gap)* @stringchar*)* { tok_string_multi_content } + -- Allow bare quotes if it's not a triple quote + (\" | \"\") / ([\n .] # \") { tok_string_multi_content } +} + +<0> { + \'\' { token ITtyQuote } + + -- The normal character match takes precedence over this because it matches + -- more characters. However, if that pattern didn't match, then this quote + -- could be a quoted identifier, like 'x. Here, just return ITsimpleQuote, + -- as the parser will lex the varid separately. + \' / ($graphic # \\ | " ") { token ITsimpleQuote } } -- Note [Whitespace-sensitive operator parsing] @@ -953,7 +993,7 @@ data Token | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" | ITstring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" - | ITmultilinestring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringMulti SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" | ITinteger IntegralLit -- Note [Literal source text] in "GHC.Types.SourceText" | ITrational FractionalLit @@ -2181,156 +2221,128 @@ lex_string_prag_comment mkTok span _buf _len _buf2 -- ----------------------------------------------------------------------------- -- Strings & Chars --- This stuff is horrible. I hates it. - -lex_string_tok :: LexStringType -> Action -lex_string_tok strType span buf _len _buf2 = do - s <- lex_string strType - - i <- getInput - case strType of - StringTypeSingle -> - lex_magic_hash i >>= \case - Just i' -> do - when (any (> '\xFF') s) $ do - pState <- getPState - let msg = PsErrPrimStringInvalidChar - let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg - addError err - - setInput i' - let (psSpan, src) = getStringLoc (buf, locStart) i' - pure $ L psSpan (ITprimstring src (unsafeMkByteString s)) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITstring src (mkFastString s)) - StringTypeMulti -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITmultilinestring src (mkFastString s)) +tok_string :: Action +tok_string span buf len _buf2 = do + s <- lex_chars ("\"", "\"") span buf (if endsInHash then len - 1 else len) + + if endsInHash + then do + when (any (> '\xFF') s) $ do + pState <- getPState + let msg = PsErrPrimStringInvalidChar + let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg + addError err + pure $ L span (ITprimstring src (unsafeMkByteString s)) + else + pure $ L span (ITstring src (mkFastString s)) where - locStart = psSpanStart span + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + +-- | Ideally, we would define this completely with Alex syntax, like normal strings. +-- Instead, this is defined as a hybrid solution by manually invoking lex states, which +-- we're doing for two reasons: +-- 1. The multiline string should all be one lexical token, not multiple +-- 2. We need to allow bare quotes, which can't be done with one regex +tok_string_multi :: Action +tok_string_multi startSpan startBuf _len _buf2 = do + -- advance to the end of the multiline string + let startLoc = psSpanStart startSpan + let i@(AI _ contentStartBuf) = + case lexDelim $ AI startLoc startBuf of + Just i -> i + Nothing -> panic "tok_string_multi did not start with a delimiter" + (AI _ contentEndBuf, i'@(AI endLoc endBuf)) <- goContent i + + -- build the values pertaining to the entire multiline string, including delimiters + let span = mkPsSpan startLoc endLoc + let len = byteDiff startBuf endBuf + let src = SourceText $ lexemeToFastString startBuf len + + -- load the content of the multiline string + let contentLen = byteDiff contentStartBuf contentEndBuf + s <- + either (throwStringLexError (AI startLoc startBuf)) pure $ + lexMultilineString contentLen contentStartBuf + + setInput i' + pure $ L span $ ITstringMulti src (mkFastString s) + where + goContent i0 = + case alexScan i0 string_multi_content of + AlexToken i1 len _ + | Just i2 <- lexDelim i1 -> pure (i1, i2) + | -- is the next token a tab character? + -- need this explicitly because there's a global rule matching $tab + Just ('\t', _) <- alexGetChar' i1 -> setInput i1 >> lexError LexError + | isEOF i1 -> checkSmartQuotes >> lexError LexError + | len == 0 -> panic $ "parsing multiline string got into infinite loop at: " ++ show i0 + | otherwise -> goContent i1 + AlexSkip i1 _ -> goContent i1 + _ -> lexError LexError + + lexDelim = + let go 0 i = Just i + go n i = + case alexGetChar' i of + Just ('"', i') -> go (n - 1) i' + _ -> Nothing + in go (3 :: Int) + + -- See Note [Bare smart quote error] + checkSmartQuotes = do + let findSmartQuote i0@(AI loc _) = + case alexGetChar' i0 of + Just ('\\', i1) | Just (_, i2) <- alexGetChar' i1 -> findSmartQuote i2 + Just (c, i1) + | isDoubleSmartQuote c -> Just (c, loc) + | otherwise -> findSmartQuote i1 + _ -> Nothing + case findSmartQuote (AI (psSpanStart startSpan) startBuf) of + Just (c, loc) -> throwSmartQuoteError c loc + Nothing -> pure () + +-- | Dummy action that should never be called. Should only be used in lex states +-- that are manually lexed in tok_string_multi. +tok_string_multi_content :: Action +tok_string_multi_content = panic "tok_string_multi_content unexpectedly invoked" + +lex_chars :: (String, String) -> PsSpan -> StringBuffer -> Int -> P String +lex_chars (startDelim, endDelim) span buf len = + either (throwStringLexError i0) pure $ + lexString contentLen contentBuf + where + i0@(AI _ contentBuf) = advanceInputBytes (length startDelim) $ AI (psSpanStart span) buf + -- assumes delimiters are ASCII, with 1 byte per Char + contentLen = len - length startDelim - length endDelim -lex_quoted_label :: Action -lex_quoted_label span buf _len _buf2 = do - s <- lex_string StringTypeSingle - (AI end bufEnd) <- getInput - let - token = ITlabelvarid (SourceText src) (mkFastString s) - src = lexemeToFastString (stepOn buf) (cur bufEnd - cur buf - 1) - start = psSpanStart span - - return $ L (mkPsSpan start end) token - - -lex_string :: LexStringType -> P String -lex_string strType = do - start <- getInput - (str, next) <- either fromStringLexError pure $ lexString strType alexGetChar' start - setInput next - pure str - - -lex_char_tok :: Action --- Here we are basically parsing character literals, such as 'x' or '\n' --- but we additionally spot 'x and ''T, returning ITsimpleQuote and --- ITtyQuote respectively, but WITHOUT CONSUMING the x or T part --- (the parser does that). --- So we have to do two characters of lookahead: when we see 'x we need to --- see if there's a trailing quote -lex_char_tok span buf _len _buf2 = do -- We've seen ' - i1 <- getInput -- Look ahead to first character - let loc = psSpanStart span - case alexGetChar' i1 of - Nothing -> lit_error i1 - - Just ('\'', i2@(AI end2 _)) -> do -- We've seen '' - setInput i2 - return (L (mkPsSpan loc end2) ITtyQuote) - - Just ('\\', i2@(AI end2 _)) -> do -- We've seen 'backslash - (lit_ch, i3) <- - either fromStringLexError pure $ - resolveEscapeCharacter alexGetChar' i2 - case alexGetChar' i3 of - Just ('\'', i4) -> do - setInput i4 - finish_char_tok buf loc lit_ch - Just (mc, _) | isSingleSmartQuote mc -> add_smart_quote_error mc end2 - _ -> lit_error i3 - - Just (c, i2@(AI end2 _)) - | not (isAnyChar c) -> lit_error i1 - | otherwise -> - - -- We've seen 'x, where x is a valid character - -- (i.e. not newline etc) but not a quote or backslash - case alexGetChar' i2 of -- Look ahead one more character - Just ('\'', i3) -> do -- We've seen 'x' - setInput i3 - finish_char_tok buf loc c - Just (c, _) | isSingleSmartQuote c -> add_smart_quote_error c end2 - _other -> do -- We've seen 'x not followed by quote - -- (including the possibility of EOF) - -- Just parse the quote only - let (AI end _) = i1 - return (L (mkPsSpan loc end) ITsimpleQuote) - --- We've already seen the closing quote --- Just need to check for trailing # -finish_char_tok :: StringBuffer -> PsLoc -> Char -> P (PsLocated Token) -finish_char_tok buf loc ch = do - i <- getInput - lex_magic_hash i >>= \case - Just i' -> do - setInput i' - -- Include the trailing # in SourceText - let (psSpan, src) = getStringLoc (buf, loc) i' - pure $ L psSpan (ITprimchar src ch) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, loc) i - pure $ L psSpan (ITchar src ch) +throwStringLexError :: AlexInput -> StringLexError -> P a +throwStringLexError i (StringLexError e pos) = setInput (advanceInputTo pos i) >> lexError e --- | Get the span and source text for a string from the given start to the given end. -getStringLoc :: (StringBuffer, PsLoc) -> AlexInput -> (PsSpan, SourceText) -getStringLoc (bufStart, locStart) (AI locEnd bufEnd) = (psSpan, SourceText src) +tok_quoted_label :: Action +tok_quoted_label span buf len _buf2 = do + s <- lex_chars ("#\"", "\"") span buf len + pure $ L span (ITlabelvarid src (mkFastString s)) where - psSpan = mkPsSpan locStart locEnd - src = lexemeToFastString bufStart (cur bufEnd - cur bufStart) - - --- Return Just if we found the magic hash, with the next input. -lex_magic_hash :: AlexInput -> P (Maybe AlexInput) -lex_magic_hash i = do - magicHash <- getBit MagicHashBit - if magicHash - then - case alexGetChar' i of - Just ('#', i') -> pure (Just i') - _other -> pure Nothing - else pure Nothing - -fromStringLexError :: StringLexError AlexInput -> P a -fromStringLexError = \case - UnexpectedEOF i squote -> checkSQuote squote >> throw i LexStringCharLitEOF - BadCharInitialLex i squote -> checkSQuote squote >> throw i LexStringCharLit - EscapeBadChar i -> throw i LexStringCharLit - EscapeUnexpectedEOF i -> throw i LexStringCharLitEOF - EscapeNumRangeError i -> throw i LexNumEscapeRange - EscapeSmartQuoteError c (AI loc _) -> add_smart_quote_error c loc + -- skip leading '#' + src = SourceText . mkFastString . drop 1 $ lexemeToString buf len + + +tok_char :: Action +tok_char span buf len _buf2 = do + c <- lex_chars ("'", "'") span buf (if endsInHash then len - 1 else len) >>= \case + [c] -> pure c + s -> panic $ "tok_char expected exactly one character, got: " ++ show s + pure . L span $ + if endsInHash + then ITprimchar src c + else ITchar src c where - throw i e = setInput i >> lexError e - checkSQuote = \case - NoSmartQuote -> pure () - SmartQuote c (AI loc _) -> add_nonfatal_smart_quote_error c loc - --- before calling lit_error, ensure that the current input is pointing to --- the position of the error in the buffer. This is so that we can report --- a correct location to the user, but also so we can detect UTF-8 decoding --- errors if they occur. -lit_error :: AlexInput -> P a -lit_error i = do setInput i; lexError LexStringCharLit + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + -- ----------------------------------------------------------------------------- -- QuasiQuote @@ -2389,32 +2401,28 @@ quasiquote_error start = do isSmartQuote :: AlexAccPred ExtsBitmap isSmartQuote _ _ _ (AI _ buf) = let c = prevChar buf ' ' in isSingleSmartQuote c || isDoubleSmartQuote c -smart_quote_error_message :: Char -> PsLoc -> MsgEnvelope PsMessage -smart_quote_error_message c loc = - let (correct_char, correct_char_name) = - if isSingleSmartQuote c then ('\'', "Single Quote") else ('"', "Quotation Mark") - err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ - PsErrUnicodeCharLooksLike c correct_char correct_char_name in - err - +throwSmartQuoteError :: Char -> PsLoc -> P a +throwSmartQuoteError c loc = addFatalError err + where + err = + mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ + PsErrUnicodeCharLooksLike c correct_char correct_char_name + (correct_char, correct_char_name) = + if isSingleSmartQuote c + then ('\'', "Single Quote") + else ('"', "Quotation Mark") + +-- | Throw a smart quote error, where the smart quote was the last character lexed smart_quote_error :: Action -smart_quote_error span buf _len _buf2 = do - let c = currentChar buf - addFatalError (smart_quote_error_message c (psSpanStart span)) - -add_smart_quote_error :: Char -> PsLoc -> P a -add_smart_quote_error c loc = addFatalError (smart_quote_error_message c loc) - -add_nonfatal_smart_quote_error :: Char -> PsLoc -> P () -add_nonfatal_smart_quote_error c loc = addError (smart_quote_error_message c loc) +smart_quote_error span _ _ buf2 = do + let c = prevChar buf2 (panic "smart_quote_error unexpectedly called on beginning of input") + throwSmartQuoteError c (psSpanStart span) -advance_to_smart_quote_character :: P () -advance_to_smart_quote_character = do - i <- getInput - case alexGetChar' i of - Just (c, _) | isDoubleSmartQuote c -> return () - Just (_, i2) -> do setInput i2; advance_to_smart_quote_character - Nothing -> return () -- should never get here +-- Note [Bare smart quote error] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- A smart quote inside of a string is allowed, but if a complete valid string +-- couldn't be lexed, we want to see if there's a smart quote that the user +-- thought ended the string, but in fact didn't. -- ----------------------------------------------------------------------------- -- Warnings @@ -2652,7 +2660,7 @@ getLastLocIncludingComments = P $ \s@(PState { prev_loc = prev_loc }) -> POk s p getLastLoc :: P PsSpan getLastLoc = P $ \s@(PState { last_loc = last_loc }) -> POk s last_loc -data AlexInput = AI !PsLoc !StringBuffer +data AlexInput = AI !PsLoc !StringBuffer deriving (Show) {- Note [Unicode in Alex] @@ -2763,6 +2771,19 @@ alexGetChar' (AI loc s) where (c,s') = nextChar s loc' = advancePsLoc loc c +-- | Advance the given input N bytes. +advanceInputBytes :: Int -> AlexInput -> AlexInput +advanceInputBytes n i0@(AI _ buf0) = advanceInputTo (cur buf0 + n) i0 + +-- | Advance the given input to the given position. +advanceInputTo :: Int -> AlexInput -> AlexInput +advanceInputTo pos = go + where + go i@(AI _ buf) + | cur buf >= pos = i + | Just (_, i') <- alexGetChar' i = go i' + | otherwise = i -- reached the end, just return the last input + getInput :: P AlexInput getInput = P $ \s at PState{ loc=l, buffer=b } -> POk s (AI l b) @@ -2770,9 +2791,10 @@ setInput :: AlexInput -> P () setInput (AI l b) = P $ \s -> POk s{ loc=l, buffer=b } () nextIsEOF :: P Bool -nextIsEOF = do - AI _ s <- getInput - return $ atEnd s +nextIsEOF = isEOF <$> getInput + +isEOF :: AlexInput -> Bool +isEOF (AI _ buf) = atEnd buf pushLexState :: Int -> P () pushLexState ls = P $ \s at PState{ lex_state=l } -> POk s{lex_state=ls:l} () @@ -3516,6 +3538,11 @@ topNoLayoutContainsCommas [] = False topNoLayoutContainsCommas (ALRLayout _ _ : ls) = topNoLayoutContainsCommas ls topNoLayoutContainsCommas (ALRNoLayout b _ : _) = b +-- If the generated alexScan/alexScanUser functions are called multiple times +-- in this file, alexScanUser gets broken out into a separate function and +-- increases memory usage. Make sure GHC inlines this function and optimizes it. +{-# INLINE alexScanUser #-} + lexToken :: P (PsLocated Token) lexToken = do inp@(AI loc1 buf) <- getInput ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -1,284 +1,199 @@ {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE PatternSynonyms #-} +{-# LANGUAGE TupleSections #-} +{-# LANGUAGE ViewPatterns #-} module GHC.Parser.String ( StringLexError (..), - ContainsSmartQuote (..), - LexStringType (..), lexString, + lexMultilineString, -- * Unicode smart quote helpers isDoubleSmartQuote, isSingleSmartQuote, - - -- * Other helpers - isAnyChar, - resolveEscapeCharacter, ) where -import GHC.Prelude +import GHC.Prelude hiding (getChar) import Control.Arrow ((>>>)) -import Control.Monad (guard, unless, when) -import Data.Char (chr, isPrint, ord) -import Data.List (unfoldr) +import Control.Monad (when) +import Data.Char (chr, ord) +import qualified Data.Foldable1 as Foldable1 +import qualified Data.List.NonEmpty as NonEmpty import Data.Maybe (listToMaybe, mapMaybe) +import GHC.Data.StringBuffer (StringBuffer) +import qualified GHC.Data.StringBuffer as StringBuffer import GHC.Parser.CharClass ( hexDigit, - is_any, is_decdigit, is_hexdigit, is_octdigit, is_space, octDecDigit, ) +import GHC.Parser.Errors.Types (LexErr (..)) import GHC.Utils.Panic (panic) -data LexStringType = StringTypeSingle | StringTypeMulti +type BufPos = Int +data StringLexError = StringLexError LexErr BufPos --- | State to accumulate while iterating through string literal. --- --- Fields are strict here to avoid space leak when iterating through large string literal --- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 -data LexStringState loc = LexStringState - { stringAcc :: !String - -- ^ The string seen so far, reversed - , multilineCommonWsPrefix :: !Int - -- ^ The common prefix for multiline strings. See Note [Multiline string literals] - , initialLoc :: !loc - -- ^ The location of the beginning of the string literal - } - --- | Get the character at the given location, with the location --- of the next character. Returns Nothing if at the end of the --- input. -type GetChar loc = loc -> Maybe (Char, loc) - -lexString :: LexStringType -> GetChar loc -> loc -> Either (StringLexError loc) (String, loc) -lexString strType getChar initialLoc = go initialState initialLoc +lexString :: Int -> StringBuffer -> Either StringLexError String +lexString = lexStringWith processChars processChars where - initialState = - LexStringState - { stringAcc = "" - , multilineCommonWsPrefix = - case strType of - StringTypeMulti -> maxBound - _ -> 0 - , initialLoc = initialLoc - } - - -- 's' is strict here to avoid space leak when iterating through large string literal - -- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 - go !s loc0 = - case getChar loc0 of - -- found closing delimiter - Just ('"', _) | Just loc1 <- checkDelimiter strType getChar loc0 -> do - let postprocess = - case strType of - StringTypeSingle -> id - StringTypeMulti -> postprocessMultiline (multilineCommonWsPrefix s) - Right (postprocess . reverse $ stringAcc s, loc1) - - -- found backslash - Just (c0@'\\', loc1) -> do - case getChar loc1 of - -- found '\&' character, which should be elided - Just ('&', loc2) -> go s loc2 - -- found start of a string gap - Just (c1, loc2) | is_space c1 -> collapseStringGap getChar s loc2 >>= go s - -- some other escape character - Just (c1, loc2) -> - case strType of - StringTypeSingle -> do - (c', loc') <- resolveEscapeCharacter getChar loc1 - go (addChar c' s) loc' - StringTypeMulti -> do - -- keep escape characters unresolved until after post-processing, - -- to distinguish between a user-newline and the user writing "\n". - -- but still process the characters here, to find any errors - _ <- resolveEscapeCharacter getChar loc1 - go (addChar c1 . addChar c0 $ s) loc2 - -- backslash at end of input - Nothing -> Left $ BadCharInitialLex loc1 (hasSQuote getChar s) - - -- found newline character in multiline string - Just (c0@'\n', loc1) | StringTypeMulti <- strType -> - uncurry go $ parseLeadingWS getChar (addChar c0 s) loc1 - - -- found some other character - Just (c0, loc1) | isAnyChar c0 -> go (addChar c0 s) loc1 - - -- found some unknown character - Just (_, _) -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - - -- reached EOF before finding end of string - Nothing -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) -{-# INLINE lexString #-} - -checkDelimiter :: LexStringType -> GetChar loc -> loc -> Maybe loc -checkDelimiter strType getChar loc0 = - case strType of - StringTypeSingle -> do - ('"', loc1) <- getChar loc0 - Just loc1 - StringTypeMulti -> do - ('"', loc1) <- getChar loc0 - ('"', loc2) <- getChar loc1 - ('"', loc3) <- getChar loc2 - Just loc3 -{-# INLINE checkDelimiter #-} - --- | A helper for adding the given character to the lexed string. -addChar :: Char -> LexStringState loc -> LexStringState loc -addChar c s = s{stringAcc = c : stringAcc s} -{-# INLINE addChar #-} - --- | Return whether the string we've parsed so far contains any smart quotes. -hasSQuote :: GetChar loc -> LexStringState loc -> ContainsSmartQuote loc -hasSQuote getChar s - | any isDoubleSmartQuote (stringAcc s) - , (c, loc) : _ <- filter (isDoubleSmartQuote . fst) allChars = - SmartQuote c loc - | otherwise = - NoSmartQuote + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps + >>> resolveEscapes + +-- ----------------------------------------------------------------------------- +-- Lexing interface + +{- +Note [Lexing strings] +~~~~~~~~~~~~~~~~~~~~~ + +After verifying if a string is lexically valid with Alex, we still need to do +some post processing of the string, namely: +1. Collapse string gaps +2. Resolve escape characters + +The problem: 'lexemeToString' is more performant than manually reading +characters from the StringBuffer. However, that completely erases the position +of each character, which we need in order to report the correct position for +error messages (e.g. when resolving escape characters). + +So what we'll do is do two passes. The first pass is optimistic; just convert +to a plain String and process it. If this results in an error, we do a second +pass, this time where each character is annotated with its position. Now, the +error has all the information it needs. + +Ideally, lexStringWith would take a single (forall c. HasChar c => ...) function, +but to help the specializer, we pass it in twice to concretize it for the two +types we actually use. +-} + +-- | See Note [Lexing strings] +lexStringWith :: + ([Char] -> Either (Char, LexErr) [Char]) + -> ([CharPos] -> Either (CharPos, LexErr) [CharPos]) + -> Int + -> StringBuffer + -> Either StringLexError String +lexStringWith processChars processCharsPos len buf = + case processChars $ bufferChars buf len of + Right s -> Right s + Left _ -> + case processCharsPos $ bufferLocatedChars buf len of + Right _ -> panic "expected lex error on second pass" + Left ((_, pos), e) -> Left $ StringLexError e pos + +class HasChar c where + getChar :: c -> Char + setChar :: Char -> c -> c + +instance HasChar Char where + getChar = id + setChar = const + +instance HasChar (Char, x) where + getChar = fst + setChar c (_, x) = (c, x) + +pattern Char :: HasChar c => Char -> c +pattern Char c <- (getChar -> c) +{-# COMPLETE Char #-} + +bufferChars :: StringBuffer -> Int -> [Char] +bufferChars = StringBuffer.lexemeToString + +type CharPos = (Char, BufPos) + +bufferLocatedChars :: StringBuffer -> Int -> [CharPos] +bufferLocatedChars initialBuf len = go initialBuf where - allChars = unfoldr getCharWithLoc (initialLoc s) - getCharWithLoc loc = - case getChar loc of - Just (c, loc') -> Just ((c, loc), loc') - Nothing -> Nothing -{-# INLINE hasSQuote #-} - --- | After parsing a backslash and a space character, consume the rest of --- the string gap and return the next location. -collapseStringGap :: GetChar loc -> LexStringState loc -> loc -> Either (StringLexError loc) loc -collapseStringGap getChar s = go + go buf + | atEnd buf = [] + | otherwise = + let (c, buf') = StringBuffer.nextChar buf + in (c, StringBuffer.cur buf) : go buf' + + atEnd buf = StringBuffer.byteDiff initialBuf buf >= len + +-- ----------------------------------------------------------------------------- +-- Lexing phases + +collapseGaps :: HasChar c => [c] -> [c] +collapseGaps = go where - go loc0 = - case getChar loc0 of - Just ('\\', loc1) -> pure loc1 - Just (c0, loc1) | is_space c0 -> go loc1 - Just _ -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - Nothing -> Left $ UnexpectedEOF loc0 (hasSQuote getChar s) -{-# INLINE collapseStringGap #-} + go = \case + c1@(Char '\\') : c2@(Char c) : cs + | is_space c -> go $ dropGap cs + | otherwise -> c1 : c2 : go cs + c : cs -> c : go cs + [] -> [] --- | See Note [Multiline string literals] -parseLeadingWS :: GetChar loc -> LexStringState loc -> loc -> (LexStringState loc, loc) -parseLeadingWS getChar = go 0 + dropGap = \case + Char '\\' : cs -> cs + _ : cs -> dropGap cs + [] -> panic "gap unexpectedly ended" + +resolveEscapes :: HasChar c => [c] -> Either (c, LexErr) [c] +resolveEscapes = go dlistEmpty where - go !col s loc = - case getChar loc of - Just (c@' ', loc') -> go (col + 1) (addChar c s) loc' - -- expand tabs - Just ('\t', loc') -> - let fill = 8 - (col `mod` 8) - s' = applyN fill (addChar ' ') s - in go (col + fill) s' loc' - -- if we see a newline or string delimiter, then this line only contained whitespace, so - -- don't include it in the common whitespace prefix - Just ('\n', _) -> (s, loc) - Just ('"', _) | Just _ <- checkDelimiter StringTypeMulti getChar loc -> (s, loc) - -- found some other character, so we're done parsing leading whitespace - _ -> - let s' = s{multilineCommonWsPrefix = min col (multilineCommonWsPrefix s)} - in (s', loc) - - applyN :: Int -> (a -> a) -> a -> a - applyN n f x0 = iterate f x0 !! n -{-# INLINE parseLeadingWS #-} - -data StringLexError loc - = UnexpectedEOF !loc !(ContainsSmartQuote loc) - -- ^ Unexpectedly hit EOF when lexing string - | BadCharInitialLex !loc !(ContainsSmartQuote loc) - -- ^ Found invalid character when initially lexing string - | EscapeBadChar !loc - -- ^ Found invalid character when parsing an escaped character - | EscapeUnexpectedEOF !loc - -- ^ Unexpectedly hit EOF when parsing an escaped character - | EscapeNumRangeError !loc - -- ^ Escaped number exceeds range - | EscapeSmartQuoteError !Char !loc - -- ^ Found escaped smart unicode chars as `\’` or `\”` - deriving (Show) - --- | When initially lexing the string, we want to track if we've --- seen a smart quote, to show a helpful "you might be accidentally --- using a smart quote" error. -data ContainsSmartQuote loc - = NoSmartQuote - | SmartQuote !Char !loc - deriving (Show) + go !acc = \case + [] -> pure $ dlistToList acc + Char '\\' : Char '&' : cs -> go acc cs + backslash@(Char '\\') : cs -> + case resolveEscapeChar cs of + Right (esc, cs') -> go (acc `dlistSnoc` setChar esc backslash) cs' + Left (c, e) -> Left (c, e) + c : cs -> go (acc `dlistSnoc` c) cs -- ----------------------------------------------------------------------------- -- Escape characters --- | After finding a backslash, parse the rest of the escape character, starting --- at the given location. -resolveEscapeCharacter :: GetChar loc -> loc -> Either (StringLexError loc) (Char, loc) -resolveEscapeCharacter getChar loc0 = do - (c0, loc1) <- expectChar loc0 - case c0 of - 'a' -> pure ('\a', loc1) - 'b' -> pure ('\b', loc1) - 'f' -> pure ('\f', loc1) - 'n' -> pure ('\n', loc1) - 'r' -> pure ('\r', loc1) - 't' -> pure ('\t', loc1) - 'v' -> pure ('\v', loc1) - '\\' -> pure ('\\', loc1) - '"' -> pure ('\"', loc1) - '\'' -> pure ('\'', loc1) - -- escape codes - 'x' -> expectNum is_hexdigit 16 hexDigit loc1 - 'o' -> expectNum is_octdigit 8 octDecDigit loc1 - _ | is_decdigit c0 -> expectNum is_decdigit 10 octDecDigit loc0 - -- control characters (e.g. '\^M') - '^' -> do - (c1, loc2) <- expectChar loc1 - unless ('@' <= c1 && c1 <= '_') $ Left $ EscapeBadChar loc1 - pure (chr $ ord c1 - ord '@', loc2) - -- long form escapes (e.g. '\NUL') - _ | Just (c1, loc2) <- parseLongEscape getChar c0 loc1 -> pure (c1, loc2) - -- check unicode smart quotes (#21843) - _ | isDoubleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - _ | isSingleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - -- unknown escape - _ -> Left $ EscapeBadChar loc0 +-- | Resolve a escape character, after having just lexed a backslash. +-- Assumes escape character is valid. +resolveEscapeChar :: HasChar c => [c] -> Either (c, LexErr) (Char, [c]) +resolveEscapeChar = \case + Char 'a' : cs -> pure ('\a', cs) + Char 'b' : cs -> pure ('\b', cs) + Char 'f' : cs -> pure ('\f', cs) + Char 'n' : cs -> pure ('\n', cs) + Char 'r' : cs -> pure ('\r', cs) + Char 't' : cs -> pure ('\t', cs) + Char 'v' : cs -> pure ('\v', cs) + Char '\\' : cs -> pure ('\\', cs) + Char '"' : cs -> pure ('\"', cs) + Char '\'' : cs -> pure ('\'', cs) + -- escape codes + Char 'x' : cs -> parseNum is_hexdigit 16 hexDigit cs + Char 'o' : cs -> parseNum is_octdigit 8 octDecDigit cs + cs@(Char c : _) | is_decdigit c -> parseNum is_decdigit 10 octDecDigit cs + -- control characters (e.g. '\^M') + Char '^' : Char c : cs -> pure (chr $ ord c - ord '@', cs) + -- long form escapes (e.g. '\NUL') + cs | Just (esc, cs') <- parseLongEscape cs -> pure (esc, cs') + -- shouldn't happen + Char c : _ -> panic $ "found unexpected escape character: " ++ show c + [] -> panic "escape character unexpectedly ended" where - expectChar loc = - case getChar loc of - Just x -> pure x - Nothing -> Left $ EscapeUnexpectedEOF loc - - expectNum isDigit base toDigit loc1 = do - (c1, loc2) <- expectChar loc1 - unless (isDigit c1) $ Left $ EscapeBadChar loc1 - let parseNum x loc = - case getChar loc of - Just (c, loc') | isDigit c -> do - let x' = x * base + toDigit c - when (x' > 0x10ffff) $ Left $ EscapeNumRangeError loc - parseNum x' loc' - _ -> - pure (chr x, loc) - parseNum (toDigit c1) loc2 -{-# INLINE resolveEscapeCharacter #-} - -parseLongEscape :: GetChar loc -> Char -> loc -> Maybe (Char, loc) -parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCodes + parseNum isDigit base toDigit = + let go x = \case + ch@(Char c) : cs | isDigit c -> do + let x' = x * base + toDigit c + when (x' > 0x10ffff) $ Left (ch, LexNumEscapeRange) + go x' cs + cs -> pure (chr x, cs) + in go 0 + +parseLongEscape :: HasChar c => [c] -> Maybe (Char, [c]) +parseLongEscape cs = listToMaybe (mapMaybe tryParse longEscapeCodes) where - tryParse (prefix, c) = do - p0 : p <- pure prefix - guard (p0 == c0) -- see if the first character matches - loc <- parsePrefix loc1 p -- see if the rest of the prefix matches - pure (c, loc) - - parsePrefix loc = \case - [] -> pure loc - p : ps -> do - (c, loc') <- getChar loc - guard (p == c) - parsePrefix loc' ps + tryParse (code, esc) = + case splitAt (length code) cs of + (pre, cs') | map getChar pre == code -> Just (esc, cs') + _ -> Nothing longEscapeCodes = [ ("NUL", '\NUL') @@ -289,14 +204,14 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("ENQ", '\ENQ') , ("ACK", '\ACK') , ("BEL", '\BEL') - , ("BS", '\BS') - , ("HT", '\HT') - , ("LF", '\LF') - , ("VT", '\VT') - , ("FF", '\FF') - , ("CR", '\CR') - , ("SO", '\SO') - , ("SI", '\SI') + , ("BS" , '\BS' ) + , ("HT" , '\HT' ) + , ("LF" , '\LF' ) + , ("VT" , '\VT' ) + , ("FF" , '\FF' ) + , ("CR" , '\CR' ) + , ("SO" , '\SO' ) + , ("SI" , '\SI' ) , ("DLE", '\DLE') , ("DC1", '\DC1') , ("DC2", '\DC2') @@ -306,17 +221,16 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("SYN", '\SYN') , ("ETB", '\ETB') , ("CAN", '\CAN') - , ("EM", '\EM') + , ("EM" , '\EM' ) , ("SUB", '\SUB') , ("ESC", '\ESC') - , ("FS", '\FS') - , ("GS", '\GS') - , ("RS", '\RS') - , ("US", '\US') - , ("SP", '\SP') + , ("FS" , '\FS' ) + , ("GS" , '\GS' ) + , ("RS" , '\RS' ) + , ("US" , '\US' ) + , ("SP" , '\SP' ) , ("DEL", '\DEL') ] -{-# INLINE parseLongEscape #-} -- ----------------------------------------------------------------------------- -- Unicode Smart Quote detection (#21843) @@ -333,6 +247,98 @@ isSingleSmartQuote = \case '’' -> True _ -> False +-- ----------------------------------------------------------------------------- +-- Multiline strings + +-- | See Note [Multiline string literals] +-- +-- Assumes string is lexically valid. Skips the steps about splitting +-- and rejoining lines, and instead manually find newline characters, +-- for performance. +lexMultilineString :: Int -> StringBuffer -> Either StringLexError String +lexMultilineString = lexStringWith processChars processChars + where + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps -- Step 1 + >>> expandLeadingTabs -- Step 3 + >>> rmCommonWhitespacePrefix -- Step 4 + >>> collapseOnlyWsLines -- Step 5 + >>> rmFirstNewline -- Step 7a + >>> rmLastNewline -- Step 7b + >>> resolveEscapes -- Step 8 + + -- expands all tabs, since the lexer will verify that tabs can only appear + -- as leading indentation + expandLeadingTabs :: HasChar c => [c] -> [c] + expandLeadingTabs = + let go !col = \case + c@(Char '\t') : cs -> + let fill = 8 - (col `mod` 8) + in replicate fill (setChar ' ' c) ++ go (col + fill) cs + c : cs -> c : go (if getChar c == '\n' then 0 else col + 1) cs + [] -> [] + in go 0 + + rmCommonWhitespacePrefix :: HasChar c => [c] -> [c] + rmCommonWhitespacePrefix cs0 = + let commonWSPrefix = getCommonWsPrefix (map getChar cs0) + go = \case + c@(Char '\n') : cs -> c : go (dropLine commonWSPrefix cs) + c : cs -> c : go cs + [] -> [] + -- drop x characters from the string, or up to a newline, whichever + -- comes first + dropLine !x = \case + cs | x <= 0 -> cs + cs@(Char '\n' : _) -> cs + _ : cs -> dropLine (x - 1) cs + [] -> [] + in go cs0 + + collapseOnlyWsLines :: HasChar c => [c] -> [c] + collapseOnlyWsLines = + let go = \case + c@(Char '\n') : cs | Just cs' <- checkAllWs cs -> c : go cs' + c : cs -> c : go cs + [] -> [] + checkAllWs = \case + -- got all the way to a newline or the end of the string, return + cs@(Char '\n' : _) -> Just cs + cs@[] -> Just cs + -- found whitespace, continue + Char c : cs | is_space c -> checkAllWs cs + -- anything else, stop + _ -> Nothing + in go + + rmFirstNewline :: HasChar c => [c] -> [c] + rmFirstNewline = \case + Char '\n' : cs -> cs + cs -> cs + + rmLastNewline :: HasChar c => [c] -> [c] + rmLastNewline = + let go = \case + [] -> [] + [Char '\n'] -> [] + c : cs -> c : go cs + in go + +-- | See step 4 in Note [Multiline string literals] +-- +-- Assumes tabs have already been expanded. +getCommonWsPrefix :: String -> Int +getCommonWsPrefix s = + case NonEmpty.nonEmpty includedLines of + Nothing -> 0 + Just ls -> Foldable1.minimum $ NonEmpty.map (length . takeWhile is_space) ls + where + includedLines = + filter (not . all is_space) -- ignore whitespace-only lines + . drop 1 -- ignore first line in calculation + $ lines s + {- Note [Multiline string literals] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -351,23 +357,13 @@ The canonical steps for post processing a multiline string are: 2. Split the string by newlines 3. Convert leading tabs into spaces * In each line, any tabs preceding non-whitespace characters are replaced with spaces up to the next tab stop -4. Remove common whitespace prefix in every line (see below) +4. Remove common whitespace prefix in every line except the first (see below) 5. If a line contains only whitespace, remove all of the whitespace 6. Join the string back with `\n` delimiters -7. If the first character of the string is a newline, remove it +7a. If the first character of the string is a newline, remove it +7b. If the last character of the string is a newline, remove it 8. Interpret escaped characters -However, for performance reasons, we do as much of this in one pass as possible: -1. As we lex the string, do the following steps as they appear: - a. Collapse string gaps - b. Keep track of the common whitespace prefix so far - c. Validate escaped characters -2. At the very end, post process the lexed string: - a. Remove the common whitespace prefix from every line - b. Remove all whitespace from all-whitespace lines - c. Remove initial newline character - d. Resolve escaped characters - The common whitespace prefix can be informally defined as "The longest prefix of whitespace shared by all lines in the string, excluding the first line and any whitespace-only lines". @@ -382,75 +378,16 @@ It's more precisely defined with the following algorithm: 3. Calculate the longest prefix of whitespace shared by all lines in the remaining list -} --- | See Note [Multiline string literals] -postprocessMultiline :: Int -> String -> String -postprocessMultiline commonWSPrefix = - rmCommonWhitespacePrefix - >>> collapseOnlyWsLines - >>> rmFirstNewline - >>> rmLastNewline - >>> resolveEscapeChars - where - rmCommonWhitespacePrefix = - let go = \case - '\n' : s -> '\n' : go (dropLine commonWSPrefix s) - c : s -> c : go s - [] -> [] - -- drop x characters from the string, or up to a newline, whichever - -- comes first - dropLine !x = \case - s | x <= 0 -> s - s@('\n' : _) -> s - _ : s -> dropLine (x - 1) s - [] -> [] - in go - - collapseOnlyWsLines = - let go = \case - '\n' : s | Just s' <- checkAllWs s -> '\n' : go s' - c : s -> c : go s - [] -> [] - checkAllWs = \case - -- got all the way to a newline or the end of the string, return - s@('\n' : _) -> Just s - s@[] -> Just s - -- found whitespace, continue - c : s | is_space c -> checkAllWs s - -- anything else, stop - _ -> Nothing - in go - - rmFirstNewline = \case - '\n' : s -> s - s -> s +-- ----------------------------------------------------------------------------- +-- DList - rmLastNewline = - let go = \case - [] -> [] - ['\n'] -> [] - c : cs -> c : go cs - in go +newtype DList a = DList ([a] -> [a]) - -- resolve escape characters, deferred from lexString. guaranteed - -- to not throw any errors, since we already checked them in lexString - resolveEscapeChars = \case - [] -> [] - '\\' : s -> - -- concretizing 'loc' to String: - -- resolveEscapeCharacter :: (String -> Maybe (Char, String)) -> String -> Either _ (Char, String) - case resolveEscapeCharacter uncons s of - Left e -> panic $ "resolving escape characters in multiline string unexpectedly found errors: " ++ show e - Right (c, s') -> c : resolveEscapeChars s' - c : s -> c : resolveEscapeChars s - - uncons = \case - c : cs -> Just (c, cs) - [] -> Nothing +dlistEmpty :: DList a +dlistEmpty = DList id --- ----------------------------------------------------------------------------- --- Helpers +dlistToList :: DList a -> [a] +dlistToList (DList f) = f [] -isAnyChar :: Char -> Bool -isAnyChar c - | c > '\x7f' = isPrint c - | otherwise = is_any c +dlistSnoc :: DList a -> a -> DList a +dlistSnoc (DList f) x = DList (f . (x :)) ===================================== testsuite/driver/testlib.py ===================================== @@ -2868,7 +2868,7 @@ def normalise_errmsg(s: str) -> str: # normalise slashes to minimise Windows/Unix filename differences, # but don't normalize backslashes in chars - s = re.sub(r"(?!')\\", '/', s) + s = re.sub(r"(?:9:1: error: [GHC-58481] parse error on input ‘+’ -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' + +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/ghci/scripts/ghci022.stderr ===================================== @@ -1,3 +1,2 @@ +ghci022.hs:2:5: error: [GHC-21231] lexical error at character '\n' -ghci022.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr ===================================== @@ -1,3 +1,3 @@ - MultilineStringsInnerTab.hs:8:5: error: [GHC-21231] - lexical error in string/character literal at character '\t' + lexical error at character '\t' + ===================================== testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr ===================================== @@ -1,6 +1,2 @@ - MultilineStringsSmartQuotes.hs:9:3: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -MultilineStringsSmartQuotes.hs:10:1: [GHC-21231] - lexical error in string/character literal at end of input ===================================== testsuite/tests/parser/should_fail/T21843c.stderr ===================================== @@ -1,6 +1,2 @@ - -T21843c.hs:3:19: [GHC-31623] +T21843c.hs:3:13: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -T21843c.hs:3:20: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/T21843e.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843e.hs:3:15: [GHC-31623] +T21843e.hs:3:13: error: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not + ===================================== testsuite/tests/parser/should_fail/T21843f.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843f.hs:3:13: [GHC-31623] +T21843f.hs:3:11: error: [GHC-31623] Unicode character '‘' ('\8216') looks like ''' (Single Quote), but it is not + ===================================== testsuite/tests/parser/should_fail/T3751.stderr ===================================== @@ -1,3 +1,2 @@ +T3751.hs:3:5: error: [GHC-21231] lexical error at character '\167' -T3751.hs:3:7: error: [GHC-21231] - lexical error in string/character literal at character '\167' ===================================== testsuite/tests/parser/should_fail/T5425.stderr ===================================== @@ -1,3 +1,2 @@ +T5425.hs:4:1: error: [GHC-21231] lexical error at character '\955' -T5425.hs:5:2: error: [GHC-21231] - lexical error in string/character literal at character '\955' ===================================== testsuite/tests/parser/should_fail/readFail002.stderr ===================================== @@ -1,3 +1,3 @@ +readFail002.hs:5:5: error: [GHC-21231] + lexical error at character '\n' -readFail002.hs:5:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/readFail004.stderr ===================================== @@ -1,3 +1,3 @@ +readFail004.hs:17:16: error: [GHC-21231] + lexical error at character '.' -readFail004.hs:19:1: error: [GHC-21231] - lexical error in string/character literal at character '.' ===================================== testsuite/tests/parser/should_fail/readFail005.stderr ===================================== @@ -1,3 +1,2 @@ - -readFail005.hs:4:7: error: [GHC-21231] - lexical error in string/character literal at character '&' +readFail005.hs:4:5: [GHC-21231] + lexical error at character '&' ===================================== testsuite/tests/parser/should_fail/readFail033.stderr ===================================== @@ -1,3 +1,3 @@ +readFail033.hs:2:5: error: [GHC-21231] + lexical error at character '\t' -readFail033.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\t' ===================================== testsuite/tests/parser/unicode/all.T ===================================== @@ -33,3 +33,6 @@ test('T18225B', normal, compile_fail, ['']) test('T18158', normal, compile, ['']) test('T18158b', normal, compile_fail, ['']) + +test('lex_unispace', normal, compile, ['']) +test('lex_unicode_ids', normal, compile, ['']) ===================================== testsuite/tests/parser/unicode/lex_unicode_ids.hs ===================================== @@ -0,0 +1,11 @@ +-- Regression tests for unicode identifiers + +{-# LANGUAGE TemplateHaskellQuotes #-} + +module ShouldCompile where + +ƞ = 1 +eta = 'ƞ + +data Ʊ +upsilon = ''Ʊ ===================================== testsuite/tests/parser/unicode/lex_unispace.hs ===================================== @@ -0,0 +1,9 @@ +-- Regression tests for unicode whitespace + +module ShouldCompile where + +-- https://github.com/fosskers/aura/blob/cdebca1f48254ebb8286d8e38591bf644282866f/haskell/aura/lib/Aura/Languages.hs#L107 +x1 = '　' -- \12288 + +-- https://github.com/jgm/pandoc/blob/98e77e02f6436e4b74a164762d0f3149ae7ecefa/src/Text/Pandoc/Writers/FB2.hs#L295C3-L295C32 +x2 = " " -- \xa0 ===================================== testsuite/tests/parser/unicode/utf8_010.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_010.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_011.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_011.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_020.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_020.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_021.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_021.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_022.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_022.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs ===================================== @@ -320,7 +320,7 @@ classify tok = ITlabelvarid{} -> TkUnknown ITchar{} -> TkChar ITstring{} -> TkString - ITmultilinestring{} -> TkString + ITstringMulti{} -> TkString ITinteger{} -> TkNumber ITrational{} -> TkNumber ITprimchar{} -> TkChar View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e7c83fe4834cbeb0c45c33a08d589133bfce7e10 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e7c83fe4834cbeb0c45c33a08d589133bfce7e10 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 06:34:32 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sun, 22 Sep 2024 02:34:32 -0400 Subject: [Git][ghc/ghc][master] rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. Message-ID: <66efba787e18c_1d20eb1015d3c1277bf@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - 1 changed file: - rts/Profiling.c Changes: ===================================== rts/Profiling.c ===================================== @@ -46,6 +46,7 @@ static Arena *prof_arena; static unsigned int CC_ID = 1; static unsigned int CCS_ID = 1; +static unsigned int DUMPED_CC_ID = 0; // we have dumped all CCs up to this id to the eventlog /* Globals for opening the profiling log file(s) */ @@ -138,10 +139,11 @@ dumpCostCentresToEventLog(void) { #if defined(PROFILING) CostCentre *cc, *next; - for (cc = CC_LIST; cc != NULL; cc = next) { + for (cc = CC_LIST; cc != NULL && cc->ccID != DUMPED_CC_ID; cc = next) { next = cc->link; traceHeapProfCostCentre(cc->ccID, cc->label, cc->module, cc->srcloc, cc->is_caf); + DUMPED_CC_ID = cc->ccID < DUMPED_CC_ID ? DUMPED_CC_ID : cc->ccID; } #endif } @@ -193,8 +195,6 @@ void initProfiling (void) if (RtsFlags.CcFlags.doCostCentres) { initTimeProfiling(); } - - traceInitEvent(dumpCostCentresToEventLog); } @@ -205,6 +205,7 @@ void initProfiling (void) void refreshProfilingCCSs (void) { ACQUIRE_LOCK(&ccs_mutex); + traceInitEvent(dumpCostCentresToEventLog); // make CCS_MAIN the parent of all the pre-defined CCSs. CostCentreStack *next; for (CostCentreStack *ccs = CCS_LIST; ccs != NULL; ) { View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ca67d7cbc328c8e991e82f6cf1f3f75c403eed67 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ca67d7cbc328c8e991e82f6cf1f3f75c403eed67 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 06:35:06 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sun, 22 Sep 2024 02:35:06 -0400 Subject: [Git][ghc/ghc][master] EPA: Replace AnnsModule am_main with EpTokens Message-ID: <66efba9a14632_1d20eb12a068813127b@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 26 changed files: - compiler/GHC/Hs.hs - compiler/GHC/Parser.y - testsuite/tests/ghc-api/exactprint/T22919.stderr - testsuite/tests/ghc-api/exactprint/Test20239.stderr - testsuite/tests/ghc-api/exactprint/ZeroWidthSemi.stderr - testsuite/tests/haddock/should_compile_flag_haddock/T17544.stderr - testsuite/tests/haddock/should_compile_flag_haddock/T17544_kw.stderr - testsuite/tests/haddock/should_compile_flag_haddock/T24221.stderr - testsuite/tests/module/mod185.stderr - testsuite/tests/parser/should_compile/DumpParsedAst.stderr - testsuite/tests/parser/should_compile/DumpParsedAstComments.stderr - testsuite/tests/parser/should_compile/DumpSemis.stderr - testsuite/tests/parser/should_compile/KindSigs.stderr - testsuite/tests/parser/should_compile/T15323.stderr - testsuite/tests/parser/should_compile/T20452.stderr - testsuite/tests/parser/should_compile/T20718.stderr - testsuite/tests/parser/should_compile/T20718b.stderr - testsuite/tests/parser/should_compile/T20846.stderr - testsuite/tests/parser/should_compile/T23315/T23315.stderr - testsuite/tests/printer/AnnotationNoListTuplePuns.stdout - testsuite/tests/printer/T18791.stderr - testsuite/tests/printer/Test20297.stdout - testsuite/tests/printer/Test24533.stdout - utils/check-exact/ExactPrint.hs - utils/check-exact/Parsers.hs - utils/check-exact/Utils.hs Changes: ===================================== compiler/GHC/Hs.hs ===================================== @@ -18,6 +18,7 @@ therefore, is almost nothing but re-exporting. {-# LANGUAGE ConstraintKinds #-} {-# LANGUAGE TypeFamilies #-} {-# LANGUAGE FlexibleInstances #-} -- For deriving instance Data +{-# LANGUAGE DataKinds #-} module GHC.Hs ( module Language.Haskell.Syntax, @@ -98,7 +99,9 @@ deriving instance Data (HsModule GhcPs) data AnnsModule = AnnsModule { - am_main :: [AddEpAnn], + am_sig :: EpToken "signature", + am_mod :: EpToken "module", + am_where :: EpToken "where", am_decls :: [TrailingAnn], -- ^ Semis before the start of top decls am_cs :: [LEpaComment], -- ^ Comments before start of top decl, -- used in exact printing only @@ -106,7 +109,7 @@ data AnnsModule } deriving (Data, Eq) instance NoAnn AnnsModule where - noAnn = AnnsModule [] [] [] Nothing + noAnn = AnnsModule NoEpTok NoEpTok NoEpTok [] [] Nothing instance Outputable (HsModule GhcPs) where ppr (HsModule { hsmodExt = XModulePs { hsmodHaddockModHeader = mbDoc } ===================================== compiler/GHC/Parser.y ===================================== @@ -911,7 +911,7 @@ signature :: { Located (HsModule GhcPs) } : 'signature' modid maybe_warning_pragma maybeexports 'where' body {% fileSrcSpan >>= \ loc -> acs loc (\loc cs-> (L loc (HsModule (XModulePs - (EpAnn (spanAsAnchor loc) (AnnsModule [mj AnnSignature $1, mj AnnWhere $5] (fstOf3 $6) [] Nothing) cs) + (EpAnn (spanAsAnchor loc) (AnnsModule (epTok $1) NoEpTok (epTok $5) (fstOf3 $6) [] Nothing) cs) (thdOf3 $6) $3 Nothing) (Just $2) $4 (fst $ sndOf3 $6) (snd $ sndOf3 $6))) @@ -921,7 +921,7 @@ module :: { Located (HsModule GhcPs) } : 'module' modid maybe_warning_pragma maybeexports 'where' body {% fileSrcSpan >>= \ loc -> acsFinal (\cs eof -> (L loc (HsModule (XModulePs - (EpAnn (spanAsAnchor loc) (AnnsModule [mj AnnModule $1, mj AnnWhere $5] (fstOf3 $6) [] eof) cs) + (EpAnn (spanAsAnchor loc) (AnnsModule NoEpTok (epTok $1) (epTok $5) (fstOf3 $6) [] eof) cs) (thdOf3 $6) $3 Nothing) (Just $2) $4 (fst $ sndOf3 $6) (snd $ sndOf3 $6)) @@ -929,7 +929,7 @@ module :: { Located (HsModule GhcPs) } | body2 {% fileSrcSpan >>= \ loc -> acsFinal (\cs eof -> (L loc (HsModule (XModulePs - (EpAnn (spanAsAnchor loc) (AnnsModule [] (fstOf3 $1) [] eof) cs) + (EpAnn (spanAsAnchor loc) (AnnsModule NoEpTok NoEpTok NoEpTok (fstOf3 $1) [] eof) cs) (thdOf3 $1) Nothing Nothing) Nothing Nothing (fst $ sndOf3 $1) (snd $ sndOf3 $1)))) } @@ -969,14 +969,14 @@ header :: { Located (HsModule GhcPs) } : 'module' modid maybe_warning_pragma maybeexports 'where' header_body {% fileSrcSpan >>= \ loc -> acs loc (\loc cs -> (L loc (HsModule (XModulePs - (EpAnn (spanAsAnchor loc) (AnnsModule [mj AnnModule $1,mj AnnWhere $5] [] [] Nothing) cs) + (EpAnn (spanAsAnchor loc) (AnnsModule NoEpTok (epTok $1) (epTok $5) [] [] Nothing) cs) EpNoLayout $3 Nothing) (Just $2) $4 $6 [] ))) } | 'signature' modid maybe_warning_pragma maybeexports 'where' header_body {% fileSrcSpan >>= \ loc -> acs loc (\loc cs -> (L loc (HsModule (XModulePs - (EpAnn (spanAsAnchor loc) (AnnsModule [mj AnnModule $1,mj AnnWhere $5] [] [] Nothing) cs) + (EpAnn (spanAsAnchor loc) (AnnsModule NoEpTok (epTok $1) (epTok $5) [] [] Nothing) cs) EpNoLayout $3 Nothing) (Just $2) $4 $6 [] ))) } ===================================== testsuite/tests/ghc-api/exactprint/T22919.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T22919.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T22919.hs:1:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T22919.hs:1:29-33 }))] + (NoEpTok) + (EpTok + (EpaSpan { T22919.hs:1:1-6 })) + (EpTok + (EpaSpan { T22919.hs:1:29-33 })) [] [] (Just ===================================== testsuite/tests/ghc-api/exactprint/Test20239.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { Test20239.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { Test20239.hs:1:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { Test20239.hs:1:18-22 }))] + (NoEpTok) + (EpTok + (EpaSpan { Test20239.hs:1:1-6 })) + (EpTok + (EpaSpan { Test20239.hs:1:18-22 })) [] [] (Just @@ -351,6 +354,6 @@ []))))))])) - Test20239.hs:5:15: error: [GHC-76037] Not in scope: type constructor or class ‘Method’ + ===================================== testsuite/tests/ghc-api/exactprint/ZeroWidthSemi.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { ZeroWidthSemi.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { ZeroWidthSemi.hs:1:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { ZeroWidthSemi.hs:1:22-26 }))] + (NoEpTok) + (EpTok + (EpaSpan { ZeroWidthSemi.hs:1:1-6 })) + (EpTok + (EpaSpan { ZeroWidthSemi.hs:1:22-26 })) [] [] (Just ===================================== testsuite/tests/haddock/should_compile_flag_haddock/T17544.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T17544.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T17544.hs:3:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T17544.hs:3:15-19 }))] + (NoEpTok) + (EpTok + (EpaSpan { T17544.hs:3:1-6 })) + (EpTok + (EpaSpan { T17544.hs:3:15-19 })) [] [] (Just ===================================== testsuite/tests/haddock/should_compile_flag_haddock/T17544_kw.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T17544_kw.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T17544_kw.hs:11:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T17544_kw.hs:13:13-17 }))] + (NoEpTok) + (EpTok + (EpaSpan { T17544_kw.hs:11:1-6 })) + (EpTok + (EpaSpan { T17544_kw.hs:13:13-17 })) [] [] (Just ===================================== testsuite/tests/haddock/should_compile_flag_haddock/T24221.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T24221.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T24221.hs:1:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T24221.hs:1:15-19 }))] + (NoEpTok) + (EpTok + (EpaSpan { T24221.hs:1:1-6 })) + (EpTok + (EpaSpan { T24221.hs:1:15-19 })) [] [] (Just ===================================== testsuite/tests/module/mod185.stderr ===================================== @@ -8,7 +8,9 @@ (EpAnn (EpaSpan { mod185.hs:1:1 }) (AnnsModule - [] + (NoEpTok) + (NoEpTok) + (NoEpTok) [] [] (Just ===================================== testsuite/tests/parser/should_compile/DumpParsedAst.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { DumpParsedAst.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { DumpParsedAst.hs:4:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { DumpParsedAst.hs:4:22-26 }))] + (NoEpTok) + (EpTok + (EpaSpan { DumpParsedAst.hs:4:1-6 })) + (EpTok + (EpaSpan { DumpParsedAst.hs:4:22-26 })) [] [] (Just ===================================== testsuite/tests/parser/should_compile/DumpParsedAstComments.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { DumpParsedAstComments.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { DumpParsedAstComments.hs:5:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { DumpParsedAstComments.hs:5:30-34 }))] + (NoEpTok) + (EpTok + (EpaSpan { DumpParsedAstComments.hs:5:1-6 })) + (EpTok + (EpaSpan { DumpParsedAstComments.hs:5:30-34 })) [] [] (Just ===================================== testsuite/tests/parser/should_compile/DumpSemis.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { DumpSemis.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { DumpSemis.hs:1:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { DumpSemis.hs:1:18-22 }))] + (NoEpTok) + (EpTok + (EpaSpan { DumpSemis.hs:1:1-6 })) + (EpTok + (EpaSpan { DumpSemis.hs:1:18-22 })) [(AddSemiAnn (EpaSpan { DumpSemis.hs:4:1 })) ,(AddSemiAnn ===================================== testsuite/tests/parser/should_compile/KindSigs.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { KindSigs.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { KindSigs.hs:6:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { KindSigs.hs:6:17-21 }))] + (NoEpTok) + (EpTok + (EpaSpan { KindSigs.hs:6:1-6 })) + (EpTok + (EpaSpan { KindSigs.hs:6:17-21 })) [] [] (Just ===================================== testsuite/tests/parser/should_compile/T15323.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T15323.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T15323.hs:3:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T15323.hs:3:15-19 }))] + (NoEpTok) + (EpTok + (EpaSpan { T15323.hs:3:1-6 })) + (EpTok + (EpaSpan { T15323.hs:3:15-19 })) [] [] (Just ===================================== testsuite/tests/parser/should_compile/T20452.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T20452.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T20452.hs:3:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T20452.hs:3:15-19 }))] + (NoEpTok) + (EpTok + (EpaSpan { T20452.hs:3:1-6 })) + (EpTok + (EpaSpan { T20452.hs:3:15-19 })) [] [] (Just ===================================== testsuite/tests/parser/should_compile/T20718.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T20718.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T20718.hs:3:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T20718.hs:3:15-19 }))] + (NoEpTok) + (EpTok + (EpaSpan { T20718.hs:3:1-6 })) + (EpTok + (EpaSpan { T20718.hs:3:15-19 })) [] [] (Just ===================================== testsuite/tests/parser/should_compile/T20718b.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T20718b.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T20718b.hs:4:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T20718b.hs:4:16-20 }))] + (NoEpTok) + (EpTok + (EpaSpan { T20718b.hs:4:1-6 })) + (EpTok + (EpaSpan { T20718b.hs:4:16-20 })) [] [] (Just ===================================== testsuite/tests/parser/should_compile/T20846.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T20846.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T20846.hs:1:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T20846.hs:1:15-19 }))] + (NoEpTok) + (EpTok + (EpaSpan { T20846.hs:1:1-6 })) + (EpTok + (EpaSpan { T20846.hs:1:15-19 })) [] [] (Just ===================================== testsuite/tests/parser/should_compile/T23315/T23315.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T23315.hsig:1:1 }) (AnnsModule - [(AddEpAnn AnnSignature (EpaSpan { T23315.hsig:1:1-9 })) - ,(AddEpAnn AnnWhere (EpaSpan { T23315.hsig:1:18-22 }))] + (EpTok + (EpaSpan { T23315.hsig:1:1-9 })) + (NoEpTok) + (EpTok + (EpaSpan { T23315.hsig:1:18-22 })) [] [] (Nothing)) ===================================== testsuite/tests/printer/AnnotationNoListTuplePuns.stdout ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { AnnotationNoListTuplePuns.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { AnnotationNoListTuplePuns.hs:3:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { AnnotationNoListTuplePuns.hs:3:34-38 }))] + (NoEpTok) + (EpTok + (EpaSpan { AnnotationNoListTuplePuns.hs:3:1-6 })) + (EpTok + (EpaSpan { AnnotationNoListTuplePuns.hs:3:34-38 })) [] [] (Just @@ -188,8 +191,11 @@ (EpAnn (EpaSpan { AnnotationNoListTuplePuns.ppr.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { AnnotationNoListTuplePuns.ppr.hs:3:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { AnnotationNoListTuplePuns.ppr.hs:3:34-38 }))] + (NoEpTok) + (EpTok + (EpaSpan { AnnotationNoListTuplePuns.ppr.hs:3:1-6 })) + (EpTok + (EpaSpan { AnnotationNoListTuplePuns.ppr.hs:3:34-38 })) [] [] (Just ===================================== testsuite/tests/printer/T18791.stderr ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { T18791.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { T18791.hs:2:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { T18791.hs:2:15-19 }))] + (NoEpTok) + (EpTok + (EpaSpan { T18791.hs:2:1-6 })) + (EpTok + (EpaSpan { T18791.hs:2:15-19 })) [] [] (Just ===================================== testsuite/tests/printer/Test20297.stdout ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { Test20297.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { Test20297.hs:2:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { Test20297.hs:2:18-22 }))] + (NoEpTok) + (EpTok + (EpaSpan { Test20297.hs:2:1-6 })) + (EpTok + (EpaSpan { Test20297.hs:2:18-22 })) [] [] (Just @@ -420,8 +423,11 @@ (EpAnn (EpaSpan { Test20297.ppr.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { Test20297.ppr.hs:2:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { Test20297.ppr.hs:2:18-22 }))] + (NoEpTok) + (EpTok + (EpaSpan { Test20297.ppr.hs:2:1-6 })) + (EpTok + (EpaSpan { Test20297.ppr.hs:2:18-22 })) [] [] (Just ===================================== testsuite/tests/printer/Test24533.stdout ===================================== @@ -8,8 +8,11 @@ (EpAnn (EpaSpan { Test24533.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { Test24533.hs:2:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { Test24533.hs:2:18-22 }))] + (NoEpTok) + (EpTok + (EpaSpan { Test24533.hs:2:1-6 })) + (EpTok + (EpaSpan { Test24533.hs:2:18-22 })) [] [] (Just @@ -640,8 +643,11 @@ (EpAnn (EpaSpan { Test24533.ppr.hs:1:1 }) (AnnsModule - [(AddEpAnn AnnModule (EpaSpan { Test24533.ppr.hs:2:1-6 })) - ,(AddEpAnn AnnWhere (EpaSpan { Test24533.ppr.hs:2:18-22 }))] + (NoEpTok) + (EpTok + (EpaSpan { Test24533.ppr.hs:2:1-6 })) + (EpTok + (EpaSpan { Test24533.ppr.hs:2:18-22 })) [] [] (Just ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -1,4 +1,5 @@ {-# LANGUAGE BangPatterns #-} +{-# LANGUAGE DataKinds #-} {-# LANGUAGE DeriveDataTypeable #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE FlexibleInstances #-} @@ -996,15 +997,21 @@ lepa k epAnn = fmap (\newAnns -> epAnn { anns = newAnns }) -- data AnnsModule -- = AnnsModule { --- am_main :: [AddEpAnn], +-- am_sig :: Maybe (EpToken "signature"), +-- am_mod :: Maybe (EpToken "module"), +-- am_where :: Maybe (EpToken "where"), -- am_decls :: [TrailingAnn], -- am_cs :: [LEpaComment], -- am_eof :: Maybe (RealSrcSpan, RealSrcSpan) -- } deriving (Data, Eq) -lam_main :: Lens AnnsModule [AddEpAnn] -lam_main k annsModule = fmap (\newAnns -> annsModule { am_main = newAnns }) - (k (am_main annsModule)) +lam_mod :: Lens AnnsModule (EpToken "module") +lam_mod k annsModule = fmap (\newAnns -> annsModule { am_mod = newAnns }) + (k (am_mod annsModule)) + +lam_where :: Lens AnnsModule (EpToken "where") +lam_where k annsModule = fmap (\newAnns -> annsModule { am_where = newAnns }) + (k (am_where annsModule)) -- lam_decls :: Lens AnnsModule AnnList -- lam_decls k annsModule = fmap (\newAnns -> annsModule { am_decls = newAnns }) @@ -1366,6 +1373,12 @@ markLensKwM' a l kw = do go Nothing = return Nothing go (Just s) = Just <$> markKwA kw s +markLensTok :: (Monad m, Monoid w, KnownSymbol sym) + => EpAnn a -> Lens a (EpToken sym) -> EP w m (EpAnn a) +markLensTok (EpAnn anc a cs) l = do + new <- markEpToken (view l a) + return (EpAnn anc (set l new a) cs) + -- --------------------------------------------------------------------- markEpAnnL' :: (Monad m, Monoid w) @@ -1676,14 +1689,14 @@ instance ExactPrint (HsModule GhcPs) where case mmn of Nothing -> return (an, mmn, mdeprec, mexports) Just m -> do - an0 <- markEpAnnL' an lam_main AnnModule + an0 <- markLensTok an lam_mod m' <- markAnnotated m mdeprec' <- setLayoutTopLevelP $ markAnnotated mdeprec mexports' <- setLayoutTopLevelP $ markAnnotated mexports - an1 <- setLayoutTopLevelP $ markEpAnnL' an0 lam_main AnnWhere + an1 <- setLayoutTopLevelP $ markLensTok an0 lam_where return (an1, Just m', mdeprec', mexports') ===================================== utils/check-exact/Parsers.hs ===================================== @@ -326,8 +326,8 @@ fixModuleHeaderComments (GHC.L l p) = GHC.L l p' -> ([GHC.LHsDecl GHC.GhcPs], GHC.EpAnn GHC.AnnsModule) rebalance (ds, GHC.EpAnn a an cs) = (ds1, GHC.EpAnn a an cs') where - (ds1,cs') = case break (\(GHC.AddEpAnn k _) -> k == GHC.AnnWhere) (GHC.am_main an) of - (_, (GHC.AddEpAnn _ whereLoc:_)) -> + (ds1,cs') = case GHC.am_where an of + GHC.EpTok whereLoc -> case GHC.hsmodDecls p of (d:ds0) -> (d':ds0, cs0) where (d',cs0) = moveComments whereLoc d cs ===================================== utils/check-exact/Utils.hs ===================================== @@ -1,3 +1,4 @@ +{-# LANGUAGE DataKinds #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE LambdaCase #-} {-# LANGUAGE NamedFieldPuns #-} @@ -292,7 +293,7 @@ insertTopLevelCppComments (HsModule (XModulePs an lo mdeprec mbDoc) mmn mexports Just _ -> -- We have a module name. Capture all comments up to the `where` let - (these, remaining) = splitOnWhere Before (am_main $ anns an) cs + (these, remaining) = splitOnWhere Before (am_where $ anns an) cs (EpAnn a anno ocs) = an :: EpAnn AnnsModule anm = EpAnn a anno (workInComments ocs these) in @@ -336,9 +337,9 @@ insertTopLevelCppComments (HsModule (XModulePs an lo mdeprec mbDoc) mmn mexports -- Either hc0i or hc0d should have comments. Combine them hc0 = hc0i ++ hc0d - (hc1,hc_cs) = if null ( am_main $ anns an3) + (hc1,hc_cs) = if NoEpTok == (am_where $ anns an3) then (hc0,[]) - else splitOnWhere After (am_main $ anns an3) hc0 + else splitOnWhere After (am_where $ anns an3) hc0 hc2 = workInComments (comments an3) hc1 an4 = an3 { anns = (anns an3) {am_cs = hc_cs}, comments = hc2 } @@ -355,15 +356,14 @@ insertTopLevelCppComments (HsModule (XModulePs an lo mdeprec mbDoc) mmn mexports (xs',rest') = allocPreceding xs rest data SplitWhere = Before | After -splitOnWhere :: SplitWhere -> [AddEpAnn] -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) -splitOnWhere _ [] csIn = (csIn,[]) -splitOnWhere w (AddEpAnn AnnWhere (EpaSpan (RealSrcSpan s _)):_) csIn = (hc, fc) + +splitOnWhere :: SplitWhere -> EpToken "where" -> [LEpaComment] -> ([LEpaComment], [LEpaComment]) +splitOnWhere w (EpTok (EpaSpan (RealSrcSpan s _))) csIn = (hc, fc) where splitFunc Before anc_pos c_pos = c_pos < anc_pos splitFunc After anc_pos c_pos = anc_pos < c_pos (hc,fc) = break (\(L ll _) -> splitFunc w (ss2pos $ anchor ll) (ss2pos s)) csIn -splitOnWhere _ (AddEpAnn AnnWhere _:_) csIn = (csIn, []) -splitOnWhere f (_:as) csIn = splitOnWhere f as csIn +splitOnWhere _ _ csIn = (csIn,[]) balanceFirstLocatedAComments :: [LocatedA a] -> ([LocatedA a], [LEpaComment]) balanceFirstLocatedAComments [] = ([],[]) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c0df5aa9999fcb7a7890973c319b56731b77aa7d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c0df5aa9999fcb7a7890973c319b56731b77aa7d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 10:46:46 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Sun, 22 Sep 2024 06:46:46 -0400 Subject: [Git][ghc/ghc][wip/25232] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66eff596e7c61_14990a24204892756@gitlab.mail> Zubin pushed to branch wip/25232 at Glasgow Haskell Compiler / GHC Commits: 2eb67c6f by Zubin Duggal at 2024-09-22T16:16:32+05:30 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 5 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - + testsuite/tests/rts/T25232.hs - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,18 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + // Not doing this lead to #25232 and #23958 + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +958,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): ===================================== testsuite/tests/rts/T25232.hs ===================================== @@ -0,0 +1,14 @@ +module Main where + +import Control.Monad +import Control.Monad.Trans.State.Strict +import Data.IORef + +main :: IO () +main = do + ref <- newIORef () + replicateM_ 1000 $ withIORef ref $ runStateT (pure ()) + +withIORef :: IORef a -> (a -> IO (b, a)) -> IO b +withIORef ref f = + readIORef ref >>= f >>= \(b, a) -> writeIORef ref a >> pure b ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [unless(have_profiling(), skip), only_ways(['normal','nonmoving','nonmoving_prof','nonmoving_thr_prof']), extra_ways(['nonmoving', 'nonmoving_prof', 'nonmoving_thr_prof'])], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2eb67c6f0664c66f9b6f34d02de891bee6f172b9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2eb67c6f0664c66f9b6f34d02de891bee6f172b9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 10:49:43 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sun, 22 Sep 2024 06:49:43 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 5 commits: Fix C calls with SIMD vectors Message-ID: <66eff647d69c4_14990a23f050935fc@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 467ad749 by sheaf at 2024-09-22T12:49:15+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 3f671281 by sheaf at 2024-09-22T12:49:16+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - 03e75c4e by sheaf at 2024-09-22T12:49:16+02:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 58a02d51 by sheaf at 2024-09-22T12:49:16+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 9d0f68ea by sheaf at 2024-09-22T12:49:16+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 12 changed files: - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToAsm/X86/Instr.hs - compiler/GHC/CmmToAsm/X86/Ppr.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs - testsuite/tests/simd/should_run/simd013.hs - testsuite/tests/simd/should_run/simd013.stdout - testsuite/tests/simd/should_run/simd013C.c Changes: ===================================== compiler/GHC/CmmToAsm/Format.hs ===================================== @@ -28,6 +28,7 @@ module GHC.CmmToAsm.Format ( scalarWidth, formatInBytes, isFloatScalarFormat, + isFloatOrFloatVecFormat, floatScalarFormat, scalarFormatFormat, VirtualRegWithFormat(..), @@ -134,6 +135,11 @@ isFloatScalarFormat = \case FmtDouble -> True _ -> False +isFloatOrFloatVecFormat :: Format -> Bool +isFloatOrFloatVecFormat = \case + VecFormat _ sFmt -> isFloatScalarFormat sFmt + fmt -> isFloatFormat fmt + floatScalarFormat :: Width -> ScalarFormat floatScalarFormat W32 = FmtFloat floatScalarFormat W64 = FmtDouble ===================================== compiler/GHC/CmmToAsm/X86/CodeGen.hs ===================================== @@ -1,6 +1,9 @@ +{-# LANGUAGE DuplicateRecordFields #-} +{-# LANGUAGE GADTs #-} {-# LANGUAGE LambdaCase #-} {-# LANGUAGE MultiWayIf #-} -{-# LANGUAGE GADTs #-} +{-# LANGUAGE ParallelListComp #-} +{-# LANGUAGE TupleSections #-} {-# LANGUAGE NondecreasingIndentation #-} {-# OPTIONS_GHC -Wno-incomplete-uni-patterns #-} @@ -73,22 +76,29 @@ import GHC.Types.Tickish ( GenTickish(..) ) import GHC.Types.SrcLoc ( srcSpanFile, srcSpanStartLine, srcSpanStartCol ) -- The rest: +import GHC.Data.Maybe ( expectJust ) import GHC.Types.ForeignCall ( CCallConv(..) ) import GHC.Data.OrdList import GHC.Utils.Outputable import GHC.Utils.Constants (debugIsOn) +import GHC.Utils.Monad ( foldMapM ) import GHC.Utils.Panic import GHC.Data.FastString import GHC.Utils.Misc import GHC.Types.Unique.Supply ( getUniqueM ) +import qualified Data.Semigroup as S + import Control.Monad +import Control.Monad.Trans.State.Strict + ( StateT, evalStateT, get, put ) +import Control.Monad.Trans.Class (lift) import Data.Foldable (fold) import Data.Int import Data.Maybe import Data.Word -import qualified Data.Map as M +import qualified Data.Map as Map is32BitPlatform :: NatM Bool is32BitPlatform = do @@ -227,7 +237,7 @@ addSpUnwindings instr@(DELTA d) = do let platform = ncgPlatform config if ncgDwarfUnwindings config then do lbl <- mkAsmTempLabel <$> getUniqueM - let unwind = M.singleton MachSp (Just $ UwReg (GlobalRegUse MachSp (bWord platform)) $ negate d) + let unwind = Map.singleton MachSp (Just $ UwReg (GlobalRegUse MachSp (bWord platform)) $ negate d) return $ toOL [ instr, UNWIND lbl unwind ] else return (unitOL instr) addSpUnwindings instr = return $ unitOL instr @@ -328,10 +338,10 @@ stmtToInstrs bid stmt = do CmmUnwind regs -> do let to_unwind_entry :: (GlobalReg, Maybe CmmExpr) -> UnwindTable - to_unwind_entry (reg, expr) = M.singleton reg (fmap (toUnwindExpr platform) expr) + to_unwind_entry (reg, expr) = Map.singleton reg (fmap (toUnwindExpr platform) expr) case foldMap to_unwind_entry regs of - tbl | M.null tbl -> return nilOL - | otherwise -> do + tbl | Map.null tbl -> return nilOL + | otherwise -> do lbl <- mkAsmTempLabel <$> getUniqueM return $ unitOL $ UNWIND lbl tbl @@ -836,12 +846,12 @@ iselExpr64ParallelBin op e1 e2 = do -- targetted for any particular type like Int8, Int32 etc data VectorArithInstns = VA_Add | VA_Sub | VA_Mul | VA_Div | VA_Min | VA_Max -getRegister :: CmmExpr -> NatM Register +getRegister :: HasDebugCallStack => CmmExpr -> NatM Register getRegister e = do platform <- getPlatform is32Bit <- is32BitPlatform getRegister' platform is32Bit e -getRegister' :: Platform -> Bool -> CmmExpr -> NatM Register +getRegister' :: HasDebugCallStack => Platform -> Bool -> CmmExpr -> NatM Register getRegister' platform is32Bit (CmmReg reg) = case reg of @@ -903,20 +913,6 @@ getRegister' _ is32Bit (CmmMachOp (MO_UU_Conv W64 W16) [x]) ro <- getNewRegNat II16 return $ Fixed II16 ro (code `appOL` toOL [ MOVZxL II16 (OpReg rlo) (OpReg ro) ]) -getRegister' _ _ (CmmLit lit@(CmmFloat f w)) = - float_const_sse2 where - float_const_sse2 - | f == 0.0 = do - -- TODO: this mishandles negative zero floating point literals. - let - format = floatFormat w - code dst = unitOL (XOR format (OpReg dst) (OpReg dst)) - -- I don't know why there are xorpd, xorps, and pxor instructions. - -- They all appear to do the same thing --SDM - return (Any format code) - - | otherwise = getFloatLitRegister lit - -- catch simple cases of zero- or sign-extended load getRegister' _ _ (CmmMachOp (MO_UU_Conv W8 W32) [CmmLoad addr _ _]) = do code <- intLoadCode (MOVZxL II8) addr @@ -1922,7 +1918,7 @@ getRegister' platform is32Bit load@(CmmLoad mem ty _) | isFloatType ty = do Amode addr mem_code <- getAmode mem - loadFloatAmode width addr mem_code + loadAmode (floatFormat width) addr mem_code | is32Bit && not (isWord64 ty) = do @@ -1950,20 +1946,6 @@ getRegister' platform is32Bit load@(CmmLoad mem ty _) format = cmmTypeFormat ty width = typeWidth ty -getRegister' _ is32Bit (CmmLit (CmmInt 0 width)) - = let - format = intFormat width - - -- x86_64: 32-bit xor is one byte shorter, and zero-extends to 64 bits - format1 = if is32Bit then format - else case format of - II64 -> II32 - _ -> format - code dst - = unitOL (XOR format1 (OpReg dst) (OpReg dst)) - in - return (Any format code) - -- Handle symbol references with LEA and %rip-relative addressing. -- See Note [%rip-relative addressing on x86-64]. getRegister' platform is32Bit (CmmLit lit) @@ -1980,80 +1962,102 @@ getRegister' platform is32Bit (CmmLit lit) is_label (CmmLabelDiffOff {}) = True is_label _ = False - -- optimisation for loading small literals on x86_64: take advantage - -- of the automatic zero-extension from 32 to 64 bits, because the 32-bit - -- instruction forms are shorter. -getRegister' platform is32Bit (CmmLit lit) - | not is32Bit, isWord64 (cmmLitType platform lit), not (isBigLit lit) - = let - imm = litToImm lit - code dst = unitOL (MOV II32 (OpImm imm) (OpReg dst)) - in - return (Any II64 code) - where - isBigLit (CmmInt i _) = i < 0 || i > 0xffffffff - isBigLit _ = False +getRegister' platform is32Bit (CmmLit lit) = do + avx <- avxEnabled + + -- NB: it is important that the code produced here (to load a literal into + -- a register) doesn't clobber any registers other than the destination + -- register; the code for generating C calls relies on this property. + -- + -- In particular, we have: + -- + -- > loadIntoRegMightClobberOtherReg (CmmLit _) = False + -- + -- which means that we assume that loading a literal into a register + -- will not clobber any other registers. + + -- TODO: this function mishandles floating-point negative zero, + -- because -0.0 == 0.0 returns True and because we represent CmmFloat as + -- Rational, which can't properly represent negative zero. + + if + -- Zero: use XOR. + | isZeroLit lit + -> let code dst + | isIntFormat fmt + = let fmt' + | is32Bit + = fmt + | otherwise + -- x86_64: 32-bit xor is one byte shorter, + -- and zero-extends to 64 bits + = case fmt of + II64 -> II32 + _ -> fmt + in unitOL (XOR fmt' (OpReg dst) (OpReg dst)) + | avx + = if float_or_floatvec + then unitOL (VXOR fmt (OpReg dst) dst dst) + else unitOL (VPXOR fmt dst dst dst) + | otherwise + = if float_or_floatvec + then unitOL (XOR fmt (OpReg dst) (OpReg dst)) + else unitOL (PXOR fmt (OpReg dst) dst) + in return $ Any fmt code + + -- Constant vector: use broadcast. + | VecFormat l sFmt <- fmt + , CmmVec (f:fs) <- lit + , all (== f) fs + -> do let w = scalarWidth sFmt + broadcast = if isFloatScalarFormat sFmt + then MO_VF_Broadcast l w + else MO_V_Broadcast l w + valCode <- getAnyReg (CmmMachOp broadcast [CmmLit f]) + return $ Any fmt valCode + + -- Optimisation for loading small literals on x86_64: take advantage + -- of the automatic zero-extension from 32 to 64 bits, because the 32-bit + -- instruction forms are shorter. + | not is32Bit, isWord64 cmmTy, not (isBigLit lit) + -> let + imm = litToImm lit + code dst = unitOL (MOV II32 (OpImm imm) (OpReg dst)) + in + return (Any II64 code) + + -- Scalar integer: use an immediate. + | isIntFormat fmt + -> let imm = litToImm lit + code dst = unitOL (MOV fmt (OpImm imm) (OpReg dst)) + in return (Any fmt code) + + -- General case: load literal from data address. + | otherwise + -> do let w = formatToWidth fmt + Amode addr addr_code <- memConstant (mkAlignment $ widthInBytes w) lit + loadAmode fmt addr addr_code + + where + cmmTy = cmmLitType platform lit + fmt = cmmTypeFormat cmmTy + float_or_floatvec = isFloatOrFloatVecFormat fmt + isZeroLit (CmmInt i _) = i == 0 + isZeroLit (CmmFloat f _) = f == 0 -- TODO: mishandles negative zero + isZeroLit (CmmVec fs) = all isZeroLit fs + isZeroLit _ = False + + isBigLit (CmmInt i _) = i < 0 || i > 0xffffffff + isBigLit _ = False -- note1: not the same as (not.is32BitLit), because that checks for -- signed literals that fit in 32 bits, but we want unsigned -- literals here. -- note2: all labels are small, because we're assuming the -- small memory model. See Note [%rip-relative addressing on x86-64]. -getRegister' platform _ (CmmLit lit) = do - avx <- avxEnabled - case fmt of - VecFormat l sFmt - | CmmVec fs <- lit - , all is_zero fs - -> let code dst - | avx - = if isFloatScalarFormat sFmt - then unitOL (VXOR fmt (OpReg dst) dst dst) - else unitOL (VPXOR fmt dst dst dst) - | otherwise - = unitOL (XOR fmt (OpReg dst) (OpReg dst)) - in return (Any fmt code) - | CmmVec (f:fs) <- lit - , all (== f) fs - -- TODO: mishandles negative zero (because -0.0 == 0.0 returns True), and because we - -- represent CmmFloat as Rational which can't properly represent negative zero. - -> do let w = scalarWidth sFmt - broadcast = if isFloatScalarFormat sFmt - then MO_VF_Broadcast l w - else MO_V_Broadcast l w - valCode <- getAnyReg (CmmMachOp broadcast [CmmLit f]) - return $ Any fmt valCode - - | otherwise - -> do - let w = formatToWidth fmt - config <- getConfig - Amode addr addr_code <- memConstant (mkAlignment $ widthInBytes w) lit - let code dst = addr_code `snocOL` - movInstr config fmt (OpAddr addr) (OpReg dst) - return (Any fmt code) - where - is_zero (CmmInt i _) = i == 0 - is_zero (CmmFloat f _) = f == 0 -- TODO: mishandles negative zero - is_zero _ = False - - _ -> let imm = litToImm lit - code dst = unitOL (MOV fmt (OpImm imm) (OpReg dst)) - in return (Any fmt code) - where - cmmTy = cmmLitType platform lit - fmt = cmmTypeFormat cmmTy - getRegister' platform _ slot@(CmmStackSlot {}) = pprPanic "getRegister(x86) CmmStackSlot" (pdoc platform slot) -getFloatLitRegister :: CmmLit -> NatM Register -getFloatLitRegister lit = do - let w :: Width - w = case lit of { CmmInt _ w -> w; CmmFloat _ w -> w; _ -> panic "getFloatLitRegister" (ppr lit) } - Amode addr code <- memConstant (mkAlignment $ widthInBytes w) lit - loadFloatAmode w addr code - intLoadCode :: (Operand -> Operand -> Instr) -> CmmExpr -> NatM (Reg -> InstrBlock) intLoadCode instr mem = do @@ -2264,7 +2268,7 @@ getNonClobberedOperand (CmmLit lit) = return (OpAddr addr, code) else do platform <- getPlatform - if is32BitLit platform lit && not (isFloatType (cmmLitType platform lit)) + if is32BitLit platform lit && isIntFormat (cmmTypeFormat (cmmLitType platform lit)) then return (OpImm (litToImm lit), nilOL) else getNonClobberedOperand_generic (CmmLit lit) @@ -2321,13 +2325,13 @@ getOperand (CmmLit lit) = do else do platform <- getPlatform - if is32BitLit platform lit && not (isFloatType (cmmLitType platform lit)) + if is32BitLit platform lit && (isIntFormat $ cmmTypeFormat (cmmLitType platform lit)) then return (OpImm (litToImm lit), nilOL) else getOperand_generic (CmmLit lit) getOperand (CmmLoad mem ty _) = do is32Bit <- is32BitPlatform - if not (isFloatType ty) && (if is32Bit then not (isWord64 ty) else True) + if isIntFormat (cmmTypeFormat ty) && (if is32Bit then not (isWord64 ty) else True) then do Amode src mem_code <- getAmode mem return (OpAddr src, mem_code) @@ -2358,7 +2362,7 @@ addAlignmentCheck align reg = where check :: Format -> Reg -> InstrBlock check fmt reg = - assert (not $ isFloatFormat fmt) $ + assert (isIntFormat fmt) $ toOL [ TEST fmt (OpImm $ ImmInt $ align-1) (OpReg reg) , JXX_GBL NE $ ImmCLbl mkBadAlignmentLabel ] @@ -2382,15 +2386,12 @@ memConstant align lit = do `consOL` addr_code return (Amode addr code) - -loadFloatAmode :: Width -> AddrMode -> InstrBlock -> NatM Register -loadFloatAmode w addr addr_code = do - let format = floatFormat w - code dst = addr_code `snocOL` - MOV format (OpAddr addr) (OpReg dst) - - return (Any format code) - +-- | Load the value at the given address into any register. +loadAmode :: Format -> AddrMode -> InstrBlock -> NatM Register +loadAmode fmt addr addr_code = do + config <- getConfig + let load dst = movInstr config fmt (OpAddr addr) (OpReg dst) + return $ Any fmt (\ dst -> addr_code `snocOL` load dst) -- if we want a floating-point literal as an operand, we can -- use it directly from memory. However, if the literal is @@ -2403,7 +2404,7 @@ isSuitableFloatingPointLit _ = False getRegOrMem :: CmmExpr -> NatM (Operand, InstrBlock) getRegOrMem e@(CmmLoad mem ty _) = do is32Bit <- is32BitPlatform - if not (isFloatType ty) && (if is32Bit then not (isWord64 ty) else True) + if isIntFormat (cmmTypeFormat ty) && (if is32Bit then not (isWord64 ty) else True) then do Amode src mem_code <- getAmode mem return (OpAddr src, mem_code) @@ -3090,10 +3091,8 @@ genSimplePrim _ op dst args = do platform <- ncgPlatform <$> getConfig pprPanic "genSimplePrim: unhandled primop" (ppr (pprCallishMachOp op, dst, fmap (pdoc platform) args)) -{- -Note [Evaluate C-call arguments before placing in destination registers] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - +{- Note [Evaluate C-call arguments before placing in destination registers] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ When producing code for C calls we must take care when placing arguments in their final registers. Specifically, we must ensure that temporary register usage due to evaluation of one argument does not clobber a register in which we @@ -3144,15 +3143,11 @@ genForeignCall{32,64}. -- | See Note [Evaluate C-call arguments before placing in destination registers] evalArgs :: BlockId -> [CmmActual] -> NatM (InstrBlock, [CmmActual]) evalArgs bid actuals - | any mightContainMachOp actuals = do + | any loadIntoRegMightClobberOtherReg actuals = do regs_blks <- mapM evalArg actuals return (concatOL $ map fst regs_blks, map snd regs_blks) | otherwise = return (nilOL, actuals) where - mightContainMachOp (CmmReg _) = False - mightContainMachOp (CmmRegOff _ _) = False - mightContainMachOp (CmmLit _) = False - mightContainMachOp _ = True evalArg :: CmmActual -> NatM (InstrBlock, CmmExpr) evalArg actual = do @@ -3166,6 +3161,16 @@ evalArgs bid actuals newLocalReg :: CmmType -> NatM LocalReg newLocalReg ty = LocalReg <$> getUniqueM <*> pure ty +-- | Might the code to put this expression into a register +-- clobber any other registers? +loadIntoRegMightClobberOtherReg :: CmmExpr -> Bool +loadIntoRegMightClobberOtherReg (CmmReg _) = False +loadIntoRegMightClobberOtherReg (CmmRegOff _ _) = False +loadIntoRegMightClobberOtherReg (CmmLit _) = False + -- NB: this last 'False' is slightly risky, because the code for loading + -- a literal into a register is not entirely trivial. +loadIntoRegMightClobberOtherReg _ = True + -- Note [DIV/IDIV for bytes] -- ~~~~~~~~~~~~~~~~~~~~~~~~~ -- IDIV reminder: @@ -3238,24 +3243,39 @@ genCCall -> [CmmFormal] -> [CmmActual] -> NatM InstrBlock -genCCall bid addr conv dest_regs args = do +genCCall bid addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do + platform <- getPlatform is32Bit <- is32BitPlatform - (instrs0, args') <- evalArgs bid args + let args_hints = zip args (argHints ++ repeat NoHint) + prom_args = map (maybePromoteCArgToW32 platform) args_hints + (instrs0, args') <- evalArgs bid prom_args instrs1 <- if is32Bit then genCCall32 addr conv dest_regs args' else genCCall64 addr conv dest_regs args' return (instrs0 `appOL` instrs1) +maybePromoteCArgToW32 :: Platform -> (CmmExpr, ForeignHint) -> CmmExpr +maybePromoteCArgToW32 platform (arg, hint) + | wfrom < wto = + -- As wto=W32, we only need to handle integer conversions, + -- never Float -> Double. + case hint of + SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] + _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] + | otherwise = arg + where + ty = cmmExprType platform arg + wfrom = typeWidth ty + wto = W32 + genCCall32 :: CmmExpr -- ^ address of the function to call -> ForeignConvention -- ^ calling convention -> [CmmFormal] -- ^ where to put the result -> [CmmActual] -- ^ arguments (of mixed type) -> NatM InstrBlock -genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do +genCCall32 addr _conv dest_regs args = do config <- getConfig let platform = ncgPlatform config - args_hints = zip args (argHints ++ repeat NoHint) - prom_args = map (maybePromoteCArg platform W32) args_hints -- If the size is smaller than the word, we widen things (see maybePromoteCArg) arg_size_bytes :: CmmType -> Int @@ -3278,7 +3298,7 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do DELTA (delta-8)] ) - | isFloatType arg_ty = do + | isFloatType arg_ty || isVecType arg_ty = do (reg, code) <- getSomeReg arg delta <- getDeltaNat setDeltaNat (delta-size) @@ -3288,11 +3308,10 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do let addr = AddrBaseIndex (EABaseReg esp) EAIndexNone (ImmInt 0) - format = floatFormat (typeWidth arg_ty) + format = cmmTypeFormat arg_ty in - -- assume SSE2 - MOV format (OpReg reg) (OpAddr addr) + movInstr config format (OpReg reg) (OpAddr addr) ] ) @@ -3316,7 +3335,7 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do let -- Align stack to 16n for calls, assuming a starting stack -- alignment of 16n - word_size on procedure entry. Which we - -- maintiain. See Note [Stack Alignment on X86] in rts/StgCRun.c. + -- maintain. See Note [Stack Alignment on X86] in rts/StgCRun.c. sizes = map (arg_size_bytes . cmmExprType platform) (reverse args) raw_arg_size = sum sizes + platformWordSizeInBytes platform arg_pad_size = (roundTo 16 $ raw_arg_size) - raw_arg_size @@ -3326,7 +3345,7 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do delta0 <- getDeltaNat setDeltaNat (delta0 - arg_pad_size) - push_codes <- mapM push_arg (reverse prom_args) + push_codes <- mapM push_arg (reverse args) delta <- getDeltaNat massert (delta == delta0 - tot_arg_size) @@ -3361,6 +3380,8 @@ genCCall32 addr (ForeignConvention _ argHints _ _) dest_regs args = do -- assign the results, if necessary assign_code [] = nilOL assign_code [dest] + | isVecType ty + = unitOL (mkRegRegMoveInstr config (cmmTypeFormat ty) xmm0 r_dest) | isFloatType ty = -- we assume SSE2 let tmp_amode = AddrBaseIndex (EABaseReg esp) @@ -3399,192 +3420,71 @@ genCCall64 :: CmmExpr -- ^ address of function to call -> [CmmFormal] -- ^ where to put the result -> [CmmActual] -- ^ arguments (of mixed type) -> NatM InstrBlock -genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do +genCCall64 addr conv dest_regs args = do config <- getConfig let platform = ncgPlatform config - -- load up the register arguments - let args_hints = zip args (argHints ++ repeat NoHint) - let prom_args = map (maybePromoteCArg platform W32) args_hints - - let load_args :: [CmmExpr] - -> [RegWithFormat] -- int regs avail for args - -> [RegWithFormat] -- FP regs avail for args - -> InstrBlock -- code computing args - -> InstrBlock -- code assigning args to ABI regs - -> NatM ([CmmExpr],[RegWithFormat],[RegWithFormat],InstrBlock,InstrBlock) - -- no more regs to use - load_args args [] [] code acode = - return (args, [], [], code, acode) - - -- no more args to push - load_args [] aregs fregs code acode = - return ([], aregs, fregs, code, acode) - - load_args (arg : rest) aregs fregs code acode - | isFloatType arg_rep = case fregs of - [] -> push_this_arg - (RegWithFormat r _fmt:rs) -> do - (code',acode') <- reg_this_arg r - load_args rest aregs rs code' acode' - | otherwise = case aregs of - [] -> push_this_arg - (RegWithFormat r _fmt:rs) -> do - (code',acode') <- reg_this_arg r - load_args rest rs fregs code' acode' - where + word_size = platformWordSizeInBytes platform + wordFmt = archWordFormat (target32Bit platform) - -- put arg into the list of stack pushed args - push_this_arg = do - (args',ars,frs,code',acode') - <- load_args rest aregs fregs code acode - return (arg:args', ars, frs, code', acode') - - -- pass the arg into the given register - reg_this_arg r - -- "operand" args can be directly assigned into r - | isOperand platform arg = do - arg_code <- getAnyReg arg - return (code, (acode `appOL` arg_code r)) - -- The last non-operand arg can be directly assigned after its - -- computation without going into a temporary register - | all (isOperand platform) rest = do - arg_code <- getAnyReg arg - return (code `appOL` arg_code r,acode) - - -- other args need to be computed beforehand to avoid clobbering - -- previously assigned registers used to pass parameters (see - -- #11792, #12614). They are assigned into temporary registers - -- and get assigned to proper call ABI registers after they all - -- have been computed. - | otherwise = do - arg_code <- getAnyReg arg - tmp <- getNewRegNat arg_fmt - let - code' = code `appOL` arg_code tmp - acode' = acode `snocOL` mkRegRegMoveInstr config arg_fmt tmp r - return (code',acode') - - arg_rep = cmmExprType platform arg - arg_fmt = cmmTypeFormat arg_rep - - load_args_win :: [CmmExpr] - -> [RegWithFormat] -- used int regs - -> [RegWithFormat] -- used FP regs - -> [(Reg, Reg)] -- (int, FP) regs avail for args - -> InstrBlock - -> NatM ([CmmExpr],[RegWithFormat],[RegWithFormat],InstrBlock,InstrBlock) - load_args_win args usedInt usedFP [] code - = return (args, usedInt, usedFP, code, nilOL) - -- no more regs to use - load_args_win [] usedInt usedFP _ code - = return ([], usedInt, usedFP, code, nilOL) - -- no more args to push - load_args_win (arg : rest) usedInt usedFP - ((ireg, freg) : regs) code - | isFloatType arg_rep = do - arg_code <- getAnyReg arg - load_args_win rest (RegWithFormat ireg II64: usedInt) (RegWithFormat freg FF64 : usedFP) regs - (code `appOL` - arg_code freg `snocOL` - -- If we are calling a varargs function - -- then we need to define ireg as well - -- as freg - MOVD FF64 (OpReg freg) (OpReg ireg)) - | otherwise = do - arg_code <- getAnyReg arg - load_args_win rest (RegWithFormat ireg II64: usedInt) usedFP regs - (code `appOL` arg_code ireg) - where - arg_rep = cmmExprType platform arg - - arg_size = 8 -- always, at the mo - - push_args [] code = return code - push_args (arg:rest) code - | isFloatType arg_rep = do - (arg_reg, arg_code) <- getSomeReg arg - delta <- getDeltaNat - setDeltaNat (delta-arg_size) - let fmt = floatFormat width - code' = code `appOL` arg_code `appOL` toOL [ - SUB (intFormat (wordWidth platform)) (OpImm (ImmInt arg_size)) (OpReg rsp), - DELTA (delta-arg_size), - MOV fmt (OpReg arg_reg) (OpAddr (spRel platform 0))] - push_args rest code' - - | otherwise = do - -- Arguments can be smaller than 64-bit, but we still use @PUSH - -- II64@ - the usual calling conventions expect integers to be - -- 8-byte aligned. - massert (width <= W64) - (arg_op, arg_code) <- getOperand arg - delta <- getDeltaNat - setDeltaNat (delta-arg_size) - let code' = code `appOL` arg_code `appOL` toOL [ - PUSH II64 arg_op, - DELTA (delta-arg_size)] - push_args rest code' - where - arg_rep = cmmExprType platform arg - width = typeWidth arg_rep - - leaveStackSpace n = do - delta <- getDeltaNat - setDeltaNat (delta - n * arg_size) - return $ toOL [ - SUB II64 (OpImm (ImmInt (n * platformWordSizeInBytes platform))) (OpReg rsp), - DELTA (delta - n * arg_size)] - - (stack_args, int_regs_used, fp_regs_used, load_args_code, assign_args_code) - <- - if platformOS platform == OSMinGW32 - then load_args_win prom_args [] [] (allArgRegs platform) nilOL - else do - let intArgRegs = map (\r -> RegWithFormat r II64) $ allIntArgRegs platform - fpArgRegs = map (\r -> RegWithFormat r FF64) $ allFPArgRegs platform - (stack_args, aregs, fregs, load_args_code, assign_args_code) - <- load_args prom_args intArgRegs fpArgRegs nilOL nilOL - let used_regs rs as = dropTail (length rs) as - fregs_used = used_regs fregs fpArgRegs - aregs_used = used_regs aregs intArgRegs - return (stack_args, aregs_used, fregs_used, load_args_code - , assign_args_code) + -- Compute the code for loading arguments into registers, + -- returning the leftover arguments that will need to be passed on the stack. + -- + -- NB: the code for loading references to data into registers is computed + -- later (in 'pushArgs'), because we don't yet know where the data will be + -- placed (due to alignment requirements). + LoadArgs + { stackArgs = proper_stack_args + , stackDataArgs = stack_data_args + , usedRegs = arg_regs_used + , assignArgsCode = assign_args_code + } + <- loadArgs config args let - wordFmt = archWordFormat (target32Bit platform) - arg_regs_used = int_regs_used ++ fp_regs_used - arg_regs = [RegWithFormat eax wordFmt] ++ arg_regs_used - -- for annotating the call instruction with - sse_regs = length fp_regs_used - arg_stack_slots = if platformOS platform == OSMinGW32 - then length stack_args + length (allArgRegs platform) - else length stack_args - tot_arg_size = arg_size * arg_stack_slots + -- Pad all arguments and data passed on stack to align them properly. + (stk_args_with_padding, args_aligned_16) = + padStackArgs platform (proper_stack_args, stack_data_args) -- Align stack to 16n for calls, assuming a starting stack -- alignment of 16n - word_size on procedure entry. Which we -- maintain. See Note [Stack Alignment on X86] in rts/StgCRun.c - let word_size = platformWordSizeInBytes platform - (real_size, adjust_rsp) <- - if (tot_arg_size + word_size) `rem` 16 == 0 - then return (tot_arg_size, nilOL) - else do -- we need to adjust... - delta <- getDeltaNat - setDeltaNat (delta - word_size) - return (tot_arg_size + word_size, toOL [ - SUB II64 (OpImm (ImmInt word_size)) (OpReg rsp), - DELTA (delta - word_size) ]) - - -- push the stack args, right to left - push_code <- push_args (reverse stack_args) nilOL - -- On Win64, we also have to leave stack space for the arguments - -- that we are passing in registers - lss_code <- if platformOS platform == OSMinGW32 - then leaveStackSpace (length (allArgRegs platform)) - else return nilOL + need_realign_call = args_aligned_16 + align_call_code <- + if need_realign_call + then addStackPadding word_size + else return nilOL + + -- Compute the code that pushes data to the stack, and also + -- the code that loads references to that data into registers, + -- when the data is passed by reference in a register. + (load_data_refs, push_code) <- + pushArgs config proper_stack_args stk_args_with_padding + + -- On Windows, leave stack space for the arguments that we are passing + -- in registers (the so-called shadow space). + let shadow_space = + if platformOS platform == OSMinGW32 + then 8 * length (allArgRegs platform) + -- NB: the shadow store is always 8 * 4 = 32 bytes large, + -- i.e. the cumulative size of rcx, rdx, r8, r9 (see 'allArgRegs'). + else 0 + shadow_space_code <- addStackPadding shadow_space + + let total_args_size + = shadow_space + + sum (map (stackArgSpace platform) stk_args_with_padding) + real_size = + total_args_size + if need_realign_call then word_size else 0 + + -- End of argument passing. + -- + -- Next step: emit the appropriate call instruction. delta <- getDeltaNat + let arg_regs = [RegWithFormat eax wordFmt] ++ arg_regs_used + -- for annotating the call instruction with + -- deal with static vs dynamic call targets (callinsns,_cconv) <- case addr of CmmLit (CmmLabel lbl) -> @@ -3594,7 +3494,7 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do return (dyn_c `snocOL` CALL (Right dyn_r) arg_regs, conv) let - -- The x86_64 ABI requires us to set %al to the number of SSE2 + -- The System V AMD64 ABI requires us to set %al to the number of SSE2 -- registers that contain arguments, if the called routine -- is a varargs function. We don't know whether it's a -- varargs function or not, so we have to assume it is. @@ -3602,7 +3502,12 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do -- It's not safe to omit this assignment, even if the number -- of SSE2 regs in use is zero. If %al is larger than 8 -- on entry to a varargs function, seg faults ensue. - assign_eax n = unitOL (MOV II32 (OpImm (ImmInt n)) (OpReg eax)) + nb_sse_regs_used = count (isFloatFormat . regWithFormat_format) arg_regs_used + assign_eax_sse_regs + | platformOS platform == OSMinGW32 + = nilOL + | otherwise + = unitOL (MOV II32 (OpImm (ImmInt nb_sse_regs_used)) (OpReg eax)) let call = callinsns `appOL` toOL ( @@ -3619,7 +3524,7 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do assign_code [] = nilOL assign_code [dest] = unitOL $ - movInstr config fmt (OpReg reg) (OpReg r_dest) + mkRegRegMoveInstr config fmt reg r_dest where reg = if isIntFormat fmt then rax else xmm0 fmt = cmmTypeFormat rep @@ -3627,24 +3532,594 @@ genCCall64 addr conv@(ForeignConvention _ argHints _ _) dest_regs args = do r_dest = getRegisterReg platform (CmmLocal dest) assign_code _many = panic "genForeignCall.assign_code many" - return (adjust_rsp `appOL` + return (align_call_code `appOL` push_code `appOL` - load_args_code `appOL` assign_args_code `appOL` - lss_code `appOL` - assign_eax sse_regs `appOL` + load_data_refs `appOL` + shadow_space_code `appOL` + assign_eax_sse_regs `appOL` call `appOL` assign_code dest_regs) +-- ----------------------------------------------------------------------------- +-- Loading arguments into registers for 64-bit C calls. + +-- | Information needed to know how to pass arguments in a C call, +-- and in particular how to load them into registers. +data LoadArgs + = LoadArgs + -- | Arguments that should be passed on the stack + { stackArgs :: [RawStackArg] + -- | Additional values to store onto the stack. + , stackDataArgs :: [CmmExpr] + -- | Which registers are we using for argument passing? + , usedRegs :: [RegWithFormat] + -- | The code to assign arguments to registers used for argument passing. + , assignArgsCode :: InstrBlock + } +instance Semigroup LoadArgs where + LoadArgs a1 d1 r1 j1 <> LoadArgs a2 d2 r2 j2 + = LoadArgs (a1 ++ a2) (d1 ++ d2) (r1 ++ r2) (j1 S.<> j2) +instance Monoid LoadArgs where + mempty = LoadArgs [] [] [] nilOL + +-- | An argument passed on the stack, either directly or by reference. +-- +-- The padding information hasn't yet been computed (see 'StackArg'). +data RawStackArg + -- | Pass the argument on the stack directly. + = RawStackArg { stackArgExpr :: CmmExpr } + -- | Pass the argument by reference. + | RawStackArgRef + { stackRef :: StackRef + -- ^ is the reference passed in a register, or on the stack? + , stackRefArgSize :: Int + -- ^ the size of the data pointed to + } + deriving ( Show ) + +-- | An argument passed on the stack, either directly or by reference, +-- with additional padding information. +data StackArg + -- | Pass the argument on the stack directly. + = StackArg + { stackArgExpr :: CmmExpr + , stackArgPadding :: Int + -- ^ padding required (in bytes) + } + -- | Pass the argument by reference. + | StackArgRef + { stackRef :: StackRef + -- ^ where the reference is passed + , stackRefArgSize :: Int + -- ^ the size of the data pointed to + , stackRefArgPadding :: Int + -- ^ padding of the data pointed to + -- (the reference itself never requires padding) + } + deriving ( Show ) + +-- | Where is a reference to data on the stack passed? +data StackRef + -- | In a register. + = InReg Reg + -- | On the stack. + | OnStack + deriving ( Eq, Ord, Show ) + +newtype Padding = Padding { paddingBytes :: Int } + deriving ( Show, Eq, Ord ) + +-- | How much space does this 'StackArg' take up on the stack? +-- +-- Only counts the "reference" part for references, not the data it points to. +stackArgSpace :: Platform -> StackArg -> Int +stackArgSpace platform = \case + StackArg arg padding -> + argSize platform arg + padding + StackArgRef { stackRef = ref } -> + case ref of + InReg {} -> 0 + OnStack {} -> 8 + +-- | Pad arguments, assuming we start aligned to a 16-byte boundary. +-- +-- Returns padded arguments, together with whether we end up aligned +-- to a 16-byte boundary. +padStackArgs :: Platform + -> ([RawStackArg], [CmmExpr]) + -> ([StackArg], Bool) +padStackArgs platform (args0, data_args0) = + let + -- Pad the direct args + (args, align_16_mid) = pad_args True args0 + + -- Pad the data section + (data_args, align_16_end) = pad_args align_16_mid (map RawStackArg data_args0) + + -- Now figure out where the data is placed relative to the direct arguments, + -- in order to resolve references. + resolve_args :: [(RawStackArg, Padding)] -> [Padding] -> [StackArg] + resolve_args [] _ = [] + resolve_args ((stk_arg, Padding pad):rest) pads = + let (this_arg, pads') = + case stk_arg of + RawStackArg arg -> (StackArg arg pad, pads) + RawStackArgRef ref size -> + let (Padding arg_pad : rest_pads) = pads + arg = + StackArgRef + { stackRef = ref + , stackRefArgSize = size + , stackRefArgPadding = arg_pad } + in (arg, rest_pads) + in this_arg : resolve_args rest pads' + + in + ( resolve_args args (fmap snd data_args) ++ + [ case data_arg of + RawStackArg arg -> StackArg arg pad + RawStackArgRef {} -> panic "padStackArgs: reference in data section" + | (data_arg, Padding pad) <- data_args + ] + , align_16_end ) -maybePromoteCArg :: Platform -> Width -> (CmmExpr, ForeignHint) -> CmmExpr -maybePromoteCArg platform wto (arg, hint) - | wfrom < wto = case hint of - SignedHint -> CmmMachOp (MO_SS_Conv wfrom wto) [arg] - _ -> CmmMachOp (MO_UU_Conv wfrom wto) [arg] - | otherwise = arg - where - wfrom = cmmExprWidth platform arg + where + pad_args :: Bool -> [RawStackArg] -> ([(RawStackArg, Padding)], Bool) + pad_args aligned_16 [] = ([], aligned_16) + pad_args aligned_16 (arg:args) + | needed_alignment > 16 + -- We don't know if the stack is aligned to 8 (mod 32) or 24 (mod 32). + -- This makes aligning the stack to a 32 or 64 byte boundary more + -- complicated, in particular with DELTA. + = sorry $ unlines + [ "X86_86 C call: unsupported argument." + , " Alignment requirement: " ++ show needed_alignment ++ " bytes." + , if platformOS platform == OSMinGW32 + then " The X86_64 NCG does not (yet) support Windows C calls with 256/512 bit vectors." + else " The X86_64 NCG cannot (yet) pass 256/512 bit vectors on the stack for C calls." + , " Please use the LLVM backend (-fllvm)." ] + | otherwise + = let ( rest, final_align_16 ) = pad_args next_aligned_16 args + in ( (arg, Padding padding) : rest, final_align_16 ) + + where + needed_alignment = case arg of + RawStackArg arg -> argSize platform arg + RawStackArgRef {} -> platformWordSizeInBytes platform + padding + | needed_alignment < 16 || aligned_16 + = 0 + | otherwise + = 8 + next_aligned_16 = not ( aligned_16 && needed_alignment < 16 ) + +-- | Load arguments into available registers. +loadArgs :: NCGConfig -> [CmmExpr] -> NatM LoadArgs +loadArgs config args + | platformOS platform == OSMinGW32 + = evalStateT (loadArgsWin config args) (allArgRegs platform) + | otherwise + = evalStateT (loadArgsSysV config args) (allIntArgRegs platform + ,allFPArgRegs platform) + where + platform = ncgPlatform config + +-- | Load arguments into available registers (System V AMD64 ABI). +loadArgsSysV :: NCGConfig + -> [CmmExpr] + -> StateT ([Reg], [Reg]) NatM LoadArgs +loadArgsSysV _ [] = return mempty +loadArgsSysV config (arg:rest) = do + (iregs, fregs) <- get + -- No available registers: pass everything on the stack (shortcut). + if null iregs && null fregs + then return $ + LoadArgs + { stackArgs = map RawStackArg (arg:rest) + , stackDataArgs = [] + , assignArgsCode = nilOL + , usedRegs = [] + } + else do + mbReg <- + if + | isIntFormat arg_fmt + , ireg:iregs' <- iregs + -> do put (iregs', fregs) + return $ Just ireg + | isFloatFormat arg_fmt || isVecFormat arg_fmt + , freg:fregs' <- fregs + -> do put (iregs, fregs') + return $ Just freg + | otherwise + -> return Nothing + this_arg <- + case mbReg of + Just reg -> do + assign_code <- lift $ loadArgIntoReg arg reg + return $ + LoadArgs + { stackArgs = [] -- passed in register + , stackDataArgs = [] + , assignArgsCode = assign_code + , usedRegs = [RegWithFormat reg arg_fmt] + } + Nothing -> do + return $ + -- No available register for this argument: pass it on the stack. + LoadArgs + { stackArgs = [RawStackArg arg] + , stackDataArgs = [] + , assignArgsCode = nilOL + , usedRegs = [] + } + others <- loadArgsSysV config rest + return $ this_arg S.<> others + + where + platform = ncgPlatform config + arg_fmt = cmmTypeFormat (cmmExprType platform arg) + +-- | Compute all things that will need to be pushed to the stack. +-- +-- On Windows, an argument passed by reference will require two pieces of data: +-- +-- - the reference (returned in the first position) +-- - the actual data (returned in the second position) +computeWinPushArgs :: Platform -> [CmmExpr] -> ([RawStackArg], [CmmExpr]) +computeWinPushArgs platform = go + where + go :: [CmmExpr] -> ([RawStackArg], [CmmExpr]) + go [] = ([], []) + go (arg:args) = + let + arg_size = argSize platform arg + (this_arg, add_this_arg) + | arg_size > 8 + = ( RawStackArgRef OnStack arg_size, (arg :) ) + | otherwise + = ( RawStackArg arg, id ) + (stk_args, stk_data) = go args + in + (this_arg:stk_args, add_this_arg stk_data) + +-- | Load arguments into available registers (Windows C X64 calling convention). +loadArgsWin :: NCGConfig -> [CmmExpr] -> StateT [(Reg,Reg)] NatM LoadArgs +loadArgsWin _ [] = return mempty +loadArgsWin config (arg:rest) = do + regs <- get + case regs of + reg:regs' -> do + put regs' + this_arg <- lift $ load_arg_win reg + rest <- loadArgsWin config rest + return $ this_arg S.<> rest + [] -> do + -- No more registers available: pass all (remaining) arguments on the stack. + let (stk_args, data_args) = computeWinPushArgs platform (arg:rest) + return $ + LoadArgs + { stackArgs = stk_args + , stackDataArgs = data_args + , assignArgsCode = nilOL + , usedRegs = [] + } + where + platform = ncgPlatform config + arg_fmt = cmmTypeFormat $ cmmExprType platform arg + load_arg_win (ireg, freg) + | isVecFormat arg_fmt + -- Vectors are passed by reference. + -- See Note [The Windows X64 C calling convention]. + = do return $ + LoadArgs + -- Pass the reference in a register, + -- and the argument data on the stack. + { stackArgs = [RawStackArgRef (InReg ireg) (argSize platform arg)] + , stackDataArgs = [arg] -- we don't yet know where the data will reside, + , assignArgsCode = nilOL -- so we defer computing the reference and storing it + -- in the register until later + , usedRegs = [RegWithFormat ireg II64] + } + | otherwise + = do let arg_reg + | isFloatFormat arg_fmt + = freg + | otherwise + = ireg + assign_code <- loadArgIntoReg arg arg_reg + -- Recall that, for varargs, we must pass floating-point + -- arguments in both fp and integer registers. + let (assign_code', regs') + | isFloatFormat arg_fmt = + ( assign_code `snocOL` MOVD FF64 (OpReg freg) (OpReg ireg), + [ RegWithFormat freg FF64 + , RegWithFormat ireg II64 ]) + | otherwise = (assign_code, [RegWithFormat ireg II64]) + return $ + LoadArgs + { stackArgs = [] -- passed in register + , stackDataArgs = [] + , assignArgsCode = assign_code' + , usedRegs = regs' + } + +-- | Load an argument into a register. +-- +-- Assumes that the expression does not contain any MachOps, +-- as per Note [Evaluate C-call arguments before placing in destination registers]. +loadArgIntoReg :: CmmExpr -> Reg -> NatM InstrBlock +loadArgIntoReg arg reg = do + when (debugIsOn && loadIntoRegMightClobberOtherReg arg) $ do + platform <- getPlatform + massertPpr False $ + vcat [ text "loadArgIntoReg: arg might contain MachOp" + , text "arg:" <+> pdoc platform arg ] + arg_code <- getAnyReg arg + return $ arg_code reg + +-- ----------------------------------------------------------------------------- +-- Pushing arguments onto the stack for 64-bit C calls. + +-- | The size of an argument (in bytes). +-- +-- Never smaller than the platform word width. +argSize :: Platform -> CmmExpr -> Int +argSize platform arg = + max (platformWordSizeInBytes platform) $ + widthInBytes (typeWidth $ cmmExprType platform arg) + +-- | Add the given amount of padding on the stack. +addStackPadding :: Int -- ^ padding (in bytes) + -> NatM InstrBlock +addStackPadding pad_bytes + | pad_bytes == 0 + = return nilOL + | otherwise + = do delta <- getDeltaNat + setDeltaNat (delta - pad_bytes) + return $ + toOL [ SUB II64 (OpImm (ImmInt pad_bytes)) (OpReg rsp) + , DELTA (delta - pad_bytes) + ] + +-- | Push one argument directly to the stack (by value). +-- +-- Assumes the current stack pointer fulfills any necessary alignment requirements. +pushArgByValue :: NCGConfig -> CmmExpr -> NatM InstrBlock +pushArgByValue config arg + -- For 64-bit integer arguments, use PUSH II64. + -- + -- Note: we *must not* do this for smaller arguments. + -- For example, if we tried to push an argument such as @CmmLoad addr W32 aln@, + -- we could end up reading unmapped memory and segfaulting. + | isIntFormat fmt + , formatInBytes fmt == 8 + = do + (arg_op, arg_code) <- getOperand arg + delta <- getDeltaNat + setDeltaNat (delta-arg_size) + return $ + arg_code `appOL` toOL + [ PUSH II64 arg_op + , DELTA (delta-arg_size) ] + + | otherwise + = do + (arg_reg, arg_code) <- getSomeReg arg + delta <- getDeltaNat + setDeltaNat (delta-arg_size) + return $ arg_code `appOL` toOL + [ SUB (intFormat (wordWidth platform)) (OpImm (ImmInt arg_size)) (OpReg rsp) + , DELTA (delta-arg_size) + , movInstr config fmt (OpReg arg_reg) (OpAddr (spRel platform 0)) ] + + where + platform = ncgPlatform config + arg_size = argSize platform arg + arg_rep = cmmExprType platform arg + fmt = cmmTypeFormat arg_rep + +-- | Load an argument into a register or push it to the stack. +loadOrPushArg :: NCGConfig -> (StackArg, Maybe Int) -> NatM (InstrBlock, InstrBlock) +loadOrPushArg config (stk_arg, mb_off) = + case stk_arg of + StackArg arg pad -> do + push_code <- pushArgByValue config arg + pad_code <- addStackPadding pad + return (nilOL, push_code `appOL` pad_code) + StackArgRef { stackRef = ref } -> + case ref of + -- Pass the reference in a register + InReg ireg -> + return (unitOL $ LEA II64 (OpAddr (spRel platform off)) (OpReg ireg), nilOL) + -- Pass the reference on the stack + OnStack {} -> do + tmp <- getNewRegNat II64 + delta <- getDeltaNat + setDeltaNat (delta-arg_ref_size) + let push_code = toOL + [ SUB (intFormat (wordWidth platform)) (OpImm (ImmInt arg_ref_size)) (OpReg rsp) + , DELTA (delta-arg_ref_size) + , LEA II64 (OpAddr (spRel platform off)) (OpReg tmp) + , MOV II64 (OpReg tmp) (OpAddr (spRel platform 0)) ] + return (nilOL, push_code) + where off = expectJust "push_arg_win offset" mb_off + where + arg_ref_size = 8 -- passing a reference to the argument + platform = ncgPlatform config + +-- | Push arguments to the stack, right to left. +-- +-- On Windows, some arguments may need to be passed by reference, +-- which requires separately passing the data and the reference. +-- See Note [The Windows X64 C calling convention]. +pushArgs :: NCGConfig + -> [RawStackArg] + -- ^ arguments proper (i.e. don't include the data for arguments passed by reference) + -> [StackArg] + -- ^ arguments we are passing on the stack + -> NatM (InstrBlock, InstrBlock) +pushArgs config proper_args all_stk_args + = do { let + vec_offs :: [Maybe Int] + vec_offs + | platformOS platform == OSMinGW32 + = go stack_arg_size all_stk_args + | otherwise + = repeat Nothing + + --------------------- + -- Windows-only code + + -- Size of the arguments we are passing on the stack, counting only + -- the reference part for arguments passed by reference. + stack_arg_size = 8 * count not_in_reg proper_args + not_in_reg (RawStackArg {}) = True + not_in_reg (RawStackArgRef { stackRef = ref }) = + case ref of + InReg {} -> False + OnStack {} -> True + + -- Check an offset is valid (8-byte aligned), for assertions. + ok off = off `rem` 8 == 0 + + -- Tricky code: compute the stack offset to the vector data + -- for this argument. + -- + -- If you're confused, Note [The Windows X64 C calling convention] + -- contains a helpful diagram. + go :: Int -> [StackArg] -> [Maybe Int] + go _ [] = [] + go off (stk_arg:args) = + assertPpr (ok off) (text "unaligned offset:" <+> ppr off) $ + case stk_arg of + StackArg {} -> + -- Only account for the stack pointer movement. + let off' = off - stackArgSpace platform stk_arg + in Nothing : go off' args + StackArgRef + { stackRefArgSize = data_size + , stackRefArgPadding = data_pad } -> + assertPpr (ok data_size) (text "unaligned data size:" <+> ppr data_size) $ + assertPpr (ok data_pad) (text "unaligned data padding:" <+> ppr data_pad) $ + let off' = off + -- Next piece of data is after the data for this reference + + data_size + data_pad + -- ... and account for the stack pointer movement. + - stackArgSpace platform stk_arg + in Just (data_pad + off) : go off' args + + -- end of Windows-only code + ---------------------------- + + -- Push the stack arguments (right to left), + -- including both the reference and the data for arguments passed by reference. + ; (load_regs, push_args) <- foldMapM (loadOrPushArg config) (reverse $ zip all_stk_args vec_offs) + ; return (load_regs, push_args) } + where + platform = ncgPlatform config + +{- Note [The Windows X64 C calling convention] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Here are a few facts about the Windows X64 C calling convention that +are important: + + - any argument larger than 8 bytes must be passed by reference, + and arguments smaller than 8 bytes are padded to 8 bytes. + + - the first four arguments are passed in registers: + - floating-point scalar arguments are passed in %xmm0, %xmm1, %xmm2, %xmm3 + - other arguments are passed in %rcx, %rdx, %r8, %r9 + (this includes vector arguments, passed by reference) + + For variadic functions, it is additionally expected that floating point + scalar arguments are copied to the corresponding integer register, e.g. + the data in xmm2 should also be copied to r8. + + There is no requirement about setting %al like there is for the + System V AMD64 ABI. + + - subsequent arguments are passed on the stack. + +There are also alignment requirements: + + - the data for vectors must be aligned to the size of the vector, + e.g. a 32 byte vector must be aligned on a 32 byte boundary, + + - the call instruction must be aligned to 16 bytes. + (This differs from the System V AMD64 ABI, which mandates that the call + instruction must be aligned to 32 bytes if there are any 32 byte vectors + passed on the stack.) + +This motivates our handling of vector values. Suppose we have a function call +with many arguments, several of them being vectors. We proceed as follows: + + - Add some padding, if necessary, to ensure the stack, when executing the call + instruction, is 16-byte aligned. Whether this padding is necessary depends + on what happens next. (Recall also that we start off at 8 (mod 16) alignment, + as per Note [Stack Alignment on X86] in rts/StgCRun.c) + - Push all the vectors to the stack first, adding padding after each one + if necessary. + - Then push the arguments: + - for non-vectors, proceed as usual, + - for vectors, push the address of the vector data we pushed above. + - Then assign the registers: + - for non-vectors, proceed as usual, + - for vectors, store the address in a general-purpose register, as opposed + to storing the data in an xmm register. + +For a concrete example, suppose we have a call of the form: + + f x1 x2 x3 x4 x5 x6 x7 + +in which: + + - x2, x3, x5 and x7 are 16 byte vectors + - the other arguments are all 8 byte wide + +Now, x1, x2, x3, x4 will get passed in registers, except that we pass +x2 and x3 by reference, because they are vectors. We proceed as follows: + + - push the vectors to the stack: x7, x5, x3, x2 (in that order) + - push the stack arguments in order: addr(x7), x6, addr(x5) + - load the remaining arguments into registers: x4, addr(x3), addr(x2), x1 + +The tricky part is to get the right offsets for the addresses of the vector +data. The following visualisation will hopefully clear things up: + + ┌──┐ + │▓▓│ ─── padding to align the call instruction + ╭─╴ ╞══╡ (ensures Sp, below, is 16-byte aligned) + │ │ │ + │ x7 ───╴ │ │ + │ ├──┤ + │ │ │ + │ x5 ───╴ │ │ + │ ├──┤ + vector data ────┤ │ │ +(individually padded) │ x3 ───╴ │ │ + │ ├──┤ + │ │ │ + │ x2 ───╴ │ │ + │ ├┄┄┤ + │ │▓▓│ ─── padding to align x2 to 16 bytes + ╭─╴ ╰─╴ ╞══╡ + │ addr(x7) ───╴ │ │ ╭─ from here: x7 is +64 + │ ├──┤ ╾──╯ = 64 (position of x5) + stack ───┤ x6 ───╴ │ │ + 16 (size of x5) + 0 (padding of x7) + arguments │ ├──┤ - 2 * 8 (x7 is 2 arguments higher than x5) + │ addr(x5) ───╴ │ │ + ╰─╴ ╭─╴ ╞══╡ ╾─── from here: + │ │ │ - x2 is +32 = 24 (stack_arg_size) + 8 (padding of x2) + shadow ───┤ │ │ - x3 is +48 = 32 (position of x2) + 16 (size of x2) + 0 (padding of x3) + space │ │ │ - x5 is +64 = 48 (position of x3) + 16 (size of x3) + 0 (padding of x5) + │ │ │ + ╰─╴ └──┘ ╾─── Sp + +This is all tested in the simd013 test. +-} -- ----------------------------------------------------------------------------- -- Generating a table-branch ===================================== compiler/GHC/CmmToAsm/X86/Instr.hs ===================================== @@ -307,6 +307,7 @@ data Instr | VMOVDQU Format Operand Operand -- logic operations + | PXOR Format Operand Reg | VPXOR Format Reg Reg Reg -- Arithmetic @@ -493,6 +494,12 @@ regUsageOfInstr platform instr MOVDQU fmt src dst -> mkRU (use_R fmt src []) (use_R fmt dst []) VMOVDQU fmt src dst -> mkRU (use_R fmt src []) (use_R fmt dst []) + PXOR fmt (OpReg src) dst + | src == dst + -> mkRU [] [mk fmt dst] + | otherwise + -> mkRU [mk fmt src, mk fmt dst] [mk fmt dst] + VPXOR fmt s1 s2 dst | s1 == s2, s1 == dst -> mkRU [] [mk fmt dst] @@ -734,6 +741,7 @@ patchRegsOfInstr platform instr env MOVDQU fmt src dst -> MOVDQU fmt (patchOp src) (patchOp dst) VMOVDQU fmt src dst -> VMOVDQU fmt (patchOp src) (patchOp dst) + PXOR fmt src dst -> PXOR fmt (patchOp src) (env dst) VPXOR fmt s1 s2 dst -> VPXOR fmt (env s1) (env s2) (env dst) VADD fmt s1 s2 dst -> VADD fmt (patchOp s1) (env s2) (env dst) @@ -887,7 +895,7 @@ mkLoadInstr config (RegWithFormat reg fmt) delta slot = -- | A move instruction for moving the entire contents of an operand -- at the given 'Format'. -movInstr :: NCGConfig -> Format -> (Operand -> Operand -> Instr) +movInstr :: HasDebugCallStack => NCGConfig -> Format -> (Operand -> Operand -> Instr) movInstr config fmt = case fmt of VecFormat _ sFmt -> ===================================== compiler/GHC/CmmToAsm/X86/Ppr.hs ===================================== @@ -1016,6 +1016,8 @@ pprInstr platform i = case i of VecFormat 64 FmtInt8 -> text "vmovdqu32" -- require the additional AVX512BW extension _ -> text "vmovdqu" + PXOR format src dst + -> pprPXor (text "pxor") format src dst VPXOR format s1 s2 dst -> pprXor (text "vpxor") format s1 s2 dst VEXTRACT format offset from to @@ -1320,6 +1322,15 @@ pprInstr platform i = case i of pprReg platform format reg3 ] + pprPXor :: Line doc -> Format -> Operand -> Reg -> doc + pprPXor name format src dst + = line $ hcat [ + pprGenMnemonic name format, + pprOperand platform format src, + comma, + pprReg platform format dst + ] + pprVxor :: Format -> Operand -> Reg -> Reg -> doc pprVxor fmt src1 src2 dst = line $ hcat [ ===================================== compiler/GHC/CmmToLlvm.hs ===================================== @@ -139,7 +139,7 @@ llvmGroupLlvmGens cmm = do Nothing -> l Just (CmmStaticsRaw info_lbl _) -> info_lbl lml <- strCLabel_llvm l' - funInsert lml =<< llvmFunTy (map globalRegUse_reg live) + funInsert lml =<< llvmFunTy live return Nothing cdata <- fmap catMaybes $ mapM split cmm ===================================== compiler/GHC/CmmToLlvm/Base.hs ===================================== @@ -12,7 +12,7 @@ module GHC.CmmToLlvm.Base ( LlvmCmmDecl, LlvmBasicBlock, - LiveGlobalRegs, + LiveGlobalRegs, LiveGlobalRegUses, LlvmUnresData, LlvmData, UnresLabel, UnresStatic, LlvmM, @@ -29,6 +29,8 @@ module GHC.CmmToLlvm.Base ( llvmFunSig, llvmFunArgs, llvmStdFunAttrs, llvmFunAlign, llvmInfAlign, llvmPtrBits, tysToParams, llvmFunSection, padLiveArgs, isFPR, + lookupRegUse, + strCLabel_llvm, getGlobalPtr, generateExternDecls, @@ -58,9 +60,11 @@ import GHC.Types.Unique.Set import GHC.Types.Unique.Supply import GHC.Utils.Logger -import Data.Maybe (fromJust) import Control.Monad.Trans.State (StateT (..)) -import Data.List (isPrefixOf) +import Control.Applicative (Alternative((<|>))) +import Data.Maybe (fromJust, mapMaybe) + +import Data.List (find, isPrefixOf) import qualified Data.List.NonEmpty as NE import Data.Ord (comparing) @@ -73,6 +77,7 @@ type LlvmBasicBlock = GenBasicBlock LlvmStatement -- | Global registers live on proc entry type LiveGlobalRegs = [GlobalReg] +type LiveGlobalRegUses = [GlobalRegUse] -- | Unresolved code. -- Of the form: (data label, data type, unresolved data) @@ -116,16 +121,16 @@ llvmGhcCC platform | otherwise = CC_Ghc -- | Llvm Function type for Cmm function -llvmFunTy :: LiveGlobalRegs -> LlvmM LlvmType +llvmFunTy :: LiveGlobalRegUses -> LlvmM LlvmType llvmFunTy live = return . LMFunction =<< llvmFunSig' live (fsLit "a") ExternallyVisible -- | Llvm Function signature -llvmFunSig :: LiveGlobalRegs -> CLabel -> LlvmLinkageType -> LlvmM LlvmFunctionDecl +llvmFunSig :: LiveGlobalRegUses -> CLabel -> LlvmLinkageType -> LlvmM LlvmFunctionDecl llvmFunSig live lbl link = do lbl' <- strCLabel_llvm lbl llvmFunSig' live lbl' link -llvmFunSig' :: LiveGlobalRegs -> LMString -> LlvmLinkageType -> LlvmM LlvmFunctionDecl +llvmFunSig' :: LiveGlobalRegUses -> LMString -> LlvmLinkageType -> LlvmM LlvmFunctionDecl llvmFunSig' live lbl link = do let toParams x | isPointer x = (x, [NoAlias, NoCapture]) | otherwise = (x, []) @@ -149,16 +154,25 @@ llvmFunSection opts lbl | otherwise = Nothing -- | A Function's arguments -llvmFunArgs :: Platform -> LiveGlobalRegs -> [LlvmVar] +llvmFunArgs :: Platform -> LiveGlobalRegUses -> [LlvmVar] llvmFunArgs platform live = - map (lmGlobalRegArg platform) (filter isPassed allRegs) + map (lmGlobalRegArg platform) (mapMaybe isPassed allRegs) where allRegs = activeStgRegs platform paddingRegs = padLiveArgs platform live - isLive r = r `elem` alwaysLive - || r `elem` live - || r `elem` paddingRegs - isPassed r = not (isFPR r) || isLive r - + isLive :: GlobalReg -> Maybe GlobalRegUse + isLive r = + lookupRegUse r (alwaysLive platform) + <|> + lookupRegUse r live + <|> + lookupRegUse r paddingRegs + isPassed r = + if not (isFPR r) + then Just $ GlobalRegUse r (globalRegSpillType platform r) + else isLive r + +lookupRegUse :: GlobalReg -> [GlobalRegUse] -> Maybe GlobalRegUse +lookupRegUse r = find ((== r) . globalRegUse_reg) isFPR :: GlobalReg -> Bool isFPR (FloatReg _) = True @@ -179,7 +193,7 @@ isFPR _ = False -- Invariant: Cmm FPR regs with number "n" maps to real registers with number -- "n" If the calling convention uses registers in a different order or if the -- invariant doesn't hold, this code probably won't be correct. -padLiveArgs :: Platform -> LiveGlobalRegs -> LiveGlobalRegs +padLiveArgs :: Platform -> LiveGlobalRegUses -> LiveGlobalRegUses padLiveArgs platform live = if platformUnregisterised platform then [] -- not using GHC's register convention for platform. @@ -188,7 +202,7 @@ padLiveArgs platform live = ---------------------------------- -- handle floating-point registers (FPR) - fprLive = filter isFPR live -- real live FPR registers + fprLive = filter (isFPR . globalRegUse_reg) live -- real live FPR registers -- we group live registers sharing the same classes, i.e. that use the same -- set of real registers to be passed. E.g. FloatReg, DoubleReg and XmmReg @@ -196,39 +210,44 @@ padLiveArgs platform live = -- classes = NE.groupBy sharesClass fprLive sharesClass a b = globalRegsOverlap platform (norm a) (norm b) -- check if mapped to overlapping registers - norm x = fpr_ctor x 1 -- get the first register of the family + norm x = globalRegUse_reg (fpr_ctor x 1) -- get the first register of the family -- For each class, we just have to fill missing registers numbers. We use -- the constructor of the greatest register to build padding registers. -- -- E.g. sortedRs = [ F2, XMM4, D5] -- output = [D1, D3] + padded :: [GlobalRegUse] padded = concatMap padClass classes + + padClass :: NE.NonEmpty GlobalRegUse -> [GlobalRegUse] padClass rs = go (NE.toList sortedRs) 1 where - sortedRs = NE.sortBy (comparing fpr_num) rs + sortedRs = NE.sortBy (comparing (fpr_num . globalRegUse_reg)) rs maxr = NE.last sortedRs ctor = fpr_ctor maxr go [] _ = [] - go (c1:c2:_) _ -- detect bogus case (see #17920) + go (GlobalRegUse c1 _: GlobalRegUse c2 _:_) _ -- detect bogus case (see #17920) | fpr_num c1 == fpr_num c2 , Just real <- globalRegMaybe platform c1 = sorryDoc "LLVM code generator" $ text "Found two different Cmm registers (" <> ppr c1 <> text "," <> ppr c2 <> text ") both alive AND mapped to the same real register: " <> ppr real <> text ". This isn't currently supported by the LLVM backend." - go (c:cs) f - | fpr_num c == f = go cs f -- already covered by a real register - | otherwise = ctor f : go (c:cs) (f + 1) -- add padding register - - fpr_ctor :: GlobalReg -> Int -> GlobalReg - fpr_ctor (FloatReg _) = FloatReg - fpr_ctor (DoubleReg _) = DoubleReg - fpr_ctor (XmmReg _) = XmmReg - fpr_ctor (YmmReg _) = YmmReg - fpr_ctor (ZmmReg _) = ZmmReg - fpr_ctor _ = error "fpr_ctor expected only FPR regs" + go (cu@(GlobalRegUse c _):cs) f + | fpr_num c == f = go cs f -- already covered by a real register + | otherwise = ctor f : go (cu:cs) (f + 1) -- add padding register + + fpr_ctor :: GlobalRegUse -> Int -> GlobalRegUse + fpr_ctor (GlobalRegUse r fmt) i = + case r of + FloatReg _ -> GlobalRegUse (FloatReg i) fmt + DoubleReg _ -> GlobalRegUse (DoubleReg i) fmt + XmmReg _ -> GlobalRegUse (XmmReg i) fmt + YmmReg _ -> GlobalRegUse (YmmReg i) fmt + ZmmReg _ -> GlobalRegUse (ZmmReg i) fmt + _ -> error "fpr_ctor expected only FPR regs" fpr_num :: GlobalReg -> Int fpr_num (FloatReg i) = i ===================================== compiler/GHC/CmmToLlvm/CodeGen.hs ===================================== @@ -37,13 +37,14 @@ import GHC.Utils.Outputable import qualified GHC.Utils.Panic as Panic import GHC.Utils.Misc +import Control.Applicative (Alternative((<|>))) import Control.Monad.Trans.Class import Control.Monad.Trans.Writer import Control.Monad import qualified Data.Semigroup as Semigroup import Data.List ( nub ) -import Data.Maybe ( catMaybes ) +import Data.Maybe ( catMaybes, isJust ) type Atomic = Maybe MemoryOrdering type LlvmStatements = OrdList LlvmStatement @@ -57,7 +58,7 @@ genLlvmProc :: RawCmmDecl -> LlvmM [LlvmCmmDecl] genLlvmProc (CmmProc infos lbl live graph) = do let blocks = toBlockListEntryFirstFalseFallthrough graph - (lmblocks, lmdata) <- basicBlocksCodeGen (map globalRegUse_reg live) blocks + (lmblocks, lmdata) <- basicBlocksCodeGen live blocks let info = mapLookup (g_entry graph) infos proc = CmmProc info lbl live (ListGraph lmblocks) return (proc:lmdata) @@ -76,7 +77,7 @@ newtype UnreachableBlockId = UnreachableBlockId BlockId -- | Generate code for a list of blocks that make up a complete -- procedure. The first block in the list is expected to be the entry -- point. -basicBlocksCodeGen :: LiveGlobalRegs -> [CmmBlock] +basicBlocksCodeGen :: LiveGlobalRegUses -> [CmmBlock] -> LlvmM ([LlvmBasicBlock], [LlvmCmmDecl]) basicBlocksCodeGen _ [] = panic "no entry block!" basicBlocksCodeGen live cmmBlocks @@ -152,7 +153,7 @@ stmtToInstrs ubid stmt = case stmt of -- Tail call CmmCall { cml_target = arg, - cml_args_regs = live } -> genJump arg $ map globalRegUse_reg live + cml_args_regs = live } -> genJump arg live _ -> panic "Llvm.CodeGen.stmtToInstrs" @@ -1050,7 +1051,7 @@ cmmPrimOpFunctions mop = do -- | Tail function calls -genJump :: CmmExpr -> [GlobalReg] -> LlvmM StmtData +genJump :: CmmExpr -> LiveGlobalRegUses -> LlvmM StmtData -- Call to known function genJump (CmmLit (CmmLabel lbl)) live = do @@ -2056,14 +2057,13 @@ getCmmReg (CmmLocal (LocalReg un _)) -- have been assigned a value at some point, triggering -- "funPrologue" to allocate it on the stack. -getCmmReg (CmmGlobal g) - = do let r = globalRegUse_reg g - onStack <- checkStackReg r +getCmmReg (CmmGlobal ru@(GlobalRegUse r _)) + = do onStack <- checkStackReg r platform <- getPlatform if onStack - then return (lmGlobalRegVar platform r) + then return (lmGlobalRegVar platform ru) else pprPanic "getCmmReg: Cmm register " $ - ppr g <> text " not stack-allocated!" + ppr r <> text " not stack-allocated!" -- | Return the value of a given register, as well as its type. Might -- need to be load from stack. @@ -2074,7 +2074,7 @@ getCmmRegVal reg = onStack <- checkStackReg (globalRegUse_reg g) platform <- getPlatform if onStack then loadFromStack else do - let r = lmGlobalRegArg platform (globalRegUse_reg g) + let r = lmGlobalRegArg platform g return (r, getVarType r, nilOL) _ -> loadFromStack where loadFromStack = do @@ -2187,8 +2187,9 @@ convertMemoryOrdering MemOrderSeqCst = SyncSeqCst -- question is never written. Therefore we skip it where we can to -- save a few lines in the output and hopefully speed compilation up a -- bit. -funPrologue :: LiveGlobalRegs -> [CmmBlock] -> LlvmM StmtData +funPrologue :: LiveGlobalRegUses -> [CmmBlock] -> LlvmM StmtData funPrologue live cmmBlocks = do + platform <- getPlatform let getAssignedRegs :: CmmNode O O -> [CmmReg] getAssignedRegs (CmmAssign reg _) = [reg] @@ -2196,7 +2197,8 @@ funPrologue live cmmBlocks = do getAssignedRegs _ = [] getRegsBlock (_, body, _) = concatMap getAssignedRegs $ blockToList body assignedRegs = nub $ concatMap (getRegsBlock . blockSplit) cmmBlocks - isLive r = r `elem` alwaysLive || r `elem` live + mbLive r = + lookupRegUse r (alwaysLive platform) <|> lookupRegUse r live platform <- getPlatform stmtss <- forM assignedRegs $ \reg -> @@ -2205,12 +2207,12 @@ funPrologue live cmmBlocks = do let (newv, stmts) = allocReg reg varInsert un (pLower $ getVarType newv) return stmts - CmmGlobal (GlobalRegUse r _) -> do - let reg = lmGlobalRegVar platform r - arg = lmGlobalRegArg platform r + CmmGlobal ru@(GlobalRegUse r _) -> do + let reg = lmGlobalRegVar platform ru + arg = lmGlobalRegArg platform ru ty = (pLower . getVarType) reg trash = LMLitVar $ LMUndefLit ty - rval = if isLive r then arg else trash + rval = if isJust (mbLive r) then arg else trash alloc = Assignment reg $ Alloca (pLower $ getVarType reg) 1 markStackReg r return $ toOL [alloc, Store rval reg Nothing []] @@ -2222,7 +2224,7 @@ funPrologue live cmmBlocks = do -- | Function epilogue. Load STG variables to use as argument for call. -- STG Liveness optimisation done here. -funEpilogue :: LiveGlobalRegs -> LlvmM ([LlvmVar], LlvmStatements) +funEpilogue :: LiveGlobalRegUses -> LlvmM ([LlvmVar], LlvmStatements) funEpilogue live = do platform <- getPlatform @@ -2248,12 +2250,16 @@ funEpilogue live = do let allRegs = activeStgRegs platform loads <- forM allRegs $ \r -> if -- load live registers - | r `elem` alwaysLive -> loadExpr (GlobalRegUse r (globalRegSpillType platform r)) - | r `elem` live -> loadExpr (GlobalRegUse r (globalRegSpillType platform r)) + | Just ru <- lookupRegUse r (alwaysLive platform) + -> loadExpr ru + | Just ru <- lookupRegUse r live + -> loadExpr ru -- load all non Floating-Point Registers - | not (isFPR r) -> loadUndef r + | not (isFPR r) + -> loadUndef (GlobalRegUse r (globalRegSpillType platform r)) -- load padding Floating-Point Registers - | r `elem` paddingRegs -> loadUndef r + | Just ru <- lookupRegUse r paddingRegs + -> loadUndef ru | otherwise -> return (Nothing, nilOL) let (vars, stmts) = unzip loads @@ -2263,7 +2269,7 @@ funEpilogue live = do -- -- This is for Haskell functions, function type is assumed, so doesn't work -- with foreign functions. -getHsFunc :: LiveGlobalRegs -> CLabel -> LlvmM ExprData +getHsFunc :: LiveGlobalRegUses -> CLabel -> LlvmM ExprData getHsFunc live lbl = do fty <- llvmFunTy live name <- strCLabel_llvm lbl ===================================== compiler/GHC/CmmToLlvm/Ppr.hs ===================================== @@ -49,9 +49,8 @@ pprLlvmCmmDecl (CmmData _ lmdata) = do return ( vcat $ map (pprLlvmData opts) lmdata , vcat $ map (pprLlvmData opts) lmdata) -pprLlvmCmmDecl (CmmProc mb_info entry_lbl liveWithUses (ListGraph blks)) - = do let live = map globalRegUse_reg liveWithUses - lbl = case mb_info of +pprLlvmCmmDecl (CmmProc mb_info entry_lbl live (ListGraph blks)) + = do let lbl = case mb_info of Nothing -> entry_lbl Just (CmmStaticsRaw info_lbl _) -> info_lbl link = if externallyVisibleCLabel lbl ===================================== compiler/GHC/CmmToLlvm/Regs.hs ===================================== @@ -14,25 +14,27 @@ import GHC.Prelude import GHC.Llvm import GHC.Cmm.Expr +import GHC.CmmToAsm.Format import GHC.Platform import GHC.Data.FastString import GHC.Utils.Panic ( panic ) import GHC.Types.Unique + -- | Get the LlvmVar function variable storing the real register -lmGlobalRegVar :: Platform -> GlobalReg -> LlvmVar +lmGlobalRegVar :: Platform -> GlobalRegUse -> LlvmVar lmGlobalRegVar platform = pVarLift . lmGlobalReg platform "_Var" -- | Get the LlvmVar function argument storing the real register -lmGlobalRegArg :: Platform -> GlobalReg -> LlvmVar +lmGlobalRegArg :: Platform -> GlobalRegUse -> LlvmVar lmGlobalRegArg platform = lmGlobalReg platform "_Arg" {- Need to make sure the names here can't conflict with the unique generated names. Uniques generated names containing only base62 chars. So using say the '_' char guarantees this. -} -lmGlobalReg :: Platform -> String -> GlobalReg -> LlvmVar -lmGlobalReg platform suf reg +lmGlobalReg :: Platform -> String -> GlobalRegUse -> LlvmVar +lmGlobalReg platform suf (GlobalRegUse reg ty) = case reg of BaseReg -> ptrGlobal $ "Base" ++ suf Sp -> ptrGlobal $ "Sp" ++ suf @@ -88,13 +90,26 @@ lmGlobalReg platform suf reg ptrGlobal name = LMNLocalVar (fsLit name) (llvmWordPtr platform) floatGlobal name = LMNLocalVar (fsLit name) LMFloat doubleGlobal name = LMNLocalVar (fsLit name) LMDouble - xmmGlobal name = LMNLocalVar (fsLit name) (LMVector 4 (LMInt 32)) - ymmGlobal name = LMNLocalVar (fsLit name) (LMVector 8 (LMInt 32)) - zmmGlobal name = LMNLocalVar (fsLit name) (LMVector 16 (LMInt 32)) + fmt = cmmTypeFormat ty + xmmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + ymmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + zmmGlobal name = LMNLocalVar (fsLit name) (formatLlvmType fmt) + +formatLlvmType :: Format -> LlvmType +formatLlvmType II8 = LMInt 8 +formatLlvmType II16 = LMInt 16 +formatLlvmType II32 = LMInt 32 +formatLlvmType II64 = LMInt 64 +formatLlvmType FF32 = LMFloat +formatLlvmType FF64 = LMDouble +formatLlvmType (VecFormat l sFmt) = LMVector l (formatLlvmType $ scalarFormatFormat sFmt) -- | A list of STG Registers that should always be considered alive -alwaysLive :: [GlobalReg] -alwaysLive = [BaseReg, Sp, Hp, SpLim, HpLim, node] +alwaysLive :: Platform -> [GlobalRegUse] +alwaysLive platform = + [ GlobalRegUse r (globalRegSpillType platform r) + | r <- [BaseReg, Sp, Hp, SpLim, HpLim, node] + ] -- | STG Type Based Alias Analysis hierarchy stgTBAA :: [(Unique, LMString, Maybe Unique)] ===================================== testsuite/tests/simd/should_run/simd013.hs ===================================== @@ -8,25 +8,26 @@ module Main where import GHC.Exts import GHC.Prim -foreign import ccall "sub" +foreign import ccall unsafe "sub" sub :: DoubleX2# -> DoubleX2# -> DoubleX2# -foreign import ccall "add6" - add6 :: DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# +foreign import ccall unsafe "add7" + add7 :: DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2# main :: IO () main = do let x1, x2 :: DoubleX2# x1 = packDoubleX2# (# 9.9##, 99.99## #) x2 = packDoubleX2# (# 1.1##, 11.11## #) - y1, y2, y3, y4, y5, y6 :: DoubleX2# - !y1 = packDoubleX2# (# 1.5##, 2.5## #) - !y2 = packDoubleX2# (# 10.25##, 20.25## #) - !y3 = packDoubleX2# (# 100.125##, 200.125## #) - !y4 = packDoubleX2# (# 1000.0625##, 2000.0625## #) - !y5 = packDoubleX2# (# 10000.03125##, 20000.03125## #) - !y6 = packDoubleX2# (# 100000.015625##, 200000.015625## #) + y1, y2, y3, y4, y5, y6, y7 :: DoubleX2# + !y1 = packDoubleX2# (# 1.5##, 2.5## #) + !y2 = packDoubleX2# (# 10.25##, 20.25## #) + !y3 = packDoubleX2# (# 100.125##, 200.125## #) + !y4 = packDoubleX2# (# 1000.0625##, 2000.0625## #) + !y5 = packDoubleX2# (# 10000.03125##, 20000.03125## #) + !y6 = packDoubleX2# (# 100000.015625##, 200000.015625## #) + !y7 = packDoubleX2# (# 1000000.0078125##, 2000000.0078125## #) !(# a, b #) = unpackDoubleX2# ( sub x1 x2 ) - !(# c, d #) = unpackDoubleX2# ( add6 y1 y2 y3 y4 y5 y6 ) + !(# c, d #) = unpackDoubleX2# ( add7 y1 y2 y3 y4 y5 y6 y7 ) print ( D# a, D# b ) print ( D# c, D# d ) ===================================== testsuite/tests/simd/should_run/simd013.stdout ===================================== @@ -1,2 +1,2 @@ (8.8,88.88) -(111111.984375,222222.984375) +(1111111.9921875,2222222.9921875) ===================================== testsuite/tests/simd/should_run/simd013C.c ===================================== @@ -1,12 +1,12 @@ -#include +#include __m128d sub(__m128d x, __m128d y) { return _mm_sub_pd(x,y); } -__m128d add6(__m128d x1, __m128d x2, __m128d x3, __m128d x4, __m128d x5, __m128d x6) +__m128d add7(__m128d x1, __m128d x2, __m128d x3, __m128d x4, __m128d x5, __m128d x6, __m128d x7) { - return _mm_add_pd(x1,_mm_add_pd(x2,_mm_add_pd(x3,_mm_add_pd(x4,_mm_add_pd(x5,x6))))); + return _mm_add_pd(x1,_mm_add_pd(x2,_mm_add_pd(x3,_mm_add_pd(x4,_mm_add_pd(x5,_mm_add_pd(x6, x7)))))); } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a7b364be1d319a7f78246060eafb3d483bd94beb...9d0f68ea956702d09115ff8c8353b9d4eee8bc7d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a7b364be1d319a7f78246060eafb3d483bd94beb...9d0f68ea956702d09115ff8c8353b9d4eee8bc7d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 11:21:15 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Sun, 22 Sep 2024 07:21:15 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/az/epa-remove-hscaseannsrest Message-ID: <66effdab4c941_14990a4979889499b@gitlab.mail> Alan Zimmerman pushed new branch wip/az/epa-remove-hscaseannsrest at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/az/epa-remove-hscaseannsrest You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 12:12:23 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sun, 22 Sep 2024 08:12:23 -0400 Subject: [Git][ghc/ghc][wip/T24978] Add Given injectivity for built-in type families Message-ID: <66f009a7dd75d_25953c192698471ca@gitlab.mail> sheaf pushed to branch wip/T24978 at Glasgow Haskell Compiler / GHC Commits: efb4dcc5 by Simon Peyton Jones at 2024-09-22T14:12:15+02:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 10 changed files: - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/efb4dcc5281aedbf79d890449c8778afa9cb1549 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/efb4dcc5281aedbf79d890449c8778afa9cb1549 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 12:15:02 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sun, 22 Sep 2024 08:15:02 -0400 Subject: [Git][ghc/ghc][wip/T24978] Add Given injectivity for built-in type families Message-ID: <66f00a46a9609_25953c137f04494c0@gitlab.mail> sheaf pushed to branch wip/T24978 at Glasgow Haskell Compiler / GHC Commits: b538e8fb by Simon Peyton Jones at 2024-09-22T14:14:55+02:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 10 changed files: - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/b538e8fbdc64cbf707d61e06b232a532e77adbd3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/b538e8fbdc64cbf707d61e06b232a532e77adbd3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 12:15:22 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Sun, 22 Sep 2024 08:15:22 -0400 Subject: [Git][ghc/ghc][wip/T24978] Add Given injectivity for built-in type families Message-ID: <66f00a5a46aaa_25953c1925305003b@gitlab.mail> sheaf pushed to branch wip/T24978 at Glasgow Haskell Compiler / GHC Commits: 6fa1e5d1 by Simon Peyton Jones at 2024-09-22T14:15:13+02:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 10 changed files: - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6fa1e5d17c3773d925ffe0e1ade39e040ec5be14 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6fa1e5d17c3773d925ffe0e1ade39e040ec5be14 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 17:41:01 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sun, 22 Sep 2024 13:41:01 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 19 commits: rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. Message-ID: <66f056adab675_1466101912c0494ef@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 85b0bdfc by sheaf at 2024-09-22T13:40:45-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - d15a73f0 by sheaf at 2024-09-22T13:40:45-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 0b063b3a by sheaf at 2024-09-22T13:40:45-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - abbf9ed2 by sheaf at 2024-09-22T13:40:45-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - aa096563 by sheaf at 2024-09-22T13:40:45-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 1a38ab24 by sheaf at 2024-09-22T13:40:45-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 280d8196 by sheaf at 2024-09-22T13:40:45-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 74e68885 by sheaf at 2024-09-22T13:40:45-04:00 Add test for C calls & SIMD vectors - - - - - 1b45e325 by sheaf at 2024-09-22T13:40:46-04:00 Add test for #25169 - - - - - f95d7ba5 by sheaf at 2024-09-22T13:40:46-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 0789fbf2 by sheaf at 2024-09-22T13:40:46-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - c6568ae3 by sheaf at 2024-09-22T13:40:46-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - b88748d6 by sheaf at 2024-09-22T13:40:46-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 2140bbe4 by sheaf at 2024-09-22T13:40:46-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - dddcb435 by sheaf at 2024-09-22T13:40:46-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 36846fdf by Brandon Chinn at 2024-09-22T13:40:47-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 43e066e7 by Alan Zimmerman at 2024-09-22T13:40:47-04:00 EPA: Remove unused hsCaseAnnsRest We never populate it, so remove it. - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a938f57cf11a116b5fe7847e75c909ee9a16e0f3...43e066e762787ea5fd3f73499d9ed3406ad53a71 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a938f57cf11a116b5fe7847e75c909ee9a16e0f3...43e066e762787ea5fd3f73499d9ed3406ad53a71 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 19:26:19 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Sun, 22 Sep 2024 15:26:19 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] ci: Allow abi-test to fail. Message-ID: <66f06f5ba3bcd_1f499c4d099060311@gitlab.mail> Zubin pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: d63782db by Zubin Duggal at 2024-09-23T00:55:20+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - 1 changed file: - .gitlab/ci.sh Changes: ===================================== .gitlab/ci.sh ===================================== @@ -796,7 +796,7 @@ function check_objects(){ } function abi_test() { - for i in {1..10}; do info "iteration $i"; run_abi_test; done + for i in {1..10}; do info "iteration $i"; ( run_abi_test ) || warn "expected failure: abi test iteration $i failed, see #12935"; done } function run_abi_test() { View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d63782db041e3d26ed032936fac1d43253c6bacf -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d63782db041e3d26ed032936fac1d43253c6bacf You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 21:56:13 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Sun, 22 Sep 2024 17:56:13 -0400 Subject: [Git][ghc/ghc][wip/interpolated-strings] 5 commits: Replace manual string lexing (#25158) Message-ID: <66f0927d993ac_1f499ca19708656ce@gitlab.mail> Brandon Chinn pushed to branch wip/interpolated-strings at Glasgow Haskell Compiler / GHC Commits: e7c83fe4 by Brandon Chinn at 2024-09-21T20:22:05-07:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 0851f078 by Brandon Chinn at 2024-09-22T14:24:58-07:00 Merge branch 'wip/strings' into HEAD - - - - - ffdd7119 by Brandon Chinn at 2024-09-22T14:24:58-07:00 Unify ITstring + ITstringMulti - - - - - a21596df by Brandon Chinn at 2024-09-22T14:29:50-07:00 Unify HsString + HsMultilineString - - - - - 65d1c307 by Brandon Chinn at 2024-09-22T14:56:06-07:00 Implement interpolated strings - - - - - 30 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Lit.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/Language/Haskell/Syntax/Expr.hs - compiler/Language/Haskell/Syntax/Extension.hs - compiler/Language/Haskell/Syntax/Lit.hs - testsuite/driver/testlib.py - testsuite/tests/ghci/prog013/prog013.stderr - testsuite/tests/ghci/scripts/ghci022.stderr - testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr - testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr - testsuite/tests/parser/should_fail/T21843c.stderr - testsuite/tests/parser/should_fail/T21843e.stderr - testsuite/tests/parser/should_fail/T21843f.stderr - testsuite/tests/parser/should_fail/T3751.stderr - testsuite/tests/parser/should_fail/T5425.stderr - testsuite/tests/parser/should_fail/readFail002.stderr - testsuite/tests/parser/should_fail/readFail004.stderr - testsuite/tests/parser/should_fail/readFail005.stderr - testsuite/tests/parser/should_fail/readFail033.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e069215631ede6dc7c4ee35fab30dbe049ee856d...65d1c307e0dbf367ccc3555a1e6411b45c8efd09 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e069215631ede6dc7c4ee35fab30dbe049ee856d...65d1c307e0dbf367ccc3555a1e6411b45c8efd09 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 22 23:01:37 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sun, 22 Sep 2024 19:01:37 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 19 commits: The X86 SIMD patch. Message-ID: <66f0a1d0e76c3_1f499ccb1f9c7953d@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 758d4978 by sheaf at 2024-09-22T19:01:18-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 004bc7c6 by sheaf at 2024-09-22T19:01:18-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - a71096a4 by sheaf at 2024-09-22T19:01:18-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - c4612dc1 by sheaf at 2024-09-22T19:01:18-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 79feb9f4 by sheaf at 2024-09-22T19:01:18-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - dcbaf7e4 by sheaf at 2024-09-22T19:01:18-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - c3a49f0e by sheaf at 2024-09-22T19:01:18-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 8717a111 by sheaf at 2024-09-22T19:01:19-04:00 Add test for C calls & SIMD vectors - - - - - f9eb1b11 by sheaf at 2024-09-22T19:01:19-04:00 Add test for #25169 - - - - - 21fe2e03 by sheaf at 2024-09-22T19:01:19-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 22e713aa by sheaf at 2024-09-22T19:01:19-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 8f55547f by sheaf at 2024-09-22T19:01:19-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - 16fe8da6 by sheaf at 2024-09-22T19:01:19-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 28c7089a by sheaf at 2024-09-22T19:01:19-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 58bd6ceb by sheaf at 2024-09-22T19:01:19-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - c3c4e738 by Brandon Chinn at 2024-09-22T19:01:20-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 4fcd96b0 by ARATA Mizuki at 2024-09-22T19:01:25-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 638b7ee5 by Zubin Duggal at 2024-09-22T19:01:26-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 7d365b16 by Alan Zimmerman at 2024-09-22T19:01:27-04:00 EPA: Remove unused hsCaseAnnsRest We never populate it, so remove it. - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/43e066e762787ea5fd3f73499d9ed3406ad53a71...7d365b16489a29127b73583cce85cfa918e6ba4e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/43e066e762787ea5fd3f73499d9ed3406ad53a71...7d365b16489a29127b73583cce85cfa918e6ba4e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 02:13:17 2024 From: gitlab at gitlab.haskell.org (Fabricio Nascimento (@fabu)) Date: Sun, 22 Sep 2024 22:13:17 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/fabu/T24848-meaningful-return Message-ID: <66f0cebd43bcc_1458847886ec446fb@gitlab.mail> Fabricio Nascimento pushed new branch wip/fabu/T24848-meaningful-return at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/fabu/T24848-meaningful-return You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 03:52:26 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sun, 22 Sep 2024 23:52:26 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 18 commits: The X86 SIMD patch. Message-ID: <66f0e5fa27564_145884cb9d8c763cf@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 2a3e73c6 by sheaf at 2024-09-22T23:52:12-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 2e072369 by sheaf at 2024-09-22T23:52:12-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - f21b16e8 by sheaf at 2024-09-22T23:52:12-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 37489472 by sheaf at 2024-09-22T23:52:12-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 7e8b1292 by sheaf at 2024-09-22T23:52:12-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 3ec87ca1 by sheaf at 2024-09-22T23:52:12-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - c2920dfd by sheaf at 2024-09-22T23:52:12-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - e943e766 by sheaf at 2024-09-22T23:52:13-04:00 Add test for C calls & SIMD vectors - - - - - 441916e8 by sheaf at 2024-09-22T23:52:13-04:00 Add test for #25169 - - - - - df5ddcb1 by sheaf at 2024-09-22T23:52:13-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 5533f870 by sheaf at 2024-09-22T23:52:13-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 82cff827 by sheaf at 2024-09-22T23:52:13-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - e0b31446 by sheaf at 2024-09-22T23:52:13-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 3b9306c8 by sheaf at 2024-09-22T23:52:13-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 89b92cb9 by sheaf at 2024-09-22T23:52:13-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 944d47ce by Brandon Chinn at 2024-09-22T23:52:14-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 6729566d by ARATA Mizuki at 2024-09-22T23:52:16-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 6985e566 by Alan Zimmerman at 2024-09-22T23:52:17-04:00 EPA: Remove unused hsCaseAnnsRest We never populate it, so remove it. - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7d365b16489a29127b73583cce85cfa918e6ba4e...6985e5669ce660fb4ad315025e0fc305317a6057 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7d365b16489a29127b73583cce85cfa918e6ba4e...6985e5669ce660fb4ad315025e0fc305317a6057 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 08:16:42 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Mon, 23 Sep 2024 04:16:42 -0400 Subject: [Git][ghc/ghc][wip/T24978] Add Given injectivity for built-in type families Message-ID: <66f123ea925c8_2a58ac24f658827d8@gitlab.mail> Zubin pushed to branch wip/T24978 at Glasgow Haskell Compiler / GHC Commits: 1bcf5ca4 by Simon Peyton Jones at 2024-09-23T13:46:30+05:30 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 10 changed files: - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1bcf5ca473627e92fbdd9ccf541595371b3ea109 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1bcf5ca473627e92fbdd9ccf541595371b3ea109 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 08:27:13 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Mon, 23 Sep 2024 04:27:13 -0400 Subject: [Git][ghc/ghc][wip/25232] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66f126614bf6a_373a24e8fe4248b8@gitlab.mail> Zubin pushed to branch wip/25232 at Glasgow Haskell Compiler / GHC Commits: e48ff3a9 by Zubin Duggal at 2024-09-23T13:57:02+05:30 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 5 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - + testsuite/tests/rts/T25232.hs - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,18 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + // Not doing this lead to #25232 and #23958 + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +958,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): @@ -1026,6 +1023,9 @@ def have_dynamic_prof( ) -> bool: def have_profiling( ) -> bool: return config.have_profiling +def have_threaded( ) -> bool: + return config.ghc_with_threaded_rts + def in_tree_compiler( ) -> bool: return config.in_tree_compiler ===================================== testsuite/tests/rts/T25232.hs ===================================== @@ -0,0 +1,14 @@ +module Main where + +import Control.Monad +import Control.Monad.Trans.State.Strict +import Data.IORef + +main :: IO () +main = do + ref <- newIORef () + replicateM_ 1000 $ withIORef ref $ runStateT (pure ()) + +withIORef :: IORef a -> (a -> IO (b, a)) -> IO b +withIORef ref f = + readIORef ref >>= f >>= \(b, a) -> writeIORef ref a >> pure b ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [unless(have_profiling(), skip), only_ways(['normal','nonmoving','nonmoving_prof','nonmoving_thr_prof']), extra_ways(['nonmoving', 'nonmoving_prof'] + (['nonmoving_thr_prof'] if have_threaded() else []))], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e48ff3a9dc1779b05f4980bf94928c7abd699941 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e48ff3a9dc1779b05f4980bf94928c7abd699941 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 08:28:23 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Mon, 23 Sep 2024 04:28:23 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 21 commits: Demand: Combine examples into Note (#25107) Message-ID: <66f126a7d02fb_373a24ccc1827462@gitlab.mail> Zubin pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 589e300f by Matthew Pickering at 2024-09-23T08:28:18+00:00 ci: Run abi-test on test-abi label - - - - - dd6a2377 by Rodrigo Mesquita at 2024-09-23T08:28:18+00:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - 91a5836c by Rodrigo Mesquita at 2024-09-23T08:28:18+00:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - c50d56ed by Rodrigo Mesquita at 2024-09-23T08:28:18+00:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 5d40f060 by Rodrigo Mesquita at 2024-09-23T08:28:18+00:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 945f1da1 by Rodrigo Mesquita at 2024-09-23T08:28:18+00:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - a699d568 by Rodrigo Mesquita at 2024-09-23T08:28:18+00:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 4578b0f5 by Rodrigo Mesquita at 2024-09-23T08:28:18+00:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - c343a7dc by Zubin Duggal at 2024-09-23T08:28:18+00:00 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - 30 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d63782db041e3d26ed032936fac1d43253c6bacf...c343a7dc48942c1cdeaba79fb0394bd9bbd59434 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/d63782db041e3d26ed032936fac1d43253c6bacf...c343a7dc48942c1cdeaba79fb0394bd9bbd59434 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 08:33:25 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Mon, 23 Sep 2024 04:33:25 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 9 commits: ci: Run abi-test on test-abi label Message-ID: <66f127d572508_373a24bf89c28563@gitlab.mail> Zubin pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: c5ce2b8e by Matthew Pickering at 2024-09-18T16:28:36+01:00 ci: Run abi-test on test-abi label - - - - - db862ba8 by Rodrigo Mesquita at 2024-09-18T16:28:37+01:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - e8881e42 by Rodrigo Mesquita at 2024-09-18T16:28:37+01:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - f0641005 by Rodrigo Mesquita at 2024-09-18T18:09:53+01:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 0a5c2dac by Rodrigo Mesquita at 2024-09-18T19:16:08+01:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 475d1f2e by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 5af98c82 by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - ed43ad42 by Rodrigo Mesquita at 2024-09-18T19:16:22+01:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - e8e33703 by Zubin Duggal at 2024-09-23T14:03:16+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/RV64/CodeGen.hs - compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c343a7dc48942c1cdeaba79fb0394bd9bbd59434...e8e33703dea320e271b5772a90bb8a14d9dd8fe3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c343a7dc48942c1cdeaba79fb0394bd9bbd59434...e8e33703dea320e271b5772a90bb8a14d9dd8fe3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 08:33:52 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Mon, 23 Sep 2024 04:33:52 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 21 commits: Demand: Combine examples into Note (#25107) Message-ID: <66f127f0b8757_373a241dcc34296db@gitlab.mail> Zubin pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - a9fc88de by Matthew Pickering at 2024-09-23T08:33:49+00:00 ci: Run abi-test on test-abi label - - - - - c4c848ff by Rodrigo Mesquita at 2024-09-23T08:33:49+00:00 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - facd2d6e by Rodrigo Mesquita at 2024-09-23T08:33:49+00:00 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 7a3614d4 by Rodrigo Mesquita at 2024-09-23T08:33:49+00:00 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 0196f628 by Rodrigo Mesquita at 2024-09-23T08:33:49+00:00 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 17033aab by Rodrigo Mesquita at 2024-09-23T08:33:49+00:00 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - a219c4be by Rodrigo Mesquita at 2024-09-23T08:33:49+00:00 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 748c87d6 by Rodrigo Mesquita at 2024-09-23T08:33:49+00:00 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - f1782d19 by Zubin Duggal at 2024-09-23T08:33:49+00:00 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - 30 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e8e33703dea320e271b5772a90bb8a14d9dd8fe3...f1782d19565fd4016a3c157e626321c611bfd3c5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e8e33703dea320e271b5772a90bb8a14d9dd8fe3...f1782d19565fd4016a3c157e626321c611bfd3c5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 09:17:39 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 23 Sep 2024 05:17:39 -0400 Subject: [Git][ghc/ghc][wip/T24807] 19 commits: RTS: expose closure_sizeW_ (#25252) Message-ID: <66f1323375c9b_373a2475969440197@gitlab.mail> Rodrigo Mesquita pushed to branch wip/T24807 at Glasgow Haskell Compiler / GHC Commits: d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 72a85e43 by Ben Gamari at 2024-09-23T10:17:30+01:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - 30 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs.hs - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Pmc/Desugar.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/Tc/Zonk/Type.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5963b8c2de54d397f1c860e011951109f9c95163...72a85e43f2fdb860916696d0823eae8b2aa04329 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5963b8c2de54d397f1c860e011951109f9c95163...72a85e43f2fdb860916696d0823eae8b2aa04329 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 10:34:48 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 23 Sep 2024 06:34:48 -0400 Subject: [Git][ghc/ghc][wip/andreask/ghc-primops] ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps Message-ID: <66f14448a6530_6c8d62f82f840074@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/ghc-primops at Glasgow Haskell Compiler / GHC Commits: 3c32c060 by Andreas Klebinger at 2024-09-23T12:16:10+02:00 ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps This will be the new place for functions that would have gone into GHC.Exts in the past but are not stable enough to do so now. Addresses #25242 - - - - - 6 changed files: - libraries/base/src/GHC/Exts.hs - libraries/ghc-experimental/ghc-experimental.cabal.in - + libraries/ghc-experimental/src/GHC/PrimOps.hs - libraries/ghc-internal/src/GHC/Internal/Exts.hs - testsuite/tests/interface-stability/ghc-experimental-exports.stdout - + testsuite/tests/interface-stability/ghc-experimental-exports.stdout-mingw32 Changes: ===================================== libraries/base/src/GHC/Exts.hs ===================================== @@ -14,6 +14,9 @@ -- -- Note: no other @base@ module should import this module. +-- See Note [Where do we export PrimOps] for details about how to expose primops +-- to users. + module GHC.Exts (-- ** Pointer types Ptr(..), ===================================== libraries/ghc-experimental/ghc-experimental.cabal.in ===================================== @@ -29,6 +29,7 @@ library exposed-modules: Data.Sum.Experimental Data.Tuple.Experimental + GHC.PrimOps GHC.Profiling.Eras GHC.TypeLits.Experimental GHC.TypeNats.Experimental ===================================== libraries/ghc-experimental/src/GHC/PrimOps.hs ===================================== @@ -0,0 +1,33 @@ +{-# LANGUAGE FlexibleInstances #-} +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE MultiParamTypeClasses #-} +{-# LANGUAGE NoImplicitPrelude #-} +{-# LANGUAGE TypeFamilies #-} +{-# LANGUAGE UnboxedTuples #-} +{-# LANGUAGE Unsafe #-} +{-# LANGUAGE DeriveDataTypeable #-} +{-# OPTIONS_HADDOCK print-explicit-runtime-reps #-} + +----------------------------------------------------------------------------- +-- | +-- Module : GHC.PrimOps +-- Copyright : Andreas Klebinger 2024 +-- License : see libraries/ghc-experimental/LICENSE +-- +-- Maintainer : ghc-devs at haskell.org +-- Stability : internal +-- Portability : non-portable (GHC Extensions) +-- +-- GHC Extensions: This is the Approved Way to get at GHC-specific extensions +-- without relying on the ghc-internal package. +----------------------------------------------------------------------------- + +-- See Note [Where do we export PrimOps] for the purpose of this module. + +module GHC.PrimOps + ( + module GHC.Internal.Exts, + ) where + +import GHC.Internal.Exts + ===================================== libraries/ghc-internal/src/GHC/Internal/Exts.hs ===================================== @@ -18,11 +18,43 @@ -- Stability : internal -- Portability : non-portable (GHC Extensions) -- --- GHC Extensions: this is the Approved Way to get at GHC-specific extensions. +-- GHC Extensions: This is a unstable way to get at GHC-specific extensions. +-- If possible prefer using GHC.PrimOps from ghc-experimental +-- or GHC.Exts from base. -- --- Note: no other base module should import this module. +-- Note: no other ghc-internal module should import this module. ----------------------------------------------------------------------------- +{- Note [Where do we export PrimOps] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Built in primops are automatically exported via the magical module +GHC.Prim (See Note [GHC.Prim]). +However we don't want users to use GHC.Prim directly. Among other reasons that +would prevent us from giving those primops a regular haskell or cmm +implementation in the future. + +So instead we provide a hirarchy of Modules through which we expose PrimOps. +* ghc-internal:GHC.Exts.Internal re-exports GHC.Prim along with some other + builtin functionality. It gives zero stability guarantee and is mostly inteded + for ghc internal use. +* ghc-experimental:GHC.PrimOps contains a more stable subset of GHC.Exts.Internal. + This module does not offer any more stability *guarantees* than ghc-internal, + but we try to keep things slightly more stable there and be backwards compatible + if it's easy to do. For example an export from ghc-experimental is likely to get + deprecated before being removed. While exports from ghc-internal could be removed + without advanced notice between ghc versions! +* base:GHC.Exts: Contains all kinds of re-exports from ghc-internal:GHC.Exts.Internal + which have been grandfathered into this module from pre ghc-9.10. + Starting with 9.10.1 only definitions with a fairly high level of expected + stability should be exposed through GHC.Exts and changing the API provided by + this module now follows the CLC process. + +If there is a desire to create a module for a specific group of primitives they +should still be exported by the catch-all modules GHC.Exts.Internal and GHC.PrimOps, +with their specific module being placed under ghc-experimental:GHC.PrimOps +in the module structure. +-} + module GHC.Internal.Exts ( -- ** Pointer types ===================================== testsuite/tests/interface-stability/ghc-experimental-exports.stdout ===================================== @@ -4316,6 +4316,1721 @@ module Data.Tuple.Experimental where data Unit# = ... getSolo :: forall a. Solo a -> a +module GHC.PrimOps where + -- Safety: Unsafe + (*#) :: Int# -> Int# -> Int# + (*##) :: Double# -> Double# -> Double# + (**##) :: Double# -> Double# -> Double# + (+#) :: Int# -> Int# -> Int# + (+##) :: Double# -> Double# -> Double# + (-#) :: Int# -> Int# -> Int# + (-##) :: Double# -> Double# -> Double# + (/##) :: Double# -> Double# -> Double# + (/=#) :: Int# -> Int# -> Int# + (/=##) :: Double# -> Double# -> Int# + (<#) :: Int# -> Int# -> Int# + (<##) :: Double# -> Double# -> Int# + (<=#) :: Int# -> Int# -> Int# + (<=##) :: Double# -> Double# -> Int# + (==#) :: Int# -> Int# -> Int# + (==##) :: Double# -> Double# -> Int# + (>#) :: Int# -> Int# -> Int# + (>##) :: Double# -> Double# -> Int# + (>=#) :: Int# -> Int# -> Int# + (>=##) :: Double# -> Double# -> Int# + type Addr# :: TYPE AddrRep + data Addr# + type Any :: forall k. k + type family Any where + type Array# :: forall {l :: Levity}. TYPE (BoxedRep l) -> UnliftedType + data Array# a + type ArrayArray# :: UnliftedType + newtype ArrayArray# = ArrayArray# (Array# ByteArray#) + type BCO :: * + data BCO + type Bool :: * + data Bool = False | True + type ByteArray# :: UnliftedType + data ByteArray# + type role CONSTRAINT nominal + type CONSTRAINT :: RuntimeRep -> * + data CONSTRAINT a + type Char :: * + data Char = C# Char# + type Char# :: TYPE WordRep + data Char# + type role Coercible representational representational + type Coercible :: forall k. k -> k -> Constraint + class Coercible a b => Coercible a b + {-# MINIMAL #-} + type Compact# :: UnliftedType + data Compact# + type Constraint :: * + type Constraint = CONSTRAINT LiftedRep + type DataToTag :: forall {lev :: Levity}. TYPE (BoxedRep lev) -> Constraint + class DataToTag a where + dataToTag# :: a -> Int# + {-# MINIMAL dataToTag# #-} + type DictBox :: Constraint -> * + data DictBox a = a => MkDictBox + type Double :: * + data Double = D# Double# + type Double# :: TYPE DoubleRep + data Double# + type DoubleBox :: TYPE DoubleRep -> * + data DoubleBox a = MkDoubleBox a + type DoubleX2# :: TYPE (VecRep Vec2 DoubleElemRep) + data DoubleX2# + type DoubleX4# :: TYPE (VecRep Vec4 DoubleElemRep) + data DoubleX4# + type DoubleX8# :: TYPE (VecRep Vec8 DoubleElemRep) + data DoubleX8# + type Down :: * -> * + newtype Down a = Down {getDown :: a} + type role FUN nominal representational representational + type FUN :: forall (n :: Multiplicity) -> forall {q :: RuntimeRep} {r :: RuntimeRep}. TYPE q -> TYPE r -> * + data FUN n a b + type Float :: * + data Float = F# Float# + type Float# :: TYPE FloatRep + data Float# + type FloatBox :: TYPE FloatRep -> * + data FloatBox a = MkFloatBox a + type FloatX16# :: TYPE (VecRep Vec16 FloatElemRep) + data FloatX16# + type FloatX4# :: TYPE (VecRep Vec4 FloatElemRep) + data FloatX4# + type FloatX8# :: TYPE (VecRep Vec8 FloatElemRep) + data FloatX8# + type role FunPtr phantom + type FunPtr :: * -> * + data FunPtr a = FunPtr Addr# + type role IOPort# nominal representational + type IOPort# :: forall {l :: Levity}. * -> TYPE (BoxedRep l) -> UnliftedType + data IOPort# a b + type Int :: * + data Int = I# Int# + type Int# :: TYPE IntRep + data Int# + type Int16# :: TYPE Int16Rep + data Int16# + type Int16X16# :: TYPE (VecRep Vec16 Int16ElemRep) + data Int16X16# + type Int16X32# :: TYPE (VecRep Vec32 Int16ElemRep) + data Int16X32# + type Int16X8# :: TYPE (VecRep Vec8 Int16ElemRep) + data Int16X8# + type Int32# :: TYPE Int32Rep + data Int32# + type Int32X16# :: TYPE (VecRep Vec16 Int32ElemRep) + data Int32X16# + type Int32X4# :: TYPE (VecRep Vec4 Int32ElemRep) + data Int32X4# + type Int32X8# :: TYPE (VecRep Vec8 Int32ElemRep) + data Int32X8# + type Int64# :: TYPE Int64Rep + data Int64# + type Int64X2# :: TYPE (VecRep Vec2 Int64ElemRep) + data Int64X2# + type Int64X4# :: TYPE (VecRep Vec4 Int64ElemRep) + data Int64X4# + type Int64X8# :: TYPE (VecRep Vec8 Int64ElemRep) + data Int64X8# + type Int8# :: TYPE Int8Rep + data Int8# + type Int8X16# :: TYPE (VecRep Vec16 Int8ElemRep) + data Int8X16# + type Int8X32# :: TYPE (VecRep Vec32 Int8ElemRep) + data Int8X32# + type Int8X64# :: TYPE (VecRep Vec64 Int8ElemRep) + data Int8X64# + type IntBox :: TYPE IntRep -> * + data IntBox a = MkIntBox a + type IsList :: * -> Constraint + class IsList l where + type Item :: * -> * + type family Item l + fromList :: [Item l] -> l + fromListN :: Int -> [Item l] -> l + toList :: l -> [Item l] + {-# MINIMAL fromList, toList #-} + type IsString :: * -> Constraint + class IsString a where + fromString :: GHC.Internal.Base.String -> a + {-# MINIMAL fromString #-} + KindRepApp :: GHC.Types.KindRep -> GHC.Types.KindRep -> GHC.Types.KindRep + KindRepFun :: GHC.Types.KindRep -> GHC.Types.KindRep -> GHC.Types.KindRep + KindRepTYPE :: RuntimeRep -> GHC.Types.KindRep + KindRepTyConApp :: GHC.Types.TyCon -> [GHC.Types.KindRep] -> GHC.Types.KindRep + KindRepTypeLitD :: GHC.Types.TypeLitSort -> [Char] -> GHC.Types.KindRep + KindRepTypeLitS :: GHC.Types.TypeLitSort -> Addr# -> GHC.Types.KindRep + KindRepVar :: GHC.Types.KindBndr -> GHC.Types.KindRep + type Levity :: * + data Levity = Lifted | Unlifted + type LiftedRep :: RuntimeRep + type LiftedRep = BoxedRep Lifted :: RuntimeRep + type List :: * -> * + data List a = ... + type role MVar# nominal representational + type MVar# :: forall {l :: Levity}. * -> TYPE (BoxedRep l) -> UnliftedType + data MVar# a b + type MultMul :: Multiplicity -> Multiplicity -> Multiplicity + type family MultMul a b where + forall (x :: Multiplicity). MultMul One x = x + forall (x :: Multiplicity). MultMul x One = x + forall (x :: Multiplicity). MultMul Many x = Many + forall (x :: Multiplicity). MultMul x Many = Many + type Multiplicity :: * + data Multiplicity = One | Many + type role MutVar# nominal representational + type MutVar# :: forall {l :: Levity}. * -> TYPE (BoxedRep l) -> UnliftedType + data MutVar# a b + type role MutableArray# nominal representational + type MutableArray# :: forall {l :: Levity}. * -> TYPE (BoxedRep l) -> UnliftedType + data MutableArray# a b + type role MutableArrayArray# nominal + type MutableArrayArray# :: * -> UnliftedType + newtype MutableArrayArray# s = MutableArrayArray# (MutableArray# s ByteArray#) + type role MutableByteArray# nominal + type MutableByteArray# :: * -> UnliftedType + data MutableByteArray# a + type Ordering :: * + data Ordering = LT | EQ | GT + type PromptTag# :: * -> UnliftedType + data PromptTag# a + type role Proxy# phantom + type Proxy# :: forall k. k -> ZeroBitType + data Proxy# a + type role Ptr phantom + type Ptr :: * -> * + data Ptr a = Ptr Addr# + type RealWorld :: * + data RealWorld + type RuntimeRep :: * + data RuntimeRep = VecRep VecCount VecElem | TupleRep [RuntimeRep] | SumRep [RuntimeRep] | BoxedRep Levity | IntRep | Int8Rep | Int16Rep | Int32Rep | Int64Rep | WordRep | Word8Rep | Word16Rep | Word32Rep | Word64Rep | AddrRep | FloatRep | DoubleRep + type SPEC :: * + data SPEC = SPEC | SPEC2 + type SmallArray# :: forall {l :: Levity}. TYPE (BoxedRep l) -> UnliftedType + data SmallArray# a + type role SmallMutableArray# nominal representational + type SmallMutableArray# :: forall {l :: Levity}. * -> TYPE (BoxedRep l) -> UnliftedType + data SmallMutableArray# a b + type SpecConstrAnnotation :: * + data SpecConstrAnnotation = NoSpecConstr | ForceSpecConstr + type role StableName# phantom + type StableName# :: forall {l :: Levity}. TYPE (BoxedRep l) -> UnliftedType + data StableName# a + type StablePtr# :: forall {l :: Levity}. TYPE (BoxedRep l) -> TYPE AddrRep + data StablePtr# a + type StackSnapshot# :: UnliftedType + data StackSnapshot# + type role State# nominal + type State# :: * -> ZeroBitType + data State# a + type Symbol :: * + data Symbol + type role TVar# nominal representational + type TVar# :: forall {l :: Levity}. * -> TYPE (BoxedRep l) -> UnliftedType + data TVar# a b + type role TYPE nominal + type TYPE :: RuntimeRep -> * + data TYPE a + type ThreadId# :: UnliftedType + data ThreadId# + TrNameD :: [Char] -> GHC.Types.TrName + TrNameS :: Addr# -> GHC.Types.TrName + TypeLitChar :: GHC.Types.TypeLitSort + TypeLitNat :: GHC.Types.TypeLitSort + TypeLitSymbol :: GHC.Types.TypeLitSort + type UnliftedRep :: RuntimeRep + type UnliftedRep = BoxedRep Unlifted :: RuntimeRep + type UnliftedType :: * + type UnliftedType = TYPE UnliftedRep + type VecCount :: * + data VecCount = Vec2 | Vec4 | Vec8 | Vec16 | Vec32 | Vec64 + type VecElem :: * + data VecElem = Int8ElemRep | Int16ElemRep | Int32ElemRep | Int64ElemRep | Word8ElemRep | Word16ElemRep | Word32ElemRep | Word64ElemRep | FloatElemRep | DoubleElemRep + type Void# :: ZeroBitType + type Void# = (# #) :: ZeroBitType + type Weak# :: forall {l :: Levity}. TYPE (BoxedRep l) -> UnliftedType + data Weak# a + type WithDict :: Constraint -> * -> Constraint + class WithDict cls meth where + withDict :: forall {rr :: RuntimeRep} (r :: TYPE rr). meth -> (cls => r) -> r + {-# MINIMAL withDict #-} + type Word :: * + data Word = W# Word# + type Word# :: TYPE WordRep + data Word# + type Word16# :: TYPE Word16Rep + data Word16# + type Word16X16# :: TYPE (VecRep Vec16 Word16ElemRep) + data Word16X16# + type Word16X32# :: TYPE (VecRep Vec32 Word16ElemRep) + data Word16X32# + type Word16X8# :: TYPE (VecRep Vec8 Word16ElemRep) + data Word16X8# + type Word32# :: TYPE Word32Rep + data Word32# + type Word32X16# :: TYPE (VecRep Vec16 Word32ElemRep) + data Word32X16# + type Word32X4# :: TYPE (VecRep Vec4 Word32ElemRep) + data Word32X4# + type Word32X8# :: TYPE (VecRep Vec8 Word32ElemRep) + data Word32X8# + type Word64# :: TYPE Word64Rep + data Word64# + type Word64X2# :: TYPE (VecRep Vec2 Word64ElemRep) + data Word64X2# + type Word64X4# :: TYPE (VecRep Vec4 Word64ElemRep) + data Word64X4# + type Word64X8# :: TYPE (VecRep Vec8 Word64ElemRep) + data Word64X8# + type Word8# :: TYPE Word8Rep + data Word8# + type Word8X16# :: TYPE (VecRep Vec16 Word8ElemRep) + data Word8X16# + type Word8X32# :: TYPE (VecRep Vec32 Word8ElemRep) + data Word8X32# + type Word8X64# :: TYPE (VecRep Vec64 Word8ElemRep) + data Word8X64# + type WordBox :: TYPE WordRep -> * + data WordBox a = MkWordBox a + type ZeroBitRep :: RuntimeRep + type ZeroBitRep = TupleRep '[] :: RuntimeRep + type ZeroBitType :: * + type ZeroBitType = TYPE ZeroBitRep + acosDouble# :: Double# -> Double# + acosFloat# :: Float# -> Float# + acoshDouble# :: Double# -> Double# + acoshFloat# :: Float# -> Float# + addCFinalizerToWeak# :: forall {k :: Levity} (b :: TYPE (BoxedRep k)). Addr# -> Addr# -> Int# -> Addr# -> Weak# b -> State# RealWorld -> (# State# RealWorld, Int# #) + addIntC# :: Int# -> Int# -> (# Int#, Int# #) + addWordC# :: Word# -> Word# -> (# Word#, Int# #) + addr2Int# :: Addr# -> Int# + addrToAny# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). Addr# -> (# a #) + and# :: Word# -> Word# -> Word# + and64# :: Word64# -> Word64# -> Word64# + andI# :: Int# -> Int# -> Int# + andWord16# :: Word16# -> Word16# -> Word16# + andWord32# :: Word32# -> Word32# -> Word32# + andWord8# :: Word8# -> Word8# -> Word8# + anyToAddr# :: forall a. a -> State# RealWorld -> (# State# RealWorld, Addr# #) + asinDouble# :: Double# -> Double# + asinFloat# :: Float# -> Float# + asinhDouble# :: Double# -> Double# + asinhFloat# :: Float# -> Float# + atanDouble# :: Double# -> Double# + atanFloat# :: Float# -> Float# + atanhDouble# :: Double# -> Double# + atanhFloat# :: Float# -> Float# + atomicCasAddrAddr# :: forall d. Addr# -> Addr# -> Addr# -> State# d -> (# State# d, Addr# #) + atomicCasWord16Addr# :: forall d. Addr# -> Word16# -> Word16# -> State# d -> (# State# d, Word16# #) + atomicCasWord32Addr# :: forall d. Addr# -> Word32# -> Word32# -> State# d -> (# State# d, Word32# #) + atomicCasWord64Addr# :: forall d. Addr# -> Word64# -> Word64# -> State# d -> (# State# d, Word64# #) + atomicCasWord8Addr# :: forall d. Addr# -> Word8# -> Word8# -> State# d -> (# State# d, Word8# #) + atomicCasWordAddr# :: forall d. Addr# -> Word# -> Word# -> State# d -> (# State# d, Word# #) + atomicExchangeAddrAddr# :: forall d. Addr# -> Addr# -> State# d -> (# State# d, Addr# #) + atomicExchangeWordAddr# :: forall d. Addr# -> Word# -> State# d -> (# State# d, Word# #) + atomicModifyMutVar# :: forall s a b c. MutVar# s a -> (a -> b) -> State# s -> (# State# s, c #) + atomicModifyMutVar2# :: forall d a c. MutVar# d a -> (a -> c) -> State# d -> (# State# d, a, c #) + atomicModifyMutVar_# :: forall d a. MutVar# d a -> (a -> a) -> State# d -> (# State# d, a, a #) + atomicReadIntArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int# #) + atomicReadWordAddr# :: forall d. Addr# -> State# d -> (# State# d, Word# #) + atomicSwapMutVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutVar# d a -> a -> State# d -> (# State# d, a #) + atomicWriteIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> State# d + atomicWriteWordAddr# :: forall d. Addr# -> Word# -> State# d -> State# d + atomically# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). (State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #) + augment :: forall a. (forall b. (a -> b -> b) -> b -> b) -> [a] -> [a] + bitReverse# :: Word# -> Word# + bitReverse16# :: Word# -> Word# + bitReverse32# :: Word# -> Word# + bitReverse64# :: Word64# -> Word64# + bitReverse8# :: Word# -> Word# + breakpoint :: forall a. a -> a + breakpointCond :: forall a. Bool -> a -> a + broadcastDoubleX2# :: Double# -> DoubleX2# + broadcastDoubleX4# :: Double# -> DoubleX4# + broadcastDoubleX8# :: Double# -> DoubleX8# + broadcastFloatX16# :: Float# -> FloatX16# + broadcastFloatX4# :: Float# -> FloatX4# + broadcastFloatX8# :: Float# -> FloatX8# + broadcastInt16X16# :: Int16# -> Int16X16# + broadcastInt16X32# :: Int16# -> Int16X32# + broadcastInt16X8# :: Int16# -> Int16X8# + broadcastInt32X16# :: Int32# -> Int32X16# + broadcastInt32X4# :: Int32# -> Int32X4# + broadcastInt32X8# :: Int32# -> Int32X8# + broadcastInt64X2# :: Int64# -> Int64X2# + broadcastInt64X4# :: Int64# -> Int64X4# + broadcastInt64X8# :: Int64# -> Int64X8# + broadcastInt8X16# :: Int8# -> Int8X16# + broadcastInt8X32# :: Int8# -> Int8X32# + broadcastInt8X64# :: Int8# -> Int8X64# + broadcastWord16X16# :: Word16# -> Word16X16# + broadcastWord16X32# :: Word16# -> Word16X32# + broadcastWord16X8# :: Word16# -> Word16X8# + broadcastWord32X16# :: Word32# -> Word32X16# + broadcastWord32X4# :: Word32# -> Word32X4# + broadcastWord32X8# :: Word32# -> Word32X8# + broadcastWord64X2# :: Word64# -> Word64X2# + broadcastWord64X4# :: Word64# -> Word64X4# + broadcastWord64X8# :: Word64# -> Word64X8# + broadcastWord8X16# :: Word8# -> Word8X16# + broadcastWord8X32# :: Word8# -> Word8X32# + broadcastWord8X64# :: Word8# -> Word8X64# + build :: forall a. (forall b. (a -> b -> b) -> b -> b) -> [a] + byteArrayContents# :: ByteArray# -> Addr# + byteSwap# :: Word# -> Word# + byteSwap16# :: Word# -> Word# + byteSwap32# :: Word# -> Word# + byteSwap64# :: Word64# -> Word64# + casArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutableArray# d a -> Int# -> a -> a -> State# d -> (# State# d, Int#, a #) + casInt16Array# :: forall d. MutableByteArray# d -> Int# -> Int16# -> Int16# -> State# d -> (# State# d, Int16# #) + casInt32Array# :: forall d. MutableByteArray# d -> Int# -> Int32# -> Int32# -> State# d -> (# State# d, Int32# #) + casInt64Array# :: forall d. MutableByteArray# d -> Int# -> Int64# -> Int64# -> State# d -> (# State# d, Int64# #) + casInt8Array# :: forall d. MutableByteArray# d -> Int# -> Int8# -> Int8# -> State# d -> (# State# d, Int8# #) + casIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> (# State# d, Int# #) + casMutVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutVar# d a -> a -> a -> State# d -> (# State# d, Int#, a #) + casSmallArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> Int# -> a -> a -> State# d -> (# State# d, Int#, a #) + castDoubleToWord64# :: Double# -> Word64# + castFloatToWord32# :: Float# -> Word32# + castWord32ToFloat# :: Word32# -> Float# + castWord64ToDouble# :: Word64# -> Double# + catch# :: forall {q :: RuntimeRep} {k :: Levity} (a :: TYPE q) (b :: TYPE (BoxedRep k)). (State# RealWorld -> (# State# RealWorld, a #)) -> (b -> State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #) + catchRetry# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). (State# RealWorld -> (# State# RealWorld, a #)) -> (State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #) + catchSTM# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) b. (State# RealWorld -> (# State# RealWorld, a #)) -> (b -> State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #) + chr# :: Int# -> Char# + clearCCS# :: forall d a. (State# d -> (# State# d, a #)) -> State# d -> (# State# d, a #) + cloneArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). Array# a -> Int# -> Int# -> Array# a + cloneMutableArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutableArray# d a -> Int# -> Int# -> State# d -> (# State# d, MutableArray# d a #) + cloneSmallArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). SmallArray# a -> Int# -> Int# -> SmallArray# a + cloneSmallMutableArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> Int# -> Int# -> State# d -> (# State# d, SmallMutableArray# d a #) + closureSize# :: forall a. a -> Int# + clz# :: Word# -> Word# + clz16# :: Word# -> Word# + clz32# :: Word# -> Word# + clz64# :: Word64# -> Word# + clz8# :: Word# -> Word# + coerce :: forall {k :: RuntimeRep} (a :: TYPE k) (b :: TYPE k). Coercible a b => a -> b + compactAdd# :: forall a. Compact# -> a -> State# RealWorld -> (# State# RealWorld, a #) + compactAddWithSharing# :: forall a. Compact# -> a -> State# RealWorld -> (# State# RealWorld, a #) + compactAllocateBlock# :: Word# -> Addr# -> State# RealWorld -> (# State# RealWorld, Addr# #) + compactContains# :: forall a. Compact# -> a -> State# RealWorld -> (# State# RealWorld, Int# #) + compactContainsAny# :: forall a. a -> State# RealWorld -> (# State# RealWorld, Int# #) + compactFixupPointers# :: Addr# -> Addr# -> State# RealWorld -> (# State# RealWorld, Compact#, Addr# #) + compactGetFirstBlock# :: Compact# -> State# RealWorld -> (# State# RealWorld, Addr#, Word# #) + compactGetNextBlock# :: Compact# -> Addr# -> State# RealWorld -> (# State# RealWorld, Addr#, Word# #) + compactNew# :: Word# -> State# RealWorld -> (# State# RealWorld, Compact# #) + compactResize# :: Compact# -> Word# -> State# RealWorld -> State# RealWorld + compactSize# :: Compact# -> State# RealWorld -> (# State# RealWorld, Word# #) + compareByteArrays# :: ByteArray# -> Int# -> ByteArray# -> Int# -> Int# -> Int# + considerAccessible :: Bool + control0# :: forall {r :: RuntimeRep} a (b :: TYPE r). PromptTag# a -> (((State# RealWorld -> (# State# RealWorld, b #)) -> State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, b #) + copyAddrToAddr# :: Addr# -> Addr# -> Int# -> State# RealWorld -> State# RealWorld + copyAddrToAddrNonOverlapping# :: Addr# -> Addr# -> Int# -> State# RealWorld -> State# RealWorld + copyAddrToByteArray# :: forall d. Addr# -> MutableByteArray# d -> Int# -> Int# -> State# d -> State# d + copyArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. Array# a -> Int# -> MutableArray# d a -> Int# -> Int# -> State# d -> State# d + copyArrayArray# :: forall s. ArrayArray# -> Int# -> MutableArrayArray# s -> Int# -> Int# -> State# s -> State# s + copyByteArray# :: forall d. ByteArray# -> Int# -> MutableByteArray# d -> Int# -> Int# -> State# d -> State# d + copyByteArrayToAddr# :: forall d. ByteArray# -> Int# -> Addr# -> Int# -> State# d -> State# d + copyMutableArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutableArray# d a -> Int# -> MutableArray# d a -> Int# -> Int# -> State# d -> State# d + copyMutableArrayArray# :: forall s. MutableArrayArray# s -> Int# -> MutableArrayArray# s -> Int# -> Int# -> State# s -> State# s + copyMutableByteArray# :: forall d. MutableByteArray# d -> Int# -> MutableByteArray# d -> Int# -> Int# -> State# d -> State# d + copyMutableByteArrayNonOverlapping# :: forall d. MutableByteArray# d -> Int# -> MutableByteArray# d -> Int# -> Int# -> State# d -> State# d + copyMutableByteArrayToAddr# :: forall d. MutableByteArray# d -> Int# -> Addr# -> Int# -> State# d -> State# d + copySmallArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. SmallArray# a -> Int# -> SmallMutableArray# d a -> Int# -> Int# -> State# d -> State# d + copySmallMutableArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> Int# -> SmallMutableArray# d a -> Int# -> Int# -> State# d -> State# d + cosDouble# :: Double# -> Double# + cosFloat# :: Float# -> Float# + coshDouble# :: Double# -> Double# + coshFloat# :: Float# -> Float# + cstringLength# :: Addr# -> Int# + ctz# :: Word# -> Word# + ctz16# :: Word# -> Word# + ctz32# :: Word# -> Word# + ctz64# :: Word64# -> Word# + ctz8# :: Word# -> Word# + currentCallStack :: GHC.Types.IO [GHC.Internal.Base.String] + deRefStablePtr# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). StablePtr# a -> State# RealWorld -> (# State# RealWorld, a #) + deRefWeak# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). Weak# a -> State# RealWorld -> (# State# RealWorld, Int#, a #) + decodeDouble_2Int# :: Double# -> (# Int#, Word#, Word#, Int# #) + decodeDouble_Int64# :: Double# -> (# Int64#, Int# #) + decodeFloat_Int# :: Float# -> (# Int#, Int# #) + delay# :: forall d. Int# -> State# d -> State# d + divideDoubleX2# :: DoubleX2# -> DoubleX2# -> DoubleX2# + divideDoubleX4# :: DoubleX4# -> DoubleX4# -> DoubleX4# + divideDoubleX8# :: DoubleX8# -> DoubleX8# -> DoubleX8# + divideFloat# :: Float# -> Float# -> Float# + divideFloatX16# :: FloatX16# -> FloatX16# -> FloatX16# + divideFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# + divideFloatX8# :: FloatX8# -> FloatX8# -> FloatX8# + double2Float# :: Double# -> Float# + double2Int# :: Double# -> Int# + eqAddr# :: Addr# -> Addr# -> Int# + eqChar# :: Char# -> Char# -> Int# + eqFloat# :: Float# -> Float# -> Int# + eqInt16# :: Int16# -> Int16# -> Int# + eqInt32# :: Int32# -> Int32# -> Int# + eqInt64# :: Int64# -> Int64# -> Int# + eqInt8# :: Int8# -> Int8# -> Int# + eqStableName# :: forall {k :: Levity} {l :: Levity} (a :: TYPE (BoxedRep k)) (b :: TYPE (BoxedRep l)). StableName# a -> StableName# b -> Int# + eqStablePtr# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). StablePtr# a -> StablePtr# a -> Int# + eqWord# :: Word# -> Word# -> Int# + eqWord16# :: Word16# -> Word16# -> Int# + eqWord32# :: Word32# -> Word32# -> Int# + eqWord64# :: Word64# -> Word64# -> Int# + eqWord8# :: Word8# -> Word8# -> Int# + expDouble# :: Double# -> Double# + expFloat# :: Float# -> Float# + expm1Double# :: Double# -> Double# + expm1Float# :: Float# -> Float# + fabsDouble# :: Double# -> Double# + fabsFloat# :: Float# -> Float# + fetchAddIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> (# State# d, Int# #) + fetchAddWordAddr# :: forall d. Addr# -> Word# -> State# d -> (# State# d, Word# #) + fetchAndIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> (# State# d, Int# #) + fetchAndWordAddr# :: forall d. Addr# -> Word# -> State# d -> (# State# d, Word# #) + fetchNandIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> (# State# d, Int# #) + fetchNandWordAddr# :: forall d. Addr# -> Word# -> State# d -> (# State# d, Word# #) + fetchOrIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> (# State# d, Int# #) + fetchOrWordAddr# :: forall d. Addr# -> Word# -> State# d -> (# State# d, Word# #) + fetchSubIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> (# State# d, Int# #) + fetchSubWordAddr# :: forall d. Addr# -> Word# -> State# d -> (# State# d, Word# #) + fetchXorIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> (# State# d, Int# #) + fetchXorWordAddr# :: forall d. Addr# -> Word# -> State# d -> (# State# d, Word# #) + finalizeWeak# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) b. Weak# a -> State# RealWorld -> (# State# RealWorld, Int#, State# RealWorld -> (# State# RealWorld, b #) #) + float2Double# :: Float# -> Double# + float2Int# :: Float# -> Int# + fmaddDouble# :: Double# -> Double# -> Double# -> Double# + fmaddFloat# :: Float# -> Float# -> Float# -> Float# + fmsubDouble# :: Double# -> Double# -> Double# -> Double# + fmsubFloat# :: Float# -> Float# -> Float# -> Float# + fnmaddDouble# :: Double# -> Double# -> Double# -> Double# + fnmaddFloat# :: Float# -> Float# -> Float# -> Float# + fnmsubDouble# :: Double# -> Double# -> Double# -> Double# + fnmsubFloat# :: Float# -> Float# -> Float# -> Float# + fork# :: forall {q :: RuntimeRep} (a :: TYPE q). (State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, ThreadId# #) + forkOn# :: forall {q :: RuntimeRep} (a :: TYPE q). Int# -> (State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, ThreadId# #) + freezeArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutableArray# d a -> Int# -> Int# -> State# d -> (# State# d, Array# a #) + freezeSmallArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> Int# -> Int# -> State# d -> (# State# d, SmallArray# a #) + geAddr# :: Addr# -> Addr# -> Int# + geChar# :: Char# -> Char# -> Int# + geFloat# :: Float# -> Float# -> Int# + geInt16# :: Int16# -> Int16# -> Int# + geInt32# :: Int32# -> Int32# -> Int# + geInt64# :: Int64# -> Int64# -> Int# + geInt8# :: Int8# -> Int8# -> Int# + geWord# :: Word# -> Word# -> Int# + geWord16# :: Word16# -> Word16# -> Int# + geWord32# :: Word32# -> Word32# -> Int# + geWord64# :: Word64# -> Word64# -> Int# + geWord8# :: Word8# -> Word8# -> Int# + getApStackVal# :: forall a b. a -> Int# -> (# Int#, b #) + getCCSOf# :: forall a d. a -> State# d -> (# State# d, Addr# #) + getCurrentCCS# :: forall a d. a -> State# d -> (# State# d, Addr# #) + getMaskingState# :: State# RealWorld -> (# State# RealWorld, Int# #) + getSizeofMutableByteArray# :: forall d. MutableByteArray# d -> State# d -> (# State# d, Int# #) + getSizeofSmallMutableArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> State# d -> (# State# d, Int# #) + getSpark# :: forall d a. State# d -> (# State# d, Int#, a #) + getThreadAllocationCounter# :: State# RealWorld -> (# State# RealWorld, Int64# #) + groupWith :: forall b a. GHC.Classes.Ord b => (a -> b) -> [a] -> [[a]] + gtAddr# :: Addr# -> Addr# -> Int# + gtChar# :: Char# -> Char# -> Int# + gtFloat# :: Float# -> Float# -> Int# + gtInt16# :: Int16# -> Int16# -> Int# + gtInt32# :: Int32# -> Int32# -> Int# + gtInt64# :: Int64# -> Int64# -> Int# + gtInt8# :: Int8# -> Int8# -> Int# + gtWord# :: Word# -> Word# -> Int# + gtWord16# :: Word16# -> Word16# -> Int# + gtWord32# :: Word32# -> Word32# -> Int# + gtWord64# :: Word64# -> Word64# -> Int# + gtWord8# :: Word8# -> Word8# -> Int# + iShiftL# :: Int# -> Int# -> Int# + iShiftRA# :: Int# -> Int# -> Int# + iShiftRL# :: Int# -> Int# -> Int# + indexAddrArray# :: ByteArray# -> Int# -> Addr# + indexAddrOffAddr# :: Addr# -> Int# -> Addr# + indexArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). Array# a -> Int# -> (# a #) + indexArrayArrayArray# :: ArrayArray# -> Int# -> ArrayArray# + indexByteArrayArray# :: ArrayArray# -> Int# -> ByteArray# + indexCharArray# :: ByteArray# -> Int# -> Char# + indexCharOffAddr# :: Addr# -> Int# -> Char# + indexDoubleArray# :: ByteArray# -> Int# -> Double# + indexDoubleArrayAsDoubleX2# :: ByteArray# -> Int# -> DoubleX2# + indexDoubleArrayAsDoubleX4# :: ByteArray# -> Int# -> DoubleX4# + indexDoubleArrayAsDoubleX8# :: ByteArray# -> Int# -> DoubleX8# + indexDoubleOffAddr# :: Addr# -> Int# -> Double# + indexDoubleOffAddrAsDoubleX2# :: Addr# -> Int# -> DoubleX2# + indexDoubleOffAddrAsDoubleX4# :: Addr# -> Int# -> DoubleX4# + indexDoubleOffAddrAsDoubleX8# :: Addr# -> Int# -> DoubleX8# + indexDoubleX2Array# :: ByteArray# -> Int# -> DoubleX2# + indexDoubleX2OffAddr# :: Addr# -> Int# -> DoubleX2# + indexDoubleX4Array# :: ByteArray# -> Int# -> DoubleX4# + indexDoubleX4OffAddr# :: Addr# -> Int# -> DoubleX4# + indexDoubleX8Array# :: ByteArray# -> Int# -> DoubleX8# + indexDoubleX8OffAddr# :: Addr# -> Int# -> DoubleX8# + indexFloatArray# :: ByteArray# -> Int# -> Float# + indexFloatArrayAsFloatX16# :: ByteArray# -> Int# -> FloatX16# + indexFloatArrayAsFloatX4# :: ByteArray# -> Int# -> FloatX4# + indexFloatArrayAsFloatX8# :: ByteArray# -> Int# -> FloatX8# + indexFloatOffAddr# :: Addr# -> Int# -> Float# + indexFloatOffAddrAsFloatX16# :: Addr# -> Int# -> FloatX16# + indexFloatOffAddrAsFloatX4# :: Addr# -> Int# -> FloatX4# + indexFloatOffAddrAsFloatX8# :: Addr# -> Int# -> FloatX8# + indexFloatX16Array# :: ByteArray# -> Int# -> FloatX16# + indexFloatX16OffAddr# :: Addr# -> Int# -> FloatX16# + indexFloatX4Array# :: ByteArray# -> Int# -> FloatX4# + indexFloatX4OffAddr# :: Addr# -> Int# -> FloatX4# + indexFloatX8Array# :: ByteArray# -> Int# -> FloatX8# + indexFloatX8OffAddr# :: Addr# -> Int# -> FloatX8# + indexInt16Array# :: ByteArray# -> Int# -> Int16# + indexInt16ArrayAsInt16X16# :: ByteArray# -> Int# -> Int16X16# + indexInt16ArrayAsInt16X32# :: ByteArray# -> Int# -> Int16X32# + indexInt16ArrayAsInt16X8# :: ByteArray# -> Int# -> Int16X8# + indexInt16OffAddr# :: Addr# -> Int# -> Int16# + indexInt16OffAddrAsInt16X16# :: Addr# -> Int# -> Int16X16# + indexInt16OffAddrAsInt16X32# :: Addr# -> Int# -> Int16X32# + indexInt16OffAddrAsInt16X8# :: Addr# -> Int# -> Int16X8# + indexInt16X16Array# :: ByteArray# -> Int# -> Int16X16# + indexInt16X16OffAddr# :: Addr# -> Int# -> Int16X16# + indexInt16X32Array# :: ByteArray# -> Int# -> Int16X32# + indexInt16X32OffAddr# :: Addr# -> Int# -> Int16X32# + indexInt16X8Array# :: ByteArray# -> Int# -> Int16X8# + indexInt16X8OffAddr# :: Addr# -> Int# -> Int16X8# + indexInt32Array# :: ByteArray# -> Int# -> Int32# + indexInt32ArrayAsInt32X16# :: ByteArray# -> Int# -> Int32X16# + indexInt32ArrayAsInt32X4# :: ByteArray# -> Int# -> Int32X4# + indexInt32ArrayAsInt32X8# :: ByteArray# -> Int# -> Int32X8# + indexInt32OffAddr# :: Addr# -> Int# -> Int32# + indexInt32OffAddrAsInt32X16# :: Addr# -> Int# -> Int32X16# + indexInt32OffAddrAsInt32X4# :: Addr# -> Int# -> Int32X4# + indexInt32OffAddrAsInt32X8# :: Addr# -> Int# -> Int32X8# + indexInt32X16Array# :: ByteArray# -> Int# -> Int32X16# + indexInt32X16OffAddr# :: Addr# -> Int# -> Int32X16# + indexInt32X4Array# :: ByteArray# -> Int# -> Int32X4# + indexInt32X4OffAddr# :: Addr# -> Int# -> Int32X4# + indexInt32X8Array# :: ByteArray# -> Int# -> Int32X8# + indexInt32X8OffAddr# :: Addr# -> Int# -> Int32X8# + indexInt64Array# :: ByteArray# -> Int# -> Int64# + indexInt64ArrayAsInt64X2# :: ByteArray# -> Int# -> Int64X2# + indexInt64ArrayAsInt64X4# :: ByteArray# -> Int# -> Int64X4# + indexInt64ArrayAsInt64X8# :: ByteArray# -> Int# -> Int64X8# + indexInt64OffAddr# :: Addr# -> Int# -> Int64# + indexInt64OffAddrAsInt64X2# :: Addr# -> Int# -> Int64X2# + indexInt64OffAddrAsInt64X4# :: Addr# -> Int# -> Int64X4# + indexInt64OffAddrAsInt64X8# :: Addr# -> Int# -> Int64X8# + indexInt64X2Array# :: ByteArray# -> Int# -> Int64X2# + indexInt64X2OffAddr# :: Addr# -> Int# -> Int64X2# + indexInt64X4Array# :: ByteArray# -> Int# -> Int64X4# + indexInt64X4OffAddr# :: Addr# -> Int# -> Int64X4# + indexInt64X8Array# :: ByteArray# -> Int# -> Int64X8# + indexInt64X8OffAddr# :: Addr# -> Int# -> Int64X8# + indexInt8Array# :: ByteArray# -> Int# -> Int8# + indexInt8ArrayAsInt8X16# :: ByteArray# -> Int# -> Int8X16# + indexInt8ArrayAsInt8X32# :: ByteArray# -> Int# -> Int8X32# + indexInt8ArrayAsInt8X64# :: ByteArray# -> Int# -> Int8X64# + indexInt8OffAddr# :: Addr# -> Int# -> Int8# + indexInt8OffAddrAsInt8X16# :: Addr# -> Int# -> Int8X16# + indexInt8OffAddrAsInt8X32# :: Addr# -> Int# -> Int8X32# + indexInt8OffAddrAsInt8X64# :: Addr# -> Int# -> Int8X64# + indexInt8X16Array# :: ByteArray# -> Int# -> Int8X16# + indexInt8X16OffAddr# :: Addr# -> Int# -> Int8X16# + indexInt8X32Array# :: ByteArray# -> Int# -> Int8X32# + indexInt8X32OffAddr# :: Addr# -> Int# -> Int8X32# + indexInt8X64Array# :: ByteArray# -> Int# -> Int8X64# + indexInt8X64OffAddr# :: Addr# -> Int# -> Int8X64# + indexIntArray# :: ByteArray# -> Int# -> Int# + indexIntOffAddr# :: Addr# -> Int# -> Int# + indexSmallArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). SmallArray# a -> Int# -> (# a #) + indexStablePtrArray# :: forall a. ByteArray# -> Int# -> StablePtr# a + indexStablePtrOffAddr# :: forall a. Addr# -> Int# -> StablePtr# a + indexWideCharArray# :: ByteArray# -> Int# -> Char# + indexWideCharOffAddr# :: Addr# -> Int# -> Char# + indexWord16Array# :: ByteArray# -> Int# -> Word16# + indexWord16ArrayAsWord16X16# :: ByteArray# -> Int# -> Word16X16# + indexWord16ArrayAsWord16X32# :: ByteArray# -> Int# -> Word16X32# + indexWord16ArrayAsWord16X8# :: ByteArray# -> Int# -> Word16X8# + indexWord16OffAddr# :: Addr# -> Int# -> Word16# + indexWord16OffAddrAsWord16X16# :: Addr# -> Int# -> Word16X16# + indexWord16OffAddrAsWord16X32# :: Addr# -> Int# -> Word16X32# + indexWord16OffAddrAsWord16X8# :: Addr# -> Int# -> Word16X8# + indexWord16X16Array# :: ByteArray# -> Int# -> Word16X16# + indexWord16X16OffAddr# :: Addr# -> Int# -> Word16X16# + indexWord16X32Array# :: ByteArray# -> Int# -> Word16X32# + indexWord16X32OffAddr# :: Addr# -> Int# -> Word16X32# + indexWord16X8Array# :: ByteArray# -> Int# -> Word16X8# + indexWord16X8OffAddr# :: Addr# -> Int# -> Word16X8# + indexWord32Array# :: ByteArray# -> Int# -> Word32# + indexWord32ArrayAsWord32X16# :: ByteArray# -> Int# -> Word32X16# + indexWord32ArrayAsWord32X4# :: ByteArray# -> Int# -> Word32X4# + indexWord32ArrayAsWord32X8# :: ByteArray# -> Int# -> Word32X8# + indexWord32OffAddr# :: Addr# -> Int# -> Word32# + indexWord32OffAddrAsWord32X16# :: Addr# -> Int# -> Word32X16# + indexWord32OffAddrAsWord32X4# :: Addr# -> Int# -> Word32X4# + indexWord32OffAddrAsWord32X8# :: Addr# -> Int# -> Word32X8# + indexWord32X16Array# :: ByteArray# -> Int# -> Word32X16# + indexWord32X16OffAddr# :: Addr# -> Int# -> Word32X16# + indexWord32X4Array# :: ByteArray# -> Int# -> Word32X4# + indexWord32X4OffAddr# :: Addr# -> Int# -> Word32X4# + indexWord32X8Array# :: ByteArray# -> Int# -> Word32X8# + indexWord32X8OffAddr# :: Addr# -> Int# -> Word32X8# + indexWord64Array# :: ByteArray# -> Int# -> Word64# + indexWord64ArrayAsWord64X2# :: ByteArray# -> Int# -> Word64X2# + indexWord64ArrayAsWord64X4# :: ByteArray# -> Int# -> Word64X4# + indexWord64ArrayAsWord64X8# :: ByteArray# -> Int# -> Word64X8# + indexWord64OffAddr# :: Addr# -> Int# -> Word64# + indexWord64OffAddrAsWord64X2# :: Addr# -> Int# -> Word64X2# + indexWord64OffAddrAsWord64X4# :: Addr# -> Int# -> Word64X4# + indexWord64OffAddrAsWord64X8# :: Addr# -> Int# -> Word64X8# + indexWord64X2Array# :: ByteArray# -> Int# -> Word64X2# + indexWord64X2OffAddr# :: Addr# -> Int# -> Word64X2# + indexWord64X4Array# :: ByteArray# -> Int# -> Word64X4# + indexWord64X4OffAddr# :: Addr# -> Int# -> Word64X4# + indexWord64X8Array# :: ByteArray# -> Int# -> Word64X8# + indexWord64X8OffAddr# :: Addr# -> Int# -> Word64X8# + indexWord8Array# :: ByteArray# -> Int# -> Word8# + indexWord8ArrayAsAddr# :: ByteArray# -> Int# -> Addr# + indexWord8ArrayAsChar# :: ByteArray# -> Int# -> Char# + indexWord8ArrayAsDouble# :: ByteArray# -> Int# -> Double# + indexWord8ArrayAsFloat# :: ByteArray# -> Int# -> Float# + indexWord8ArrayAsInt# :: ByteArray# -> Int# -> Int# + indexWord8ArrayAsInt16# :: ByteArray# -> Int# -> Int16# + indexWord8ArrayAsInt32# :: ByteArray# -> Int# -> Int32# + indexWord8ArrayAsInt64# :: ByteArray# -> Int# -> Int64# + indexWord8ArrayAsStablePtr# :: forall a. ByteArray# -> Int# -> StablePtr# a + indexWord8ArrayAsWideChar# :: ByteArray# -> Int# -> Char# + indexWord8ArrayAsWord# :: ByteArray# -> Int# -> Word# + indexWord8ArrayAsWord16# :: ByteArray# -> Int# -> Word16# + indexWord8ArrayAsWord32# :: ByteArray# -> Int# -> Word32# + indexWord8ArrayAsWord64# :: ByteArray# -> Int# -> Word64# + indexWord8ArrayAsWord8X16# :: ByteArray# -> Int# -> Word8X16# + indexWord8ArrayAsWord8X32# :: ByteArray# -> Int# -> Word8X32# + indexWord8ArrayAsWord8X64# :: ByteArray# -> Int# -> Word8X64# + indexWord8OffAddr# :: Addr# -> Int# -> Word8# + indexWord8OffAddrAsAddr# :: Addr# -> Int# -> Addr# + indexWord8OffAddrAsChar# :: Addr# -> Int# -> Char# + indexWord8OffAddrAsDouble# :: Addr# -> Int# -> Double# + indexWord8OffAddrAsFloat# :: Addr# -> Int# -> Float# + indexWord8OffAddrAsInt# :: Addr# -> Int# -> Int# + indexWord8OffAddrAsInt16# :: Addr# -> Int# -> Int16# + indexWord8OffAddrAsInt32# :: Addr# -> Int# -> Int32# + indexWord8OffAddrAsInt64# :: Addr# -> Int# -> Int64# + indexWord8OffAddrAsStablePtr# :: forall a. Addr# -> Int# -> StablePtr# a + indexWord8OffAddrAsWideChar# :: Addr# -> Int# -> Char# + indexWord8OffAddrAsWord# :: Addr# -> Int# -> Word# + indexWord8OffAddrAsWord16# :: Addr# -> Int# -> Word16# + indexWord8OffAddrAsWord32# :: Addr# -> Int# -> Word32# + indexWord8OffAddrAsWord64# :: Addr# -> Int# -> Word64# + indexWord8OffAddrAsWord8X16# :: Addr# -> Int# -> Word8X16# + indexWord8OffAddrAsWord8X32# :: Addr# -> Int# -> Word8X32# + indexWord8OffAddrAsWord8X64# :: Addr# -> Int# -> Word8X64# + indexWord8X16Array# :: ByteArray# -> Int# -> Word8X16# + indexWord8X16OffAddr# :: Addr# -> Int# -> Word8X16# + indexWord8X32Array# :: ByteArray# -> Int# -> Word8X32# + indexWord8X32OffAddr# :: Addr# -> Int# -> Word8X32# + indexWord8X64Array# :: ByteArray# -> Int# -> Word8X64# + indexWord8X64OffAddr# :: Addr# -> Int# -> Word8X64# + indexWordArray# :: ByteArray# -> Int# -> Word# + indexWordOffAddr# :: Addr# -> Int# -> Word# + inline :: forall a. a -> a + insertDoubleX2# :: DoubleX2# -> Double# -> Int# -> DoubleX2# + insertDoubleX4# :: DoubleX4# -> Double# -> Int# -> DoubleX4# + insertDoubleX8# :: DoubleX8# -> Double# -> Int# -> DoubleX8# + insertFloatX16# :: FloatX16# -> Float# -> Int# -> FloatX16# + insertFloatX4# :: FloatX4# -> Float# -> Int# -> FloatX4# + insertFloatX8# :: FloatX8# -> Float# -> Int# -> FloatX8# + insertInt16X16# :: Int16X16# -> Int16# -> Int# -> Int16X16# + insertInt16X32# :: Int16X32# -> Int16# -> Int# -> Int16X32# + insertInt16X8# :: Int16X8# -> Int16# -> Int# -> Int16X8# + insertInt32X16# :: Int32X16# -> Int32# -> Int# -> Int32X16# + insertInt32X4# :: Int32X4# -> Int32# -> Int# -> Int32X4# + insertInt32X8# :: Int32X8# -> Int32# -> Int# -> Int32X8# + insertInt64X2# :: Int64X2# -> Int64# -> Int# -> Int64X2# + insertInt64X4# :: Int64X4# -> Int64# -> Int# -> Int64X4# + insertInt64X8# :: Int64X8# -> Int64# -> Int# -> Int64X8# + insertInt8X16# :: Int8X16# -> Int8# -> Int# -> Int8X16# + insertInt8X32# :: Int8X32# -> Int8# -> Int# -> Int8X32# + insertInt8X64# :: Int8X64# -> Int8# -> Int# -> Int8X64# + insertWord16X16# :: Word16X16# -> Word16# -> Int# -> Word16X16# + insertWord16X32# :: Word16X32# -> Word16# -> Int# -> Word16X32# + insertWord16X8# :: Word16X8# -> Word16# -> Int# -> Word16X8# + insertWord32X16# :: Word32X16# -> Word32# -> Int# -> Word32X16# + insertWord32X4# :: Word32X4# -> Word32# -> Int# -> Word32X4# + insertWord32X8# :: Word32X8# -> Word32# -> Int# -> Word32X8# + insertWord64X2# :: Word64X2# -> Word64# -> Int# -> Word64X2# + insertWord64X4# :: Word64X4# -> Word64# -> Int# -> Word64X4# + insertWord64X8# :: Word64X8# -> Word64# -> Int# -> Word64X8# + insertWord8X16# :: Word8X16# -> Word8# -> Int# -> Word8X16# + insertWord8X32# :: Word8X32# -> Word8# -> Int# -> Word8X32# + insertWord8X64# :: Word8X64# -> Word8# -> Int# -> Word8X64# + int16ToInt# :: Int16# -> Int# + int16ToWord16# :: Int16# -> Word16# + int2Addr# :: Int# -> Addr# + int2Double# :: Int# -> Double# + int2Float# :: Int# -> Float# + int2Word# :: Int# -> Word# + int32ToInt# :: Int32# -> Int# + int32ToWord32# :: Int32# -> Word32# + int64ToInt# :: Int64# -> Int# + int64ToWord64# :: Int64# -> Word64# + int8ToInt# :: Int8# -> Int# + int8ToWord8# :: Int8# -> Word8# + intToInt16# :: Int# -> Int16# + intToInt32# :: Int# -> Int32# + intToInt64# :: Int# -> Int64# + intToInt8# :: Int# -> Int8# + isByteArrayPinned# :: ByteArray# -> Int# + isByteArrayWeaklyPinned# :: ByteArray# -> Int# + isCurrentThreadBound# :: State# RealWorld -> (# State# RealWorld, Int# #) + isEmptyMVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MVar# d a -> State# d -> (# State# d, Int# #) + isMutableByteArrayPinned# :: forall d. MutableByteArray# d -> Int# + isMutableByteArrayWeaklyPinned# :: forall d. MutableByteArray# d -> Int# + isTrue# :: Int# -> Bool + keepAlive# :: forall {l :: Levity} {r :: RuntimeRep} (a :: TYPE (BoxedRep l)) d (b :: TYPE r). a -> State# d -> (State# d -> b) -> b + killThread# :: forall a. ThreadId# -> a -> State# RealWorld -> State# RealWorld + labelThread# :: ThreadId# -> ByteArray# -> State# RealWorld -> State# RealWorld + lazy :: forall a. a -> a + leAddr# :: Addr# -> Addr# -> Int# + leChar# :: Char# -> Char# -> Int# + leFloat# :: Float# -> Float# -> Int# + leInt16# :: Int16# -> Int16# -> Int# + leInt32# :: Int32# -> Int32# -> Int# + leInt64# :: Int64# -> Int64# -> Int# + leInt8# :: Int8# -> Int8# -> Int# + leWord# :: Word# -> Word# -> Int# + leWord16# :: Word16# -> Word16# -> Int# + leWord32# :: Word32# -> Word32# -> Int# + leWord64# :: Word64# -> Word64# -> Int# + leWord8# :: Word8# -> Word8# -> Int# + leftSection :: forall {q :: RuntimeRep} {r :: RuntimeRep} (a :: TYPE q) (b :: TYPE r). (a -> b) -> a -> b + listThreads# :: State# RealWorld -> (# State# RealWorld, Array# ThreadId# #) + log1pDouble# :: Double# -> Double# + log1pFloat# :: Float# -> Float# + logDouble# :: Double# -> Double# + logFloat# :: Float# -> Float# + ltAddr# :: Addr# -> Addr# -> Int# + ltChar# :: Char# -> Char# -> Int# + ltFloat# :: Float# -> Float# -> Int# + ltInt16# :: Int16# -> Int16# -> Int# + ltInt32# :: Int32# -> Int32# -> Int# + ltInt64# :: Int64# -> Int64# -> Int# + ltInt8# :: Int8# -> Int8# -> Int# + ltWord# :: Word# -> Word# -> Int# + ltWord16# :: Word16# -> Word16# -> Int# + ltWord32# :: Word32# -> Word32# -> Int# + ltWord64# :: Word64# -> Word64# -> Int# + ltWord8# :: Word8# -> Word8# -> Int# + makeStableName# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). a -> State# RealWorld -> (# State# RealWorld, StableName# a #) + makeStablePtr# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). a -> State# RealWorld -> (# State# RealWorld, StablePtr# a #) + maskAsyncExceptions# :: forall {q :: RuntimeRep} (a :: TYPE q). (State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #) + maskUninterruptible# :: forall {q :: RuntimeRep} (a :: TYPE q). (State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #) + maxTupleSize :: Int + minusAddr# :: Addr# -> Addr# -> Int# + minusDoubleX2# :: DoubleX2# -> DoubleX2# -> DoubleX2# + minusDoubleX4# :: DoubleX4# -> DoubleX4# -> DoubleX4# + minusDoubleX8# :: DoubleX8# -> DoubleX8# -> DoubleX8# + minusFloat# :: Float# -> Float# -> Float# + minusFloatX16# :: FloatX16# -> FloatX16# -> FloatX16# + minusFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# + minusFloatX8# :: FloatX8# -> FloatX8# -> FloatX8# + minusInt16X16# :: Int16X16# -> Int16X16# -> Int16X16# + minusInt16X32# :: Int16X32# -> Int16X32# -> Int16X32# + minusInt16X8# :: Int16X8# -> Int16X8# -> Int16X8# + minusInt32X16# :: Int32X16# -> Int32X16# -> Int32X16# + minusInt32X4# :: Int32X4# -> Int32X4# -> Int32X4# + minusInt32X8# :: Int32X8# -> Int32X8# -> Int32X8# + minusInt64X2# :: Int64X2# -> Int64X2# -> Int64X2# + minusInt64X4# :: Int64X4# -> Int64X4# -> Int64X4# + minusInt64X8# :: Int64X8# -> Int64X8# -> Int64X8# + minusInt8X16# :: Int8X16# -> Int8X16# -> Int8X16# + minusInt8X32# :: Int8X32# -> Int8X32# -> Int8X32# + minusInt8X64# :: Int8X64# -> Int8X64# -> Int8X64# + minusWord# :: Word# -> Word# -> Word# + minusWord16X16# :: Word16X16# -> Word16X16# -> Word16X16# + minusWord16X32# :: Word16X32# -> Word16X32# -> Word16X32# + minusWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# + minusWord32X16# :: Word32X16# -> Word32X16# -> Word32X16# + minusWord32X4# :: Word32X4# -> Word32X4# -> Word32X4# + minusWord32X8# :: Word32X8# -> Word32X8# -> Word32X8# + minusWord64X2# :: Word64X2# -> Word64X2# -> Word64X2# + minusWord64X4# :: Word64X4# -> Word64X4# -> Word64X4# + minusWord64X8# :: Word64X8# -> Word64X8# -> Word64X8# + minusWord8X16# :: Word8X16# -> Word8X16# -> Word8X16# + minusWord8X32# :: Word8X32# -> Word8X32# -> Word8X32# + minusWord8X64# :: Word8X64# -> Word8X64# -> Word8X64# + mkApUpd0# :: forall a. BCO -> (# a #) + mkWeak# :: forall {l :: Levity} {k :: Levity} (a :: TYPE (BoxedRep l)) (b :: TYPE (BoxedRep k)) c. a -> b -> (State# RealWorld -> (# State# RealWorld, c #)) -> State# RealWorld -> (# State# RealWorld, Weak# b #) + mkWeakNoFinalizer# :: forall {l :: Levity} {k :: Levity} (a :: TYPE (BoxedRep l)) (b :: TYPE (BoxedRep k)). a -> b -> State# RealWorld -> (# State# RealWorld, Weak# b #) + mulIntMayOflo# :: Int# -> Int# -> Int# + mutableByteArrayContents# :: forall d. MutableByteArray# d -> Addr# + myThreadId# :: State# RealWorld -> (# State# RealWorld, ThreadId# #) + narrow16Int# :: Int# -> Int# + narrow16Word# :: Word# -> Word# + narrow32Int# :: Int# -> Int# + narrow32Word# :: Word# -> Word# + narrow8Int# :: Int# -> Int# + narrow8Word# :: Word# -> Word# + neAddr# :: Addr# -> Addr# -> Int# + neChar# :: Char# -> Char# -> Int# + neFloat# :: Float# -> Float# -> Int# + neInt16# :: Int16# -> Int16# -> Int# + neInt32# :: Int32# -> Int32# -> Int# + neInt64# :: Int64# -> Int64# -> Int# + neInt8# :: Int8# -> Int8# -> Int# + neWord# :: Word# -> Word# -> Int# + neWord16# :: Word16# -> Word16# -> Int# + neWord32# :: Word32# -> Word32# -> Int# + neWord64# :: Word64# -> Word64# -> Int# + neWord8# :: Word8# -> Word8# -> Int# + negateDouble# :: Double# -> Double# + negateDoubleX2# :: DoubleX2# -> DoubleX2# + negateDoubleX4# :: DoubleX4# -> DoubleX4# + negateDoubleX8# :: DoubleX8# -> DoubleX8# + negateFloat# :: Float# -> Float# + negateFloatX16# :: FloatX16# -> FloatX16# + negateFloatX4# :: FloatX4# -> FloatX4# + negateFloatX8# :: FloatX8# -> FloatX8# + negateInt# :: Int# -> Int# + negateInt16# :: Int16# -> Int16# + negateInt16X16# :: Int16X16# -> Int16X16# + negateInt16X32# :: Int16X32# -> Int16X32# + negateInt16X8# :: Int16X8# -> Int16X8# + negateInt32# :: Int32# -> Int32# + negateInt32X16# :: Int32X16# -> Int32X16# + negateInt32X4# :: Int32X4# -> Int32X4# + negateInt32X8# :: Int32X8# -> Int32X8# + negateInt64# :: Int64# -> Int64# + negateInt64X2# :: Int64X2# -> Int64X2# + negateInt64X4# :: Int64X4# -> Int64X4# + negateInt64X8# :: Int64X8# -> Int64X8# + negateInt8# :: Int8# -> Int8# + negateInt8X16# :: Int8X16# -> Int8X16# + negateInt8X32# :: Int8X32# -> Int8X32# + negateInt8X64# :: Int8X64# -> Int8X64# + newAlignedPinnedByteArray# :: forall d. Int# -> Int# -> State# d -> (# State# d, MutableByteArray# d #) + newArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. Int# -> a -> State# d -> (# State# d, MutableArray# d a #) + newArrayArray# :: forall s. Int# -> State# s -> (# State# s, MutableArrayArray# s #) + newBCO# :: forall a d. ByteArray# -> ByteArray# -> Array# a -> Int# -> ByteArray# -> State# d -> (# State# d, BCO #) + newByteArray# :: forall d. Int# -> State# d -> (# State# d, MutableByteArray# d #) + newIOPort# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). State# d -> (# State# d, IOPort# d a #) + newMVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). State# d -> (# State# d, MVar# d a #) + newMutVar# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. a -> State# d -> (# State# d, MutVar# d a #) + newPinnedByteArray# :: forall d. Int# -> State# d -> (# State# d, MutableByteArray# d #) + newPromptTag# :: forall a. State# RealWorld -> (# State# RealWorld, PromptTag# a #) + newSmallArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. Int# -> a -> State# d -> (# State# d, SmallMutableArray# d a #) + newTVar# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. a -> State# d -> (# State# d, TVar# d a #) + noDuplicate# :: forall d. State# d -> State# d + noinline :: forall a. a -> a + not# :: Word# -> Word# + not64# :: Word64# -> Word64# + notI# :: Int# -> Int# + notWord16# :: Word16# -> Word16# + notWord32# :: Word32# -> Word32# + notWord8# :: Word8# -> Word8# + nullAddr# :: Addr# + numSparks# :: forall d. State# d -> (# State# d, Int# #) + oneShot :: forall {q :: RuntimeRep} {r :: RuntimeRep} (a :: TYPE q) (b :: TYPE r). (a -> b) -> a -> b + or# :: Word# -> Word# -> Word# + or64# :: Word64# -> Word64# -> Word64# + orI# :: Int# -> Int# -> Int# + orWord16# :: Word16# -> Word16# -> Word16# + orWord32# :: Word32# -> Word32# -> Word32# + orWord8# :: Word8# -> Word8# -> Word8# + ord# :: Char# -> Int# + packDoubleX2# :: (# Double#, Double# #) -> DoubleX2# + packDoubleX4# :: (# Double#, Double#, Double#, Double# #) -> DoubleX4# + packDoubleX8# :: (# Double#, Double#, Double#, Double#, Double#, Double#, Double#, Double# #) -> DoubleX8# + packFloatX16# :: (# Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float# #) -> FloatX16# + packFloatX4# :: (# Float#, Float#, Float#, Float# #) -> FloatX4# + packFloatX8# :: (# Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float# #) -> FloatX8# + packInt16X16# :: (# Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16# #) -> Int16X16# + packInt16X32# :: (# Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16# #) -> Int16X32# + packInt16X8# :: (# Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16# #) -> Int16X8# + packInt32X16# :: (# Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32# #) -> Int32X16# + packInt32X4# :: (# Int32#, Int32#, Int32#, Int32# #) -> Int32X4# + packInt32X8# :: (# Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32# #) -> Int32X8# + packInt64X2# :: (# Int64#, Int64# #) -> Int64X2# + packInt64X4# :: (# Int64#, Int64#, Int64#, Int64# #) -> Int64X4# + packInt64X8# :: (# Int64#, Int64#, Int64#, Int64#, Int64#, Int64#, Int64#, Int64# #) -> Int64X8# + packInt8X16# :: (# Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8# #) -> Int8X16# + packInt8X32# :: (# Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8# #) -> Int8X32# + packInt8X64# :: (# Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8# #) -> Int8X64# + packWord16X16# :: (# Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16# #) -> Word16X16# + packWord16X32# :: (# Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16# #) -> Word16X32# + packWord16X8# :: (# Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16# #) -> Word16X8# + packWord32X16# :: (# Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32# #) -> Word32X16# + packWord32X4# :: (# Word32#, Word32#, Word32#, Word32# #) -> Word32X4# + packWord32X8# :: (# Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32# #) -> Word32X8# + packWord64X2# :: (# Word64#, Word64# #) -> Word64X2# + packWord64X4# :: (# Word64#, Word64#, Word64#, Word64# #) -> Word64X4# + packWord64X8# :: (# Word64#, Word64#, Word64#, Word64#, Word64#, Word64#, Word64#, Word64# #) -> Word64X8# + packWord8X16# :: (# Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8# #) -> Word8X16# + packWord8X32# :: (# Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8# #) -> Word8X32# + packWord8X64# :: (# Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8# #) -> Word8X64# + par# :: forall a. a -> Int# + pdep# :: Word# -> Word# -> Word# + pdep16# :: Word# -> Word# -> Word# + pdep32# :: Word# -> Word# -> Word# + pdep64# :: Word64# -> Word64# -> Word64# + pdep8# :: Word# -> Word# -> Word# + pext# :: Word# -> Word# -> Word# + pext16# :: Word# -> Word# -> Word# + pext32# :: Word# -> Word# -> Word# + pext64# :: Word64# -> Word64# -> Word64# + pext8# :: Word# -> Word# -> Word# + plusAddr# :: Addr# -> Int# -> Addr# + plusDoubleX2# :: DoubleX2# -> DoubleX2# -> DoubleX2# + plusDoubleX4# :: DoubleX4# -> DoubleX4# -> DoubleX4# + plusDoubleX8# :: DoubleX8# -> DoubleX8# -> DoubleX8# + plusFloat# :: Float# -> Float# -> Float# + plusFloatX16# :: FloatX16# -> FloatX16# -> FloatX16# + plusFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# + plusFloatX8# :: FloatX8# -> FloatX8# -> FloatX8# + plusInt16# :: Int16# -> Int16# -> Int16# + plusInt16X16# :: Int16X16# -> Int16X16# -> Int16X16# + plusInt16X32# :: Int16X32# -> Int16X32# -> Int16X32# + plusInt16X8# :: Int16X8# -> Int16X8# -> Int16X8# + plusInt32# :: Int32# -> Int32# -> Int32# + plusInt32X16# :: Int32X16# -> Int32X16# -> Int32X16# + plusInt32X4# :: Int32X4# -> Int32X4# -> Int32X4# + plusInt32X8# :: Int32X8# -> Int32X8# -> Int32X8# + plusInt64# :: Int64# -> Int64# -> Int64# + plusInt64X2# :: Int64X2# -> Int64X2# -> Int64X2# + plusInt64X4# :: Int64X4# -> Int64X4# -> Int64X4# + plusInt64X8# :: Int64X8# -> Int64X8# -> Int64X8# + plusInt8# :: Int8# -> Int8# -> Int8# + plusInt8X16# :: Int8X16# -> Int8X16# -> Int8X16# + plusInt8X32# :: Int8X32# -> Int8X32# -> Int8X32# + plusInt8X64# :: Int8X64# -> Int8X64# -> Int8X64# + plusWord# :: Word# -> Word# -> Word# + plusWord16# :: Word16# -> Word16# -> Word16# + plusWord16X16# :: Word16X16# -> Word16X16# -> Word16X16# + plusWord16X32# :: Word16X32# -> Word16X32# -> Word16X32# + plusWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# + plusWord2# :: Word# -> Word# -> (# Word#, Word# #) + plusWord32# :: Word32# -> Word32# -> Word32# + plusWord32X16# :: Word32X16# -> Word32X16# -> Word32X16# + plusWord32X4# :: Word32X4# -> Word32X4# -> Word32X4# + plusWord32X8# :: Word32X8# -> Word32X8# -> Word32X8# + plusWord64# :: Word64# -> Word64# -> Word64# + plusWord64X2# :: Word64X2# -> Word64X2# -> Word64X2# + plusWord64X4# :: Word64X4# -> Word64X4# -> Word64X4# + plusWord64X8# :: Word64X8# -> Word64X8# -> Word64X8# + plusWord8# :: Word8# -> Word8# -> Word8# + plusWord8X16# :: Word8X16# -> Word8X16# -> Word8X16# + plusWord8X32# :: Word8X32# -> Word8X32# -> Word8X32# + plusWord8X64# :: Word8X64# -> Word8X64# -> Word8X64# + popCnt# :: Word# -> Word# + popCnt16# :: Word# -> Word# + popCnt32# :: Word# -> Word# + popCnt64# :: Word64# -> Word# + popCnt8# :: Word# -> Word# + powerFloat# :: Float# -> Float# -> Float# + prefetchAddr0# :: forall d. Addr# -> Int# -> State# d -> State# d + prefetchAddr1# :: forall d. Addr# -> Int# -> State# d -> State# d + prefetchAddr2# :: forall d. Addr# -> Int# -> State# d -> State# d + prefetchAddr3# :: forall d. Addr# -> Int# -> State# d -> State# d + prefetchByteArray0# :: forall d. ByteArray# -> Int# -> State# d -> State# d + prefetchByteArray1# :: forall d. ByteArray# -> Int# -> State# d -> State# d + prefetchByteArray2# :: forall d. ByteArray# -> Int# -> State# d -> State# d + prefetchByteArray3# :: forall d. ByteArray# -> Int# -> State# d -> State# d + prefetchMutableByteArray0# :: forall d. MutableByteArray# d -> Int# -> State# d -> State# d + prefetchMutableByteArray1# :: forall d. MutableByteArray# d -> Int# -> State# d -> State# d + prefetchMutableByteArray2# :: forall d. MutableByteArray# d -> Int# -> State# d -> State# d + prefetchMutableByteArray3# :: forall d. MutableByteArray# d -> Int# -> State# d -> State# d + prefetchValue0# :: forall a d. a -> State# d -> State# d + prefetchValue1# :: forall a d. a -> State# d -> State# d + prefetchValue2# :: forall a d. a -> State# d -> State# d + prefetchValue3# :: forall a d. a -> State# d -> State# d + prompt# :: forall a. PromptTag# a -> (State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #) + proxy# :: forall {k} (a :: k). Proxy# a + putMVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MVar# d a -> a -> State# d -> State# d + quotInt# :: Int# -> Int# -> Int# + quotInt16# :: Int16# -> Int16# -> Int16# + quotInt16X16# :: Int16X16# -> Int16X16# -> Int16X16# + quotInt16X32# :: Int16X32# -> Int16X32# -> Int16X32# + quotInt16X8# :: Int16X8# -> Int16X8# -> Int16X8# + quotInt32# :: Int32# -> Int32# -> Int32# + quotInt32X16# :: Int32X16# -> Int32X16# -> Int32X16# + quotInt32X4# :: Int32X4# -> Int32X4# -> Int32X4# + quotInt32X8# :: Int32X8# -> Int32X8# -> Int32X8# + quotInt64# :: Int64# -> Int64# -> Int64# + quotInt64X2# :: Int64X2# -> Int64X2# -> Int64X2# + quotInt64X4# :: Int64X4# -> Int64X4# -> Int64X4# + quotInt64X8# :: Int64X8# -> Int64X8# -> Int64X8# + quotInt8# :: Int8# -> Int8# -> Int8# + quotInt8X16# :: Int8X16# -> Int8X16# -> Int8X16# + quotInt8X32# :: Int8X32# -> Int8X32# -> Int8X32# + quotInt8X64# :: Int8X64# -> Int8X64# -> Int8X64# + quotRemInt# :: Int# -> Int# -> (# Int#, Int# #) + quotRemInt16# :: Int16# -> Int16# -> (# Int16#, Int16# #) + quotRemInt32# :: Int32# -> Int32# -> (# Int32#, Int32# #) + quotRemInt8# :: Int8# -> Int8# -> (# Int8#, Int8# #) + quotRemWord# :: Word# -> Word# -> (# Word#, Word# #) + quotRemWord16# :: Word16# -> Word16# -> (# Word16#, Word16# #) + quotRemWord2# :: Word# -> Word# -> Word# -> (# Word#, Word# #) + quotRemWord32# :: Word32# -> Word32# -> (# Word32#, Word32# #) + quotRemWord8# :: Word8# -> Word8# -> (# Word8#, Word8# #) + quotWord# :: Word# -> Word# -> Word# + quotWord16# :: Word16# -> Word16# -> Word16# + quotWord16X16# :: Word16X16# -> Word16X16# -> Word16X16# + quotWord16X32# :: Word16X32# -> Word16X32# -> Word16X32# + quotWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# + quotWord32# :: Word32# -> Word32# -> Word32# + quotWord32X16# :: Word32X16# -> Word32X16# -> Word32X16# + quotWord32X4# :: Word32X4# -> Word32X4# -> Word32X4# + quotWord32X8# :: Word32X8# -> Word32X8# -> Word32X8# + quotWord64# :: Word64# -> Word64# -> Word64# + quotWord64X2# :: Word64X2# -> Word64X2# -> Word64X2# + quotWord64X4# :: Word64X4# -> Word64X4# -> Word64X4# + quotWord64X8# :: Word64X8# -> Word64X8# -> Word64X8# + quotWord8# :: Word8# -> Word8# -> Word8# + quotWord8X16# :: Word8X16# -> Word8X16# -> Word8X16# + quotWord8X32# :: Word8X32# -> Word8X32# -> Word8X32# + quotWord8X64# :: Word8X64# -> Word8X64# -> Word8X64# + raise# :: forall {l :: Levity} {r :: RuntimeRep} (a :: TYPE (BoxedRep l)) (b :: TYPE r). a -> b + raiseDivZero# :: forall {r :: RuntimeRep} (b :: TYPE r). (# #) -> b + raiseIO# :: forall {l :: Levity} {r :: RuntimeRep} (a :: TYPE (BoxedRep l)) (b :: TYPE r). a -> State# RealWorld -> (# State# RealWorld, b #) + raiseOverflow# :: forall {r :: RuntimeRep} (b :: TYPE r). (# #) -> b + raiseUnderflow# :: forall {r :: RuntimeRep} (b :: TYPE r). (# #) -> b + readAddrArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Addr# #) + readAddrOffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Addr# #) + readArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutableArray# d a -> Int# -> State# d -> (# State# d, a #) + readArrayArrayArray# :: forall s. MutableArrayArray# s -> Int# -> State# s -> (# State# s, ArrayArray# #) + readByteArrayArray# :: forall s. MutableArrayArray# s -> Int# -> State# s -> (# State# s, ByteArray# #) + readCharArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Char# #) + readCharOffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Char# #) + readDoubleArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Double# #) + readDoubleArrayAsDoubleX2# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, DoubleX2# #) + readDoubleArrayAsDoubleX4# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, DoubleX4# #) + readDoubleArrayAsDoubleX8# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, DoubleX8# #) + readDoubleOffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Double# #) + readDoubleOffAddrAsDoubleX2# :: forall d. Addr# -> Int# -> State# d -> (# State# d, DoubleX2# #) + readDoubleOffAddrAsDoubleX4# :: forall d. Addr# -> Int# -> State# d -> (# State# d, DoubleX4# #) + readDoubleOffAddrAsDoubleX8# :: forall d. Addr# -> Int# -> State# d -> (# State# d, DoubleX8# #) + readDoubleX2Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, DoubleX2# #) + readDoubleX2OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, DoubleX2# #) + readDoubleX4Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, DoubleX4# #) + readDoubleX4OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, DoubleX4# #) + readDoubleX8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, DoubleX8# #) + readDoubleX8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, DoubleX8# #) + readFloatArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Float# #) + readFloatArrayAsFloatX16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, FloatX16# #) + readFloatArrayAsFloatX4# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, FloatX4# #) + readFloatArrayAsFloatX8# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, FloatX8# #) + readFloatOffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Float# #) + readFloatOffAddrAsFloatX16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, FloatX16# #) + readFloatOffAddrAsFloatX4# :: forall d. Addr# -> Int# -> State# d -> (# State# d, FloatX4# #) + readFloatOffAddrAsFloatX8# :: forall d. Addr# -> Int# -> State# d -> (# State# d, FloatX8# #) + readFloatX16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, FloatX16# #) + readFloatX16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, FloatX16# #) + readFloatX4Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, FloatX4# #) + readFloatX4OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, FloatX4# #) + readFloatX8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, FloatX8# #) + readFloatX8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, FloatX8# #) + readIOPort# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). IOPort# d a -> State# d -> (# State# d, a #) + readInt16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int16# #) + readInt16ArrayAsInt16X16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int16X16# #) + readInt16ArrayAsInt16X32# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int16X32# #) + readInt16ArrayAsInt16X8# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int16X8# #) + readInt16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int16# #) + readInt16OffAddrAsInt16X16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int16X16# #) + readInt16OffAddrAsInt16X32# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int16X32# #) + readInt16OffAddrAsInt16X8# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int16X8# #) + readInt16X16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int16X16# #) + readInt16X16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int16X16# #) + readInt16X32Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int16X32# #) + readInt16X32OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int16X32# #) + readInt16X8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int16X8# #) + readInt16X8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int16X8# #) + readInt32Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int32# #) + readInt32ArrayAsInt32X16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int32X16# #) + readInt32ArrayAsInt32X4# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int32X4# #) + readInt32ArrayAsInt32X8# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int32X8# #) + readInt32OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int32# #) + readInt32OffAddrAsInt32X16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int32X16# #) + readInt32OffAddrAsInt32X4# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int32X4# #) + readInt32OffAddrAsInt32X8# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int32X8# #) + readInt32X16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int32X16# #) + readInt32X16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int32X16# #) + readInt32X4Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int32X4# #) + readInt32X4OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int32X4# #) + readInt32X8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int32X8# #) + readInt32X8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int32X8# #) + readInt64Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int64# #) + readInt64ArrayAsInt64X2# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int64X2# #) + readInt64ArrayAsInt64X4# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int64X4# #) + readInt64ArrayAsInt64X8# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int64X8# #) + readInt64OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int64# #) + readInt64OffAddrAsInt64X2# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int64X2# #) + readInt64OffAddrAsInt64X4# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int64X4# #) + readInt64OffAddrAsInt64X8# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int64X8# #) + readInt64X2Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int64X2# #) + readInt64X2OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int64X2# #) + readInt64X4Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int64X4# #) + readInt64X4OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int64X4# #) + readInt64X8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int64X8# #) + readInt64X8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int64X8# #) + readInt8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int8# #) + readInt8ArrayAsInt8X16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int8X16# #) + readInt8ArrayAsInt8X32# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int8X32# #) + readInt8ArrayAsInt8X64# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int8X64# #) + readInt8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int8# #) + readInt8OffAddrAsInt8X16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int8X16# #) + readInt8OffAddrAsInt8X32# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int8X32# #) + readInt8OffAddrAsInt8X64# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int8X64# #) + readInt8X16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int8X16# #) + readInt8X16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int8X16# #) + readInt8X32Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int8X32# #) + readInt8X32OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int8X32# #) + readInt8X64Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int8X64# #) + readInt8X64OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int8X64# #) + readIntArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int# #) + readIntOffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int# #) + readMVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MVar# d a -> State# d -> (# State# d, a #) + readMutVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutVar# d a -> State# d -> (# State# d, a #) + readMutableArrayArrayArray# :: forall s. MutableArrayArray# s -> Int# -> State# s -> (# State# s, MutableArrayArray# s #) + readMutableByteArrayArray# :: forall s. MutableArrayArray# s -> Int# -> State# s -> (# State# s, MutableByteArray# s #) + readSmallArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> Int# -> State# d -> (# State# d, a #) + readStablePtrArray# :: forall d a. MutableByteArray# d -> Int# -> State# d -> (# State# d, StablePtr# a #) + readStablePtrOffAddr# :: forall d a. Addr# -> Int# -> State# d -> (# State# d, StablePtr# a #) + readTVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). TVar# d a -> State# d -> (# State# d, a #) + readTVarIO# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). TVar# d a -> State# d -> (# State# d, a #) + readWideCharArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Char# #) + readWideCharOffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Char# #) + readWord16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word16# #) + readWord16ArrayAsWord16X16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word16X16# #) + readWord16ArrayAsWord16X32# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word16X32# #) + readWord16ArrayAsWord16X8# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word16X8# #) + readWord16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word16# #) + readWord16OffAddrAsWord16X16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word16X16# #) + readWord16OffAddrAsWord16X32# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word16X32# #) + readWord16OffAddrAsWord16X8# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word16X8# #) + readWord16X16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word16X16# #) + readWord16X16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word16X16# #) + readWord16X32Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word16X32# #) + readWord16X32OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word16X32# #) + readWord16X8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word16X8# #) + readWord16X8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word16X8# #) + readWord32Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word32# #) + readWord32ArrayAsWord32X16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word32X16# #) + readWord32ArrayAsWord32X4# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word32X4# #) + readWord32ArrayAsWord32X8# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word32X8# #) + readWord32OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word32# #) + readWord32OffAddrAsWord32X16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word32X16# #) + readWord32OffAddrAsWord32X4# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word32X4# #) + readWord32OffAddrAsWord32X8# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word32X8# #) + readWord32X16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word32X16# #) + readWord32X16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word32X16# #) + readWord32X4Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word32X4# #) + readWord32X4OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word32X4# #) + readWord32X8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word32X8# #) + readWord32X8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word32X8# #) + readWord64Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word64# #) + readWord64ArrayAsWord64X2# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word64X2# #) + readWord64ArrayAsWord64X4# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word64X4# #) + readWord64ArrayAsWord64X8# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word64X8# #) + readWord64OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word64# #) + readWord64OffAddrAsWord64X2# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word64X2# #) + readWord64OffAddrAsWord64X4# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word64X4# #) + readWord64OffAddrAsWord64X8# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word64X8# #) + readWord64X2Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word64X2# #) + readWord64X2OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word64X2# #) + readWord64X4Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word64X4# #) + readWord64X4OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word64X4# #) + readWord64X8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word64X8# #) + readWord64X8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word64X8# #) + readWord8Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word8# #) + readWord8ArrayAsAddr# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Addr# #) + readWord8ArrayAsChar# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Char# #) + readWord8ArrayAsDouble# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Double# #) + readWord8ArrayAsFloat# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Float# #) + readWord8ArrayAsInt# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int# #) + readWord8ArrayAsInt16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int16# #) + readWord8ArrayAsInt32# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int32# #) + readWord8ArrayAsInt64# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Int64# #) + readWord8ArrayAsStablePtr# :: forall d a. MutableByteArray# d -> Int# -> State# d -> (# State# d, StablePtr# a #) + readWord8ArrayAsWideChar# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Char# #) + readWord8ArrayAsWord# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word# #) + readWord8ArrayAsWord16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word16# #) + readWord8ArrayAsWord32# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word32# #) + readWord8ArrayAsWord64# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word64# #) + readWord8ArrayAsWord8X16# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word8X16# #) + readWord8ArrayAsWord8X32# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word8X32# #) + readWord8ArrayAsWord8X64# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word8X64# #) + readWord8OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word8# #) + readWord8OffAddrAsAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Addr# #) + readWord8OffAddrAsChar# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Char# #) + readWord8OffAddrAsDouble# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Double# #) + readWord8OffAddrAsFloat# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Float# #) + readWord8OffAddrAsInt# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int# #) + readWord8OffAddrAsInt16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int16# #) + readWord8OffAddrAsInt32# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int32# #) + readWord8OffAddrAsInt64# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Int64# #) + readWord8OffAddrAsStablePtr# :: forall d a. Addr# -> Int# -> State# d -> (# State# d, StablePtr# a #) + readWord8OffAddrAsWideChar# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Char# #) + readWord8OffAddrAsWord# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word# #) + readWord8OffAddrAsWord16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word16# #) + readWord8OffAddrAsWord32# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word32# #) + readWord8OffAddrAsWord64# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word64# #) + readWord8OffAddrAsWord8X16# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word8X16# #) + readWord8OffAddrAsWord8X32# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word8X32# #) + readWord8OffAddrAsWord8X64# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word8X64# #) + readWord8X16Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word8X16# #) + readWord8X16OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word8X16# #) + readWord8X32Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word8X32# #) + readWord8X32OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word8X32# #) + readWord8X64Array# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word8X64# #) + readWord8X64OffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word8X64# #) + readWordArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, Word# #) + readWordOffAddr# :: forall d. Addr# -> Int# -> State# d -> (# State# d, Word# #) + realWorld# :: State# RealWorld + reallyUnsafePtrEquality :: forall a. a -> a -> Int# + reallyUnsafePtrEquality# :: forall {l :: Levity} {k :: Levity} (a :: TYPE (BoxedRep l)) (b :: TYPE (BoxedRep k)). a -> b -> Int# + remAddr# :: Addr# -> Int# -> Int# + remInt# :: Int# -> Int# -> Int# + remInt16# :: Int16# -> Int16# -> Int16# + remInt16X16# :: Int16X16# -> Int16X16# -> Int16X16# + remInt16X32# :: Int16X32# -> Int16X32# -> Int16X32# + remInt16X8# :: Int16X8# -> Int16X8# -> Int16X8# + remInt32# :: Int32# -> Int32# -> Int32# + remInt32X16# :: Int32X16# -> Int32X16# -> Int32X16# + remInt32X4# :: Int32X4# -> Int32X4# -> Int32X4# + remInt32X8# :: Int32X8# -> Int32X8# -> Int32X8# + remInt64# :: Int64# -> Int64# -> Int64# + remInt64X2# :: Int64X2# -> Int64X2# -> Int64X2# + remInt64X4# :: Int64X4# -> Int64X4# -> Int64X4# + remInt64X8# :: Int64X8# -> Int64X8# -> Int64X8# + remInt8# :: Int8# -> Int8# -> Int8# + remInt8X16# :: Int8X16# -> Int8X16# -> Int8X16# + remInt8X32# :: Int8X32# -> Int8X32# -> Int8X32# + remInt8X64# :: Int8X64# -> Int8X64# -> Int8X64# + remWord# :: Word# -> Word# -> Word# + remWord16# :: Word16# -> Word16# -> Word16# + remWord16X16# :: Word16X16# -> Word16X16# -> Word16X16# + remWord16X32# :: Word16X32# -> Word16X32# -> Word16X32# + remWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# + remWord32# :: Word32# -> Word32# -> Word32# + remWord32X16# :: Word32X16# -> Word32X16# -> Word32X16# + remWord32X4# :: Word32X4# -> Word32X4# -> Word32X4# + remWord32X8# :: Word32X8# -> Word32X8# -> Word32X8# + remWord64# :: Word64# -> Word64# -> Word64# + remWord64X2# :: Word64X2# -> Word64X2# -> Word64X2# + remWord64X4# :: Word64X4# -> Word64X4# -> Word64X4# + remWord64X8# :: Word64X8# -> Word64X8# -> Word64X8# + remWord8# :: Word8# -> Word8# -> Word8# + remWord8X16# :: Word8X16# -> Word8X16# -> Word8X16# + remWord8X32# :: Word8X32# -> Word8X32# -> Word8X32# + remWord8X64# :: Word8X64# -> Word8X64# -> Word8X64# + resizeMutableByteArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> (# State# d, MutableByteArray# d #) + resizeSmallMutableArray# :: forall s a. SmallMutableArray# s a -> Int# -> a -> State# s -> (# State# s, SmallMutableArray# s a #) + retry# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). State# RealWorld -> (# State# RealWorld, a #) + rightSection :: forall {q :: RuntimeRep} {r :: RuntimeRep} {s :: RuntimeRep} (a :: TYPE q) (b :: TYPE r) (c :: TYPE s). (a -> b -> c) -> b -> a -> c + runRW# :: forall (r :: RuntimeRep) (o :: TYPE r). (State# RealWorld -> o) -> o + sameArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). Array# a -> Array# a -> Int# + sameArrayArray# :: ArrayArray# -> ArrayArray# -> Int# + sameByteArray# :: ByteArray# -> ByteArray# -> Int# + sameIOPort# :: forall {l :: Levity} s (a :: TYPE (BoxedRep l)). IOPort# s a -> IOPort# s a -> Int# + sameMVar# :: forall {l :: Levity} s (a :: TYPE (BoxedRep l)). MVar# s a -> MVar# s a -> Int# + sameMutVar# :: forall {l :: Levity} s (a :: TYPE (BoxedRep l)). MutVar# s a -> MutVar# s a -> Int# + sameMutableArray# :: forall {l :: Levity} s (a :: TYPE (BoxedRep l)). MutableArray# s a -> MutableArray# s a -> Int# + sameMutableArrayArray# :: forall s. MutableArrayArray# s -> MutableArrayArray# s -> Int# + sameMutableByteArray# :: forall s. MutableByteArray# s -> MutableByteArray# s -> Int# + samePromptTag# :: forall a. PromptTag# a -> PromptTag# a -> Int# + sameSmallArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). SmallArray# a -> SmallArray# a -> Int# + sameSmallMutableArray# :: forall {l :: Levity} s (a :: TYPE (BoxedRep l)). SmallMutableArray# s a -> SmallMutableArray# s a -> Int# + sameTVar# :: forall {l :: Levity} s (a :: TYPE (BoxedRep l)). TVar# s a -> TVar# s a -> Int# + seq :: forall {r :: RuntimeRep} a (b :: TYPE r). a -> b -> b + seq# :: forall a s. a -> State# s -> (# State# s, a #) + setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld + setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld + shiftL# :: Word# -> Int# -> Word# + shiftRL# :: Word# -> Int# -> Word# + shrinkMutableByteArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> State# d + shrinkSmallMutableArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> Int# -> State# d -> State# d + sinDouble# :: Double# -> Double# + sinFloat# :: Float# -> Float# + sinhDouble# :: Double# -> Double# + sinhFloat# :: Float# -> Float# + sizeofArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). Array# a -> Int# + sizeofArrayArray# :: ArrayArray# -> Int# + sizeofByteArray# :: ByteArray# -> Int# + sizeofMutableArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutableArray# d a -> Int# + sizeofMutableArrayArray# :: forall s. MutableArrayArray# s -> Int# + sizeofMutableByteArray# :: forall d. MutableByteArray# d -> Int# + sizeofSmallArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). SmallArray# a -> Int# + sizeofSmallMutableArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> Int# + sortWith :: forall b a. GHC.Classes.Ord b => (a -> b) -> [a] -> [a] + spark# :: forall a d. a -> State# d -> (# State# d, a #) + sqrtDouble# :: Double# -> Double# + sqrtFloat# :: Float# -> Float# + stableNameToInt# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)). StableName# a -> Int# + subInt16# :: Int16# -> Int16# -> Int16# + subInt32# :: Int32# -> Int32# -> Int32# + subInt64# :: Int64# -> Int64# -> Int64# + subInt8# :: Int8# -> Int8# -> Int8# + subIntC# :: Int# -> Int# -> (# Int#, Int# #) + subWord16# :: Word16# -> Word16# -> Word16# + subWord32# :: Word32# -> Word32# -> Word32# + subWord64# :: Word64# -> Word64# -> Word64# + subWord8# :: Word8# -> Word8# -> Word8# + subWordC# :: Word# -> Word# -> (# Word#, Int# #) + tagToEnum# :: forall a. Int# -> a + takeMVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MVar# d a -> State# d -> (# State# d, a #) + tanDouble# :: Double# -> Double# + tanFloat# :: Float# -> Float# + tanhDouble# :: Double# -> Double# + tanhFloat# :: Float# -> Float# + thawArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. Array# a -> Int# -> Int# -> State# d -> (# State# d, MutableArray# d a #) + thawSmallArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. SmallArray# a -> Int# -> Int# -> State# d -> (# State# d, SmallMutableArray# d a #) + the :: forall a. GHC.Classes.Eq a => [a] -> a + threadLabel# :: ThreadId# -> State# RealWorld -> (# State# RealWorld, Int#, ByteArray# #) + threadStatus# :: ThreadId# -> State# RealWorld -> (# State# RealWorld, Int#, Int#, Int# #) + timesDoubleX2# :: DoubleX2# -> DoubleX2# -> DoubleX2# + timesDoubleX4# :: DoubleX4# -> DoubleX4# -> DoubleX4# + timesDoubleX8# :: DoubleX8# -> DoubleX8# -> DoubleX8# + timesFloat# :: Float# -> Float# -> Float# + timesFloatX16# :: FloatX16# -> FloatX16# -> FloatX16# + timesFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# + timesFloatX8# :: FloatX8# -> FloatX8# -> FloatX8# + timesInt16# :: Int16# -> Int16# -> Int16# + timesInt16X16# :: Int16X16# -> Int16X16# -> Int16X16# + timesInt16X32# :: Int16X32# -> Int16X32# -> Int16X32# + timesInt16X8# :: Int16X8# -> Int16X8# -> Int16X8# + timesInt2# :: Int# -> Int# -> (# Int#, Int#, Int# #) + timesInt32# :: Int32# -> Int32# -> Int32# + timesInt32X16# :: Int32X16# -> Int32X16# -> Int32X16# + timesInt32X4# :: Int32X4# -> Int32X4# -> Int32X4# + timesInt32X8# :: Int32X8# -> Int32X8# -> Int32X8# + timesInt64# :: Int64# -> Int64# -> Int64# + timesInt64X2# :: Int64X2# -> Int64X2# -> Int64X2# + timesInt64X4# :: Int64X4# -> Int64X4# -> Int64X4# + timesInt64X8# :: Int64X8# -> Int64X8# -> Int64X8# + timesInt8# :: Int8# -> Int8# -> Int8# + timesInt8X16# :: Int8X16# -> Int8X16# -> Int8X16# + timesInt8X32# :: Int8X32# -> Int8X32# -> Int8X32# + timesInt8X64# :: Int8X64# -> Int8X64# -> Int8X64# + timesWord# :: Word# -> Word# -> Word# + timesWord16# :: Word16# -> Word16# -> Word16# + timesWord16X16# :: Word16X16# -> Word16X16# -> Word16X16# + timesWord16X32# :: Word16X32# -> Word16X32# -> Word16X32# + timesWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# + timesWord2# :: Word# -> Word# -> (# Word#, Word# #) + timesWord32# :: Word32# -> Word32# -> Word32# + timesWord32X16# :: Word32X16# -> Word32X16# -> Word32X16# + timesWord32X4# :: Word32X4# -> Word32X4# -> Word32X4# + timesWord32X8# :: Word32X8# -> Word32X8# -> Word32X8# + timesWord64# :: Word64# -> Word64# -> Word64# + timesWord64X2# :: Word64X2# -> Word64X2# -> Word64X2# + timesWord64X4# :: Word64X4# -> Word64X4# -> Word64X4# + timesWord64X8# :: Word64X8# -> Word64X8# -> Word64X8# + timesWord8# :: Word8# -> Word8# -> Word8# + timesWord8X16# :: Word8X16# -> Word8X16# -> Word8X16# + timesWord8X32# :: Word8X32# -> Word8X32# -> Word8X32# + timesWord8X64# :: Word8X64# -> Word8X64# -> Word8X64# + touch# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. a -> State# d -> State# d + traceBinaryEvent# :: forall d. Addr# -> Int# -> State# d -> State# d + traceEvent :: GHC.Internal.Base.String -> GHC.Types.IO () + traceEvent# :: forall d. Addr# -> State# d -> State# d + traceMarker# :: forall d. Addr# -> State# d -> State# d + tryPutMVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MVar# d a -> a -> State# d -> (# State# d, Int# #) + tryReadMVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MVar# d a -> State# d -> (# State# d, Int#, a #) + tryTakeMVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MVar# d a -> State# d -> (# State# d, Int#, a #) + uncheckedIShiftL# :: Int# -> Int# -> Int# + uncheckedIShiftL64# :: Int64# -> Int# -> Int64# + uncheckedIShiftRA# :: Int# -> Int# -> Int# + uncheckedIShiftRA64# :: Int64# -> Int# -> Int64# + uncheckedIShiftRL# :: Int# -> Int# -> Int# + uncheckedIShiftRL64# :: Int64# -> Int# -> Int64# + uncheckedShiftL# :: Word# -> Int# -> Word# + uncheckedShiftL64# :: Word64# -> Int# -> Word64# + uncheckedShiftLInt16# :: Int16# -> Int# -> Int16# + uncheckedShiftLInt32# :: Int32# -> Int# -> Int32# + uncheckedShiftLInt8# :: Int8# -> Int# -> Int8# + uncheckedShiftLWord16# :: Word16# -> Int# -> Word16# + uncheckedShiftLWord32# :: Word32# -> Int# -> Word32# + uncheckedShiftLWord8# :: Word8# -> Int# -> Word8# + uncheckedShiftRAInt16# :: Int16# -> Int# -> Int16# + uncheckedShiftRAInt32# :: Int32# -> Int# -> Int32# + uncheckedShiftRAInt8# :: Int8# -> Int# -> Int8# + uncheckedShiftRL# :: Word# -> Int# -> Word# + uncheckedShiftRL64# :: Word64# -> Int# -> Word64# + uncheckedShiftRLInt16# :: Int16# -> Int# -> Int16# + uncheckedShiftRLInt32# :: Int32# -> Int# -> Int32# + uncheckedShiftRLInt8# :: Int8# -> Int# -> Int8# + uncheckedShiftRLWord16# :: Word16# -> Int# -> Word16# + uncheckedShiftRLWord32# :: Word32# -> Int# -> Word32# + uncheckedShiftRLWord8# :: Word8# -> Int# -> Word8# + unmaskAsyncExceptions# :: forall {q :: RuntimeRep} (a :: TYPE q). (State# RealWorld -> (# State# RealWorld, a #)) -> State# RealWorld -> (# State# RealWorld, a #) + unpackAppendCString# :: Addr# -> [Char] -> [Char] + unpackCString# :: Addr# -> [Char] + unpackCStringUtf8# :: Addr# -> [Char] + unpackClosure# :: forall a b. a -> (# Addr#, ByteArray#, Array# b #) + unpackDoubleX2# :: DoubleX2# -> (# Double#, Double# #) + unpackDoubleX4# :: DoubleX4# -> (# Double#, Double#, Double#, Double# #) + unpackDoubleX8# :: DoubleX8# -> (# Double#, Double#, Double#, Double#, Double#, Double#, Double#, Double# #) + unpackFloatX16# :: FloatX16# -> (# Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float# #) + unpackFloatX4# :: FloatX4# -> (# Float#, Float#, Float#, Float# #) + unpackFloatX8# :: FloatX8# -> (# Float#, Float#, Float#, Float#, Float#, Float#, Float#, Float# #) + unpackFoldrCString# :: forall a. Addr# -> (Char -> a -> a) -> a -> a + unpackInt16X16# :: Int16X16# -> (# Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16# #) + unpackInt16X32# :: Int16X32# -> (# Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16# #) + unpackInt16X8# :: Int16X8# -> (# Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16#, Int16# #) + unpackInt32X16# :: Int32X16# -> (# Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32# #) + unpackInt32X4# :: Int32X4# -> (# Int32#, Int32#, Int32#, Int32# #) + unpackInt32X8# :: Int32X8# -> (# Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32#, Int32# #) + unpackInt64X2# :: Int64X2# -> (# Int64#, Int64# #) + unpackInt64X4# :: Int64X4# -> (# Int64#, Int64#, Int64#, Int64# #) + unpackInt64X8# :: Int64X8# -> (# Int64#, Int64#, Int64#, Int64#, Int64#, Int64#, Int64#, Int64# #) + unpackInt8X16# :: Int8X16# -> (# Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8# #) + unpackInt8X32# :: Int8X32# -> (# Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8# #) + unpackInt8X64# :: Int8X64# -> (# Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8#, Int8# #) + unpackNBytes# :: Addr# -> Int# -> [Char] + unpackWord16X16# :: Word16X16# -> (# Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16# #) + unpackWord16X32# :: Word16X32# -> (# Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16# #) + unpackWord16X8# :: Word16X8# -> (# Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16#, Word16# #) + unpackWord32X16# :: Word32X16# -> (# Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32# #) + unpackWord32X4# :: Word32X4# -> (# Word32#, Word32#, Word32#, Word32# #) + unpackWord32X8# :: Word32X8# -> (# Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32#, Word32# #) + unpackWord64X2# :: Word64X2# -> (# Word64#, Word64# #) + unpackWord64X4# :: Word64X4# -> (# Word64#, Word64#, Word64#, Word64# #) + unpackWord64X8# :: Word64X8# -> (# Word64#, Word64#, Word64#, Word64#, Word64#, Word64#, Word64#, Word64# #) + unpackWord8X16# :: Word8X16# -> (# Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8# #) + unpackWord8X32# :: Word8X32# -> (# Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8# #) + unpackWord8X64# :: Word8X64# -> (# Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8#, Word8# #) + unsafeCoerce# :: forall (q :: RuntimeRep) (r :: RuntimeRep) (a :: TYPE q) (b :: TYPE r). a -> b + unsafeFreezeArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutableArray# d a -> State# d -> (# State# d, Array# a #) + unsafeFreezeArrayArray# :: forall s. MutableArrayArray# s -> State# s -> (# State# s, ArrayArray# #) + unsafeFreezeByteArray# :: forall d. MutableByteArray# d -> State# d -> (# State# d, ByteArray# #) + unsafeFreezeSmallArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> State# d -> (# State# d, SmallArray# a #) + unsafePtrEquality# :: forall (a :: UnliftedType) (b :: UnliftedType). a -> b -> Int# + unsafeThawArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. Array# a -> State# d -> (# State# d, MutableArray# d a #) + unsafeThawByteArray# :: forall d. ByteArray# -> State# d -> (# State# d, MutableByteArray# d #) + unsafeThawSmallArray# :: forall {l :: Levity} (a :: TYPE (BoxedRep l)) d. SmallArray# a -> State# d -> (# State# d, SmallMutableArray# d a #) + void# :: (# #) + waitRead# :: forall d. Int# -> State# d -> State# d + waitWrite# :: forall d. Int# -> State# d -> State# d + word16ToInt16# :: Word16# -> Int16# + word16ToWord# :: Word16# -> Word# + word2Double# :: Word# -> Double# + word2Float# :: Word# -> Float# + word2Int# :: Word# -> Int# + word32ToInt32# :: Word32# -> Int32# + word32ToWord# :: Word32# -> Word# + word64ToInt64# :: Word64# -> Int64# + word64ToWord# :: Word64# -> Word# + word8ToInt8# :: Word8# -> Int8# + word8ToWord# :: Word8# -> Word# + wordToWord16# :: Word# -> Word16# + wordToWord32# :: Word# -> Word32# + wordToWord64# :: Word# -> Word64# + wordToWord8# :: Word# -> Word8# + writeAddrArray# :: forall d. MutableByteArray# d -> Int# -> Addr# -> State# d -> State# d + writeAddrOffAddr# :: forall d. Addr# -> Int# -> Addr# -> State# d -> State# d + writeArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutableArray# d a -> Int# -> a -> State# d -> State# d + writeArrayArrayArray# :: forall s. MutableArrayArray# s -> Int# -> ArrayArray# -> State# s -> State# s + writeByteArrayArray# :: forall s. MutableArrayArray# s -> Int# -> ByteArray# -> State# s -> State# s + writeCharArray# :: forall d. MutableByteArray# d -> Int# -> Char# -> State# d -> State# d + writeCharOffAddr# :: forall d. Addr# -> Int# -> Char# -> State# d -> State# d + writeDoubleArray# :: forall d. MutableByteArray# d -> Int# -> Double# -> State# d -> State# d + writeDoubleArrayAsDoubleX2# :: forall d. MutableByteArray# d -> Int# -> DoubleX2# -> State# d -> State# d + writeDoubleArrayAsDoubleX4# :: forall d. MutableByteArray# d -> Int# -> DoubleX4# -> State# d -> State# d + writeDoubleArrayAsDoubleX8# :: forall d. MutableByteArray# d -> Int# -> DoubleX8# -> State# d -> State# d + writeDoubleOffAddr# :: forall d. Addr# -> Int# -> Double# -> State# d -> State# d + writeDoubleOffAddrAsDoubleX2# :: forall d. Addr# -> Int# -> DoubleX2# -> State# d -> State# d + writeDoubleOffAddrAsDoubleX4# :: forall d. Addr# -> Int# -> DoubleX4# -> State# d -> State# d + writeDoubleOffAddrAsDoubleX8# :: forall d. Addr# -> Int# -> DoubleX8# -> State# d -> State# d + writeDoubleX2Array# :: forall d. MutableByteArray# d -> Int# -> DoubleX2# -> State# d -> State# d + writeDoubleX2OffAddr# :: forall d. Addr# -> Int# -> DoubleX2# -> State# d -> State# d + writeDoubleX4Array# :: forall d. MutableByteArray# d -> Int# -> DoubleX4# -> State# d -> State# d + writeDoubleX4OffAddr# :: forall d. Addr# -> Int# -> DoubleX4# -> State# d -> State# d + writeDoubleX8Array# :: forall d. MutableByteArray# d -> Int# -> DoubleX8# -> State# d -> State# d + writeDoubleX8OffAddr# :: forall d. Addr# -> Int# -> DoubleX8# -> State# d -> State# d + writeFloatArray# :: forall d. MutableByteArray# d -> Int# -> Float# -> State# d -> State# d + writeFloatArrayAsFloatX16# :: forall d. MutableByteArray# d -> Int# -> FloatX16# -> State# d -> State# d + writeFloatArrayAsFloatX4# :: forall d. MutableByteArray# d -> Int# -> FloatX4# -> State# d -> State# d + writeFloatArrayAsFloatX8# :: forall d. MutableByteArray# d -> Int# -> FloatX8# -> State# d -> State# d + writeFloatOffAddr# :: forall d. Addr# -> Int# -> Float# -> State# d -> State# d + writeFloatOffAddrAsFloatX16# :: forall d. Addr# -> Int# -> FloatX16# -> State# d -> State# d + writeFloatOffAddrAsFloatX4# :: forall d. Addr# -> Int# -> FloatX4# -> State# d -> State# d + writeFloatOffAddrAsFloatX8# :: forall d. Addr# -> Int# -> FloatX8# -> State# d -> State# d + writeFloatX16Array# :: forall d. MutableByteArray# d -> Int# -> FloatX16# -> State# d -> State# d + writeFloatX16OffAddr# :: forall d. Addr# -> Int# -> FloatX16# -> State# d -> State# d + writeFloatX4Array# :: forall d. MutableByteArray# d -> Int# -> FloatX4# -> State# d -> State# d + writeFloatX4OffAddr# :: forall d. Addr# -> Int# -> FloatX4# -> State# d -> State# d + writeFloatX8Array# :: forall d. MutableByteArray# d -> Int# -> FloatX8# -> State# d -> State# d + writeFloatX8OffAddr# :: forall d. Addr# -> Int# -> FloatX8# -> State# d -> State# d + writeIOPort# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). IOPort# d a -> a -> State# d -> (# State# d, Int# #) + writeInt16Array# :: forall d. MutableByteArray# d -> Int# -> Int16# -> State# d -> State# d + writeInt16ArrayAsInt16X16# :: forall d. MutableByteArray# d -> Int# -> Int16X16# -> State# d -> State# d + writeInt16ArrayAsInt16X32# :: forall d. MutableByteArray# d -> Int# -> Int16X32# -> State# d -> State# d + writeInt16ArrayAsInt16X8# :: forall d. MutableByteArray# d -> Int# -> Int16X8# -> State# d -> State# d + writeInt16OffAddr# :: forall d. Addr# -> Int# -> Int16# -> State# d -> State# d + writeInt16OffAddrAsInt16X16# :: forall d. Addr# -> Int# -> Int16X16# -> State# d -> State# d + writeInt16OffAddrAsInt16X32# :: forall d. Addr# -> Int# -> Int16X32# -> State# d -> State# d + writeInt16OffAddrAsInt16X8# :: forall d. Addr# -> Int# -> Int16X8# -> State# d -> State# d + writeInt16X16Array# :: forall d. MutableByteArray# d -> Int# -> Int16X16# -> State# d -> State# d + writeInt16X16OffAddr# :: forall d. Addr# -> Int# -> Int16X16# -> State# d -> State# d + writeInt16X32Array# :: forall d. MutableByteArray# d -> Int# -> Int16X32# -> State# d -> State# d + writeInt16X32OffAddr# :: forall d. Addr# -> Int# -> Int16X32# -> State# d -> State# d + writeInt16X8Array# :: forall d. MutableByteArray# d -> Int# -> Int16X8# -> State# d -> State# d + writeInt16X8OffAddr# :: forall d. Addr# -> Int# -> Int16X8# -> State# d -> State# d + writeInt32Array# :: forall d. MutableByteArray# d -> Int# -> Int32# -> State# d -> State# d + writeInt32ArrayAsInt32X16# :: forall d. MutableByteArray# d -> Int# -> Int32X16# -> State# d -> State# d + writeInt32ArrayAsInt32X4# :: forall d. MutableByteArray# d -> Int# -> Int32X4# -> State# d -> State# d + writeInt32ArrayAsInt32X8# :: forall d. MutableByteArray# d -> Int# -> Int32X8# -> State# d -> State# d + writeInt32OffAddr# :: forall d. Addr# -> Int# -> Int32# -> State# d -> State# d + writeInt32OffAddrAsInt32X16# :: forall d. Addr# -> Int# -> Int32X16# -> State# d -> State# d + writeInt32OffAddrAsInt32X4# :: forall d. Addr# -> Int# -> Int32X4# -> State# d -> State# d + writeInt32OffAddrAsInt32X8# :: forall d. Addr# -> Int# -> Int32X8# -> State# d -> State# d + writeInt32X16Array# :: forall d. MutableByteArray# d -> Int# -> Int32X16# -> State# d -> State# d + writeInt32X16OffAddr# :: forall d. Addr# -> Int# -> Int32X16# -> State# d -> State# d + writeInt32X4Array# :: forall d. MutableByteArray# d -> Int# -> Int32X4# -> State# d -> State# d + writeInt32X4OffAddr# :: forall d. Addr# -> Int# -> Int32X4# -> State# d -> State# d + writeInt32X8Array# :: forall d. MutableByteArray# d -> Int# -> Int32X8# -> State# d -> State# d + writeInt32X8OffAddr# :: forall d. Addr# -> Int# -> Int32X8# -> State# d -> State# d + writeInt64Array# :: forall d. MutableByteArray# d -> Int# -> Int64# -> State# d -> State# d + writeInt64ArrayAsInt64X2# :: forall d. MutableByteArray# d -> Int# -> Int64X2# -> State# d -> State# d + writeInt64ArrayAsInt64X4# :: forall d. MutableByteArray# d -> Int# -> Int64X4# -> State# d -> State# d + writeInt64ArrayAsInt64X8# :: forall d. MutableByteArray# d -> Int# -> Int64X8# -> State# d -> State# d + writeInt64OffAddr# :: forall d. Addr# -> Int# -> Int64# -> State# d -> State# d + writeInt64OffAddrAsInt64X2# :: forall d. Addr# -> Int# -> Int64X2# -> State# d -> State# d + writeInt64OffAddrAsInt64X4# :: forall d. Addr# -> Int# -> Int64X4# -> State# d -> State# d + writeInt64OffAddrAsInt64X8# :: forall d. Addr# -> Int# -> Int64X8# -> State# d -> State# d + writeInt64X2Array# :: forall d. MutableByteArray# d -> Int# -> Int64X2# -> State# d -> State# d + writeInt64X2OffAddr# :: forall d. Addr# -> Int# -> Int64X2# -> State# d -> State# d + writeInt64X4Array# :: forall d. MutableByteArray# d -> Int# -> Int64X4# -> State# d -> State# d + writeInt64X4OffAddr# :: forall d. Addr# -> Int# -> Int64X4# -> State# d -> State# d + writeInt64X8Array# :: forall d. MutableByteArray# d -> Int# -> Int64X8# -> State# d -> State# d + writeInt64X8OffAddr# :: forall d. Addr# -> Int# -> Int64X8# -> State# d -> State# d + writeInt8Array# :: forall d. MutableByteArray# d -> Int# -> Int8# -> State# d -> State# d + writeInt8ArrayAsInt8X16# :: forall d. MutableByteArray# d -> Int# -> Int8X16# -> State# d -> State# d + writeInt8ArrayAsInt8X32# :: forall d. MutableByteArray# d -> Int# -> Int8X32# -> State# d -> State# d + writeInt8ArrayAsInt8X64# :: forall d. MutableByteArray# d -> Int# -> Int8X64# -> State# d -> State# d + writeInt8OffAddr# :: forall d. Addr# -> Int# -> Int8# -> State# d -> State# d + writeInt8OffAddrAsInt8X16# :: forall d. Addr# -> Int# -> Int8X16# -> State# d -> State# d + writeInt8OffAddrAsInt8X32# :: forall d. Addr# -> Int# -> Int8X32# -> State# d -> State# d + writeInt8OffAddrAsInt8X64# :: forall d. Addr# -> Int# -> Int8X64# -> State# d -> State# d + writeInt8X16Array# :: forall d. MutableByteArray# d -> Int# -> Int8X16# -> State# d -> State# d + writeInt8X16OffAddr# :: forall d. Addr# -> Int# -> Int8X16# -> State# d -> State# d + writeInt8X32Array# :: forall d. MutableByteArray# d -> Int# -> Int8X32# -> State# d -> State# d + writeInt8X32OffAddr# :: forall d. Addr# -> Int# -> Int8X32# -> State# d -> State# d + writeInt8X64Array# :: forall d. MutableByteArray# d -> Int# -> Int8X64# -> State# d -> State# d + writeInt8X64OffAddr# :: forall d. Addr# -> Int# -> Int8X64# -> State# d -> State# d + writeIntArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> State# d + writeIntOffAddr# :: forall d. Addr# -> Int# -> Int# -> State# d -> State# d + writeMutVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). MutVar# d a -> a -> State# d -> State# d + writeMutableArrayArrayArray# :: forall s. MutableArrayArray# s -> Int# -> MutableArrayArray# s -> State# s -> State# s + writeMutableByteArrayArray# :: forall s. MutableArrayArray# s -> Int# -> MutableByteArray# s -> State# s -> State# s + writeSmallArray# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). SmallMutableArray# d a -> Int# -> a -> State# d -> State# d + writeStablePtrArray# :: forall d a. MutableByteArray# d -> Int# -> StablePtr# a -> State# d -> State# d + writeStablePtrOffAddr# :: forall a d. Addr# -> Int# -> StablePtr# a -> State# d -> State# d + writeTVar# :: forall {l :: Levity} d (a :: TYPE (BoxedRep l)). TVar# d a -> a -> State# d -> State# d + writeWideCharArray# :: forall d. MutableByteArray# d -> Int# -> Char# -> State# d -> State# d + writeWideCharOffAddr# :: forall d. Addr# -> Int# -> Char# -> State# d -> State# d + writeWord16Array# :: forall d. MutableByteArray# d -> Int# -> Word16# -> State# d -> State# d + writeWord16ArrayAsWord16X16# :: forall d. MutableByteArray# d -> Int# -> Word16X16# -> State# d -> State# d + writeWord16ArrayAsWord16X32# :: forall d. MutableByteArray# d -> Int# -> Word16X32# -> State# d -> State# d + writeWord16ArrayAsWord16X8# :: forall d. MutableByteArray# d -> Int# -> Word16X8# -> State# d -> State# d + writeWord16OffAddr# :: forall d. Addr# -> Int# -> Word16# -> State# d -> State# d + writeWord16OffAddrAsWord16X16# :: forall d. Addr# -> Int# -> Word16X16# -> State# d -> State# d + writeWord16OffAddrAsWord16X32# :: forall d. Addr# -> Int# -> Word16X32# -> State# d -> State# d + writeWord16OffAddrAsWord16X8# :: forall d. Addr# -> Int# -> Word16X8# -> State# d -> State# d + writeWord16X16Array# :: forall d. MutableByteArray# d -> Int# -> Word16X16# -> State# d -> State# d + writeWord16X16OffAddr# :: forall d. Addr# -> Int# -> Word16X16# -> State# d -> State# d + writeWord16X32Array# :: forall d. MutableByteArray# d -> Int# -> Word16X32# -> State# d -> State# d + writeWord16X32OffAddr# :: forall d. Addr# -> Int# -> Word16X32# -> State# d -> State# d + writeWord16X8Array# :: forall d. MutableByteArray# d -> Int# -> Word16X8# -> State# d -> State# d + writeWord16X8OffAddr# :: forall d. Addr# -> Int# -> Word16X8# -> State# d -> State# d + writeWord32Array# :: forall d. MutableByteArray# d -> Int# -> Word32# -> State# d -> State# d + writeWord32ArrayAsWord32X16# :: forall d. MutableByteArray# d -> Int# -> Word32X16# -> State# d -> State# d + writeWord32ArrayAsWord32X4# :: forall d. MutableByteArray# d -> Int# -> Word32X4# -> State# d -> State# d + writeWord32ArrayAsWord32X8# :: forall d. MutableByteArray# d -> Int# -> Word32X8# -> State# d -> State# d + writeWord32OffAddr# :: forall d. Addr# -> Int# -> Word32# -> State# d -> State# d + writeWord32OffAddrAsWord32X16# :: forall d. Addr# -> Int# -> Word32X16# -> State# d -> State# d + writeWord32OffAddrAsWord32X4# :: forall d. Addr# -> Int# -> Word32X4# -> State# d -> State# d + writeWord32OffAddrAsWord32X8# :: forall d. Addr# -> Int# -> Word32X8# -> State# d -> State# d + writeWord32X16Array# :: forall d. MutableByteArray# d -> Int# -> Word32X16# -> State# d -> State# d + writeWord32X16OffAddr# :: forall d. Addr# -> Int# -> Word32X16# -> State# d -> State# d + writeWord32X4Array# :: forall d. MutableByteArray# d -> Int# -> Word32X4# -> State# d -> State# d + writeWord32X4OffAddr# :: forall d. Addr# -> Int# -> Word32X4# -> State# d -> State# d + writeWord32X8Array# :: forall d. MutableByteArray# d -> Int# -> Word32X8# -> State# d -> State# d + writeWord32X8OffAddr# :: forall d. Addr# -> Int# -> Word32X8# -> State# d -> State# d + writeWord64Array# :: forall d. MutableByteArray# d -> Int# -> Word64# -> State# d -> State# d + writeWord64ArrayAsWord64X2# :: forall d. MutableByteArray# d -> Int# -> Word64X2# -> State# d -> State# d + writeWord64ArrayAsWord64X4# :: forall d. MutableByteArray# d -> Int# -> Word64X4# -> State# d -> State# d + writeWord64ArrayAsWord64X8# :: forall d. MutableByteArray# d -> Int# -> Word64X8# -> State# d -> State# d + writeWord64OffAddr# :: forall d. Addr# -> Int# -> Word64# -> State# d -> State# d + writeWord64OffAddrAsWord64X2# :: forall d. Addr# -> Int# -> Word64X2# -> State# d -> State# d + writeWord64OffAddrAsWord64X4# :: forall d. Addr# -> Int# -> Word64X4# -> State# d -> State# d + writeWord64OffAddrAsWord64X8# :: forall d. Addr# -> Int# -> Word64X8# -> State# d -> State# d + writeWord64X2Array# :: forall d. MutableByteArray# d -> Int# -> Word64X2# -> State# d -> State# d + writeWord64X2OffAddr# :: forall d. Addr# -> Int# -> Word64X2# -> State# d -> State# d + writeWord64X4Array# :: forall d. MutableByteArray# d -> Int# -> Word64X4# -> State# d -> State# d + writeWord64X4OffAddr# :: forall d. Addr# -> Int# -> Word64X4# -> State# d -> State# d + writeWord64X8Array# :: forall d. MutableByteArray# d -> Int# -> Word64X8# -> State# d -> State# d + writeWord64X8OffAddr# :: forall d. Addr# -> Int# -> Word64X8# -> State# d -> State# d + writeWord8Array# :: forall d. MutableByteArray# d -> Int# -> Word8# -> State# d -> State# d + writeWord8ArrayAsAddr# :: forall d. MutableByteArray# d -> Int# -> Addr# -> State# d -> State# d + writeWord8ArrayAsChar# :: forall d. MutableByteArray# d -> Int# -> Char# -> State# d -> State# d + writeWord8ArrayAsDouble# :: forall d. MutableByteArray# d -> Int# -> Double# -> State# d -> State# d + writeWord8ArrayAsFloat# :: forall d. MutableByteArray# d -> Int# -> Float# -> State# d -> State# d + writeWord8ArrayAsInt# :: forall d. MutableByteArray# d -> Int# -> Int# -> State# d -> State# d + writeWord8ArrayAsInt16# :: forall d. MutableByteArray# d -> Int# -> Int16# -> State# d -> State# d + writeWord8ArrayAsInt32# :: forall d. MutableByteArray# d -> Int# -> Int32# -> State# d -> State# d + writeWord8ArrayAsInt64# :: forall d. MutableByteArray# d -> Int# -> Int64# -> State# d -> State# d + writeWord8ArrayAsStablePtr# :: forall d a. MutableByteArray# d -> Int# -> StablePtr# a -> State# d -> State# d + writeWord8ArrayAsWideChar# :: forall d. MutableByteArray# d -> Int# -> Char# -> State# d -> State# d + writeWord8ArrayAsWord# :: forall d. MutableByteArray# d -> Int# -> Word# -> State# d -> State# d + writeWord8ArrayAsWord16# :: forall d. MutableByteArray# d -> Int# -> Word16# -> State# d -> State# d + writeWord8ArrayAsWord32# :: forall d. MutableByteArray# d -> Int# -> Word32# -> State# d -> State# d + writeWord8ArrayAsWord64# :: forall d. MutableByteArray# d -> Int# -> Word64# -> State# d -> State# d + writeWord8ArrayAsWord8X16# :: forall d. MutableByteArray# d -> Int# -> Word8X16# -> State# d -> State# d + writeWord8ArrayAsWord8X32# :: forall d. MutableByteArray# d -> Int# -> Word8X32# -> State# d -> State# d + writeWord8ArrayAsWord8X64# :: forall d. MutableByteArray# d -> Int# -> Word8X64# -> State# d -> State# d + writeWord8OffAddr# :: forall d. Addr# -> Int# -> Word8# -> State# d -> State# d + writeWord8OffAddrAsAddr# :: forall d. Addr# -> Int# -> Addr# -> State# d -> State# d + writeWord8OffAddrAsChar# :: forall d. Addr# -> Int# -> Char# -> State# d -> State# d + writeWord8OffAddrAsDouble# :: forall d. Addr# -> Int# -> Double# -> State# d -> State# d + writeWord8OffAddrAsFloat# :: forall d. Addr# -> Int# -> Float# -> State# d -> State# d + writeWord8OffAddrAsInt# :: forall d. Addr# -> Int# -> Int# -> State# d -> State# d + writeWord8OffAddrAsInt16# :: forall d. Addr# -> Int# -> Int16# -> State# d -> State# d + writeWord8OffAddrAsInt32# :: forall d. Addr# -> Int# -> Int32# -> State# d -> State# d + writeWord8OffAddrAsInt64# :: forall d. Addr# -> Int# -> Int64# -> State# d -> State# d + writeWord8OffAddrAsStablePtr# :: forall a d. Addr# -> Int# -> StablePtr# a -> State# d -> State# d + writeWord8OffAddrAsWideChar# :: forall d. Addr# -> Int# -> Char# -> State# d -> State# d + writeWord8OffAddrAsWord# :: forall d. Addr# -> Int# -> Word# -> State# d -> State# d + writeWord8OffAddrAsWord16# :: forall d. Addr# -> Int# -> Word16# -> State# d -> State# d + writeWord8OffAddrAsWord32# :: forall d. Addr# -> Int# -> Word32# -> State# d -> State# d + writeWord8OffAddrAsWord64# :: forall d. Addr# -> Int# -> Word64# -> State# d -> State# d + writeWord8OffAddrAsWord8X16# :: forall d. Addr# -> Int# -> Word8X16# -> State# d -> State# d + writeWord8OffAddrAsWord8X32# :: forall d. Addr# -> Int# -> Word8X32# -> State# d -> State# d + writeWord8OffAddrAsWord8X64# :: forall d. Addr# -> Int# -> Word8X64# -> State# d -> State# d + writeWord8X16Array# :: forall d. MutableByteArray# d -> Int# -> Word8X16# -> State# d -> State# d + writeWord8X16OffAddr# :: forall d. Addr# -> Int# -> Word8X16# -> State# d -> State# d + writeWord8X32Array# :: forall d. MutableByteArray# d -> Int# -> Word8X32# -> State# d -> State# d + writeWord8X32OffAddr# :: forall d. Addr# -> Int# -> Word8X32# -> State# d -> State# d + writeWord8X64Array# :: forall d. MutableByteArray# d -> Int# -> Word8X64# -> State# d -> State# d + writeWord8X64OffAddr# :: forall d. Addr# -> Int# -> Word8X64# -> State# d -> State# d + writeWordArray# :: forall d. MutableByteArray# d -> Int# -> Word# -> State# d -> State# d + writeWordOffAddr# :: forall d. Addr# -> Int# -> Word# -> State# d -> State# d + xor# :: Word# -> Word# -> Word# + xor64# :: Word64# -> Word64# -> Word64# + xorI# :: Int# -> Int# -> Int# + xorWord16# :: Word16# -> Word16# -> Word16# + xorWord32# :: Word32# -> Word32# -> Word32# + xorWord8# :: Word8# -> Word8# -> Word8# + yield# :: State# RealWorld -> State# RealWorld + type (~) :: forall k. k -> k -> Constraint + class (a ~ b) => (~) a b + {-# MINIMAL #-} + type (~~) :: forall k0 k1. k0 -> k1 -> Constraint + class (a ~~ b) => (~~) a b + {-# MINIMAL #-} + module GHC.Profiling.Eras where -- Safety: Trustworthy getUserEra :: GHC.Types.IO GHC.Types.Word @@ -8657,6 +10372,92 @@ module Prelude.Experimental where -- Instances: +instance GHC.Internal.Base.Alternative GHC.Types.IO -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Alternative [] -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Alternative GHC.Internal.Maybe.Maybe -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Applicative GHC.Internal.Data.Ord.Down -- Defined in ‘GHC.Internal.Data.Ord’ +instance GHC.Internal.Base.Applicative GHC.Types.IO -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Applicative [] -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Applicative GHC.Internal.Maybe.Maybe -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Applicative GHC.Internal.Base.NonEmpty -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Applicative Solo -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Applicative ((,) a) -- Defined in ‘GHC.Internal.Base’ +instance forall a b. (GHC.Internal.Base.Monoid a, GHC.Internal.Base.Monoid b) => GHC.Internal.Base.Applicative ((,,) a b) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c. (GHC.Internal.Base.Monoid a, GHC.Internal.Base.Monoid b, GHC.Internal.Base.Monoid c) => GHC.Internal.Base.Applicative ((,,,) a b c) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Functor GHC.Internal.Data.Ord.Down -- Defined in ‘GHC.Internal.Data.Ord’ +instance GHC.Internal.Base.Functor GHC.Types.IO -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Functor [] -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Functor GHC.Internal.Maybe.Maybe -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Functor GHC.Internal.Base.NonEmpty -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Functor Solo -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Functor ((,) a) -- Defined in ‘GHC.Internal.Base’ +instance forall a b. GHC.Internal.Base.Functor ((,,) a b) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c. GHC.Internal.Base.Functor ((,,,) a b c) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c d. GHC.Internal.Base.Functor ((,,,,) a b c d) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c d e. GHC.Internal.Base.Functor ((,,,,,) a b c d e) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c d e f. GHC.Internal.Base.Functor ((,,,,,,) a b c d e f) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Monad GHC.Internal.Data.Ord.Down -- Defined in ‘GHC.Internal.Data.Ord’ +instance GHC.Internal.Base.Monad GHC.Types.IO -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Monad [] -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Monad GHC.Internal.Maybe.Maybe -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Monad GHC.Internal.Base.NonEmpty -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Monad Solo -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monad ((,) a) -- Defined in ‘GHC.Internal.Base’ +instance forall a b. (GHC.Internal.Base.Monoid a, GHC.Internal.Base.Monoid b) => GHC.Internal.Base.Monad ((,,) a b) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c. (GHC.Internal.Base.Monoid a, GHC.Internal.Base.Monoid b, GHC.Internal.Base.Monoid c) => GHC.Internal.Base.Monad ((,,,) a b c) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.MonadPlus GHC.Types.IO -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.MonadPlus [] -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.MonadPlus GHC.Internal.Maybe.Maybe -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.Types.IO a) -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Monoid [a] -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Semigroup a => GHC.Internal.Base.Monoid (GHC.Internal.Maybe.Maybe a) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Monoid GHC.Types.Ordering -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (Solo a) -- Defined in ‘GHC.Internal.Base’ +instance forall a b. (GHC.Internal.Base.Monoid a, GHC.Internal.Base.Monoid b) => GHC.Internal.Base.Monoid (a, b) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c. (GHC.Internal.Base.Monoid a, GHC.Internal.Base.Monoid b, GHC.Internal.Base.Monoid c) => GHC.Internal.Base.Monoid (a, b, c) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c d. (GHC.Internal.Base.Monoid a, GHC.Internal.Base.Monoid b, GHC.Internal.Base.Monoid c, GHC.Internal.Base.Monoid d) => GHC.Internal.Base.Monoid (a, b, c, d) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c d e. (GHC.Internal.Base.Monoid a, GHC.Internal.Base.Monoid b, GHC.Internal.Base.Monoid c, GHC.Internal.Base.Monoid d, GHC.Internal.Base.Monoid e) => GHC.Internal.Base.Monoid (a, b, c, d, e) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Monoid () -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Semigroup a => GHC.Internal.Base.Semigroup (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Base.Semigroup a => GHC.Internal.Base.Semigroup (GHC.Types.IO a) -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Semigroup [a] -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Semigroup a => GHC.Internal.Base.Semigroup (GHC.Internal.Maybe.Maybe a) -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Semigroup (GHC.Internal.Base.NonEmpty a) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Semigroup GHC.Types.Ordering -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Base.Semigroup a => GHC.Internal.Base.Semigroup (Solo a) -- Defined in ‘GHC.Internal.Base’ +instance forall a b. (GHC.Internal.Base.Semigroup a, GHC.Internal.Base.Semigroup b) => GHC.Internal.Base.Semigroup (a, b) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c. (GHC.Internal.Base.Semigroup a, GHC.Internal.Base.Semigroup b, GHC.Internal.Base.Semigroup c) => GHC.Internal.Base.Semigroup (a, b, c) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c d. (GHC.Internal.Base.Semigroup a, GHC.Internal.Base.Semigroup b, GHC.Internal.Base.Semigroup c, GHC.Internal.Base.Semigroup d) => GHC.Internal.Base.Semigroup (a, b, c, d) -- Defined in ‘GHC.Internal.Base’ +instance forall a b c d e. (GHC.Internal.Base.Semigroup a, GHC.Internal.Base.Semigroup b, GHC.Internal.Base.Semigroup c, GHC.Internal.Base.Semigroup d, GHC.Internal.Base.Semigroup e) => GHC.Internal.Base.Semigroup (a, b, c, d, e) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Semigroup () -- Defined in ‘GHC.Internal.Base’ +instance GHC.Internal.Base.Semigroup GHC.Internal.Base.Void -- Defined in ‘GHC.Internal.Base’ +instance forall a. GHC.Internal.Bits.Bits a => GHC.Internal.Bits.Bits (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Bits.FiniteBits a => GHC.Internal.Bits.FiniteBits (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance GHC.Internal.Data.Data.Data GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ +instance forall a k (b :: k). GHC.Internal.Data.String.IsString a => GHC.Internal.Data.String.IsString (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.String’ +instance forall a. GHC.Internal.Data.String.IsString a => GHC.Internal.Data.String.IsString (GHC.Internal.Data.Functor.Identity.Identity a) -- Defined in ‘GHC.Internal.Data.String’ +instance forall a. (a ~ GHC.Types.Char) => GHC.Internal.Data.String.IsString [a] -- Defined in ‘GHC.Internal.Data.String’ +instance forall a. GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. (GHC.Internal.Enum.Enum a, GHC.Internal.Enum.Bounded a, GHC.Classes.Eq a) => GHC.Internal.Enum.Enum (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Float.Floating a => GHC.Internal.Float.Floating (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Float.RealFloat a => GHC.Internal.Float.RealFloat (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Foreign.Storable.Storable a => GHC.Internal.Foreign.Storable.Storable (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance GHC.Internal.IsList.IsList GHC.Internal.Stack.Types.CallStack -- Defined in ‘GHC.Internal.IsList’ +instance forall a. GHC.Internal.IsList.IsList [a] -- Defined in ‘GHC.Internal.IsList’ +instance forall a. GHC.Internal.IsList.IsList (GHC.Internal.Base.NonEmpty a) -- Defined in ‘GHC.Internal.IsList’ +instance GHC.Internal.IsList.IsList GHC.Internal.Data.Version.Version -- Defined in ‘GHC.Internal.IsList’ +instance forall a. GHC.Internal.IsList.IsList (GHC.Internal.Functor.ZipList.ZipList a) -- Defined in ‘GHC.Internal.IsList’ +instance forall a. GHC.Internal.Ix.Ix a => GHC.Internal.Ix.Ix (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Num.Num a => GHC.Internal.Num.Num (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Read.Read a => GHC.Internal.Read.Read (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Real.Fractional a => GHC.Internal.Real.Fractional (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Real.Real a => GHC.Internal.Real.Real (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Real.RealFrac a => GHC.Internal.Real.RealFrac (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Show.Show a => GHC.Internal.Show.Show (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Internal.Show.Show (GHC.Internal.Ptr.FunPtr a) -- Defined in ‘GHC.Internal.Ptr’ +instance forall a. GHC.Internal.Show.Show (GHC.Internal.Ptr.Ptr a) -- Defined in ‘GHC.Internal.Ptr’ +instance GHC.Internal.Show.Show GHC.Internal.IO.MaskingState -- Defined in ‘GHC.Internal.IO’ instance GHC.Classes.Eq GHC.Types.Bool -- Defined in ‘GHC.Classes’ instance GHC.Classes.Eq GHC.Types.Char -- Defined in ‘GHC.Classes’ instance GHC.Classes.Eq GHC.Types.Double -- Defined in ‘GHC.Classes’ @@ -8684,6 +10485,13 @@ instance forall a b c d e f g h i. (GHC.Classes.Eq a, GHC.Classes.Eq b, GHC.Clas instance GHC.Classes.Eq GHC.Types.TyCon -- Defined in ‘GHC.Classes’ instance GHC.Classes.Eq () -- Defined in ‘GHC.Classes’ instance GHC.Classes.Eq GHC.Types.Word -- Defined in ‘GHC.Classes’ +instance forall a. GHC.Classes.Eq a => GHC.Classes.Eq (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance GHC.Classes.Eq GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ +instance forall a. GHC.Classes.Eq (GHC.Internal.Ptr.FunPtr a) -- Defined in ‘GHC.Internal.Ptr’ +instance forall a. GHC.Classes.Eq (GHC.Internal.Ptr.Ptr a) -- Defined in ‘GHC.Internal.Ptr’ +instance forall a. GHC.Classes.Eq a => GHC.Classes.Eq (GHC.Internal.Base.NonEmpty a) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Classes.Eq GHC.Internal.Base.Void -- Defined in ‘GHC.Internal.Base’ +instance GHC.Classes.Eq GHC.Internal.IO.MaskingState -- Defined in ‘GHC.Internal.IO’ instance GHC.Classes.Ord GHC.Types.Bool -- Defined in ‘GHC.Classes’ instance GHC.Classes.Ord GHC.Types.Char -- Defined in ‘GHC.Classes’ instance GHC.Classes.Ord GHC.Types.Double -- Defined in ‘GHC.Classes’ @@ -8709,3 +10517,8 @@ instance forall a b c d e f g h i. (GHC.Classes.Ord a, GHC.Classes.Ord b, GHC.Cl instance GHC.Classes.Ord GHC.Types.TyCon -- Defined in ‘GHC.Classes’ instance GHC.Classes.Ord () -- Defined in ‘GHC.Classes’ instance GHC.Classes.Ord GHC.Types.Word -- Defined in ‘GHC.Classes’ +instance forall a. GHC.Classes.Ord a => GHC.Classes.Ord (GHC.Internal.Data.Ord.Down a) -- Defined in ‘GHC.Internal.Data.Ord’ +instance forall a. GHC.Classes.Ord (GHC.Internal.Ptr.FunPtr a) -- Defined in ‘GHC.Internal.Ptr’ +instance forall a. GHC.Classes.Ord (GHC.Internal.Ptr.Ptr a) -- Defined in ‘GHC.Internal.Ptr’ +instance forall a. GHC.Classes.Ord a => GHC.Classes.Ord (GHC.Internal.Base.NonEmpty a) -- Defined in ‘GHC.Internal.Base’ +instance GHC.Classes.Ord GHC.Internal.Base.Void -- Defined in ‘GHC.Internal.Base’ ===================================== testsuite/tests/interface-stability/ghc-experimental-exports.stdout-mingw32 ===================================== The diff for this file was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3c32c0602eac46602fb8c86088ba7e430d5f2fc6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3c32c0602eac46602fb8c86088ba7e430d5f2fc6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 11:04:49 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 23 Sep 2024 07:04:49 -0400 Subject: [Git][ghc/ghc][wip/romes/exceptions-layout] 14 commits: Demand: Combine examples into Note (#25107) Message-ID: <66f14b51e172e_6c8d64ce76c425e4@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exceptions-layout at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 72a85e43 by Ben Gamari at 2024-09-23T10:17:30+01:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - 90570d2c by Rodrigo Mesquita at 2024-09-23T12:04:28+01:00 exceptions: Improve the message layout as per #285 This commit fixes the layout of the additional information included when displaying an exception, namely the type of the exception. It also fixes the default handler's heading message to work well together with the improved display message of SomeException. - - - - - 27 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/extended_literals.rst - hadrian/bootstrap/bootstrap.py - hadrian/bootstrap/generate_bootstrap_plans - hadrian/bootstrap/plan-9_10_1.json - − hadrian/bootstrap/plan-9_4_1.json - − hadrian/bootstrap/plan-9_4_2.json The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/85ddf1c8227b893440e772e72b184344bb3403fe...90570d2c09e0be45878c2c01a6dce26c334a48ad -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/85ddf1c8227b893440e772e72b184344bb3403fe...90570d2c09e0be45878c2c01a6dce26c334a48ad You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 11:41:52 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 23 Sep 2024 07:41:52 -0400 Subject: [Git][ghc/ghc][wip/andreask/spec-constr-args] 82 commits: JS: add basic support for POSIX *at functions (#25190) Message-ID: <66f154006aefa_148aa7bc458636e2@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/spec-constr-args at Glasgow Haskell Compiler / GHC Commits: 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 870a8d7b by Andreas Klebinger at 2024-09-23T13:23:02+02:00 SpecConstr: Introduce a separate argument limit for forced specs. We used to put no limit at all on specializations forced via the SPEC argument. This isn't always reasonable so we introduce a very high limit that applies to forced specializations, a flag to control it, and we now emit a warning if we fail a specialization because we exceed the warning. Fixes #25197 - - - - - 22 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/bfabc66c04e9bd3fe62ba219a64f94a198b9a663...870a8d7be045c3f2575699efb8ef9eb2b3539872 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/bfabc66c04e9bd3fe62ba219a64f94a198b9a663...870a8d7be045c3f2575699efb8ef9eb2b3539872 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 11:47:43 2024 From: gitlab at gitlab.haskell.org (Sjoerd Visscher (@trac-sjoerd_visscher)) Date: Mon, 23 Sep 2024 07:47:43 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/parallel-depanal-downsweep] 106 commits: Bump array submodule Message-ID: <66f1555f99b97_148aa7bc4d0655dd@gitlab.mail> Sjoerd Visscher pushed to branch wip/torsten.schmits/parallel-depanal-downsweep at Glasgow Haskell Compiler / GHC Commits: 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - f3bf7876 by Cheng Shao at 2024-07-12T10:16:24-07:00 Oneshot bytecode linking - - - - - 7d129149 by Torsten Schmits at 2024-07-12T10:17:14-07:00 Package deps bytecode linking - - - - - 0df7982d by Cheng Shao at 2024-07-17T11:22:03-07:00 driver: add -dep-json -opt-json flags to ghc -M docs: document -dep-json -opt-json flags Rework protocol merge opt-json into dep-json normalize source filenames include more canonical variant of the package id choose the main library when -package matches multiple units with the same version allow specifying library deps with -package pkg:lib OsString <-> FilePath change - - - - - 76e97b66 by Cheng Shao at 2024-07-17T11:23:00-07:00 debug print - - - - - 52562340 by Matthew Pickering at 2024-07-18T16:00:44-07:00 Run on test-abi label - - - - - a58642ef by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Write a test for object determinism Extend abi_test with object determinism check Standalone run abi test Disable local test on CI - - - - - d7f945d1 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Remame uniques straight off stgtocmm, before cmm pipeline WIP Progress Work around LLVM assembler bug! In a really stupid way) Fix ordering of CLabels for IdLabels Local test script tweaks Do uniq renaming before SRTs Revert "Do uniq renaming before SRTs" This reverts commit db38b635d626106e40b3ab18091e0a24046c30c5. Do on CmmGroup Do uniq-renaming pass right at `codeGen` not better Revert "Do uniq-renaming pass right at `codeGen`" This reverts commit 74e9068aaaf736bf815a36bf74a0dde19a074a7a. Reapply "Do uniq renaming before SRTs" This reverts commit 682f89732fc2a95fa011f530c0c6922bf576d229. Try ALSO after SRT Revert "Try ALSO after SRT" This reverts commit c5dd7b426cde768126402aac3f39617ccb99f5c5. Renaming before and after SRTs bc of procs and srts and ... Wait no that was way too slow... cleaner approach, same idea Revert "Reapply "Do uniq renaming before SRTs"" This reverts commit 70ff49b7efc8c1fca46cba6eff630c5d39a99213. Finfixes Add traces Fix bug triggered by static data generated during SRT Wait, that was still there?! StableNmCmp is invalid for internal names... also, rename all Id uniques This fixes a really awful bug. Tweaks - - - - - 6812c22d by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 cmm: Back LabelMap with UDFM Use a deterministic unique map to back the implementation of `LabelMap`. This is necessary towards the goal of object code determinism in #12935. Our intended solution requires renaming uniques in a deterministic order (which will be the order in which they were created), but storing them label map makes us lose this order. Backing it with a UDFM fixes this issue. - - - - - 634ac689 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - 1314f380 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 UniqDSM det uniques + use in Cmm.Info Now for SRTs SRT generation using deterministic uniq supply Back LabelMap with deterministic UDFM TSAN uniq rename hard Revert "TSAN uniq rename hard" This reverts commit 7ca5ab3036c15f38c6d4cbcb616d415958c6bcda. improvements to uniqdsm UniqDSM ProcPoint CmmLayoutStack UniqDet 90% of cpsTop UniqDSM Major progress in using UniqDSM in CmmToAsm and Ncg backends Fix imports Un-back label map with udfm Revert "Un-back label map with udfm" This reverts commit f5d2e4257214a3f7b7d845651e6662c5babfd6a3. - - - - - 8de2be45 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Tweaks - - - - - 98ee4a92 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Make UDSM oneshot deriving via state - - - - - 72877779 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Put deterministic renaming behind a flag - - - - - 994c7e9c by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Introduce back LabelMap non deterministic - - - - - aba296e5 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Use NonDeterministic Label map in multiple passes (TODO: More could be available. Look through Det LabelMap uses again) - - - - - e0794abd by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Drop dumps - - - - - e6500cd0 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Make FactBase deterministic again - - - - - 82a70119 by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Undo a bit more NonDet LblMap - - - - - 809cc44a by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Revert "Undo a bit more NonDet LblMap" This reverts commit f526e1aee078712a5ae611d73fe90afa5e5095cb. - - - - - df575b9a by Rodrigo Mesquita at 2024-07-18T16:00:44-07:00 Revert "Make FactBase deterministic again" This reverts commit 315f05c001f41cf27b75870aa60d55f15a725421. - - - - - 4081878f by Matthew Pickering at 2024-07-18T16:00:44-07:00 MP fixes, don't rename external names - - - - - 447414e7 by Matthew Pickering at 2024-07-18T16:00:45-07:00 Use NonDet for CFG - - - - - d91fa729 by Matthew Pickering at 2024-07-18T16:00:45-07:00 fix Ord instance - - - - - 49898340 by Matthew Pickering at 2024-07-18T16:00:45-07:00 More NonDet - - - - - 21225d26 by Matthew Pickering at 2024-07-18T16:00:45-07:00 More explicit - - - - - be5864f7 by Matthew Pickering at 2024-07-18T16:00:45-07:00 Introduce DCmmDecl, start - - - - - 2ed04290 by Matthew Pickering at 2024-07-18T16:00:45-07:00 Removing more maps - - - - - a73568e4 by Matthew Pickering at 2024-07-18T16:00:45-07:00 Fix warnings - - - - - 33bb563c by Matthew Pickering at 2024-07-18T16:00:45-07:00 tests - - - - - ae9ca59a by Matthew Pickering at 2024-07-18T16:00:45-07:00 Fix tests - - - - - 222fa818 by Matthew Pickering at 2024-07-18T16:00:45-07:00 undo - - - - - e4721154 by Matthew Pickering at 2024-07-18T16:00:45-07:00 undo - - - - - 94486951 by Simon Peyton Jones at 2024-08-06T13:05:05-07:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 9c3b8ba0 by Simon Peyton Jones at 2024-08-09T08:01:40-07:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. * In GHC.Core.InstEnv, change the type synonym `Canonical` to a data type `CanonicalEvidence`; and document it better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. * Fix a bug I found when desugaring RULE left hand sides; see (NC1) in Note [Desugaring non-canonical evidence] in GHC.HsToCore.Binds This means giving a boolean flag to dsHsWrapper, alas. But I think it will go away again when I have finished with my (entirely separate) speciaise-on-values patch. - - - - - 0dc954f7 by Cheng Shao at 2024-08-21T16:10:54-07:00 compiler: implement --show-iface-abi-hash major mode - - - - - d9f06653 by Torsten Schmits at 2024-08-28T12:32:00-07:00 store boot module data separately - - - - - db5ba983 by Torsten Schmits at 2024-08-28T12:34:29-07:00 treat boot files in CLI targets properly - - - - - ae5c6d7d by Ian-Woo Kim at 2024-08-28T12:41:36-07:00 fix OsString <-> FilePath in cherry-picked code - - - - - 64c085bf by Ian-Woo Kim at 2024-09-13T09:42:54-07:00 added many debug prints - - - - - 6411a5dc by Torsten Schmits at 2024-09-20T14:41:53-07:00 Parallelize getRootSummary computations in dep analysis downsweep This reuses the upsweep step's infrastructure to process batches of modules in parallel. I benchmarked this by running `ghc -M` on two sets of 10,000 modules; one with a linear dependency chain and the other with a binary tree. Comparing different values for the number of modules per thread suggested an optimum at `length targets `div` (n_cap * 2)`, with results similar to this one (6 cores, 12 threads): ``` Benchmark 1: linear 1 jobs Time (mean ± σ): 1.775 s ± 0.026 s [User: 1.377 s, System: 0.399 s] Range (min … max): 1.757 s … 1.793 s 2 runs Benchmark 2: linear 6 jobs Time (mean ± σ): 876.2 ms ± 20.9 ms [User: 1833.2 ms, System: 518.6 ms] Range (min … max): 856.2 ms … 898.0 ms 3 runs Benchmark 3: linear 12 jobs Time (mean ± σ): 793.5 ms ± 23.2 ms [User: 2318.9 ms, System: 718.6 ms] Range (min … max): 771.9 ms … 818.0 ms 3 runs ``` Results don't differ much when the batch size is reduced to a quarter of that, but there's significant thread scheduling overhead for a size of 1: ``` Benchmark 1: linear 1 jobs Time (mean ± σ): 2.611 s ± 0.029 s [User: 2.851 s, System: 0.783 s] Range (min … max): 2.591 s … 2.632 s 2 runs Benchmark 2: linear 6 jobs Time (mean ± σ): 1.189 s ± 0.007 s [User: 2.707 s, System: 1.103 s] Range (min … max): 1.184 s … 1.194 s 2 runs Benchmark 3: linear 12 jobs Time (mean ± σ): 1.097 s ± 0.006 s [User: 2.938 s, System: 1.300 s] Range (min … max): 1.093 s … 1.101 s 2 runs ``` Larger batches also slightly worsen performance. - - - - - 60ae7968 by Torsten Schmits at 2024-09-20T14:41:53-07:00 use diag_wrapper argument like upsweep does - - - - - 4139f44d by Torsten Schmits at 2024-09-20T14:41:53-07:00 restructure a bit - - - - - 7d3f785d by Torsten Schmits at 2024-09-20T14:41:53-07:00 also pass down the Messager - - - - - 77c1b533 by Torsten Schmits at 2024-09-20T14:41:53-07:00 propagate exceptions to preserve current behavior - - - - - 7a8e215c by Torsten Schmits at 2024-09-20T14:41:53-07:00 use fixed bundle size - - - - - 64bc18c8 by Ian-Woo Kim at 2024-09-20T14:44:48-07:00 Make addToFinderCache monotonic - - - - - 5fdb31ab by Sjoerd Visscher at 2024-09-20T14:45:00-07:00 Fix tests that call downsweep - - - - - 30 changed files: - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - compiler/GHC.hs - compiler/GHC/Builtin/PrimOps.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/Utils.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Linker.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/GenericOpt.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/04a3a9e61d0cc14865b445af3220d93e15a1e29f...5fdb31abb5f104c438e6b5f34ed607641db4dae4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/04a3a9e61d0cc14865b445af3220d93e15a1e29f...5fdb31abb5f104c438e6b5f34ed607641db4dae4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 11:55:32 2024 From: gitlab at gitlab.haskell.org (Sjoerd Visscher (@trac-sjoerd_visscher)) Date: Mon, 23 Sep 2024 07:55:32 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/parallel-depanal-downsweep] 160 commits: Testsuite: use py-cpuinfo to compute CPU features Message-ID: <66f15734cf042_148aa730df18661bf@gitlab.mail> Sjoerd Visscher pushed to branch wip/torsten.schmits/parallel-depanal-downsweep at Glasgow Haskell Compiler / GHC Commits: 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - f1cce48d by Torsten Schmits at 2024-09-23T13:55:00+02:00 Parallelize getRootSummary computations in dep analysis downsweep This reuses the upsweep step's infrastructure to process batches of modules in parallel. I benchmarked this by running `ghc -M` on two sets of 10,000 modules; one with a linear dependency chain and the other with a binary tree. Comparing different values for the number of modules per thread suggested an optimum at `length targets `div` (n_cap * 2)`, with results similar to this one (6 cores, 12 threads): ``` Benchmark 1: linear 1 jobs Time (mean ± σ): 1.775 s ± 0.026 s [User: 1.377 s, System: 0.399 s] Range (min … max): 1.757 s … 1.793 s 2 runs Benchmark 2: linear 6 jobs Time (mean ± σ): 876.2 ms ± 20.9 ms [User: 1833.2 ms, System: 518.6 ms] Range (min … max): 856.2 ms … 898.0 ms 3 runs Benchmark 3: linear 12 jobs Time (mean ± σ): 793.5 ms ± 23.2 ms [User: 2318.9 ms, System: 718.6 ms] Range (min … max): 771.9 ms … 818.0 ms 3 runs ``` Results don't differ much when the batch size is reduced to a quarter of that, but there's significant thread scheduling overhead for a size of 1: ``` Benchmark 1: linear 1 jobs Time (mean ± σ): 2.611 s ± 0.029 s [User: 2.851 s, System: 0.783 s] Range (min … max): 2.591 s … 2.632 s 2 runs Benchmark 2: linear 6 jobs Time (mean ± σ): 1.189 s ± 0.007 s [User: 2.707 s, System: 1.103 s] Range (min … max): 1.184 s … 1.194 s 2 runs Benchmark 3: linear 12 jobs Time (mean ± σ): 1.097 s ± 0.006 s [User: 2.938 s, System: 1.300 s] Range (min … max): 1.093 s … 1.101 s 2 runs ``` Larger batches also slightly worsen performance. - - - - - 28 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/GenericOpt.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5fdb31abb5f104c438e6b5f34ed607641db4dae4...f1cce48d8cec226cc511107494b69843f09d1503 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5fdb31abb5f104c438e6b5f34ed607641db4dae4...f1cce48d8cec226cc511107494b69843f09d1503 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 12:02:22 2024 From: gitlab at gitlab.haskell.org (Sjoerd Visscher (@trac-sjoerd_visscher)) Date: Mon, 23 Sep 2024 08:02:22 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/parallel-depanal-downsweep] Parallelize getRootSummary computations in dep analysis downsweep Message-ID: <66f158ceb4080_148aa7304940718f6@gitlab.mail> Sjoerd Visscher pushed to branch wip/torsten.schmits/parallel-depanal-downsweep at Glasgow Haskell Compiler / GHC Commits: c8e57a72 by Torsten Schmits at 2024-09-23T14:02:10+02:00 Parallelize getRootSummary computations in dep analysis downsweep This reuses the upsweep step's infrastructure to process batches of modules in parallel. I benchmarked this by running `ghc -M` on two sets of 10,000 modules; one with a linear dependency chain and the other with a binary tree. Comparing different values for the number of modules per thread suggested an optimum at `length targets `div` (n_cap * 2)`, with results similar to this one (6 cores, 12 threads): ``` Benchmark 1: linear 1 jobs Time (mean ± σ): 1.775 s ± 0.026 s [User: 1.377 s, System: 0.399 s] Range (min … max): 1.757 s … 1.793 s 2 runs Benchmark 2: linear 6 jobs Time (mean ± σ): 876.2 ms ± 20.9 ms [User: 1833.2 ms, System: 518.6 ms] Range (min … max): 856.2 ms … 898.0 ms 3 runs Benchmark 3: linear 12 jobs Time (mean ± σ): 793.5 ms ± 23.2 ms [User: 2318.9 ms, System: 718.6 ms] Range (min … max): 771.9 ms … 818.0 ms 3 runs ``` Results don't differ much when the batch size is reduced to a quarter of that, but there's significant thread scheduling overhead for a size of 1: ``` Benchmark 1: linear 1 jobs Time (mean ± σ): 2.611 s ± 0.029 s [User: 2.851 s, System: 0.783 s] Range (min … max): 2.591 s … 2.632 s 2 runs Benchmark 2: linear 6 jobs Time (mean ± σ): 1.189 s ± 0.007 s [User: 2.707 s, System: 1.103 s] Range (min … max): 1.184 s … 1.194 s 2 runs Benchmark 3: linear 12 jobs Time (mean ± σ): 1.097 s ± 0.006 s [User: 2.938 s, System: 1.300 s] Range (min … max): 1.093 s … 1.101 s 2 runs ``` Larger batches also slightly worsen performance. - - - - - 5 changed files: - compiler/GHC/Driver/Make.hs - compiler/GHC/Unit/Finder.hs - testsuite/tests/ghc-api/downsweep/OldModLocation.hs - testsuite/tests/ghc-api/downsweep/PartialDownsweep.hs - utils/haddock/haddock-api/src/Haddock/Interface.hs Changes: ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -5,6 +5,8 @@ {-# LANGUAGE ApplicativeDo #-} {-# LANGUAGE MultiWayIf #-} {-# LANGUAGE RecordWildCards #-} +{-# LANGUAGE LambdaCase #-} +{-# LANGUAGE BlockArguments #-} -- ----------------------------------------------------------------------------- -- @@ -122,7 +124,7 @@ import qualified Control.Monad.Catch as MC import Data.IORef import Data.Maybe import Data.Time -import Data.List (sortOn) +import Data.List (sortOn, unfoldr) import Data.Bifunctor (first) import System.Directory import System.FilePath @@ -169,7 +171,7 @@ depanal :: GhcMonad m => -> Bool -- ^ allow duplicate roots -> m ModuleGraph depanal excluded_mods allow_dup_roots = do - (errs, mod_graph) <- depanalE excluded_mods allow_dup_roots + (errs, mod_graph) <- depanalE mkUnknownDiagnostic Nothing excluded_mods allow_dup_roots if isEmptyMessages errs then pure mod_graph else throwErrors (fmap GhcDriverMessage errs) @@ -177,12 +179,14 @@ depanal excluded_mods allow_dup_roots = do -- | Perform dependency analysis like in 'depanal'. -- In case of errors, the errors and an empty module graph are returned. depanalE :: GhcMonad m => -- New for #17459 - [ModuleName] -- ^ excluded modules + (GhcMessage -> AnyGhcDiagnostic) + -> Maybe Messager + -> [ModuleName] -- ^ excluded modules -> Bool -- ^ allow duplicate roots -> m (DriverMessages, ModuleGraph) -depanalE excluded_mods allow_dup_roots = do +depanalE diag_wrapper msg excluded_mods allow_dup_roots = do hsc_env <- getSession - (errs, mod_graph) <- depanalPartial excluded_mods allow_dup_roots + (errs, mod_graph) <- depanalPartial diag_wrapper msg excluded_mods allow_dup_roots if isEmptyMessages errs then do hsc_env <- getSession @@ -220,11 +224,13 @@ depanalE excluded_mods allow_dup_roots = do -- new module graph. depanalPartial :: GhcMonad m - => [ModuleName] -- ^ excluded modules + => (GhcMessage -> AnyGhcDiagnostic) + -> Maybe Messager + -> [ModuleName] -- ^ excluded modules -> Bool -- ^ allow duplicate roots -> m (DriverMessages, ModuleGraph) -- ^ possibly empty 'Bag' of errors and a module graph. -depanalPartial excluded_mods allow_dup_roots = do +depanalPartial diag_wrapper msg excluded_mods allow_dup_roots = do hsc_env <- getSession let targets = hsc_targets hsc_env @@ -243,7 +249,7 @@ depanalPartial excluded_mods allow_dup_roots = do liftIO $ flushFinderCaches (hsc_FC hsc_env) (hsc_unit_env hsc_env) (errs, graph_nodes) <- liftIO $ downsweep - hsc_env (mgModSummaries old_graph) + hsc_env diag_wrapper msg (mgModSummaries old_graph) excluded_mods allow_dup_roots let mod_graph = mkModuleGraph graph_nodes @@ -497,8 +503,8 @@ loadWithCache :: GhcMonad m => Maybe ModIfaceCache -- ^ Instructions about how t -> LoadHowMuch -- ^ How much `loadWithCache` should load -> m SuccessFlag loadWithCache cache diag_wrapper how_much = do - (errs, mod_graph) <- depanalE [] False -- #17459 msg <- mkBatchMsg <$> getSession + (errs, mod_graph) <- depanalE diag_wrapper (Just msg) [] False -- #17459 success <- load' cache how_much diag_wrapper (Just msg) mod_graph if isEmptyMessages errs then pure success @@ -506,7 +512,7 @@ loadWithCache cache diag_wrapper how_much = do -- Note [Unused packages] -- ~~~~~~~~~~~~~~~~~~~~~~ --- Cabal passes `--package-id` flag for each direct dependency. But GHC +-- Cabal passes `-package-id` flag for each direct dependency. But GHC -- loads them lazily, so when compilation is done, we have a list of all -- actually loaded packages. All the packages, specified on command line, -- but never loaded, are probably unused dependencies. @@ -1551,6 +1557,8 @@ type DownsweepCache = M.Map (UnitId, PkgQual, ModuleNameWithIsBoot) [Either Driv -- module, plus one for any hs-boot files. The imports of these nodes -- are all there, including the imports of non-home-package modules. downsweep :: HscEnv + -> (GhcMessage -> AnyGhcDiagnostic) + -> Maybe Messager -> [ModSummary] -- ^ Old summaries -> [ModuleName] -- Ignore dependencies on these; treat @@ -1562,17 +1570,38 @@ downsweep :: HscEnv -- The non-error elements of the returned list all have distinct -- (Modules, IsBoot) identifiers, unless the Bool is true in -- which case there can be repeats -downsweep hsc_env old_summaries excl_mods allow_dup_roots +downsweep hsc_env diag_wrapper msg old_summaries excl_mods allow_dup_roots = do + n_jobs <- mkWorkerLimit (hsc_dflags hsc_env) + new <- rootSummariesParallel n_jobs hsc_env diag_wrapper msg summary + downsweep_imports hsc_env old_summary_map excl_mods allow_dup_roots new + where + summary = getRootSummary excl_mods old_summary_map + + -- A cache from file paths to the already summarised modules. The same file + -- can be used in multiple units so the map is also keyed by which unit the + -- file was used in. + -- Reuse these if we can because the most expensive part of downsweep is + -- reading the headers. + old_summary_map :: M.Map (UnitId, FilePath) ModSummary + old_summary_map = + M.fromList [((ms_unitid ms, msHsFilePath ms), ms) | ms <- old_summaries] + +downsweep_imports :: HscEnv + -> M.Map (UnitId, FilePath) ModSummary + -> [ModuleName] + -> Bool + -> ([(UnitId, DriverMessages)], [ModSummary]) + -> IO ([DriverMessages], [ModuleGraphNode]) +downsweep_imports hsc_env old_summaries excl_mods allow_dup_roots (root_errs, rootSummariesOk) = do - (root_errs, rootSummariesOk) <- partitionWithM getRootSummary roots -- #17549 let root_map = mkRootMap rootSummariesOk checkDuplicates root_map (deps, map0) <- loopSummaries rootSummariesOk (M.empty, root_map) - let closure_errs = checkHomeUnitsClosed (hsc_unit_env hsc_env) - let unit_env = hsc_unit_env hsc_env - let tmpfs = hsc_tmpfs hsc_env + let closure_errs = checkHomeUnitsClosed unit_env + unit_env = hsc_unit_env hsc_env + tmpfs = hsc_tmpfs hsc_env - let downsweep_errs = lefts $ concat $ M.elems map0 + downsweep_errs = lefts $ concat $ M.elems map0 downsweep_nodes = M.elems deps (other_errs, unit_nodes) = partitionEithers $ unitEnv_foldWithKey (\nodes uid hue -> nodes ++ unitModuleNodes downsweep_nodes uid hue) [] (hsc_HUG hsc_env) @@ -1604,46 +1633,6 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots [(ms_unitid ms, b, c) | (b, c) <- msDeps ms ] logger = hsc_logger hsc_env - roots = hsc_targets hsc_env - - -- A cache from file paths to the already summarised modules. The same file - -- can be used in multiple units so the map is also keyed by which unit the - -- file was used in. - -- Reuse these if we can because the most expensive part of downsweep is - -- reading the headers. - old_summary_map :: M.Map (UnitId, FilePath) ModSummary - old_summary_map = M.fromList [((ms_unitid ms, msHsFilePath ms), ms) | ms <- old_summaries] - - getRootSummary :: Target -> IO (Either (UnitId, DriverMessages) ModSummary) - getRootSummary Target { targetId = TargetFile file mb_phase - , targetContents = maybe_buf - , targetUnitId = uid - } - = do let offset_file = augmentByWorkingDirectory dflags file - exists <- liftIO $ doesFileExist offset_file - if exists || isJust maybe_buf - then first (uid,) <$> - summariseFile hsc_env home_unit old_summary_map offset_file mb_phase - maybe_buf - else return $ Left $ (uid,) $ singleMessage - $ mkPlainErrorMsgEnvelope noSrcSpan (DriverFileNotFound offset_file) - where - dflags = homeUnitEnv_dflags (ue_findHomeUnitEnv uid (hsc_unit_env hsc_env)) - home_unit = ue_unitHomeUnit uid (hsc_unit_env hsc_env) - getRootSummary Target { targetId = TargetModule modl - , targetContents = maybe_buf - , targetUnitId = uid - } - = do maybe_summary <- summariseModule hsc_env home_unit old_summary_map NotBoot - (L rootLoc modl) (ThisPkg (homeUnitId home_unit)) - maybe_buf excl_mods - case maybe_summary of - FoundHome s -> return (Right s) - FoundHomeWithError err -> return (Left err) - _ -> return $ Left $ (uid, moduleNotFoundErr modl) - where - home_unit = ue_unitHomeUnit uid (hsc_unit_env hsc_env) - rootLoc = mkGeneralSrcSpan (fsLit "") -- In a root module, the filename is allowed to diverge from the module -- name, so we have to check that there aren't multiple root files @@ -1711,7 +1700,7 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots loopImports ss done summarised | otherwise = do - mb_s <- summariseModule hsc_env home_unit old_summary_map + mb_s <- summariseModule hsc_env home_unit old_summaries is_boot wanted_mod mb_pkg Nothing excl_mods case mb_s of @@ -1736,6 +1725,90 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots GWIB { gwib_mod = L loc mod, gwib_isBoot = is_boot } = gwib wanted_mod = L loc mod +getRootSummary :: + [ModuleName] -> + M.Map (UnitId, FilePath) ModSummary -> + HscEnv -> + Target -> + IO (Either (UnitId, DriverMessages) ModSummary) +getRootSummary excl_mods old_summary_map hsc_env target + | TargetFile file mb_phase <- targetId + = do + let offset_file = augmentByWorkingDirectory dflags file + exists <- liftIO $ doesFileExist offset_file + if exists || isJust maybe_buf + then first (uid,) <$> + summariseFile hsc_env home_unit old_summary_map offset_file mb_phase + maybe_buf + else + return $ Left $ (uid,) $ singleMessage $ + mkPlainErrorMsgEnvelope noSrcSpan (DriverFileNotFound offset_file) + | TargetModule modl <- targetId + = do + maybe_summary <- summariseModule hsc_env home_unit old_summary_map NotBoot + (L rootLoc modl) (ThisPkg (homeUnitId home_unit)) + maybe_buf excl_mods + pure case maybe_summary of + FoundHome s -> Right s + FoundHomeWithError err -> Left err + _ -> Left (uid, moduleNotFoundErr modl) + where + Target {targetId, targetContents = maybe_buf, targetUnitId = uid} = target + home_unit = ue_unitHomeUnit uid (hsc_unit_env hsc_env) + rootLoc = mkGeneralSrcSpan (fsLit "") + dflags = homeUnitEnv_dflags (ue_findHomeUnitEnv uid (hsc_unit_env hsc_env)) + +-- | Execute 'getRootSummary' for the 'Target's using the parallelism pipeline +-- system. +-- Create bundles of 'Target's wrapped in a 'MakeAction' that uses +-- 'withAbstractSem' to wait for a free slot, limiting the number of +-- concurrently computed summaries to the value of the @-j@ option or the slots +-- allocated by the job server, if that is used. +-- +-- The 'MakeAction' returns 'Maybe', which is not handled as an error, because +-- 'runLoop' only sets it to 'Nothing' when an exception was thrown, so the +-- result won't be read anyway here. +-- +-- To emulate the current behavior, we funnel exceptions past the concurrency +-- barrier and rethrow the first one afterwards. +rootSummariesParallel :: + WorkerLimit -> + HscEnv -> + (GhcMessage -> AnyGhcDiagnostic) -> + Maybe Messager -> + (HscEnv -> Target -> IO (Either (UnitId, DriverMessages) ModSummary)) -> + IO ([(UnitId, DriverMessages)], [ModSummary]) +rootSummariesParallel n_jobs hsc_env diag_wrapper msg get_summary = do + (actions, get_results) <- unzip <$> mapM action_and_result (zip [1..] bundles) + runPipelines n_jobs hsc_env diag_wrapper msg actions + (sequence . catMaybes <$> sequence get_results) >>= \case + Right results -> pure (partitionEithers (concat results)) + Left exc -> throwIO exc + where + bundles = mk_bundles targets + + mk_bundles = unfoldr \case + [] -> Nothing + ts -> Just (splitAt bundle_size ts) + + bundle_size = 20 + + targets = hsc_targets hsc_env + + action_and_result (log_queue_id, ts) = do + res_var <- liftIO newEmptyMVar + pure $! (MakeAction (action log_queue_id ts) res_var, readMVar res_var) + + action log_queue_id target_bundle = do + env at MakeEnv {compile_sem} <- ask + lift $ lift $ + withAbstractSem compile_sem $ + withLoggerHsc log_queue_id env \ lcl_hsc_env -> + MC.try (mapM (get_summary lcl_hsc_env) target_bundle) >>= \case + Left e | Just (_ :: SomeAsyncException) <- fromException e -> + throwIO e + a -> pure a + -- | This function checks then important property that if both p and q are home units -- then any dependency of p, which transitively depends on q is also a home unit. -- @@ -2453,12 +2526,12 @@ wrapAction msg_wrapper hsc_env k = do let lcl_logger = hsc_logger hsc_env lcl_dynflags = hsc_dflags hsc_env print_config = initPrintConfig lcl_dynflags - let logg err = printMessages lcl_logger print_config (initDiagOpts lcl_dynflags) (msg_wrapper <$> srcErrorMessages err) + logg err = printMessages lcl_logger print_config (initDiagOpts lcl_dynflags) (msg_wrapper <$> srcErrorMessages err) -- MP: It is a bit strange how prettyPrintGhcErrors handles some errors but then we handle -- SourceError and ThreadKilled differently directly below. TODO: Refactor to use `catches` -- directly. MP should probably use safeTry here to not catch async exceptions but that will regress performance due to -- internally using forkIO. - mres <- MC.try $ liftIO $ prettyPrintGhcErrors lcl_logger $ k + mres <- MC.try $ prettyPrintGhcErrors lcl_logger $ k case mres of Right res -> return $ Just res Left exc -> do @@ -2657,7 +2730,7 @@ R.hs: module R where == Why we need to rehydrate A's ModIface before compiling R.hs After compiling A.hs we'll have a TypeEnv in which the Id for `f` has a type -type uses the AbstractTyCon T; and a TyCon for `S` that also mentions that same +that uses the AbstractTyCon T; and a TyCon for `S` that also mentions that same AbstractTyCon. (Abstract because it came from R.hs-boot; we know nothing about it.) ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -86,6 +86,20 @@ type BaseName = OsPath -- Basename of file -- ----------------------------------------------------------------------------- -- The finder's cache +{- +[Note: Monotonic addToFinderCache] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +addToFinderCache is only used by functions that return the cached value +if there is one, or by functions that always write an InstalledFound value. +Without multithreading it is then safe to always directly write the value +without checking the previously cached value. + +However, with multithreading, it is possible that another function has +written a value into cache between the lookup and the addToFinderCache call. +in this case we should check to not overwrite an InstalledFound with an +InstalledNotFound. +-} initFinderCache :: IO FinderCache initFinderCache = do @@ -100,7 +114,12 @@ initFinderCache = do addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () addToFinderCache key val = - atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleWithIsBootEnv c key val, ()) + atomicModifyIORef' mod_cache $ \c -> + case (lookupInstalledModuleEnv c key, val) of + -- Don't overwrite an InstalledFound with an InstalledNotFound + -- See [Note Monotonic addToFinderCache] + (Just InstalledFound{}, InstalledNotFound{}) -> (c, ()) + _ -> (extendInstalledModuleEnv c key val, ()) lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) lookupFinderCache key = do ===================================== testsuite/tests/ghc-api/downsweep/OldModLocation.hs ===================================== @@ -6,6 +6,7 @@ import GHC import GHC.Driver.Make import GHC.Driver.Session import GHC.Driver.Env +import GHC.Types.Error (mkUnknownDiagnostic) import GHC.Unit.Module.Graph import GHC.Unit.Finder @@ -47,13 +48,13 @@ main = do liftIO $ do - _emss <- downsweep hsc_env [] [] False + _emss <- downsweep hsc_env mkUnknownDiagnostic Nothing [] [] False flushFinderCaches (hsc_FC hsc_env) (hsc_unit_env hsc_env) createDirectoryIfMissing False "mydir" renameFile "B.hs" "mydir/B.hs" - (_, nodes) <- downsweep hsc_env [] [] False + (_, nodes) <- downsweep hsc_env mkUnknownDiagnostic Nothing [] [] False -- If 'checkSummaryTimestamp' were to call 'addHomeModuleToFinder' with -- (ms_location old_summary) like summariseFile used to instead of ===================================== testsuite/tests/ghc-api/downsweep/PartialDownsweep.hs ===================================== @@ -6,6 +6,7 @@ import GHC import GHC.Driver.Make import GHC.Driver.Session +import GHC.Types.Error (mkUnknownDiagnostic) import GHC.Utils.Outputable import GHC.Utils.Exception (ExceptionMonad) import GHC.Data.Bag @@ -168,7 +169,7 @@ go label mods cnd = setTargets [tgt] hsc_env <- getSession - (_, nodes) <- liftIO $ downsweep hsc_env [] [] False + (_, nodes) <- liftIO $ downsweep hsc_env mkUnknownDiagnostic Nothing [] [] False it label $ cnd (mapMaybe moduleGraphNodeModSum nodes) ===================================== utils/haddock/haddock-api/src/Haddock/Interface.hs ===================================== @@ -170,7 +170,7 @@ createIfaces verbosity modules flags instIfaceMap = do _ <- setSessionDynFlags dflags'' targets <- mapM (\(filePath, _) -> guessTarget filePath Nothing Nothing) hs_srcs setTargets targets - (_errs, modGraph) <- depanalE [] False + (_errs, modGraph) <- depanalE mkUnknownDiagnostic (Just batchMsg) [] False -- Create (if necessary) and load .hi-files. With --no-compilation this happens later. when (Flag_NoCompilation `notElem` flags) $ do View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c8e57a72b9a504c781d8d7900766a9eab3a9db3f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c8e57a72b9a504c781d8d7900766a9eab3a9db3f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 12:20:46 2024 From: gitlab at gitlab.haskell.org (Ryan Scott (@RyanGlScott)) Date: Mon, 23 Sep 2024 08:20:46 -0400 Subject: [Git][ghc/ghc][wip/T25148] Even more review comments Message-ID: <66f15d1d594e4_148aa75d18a8759f3@gitlab.mail> Ryan Scott pushed to branch wip/T25148 at Glasgow Haskell Compiler / GHC Commits: 4776f2e3 by Ryan Scott at 2024-09-23T08:15:28-04:00 Even more review comments - - - - - 2 changed files: - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/TyCl/Instance.hs Changes: ===================================== compiler/GHC/Tc/Deriv/Generate.hs ===================================== @@ -1920,9 +1920,12 @@ ambiguous type variables. Here are a couple of examples to illustrate this: * In this example (from #15637), the class-bound type variable `a` is ambiguous in the type of `f`: - class C a where f :: String - instance C () where f = "foo" - newtype T = T () deriving C + class C a where + f :: String -- f :: forall a. C a => String + instance C () + where f = "foo" + newtype T = T () + deriving C A naïve attempt and generating a C T instance would be: @@ -1966,7 +1969,7 @@ ambiguous type variables. Here are a couple of examples to illustrate this: Note that we use this same approach for resolving ambiguity in default methods, as described in Note [Default methods in instances] (Wrinkle: - Ambiguous types from method type signatures) in GHC.Tc.TyCl.Instance. + Ambiguous types from vanilla method type signatures) in GHC.Tc.TyCl.Instance. -} gen_Newtype_binds :: SrcSpan ===================================== compiler/GHC/Tc/TyCl/Instance.hs ===================================== @@ -2230,15 +2230,19 @@ mkDefMethBind loc dfun_id clas sel_id dm_name dm_spec [ ty | (tcb, ty) <- tyConBinders (classTyCon clas) `zip` inst_tys , tyConBinderForAllTyFlag tcb /= Inferred ] - -- When dealing with a vanilla default method, compute the type variables - -- from the method's type signature. That way, we can bind them with - -- TypeAbstractions (visible_sel_pats) and use them in the visible type - -- application (visible_sel_tys). If we are dealing with a generic default - -- method, don't bother doing any of this. See Note [Default methods in - -- instances] (Wrinkle: Ambiguous types from method type signatures). visible_sel_tvbs = case dm_spec of - VanillaDM -> filter (\tvb -> binderFlag tvb /= InferredSpec) sel_tvbs + -- When dealing with a vanilla default method, compute the type + -- variables from the method's type signature. That way, we can bind + -- them with TypeAbstractions (visible_sel_pats) and use them in the + -- visible type application (visible_sel_tys). See Note [Default methods + -- in instances] (Wrinkle: Ambiguous types from vanilla method type + -- signatures). + VanillaDM -> filter (\tvb -> binderFlag tvb /= InferredSpec) sel_tvbs + -- If we are dealing with a generic default method, on the other hand, + -- don't bother doing any of this. See Note [Default methods + -- in instances] (Wrinkle: Ambiguous types from generic default method + -- type signatures). GenericDM {} -> [] visible_sel_pats = map mk_ty_pat visible_sel_tvbs visible_sel_tys = map (mkTyVarTy . binderVar) visible_sel_tvbs @@ -2324,7 +2328,7 @@ post-type-checked code, which took a lot more code, and didn't work for generic default methods. ----- --- Wrinkle: Ambiguous types from method type signatures +-- Wrinkle: Ambiguous types from vanilla method type signatures ----- In the Bar example above, the ambiguity arises from `v`, a type variable @@ -2387,19 +2391,37 @@ This resolves the ambiguity and avoids the need for a subtype check. (We also use a similar trick for resolving ambiguity in GeneralizedNewtypeDeriving: see also Note [GND and ambiguity] in GHC.Tc.Deriv.Generate.) -Note that this approach will only work for vanilla default methods and /not/ -for generic default methods. This is because for vanilla default methods, the -type of the generated $dm* function will always quantify the same type -variables as the method's original type signature, in the same order and with -the same specificities. For example, the type of the $dmf function will be: +----- +-- Wrinkle: Ambiguous types from generic default method type signatures +----- + +Note that the approach described above (in Wrinkle: Ambiguous types from +vanilla method type signatures) will only work for vanilla default methods and +/not/ for generic default methods (i.e., methods using DefaultSignatures). This +is because for vanilla default methods, the type of the generated $dm* function +will always quantify the same type variables as the method's original type +signature, in the same order and with the same specificities. For example, the +type of the $dmf function will be: $dmf :: forall t. A t => forall x m. Monoid x => t m -> m As such, it is guaranteed that the type variables from the method's original type signature will line up exactly with the type variables from the $dm* -function (after instantiating all of the class variables). We cannot guarantee -this property for generic default methods, however. There are a number of -reasons why this would not work: +function (after instantiating all of the class variables): + + instance A [] where + f @x @m = $dmf @[] @x @m + +We cannot guarantee this property for generic default methods, however. As +such, we must be more conservative and generate code without instantiating any +of the type variables bound by the method's type signature (only the type +variables bound by the class header): + + instance A [] where + f = $dmf @[] + +There are a number of reasons why we cannot reliably instantiate the type +variables bound by a generic default method's type signature: * Default methods can quantify type variables in a different order, e.g., View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4776f2e351147e373adc15430f00d26796c9f2f9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4776f2e351147e373adc15430f00d26796c9f2f9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 12:21:30 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 23 Sep 2024 08:21:30 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] 55 commits: JS: fake support for native adjustors (#25159) Message-ID: <66f15d4a8fe89_148aa764da3477697@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 73753020 by Cheng Shao at 2024-09-23T14:21:16+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). Metric Decrease: MultiLayerModules T13701 - - - - - 16 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6c1a78cb43b63e7d9a21a04fb0530baf59de10c3...7375302034669a48d939c407a0d9669218b80dfe -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6c1a78cb43b63e7d9a21a04fb0530baf59de10c3...7375302034669a48d939c407a0d9669218b80dfe You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 13:10:34 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 23 Sep 2024 09:10:34 -0400 Subject: [Git][ghc/ghc][wip/clc275] 76 commits: simplifier: Fix space leak during demand analysis Message-ID: <66f168ca50042_1f095413c2842a3@gitlab.mail> Andreas Klebinger pushed to branch wip/clc275 at Glasgow Haskell Compiler / GHC Commits: 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 85b92a49 by Ben Gamari at 2024-09-23T13:10:28+00:00 base: Add `HasCallStack` constraint to `ioError` As proposed in core-libraries-committee#275. - - - - - 22 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/eb462c9e84b1ada45c61b30b2567a69685cf6184...85b92a493143ca704a2d381896fd71e3da532567 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/eb462c9e84b1ada45c61b30b2567a69685cf6184...85b92a493143ca704a2d381896fd71e3da532567 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 13:28:01 2024 From: gitlab at gitlab.haskell.org (Cheng Shao (@TerrorJack)) Date: Mon, 23 Sep 2024 09:28:01 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/wasm-bump Message-ID: <66f16ce1734fd_1f0954328bb04349@gitlab.mail> Cheng Shao pushed new branch wip/wasm-bump at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/wasm-bump You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 13:43:05 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 23 Sep 2024 09:43:05 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/andreask/expose-overloaded-docs Message-ID: <66f17069dc0b9_1f095431ef0c103f@gitlab.mail> Andreas Klebinger pushed new branch wip/andreask/expose-overloaded-docs at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/andreask/expose-overloaded-docs You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 13:43:42 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 23 Sep 2024 09:43:42 -0400 Subject: [Git][ghc/ghc][wip/andreask/expose-overloaded-docs] Reword -fexpose-overloaded-unfoldings docs. Message-ID: <66f1708e5e53f_1f095431d4541057f@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/expose-overloaded-docs at Glasgow Haskell Compiler / GHC Commits: 07e98bde by Andreas Klebinger at 2024-09-23T15:25:07+02:00 Reword -fexpose-overloaded-unfoldings docs. This should make them slightly clearer. Fixes #24844 - - - - - 1 changed file: - docs/users_guide/using-optimisation.rst Changes: ===================================== docs/users_guide/using-optimisation.rst ===================================== @@ -531,7 +531,8 @@ as such you shouldn't need to set any of them explicitly. A flag * Use of the magic `inline` function to force inlining. .. ghc-flag:: -fexpose-overloaded-unfoldings - :shortdesc: Expose unfoldings carrying constraints, even for very large or recursive functions. + :shortdesc: Expose function unfoldings whos type contains constraints, + even for very large or recursive functions. :type: dynamic :reverse: -fno-expose-overloaded-unfoldings :category: @@ -541,19 +542,24 @@ as such you shouldn't need to set any of them explicitly. A flag This experimental flag is a slightly less heavy weight alternative to :ghc-flag:`-fexpose-all-unfoldings`. - Instead of exposing all functions it only aims at exposing constrained functions. + Instead of exposing all functions it exposes only those functions which + contain constraints within their type. This is intended to be used for cases where specialization is considered crucial but :ghc-flag:`-fexpose-all-unfoldings` imposes too much compile time cost. - Currently this won't expose unfoldings where a type class is hidden under a - newtype. That is for cases like: :: + This doesn't guarantee *all* functions which might benefit from specialization + will be exposed, for example currently this won't expose unfoldings where a + type class is hidden under a newtype. + + That is for cases like: :: newtype NT a = NT (Integral a => a) foo :: NT a -> T1 -> TR - GHC won't recognise `foo` as specialisable and won't expose the unfolding + GHC won't look under `NT` for constraints and therefore `foo` won't be + recognized as specialisable and this flag won't expose the unfolding even with :ghc-flag:`-fexpose-overloaded-unfoldings` enabled. All the other caveats about :ghc-flag:`-fexpose-overloaded-unfoldings` View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/07e98bde2989adb29077bae8e29fe073e5948001 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/07e98bde2989adb29077bae8e29fe073e5948001 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 13:59:55 2024 From: gitlab at gitlab.haskell.org (Jade (@Jade)) Date: Mon, 23 Sep 2024 09:59:55 -0400 Subject: [Git][ghc/ghc][wip/structured-ghci-errors] Add structural errors to GHCi (#23338) Message-ID: <66f1745b9654b_1f09546138ac16129@gitlab.mail> Jade pushed to branch wip/structured-ghci-errors at Glasgow Haskell Compiler / GHC Commits: e8f33c12 by Jade at 2024-09-23T15:59:33+02:00 Add structural errors to GHCi (#23338) This patch creates the 'GhciCommandErrorMessage' data type which implents the 'Diagnostic' class and also provides error code for these error conditions. - - - - - 30 changed files: - compiler/GHC/Driver/Errors.hs - compiler/GHC/Driver/Errors/Ppr.hs - compiler/GHC/Driver/Errors/Types.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Ppr.hs - compiler/GHC/HsToCore/Errors/Ppr.hs - compiler/GHC/HsToCore/Errors/Types.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Types/Error.hs - compiler/GHC/Types/Error/Codes.hs - docs/users_guide/9.12.1-notes.rst - ghc/GHCi/UI.hs - ghc/GHCi/UI/Exception.hs - ghc/GHCi/UI/Info.hs - ghc/GHCi/UI/Monad.hs - + ghc/GHCi/UI/Print.hs - ghc/ghc-bin.cabal.in - testsuite/tests/driver/multipleHomeUnits/multiGHCi.stderr - testsuite/tests/ghc-e/should_fail/T18441fail0.stderr - testsuite/tests/ghc-e/should_fail/T18441fail1.stderr - testsuite/tests/ghc-e/should_fail/T18441fail10.stderr - testsuite/tests/ghc-e/should_fail/T18441fail11.stderr - testsuite/tests/ghc-e/should_fail/T18441fail14.stderr - testsuite/tests/ghc-e/should_fail/T18441fail15.stderr - testsuite/tests/ghc-e/should_fail/T18441fail2.stderr - testsuite/tests/ghc-e/should_fail/T18441fail6.stderr - testsuite/tests/ghc-e/should_fail/T18441fail8.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e8f33c1214ec1139a0928e24b8bbf3e02231faff -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e8f33c1214ec1139a0928e24b8bbf3e02231faff You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 14:08:57 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 23 Sep 2024 10:08:57 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] 73 commits: rts: fix checkClosure error message Message-ID: <66f17679a2fd8_2d8bc4bc7001953c@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 72a85e43 by Ben Gamari at 2024-09-23T10:17:30+01:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - 90570d2c by Rodrigo Mesquita at 2024-09-23T12:04:28+01:00 exceptions: Improve the message layout as per #285 This commit fixes the layout of the additional information included when displaying an exception, namely the type of the exception. It also fixes the default handler's heading message to work well together with the improved display message of SomeException. - - - - - dbefa376 by Matthew Pickering at 2024-09-23T14:45:05+01:00 Fix toException method for ExceptionWithContext Fixes #25235 - - - - - 839fcac3 by Matthew Pickering at 2024-09-23T15:07:26+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. - - - - - 21 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b47b517b4ae17cca113c3462e79708a58941a012...839fcac3858569ad589e6548cc11d918da0c95be -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b47b517b4ae17cca113c3462e79708a58941a012...839fcac3858569ad589e6548cc11d918da0c95be You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 14:08:57 2024 From: gitlab at gitlab.haskell.org (Jade (@Jade)) Date: Mon, 23 Sep 2024 10:08:57 -0400 Subject: [Git][ghc/ghc][wip/structured-ghci-errors] 126 commits: haddock: decrease margin on top of small headings Message-ID: <66f17679f0b5b_2d8bc4bc5201969b@gitlab.mail> Jade pushed to branch wip/structured-ghci-errors at Glasgow Haskell Compiler / GHC Commits: af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 09a2a5eb by sheaf at 2024-09-23T16:06:39+02:00 Generalise GHC diagnostic code infrastructure This commit generalises the infrastructure used for diagnostic codes, allowing it to be used for other namespaces than the GHC namespace. In particular, this enables GHCi to re-use the same infrastructure to emit error messages. - - - - - 126ed527 by Jade at 2024-09-23T16:06:39+02:00 Add structural errors to GHCi (#23338) This patch creates the 'GhciCommandErrorMessage' data type which implents the 'Diagnostic' class and also provides error code for these error conditions. - - - - - 25 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e8f33c1214ec1139a0928e24b8bbf3e02231faff...126ed527bb60384f08b4f385749ce3da42978e1f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e8f33c1214ec1139a0928e24b8bbf3e02231faff...126ed527bb60384f08b4f385749ce3da42978e1f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 14:10:21 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Mon, 23 Sep 2024 10:10:21 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25282 Message-ID: <66f176cd7be3d_2d8bc4bc728200a@gitlab.mail> Teo Camarasu pushed new branch wip/T25282 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25282 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 14:17:13 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Mon, 23 Sep 2024 10:17:13 -0400 Subject: [Git][ghc/ghc][wip/T25282] 2 commits: hadrian: Add non-reinstallable package list to generated settings file Message-ID: <66f178692313_2d8bc4bc520295f@gitlab.mail> Teo Camarasu pushed to branch wip/T25282 at Glasgow Haskell Compiler / GHC Commits: fdc96b1e by Teo Camarasu at 2024-09-23T15:16:32+01:00 hadrian: Add non-reinstallable package list to generated settings file This lets Cabal know which packages are not reinstallable. Previously this was a fixed list that was hard-coded into Cabal. See: <https://github.com/haskell/cabal/issues/10087> Resolvces #25282 - - - - - a2bfa54b by Teo Camarasu at 2024-09-23T15:16:55+01:00 Mark template-haskell as reinstallable - - - - - 3 changed files: - hadrian/src/Packages.hs - hadrian/src/Rules/CabalReinstall.hs - hadrian/src/Rules/Generate.hs Changes: ===================================== hadrian/src/Packages.hs ===================================== @@ -12,7 +12,7 @@ module Packages ( runGhc, semaphoreCompat, stm, templateHaskell, terminfo, text, time, timeout, transformers, unlit, unix, win32, xhtml, lintersCommon, lintNotes, lintCodes, lintCommitMsg, lintSubmoduleRefs, lintWhitespace, - ghcPackages, isGhcPackage, + ghcPackages, isGhcPackage, nonReinstallablePackages, -- * Package information crossPrefix, programName, nonHsMainPackage, programPath, timeoutPath, @@ -248,3 +248,29 @@ Cabal requires a few headstands: its face when a relative path is used (e.g. `../cwrapper.c`). Consequently we copy the files into `driver/ghci` in the configure script. -} + + +-- | The set of non-reinstallable packages includes those which cannot be +-- rebuilt using a GHC installation and Hackage-published source distribution. +-- There are a few reasons why this might be true: +-- +-- * the package overrides its unit ID (e.g. with ghc's @-this-unit-id@ flag), +-- which can result in multiple indistinguishable packages (having potentially +-- different ABIs) with the same unit ID. +-- +-- * the package contains definitions of wired-in declarations which tie +-- it to a particular compiler (e.g. we can't build link against +-- @base-4.18.0.0@ using GHC 9.6.1). +-- +-- * the package does not have a complete (that is, buildable) source distribution. +-- For instance, some packages provided by GHC rely on files outside of the +-- source tree generated by GHC's build system. +nonReinstallablePackages :: [Package] +nonReinstallablePackages = + [ base + , ghcBignum + , ghcPrim + , ghc + , integerGmp + , integerSimple + ] ===================================== hadrian/src/Rules/CabalReinstall.hs ===================================== @@ -24,7 +24,7 @@ The libdir of the reinstalled GHC points to the libdir of the stage 2 compiler ( -- | We don't support reinstalling these cabalExcludedPackages :: [Package] -cabalExcludedPackages = [array, base, deepseq, filepath, ghcBignum, ghcBootTh, ghcPrim, integerGmp, integerSimple, pretty, templateHaskell] +cabalExcludedPackages = [array, base, deepseq, filepath, ghcBignum, ghcBootTh, ghcPrim, integerGmp, integerSimple, pretty] cabalBuildRules :: Rules () ===================================== hadrian/src/Rules/Generate.hs ===================================== @@ -505,6 +505,7 @@ generateSettings settingsFile = do , ("Use LibFFI", expr $ yesNo <$> useLibffiForAdjustors) , ("RTS expects libdw", yesNo <$> getFlag UseLibdw) , ("Relative Global Package DB", pure rel_pkg_db) + , ("Non-reinstallable packages", pure . unwords . map pkgName $ nonReinstallablePackages) ] let showTuple (k, v) = "(" ++ show k ++ ", " ++ show v ++ ")" pure $ case settings of View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4e149694f95b6520a1509a06f099659854aacd24...a2bfa54bd8e4dfbebe0d961747fbbfe3ddd23414 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4e149694f95b6520a1509a06f099659854aacd24...a2bfa54bd8e4dfbebe0d961747fbbfe3ddd23414 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 15:08:12 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Mon, 23 Sep 2024 11:08:12 -0400 Subject: [Git][ghc/ghc][wip/T25282] 2 commits: hadrian: Add non-reinstallable package list to generated settings file Message-ID: <66f1845ca2db6_2d8bc4535c78298bd@gitlab.mail> Teo Camarasu pushed to branch wip/T25282 at Glasgow Haskell Compiler / GHC Commits: dbb4b707 by Teo Camarasu at 2024-09-23T16:07:59+01:00 hadrian: Add non-reinstallable package list to generated settings file This lets Cabal know which packages are not reinstallable. Previously this was a fixed list that was hard-coded into Cabal. See: <https://github.com/haskell/cabal/issues/10087> Resolvces #25282 - - - - - 02ce16f9 by Teo Camarasu at 2024-09-23T16:07:59+01:00 Mark template-haskell as reinstallable - - - - - 3 changed files: - hadrian/src/Packages.hs - hadrian/src/Rules/CabalReinstall.hs - hadrian/src/Rules/Generate.hs Changes: ===================================== hadrian/src/Packages.hs ===================================== @@ -12,7 +12,7 @@ module Packages ( runGhc, semaphoreCompat, stm, templateHaskell, terminfo, text, time, timeout, transformers, unlit, unix, win32, xhtml, lintersCommon, lintNotes, lintCodes, lintCommitMsg, lintSubmoduleRefs, lintWhitespace, - ghcPackages, isGhcPackage, + ghcPackages, isGhcPackage, nonReinstallablePackages, -- * Package information crossPrefix, programName, nonHsMainPackage, programPath, timeoutPath, @@ -248,3 +248,30 @@ Cabal requires a few headstands: its face when a relative path is used (e.g. `../cwrapper.c`). Consequently we copy the files into `driver/ghci` in the configure script. -} + + +-- | The set of non-reinstallable packages includes those which cannot be +-- rebuilt using a GHC installation and Hackage-published source distribution. +-- There are a few reasons why this might be true: +-- +-- * the package overrides its unit ID (e.g. with ghc's @-this-unit-id@ flag), +-- which can result in multiple indistinguishable packages (having potentially +-- different ABIs) with the same unit ID. +-- +-- * the package contains definitions of wired-in declarations which tie +-- it to a particular compiler (e.g. we can't build link against +-- @base-4.18.0.0@ using GHC 9.6.1). +-- +-- * the package does not have a complete (that is, buildable) source distribution. +-- For instance, some packages provided by GHC rely on files outside of the +-- source tree generated by GHC's build system. +nonReinstallablePackages :: [Package] +nonReinstallablePackages = + [ base + , ghcBignum + , ghcInternal + , ghcPrim + , ghc + , integerGmp + , integerSimple + ] ===================================== hadrian/src/Rules/CabalReinstall.hs ===================================== @@ -24,7 +24,7 @@ The libdir of the reinstalled GHC points to the libdir of the stage 2 compiler ( -- | We don't support reinstalling these cabalExcludedPackages :: [Package] -cabalExcludedPackages = [array, base, deepseq, filepath, ghcBignum, ghcBootTh, ghcPrim, integerGmp, integerSimple, pretty, templateHaskell] +cabalExcludedPackages = [array, base, deepseq, filepath, ghcBignum, ghcBootTh, ghcPrim, integerGmp, integerSimple, pretty] cabalBuildRules :: Rules () ===================================== hadrian/src/Rules/Generate.hs ===================================== @@ -505,6 +505,7 @@ generateSettings settingsFile = do , ("Use LibFFI", expr $ yesNo <$> useLibffiForAdjustors) , ("RTS expects libdw", yesNo <$> getFlag UseLibdw) , ("Relative Global Package DB", pure rel_pkg_db) + , ("Non-reinstallable packages", pure . unwords . map pkgName $ nonReinstallablePackages) ] let showTuple (k, v) = "(" ++ show k ++ ", " ++ show v ++ ")" pure $ case settings of View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a2bfa54bd8e4dfbebe0d961747fbbfe3ddd23414...02ce16f941f5e4d3e948f8810b3ce5637178e21e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a2bfa54bd8e4dfbebe0d961747fbbfe3ddd23414...02ce16f941f5e4d3e948f8810b3ce5637178e21e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 15:47:06 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Mon, 23 Sep 2024 11:47:06 -0400 Subject: [Git][ghc/ghc][wip/T25282] Remove integer-gmp, integer-simple from non-reinst list Message-ID: <66f18d7a29c49_2d8bc47777e8438e6@gitlab.mail> Teo Camarasu pushed to branch wip/T25282 at Glasgow Haskell Compiler / GHC Commits: 09b12cf1 by Teo Camarasu at 2024-09-23T16:46:45+01:00 Remove integer-gmp, integer-simple from non-reinst list - - - - - 2 changed files: - hadrian/src/Packages.hs - hadrian/src/Rules/CabalReinstall.hs Changes: ===================================== hadrian/src/Packages.hs ===================================== @@ -272,6 +272,4 @@ nonReinstallablePackages = , ghcInternal , ghcPrim , ghc - , integerGmp - , integerSimple ] ===================================== hadrian/src/Rules/CabalReinstall.hs ===================================== @@ -24,7 +24,7 @@ The libdir of the reinstalled GHC points to the libdir of the stage 2 compiler ( -- | We don't support reinstalling these cabalExcludedPackages :: [Package] -cabalExcludedPackages = [array, base, deepseq, filepath, ghcBignum, ghcBootTh, ghcPrim, integerGmp, integerSimple, pretty] +cabalExcludedPackages = [array, base, deepseq, filepath, ghcBignum, ghcBootTh, ghcPrim, pretty] cabalBuildRules :: Rules () View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/09b12cf11379463283f0ff42c13fc5071a17a1cb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/09b12cf11379463283f0ff42c13fc5071a17a1cb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 15:51:59 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 23 Sep 2024 11:51:59 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] 2 commits: Add unicode drawing lines Message-ID: <66f18e9ef384b_2d8bc476ee5444196@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: c2a4ffed by Rodrigo Mesquita at 2024-09-23T16:43:51+01:00 Add unicode drawing lines - - - - - 52a6fde3 by Rodrigo Mesquita at 2024-09-23T16:51:45+01:00 Accept tests? - - - - - 25 changed files: - libraries/base/tests/IO/T4808.stderr - libraries/base/tests/IO/mkdirExists.stderr - libraries/base/tests/IO/openFile002.stderr - libraries/base/tests/IO/withBinaryFile001.stderr - libraries/base/tests/IO/withBinaryFile002.stderr - libraries/base/tests/IO/withFile001.stderr - libraries/base/tests/IO/withFile002.stderr - libraries/base/tests/IO/withFileBlocking001.stderr - libraries/base/tests/IO/withFileBlocking002.stderr - libraries/base/tests/T15349.stderr - libraries/base/tests/T24807.stderr - libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs - testsuite/tests/codeGen/should_run/cgrun016.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/ffi/should_run/T7170.stderr - testsuite/tests/ghc-e/should_fail/T18441fail2.stderr - testsuite/tests/ghc-e/should_fail/T18441fail7.stderr - testsuite/tests/ghc-e/should_fail/T18441fail8.stderr - testsuite/tests/ghc-e/should_fail/T23663.stderr - testsuite/tests/ghc-e/should_fail/T9930fail.stderr - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/mdo/should_fail/mdofail006.stderr - testsuite/tests/runghc/T7859.stderr - testsuite/tests/typecheck/should_compile/holes.stderr - testsuite/tests/typecheck/should_compile/holes3.stderr Changes: ===================================== libraries/base/tests/IO/T4808.stderr ===================================== @@ -4,7 +4,7 @@ T4808.test: hGetLine: illegal operation (handle is closed) HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception - ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception + ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs:347:20 in ghc-internal:GHC.Internal.IO.Handle.Internals ===================================== libraries/base/tests/IO/mkdirExists.stderr ===================================== @@ -4,7 +4,8 @@ foo: createDirectory: already exists (File exists) HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception + ioError, called at libraries/unix/System/Posix/PosixPath/FilePath.hsc:106:5 in unix-2.8.5.1-inplace:System.Posix.PosixPath.FilePath ===================================== libraries/base/tests/IO/openFile002.stderr ===================================== @@ -4,7 +4,8 @@ nonexistent: openFile: does not exist (No such file or directory) HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception + ioError, called at libraries/ghc-internal/src/GHC/Internal/Foreign/C/Error.hs:291:5 in ghc-internal:GHC.Internal.Foreign.C.Error ===================================== libraries/base/tests/IO/withBinaryFile001.stderr ===================================== @@ -4,7 +4,8 @@ test.bin: withBinaryFile: does not exist (No such file or directory) HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception + ioError, called at libraries/ghc-internal/src/GHC/Internal/Foreign/C/Error.hs:291:5 in ghc-internal:GHC.Internal.Foreign.C.Error ===================================== libraries/base/tests/IO/withBinaryFile002.stderr ===================================== @@ -1,10 +1,3 @@ withBinaryFile002: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.IOException: user error (test) - -HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO - throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception - ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception - ===================================== libraries/base/tests/IO/withFile001.stderr ===================================== @@ -4,7 +4,8 @@ test.txt: withFile: does not exist (No such file or directory) HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception + ioError, called at libraries/ghc-internal/src/GHC/Internal/Foreign/C/Error.hs:291:5 in ghc-internal:GHC.Internal.Foreign.C.Error ===================================== libraries/base/tests/IO/withFile002.stderr ===================================== @@ -1,10 +1,3 @@ withFile002: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.IOException: user error (test) - -HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO - throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception - ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception - ===================================== libraries/base/tests/IO/withFileBlocking001.stderr ===================================== @@ -4,7 +4,8 @@ test.txt: withFileBlocking: does not exist (No such file or directory) HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception + ioError, called at libraries/ghc-internal/src/GHC/Internal/Foreign/C/Error.hs:291:5 in ghc-internal:GHC.Internal.Foreign.C.Error ===================================== libraries/base/tests/IO/withFileBlocking002.stderr ===================================== @@ -1,10 +1,3 @@ withFileBlocking002: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.IOException: user error (test) - -HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO - throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception - ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception - ===================================== libraries/base/tests/T15349.stderr ===================================== @@ -2,8 +2,12 @@ T15349: Uncaught exception ghc-internal:GHC.Internal.Control.Exception.Base.NonT <> +While handling ghc-internal:GHC.Internal.IO.Exception.BlockedIndefinitelyOnMVar: + │ + │ thread blocked indefinitely in an MVar operation + HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at libraries/ghc-internal/src/GHC/Internal/Control/Monad/ST/Imp.hs:58:37 in ghc-internal:GHC.Internal.Control.Monad.ST.Imp ===================================== libraries/base/tests/T24807.stderr ===================================== @@ -1,11 +1,11 @@ -T24807: Exception: +T24807: Uncaught exception ghc-internal:GHC.Internal.Exception.ErrorCall: + hi CallStack (from HasCallStack): error, called at T24807.hs:1:8 in main:Main -Package: ghc-internal -Module: GHC.Internal.Exception -Type: ErrorCall + HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:127:5 in ghc-internal:GHC.Internal.Exception error, called at T24807.hs:1:8 in main:Main + ===================================== libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs ===================================== @@ -86,7 +86,7 @@ instance ExceptionAnnotation WhileHandling where (l1:ls) -> -- Indent lines forward. -- displayException may be ill prepared for this?... - unlines $ l1:[if null l then l else " " ++ l | l <- ls] + unlines $ l1:[if null l then " │" else " │ " ++ l | l <- ls] -- | Create 'SomeException' from an explicit context and exception. ===================================== testsuite/tests/codeGen/should_run/cgrun016.stderr ===================================== @@ -3,3 +3,9 @@ cgrun016: Uncaught exception ghc-internal:GHC.Internal.Exception.ErrorCall: 6th call to error CallStack (from HasCallStack): error, called at cgrun016.hs:8:8 in main:Main + +HasCallStack backtrace: + collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:127:5 in ghc-internal:GHC.Internal.Exception + error, called at cgrun016.hs:8:8 in main:Main + ===================================== testsuite/tests/codeGen/should_run/cgrun025.stderr ===================================== @@ -33,6 +33,16 @@ hello, error CallStack (from HasCallStack): error, called at cgrun025.hs:25:75 in main:Main +While handling ghc-internal:GHC.Internal.IO.Exception.IOException: + │ + │ __WURBLE__: getEnv: does not exist (no environment variable) + │ + │ HasCallStack backtrace: + │ collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + │ toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO + │ throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception + │ ioException, called at libraries/ghc-internal/src/GHC/Internal/System/Environment.hs:192:26 in ghc-internal:GHC.Internal.System.Environment + HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:127:5 in ghc-internal:GHC.Internal.Exception ===================================== testsuite/tests/ffi/should_run/T7170.stderr ===================================== @@ -1,9 +1,3 @@ T7170: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.BlockedIndefinitelyOnMVar: thread blocked indefinitely in an MVar operation - -HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO - throwIO, called at libraries/ghc-internal/src/GHC/Internal/Control/Exception/Base.hs:195:43 in ghc-internal:GHC.Internal.Control.Exception.Base - ===================================== testsuite/tests/ghc-e/should_fail/T18441fail2.stderr ===================================== @@ -3,9 +3,30 @@ unrecognised flag: -Xabcde +While handling ghc-9.11-inplace:GHC.Utils.Panic.GhcException: + │ + │ unrecognised flag: -Xabcde + │ + │ + │ HasCallStack backtrace: + │ collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + │ toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO + │ throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at compiler/GHC/Driver/Monad.hs:167:54 in ghc-9.11-inplace:GHC.Driver.Monad + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at ghc/GHCi/UI/Monad.hs:288:15 in ghc-bin-9.11.20240923-inplace:GHCi.UI.Monad + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/haskeline/System/Console/Haskeline/InputT.hs:53:39 in haskeline-0.8.2.1-inplace:System.Console.Haskeline.InputT + │ throwM, called at ghc/GHCi/UI/Monad.hs:215:52 in ghc-bin-9.11.20240923-inplace:GHCi.UI.Monad + HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at compiler/GHC/Utils/Error.hs:512:19 in ghc-9.11-inplace:GHC.Utils.Error 1 ===================================== testsuite/tests/ghc-e/should_fail/T18441fail7.stderr ===================================== @@ -2,9 +2,29 @@ IO error: "Abcde" does not exist +While handling ghc-9.11-inplace:GHC.Utils.Panic.GhcException: + │ + │ IO error: "Abcde" does not exist + │ + │ HasCallStack backtrace: + │ collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + │ toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO + │ throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at compiler/GHC/Driver/Monad.hs:167:54 in ghc-9.11-inplace:GHC.Driver.Monad + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at ghc/GHCi/UI/Monad.hs:288:15 in ghc-bin-9.11.20240923-inplace:GHCi.UI.Monad + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/haskeline/System/Console/Haskeline/InputT.hs:53:39 in haskeline-0.8.2.1-inplace:System.Console.Haskeline.InputT + │ throwM, called at ghc/GHCi/UI/Monad.hs:215:52 in ghc-bin-9.11.20240923-inplace:GHCi.UI.Monad + HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at compiler/GHC/Utils/Error.hs:512:19 in ghc-9.11-inplace:GHC.Utils.Error 1 ===================================== testsuite/tests/ghc-e/should_fail/T18441fail8.stderr ===================================== @@ -2,9 +2,29 @@ syntax: :script +While handling ghc-9.11-inplace:GHC.Utils.Panic.GhcException: + │ + │ syntax: :script + │ + │ HasCallStack backtrace: + │ collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + │ toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO + │ throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at compiler/GHC/Driver/Monad.hs:167:54 in ghc-9.11-inplace:GHC.Driver.Monad + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at ghc/GHCi/UI/Monad.hs:288:15 in ghc-bin-9.11.20240923-inplace:GHCi.UI.Monad + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/haskeline/System/Console/Haskeline/InputT.hs:53:39 in haskeline-0.8.2.1-inplace:System.Console.Haskeline.InputT + │ throwM, called at ghc/GHCi/UI/Monad.hs:215:52 in ghc-bin-9.11.20240923-inplace:GHCi.UI.Monad + HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at compiler/GHC/Utils/Error.hs:512:19 in ghc-9.11-inplace:GHC.Utils.Error 1 ===================================== testsuite/tests/ghc-e/should_fail/T23663.stderr ===================================== @@ -5,9 +5,32 @@ did you mean one of: -XCUSKs +While handling ghc-9.11-inplace:GHC.Utils.Panic.GhcException: + │ + │ unrecognised flag: -XCUSKS + │ did you mean one of: + │ -XCUSKs + │ + │ + │ HasCallStack backtrace: + │ collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + │ toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO + │ throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at compiler/GHC/Driver/Monad.hs:167:54 in ghc-9.11-inplace:GHC.Driver.Monad + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at ghc/GHCi/UI/Monad.hs:288:15 in ghc-bin-9.11.20240923-inplace:GHCi.UI.Monad + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:490:21 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/haskeline/System/Console/Haskeline/InputT.hs:53:39 in haskeline-0.8.2.1-inplace:System.Console.Haskeline.InputT + │ throwM, called at ghc/GHCi/UI/Monad.hs:215:52 in ghc-bin-9.11.20240923-inplace:GHCi.UI.Monad + HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at compiler/GHC/Utils/Error.hs:512:19 in ghc-9.11-inplace:GHC.Utils.Error 1 ===================================== testsuite/tests/ghc-e/should_fail/T9930fail.stderr ===================================== @@ -3,6 +3,36 @@ ghc: Uncaught exception ghc-9.11-inplace:GHC.Utils.Panic.GhcException: default output name would overwrite the input file; must specify -o explicitly Usage: For basic information, try the `--help' option. +While handling ghc-9.11-inplace:GHC.Utils.Panic.GhcException: + │ + │ default output name would overwrite the input file; must specify -o explicitly + │ Usage: For basic information, try the `--help' option. + │ + │ While handling ghc-9.11-inplace:GHC.Utils.Panic.GhcException: + │ │ + │ │ default output name would overwrite the input file; must specify -o explicitly + │ │ Usage: For basic information, try the `--help' option. + │ │ + │ │ While handling ghc-9.11-inplace:GHC.Utils.Panic.GhcException: + │ │ │ + │ │ │ default output name would overwrite the input file; must specify -o explicitly + │ │ │ Usage: For basic information, try the `--help' option. + │ │ │ + │ │ │ HasCallStack backtrace: + │ │ │ collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + │ │ │ toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:84:32 in ghc-internal:GHC.Internal.Exception + │ │ │ throw, called at compiler/GHC/Utils/Panic.hs:180:21 in ghc-9.11-inplace:GHC.Utils.Panic + │ │ + │ │ HasCallStack backtrace: + │ │ bracket_, called at libraries/semaphore-compat/src/System/Semaphore.hs:320:23 in semaphore-compat-1.0.0-inplace:System.Semaphore + │ + │ HasCallStack backtrace: + │ collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + │ toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO + │ throwIO, called at libraries/exceptions/src/Control/Monad/Catch.hs:371:12 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ throwM, called at libraries/exceptions/src/Control/Monad/Catch.hs:860:84 in exceptions-0.10.7-inplace:Control.Monad.Catch + │ onException, called at compiler/GHC/Driver/Make.hs:2986:23 in ghc-9.11-inplace:GHC.Driver.Make + HasCallStack backtrace: bracket, called at compiler/GHC/Driver/Make.hs:2953:3 in ghc-9.11-inplace:GHC.Driver.Make ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -234,7 +234,7 @@ module Control.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -268,6 +268,8 @@ module Control.Exception where data SomeException = forall e. (Exception e, GHC.Internal.Exception.Type.HasExceptionContext) => SomeException e type TypeError :: * newtype TypeError = TypeError GHC.Internal.Base.String + type WhileHandling :: * + data WhileHandling = WhileHandling SomeException addExceptionContext :: forall a. GHC.Internal.Exception.Context.ExceptionAnnotation a => a -> SomeException -> SomeException allowInterrupt :: GHC.Types.IO () annotateIO :: forall e a. GHC.Internal.Exception.Context.ExceptionAnnotation e => e -> GHC.Types.IO a -> GHC.Types.IO a @@ -279,6 +281,7 @@ module Control.Exception where bracket_ :: forall a b c. GHC.Types.IO a -> GHC.Types.IO b -> GHC.Types.IO c -> GHC.Types.IO c catch :: forall e a. Exception e => GHC.Types.IO a -> (e -> GHC.Types.IO a) -> GHC.Types.IO a catchJust :: forall e b a. Exception e => (e -> GHC.Internal.Maybe.Maybe b) -> GHC.Types.IO a -> (b -> GHC.Types.IO a) -> GHC.Types.IO a + catchNoPropagate :: forall e a. Exception e => GHC.Types.IO a -> (ExceptionWithContext e -> GHC.Types.IO a) -> GHC.Types.IO a catches :: forall a. GHC.Types.IO a -> [Handler a] -> GHC.Types.IO a evaluate :: forall a. a -> GHC.Types.IO a finally :: forall a b. GHC.Types.IO a -> GHC.Types.IO b -> GHC.Types.IO a @@ -286,24 +289,26 @@ module Control.Exception where handle :: forall e a. Exception e => (e -> GHC.Types.IO a) -> GHC.Types.IO a -> GHC.Types.IO a handleJust :: forall e b a. Exception e => (e -> GHC.Internal.Maybe.Maybe b) -> (b -> GHC.Types.IO a) -> GHC.Types.IO a -> GHC.Types.IO a interruptible :: forall a. GHC.Types.IO a -> GHC.Types.IO a - ioError :: forall a. GHC.Internal.IO.Exception.IOError -> GHC.Types.IO a + ioError :: forall a. GHC.Internal.Stack.Types.HasCallStack => GHC.Internal.IO.Exception.IOError -> GHC.Types.IO a mapException :: forall e1 e2 a. (Exception e1, Exception e2) => (e1 -> e2) -> a -> a mask :: forall b. ((forall a. GHC.Types.IO a -> GHC.Types.IO a) -> GHC.Types.IO b) -> GHC.Types.IO b mask_ :: forall a. GHC.Types.IO a -> GHC.Types.IO a onException :: forall a b. GHC.Types.IO a -> GHC.Types.IO b -> GHC.Types.IO a + rethrowIO :: forall e a. Exception e => ExceptionWithContext e -> GHC.Types.IO a someExceptionContext :: SomeException -> GHC.Internal.Exception.Context.ExceptionContext throw :: forall (r :: GHC.Types.RuntimeRep) (a :: TYPE r) e. (GHC.Internal.Stack.Types.HasCallStack, Exception e) => e -> a throwIO :: forall e a. (GHC.Internal.Stack.Types.HasCallStack, Exception e) => e -> GHC.Types.IO a throwTo :: forall e. Exception e => GHC.Internal.Conc.Sync.ThreadId -> e -> GHC.Types.IO () try :: forall e a. Exception e => GHC.Types.IO a -> GHC.Types.IO (GHC.Internal.Data.Either.Either e a) tryJust :: forall e b a. Exception e => (e -> GHC.Internal.Maybe.Maybe b) -> GHC.Types.IO a -> GHC.Types.IO (GHC.Internal.Data.Either.Either b a) + tryWithContext :: forall e a. Exception e => GHC.Types.IO a -> GHC.Types.IO (GHC.Internal.Data.Either.Either (ExceptionWithContext e) a) uninterruptibleMask :: forall b. ((forall a. GHC.Types.IO a -> GHC.Types.IO a) -> GHC.Types.IO b) -> GHC.Types.IO b uninterruptibleMask_ :: forall a. GHC.Types.IO a -> GHC.Types.IO a module Control.Exception.Annotation where -- Safety: None type ExceptionAnnotation :: * -> Constraint - class ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where + class ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => ExceptionAnnotation a where displayExceptionAnnotation :: a -> GHC.Internal.Base.String default displayExceptionAnnotation :: GHC.Internal.Show.Show a => a -> GHC.Internal.Base.String {-# MINIMAL #-} @@ -345,7 +350,7 @@ module Control.Exception.Base where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -394,7 +399,7 @@ module Control.Exception.Base where handleJust :: forall e b a. Exception e => (e -> GHC.Internal.Maybe.Maybe b) -> (b -> GHC.Types.IO a) -> GHC.Types.IO a -> GHC.Types.IO a impossibleConstraintError :: forall (q :: GHC.Types.RuntimeRep) (a :: GHC.Prim.CONSTRAINT q). GHC.Prim.Addr# -=> a impossibleError :: forall (q :: GHC.Types.RuntimeRep) (a :: TYPE q). GHC.Prim.Addr# -> a - ioError :: forall a. GHC.Internal.IO.Exception.IOError -> GHC.Types.IO a + ioError :: forall a. GHC.Internal.Stack.Types.HasCallStack => GHC.Internal.IO.Exception.IOError -> GHC.Types.IO a mapException :: forall e1 e2 a. (Exception e1, Exception e2) => (e1 -> e2) -> a -> a mask :: forall b. ((forall a. GHC.Types.IO a -> GHC.Types.IO a) -> GHC.Types.IO b) -> GHC.Types.IO b mask_ :: forall a. GHC.Types.IO a -> GHC.Types.IO a @@ -850,11 +855,11 @@ module Data.Data where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b constrFields :: Constr -> [GHC.Internal.Base.String] constrFixity :: Constr -> Fixity @@ -897,7 +902,7 @@ module Data.Data where showConstr :: Constr -> GHC.Internal.Base.String showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -921,14 +926,14 @@ module Data.Dynamic where -- Safety: Safe type Dynamic :: * data Dynamic where - Dynamic :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic + Dynamic :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a -> a -> Dynamic type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} dynApp :: Dynamic -> Dynamic -> Dynamic dynApply :: Dynamic -> Dynamic -> GHC.Internal.Maybe.Maybe Dynamic - dynTypeRep :: Dynamic -> ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + dynTypeRep :: Dynamic -> ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep fromDyn :: forall a. Typeable a => Dynamic -> a -> a fromDynamic :: forall a. Typeable a => Dynamic -> GHC.Internal.Maybe.Maybe a toDyn :: forall a. Typeable a => a -> Dynamic @@ -1815,11 +1820,11 @@ module Data.Typeable where type TyCon :: * data TyCon = ... type TypeRep :: * - type TypeRep = ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep + type TypeRep = ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} cast :: forall a b. (Typeable a, Typeable b) => a -> GHC.Internal.Maybe.Maybe b decT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Data.Either.Either ((a :~: b) -> GHC.Internal.Base.Void) (a :~: b) eqT :: forall {k} (a :: k) (b :: k). (Typeable a, Typeable b) => GHC.Internal.Maybe.Maybe (a :~: b) @@ -1834,7 +1839,7 @@ module Data.Typeable where rnfTypeRep :: TypeRep -> () showsTypeRep :: TypeRep -> GHC.Internal.Show.ShowS splitTyConApp :: TypeRep -> (TyCon, [TypeRep]) - trLiftedRep :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep + trLiftedRep :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep GHC.Types.LiftedRep tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConModule :: TyCon -> GHC.Internal.Base.String tyConName :: TyCon -> GHC.Internal.Base.String @@ -5255,7 +5260,7 @@ module GHC.Event where data EventManager = ... type FdKey :: * data FdKey - = ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] + = ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey {keyFd :: ! {-# UNPACK #-}(GHC.Internal.System.Posix.Types.N:Fd[0] ; GHC.Internal.Foreign.C.Types.N:CInt[0])GHC.Internal.System.Posix.Types.Fd, ...} type IOCallback :: * @@ -5288,9 +5293,9 @@ module GHC.Event.TimeOut where type TimeoutEdit :: * type TimeoutEdit = TimeoutQueue -> TimeoutQueue type TimeoutKey :: * - newtype TimeoutKey = TK ghc-internal-9.1001.0:GHC.Internal.Event.Unique.Unique + newtype TimeoutKey = TK ghc-internal-9.1100.0:GHC.Internal.Event.Unique.Unique type TimeoutQueue :: * - type TimeoutQueue = ghc-internal-9.1001.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback + type TimeoutQueue = ghc-internal-9.1100.0:GHC.Internal.Event.PSQ.PSQ TimeoutCallback module GHC.Exception where -- Safety: Safe @@ -5302,7 +5307,7 @@ module GHC.Exception where type ErrorCall :: * data ErrorCall = ErrorCallWithLocation GHC.Internal.Base.String GHC.Internal.Base.String type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -5331,7 +5336,7 @@ module GHC.Exception.Type where type ArithException :: * data ArithException = Overflow | Underflow | LossOfPrecision | DivideByZero | Denormal | RatioZeroDenominator type Exception :: * -> Constraint - class (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where + class (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable e, GHC.Internal.Show.Show e) => Exception e where toException :: e -> SomeException fromException :: SomeException -> GHC.Internal.Maybe.Maybe e displayException :: e -> GHC.Internal.Base.String @@ -7503,7 +7508,7 @@ module GHC.IO where data MaskingState = Unmasked | MaskedInterruptible | MaskedUninterruptible bracket :: forall a b c. IO a -> (a -> IO b) -> (a -> IO c) -> IO c catch :: forall e a. GHC.Internal.Exception.Type.Exception e => IO a -> (e -> IO a) -> IO a - catchAny :: forall a. IO a -> (forall e. GHC.Internal.Exception.Type.Exception e => e -> IO a) -> IO a + catchAny :: forall a. IO a -> (forall e. (GHC.Internal.Exception.Type.HasExceptionContext, GHC.Internal.Exception.Type.Exception e) => e -> IO a) -> IO a catchException :: forall e a. GHC.Internal.Exception.Type.Exception e => IO a -> (e -> IO a) -> IO a evaluate :: forall a. a -> IO a finally :: forall a b. IO a -> IO b -> IO a @@ -7800,7 +7805,7 @@ module GHC.IO.Exception where cannotCompactMutable :: GHC.Internal.Exception.Type.SomeException cannotCompactPinned :: GHC.Internal.Exception.Type.SomeException heapOverflow :: GHC.Internal.Exception.Type.SomeException - ioError :: forall a. IOError -> GHC.Types.IO a + ioError :: forall a. GHC.Internal.Stack.Types.HasCallStack => IOError -> GHC.Types.IO a ioException :: forall a. GHC.Internal.Stack.Types.HasCallStack => IOException -> GHC.Types.IO a stackOverflow :: GHC.Internal.Exception.Type.SomeException unsupportedOperation :: IOError @@ -7884,8 +7889,8 @@ module GHC.IO.Handle where hTryLock :: Handle -> LockMode -> GHC.Types.IO GHC.Types.Bool hWaitForInput :: Handle -> GHC.Types.Int -> GHC.Types.IO GHC.Types.Bool isEOF :: GHC.Types.IO GHC.Types.Bool - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> NewlineMode -> GHC.Types.IO Handle nativeNewline :: Newline nativeNewlineMode :: NewlineMode noNewlineTranslation :: NewlineMode @@ -7935,11 +7940,11 @@ module GHC.IO.Handle.Internals where ioe_notReadable :: forall a. GHC.Types.IO a ioe_notWritable :: forall a. GHC.Types.IO a ioe_semiclosedHandle :: forall a. GHC.Types.IO a - mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle - mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkDuplexHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkFileHandleNoFinalizer :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.IOMode.IOMode -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle + mkHandle :: forall dev. (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => dev -> GHC.Internal.IO.FilePath -> GHC.Internal.IO.Handle.Types.HandleType -> GHC.Types.Bool -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.NewlineMode -> GHC.Internal.Maybe.Maybe HandleFinalizer -> GHC.Internal.Maybe.Maybe (GHC.Internal.MVar.MVar GHC.Internal.IO.Handle.Types.Handle__) -> GHC.Types.IO GHC.Internal.IO.Handle.Types.Handle openTextEncoding :: forall a. GHC.Internal.Maybe.Maybe GHC.Internal.IO.Encoding.Types.TextEncoding -> GHC.Internal.IO.Handle.Types.HandleType -> (forall es ds. GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextEncoder es) -> GHC.Internal.Maybe.Maybe (GHC.Internal.IO.Encoding.Types.TextDecoder ds) -> GHC.Types.IO a) -> GHC.Types.IO a readTextDevice :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer readTextDeviceNonBlocking :: GHC.Internal.IO.Handle.Types.Handle__ -> GHC.Internal.IO.Buffer.CharBuffer -> GHC.Types.IO GHC.Internal.IO.Buffer.CharBuffer @@ -8003,7 +8008,7 @@ module GHC.IO.Handle.Types where type Handle__ :: * data Handle__ = forall dev enc_state dec_state. - (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => + (GHC.Internal.IO.Device.RawIO dev, GHC.Internal.IO.Device.IODevice dev, GHC.Internal.IO.BufferedIO.BufferedIO dev, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable dev) => Handle__ {haDevice :: !dev, haType :: HandleType, haByteBuffer :: ! {-# UNPACK #-}(GHC.Internal.IORef.N:IORef[0] _N)(GHC.Internal.IORef.IORef (GHC.Internal.IO.Buffer.Buffer GHC.Internal.Word.Word8)), @@ -9417,7 +9422,7 @@ module GHC.StaticPtr where -- Safety: None type IsStatic :: (* -> *) -> Constraint class IsStatic p where - fromStaticPtr :: forall a. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a + fromStaticPtr :: forall a. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a => StaticPtr a -> p a {-# MINIMAL fromStaticPtr #-} type StaticKey :: * type StaticKey = GHC.Internal.Fingerprint.Type.Fingerprint @@ -10133,7 +10138,7 @@ module Prelude where id :: forall a. a -> a init :: forall a. GHC.Internal.Stack.Types.HasCallStack => [a] -> [a] interact :: (String -> String) -> IO () - ioError :: forall a. IOError -> IO a + ioError :: forall a. GHC.Internal.Stack.Types.HasCallStack => IOError -> IO a iterate :: forall a. (a -> a) -> a -> [a] last :: forall a. GHC.Internal.Stack.Types.HasCallStack => [a] -> a lcm :: forall a. Integral a => a -> a -> a @@ -10372,7 +10377,7 @@ module System.IO.Error where eofErrorType :: IOErrorType fullErrorType :: IOErrorType illegalOperationErrorType :: IOErrorType - ioError :: forall a. IOError -> GHC.Types.IO a + ioError :: forall a. GHC.Internal.Stack.Types.HasCallStack => IOError -> GHC.Types.IO a ioeGetErrorString :: IOError -> GHC.Internal.Base.String ioeGetErrorType :: IOError -> IOErrorType ioeGetFileName :: IOError -> GHC.Internal.Maybe.Maybe GHC.Internal.IO.FilePath @@ -10814,8 +10819,8 @@ module Type.Reflection where data (:~~:) a b where HRefl :: forall {k1} (a :: k1). (:~~:) a a pattern App :: forall k2 (t :: k2). () => forall k1 (a :: k1 -> k2) (b :: k1). (t ~ a b) => TypeRep a -> TypeRep b -> TypeRep t - pattern Con :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a - pattern Con' :: forall k (a :: k). () => ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a + pattern Con :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> TypeRep a + pattern Con' :: forall k (a :: k). () => ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.NotApplication a => TyCon -> [SomeTypeRep] -> TypeRep a pattern Fun :: forall k (fun :: k). () => forall (r1 :: GHC.Types.RuntimeRep) (r2 :: GHC.Types.RuntimeRep) (arg :: TYPE r1) (res :: TYPE r2). (k ~ *, fun ~~ (arg -> res)) => TypeRep arg -> TypeRep res -> TypeRep fun type Module :: * data Module = ... @@ -10832,7 +10837,7 @@ module Type.Reflection where type Typeable :: forall k. k -> Constraint class Typeable a where ... - {-# MINIMAL ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} + {-# MINIMAL ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.typeRep# #-} decTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Data.Either.Either ((a :~~: b) -> GHC.Internal.Base.Void) (a :~~: b) eqTypeRep :: forall k1 k2 (a :: k1) (b :: k2). TypeRep a -> TypeRep b -> GHC.Internal.Maybe.Maybe (a :~~: b) moduleName :: Module -> GHC.Internal.Base.String @@ -10867,9 +10872,9 @@ module Type.Reflection.Unsafe where data TypeRep a where ... mkTrApp :: forall k1 k2 (a :: k1 -> k2) (b :: k1). TypeRep a -> TypeRep b -> TypeRep (a b) - mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a + mkTrCon :: forall k (a :: k). TyCon -> [ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep] -> TypeRep a mkTyCon :: GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Internal.Base.String -> GHC.Types.Int -> KindRep -> TyCon - someTypeRepFingerprint :: ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint + someTypeRepFingerprint :: ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -> GHC.Internal.Fingerprint.Type.Fingerprint tyConFingerprint :: TyCon -> GHC.Internal.Fingerprint.Type.Fingerprint tyConKindArgs :: TyCon -> GHC.Types.Int tyConKindRep :: TyCon -> KindRep @@ -11203,9 +11208,9 @@ instance forall a. GHC.Internal.Base.Monoid a => GHC.Internal.Base.Monoid (GHC.I instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. (GHC.Classes.Ord a, GHC.Internal.Enum.Bounded a) => GHC.Internal.Base.Monoid (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Monoid (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Monoid ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Monoid ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Monoid (f p), GHC.Internal.Base.Monoid (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Monoid (f (g p)) => GHC.Internal.Base.Monoid ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Monoid (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Monoid (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11261,9 +11266,9 @@ instance forall a. GHC.Internal.Base.Semigroup (Data.Semigroup.Last a) -- Define instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Max a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Classes.Ord a => GHC.Internal.Base.Semigroup (Data.Semigroup.Min a) -- Defined in ‘Data.Semigroup’ instance forall m. GHC.Internal.Base.Monoid m => GHC.Internal.Base.Semigroup (Data.Semigroup.WrappedMonoid m) -- Defined in ‘Data.Semigroup’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Base.Semigroup ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Base.Semigroup ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ instance forall k (f :: k -> *) (p :: k) (g :: k -> *). (GHC.Internal.Base.Semigroup (f p), GHC.Internal.Base.Semigroup (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:*:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall k2 k1 (f :: k2 -> *) (g :: k1 -> k2) (p :: k1). GHC.Internal.Base.Semigroup (f (g p)) => GHC.Internal.Base.Semigroup ((GHC.Internal.Generics.:.:) f g p) -- Defined in ‘GHC.Internal.Generics’ instance forall a. (GHC.Internal.Generics.Generic a, GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Rep a ())) => GHC.Internal.Base.Semigroup (GHC.Internal.Generics.Generically a) -- Defined in ‘GHC.Internal.Generics’ @@ -11393,20 +11398,20 @@ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Last -- Defined instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Max -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.Fix.MonadFix Data.Semigroup.Min -- Defined in ‘Data.Semigroup’ instance GHC.Internal.Control.Monad.IO.Class.MonadIO GHC.Types.IO -- Defined in ‘GHC.Internal.Control.Monad.IO.Class’ -instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ -instance forall (m :: * -> *) a. (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ +instance forall (a :: * -> * -> *) b c. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable c, GHC.Internal.Data.Data.Data (a b c)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedArrow a b c) -- Defined in ‘Control.Applicative’ +instance forall (m :: * -> *) a. (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable m, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, GHC.Internal.Data.Data.Data (m a)) => GHC.Internal.Data.Data.Data (Control.Applicative.WrappedMonad m a) -- Defined in ‘Control.Applicative’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Functor.ZipList.ZipList a) -- Defined in ‘GHC.Internal.Functor.ZipList’ instance GHC.Internal.Data.Data.Data Data.Array.Byte.ByteArray -- Defined in ‘Data.Array.Byte’ -instance forall s. ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ +instance forall s. ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable s => GHC.Internal.Data.Data.Data (Data.Array.Byte.MutableByteArray s) -- Defined in ‘Data.Array.Byte’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Complex.Complex a) -- Defined in ‘Data.Complex’ -instance forall i j (a :: i) (b :: j). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall i j (a :: i) (b :: j). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable i, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable j, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b, a ~~ b) => GHC.Internal.Data.Data.Data (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.All -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Data.Semigroup.Internal.Any -- Defined in ‘GHC.Internal.Data.Data’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b, GHC.Internal.Ix.Ix a) => GHC.Internal.Data.Data.Data (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.Associativity -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Bool -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Char -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k a (b :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ +instance forall k a (b :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable b) => GHC.Internal.Data.Data.Data (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Data’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Generics.DecidedStrictness -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Types.Double -- Defined in ‘GHC.Internal.Data.Data’ @@ -11454,10 +11459,10 @@ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word32 -- Defined in ‘G instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word64 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Word.Word8 -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Data.Data.Data GHC.Internal.Foreign.Ptr.WordPtr -- Defined in ‘GHC.Internal.Data.Data’ -instance forall k (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ -instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ -instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ +instance forall k (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a) => GHC.Internal.Data.Data.Data (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ +instance forall k1 (f :: k1 -> *) k2 (g :: k2 -> k1) (a :: k2). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k1, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k2, GHC.Internal.Data.Data.Data (f (g a))) => GHC.Internal.Data.Data.Data (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Product.Product f g a) -- Defined in ‘Data.Functor.Product’ +instance [safe] forall k (f :: k -> *) (g :: k -> *) (a :: k). (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable a, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable f, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable g, ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.Typeable k, GHC.Internal.Data.Data.Data (f a), GHC.Internal.Data.Data.Data (g a)) => GHC.Internal.Data.Data.Data (Data.Functor.Sum.Sum f g a) -- Defined in ‘Data.Functor.Sum’ instance forall a b. (GHC.Internal.Data.Data.Data a, GHC.Internal.Data.Data.Data b) => GHC.Internal.Data.Data.Data (Data.Semigroup.Arg a b) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.First a) -- Defined in ‘Data.Semigroup’ instance forall a. GHC.Internal.Data.Data.Data a => GHC.Internal.Data.Data.Data (Data.Semigroup.Last a) -- Defined in ‘Data.Semigroup’ @@ -11523,7 +11528,7 @@ instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SCha instance GHC.Internal.Data.Type.Equality.TestEquality GHC.Internal.TypeLits.SSymbol -- Defined in ‘GHC.Internal.TypeLits’ instance forall k (a :: k). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k (a :: k1). GHC.Internal.Data.Type.Equality.TestEquality ((GHC.Internal.Data.Type.Equality.:~~:) a) -- Defined in ‘GHC.Internal.Data.Type.Equality’ -instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k. GHC.Internal.Data.Type.Equality.TestEquality ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1). GHC.Internal.Data.Type.Equality.TestEquality f => GHC.Internal.Data.Type.Equality.TestEquality (Data.Functor.Compose.Compose f g) -- Defined in ‘Data.Functor.Compose’ instance forall a k (b :: k). GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ instance forall a. GHC.Internal.Enum.Bounded a => GHC.Internal.Enum.Bounded (GHC.Internal.Data.Bits.And a) -- Defined in ‘GHC.Internal.Data.Bits’ @@ -11691,6 +11696,7 @@ instance GHC.Internal.Enum.Enum GHC.Internal.RTS.Flags.DoHeapProfile -- Defined instance GHC.Internal.Enum.Enum GHC.Internal.RTS.Flags.DoTrace -- Defined in ‘GHC.Internal.RTS.Flags’ instance GHC.Internal.Enum.Enum GHC.Internal.RTS.Flags.GiveGCStats -- Defined in ‘GHC.Internal.RTS.Flags’ instance GHC.Internal.Enum.Enum GHC.Internal.RTS.Flags.IoManagerFlag -- Defined in ‘GHC.Internal.RTS.Flags’ +instance GHC.Internal.Exception.Context.ExceptionAnnotation GHC.Internal.Exception.Type.WhileHandling -- Defined in ‘GHC.Internal.Exception.Type’ instance GHC.Internal.Exception.Context.ExceptionAnnotation GHC.Internal.Exception.Backtrace.Backtraces -- Defined in ‘GHC.Internal.Exception.Backtrace’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IO.Exception.AllocationLimitExceeded -- Defined in ‘GHC.Internal.IO.Exception’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IO.Exception.ArrayException -- Defined in ‘GHC.Internal.IO.Exception’ @@ -11720,7 +11726,7 @@ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Ba instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.RecUpdError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Control.Exception.Base.TypeError -- Defined in ‘GHC.Internal.Control.Exception.Base’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ -instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Exception.Type.Exception ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Exception.Type.Exception GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ instance [safe] GHC.Internal.Exception.Type.Exception System.Timeout.Timeout -- Defined in ‘System.Timeout’ instance forall a k (b :: k). GHC.Internal.Float.Floating a => GHC.Internal.Float.Floating (GHC.Internal.Data.Functor.Const.Const a b) -- Defined in ‘GHC.Internal.Data.Functor.Const’ @@ -12326,6 +12332,7 @@ instance GHC.Internal.Show.Show GHC.Internal.Exception.Type.ArithException -- De instance forall a. GHC.Internal.Show.Show a => GHC.Internal.Show.Show (GHC.Internal.Exception.Type.ExceptionWithContext a) -- Defined in ‘GHC.Internal.Exception.Type’ instance forall e. GHC.Internal.Show.Show e => GHC.Internal.Show.Show (GHC.Internal.Exception.Type.NoBacktrace e) -- Defined in ‘GHC.Internal.Exception.Type’ instance GHC.Internal.Show.Show GHC.Internal.Exception.Type.SomeException -- Defined in ‘GHC.Internal.Exception.Type’ +instance GHC.Internal.Show.Show GHC.Internal.Exception.Type.WhileHandling -- Defined in ‘GHC.Internal.Exception.Type’ instance GHC.Internal.Show.Show GHC.Internal.Exception.ErrorCall -- Defined in ‘GHC.Internal.Exception’ instance GHC.Internal.Show.Show GHC.Internal.IO.MaskingState -- Defined in ‘GHC.Internal.IO’ instance GHC.Internal.Show.Show GHC.Internal.Control.Exception.Base.NestedAtomically -- Defined in ‘GHC.Internal.Control.Exception.Base’ @@ -12391,8 +12398,8 @@ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataRep -- Defined in ‘ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.DataType -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Internal.Show.Show GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Internal.Show.Show (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Internal.Show.Show (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance GHC.Internal.Show.Show GHC.Internal.Data.Dynamic.Dynamic -- Defined in ‘GHC.Internal.Data.Dynamic’ instance forall a b. (GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). Data.Fixed.HasResolution a => GHC.Internal.Show.Show (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ @@ -12463,13 +12470,13 @@ instance GHC.Internal.Show.Show GHC.Internal.Foreign.C.Types.CWchar -- Defined i instance forall a. GHC.Internal.Show.Show (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) -- Defined in ‘GHC.Internal.Foreign.C.ConstPtr’ instance forall a b. (GHC.Internal.Ix.Ix a, GHC.Internal.Show.Show a, GHC.Internal.Show.Show b) => GHC.Internal.Show.Show (GHC.Internal.Arr.Array a b) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Internal.Show.Show GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Timeout -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Internal.Show.Show GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ instance GHC.Internal.Show.Show GHC.Types.Double -- Defined in ‘GHC.Internal.Float’ instance GHC.Internal.Show.Show GHC.Types.Float -- Defined in ‘GHC.Internal.Float’ @@ -12503,7 +12510,7 @@ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Handle -- Defined i instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.HandleType -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.Newline -- Defined in ‘GHC.Internal.IO.Handle.Types’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.Types.NewlineMode -- Defined in ‘GHC.Internal.IO.Handle.Types’ -instance [safe] GHC.Internal.Show.Show ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.IO.Handle.Lock.Common’ +instance [safe] GHC.Internal.Show.Show ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common.FileLockingNotSupported -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.IO.Handle.Lock.Common’ instance GHC.Internal.Show.Show GHC.Internal.IO.Handle.HandlePosn -- Defined in ‘GHC.Internal.IO.Handle’ instance GHC.Internal.Show.Show GHC.Internal.IO.IOMode.IOMode -- Defined in ‘GHC.Internal.IO.IOMode’ instance GHC.Internal.Show.Show GHC.Internal.IOPort.IOPortException -- Defined in ‘GHC.Internal.IOPort’ @@ -12600,8 +12607,8 @@ instance GHC.Classes.Eq GHC.Internal.Data.Data.ConstrRep -- Defined in ‘GHC.In instance GHC.Classes.Eq GHC.Internal.Data.Data.DataRep -- Defined in ‘GHC.Internal.Data.Data’ instance GHC.Classes.Eq GHC.Internal.Data.Data.Fixity -- Defined in ‘GHC.Internal.Data.Data’ instance forall k (s :: k). GHC.Classes.Eq (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Eq (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Eq a, GHC.Classes.Eq b) => GHC.Classes.Eq (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Eq (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Eq (f (g a)) => GHC.Classes.Eq (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ @@ -12677,13 +12684,13 @@ instance forall a. GHC.Classes.Eq (GHC.Internal.Foreign.C.ConstPtr.ConstPtr a) - instance forall i e. (GHC.Internal.Ix.Ix i, GHC.Classes.Eq e) => GHC.Classes.Eq (GHC.Internal.Arr.Array i e) -- Defined in ‘GHC.Internal.Arr’ instance forall s i e. GHC.Classes.Eq (GHC.Internal.Arr.STArray s i e) -- Defined in ‘GHC.Internal.Arr’ instance GHC.Classes.Eq GHC.Internal.ByteOrder.ByteOrder -- Defined in ‘GHC.Internal.ByteOrder’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Internal.Types’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Event -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.EventLifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types.Lifetime -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Internal.Types’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.FdKey -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.Manager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.Manager’ instance GHC.Classes.Eq GHC.Internal.Event.TimeOut.TimeoutKey -- Defined in ‘GHC.Internal.Event.TimeOut’ -instance GHC.Classes.Eq ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Event.TimerManager’ +instance GHC.Classes.Eq ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager.State -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Event.TimerManager’ instance GHC.Classes.Eq GHC.Internal.Stack.Types.SrcLoc -- Defined in ‘GHC.Internal.Stack.Types’ instance GHC.Classes.Eq GHC.Internal.Exts.SpecConstrAnnotation -- Defined in ‘GHC.Internal.Exts’ instance GHC.Classes.Eq GHC.Internal.Fingerprint.Type.Fingerprint -- Defined in ‘GHC.Internal.Fingerprint.Type’ @@ -12779,8 +12786,8 @@ instance GHC.Classes.Ord GHC.Internal.Unicode.GeneralCategory -- Defined in ‘G instance forall k (a :: k) (b :: k). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k1 k2 (a :: k1) (b :: k2). GHC.Classes.Ord (a GHC.Internal.Data.Type.Equality.:~~: b) -- Defined in ‘GHC.Internal.Data.Type.Equality’ instance forall k (s :: k). GHC.Classes.Ord (GHC.Internal.Data.Proxy.Proxy s) -- Defined in ‘GHC.Internal.Data.Proxy’ -instance GHC.Classes.Ord ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ -instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1001.0:GHC.Internal.Data.Typeable.Internal’ +instance GHC.Classes.Ord ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.SomeTypeRep -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ +instance forall k (a :: k). GHC.Classes.Ord (ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal.TypeRep a) -- Defined in ‘ghc-internal-9.1100.0:GHC.Internal.Data.Typeable.Internal’ instance forall a b. (GHC.Classes.Ord a, GHC.Classes.Ord b) => GHC.Classes.Ord (GHC.Internal.Data.Either.Either a b) -- Defined in ‘GHC.Internal.Data.Either’ instance forall k (a :: k). GHC.Classes.Ord (Data.Fixed.Fixed a) -- Defined in ‘Data.Fixed’ instance forall k1 k2 (f :: k1 -> *) (g :: k2 -> k1) (a :: k2). GHC.Classes.Ord (f (g a)) => GHC.Classes.Ord (Data.Functor.Compose.Compose f g a) -- Defined in ‘Data.Functor.Compose’ ===================================== testsuite/tests/mdo/should_fail/mdofail006.stderr ===================================== @@ -2,8 +2,12 @@ mdofail006: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.FixIOExcep cyclic evaluation in fixIO +While handling ghc-internal:GHC.Internal.IO.Exception.BlockedIndefinitelyOnMVar: + │ + │ thread blocked indefinitely in an MVar operation + HasCallStack backtrace: collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:260:11 in ghc-internal:GHC.Internal.IO + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO throwIO, called at libraries/ghc-internal/src/GHC/Internal/System/IO.hs:641:37 in ghc-internal:GHC.Internal.System.IO ===================================== testsuite/tests/runghc/T7859.stderr ===================================== @@ -1,3 +1,11 @@ runghc: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.IOException: defer-type-errors: executeFile: does not exist (No such file or directory) + +HasCallStack backtrace: + collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/IO.hs:284:11 in ghc-internal:GHC.Internal.IO + throwIO, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:315:19 in ghc-internal:GHC.Internal.IO.Exception + ioException, called at libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs:319:20 in ghc-internal:GHC.Internal.IO.Exception + ioError, called at libraries/ghc-internal/src/GHC/Internal/Foreign/C/Error.hs:428:5 in ghc-internal:GHC.Internal.Foreign.C.Error + ===================================== testsuite/tests/typecheck/should_compile/holes.stderr ===================================== @@ -1,4 +1,3 @@ - holes.hs:3:5: warning: [GHC-88464] [-Wtyped-holes (in -Wdefault)] • Found hole: _ :: p Where: ‘p’ is a rigid type variable bound by @@ -92,7 +91,9 @@ holes.hs:11:15: warning: [GHC-88464] [-Wtyped-holes (in -Wdefault)] asTypeOf :: forall a. a -> a -> a id :: forall a. a -> a until :: forall a. (a -> Bool) -> (a -> a) -> a -> a - ioError :: forall a. IOError -> IO a + ioError :: forall a. + GHC.Internal.Stack.Types.HasCallStack => + IOError -> IO a (!!) :: forall a. GHC.Internal.Stack.Types.HasCallStack => [a] -> Int -> a @@ -204,3 +205,4 @@ holes.hs:11:15: warning: [GHC-88464] [-Wtyped-holes (in -Wdefault)] flip :: forall a b c. (a -> b -> c) -> b -> a -> c zipWith3 :: forall a b c d. (a -> b -> c -> d) -> [a] -> [b] -> [c] -> [d] + ===================================== testsuite/tests/typecheck/should_compile/holes3.stderr ===================================== @@ -1,4 +1,3 @@ - holes3.hs:3:5: error: [GHC-88464] • Found hole: _ :: p Where: ‘p’ is a rigid type variable bound by @@ -95,7 +94,9 @@ holes3.hs:11:15: error: [GHC-88464] asTypeOf :: forall a. a -> a -> a id :: forall a. a -> a until :: forall a. (a -> Bool) -> (a -> a) -> a -> a - ioError :: forall a. IOError -> IO a + ioError :: forall a. + GHC.Internal.Stack.Types.HasCallStack => + IOError -> IO a (!!) :: forall a. GHC.Internal.Stack.Types.HasCallStack => [a] -> Int -> a @@ -207,3 +208,4 @@ holes3.hs:11:15: error: [GHC-88464] flip :: forall a b c. (a -> b -> c) -> b -> a -> c zipWith3 :: forall a b c d. (a -> b -> c -> d) -> [a] -> [b] -> [c] -> [d] + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/839fcac3858569ad589e6548cc11d918da0c95be...52a6fde3eae77cb98be9e2569cf8c2e7669f50fe -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/839fcac3858569ad589e6548cc11d918da0c95be...52a6fde3eae77cb98be9e2569cf8c2e7669f50fe You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 16:37:12 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 23 Sep 2024 12:37:12 -0400 Subject: [Git][ghc/ghc][wip/spj-unf-size] 253 commits: ci: Enable strict ghc-toolchain setting for bindists Message-ID: <66f19938268f8_2d8bc4b8c6d0494c9@gitlab.mail> Simon Peyton Jones pushed to branch wip/spj-unf-size at Glasgow Haskell Compiler / GHC Commits: db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 8f86a3a2 by Simon Peyton Jones at 2024-09-23T12:43:32+01:00 Work in progress on unfoldings re-engineering - - - - - 5d9f09d4 by Simon Peyton Jones at 2024-09-23T12:43:32+01:00 Fix a bad, subtle bug in exprIsConApp_maybe In extend_in_scope We were simply overwriting useful bindings in the in-scope set, notably ones that had unfoldings. That could lead to repeated simplifier iterations. - - - - - a18346c0 by Simon Peyton Jones at 2024-09-23T12:43:32+01:00 Minor refactoring... Plus: don't be so eager to inline when argument is a non-value, but has some struture. We want *some* incentive though. - - - - - 94e5b2d2 by Simon Peyton Jones at 2024-09-23T12:43:32+01:00 Adjust * Reduce caseElimDiscount to 10 Example: f_nand in spectral/hartel/event is quite big but was still getting inlined; that make f_simulate too big for SpecConstr * Increase jumpSize. Not so much cheaper than tail calls. I'm trying making them the same size. - - - - - 41f8a050 by Simon Peyton Jones at 2024-09-23T12:43:32+01:00 Typo - - - - - 24 changed files: - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC.hs - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/PrimOps.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/Utils.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/ByteCode/Types.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5d42a398137563837682312f2d4b409556188029...41f8a0503ef23d883c6ce137ac6d982b87f228bf -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5d42a398137563837682312f2d4b409556188029...41f8a0503ef23d883c6ce137ac6d982b87f228bf You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 17:20:43 2024 From: gitlab at gitlab.haskell.org (Ryan Scott (@RyanGlScott)) Date: Mon, 23 Sep 2024 13:20:43 -0400 Subject: [Git][ghc/ghc][wip/T25148] Note [Newtype-deriving instances]: List children Notes Message-ID: <66f1a36b7ec5a_2d8bc4da0aac563da@gitlab.mail> Ryan Scott pushed to branch wip/T25148 at Glasgow Haskell Compiler / GHC Commits: 7d6e30e6 by Ryan Scott at 2024-09-23T13:20:17-04:00 Note [Newtype-deriving instances]: List children Notes - - - - - 1 changed file: - compiler/GHC/Tc/Deriv/Generate.hs Changes: ===================================== compiler/GHC/Tc/Deriv/Generate.hs ===================================== @@ -1737,6 +1737,26 @@ However, to allow VTA with polytypes we must switch on -XImpredicativeTypes locally in GHC.Tc.Deriv.genInst. See #8503 for more discussion. +The following Notes describe further nuances of GeneralizedNewtypeDeriving: + +----- +-- In GHC.Tc.Deriv +----- + +* Note [Newtype deriving] +* Note [Newtype representation] +* Note [Recursive newtypes] +* Note [Determining whether newtype-deriving is appropriate] +* Note [GND and associated type families] +* Note [Bindings for Generalised Newtype Deriving] + +----- +-- In GHC.Tc.Deriv.Generate +----- + +* Note [Newtype-deriving trickiness] +* Note [GND and ambiguity] + Note [Inferred invisible patterns] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Consider the following: View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7d6e30e63b9c3d40d3ef1395924fd5d565582af5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7d6e30e63b9c3d40d3ef1395924fd5d565582af5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 18:27:04 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 23 Sep 2024 14:27:04 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/torsten.schmits/package-deps-bytecode-2024-09-23 Message-ID: <66f1b2f819a8e_173fbabc4804115@gitlab.mail> Torsten Schmits pushed new branch wip/torsten.schmits/package-deps-bytecode-2024-09-23 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/torsten.schmits/package-deps-bytecode-2024-09-23 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 18:35:57 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Mon, 23 Sep 2024 14:35:57 -0400 Subject: [Git][ghc/ghc][wip/bump-process] 7118 commits: [haddock @ 2002-04-04 16:23:43 by simonmar] Message-ID: <66f1b50d95fad_173fba20e0cc4327@gitlab.mail> Ben Gamari pushed to branch wip/bump-process at Glasgow Haskell Compiler / GHC Commits: 2b39cd94 by Simon Marlow at 2002-04-04T16:23:43+00:00 [haddock @ 2002-04-04 16:23:43 by simonmar] This is Haddock, my stab at a Haskell documentation tool. It's not quite ready for release yet, but I'm putting it in the repository so others can take a look. It uses a locally modified version of the hssource parser, extended with support for GHC extensions and documentation annotations. - - - - - 99ede94f by Simon Marlow at 2002-04-04T16:24:10+00:00 [haddock @ 2002-04-04 16:24:10 by simonmar] forgot one file - - - - - 8363294c by Simon Marlow at 2002-04-05T13:58:15+00:00 [haddock @ 2002-04-05 13:58:15 by simonmar] Remap names in the exported declarations to be "closer" to the current module. eg. if an exported declaration mentions a type 'T' which is imported from module A then re-exported from the current module, then links from the type or indeed the documentation will point to the current module rather than module A. This is to support better hiding: module A won't be referred to in the generated output. - - - - - 1570cbc1 by Simon Marlow at 2002-04-05T13:58:23+00:00 [haddock @ 2002-04-05 13:58:23 by simonmar] update the TODO list - - - - - 3a62f96b by Simon Marlow at 2002-04-05T14:11:51+00:00 [haddock @ 2002-04-05 14:11:51 by simonmar] Fix the anchor for a class declaration - - - - - c5d9a471 by Simon Marlow at 2002-04-05T14:18:41+00:00 [haddock @ 2002-04-05 14:18:41 by simonmar] remove underlines on visited links - - - - - 97280525 by Simon Marlow at 2002-04-05T16:11:47+00:00 [haddock @ 2002-04-05 16:11:47 by simonmar] - Update to generate more correct HTML. - Use our own non-overloaded table combinators, as the overloaded versions were giving me a headache. The improved type safety caught several errors in the HTML generation. - - - - - 9acd3a4d by Simon Marlow at 2002-04-05T16:32:19+00:00 [haddock @ 2002-04-05 16:32:19 by simonmar] Add width property to the title, and add TD.children for the module contents page. - - - - - ec9a0847 by Simon Marlow at 2002-04-08T16:39:56+00:00 [haddock @ 2002-04-08 16:39:56 by simonmar] Fix a problem with exports of the form T(..). - - - - - e4627dc8 by Simon Marlow at 2002-04-08T16:41:38+00:00 [haddock @ 2002-04-08 16:41:37 by simonmar] - Add our own versions of Html & BlockTable for the time being. - Add support for generating an index to the HTML backend - - - - - 2d73fd75 by Simon Marlow at 2002-04-09T11:23:24+00:00 [haddock @ 2002-04-09 11:23:24 by simonmar] Add '-- /' as a synonym for '-- |', for compatibility with IDoc. - - - - - 3675464e by Simon Marlow at 2002-04-09T11:33:55+00:00 [haddock @ 2002-04-09 11:33:54 by simonmar] - add the <...> syntax for marking up URLs in documentation - Make the output for data & class declarations more compact when there aren't any documentation annotations on the individual methods or constructors respectively. - - - - - 5077f5b1 by Simon Marlow at 2002-04-09T11:36:04+00:00 [haddock @ 2002-04-09 11:36:04 by simonmar] Update the TODO list - - - - - 9e83c54d by Simon Marlow at 2002-04-10T10:50:06+00:00 [haddock @ 2002-04-10 10:50:06 by simonmar] Use explicit 'px' suffix on pixel sizes; IE seems to prefer them - - - - - 052de51c by Simon Marlow at 2002-04-10T13:23:13+00:00 [haddock @ 2002-04-10 13:23:13 by simonmar] Lex URLs as a single token to avoid having to escape special characters inside the URL string. - - - - - 47187edb by Simon Marlow at 2002-04-10T13:23:55+00:00 [haddock @ 2002-04-10 13:23:55 by simonmar] Not sure why I made the constructor name for a record declaration into a TyCls name, but change it back into a Var name anyhow. - - - - - 3dc6aa81 by Simon Marlow at 2002-04-10T13:26:10+00:00 [haddock @ 2002-04-10 13:26:09 by simonmar] Lots of changes, including: - add index support to the HTML backend - clean up the renamer, put it into a monad - propogate unresolved names to the top level and report them in a nicer way - various bugfixes - - - - - c2a70a72 by Simon Marlow at 2002-04-10T13:32:39+00:00 [haddock @ 2002-04-10 13:32:39 by simonmar] Skeleton documentation - - - - - 50c98d17 by Simon Marlow at 2002-04-10T13:37:23+00:00 [haddock @ 2002-04-10 13:37:23 by simonmar] Update the TODO list, separate into pre-1.0 and post-1.0 items - - - - - f3778be6 by Simon Marlow at 2002-04-10T14:30:58+00:00 [haddock @ 2002-04-10 14:30:58 by simonmar] Add an introduction - - - - - cfbaf9f7 by Simon Marlow at 2002-04-10T14:59:51+00:00 [haddock @ 2002-04-10 14:59:51 by simonmar] Sort the module tree - - - - - 76bd7b34 by Simon Marlow at 2002-04-10T15:50:11+00:00 [haddock @ 2002-04-10 15:50:10 by simonmar] Generate a little table of contents at the top of the module doc (only if the module actually contains some section headings, though). - - - - - bb8560a1 by Simon Marlow at 2002-04-10T16:10:26+00:00 [haddock @ 2002-04-10 16:10:26 by simonmar] Now we understand (or at least don't barf on) type signatures in patterns such as you might find when scoped type variables are in use. - - - - - 86c2a026 by Simon Marlow at 2002-04-10T16:10:49+00:00 [haddock @ 2002-04-10 16:10:49 by simonmar] more updates - - - - - 1c052b0e by Simon Marlow at 2002-04-10T16:28:05+00:00 [haddock @ 2002-04-10 16:28:05 by simonmar] Parse errors in doc strings are now reported as warnings rather that causing the whole thing to fall over. It still needs cleaning up (the warning is emitted with trace) but this will do for the time being. - - - - - ace03e8f by Simon Marlow at 2002-04-10T16:38:03+00:00 [haddock @ 2002-04-10 16:38:03 by simonmar] update again - - - - - 69006c3e by Simon Marlow at 2002-04-11T13:38:02+00:00 [haddock @ 2002-04-11 13:38:02 by simonmar] mention Opera - - - - - fe9b10f8 by Simon Marlow at 2002-04-11T13:40:31+00:00 [haddock @ 2002-04-11 13:40:30 by simonmar] - copy haddock.css into the same place as the generated HTML - new option: --css <file> specifies the style sheet to use - new option: -o <dir> specifies the directory in which to generate the output. - because Haddock now needs to know where to find its default stylesheet, we have to have a wrapper script and do the haddock-inplace thing (Makefile code copied largely from fptools/happy). - - - - - 106adbbe by Simon Marlow at 2002-04-24T15:12:41+00:00 [haddock @ 2002-04-24 15:12:41 by simonmar] Stop slurping comment lines when we see a row of dashes longer than length 2: these are useful as separators. - - - - - 995d3f9e by Simon Marlow at 2002-04-24T15:14:12+00:00 [haddock @ 2002-04-24 15:14:11 by simonmar] Grok the kind of module headers we use in fptools/libraries, and pass the "portability", "stability", and "maintainer" strings through into the generated HTML. If the module header doesn't match the pattern, then we don't include the info in the HTML. - - - - - e14da136 by Simon Marlow at 2002-04-24T15:16:57+00:00 [haddock @ 2002-04-24 15:16:57 by simonmar] Done module headers now. - - - - - 2ca8dfd4 by Simon Marlow at 2002-04-24T15:57:48+00:00 [haddock @ 2002-04-24 15:57:47 by simonmar] Handle gcons in export lists (a common extension). - - - - - 044cea81 by Simon Marlow at 2002-04-25T14:20:12+00:00 [haddock @ 2002-04-25 14:20:12 by simonmar] Add the little lambda icon - - - - - 63955027 by Simon Marlow at 2002-04-25T14:40:05+00:00 [haddock @ 2002-04-25 14:40:05 by simonmar] - Add support for named chunks of documentation which can be referenced from the export list. - Copy the icon from $libdir to the destination in HTML mode. - - - - - 36e3f913 by Simon Marlow at 2002-04-25T16:48:36+00:00 [haddock @ 2002-04-25 16:48:36 by simonmar] More keyboard bashing - - - - - 7ae18dd0 by Simon Marlow at 2002-04-26T08:43:33+00:00 [haddock @ 2002-04-26 08:43:33 by simonmar] Package util reqd. to compile with 4.08.2 - - - - - bbd5fbab by Simon Marlow at 2002-04-26T10:13:00+00:00 [haddock @ 2002-04-26 10:13:00 by simonmar] Include $(GHC_HAPPY_OPTS) when compiling HsParser - - - - - 31c53d79 by Simon Marlow at 2002-04-26T11:18:57+00:00 [haddock @ 2002-04-26 11:18:56 by simonmar] - support for fundeps (partially contributed by Brett Letner - thanks Brett). - make it build with GHC 4.08.2 - - - - - c415ce76 by Simon Marlow at 2002-04-26T13:15:02+00:00 [haddock @ 2002-04-26 13:15:02 by simonmar] Move the explicit formatting of the little table for the stability/portability/maintainer info from the HTML into the CSS, and remove the explicit table size (just right-align it). - - - - - 520ee21a by Simon Marlow at 2002-04-26T16:01:44+00:00 [haddock @ 2002-04-26 16:01:44 by simonmar] Yet more keyboard bashing - this is pretty much complete now. - - - - - 2ae37179 by Simon Marlow at 2002-04-26T16:02:14+00:00 [haddock @ 2002-04-26 16:02:14 by simonmar] Add a couple of things I forgot about - - - - - b7211e04 by Simon Marlow at 2002-04-29T15:28:12+00:00 [haddock @ 2002-04-29 15:28:12 by simonmar] bugfix for declBinders on a NewTypeDecl - - - - - 640c154a by Simon Marlow at 2002-04-29T15:28:54+00:00 [haddock @ 2002-04-29 15:28:54 by simonmar] Allow '-- |' style annotations on constructors and record fields. - - - - - 393f258a by Simon Marlow at 2002-04-29T15:37:32+00:00 [haddock @ 2002-04-29 15:37:32 by simonmar] syntax fix - - - - - 8a2c2549 by Simon Marlow at 2002-04-29T15:37:48+00:00 [haddock @ 2002-04-29 15:37:48 by simonmar] Add an example - - - - - db88f8a2 by Simon Marlow at 2002-04-29T15:55:46+00:00 [haddock @ 2002-04-29 15:55:46 by simonmar] remove a trace - - - - - 2b0248e0 by Simon Marlow at 2002-04-29T15:56:19+00:00 [haddock @ 2002-04-29 15:56:19 by simonmar] Fix for 'make install' - - - - - 120453a0 by Simon Marlow at 2002-04-29T15:56:39+00:00 [haddock @ 2002-04-29 15:56:39 by simonmar] Install the auxilliary bits - - - - - 950e6dbb by Simon Marlow at 2002-04-29T15:57:30+00:00 [haddock @ 2002-04-29 15:57:30 by simonmar] Add BinDist bits - - - - - 154b9d71 by Simon Marlow at 2002-05-01T11:02:52+00:00 [haddock @ 2002-05-01 11:02:52 by simonmar] update - - - - - ba6c39fa by Simon Marlow at 2002-05-01T11:03:26+00:00 [haddock @ 2002-05-01 11:03:26 by simonmar] Add another item - - - - - bacb5e33 by Simon Marlow at 2002-05-03T08:50:00+00:00 [haddock @ 2002-05-03 08:50:00 by simonmar] Fix some typos. - - - - - 54c87895 by Sven Panne at 2002-05-05T19:40:51+00:00 [haddock @ 2002-05-05 19:40:51 by panne] As a temporary hack/workaround for a bug in GHC's simplifier, don't pass Happy the -c option for generating the parsers in this subdir. Furthermore, disable -O for HaddocParse, too. - - - - - e6c08703 by Simon Marlow at 2002-05-06T09:51:10+00:00 [haddock @ 2002-05-06 09:51:10 by simonmar] Add RPM spec file (thanks to Tom Moertel <tom-rpms at moertel.com>) - - - - - 7b8fa8e7 by Simon Marlow at 2002-05-06T12:29:26+00:00 [haddock @ 2002-05-06 12:29:26 by simonmar] Add missing type signature (a different workaround for the bug in GHC's simplifier). - - - - - cd0e300d by Simon Marlow at 2002-05-06T12:30:09+00:00 [haddock @ 2002-05-06 12:30:09 by simonmar] Remove workaround for simplifier bug in previous revision. - - - - - 687e68fa by Simon Marlow at 2002-05-06T12:32:32+00:00 [haddock @ 2002-05-06 12:32:32 by simonmar] Allow empty data declarations (another GHC extension). - - - - - 8f29f696 by Simon Marlow at 2002-05-06T12:49:21+00:00 [haddock @ 2002-05-06 12:49:21 by simonmar] Fix silly bug in named documentation block lookup. - - - - - 8e0059af by Simon Marlow at 2002-05-06T13:02:42+00:00 [haddock @ 2002-05-06 13:02:42 by simonmar] Add another named chunk with a different name - - - - - 68f8a896 by Simon Marlow at 2002-05-06T13:32:32+00:00 [haddock @ 2002-05-06 13:32:32 by simonmar] Be more lenient about extra paragraph breaks - - - - - 65fc31db by Simon Marlow at 2002-05-07T15:36:36+00:00 [haddock @ 2002-05-07 15:36:36 by simonmar] DocEmpty is a right and left-unit of DocAppend (remove it in the smart constructor). - - - - - adc81078 by Simon Marlow at 2002-05-07T15:37:15+00:00 [haddock @ 2002-05-07 15:37:15 by simonmar] Allow code blocks to be denoted with bird-tracks in addition to [...]. - - - - - 1283a3c1 by Simon Marlow at 2002-05-08T11:21:56+00:00 [haddock @ 2002-05-08 11:21:56 by simonmar] Add a facility for specifying options that affect Haddock's treatment of the module. Options are given at the top of the module in a comma-separated list, beginning with '-- #'. eg. -- # prune, hide, ignore-exports Options currently available, with their meanings: prune: ignore declarations which have no documentation annotations ignore-exports: act as if the export list were not specified (i.e. export everything local to the module). hide: do not include this module in the generated documentation, but propagate any exported definitions to modules which re-export them. There's a slight change in the semantics for re-exporting a full module by giving 'module M' in the export list: if module M does not have the 'hide' option, then the documentation will now just contain a reference to module M rather than the full inlined contents of that module. These features, and some other changes in the pipeline, are the result of discussions between myself and Manuel Chakravarty <chak at cse.unsw.edu.au> (author of IDoc) yesterday. Also: some cleanups, use a Writer monad to collect error messages in some places instead of just printing them with trace. - - - - - a2239cf5 by Simon Marlow at 2002-05-08T11:22:30+00:00 [haddock @ 2002-05-08 11:22:30 by simonmar] Update to test new features. - - - - - 6add955f by Simon Marlow at 2002-05-08T13:37:25+00:00 [haddock @ 2002-05-08 13:37:25 by simonmar] Change the markup for typewriter-font from [...] to @... at . The reasoning is that the '@' symbol is much less likely to be needed than square brackets, and we don't want to have to escape square brackets in code fragments. This will be mildly painful in the short term, but it's better to get the change out of the way as early as possible. - - - - - cda06447 by Simon Marlow at 2002-05-08T13:39:56+00:00 [haddock @ 2002-05-08 13:39:56 by simonmar] Allow nested-style comments to be used as documentation annotations too. eg. {-| ... -} is equivalent to -- | ... An extra space can also be left after the comment opener: {- | ... -}. The only version that isn't allowed is {-# ... -}, because this syntax overlaps with Haskell pragmas; use {- # ... -} instead. - - - - - db23f65e by Simon Marlow at 2002-05-08T14:48:41+00:00 [haddock @ 2002-05-08 14:48:39 by simonmar] Add support for existential quantifiers on constructors. - - - - - adce3794 by Simon Marlow at 2002-05-08T15:43:25+00:00 [haddock @ 2002-05-08 15:43:25 by simonmar] update - - - - - 62a1f436 by Simon Marlow at 2002-05-08T15:44:10+00:00 [haddock @ 2002-05-08 15:44:10 by simonmar] Update to version 0.2 - - - - - f6a24ba3 by Simon Marlow at 2002-05-09T08:48:29+00:00 [haddock @ 2002-05-09 08:48:29 by simonmar] typo - - - - - 9f9522a4 by Simon Marlow at 2002-05-09T10:33:14+00:00 [haddock @ 2002-05-09 10:33:14 by simonmar] oops, left out '/' from the special characters in the last change. - - - - - 14abcb39 by Simon Marlow at 2002-05-09T10:34:44+00:00 [haddock @ 2002-05-09 10:34:44 by simonmar] Fix buglet - - - - - b8d878be by Simon Marlow at 2002-05-09T10:35:00+00:00 [haddock @ 2002-05-09 10:35:00 by simonmar] Give a more useful instance of Show for Module. - - - - - f7bfd626 by Simon Marlow at 2002-05-09T10:37:07+00:00 [haddock @ 2002-05-09 10:37:07 by simonmar] The last commit to Main.lhs broke the delicate balance of laziness which was being used to avoid computing the dependency graph of modules. So I finally bit the bullet and did a proper topological sort of the module graph, which turned out to be easy (stealing the Digraph module from GHC - this really ought to be in the libraries somewhere). - - - - - b481c1d0 by Simon Marlow at 2002-05-09T10:37:25+00:00 [haddock @ 2002-05-09 10:37:25 by simonmar] another item done - - - - - 032e2b42 by Simon Marlow at 2002-05-09T10:44:15+00:00 [haddock @ 2002-05-09 10:44:15 by simonmar] Don't consider a module re-export as having documentation, for the purposes of deciding whether we need a Synopsis section or not. - - - - - 5fb45e92 by Simon Marlow at 2002-05-09T11:10:55+00:00 [haddock @ 2002-05-09 11:10:55 by simonmar] Add a special case for list types in ppHsAType - - - - - 1937e428 by Simon Marlow at 2002-05-09T12:43:06+00:00 [haddock @ 2002-05-09 12:43:06 by simonmar] Type synonyms can accept a ctype on the RHS, to match GHC. - - - - - 0f16ce56 by Simon Marlow at 2002-05-09T12:45:19+00:00 [haddock @ 2002-05-09 12:45:19 by simonmar] Add 'stdcall' keyword - - - - - 29b0d7d2 by Simon Marlow at 2002-05-09T13:35:45+00:00 [haddock @ 2002-05-09 13:35:45 by simonmar] Add System Requirements section - - - - - bf14dddd by Simon Marlow at 2002-05-09T13:36:11+00:00 [haddock @ 2002-05-09 13:36:11 by simonmar] Test existential types, amongst other things - - - - - 502f8f6f by Simon Marlow at 2002-05-09T13:37:35+00:00 [haddock @ 2002-05-09 13:37:35 by simonmar] Print the module name in a doc-string parse error - - - - - ca1f8d49 by Simon Marlow at 2002-05-09T13:38:04+00:00 [haddock @ 2002-05-09 13:38:04 by simonmar] Add dependency - - - - - 8d3d91ff by Simon Marlow at 2002-05-09T15:37:57+00:00 [haddock @ 2002-05-09 15:37:57 by simonmar] Add the changelog/release notes - - - - - f3960959 by Simon Marlow at 2002-05-09T15:47:47+00:00 [haddock @ 2002-05-09 15:47:47 by simonmar] mention the backquote-style of markup - - - - - 089fb6e6 by Simon Marlow at 2002-05-09T15:59:45+00:00 [haddock @ 2002-05-09 15:59:45 by simonmar] update - - - - - bdd3be0b by Simon Marlow at 2002-05-09T15:59:56+00:00 [haddock @ 2002-05-09 15:59:56 by simonmar] Document changes since 0.1 - - - - - 00fc4af8 by Simon Marlow at 2002-05-10T08:22:48+00:00 [haddock @ 2002-05-10 08:22:48 by simonmar] oops, update to version 0.2 - - - - - a8a79041 by Simon Marlow at 2002-05-10T16:05:08+00:00 [haddock @ 2002-05-10 16:05:08 by simonmar] Only include a mini-contents if there are 2 or more sections - - - - - 06653319 by Simon Marlow at 2002-05-13T09:13:12+00:00 [haddock @ 2002-05-13 09:13:12 by simonmar] fix typos - - - - - 1402b19b by Simon Marlow at 2002-05-13T10:14:22+00:00 [haddock @ 2002-05-13 10:14:22 by simonmar] Allow backquote as the right-hand quote as well as the left-hand quote, as suggested by Dean Herrington. Clean up the grammar a litte. - - - - - dcd5320d by Simon Marlow at 2002-05-13T10:44:10+00:00 [haddock @ 2002-05-13 10:44:10 by simonmar] a couple more things, prioritise a bit - - - - - a90130c4 by Simon Marlow at 2002-05-13T15:19:03+00:00 [haddock @ 2002-05-13 15:19:03 by simonmar] Cope with datatypes which have documentation on the constructor but not the type itself, and records which have documentation on the fields but not the constructor. (Thanks to Ross Paterson for pointing out the bugs). - - - - - a774d432 by Simon Marlow at 2002-05-13T15:20:54+00:00 [haddock @ 2002-05-13 15:20:54 by simonmar] Fix one of the record examples - - - - - 2d1d5218 by Simon Marlow at 2002-05-15T12:44:35+00:00 [haddock @ 2002-05-15 12:44:35 by simonmar] Preserve the newline before a bird-track, but only within a paragraph. - - - - - 1554c09a by Simon Marlow at 2002-05-15T13:03:02+00:00 [haddock @ 2002-05-15 13:03:01 by simonmar] Reworking of the internals to support documenting function arguments (the Most Wanted new feature by the punters). The old method of keeping parsed documentation in a Name -> Doc mapping wasn't going to cut it for anntations on type components, where there's no name to attach the documentation to, so I've moved to storing all the documentation in the abstract syntax. Previously some of the documentation was left in the abstract syntax by the parser, but was later extracted into the mapping. In order to avoid having to parameterise the abstract syntax over the type of documentation stored in it, we have to parse the documentation at the same time as we parse the Haskell source (well, I suppose we could store 'Either String Doc' in the HsSyn, but that's clunky). One upshot is that documentation is now parsed eagerly, and documentation parse errors are fatal (but have better line numbers in the error message). The new story simplifies matters for the code that processes the source modules, because we don't have to maintain the extra Name->Doc mapping, and it should improve efficiency a little too. New features: - Function arguments and return values can now have doc annotations. - If you refer to a qualified name in a doc string, eg. 'IO.putStr', then Haddock will emit a hyperlink even if the identifier is not in scope, so you don't have to make sure everything referred to from the documentation is imported. - several bugs & minor infelicities fixed. - - - - - 57344dc3 by Simon Marlow at 2002-05-15T13:03:19+00:00 [haddock @ 2002-05-15 13:03:19 by simonmar] Bump to version 0.3 - - - - - b2791812 by Simon Marlow at 2002-05-15T13:03:41+00:00 [haddock @ 2002-05-15 13:03:41 by simonmar] update - - - - - fead183e by Simon Marlow at 2002-05-15T13:10:15+00:00 [haddock @ 2002-05-15 13:10:15 by simonmar] Rename Foo.hs to Test.hs, and add a Makefile - - - - - b0b1f89f by Simon Marlow at 2002-05-15T13:16:07+00:00 [haddock @ 2002-05-15 13:16:07 by simonmar] - Remove the note about function argument docs not being implemented - Note that qualified identifiers can be used to point to entities that aren't in scope. - - - - - 5665f31a by Simon Marlow at 2002-05-15T13:28:46+00:00 [haddock @ 2002-05-15 13:28:46 by simonmar] Patch to add support for GHC-style primitive strings ".."#, from Ross Paterson. - - - - - 0564505d by Simon Marlow at 2002-05-17T10:51:57+00:00 [haddock @ 2002-05-17 10:51:57 by simonmar] Fix bugs in qualified name handling (A.B.f was returned as B.f) - - - - - 10e7311c by Simon Marlow at 2002-05-21T10:24:52+00:00 [haddock @ 2002-05-21 10:24:52 by simonmar] - Use an alternate tabular layout for datatypes, which is more compact - Fix some problems with the function argument documentation - - - - - 2f91c2a6 by Simon Marlow at 2002-05-21T10:27:40+00:00 [haddock @ 2002-05-21 10:27:40 by simonmar] add a few more test cases - - - - - 01c2ddd2 by Simon Marlow at 2002-05-21T10:28:33+00:00 [haddock @ 2002-05-21 10:28:33 by simonmar] Rearrange a bit, and add support for tabular datatype rendering - - - - - a4e4c5f8 by Simon Marlow at 2002-05-27T09:03:52+00:00 [haddock @ 2002-05-27 09:03:51 by simonmar] Lots of changes: - instances of a class are listed with the class, and instances involving a datatype are listed with that type. Derived instances aren't included at the moment: the calculation to find the instance head for a derived instance is non-trivial. - some formatting changes; use rows with specified height rather than cellspacing in some places. - various fixes (source file links were wrong, amongst others) - - - - - 48722e68 by Simon Marlow at 2002-05-27T12:30:38+00:00 [haddock @ 2002-05-27 12:30:37 by simonmar] - Put function arguments *before* the doc for the function, as suggested by Sven Panne. This looks nicer when the function documentation is long. - Switch to using bold for binders at the definition site, and use underline for keywords. This makes the binder stand out more. - - - - - 657204d2 by Simon Marlow at 2002-05-27T13:19:49+00:00 [haddock @ 2002-05-27 13:19:49 by simonmar] Fix bug: we weren't renaming HsDocCommentNamed in renameDecl - - - - - 592aae66 by Simon Marlow at 2002-05-27T14:10:27+00:00 [haddock @ 2002-05-27 14:10:27 by simonmar] Fix some bugs in the rendering of qualified type signatures. - - - - - 69c8f763 by Simon Marlow at 2002-05-27T14:36:45+00:00 [haddock @ 2002-05-27 14:36:45 by simonmar] warning message tweak - - - - - 16e64e21 by Simon Marlow at 2002-05-27T14:53:53+00:00 [haddock @ 2002-05-27 14:53:53 by simonmar] hyperlinked identifiers should be in <tt> - - - - - 8d5e4783 by Simon Marlow at 2002-05-27T15:56:45+00:00 [haddock @ 2002-05-27 15:56:45 by simonmar] Do something sensible for modules which don't export anything (except instances). - - - - - 9d3ef811 by Simon Marlow at 2002-05-28T10:12:50+00:00 [haddock @ 2002-05-28 10:12:50 by simonmar] Rename the module documentation properly (bug reported by Sven Panne). - - - - - ef03a1cc by Simon Marlow at 2002-05-28T10:13:04+00:00 [haddock @ 2002-05-28 10:13:04 by simonmar] Add some more test cases - - - - - 92baa0e8 by Simon Marlow at 2002-05-28T11:17:55+00:00 [haddock @ 2002-05-28 11:17:55 by simonmar] If an identifier doesn't lex, then just replace it by a DocString. - - - - - a3156213 by Simon Marlow at 2002-05-28T16:16:19+00:00 [haddock @ 2002-05-28 16:16:19 by simonmar] Only link to names in the current module which are actually listed in the documentation. A name may be exported but not present in the documentation if it is exported as part of a 'module M' export specifier. - - - - - 31acf941 by Simon Marlow at 2002-05-28T16:17:11+00:00 [haddock @ 2002-05-28 16:17:11 by simonmar] update - - - - - 7e474ebf by Sigbjorn Finne at 2002-05-28T22:42:08+00:00 [haddock @ 2002-05-28 22:42:08 by sof] Handle lone occurrences of '/', e.g., -- | This/that. [did this in the lexer rather than in the parser, as I couldn't see a way not to introduce an S/R conflict that way.] - - - - - 093f7e53 by Simon Marlow at 2002-05-29T09:09:49+00:00 [haddock @ 2002-05-29 09:09:49 by simonmar] Back out previous change until we can find a better way to do this. - - - - - 9234389c by Simon Marlow at 2002-05-29T13:19:06+00:00 [haddock @ 2002-05-29 13:19:06 by simonmar] Make the markup syntax a little more friendly: - single quotes are now interpreted literally unless they surround a valid Haskell identifier. So for example now there's no need to escape a single quote used as an apostrophe. - text to the right of a bird track is now literal (if you want marked-up text in a code block, use @...@). - - - - - b3333526 by Simon Marlow at 2002-05-29T13:38:51+00:00 [haddock @ 2002-05-29 13:38:51 by simonmar] Document recent changes to markup syntax - - - - - f93641d6 by Simon Marlow at 2002-05-29T15:27:18+00:00 [haddock @ 2002-05-29 15:27:18 by simonmar] Include the instances in abstract data types too - - - - - 613f21e3 by Simon Marlow at 2002-06-03T13:05:58+00:00 [haddock @ 2002-06-03 13:05:57 by simonmar] Allow exporting of individual class methods and record selectors. For these we have to invent the correct type signature, which we do in the simplest possible way (i.e. no context reduction nonsense in the class case). - - - - - 14b36807 by Simon Marlow at 2002-06-03T13:20:00+00:00 [haddock @ 2002-06-03 13:20:00 by simonmar] Fix linking to qualified names again (thanks to Sven Panne for pointing out the bug). - - - - - 95b10eac by Simon Marlow at 2002-06-03T13:46:48+00:00 [haddock @ 2002-06-03 13:46:48 by simonmar] Fix for exporting record selectors from a newtype declaration - - - - - 272f932e by Simon Marlow at 2002-06-03T13:56:38+00:00 [haddock @ 2002-06-03 13:56:38 by simonmar] update to version 0.3 - - - - - 1c0a3bed by Simon Marlow at 2002-06-03T14:05:07+00:00 [haddock @ 2002-06-03 14:05:07 by simonmar] Add changes in version 0.3 - - - - - 145b4626 by Simon Marlow at 2002-06-03T14:12:38+00:00 [haddock @ 2002-06-03 14:12:38 by simonmar] Render class names as proper binders - - - - - 052106b3 by Simon Marlow at 2002-06-03T14:15:10+00:00 [haddock @ 2002-06-03 14:15:10 by simonmar] update, and separate into bugs, features, and cosmetic items. - - - - - 854f4914 by Simon Marlow at 2002-06-03T14:16:13+00:00 [haddock @ 2002-06-03 14:16:13 by simonmar] More test cases - - - - - 466922c8 by Simon Marlow at 2002-06-03T14:16:56+00:00 [haddock @ 2002-06-03 14:16:56 by simonmar] Example from the paper - - - - - 9962a045 by Simon Marlow at 2002-06-03T14:17:49+00:00 [haddock @ 2002-06-03 14:17:49 by simonmar] A debugging version of the style-sheet, which gives some tables coloured backgrounds so we can see what's going on. - - - - - f16b79db by Simon Marlow at 2002-06-03T14:19:46+00:00 [haddock @ 2002-06-03 14:19:46 by simonmar] typo - - - - - 620db27b by Simon Marlow at 2002-06-03T14:48:32+00:00 [haddock @ 2002-06-03 14:48:32 by simonmar] oops, fix markup bugs - - - - - 53fd105c by Simon Marlow at 2002-06-05T09:05:07+00:00 [haddock @ 2002-06-05 09:05:07 by simonmar] Keep foreign imports when there is no export list (bug reported by Sven Panne). - - - - - 6d98989c by Simon Marlow at 2002-06-05T09:12:02+00:00 [haddock @ 2002-06-05 09:12:02 by simonmar] Identifiers in single quotes can be symbol names too (bug reported by Hal Daume). - - - - - 001811e5 by Sven Panne at 2002-06-08T14:03:36+00:00 [haddock @ 2002-06-08 14:03:36 by panne] Tiny workaround for the fact that Haddock currently ignores HsImportSpecs: Let the local_orig_env take precedence. This is no real solution at all, but improves things sometimes, e.g. in my GLUT documentation. :-) - - - - - 504d19c9 by Simon Marlow at 2002-06-11T09:23:25+00:00 [haddock @ 2002-06-11 09:23:25 by simonmar] portability nit - - - - - e13b5af4 by Simon Marlow at 2002-06-20T12:38:07+00:00 [haddock @ 2002-06-20 12:38:07 by simonmar] Empty declaration fixes. - - - - - f467a9b6 by Simon Marlow at 2002-06-20T12:39:02+00:00 [haddock @ 2002-06-20 12:39:01 by simonmar] Add support for a "prologue" - a description for the whole library, placed on the contents page before the module list. - - - - - b8dbfe20 by Simon Marlow at 2002-06-21T12:43:06+00:00 [haddock @ 2002-06-21 12:43:06 by simonmar] When we have a single code block paragraph, don't place it in <pre>..</pre>, just use <tt>..</tt> to avoid generating extra vertical white space in some browsers. - - - - - 4831dbbd by Simon Marlow at 2002-06-21T15:50:42+00:00 [haddock @ 2002-06-21 15:50:42 by simonmar] Add support for reading and writing interface files(!) This turned out to be quite easy, and necessary to get decent hyperlinks between the documentation for separate packages in the libraries. The functionality isn't quite complete yet: for a given package of modules, you'd like to say "the HTML for these modules lives in directory <dir>" (currently they are assumed to be all in the same place). Two new flags: --dump-interface=FILE dump an interface file in FILE --read-interface=FILE read interface from FILE an interface file describes *all* the modules being processed. Only the exported names are kept in the interface: if you re-export a name from a module in another interface the signature won't be copied. This is a compromise to keep the size of the interfaces sensible. Also, I added another useful option: --no-implicit-prelude avoids trying to import the Prelude. Previously this was the default, but now importing the Prelude from elsewhere makes sense if you also read in an interface containing the Prelude module, so Haddock imports the Prelude implicitly according to the Haskell spec. - - - - - d3640a19 by Sven Panne at 2002-06-23T14:54:00+00:00 [haddock @ 2002-06-23 14:54:00 by panne] Make it compile with newer GHCs - - - - - 780c506b by Sven Panne at 2002-06-23T15:44:31+00:00 [haddock @ 2002-06-23 15:44:31 by panne] Cleaned up build root handling and added more docs - - - - - 45290d2e by Simon Marlow at 2002-06-24T14:37:43+00:00 [haddock @ 2002-06-24 14:37:42 by simonmar] When reading an interface, allow a file path offset to be specified which represents the path to the HTML files for the modules specified by that interface. The path may be either relative (to the location of the HTML for this package), or absolute. The syntax is --read-interface=PATH,FILE where PATH is the path to the HTML, and FILE is the filename containing the interface. - - - - - 4e2b9ae6 by Simon Marlow at 2002-07-03T16:01:08+00:00 [haddock @ 2002-07-03 16:01:07 by simonmar] Handle import specs properly, include 'hiding'. Haddock now has a complete implementation of the Haskell module system (more or less; I won't claim it's 100% correct). - - - - - 9a9aa1a8 by Simon Marlow at 2002-07-03T16:18:16+00:00 [haddock @ 2002-07-03 16:18:16 by simonmar] Update - - - - - 560c3026 by Simon Marlow at 2002-07-04T14:56:10+00:00 [haddock @ 2002-07-04 14:56:10 by simonmar] Clean up the code that constructs the exported declarations, and fix a couple of bugs along the way. Now if you import a class hiding one of the methods, then re-export the class, the version in the documentation will correctly have the appropriate method removed. - - - - - 2c26e77d by Simon Marlow at 2002-07-04T15:26:13+00:00 [haddock @ 2002-07-04 15:26:13 by simonmar] More bugfixes to the export handling - - - - - 03e0710d by Simon Marlow at 2002-07-09T10:12:10+00:00 [haddock @ 2002-07-09 10:12:10 by simonmar] Don't require that the list type comes from "Prelude" for it to be treated as special syntax (sometimes it comes from Data.List or maybe even GHC.Base). - - - - - 44f3891a by Simon Marlow at 2002-07-09T10:12:51+00:00 [haddock @ 2002-07-09 10:12:51 by simonmar] commented-out debugging code - - - - - 97280873 by Krasimir Angelov at 2002-07-09T16:33:33+00:00 [haddock @ 2002-07-09 16:33:31 by krasimir] 'Microsoft HTML Help' support - - - - - 3dc04655 by Simon Marlow at 2002-07-10T09:40:56+00:00 [haddock @ 2002-07-10 09:40:56 by simonmar] Fix for rendering of the (->) type constructor, from Ross Paterson. - - - - - c9f149c6 by Simon Marlow at 2002-07-10T10:26:11+00:00 [haddock @ 2002-07-10 10:26:11 by simonmar] Tweaks to the MS Help support: the extra files are now only generated if you ask for them (--ms-help). - - - - - e8acc1e6 by Simon Marlow at 2002-07-10T10:57:10+00:00 [haddock @ 2002-07-10 10:57:10 by simonmar] Document all the new options since 0.3 - - - - - 8bb85544 by Simon Marlow at 2002-07-10T10:58:31+00:00 [haddock @ 2002-07-10 10:58:31 by simonmar] Sort the options a bit - - - - - abc0dd59 by Simon Marlow at 2002-07-15T09:19:38+00:00 [haddock @ 2002-07-15 09:19:38 by simonmar] Fix a bug in mkExportItems when processing a module without an explicit export list. We were placing one copy of a declaration for each binder in the declaration, which for a data type would mean one copy of the whole declaration per constructor or record selector. - - - - - dde65bb9 by Simon Marlow at 2002-07-15T09:54:16+00:00 [haddock @ 2002-07-15 09:54:16 by simonmar] merge rev. 1.35 - - - - - bd7eb8c4 by Simon Marlow at 2002-07-15T10:14:31+00:00 [haddock @ 2002-07-15 10:14:30 by simonmar] Be a bit more liberal in the kind of commenting styles we allow, as suggested by Malcolm Wallace. Mostly this consists of allowing doc comments either side of a separator token. In an export list, a section heading is now allowed before the comma, as well as after it. eg. module M where ( T(..) -- * a section heading , f -- * another section heading , g ) In record fields, doc comments are allowed anywhere (previously a doc-next was allowed only after the comma, and a doc-before was allowed only before the comma). eg. data R = C { -- | describes 'f' f :: Int -- | describes 'g' , g :: Int } - - - - - 8f6dfe34 by Simon Marlow at 2002-07-15T10:21:56+00:00 [haddock @ 2002-07-15 10:21:56 by simonmar] Mention alternative commenting styles. - - - - - fc515bb7 by Simon Marlow at 2002-07-15T16:16:50+00:00 [haddock @ 2002-07-15 16:16:50 by simonmar] Allow multiple sections/subsections before and after a comma in the export list. Also at the same time I made the syntax a little stricter (multiple commas now aren't allowed between export specs). - - - - - 80a97e74 by Simon Marlow at 2002-07-19T09:13:10+00:00 [haddock @ 2002-07-19 09:13:10 by simonmar] Allow special id's ([], (), etc.) to be used in an import declaration. - - - - - a69d7378 by Simon Marlow at 2002-07-19T09:59:02+00:00 [haddock @ 2002-07-19 09:59:02 by simonmar] Allow special id's ([], (), etc.) to be used in an import declarations. - - - - - d205fa60 by Simon Marlow at 2002-07-19T10:00:16+00:00 [haddock @ 2002-07-19 10:00:16 by simonmar] Relax the restrictions which require doc comments to be followed by semi colons - in some cases this isn't necessary. Now you can write module M where { -- | some doc class C where {} } without needing to put a semicolon before the class declaration. - - - - - e9301e14 by Simon Marlow at 2002-07-23T08:24:09+00:00 [haddock @ 2002-07-23 08:24:09 by simonmar] A new TODO list item - - - - - e5d77586 by Simon Marlow at 2002-07-23T08:40:56+00:00 [haddock @ 2002-07-23 08:40:56 by simonmar] - update the acknowledgements - remove the paragraph that described how to use explicit layout with doc comments; it isn't relevant any more. - - - - - 78a94137 by Simon Marlow at 2002-07-23T08:43:02+00:00 [haddock @ 2002-07-23 08:43:02 by simonmar] more tests - - - - - 5c320927 by Simon Marlow at 2002-07-23T08:43:26+00:00 [haddock @ 2002-07-23 08:43:26 by simonmar] Updates for version 0.4 - - - - - 488e99ae by Simon Marlow at 2002-07-23T09:10:46+00:00 [haddock @ 2002-07-23 09:10:46 by simonmar] Fix the %changelog (rpm complained that it wasn't in the right order) - - - - - a77bb373 by Simon Marlow at 2002-07-23T09:12:38+00:00 [haddock @ 2002-07-23 09:12:38 by simonmar] Another item for the TODO list - - - - - f1ec1813 by Simon Marlow at 2002-07-23T10:18:46+00:00 [haddock @ 2002-07-23 10:18:46 by simonmar] Add a version banner when invoked with -v - - - - - 1d44cadf by Simon Marlow at 2002-07-24T09:28:19+00:00 [haddock @ 2002-07-24 09:28:19 by simonmar] Remove ^Ms - - - - - 4d8d5e94 by Simon Marlow at 2002-07-24T09:42:18+00:00 [haddock @ 2002-07-24 09:42:17 by simonmar] Patches to quieten ghc -Wall, from those nice folks at Galois. - - - - - d6edc43e by Simon Marlow at 2002-07-25T14:37:29+00:00 [haddock @ 2002-07-25 14:37:28 by simonmar] Patch to allow simple hyperlinking to an arbitrary location in another module's documentation, from Volker Stolz. Now in a doc comment: #foo# creates <a name="foo"></a> And you can use the form "M\#foo" to hyperlink to the label 'foo' in module 'M'. Note that the backslash is necessary for now. - - - - - b34d18fa by Simon Marlow at 2002-08-02T09:08:22+00:00 [haddock @ 2002-08-02 09:08:22 by simonmar] The <TT> and <PRE> environments seem to use a font that is a little too small in IE. Compensate. (suggestion from Daan Leijen). - - - - - 8106b086 by Simon Marlow at 2002-08-02T09:25:23+00:00 [haddock @ 2002-08-02 09:25:20 by simonmar] Remove <P>..</P> from around list items, to reduce excess whitespace between the items of bulleted and ordered lists. (Suggestion from Daan Leijen). - - - - - c1acff8f by Simon Marlow at 2002-08-05T09:03:49+00:00 [haddock @ 2002-08-05 09:03:49 by simonmar] update - - - - - f968661c by Simon Marlow at 2002-11-11T09:32:57+00:00 [haddock @ 2002-11-11 09:32:57 by simonmar] Fix cut-n-pasto - - - - - 12d02619 by Simon Marlow at 2002-11-13T09:49:46+00:00 [haddock @ 2002-11-13 09:49:46 by simonmar] Small bugfix in the --read-interface option parsing from Brett Letner. - - - - - 30e32d5e by Ross Paterson at 2003-01-16T15:07:57+00:00 [haddock @ 2003-01-16 15:07:57 by ross] Adjust for the new exception libraries (as well as the old ones). - - - - - 871f65df by Sven Panne at 2003-02-20T21:31:40+00:00 [haddock @ 2003-02-20 21:31:40 by panne] * Add varsyms and consyms to index * Exclude empty entries from index - - - - - bc42cc87 by Sven Panne at 2003-02-24T21:26:29+00:00 [haddock @ 2003-02-24 21:26:29 by panne] Don't convert a "newtype" to a single-constructor "data" for non-abstractly exported types, they are quite different regarding strictness/pattern matching. Now a "data" without any constructors is only emitted for an abstractly exported type, regardless if it is actually a "newtype" or a "data". - - - - - 0c2a1d99 by Sven Panne at 2003-03-08T19:02:38+00:00 [haddock @ 2003-03-08 19:02:38 by panne] Fixed some broken/redirected/canonicalized links found by a very picky link checker. - - - - - 25459269 by Sven Panne at 2003-03-09T21:13:43+00:00 [haddock @ 2003-03-09 21:13:43 by panne] Don't append a fragment to non-defining index entries, only documents with a defining occurrence have a name anchor. - - - - - 6be4db86 by Sven Panne at 2003-03-10T21:34:25+00:00 [haddock @ 2003-03-10 21:34:24 by panne] Escape fragments. This fixes e.g. links to operators. - - - - - eb12972c by Ross Paterson at 2003-04-25T10:50:06+00:00 [haddock @ 2003-04-25 10:50:05 by ross] An 80% solution to generating derived instances. A complete solution would duplicate the instance inference logic, but if a type variable occurs as a constructor argument, then we can just propagate the derived class to the variable. But we know nothing of the constraints on any type variables that occur elsewhere. For example, the declarations data Either a b = Left a | Right b deriving (Eq, Ord) data Ptr a = Ptr Addr# deriving (Eq, Ord) newtype IORef a = IORef (STRef RealWorld a) deriving Eq yield the instances (Eq a, Eq b) => Eq (Either a b) (Ord a, Ord b) => Ord (Either a b) Eq (Ptr a) Ord (Ptr a) (??? a) => Eq (IORef a) The last example shows the limits of this local analysis. Note that a type variable may be in both categories: then we know a constraint, but there may be more, or a stronger constraint, e.g. data Tree a = Node a [Tree a] deriving Eq yields (Eq a, ??? a) => Eq (Tree a) - - - - - de886f78 by Simon Marlow at 2003-04-25T11:17:55+00:00 [haddock @ 2003-04-25 11:17:55 by simonmar] Some updates, including moving the derived instance item down to the bottom of the list now that Ross has contributed some code that does the job for common cases. - - - - - 1b52cffd by Simon Marlow at 2003-04-30T14:02:32+00:00 [haddock @ 2003-04-30 14:02:32 by simonmar] When installing on Windows, run cygpath over $(HADDOCKLIB) so that haddock (a mingw program, built by GHC) can understand it. You still need to be in a cygwin environment to run Haddock, because of the shell script wrapper. - - - - - d4f638de by Simon Marlow at 2003-05-06T10:04:47+00:00 [haddock @ 2003-05-06 10:04:47 by simonmar] Catch another case of a paragraph containing just a DocMonospaced that should turn into a DocCodeBlock. - - - - - 4162b2b9 by Simon Marlow at 2003-05-06T10:11:44+00:00 [haddock @ 2003-05-06 10:11:44 by simonmar] Add some more code-block tests. - - - - - 4f5802c8 by Simon Marlow at 2003-05-06T10:14:52+00:00 [haddock @ 2003-05-06 10:14:52 by simonmar] Don't turn a single DocCodeBlock into a DocMonospaced, because that tends to remove the line breaks in the code. - - - - - ef8c45f7 by Simon Marlow at 2003-05-21T15:07:21+00:00 [haddock @ 2003-05-21 15:07:21 by simonmar] Only omit the module contents when there are no section headings at all. - - - - - bcee1e75 by Sigbjorn Finne at 2003-05-30T16:50:45+00:00 [haddock @ 2003-05-30 16:50:45 by sof] cygpath: for now, steer clear of --mixed - - - - - 30567af3 by Sigbjorn Finne at 2003-05-30T17:59:28+00:00 [haddock @ 2003-05-30 17:59:28 by sof] oops, drop test defn from prev commit - - - - - b0856e7d by Simon Marlow at 2003-06-03T09:55:26+00:00 [haddock @ 2003-06-03 09:55:26 by simonmar] Two small fixes to make the output valid HTML 4.01 (transitional). Thanks to Malcolm Wallace for pointing out the problems. - - - - - 70e137ea by Simon Marlow at 2003-07-28T13:30:35+00:00 [haddock @ 2003-07-28 13:30:35 by simonmar] Add tests for a couple of bugs. - - - - - 122bd578 by Simon Marlow at 2003-07-28T13:31:25+00:00 [haddock @ 2003-07-28 13:31:25 by simonmar] Add documentation for anchors. - - - - - 0bd27cb2 by Simon Marlow at 2003-07-28T13:31:46+00:00 [haddock @ 2003-07-28 13:31:46 by simonmar] Update - - - - - 08052d42 by Simon Marlow at 2003-07-28T13:32:12+00:00 [haddock @ 2003-07-28 13:32:12 by simonmar] layout tweak. - - - - - 13942749 by Simon Marlow at 2003-07-28T13:33:03+00:00 [haddock @ 2003-07-28 13:33:03 by simonmar] Differentiate links to types/classes from links to variables/constructors with a prefix ("t:" and "v:" respectively). - - - - - d7f493b9 by Simon Marlow at 2003-07-28T13:35:17+00:00 [haddock @ 2003-07-28 13:35:16 by simonmar] When a module A exports another module's contents via 'module B', then modules which import entities from B re-exported by A should link to B.foo rather than A.foo. See examples/Bug2.hs. - - - - - d94cf705 by Simon Marlow at 2003-07-28T13:36:14+00:00 [haddock @ 2003-07-28 13:36:14 by simonmar] Update to version 0.5 - - - - - dbb776cd by Sven Panne at 2003-07-28T14:02:43+00:00 [haddock @ 2003-07-28 14:02:43 by panne] * Updated to version 0.5 * Automagically generate configure if it is not there - - - - - 6cfeee53 by Simon Marlow at 2003-07-28T14:32:43+00:00 [haddock @ 2003-07-28 14:32:42 by simonmar] Update to avoid using hslibs with GHC >= 5.04 - - - - - a1ce838f by Simon Marlow at 2003-07-28T14:33:37+00:00 [haddock @ 2003-07-28 14:33:37 by simonmar] Update for 0.5 - - - - - c0fe6493 by Simon Marlow at 2003-07-28T14:53:22+00:00 [haddock @ 2003-07-28 14:53:22 by simonmar] Markup fix - - - - - 6ea31596 by Sven Panne at 2003-07-28T16:40:45+00:00 [haddock @ 2003-07-28 16:40:45 by panne] Make it compile with GHC >= 6.01 - - - - - afcd30fc by Simon Marlow at 2003-07-30T15:04:52+00:00 [haddock @ 2003-07-30 15:04:52 by simonmar] Pay attention to import specs when building the the import env, as well as the orig env. This may fix some wrong links in documentation when import specs are being used. - - - - - 17c3137f by Simon Marlow at 2003-07-30T16:05:41+00:00 [haddock @ 2003-07-30 16:05:40 by simonmar] Rename instances based on the import_env for the module in which they are to be displayed. This should give, in many cases, better links for the types and classes mentioned in the instance head. This involves keeping around the import_env in the iface until the end, because instances are not collected up until all the modules have been processed. Fortunately it doesn't seem to affect performance much. Instance heads are now attached to ExportDecls, rather than the HTML backend passing around a separate mapping for instances. This is a cleanup. - - - - - 3d3b5c87 by Sven Panne at 2003-08-04T10:18:24+00:00 [haddock @ 2003-08-04 10:18:24 by panne] Don't print parentheses around one-element contexts - - - - - 9e3f3f2d by Simon Marlow at 2003-08-04T12:59:47+00:00 [haddock @ 2003-08-04 12:59:47 by simonmar] A couple of TODOs. - - - - - e9d8085c by Simon Marlow at 2003-08-05T14:10:31+00:00 [haddock @ 2003-08-05 14:10:31 by simonmar] I'm not sure why, but it seems that the index entries for non-defining occurrences of entities did not have an anchor - the link just pointed to the module. This fixes it. - - - - - ff5c7d6d by Simon Marlow at 2003-08-15T14:42:59+00:00 [haddock @ 2003-08-15 14:42:59 by simonmar] Convert the lexer to Alex, and fix a bug in the process. - - - - - 1aa077bf by Simon Marlow at 2003-08-15T15:00:18+00:00 [haddock @ 2003-08-15 15:00:18 by simonmar] Update - - - - - d3de1e38 by Simon Marlow at 2003-08-15T15:01:03+00:00 [haddock @ 2003-08-15 15:01:03 by simonmar] wibbles - - - - - b40ece3b by Simon Marlow at 2003-08-18T10:04:47+00:00 [haddock @ 2003-08-18 10:04:47 by simonmar] Lex the 'mdo' keyword as 'do'. - - - - - 8f9a1146 by Simon Marlow at 2003-08-18T11:48:24+00:00 [haddock @ 2003-08-18 11:48:24 by simonmar] Two bugs from Sven. - - - - - ea54ebc0 by Simon Marlow at 2003-08-18T11:48:46+00:00 [haddock @ 2003-08-18 11:48:46 by simonmar] Fixes to the new lexer. - - - - - d5f6a4b5 by Simon Marlow at 2003-08-19T09:09:03+00:00 [haddock @ 2003-08-19 09:09:03 by simonmar] Further wibbles to the syntax. - - - - - 6bbdadb7 by Sven Panne at 2003-08-26T18:45:35+00:00 [haddock @ 2003-08-26 18:45:35 by panne] Use autoreconf instead of autoconf - - - - - 32e889cb by Sven Panne at 2003-08-26T19:01:19+00:00 [haddock @ 2003-08-26 19:01:18 by panne] Made option handling a bit more consistent with other tools, in particular: Every program in fptools should output * version info on stdout and terminate successfully when -V or --version * usage info on stdout and terminate successfully when -? or --help * usage info on stderr and terminate unsuccessfully when an unknown option is given. - - - - - 5d156a91 by Sven Panne at 2003-08-26T19:20:55+00:00 [haddock @ 2003-08-26 19:20:55 by panne] Make it *very* clear that we terminate when given a -V/--version flag - - - - - e6577265 by Sven Panne at 2003-08-27T07:50:03+00:00 [haddock @ 2003-08-27 07:50:02 by panne] * Made -D a short option for --dump-interface. * Made -m a short option for --ms-help. * Made -n a short option for --no-implicit-prelude. * Made -c a short option for --css. * Removed DocBook options from executable (they didn't do anything), but mark them as reserved in the docs. Note that the short option for DocBook output is now -S (from SGML) instead of -d. The latter is now a short option for --debug. * The order of the Options in the documentation now matches the order printed by Haddock itself. Note: Although changing the names of options is often a bad idea, I'd really like to make the options for the programs in fptools more consistent and compatible to the ones used in common GNU programs. - - - - - d303ff98 by Simon Marlow at 2003-09-10T08:23:48+00:00 [haddock @ 2003-09-10 08:23:48 by simonmar] Add doc subdir. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 9a70e46a by Simon Marlow at 2003-09-10T08:24:32+00:00 [haddock @ 2003-09-10 08:24:32 by simonmar] Install these files in $(datadir), not $(libdir), since they're architecture independent. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - bbb87e7a by Simon Marlow at 2003-09-10T08:25:31+00:00 [haddock @ 2003-09-10 08:25:31 by simonmar] Haddock's supplementary HTML bits now live in $(datadir), not $(libdir). Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 3587c24b by Simon Marlow at 2003-09-22T10:34:38+00:00 [haddock @ 2003-09-22 10:34:38 by simonmar] Allow installing of docs. - - - - - d510b517 by Sven Panne at 2003-10-11T08:10:44+00:00 [haddock @ 2003-10-11 08:10:44 by panne] Include architecture-independent files in file list - - - - - 187d7618 by Sigbjorn Finne at 2003-10-20T17:19:24+00:00 [haddock @ 2003-10-20 17:19:22 by sof] support for i-parameters + zip comprehensions - - - - - b6c7a273 by Simon Marlow at 2003-11-03T14:24:24+00:00 [haddock @ 2003-11-03 14:24:24 by simonmar] Update TODO file. - - - - - 58513e33 by Simon Marlow at 2003-11-05T11:22:04+00:00 [haddock @ 2003-11-05 11:22:04 by simonmar] Remove the last of the uses of 'trace' to emit warnings, and tidy up a couple of places where duplicate warnings were being emitted. - - - - - 33a78846 by Simon Marlow at 2003-11-05T11:30:53+00:00 [haddock @ 2003-11-05 11:30:52 by simonmar] - Suppress warnings about unknown imported modules by default. - Add a -v/--verbose flag to re-enable these warnings. The general idea is to suppress the "Warning: unknown module: Prelude" warnings which most Haddock users will see every time, and which aren't terribly useful. - - - - - a969de7f by Simon Marlow at 2003-11-05T12:30:28+00:00 [haddock @ 2003-11-05 12:30:28 by simonmar] - Remove the emboldening of index entries for defining locations. This isn't useful, and breaks abstractions. - If an entity is re-exported by a module but the module doesn't include documentation for that entity (perhaps because it is re-exported by 'module M'), then don't attempt to hyperlink to the documentation from the index. Instead, just list that module in the index, to indicate that the entity is exported from there. - - - - - f14ea82a by Simon Marlow at 2003-11-05T15:15:59+00:00 [haddock @ 2003-11-05 15:15:59 by simonmar] Index overhaul: - no more separate type/class and variable/function indices - the index now makes a distinction between different entities with the same name. One example is a type constructor with the same name as a data constructor, but another example is simply a function with the same name exported by two different modules. For example, the index entry for 'catch' now looks like this: catch 1 (Function) Control.Exception 2 (Function) GHC.Exception, Prelude, System.IO, System.IO.Error making it clear that there are two different 'catch'es, but one of them is exported by several modules. - Each index page now has the index contents (A B C ...) at the top. Please let me know if you really hate any of this. - - - - - 01a25ca6 by Simon Marlow at 2003-11-05T15:16:38+00:00 [haddock @ 2003-11-05 15:16:38 by simonmar] Update - - - - - 1a7ccb86 by Simon Marlow at 2003-11-05T17:16:05+00:00 [haddock @ 2003-11-05 17:16:04 by simonmar] Support for generating a single unified index for several packages. --use-index=URL turns off normal index generation, causes Index links to point to URL. --gen-index generates an combined index from the specified interfaces. Currently doesn't work exactly right, because the interfaces don't contain the iface_reexported info. I'll need to fix that up. - - - - - a2bca16d by Simon Marlow at 2003-11-06T10:44:52+00:00 [haddock @ 2003-11-06 10:44:52 by simonmar] Include iface_reexported in the .haddock file. This unfortunately bloats the file (40% for base). If this gets to be a problem we can always apply the dictionary trick that GHC uses for squashing .hi files. - - - - - 0a09c293 by Simon Marlow at 2003-11-06T12:39:47+00:00 [haddock @ 2003-11-06 12:39:46 by simonmar] - Add definition lists, marked up like this: -- | This is a definition list: -- -- [@foo@] The description of @foo at . -- -- [@bar@] The description of @bar at . Cunningly, the [] characters are not treated specially unless a [ is found at the beginning of a paragraph, in which case the ] becomes special in the following text. - Add --use-contents and --gen-contents, along the lines of --use-index and --gen-index added yesterday. Now we can generate a combined index and contents for the whole of the hierarchical libraries, and in theory the index/contents on the system could be updated as new packages are added. - - - - - fe1b3460 by Simon Marlow at 2003-11-06T14:47:36+00:00 [haddock @ 2003-11-06 14:47:36 by simonmar] Remove the 'Parent' button - it is of dubious use, and often points into thin air. - - - - - db6d762f by Simon Marlow at 2003-11-06T16:48:14+00:00 [haddock @ 2003-11-06 16:48:11 by simonmar] - Include the OptHide setting in the interface, so we don't include hidden modules in the combined index/contents. - Add a -k/--package flag to set the package name for the current set of modules. The package name for each module is now shown in the right-hand column of the contents, in a combined contents page. - - - - - 7d71718b by Simon Marlow at 2003-11-06T16:50:28+00:00 [haddock @ 2003-11-06 16:50:28 by simonmar] Add -k/--package docs - - - - - ef43949d by Simon Marlow at 2003-11-06T16:51:23+00:00 [haddock @ 2003-11-06 16:51:23 by simonmar] Bump to 0.6 - - - - - 1c419e06 by Simon Marlow at 2003-11-06T16:51:50+00:00 [haddock @ 2003-11-06 16:51:50 by simonmar] update - - - - - 69422327 by Simon Marlow at 2003-11-10T14:41:06+00:00 [haddock @ 2003-11-10 14:41:05 by simonmar] Re-exporting names from a different package is problematic, because we don't have access to the full documentation for the entity. Currently Haddock just ignores entities with no documentation, but this results in bogus-looking empty documentation for many of the modules in the haskell98 package. So: - the documentation will now just list the name, as a link pointing to the location of the actual documentation. - now we don't attempt to link to these re-exported entities if they are referred to by the current module. Additionally: - If there is no documentation in the current module, include just the Synopsis section (rather than just the documentation section, as it was before). This just looks nicer and was on the TODO list. - - - - - 3c3fc433 by Simon Marlow at 2003-11-10T14:51:59+00:00 [haddock @ 2003-11-10 14:51:59 by simonmar] Fix for getReExports: take into account names which are not visible because they are re-exported from a different package. - - - - - 31c8437b by Simon Marlow at 2003-11-10T15:10:53+00:00 [haddock @ 2003-11-10 15:10:53 by simonmar] Version 0.6 changes - - - - - a7c2430b by Simon Marlow at 2003-11-10T15:15:58+00:00 [haddock @ 2003-11-10 15:15:58 by simonmar] getReExports: one error case that isn't - - - - - 00cc459c by Simon Marlow at 2003-11-10T16:15:19+00:00 [haddock @ 2003-11-10 16:15:18 by simonmar] copyright update - - - - - ca62408d by Simon Marlow at 2003-11-11T09:57:25+00:00 [haddock @ 2003-11-11 09:57:25 by simonmar] Version 0.6 - - - - - 3acbf818 by Simon Marlow at 2003-11-11T12:10:44+00:00 [haddock @ 2003-11-11 12:10:44 by simonmar] Go back to producing just the documentation section, rather than just the synopsis section, for a module with no documentation annotations. One reason is that the synopsis section tries to link each entity to its documentation on the same page. Also, the doc section anchors each entity, and it lists instances which the synopsis doesn't. - - - - - 6c90abc2 by Simon Marlow at 2003-11-12T10:03:39+00:00 [haddock @ 2003-11-12 10:03:39 by simonmar] 2002 -> 2003 - - - - - 090bbc4c by Simon Marlow at 2003-11-28T12:08:00+00:00 [haddock @ 2003-11-28 12:08:00 by simonmar] update - - - - - 8096a832 by Simon Marlow at 2003-11-28T12:09:58+00:00 [haddock @ 2003-11-28 12:09:58 by simonmar] Fix some of the problems with Haddock generating pages that are too wide. Now we only specify 'nowrap' when it is necessary to avoid a code box getting squashed up by the text to the right of it. - - - - - 35294929 by Sven Panne at 2003-12-29T17:16:31+00:00 [haddock @ 2003-12-29 17:16:31 by panne] Updated my email address - - - - - cdb697bf by Simon Marlow at 2004-01-08T10:14:24+00:00 [haddock @ 2004-01-08 10:14:24 by simonmar] Add instructions for using GHC to pre-process source for feeding to Haddock. - - - - - 8dfc491f by Simon Marlow at 2004-01-09T12:45:46+00:00 [haddock @ 2004-01-09 12:45:46 by simonmar] Add -optP-P to example ghc command line. - - - - - ac41b820 by Simon Marlow at 2004-02-03T11:02:03+00:00 [haddock @ 2004-02-03 11:02:03 by simonmar] Fix bug in index generation - - - - - f4e7edcb by Simon Marlow at 2004-02-10T11:51:16+00:00 [haddock @ 2004-02-10 11:51:16 by simonmar] Don't throw away whitespace at the beginning of a line (experimental fix). - - - - - 68e212d2 by Simon Marlow at 2004-02-10T12:10:08+00:00 [haddock @ 2004-02-10 12:10:08 by simonmar] Fix for previous commit: I now realise why the whitespace was stripped from the beginning of the line. Work around it. - - - - - e7d7f2df by Sven Panne at 2004-02-10T18:38:45+00:00 [haddock @ 2004-02-10 18:38:45 by panne] Make Haddock link with the latest relocated monad transformer package - - - - - 992d4225 by Simon Marlow at 2004-02-16T10:21:35+00:00 [haddock @ 2004-02-16 10:21:35 by simonmar] Add a TODO - - - - - 1ac55326 by Simon Marlow at 2004-03-12T11:33:39+00:00 [haddock @ 2004-03-12 11:33:39 by simonmar] Add an item. - - - - - 0478e903 by Simon Marlow at 2004-03-15T12:24:05+00:00 [haddock @ 2004-03-15 12:24:05 by simonmar] Add an item. - - - - - 6f26d21a by Simon Marlow at 2004-03-18T14:21:29+00:00 [haddock @ 2004-03-18 14:21:29 by simonmar] Fix URL - - - - - 19b6bb99 by Simon Marlow at 2004-03-22T14:09:03+00:00 [haddock @ 2004-03-22 14:09:03 by simonmar] getReExports was bogus: we should really look in the import_env to find the documentation for an entity which we are re-exporting without documentation. Suggested by: Ross Paterson (patch modified by me). - - - - - 5c756031 by Simon Marlow at 2004-03-24T09:42:11+00:00 [haddock @ 2004-03-24 09:42:10 by simonmar] hiding bug from Ross Paterson (fixed in rev 1.59 of Main.hs) - - - - - 1b692e6c by Simon Marlow at 2004-03-24T10:10:50+00:00 [haddock @ 2004-03-24 10:10:50 by simonmar] mkExportItems fix & simplification: we should be looking at the actual exported names (calculated earlier) to figure out which subordinates of a declaration are exported. This means that if you export a record, and name its fields separately in the export list, the fields will still be visible in the documentation for the constructor. - - - - - 90e5e294 by Simon Marlow at 2004-03-24T10:12:08+00:00 [haddock @ 2004-03-24 10:12:08 by simonmar] Make restrictCons take into account record field names too (removing a ToDo). - - - - - 2600efa4 by Simon Marlow at 2004-03-24T10:16:17+00:00 [haddock @ 2004-03-24 10:16:17 by simonmar] Record export tests. - - - - - 6a8575c7 by Simon Marlow at 2004-03-25T09:35:14+00:00 [haddock @ 2004-03-25 09:35:14 by simonmar] restrictTo: fix for restricting a newtype with a record field. - - - - - dcf55a8d by Simon Marlow at 2004-03-25T10:01:42+00:00 [haddock @ 2004-03-25 10:01:42 by simonmar] Fix duplicate instance bug - - - - - f49aa758 by Simon Marlow at 2004-03-25T10:02:41+00:00 [haddock @ 2004-03-25 10:02:41 by simonmar] Duplicate instance bug. - - - - - 7b87344c by Simon Marlow at 2004-03-25T10:29:56+00:00 [haddock @ 2004-03-25 10:29:56 by simonmar] If a name is imported from two places, one hidden and one not, choose the unhidden one to link to. Also, when there's only a hidden module to link to, don't try linking to it. - - - - - 40f44d7b by Simon Marlow at 2004-03-25T15:17:24+00:00 [haddock @ 2004-03-25 15:17:23 by simonmar] Add support for collaspible parts of the page, with a +/- button and a bit of JavaScript. Make the instances collapsible, and collapse them by default. This makes documentation with long lists of instances (eg. the Prelude) much easier to read. Maybe we should give other documentation sections the same treatment. - - - - - 9b64dc0f by Simon Marlow at 2004-03-25T15:20:55+00:00 [haddock @ 2004-03-25 15:20:55 by simonmar] Update - - - - - c2fff7f2 by Simon Marlow at 2004-03-25T15:45:10+00:00 [haddock @ 2004-03-25 15:45:10 by simonmar] Eliminate some unnecessary spaces in the HTML rendering - - - - - b7948ff0 by Simon Marlow at 2004-03-25T16:00:37+00:00 [haddock @ 2004-03-25 16:00:36 by simonmar] Remove all that indentation in the generated HTML to keep the file sizes down. - - - - - da2bb4ca by Sven Panne at 2004-03-27T09:57:58+00:00 [haddock @ 2004-03-27 09:57:57 by panne] Added the new-born haddock.js to the build process and the documentation. - - - - - b99e6f8c by Sven Panne at 2004-03-27T10:32:20+00:00 [haddock @ 2004-03-27 10:32:20 by panne] "type" is a required attribute of the "script" element - - - - - 562b185a by Sven Panne at 2004-03-27T12:52:34+00:00 [haddock @ 2004-03-27 12:52:34 by panne] Add a doctype for the contents page, too. - - - - - f6a99c2d by Simon Marlow at 2004-04-14T10:03:25+00:00 [haddock @ 2004-04-14 10:03:25 by simonmar] fix for single-line comment syntax - - - - - de366303 by Simon Marlow at 2004-04-20T13:08:04+00:00 [haddock @ 2004-04-20 13:08:04 by simonmar] Allow a 'type' declaration to include documentation comments. These will be ignored by Haddock, but at least one user (Johannes Waldmann) finds this feature useful, and it's easy to add. - - - - - fd78f51e by Simon Marlow at 2004-05-07T15:14:56+00:00 [haddock @ 2004-05-07 15:14:56 by simonmar] - update copyright - add version to abstract - - - - - 59f53e32 by Sven Panne at 2004-05-09T14:39:53+00:00 [haddock @ 2004-05-09 14:39:53 by panne] Fix the fix for single-line comment syntax, ------------------------------------------- is now a valid comment line again. - - - - - 8b18f2fe by Simon Marlow at 2004-05-10T10:11:51+00:00 [haddock @ 2004-05-10 10:11:51 by simonmar] Update - - - - - 225a491d by Ross Paterson at 2004-05-19T13:10:23+00:00 [haddock @ 2004-05-19 13:10:23 by ross] Make the handling of "deriving" slightly smarter, by ignoring data constructor arguments that are identical to the lhs. Now handles things like data Tree a = Leaf a | Branch (Tree a) (Tree a) deriving ... - - - - - 37588686 by Mike Thomas at 2004-05-21T06:38:14+00:00 [haddock @ 2004-05-21 06:38:14 by mthomas] Windows exe extensions (bin remains for Unix). - - - - - cf2b9152 by Simon Marlow at 2004-05-25T09:34:54+00:00 [haddock @ 2004-05-25 09:34:54 by simonmar] Add some TODO items - - - - - 4d29cdfc by Simon Marlow at 2004-05-25T10:41:46+00:00 [haddock @ 2004-05-25 10:41:46 by simonmar] Complain if -h is used with --gen-index or --gen-contents, because it'll overwrite the new index/contents. - - - - - 2e0771e0 by Mike Thomas at 2004-05-28T20:17:55+00:00 [haddock @ 2004-05-28 20:17:55 by mthomas] Windows: search for templates in executable directory. Unix: Haddock tries cwd first rather than error if no -l arg. - - - - - 8d10bde1 by Sven Panne at 2004-06-05T16:53:34+00:00 [haddock @ 2004-06-05 16:53:34 by panne] Misc. rpm spec file cleanup, including: * make BuildRoot handling more consistent * added default file attributes * consistent defines and tags - - - - - 59974349 by Sven Panne at 2004-06-05T18:01:00+00:00 [haddock @ 2004-06-05 18:01:00 by panne] More rpm spec file cleanup, including: * added some BuildRequires * changed packager to me, so people can complain at the right place :-] * consistently refer to haskell.org instead of www.haskell.org - - - - - b94d4903 by Simon Marlow at 2004-07-01T11:08:58+00:00 [haddock @ 2004-07-01 11:08:57 by simonmar] Update to the +/- buttons: use a resized image rather than a <button>. Still seeing some strange effects in Konqueror, so might need to use a fixed-size image instead. - - - - - d5278f67 by Sven Panne at 2004-07-04T15:15:55+00:00 [haddock @ 2004-07-04 15:15:55 by panne] Install pictures for +/- pictures, too (JPEG is a strange format for graphics like this, I would have expected GIF or PNG here.) Things look fine with Konqueror and Netscape on Linux now, the only downside is that the cursor doesn't change when positioned above the "button". - - - - - 46dec6c5 by Sven Panne at 2004-07-13T17:59:28+00:00 [haddock @ 2004-07-13 17:59:28 by panne] A quote is a valid part of a Haskell identifier, but it would interfere with an ECMA script string delimiter, so escape it there. - - - - - 1d7bc432 by Simon Marlow at 2004-07-22T08:54:06+00:00 [haddock @ 2004-07-22 08:54:06 by simonmar] Add single quote to $ident, so you can say eg. 'foldl'' to refer to foldl' (the longest match rule is our friend). Bug reported by Adrian Hey <ahey at iee.org> - - - - - f183618b by Krasimir Angelov at 2004-07-27T22:59:35+00:00 [haddock @ 2004-07-27 22:58:23 by krasimir] Add basic support for Microsoft HTML Help 2.0 - - - - - d515d0c2 by Krasimir Angelov at 2004-07-27T23:02:36+00:00 [haddock @ 2004-07-27 23:02:36 by krasimir] escape names in the index - - - - - a5f1be23 by Krasimir Angelov at 2004-07-27T23:05:21+00:00 [haddock @ 2004-07-27 23:05:21 by krasimir] Add jsFile, plusFile and minusFile to the file list - - - - - c4fb4881 by Krasimir Angelov at 2004-07-28T22:12:10+00:00 [haddock @ 2004-07-28 22:12:09 by krasimir] bugfix. Move contentsHtmlFile, indexHtmlFile and subIndexHtmlFile functions to HaddockUtil.hs module to make them accessible from HaddockHH2.hs - - - - - 64d30b1d by Krasimir Angelov at 2004-07-30T22:15:47+00:00 [haddock @ 2004-07-30 22:15:45 by krasimir] more stuffs - support for separated compilation of packages - the contents page now uses DHTML TreeView - fixed copyFile bug - - - - - 133c8c5c by Krasimir Angelov at 2004-07-31T12:04:38+00:00 [haddock @ 2004-07-31 12:04:37 by krasimir] make the DHtmlTree in contents page more portable. The +/- buttons are replaced with new images which looks more beatiful. - - - - - 79040963 by Krasimir Angelov at 2004-07-31T13:10:20+00:00 [haddock @ 2004-07-31 13:10:20 by krasimir] Make DHtmlTree compatible with Mozila browser - - - - - 1a55dc90 by Krasimir Angelov at 2004-07-31T14:52:55+00:00 [haddock @ 2004-07-31 14:52:55 by krasimir] fix - - - - - 85ce0237 by Krasimir Angelov at 2004-07-31T14:53:28+00:00 [haddock @ 2004-07-31 14:53:28 by krasimir] HtmlHelp 1.x - - - - - 3c0c53ba by Krasimir Angelov at 2004-07-31T20:35:21+00:00 [haddock @ 2004-07-31 20:35:21 by krasimir] Added support for DevHelp - - - - - d42b5af1 by Krasimir Angelov at 2004-07-31T21:17:51+00:00 [haddock @ 2004-07-31 21:17:51 by krasimir] Document new features in HtmlHelp - - - - - 790fe21e by Krasimir Angelov at 2004-08-01T15:14:02+00:00 [haddock @ 2004-08-01 15:14:02 by krasimir] add missing imports - - - - - fd7cc6bc by Krasimir Angelov at 2004-08-01T19:52:08+00:00 [haddock @ 2004-08-01 19:52:06 by krasimir] fix some bugs. Now I have got the entire libraries documentation in HtmlHelp 2.0 format. - - - - - 94ad7ac8 by Krasimir Angelov at 2004-08-01T19:53:50+00:00 [haddock @ 2004-08-01 19:53:50 by krasimir] I forgot to add the new +/- images - - - - - f0c65388 by Krasimir Angelov at 2004-08-02T16:25:53+00:00 [haddock @ 2004-08-02 16:25:53 by krasimir] Add root node to the table of contents. All modules in tree are not children of the root - - - - - f50bd85d by Sven Panne at 2004-08-02T18:17:46+00:00 [haddock @ 2004-08-02 18:17:46 by panne] Mainly DocBook fixes - - - - - 09527ce3 by Sven Panne at 2004-08-02T20:02:29+00:00 [haddock @ 2004-08-02 20:02:29 by panne] Fixed -o/--odir handling. Generating the output, especially the directory handling, is getting a bit convoluted nowadays... - - - - - c8fbacfa by Sven Panne at 2004-08-02T20:31:13+00:00 [haddock @ 2004-08-02 20:31:13 by panne] Warning police - - - - - 37830bff by Sven Panne at 2004-08-02T20:32:29+00:00 [haddock @ 2004-08-02 20:32:28 by panne] Nuked dead code - - - - - 13847171 by Sven Panne at 2004-08-02T21:12:27+00:00 [haddock @ 2004-08-02 21:12:25 by panne] Use pathJoin instead of low-level list-based manipulation for FilePaths - - - - - c711d61e by Sven Panne at 2004-08-02T21:16:02+00:00 [haddock @ 2004-08-02 21:16:02 by panne] Removed WinDoze CRs - - - - - b1f7dc88 by Sven Panne at 2004-08-03T19:35:59+00:00 [haddock @ 2004-08-03 19:35:59 by panne] Fixed spelling of "http-equiv" attribute - - - - - dd5f394e by Sven Panne at 2004-08-03T19:44:03+00:00 [haddock @ 2004-08-03 19:44:03 by panne] Pacify W3C validator: * Added document encoding (currently UTF-8, not sure if this is completely correct) * Fixed syntax of `id' attributes * Added necessary `alt' attribute for +/- images Small layout improvement: * Added space after +/- images (still not perfect, but better than before) - - - - - 919c47c6 by Sigbjorn Finne at 2004-08-03T19:45:11+00:00 [haddock @ 2004-08-03 19:45:11 by sof] make it compile with <= ghc-6.1 - - - - - 4d6f01d8 by Sigbjorn Finne at 2004-08-03T19:45:30+00:00 [haddock @ 2004-08-03 19:45:30 by sof] ffi wibble - - - - - 4770643a by Sven Panne at 2004-08-03T20:47:46+00:00 [haddock @ 2004-08-03 20:47:46 by panne] Fixed CSS for button style. Note that only "0" is a valid measure without a unit! - - - - - 14aaf2e5 by Sven Panne at 2004-08-03T21:07:59+00:00 [haddock @ 2004-08-03 21:07:58 by panne] Improved spacing of dynamic module tree - - - - - 97c3579a by Simon Marlow at 2004-08-09T11:03:04+00:00 [haddock @ 2004-08-09 11:03:04 by simonmar] Add FormatVersion Patch submitted by: George Russell <ger at informatik.uni-bremen.de> - - - - - af7f8c03 by Simon Marlow at 2004-08-09T11:55:07+00:00 [haddock @ 2004-08-09 11:55:05 by simonmar] Add support for a short description for each module, which is included in the contents. The short description should be given in a "Description: " field of the header. Included in this patch are changes that make the format of the header a little more flexible. From the comments: -- all fields in the header are optional and have the form -- -- [spaces1][field name][spaces] ":" -- [text]"\n" ([spaces2][space][text]"\n" | [spaces]"\n")* -- where each [spaces2] should have [spaces1] as a prefix. -- -- Thus for the key "Description", -- -- > Description : this is a -- > rather long -- > -- > description -- > -- > The module comment starts here -- -- the value will be "this is a .. description" and the rest will begin -- at "The module comment". The header fields must be in the following order: Module, Description, Copyright, License, Maintainer, Stability, Portability. Patches submitted by: George Russell <ger at informatik.uni-bremen.de>, with a few small changes be me, mostly to merge with other recent changes. ToDo: document the module header. - - - - - 7b865ad3 by Simon Marlow at 2004-08-10T14:09:57+00:00 [haddock @ 2004-08-10 14:09:57 by simonmar] Fixes for DevHelp/HtmlHelp following introduction of short module description. - - - - - 814766cd by Simon Marlow at 2004-08-10T14:33:46+00:00 [haddock @ 2004-08-10 14:33:45 by simonmar] Fixes to installation under Windows. - - - - - 39cf9ede by Simon Marlow at 2004-08-12T12:08:23+00:00 [haddock @ 2004-08-12 12:08:23 by simonmar] Avoid using string-gap tricks. - - - - - b6d78551 by Simon Marlow at 2004-08-13T10:53:21+00:00 [haddock @ 2004-08-13 10:53:21 by simonmar] Update - - - - - eaae7417 by Simon Marlow at 2004-08-13T10:53:50+00:00 [haddock @ 2004-08-13 10:53:50 by simonmar] Test for primes in quoted links - - - - - 68c34f06 by Sven Panne at 2004-08-16T19:59:38+00:00 [haddock @ 2004-08-16 19:59:36 by panne] XMLification - - - - - 7f45a6f9 by Sven Panne at 2004-08-18T16:42:54+00:00 [haddock @ 2004-08-18 16:42:54 by panne] Re-added indices + minor fixes - - - - - 8a5dd97c by Sigbjorn Finne at 2004-08-25T17:15:42+00:00 [haddock @ 2004-08-25 17:15:42 by sof] backquote HADDOCK_VERSION defn for <= ghc-6.0.x; believe this is only needed under mingw - - - - - 4b1b42ea by Sven Panne at 2004-08-26T20:08:50+00:00 [haddock @ 2004-08-26 20:08:49 by panne] SGML is dead, long live DocBook XML! Note: The BuildRequires tags in the spec files are still incomplete and the documentation about the DocBook tools needs to be updated, too. Stay tuned... - - - - - 8d52cedb by Sven Panne at 2004-08-26T21:03:19+00:00 [haddock @ 2004-08-26 21:03:19 by panne] Updated BuildRequires tags. Alas, there seems to be no real standard here, so your mileage may vary... At least the current specs should work on SuSE Linux. - - - - - e6982912 by Sigbjorn Finne at 2004-08-30T15:44:59+00:00 [haddock @ 2004-08-30 15:44:59 by sof] escape HADDOCK_VERSION double quotes on all platforms when compiling with <=6.0.x - - - - - b3fbc867 by Simon Marlow at 2004-08-31T13:09:42+00:00 [haddock @ 2004-08-31 13:09:42 by simonmar] Avoid GHC/shell versionitis and create Version.hs - - - - - c359e16a by Sven Panne at 2004-09-05T19:12:33+00:00 [haddock @ 2004-09-05 19:12:32 by panne] * HTML documentation for "foo.xml" goes into directory "foo" again, not "foo-html". This is nicer and consistent with the behaviour for building the docs from SGML. * Disabled building PostScript documentation in the spec files for now, there are some strange issues with the FO->PS conversion for some files which have to be clarified first. - - - - - c68b1eba by Sven Panne at 2004-09-24T07:04:38+00:00 [haddock @ 2004-09-24 07:04:38 by panne] Switched the default state for instances and the module hierarchy to non-collapsed. This can be reversed when we finally use cookies from JavaScript to have a more persistent state. Previously going back and forth in the documentation was simply too annoying because everything was collapsed again and therefore the documentation was not easily navigatable. - - - - - dfb32615 by Simon Marlow at 2004-09-30T08:21:29+00:00 [haddock @ 2004-09-30 08:21:29 by simonmar] Add a feature request - - - - - 45ff783c by Sven Panne at 2004-10-23T19:54:00+00:00 [haddock @ 2004-10-23 19:54:00 by panne] Improved the Cygwin/MinGW chaos a little bit. There is still confusion about host platform vs. target platform... - - - - - 5f644714 by Krasimir Angelov at 2004-10-28T16:01:51+00:00 [haddock @ 2004-10-28 16:01:51 by krasimir] update for ghc-6.3+ - - - - - 92d9753e by Sven Panne at 2004-11-01T16:39:01+00:00 [haddock @ 2004-11-01 16:39:01 by panne] Revert previous commit: It's Network.URI which should be changed, not Haddock. - - - - - 05f70f6e by Simon Marlow at 2005-01-04T16:15:51+00:00 [haddock @ 2005-01-04 16:15:51 by simonmar] parser fix: allow qualified specialids. - - - - - 47870837 by Simon Marlow at 2005-01-04T16:16:54+00:00 [haddock @ 2005-01-04 16:16:54 by simonmar] Add a test - - - - - ff11fc2c by Ross Paterson at 2005-01-10T19:18:22+00:00 [haddock @ 2005-01-10 19:18:22 by ross] Render non-ASCII characters using numeric character references, to simplify charset issues. There's a META tag saying the charset is UTF-8, but GHC outputs characters as raw bytes. Ideally we need an encoding on the input side too, primarily in comments, because source files containing non-ASCII characters aren't portable between locales. - - - - - eba2fc4e by Simon Marlow at 2005-01-11T10:44:37+00:00 [haddock @ 2005-01-11 10:44:37 by simonmar] Remove string gap - - - - - b899a381 by Ross Paterson at 2005-01-13T11:41:33+00:00 [haddock @ 2005-01-13 11:41:33 by ross] recognize SGML-style numeric character references &#ddd; or &#xhhhh; and translate them into Chars. - - - - - 106e3cf0 by Ross Paterson at 2005-01-13T14:43:41+00:00 [haddock @ 2005-01-13 14:43:41 by ross] also allow uppercase X in hexadecimal character references (like SGML) - - - - - e8f54f25 by Ross Paterson at 2005-01-13T14:44:24+00:00 [haddock @ 2005-01-13 14:44:24 by ross] Describe numeric character references. - - - - - 914ccdce by Sven Panne at 2005-01-15T18:44:48+00:00 [haddock @ 2005-01-15 18:44:45 by panne] Make Haddock compile again after the recent base package changed. The Map/Set legacy hell has been factored out, so that all modules can simply use the new non-deprecated interfaces. Probably a lot of things can be improved by a little bit of Map/Set/List algebra, this can be done later if needed. Small note: Currently the list of instances in HTML code is reversed. This will hopefully be fixed later. - - - - - 6ab20e84 by Sven Panne at 2005-01-16T12:18:26+00:00 [haddock @ 2005-01-16 12:18:26 by panne] Trim imports - - - - - efb81da9 by Sven Panne at 2005-01-16T12:58:08+00:00 [haddock @ 2005-01-16 12:58:03 by panne] Correctly handle the new order of arguments for the combining function given to fromListWith. - - - - - e27b5834 by Sven Panne at 2005-01-16T14:14:41+00:00 [haddock @ 2005-01-16 14:14:39 by panne] Data.Map.unions is left-biased. - - - - - dae3cc3e by Sven Panne at 2005-01-16T14:22:44+00:00 [haddock @ 2005-01-16 14:22:44 by panne] Added the last missing "flip" to get identical HTML output as previous versions. - - - - - 951d8408 by Sven Panne at 2005-01-16T14:37:10+00:00 [haddock @ 2005-01-16 14:37:10 by panne] Refactored Text.PrettyPrint legacy hell into a separate module. - - - - - f1c4b892 by Sven Panne at 2005-01-16T15:41:25+00:00 [haddock @ 2005-01-16 15:41:21 by panne] Cleaned up imports and dropped support for GHC < 5.03, it never worked, anyway. - - - - - 60824c6e by Simon Marlow at 2005-01-18T10:02:48+00:00 [haddock @ 2005-01-18 10:02:48 by simonmar] Add a TODO - - - - - a8c82f23 by Krasimir Angelov at 2005-01-28T23:19:39+00:00 [haddock @ 2005-01-28 23:19:39 by krasimir] import Foreign/Foreign.C are required for Windows - - - - - d8450a23 by Simon Marlow at 2005-02-02T16:23:04+00:00 [haddock @ 2005-02-02 16:23:00 by simonmar] Revamp the linking strategy in Haddock. Now name resolution is done in two phases: - first resolve everything to original names, like a Haskell compiler would. - then, figure out the "home" location for every entity, and point all the links to there. The home location is the lowest non-hidden module in the import hierarchy that documents the entity. If there are multiple candidates, one is chosen at random. Also: - Haddock should not generate any HTML with dangling links any more. Unlinked references are just rendered as plain text. - Error reporting is better: if we can't find a link destination for an entity reference, we now emit a warning. - - - - - 1cce71d0 by Simon Marlow at 2005-02-03T13:42:19+00:00 [haddock @ 2005-02-03 13:42:19 by simonmar] - add --ignore-all-exports flag, which behaves as if every module has the ignore-exports attribute (requested by Chris Ryder). - add --hide option to hide a module on the command line. - add --use-package option to get Haddock info for a package from ghc-pkg (largely untested). - remove reexports from the .haddock file, they aren't used any more. - - - - - 767123ef by Ross Paterson at 2005-02-03T16:17:37+00:00 [haddock @ 2005-02-03 16:17:37 by ross] fix typo for < 6.3 - - - - - 0c680c04 by Simon Marlow at 2005-02-04T12:03:31+00:00 [haddock @ 2005-02-04 12:03:31 by simonmar] Fix bug in renameExportItems that meant links in instances weren't being renamed properly. - - - - - ff7abe5f by Simon Marlow at 2005-02-04T12:15:53+00:00 [haddock @ 2005-02-04 12:15:52 by simonmar] Add attribute #not-home, to indicate that the current module should not be considered to be a home module for the each entity it exports, unless there is no other module that exports the entity. - - - - - fc2cfd27 by Simon Marlow at 2005-02-04T12:40:02+00:00 [haddock @ 2005-02-04 12:40:02 by simonmar] Update the documentation w.r.t. home modules and the not-home attribute. - - - - - 26b8ddf7 by Ross Paterson at 2005-02-04T13:36:06+00:00 [haddock @ 2005-02-04 13:36:05 by ross] sort lists of instances by - arity of the type constructors (so higher-kinded instances come first) - name of the class - argument types - - - - - 26bfb19c by Simon Marlow at 2005-02-23T15:57:12+00:00 [haddock @ 2005-02-23 15:57:12 by simonmar] Fix documentation regarding the module attributes. - - - - - 9c3afd02 by Simon Marlow at 2005-02-28T16:18:17+00:00 [haddock @ 2005-02-28 16:18:17 by simonmar] version 0.7 - - - - - a95fd63f by Simon Marlow at 2005-02-28T16:22:08+00:00 [haddock @ 2005-02-28 16:22:08 by simonmar] Attempt to fix the layout of the package names in the contents. Having tried just about everything, the only thing I can get to work reliably is to make the package names line up on a fixed offset from the left margin. This obviously isn't ideal, so anyone else that would like to have a go at improving it is welcome. One option is to remove the +/- buttons from the contents list and go back to a plain table. The contents page now uses CSS for layout rather than tables. It seems that most browsers have different interpretations of CSS layout, so only the simplest things lead to consistent results. - - - - - 905d42f7 by Simon Marlow at 2005-03-01T17:16:42+00:00 [haddock @ 2005-03-01 17:16:40 by simonmar] Another attempt at lining up the package names on the contents page. Now, they line up with Konqueror, and almost line up with Firefox & IE (different layout in each case). - - - - - a0e1d178 by Wolfgang Thaller at 2005-03-09T08:28:39+00:00 [haddock @ 2005-03-09 08:28:39 by wolfgang] Hack haddock's lexer to accept the output from Apple's broken version of cpp (Apple's cpp leaves #pragma set_debug_pwd directives in it's output). - - - - - 9e1eb784 by Simon Marlow at 2005-04-22T14:27:15+00:00 [haddock @ 2005-04-22 14:27:15 by simonmar] Add a TODO item - - - - - 23281f78 by Ross Paterson at 2005-05-18T12:41:59+00:00 [haddock @ 2005-05-18 12:41:59 by ross] fix 3 bugs in --use-package, and document it. - - - - - 00074a68 by Sven Panne at 2005-05-21T12:35:29+00:00 [haddock @ 2005-05-21 12:35:29 by panne] Warning/versionitis police - - - - - 341fa822 by Simon Marlow at 2005-06-15T15:43:21+00:00 [haddock @ 2005-06-15 15:43:21 by simonmar] Allow "licence" as an alternate spelling of "license" - - - - - 3b953f8b by Simon Marlow at 2005-06-16T08:14:12+00:00 [haddock @ 2005-06-16 08:14:12 by simonmar] wibble - - - - - abfd9826 by Simon Marlow at 2005-06-27T14:46:40+00:00 [haddock @ 2005-06-27 14:46:40 by simonmar] name hierarchical HTML files as A-B-C.html instead of A.B.C.html. The old way confused Apache because the extensions are sometimes interpreted as having special meanings. - - - - - a01eea00 by Simon Marlow at 2005-08-04T13:59:40+00:00 [haddock @ 2005-08-04 13:59:40 by simonmar] 0.7 changes - - - - - 170ef87e by Simon Marlow at 2005-08-04T15:08:03+00:00 [haddock @ 2005-08-04 15:08:03 by simonmar] spec file from Jens Peterson - - - - - 7621fde4 by Simon Marlow at 2005-08-04T15:59:30+00:00 [haddock @ 2005-08-04 15:59:30 by simonmar] replace mingw tests with $(Windows) - - - - - a20739bb by Sven Panne at 2005-08-05T07:01:12+00:00 [haddock @ 2005-08-05 07:01:12 by panne] Reverted to previous version (but with bumped version number), the last commit broke RPM building on SuSE systems due to differently named dependencies. As a clarification: All .spec files in the repository have to work at least on SuSE, because that's the system I'm using. And as "Mr. Building Police", I reserve me the right to keep them that way... >:-) It might very well be the case that we need different .spec files for different platforms, so packagers which are unhappy with the current .spec files should contact me, stating the actual problems. - - - - - 4afb15cf by Simon Marlow at 2005-10-05T10:51:45+00:00 [haddock @ 2005-10-05 10:51:45 by simonmar] Add a bug - - - - - 60f69f82 by Simon Marlow at 2005-10-05T12:52:03+00:00 [haddock @ 2005-10-05 12:52:03 by simonmar] Document new behaviour of -s option - - - - - f7e520ca by Simon Marlow at 2005-10-10T15:02:55+00:00 [haddock @ 2005-10-10 15:02:55 by simonmar] extractRecSel: ignore non-record constructors (fixes a crash when using datatypes with a mixture of record and non-record style constructors). - - - - - b2edbedb by Simon Marlow at 2005-10-14T09:44:21+00:00 Start CHANGES for 0.8 - - - - - 21c7ac8d by Simon Marlow at 2005-10-14T23:11:19+00:00 First cut of Cabal build system - - - - - 766cecdd by Simon Marlow at 2005-10-29T08:14:43+00:00 Add configure script and Makefile for the docs Add a separate configure script and build system for building the documentation. The configure and Makefile code is stolen from fptools. This is left as a separate build system so that the main Cabal setup doesn't require a Unix build environment or DocBook XML tools. - - - - - aa36c783 by Duncan Coutts at 2006-01-17T19:29:55+00:00 Add a --wiki=URL flag to add a per-module link to a correspondng wiki page. So each html page gets an extra link (placed next to the source code and contents links) to a corresponding wiki page. The idea is to let readers contribute their own notes, examples etc to the documentation. Also slightly tidy up the code for the --source option. - - - - - e06e2da2 by Simon Marlow at 2006-01-18T09:28:15+00:00 TODO: documnet --wiki - - - - - 17adfda9 by Duncan Coutts at 2006-01-19T20:17:59+00:00 Add an optional wiki link for each top level exported name. In each module, for each "top level" exported entity we add a hyper link to a corresponding wiki page. The link url gets the name of the exported entity as a '#'-style anchor, so if there is an anchor in the page with that name then the users browser should jump directly to it. By "top level" we mean functions, classes, class members and data types (data, type, newtype), but not data constructors, class instances or data type class membership. The link is added at the right of the page and in a small font. Hopefully this is the right balance of visibility/distraction. We also include a link to the wiki base url in the contents and index pages. - - - - - f52324bb by Duncan Coutts at 2006-01-19T20:28:27+00:00 Rewrite pathJoin to only add a path separator when necessary. When the path ends in a file seperator there is no need to add another. Now using "--wiki=http://blah.com/foo/" should do the right thing. (Code snippet adapted from Isaac's FilePath package.) - - - - - 43bb89fa by Duncan Coutts at 2006-01-21T17:15:27+00:00 Teach haddock about line pragmas and add accurate source code links Teach haddock about C and Haskell style line pragmas. Extend the lexer/parser's source location tracking to include the file name as well as line/column. This way each AST item that is tagged with a SrcLoc gets the original file name too. Use this original file name to add source links to each exported item, in the same visual style as the wiki links. Note that the per-export source links are to the defining module rather than whichever module haddock pretends it is exported from. This is what we want for source code links. The source code link URL can also contain the name of the export so one could implement jumping to the actual location of the function in the file if it were linked to an html version of the source rather than just plain text. The name can be selected with the %N wild card. So for linking to the raw source code one might use: --source=http://darcs/haskell.org/foo/%F Or for linking to html syntax highlighted code: --source=http://darcs/haskell.org/foo/%M.html#%N - - - - - edd9f229 by Duncan Coutts at 2006-01-22T00:02:00+00:00 Extend URL variable expansion syntax and add source links to the contents page Like the wiki link on the contents and index page, add a source code link too. Extend the wiki & source URL variable expansion syntax. The original syntax was: %F for the source file name (the .hs version only, not the .lhs or .hs.pp one) %M for the module name (with '.' replaced by '/') The new syntax is: %F or %{FILE} for the original source file name %M or %{MODULE} for the module name (no replacements) %N or %{NAME} for the function/type export name %K or %{KIND} for a type/value flag "t" or "v" with these extensions: %{MODULE/./c} to replace the '.' module seperator with any other char c %{VAR|some text with the % char in it} which means if the VAR is not in use in this URL context then "" else replace the given text with the '%' char replaced by the string value of the VAR. This extension allows us to construct URLs wit optional parts, since the module/file name is not available for the URL in the contents/index pages and the value/type name is not available for the URL at the top level of each module. - - - - - eb3c6ada by Duncan Coutts at 2006-01-23T13:42:34+00:00 Remove the complex substitutions and add more command line flags instead. Instead of incomprehensable URL substitutions like ${MODULE/./-|?m=%} we now use three seperate command line flags for the top level, per-module and per-entity source and wiki links. They are: --source-base, --source-module, --source-entity --comments-base, --comments-module, --comments-entity We leave -s, --source as an alias for --source-module which is how that option behaved previously. The long forms of the substitutions are still available, ${FILE} ${MODULE} etc and the only non-trivial substitution is ${MODULE/./c} to replace the '.' characters in the module name with any other character c. eg ${MODULE/./-} Seperating the source and wiki url flags has the added bonus that they can be turned on or off individually. So users can have per-module links for example without having to also have per-entity links.` - - - - - a2f0f2af by Duncan Coutts at 2006-01-23T13:54:52+00:00 Make the --help output fit in 80 columns. This is a purely cosmetic patch, feel free to ignore it. The only trickery going on is that we don't display the deprecated -s, --source flags in the help message, but we do still accept them. - - - - - 2d3a4b0c by Duncan Coutts at 2006-01-23T14:12:16+00:00 Add documentation for the new --source-* and --comments-* command line options - - - - - 1a82a297 by Simon Marlow at 2006-01-23T17:03:27+00:00 fix markup - - - - - 100d464a by Duncan Coutts at 2006-01-23T18:31:13+00:00 remove a couple TODO items that have been done The --wiki, or rather the --comment-* options are now documented. There is probably no need to have haddock invoke unlit or cpp itself since it can now pick up the line pragmas to get the source locations right. Tools like Cabal will arrange for preprocessors to be run so there is less of a need for tools like haddock to do it themselves. - - - - - 3162fa91 by Simon Marlow at 2006-01-24T14:21:56+00:00 add a test I had lying around - - - - - 98947063 by Simon Marlow at 2006-01-31T13:52:54+00:00 add scabal-version field - - - - - c41876e6 by Neil Mitchell at 2006-02-26T17:48:21+00:00 Add Hoogle output option - - - - - f86fb9c0 by Simon Marlow at 2006-03-08T09:15:20+00:00 add haskell.vim Contributed by Brad Bowman <bsb at bereft.net>, thanks! - - - - - 35d3c511 by benjamin.franksen at 2006-03-03T22:39:54+00:00 fixed libdir (/html was missing) - - - - - 4d08fd7d by Simon Marlow at 2006-03-10T11:13:31+00:00 add PatternGuards extension - - - - - 3f095e70 by Simon Marlow at 2006-03-13T11:40:42+00:00 bug fixes from Brad Bowman - - - - - 8610849d by Sven Panne at 2006-03-19T17:02:56+00:00 Fixed Cabal/RPM build - - - - - 34a994d6 by sven.panne at 2006-04-20T12:39:23+00:00 Avoid pattern guards Due to the use of pattern guards in Haddock, GHC was called with -fglasgow-exts. This in turn enables bang patterns, too, which broke the Haddock build. Removing some unnecessary pattern guards seemed to be the better way of fixing this instead of using a pragma to disable pattern guards. - - - - - bb523f51 by Ross Paterson at 2006-04-24T09:03:25+00:00 extend 'deriving' heuristic a little If an argument of a data constructor has a type variable head, it is irreducible and the same type class can be copied into the constraint. (Formerly we just did this for type variable arguments.) - - - - - dab9fe7a by Simon Marlow at 2006-04-26T10:02:31+00:00 record an idea - - - - - 748b7078 by Simon Marlow at 2006-05-08T08:28:53+00:00 add section about deriving - - - - - 11252ea1 by Simon Marlow at 2006-05-24T15:43:10+00:00 replace a fatal error in lexChar with a parseError - - - - - 382c9411 by Simon Marlow at 2006-05-24T15:45:47+00:00 add a bug - - - - - b79272f5 by Simon Marlow at 2006-05-24T15:46:29+00:00 add a bug report - - - - - 912edf65 by David Waern at 2006-07-10T19:09:23+00:00 Initial modifications -- doesn't compile - - - - - a3c7ba99 by David Waern at 2006-07-11T00:54:19+00:00 More porting work -- doesn't compile - - - - - 0a173d19 by David Waern at 2006-07-11T11:30:03+00:00 Make the repos temporarily compile and illustrate a problem - - - - - bad316de by David Waern at 2006-07-11T15:43:47+00:00 Progress on the porting process - - - - - bbf12d02 by David Waern at 2006-07-11T23:07:44+00:00 More progress on the porting -- first pass starting to shape up - - - - - de580ba2 by David Waern at 2006-07-20T17:48:30+00:00 More progress -- still on phase1 - - - - - 75a917a2 by David Waern at 2006-07-23T18:22:43+00:00 More work on pass1 -- mostly done - - - - - 6697b3f7 by David Waern at 2006-07-23T22:17:40+00:00 More work, started working on the renaming phase -- this code will need a cleanup soon :) - - - - - 82a5bcbb by David Waern at 2006-07-29T16:16:43+00:00 Add instances, build renaming environment, start on the renamer - - - - - c3f8f4f1 by David Waern at 2006-07-29T21:37:48+00:00 Complete the renamer - - - - - 7e00d464 by David Waern at 2006-07-30T21:01:57+00:00 Start porting the Html renderer - - - - - f04ce121 by David Waern at 2006-08-09T20:04:56+00:00 More Html rendering progress - - - - - 20c21b53 by David Waern at 2006-08-10T17:37:47+00:00 More progress - - - - - d7097e0d by David Waern at 2006-08-11T20:31:51+00:00 Cleanup - - - - - a7351e86 by David Waern at 2006-08-12T11:44:47+00:00 Render H98 Data declarations - - - - - 3fb2208e by David Waern at 2006-08-12T17:15:34+00:00 Perfect rendering of Test.hs - - - - - 454fd062 by David Waern at 2006-08-13T21:57:08+00:00 Misc fixes and interface load/save - - - - - 7ef7e7be by David Waern at 2006-08-14T00:56:07+00:00 Some refactoring - - - - - a7d3efef by David Waern at 2006-08-19T20:07:55+00:00 Adapt to latest GHC - - - - - 5fc3c0d7 by David Waern at 2006-08-20T21:28:11+00:00 Move interface read/write to its own module + some cleanup - - - - - 037e011c by David Waern at 2006-08-20T21:38:24+00:00 Small cleanup - - - - - da3a1023 by David Waern at 2006-09-03T16:05:22+00:00 Change mode to BatchCompile to avoid GHC API bug - - - - - 3cc9be3b by David Waern at 2006-09-03T16:06:59+00:00 Starting work on GADT rendering - - - - - 94506037 by David Waern at 2006-09-03T20:02:48+00:00 Compensate for change of export list order in GHC - - - - - c2cec4eb by David Waern at 2006-09-04T20:53:01+00:00 Rename a function - - - - - 9a9735ba by David Waern at 2006-09-05T15:51:21+00:00 Change version number to 2.0 - - - - - 3758a714 by David Waern at 2006-09-05T15:51:49+00:00 Align comment properly - - - - - 68478d9e by David Waern at 2006-09-15T18:03:00+00:00 Remove interface reading/writing code and use the GHC api for creating package environments instead - - - - - d2eedd95 by David Waern at 2006-09-15T18:05:29+00:00 Change the executable name to haddock-ghc-nolib - - - - - fcfbcf66 by David Waern at 2006-09-15T18:05:45+00:00 Small source code cleanup - - - - - d08eb017 by David Waern at 2006-09-15T18:06:21+00:00 Remove handling of --package flag - - - - - b8a4cf53 by David Waern at 2006-09-15T18:07:16+00:00 Remove commented-out code - - - - - bef0a684 by David Waern at 2006-09-15T18:37:57+00:00 Don't warn about missing links to () - - - - - e7d25fd7 by David Waern at 2006-09-15T19:50:49+00:00 Remove Interface and Binary2 modules - - - - - 9894f2a1 by David Waern at 2006-09-15T19:53:43+00:00 Remove debug printing from HaddockHtml - - - - - a0e7455d by David Waern at 2006-09-16T00:16:29+00:00 Comments only - - - - - d5b26fa7 by David Waern at 2006-09-16T00:16:57+00:00 Refactor PackageData creation code and start on building the doc env propery (unfinished) - - - - - 06aaa779 by David Waern at 2006-09-16T00:19:25+00:00 Better comments in Main.hs - - - - - 1a52d1b4 by David Waern at 2006-09-18T22:17:11+00:00 Comments and spacing change - - - - - e5a97767 by David Waern at 2006-09-21T17:02:45+00:00 Remove unnecessary fmapM import in Main - - - - - 9d0f9d3a by David Waern at 2006-09-22T18:07:07+00:00 Make import list in HaddockHtml prettier - - - - - 3452f662 by David Waern at 2006-09-22T18:08:47+00:00 Refactor context rendering - - - - - 12d0a6d0 by David Waern at 2006-09-22T18:09:52+00:00 Do proper HsType rendering (inser parentheses correctly) - - - - - 2c20c2f9 by David Waern at 2006-09-22T18:10:45+00:00 Fix a bug in Main.toHsType - - - - - c5396443 by David Waern at 2006-09-22T18:11:16+00:00 Skip external package modules sort for now - - - - - 3fb95547 by David Waern at 2006-09-22T20:35:40+00:00 Take away trailin "2" on all previously clashing type names - - - - - 2174755f by David Waern at 2006-09-22T20:51:43+00:00 Remove unused imports in Main - - - - - 1e9f7a39 by David Waern at 2006-09-22T20:52:11+00:00 Fix a comment in Main - - - - - 32d9e028 by David Waern at 2006-10-05T16:40:11+00:00 Merge with changes to ghc HEAD - - - - - 3058c8f5 by David Waern at 2006-10-05T16:41:02+00:00 Comment fixes - - - - - b9c217ec by David Waern at 2006-10-05T16:49:59+00:00 Filter out more builtin type constructors from warning messages - - - - - 67e7d252 by David Waern at 2006-10-05T19:38:22+00:00 Refactoring -- better structured pass1 - - - - - cd21c0c1 by David Waern at 2006-10-05T19:44:42+00:00 Remove read/dump interface flags - - - - - 313f9e69 by David Waern at 2006-10-05T19:49:26+00:00 Remove unused pretty printing - - - - - 480f09d1 by David Waern at 2006-12-28T13:22:24+00:00 Update to build with latest GHC HEAD - - - - - 63dccfcb by David Waern at 2007-01-05T01:38:45+00:00 Fixed a bug so that --ghc-flag works correctly - - - - - 3117dadc by David Waern at 2006-12-29T18:53:39+00:00 Automatically get the GHC lib dir - - - - - 9dc84a5c by David Waern at 2006-12-29T19:58:53+00:00 Comments - - - - - 0b0237cc by David Waern at 2007-01-05T16:48:30+00:00 Collect docs based on SrcLoc, syncing with removal of DeclEntity from GHC - - - - - a962c256 by David Waern at 2007-01-05T17:02:47+00:00 Add tabs in haddock.cabal - - - - - 0ca30c97 by David Waern at 2007-01-05T17:04:11+00:00 Add GHCUtils.hs - - - - - c0ab9abe by David Waern at 2007-01-10T11:43:08+00:00 Change package name to haddock-ghc, version 0.1 - - - - - 38e18b27 by David Waern at 2007-01-12T12:03:52+00:00 No binder name for foreign exports - - - - - d18587ab by David Waern at 2007-01-12T12:08:15+00:00 Temp record - - - - - ba6251a0 by David Waern at 2007-01-12T18:27:55+00:00 Remove read/dump-interface (again) - - - - - f4ba2b39 by David Waern at 2007-01-12T18:31:36+00:00 Remove DocOption, use the GHC type - - - - - 511be8bd by David Waern at 2007-01-12T18:32:41+00:00 Use exceptions instead of Either when loading package info - - - - - 0f2144d8 by David Waern at 2007-01-12T18:33:23+00:00 Small type change - - - - - 77507eb7 by David Waern at 2007-01-12T18:33:59+00:00 Remove interface file read/write - - - - - 0ea1e14f by David Waern at 2007-01-17T21:40:26+00:00 Add trace_ppr to GHCUtils - - - - - 3878b493 by David Waern at 2007-01-17T21:40:53+00:00 Sort external package modules and build a doc env - - - - - 8dc323fc by David Waern at 2007-01-17T21:42:41+00:00 Remove comment - - - - - f4c5b097 by David Waern at 2007-01-18T23:22:18+00:00 Add haddock-ghc.cabal and remove ghc option pragma in source file - - - - - da242b2c by David Waern at 2007-01-18T23:22:46+00:00 Remove some tabs - - - - - 288ed096 by David Waern at 2007-01-18T23:39:28+00:00 Moved the defaultErrorHandler to scope only over sortAndCheckModules for now - - - - - 4dd150fe by David Waern at 2007-02-03T21:23:56+00:00 Let restrictCons handle infix constructors - - - - - 97893442 by David Waern at 2007-02-04T16:26:00+00:00 Render infix data constructors - - - - - da89db72 by David Waern at 2007-02-04T16:26:33+00:00 CHange project name to Haddock-GHC - - - - - e93d48af by David Waern at 2007-02-04T16:59:08+00:00 Render infix type constructors properly - - - - - 357bc99b by David Waern at 2007-02-04T17:37:08+00:00 Insert spaces around infix function names - - - - - ab6cfc49 by David Waern at 2007-02-04T17:59:54+00:00 Do not list entities without documentation - - - - - 04249c7e by David Waern at 2007-02-04T19:16:25+00:00 Add GADT support (quite untested) - - - - - 2c223f8d by David Waern at 2007-02-04T19:25:10+00:00 Add package file write/save again! - - - - - b07ed218 by David Waern at 2007-02-04T19:33:02+00:00 Comment out minf_iface based stuff - - - - - 953d1fa7 by David Waern at 2007-02-05T00:12:23+00:00 Solve conflicts - - - - - 593247fc by David Waern at 2007-02-06T19:48:48+00:00 Remove -package flag, GHC's can be used instead - - - - - f658ded2 by David Waern at 2007-02-06T20:50:44+00:00 Start for support of ATs - - - - - 97f9e913 by David Waern at 2007-02-06T20:52:27+00:00 Wibble - - - - - 2ce8e4cf by David Waern at 2007-02-16T12:09:49+00:00 Add the DocOptions change - - - - - dee4a9b5 by David Waern at 2007-03-06T01:24:48+00:00 Wibble - - - - - 7cb99d18 by David Waern at 2007-03-06T01:24:58+00:00 Change version to 2.0 and executable name to haddock - - - - - c5aa02bc by David Waern at 2007-03-08T15:59:49+00:00 Go back to -B flag - - - - - 3a349201 by David Waern at 2007-03-09T13:31:59+00:00 Better exception handling and parsing of GHC flags - - - - - 05a69b71 by David Waern at 2007-03-09T17:45:44+00:00 Remove commented-out DocEntity printing - - - - - 755032cb by davve at dtek.chalmers.se at 2007-03-23T23:30:20+00:00 Remove a file that shouldn't be here - - - - - a7077e5f by davve at dtek.chalmers.se at 2007-03-24T03:58:48+00:00 Remove an import - - - - - 6f55aa8b by davve at dtek.chalmers.se at 2007-03-25T00:46:48+00:00 Start work on Haddock API - - - - - f0199480 by davve at dtek.chalmers.se at 2007-03-25T00:56:36+00:00 Prettify some comments - - - - - f952f9d1 by davve at dtek.chalmers.se at 2007-03-25T00:56:53+00:00 Remove ppr in HaddockTypes - - - - - bc594904 by davve at dtek.chalmers.se at 2007-03-25T00:57:53+00:00 Remove commented out doc env inference - - - - - 11ebf08d by davve at dtek.chalmers.se at 2007-03-25T01:23:25+00:00 De-flatten the namespace - - - - - f696b4bc by davve at dtek.chalmers.se at 2007-03-25T03:21:48+00:00 Add missing stuff to API - - - - - 9a2a04c3 by davve at dtek.chalmers.se at 2007-03-25T03:22:02+00:00 Wibble - - - - - 7d04a6d5 by davve at dtek.chalmers.se at 2007-03-25T03:22:08+00:00 Avoid a GHC bug with parseStaticFlags [] - - - - - 4d2820ba by davve at dtek.chalmers.se at 2007-03-26T04:57:01+00:00 Add fall-through case to mkExportItem - - - - - 6ebc8950 by Stefan O'Rear at 2007-03-26T04:14:53+00:00 Add shebang line to Setup.lhs - - - - - 80966ec5 by davve at dtek.chalmers.se at 2007-03-26T05:24:26+00:00 Fix stupid compile error - - - - - 1ea1385d by davve at dtek.chalmers.se at 2007-04-05T17:19:56+00:00 Do save/read of interface files properly - - - - - 0e4f6541 by David Waern at 2007-04-10T21:08:36+00:00 Add version to ghc dependency - - - - - b0499b63 by David Waern at 2007-04-10T21:37:08+00:00 Change package name to haddock - - - - - 9d50d27e by David Waern at 2007-04-24T00:22:14+00:00 Use filepath package instead of FilePath - - - - - 87c7fcdf by David Waern at 2007-07-10T21:03:04+00:00 Add new package dependencies - - - - - 4768709c by David Waern at 2007-07-11T20:37:11+00:00 Follow changes to record constructor representation - - - - - b9a02fee by Simon Marlow at 2007-05-30T14:00:48+00:00 update to compile with the latest GHC & Cabal - - - - - c0ebdc01 by David Waern at 2007-07-11T21:35:45+00:00 Fix conflicts - - - - - 97f7afd4 by David Waern at 2007-07-11T21:52:38+00:00 Follow changes to the GHC API - - - - - a5b7b58f by David Waern at 2007-07-12T20:36:48+00:00 Call parseStaticFlags before newSession - - - - - f7f50dbc by David Waern at 2007-08-01T21:52:58+00:00 Better indentation in haddock.cabal - - - - - d84e52ad by David Waern at 2007-08-02T00:08:18+00:00 Wibble - - - - - a23f494a by David Waern at 2007-08-02T00:08:24+00:00 Be better at trying to load all module dependencies (debugging) - - - - - ee917f13 by David Waern at 2007-08-03T18:48:08+00:00 Load all targets explicitly (checkModule doesn't chase dependencies anymore) - - - - - 5182d631 by David Waern at 2007-08-16T16:48:55+00:00 Finalize support for links to other packages - - - - - dfd1e3da by David Waern at 2007-08-16T16:51:11+00:00 Fix haddock comment errors in Haddock.Types - - - - - 50c0d83e by David Waern at 2007-08-16T16:51:37+00:00 Remove a debug import - - - - - d84b7c2b by David Waern at 2007-08-16T17:06:30+00:00 Rename PackageData to HaddockPackage - - - - - 3b52cb9f by David Waern at 2007-08-16T17:09:42+00:00 Simplify some comments - - - - - 66fa68d9 by David Waern at 2007-08-16T17:11:38+00:00 Comment the HaddockPackage definition - - - - - 8674c761 by David Waern at 2007-08-16T17:25:54+00:00 Improve code layout in Main - - - - - 571a3a0b by David Waern at 2007-08-16T17:32:13+00:00 Remove explict module imports in Main - - - - - d31b3cb0 by David Waern at 2007-08-16T17:36:23+00:00 Correct comments - - - - - 7f8a9f2b by David Waern at 2007-08-16T17:39:50+00:00 Fix layout problems in Haddock.Types - - - - - 9f421d7f by David Waern at 2007-08-17T11:16:48+00:00 Move options out of Main into Haddock.Options - - - - - 80042b63 by David Waern at 2007-08-17T11:26:59+00:00 Small comment/layout fixes - - - - - b141b982 by David Waern at 2007-08-17T11:28:28+00:00 Change project name from Haddock-GHC to Haddock - - - - - dbeb4a81 by David Waern at 2007-08-17T11:41:05+00:00 Add top module comment to all files - - - - - ce99cc9e by David Waern at 2007-08-17T14:53:04+00:00 Factor out typechecking phase into Haddock.Typecheck - - - - - 6bf75d9e by David Waern at 2007-08-17T16:55:35+00:00 Factor out package code to Haddock.Packages - - - - - b396db37 by David Waern at 2007-08-29T22:40:23+00:00 Major refactoring - - - - - 3d4f95ee by David Waern at 2007-08-29T23:26:24+00:00 Rename HaddockModule to Interface and a few more refactorings - - - - - c55326db by David Waern at 2007-08-29T23:48:03+00:00 Some comment cleanup - - - - - 9a84fc46 by David Waern at 2007-08-29T23:49:29+00:00 Add some modules that I forgot to add earlier - - - - - 4536dce2 by David Waern at 2007-08-29T23:55:24+00:00 Wibble - - - - - 9b7f0206 by David Waern at 2007-08-30T16:03:29+00:00 Wibble - - - - - c52c050a by David Waern at 2007-08-30T16:30:37+00:00 Rename HaddockModule to Interface - - - - - eae2995f by David Waern at 2007-08-30T16:42:59+00:00 Simplify createInterfaces - - - - - 53f99caa by David Waern at 2007-09-29T00:04:31+00:00 Add build-type: Simple to the cabal file - - - - - 0d3103a8 by David Waern at 2007-09-29T00:04:58+00:00 Add containers and array dependency - - - - - 6acf5f30 by David Waern at 2007-09-29T00:13:36+00:00 Prettify the cabal file - - - - - 87c1e378 by David Waern at 2007-09-29T13:16:39+00:00 FIX: consym data headers with more than two variables - - - - - b67fc16a by David Waern at 2007-09-29T14:01:32+00:00 FIX: prefix types used as operators should be quoted - - - - - a8f925bc by David Waern at 2007-09-29T14:02:26+00:00 Use isSymOcc from OccName instead of isConSym - - - - - fc330701 by David Waern at 2007-09-29T14:15:37+00:00 Use isLexConSym/isLexVarSym from OccName - - - - - e4f3dbad by David Waern at 2007-09-29T15:01:08+00:00 FIX: do not quote varsym type operators - - - - - 402207d2 by David Waern at 2007-09-29T15:01:50+00:00 Wibble - - - - - f9d89ef0 by David Waern at 2007-09-29T15:17:40+00:00 Take care when pp tyvars - add parens on syms - - - - - 849e2a77 by David Waern at 2007-10-01T21:56:39+00:00 Go back to using a ModuleMap instead of LookupMod - fixes a bug - - - - - 549dbac6 by David Waern at 2007-10-02T01:05:19+00:00 Improve parsing of doc options - - - - - a36021b8 by David Waern at 2007-10-02T23:05:00+00:00 FIX: double arrows in constructor contexts - - - - - d03bf347 by David Waern at 2007-10-09T16:14:05+00:00 Add a simple test suite - - - - - c252c140 by David Waern at 2007-10-17T16:02:28+00:00 Add --optghc=.. style flag passing to GHC - - - - - cce6c1b3 by David Waern at 2007-10-18T22:03:20+00:00 Add support for --read-interface again - - - - - 33d059c0 by David Waern at 2007-10-18T22:30:18+00:00 Refactoring -- get rid of Haddock.Packages - - - - - f9ed0a4c by David Waern at 2007-10-18T22:34:36+00:00 Name changes - - - - - 8a1c816f by David Waern at 2007-10-20T14:24:23+00:00 Add --ghc-version option - - - - - 4925aaa1 by David Waern at 2007-10-21T14:34:26+00:00 Add some Outputable utils - - - - - 69e7e47f by David Waern at 2007-10-21T14:35:49+00:00 FIX: Ord for OrdName was not comparing modules - - - - - 5a4ae535 by David Waern at 2007-10-21T21:18:48+00:00 Wibble - - - - - 03d48e20 by David Waern at 2007-10-24T15:52:56+00:00 Remove Main from "other modules" - - - - - c66f6d82 by David Waern at 2007-10-24T16:37:18+00:00 Make it possible to run haddock on itself - - - - - 21d156d8 by David Waern at 2007-10-25T14:02:14+00:00 Don't set boot modules as targets - - - - - f8bcf91c by David Waern at 2007-10-31T22:11:17+00:00 Add optimisation flags - - - - - 7ac758f2 by David Waern at 2007-11-04T09:48:28+00:00 Go back to loading only targets (seems to work now) - - - - - 4862aae1 by David Waern at 2007-11-05T22:24:57+00:00 Do full compilation of modules -- temporary fix for GHC API problem - - - - - 697e1517 by David Waern at 2007-11-05T22:25:50+00:00 Don't warn about not being able to link to wired/system/builtin-names - - - - - 892186da by David Waern at 2007-11-06T00:49:21+00:00 Filter out instances with TyCons that are not exported - - - - - 9548314c by David Waern at 2007-11-06T09:37:14+00:00 Wibble - - - - - 5cafd627 by David Waern at 2007-11-08T01:43:07+00:00 Filter out all non-vanilla type sigs - - - - - 04621830 by David Waern at 2007-11-08T01:45:13+00:00 Synch loading of names from .haddock files with GHC's name cache - - - - - 88d37f77 by David Waern at 2007-11-08T01:46:21+00:00 Remove commented-out code - - - - - 6409c911 by David Waern at 2007-11-08T01:56:00+00:00 Small bugfix and cleanup in getDeclFromTyCls - - - - - af59d9c2 by David Waern at 2007-11-08T02:08:44+00:00 Remove OrdName stuff - - - - - 3a615e2e by David Waern at 2007-11-08T02:13:41+00:00 Update runtests.hs following changes to haddock - - - - - 01f3314e by David Waern at 2007-11-08T02:33:01+00:00 Complain if we can't link to wired-in names - - - - - fcafb5d1 by David Waern at 2007-11-09T02:40:16+00:00 Don't exit when there are no file arguments - - - - - 194bc332 by David Waern at 2007-11-09T02:55:37+00:00 Wibble - - - - - dbe4cb55 by David Waern at 2007-11-09T02:56:14+00:00 Wibble - - - - - 82869fda by David Waern at 2007-11-10T17:01:43+00:00 Introduce InstalledInterface structure and add more stuff to the .haddock files We introduce InstalledInterface capturing the part of Interface that is stored in the interface files. We change the ppHtmlContents and ppHtmllIndex to take this structure instead of a partial Interface. We add stuff like the doc map and exported names to the .haddock file (via InstalledInterface). - - - - - d6bb57bf by David Waern at 2007-11-10T17:19:48+00:00 FIX: contents and index should include external package modules when --gen-contents/--gen-index - - - - - e8814716 by David Waern at 2007-11-11T00:29:27+00:00 Remove lDocLinkName and its use in Html backend - - - - - 6f9bd702 by David Waern at 2007-11-11T00:50:57+00:00 Do some refactoring in the html backend This also merges an old patch by Augustsson: Wed Jul 12 19:54:36 CEST 2006 lennart.augustsson at credit-suisse.com * Print type definitions like signatures if given arrows. - - - - - 09d0ce24 by Malcolm.Wallace at 2006-07-20T13:13:57+00:00 mention HsColour in the docs, next to option flags for linking to source code - - - - - 24da6c34 by Malcolm.Wallace at 2006-07-20T13:14:50+00:00 change doc references to CVS to give darcs repository location instead - - - - - 74d52cd6 by David Waern at 2007-11-11T00:55:33+00:00 Update copyright - - - - - fcaa3b4f by Duncan Coutts at 2006-09-08T13:41:00+00:00 Eliminate dep on network by doing a little cut'n'paste haddock depending on the network causes a circular dependency at least if you want to build the network lib with haddock docs. - - - - - 10cc9bda by David Waern at 2007-11-11T02:09:41+00:00 Fix conflicts - - - - - 4e3acd39 by David Waern at 2007-11-11T02:21:19+00:00 Manual merge of a patch from Duncan Coutts that removes the dependency on mtl - - - - - fa9070da by Neil Mitchell at 2006-09-29T15:52:03+00:00 Do not generate an empty table if there are no exports, this fixes a <table></table> tag being generated, which is not valid HTML 4.01 - - - - - d7431c85 by David Waern at 2007-11-11T02:28:50+00:00 Fix conflicts - - - - - f87e8f98 by Simon Marlow at 2006-10-10T11:37:16+00:00 changes for 0.8 - - - - - db929565 by Simon Marlow at 2006-10-10T12:07:12+00:00 fix the name of the source file - - - - - 8220aa4b by Simon Marlow at 2006-10-11T14:17:37+00:00 Rename haddock.js to haddock-util.js haddock.js will be run automatically by Windows when you type 'haddock' if it is found on the PATH, so rename to avoid confusion. Spotted by Adrian Hey. - - - - - 6bccdaa1 by sven.panne at 2006-10-12T15:28:23+00:00 Cabal's sdist does not generate "-src.tar.gz" files, but ".tar.gz" ones - - - - - d3f3fc19 by Simon Marlow at 2006-12-06T16:05:07+00:00 add todo item for --maintainer - - - - - 2da7e269 by Simon Marlow at 2006-12-15T15:52:00+00:00 TODO: do something better about re-exported symbols from another package - - - - - 42d85549 by David Waern at 2007-11-11T02:30:59+00:00 Fix conflicts - - - - - 5e7ef6e5 by Neil Mitchell at 2007-01-11T15:41:15+00:00 Never do spliting index files into many - - - - - f3d4aebe by Neil Mitchell at 2007-01-11T17:07:09+00:00 Add searching on the index page - - - - - bad3ab66 by Neil Mitchell at 2007-01-11T18:17:46+00:00 Delete dead code, now there is only one index page - - - - - cd09eedb by Neil Mitchell at 2007-01-11T18:21:19+00:00 Delete more stuff that is no longer required - - - - - e2806646 by David Waern at 2007-11-11T02:41:53+00:00 Fix conflicts - - - - - a872a823 by Neil Mitchell at 2007-01-11T18:51:43+00:00 Make the index be in case-insensitive alphabetic order - - - - - 8bddd9d7 by Neil Mitchell at 2007-02-06T17:49:12+00:00 Do not create empty tables for data declarations which don't have any constructors, instances or comments. Gets better HTML 4.01 compliance - - - - - 036b8120 by David Waern at 2007-11-11T02:56:58+00:00 Fix conflicts - - - - - f50c1639 by Conal Elliott at 2007-02-14T21:54:00+00:00 added substitution %{FILE///c} - - - - - 402e166a by David Waern at 2007-11-11T03:35:46+00:00 Manual merge of old patch: Sat Apr 21 04:36:43 CEST 2007 Roberto Zunino <zunrob at users.sf.net> * URL expansion for %%, %L, %{LINE} - - - - - 2f264fbd by David Waern at 2007-11-11T03:40:33+00:00 Manual merge of an old patch: Thu Apr 19 20:23:40 CEST 2007 Wolfgang Jeltsch <g9ks157k at acme.softbase.org> * bug fix When Haddock was invoked with the --ignore-all-exports flag but the ignore-exports module attribute wasn't used, hyperlinks weren't created for non-exported names. This fix might not be as clean as one would wish (since --ignore-all-exports now results in ignore_all_exports = True *and* an additional OptIgnoreExports option for every module) but at least the bug seems to be resolved now. - - - - - 7d7ae106 by sven.panne at 2007-09-02T12:18:02+00:00 Install LICENSE in the correct place - - - - - 66eaa924 by David Waern at 2007-11-11T19:02:46+00:00 Fix a bug that made haddock loop - - - - - 4ed47b58 by David Waern at 2007-11-11T19:03:09+00:00 Rename java-script file (this wasn't merge correctly) - - - - - d569534a by David Waern at 2007-11-11T19:06:44+00:00 Don't require -B <ghc-libdir> when no argument files Change readInterfaceFile to take a Maybe Session, to avoid having to pass -B <ghc-libdir> to Haddock when there're no source files to process. This is nice when computing contents/index for external packages. - - - - - 373368bc by Neil Mitchell at 2007-01-11T18:22:44+00:00 Change from tabs to spaces in the ppHtmlIndex function - - - - - 6b063a77 by Neil Mitchell at 2007-01-12T12:17:46+00:00 Rewrite much of the index searching code, previously was too slow to execute on the base library with IE, the new version guarantees less than O(log n) operations be performed, where n is the number in the list (before was always O(n)) - - - - - bfad00b7 by David Waern at 2007-11-11T23:33:53+00:00 Fix conflicts - - - - - cd2dcc09 by Neil Mitchell at 2007-01-12T12:25:01+00:00 Make the max number of results 75 instead of 50, to allow map searching in the base library to work - - - - - 3ae74764 by Neil Mitchell at 2007-01-12T12:58:17+00:00 Make the search box in a form so that enter does the default search - - - - - 142103e5 by David Waern at 2007-11-12T00:03:18+00:00 Merge patch from the old branch: Fri Aug 31 13:21:45 CEST 2007 Duncan Coutts <duncan at haskell.org> * Add category: Development to .cabal file Otherwise it appears on the hackage website in the "Unclassified" category. - - - - - 22ec2ddb by David Waern at 2007-11-25T01:55:29+00:00 A a list of small improvements to the TODO file - - - - - eb0129f4 by Wolfgang Jeltsch at 2007-12-03T23:47:55+00:00 addition of type equality support (at least for HTML generation) - - - - - 816a7e22 by David Waern at 2007-12-08T15:46:26+00:00 Handle class operators correctly when rendering predicates - - - - - 68baaad2 by David Waern at 2007-12-08T16:15:54+00:00 Code layout changes - - - - - 09b77fb4 by David Waern at 2007-12-08T16:16:03+00:00 Handle infix operators correctly in the Type -> HsType translation - - - - - 31c36da2 by David Waern at 2007-12-08T16:24:27+00:00 Add ppLParendTypes/ppLParendType - - - - - b17cc818 by David Waern at 2007-12-08T16:26:12+00:00 Use ppParendType when printing types args in predicates - - - - - ffd1f2cf by David Waern at 2007-12-08T16:45:06+00:00 Fix rendering of instance heads to handle infix operators This is also a refactoring to share this code for rendering predicates. - - - - - ff886d45 by David Waern at 2007-12-08T17:27:46+00:00 Fix rendering of class operators - - - - - e2fcbb9e by David Waern at 2007-12-08T17:59:28+00:00 Fix a bug (use ppTyName instead of ppName to print names in type apps) - - - - - 79a1056e by David Waern at 2007-12-08T21:25:18+00:00 Update tests - - - - - 867741ac by David Waern at 2007-12-08T21:25:49+00:00 Give a diff on test failure - - - - - 7e5eb274 by David Waern at 2008-01-05T14:33:45+00:00 Add DrIFT commands - - - - - 3656454d by David Waern at 2008-01-05T20:26:00+00:00 Add "cabal-version: >= 1.2" to the cabal file - - - - - 77974efc by Simon Marlow at 2007-12-20T09:52:44+00:00 add an item - - - - - f6ac1708 by Simon Marlow at 2007-12-06T14:00:10+00:00 Source links must point to the original module, not the referring module - - - - - eda1d5c9 by David Waern at 2008-01-06T14:40:52+00:00 Manual merge of a patch to the 0.8 branch Thu Dec 6 15:00:10 CET 2007 Simon Marlow <simonmar at microsoft.com> * Source links must point to the original module, not the referring module - - - - - 378f4085 by David Waern at 2008-01-06T16:03:45+00:00 Change stability from stable to experimental - - - - - 8bdafe44 by David Waern at 2008-01-06T16:14:22+00:00 Add haskell.vim (it had been removed somehow) - - - - - ea34d02e by David Waern at 2008-01-06T16:36:57+00:00 Change version to 2.0.0.0 - - - - - 34631ac0 by David Waern at 2008-01-06T16:44:57+00:00 Add missing modules to the cabal file - - - - - 9e142935 by David Waern at 2008-01-06T17:25:42+00:00 Depend on ghc >= 6.8.2 && < 6.9 - - - - - 59f9eeaa by Simon Marlow at 2007-12-20T10:43:04+00:00 add build scripts - - - - - 1c29ae30 by Simon Marlow at 2007-12-20T10:47:07+00:00 update version number - - - - - fe16a3e4 by Simon Marlow at 2007-12-20T10:48:03+00:00 update version - - - - - f688530f by Simon Marlow at 2007-12-20T10:48:29+00:00 doc updates - - - - - ce71b611 by David Waern at 2008-01-07T13:46:32+00:00 Change version in docs and spec - - - - - 03ab8d6f by David Waern at 2008-01-07T13:47:38+00:00 Manually merge over changes to CHANGES for 0.9 - - - - - 39f1b042 by David Waern at 2008-01-07T15:17:41+00:00 Remove the -use-package flag, we don't support it anyway - - - - - 7274a544 by David Waern at 2008-01-07T15:33:05+00:00 Update CHANGES for 2.0.0.0 - - - - - 96594f5d by David Waern at 2008-01-07T15:46:49+00:00 Wibble - - - - - f4c5a4c4 by David Waern at 2008-01-07T15:55:36+00:00 Change url to repo in documentation - - - - - 8a4c77f0 by David Waern at 2008-01-07T16:00:54+00:00 Update CHANGES - - - - - cb3a9288 by David Waern at 2008-01-07T16:02:55+00:00 Documentation fix - - - - - d8e45539 by David Waern at 2008-01-07T16:12:00+00:00 Update docs to say that Haddock accets .lhs files and module names - - - - - 4b5ce824 by David Waern at 2008-01-07T16:12:25+00:00 Document -B option - - - - - 47274262 by David Waern at 2008-01-07T16:23:07+00:00 Update CHANGES - - - - - 7ff314a9 by David Waern at 2008-01-07T16:23:20+00:00 Remove --use-package, --package & --no-implicit.. flags from docs - - - - - 6c3819c0 by David Waern at 2008-01-07T16:23:52+00:00 Remove --no-implicit-prelide flag - - - - - 1b14ae40 by David Waern at 2008-01-07T16:32:26+00:00 Update the "Using literate or pre-processed source" section - - - - - 0117f620 by David Waern at 2008-01-07T16:41:55+00:00 Document the --optghc flag - - - - - 087ab1cf by David Waern at 2008-01-07T16:42:10+00:00 Remove the documenation section on derived instances The problem mentioned there doesn't exist in Haddock 2.0.0.0 - - - - - 7253951e by David Waern at 2008-01-07T16:48:40+00:00 Document OPTIONS_HADDOCK - - - - - 3b6bdcf6 by David Waern at 2008-01-07T16:56:54+00:00 Wibble - - - - - 3025adf9 by David Waern at 2008-01-07T17:08:14+00:00 Wibble - - - - - 5f30f1a0 by David Waern at 2008-01-07T17:15:44+00:00 Change synopsis field to description - - - - - 1673f54b by David Waern at 2008-01-07T17:18:21+00:00 Change my email address in the cabal file - - - - - 55aa9808 by David Waern at 2008-01-07T18:18:02+00:00 Add documentation for readInterfaceFile - - - - - eaea417f by David Waern at 2008-01-07T18:21:30+00:00 Export necessary stuff from Distribution.Haddock - - - - - 7ea18759 by David Waern at 2008-01-07T18:31:49+00:00 Remove dep on Cabal - - - - - 7b79c74e by David Waern at 2008-01-07T18:33:49+00:00 Remove dep on process - - - - - ce3054e6 by David Waern at 2008-01-16T23:01:21+00:00 Add feature-requsts from Henning Thielemann to TODO - - - - - 0c08f1ec by David Waern at 2008-01-16T23:03:02+00:00 Record a bug in TODO - - - - - b04605f3 by David Waern at 2008-01-23T16:59:06+00:00 Add a bug reported by Ross to TODO - - - - - 5b17c030 by David Waern at 2008-01-23T18:05:53+00:00 A a bug report to TODO - - - - - 1c993b0d by David Waern at 2008-01-25T16:30:25+00:00 Accept test output - - - - - c22fc0d0 by David Waern at 2008-01-25T16:34:49+00:00 Accept test output - - - - - 4b795811 by David Waern at 2008-01-25T16:38:37+00:00 Change Hidden.hs (test) to use OPTIONS_HADDOCK - - - - - c124dbd9 by David Waern at 2008-01-25T16:39:23+00:00 Accept test output - - - - - ec6f6eea by David Waern at 2008-01-25T16:42:08+00:00 Add Hidden.html.ref to tests - - - - - 1dc9610c by David Waern at 2008-02-02T20:50:51+00:00 Add a comment about UNPACK bug in TODO - - - - - 2d3f7081 by David Waern at 2008-02-09T22:33:24+00:00 Change the representation of DocNames Ross Paterson reported a bug where links would point to the defining module instead of the "best" module for an identifier (e.g Int pointing to GHC.Base instead of Data.Int). This patch fixes this problem by refactoring the way renamed names are represented. Instead of representing them by: > data DocName = Link Name | NoLink Name they are now represented as such: > data DocName = Documented Name Module | Undocumented Name and the the link-env looks like this: > type LinkEnv = Map Name Module There are several reasons for this. First of all, the bug was caused by changing the module part of Names during the renaming process, without changing the Unique field. This caused names to be overwritten during the loading of .haddock files (which caches names using the NameCache of the GHC session). So we might create new Uniques during renaming to fix this (but I'm not sure that would be problem-free). Instead, we just keep the Name and add the Module where the name is best documented, since it can be useful to keep the original Name around (for e.g. source-code location info and for users of the Haddock API). Also, the names Link/NoLink don't really make sense, since wether to use links or not is entirely up to the users of DocName. In the process of following this change into H.Backends.Html I removed the assumption that binder names are Undocumented (which was just an unnecessary assumption, the OccName is the only thing needed to render these). This will probably make it possible to get rid of the renamer and replace it with a traversal from SYB or Uniplate. Since DocName has changed, InterfaceFile has changed so this patch also increments the file-format version. No backwards-compatibility is implemented. - - - - - 0f28c921 by David Waern at 2008-02-09T23:00:36+00:00 H.GHC.Utils: remove unused imports/exports - - - - - 0c44cad5 by David Waern at 2008-02-10T00:28:13+00:00 H.GHC.Utils: add some functions that were removed by mistake - - - - - e3452f49 by David Waern at 2008-02-10T00:28:48+00:00 Fix some trivial warnings in H.InterfaceFile - - - - - a6d74644 by David Waern at 2008-02-10T00:48:06+00:00 Update the version message to fit in small terminals - - - - - 76c9cd3e by David Waern at 2008-02-10T14:47:39+00:00 Remove bugs from TODO that don't apply anymore since the port - - - - - 5e10e090 by David Waern at 2008-02-10T15:22:47+00:00 Remove bugs from TODO that weren't actual bugs - - - - - fef70878 by David Waern at 2008-02-10T15:23:44+00:00 Remove yet another item from TODO that was not an actual bug - - - - - e1af47b8 by David Waern at 2008-02-11T10:25:57+00:00 Bump the version number to 2.1.0 Since the exported datatype DocName has changed, we need to bump the major version number. Let's also drop the fourth version component, it's not that useful. - - - - - e3be7825 by David Waern at 2008-04-11T14:29:04+00:00 Add a bug to TODO - - - - - cb6574be by David Waern at 2008-04-11T16:00:45+00:00 Use the in-place haddock when running tests - - - - - c6d7af0d by David Waern at 2008-04-11T16:09:16+00:00 Turn off GHC warnings when running tests - - - - - 7f61b546 by David Waern at 2008-04-11T17:24:00+00:00 Add a flag for turning off all warnings - - - - - 883b8422 by David Waern at 2008-04-12T14:02:18+00:00 Fix printing of data binders - - - - - 2a0db8fc by David Waern at 2008-04-12T18:52:46+00:00 Fix missing parenthesis in constructor args bug - - - - - 1b3ac3f9 by David Waern at 2008-04-12T18:57:23+00:00 Simplify test suite and add tests I move all tests into one single directory to simplify things, and add a test for the last bug that was fixed. - - - - - 8f178376 by David Waern at 2008-04-12T19:00:15+00:00 Add a script for copying test output to "expected" output - - - - - 193e3a03 by David Waern at 2008-04-12T19:16:37+00:00 Remove two fixed bugs from TODO - - - - - ddc9130c by David Waern at 2008-04-12T19:37:06+00:00 Update test README - - - - - 956069c0 by David Waern at 2008-05-01T12:16:14+00:00 Update version number in spec and docs - - - - - 5478621c by David Waern at 2008-05-01T12:28:12+00:00 Remove claim of backwards compatibility from docs for readInterfaceFile - - - - - 4a16dea9 by David Waern at 2008-05-01T12:33:04+00:00 Update CHANGES - - - - - 804216fb by David Waern at 2008-05-01T12:43:16+00:00 Add a synopsis - - - - - fd0c84d5 by David Waern at 2008-05-01T12:44:44+00:00 Add Haddock.DocName to the cabal file - - - - - 9f4a7439 by David Waern at 2008-05-01T12:45:53+00:00 Remove -fglasgow-exts and -fasm - - - - - aee7c145 by David Waern at 2008-05-01T12:54:01+00:00 Add LANGUAGE pragmas to source files - - - - - 9a58428b by David Waern at 2008-05-01T12:54:19+00:00 Add extensions to cabal file - - - - - 494f1bee by David Waern at 2008-05-01T13:12:09+00:00 Export DocName in the API - - - - - c938196b by David Waern at 2008-05-01T13:12:19+00:00 Add hide options to some source files - - - - - 236e86af by Neil Mitchell at 2008-06-07T20:45:10+00:00 Rewrite the --hoogle flag support - - - - - 6d910950 by Neil Mitchell at 2008-06-14T10:56:50+00:00 Simplify the newtype/data outputting in Hoogle, as haddock does it automatically - - - - - f87a95a8 by Neil Mitchell at 2008-06-14T12:10:18+00:00 Add initial structure for outputting documentation as well, but does not yet output anything - - - - - 7c3bce54 by Neil Mitchell at 2008-06-14T12:27:07+00:00 Remove <document comment> from the Hoogle output - - - - - 9504a325 by Neil Mitchell at 2008-06-16T06:33:21+00:00 Default to "main" if there is no package, otherwise will clobber hoogle's hoogle info - - - - - 4a794a79 by Neil Mitchell at 2008-06-16T06:53:29+00:00 Change packageName to packageStr, as it better reflects the information stored in it - - - - - 7abc9baf by Neil Mitchell at 2008-06-16T07:09:49+00:00 Add modulePkgInfo to Haddock.GHC.Utils, which gives back package name and version info - - - - - 8ca11514 by Neil Mitchell at 2008-06-16T07:13:48+00:00 Change Hoogle to take the package name and package version separately - - - - - a6da452d by Neil Mitchell at 2008-06-18T11:29:46+00:00 In Hoogle do not list things that are not local to this module - - - - - 974b76b7 by David Waern at 2008-06-19T18:40:13+00:00 Be more consistent with GHC API naming in H.GHC.Utils - - - - - 2facb4eb by David Waern at 2008-06-19T19:03:03+00:00 Update test output - - - - - c501de72 by David Waern at 2008-06-26T20:26:49+00:00 Use ghc-paths to get the lib dir The path can still be overridden using the -B flag. It's not longer required to pass the lib dir to the program that runs the test suite. - - - - - ac4c6836 by David Waern at 2008-06-26T20:33:08+00:00 Update CHANGES - - - - - 9d21c60a by David Waern at 2008-06-26T20:34:53+00:00 Update README - - - - - 741448f0 by David Waern at 2008-06-26T21:12:57+00:00 Improve wording in the help message - - - - - b1b42b11 by David Waern at 2008-06-30T10:16:17+00:00 Rename ForeignType - - - - - 6d6c2b34 by David Waern at 2008-06-30T10:25:09+00:00 Rename TyFamily - - - - - 8d1125ed by David Waern at 2008-06-30T10:37:21+00:00 Rename type patterns - - - - - 7610a4cb by David Waern at 2008-06-30T10:45:07+00:00 Rename associated types - - - - - 8eeba14c by David Waern at 2008-06-30T10:47:41+00:00 Remove the TODO file now that we have a trac - - - - - 1af5b25b by David Waern at 2008-07-02T18:19:28+00:00 Render type family declarations (untested) - - - - - ceb99797 by David Waern at 2008-07-02T18:24:06+00:00 Remove redundant check for summary when rendering data types - - - - - b36a58e0 by David Waern at 2008-07-02T22:01:38+00:00 More support for type families and associated types Now we just need to render the instances - - - - - 78784879 by David Waern at 2008-07-07T22:13:58+00:00 Remove filtering of instances We were filtering out all instances for types with unknown names. This was probably an attempt to filter out instances for internal types. I am removing the filtering for the moment, and will try to fix this properly later. - - - - - 3e758dad by David Waern at 2008-06-30T18:50:30+00:00 Run haddock in-place during testing - - - - - d9dab0ce by David Waern at 2008-07-08T21:04:32+00:00 Remove index.html and doc-index.html from output, they should not be versioned - - - - - 3e6c4681 by David Waern at 2008-07-08T21:06:42+00:00 Update test output following change to instance filtering - - - - - e34a3f14 by David Waern at 2008-07-12T16:48:28+00:00 Stop using the map from exported names to declarations During creation of the interface, we were using two maps: one from exported names to declarations, and one from all defined names in the module to declarations. The first contained subordinate names while the second one didn't. The first map was never used to look up names not defined in the associated module, so if we add subordinate names to the second map, we could use it everywhere. That's that this patch does. This simplifies code because we don't have to pass around two maps everywhere. We now store the map from locally defined things in the interface structure instead of the one from exported names. - - - - - 2e1d2766 by David Waern at 2008-07-12T16:55:21+00:00 Get the all locally defined names from GHC API We previously had some code to compute all locally defined names in a module including subordinate names. We don't need it since we can get the names from modInfoTyThings in the GHC API. - - - - - bf637994 by David Waern at 2008-07-13T13:09:16+00:00 Refactoring in H.Interface.Create We were creating a doc map, a declaration map and a list of entities separately by going through the HsGroup. These structures were all used to build the interface of a module. Instead of doing this, we can start by creating a list of declarations from the HsGroup, then collect the docs directly from this list (instead of using the list of entities), creating a documentation map. We no longer need the Entity data type, and we can store a single map from names to declarations and docs in the interface, instead of the declaration map and the doc map. This way, there is only one place where we filter out the declarations that we don't want, and we can remove a lot of code. Another advantage of this is that we can create the exports directly out of the list of declarations when we export the full module contents. (Previously we did a look up for each name to find the declarations). This is faster and removes another point where we depend on names to identify exported declarations, which is good because it eliminates problems with instances (which don't have names). - - - - - 547e410e by David Waern at 2008-07-13T13:34:51+00:00 Remove FastString import and FSLIT macro in H.I.Create -- they were unused - - - - - 693759d1 by David Waern at 2008-07-13T13:36:23+00:00 Remove unused import from H.I.Create - - - - - cde6e7fb by David Waern at 2008-07-13T13:51:54+00:00 Small touches - - - - - 96de8f1d by David Waern at 2008-07-20T11:21:46+00:00 Preparation for rendering instances as separate declarations We want to be able to render instances as separate declarations. So we remove the Name argument of ExportDecl, since instances are nameless. This patch also contains the first steps needed to gather type family instances and display them in the backend, but the implementation is far from complete. Because of this, we don't actually show the instances yet. - - - - - b0f824fb by David Waern at 2008-07-20T15:53:08+00:00 Follow changes to ExportDecl in Hoogle - - - - - 1192eff3 by Neil Mitchell at 2008-06-26T00:28:10+00:00 Change how the Hoogle backend outputs classes, adding the context in - - - - - 7a0d1464 by Neil Mitchell at 2008-06-26T00:28:46+00:00 Remove the indent utility function from Hoogle backend - - - - - 3361241b by Neil Mitchell at 2008-06-26T09:45:09+00:00 Add support for Hoogle writing ForeignImport/ForeignExport properly - - - - - 795ad3bf by Neil Mitchell at 2008-06-26T12:15:25+00:00 Flesh out the Hoogle code to render documentation - - - - - 23277995 by Neil Mitchell at 2008-06-26T14:56:41+00:00 Fix a bug in the Hoogle backend, unordered lists were being written out <ul>...</u> - - - - - db739b27 by Neil Mitchell at 2008-06-26T15:09:54+00:00 Remove any white space around a <li> element - - - - - f2e6bb8c by Neil Mitchell at 2008-07-10T15:30:47+00:00 Remove the TODO in the Hoogle HTML generation, was already done - - - - - 693ec9a3 by Neil Mitchell at 2008-07-10T15:53:00+00:00 Put brackets round operators in more places in the Hoogle output - - - - - 842313aa by Neil Mitchell at 2008-07-10T16:01:25+00:00 Print type signatures with brackets around the name - - - - - cf93deb0 by David Waern at 2008-07-20T17:04:22+00:00 Bump version number to 2.2.0 - - - - - 30e6a8d1 by David Waern at 2008-07-20T17:04:41+00:00 Resolve conflicts in H.B.Hoogle - - - - - 1f0071c9 by David Waern at 2008-07-23T23:05:01+00:00 Add "all" command to runtests.hs that runs all tests despite failures - - - - - f2723023 by David Waern at 2008-07-23T23:08:39+00:00 Update tests/README - - - - - c0304a11 by David Waern at 2008-07-23T23:21:15+00:00 Be compatible with GHC 6.8.3 The cabal file is converted to use the "new" syntax with explicit Library and Executable sections. We define the __GHC_PATCHLEVEL__ symbol using a conditinal cpp-options field in the cabal file. (Ideally, Cabal would define the symbol for us, like it does for __GLASGOW_HASKELL__). We use these symbols to #ifdef around a small difference between 6.8.2 and 6.8.3. Previously, we only supported GHC 6.8.2 officially but the dependencies field said "ghc <= 6.9". This was just for convenience when testing against the (then compatible) HEAD version of GHC, and was left in the release by mistake. Now, we support both GHC 6.8.2 and 6.8.3 and the dependencies field correctly reflects this. - - - - - 88a5fe71 by David Waern at 2008-07-23T23:54:16+00:00 Depend on the currently available ghc-paths versions only - - - - - 8738d97b by David Waern at 2008-07-24T10:50:44+00:00 FIX haskell/haddock#44: Propagate parenthesis level when printing documented types - - - - - 05339119 by David Waern at 2008-07-24T16:06:18+00:00 Drop unnecessary parenthesis in types, put in by the user We were putting in parenthesis were the user did. Let's remove this since it just clutters up the types. The types are readable anyway since we print parens around infix operators and do not rely on fixity levels. When doing this I discovered that we were relying on user parenthesis when printin types like (a `O` b) c. This patchs fixes this problem so that parenthesis are always inserted around an infix op application in case it is applied to further arguments, or if it's an arguments to a type constructor. Tests are updated. - - - - - b3a99828 by David Waern at 2008-07-24T10:19:43+00:00 Print parenthesis around non-atomic banged types Fixes half of haskell/haddock#44 - - - - - ab5238e0 by David Waern at 2008-07-24T22:07:49+00:00 Add a reference file for the TypeFamilies test - - - - - 1941cc11 by David Waern at 2008-07-25T17:15:53+00:00 Simplify definition of pretty and trace_ppr - - - - - e3bfa33c by David Waern at 2008-07-25T17:18:27+00:00 Warning messages Output a warning when filtering out data/type instances and associated types in instances. We don't show these in the documentation yet, and we need to let the user know. - - - - - 9b85fc89 by David Waern at 2008-07-25T17:45:40+00:00 Doc: Mention Hoogle in the Introduction - - - - - afb2dd60 by David Waern at 2008-07-25T17:49:00+00:00 Doc: update -B description - - - - - 584c0c91 by David Waern at 2008-07-25T18:11:38+00:00 Doc: describe -w flag - - - - - 77619c24 by David Waern at 2008-07-28T12:29:07+00:00 Remove TODO from cabal file - - - - - 96717d5f by David Waern at 2008-07-28T12:29:27+00:00 Support type equality predicates - - - - - c2fd2330 by David Waern at 2008-07-29T19:45:14+00:00 Move unL from H.B.Hoogle to H.GHC.Utils I like Neil's shorter unL better than unLoc from the GHC API. - - - - - c4c3bf6a by David Waern at 2008-07-29T19:47:36+00:00 Do not export ATs when not in list of subitems - - - - - bf9a7b85 by David Waern at 2008-08-03T11:42:59+00:00 Filter out ForeignExports - - - - - df59fcb0 by David Waern at 2008-08-03T14:02:51+00:00 Filter out more declarations The previous refactorings in H.I.Create introduced a few bugs. Filtering of some types of declarations that we don't handle was removed. This patch fixes this. - - - - - 2f8a958b by David Waern at 2008-08-03T15:24:07+00:00 Move reL to H.GHC.Utils so we can use it everywhere - - - - - 8ec15efd by David Waern at 2008-08-03T15:25:00+00:00 Use isVanillaLSig from GHC API instead of home brewn function - - - - - 300f93a2 by David Waern at 2008-08-03T15:25:27+00:00 Filter out separately exported ATs This is a quick and dirty hack to get rid of separately exported ATs. We haven't decided how to handle them yet. No warning message is given. - - - - - 8776d1ec by David Waern at 2008-08-03T16:21:21+00:00 Filter out more declarations and keep only vanilla type sigs in classes - - - - - ea07eada by David Waern at 2008-08-03T16:48:00+00:00 Fix layout - - - - - dd5e8199 by David Waern at 2008-08-03T16:50:52+00:00 Move some utility functions from H.I.Create to H.GHC.Utils - - - - - 4a1dbd72 by David Waern at 2008-08-03T17:39:55+00:00 Do not filter out doc declarations - - - - - 0bc8dca4 by David Waern at 2008-08-03T17:47:26+00:00 Filter out separately exported ATs (take two) - - - - - af970fe8 by David Waern at 2008-08-03T22:39:17+00:00 Update CHANGES - - - - - 5436ad24 by David Waern at 2008-08-03T22:40:20+00:00 Bump version number to 2.2.1 - - - - - d66de448 by David Waern at 2008-08-05T19:00:32+00:00 Remove version restriction on ghc-paths - - - - - 534b1364 by David Waern at 2008-08-05T19:04:35+00:00 Bump version to 2.2.2 and update CHANGES - - - - - 549188ff by David Waern at 2008-08-05T19:16:49+00:00 Fix CHANGES - - - - - 0d156bb4 by Luke Plant at 2008-08-11T15:20:59+00:00 invoking haddock clarification and help - - - - - 748295cc by David Waern at 2008-08-11T18:56:37+00:00 Doc: say that the --hoogle option is functional - - - - - 43301db4 by David Waern at 2008-08-05T19:26:08+00:00 Change ghc version dependency to >= 6.8.2 - - - - - 3e5a53b6 by David Waern at 2008-08-10T22:42:05+00:00 Make H.GHC.Utils build with GHC HEAD - - - - - 7568ace0 by David Waern at 2008-08-11T19:41:54+00:00 Import Control.OldException instead of C.Exception when using ghc >= 6.9 We should really test for base version instead, but I don't currently know which version to test for. - - - - - b71ae991 by David Waern at 2008-08-12T22:40:39+00:00 Make our .haddock file version number depend on the GHC version We need to do this, since our .haddock format can potentially change whenever GHC's version changes (even when only the patchlevel changes). - - - - - 6307ce3f by David Waern at 2008-08-12T22:49:57+00:00 Remove matching on NoteTy in AttachInstances, it has been removed - - - - - 2dbcfd5f by David Waern at 2008-08-12T23:02:02+00:00 Comment out H.GHC.loadPackages - it is unused and doesn't build with ghc >= 6.9 - - - - - c74db5c2 by David Waern at 2008-08-12T23:03:58+00:00 Hide <.> from GHC import in Hoogle only for ghc <= 6.8.3 - - - - - 69a44ebb by David Waern at 2008-08-12T23:11:12+00:00 Follow changes to parseDynamic/StaticFlags - - - - - 5881f3f0 by David Waern at 2008-08-13T21:43:58+00:00 Add __GHC_PATCHLEVEL__ symbol also when building the library - - - - - 8574dc11 by David Waern at 2008-08-13T21:44:17+00:00 Follow move of package string functions from PackageConfig to Module - - - - - c9baa77f by David Waern at 2008-08-13T21:45:29+00:00 Follow extensible exceptions changes - - - - - 9092de15 by David Waern at 2008-08-13T21:46:20+00:00 Update test following Haddock version change - - - - - ebe569a4 by David Waern at 2008-08-13T21:46:54+00:00 Follow changes to parseDynamic- parseStaticFlags in GHC - - - - - b8a5ffd3 by David Waern at 2008-08-13T21:47:36+00:00 Follow changes to Binary in GHC 6.9 - - - - - edfda1cc by David Waern at 2008-08-13T21:50:17+00:00 Change ghc version dependency to >= 6.8.2 && <= 6.9 - - - - - d59be1cf by Neil Mitchell at 2008-08-12T16:02:53+00:00 Output all items, even if they are not defined in this module - ensures map comes from Prelude, not just GHC.Base - - - - - dda93b9f by Neil Mitchell at 2008-08-12T21:37:32+00:00 Add support for type synonyms to Hoogle, was accidentally missing before (woops!) - - - - - b6ee795c by Neil Mitchell at 2008-08-13T14:03:24+00:00 Generalise Hoogle.doc and add a docWith - - - - - 415e1bb2 by Neil Mitchell at 2008-08-13T14:03:46+00:00 Make Hoogle add documentation to a package - - - - - 790a1202 by Neil Mitchell at 2008-08-18T12:52:43+00:00 Use the same method to put out signatures as class methods in the Hoogle backend - - - - - ded37eba by Neil Mitchell at 2008-08-18T12:53:04+00:00 Remove Explicit top-level forall's when pretty-printing signatures - - - - - 6468c722 by Neil Mitchell at 2008-08-20T07:59:13+00:00 Simplify the code by removing not-to-important use of <.> in the Hoogle back end - - - - - 788c3a8b by Neil Mitchell at 2008-08-21T18:20:24+00:00 In the hoogle back end, markup definition lists using <i>, not <b> - - - - - 77d4b000 by Ian Lynagh at 2008-08-14T10:49:14+00:00 Add a Makefile for GHC's build system. Still won't work yet, but we're closer - - - - - 920440d7 by Ian Lynagh at 2008-08-27T18:06:46+00:00 Add haddock.wrapper - - - - - bcda925f by Ian Lynagh at 2008-08-27T18:07:02+00:00 Add a manual Cabal flag to control the ghc-paths dependency - - - - - 04d194e2 by Ian Lynagh at 2008-08-27T20:41:27+00:00 Update extensions in Cabal file Use ScopedTypeVariables instead of PatternSignatures - - - - - 12480043 by Ian Lynagh at 2008-08-27T20:41:55+00:00 Increase the upper bound on the GHC version number - - - - - b1f809a5 by Ian Lynagh at 2008-08-27T21:32:22+00:00 Fix some warnings - - - - - aea0453d by Ian Lynagh at 2008-08-28T14:22:29+00:00 Fixes for using haddock in a GHC build tree - - - - - ad23bf86 by Ian Lynagh at 2008-08-28T21:14:27+00:00 Don't use Cabal wrappers on Windows - - - - - 35858e4c by Ian Lynagh at 2008-08-29T00:07:42+00:00 Fix in-tree haddock on Windows - - - - - c2642066 by Ian Lynagh at 2008-09-03T22:35:53+00:00 follow library changes - - - - - 2eb55d50 by Ian Lynagh at 2008-09-07T18:52:51+00:00 bindist fixes - - - - - 3daa5b59 by Ian Lynagh at 2008-09-10T16:58:18+00:00 We need to tell haddock that its datasubdir is . or it can't find package.conf - - - - - 388fd8c2 by Ian Lynagh at 2008-09-10T19:47:44+00:00 Fix haddock inplace on Windows - - - - - 70a641c1 by Ian Lynagh at 2008-09-10T22:15:44+00:00 Fix installed haddock on Windows - - - - - 83c1e997 by Neil Mitchell at 2008-09-11T10:48:55+00:00 Import GHC.Paths if not IN_GHC_TREE, seems to match the use of GHC.Paths functions much better - - - - - b452519b by Ian Lynagh at 2008-09-12T12:58:24+00:00 Add a LANGUAGE ForeignFunctionInterface pragma - - - - - afbd592c by Ian Lynagh at 2008-09-12T12:59:13+00:00 Wibble imports - - - - - 547ac4ad by Ian Lynagh at 2008-09-14T15:34:22+00:00 Add a "#!/bin/sh" to haddock.wrapper - - - - - f207a807 by Ian Lynagh at 2008-09-15T10:02:32+00:00 Use "exec" when calling haddock in the wrapper - - - - - 2ee68509 by Thomas Schilling at 2008-09-15T09:09:16+00:00 Port Haddock.Interface to new GHC API. This required one bigger change: 'readInterfaceFile' used to take an optional 'Session' argument. This was used to optionally update the name cache of an existing GHC session. This does not work with the new GHC API, because an active session requires the function to return a 'GhcMonad' action, but this is not possible if no session is provided. The solution is to use an argument of functions for reading and updating the name cache and to make the function work for any monad that embeds IO, so it's result type can adapt to the calling context. While refactoring, I tried to make the code a little more self-documenting, mostly turning comments into function names. - - - - - 3bb96431 by Thomas Schilling at 2008-09-15T09:09:37+00:00 Reflect GHC API changes. - - - - - 2e60f714 by Thomas Schilling at 2008-09-15T09:10:37+00:00 Port Haddock.GHC.Typecheck to new GHC API. - - - - - 9cfd4cff by Thomas Schilling at 2008-09-15T09:11:00+00:00 Port Haddock.GHC to new GHC API. - - - - - caffa003 by Thomas Schilling at 2008-09-15T09:11:25+00:00 Port Main to new GHC API. - - - - - 069a4608 by Ian Lynagh at 2008-09-21T11:19:00+00:00 Fix paths used on Windows frmo a GHC tree: There is no whare directory - - - - - 7ceee1f7 by Ian Lynagh at 2008-09-21T12:20:16+00:00 Fix the in-tree haddock on Windows - - - - - 0d486514 by Ian Lynagh at 2008-09-23T18:06:58+00:00 Increase the GHC upper bound from 6.11 to 6.13 - - - - - f092c414 by Neil Mitchell at 2008-09-11T14:56:07+00:00 Do not wrap __ in brackets - - - - - 036bdd13 by Ian Lynagh at 2008-09-28T01:42:35+00:00 Fix building haddock when GhcProfiled=YES - - - - - 01434a89 by David Waern at 2008-09-24T20:24:21+00:00 Add PatternSignatures LANGUAGE pragma to Main and Utils - - - - - 1671a750 by David Waern at 2008-10-02T22:57:25+00:00 For source links, get original module from declaration name instead of environment. Getting it from the environment must have been a remnant from the times when we were using unqualified names (versions 0.x). - - - - - a25dde99 by David Waern at 2008-10-02T22:59:57+00:00 Remove ifaceEnv from Interface - it's no longer used - - - - - 610993da by David Waern at 2008-10-02T23:04:58+00:00 Write a comment about source links for type instance declarations - - - - - 5a96b5d5 by Thomas Schilling at 2008-10-03T10:45:08+00:00 Follow GHC API change of parseModule. - - - - - 5a943ae5 by Ian Lynagh at 2008-10-03T15:56:58+00:00 TAG 2008-10-03 - - - - - 76cdd6ae by Thomas Schilling at 2008-10-08T12:29:50+00:00 Only load modules once when typechecking with GHC. This still doesn't fix the memory leak since the typechecked source is retained and then processed separately. To fix the leak, modules must be processed directly after typechecking. - - - - - 7074d251 by David Waern at 2008-10-09T23:53:54+00:00 Interleave typechecking with interface creation At the same time, we fix a bug where the list of interfaces were processed in the wrong order, when building the links and renaming the interfaces. - - - - - 4b9b2b2d by David Waern at 2008-10-09T23:54:49+00:00 Add some strictness annotations in Interface We add some strictness annotations to the fields of Interface, so that less GHC data is hold on to during processing. - - - - - 22035628 by David Waern at 2008-10-10T20:02:31+00:00 Remove typecheckFiles and MonadUtils import from H.GHC.Typeccheck - - - - - be637ad3 by David Waern at 2008-10-10T20:33:38+00:00 Make Haddock build with GHC 6.8.2 - - - - - 523b3404 by David Waern at 2008-10-10T21:08:09+00:00 Fix documentation for createInterfaces - - - - - e1556702 by David Waern at 2008-10-10T21:26:19+00:00 Hide H.Utils in library - - - - - a8e751c3 by David Waern at 2008-10-10T21:34:59+00:00 Add back .haddock file versioning based on GHC version It was accidentally removed in the patch for GHC 6.8.2 compatibility - - - - - 06fb3c01 by David Waern at 2008-10-10T21:47:15+00:00 Bump version number to 2.3.0 - - - - - ff087fce by David Waern at 2008-10-10T22:35:49+00:00 Add support for DocPic The support for DocPic was merged into the GHC source long ago, but the support in Haddock was forgotten. Thanks Peter Gavin for submitting this fix! - - - - - 3af85bf6 by David Waern at 2008-10-10T23:34:05+00:00 Update tests - - - - - 0966873c by Simon Marlow at 2008-10-10T14:43:04+00:00 no need for handleErrMsg now, we don't throw any ErrMsgs - - - - - f1870de3 by Clemens Fruhwirth at 2008-10-10T13:29:36+00:00 Compile with wrapper but remove it for dist-install - - - - - 7b440dc2 by David Waern at 2008-10-11T14:02:25+00:00 Remove interface from LinksInfo It was there to know the documentation home module when creating a wiki link, but we already know this since we have the DocName. - - - - - e5729e6a by David Waern at 2008-10-15T20:49:18+00:00 Wibble - - - - - b2a8e01a by David Waern at 2008-10-15T21:03:36+00:00 Use type synonyms for declarations and docs in H.I.Create - - - - - be71a15b by David Waern at 2008-10-15T21:12:17+00:00 Comment out unused type family stuff completely - - - - - 91aaf075 by David Waern at 2008-10-15T21:49:04+00:00 Wibble - - - - - 42ba4eb4 by David Waern at 2008-10-15T21:53:53+00:00 Move convenient type synonym to H.Types - - - - - db11b723 by David Waern at 2008-10-15T22:14:07+00:00 Add DeclInfo to H.Types - - - - - 193552b6 by David Waern at 2008-10-15T22:15:01+00:00 Add subordinates with docs to the declaration map The only place in the code where we want the subordinates for a declaration is right after having looked up the declaration in the map. And since we include subordinates in the map, we might as well take the opportunity to store those subordinates that belong to a particular declaration together with that declaration. We also store the documentation for each subordinate. - - - - - 31e6eebc by David Waern at 2008-10-16T17:18:47+00:00 Wibble - - - - - 0dcbd79f by David Waern at 2008-10-16T20:58:42+00:00 Fix haskell/haddock#61 We were not getting docs for re-exported class methods. This was because we were looking up the docs in a map made from the declarations in the current module being rendered. Obviously, re-exported class methods come from another module. Class methods and ATs were the only thing we were looking up using the doc map, everything else we found in the ExporItems. So now I've put subordinate docs in the ExportItem's directly, to make things a bit more consistent. To do this, I added subordinates to the the declarations in the declaration map. This was easy since we were computing subordinates anyway, to store stand-alone in the map. I added a new type synonym 'DeclInfo', which is what we call what is now stored in the map. This little refactoring removes duplicate code to retrieve subordinates and documentation from the HsGroup. - - - - - de47f20a by David Waern at 2008-10-16T22:06:35+00:00 Document function and improve its layout - - - - - e74e625a by Thomas Schilling at 2008-10-20T11:12:57+00:00 Force interface more aggressively. For running Haddock on GHC this reduces memory usage by about 50 MB on a 32 bit system. A heap profile shows total memory usage peak at about 100 MB, but actual usage is at around 300 MB even with compacting GC (+RTS -c). - - - - - b63ac9a1 by David Waern at 2008-10-20T20:25:50+00:00 Make renamer consistent Instead of explicitly making some binders Undocumented, treat all names the same way (that is, try to find a Documented name). - - - - - f6de0bb0 by Ian Lynagh at 2008-09-19T00:54:43+00:00 TAG GHC 6.10 fork - - - - - 74599cd0 by David Waern at 2008-10-20T21:13:24+00:00 Do not save hidden modules in the .haddock file We were saving interfaces of all processed modules including those hidden using {-# OPTIONS_HADDOCK hide #-} in the .haddock file. This caused broken links when generating the index for the libraries that come with GHC. This patch excludes modules with hidden documentation when writing .haddock files. It should fix the above problem. - - - - - 7b6742e9 by David Waern at 2008-10-21T19:54:52+00:00 Do not save hidden modules in the .haddock file (also for ghc >= 6.9) When writing the first patch, I forgot to do the fix in both branches of an #if macro. - - - - - b99b1951 by David Waern at 2008-10-22T20:04:18+00:00 Remove subordinate map and its usage It is not needed now that we store subordinate names in the DeclInfo map. - - - - - da97cddc by David Waern at 2008-10-22T20:11:46+00:00 Tidy up code in H.I.Create a little Remove commented out half-done type instance support, and remove DeclWithDoc synonym. - - - - - 6afa76f3 by David Waern at 2008-10-22T21:17:29+00:00 Fix warnings in H.GHC.Utils - - - - - 171ea1e8 by David Waern at 2008-10-22T21:35:04+00:00 Fix warnings in H.Utils - - - - - c8cb3b91 by David Waern at 2008-10-22T21:36:49+00:00 Wibble - - - - - 767fa06a by David Waern at 2008-10-27T19:59:04+00:00 Make named doc comments into ExportDoc instead of ExportDecl Fixes a crash when processing modules without export lists containing named docs. - - - - - e638bbc6 by David Waern at 2008-11-02T22:21:10+00:00 Add HCAR entry - - - - - 92b4ffcf by David Waern at 2008-11-02T22:44:19+00:00 Update CHANGES - - - - - 84d4da6e by David Waern at 2008-11-03T11:25:04+00:00 Add failing test for template haskell crash - - - - - 2a9cd2b1 by David Waern at 2008-11-04T21:13:44+00:00 Add tests/TH.hs - - - - - 8a59348e by David Waern at 2008-11-04T21:30:26+00:00 TAG 2.3.0 - - - - - 54f70d31 by Thomas Schilling at 2008-10-24T17:04:08+00:00 Enable framed view of the HTML documentation. This patch introduces: - A page that displays the documentation in a framed view. The left side will show a full module index. Clicking a module name will show it in the right frame. If Javascript is enabled, the left side is split again to show the modules at the top and a very short synopsis for the module currently displayed on the right. - Code to generate the mini-synopsis for each module and the mini module index ("index-frames.html"). - CSS rules for the mini-synopsis. - A very small amount of javascript to update the mini-synopsis (but only if inside a frame.) Some perhaps controversial things: - Sharing code was very difficult, so there is a small amount of code duplication. - The amount of generated pages has been doubled, since every module now also gets a mini-synopsis. The overhead should not be too much, but I haven't checked. Alternatively, the mini-synopsis could also be generated using Javascript if we properly annotate the actual synopsis. - - - - - 5d7ea5a6 by David Waern at 2008-11-04T23:20:17+00:00 Follow change to ExportDecl in frames code - - - - - 60e16308 by David Waern at 2008-11-04T23:35:26+00:00 Update CHANGES - - - - - d63fd26d by David Waern at 2008-11-04T23:37:43+00:00 Bump version number - - - - - c1660c39 by David Waern at 2008-11-04T23:44:46+00:00 Update CHANGES - - - - - 995ab384 by David Waern at 2008-11-04T23:55:21+00:00 Remove .ref files from tests/output/ - - - - - 1abbbe75 by David Waern at 2008-11-04T23:57:41+00:00 Output version info before running tests - - - - - 649b182f by David Waern at 2008-11-05T22:45:37+00:00 Add ANNOUNCE message - - - - - c36ae0bb by David Waern at 2008-11-05T23:15:35+00:00 Update ANNOUNCE - - - - - 9c4f3d40 by David Waern at 2008-11-05T23:18:30+00:00 Wibble - - - - - 5aac87ce by David Waern at 2008-11-06T21:07:48+00:00 Depend on base 4.* when using GHC >= 6.9, otherwise 3.* - - - - - b9796a74 by David Waern at 2008-11-06T21:13:40+00:00 Bump version to 2.4.1 and update CHANGES - - - - - d4b26baa by David Waern at 2008-11-06T21:26:33+00:00 Depend on base 4.0.* instead of 4.* - - - - - 2cb0903c by David Waern at 2008-11-06T21:46:53+00:00 Fix warnings in H.B.HH and H.B.HH2 - - - - - e568e89a by David Waern at 2008-11-06T21:47:12+00:00 Fix warnings in Haddock.ModuleTree - - - - - 9dc14fbd by David Waern at 2008-11-06T21:47:52+00:00 Fix warnings in Haddock.Version - - - - - 02ac197c by David Waern at 2008-11-06T21:51:31+00:00 Fix warnings in H.InterfaceFile and H.Options - - - - - 63e7439a by David Waern at 2008-11-06T21:59:45+00:00 Fix warnings in H.GHC.Typecheck - - - - - 4bca5b68 by David Waern at 2008-11-08T13:43:42+00:00 Set HscTarget to HscNothing instead of HscAsm There used to be a bug in the GHC API that prevented us from setting this value. - - - - - 07357aec by David Waern at 2008-11-09T22:27:00+00:00 Re-export NameCache and friends from Distribution.Haddock - - - - - ea554b5a by David Waern at 2008-11-09T23:14:10+00:00 Add Haddock.GHC.Utils to other-modules in library - - - - - 74aecfd7 by David Waern at 2008-11-10T01:18:57+00:00 Export DocName in the library - - - - - 241a58b3 by David Waern at 2008-11-10T01:19:18+00:00 Document the functions in H.DocName - - - - - edc2ef1b by David Waern at 2008-11-10T01:20:52+00:00 Export H.DocName in the library - - - - - 4f588d55 by David Waern at 2008-11-10T01:29:14+00:00 Make DocName an instance of NamedThing - - - - - b4647244 by David Waern at 2008-11-15T22:58:18+00:00 Reflect version bump in test suite - - - - - 4bee8ce2 by David Waern at 2008-11-15T22:58:45+00:00 Update tests For unknown reasons, test output for Bug1 and Test has changed for the better. - - - - - 1690e2f9 by David Waern at 2008-11-15T22:59:33+00:00 Store hidden modules in .haddock files We store documentation for an entity in the 'InstalledInterface' of the definition site module, and never in the same structure for a module which re-exports the entity. So when a client of the Haddock library wants to look up some documentation, he/she might need to access a hidden module. But we currently don't store hidden modules in the .haddock files. So we add the hidden modules and the Haddock options to the .haddock files. The options will be used to filter the module list to obtain the visible modules only, which is necessary for generating the contents and index for installed packages. - - - - - 8add6435 by David Waern at 2008-11-16T14:35:50+00:00 Bump major version number due to .haddock file format change - - - - - 48bfcf82 by David Waern at 2008-11-23T14:32:52+00:00 Update tests to account for version number bump - - - - - 0bbd1738 by David Waern at 2008-11-23T14:33:31+00:00 HADDOCK_DATA_DIR changed to haddock_datadir - - - - - 5088b78c by David Waern at 2008-11-23T17:13:21+00:00 FIX haskell/haddock#45: generate two anchors for each name We generate two anchor tags for each name, one where we don't escape the name and one where we URI-encode it. This is for compatibility between IE and Opera. Test output is updated. - - - - - 5ee5ca3b by Neil Mitchell at 2008-11-27T14:38:11+00:00 Drop HsDocTy annotations, they mess up pretty printing and also have a bracketing bug (#2584) - - - - - 51c014e9 by Roman Cheplyaka at 2008-11-27T22:27:36+00:00 Allow referring to a specific section within a module in a module link Fixes haskell/haddock#65 - - - - - 4094bdc5 by David Waern at 2008-11-28T21:13:33+00:00 Update tests following anchor change - - - - - f89552dd by Thomas Schilling at 2008-11-29T16:16:20+00:00 Haddock really shouldn't try to overwrite files. - - - - - 98127499 by David Waern at 2008-12-07T14:09:15+00:00 Solve conflict - - - - - 319356c5 by David Waern at 2008-10-22T21:16:55+00:00 Add -Wall -Werror to ghc-options - - - - - 3c4968c9 by David Waern at 2008-11-04T23:38:56+00:00 TAG 2.4.0 - - - - - 4b21e003 by David Waern at 2008-11-06T21:14:04+00:00 TAG 2.4.1 - - - - - 8e0cad5c by David Waern at 2008-12-07T14:12:54+00:00 Remove -Werror - - - - - 299d6deb by David Waern at 2008-12-07T14:25:18+00:00 Remove -Wall, we'll focus on warnings after 6.10.2 is out - - - - - 5f4216b6 by David Waern at 2008-12-07T20:58:05+00:00 Resolve conflict properly - - - - - 67d774e7 by Neil Mitchell at 2008-12-15T11:44:26+00:00 Make forall's in constructors explicit, i.e. data Foo = Foo {foo :: Eq a => a} - - - - - 61851792 by Neil Mitchell at 2008-12-18T15:39:39+00:00 Try and find a better package name than "main" for Hoogle, goes wrong when working on an executable rather than a library - - - - - 2fab8554 by David Waern at 2008-12-08T23:19:48+00:00 Make visible names from ExportItems Instead of a complicated calculation of visible names out of GHC's export items, we can get them straight out of the already calculated ExportItems. The ExportItems should represent exactly those items that are visible in an interface. If store all the exported sub-names in ExportDecl instead of only those with documentation, the calculation becomes very simple. So we do this change as well (should perhaps have been a separate patch). This should fix the problem with names from ghc-prim not appearing in the link environment. - - - - - 7caadd8c by Ian Lynagh at 2008-12-11T17:01:04+00:00 Wrap the GHC usage with defaultCleanupHandler This fixes a bug where haddock leaves /tmp/ghc* directories uncleaned. - - - - - 7c9fc9a5 by David Waern at 2009-01-02T21:38:27+00:00 Show re-exported names from external packages again This fixes GHC ticket 2746. In order to also link to the exported subordinate names of a declaration, we need to re-introduce the sub map in the .haddock files. - - - - - 119e4e05 by David Waern at 2009-01-06T23:34:17+00:00 Do not process boot modules We should of course not try to produce documentation for boot modules! The reason this has worked in the past is that the output of "real" modules overwrites the output of boot modules later in the process. However, this causes a subtle link environment problem. So let's get rid of this stupid behaviour. We avoid processing boot modules, but we continue to typecheck them. - - - - - c285b9d2 by David Waern at 2009-01-08T18:03:36+00:00 Export modules also when coming from external packages This seems to have regressed since a refactoring that was part of the 2.3.0 release. - - - - - 24031c17 by David Waern at 2009-01-10T15:26:26+00:00 Change version to 2.4.2 - no need to go to 2.5.0 - - - - - 864d1c3f by David Waern at 2009-01-10T15:35:20+00:00 Update tests to account for version number change - - - - - 524ba886 by David Waern at 2009-01-10T18:29:17+00:00 Add test for Template Haskell splicing - - - - - 05e6e003 by David Waern at 2009-01-10T19:35:42+00:00 Fix Trac haskell/haddock#68: Turn on compilation via C for Template Haskell packages We can't use HscNothing if we need to run code coming from modules inside the processed package during typechecking, which is the case for some packages using Template Haskell. This could be improved, to e.g. use HscInterpreted and HscNothing where possible, instead of using HscC for all modules in the package. - - - - - 2b2bafa1 by David Waern at 2009-01-10T20:22:25+00:00 Only use needsTemplateHaskell when compiling with GHC 6.10.2 or above - - - - - bedc3a93 by Ian Lynagh at 2009-01-11T14:58:41+00:00 Fix the location of INPLACE_PKG_CONF; fixes the build Spotted by Conal Elliott - - - - - 943107c8 by David Waern at 2009-01-20T19:27:39+00:00 Document H.I.Create.collectDocs better - - - - - c6252e37 by David Waern at 2009-01-20T19:29:51+00:00 Fix Trac haskell/haddock#59: TH-generated declarations disappearing This patch was contributed by Joachim Breitner (nomeata). - - - - - 3568a6af by David Waern at 2009-01-21T21:41:48+00:00 Do not indicate that a constructor argument is unboxed We only show the strictness annotation for an unboxed constructor argument. The fact that it is unboxed is an implementation detail and should not be part of the module interface. - - - - - 562a4523 by David Waern at 2009-01-22T18:53:49+00:00 Fix Trac haskell/haddock#50: do not attach docs to pragmas or other kinds of non-declarations We now filter out everything that is not a proper Haskell declaration before collecting the docs and attaching them to declarations. - - - - - 6fdf21c2 by David Waern at 2009-01-22T19:48:09+00:00 Add test for quasi quotation. No reference output yet. - - - - - dc4100fd by David Waern at 2009-01-22T19:57:47+00:00 Improve quasi-quotation test and add reference output - - - - - 908b74bb by David Waern at 2009-01-23T23:22:03+00:00 Filter out separately exported associated types in a smarter way - - - - - f6b42ecb by David Waern at 2009-01-24T16:54:39+00:00 Correct spelling mistake in error message - - - - - 24e4245d by David Waern at 2009-01-24T17:48:03+00:00 Correct comment - - - - - b5e8462f by David Waern at 2009-02-07T13:22:29+00:00 Do not show a subordinate at the top level if its parent is also exported See note in the source code for more info. - - - - - 4b09de57 by David Waern at 2009-02-07T13:53:53+00:00 Update test following change to top level subordinates - - - - - 76379896 by David Waern at 2009-02-07T13:58:04+00:00 Remove html files in the tests/output/ directory which have been accidentally added - - - - - 1a6d8b10 by Joachim Breitner at 2009-02-20T10:29:43+00:00 Typo in comment - - - - - fec367d0 by David Waern at 2009-02-24T20:21:17+00:00 Fix small bug The rule is to prefer type constructors to other things when an identifier in a doc string can refer to multiple things. This stopped working with newer GHC versions (due to a tiny change in the GHC renamer). We implement this rule in the HTML backend for now, instead of fixing it in GHC, since we will move renaming of doc strings to Haddock in the future anyway. - - - - - 9b4172eb by David Waern at 2009-02-25T20:04:38+00:00 Fix bad error handling with newer GHCs When support for GHC 6.10 was added, an error handler was installed only around the typechecking phase. This had the effect that errors thrown during dependency chasing were caught in the top-level exception handler and not printed with enough detail. With this patch we wrap the error handler around all our usage of the Ghc monad. - - - - - de2df363 by Simon Peyton Jones at 2009-02-02T16:47:42+00:00 Hide funTyConName, now exported by TypeRep - - - - - 4d40a29f by Ian Lynagh at 2009-02-12T18:57:49+00:00 Don't build the library when building in the GHC tree - - - - - 1cd0abe4 by Ian Lynagh at 2009-02-13T13:58:53+00:00 Add a ghc.mk - - - - - 3d814eeb by Ian Lynagh at 2009-02-13T18:50:28+00:00 do .depend generation for haddock with the stage1 compiler This is a bit of a hack. We mkdepend with stage1 as if .depend depends on the stage2 compiler then make goes wrong: haddock's .depend gets included, which means that make won't reload until it's built, but we can't build it without the stage2 compiler. We therefore build the stage2 compiler before its .depend file is available, and so compilation fails. - - - - - b55036a4 by Ian Lynagh at 2009-02-25T01:38:13+00:00 Give haddock a wrapper on unix in the new GHC build system - - - - - 9eabfe68 by Ian Lynagh at 2009-02-25T19:21:32+00:00 Create inplace/lib/html in the new GHC build system - - - - - 93af30c7 by Ian Lynagh at 2008-11-07T19:18:23+00:00 TAG GHC 6.10.1 release - - - - - 06e6e34a by Thomas Schilling at 2009-02-24T18:11:00+00:00 Define __GHC_PATCHLEVEL__ for recent version of GHC (stable). - - - - - 680e6ed8 by Thomas Schilling at 2009-02-24T18:12:26+00:00 'needsTemplateHaskell' is not defined in current stable GHC. - - - - - 6c5619df by David Waern at 2009-02-25T22:15:23+00:00 Hide fynTyConName only for recent GHC versions - - - - - 6b2344f1 by Ian Lynagh at 2009-02-26T00:49:56+00:00 Add the module to one of haddocks warnings - - - - - e5d11c70 by David Waern at 2009-02-27T21:37:20+00:00 Bug fix We tried to filter out subordinates that were already exported through their parent. This didn't work properly since we were in some cases looking at the grand-parent and not the parent. We now properly compute all the parent-child relations of a declaration, and use this information to get the parent of a subordinate. We also didn't consider record fields with multiple parents. This is now handled correctly. We don't currently support separately exported associated types. But when we do, they should be handled correctly by this process too. Also slightly improved the warning message that we give when filtering out subordinates. - - - - - 10a79a60 by David Waern at 2009-02-27T22:08:08+00:00 Fix error message conflict The module name is already written in the beginning of the message, as seems to be the convention in Haddock. Perhaps not so clear, but we should change it everywhere in that case. Leaving it as it is for now. - - - - - c5055c7f by David Waern at 2009-02-27T22:15:17+00:00 Shorten warning message - - - - - a72fed3a by David Waern at 2009-02-28T00:53:55+00:00 Do not show package name in warning message - - - - - a5daccb2 by Ian Lynagh at 2009-03-01T14:59:35+00:00 Install haddock in the new GHC build system - - - - - dfdb025c by Ian Lynagh at 2009-03-07T23:56:29+00:00 Relax base dependency to < 4.2, not < 4.1 - - - - - 5769c8b4 by David Waern at 2009-03-21T14:58:52+00:00 Bump .haddock file version number (due to change of format) - - - - - f1b8f67b by David Waern at 2009-03-21T14:59:26+00:00 Define __GHC_PATCHLEVEL__=1 when using ghc-6.10.1 - - - - - 23f78831 by David Waern at 2009-03-21T16:40:52+00:00 Update CHANGES - - - - - 7d2735e9 by David Waern at 2009-03-21T16:50:33+00:00 Update ANNOUNCE - - - - - 0771e00a by David Waern at 2009-03-21T16:54:40+00:00 Update ANNOUNCE, again - - - - - 81a6942a by David Waern at 2009-03-21T17:50:06+00:00 Don't be too verbose in CHANGES - - - - - 29861dcf by David Waern at 2009-03-21T18:03:31+00:00 TAG 2.4.2 - - - - - a585f285 by David Waern at 2009-03-21T19:20:29+00:00 Require Cabal >= 1.2.3 - - - - - 7c611662 by David Waern at 2009-03-21T19:21:48+00:00 TAG 2.4.2 with cabal-version >= 1.2.3 - - - - - 23b7deff by Simon Marlow at 2009-03-20T15:43:42+00:00 new GHC build system: use shell-wrappers macro - - - - - 25f8afe7 by Ian Lynagh at 2009-03-21T19:13:53+00:00 Fix (with a hack?) haddock in teh new build system - - - - - 6a29a37e by David Waern at 2009-03-24T22:10:15+00:00 Remove unnecessary LANGUAGE pragma - - - - - 954da57d by David Waern at 2009-03-24T22:21:23+00:00 Fix warnings in H.B.DevHelp - - - - - 1619f1df by David Waern at 2009-03-26T23:20:44+00:00 -Wall police in H.B.Html - - - - - b211e13b by Simon Marlow at 2009-03-24T13:00:56+00:00 install Haddock's html stuff - - - - - 78e0b107 by David Waern at 2008-12-07T19:58:53+00:00 Add verbosity flag and utils, remove "verbose" flag - - - - - 913dae06 by David Waern at 2008-12-07T20:01:05+00:00 Add some basic "verbose" mode logging in H.Interface - - - - - 1cbff3bf by David Waern at 2009-03-27T00:07:26+00:00 Fix conflicts - - - - - 22f82032 by David Waern at 2009-03-27T21:15:11+00:00 Remove H.GHC.Typecheck - - - - - 81557804 by David Waern at 2009-03-27T21:19:22+00:00 Remove docNameOrig and use getName everywhere instead - - - - - d8267213 by David Waern at 2009-03-27T21:21:46+00:00 Use docNameOcc instead of nameOccName . getName - - - - - 5d55deab by David Waern at 2009-03-27T21:33:04+00:00 Remove H.DocName and put DocName in H.Types - - - - - 8ba72611 by David Waern at 2009-03-27T22:06:26+00:00 Document DocName - - - - - 605f8ca5 by David Waern at 2009-03-27T22:45:21+00:00 -Wall police - - - - - e4da93ae by David Waern at 2009-03-27T23:12:53+00:00 -Wall police in H.B.Hoogle - - - - - bb255519 by David Waern at 2009-03-27T23:41:28+00:00 Define Foldable and Traversable instances for Located - - - - - f1195cfe by David Waern at 2009-03-27T23:51:34+00:00 Wibble - - - - - 23818d7c by David Waern at 2009-03-28T00:03:55+00:00 -Wall police in H.I.Rename - - - - - 0f050d67 by David Waern at 2009-03-28T00:15:15+00:00 -Wall police in H.I.AttachInstances - - - - - 0f3fe038 by David Waern at 2009-03-28T21:09:41+00:00 Wibble - - - - - 275d4865 by David Waern at 2009-03-28T21:27:06+00:00 Layout fix - - - - - 54ff0ef8 by David Waern at 2009-03-28T21:59:07+00:00 -Wall police in H.I.Create - - - - - 7f58b117 by David Waern at 2009-03-28T22:10:19+00:00 -Wall police in H.Interface - - - - - f0c03b44 by David Waern at 2009-03-28T22:22:59+00:00 -Wall police in Main - - - - - 29da355c by David Waern at 2009-03-28T22:23:39+00:00 Turn on -Wall -Werror - - - - - 446d3060 by David Waern at 2009-04-01T20:40:30+00:00 hlint police - - - - - 3867c9fc by David Waern at 2009-04-01T20:48:42+00:00 hlint police - - - - - bd1f1600 by David Waern at 2009-04-01T20:58:02+00:00 hlint police - - - - - e0e90866 by David Waern at 2009-04-05T12:42:53+00:00 Move H.GHC.Utils to H.GhcUtils - - - - - 9cbd426b by David Waern at 2009-04-05T12:57:21+00:00 Remove Haddock.GHC and move its (small) contents to Main - - - - - b5c2cbfd by David Waern at 2009-04-05T13:07:04+00:00 Fix whitespace and stylistic issues in Main - - - - - 3c04aa56 by porges at 2008-12-07T08:22:19+00:00 add unicode output - - - - - 607918da by David Waern at 2009-04-26T15:09:43+00:00 Resolve conflict - - - - - 4bec6b6b by Simon Marlow at 2009-05-13T10:00:31+00:00 fix markup - - - - - 436ad6f4 by Simon Marlow at 2009-03-23T11:54:45+00:00 clean up - - - - - bdcd1398 by Simon Marlow at 2009-03-24T10:36:45+00:00 new GHC build system: add $(exeext) - - - - - 9c0972f3 by Simon Marlow at 2009-03-24T11:04:31+00:00 update for new GHC build system layout - - - - - d0f3f83a by Ian Lynagh at 2009-03-29T15:31:43+00:00 GHC new build system fixes - - - - - 5a8245c2 by Ian Lynagh at 2009-04-04T20:44:23+00:00 Tweak new build system - - - - - 9c6f2d7b by Simon Marlow at 2009-05-13T10:01:27+00:00 add build instructions for GHC - - - - - 66d07c76 by Ian Lynagh at 2009-05-31T00:37:53+00:00 Quote program paths in ghc.mk - - - - - bb7de2cd by Ian Lynagh at 2009-06-03T22:57:55+00:00 Use a bang pattern on an unlifted binding - - - - - 3ad283fc by Ian Lynagh at 2009-06-13T16:17:50+00:00 Include haddock in GHC bindists - - - - - ac447ff4 by David Waern at 2009-06-24T21:07:50+00:00 Delete Haddock.Exception and move contents to Haddock.Types Only a few lines of code that mainly declares a type - why not just put it in Haddock.Types. - - - - - 4464fb9b by David Waern at 2009-06-24T22:23:23+00:00 Add Haddock module headers Add a proper Haddock module header to each module, with a more finegrained copyright. If you feel mis-accreditted, please correct any copyright notice! The maintainer field is set to haddock at projects.haskell.org. Next step is to add a brief description to each module. - - - - - 5f4c95dd by David Waern at 2009-06-24T22:39:44+00:00 Fix spelling error - - - - - 6d074cdb by David Waern at 2009-06-25T21:53:56+00:00 Document Interface and InstalledInterface better - - - - - d0cbd183 by David Waern at 2009-06-27T12:46:46+00:00 Remove misplaced whitespace in H.I.Rename - - - - - fa381c49 by David Waern at 2009-06-27T13:26:03+00:00 Fix haskell/haddock#104 - create output directory if missing - - - - - 91fb77ae by Ian Lynagh at 2009-06-25T15:59:50+00:00 TAG 2009-06-25 - - - - - 0d853f40 by Simon Peyton Jones at 2009-07-02T15:35:22+00:00 Follow extra field in ConDecl - - - - - b201735d by Ian Lynagh at 2009-07-05T16:50:35+00:00 Update Makefile for the new GHC build system - - - - - df6c0092 by Ian Lynagh at 2009-07-05T17:01:13+00:00 Resolve conflicts - - - - - 1066870a by Ian Lynagh at 2009-07-05T17:01:48+00:00 Remove the -Wwarn hack in the GHC build system - - - - - 7e856076 by Ian Lynagh at 2009-07-05T17:17:59+00:00 Fix warnings - - - - - 5d4cd958 by Ian Lynagh at 2009-07-05T19:35:40+00:00 Bump version number Cabal needs to distinguish between haddocks having a --verbose and --verbosity flag - - - - - 6ee07c99 by David Waern at 2009-07-06T20:14:57+00:00 Wibble - - - - - 2308b66f by David Waern at 2009-07-06T20:24:20+00:00 Clearer printing of versions by runtests.hs - - - - - d4b5d9ab by David Waern at 2009-07-06T21:22:42+00:00 Fix (invisible) bug introduced by unicode patch - - - - - 2caca8d8 by David Waern at 2009-07-06T21:44:10+00:00 Use HscAsm instead of HscC when using TH - - - - - 18f3b755 by David Waern at 2009-07-06T22:10:22+00:00 Update HCAR entry (by Janis) - - - - - a72ac9db by David Waern at 2009-07-06T23:01:35+00:00 Follow HsRecTy change with an #if __GLASGOW_HASKEL__ >= 611 - - - - - 549135d2 by David Waern at 2009-07-06T23:11:41+00:00 Remove unused functions from Haddock.Utils - - - - - b450134a by Isaac Dupree at 2009-07-11T14:59:00+00:00 revert to split-index for large indices - remove the search-box, because browsers have search-for-text abilities anyway. - pick 150 items in index as the arbitrary time at which to split it - notice the bug that identifiers starting with non-ASCII characters won't be listed in split-index, but don't bother to fix it yet (see ticket haskell/haddock#116, http://trac.haskell.org/haddock/ticket/116 ) - - - - - 78a5661e by Isaac Dupree at 2009-07-20T15:37:18+00:00 Implement GADT records in HTML backend - - - - - 4e163555 by Isaac Dupree at 2009-07-21T22:03:25+00:00 add test for GADT records - - - - - 79aa4d6e by David Waern at 2009-07-23T20:40:37+00:00 Update test suite following version bump - - - - - 5932c011 by David Waern at 2009-08-02T10:25:39+00:00 Fix documentation bug - - - - - a6970fca by David Waern at 2009-08-12T23:08:53+00:00 Remove support for ghc 6.8.* from .cabal file - - - - - c1695902 by Ian Lynagh at 2009-07-07T13:35:45+00:00 Fix unused import warnings - - - - - fb6df7f9 by Ian Lynagh at 2009-07-16T00:20:31+00:00 Use cProjectVersion directly rather than going through compilerInfo Fixes the build after changes in GHC - - - - - 548cdd66 by Simon Marlow at 2009-07-28T14:27:04+00:00 follow changes in GHC's ForeignType - - - - - 9395aaa0 by David Waern at 2009-08-13T22:17:33+00:00 Switch from PatternSignatures to ScopedTypeVariables in Main - - - - - eebf39bd by David Waern at 2009-08-14T17:14:28+00:00 Version .haddock files made with GHC 6.10.3/4 correclty - - - - - 58f3e735 by David Waern at 2009-08-14T17:19:37+00:00 Support GHC 6.10.* and 6.11.* only - - - - - 5f63cecc by David Waern at 2009-08-14T22:03:20+00:00 Do not version .haddock file based on GHC patchlevel version We require that the instances of Binary that we use from GHC will not change between patchlevel versions. - - - - - d519de9f by David Waern at 2009-08-14T23:50:00+00:00 Update CHANGES - - - - - 35dccf5c by David Waern at 2009-08-14T23:51:38+00:00 Update version number everywhere - - - - - 6d363fea by David Waern at 2009-08-15T09:46:49+00:00 Update ANNOUNCE - - - - - c7ee6bc2 by David Waern at 2009-08-15T09:47:13+00:00 Remove -Werror Forgot that Hackage doesn't like it. - - - - - a125c12b by David Waern at 2009-08-15T09:49:50+00:00 Require Cabal >= 1.6 - - - - - adb2f560 by Isaac Dupree at 2009-08-12T03:47:14+00:00 Cross-Package Documentation version 4 - - - - - 3d6dc04d by David Waern at 2009-08-15T23:42:57+00:00 Put all the IN_GHC_TREE stuff inside getGhcLibDir - - - - - 56624097 by David Waern at 2009-08-15T23:52:03+00:00 Add --print-ghc-libdir - - - - - f15d3ccb by David Waern at 2009-08-16T00:37:52+00:00 Read base.haddock when running tests We can now test cross-package docs. - - - - - 283f0fb9 by David Waern at 2009-08-16T00:50:59+00:00 Update test output - we now have more links - - - - - 673d1004 by David Waern at 2009-08-16T01:26:08+00:00 Read process.haddock when running tests - - - - - 0d127f82 by David Waern at 2009-08-16T01:43:04+00:00 Add a test for cross-package documentation - - - - - f94db967 by Ian Lynagh at 2009-08-16T18:42:44+00:00 Follow GHC build system changes - - - - - 5151278a by Isaac Dupree at 2009-08-16T19:58:05+00:00 make cross-package list types look nicer - - - - - c41e8228 by Isaac Dupree at 2009-08-18T01:47:47+00:00 Haddock.Convert: export more functions This lets us remove some code in Haddock.Interface.AttachInstances - - - - - 2e5fa398 by Isaac Dupree at 2009-08-18T02:11:05+00:00 switch AttachInstances to use synify code It changed an instance from showing ((,) a b) to (a, b) because my synify code is more sophisticated; I hope the latter is a good thing rather than a bad thing aesthetically, here. But this definitely reduces code duplication! - - - - - b8b07123 by Isaac Dupree at 2009-08-18T02:23:31+00:00 Find instances using GHC, which is more complete. In particular, it works cross-package. An intermediate patch also moved the instance-finding into createInterface, but that move turned out not to be necessary, so if we want to do that, it'd go in a separate patch. (Is that possible? Or will we need GHC to have loaded all the modules first, before we can go searching for the instances (e.g. if the modules are recursive or something)?) - - - - - 6959b451 by Isaac Dupree at 2009-08-17T00:37:18+00:00 fix preprocessor conditional sense - - - - - 942823af by Isaac Dupree at 2009-08-16T22:46:48+00:00 remove ghc 6.8 conditionals from Haddock.Interface - - - - - 4b3ad888 by Isaac Dupree at 2009-08-18T20:24:38+00:00 Fix GHC 6.11 build in Haddock.Convert - - - - - 0a89c5ab by Isaac Dupree at 2009-08-23T00:08:58+00:00 hacks to make it compile without fnArgDocsn - - - - - 7b3bed43 by Isaac Dupree at 2009-08-23T03:01:28+00:00 less big-Map-based proper extraction of constructor subdocs - - - - - b21c279a by Isaac Dupree at 2009-08-23T03:02:06+00:00 Html: remove unnecessary+troublesome GHC. qualifications - - - - - 96c97115 by Isaac Dupree at 2009-08-23T03:08:03+00:00 Move doc parsing/lexing into Haddock for ghc>=6.11 - - - - - e1cec02d by Isaac Dupree at 2009-08-23T05:08:14+00:00 get rid of unused DocMap parameter in Html - - - - - 66960c59 by Isaac Dupree at 2009-08-23T05:54:20+00:00 fix horrible named-docs-disappearing bug :-) - - - - - a9d7eff3 by Isaac Dupree at 2009-08-23T06:26:36+00:00 re-implement function-argument docs ..on top of the lexParseRn work. This patch doesn't change the InstalledInterface format, and thus, it does not work cross-package, but that will be easy to add subsequently. - - - - - 8bf6852c by Isaac Dupree at 2009-08-23T07:26:05+00:00 cross-package fnArgDocs. WARNING: changes .haddock binary format While breaking the format, I took the opportunity to unrename the DocMap that's saved to disk, because there's really no reason that we want to know what *another* package's favorite place to link a Name to was. (Is that true? Or might we want to know, someday?) Also, I added instance Binary Map in InterfaceFile. It makes the code a little simpler without changing anything of substance. Also it lets us add another Map hidden inside another Map (fnArgsDocs in instDocMap) without having really-convoluted serialization code. Instances are neat! I don't understand why this change to InterfaceFile seemed to subtly break binary compatibility all by itself, but no matter, I'll just roll it into the greater format-changing patch. Done! - - - - - 30115a64 by Isaac Dupree at 2009-08-23T18:22:47+00:00 Improve behavior for unfindable .haddock - - - - - aa364bda by Isaac Dupree at 2009-08-23T18:28:16+00:00 add comment for FnArgsDoc type - - - - - 49b23a99 by Isaac Dupree at 2009-08-23T21:52:48+00:00 bugfix: restore fnArgDocs for type-synonyms - - - - - f65f9467 by Isaac Dupree at 2009-08-23T22:06:55+00:00 Backends.Hoogle: eliminate warnings - - - - - a292d216 by Isaac Dupree at 2009-08-23T22:10:24+00:00 Haddock.Convert: eliminate warnings - - - - - 5546cd20 by Isaac Dupree at 2009-08-23T22:12:31+00:00 Haddock.Interface.Rename: eliminate warnings - - - - - 0a9798b6 by Isaac Dupree at 2009-08-23T22:18:47+00:00 Main.hs: remove ghc<6.9 conditionals - - - - - e8f9867f by Isaac Dupree at 2009-08-23T22:27:46+00:00 Main.hs: eliminate warnings (except for OldException) - - - - - 61c64247 by Isaac Dupree at 2009-08-23T22:41:01+00:00 move get*LibDir code in Main.hs, to +consistent code, -duplication - - - - - 948f1e69 by Isaac Dupree at 2009-08-23T23:14:26+00:00 Main.hs: OldException->Exception: which eliminates warnings - - - - - 3d5d5e03 by Isaac Dupree at 2009-08-23T23:20:11+00:00 GhcUtils: ghc >= 6.10 - - - - - 2771d657 by Isaac Dupree at 2009-08-23T23:21:55+00:00 InterfaceFile: ghc >= 6.10 - - - - - d9f2b9d1 by Isaac Dupree at 2009-08-23T23:22:58+00:00 Types: ghc >= 6.10 - - - - - ca39210e by Isaac Dupree at 2009-08-23T23:23:26+00:00 ModuleTree: ghc >= 6.10 - - - - - 883c4e59 by Isaac Dupree at 2009-08-23T23:24:04+00:00 Backends.DevHelp: ghc >= 6.10 - - - - - 04667df5 by Isaac Dupree at 2009-08-23T23:24:37+00:00 Backends.Html: ghc >= 6.10 - - - - - a9f7f25f by Isaac Dupree at 2009-08-23T23:25:24+00:00 Utils: ghc >= 6.10 - - - - - b7105022 by Isaac Dupree at 2009-08-23T23:37:47+00:00 eliminate haskell98 dependency, following GHC's example It turns out I/we already had, and it was only a matter of deleting it from the cabal file. - - - - - 292e0911 by Isaac Dupree at 2009-08-24T01:22:44+00:00 refactor out subordinatesWithNoDocs dep of inferenced-decls fix - - - - - c2ed46a2 by Isaac Dupree at 2009-08-24T01:24:03+00:00 Eradicate wrong runtime warning for type-inferenced exported-functions see the long comment in the patch for why I did it this way :-) - - - - - 4ac0b57c by David Waern at 2009-09-04T22:56:20+00:00 Clean up tyThingToHsSynSig a little Factor out noLoc and use the case construct. Also rename the function to tyThingToLHsDecl, since it doesn't just create type signatures. - - - - - 28ab9201 by David Waern at 2009-09-04T22:58:50+00:00 Wibble - - - - - 0d9fe6d0 by David Waern at 2009-09-06T18:39:30+00:00 Add more copyright owners to H.I.AttachInstances - - - - - 122441b1 by David Waern at 2009-09-06T18:44:12+00:00 Style police - - - - - 1fa79463 by David Waern at 2009-09-06T18:57:45+00:00 Move toHsInstHead to Haddock.Convert and call it synifyInstHead - - - - - 0d42a8aa by David Waern at 2009-09-06T21:11:38+00:00 Use colordiff to display test results if available - - - - - ea9d8e03 by Simon Marlow at 2009-08-24T08:46:14+00:00 Follow changes in GHC's interface file format Word32 instead of Int for FastString and Name offsets - - - - - 537e051e by Simon Marlow at 2009-07-29T14:16:53+00:00 define unpackPackageId (it was removed from GHC) - - - - - 50c63aa7 by David Waern at 2009-09-09T23:18:03+00:00 Remove commented-out code - - - - - 511631fe by David Waern at 2009-09-09T23:19:05+00:00 Correct copyright in H.I.ParseModuleHeader - - - - - 898ec768 by David Waern at 2009-09-11T11:22:29+00:00 Use Map.fromList/toList intead of fromAscList/toAscList when serializing Maps This fixes the missing docs problem. The Eq and Ord instances for Name uses the unique number in Name. This number is created at deserialization time by GHC's magic Binary instance for Name, and it is random. Thus, fromAscList can't be used at deserialization time, even though toAscList was used at serialization time. - - - - - 37bec0d5 by Simon Peyton Jones at 2009-09-11T08:28:04+00:00 Track change in HsType - - - - - eb3a97c3 by Ian Lynagh at 2009-09-11T16:07:09+00:00 Allow building with base 4.2 - - - - - bb4205ed by Ian Lynagh at 2009-09-22T13:50:02+00:00 Loosen the GHC dependency - - - - - 5c75deb2 by Ian Lynagh at 2009-09-22T14:08:39+00:00 Fix building with GHC >= 6.12 - - - - - fb131481 by David Waern at 2009-09-11T11:24:48+00:00 Update runtests.hs to work with GHC 6.11 - - - - - ac3a419d by David Waern at 2009-09-11T11:25:14+00:00 Update CrossPackageDocs test - - - - - ec65c3c6 by David Waern at 2009-09-11T11:25:40+00:00 Add reference output for CrossPackageDocs - - - - - 520c2758 by Ian Lynagh at 2009-10-25T17:26:40+00:00 Fix installation in the GHC build system - - - - - 28b3d7df by Ian Lynagh at 2009-11-05T15:57:27+00:00 GHC build system: Make *nix installation work in paths containing spaces - - - - - 5c9bb541 by David Waern at 2009-11-14T11:56:39+00:00 Track change in HsType for the right compiler version - - - - - 905097ce by David Waern at 2009-11-14T12:10:47+00:00 hlint police - - - - - 04920630 by Ian Lynagh at 2009-11-20T13:46:30+00:00 Use defaultObjectTarget rather than HscAsm This fixes haddock when we don't have a native code generator - - - - - 966eb079 by David Waern at 2009-11-15T12:32:21+00:00 Remove commented-out code - - - - - 37f00fc4 by David Waern at 2009-11-22T13:58:48+00:00 Make runtests.hs strip links before diffing Generates easier to read diffs when tests fail. The content of the links is not important anyway since it is not taken into account by the tests. - - - - - 3a9bb8ef by David Waern at 2009-11-22T14:05:06+00:00 Follow findProgramOnPath signature change in runtests.hs - - - - - b26b9e5a by David Waern at 2009-11-22T14:08:40+00:00 Follow removal of GHC.MVar from base in CrossPackageDocs - - - - - f4d90ae4 by David Waern at 2009-11-22T14:48:47+00:00 Make copy.hs strip link contents before copying No more updating of reference files when URLs in links changes. - - - - - 4c9c420d by David Waern at 2009-11-22T15:26:41+00:00 Update test reference output * More links (Int, Float etc) * Stripped link contents - - - - - a62b80e3 by David Waern at 2009-11-23T23:19:39+00:00 Update CrossPackageDocs reference output - Remove GHC.MVar import (removed from base) - Strip link contents - - - - - 43491394 by David Waern at 2009-11-23T23:20:00+00:00 Update test reference files with comments on instances - - - - - 0d370a0b by David Waern at 2009-11-23T23:25:16+00:00 Bump version number - - - - - 2293113e by David Waern at 2009-11-24T20:55:49+00:00 Comments on instances Implementing this was a little trickier than I thought, since we need to match up instances from the renamed syntax with instances represented by InstEnv.Instance. This is due to the current design of Haddock, which matches comments with declarations from the renamed syntax, while getting the list of instances of a class/family directly using the GHC API. - Works for class instances only (Haddock has no support for type family instances yet) - The comments are rendered to the right of the instance head in the HTML output - No change to the .haddock file format - Works for normal user-written instances only. No comments are added on derived or TH-generated instances - - - - - bf586f29 by David Waern at 2009-11-27T22:05:15+00:00 Whitespace police - - - - - b8f03afa by David Waern at 2009-11-27T22:11:46+00:00 Remove bad whitespace and commented-out pieces - - - - - 90b8ee90 by David Waern at 2009-11-27T22:15:04+00:00 Whitespace police - - - - - b5ede900 by David Waern at 2009-11-27T22:15:50+00:00 Whitespace police - - - - - e3fddbfe by David Waern at 2009-11-28T13:37:59+00:00 Remove Name from DocInstance It's not used. - - - - - 9502786c by David Waern at 2009-11-28T13:56:54+00:00 Require at least GHC 6.12 While regression testing Haddock, I found a bug that happens with GHC 6.10.3, but not with GHC 6.12-rc2 (haven't tried 6.10.4). I don't have time to track it down. I think we should just always require the latest major GHC version. The time spent on making Haddock work with older versions is too high compared to the time spent on bugfixing, refactoring and features. - - - - - 8fa688d8 by David Waern at 2009-11-28T15:05:03+00:00 Remove cruft due to compatibility with older GHCs - - - - - 46fbbe9d by David Waern at 2009-11-28T15:07:50+00:00 Add a documentation header to Haddock.Convert - - - - - c3d2cc4a by David Waern at 2009-11-28T15:10:14+00:00 Remove unused H.Utils.FastMutInt2 - - - - - 490aba80 by David Waern at 2009-11-28T15:36:36+00:00 Rename Distribution.Haddock into Documentation.Haddock - - - - - 33ee2397 by David Waern at 2009-11-28T15:36:47+00:00 Fix error message - - - - - a5a3b950 by David Waern at 2009-11-28T16:58:39+00:00 Add a test flag that brings in QuickCheck - - - - - fa049e13 by David Waern at 2009-11-28T19:32:18+00:00 Say that we want quickcheck 2 - - - - - f32b0d9b by David Waern at 2009-11-28T19:32:40+00:00 Add an Arbitrary instance for HsDoc - - - - - da9a8bd7 by David Waern at 2009-11-28T20:15:30+00:00 Rename HsDoc back into Doc - - - - - edb60101 by David Waern at 2009-11-28T22:16:16+00:00 Move H.Interface.Parse/Lex to H.Parse/Lex These are not just used to build Interfaces. - - - - - 0656a9b8 by David Waern at 2009-11-28T23:12:14+00:00 Update version number in test suite - - - - - 5e8c6f4a by David Waern at 2009-12-21T14:12:41+00:00 Improve doc of DocName - - - - - 7868e551 by Ian Lynagh at 2009-09-22T10:43:03+00:00 TAG GHC 6.12-branch created - - - - - 0452a3ea by Ian Lynagh at 2009-12-15T12:46:07+00:00 TAG GHC 6.12.1 release - - - - - 65e9be62 by David Waern at 2009-12-21T16:58:58+00:00 Update CHANGES - - - - - 145cee32 by David Waern at 2009-12-21T16:59:09+00:00 TAG 2.6.0 - - - - - 3c552008 by David Waern at 2009-12-22T17:11:14+00:00 Update ANNOUNCE - - - - - 931f9db4 by David Waern at 2010-01-22T19:57:17+00:00 Convert haddock.vim to use unix newlines - - - - - 4e56588f by David Waern at 2010-01-22T22:11:17+00:00 Remove unnecessary (and inexplicable) uses of nub - - - - - 744bb4d1 by David Waern at 2010-01-22T22:12:14+00:00 Follow move of parser and lexer - - - - - e34bab14 by David Waern at 2010-01-22T22:49:13+00:00 Use findProgramLocation instead of findProgramOnPath in runtests.hs - - - - - 8d39891b by Isaac Dupree at 2010-01-14T18:53:18+00:00 fix html arg-doc off-by-one and silliness - - - - - 9401f2e9 by David Waern at 2010-01-22T22:57:03+00:00 Create a test for function argument docs - - - - - 507a82d7 by David Waern at 2010-01-22T23:24:47+00:00 Put parenthesis around type signature arguments of function type - - - - - 8a305c28 by David Waern at 2010-01-23T17:26:59+00:00 Add reference file for the FunArgs test - - - - - 1309d5e1 by David Waern at 2010-01-24T16:05:08+00:00 Improve FunArg test and update Test.html.ref - - - - - 2990f055 by Yitzchak Gale at 2010-02-14T16:03:46+00:00 Do not generate illegal character in HTML ID attribute. - - - - - c5bcab7a by David Waern at 2010-02-22T22:10:30+00:00 Fix Haddock markup error in comment - - - - - c6416a73 by David Waern at 2010-02-24T22:55:08+00:00 Large additions to the Haddock API Also improved and added more doc comments. - - - - - 57d289d7 by David Waern at 2010-02-24T22:58:02+00:00 Remove unused ifaceLocals - - - - - 80528d93 by David Waern at 2010-02-25T21:05:09+00:00 Add HaddockModInfo to the API - - - - - 82806848 by David Waern at 2010-02-25T21:05:27+00:00 Wibble - - - - - 744cad4c by David Waern at 2010-02-25T23:30:59+00:00 Make it possible to run a single test - - - - - 6a806e4c by David Waern at 2010-03-14T14:19:39+00:00 Bump version number - - - - - a5a8e4a7 by David Waern at 2010-03-14T14:36:35+00:00 Update ANNOUNCE - - - - - 6f05435e by Simon Hengel at 2010-03-15T20:52:42+00:00 Add missing dependencies for 'library' in haddock.cabal - - - - - faefe2bd by David Waern at 2010-03-15T22:29:37+00:00 Solve conflicts - - - - - 9808ad52 by David Waern at 2010-03-15T22:51:21+00:00 Bump version number - - - - - eb0bf60b by David Waern at 2010-03-15T22:52:32+00:00 Update CHANGES - - - - - f95cd891 by David Waern at 2010-03-15T23:01:06+00:00 Add Paths_haddock to other-modules of library - - - - - 65997b0a by David Waern at 2010-03-15T23:14:59+00:00 Update CHANGES - - - - - 7e251731 by David Waern at 2010-03-15T23:15:30+00:00 Bump version number - - - - - c9cd0ddc by David Waern at 2010-03-16T00:28:34+00:00 Fix warning - - - - - 1cac2d93 by Simon Peyton Jones at 2010-01-04T15:22:16+00:00 Fix imports for new location of splitKindFunTys - - - - - 474f26f6 by Simon Peyton Jones at 2010-02-10T14:36:06+00:00 Update Haddock for quasiquotes - - - - - 0dcc06c0 by Simon Peyton Jones at 2010-02-10T10:59:45+00:00 Track changes in HsTyVarBndr - - - - - 2d84733a by Simon Peyton Jones at 2010-02-10T14:52:44+00:00 Track HsSyn chnages - - - - - 9e3adb8b by Ian Lynagh at 2010-02-20T17:09:42+00:00 Resolve conflicts - - - - - a3e72ff8 by Simon Peyton Jones at 2010-03-04T13:05:16+00:00 Track change in HsUtils; and use a nicer function not an internal one - - - - - 27994854 by David Waern at 2010-03-18T22:22:27+00:00 Fix build with GHC 6.12.1 - - - - - 11f6e488 by David Waern at 2010-03-18T22:24:09+00:00 Bump version in test reference files - - - - - 0ef2f11b by David Waern at 2010-03-20T00:56:30+00:00 Fix library part of cabal file when in ghc tree - - - - - 3f6146ff by Mark Lentczner at 2010-03-20T22:30:11+00:00 First, experimental XHTML rendering switch to using the xhtml package copied Html.hs to Xhtml.hs and split into sub-modules under Haddock/Backends/Xhtml and detabify moved footer into div, got ready for iface change headers converted to semantic markup contents in semantic markup summary as semantic markup description in semantic markup, info block in header fixed factored out rendering so during debug it can be readable (see renderToString) - - - - - b8ab329b by Mark Lentczner at 2010-03-20T22:54:01+00:00 apply changes to Html.hs to Xhtml/*.hs incorporate changes that were made between the time Html.hs was copied and split into Xhtml.hs and Xhtml/*.hs includes patchs after "Wibble" (!) through "Fix build with GHC 6.12.1" - - - - - 73df2433 by Ian Lynagh at 2010-03-20T21:56:37+00:00 Follow LazyUniqFM->UniqFM in GHC - - - - - db4f602b by David Waern at 2010-03-29T22:00:01+00:00 Fix build with GHC 6.12 - - - - - d8dca088 by Simon Hengel at 2010-04-02T16:39:55+00:00 Add missing dependencies to cabal file - - - - - e2adc437 by Simon Hengel at 2010-04-02T14:08:40+00:00 Add markup support for interactive examples - - - - - e882ac05 by Simon Hengel at 2010-04-02T14:11:53+00:00 Add tests for interactive examples - - - - - 5a07a6d3 by David Waern at 2010-04-07T17:05:20+00:00 Propagate source positions from Lex.x to Parse.y - - - - - 6493b46f by David Waern at 2010-04-07T21:48:57+00:00 Let runtests.hs die when haddock has not been built - - - - - 5e34423e by David Waern at 2010-04-07T22:01:13+00:00 Make runtests.hs slightly more readable - - - - - 321d59b3 by David Waern at 2010-04-07T22:13:27+00:00 Fix haskell/haddock#75 Add colons to the $ident character set. - - - - - 37b08b8d by David Waern at 2010-04-08T00:32:52+00:00 Fix haskell/haddock#118 Avoid being too greedy when lexing URL markup (<..>), in order to allow multiple URLs on the same line. Do the same thing with <<..>> and #..#. - - - - - df8feac9 by David Waern at 2010-04-08T00:57:33+00:00 Make it easier to add new package deps to test suite This is a hack - we should use Cabal to get the package details instead. - - - - - 1ca6f84b by David Waern at 2010-04-08T01:03:06+00:00 Add ghc-prim to test suite deps - - - - - 27371e3a by Simon Hengel at 2010-04-08T19:26:34+00:00 Let parsing fails on paragraphs that are immediately followed by an example This is more consistent with the way we treat code blocks. - - - - - 83096e4a by David Waern at 2010-04-08T21:20:00+00:00 Improve function name - - - - - 439983ce by David Waern at 2010-04-10T10:46:14+00:00 Fix haskell/haddock#112 No link was generated for 'Addr#' in a doc comment. The reason was simply that the identifier didn't parse. We were using parseIdentifier from the GHC API, with a parser state built from 'defaultDynFlags'. If we pass the dynflags of the module instead, the right options are turned on on while parsing the identifer (in this case -XMagicHash), and the parse succeeds. - - - - - 5c0d35d7 by David Waern at 2010-04-10T10:54:06+00:00 Rename startGhc into withGhc - - - - - dca081fa by Simon Hengel at 2010-04-12T19:09:16+00:00 Add documentation for interactive examples - - - - - c7f26bfa by David Waern at 2010-04-13T00:51:51+00:00 Slight fix to the documentation of examples - - - - - 06eb7c4c by David Waern at 2010-04-13T00:57:05+00:00 Rename Interactive Examples into Examples (and simplify explanation) - - - - - 264830cb by David Waern at 2010-05-10T20:07:27+00:00 Update CHANGES with info about 2.6.1 - - - - - 8e5d4514 by Simon Hengel at 2010-04-18T18:16:54+00:00 Add unit tests for parser - - - - - 68297f40 by David Waern at 2010-05-10T21:53:37+00:00 Improve testsuite README - - - - - f04eb6e4 by David Waern at 2010-05-11T19:14:31+00:00 Re-organise the testsuite structure - - - - - a360f710 by David Waern at 2010-05-11T19:18:03+00:00 Shorten function name - - - - - 1d5dd359 by David Waern at 2010-05-11T21:40:02+00:00 Update runtests.hs following testsuite re-organisation - - - - - ffebe217 by David Waern at 2010-05-11T21:40:10+00:00 Update runtests.hs to use base-4.2.0.1 - - - - - 635de402 by David Waern at 2010-05-11T21:41:11+00:00 Update runparsetests.hs following testsuite reorganisation - - - - - 72137910 by Ian Lynagh at 2010-05-06T20:43:06+00:00 Fix build - - - - - 1a80b76e by Ian Lynagh at 2010-05-06T22:25:29+00:00 Remove redundant import - - - - - 1031a80c by Simon Peyton Jones at 2010-05-07T13:21:09+00:00 Minor wibbles to HsBang stuff - - - - - dd8e7fe5 by Ian Lynagh at 2010-05-08T15:22:00+00:00 GHC build system: Follow "rm" variable changes - - - - - 7f5e6748 by David Waern at 2010-05-13T11:53:02+00:00 Fix build with GHC 6.12.2 - - - - - 7953d4d8 by David Waern at 2010-05-13T18:45:01+00:00 Fixes to comments only - - - - - 8ae8eb64 by David Waern at 2010-05-13T18:57:26+00:00 ModuleMap -> IfaceMap - - - - - 1c3eadc6 by David Waern at 2010-05-13T19:03:13+00:00 Fix whitespace style issues - - - - - e96783c0 by David Waern at 2010-05-13T19:08:53+00:00 Fix comment - - - - - c998a78b by David Waern at 2010-05-13T19:39:00+00:00 Position the module header the same way everywhere Silly, but nice with some consistency :-) - - - - - b48a714e by David Waern at 2010-05-13T19:41:32+00:00 Position of module header, this time in the HTML backends - - - - - f9bfb12e by David Waern at 2010-05-13T19:43:05+00:00 Two newlines between declarations in Main - - - - - 071d44c7 by David Waern at 2010-05-13T19:44:21+00:00 Newlines in Convert - - - - - 036346db by David Waern at 2010-05-13T19:46:47+00:00 Fix a few stylistic issues in H.InterfaceFile - - - - - f0b8379e by David Waern at 2010-05-13T19:47:53+00:00 Add newlines to H.ModuleTree - - - - - 27409f8e by David Waern at 2010-05-13T19:51:10+00:00 Fix stylistic issues in H.Utils - - - - - 24774a11 by David Waern at 2010-05-13T20:00:43+00:00 Structure H.Types better - - - - - 7b6f5e40 by David Waern at 2010-05-13T20:01:04+00:00 Remove bad Arbitrary instance - - - - - fac9f1f6 by David Waern at 2010-05-13T20:05:50+00:00 Get rid of H.Utils.pathJoin and use System.FilePath.joinPath instead - - - - - fe6d00c4 by David Waern at 2010-05-13T20:51:55+00:00 Export a couple of more types from the API - - - - - b2e33a5f by David Waern at 2010-05-13T21:27:51+00:00 Improve doc comment for Interface - - - - - c585f2ce by David Waern at 2010-05-13T21:30:14+00:00 Improve documentation of Haddock.Interface - - - - - e6791db2 by David Waern at 2010-05-13T22:07:35+00:00 Remove meaningless comments - - - - - 7801b390 by David Waern at 2010-05-14T17:53:33+00:00 Remove unused modules - - - - - f813e937 by David Waern at 2010-05-14T17:55:17+00:00 Re-direct compilation output to a temporary directory Also add a flag --no-tmp-comp-dir that can be used to get the old behaviour of writing compilation files to GHC's output directory (default "."). - - - - - e56737ec by David Waern at 2010-05-14T18:06:11+00:00 Wibble - - - - - e40b0447 by David Waern at 2010-05-14T19:01:52+00:00 Move flag evaluation code from Main to Haddock.Options Determining the value of "singular" flags (by e.g. taking the last occurrence of the flag) and other flag evaluation should done in Haddock.Options which is the module that is supposed to define the command line interface. This makes Main a bit easier on the eyes as well. - - - - - 27091f57 by David Waern at 2010-05-14T19:05:10+00:00 Wibble - - - - - c658cf61 by David Waern at 2010-05-14T19:06:49+00:00 Re-order things in Haddock.Options a bit - - - - - 8cfdd342 by David Waern at 2010-05-14T19:20:29+00:00 De-tabify Haddock.Options and fix other whitespace issues - - - - - 0df16b62 by David Waern at 2010-05-14T19:25:07+00:00 Improve comments - - - - - 80b38e2b by David Waern at 2010-05-14T19:26:42+00:00 Whitespace police - - - - - fe580255 by David Waern at 2010-05-14T19:31:23+00:00 Wibbles to comments - - - - - a2b43fad by David Waern at 2010-05-14T20:24:32+00:00 Move some more flag functions to Haddock.Options - - - - - 3f895547 by David Waern at 2010-05-14T20:37:12+00:00 Make renderStep a top-level function in Main - - - - - 5cdca11d by David Waern at 2010-05-14T20:39:27+00:00 Spelling in comment - - - - - ad98d14c by David Waern at 2010-05-14T20:40:26+00:00 Comment fixes - - - - - 0bb9218f by David Waern at 2010-05-14T20:49:01+00:00 Whitespace police - - - - - 0f0a533f by David Waern at 2010-05-15T16:42:29+00:00 Improve description of --dump-interface - - - - - 5b2833ac by David Waern at 2010-05-15T17:16:53+00:00 Document --no-tmp-comp-dir - - - - - 8160b170 by David Waern at 2010-05-15T17:18:59+00:00 Wibble - - - - - 570dbe33 by David Waern at 2010-05-18T21:15:38+00:00 HLint police - - - - - 204e425f by David Waern at 2010-05-18T21:16:30+00:00 HLint police - - - - - 6db657ac by David Waern at 2010-05-18T21:16:37+00:00 Wibble - - - - - b942ccd7 by Simon Marlow at 2010-06-02T08:27:30+00:00 Interrupted disappeared in GHC 6.13 (GHC ticket haskell/haddock#4100) - - - - - 3b94a819 by Simon Marlow at 2010-06-02T08:45:08+00:00 Allow base-4.3 - - - - - c5a1fb7c by Simon Marlow at 2010-06-02T09:03:04+00:00 Fix compilation with GHC 6.13 - - - - - 6181296c by David Waern at 2010-06-08T21:09:05+00:00 Display name of prologue file when parsing it fails - - - - - 7cbc6f60 by Ian Lynagh at 2010-06-13T16:20:25+00:00 Remove redundant imports - - - - - 980c804b by Simon Marlow at 2010-06-22T08:41:50+00:00 isLocalAndTypeInferenced: fix for local module names overlapping package modules - - - - - d74d4a12 by Simon Marlow at 2010-06-23T12:03:27+00:00 Unresolved identifiers in Doc get replaced with DocMonospaced rather than plain strings - - - - - d8546783 by Simon Marlow at 2010-06-30T12:45:17+00:00 LaTeX backend (new options: --latex, --latex-style=<style>) - - - - - 437afa9e by David Waern at 2010-07-01T12:02:44+00:00 Fix a few stylistic whitespace issues in LaTeX backend - - - - - 85bc1fae by David Waern at 2010-07-01T15:42:45+00:00 Make runtest.hs work with GHC 6.12.3 (we should really stop hard coding this) - - - - - 7d2eb86f by David Waern at 2010-07-01T15:43:33+00:00 Update test following Simon's patch to render unresolved names in monospaced font - - - - - 08fcbcd2 by David Waern at 2010-07-01T16:12:18+00:00 Warning police - - - - - d04a8d7a by David Waern at 2010-07-04T14:53:39+00:00 Fix a bug in attachInstances We didn't look for instance docs in all the interfaces of the package. This had the effect of instance docs not always showing up under a declaration. I took the opportunity to clean up the code in H.I.AttachInstances a bit as well. More cleanup is needed, however. - - - - - d10344eb by Simon Hengel at 2010-07-10T09:19:04+00:00 Add missing dependencies to cabal file - - - - - 24090531 by Mark Lentczner at 2010-03-21T04:51:16+00:00 add exports to Xhtml modules - - - - - 84f9a333 by Mark Lentczner at 2010-04-03T19:14:22+00:00 clean up Doc formatting code - add CSS for lists - renderToString now uses showHtml since prettyHtml messes up <pre> sections - - - - - bebccf52 by Mark Lentczner at 2010-04-04T04:51:08+00:00 tweak list css - - - - - 0c2aeb5e by Mark Lentczner at 2010-04-04T06:24:14+00:00 all decls now generate Html not HtmlTable - ppDecl return Html, and so now do all of the functions it calls - added some internal tables to some decls, which is wrong, and will have to be fixed - decl "Box" functions became "Elem" functions to make clear they aren't in a table anymore (see Layout.hs) - docBox went away, as only used in one place (and its days are numbered) - cleaned up logic in a number of places, removed dead code - added maybeDocToHtml which simplified a number of places in the code - - - - - dbf73e6e by Mark Lentczner at 2010-04-05T05:02:43+00:00 clean up processExport and place a div around each decl - - - - - e25b7e9f by Mark Lentczner at 2010-04-10T21:23:21+00:00 data decls are now a sequence of paragraphs, not a table - - - - - 89ee0294 by Mark Lentczner at 2010-04-10T21:29:16+00:00 removed commented out code that can't be maintained - - - - - d466f536 by Mark Lentczner at 2010-04-12T04:56:27+00:00 removed declWithDoc and cleaned up data decls in summary - - - - - ed755832 by Mark Lentczner at 2010-04-12T05:07:53+00:00 merge in markupExample changes - - - - - c36f51fd by Mark Lentczner at 2010-04-25T04:56:37+00:00 made record fields be an unordList, not a table - - - - - ed3a28d6 by Mark Lentczner at 2010-04-25T05:23:28+00:00 fixed surround of instance and constructor tables - - - - - 0e35bbc4 by Mark Lentczner at 2010-04-25T05:36:59+00:00 fix class member boxes in summary - - - - - 5041749b by Mark Lentczner at 2010-04-25T05:38:35+00:00 remove unused bodyBox - - - - - e91724db by Mark Lentczner at 2010-04-25T06:26:10+00:00 fixed javascript quoting/escpaing issue - - - - - f4abbb73 by Mark Lentczner at 2010-05-03T23:04:31+00:00 adjust css for current markup - - - - - e75fec4c by Mark Lentczner at 2010-05-04T06:14:34+00:00 added assoicated types and methods back into class decls - - - - - 84169323 by Mark Lentczner at 2010-05-24T13:13:42+00:00 merge in changes from the big-whitespace cleanup - - - - - 3c1c872e by Mark Lentczner at 2010-06-11T21:03:58+00:00 adjust synopsis and bottom bar spacing - - - - - 3c1f9ef7 by Mark Lentczner at 2010-06-11T21:14:44+00:00 fix missing space in "module" lines in synoposis - - - - - 9a137e6d by Mark Lentczner at 2010-06-11T21:34:08+00:00 changed tt elements to code elements - - - - - 50f71ef1 by Mark Lentczner at 2010-06-11T23:27:46+00:00 factored out ppInstances - - - - - 3b9a9de5 by Mark Lentczner at 2010-06-17T17:36:01+00:00 push single constructors (newtype) onto line with decl - - - - - e0f8f2ec by Mark Lentczner at 2010-06-17T22:20:56+00:00 remove <++> connector - - - - - 56c075dd by Mark Lentczner at 2010-07-13T05:26:21+00:00 change to new page structure - - - - - 04be6ca7 by Mark Lentczner at 2010-07-14T04:21:55+00:00 constructors and args as dl lists, built in Layout.hs - - - - - 65aeafc2 by Mark Lentczner at 2010-07-14T05:38:32+00:00 better interface to subDecls - - - - - 72032189 by Mark Lentczner at 2010-07-14T07:04:10+00:00 made subDecl tables looks just so - - - - - b782eca2 by Mark Lentczner at 2010-07-14T16:00:54+00:00 convert args to SubDecl format - - - - - cc75e98f by Mark Lentczner at 2010-07-14T16:28:53+00:00 convert instances to SubDecl - - - - - 34e2aa5a by Mark Lentczner at 2010-07-14T21:07:32+00:00 removing old table cruft from Layout.hs - - - - - d5810d95 by Mark Lentczner at 2010-07-14T21:54:58+00:00 methods and associated types in new layout scheme - - - - - 65ef9579 by Mark Lentczner at 2010-07-14T23:43:42+00:00 clean up synopsis lists - - - - - e523318f by Mark Lentczner at 2010-07-15T05:02:26+00:00 clean up of anchors - - - - - 1215dfc5 by Mark Lentczner at 2010-07-15T23:53:01+00:00 added two new themes and rough css switcher - - - - - 7f0fd36f by Mark Lentczner at 2010-07-16T04:57:38+00:00 fixed package catpion, added style menu - - - - - 0dd4999c by Mark Lentczner at 2010-07-16T20:12:39+00:00 new output for mini_ pages - - - - - 64b2810b by Mark Lentczner at 2010-07-16T20:58:41+00:00 reformat index-frames - - - - - 3173f555 by Mark Lentczner at 2010-07-16T22:41:53+00:00 convert index to new markup - - - - - b0a4b7c9 by Mark Lentczner at 2010-07-17T04:07:22+00:00 convert index.html to new markup, adjust module markup - - - - - 8261ae1e by Mark Lentczner at 2010-07-17T05:07:29+00:00 classing styling of ancillary pages - - - - - 2a4fb025 by Mark Lentczner at 2010-07-17T05:11:45+00:00 clean up Layout.hs: no more vanillaTable - - - - - 87eec685 by Mark Lentczner at 2010-07-17T05:35:16+00:00 clean up Util.hs - - - - - d304e9b0 by Mark Lentczner at 2010-07-17T05:38:50+00:00 qualify import of XHtml as XHtml - - - - - 7dc05807 by Mark Lentczner at 2010-07-17T06:17:53+00:00 factored out head element generation - - - - - 9cdaec9e by Mark Lentczner at 2010-07-17T06:44:54+00:00 refactored out main page body generation - - - - - 8a51019e by Mark Lentczner at 2010-07-17T06:48:20+00:00 moved footer into only place that used it - - - - - efa479da by Mark Lentczner at 2010-07-17T18:48:30+00:00 styling auxillary pages for tibbe and snappy themes - - - - - 81de5509 by Mark Lentczner at 2010-07-18T04:41:38+00:00 fixed alphabet on index page, and styling of it and packages in module lists - - - - - 20718c1a by Mark Lentczner at 2010-07-18T05:34:29+00:00 cleaned up div functions in Layout.hs - - - - - 60d50453 by Mark Lentczner at 2010-07-18T05:48:39+00:00 added content div to main pages - - - - - ed16561c by Mark Lentczner at 2010-07-18T06:12:22+00:00 add .doc class to documentation blocks - - - - - f5c781b0 by Mark Lentczner at 2010-07-19T05:20:53+00:00 refactoring of anchor ID and fragment handling - - - - - a69a93bf by Mark Lentczner at 2010-07-19T05:35:55+00:00 remove an explicit bold tag - replace with .def class - - - - - d76c7225 by Mark Lentczner at 2010-07-19T06:56:15+00:00 rename Haddock.Backends.Xhtml.Util to Utils - - - - - 5a58c0da by David Waern at 2010-07-21T13:30:54+00:00 Remove trailing whitespace in Haddock.Backends.Xhtml - - - - - 0652aa17 by David Waern at 2010-07-21T13:33:21+00:00 Align a few comments - - - - - 785776c3 by David Waern at 2010-07-21T13:39:04+00:00 Remove trailing whitespace in H.B.X.Decl - - - - - 71a30710 by David Waern at 2010-07-21T13:44:27+00:00 Remove more trailing whitespace - - - - - 38750394 by David Waern at 2010-07-21T13:50:43+00:00 Style police - - - - - 3023d940 by David Waern at 2010-07-21T14:01:22+00:00 Style police in H.B.X.Decl - - - - - df16e9e6 by David Waern at 2010-07-21T14:14:45+00:00 Style police in H.B.X.DocMarkup - - - - - 6020e321 by David Waern at 2010-07-21T14:17:32+00:00 More style police - - - - - 86ad8bf5 by David Waern at 2010-07-21T14:21:02+00:00 Style police in H.B.Xhtml - - - - - aea27d03 by David Waern at 2010-07-21T14:42:03+00:00 Fix warnings in LaTeX backend - - - - - 2aff34a9 by David Waern at 2010-07-21T14:50:46+00:00 Style police in LaTeX backend (mainly more newlines) - - - - - e517162d by David Waern at 2010-07-21T15:05:47+00:00 Doc sections in Main - - - - - b971aa0c by David Waern at 2010-07-21T15:06:17+00:00 Trailing whitespace in Documentation.Haddock - - - - - f11628fb by David Waern at 2010-07-21T15:07:06+00:00 Trailing whitespace in Haddock.Convert - - - - - cbaf284c by David Waern at 2010-07-21T15:08:11+00:00 Style police in Haddock.GhcUtils - - - - - 71feb77b by David Waern at 2010-07-21T15:09:06+00:00 Style police in Haddock.InterfaceFile - - - - - 0a9c80e6 by David Waern at 2010-07-21T15:11:33+00:00 Whitespace police - - - - - 6168376c by David Waern at 2010-07-21T15:16:35+00:00 Style police in Haddock.Utils - - - - - 9fe4dd90 by David Waern at 2010-07-21T15:19:31+00:00 Add -fwarn-tabs - - - - - a000d752 by Mark Lentczner at 2010-07-20T17:25:52+00:00 move CSS Theme functions into Themes.hs - - - - - b52b440f by Mark Lentczner at 2010-07-20T17:29:35+00:00 add Thomas Schilling's theme - - - - - e43fa7e8 by Mark Lentczner at 2010-07-21T04:49:34+00:00 correct icon used with Snappy theme - - - - - ba5092d3 by Mark Lentczner at 2010-07-21T04:56:47+00:00 apply Tibbe's updates to his theme - - - - - 7804eef6 by Mark Lentczner at 2010-07-21T05:15:49+00:00 space between "Style" and the downward triangle - - - - - 7131d4c6 by Mark Lentczner at 2010-07-21T17:43:35+00:00 merge with David's source cleanups - - - - - ee65f1cb by David Waern at 2010-07-22T16:50:46+00:00 Fix a bug where we allowed --hoogle, --latex, etc without input files - - - - - e413ff7a by David Waern at 2010-07-22T17:21:58+00:00 Improve function name - - - - - a0fd14f3 by Simon Marlow at 2010-06-30T15:34:32+00:00 fix warnings - - - - - 31f73d2a by David Waern at 2010-07-22T19:29:41+00:00 Solve conflicts - - - - - d563b4a5 by Simon Marlow at 2010-06-30T15:34:37+00:00 fix warning - - - - - 412b6469 by David Waern at 2010-07-22T19:31:28+00:00 Solve conflict - - - - - 35174b94 by Ian Lynagh at 2010-07-06T17:27:16+00:00 Follow mkPState argument order change - - - - - b5c3585c by Simon Marlow at 2010-07-14T08:49:21+00:00 common up code for instance rendering - - - - - d8009560 by Simon Marlow at 2010-07-14T12:37:11+00:00 fix warnings - - - - - a6d88695 by David Waern at 2010-07-24T15:33:33+00:00 Fix build with ghc < 6.13 - - - - - 94cf9de1 by David Waern at 2010-07-24T15:34:37+00:00 Remove conflict left-over - - - - - 313b15c0 by Mark Lentczner at 2010-07-21T22:09:04+00:00 reorganization of nhaddock.css with tibbe - - - - - 9defed80 by Mark Lentczner at 2010-07-21T22:42:14+00:00 further cleanup of nhaddock.css, float TOC, support aux. pages - - - - - 6d944c1b by Mark Lentczner at 2010-07-22T06:22:23+00:00 remove old HTML backend - - - - - b3e8cba5 by Mark Lentczner at 2010-07-22T06:43:32+00:00 remove --html-help support - it was old, out-of-date, and mostly missing - - - - - d2654a08 by Mark Lentczner at 2010-07-22T21:45:34+00:00 tweaks to nhaddock.css - - - - - f73b285c by Mark Lentczner at 2010-07-23T06:19:35+00:00 command like processing for theme selection The bulk of the change is threadnig the selected theme set through functions in Xhtml.hs so that the selected themes can be used when generating the page output. There isn't much going on in most of these changes, just passing it along. The real work is all done in Themes.hs. - - - - - 8bddc90d by Mark Lentczner at 2010-07-23T06:58:31+00:00 drop --themes support, add named theme support decided that --themes was silly - no one would do that, just use multiple --theme arguments made --theme a synonym for --css and -c made those arguments, if no file is found, look up the argument as the name of a built in theme all of this let's haddock be invoked with "--theme=classic" for example. - - - - - 20cafd4f by Mark Lentczner at 2010-07-23T17:44:29+00:00 rename --default-themes to --built-in-themes - - - - - 0fe41307 by Mark Lentczner at 2010-07-23T18:33:02+00:00 tweaks to theme for info table, headings, and tables - - - - - cba4fee0 by Mark Lentczner at 2010-07-23T19:13:59+00:00 tweaks for dl layout, though still not used - - - - - 463fa294 by Mark Lentczner at 2010-07-23T21:07:19+00:00 tweak look of mini pages, keywords, and preblocks - - - - - 5472fc02 by Mark Lentczner at 2010-07-24T05:36:15+00:00 slide out Synopsis drawer - - - - - 9d5d5de5 by Mark Lentczner at 2010-07-24T06:02:42+00:00 extend package header and footer to edges of page - - - - - a47c91a2 by Mark Lentczner at 2010-07-24T06:28:44+00:00 fields are def lists, tweak css for style menu, mini pages, arguments - - - - - ca20f23b by Mark Lentczner at 2010-07-24T16:55:22+00:00 excisting last vestiges of the --xhtml flag - - - - - 71fb012e by Mark Lentczner at 2010-07-25T18:47:49+00:00 change how collapsing sections are done make whole .caption be the target improve javascript for class toggling have plus/minus images come from .css, not img tags - - - - - c168c8d3 by Mark Lentczner at 2010-07-26T00:32:05+00:00 reorganize files in the html lib data dir - - - - - 93324301 by Mark Lentczner at 2010-07-26T01:27:42+00:00 cleaned up Themes.hs - - - - - ad3b5dd4 by Mark Lentczner at 2010-07-26T02:39:15+00:00 make module list use new collapsers - - - - - 1df9bfc6 by Mark Lentczner at 2010-07-27T19:09:25+00:00 remove Tibbe theme - - - - - 8b9b01b3 by Mark Lentczner at 2010-07-27T20:04:03+00:00 move themes into html dir with .theme and .std-theme extensions - - - - - a7beb965 by Mark Lentczner at 2010-07-27T21:06:34+00:00 give a class to empty dd elements so they can be hidden - - - - - a258c117 by Mark Lentczner at 2010-07-27T21:23:58+00:00 remove custom version of copyFile in Xhtml.hs - - - - - b70dba6e by Mark Lentczner at 2010-07-27T22:12:45+00:00 apply margin changes to pre and headings as per group decision, and small cleanups - - - - - e6f722a2 by Mark Lentczner at 2010-07-28T00:03:12+00:00 make info block and package bar links be floatable by placing them first in the dom tree - - - - - c8278867 by Mark Lentczner at 2010-07-28T19:01:18+00:00 styling source links on declarations - - - - - 88fdc399 by Mark Lentczner at 2010-07-29T01:12:46+00:00 styling tweaks don't generate an empty li for absent style menu in links area update css for Classic and Snappy to handle: dl lists links in package header and in declarations floating of links and info block in package and module headers - - - - - 8a75b213 by Ian Lynagh at 2010-07-30T20:21:46+00:00 Fix build in GHC tree - - - - - ce8e18b3 by Simon Hengel at 2010-08-03T18:37:26+00:00 Adapt paths to data files in cabal file - - - - - 9701a455 by Simon Hengel at 2010-08-07T13:20:27+00:00 Add missing dependency to cabal file - - - - - 01b838d1 by Mark Lentczner at 2010-07-30T20:19:40+00:00 improved synopsis drawer: on click, not hover - - - - - 7b6f3e59 by Mark Lentczner at 2010-07-30T23:38:55+00:00 put the synopsis back in the other themes - - - - - 7b2904c9 by Mark Lentczner at 2010-08-11T11:11:26+00:00 close arrows on expanded synopsis drawer - - - - - ea19e177 by Mark Lentczner at 2010-08-12T21:16:45+00:00 width and font changes removed the max width restrictions on the page as a whole and the synopsis made the main font size smaller (nominally 14pt) and then tweaked most font sizes (relative) to be more consistent - - - - - 5ced00c0 by Mark Lentczner at 2010-08-13T15:09:55+00:00 implemented YUI's CSS font approach - - - - - 2799c548 by Mark Lentczner at 2010-08-13T15:11:59+00:00 adjusted margin to 2em, 1 wasn't enough - - - - - 58f06893 by Mark Lentczner at 2010-08-13T15:48:44+00:00 removed underlining on hover for named anchors headings in interface lost thier a element, no need, just put id on heading css for a elements now only applies to those with href attribute - - - - - 7aced4c4 by Mark Lentczner at 2010-08-13T15:50:22+00:00 more space between elements - - - - - 5a3c1cce by Mark Lentczner at 2010-08-13T16:43:43+00:00 adjusted font sizes of auxilary pages per new scheme - - - - - 487539ef by Mark Lentczner at 2010-08-13T21:43:41+00:00 add Frames button and clean up frames.html - - - - - c1a140b6 by Mark Lentczner at 2010-08-13T22:17:48+00:00 move frames button to js - - - - - b0bdb68e by Mark Lentczner at 2010-08-14T03:44:46+00:00 build style menu in javascript moved to javascript, so as to not polute the content with the style menu removed menu building code in Themes.hs removed onclick in Utils.hs changed text of button in header from "Source code" to "Source" more consistent with links in rest of page - - - - - 43ab7120 by Mark Lentczner at 2010-08-16T15:15:37+00:00 font size and margin tweaks - - - - - c0b68652 by Mark Lentczner at 2010-08-17T18:19:52+00:00 clean up collapser logics javascript code for collapasble sections cleaned up rewrote class utilities in javascript to be more robust refactored utilities for generating collapsable sections made toc be same color as synopsis module list has needed clear attribute in CSS - - - - - 5d573427 by Mark Lentczner at 2010-08-17T23:06:02+00:00 don't collapse entries in module list when clicking on links - - - - - 8c307c4a by Mark Lentczner at 2010-08-17T23:21:43+00:00 add missing data file to .cabal - - - - - 414bcfcf by Mark Lentczner at 2010-08-17T23:28:47+00:00 remove synopsis when in frames - - - - - ba0fa98a by Mark Lentczner at 2010-08-18T16:16:11+00:00 layout tweeks - mini page font size, toc color, etc. - - - - - 63c1bed1 by Mark Lentczner at 2010-08-18T19:50:02+00:00 margin fiddling - - - - - c311c094 by Mark Lentczner at 2010-08-20T01:37:55+00:00 better synopsis handling logic - no flashing - - - - - f1fe5fa8 by Mark Lentczner at 2010-08-20T01:41:06+00:00 fix small layout issues mini frames should have same size top heading give info block dts some padding so they don't collide in some browsers - - - - - 0de84d77 by Mark Lentczner at 2010-08-20T02:13:09+00:00 made style changing and cookies storage robust - - - - - 1ef064f9 by Thomas Schilling at 2010-08-04T13:12:22+00:00 Make synopsis frame behave properly in Firefox. In Firefox, pressing the back button first reverted the synopsis frame, and only clicking the back button a second time would update the main frame. - - - - - dd1c9a94 by Mark Lentczner at 2010-08-21T01:46:19+00:00 remove Snappy theme - - - - - 2353a90d by Mark Lentczner at 2010-08-25T05:16:19+00:00 fix occasional v.scroll bars on pre blocks (I think) - - - - - 459b8bf1 by Simon Hengel at 2010-08-08T10:12:45+00:00 Add createInterfaces' (a more high-level alternative to createInterfaces) to Haddock API - - - - - b1b68675 by David Waern at 2010-08-26T20:31:58+00:00 Follow recent API additions with some refactorings Simon Hegel's patch prompted me to do some refactorings in Main, Haddock.Documentation and Haddock.Interface. - - - - - 264d4d67 by David Waern at 2010-08-26T21:40:59+00:00 Get rid of GhcModule and related cruft We can get everything we need directly from TypecheckedModule. - - - - - 0feacec2 by Mark Lentczner at 2010-08-26T23:44:13+00:00 fixed CSS for ordered lists and def lists in doc blocks - - - - - 2997e0c2 by Mark Lentczner at 2010-08-26T23:45:03+00:00 support both kinds of enumerated lists in doc markup The documentation for Haddock says enumerated lists can use either of (1) first item 2. second item The second form wasn't actually supported - - - - - 5d4ddeec by Mark Lentczner at 2010-08-27T21:29:48+00:00 fix broken header link margins - - - - - 614456ba by Mark Lentczner at 2010-08-27T22:16:19+00:00 fix table of contents CSS - - - - - 03f329a2 by David Waern at 2010-08-28T16:36:09+00:00 Update tests following switch to the Xhtml backend - - - - - ca689fa2 by Mark Lentczner at 2010-08-28T18:25:16+00:00 fix def lists - - - - - 18e1d3d2 by Mark Lentczner at 2010-08-28T18:26:18+00:00 push footer to bottom of window - - - - - b0ab8d82 by David Waern at 2010-08-28T22:04:32+00:00 Whitespace police - - - - - 2d217977 by David Waern at 2010-08-29T12:44:45+00:00 Remove Snappy data files - - - - - 01e27d5f by David Waern at 2010-08-29T13:03:28+00:00 Add source entity path to --read-interface You can now use this flag like this: --read-interface=<html path>,<source entity path>,<.haddock file> By "source entity path" I mean the same thing that is specified with the --source-entity flag. The purpose of this is to be able to specify the source entity path per package, to allow source links to work in the presence of cross-package documentation. When given two arguments or less the --read-interface flag behaves as before. - - - - - 20bf4aaa by David Waern at 2010-08-29T13:11:03+00:00 Naming wibbles - - - - - ad22463f by Mark Lentczner at 2010-08-29T15:14:54+00:00 make portability block be a table - solves layout issues - - - - - 97bd1ae6 by Mark Lentczner at 2010-08-29T15:17:42+00:00 update golden test for Test due to portability box change - - - - - d37e139e by Mark Lentczner at 2010-08-29T17:07:17+00:00 move TOC and Info blocks down 0.5em to improve layout issue w/Test.hs - - - - - acf52501 by David Waern at 2010-08-29T17:32:36+00:00 Allow building with ghc < 6.16 - - - - - 1cb34ed8 by Ian Lynagh at 2010-07-24T23:18:49+00:00 Flatten the dynflags before parsing - - - - - b36845b4 by Ian Lynagh at 2010-07-24T23:26:49+00:00 Follow flattenLanguageFlags -> flattenExtensionFlags rename - - - - - 7f7fcc7e by David Waern at 2010-08-29T17:46:23+00:00 Use flattenExtensionFlags with ghc >= 6.13 only - - - - - 13cf9411 by Ian Lynagh at 2010-08-01T18:09:54+00:00 Make the main haddock script versioned, and make plain "haddock" a symlink - - - - - 495cbff2 by Ian Lynagh at 2010-08-18T18:57:24+00:00 Fix installation in the GHC build system Data-files are now in subdirectories, so we need to handle that - - - - - 88ebab0a by Ian Lynagh at 2010-08-18T19:43:53+00:00 GHC build system: Add all the data files to BINDIST_EXTRAS - - - - - 65837172 by David Waern at 2010-08-29T20:12:34+00:00 Update Test - - - - - 094bbaa2 by David Waern at 2010-08-29T20:55:14+00:00 Revert update to Test - - - - - a881cfb3 by David Waern at 2010-08-31T18:24:15+00:00 Bump version number - - - - - 1fc8a3eb by David Waern at 2010-08-31T22:32:27+00:00 Update ANNOUNCE - - - - - ee1df9d0 by David Waern at 2010-08-31T22:33:11+00:00 Update CHANGES - - - - - 394cc854 by David Waern at 2010-08-31T22:33:23+00:00 Update interface file versioning to work with ghc 6.14/15 - - - - - 7d03b79b by David Waern at 2010-08-31T22:36:00+00:00 Update test output following version change - - - - - a48d82d1 by Mark Lentczner at 2010-09-01T04:29:35+00:00 sort options in doc to match --help output removed --html-help option, as it is no longer supported - - - - - 06561aeb by Mark Lentczner at 2010-09-01T05:29:32+00:00 update options documentation rewrote doc for --html added doc for --theme and --built-in-themes added --use-contents and --gen-contents - - - - - 57dea832 by Mark Lentczner at 2010-09-01T05:31:27+00:00 slight wording change about Frames mode - - - - - fa1f6da3 by David Waern at 2010-09-01T10:57:44+00:00 Update doc configure script to find docbook stylesheets on arch linux - - - - - addff770 by David Waern at 2010-09-01T11:02:29+00:00 Wibble - - - - - 8399006d by David Waern at 2010-09-01T11:19:21+00:00 Replace ghci> with >>> in example syntax - - - - - 35074cf8 by David Waern at 2010-09-01T19:03:27+00:00 Improve docs for --no-tmp-comp-dir - - - - - 0f8f8cfd by David Waern at 2010-09-02T11:22:27+00:00 Add a list of contributors to the user guide Break out everyone thanked in the `Acknowledgements` chapter into a separate contributor list and add everyone from `darcs show authors`. We consider everyone who is thanked to be a contributor as a conservative estimation :-) I have added some more contributors that I know about, who were not in the darcs history, but others may be missing. So please add anyone that you think is missing from the list. - - - - - 42ccf099 by David Waern at 2010-09-02T11:29:22+00:00 Update copyright years in license - - - - - 0d560479 by David Waern at 2010-09-02T11:38:52+00:00 Update release instructions - - - - - 72ab7796 by David Waern at 2010-09-02T19:27:08+00:00 Add a note to ANNOUNCE - - - - - bf9d9c5d by David Waern at 2010-09-02T19:27:48+00:00 H.Utils needs FFI on Win+MinGW - - - - - 048ae44a by Mark Lentczner at 2010-09-04T23:19:47+00:00 make TOC group header identifiers validate - - - - - 8c6faf36 by Simon Michael at 2010-09-22T07:12:34+00:00 add hints for cleaner darcs show authors output - - - - - 9909bd17 by Simon Michael at 2010-09-22T17:58:06+00:00 print haddock coverage info on stdout when generating docs A module's haddockable items are its exports and the module itself. The output is lightly formatted so you can align the :'s and sort for readability. - - - - - 6da72171 by David Waern at 2010-10-03T21:31:24+00:00 Style wibble - - - - - 2f8d8e4d by Tobias Brandt at 2010-08-27T07:01:21+00:00 adding the option to fully qualify identifiers - - - - - 833be6c6 by Tobias Brandt at 2010-08-27T15:50:28+00:00 adding support for local and relative name qualification - - - - - df15c4e9 by Tobias Brandt at 2010-08-27T15:56:37+00:00 corrected qualification help message - - - - - 449e9ce1 by David Waern at 2010-10-16T17:34:30+00:00 Solve conflicts - - - - - 3469bda5 by David Waern at 2010-10-16T18:42:40+00:00 Use "qual" as an abbreviation for qualification instead of "quali" for consistency - - - - - 97c2d728 by David Waern at 2010-10-16T18:47:07+00:00 Style police - - - - - ce14fbea by David Waern at 2010-10-16T21:15:25+00:00 Style police - - - - - fdf29e9d by David Waern at 2010-10-17T00:30:44+00:00 Add a pointer to the style guide - - - - - 8e6b44e8 by rrnewton at 2010-10-24T03:19:28+00:00 Change to index pages: include an 'All' option even when subdividing A-Z. - - - - - 755b131c by David Waern at 2010-11-14T19:39:36+00:00 Bump version - - - - - d0345a04 by David Waern at 2010-11-14T19:41:59+00:00 TAG 2.8.1 - - - - - f6221508 by Simon Peyton Jones at 2010-09-13T09:53:00+00:00 Adapt to minor changes in internal GHC functions - - - - - 1290713d by Ian Lynagh at 2010-09-15T10:37:18+00:00 Remove duplicate Outputable instance for Data.Map.Map - - - - - 87f69eef by Ian Lynagh at 2010-09-21T15:01:10+00:00 Bump GHC dep upper bound - - - - - af36e087 by Ian Lynagh at 2010-09-21T15:12:02+00:00 Fix up __GLASGOW_HASKELL__ tests - - - - - ad67716c by Ian Lynagh at 2010-09-21T20:31:35+00:00 Don't build haddock is HADDOCK_DOCS is NO - - - - - 63b3f1f5 by Ian Lynagh at 2010-09-21T21:39:51+00:00 Fixes for when HADDOCK_DOCS=NO - - - - - e92bfa42 by Ian Lynagh at 2010-09-29T21:15:38+00:00 Fix URL creation on Windows: Use / not \ in URLs. Fixes haskell/haddock#4353 - - - - - 66c55e05 by Ian Lynagh at 2010-09-30T17:03:34+00:00 Tidy up haddock symlink installation In particular, it now doesn't get created if we aren't installing haddock. - - - - - 549b5556 by Ian Lynagh at 2010-10-23T21:17:14+00:00 Follow extension-flattening change in GHC - - - - - d7c2f72b by David Waern at 2010-11-14T20:17:55+00:00 Bump version to 2.8.2 - - - - - 6989a3a9 by David Waern at 2010-11-14T20:26:01+00:00 Solve conflict - - - - - 055c6910 by Ian Lynagh at 2010-09-22T15:36:20+00:00 Bump GHC dep - - - - - c96c0763 by Simon Marlow at 2010-10-27T11:09:44+00:00 follow changes in the GHC API - - - - - 45907129 by David Waern at 2010-11-07T14:00:58+00:00 Update the HCAR entry - - - - - 61940b95 by David Waern at 2010-11-07T14:07:34+00:00 Make the HCAR entry smaller - - - - - aa590b7d by David Waern at 2010-11-14T21:30:59+00:00 Update HCAR entry with November 2010 version - - - - - 587f9847 by David Waern at 2010-11-14T23:48:17+00:00 Require ghc >= 7.0 - - - - - ff5c647c by David Waern at 2010-11-14T23:49:09+00:00 TAG 2.8.2 - - - - - 937fcb4f by David Waern at 2010-11-14T23:49:45+00:00 Solve conflict - - - - - 8e5d0c1a by David Waern at 2010-11-15T21:09:50+00:00 Remove code for ghc < 7 - - - - - 3d47b70a by David Waern at 2010-11-15T21:11:06+00:00 Fix bad merge - - - - - 7f4a0d8a by David Waern at 2010-11-15T21:13:57+00:00 Remove more ghc < 7 code - - - - - 9ee34b50 by David Waern at 2010-11-15T21:31:25+00:00 Match all AsyncExceptions in exception handler - - - - - 42849c70 by David Waern at 2010-11-15T21:35:31+00:00 Just say "internal error" instead of "internal Haddock or GHC error" - - - - - c88c809b by David Waern at 2010-11-15T21:44:19+00:00 Remove docNameOcc under the motto "don't name compositions" - - - - - b798fc7c by David Waern at 2010-11-15T23:27:13+00:00 Wibble - - - - - 2228197e by David Waern at 2010-11-15T23:28:24+00:00 Rename the HCAR entry file - - - - - 8a3f9090 by David Waern at 2010-11-16T00:05:29+00:00 Remove Haskell 2010 extensions from .cabal file - - - - - c7a0c597 by David Waern at 2010-11-16T00:10:28+00:00 Style wibbles - - - - - cde707a5 by David Waern at 2010-11-16T00:12:00+00:00 Remove LANGUAGE ForeignFunctionInterface pragmas - - - - - 1dbda8ed by David Waern at 2010-11-16T00:17:21+00:00 Make a little more use of DoAndIfThenElse - - - - - 4c45ff6e by David Waern at 2010-11-16T00:59:41+00:00 hlint police - - - - - d2feaf09 by David Waern at 2010-11-16T01:14:15+00:00 hlint police - - - - - 99876e97 by David Waern at 2010-11-20T19:06:00+00:00 Haddock documentation updates - - - - - 65ce6987 by David Waern at 2010-11-20T19:42:51+00:00 Follow the style guide closer in Haddock.Types and improve docs - - - - - 28ca304a by tob.brandt at 2010-11-20T17:04:40+00:00 add full qualification for undocumented names - - - - - d61341e3 by David Waern at 2010-11-20T20:04:15+00:00 Re-structure qualification code a little - - - - - 0057e4d6 by David Waern at 2010-11-20T20:07:55+00:00 Re-order functions - - - - - d7279afd by David Waern at 2010-11-21T03:39:54+00:00 Add BangPatterns to alex and happy source files - - - - - 629fe60e by tob.brandt at 2010-11-23T23:35:11+00:00 documentation for qualification - - - - - 37031cee by David Waern at 2010-11-23T21:06:44+00:00 Update CHANGES - don't mention 2.8.2, we won't release it - - - - - f2489e19 by David Waern at 2010-12-01T21:57:11+00:00 Update deps of runtests.hs to work with ghc 7.0.1 - - - - - d3657e9a by David Waern at 2010-12-01T22:04:57+00:00 Make tests compile with ghc 7.0.1 - - - - - a2f09d9b by David Waern at 2010-12-01T22:06:59+00:00 Update tests following version bump - - - - - 50883ebb by David Waern at 2010-12-06T14:09:18+00:00 Update tests following recent changes - - - - - fc2fadeb by David Waern at 2010-12-06T14:17:29+00:00 Add a flag --pretty-html for rendering indented html with newlines - - - - - 30832ef2 by David Waern at 2010-12-06T14:17:35+00:00 Use --pretty-html when running the test suite. Makes it easier to compare output - - - - - a0b81b31 by David Waern at 2010-12-06T14:18:27+00:00 Wibble - - - - - 3aaa23fe by David Waern at 2010-12-06T14:19:29+00:00 Haddockify ppHtml comments - - - - - 24bb24f0 by David Waern at 2010-12-06T14:23:15+00:00 Remove --debug. It was't used, and --verbosity should take its place - - - - - 6bc076e5 by David Waern at 2010-12-06T14:25:37+00:00 Rename golden-tests into html-tests. "golden tests" sounds strange - - - - - 53301e55 by David Waern at 2010-12-06T14:26:26+00:00 QUALI -> QUAL in the description --qual for consistency - - - - - 98b6affb by David Waern at 2010-12-06T21:54:02+00:00 Bump version - - - - - 371bf1b3 by David Waern at 2010-12-06T22:08:55+00:00 Update tests following version bump - - - - - 25be762d by David Waern at 2010-12-06T22:21:03+00:00 Update CHANGES - - - - - 7c7dac71 by David Waern at 2010-12-06T22:33:43+00:00 Update ANNOUNCE - - - - - 30d7a5f2 by Simon Peyton Jones at 2010-11-15T08:38:38+00:00 Alex generates BangPatterns, so make Lex.x accept them (It'd be better for Alex to generate this pragma.) - - - - - 605e8018 by Simon Marlow at 2010-11-17T11:37:24+00:00 Add {-# LANGUAGE BangPatterns #-} to mollify GHC - - - - - a46607ba by David Waern at 2010-12-07T14:08:10+00:00 Solve conflicts - - - - - b28cda66 by David Waern at 2010-12-09T20:41:35+00:00 Docs: Mention that \ is a special character in markup - - - - - a435bfdd by Ian Lynagh at 2010-11-17T14:01:19+00:00 TAG GHC 7.0.1 release - - - - - 5a15a05a by David Waern at 2010-12-11T17:51:19+00:00 Fix indentation problem - - - - - 4232289a by Lennart Kolmodin at 2010-12-17T18:32:03+00:00 Revise haddock.cabal given that we now require ghc-7 default-language should be Haskell2010, slight new semantics for extensions. Rewrite into clearer dependencies of base and Cabal. - - - - - a36302dc by David Waern at 2010-12-19T17:12:37+00:00 Update CHANGES - - - - - 7c8b85b3 by David Waern at 2010-12-19T17:14:24+00:00 Bump version - - - - - cff22813 by Ian Lynagh at 2011-01-05T18:24:27+00:00 Write hoogle output in utf8; fixes GHC build on Windows - - - - - c7e762ea by David Waern at 2011-01-22T00:00:35+00:00 Put title outside doc div when HTML:fying title+prologue Avoids indenting the title, and makes more sense since the title is not a doc string anyway. - - - - - 5f639054 by David Waern at 2011-01-22T16:09:44+00:00 Fix spelling error - contributed by Marco Silva - - - - - c11dce78 by Ian Lynagh at 2011-01-07T02:33:11+00:00 Follow GHC build system changes - - - - - 101cfaf5 by David Waern at 2011-01-08T14:06:44+00:00 Bump version - - - - - af62348b by David Waern at 2011-01-08T14:07:07+00:00 TAG 2.9.2 - - - - - 4d1f6461 by Ian Lynagh at 2011-01-07T23:06:57+00:00 Name the haddock script haddock-ghc-7.0.2 instead of haddock-7.0.2; haskell/haddock#4882 "7.0.2" looked like a haddock version number before - - - - - 8ee4d5d3 by Simon Peyton Jones at 2011-01-10T17:31:12+00:00 Update Haddock to reflect change in hs_tyclds field of HsGroup - - - - - 06f3e3db by Ian Lynagh at 2011-03-03T15:02:37+00:00 TAG GHC 7.0.2 release - - - - - 7de0667d by David Waern at 2011-03-10T22:47:13+00:00 Update CHANGES - - - - - 33a9f1c8 by David Waern at 2011-03-10T22:47:31+00:00 Fix build with ghc 7.0.1 - - - - - 4616f861 by David Waern at 2011-03-10T22:47:50+00:00 TAG 2.9.2-actual - - - - - 0dab5e3c by Simon Hengel at 2011-04-08T15:53:01+00:00 Set shell script for unit tests back to work - - - - - 85c54dee by Simon Hengel at 2011-04-08T16:01:24+00:00 Set unit tests back to work Here "ghci>" was still used instead of ">>>". - - - - - 1cea9b78 by Simon Hengel at 2011-04-08T16:25:36+00:00 Update runtests.hs for GHC 7.0.2 - - - - - 8e5b3bbb by Simon Hengel at 2011-04-08T16:28:49+00:00 Update Haddock version in *.html.ref - - - - - 2545e955 by Simon Hengel at 2011-04-08T17:09:28+00:00 Add support for blank lines in the result of examples Result lines that only contain the string "<BLANKLINE>" are treated as a blank line. - - - - - adf64d2e by Simon Hengel at 2011-04-08T17:36:50+00:00 Add documentation for "support for blank lines in the result of examples" - - - - - c51352ca by David Waern at 2011-05-21T23:57:56+00:00 Improve a haddock comment - - - - - 7419cf2c by David Waern at 2011-05-22T15:41:52+00:00 Use cabal's test suite support to run the test suite This gives up proper dependency tracking of the test script. - - - - - 7770070c by David Waern at 2011-05-22T01:45:44+00:00 We don't need to send DocOptions nor a flag to mkExportItems - - - - - 9d95b7b6 by David Waern at 2011-05-22T21:39:03+00:00 Fix a bug - - - - - 1f93699b by David Waern at 2011-05-22T21:40:21+00:00 Break out fullContentsOf, give it a better name and some documentation The documentation describes how we want this function to eventually behave, once we have fixed a few problems with the current implementation. - - - - - 9a86432f by David Waern at 2011-05-22T21:53:52+00:00 Fix some stylistic issues in mkExportItems - - - - - c271ff0c by David Waern at 2011-05-22T22:09:11+00:00 Indentation - - - - - 93e602b1 by David Waern at 2011-06-10T01:35:31+00:00 Add git commits since switchover: darcs format (followed by a conflict resolution): commit 6f92cdd12d1354dfbd80f8323ca333bea700896a Merge: f420cc4 28df3a1 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Thu May 19 17:54:34 2011 +0100 Merge remote branch 'origin/master' into ghc-generics commit 28df3a119f770fdfe85c687dd73d5f6712b8e7d0 Author: Max Bolingbroke <batterseapower at hotmail.com> Date: Sat May 14 22:37:02 2011 +0100 Unicode fix for getExecDir on Windows commit 89813e729be8bce26765b95419a171a7826f6d70 Merge: 6df3a04 797ab27 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 9 11:55:17 2011 +0100 Merge branch 'ghc-new-co' commit 6df3a040da3dbddee67c6e30a892f87e6b164383 Author: Ian Lynagh <igloo at earth.li> Date: Sun May 8 17:05:50 2011 +0100 Follow changes in SDoc commit f420cc48b9259f0b1afd2438b12f9a2bde57053d Author: Jose Pedro Magalhaes <jpm at cs.uu.nl> Date: Wed May 4 17:31:52 2011 +0200 Adapt haddock to the removal of HsNumTy and TypePat. commit 797ab27bdccf39c73ccad374fea265f124cb52ea Merge: 1d81436 5a91450 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:05:03 2011 +0100 Merge remote branch 'origin/master' into ghc-new-co commit 1d8143659a81cf9611668348e33fd0775c7ab1d2 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:03:46 2011 +0100 Wibbles for ghc-new-co branch commit 5a91450e2ea5a93c70bd3904b022445c9cc82488 Author: Ian Lynagh <igloo at earth.li> Date: Fri Apr 22 00:51:56 2011 +0100 Follow defaultDynFlags change in GHC - - - - - 498da5ae by David Waern at 2011-06-11T00:33:33+00:00 * Merge in git patch from Michal Terepeta >From 6fc71d067738ef4b7de159327bb6dc3d0596be29 Mon Sep 17 00:00:00 2001 From: Michal Terepeta <michal.terepeta at gmail.com> Date: Sat, 14 May 2011 19:18:22 +0200 Subject: [PATCH] Follow the change of TypeSig in GHC. This follows the change in GHC to make TypeSig take a list of names (instead of just one); GHC ticket haskell/haddock#1595. This should also improve the Haddock output in case the user writes a type signature that refers to many names: -- | Some comment.. foo, bar :: ... will now generate the expected output with one signature for both names. - - - - - 094607fe by Ian Lynagh at 2011-06-17T19:10:29+01:00 Fix build - - - - - 8fa35740 by Ian Lynagh at 2011-06-26T21:06:40+01:00 Bump GHC dep to allow 7.2 - - - - - e4d2ca3c by Ian Lynagh at 2011-07-07T23:06:28+01:00 Relax base dep - - - - - b948fde9 by Ian Lynagh at 2011-07-28T16:39:45+01:00 GHC build system: Don't install the datafiles twice - - - - - f82f6d70 by Simon Marlow at 2011-08-11T12:08:15+01:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - b341cc12 by Max Bolingbroke at 2011-08-22T20:25:27+01:00 Fix compilation with no-pred-ty GHC - - - - - 30494581 by Max Bolingbroke at 2011-08-23T10:20:54+01:00 Remaining fixes for PredTy removal - - - - - 0b197138 by Max Bolingbroke at 2011-08-26T08:27:45+01:00 Rename factKind to constraintKind - - - - - a379bec5 by Max Bolingbroke at 2011-09-04T12:54:47+01:00 Deal with change to IParam handling in GHC - - - - - f94e421b by Max Bolingbroke at 2011-09-06T17:34:31+01:00 Adapt Haddock for the ConstraintKind extension changes - - - - - 8821e5cc by Max Bolingbroke at 2011-09-09T08:24:59+01:00 Ignore associated type defaults (just as we ignore default methods) - - - - - 31a0afd4 by Max Bolingbroke at 2011-09-09T09:06:00+01:00 Merge branch 'no-pred-ty' of ssh://darcs.haskell.org/srv/darcs/haddock into no-pred-ty - - - - - dd3b530a by Max Bolingbroke at 2011-09-09T14:10:25+01:00 Merge branch 'no-pred-ty' Conflicts: src/Haddock/Convert.hs - - - - - 5f25ec96 by Max Bolingbroke at 2011-09-09T14:10:40+01:00 Replace FactTuple with ConstraintTuple - - - - - cd30b9cc by David Waern at 2011-09-26T02:17:55+02:00 Bump to version 2.9.3 - - - - - 4fbfd397 by Max Bolingbroke at 2011-09-27T14:55:21+01:00 Follow changes to BinIface Name serialization - - - - - 92257d90 by David Waern at 2011-09-30T23:45:07+02:00 Fix problem with test files not added to distribution tarball - - - - - 00255bda by David Waern at 2011-09-30T23:48:24+02:00 Merge branch 'development' - - - - - 5421264f by David Waern at 2011-10-01T01:25:39+02:00 Merge in darcs patch from Simon Meier: Wed Jun 1 19:41:16 CEST 2011 iridcode at gmail.com * prettier haddock coverage info The new coverage info rendering uses less horizontal space. This reduces the number of unnecessary line-wrappings. Moreover, the most important information, how much has been documented already, is now put up front. Hopefully, this makes it more likely that a library author is bothered by the low coverage of his modules and fixes that issue ;-) - - - - - 07d318ef by David Waern at 2011-10-01T01:34:10+02:00 Use printException instead of deprecated printExceptionAndWarnings - - - - - 40d52ee4 by David Waern at 2011-10-01T01:41:13+02:00 Merge in darcs pach: Mon Apr 11 18:09:54 JST 2011 Liyang HU <haddock at liyang.hu> * Remember collapsed sections in index.html / haddock-util.js - - - - - 279d6dd4 by David Waern at 2011-10-01T01:55:45+02:00 Merge in darcs patch: Joachim Breitner <mail at joachim-breitner.de>**20110619201645 Ignore-this: f6c51228205b0902ad5bfad5040b989a As reported on http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=578301, generating the global index takes much too long if type-level (with lots of auto-generated types) is installed. The patch avoids a quadratic runtime in the subfunction getIfaceIndex of ppHtmlIndex by using a temporary set. Runtime improvement observed here from 25.36s to 2.86s. - - - - - d1612383 by David Waern at 2011-10-01T01:56:48+02:00 Merge branch 'development' - - - - - 347520c1 by David Waern at 2011-10-01T01:56:54+02:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9a0c95e8 by David Waern at 2011-10-01T02:19:10+02:00 Improve .cabal file - - - - - 6967dc64 by Ian Lynagh at 2011-10-01T01:34:06+01:00 Follow changes to ForeignImport/ForeignExport in GHC - - - - - 565cb26b by Simon Marlow at 2011-10-04T00:15:04+02:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - 8b74f512 by David Waern at 2011-10-04T00:18:17+02:00 Requre ghc >= 7.2 - - - - - 271d360c by David Waern at 2011-10-04T00:22:50+02:00 Bump version to 2.9.4 - - - - - 37f3edb0 by David Waern at 2011-10-06T02:30:21+02:00 Add alex and happy to build-tools. - - - - - 7ac2bb6e by David Terei at 2011-10-12T14:02:55-07:00 Add safe haskell indication to haddock output - - - - - 42c91a47 by David Terei at 2011-10-12T14:06:03-07:00 Fix CSS issue with info table not being contained in module header - - - - - 0eddab6c by David Terei at 2011-10-12T14:06:58-07:00 Add safe haskell indication to haddock output - - - - - 3df058eb by David Terei at 2011-10-12T14:07:07-07:00 Fix CSS issue with info table not being contained in module header - - - - - a40a6c3f by David Waern at 2011-10-22T11:29:06+02:00 Bump .haddock file version since the format has changed recently - - - - - 8a6254be by David Waern at 2011-10-22T11:30:42+02:00 Merge branch 'development' - - - - - 642e3e02 by David Waern at 2011-10-23T21:23:39+02:00 Sort import list - - - - - 36371cf8 by David Waern at 2011-10-23T22:48:18+02:00 Remove NEW_GHC_LAYOUT conditional. - - - - - 5604b499 by David Waern at 2011-10-27T00:15:03+02:00 Add --print-ghc-path. - - - - - 463499fa by David Waern at 2011-10-27T00:16:22+02:00 Make testsuite able to find its dependencies automatically. - - - - - a3506172 by Ryan Newton at 2011-11-05T05:59:58-04:00 Improved declNames internal error. Added a case to handle DocD. - - - - - 001b8baf by David Waern at 2011-11-05T20:37:29+01:00 Rename copy.hs -> accept.hs. - - - - - 55d808d3 by David Waern at 2011-11-05T23:30:02+01:00 Fix build. - - - - - deb5c3be by David Waern at 2011-11-06T00:01:47+01:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9b663554 by David Waern at 2011-11-06T00:03:45+01:00 Merge https://github.com/rrnewton/haddock - - - - - 1abb0ff6 by David Waern at 2011-11-06T01:20:37+01:00 Use getDeclMainBinder instead of declNames. - - - - - 4b005c01 by David Waern at 2011-11-06T19:09:53+01:00 Fix build. - - - - - c2c51bc7 by Ian Lynagh at 2011-11-06T23:01:33+00:00 Remove -DNEW_GHC_LAYOUT in ghc.mk - - - - - f847d703 by Jose Pedro Magalhaes at 2011-11-11T09:07:39+00:00 New kind-polymorphic core This big patch implements a kind-polymorphic core for GHC. The current implementation focuses on making sure that all kind-monomorphic programs still work in the new core; it is not yet guaranteed that kind-polymorphic programs (using the new -XPolyKinds flag) will work. For more information, see http://haskell.org/haskellwiki/GHC/Kinds - - - - - 7d7c3b09 by Jose Pedro Magalhaes at 2011-11-16T21:42:22+01:00 Follow changes to tuple sorts in master - - - - - 8430e03e by Simon Peyton Jones at 2011-11-17T10:20:27+00:00 Remove redundant imports - - - - - d1b06832 by Ian Lynagh at 2011-11-19T01:33:21+00:00 Follow GHC build system change to the way we call rm - - - - - 9e2230ed by David Waern at 2011-11-24T15:00:24+01:00 Fix a bug in test runner and get rid of regex-compat dependency. - - - - - 52039b21 by David Waern at 2011-11-24T23:55:36+01:00 Avoid haskell98 dependency in test - - - - - 92e1220d by David Waern at 2011-11-25T00:03:33+01:00 Avoid depency on regex-compat also in accept.hs. - - - - - ddac6b6f by David Waern at 2011-11-25T02:13:38+01:00 Accept test output. - - - - - 5a720455 by David Waern at 2011-11-25T02:16:20+01:00 Some more changes to test scripts. - - - - - 170a9004 by David Waern at 2011-11-25T02:30:41+01:00 Add flag --interface-version. - - - - - d225576c by David Waern at 2011-11-25T02:39:26+01:00 Remove #ifs for older compiler versions. - - - - - f0d0a4f5 by David Waern at 2011-11-26T04:20:12+01:00 Give preference to type over data constructors for doc comment links at renaming time. Previously this was done in the backends. Also, warn when a doc comment refers to something that is in scope but which we don't have the .haddock file for. These changes mean we can make DocIdentifier [a] into DocIdentifier a. - - - - - eef0e776 by David Waern at 2011-11-26T17:01:06+01:00 Allow doc comments to link to out-of-scope things (#78). (A bug that should have been fixed long ago.) - - - - - 565ad529 by David Waern at 2011-11-26T19:56:21+01:00 Update tests. - - - - - fb3ce7b9 by David Waern at 2011-11-26T21:44:28+01:00 Cleanup. - - - - - d0328126 by David Waern at 2011-11-26T22:10:28+01:00 Fix module reference bug. - - - - - c03765f8 by David Waern at 2011-12-03T05:20:20+01:00 Slightly better behaviour on top-levels without type signatures. - Docs don't get attached to the next top-level with signature by mistake. - If there's an export list and the top-level is part of it, its doc comment shows up in the documentation. - - - - - 48461d31 by David Waern at 2011-12-03T05:38:10+01:00 Add a test for Unicode doc comments. - - - - - 549c4b4e by David Waern at 2011-12-03T19:07:55+01:00 Cleanup. - - - - - 7bfecf91 by David Waern at 2011-12-03T20:13:08+01:00 More cleanup. - - - - - 14fab722 by Ian Lynagh at 2011-12-12T21:21:35+00:00 Update dependencies and binaryInterfaceVersion - - - - - 469e6568 by Ian Lynagh at 2011-12-18T12:56:16+00:00 Fix (untested) building from source tarball without alex/happy haddock's .cabal file was declaring that it needed alex and happy to build, but in the GHC source tarballs it doesn't. - - - - - 895c9a8c by David Waern at 2011-12-27T12:57:43+01:00 Go back to having a doc, sub and decl map instead of one big decl map. This setup makes more sense since when we add value bindings to the processed declarations (for type inference), we will have multiple declarations which should share documentation. Also, we already have a separate doc map for instances which we can now merge into the main doc map. Another benefit is that we don't need the DeclInfo type any longer. - - - - - 736767d9 by David Waern at 2011-12-27T13:33:41+01:00 Merge ../../../haddock Conflicts: src/Haddock/InterfaceFile.hs - - - - - 20016f79 by David Waern at 2011-12-27T13:57:23+01:00 Bump version. - - - - - 31f276fb by David Waern at 2011-12-27T13:57:32+01:00 Merge ../ghc/utils/haddock - - - - - 95b367cd by David Waern at 2011-12-27T14:57:29+01:00 Update tests following version bump. - - - - - fa3c94cd by David Waern at 2011-12-27T14:57:51+01:00 Get rid of quite unnecessary use of different lists. - - - - - 9c4d3c54 by David Waern at 2011-12-27T15:26:42+01:00 Cleanup. - - - - - 2caf9f90 by David Waern at 2011-12-27T16:18:05+01:00 Wibbles. - - - - - 3757d09b by David Waern at 2011-12-27T20:50:26+01:00 Complete support for inferring types for top-level bindings. - - - - - 53418734 by David Waern at 2011-12-28T15:02:13+01:00 Minor fixes and cleanup. - - - - - 0c9d0385 by Ian Lynagh at 2012-01-03T18:31:29+00:00 Follow rename of Instance to ClsInst in GHC - - - - - c9bc969a by Simon Hengel at 2012-01-12T21:28:14+01:00 Make sure that generated xhtml is valid (close haskell/haddock#186) Thanks to Phyx. - - - - - 836a0b9a by David Waern at 2012-02-01T02:30:05+01:00 Fix bug introduced in my recent refactoring. - - - - - c7d733eb by David Waern at 2012-02-01T02:30:26+01:00 Cleanup mkMaps and avoid quadratic behaviour. - - - - - da3cda8f by David Waern at 2012-02-01T02:56:56+01:00 Require ghc >= 7.4. - - - - - 83a3287e by David Waern at 2012-02-01T02:57:36+01:00 Update CHANGES. - - - - - 93408f0b by Simon Hengel at 2012-02-04T00:48:04+01:00 Add reference renderings - - - - - 49d00d2c by Simon Hengel at 2012-02-04T00:48:25+01:00 Set unit tests for parser back to work - - - - - eb450980 by Simon Hengel at 2012-02-04T00:49:07+01:00 Add .gitignore - - - - - a841602c by Simon Hengel at 2012-02-04T00:49:16+01:00 Add .ghci file - - - - - 8861199d by Simon Hengel at 2012-02-04T00:49:29+01:00 tests/html-tests/copy.hs: Use mapM_ instead of mapM So we do net get a list of () on stdout when running with runhaskell. - - - - - b477d9b5 by Simon Hengel at 2012-02-04T00:49:46+01:00 Remove index files from golden tests - - - - - 9dbda34e by Simon Hengel at 2012-02-04T00:49:57+01:00 Add /tests/html-tests/tests/*index*.ref to .gitignore - - - - - a9434817 by Simon Hengel at 2012-02-04T00:50:04+01:00 Add DocWarning to Doc The Xhtml backend has special markup for that, Hoogle and LaTeX reuse what we have for DocEmphasis. - - - - - de2fb6fa by Simon Hengel at 2012-02-04T00:50:13+01:00 Add support for module warnings - - - - - 0640920e by Simon Hengel at 2012-02-04T00:50:21+01:00 Add tests for module warnings - - - - - 30ce0d77 by Simon Hengel at 2012-02-04T00:50:29+01:00 Add support for warnings - - - - - bb367960 by Simon Hengel at 2012-02-04T00:50:37+01:00 Add tests for warnings - - - - - 6af1dc2d by Simon Hengel at 2012-02-04T00:50:50+01:00 Expand type signatures in export list (fixes haskell/haddock#192) - - - - - a06cbf25 by Simon Hengel at 2012-02-04T00:51:04+01:00 Expand type signatures for modules without explicit export list - - - - - 57dda796 by Simon Hengel at 2012-02-04T00:51:15+01:00 Remove obsolete TODO - - - - - 270c3253 by David Waern at 2012-02-04T00:51:24+01:00 Fix issues in support for warnings. * Match against local names only. * Simplify (it's OK to map over the warnings). - - - - - 683634bd by David Waern at 2012-02-04T00:55:11+01:00 Some cleanup and make sure we filter warnings through exports. - - - - - 210cb4ca by David Waern at 2012-02-04T03:01:30+01:00 Merge branch 'fix-for-186' of https://github.com/sol/haddock into ghc-7.4 - - - - - e8db9031 by David Waern at 2012-02-04T03:07:51+01:00 Style police. - - - - - 261f9462 by David Waern at 2012-02-04T03:20:16+01:00 Update tests. - - - - - 823cfc7c by David Waern at 2012-02-04T03:21:12+01:00 Use mapM_ in accept.hs as well. - - - - - 873dd619 by David Waern at 2012-02-04T03:21:33+01:00 Remove copy.hs - use accept.hs instead. - - - - - 0e31a14a by David Waern at 2012-02-04T03:47:33+01:00 Use <> instead of mappend. - - - - - 2ff7544f by David Waern at 2012-02-04T03:48:55+01:00 Remove code for older ghc versions. - - - - - dacf2786 by David Waern at 2012-02-04T15:52:51+01:00 Clean up some code from last SoC project. - - - - - 00cbb117 by David Waern at 2012-02-04T21:43:49+01:00 Mostly hlint-inspired cleanup. - - - - - 7dc86cc2 by Simon Peyton Jones at 2012-02-06T09:14:41+00:00 Track changes in HsDecls - - - - - f91f82fe by Ian Lynagh at 2012-02-16T13:40:11+00:00 Follow changes in GHC caused by the CAPI CTYPE pragma - - - - - a0ea6b0b by Ian Lynagh at 2012-02-22T02:26:12+00:00 Follow changes in GHC - - - - - b23b07d1 by Simon Peyton Jones at 2012-03-02T16:36:41+00:00 Follow changes in data representation from the big PolyKinds commit - - - - - 43406022 by Simon Hengel at 2012-03-05T11:18:34+01:00 Save/restore global state for static flags when running GHC actions This is necessary if we want to run createInterfaces (from Documentation.Haddock) multiple times in the same process. - - - - - 9fba16fe by Paolo Capriotti at 2012-03-06T10:57:33+00:00 Update .gitignore. - - - - - a9325044 by Simon Peyton Jones at 2012-03-14T17:35:42+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - fd48065a by Iavor Diatchki at 2012-03-15T22:43:35-07:00 Add support for type-level literals. - - - - - 2e8206dd by Simon Peyton Jones at 2012-03-16T14:18:22+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - 93e13319 by Simon Peyton Jones at 2012-03-17T01:04:05+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock Conflicts: src/Haddock/Convert.hs - - - - - d253fa71 by Iavor Diatchki at 2012-03-19T20:12:18-07:00 Merge remote-tracking branch 'origin/master' into type-nats - - - - - fc40acc8 by Iavor Diatchki at 2012-03-19T20:31:27-07:00 Add a missing case for type literals. - - - - - fd2ad699 by Iavor Diatchki at 2012-03-24T13:28:29-07:00 Rename variable to avoid shadowing warning. - - - - - 9369dd3c by Simon Peyton Jones at 2012-03-26T09:14:23+01:00 Follow refactoring of TyClDecl/HsTyDefn - - - - - 38825ca5 by Simon Peyton Jones at 2012-03-26T09:14:37+01:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - 4324ac0f by David Waern at 2012-04-01T01:51:19+02:00 Disable unicode test. - - - - - 3165b750 by David Waern at 2012-04-01T01:51:34+02:00 Take reader environment directly from TypecheckedSource. - - - - - 213b644c by David Waern at 2012-04-01T01:55:20+02:00 Cleanup. - - - - - 3118b4ba by David Waern at 2012-04-01T02:16:15+02:00 Don't filter out unexported names from the four maps - fixes a regression. - - - - - d6524e17 by David Waern at 2012-04-01T02:40:34+02:00 Fix crash when using --qual. Naughty GHC API! - - - - - ea3c43d8 by Henning Thielemann at 2012-04-01T13:03:07+02:00 add QualOption type for distinction between qualification argument given by the user and the actual qualification for a concrete module - - - - - 5422ff05 by Henning Thielemann at 2012-04-01T16:25:02+02:00 emit an error message when the --qual option is used incorrectly - - - - - 026e3404 by David Waern at 2012-04-01T18:10:30+02:00 Don't crash on unicode strings in doc comments. - - - - - ce006632 by David Waern at 2012-04-01T20:13:35+02:00 Add test for --ignore-all-exports flag/ignore-exports pragma. - - - - - 6e4dd33c by David Waern at 2012-04-01T20:21:03+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.4 - - - - - 734ae124 by Henning Thielemann at 2012-04-01T20:22:10+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - 622f9ba5 by David Waern at 2012-04-01T21:26:13+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 55ce17cb by Henning Thielemann at 2012-04-01T22:03:25+02:00 'abbreviate' qualification style - basic support Currently we ignore the package a module is imported from. This means that a module import would shadow another one with the same module name from a different package. - - - - - c85314ef by David Waern at 2012-04-01T22:05:12+02:00 Check qualification option before processing modules. - - - - - ae4b626c by Henning Thielemann at 2012-04-02T00:19:36+02:00 abbreviated qualification: use Packages.lookupModuleInAllPackages for finding the package that a module belongs to - - - - - 60bdbcf5 by Henning Thielemann at 2012-04-02T00:25:31+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - df44301d by Henning Thielemann at 2012-04-02T00:29:05+02:00 qualification style 'abbreviated' -> 'aliased' - - - - - f4192a64 by David Waern at 2012-04-02T01:05:47+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 7ba09067 by David Terei at 2012-04-04T15:08:21-07:00 Fix reporting of modules safe haskell mode (#5989) - - - - - d0cc33d0 by David Terei at 2012-04-06T15:50:41+01:00 Fix reporting of modules safe haskell mode (#5989) - - - - - 6e3434c5 by Simon Peyton Jones at 2012-04-20T18:37:46+01:00 Track changes in HsSyn - - - - - 22014ed0 by Simon Peyton Jones at 2012-05-11T22:45:15+01:00 Follow changes to LHsTyVarBndrs - - - - - d9a07b24 by David Waern at 2012-05-15T01:46:35+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - a6c4ebc6 by David Waern at 2012-05-16T02:18:32+02:00 Update CHANGES. - - - - - 8e181d29 by David Waern at 2012-05-16T02:27:56+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - e358210d by David Waern at 2012-05-16T02:35:33+02:00 Mention the new aliased --qual mode in CHANGES. - - - - - efd36a28 by David Waern at 2012-05-16T21:33:13+02:00 Bump version number. - - - - - d6b3af14 by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for deprecated record field - - - - - 927f800e by Simon Hengel at 2012-05-17T19:08:20+02:00 Use >>= instead of fmap and join - - - - - 048b41d5 by Simon Hengel at 2012-05-17T19:08:20+02:00 newtype-wrap Doc nodes for things that may have warnings attached - - - - - e3a89fc3 by Simon Hengel at 2012-05-17T19:08:20+02:00 Attach warnings to `Documentation` type - - - - - 5d4cc43d by Simon Hengel at 2012-05-17T19:08:20+02:00 Simplify lookupWarning - - - - - cf8ae69d by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for haskell/haddock#205 - - - - - cb409b19 by Simon Peyton Jones at 2012-05-25T08:30:11+01:00 Follow changes in LHsTyVarBndrs - - - - - 2d5f4179 by Simon Hengel at 2012-05-26T19:21:29+02:00 Add Applicative instance for (GenRnM a) - - - - - e4373060 by Simon Hengel at 2012-05-26T19:21:33+02:00 Use a map for warnings, as suggested by @waern - - - - - 597a68c7 by Simon Hengel at 2012-05-27T08:48:24+02:00 Add an optional label to URLs - - - - - ef1ac7fe by Simon Hengel at 2012-05-27T08:48:24+02:00 Add support for hyperlink labels to parser - - - - - 41f2adce by Simon Hengel at 2012-05-27T08:48:24+02:00 Add golden test for hyperlinks - - - - - 83d5e764 by Simon Hengel at 2012-05-27T08:50:02+02:00 Use LANGUAGE pragmas instead of default-extensions in cabal file - - - - - ddb755e5 by Simon Hengel at 2012-05-27T08:50:02+02:00 Fix typo in comment - - - - - 110676b4 by Simon Hengel at 2012-05-27T08:50:02+02:00 Add a type signature for a where-binding - - - - - 7d9ba2a0 by Ian Lynagh at 2012-06-12T14:38:01+01:00 Follow changes in GHC - - - - - 47c704f2 by Ian Lynagh at 2012-06-12T18:52:16+01:00 Follow changes in GHC - - - - - e1efe1ab by Simon Peyton Jones at 2012-06-13T17:25:29+01:00 Follow changes for the implementation of implicit parameters - - - - - 69abc81c by Ian Lynagh at 2012-06-19T22:52:58+01:00 Follow changes in base - - - - - 9d074a21 by Paolo Capriotti at 2012-06-22T18:26:47+01:00 Use right docMap to get decl documentation. - - - - - e3292ef6 by Ian Lynagh at 2012-07-15T01:31:19+01:00 Follow changes in GHC - - - - - ceae56b0 by Ian Lynagh at 2012-07-16T21:22:48+01:00 Fix haddock following some GHC changes Passing _|_ as the Settings for defaultDynFlags no longer works well enough - - - - - 9df72735 by Paolo Capriotti at 2012-07-19T16:49:32+01:00 Forward port changes from stable. - - - - - 572f5fcf by Ian Lynagh at 2012-07-19T20:38:26+01:00 Merge branch 'master' of darcs.haskell.org:/srv/darcs//haddock - - - - - 9195aca4 by Paolo Capriotti at 2012-07-20T10:27:28+01:00 Update dependencies. - - - - - 33db3923 by Ian Lynagh at 2012-07-20T17:54:43+01:00 Build with GHC 7.7 - - - - - 925a2cea by David Waern at 2012-07-23T16:50:40+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.6 Conflicts: src/Haddock/InterfaceFile.hs - - - - - d710ef97 by David Waern at 2012-07-23T16:52:07+02:00 Bump version number. - - - - - eb0c2f83 by David Waern at 2012-07-23T16:57:58+02:00 Update CHANGES. - - - - - b3f56943 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Hide "internal" instances This fixes haskell/haddock#37 (http://trac.haskell.org/haddock/ticket/37) Precisely, we show an instance iff its class and all the types are exported by non-hidden modules. - - - - - a70aa412 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Tests for hiding instances (#37) - - - - - c0f4aa58 by Simon Hengel at 2012-07-27T13:00:13+03:00 Add an other test for hiding instances (#37) - - - - - a7ed6268 by Ian Lynagh at 2012-08-07T14:48:13+01:00 Follow changes in GHC - - - - - 0ab30d38 by Ian Lynagh at 2012-08-13T22:12:27+01:00 Improve haddock memory usage - - - - - 0eaa4e30 by Ian Lynagh at 2012-08-13T23:58:46+01:00 Improve haddock memory usage - - - - - 659d26cf by Ian Lynagh at 2012-08-14T13:16:48+01:00 Remove some temporary pragmas I accidentally recorded - - - - - d97fceb6 by Simon Hengel at 2012-08-25T13:19:34+02:00 Add missing dependency to library - - - - - 4c910697 by Simon Hengel at 2012-08-28T07:39:14+02:00 Move .ghci to project root - - - - - fc3c601a by Simon Hengel at 2012-08-28T07:39:14+02:00 accept.hs: Ignore some files - - - - - 1af9b984 by Simon Hengel at 2012-08-28T07:40:04+02:00 Update reference renderings (bump version) - - - - - 980dc253 by Simon Hengel at 2012-08-28T07:40:32+02:00 Update reference renderings (remove links for ()) - - - - - 33651dbf by Simon Hengel at 2012-08-28T07:41:50+02:00 Update documentation of `runInteractiveProcess` in reference rendering - - - - - 7ab25078 by David Waern at 2012-09-07T10:38:50+02:00 Merge branch 'hiddenInstances2' of http://github.com/feuerbach/haddock into ghc-7.6 - - - - - c3de3a4b by David Waern at 2012-09-07T14:29:27+02:00 Follow changes in GHC. - - - - - 298c43ac by David Waern at 2012-09-07T14:59:24+02:00 Update CHANGES. - - - - - e797993a by David Waern at 2012-09-07T15:21:30+02:00 Update ANNOUNCE. - - - - - d0b44790 by David Waern at 2012-09-07T15:22:43+02:00 Merge branch 'hidden-instances' into ghc-7.6 - - - - - 41a4adc8 by Simon Hengel at 2012-09-08T12:08:37+02:00 Update doc/README - - - - - 71ad1040 by Simon Hengel at 2012-09-08T12:17:17+02:00 Add documentation for URL labels - - - - - 9bb41afd by Simon Peyton Jones at 2012-09-20T18:14:26+01:00 Follow data type changes in the tc-untouchables branch Relating entirely to SynTyConRhs - - - - - b8139bfa by Simon Hengel at 2012-09-21T14:24:16+02:00 Disable Unicode test for now - - - - - a5fafdd7 by Simon Hengel at 2012-09-21T14:35:45+02:00 Update TypeOperators test for GHC 7.6.1 Type operators can't be used as type variables anymore! - - - - - 6ccf0025 by Simon Hengel at 2012-09-21T16:02:24+02:00 Remove (Monad (Either e)) instance from ref. rendering of CrossPackageDocs I do not really understand why the behavior changed, so I'll open a ticket, so that we can further investigate. - - - - - b5c6c138 by Ian Lynagh at 2012-09-27T02:00:57+01:00 Follow changes in GHC build system - - - - - b98eded0 by David Waern at 2012-09-27T15:37:02+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 76cc2051 by David Waern at 2012-09-27T15:48:19+02:00 Update hidden instances tests. - - - - - aeaa1c59 by David Waern at 2012-09-28T10:21:32+02:00 Make API buildable with GHC 7.6. - - - - - d76be1b0 by Simon Peyton Jones at 2012-09-28T15:57:05+01:00 Merge remote-tracking branch 'origin/master' into tc-untouchables - - - - - a1922af8 by David Waern at 2012-09-28T19:50:20+02:00 Fix spurious superclass constraints bug. - - - - - bc41bdbb by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove old examples - - - - - bed7d3dd by Simon Hengel at 2012-10-01T11:30:51+02:00 Adapt parsetests for GHC 7.6.1 - - - - - dcdb22bb by Simon Hengel at 2012-10-01T11:30:51+02:00 Add test-suite section for parsetests to cabal file + get rid of HUnit dependency - - - - - 1e5263c9 by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove test flag from cabal file This was not really used. - - - - - 4beee98b by David Waern at 2012-09-28T23:42:28+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 11dd2256 by Ian Lynagh at 2012-10-03T16:17:35+01:00 Follow change in GHC build system - - - - - fbd77962 by Simon Hengel at 2012-10-03T18:49:40+02:00 Remove redundant dependency from cabal file - - - - - 09218989 by Simon Hengel at 2012-10-04T16:03:05+02:00 Fix typo - - - - - 93a2d5f9 by Simon Hengel at 2012-10-04T16:11:41+02:00 Remove trailing whitespace from cabal file - - - - - c8b46cd3 by Simon Hengel at 2012-10-04T16:12:17+02:00 Export Haddock's main entry point from library - - - - - b411e77b by Simon Hengel at 2012-10-04T16:29:46+02:00 Depend on library for executable The main motivation for this is to increase build speed. In GHC's source tree the library is not build, but all modules are now required for the executable, so that GHC's validate will now detect build failures for the library. - - - - - f8f0979f by Simon Hengel at 2012-10-05T00:32:57+02:00 Set executable flag for Setup.lhs - - - - - dd045998 by Simon Hengel at 2012-10-07T16:44:06+02:00 Extend rather than set environment when running HTML tests On some platforms (e.g. ppc64) GHC requires gcc in the path. - - - - - 7b39c3ae by Simon Hengel at 2012-10-07T17:05:45+02:00 cross-package test: re-export IsString instead of Monad There is a monad instance for Q, which is not available on platforms that do not have GHCi support. This caused CrossPackageDocs to fail on those platforms. Re-exporting IsString should test the same thing, but it works on all platforms. - - - - - 0700c605 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Fix some warnings - - - - - f78eca79 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Make -Wall proof - - - - - 6beec041 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Use listToMaybe/fromMaybe instead of safeHead/maybe - - - - - 44b8ce86 by Ian Lynagh at 2012-10-08T21:59:46+01:00 Follow changes in GHC - - - - - 6da5f702 by Simon Hengel at 2012-10-09T11:16:19+02:00 Update .ghci - - - - - 9ac1a1b9 by Kazu Yamamoto at 2012-10-09T12:45:31+02:00 Add markup support for properties - - - - - 1944cb42 by Simon Hengel at 2012-10-09T12:45:31+02:00 Simplify lexing/parsing of properties In contrast to what we do for examples, we do not really need to capture the "prompt" here. - - - - - bffd8e62 by Simon Hengel at 2012-10-09T13:40:14+02:00 Add HTML test for properties - - - - - 2fe9c5cb by Simon Hengel at 2012-10-09T13:40:21+02:00 Add unit tests for properties - - - - - 874e361b by Simon Hengel at 2012-10-09T13:40:33+02:00 Bump interface version - - - - - 2506cc37 by Simon Hengel at 2012-10-09T15:15:04+02:00 Fix parser bug - - - - - 743d2b7d by Simon Hengel at 2012-10-09T15:31:06+02:00 Allow to load interface files with compatible versions - - - - - 981a1660 by Simon Hengel at 2012-10-10T10:32:05+02:00 Export more types from Documentation.Haddock (fixes haskell/haddock#216) - - - - - dff7dc76 by Simon Hengel at 2012-10-10T11:15:19+02:00 Update ANNOUNCE and CHANGES - - - - - edd2bb01 by Simon Hengel at 2012-10-10T11:22:50+02:00 Bump version - - - - - 5039163b by Simon Hengel at 2012-10-10T13:56:04+02:00 Fix typo in documentation - - - - - e4ce34da by Simon Hengel at 2012-10-10T14:28:35+02:00 Add documentation for properties - - - - - 9555ebca by Simon Hengel at 2012-10-11T10:49:04+02:00 Remove redundant if-defs, more source documentation - - - - - 87aa67e1 by Simon Hengel at 2012-10-11T12:32:51+02:00 Adapt cabal file - - - - - c44c1dee by Simon Hengel at 2012-10-11T12:41:58+02:00 Require ghc 7.6 - - - - - 8383bc34 by Simon Hengel at 2012-10-11T12:50:24+02:00 Bump version - - - - - 1030eb38 by Simon Hengel at 2012-10-11T12:55:44+02:00 Update ANNOUNCE and CHANGES - - - - - 74955088 by Simon Hengel at 2012-10-12T09:49:31+02:00 Improve note about `binaryInterfaceVersion` (thanks David) - - - - - ee30f6b7 by Simon Hengel at 2012-10-13T13:40:59+02:00 Update version in html tests, rpm spec file, and user manual - - - - - f2861f18 by Simon Hengel at 2012-10-13T14:40:33+02:00 Remove unused MonadFix constraint - - - - - dfdf1a74 by Simon Hengel at 2012-10-13T15:15:38+02:00 Minor code simplification - - - - - 4ecd1e70 by Simon Hengel at 2012-10-13T15:33:43+02:00 Increase code locality - - - - - f7df5cc9 by Simon Hengel at 2012-10-13T16:03:12+02:00 Minor code simplification - - - - - e737eb6e by Simon Hengel at 2012-10-13T19:03:04+02:00 Handle HsExplicitListTy in renameer (fixes haskell/haddock#213) - - - - - c2dc8f17 by Simon Hengel at 2012-10-13T20:46:31+02:00 Better error messages - - - - - 14d48b4c by Simon Hengel at 2012-10-14T00:21:07+02:00 Simplify RnM type - - - - - 6c2cc547 by Simon Hengel at 2012-10-14T00:23:35+02:00 Simplify lookupRn - - - - - bc77ce85 by Simon Hengel at 2012-10-14T01:51:32+02:00 Organize unite tests hierarchically - - - - - 2306d117 by Simon Hengel at 2012-10-14T10:34:58+02:00 Handle more cases in renameType - - - - - 8a864203 by Simon Hengel at 2012-10-14T11:47:59+02:00 Add mini_HiddenInstances.html.ref and mini_HiddenInstancesB.html.ref - - - - - 3a978eca by Simon Hengel at 2012-10-14T11:49:28+02:00 Add /tests/html-tests/output/ to .gitignore - - - - - db18888a by Simon Hengel at 2012-10-14T13:38:21+02:00 Allow haddock markup in deprecation messages - - - - - e7cfee9f by Simon Hengel at 2012-10-14T14:00:23+02:00 If parsing of deprecation message fails, include it verbatim - - - - - 242a85be by Simon Hengel at 2012-10-14T14:13:24+02:00 Add description for PruneWithWarning test - - - - - 43d33df1 by Simon Hengel at 2012-10-14T15:40:53+02:00 Minor formatting change - - - - - 22768c44 by Simon Hengel at 2012-10-14T16:03:43+02:00 Properly handle deprecation messages for re-exported things (fixes haskell/haddock#220) - - - - - cb4b9111 by Simon Hengel at 2012-10-14T17:30:28+02:00 Add build artifacts for documentation to .gitignore - - - - - 854cd8de by Simon Hengel at 2012-10-14T23:34:51+02:00 unit-tests: Improve readability Add IsString instance for (Doc RdrName) + use <> instead of DocAppend. - - - - - c4446d54 by Simon Hengel at 2012-10-14T23:37:21+02:00 unit-tests: Minor refactoring Rename parse to parseParas. - - - - - 04f2703c by Simon Hengel at 2012-10-15T00:36:42+02:00 Fix typo - - - - - 3d109e44 by Simon Hengel at 2012-10-15T10:30:07+02:00 Add description for DeprecatedReExport test - - - - - 84f0985c by Simon Hengel at 2012-10-15T14:54:19+02:00 Move resources to /resources directory - - - - - a5de7ca6 by Simon Hengel at 2012-10-15T15:46:18+02:00 Move HTML tests to directory /html-test/ - - - - - e21f727d by Simon Hengel at 2012-10-15T19:32:42+02:00 Move HTML reference renderings to /html-test/ref/ - - - - - 3a3c6c75 by Simon Hengel at 2012-10-15T19:32:42+02:00 Copy css, images, etc. on accept - - - - - 40ead6dc by Simon Hengel at 2012-10-15T19:32:42+02:00 Move unit tests to /test directory - - - - - 99a28231 by Simon Hengel at 2012-10-15T19:32:42+02:00 Fix Setup.lhs /usr/bin/runhaskell is not installed on all systems. - - - - - 95faf45e by Simon Hengel at 2012-10-15T19:32:42+02:00 Make test management scripts more robust * They are now independent from the current directory, and hence can be called from everywhere * On UNIX/Linux they can now be run as scripts - - - - - 027aaa2d by Simon Hengel at 2012-10-15T19:53:40+02:00 Add 'dev' flag to cabal file, that builds without -O2 That way --disable-optimization can be used, which decreases build time considerably. - - - - - e0266ede by Simon Hengel at 2012-10-15T20:03:43+02:00 Add test case for "spurious superclass constraints bug" - - - - - 52a2aa92 by Simon Hengel at 2012-10-15T20:28:55+02:00 Adapt accept.lhs, so that it ignores more index files - - - - - 53530781 by Simon Hengel at 2012-10-15T20:49:39+02:00 Rename html-test/runtests.lhs to html-test/run.lhs - - - - - 84518797 by Simon Hengel at 2012-10-15T20:49:39+02:00 Move source files for HTML tests to html-test/src - - - - - a911dc6c by Simon Hengel at 2012-10-15T20:49:39+02:00 Adapt output directory for HTML tests - - - - - d3c15857 by Ian Lynagh at 2012-10-16T16:54:43+01:00 Follow dopt->gopt rename - - - - - 956665a5 by Simon Hengel at 2012-10-18T08:42:48+02:00 Update html-test/README - - - - - 903b1029 by Simon Hengel at 2012-10-18T08:50:26+02:00 Use markdown for html-test/README - - - - - 150b4d63 by Ian Lynagh at 2012-10-18T16:36:00+01:00 Follow changes in GHC: 'flags' has been renamed 'generalFlags' - - - - - 41e04ff9 by Simon Hengel at 2012-11-28T09:54:35+01:00 Export missing types from Documentation.Haddock - - - - - 9be59237 by Ian Lynagh at 2012-11-30T23:20:47+00:00 Update dependencies - - - - - e06842f5 by Simon Hengel at 2012-12-07T20:58:05+01:00 Bump version - - - - - e3dbede0 by Simon Hengel at 2012-12-07T20:58:05+01:00 Add missing test files to cabal file (fixes haskell/haddock#230) - - - - - ee0dcca7 by Simon Hengel at 2012-12-07T20:58:05+01:00 Update CHANGES - - - - - 51601bdb by Simon Peyton Jones at 2012-12-19T17:28:35+00:00 Track changes in UNPACK pragma stuff - - - - - f2573bc1 by Richard Eisenberg at 2012-12-21T20:56:25-05:00 Implement overlapping type family instances. An ordered, overlapping type family instance is introduced by 'type instance where', followed by equations. See the new section in the user manual (7.7.2.2) for details. The canonical example is Boolean equality at the type level: type family Equals (a :: k) (b :: k) :: Bool type instance where Equals a a = True Equals a b = False A branched family instance, such as this one, checks its equations in order and applies only the first the matches. As explained in the note [Instance checking within groups] in FamInstEnv.lhs, we must be careful not to simplify, say, (Equals Int b) to False, because b might later unify with Int. This commit includes all of the commits on the overlapping-tyfams branch. SPJ requested that I combine all my commits over the past several months into one monolithic commit. The following GHC repos are affected: ghc, testsuite, utils/haddock, libraries/template-haskell, and libraries/dph. Here are some details for the interested: - The definition of CoAxiom has been moved from TyCon.lhs to a new file CoAxiom.lhs. I made this decision because of the number of definitions necessary to support BranchList. - BranchList is a GADT whose type tracks whether it is a singleton list or not-necessarily-a-singleton-list. The reason I introduced this type is to increase static checking of places where GHC code assumes that a FamInst or CoAxiom is indeed a singleton. This assumption takes place roughly 10 times throughout the code. I was worried that a future change to GHC would invalidate the assumption, and GHC might subtly fail to do the right thing. By explicitly labeling CoAxioms and FamInsts as being Unbranched (singleton) or Branched (not-necessarily-singleton), we make this assumption explicit and checkable. Furthermore, to enforce the accuracy of this label, the list of branches of a CoAxiom or FamInst is stored using a BranchList, whose constructors constrain its type index appropriately. I think that the decision to use BranchList is probably the most controversial decision I made from a code design point of view. Although I provide conversions to/from ordinary lists, it is more efficient to use the brList... functions provided in CoAxiom than always to convert. The use of these functions does not wander far from the core CoAxiom/FamInst logic. BranchLists are motivated and explained in the note [Branched axioms] in CoAxiom.lhs. - The CoAxiom type has changed significantly. You can see the new type in CoAxiom.lhs. It uses a CoAxBranch type to track branches of the CoAxiom. Correspondingly various functions producing and consuming CoAxioms had to change, including the binary layout of interface files. - To get branched axioms to work correctly, it is important to have a notion of type "apartness": two types are apart if they cannot unify, and no substitution of variables can ever get them to unify, even after type family simplification. (This is different than the normal failure to unify because of the type family bit.) This notion in encoded in tcApartTys, in Unify.lhs. Because apartness is finer-grained than unification, the tcUnifyTys now calls tcApartTys. - CoreLinting axioms has been updated, both to reflect the new form of CoAxiom and to enforce the apartness rules of branch application. The formalization of the new rules is in docs/core-spec/core-spec.pdf. - The FamInst type (in types/FamInstEnv.lhs) has changed significantly, paralleling the changes to CoAxiom. Of course, this forced minor changes in many files. - There are several new Notes in FamInstEnv.lhs, including one discussing confluent overlap and why we're not doing it. - lookupFamInstEnv, lookupFamInstEnvConflicts, and lookup_fam_inst_env' (the function that actually does the work) have all been more-or-less completely rewritten. There is a Note [lookup_fam_inst_env' implementation] describing the implementation. One of the changes that affects other files is to change the type of matches from a pair of (FamInst, [Type]) to a new datatype (which now includes the index of the matching branch). This seemed a better design. - The TySynInstD constructor in Template Haskell was updated to use the new datatype TySynEqn. I also bumped the TH version number, requiring changes to DPH cabal files. (That's why the DPH repo has an overlapping-tyfams branch.) - As SPJ requested, I refactored some of the code in HsDecls: * splitting up TyDecl into SynDecl and DataDecl, correspondingly changing HsTyDefn to HsDataDefn (with only one constructor) * splitting FamInstD into TyFamInstD and DataFamInstD and splitting FamInstDecl into DataFamInstDecl and TyFamInstDecl * making the ClsInstD take a ClsInstDecl, for parallelism with InstDecl's other constructors * changing constructor TyFamily into FamDecl * creating a FamilyDecl type that stores the details for a family declaration; this is useful because FamilyDecls can appear in classes but other decls cannot * restricting the associated types and associated type defaults for a * class to be the new, more restrictive types * splitting cid_fam_insts into cid_tyfam_insts and cid_datafam_insts, according to the new types * perhaps one or two more that I'm overlooking None of these changes has far-reaching implications. - The user manual, section 7.7.2.2, is updated to describe the new type family instances. - - - - - f788d0fb by Simon Peyton Jones at 2012-12-23T15:49:58+00:00 Track changes in HsBang - - - - - ca460a0c by Simon Peyton Jones at 2012-12-23T15:50:28+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - f078fea6 by Simon Peyton Jones at 2013-01-02T08:33:13+00:00 Use InstEnv.instanceSig rather than instanceHead (name change) - - - - - 88e41305 by Simon Peyton Jones at 2013-01-14T17:10:27+00:00 Track change to HsBang type - - - - - e1ad4e19 by Kazu Yamamoto at 2013-02-01T11:59:24+09:00 Merge branch 'ghc-7.6' into ghc-7.6-merge-2 Conflicts: haddock.cabal src/Haddock/Interface/AttachInstances.hs src/Haddock/Interface/Create.hs src/Haddock/Interface/LexParseRn.hs src/Haddock/InterfaceFile.hs src/Haddock/Types.hs Only GHC HEAD can compile this. GHC 7.6.x cannot compile this. Some test fail. - - - - - 62bec012 by Kazu Yamamoto at 2013-02-06T11:12:28+09:00 Using tcSplitSigmaTy in instanceHead' (FIXME is resolved.) - - - - - 013fd2e4 by Kazu Yamamoto at 2013-02-06T17:56:21+09:00 Refactoring instanceHead'. - - - - - 3148ce0e by Kazu Yamamoto at 2013-02-07T17:45:10+09:00 Using new syntax in html-test/src/GADTRecords.hs. - - - - - 626dabe7 by Gabor Greif at 2013-02-15T22:42:01+01:00 Typo - - - - - 1eb667ae by Ian Lynagh at 2013-02-16T17:02:07+00:00 Follow changes in base - - - - - 3ef8253a by Ian Lynagh at 2013-03-01T23:23:57+00:00 Follow changes in GHC's build system - - - - - 1a265a3c by Ian Lynagh at 2013-03-03T23:12:07+00:00 Follow changes in GHC build system - - - - - 69941c79 by Max Bolingbroke at 2013-03-10T09:38:28-07:00 Use Alex 3's Unicode support to properly lex source files as UTF-8 Signed-off-by: David Waern <david.waern at gmail.com> - - - - - ea687dad by Simon Peyton Jones at 2013-03-15T14:16:10+00:00 Adapt to tcRnGetInfo returning family instances too This API change was part of the fix to Trac haskell/haddock#4175. But it offers new information to Haddock: the type-family instances, as well as the class instances, of this type. This patch just drops the new information on the floor, but there's an open opportunity to use it in the information that Haddock displays. - - - - - 971a30b0 by Andreas Voellmy at 2013-05-19T20:47:39+01:00 Fix for haskell/haddock#7879. Changed copy of utils/haddock/html/resources/html to use "cp -RL" rather than "cp -R". This allows users to run validate in a build tree, where the build tree was setup using lndir with a relative path to the source directory. - - - - - 31fb7694 by Ian Lynagh at 2013-05-19T20:47:49+01:00 Use "cp -L" when making $(INPLACE_LIB)/latex too - - - - - e9952233 by Simon Hengel at 2013-06-01T18:06:50+02:00 Add -itest to .ghci - - - - - b06873b3 by Mateusz Kowalczyk at 2013-06-01T18:06:50+02:00 Workaround for a failing build with --enable-tests. - - - - - e7858d16 by Simon Hengel at 2013-06-01T19:29:28+02:00 Fix broken test - - - - - 0690acb1 by Richard Eisenberg at 2013-06-21T14:08:25+01:00 Updates to reflect changes in HsDecls to support closed type families. - - - - - 7fd347ec by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 53ed81b6 by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 931c4f4f by Richard Eisenberg at 2013-07-24T13:15:59+01:00 Remove (error "synifyKind") to use WithinType, to allow haddock to process base. - - - - - 55a9c804 by Richard Eisenberg at 2013-08-02T15:54:55+01:00 Changes to reflect changes in GHC's type HsTyVarBndr - - - - - b6e9226c by Mathieu Boespflug at 2013-08-04T10:39:43-07:00 Output Copright and License keys in Xhtml backend. This information is as relevant in the documentation as it is in the source files themselves. Signed-off-by: David Waern <david.waern at gmail.com> - - - - - 4c66028a by David Waern at 2013-08-04T15:27:36-07:00 Bump interface file version. - - - - - 67340163 by David Waern at 2013-08-09T16:12:51-07:00 Update tests. - - - - - 2087569b by Mateusz Kowalczyk at 2013-08-25T09:24:13+02:00 Add spec tests. This adds tests for all elements we can create during regular parsing. This also adds tests for text with unicode in it. - - - - - 97f36a11 by Mateusz Kowalczyk at 2013-08-27T06:59:12+01:00 Fix ticket haskell/haddock#247. I do the same thing that the XHTML backend does: give these no special treatment and just act as if they are regular functions. - - - - - 60681b4f by Mateusz Kowalczyk at 2013-08-27T21:22:48+02:00 LaTeX tests setup - - - - - fa4c27b2 by Mateusz Kowalczyk at 2013-09-02T23:21:43+01:00 Fixes haskell/haddock#253 - - - - - 1a202490 by Mateusz Kowalczyk at 2013-09-03T01:12:50+01:00 Use Hspec instead of nanospec This is motivated by the fact that Haddock tests are not ran by the GHC's ‘validate’ script so we're pretty liberal on dependencies in that area. Full Hspec gives us some nice features such as Quickcheck integration. - - - - - 8cde3b20 by David Luposchainsky at 2013-09-08T07:27:28-05:00 Fix AMP warnings Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - d10661f2 by Herbert Valerio Riedel at 2013-09-11T15:15:01+02:00 Update Git repo URL in `.cabal` file - - - - - 16a44eb5 by Richard Eisenberg at 2013-09-17T09:34:26-04:00 Revision to reflect new role annotation syntax in GHC. - - - - - 4b9833b9 by Herbert Valerio Riedel at 2013-09-18T10:15:28+02:00 Add missing `traverse` method for `GenLocated` As `Traversable` needs at least one of `traverse` or `sequenceA` to be overridden. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - b71fed5d by Simon Hengel at 2013-09-18T22:43:34+02:00 Add test helper - - - - - 4fc1ea86 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#231 - - - - - 435872f6 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#256 We inject -dynamic-too into flags before we run all our actions in the GHC monad. - - - - - b8b24abb by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Add new field to DynFlags - - - - - 49558795 by Simon Hengel at 2013-09-18T22:43:35+02:00 Fallback to ./resources when Cabal data is not found (so that themes are found during development) - - - - - bf79d05c by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Fixes haskell/haddock#5 - - - - - e1baebc2 by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Print missing documentation. Fixes haskell/haddock#258. - - - - - 02ea74de by Austin Seipp at 2013-10-09T10:52:22-05:00 Don't consider StaticFlags when parsing arguments. Instead, discard any static flags before parsing the command line using GHC's DynFlags parser. See http://ghc.haskell.org/trac/ghc/ticket/8276 Based off a patch from Simon Hengel. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 704fd5bb by Simon Hengel at 2013-11-09T00:15:13+01:00 Update HTML tests - - - - - f9fed49e by Simon Hengel at 2013-11-10T18:43:58+01:00 Bump version - - - - - 97ae1999 by Simon Peyton Jones at 2013-11-25T17:25:14+00:00 Track changes in HsSpliceTy data constructor - - - - - 59ad8268 by Simon Peyton Jones at 2014-01-10T18:17:43+00:00 Adapt to small change in Pretty's exports - - - - - 8b12e6aa by Simon Hengel at 2014-01-12T14:48:35-06:00 Some code simplification by using traverse - - - - - fc5ea9a2 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix warnings in test helper - - - - - 6dbb3ba5 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Add ByteString version of Attoparsec - - - - - 968d7774 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 One pass parser and tests. We remove the HTML test as it is no longer necessary. We cover the test case in spec tests and other HTML tests but keeping this around fails: this is because the new parser has different semantics there. In fact, I suspect the original behaviour was a bug that wasn't caught/fixed but simply included as-is during the testing. - - - - - 37a07c9c by Simon Hengel at 2014-01-12T14:48:35-06:00 Rename Haddock.ParseSpec to Haddock.ParserSpec - - - - - f0f68fe9 by Simon Hengel at 2014-01-12T14:48:35-06:00 Don't append newline to parseString input We also check that we have parsed everything with endOfInput. - - - - - 95d60093 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix totality, unicode, examples, paragraph parsing Also simplify specs and parsers while we're at it. Some parsers were made more generic. This commit is a part of GHC pre-merge squash, email fuuzetsu at fuuzetsu.co.uk if you need the full commit history. - - - - - 7d99108c by Simon Hengel at 2014-01-12T14:48:35-06:00 Update acceptance tests - - - - - d1b59640 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Support for bold. Conflicts: src/Haddock/Backends/Hoogle.hs src/Haddock/Interface/Rename.hs src/Haddock/Parser.hs - - - - - 4b412b39 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Allow for headings inside function documentation. LaTeX will treat the h3-h6 headings the same as we'd have to hack the style file heavily otherwise and it would make the headings tiny anyway. Hoogle upstream said they will put in the functionality on their end. Conflicts: src/Haddock/Interface/Rename.hs src/Haddock/Types.hs test/Haddock/ParserSpec.hs - - - - - fdcca428 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Per-module extension flags and language listing. Any extensions that are not enabled by a used language (Haskell2010 &c) will be shown. Furthermore, any implicitly enabled are also going to be shown. While we could eliminate this either by using the GHC API or a dirty hack, I opted not to: if a user doesn't want the implied flags to show, they are recommended to use enable extensions more carefully or individually. Perhaps this will encourage users to not enable the most powerful flags needlessly. Enabled with show-extensions. Conflicts: src/Haddock/InterfaceFile.hs - - - - - 368942a2 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Bump interface version There were some breaking changes over the last few patches so we bump the interface version. This causes a big problem with testing: 1. To generate cross package docs, we first need to generate docs for the package used. 2. To generate package docs with new interface version, we need to use Haddock which has the version bumped. 3. To get Haddock with the version bump, we first need to test cross package docs 4. GOTO 1 So the problem is the chicken and the egg problem. It seems that the only solution would be to generate some interface files on the fly but it is non-trivial. To run this test, you'll have to: * build Haddock without the test (make sure everything else passes) * rebuild the packages used in the test with your shiny new binary making sure they are visible to Haddock * remove the ‘_hidden’ suffix and re-run the tests Note: because the packages currently used for this test are those provided by GHC, it's probably non-trivial to just re-build them. Preferably something less tedious to rebuild should be used and something that is not subject to change. - - - - - 124ae7a9 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow for nesting of paragraphs under lists. The nesting rules are similar to Markdown's with the exception that we can not simply indent the first line of a hard wrapped indented paragraph and have it treated as if it was fully indented. The reason is differences in markup as some of our constructs care about whitespace while others just swallow everything up so it's just a lot easier to not bother with it rather than making arbitrary rules. Note that we now drop trailing for string entities inside of lists. They weren't needed and it makes the output look uniform whether we use a single or double newline between list elements. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - c7913535 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow escaping in URLs and pictures. Some tests were moved under parseString as they weren't about paragraph level markup. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - 32326680 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update documentation. - - - - - fbef6406 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update maintainer - - - - - b40e82f4 by Mateusz Kowalczyk at 2014-01-13T02:39:25-06:00 Fixes haskell/haddock#271 Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - f4eafbf8 by Gergő Érdi at 2014-01-19T15:35:16-06:00 Support for -XPatternSynonyms Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - a8939591 by Austin Seipp at 2014-01-29T08:09:04-06:00 Update CPP check for __GLASGOW_HASKELL__ Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 30d7e9d5 by Gergő Érdi at 2014-01-31T00:15:01+08:00 <+>: Don't insert a space when concatenating empty nodes - - - - - a25ccd4d by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Fix @ code blocks In cases where we had some horizontal space before the closing ‘@’, the parser would not accept the block as a code block and we'd get ugly output. - - - - - 0f67305a by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Update tests This updates tests due to Haddock Trac haskell/haddock#271 fix and due to removal of TypeHoles as an extension from GHC. - - - - - 157322a7 by Gergő Érdi at 2014-01-31T01:03:17+08:00 Handle infix vs prefix names correctly everywhere, by explicitly specifying the context The basic idea is that "a" and "+" are either pretty-printed as "a" and "(+)" or "`a`" and "+" - - - - - aa6d9685 by Mateusz Kowalczyk at 2014-01-30T17:21:50+00:00 Correct whitespace in ‘hidden’ test for <+> change - - - - - 121872f0 by Mateusz Kowalczyk at 2014-02-09T17:59:12+00:00 Document module header. Fixes Haddock Trac haskell/haddock#270. - - - - - e3253746 by Mateusz Kowalczyk at 2014-02-10T21:37:48+00:00 Insert a space between module link and description Fixes Haddock Trac haskell/haddock#277. - - - - - 771d2384 by Mateusz Kowalczyk at 2014-02-10T23:27:21+00:00 Ensure a space between type signature and ‘Source’ This is briefly related to Haddock Trac haskell/haddock#249 and employs effectively the suggested fix _but_ it doesn't actually fix the reported issue. This commit simply makes copying the full line a bit less of a pain. - - - - - 8cda9eff by nand at 2014-02-11T15:48:30+00:00 Add support for type/data families This adds support for type/data families with their respective instances, as well as closed type families and associated type/data families. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 3f22c510 by nand at 2014-02-11T15:53:50+00:00 Improve display of poly-kinded type operators This now displays them as (==) k a b c ... to mirror GHC's behavior, instead of the old (k == a) b c ... which was just wrong. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - effb2d6b by nand at 2014-02-11T15:56:50+00:00 Add test case for PatternSynonyms This just tests various stuff including poly-kinded patterns and operator patterns to make sure the rendering isn't broken. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - b38faf0d by Niklas Haas at 2014-02-13T21:53:32+00:00 Get rid of re-implementation of sortBy I have no idea what this was doing lying around here, and due to the usage of tuples it's actually slower, too. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - ac1e0413 by Mateusz Kowalczyk at 2014-02-13T23:57:16+00:00 Only warn about missing docs when docs are missing This fixes the ‘Missing documentation for…’ message for modules with 100% coverage. - - - - - cae2e36a by Niklas Haas at 2014-02-15T21:56:18+00:00 Add test case for inter-module type/data family instances These should show up in every place where the class is visible, and indeed they do right now. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 8bea5c3a by Mateusz Kowalczyk at 2014-02-19T05:11:34+00:00 Use a bespoke data type to indicate fixity This deals with what I imagine was an ancient TODO and makes it much clearer what the argument actually does rather than having the user chase down the comment. - - - - - 5b52d57c by Niklas Haas at 2014-02-22T21:31:03+01:00 Strip a single leading space from bird tracks (#201) This makes bird tracks in the form > foo > bar > bat parse as if they had been written as >foo >bar >bat ie. without the leading whitespace in front of every line. Ideally we also want to look into how leading whitespace affects code blocks written using the @ @ syntax, which are currently unaffected by this patch. - - - - - 5a1315a5 by Simon Hengel at 2014-02-22T21:55:35+01:00 Turn a source code comment into specs - - - - - 784cfe58 by Mateusz Kowalczyk at 2014-02-23T05:02:22+00:00 Update test case for lifted GADT type rendering The parsing of these seems to have been fixed by GHC folk and it now renders differently. IMHO it now renders in a better way so I'm updating the test to reflect this. - - - - - c3c88c2f by Mateusz Kowalczyk at 2014-02-23T06:37:14+00:00 Don't shadow ‘strip’. -Wall complains - - - - - 293031d8 by Niklas Haas at 2014-02-23T15:21:52+01:00 Make ImplicitParams render correctly (#260) This introduces a new precedence level for single contexts (because implicit param contexts always need parens around them, but other types of contexts don't necessarily, even when alone) - - - - - 4200842d by Niklas Haas at 2014-02-23T15:37:13+01:00 Lower precedence of equality constraints This drops them to the new precedence pREC_CTX, which makes single eqaulity constraints show up as (a ~ b) => ty, in line with GHC's rendering. Additional tests added to make sure other type operators render as intended. Current behavior matches GHC - - - - - b59e3227 by Niklas Haas at 2014-02-23T16:11:22+01:00 Add RankNTypes test case to ImplicitParams.hs This test actually tests what haskell/haddock#260 originally reported - I omitted the RankNTypes scenario from the original fix because I realized it's not relevant to the underlying issue and indeed, this renders as intended now. Still good to have more tests. - - - - - c373dbf7 by Mateusz Kowalczyk at 2014-02-24T06:09:54+00:00 Fix rendering of Contents when links are present Fixes Haddock Trac haskell/haddock#267. - - - - - 9ecb0e56 by Mateusz Kowalczyk at 2014-02-24T06:26:50+00:00 Fix wording in the docs - - - - - 4f4dcd8e by Mateusz Kowalczyk at 2014-02-27T03:00:33+00:00 Change rendering of duplicate record field docs See Haddock Trac haskell/haddock#195. We now change this behaviour to only rendering the documentation attached to the first instance of a duplicate field. Perhaps we could improve this by rendering the first instance that has documentation attached to it but for now, we'll stick with this. - - - - - ad8aa609 by Niklas Haas at 2014-03-08T09:43:26+01:00 Render fixity information Affects functions, type synonyms, type families, class names, data type names, constructors, data families, associated TFs/DFs, type synonyms, pattern synonyms and everything else I could think of. - - - - - 6a39c917 by Niklas Haas at 2014-03-09T07:43:39+01:00 Reorder topDeclElem to move the source/wiki links to the top They appear in the same position due to the float: right attribute but now they're always at the top of the box instead of at the bottom. - - - - - 2d34b3b4 by Niklas Haas at 2014-03-09T07:53:46+01:00 Use optLast instead of listToMaybe for sourceUrls/wikiUrls This lets you override them using eg. cabal haddock --haddock-options, which can come in handy if you want to use a different layout or URL for your source code links than cabal-install generates. - - - - - 0eff4624 by Niklas Haas at 2014-03-09T07:53:46+01:00 Differentiate between TH splices (line-links) and regular names This adds a new type of source code link, to a specific line rather than a specific declaration/name - this is used to link to the location of a TH splice that defines a certain name. Rather hefty changes throughout and still one unresolved issue (the line URLs aren't parsed from the third form of --read-interface which means they're currently restricted to same-interface links). Not sure if this issue is really worth all the hassle, especially since we could just use line links in general. This commit also contains some cleanup/clarification of the types in Haddock.Backends.Xhtml.Decl and shortens some overlong lines in the process. Notably, the Bool parameter was replaced by a Unicode type synonym to help clarify its presence in type signatures. - - - - - 66d6f77b by Niklas Haas at 2014-03-09T20:02:43+01:00 Group similar fixities together Identical fixities declared for the same line should now render using syntax like: infix 4 <, >=, >, <= - - - - - 6587f9f5 by Mateusz Kowalczyk at 2014-03-10T04:24:18+00:00 Update changelog - - - - - 7387ddad by Niklas Haas at 2014-03-11T10:26:04+01:00 Include fixity information in the Interface file This resolves fixity information not appearing across package borders. The binary file version has been increased accordingly. - - - - - ab46ef44 by Niklas Haas at 2014-03-11T10:26:04+01:00 Update changelog - - - - - 565cab6f by Niklas Haas at 2014-03-11T10:26:04+01:00 Update appearance of fixity annotations This moves them in-line with their corresponding lines, similar to a presentation envision by @hvr and described in #ghc. Redundant operator names are also omitted when no ambiguity is present. - - - - - 5d7afd67 by Niklas Haas at 2014-03-11T10:26:05+01:00 Filter family instances of hidden types Currently, this check does not extend to hidden right hand sides, although it probably should hide them in that case. - - - - - ec291b0c by Niklas Haas at 2014-03-11T10:26:05+01:00 Add documentation for --source-entity-line - - - - - 0922e581 by Niklas Haas at 2014-03-11T10:37:32+01:00 Revert "Reorder topDeclElem to move the source/wiki links to the top" This reverts commit 843c42c4179526a2ad3526e4c7d38cbf4d50001d. This change is no longer needed with the new rendering style, and it messes with copy/pasting lines. - - - - - 30618e8b by Mateusz Kowalczyk at 2014-03-11T09:41:07+00:00 Bump version to 2.15.0 - - - - - adf3f1bb by Mateusz Kowalczyk at 2014-03-11T09:41:09+00:00 Fix up some whitespace - - - - - 8905f57d by Niklas Haas at 2014-03-13T19:18:06+00:00 Hide RHS of TFs with non-exported right hand sides Not sure what to do about data families yet, since technically it would not make a lot of sense to display constructors that cannot be used by the user. - - - - - 5c44d5c2 by Niklas Haas at 2014-03-13T19:18:08+00:00 Add UnicodeSyntax alternatives for * and -> I could not find a cleaner way to do this other than checking for string equality with the given built-in types. But seeing as it's actually equivalent to string rewriting in GHC's implementation of UnicodeSyntax, it's probably fitting. - - - - - b04a63e6 by Niklas Haas at 2014-03-13T19:18:10+00:00 Display minimal complete definitions for type classes This corresponds to the new {-# MINIMAL #-} pragma present in GHC 7.8+. I also cleaned up some of the places in which ExportDecl is used to make adding fields easier in the future. Lots of test cases have been updated since they now render with minimality information. - - - - - a4a20b16 by Niklas Haas at 2014-03-13T19:18:12+00:00 Strip links from recently added html tests These were accidentally left there when the tests were originally added - - - - - d624f315 by Mateusz Kowalczyk at 2014-03-13T19:19:31+00:00 Update changelog - - - - - d27a21ac by Mateusz Kowalczyk at 2014-03-13T21:19:07+00:00 Always read in prologue files as UTF8 (#286). - - - - - 54b2fd78 by Mateusz Kowalczyk at 2014-03-13T21:28:09+00:00 Style only - - - - - fa4fe650 by Simon Hengel at 2014-03-15T09:04:18+01:00 Add Fuuzetsu maintainers field in cabal file - - - - - f83484b7 by Niklas Haas at 2014-03-15T18:20:24+00:00 Hide minimal definition for only-method classes Previously this was not covered by the All xs check since here it is not actually an All, rather a single Var n. This also adds the previously missing html-test/src/Minimal.hs. - - - - - 0099d276 by Niklas Haas at 2014-03-15T18:20:26+00:00 Fix issue haskell/haddock#281 This is a regression from the data family instances change. Data instances are now distinguished from regular lists by usage of the new class "inst", and the style has been updated to only apply to those. I've also updated the appropriate test case to test this a bit better, including GADT instances with GADT-style records. - - - - - 1f9687bd by Mateusz Kowalczyk at 2014-03-21T17:48:37+00:00 Please cabal sdist - - - - - 75542693 by Mateusz Kowalczyk at 2014-03-22T16:36:16+00:00 Drop needless --split-objs which slows us down. Involves tiny cleanup of all the dynflag bindings. Fixes haskell/haddock#292. - - - - - 31214dc3 by Herbert Valerio Riedel at 2014-03-23T18:01:01+01:00 Fix a few typos Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 0b73e638 by Mateusz Kowalczyk at 2014-03-31T05:34:36+01:00 Print kind signatures on GADTs - - - - - 2bab42f3 by Mateusz Kowalczyk at 2014-03-31T16:53:25+01:00 Add default for new PlatformConstraints field - - - - - 42647c5f by Mateusz Kowalczyk at 2014-03-31T18:29:04+01:00 Drop leading whitespace in @-style blocks. Fixes haskell/haddock#201. - - - - - 98208294 by Niklas Haas at 2014-03-31T20:09:58+02:00 Crash when exporting record selectors of data family instances This fixes bug haskell/haddock#294. This also fixes a related but never-before-mentioned bug about the display of GADT record selectors with non-polymorphic type signatures. Note: Associated data type constructors fail to show up if nothing is exported that they could be attached to. Exporting any of the data types in the instance head, or the class + data family itself, causes them to show up, but in the absence of either of these, exporting just the associated data type with the constructor itself will result in it being hidden. The only scenario I can come up that would involve this kind of situation involved OverlappingInstances, and even then it can be mitigated by just exporting the class itself, so I'm not going to solve it since the logic would most likely be very complicated. - - - - - 3832d171 by Mateusz Kowalczyk at 2014-04-01T19:07:33+01:00 Make CHANGES consistent with what's now in 2.14.2 - - - - - c386ae89 by Mateusz Kowalczyk at 2014-04-01T19:18:36+01:00 Actually bundle extra spec tests in sdist - - - - - bd57a6d3 by Mateusz Kowalczyk at 2014-04-03T21:13:48+01:00 Update test cases for GHC bug haskell/haddock#8945, Haddock haskell/haddock#188 The order of signature groups has been corrected upstream. Here we add a test case and update some existing test-cases to reflect this change. We remove grouped signature in test cases that we can (Minimal, BugDeprecated &c) so that the test is as self-contained as possible. - - - - - 708b88b1 by Mateusz Kowalczyk at 2014-04-03T21:16:07+01:00 Enforce strict GHC version in cabal file This stops people with 7.6.3 trying to install 2.15.x which clearly won't work. Unfortunately we shipped 2.14.x without realising this. - - - - - 60334f7c by Mateusz Kowalczyk at 2014-04-03T21:19:24+01:00 Initialise some new PlatformConstants fields - - - - - ea77f668 by Mateusz Kowalczyk at 2014-04-11T16:52:23+01:00 We don't actually want unicode here - - - - - 0b651cae by Mateusz Kowalczyk at 2014-04-11T18:13:30+01:00 Parse identifiers with ^ and ⋆ in them. Fixes haskell/haddock#298. - - - - - e8ad0f5f by Mateusz Kowalczyk at 2014-04-11T18:47:41+01:00 Ignore version string during HTML tests. - - - - - de489089 by Mateusz Kowalczyk at 2014-04-11T18:59:30+01:00 Update CHANGES to follow 2.14.3 - - - - - beb464a9 by Gergő Érdi at 2014-04-13T16:31:10+08:00 remove Origin flag from LHsBindsLR - - - - - cb16f07c by Herbert Valerio Riedel at 2014-04-21T17:16:50+02:00 Replace local `die` by new `System.Exit.die` Starting with GHC 7.10, System.Exit exports the new `die` which is essentially the same as Haddock.Util.die, so this commit changes Haddock.Util.die to be a simple re-export of System.Exit.die. See also https://ghc.haskell.org/trac/ghc/ticket/9016 for more details. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 9b9b23c7 by Mateusz Kowalczyk at 2014-05-03T15:40:11+02:00 Disambiguate ‘die’ in test runners. - - - - - 5d28a2b8 by Mateusz Kowalczyk at 2014-05-05T09:19:49+02:00 Prepare modules for parser split. We have to generalise the Doc (now DocH) slightly to remove the dependency on GHC-supplied type. - - - - - d3967ff3 by Mateusz Kowalczyk at 2014-05-05T11:00:41+02:00 Move parser + parser tests out to own package. We move some types out that are necessary as well and then re-export and specialise them in the core Haddock. Reason for moving out spec tests is that if we're working on the parser, we can simply work on that and we can ignore the rest of Haddock. The downside is that it's a little inconvenient if at the end of the day we want to see that everything passes. - - - - - 522a448d by Mateusz Kowalczyk at 2014-05-05T11:14:47+02:00 Move out Show and Eq instances to Types They are much more useful to the users here. - - - - - 11a6f0f2 by Mateusz Kowalczyk at 2014-05-06T13:50:31+02:00 Remove no longer necessary parser error handling. We can now drop some Maybe tests and even lets us strip an error handling monad away in a few places. - - - - - 6992c924 by Mateusz Kowalczyk at 2014-05-14T02:23:55+02:00 Please the GHC build-system. As I can not figure out how to do this properly, if we're in GHC tree, we treat the library as being the same package. If we're not in the tree, we require that the library be installed separately. - - - - - 7a8ad763 by Mateusz Kowalczyk at 2014-05-14T14:50:25+02:00 Update issue tracker URL - - - - - f616c521 by Mateusz Kowalczyk at 2014-05-14T14:53:32+02:00 Update issue tracker URL for haddock-library - - - - - 66580ded by Gergő Érdi at 2014-05-25T14:24:16+08:00 Accomodate change in PatSyn representation - - - - - 0e43b988 by Mateusz Kowalczyk at 2014-05-29T03:15:29+02:00 Revert "Accomodate change in PatSyn representation" This reverts commit 57aa591362d7c8ba21285fccd6a958629a422091. I am reverting this because I pushed it to master when it was meant to stay on a wip-branch. Sorry Gergo and everyone who had trouble due to this. - - - - - e10d7ec8 by Mateusz Kowalczyk at 2014-05-29T03:24:11+02:00 Revert "Revert "Accomodate change in PatSyn representation"" This reverts commit e110e6e70e40eed06c06676fd2e62578da01d295. Apparently as per GHC commit ac2796e6ddbd54c5762c53e2fcf29f20ea162fd5 this was actually intended. Embarrasing for me. - - - - - 5861aca9 by Mateusz Kowalczyk at 2014-06-05T19:49:27+02:00 Clear up highlighting of identifiers with ‘'’s. - - - - - d7cc420f by Simon Peyton Jones at 2014-06-06T12:41:09+01:00 Follow change in patSynSig - - - - - 938b4fd8 by Mateusz Kowalczyk at 2014-06-12T07:24:29+02:00 Slightly update the readme. Style-sheets are no longer a recent thing, dead links, old maintainers, different formats. - - - - - c7799dea by Mateusz Kowalczyk at 2014-06-18T00:05:56+02:00 Update cabal files Update repository urls, use subdir property for haddock-library and use a separate versioning scheme for haddock-library in preparation for release. - - - - - a2750b6a by Simon Hengel at 2014-06-18T11:01:18+08:00 Compatibility with older versions of base and bytestring - - - - - 009b4b03 by Simon Hengel at 2014-06-18T11:14:01+08:00 Enable travis-ci for haddock-library - - - - - 9b5862eb by Simon Hengel at 2014-06-18T11:14:01+08:00 haddock-library: Do not depend on haddock-library in test suite I think you either add src to hs-source-dirs or the library to build-depends. But doing both does not make sense (AFAICT). - - - - - fb1f3279 by Simon Hengel at 2014-06-18T11:49:05+08:00 haddock-library: Use -Wall for specs - - - - - 649340e1 by Mateusz Kowalczyk at 2014-06-18T06:58:54+02:00 Use Travis with multiple GHC versions When using HEAD, we build haddock-library directly from repository as a dependency (and thanks to --enable-tests, the tests get ran anyway). In all other cases, we manually run the tests on haddock-library only and don't test the main project. - - - - - d7eeeec2 by Mateusz Kowalczyk at 2014-06-18T07:49:04+02:00 Comment improvements + few words in cabal file - - - - - 0f8db914 by Simon Hengel at 2014-06-18T13:52:23+08:00 Use doctest to check examples in documentation - - - - - 2888a8dc by Simon Hengel at 2014-06-18T14:16:48+08:00 Remove doctest dependency (so that we can use haddock-library with doctest) - - - - - 626d5e85 by Mateusz Kowalczyk at 2014-06-18T08:41:25+02:00 Travis tweaks - - - - - 41d4f9cc by Mateusz Kowalczyk at 2014-06-18T08:58:43+02:00 Don't actually forget to install specified GHC. - - - - - c6aa512a by John MacFarlane at 2014-06-18T10:43:57-07:00 Removed reliance on LambdaCase (which breaks build with ghc 7.4). - - - - - b9b93b6f by John MacFarlane at 2014-06-18T10:54:56-07:00 Fixed haddock warnings. - - - - - a41b0ab5 by Mateusz Kowalczyk at 2014-06-19T01:20:10+02:00 Update Travis, bump version - - - - - 864bf62a by Mateusz Kowalczyk at 2014-06-25T10:36:54+02:00 Fix anchors. Closes haskell/haddock#308. - - - - - 53df91bb by Mateusz Kowalczyk at 2014-06-25T15:04:49+02:00 Drop DocParagraph from front of headers I can not remember why they were wrapped in paragraphs to begin with and it seems unnecessary now that I test it. Closes haskell/haddock#307. - - - - - 29b5f2fa by Mateusz Kowalczyk at 2014-06-25T15:17:20+02:00 Don't mangle append order for nested lists. The benefit of this is that the ‘top-level’ element of such lists is properly wrapped in <p> tags so any CSS working with these will be applied properly. It also just makes more sense. Pointed out at jgm/pandoc#1346. - - - - - 05cb6e9c by Mateusz Kowalczyk at 2014-06-25T15:19:45+02:00 Bump haddock-library to 1.1.0 for release - - - - - 70feab15 by Iavor Diatchki at 2014-07-01T03:37:07-07:00 Propagate overloading-mode for instance declarations in haddock (#9242) - - - - - d4ca34a7 by Simon Peyton Jones at 2014-07-14T16:23:15+01:00 Adapt to new definition of HsDecls.TyFamEqn This is a knock-on from the refactoring from Trac haskell/haddock#9063. I'll push the corresponding changes to GHC shortly. - - - - - f91e2276 by Edward Z. Yang at 2014-07-21T08:14:19-07:00 Track GHC PackageId to PackageKey renaming. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - b010f9ef by Edward Z. Yang at 2014-07-25T16:28:46-07:00 Track changes for module reexports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - 8b85f9f9 by Mateusz Kowalczyk at 2014-07-28T13:25:43+02:00 Catch mid-line URLs. Fixes haskell/haddock#314. - - - - - 4c613a78 by Edward Z. Yang at 2014-08-05T03:11:00-07:00 Track type signature change of lookupModuleInAllPackages Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - e80b051c by Edward Z. Yang at 2014-08-05T17:34:26+01:00 If GhcProfiled, also build Haddock profiled. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f9cccd29 by Edward Z. Yang at 2014-08-07T14:23:35+01:00 Ignore TAGS files. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 00b3af52 by Mateusz Kowalczyk at 2014-08-08T04:58:19+02:00 Update to attoparsec-0.12.1.1 There seems to be memory and speed improvement. - - - - - 5457dc71 by Mateusz Kowalczyk at 2014-08-08T18:24:02+02:00 Fix forgotten src - - - - - 3520cb04 by Mateusz Kowalczyk at 2014-08-14T20:19:07+01:00 Bump down the version for master to 2.14.4 - - - - - dc98c21b by Mateusz Kowalczyk at 2014-08-14T20:23:27+01:00 Revert "Track type signature change of lookupModuleInAllPackages" This reverts commit d59fec2c9551b5662a3507c0011e32a09a9c118f. - - - - - 3f2038c0 by Mateusz Kowalczyk at 2014-08-14T20:23:31+01:00 Revert "Track changes for module reexports." This reverts commit b99b57c0df072d12b67816b45eca2a03cb1da96d. - - - - - 56d4e49e by Mateusz Kowalczyk at 2014-08-14T20:23:42+01:00 Revert "Track GHC PackageId to PackageKey renaming." This reverts commit 8ac42d3327473939c013551750425cac191ff0fd. - - - - - 726ea3cb by Mateusz Kowalczyk at 2014-08-14T20:23:47+01:00 Revert "Adapt to new definition of HsDecls.TyFamEqn" This reverts commit cb96b4f1ed0462b4a394b9fda6612c3bea9886bd. - - - - - 61a88ff0 by Mateusz Kowalczyk at 2014-08-14T20:23:52+01:00 Revert "Propagate overloading-mode for instance declarations in haddock (#9242)" This reverts commit 8d20ca8d5a9bee73252ff2035ec45f9c03d0820c. - - - - - a32ba674 by Mateusz Kowalczyk at 2014-08-14T20:26:03+01:00 Revert "Disambiguate ‘die’ in test runners." This reverts commit dba02d6df32534aac5d257f2d28596238d248942. - - - - - f335820f by Mateusz Kowalczyk at 2014-08-14T20:26:09+01:00 Revert "Replace local `die` by new `System.Exit.die`" This reverts commit 08aa509ebac58bfb202ea79c7c41291ec280a1c5. - - - - - 107078e4 by Mateusz Kowalczyk at 2014-08-14T20:27:34+01:00 Merge branch 'reverts' This reverts any changes that were made to have Haddock compile with 7.9. When 7.10 release comes, we can simply re-apply all the patches and any patches that occur on ghc-head branch from now on. This allows us to build master with 7.8.3 - - - - - b44b3871 by Mateusz Kowalczyk at 2014-08-15T02:47:40+01:00 Fix haskell/haddock#313 by doing some list munging. I get rid of the Monoid instance because we weren't satisfying the laws. Convenience of having <> didn't outweigh the shock-factor of having it behave badly. - - - - - e1a62cde by Mateusz Kowalczyk at 2014-08-15T02:52:56+01:00 Stop testing haskell/haddock#188. Because the change is in GHC 7.9 and we now work against 7.8.3, this test no longer makes sense. We revert it until 7.10 becomes the standard version. If anything, there should be a test for this in GHC itself. - - - - - 54e8286d by Mateusz Kowalczyk at 2014-08-15T05:31:57+01:00 Add haskell/haddock#313 to CHANGES - - - - - 9df7ad5d by Simon Hengel at 2014-08-20T11:25:32+08:00 Fix warning - - - - - ee2574d6 by Simon Hengel at 2014-08-20T12:07:01+08:00 Fix travis builds - - - - - 384cf2e6 by Simon Hengel at 2014-08-20T12:14:31+08:00 Require GHC 7.8.3 - - - - - d4779863 by Simon Hengel at 2014-08-22T12:14:16+08:00 Move Haddock API to a separate package - - - - - 80f3e0e1 by Simon Hengel at 2014-08-22T14:57:38+08:00 Bump version to 2.15.0 and add version constraints - - - - - 309a94ce by Simon Hengel at 2014-08-22T15:18:06+08:00 Add deprecated compatibility module - - - - - 4d1e4e3f by Luite Stegeman at 2014-08-22T20:46:45+02:00 export things to allow customizing how the Ghc session is run - - - - - 47884591 by Luite Stegeman at 2014-08-22T20:46:51+02:00 ghc 7.8.2 compatibility - - - - - 5ea94e2c by Luite Stegeman at 2014-08-22T22:08:58+02:00 install dependencies for haddock-api on travis - - - - - 9fb845b2 by Mateusz Kowalczyk at 2014-08-23T10:09:34+01:00 Move sources under haddock-api/src - - - - - 85817dc4 by Mateusz Kowalczyk at 2014-08-23T10:10:48+01:00 Remove compat stuff - - - - - 151c6169 by Niklas Haas at 2014-08-24T08:14:10+02:00 Fix extra whitespace on signatures and update all test cases This was long overdue, now running ./accept.lhs on a clean test from master will not generate a bunch of changes. - - - - - d320e0d2 by Niklas Haas at 2014-08-24T08:14:35+02:00 Omit unnecessary foralls and fix haskell/haddock#315 This also fixes haskell/haddock#86. - - - - - bdafe108 by Mateusz Kowalczyk at 2014-08-24T15:06:46+01:00 Update CHANGES - - - - - fafa6d6e by Mateusz Kowalczyk at 2014-08-24T15:14:23+01:00 Delete few unused/irrelevant/badly-place files. - - - - - 3634923d by Duncan Coutts at 2014-08-27T13:49:31+01:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. - - - - - 539a7e70 by Herbert Valerio Riedel at 2014-08-31T11:36:32+02:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 9e3a0e5b by Mateusz Kowalczyk at 2014-08-31T12:54:43+01:00 Bump version in doc - - - - - 4a177525 by Mateusz Kowalczyk at 2014-08-31T13:01:23+01:00 Bump haddock-library version - - - - - f99c1384 by Mateusz Kowalczyk at 2014-08-31T13:05:25+01:00 Remove references to deleted files - - - - - 5e51a247 by Mateusz Kowalczyk at 2014-08-31T14:18:44+01:00 Make the doc parser not complain - - - - - 2cedb49a by Mateusz Kowalczyk at 2014-09-03T03:33:15+01:00 CONTRIBUTING file for issues - - - - - 88027143 by Mateusz Kowalczyk at 2014-09-04T00:46:59+01:00 Mention --print-missing-docs - - - - - 42f6754f by Alan Zimmerman at 2014-09-05T18:13:24-05:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e712719e by Austin Seipp at 2014-09-09T01:03:27-05:00 Fix import of 'empty' due to AMP. Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 71c29755 by Herbert Valerio Riedel at 2014-09-09T17:35:20+02:00 Bump `base` constraint for AMP - - - - - 0bf9f3ed by Mateusz Kowalczyk at 2014-09-12T19:18:32+01:00 Delete stale ANNOUNCE - - - - - cac89ee6 by Krzysztof Gogolewski at 2014-09-14T17:17:09+02:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 4d683426 by Edward Z. Yang at 2014-09-18T13:38:11-07:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 80697fd5 by Herbert Valerio Riedel at 2014-09-19T00:07:52+02:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. - - - - - c015eb70 by Herbert Valerio Riedel at 2014-09-19T00:10:36+02:00 Revert "Followup changes to addition of -fwarn-context-quantification" This reverts commit 4023817d7c0e46db012ba2eea28022626841ca9b temporarily as the respective feature hasn't landed in GHC HEAD yet, but this commit blocks later commits from being referenced in GHC HEAD. - - - - - 38ded784 by Edward Z. Yang at 2014-09-18T15:32:15-07:00 Revert "Revert "Followup changes to addition of -fwarn-context-quantification"" This reverts commit db14fd8ab4fab43694139bc203808b814eafb2dc. It's in HEAD now. - - - - - f55d59c9 by Herbert Valerio Riedel at 2014-09-26T19:18:28+02:00 Revert "Fix import of 'empty' due to AMP." This reverts commit 0cc5bc85e9fca92ab712b68a2ba2c0dd9d3d79f4 since it turns out we don't need to re-export `empty` from Control.Monad after all. - - - - - 467050f1 by David Feuer at 2014-10-09T20:07:36-04:00 Fix improper lazy IO use Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. - - - - - cc47b699 by Edward Z. Yang at 2014-10-09T21:38:19-07:00 Fix use-after-close lazy IO bug Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. Signed-off-by: David Feuer <David.Feuer at gmail.com> Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 87babcbe by Austin Seipp at 2014-10-20T20:05:27-05:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - ab259516 by Austin Seipp at 2014-10-20T20:07:01-05:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - b918093c by Mateusz Kowalczyk at 2014-10-29T03:59:39+00:00 Experimental support for collapsable headers Closes haskell/haddock#335 - - - - - 849db129 by Mateusz Kowalczyk at 2014-10-29T10:07:26+01:00 Experimental support for collapsable headers (cherry picked from commit e2ed3b9d8dfab09f1b1861dbc8e74f08e137ebcc) - - - - - a4cc4789 by Herbert Valerio Riedel at 2014-10-31T11:08:26+01:00 Collapse user-defined section by default (re haskell/haddock#335) - - - - - 9da1b33e by Yuras Shumovich at 2014-10-31T16:11:04-05:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - c625aefc by Austin Seipp at 2014-10-31T19:34:10-05:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - c7738e5e by Simon Hengel at 2014-11-02T07:25:30+08:00 Remove -fobject-code from .ghci (this slows down reloads on modifications) - - - - - d4a86e95 by Simon Hengel at 2014-11-03T09:26:11+08:00 Get rid of StandaloneDeriving - - - - - a974e311 by Simon Hengel at 2014-11-03T09:26:11+08:00 Derive more instances - - - - - 8aa0c4d7 by Simon Hengel at 2014-11-03T09:27:08+08:00 Remove unused language extensions - - - - - 3052d46a by Simon Hengel at 2014-11-03T09:30:46+08:00 Minor refactoring - - - - - 4281d3cb by Simon Hengel at 2014-11-03T09:30:46+08:00 parser: Try to parse definition lists right before text paragraphs - - - - - 8ba12bf9 by Simon Hengel at 2014-11-03T09:34:19+08:00 Add support for markdown links (closes haskell/haddock#336) - - - - - a2f8d747 by Simon Hengel at 2014-11-03T09:34:19+08:00 Allow markdown links at the beginning of a paragraph - - - - - 53b11207 by Simon Hengel at 2014-11-03T09:34:20+08:00 Update documentation - - - - - 652267c6 by Simon Hengel at 2014-11-03T09:34:20+08:00 Add support for markdown images - - - - - 9d667502 by Simon Hengel at 2014-11-03T09:34:20+08:00 Allow an optional colon after the closing bracket of definition lists This is to disambiguate them from markdown links and will be require with a future release. - - - - - 8167fc32 by Mateusz Kowalczyk at 2014-11-04T01:16:51+00:00 whitespace only - - - - - 3da62981 by Mateusz Kowalczyk at 2014-11-04T01:17:31+00:00 Fix re-exports of built-in type families Fixes haskell/haddock#310 - - - - - edc76b34 by Mateusz Kowalczyk at 2014-11-04T02:54:28+00:00 Turn some uses of error into recoverable warnings This should at the very least not abort when something weird happens. It does feel like we should have a type that carries these errors until the end however as the user might not see them unless they are printed at the end. - - - - - 0a137400 by Mateusz Kowalczyk at 2014-11-04T04:09:44+00:00 Fix warnings - - - - - d068fc21 by Mateusz Kowalczyk at 2014-11-04T21:04:07+00:00 Fix parsing of identifiers written in infix way - - - - - 1a9f2f3d by Simon Hengel at 2014-11-08T11:32:42+08:00 Minor code simplification - - - - - 6475e9b1 by Simon Hengel at 2014-11-08T17:28:33+08:00 newtype-wrap parser monad - - - - - dc1ea105 by Herbert Valerio Riedel at 2014-11-15T11:55:43+01:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - fbb1aca4 by Simon Hengel at 2014-11-16T08:51:38+08:00 State intention rather than implementation details in Haddock comment - - - - - 97851ab2 by Simon Hengel at 2014-11-16T10:20:19+08:00 (wip) Add support for @since (closes haskell/haddock#26) - - - - - 34bcd18e by Gergő Érdi at 2014-11-20T22:35:38+08:00 Update Haddock to new pattern synonym type signature syntax - - - - - 304b7dc3 by Jan Stolarek at 2014-11-20T17:48:43+01:00 Follow changes from haskell/haddock#9812 - - - - - 920f9b03 by Richard Eisenberg at 2014-11-20T16:52:50-05:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 0bfe4e78 by Alan Zimmerman at 2014-11-21T11:23:09-06:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 356ed45a by Thomas Winant at 2014-11-28T16:11:22-06:00 Support for PartialTypeSignatures - - - - - 5dc8f3b1 by Gergő Érdi at 2014-11-29T15:39:09+08:00 For pattern synonyms, render "pattern" as a keyword - - - - - fe704480 by Mateusz Kowalczyk at 2014-12-09T03:38:32+00:00 List new module in cabal file - - - - - b9ad5a29 by Mateusz Kowalczyk at 2014-12-10T00:58:24+00:00 Allow the parser to spit out meta-info Currently we only use it only for ‘since’ annotations but with these patches it should be fairly simple to add new attributes if we wish to. Closes haskell/haddock#26. It seems to work fine but due to 7.10 rush I don't have the chance to do more exhaustive testing right now. The way the meta is output (emphasis at the end of the whole comment) is fairly arbitrary and subject to bikeshedding. Note that this makes test for Bug310 fail due to interface version bump: it can't find the docs for base with this interface version so it fails. There is not much we can do to help this because it tests for ’built-in’ identifier, not something we can provide ourselves. - - - - - 765af0e3 by Mateusz Kowalczyk at 2014-12-10T01:17:19+00:00 Update doctest parts of comments - - - - - 8670272b by jpmoresmau at 2014-12-10T01:35:31+00:00 header could contain several lines Closes haskell/haddock#348 - - - - - 4f9ae4f3 by Mateusz Kowalczyk at 2014-12-12T06:22:31+00:00 Revert "Merge branch 'reverts'" This reverts commit 5c93cc347773c7634321edd5f808d5b55b46301f, reversing changes made to 5b81a9e53894d2ae591ca0c6c96199632d39eb06. Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - e974ac94 by Duncan Coutts at 2014-12-12T06:26:11+00:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. Conflicts: haddock-api/src/Haddock.hs - - - - - 2f3a2365 by Herbert Valerio Riedel at 2014-12-12T06:26:51+00:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 1dbd6390 by Alan Zimmerman at 2014-12-12T06:32:07+00:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - bb6ff1f4 by Mateusz Kowalczyk at 2014-12-12T06:35:07+00:00 Bump ‘base’ constraint Follows the similar commit made on ghc-head branch - - - - - 466fe4ab by Krzysztof Gogolewski at 2014-12-12T06:37:42+00:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 97e080c9 by Edward Z. Yang at 2014-12-12T06:39:35+00:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: haddock-api/src/Haddock/ModuleTree.hs - - - - - 20b2af56 by Herbert Valerio Riedel at 2014-12-12T06:42:50+00:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. Conflicts: haddock-library/src/Documentation/Haddock/Parser.hs - - - - - b3ad269d by Austin Seipp at 2014-12-12T06:44:14+00:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 072df0dd by Austin Seipp at 2014-12-12T06:45:01+00:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - dbb9294a by Herbert Valerio Riedel at 2014-12-12T06:46:17+00:00 Collapse user-defined section by default (re haskell/haddock#335) Conflicts: haddock-api/src/Haddock/Backends/Xhtml/DocMarkup.hs - - - - - f23ab545 by Yuras Shumovich at 2014-12-12T06:46:41+00:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - 753a4b67 by Austin Seipp at 2014-12-12T06:46:51+00:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8954e8f5 by Herbert Valerio Riedel at 2014-12-12T06:50:53+00:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - d2b06d61 by Gergő Érdi at 2014-12-12T07:07:30+00:00 Update Haddock to new pattern synonym type signature syntax Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 1ff02426 by Jan Stolarek at 2014-12-12T07:13:24+00:00 Follow changes from haskell/haddock#9812 Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - 06ad7600 by Richard Eisenberg at 2014-12-12T07:13:43+00:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 8fd2aa8b by Alan Zimmerman at 2014-12-12T07:22:25+00:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Backends/LaTeX.hs haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 95c3db98 by Thomas Winant at 2014-12-12T07:35:49+00:00 Support for PartialTypeSignatures Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs haddock-api/src/Haddock/Interface/Create.hs - - - - - 45494428 by Gergő Érdi at 2014-12-12T07:36:18+00:00 For pattern synonyms, render "pattern" as a keyword - - - - - a237e3eb by Mateusz Kowalczyk at 2014-12-12T12:27:13+00:00 Various fixups and bumps for next release - - - - - 22918bcd by Herbert Valerio Riedel at 2014-12-14T10:11:47+01:00 Remove redundant wild-card pattern match (this would otherwise cause a build-failure with `-Werror`) - - - - - 1d6ce947 by Herbert Valerio Riedel at 2014-12-14T10:17:06+01:00 Treat GHC 7.10 the same as GHC 7.9 ...since the current GHC 7.9 is going to become GHC 7.10 real-soon-now anyway - - - - - f434ea89 by Herbert Valerio Riedel at 2014-12-14T18:26:50+01:00 Fixup ghc.mk (follow-up to 1739375eb23342) This makes the GHC build-system aware of the data-files to be copied into the bindist (as haddock.cabal doesn't list those anymore) - - - - - 6fb839eb by Mateusz Kowalczyk at 2014-12-17T09:28:59+00:00 Only keep one Version instead of blindly appending - - - - - 40645489 by Mateusz Kowalczyk at 2014-12-18T07:09:44+00:00 Fix dependency version - - - - - 8b3b927b by Mateusz Kowalczyk at 2014-12-18T07:14:23+00:00 Print missing docs by default Adds --no-print-missing-docs - - - - - 59666694 by Mateusz Kowalczyk at 2014-12-18T07:21:37+00:00 update changelog - - - - - aa6d168e by Mateusz Kowalczyk at 2014-12-18T07:30:58+00:00 Update docs for @since - - - - - 2d7043ee by Luite Stegeman at 2014-12-19T18:29:35-06:00 hide projectVersion from DynFlags since it clashes with Haddock.Version.projectVersion - - - - - aaa70fc0 by Luite Stegeman at 2014-12-22T15:58:43+01:00 Add missing import for standalone haddock-api package - - - - - 9ce01269 by Herbert Valerio Riedel at 2014-12-22T17:48:45+01:00 Reset ghc-head with master's tree (this is an overwriting git merge of master into ghc-head) - - - - - fcd6fec1 by Herbert Valerio Riedel at 2014-12-22T17:51:52+01:00 Bump versions for ghc-7.11 - - - - - 525ec900 by Mateusz Kowalczyk at 2014-12-23T13:36:24+00:00 travis-ci: test with HEAD - - - - - cbf494b5 by Simon Peyton Jones at 2014-12-23T15:22:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 50e01c99 by Mateusz Kowalczyk at 2014-12-29T15:28:47+00:00 Make travis use 7.10.x - - - - - 475e60b0 by Njagi Mwaniki at 2014-12-29T15:30:44+00:00 Turn the README into GitHub Markdown format. Closes haskell/haddock#354 - - - - - 8cacf48e by Luite Stegeman at 2015-01-05T16:25:37+01:00 bump haddock-api ghc dependency to allow release candidate and first release - - - - - 6ed6cf1f by Simon Peyton Jones at 2015-01-06T16:37:47+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 8b484f33 by Simon Peyton Jones at 2015-01-08T15:50:22+00:00 Track naming change in DataCon - - - - - 23c5c0b5 by Alan Zimmerman at 2015-01-16T10:15:11-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e7a5532c by JP Moresmau at 2015-01-22T17:19:03+00:00 Ignore warnings, install Cabal 1.22 - - - - - 86942c84 by jpmoresmau at 2015-01-22T17:19:04+00:00 solve dataDir ambiguity - - - - - 5ceb743e by jpmoresmau at 2015-01-22T19:17:32+00:00 support GHC 7.10: no Safe-Inferred, Foldable instance - - - - - 6a3b3fb5 by Mateusz Kowalczyk at 2015-01-22T19:32:10+00:00 Update test files Test: a correct behaviour for fields comma-separating values. I'm surprised we had no bug open for this. Maybe it affects how haskell/haddock#301 renders now but I doubt. Operators: Seems GHC is giving us a new order for operators, something must have changed on their side again. cc @haasn , this makes the fixity to the side not match the order on the LHS which is a bit unpleasant. Maybe the fixity can be made to match the GHC order? Bug335: We expand examples by default now. Bug310: Now inferred safe. - - - - - 708f8b2f by jpmoresmau at 2015-01-22T19:36:59+00:00 Links to source location of class instance definitions - - - - - 5cf8a6da by Vincent Berthoux at 2015-01-22T19:59:58+00:00 Filter '\r' from comments due to Windows problems. On Windows this was causing newline to be rendered twice in code blocks. Closes haskell/haddock#359, fixes haskell/haddock#356. - - - - - 1749e6f0 by Mateusz Kowalczyk at 2015-01-22T20:31:27+00:00 Changelog only - - - - - c8145f90 by Mateusz Kowalczyk at 2015-01-22T23:34:05+00:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. - - - - - 14248254 by Mateusz Kowalczyk at 2015-01-22T23:43:18+00:00 Sort out some module import warnings - - - - - d8a38989 by Simon Peyton Jones at 2015-01-23T07:10:16-06:00 Track naming change in DataCon (cherry picked from commit 04cf63d0195837ed52075ed7d2676e71831e8a0b) - - - - - d3ac6ae4 by Alan Zimmerman at 2015-01-23T07:17:19-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit d61bbc75890e4eb0ad508b9c2a27b91f691213e6) - - - - - 4c1ffeb0 by Simon Peyton Jones at 2015-02-10T12:10:33+00:00 Track changes in HsSyn for quasi-quotes - - - - - 775d20f7 by Mateusz Kowalczyk at 2015-03-15T08:11:48+01:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. (cherry picked from commit 8e06728afb0784128ab2df0be7a5d7a191d30ff4) - - - - - f9245e72 by Phil Ruffwind at 2015-03-16T04:32:01-04:00 Prevent Synopsis from using up too much horizontal space When long type signatures occur in the Synopsis, the element is stretched beyond the width of the window. Scrollbars don't appear, so it's impossible to read anything when this happens. - - - - - cd8fa415 by Mateusz Kowalczyk at 2015-03-17T21:59:39+00:00 Update changelog Closes haskell/haddock#151 due to 71170fc77962f10d7d001e3b8bc8b92bfeda99bc - - - - - b5248b47 by Ben Gamari at 2015-03-25T17:12:17+00:00 Make the error encountered when a package can't be found more user-friendly Closes haskell/haddock#369 - - - - - b756b772 by Mateusz Kowalczyk at 2015-03-26T16:31:40+00:00 Remove now redundant imports - - - - - 5ea5e8dd by Mateusz Kowalczyk at 2015-03-26T16:45:52+00:00 Update test to account for \r filtering - - - - - 6539bfb3 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Test for anchor defaulting I delete the old tests because it turns out that: * test runner would never put them in scope of each other even with imports so just one would suffice * test runner actually needed some hacking to keep links so in the end we would end up with no anchors making them useless - - - - - 1a01d950 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Clearly default to variables in out of scope case - - - - - 7943abe8 by Mateusz Kowalczyk at 2015-03-27T01:14:11+00:00 Fix Hoogle display of constructors Fixes haskell/haddock#361 - - - - - 6d6e587e by Mateusz Kowalczyk at 2015-03-27T01:45:18+00:00 Fully qualify names in Hoogle instances output Closes haskell/haddock#263 - - - - - 52dac365 by Mateusz Kowalczyk at 2015-03-27T01:55:01+00:00 Update changelog - - - - - ca5af9a8 by Mateusz Kowalczyk at 2015-03-27T02:43:55+00:00 Output method documentation in Hoogle backend One thing of note is that we no longer preserve grouping of methods and print each method on its own line. We could preserve it if no documentation is present for any methods in the group if someone asks for it though. Fixes haskell/haddock#259 - - - - - a33f0c10 by Mateusz Kowalczyk at 2015-03-27T03:04:21+00:00 Don't print instance safety information in Hoogle Fixes haskell/haddock#168 - - - - - df6c935a by Mateusz Kowalczyk at 2015-03-28T00:11:47+00:00 Post-release version bumps and changelog - - - - - dde8f7c0 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Loosen bounds on haddock-* - - - - - de93bf89 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Expand response files in arguments Closes haskell/haddock#285 - - - - - 1f0b0856 by Zejun Wu at 2015-04-26T16:35:35-07:00 Do not insert anchor for section headings in contents box - - - - - 860439d7 by Simon Peyton Jones at 2015-05-01T09:36:47+01:00 Track change in API of TyCon - - - - - a32f3e5f by Adam Gundry at 2015-05-04T15:32:59+01:00 Track API changes to support empty closed type familes - - - - - 77e98bee by Ben Gamari at 2015-05-06T20:17:08+01:00 Ignore doc/haddock.{ps,pdf} - - - - - 663d0204 by Murray Campbell at 2015-05-11T04:47:37-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8bb0dcf5 by Murray Campbell at 2015-05-11T06:35:06-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit 2380f07c430c525b205ce2eae6dab23c8388d899) - - - - - bad900ea by Adam Bergmark at 2015-05-11T15:29:39+01:00 haddock-library: require GHC >= 7.4 `Data.Monoid.<>` was added in base-4.5/GHC-7.4 Closes haskell/haddock#394 Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - daceff85 by Simon Peyton Jones at 2015-05-13T12:04:21+01:00 Track the new location of setRdrNameSpace - - - - - 1937d1c4 by Alan Zimmerman at 2015-05-25T21:27:15+02:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - ee0fb6c2 by Łukasz Hanuszczak at 2015-05-27T11:51:31+02:00 Create simple method for indentation parsing. - - - - - 7d6fcad5 by Łukasz Hanuszczak at 2015-05-27T21:36:13+02:00 Make nested lists count indentation according to first item. - - - - - d6819398 by Łukasz Hanuszczak at 2015-05-27T22:46:13+02:00 Add simple test case for arbitrary-depth list nesting. - - - - - 2929c54d by Łukasz Hanuszczak at 2015-06-03T02:11:31+02:00 Add arbitrary-indent spec test for parser. - - - - - 9a0a9bb0 by Mateusz Kowalczyk at 2015-06-03T05:25:29+01:00 Update docs with info on new list nesting rule Fixes haskell/haddock#278 through commits from PR haskell/haddock#401 - - - - - 12efc92c by Mateusz Kowalczyk at 2015-06-03T05:29:26+01:00 Update some meta data at the top of the docs - - - - - 765ee49f by Bartosz Nitka at 2015-06-07T08:40:59-07:00 Add some Hacking docs for getting started - - - - - 19aaf851 by Bartosz Nitka at 2015-06-07T08:44:30-07:00 Fix markdown - - - - - 2a90cb70 by Mateusz Kowalczyk at 2015-06-08T15:08:36+01:00 Refine hacking instructions slightly - - - - - 0894da6e by Thomas Winant at 2015-06-08T23:47:28-05:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 10a9bb76 by Emanuel Borsboom at 2015-06-12T02:46:23+01:00 Build executable with '-threaded' (fixes haskell/haddock#399) - - - - - 7696b94f by Mateusz Kowalczyk at 2015-06-12T02:59:19+01:00 Update changelog for -threaded Closes haskell/haddock#400 - - - - - d3c118ec by Bartosz Nitka at 2015-06-12T03:00:58+01:00 Fix haddock: internal error: spliceURL UnhelpfulSpan (#207) Inferred type signatures don't have SrcSpans, so let's use the one from the declaration. I've tested this manually on the test-case from haskell/haddock#207, but I got stuck at trying to run the test-suite. - - - - - b67e843b by Mateusz Kowalczyk at 2015-06-12T03:01:50+01:00 Changelog for haskell/haddock#207 Fixes haskell/haddock#207, closes haskell/haddock#402 - - - - - 841d785e by jpmoresmau at 2015-06-12T16:03:16+01:00 Attach to instance location the name that has the same location file Fixes haskell/haddock#383 - - - - - 98791cae by Mateusz Kowalczyk at 2015-06-12T16:08:27+01:00 Update changelog Closes haskell/haddock#398 - - - - - 7c0b5a87 by Phil Ruffwind at 2015-06-12T13:07:25-04:00 Fix alignment of Source links in instance table in Firefox Due to a Firefox bug [1], a combination of 'whitespace: nowrap' on the parent element with 'float: right' on the inner element can cause the floated element to be displaced downwards for no apparent reason. To work around this, the left side is wrapped in its own <span> and set to 'float: left'. As a precautionary measure to prevent the parent element from collapsing entirely, we also add the classic "clearfix" hack. The latter is not strictly needed but it helps prevent bugs if the layout is altered again in the future. Fixes haskell/haddock#384. Remark: line 159 of src/Haddock/Backends/Xhtml/Layout.hs was indented to prevent confusion over the operator precedence of (<+>) vs (<<). [1]: https://bugzilla.mozilla.org/show_bug.cgi?id=488725 - - - - - cfe86e73 by Mateusz Kowalczyk at 2015-06-14T10:49:01+01:00 Update tests for the CSS changes - - - - - 2d4983c1 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding for Haskell source parser module. - - - - - 29548785 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement function for tagging parsed chunks with source spans. - - - - - 6a5e4074 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement simple string chunking based on HsColour library. - - - - - 6e52291f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create basic token classification method. - - - - - da971a27 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Adapt source span tagging to work with current whitespace handling. - - - - - 4feb5a22 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add record accessors to exports of hyperlinker parser module. - - - - - a8cc4e39 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Make parser module export all types and associated accessors. - - - - - fb8d468f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple HTML renderer for parsed source file. - - - - - 80747822 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for specifying the CSS file path in HTML source renderer. - - - - - 994dc1f5 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix identifier recognition in Haskell source parser. - - - - - b1bd0430 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix comment recognition in Haskell source parser. - - - - - 11db85ae by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for recognizing compiler pragmas in source parser. - - - - - 736c7bd3 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding of module for associating tokens with AST names. - - - - - 7e149bc2 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement utility method for extracting variable identifiers from AST. - - - - - 32eb640a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple mechanism for associating tokens with AST names. - - - - - d4eba5bc by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add dummy support for hyperlinking named tokens. - - - - - 2b76141f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matcher bug causing wrong items being hyperlinked. - - - - - 2d48002e by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Constrain elements exported by hyperlinker modules. - - - - - 9715eec6 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for type token recognition. - - - - - 8fa401cb by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for binding token recognition. - - - - - d062400b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement go-to-definition mechanism for local bindings. - - - - - f4dc229b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement module export- and import-list item hyperlinking. - - - - - c9a46d58 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matching to allow parenthesized operators hyperlinking. - - - - - 03aad95a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix weird hyperlinking of parenthesized operators. - - - - - b4694a7d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for type declaration anchors. - - - - - 7358d2d2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for top-level function declaration anchors. - - - - - dfc24b24 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix external anchors to contain HTML file extension. - - - - - a045926c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Refactor the way AST names are handled within detailed tokens. - - - - - c76049b4 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement hyperlinking of imported module names. - - - - - 2d2a1572 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix parsing of single line comments with broken up newlines. - - - - - 11afdcf2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix bug with improper newline handling. - - - - - 8137f104 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issues with escaped newlines in comments. - - - - - 34759b19 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for parsing C preprocessor macros. - - - - - 09f0f847 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for parser module of source hyperlinker. - - - - - 709a8389 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for AST module of source hyperlinker. - - - - - 4df5c227 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add command line option for generating hyperlinked source. - - - - - 7a755ea2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Extend module interface with rich source token stream field. - - - - - 494f4ab1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement source tokenization during interface creation process. - - - - - 5f21c953 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create hyperlinker module and plug it into the Haddock pipeline. - - - - - 0cc8a216 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for providing custom CSS files for hyperlinked source. - - - - - a32bbdc1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for fancy highlighting upon hovering over identifier. - - - - - d16d642a by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make source hyperlinker generate output in apropriate directory. - - - - - ae12953d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create module with hyperlinker utility functions. - - - - - 6d4952c5 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make external hyperlinks point to locations specified by source URLs. - - - - - 8417555d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Rewrite source generation to fixed links and directory structure. - - - - - ce9cec01 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add basic support for cross-package hyperlink generation. - - - - - 7eaf025c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Disable generating hyperlinks for module references. - - - - - a50bf92e by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make Haddock generate source for all interfaces (also hidden ones). - - - - - f5ae2838 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Prevent source parser from throwing exception when lexing fails. - - - - - db9ffbe0 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement workaround for Chrome highlighting issues. - - - - - 0b6b453b by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate correct anchors for data constructors. - - - - - c86d38bc by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate anchors for record field declarations. - - - - - 063abf7f by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issue with hyperlink highlight styling in Chrome browser. - - - - - 880fc611 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking constructor names in patters. - - - - - c9e89b95 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record patterns. - - - - - 17a11996 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record expressions. - - - - - 0eef932d by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Make hyperlinker respect pretty-printer flag and add documentation. - - - - - f87c1776 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Unexpose hyperlinker modules in Cabal configuration. - - - - - 4c9e2b06 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Setup HSpec framework for Haddock API package. - - - - - 4b20cb30 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add basic tests related to comment parsing. - - - - - 6842e919 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add tests related to parsing basic language constructs. - - - - - 87bffb35 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add simple tests for do-notation parsing. - - - - - e7af1841 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add very simple QuickCheck properties for source parser spec. - - - - - c84efcf1 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Create simple test runner for hyperlinker tests. - - - - - 76b90447 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for basic identifier hyperlinking. - - - - - 0fbf4df6 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for operator hyperlinking. - - - - - 731aa039 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for constructor hyperlinking. - - - - - 995a78a2 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for record expressions and patterns hyperlinking. - - - - - 3566875a by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for literal syntax highlighting. - - - - - 68469a35 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add hyperlinker test runner to .cabal and .gitignore files. - - - - - aa946c93 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Adapt hyperlinker test runner to have the same interface as HTML one. - - - - - ce34da16 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Fix hyperlinker test runner file paths and add pretty-printing option. - - - - - 0d7dd65e by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add reference files for hyperlinker test cases. - - - - - efa4a1e0 by Łukasz Hanuszczak at 2015-07-01T00:47:32+02:00 Make hyperlinker test runner strip local links from generated source. - - - - - 3e96e584 by Łukasz Hanuszczak at 2015-07-01T01:14:59+02:00 Create simple script for accepting hyperlinker test case references. - - - - - 526fe610 by Łukasz Hanuszczak at 2015-07-01T01:16:41+02:00 Re-accept hyperlinker test cases with local references stripped out. - - - - - 892e2cb3 by Łukasz Hanuszczak at 2015-07-01T01:22:09+02:00 Fix bug with diffing wrong files in hyperlinker test runner. - - - - - 9ff46039 by Łukasz Hanuszczak at 2015-07-01T18:04:46+02:00 Remove unused dependencies in Haddock API spec configuration. - - - - - 47969c07 by Łukasz Hanuszczak at 2015-07-01T18:32:19+02:00 Add support for hyperlinking synonyms in patterns. - - - - - a73449e0 by Łukasz Hanuszczak at 2015-07-01T18:33:44+02:00 Create test case for hyperlinking @-patterns. - - - - - c2077ed8 by Łukasz Hanuszczak at 2015-07-01T19:06:04+02:00 Add support for hyperlinking universally quantified type variables. - - - - - 68017342 by Łukasz Hanuszczak at 2015-07-01T19:28:32+02:00 Create hyperlinker test case with quantified type variables. - - - - - 51c01a78 by Łukasz Hanuszczak at 2015-07-01T19:34:22+02:00 Add scoped type variables test for polymorphism test case. - - - - - 13181ae2 by Łukasz Hanuszczak at 2015-07-01T19:56:27+02:00 Add record wildcards test for records hyperlinking test case. - - - - - 991b81dd by Łukasz Hanuszczak at 2015-07-01T21:01:42+02:00 Document some functions in XHTML utlity module. - - - - - 98c8dfe5 by Łukasz Hanuszczak at 2015-07-01T22:25:21+02:00 Make hyperlinker render qualified names as one entity. - - - - - 75e13b9b by Łukasz Hanuszczak at 2015-07-01T22:27:38+02:00 Add qualified name test for identifiers hyperlinking test case. - - - - - de1e143f by Łukasz Hanuszczak at 2015-07-02T12:32:59+02:00 Fix crash happening when hyperlinking type family declarations. - - - - - 7a8fb175 by Łukasz Hanuszczak at 2015-07-02T12:47:03+02:00 Add support for anchoring data family constructor declarations. - - - - - 3b404e49 by Łukasz Hanuszczak at 2015-07-02T13:31:05+02:00 Improve support for hyperlinking type families. - - - - - 59eb7143 by Łukasz Hanuszczak at 2015-07-02T13:33:34+02:00 Add hyperlinker test case for checking type and type family declarations. - - - - - d1cda0c0 by Łukasz Hanuszczak at 2015-07-02T13:41:38+02:00 Fix issue with operators being recognized as preprocessor directives. - - - - - da206c9d by Łukasz Hanuszczak at 2015-07-02T17:18:12+02:00 Fix broken tests for parsing and hyperlinking hash operators. - - - - - 53750d1b by Łukasz Hanuszczak at 2015-07-02T18:53:28+02:00 Add support for anchoring signatures in type class declarations. - - - - - 1fa5bb10 by Łukasz Hanuszczak at 2015-07-02T19:04:47+02:00 Make hyperlinker generate anchors only to top-level value bindings. - - - - - a542305c by Łukasz Hanuszczak at 2015-07-02T19:05:58+02:00 Create hyperlinker test case for type classes. - - - - - b0dd4581 by Łukasz Hanuszczak at 2015-07-04T16:28:26+02:00 Update docs with information about source hyperlinking. - - - - - 9795302a by Łukasz Hanuszczak at 2015-07-04T16:52:15+02:00 Update docs on using `--read-interface` option. - - - - - 9acdc002 by Łukasz Hanuszczak at 2015-07-04T17:15:26+02:00 Remove potentially dangerous record access in hyperlinker AST module. - - - - - fb3ab7be by Łukasz Hanuszczak at 2015-07-04T17:40:10+02:00 Make Haddock generate warnings about potential misuse of hyperlinker. - - - - - a324c504 by Łukasz Hanuszczak at 2015-07-04T17:43:22+02:00 Fix incorrect specification of source style option in doc file. - - - - - 3f01a8e4 by Łukasz Hanuszczak at 2015-07-05T17:06:36+02:00 Refactor source path mapping to use modules as indices. - - - - - ac70f5b1 by Łukasz Hanuszczak at 2015-07-05T17:47:34+02:00 Fix bug where not all module interfaces were added to source mapping. - - - - - f5e57da9 by Łukasz Hanuszczak at 2015-07-06T16:39:57+02:00 Extract main hyperlinker types to separate module. - - - - - 43974905 by Łukasz Hanuszczak at 2015-07-06T16:52:13+02:00 Move source paths types to hyperlinker types module. - - - - - 3e236055 by Łukasz Hanuszczak at 2015-07-06T17:06:19+02:00 Add support for hyperlinking modules in import lists. - - - - - 58233d9f by Łukasz Hanuszczak at 2015-07-06T17:26:49+02:00 Add short documentation for hyperlinker source map type. - - - - - 14da016d by Łukasz Hanuszczak at 2015-07-06T18:07:20+02:00 Fix bug with module name being hyperlinked to `Prelude`. - - - - - 8f79db52 by Łukasz Hanuszczak at 2015-07-06T18:23:47+02:00 Fix problem with spec build in Haddock API configuration. - - - - - e7cc056c by Adam Sandberg Eriksson at 2015-07-07T23:22:21+01:00 StrictData: print correct strictness marks - - - - - e8253ca8 by Mateusz Kowalczyk at 2015-07-07T23:58:28+01:00 Update changelog - - - - - 0aba676b by Mateusz Kowalczyk at 2015-07-07T23:58:33+01:00 Relax upper bound on GHC a bit - - - - - 7a595381 by Mateusz Kowalczyk at 2015-07-07T23:58:52+01:00 Delete trailing whitespace - - - - - 50976d5e by Adam Sandberg Eriksson at 2015-07-08T15:03:04+02:00 StrictData: changes in HsBang type - - - - - 83b045fa by Mateusz Kowalczyk at 2015-07-11T14:35:18+01:00 Fix expansion icon for user-collapsible sections Closes haskell/haddock#412 - - - - - b2a3b0d1 by Mateusz Kowalczyk at 2015-07-22T22:03:21+01:00 Make some version changes after 2.16.1 release - - - - - a8294423 by Ben Gamari at 2015-07-27T13:16:07+02:00 Merge pull request haskell/haddock#422 from adamse/adamse-D1033 Merge for GHC D1033 - - - - - c0173f17 by randen at 2015-07-30T14:49:08-07:00 Break the response file by line termination rather than spaces, since spaces may be within the parameters. This simple approach avoids having the need for any quoting and/or escaping (although a newline char will not be possible in a parameter and has no escape mechanism to allow it). - - - - - 47c0ca14 by Alan Zimmerman at 2015-07-31T10:41:52+02:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - 45a9d770 by Mateusz Kowalczyk at 2015-07-31T09:47:43+01:00 Update changelog - - - - - 347a20a3 by Phil Ruffwind at 2015-08-02T23:15:26+01:00 Avoid JavaScript error during page load in non-frame mode In non-frame mode, parent.window.synopsis refers to the synopsis div rather than the nonexistent frame. Unfortunately, the script wrongly assumes that if it exists it must be a frame, leading to an error where it tries to access the nonexistent attribute 'replace' of an undefined value (synopsis.location). Closes haskell/haddock#406 - - - - - 54ebd519 by Phil Ruffwind at 2015-08-02T23:27:10+01:00 Link to the definitions to themselves Currently, the definitions already have an anchor tag that allows URLs with fragment identifiers to locate them, but it is rather inconvenient to obtain such a URL (so-called "permalink") as it would require finding the a link to the corresponding item in the Synopsis or elsewhere. This commit adds hyperlinks to the definitions themselves, allowing users to obtain links to them easily. To preserve the original aesthetics of the definitions, we alter the color of the link so as to be identical to what it was, except it now has a hover effect indicating that it is clickable. Additionally, the anchor now uses the 'id' attribute instead of the (obsolete) 'name' attribute. Closes haskell/haddock#407 - - - - - 02cc8bb7 by Phil Ruffwind at 2015-08-02T23:28:02+01:00 Fix typo in Haddock.Backends.Xhtml.Layout: divSynposis -> divSynopsis Closes haskell/haddock#408 - - - - - 2eb0a458 by Phil Ruffwind at 2015-08-02T23:30:07+01:00 Fix record field alignment when name is too long Change <dl> to <ul> and use display:table rather than floats to layout the record fields. This avoids bug haskell/haddock#301 that occurs whenever the field name gets too long. Slight aesthetic change: the entire cell of the field's source code is now shaded gray rather than just the area where text exists. Fixes haskell/haddock#301. Closes haskell/haddock#421 - - - - - 7abb3402 by Łukasz Hanuszczak at 2015-08-02T23:32:14+01:00 Add some utility definitions for generating line anchors. - - - - - e0b1d79b by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Make hyperlinked source renderer generate line anchors. - - - - - 24dd4c9f by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Re-accept test cases after adding line anchors for each of them. - - - - - 0372cfcb by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Override source line flags when source hyperlinker is enabled. - - - - - a81bcd07 by Mateusz Kowalczyk at 2015-08-02T23:58:25+01:00 Update tests to follow HTML changes - - - - - d2d7426f by Łukasz Hanuszczak at 2015-08-06T20:54:59+02:00 Fix quote syntax for promoted types. - - - - - 668cf029 by Łukasz Hanuszczak at 2015-08-06T21:12:00+02:00 Apply promoted type quoting to type-level consing. - - - - - 89f8e7c6 by Łukasz Hanuszczak at 2015-08-06T21:17:10+02:00 Extend advanced types test case with other examples. - - - - - 86494bca by Łukasz Hanuszczak at 2015-08-06T21:22:06+02:00 Rename advanced types test case and accept new output. - - - - - dbb7c7c0 by Adam Sandberg Eriksson at 2015-08-09T23:01:05+02:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 2a7704fa by Ben Gamari at 2015-08-10T13:18:05+02:00 Merge pull request haskell/haddock#433 from adamse/split-hsbang HsBang is split into HsSrcBang and HsImplBang - - - - - 891954bc by Thomas Miedema at 2015-08-15T14:51:18+02:00 Follow changes in GHC build system - - - - - b55d32ab by Mateusz Kowalczyk at 2015-08-21T18:06:09+01:00 Make Travis use 7.10.2 - - - - - 97348b51 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Move SYB utilities to standalone module. - - - - - 748ec081 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement `everywhere` transformation in SYB module. - - - - - 011cc543 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement generic transformation constructor. - - - - - b9510db2 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Create simple utility module for type specialization. - - - - - 43229fa6 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Make type of type specialization function more general. - - - - - fd844e90 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Add basic HTML test case for checking instance specialization. - - - - - 6ea0ad04 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Make HTML class instance printer take optional signature argument. - - - - - 65aa41b6 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Refactor instance head type to record instead of a meaningless tuple. - - - - - 3fc3bede by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add expandable method section for each class instance declaration. - - - - - 99ceb107 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Move dummy post-family instances for `DocName` to `Types` module. - - - - - e98f4708 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create convenience functions for type specialization module. - - - - - b947552f by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Hook type specialization logic with HTML pretty-printer. - - - - - dcaa8030 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create stub functions for sugaring specialized types. - - - - - fa84bc65 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement list syntax sugaring logic for specialized types. - - - - - e8b05b07 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement tuple syntax sugaring logic for specialized types. - - - - - 68a2e5bc by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Get rid of code duplication in type specialization module. - - - - - 4721c336 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create scaffolding of a framework for renaming specialized types. - - - - - 271b488d by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in missing cases in specialized type renaming function. - - - - - bfa5f2a4 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Remove code duplication in specialized type renamer. - - - - - ea6bd0e8 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Change state of the type renaming monad. - - - - - 77c5496e by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement simple mechanism for generating new type names. - - - - - 91bfb48b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in stub behaviour with actual environment renaming. - - - - - d244517b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix logic behind binder type renaming. - - - - - f3c5e360 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add SYB-like utility function for performing stateful queries. - - - - - eb3f9154 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create function for retrieving free variables from given type. - - - - - a94561d3 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix compilation error caused by incorrect type signature. - - - - - 8bb707cf by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Move `SetName` class definition to types module. - - - - - 5800b13b by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Hook type renamer with instance method HTML pretty-printer. - - - - - 6a480164 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some test cases for type renamer. - - - - - 839842f7 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized signatures refer to original signature declaration. - - - - - 4880f7c9 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized methods be nicely formatted again. - - - - - ab5a6a2e by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Attach source locations to the specialized class methods. - - - - - 43f8a559 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Extend instances test case to also test multi-name type signatures. - - - - - 59bc751c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix tab-based indentation in instances test case. - - - - - c2126815 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Improve placement of instance methods expander button. - - - - - 0a32e287 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add new data type declaration to instance specialization test case. - - - - - 5281af1f by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make type renamer first try single-letter names as alternatives. - - - - - 7d509475 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix type renamer bug with incorrect names being generated. - - - - - 0f35bf7c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some documentation and refactor type specialization module. - - - - - da1d0803 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix another bug where type renamer was generating incorrect names. - - - - - cd39b5cb by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Refactor type renamer to rebinding and pure renaming phases. - - - - - 850251f4 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix unwitting compilation bug. - - - - - e5e9fc01 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Integrate instance specification type into class instance definition. - - - - - 825b0ea0 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer neccessary instance specification type. - - - - - cdba44eb by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix declaration converter to use more appropriate mode for methods. - - - - - bc45c309 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug with types not being specialized at all. - - - - - 5d8e5d89 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug where instance expander was opening wrong section. - - - - - 6001ee41 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix another type renamer bug where not all names were rebound. - - - - - 5f58ce2a by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix yet another renamer bug where some names were not unique. - - - - - 8265e521 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split instance subsection layout method to top-level declarations. - - - - - e5e66298 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Rearrange layout of instance methods in generated documentation. - - - - - a50b4eea by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer used layout method. - - - - - 2ff36ec2 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach section title to the instance methods block. - - - - - 7ac15300 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Add basic tests for associated types in instances test case. - - - - - db0ea2f9 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach associated types information to instance header. - - - - - 71cad4d5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details section contain associated types information. - - - - - deee2809 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Improve look of rendered associated families in instance details. - - - - - 839d13a5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Introduce alternative type for family declarations. - - - - - d397f03f by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details record use new type for family declarations. - - - - - 2b23fe97 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split printer of type family header to separate functions. - - - - - c3498cdc by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Implement HTML renderer for pseudo-family declarations. - - - - - c12bbb04 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Apply type specializer to associated type family declarations. - - - - - 2fd69ff2 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Create helper method for specializing type signatures. - - - - - 475826e7 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Refactor specializer module to be independent from XHTML backend. - - - - - f00b431c by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add some documentation for instance head specializer. - - - - - a9fef2dc by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix bug with missing space in documentation for associated types. - - - - - 50e29056 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with incorrect instance details sections being expanded. - - - - - e6dfdd03 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by adding instance details section. - - - - - 75565b2a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make section identifier of instance details more GHC-independent. - - - - - add0c23e by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Re-accept tests after applying deterministic section identifiers. - - - - - 878f2534 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make identifier generation also architecture-independent. - - - - - 48be69f8 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with instance expander hijacking type hyperlink click. - - - - - 47830c1f by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Get rid of dreadful hashing function for generating identifiers. - - - - - 956cd5af by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Move `InstOrigin` type declaration to more appropriate module. - - - - - bf672ed3 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by changes related to instance expander. - - - - - 8f2a949a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add examples with type operators to the instances test case. - - - - - 64600a84 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add basic support for sugaring infix type operators. - - - - - 747d71b8 by Łukasz Hanuszczak at 2015-08-21T18:22:34+01:00 Add support for sugaring built-in function syntax. - - - - - d4696ffb by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Remove default methods from Hoogle class output. - - - - - bf0e09d7 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add fixity declarations in Hoogle backend output. - - - - - 90e91a51 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix bug with incorrect fixities being generated in Hoogle backend. - - - - - 48f11d35 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve class type family declarations output in Hoogle backend. - - - - - 661e8e8f by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add missing default family equations in Hoogle output. - - - - - e2d64103 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve formatting of class details output in Hoogle backend. - - - - - 490fc377 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix weird-looking Hoogle output for familyless classes. - - - - - ea115b64 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create script file for new HTML test runner. - - - - - 609913d3 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Set default behaviour if no arguments given. - - - - - dc115f67 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for providing optional arguments for test runner. - - - - - d93ec867 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve output of test runner error messages. - - - - - 0be9fe12 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for executing Haddock process in test runner. - - - - - 4e4d00d9 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add GHC path to test runner configuration. - - - - - d67a2086 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make GHC path a test runner command-line argument. - - - - - c810079a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Extend test runner configuration with Haddock arguments. - - - - - fee18845 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor test runner and create stub functions. - - - - - ff7c161f by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make test runner actually run Haddock executable. - - - - - 391f73e6 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with test runner not producing any output files. - - - - - 81a74e2d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Setup skeleton of framework for running tests. - - - - - f8a79ec4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with modules not being found in global search mode. - - - - - 7e700b4d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make Haddock standard output redirection be more configurable. - - - - - 53b4c17a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Incorporate old, ugly functions for comparing output files. - - - - - 8277c8aa by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor architecture of test runner output checking functions. - - - - - 587bb414 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement actual diffing mechanism. - - - - - 9ed2b5e4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve code style to match popular guidelines. - - - - - 14bffaf8 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make it possible to choose alternative diff tool. - - - - - 5cdfb005 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create stub methods for processing test output as XML documents. - - - - - 7ef8e12e by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement link-stripping logic as simple SYB transformation. - - - - - 8a1fcd4f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Incorporate link stripping to output diffing mechanism. - - - - - 37dba2bc by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement footer-stripping logic. - - - - - 9cd52120 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Add missing dependencies in Cabal configuration file. - - - - - e0f83c6e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix issue with output being printed in incorrect order. - - - - - 0a94fbb0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to run tests without generating diff. - - - - - 76a58c6f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor HTML test suite boilerplate to external package. - - - - - af41e6b0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utilities for storing directory configuration. - - - - - d8f0698f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move IO-dependent config of HTML test suite to test package. - - - - - 17369fa0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Enable all compiler warnings in Haddock test package configuration. - - - - - 9d03b47a by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move Haddock runner of HTML test suite to Haddock test package. - - - - - 4b3483c5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 03754194 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience wrappers to simplify in test entry points. - - - - - 27476ab7 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adjust module visibility and items they export. - - - - - c40002ba by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Remove no longer useful test option. - - - - - 55ab2541 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Change extension of test files used for diffing. - - - - - 136bf4e4 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor and simplify XHTML helper module of test package. - - - - - 69f7e3df by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix typo in link stripper of HTML test suite runner. - - - - - 0c3c1c6b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience script for running specific HTML tests. - - - - - 489e1b05 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement utility functions for conditional link stripping. - - - - - 0f985dc3 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt `hypsrc-test` module to work with new testing framework. - - - - - 927406f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement output accepting mechanism in test package. - - - - - 8545715e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utility function for recursive obtaining directory contents. - - - - - cb70381f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 019599b5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix path handling in test runner. - - - - - 399b985b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to specify ignored files for test output. - - - - - 41b3d93d by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt HTML test runner to use new ignoring functionality. - - - - - e2091c8b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix bug with not all test output files being checked. - - - - - b22134f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Specify ignored files for hyperlinker source test runner. - - - - - 3301dfa1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Copy test runner script for hyperlinked source case. - - - - - d39a6dfa by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner invoking Haddock in incorrect mode. - - - - - f32c8ff3 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix path handling in test module loader. - - - - - 10f94ee9 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make test runner ignore test packages with no modules. - - - - - 5dc4239c by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create test runner entry points for LaTeX test suite. - - - - - 58d1f7cf by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with unnecessary checking old test output. - - - - - c7ce76e1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Re-implement test acceptance functionality. - - - - - 13bbabe8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix warning about no longer needed definition. - - - - - 958a99b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Adapt Cabal configuration to execute LaTeX suite with new runner. - - - - - 550ff663 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Setup test suite for Hoogle backend. - - - - - 3aa969c4 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make Hoogle backend create output directory if needed. - - - - - eb085b02 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Add appropriate .gitignore entry and configure Hoogle test suite. - - - - - a50bf915 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner failing when run on multiple test packages. - - - - - bf5368b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create simple test cases for Hoogle backend. - - - - - 6121ba4b by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create helper function for conversion between XML and XHTML. - - - - - cb516061 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Refactor existing code to use XHTML printer instead of XML one. - - - - - e2de8c82 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Improve portability of test runner scripts. - - - - - 9563e774 by Łukasz Hanuszczak at 2015-08-22T23:43:16+02:00 Remove redundant import statement. - - - - - 55353df1 by Łukasz Hanuszczak at 2015-08-24T23:09:20+02:00 Fix bug with accepting to non-existing directory. - - - - - 00a334ca by Łukasz Hanuszczak at 2015-08-24T23:09:47+02:00 Accept output for Hoogle and LaTeX backends. - - - - - 29191d8b by Łukasz Hanuszczak at 2015-08-24T23:14:18+02:00 Get rid of obsolete testing utilities. - - - - - bbb25db3 by Łukasz Hanuszczak at 2015-08-24T23:18:50+02:00 Update sandbox setup guide to work with Haddock test package. - - - - - cfd45248 by Łukasz Hanuszczak at 2015-08-24T23:51:30+02:00 Make Travis aware of Haddock test package. - - - - - 74185b7a by Łukasz Hanuszczak at 2015-08-25T17:41:59+02:00 Fix test suite failure when used with Stack. - - - - - 18769697 by Łukasz Hanuszczak at 2015-08-25T18:02:09+02:00 Add sample Stack setup to the hacking guide. - - - - - 22715eeb by Łukasz Hanuszczak at 2015-08-25T18:04:47+02:00 Fix Markdown formatting of README file. - - - - - b49ec386 by Łukasz Hanuszczak at 2015-08-25T18:13:36+02:00 Setup Haddock executable path in Travis configuration. - - - - - 5d29eb03 by Eric Seidel at 2015-08-30T09:55:58-07:00 account for changes to ipClass - - - - - f111740a by Ben Gamari at 2015-09-02T13:20:37+02:00 Merge pull request haskell/haddock#443 from bgamari/ghc-head account for changes to ipClass - - - - - a2654bf6 by Jan Stolarek at 2015-09-03T01:32:57+02:00 Follow changes from haskell/haddock#6018 - - - - - 2678bafe by Richard Eisenberg at 2015-09-21T12:00:47-04:00 React to refactoring CoAxiom branch lists. - - - - - ebc56e24 by Edward Z. Yang at 2015-09-21T11:53:46-07:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4a8c4198 by Tamar Christina at 2015-09-27T13:59:08+02:00 Create Process: removed PhaseFailed - - - - - 7e99b790 by Oleg Grenrus at 2015-09-27T20:52:10+03:00 Generate docs for orphan instances - - - - - 32e932e2 by Oleg Grenrus at 2015-09-28T07:21:11+03:00 Have source links for orphan instances - - - - - c2eb9f4f by Oleg Grenrus at 2015-09-28T07:24:58+03:00 Print orphan instances header only if required - - - - - ff96f978 by Oleg Grenrus at 2015-09-28T07:40:54+03:00 Add orphan instances link to contents box - - - - - d72490a6 by Oleg Grenrus at 2015-09-28T16:37:44+03:00 Fix orphan instance collapsing - - - - - 25d3dfe5 by Ben Gamari at 2015-10-03T12:38:09+02:00 Merge pull request haskell/haddock#448 from Mistuke/fix-silent-death-of-runInteractive Remove PhaseFailed - - - - - 1e45e43b by Edward Z. Yang at 2015-10-11T13:10:10-07:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b1370ac1 by Adam Gundry at 2015-10-16T16:26:42+01:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 60bef421 by Simon Peyton Jones at 2015-10-26T12:52:36+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 4c1898ca by Simon Peyton Jones at 2015-10-27T14:24:56+00:00 Track change to PatSyn.patSynSig - - - - - 25108e85 by Simon Peyton Jones at 2015-10-27T17:34:18+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - 693643ac by Ben Gamari at 2015-10-28T14:33:06+01:00 Account for Typeable changes The treatment of type families changed. - - - - - cd7c2221 by Simon Peyton Jones at 2015-10-30T13:03:51+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - 712032cb by Herbert Valerio Riedel at 2015-10-31T11:01:45+01:00 Relax upper bound on `base` to allow base-4.9 - - - - - 0bfa0475 by Simon Peyton Jones at 2015-10-31T19:08:13+00:00 More adaption to wildcard-refactor - - - - - 0a3c0cb7 by Simon Peyton Jones at 2015-10-31T22:14:43+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - c4fd4ec9 by Alan Zimmerman at 2015-11-01T11:16:34+01:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 42cdd882 by Matthew Pickering at 2015-11-06T20:02:16+00:00 Change for IEThingWith - - - - - f368b7be by Ben Gamari at 2015-11-11T11:35:51+01:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - e32965b8 by Simon Peyton Jones at 2015-11-13T12:18:17+00:00 Merge with origin/head - - - - - ebcf795a by Edward Z. Yang at 2015-11-13T21:56:27-08:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4e23989f by Simon Peyton Jones at 2015-11-18T11:32:54+00:00 Wibbles to Haddock - - - - - 2289cd4a by Simon Peyton Jones at 2015-11-20T23:12:49+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor - - - - - 695975a6 by Alan Zimmerman at 2015-11-21T21:16:12+02:00 Update to match GHC wip/T11019 - - - - - bbba21e7 by Simon Peyton Jones at 2015-11-23T13:54:31+00:00 merge with origin/ghc-head - - - - - 3d664258 by Simon Peyton Jones at 2015-11-23T17:17:18+00:00 Wibble - - - - - e64cf586 by Herbert Valerio Riedel at 2015-12-05T00:29:55+01:00 Canonicalise Monad instances - - - - - a2de15a7 by Alan Zimmerman at 2015-12-05T17:33:52+02:00 Matching changes for haskell/haddock#11028 - - - - - cc29a3e4 by Alan Zimmerman at 2015-12-05T19:45:33+02:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 95dd15d1 by Richard Eisenberg at 2015-12-11T17:33:39-06:00 Update for type=kinds - - - - - cb5fd9ed by Herbert Valerio Riedel at 2015-12-14T15:07:30+00:00 Bump versions for ghc-7.11 - - - - - 4f286d96 by Simon Peyton Jones at 2015-12-14T15:10:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 13ea2733 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 098df8b8 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track changes in HsSyn for quasi-quotes - - - - - 716a64de by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track change in API of TyCon - - - - - 77a66bca by Adam Gundry at 2015-12-14T15:10:58+00:00 Track API changes to support empty closed type familes - - - - - f2808305 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track the new location of setRdrNameSpace - - - - - ba8b08a4 by Alan Zimmerman at 2015-12-14T15:10:59+00:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - a4ded87e by Thomas Winant at 2015-12-14T15:14:05+00:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 25c78107 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: print correct strictness marks - - - - - 6cbc41c4 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: changes in HsBang type - - - - - ad46821a by Alan Zimmerman at 2015-12-14T15:14:06+00:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - abc0ae5b by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 3308d06c by Thomas Miedema at 2015-12-14T15:14:07+00:00 Follow changes in GHC build system - - - - - 6c763deb by Eric Seidel at 2015-12-14T15:14:07+00:00 account for changes to ipClass - - - - - ae5b4eac by Jan Stolarek at 2015-12-14T15:17:00+00:00 Follow changes from haskell/haddock#6018 - - - - - ffbc40e0 by Richard Eisenberg at 2015-12-14T15:17:02+00:00 React to refactoring CoAxiom branch lists. - - - - - d1f531e9 by Edward Z. Yang at 2015-12-14T15:17:02+00:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 79f73754 by Tamar Christina at 2015-12-14T15:17:02+00:00 Create Process: removed PhaseFailed - - - - - 3d37bebb by Edward Z. Yang at 2015-12-14T15:20:46+00:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 5f8a9e44 by Adam Gundry at 2015-12-14T15:20:48+00:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 79dda70f by Simon Peyton Jones at 2015-12-14T15:26:02+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 959930fb by Simon Peyton Jones at 2015-12-14T15:37:50+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - e18a8df5 by Simon Peyton Jones at 2015-12-14T15:37:52+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - aa35ab52 by Simon Peyton Jones at 2015-12-14T15:40:18+00:00 More adaption to wildcard-refactor - - - - - 8ceef94b by Simon Peyton Jones at 2015-12-14T15:46:04+00:00 Track change to PatSyn.patSynSig - - - - - cd81e83d by Ben Gamari at 2015-12-14T15:46:06+00:00 Account for Typeable changes The treatment of type families changed. - - - - - 63c9117c by Herbert Valerio Riedel at 2015-12-14T15:46:34+00:00 Relax upper bound on `base` to allow base-4.9 - - - - - a484c613 by Alan Zimmerman at 2015-12-14T15:47:46+00:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 2c26fa51 by Matthew Pickering at 2015-12-14T15:47:47+00:00 Change for IEThingWith - - - - - 593baa0f by Ben Gamari at 2015-12-14T15:49:21+00:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - b6b5ca78 by Edward Z. Yang at 2015-12-14T15:49:54+00:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b5b0e072 by Alan Zimmerman at 2015-12-14T15:54:20+00:00 Update to match GHC wip/T11019 - - - - - 14ddeb68 by Simon Peyton Jones at 2015-12-14T15:54:22+00:00 Wibble - - - - - 10a90ad8 by Herbert Valerio Riedel at 2015-12-14T15:54:22+00:00 Canonicalise Monad instances - - - - - ed68ac50 by Alan Zimmerman at 2015-12-14T15:55:48+00:00 Matching changes for haskell/haddock#11028 - - - - - 3f7e5a2d by Alan Zimmerman at 2015-12-14T15:55:49+00:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 6543a73f by Richard Eisenberg at 2015-12-14T15:59:55+00:00 Update for type=kinds - - - - - 193a5c48 by Matthew Pickering at 2015-12-14T18:17:00+00:00 Changes to compile with 8.0 - - - - - add669ec by Matthew Pickering at 2015-12-14T18:47:12+00:00 Warnings - - - - - 223f3fb4 by Ben Gamari at 2015-12-15T23:45:05+01:00 Update for D1200 - - - - - d058388f by Ben Gamari at 2015-12-16T05:40:17-05:00 Types: Add Outputable[Bndr] DocName instances - - - - - 62ecd7fb by Ben Gamari at 2015-12-16T09:23:09-05:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ddbc187a by Ben Gamari at 2015-12-16T17:54:55+01:00 Update for D1200 - - - - - cec83b52 by Ben Gamari at 2015-12-16T17:54:55+01:00 Types: Add Outputable[Bndr] DocName instances - - - - - d12ecc98 by Ben Gamari at 2015-12-16T17:54:55+01:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ada1616f by Ben Gamari at 2015-12-16T17:54:58+01:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a4f0383d by Ben Gamari at 2015-12-16T23:32:38+01:00 Fix Hyperlinker GHC.con_names is now GHC.getConNames - - - - - a10e6849 by Ben Gamari at 2015-12-20T00:54:11+01:00 Merge remote-tracking branch 'mrhania/testing-framework-improvements' into ghc-head - - - - - f078b4fd by Ben Gamari at 2015-12-20T00:59:51+01:00 test: Compatibility with Cabal 1.23 - - - - - 88a511a9 by Ben Gamari at 2015-12-20T01:14:35+01:00 Merge remote-tracking branch 'phadej/orphans' into ghc-head - - - - - 4e250f36 by Ben Gamari at 2015-12-20T01:14:52+01:00 Add html-test for orphan instances output - - - - - 87fffbad by Alan Zimmerman at 2015-12-20T09:50:42+02:00 Update for GHC trac#11258 Adding locations to RdrName in FieldOcc and AmbiguousFieldOcc - - - - - 6b7e51c9 by idontgetoutmuch at 2015-12-20T21:01:47+00:00 Merge pull request haskell/haddock#1 from haskell/ghc-head Ghc head - - - - - 229c1fb5 by Dominic Steinitz at 2015-12-21T07:19:16+00:00 Handle inline math with mathjax. - - - - - 57902d66 by Dominic Steinitz at 2015-12-21T08:07:11+00:00 Fix the documentation for haddock itself. Change notation and add support for inline math. Allow newlines in display math. Add a command line option for the mathjax url (you might want to use a locally installed version). Rebase tests because of extra url and version change. Respond to (some of the) comments. Fix warnings in InterfaceFile.hs - - - - - 0e69f236 by Herbert Valerio Riedel at 2015-12-21T18:30:43+01:00 Fix-up left-over assumptions of GHC 7.12 into GHC 8.0 - - - - - c67f8444 by Simon Peyton Jones at 2015-12-22T16:26:56+00:00 Follow removal of NamedWildCard from HsType - - - - - da40327a by Ben Gamari at 2015-12-23T14:15:28+01:00 html-test/Operators: Clear up ambiguous types For reasons that aren't entirely clear a class with ambiguous types was accepted by GHC <8.0. I've added a functional dependency to clear up this ambiguity. - - - - - 541b7fa4 by Ben Gamari at 2015-12-23T14:18:51+01:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 0febc947 by Ben Gamari at 2015-12-24T00:30:20+01:00 hoogle-test/AssocTypes: Allow AmbiguousTypes GHC 8.0 complains otherwise - - - - - 25810841 by Ben Gamari at 2015-12-24T00:33:18+01:00 OrphanInstances: Accept test output - - - - - 841987f3 by Ben Gamari at 2015-12-25T11:03:11+01:00 Merge remote-tracking branch 'idontgetoutmuch/ghc-head' into ghc-head - - - - - 358391f0 by Ben Gamari at 2015-12-26T10:44:50+01:00 Add missing import - - - - - a8896885 by Ben Gamari at 2015-12-26T10:45:27+01:00 travis: Use Travis containers - - - - - 85e82134 by Herbert Valerio Riedel at 2015-12-30T17:25:39+01:00 tweak version bounds for GHC-8.1 - - - - - 672a5f75 by randen at 2016-01-01T23:45:25-08:00 The Haddock part for fully gcc-like response files " driver/Main.hs * Moved the response file handling into ResponseFile.hs, updating import section as appropriate. * driver/ResponseFile.hs * New file. In anticipation that maybe some day this could be provided by another library, and to make it possible to unit test, this functionality is pulled out of the Main.hs module, and expanded to support the style/format of response files which gcc uses. * The specification for the format of response files which gcc generates and consumes, seems to be best derived from the gcc code itself (libiberty/argv.c), so that is what has been done here. * This is intended to fix haskell/haddock#379 * driver-test/Main.hs * New file for testing code in the driver source tree * driver-test/ResponseFileSpec.hs * Tests, adapted/adopted from the same gcc code where the escaping/unescaping is from, in the hspec style of unit tests * haddock.cabal * Add the driver-test test-suite. Introduces a new library dependency (upon hspec) for the haddock driver target in the haddock.cabal file, but practically, this should not be a problem as the haddock-api tests already depend on hspec. - - - - - 498781df by Ben Gamari at 2016-01-06T13:41:04+01:00 Version bumps and changelog - - - - - 8451e46a by Ben Gamari at 2016-01-06T13:47:17+01:00 Merge remote-tracking branch 'randen/bug468' - - - - - fb2d9181 by Ben Gamari at 2016-01-06T08:14:42-05:00 Add ResponseFile to OtherModules - - - - - 2cb2d2e3 by Ben Gamari at 2016-01-06T14:35:00+01:00 Merge branch 'master' into ghc-head - - - - - 913477d4 by Eric Seidel at 2016-01-11T14:57:57-08:00 deal with un-wiring of IP class - - - - - c557a4b3 by Alan Zimmerman at 2016-01-15T11:14:35+02:00 Update to match wip/T11430 in GHC - - - - - 3e135093 by Alan Zimmerman at 2016-01-16T18:21:59+01:00 Update to match wip/T11430 in GHC - - - - - c48ef2f9 by Ben Gamari at 2016-01-18T09:50:06+01:00 Merge remote-tracking branch 'gridaphobe/ghc-head' into ghc-head - - - - - 9138a1b0 by Eric Seidel at 2016-01-18T12:50:15+01:00 deal with un-wiring of IP class (cherry picked from commit 17388b0f0029d969d79353be7737eb01c7b8dc5f) - - - - - b48c172e by Joachim Breitner at 2016-01-19T00:11:38+01:00 Make sure --mathjax affects all written HTML files This fixes haskell/haddock#475. - - - - - af61fe63 by Ryan Scott at 2016-02-07T23:25:57+01:00 Render */# instead of TYPE 'Lifted/TYPE 'Unlifted (fixes haskell/haddock#473) - - - - - b6458693 by Ben Gamari at 2016-02-07T23:29:27+01:00 Merge pull request haskell/haddock#477 from haskell/issue-475 Make sure --mathjax affects all written HTML files - - - - - adcc0071 by Ben Gamari at 2016-02-07T23:34:52+01:00 Merge branch 'master' into ghc-head - - - - - d0404e61 by Ben Gamari at 2016-02-08T12:46:49+01:00 doc: Switch to Sphinx - - - - - acb153b3 by Ben Gamari at 2016-02-08T12:46:56+01:00 Document --use-unicode flag - - - - - c20bdf1d by Ben Gamari at 2016-02-08T13:41:24+01:00 Fix GHC and haddock-library dependency bounds - - - - - 8d946801 by Ben Gamari at 2016-02-08T14:54:56+01:00 testsuite: Rework handling of output sanitization Previously un-cleaned artifacts were kept as reference output, making it difficult to tell what has changed and causing spurious changes in the version control history. Here we rework this, cleaning the output during acceptance. To accomplish this it was necessary to move to strict I/O to ensure the reference handle was closed before accept attempts to open the reference file. - - - - - c465705d by Ben Gamari at 2016-02-08T15:36:05+01:00 test: Compare on dump For reasons I don't understand the Xml representations differ despite their textual representations being identical. - - - - - 1ec0227a by Ben Gamari at 2016-02-08T15:36:05+01:00 html-test: Accept test output - - - - - eefbd63a by Ben Gamari at 2016-02-08T15:36:08+01:00 hypsrc-test: Accept test output And fix impredicative Polymorphism testcase. - - - - - d1df4372 by Ben Gamari at 2016-02-08T15:40:44+01:00 Merge branch 'fix-up-testsuite' - - - - - 206a3859 by Phil Ruffwind at 2016-02-08T17:51:21+01:00 Move the permalinks to "#" on the right side Since pull request haskell/haddock#407, the identifiers have been permalinked to themselves, but this makes it difficult to copy the identifier by double-clicking. To work around this usability problem, the permalinks are now placed on the far right adjacent to "Source", indicated by "#". Also, 'namedAnchor' now uses 'id' instead of 'name' (which is obsolete). - - - - - 6c89fa03 by Phil Ruffwind at 2016-02-08T17:54:44+01:00 Update tests for previous commit - - - - - effaa832 by Ben Gamari at 2016-02-08T17:56:17+01:00 Merge branch 'anchors-redux' - - - - - 9a2bec90 by Ben Gamari at 2016-02-08T17:58:40+01:00 Use -fprint-unicode-syntax when --use-unicode is enabled This allows GHC to render `*` as its Unicode representation, among other things. - - - - - 28ecac5b by Ben Gamari at 2016-02-11T18:53:03+01:00 Merge pull request haskell/haddock#480 from bgamari/sphinx Move documentation to ReStructuredText - - - - - 222e5920 by Ryan Scott at 2016-02-11T15:42:42-05:00 Collapse type/data family instances by default - - - - - a80ac03b by Ryan Scott at 2016-02-11T20:17:09-05:00 Ensure expanded family instances render correctly - - - - - 7f985231 by Ben Gamari at 2016-02-12T10:04:22+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - d4eda086 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Various cleanups - - - - - 79bee48d by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show kind signatures for type family variables Addresses GHC haskell/haddock#11588. - - - - - b2981d98 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show 'where ...' after closed type family Seems like we should ideally show the actual equations as well but that seems like it would be a fair amount of work - - - - - cfc0e621 by Ben Gamari at 2016-02-18T22:48:12+01:00 Merge pull request haskell/haddock#483 from bgamari/T11588 Fix GHC haskell/haddock#11588 This fixes GHC haskell/haddock#11588: * Show where ... after closed type families * Show kind signatures on type family type variables - - - - - 256e8a0d by Ben Gamari at 2016-02-18T23:15:39+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 32402036 by Richard Eisenberg at 2016-02-24T13:21:44-05:00 Follow-on changes to support RuntimeRep - - - - - 2b1c572d by Matthew Pickering at 2016-03-04T21:04:02+00:00 Remove unused functions - - - - - eb906f50 by Richard Eisenberg at 2016-03-13T21:17:20+01:00 Follow-on changes to support RuntimeRep (cherry picked from commit ab954263a793d8ced734459d6194a5d89214b66c) - - - - - 8c34ef34 by Richard Eisenberg at 2016-03-14T23:47:23-04:00 Changes due to fix for GHC#11648. - - - - - 0e022014 by Richard Eisenberg at 2016-03-15T14:06:45+01:00 Changes due to fix for GHC#11648. (cherry picked from commit bb994de1ab0c76d1aaf1e39c54158db2526d31f1) - - - - - ed3f78ab by Rik Steenkamp at 2016-04-02T22:20:36+01:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 - - - - - d3210042 by Rik Steenkamp at 2016-04-04T15:43:32+02:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 (cherry picked from commit 3ddcbd6b8e6884bd95028381176eb33bee6896fb) - - - - - 236eec90 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 692ee7e0 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. - - - - - 79619f57 by Ben Gamari at 2016-04-10T23:46:22+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. (cherry picked from commit c6d6a18d85e5e2d9bb5904e6919e8a8d7e31c4c5) - - - - - 3358ccb4 by Ben Gamari at 2016-04-10T23:47:27+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 264949b1 by Ben Gamari at 2016-04-16T17:50:23+02:00 Merge pull request haskell/haddock#482 from RyanGlScott/ghc-head Collapse type/data family instances by default - - - - - 478c483a by Ben Gamari at 2016-04-16T17:51:09+02:00 Merge pull request haskell/haddock#489 from mpickering/unused-functions Remove some unused functions - - - - - c94e55f0 by Ryan Scott at 2016-04-16T17:57:54+02:00 Collapse type/data family instances by default (cherry picked from commit 2da130a8db8f995c119b544fad807533236cf088) - - - - - 31e633d3 by Ryan Scott at 2016-04-16T17:58:06+02:00 Ensure expanded family instances render correctly (cherry picked from commit 1338b5d7c32939de6bbc31af0049477e4f847103) - - - - - 03e4d197 by Matthew Pickering at 2016-04-16T17:58:21+02:00 Remove unused functions (cherry picked from commit b89d1c2456bdb2d4208d94ded56155f7088a37d0) - - - - - ed4116f6 by Ben Gamari at 2016-04-20T10:46:57+02:00 ghc: Install files for needed --hyperlinked-source - - - - - 0be999c4 by Ben Gamari at 2016-04-20T11:37:54+02:00 ghc: Install files for needed --hyperlinked-source (cherry picked from commit 5c82c9fc2d21ddaae4a2470f1c375426968f19c6) - - - - - 4d17544c by Simon Peyton Jones at 2016-04-20T12:42:28+01:00 Track change to HsGroup This relates to a big GHC patch for Trac haskell/haddock#11348 - - - - - 1700a50d by Ben Gamari at 2016-05-01T13:19:27+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. - - - - - 0b7c8125 by Ben Gamari at 2016-05-01T13:21:43+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. (cherry picked from commit 609018dd09c4ffe27f9248b2d8b50f6196cd42b9) - - - - - af115ce0 by Ryan Scott at 2016-05-04T22:15:50-04:00 Render Haddocks for derived instances Currently, one can document top-level instance declarations, but derived instances (both those in `deriving` clauses and standalone `deriving` instances) do not enjoy the same privilege. This makes the necessary changes to the Haddock API to enable rendering Haddock comments for derived instances. This is part of a fix for Trac haskell/haddock#11768. - - - - - 76fa1edc by Ben Gamari at 2016-05-10T18:13:25+02:00 haddock-test: A bit of refactoring for debuggability - - - - - 7d4c4b20 by Ben Gamari at 2016-05-10T18:13:25+02:00 Create: Mark a comment as TODO - - - - - 2a6d0c90 by Ben Gamari at 2016-05-10T18:13:25+02:00 html-test: Update reference output - - - - - bd60913d by Ben Gamari at 2016-05-10T18:13:25+02:00 hypsrc-test: Fix reference file path in cabal file It appears the haddock insists on prefixing --hyperlinked-sourcer output with directory which the source appeared in. - - - - - c1548057 by Ben Gamari at 2016-05-10T18:22:12+02:00 doc: Update extra-source-files in Cabal file - - - - - 41d5bae3 by Ben Gamari at 2016-05-10T18:29:21+02:00 Bump versions - - - - - ca75b779 by Ben Gamari at 2016-05-11T16:03:44+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 4e3cfd62 by Ben Gamari at 2016-05-11T16:06:45+02:00 Merge remote-tracking branch 'RyanGlScott/ghc-head' into ghc-head - - - - - a2379970 by Ben Gamari at 2016-05-11T23:15:11+02:00 doc: Add clean targets - - - - - f275212e by Ben Gamari at 2016-05-11T23:15:14+02:00 doc: Add html as an all-target for ghc Otherwise the html documentation won't be installed for binary-dist. - - - - - 388fc0af by Ben Gamari at 2016-05-12T09:49:12+02:00 Update CHANGES - - - - - bad81ad5 by Ben Gamari at 2016-05-12T09:49:38+02:00 Version bump - - - - - c01688a7 by Ben Gamari at 2016-05-12T10:04:58+02:00 Revert "Version bump" This bump was a bit premature. This reverts commit 7b238d9c5be9b07aa2d10df323b5c7b8d1634dc8. - - - - - 7ed05724 by Ben Gamari at 2016-05-12T10:05:33+02:00 doc: Fix GHC clean rule Apparently GHC's build system doesn't permit wildcards in clean paths. - - - - - 5d9611f4 by Ben Gamari at 2016-05-12T17:43:50+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 653566b2 by Ben Gamari at 2016-05-14T09:57:31+02:00 Version bump to 2.17.2 - - - - - b355c439 by Ben Gamari at 2016-05-14T09:57:51+02:00 doc: Use `$(MAKE)` instead of `make` This is necessary to ensure we use gmake. - - - - - 8a18537d by Ben Gamari at 2016-05-14T10:15:45+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - b3290ef1 by Sebastian Meric de Bellefon at 2016-05-14T11:29:47-04:00 Fix haskell/haddock#303. Hide footer when printing The "Produced by Haddock" footer was overlapping the page's body when printing. This patch hides the footer with a css media rule. - - - - - b4a76f89 by Sebastian Meric de Bellefon at 2016-05-15T02:12:46-04:00 Fix haskell/haddock#280. Parsing of module header The initial newlines were counted as indentation spaces, thus disturbing the parsing of next lines - - - - - ba797c9e by Ben Gamari at 2016-05-16T14:53:46+02:00 doc: Vendorize alabaster Sphinx theme Alabaster is now the default sphinx theme and is a significant improvement over the previous default that it's worthproviding it when unavailable (e.g. Sphinx <1.3). - - - - - c9283e44 by Ben Gamari at 2016-05-16T14:55:17+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 1c9ea198 by Sebastian Méric de Bellefon at 2016-05-16T12:30:40-04:00 Merge pull request haskell/haddock#502 from Helkafen/master Fix haskell/haddock#303. Hide footer when printing - - - - - 33631016 by Ben Gamari at 2016-05-16T19:56:11+02:00 Revert "doc: Vendorize alabaster Sphinx theme" This ended up causes far too many issues to be worthwhile. We'll just have to live with inconsistent haddock documentation. This reverts commit cec21957001143794e71bcd9420283df18e7de40. - - - - - 93317d26 by Ben Gamari at 2016-05-16T19:56:11+02:00 cabal: Fix README path - - - - - c8695b22 by Ben Gamari at 2016-05-16T19:58:51+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 0b50eaaa by Ben Gamari at 2016-05-16T21:02:08+02:00 doc: Use whichever theme sphinx deems appropriate - - - - - 857c1c9c by Ben Gamari at 2016-05-16T21:07:08+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 15fc5637 by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Remove redundant imports - - - - - 132ddc6a by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Better debug output For tracking down haskell/haddock#505 - - - - - 2252a149 by Ben Gamari at 2016-05-22T12:43:59+02:00 Don't consider default class ops when looking for decls When we are looking for an operation within a class we don't care about `default`-type declarations. This was the cause of haskell/haddock#505. - - - - - 4886b2ec by Oleg Grenrus at 2016-05-24T16:19:48+03:00 UnfelpfulSpan line number omitted Kind of resolves https://github.com/haskell/haddock/issues/508 - - - - - a4befd36 by Oleg Grenrus at 2016-05-24T16:53:35+03:00 Change Hyperlinked lexer to know about DataKinds ticks - - - - - f45cb52e by David Feuer at 2016-05-24T18:48:53-04:00 Make parser state a newtype Previously, it was `data` wrapping a `Maybe`, which seems a bit silly. Obviously, this can be changed back if anyone wants to add more fields some day. - - - - - 05013dd7 by Sebastian Meric de Bellefon at 2016-05-24T22:03:55-04:00 remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) Frames are a bit broken, ignored by Hackage, and considered obsolete in general. This patch disables frames generation. The mini_*.html files are still used in the synopsis. - - - - - b8163a88 by Ben Gamari at 2016-05-25T14:44:15+02:00 Merge pull request haskell/haddock#507 from bgamari/T505 Fix haskell/haddock#505 - - - - - ea1b30c6 by Sebastian Meric de Bellefon at 2016-05-25T14:17:00-04:00 Update CHANGES - - - - - eddfc258 by Sebastian Méric de Bellefon at 2016-05-25T15:17:40-04:00 Merge pull request haskell/haddock#514 from Helkafen/frames remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) - - - - - 0e506818 by Alex Biehl at 2016-05-26T12:43:09+02:00 Remove misplaced haddock comment - - - - - a07d28c0 by Ben Gamari at 2016-05-27T11:34:59+02:00 Merge pull request haskell/haddock#515 from alexbiehl/master Remove misplaced haddock comment - - - - - 9001d267 by Ben Gamari at 2016-05-27T11:35:46+02:00 Merge pull request haskell/haddock#513 from treeowl/newtype-since Make parser state a newtype - - - - - 74e1a018 by Sebastian Méric de Bellefon at 2016-05-28T17:28:15-04:00 Merge pull request haskell/haddock#504 from Helkafen/issue-280 Fix haskell/haddock#280. Parsing of module header - - - - - 37557f4f by Alan Zimmerman at 2016-05-29T23:36:50+02:00 Matching changes for haskell/haddock#12105 - - - - - 7d09e5d6 by Sebastian Meric de Bellefon at 2016-06-03T18:07:48-04:00 Version bumps (2.17.3, 1.4.2) - - - - - 85b4bc15 by Sebastian Méric de Bellefon at 2016-06-06T18:35:13-04:00 Merge pull request haskell/haddock#521 from Helkafen/master Version bumps (2.17.3, 1.4.2) - - - - - e95f0dee by Sebastian Meric de Bellefon at 2016-06-06T19:11:35-04:00 publish haddock-test library - - - - - 4de40586 by Sebastian Méric de Bellefon at 2016-06-06T20:26:30-04:00 Merge pull request haskell/haddock#512 from phadej/oleg-fixes Fixes for haskell/haddock#508 and haskell/haddock#510 - - - - - ddfd0789 by Dominic Steinitz at 2016-06-09T09:27:28+01:00 Documentation for LaTeX markup. - - - - - 697a503a by Dominic Steinitz at 2016-06-09T09:33:59+01:00 Fix spelling mistake. - - - - - 246f6fff by Dominic Steinitz at 2016-06-09T09:37:15+01:00 Camel case MathJax. - - - - - 4684bd23 by Dominic Steinitz at 2016-06-09T09:44:53+01:00 Fix math typo and add link. - - - - - f20c037c by Simon Peyton Jones at 2016-06-13T18:26:03+01:00 Follow changes to LHsSigWcType - - - - - 0c58996d by Simon Peyton Jones at 2016-06-15T12:56:01+01:00 Follow GHC re-adding FunTy - - - - - 401b5ca7 by Sebastian Méric de Bellefon at 2016-06-15T12:16:47-04:00 Merge pull request haskell/haddock#525 from idontgetoutmuch/master Documentation for LaTeX markup. - - - - - 92d263b7 by Sebastian Méric de Bellefon at 2016-06-15T12:17:29-04:00 Merge pull request haskell/haddock#522 from Helkafen/master publish haddock-test library - - - - - 0953a2ca by Sebastian Meric de Bellefon at 2016-06-16T00:46:46-04:00 Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 65453e14 by Ben Gamari at 2016-06-16T11:16:32+02:00 ocean: Ensure that synopsis fully covers other content Previously MathJax content was being rendered on top of the synopsis due to ambiguous z-ordering. Here we explicitly give the synopsis block a higher z-index to ensure it is rendered on top. Fixes haskell/haddock#531. - - - - - 68e411a1 by Sebastian Méric de Bellefon at 2016-06-16T23:34:39-04:00 Merge pull request haskell/haddock#534 from bgamari/T531 ocean: Ensure that synopsis fully covers other content - - - - - fad6491b by Sebastian Méric de Bellefon at 2016-06-18T23:57:20-04:00 Merge pull request haskell/haddock#533 from Helkafen/master Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 6108e21b by Sebastian Meric de Bellefon at 2016-06-22T23:08:28-04:00 do not create empty src directory Fix haskell/haddock#536. - - - - - 1ef23823 by Sebastian Méric de Bellefon at 2016-06-24T00:04:48-04:00 Merge pull request haskell/haddock#537 from Helkafen/master do not create empty src directory - - - - - 966baa96 by Omari Norman at 2016-06-29T21:59:34-04:00 Add $ as a special character If this character is not escaped, documentation built with Haddock 2.17.2 will fail. This was not an issue with 2.16 series, which causes builds to fail and there is nothing in the docs or error message giving a clue about why builds that used to succeed now don't. - - - - - 324adb60 by Ben Gamari at 2016-07-01T12:18:51+02:00 GhcUtils: Changes for multi-pattern signatures - - - - - d7571675 by Ömer Sinan Ağacan at 2016-07-21T13:30:47+02:00 Add support for unboxed sums - - - - - 29d0907b by Simon Marlow at 2016-07-22T13:55:48+01:00 Disable NFData instances for GHC types when GHC >= 8.2 - - - - - 702d95f3 by Simon Marlow at 2016-08-02T15:57:30+02:00 Disable NFData instances for GHC types when GHC >= 8.0.2 (cherry picked from commit a3309e797c42dae9bccdeb17ce52fcababbaff8a) - - - - - f4fa79c3 by Ben Gamari at 2016-08-07T13:51:18+02:00 ghc.mk: Don't attempt to install html/frames.html The frames business has been removed. - - - - - 9cd63daf by Ben Gamari at 2016-08-07T13:51:40+02:00 Haddock.Types: More precise version guard This allows haddock to be built with GHC 8.0.2 pre-releases. - - - - - f3d7e03f by Mateusz Kowalczyk at 2016-08-29T20:47:45+01:00 Merge pull request haskell/haddock#538 from massysett/master Add $ as a special character - - - - - 16dbf7fd by Bartosz Nitka at 2016-09-20T19:44:04+01:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 - - - - - 7c31c1ff by Bartosz Nitka at 2016-09-27T17:32:22-04:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 (cherry picked from commit 073d899a8f94ddec698f617a38d3420160a7fd0b) - - - - - 33a90dce by Ryan Scott at 2016-09-30T20:53:41-04:00 Haddock changes for T10598 See https://ghc.haskell.org/trac/ghc/ticket/10598 - - - - - 1f32f7cb by Ben Gamari at 2016-10-13T20:01:26-04:00 Update for refactoring of NameCache - - - - - 1678ff2e by Ben Gamari at 2016-11-15T17:42:48-05:00 Bump upper bound on base - - - - - 9262a7c5 by Alan Zimmerman at 2016-12-07T21:14:28+02:00 Match changes in GHC wip/T3384 branch - - - - - ac0eaf1a by Ben Gamari at 2016-12-09T09:48:41-05:00 haddock-api: Don't use stdcall calling convention on 64-bit Windows See GHC haskell/haddock#12890. - - - - - 04afe4f7 by Alan Zimmerman at 2016-12-12T20:07:21+02:00 Matching changes for GHC wip/T12942 - - - - - e1d1701d by Ben Gamari at 2016-12-13T16:50:41-05:00 Bump base upper bound - - - - - 3d3eacd1 by Alan Zimmerman at 2017-01-10T16:59:38+02:00 HsIParamTy now has a Located name - - - - - 7dbceefd by Kyrill Briantsev at 2017-01-12T13:23:50+03:00 Prevent GHC API from doing optimization passes. - - - - - d48d1e33 by Richard Eisenberg at 2017-01-19T08:41:41-05:00 Upstream changes re levity polymorphism - - - - - 40c25ed6 by Alan Zimmerman at 2017-01-26T15:16:18+02:00 Changes to match haskell/haddock#13163 in GHC - - - - - 504f586d by Ben Gamari at 2017-02-02T17:19:37-05:00 Kill remaining static flags - - - - - 49147ea0 by Justus Adam at 2017-03-02T15:33:34+01:00 Adding MDoc to exports of Documentation.Haddock - - - - - 1cfba9b4 by Justus Adam at 2017-03-09T11:41:44+01:00 Also exposing toInstalledIface - - - - - 53f0c0dd by Ben Gamari at 2017-03-09T13:10:08-05:00 Bump for GHC 8.3 - - - - - c7902d2e by Ben Gamari at 2017-03-09T23:46:02-05:00 Bump for GHC 8.2 - - - - - 4f3a74f8 by Ben Gamari at 2017-03-10T10:21:55-05:00 Merge branch 'ghc-head' - - - - - e273b72f by Richard Eisenberg at 2017-03-14T13:34:04-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - 6ec3d436 by Richard Eisenberg at 2017-03-14T15:15:52-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - eee3cda1 by Ben Gamari at 2017-03-15T15:19:59-04:00 Adapt to EnumSet - - - - - 017cf58e by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 736d6773 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 475f84a0 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 13240b53 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - cd16d529 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 3bea97ae by Edward Z. Yang at 2017-03-15T22:50:46-07:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b2b051ce by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 0f082795 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 20ef63c9 by Edward Z. Yang at 2017-03-22T13:48:12-07:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 45692dcb by Edward Z. Yang at 2017-03-22T14:11:25-07:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4eae8caf by Ben Gamari at 2017-03-23T09:25:33-04:00 Merge commit '240bc38b94ed2d0af27333b23392d03eeb615e82' into HEAD - - - - - 0bbe03f5 by Ben Gamari at 2017-03-23T09:27:28-04:00 haddock-api: Bump bound on GHC - - - - - 65f3ac9d by Alex Biehl at 2017-03-23T17:36:11+01:00 Merge pull request haskell/haddock#581 from JustusAdam/master Adding more exports to Documentation.Haddock - - - - - 37d49a47 by Alex Biehl at 2017-03-23T17:39:14+01:00 Merge pull request haskell/haddock#568 from awson/ghc-head Prevent GHC API from doing optimization passes. - - - - - 1ed047e4 by Brian Huffman at 2017-03-23T17:45:58+01:00 Print any user-supplied kind signatures on type parameters. This applies to type parameters on data, newtype, type, and class declarations, and also to forall-bound type vars in type signatures. - - - - - 1b78ca5c by Brian Huffman at 2017-03-23T17:45:58+01:00 Update test suite to expect kind annotations on type parameters. - - - - - a856b162 by Alex Biehl at 2017-03-23T17:49:32+01:00 Include travis build indication badge - - - - - 8e2e2c56 by Ben Gamari at 2017-03-23T17:20:08-04:00 haddock-api: Bump bound on GHC - - - - - 4d2d9995 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 26d6c150b31bc4580ab17cfd07b6e7f9afe10737) - - - - - a650e20f by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 930cfbe58e2e87f5a4d431d89a3c204934e6e858) - - - - - caa282c2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 2067a2d0afa9cef381d26fb7140b67c62f433fc0) - - - - - 49684884 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 0671abfe7e8ceae2269467a30b77ed9d9656e2cc) - - - - - 4dcfeb1a by Edward Z. Yang at 2017-03-23T17:20:08-04:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 3d77b373dd5807d5d956719dd7c849a11534fa6a) - - - - - 74dd19d2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 94610e9b446324f4231fa6ad4c6ac51e4eba8c0e) - - - - - a9b19a23 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 6cc832dfb1de6088a4abcaae62b25a7e944d55c3) - - - - - d3631064 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit bf3c4d72a0fda38561376eac7eda216158783267) - - - - - ef2148fc by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 07b88c5d4e79b87a319fbb08f8ea01dbb41063c1) - - - - - 2f29518b by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 4eb765ca4205c79539d60b7afa9b7e261a4a49fe) - - - - - 37de047d by Phil Ruffwind at 2017-04-03T11:57:14+02:00 Update MathJax URL MathJax is shutting down their CDN: https://www.mathjax.org/cdn-shutting-down/ They recommend migrating to cdnjs. - - - - - e9d24ba8 by David C. Turner at 2017-04-03T14:58:01+02:00 Add highlight for :target to ocean.css - - - - - 4819a202 by Alex Biehl at 2017-04-11T19:36:48+02:00 Allow base-4.10 for haddock-test - - - - - 44cec69c by Alex Biehl at 2017-04-11T19:39:22+02:00 cabal.project for haddock-api, haddock-library and haddock-test - - - - - 935d0f6a by Alex Biehl at 2017-04-11T19:46:29+02:00 Move dist scripts to scripts/ - - - - - 128e150c by Alex Biehl at 2017-04-11T20:34:46+02:00 Add haddock to cabal.project - - - - - cc8e08ea by Alex Biehl at 2017-04-11T20:35:08+02:00 Read files for hyperlinker eagerly This also exposes Documentation.Haddock.Utf8 - - - - - 152dda78 by Alex Biehl at 2017-04-11T20:37:06+02:00 Explicit import list ofr Control.DeepSeq in Haddock.Interface.Create - - - - - 501b33c4 by Kyrill Briantsev at 2017-04-11T21:01:42+02:00 Prevent GHC API from doing optimization passes. - - - - - c9f3f5ff by Alexander Biehl at 2017-04-12T16:36:53+02:00 Add @alexbiehl as maintaner - - - - - 76f214cc by Alex Biehl at 2017-04-13T07:27:18+02:00 Disable doctest with ghc-8.3 Currently doctest doesn't support ghc-head - - - - - 46b4f5fc by Edward Z. Yang at 2017-04-22T20:38:26-07:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f0555235 by Alex Biehl at 2017-04-25T10:08:48+02:00 Travis: Use ghc-8.2.1 on master - - - - - 966ea348 by Alex Biehl at 2017-04-25T10:32:01+02:00 Travis: Verbose cabal output cf. https://travis-ci.org/haskell/haddock/jobs/225512194#L377 - - - - - 36972bcd by Alex Biehl at 2017-04-25T10:40:43+02:00 Use travis_retry for cabal invocations - - - - - b3a09d2c by Alex Biehl at 2017-04-25T17:02:20+02:00 Use new MathJax URL in html-test 18ed871afb82560d5433b2f53e31b4db9353a74e switched to a new MathJax URL but didn't update the tests. - - - - - ae331e5f by Alexander Biehl at 2017-04-25T17:02:20+02:00 Expand signatures for class declarations - - - - - e573c65a by Alexander Biehl at 2017-04-25T17:02:20+02:00 Hoogle: Correctly print classes with associated data types - - - - - 3fc6be9b by Edward Z. Yang at 2017-04-25T17:02:20+02:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit a0c4790e15a2d3fab8d830eee8fcd639fe6d39c9) - - - - - 6725c060 by Herbert Valerio Riedel at 2017-04-25T17:02:20+02:00 `html-test --accept` deltas to reference samples - - - - - 7d444d61 by Alex Biehl at 2017-04-26T07:13:50+02:00 Remove anything related to obsolete frames mode - - - - - b888972c by Alex Biehl at 2017-04-26T07:49:10+02:00 Cherry-picked remaining commits from haddock-2.17.4-release (#603) * Release haddock/haddock-api 2.17.4 and haddock-library 1.4.3 * Set version bounds for haddock-library NB: This allows GHC 8.2.1's base * Set version bounds for haddock & haddock-api The version bounds support GHC 8.2 * Merge (temporary) v2.17.3 branch into v2.17 This allows us to delete the v2.17.3 branch * Fixup changelog * Pin down haddock-api to a single version as otherwise `haddock`'s package version has no proper meaning * fix source-repo spec for haddock-api - - - - - 4161099b by Alex Biehl at 2017-04-26T11:11:20+02:00 Update changelog to reflect news in HEAD - - - - - eed72cb8 by Alex Biehl at 2017-04-26T11:11:20+02:00 Markdownify changelog - - - - - 5815cea1 by Alex Biehl at 2017-04-26T11:32:33+02:00 Bump to 2.18.0 (#605) - - - - - a551d558 by Alex Biehl at 2017-04-29T22:00:25+02:00 Update attoparsec-0.12.1.1 to attoparsec-0.13.1.0 - - - - - ea164a8d by Sergey Vinokurov at 2017-04-29T22:42:36+02:00 Improve error message - - - - - 2e10122f by Alex Biehl at 2017-04-30T10:07:46+02:00 Correctly remember collapsed sections (#608) Now the "collapsed" cookie stores which sections have changed state instead of which are collapsed. - - - - - f9b24d99 by Alex Biehl at 2017-05-01T17:40:36+02:00 Lazily decode docMap and argMap (#610) These are only used in case of a doc reexport so most of the time decoding these is wasted work. - - - - - 2372af62 by Alex Biehl at 2017-05-01T21:59:23+02:00 Fix Binary instance for InstalledInterface (#611) (#610) introduced lazy decoding for docs from InstalledInterface but forgot to remove the original calls to get and put_ - - - - - 6c633c13 by Nathan Collins at 2017-05-11T11:47:55+02:00 Improve documenation of Haddock markup (#614) * Improve documentation of Haddock markup. - document that Haddock supports inferring types top-level functions with without type signatures, but also explain why using this feature is discouraged. Looks like this feature has been around since version 2.0.0.0 in 2008! - rework the "Module description" section: - move the general discussion of field formatting to the section intro and add examples illustrating the prose for multiline fields. - mention that newlines are preserved in some multiline fields, but not in others (I also noticed that commas in the `Copyright` field are not preserved; I'll look into this bug later). - add a subsection for the module description fields documentation, and put the field keywords in code formatting (double back ticks) instead of double quotes, to be consistent with the typesetting of keywords in other parts of the documentation. - mention that "Named chunks" are not supported in the long-form "Module description" documentation. - fix formatting of keywords in the "Module attributes" section. Perhaps these errors were left over from an automatic translation to ReST from some other format as part of the transition to using Sphinx for Haddock documentation? Also, add a missing reference here; it just said "See ?"! - update footnote about special treatment for re-exporting partially imported modules not being implemented. In my tests it's not implemented at all -- I tried re-exporting both `import B hiding (f)` and `import B (a, b)` style partial imports, and in both cases got the same result as with full imports `import B`: I only get a module reference. * Rework the `Controlling the documentation structure` section. My main goal was to better explain how to use Haddock without an export list, since that's my most common use case, but I hope I improved the section overall: - remove the incomplete `Omitting the export list` section and fold it into the other sections. In particular, summarize the differences between using and not using an export list -- i.e. control over what and in what order is documented -- in the section lead. - add "realistic" examples that use the structure markup, both with and without an export list. I wanted a realistic example here to capture how it can be useful to explain the relationship between a group of functions in a section, in addition to documenting their individual APIs. - make it clear that you can associate documentation chunks with documentation sections when you aren't using an export list, and that doing it in the most obvious way -- i.e. with `-- |`, as you can in the export list -- doesn't work without an export list. It took me a while to figure this out the first time, since the docs didn't explain it at all before. - add a "no export list" example to the section header section. - add more cross references. * Add examples of gotchas for markup in `@...@`. I'm not sure this will help anyone, since I think most people first learn about `@...@` by reading other people's Haddocks, but I've documented the mistakes which I've made and then gotten confused by. * Use consistent Capitalization of Titles. Some titles were in usual title caps, and others only had the first word capitalized. I chose making them all use title caps because that seems to make the cross references look better. - - - - - d4734f45 by Ben Gamari at 2017-05-12T20:36:08+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 93883f37 by Alex Biehl at 2017-05-12T21:02:33+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 5b8f179c by Alex Biehl at 2017-05-13T12:48:10+02:00 Consequently use inClass and notInClass in haddock-library (#617) These allow attoparsec to do some clever lookup optimization - - - - - 77984b82 by Doug Wilson at 2017-05-27T17:37:38+02:00 Don't enable compilation for template haskell (#624) This is no longer necessary after ghc commit 53c78be0aab76a3107c4dacbb1d177afacdd37fa - - - - - 5a3de2b4 by Doug Wilson at 2017-05-27T19:54:53+02:00 Improve Syb code (#621) Specialize.hs and Ast.hs are modified to have their Syb code not recurse into Name or Id in HsSyn types. Specialize.hs is refactored to have fewer calls to Syb functions. Syb.hs has some foldl calls replaced with foldl' calls. There is still a lot of performance on the floor of Ast.hs. The RenamedSource is traversed many times, and lookupBySpan is very inefficient. everywhereBut and lookupBySpan dominate the runtime whenever --hyperlinked-source is passed. - - - - - 3d35a949 by Alex Biehl at 2017-05-30T19:01:37+02:00 Clear fixme comment (#625) - - - - - 2a44bd0c by Alex Biehl at 2017-05-30T19:02:12+02:00 Make haddock-library and haddock-api warning free (#626) - - - - - bd1a0e42 by Alex Biehl at 2017-06-01T10:40:33+02:00 Include `driver-test/*.hs` sdist (#630) This lead to haskell/haddock#629. - - - - - 184a3ab6 by Doug Wilson at 2017-06-03T12:02:08+02:00 Disable pattern match warnings (#628) This disables the pattern match checker which can be very expensive in some cases. The disabled warnings include: * Opt_WarnIncompletePatterns * Opt_WarnIncompleteUniPatterns * Opt_WarnIncompletePatternsRecUpd * Opt_WarnOverlappingPatterns - - - - - 0cf68004 by Alex Biehl at 2017-06-03T20:37:28+02:00 Allow user defined signatures for pattern synonyms (#631) - - - - - 7f51a58a by Alex Biehl at 2017-06-04T11:56:38+02:00 Use NameSet for isExported check (#632) - - - - - d8f044a9 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Match new AST as per GHC wip/new-tree-one-param See https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow - - - - - da1254e3 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Rename extension index tags - - - - - 538c7514 by Christiaan Baaij at 2017-06-09T08:26:43+02:00 Haddock support for bundled pattern synonyms (#627) * Haddock support for bundled pattern synonyms * Add fixities to bundled pattern synonyms * Add bundled pattern synonyms to the synopsis * Store bundled pattern fixities in expItemFixities * Add test for bundled pattern synonyms * Stop threading fixities * Include bundled pattern synonyms for re-exported data types Sadly, fixity information isn't found for re-exported data types * Support for pattern synonyms * Modify tests after haskell/haddock#631 * Test some reexport variations * Also lookup bundled pattern synonyms from `InstalledInterface`s * Check isExported for bundled pattern synonyms * Pattern synonym is exported check * Always look for pattern synonyms in the current module Another overlooked cornercase * Account for types named twice in export lists Also introduce a fast function for nubbing on a `Name` and use it throughout the code base. * correct fixities for reexported pattern synonyms * Fuse concatMap and map * Remove obsolete import * Add pattern synonyms to visible exports * Fix test * Remove corner case - - - - - a050bffd by Doug Wilson at 2017-06-21T09:27:33+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#636) There is some performance improvement. GHC compiler: | version | bytes allocated | cpu_seconds --------------------------------- | before | 56057108648 | 41.0 | after | 51592019560 | 35.1 base: | version | bytes allocated | cpu_seconds --------------------------------- | before | 25174011784 | 14.6 | after | 23712637272 | 13.1 Cabal: | version | bytes allocated | cpu_seconds --------------------------------- | before | 18754966920 | 12.6 | after | 18198208864 | 11.6 - - - - - 5d06b871 by Doug Wilson at 2017-06-22T20:23:29+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#639) * Use new function getNameToInstancesIndex instead of tcRnGetInfo There is some significant performance improvement in the ghc testsuite. haddock.base: -23.3% haddock.Cabal: -16.7% haddock.compiler: -19.8% * Remove unused imports - - - - - b11bb73a by Alex Biehl at 2017-06-23T14:44:41+02:00 Lookup fixities for reexports without subordinates (#642) So we agree that reexported declarations which do not have subordinates (for example top-level functions) shouldn't have gotten fixities reexported according to the current logic. I wondered why for example Prelude.($) which is obviously reexported from GHC.Base has fixities attached (c.f. http://hackage.haskell.org/package/base-4.9.1.0/docs/Prelude.html#v:-36-). The reason is this: In mkMaps we lookup all the subordinates of top-level declarations, of course top-level functions don't have subordinates so for them the resulting list is empty. In haskell/haddock#644 I established the invariant that there won't be any empty lists in the subordinate map. Without the patch from haskell/haddock#642 top-level functions now started to fail reexporting their fixities. - - - - - d2a6dad6 by Alex Biehl at 2017-06-23T18:30:45+02:00 Don't include names with empty subordinates in maps (#644) These are unecessary anyway and just blow up interface size - - - - - 69c2aac4 by Alex Biehl at 2017-06-29T19:54:49+02:00 Make per-argument docs for class methods work again (#648) * Make per-argument docs for class methods work again * Test case - - - - - c9448d54 by Bartosz Nitka at 2017-07-02T12:12:01+02:00 Fix haddock: internal error: links: UnhelpfulSpan (#561) * Fix haddock: internal error: links: UnhelpfulSpan This fixes haskell/haddock#554 for me. I believe this is another fall out of `wildcard-refactor`, like haskell/haddock#549. * Comment to clarify why we take the methods name location - - - - - d4f29eb7 by Alex Biehl at 2017-07-03T19:43:04+02:00 Document record fields when DuplicateRecordFields is enabled (#649) - - - - - 9d6e3423 by Yuji Yamamoto at 2017-07-03T22:37:58+02:00 Fix test failures on Windows (#564) * Ignore .stack-work * Fix for windows: use nul instead of /dev/null * Fix for windows: canonicalize line separator * Also normalize osx line endings - - - - - 7d81e8b3 by Yuji Yamamoto at 2017-07-04T16:13:12+02:00 Avoid errors on non UTF-8 Windows (#566) * Avoid errors on non UTF-8 Windows Problem ==== haddock exits with errors like below: `(1)` ``` haddock: internal error: <stderr>: hPutChar: invalid argument (invalid character) ``` `(2)` ``` haddock: internal error: Language\Haskell\HsColour\Anchors.hs: hGetContents: invalid argument (invalid byte sequence) ``` `(1)` is caused by printing [the "bullet" character](http://www.fileformat.info/info/unicode/char/2022/index.htm) onto stderr. For example, this warning contains it: ``` Language\Haskell\HsColour\ANSI.hs:62:10: warning: [-Wmissing-methods] • No explicit implementation for ‘toEnum’ • In the instance declaration for ‘Enum Highlight’ ``` `(2)` is caused when the input file of `readFile` contains some Unicode characters. In the case above, '⇒' is the cause. Environment ---- OS: Windows 10 haddock: 2.17.3 GHC: 8.0.1 Solution ==== Add `hSetEncoding handle utf8` to avoid the errors. Note ==== - I found the detailed causes by these changes for debugging: - https://github.com/haskell/haddock/commit/8f29edb6b02691c1cf4c479f6c6f3f922b35a55b - https://github.com/haskell/haddock/commit/1dd23bf2065a1e1f2c14d0f4abd847c906b4ecb4 - These errors happen even after executing `chcp 65001` on the console. According to the debug code, `hGetEncoding stderr` returns `CP932` regardless of the console encoding. * Avoid 'internal error: <stderr>: hPutChar: invalid argument (invalid character)' non UTF-8 Windows Better solution for 59411754a6db41d17820733c076e6a72bcdbd82b's (1) - - - - - eded67d2 by Alex Biehl at 2017-07-07T19:17:15+02:00 Remove redudant import warning (#651) - - - - - 05114757 by Alex Biehl at 2017-07-08T00:33:12+02:00 Avoid missing home module warning (#652) * Avoid missing home module warning * Update haddock-library.cabal - - - - - e9cfc902 by Bryn Edwards at 2017-07-17T07:51:20+02:00 Fix haskell/haddock#249 (#655) - - - - - eb02792b by Herbert Valerio Riedel at 2017-07-20T09:09:15+02:00 Fix compilation of lib:haddock-library w/ GHC < 8 - - - - - 9200bfbc by Alex Biehl at 2017-07-20T09:20:38+02:00 Prepare 2.18.1 release (#657) - - - - - 46ddd22c by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Tweak haddock-api.cabal for pending release - - - - - 85e33d29 by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Avoid trivial use of LambdaCase otherwise we can't test w/ e.g. GHC 7.4.2 - - - - - 3afb4bfe by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Refactor .cabal to use sub-lib for vendored lib A practical benefit is that we can control the build-depends and also avoid some recompilation between library and test-suite. - - - - - e56a552e by Herbert Valerio Riedel at 2017-07-20T10:17:48+02:00 haddock-api: add changelog pointing to haddock's changelog This addresses https://github.com/haskell/haddock/issues/638#issuecomment-309283297 - - - - - 2222ff0d by Herbert Valerio Riedel at 2017-07-20T10:19:56+02:00 Drop obsolete/misleading `stability: experimental` This .cabal property has long been considered obsolete - - - - - 9b882905 by Alex Biehl at 2017-07-20T11:25:54+02:00 Beef up haddock description (#658) * Beef up haddock description * Handle empty lines - - - - - bb60e95c by Herbert Valerio Riedel at 2017-07-20T12:08:53+02:00 Import @aisamanra's Haddock cheatsheet from https://github.com/aisamanra/haddock-cheatsheet - - - - - 0761e456 by Herbert Valerio Riedel at 2017-07-20T12:12:55+02:00 Add cheatsheet to haddock.cabal - - - - - 2ece0f0f by Herbert Valerio Riedel at 2017-07-20T12:18:38+02:00 Mention new-build in README - - - - - 947b7865 by Herbert Valerio Riedel at 2017-07-20T12:32:16+02:00 Update README Also improves markup and removes/fixes redundant/obsolete parts [skip ci] - - - - - 785e09ad by Alex Biehl at 2017-07-27T07:28:57+02:00 Bump haddock to 2.18.2, haddock-library to 1.4.5 - - - - - e3ff1ca3 by Alex Biehl at 2017-07-31T20:15:32+02:00 Move `DocMarkup` from haddock-api to haddock-library (#659) * Move `DocMarkup` from haddock-api to haddock-library * Move more markup related functions * Markup module * CHANGELOG - - - - - cda7c20c by Alex Biehl at 2017-07-31T20:35:49+02:00 Fixup haddock - - - - - 583b6812 by Alex Biehl at 2017-07-31T21:20:45+02:00 Changelog for haddock-library - - - - - bac6a0eb by Alex Biehl at 2017-07-31T21:50:24+02:00 Prepare haddock-library-1.4.5 release - - - - - 58ce6877 by Moritz Drexl at 2017-08-05T16:44:40+02:00 Fix renaming after instance signature specializing (#660) * rework rename * Add regression test for Bug 613 * update tests * update changelog - - - - - b8137ec8 by Tim Baumann at 2017-08-06T11:33:38+02:00 Fix: Generate pattern signatures for constructors exported as patterns (#663) * Fix pretty-printing of pattern signatures Pattern synonyms can have up to two contexts, both having a different semantic meaning: The first holds the constraints required to perform the matching, the second contains the constraints provided by a successful pattern match. When the first context is empty but the second is not it is necessary to render the first, empty context. * Generate pattern synonym signatures for ctors exported as patterns This fixes haskell/haddock#653. * Simplify extractPatternSyn It is not necessary to generate the simplest type signature since it will be simplified when pretty-printed. * Add changelog entries for PR haskell/haddock#663 * Fix extractPatternSyn error message - - - - - d037086b by Alex Biehl at 2017-08-06T12:43:25+02:00 Bump haddock-library - - - - - 99d7e792 by Alex Biehl at 2017-08-06T12:44:07+02:00 Bump haddock-library in haddock-api - - - - - 94802a5b by Alex Biehl at 2017-08-06T13:18:02+02:00 Provide --show-interface option to dump interfaces (#645) * WIP: Provide --show-interface option to dump interfaces Like ghcs own --show-iface this flag dumps a binary interface file to stdout in a human (and machine) readable fashion. Currently it uses json as output format. * Fill all the jsonNull stubs * Rework Bifunctor instance of DocH, update changelog and documentation * replace changelog, bring DocMarkupH doc back * Update CHANGES.md * Update CHANGES.md * Move Control.Arrow up It would result in unused import if the Bifunctor instance is not generated. - - - - - c662e476 by Ryan Scott at 2017-08-14T21:00:21-04:00 Adapt to haskell/haddock#14060 - - - - - b891eb73 by Alex Biehl at 2017-08-16T08:24:48+02:00 Bifoldable and Bitraversable for DocH and MetaDoc - - - - - 021bb56c by Alex Biehl at 2017-08-16T09:06:40+02:00 Refactoring: Make doc renaming monadic This allows us to later throw warnings if can't find an identifier - - - - - 39fbf022 by Alex Biehl at 2017-08-19T20:35:27+02:00 Hyperlinker: Avoid linear lookup in enrichToken (#669) * Make Span strict in Position * Hyperlinker: Use a proper map to enrich tokens - - - - - e13baedd by Alex Biehl at 2017-08-21T20:05:42+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 27dd6e87 by Alex Biehl at 2017-08-21T22:06:35+02:00 Drop Avails from export list - - - - - 86b247e2 by Alex Biehl at 2017-08-22T08:44:22+02:00 Bump ghc version for haddock-api tests - - - - - d4607ca0 by Alex Biehl at 2017-08-22T08:45:17+02:00 Revert "Drop Avails from export list" This reverts commit a850ba86d88a4fb9c0bd175453a2580e544e3def. - - - - - c9c54c30 by Alex Biehl at 2017-08-22T09:26:01+02:00 IntefaceFile version - - - - - a85b7c02 by Ben Gamari at 2017-08-22T09:29:52-04:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 34e976f5 by Ben Gamari at 2017-08-22T17:40:06+02:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 577abf06 by Ryan Scott at 2017-08-23T14:47:29-04:00 Update for haskell/haddock#14131 - - - - - da68fc55 by Florian Eggenhofer at 2017-08-27T18:21:56+02:00 Generate an index for package content search (#662) Generate an index for package content search - - - - - 39e62302 by Alex Biehl at 2017-08-27T18:50:16+02:00 Content search for haddock html doc - - - - - 91fd6fb2 by Alex Biehl at 2017-08-28T18:39:58+02:00 Fix tests for content search - - - - - b4a3798a by Alex Biehl at 2017-08-28T18:44:08+02:00 Add search button to #page-menu - - - - - 25a7ca65 by Alex Biehl at 2017-08-28T18:47:43+02:00 Load javascript below the fold - - - - - 8d323c1a by Alex Biehl at 2017-08-28T18:49:22+02:00 Accept tests - - - - - c5dac557 by Alex Biehl at 2017-08-28T19:14:55+02:00 Content search css - - - - - 89a5af57 by Paolo Veronelli at 2017-08-29T07:42:13+02:00 Removed `nowrap` for interface method sigs (#674) with nowrap the interfaces method sigs would expand at libitum - - - - - a505f6f7 by Alex Biehl at 2017-08-29T08:05:33+02:00 Include subordinates in content index - - - - - 4bb698c4 by Alexander Biehl at 2017-08-29T11:40:19+02:00 QuickNav: Make docbase configurable - - - - - c783bf44 by Alexander Biehl at 2017-08-29T11:48:36+02:00 QuickNav: Also use baseUrl for doc-index.json request - - - - - 47017510 by Alex Biehl at 2017-08-29T17:56:47+02:00 Fix test fallout (again) - - - - - 924fc318 by Alex Biehl at 2017-08-30T09:24:56+02:00 Write meta.json when generating html output (#676) - - - - - 717dea52 by Alex Biehl at 2017-09-01T09:20:34+02:00 Use relative URL when no docBaseUrl given - - - - - e5d85f3b by Alex Biehl at 2017-09-01T09:35:19+02:00 Add missing js files to data-files (#677) - - - - - 95b9231a by Alex Biehl at 2017-09-01T11:01:36+02:00 Rename "Search" tab to "Quick Jump" - - - - - da0ead0b by Alex Biehl at 2017-09-01T13:03:49+02:00 Make trigger link configurable (#678) QuickNav: Configurable show/hide trigger - - - - - de7da594 by Ben Gamari at 2017-09-05T06:49:55-04:00 Account for "Remember the AvailInfo for each IE" As of GHC commit f609374a55bdcf3b79f3a299104767aae2ffbf21 GHC retains the AvailInfo associated with each IE. @alexbiehl has a patch making proper use of this change, but this is just to keep things building. - - - - - b05cd3b3 by Ben Gamari at 2017-09-14T07:55:07-04:00 Bump upper bound on base - - - - - 79db899e by Herbert Valerio Riedel at 2017-09-21T23:27:52+02:00 Make compatible with Prelude.<> export in GHC 8.4/base-4.11 - - - - - 3405dd52 by Tim Baumann at 2017-09-23T22:02:01+02:00 Add compile step that bundles and compresses JS files (#684) * Add compile step that bundles and compresses JS files Also, manage dependencies on third-party JS libraries using NPM. * Compile JS from TypeScript * Enable 'noImplicitAny' in TypeScript * QuickJump: use JSX syntax * Generate source maps from TypeScript for easier debugging * TypeScript: more accurate type * Separate quick jump css file from ocean theme - - - - - df0b5742 by Alex Biehl at 2017-09-29T21:15:40+02:00 Bump base for haddock-library and haddock-test - - - - - 62b12ea0 by Merijn Verstraaten at 2017-10-04T16:03:13+02:00 Inhibit output of coverage information for hidden modules. (#687) * Inhibit output of coverage information for hidden modules. * Add changelog entry. - - - - - 8daf8bc1 by Alexander Biehl at 2017-10-05T11:27:05+02:00 Don't use subMap in attachInstances - - - - - ad75114e by Alexander Biehl at 2017-10-05T11:27:58+02:00 Revert "Don't use subMap in attachInstances" This reverts commit 3adf5bcb1a6c5326ab33dc77b4aa229a91d91ce9. - - - - - 7d4aa02f by Alex Biehl at 2017-10-08T15:32:28+02:00 Precise Haddock: Use Avails for export resolution (#688) * Use Avails for export resolution * Support reexported modules * Factor out availExportItem * Use avails for fullModuleExports * Don't use subMap in attachInstances * lookupDocs without subMap * Completely remove subMap * Only calculate unqualified modules when explicit export list is given * Refactor * Refine comment * return * Fix * Refactoring * Split avail if declaration is not exported itself * Move avail splitting - - - - - b9b4faa8 by Alex Biehl at 2017-10-08T19:38:21+02:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 43325295 by Alex Biehl at 2017-10-08T20:18:46+02:00 Fix merge fallout - - - - - c6423cc0 by Alex Biehl at 2017-10-08T20:36:12+02:00 Copy QuickJump files over - - - - - 1db587c3 by Tim Baumann at 2017-10-09T18:33:09+02:00 Use <details> element for collapsibles (#690) * Remove unnecessary call to 'collapseSection' The call is unnecessary since there is no corresponding toggle for hiding the section of orphan instances. * Use <details> for collapsibles This makes them work even when JS is disabled. Closes haskell/haddock#560. - - - - - 1b54c64b by Tim Baumann at 2017-10-10T09:50:59+02:00 Quick Jump: Show error when loading 'doc-index.json' failed (#691) - - - - - 910f716d by Veronika Romashkina at 2017-10-24T07:36:20+02:00 Fix tiny typo in docs (#693) - - - - - b21de7e5 by Ryan Scott at 2017-10-24T13:07:15+02:00 Overhaul Haddock's rendering of kind signatures (#681) * Overhaul Haddock's rendering of kind signatures * Strip off kind signatures when specializing As an added bonus, this lets us remove an ugly hack specifically for `(->)`. Yay! * Update due to 0390e4a0f61e37bd1dcc24a36d499e92f2561b67 * @alexbiehl's suggestions * Import injectiveVarsOfBinder from GHC - - - - - 6704405c by Ryan Scott at 2017-10-28T07:10:27+02:00 Fix Haddock rendering of kind-indexed data family instances (#694) - - - - - 470f6b9c by Alex Biehl at 2017-10-30T08:45:51+01:00 Add QuickJump version to meta.json (#696) - - - - - b89eccdf by Alex Biehl at 2017-10-30T10:15:49+01:00 Put Quickjump behind --quickjump flag (#697) - - - - - 3095fb58 by Alex Biehl at 2017-10-30T19:09:06+01:00 Add build command to package.json - - - - - f223fda9 by Alex Biehl at 2017-10-30T19:10:39+01:00 Decrease threshold for fuzzy matching - - - - - 80245dda by Edward Z. Yang at 2017-10-31T20:35:05+01:00 Supported reexported-modules via --reexport flag. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 7e389742 by Alex Biehl at 2017-10-31T20:37:56+01:00 Correct missing title in changelog - - - - - 1a2a1c03 by Alex Biehl at 2017-10-31T20:59:07+01:00 Copy quickjump.css for nicer error messages - - - - - db234bb9 by Alex Biehl at 2017-10-31T21:31:18+01:00 Reexported modules: Report warnings if argument cannot be parsed or ... module cannot be found - - - - - eea8a205 by Carlo Hamalainen at 2017-10-31T21:43:14+01:00 More general type for nameCacheFromGhc. (#539) - - - - - 580eb42a by Alex Biehl at 2017-10-31T21:46:52+01:00 Remote tab - - - - - 0e599498 by Alex Biehl at 2017-10-31T21:48:55+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 7b8539bb by Alex Biehl at 2017-10-31T22:28:34+01:00 fullModuleContents traverses exports in declaration order - - - - - 0c91fbf2 by Alex Biehl at 2017-10-31T22:32:31+01:00 Remove excessive use of list comprehensions - - - - - f7356e02 by Alex Biehl at 2017-11-01T19:11:03+01:00 Make better use of AvailInfo - - - - - f3e512d5 by Alex Biehl at 2017-11-02T12:16:22+01:00 Always return documentation for exported subordinates ... event if they have no documentation (e.g. noDocForDecl) By using the information in the AvailInfo we don't need additional export checks. - - - - - 7cf58898 by Alan Zimmerman at 2017-11-07T08:28:03+02:00 Match changes for Trees that Grow in GHC - - - - - e5105a41 by Alan Zimmerman at 2017-11-08T17:21:58+02:00 Match Trees That Grow - - - - - 55178266 by Alan Zimmerman at 2017-11-11T22:20:31+02:00 Match Trees that Grow in GHC for HsExpr - - - - - 2082ab02 by Ryan Scott at 2017-11-14T15:27:03+01:00 Actually render infix type operators as infix (#703) * Actually render infix type operators as infix * Account for things like `(f :*: g) p`, too - - - - - c52ab7d0 by Alan Zimmerman at 2017-11-14T23:14:26+02:00 Clean up use of PlaceHolder, to match TTG - - - - - 81cc9851 by Moritz Angermann at 2017-11-20T07:52:49+01:00 Declare use of `Paths_haddock` module in other-modules (#705) This was detected by `-Wmissing-home-modules` - - - - - f9d27598 by Moritz Angermann at 2017-11-20T12:47:34+01:00 Drop Paths_haddock from ghc.mk (#707) With haskell/haddock#705 and haskell/haddock#706, the custom addition should not be necessary any more. # Conflicts: # ghc.mk - - - - - f34818dc by Moritz Angermann at 2017-11-20T12:47:59+01:00 Add autogen-modules (#706) > Packages using 'cabal-version: >= 1.25' and the autogenerated module Paths_* must include it also on the 'autogen-modules' field besides 'exposed-modules' and 'other-modules'. This specifies that the module does not come with the package and is generated on setup. Modules built with a custom Setup.hs script also go here to ensure that commands like sdist don't fail. # Conflicts: # haddock.cabal - - - - - bb43a0aa by Ben Gamari at 2017-11-21T15:50:12-05:00 Revert "Clean up use of PlaceHolder, to match TTG" This reverts commit 134a7bb054ea730b13c8629a76232d73e3ace049. - - - - - af9ebb2b by Ben Gamari at 2017-11-21T15:50:14-05:00 Revert "Match Trees that Grow in GHC for HsExpr" This reverts commit 9f054dc365379c66668de6719840918190ae6e44. - - - - - 5d35c3af by Ben Gamari at 2017-11-21T15:50:15-05:00 Revert "Match Trees That Grow" This reverts commit 73a26af844ac50b8bec39de11d64452a6286b00c. - - - - - 99a8e43b by Ben Gamari at 2017-11-21T16:36:06-05:00 Revert "Match changes for Trees that Grow in GHC" This reverts commit 01eeeb048acd2dd05ff6471ae148a97cf0720547. - - - - - c4d650c2 by Ben Gamari at 2017-12-04T15:06:07-05:00 Bump GHC version - - - - - 027b2274 by Ben Gamari at 2017-12-04T17:06:31-05:00 Bump GHC bound to 8.4.* - - - - - 58eaf755 by Alex Biehl at 2017-12-06T15:44:24+01:00 Update changelog - - - - - d68f5584 by Simon Peyton Jones at 2017-12-07T14:39:56+00:00 Track changes to follow Trac haskell/haddock#14529 This tracks the refactoring of HsDecl.ConDecl. - - - - - dc519d6b by Alec Theriault at 2018-01-06T08:20:43-08:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - 8285118c by Alec Theriault at 2018-01-13T12:12:37+01:00 Constructor and pattern synonym argument docs (#709) * Support Haddocks on constructor arguments This is in conjunction with https://phabricator.haskell.org/D4094. Adds support for rendering Haddock's on (non-record) constructor arguments, both for regular and GADT constructors. * Support haddocks on pattern synonym arguments It appears that GHC already parsed these - we just weren't using them. In the process of doing this, I tried to deduplicate some code around handling patterns. * Update the markup guide Add some information about the new support for commenting constructor arguments, and mention pattern synonyms and GADT-style constructors. * Overhaul LaTeX support for data/pattern decls This includes at least * fixing several bugs that resulted in invalid LaTeX * fixing GADT data declaration headers * overhaul handling of record fields * overhaul handling of GADT constructors * overhaul handling of bundled patterns * add support for constructor argument docs * Support GADT record constructors This means changes what existing HTML docs look like. As for LaTeX, looks like GADT records were never even supported. Now they are. * Clean up code/comments Made code/comments consistent between the LaTeX and XHTML backend when possible. * Update changelog * Patch post-rebase regressions * Another post-rebase change We want return values to be documentable on record GADT constructors. - - - - - ca4fabb4 by Alec Theriault at 2018-01-15T17:12:18-08:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 4c472fea by Ryan Scott at 2018-01-19T10:44:02+01:00 Fix haskell/haddock#732 (#733) - - - - - bff14dbd by Alex Biehl at 2018-01-19T15:33:30+01:00 extractDecl: Extract associated types correctly (#736) - - - - - a2a94a73 by Alex Biehl at 2018-01-19T15:34:40+01:00 extractDecl: Extract associated types correctly (#736) - - - - - 26df93dc by Alex Biehl at 2018-01-20T10:18:22+01:00 haddock-api: bump ghc to ^>= 8.4 - - - - - f65aeb1d by Alex Biehl at 2018-01-20T19:18:20+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - 0e721b97 by Alex Biehl at 2018-01-20T19:20:19+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - cb6234f6 by Ben Gamari at 2018-01-26T13:40:55-05:00 Merge remote-tracking branch 'harpocrates/fix/missing-orphan-instances' into ghc-head - - - - - 0fc28554 by Alec Theriault at 2018-02-01T14:58:18+01:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - b9123772 by Alec Theriault at 2018-02-01T14:58:18+01:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 0c12e274 by Ryan Scott at 2018-02-01T14:58:18+01:00 Fix haskell/haddock#548 by rendering datatype kinds more carefully (#702) - - - - - 8876d20b by Alec Theriault at 2018-02-01T14:58:18+01:00 Use the GHC lexer for the Hyperlinker backend (#714) * Start changing to use GHC lexer * better cpp * Change SrcSpan to RealSrcSpan * Remove error * Try to stop too many open files * wip * wip * Revert "wip" This reverts commit b605510a195f26315e3d8ca90e6d95a6737553e1. Conflicts: haddock-api/haddock-api.cabal haddock-api/src/Haddock/Interface.hs * Remove pointless 'caching' * Use dlist rather than lists when finding vars * Use a map rather than list * Delete bogus comment * Rebase followup Things now run using the GHC lexer. There are still - stray debug statements - unnecessary changes w.r.t. master * Cleaned up differences w.r.t. current Haddock HEAD Things are looking good. quasiquotes in particular look beautiful: the TH ones (with Haskell source inside) colour/link their contents too! Haven't yet begun to check for possible performance problems. * Support CPP and top-level pragmas The support for these is hackier - but no more hacky than the existing support. * Tests pass, CPP is better recognized The tests were in some cases altered: I consider the new output to be more correct than the old one.... * Fix shrinking of source without tabs in test * Replace 'Position'/'Span' with GHC counterparts Replaces 'Position' -> 'GHC.RealSrcLoc' and 'Span' -> 'GHC.RealSrcSpan'. * Nits * Forgot entry in .cabal * Update changelog - - - - - 95c6a771 by Alec Theriault at 2018-02-01T14:58:18+01:00 Clickable anchors for headings (#716) See haskell/haddock#579. This just adds an <a> tag around the heading, pointing to the heading itself. - - - - - 21463d28 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump: Matches on function names weight more than matches in ... module names. - - - - - 8023af39 by Alex Biehl at 2018-02-01T14:58:18+01:00 Treat escaped \] better in definition lists (#717) This fixes haskell/haddock#546. - - - - - e4866dc1 by Alex Biehl at 2018-02-01T14:58:18+01:00 Remove scanner, takeWhile1_ already takes care of escaping - - - - - 9bcaa49d by Alex Biehl at 2018-02-01T14:58:18+01:00 Take until line feed - - - - - 01d2af93 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Add simple framework for running parser fixtures (#668) * Add simple framework for running parser fixtures * Compatible with tree-diff-0.0.0.1 * Use parseParas to parse fixtures This allows to test all syntactic constructs available in haddock markup. - - - - - 31128417 by Alec Theriault at 2018-02-01T14:58:18+01:00 Patch flaky parser test (#720) * Patch flaky parser test This test was a great idea, but it doesn't port over too well to using the GHC lexer. GHC rewrites its input a bit - nothing surprising, but we need to guard against those cases for the test. * Change instance head * Change use site - - - - - 9704f214 by Herbert Valerio Riedel at 2018-02-01T14:58:18+01:00 Include secondary LICENSE file in source dist - - - - - 51f25074 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Grid Tables (#718) * Add table examples * Add table types and adopt simple parser Simple parser is done by Giovanni Cappellotto (@potomak) in https://github.com/haskell/haddock/pull/577 It seems to support single fine full tables, so far from full RST-grid tables, but it's good start. Table type support row- and colspans, but obviously parser is lacking. Still TODO: - Latex backend. Should we use multirow package https://ctan.org/pkg/multirow?lang=en? - Hoogle backend: ? * Implement grid-tables * Refactor table parser * Add two ill-examples * Update CHANGES.md * Basic documentation for tables * Fix documentation example - - - - - 670d6200 by Alex Biehl at 2018-02-01T14:58:18+01:00 Add grid table example to cheatsheet (pdf and svg need to be regenerated thought) - - - - - 4262dec9 by Alec Theriault at 2018-02-01T14:58:18+01:00 Fix infinite loop when specializing instance heads (#723) * Fix infinite loop when specializing instance heads The bug can only be triggered from TH, hence why it went un-noticed for so long. * Add test for haskell/haddock#679 and haskell/haddock#710 - - - - - 67ecd803 by Alec Theriault at 2018-02-01T14:58:18+01:00 Filter RTS arguments from 'ghc-options' arguments (#725) This fixes haskell/haddock#666. - - - - - 7db26992 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump Scrollable overlay - - - - - da9ff634 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Hyperlinker: Adjust parser to new PFailed constructor - - - - - 7b7cf8cb by Alexander Biehl at 2018-02-01T14:58:18+01:00 Specialize: Add missing IdP annotations - - - - - 78cd7231 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Convert: Correct pass type - - - - - a2d0f590 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Warning free compilation - - - - - cd861cf3 by Alexander Biehl at 2018-02-01T14:58:18+01:00 hadock-2.19.0 / haddock-api-2.19.0 / haddock-library-1.5.0 - - - - - c6651b72 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Adjust changelogs - - - - - 1e93da0b by Alexander Biehl at 2018-02-01T14:58:18+01:00 haddock-library: Info about breaking changes - - - - - f9b11db8 by Alec Theriault at 2018-02-02T12:36:02+01:00 Properly color pragma contents in hyperlinker The hyperlinker backend now classifies the content of pragmas as 'TkPragma'. That means that in something like '{-# INLINE foo #-}', 'foo' still gets classified as a pragma token. - - - - - c40b0043 by Alec Theriault at 2018-02-02T12:36:02+01:00 Support the new 'ITcolumn_prag' token - - - - - 4a2a4d39 by Alex Biehl at 2018-02-03T12:11:55+01:00 QuickJump: Mitigate encoding problems on Windows - - - - - bb34503a by Alex Biehl at 2018-02-04T18:39:31+01:00 Use withBinaryFile - - - - - 637605bf by Herbert Valerio Riedel at 2018-02-05T09:48:32+01:00 Try GHC 8.4.1 for Travis CI job - - - - - 7abb67e4 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 try harder to build w/ GHC 8.4.1 - - - - - 8255cc98 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 Add `SPDX-License-Identifier` as alised for "license" module header tokens C.f. SPDX 2.1 - Appendix V https://spdx.org/spdx-specification-21-web-version#h.twlc0ztnng3b The tag should appear on its own line in the source file, generally as part of a comment. SPDX-License-Identifier: <SPDX License Expression> Cherry-picked from haskell/haddock#743 - - - - - 267cd23d by Herbert Valerio Riedel at 2018-02-05T10:24:34+01:00 Make test-suite SMP compatible - - - - - 95d4bf40 by Alec Theriault at 2018-02-05T22:01:04+01:00 Hyperlink pattern synonyms and 'module' imports (#744) Links to pattern synonyms are now generated, as well as links from modules in import lists. Fixes haskell/haddock#731. - - - - - 67838dcd by Alec Theriault at 2018-02-06T08:23:36+01:00 Don't warn about missing '~' (#746) This manually filters out '~' from the list of things to warn about. It truly makes no sense to warn on this since '~' has nothing it could link to - it is magical. This fixes haskell/haddock#532. - - - - - ab6c3f9f by Alec Theriault at 2018-02-06T08:24:47+01:00 Don't barf on 'HsSpliceTy' (#745) This handles 'HsSpliceTy's by replacing them with what they expand to. IIUC everything that is happening, 'renameHsSpliceTy' should not be able to fail for the inputs we feed it from GHC. This fixes haskell/haddock#574. - - - - - 92bf95ad by Alex Biehl at 2018-02-06T08:28:23+01:00 Rename: renameHsSpliceTy ttg - - - - - 3130b1e1 by Alex Biehl at 2018-02-06T09:02:14+01:00 Expand SigDs - - - - - c72adae5 by Alex Biehl at 2018-02-06T09:20:51+01:00 fullModuleContents: support named docs - - - - - de2e4dbf by Alex Biehl at 2018-02-06T13:56:17+01:00 Hyperlinker: Also link pattern synonym arguments - - - - - b7c98237 by Alex Biehl at 2018-02-09T18:44:23+01:00 Expand SigD in a better place In https://github.com/haskell/haddock/issues/287 we found that haddock-2.19.0 would miss documentation on class methods with multiples names. This patch uses expandSigDecls in a more sensible place. - - - - - 8f598b27 by Alec Theriault at 2018-02-11T12:29:56+01:00 Add module tooltips to linked identifiers (#753) No more clicking to figure out whether your bytestring is strict or lazy! - - - - - d812e65d by Alec Theriault at 2018-02-11T12:31:44+01:00 Add 'show' option to complement 'hide' (#752) * Add 'show' option to complement 'hide' The behaviour is for flags passed in the command line to override flags in file headers. In the command line, later flags override earlier ones. Fixes haskell/haddock#751 and haskell/haddock#266. * Add a '--show-all' option - - - - - 6676cecb by Alex Biehl at 2018-02-18T11:07:15-05:00 QuickJump: Mitigate encoding problems on Windows (cherry picked from commit 86292c54bfee2343aee84559ec01f1fc68f52231) - - - - - e753dd88 by Alex Biehl at 2018-02-18T17:59:54+01:00 Use withBinaryFile - - - - - 724dc881 by Tamar Christina at 2018-02-19T05:34:49+01:00 Haddock: support splitted include paths. (#689) - - - - - 9b6d6f50 by Alex Biehl at 2018-02-19T05:57:02+01:00 Teach the HTML backend how to render methods with multiple names - - - - - a74aa754 by Alexander Biehl at 2018-02-19T10:04:34+01:00 Hoogle/Latex: Remove use of partial function - - - - - 66d8bb0e by Alec Theriault at 2018-02-25T16:04:01+01:00 Fix file handle leak (#763) (#764) Brought back some mistakenly deleted code for handling encoding and eager reading of files from e0ada1743cb722d2f82498a95b201f3ffb303137. - - - - - bb92d03d by Alex Biehl at 2018-03-02T14:21:23+01:00 Enable running test suite with stock haddock and ghc using ``` $ cabal new-run -- html-test --haddock-path=$(which haddock) --ghc-path=$(which ghc) ``` - - - - - dddb3cb2 by Alex Biehl at 2018-03-02T15:43:21+01:00 Make testsuite work with haddock-1.19.0 release (#766) - - - - - f38636ed by Alec Theriault at 2018-03-02T15:48:36+01:00 Support unicode operators, proper modules Unicode operators are a pretty big thing in Haskell, so supporting linking them seems like it outweighs the cost of the extra machinery to force Attoparsec to look for unicode. Fixes haskell/haddock#458. - - - - - 09d89f7c by Alec Theriault at 2018-03-02T15:48:43+01:00 Remove bang pattern - - - - - d150a687 by Alex Biehl at 2018-03-02T15:48:48+01:00 fix test - - - - - d6fd71a5 by Alex Biehl at 2018-03-02T16:22:38+01:00 haddock-test: Be more explicit which packages to pass We now pass `-hide-all-packages` to haddock when invoking the testsuite. This ensures we don't accidentally pick up any dependencies up through ghc.env files. - - - - - 0932c78c by Alex Biehl at 2018-03-02T17:50:38+01:00 Revert "fix test" This reverts commit 1ac2f9569242f6cb074ba6e577285a4c33ae1197. - - - - - 52516029 by Alex Biehl at 2018-03-02T18:16:50+01:00 Fix Bug548 for real - - - - - 89df9eb5 by Alex Biehl at 2018-03-05T18:28:19+01:00 Hyperlinker: Links for TyOps, class methods and associated types - - - - - d019a4cb by Ryan Scott at 2018-03-06T13:43:56-05:00 Updates for haskell/haddock#13324 - - - - - 6d5a42ce by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump haddock-2.19.0.1, haddock-api-2.19.0.1, haddock-library-1.5.0.1 - - - - - c0e6f380 by Alex Biehl at 2018-03-10T18:25:57+01:00 Update changelogs for haddock-2.19.0.1 and haddock-library-1.5.0.1 - - - - - 500da489 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Update to QC 2.11 - - - - - ce8362e9 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Restore backward-compat with base-4.5 through base-4.8 - - - - - baae4435 by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump lower bound for haddock-library - - - - - 10b7a73e by Alex Biehl at 2018-03-10T18:25:57+01:00 Haddock: Straighten out base bound - - - - - a6096f7b by Alex Biehl at 2018-03-13T08:45:06+01:00 extractDecl: Extract constructor patterns from data family instances (#776) * extractDecl: Allow extraction of data family instance constructors * extractDecl: extract data family instance constructors - - - - - ba4a0744 by Simon Jakobi at 2018-03-14T08:26:42+01:00 Readme: Update GHC version (#778) - - - - - 8de157d4 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for definition lists - - - - - 425b46f9 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for links - - - - - d53945d8 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for inline links - - - - - f1dc7c99 by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Slightly unmangle output - - - - - 0879d31c by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Prevent stdout buffering - - - - - 1f9e5f1b by Simon Jakobi at 2018-03-14T20:39:29+01:00 haddock-library.cabal: Clean up GHC options - - - - - 066b891a by Simon Jakobi at 2018-03-14T20:39:29+01:00 Make a proper definition for the <link> parser - - - - - 573d6ba7 by Alec Theriault at 2018-03-21T09:16:57+01:00 Show where instances are defined (#748) * Indicate source module of instances Above instance, we now also display a link to the module where the instance was defined. This is sometimes helpful in figuring out what to import. * Source module for type/data families too * Remove parens * Accept tests - - - - - 99b5d28b by Alex Biehl at 2018-03-21T09:20:36+01:00 Prepare changelog for next release - - - - - 482d3a93 by Alex Biehl at 2018-03-23T15:57:36+01:00 Useful cost centres, timers and allocation counters (#785) * Add some useful cost-centres for profiling * Add withTiming for each haddock phase Invoking haddock with `--optghc=-ddump-timings` now shows the amount of time spent and the number of allocated bytes for each phase. - - - - - 773b41bb by Alec Theriault at 2018-03-27T08:35:59+02:00 @since includes package name (#749) * Metadoc stores a package name This means that '@since' annotations can be package aware. * Get the package name the right way This should extract the package name for `@since` annotations the right way. I had to move `modulePackageInfo` around to do this and, in the process, I took the liberty to update it. Since it appears that finding the package name is something that can fail, I added a warning for this case. * Silence warnings * Hide package for local 'since' annotations As discussed, this is still the usual case (and we should avoid being noisy for it). Although this commit is large, it is basically only about threading a 'Maybe Package' from 'Haddock.render' all the way to 'Haddock.Backends.Xhtml.DocMarkup.renderMeta'. * Bump binary interface version * Add a '--since-qual' option This controls when to qualify since annotations with the package they come from. The default is always, but I've left an 'external' variant where only those annotations coming from outside of the current package are qualified. * Make ParserSpec work * Make Fixtures work * Use package name even if package version is not available The @since stuff needs only the package name passed in, so it makes sense to not be forced to pass in a version too. - - - - - e42c57bc by Alex Biehl at 2018-03-27T08:42:50+02:00 haddock-2.19.1, haddock-api-2.19.1, haddock-library-1.6.0 - - - - - 8373a529 by Alex Biehl at 2018-03-28T10:17:11+02:00 Bump haddock and haddock-api to 2.20.0 - - - - - 5038eddd by Jack Henahan at 2018-04-03T13:28:12+02:00 Clear search string on hide for haskell/haddock#781 (#789) - - - - - 920ca1eb by Alex Biehl at 2018-04-03T16:35:50+02:00 Travis: Build with ghc-8.4.2 (#793) - - - - - a232f0eb by Alan Zimmerman at 2018-04-07T14:14:32+02:00 Match changes in GHC for D4199 Removing HasSourceText and SourceTextX classes. - - - - - ab85060b by Alan Zimmerman at 2018-04-09T21:20:24+02:00 Match GHC changes for TTG - - - - - 739302b6 by Alan Zimmerman at 2018-04-13T13:31:44+02:00 Match GHC for TTG implemented on HsBinds, D4581 - - - - - 2f56d3cb by Ryan Scott at 2018-04-19T11:42:58-04:00 Bump upper bound on base to < 4.13 See https://ghc.haskell.org/trac/ghc/ticket/15018. - - - - - a49df92a by Alex Biehl at 2018-04-20T07:31:44+02:00 Don't treat fixity signatures like declarations - - - - - d02c103b by Ryan Scott at 2018-04-24T11:20:11-04:00 Add regression test for haskell/haddock#413 Fixes haskell/haddock#413. - - - - - c7577f52 by Ryan Scott at 2018-04-24T13:51:06-07:00 Improve the Hoogle backend's treatment of type families (#808) Fixes parts 1 and 2 of haskell/haddock#806. - - - - - d88f85b1 by Alec Theriault at 2018-04-25T11:24:07-07:00 Replace 'attoparsec' with 'parsec' (#799) * Remove attoparsec with parsec and start fixing failed parses * Make tests pass * Fix encoding issues The Haddock parser no longer needs to worry about bytestrings. All the internal parsing work in haddock-library happens over 'Text'. * Remove attoparsec vendor * Fix stuff broken in 'attoparsec' -> 'parsec' * hyperlinks * codeblocks * examples Pretty much all issues are due to attoparsec's backtracking failure behaviour vs. parsec's non-backtracking failure behaviour. * Fix small TODOs * Missing quote + Haddocks * Better handle spaces before/after paragraphs * Address review comments - - - - - fc25e2fe by Alan Zimmerman at 2018-04-27T15:36:53+02:00 Match changes in GHC for TTG - - - - - 06175f91 by Herbert Valerio Riedel at 2018-05-01T18:11:09+02:00 Merge branch 'ghc-head' with 'ghc-8.4' - - - - - 879caaa8 by Alec Theriault at 2018-05-07T18:53:15-07:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 3e0120cb by Simon Jakobi at 2018-05-07T19:00:18-07:00 Add docs for some DocH constructors (#814) - - - - - 0a32c6db by Alec Theriault at 2018-05-08T02:15:45-07:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 8816e783 by Simon Jakobi at 2018-05-08T10:48:11-07:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - ad60366f by Ryan Scott at 2018-05-10T11:19:47-04:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - 03b7cc3b by Ryan Scott at 2018-05-10T11:24:38-04:00 Wibbles - - - - - b03dd563 by Chaitanya Koparkar at 2018-05-10T11:44:58-04:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 9f298a40 by Ben Gamari at 2018-05-13T17:36:04-04:00 Account for refactoring of LitString - - - - - ea3dabe7 by Ryan Scott at 2018-05-16T09:21:43-04:00 Merge pull request haskell/haddock#826 from haskell/T825 Remove Hoogle backend hack that butchers infix datatype names - - - - - 0d234f7c by Alec Theriault at 2018-05-23T11:29:05+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 15fc9712 by Simon Jakobi at 2018-05-31T04:17:47+02:00 Adjust to new HsDocString internals - - - - - 6f1e19a8 by Ben Gamari at 2018-06-02T16:18:58-04:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 0d0355d9 by Ryan Scott at 2018-06-04T21:26:59-04:00 DerivingVia changes - - - - - 0d93475a by Simon Jakobi at 2018-06-05T19:47:05+02:00 Bump a few dependency bounds (#845) - - - - - 5cbef804 by Alec Theriault at 2018-06-05T19:47:16+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 9094c56f by Alec Theriault at 2018-06-05T22:53:25+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 70188719 by Simon Jakobi at 2018-06-08T22:20:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - 495cd1fc by Chaitanya Koparkar at 2018-06-13T23:01:34+02:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 81088732 by Ben Gamari at 2018-06-13T23:01:34+02:00 Account for refactoring of LitString - - - - - 7baf6587 by Simon Jakobi at 2018-06-13T23:05:08+02:00 Adjust to new HsDocString internals - - - - - bb61464d by Ben Gamari at 2018-06-13T23:05:22+02:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 5d8cb87f by Ryan Scott at 2018-06-13T23:39:30+02:00 DerivingVia changes - - - - - 73d373a3 by Alec Theriault at 2018-06-13T23:39:30+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 4865e254 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Remove `ITtildehsh` token - - - - - b867db54 by Alec Theriault at 2018-06-13T23:39:30+02:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 9598e392 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Add docs for some DocH constructors (#814) - - - - - 8a59035b by Alec Theriault at 2018-06-13T23:39:30+02:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 29350fc8 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - 2590bbd9 by Ryan Scott at 2018-06-13T23:39:30+02:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - a9939fdc by Ryan Scott at 2018-06-13T23:39:30+02:00 Wibbles - - - - - a22f7df4 by Alec Theriault at 2018-06-13T23:39:30+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 8741015d by Simon Jakobi at 2018-06-13T23:39:30+02:00 Bump a few dependency bounds (#845) - - - - - 4791e1cc by Alec Theriault at 2018-06-13T23:39:30+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 311d3216 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - d0577817 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Complete FixitySig and FamilyDecl pattern matches - - - - - 055b3aa7 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Fix redundant import warnings - - - - - f9ce19b1 by Simon Jakobi at 2018-06-13T23:49:52+02:00 html-test: Accept output - - - - - 04604ea7 by Simon Jakobi at 2018-06-13T23:54:37+02:00 Bump bounds on Cabal - - - - - 0713b692 by Simon Jakobi at 2018-06-14T00:00:12+02:00 Merge branch 'ghc-head' into ghc-head-update-3 - - - - - c6a56bfd by Simon Jakobi at 2018-06-14T02:33:27+02:00 Bump ghc bound for haddock-api spec test-suite - - - - - 119d04b2 by Simon Jakobi at 2018-06-14T12:37:48+02:00 Travis: `--allow-newer` for all packages - - - - - 0e876e2c by Alex Biehl at 2018-06-14T15:28:52+02:00 Merge pull request haskell/haddock#857 from sjakobi/ghc-head-update-3 Update ghc-head - - - - - 5be46454 by Alec Theriault at 2018-06-14T21:42:45+02:00 Improved handling of interfaces in 'haddock-test' (#851) This should now work with an inplace GHC where (for instance) HTML directories may not be properly recorded in the package DB. - - - - - 96ab1387 by Vladislav Zavialov at 2018-06-14T17:06:21-04:00 Handle -XStarIsType - - - - - e518f8c4 by Ben Gamari at 2018-06-14T17:48:00-04:00 Revert unintentional reversion of fix of haskell/haddock#548 - - - - - 01b9f96d by Alan Zimmerman at 2018-06-19T11:52:22+02:00 Match changes in GHC for haskell/haddock#14259 - - - - - 7f8c8298 by Ben Gamari at 2018-06-19T18:14:27-04:00 Bump GHC version to 8.6 - - - - - 11c6b5d2 by Ryan Scott at 2018-06-19T23:17:31-04:00 Remove HsEqTy and XEqTy - - - - - b33347c2 by Herbert Valerio Riedel at 2018-06-20T23:14:52+02:00 Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. - - - - - f0d2460e by Herbert Valerio Riedel at 2018-06-20T23:28:46+02:00 Update Travis CI job - - - - - ef239223 by Herbert Valerio Riedel at 2018-06-20T23:32:41+02:00 Drop GHC HEAD from CI and update GHC to 8.4.3 It's a waste of resource to even try to build this branch w/ ghc-head; so let's not do that... - - - - - 41c4a9fa by Ben Gamari at 2018-06-20T18:26:20-04:00 Bump GHC version to 8.7 - - - - - 8be593dc by Herbert Valerio Riedel at 2018-06-21T22:32:15+02:00 Update CI job to use GHC 8.7.* - - - - - b91d334a by Simon Jakobi at 2018-06-30T13:41:38+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section - - - - - f707d848 by Alec Theriault at 2018-07-05T10:43:35-04:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. - - - - - a6d2b8dc by Alec Theriault at 2018-07-06T10:06:32-04:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case - - - - - 13819f71 by Alan Zimmerman at 2018-07-15T19:33:51+02:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 - - - - - c346aa78 by Simon Jakobi at 2018-07-19T12:29:32+02:00 haddock-library: Bump bounds for containers - - - - - 722e733c by Simon Jakobi at 2018-07-19T13:36:45+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] - - - - - f0bd83fd by Alec Theriault at 2018-07-19T14:39:57+02:00 Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. - - - - - 394053a8 by Simon Jakobi at 2018-07-19T14:58:07+02:00 haddock-library: Bump bounds for containers - - - - - 1bda11a2 by Alec Theriault at 2018-07-20T09:04:03+02:00 Add HEAD.hackage overlay (#887) * Add HEAD.hackage overlay * Add HCPKG variable - - - - - c7b4ab45 by Alec Theriault at 2018-07-20T12:01:16+02:00 Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test - - - - - c05d32ad by Alec Theriault at 2018-07-20T12:01:49+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output - - - - - 24b39ee4 by Alec Theriault at 2018-07-20T12:02:16+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. - - - - - cb9d2099 by Simon Jakobi at 2018-07-20T13:39:29+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) - - - - - 133f24f5 by Alec Theriault at 2018-07-20T13:39:29+02:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) - - - - - 11024149 by Alec Theriault at 2018-07-20T13:39:29+02:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) - - - - - de0c139e by Simon Jakobi at 2018-07-20T13:39:29+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) - - - - - 6435e952 by Alec Theriault at 2018-07-20T13:39:29+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) - - - - - 1461af39 by Alec Theriault at 2018-07-20T13:39:29+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) - - - - - 69d3bde1 by Alec Theriault at 2018-07-20T13:49:47+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) - - - - - 6a5c73c7 by Alec Theriault at 2018-07-20T13:50:00+02:00 Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests - - - - - 92ca94c6 by Alec Theriault at 2018-07-20T13:55:36+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) - - - - - 981bc7fa by Simon Jakobi at 2018-07-20T15:06:06+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers - - - - - 27e7c0c5 by Simon Jakobi at 2018-07-20T15:09:05+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 49e1a415 by Simon Jakobi at 2018-07-20T16:02:02+02:00 Update the ghc-8.6 branch (#889) * Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. * README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) * Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) * Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) * tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) * Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. * Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test * Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) * Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) * Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) * Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests * Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 5ca14bed by Simon Jakobi at 2018-07-20T16:05:47+02:00 Revert "Revert "Bump GHC version to 8.6"" That commit didn't belong onto the ghc-8.6 branch. This reverts commit acbaef3b9daf1d2dea10017964bf886e77a8e967. - - - - - 2dd600dd by Simon Jakobi at 2018-07-20T16:18:21+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. - - - - - fceb2422 by Simon Jakobi at 2018-07-20T16:18:21+02:00 outOfScope: Recommend qualifying the identifier - - - - - acea5d23 by Simon Jakobi at 2018-07-20T16:19:35+02:00 outOfScope: Recommend qualifying the identifier (cherry picked from commit 73707ed58d879cc04cb644c5dab88c39ca1465b7) - - - - - 1a83ca55 by Simon Jakobi at 2018-07-20T16:19:35+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. (cherry picked from commit d504a2864a4e1982e142cf88c023e7caeea3b76f) - - - - - 48374451 by Masahiro Sakai at 2018-07-20T17:06:42+02:00 Add # as a special character (#884) '#' has special meaning used for anchors and can be escaped using backslash. Therefore it would be nice to be listed as special characters. - - - - - 5e1a5275 by Alec Theriault at 2018-07-20T23:37:24+02:00 Let `haddock-test` bypass interface version check (#890) This means `haddock-test` might * crash during deserialization * deserialize incorrectly Still - it means things _might_ work where they were previously sure not to. - - - - - 27286754 by Yuji Yamamoto at 2018-07-23T08:16:01+02:00 Avoid "invalid argument (invalid character)" on non-unicode Windows (#892) Steps to reproduce and the error message ==== ``` > stack haddock basement ... snip ... Warning: 'A' is out of scope. Warning: 'haddock: internal error: <stdout>: commitBuffer: invalid argument (invalid character) ``` Environment ==== OS: Windows 10 ver. 1709 haddock: [HEAD of ghc-8.4 when I reproduce the error](https://github.com/haskell/haddock/commit/532b209d127e4cecdbf7e9e3dcf4f653a5605b5a). (I had to use this version to avoid another probrem already fixed in HEAD) GHC: 8.4.3 stack: Version 1.7.1, Git revision 681c800873816c022739ca7ed14755e85a579565 (5807 commits) x86_64 hpack-0.28.2 Related pull request ==== https://github.com/haskell/haddock/pull/566 - - - - - 6729d361 by Alec Theriault at 2018-07-23T13:52:56-07:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). - - - - - 8cf4e6b5 by Ryan Scott at 2018-07-27T11:28:03-04:00 eqTyCon_RDR now lives in TysWiredIn After GHC commit http://git.haskell.org/ghc.git/commit/f265008fb6f70830e7e92ce563f6d83833cef071 - - - - - 1ad251a6 by Alan Zimmerman at 2018-07-30T13:28:09-04:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 (cherry picked from commit e3926b50ab8a7269fd6904b06e881745f08bc5d6) - - - - - 8aea2492 by Richard Eisenberg at 2018-08-02T10:54:17-04:00 Update against new HsImplicitBndrs - - - - - e42cada9 by Alec Theriault at 2018-08-04T17:51:30+02:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 0e852512 by Alex Biehl at 2018-08-06T13:04:02+02:00 Make --package-version optional for --hoogle generation (#899) * Make --package-version optional for --hoogle generation * Import mkVersion * It's makeVersion not mkVersion - - - - - d2abd684 by Noel Bourke at 2018-08-21T09:34:18+02:00 Remove unnecessary backslashes from docs (#908) On https://haskell-haddock.readthedocs.io/en/latest/markup.html#special-characters the backslash and backtick special characters showed up with an extra backslash before them – I think the escaping is not (or no longer) needed for those characters in rst. - - - - - 7a578a9e by Matthew Pickering at 2018-08-21T09:34:50+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - aa3d4db3 by Matthew Pickering at 2018-08-21T09:37:34+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - ede91744 by Alec Theriault at 2018-08-21T09:42:52+02:00 Better test output when Haddock crashes on a test (#902) In particular: we report the tests that crashed seperately from the tests that produced incorrect output. In order for tests to pass (and exit 0), they must not crash and must produce the right output. - - - - - 4a872b84 by Guillaume Bouchard at 2018-08-21T09:45:57+02:00 Fix a typo (#878) - - - - - 4dbf7595 by Ben Sklaroff at 2018-08-21T12:04:09-04:00 Add ITcomment_line_prag token to Hyperlinker Parser This token is necessary for parsing #line pragmas inside nested comments. Reviewers: bgamari Reviewed By: bgamari Differential Revision: https://phabricator.haskell.org/D4935 - - - - - 9170b2a9 by Ben Gamari at 2018-08-21T17:55:15-04:00 Merge pull request haskell/haddock#893 from harpocrates/get-name-to-instances Accumulate explicitly which modules to load for 'attachInstances' - - - - - d57b57cc by Ben Gamari at 2018-08-21T17:59:13-04:00 Merge branch 'ghc-head' of github.com:haskell/haddock into ghc-head - - - - - 14601ca2 by Alec Theriault at 2018-08-21T19:09:37-04:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). (cherry picked from commit 5c7c596c51d69b92164e9ba920157b36ce2b2ec1) - - - - - 438c645e by Matthew Pickering at 2018-08-21T19:12:39-04:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 (cherry picked from commit e6aa8fb47b9477cc5ef5e46097524fe83e080f6d) - - - - - a80c5161 by Alec Theriault at 2018-08-21T22:06:40-07:00 Better rendering of unboxed sums/tuples * adds space after/before the '#' marks * properly reify 'HsSumTy' in 'synifyType' - - - - - 88456cc1 by Alec Theriault at 2018-08-21T22:06:40-07:00 Handle promoted tuples in 'synifyType' When we have a fully applied promoted tuple, we can expand it out properly. - - - - - fd1c1094 by Alec Theriault at 2018-08-21T22:19:34-07:00 Accept test cases - - - - - 6e80d9e0 by Alec Theriault at 2018-08-21T22:24:03-07:00 Merge pull request haskell/haddock#914 from harpocrates/feature/unboxed-stuff Better rendering of unboxed sums, unboxed tuples, promoted tuples. - - - - - 181a23f1 by Ben Gamari at 2018-08-23T15:53:48-04:00 Merge remote-tracking branch 'origin/ghc-8.6' into ghc-8.6 - - - - - 3a18c1d8 by Alec Theriault at 2018-08-27T14:15:25-07:00 Properly synify promoted list types We reconstruct promoted list literals whenever possible. That means that 'synifyType' produces '[Int, Bool, ()] instead of (Int ': (() ': (Bool ': ([] :: [Type])))) - - - - - b4794946 by Alec Theriault at 2018-09-03T07:19:55-07:00 Only look at visible types when synifying a 'HsListTy' The other types are still looked at when considering whether to make a kind signature or not. - - - - - a231fce2 by Alec Theriault at 2018-09-03T07:38:10-07:00 Merge pull request haskell/haddock#922 from harpocrates/promoted-lists Properly synify promoted list types - - - - - 0fdf044e by Ningning Xie at 2018-09-15T10:25:58-04:00 Update according to GHC Core changes - - - - - 7379b115 by Ningning Xie at 2018-09-15T15:40:18-04:00 update dataFullSig to work with Co Quantification This should have been in the previous patch, but wasn't. - - - - - cf84a046 by Alec Theriault at 2018-09-17T20:12:18-07:00 Fix/add to various docs * Add documentation for a bunch of previously undocumented options (fixes haskell/haddock#870) * Extend the documentation of `--hoogle` considerably (see haskell/haddock#807) * Describe how to add docs to `deriving` clauses (fixes haskell/haddock#912) * Fix inaccurate docs about hyperlinking infix identifiers (fixes haskell/haddock#780) - - - - - ae017935 by Alec Theriault at 2018-09-22T08:32:16-07:00 Update Travis - - - - - d95ae753 by Alec Theriault at 2018-09-22T09:34:10-07:00 Accept failing tests Also silence orphan warnings. - - - - - f3e67024 by Alec Theriault at 2018-09-22T09:41:23-07:00 Bump haddock-api-2.21.0, haddock-library-1.7.0 * Update CHANGELOGS * Update new versions in Cabal files * Purge references to ghc-8.4/master branches in README - - - - - 3f136d4a by Alec Theriault at 2018-09-22T10:53:31-07:00 Turn haddock-library into a minor release Fix some version bounds in haddock-library too. - - - - - b9def006 by Alec Theriault at 2018-09-22T13:07:35-07:00 keep cabal.project file - - - - - 4909aca7 by Alec Theriault at 2018-10-16T09:36:30-07:00 Build on 7.4 and 7.8 - - - - - 99d20a28 by Herbert Valerio Riedel at 2018-10-16T18:45:52+02:00 Minor tweak to package description - - - - - a8059618 by Herbert Valerio Riedel at 2018-10-16T18:47:24+02:00 Merge pull request haskell/haddock#945 haddock-api 2.21.0 and haddock-library 1.6.1 release - - - - - 2d9bdfc1 by Alec Theriault at 2018-10-16T10:54:21-07:00 Bump haddock-library to 1.7.0 The 1.6.1 release should've been a major bump, since types in the `Documentation.Haddock.Parser.Monad` module changed. This version makes that module internal (as it morally should be). - - - - - ed340cef by Alec Theriault at 2018-10-16T14:59:13-07:00 Merge branch 'ghc-8.4' into ghc-8.6 - - - - - 2821a8df by Alec Theriault at 2018-10-16T15:14:48-07:00 Merge branch 'ghc-8.6' into ghc-head - - - - - a722dc84 by Alec Theriault at 2018-10-16T16:28:55-07:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 63377496 by Alec Theriault at 2018-10-16T16:39:07-07:00 Update changelog - - - - - 099a0110 by Alec Theriault at 2018-10-16T16:49:28-07:00 Merge pull request haskell/haddock#942 from harpocrates/update-docs Fix & add to documentation - - - - - 0927416f by Alec Theriault at 2018-10-16T16:50:14-07:00 Set UTF-8 encoding before writing files (#934) This should fix haskell/haddock#929, as well as guard against future problems of this sort in other places. Basically replaces 'writeFile' (which selects the users default locale) with 'writeUtf8File' (which always uses utf8). - - - - - 83b7b017 by Alec Theriault at 2018-10-16T17:42:05-07:00 Output pattern synonyms in Hoogle backend (#947) * Output pattern synonyms in Hoogle backend We were previously weren't outputting _any_ pattern synonyms, bundled or not. Now, we output both. Fixes haskell/haddock#946. * Update changelog - - - - - 81e5033d by Alec Theriault at 2018-10-16T18:04:40-07:00 Release `haddock{,-api}-2.22.0` This version will accompany ghc-8.6.2 - - - - - 9661744e by Alex Biehl at 2018-10-18T08:14:32-07:00 Add NewOcean theme And make it the default theme. - - - - - 7ae6d722 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Improve appearance and readability These changes include: - use latest Haskell's logo colors - decrease #content width to improve readability - use nicer font - improve sizes and distances - - - - - 37f8703d by NunoAlexandre at 2018-10-18T08:14:32-07:00 Include custom font in the html head - - - - - 1d5e1d79 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Update html test reference files - - - - - 53b7651f by NunoAlexandre at 2018-10-18T08:14:32-07:00 Make it responsive - It makes small screens taking more space than larger ones - fixes a few issues present in small screens currently - make it look good across different screen sizes. - - - - - 6aa1aeb1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make the style consistent with hackage Several things are addressed here: - better responsive behaviour on the header - better space usage - consistent colors overall - other nit PR comments - - - - - 3a250c5c by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Place the package name before the menu links This supports the expected responsive menu design, where the package name appears above the menu links. - - - - - cae699b3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update html-test reference files The package name element in the package-header is now a div instead of a paragraph, and it is now above the menu ul.links instead of below. - - - - - 2ec7fd2d by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve synopsis style and code - Use CSS3 instead of loading pictures to show "+" and "-" symbols - Drop redundant code - - - - - 0c874c01 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Decrease space between code blocks There was too much space between code blocks as pointed out by reviewers. - - - - - 85568ce2 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Add an initial-scale property to all haddock pages This solves an issue reported about the content looking incredibly small on mobile devices. - - - - - c1538926 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Address several PR comments - Darken text color like done for hackage - Move synopsis to left side - Make table of contents stick to the left on wide screens - Wrap links to avoid page overflow - Improve expand/collapse buttons - Fix issue with content size on mobile devices - Fix issue with font-size on landscape mode - Increase width of the content - Change colors of table of contents and synopsis - Etc - - - - - e6639e5f by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make synopsis scrollable on wide screens When the synopsis is longer than the screen, you can’t see its end and you can't scroll down either, making the content unreachable. - - - - - 1f0591ff by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve information density - Reduce font size - Improve space between and within code blocks - Improve alignments - Improve spacing within sub-blocks - - - - - bf083097 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Minor adjustments Bring in some adjustments made to hackage: - link colors - page header show everything when package title is too long - - - - - 10375fc7 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix responsive triggers overlap issue The min and max width triggers have the same values, which caused the style resolution to take an intersection of both style declarations when the screen resolution had the size of the limts (say 1280px), causing an odd behaviour and look. - - - - - 95ff2f95 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix issue with menu alignment on firefox Reported and described here: https://github.com/haskell/haddock/pull/721#issuecomment-374668869 - - - - - dc86587e by Alex Biehl at 2018-10-18T08:14:32-07:00 Changelog entry for NewOcean - - - - - 27195e47 by Herbert Valerio Riedel at 2018-10-18T08:14:32-07:00 html-test --accept - - - - - 83f4f9c0 by Alex Biehl at 2018-10-18T08:14:32-07:00 Avoid name shadowing - - - - - 231487f1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font to PT Sans Also migrate some general text related changes from hackage. - - - - - 313db81a by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Use 'flex' to fix header alignment - - - - - 5087367b by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Misc of tweaks - Update link colors to hackage scheme - Tune spacing between content elements - Update footer style - Fix and improve code blocks identation - - - - - b08020df by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font in Xhtml.hs to PT Sans - - - - - 78ce06e3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve code blocks styling - Fix and improve spacing - Improve colors and borders - - - - - 81262d20 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make package-header caption backward-compatible The current html generator of this branch wraps the package-header caption as a div, which does not work (without style adjustments) with the old themes. Changing it from div to span does the trick, without needing to adjust the old stylesheets. - - - - - dc4475cb by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update test-suite reference html pages - - - - - 393d35d8 by Alec Theriault at 2018-10-18T08:25:36-07:00 Accept tests - - - - - a94484ba by Alec Theriault at 2018-10-21T10:29:29-07:00 Fix CHANGELOG - - - - - 8797eca3 by Alec Theriault at 2018-10-21T10:36:19-07:00 Update 'data-files' to include NewOcean stuff - - - - - 1ae51e4a by Simon Jakobi at 2018-10-23T11:29:14+02:00 Fix typo in a warning - - - - - 009ad8e8 by Alec Theriault at 2018-10-24T12:47:47-07:00 Update JS dependencies This was done via `npm audit fix`. I think this fixes haskell/haddock#903 along with some more serious vulnerabilities that nobody seems to have noticed. - - - - - 051994db by Alec Theriault at 2018-10-24T17:31:09-07:00 Resurrect the style-switcher This fixes haskell/haddock#810. Looks like things were broken during the quickjump refactor of the JS. For the (git) record: I do not think the style switcher is a good idea. I'm fixing it for the same reason @mzero added it; as an answer to "rumblings from some that they didn't want their pixels changed on bit" - - - - - 2a1d620f by Alec Theriault at 2018-10-24T17:38:07-07:00 Fix copy-pasta error in data-files - - - - - ed5bfb7f by Alec Theriault at 2018-10-24T20:42:14-07:00 Fix the synopsis button Here's these changes are supposed to do: * put the synopsis back on the right side * properly have it on the edge of the screen on wide screens * adjust the background of the synopsis to match the button (otherwise the grey blends in with what is underneath) * get rid of the dotted purple line * the synopsis contents are now scrollable even when in wide screens (this has been a long-standing bug) - - - - - 883fd74b by Alec Theriault at 2018-10-25T20:16:46-07:00 Avoid more conflicts in generated ids (#954) This fixes haskell/haddock#953 by passing more names into the generated ids. - - - - - ea54e331 by Alec Theriault at 2018-10-25T21:07:12-07:00 Don't hide bullets in method docs I think thst CSS was meant only to deal with fields and the effect on bullets was accidental. Fixes haskell/haddock#926. - - - - - 9a14ef4a by Alec Theriault at 2018-10-25T22:02:07-07:00 Indent more things + slightly smaller font - - - - - b9f17e29 by Alec Theriault at 2018-10-25T22:10:01-07:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 096a3cfa by Alec Theriault at 2018-10-25T22:24:38-07:00 Accept HTML output - - - - - 2669517d by Alec Theriault at 2018-10-26T09:02:35-07:00 User manual + stuff for building GHC docs - - - - - 46b27687 by Alec Theriault at 2018-10-26T09:10:59-07:00 Make 'Contents' in NewOcean scrollable This only happens if the contents block on the left is so big that it doesn't fit (vertically) on the page. If that happens, we want it to be scrollable. - - - - - 3443dd94 by Alec Theriault at 2018-10-26T09:36:46-07:00 Revert "Make 'Contents' in NewOcean scrollable" This reverts commit f909ffd8353d6463fd5dd184998a32aa98d5c922. I missed the fact this also forces the 'Contents' to always go down to the bottom of the page. - - - - - ed081424 by Alec Theriault at 2018-10-26T14:22:23-07:00 Avoid some partiality AFAICT this wasn't causing any crashes, but that's mostly because we happen not to be forcing `pkgStr` when it would diverge. We come dangerously close to doing that in `ppHtmlIndex`. Fixes haskell/haddock#569. - - - - - 6a5bec41 by Alec Theriault at 2018-10-27T10:05:04-07:00 Fix documentation in `haddock-api` (#957) * Fix misplaced Haddocks in Haddock itself Haddock should be able to generate documentation for 'haddock-api' again. * Make CI check that documentation can be built. * Add back a doc that is OK - - - - - 5100450a by Matthew Yacavone at 2018-10-27T14:51:38-04:00 More explicit foralls (GHC Proposal 0007) - - - - - 8771a6b0 by Alec Theriault at 2018-11-05T13:58:11-08:00 Only run MathJax on entities with "mathjax" class (#960) Correspondingly, we wrap all inline/diplay math in <span class="mathjax"> ... the math .... </span> This fixes haskell/haddock#959. - - - - - bd7ff5c5 by Alec Theriault at 2018-11-05T15:54:22-08:00 Deduplicate some work in 'AttachInstances' Perf only change: * avoid needlessly union-ing maps * avoid synify-ing instances twice Took this opportunity to add some docs too - - - - - cf99fd8f by Alec Theriault at 2018-11-05T15:54:22-08:00 Specialize some SYB functions Perf only change: * Add a 'SPECIALIZE' pragma to help GHC optimize a 'Data a =>' constraint * Manually specialize the needlessly general type of 'specializeTyVarBndrs' - - - - - 4f91c473 by Alec Theriault at 2018-11-05T15:54:22-08:00 Improve perf of renaming Perf only change: * don't look up type variable names (they're never in the environment) * use a difference list for accumulating missing names * more efficient 'Functor'/'Applicative' instances for 'RnM' - - - - - 4bbab0d4 by Alec Theriault at 2018-11-05T15:54:22-08:00 Faster 'Text' driven parser combinators Perf only change: * use 'getParserState'/'setParserState' to make 'Text'-optimized parser combinators * minimize uses of 'Data.Text.{pack,unpack,cons,snoc}' - - - - - fa430c02 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support hyperlink labels with inline markup The parser for pictures hasn't been properly adjusted yet. - - - - - c1431035 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support (and flatten) inline markup in image links Inline markup is supported in image links but, as per the [commonmark recommendation][0], it is stripped back to a plain text representation. [0]: https://spec.commonmark.org/0.28/#example-547 - - - - - d4ee1ba5 by Alec Theriault at 2018-11-06T12:03:24-08:00 Accept test case - - - - - 8088aeb1 by Alec Theriault at 2018-11-06T12:03:24-08:00 Fix/add to haddock-library test suite - - - - - e78f644d by Alec Theriault at 2018-11-06T13:26:31-08:00 Bump version bounds - - - - - 644335eb by Alec Theriault at 2018-11-06T13:53:30-08:00 Merge pull request haskell/haddock#875 from harpocrates/feature/markup-in-hyperlinks Inline markup in markdown-style links and images - - - - - e173ed0d by Alec Theriault at 2018-11-07T12:37:18-08:00 Fix issues around plus/minus * swap the minimize unicode to something more intuitive * use new unicode expander/collapser for instance lists * address some alignment issues in the "index" page - - - - - b2d92df7 by Alec Theriault at 2018-11-07T13:41:57-08:00 Allow "Contents" summary to scroll in a fixed div In the unfortunate event that the "Contents" summary doesn't fit vertically (like in the "Prelude"), it will be scrollable. - - - - - ca704c23 by Alec Theriault at 2018-11-07T13:45:15-08:00 Accept HTML output changes - - - - - 82c0ec6d by Alec Theriault at 2018-11-07T18:12:54-08:00 overflow-y 'scroll' -> 'auto' - - - - - 571d7657 by Alec Theriault at 2018-11-08T19:44:12-08:00 Clicking on "Contents" navigates to top of page - - - - - 8065a012 by Alec Theriault at 2018-11-08T19:44:17-08:00 Space out functions more Also, functions and data decls now have the same space before and after them. - - - - - cc650ede by Alec Theriault at 2018-11-09T08:13:35-08:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 65f8c17f by Alec Theriault at 2018-11-10T14:04:06-08:00 Update changelog - - - - - 20473847 by Alec Theriault at 2018-11-10T14:21:40-08:00 Replace oplus/ominus expander/collapser icons with triangles - - - - - 16592957 by Alec Theriault at 2018-11-10T14:35:10-08:00 Merge pull request haskell/haddock#949 from haskell/wip/new-ocean Introduce NewOcean theme. - - - - - 357cefe1 by Alec Theriault at 2018-11-10T16:02:13-08:00 Merge branch 'ghc-8.6' into ghc-head - - - - - de612267 by Alec Theriault at 2018-11-11T20:01:21-08:00 Rename 'NewOcean' theme to 'Linuwial' - - - - - 954b5baa by Alec Theriault at 2018-11-12T08:33:18-08:00 Add blockquote styling Matches b71da1feabf33efbbc517ac376bb690b5a604c2f from hackage-server. Fixes haskell/haddock#967. - - - - - d32c0b0b by Fangyi Zhou at 2018-11-12T10:24:13-08:00 Fix some broken links (#15733) Summary: For links in subpackages as well. https://phabricator.haskell.org/D5257 Test Plan: Manually verify links Reviewers: mpickering, bgamari, osa1 Reviewed By: osa1 GHC Trac Issues: haskell/haddock#15733 Differential Revision: https://phabricator.haskell.org/D5262 - - - - - 41098b1f by Alp Mestanogullari at 2018-11-15T22:40:09+01:00 Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change It got introduced in ghc/ghc at ae2c9b40f5b6bf272251d1f4107c60003f541b62. - - - - - c5c1c7e0 by Alec Theriault at 2018-11-15T13:48:13-08:00 Merge pull request haskell/haddock#970 from alpmestan/alp/fix-promotionflag Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change - - - - - 6473d3a4 by Shayan-Najd at 2018-11-23T01:38:49+01:00 [TTG: Handling Source Locations] Foundation and Pat Trac Issues haskell/haddock#15495 This patch removes the ping-pong style from HsPat (only, for now), using the plan laid out at https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow/HandlingSourceLocations (solution A). - the class `HasSrcSpan`, and its functions (e.g., `cL` and `dL`), are introduced - some instances of `HasSrcSpan` are introduced - some constructors `L` are replaced with `cL` - some patterns `L` are replaced with `dL->L` view pattern - some type annotation are necessarily updated (e.g., `Pat p` --> `Pat (GhcPass p)`) - - - - - 7a088dfe by Alec Theriault at 2018-11-26T11:11:28-08:00 More uniform handling of `forall`'s in HTML/LaTeX * don't forget to print explicit `forall`'s when there are arg docs * when printing an explicit `forall`, print all tyvars Fixes haskell/haddock#973 - - - - - d735e570 by Alec Theriault at 2018-12-12T08:42:09-08:00 Fix warnings, accept output * remove redundant imports (only brought to light due to recent work for improving redundant import detection) * fix a bug that was casuing exports to appear in reverse order * fix something in haddock-library that prevented compilation on old GHC's - - - - - a3852f8a by Zejun Wu at 2018-12-14T09:37:47-05:00 Output better debug infromation on internal error in extractDecl This will make investigation of haskell/haddock#979 easier - - - - - 2eccb5b9 by Alec Theriault at 2018-12-17T09:25:10-05:00 Refactor names + unused functions (#982) This commit should not introduce any change in functionality! * consistently use `getOccString` to convert `Name`s to strings * compare names directly when possible (instead of comparing strings) * get rid of unused utility functions - - - - - e82e4df8 by Alec Theriault at 2018-12-20T16:16:30-05:00 Load plugins when compiling each module (#983) * WIP: Load (typechecker) plugins from language pragmas * Revert "Load plugins when starting a GHC session (#905)" This reverts commit 72d82e52f2a6225686d9668790ac33c1d1743193. * Simplify plugin initialization code - - - - - 96e86f38 by Alec Theriault at 2018-12-23T10:23:20-05:00 Properly synify and render promoted type variables (#985) * Synify and render properly promoted type variables Fixes haskell/haddock#923. * Accept output - - - - - 23343345 by Alec Theriault at 2018-12-27T16:39:38-05:00 Remove `haddock-test`'s dep. on `syb` (#987) The functionality is easily inlined into one short function: `gmapEverywhere`. This doesn't warrant pulling in another package. - - - - - d0734f21 by Alec Theriault at 2018-12-27T16:39:52-05:00 Address deprecation warnings in `haddock-test` (#988) Fixes haskell/haddock#885. - - - - - 4d9f144e by mynguyen at 2018-12-30T23:42:26-05:00 Visible kind application haddock update - - - - - ffe0e9ed by Alec Theriault at 2019-01-07T13:55:22-08:00 Print kinded tyvars in constructors for Hoogle (#993) Fixes haskell/haddock#992 - - - - - 2e18b55d by Alec Theriault at 2019-01-10T16:42:45-08:00 Accept new output `GHC.Maybe` -> `Data.Maybe` (#996) Since 53874834b779ad0dfbcde6650069c37926da1b79 in GHC, "GHC.Maybe" is marked as `not-home`. That changes around some test output. - - - - - 055da666 by Gabor Greif at 2019-01-22T14:41:51+01:00 Lone typofix - - - - - 01bb71c9 by Alec Theriault at 2019-01-23T11:46:46-08:00 Keep forall on H98 existential data constructors (#1003) The information about whether or not there is a source-level `forall` is already available on a `ConDecl` (as `con_forall`), so we should use it instead of always assuming `False`! Fixes haskell/haddock#1002. - - - - - f9b9bc0e by Ryan Scott at 2019-01-27T09:28:12-08:00 Fix haskell/haddock#1004 with a pinch of dropForAlls - - - - - 5cfcdd0a by Alec Theriault at 2019-01-28T16:49:57-08:00 Loosen 'QuickCheck' and 'hspec' bounds It looks like the new versions don't cause any breakage and loosening the bounds helps deps fit in one stack resolver. - - - - - 3545d3dd by Alec Theriault at 2019-01-31T01:37:25-08:00 Use `.hie` files for the Hyperlinker backend (#977) # Summary This is a large architectural change to the Hyperlinker. * extract link (and now also type) information from `.hie` instead of doing ad-hoc SYB traversals of the `RenamedSource`. Also adds a superb type-on-hover feature (#715). * re-engineer the lexer to avoid needless string conversions. By going directly through GHC's `P` monad and taking bytestring slices, we avoid a ton of allocation and have better handling of position pragmas and CPP. In terms of performance, the Haddock side of things has gotten _much_ more efficient. Unfortunately, much of this is cancelled out by the increased GHC workload for generating `.hie` files. For the full set of boot libs (including `ghc`-the-library) * the sum of total time went down by 9-10% overall * the sum of total allocations went down by 6-7% # Motivation Haddock is moving towards working entirely over `.hi` and `.hie` files. This change means we no longer need the `RenamedSource` from `TypecheckedModule` (something which is _not_ in `.hi` files). # Details Along the way a bunch of things were fixed: * Cross package (and other) links are now more reliable (#496) * The lexer tries to recover from errors on every line (instead of at CPP boundaries) * `LINE`/`COLUMN` pragmas are taken into account * filter out zero length tokens before rendering * avoid recomputing the `ModuleName`-based `SrcMap` * remove the last use of `Documentation.Haddock.Utf8` (see haskell/haddock#998) * restructure temporary folder logic for `.hi`/`.hie` model - - - - - 2ded3359 by Herbert Valerio Riedel at 2019-02-02T12:06:12+01:00 Update/modernise haddock-library.cabal file - - - - - 62b93451 by Herbert Valerio Riedel at 2019-02-02T12:19:31+01:00 Tentatively declare support for unreleased base-4.13/ghc-8.8 - - - - - 6041e767 by Herbert Valerio Riedel at 2019-02-02T16:04:32+01:00 Normalise LICENSE text w/ cabal's BSD2 template Also, correct the `.cabal` files to advertise `BSD2` instead of the incorrect `BSD3` license. - - - - - 0b459d7f by Alec Theriault at 2019-02-02T18:06:12-08:00 CI: fetch GHC from validate artifact Should help make CI be less broken - - - - - 6b5c07cf by Alec Theriault at 2019-02-02T18:06:12-08:00 Fix some Hyperlinker test suite fallout * Amend `ParserSpec` to match new Hyperlinker API - pass in compiler info - strip out null tokens * Make `hypsrc-test` pass reliably - strip out `local-*` ids - strip out `line-*` ids from the `ClangCppBug` test - re-accept output - - - - - ded34791 by Nathan Collins at 2019-02-02T18:31:23-08:00 Update README instructions for Stack No need to `stack install` Haddock to test it. Indeed, `stack install` changes the `haddock` on user's `PATH` if `~/.local/bin` is on user's `PATH` which may not be desirable when hacking on Haddock. - - - - - 723298c9 by Alec Theriault at 2019-02-03T09:11:05-08:00 Remove `Documentation.Haddock.Utf8` The circumstances under which this module appeared are completely gone. The Hyperlinker backend no longer needs this module (it uses the more efficient `Encoding` module from `ghc`). Why no deprecation? Because this module really shouldn't exist! - It isn't used in `haddock-library`/`haddock-api` anymore - It was copy pasted directly from `utf8-string` - Folks seeking a boot-lib only solution can use `ghc`'s `Encoding` - - - - - 51050006 by Alec Theriault at 2019-02-03T22:58:58-08:00 Miscellaneous improvements to `Convert` (#1020) Now that Haddock is moving towards working entirely over `.hi` and `.hie` files, all declarations and types are going to be synthesized via the `Convert` module. In preparation for this change, here are a bunch of fixes to this module: * Add kind annotations to type variables in `forall`'s whose kind is not `Type`, unless the kind can be inferred from some later use of the variable. See `implicitForAll` and `noKindTyVars` in particular if you wish to dive into this. * Properly detect `HsQualTy` in `synifyType`. This is done by following suit with what GHC's `toIfaceTypeX` does and checking the first argument of `FunTy{} :: Type` to see if it classified as a given/wanted in the typechecker (see `isPredTy`). * Beef up the logic around figuring out when an explicit `forall` is needed. This includes: observing if any of the type variables will need kind signatures, if the inferred type variable order _without_ a forall will still match the one GHC claims, and some other small things. * Add some (not yet used) functionality for default levity polymorphic type signatures. This functionality similar to `fprint-explicit-runtime-reps`. Couple other smaller fixes only worth mentioning: * Show the family result signature only when it isn't `Type` * Fix rendering of implicit parameters in the LaTeX and Hoogle backends * Better handling of the return kind of polykinded H98 data declarations * Class decls produced by `tyThingToLHsDecl` now contain associated type defaults and default method signatures when appropriate * Filter out more `forall`'s in pattern synonyms - - - - - 841980c4 by Oleg Grenrus at 2019-02-04T08:44:25-08:00 Make a fixture of weird parsing of lists (#997) The second example is interesting. If there's a list directly after the header, and that list has deeper structure, the parser is confused: It finds two lists: - One with the first nested element, - everything after it I'm not trying to fix this, as I'm not even sure this is a bug, and not a feature. - - - - - 7315c0c8 by Ryan Scott at 2019-02-04T12:17:56-08:00 Fix haskell/haddock#1015 with dataConUserTyVars (#1022) The central trick in this patch is to use `dataConUserTyVars` instead of `univ_tvs ++ ex_tvs`, which displays the foralls in a GADT constructor in a way that's more faithful to how the user originally wrote it. Fixes haskell/haddock#1015. - - - - - ee0b49a3 by Ryan Scott at 2019-02-04T15:25:17-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. - - - - - 1c850dc8 by Alan Zimmerman at 2019-02-05T21:54:18+02:00 Matching changes in GHC for haskell/haddock#16236 - - - - - ab03c38e by Simon Marlow at 2019-02-06T08:07:33+00:00 Merge pull request haskell/haddock#1014 from hvr/pr/bsd2-normalise Normalise LICENSE text w/ cabal's BSD2 template - - - - - 5a92ccae by Alec Theriault at 2019-02-10T06:21:55-05:00 Merge remote-tracking branch 'gitlab/wip/T16236-2' into ghc-head - - - - - c0485a1d by Alec Theriault at 2019-02-10T03:32:52-08:00 Removes `haddock-test`s dependency on `xml`/`xhtml` (#1027) This means that `html-test`, `latex-test`, `hoogle-test`, and `hypsrc-test` now only depend on GHC boot libs. So we should now be able to build and run these as part of GHC's testsuite. \o/ The reference output has changed very slightly, in three ways: * we don't convert quotes back into `&quot;` as the `xml` lib did * we don't add extra `&nbsp;` as the `xml` lib did * we now remove the entire footer `div` (instead of just emptying it) - - - - - 65a448e3 by Alec Theriault at 2019-02-11T12:27:41-05:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - 360ca937 by Alec Theriault at 2019-02-13T11:36:11-05:00 Clean up logic for guessing `-B` and `--lib` (#1026) Haddock built with the `in-ghc-tree` flag tries harder to find the GHC lib folder and its own resources. This should make it possible to use `in-ghc-tree`-built Haddock without having to specify the `-B` and `--lib` options (just how you can use in-tree GHC without always specifying the `-B` option). The logic to do this relies on `getExecutablePath`, so we only get this auto-detection on platforms where this function works. - - - - - d583e364 by Alec Theriault at 2019-02-16T10:41:22-05:00 Fix tests broken by GHC Changes in 19626218566ea709b5f6f287d3c296b0c4021de2 affected some of the hyperlinker output. Accepted the new output (hovering over a `..` now shows you what that wildcard binds). Also fixed some stray deprecation warnings. - - - - - da0c42cc by Vladislav Zavialov at 2019-02-17T11:39:19+03:00 Parser changes to match !380 - - - - - ab96bed7 by Ryan Scott at 2019-02-18T04:44:08-05:00 Bump ghc version to 8.9 - - - - - 44b7c714 by Alec Theriault at 2019-02-22T05:49:43-08:00 Match GHC changes for T16185 `FunTy` now has an `AnonArgFlag` that indicates whether the arrow is a `t1 => t2` or `t1 -> t2`. This commit shouldn't change any functionality in Haddock. - - - - - 2ee653b1 by Alec Theriault at 2019-02-24T18:53:33-08:00 Update .travis.yml Points to the new GHC CI artifact. - - - - - 90939d71 by Alec Theriault at 2019-02-25T00:42:41-08:00 Support value/type namespaces on identifier links Identifier links can be prefixed with a 'v' or 't' to indicate the value or type namespace of the desired identifier. For example: -- | Some link to a value: v'Data.Functor.Identity' -- -- Some link to a type: t'Data.Functor.Identity' The default is still the type (with a warning about the ambiguity) - - - - - d6ed496c by Alec Theriault at 2019-02-25T00:42:46-08:00 Better identifier parsing * '(<|>)' and '`elem`' now get parsed and rendered properly as links * 'DbModule'/'DbUnitId' now properly get split apart into two links * tuple names now get parsed properly * some more small niceties... The identifier parsing code is more precise and more efficient (although to be fair: it is also longer and in its own module). On the rendering side, we need to pipe through information about backticks/parens/neither all the way through from renaming to the backends. In terms of impact: a total of 35 modules in the entirety of the bootlib + ghc lib docs change. The only "regression" is things like '\0'. These should be changed to @\\0@ (the path by which this previously worked seems accidental). - - - - - 3c3b404c by Alec Theriault at 2019-02-25T22:12:11-08:00 Fix standalone deriving docs Docs on standalone deriving decls for classes with associated types should be associated with the class instance, not the associated type instance. Fixes haskell/haddock#1033 - - - - - d51ef69e by Alec Theriault at 2019-02-26T19:14:59-08:00 Fix bogus identifier defaulting This avoids a situation in which an identifier would get defaulted to a completely different identifier. Prior to this commit, the 'Bug1035' test case would hyperlink 'Foo' into 'Bar'! Fixes haskell/haddock#1035. - - - - - 88cbbdc7 by Ryan Scott at 2019-02-27T10:14:03-05:00 Visible dependent quantification (#16326) changes - - - - - 0dcf6cee by Xia Li-yao at 2019-02-27T21:53:27-05:00 Menu item controlling which instances are expanded/collapsed (#1007) Adds a menu item (like "Quick Jump") for options related to displaying instances. This provides functionality for: * expanding/collapsing all instances on the currently opened page * controlling whether instances are expanded/collapsed by default * controlling whether the state of instances should be "remembered" This new functionality is implemented in Typescript in `details-helper`. The built-in-themes style switcher also got a revamp so that all three of QuickJump, the style switcher, and instance preferences now have the same style and implementation structure. See also: https://mail.haskell.org/pipermail/haskell-cafe/2019-January/130495.html Fixes haskell/haddock#698. Co-authored-by: Lysxia <lysxia at gmail.com> Co-authored-by: Nathan Collins <conathan at galois.com> - - - - - 3828c0fb by Alec Theriault at 2019-02-28T12:42:49-05:00 `--show-interface` should output to stdout. (#1040) Fixes haskell/haddock#864. - - - - - a50f4cda by gbaz at 2019-03-01T07:43:16-08:00 Increase contrast of Linuwal theme (#1037) This is to address the concern that, on less nice and older screens, some of the shades of grey blend in too easily with the white background. * darken the font slightly * darken slightly the grey behind type signatures and such * add a border and round the corners on code blocks * knock the font down by one point - - - - - ab4d41de by Alec Theriault at 2019-03-03T09:23:26-08:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 12f509eb by Ben Gamari at 2019-03-04T22:13:20-05:00 Remove reference to Opt_SplitObjs flag Split-objects has been removed. - - - - - 5b3e4c9a by Ryan Scott at 2019-03-06T19:16:24-05:00 Update html-test output to reflect haskell/haddock#16391 changes - - - - - fc228af1 by Alec Theriault at 2019-03-09T08:29:23-08:00 Match changes for "Stop inferring over-polymorphic kinds" The `hsq_ext` field of `HsQTvs` is now just the implicit variables (instead of also including information about which of these variables are dependent). This commit shouldn't change any functionality in Haddock. - - - - - 6ac109eb by Alec Theriault at 2019-03-09T11:22:55-08:00 Add .hi, .dyn_hi, etc files to .gitignore Fixes haskell/haddock#1030. - - - - - b55f0c05 by Alec Theriault at 2019-03-09T11:22:55-08:00 Better support for default methods in classes * default methods now get rendered differently * default associated types get rendered * fix a forgotten `s/TypeSig/ClassOpSig/` refactor in LaTeX backend * LaTeX backend now renders default method signatures NB: there is still no way to document default class members and the NB: LaTeX backend still crashes on associated types - - - - - 10aea0cf by Alec Theriault at 2019-03-09T11:22:55-08:00 Avoid multi-line `emph` in LaTeX backend `markupWarning` often processes inputs which span across paragraphs. Unfortunately, LaTeX's `emph` is not made to handle this (and will crash). Fixes haskell/haddock#936. - - - - - d22dc2c9 by Alec Theriault at 2019-03-09T11:22:55-08:00 Many LaTeX backend fixes After this commit, we can run with `--latex` on all boot libraries without crashing (although the generated LaTeX still fails to compile in a handful of larger packages like `ghc` and `base`). * Add newlines after all block elements in LaTeX. This is important to prevent the final output from being more an more indented. See the `latext-test/src/Example` test case for a sample of this. * Support associated types in class declarations (but not yet defaults) * Several small issues for producing compiling LaTeX; - avoid empy `\haddockbeginargs` lists (ex: `type family Any`) - properly escape identifiers depending on context (ex: `Int#`) - add `vbox` around `itemize`/`enumerate` (so they can be in tables) * Several spacing fixes: - limit the width of `Pretty`-arranged monospaced code - cut out extra space characters in export lists - only escape spaces if there are _multiple_ spaces - allow type signatures to be multiline (even without docs) * Remove uninteresting and repetitive `main.tex`/`haddock.sty` files from `latex-test` test reference output. Fixes haskell/haddock#935, haskell/haddock#929 (LaTeX docs for `text` build & compile) Fixes haskell/haddock#727, haskell/haddock#930 (I think both are really about type families...) - - - - - 0e6cee00 by Alec Theriault at 2019-03-29T12:11:56-07:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - ce05434d by Alan Zimmerman at 2019-03-29T12:12:11-07:00 Matching changes in GHC for haskell/haddock#16236 (cherry picked from commit 3ee6526d4ae7bf4deb7cd1caf24b3d7355573576) - - - - - d85766b2 by Ben Gamari at 2019-03-29T12:14:04-07:00 Bump GHC to 8.8 - - - - - 5a82cbaf by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Redo ParseModuleHeader - - - - - b9033348 by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - bb55c8f4 by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove outdated `.ghci` files and `scripts` The `.ghci` files are actively annoying when trying to `cabal v2-repl`. As for the `scripts`, the distribution workflow is completely different. - - - - - 5ee244dc by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove obsolete arcanist files + STYLE Now that GHC is hosted on Gitlab, the arcanist files don't make sense anymore. The STYLE file contains nothing more than a dead link too. - - - - - d07c1928 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Redo ParseModuleHeader - - - - - 492762d2 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - af2ac773 by Ryan Scott at 2019-05-14T17:22:13-04:00 Changes for haskell/haddock#16110/#16356 - - - - - 6820ed0d by Alec Theriault at 2019-05-17T08:51:27-07:00 Unbreak haskell/haddock#1004 test case `fail` is no longer part of `Monad`. - - - - - 6bf7be98 by Alec Theriault at 2019-05-17T08:51:27-07:00 Fix haskell/haddock#1063 with better parenthesization logic for contexts The only other change in html/hoogle/hyperlinker output for the boot libraries that this caused is a fix to some Hoogle output for implicit params. ``` $ diff -r _build/docs/ old_docs diff -r _build/docs/html/libraries/base/base.txt old_docs/html/libraries/base/base.txt 13296c13296 < assertError :: (?callStack :: CallStack) => Bool -> a -> a --- > assertError :: ?callStack :: CallStack => Bool -> a -> a ``` - - - - - b5716b61 by Ryan Scott at 2019-05-22T17:24:32-04:00 Match changes with haskell/haddock#14332 - - - - - c115abf6 by Alec Theriault at 2019-05-26T16:01:58-04:00 Remove Haddock's dependency on `Cabal` At this point, Haddock depended on Cabal-the-library solely for a verbosity parser (which misleadingly accepts all sorts of verbosity options that Haddock never uses). Now, the only dependency on Cabal is for `haddock-test` (which uses Cabal to locate the Haddock interface files of a couple boot libraries). - - - - - e5b2d4a3 by Alec Theriault at 2019-05-26T16:16:25-04:00 Regression test: promoted lists in associated types When possible, associated types with promoted lists should use the promoted list literal syntax (instead of repeated applications of ': and '[]). This was fixed in 2122de5473fd5b434af690ff9ccb1a2e58491f8c. Closes haskell/haddock#466, - - - - - cc5ad5d3 by Alec Theriault at 2019-05-26T17:55:54-04:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 4b3301a6 by Alec Theriault at 2019-05-26T17:57:52-04:00 Release haddock-2.23, haddock-library-1.8.0 Tentatively adjust bounds and changelogs for the release to be bundled with GHC 8.8.1. - - - - - 69c7cfce by Matthew Pickering at 2019-05-30T10:54:27+01:00 Update hyperlinker tests for new types in .hie files - - - - - 29b7e738 by Zubin Duggal at 2019-05-30T10:57:51+01:00 update for new way to store hiefile headers - - - - - aeca5d5f by Zubin Duggal at 2019-06-04T18:57:42-04:00 update for new way to store hiefile headers - - - - - ba2ca518 by Ben Gamari at 2019-06-07T23:11:14+00:00 Update test output for introduction of Safe-Inferred - - - - - 3a975a6c by Ryan Scott at 2019-07-03T12:06:27-04:00 Changes for haskell/haddock#15247 - - - - - 0df46555 by Zubin Duggal at 2019-07-22T10:52:50+01:00 Fix haddockHypsrcTest - - - - - 2688686b by Sylvain Henry at 2019-09-12T23:19:39+02:00 Fix for GHC module renaming - - - - - 9ec0f3fc by Alec Theriault at 2019-09-20T03:21:00-04:00 Fix Travis CI, loosen .cabal bounds (#1089) Tentatively for the 2.23 release: * updated Travis CI to work again * tweaked bounds in the `.cabal` files * adjusted `extra-source-files` to properly identify test files - - - - - ca559beb by Matthías Páll Gissurarson at 2019-09-28T12:14:40-04:00 Small change in to facilitate extended typed-holes (#1090) This change has no functional effect on haddock itself, it just changes one pattern to use `_ (` rather than `_(`, so that we may use `_(` as a token for extended typed-holes later. - - - - - 02e28976 by Vladislav Zavialov at 2019-09-28T12:17:45-04:00 Remove spaces around @-patterns (#1093) This is needed to compile `haddock` when [GHC Proposal haskell/haddock#229](https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0229-whitespace-bang-patterns.rst) is implemented. - - - - - 83cbbf55 by Alexis King at 2019-09-30T21:12:42-04:00 Fix the ignore-exports option (#1082) The `ignore-exports` option has been broken since haskell/haddock#688, as mentioned in https://github.com/haskell/haddock/pull/766#issue-172505043. This PR fixes it. - - - - - e127e0ab by Ben Gamari at 2019-10-06T15:12:06-04:00 Fix a few haddock issues - - - - - 3a0f5c89 by Zubin Duggal at 2019-10-07T17:56:13-04:00 Fix crash when there are no srcspans in the file due to CPP - - - - - 339c5ff8 by Alec Theriault at 2019-10-07T17:56:13-04:00 Prefer un-hyperlinked sources to no sources It is possible to fail to extract an HIE ast. This is however not a reason to produce _no_ output - we should still make a colorized HTML page. - - - - - d47ef478 by Alec Theriault at 2019-10-07T17:56:13-04:00 Add a regression test for haskell/haddock#1091 Previously, this input would crash Haddock. - - - - - ed7c8b0f by Alec Theriault at 2019-10-07T20:56:48-04:00 Add Hyperlinker test cases for TH-related stuff Hopefully this will guard against regressions around quasiquotes, TH quotes, and TH splices. - - - - - d00436ab by Andreas Klebinger at 2019-10-21T15:53:03+02:00 Refactor for withTiming changes. - - - - - 4230e712 by Ben Gamari at 2019-10-22T09:36:37-04:00 Merge pull request haskell/haddock#1101 from AndreasPK/withTimingRefactor Refactor for withTiming changes. - - - - - d155c5f4 by Ryan Scott at 2019-10-23T10:37:17-04:00 Reify oversaturated data family instances correctly (#1103) This fixes haskell/haddock#1103 by adapting the corresponding patch for GHC (see https://gitlab.haskell.org/ghc/ghc/issues/17296 and https://gitlab.haskell.org/ghc/ghc/merge_requests/1877). - - - - - 331a5adf by Sebastian Graf at 2019-10-25T17:14:40+02:00 Refactor for OutputableBndrId changes - - - - - 48a490e0 by Ben Gamari at 2019-10-27T10:16:16-04:00 Merge pull request haskell/haddock#1105 from sgraf812/wip/flexible-outputable Refactor for OutputableBndrId changes - - - - - f62a7dfc by Sebastian Graf at 2019-11-01T11:54:16+00:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - d9b242ed by Ryan Scott at 2019-11-03T13:20:03-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. (cherry picked from commit cfd682c5fd03b099a3d78c44f9279faf56a0ac70) - - - - - dfd42406 by Sebastian Graf at 2019-11-04T07:02:14-05:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - 0b15be7c by Ben Gamari at 2019-11-09T13:21:33-05:00 Import isRuntimeRepVar from Type rather than TyCoRep isRuntimeRepVar is not longer exported from TyCoRep due to ghc#17441. - - - - - 091f7283 by Ben Gamari at 2019-11-10T12:47:06-05:00 Bump to GHC 8.10 - - - - - e88c71f2 by Ben Gamari at 2019-11-14T00:22:24-05:00 Merge pull request haskell/haddock#1110 from haskell/wip/T17441 Import isRuntimeRepVar from Type rather than TyCoRep - - - - - 4e0bbc17 by Ben Gamari at 2019-11-14T00:22:45-05:00 Version bumps for GHC 8.11 - - - - - 0e85ceb4 by Ben Gamari at 2019-11-15T11:59:45-05:00 Bump to GHC 8.10 - - - - - 00d6d68b by Ben Gamari at 2019-11-16T18:35:58-05:00 Bump ghc version to 8.11 - - - - - dde1fc3f by Ben Gamari at 2019-11-16T20:40:37-05:00 Drop support for base 4.13 - - - - - f52e331d by Vladislav Zavialov at 2019-11-24T13:02:28+03:00 Update Hyperlinker.Parser.classify to use ITdollar - - - - - 1ad96198 by Vladislav Zavialov at 2019-11-28T16:12:33+03:00 Remove HasSrcSpan (#17494) - - - - - 651afd70 by Herbert Valerio Riedel at 2019-12-08T12:08:16+01:00 Document error-prone conditional definition of instances This can easily trip up people if one isn't aware of it. Usually it's better to avoid this kind of conditionality especially for typeclasses for which there's an compat-package as conditional instances like these tend to fragment the ecosystem into those packages that go the extra mile to provide backward compat via those compat-packages and those that fail to do so. - - - - - b521af56 by Herbert Valerio Riedel at 2019-12-08T12:09:54+01:00 Fix build-failure regression for base < 4.7 The `$>` operator definition is available only since base-4.7 which unfortunately wasn't caught before release to Hackage (but has been fixed up by a metadata-revision) This commit introduces a `CompatPrelude` module which allows to reduce the amount of CPP by ousting it to a central location, i.e. the new `CompatPrelude` module. This pattern also tends to reduce the tricks needed to silence unused import warnings. Addresses haskell/haddock#1119 - - - - - 556c375d by Sylvain Henry at 2020-01-02T19:01:55+01:00 Fix after Iface modules renaming - - - - - bd6c53e5 by Sylvain Henry at 2020-01-07T00:48:48+01:00 hsyl20-modules-renamer - - - - - fb23713b by Ryan Scott at 2020-01-08T07:41:13-05:00 Changes for GHC#17608 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2372 - - - - - 4a4dd382 by Ryan Scott at 2020-01-25T08:08:26-05:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - e782a44d by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename PackageConfig into UnitInfo - - - - - ba3c9f05 by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename lookupPackage - - - - - ab37f9b3 by Ben Gamari at 2020-01-29T13:00:44-05:00 Merge pull request haskell/haddock#1125 from haskell/wip/T17566-take-two Changes for GHC#17566 - - - - - 3ebd5ae0 by Ryan Scott at 2020-01-31T05:56:50-05:00 Merge branch 'wip-hsyl20-package-refactor' into ghc-head - - - - - 602a747e by Richard Eisenberg at 2020-02-04T09:05:43+00:00 Echo GHC's removal of PlaceHolder module This goes with GHC's !2083. - - - - - ccfe5679 by Sylvain Henry at 2020-02-10T10:13:56+01:00 Module hierarchy: runtime (cf haskell/haddock#13009) - - - - - 554914ce by Cale Gibbard at 2020-02-10T16:10:39-05:00 Fix build of haddock in stage1 We have to use the correct version of the GHC API, but the version of the compiler itself doesn't matter. - - - - - 5b6fa2a7 by John Ericson at 2020-02-10T16:18:07-05:00 Noramlize `tested-with` fields in cabal files - - - - - e6eb3ebe by Vladislav Zavialov at 2020-02-16T13:25:26+03:00 No MonadFail/Alternative for P - - - - - 90e181f7 by Ben Gamari at 2020-02-18T14:13:47-05:00 Merge pull request haskell/haddock#1129 from obsidiansystems/wip/fix-stage1-build Fix build of haddock in stage1 - - - - - 93b64636 by Sylvain Henry at 2020-02-19T11:20:27+01:00 Modules: Driver (#13009) - - - - - da4f6c7b by Vladislav Zavialov at 2020-02-22T15:33:02+03:00 Use RealSrcSpan in InstMap - - - - - 479b1b50 by Ben Gamari at 2020-02-23T10:28:13-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 55ecacf0 by Sylvain Henry at 2020-02-25T15:18:27+01:00 Modules: Core (#13009) - - - - - 60867b3b by Vladislav Zavialov at 2020-02-28T15:53:52+03:00 Ignore the BufLoc/BufSpan added in GHC's !2516 - - - - - 1e5506d3 by Sylvain Henry at 2020-03-02T12:32:43+01:00 Modules: Core (#13009) - - - - - 6fb53177 by Richard Eisenberg at 2020-03-09T14:49:40+00:00 Changes in GHC's !1913. - - - - - 30b792ea by Ben Gamari at 2020-03-16T12:45:02-04:00 Merge pull request haskell/haddock#1130 from hsyl20/wip/hsyl20-modules-core2 Modules: Core (#13009) - - - - - cd761ffa by Sylvain Henry at 2020-03-18T15:24:00+01:00 Modules: Types - - - - - b6646486 by Ben Gamari at 2020-03-18T14:42:43-04:00 Merge pull request haskell/haddock#1133 from hsyl20/wip/hsyl20/modules/types Modules: Types - - - - - 9325d734 by Kleidukos at 2020-03-19T12:38:31-04:00 Replace the 'caption' class so that the collapsible sections are shown - - - - - 5e2bb555 by Kleidukos at 2020-03-19T12:38:31-04:00 Force ghc-8.8.3 - - - - - c6fcd0aa by Kleidukos at 2020-03-19T12:38:31-04:00 Update test fixtures - - - - - 5c849cb1 by Sylvain Henry at 2020-03-20T09:34:39+01:00 Modules: Types - - - - - 7f439155 by Alec Theriault at 2020-03-20T20:17:01-04:00 Merge branch 'ghc-8.8' into ghc-8.10 - - - - - b7904e5c by Alina Banerjee at 2020-03-20T20:24:17-04:00 Update parsing to strip whitespace from table cells (#1074) * Update parsing to strip leading & trailing whitespace from table cells * Update fixture data to disallow whitespaces at both ends in table cells * Add test case for whitespaces stripped from both ends of table cells * Update table reference test data for html tests - - - - - b9d60a59 by Alec Theriault at 2020-03-22T11:46:42-04:00 Clean up warnings * unused imports * imports of `Data.List` without import lists * missing `CompatPrelude` file in `.cabal` - - - - - 0c317dbe by Alec Theriault at 2020-03-22T18:46:54-04:00 Fix NPM security warnings This was done by calling `npm audit fix`. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. - - - - - 6e306242 by Alec Theriault at 2020-03-22T20:10:52-04:00 Tentative 2.24 release Adjusted changelogs and versions in `.cabal` files in preparation for the upcoming release bundled with GHC 8.10. - - - - - 1bfb4645 by Ben Gamari at 2020-03-23T16:40:54-04:00 Merge commit '3c2944c037263b426c4fe60a3424c27b852ea71c' into HEAD More changes from the GHC types module refactoring. - - - - - be8c6f3d by Alec Theriault at 2020-03-26T20:10:53-04:00 Update `.travis.yml` to work with GHC 8.10.1 * Regenerated the Travis file with `haskell-ci` * Beef up `.cabal` files with more `tested-with` information - - - - - b025a9c6 by Alec Theriault at 2020-03-26T20:10:53-04:00 Update README Removed some out of date links/info, added some more useful links. * badge to Hackage * update old trac link * `ghc-head` => `ghc-8.10` * `cabal new-*` is now `cabal v2-*` and it should Just Work * `--test-option='--accept'` is the way to accept testsuite output - - - - - 564d889a by Alec Theriault at 2020-03-27T20:34:33-04:00 Fix crash in `haddock-library` on unicode space Our quickcheck tests for `haddock-library` stumbled across an edge case input that was causing Haddock to crash: it was a unicode space character. The root cause of the crash is that we were implicitly assuming that if a space character was not " \t\f\v\r", it would have to be "\n". We fix this by instead defining horizontal space as: any space character that is not '\n'. Fixes haskell/haddock#1142 - - - - - 2d360ba1 by Alec Theriault at 2020-03-27T21:57:32-04:00 Disallow qualified uses of reserved identifiers This a GHC bug (https://gitlab.haskell.org/ghc/ghc/issues/14109) too, but it is a relatively easy fix in Haddock. Note that the fix must live in `haddock-api` instead of `haddock-library` because we can only really decide if an identifier is a reserved one by asking the GHC lexer. Fixes haskell/haddock#952 - - - - - 47ae22ed by Alec Theriault at 2020-03-28T13:36:25-04:00 Remove unused `Haddock.Utils` functions * removed functions in `Haddock.Utils` that were not used anywhere (or exported from the `haddock-api` package) * moved GHC-specific utils from `Haddock.Utils` to `Haddock.GhcUtils` - - - - - c0291245 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use TTG empty extensions to remove some `error`'s None of these error cases should ever have been reachable, so this is just a matter of leveraging the type system to assert this. * Use the `NoExtCon` and `noExtCon` to handle case matches for no extension constructors, instead of throwing an `error`. * Use the extension field of `HsSpliceTy` to ensure that this variant of `HsType` cannot exist in an `HsType DocNameI`. - - - - - 0aff8dc4 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use `unLoc`/`noLoc` from GHC instead of `unL`/`reL` * `unL` is already defined by GHC as `unLoc` * `reL` is already defined by GHC as `noLoc` (in a safer way too!) * Condense `setOutputDir` and add a about exporting from GHC Fixes haskell/haddock#978 - - - - - bf6f2fb7 by Alec Theriault at 2020-03-28T13:36:25-04:00 Cleanup up GHC flags in `.cabal` files * enable more useful warning flags in `haddock-api`, handle the new warnings generated * remove `-fwarn-tabs` (now we'd use `-Wtabs`, but this has been in `-Wall` for a while now) - - - - - c576fbf1 by Alec Theriault at 2020-03-28T13:36:25-04:00 `haddock-library` document header level Document the fact the header level is going to always be between 1 and 6 inclusive. Along the way, I also optimized the parsing code a bit. - - - - - 71bce0ee by Alec Theriault at 2020-03-28T14:26:27-04:00 Disallow links in section headers This is quite straightforward to implement, since we already had a function `docToHtmlNoAnchors` (which we used to generate the link in the sidebar "Contents"). This breaks test `Bug387`, but that test case has aged badly: we now automatically generate anchors for all headings, so manually adding an anchor in a section makes no sense. Nested anchors are, as pointed out in haskell/haddock#1054, disallowed by the HTML standard. Fixes haskell/haddock#1054 - - - - - b461b0ed by Sylvain Henry at 2020-03-30T10:34:23+02:00 Modules: type checker - - - - - cd8cd1ee by Ben Gamari at 2020-03-31T12:45:02-04:00 Merge pull request haskell/haddock#1152 from hsyl20/wip/hsyl20/modules/tc Module renaming - - - - - 5e8f8ea7 by Felix Yan at 2020-04-01T17:58:06-07:00 Allow QuickCheck 2.14 Builds fine and all tests pass. - - - - - dc6b1633 by Sylvain Henry at 2020-04-05T16:43:44+02:00 Module renaming: amend previous patch - - - - - eee2f4ae by Ryan Scott at 2020-04-05T09:04:43-07:00 Fix haskell/haddock#1050 by filtering out invisible AppTy arguments This makes the `synifyType` case for `AppTy` more intelligent by taking into consideration the visibilities of each `AppTy` argument and filtering out any invisible arguments, as they aren't intended to be displayed in the source code. (See haskell/haddock#1050 for an example of what can happen if you fail to filter these out.) Along the way, I noticed that a special `synifyType` case for `AppTy t1 (CoercionTy {})` could be consolidated with the case below it, so I took the opportunity to tidy this up. - - - - - 23eb99e8 by Ben Gamari at 2020-04-07T11:19:58-04:00 Merge pull request haskell/haddock#1154 from hsyl20/wip/hsyl20/modules/tc Module renaming: amend previous patch - - - - - 072d994d by Ryan Scott at 2020-04-07T19:32:47-04:00 Make NoExtCon fields strict These changes are a part of a fix for [GHC#17992](https://gitlab.haskell.org/ghc/ghc/issues/17992). - - - - - d8ebf6c8 by Ignat Insarov at 2020-04-09T21:15:01-04:00 Recode Doc to Json. (#1159) * Recode Doc to Json. * More descriptive field labels. - - - - - 52df4b4e by Sylvain Henry at 2020-04-10T12:39:18+02:00 Module renaming - - - - - d9ab8ec8 by Cale Gibbard at 2020-04-14T11:43:34-04:00 Add instance of XCollectPat for DocNameI - - - - - 323d221d by Cale Gibbard at 2020-04-14T11:43:34-04:00 Rename XCollectPat -> CollectPass - - - - - 2df80867 by Alec Theriault at 2020-04-15T07:30:51-07:00 Prune docstrings that are never rendered When first creating a Haddock interface, trim `ifaceDocMap` and `ifaceArgMap` to not include docstrings that can never appear in the final output. Besides checking with GHC which names are exported, we also need to keep all the docs attached to instance declarations (it is much tougher to detect when an instance is fully private). This change means: * slightly smaller interface files (7% reduction on boot libs) * slightly less work to do processing docstrings that aren't used * no warnings in Haddock's output about private docstrings (see haskell/haddock#1070) I've tested manually that this does not affect any of the boot library generated docs (the only change in output was some small re-ordering in a handful of instance lists). This should mean no docstrings have been incorrectly dropped. - - - - - f49c90cc by Alec Theriault at 2020-04-15T07:30:51-07:00 Don't warn about missing links in miminal sigs When renaming the Haddock interface, never emit warnings when renaming a minimal signature. Also added some documention around `renameInterface`. Minimal signatures intentionally include references to potentially un-exported methods (see the discussion in haskell/haddock#330), so it is expected that they will not always have a link destination. On the principle that warnings should always be resolvable, this shouldn't produce a warning. See haskell/haddock#1070. - - - - - a9eda64d by Ben Gamari at 2020-04-17T09:27:35-04:00 Merge pull request haskell/haddock#1160 from hsyl20/wip/hsyl20/modules/systools Module renaming - - - - - f40d7879 by Cale Gibbard at 2020-04-20T11:30:38-04:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ttg-con-pat - - - - - a50e7753 by Ben Gamari at 2020-04-20T11:36:10-04:00 Merge pull request haskell/haddock#1165 from obsidiansystems/wip/ttg-con-pat Trees that Grow refactor (GHC !2553) - - - - - 6a24795c by Alec Theriault at 2020-04-21T08:06:45-07:00 Fallback to `hiDecl` when `extractDecl` fails Sometimes, the declaration being exported is a subdecl (for instance, a record accessor getting exported at the top-level). For these cases, Haddock has to find a way to produce some synthetic sensible top-level declaration. This is done with `extractDecl`. As is shown by haskell/haddock#1067, this is sometimes impossible to do just at a syntactic level (for instance when the subdecl is re-exported). In these cases, the only sensible thing to do is to try to reify a declaration based on a GHC `TyThing` via `hiDecl`. - - - - - eee1a8b7 by Sylvain Henry at 2020-04-24T15:46:05+02:00 Module structure - - - - - 50b9259c by Iñaki at 2020-04-25T18:38:11-04:00 Add support for custom section anchors (#1179) This allows to have stable anchors for groups, even if the set of groups in the documentation is altered. The syntax for setting the anchor of a group is -- * Group name #desiredAnchor# Which will produce an html anchor of the form '#g:desiredAnchor' Co-authored-by: Iñaki García Etxebarria <git at inaki.blueleaf.cc> - - - - - 4003c97a by Ben Gamari at 2020-04-26T09:35:15-04:00 Merge pull request haskell/haddock#1166 from hsyl20/wip/hsyl20/modules/utils Module structure - - - - - 5206ab60 by Sylvain Henry at 2020-04-27T16:47:39+02:00 Renamed UnitInfo fields - - - - - c32c333b by Sylvain Henry at 2020-04-27T17:32:58+02:00 UnitId has been renamed into Unit - - - - - 3e87db64 by Sylvain Henry at 2020-04-27T17:36:00+02:00 Fix for GHC.Unit.* modules - - - - - ae3323a7 by Ben Gamari at 2020-04-29T12:36:37-04:00 Merge pull request haskell/haddock#1183 from hsyl20/wip/hsyl20/unitid Refactoring of Unit code - - - - - b105564a by Artem Pelenitsyn at 2020-05-03T08:14:10+01:00 add dependency on exceptions because GHC.Exception was boiled down (ghc haskell/haddock#18075) - - - - - 9857eff3 by Zubin Duggal at 2020-05-04T18:48:25+01:00 Atomic update of NameCache in readHieFile - - - - - 86bbb226 by Sylvain Henry at 2020-05-14T16:36:27+02:00 Fix after Config module renaming - - - - - a4bbdbc2 by Gert-Jan Bottu at 2020-05-15T22:09:44+02:00 Explicit Specificity Support for Haddock - - - - - 46199daf by Ben Gamari at 2020-05-19T09:59:56-04:00 Merge pull request haskell/haddock#1192 from hsyl20/hsyl20/modules-config Fix after Config module renaming - - - - - f9a9d2ba by Gert-Jan Bottu at 2020-05-20T16:48:38-04:00 Explicit Specificity Support for Haddock - - - - - 55c5b7ea by Ben Gamari at 2020-05-21T00:32:02-04:00 Merge commit 'a8d7e66da4dcc3b242103271875261604be42d6e' into ghc-head - - - - - a566557f by Cale Gibbard at 2020-05-21T16:02:06-04:00 isBootSummary now produces a result of type IsBootInterface - - - - - ea52f905 by Zubin Duggal at 2020-05-24T17:55:48+01:00 update for hiefile-typeclass-info - - - - - 49ba7a67 by Willem Van Onsem at 2020-05-25T12:23:01-04:00 Use floor over round to calculate the percentage (#1195) If we compile documentation where only a small fraction is undocumented, it is misleading to see 100% coverage - 99% is more intuitive. Fixes haskell/haddock#1194 - - - - - c025ebf1 by Ben Gamari at 2020-05-29T14:32:42-04:00 Merge pull request haskell/haddock#1185 from obsidiansystems/boot-disambig isBootSummary now produces a result of type IsBootInterface - - - - - 74ab9415 by Ben Gamari at 2020-05-29T20:23:39-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - b40be944 by Ben Gamari at 2020-06-03T17:02:31-04:00 testsuite: Update expected output for simplified subsumption - - - - - 624be71c by Ryan Scott at 2020-06-05T12:43:23-04:00 Changes for GHC#18191 See https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3337. - - - - - fbd8f7ce by Sylvain Henry at 2020-06-08T15:31:47+02:00 Fix after unit refactoring - - - - - 743fda4d by Ben Gamari at 2020-06-09T12:09:58-04:00 Merge pull request haskell/haddock#1202 from hsyl20/wip/hsyl20/unitid-ii Fix after unit refactoring - - - - - d07a06a9 by Ryan Scott at 2020-06-13T07:16:55-04:00 Use HsForAllTelescope (GHC#18235) - - - - - 389bb60d by Ben Gamari at 2020-06-13T15:30:52-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - 7a377f5f by Ben Gamari at 2020-06-17T14:53:16-04:00 Merge pull request haskell/haddock#1199 from bgamari/wip/ghc-8.12 haddock: Bounds bumps for GHC 8.12 - - - - - 9fd9e586 by Krzysztof Gogolewski at 2020-06-17T16:09:07-04:00 Adapt Haddock to LinearTypes See ghc/ghc!852. - - - - - 46fe7636 by Ben Gamari at 2020-06-18T14:20:02-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 35a3c9e2 by Zubin Duggal at 2020-06-21T21:19:18+05:30 Use functions exported from HsToCore - - - - - 8abe3928 by Ben Gamari at 2020-06-24T13:53:39-04:00 Merge pull request haskell/haddock#1204 from wz1000/wip/haddock-hstocore Use functions exported from GHC.HsToCore.Docs - - - - - 22f2c937 by Matthías Páll Gissurarson at 2020-06-26T19:07:03+02:00 Adapt Haddock for QualifiedDo - - - - - 3f6208d7 by Vladislav Zavialov at 2020-06-28T14:28:16+03:00 Handle LexicalNegation's ITprefixminus - - - - - 03a19f41 by Sylvain Henry at 2020-07-02T09:37:38+02:00 Rename hsctarget into backend - - - - - ea17ff23 by Andreas Klebinger at 2020-07-02T17:44:18+02:00 Update for UniqFM changes. - - - - - 9872f2f3 by Ben Gamari at 2020-07-09T10:39:19-04:00 Merge pull request haskell/haddock#1209 from AndreasPK/wip/typed_uniqfm Update for UniqFM changes. - - - - - 68f7b668 by Krzysztof Gogolewski at 2020-07-12T18:16:57+02:00 Sync with GHC removing {-# CORE #-} pragma See ghc ticket haskell/haddock#18048 - - - - - eb372681 by Sylvain Henry at 2020-07-20T11:41:30+02:00 Rename hscTarget into backend - - - - - fb7f78bf by Ben Gamari at 2020-07-21T12:15:25-04:00 Merge pull request haskell/haddock#1214 from hsyl20/wip/hsyl20/hadrian/ncg Rename hscTarget into backend - - - - - 1e8f5b56 by Ben Gamari at 2020-07-23T09:11:50-04:00 Merge commit '904dce0cafe0a241dd3ef355775db47fc12f434d' into ghc-head - - - - - d8fd1775 by Zubin Duggal at 2020-07-23T18:46:40+05:30 Update for modular ping pong - - - - - 8416f872 by Ben Gamari at 2020-07-23T09:35:03-04:00 Merge pull request haskell/haddock#1200 from wz1000/wip/wz1000-modular-ping-pong Modular ping pong - - - - - a24a8577 by Ben Gamari at 2020-07-28T15:23:36-04:00 Bump GHC version to 9.0 - - - - - 6a51c9dd by Sylvain Henry at 2020-08-05T18:47:05+02:00 Fix after Outputable refactoring - - - - - c05e1c99 by Ben Gamari at 2020-08-10T14:41:41-04:00 Merge pull request haskell/haddock#1223 from hsyl20/wip/hsyl20/dynflags/exception Fix after Outputable refactoring - - - - - d964f15b by Sylvain Henry at 2020-08-12T11:58:49+02:00 Fix after HomeUnit - - - - - 8e6d5b23 by Ben Gamari at 2020-08-12T14:25:30-04:00 Merge pull request haskell/haddock#1225 from hsyl20/wip/hsyl20/plugins/homeunit Fix after HomeUnit - - - - - 8c7880fe by Sylvain Henry at 2020-08-17T14:13:29+02:00 Remove Ord FastString instance - - - - - 8ea410db by Alex Biehl at 2020-08-19T10:56:32+02:00 Another round of `npm audit fix` (#1228) This should shut down the warnings on Github. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. Last NPM dependency audit happend in d576b2327e2bc117f912fe0a9d595e9ae62614e0 Co-authored-by: Alex Biehl <alex.biehl at target.com> - - - - - 7af6e2a8 by Ben Gamari at 2020-08-31T13:59:34-04:00 Merge pull request haskell/haddock#1226 from hsyl20/wip/hsyl20/fs_ord Remove Ord FastString instance - - - - - ffbc8702 by Alan Zimmerman at 2020-09-07T21:47:41+01:00 Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - a93f1268 by Alan Zimmerman at 2020-09-07T23:11:38+01:00 Merge pull request haskell/haddock#1232 from haskell/wip/T18639-remove-generated-pragma, Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - 1f605d50 by Ben Gamari at 2020-09-14T18:30:01-04:00 Bump GHC version to 9.1 - - - - - 6599df62 by Vladislav Zavialov at 2020-09-18T14:05:15+03:00 Bump base upper bound to 4.16 - - - - - a01b3c43 by Ben Gamari at 2020-09-22T15:41:48-04:00 Update hypsrc-test for QuickLook This appears to be a spurious change. - - - - - e9cc6cac by Vladislav Zavialov at 2020-09-26T21:00:12+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 30e3ca7c by Sylvain Henry at 2020-09-29T11:18:32-04:00 Update for parser (#1234) - - - - - b172f3e3 by Vladislav Zavialov at 2020-09-30T01:01:30+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 0b9c08d3 by Sylvain Henry at 2020-09-30T11:02:33+02:00 Adapt to GHC parser changes - - - - - b9540b7a by Sylvain Henry at 2020-10-12T09:13:38-04:00 Don't pass the HomeUnitId (#1239) - - - - - 34762e80 by HaskellMouse at 2020-10-13T12:58:04+03:00 Changed tests due to unification of `Nat` and `Natural` in the follwing merge request: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3583 - - - - - 256f86b6 by Vladislav Zavialov at 2020-10-15T10:48:03+03:00 Add whitespace in: map ($ v) - - - - - 4a3f711b by Alan Zimmerman at 2020-10-19T08:57:27+01:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled - - - - - 072cdd21 by Alan Zimmerman at 2020-10-21T14:48:28-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 9e09a445 by Alan Zimmerman at 2020-10-21T23:53:34-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 636d7de3 by Sylvain Henry at 2020-10-26T14:31:54-04:00 GHC.Driver.Types refactoring (#1242) - - - - - a597f000 by Ryan Scott at 2020-10-29T04:18:05-04:00 Adapt to the removal of Hs{Boxed,Constraint}Tuple See ghc/ghc!4097 and GHC#18723. - - - - - b96660fb by Ryan Scott at 2020-10-30T04:53:05-04:00 Adapt to HsConDecl{H98,GADT}Details split Needed for GHC#18844. - - - - - c287d82c by Ryan Scott at 2020-10-30T19:35:59-04:00 Adapt to HsOuterTyVarBndrs These changes accompany ghc/ghc!4107, which aims to be a fix for haskell/haddock#16762. - - - - - a34c31a1 by Ryan Scott at 2020-11-13T13:38:34-05:00 Adapt to splitPiTysInvisible being renamed to splitInvisPiTys This is a part of !4434, a fix for GHC#18939. - - - - - 66ea459d by Sylvain Henry at 2020-11-16T10:59:30+01:00 Fix after Plugins moved into HscEnv - - - - - 508556d8 by Ben Gamari at 2020-11-18T15:47:40-05:00 Merge pull request haskell/haddock#1253 from hsyl20/wip/hsyl20/plugins/hscenv Fix after Plugins moved into HscEnv - - - - - 620fec1a by Andreas Klebinger at 2020-11-24T20:51:59+01:00 Update for changes in GHC's Pretty - - - - - 01cc13ab by Richard Eisenberg at 2020-11-25T23:18:35-05:00 Avoid GHC#18932. - - - - - 8d29ba21 by Cale Gibbard at 2020-11-25T23:18:35-05:00 Add type arguments to PrefixCon - - - - - 414d5f87 by Sylvain Henry at 2020-11-30T17:06:04+01:00 DynFlags's unit fields moved to HscEnv - - - - - e356668c by Ben Gamari at 2020-11-30T11:11:37-05:00 Merge pull request haskell/haddock#1258 from hsyl20/wip/hsyl20/hscenv/unitstate Unit fields moved from DynFlags to HscEnv - - - - - 7cf552f1 by Ben Gamari at 2020-12-03T10:31:27-05:00 Merge pull request haskell/haddock#1257 from AndreasPK/wip/andreask/opt_dumps Update for changes in GHC's Pretty - - - - - fc0871c3 by Veronika Romashkina at 2020-12-08T16:35:33+01:00 Fix docs links from Darcs to GitHub in intro (#1262) - - - - - 7059e808 by Veronika Romashkina at 2020-12-08T16:36:16+01:00 Use gender neutral word in docs (#1260) - - - - - 1b16e5ee by Maximilian Tagher at 2020-12-08T16:40:03+01:00 Allow scrolling search results (#1235) Closes https://github.com/haskell/haddock/issues/1231 - - - - - 8a118c01 by dependabot[bot] at 2020-12-08T16:40:25+01:00 Bump bl from 1.2.2 to 1.2.3 in /haddock-api/resources/html (#1255) Bumps [bl](https://github.com/rvagg/bl) from 1.2.2 to 1.2.3. - [Release notes](https://github.com/rvagg/bl/releases) - [Commits](https://github.com/rvagg/bl/compare/v1.2.2...v1.2.3) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - c89ff587 by Xia Li-yao at 2020-12-08T16:42:17+01:00 Allow more characters in anchor following module reference (#1220) - - - - - 14af7d64 by Xia Li-yao at 2020-12-08T16:43:05+01:00 Add dangling changes from branches ghc-8.6 and ghc-8.8 (#1243) * Fix multiple typos and inconsistencies in doc/markup.rst Note: I noticed some overlap with haskell/haddock#1112 from @wygulmage and haskell/haddock#1081 from @parsonsmatt after creating these proposed changes - mea culpa for not looking at the open PRs sooner. * Fix haskell/haddock#1113 If no Signatures, no section of index.html * Change the formatting of missing link destinations The current formatting of the missing link destination does not really help user to understand the reasons of the missing link. To address this, I've changed the formatting in two ways: - the missing link symbol name is now fully qualified. This way you immediately know which haskell module cannot be linked. It is then easier to understand why this module does not have documentation (hidden module or broken documentation). - one line per missing link, that's more readable now that symbol name can be longer due to qualification. For example, before haddock was listing missing symbol such as: ``` could not find link destinations for: Word8 Word16 mapMaybe ``` Now it is listed as: ``` could not find link destinations for: - Data.Word.Word8 - Data.Word.Word16 - Data.Maybe.mapMaybe ``` * Add `--ignore-link-symbol` command line argument This argument can be used multiples time. A missing link to a symbol listed by `--ignore-link-symbol` won't trigger "missing link" warning. * Forbid spaces in anchors (#1148) * Improve error messages with context information (#1060) Co-authored-by: Matt Audesse <matt at mattaudesse.com> Co-authored-by: Mike Pilgrem <mpilgrem at users.noreply.github.com> Co-authored-by: Guillaume Bouchard <guillaume.bouchard at tweag.io> Co-authored-by: Pepe Iborra <pepeiborra at gmail.com> - - - - - 89e3af13 by tomjaguarpaw at 2020-12-08T18:00:04+01:00 Enable two warnings (#1245) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - c3320f8d by Willem Van Onsem at 2020-12-08T18:26:55+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 685df308 by Alex Biehl at 2020-12-08T20:06:26+01:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - be3ec3c0 by Alex Biehl at 2020-12-08T20:06:26+01:00 Import intercalate - - - - - 32c33912 by Matthías Páll Gissurarson at 2020-12-08T21:15:30+01:00 Adapt Haddock for QualifiedDo - - - - - 31696088 by Alex Biehl at 2020-12-08T22:06:02+01:00 Fix haddock-library tests - - - - - fbc0998a by Alex Biehl at 2020-12-08T23:08:23+01:00 Move to GitHub CI (#1266) * Initial version of ci.yml This is a straight copy from Dmitrii Kovanikov's blog post at https://kodimensional.dev/github-actions. Will adapt to haddock in successive commits. * Delete .travis.yml * Modify to only test on ghc-8.10.{1,2} * Use actions/setup-haskell at v1.1.4 * Relax QuickCheck bound on haddock-api * Remove stack matrix for now * Nail down to ghc-8.10 branch for now * Pin index state to 2020-12-08T20:13:44Z for now * Disable macOS and Windows tests for now for speed up - - - - - 5b946b9a by tomjaguarpaw at 2020-12-10T19:01:41+01:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - bc5a408f by dependabot[bot] at 2020-12-10T19:02:16+01:00 Bump ini from 1.3.5 to 1.3.7 in /haddock-api/resources/html (#1269) Bumps [ini](https://github.com/isaacs/ini) from 1.3.5 to 1.3.7. - [Release notes](https://github.com/isaacs/ini/releases) - [Commits](https://github.com/isaacs/ini/compare/v1.3.5...v1.3.7) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - d02995f1 by Andrew Martin at 2020-12-14T16:48:40-05:00 Update for boxed rep - - - - - a381aeff by Ben Gamari at 2020-12-15T15:13:30-05:00 Revert "Enable two warnings (#1245) (#1268)" As this does not build on GHC `master`. This reverts commit 7936692badfe38f23ae95b51fb7bd7c2ff7e9bce. - - - - - a63c0a9e by Ben Gamari at 2020-12-15T15:17:59-05:00 Revert "Update for boxed rep" This reverts commit 4ffb30d8b637ccebecc81ce610f0af451ac8088d. - - - - - 53bfbb29 by Ben Gamari at 2020-12-15T15:37:24-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - bae76a30 by Ben Gamari at 2020-12-16T02:44:42+00:00 Update output for nullary TyConApp optimisation (ghc/ghc!2952) - - - - - 4b733b57 by Krzysztof Gogolewski at 2020-12-16T20:03:14+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. - - - - - ee463bd3 by Ryan Scott at 2020-12-16T16:55:23-05:00 Adapt to HsCoreTy (formerly NewHsTypeX) becoming a type synonym Needed for !4417, the fix for GHC#15706 and GHC#18914. - - - - - ed0b02f8 by tomjaguarpaw at 2020-12-19T10:17:19+00:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - d80bf8f5 by Sylvain Henry at 2020-12-21T10:09:25+01:00 Fix after binder collect changes - - - - - bf4c9d32 by Adam Gundry at 2020-12-23T21:35:01+00:00 Adapt to changes to GlobalRdrElt and AvailInfo Needed for ghc/ghc!4467 - - - - - 37736c4c by John Ericson at 2020-12-28T12:27:02-05:00 Support a new ghc --make node type for parallel backpack upsweep - - - - - 717bdeac by Vladislav Zavialov at 2020-12-29T10:50:02+03:00 Inline and fix getGADTConTypeG The getGADTConTypeG used HsRecTy, which is at odds with GHC issue haskell/haddock#18782. I noticed that getGADTConTypeG was only used in the Hoogle backend. Interestingly, when handling H98 constructors, Hoogle converts RecCon to PrefixCon (see Haddock.Backends.Hoogle.ppCtor). So I changed getGADTConTypeG to handle RecConGADT in the same manner as PrefixConGADT, and after this simplification moved it into the 'where' clause of ppCtor, to the only place where it is used. The practical effect of this change is as follows. Consider this example: data TestH98 = T98 { bar::Int } data TestGADT where TG :: { foo :: Int } -> TestGADT Before this patch, haddock --hoogle used to produce: T98 :: Int -> TestH98 [TG] :: {foo :: Int} -> TestGADT Notice how the record syntax was discarded in T98 but not TG. With this patch, we always produce signatures without record syntax: T98 :: Int -> TestH98 [TG] :: Int -> TestGADT I suspect this might also be a bugfix, as currently Hoogle doesn't seem to render GADT record constructors properly. - - - - - cb1b8c56 by Andreas Abel at 2020-12-30T21:12:37+01:00 Build instructions: haddock-library and -api first! - - - - - b947f6ad by Ben Gamari at 2020-12-31T13:04:19-05:00 Merge pull request haskell/haddock#1281 from obsidiansystems/wip/backpack-j Changes to support -j with backpack - - - - - 120e1cfd by Hécate Moonlight at 2021-01-04T19:54:58+01:00 Merge pull request haskell/haddock#1282 from andreasabel/master Build instructions: haddock-library and -api first! - - - - - fd45e41a by Ben Gamari at 2021-01-05T16:14:31-05:00 Merge remote-tracking branch 'origin/ghc-8.10' into ghc-9.0 - - - - - b471bdec by Ben Gamari at 2021-01-05T16:23:02-05:00 Merge commit '1e56f63c3197e7ca1c1e506e083c2bad25d08793' into ghc-9.0 - - - - - 81cdbc41 by Alex Biehl at 2021-01-09T12:14:41+01:00 Prepare Haddock for being a GHC Plugin - - - - - b646d952 by Alex Biehl at 2021-01-09T12:14:41+01:00 Make Haddock a GHC Plugin - - - - - cc044674 by Alex Biehl at 2021-01-09T12:14:41+01:00 Add -j[n] CLI param to Haddock executable It translates to `--ghcopt=-j[n]` - - - - - 84a04073 by Alex Biehl at 2021-01-09T12:14:41+01:00 Abstract Monad for interface creation I found that when running as a plugin the lookupName function (which runs in Ghc monad) does not work correctly from the typeCheckResultAction hook. Instead, we abstracted the monad used when creating interfaces, so that access to GHC session specific parts is explicit and so that the TcM can provide their (correct) implementation of lookupName. - - - - - 5be2c4f7 by Alex Biehl at 2021-01-09T12:14:41+01:00 Accept tests - - - - - 8cefee9d by Alex Biehl at 2021-01-09T16:10:47+01:00 Add missing dependency for mtl - - - - - 3681f919 by Ben Gamari at 2021-01-13T18:39:25-05:00 Merge remote-tracking branch 'origin/ghc-9.0' into ghc-head - - - - - 33c6b152 by Hécate Moonlight at 2021-01-14T16:04:20+01:00 Merge pull request haskell/haddock#1273 from hsyl20/wip/hsyl20/arrows Fix after binder collect changes - - - - - 70d13e8e by Joachim Breitner at 2021-01-22T19:03:45+01:00 Make haddock more robust to changes to the `Language` data type With the introduction of GHC2021, the `Languages` data type in GHC will grow. In preparation of that (and to avoid changing haddock with each new language), this change makes the code handle extensions to that data type gracefully. (cherry picked from commit c341dd7c9c3fc5ebc83a2d577c5a726f3eb152a5) - - - - - 7d6dd57a by John Ericson at 2021-01-22T22:02:02+00:00 Add `NoGhcTc` instance now that it's not closed - - - - - e5fdaf0a by Alan Zimmerman at 2021-01-23T22:57:44+00:00 Merge pull request haskell/haddock#1293 from obsidiansystems/wip/fix-18936 Add `NoGhcTc` instance now that it's not closed - - - - - 989a1e05 by Oleg Grenrus at 2021-01-24T16:11:46+03:00 Add import list to Data.List - - - - - 368e144a by Ben Gamari at 2021-01-28T22:15:48+01:00 Adapt to "Make PatSyn immutable" - - - - - abe66c21 by Alfredo Di Napoli at 2021-02-01T08:05:35+01:00 Rename pprLogErrMsg to new name - - - - - e600e75c by Hécate Moonlight at 2021-02-05T14:53:00+01:00 Move CI to ghc-9.0 - - - - - dd492961 by Vladislav Zavialov at 2021-02-05T14:53:00+01:00 Update cabal.project and README build instructions - - - - - 31bd292a by Hécate Moonlight at 2021-02-05T15:03:56+01:00 Merge pull request haskell/haddock#1296 from Kleidukos/ghc-9.0 Merge the late additions to ghc-8.10 into ghc-9.0 - - - - - 6388989e by Vladislav Zavialov at 2021-02-05T17:41:57+03:00 Cleanup: fix build warnings - - - - - f99407ef by Daniel Rogozin at 2021-02-05T18:11:48+03:00 type level characters support for haddock (required for haskell/haddock#11342) - - - - - d8c6b26f by Hécate Moonlight at 2021-02-05T17:44:50+01:00 Add a CONTRIBUTING.md file - - - - - 6a01ad98 by Hécate Moonlight at 2021-02-05T17:58:16+01:00 Merge pull request haskell/haddock#1312 from Kleidukos/proper-branch-etiquette Add a CONTRIBUTING.md file - - - - - 955eecc4 by Vladislav Zavialov at 2021-02-05T20:29:00+03:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into ghc-head - - - - - 47b3d6ab by Hécate Moonlight at 2021-02-05T19:09:38+01:00 Amend the CONTRIBUTING.md file - - - - - 23de6137 by Hécate Moonlight at 2021-02-05T19:16:49+01:00 Merge pull request haskell/haddock#1313 from Kleidukos/amend-contributing Amend the CONTRIBUTING.md file - - - - - 69026b59 by Krzysztof Gogolewski at 2021-02-05T23:05:56+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. (cherry picked from commit b4b4d896d2d68d6c48e7db7bfe95c185ca0709cb) - - - - - ea026b78 by Oleg Grenrus at 2021-02-06T17:14:45+01:00 Add import list to Data.List - - - - - 5204326f by Hécate Moonlight at 2021-02-06T17:15:44+01:00 Merge pull request haskell/haddock#1316 from Kleidukos/explicit-imports-to-data-list Add import list to Data.List - - - - - 1f4d2136 by Ben Gamari at 2021-02-06T11:53:31-05:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ghc-head-merge - - - - - 13f0d09a by Ben Gamari at 2021-02-06T11:53:45-05:00 Fix partial record selector warning - - - - - 5c115f7e by Ben Gamari at 2021-02-06T11:55:52-05:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into wip/ghc-head-merge - - - - - b6fd8b75 by Ben Gamari at 2021-02-06T12:01:31-05:00 Merge commit '41964cb2fd54b5a10f8c0f28147015b7d5ad2c02' into wip/ghc-head-merge - - - - - a967194c by Ben Gamari at 2021-02-06T18:30:35-05:00 Merge branch 'wip/ghc-head-merge' into ghc-head - - - - - 1f4c3a91 by MorrowM at 2021-02-07T01:52:33+02:00 Fix search div not scrolling - - - - - 684b1287 by Iñaki García Etxebarria at 2021-02-07T16:13:04+01:00 Add support for labeled module references Support a markdown-style way of annotating module references. For instance -- | [label]("Module.Name#anchor") will create a link that points to the same place as the module reference "Module.Name#anchor" but the text displayed on the link will be "label". - - - - - bdb55a5d by Hécate Moonlight at 2021-02-07T16:18:10+01:00 Merge pull request haskell/haddock#1319 from alexbiehl/alex/compat Backward compat: Add support for labeled module references - - - - - 6ca70991 by Hécate Moonlight at 2021-02-07T16:21:29+01:00 Merge pull request haskell/haddock#1314 from tweag/show-linear-backport Backport haskell/haddock#1238 (linear types) to ghc-9.0 - - - - - d9d73298 by Alex Biehl at 2021-02-07T17:46:25+01:00 Remove dubious parseModLink Instead construct the ModLink value directly when parsing. - - - - - 33b4d020 by Hécate Moonlight at 2021-02-07T17:52:05+01:00 Merge pull request haskell/haddock#1320 from haskell/alex/fix Remove dubious parseModLink - - - - - 54211316 by Hécate Moonlight at 2021-02-07T18:12:07+01:00 Merge pull request haskell/haddock#1318 from MorrowM/ghc-9.0 Fix search div not scrolling - - - - - 19db679e by alexbiehl-gc at 2021-02-07T18:14:46+01:00 Merge pull request haskell/haddock#1317 from bgamari/wip/ghc-head-merge Merge ghc-8.10 into ghc-head - - - - - 6bc1e9e4 by Willem Van Onsem at 2021-02-07T18:25:30+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - c8537cf8 by alexbiehl-gc at 2021-02-07T18:30:40+01:00 Merge pull request haskell/haddock#1322 from haskell/alex/forward-port simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 2d47ae4e by alexbiehl-gc at 2021-02-07T18:39:59+01:00 Merge branch 'ghc-head' into ghc-9.0 - - - - - 849e4733 by Hécate Moonlight at 2021-02-07T18:43:19+01:00 Merge pull request haskell/haddock#1321 from Kleidukos/ghc-9.0 Merge ghc-9.0 into ghc-head - - - - - ee6095d7 by Sylvain Henry at 2021-02-08T11:36:38+01:00 Update for Logger - - - - - 4ad688c9 by Alex Biehl at 2021-02-08T18:11:24+01:00 Merge pull request haskell/haddock#1310 from hsyl20/wip/hsyl20/logger2 Logger refactoring - - - - - 922a9e0e by Ben Gamari at 2021-02-08T12:54:33-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - 991649d2 by Sylvain Henry at 2021-02-09T10:55:17+01:00 Fix to build with HEAD - - - - - a8348dc2 by Hécate Moonlight at 2021-02-09T10:58:51+01:00 Merge pull request haskell/haddock#1327 from hsyl20/wip/hsyl20/logger2 Fix to build with HEAD - - - - - 0abdbca6 by Fendor at 2021-02-09T20:06:15+01:00 Add UnitId to Target record - - - - - d5790a0e by Alex Biehl at 2021-02-11T10:32:32+01:00 Stable sort for (data/newtype) instances - - - - - 8e6036f5 by Alex Biehl at 2021-02-11T10:32:32+01:00 Also make TyLit deterministic - - - - - f76d2945 by Hécate Moonlight at 2021-02-11T11:00:31+01:00 Merge pull request haskell/haddock#1329 from hsyl20/hsyl20/stabe_iface Stable sort for instances - - - - - 5e0469ea by Oleg Grenrus at 2021-02-14T15:28:15+02:00 Add import list to Data.List in Haddock.Interface.Create - - - - - fa57cd24 by Hécate Moonlight at 2021-02-14T17:19:27+01:00 Merge pull request haskell/haddock#1331 from phadej/more-explicit-data-list-imports Add import list to Data.List in Haddock.Interface.Create - - - - - f0cd629c by Hécate Moonlight at 2021-02-21T00:22:01+01:00 Merge pull request haskell/haddock#1311 from fendor/wip/add-targetUnitId-to-target Add UnitId to Target record - - - - - 674ef723 by Joachim Breitner at 2021-02-22T10:39:18+01:00 html-test: Always set language from ghc-9.2 on, the “default” langauge of GHC is expected to change more wildly. To prepare for that (and unblock https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4853), this sets the language for all the test files to `Haskell2010`. This should insolate this test suite against changes to the default. Cherry-picked from https://github.com/haskell/haddock/pull/1341 - - - - - f072d623 by Hécate Moonlight at 2021-02-22T10:56:51+01:00 Merge pull request haskell/haddock#1342 from nomeata/joachim/haskell2010-in-tests-ghc-head html-test: Always set language - - - - - caebbfca by Hécate Moonlight at 2021-02-22T11:53:07+01:00 Clean-up of Interface and Interface.Create's imports and pragmata - - - - - f6caa89d by Hécate Moonlight at 2021-02-22T11:54:58+01:00 Merge pull request haskell/haddock#1345 from Kleidukos/head/fix-interface-imports [ghc-head] Clean-up of Interface and Interface.Create's imports and pragmata - - - - - 7395c9cb by Hécate Moonlight at 2021-02-22T18:44:57+01:00 Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 6e9fb5d5 by Hécate Moonlight at 2021-02-22T18:45:28+01:00 Merge pull request haskell/haddock#1348 from Kleidukos/head/explicit-imports-interface Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 9198b118 by Alan Zimmerman at 2021-02-22T20:04:24+00:00 Context becomes a Maybe in the GHC AST This prevents noLoc's appearing in the ParsedSource. Match the change in GHC. - - - - - 0af20f64 by Hécate Moonlight at 2021-02-23T12:36:12+01:00 Fix the call-site of guessTarget in Interface.hs Explicit the imports from GHC.HsToCore.Docs - - - - - b7886885 by Hécate Moonlight at 2021-02-23T12:37:54+01:00 Merge pull request haskell/haddock#1349 from Kleidukos/fix-interface-guesstarget-call Fix the call-site of guessTarget in Interface.hs - - - - - 9cf041ba by Sylvain Henry at 2021-02-24T11:08:20+01:00 Fix haddockHypsrcTest output in ghc-head - - - - - b194182a by Hécate Moonlight at 2021-02-24T11:12:36+01:00 Merge pull request haskell/haddock#1351 from hsyl20/wip/hsyl20/fix-head Fix haddockHypsrcTest output in ghc-head - - - - - 3ce8b375 by Shayne Fletcher at 2021-03-06T09:55:03-05:00 Add ITproj to parser - - - - - d2abf762 by Ben Gamari at 2021-03-06T19:26:49-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a0f6047d by Andrew Martin at 2021-03-07T11:25:23-05:00 Update for boxed rep - - - - - 6f63c99e by Ben Gamari at 2021-03-10T13:20:21-05:00 Update for "FastString: Use FastMutInt instead of IORef Int" - - - - - e13f01df by Luke Lau at 2021-03-10T15:38:40-05:00 Implement template-haskell's putDoc This catches up to GHC using the new extractTHDocs function, which returns documentation added via the putDoc function (provided it was compiled with Opt_Haddock). Since it's already a map from names -> docs, there's no need to do traversal etc. It also matches the change from the argument map being made an IntMap rather than a Map Int - - - - - 89263d94 by Alan Zimmerman at 2021-03-15T17:15:26+00:00 Match changes in GHC AST for in-tree API Annotations As landed via https://gitlab.haskell.org/ghc/ghc/-/merge_requests/2418 - - - - - 28db1934 by Alan Zimmerman at 2021-03-15T20:40:09+00:00 Change some type family test results. It is not clear to me whether the original was printing incorrectly (since we did not have the TopLevel flag before now), or if this behaviour is expected. For the time being I am assuming the former. - - - - - 7c11c989 by Sylvain Henry at 2021-03-22T10:05:19+01:00 Fix after NameCache changes - - - - - addbde15 by Sylvain Henry at 2021-03-22T10:05:19+01:00 NameCache doesn't store a UniqSupply anymore - - - - - 15ec6cec by Ben Gamari at 2021-03-22T17:53:44-04:00 Bump GHC version to 9.2 - - - - - dbd6aa63 by Hécate Moonlight at 2021-03-24T14:28:36+01:00 Merge pull request haskell/haddock#1365 from hsyl20/wip/hsyl20/iface1 NameCache refactoring - - - - - 2d32da7e by Oleg Grenrus at 2021-03-27T01:12:00+02:00 Specialization of Data.List - - - - - 32b84fa6 by Fendor at 2021-03-27T10:50:17+01:00 Add UnitId to Target record This way we always know to which home-unit a given target belongs to. So far, there only exists a single home-unit at a time, but it enables having multiple home-units at the same time. - - - - - 54bf9f0e by Hécate Moonlight at 2021-03-28T14:08:35+02:00 Merge pull request haskell/haddock#1368 from fendor/target-unit-id-revert Add UnitId to Target record - - - - - 7dea168a by Alan Zimmerman at 2021-03-29T08:45:52+01:00 EPA : Rename ApiAnn to EpAnn - - - - - 72967f65 by Alfredo Di Napoli at 2021-03-29T09:47:01+02:00 pprError changed name in GHC - - - - - 4bc61035 by Alan Zimmerman at 2021-03-29T16:16:27-04:00 EPA : Rename ApiAnn to EpAnn - - - - - 108d031d by Ben Gamari at 2021-03-29T18:49:36-04:00 Merge commit '36418c4f70d7d2b179a77925b3ad5caedb08c9b5' into HEAD - - - - - 1444f700 by Ben Gamari at 2021-03-31T09:18:39-04:00 Merge pull request haskell/haddock#1370 from adinapoli/wip/adinapoli-diag-reason-severity Rename pprError to mkParserErr - - - - - d3087b79 by Ben Gamari at 2021-03-31T11:34:17-04:00 Merge commit 'd8d8024ad6796549a8d3b5512dabf3288d14e30f' into ghc-head - - - - - 170b79e9 by Ben Gamari at 2021-03-31T12:24:56-04:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - db0d6bae by Ben Gamari at 2021-04-10T09:34:35-04:00 Bump GHC version to 9.3 - - - - - a9f2c421 by Alan Zimmerman at 2021-04-19T18:26:46-04:00 Update for EPA changes in GHC (cherry picked from commit cafb48118f7c111020663776845897e225607b41) - - - - - 1ee4b7c7 by Sylvain Henry at 2021-05-11T10:00:06+02:00 Removal of HsVersions.h (#1388) * Update for EPA changes in GHC * Account for HsVersions.h removal Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - 79e819e9 by Hécate Moonlight at 2021-05-11T10:14:47+02:00 Revert "Removal of HsVersions.h (#1388)" This reverts commit 72118896464f94d81f10c52f5d9261efcacc57a6. - - - - - 3dbd3f8b by Alan Zimmerman at 2021-05-11T10:15:17+02:00 Update for EPA changes in GHC - - - - - 2ce80c17 by Sylvain Henry at 2021-05-11T10:15:19+02:00 Account for HsVersions.h removal - - - - - 00e4c918 by Christiaan Baaij at 2021-05-13T08:21:56+02:00 Add Haddock support for the OPAQUE pragma (#1380) - - - - - 8f9049b2 by Hécate Moonlight at 2021-05-13T08:40:22+02:00 fixup! Use GHC 9.2 in CI runner - - - - - 27ddec38 by Alan Zimmerman at 2021-05-13T22:51:20+01:00 EPA: match changes from GHC T19834 - - - - - f8a1d714 by Felix Yan at 2021-05-14T17:10:04+02:00 Allow hspec 2.8 (#1389) All tests are passing. - - - - - df44453b by Divam Narula at 2021-05-20T15:42:42+02:00 Update ref, the variables got renamed. (#1391) This is due to ghc/ghc!5555 which caused a change in ModDetails in case of NoBackend. Now the initModDetails is used to recreate the ModDetails from interface and in-memory ModDetails is not used. - - - - - e46bfc87 by Alan Zimmerman at 2021-05-20T19:05:09+01:00 Remove Maybe from HsQualTy Match changes in GHC for haskell/haddock#19845 - - - - - 79bd7b62 by Shayne Fletcher at 2021-05-22T08:20:39+10:00 FieldOcc: rename extFieldOcc to foExt - - - - - 6ed68c74 by Ben Gamari at 2021-05-21T22:29:30-04:00 Merge commit '3b6a8774bdb543dad59b2618458b07feab8a55e9' into ghc-head - - - - - f9a02d34 by Alfredo Di Napoli at 2021-05-24T13:53:00+02:00 New Parser diagnostics interface - - - - - 392807d0 by Ben Gamari at 2021-05-24T09:57:40-04:00 Merge pull request haskell/haddock#1394 from adinapoli/wip/adinapoli-align-ps-messages Align Haddock to use the new Parser diagnostics interface - - - - - 33023cd8 by Ben Gamari at 2021-05-24T11:19:16-04:00 Revert "Add Haddock support for the OPAQUE pragma (#1380)" This reverts commit a1337c599ef7720b0482a25c55f11794112496dc. The GHC patch associated with this change is not yet ready to be merged. - - - - - 8c005af7 by Ben Simms at 2021-05-28T07:56:20+02:00 CI configuration for ghc-head (#1395) - - - - - 1e947612 by Hécate Moonlight at 2021-05-28T12:27:35+02:00 Use GHC 9.2 in CI runner (#1378) - - - - - e6fa10ab by CGenie at 2021-05-31T09:02:13+02:00 Add page about common errors (#1396) * Update index.rst Common errors page * Create common-errors.rst * Update common-errors.rst * Use GHC 9.2 in CI runner (#1378) * [haddock-api] remove .hspec-failures Co-authored-by: Hécate Moonlight <Kleidukos at users.noreply.github.com> - - - - - abc72a8d by Sylvain Henry at 2021-06-01T10:02:06+02:00 Adapt Haddock to Logger and Parser changes (#1399) - - - - - 91373656 by Zubin Duggal at 2021-06-01T20:45:10+02:00 Update haddockHypSrc tests since we now compute slighly more type info (#1397) - - - - - ed712822 by Marcin Szamotulski at 2021-06-02T08:54:33+02:00 Added myself to contributors - - - - - 49fdbcb7 by Marcin Szamotulski at 2021-06-02T08:57:24+02:00 Document multi component support - - - - - 9ddc8d7d by Hécate Moonlight at 2021-06-02T09:35:55+02:00 Merge pull request haskell/haddock#1379 from coot/coot/document-multi-component-support Document multi component support - - - - - 585b5c5e by Ben Simms at 2021-06-02T19:46:54+02:00 Update CONTRIBUTING.md (#1402) - - - - - 1df4a605 by Ben Simms at 2021-06-02T19:47:14+02:00 Update CONTRIBUTING.md (#1403) - - - - - 58ea43d2 by sheaf at 2021-06-02T22:09:06+02:00 Update Haddock Bug873 to account for renaming - - - - - c5d0ab23 by Vladislav Zavialov at 2021-06-10T13:35:42+03:00 HsToken in FunTy, RecConGADT - - - - - 1ae2f40c by Hécate Moonlight at 2021-06-11T11:19:09+02:00 Update the CI badges - - - - - 6fdc4de2 by Sylvain Henry at 2021-06-28T19:21:17+02:00 Fix mkParserOpts (#1411) - - - - - 18201670 by Alfredo Di Napoli at 2021-07-05T07:55:12+02:00 Rename getErrorMessages Lexer import This commit renames the Lexer import in `Hyperlinker.Parser` from `getErrorMessages` to `getPsErrorMessages` to eliminate the ambiguity with the `getErrorMessages` function defined in `GHC.Types.Error`. - - - - - 23173ca3 by Ben Gamari at 2021-07-07T11:31:44-04:00 Merge pull request haskell/haddock#1413 from adinapoli/wip/adinapoli-issue-19920 Rename getErrorMessages Lexer import - - - - - b3dc4ed8 by Alan Zimmerman at 2021-07-28T22:30:59+01:00 EPA: match changes from GHC T19834 (cherry picked from commit 2fec1b44e0ee7e263286709aa528b4ecb99ac6c2) - - - - - 5f177278 by Ben Gamari at 2021-08-06T01:17:37-04:00 Merge commit '2a966c8ca37' into HEAD - - - - - cdd81d08 by Marcin Szamotulski at 2021-08-08T17:19:06+02:00 coot/multiple packages (ghc-9.2) (#1418) - - - - - be0d71f1 by Marcin Szamotulski at 2021-08-16T08:46:03+02:00 coot/multiple package (ghc-head) (#1419) * FromJSON class Aeson style FromJSON class with Parsec based json parser. * doc-index.json file for multiple packages When creating haddock summary page for multiple packages render doc-index.json file using contents of all found 'doc-index.json' files. * Render doc-index.json When rendering html, render doc-index.json file independently of maybe_index_url option. doc-index.json file is useful now even if maybe_index_url is not `Nothing`. * base url option New `Flag_BaseURL` which configures from where static files are loaded (--base-url). If given and not equal "." static files are not coppied, as this indicates that they are not read from the the directory where we'd copy them. The default value is ".". - - - - - 3b09dbdf by Hécate Moonlight at 2021-10-07T23:26:03+02:00 Update GHC 9.2 to latest pre-release in CI - - - - - 7ac55417 by Zubin Duggal at 2021-10-11T12:10:19+02:00 Enable Haddock tests in GHC windows CI (#1428) * testsuite: strip windows line endings for haddock * hyperlinker: Work around double escaping (#19236) * deterministic SCC - - - - - 1cb81f25 by Andrew Lelechenko at 2021-10-12T15:23:19+02:00 haddock-library does not depend on bytestring or transformers (#1426) - - - - - a890b9aa by sheaf at 2021-10-15T22:19:42+02:00 update haddockHypsrcTest for GHC MR !6705 (#1430) - - - - - 42a55c6c by Sylvain Henry at 2021-10-15T22:20:10+02:00 Fix after PkgQual refactoring (#1429) - - - - - 91659238 by Alan Zimmerman at 2021-10-28T18:57:10+01:00 Update for changes in GHC for branch wip/az/no-srcspan-anno-instances - - - - - acf23e60 by Vladislav Zavialov at 2021-11-05T02:09:47+03:00 Do not use forall as an identifier See GHC ticket haskell/haddock#20609 - - - - - c565db0e by Krzysztof Gogolewski at 2021-11-27T02:42:35+01:00 Update after NoExtCon -> DataConCantHappen rename - - - - - b5f55590 by Artem Pelenitsyn at 2021-11-27T11:14:17+01:00 fix CI for 9.2 (#1436) - - - - - 25cd621e by Matthew Pickering at 2021-12-02T11:46:54+00:00 Update html-test for Data.List revert - - - - - 1d5ff85f by malteneuss at 2021-12-15T07:56:55+01:00 Add hint about inline link issue (#1444) - - - - - 791fde81 by Sylvain Henry at 2021-12-16T09:29:51+01:00 Bump ghc-head (#1445) * Update after NoExtCon -> DataConCantHappen rename * Update html-test for Data.List revert * Fix for new Plugins datatype Co-authored-by: Krzysztof Gogolewski <krzysztof.gogolewski at tweag.io> Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 44236317 by Sylvain Henry at 2021-12-17T09:39:00+01:00 Fix for new Plugins datatype - - - - - 80ada0fa by Hécate Moonlight at 2021-12-17T17:28:48+01:00 Remove ghc-head workflow (#1446) Contributions of GHC glue code are now done on the GHC gitlab, not in the GitHub repo anymore. - - - - - 49e171cd by Matthew Pickering at 2021-12-28T09:47:09+00:00 Remove use of ExtendedModSummary - - - - - 0e91b5ea by askeblad at 2022-01-04T09:18:35+01:00 update URLs - - - - - 9f13c212 by Hécate Moonlight at 2022-02-25T10:19:46+01:00 Fix solver for GHC 9.2 - - - - - 386751a1 by Meng Weng Wong at 2022-02-25T19:19:11+01:00 IDoc link has bitrotted; replaced with web.archive.org cache. (#1454) - - - - - d877cbe6 by Hécate Moonlight at 2022-02-25T19:21:58+01:00 Fix haddock user guide (#1456) - - - - - cc47f036 by Andrew Lelechenko at 2022-03-04T17:29:36+01:00 Allow text-2.0 in haddock-library (#1459) - - - - - 7b3685a3 by malteneuss at 2022-03-07T19:27:24+01:00 Add multi-line style hint to style section (#1460) - - - - - c51088b8 by John Ericson at 2022-03-11T16:46:26+01:00 Fix CollectPass instance to match TTG refactor Companion to GHC !7614 (https://gitlab.haskell.org/ghc/ghc/-/merge_requests/7614) - - - - - b882195b by Vladislav Zavialov at 2022-03-14T20:32:30+01:00 Link to (~) - - - - - 877349b8 by Christiaan Baaij at 2022-03-16T09:20:43+01:00 Add Haddock support for the OPAQUE pragma - - - - - 0ea22721 by askeblad at 2022-03-16T09:44:27+01:00 typos (#1464) - - - - - a6d13da1 by Matthew Pickering at 2022-03-22T13:41:17+00:00 Minimum changes needed for compilation with hi-haddock With hi-haddock, of course there is a much large refactoring of haddock which could be achieved but that is left for a future patch which can implemented at any time independently of GHC. - - - - - e7ac9129 by Matthew Pickering at 2022-03-22T21:17:50+00:00 Update test output - - - - - 6d916214 by Matthew Pickering at 2022-03-24T15:06:26+00:00 Merge branch 'wip/opaque_pragma' into 'ghc-head' Add Haddock support for the OPAQUE pragma See merge request ghc/haddock!2 - - - - - 42208183 by Steve Hart at 2022-03-25T20:43:50+01:00 Fix CI (#1467) * CI: Reinstall GHC with docs CI tests were failing because the GHC preinstalled to the CI environment does not include documentation, which is required for running the Haddock tests. This commit causes the CI workflow to reinstall GHC with docs so that tests can succeed. - - - - - 9676fd79 by Steve Hart at 2022-03-25T21:33:34+01:00 Make links in Synopsis functional again (#1458) Commit e41c1cbe9f0476997eac7b4a3f17cbc6b2262faf added a call to e.preventDefault() when handling click events that reach a toggle element. This prevents the browser from following hyperlinks within the Synopsis section when they are clicked by a user. This commit restores functioning hyperlinks within the Synopsis section by removing the call to e.preventDefault(), as it does not appear to be necessary, and removing it increases the flexibility of the details-helper code. - - - - - d1edd637 by sheaf at 2022-04-01T12:02:02+02:00 Keep track of promotion ticks in HsOpTy Keeping track of promotion ticks in HsOpTy allows us to properly pretty-print promoted constructors such as lists. - - - - - 9dcb2dfc by Jakob Brünker at 2022-04-01T15:46:22+00:00 Add support for \cases See merge request ghc/ghc!7873 - - - - - b0412ee5 by askeblad at 2022-04-06T17:47:57+02:00 spelling errors (#1471) - - - - - 6b18829b by Vladislav Zavialov at 2022-04-06T18:53:58+02:00 Rename [] to List - - - - - 2d046691 by Vladislav Zavialov at 2022-04-07T20:25:54+03:00 HsToken ConDeclGADT con_dcolon - - - - - 90b43da4 by Steve Hart at 2022-04-12T13:29:46+02:00 Parse Markdown links at beginning of line within a paragraph (#1470) * Catch Markdown links at beginning of line within paragraph Per Issue haskell/haddock#774, Markdown links were being parsed as ordinary text when they occurred at the beginning of a line other than the first line of the paragraph. This occurred because the parser was not interpreting a left square bracket as a special character that could delimit special markup. A space character was considered a special character, so, if a space occurred at the beginning of the new line, then the parser would interpret the space by itself and then continue parsing, thereby catching the Markdown link. '\n' was not treated as a special character, so the parser did not catch a Markdown link that may have followed. Note that this will allow for Markdown links that are not surrounded by spaces. For example, the following text includes a Markdown link that will be parsed: Hello, world[label](url) This is consistent with how the parser handles other types of markup. * Remove obsolete documentation hint Commit 6b9aeafddf20efc65d3725c16e3fc43a20aac343 should eliminate the need for the workaround suggested in the documentation. - - - - - 5b08312d by Hécate Moonlight at 2022-04-12T13:36:38+02:00 Force ghc-9.2 in the cabal.project - - - - - 0d0ea349 by dependabot[bot] at 2022-04-12T13:57:41+02:00 Bump path-parse from 1.0.5 to 1.0.7 in /haddock-api/resources/html (#1469) Bumps [path-parse](https://github.com/jbgutierrez/path-parse) from 1.0.5 to 1.0.7. - [Release notes](https://github.com/jbgutierrez/path-parse/releases) - [Commits](https://github.com/jbgutierrez/path-parse/commits/v1.0.7) --- updated-dependencies: - dependency-name: path-parse dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 2b9fc65e by dependabot[bot] at 2022-04-12T13:57:54+02:00 Bump copy-props from 2.0.4 to 2.0.5 in /haddock-api/resources/html (#1468) Bumps [copy-props](https://github.com/gulpjs/copy-props) from 2.0.4 to 2.0.5. - [Release notes](https://github.com/gulpjs/copy-props/releases) - [Changelog](https://github.com/gulpjs/copy-props/blob/master/CHANGELOG.md) - [Commits](https://github.com/gulpjs/copy-props/compare/2.0.4...2.0.5) --- updated-dependencies: - dependency-name: copy-props dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - ea98a6fb by Ben Gamari at 2022-04-23T22:54:37-04:00 Update for GHC 9.4 - - - - - 9e11864f by Ben Gamari at 2022-04-25T16:24:31-04:00 Merge remote-tracking branch 'upstream/ghc-9.2' into ghc-head - - - - - f83cc506 by Ben Gamari at 2022-04-25T17:00:25-04:00 Bump ghc version to 9.5 - - - - - e01c2e7d by Ben Gamari at 2022-04-28T16:19:04-04:00 Revert "Bump ghc-head (#1445)" This reverts commit b29a78ef6926101338f62e84f456dac8659dc9d2. This should not have been merged. - - - - - a2b5ee8c by Ben Gamari at 2022-04-28T16:19:24-04:00 Merge commit '2627a86c' into ghc-head - - - - - 0c6fe4f9 by Ben Gamari at 2022-04-29T10:05:54-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-9.4 - - - - - b6e5cb0a by Ben Gamari at 2022-04-29T11:46:06-04:00 Revert "HsToken ConDeclGADT con_dcolon" This reverts commit 24208496649a02d5f87373052c430ea4a97842c5. - - - - - 15a62888 by Ben Gamari at 2022-04-29T15:12:55-04:00 Bump base upper bound - - - - - 165b9031 by Ben Gamari at 2022-04-29T23:58:38-04:00 Update test output - - - - - e0c3e5da by Phil de Joux at 2022-05-02T14:46:38+02:00 Add hlint action .hlint.yaml with ignores & CPP. (#1475) - - - - - ead1158d by Raphael Das Gupta at 2022-05-02T14:46:48+02:00 fix grammar in docs: "can the" → "can be" (#1477) - - - - - cff97944 by Ben Gamari at 2022-05-02T18:38:56-04:00 Allow base-4.17 - - - - - e4ecb201 by Phil de Joux at 2022-05-03T13:14:55+02:00 Remove unused imports that GHC warned about. (#1480) - - - - - 222890b1 by Phil de Joux at 2022-05-03T13:15:46+02:00 Follow hlint suggestion to remove redundant bang. (#1479) - - - - - 058b671f by Phil de Joux at 2022-05-03T13:34:04+02:00 Follow hlint, remove language pragmas in libs. (#1478) - - - - - 0a645049 by Ben Simms at 2022-05-03T14:19:24+02:00 Keep track of ordered list indexes and render them (#1407) * Keep track of ordered list indexes and render them * Rename some identifiers to clarify - - - - - f0433304 by Norman Ramsey at 2022-05-04T15:13:34-04:00 update for changes in GHC API - - - - - 3740cf71 by Emily Martins at 2022-05-06T18:23:48+02:00 Add link to the readthedocs in cabal description to show on hackage. (cherry picked from commit 52e2d40d47295c02d3181aac0c53028e730f1e3b) - - - - - 5d754f1e by Hécate Moonlight at 2022-05-06T18:44:57+02:00 remove Bug873 - - - - - 968fc267 by Hécate Moonlight at 2022-05-06T18:48:28+02:00 Ignore "Use second" HLint suggestion. It increases laziness. - - - - - 02d14e97 by Jade Lovelace at 2022-05-07T17:42:08+02:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` - - - - - b22b87ed by Artem Pelenitsyn at 2022-05-08T16:19:47+02:00 fix parsing trailing quotes in backticked identifiers (#1408) (#1483) - - - - - 80ae107b by Alex Biehl at 2022-05-08T16:37:16+02:00 Fix "Defined by not used" error (cherry picked from commit 6e02a620a26c3a44f98675dd1b93b08070c36c0a) - - - - - 4c838e84 by Hécate Moonlight at 2022-05-08T16:37:16+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - fc9827b4 by Hécate Moonlight at 2022-05-08T16:40:40+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - b153b555 by Xia Li-yao at 2022-05-20T17:52:42+02:00 Hide synopsis from search when hidden (#1486) Fix haskell/haddock#1451 - - - - - f3e38b85 by Marcin Szamotulski at 2022-05-21T23:32:31+02:00 Allow to hide interfaces when rendering multiple components (#1487) This is useful when one wishes to `--gen-contents` when rendering multiple components, but one does not want to render all modules. This is in particular useful when adding base package. - - - - - f942863b by Marcin Szamotulski at 2022-05-24T08:29:59+02:00 Check if doc-index.json exists before reading it (#1488) - - - - - 31e92982 by Marcin Szamotulski at 2022-05-25T16:22:13+02:00 Version bump 2.26.1 (#1489) * Version bump 2.26.1 We extended format accepted by `--read-interface` option, which requires updating the minor version. * Update documentation of --read-interface option - - - - - 7cc873e0 by sheaf at 2022-05-25T16:42:31+02:00 Updated HaddockHypsrcTest output for record update changes (MR !7981) - - - - - cd196942 by Marcin Szamotulski at 2022-05-25T20:28:47+02:00 Use visibility to decide which interfaces are included in quickjump (#1490) This is also consistent with how html index is build. See haskell/cabal#7669 for rationale behind this decision. - - - - - 00c713c5 by Hécate Moonlight at 2022-05-26T17:09:15+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 2f3039f1 by Hécate Moonlight at 2022-05-26T17:10:59+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 63a5650c by romes at 2022-05-31T12:43:22+01:00 TTG: Match new GHC AST - - - - - dd7d1617 by romes at 2022-06-02T16:11:00+01:00 Update for IE changes in !8228 - - - - - c23aaab7 by cydparser at 2022-06-06T08:48:14+02:00 Fix and improve CI (#1495) * Pin GHC version before creating the freeze file * Use newest action versions * Improve caching * Avoid unnecessarily reinstalling GHC * Use GHC 9.2.2 for CI Co-authored-by: Cyd Wise <cwise at tripshot.com> - - - - - c156fa77 by Hécate Moonlight at 2022-06-06T11:59:35+02:00 Add Mergify configuration (#1496) - - - - - 2dba4188 by Hécate Moonlight at 2022-06-06T16:12:50+02:00 Bump haddock's version in cabal file to 2.26.1 (#1497) - - - - - d7d4b8b9 by Marcin Szamotulski at 2022-06-07T06:09:40+00:00 Render module tree per package in the content page (#1492) * Render module tree per package in the content page When rendering content page for multiple packages it is useful to split the module tree per package. Package names in this patch are inferred from haddock's interface file names. * Write PackageInfo into interface file To keep interface file format backward compatible, instead of using `Binary` instance for `InterfaceFile` we introduce functions to serialise and deserialise, which depends on the interface file version. - - - - - 77765665 by Mike Pilgrem at 2022-06-12T21:57:19+01:00 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - b0e079b0 by mergify[bot] at 2022-06-13T11:49:37+00:00 Merge pull request haskell/haddock#1108 from mpilgrem/fix783 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - 6c0292b1 by Hécate Moonlight at 2022-06-21T17:21:08+02:00 Update the contribution guide - - - - - e413b9fa by dependabot[bot] at 2022-06-21T23:38:19+02:00 Bump shell-quote from 1.6.1 to 1.7.3 in /haddock-api/resources/html (#1500) Bumps [shell-quote](https://github.com/substack/node-shell-quote) from 1.6.1 to 1.7.3. - [Release notes](https://github.com/substack/node-shell-quote/releases) - [Changelog](https://github.com/substack/node-shell-quote/blob/master/CHANGELOG.md) - [Commits](https://github.com/substack/node-shell-quote/compare/1.6.1...1.7.3) --- updated-dependencies: - dependency-name: shell-quote dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 29d0ef70 by romes at 2022-07-06T11:29:39+02:00 TTG: AST Updates for !8308 - - - - - 1bae7c87 by Alan Zimmerman at 2022-07-06T22:50:43+01:00 Match GHC changes for T21805 This brings in a newtype for FieldLabelString - - - - - 6fe8b988 by Phil de Joux at 2022-07-16T20:54:26+00:00 Bump hlint version to 3.4.1, the version with counts. (#1503) Redo the counts available with the --default option. - - - - - 48fb43af by Phil de Joux at 2022-07-19T09:32:55+02:00 Follow hlint suggestion: unused LANGUAGE pragma. (#1504) * Follow hlint suggestion: unused LANGUAGE pragma. * Ignore within modules to pass linting and pass tests. - - - - - c1cf1fa7 by Phil de Joux at 2022-07-24T13:45:59+02:00 Follow hlint suggestion: redundant $. (#1505) * Follow hlint suggestion: redundant $. * Remove $ and surplus blank lines in Operators. - - - - - 74777eb2 by Jade Lovelace at 2022-07-29T11:02:41+01:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` (cherry picked from commit ab53ccf089ea703b767581ac14be0f6c78a7678a) - - - - - faa4cfcf by Hécate Moonlight at 2022-07-29T20:31:20+02:00 Merge pull request haskell/haddock#1516 from duog/9-4-backport-fix-hyperlinks Backport 9-4: Fix hyperlinks to external items and modules (#1482) - - - - - 5d2450f3 by Ben Gamari at 2022-08-05T17:41:15-04:00 Merge remote-tracking branch 'origin/ghc-9.4' - - - - - 63954f73 by Ben Gamari at 2022-08-05T19:08:36-04:00 Clean up build and testsuite for GHC 9.4 - - - - - d4568cb8 by Hécate Moonlight at 2022-08-05T19:10:49-04:00 Bump the versions - - - - - 505583a4 by Ben Gamari at 2022-08-06T13:58:27-04:00 Merge pull request haskell/haddock#1518 from bgamari/wip/ghc-9.4-merge Merge GHC 9.4 into `main` - - - - - 5706f6a4 by Ben Gamari at 2022-08-06T22:57:21-04:00 html-test: Testsuite changes for GHC 9.4.1 - - - - - 5f2a45a2 by Ben Gamari at 2022-08-15T14:33:05-04:00 doc: Fix a few minor ReST issues Sphinx was complaining about too-short title underlines. - - - - - 220e6410 by Ben Gamari at 2022-08-15T14:41:24-04:00 Merge branch 'main' into ghc-head - - - - - fbeb1b02 by Ben Gamari at 2022-08-15T14:45:16-04:00 Updates for GHC 9.5 - - - - - eee562eb by Vladislav Zavialov at 2022-08-15T14:46:13-04:00 HsToken ConDeclGADT con_dcolon - - - - - c5f073db by Ben Gamari at 2022-08-15T16:55:35-04:00 Updates for GHC 9.5 - - - - - 3f7ab242 by Vladislav Zavialov at 2022-08-15T16:55:35-04:00 HsToken ConDeclGADT con_dcolon - - - - - a18e473d by Ben Gamari at 2022-08-16T08:35:19-04:00 Merge branch 'wip/ghc-head-bump' into ghc-head - - - - - af0ff3a4 by M Farkas-Dyck at 2022-09-15T21:16:05+00:00 Disuse `mapLoc`. - - - - - a748fc38 by Matthew Farkas-Dyck at 2022-09-17T10:44:18+00:00 Scrub partiality about `NewOrData`. - - - - - 2758fb6c by John Ericson at 2022-09-18T03:27:37+02:00 Test output changed because of change to `base` Spooky, but I guess that is intended? - - - - - a7eec128 by Torsten Schmits at 2022-09-21T11:06:55+02:00 update tests for the move of tuples to GHC.Tuple.Prim - - - - - 461e7b9d by Ross Paterson at 2022-09-24T22:01:25+00:00 match implementation of GHC proposal haskell/haddock#106 (Define Kinds Without Promotion) - - - - - f7fd77ef by sheaf at 2022-10-17T14:53:01+02:00 Update Haddock for GHC MR !8563 (configuration of diagnostics) - - - - - ff6f2228 by M Farkas-Dyck at 2022-10-20T16:15:51-04:00 CoreToStg: purge `DynFlags`. - - - - - 1ebd521f by Matthew Pickering at 2022-10-20T16:16:27-04:00 ci: Make fat014 test robust For some reason I implemented this as a makefile test rather than a ghci_script test. Hopefully making it a ghci_script test makes it more robust. Fixes #22313 - - - - - 8cd6f435 by Curran McConnell at 2022-10-21T02:58:01-04:00 remove a no-warn directive from GHC.Cmm.ContFlowOpt This patch is motivated by the desire to remove the {-# OPTIONS_GHC -fno-warn-incomplete-patterns #-} directive at the top of GHC.Cmm.ContFlowOpt. (Based on the text in this coding standards doc, I understand it's a goal of the project to remove such directives.) I chose this task because I'm a new contributor to GHC, and it seemed like a good way to get acquainted with the patching process. In order to address the warning that arose when I removed the no-warn directive, I added a case to removeUnreachableBlocksProc to handle the CmmData constructor. Clearly, since this partial function has not been erroring out in the wild, its inputs are always in practice wrapped by the CmmProc constructor. Therefore the CmmData case is handled by a precise panic (which is an improvement over the partial pattern match from before). - - - - - a2af7c4c by Nicolas Trangez at 2022-10-21T02:58:39-04:00 build: get rid of `HAVE_TIME_H` As advertized by `autoreconf`: > All current systems provide time.h; it need not be checked for. Hence, remove the check for it in `configure.ac` and remove conditional inclusion of the header in `HAVE_TIME_H` blocks where applicable. The `time.h` header was being included in various source files without a `HAVE_TIME_H` guard already anyway. - - - - - 25cdc630 by Nicolas Trangez at 2022-10-21T02:58:39-04:00 rts: remove use of `TIME_WITH_SYS_TIME` `autoreconf` will insert an `m4_warning` when the obsolescent `AC_HEADER_TIME` macro is used: > Update your code to rely only on HAVE_SYS_TIME_H, > then remove this warning and the obsolete code below it. > All current systems provide time.h; it need not be checked for. > Not all systems provide sys/time.h, but those that do, all allow > you to include it and time.h simultaneously. Presence of `sys/time.h` was already checked in an earlier `AC_CHECK_HEADERS` invocation, so `AC_HEADER_TIME` can be dropped and guards relying on `TIME_WITH_SYS_TIME` can be reworked to (unconditionally) include `time.h` and include `sys/time.h` based on `HAVE_SYS_TIME_H`. Note the documentation of `AC_HEADER_TIME` in (at least) Autoconf 2.67 says > This macro is obsolescent, as current systems can include both files > when they exist. New programs need not use this macro. - - - - - 1fe7921c by Eric Lindblad at 2022-10-21T02:59:21-04:00 runhaskell - - - - - e3b3986e by David Feuer at 2022-10-21T03:00:00-04:00 Document how to quote certain names with spaces Quoting a name for Template Haskell is a bit tricky if the second character of that name is a single quote. The User's Guide falsely claimed that it was impossible. Document how to do it. Fixes #22236 - - - - - 0eba81e8 by Krzysztof Gogolewski at 2022-10-21T03:00:00-04:00 Fix syntax - - - - - a4dbd102 by Ben Gamari at 2022-10-21T09:11:12-04:00 Fix manifest filename when writing Windows .rc files As noted in #12971, we previously used `show` which resulted in inappropriate escaping of non-ASCII characters. - - - - - 30f0d9a9 by Ben Gamari at 2022-10-21T09:11:12-04:00 Write response files in UTF-8 on Windows This reverts the workaround introduced in f63c8ef33ec9666688163abe4ccf2d6c0428a7e7, which taught our response file logic to write response files with the `latin1` encoding to workaround `gcc`'s lacking Unicode support. This is now no longer necessary (and in fact actively unhelpful) since we rather use Clang. - - - - - b8304648 by Matthew Farkas-Dyck at 2022-10-21T09:11:56-04:00 Scrub some partiality in `GHC.Core.Opt.Simplify.Utils`. - - - - - 09ec7de2 by Teo Camarasu at 2022-10-21T13:23:07-04:00 template-haskell: Improve documentation of strictness annotation types Before it was undocumentated that DecidedLazy can be returned by reifyConStrictness for strict fields. This can happen when a field has an unlifted type or its the single field of a newtype constructor. Fixes #21380 - - - - - 88172069 by M Farkas-Dyck at 2022-10-21T13:23:51-04:00 Delete `eqExpr`, since GHC 9.4 has been released. - - - - - 86e6549e by Ömer Sinan Ağacan at 2022-10-22T07:41:30-04:00 Introduce a standard thunk for allocating strings Currently for a top-level closure in the form hey = unpackCString# x we generate code like this: Main.hey_entry() // [R1] { info_tbls: [(c2T4, label: Main.hey_info rep: HeapRep static { Thunk } srt: Nothing)] stack_info: arg_space: 8 updfr_space: Just 8 } {offset c2T4: // global _rqm::P64 = R1; if ((Sp + 8) - 24 < SpLim) (likely: False) goto c2T5; else goto c2T6; c2T5: // global R1 = _rqm::P64; call (stg_gc_enter_1)(R1) args: 8, res: 0, upd: 8; c2T6: // global (_c2T1::I64) = call "ccall" arg hints: [PtrHint, PtrHint] result hints: [PtrHint] newCAF(BaseReg, _rqm::P64); if (_c2T1::I64 == 0) goto c2T3; else goto c2T2; c2T3: // global call (I64[_rqm::P64])() args: 8, res: 0, upd: 8; c2T2: // global I64[Sp - 16] = stg_bh_upd_frame_info; I64[Sp - 8] = _c2T1::I64; R2 = hey1_r2Gg_bytes; Sp = Sp - 16; call GHC.CString.unpackCString#_info(R2) args: 24, res: 0, upd: 24; } } This code is generated for every string literal. Only difference between top-level closures like this is the argument for the bytes of the string (hey1_r2Gg_bytes in the code above). With this patch we introduce a standard thunk in the RTS, called stg_MK_STRING_info, that does what `unpackCString# x` does, except it gets the bytes address from the payload. Using this, for the closure above, we generate this: Main.hey_closure" { Main.hey_closure: const stg_MK_STRING_info; const 0; // padding for indirectee const 0; // static link const 0; // saved info const hey1_r1Gg_bytes; // the payload } This is much smaller in code. Metric Decrease: T10421 T11195 T12150 T12425 T16577 T18282 T18698a T18698b Co-Authored By: Ben Gamari <ben at well-typed.com> - - - - - 1937016b by Andreas Klebinger at 2022-10-22T07:42:06-04:00 hadrian: Improve error for wrong key/value errors. - - - - - 3d3e85ab by Vladislav Zavialov at 2022-10-22T23:04:06+03:00 Class layout info - - - - - 11fe42d8 by Vladislav Zavialov at 2022-10-23T00:11:50+03:00 Class layout info (#19623) Updates the haddock submodule. - - - - - f0a90c11 by Sven Tennie at 2022-10-24T00:12:51-04:00 Pin used way for test cloneMyStack (#21977) cloneMyStack checks the order of closures on the cloned stack. This may change for different ways. Thus we limit this test to one way (normal). - - - - - 0614e74d by Aaron Allen at 2022-10-24T17:11:21+02:00 Convert Diagnostics in GHC.Tc.Gen.Splice (#20116) Replaces uses of `TcRnUnknownMessage` in `GHC.Tc.Gen.Splice` with structured diagnostics. closes #20116 - - - - - 8d2dbe2d by Andreas Klebinger at 2022-10-24T15:59:41-04:00 Improve stg lint for unboxed sums. It now properly lints cases where sums end up distributed over multiple args after unarise. Fixes #22026. - - - - - 41406da5 by Simon Peyton Jones at 2022-10-25T18:07:03-04:00 Fix binder-swap bug This patch fixes #21229 properly, by avoiding doing a binder-swap on dictionary Ids. This is pretty subtle, and explained in Note [Care with binder-swap on dictionaries]. Test is already in simplCore/should_run/T21229 This allows us to restore a feature to the specialiser that we had to revert: see Note [Specialising polymorphic dictionaries]. (This is done in a separate patch.) I also modularised things, using a new function scrutBinderSwap_maybe in all the places where we are (effectively) doing a binder-swap, notably * Simplify.Iteration.addAltUnfoldings * SpecConstr.extendCaseBndrs In Simplify.Iteration.addAltUnfoldings I also eliminated a guard Many <- idMult case_bndr because we concluded, in #22123, that it was doing no good. - - - - - 5a997e16 by Simon Peyton Jones at 2022-10-25T18:07:03-04:00 Make the specialiser handle polymorphic specialisation Ticket #13873 unexpectedly showed that a SPECIALISE pragma made a program run (a lot) slower, because less specialisation took place overall. It turned out that the specialiser was missing opportunities because of quantified type variables. It was quite easy to fix. The story is given in Note [Specialising polymorphic dictionaries] Two other minor fixes in the specialiser * There is no benefit in specialising data constructor /wrappers/. (They can appear overloaded because they are given a dictionary to store in the constructor.) Small guard in canSpecImport. * There was a buglet in the UnspecArg case of specHeader, in the case where there is a dead binder. We need a LitRubbish filler for the specUnfolding stuff. I expanded Note [Drop dead args from specialisations] to explain. There is a 4% increase in compile time for T15164, because we generate more specialised code. This seems OK. Metric Increase: T15164 - - - - - 7f203d00 by Sylvain Henry at 2022-10-25T18:07:43-04:00 Numeric exceptions: replace FFI calls with primops ghc-bignum needs a way to raise numerical exceptions defined in base package. At the time we used FFI calls into primops defined in the RTS. These FFI calls had to be wrapped into hacky bottoming functions because "foreign import prim" syntax doesn't support giving a bottoming demand to the foreign call (cf #16929). These hacky wrapper functions trip up the JavaScript backend (#21078) because they are polymorphic in their return type. This commit replaces them with primops very similar to raise# but raising predefined exceptions. - - - - - 0988a23d by Sylvain Henry at 2022-10-25T18:08:24-04:00 Enable popcount rewrite rule when cross-compiling The comment applies only when host's word size < target's word size. So we can relax the guard. - - - - - a2f53ac8 by Sylvain Henry at 2022-10-25T18:09:05-04:00 Add GHC.SysTools.Cpp module Move doCpp out of the driver to be able to use it in the upcoming JS backend. - - - - - 1fd7f201 by Ben Gamari at 2022-10-25T18:09:42-04:00 llvm-targets: Add datalayouts for big-endian AArch64 targets Fixes #22311. Thanks to @zeldin for the patch. - - - - - f5a486eb by Krzysztof Gogolewski at 2022-10-25T18:10:19-04:00 Cleanup String/FastString conversions Remove unused mkPtrString and isUnderscoreFS. We no longer use mkPtrString since 1d03d8bef96. Remove unnecessary conversions between FastString and String and back. - - - - - cbde4cb0 by Simon Peyton Jones at 2022-10-25T23:19:18+01:00 Adapt to Constraint-vs-Type See haskell/haddock#21623 and !8750 - - - - - f7bfb40c by Ryan Scott at 2022-10-26T00:01:24-04:00 Broaden the in-scope sets for liftEnvSubst and composeTCvSubst This patch fixes two distinct (but closely related) buglets that were uncovered in #22235: * `liftEnvSubst` used an empty in-scope set, which was not wide enough to cover the variables in the range of the substitution. This patch fixes this by populating the in-scope set from the free variables in the range of the substitution. * `composeTCvSubst` applied the first substitution argument to the range of the second substitution argument, but the first substitution's in-scope set was not wide enough to cover the range of the second substutition. We similarly fix this issue in this patch by widening the first substitution's in-scope set before applying it. Fixes #22235. - - - - - 0270cc54 by Vladislav Zavialov at 2022-10-26T00:02:01-04:00 Introduce TcRnWithHsDocContext (#22346) Before this patch, GHC used withHsDocContext to attach an HsDocContext to an error message: addErr $ mkTcRnUnknownMessage $ mkPlainError noHints (withHsDocContext ctxt msg) The problem with this approach is that it only works with TcRnUnknownMessage. But could we attach an HsDocContext to a structured error message in a generic way? This patch solves the problem by introducing a new constructor to TcRnMessage: data TcRnMessage where ... TcRnWithHsDocContext :: !HsDocContext -> !TcRnMessage -> TcRnMessage ... - - - - - 9ab31f42 by Sylvain Henry at 2022-10-26T09:32:20+02:00 Testsuite: more precise test options Necessary for newer cross-compiling backends (JS, Wasm) that don't support TH yet. - - - - - f60a1a62 by Vladislav Zavialov at 2022-10-26T12:17:14-04:00 Use TcRnVDQInTermType in noNestedForallsContextsErr (#20115) When faced with VDQ in the type of a term, GHC generates the following error message: Illegal visible, dependent quantification in the type of a term (GHC does not yet support this) Prior to this patch, there were two ways this message could have been generated and represented: 1. with the dedicated constructor TcRnVDQInTermType (see check_type in GHC.Tc.Validity) 2. with the transitional constructor TcRnUnknownMessage (see noNestedForallsContextsErr in GHC.Rename.Utils) Not only this led to duplication of code generating the final SDoc, it also made it tricky to track the origin of the error message. This patch fixes the problem by using TcRnVDQInTermType exclusively. - - - - - 223e159d by Owen Shepherd at 2022-10-27T13:54:33-04:00 Remove source location information from interface files This change aims to minimize source location information leaking into interface files, which makes ABI hashes dependent on the build location. The `Binary (Located a)` instance has been removed completely. It seems that the HIE interface still needs the ability to serialize SrcSpans, but by wrapping the instances, it should be a lot more difficult to inadvertently add source location information. - - - - - 22e3deb9 by Simon Peyton Jones at 2022-10-27T13:55:37-04:00 Add missing dict binds to specialiser I had forgotten to add the auxiliary dict bindings to the /unfolding/ of a specialised function. This caused #22358, which reports failures when compiling Hackage packages fixed-vector indexed-traversable Regression test T22357 is snarfed from indexed-traversable - - - - - a8ed36f9 by Evan Relf at 2022-10-27T13:56:36-04:00 Fix broken link to `async` package - - - - - 750846cd by Zubin Duggal at 2022-10-28T00:49:22-04:00 Pass correct package db when testing stage1. It used to pick the db for stage-2 which obviously didn't work. - - - - - ad612f55 by Krzysztof Gogolewski at 2022-10-28T00:50:00-04:00 Minor SDoc-related cleanup * Rename pprCLabel to pprCLabelStyle, and use the name pprCLabel for a function using CStyle (analogous to pprAsmLabel) * Move LabelStyle to the CLabel module, it no longer needs to be in Outputable. * Move calls to 'text' right next to literals, to make sure the text/str rule is triggered. * Remove FastString/String roundtrip in Tc.Deriv.Generate * Introduce showSDocForUser', which abstracts over a pattern in GHCi.UI - - - - - c2872f3f by Bryan Richter at 2022-10-28T11:36:34+03:00 CI: Don't run lint-submods on nightly Fixes #22325 - - - - - 270037fa by Hécate Moonlight at 2022-10-28T19:46:12-04:00 Start the deprecation process for GHC.Pack - - - - - d45d8cb3 by M Farkas-Dyck at 2022-11-01T12:47:21-04:00 Drop a kludge for binutils<2.17, which is now over 10 years old. - - - - - 8ee8b418 by Nicolas Trangez at 2022-11-01T12:47:58-04:00 rts: `name` argument of `createOSThread` can be `const` Since we don't intend to ever change the incoming string, declare this to be true. Also, in the POSIX implementation, the argument is no longer `STG_UNUSED` (since ee0deb8054da2a597fc5624469b4c44fd769ada2) in any code path. See: https://gitlab.haskell.org/ghc/ghc/-/commit/ee0deb8054da2a597fc5624469b4c44fd769ada2#note_460080 - - - - - 13b5f102 by Nicolas Trangez at 2022-11-01T12:47:58-04:00 rts: fix lifetime of `start_thread`s `name` value Since, unlike the code in ee0deb8054da2^, usage of the `name` value passed to `createOSThread` now outlives said function's lifetime, and could hence be released by the caller by the time the new thread runs `start_thread`, it needs to be copied. See: https://gitlab.haskell.org/ghc/ghc/-/commit/ee0deb8054da2a597fc5624469b4c44fd769ada2#note_460080 See: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/9066 - - - - - edd175c9 by Nicolas Trangez at 2022-11-01T12:47:58-04:00 rts: fix OS thread naming in ticker Since ee0deb805, the use of `pthread_setname_np` on Darwin was fixed when invoking `createOSThread`. However, the 'ticker' has some thread-creation code which doesn't rely on `createOSThread`, yet also uses `pthread_setname_np`. This patch enforces all thread creation to go through a single function, which uses the (correct) thread-naming code introduced in ee0deb805. See: https://gitlab.haskell.org/ghc/ghc/-/commit/ee0deb8054da2a597fc5624469b4c44fd769ada2 See: https://gitlab.haskell.org/ghc/ghc/-/issues/22206 See: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/9066 - - - - - b7a00113 by Krzysztof Gogolewski at 2022-11-01T12:48:35-04:00 Typo: rename -fwrite-if-simplfied-core to -fwrite-if-simplified-core - - - - - 30e625e6 by Vladislav Zavialov at 2022-11-01T12:49:10-04:00 ThToHs: fix overzealous parenthesization Before this patch, when converting from TH.Exp to LHsExpr GhcPs, the compiler inserted more parentheses than required: ((f a) (b + c)) d This was happening because the LHS of the function application was parenthesized as if it was the RHS. Now we use funPrec and appPrec appropriately and produce sensibly parenthesized expressions: f a (b + c) d I also took the opportunity to remove the special case for LamE, which was not special at all and simply duplicated code. - - - - - 0560821f by Simon Peyton Jones at 2022-11-01T12:49:47-04:00 Add accurate skolem info when quantifying Ticket #22379 revealed that skolemiseQuantifiedTyVar was dropping the passed-in skol_info on the floor when it encountered a SkolemTv. Bad! Several TyCons thereby share a single SkolemInfo on their binders, which lead to bogus error reports. - - - - - 38d19668 by Fendor at 2022-11-01T12:50:25-04:00 Expose UnitEnvGraphKey for user-code - - - - - 77e24902 by Simon Peyton Jones at 2022-11-01T12:51:00-04:00 Shrink test case for #22357 Ryan Scott offered a cut-down repro case (60 lines instead of more than 700 lines) - - - - - 4521f649 by Simon Peyton Jones at 2022-11-01T12:51:00-04:00 Add two tests for #17366 - - - - - 7108ba96 by Tom Smeding at 2022-11-01T22:33:23+01:00 Remove outdated footnote about module re-exports The footnote is invalid with GHC 9.2.4 (and possibly earlier): the described behaviour in the main text works fine. - - - - - 206c6bc7 by Hécate Moonlight at 2022-11-01T23:00:46+01:00 Merge pull request haskell/haddock#1534 from tomsmeding/patch-1 - - - - - 6b400d26 by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: introduce (and use) `STG_NORETURN` Instead of sprinkling the codebase with `GNU(C3)_ATTRIBUTE(__noreturn__)`, add a `STG_NORETURN` macro (for, basically, the same thing) similar to `STG_UNUSED` and others, and update the code to use this macro where applicable. - - - - - f9638654 by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: consistently use `STG_UNUSED` - - - - - 81a58433 by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: introduce (and use) `STG_USED` Similar to `STG_UNUSED`, have a specific macro for `__attribute__(used)`. - - - - - 41e1f748 by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: introduce (and use) `STG_MALLOC` Instead of using `GNUC3_ATTRIBUTE(__malloc__)`, provide a `STG_MALLOC` macro definition and use it instead. - - - - - 3a9a8bde by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: use `STG_UNUSED` - - - - - 9ab999de by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: specify deallocator of allocating functions This patch adds a new `STG_MALLOC1` macro (and its counterpart `STG_MALLOC2` for completeness) which allows to specify the deallocation function to be used with allocations of allocating functions, and applies it to `stg*allocBytes`. It also fixes a case where `free` was used to free up an `stgMallocBytes` allocation, found by the above change. See: https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-malloc-function-attribute See: https://gitlab.haskell.org/ghc/ghc/-/issues/22381 - - - - - 81c0c7c9 by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: use `alloc_size` attribute This patch adds the `STG_ALLOC_SIZE1` and `STG_ALLOC_SIZE2` macros which allow to set the `alloc_size` attribute on functions, when available. See: https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-alloc_005fsize-function-attribute See: https://gitlab.haskell.org/ghc/ghc/-/issues/22381 - - - - - 99a1d896 by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: add and use `STG_RETURNS_NONNULL` See: https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-returns_005fnonnull-function-attribute See: https://gitlab.haskell.org/ghc/ghc/-/issues/22381 - - - - - c235b399 by Nicolas Trangez at 2022-11-02T12:06:48-04:00 rts: tag `stgStrndup` as `STG_MALLOC` See: https://gitlab.haskell.org/ghc/ghc/-/issues/22381 - - - - - ed81b448 by Oleg Grenrus at 2022-11-02T12:07:27-04:00 Move Symbol implementation note out of public haddock - - - - - 284fd39c by Ben Gamari at 2022-11-03T01:58:54-04:00 gen-dll: Drop it Currently it is only used by the make build system, which is soon to be retired, and it has not built since 41cf758b. We may need to reintroduce it when dynamic-linking support is introduced on Windows, but we will cross that bridge once we get there. Fixes #21753. - - - - - 24f4f54f by Matthew Pickering at 2022-11-03T01:59:30-04:00 Port foundation numeric tests to GHC testsuite This commit ports the numeric tests which found a regression in GHC-9.4. https://github.com/haskell-foundation/foundation/issues/571 Included in the commit is a simple random number generator and simplified QuickCheck implementation. In future these could be factored out of this standalone file and reused as a general purpose library which could be used for other QuickCheck style tests in the testsuite. See #22282 - - - - - d51bf7bd by M Farkas-Dyck at 2022-11-03T02:00:13-04:00 git: ignore HIE files. Cleans up git status if one sets -fwrite-ide-info in hadrian/ghci. - - - - - a9fc15b1 by Matthew Pickering at 2022-11-03T02:00:49-04:00 Clarify status of bindings in WholeCoreBindings Gergo points out that these bindings are tidied, rather than prepd as the variable claims. Therefore we update the name of the variable to reflect reality and add a comment to the data type to try to erase any future confusion. Fixes #22307 - - - - - 634da448 by Andrew Lelechenko at 2022-11-03T21:25:02+00:00 Fix haddocks for GHC.IORef - - - - - 31125154 by Andreas Klebinger at 2022-11-03T23:08:09-04:00 Export pprTrace and friends from GHC.Prelude. Introduces GHC.Prelude.Basic which can be used in modules which are a dependency of the ppr code. - - - - - bdc8cbb3 by Bryan Richter at 2022-11-04T10:27:37+02:00 CI: Allow hadrian-ghc-in-ghci to run in nightlies Since lint-submods doesn't run in nightlies, hadrian-ghc-in-ghci needs to mark it as "optional" so it can run if the job doesn't exist. Fixes #22396. - - - - - 3c0e3793 by Krzysztof Gogolewski at 2022-11-05T00:29:57-04:00 Minor refactor around FastStrings Pass FastStrings to functions directly, to make sure the rule for fsLit "literal" fires. Remove SDoc indirection in GHCi.UI.Tags and GHC.Unit.Module.Graph. - - - - - e41b2f55 by Matthew Pickering at 2022-11-05T14:18:10+00:00 Bump unix submodule to 2.8.0.0 Also bumps process and ghc-boot bounds on unix. For hadrian, when cross-compiling, we add -Wwarn=unused-imports -Wwarn=unused-top-binds to validation flavour. Further fixes in unix and/or hsc2hs is needed to make it completely free of warnings; for the time being, this change is needed to unblock other cross-compilation related work. - - - - - 42938a58 by Matthew Pickering at 2022-11-05T14:18:10+00:00 Bump Win32 submodule to 2.13.4.0 Fixes #22098 - - - - - e7372bc5 by Cheng Shao at 2022-11-06T13:15:22+00:00 Bump ci-images revision ci-images has recently been updated, including changes needed for wasm32-wasi CI. - - - - - 88cb9492 by Cheng Shao at 2022-11-06T13:15:22+00:00 Bump gmp-tarballs submodule Includes a fix for wasm support, doesn't impact other targets. - - - - - 69427ce9 by Cheng Shao at 2022-11-06T13:15:22+00:00 Bump haskeline submodule Includes a fix for wasm support, doesn't impact other targets. - - - - - 5fe11fe6 by Carter Schonwald at 2022-11-07T13:22:14-05:00 bump llvm upper bound - - - - - 68f49874 by M Farkas-Dyck at 2022-11-08T12:53:55-05:00 Define `Infinite` list and use where appropriate. Also add perf test for infinite list fusion. In particular, in `GHC.Core`, often we deal with infinite lists of roles. Also in a few locations we deal with infinite lists of names. Thanks to simonpj for helping to write the Note [Fusion for `Infinite` lists]. - - - - - ce726cd2 by Ross Paterson at 2022-11-08T12:54:34-05:00 Fix TypeData issues (fixes #22315 and #22332) There were two bugs here: 1. Treating type-level constructors as PromotedDataCon doesn't always work, in particular because constructors promoted via DataKinds are called both T and 'T. (Tests T22332a, T22332b, T22315a, T22315b) Fix: guard these cases with isDataKindsPromotedDataCon. 2. Type-level constructors were sent to the code generator, producing things like constructor wrappers. (Tests T22332a, T22332b) Fix: test for them in isDataTyCon. Other changes: * changed the marking of "type data" DataCon's as suggested by SPJ. * added a test TDGADT for a type-level GADT. * comment tweaks * change tcIfaceTyCon to ignore IfaceTyConInfo, so that IfaceTyConInfo is used only for pretty printing, not for typechecking. (SPJ) - - - - - 132f8908 by Jade Lovelace at 2022-11-08T12:55:18-05:00 Clarify msum/asum documentation - - - - - bb5888c5 by Jade Lovelace at 2022-11-08T12:55:18-05:00 Add example for (<$) - - - - - 080fffa1 by Jade Lovelace at 2022-11-08T12:55:18-05:00 Document what Alternative/MonadPlus instances actually do - - - - - 92ccb8de by Giles Anderson at 2022-11-09T09:27:52-05:00 Use TcRnDiagnostic in GHC.Tc.TyCl.Instance (#20117) The following `TcRnDiagnostic` messages have been introduced: TcRnWarnUnsatisfiedMinimalDefinition TcRnMisplacedInstSig TcRnBadBootFamInstDeclErr TcRnIllegalFamilyInstance TcRnAssocInClassErr TcRnBadFamInstDecl TcRnNotOpenFamily - - - - - 90c5abd4 by Hécate Moonlight at 2022-11-09T09:28:30-05:00 GHCi tags generation phase 2 see #19884 - - - - - f9f17b68 by Simon Peyton Jones at 2022-11-10T12:20:03+00:00 Fire RULES in the Specialiser The Specialiser has, for some time, fires class-op RULES in the specialiser itself: see Note [Specialisation modulo dictionary selectors] This MR beefs it up a bit, so that it fires /all/ RULES in the specialiser, not just class-op rules. See Note [Fire rules in the specialiser] The result is a bit more specialisation; see test simplCore/should_compile/T21851_2 This pushed me into a bit of refactoring. I made a new data types GHC.Core.Rules.RuleEnv, which combines - the several source of rules (local, home-package, external) - the orphan-module dependencies in a single record for `getRules` to consult. That drove a bunch of follow-on refactoring, including allowing me to remove cr_visible_orphan_mods from the CoreReader data type. I moved some of the RuleBase/RuleEnv stuff into GHC.Core.Rule. The reorganisation in the Simplifier improve compile times a bit (geom mean -0.1%), but T9961 is an outlier Metric Decrease: T9961 - - - - - 2b3d0bee by Simon Peyton Jones at 2022-11-10T12:21:13+00:00 Make indexError work better The problem here is described at some length in Note [Boxity for bottoming functions] and Note [Reboxed crud for bottoming calls] in GHC.Core.Opt.DmdAnal. This patch adds a SPECIALISE pragma for indexError, which makes it much less vulnerable to the problem described in these Notes. (This came up in another line of work, where a small change made indexError do reboxing (in nofib/spectral/simple/table_sort) that didn't happen before my change. I've opened #22404 to document the fagility. - - - - - 399e921b by Simon Peyton Jones at 2022-11-10T12:21:14+00:00 Fix DsUselessSpecialiseForClassMethodSelector msg The error message for DsUselessSpecialiseForClassMethodSelector was just wrong (a typo in some earlier work); trivial fix - - - - - dac0682a by Sebastian Graf at 2022-11-10T21:16:01-05:00 WorkWrap: Unboxing unboxed tuples is not always useful (#22388) See Note [Unboxing through unboxed tuples]. Fixes #22388. - - - - - 1230c268 by Sebastian Graf at 2022-11-10T21:16:01-05:00 Boxity: Handle argument budget of unboxed tuples correctly (#21737) Now Budget roughly tracks the combined width of all arguments after unarisation. See the changes to `Note [Worker argument budgets]`. Fixes #21737. - - - - - 2829fd92 by Cheng Shao at 2022-11-11T00:26:54-05:00 autoconf: check getpid getuid raise This patch adds checks for getpid, getuid and raise in autoconf. These functions are absent in wasm32-wasi and thus needs to be checked. - - - - - f5dfd1b4 by Cheng Shao at 2022-11-11T00:26:55-05:00 hadrian: add -Wwarn only for cross-compiling unix - - - - - 2e6ab453 by Cheng Shao at 2022-11-11T00:26:55-05:00 hadrian: add targetSupportsThreadedRts flag This patch adds a targetSupportsThreadedRts flag to indicate whether the target supports the threaded rts at all, different from existing targetSupportsSMP that checks whether -N is supported by the RTS. All existing flavours have also been updated accordingly to respect this flags. Some targets (e.g. wasm32-wasi) does not support the threaded rts, therefore this flag is needed for the default flavours to work. It makes more sense to have proper autoconf logic to check for threading support, but for the time being, we just set the flag to False iff the target is wasm32. - - - - - 8104f6f5 by Cheng Shao at 2022-11-11T00:26:55-05:00 Fix Cmm symbol kind - - - - - b2035823 by Norman Ramsey at 2022-11-11T00:26:55-05:00 add the two key graph modules from Martin Erwig's FGL Martin Erwig's FGL (Functional Graph Library) provides an "inductive" representation of graphs. A general graph has labeled nodes and labeled edges. The key operation on a graph is to decompose it by removing one node, together with the edges that connect the node to the rest of the graph. There is also an inverse composition operation. The decomposition and composition operations make this representation of graphs exceptionally well suited to implement graph algorithms in which the graph is continually changing, as alluded to in #21259. This commit adds `GHC.Data.Graph.Inductive.Graph`, which defines the interface, and `GHC.Data.Graph.Inductive.PatriciaTree`, which provides an implementation. Both modules are taken from `fgl-5.7.0.3` on Hackage, with these changes: - Copyright and license text have been copied into the files themselves, not stored separately. - Some calls to `error` have been replaced with calls to `panic`. - Conditional-compilation support for older versions of GHC, `containers`, and `base` has been removed. - - - - - 3633a5f5 by Norman Ramsey at 2022-11-11T00:26:55-05:00 add new modules for reducibility and WebAssembly translation - - - - - df7bfef8 by Cheng Shao at 2022-11-11T00:26:55-05:00 Add support for the wasm32-wasi target tuple This patch adds the wasm32-wasi tuple support to various places in the tree: autoconf, hadrian, ghc-boot and also the compiler. The codegen logic will come in subsequent commits. - - - - - 32ae62e6 by Cheng Shao at 2022-11-11T00:26:55-05:00 deriveConstants: parse .ll output for wasm32 due to broken nm This patch makes deriveConstants emit and parse an .ll file when targeting wasm. It's a necessary workaround for broken llvm-nm on wasm, which isn't capable of reporting correct constant values when parsing an object. - - - - - 07e92c92 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: workaround cmm's improper variadic ccall breaking wasm32 typechecking Unlike other targets, wasm requires the function signature of the call site and callee to strictly match. So in Cmm, when we call a C function that actually returns a value, we need to add an _unused local variable to receive it, otherwise type error awaits. An even bigger problem is calling variadic functions like barf() and such. Cmm doesn't support CAPI calling convention yet, so calls to variadic functions just happen to work in some cases with some target's ABI. But again, it doesn't work with wasm. Fortunately, the wasm C ABI lowers varargs to a stack pointer argument, and it can be passed NULL when no other arguments are expected to be passed. So we also add the additional unused NULL arguments to those functions, so to fix wasm, while not affecting behavior on other targets. - - - - - 00124d12 by Cheng Shao at 2022-11-11T00:26:55-05:00 testsuite: correct sleep() signature in T5611 In libc, sleep() returns an integer. The ccall type signature should match the libc definition, otherwise it causes linker error on wasm. - - - - - d72466a9 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: prefer ffi_type_void over FFI_TYPE_VOID This patch uses ffi_type_void instead of FFI_TYPE_VOID in the interpreter code, since the FFI_TYPE_* macros are not available in libffi-wasm32 yet. The libffi public documentation also only mentions the lower-case ffi_type_* symbols, so we should prefer the lower-case API here. - - - - - 4d36a1d3 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: don't define RTS_USER_SIGNALS when signal.h is not present In the rts, we have a RTS_USER_SIGNALS macro, and most signal-related logic is guarded with RTS_USER_SIGNALS. This patch extends the range of code guarded with RTS_USER_SIGNALS, and define RTS_USER_SIGNALS iff signal.h is actually detected by autoconf. This is required for wasm32-wasi to work, which lacks signals. - - - - - 3f1e164f by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: use HAVE_GETPID to guard subprocess related logic We've previously added detection of getpid() in autoconf. This patch uses HAVE_GETPID to guard some subprocess related logic in the RTS. This is required for certain targets like wasm32-wasi, where there isn't a process model at all. - - - - - 50bf5e77 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: IPE.c: don't do mutex stuff when THREADED_RTS is not defined This patch adds the missing THREADED_RTS CPP guard to mutex logic in IPE.c. - - - - - ed3b3da0 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: genericRaise: use exit() instead when not HAVE_RAISE We check existence of raise() in autoconf, and here, if not HAVE_RAISE, we should use exit() instead in genericRaise. - - - - - c0ba1547 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: checkSuid: don't do it when not HAVE_GETUID When getuid() is not present, don't do checkSuid since it doesn't make sense anyway on that target. - - - - - d2d6dfd2 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: wasm32 placeholder linker This patch adds minimal placeholder linker logic for wasm32, just enough to unblock compiling rts on wasm32. RTS linker functionality is not properly implemented yet for wasm32. - - - - - 65ba3285 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: RtsStartup: chdir to PWD on wasm32 This patch adds a wasm32-specific behavior to RtsStartup logic. When the PWD environment variable is present, we chdir() to it first. The point is to workaround an issue in wasi-libc: it's currently not possible to specify the initial working directory, it always defaults to / (in the virtual filesystem mapped from some host directory). For some use cases this is sufficient, but there are some other cases (e.g. in the testsuite) where the program needs to access files outside. - - - - - 65b82542 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: no timer for wasm32 Due to the lack of threads, on wasm32 there can't be a background timer that periodically resets the context switch flag. This patch disables timer for wasm32, and also makes the scheduler default to -C0 on wasm32 to avoid starving threads. - - - - - e007586f by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: RtsSymbols: empty RTS_POSIX_ONLY_SYMBOLS for wasm32 The default RTS_POSIX_ONLY_SYMBOLS doesn't make sense on wasm32. - - - - - 0e33f667 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: Schedule: no FORKPROCESS_PRIMOP_SUPPORTED on wasm32 On wasm32 there isn't a process model at all, so no FORKPROCESS_PRIMOP_SUPPORTED. - - - - - 88bbdb31 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: LibffiAdjustor: adapt to ffi_alloc_prep_closure interface for wasm32 libffi-wasm32 only supports non-standard libffi closure api via ffi_alloc_prep_closure(). This patch implements ffi_alloc_prep_closure() via standard libffi closure api on other targets, and uses it to implement adjustor functionality. - - - - - 15138746 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: don't return memory to OS on wasm32 This patch makes the storage manager not return any memory on wasm32. The detailed reason is described in Note [Megablock allocator on wasm]. - - - - - 631af3cc by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: make flushExec a no-op on wasm32 This patch makes flushExec a no-op on wasm32, since there's no such thing as executable memory on wasm32 in the first place. - - - - - 654a3d46 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: RtsStartup: don't call resetTerminalSettings, freeThreadingResources on wasm32 This patch prevents resetTerminalSettings and freeThreadingResources to be called on wasm32, since there is no TTY or threading on wasm32 at all. - - - - - f271e7ca by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: OSThreads.h: stub types for wasm32 This patch defines stub Condition/Mutex/OSThreadId/ThreadLocalKey types for wasm32, just enough to unblock compiling RTS. Any threading-related functionality has been patched to be disabled on wasm32. - - - - - a6ac67b0 by Cheng Shao at 2022-11-11T00:26:55-05:00 Add register mapping for wasm32 This patch adds register mapping logic for wasm32. See Note [Register mapping on WebAssembly] in wasm32 NCG for more description. - - - - - d7b33982 by Cheng Shao at 2022-11-11T00:26:55-05:00 rts: wasm32 specific logic This patch adds the rest of wasm32 specific logic in rts. - - - - - 7f59b0f3 by Cheng Shao at 2022-11-11T00:26:55-05:00 base: fall back to using monotonic clock to emulate cputime on wasm32 On wasm32, we have to fall back to using monotonic clock to emulate cputime, since there's no native support for cputime as a clock id. - - - - - 5fcbae0b by Cheng Shao at 2022-11-11T00:26:55-05:00 base: more autoconf checks for wasm32 This patch adds more autoconf checks to base, since those functions and headers may exist on other POSIX systems but don't exist on wasm32. - - - - - 00a9359f by Cheng Shao at 2022-11-11T00:26:55-05:00 base: avoid using unsupported posix functionality on wasm32 This base patch avoids using unsupported posix functionality on wasm32. - - - - - 34b8f611 by Cheng Shao at 2022-11-11T00:26:55-05:00 autoconf: set CrossCompiling=YES in cross bindist configure This patch fixes the bindist autoconf logic to properly set CrossCompiling=YES when it's a cross GHC bindist. - - - - - 5ebeaa45 by Cheng Shao at 2022-11-11T00:26:55-05:00 compiler: add util functions for UniqFM and UniqMap This patch adds addToUFM_L (backed by insertLookupWithKey), addToUniqMap_L and intersectUniqMap_C. These UniqFM/UniqMap util functions are used by the wasm32 NCG. - - - - - 177c56c1 by Cheng Shao at 2022-11-11T00:26:55-05:00 driver: avoid -Wl,--no-as-needed for wasm32 The driver used to pass -Wl,--no-as-needed for LLD linking. This is actually only supported for ELF targets, and must be avoided when linking for wasm32. - - - - - 06f01c74 by Cheng Shao at 2022-11-11T00:26:55-05:00 compiler: allow big arith for wasm32 This patch enables Cmm big arithmetic on wasm32, since 64-bit arithmetic can be efficiently lowered to wasm32 opcodes. - - - - - df6bb112 by Cheng Shao at 2022-11-11T00:26:55-05:00 driver: pass -Wa,--no-type-check for wasm32 when runAsPhase This patch passes -Wa,--no-type-check for wasm32 when compiling assembly. See the added note for more detailed explanation. - - - - - c1fe4ab6 by Cheng Shao at 2022-11-11T00:26:55-05:00 compiler: enforce cmm switch planning for wasm32 This patch forcibly enable Cmm switch planning for wasm32, since otherwise the switch tables we generate may exceed the br_table maximum allowed size. - - - - - a8adc71e by Cheng Shao at 2022-11-11T00:26:55-05:00 compiler: annotate CmmFileEmbed with blob length This patch adds the blob length field to CmmFileEmbed. The wasm32 NCG needs to know the precise size of each data segment. - - - - - 36340328 by Cheng Shao at 2022-11-11T00:26:55-05:00 compiler: wasm32 NCG This patch adds the wasm32 NCG. - - - - - 435f42ea by Cheng Shao at 2022-11-11T00:26:55-05:00 ci: add wasm32-wasi release bindist job - - - - - d8262fdc by Cheng Shao at 2022-11-11T00:26:55-05:00 ci: add a stronger test for cross bindists This commit adds a simple GHC API program that parses and reprints the original hello world program used for basic testing of cross bindists. Before there's full cross-compilation support in the test suite driver, this provides better coverage than the original test. - - - - - 8e6ae882 by Cheng Shao at 2022-11-11T00:26:55-05:00 CODEOWNERS: add wasm-specific maintainers - - - - - 707d5651 by Zubin Duggal at 2022-11-11T00:27:31-05:00 Clarify that LLVM upper bound is non-inclusive during configure (#22411) - - - - - 430eccef by Ben Gamari at 2022-11-11T13:16:45-05:00 rts: Check for program_invocation_short_name via autoconf Instead of assuming support on all Linuxes. - - - - - 6dab0046 by Matthew Pickering at 2022-11-11T13:17:22-05:00 driver: Fix -fdefer-diagnostics flag The `withDeferredDiagnostics` wrapper wasn't doing anything because the session it was modifying wasn't used in hsc_env. Therefore the fix is simple, just push the `getSession` call into the scope of `withDeferredDiagnostics`. Fixes #22391 - - - - - d0c691b6 by Simon Peyton Jones at 2022-11-11T13:18:07-05:00 Add a fast path for data constructor workers See Note [Fast path for data constructors] in GHC.Core.Opt.Simplify.Iteration This bypasses lots of expensive logic, in the special case of applications of data constructors. It is a surprisingly worthwhile improvement, as you can see in the figures below. Metrics: compile_time/bytes allocated ------------------------------------------------ CoOpt_Read(normal) -2.0% CoOpt_Singletons(normal) -2.0% ManyConstructors(normal) -1.3% T10421(normal) -1.9% GOOD T10421a(normal) -1.5% T10858(normal) -1.6% T11545(normal) -1.7% T12234(optasm) -1.3% T12425(optasm) -1.9% GOOD T13035(normal) -1.0% GOOD T13056(optasm) -1.8% T13253(normal) -3.3% GOOD T15164(normal) -1.7% T15304(normal) -3.4% T15630(normal) -2.8% T16577(normal) -4.3% GOOD T17096(normal) -1.1% T17516(normal) -3.1% T18282(normal) -1.9% T18304(normal) -1.2% T18698a(normal) -1.2% GOOD T18698b(normal) -1.5% GOOD T18923(normal) -1.3% T1969(normal) -1.3% GOOD T19695(normal) -4.4% GOOD T21839c(normal) -2.7% GOOD T21839r(normal) -2.7% GOOD T4801(normal) -3.8% GOOD T5642(normal) -3.1% GOOD T6048(optasm) -2.5% GOOD T9020(optasm) -2.7% GOOD T9630(normal) -2.1% GOOD T9961(normal) -11.7% GOOD WWRec(normal) -1.0% geo. mean -1.1% minimum -11.7% maximum +0.1% Metric Decrease: T10421 T12425 T13035 T13253 T16577 T18698a T18698b T1969 T19695 T21839c T21839r T4801 T5642 T6048 T9020 T9630 T9961 - - - - - 3c37d30b by Krzysztof Gogolewski at 2022-11-11T19:18:39+01:00 Use a more efficient printer for code generation (#21853) The changes in `GHC.Utils.Outputable` are the bulk of the patch and drive the rest. The types `HLine` and `HDoc` in Outputable can be used instead of `SDoc` and support printing directly to a handle with `bPutHDoc`. See Note [SDoc versus HDoc] and Note [HLine versus HDoc]. The classes `IsLine` and `IsDoc` are used to make the existing code polymorphic over `HLine`/`HDoc` and `SDoc`. This is done for X86, PPC, AArch64, DWARF and dependencies (printing module names, labels etc.). Co-authored-by: Alexis King <lexi.lambda at gmail.com> Metric Decrease: CoOpt_Read ManyAlternatives ManyConstructors T10421 T12425 T12707 T13035 T13056 T13253 T13379 T18140 T18282 T18698a T18698b T1969 T20049 T21839c T21839r T3064 T3294 T4801 T5321FD T5321Fun T5631 T6048 T783 T9198 T9233 - - - - - 6b92b47f by Matthew Craven at 2022-11-11T18:32:14-05:00 Weaken wrinkle 1 of Note [Scrutinee Constant Folding] Fixes #22375. Co-authored-by: Simon Peyton Jones <simon.peytonjones at gmail.com> - - - - - 154c70f6 by Simon Peyton Jones at 2022-11-11T23:40:10+00:00 Fix fragile RULE setup in GHC.Float In testing my type-vs-constraint patch I found that the handling of Natural literals was very fragile -- and I somehow tripped that fragility in my work. So this patch fixes the fragility. See Note [realToFrac natural-to-float] This made a big (9%) difference in one existing test in perf/should_run/T1-359 Metric Decrease: T10359 - - - - - 778c6adc by Simon Peyton Jones at 2022-11-11T23:40:10+00:00 Type vs Constraint: finally nailed This big patch addresses the rats-nest of issues that have plagued us for years, about the relationship between Type and Constraint. See #11715/#21623. The main payload of the patch is: * To introduce CONSTRAINT :: RuntimeRep -> Type * To make TYPE and CONSTRAINT distinct throughout the compiler Two overview Notes in GHC.Builtin.Types.Prim * Note [TYPE and CONSTRAINT] * Note [Type and Constraint are not apart] This is the main complication. The specifics * New primitive types (GHC.Builtin.Types.Prim) - CONSTRAINT - ctArrowTyCon (=>) - tcArrowTyCon (-=>) - ccArrowTyCon (==>) - funTyCon FUN -- Not new See Note [Function type constructors and FunTy] and Note [TYPE and CONSTRAINT] * GHC.Builtin.Types: - New type Constraint = CONSTRAINT LiftedRep - I also stopped nonEmptyTyCon being built-in; it only needs to be wired-in * Exploit the fact that Type and Constraint are distinct throughout GHC - Get rid of tcView in favour of coreView. - Many tcXX functions become XX functions. e.g. tcGetCastedTyVar --> getCastedTyVar * Kill off Note [ForAllTy and typechecker equality], in (old) GHC.Tc.Solver.Canonical. It said that typechecker-equality should ignore the specified/inferred distinction when comparein two ForAllTys. But that wsa only weakly supported and (worse) implies that we need a separate typechecker equality, different from core equality. No no no. * GHC.Core.TyCon: kill off FunTyCon in data TyCon. There was no need for it, and anyway now we have four of them! * GHC.Core.TyCo.Rep: add two FunTyFlags to FunCo See Note [FunCo] in that module. * GHC.Core.Type. Lots and lots of changes driven by adding CONSTRAINT. The key new function is sORTKind_maybe; most other changes are built on top of that. See also `funTyConAppTy_maybe` and `tyConAppFun_maybe`. * Fix a longstanding bug in GHC.Core.Type.typeKind, and Core Lint, in kinding ForAllTys. See new tules (FORALL1) and (FORALL2) in GHC.Core.Type. (The bug was that before (forall (cv::t1 ~# t2). blah), where blah::TYPE IntRep, would get kind (TYPE IntRep), but it should be (TYPE LiftedRep). See Note [Kinding rules for types] in GHC.Core.Type. * GHC.Core.TyCo.Compare is a new module in which we do eqType and cmpType. Of course, no tcEqType any more. * GHC.Core.TyCo.FVs. I moved some free-var-like function into this module: tyConsOfType, visVarsOfType, and occCheckExpand. Refactoring only. * GHC.Builtin.Types. Compiletely re-engineer boxingDataCon_maybe to have one for each /RuntimeRep/, rather than one for each /Type/. This dramatically widens the range of types we can auto-box. See Note [Boxing constructors] in GHC.Builtin.Types The boxing types themselves are declared in library ghc-prim:GHC.Types. GHC.Core.Make. Re-engineer the treatment of "big" tuples (mkBigCoreVarTup etc) GHC.Core.Make, so that it auto-boxes unboxed values and (crucially) types of kind Constraint. That allows the desugaring for arrows to work; it gathers up free variables (including dictionaries) into tuples. See Note [Big tuples] in GHC.Core.Make. There is still work to do here: #22336. But things are better than before. * GHC.Core.Make. We need two absent-error Ids, aBSENT_ERROR_ID for types of kind Type, and aBSENT_CONSTRAINT_ERROR_ID for vaues of kind Constraint. Ditto noInlineId vs noInlieConstraintId in GHC.Types.Id.Make; see Note [inlineId magic]. * GHC.Core.TyCo.Rep. Completely refactor the NthCo coercion. It is now called SelCo, and its fields are much more descriptive than the single Int we used to have. A great improvement. See Note [SelCo] in GHC.Core.TyCo.Rep. * GHC.Core.RoughMap.roughMatchTyConName. Collapse TYPE and CONSTRAINT to a single TyCon, so that the rough-map does not distinguish them. * GHC.Core.DataCon - Mainly just improve documentation * Some significant renamings: GHC.Core.Multiplicity: Many --> ManyTy (easier to grep for) One --> OneTy GHC.Core.TyCo.Rep TyCoBinder --> GHC.Core.Var.PiTyBinder GHC.Core.Var TyCoVarBinder --> ForAllTyBinder AnonArgFlag --> FunTyFlag ArgFlag --> ForAllTyFlag GHC.Core.TyCon TyConTyCoBinder --> TyConPiTyBinder Many functions are renamed in consequence e.g. isinvisibleArgFlag becomes isInvisibleForAllTyFlag, etc * I refactored FunTyFlag (was AnonArgFlag) into a simple, flat data type data FunTyFlag = FTF_T_T -- (->) Type -> Type | FTF_T_C -- (-=>) Type -> Constraint | FTF_C_T -- (=>) Constraint -> Type | FTF_C_C -- (==>) Constraint -> Constraint * GHC.Tc.Errors.Ppr. Some significant refactoring in the TypeEqMisMatch case of pprMismatchMsg. * I made the tyConUnique field of TyCon strict, because I saw code with lots of silly eval's. That revealed that GHC.Settings.Constants.mAX_SUM_SIZE can only be 63, because we pack the sum tag into a 6-bit field. (Lurking bug squashed.) Fixes * #21530 Updates haddock submodule slightly. Performance changes ~~~~~~~~~~~~~~~~~~~ I was worried that compile times would get worse, but after some careful profiling we are down to a geometric mean 0.1% increase in allocation (in perf/compiler). That seems fine. There is a big runtime improvement in T10359 Metric Decrease: LargeRecord MultiLayerModulesTH_OneShot T13386 T13719 Metric Increase: T8095 - - - - - 360f5fec by Simon Peyton Jones at 2022-11-11T23:40:11+00:00 Indent closing "#-}" to silence HLint - - - - - e160cf47 by Krzysztof Gogolewski at 2022-11-12T08:05:28-05:00 Fix merge conflict in T18355.stderr Fixes #22446 - - - - - 294f9073 by Simon Peyton Jones at 2022-11-12T23:14:13+00:00 Fix a trivial typo in dataConNonlinearType Fixes #22416 - - - - - 268a3ce9 by Ben Gamari at 2022-11-14T09:36:57-05:00 eventlog: Ensure that IPE output contains actual info table pointers The refactoring in 866c736e introduced a rather subtle change in the semantics of the IPE eventlog output, changing the eventlog field from encoding info table pointers to "TNTC pointers" (which point to entry code when tables-next-to-code is enabled). Fix this. Fixes #22452. - - - - - d91db679 by Matthew Pickering at 2022-11-14T16:48:10-05:00 testsuite: Add tests for T22347 These are fixed in recent versions but might as well add regression tests. See #22347 - - - - - 8f6c576b by Matthew Pickering at 2022-11-14T16:48:45-05:00 testsuite: Improve output from tests which have failing pre_cmd There are two changes: * If a pre_cmd fails, then don't attempt to run the test. * If a pre_cmd fails, then print the stdout and stderr from running that command (which hopefully has a nice error message). For example: ``` =====> 1 of 1 [0, 0, 0] *** framework failure for test-defaulting-plugin(normal) pre_cmd failed: 2 ** pre_cmd was "$MAKE -s --no-print-directory -C defaulting-plugin package.test-defaulting-plugin TOP={top}". stdout: stderr: DefaultLifted.hs:19:13: error: [GHC-76037] Not in scope: type constructor or class ‘Typ’ Suggested fix: Perhaps use one of these: ‘Type’ (imported from GHC.Tc.Utils.TcType), data constructor ‘Type’ (imported from GHC.Plugins) | 19 | instance Eq Typ where | ^^^ make: *** [Makefile:17: package.test-defaulting-plugin] Error 1 Performance Metrics (test environment: local): ``` Fixes #22329 - - - - - 2b7d5ccc by Madeline Haraj at 2022-11-14T22:44:17+00:00 Implement UNPACK support for sum types. This is based on osa's unpack_sums PR from ages past. The meat of the patch is implemented in dataConArgUnpackSum and described in Note [UNPACK for sum types]. - - - - - 78f7ecb0 by Andreas Klebinger at 2022-11-14T22:20:29-05:00 Expand on the need to clone local binders. Fixes #22402. - - - - - 65ce43cc by Krzysztof Gogolewski at 2022-11-14T22:21:05-05:00 Fix :i Constraint printing "type Constraint = Constraint" Since Constraint became a synonym for CONSTRAINT 'LiftedRep, we need the same code for handling printing as for the synonym Type = TYPE 'LiftedRep. This addresses the same bug as #18594, so I'm reusing the test. - - - - - 94549f8f by ARATA Mizuki at 2022-11-15T21:36:03-05:00 configure: Don't check for an unsupported version of LLVM The upper bound is not inclusive. Fixes #22449 - - - - - 02d3511b by Andrew Lelechenko at 2022-11-15T21:36:41-05:00 Fix capitalization in haddock for TestEquality - - - - - 08bf2881 by Cheng Shao at 2022-11-16T09:16:29+00:00 base: make Foreign.Marshal.Pool use RTS internal arena for allocation `Foreign.Marshal.Pool` used to call `malloc` once for each allocation request. Each `Pool` maintained a list of allocated pointers, and traverses the list to `free` each one of those pointers. The extra O(n) overhead is apparently bad for a `Pool` that serves a lot of small allocation requests. This patch uses the RTS internal arena to implement `Pool`, with these benefits: - Gets rid of the extra O(n) overhead. - The RTS arena is simply a bump allocator backed by the block allocator, each allocation request is likely faster than a libc `malloc` call. Closes #14762 #18338. - - - - - 37cfe3c0 by Krzysztof Gogolewski at 2022-11-16T14:50:06-05:00 Misc cleanup * Replace catMaybes . map f with mapMaybe f * Use concatFS to concatenate multiple FastStrings * Fix documentation of -exclude-module * Cleanup getIgnoreCount in GHCi.UI - - - - - b0ac3813 by Lawton Nichols at 2022-11-19T03:22:14-05:00 Give better errors for code corrupted by Unicode smart quotes (#21843) Previously, we emitted a generic and potentially confusing error during lexical analysis on programs containing smart quotes (“/”/‘/’). This commit adds smart quote-aware lexer errors. - - - - - cb8430f8 by Sebastian Graf at 2022-11-19T03:22:49-05:00 Make OpaqueNo* tests less noisy to unrelated changes - - - - - b1a8af69 by Sebastian Graf at 2022-11-19T03:22:49-05:00 Simplifier: Consider `seq` as a `BoringCtxt` (#22317) See `Note [Seq is boring]` for the rationale. Fixes #22317. - - - - - 9fd11585 by Sebastian Graf at 2022-11-19T03:22:49-05:00 Make T21839c's ghc/max threshold more forgiving - - - - - 4b6251ab by Simon Peyton Jones at 2022-11-19T03:23:24-05:00 Be more careful when reporting unbound RULE binders See Note [Variables unbound on the LHS] in GHC.HsToCore.Binds. Fixes #22471. - - - - - e8f2b80d by Peter Trommler at 2022-11-19T03:23:59-05:00 PPC NCG: Fix generating assembler code Fixes #22479 - - - - - f2f9ef07 by Andrew Lelechenko at 2022-11-20T18:39:30-05:00 Extend documentation for Data.IORef - - - - - ef511b23 by Simon Peyton Jones at 2022-11-20T18:40:05-05:00 Buglet in GHC.Tc.Module.checkBootTyCon This lurking bug used the wrong function to compare two types in GHC.Tc.Module.checkBootTyCon It's hard to trigger the bug, which only came up during !9343, so there's no regression test in this MR. - - - - - 451aeac3 by Andrew Lelechenko at 2022-11-20T18:40:44-05:00 Add since pragmas for c_interruptible_open and hostIsThreaded - - - - - a57b4c4b by Andrew Lelechenko at 2022-11-21T00:39:52+00:00 Support mtl-2.3 - - - - - 8d6aaa49 by Duncan Coutts at 2022-11-22T02:06:16-05:00 Introduce CapIOManager as the per-cap I/O mangager state Rather than each I/O manager adding things into the Capability structure ad-hoc, we should have a common CapIOManager iomgr member of the Capability structure, with a common interface to initialise etc. The content of the CapIOManager struct will be defined differently for each I/O manager implementation. Eventually we should be able to have the CapIOManager be opaque to the rest of the RTS, and known just to the I/O manager implementation. We plan for that by making the Capability contain a pointer to the CapIOManager rather than containing the structure directly. Initially just move the Unix threaded I/O manager's control FD. - - - - - 8901285e by Duncan Coutts at 2022-11-22T02:06:17-05:00 Add hook markCapabilityIOManager To allow I/O managers to have GC roots in the Capability, within the CapIOManager structure. Not yet used in this patch. - - - - - 5cf709c5 by Duncan Coutts at 2022-11-22T02:06:17-05:00 Move APPEND_TO_BLOCKED_QUEUE from cmm to C The I/O and delay blocking primitives for the non-threaded way currently access the blocked_queue and sleeping_queue directly. We want to move where those queues are to make their ownership clearer: to have them clearly belong to the I/O manager impls rather than to the scheduler. Ultimately we will want to change their representation too. It's inconvenient to do that if these queues are accessed directly from cmm code. So as a first step, replace the APPEND_TO_BLOCKED_QUEUE with a C version appendToIOBlockedQueue(), and replace the open-coded sleeping_queue insertion with insertIntoSleepingQueue(). - - - - - ced9acdb by Duncan Coutts at 2022-11-22T02:06:17-05:00 Move {blocked,sleeping}_queue from scheduler global vars to CapIOManager The blocked_queue_{hd,tl} and the sleeping_queue are currently cooperatively managed between the scheduler and (some but not all of) the non-threaded I/O manager implementations. They lived as global vars with the scheduler, but are poked by I/O primops and the I/O manager backends. This patch is a step on the path towards making the management of I/O or timer blocking belong to the I/O managers and not the scheduler. Specifically, this patch moves the {blocked,sleeping}_queue from being global vars in the scheduler to being members of the CapIOManager struct within each Capability. They are not yet exclusively used by the I/O managers: they are still poked from a couple other places, notably in the scheduler before calling awaitEvent. - - - - - 0f68919e by Duncan Coutts at 2022-11-22T02:06:17-05:00 Remove the now-unused markScheduler The global vars {blocked,sleeping}_queue are now in the Capability and so get marked there via markCapabilityIOManager. - - - - - 39a91f60 by Duncan Coutts at 2022-11-22T02:06:17-05:00 Move macros for checking for pending IO or timers from Schedule.h to Schedule.c and IOManager.h This is just moving, the next step will be to rejig them slightly. For the non-threaded RTS the scheduler needs to be able to test for there being pending I/O operation or pending timers. The implementation of these tests should really be considered to be part of the I/O managers and not part of the scheduler. - - - - - 664b034b by Duncan Coutts at 2022-11-22T02:06:17-05:00 Replace EMPTY_{BLOCKED,SLEEPING}_QUEUE macros by function These are the macros originaly from Scheduler.h, previously moved to IOManager.h, and now replaced with a single inline function anyPendingTimeoutsOrIO(). We can use a single function since the two macros were always checked together. Note that since anyPendingTimeoutsOrIO is defined for all IO manager cases, including threaded, we do not need to guard its use by cpp #if !defined(THREADED_RTS) - - - - - 32946220 by Duncan Coutts at 2022-11-22T02:06:17-05:00 Expand emptyThreadQueues inline for clarity It was not really adding anything. The name no longer meant anything since those I/O and timeout queues do not belong to the scheuler. In one of the two places it was used, the comments already had to explain what it did, whereas now the code matches the comment nicely. - - - - - 9943baf9 by Duncan Coutts at 2022-11-22T02:06:17-05:00 Move the awaitEvent declaration into IOManager.h And add or adjust comments at the use sites of awaitEvent. - - - - - 054dcc9d by Duncan Coutts at 2022-11-22T02:06:17-05:00 Pass the Capability *cap explicitly to awaitEvent It is currently only used in the non-threaded RTS so it works to use MainCapability, but it's a bit nicer to pass the cap anyway. It's certainly shorter. - - - - - 667fe5a4 by Duncan Coutts at 2022-11-22T02:06:17-05:00 Pass the Capability *cap explicitly to appendToIOBlockedQueue And to insertIntoSleepingQueue. Again, it's a bit cleaner and simpler though not strictly necessary given that these primops are currently only used in the non-threaded RTS. - - - - - 7181b074 by Duncan Coutts at 2022-11-22T02:06:17-05:00 Reveiew feedback: improve one of the TODO comments The one about the nonsense (const False) test on WinIO for there being any IO or timers pending, leading to unnecessary complication later in the scheduler. - - - - - e5b68183 by Andreas Klebinger at 2022-11-22T02:06:52-05:00 Optimize getLevity. Avoid the intermediate data structures allocated by splitTyConApp. This avoids ~0.5% of allocations for a build using -O2. Fixes #22254 - - - - - de5fb348 by Andreas Klebinger at 2022-11-22T02:07:28-05:00 hadrian:Set TNTC when running testsuite. - - - - - 9d61c182 by Oleg Grenrus at 2022-11-22T15:59:34-05:00 Add unsafePtrEquality# restricted to UnliftedTypes - - - - - e817c871 by Jonathan Dowland at 2022-11-22T16:00:14-05:00 utils/unlit: adjust parser to match Report spec The Haskell 2010 Report says that, for Latex-style Literate format, "Program code begins on the first line following a line that begins \begin{code}". (This is unchanged from the 98 Report) However the unlit.c implementation only matches a line that contains "\begin{code}" and nothing else. One consequence of this is that one cannot suffix Latex options to the code environment. I.e., this does not work: \begin{code}[label=foo,caption=Foo Code] Adjust the matcher to conform to the specification from the Report. The Haskell Wiki currently recommends suffixing a '%' to \begin{code} in order to deliberately hide a code block from Haskell. This is bad advice, as it's relying on an implementation quirk rather than specified behaviour. None-the-less, some people have tried to use it, c.f. <https://mail.haskell.org/pipermail/haskell-cafe/2009-September/066780.html> An alternative solution is to define a separate, equivalent Latex environment to "code", that is functionally identical in Latex but ignored by unlit. This should not be a burden: users are required to manually define the code environment anyway, as it is not provided by the Latex verbatim or lstlistings packages usually used for presenting code in documents. Fixes #3549. - - - - - 0b7fef11 by Teo Camarasu at 2022-11-23T12:44:33-05:00 Fix eventlog all option Previously it didn't enable/disable nonmoving_gc and ticky event types Fixes #21813 - - - - - 04d0618c by Arnaud Spiwack at 2022-11-23T12:45:14-05:00 Expand Note [Linear types] with the stance on linting linearity Per the discussion on #22123 - - - - - e1538516 by Lawton Nichols at 2022-11-23T12:45:55-05:00 Add documentation on custom Prelude modules (#22228) Specifically, custom Prelude modules that are named `Prelude`. - - - - - b5c71454 by Sylvain Henry at 2022-11-23T12:46:35-05:00 Don't let configure perform trivial substitutions (#21846) Hadrian now performs substitutions, especially to generate .cabal files from .cabal.in files. Two benefits: 1. We won't have to re-configure when we modify thing.cabal.in. Hadrian will take care of this for us. 2. It paves the way to allow the same package to be configured differently by Hadrian in the same session. This will be useful to fix #19174: we want to build a stage2 cross-compiler for the host platform and a stage1 compiler for the cross target platform in the same Hadrian session. - - - - - 99aca26b by nineonine at 2022-11-23T12:47:11-05:00 CApiFFI: add ConstPtr for encoding const-qualified pointer return types (#22043) Previously, when using `capi` calling convention in foreign declarations, code generator failed to handle const-cualified pointer return types. This resulted in CC toolchain throwing `-Wincompatible-pointer-types-discards-qualifiers` warning. `Foreign.C.Types.ConstPtr` newtype was introduced to handle these cases - special treatment was put in place to generate appropritetly qualified C wrapper that no longer triggers the above mentioned warning. Fixes #22043 - - - - - 040bfdc3 by M Farkas-Dyck at 2022-11-23T21:59:03-05:00 Scrub some no-warning pragmas. - - - - - 178c1fd8 by Vladislav Zavialov at 2022-11-23T21:59:39-05:00 Check if the SDoc starts with a single quote (#22488) This patch fixes pretty-printing of character literals inside promoted lists and tuples. When we pretty-print a promoted list or tuple whose first element starts with a single quote, we want to add a space between the opening bracket and the element: '[True] -- ok '[ 'True] -- ok '['True] -- not ok If we don't add the space, we accidentally produce a character literal '['. Before this patch, pprSpaceIfPromotedTyCon inspected the type as an AST and tried to guess if it would be rendered with a single quote. However, it missed the case when the inner type was itself a character literal: '[ 'x'] -- ok '['x'] -- not ok Instead of adding this particular case, I opted for a more future-proof solution: check the SDoc directly. This way we can detect if the single quote is actually there instead of trying to predict it from the AST. The new function is called spaceIfSingleQuote. - - - - - 11627c42 by Matthew Pickering at 2022-11-23T22:00:15-05:00 notes: Fix references to HPT space leak note Updating this note was missed when updating the HPT to the HUG. Fixes #22477 - - - - - 86ff1523 by Andrei Borzenkov at 2022-11-24T17:24:51-05:00 Convert diagnostics in GHC.Rename.Expr to proper TcRnMessage (#20115) Problem: avoid usage of TcRnMessageUnknown Solution: The following `TcRnMessage` messages has been introduced: TcRnNoRebindableSyntaxRecordDot TcRnNoFieldPunsRecordDot TcRnIllegalStaticExpression TcRnIllegalStaticFormInSplice TcRnListComprehensionDuplicateBinding TcRnEmptyStmtsGroup TcRnLastStmtNotExpr TcRnUnexpectedStatementInContext TcRnIllegalTupleSection TcRnIllegalImplicitParameterBindings TcRnSectionWithoutParentheses Co-authored-by: sheaf <sam.derbyshire at gmail.com> - - - - - d198a19a by Cheng Shao at 2022-11-24T17:25:29-05:00 rts: fix missing Arena.h symbols in RtsSymbols.c It was an unfortunate oversight in !8961 and broke devel2 builds. - - - - - 5943e739 by Andrew Lelechenko at 2022-11-25T04:38:28-05:00 Assorted fixes to avoid Data.List.{head,tail} - - - - - 1f1b99b8 by sheaf at 2022-11-25T04:38:28-05:00 Review suggestions for assorted fixes to avoid Data.List.{head,tail} - - - - - 13d627bb by Vladislav Zavialov at 2022-11-25T04:39:04-05:00 Print unticked promoted data constructors (#20531) Before this patch, GHC unconditionally printed ticks before promoted data constructors: ghci> type T = True -- unticked (user-written) ghci> :kind! T T :: Bool = 'True -- ticked (compiler output) After this patch, GHC prints ticks only when necessary: ghci> type F = False -- unticked (user-written) ghci> :kind! F F :: Bool = False -- unticked (compiler output) ghci> data False -- introduce ambiguity ghci> :kind! F F :: Bool = 'False -- ticked by necessity (compiler output) The old behavior can be enabled by -fprint-redundant-promotion-ticks. Summary of changes: * Rename PrintUnqualified to NamePprCtx * Add QueryPromotionTick to it * Consult the GlobalRdrEnv to decide whether to print a tick (see mkPromTick) * Introduce -fprint-redundant-promotion-ticks Co-authored-by: Artyom Kuznetsov <hi at wzrd.ht> - - - - - e9d62453 by Simon Peyton Jones at 2022-11-25T13:49:12+01:00 Track small API change in TyCon.hs - - - - - d10dc6bd by Simon Peyton Jones at 2022-11-25T22:31:27+00:00 Fix decomposition of TyConApps Ticket #22331 showed that we were being too eager to decompose a Wanted TyConApp, leading to incompleteness in the solver. To understand all this I ended up doing a substantial rewrite of the old Note [Decomposing equalities], now reborn as Note [Decomposing TyConApp equalities]. Plus rewrites of other related Notes. The actual fix is very minor and actually simplifies the code: in `can_decompose` in `GHC.Tc.Solver.Canonical.canTyConApp`, we now call `noMatchableIrreds`. A closely related refactor: we stop trying to use the same "no matchable givens" function here as in `matchClassInst`. Instead split into two much simpler functions. - - - - - 2da5c38a by Will Hawkins at 2022-11-26T04:05:04-05:00 Redirect output of musttail attribute test Compilation output from test for support of musttail attribute leaked to the console. - - - - - 0eb1c331 by Cheng Shao at 2022-11-28T08:55:53+00:00 Move hs_mulIntMayOflo cbits to ghc-prim It's only used by wasm NCG at the moment, but ghc-prim is a more reasonable place for hosting out-of-line primops. Also, we only need a single version of hs_mulIntMayOflo. - - - - - 36b53a9d by Cheng Shao at 2022-11-28T09:05:57+00:00 compiler: generate ccalls for clz/ctz/popcnt in wasm NCG We used to generate a single wasm clz/ctz/popcnt opcode, but it's wrong when it comes to subwords, so might as well generate ccalls for them. See #22470 for details. - - - - - d4134e92 by Cheng Shao at 2022-11-28T23:48:14-05:00 compiler: remove unused MO_U_MulMayOflo We actually only emit MO_S_MulMayOflo and never emit MO_U_MulMayOflo anywhere. - - - - - 8d15eadc by Apoorv Ingle at 2022-11-29T03:09:31-05:00 Killing cc_fundeps, streamlining kind equality orientation, and type equality processing order Fixes: #217093 Associated to #19415 This change * Flips the orientation of the the generated kind equality coercion in canEqLHSHetero; * Removes `cc_fundeps` in CDictCan as the check was incomplete; * Changes `canDecomposableTyConAppOk` to ensure we process kind equalities before type equalities and avoiding a call to `canEqLHSHetero` while processing wanted TyConApp equalities * Adds 2 new tests for validating the change - testsuites/typecheck/should_compile/T21703.hs and - testsuites/typecheck/should_fail/T19415b.hs (a simpler version of T19415.hs) * Misc: Due to the change in the equality direction some error messages now have flipped type mismatch errors * Changes in Notes: - Note [Fundeps with instances, and equality orientation] supercedes Note [Fundeps with instances] - Added Note [Kind Equality Orientation] to visualize the kind flipping - Added Note [Decomposing Dependent TyCons and Processing Wanted Equalties] - - - - - 646969d4 by Krzysztof Gogolewski at 2022-11-29T03:10:13-05:00 Change printing of sized literals to match the proposal Literals in Core were printed as e.g. 0xFF#16 :: Int16#. The proposal 451 now specifies syntax 0xFF#Int16. This change affects the Core printer only - more to be done later. Part of #21422. - - - - - 02e282ec by Simon Peyton Jones at 2022-11-29T03:10:48-05:00 Be a bit more selective about floating bottoming expressions This MR arranges to float a bottoming expression to the top only if it escapes a value lambda. See #22494 and Note [Floating to the top] in SetLevels. This has a generally beneficial effect in nofib +-------------------------------++----------+ | ||tsv (rel) | +===============================++==========+ | imaginary/paraffins || -0.93% | | imaginary/rfib || -0.05% | | real/fem || -0.03% | | real/fluid || -0.01% | | real/fulsom || +0.05% | | real/gamteb || -0.27% | | real/gg || -0.10% | | real/hidden || -0.01% | | real/hpg || -0.03% | | real/scs || -11.13% | | shootout/k-nucleotide || -0.01% | | shootout/n-body || -0.08% | | shootout/reverse-complement || -0.00% | | shootout/spectral-norm || -0.02% | | spectral/fibheaps || -0.20% | | spectral/hartel/fft || -1.04% | | spectral/hartel/solid || +0.33% | | spectral/hartel/wave4main || -0.35% | | spectral/mate || +0.76% | +===============================++==========+ | geom mean || -0.12% | The effect on compile time is generally slightly beneficial Metrics: compile_time/bytes allocated ---------------------------------------------- MultiLayerModulesTH_OneShot(normal) +0.3% PmSeriesG(normal) -0.2% PmSeriesT(normal) -0.1% T10421(normal) -0.1% T10421a(normal) -0.1% T10858(normal) -0.1% T11276(normal) -0.1% T11303b(normal) -0.2% T11545(normal) -0.1% T11822(normal) -0.1% T12150(optasm) -0.1% T12234(optasm) -0.3% T13035(normal) -0.2% T16190(normal) -0.1% T16875(normal) -0.4% T17836b(normal) -0.2% T17977(normal) -0.2% T17977b(normal) -0.2% T18140(normal) -0.1% T18282(normal) -0.1% T18304(normal) -0.2% T18698a(normal) -0.1% T18923(normal) -0.1% T20049(normal) -0.1% T21839r(normal) -0.1% T5837(normal) -0.4% T6048(optasm) +3.2% BAD T9198(normal) -0.2% T9630(normal) -0.1% TcPlugin_RewritePerf(normal) -0.4% hard_hole_fits(normal) -0.1% geo. mean -0.0% minimum -0.4% maximum +3.2% The T6048 outlier is hard to pin down, but it may be the effect of reading in more interface files definitions. It's a small program for which compile time is very short, so I'm not bothered about it. Metric Increase: T6048 - - - - - ab23dc5e by Ben Gamari at 2022-11-29T03:11:25-05:00 testsuite: Mark unpack_sums_6 as fragile due to #22504 This test is explicitly dependent upon runtime, which is generally not appropriate given that the testsuite is run in parallel and generally saturates the CPU. - - - - - def47dd3 by Ben Gamari at 2022-11-29T03:11:25-05:00 testsuite: Don't use grep -q in unpack_sums_7 `grep -q` closes stdin as soon as it finds the pattern it is looking for, resulting in #22484. - - - - - cc25d52e by Sylvain Henry at 2022-11-29T09:44:31+01:00 Add Javascript backend Add JS backend adapted from the GHCJS project by Luite Stegeman. Some features haven't been ported or implemented yet. Tests for these features have been disabled with an associated gitlab ticket. Bump array submodule Work funded by IOG. Co-authored-by: Jeffrey Young <jeffrey.young at iohk.io> Co-authored-by: Luite Stegeman <stegeman at gmail.com> Co-authored-by: Josh Meredith <joshmeredith2008 at gmail.com> - - - - - 68c966cd by sheaf at 2022-11-30T09:31:25-05:00 Fix @since annotations on WithDict and Coercible Fixes #22453 - - - - - a3a8e9e9 by Simon Peyton Jones at 2022-11-30T09:32:03-05:00 Be more careful in GHC.Tc.Solver.Interact.solveOneFromTheOther We were failing to account for the cc_pend_sc flag in this important function, with the result that we expanded superclasses forever. Fixes #22516. - - - - - a9d9b8c0 by Simon Peyton Jones at 2022-11-30T09:32:03-05:00 Use mkNakedFunTy in tcPatSynSig As #22521 showed, in tcPatSynSig we make a "fake type" to kind-generalise; and that type has unzonked type variables in it. So we must not use `mkFunTy` (which checks FunTy's invariants) via `mkPhiTy` when building this type. Instead we need to use `mkNakedFunTy`. Easy fix. - - - - - 31462d98 by Andreas Klebinger at 2022-11-30T14:50:58-05:00 Properly cast values when writing/reading unboxed sums. Unboxed sums might store a Int8# value as Int64#. This patch makes sure we keep track of the actual value type. See Note [Casting slot arguments] for the details. - - - - - 10a2a7de by Oleg Grenrus at 2022-11-30T14:51:39-05:00 Move Void to GHC.Base... This change would allow `Void` to be used deeper in module graph. For example exported from `Prelude` (though that might be already possible). Also this change includes a change `stimes @Void _ x = x`, https://github.com/haskell/core-libraries-committee/issues/95 While the above is not required, maintaining old stimes behavior would be tricky as `GHC.Base` doesn't know about `Num` or `Integral`, which would require more hs-boot files. - - - - - b4cfa8e2 by Sebastian Graf at 2022-11-30T14:52:24-05:00 DmdAnal: Reflect the `seq` of strict fields of a DataCon worker (#22475) See the updated `Note [Data-con worker strictness]` and the new `Note [Demand transformer for data constructors]`. Fixes #22475. - - - - - d87f28d8 by Baldur Blöndal at 2022-11-30T21:16:36+01:00 Make Functor a quantified superclass of Bifunctor. See https://github.com/haskell/core-libraries-committee/issues/91 for discussion. This change relates Bifunctor with Functor by requiring second = fmap. Moreover this change is a step towards unblocking the major version bump of bifunctors and profunctors to major version 6. This paves the way to move the Profunctor class into base. For that Functor first similarly becomes a superclass of Profunctor in the new major version 6. - - - - - 72cf4c5d by doyougnu at 2022-12-01T12:36:44-05:00 FastString: SAT bucket_match Metric Decrease: MultiLayerModulesTH_OneShot - - - - - afc2540d by Simon Peyton Jones at 2022-12-01T12:37:20-05:00 Add a missing varToCoreExpr in etaBodyForJoinPoint This subtle bug showed up when compiling a library with 9.4. See #22491. The bug is present in master, but it is hard to trigger; the new regression test T22491 fails in 9.4. The fix was easy: just add a missing varToCoreExpr in etaBodyForJoinPoint. The fix is definitely right though! I also did some other minor refatoring: * Moved the preInlineUnconditionally test in simplExprF1 to before the call to joinPointBinding_maybe, to avoid fruitless eta-expansion. * Added a boolean from_lam flag to simplNonRecE, to avoid two fruitless tests, and commented it a bit better. These refactorings seem to save 0.1% on compile-time allocation in perf/compiler; with a max saving of 1.4% in T9961 Metric Decrease: T9961 - - - - - 81eeec7f by M Farkas-Dyck at 2022-12-01T12:37:56-05:00 CI: Forbid the fully static build on Alpine to fail. To do so, we mark some tests broken in this configuration. - - - - - c5d1bf29 by Bryan Richter at 2022-12-01T12:37:56-05:00 CI: Remove ARMv7 jobs These jobs fail (and are allowed to fail) nearly every time. Soon they won't even be able to run at all, as we won't currently have runners that can run them. Fixing the latter problem is tracked in #22409. I went ahead and removed all settings and configurations. - - - - - d82992fd by Bryan Richter at 2022-12-01T12:37:56-05:00 CI: Fix CI lint Failure was introduced by conflicting changes to gen_ci.hs that did *not* trigger git conflicts. - - - - - ce126993 by Simon Peyton Jones at 2022-12-02T01:22:12-05:00 Refactor TyCon to have a top-level product This patch changes the representation of TyCon so that it has a top-level product type, with a field that gives the details (newtype, type family etc), #22458. Not much change in allocation, but execution seems to be a bit faster. Includes a change to the haddock submodule to adjust for API changes. - - - - - 74c767df by Matthew Pickering at 2022-12-02T01:22:48-05:00 ApplicativeDo: Set pattern location before running exhaustiveness checker This improves the error messages of the exhaustiveness checker when checking statements which have been moved around with ApplicativeDo. Before: Test.hs:2:3: warning: [GHC-62161] [-Wincomplete-uni-patterns] Pattern match(es) are non-exhaustive In a pattern binding: Patterns of type ‘Maybe ()’ not matched: Nothing | 2 | let x = () | ^^^^^^^^^^ After: Test.hs:4:3: warning: [GHC-62161] [-Wincomplete-uni-patterns] Pattern match(es) are non-exhaustive In a pattern binding: Patterns of type ‘Maybe ()’ not matched: Nothing | 4 | ~(Just res1) <- seq x (pure $ Nothing @()) | Fixes #22483 - - - - - 85ecc1a0 by Matthew Pickering at 2022-12-02T19:46:43-05:00 Add special case for :Main module in `GHC.IfaceToCore.mk_top_id` See Note [Root-main Id] The `:Main` special binding is actually defined in the current module (hence don't go looking for it externally) but the module name is rOOT_MAIN rather than the current module so we need this special case. There was already some similar logic in `GHC.Rename.Env` for External Core, but now the "External Core" is in interface files it needs to be moved here instead. Fixes #22405 - - - - - 108c319f by Krzysztof Gogolewski at 2022-12-02T19:47:18-05:00 Fix linearity checking in Lint Lint was not able to see that x*y <= x*y, because this inequality was decomposed to x <= x*y && y <= x*y, but there was no rule to see that x <= x*y. Fixes #22546. - - - - - bb674262 by Bryan Richter at 2022-12-03T04:38:46-05:00 Mark T16916 fragile See https://gitlab.haskell.org/ghc/ghc/-/issues/16966 - - - - - 5d267d46 by Vladislav Zavialov at 2022-12-03T04:39:22-05:00 Refactor: FreshOrReuse instead of addTyClTyVarBinds This is a refactoring that should have no effect on observable behavior. Prior to this change, GHC.HsToCore.Quote contained a few closely related functions to process type variable bindings: addSimpleTyVarBinds, addHsTyVarBinds, addQTyVarBinds, and addTyClTyVarBinds. We can classify them by their input type and name generation strategy: Fresh names only Reuse bound names +---------------------+-------------------+ [Name] | addSimpleTyVarBinds | | [LHsTyVarBndr flag GhcRn] | addHsTyVarBinds | | LHsQTyVars GhcRn | addQTyVarBinds | addTyClTyVarBinds | +---------------------+-------------------+ Note how two functions are missing. Because of this omission, there were two places where a LHsQTyVars value was constructed just to be able to pass it to addTyClTyVarBinds: 1. mk_qtvs in addHsOuterFamEqnTyVarBinds -- bad 2. mkHsQTvs in repFamilyDecl -- bad This prevented me from making other changes to LHsQTyVars, so the main goal of this refactoring is to get rid of those workarounds. The most direct solution would be to define the missing functions. But that would lead to a certain amount of code duplication. To avoid code duplication, I factored out the name generation strategy into a function parameter: data FreshOrReuse = FreshNamesOnly | ReuseBoundNames addSimpleTyVarBinds :: FreshOrReuse -> ... addHsTyVarBinds :: FreshOrReuse -> ... addQTyVarBinds :: FreshOrReuse -> ... - - - - - c189b831 by Vladislav Zavialov at 2022-12-03T04:39:22-05:00 addHsOuterFamEqnTyVarBinds: use FreshNamesOnly for explicit binders Consider this example: [d| instance forall a. C [a] where type forall b. G [a] b = Proxy b |] When we process "forall b." in the associated type instance, it is unambiguously the binding site for "b" and we want a fresh name for it. Therefore, FreshNamesOnly is more fitting than ReuseBoundNames. This should not have any observable effect but it avoids pointless lookups in the MetaEnv. - - - - - 42512264 by Ross Paterson at 2022-12-03T10:32:45+00:00 Handle type data declarations in Template Haskell quotations and splices (fixes #22500) This adds a TypeDataD constructor to the Template Haskell Dec type, and ensures that the constructors it contains go in the TyCls namespace. - - - - - 1a767fa3 by Vladislav Zavialov at 2022-12-05T05:18:50-05:00 Add BufSpan to EpaLocation (#22319, #22558) The key part of this patch is the change to mkTokenLocation: - mkTokenLocation (RealSrcSpan r _) = TokenLoc (EpaSpan r) + mkTokenLocation (RealSrcSpan r mb) = TokenLoc (EpaSpan r mb) mkTokenLocation used to discard the BufSpan, but now it is saved and can be retrieved from LHsToken or LHsUniToken. This is made possible by the following change to EpaLocation: - data EpaLocation = EpaSpan !RealSrcSpan + data EpaLocation = EpaSpan !RealSrcSpan !(Strict.Maybe BufSpan) | ... The end goal is to make use of the BufSpan in Parser/PostProcess/Haddock. - - - - - cd31acad by sheaf at 2022-12-06T15:45:58-05:00 Hadrian: fix ghcDebugAssertions off-by-one error Commit 6b2f7ffe changed the logic that decided whether to enable debug assertions. However, it had an off-by-one error, as the stage parameter to the function inconsistently referred to the stage of the compiler being used to build or the stage of the compiler we are building. This patch makes it consistent. Now the parameter always refers to the the compiler which is being built. In particular, this patch re-enables assertions in the stage 2 compiler when building with devel2 flavour, and disables assertions in the stage 2 compiler when building with validate flavour. Some extra performance tests are now run in the "validate" jobs because the stage2 compiler no longer contains assertions. ------------------------- Metric Decrease: CoOpt_Singletons MultiComponentModules MultiComponentModulesRecomp MultiLayerModulesTH_OneShot T11374 T12227 T12234 T13253-spj T13701 T14683 T14697 T15703 T17096 T17516 T18304 T18478 T18923 T5030 T9872b TcPlugin_RewritePerf Metric Increase: MultiComponentModules MultiComponentModulesRecomp MultiLayerModules MultiLayerModulesRecomp MultiLayerModulesTH_Make T13386 T13719 T3294 T9233 T9675 parsing001 ------------------------- - - - - - 21d66db1 by mrkun at 2022-12-06T15:46:38-05:00 Push DynFlags out of runInstallNameTool - - - - - aaaaa79b by mrkun at 2022-12-06T15:46:38-05:00 Push DynFlags out of askOtool - - - - - 4e28f49e by mrkun at 2022-12-06T15:46:38-05:00 Push DynFlags out of runInjectRPaths - - - - - a7422580 by mrkun at 2022-12-06T15:46:38-05:00 Push DynFlags out of Linker.MacOS - - - - - eb1c73f7 by Ben Gamari at 2022-12-07T08:46:21-05:00 Update for GhC 9.6 - - - - - 063268dd by Ben Gamari at 2022-12-07T11:26:32-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - e902d771 by Matthew Craven at 2022-12-08T08:30:23-05:00 Fix bounds-checking buglet in Data.Array.Byte ...another manifestation of #20851 which I unfortunately missed in my first pass. - - - - - 8d36c0c6 by Gergő Érdi at 2022-12-08T08:31:03-05:00 Remove copy-pasted definitions of `graphFromEdgedVertices*` - - - - - c5d8ed3a by Gergő Érdi at 2022-12-08T08:31:03-05:00 Add version of `reachableGraph` that avoids loop for cyclic inputs by building its result connected component by component Fixes #22512 - - - - - 90cd5396 by Krzysztof Gogolewski at 2022-12-08T08:31:39-05:00 Mark Type.Reflection.Unsafe as Unsafe This module can be used to construct ill-formed TypeReps, so it should be Unsafe. - - - - - 2057c77d by Ian-Woo Kim at 2022-12-08T08:32:19-05:00 Truncate eventlog event for large payload (#20221) RTS eventlog events for postCapsetVecEvent are truncated if payload is larger than EVENT_PAYLOAD_SIZE_MAX Previously, postCapsetVecEvent records eventlog event with payload of variable size larger than EVENT_PAYLOAD_SIZE_MAX (2^16) without any validation, resulting in corrupted data. For example, this happens when a Haskell binary is invoked with very long command line arguments exceeding 2^16 bytes (see #20221). Now we check the size of accumulated payload messages incrementally, and truncate the message just before the payload size exceeds EVENT_PAYLOAD_SIZE_MAX. RTS will warn the user with a message showing how many arguments are truncated. - - - - - 9ec76f61 by Cheng Shao at 2022-12-08T08:32:59-05:00 hadrian: don't add debug info to non-debug ways of rts Hadrian used to pass -g when building all ways of rts. It makes output binaries larger (especially so for wasm backend), and isn't needed by most users out there, so this patch removes that flag. In case the debug info is desired, we still pass -g3 when building the debug way, and there's also the debug_info flavour transformer which ensures -g3 is passed for all rts ways. - - - - - 7658cdd4 by Krzysztof Gogolewski at 2022-12-08T08:33:36-05:00 Restore show (typeRep @[]) == "[]" The Show instance for TypeRep [] has changed in 9.5 to output "List" because the name of the type constructor changed. This seems to be accidental and is inconsistent with TypeReps of saturated lists, which are printed as e.g. "[Int]". For now, I'm restoring the old behavior; in the future, maybe we should show TypeReps without puns (List, Tuple, Type). - - - - - 4ca722fe by Ben Gamari at 2022-12-08T14:43:26-05:00 Bump bounds to accomodate base-4.18 - - - - - 216deefd by Matthew Pickering at 2022-12-08T22:45:27-05:00 Add test for #22162 - - - - - 5d0a311f by Matthew Pickering at 2022-12-08T22:45:27-05:00 ci: Add job to test interface file determinism guarantees In this job we can run on every commit we add a test which builds the Cabal library twice and checks that the ABI hash and interface hash is stable across the two builds. * We run the test 20 times to try to weed out any race conditions due to `-j` * We run the builds in different temporary directories to try to weed out anything related to build directory affecting ABI or interface file hash. Fixes #22180 - - - - - 0a76d7d4 by Matthew Pickering at 2022-12-08T22:45:27-05:00 ci: Add job for testing interface stability across builds The idea is that both the bindists should product libraries with the same ABI and interface hash. So the job checks with ghc-pkg to make sure the computed ABI is the same. In future this job can be extended to check for the other facets of interface determinism. Fixes #22180 - - - - - 74c9bf91 by Matthew Pickering at 2022-12-08T22:45:27-05:00 backpack: Be more careful when adding together ImportAvails There was some code in the signature merging logic which added together the ImportAvails of the signature and the signature which was merged into it. This had the side-effect of making the merged signature depend on the signature (via a normal module dependency). The intention was to propagate orphan instances through the merge but this also messed up recompilation logic because we shouldn't be attempting to load B.hi when mergeing it. The fix is to just combine the part of ImportAvails that we intended to (transitive info, orphan instances and type family instances) rather than the whole thing. - - - - - d122e022 by Matthew Pickering at 2022-12-08T22:45:27-05:00 Fix mk_mod_usage_info if the interface file is not already loaded In #22217 it was observed that the order modules are compiled in affects the contents of an interface file. This was because a module dependended on another module indirectly, via a re-export but the interface file for this module was never loaded because the symbol was never used in the file. If we decide that we depend on a module then we jolly well ought to record this fact in the interface file! Otherwise it could lead to very subtle recompilation bugs if the dependency is not tracked and the module is updated. Therefore the best thing to do is just to make sure the file is loaded by calling the `loadSysInterface` function. This first checks the caches (like we did before) but then actually goes to find the interface on disk if it wasn't loaded. Fixes #22217 - - - - - ea25088d by lrzlin at 2022-12-08T22:46:06-05:00 Add initial support for LoongArch Architecture. - - - - - 9eb9d2f4 by Andrew Lelechenko at 2022-12-08T22:46:47-05:00 Update submodule mtl to 2.3.1, parsec to 3.1.15.1, haddock and Cabal to HEAD - - - - - 08d8fe2a by Andrew Lelechenko at 2022-12-08T22:46:47-05:00 Allow mtl-2.3 in hadrian - - - - - 3807a46c by Andrew Lelechenko at 2022-12-08T22:46:47-05:00 Support mtl-2.3 in check-exact - - - - - ef702a18 by Andrew Lelechenko at 2022-12-08T22:46:47-05:00 Fix tests - - - - - 3144e8ff by Sebastian Graf at 2022-12-08T22:47:22-05:00 Make (^) INLINE (#22324) So that we get to cancel away the allocation for the lazily used base. We can move `powImpl` (which *is* strict in the base) to the top-level so that we don't duplicate too much code and move the SPECIALISATION pragmas onto `powImpl`. The net effect of this change is that `(^)` plays along much better with inlining thresholds and loopification (#22227), for example in `x2n1`. Fixes #22324. - - - - - 1d3a8b8e by Matthew Pickering at 2022-12-08T22:47:59-05:00 Typeable: Fix module locations of some definitions in GHC.Types There was some confusion in Data.Typeable about which module certain wired-in things were defined in. Just because something is wired-in doesn't mean it comes from GHC.Prim, in particular things like LiftedRep and RuntimeRep are defined in GHC.Types and that's the end of the story. Things like Int#, Float# etc are defined in GHC.Prim as they have no Haskell definition site at all so we need to generate type representations for them (which live in GHC.Types). Fixes #22510 - - - - - 0f7588b5 by Sebastian Graf at 2022-12-08T22:48:34-05:00 Make `drop` and `dropWhile` fuse (#18964) I copied the fusion framework we have in place for `take`. T18964 asserts that we regress neither when fusion fires nor when it doesn't. Fixes #18964. - - - - - 26e71562 by Sebastian Graf at 2022-12-08T22:49:10-05:00 Do not strictify a DFun's parameter dictionaries (#22549) ... thus fixing #22549. The details are in the refurbished and no longer dead `Note [Do not strictify a DFun's parameter dictionaries]`. There's a regression test in T22549. - - - - - 36093407 by John Ericson at 2022-12-08T22:49:45-05:00 Delete `rts/package.conf.in` It is a relic of the Make build system. The RTS now uses a `package.conf` file generated the usual way by Cabal. - - - - - b0cc2fcf by Krzysztof Gogolewski at 2022-12-08T22:50:21-05:00 Fixes around primitive literals * The SourceText of primitive characters 'a'# did not include the #, unlike for other primitive literals 1#, 1##, 1.0#, 1.0##, "a"#. We can now remove the function pp_st_suffix, which was a hack to add the # back. * Negative primitive literals shouldn't use parentheses, as described in Note [Printing of literals in Core]. Added a testcase to T14681. - - - - - aacf616d by Bryan Richter at 2022-12-08T22:50:56-05:00 testsuite: Mark conc024 fragile on Windows - - - - - ed239a24 by Ryan Scott at 2022-12-09T09:42:16-05:00 Document TH splices' interaction with INCOHERENT instances Top-level declaration splices can having surprising interactions with `INCOHERENT` instances, as observed in #22492. This patch resolves #22492 by documenting this strange interaction in the GHC User's Guide. [ci skip] - - - - - 1023b432 by Mike Pilgrem at 2022-12-09T09:42:56-05:00 Fix #22300 Document GHC's extensions to valid whitespace - - - - - 79b0cec0 by Luite Stegeman at 2022-12-09T09:43:38-05:00 Add support for environments that don't have setImmediate - - - - - 5b007ec5 by Luite Stegeman at 2022-12-09T09:43:38-05:00 Fix bound thread status - - - - - 65335d10 by Matthew Pickering at 2022-12-09T20:15:45-05:00 Update containers submodule This contains a fix necessary for the multi-repl to work on GHC's code base where we try to load containers and template-haskell into the same session. - - - - - 4937c0bb by Matthew Pickering at 2022-12-09T20:15:45-05:00 hadrian-multi: Put interface files in separate directories Before we were putting all the interface files in the same directory which was leading to collisions if the files were called the same thing. - - - - - 8acb5b7b by Matthew Pickering at 2022-12-09T20:15:45-05:00 hadrian-toolargs: Add filepath to allowed repl targets - - - - - 5949d927 by Matthew Pickering at 2022-12-09T20:15:45-05:00 driver: Set correct UnitId when rehydrating modules We were not setting the UnitId before rehydrating modules which just led to us attempting to find things in the wrong HPT. The test for this is the hadrian-multi command (which is now added as a CI job). Fixes #22222 - - - - - ab06c0f0 by Matthew Pickering at 2022-12-09T20:15:45-05:00 ci: Add job to test hadrian-multi command I am not sure this job is good because it requires booting HEAD with HEAD, but it should be fine. - - - - - fac3e568 by Matthew Pickering at 2022-12-09T20:16:20-05:00 hadrian: Update bootstrap plans to 9.2.* series and 9.4.* series. This updates the build plans for the most recent compiler versions, as well as fixing the hadrian-bootstrap-gen script to a specific GHC version. - - - - - 195b08b4 by Matthew Pickering at 2022-12-09T20:16:20-05:00 ci: Bump boot images to use ghc-9.4.3 Also updates the bootstrap jobs to test booting 9.2 and 9.4. - - - - - c658c580 by Matthew Pickering at 2022-12-09T20:16:20-05:00 hlint: Removed redundant UnboxedSums pragmas UnboxedSums is quite confusingly implied by UnboxedTuples, alas, just the way it is. See #22485 - - - - - 340b7511 by Vladislav Zavialov at 2022-12-10T12:31:28+00:00 HsToken in HsAppKindTy - - - - - b3e98a92 by Oleg Grenrus at 2022-12-11T12:26:17-05:00 Add heqT, a kind-heterogeneous variant of heq CLC proposal https://github.com/haskell/core-libraries-committee/issues/99 - - - - - bfd7c1e6 by Andrew Lelechenko at 2022-12-11T12:26:55-05:00 Document that Bifunctor instances for tuples are lawful only up to laziness - - - - - 5d1a1881 by Bryan Richter at 2022-12-12T16:22:36-05:00 Mark T21336a fragile - - - - - c30accc2 by Matthew Pickering at 2022-12-12T16:23:11-05:00 Add test for #21476 This issues seems to have been fixed since the ticket was made, so let's add a test and move on. Fixes #21476 - - - - - 946226ec by Ben Gamari at 2022-12-13T20:12:56-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - e9d74a3e by Sebastian Graf at 2022-12-13T22:18:39-05:00 Respect -XStrict in the pattern-match checker (#21761) We were missing a call to `decideBangHood` in the pattern-match checker. There is another call in `matchWrapper.mk_eqn_info` which seems redundant but really is not; see `Note [Desugaring -XStrict matches in Pmc]`. Fixes #21761. - - - - - 884790e2 by Gergő Érdi at 2022-12-13T22:19:14-05:00 Fix loop in the interface representation of some `Unfolding` fields As discovered in #22272, dehydration of the unfolding info of a recursive definition used to involve a traversal of the definition itself, which in turn involves traversing the unfolding info. Hence, a loop. Instead, we now store enough data in the interface that we can produce the unfolding info without this traversal. See Note [Tying the 'CoreUnfolding' knot] for details. Fixes #22272 Co-authored-by: Simon Peyton Jones <simon.peytonjones at gmail.com> - - - - - 9f301189 by Alan Zimmerman at 2022-12-13T22:19:50-05:00 EPA: When splitting out header comments, keep ones for first decl Any comments immediately preceding the first declaration are no longer kept as header comments, but attach to the first declaration instead. - - - - - 8b1f1b45 by Sylvain Henry at 2022-12-13T22:20:28-05:00 JS: fix object file name comparison (#22578) - - - - - e9e161bb by Bryan Richter at 2022-12-13T22:21:03-05:00 configure: Bump min bootstrap GHC version to 9.2 - - - - - 75855643 by Ben Gamari at 2022-12-15T03:54:02-05:00 hadrian: Don't enable TSAN in stage0 build - - - - - da7b51d8 by Ben Gamari at 2022-12-15T03:54:02-05:00 cmm: Introduce blockConcat - - - - - 34f6b09c by Ben Gamari at 2022-12-15T03:54:02-05:00 cmm: Introduce MemoryOrderings - - - - - 43beaa7b by Ben Gamari at 2022-12-15T03:54:02-05:00 llvm: Respect memory specified orderings - - - - - 8faf74fc by Ben Gamari at 2022-12-15T03:54:02-05:00 Codegen/x86: Eliminate barrier for relaxed accesses - - - - - 6cc3944a by Ben Gamari at 2022-12-15T03:54:02-05:00 cmm/Parser: Reduce some repetition - - - - - 6c9862c4 by Ben Gamari at 2022-12-15T03:54:02-05:00 cmm/Parser: Add syntax for ordered loads and stores - - - - - 748490d2 by Ben Gamari at 2022-12-15T03:54:02-05:00 cmm/Parser: Atomic load syntax Originally I had thought I would just use the `prim` call syntax instead of introducing new syntax for atomic loads. However, it turns out that `prim` call syntax tends to make things quite unreadable. This new syntax seems quite natural. - - - - - 28c6781a by Ben Gamari at 2022-12-15T03:54:02-05:00 codeGen: Introduce ThreadSanitizer instrumentation This introduces a new Cmm pass which instruments the program with ThreadSanitizer annotations, allowing full tracking of mutator memory accesses via TSAN. - - - - - d97aa311 by Ben Gamari at 2022-12-15T03:54:02-05:00 Hadrian: Drop TSAN_ENABLED define from flavour This is redundant since the TSANUtils.h already defines it. - - - - - 86974ef1 by Ben Gamari at 2022-12-15T03:54:02-05:00 hadrian: Enable Cmm instrumentation in TSAN flavour - - - - - 93723290 by Ben Gamari at 2022-12-15T03:54:02-05:00 rts: Ensure that global regs are never passed as fun call args This is in general unsafe as they may be clobbered if they are mapped to caller-saved machine registers. See Note [Register parameter passing]. - - - - - 2eb0fb87 by Matthew Pickering at 2022-12-15T03:54:39-05:00 Package Imports: Get candidate packages also from re-exported modules Previously we were just looking at the direct imports to try and work out what a package qualifier could apply to but #22333 pointed out we also needed to look for reexported modules. Fixes #22333 - - - - - 552b7908 by Ben Gamari at 2022-12-15T03:55:15-05:00 compiler: Ensure that MutVar operations have necessary barriers Here we add acquire and release barriers in readMutVar# and writeMutVar#, which are necessary for soundness. Fixes #22468. - - - - - 933d61a4 by Simon Peyton Jones at 2022-12-15T03:55:51-05:00 Fix bogus test in Lint The Lint check for branch compatiblity within an axiom, in GHC.Core.Lint.compatible_branches was subtly different to the check made when contructing an axiom, in GHC.Core.FamInstEnv.compatibleBranches. The latter is correct, so I killed the former and am now using the latter. On the way I did some improvements to pretty-printing and documentation. - - - - - 03ed0b95 by Ryan Scott at 2022-12-15T03:56:26-05:00 checkValidInst: Don't expand synonyms when splitting sigma types Previously, the `checkValidInst` function (used when checking that an instance declaration is headed by an actual type class, not a type synonym) was using `tcSplitSigmaTy` to split apart the `forall`s and instance context. This is incorrect, however, as `tcSplitSigmaTy` expands type synonyms, which can cause instances headed by quantified constraint type synonyms to be accepted erroneously. This patch introduces `splitInstTyForValidity`, a variant of `tcSplitSigmaTy` specialized for validity checking that does _not_ expand type synonyms, and uses it in `checkValidInst`. Fixes #22570. - - - - - ed056bc3 by Ben Gamari at 2022-12-16T16:12:44-05:00 rts/Messages: Refactor This doesn't change behavior but makes the code a bit easier to follow. - - - - - 7356f8e0 by Ben Gamari at 2022-12-16T16:12:44-05:00 rts/ThreadPaused: Ordering fixes - - - - - 914f0025 by Ben Gamari at 2022-12-16T16:12:44-05:00 eventlog: Silence spurious data race - - - - - fbc84244 by Ben Gamari at 2022-12-16T16:12:44-05:00 Introduce SET_INFO_RELEASE for Cmm - - - - - 821b5472 by Ben Gamari at 2022-12-16T16:12:44-05:00 rts: Use fences instead of explicit barriers - - - - - 2228c999 by Ben Gamari at 2022-12-16T16:12:44-05:00 rts/stm: Fix memory ordering in readTVarIO# See #22421. - - - - - 99269b9f by Ben Gamari at 2022-12-16T16:12:44-05:00 Improve heap memory barrier Note Also introduce MUT_FIELD marker in Closures.h to document mutable fields. - - - - - 70999283 by Ben Gamari at 2022-12-16T16:12:44-05:00 rts: Introduce getNumCapabilities And ensure accesses to n_capabilities are atomic (although with relaxed ordering). This is necessary as RTS API callers may concurrently call into the RTS without holding a capability. - - - - - 98689f77 by Ben Gamari at 2022-12-16T16:12:44-05:00 ghc: Fix data race in dump file handling Previously the dump filename cache would use a non-atomic update which could potentially result in lost dump contents. Note that this is still a bit racy since the first writer may lag behind a later appending writer. - - - - - 605d9547 by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Always use atomics for context_switch and interrupt Since these are modified by the timer handler. - - - - - 86f20258 by Ben Gamari at 2022-12-16T16:12:45-05:00 rts/Timer: Always use atomic operations As noted in #22447, the existence of the pthread-based ITimer implementation means that we cannot assume that the program is single-threaded. - - - - - f8e901dc by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Encapsulate recent_activity access This makes it easier to ensure that it is accessed using the necessary atomic operations. - - - - - e0affaa9 by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Encapsulate access to capabilities array - - - - - 7ca683e4 by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Encapsulate sched_state - - - - - 1cf13bd0 by Ben Gamari at 2022-12-16T16:12:45-05:00 PrimOps: Fix benign MutVar race Relaxed ordering is fine here since the later CAS implies a release. - - - - - 3d2a7e08 by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Style fix - - - - - 82c62074 by Ben Gamari at 2022-12-16T16:12:45-05:00 compiler: Use release store in eager blackholing - - - - - eb1a0136 by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Fix ordering of makeStableName - - - - - ad0e260a by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Use ordered accesses instead of explicit barriers - - - - - a3eccf06 by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Statically allocate capabilities This is a rather simplistic way of solving #17289. - - - - - 287fa3fb by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Ensure that all accesses to pending_sync are atomic - - - - - 351eae58 by Ben Gamari at 2022-12-16T16:12:45-05:00 rts: Note race with wakeBlockingQueue - - - - - 5acf33dd by Andrew Lelechenko at 2022-12-16T16:13:22-05:00 Bump submodule directory to 1.3.8.0 and hpc to HEAD - - - - - 0dd95421 by Andrew Lelechenko at 2022-12-16T16:13:22-05:00 Accept allocations increase on Windows This is because of `filepath-1.4.100.0` and AFPP, causing increasing round-trips between lists and ByteArray. See #22625 for discussion. Metric Increase: MultiComponentModules MultiComponentModulesRecomp MultiLayerModules MultiLayerModulesRecomp T10421 T10547 T12150 T12227 T12234 T12425 T13035 T13253 T13253-spj T13701 T13719 T15703 T16875 T18140 T18282 T18304 T18698a T18698b T18923 T20049 T21839c T21839r T5837 T6048 T9198 T9961 TcPlugin_RewritePerf hard_hole_fits - - - - - ef9ac9d2 by Cheng Shao at 2022-12-16T16:13:59-05:00 testsuite: Mark T9405 as fragile instead of broken on Windows It's starting to pass again, and the unexpected pass blocks CI. - - - - - 1f3abd85 by Cheng Shao at 2022-12-16T21:16:28+00:00 compiler: remove obsolete commented code in wasm NCG It was just a temporary hack to workaround a bug in the relooper, that bug has been fixed long before the wasm backend is merged. - - - - - e3104eab by Cheng Shao at 2022-12-16T21:16:28+00:00 compiler: add missing export list of GHC.CmmToAsm.Wasm.FromCmm Also removes some unreachable code here. - - - - - 1c6930bf by Cheng Shao at 2022-12-16T21:16:28+00:00 compiler: change fallback function signature to Cmm function signature in wasm NCG In the wasm NCG, when handling a `CLabel` of undefined function without knowing its function signature, we used to fallback to `() -> ()` which is accepted by `wasm-ld`. This patch changes it to the signature of Cmm functions, which equally works, but would be required when we emit tail call instructions. - - - - - 8a81d9d9 by Cheng Shao at 2022-12-16T21:16:28+00:00 compiler: add optional tail-call support in wasm NCG When the `-mtail-call` clang flag is passed at configure time, wasm tail-call extension is enabled, and the wasm NCG will emit `return_call`/`return_call_indirect` instructions to take advantage of it and avoid the `StgRun` trampoline overhead. Closes #22461. - - - - - d1431cc0 by Cheng Shao at 2022-12-17T08:07:15-05:00 base: add missing autoconf checks for waitpid/umask These are not present in wasi-libc. Required for fixing #22589 - - - - - da3f1e91 by Cheng Shao at 2022-12-17T08:07:51-05:00 compiler: make .wasm the default executable extension on wasm32 Following convention as in other wasm toolchains. Fixes #22594. - - - - - ad21f4ef by Cheng Shao at 2022-12-17T08:07:51-05:00 ci: support hello.wasm in ci.sh cross testing logic - - - - - 6fe2d778 by amesgen at 2022-12-18T19:33:49-05:00 Correct `exitWith` Haddocks The `IOError`-specific `catch` in the Prelude is long gone. - - - - - b3eacd64 by Ben Gamari at 2022-12-18T19:34:24-05:00 rts: Drop racy assertion 0e274c39bf836d5bb846f5fa08649c75f85326ac added an assertion in `dirty_MUT_VAR` checking that the MUT_VAR being dirtied was clean. However, this isn't necessarily the case since another thread may have raced us to dirty the object. - - - - - 761c1f49 by Ben Gamari at 2022-12-18T19:35:00-05:00 rts/libdw: Silence uninitialized usage warnings As noted in #22538, previously some GCC versions warned that various locals in Libdw.c may be used uninitialized. Although this wasn't strictly true (since they were initialized in an inline assembler block) we fix this by providing explicit empty initializers. Fixes #22538 - - - - - 5e047eff by Matthew Pickering at 2022-12-20T15:12:04+00:00 testsuite: Mark T16392 as fragile on windows See #22649 - - - - - 703a4665 by M Farkas-Dyck at 2022-12-20T21:14:46-05:00 Scrub some partiality in `GHC.Cmm.Info.Build`: `doSRTs` takes a `[(CAFSet, CmmDecl)]` but truly wants a `[(CAFSet, CmmStatics)]`. - - - - - 9736ab74 by Matthew Pickering at 2022-12-20T21:15:22-05:00 packaging: Fix upload_ghc_libs.py script This change reflects the changes where .cabal files are now generated by hadrian rather than ./configure. Fixes #22518 - - - - - 7c6de18d by Ben Gamari at 2022-12-20T21:15:57-05:00 configure: Drop uses of AC_PROG_CC_C99 As noted in #22566, this macro is deprecated as of autoconf-2.70 `AC_PROG_CC` now sets `ac_cv_prog_cc_c99` itself. Closes #22566. - - - - - 36c5d98e by Ben Gamari at 2022-12-20T21:15:57-05:00 configure: Use AS_HELP_STRING instead of AC_HELP_STRING The latter has been deprecated. See #22566. - - - - - befe6ff8 by Andrew Lelechenko at 2022-12-20T21:16:37-05:00 GHCi.UI: fix various usages of head and tail - - - - - 666d0ba7 by Andrew Lelechenko at 2022-12-20T21:16:37-05:00 GHCi.UI: avoid head and tail in parseCallEscape and around - - - - - 5d96fd50 by Andrew Lelechenko at 2022-12-20T21:16:37-05:00 Make GHC.Driver.Main.hscTcRnLookupRdrName to return NonEmpty - - - - - 3ce2ab94 by Andrew Lelechenko at 2022-12-21T06:17:56-05:00 Allow transformers-0.6 in ghc, ghci, ghc-bin and hadrian - - - - - 954de93a by Andrew Lelechenko at 2022-12-21T06:17:56-05:00 Update submodule haskeline to HEAD (to allow transformers-0.6) - - - - - cefbeec3 by Andrew Lelechenko at 2022-12-21T06:17:56-05:00 Update submodule transformers to 0.6.0.4 - - - - - b4730b62 by Andrew Lelechenko at 2022-12-21T06:17:56-05:00 Fix tests T13253 imports MonadTrans, which acquired a quantified constraint in transformers-0.6, thus increase in allocations Metric Increase: T13253 - - - - - 0be75261 by Simon Peyton Jones at 2022-12-21T06:18:32-05:00 Abstract over the right free vars Fix #22459, in two ways: (1) Make the Specialiser not create a bogus specialisation if it is presented by strangely polymorphic dictionary. See Note [Weird special case in SpecDict] in GHC.Core.Opt.Specialise (2) Be more careful in abstractFloats See Note [Which type variables to abstract over] in GHC.Core.Opt.Simplify.Utils. So (2) stops creating the excessively polymorphic dictionary in abstractFloats, while (1) stops crashing if some other pass should nevertheless create a weirdly polymorphic dictionary. - - - - - df7bc6b3 by Ying-Ruei Liang (TheKK) at 2022-12-21T14:31:54-05:00 rts: explicitly store return value of ccall checkClosure to prevent type error (#22617) - - - - - e193e537 by Simon Peyton Jones at 2022-12-21T14:32:30-05:00 Fix shadowing lacuna in OccurAnal Issue #22623 demonstrated another lacuna in the implementation of wrinkle (BS3) in Note [The binder-swap substitution] in the occurrence analyser. I was failing to add TyVar lambda binders using addInScope/addOneInScope and that led to a totally bogus binder-swap transformation. Very easy to fix. - - - - - 3d55d8ab by Simon Peyton Jones at 2022-12-21T14:32:30-05:00 Fix an assertion check in addToEqualCtList The old assertion saw that a constraint ct could rewrite itself (of course it can) and complained (stupid). Fixes #22645 - - - - - ceb2e9b9 by Ben Gamari at 2022-12-21T15:26:08-05:00 configure: Bump version to 9.6 - - - - - fb4d36c4 by Ben Gamari at 2022-12-21T15:27:49-05:00 base: Bump version to 4.18 Requires various submodule bumps. - - - - - 93ee7e90 by Ben Gamari at 2022-12-21T15:27:49-05:00 ghc-boot: Fix bootstrapping - - - - - fd8faa66 by Ben Gamari at 2022-12-22T13:44:28-05:00 Bump GHC version to 9.7 - - - - - fc3a2232 by Ben Gamari at 2022-12-22T13:45:06-05:00 Bump GHC version to 9.7 - - - - - 2958aa9c by Ben Gamari at 2022-12-22T14:49:16-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 914f7fe3 by Andreas Klebinger at 2022-12-22T23:36:10-05:00 Don't consider large byte arrays/compact regions pinned. Workaround for #22255 which showed how treating large/compact regions as pinned could cause segfaults. - - - - - 32b32d7f by Matthew Pickering at 2022-12-22T23:36:46-05:00 hadrian bindist: Install manpages to share/man/man1/ghc.1 When the installation makefile was copied over the manpages were no longer installed in the correct place. Now we install it into share/man/man1/ghc.1 as the make build system did. Fixes #22371 - - - - - b3ddf803 by Ben Gamari at 2022-12-22T23:37:23-05:00 rts: Drop paths from configure from cabal file A long time ago we would rely on substitutions from the configure script to inject paths of the include and library directories of libffi and libdw. However, now these are instead handled inside Hadrian when calling Cabal's `configure` (see the uses of `cabalExtraDirs` in Hadrian's `Settings.Packages.packageArgs`). While the occurrences in the cabal file were redundant, they did no harm. However, since b5c714545abc5f75a1ffdcc39b4bfdc7cd5e64b4 they have no longer been interpolated. @mpickering noticed the suspicious uninterpolated occurrence of `@FFIIncludeDir@` in #22595, prompting this commit to finally remove them. - - - - - b2c7523d by Ben Gamari at 2022-12-22T23:37:59-05:00 Bump libffi-tarballs submodule We will now use libffi-3.4.4. - - - - - 3699a554 by Alan Zimmerman at 2022-12-22T23:38:35-05:00 EPA: Make EOF position part of AnnsModule Closes #20951 Closes #19697 - - - - - 99757ce8 by Sylvain Henry at 2022-12-22T23:39:13-05:00 JS: fix support for -outputdir (#22641) The `-outputdir` option wasn't correctly handled with the JS backend because the same code path was used to handle both objects produced by the JS backend and foreign .js files. Now we clearly distinguish the two in the pipeline, fixing the bug. - - - - - 02ed7d78 by Simon Peyton Jones at 2022-12-22T23:39:49-05:00 Refactor mkRuntimeError This patch fixes #22634. Because we don't have TYPE/CONSTRAINT polymorphism, we need two error functions rather than one. I took the opportunity to rname runtimeError to impossibleError, to line up with mkImpossibleExpr, and avoid confusion with the genuine runtime-error-constructing functions. - - - - - 35267f07 by Ben Gamari at 2022-12-22T23:40:32-05:00 base: Fix event manager shutdown race on non-Linux platforms During shutdown it's possible that we will attempt to use a closed fd to wakeup another capability's event manager. On the Linux eventfd path we were careful to handle this. However on the non-Linux path we failed to do so. Fix this. - - - - - 317f45c1 by Simon Peyton Jones at 2022-12-22T23:41:07-05:00 Fix unifier bug: failing to decompose over-saturated type family This simple patch fixes #22647 - - - - - 14b2e3d3 by Ben Gamari at 2022-12-22T23:41:42-05:00 rts/m32: Fix sanity checking Previously we would attempt to clear pages which were marked as read-only. Fix this. - - - - - 16a1bcd1 by Matthew Pickering at 2022-12-23T09:15:24+00:00 ci: Move wasm pipelines into nightly rather than master See #22664 for the changes which need to be made to bring one of these back to the validate pipeline. - - - - - 18d2acd2 by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Fix race in marking of blackholes We must use an acquire-fence when marking to ensure that the indirectee is visible. - - - - - 11241efa by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Fix segment list races - - - - - 602455c9 by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Use atomic when looking at bd->gen Since it may have been mutated by a moving GC. - - - - - 9d63b160 by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Eliminate race in bump_static_flag To ensure that we don't race with a mutator entering a new CAF we take the SM mutex before touching static_flag. The other option here would be to instead modify newCAF to use a CAS but the present approach is a bit safer. - - - - - 26837523 by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Ensure that mutable fields have acquire barrier - - - - - 8093264a by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Fix races in collector status tracking Mark a number of accesses to do with tracking of the status of the concurrent collection thread as atomic. No interesting races here, merely necessary to satisfy TSAN. - - - - - 387d4fcc by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Make segment state updates atomic - - - - - 543cae00 by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Refactor update remembered set initialization This avoids a lock inversion between the storage manager mutex and the stable pointer table mutex by not dropping the SM_MUTEX in nonmovingCollect. This requires quite a bit of rejiggering but it does seem like a better strategy. - - - - - c9936718 by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Ensure that we aren't holding locks when closing them TSAN complains about this sort of thing. - - - - - 0cd31f7d by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Make bitmap accesses atomic This is a benign race on any sensible hard since these are byte accesses. Nevertheless, atomic accesses are necessary to satisfy TSAN. - - - - - d3fe110a by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Fix benign race in update remembered set check Relaxed load is fine here since we will take the lock before looking at the list. - - - - - ab6cf893 by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Fix race in shortcutting We must use an acquire load to read the info table pointer since if we find an indirection we must be certain that we see the indirectee. - - - - - 36c9f23c by Ben Gamari at 2022-12-23T19:09:30-05:00 nonmoving: Make free list counter accesses atomic Since these may race with the allocator(s). - - - - - aebef31c by doyougnu at 2022-12-23T19:10:09-05:00 add GHC.Utils.Binary.foldGet' and use for Iface A minor optimization to remove lazy IO and a lazy accumulator strictify foldGet' IFace.Binary: use strict foldGet' remove superfluous bang - - - - - 5eb357d9 by Ben Gamari at 2022-12-24T00:41:05-05:00 compiler: Ensure that GHC toolchain is first in search path As noted in #22561, it is important that GHC's toolchain look first for its own headers and libraries to ensure that the system's are not found instead. If this happens things can break in surprising ways (e.g. see #22561). - - - - - cbaebfb9 by Matthew Pickering at 2022-12-24T00:41:40-05:00 head.hackage: Use slow-validate bindist for linting jobs This enables the SLOW_VALIDATE env var for the linting head.hackage jobs, namely the jobs enabled manually, by the label or on the nightly build now use the deb10-numa-slow-validate bindist which has assertions enabled. See #22623 for a ticket which was found by using this configuration already! The head.hackage jobs triggered by upstream CI are now thusly: hackage-lint: Can be triggered on any MR, normal validate pipeline or nightly build. Runs head.hackage with -dlint and a slow-validate bindist hackage-label-lint: Trigged on MRs with "user-facing" label, runs the slow-validate head.hackage build with -dlint. nightly-hackage-lint: Runs automatically on nightly pipelines with slow-validate + dlint config. nightly-hackage-perf: Runs automaticaly on nightly pipelines with release build and eventlogging enabled. release-hackage-lint: Runs automatically on release pipelines with -dlint on a release bindist. - - - - - f4850f36 by Matthew Pickering at 2022-12-24T00:41:40-05:00 ci: Don't run abi-test-nightly on release jobs The test is not configured to get the correct dependencies for the release pipelines (and indeed stops the release pipeline being run at all) - - - - - c264b06b by Matthew Pickering at 2022-12-24T00:41:40-05:00 ci: Run head.hackage jobs on upstream-testing branch rather than master This change allows less priviledged users to trigger head.hackage jobs because less permissions are needed to trigger jobs on the upstream-testing branch, which is not protected. There is a CI job which updates upstream-testing each hour to the state of the master branch so it should always be relatively up-to-date. - - - - - 63b97430 by Ben Gamari at 2022-12-24T00:42:16-05:00 llvmGen: Fix relaxed ordering Previously I used LLVM's `unordered` ordering for the C11 `relaxed` ordering. However, this is wrong and should rather use the LLVM `monotonic` ordering. Fixes #22640 - - - - - f42ba88f by Ben Gamari at 2022-12-24T00:42:16-05:00 gitlab-ci: Introduce aarch64-linux-llvm job This nightly job will ensure that we don't break the LLVM backend on AArch64/Linux by bootstrapping GHC. This would have caught #22640. - - - - - 6d62f6bf by Matthew Pickering at 2022-12-24T00:42:51-05:00 Store RdrName rather than OccName in Holes In #20472 it was pointed out that you couldn't defer out of scope but the implementation collapsed a RdrName into an OccName to stuff it into a Hole. This leads to the error message for a deferred qualified name dropping the qualification which affects the quality of the error message. This commit adds a bit more structure to a hole, so a hole can replace a RdrName without losing information about what that RdrName was. This is important when printing error messages. I also added a test which checks the Template Haskell deferral of out of scope qualified names works properly. Fixes #22130 - - - - - 3c3060e4 by Richard Eisenberg at 2022-12-24T17:34:19+00:00 Drop support for kind constraints. This implements proposal 547 and closes ticket #22298. See the proposal and ticket for motivation. Compiler perf improves a bit Metrics: compile_time/bytes allocated ------------------------------------- CoOpt_Singletons(normal) -2.4% GOOD T12545(normal) +1.0% T13035(normal) -13.5% GOOD T18478(normal) +0.9% T9872d(normal) -2.2% GOOD geo. mean -0.2% minimum -13.5% maximum +1.0% Metric Decrease: CoOpt_Singletons T13035 T9872d - - - - - 6d7d4393 by Ben Gamari at 2022-12-24T21:09:56-05:00 hadrian: Ensure that linker scripts are used when merging objects In #22527 @rui314 inadvertantly pointed out a glaring bug in Hadrian's implementation of the object merging rules: unlike the old `make` build system we utterly failed to pass the needed linker scripts. Fix this. - - - - - a5bd0eb8 by Andrew Lelechenko at 2022-12-24T21:10:34-05:00 Document infelicities of instance Ord Double and workarounds - - - - - 62b9a7b2 by Zubin Duggal at 2023-01-03T12:22:11+00:00 Force the Docs structure to prevent leaks in GHCi with -haddock without -fwrite-interface Involves adding many new NFData instances. Without forcing Docs, references to the TcGblEnv for each module are retained by the Docs structure. Usually these are forced when the ModIface is serialised but not when we aren't writing the interface. - - - - - 21bedd84 by Facundo Domínguez at 2023-01-03T23:27:30-05:00 Explain the auxiliary functions of permutations - - - - - 32255d05 by Matthew Pickering at 2023-01-04T11:58:42+00:00 compiler: Add -f[no-]split-sections flags Here we add a `-fsplit-sections` flag which may some day replace `-split-sections`. This has the advantage of automatically providing a `-fno-split-sections` flag, which is useful for our packaging because we enable `-split-sections` by default but want to disable it in certain configurations. - - - - - e640940c by Matthew Pickering at 2023-01-04T11:58:42+00:00 hadrian: Fix computation of tables_next_to_code for outOfTreeCompiler This copy-pasto was introduced in de5fb3489f2a9bd6dc75d0cb8925a27fe9b9084b - - - - - 15bee123 by Matthew Pickering at 2023-01-04T11:58:42+00:00 hadrian: Add test:all_deps to build just testsuite dependencies Fixes #22534 - - - - - fec6638e by Matthew Pickering at 2023-01-04T11:58:42+00:00 hadrian: Add no_split_sections tranformer This transformer reverts the effect of `split_sections`, which we intend to use for platforms which don't support split sections. In order to achieve this we have to modify the implemntation of the split_sections transformer to store whether we are enabling split_sections directly in the `Flavour` definition. This is because otherwise there's no convenient way to turn off split_sections due to having to pass additional linker scripts when merging objects. - - - - - 3dc05726 by Matthew Pickering at 2023-01-04T11:58:42+00:00 check-exact: Fix build with -Werror - - - - - 53a6ae7a by Matthew Pickering at 2023-01-04T11:58:42+00:00 ci: Build all test dependencies with in-tree compiler This means that these executables will honour flavour transformers such as "werror". Fixes #22555 - - - - - 32e264c1 by Matthew Pickering at 2023-01-04T11:58:42+00:00 hadrian: Document using GHC environment variable to select boot compiler Fixes #22340 - - - - - be9dd9b0 by Matthew Pickering at 2023-01-04T11:58:42+00:00 packaging: Build perf builds with -split-sections In 8f71d958 the make build system was made to use split-sections on linux systems but it appears this logic never made it to hadrian. There is the split_sections flavour transformer but this doesn't appear to be used for perf builds on linux. This is disbled on deb9 and windows due to #21670 Closes #21135 - - - - - 00dc5106 by Matthew Pickering at 2023-01-04T14:32:45-05:00 sphinx: Use modern syntax for extlinks This fixes the following build error: ``` Command line: /opt/homebrew/opt/sphinx-doc/bin/sphinx-build -b man -d /private/tmp/extra-dir-55768274273/.doctrees-man -n -w /private/tmp/extra-dir-55768274273/.log docs/users_guide /private/tmp/extra-dir-55768274273 ===> Command failed with error code: 2 Exception occurred: File "/opt/homebrew/Cellar/sphinx-doc/6.0.0/libexec/lib/python3.11/site-packages/sphinx/ext/extlinks.py", line 101, in role title = caption % part ~~~~~~~~^~~~~~ TypeError: not all arguments converted during string formatting ``` I tested on Sphinx-5.1.1 and Sphinx-6.0.0 Thanks for sterni for providing instructions about how to test using sphinx-6.0.0. Fixes #22690 - - - - - 541aedcd by Krzysztof Gogolewski at 2023-01-05T10:48:34-05:00 Misc cleanup - Remove unused uniques and hs-boot declarations - Fix types of seq and unsafeCoerce# - Remove FastString/String roundtrip in JS - Use TTG to enforce totality - Remove enumeration in Heap/Inspect; the 'otherwise' clause serves the primitive types well. - - - - - 22bb8998 by Alan Zimmerman at 2023-01-05T10:49:09-05:00 EPA: Do not collect comments from end of file In Parser.y semis1 production triggers for the virtual semi at the end of the file. This is detected by it being zero length. In this case, do not extend the span being used to gather comments, so any final comments are allocated at the module level instead. - - - - - 9e077999 by Vladislav Zavialov at 2023-01-05T23:01:55-05:00 HsToken in TypeArg (#19623) Updates the haddock submodule. - - - - - b2a2db04 by Matthew Pickering at 2023-01-05T23:02:30-05:00 Revert "configure: Drop uses of AC_PROG_CC_C99" This reverts commit 7c6de18dd3151ead954c210336728e8686c91de6. Centos7 using a very old version of the toolchain (autotools-2.69) where the behaviour of these macros has not yet changed. I am reverting this without haste as it is blocking the 9.6 branch. Fixes #22704 - - - - - 28f8c0eb by Luite Stegeman at 2023-01-06T18:16:24+09:00 Add support for sized literals in the bytecode interpreter. The bytecode interpreter only has branching instructions for word-sized values. These are used for pattern matching. Branching instructions for other types (e.g. Int16# or Word8#) weren't needed, since unoptimized Core or STG never requires branching on types like this. It's now possible for optimized STG to reach the bytecode generator (e.g. fat interface files or certain compiler flag combinations), which requires dealing with various sized literals in branches. This patch improves support for generating bytecode from optimized STG by adding the following new bytecode instructions: TESTLT_I64 TESTEQ_I64 TESTLT_I32 TESTEQ_I32 TESTLT_I16 TESTEQ_I16 TESTLT_I8 TESTEQ_I8 TESTLT_W64 TESTEQ_W64 TESTLT_W32 TESTEQ_W32 TESTLT_W16 TESTEQ_W16 TESTLT_W8 TESTEQ_W8 Fixes #21945 - - - - - ac39e8e9 by Matthew Pickering at 2023-01-06T13:47:00-05:00 Only store Name in FunRhs rather than Id with knot-tied fields All the issues here have been caused by #18758. The goal of the ticket is to be able to talk about things like `LTyClDecl GhcTc`. In the case of HsMatchContext, the correct "context" is whatever we want, and in fact storing just a `Name` is sufficient and correct context, even if the rest of the AST is storing typechecker Ids. So this reverts (#20415, !5579) which intended to get closed to #18758 but didn't really and introduced a few subtle bugs. Printing of an error message in #22695 would just hang, because we would attempt to print the `Id` in debug mode to assertain whether it was empty or not. Printing the Name is fine for the error message. Another consequence is that when `-dppr-debug` was enabled the compiler would hang because the debug printing of the Id would try and print fields which were not populated yet. This also led to 32070e6c2e1b4b7c32530a9566fe14543791f9a6 having to add a workaround for the `checkArgs` function which was probably a very similar bug to #22695. Fixes #22695 - - - - - c306d939 by Matthew Pickering at 2023-01-06T22:08:53-05:00 ci: Upgrade darwin, windows and freebsd CI to use GHC-9.4.3 Fixes #22599 - - - - - 0db496ff by Matthew Pickering at 2023-01-06T22:08:53-05:00 darwin ci: Explicitly pass desired build triple to configure On the zw3rk machines for some reason the build machine was inferred to be arm64. Setting the build triple appropiately resolve this confusion and we produce x86 binaries. - - - - - 2459c358 by Ben Gamari at 2023-01-06T22:09:29-05:00 rts: MUT_VAR is not a StgMutArrPtrs There was previously a comment claiming that the MUT_VAR closure type had the layout of StgMutArrPtrs. - - - - - 6206cb92 by Simon Peyton Jones at 2023-01-07T12:14:40-05:00 Make FloatIn robust to shadowing This MR fixes #22622. See the new Note [Shadowing and name capture] I did a bit of refactoring in sepBindsByDropPoint too. The bug doesn't manifest in HEAD, but it did show up in 9.4, so we should backport this patch to 9.4 - - - - - a960ca81 by Matthew Pickering at 2023-01-07T12:15:15-05:00 T10955: Set DYLD_LIBRARY_PATH for darwin The correct path to direct the dynamic linker on darwin is DYLD_LIBRARY_PATH rather than LD_LIBRARY_PATH. On recent versions of OSX using LD_LIBRARY_PATH seems to have stopped working. For more reading see: https://stackoverflow.com/questions/3146274/is-it-ok-to-use-dyld-library-path-on-mac-os-x-and-whats-the-dynamic-library-s - - - - - 73484710 by Matthew Pickering at 2023-01-07T12:15:15-05:00 Skip T18623 on darwin (to add to the long list of OSs) On recent versions of OSX, running `ulimit -v` results in ``` ulimit: setrlimit failed: invalid argument ``` Time is too short to work out what random stuff Apple has been doing with ulimit, so just skip the test like we do for other platforms. - - - - - 8c0ea25f by Matthew Pickering at 2023-01-07T12:15:15-05:00 Pass -Wl,-no_fixup_chains to ld64 when appropiate Recent versions of MacOS use a version of ld where `-fixup_chains` is on by default. This is incompatible with our usage of `-undefined dynamic_lookup`. Therefore we explicitly disable `fixup-chains` by passing `-no_fixup_chains` to the linker on darwin. This results in a warning of the form: ld: warning: -undefined dynamic_lookup may not work with chained fixups The manual explains the incompatible nature of these two flags: -undefined treatment Specifies how undefined symbols are to be treated. Options are: error, warning, suppress, or dynamic_lookup. The default is error. Note: dynamic_lookup that depends on lazy binding will not work with chained fixups. A relevant ticket is #22429 Here are also a few other links which are relevant to the issue: Official comment: https://developer.apple.com/forums/thread/719961 More relevant links: https://openradar.appspot.com/radar?id=5536824084660224 https://github.com/python/cpython/issues/97524 Note in release notes: https://developer.apple.com/documentation/xcode-release-notes/xcode-13-releas e-notes - - - - - 365b3045 by Matthew Pickering at 2023-01-09T02:36:20-05:00 Disable split sections on aarch64-deb10 build See #22722 Failure on this job: https://gitlab.haskell.org/ghc/ghc/-/jobs/1287852 ``` Unexpected failures: /builds/ghc/ghc/tmp/ghctest-s3d8g1hj/test spaces/testsuite/tests/th/T10828.run T10828 [exit code non-0] (ext-interp) /builds/ghc/ghc/tmp/ghctest-s3d8g1hj/test spaces/testsuite/tests/th/T13123.run T13123 [exit code non-0] (ext-interp) /builds/ghc/ghc/tmp/ghctest-s3d8g1hj/test spaces/testsuite/tests/th/T20590.run T20590 [exit code non-0] (ext-interp) Appending 232 stats to file: /builds/ghc/ghc/performance-metrics.tsv ``` ``` Compile failed (exit code 1) errors were: data family D_0 a_1 :: * -> * data instance D_0 GHC.Types.Int GHC.Types.Bool :: * where DInt_2 :: D_0 GHC.Types.Int GHC.Types.Bool data E_3 where MkE_4 :: a_5 -> E_3 data Foo_6 a_7 b_8 where MkFoo_9, MkFoo'_10 :: a_11 -> Foo_6 a_11 b_12 newtype Bar_13 :: * -> GHC.Types.Bool -> * where MkBar_14 :: a_15 -> Bar_13 a_15 b_16 data T10828.T (a_0 :: *) where T10828.MkT :: forall (a_1 :: *) . a_1 -> a_1 -> T10828.T a_1 T10828.MkC :: forall (a_2 :: *) (b_3 :: *) . (GHC.Types.~) a_2 GHC.Types.Int => {T10828.foo :: a_2, T10828.bar :: b_3} -> T10828.T GHC.Types.Int T10828.hs:1:1: error: [GHC-87897] Exception when trying to run compile-time code: ghc-iserv terminated (-4) Code: (do TyConI dec <- runQ $ reify (mkName "T") runIO $ putStrLn (pprint dec) >> hFlush stdout d <- runQ $ [d| data T' a :: Type where MkT' :: a -> a -> T' a MkC' :: forall a b. (a ~ Int) => {foo :: a, bar :: b} -> T' Int |] runIO $ putStrLn (pprint d) >> hFlush stdout ....) *** unexpected failure for T10828(ext-interp) =====> 7000 of 9215 [0, 1, 0] =====> 7000 of 9215 [0, 1, 0] =====> 7000 of 9215 [0, 1, 0] =====> 7000 of 9215 [0, 1, 0] Compile failed (exit code 1) errors were: T13123.hs:1:1: error: [GHC-87897] Exception when trying to run compile-time code: ghc-iserv terminated (-4) Code: ([d| data GADT where MkGADT :: forall k proxy (a :: k). proxy a -> GADT |]) *** unexpected failure for T13123(ext-interp) =====> 7100 of 9215 [0, 2, 0] =====> 7100 of 9215 [0, 2, 0] =====> 7200 of 9215 [0, 2, 0] Compile failed (exit code 1) errors were: T20590.hs:1:1: error: [GHC-87897] Exception when trying to run compile-time code: ghc-iserv terminated (-4) Code: ([d| data T where MkT :: forall a. a -> T |]) *** unexpected failure for T20590(ext-interp) ``` Looks fairly worrying to me. - - - - - 965a2735 by Alan Zimmerman at 2023-01-09T02:36:20-05:00 EPA: exact print HsDocTy To match ghc-exactprint https://github.com/alanz/ghc-exactprint/pull/121 - - - - - 5d65773e by John Ericson at 2023-01-09T20:39:27-05:00 Remove RTS hack for configuring See the brand new Note [Undefined symbols in the RTS] for additional details. - - - - - e3fff751 by Sebastian Graf at 2023-01-09T20:40:02-05:00 Handle shadowing in DmdAnal (#22718) Previously, when we had a shadowing situation like ```hs f x = ... -- demand signature <1L><1L> main = ... \f -> f 1 ... ``` we'd happily use the shadowed demand signature at the call site inside the lambda. Of course, that's wrong and solution is simply to remove the demand signature from the `AnalEnv` when we enter the lambda. This patch does so for all binding constructs Core. In #22718 the issue was caused by LetUp not shadowing away the existing demand signature for the let binder in the let body. The resulting absent error is fickle to reproduce; hence no reproduction test case. #17478 would help. Fixes #22718. It appears that TcPlugin_Rewrite regresses by ~40% on Darwin. It is likely that DmdAnal was exploiting ill-scoped analysis results. Metric increase ['bytes allocated'] (test_env=x86_64-darwin-validate): TcPlugin_Rewrite - - - - - d53f6f4d by Oleg Grenrus at 2023-01-09T21:11:02-05:00 Add safe list indexing operator: !? With Joachim's amendments. Implements https://github.com/haskell/core-libraries-committee/issues/110 - - - - - cfaf1ad7 by Nicolas Trangez at 2023-01-09T21:11:03-05:00 rts, tests: limit thread name length to 15 bytes On Linux, `pthread_setname_np` (or rather, the kernel) only allows for thread names up to 16 bytes, including the terminating null byte. This commit adds a note pointing this out in `createOSThread`, and fixes up two instances where a thread name of more than 15 characters long was used (in the RTS, and in a test-case). Fixes: #22366 Fixes: https://gitlab.haskell.org/ghc/ghc/-/issues/22366 See: https://gitlab.haskell.org/ghc/ghc/-/issues/22366#note_460796 - - - - - 64286132 by Matthew Pickering at 2023-01-09T21:11:03-05:00 Store bootstrap_llvm_target and use it to set LlvmTarget in bindists This mirrors some existing logic for the bootstrap_target which influences how TargetPlatform is set. As described on #21970 not storing this led to `LlvmTarget` being set incorrectly and hence the wrong `--target` flag being passed to the C compiler. Towards #21970 - - - - - 4724e8d1 by Matthew Pickering at 2023-01-09T21:11:04-05:00 Check for FP_LD_NO_FIXUP_CHAINS in installation configure script Otherwise, when installing from a bindist the C flag isn't passed to the C compiler. This completes the fix for #22429 - - - - - 2e926b88 by Georgi Lyubenov at 2023-01-09T21:11:07-05:00 Fix outdated link to Happy section on sequences - - - - - 146a1458 by Matthew Pickering at 2023-01-09T21:11:07-05:00 Revert "NCG(x86): Compile add+shift as lea if possible." This reverts commit 20457d775885d6c3df020d204da9a7acfb3c2e5a. See #22666 and #21777 - - - - - 6e6adbe3 by Jade Lovelace at 2023-01-11T00:55:30-05:00 Fix tcPluginRewrite example - - - - - faa57138 by Jade Lovelace at 2023-01-11T00:55:31-05:00 fix missing haddock pipe - - - - - 0470ea7c by Florian Weimer at 2023-01-11T00:56:10-05:00 m4/fp_leading_underscore.m4: Avoid implicit exit function declaration And switch to a new-style function definition. Fixes build issues with compilers that do not accept implicit function declarations. - - - - - b2857df4 by HaskellMouse at 2023-01-11T00:56:52-05:00 Added a new warning about compatibility with RequiredTypeArguments This commit introduces a new warning that indicates code incompatible with future extension: RequiredTypeArguments. Enabling this extension may break some code and the warning will help to make it compatible in advance. - - - - - 5f17e21a by Ben Gamari at 2023-01-11T00:57:27-05:00 testsuite: Drop testheapalloced.c As noted in #22414, this file (which appears to be a benchmark for characterising the one-step allocator's MBlock cache) is currently unreferenced. Remove it. Closes #22414. - - - - - bc125775 by Vladislav Zavialov at 2023-01-11T00:58:03-05:00 Introduce the TypeAbstractions language flag GHC Proposals #448 "Modern scoped type variables" and #425 "Invisible binders in type declarations" introduce a new language extension flag: TypeAbstractions. Part of the functionality guarded by this flag has already been implemented, namely type abstractions in constructor patterns, but it was guarded by a combination of TypeApplications and ScopedTypeVariables instead of a dedicated language extension flag. This patch does the following: * introduces a new language extension flag TypeAbstractions * requires TypeAbstractions for @a-syntax in constructor patterns instead of TypeApplications and ScopedTypeVariables * creates a User's Guide page for TypeAbstractions and moves the "Type Applications in Patterns" section there To avoid a breaking change, the new flag is implied by ScopedTypeVariables and is retroactively added to GHC2021. Metric Decrease: MultiLayerModulesTH_OneShot - - - - - 083f7015 by Krzysztof Gogolewski at 2023-01-11T00:58:38-05:00 Misc cleanup - Remove unused mkWildEvBinder - Use typeTypeOrConstraint - more symmetric and asserts that that the type is Type or Constraint - Fix escape sequences in Python; they raise a deprecation warning with -Wdefault - - - - - aed1974e by Richard Eisenberg at 2023-01-11T08:30:42+00:00 Refactor the treatment of loopy superclass dicts This patch completely re-engineers how we deal with loopy superclass dictionaries in instance declarations. It fixes #20666 and #19690 The highlights are * Recognise that the loopy-superclass business should use precisely the Paterson conditions. This is much much nicer. See Note [Recursive superclasses] in GHC.Tc.TyCl.Instance * With that in mind, define "Paterson-smaller" in Note [Paterson conditions] in GHC.Tc.Validity, and the new data type `PatersonSize` in GHC.Tc.Utils.TcType, along with functions to compute and compare PatsonSizes * Use the new PatersonSize stuff when solving superclass constraints See Note [Solving superclass constraints] in GHC.Tc.TyCl.Instance * In GHC.Tc.Solver.Monad.lookupInInerts, add a missing call to prohibitedSuperClassSolve. This was the original cause of #20666. * Treat (TypeError "stuff") as having PatersonSize zero. See Note [Paterson size for type family applications] in GHC.Tc.Utils.TcType. * Treat the head of a Wanted quantified constraint in the same way as the superclass of an instance decl; this is what fixes #19690. See GHC.Tc.Solver.Canonical Note [Solving a Wanted forall-constraint] (Thanks to Matthew Craven for this insight.) This entailed refactoring the GivenSc constructor of CtOrigin a bit, to say whether it comes from an instance decl or quantified constraint. * Some refactoring way in which redundant constraints are reported; we don't want to complain about the extra, apparently-redundant constraints that we must add to an instance decl because of the loopy-superclass thing. I moved some work from GHC.Tc.Errors to GHC.Tc.Solver. * Add a new section to the user manual to describe the loopy superclass issue and what rules it follows. - - - - - 300bcc15 by HaskellMouse at 2023-01-11T13:43:36-05:00 Parse qualified terms in type signatures This commit allows qualified terms in type signatures to pass the parser and to be cathced by renamer with more informative error message. Adds a few tests. Fixes #21605 - - - - - 964284fc by Simon Peyton Jones at 2023-01-11T13:44:12-05:00 Fix void-arg-adding mechanism for worker/wrapper As #22725 shows, in worker/wrapper we must add the void argument /last/, not first. See GHC.Core.Opt.WorkWrap.Utils Note [Worker/wrapper needs to add void arg last]. That led me to to study GHC.Core.Opt.SpecConstr Note [SpecConstr needs to add void args first] which suggests the opposite! And indeed I think it's the other way round for SpecConstr -- or more precisely the void arg must precede the "extra_bndrs". That led me to some refactoring of GHC.Core.Opt.SpecConstr.calcSpecInfo. - - - - - f7ceafc9 by Krzysztof Gogolewski at 2023-01-11T22:36:59-05:00 Add 'docWithStyle' to improve codegen This new combinator docWithStyle :: IsOutput doc => doc -> (PprStyle -> SDoc) -> doc let us remove the need for code to be polymorphic in HDoc when not used in code style. Metric Decrease: ManyConstructors T13035 T1969 - - - - - b3be0d18 by Simon Peyton Jones at 2023-01-11T22:37:35-05:00 Fix finaliseArgBoxities for OPAQUE function We never do worker wrapper for OPAQUE functions, so we must zap the unboxing info during strictness analysis. This patch fixes #22502 - - - - - db11f358 by Ben Gamari at 2023-01-12T07:49:04-05:00 Revert "rts: Drop racy assertion" The logic here was inverted. Reverting the commit to avoid confusion when examining the commit history. This reverts commit b3eacd64fb36724ed6c5d2d24a81211a161abef1. - - - - - 3242139f by Ben Gamari at 2023-01-12T07:49:04-05:00 rts: Drop racy assertion 0e274c39bf836d5bb846f5fa08649c75f85326ac added an assertion in `dirty_MUT_VAR` checking that the MUT_VAR being dirtied was clean. However, this isn't necessarily the case since another thread may have raced us to dirty the object. - - - - - 9ffd5d57 by Ben Gamari at 2023-01-12T07:49:41-05:00 configure: Fix escaping of `$tooldir` In !9547 I introduced `$tooldir` directories into GHC's default link and compilation flags to ensure that our C toolchain finds its own headers and libraries before others on the system. However, the patch was subtly wrong in the escaping of `$tooldir`. Fix this. Fixes #22561. - - - - - 905d0b6e by Sebastian Graf at 2023-01-12T15:51:47-05:00 Fix contification with stable unfoldings (#22428) Many functions now return a `TailUsageDetails` that adorns a `UsageDetails` with a `JoinArity` that reflects the number of join point binders around the body for which the `UsageDetails` was computed. `TailUsageDetails` is now returned by `occAnalLamTail` as well as `occAnalUnfolding` and `occAnalRules`. I adjusted `Note [Join points and unfoldings/rules]` and `Note [Adjusting right-hand sides]` to account for the new machinery. I also wrote a new `Note [Join arity prediction based on joinRhsArity]` and refer to it when we combine `TailUsageDetails` for a recursive RHS. I also renamed * `occAnalLam` to `occAnalLamTail` * `adjustRhsUsage` to `adjustTailUsage` * a few other less important functions and properly documented the that each call of `occAnalLamTail` must pair up with `adjustTailUsage`. I removed `Note [Unfoldings and join points]` because it was redundant with `Note [Occurrences in stable unfoldings]`. While in town, I refactored `mkLoopBreakerNodes` so that it returns a condensed `NodeDetails` called `SimpleNodeDetails`. Fixes #22428. The refactoring seems to have quite beneficial effect on ghc/alloc performance: ``` CoOpt_Read(normal) ghc/alloc 784,778,420 768,091,176 -2.1% GOOD T12150(optasm) ghc/alloc 77,762,270 75,986,720 -2.3% GOOD T12425(optasm) ghc/alloc 85,740,186 84,641,712 -1.3% GOOD T13056(optasm) ghc/alloc 306,104,656 299,811,632 -2.1% GOOD T13253(normal) ghc/alloc 350,233,952 346,004,008 -1.2% T14683(normal) ghc/alloc 2,800,514,792 2,754,651,360 -1.6% T15304(normal) ghc/alloc 1,230,883,318 1,215,978,336 -1.2% T15630(normal) ghc/alloc 153,379,590 151,796,488 -1.0% T16577(normal) ghc/alloc 7,356,797,056 7,244,194,416 -1.5% T17516(normal) ghc/alloc 1,718,941,448 1,692,157,288 -1.6% T19695(normal) ghc/alloc 1,485,794,632 1,458,022,112 -1.9% T21839c(normal) ghc/alloc 437,562,314 431,295,896 -1.4% GOOD T21839r(normal) ghc/alloc 446,927,580 440,615,776 -1.4% GOOD geo. mean -0.6% minimum -2.4% maximum -0.0% ``` Metric Decrease: CoOpt_Read T10421 T12150 T12425 T13056 T18698a T18698b T21839c T21839r T9961 - - - - - a1491c87 by Andreas Klebinger at 2023-01-12T15:52:23-05:00 Only gc sparks locally when we can ensure marking is done. When performing GC without work stealing there was no guarantee that spark pruning was happening after marking of the sparks. This could cause us to GC live sparks under certain circumstances. Fixes #22528. - - - - - 8acfe930 by Cheng Shao at 2023-01-12T15:53:00-05:00 Change MSYSTEM to CLANG64 uniformly - - - - - 73bc162b by M Farkas-Dyck at 2023-01-12T15:53:42-05:00 Make `GHC.Tc.Errors.Reporter` take `NonEmpty ErrorItem` rather than `[ErrorItem]`, which lets us drop some panics. Also use the `BasicMismatch` constructor rather than `mkBasicMismatchMsg`, which lets us drop the "-Wno-incomplete-record-updates" flag. - - - - - 1b812b69 by Oleg Grenrus at 2023-01-12T15:54:21-05:00 Fix #22728: Not all diagnostics in safe check are fatal Also add tests for the issue and -Winferred-safe-imports in general - - - - - c79b2b65 by Matthew Pickering at 2023-01-12T15:54:58-05:00 Don't run hadrian-multi on fast-ci label Fixes #22667 - - - - - 9a3d6add by Andrew Lelechenko at 2023-01-13T00:46:36-05:00 Bump submodule bytestring to 0.11.4.0 Metric Decrease: T21839c T21839r - - - - - df33c13c by Ben Gamari at 2023-01-13T00:47:12-05:00 gitlab-ci: Bump Darwin bootstrap toolchain This updates the bootstrap compiler on Darwin from 8.10.7 to 9.2.5, ensuring that we have the fix for #21964. - - - - - 756a66ec by Ben Gamari at 2023-01-13T00:47:12-05:00 gitlab-ci: Pass -w to cabal update Due to cabal#8447, cabal-install 3.8.1.0 requires a compiler to run `cabal update`. - - - - - 1142f858 by Cheng Shao at 2023-01-13T11:04:00+00:00 Bump hsc2hs submodule - - - - - d4686729 by Cheng Shao at 2023-01-13T11:04:00+00:00 Bump process submodule - - - - - 84ae6573 by Cheng Shao at 2023-01-13T11:06:58+00:00 ci: Bump DOCKER_REV - - - - - d53598c5 by Cheng Shao at 2023-01-13T11:06:58+00:00 ci: enable xz parallel compression for x64 jobs - - - - - d31fcbca by Cheng Shao at 2023-01-13T11:06:58+00:00 ci: use in-image emsdk for js jobs - - - - - 93b9bbc1 by Cheng Shao at 2023-01-13T11:47:17+00:00 ci: improve nix-shell for gen_ci.hs and fix some ghc/hlint warnings - Add a ghc environment including prebuilt dependencies to the nix-shell. Get rid of the ad hoc cabal cache and all dependencies are now downloaded from the nixos binary cache. - Make gen_ci.hs a cabal package with HLS integration, to make future hacking of gen_ci.hs easier. - Fix some ghc/hlint warnings after I got HLS to work. - For the lint-ci-config job, do a shallow clone to save a few minutes of unnecessary git checkout time. - - - - - 8acc56c7 by Cheng Shao at 2023-01-13T11:47:17+00:00 ci: source the toolchain env file in wasm jobs - - - - - 87194df0 by Cheng Shao at 2023-01-13T11:47:17+00:00 ci: add wasm ci jobs via gen_ci.hs - There is one regular wasm job run in validate pipelines - Additionally, int-native/unreg wasm jobs run in nightly/release pipelines Also, remove the legacy handwritten wasm ci jobs in .gitlab-ci.yml. - - - - - b6eb9bcc by Matthew Pickering at 2023-01-13T11:52:16+00:00 wasm ci: Remove wasm release jobs This removes the wasm release jobs, as we do not yet intend to distribute these binaries. - - - - - 496607fd by Simon Peyton Jones at 2023-01-13T16:52:07-05:00 Add a missing checkEscapingKind Ticket #22743 pointed out that there is a missing check, for type-inferred bindings, that the inferred type doesn't have an escaping kind. The fix is easy. - - - - - 7a9a1042 by Andreas Klebinger at 2023-01-16T20:48:19-05:00 Separate core inlining logic from `Unfolding` type. This seems like a good idea either way, but is mostly motivated by a patch where this avoids a module loop. - - - - - 33b58f77 by sheaf at 2023-01-16T20:48:57-05:00 Hadrian: generalise &%> to avoid warnings This patch introduces a more general version of &%> that works with general traversable shapes, instead of lists. This allows us to pass along the information that the length of the list of filepaths passed to the function exactly matches the length of the input list of filepath patterns, avoiding pattern match warnings. Fixes #22430 - - - - - 8c7a991c by Andreas Klebinger at 2023-01-16T20:49:34-05:00 Add regression test for #22611. A case were a function used to fail to specialize, but now does. - - - - - 6abea760 by Andreas Klebinger at 2023-01-16T20:50:10-05:00 Mark maximumBy/minimumBy as INLINE. The RHS was too large to inline which often prevented the overhead of the Maybe from being optimized away. By marking it as INLINE we can eliminate the overhead of both the maybe and are able to unpack the accumulator when possible. Fixes #22609 - - - - - 99d151bb by Matthew Pickering at 2023-01-16T20:50:50-05:00 ci: Bump CACHE_REV so that ghc-9.6 branch and HEAD have different caches Having the same CACHE_REV on both branches leads to issues where the darwin toolchain is different on ghc-9.6 and HEAD which leads to long darwin build times. In general we should ensure that each branch has a different CACHE_REV. - - - - - 6a5845fb by Matthew Pickering at 2023-01-16T20:51:25-05:00 ci: Change owner of files in source-tarball job This fixes errors of the form: ``` fatal: detected dubious ownership in repository at '/builds/ghc/ghc' To add an exception for this directory, call: git config --global --add safe.directory /builds/ghc/ghc inferred 9.7.20230113 checking for GHC Git commit id... fatal: detected dubious ownership in repository at '/builds/ghc/ghc' To add an exception for this directory, call: git config --global --add safe.directory /builds/ghc/ghc ``` - - - - - 4afb952c by Matthew Pickering at 2023-01-16T20:51:25-05:00 ci: Don't build aarch64-deb10-llvm job on release pipelines Closes #22721 - - - - - 8039feb9 by Matthew Pickering at 2023-01-16T20:51:25-05:00 ci: Change owner of files in test-bootstrap job - - - - - 0b358d0c by Matthew Pickering at 2023-01-16T20:51:25-05:00 rel_eng: Add release engineering scripts into ghc tree It is better to keep these scripts in the tree as they depend on the CI configuration and so on. By keeping them in tree we can keep them up-to-date as the CI config changes and also makes it easier to backport changes to the release script between release branches in future. The final motivation is that it makes generating GHCUp metadata possible. - - - - - 28cb2ed0 by Matthew Pickering at 2023-01-16T20:51:25-05:00 ci: Don't use complicated image or clone in not-interruptible job This job exists only for the meta-reason of not allowing nightly pipelines to be cancelled. It was taking two minutes to run as in order to run "true" we would also clone the whole GHC repo. - - - - - eeea59bb by Matthew Pickering at 2023-01-16T20:51:26-05:00 Add scripts to generate ghcup metadata on nightly and release pipelines 1. A python script in .gitlab/rel_eng/mk-ghcup-metadata which generates suitable metadata for consumption by GHCUp for the relevant pipelines. - The script generates the metadata just as the ghcup maintainers want, without taking into account platform/library combinations. It is updated manually when the mapping changes. - The script downloads the bindists which ghcup wants to distribute, calculates the hash and generates the yaml in the correct structure. - The script is documented in the .gitlab/rel_eng/mk-ghcup-metadata/README.mk file 1a. The script requires us to understand the mapping from platform -> job. To choose the preferred bindist for each platform the .gitlab/gen_ci.hs script is modified to allow outputting a metadata file which answers the question about which job produces the bindist which we want to distribute to users for a specific platform. 2. Pipelines to run on nightly and release jobs to generate metadata - ghcup-metadata-nightly: Generates metadata which points directly to artifacts in the nightly job. - ghcup-metadata-release: Generates metadata suitable for inclusion directly in ghcup by pointing to the downloads folder where the bindist will be uploaded to. 2a. Trigger jobs which test the generated metadata in the downstream `ghccup-ci` repo. See that repo for documentation about what is tested and how but essentially we test in a variety of clean images that ghcup can download and install the bindists we say exist in our metadata. - - - - - 97bd4d8c by Andrew Lelechenko at 2023-01-16T20:52:04-05:00 Bump submodule parsec to 3.1.16.1 - - - - - 97ac8230 by Alan Zimmerman at 2023-01-16T20:52:39-05:00 EPA: Add annotation for 'type' in DataDecl Closes #22765 - - - - - dbbab95d by Ben Gamari at 2023-01-17T06:36:06-05:00 compiler: Small optimisation of assertM In #22739 @AndreasK noticed that assertM performed the action to compute the asserted predicate regardless of whether DEBUG is enabled. This is inconsistent with the other assertion operations and general convention. Fix this. Closes #22739. - - - - - fc02f3bb by Viktor Dukhovni at 2023-01-17T06:36:47-05:00 Avoid unnecessary printf warnings in EventLog.c Fixes #22778 - - - - - 003b6d44 by Simon Peyton Jones at 2023-01-17T16:33:05-05:00 Document the semantics of pattern bindings a bit better This MR is in response to the discussion on #22719 - - - - - f4d50baf by Vladislav Zavialov at 2023-01-17T16:33:41-05:00 Hadrian: fix warnings (#22783) This change fixes the following warnings when building Hadrian: src/Hadrian/Expression.hs:38:10: warning: [-Wredundant-constraints] src/Hadrian/Expression.hs:84:13: warning: [-Wtype-equality-requires-operators] src/Hadrian/Expression.hs:84:21: warning: [-Wtype-equality-requires-operators] src/Hadrian/Haskell/Cabal/Parse.hs:67:1: warning: [-Wunused-imports] - - - - - 06036d93 by Sylvain Henry at 2023-01-18T01:55:10-05:00 testsuite: req_smp --> req_target_smp, req_ghc_smp See #22630 and !9552 This commit: - splits req_smp into req_target_smp and req_ghc_smp - changes the testsuite driver to calculate req_ghc_smp - changes a handful of tests to use req_target_smp instead of req_smp - changes a handful of tests to use req_host_smp when needed The problem: - the problem this solves is the ambiguity surrounding req_smp - on master req_smp was used to express the constraint that the program being compiled supports smp _and_ that the host RTS (i.e., the RTS used to compile the program) supported smp. Normally that is fine, but in cross compilation this is not always the case as was discovered in #22630. The solution: - Differentiate the two constraints: - use req_target_smp to say the RTS the compiled program is linked with (and the platform) supports smp - use req_host_smp to say the RTS the host is linked with supports smp WIP: fix req_smp (target vs ghc) add flag to separate bootstrapper split req_smp -> req_target_smp and req_ghc_smp update tests smp flags cleanup and add some docstrings only set ghc_with_smp to bootstrapper on S1 or CC Only set ghc_with_smp to bootstrapperWithSMP of when testing stage 1 and cross compiling test the RTS in config/ghc not hadrian re-add ghc_with_smp fix and align req names fix T11760 to use req_host_smp test the rts directly, avoid python 3.5 limitation test the compiler in a try block align out of tree and in tree withSMP flags mark failing tests as host req smp testsuite: req_host_smp --> req_ghc_smp Fix ghc vs host, fix ghc_with_smp leftover - - - - - ee9b78aa by Krzysztof Gogolewski at 2023-01-18T01:55:45-05:00 Use -Wdefault when running Python testdriver (#22727) - - - - - e9c0537c by Vladislav Zavialov at 2023-01-18T01:56:22-05:00 Enable -Wstar-is-type by default (#22759) Following the plan in GHC Proposal #143 "Remove the * kind syntax", which states: In the next release (or 3 years in), enable -fwarn-star-is-type by default. The "next release" happens to be 9.6.1 I also moved the T21583 test case from should_fail to should_compile, because the only reason it was failing was -Werror=compat in our test suite configuration. - - - - - 4efee43d by Ryan Scott at 2023-01-18T01:56:59-05:00 Add missing parenthesizeHsType in cvtSigTypeKind We need to ensure that the output of `cvtSigTypeKind` is parenthesized (at precedence `sigPrec`) so that any type signatures with an outermost, explicit kind signature can parse correctly. Fixes #22784. - - - - - f891a442 by Ben Gamari at 2023-01-18T07:28:00-05:00 Bump ghc-tarballs to fix #22497 It turns out that gmp 6.2.1 uses the platform-reserved `x18` register on AArch64/Darwin. This was fixed in upstream changeset 18164:5f32dbc41afc, which was merged in 2020. Here I backport this patch although I do hope that a new release is forthcoming soon. Bumps gmp-tarballs submodule. Fixes #22497. - - - - - b13c6ea5 by Ben Gamari at 2023-01-18T07:28:00-05:00 Bump gmp-tarballs submodule This backports the upstream fix for CVE-2021-43618, fixing #22789. - - - - - c45a5fff by Cheng Shao at 2023-01-18T07:28:37-05:00 Fix typo in recent darwin tests fix Corrects a typo in !9647. Otherwise T18623 will still fail on darwin and stall other people's work. - - - - - b4c14c4b by Luite Stegeman at 2023-01-18T14:21:42-05:00 Add PrimCallConv support to GHCi This adds support for calling Cmm code from bytecode using the native calling convention, allowing modules that use `foreign import prim` to be loaded and debugged in GHCi. This patch introduces a new `PRIMCALL` bytecode instruction and a helper stack frame `stg_primcall`. The code is based on the existing functionality for dealing with unboxed tuples in bytecode, which has been generalised to handle arbitrary calls. Fixes #22051 - - - - - d0a63ef8 by Adam Gundry at 2023-01-18T14:22:26-05:00 Refactor warning flag parsing to add missing flags This adds `-Werror=<group>` and `-fwarn-<group>` flags for warning groups as well as individual warnings. Previously these were defined on an ad hoc basis so for example we had `-Werror=compat` but not `-Werror=unused-binds`, whereas we had `-fwarn-unused-binds` but not `-fwarn-compat`. Fixes #22182. - - - - - 7ed1b8ef by Adam Gundry at 2023-01-18T14:22:26-05:00 Minor corrections to comments - - - - - 5389681e by Adam Gundry at 2023-01-18T14:22:26-05:00 Revise warnings documentation in user's guide - - - - - ab0d5cda by Adam Gundry at 2023-01-18T14:22:26-05:00 Move documentation of deferred type error flags out of warnings section - - - - - eb5a6b91 by John Ericson at 2023-01-18T22:24:10-05:00 Give the RTS it's own configure script Currently it doesn't do much anything, we are just trying to introduce it without breaking the build. Later, we will move functionality from the top-level configure script over to it. We need to bump Cabal for https://github.com/haskell/cabal/pull/8649; to facilitate and existing hack of skipping some configure checks for the RTS we now need to skip just *part* not *all* of the "post configure" hook, as running the configure script (which we definitely want to do) is also implemented as part of the "post configure" hook. But doing this requires exposing functionality that wasn't exposed before. - - - - - 32ab07bf by Andrew Lelechenko at 2023-01-18T22:24:51-05:00 ghc package does not have to depend on terminfo - - - - - 981ff7c4 by Andrew Lelechenko at 2023-01-18T22:24:51-05:00 ghc-pkg does not have to depend on terminfo - - - - - f058e367 by Ben Gamari at 2023-01-18T22:25:27-05:00 nativeGen/X86: MFENCE is unnecessary for release semantics In #22764 a user noticed that a program implementing a simple atomic counter via an STRef regressed significantly due to the introduction of necessary atomic operations in the MutVar# primops (#22468). This regression was caused by a bug in the NCG, which emitted an unnecessary MFENCE instruction for a release-ordered atomic write. MFENCE is rather only needed to achieve sequentially consistent ordering. Fixes #22764. - - - - - 154889db by Ryan Scott at 2023-01-18T22:26:03-05:00 Add regression test for #22151 Issue #22151 was coincidentally fixed in commit aed1974e92366ab8e117734f308505684f70cddf (`Refactor the treatment of loopy superclass dicts`). This adds a regression test to ensure that the issue remains fixed. Fixes #22151. - - - - - 14b5982a by Andrei Borzenkov at 2023-01-18T22:26:43-05:00 Fix printing of promoted MkSolo datacon (#22785) Problem: In 2463df2f, the Solo data constructor was renamed to MkSolo, and Solo was turned into a pattern synonym for backwards compatibility. Since pattern synonyms can not be promoted, the old code that pretty-printed promoted single-element tuples started producing ill-typed code: t :: Proxy ('Solo Int) This fails with "Pattern synonym ‘Solo’ used as a type" The solution is to track the distinction between type constructors and data constructors more carefully when printing single-element tuples. - - - - - 1fe806d3 by Cheng Shao at 2023-01-23T04:48:47-05:00 hadrian: add hi_core flavour transformer The hi_core flavour transformer enables -fwrite-if-simplified-core for stage1 libraries, which emit core into interface files to make it possible to restart code generation. Building boot libs with it makes it easier to use GHC API to prototype experimental backends that needs core/stg at link time. - - - - - 317cad26 by Cheng Shao at 2023-01-23T04:48:47-05:00 hadrian: add missing docs for recently added flavour transformers - - - - - 658f4446 by Ben Gamari at 2023-01-23T04:49:23-05:00 gitlab-ci: Add Rocky8 jobs Addresses #22268. - - - - - a83ec778 by Vladislav Zavialov at 2023-01-23T04:49:58-05:00 Set "since: 9.8" for TypeAbstractions and -Wterm-variable-capture These flags did not make it into the 9.6 release series, so the "since" annotations must be corrected. - - - - - fec7c2ea by Alan Zimmerman at 2023-01-23T04:50:33-05:00 EPA: Add SourceText to HsOverLabel To be able to capture string literals with possible escape codes as labels. Close #22771 - - - - - 3efd1e99 by Ben Gamari at 2023-01-23T04:51:08-05:00 template-haskell: Bump version to 2.20.0.0 Updates `text` and `exceptions` submodules for bounds bumps. Addresses #22767. - - - - - 0900b584 by Cheng Shao at 2023-01-23T04:51:45-05:00 hadrian: disable alloca for in-tree GMP on wasm32 When building in-tree GMP for wasm32, disable its alloca usage, since it may potentially cause stack overflow (e.g. #22602). - - - - - db0f1bfd by Cheng Shao at 2023-01-23T04:52:21-05:00 Bump process submodule Includes a critical fix for wasm32, see https://github.com/haskell/process/pull/272 for details. Also changes the existing cross test to include process stuff and avoid future regression here. - - - - - 9222b167 by Matthew Pickering at 2023-01-23T04:52:57-05:00 ghcup metadata: Fix subdir for windows bindist - - - - - 9a9bec57 by Matthew Pickering at 2023-01-23T04:52:57-05:00 ghcup metadata: Remove viPostRemove field from generated metadata This has been removed from the downstream metadata. - - - - - 82884ce0 by Simon Peyton Jones at 2023-01-23T04:53:32-05:00 Fix #22742 runtimeRepLevity_maybe was panicing unnecessarily; and the error printing code made use of the case when it should return Nothing rather than panicing. For some bizarre reason perf/compiler/T21839r shows a 10% bump in runtime peak-megagbytes-used, on a single architecture (alpine). See !9753 for commentary, but I'm going to accept it. Metric Increase: T21839r - - - - - 2c6deb18 by Bryan Richter at 2023-01-23T14:12:22+02:00 codeowners: Add Ben, Matt, and Bryan to CI - - - - - eee3bf05 by Matthew Craven at 2023-01-23T21:46:41-05:00 Do not collect compile-time metrics for T21839r ...the testsuite doesn't handle this properly since it also collects run-time metrics. Compile-time metrics for this test are already tracked via T21839c. Metric Decrease: T21839r - - - - - 1d1dd3fb by Matthew Pickering at 2023-01-24T05:37:52-05:00 Fix recompilation checking for multiple home units The key part of this change is to store a UnitId in the `UsageHomeModule` and `UsageHomeModuleInterface`. * Fine-grained dependency tracking is used if the dependency comes from any home unit. * We actually look up the right module when checking whether we need to recompile in the `UsageHomeModuleInterface` case. These scenarios are both checked by the new tests ( multipleHomeUnits_recomp and multipleHomeUnits_recomp_th ) Fixes #22675 - - - - - 7bfb30f9 by Matthew Pickering at 2023-01-24T05:37:52-05:00 Augment target filepath by working directory when checking if module satisfies target This fixes a spurious warning in -Wmissing-home-modules. This is a simple oversight where when looking for the target in the first place we augment the search by the -working-directory flag but then fail to do so when checking this warning. Fixes #22676 - - - - - 69500dd4 by Matthew Pickering at 2023-01-24T05:37:52-05:00 Use NodeKey rather than ModuleName in pruneCache The `pruneCache` function assumes that the list of `CachedInfo` all have unique `ModuleName`, this is not true: * In normal compilation, the same module name can appear for a file and it's boot file. * In multiple home unit compilation the same ModuleName can appear in different units The fix is to use a `NodeKey` as the actual key for the interfaces which includes `ModuleName`, `IsBoot` and `UnitId`. Fixes #22677 - - - - - 336b2b1c by Matthew Pickering at 2023-01-24T05:37:52-05:00 Recompilation checking: Don't try to find artefacts for Interactive & hs-boot combo In interactive mode we don't produce any linkables for hs-boot files. So we also need to not going looking for them when we check to see if we have all the right objects needed for recompilation. Ticket #22669 - - - - - 6469fea7 by Matthew Pickering at 2023-01-24T05:37:52-05:00 Don't write o-boot files in Interactive mode We should not be producing object files when in interactive mode but we still produced the dummy o-boot files. These never made it into a `Linkable` but then confused the recompilation checker. Fixes #22669 - - - - - 06cc0a95 by Matthew Pickering at 2023-01-24T05:37:52-05:00 Improve driver diagnostic messages by including UnitId in message Currently the driver diagnostics don't give any indication about which unit they correspond to. For example `-Wmissing-home-modules` can fire multiple times for each different home unit and gives no indication about which unit it's actually reporting about. Perhaps a longer term fix is to generalise the providence information away from a SrcSpan so that these kind of whole project errors can be reported with an accurate provenance. For now we can just include the `UnitId` in the error message. Fixes #22678 - - - - - 4fe9eaff by Matthew Pickering at 2023-01-24T05:37:52-05:00 Key ModSummary cache by UnitId as well as FilePath Multiple units can refer to the same files without any problem. Just another assumption which needs to be updated when we may have multiple home units. However, there is the invariant that within each unit each file only maps to one module, so as long as we also key the cache by UnitId then we are all good. This led to some confusing behaviour in GHCi when reloading, multipleHomeUnits_shared distils the essence of what can go wrong. Fixes #22679 - - - - - ada29f5c by Matthew Pickering at 2023-01-24T05:37:52-05:00 Finder: Look in current unit before looking in any home package dependencies In order to preserve existing behaviour it's important to look within the current component before consideirng a module might come from an external component. This already happened by accident in `downsweep`, (because roots are used to repopulated the cache) but in the `Finder` the logic was the wrong way around. Fixes #22680 ------------------------- Metric Decrease: MultiComponentModules MultiComponentModulesRecomp -------------------------p - - - - - be701cc6 by Matthew Pickering at 2023-01-24T05:37:52-05:00 Debug: Print full NodeKey when pretty printing ModuleGraphNode This is helpful when debugging multiple component issues. - - - - - 34d2d463 by Krzysztof Gogolewski at 2023-01-24T05:38:32-05:00 Fix Lint check for duplicate external names Lint was checking for duplicate external names by calling removeDups, which needs a comparison function that is passed to Data.List.sortBy. But the comparison was not a valid ordering - it returned LT if one of the names was not external. For example, the previous implementation won't find a duplicate in [M.x, y, M.x]. Instead, we filter out non-external names before looking for duplicates. - - - - - 1c050ed2 by Matthew Pickering at 2023-01-24T05:39:08-05:00 Add test for T22671 This was fixed by b13c6ea5 Closes #22671 - - - - - 05e6a2d9 by Tom Ellis at 2023-01-24T12:10:52-05:00 Clarify where `f` is defined - - - - - d151546e by Cheng Shao at 2023-01-24T12:11:29-05:00 CmmToC: fix CmmRegOff for 64-bit register on a 32-bit target We used to print the offset value to a platform word sized integer. This is incorrect when the offset is negative (e.g. output of cmm constant folding) and the register is 64-bit but on a 32-bit target, and may lead to incorrect runtime result (e.g. #22607). The fix is simple: just treat it as a proper MO_Add, with the correct width info inferred from the register itself. Metric Increase: T12707 T13379 T4801 T5321FD T5321Fun - - - - - e5383a29 by Wander Hillen at 2023-01-24T20:02:26-05:00 Allow waiting for timerfd to be interrupted during rts shutdown - - - - - 1957eda1 by Ryan Scott at 2023-01-24T20:03:01-05:00 Restore Compose's Read/Show behavior to match Read1/Show1 instances Fixes #22816. - - - - - 30972827 by Matthew Pickering at 2023-01-25T03:54:14-05:00 docs: Update INSTALL.md Removes references to make. Fixes #22480 - - - - - bc038c3b by Cheng Shao at 2023-01-25T03:54:50-05:00 compiler: fix handling of MO_F_Neg in wasm NCG In the wasm NCG, we used to compile MO_F_Neg to 0.0-x. It was an oversight, there actually exists f32.neg/f64.neg opcodes in the wasm spec and those should be used instead! The old behavior almost works, expect when GHC compiles the -0.0 literal, which will incorrectly become 0.0. - - - - - e987e345 by Sylvain Henry at 2023-01-25T14:47:41-05:00 Hadrian: correctly detect AR at-file support Stage0's ar may not support at-files. Take it into account. Found while cross-compiling from Darwin to Windows. - - - - - 48131ee2 by Sylvain Henry at 2023-01-25T14:47:41-05:00 Hadrian: fix Windows cross-compilation Decision to build either unix or Win32 package must be stage specific for cross-compilation to be supported. - - - - - 288fa017 by Sylvain Henry at 2023-01-25T14:47:41-05:00 Fix RTS build on Windows This change fixes a cross-compilation issue from ArchLinux to Windows because these symbols weren't found. - - - - - 2fdf22ae by Sylvain Henry at 2023-01-25T14:47:41-05:00 configure: support "windows" as an OS - - - - - 13a0566b by Simon Peyton Jones at 2023-01-25T14:48:16-05:00 Fix in-scope set in specImports Nothing deep here; I had failed to bring some floated dictionary binders into scope. Exposed by -fspecialise-aggressively Fixes #22715. - - - - - b7efdb24 by Matthew Pickering at 2023-01-25T14:48:51-05:00 ci: Disable HLint job due to excessive runtime The HLint jobs takes much longer to run (20 minutes) after "Give the RTS it's own configure script" eb5a6b91 Now the CI job will build the stage0 compiler before it generates the necessary RTS headers. We either need to: * Fix the linting rules so they take much less time * Revert the commit * Remove the linting of base from the hlint job * Remove the hlint job This is highest priority as it is affecting all CI pipelines. For now I am just disabling the job because there are many more pressing matters at hand. Ticket #22830 - - - - - 1bd32a35 by Sylvain Henry at 2023-01-26T12:34:21-05:00 Factorize hptModulesBelow Create and use moduleGraphModulesBelow in GHC.Unit.Module.Graph that doesn't need anything from the driver to be used. - - - - - 1262d3f8 by Matthew Pickering at 2023-01-26T12:34:56-05:00 Store dehydrated data structures in CgModBreaks This fixes a tricky leak in GHCi where we were retaining old copies of HscEnvs when reloading. If not all modules were recompiled then these hydrated fields in break points would retain a reference to the old HscEnv which could double memory usage. Fixes #22530 - - - - - e27eb80c by Matthew Pickering at 2023-01-26T12:34:56-05:00 Force more in NFData Name instance Doesn't force the lazy `OccName` field (#19619) which is already known as a really bad source of leaks. When we slam the hammer storing Names on disk (in interface files or the like), all this should be forced as otherwise a `Name` can easily retain an `Id` and hence the entire world. Fixes #22833 - - - - - 3d004d5a by Matthew Pickering at 2023-01-26T12:34:56-05:00 Force OccName in tidyTopName This occname has just been derived from an `Id`, so need to force it promptly so we can release the Id back to the world. Another symptom of the bug caused by #19619 - - - - - f2a0fea0 by Matthew Pickering at 2023-01-26T12:34:56-05:00 Strict fields in ModNodeKey (otherwise retains HomeModInfo) Towards #22530 - - - - - 5640cb1d by Sylvain Henry at 2023-01-26T12:35:36-05:00 Hadrian: fix doc generation Was missing dependencies on files generated by templates (e.g. ghc.cabal) - - - - - 3e827c3f by Richard Eisenberg at 2023-01-26T20:06:53-05:00 Do newtype unwrapping in the canonicaliser and rewriter See Note [Unwrap newtypes first], which has the details. Close #22519. - - - - - b3ef5c89 by doyougnu at 2023-01-26T20:07:48-05:00 tryFillBuffer: strictify more speculative bangs - - - - - d0d7ba0f by Vladislav Zavialov at 2023-01-26T20:08:25-05:00 base: NoImplicitPrelude in Data.Void and Data.Kind This change removes an unnecessary dependency on Prelude from two modules in the base package. - - - - - fa1db923 by Matthew Pickering at 2023-01-26T20:09:00-05:00 ci: Add ubuntu18_04 nightly and release jobs This adds release jobs for ubuntu18_04 which uses glibc 2.27 which is older than the 2.28 which is used by Rocky8 bindists. Ticket #22268 - - - - - 807310a1 by Matthew Pickering at 2023-01-26T20:09:00-05:00 rel-eng: Add missing rocky8 bindist We intend to release rocky8 bindist so the fetching script needs to know about them. - - - - - c7116b10 by Ben Gamari at 2023-01-26T20:09:35-05:00 base: Make changelog proposal references more consistent Addresses #22773. - - - - - 6932cfc7 by Sylvain Henry at 2023-01-26T20:10:27-05:00 Fix spurious change from !9568 - - - - - e480fbc2 by Ben Gamari at 2023-01-27T05:01:24-05:00 rts: Use C11-compliant static assertion syntax Previously we used `static_assert` which is only available in C23. By contrast, C11 only provides `_Static_assert`. Fixes #22777 - - - - - 2648c09c by Andrei Borzenkov at 2023-01-27T05:02:07-05:00 Replace errors from badOrigBinding with new one (#22839) Problem: in 02279a9c the type-level [] syntax was changed from a built-in name to an alias for the GHC.Types.List constructor. badOrigBinding assumes that if a name is not built-in then it must have come from TH quotation, but this is not necessarily the case with []. The outdated assumption in badOrigBinding leads to incorrect error messages. This code: data [] Fails with "Cannot redefine a Name retrieved by a Template Haskell quote: []" Unfortunately, there is not enough information in RdrName to directly determine if the name was constructed via TH or by the parser, so this patch changes the error message instead. It unifies TcRnIllegalBindingOfBuiltIn and TcRnNameByTemplateHaskellQuote into a new error TcRnBindingOfExistingName and changes its wording to avoid guessing the origin of the name. - - - - - 545bf8cf by Matthew Pickering at 2023-01-27T14:58:53+00:00 Revert "base: NoImplicitPrelude in Data.Void and Data.Kind" Fixes CI errors of the form. ``` ===> Command failed with error code: 1 ghc: panic! (the 'impossible' happened) GHC version 9.7.20230127: lookupGlobal Failed to load interface for ‘GHC.Num.BigNat’ There are files missing in the ‘ghc-bignum’ package, try running 'ghc-pkg check'. Use -v (or `:set -v` in ghci) to see a list of the files searched for. Call stack: CallStack (from HasCallStack): callStackDoc, called at compiler/GHC/Utils/Panic.hs:189:37 in ghc:GHC.Utils.Panic pprPanic, called at compiler/GHC/Tc/Utils/Env.hs:154:32 in ghc:GHC.Tc.Utils.Env CallStack (from HasCallStack): panic, called at compiler/GHC/Utils/Error.hs:454:29 in ghc:GHC.Utils.Error Please report this as a GHC bug: https://www.haskell.org/ghc/reportabug ``` This reverts commit d0d7ba0fb053ebe7f919a5932066fbc776301ccd. The module now lacks a dependency on GHC.Num.BigNat which it implicitly depends on. It is causing all CI jobs to fail so we revert without haste whilst the patch can be fixed. Fixes #22848 - - - - - 638277ba by Simon Peyton Jones at 2023-01-27T23:54:55-05:00 Detect family instance orphans correctly We were treating a type-family instance as a non-orphan if there was a type constructor on its /right-hand side/ that was local. Boo! Utterly wrong. With this patch, we correctly check the /left-hand side/ instead! Fixes #22717 - - - - - 46a53bb2 by Simon Peyton Jones at 2023-01-27T23:54:55-05:00 Report family instance orphans correctly This fixes the fact that we were not reporting orphan family instances at all. The fix here is easy, but touches a bit of code. I refactored the code to be much more similar to the way that class instances are done: - Add a fi_orphan field to FamInst, like the is_orphan field in ClsInst - Make newFamInst initialise this field, just like newClsInst - And make newFamInst report a warning for an orphan, just like newClsInst - I moved newFamInst from GHC.Tc.Instance.Family to GHC.Tc.Utils.Instantiate, just like newClsInst. - I added mkLocalFamInst to FamInstEnv, just like mkLocalClsInst in InstEnv - TcRnOrphanInstance and SuggestFixOrphanInstance are now parametrised over class instances vs type/data family instances. Fixes #19773 - - - - - faa300fb by Simon Peyton Jones at 2023-01-27T23:54:55-05:00 Avoid orphans in STG This patch removes some orphan instances in the STG namespace by introducing the GHC.Stg.Lift.Types module, which allows various type family instances to be moved to GHC.Stg.Syntax, avoiding orphan instances. - - - - - 0f25a13b by Simon Peyton Jones at 2023-01-27T23:54:55-05:00 Avoid orphans in the parser This moves Anno instances for PatBuilder from GHC.Parser.PostProcess to GHC.Parser.Types to avoid orphans. - - - - - 15750d33 by Simon Peyton Jones at 2023-01-27T23:54:55-05:00 Accept an orphan declaration (sadly) This accepts the orphan type family instance type instance DsForeignHook = ... in GHC.HsToCore.Types. See Note [The Decoupling Abstract Data Hack] in GHC.Driver.Hooks - - - - - c9967d13 by Zubin Duggal at 2023-01-27T23:55:31-05:00 bindist configure: Fail if find not found (#22691) - - - - - ad8cfed4 by John Ericson at 2023-01-27T23:56:06-05:00 Put hadrian bootstrap plans through `jq` This makes it possible to review changes with conventional diffing tools. - - - - - d0ddc01b by Ben Gamari at 2023-01-27T23:56:42-05:00 testsuite: Introduce threaded2_sanity way Incredibly, we previously did not have a single way which would test the threaded RTS with multiple capabilities and the sanity-checker enabled. - - - - - 38ad8351 by Ben Gamari at 2023-01-27T23:56:42-05:00 rts: Relax Messages assertion `doneWithMsgThrowTo` was previously too strict in asserting that the `Message` is locked. Specifically, it failed to consider that the `Message` may not be locked if we are deleting all threads during RTS shutdown. - - - - - a9fe81af by Ben Gamari at 2023-01-27T23:56:42-05:00 testsuite: Fix race in UnliftedTVar2 Previously UnliftedTVar2 would fail when run with multiple capabilities (and possibly even with one capability) as it would assume that `killThread#` would immediately kill the "increment" thread. Also, refactor the the executable to now succeed with no output and fails with an exit code. - - - - - 8519af60 by Ben Gamari at 2023-01-27T23:56:42-05:00 testsuite: Make listThreads more robust Previously it was sensitive to the labels of threads which it did not create (e.g. the IO manager event loop threads). Fix this. - - - - - 55a81995 by Ben Gamari at 2023-01-27T23:56:43-05:00 rts: Fix non-atomic mutation of enabled_capabilities - - - - - b5c75f1d by Ben Gamari at 2023-01-27T23:56:43-05:00 rts: Fix C++ compilation issues Make the RTS compilable with a C++ compiler by inserting necessary casts. - - - - - c261b62f by Ben Gamari at 2023-01-27T23:56:43-05:00 rts: Fix typo "tracingAddCapabilities" was mis-named - - - - - 77fdbd3f by Ben Gamari at 2023-01-27T23:56:43-05:00 rts: Drop long-dead fallback definitions for INFINITY & NAN These are no longer necessary since we now compile as C99. - - - - - 56c1bd98 by Ben Gamari at 2023-01-28T02:57:59-05:00 Revert "CApiFFI: add ConstPtr for encoding const-qualified pointer return types (#22043)" This reverts commit 99aca26b652603bc62953157a48e419f737d352d. - - - - - b3a3534b by nineonine at 2023-01-28T02:57:59-05:00 CApiFFI: add ConstPtr for encoding const-qualified pointer return types Previously, when using `capi` calling convention in foreign declarations, code generator failed to handle const-cualified pointer return types. This resulted in CC toolchain throwing `-Wincompatible-pointer-types-discards-qualifiers` warning. `Foreign.C.Types.ConstPtr` newtype was introduced to handle these cases - special treatment was put in place to generate appropritetly qualified C wrapper that no longer triggers the above mentioned warning. Fixes #22043. - - - - - 082b7d43 by Oleg Grenrus at 2023-01-28T02:58:38-05:00 Add Foldable1 Solo instance - - - - - 50b1e2e8 by Andrei Borzenkov at 2023-01-28T02:59:18-05:00 Convert diagnostics in GHC.Rename.Bind to proper TcRnMessage (#20115) I removed all occurrences of TcRnUnknownMessage in GHC.Rename.Bind module. Instead, these TcRnMessage messages were introduced: TcRnMultipleFixityDecls TcRnIllegalPatternSynonymDecl TcRnIllegalClassBiding TcRnOrphanCompletePragma TcRnEmptyCase TcRnNonStdGuards TcRnDuplicateSigDecl TcRnMisplacedSigDecl TcRnUnexpectedDefaultSig TcRnBindInBootFile TcRnDuplicateMinimalSig - - - - - 3330b819 by Matthew Pickering at 2023-01-28T02:59:54-05:00 hadrian: Fix library-dirs, dynamic-library-dirs and static-library-dirs in inplace .conf files Previously we were just throwing away the contents of the library-dirs fields but really we have to do the same thing as for include-dirs, relativise the paths into the current working directory and maintain any extra libraries the user has specified. Now the relevant section of the rts.conf file looks like: ``` library-dirs: ${pkgroot}/../rts/build ${pkgroot}/../../..//_build/stage1/rts/build /nix/store/av4c0fib4rkmb6sa1074z0rb1ciria5b-gperftools-2.10/lib /nix/store/2infxahfp9lj084xn3q9ib5ajks8447i-libffi-3.4.4/lib library-dirs-static: ${pkgroot}/../rts/build ${pkgroot}/../../..//_build/stage1/rts/build /nix/store/av4c0fib4rkmb6sa1074z0rb1ciria5b-gperftools-2.10/lib /nix/store/2infxahfp9lj084xn3q9ib5ajks8447i-libffi-3.4.4/lib dynamic-library-dirs: ${pkgroot}/../rts/build ${pkgroot}/../../..//_build/stage1/rts/build /nix/store/av4c0fib4rkmb6sa1074z0rb1ciria5b-gperftools-2.10/lib /nix/store/2infxahfp9lj084xn3q9ib5ajks8447i-libffi-3.4.4/lib ``` Fixes #22209 - - - - - c9ad8852 by Andrew Lelechenko at 2023-01-28T03:00:33-05:00 Document differences between Data.{Monoid,Semigroup}.{First,Last} - - - - - 7e11c6dc by Cheng Shao at 2023-01-28T03:01:09-05:00 compiler: fix subword literal narrowing logic in the wasm NCG This patch fixes the W8/W16 literal narrowing logic in the wasm NCG, which used to lower it to something like i32.const -1, without properly zeroing-out the unused higher bits. Fixes #22608. - - - - - 6ea2aa02 by Cheng Shao at 2023-01-28T03:01:46-05:00 compiler: fix lowering of CmmBlock in the wasm NCG The CmmBlock datacon was not handled in lower_CmmLit, since I thought it would have been eliminated after proc-point splitting. Turns out it still occurs in very rare occasions, and this patch is needed to fix T9329 for wasm. - - - - - 2b62739d by Andrew Lelechenko at 2023-01-28T17:16:11-05:00 Assorted changes to avoid Data.List.{head,tail} - - - - - 78c07219 by Cheng Shao at 2023-01-28T17:16:48-05:00 compiler: properly handle ForeignHints in the wasm NCG Properly handle ForeignHints of ccall arguments/return value, insert sign extends and truncations when handling signed subwords. Fixes #22852. - - - - - 9e0fefd8 by Andrei Borzenkov at 2023-01-30T14:02:04+04:00 Rename () to Unit, Rename (,,...,,) to Tuple<n> - - - - - 8bed166b by Ben Gamari at 2023-01-30T05:06:26-05:00 nativeGen: Disable asm-shortcutting on Darwin Asm-shortcutting may produce relative references to symbols defined in other compilation units. This is not something that MachO relocations support (see #21972). For this reason we disable the optimisation on Darwin. We do so without a warning since this flag is enabled by `-O2`. Another way to address this issue would be to rather implement a PLT-relocatable jump-table strategy. However, this would only benefit Darwin and does not seem worth the effort. Closes #21972. - - - - - da468391 by Cheng Shao at 2023-01-30T05:07:03-05:00 compiler: fix data section alignment in the wasm NCG Previously we tried to lower the alignment requirement as far as possible, based on the section kind inferred from the CLabel. For info tables, .p2align 1 was applied given the GC should only need the lowest bit to tag forwarding pointers. But this would lead to unaligned loads/stores, which has a performance penalty even if the wasm spec permits it. Furthermore, the test suite has shown memory corruption in a few cases when compacting gc is used. This patch takes a more conservative approach: all data sections except C strings align to word size. - - - - - 08ba8720 by Andreas Klebinger at 2023-01-30T21:18:45-05:00 ghc-the-library: Retain cafs in both static in dynamic builds. We use keepCAFsForGHCi.c to force -fkeep-cafs behaviour by using a __attribute__((constructor)) function. This broke for static builds where the linker discarded the object file since it was not reverenced from any exported code. We fix this by asserting that the flag is enabled using a function in the same module as the constructor. Which causes the object file to be retained by the linker, which in turn causes the constructor the be run in static builds. This changes nothing for dynamic builds using the ghc library. But causes static to also retain CAFs (as we expect them to). Fixes #22417. ------------------------- Metric Decrease: T21839r ------------------------- - - - - - 20598ef6 by Ryan Scott at 2023-01-30T21:19:20-05:00 Handle `type data` properly in tyThingParent_maybe Unlike most other data constructors, data constructors declared with `type data` are represented in `TyThing`s as `ATyCon` rather than `ADataCon`. The `ATyCon` case in `tyThingParent_maybe` previously did not consider the possibility of the underlying `TyCon` being a promoted data constructor, which led to the oddities observed in #22817. This patch adds a dedicated special case in `tyThingParent_maybe`'s `ATyCon` case for `type data` data constructors to fix these oddities. Fixes #22817. - - - - - 2f145052 by Ryan Scott at 2023-01-30T21:19:56-05:00 Fix two bugs in TypeData TH reification This patch fixes two issues in the way that `type data` declarations were reified with Template Haskell: * `type data` data constructors are now properly reified using `DataConI`. This is accomplished with a special case in `reifyTyCon`. Fixes #22818. * `type data` type constructors are now reified in `reifyTyCon` using `TypeDataD` instead of `DataD`. Fixes #22819. - - - - - d0f34f25 by Simon Peyton Jones at 2023-01-30T21:20:35-05:00 Take account of loop breakers in specLookupRule The key change is that in GHC.Core.Opt.Specialise.specLookupRule we were using realIdUnfolding, which ignores the loop-breaker flag. When given a loop breaker, rule matching therefore looped infinitely -- #22802. In fixing this I refactored a bit. * Define GHC.Core.InScopeEnv as a data type, and use it. (Previously it was a pair: hard to grep for.) * Put several functions returning an IdUnfoldingFun into GHC.Types.Id, namely idUnfolding alwaysActiveUnfoldingFun, whenActiveUnfoldingFun, noUnfoldingFun and use them. (The are all loop-breaker aware.) - - - - - de963cb6 by Matthew Pickering at 2023-01-30T21:21:11-05:00 ci: Remove FreeBSD job from release pipelines We no longer attempt to build or distribute this release - - - - - f26d27ec by Matthew Pickering at 2023-01-30T21:21:11-05:00 rel_eng: Add check to make sure that release jobs are downloaded by fetch-gitlab This check makes sure that if a job is a prefixed by "release-" then the script downloads it and understands how to map the job name to the platform. - - - - - 7619c0b4 by Matthew Pickering at 2023-01-30T21:21:11-05:00 rel_eng: Fix the name of the ubuntu-* jobs These were not uploaded for alpha1 Fixes #22844 - - - - - 68eb8877 by Matthew Pickering at 2023-01-30T21:21:11-05:00 gen_ci: Only consider release jobs for job metadata In particular we do not have a release job for FreeBSD so the generation of the platform mapping was failing. - - - - - b69461a0 by Jason Shipman at 2023-01-30T21:21:50-05:00 User's guide: Clarify overlapping instance candidate elimination This commit updates the user's guide section on overlapping instance candidate elimination to use "or" verbiage instead of "either/or" in regards to the current pair of candidates' being overlappable or overlapping. "Either IX is overlappable, or IY is overlapping" can cause confusion as it suggests "Either IX is overlappable, or IY is overlapping, but not both". This was initially discussed on this Discourse topic: https://discourse.haskell.org/t/clarification-on-overlapping-instance-candidate-elimination/5677 - - - - - 7cbdaad0 by Matthew Pickering at 2023-01-31T07:53:53-05:00 Fixes for cabal-reinstall CI job * Allow filepath to be reinstalled * Bump some version bounds to allow newer versions of libraries * Rework testing logic to avoid "install --lib" and package env files Fixes #22344 - - - - - fd8f32bf by Cheng Shao at 2023-01-31T07:54:29-05:00 rts: prevent potential divide-by-zero when tickInterval=0 This patch fixes a few places in RtsFlags.c that may result in divide-by-zero error when tickInterval=0, which is the default on wasm. Fixes #22603. - - - - - 085a6db6 by Joachim Breitner at 2023-01-31T07:55:05-05:00 Update note at beginning of GHC.Builtin.NAmes some things have been renamed since it was written, it seems. - - - - - 7716cbe6 by Cheng Shao at 2023-01-31T07:55:41-05:00 testsuite: use tgamma for cg007 gamma is a glibc-only deprecated function, use tgamma instead. It's required for fixing cg007 when testing the wasm unregisterised codegen. - - - - - 19c1fbcd by doyougnu at 2023-01-31T13:08:03-05:00 InfoTableProv: ShortText --> ShortByteString - - - - - 765fab98 by doyougnu at 2023-01-31T13:08:03-05:00 FastString: add fastStringToShorText - - - - - a83c810d by Simon Peyton Jones at 2023-01-31T13:08:38-05:00 Improve exprOkForSpeculation for classops This patch fixes #22745 and #15205, which are about GHC's failure to discard unnecessary superclass selections that yield coercions. See GHC.Core.Utils Note [exprOkForSpeculation and type classes] The main changes are: * Write new Note [NON-BOTTOM_DICTS invariant] in GHC.Core, and refer to it * Define new function isTerminatingType, to identify those guaranteed-terminating dictionary types. * exprOkForSpeculation has a new (very simple) case for ClassOpId * ClassOpId has a new field that says if the return type is an unlifted type, or a terminating type. This was surprisingly tricky to get right. In particular note that unlifted types are not terminating types; you can write an expression of unlifted type, that diverges. Not so for dictionaries (or, more precisely, for the dictionaries that GHC constructs). Metric Decrease: LargeRecord - - - - - f83374f8 by Krzysztof Gogolewski at 2023-01-31T13:09:14-05:00 Support "unusable UNPACK pragma" warning with -O0 Fixes #11270 - - - - - a2d814dc by Ben Gamari at 2023-01-31T13:09:50-05:00 configure: Always create the VERSION file Teach the `configure` script to create the `VERSION` file. This will serve as the stable interface to allow the user to determine the version number of a working tree. Fixes #22322. - - - - - 5618fc21 by sheaf at 2023-01-31T15:51:06-05:00 Cmm: track the type of global registers This patch tracks the type of Cmm global registers. This is needed in order to lint uses of polymorphic registers, such as SIMD vector registers that can be used both for floating-point and integer values. This changes allows us to refactor VanillaReg to not store VGcPtr, as that information is instead stored in the type of the usage of the register. Fixes #22297 - - - - - 78b99430 by sheaf at 2023-01-31T15:51:06-05:00 Revert "Cmm Lint: relax SIMD register assignment check" This reverts commit 3be48877, which weakened a Cmm Lint check involving SIMD vectors. Now that we keep track of the type a global register is used at, we can restore the original stronger check. - - - - - be417a47 by Ben Gamari at 2023-01-31T15:51:45-05:00 nativeGen/AArch64: Fix debugging output Previously various panics would rely on a half-written Show instance, leading to very unhelpful errors. Fix this. See #22798. - - - - - 30989d13 by Ben Gamari at 2023-01-31T15:51:45-05:00 nativeGen: Teach graph-colouring allocator that x18 is unusable Previously trivColourable for AArch64 claimed that at 18 registers were trivially-colourable. This is incorrect as x18 is reserved by the platform on AArch64/Darwin. See #22798. - - - - - 7566fd9d by Ben Gamari at 2023-01-31T15:51:45-05:00 nativeGen/AArch64: Fix graph-colouring allocator Previously various `Instr` queries used by the graph-colouring allocator failed to handle a few pseudo-instructions. This manifested in compiler panicks while compiling `SHA`, which uses `-fregs-graph`. Fixes #22798. - - - - - 2cb500a5 by Ben Gamari at 2023-01-31T15:51:45-05:00 testsuite: Add regression test for #22798 - - - - - 03d693b2 by Ben Gamari at 2023-01-31T15:52:32-05:00 Revert "Hadrian: fix doc generation" This is too large of a hammer. This reverts commit 5640cb1d84d3cce4ce0a9e90d29b2b20d2b38c2f. - - - - - f838815c by Ben Gamari at 2023-01-31T15:52:32-05:00 hadrian: Sphinx docs require templated cabal files The package-version discovery logic in `doc/users_guide/package_versions.py` uses packages' cabal files to determine package versions. Teach Sphinx about these dependencies in cases where the cabal files are generated by templates. - - - - - 2e48c19a by Ben Gamari at 2023-01-31T15:52:32-05:00 hadrian: Refactor templating logic This refactors Hadrian's autoconf-style templating logic to be explicit about which interpolation variables should be substituted in which files. This clears the way to fix #22714 without incurring rule cycles. - - - - - 93f0e3c4 by Ben Gamari at 2023-01-31T15:52:33-05:00 hadrian: Substitute LIBRARY_*_VERSION variables This teaches Hadrian to substitute the `LIBRARY_*_VERSION` variables in `libraries/prologue.txt`, fixing #22714. Fixes #22714. - - - - - 22089f69 by Ben Gamari at 2023-01-31T20:46:27-05:00 Bump transformers submodule to 0.6.0.6 Fixes #22862. - - - - - f0eefa3c by Cheng Shao at 2023-01-31T20:47:03-05:00 compiler: properly handle non-word-sized CmmSwitch scrutinees in the wasm NCG Currently, the wasm NCG has an implicit assumption: all CmmSwitch scrutinees are 32-bit integers. This is not always true; #22864 is one counter-example with a 64-bit scrutinee. This patch fixes the logic by explicitly converting the scrutinee to a word that can be used as a br_table operand. Fixes #22871. Also includes a regression test. - - - - - 9f95db54 by Simon Peyton Jones at 2023-02-01T08:55:08+00:00 Improve treatment of type applications in patterns This patch fixes a subtle bug in the typechecking of type applications in patterns, e.g. f (MkT @Int @a x y) = ... See Note [Type applications in patterns] in GHC.Tc.Gen.Pat. This fixes #19847, #22383, #19577, #21501 - - - - - 955a99ea by Simon Peyton Jones at 2023-02-01T12:31:23-05:00 Treat existentials correctly in dubiousDataConInstArgTys Consider (#22849) data T a where MkT :: forall k (t::k->*) (ix::k). t ix -> T @k a Then dubiousDataConInstArgTys MkT [Type, Foo] should return [Foo (ix::Type)] NOT [Foo (ix::k)] A bit of an obscure case, but it's an outright bug, and the fix is easy. - - - - - 0cc16aaf by Matthew Pickering at 2023-02-01T12:31:58-05:00 Bump supported LLVM range from 10 through 15 to 11 through 16 LLVM 15 turns on the new pass manager by default, which we have yet to migrate to so for new we pass the `-enable-new-pm-0` flag in our llvm-passes flag. LLVM 11 was the first version to support the `-enable-new-pm` flag so we bump the lowest supported version to 11. Our CI jobs are using LLVM 12 so they should continue to work despite this bump to the lower bound. Fixes #21936 - - - - - f94f1450 by Matthew Pickering at 2023-02-01T12:31:58-05:00 Bump DOCKER_REV to use alpine image without LLVM installed alpine_3_12 only supports LLVM 10, which is now outside the supported version range. - - - - - 083e26ed by Matthew Pickering at 2023-02-01T17:43:21-05:00 Remove tracing OPTIONS_GHC These were accidentally left over from !9542 - - - - - 354aa47d by Teo Camarasu at 2023-02-01T17:44:00-05:00 doc: fix gcdetails_block_fragmentation_bytes since annotation - - - - - 61ce5bf6 by Jaro Reinders at 2023-02-02T00:15:30-05:00 compiler: Implement higher order patterns in the rule matcher This implements proposal 555 and closes ticket #22465. See the proposal and ticket for motivation. The core changes of this patch are in the GHC.Core.Rules.match function and they are explained in the Note [Matching higher order patterns]. - - - - - 394b91ce by doyougnu at 2023-02-02T00:16:10-05:00 CI: JavaScript backend runs testsuite This MR runs the testsuite for the JS backend. Note that this is a temporary solution until !9515 is merged. Key point: The CI runs hadrian on the built cross compiler _but not_ on the bindist. Other Highlights: - stm submodule gets a bump to mark tests as broken - several tests are marked as broken or are fixed by adding more - conditions to their test runner instance. List of working commit messages: CI: test cross target _and_ emulator CI: JS: Try run testsuite with hadrian JS.CI: cleanup and simplify hadrian invocation use single bracket, print info JS CI: remove call to test_compiler from hadrian don't build haddock JS: mark more tests as broken Tracked in https://gitlab.haskell.org/ghc/ghc/-/issues/22576 JS testsuite: don't skip sum_mod test Its expected to fail, yet we skipped it which automatically makes it succeed leading to an unexpected success, JS testsuite: don't mark T12035j as skip leads to an unexpected pass JS testsuite: remove broken on T14075 leads to unexpected pass JS testsuite: mark more tests as broken JS testsuite: mark T11760 in base as broken JS testsuite: mark ManyUnbSums broken submodules: bump process and hpc for JS tests Both submodules has needed tests skipped or marked broken for th JS backend. This commit now adds these changes to GHC. See: HPC: https://gitlab.haskell.org/hpc/hpc/-/merge_requests/21 Process: https://github.com/haskell/process/pull/268 remove js_broken on now passing tests separate wasm and js backend ci test: T11760: add threaded, non-moving only_ways test: T10296a add req_c T13894: skip for JS backend tests: jspace, T22333: mark as js_broken(22573) test: T22513i mark as req_th stm submodule: mark stm055, T16707 broken for JS tests: js_broken(22374) on unpack_sums_6, T12010 dont run diff on JS CI, cleanup fixup: More CI cleanup fix: align text to master fix: align exceptions submodule to master CI: Bump DOCKER_REV Bump to ci-images commit that has a deb11 build with node. Required for !9552 testsuite: mark T22669 as js_skip See #22669 This test tests that .o-boot files aren't created when run in using the interpreter backend. Thus this is not relevant for the JS backend. testsuite: mark T22671 as broken on JS See #22835 base.testsuite: mark Chan002 fragile for JS see #22836 revert: submodule process bump bump stm submodule New hash includes skips for the JS backend. testsuite: mark RnPatternSynonymFail broken on JS Requires TH: - see !9779 - and #22261 compiler: GHC.hs ifdef import Utils.Panic.Plain - - - - - 1ffe770c by Cheng Shao at 2023-02-02T09:40:38+00:00 docs: 9.6 release notes for wasm backend - - - - - 0ada4547 by Matthew Pickering at 2023-02-02T11:39:44-05:00 Disable unfolding sharing for interface files with core definitions Ticket #22807 pointed out that the RHS sharing was not compatible with -fignore-interface-pragmas because the flag would remove unfoldings from identifiers before the `extra-decls` field was populated. For the 9.6 timescale the only solution is to disable this sharing, which will make interface files bigger but this is acceptable for the first release of `-fwrite-if-simplified-core`. For 9.8 it would be good to fix this by implementing #20056 due to the large number of other bugs that would fix. I also improved the error message in tc_iface_binding to avoid the "no match in record selector" error but it should never happen now as the entire sharing logic is disabled. Also added the currently broken test for #22807 which could be fixed by !6080 Fixes #22807 - - - - - 7e2d3eb5 by lrzlin at 2023-02-03T05:23:27-05:00 Enable tables next to code for LoongArch64 - - - - - 2931712a by Wander Hillen at 2023-02-03T05:24:06-05:00 Move pthread and timerfd ticker implementations to separate files - - - - - 41c4baf8 by Ben Gamari at 2023-02-03T05:24:44-05:00 base: Fix Note references in GHC.IO.Handle.Types - - - - - 31358198 by Andrew Lelechenko at 2023-02-03T05:25:22-05:00 Bump submodule containers to 0.6.7 Metric Decrease: ManyConstructors T10421 T12425 T12707 T13035 T13379 T15164 T1969 T783 T9198 T9961 WWRec - - - - - 8feb9301 by Ben Gamari at 2023-02-03T05:25:59-05:00 gitlab-ci: Eliminate redundant ghc --info output Previously ci.sh would emit the output of `ghc --info` every time it ran when using the nix toolchain. This produced a significant amount of noise. See #22861. - - - - - de1d1512 by Ryan Scott at 2023-02-03T14:07:30-05:00 Windows: Remove mingwex dependency The clang based toolchain uses ucrt as its math library and so mingwex is no longer needed. In fact using mingwex will cause incompatibilities as the default routines in both have differing ULPs and string formatting modifiers. ``` $ LIBRARY_PATH=/mingw64/lib ghc/_build/stage1/bin/ghc Bug.hs -fforce-recomp && ./Bug.exe [1 of 2] Compiling Main ( Bug.hs, Bug.o ) ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `__imp___p__environ' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `__hscore_get_errno' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_ForeignziCziError_errnoToIOError_info' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziWindows_failIf2_closure' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziIOziEncodingziCodePageziAPI_mkCodePageEncoding_info' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziIOziEncodingziCodePage_currentCodePage_closure' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziIOziEncoding_getForeignEncoding_closure' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_ForeignziCziString_withCStringLen1_info' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziIOziHandleziInternals_zdwflushCharReadBuffer_info' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziIOziHandleziText_hGetBuf1_info' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziFingerprint_fingerprintString_closure' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_DataziTypeableziInternal_mkTrCon_closure' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziException_errorCallWithCallStackException_closure' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\base-4.17.0.0\libHSbase-4.17.0.0.a: unknown symbol `base_GHCziErr_error_info' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\template-haskell-2.19.0.0\libHStemplate-haskell-2.19.0.0.a: unknown symbol `base_DataziMaybe_fromJust1_info' ghc.exe: | C:\Users\winferno\Software\ghc\_build\stage1\lib\x86_64-windows-ghc-9.5.20220908\template-haskell-2.19.0.0\libHStemplate-haskell-2.19.0.0.a: unknown symbol `templatezmhaskell_LanguageziHaskellziTHziSyntax_IntPrimL_con_info' ghc.exe: ^^ Could not load 'templatezmhaskell_LanguageziHaskellziTHziLibziInternal_stringL_closure', dependency unresolved. See top entry above. <no location info>: error: GHC.ByteCode.Linker.lookupCE During interactive linking, GHCi couldn't find the following symbol: templatezmhaskell_LanguageziHaskellziTHziLibziInternal_stringL_closure This may be due to you not asking GHCi to load extra object files, archives or DLLs needed by your current session. Restart GHCi, specifying the missing library using the -L/path/to/object/dir and -lmissinglibname flags, or simply by naming the relevant files on the GHCi command line. Alternatively, this link failure might indicate a bug in GHCi. If you suspect the latter, please report this as a GHC bug: https://www.haskell.org/ghc/reportabug ``` - - - - - 48e39195 by Tamar Christina at 2023-02-03T14:07:30-05:00 linker: Fix BFD import libraries This commit fixes the BFD style import library support in the runtime linker. This was accidentally broken during the refactoring to clang and went unnoticed because clang itself is unable to generate the BFD style import libraries. With this change we can not link against both GCC or Clang produced libraries again and intermix code produced by both compilers. - - - - - b2bb3e62 by Ben Gamari at 2023-02-03T14:07:30-05:00 Bump Windows toolchain Updates to LLVM 14, hopefully fixing #21964. - - - - - bf3f88a1 by Andreas Klebinger at 2023-02-03T14:08:07-05:00 Fix CallerCC potentially shadowing other cost centres. Add a CallerCC cost centre flavour for cost centres added by the CallerCC pass. This avoids potential accidental shadowing between CCs added by user annotations and ones added by CallerCC. - - - - - faea4bcd by j at 2023-02-03T14:08:47-05:00 Disable several ignore-warning flags in genapply. - - - - - 25537dfd by Ben Gamari at 2023-02-04T04:12:57-05:00 Revert "Use fix-sized bit-fiddling primops for fixed size boxed types" This reverts commit 4512ad2d6a8e65ea43c86c816411cb13b822f674. This was never applied to master/9.6 originally. (cherry picked from commit a44bdc2720015c03d57f470b759ece7fab29a57a) - - - - - 7612dc71 by Krzysztof Gogolewski at 2023-02-04T04:13:34-05:00 Minor refactor * Introduce refactorDupsOn f = refactorDups (comparing f) * Make mkBigTupleCase and coreCaseTuple monadic. Every call to those functions was preceded by calling newUniqueSupply. * Use mkUserLocalOrCoVar, which is equivalent to combining mkLocalIdOrCoVar with mkInternalName. - - - - - 5a54ac0b by Andrew Lelechenko at 2023-02-04T18:48:32-05:00 Fix colors in emacs terminal - - - - - 3c0f0c6d by Andrew Lelechenko at 2023-02-04T18:49:11-05:00 base changelog: move entries which were not backported to ghc-9.6 to base-4.19 section - - - - - b18fbf52 by Josh Meredith at 2023-02-06T07:47:57+00:00 Update JavaScript fileStat to match Emscripten layout - - - - - 6636b670 by Sylvain Henry at 2023-02-06T09:43:21-05:00 JS: replace "js" architecture with "javascript" Despite Cabal supporting any architecture name, `cabal --check` only supports a few built-in ones. Sadly `cabal --check` is used by Hackage hence using any non built-in name in a package (e.g. `arch(js)`) is rejected and the package is prevented from being uploaded on Hackage. Luckily built-in support for the `javascript` architecture was added for GHCJS a while ago. In order to allow newer `base` to be uploaded on Hackage we make the switch from `js` to `javascript` architecture. Fixes #22740. Co-authored-by: Ben Gamari <ben at smart-cactus.org> - - - - - 77a8234c by Luite Stegeman at 2023-02-06T09:43:59-05:00 Fix marking async exceptions in the JS backend Async exceptions are posted as a pair of the exception and the thread object. This fixes the marking pass to correctly follow the two elements of the pair. Potentially fixes #22836 - - - - - 3e09cf82 by Jan Hrček at 2023-02-06T09:44:38-05:00 Remove extraneous word in Roles user guide - - - - - b17fb3d9 by sheaf at 2023-02-07T10:51:33-05:00 Don't allow . in overloaded labels This patch removes . from the list of allowed characters in a non-quoted overloaded label, as it was realised this steals syntax, e.g. (#.). Users who want this functionality will have to add quotes around the label, e.g. `#"17.28"`. Fixes #22821 - - - - - 5dce04ee by romes at 2023-02-07T10:52:10-05:00 Update kinds in comments in GHC.Core.TyCon Use `Type` instead of star kind (*) Fix comment with incorrect kind * to have kind `Constraint` - - - - - 92916194 by Ben Gamari at 2023-02-07T10:52:48-05:00 Revert "Use fix-sized equality primops for fixed size boxed types" This reverts commit 024020c38126f3ce326ff56906d53525bc71690c. This was never applied to master/9.6 originally. See #20405 for why using these primops is a bad idea. (cherry picked from commit b1d109ad542e4c37ae5af6ace71baf2cb509d865) - - - - - c1670c6b by Sylvain Henry at 2023-02-07T21:25:18-05:00 JS: avoid head/tail and unpackFS - - - - - a9912de7 by Krzysztof Gogolewski at 2023-02-07T21:25:53-05:00 testsuite: Fix Python warnings (#22856) - - - - - 9ee761bf by sheaf at 2023-02-08T14:40:40-05:00 Fix tyvar scoping within class SPECIALISE pragmas Type variables from class/instance headers scope over class/instance method type signatures, but DO NOT scope over the type signatures in SPECIALISE and SPECIALISE instance pragmas. The logic in GHC.Rename.Bind.rnMethodBinds correctly accounted for SPECIALISE inline pragmas, but forgot to apply the same treatment to method SPECIALISE pragmas, which lead to a Core Lint failure with an out-of-scope type variable. This patch makes sure we apply the same logic for both cases. Fixes #22913 - - - - - 7eac2468 by Matthew Pickering at 2023-02-08T14:41:17-05:00 Revert "Don't keep exit join points so much" This reverts commit caced75765472a1a94453f2e5a439dba0d04a265. It seems the patch "Don't keep exit join points so much" is causing wide-spread regressions in the bytestring library benchmarks. If I revert it then the 9.6 numbers are better on average than 9.4. See https://gitlab.haskell.org/ghc/ghc/-/issues/22893#note_479525 ------------------------- Metric Decrease: MultiComponentModules MultiComponentModulesRecomp MultiLayerModules MultiLayerModulesRecomp MultiLayerModulesTH_Make T12150 T13386 T13719 T21839c T3294 parsing001 ------------------------- - - - - - 633f2799 by Cheng Shao at 2023-02-08T18:42:16-05:00 testsuite: remove config.use_threads This patch simplifies the testsuite driver by removing the use_threads config field. It's just a degenerate case of threads=1. - - - - - ca6673e3 by Cheng Shao at 2023-02-08T18:42:16-05:00 testsuite: use concurrent.futures.ThreadPoolExecutor in the driver The testsuite driver used to create one thread per test case, and explicitly use semaphore and locks for rate limiting and synchronization. This is a bad practice in any language, and occasionally may result in livelock conditions (e.g. #22889). This patch uses concurrent.futures.ThreadPoolExecutor for scheduling test case runs, which is simpler and more robust. - - - - - f22cce70 by Alan Zimmerman at 2023-02-08T18:42:51-05:00 EPA: Comment between module and where should be in header comments Do not apply the heuristic to associate a comment with a prior declaration for the first declaration in the file. Closes #22919 - - - - - d69ecac2 by Josh Meredith at 2023-02-09T03:24:05-05:00 JS generated refs: update testsuite conditions - - - - - 2ea1a6bc by sheaf at 2023-02-09T03:24:44-05:00 Bump transformers to 0.6.1.0 This allows us to avoid orphans for Foldable1 instances, fixing #22898. Updates transformers submodule. - - - - - d9d0c28d by konsumlamm at 2023-02-09T14:07:48-05:00 Update `Data.List.singleton` doc comment - - - - - fe9cd6ef by Ben Gamari at 2023-02-09T14:08:23-05:00 gitlab-template: Emphasize `user facing` label My sense is that the current mention of the ~"user facing" label is overlooked by many MR authors. Let's move this point up in the list to make it more likely that it is seen. Also rephrase some of the points. - - - - - e45eb828 by Simon Peyton Jones at 2023-02-10T06:51:28-05:00 Refactor the simplifier a bit to fix #22761 The core change in this commit, which fixes #22761, is that * In a Core rule, ru_rhs is always occ-analysed. This means adding a couple of calls to occurAnalyseExpr when building a Rule, in * GHC.Core.Rules.mkRule * GHC.Core.Opt.Simplify.Iteration.simplRules But diagosing the bug made me stare carefully at the code of the Simplifier, and I ended up doing some only-loosely-related refactoring. * I think that RULES could be lost because not every code path did addBndrRules * The code around lambdas was very convoluted It's mainly moving deck chairs around, but I like it more now. - - - - - 11e0cacb by Rebecca Turner at 2023-02-10T06:52:09-05:00 Detect the `mold` linker Enables support for the `mold` linker by rui314. - - - - - 59556235 by parsonsmatt at 2023-02-10T09:53:11-05:00 Add Lift instance for Fixed - - - - - c44e5f30 by Sylvain Henry at 2023-02-10T09:53:51-05:00 Testsuite: decrease length001 timeout for JS (#22921) - - - - - 133516af by Zubin Duggal at 2023-02-10T09:54:27-05:00 compiler: Use NamedFieldPuns for `ModIface_` and `ModIfaceBackend` `NFData` instances This is a minor refactor that makes it easy to add and remove fields from `ModIface_` and `ModIfaceBackend`. Also change the formatting to make it clear exactly which fields are fully forced with `rnf` - - - - - 1e9eac1c by Matthew Pickering at 2023-02-13T11:36:41+01:00 Refresh profiling docs I went through the whole of the profiling docs and tried to amend them to reflect current best practices and tooling. In particular I removed some old references to tools such as hp2any and replaced them with references to eventlog2html. - - - - - da208b9a by Matthew Pickering at 2023-02-13T11:36:41+01:00 docs: Add section about profiling and foreign calls Previously there was no documentation for how foreign calls interacted with the profiler. This can be quite confusing for users so getting it into the user guide is the first step to a potentially better solution. See the ticket for more insightful discussion. Fixes #21764 - - - - - 081640f1 by Andrew Lelechenko at 2023-02-13T12:51:52-05:00 Document that -fproc-alignment was introduced only in GHC 8.6 - - - - - 16adc349 by Sven Tennie at 2023-02-14T11:26:31-05:00 Add clangd flag to include generated header files This enables clangd to correctly check C files that import Rts.h. (The added include directory contains ghcautoconf.h et. al.) - - - - - c399ccd9 by amesgen at 2023-02-14T11:27:14-05:00 Mention new `Foreign.Marshal.Pool` implementation in User's Guide - - - - - b9282cf7 by Ben Gamari at 2023-02-14T11:27:50-05:00 upload_ghc_libs: More control over which packages to operate on Here we add a `--skip` flag to `upload_ghc_libs`, making it easier to limit which packages to upload. This is often necessary when one package is not uploadable (e.g. see #22740). - - - - - aa3a262d by PHO at 2023-02-14T11:28:29-05:00 Assume platforms support rpaths if they use either ELF or Mach-O Not only Linux, Darwin, and FreeBSD support rpaths. Determine the usability of rpaths based on the object format, not on OS. - - - - - 47716024 by PHO at 2023-02-14T11:29:09-05:00 RTS linker: Improve compatibility with NetBSD 1. Hint address to NetBSD mmap(2) has a different semantics from that of Linux. When a hint address is provided, mmap(2) searches for a free region at or below the hint but *never* above it. This means we can't reliably search for free regions incrementally on the userland, especially when ASLR is enabled. Let the kernel do it for us if we don't care where the mapped address is going to be. 2. NetBSD not only hates to map pages as rwx, but also disallows to switch pages from rw- to r-x unless the intention is declared when pages are initially requested. This means we need a new MemoryAccess mode for pages that are going to be changed to r-x. - - - - - 11de324a by Li-yao Xia at 2023-02-14T11:29:49-05:00 base: Move changelog entry to its place - - - - - 75930424 by Ben Gamari at 2023-02-14T11:30:27-05:00 nativeGen/AArch64: Emit Atomic{Read,Write} inline Previously the AtomicRead and AtomicWrite operations were emitted as out-of-line calls. However, these tend to be very important for performance, especially the RELAXED case (which only exists for ThreadSanitizer checking). Fixes #22115. - - - - - d6411d6c by Andreas Klebinger at 2023-02-14T11:31:04-05:00 Fix some correctness issues around tag inference when targeting the bytecode generator. * Let binders are now always assumed untagged for bytecode. * Imported referenced are now always assumed to be untagged for bytecode. Fixes #22840 - - - - - 9fb4ca89 by sheaf at 2023-02-14T11:31:49-05:00 Introduce warning for loopy superclass solve Commit aed1974e completely re-engineered the treatment of loopy superclass dictionaries in instance declarations. Unfortunately, it has the potential to break (albeit in a rather minor way) user code. To alleviate migration concerns, this commit re-introduces the old behaviour. Any reliance on this old behaviour triggers a warning, controlled by `-Wloopy-superclass-solve`. The warning text explains that GHC might produce bottoming evidence, and provides a migration strategy. This allows us to provide a graceful migration period, alerting users when they are relying on this unsound behaviour. Fixes #22912 #22891 #20666 #22894 #22905 - - - - - 1928c7f3 by Cheng Shao at 2023-02-14T11:32:26-05:00 rts: make it possible to change mblock size on 32-bit targets The MBLOCK_SHIFT macro must be the single source of truth for defining the mblock size, and changing it should only affect performance, not correctness. This patch makes it truly possible to reconfigure mblock size, at least on 32-bit targets, by fixing places which implicitly relied on the previous MBLOCK_SHIFT constant. Fixes #22901. - - - - - 78aa3b39 by Simon Hengel at 2023-02-14T11:33:06-05:00 Update outdated references to notes - - - - - e8baecd2 by meooow25 at 2023-02-14T11:33:49-05:00 Documentation: Improve Foldable1 documentation * Explain foldrMap1, foldlMap1, foldlMap1', and foldrMap1' in greater detail, the text is mostly adapted from documentation of Foldable. * Describe foldr1, foldl1, foldl1' and foldr1' in terms of the above functions instead of redoing the full explanation. * Small updates to documentation of fold1, foldMap1 and toNonEmpty, again adapting from Foldable. * Update the foldMap1 example to lists instead of Sum since this is recommended for lazy right-associative folds. Fixes #22847 - - - - - 85a1a575 by romes at 2023-02-14T11:34:25-05:00 fix: Mark ghci Prelude import as implicit Fixes #22829 In GHCi, we were creating an import declaration for Prelude but we were not setting it as an implicit declaration. Therefore, ghci's import of Prelude triggered -Wmissing-import-lists. Adds regression test T22829 to testsuite - - - - - 3b019a7a by Cheng Shao at 2023-02-14T11:35:03-05:00 compiler: fix generateCgIPEStub for no-tables-next-to-code builds generateCgIPEStub already correctly implements the CmmTick finding logic for when tables-next-to-code is on/off, but it used the wrong predicate to decide when to switch between the two. Previously it switches based on whether the codegen is unregisterised, but there do exist registerised builds that disable tables-next-to-code! This patch corrects that problem. Fixes #22896. - - - - - 08c0822c by doyougnu at 2023-02-15T00:16:39-05:00 docs: release notes, user guide: add js backend Follow up from #21078 - - - - - 79d8fd65 by Bryan Richter at 2023-02-15T00:17:15-05:00 Allow failure in nightly-x86_64-linux-deb10-no_tntc-validate See #22343 - - - - - 9ca51f9e by Cheng Shao at 2023-02-15T00:17:53-05:00 rts: add the rts_clearMemory function This patch adds the rts_clearMemory function that does its best to zero out unused RTS memory for a wasm backend use case. See the comment above rts_clearMemory() prototype declaration for more detailed explanation. Closes #22920. - - - - - 26df73fb by Oleg Grenrus at 2023-02-15T22:20:57-05:00 Add -single-threaded flag to force single threaded rts This is the small part of implementing https://github.com/ghc-proposals/ghc-proposals/pull/240 - - - - - 631c6c72 by Cheng Shao at 2023-02-16T06:43:09-05:00 docs: add a section for the wasm backend Fixes #22658 - - - - - 1878e0bd by Bryan Richter at 2023-02-16T06:43:47-05:00 tests: Mark T12903 fragile everywhere See #21184 - - - - - b9420eac by Bryan Richter at 2023-02-16T06:43:47-05:00 Mark all T5435 variants as fragile See #22970. - - - - - df3d94bd by Sylvain Henry at 2023-02-16T06:44:33-05:00 Testsuite: mark T13167 as fragile for JS (#22921) - - - - - 324e925b by Sylvain Henry at 2023-02-16T06:45:15-05:00 JS: disable debugging info for heap objects - - - - - 518af814 by Josh Meredith at 2023-02-16T10:16:32-05:00 Factor JS Rts generation for h$c{_,0,1,2} into h$c{n} and improve name caching - - - - - 34cd308e by Ben Gamari at 2023-02-16T10:17:08-05:00 base: Note move of GHC.Stack.CCS.whereFrom to GHC.InfoProv in changelog Fixes #22883. - - - - - 12965aba by Simon Peyton Jones at 2023-02-16T10:17:46-05:00 Narrow the dont-decompose-newtype test Following #22924 this patch narrows the test that stops us decomposing newtypes. The key change is the use of noGivenNewtypeReprEqs in GHC.Tc.Solver.Canonical.canTyConApp. We went to and fro on the solution, as you can see in #22924. The result is carefully documented in Note [Decomoposing newtype equalities] On the way I had revert most of commit 3e827c3f74ef76d90d79ab6c4e71aa954a1a6b90 Author: Richard Eisenberg <rae at cs.brynmawr.edu> Date: Mon Dec 5 10:14:02 2022 -0500 Do newtype unwrapping in the canonicaliser and rewriter See Note [Unwrap newtypes first], which has the details. It turns out that (a) 3e827c3f makes GHC behave worse on some recursive newtypes (see one of the tests on this commit) (b) the finer-grained test (namely noGivenNewtypeReprEqs) renders 3e827c3f unnecessary - - - - - 5b038888 by Andrew Lelechenko at 2023-02-16T10:18:24-05:00 Documentation: add an example of SPEC usage - - - - - 681e0e8c by sheaf at 2023-02-16T14:09:56-05:00 No default finalizer exception handler Commit cfc8e2e2 introduced a mechanism for handling of exceptions that occur during Handle finalization, and 372cf730 set the default handler to print out the error to stderr. However, #21680 pointed out we might not want to set this by default, as it might pollute users' terminals with unwanted information. So, for the time being, the default handler discards the exception. Fixes #21680 - - - - - b3ac17ad by Matthew Pickering at 2023-02-16T14:10:31-05:00 unicode: Don't inline bitmap in generalCategory generalCategory contains a huge literal string but is marked INLINE, this will duplicate the string into any use site of generalCategory. In particular generalCategory is used in functions like isSpace and the literal gets inlined into this function which makes it massive. https://github.com/haskell/core-libraries-committee/issues/130 Fixes #22949 ------------------------- Metric Decrease: T4029 T18304 ------------------------- - - - - - 8988eeef by sheaf at 2023-02-16T20:32:27-05:00 Expand synonyms in RoughMap We were failing to expand type synonyms in the function GHC.Core.RoughMap.typeToRoughMatchLookupTc, even though the RoughMap infrastructure crucially relies on type synonym expansion to work. This patch adds the missing type-synonym expansion. Fixes #22985 - - - - - 3dd50e2f by Matthew Pickering at 2023-02-16T20:33:03-05:00 ghcup-metadata: Add test artifact Add the released testsuite tarball to the generated ghcup metadata. - - - - - c6a967d9 by Matthew Pickering at 2023-02-16T20:33:03-05:00 ghcup-metadata: Use Ubuntu and Rocky bindists Prefer to use the Ubuntu 20.04 and 18.04 binary distributions on Ubuntu and Linux Mint. Prefer to use the Rocky 8 binary distribution on unknown distributions. - - - - - be0b7209 by Matthew Pickering at 2023-02-17T09:37:16+00:00 Add INLINABLE pragmas to `generic*` functions in Data.OldList These functions are * recursive * overloaded So it's important to add an `INLINABLE` pragma to each so that they can be specialised at the use site when the specific numeric type is known. Adding these pragmas improves the LazyText replicate benchmark (see https://gitlab.haskell.org/ghc/ghc/-/issues/22886#note_481020) https://github.com/haskell/core-libraries-committee/issues/129 - - - - - a203ad85 by Sylvain Henry at 2023-02-17T15:59:16-05:00 Merge libiserv with ghci `libiserv` serves no purpose. As it depends on `ghci` and doesn't have more dependencies than the `ghci` package, its code could live in the `ghci` package too. This commit also moves most of the code from the `iserv` program into the `ghci` package as well so that it can be reused. This is especially useful for the implementation of TH for the JS backend (#22261, !9779). - - - - - 7080a93f by Simon Peyton Jones at 2023-02-20T12:06:32+01:00 Improve GHC.Tc.Gen.App.tcInstFun It wasn't behaving right when inst_final=False, and the function had no type variables f :: Foo => Int Rather a corner case, but we might as well do it right. Fixes #22908 Unexpectedly, three test cases (all using :type in GHCi) got slightly better output as a result: T17403, T14796, T12447 - - - - - 2592ab69 by Cheng Shao at 2023-02-20T10:35:30-05:00 compiler: fix cost centre profiling breakage in wasm NCG due to incorrect register mapping The wasm NCG used to map CCCS to a wasm global, based on the observation that CCCS is a transient register that's already handled by thread state load/store logic, so it doesn't need to be backed by the rCCCS field in the register table. Unfortunately, this is wrong, since even when Cmm execution hasn't yielded back to the scheduler, the Cmm code may call enterFunCCS, which does use rCCCS. This breaks cost centre profiling in a subtle way, resulting in inaccurate stack traces in some test cases. The fix is simple though: just remove the CCCS mapping. - - - - - 26243de1 by Alexis King at 2023-02-20T15:27:17-05:00 Handle top-level Addr# literals in the bytecode compiler Fixes #22376. - - - - - 0196cc2b by romes at 2023-02-20T15:27:52-05:00 fix: Explicitly flush stdout on plugin Because of #20791, the plugins tests often fail. This is a temporary fix to stop the tests from failing due to unflushed outputs on windows and the explicit flush should be removed when #20791 is fixed. - - - - - 4327d635 by Ryan Scott at 2023-02-20T20:44:34-05:00 Don't generate datacon wrappers for `type data` declarations Data constructor wrappers only make sense for _value_-level data constructors, but data constructors for `type data` declarations only exist at the _type_ level. This patch does the following: * The criteria in `GHC.Types.Id.Make.mkDataConRep` for whether a data constructor receives a wrapper now consider whether or not its parent data type was declared with `type data`, omitting a wrapper if this is the case. * Now that `type data` data constructors no longer receive wrappers, there is a spot of code in `refineDefaultAlt` that panics when it encounters a value headed by a `type data` type constructor. I've fixed this with a special case in `refineDefaultAlt` and expanded `Note [Refine DEFAULT case alternatives]` to explain why we do this. Fixes #22948. - - - - - 96dc58b9 by Ryan Scott at 2023-02-20T20:44:35-05:00 Treat type data declarations as empty when checking pattern-matching coverage The data constructors for a `type data` declaration don't exist at the value level, so we don't want GHC to warn users to match on them. Fixes #22964. - - - - - ff8e99f6 by Ryan Scott at 2023-02-20T20:44:35-05:00 Disallow `tagToEnum#` on `type data` types We don't want to allow users to conjure up values of a `type data` type using `tagToEnum#`, as these simply don't exist at the value level. - - - - - 8e765aff by Andrew Lelechenko at 2023-02-21T12:03:24-05:00 Bump submodule text to 2.0.2 - - - - - 172ff88f by Georgi Lyubenov at 2023-02-21T18:35:56-05:00 GHC proposal 496 - Nullary record wildcards This patch implements GHC proposal 496, which allows record wildcards to be used for nullary constructors, e.g. data A = MkA1 | MkA2 { fld1 :: Int } f :: A -> Int f (MkA1 {..}) = 0 f (MkA2 {..}) = fld1 To achieve this, we add arity information to the record field environment, so that we can accept a constructor which has no fields while continuing to reject non-record constructors with more than 1 field. See Note [Nullary constructors and empty record wildcards], as well as the more general overview in Note [Local constructor info in the renamer], both in the newly introduced GHC.Types.ConInfo module. Fixes #22161 - - - - - f70a0239 by sheaf at 2023-02-21T18:36:35-05:00 ghc-prim: levity-polymorphic array equality ops This patch changes the pointer-equality comparison operations in GHC.Prim.PtrEq to work with arrays of unlifted values, e.g. sameArray# :: forall {l} (a :: TYPE (BoxedRep l)). Array# a -> Array# a -> Int# Fixes #22976 - - - - - 9296660b by Andreas Klebinger at 2023-02-21T23:58:05-05:00 base: Correct @since annotation for FP<->Integral bit cast operations. Fixes #22708 - - - - - f11d9c27 by romes at 2023-02-21T23:58:42-05:00 fix: Update documentation links Closes #23008 Additionally batches some fixes to pointers to the Note [Wired-in units], and a typo in said note. - - - - - fb60339f by Bryan Richter at 2023-02-23T14:45:17+02:00 Propagate failure if unable to push notes - - - - - 8e170f86 by Alexis King at 2023-02-23T16:59:22-05:00 rts: Fix `prompt#` when profiling is enabled This commit also adds a new -Dk RTS option to the debug RTS to assist debugging continuation captures. Currently, the printed information is quite minimal, but more can be added in the future if it proves to be useful when debugging future issues. fixes #23001 - - - - - e9e7a00d by sheaf at 2023-02-23T17:00:01-05:00 Explicit migration timeline for loopy SC solving This patch updates the warning message introduced in commit 9fb4ca89bff9873e5f6a6849fa22a349c94deaae to specify an explicit migration timeline: GHC will no longer support this constraint solving mechanism starting from GHC 9.10. Fixes #22912 - - - - - 4eb9c234 by Sylvain Henry at 2023-02-24T17:27:45-05:00 JS: make some arithmetic primops faster (#22835) Don't use BigInt for wordAdd2, mulWord32, and timesInt32. Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 92e76483 by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump terminfo submodule to 0.4.1.6 - - - - - f229db14 by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump unix submodule to 2.8.1.0 - - - - - 47bd48c1 by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump deepseq submodule to 1.4.8.1 - - - - - d2012594 by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump directory submodule to 1.3.8.1 - - - - - df6f70d1 by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump process submodule to v1.6.17.0 - - - - - 4c869e48 by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump hsc2hs submodule to 0.68.8 - - - - - 81d96642 by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump array submodule to 0.5.4.0 - - - - - 6361f771 by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump Cabal submodule to 3.9 pre-release - - - - - 4085fb6c by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump filepath submodule to 1.4.100.1 - - - - - 2bfad50f by Ben Gamari at 2023-02-24T17:28:20-05:00 Bump haskeline submodule to 0.8.2.1 - - - - - fdc89a8d by Ben Gamari at 2023-02-24T21:29:32-05:00 gitlab-ci: Run nix-build with -v0 This significantly cuts down on the amount of noise in the job log. Addresses #22861. - - - - - 69fb0b13 by Aaron Allen at 2023-02-24T21:30:10-05:00 Fix ParallelListComp out of scope suggestion This patch makes it so vars from one block of a parallel list comprehension are not in scope in a subsequent block during type checking. This was causing GHC to emit a faulty suggestion when an out of scope variable shared the occ name of a var from a different block. Fixes #22940 - - - - - ece092d0 by Simon Peyton Jones at 2023-02-24T21:30:45-05:00 Fix shadowing bug in prepareAlts As #23012 showed, GHC.Core.Opt.Simplify.Utils.prepareAlts was using an OutType to construct an InAlt. When shadowing is in play, this is outright wrong. See Note [Shadowing in prepareAlts]. - - - - - 7825fef9 by Sylvain Henry at 2023-02-24T21:31:25-05:00 JS: Store CI perf results (fix #22923) - - - - - b56025f4 by Gergő Érdi at 2023-02-27T13:34:22+00:00 Don't specialise incoherent instance applications Using incoherent instances, there can be situations where two occurrences of the same overloaded function at the same type use two different instances (see #22448). For incoherently resolved instances, we must mark them with `nospec` to avoid the specialiser rewriting one to the other. This marking is done during the desugaring of the `WpEvApp` wrapper. Fixes #22448 Metric Increase: T15304 - - - - - d0c7bbed by Tom Ellis at 2023-02-27T20:04:07-05:00 Fix SCC grouping example - - - - - f84a8cd4 by Bryan Richter at 2023-02-28T05:58:37-05:00 Mark setnumcapabilities001 fragile - - - - - 29a04d6e by Bryan Richter at 2023-02-28T05:58:37-05:00 Allow nightly-x86_64-linux-deb10-validate+thread_sanitizer to fail See #22520 - - - - - 9fa54572 by Cheng Shao at 2023-02-28T05:59:15-05:00 ghc-prim: fix hs_cmpxchg64 function prototype hs_cmpxchg64 must return a StgWord64, otherwise incorrect runtime results of 64-bit MO_Cmpxchg will appear in 32-bit unregisterised builds, which go unnoticed at compile-time due to C implicit casting in .hc files. - - - - - 0c200ab7 by Simon Peyton Jones at 2023-02-28T11:10:31-05:00 Account for local rules in specImports As #23024 showed, in GHC.Core.Opt.Specialise.specImports, we were generating specialisations (a locally-define function) for imported functions; and then generating specialisations for those locally-defined functions. The RULE for the latter should be attached to the local Id, not put in the rules-for-imported-ids set. Fix is easy; similar to what happens in GHC.HsToCore.addExportFlagsAndRules - - - - - 8b77f9bf by Sylvain Henry at 2023-02-28T11:11:21-05:00 JS: fix for overlap with copyMutableByteArray# (#23033) The code wasn't taking into account some kind of overlap. cgrun070 has been extended to test the missing case. - - - - - 239202a2 by Sylvain Henry at 2023-02-28T11:12:03-05:00 Testsuite: replace some js_skip with req_cmm req_cmm is more informative than js_skip - - - - - 7192ef91 by Simon Peyton Jones at 2023-02-28T18:54:59-05:00 Take more care with unlifted bindings in the specialiser As #22998 showed, we were floating an unlifted binding to top level, which breaks a Core invariant. The fix is easy, albeit a little bit conservative. See Note [Care with unlifted bindings] in GHC.Core.Opt.Specialise - - - - - bb500e2a by Simon Peyton Jones at 2023-02-28T18:55:35-05:00 Account for TYPE vs CONSTRAINT in mkSelCo As #23018 showed, in mkRuntimeRepCo we need to account for coercions between TYPE and COERCION. See Note [mkRuntimeRepCo] in GHC.Core.Coercion. - - - - - 79ffa170 by Ben Gamari at 2023-03-01T04:17:20-05:00 hadrian: Add dependency from lib/settings to mk/config.mk In 81975ef375de07a0ea5a69596b2077d7f5959182 we attempted to fix #20253 by adding logic to the bindist Makefile to regenerate the `settings` file from information gleaned by the bindist `configure` script. However, this fix had no effect as `lib/settings` is shipped in the binary distribution (to allow in-place use of the binary distribution). As `lib/settings` already existed and its rule declared no dependencies, `make` would fail to use the added rule to regenerate it. Fix this by explicitly declaring a dependency from `lib/settings` on `mk/config.mk`. Fixes #22982. - - - - - a2a1a1c0 by Sebastian Graf at 2023-03-01T04:17:56-05:00 Revert the main payload of "Make `drop` and `dropWhile` fuse (#18964)" This reverts the bits affecting fusion of `drop` and `dropWhile` of commit 0f7588b5df1fc7a58d8202761bf1501447e48914 and keeps just the small refactoring unifying `flipSeqTake` and `flipSeqScanl'` into `flipSeq`. It also adds a new test for #23021 (which was the reason for reverting) as well as adds a clarifying comment to T18964. Fixes #23021, unfixes #18964. Metric Increase: T18964 Metric Decrease: T18964 - - - - - cf118e2f by Simon Peyton Jones at 2023-03-01T04:18:33-05:00 Refine the test for naughty record selectors The test for naughtiness in record selectors is surprisingly subtle. See the revised Note [Naughty record selectors] in GHC.Tc.TyCl.Utils. Fixes #23038. - - - - - 86f240ca by romes at 2023-03-01T04:19:10-05:00 fix: Consider strictness annotation in rep_bind Fixes #23036 - - - - - 1ed573a5 by Richard Eisenberg at 2023-03-02T22:42:06-05:00 Don't suppress *all* Wanteds Code in GHC.Tc.Errors.reportWanteds suppresses a Wanted if its rewriters have unfilled coercion holes; see Note [Wanteds rewrite Wanteds] in GHC.Tc.Types.Constraint. But if we thereby suppress *all* errors that's really confusing, and as #22707 shows, GHC goes on without even realising that the program is broken. Disaster. This MR arranges to un-suppress them all if they all get suppressed. Close #22707 - - - - - 8919f341 by Luite Stegeman at 2023-03-02T22:42:45-05:00 Check for platform support for JavaScript foreign imports GHC was accepting `foreign import javascript` declarations on non-JavaScript platforms. This adds a check so that these are only supported on an platform that supports the JavaScript calling convention. Fixes #22774 - - - - - db83f8bb by Ben Gamari at 2023-03-02T22:43:22-05:00 rts: Statically assert alignment of Capability In #22965 we noticed that changes in the size of `Capability` can result in unsound behavior due to the `align` pragma claiming an alignment which we don't in practice observe. Avoid this by statically asserting that the size is a multiple of the alignment. - - - - - 5f7a4a6d by Ben Gamari at 2023-03-02T22:43:22-05:00 rts: Introduce stgMallocAlignedBytes - - - - - 8a6f745d by Ben Gamari at 2023-03-02T22:43:22-05:00 rts: Correctly align Capability allocations Previously we failed to tell the C allocator that `Capability`s needed to be aligned, resulting in #22965. Fixes #22965. Fixes #22975. - - - - - 5464c73f by Ben Gamari at 2023-03-02T22:43:22-05:00 rts: Drop no-alignment special case for Windows For reasons that aren't clear, we were previously not giving Capability the same favorable alignment on Windows that we provided on other platforms. Fix this. - - - - - a86aae8b by Matthew Pickering at 2023-03-02T22:43:59-05:00 constant folding: Correct type of decodeDouble_Int64 rule The first argument is Int64# unconditionally, so we better produce something of that type. This fixes a core lint error found in the ad package. Fixes #23019 - - - - - 68dd64ff by Zubin Duggal at 2023-03-02T22:44:35-05:00 ncg/aarch64: Handle MULTILINE_COMMENT identically as COMMENTs Commit 7566fd9de38c67360c090f828923d41587af519c with the fix for #22798 was incomplete as it failed to handle MULTILINE_COMMENT pseudo-instructions, and didn't completly fix the compiler panics when compiling with `-fregs-graph`. Fixes #23002 - - - - - 2f97c861 by Simon Peyton Jones at 2023-03-02T22:45:11-05:00 Get the right in-scope set in etaBodyForJoinPoint Fixes #23026 - - - - - 45af8482 by David Feuer at 2023-03-03T11:40:47-05:00 Export getSolo from Data.Tuple Proposed in [CLC proposal #113](https://github.com/haskell/core-libraries-committee/issues/113) and [approved by the CLC](https://github.com/haskell/core-libraries-committee/issues/113#issuecomment-1452452191) - - - - - 0c694895 by David Feuer at 2023-03-03T11:40:47-05:00 Document getSolo - - - - - bd0536af by Simon Peyton Jones at 2023-03-03T11:41:23-05:00 More fixes for `type data` declarations This MR fixes #23022 and #23023. Specifically * Beef up Note [Type data declarations] in GHC.Rename.Module, to make invariant (I1) explicit, and to name the several wrinkles. And add references to these specific wrinkles. * Add a Lint check for invariant (I1) above. See GHC.Core.Lint.checkTypeDataConOcc * Disable the `caseRules` for dataToTag# for `type data` values. See Wrinkle (W2c) in the Note above. Fixes #23023. * Refine the assertion in dataConRepArgTys, so that it does not complain about the absence of a wrapper for a `type data` constructor Fixes #23022. Acked-by: Simon Peyton Jones <simon.peytonjones at gmail.com> - - - - - 858f34d5 by Oleg Grenrus at 2023-03-04T01:13:55+02:00 Add decideSymbol, decideChar, decideNat, decTypeRep, decT and hdecT These all type-level equality decision procedures. Implementes a CLC proposal https://github.com/haskell/core-libraries-committee/issues/98 - - - - - bf43ba92 by Simon Peyton Jones at 2023-03-04T01:18:23-05:00 Add test for T22793 - - - - - c6e1f3cd by Chris Wendt at 2023-03-04T03:35:18-07:00 Fix typo in docs referring to threadLabel - - - - - 232cfc24 by Simon Peyton Jones at 2023-03-05T19:57:30-05:00 Add regression test for #22328 - - - - - 5ed77deb by Gabriella Gonzalez at 2023-03-06T17:06:50-05:00 Enable response files for linker if supported - - - - - 1e0f6c89 by Gabriella Gonzalez at 2023-03-06T17:06:50-05:00 Synchronize `configure.ac` and `distrib/configure.ac.in` - - - - - 70560952 by Gabriella Gonzalez at 2023-03-06T17:06:50-05:00 Fix `hadrian/bindist/config.mk.in` … as suggested by @bgamari - - - - - b042b125 by sheaf at 2023-03-06T17:06:50-05:00 Apply 1 suggestion(s) to 1 file(s) - - - - - 674b6b81 by Gabriella Gonzalez at 2023-03-06T17:06:50-05:00 Try to create somewhat portable `ld` command I cannot figure out a good way to generate an `ld` command that works on both Linux and macOS. Normally you'd use something like `AC_LINK_IFELSE` for this purpose (I think), but that won't let us test response file support. - - - - - 83b0177e by Gabriella Gonzalez at 2023-03-06T17:06:50-05:00 Quote variables … as suggested by @bgamari - - - - - 845f404d by Gabriella Gonzalez at 2023-03-06T17:06:50-05:00 Fix configure failure on alpine linux - - - - - c56a3ae6 by Gabriella Gonzalez at 2023-03-06T17:06:50-05:00 Small fixes to configure script - - - - - cad5c576 by Andrei Borzenkov at 2023-03-06T17:07:33-05:00 Convert diagnostics in GHC.Rename.Module to proper TcRnMessage (#20115) I've turned almost all occurrences of TcRnUnknownMessage in GHC.Rename.Module module into a proper TcRnMessage. Instead, these TcRnMessage messages were introduced: TcRnIllegalInstanceHeadDecl TcRnUnexpectedStandaloneDerivingDecl TcRnUnusedVariableInRuleDecl TcRnUnexpectedStandaloneKindSig TcRnIllegalRuleLhs TcRnBadAssocRhs TcRnDuplicateRoleAnnot TcRnDuplicateKindSig TcRnIllegalDerivStrategy TcRnIllegalMultipleDerivClauses TcRnNoDerivStratSpecified TcRnStupidThetaInGadt TcRnBadImplicitSplice TcRnShadowedTyVarNameInFamResult TcRnIncorrectTyVarOnLhsOfInjCond TcRnUnknownTyVarsOnRhsOfInjCond Was introduced one helper type: RuleLhsErrReason - - - - - c6432eac by Apoorv Ingle at 2023-03-06T23:26:12+00:00 Constraint simplification loop now depends on `ExpansionFuel` instead of a boolean flag for `CDictCan.cc_pend_sc`. Pending givens get a fuel of 3 while Wanted and quantified constraints get a fuel of 1. This helps pending given constraints to keep up with pending wanted constraints in case of `UndecidableSuperClasses` and superclass expansions while simplifying the infered type. Adds 3 dynamic flags for controlling the fuels for each type of constraints `-fgivens-expansion-fuel` for givens `-fwanteds-expansion-fuel` for wanteds and `-fqcs-expansion-fuel` for quantified constraints Fixes #21909 Added Tests T21909, T21909b Added Note [Expanding Recursive Superclasses and ExpansionFuel] - - - - - a5afc8ab by Andrew Lelechenko at 2023-03-06T22:51:01-05:00 Documentation: describe laziness of several function from Data.List - - - - - fa559c28 by Ollie Charles at 2023-03-07T20:56:21+00:00 Add `Data.Functor.unzip` This function is currently present in `Data.List.NonEmpty`, but `Data.Functor` is a better home for it. This change was discussed and approved by the CLC at https://github.com/haskell/core-libraries-committee/issues/88. - - - - - 2aa07708 by MorrowM at 2023-03-07T21:22:22-05:00 Fix documentation for traceWith and friends - - - - - f3ff7cb1 by David Binder at 2023-03-08T01:24:17-05:00 Remove utils/hpc subdirectory and its contents - - - - - cf98e286 by David Binder at 2023-03-08T01:24:17-05:00 Add git submodule for utils/hpc - - - - - 605fbbb2 by David Binder at 2023-03-08T01:24:18-05:00 Update commit for utils/hpc git submodule - - - - - 606793d4 by David Binder at 2023-03-08T01:24:18-05:00 Update commit for utils/hpc git submodule - - - - - 4158722a by Sylvain Henry at 2023-03-08T01:24:58-05:00 linker: fix linking with aligned sections (#23066) Take section alignment into account instead of assuming 16 bytes (which is wrong when the section requires 32 bytes, cf #23066). - - - - - 1e0d8fdb by Greg Steuck at 2023-03-08T08:59:05-05:00 Change hostSupportsRPaths to report False on OpenBSD OpenBSD does support -rpath but ghc build process relies on some related features that don't work there. See ghc/ghc#23011 - - - - - bed3a292 by Alexis King at 2023-03-08T08:59:53-05:00 bytecode: Fix bitmaps for BCOs used to tag tuples and prim call args fixes #23068 - - - - - 321d46d9 by Ben Gamari at 2023-03-08T15:02:30-05:00 rts: Drop redundant prototype - - - - - abb6070f by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Fix style - - - - - be278901 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Deduplicate assertion - - - - - b9034639 by Ben Gamari at 2023-03-08T15:02:30-05:00 rts: Fix type issues in Sparks.h Adds explicit casts to satisfy a C++ compiler. - - - - - da7b2b94 by Ben Gamari at 2023-03-08T15:02:30-05:00 rts: Use release ordering when storing thread labels Since this makes the ByteArray# visible from other cores. - - - - - 5b7f6576 by Ben Gamari at 2023-03-08T15:02:30-05:00 rts/BlockAlloc: Allow disabling of internal assertions These can be quite expensive and it is sometimes useful to compile a DEBUG RTS without them. - - - - - 6283144f by Ben Gamari at 2023-03-08T15:02:30-05:00 rts/Sanity: Mark pinned_object_blocks - - - - - 9b528404 by Ben Gamari at 2023-03-08T15:02:30-05:00 rts/Sanity: Look at nonmoving saved_filled lists - - - - - 0edc5438 by Ben Gamari at 2023-03-08T15:02:30-05:00 Evac: Squash data race in eval_selector_chain - - - - - 7eab831a by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Clarify implementation This makes the intent of this implementation a bit clearer. - - - - - 532262b9 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Clarify comment - - - - - bd9cd84b by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Add missing no-op in busy-wait loop - - - - - c4e6bfc8 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Don't push empty arrays to update remembered set Previously the write barrier of resizeSmallArray# incorrectly handled resizing of zero-sized arrays, pushing an invalid pointer to the update remembered set. Fixes #22931. - - - - - 92227b60 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Fix handling of weak pointers This fixes an interaction between aging and weak pointer handling which prevented the finalization of some weak pointers. In particular, weak pointers could have their keys incorrectly marked by the preparatory collector, preventing their finalization by the subsequent concurrent collection. While in the area, we also significantly improve the assertions regarding weak pointers. Fixes #22327. - - - - - ba7e7972 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Sanity check nonmoving large objects and compacts - - - - - 71b038a1 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Sanity check mutable list Assert that entries in the nonmoving generation's generational remembered set (a.k.a. mutable list) live in nonmoving generation. - - - - - 99d144d5 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Don't show occupancy if we didn't collect live words - - - - - 81d6cc55 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Fix tracking of FILLED_SWEEPING segments Previously we only updated the state of the segment at the head of each allocator's filled list. - - - - - 58e53bc4 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Assert state of swept segments - - - - - 2db92e01 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Handle new closures in nonmovingIsNowAlive We must conservatively assume that new closures are reachable since we are not guaranteed to mark such blocks. - - - - - e4c3249f by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Don't clobber update rem sets of old capabilities Previously `storageAddCapabilities` (called by `setNumCapabilities`) would clobber the update remembered sets of existing capabilities when increasing the capability count. Fix this by only initializing the update remembered sets of the newly-created capabilities. Fixes #22927. - - - - - 1b069671 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Add missing write barriers in selector optimisation This fixes the selector optimisation, adding a few write barriers which are necessary for soundness. See the inline comments for details. Fixes #22930. - - - - - d4032690 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Post-sweep sanity checking - - - - - 0baa8752 by Ben Gamari at 2023-03-08T15:02:30-05:00 nonmoving: Avoid n_caps race - - - - - 5d3232ba by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Don't push if nonmoving collector isn't enabled - - - - - 0a7eb0aa by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Be more paranoid in segment tracking Previously we left various segment link pointers dangling. None of this wrong per se, but it did make it harder than necessary to debug. - - - - - 7c817c0a by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Sync-phase mark budgeting Here we significantly improve the bound on sync phase pause times by imposing a limit on the amount of work that we can perform during the sync. If we find that we have exceeded our marking budget then we allow the mutators to resume, return to concurrent marking, and try synchronizing again later. Fixes #22929. - - - - - ce22a3e2 by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Allow pinned gen0 objects to be WEAK keys - - - - - 78746906 by Ben Gamari at 2023-03-08T15:02:31-05:00 rts: Reenable assertion - - - - - b500867a by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Move current segment array into Capability The current segments are conceptually owned by the mutator, not the collector. Consequently, it was quite tricky to prove that the mutator would not race with the collect due to this shared state. It turns out that such races are possible: when resizing the current segment array we may concurrently try to take a heap census. This will attempt to walk the current segment array, causing a data race. Fix this by moving the current segment array into `Capability`, where it belongs. Fixes #22926. - - - - - 56e669c1 by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Fix Note references Some references to Note [Deadlock detection under the non-moving collector] were missing an article. - - - - - 4a7650d7 by Ben Gamari at 2023-03-08T15:02:31-05:00 rts/Sanity: Fix block count assertion with non-moving collector The nonmoving collector does not use `oldest_gen->blocks` to track its block list. However, it nevertheless updates `oldest_gen->n_blocks` to ensure that its size is accounted for by the storage manager. Consequently, we must not attempt to assert consistency between the two. - - - - - 96a5aaed by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Don't call prepareUnloadCheck When the nonmoving GC is in use we do not call `checkUnload` (since we don't unload code) and therefore should not call `prepareUnloadCheck`, lest we run into assertions. - - - - - 6c6674ca by Ben Gamari at 2023-03-08T15:02:31-05:00 rts: Encapsulate block allocator spinlock This makes it a bit easier to add instrumentation on this spinlock while debugging. - - - - - e84f7167 by Ben Gamari at 2023-03-08T15:02:31-05:00 testsuite: Skip some tests when sanity checking is enabled - - - - - 3ae0f368 by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Fix unregisterised build - - - - - 4eb9d06b by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Ensure that sanity checker accounts for saved_filled segments - - - - - f0cf384d by Ben Gamari at 2023-03-08T15:02:31-05:00 hadrian: Add +boot_nonmoving_gc flavour transformer For using GHC bootstrapping to validate the non-moving GC. - - - - - 581e58ac by Ben Gamari at 2023-03-08T15:02:31-05:00 gitlab-ci: Add job bootstrapping with nonmoving GC - - - - - 487a8b58 by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Move allocator into new source file - - - - - 8f374139 by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Split out nonmovingAllocateGC - - - - - 662b6166 by Ben Gamari at 2023-03-08T15:02:31-05:00 testsuite: Only run T22795* in the normal way It doesn't make sense to run these in multiple ways as they merely test whether `-threaded`/`-single-threaded` flags. - - - - - 0af21dfa by Ben Gamari at 2023-03-08T15:02:31-05:00 rts: Rename clear_segment(_free_blocks)? To reflect the fact that these are to do with the nonmoving collector, now since they are exposed no longer static. - - - - - 7bcb192b by Ben Gamari at 2023-03-08T15:02:31-05:00 rts: Fix incorrect STATIC_INLINE This should be INLINE_HEADER lest we get unused declaration warnings. - - - - - f1fd3ffb by Ben Gamari at 2023-03-08T15:02:31-05:00 testsuite: Mark ffi023 as broken due to #23089 - - - - - a57f12b3 by Ben Gamari at 2023-03-08T15:02:31-05:00 testsuite: Skip T7160 in the nonmoving way Finalization order is different under the nonmoving collector. - - - - - f6f12a36 by Ben Gamari at 2023-03-08T15:02:31-05:00 rts: Capture GC configuration in a struct The number of distinct arguments passed to GarbageCollect was getting a bit out of hand. - - - - - ba73a807 by Ben Gamari at 2023-03-08T15:02:31-05:00 nonmoving: Non-concurrent collection - - - - - 7c813d06 by Alexis King at 2023-03-08T15:03:10-05:00 hadrian: Fix flavour compiler stage options off-by-one error !9193 pointed out that ghcDebugAssertions was supposed to be a predicate on the stage of the built compiler, but in practice it was a predicate on the stage of the compiler used to build. Unfortunately, while it fixed that issue for ghcDebugAssertions, it documented every other similar option as behaving the same way when in fact they all used the old behavior. The new behavior of ghcDebugAssertions seems more intuitive, so this commit changes the interpretation of every other option to match. It also improves the enableProfiledGhc and debugGhc flavour transformers by making them more selective about which stages in which they build additional library/RTS ways. - - - - - f97c7f6d by Luite Stegeman at 2023-03-09T09:52:09-05:00 Delete created temporary subdirectories at end of session. This patch adds temporary subdirectories to the list of paths do clean up at the end of the GHC session. This fixes warnings about non-empty temporary directories. Fixes #22952 - - - - - 9ea719f2 by Apoorv Ingle at 2023-03-09T09:52:45-05:00 Fixes #19627. Previously the solver failed with an unhelpful "solver reached too may iterations" error. With the fix for #21909 in place we no longer have the possibility of generating such an error if we have `-fconstraint-solver-iteration` > `-fgivens-fuel > `-fwanteds-fuel`. This is true by default, and the said fix also gives programmers a knob to control how hard the solver should try before giving up. This commit adds: * Reference to ticket #19627 in the Note [Expanding Recursive Superclasses and ExpansionFuel] * Test `typecheck/should_fail/T19627.hs` for regression purposes - - - - - eb3968b5 by Ben Gamari at 2023-03-10T02:32:43-05:00 Bump versions for ghc-9.6 release - - - - - ec2d93eb by Sebastian Graf at 2023-03-10T10:18:54-05:00 DmdAnal: Fix a panic on OPAQUE and trivial/PAP RHS (#22997) We should not panic in `add_demands` (now `set_lam_dmds`), because that code path is legimitely taken for OPAQUE PAP bindings, as in T22997. Fixes #22997. - - - - - 5b4628ae by Sylvain Henry at 2023-03-10T10:19:34-05:00 JS: remove dead code for old integer-gmp - - - - - bab23279 by Josh Meredith at 2023-03-10T23:24:49-05:00 JS: Fix implementation of MK_JSVAL - - - - - ec263a59 by Sebastian Graf at 2023-03-10T23:25:25-05:00 Simplify: Move `wantEtaExpansion` before expensive `do_eta_expand` check There is no need to run arity analysis and what not if we are not in a Simplifier phase that eta-expands or if we don't want to eta-expand the expression in the first place. Purely a refactoring with the goal of improving compiler perf. - - - - - 047e9d4f by Josh Meredith at 2023-03-13T03:56:03+00:00 JS: fix implementation of forceBool to use JS backend syntax - - - - - 559a4804 by Sebastian Graf at 2023-03-13T07:31:23-04:00 Simplifier: `countValArgs` should not count Type args (#23102) I observed miscompilations while working on !10088 caused by this. Fixes #23102. Metric Decrease: T10421 - - - - - 536d1f90 by Matthew Pickering at 2023-03-13T14:04:49+00:00 Bump Win32 to 2.13.4.0 Updates Win32 submodule - - - - - ee17001e by Ben Gamari at 2023-03-13T21:18:24-04:00 ghc-bignum: Drop redundant include-dirs field - - - - - c9c26cd6 by Teo Camarasu at 2023-03-16T12:17:50-04:00 Fix BCO creation setting caps when -j > -N * Remove calls to 'setNumCapabilities' in 'createBCOs' These calls exist to ensure that 'createBCOs' can benefit from parallelism. But this is not the right place to call `setNumCapabilities`. Furthermore the logic differs from that in the driver causing the capability count to be raised and lowered at each TH call if -j > -N. * Remove 'BCOOpts' No longer needed as it was only used to thread the job count down to `createBCOs` Resolves #23049 - - - - - 5ddbf5ed by Teo Camarasu at 2023-03-16T12:17:50-04:00 Add changelog entry for #23049 - - - - - 6e3ce9a4 by Ben Gamari at 2023-03-16T12:18:26-04:00 configure: Fix FIND_CXX_STD_LIB test on Darwin Annoyingly, Darwin's <cstddef> includes <version> and APFS is case-insensitive. Consequently, it will end up #including the `VERSION` file generated by the `configure` script on the second and subsequent runs of the `configure` script. See #23116. - - - - - 19d6d039 by sheaf at 2023-03-16T21:31:22+01:00 ghci: only keep the GlobalRdrEnv in ModInfo The datatype GHC.UI.Info.ModInfo used to store a ModuleInfo, which includes a TypeEnv. This can easily cause space leaks as we have no way of forcing everything in a type environment. In GHC, we only use the GlobalRdrEnv, which we can force completely. So we only store that instead of a fully-fledged ModuleInfo. - - - - - 73d07c6e by Torsten Schmits at 2023-03-17T14:36:49-04:00 Add structured error messages for GHC.Tc.Utils.Backpack Tracking ticket: #20119 MR: !10127 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. One occurrence, when handing a nested error from the interface loading machinery, was omitted. It will be handled by a subsequent changeset that addresses interface errors. - - - - - a13affce by Andrei Borzenkov at 2023-03-21T11:17:17-04:00 Rename () into Unit, (,,...,,) into Tuple<n> (#21294) This patch implements a part of GHC Proposal #475. The key change is in GHC.Tuple.Prim: - data () = () - data (a,b) = (a,b) - data (a,b,c) = (a,b,c) ... + data Unit = () + data Tuple2 a b = (a,b) + data Tuple3 a b c = (a,b,c) ... And the rest of the patch makes sure that Unit and Tuple<n> are pretty-printed as () and (,,...,,) in various contexts. Updates the haddock submodule. Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 23642bf6 by Adam Sandberg Ericsson at 2023-03-21T11:17:53-04:00 docs: fix some wrongs in the eventlog format documentation - - - - - 90159773 by Adam Sandberg Ericsson at 2023-03-21T11:17:53-04:00 docs: explain the BLOCK_MARKER event - - - - - ab1c25e8 by Adam Sandberg Ericsson at 2023-03-21T11:17:53-04:00 docs: add BlockedOnMVarRead thread status in eventlog encodings - - - - - 898afaef by Adam Sandberg Ericsson at 2023-03-21T11:17:53-04:00 docs: add TASK_DELETE event in eventlog encodings - - - - - bb05b4cc by Adam Sandberg Ericsson at 2023-03-21T11:17:53-04:00 docs: add WALL_CLOCK_TIME event in eventlog encodings - - - - - eeea0343 by Torsten Schmits at 2023-03-21T11:18:34-04:00 Add structured error messages for GHC.Tc.Utils.Env Tracking ticket: #20119 MR: !10129 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. - - - - - be1d4be8 by Andrew Lelechenko at 2023-03-21T11:19:13-04:00 Document pdep / pext primops - - - - - e8b4aac4 by Alex Mason at 2023-03-21T18:11:04-04:00 Allow LLVM backend to use HDoc for faster file generation. Also remove the MetaStmt constructor from LlvmStatement and places the annotations into the Store statement. Includes “Implement a workaround for -no-asm-shortcutting bug“ (https://gitlab.haskell.org/ghc/ghc/-/commit/2fda9e0df886cc551e2cd6b9c2a384192bdc3045) - - - - - ea24360d by Luite Stegeman at 2023-03-21T18:11:44-04:00 Compute LambdaFormInfo when using JavaScript backend. CmmCgInfos is needed to write interface files, but the JavaScript backend does not generate it, causing "Name without LFInfo" warnings. This patch adds a conservative but always correct CmmCgInfos when the JavaScript backend is used. Fixes #23053 - - - - - 926ad6de by Simon Peyton Jones at 2023-03-22T01:03:08-04:00 Be more careful about quantification This MR is driven by #23051. It does several things: * It is guided by the generalisation plan described in #20686. But it is still far from a complete implementation of that plan. * Add Note [Inferred type with escaping kind] to GHC.Tc.Gen.Bind. This explains that we don't (yet, pending #20686) directly prevent generalising over escaping kinds. * In `GHC.Tc.Utils.TcMType.defaultTyVar` we default RuntimeRep and Multiplicity variables, beause we don't want to quantify over them. We want to do the same for a Concrete tyvar, but there is nothing sensible to default it to (unless it has kind RuntimeRep, in which case it'll be caught by an earlier case). So we promote instead. * Pure refactoring in GHC.Tc.Solver: * Rename decideMonoTyVars to decidePromotedTyVars, since that's what it does. * Move the actual promotion of the tyvars-to-promote from `defaultTyVarsAndSimplify` to `decidePromotedTyVars`. This is a no-op; just tidies up the code. E.g then we don't need to return the promoted tyvars from `decidePromotedTyVars`. * A little refactoring in `defaultTyVarsAndSimplify`, but no change in behaviour. * When making a TauTv unification variable into a ConcreteTv (in GHC.Tc.Utils.Concrete.makeTypeConcrete), preserve the occ-name of the type variable. This just improves error messages. * Kill off dead code: GHC.Tc.Utils.TcMType.newConcreteHole - - - - - 0ab0cc11 by Sylvain Henry at 2023-03-22T01:03:48-04:00 Testsuite: use appropriate predicate for ManyUbxSums test (#22576) - - - - - 048c881e by romes at 2023-03-22T01:04:24-04:00 fix: Incorrect @since annotations in GHC.TypeError Fixes #23128 - - - - - a1528b68 by Sylvain Henry at 2023-03-22T01:05:04-04:00 Testsuite: use req_interp predicate for T16318 (#22370) - - - - - ad765b6f by Sylvain Henry at 2023-03-22T01:05:04-04:00 Testsuite: use req_interp predicate for T20214 - - - - - e0b8eaf3 by Simon Peyton Jones at 2023-03-22T09:50:13+00:00 Refactor the constraint solver pipeline The big change is to put the entire type-equality solver into GHC.Tc.Solver.Equality, rather than scattering it over Canonical and Interact. Other changes * EqCt becomes its own data type, a bit like QCInst. This is great because EqualCtList is then just [EqCt] * New module GHC.Tc.Solver.Dict has come of the class-contraint solver. In due course it will be all. One step at a time. This MR is intended to have zero change in behaviour: it is a pure refactor. It opens the way to subsequent tidying up, we believe. - - - - - cedf9a3b by Torsten Schmits at 2023-03-22T15:31:18-04:00 Add structured error messages for GHC.Tc.Utils.TcMType Tracking ticket: #20119 MR: !10138 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. - - - - - 30d45e97 by Sylvain Henry at 2023-03-22T15:32:01-04:00 Testsuite: use js_skip for T2615 (#22374) - - - - - 4aeead36 by Adam Gundry at 2023-03-23T13:53:47+01:00 Adapt to warning categories changes - - - - - 8c98deba by Armando Ramirez at 2023-03-23T09:19:32-04:00 Optimized Foldable methods for Data.Functor.Compose Explicitly define length, elem, etc. in Foldable instance for Data.Functor.Compose Implementation of https://github.com/haskell/core-libraries-committee/issues/57 - - - - - bc066108 by Armando Ramirez at 2023-03-23T09:19:32-04:00 Additional optimized versions - - - - - 80fce576 by Andrew Lelechenko at 2023-03-23T09:19:32-04:00 Simplify minimum/maximum in instance Foldable (Compose f g) - - - - - 8cb88a5a by Andrew Lelechenko at 2023-03-23T09:19:32-04:00 Update changelog to mention changes to instance Foldable (Compose f g) - - - - - e1c8c41d by Torsten Schmits at 2023-03-23T09:20:13-04:00 Add structured error messages for GHC.Tc.TyCl.PatSyn Tracking ticket: #20117 MR: !10158 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. - - - - - f932c589 by Adam Gundry at 2023-03-24T02:36:09-04:00 Allow WARNING pragmas to be controlled with custom categories Closes #17209. This implements GHC Proposal 541, allowing a WARNING pragma to be annotated with a category like so: {-# WARNING in "x-partial" head "This function is undefined on empty lists." #-} The user can then enable, disable and set the severity of such warnings using command-line flags `-Wx-partial`, `-Werror=x-partial` and so on. There is a new warning group `-Wextended-warnings` containing all these warnings. Warnings without a category are treated as if the category was `deprecations`, and are (still) controlled by the flags `-Wdeprecations` and `-Wwarnings-deprecations`. Updates Haddock submodule. - - - - - 0426515b by Adam Gundry at 2023-03-24T02:36:09-04:00 Move mention of warning groups change to 9.8.1 release notes - - - - - b8d783d2 by Ben Gamari at 2023-03-24T02:36:45-04:00 nativeGen/AArch64: Fix bitmask immediate predicate Previously the predicate for determining whether a logical instruction operand could be encoded as a bitmask immediate was far too conservative. This meant that, e.g., pointer untagged required five instructions whereas it should only require one. Fixes #23030. - - - - - 46120bb6 by Joachim Breitner at 2023-03-24T13:09:43-04:00 User's guide: Improve docs for -Wall previously it would list the warnings _not_ enabled by -Wall. That’s unnecessary round-about and was out of date. So let's just name the relevant warnings (based on `compiler/GHC/Driver/Flags.hs`). - - - - - 509d1f11 by Ben Gamari at 2023-03-24T13:10:20-04:00 codeGen/tsan: Disable instrumentation of unaligned stores There is some disagreement regarding the prototype of `__tsan_unaligned_write` (specifically whether it takes just the written address, or the address and the value as an argument). Moreover, I have observed crashes which appear to be due to it. Disable instrumentation of unaligned stores as a temporary mitigation. Fixes #23096. - - - - - 6a73655f by Li-yao Xia at 2023-03-25T00:02:44-04:00 base: Document GHC versions associated with past base versions in the changelog - - - - - 43bd7694 by Teo Camarasu at 2023-03-25T00:03:24-04:00 Add regression test for #17574 This test currently fails in the nonmoving way - - - - - f2d56bf7 by Teo Camarasu at 2023-03-25T00:03:24-04:00 fix: account for large and compact object stats with nonmoving gc Make sure that we keep track of the size of large and compact objects that have been moved onto the nonmoving heap. We keep track of their size and add it to the amount of live bytes in nonmoving segments to get the total size of the live nonmoving heap. Resolves #17574 - - - - - 7131b705 by David Feuer at 2023-03-25T00:04:04-04:00 Modify ThreadId documentation and comments For a long time, `GHC.Conc.Sync` has said ```haskell -- ToDo: data ThreadId = ThreadId (Weak ThreadId#) -- But since ThreadId# is unlifted, the Weak type must use open -- type variables. ``` We are now actually capable of using `Weak# ThreadId#`, but the world has moved on. To support the `Show` and `Ord` instances, we'd need to store the thread ID number in the `ThreadId`. And it seems very difficult to continue to support `threadStatus` in that regime, since it needs to be able to explain how threads died. In addition, garbage collection of weak references can be quite expensive, and it would be hard to evaluate the cost over he whole ecosystem. As discussed in [this CLC issue](https://github.com/haskell/core-libraries-committee/issues/125), it doesn't seem very likely that we'll actually switch to weak references here. - - - - - c421bbbb by Ben Gamari at 2023-03-25T00:04:41-04:00 rts: Fix barriers of IND and IND_STATIC Previously IND and IND_STATIC lacked the acquire barriers enjoyed by BLACKHOLE. As noted in the (now updated) Note [Heap memory barriers], this barrier is critical to ensure that the indirectee is visible to the entering core. Fixes #22872. - - - - - 62fa7faa by Andrew Lelechenko at 2023-03-25T00:05:22-04:00 Improve documentation of atomicModifyMutVar2# - - - - - b2d14d0b by Cheng Shao at 2023-03-25T03:46:43-04:00 rts: use performBlockingMajorGC in hs_perform_gc and fix ffi023 This patch does a few things: - Add the missing RtsSymbols.c entry of performBlockingMajorGC - Make hs_perform_gc call performBlockingMajorGC, which restores previous behavior - Use hs_perform_gc in ffi023 - Remove rts_clearMemory() call in ffi023, it now works again in some test ways previously marked as broken. Fixes #23089 - - - - - d9ae24ad by Cheng Shao at 2023-03-25T03:46:44-04:00 testsuite: add the rts_clearMemory test case This patch adds a standalone test case for rts_clearMemory that mimics how it's typically used by wasm backend users and ensures this RTS API isn't broken by future RTS refactorings. Fixes #23901. - - - - - 80729d96 by Andrew Lelechenko at 2023-03-25T03:47:22-04:00 Improve documentation for resizing of byte arrays - - - - - c6ec4cd1 by Ben Gamari at 2023-03-25T20:23:47-04:00 rts: Don't rely on EXTERN_INLINE for slop-zeroing logic Previously we relied on calling EXTERN_INLINE functions defined in ClosureMacros.h from Cmm to zero slop. However, as far as I can tell, this is no longer safe to do in C99 as EXTERN_INLINE definitions may be emitted in each compilation unit. Fix this by explicitly declaring a new set of non-inline functions in ZeroSlop.c which can be called from Cmm and marking the ClosureMacros.h definitions as INLINE_HEADER. In the future we should try to eliminate EXTERN_INLINE. - - - - - c32abd4b by Ben Gamari at 2023-03-25T20:23:48-04:00 rts: Fix capability-count check in zeroSlop Previously `zeroSlop` examined `RtsFlags` to determine whether the program was single-threaded. This is wrong; a program may be started with `+RTS -N1` yet the process may later increase the capability count with `setNumCapabilities`. This lead to quite subtle and rare crashes. Fixes #23088. - - - - - 656d4cb3 by Ryan Scott at 2023-03-25T20:24:23-04:00 Add Eq/Ord instances for SSymbol, SChar, and SNat This implements [CLC proposal #148](https://github.com/haskell/core-libraries-committee/issues/148). - - - - - 4f93de88 by David Feuer at 2023-03-26T15:33:02-04:00 Update and expand atomic modification Haddocks * The documentation for `atomicModifyIORef` and `atomicModifyIORef'` were incomplete, and the documentation for `atomicModifyIORef` was out of date. Update and expand. * Remove a useless lazy pattern match in the definition of `atomicModifyIORef`. The pair it claims to match lazily was already forced by `atomicModifyIORef2`. - - - - - e1fb56b2 by David Feuer at 2023-03-26T15:33:41-04:00 Document the constructor name for lists Derived `Data` instances use raw infix constructor names when applicable. The `Data.Data [a]` instance, if derived, would have a constructor name of `":"`. However, it actually uses constructor name `"(:)"`. Document this peculiarity. See https://github.com/haskell/core-libraries-committee/issues/147 - - - - - c1f755c4 by Simon Peyton Jones at 2023-03-27T22:09:41+01:00 Make exprIsConApp_maybe a bit cleverer Addresses #23159. See Note Note [Exploit occ-info in exprIsConApp_maybe] in GHC.Core.SimpleOpt. Compile times go down very slightly, but always go down, never up. Good! Metrics: compile_time/bytes allocated ------------------------------------------------ CoOpt_Singletons(normal) -1.8% T15703(normal) -1.2% GOOD geo. mean -0.1% minimum -1.8% maximum +0.0% Metric Decrease: CoOpt_Singletons T15703 - - - - - 76bb4c58 by Ryan Scott at 2023-03-28T08:12:08-04:00 Add COMPLETE pragmas to TypeRep, SSymbol, SChar, and SNat This implements [CLC proposal #149](https://github.com/haskell/core-libraries-committee/issues/149). - - - - - 642d8d60 by sheaf at 2023-03-29T13:35:56+02:00 Adapt to record field refactor This commit adapts to the changes in GHC MR !8686, which overhauls the treatment of record fields in the renamer, adding separate record field namespaces and entirely removing the GreName datatype. - - - - - 3f374399 by sheaf at 2023-03-29T13:57:33+02:00 Handle records in the renamer This patch moves the field-based logic for disambiguating record updates to the renamer. The type-directed logic, scheduled for removal, remains in the typechecker. To do this properly (and fix the myriad of bugs surrounding the treatment of duplicate record fields), we took the following main steps: 1. Create GREInfo, a renamer-level equivalent to TyThing which stores information pertinent to the renamer. This allows us to uniformly treat imported and local Names in the renamer, as described in Note [GREInfo]. 2. Remove GreName. Instead of a GlobalRdrElt storing GreNames, which distinguished between normal names and field names, we now store simple Names in GlobalRdrElt, along with the new GREInfo information which allows us to recover the FieldLabel for record fields. 3. Add namespacing for record fields, within the OccNames themselves. This allows us to remove the mangling of duplicate field selectors. This change ensures we don't print mangled names to the user in error messages, and allows us to handle duplicate record fields in Template Haskell. 4. Move record disambiguation to the renamer, and operate on the level of data constructors instead, to handle #21443. The error message text for ambiguous record updates has also been changed to reflect that type-directed disambiguation is on the way out. (3) means that OccEnv is now a bit more complex: we first key on the textual name, which gives an inner map keyed on NameSpace: OccEnv a ~ FastStringEnv (UniqFM NameSpace a) Note that this change, along with (2), both increase the memory residency of GlobalRdrEnv = OccEnv [GlobalRdrElt], which causes a few tests to regress somewhat in compile-time allocation. Even though (3) simplified a lot of code (in particular the treatment of field selectors within Template Haskell and in error messages), it came with one important wrinkle: in the situation of -- M.hs-boot module M where { data A; foo :: A -> Int } -- M.hs module M where { data A = MkA { foo :: Int } } we have that M.hs-boot exports a variable foo, which is supposed to match with the record field foo that M exports. To solve this issue, we add a new impedance-matching binding to M foo{var} = foo{fld} This mimics the logic that existed already for impedance-binding DFunIds, but getting it right was a bit tricky. See Note [Record field impedance matching] in GHC.Tc.Module. We also needed to be careful to avoid introducing space leaks in GHCi. So we dehydrate the GlobalRdrEnv before storing it anywhere, e.g. in ModIface. This means stubbing out all the GREInfo fields, with the function forceGlobalRdrEnv. When we read it back in, we rehydrate with rehydrateGlobalRdrEnv. This robustly avoids any space leaks caused by retaining old type environments. Fixes #13352 #14848 #17381 #17551 #19664 #21443 #21444 #21720 #21898 #21946 #21959 #22125 #22160 #23010 #23062 #23063 Updates haddock submodule ------------------------- Metric Increase: MultiComponentModules MultiLayerModules MultiLayerModulesDefsGhci MultiLayerModulesNoCode T13701 T14697 hard_hole_fits ------------------------- - - - - - 4f1940f0 by sheaf at 2023-03-29T13:57:33+02:00 Avoid repeatedly shadowing in shadowNames This commit refactors GHC.Type.Name.Reader.shadowNames to first accumulate all the shadowing arising from the introduction of a new set of GREs, and then applies all the shadowing to the old GlobalRdrEnv in one go. - - - - - d246049c by sheaf at 2023-03-29T13:57:34+02:00 igre_prompt_env: discard "only-qualified" names We were unnecessarily carrying around names only available qualified in igre_prompt_env, violating the icReaderEnv invariant. We now get rid of these, as they aren't needed for the shadowing computation that igre_prompt_env exists for. Fixes #23177 ------------------------- Metric Decrease: T14052 T14052Type ------------------------- - - - - - ac8d4333 by doyougnu at 2023-03-29T11:11:44-04:00 Update UniqMap API - - - - - 41a572f6 by Matthew Pickering at 2023-03-29T16:17:21-04:00 hadrian: Fix path to HpcParser.y The source for this project has been moved into a src/ folder so we also need to update this path. Fixes #23187 - - - - - b159e0e9 by doyougnu at 2023-03-30T01:40:08-04:00 js: split JMacro into JS eDSL and JS syntax This commit: Splits JExpr and JStat into two nearly identical DSLs: - GHC.JS.Syntax is the JMacro based DSL without unsaturation, i.e., a value cannot be unsaturated, or, a value of this DSL is a witness that a value of GHC.JS.Unsat has been saturated - GHC.JS.Unsat is the JMacro DSL from GHCJS with Unsaturation. Then all binary and outputable instances are changed to use GHC.JS.Syntax. This moves us closer to closing out #22736 and #22352. See #22736 for roadmap. ------------------------- Metric Increase: CoOpt_Read LargeRecord ManyAlternatives PmSeriesS PmSeriesT PmSeriesV T10421 T10858 T11195 T11374 T11822 T12227 T12707 T13035 T13253 T13253-spj T13379 T14683 T15164 T15703 T16577 T17096 T17516 T17836 T18140 T18282 T18304 T18478 T18698a T18698b T18923 T1969 T19695 T20049 T21839c T3064 T4801 T5321FD T5321Fun T5631 T5642 T783 T9198 T9233 T9630 TcPlugin_RewritePerf WWRec ------------------------- - - - - - f4f1f14f by Sylvain Henry at 2023-03-30T01:40:49-04:00 ghc-heap: remove wrong Addr# coercion (#23181) Conversion from Addr# to I# isn't correct with the JS backend. Also used the opportunity to reenable 64-bit Word/Int tests - - - - - a5360490 by Ben Gamari at 2023-03-30T01:41:25-04:00 testsuite: Fix racing prints in T21465 As noted in #23155, we previously failed to add flushes necessary to ensure predictable output. Fixes #23155. - - - - - 98b5cf67 by Matthew Pickering at 2023-03-30T09:58:40+01:00 Revert "ghc-heap: remove wrong Addr# coercion (#23181)" This reverts commit f4f1f14f8009c3c120b8b963ec130cbbc774ec02. This fails to build with GHC-9.2 as a boot compiler. See #23195 for tracking this issue. - - - - - 61a2dfaa by Andrew Lelechenko at 2023-03-30T14:35:57-04:00 Add {-# WARNING #-} to Data.List.{head,tail} - - - - - 8f15c47c by Andrew Lelechenko at 2023-03-30T14:35:57-04:00 Fixes to accomodate Data.List.{head,tail} with {-# WARNING #-} - - - - - 7c7dbade by Andrew Lelechenko at 2023-03-30T14:35:57-04:00 Bump submodules - - - - - d2d8251b by Andrew Lelechenko at 2023-03-30T14:35:57-04:00 Fix tests - - - - - 3d38dcb6 by sheaf at 2023-03-30T14:35:57-04:00 Proxies for head and tail: review suggestions - - - - - 930edcfd by sheaf at 2023-03-30T14:36:33-04:00 docs: move RecordUpd changelog entry to 9.8 This was accidentally included in the 9.6 changelog instead of the 9.6 changelog. - - - - - 6f885e65 by sheaf at 2023-03-30T14:37:09-04:00 Add LANGUAGE GADTs to GHC.Rename.Env We need to enable this extension for the file to compile with ghc 9.2, as we are pattern matching on a GADT and this required the GADT extension to be enabled until 9.4. - - - - - 6d6a37a8 by Cheng Shao at 2023-03-30T18:42:56+00:00 ci: make lint-ci-config job fast again We don't pin our nixpkgs revision and tracks the default nixpkgs-unstable channel anyway. Instead of using haskell.packages.ghc924, we should be using haskell.packages.ghc92 to maximize the binary cache hit rate and make lint-ci-config job fast again. Also bumps the nix docker image to the latest revision. - - - - - ef1548c4 by Cheng Shao at 2023-03-30T18:42:56+00:00 ci: ensure that all non-i386 pipelines do parallel xz compression We can safely enable parallel xz compression for non-i386 pipelines. However, previously we didn't export XZ_OPT, so the xz process won't see it if XZ_OPT hasn't already been set in the current job. - - - - - 20432d16 by Cheng Shao at 2023-03-30T18:42:56+00:00 ci: unset CROSS_EMULATOR for js job - - - - - 4a24dbbe by Cheng Shao at 2023-03-30T18:42:56+00:00 ci: fix lint-testsuite job The list_broken make target will transitively depend on the calibrate.out target, which used STAGE1_GHC instead of TEST_HC. It really should be TEST_HC since that's what get passed in the gitlab CI config. - - - - - cea56ccc by Cheng Shao at 2023-03-30T18:42:56+00:00 ci: use alpine3_17-wasm image for wasm jobs Bump the ci-images dependency and use the new alpine3_17-wasm docker image for wasm jobs. - - - - - 79d0cb32 by Ben Gamari at 2023-03-30T18:43:53+00:00 testsuite/driver: Add basic support for testing cross-compilers - - - - - e7392b4e by Ben Gamari at 2023-03-30T18:43:53+00:00 testsuite/driver: Normalize away differences in ghc executable name - - - - - ee160d06 by Ben Gamari at 2023-03-30T18:43:53+00:00 hadrian: Pass CROSS_EMULATOR to runtests.py - - - - - 30c84511 by Cheng Shao at 2023-03-30T18:43:53+00:00 testsuite: don't add optllvm way for wasm32 - - - - - f1beee36 by Cheng Shao at 2023-03-30T18:43:53+00:00 testsuite: normalize the .wasm extension - - - - - a984a103 by Cheng Shao at 2023-03-30T18:43:53+00:00 testsuite: strip the cross ghc prefix in output and error message - - - - - f7478d95 by Cheng Shao at 2023-03-30T18:43:53+00:00 testsuite: handle target executable extension - - - - - 8fe8b653 by Cheng Shao at 2023-03-30T18:43:53+00:00 testsuite: mypy typing error fixes This patch fixes some mypy typing errors which weren't caught in previous linting jobs. - - - - - 0149f32f by Cheng Shao at 2023-03-30T18:43:53+00:00 testsuite: use context variable instead of thread-local variable This patch changes a thread-local variable to context variable instead, which works as intended when the testsuite transitions to use asyncio & coroutines instead of multi-threading to concurrently run test cases. Note that this also raises the minimum Python version to 3.7. - - - - - ea853ff0 by Cheng Shao at 2023-03-30T18:43:53+00:00 testsuite: asyncify the testsuite driver This patch refactors the testsuite driver, gets rid of multi-threading logic for running test cases concurrently, and uses asyncio & coroutines instead. This is not yak shaving for its own sake; the previous multi-threading logic is prone to livelock/deadlock conditions for some reason, even if the total number of threads is bounded to a thread pool's capacity. The asyncify change is an internal implementation detail of the testsuite driver and does not impact most GHC maintainers out there. The patch does not touch the .T files, test cases can be added/modified the exact same way as before. - - - - - 0077cb22 by Matthew Pickering at 2023-03-31T21:28:28-04:00 Add test for T23184 There was an outright bug, which Simon fixed in July 2021, as a little side-fix on a complicated patch: ``` commit 6656f0165a30fc2a22208532ba384fc8e2f11b46 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Fri Jul 23 23:57:01 2021 +0100 A bunch of changes related to eta reduction This is a large collection of changes all relating to eta reduction, originally triggered by #18993, but there followed a long saga. Specifics: ...lots of lines omitted... Other incidental changes * Fix a fairly long-standing outright bug in the ApplyToVal case of GHC.Core.Opt.Simplify.mkDupableContWithDmds. I was failing to take the tail of 'dmds' in the recursive call, which meant the demands were All Wrong. I have no idea why this has not caused problems before now. ``` Note this "Fix a fairly longstanding outright bug". This is the specific fix ``` @@ -3552,8 +3556,8 @@ mkDupableContWithDmds env dmds -- let a = ...arg... -- in [...hole...] a -- NB: sc_dup /= OkToDup; that is caught earlier by contIsDupable - do { let (dmd:_) = dmds -- Never fails - ; (floats1, cont') <- mkDupableContWithDmds env dmds cont + do { let (dmd:cont_dmds) = dmds -- Never fails + ; (floats1, cont') <- mkDupableContWithDmds env cont_dmds cont ; let env' = env `setInScopeFromF` floats1 ; (_, se', arg') <- simplArg env' dup se arg ; (let_floats2, arg'') <- makeTrivial env NotTopLevel dmd (fsLit "karg") arg' ``` Ticket #23184 is a report of the bug that this diff fixes. - - - - - 62d25071 by mangoiv at 2023-04-01T04:20:01-04:00 [feat] make ($) representation polymorphic - this change was approved by the CLC in [1] following a CLC proposal [2] - make ($) representation polymorphic (adjust the type signature) - change ($) implementation to allow additional polymorphism - adjust the haddock of ($) to reflect these changes - add additional documentation to document these changes - add changelog entry - adjust tests (move now succeeding tests and adjust stdout of some tests) [1] https://github.com/haskell/core-libraries-committee/issues/132#issuecomment-1487456854 [2] https://github.com/haskell/core-libraries-committee/issues/132 - - - - - 77c33fb9 by Artem Pelenitsyn at 2023-04-01T04:20:41-04:00 User Guide: update copyright year: 2020->2023 - - - - - 3b5be05a by doyougnu at 2023-04-01T09:42:31-04:00 driver: Unit State Data.Map -> GHC.Unique.UniqMap In pursuit of #22426. The driver and unit state are major contributors. This commit also bumps the haddock submodule to reflect the API changes in UniqMap. ------------------------- Metric Decrease: MultiComponentModules MultiComponentModulesRecomp T10421 T10547 T12150 T12234 T12425 T13035 T16875 T18140 T18304 T18698a T18698b T18923 T20049 T5837 T6048 T9198 ------------------------- - - - - - a84fba6e by Torsten Schmits at 2023-04-01T09:43:12-04:00 Add structured error messages for GHC.Tc.TyCl Tracking ticket: #20117 MR: !10183 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. - - - - - 6e2eb275 by doyougnu at 2023-04-01T18:27:56-04:00 JS: Linker: use saturated JExpr Follow on to MR!10142 in pursuit of #22736 - - - - - 3da69346 by sheaf at 2023-04-01T18:28:37-04:00 Improve haddocks of template-haskell Con datatype This adds a bit more information, in particular about the lists of constructors in the GadtC and RecGadtC cases. - - - - - 3b7bbb39 by sheaf at 2023-04-01T18:28:37-04:00 TH: revert changes to GadtC & RecGadtC Commit 3f374399 included a breaking-change to the template-haskell library when it made the GadtC and RecGadtC constructors take non-empty lists of names. As this has the potential to break many users' packages, we decided to revert these changes for now. - - - - - f60f6110 by Andrew Lelechenko at 2023-04-02T18:59:30-04:00 Rework documentation for data Char - - - - - 43ebd5dc by Andrew Lelechenko at 2023-04-02T19:00:09-04:00 cmm: implement parsing of MO_AtomicRMW from hand-written CMM files Fixes #23206 - - - - - ab9cd52d by Sylvain Henry at 2023-04-03T08:15:21-04:00 ghc-heap: remove wrong Addr# coercion (#23181) Conversion from Addr# to I# isn't correct with the JS backend. - - - - - 2b2afff3 by Matthew Pickering at 2023-04-03T08:15:58-04:00 hadrian: Update bootstrap plans for 9.2.6, 9.2.7, 9.4.4, 9.4.5, 9.6.1 Also fixes the ./generate_bootstrap_plans script which was recently broken We can hopefully drop the 9.2 plans soon but they still work so kept them around for now. - - - - - c2605e25 by Matthew Pickering at 2023-04-03T08:15:58-04:00 ci: Add job to test 9.6 bootstrapping - - - - - 53e4d513 by Krzysztof Gogolewski at 2023-04-03T08:16:35-04:00 hadrian: Improve option parsing Several options in Hadrian had their argument marked as optional (`OptArg`), but if the argument wasn't there they were just giving an error. It's more idiomatic to mark the argument as required instead; the code uses less Maybes, the parser can enforce that the argument is present, --help gives better output. - - - - - a8e36892 by Sylvain Henry at 2023-04-03T08:17:16-04:00 JS: fix issues with FD api support - Add missing implementations for fcntl_read/write/lock - Fix fdGetMode These were found while implementing TH in !9779. These functions must be used somehow by the external interpreter code. - - - - - 8b092910 by Haskell-mouse at 2023-04-03T19:31:26-04:00 Convert diagnostics in GHC.Rename.HsType to proper TcRnMessage I've turned all occurrences of TcRnUnknownMessage in GHC.Rename.HsType module into a proper TcRnMessage. Instead, these TcRnMessage messages were introduced: TcRnDataKindsError TcRnUnusedQuantifiedTypeVar TcRnIllegalKindSignature TcRnUnexpectedPatSigType TcRnSectionPrecedenceError TcRnPrecedenceParsingError TcRnIllegalKind TcRnNegativeNumTypeLiteral TcRnUnexpectedKindVar TcRnBindMultipleVariables TcRnBindVarAlreadyInScope - - - - - 220a7a48 by Krzysztof Gogolewski at 2023-04-03T19:32:02-04:00 Fixes around unsafeCoerce# 1. `unsafeCoerce#` was documented in `GHC.Prim`. But since the overhaul in 74ad75e87317, `unsafeCoerce#` is no longer defined there. I've combined the documentation in `GHC.Prim` with the `Unsafe.Coerce` module. 2. The documentation of `unsafeCoerce#` stated that you should not cast a function to an algebraic type, even if you later cast it back before applying it. But ghci was doing that type of cast, as can be seen with 'ghci -ddump-ds' and typing 'x = not'. I've changed it to use Any following the documentation. - - - - - 9095e297 by Matthew Craven at 2023-04-04T01:04:10-04:00 Add a few more memcpy-ish primops * copyMutableByteArrayNonOverlapping# * copyAddrToAddr# * copyAddrToAddrNonOverlapping# * setAddrRange# The implementations of copyBytes, moveBytes, and fillBytes in base:Foreign.Marshal.Utils now use these new primops, which can cause us to work a bit harder generating code for them, resulting in the metric increase in T21839c observed by CI on some architectures. But in exchange, we get better code! Metric Increase: T21839c - - - - - f7da530c by Matthew Craven at 2023-04-04T01:04:10-04:00 StgToCmm: Upgrade -fcheck-prim-bounds behavior Fixes #21054. Additionally, we can now check for range overlap when generating Cmm for primops that use memcpy internally. - - - - - cd00e321 by sheaf at 2023-04-04T01:04:50-04:00 Relax assertion in varToRecFieldOcc When using Template Haskell, it is possible to re-parent a field OccName belonging to one data constructor to another data constructor. The lsp-types package did this in order to "extend" a data constructor with additional fields. This ran into an assertion in 'varToRecFieldOcc'. This assertion can simply be relaxed, as the resulting splices are perfectly sound. Fixes #23220 - - - - - eed0d930 by Sylvain Henry at 2023-04-04T11:09:15-04:00 GHCi.RemoteTypes: fix doc and avoid unsafeCoerce (#23201) - - - - - 071139c3 by Ryan Scott at 2023-04-04T11:09:51-04:00 Make INLINE pragmas for pattern synonyms work with TH Previously, the code for converting `INLINE <name>` pragmas from TH splices used `vNameN`, which assumed that `<name>` must live in the variable namespace. Pattern synonyms, on the other hand, live in the constructor namespace. I've fixed the issue by switching to `vcNameN` instead, which works for both the variable and constructor namespaces. Fixes #23203. - - - - - 7c16f3be by Krzysztof Gogolewski at 2023-04-04T17:13:00-04:00 Fix unification with oversaturated type families unify_ty was incorrectly saying that F x y ~ T x are surely apart, where F x y is an oversaturated type family and T x is a tyconapp. As a result, the simplifier dropped a live case alternative (#23134). - - - - - c165f079 by sheaf at 2023-04-04T17:13:40-04:00 Add testcase for #23192 This issue around solving of constraints arising from superclass expansion using other constraints also borned from superclass expansion was the topic of commit aed1974e. That commit made sure we don't emit a "redundant constraint" warning in a situation in which removing the constraint would cause errors. Fixes #23192 - - - - - d1bb16ed by Ben Gamari at 2023-04-06T03:40:45-04:00 nonmoving: Disable slop-zeroing As noted in #23170, the nonmoving GC can race with a mutator zeroing the slop of an updated thunk (in much the same way that two mutators would race). Consequently, we must disable slop-zeroing when the nonmoving GC is in use. Closes #23170 - - - - - 04b80850 by Brandon Chinn at 2023-04-06T03:41:21-04:00 Fix reverse flag for -Wunsupported-llvm-version - - - - - 0c990e13 by Pierre Le Marre at 2023-04-06T10:16:29+00:00 Add release note for GHC.Unicode refactor in base-4.18. Also merge CLC proposal 130 in base-4.19 with CLC proposal 59 in base-4.18 and add proper release date. - - - - - cbbfb283 by Alex Dixon at 2023-04-07T18:27:45-04:00 Improve documentation for ($) (#22963) - - - - - 5193c2b0 by Alex Dixon at 2023-04-07T18:27:45-04:00 Remove trailing whitespace from ($) commentary - - - - - b384523b by Sebastian Graf at 2023-04-07T18:27:45-04:00 Adjust wording wrt representation polymorphism of ($) - - - - - 6a788f0a by Torsten Schmits at 2023-04-07T22:29:28-04:00 Add structured error messages for GHC.Tc.TyCl.Utils Tracking ticket: #20117 MR: !10251 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. - - - - - 3ba77b36 by sheaf at 2023-04-07T22:30:07-04:00 Renamer: don't call addUsedGRE on an exact Name When looking up a record field in GHC.Rename.Env.lookupRecFieldOcc, we could end up calling addUsedGRE on an exact Name, which would then lead to a panic in the bestImport function: it would be incapable of processing a GRE which is not local but also not brought into scope by any imports (as it is referred to by its unique instead). Fixes #23240 - - - - - bc4795d2 by Krzysztof Gogolewski at 2023-04-11T19:24:54-04:00 Add support for -debug in the testsuite Confusingly, GhcDebugged referred to GhcDebugAssertions. - - - - - b7474b57 by Krzysztof Gogolewski at 2023-04-11T19:24:54-04:00 Add missing cases in -Di prettyprinter Fixes #23142 - - - - - 6c392616 by Cheng Shao at 2023-04-11T19:25:31-04:00 compiler: make WasmCodeGenM an instance of MonadUnique - - - - - 05d26a65 by Cheng Shao at 2023-04-11T19:25:31-04:00 compiler: apply cmm node-splitting for wasm backend This patch applies cmm node-splitting for wasm32 NCG, which is required when handling irreducible CFGs. Fixes #23237. - - - - - f1892cc0 by Andrew Lelechenko at 2023-04-11T19:26:09-04:00 Set base 'maintainer' field to CLC - - - - - ecf22da3 by Simon Peyton Jones at 2023-04-11T19:26:45-04:00 Clarify a couple of Notes about 'nospec' - - - - - ebd8918b by Oleg Grenrus at 2023-04-12T12:32:57-04:00 Allow generation of TTH syntax with TH In other words allow generation of typed splices and brackets with Untyped Template Haskell. That is useful in cases where a library is build with TTH in mind, but we still want to generate some auxiliary declarations, where TTH cannot help us, but untyped TH can. Such example is e.g. `staged-sop` which works with TTH, but we would like to derive `Generic` declarations with TH. An alternative approach is to use `unsafeCodeCoerce`, but then the derived `Generic` instances would be type-checked only at use sites, i.e. much later. Also `-ddump-splices` output is quite ugly: user-written instances would use TTH brackets, not `unsafeCodeCoerce`. This commit doesn't allow generating of untyped template splices and brackets with untyped TH, as I don't know why one would want to do that (instead of merging the splices, e.g.) - - - - - 690d0225 by Rodrigo Mesquita at 2023-04-12T12:33:33-04:00 Add regression test for #23229 - - - - - 59321879 by Sylvain Henry at 2023-04-13T08:50:33-04:00 Add quotRem rules (#22152) case quotRemInt# x y of (# q, _ #) -> body ====> case quotInt# x y of q -> body case quotRemInt# x y of (# _, r #) -> body ====> case remInt# x y of r -> body - - - - - 4dd02122 by Sylvain Henry at 2023-04-13T08:50:33-04:00 Add quot folding rule (#22152) (x / l1) / l2 l1 and l2 /= 0 l1*l2 doesn't overflow ==> x / (l1 * l2) - - - - - 1148ac72 by Sylvain Henry at 2023-04-13T08:50:33-04:00 Make Int64/Word64 division ok for speculation too. Only when the divisor is definitely non-zero. - - - - - 8af401cc by Sylvain Henry at 2023-04-13T08:50:33-04:00 Make WordQuotRem2Op ok-for-speculation too - - - - - 27d2978e by Josh Meredith at 2023-04-13T08:51:09-04:00 Base/JS: GHC.JS.Foreign.Callback module (issue 23126) * Add the Callback module for "exporting" Haskell functions to be available to plain JavaScript code * Fix some primitives defined in GHC.JS.Prim * Add a JavaScript section to the user guide with instructions on how to use the JavaScript FFI, building up to using Callbacks to interact with the browser * Add tests for the JavaScript FFI and Callbacks - - - - - a34aa8da by Adam Sandberg Ericsson at 2023-04-14T04:17:52-04:00 rts: improve memory ordering and add some comments in the StablePtr implementation - - - - - d7a768a4 by Matthew Pickering at 2023-04-14T04:18:28-04:00 docs: Generate docs/index.html with version number * Generate docs/index.html to include the version of the ghc library * This also fixes the packageVersions interpolations which were - Missing an interpolation for `LIBRARY_ghc_VERSION` - Double quoting the version so that "9.7" was being inserted. Fixes #23121 - - - - - d48fbfea by Simon Peyton Jones at 2023-04-14T04:19:05-04:00 Stop if type constructors have kind errors Otherwise we get knock-on errors, such as #23252. This makes GHC fail a bit sooner, and I have not attempted to add recovery code, to add a fake TyCon place of the erroneous one, in an attempt to get more type errors in one pass. We could do that (perhaps) if there was a call for it. - - - - - 2371d6b2 by Simon Peyton Jones at 2023-04-14T20:01:02+02:00 Major refactor in the handling of equality constraints This MR substantially refactors the way in which the constraint solver deals with equality constraints. The big thing is: * Intead of a pipeline in which we /first/ canonicalise and /then/ interact (the latter including performing unification) the two steps are more closely integreated into one. That avoids the current rather indirect communication between the two steps. The proximate cause for this refactoring is fixing #22194, which involve solving [W] alpha[2] ~ Maybe (F beta[4]) by doing this: alpha[2] := Maybe delta[2] [W] delta[2] ~ F beta[4] That is, we don't promote beta[4]! This is very like introducing a cycle breaker, and was very awkward to do before, but now it is all nice. See GHC.Tc.Utils.Unify Note [Promotion and level-checking] and Note [Family applications in canonical constraints]. The big change is this: * Several canonicalisation checks (occurs-check, cycle-breaking, checking for concreteness) are combined into one new function: GHC.Tc.Utils.Unify.checkTyEqRhs This function is controlled by `TyEqFlags`, which says what to do for foralls, type families etc. * `canEqCanLHSFinish` now sees if unification is possible, and if so, actually does it: see `canEqCanLHSFinish_try_unification`. There are loads of smaller changes: * The on-the-fly unifier `GHC.Tc.Utils.Unify.unifyType` has a cheap-and-cheerful version of `checkTyEqRhs`, called `simpleUnifyCheck`. If `simpleUnifyCheck` succeeds, it can unify, otherwise it defers by emitting a constraint. This is simpler than before. * I simplified the swapping code in `GHC.Tc.Solver.Equality.canEqCanLHS`. Especially the nasty stuff involving `swap_for_occurs` and `canEqTyVarFunEq`. Much nicer now. See Note [Orienting TyVarLHS/TyFamLHS] Note [Orienting TyFamLHS/TyFamLHS] * Added `cteSkolemOccurs`, `cteConcrete`, and `cteCoercionHole` to the problems that can be discovered by `checkTyEqRhs`. * I fixed #23199 `pickQuantifiablePreds`, which actually allows GHC to to accept both cases in #22194 rather than rejecting both. Yet smaller: * Added a `synIsConcrete` flag to `SynonymTyCon` (alongside `synIsFamFree`) to reduce the need for synonym expansion when checking concreteness. Use it in `isConcreteType`. * Renamed `isConcrete` to `isConcreteType` * Defined `GHC.Core.TyCo.FVs.isInjectiveInType` as a more efficient way to find if a particular type variable is used injectively than finding all the injective variables. It is called in `GHC.Tc.Utils.Unify.definitely_poly`, which in turn is used quite a lot. * Moved `rewriterView` to `GHC.Core.Type`, so we can use it from the constraint solver. Fixes #22194, #23199 Compile times decrease by an average of 0.1%; but there is a 7.4% drop in compiler allocation on T15703. Metric Decrease: T15703 - - - - - 99b2734b by Simon Peyton Jones at 2023-04-14T20:01:02+02:00 Add some documentation about redundant constraints - - - - - 3f2d0eb8 by Simon Peyton Jones at 2023-04-14T20:01:02+02:00 Improve partial signatures This MR fixes #23223. The changes are in two places: * GHC.Tc.Bind.checkMonomorphismRestriction See the new `Note [When the MR applies]` We now no longer stupidly attempt to apply the MR when the user specifies a context, e.g. f :: Eq a => _ -> _ * GHC.Tc.Solver.decideQuantification See rewritten `Note [Constraints in partial type signatures]` Fixing this bug apparently breaks three tests: * partial-sigs/should_compile/T11192 * partial-sigs/should_fail/Defaulting1MROff * partial-sigs/should_fail/T11122 However they are all symptoms of #23232, so I'm marking them as expect_broken(23232). I feel happy about this MR. Nice. - - - - - 23e2a8a0 by Simon Peyton Jones at 2023-04-14T20:01:02+02:00 Make approximateWC a bit cleverer This MR fixes #23224: making approximateWC more clever See the long `Note [ApproximateWC]` in GHC.Tc.Solver All this is delicate and ad-hoc -- but it /has/ to be: we are talking about inferring a type for a binding in the presence of GADTs, type families and whatnot: known difficult territory. We just try as hard as we can. - - - - - 2c040246 by Matthew Pickering at 2023-04-15T00:57:14-04:00 docs: Update template-haskell docs to use Code Q a rather than Q (TExp a) Since GHC Proposal #195, the type of [|| ... ||] has been Code Q a rather than Q (TExp a). The documentation in the `template-haskell` library wasn't updated to reflect this change. Fixes #23148 - - - - - 0da18eb7 by Krzysztof Gogolewski at 2023-04-15T14:35:53+02:00 Show an error when we cannot default a concrete tyvar Fixes #23153 - - - - - bad2f8b8 by sheaf at 2023-04-15T15:14:36+02:00 Handle ConcreteTvs in inferResultToType inferResultToType was discarding the ir_frr information, which meant some metavariables ended up being MetaTvs instead of ConcreteTvs. This function now creates new ConcreteTvs as necessary, instead of always creating MetaTvs. Fixes #23154 - - - - - 3b0ea480 by Simon Peyton Jones at 2023-04-16T18:12:20-04:00 Transfer DFunId_ness onto specialised bindings Whether a binding is a DFunId or not has consequences for the `-fdicts-strict` flag, essentially if we are doing demand analysis for a DFunId then `-fdicts-strict` does not apply because the constraint solver can create recursive groups of dictionaries. In #22549 this was fixed for the "normal" case, see Note [Do not strictify the argument dictionaries of a dfun]. However the loop still existed if the DFunId was being specialised. The problem was that the specialiser would specialise a DFunId and turn it into a VanillaId and so the demand analyser didn't know to apply special treatment to the binding anymore and the whole recursive group was optimised to bottom. The solution is to transfer over the DFunId-ness of the binding in the specialiser so that the demand analyser knows not to apply the `-fstrict-dicts`. Fixes #22549 - - - - - a1371ebb by Oleg Grenrus at 2023-04-16T18:12:59-04:00 Add import lists to few GHC.Driver.Session imports Related to https://gitlab.haskell.org/ghc/ghc/-/issues/23261. There are a lot of GHC.Driver.Session which only use DynFlags, but not the parsing code. - - - - - 51479ceb by Matthew Pickering at 2023-04-17T08:08:48-04:00 Account for special GHC.Prim import in warnUnusedPackages The GHC.Prim import is treated quite specially primarily because there isn't an interface file for GHC.Prim. Therefore we record separately in the ModSummary if it's imported or not so we don't go looking for it. This logic hasn't made it's way to `-Wunused-packages` so if you imported GHC.Prim then the warning would complain you didn't use `-package ghc-prim`. Fixes #23212 - - - - - 1532a8b2 by Simon Peyton Jones at 2023-04-17T08:09:24-04:00 Add regression test for #23199 - - - - - 0158c5f1 by Ryan Scott at 2023-04-17T18:43:27-04:00 validDerivPred: Reject exotic constraints in IrredPreds This brings the `IrredPred` case in sync with the treatment of `ClassPred`s as described in `Note [Valid 'deriving' predicate]` in `GHC.Tc.Validity`. Namely, we should reject `IrredPred`s that are inferred from `deriving` clauses whose arguments contain other type constructors, as described in `(VD2) Reject exotic constraints` of that Note. This has the nice property that `deriving` clauses whose inferred instance context mention `TypeError` will now emit the type error in the resulting error message, which better matches existing intuitions about how `TypeError` should work. While I was in town, I noticed that much of `Note [Valid 'deriving' predicate]` was duplicated in a separate `Note [Exotic derived instance contexts]` in `GHC.Tc.Deriv.Infer`. I decided to fold the latter Note into the former so that there is a single authority on describing the conditions under which an inferred `deriving` constraint can be considered valid. This changes the behavior of `deriving` in a way that existing code might break, so I have made a mention of this in the GHC User's Guide. It seems very, very unlikely that much code is relying on this strange behavior, however, and even if there is, there is a clear, backwards-compatible migration path using `StandaloneDeriving`. Fixes #22696. - - - - - 10364818 by Krzysztof Gogolewski at 2023-04-17T18:44:03-04:00 Misc cleanup - Use dedicated list functions - Make cloneBndrs and cloneRecIdBndrs monadic - Fix invalid haddock comments in libraries/base - - - - - 5e1d33d7 by Matthew Pickering at 2023-04-18T10:31:02-04:00 Convert interface file loading errors into proper diagnostics This patch converts all the errors to do with loading interface files into proper structured diagnostics. * DriverMessage: Sometimes in the driver we attempt to load an interface file so we embed the IfaceMessage into the DriverMessage. * TcRnMessage: Most the time we are loading interface files during typechecking, so we embed the IfaceMessage This patch also removes the TcRnInterfaceLookupError constructor which is superceded by the IfaceMessage, which is now structured compared to just storing an SDoc before. - - - - - df1a5811 by sheaf at 2023-04-18T10:31:43-04:00 Don't panic in ltPatersonSize The function GHC.Tc.Utils.TcType.ltPatersonSize would panic when it encountered a type family on the RHS, as usually these are not allowed (type families are not allowed on the RHS of class instances or of quantified constraints). However, it is possible to still encounter type families on the RHS after doing a bit of constraint solving, as seen in test case T23171. This could trigger the panic in the call to ltPatersonSize in GHC.Tc.Solver.Canonical.mk_strict_superclasses, which is involved in avoiding loopy superclass constraints. This patch simply changes ltPatersonSize to return "I don't know, because there's a type family involved" in these cases. Fixes #23171 - - - - - d442ac05 by Sylvain Henry at 2023-04-19T20:04:35-04:00 JS: fix thread-related primops - - - - - 7a96f90b by Bryan Richter at 2023-04-19T20:05:11-04:00 CI: Disable abi-test-nightly See #23269 - - - - - ab6c1d29 by Sylvain Henry at 2023-04-19T20:05:50-04:00 Testsuite: don't use obsolescent egrep (#22351) Recent egrep displays the following message, breaking golden tests: egrep: warning: egrep is obsolescent; using grep -E Switch to using "grep -E" instead - - - - - 7866fc86 by Ben Orchard at 2023-04-20T11:29:33+02:00 update classify with new tokens - - - - - f15b0ce5 by Matthew Pickering at 2023-04-20T11:01:06-04:00 hadrian: Pass haddock file arguments in a response file In !10119 CI was failing on windows because the command line was too long. We can mitigate this by passing the file arguments to haddock in a response file. We can't easily pass all the arguments in a response file because the `+RTS` arguments can't be placed in the response file. Fixes #23273 - - - - - 7012ec2f by tocic at 2023-04-20T11:01:42-04:00 Fix doc typo in GHC.Read.readList - - - - - 5c873124 by sheaf at 2023-04-20T18:33:34-04:00 Implement -jsem: parallelism controlled by semaphores See https://github.com/ghc-proposals/ghc-proposals/pull/540/ for a complete description for the motivation for this feature. The `-jsem` option allows a build tool to pass a semaphore to GHC which GHC can use in order to control how much parallelism it requests. GHC itself acts as a client in the GHC jobserver protocol. ``` GHC Jobserver Protocol ~~~~~~~~~~~~~~~~~~~~~~ This proposal introduces the GHC Jobserver Protocol. This protocol allows a server to dynamically invoke many instances of a client process, while restricting all of those instances to use no more than <n> capabilities. This is achieved by coordination over a system semaphore (either a POSIX semaphore [6]_ in the case of Linux and Darwin, or a Win32 semaphore [7]_ in the case of Windows platforms). There are two kinds of participants in the GHC Jobserver protocol: - The *jobserver* creates a system semaphore with a certain number of available tokens. Each time the jobserver wants to spawn a new jobclient subprocess, it **must** first acquire a single token from the semaphore, before spawning the subprocess. This token **must** be released once the subprocess terminates. Once work is finished, the jobserver **must** destroy the semaphore it created. - A *jobclient* is a subprocess spawned by the jobserver or another jobclient. Each jobclient starts with one available token (its *implicit token*, which was acquired by the parent which spawned it), and can request more tokens through the Jobserver Protocol by waiting on the semaphore. Each time a jobclient wants to spawn a new jobclient subprocess, it **must** pass on a single token to the child jobclient. This token can either be the jobclient's implicit token, or another token which the jobclient acquired from the semaphore. Each jobclient **must** release exactly as many tokens as it has acquired from the semaphore (this does not include the implicit tokens). ``` Build tools such as cabal act as jobservers in the protocol and are responsibile for correctly creating, cleaning up and managing the semaphore. Adds a new submodule (semaphore-compat) for managing and interacting with semaphores in a cross-platform way. Fixes #19349 - - - - - 52d3e9b4 by Ben Gamari at 2023-04-20T18:34:11-04:00 rts: Initialize Array# header in listThreads# Previously the implementation of listThreads# failed to initialize the header of the created array, leading to various nastiness. Fixes #23071 - - - - - 1db30fe1 by Ben Gamari at 2023-04-20T18:34:11-04:00 testsuite: Add test for #23071 - - - - - dae514f9 by tocic at 2023-04-21T13:31:21-04:00 Fix doc typos in libraries/base/GHC - - - - - 113e21d7 by Sylvain Henry at 2023-04-21T13:32:01-04:00 Testsuite: replace some js_broken/js_skip predicates with req_c Using req_c is more precise. - - - - - 038bb031 by Krzysztof Gogolewski at 2023-04-21T18:03:04-04:00 Minor doc fixes - Add docs/index.html to .gitignore. It is created by ./hadrian/build docs, and it was the only file in Hadrian's templateRules not present in .gitignore. - Mention that MultiWayIf supports non-boolean guards - Remove documentation of optdll - removed in 2007, 763daed95 - Fix markdown syntax - - - - - e826cdb2 by amesgen at 2023-04-21T18:03:44-04:00 User's guide: DeepSubsumption is implied by Haskell{98,2010} - - - - - 499a1c20 by PHO at 2023-04-23T13:39:32-04:00 Implement executablePath for Solaris and make getBaseDir less platform-dependent Use base-4.17 executablePath when possible, and fall back on getExecutablePath when it's not available. The sole reason why getBaseDir had #ifdef's was apparently that getExecutablePath wasn't reliable, and we could reduce the number of CPP conditionals by making use of executablePath instead. Also export executablePath on js_HOST_ARCH. - - - - - 97a6f7bc by tocic at 2023-04-23T13:40:08-04:00 Fix doc typos in libraries/base - - - - - ffcdd683 by Finley McIlwaine at 2023-04-24T09:36:18-06:00 Remove index-state - - - - - 787c6e8c by Ben Gamari at 2023-04-24T12:19:06-04:00 testsuite/T20137: Avoid impl.-defined behavior Previously we would cast pointers to uint64_t. However, implementations are allowed to either zero- or sign-extend such casts. Instead cast to uintptr_t to avoid this. Fixes #23247. - - - - - 87095f6a by Cheng Shao at 2023-04-24T12:19:44-04:00 rts: always build 64-bit atomic ops This patch does a few things: - Always build 64-bit atomic ops in rts/ghc-prim, even on 32-bit platforms - Remove legacy "64bit" cabal flag of rts package - Fix hs_xchg64 function prototype for 32-bit platforms - Fix AtomicFetch test for wasm32 - - - - - 2685a12d by Cheng Shao at 2023-04-24T12:20:21-04:00 compiler: don't install signal handlers when the host platform doesn't have signals Previously, large parts of GHC API will transitively invoke withSignalHandlers, which doesn't work on host platforms without signal functionality at all (e.g. wasm32-wasi). By making withSignalHandlers a no-op on those platforms, we can make more parts of GHC API work out of the box when signals aren't supported. - - - - - 1338b7a3 by Cheng Shao at 2023-04-24T16:21:30-04:00 hadrian: fix non-ghc program paths passed to testsuite driver when testing cross GHC - - - - - 1a10f556 by Andrew Lelechenko at 2023-04-24T16:22:09-04:00 Add since pragma to Data.Functor.unzip - - - - - 0da9e882 by Soham Chowdhury at 2023-04-25T00:15:22-04:00 More informative errors for bad imports (#21826) - - - - - ebd5b078 by Josh Meredith at 2023-04-25T00:15:58-04:00 JS/base: provide implementation for mkdir (issue 22374) - - - - - 8f656188 by Josh Meredith at 2023-04-25T18:12:38-04:00 JS: Fix h$base_access implementation (issue 22576) - - - - - 74c55712 by Andrei Borzenkov at 2023-04-25T18:13:19-04:00 Give more guarntees about ImplicitParams (#23289) - Added new section in the GHC user's guide that legends behavior of nested implicit parameter bindings in these two cases: let ?f = 1 in let ?f = 2 in ?f and data T where MkT :: (?f :: Int) => T f :: T -> T -> Int f MkT MkT = ?f - Added new test case to examine this behavior. - - - - - 05b70982 by Finley McIlwaine at 2023-04-26T08:16:31-06:00 `renameInterface` space leak fixes - Change logic for accumulation of names for which link warnings will be generated - Change handling of `--ignore-link-symbol` to allow qualified and unqualified names. Added to CHANGES.md - Some formatting changes and comments here and there - - - - - c30ac25f by Sebastian Graf at 2023-04-26T14:50:51-04:00 DmdAnal: Unleash demand signatures of free RULE and unfolding binders (#23208) In #23208 we observed that the demand signature of a binder occuring in a RULE wasn't unleashed, leading to a transitively used binder being discarded as absent. The solution was to use the same code path that we already use for handling exported bindings. See the changes to `Note [Absence analysis for stable unfoldings and RULES]` for more details. I took the chance to factor out the old notion of a `PlusDmdArg` (a pair of a `VarEnv Demand` and a `Divergence`) into `DmdEnv`, which fits nicely into our existing framework. As a result, I had to touch quite a few places in the code. This refactoring exposed a few small bugs around correct handling of bottoming demand environments. As a result, some strictness signatures now mention uniques that weren't there before which caused test output changes to T13143, T19969 and T22112. But these tests compared whole -ddump-simpl listings which is a very fragile thing to begin with. I changed what exactly they test for based on the symptoms in the corresponding issues. There is a single regression in T18894 because we are more conservative around stable unfoldings now. Unfortunately it is not easily fixed; let's wait until there is a concrete motivation before invest more time. Fixes #23208. - - - - - 77f506b8 by Josh Meredith at 2023-04-26T14:51:28-04:00 Refactor GenStgRhs to include the Type in both constructors (#23280, #22576, #22364) Carry the actual type of an expression through the PreStgRhs and into GenStgRhs for use in later stages. Currently this is used in the JavaScript backend to fix some tests from the above mentioned issues: EtaExpandLevPoly, RepPolyWrappedVar2, T13822, T14749. - - - - - 052e2bb6 by Alan Zimmerman at 2023-04-26T14:52:05-04:00 EPA: Use ExplicitBraces only in HsModule !9018 brought in exact print annotations in LayoutInfo for open and close braces at the top level. But it retained them in the HsModule annotations too. Remove the originals, so exact printing uses LayoutInfo - - - - - d5c4629b by Cheng Shao at 2023-04-27T16:00:35-04:00 ci: update ci.sh to actually run the entire testsuite for wasm backend For the time being, we still need to use in-tree mode and can't test the bindist yet. - - - - - 533d075e by Cheng Shao at 2023-04-27T16:00:35-04:00 ci: additional wasm32 manual jobs in validate pipelines This patch enables bignum native & unregisterised wasm32 jobs as manual jobs in validate pipelines, which can be useful to prevent breakage when working on wasm32 related patches. - - - - - b5f00811 by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: fix cross prefix stripping This patch fixes cross prefix stripping in the testsuite driver. The normalization logic used to only handle prefixes of the triple form <arch>-<vendor>-<os>, now it's relaxed to allow any number of tokens in the prefix tuple, so the cross prefix stripping logic would work when ghc is configured with something like --target=wasm32-wasi. - - - - - 6f511c36 by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: include target exe extension in heap profile filenames This patch fixes hp2ps related framework failures when testing the wasm backend by including target exe extension in heap profile filenames. - - - - - e6416b10 by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: exclude ghci ways if no rts linker is present This patch implements logic to automatically exclude ghci ways when there is no rts linker. It's way better than having to annotate individual test cases. - - - - - 791cce64 by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: fix permission bits in copy_files When the testsuite driver copy files instead of symlinking them, it should also copy the permission bits, otherwise there'll be permission denied errors. Also, enforce file copying when testing wasm32, since wasmtime doesn't handle host symlinks quite well (https://github.com/bytecodealliance/wasmtime/issues/6227). - - - - - aa6afe8a by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: add the req_ghc_with_threaded_rts predicate This patch adds the req_ghc_with_threaded_rts predicate to the testsuite to assert the platform has threaded RTS, and mark some tests as req_ghc_with_threaded_rts. Also makes ghc_with_threaded_rts a config field instead of a global variable. - - - - - ce580426 by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: add the req_process predicate This patch adds the req_process predicate to the testsuite to assert the platform has a process model, also marking tests that involve spawning processes as req_process. Also bumps hpc & process submodule. - - - - - cb933665 by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: add the req_host_target_ghc predicate This patch adds the req_host_target_ghc predicate to the testsuite to assert the ghc compiler being tested can compile both host/target code. When testing cross GHCs this is not supported yet, but it may change in the future. - - - - - b174a110 by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: add missing annotations for some tests This patch adds missing annotations (req_th, req_dynamic_lib_support, req_rts_linker) to some tests. They were discovered when testing wasm32, though it's better to be explicit about what features they require, rather than simply adding when(arch('wasm32'), skip). - - - - - bd2bfdec by Cheng Shao at 2023-04-27T16:00:35-04:00 testsuite: wasm32-specific fixes This patch includes all wasm32-specific testsuite fixes. - - - - - 4eaf2c2a by Josh Meredith at 2023-04-27T16:01:11-04:00 JS: change GHC.JS.Transform.identsS/E/V to take a saturated IR (#23304) - - - - - e5697d7c by Finley McIlwaine at 2023-04-27T18:46:36-06:00 Messy things - ghc-debug dependency and instrumentation - cabal.project custom with-compiler - hie.yaml files - traces and such - - - - - 57277662 by sheaf at 2023-04-29T20:23:06+02:00 Add the Unsatisfiable class This commit implements GHC proposal #433, adding the Unsatisfiable class to the GHC.TypeError module. This provides an alternative to TypeError for which error reporting is more predictable: we report it when we are reporting unsolved Wanted constraints. Fixes #14983 #16249 #16906 #18310 #20835 - - - - - 00a8a5ff by Torsten Schmits at 2023-04-30T03:45:09-04:00 Add structured error messages for GHC.Rename.Names Tracking ticket: #20115 MR: !10336 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. - - - - - 0b8ef80b by Finley McIlwaine at 2023-05-02T18:08:52-06:00 Stop retaining GRE closures GRE closures should never be necessary to Haddock, so we never want to keep them on the heap. Despite that, they are retained by a lot of the data structures that Haddock makes use of. - Attempt to fix that situation by adding strictness to various computations and pruning the `ifaceInstances` field of `Interface` to a much thinner data type. - Removes the `ifaceFamInstances` field, as it was never used. - Move some of the attach instances types (e.g. `SimpleType`) to the types module - - - - - 931c8d82 by Ben Orchard at 2023-05-03T20:16:18-04:00 Add sized primitive literal syntax Adds a new LANGUAGE pragma ExtendedLiterals, which enables defining unboxed numeric literals such as `0xFF#Word8 :: Word8#`. Implements GHC proposal 0451: https://github.com/ghc-proposals/ghc-proposals/blob/b384a538b34f79d18a0201455b7b3c473bc8c936/proposals/0451-sized-literals.rst Fixes #21422. Bumps haddock submodule. Co-authored-by: Krzysztof Gogolewski <krzysztof.gogolewski at tweag.io> - - - - - f3460845 by Andrew Lelechenko at 2023-05-03T20:16:57-04:00 Document instances of Double - - - - - 1e9caa1a by Sylvain Henry at 2023-05-03T20:17:37-04:00 Bump Cabal submodule (#22356) - - - - - 4eafb52a by sheaf at 2023-05-03T20:18:16-04:00 Don't forget to check the parent in an export list Commit 3f374399 introduced a bug which caused us to forget to include the parent of an export item of the form T(..) (that is, IEThingAll) when checking for duplicate exports. Fixes #23318 - - - - - 8fde4ac8 by amesgen at 2023-05-03T20:18:57-04:00 Fix unlit path in cross bindists - - - - - 8cc9a534 by Matthew Pickering at 2023-05-04T14:58:14-04:00 hadrian: Flavour: Change args -> extraArgs Previously in a flavour definition you could override all the flags which were passed to GHC. This causes issues when needed to compute a package hash because we need to know what these extra arguments are going to be before computing the hash. The solution is to modify flavour so that the arguments you pass here are just extra ones rather than all the arguments that you need to compile something. This makes things work more like how cabal.project files work when you give extra arguments to a package and also means that flavour transformers correctly affect the hash. - - - - - 3fdb18f8 by romes at 2023-05-04T14:58:14-04:00 Hardwire a better unit-id for ghc Previously, the unit-id of ghc-the-library was fixed as `ghc`. This was done primarily because the compiler must know the unit-id of some packages (including ghc) a-priori to define wired-in names. However, as seen in #20742, a reinstallable `ghc` whose unit-id is fixed to `ghc` might result in subtle bugs when different ghc's interact. A good example of this is having GHC_A load a plugin compiled by GHC_B, where GHC_A and GHC_B are linked to ghc-libraries that are ABI incompatible. Without a distinction between the unit-id of the ghc library GHC_A is linked against and the ghc library the plugin it is loading was compiled against, we can't check compatibility. This patch gives a slightly better unit-id to ghc (ghc-version) by (1) Not setting -this-unit-id to ghc, but rather to the new unit-id (modulo stage0) (2) Adding a definition to `GHC.Settings.Config` whose value is the new unit-id. (2.1) `GHC.Settings.Config` is generated by Hadrian (2.2) and also by cabal through `compiler/Setup.hs` This unit-id definition is imported by `GHC.Unit.Types` and used to set the wired-in unit-id of "ghc", which was previously fixed to "ghc" The commits following this one will improve the unit-id with a cabal-style package hash and check compatibility when loading plugins. Note that we also ensure that ghc's unit key matches unit id both when hadrian or cabal builds ghc, and in this way we no longer need to add `ghc` to the WiringMap. - - - - - 6689c9c6 by romes at 2023-05-04T14:58:14-04:00 Validate compatibility of ghcs when loading plugins Ensure, when loading plugins, that the ghc the plugin depends on is the ghc loading the plugin -- otherwise fail to load the plugin. Progress towards #20742. - - - - - db4be339 by romes at 2023-05-04T14:58:14-04:00 Add hashes to unit-ids created by hadrian This commit adds support for computing an inputs hash for packages compiled by hadrian. The result is that ABI incompatible packages should be given different hashes and therefore be distinct in a cabal store. Hashing is enabled by the `--flag`, and is off by default as the hash contains a hash of the source files. We enable it when we produce release builds so that the artifacts we distribute have the right unit ids. - - - - - 944a9b94 by Matthew Pickering at 2023-05-04T14:58:14-04:00 Use hash-unit-ids in release jobs Includes fix upload_ghc_libs glob - - - - - 116d7312 by Josh Meredith at 2023-05-04T14:58:51-04:00 JS: fix bounds checking (Issue 23123) * For ByteArray-based bounds-checking, the JavaScript backend must use the `len` field, instead of the inbuild JavaScript `length` field. * Range-based operations must also check both the start and end of the range for bounds * All indicies are valid for ranges of size zero, since they are essentially no-ops * For cases of ByteArray accesses (e.g. read as Int), the end index is (i * sizeof(type) + sizeof(type) - 1), while the previous implementation uses (i + sizeof(type) - 1). In the Int32 example, this is (i * 4 + 3) * IndexByteArrayOp_Word8As* primitives use byte array indicies (unlike the previous point), but now check both start and end indicies * Byte array copies now check if the arrays are the same by identity and then if the ranges overlap. - - - - - 2d5c1dde by Sylvain Henry at 2023-05-04T14:58:51-04:00 Fix remaining issues with bound checking (#23123) While fixing these I've also changed the way we store addresses into ByteArray#. Addr# are composed of two parts: a JavaScript array and an offset (32-bit number). Suppose we want to store an Addr# in a ByteArray# foo at offset i. Before this patch, we were storing both fields as a tuple in the "arr" array field: foo.arr[i] = [addr_arr, addr_offset]; Now we only store the array part in the "arr" field and the offset directly in the array: foo.dv.setInt32(i, addr_offset): foo.arr[i] = addr_arr; It avoids wasting space for the tuple. - - - - - 98c5ee45 by Luite Stegeman at 2023-05-04T14:59:31-04:00 JavaScript: Correct arguments to h$appendToHsStringA fixes #23278 - - - - - ca611447 by Josh Meredith at 2023-05-04T15:00:07-04:00 base/encoding: add an allocations performance test (#22946) - - - - - e3ddf58d by Krzysztof Gogolewski at 2023-05-04T15:00:44-04:00 linear types: Don't add external names to the usage env This has no observable effect, but avoids storing useless data. - - - - - b3226616 by Andrei Borzenkov at 2023-05-04T15:01:25-04:00 Improved documentation for the Data.OldList.nub function There was recomentation to use map head . group . sort instead of nub function, but containers library has more suitable and efficient analogue - - - - - e8b72ff6 by Ryan Scott at 2023-05-04T15:02:02-04:00 Fix type variable substitution in gen_Newtype_fam_insts Previously, `gen_Newtype_fam_insts` was substituting the type variable binders of a type family instance using `substTyVars`, which failed to take type variable dependencies into account. There is similar code in `GHC.Tc.TyCl.Class.tcATDefault` that _does_ perform this substitution properly, so this patch: 1. Factors out this code into a top-level `substATBndrs` function, and 2. Uses `substATBndrs` in `gen_Newtype_fam_insts`. Fixes #23329. - - - - - 275836d2 by Torsten Schmits at 2023-05-05T08:43:02+00:00 Add structured error messages for GHC.Rename.Utils Tracking ticket: #20115 MR: !10350 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. - - - - - 983ce558 by Oleg Grenrus at 2023-05-05T13:11:29-04:00 Use TemplateHaskellQuotes in TH.Syntax to construct Names - - - - - a5174a59 by Matthew Pickering at 2023-05-05T18:42:31-04:00 driver: Use hooks from plugin_hsc_env This fixes a bug in oneshot mode where hooks modified in a plugin wouldn't be used in oneshot mode because we neglected to use the right hsc_env. This was observed by @csabahruska. - - - - - 18a7d03d by Aaron Allen at 2023-05-05T18:42:31-04:00 Rework plugin initialisation points In general this patch pushes plugin initialisation points to earlier in the pipeline. As plugins can modify the `HscEnv`, it's imperative that the plugins are initialised as soon as possible and used thereafter. For example, there are some new tests which modify hsc_logger and other hooks which failed to fire before (and now do) One consequence of this change is that the error for specifying the usage of a HPT plugin from the command line has changed, because it's now attempted to be loaded at initialisation rather than causing a cyclic module import. Closes #21279 Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 6e776ed3 by Matthew Pickering at 2023-05-05T18:42:31-04:00 docs: Add Note [Timing of plugin initialization] - - - - - e1df8511 by Matthew Pickering at 2023-05-05T18:43:07-04:00 Incrementally update ghcup metadata in ghc/ghcup-metadata This job paves the way for distributing nightly builds * A new repo https://gitlab.haskell.org/ghc/ghcup-metadata stores the metadata on the "updates" branch. * Each night this metadata is downloaded and the nightly builds are appended to the end of the metadata. * The update job only runs on the scheduled nightly pipeline, not just when NIGHTLY=1. Things which are not done yet * Modify the retention policy for nightly jobs * Think about building release flavour compilers to distribute nightly. Fixes #23334 - - - - - 8f303d27 by Rodrigo Mesquita at 2023-05-05T22:04:31-04:00 docs: Remove mentions of ArrayArray# from unlifted FFI section Fixes #23277 - - - - - 994bda56 by Torsten Schmits at 2023-05-05T22:05:12-04:00 Add structured error messages for GHC.Rename.Module Tracking ticket: #20115 MR: !10361 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. Only addresses the single warning missing from the previous MR. - - - - - 3e3a6be4 by Ben Gamari at 2023-05-08T12:15:19+00:00 rts: Fix data-race in hs_init_ghc As noticed by @Terrorjack, `hs_init_ghc` previously used non-atomic increment/decrement on the RTS's initialization count. This may go wrong in a multithreaded program which initializes the runtime multiple times. Closes #22756. - - - - - 8bda991b by Finley McIlwaine at 2023-05-08T16:07:51-06:00 Memory usage fixes - Refactor `ifaceDeclMap` to drastically reduce memory footprint. We no longer store all declarations associated with a given name, since we only cared to determine if the only declaration associated with a name was a value declaration. Change the `DeclMap` type to better reflect this. - Drop pre-renaming export items after the renaming step. Since the Hoogle backend used the pre-renamed export items, this isn't trivial. We now generate Hoogle output for exported declarations during the renaming step (if Hoogle output /should/ be generated), and store that with the renamed export item. - Slightly refactor Hoogle backend to handle the above change and allow for early generation of Hoogle output. - Remove the `ifaceRnDocMap` and `ifaceRnArgMap` fields of the `Interface` type, as they were never used. - Remove some unnecessary strictness - Remove a lot of dead code from `Syb` module - - - - - 78c8dc50 by Torsten Schmits at 2023-05-08T21:41:51-04:00 Add structured error messages for GHC.IfaceToCore Tracking ticket: #20114 MR: !10390 This converts uses of `mkTcRnUnknownMessage` to newly added constructors of `TcRnMessage`. - - - - - 0e2df4c9 by Bryan Richter at 2023-05-09T12:03:35+03:00 Fix up rules for ghcup-metadata-nightly-push - - - - - b970e64f by Ben Gamari at 2023-05-09T08:41:33-04:00 testsuite: Add test for atomicSwapIORef - - - - - 81cfefd2 by Ben Gamari at 2023-05-09T08:41:53-04:00 compiler: Implement atomicSwapIORef with xchg As requested by @treeowl in CLC#139. - - - - - 6b29154d by Ben Gamari at 2023-05-09T08:41:53-04:00 Make atomicSwapMutVar# an inline primop - - - - - 1611ac0c by Finley McIlwaine at 2023-05-09T11:51:57-06:00 Unify ErrMsgM and IfM - Delete ErrMsgM, stop accumulating warnings in a writer - Make IfM a state monad, print warnings directly to stdout, move IfM type into types module - Drop ErrMsg = String synonym - Unset IORefs from plugin after they are read, preventing unnecessary retention of interfaces - - - - - 64064cfe by doyougnu at 2023-05-09T18:40:01-04:00 JS: add GHC.JS.Optimizer, remove RTS.Printer, add Linker.Opt This MR changes some simple optimizations and is a first step in re-architecting the JS backend pipeline to add the optimizer. In particular it: - removes simple peep hole optimizations from `GHC.StgToJS.Printer` and removes that module - adds module `GHC.JS.Optimizer` - defines the same peep hole opts that were removed only now they are `Syntax -> Syntax` transformations rather than `Syntax -> JS code` optimizations - hooks the optimizer into code gen - adds FuncStat and ForStat constructors to the backend. Working Ticket: - #22736 Related MRs: - MR !10142 - MR !10000 ------------------------- Metric Decrease: CoOpt_Read ManyAlternatives PmSeriesS PmSeriesT PmSeriesV T10421 T12707 T13253 T13253-spj T15164 T17516 T18140 T18282 T18698a T18698b T18923 T1969 T19695 T20049 T3064 T5321FD T5321Fun T783 T9198 T9233 T9630 ------------------------- - - - - - 6738c01d by Krzysztof Gogolewski at 2023-05-09T18:40:38-04:00 Add a regression test for #21050 - - - - - b2cdb7da by Ben Gamari at 2023-05-09T18:41:14-04:00 nonmoving: Account for mutator allocations in bytes_allocated Previously we failed to account direct mutator allocations into the nonmoving heap against the mutator's allocation limit and `cap->total_allocated`. This only manifests during CAF evaluation (since we allocate the CAF's blackhole directly into the nonmoving heap). Fixes #23312. - - - - - 0657b482 by Sven Tennie at 2023-05-09T22:22:42-04:00 Adjust AArch64 stackFrameHeaderSize The prologue of each stack frame are the saved LR and FP registers, 8 byte each. I.e. the size of the stack frame header is 2 * 8 byte. - - - - - 7788c09c by konsumlamm at 2023-05-09T22:23:23-04:00 Make `(&)` representation polymorphic in the return type - - - - - b3195922 by Ben Gamari at 2023-05-10T05:06:45-04:00 ghc-prim: Generalize keepAlive#/touch# in state token type Closes #23163. - - - - - 1e6861dd by Cheng Shao at 2023-05-10T05:07:25-04:00 Bump hsc2hs submodule Fixes #22981. - - - - - 0a513952 by Ben Gamari at 2023-05-11T04:10:17-04:00 base: Export GHC.Conc.Sync.fromThreadId Closes #22706. - - - - - 29be39ba by Matthew Pickering at 2023-05-11T04:10:54-04:00 Build vanilla alpine bindists We currently attempt to build and distribute fully static alpine bindists (ones which could be used on any linux platform) but most people who use the alpine bindists want to use alpine to build their own static applications (for which a fully static bindist is not necessary). We should build and distribute these bindists for these users whilst the fully-static bindist is still unusable. Fixes #23349 - - - - - 40c7daed by Simon Peyton Jones at 2023-05-11T04:11:30-04:00 Look both ways when looking for quantified equalities When looking up (t1 ~# t2) in the quantified constraints, check both orientations. Forgetting this led to #23333. - - - - - c17bb82f by Rodrigo Mesquita at 2023-05-11T04:12:07-04:00 Move "target has RTS linker" out of settings We move the "target has RTS linker" information out of configure into a predicate in GHC, and remove this option from the settings file where it is unnecessary -- it's information statically known from the platform. Note that previously we would consider `powerpc`s and `s390x`s other than `powerpc-ibm-aix*` and `s390x-ibm-linux` to have an RTS linker, but the RTS linker supports neither platform. Closes #23361 - - - - - bd0b056e by Krzysztof Gogolewski at 2023-05-11T04:12:44-04:00 Add a test for #17284 Since !10123 we now reject this program. - - - - - 630b1fea by Andrew Lelechenko at 2023-05-11T04:13:24-04:00 Document unlawfulness of instance Num Fixed Fixes #22712 - - - - - 87eebf98 by sheaf at 2023-05-11T11:55:22-04:00 Add fused multiply-add instructions This patch adds eight new primops that fuse a multiplication and an addition or subtraction: - `{fmadd,fmsub,fnmadd,fnmsub}{Float,Double}#` fmadd x y z is x * y + z, computed with a single rounding step. This patch implements code generation for these primops in the following backends: - X86, AArch64 and PowerPC NCG, - LLVM - C WASM uses the C implementation. The primops are unsupported in the JavaScript backend. The following constant folding rules are also provided: - compute a * b + c when a, b, c are all literals, - x * y + 0 ==> x * y, - ±1 * y + z ==> z ± y and x * ±1 + z ==> z ± x. NB: the constant folding rules incorrectly handle signed zero. This is a known limitation with GHC's floating-point constant folding rules (#21227), which we hope to resolve in the future. - - - - - ad16a066 by Krzysztof Gogolewski at 2023-05-11T11:55:59-04:00 Add a test for #21278 - - - - - 05cea68c by Matthew Pickering at 2023-05-11T11:56:36-04:00 rts: Refine memory retention behaviour to account for pinned/compacted objects When using the copying collector there is still a lot of data which isn't copied (such as pinned, compacted, large objects etc). The logic to decide how much memory to retain didn't take into account that these wouldn't be copied. Therefore we pessimistically retained 2* the amount of memory for these blocks even though they wouldn't be copied by the collector. The solution is to split up the heap into two parts, the parts which will be copied and the parts which won't be copied. Then the appropiate factor is applied to each part individually (2 * for copying and 1.2 * for not copying). The T23221 test demonstrates this improvement with a program which first allocates many unpinned ByteArray# followed by many pinned ByteArray# and observes the difference in the ultimate memory baseline between the two. There are some charts on #23221. Fixes #23221 - - - - - 1bb24432 by Cheng Shao at 2023-05-11T11:57:15-04:00 hadrian: fix no_dynamic_libs flavour transformer This patch fixes the no_dynamic_libs flavour transformer and make fully_static reuse it. Previously building with no_dynamic_libs fails since ghc program is still dynamic and transitively brings in dyn ways of rts which are produced by no rules. - - - - - 42d696ab by Finley McIlwaine at 2023-05-11T15:52:07-06:00 Thunk leak fixes The strictness introduced in this commit was motivated by observing thunk leaks in the eventlog2html output. - Refactor attach instances list comprehension to avoid large intermediate thunks - Refactor some HTML backend list comprehensions to avoid large intermediate thunks - Avoid thunks accumulating in documentation types or documentation parser - A lot of orphan NFData instances to allow us to force documentation values - - - - - 68561cf6 by Finley McIlwaine at 2023-05-11T17:02:10-06:00 Remove GHC debug dep - - - - - 0ed493a3 by Josh Meredith at 2023-05-11T23:08:27-04:00 JS: refactor jsSaturate to return a saturated JStat (#23328) - - - - - a856d98e by Pierre Le Marre at 2023-05-11T23:09:08-04:00 Doc: Fix out-of-sync using-optimisation page - Make explicit that default flag values correspond to their -O0 value. - Fix -fignore-interface-pragmas, -fstg-cse, -fdo-eta-reduction, -fcross-module-specialise, -fsolve-constant-dicts, -fworker-wrapper. - - - - - c176ad18 by sheaf at 2023-05-12T06:10:57-04:00 Don't panic in mkNewTyConRhs This function could come across invalid newtype constructors, as we only perform validity checking of newtypes once we are outside the knot-tied typechecking loop. This patch changes this function to fake up a stub type in the case of an invalid newtype, instead of panicking. This patch also changes "checkNewDataCon" so that it reports as many errors as possible at once. Fixes #23308 - - - - - ab63daac by Krzysztof Gogolewski at 2023-05-12T06:11:38-04:00 Allow Core optimizations when interpreting bytecode Tracking ticket: #23056 MR: !10399 This adds the flag `-funoptimized-core-for-interpreter`, permitting use of the `-O` flag to enable optimizations when compiling with the interpreter backend, like in ghci. - - - - - c6cf9433 by Ben Gamari at 2023-05-12T06:12:14-04:00 hadrian: Fix mention of non-existent removeFiles function Previously Hadrian's bindist Makefile referred to a `removeFiles` function that was previously defined by the `make` build system. Since the `make` build system is no longer around, this function is now undefined. Naturally, make being make, this appears to be silently ignored instead of producing an error. Fix this by rewriting it to `rm -f`. Closes #23373. - - - - - eb60ec18 by Andrew Lelechenko at 2023-05-12T06:12:54-04:00 Mention new implementation of GHC.IORef.atomicSwapIORef in the changelog - - - - - aa84cff4 by Teo Camarasu at 2023-05-12T19:27:23-04:00 rts: Ensure non-moving gc is not running when pausing - - - - - 5ad776ab by Teo Camarasu at 2023-05-12T19:27:23-04:00 rts: Teach listAllBlocks about nonmoving heap List all blocks on the non-moving heap. Resolves #22627 - - - - - d683b2e5 by Krzysztof Gogolewski at 2023-05-12T19:28:00-04:00 Fix coercion optimisation for SelCo (#23362) setNominalRole_maybe is supposed to output a nominal coercion. In the SelCo case, it was not updating the stored role to Nominal, causing #23362. - - - - - 59aa4676 by Alexis King at 2023-05-12T19:28:47-04:00 hadrian: Fix linker script flag for MergeObjects builder This fixes what appears to have been a typo in !9530. The `-t` flag just enables tracing on all versions of `ld` I’ve looked at, while `-T` is used to specify a linker script. It seems that this worked anyway for some reason on some `ld` implementations (perhaps because they automatically detect linker scripts), but the missing `-T` argument causes `gold` to complain. - - - - - 4bf9fa0f by Adam Gundry at 2023-05-12T23:49:49-04:00 Less coercion optimization for non-newtype axioms See Note [Push transitivity inside newtype axioms only] for an explanation of the change here. This change substantially improves the performance of coercion optimization for programs involving transitive type family reductions. ------------------------- Metric Decrease: CoOpt_Singletons LargeRecord T12227 T12545 T13386 T15703 T5030 T8095 ------------------------- - - - - - dc0c9574 by Adam Gundry at 2023-05-12T23:49:49-04:00 Move checkAxInstCo to GHC.Core.Lint A consequence of the previous change is that checkAxInstCo is no longer called during coercion optimization, so it can be moved back where it belongs. Also includes some edits to Note [Conflict checking with AxiomInstCo] as suggested by @simonpj. - - - - - 8b9b7dbc by Simon Peyton Jones at 2023-05-12T23:50:25-04:00 Use the eager unifier in the constraint solver This patch continues the refactoring of the constraint solver described in #23070. The Big Deal in this patch is to call the regular, eager unifier from the constraint solver, when we want to create new equalities. This replaces the existing, unifyWanted which amounted to yet-another-unifier, so it reduces duplication of a rather subtle piece of technology. See * Note [The eager unifier] in GHC.Tc.Utils.Unify * GHC.Tc.Solver.Monad.wrapUnifierTcS I did lots of other refactoring along the way * I simplified the treatment of right hand sides that contain CoercionHoles. Now, a constraint that contains a hetero-kind CoercionHole is non-canonical, and cannot be used for rewriting or unification alike. This required me to add the ch_hertero_kind flag to CoercionHole, with consequent knock-on effects. See wrinkle (2) of `Note [Equalities with incompatible kinds]` in GHC.Tc.Solver.Equality. * I refactored the StopOrContinue type to add StartAgain, so that after a fundep improvement (for example) we can simply start the pipeline again. * I got rid of the unpleasant (and inefficient) rewriterSetFromType/Co functions. With Richard I concluded that they are never needed. * I discovered Wrinkle (W1) in Note [Wanteds rewrite Wanteds] in GHC.Tc.Types.Constraint, and therefore now prioritise non-rewritten equalities. Quite a few error messages change, I think always for the better. Compiler runtime stays about the same, with one outlier: a 17% improvement in T17836 Metric Decrease: T17836 T18223 - - - - - 5cad28e7 by Bartłomiej Cieślar at 2023-05-12T23:51:06-04:00 Cleanup of dynflags override in export renaming The deprecation warnings are normally emitted whenever the name's GRE is being looked up, which calls the GHC.Rename.Env.addUsedGRE function. We do not want those warnings to be emitted when renaming export lists, so they are artificially turned off by removing all warning categories from DynFlags at the beginning of GHC.Tc.Gen.Export.rnExports. This commit removes that dependency by unifying the function used for GRE lookup in lookup_ie to lookupGreAvailRn and disabling the call to addUsedGRE in said function (the warnings are also disabled in a call to lookupSubBndrOcc_helper in lookupChildrenExport), as per #17957. This commit also changes the setting for whether to warn about deprecated names in addUsedGREs to be an explicit enum instead of a boolean. - - - - - d85ed900 by Alexis King at 2023-05-13T08:45:18-04:00 Use a uniform return convention in bytecode for unary results fixes #22958 - - - - - 8a0d45f7 by Andrew Lelechenko at 2023-05-13T08:45:58-04:00 Add more instances for Compose: Enum, Bounded, Num, Real, Integral See https://github.com/haskell/core-libraries-committee/issues/160 for discussion - - - - - 902f0730 by Simon Peyton Jones at 2023-05-13T14:58:34-04:00 Make GHC.Types.Id.Make.shouldUnpackTy a bit more clever As #23307, GHC.Types.Id.Make.shouldUnpackTy was leaving money on the table, failing to unpack arguments that are perfectly unpackable. The fix is pretty easy; see Note [Recursive unboxing] - - - - - a5451438 by sheaf at 2023-05-13T14:59:13-04:00 Fix bad multiplicity role in tyConAppFunCo_maybe The function tyConAppFunCo_maybe produces a multiplicity coercion for the multiplicity argument of the function arrow, except that it could be at the wrong role if asked to produce a representational coercion. We fix this by using the 'funRole' function, which computes the right roles for arguments to the function arrow TyCon. Fixes #23386 - - - - - 5b9e9300 by sheaf at 2023-05-15T11:26:59-04:00 Turn "ambiguous import" error into a panic This error should never occur, as a lookup of a type or data constructor should never be ambiguous. This is because a single module cannot export multiple Names with the same OccName, as per item (1) of Note [Exporting duplicate declarations] in GHC.Tc.Gen.Export. This code path was intended to handle duplicate record fields, but the rest of the code had since been refactored to handle those in a different way. We also remove the AmbiguousImport constructor of IELookupError, as it is no longer used. Fixes #23302 - - - - - e305e60c by M Farkas-Dyck at 2023-05-15T11:27:41-04:00 Unbreak some tests with latest GNU grep, which now warns about stray '\'. Confusingly, the testsuite mangled the error to say "stray /". We also migrate some tests from grep to grep -E, as it seems the author actually wanted an "POSIX extended" (a.k.a. sane) regex. Background: POSIX specifies 2 "regex" syntaxen: "basic" and "extended". Of these, only "extended" syntax is actually a regular expression. Furthermore, "basic" syntax is inconsistent in its use of the '\' character — sometimes it escapes a regex metacharacter, but sometimes it unescapes it, i.e. it makes an otherwise normal character become a metacharacter. This baffles me and it seems also the authors of these tests. Also, the regex(7) man page (at least on Linux) says "basic" syntax is obsolete. Nearly all modern tools and libraries are consistent in this use of the '\' character (of which many use "extended" syntax by default). - - - - - 10519e3d by Finley McIlwaine at 2023-05-15T12:40:48-06:00 Force HIE file path Removes a potential retainer of `ModSummary`s - - - - - 5ae81842 by sheaf at 2023-05-15T14:49:17-04:00 Improve "ambiguous occurrence" error messages This error was sometimes a bit confusing, especially when data families were involved. This commit improves the general presentation of the "ambiguous occurrence" error, and adds a bit of extra context in the case of data families. Fixes #23301 - - - - - 2f571afe by Sylvain Henry at 2023-05-15T14:50:07-04:00 Fix GHCJS OS platform (fix #23346) - - - - - 86aae570 by Oleg Grenrus at 2023-05-15T14:50:43-04:00 Split DynFlags structure into own module This will allow to make command line parsing to depend on diagnostic system (which depends on dynflags) - - - - - 1e4a6ec6 by Finley McIlwaine at 2023-05-15T14:20:34-06:00 Re-add index-state, with-compiler, delete hie.yamls - - - - - a2363fe9 by Hécate Moonlight at 2023-05-15T22:45:16+02:00 Merge pull request haskell/haddock#1594 from FinleyMcIlwaine/finley/ghc-9.6-mem-fixes Reduce memory usage - - - - - fbe3fe00 by Josh Meredith at 2023-05-15T18:01:43-04:00 Replace the implementation of CodeBuffers with unboxed types - - - - - 21f3aae7 by Josh Meredith at 2023-05-15T18:01:43-04:00 Use unboxed codebuffers in base Metric Decrease: encodingAllocations - - - - - 18ea2295 by Ben Gamari at 2023-05-15T18:02:20-04:00 rts: Weak pointer cleanups Various stylistic cleanups. No functional changes. - - - - - c343112f by Ben Gamari at 2023-05-15T18:02:20-04:00 rts: Don't force debug output to stderr Previously `+RTS -Dw -l` would emit debug output to the eventlog while `+RTS -l -Dw` would emit it to stderr. This was because the parser for `-D` would unconditionally override the debug output target. Now we instead only do so if no it is currently `TRACE_NONE`. - - - - - a5f5f067 by Ben Gamari at 2023-05-15T18:02:20-04:00 rts: Forcibly flush eventlog on barf Previously we would attempt to flush via `endEventLogging` which can easily deadlock, e.g., if `barf` fails during GC. Using `flushEventLog` directly may result in slightly less consistent eventlog output (since we don't take all capabilities before flushing) but avoids deadlocking. - - - - - 73b1e87c by Ben Gamari at 2023-05-15T18:02:20-04:00 rts: Assert that pointers aren't cleared by -DZ This turns many segmentation faults into much easier-to-debug assertion failures by ensuring that LOOKS_LIKE_*_PTR checks recognize bit-patterns produced by `+RTS -DZ` clearing as invalid pointers. This is a bit ad-hoc but this is the debug runtime. - - - - - 37fb61d8 by Ben Gamari at 2023-05-15T18:02:20-04:00 rts: Introduce printGlobalThreads - - - - - 451d65a6 by Ben Gamari at 2023-05-15T18:02:20-04:00 rts: Don't sanity-check StgTSO.global_link See Note [Avoid dangling global_link pointers]. Fixes #19146. - - - - - d69cbd78 by sheaf at 2023-05-15T18:03:00-04:00 Split up tyThingToIfaceDecl from GHC.Iface.Make This commit moves tyThingToIfaceDecl and coAxiomToIfaceDecl from GHC.Iface.Make into GHC.Iface.Decl. This avoids GHC.Types.TyThing.Ppr, which needs tyThingToIfaceDecl, transitively depending on e.g. GHC.Iface.Load and GHC.Tc.Utils.Monad. - - - - - 4d29ecdf by sheaf at 2023-05-15T18:03:00-04:00 Migrate errors to diagnostics in GHC.Tc.Module This commit migrates the errors in GHC.Tc.Module to use the new diagnostic infrastructure. It required a significant overhaul of the compatibility checks between an hs-boot or signature module and its implementation; we now use a Writer monad to accumulate errors; see the BootMismatch datatype in GHC.Tc.Errors.Types, with its panoply of subtypes. For the sake of readability, several local functions inside the 'checkBootTyCon' function were split off into top-level functions. We split off GHC.Types.HscSource into a "boot or sig" vs "normal hs file" datatype, as this mirrors the logic in several other places where we want to treat hs-boot and hsig files in a similar fashion. This commit also refactors the Backpack checks for type synonyms implementing abstract data, to correctly reject implementations that contain qualified or quantified types (this fixes #23342 and #23344). - - - - - d986c98e by Rodrigo Mesquita at 2023-05-16T00:14:04-04:00 configure: Drop unused AC_PROG_CPP In configure, we were calling `AC_PROG_CPP` but never making use of the $CPP variable it sets or reads. The issue is $CPP will show up in the --help output of configure, falsely advertising a configuration option that does nothing. The reason we don't use the $CPP variable is because HS_CPP_CMD is expected to be a single command (without flags), but AC_PROG_CPP, when CPP is unset, will set said variable to something like `/usr/bin/gcc -E`. Instead, we configure HS_CPP_CMD through $CC. - - - - - a8f0435f by Cheng Shao at 2023-05-16T00:14:42-04:00 rts: fix --disable-large-address-space This patch moves ACQUIRE_ALLOC_BLOCK_SPIN_LOCK/RELEASE_ALLOC_BLOCK_SPIN_LOCK from Storage.h to HeapAlloc.h. When --disable-large-address-space is passed to configure, the code in HeapAlloc.h makes use of these two macros. Fixes #23385. - - - - - bdb93cd2 by Oleg Grenrus at 2023-05-16T07:59:21+03:00 Add -Wmissing-role-annotations Implements #22702 - - - - - 41ecfc34 by Ben Gamari at 2023-05-16T07:28:15-04:00 base: Export {get,set}ExceptionFinalizer from System.Mem.Weak As proposed in CLC Proposal #126 [1]. [1]: https://github.com/haskell/core-libraries-committee/issues/126 - - - - - 67330303 by Ben Gamari at 2023-05-16T07:28:16-04:00 base: Introduce printToHandleFinalizerExceptionHandler - - - - - 5e3f9bb5 by Josh Meredith at 2023-05-16T13:59:22-04:00 JS: Implement h$clock_gettime in the JavaScript RTS (#23360) - - - - - 90e69d5d by Zubin Duggal at 2023-05-16T14:00:00-04:00 compiler: Use compact representation for SourceText SourceText is serialized along with INLINE pragmas into interface files. Many of these SourceTexts are identical, for example "{-# INLINE#". When deserialized, each such SourceText was previously expanded out into a [Char], which is highly wasteful of memory, and each such instance of the text would allocate an independent list with its contents as deserializing breaks any sharing that might have existed. Instead, we use a `FastString` to represent these, so that each instance unique text will be interned and stored in a memory efficient manner. - - - - - b70bc690 by Zubin Duggal at 2023-05-16T14:00:00-04:00 compiler: Use compact representation/FastStrings for `SourceNote`s `SourceNote`s should not be stored as [Char] as this is highly wasteful and in certain scenarios can be highly duplicated. Metric Decrease: hard_hole_fits - - - - - 6231a126 by Zubin Duggal at 2023-05-16T14:00:00-04:00 compiler: Use compact representation for UsageFile (#22744) Use FastString to store filepaths in interface files, as this data is highly redundant so we want to share all instances of filepaths in the compiler session. - - - - - 47a58150 by Zubin Duggal at 2023-05-16T14:00:00-04:00 testsuite: add test for T22744 This test checks for #22744 by compiling 100 modules which each have a dependency on 1000 distinct external files. Previously, when loading these interfaces from disk, each individual instance of a filepath in the interface will would be allocated as an individual object on the heap, meaning we have heap objects for 100*1000 files, when there are only 1000 distinct files we care about. This test checks this by first compiling the module normally, then measuring the peak memory usage in a no-op recompile, as the recompilation checking will force the allocation of all these filepaths. - - - - - 0451bdc9 by Ben Gamari at 2023-05-16T21:31:40-04:00 users guide: Add glossary Currently this merely explains the meaning of "technology preview" in the context of released features. - - - - - 0ba52e4e by Ben Gamari at 2023-05-16T21:31:40-04:00 Update glossary.rst - - - - - 3d23060c by Ben Gamari at 2023-05-16T21:31:40-04:00 Use glossary directive - - - - - 2972fd66 by Sylvain Henry at 2023-05-16T21:32:20-04:00 JS: fix getpid (fix #23399) - - - - - e8a78383 by Finley McIlwaine at 2023-05-17T12:19:16-06:00 Merge branch ghc-9.6 into ghc-head - - - - - 22e25581 by Finley McIlwaine at 2023-05-17T12:20:23-06:00 Merge branch 'ghc-head' of gitlab.haskell.org:ghc/haddock into ghc-head - - - - - 5fe1d3e6 by Matthew Pickering at 2023-05-17T21:42:00-04:00 Use setSrcSpan rather than setLclEnv in solveForAll In subsequent MRs (#23409) we want to remove the TcLclEnv argument from a CtLoc. This MR prepares us for that by removing the one place where the entire TcLclEnv is used, by using it more precisely to just set the contexts source location. Fixes #23390 - - - - - 385edb65 by Torsten Schmits at 2023-05-17T21:42:40-04:00 Update the users guide paragraph on -O in GHCi In relation to #23056 - - - - - 87626ef0 by Ben Gamari at 2023-05-18T15:18:53-04:00 base: Add test for #13660 - - - - - 9eef53b1 by Ben Gamari at 2023-05-18T15:18:53-04:00 base: Move implementation of GHC.Foreign to GHC.Internal - - - - - 174ea2fa by Ben Gamari at 2023-05-18T15:18:53-04:00 base: Introduce {new,with}CStringLen0 These are useful helpers for implementing the internal-NUL code unit check needed to fix #13660. - - - - - a46ced16 by Ben Gamari at 2023-05-18T15:18:53-04:00 base: Clean up documentation - - - - - b98d99cc by Ben Gamari at 2023-05-18T15:18:53-04:00 base: Ensure that FilePaths don't contain NULs POSIX filepaths may not contain the NUL octet but previously we did not reject such paths. This could be exploited by untrusted input to cause discrepancies between various `FilePath` queries and the opened filename. For instance, `readFile "hello.so\x00.txt"` would open the file `"hello.so"` yet `takeFileExtension` would return `".txt"`. The same argument applies to Windows FilePaths Fixes #13660. - - - - - 7ae45459 by Simon Peyton Jones at 2023-05-18T15:19:29-04:00 Allow the demand analyser to unpack tuple and equality dictionaries Addresses #23398. The demand analyser usually does not unpack class dictionaries: see Note [Do not unbox class dictionaries] in GHC.Core.Opt.DmdAnal. This patch makes an exception for tuple dictionaries and equality dictionaries, for reasons explained in wrinkles (DNB1) and (DNB2) of the above Note. Compile times fall by 0.1% for some reason (max 0.7% on T18698b). - - - - - b53a9086 by Greg Steuck at 2023-05-18T15:20:08-04:00 Use a simpler and more portable construct in ld.ldd check printf '%q\n' is a bash extension which led to incorrectly failing an ld.lld test on OpenBSD which uses pdksh as /bin/sh - - - - - dd5710af by Torsten Schmits at 2023-05-18T15:20:50-04:00 Update the warning about interpreter optimizations to reflect that they're not incompatible anymore, but guarded by a flag - - - - - 4f6dd999 by Matthew Pickering at 2023-05-18T15:21:26-04:00 Remove stray dump flags in GHC.Rename.Names - - - - - 4bca0486 by Oleg Grenrus at 2023-05-19T11:51:33+03:00 Make Warn = Located DriverMessage This change makes command line argument parsing use diagnostic framework for producing warnings. - - - - - 525ed554 by Simon Peyton Jones at 2023-05-19T10:09:15-04:00 Type inference for data family newtype instances This patch addresses #23408, a tricky case with data family newtype instances. Consider type family TF a where TF Char = Bool data family DF a newtype instance DF Bool = MkDF Int and [W] Int ~R# DF (TF a), with a Given (a ~# Char). We must fully rewrite the Wanted so the tpye family can fire; that wasn't happening. - - - - - c6fb6690 by Peter Trommler at 2023-05-20T03:16:08-04:00 testsuite: fix predicate on rdynamic test Test rdynamic requires dynamic linking support, which is orthogonal to RTS linker support. Change the predicate accordingly. Fixes #23316 - - - - - 735d504e by Matthew Pickering at 2023-05-20T03:16:44-04:00 docs: Use ghc-ticket directive where appropiate in users guide Using the directive automatically formats and links the ticket appropiately. - - - - - b56d7379 by Sylvain Henry at 2023-05-22T14:21:22-04:00 NCG: remove useless .align directive (#20758) - - - - - 15b93d2f by Simon Peyton Jones at 2023-05-22T14:21:58-04:00 Add test for #23156 This program had exponential typechecking time in GHC 9.4 and 9.6 - - - - - 2b53f206 by Greg Steuck at 2023-05-22T20:23:11-04:00 Revert "Change hostSupportsRPaths to report False on OpenBSD" This reverts commit 1e0d8fdb55a38ece34fa6cf214e1d2d46f5f5bf2. - - - - - 882e43b7 by Greg Steuck at 2023-05-22T20:23:11-04:00 Disable T17414 on OpenBSD Like on other systems it's not guaranteed that there's sufficient space in /tmp to write 2G out. - - - - - 9d531f9a by Greg Steuck at 2023-05-22T20:23:11-04:00 Bring back getExecutablePath to getBaseDir on OpenBSD Fix #18173 - - - - - 9db0eadd by Krzysztof Gogolewski at 2023-05-22T20:23:47-04:00 Add an error origin for impedance matching (#23427) - - - - - 33cf4659 by Ben Gamari at 2023-05-23T03:46:20-04:00 testsuite: Add tests for #23146 Both lifted and unlifted variants. - - - - - 76727617 by Ben Gamari at 2023-05-23T03:46:21-04:00 codeGen: Fix some Haddocks - - - - - 33a8c348 by Ben Gamari at 2023-05-23T03:46:21-04:00 codeGen: Give proper LFInfo to datacon wrappers As noted in `Note [Conveying CAF-info and LFInfo between modules]`, when importing a binding from another module we must ensure that it gets the appropriate `LambdaFormInfo` if it is in WHNF to ensure that references to it are tagged correctly. However, the implementation responsible for doing this, `GHC.StgToCmm.Closure.mkLFImported`, only dealt with datacon workers and not wrappers. This lead to the crash of this program in #23146: module B where type NP :: [UnliftedType] -> UnliftedType data NP xs where UNil :: NP '[] module A where import B fieldsSam :: NP xs -> NP xs -> Bool fieldsSam UNil UNil = True x = fieldsSam UNil UNil Due to its GADT nature, `UNil` produces a trivial wrapper $WUNil :: NP '[] $WUNil = UNil @'[] @~(<co:1>) which is referenced in the RHS of `A.x`. Due to the above-mentioned bug in `mkLFImported`, the references to `$WUNil` passed to `fieldsSam` were not tagged. This is problematic as `fieldsSam` expected its arguments to be tagged as they are unlifted. The fix is straightforward: extend the logic in `mkLFImported` to cover (nullary) datacon wrappers as well as workers. This is safe because we know that the wrapper of a nullary datacon will be in WHNF, even if it includes equalities evidence (since such equalities are not runtime relevant). Thanks to @MangoIV for the great ticket and @alt-romes for his minimization and help debugging. Fixes #23146. - - - - - 2fc18e9e by Rodrigo Mesquita at 2023-05-23T03:46:21-04:00 codeGen: Fix LFInfo of imported datacon wrappers As noted in #23231 and in the previous commit, we were failing to give a an LFInfo of LFCon to a nullary datacon wrapper from another module, failing to properly tag pointers which ultimately led to the segmentation fault in #23146. On top of the previous commit which now considers wrappers where we previously only considered workers, we change the order of the guards so that we check for the arity of the binding before we check whether it is a constructor. This allows us to (1) Correctly assign `LFReEntrant` to imported wrappers whose worker was nullary, which we previously would fail to do (2) Remove the `isNullaryRepDataCon` predicate: (a) which was previously wrong, since it considered wrappers whose workers had zero-width arguments to be non-nullary and would fail to give `LFCon` to them (b) is now unnecessary, since arity == 0 guarantees - that the worker takes no arguments at all - and the wrapper takes no arguments and its RHS must be an application of the worker to zero-width-args only. - we lint these two items with an assertion that the datacon `hasNoNonZeroWidthArgs` We also update `isTagged` to use the new logic in determining the LFInfos of imported Ids. The creation of LFInfos for imported Ids and this detail are explained in Note [The LFInfo of Imported Ids]. Note that before the patch to those issues we would already consider these nullary wrappers to have `LFCon` lambda form info; but failed to re-construct that information in `mkLFImported` Closes #23231, #23146 (I've additionally batched some fixes to documentation I found while investigating this issue) - - - - - 0598f7f0 by Rodrigo Mesquita at 2023-05-23T03:46:21-04:00 Make LFInfos for DataCons on construction As a result of the discussion in !10165, we decided to amend the previous commit which fixed the logic of `mkLFImported` with regard to datacon workers and wrappers. Instead of having the logic for the LFInfo of datacons be in `mkLFImported`, we now construct an LFInfo for all data constructors on GHC.Types.Id.Make and store it in the `lfInfo` field. See the new Note [LFInfo of DataCon workers and wrappers] and ammendments to Note [The LFInfo of Imported Ids] - - - - - 12294b22 by Rodrigo Mesquita at 2023-05-23T03:46:21-04:00 Update Note [Core letrec invariant] Authored by @simonpj - - - - - e93ab972 by Rodrigo Mesquita at 2023-05-23T03:46:21-04:00 Rename mkLFImported to importedIdLFInfo The `mkLFImported` sounded too much like a constructor of sorts, when really it got the `LFInfo` of an imported Id from its `lf_info` field when this existed, and otherwise returned a conservative estimate of that imported Id's LFInfo. This in contrast to functions such as `mkLFReEntrant` which really are about constructing an `LFInfo`. - - - - - e54d9259 by Rodrigo Mesquita at 2023-05-23T03:46:21-04:00 Enforce invariant on typePrimRepArgs in the types As part of the documentation effort in !10165 I came across this invariant on 'typePrimRepArgs' which is easily expressed at the type-level through a NonEmpty list. It allowed us to remove one panic. - - - - - b8fe6a0c by Rodrigo Mesquita at 2023-05-23T03:46:21-04:00 Merge outdated Note [Data con representation] into Note [Data constructor representation] Introduce new Note [Constructor applications in STG] to better support the merge, and reference it from the relevant bits in the STG syntax. - - - - - e1590ddc by Simon Peyton Jones at 2023-05-23T03:46:57-04:00 Add the SolverStage monad This refactoring makes a substantial improvement in the structure of the type-checker's constraint solver: #23070. Specifically: * Introduced the SolverStage monad. See GHC.Tc.Solver.Monad Note [The SolverStage monad] * Make each solver pipeline (equalities, dictionaries, irreds etc) deal with updating the inert set, as a separate SolverStage. There is sometimes special stuff to do, and it means that each full pipeline can have type SolverStage Void, indicating that they never return anything. * Made GHC.Tc.Solver.Equality.zonkEqTypes into a SolverStage. Much nicer. * Combined the remnants of GHC.Tc.Solver.Canonical and GHC.Tc.Solver.Interact into a new module GHC.Tc.Solver.Solve. (Interact and Canonical are removed.) * Gave the same treatment to dictionary and irred constraints as I have already done for equality constraints: * New types (akin to EqCt): IrredCt and DictCt * Ct is now just a simple sum type data Ct = CDictCan DictCt | CIrredCan IrredCt | CEqCan EqCt | CQuantCan QCInst | CNonCanonical CtEvidence * inert_dicts can now have the better type DictMap DictCt, instead of DictMap Ct; and similarly inert_irreds. * Significantly simplified the treatment of implicit parameters. Previously we had a number of special cases * interactGivenIP, an entire function * special case in maybeKickOut * special case in findDict, when looking up dictionaries But actually it's simpler than that. When adding a new Given, implicit parameter constraint to the InertSet, we just need to kick out any existing inert constraints that mention that implicit parameter. The main work is done in GHC.Tc.Solver.InertSet.delIPDict, along with its auxiliary GHC.Core.Predicate.mentionsIP. See Note [Shadowing of implicit parameters] in GHC.Tc.Solver.Dict. * Add a new fast-path in GHC.Tc.Errors.Hole.tcCheckHoleFit. See Note [Fast path for tcCheckHoleFit]. This is a big win in some cases: test hard_hole_fits gets nearly 40% faster (at compile time). * Add a new fast-path for solving /boxed/ equality constraints (t1 ~ t2). See Note [Solving equality classes] in GHC.Tc.Solver.Dict. This makes a big difference too: test T17836 compiles 40% faster. * Implement the PermissivePlan of #23413, which concerns what happens with insoluble Givens. Our previous treatment was wildly inconsistent as that ticket pointed out. A part of this, I simplified GHC.Tc.Validity.checkAmbiguity: now we simply don't run the ambiguity check at all if -XAllowAmbiguousTypes is on. Smaller points: * In `GHC.Tc.Errors.misMatchOrCND` instead of having a special case for insoluble /occurs/ checks, broaden in to all insouluble constraints. Just generally better. See Note [Insoluble mis-match] in that module. As noted above, compile time perf gets better. Here are the changes over 0.5% on Fedora. (The figures are slightly larger on Windows for some reason.) Metrics: compile_time/bytes allocated ------------------------------------- LargeRecord(normal) -0.9% MultiLayerModulesTH_OneShot(normal) +0.5% T11822(normal) -0.6% T12227(normal) -1.8% GOOD T12545(normal) -0.5% T13035(normal) -0.6% T15703(normal) -1.4% GOOD T16875(normal) -0.5% T17836(normal) -40.7% GOOD T17836b(normal) -12.3% GOOD T17977b(normal) -0.5% T5837(normal) -1.1% T8095(normal) -2.7% GOOD T9020(optasm) -1.1% hard_hole_fits(normal) -37.0% GOOD geo. mean -1.3% minimum -40.7% maximum +0.5% Metric Decrease: T12227 T15703 T17836 T17836b T8095 hard_hole_fits LargeRecord T9198 T13035 - - - - - 6abf3648 by Simon Peyton Jones at 2023-05-23T03:46:57-04:00 Avoid an assertion failure in abstractFloats The function GHC.Core.Opt.Simplify.Utils.abstractFloats was carelessly calling lookupIdSubst_maybe on a CoVar; but a precondition of the latter is being given an Id. In fact it's harmless to call it on a CoVar, but still, the precondition on lookupIdSubst_maybe makes sense, so I added a test for CoVars. This avoids a crash in a DEBUG compiler, but otherwise has no effect. Fixes #23426. - - - - - 41bbf0df by Bartłomiej Cieślar at 2023-05-24T08:57:58+02:00 changes to the WarningTxt cases Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - 838aaf4b by hainq at 2023-05-24T12:41:19-04:00 Migrate errors in GHC.Tc.Validity This patch migrates the error messages in GHC.Tc.Validity to use the new diagnostic infrastructure. It adds the constructors: - TcRnSimplifiableConstraint - TcRnArityMismatch - TcRnIllegalInstanceDecl, with sub-datatypes for HasField errors and fundep coverage condition errors. - - - - - 8539764b by Krzysztof Gogolewski at 2023-05-24T12:41:56-04:00 linear lint: Add missing processing of DEFAULT In this correct program f :: a %1 -> a f x = case x of x { _DEFAULT -> x } after checking the alternative we weren't popping the case binder 'x' from the usage environment, which meant that the lambda-bound 'x' was counted twice: in the scrutinee and (incorrectly) in the alternative. In fact, we weren't checking the usage of 'x' at all. Now the code for handling _DEFAULT is similar to the one handling data constructors. Fixes #23025. - - - - - ae683454 by Matthew Pickering at 2023-05-24T12:42:32-04:00 Remove outdated "Don't check hs-boot type family instances too early" note This note was introduced in 25b70a29f623 which delayed performing some consistency checks for type families. However, the change was reverted later in 6998772043a7f0b0360116eb5ffcbaa5630b21fb but the note was not removed. I found it confusing when reading to code to try and work out what special behaviour there was for hs-boot files (when in-fact there isn't any). - - - - - 44af57de by Matthew Pickering at 2023-05-24T12:43:08-04:00 rts: Define ticky macro stubs These macros have long been undefined which has meant we were missing reporting these allocations in ticky profiles. The most critical missing definition was TICK_ALLOC_HEAP_NOCTR which was missing all the RTS calls to allocate, this leads to a the overall ALLOC_RTS_tot number to be severaly underreported. Of particular interest though is the ALLOC_STACK_ctr and ALLOC_STACK_tot counters which are useful to tracking stack allocations. Fixes #23421 - - - - - b2dabe3a by Matthew Pickering at 2023-05-24T12:43:08-04:00 rts: ticky: Rename TICK_ALLOC_HEAP_NOCTR to TICK_ALLOC_RTS This macro increments the ALLOC_HEAP_tot and ALLOC_HEAP_ctr so it makes more sense to name it after that rather than the suffix NOCTR, whose meaning has been lost to the mists of time. - - - - - eac4420a by Ben Gamari at 2023-05-24T12:43:45-04:00 users guide: A few small mark-up fixes - - - - - a320ca76 by Rodrigo Mesquita at 2023-05-24T12:44:20-04:00 configure: Fix support check for response files. In failing to escape the '-o' in '-o\nconftest\nconftest.o\n' argument to printf, the writing of the arguments response file always failed. The fix is to pass the arguments after `--` so that they are treated positional arguments rather than flags to printf. Closes #23435 - - - - - f21ce0e4 by mangoiv at 2023-05-24T12:45:00-04:00 [feat] add .direnv to the .gitignore file - - - - - 36d5944d by Andrew Lelechenko at 2023-05-24T20:58:34-04:00 Add Data.List.unsnoc See https://github.com/haskell/core-libraries-committee/issues/165 for discussion - - - - - c0f2f9e3 by Bartłomiej Cieślar at 2023-05-24T20:59:14-04:00 Fix crash in backpack signature merging with -ddump-rn-trace In some cases, backpack signature merging could crash in addUsedGRE when -ddump-rn-trace was enabled, as pretty-printing the GREInfo would cause unavailable interfaces to be loaded. This commit fixes that issue by not pretty-printing the GREInfo in addUsedGRE when -ddump-rn-trace is enabled. Fixes #23424 Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - 5a07d94a by Krzysztof Gogolewski at 2023-05-25T03:30:20-04:00 Add a regression test for #13981 The panic was fixed by 6998772043a7f0b. Fixes #13981. - - - - - 182df90e by Krzysztof Gogolewski at 2023-05-25T03:30:57-04:00 Add a test for #23355 It was fixed by !10061, so I'm adding it in the same group. - - - - - 1b31b039 by uhbif19 at 2023-05-25T12:08:28+02:00 Migrate errors in GHC.Rename.Splice GHC.Rename.Pat This commit migrates the errors in GHC.Rename.Splice and GHC.Rename.Pat to use the new diagnostic infrastructure. - - - - - 56abe494 by sheaf at 2023-05-25T12:09:55+02:00 Common up Template Haskell errors in TcRnMessage This commit commons up the various Template Haskell errors into a single constructor, TcRnTHError, of TcRnMessage. - - - - - a487ba9e by Krzysztof Gogolewski at 2023-05-25T14:35:56-04:00 Enable ghci tests for unboxed tuples The tests were originally skipped because ghci used not to support unboxed tuples/sums. - - - - - dc3422d4 by Matthew Pickering at 2023-05-25T18:57:19-04:00 rts: Build ticky GHC with single-threaded RTS The threaded RTS allows you to use ticky profiling but only for the counters in the generated code. The counters used in the C portion of the RTS are disabled. Updating the counters is also racy using the threaded RTS which can lead to misleading or incorrect ticky results. Therefore we change the hadrian flavour to build using the single-threaded RTS (mainly in order to get accurate C code counter increments) Fixes #23430 - - - - - fbc8e04e by sheaf at 2023-05-25T18:58:00-04:00 Propagate long-distance info in generated code When desugaring generated pattern matches, we skip pattern match checks. However, this ended up also discarding long-distance information, which might be needed for user-written sub-expressions. Example: ```haskell okay (GADT di) cd = let sr_field :: () sr_field = case getFooBar di of { Foo -> () } in case cd of { SomeRec _ -> SomeRec sr_field } ``` With sr_field a generated FunBind, we still want to propagate the outer long-distance information from the GADT pattern match into the checks for the user-written RHS of sr_field. Fixes #23445 - - - - - f8ced241 by Matthew Pickering at 2023-05-26T15:26:21-04:00 Introduce GHCiMessage to wrap GhcMessage By introducing a wrapped message type we can control how certain messages are printed in GHCi (to add extra information for example) - - - - - 58e554c1 by Matthew Pickering at 2023-05-26T15:26:22-04:00 Generalise UnknownDiagnostic to allow embedded diagnostics to access parent diagnostic options. * Split default diagnostic options from Diagnostic class into HasDefaultDiagnosticOpts class. * Generalise UnknownDiagnostic to allow embedded diagnostics to access options. The principle idea here is that when wrapping an error message (such as GHCMessage to make GHCiMessage) then we need to also be able to lift the configuration when overriding how messages are printed (see load' for an example). - - - - - b112546a by Matthew Pickering at 2023-05-26T15:26:22-04:00 Allow API users to wrap error messages created during 'load' This allows API users to configure how messages are rendered when they are emitted from the load function. For an example see how 'loadWithCache' is used in GHCi. - - - - - 2e4cf0ee by Matthew Pickering at 2023-05-26T15:26:22-04:00 Abstract cantFindError and turn Opt_BuildingCabal into a print-time option * cantFindError is abstracted so that the parts which mention specific things about ghc/ghci are parameters. The intention being that GHC/GHCi can specify the right values to put here but otherwise display the same error message. * The BuildingCabalPackage argument from GenericMissing is removed and turned into a print-time option. The reason for the error is not dependent on whether `-fbuilding-cabal-package` is passed, so we don't want to store that in the error message. - - - - - 34b44f7d by Matthew Pickering at 2023-05-26T15:26:22-04:00 error messages: Don't display ghci specific hints for missing packages Tickets like #22884 suggest that it is confusing that GHC used on the command line can suggest options which only work in GHCi. This ticket uses the error message infrastructure to override certain error messages which displayed GHCi specific information so that this information is only showed when using GHCi. The main annoyance is that we mostly want to display errors in the same way as before, but with some additional information. This means that the error rendering code has to be exported from the Iface/Errors/Ppr.hs module. I am unsure about whether the approach taken here is the best or most maintainable solution. Fixes #22884 - - - - - 05a1b626 by Matthew Pickering at 2023-05-26T15:26:58-04:00 ghcup-metadata: Don't override existing metadata if version already exists. If a nightly pipeline runs twice for some reason for the same version then we really don't want to override an existing entry with new bindists. This could cause ABI compatability issues for users or break ghcup's caching logic. - - - - - fcbcb3cc by Matthew Pickering at 2023-05-26T15:26:58-04:00 ghcup-metadata: Use proper API url for bindist download Previously we were using links from the web interface, but it's more robust and future-proof to use the documented links to the artifacts. https://docs.gitlab.com/ee/api/job_artifacts.html - - - - - 5b59c8fe by Matthew Pickering at 2023-05-26T15:26:58-04:00 ghcup-metadata: Set Nightly and LatestNightly tags The latest nightly release needs the LatestNightly tag, and all other nightly releases need the Nightly tag. Therefore when the metadata is updated we need to replace all LatestNightly with Nightly.` - - - - - 914e1468 by Matthew Pickering at 2023-05-26T15:26:58-04:00 ghcup-metadata: Download nightly metadata for correct date The metadata now lives in https://gitlab.haskell.org/ghc/ghcup-metadata with one metadata file per year. When we update the metadata we download and update the right file for the current year. - - - - - 16cf7d2e by Matthew Pickering at 2023-05-26T15:26:58-04:00 ghcup-metadata: Download metadata and update for correct year something about pipeline date - - - - - 14792c4b by Matthew Pickering at 2023-05-26T15:26:58-04:00 ghcup-metadata: Don't skip CI On a push we now have a CI job which updates gitlab pages with the metadata files. - - - - - 1121bdd8 by Matthew Pickering at 2023-05-26T15:26:59-04:00 ghcup-metadata: Add --date flag to specify the release date The ghcup-metadata now has a viReleaseDay field which needs to be populated with the day of the release. - - - - - bc478bee by Matthew Pickering at 2023-05-26T15:26:59-04:00 ghcup-metadata: Add dlOutput field ghcup now requires us to add this field which specifies where it should download the bindist to. See https://gitlab.haskell.org/ghc/ghcup-metadata/-/issues/1 for some more discussion. - - - - - 2bdbd9da by Josh Meredith at 2023-05-26T15:27:35-04:00 JS: Convert rendering to use HLine instead of SDoc (#22455) - - - - - abd9e37c by Norman Ramsey at 2023-05-26T15:28:12-04:00 testsuite: add WasmControlFlow test This patch adds the WasmControlFlow test to test the wasm backend's relooper component. - - - - - 07f858eb by Sylvain Henry at 2023-05-26T15:28:53-04:00 Factorize getLinkDeps Prepare reuse of getLinkDeps for TH implementation in the JS backend (cf #22261 and review of !9779). - - - - - fad9d092 by Oleg Grenrus at 2023-05-27T13:38:08-04:00 Change GHC.Driver.Session import to .DynFlags Also move targetPlatform selector Plenty of GHC needs just DynFlags. Even more can be made to use .DynFlags if more selectors is migrated. This is a low hanging fruit. - - - - - 69fdbece by Alan Zimmerman at 2023-05-27T13:38:45-04:00 EPA: Better fix for #22919 The original fix for #22919 simply removed the ability to match up prior comments with the first declaration in the file. Restore it, but add a check that the comment is on a single line, by ensuring that it comes immediately prior to the next thing (comment or start of declaration), and that the token preceding it is not on the same line. closes #22919 - - - - - 0350b186 by Josh Meredith at 2023-05-29T12:46:27+00:00 Remove JavaScriptFFI from --supported-extensions for non-JS targets (#11214) - - - - - b4816919 by Matthew Pickering at 2023-05-30T17:07:43-04:00 testsuite: Pass -kb16k -kc128k for performance tests Setting a larger stack chunk size gives a greater protection from stack thrashing (where the repeated overflow/underflow allocates a lot of stack chunks which sigificantly impact allocations). This stabilises some tests against differences cause by more things being pushed onto the stack. The performance tests are generally testing work done by the compiler, using allocation as a proxy, so removing/stabilising the allocations due to the stack gives us more stable tests which are also more sensitive to actual changes in compiler performance. The tests which increase are ones where we compile a lot of modules, and for each module we spawn a thread to compile the module in. Therefore increasing these numbers has a multiplying effect on these tests because there are many more stacks which we can increase in size. The most significant improvements though are cases such as T8095 which reduce significantly in allocations (30%). This isn't a performance improvement really but just helps stabilise the test against this threshold set by the defaults. Fixes #23439 ------------------------- Metric Decrease: InstanceMatching T14683 T8095 T9872b_defer T9872d T9961 hie002 T19695 T3064 Metric Increase: MultiLayerModules T13701 T14697 ------------------------- - - - - - 6629f1c5 by Ben Gamari at 2023-05-30T17:08:20-04:00 Move via-C flags into GHC These were previously hardcoded in configure (with no option for overriding them) and simply passed onto ghc through the settings file. Since configure already guarantees gcc supports those flags, we simply move them into GHC. - - - - - 981e5e11 by Arnaud Spiwack at 2023-05-31T08:34:33-04:00 Allow CPR on unrestricted constructors Per the new `Note [Linting linearity]`, we want optimisations over trying to preserve linearity. This will allow CPR to handle `Ur`, in particular. - - - - - bf9344d2 by Arnaud Spiwack at 2023-05-31T08:34:33-04:00 Push coercions across multiplicity boundaries Per the new `Note [Linting linearity]`, we want optimisations over trying to preserve linearity. This will avoid preventing inlinings and reductions and make linear programs more efficient. - - - - - d56dd695 by sheaf at 2023-05-31T11:37:12-04:00 Data.Bag: add INLINEABLE to polymorphic functions This commit allows polymorphic methods in GHC.Data.Bag to be specialised, avoiding having to pass explicit dictionaries when they are instantiated with e.g. a known monad. - - - - - 5366cd35 by sheaf at 2023-05-31T11:37:12-04:00 Split off TcBinderStack into its own module This commit splits off TcBinderStack into its own module, to avoid module cycles: we might want to refer to it without also pulling in the TcM monad. - - - - - 09d4d307 by sheaf at 2023-05-31T11:37:12-04:00 Split off TcRef into its own module This helps avoid pull in the full TcM monad when we just want access to mutable references in the typechecker. This facilitates later patches which introduce a slimmed down TcM monad for zonking. - - - - - 88cc19b3 by sheaf at 2023-05-31T11:37:12-04:00 Introduce Codensity monad The Codensity monad is useful to write state-passing computations in continuation-passing style, e.g. to implement a State monad as continuation-passing style over a Reader monad. - - - - - f62d8195 by sheaf at 2023-05-31T11:37:12-04:00 Restructure the zonker This commit splits up the zonker into a few separate components, described in Note [The structure of the zonker] in `GHC.Tc.Zonk.Type`. 1. `GHC.Tc.Zonk.Monad` introduces a pared-down `TcM` monad, `ZonkM`, which has enough information for zonking types. This allows us to refactor `ErrCtxt` to use `ZonkM` instead of `TcM`, which guarantees we don't throw an error while reporting an error. 2. `GHC.Tc.Zonk.Env` is the new home of `ZonkEnv`, and also defines two zonking monad transformers, `ZonkT` and `ZonkBndrT`. `ZonkT` is a reader monad transformer over `ZonkEnv`. `ZonkBndrT m` is the codensity monad over `ZonkT m`. `ZonkBndrT` is used for computations that accumulate binders in the `ZonkEnv`. 3. `GHC.Tc.Zonk.TcType` contains the code for zonking types, for use in the typechecker. It uses the `ZonkM` monad. 4. `GHC.Tc.Zonk.Type` contains the code for final zonking to `Type`, which has been refactored to use `ZonkTcM = ZonkT TcM` and `ZonkBndrTcM = ZonkBndrT TcM`. Allocations slightly decrease on the whole due to using continuation-passing style instead of manual state passing of ZonkEnv in the final zonking to Type. ------------------------- Metric Decrease: T4029 T8095 T14766 T15304 hard_hole_fits RecordUpdPerf Metric Increase: T10421 ------------------------- - - - - - 70526f5b by mimi.vx at 2023-05-31T11:37:53-04:00 Update rdt-theme to latest upstream version Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/23444 - - - - - f3556d6c by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Restructure IPE buffer layout Reference ticket #21766 This commit restructures IPE buffer list entries to not contain references to their corresponding info tables. IPE buffer list nodes now point to two lists of equal length, one holding the list of info table pointers and one holding the corresponding entries for each info table. This will allow the entry data to be compressed without losing the references to the info tables. - - - - - 5d1f2411 by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Add IPE compression to configure Reference ticket #21766 Adds an `--enable-ipe-data-compreesion` flag to the configure script which will check for libzstd and set the appropriate flags to allow for IPE data compression in the compiler - - - - - b7a640ac by Finley McIlwaine at 2023-06-01T04:53:12-04:00 IPE data compression Reference ticket #21766 When IPE data compression is enabled, compress the emitted IPE buffer entries and decompress them in the RTS. - - - - - 5aef5658 by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Fix libzstd detection in configure and RTS Ensure that `HAVE_LIBZSTD` gets defined to either 0 or 1 in all cases and properly check that before IPE data decompression in the RTS. See ticket #21766. - - - - - 69563c97 by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Add note describing IPE data compression See ticket #21766 - - - - - 7872e2b6 by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Fix byte order of IPE data, fix IPE tests Make sure byte order of written IPE buffer entries matches target. Make sure the IPE-related tests properly access the fields of IPE buffer entry nodes with the new IPE layout. This commit also introduces checks to avoid importing modules if IPE compression is not enabled. See ticket #21766. - - - - - 0e85099b by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Fix IPE data decompression buffer allocation Capacity of buffers allocated for decompressed IPE data was incorrect due to a misuse of the `ZSTD_findFrameCompressedSize` function. Fix by always storing decompressed size of IPE data in IPE buffer list nodes and using `ZSTD_findFrameCompressedSize` to determine the size of the compressed data. See ticket #21766 - - - - - a0048866 by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Add optional dependencies to ./configure output Changes the configure script to indicate whether libnuma, libzstd, or libdw are being used as dependencies due to their optional features being enabled. - - - - - 09d93bd0 by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Add IPE-enabled builds to CI - Adds an IPE job to the CI pipeline which is triggered by the ~IPE label - Introduces CI logic to enable IPE data compression - Enables uncompressed IPE data on debug CI job - Regenerates jobs.yaml MR https://gitlab.haskell.org/ghc/ci-images/-/merge_requests/112 on the images repository is meant to ensure that the proper images have libzstd-dev installed. - - - - - 3ded9a1c by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Update user's guide and release notes, small fixes Add mention of IPE data compression to user's guide and the release notes for 9.8.1. Also note the impact compression has on binary size in both places. Change IpeBufferListNode compression check so only the value `1` indicates compression. See ticket #21766 - - - - - 41b41577 by Finley McIlwaine at 2023-06-01T04:53:12-04:00 Remove IPE enabled builds from CI We don't need to explicitly specify the +ipe transformer to test IPE data since there are tests which manually enable IPE information. This commit does leave zstd IPE data compression enabled on the debian CI jobs. - - - - - 982bef3a by Krzysztof Gogolewski at 2023-06-01T04:53:49-04:00 Fix build with 9.2 GHC.Tc.Zonk.Type uses an equality constraint. ghc.nix currently provides 9.2. - - - - - 1c96bc3d by Krzysztof Gogolewski at 2023-06-01T10:56:11-04:00 Output Lint errors to stderr instead of stdout This is a continuation of 7b095b99, which fixed warnings but not errors. Refs #13342 - - - - - 8e81f140 by sheaf at 2023-06-01T10:56:51-04:00 Refactor lookupExactOrOrig & friends This refactors the panoply of renamer lookup functions relating to lookupExactOrOrig to more graciously handle Exact and Orig names. In particular, we avoid the situation in which we would add Exact/Orig GREs to the tcg_used_gres field, which could cause a panic in bestImport like in #23240. Fixes #23428 - - - - - 5d415bfd by Krzysztof Gogolewski at 2023-06-01T10:57:31-04:00 Use the one-shot trick for UM and RewriteM functors As described in Note [The one-shot state monad trick], we shouldn't use derived Functor instances for monads using one-shot. This was done for most of them, but UM and RewriteM were missed. - - - - - 2c38551e by Krzysztof Gogolewski at 2023-06-01T10:58:08-04:00 Fix testsuite skipping Lint setTestOpts() is used to modify the test options for an entire .T file, rather than a single test. If there was a test using collect_compiler_stats, all of the tests in the same file had lint disabled. Fixes #21247 - - - - - 00a1e50b by Krzysztof Gogolewski at 2023-06-01T10:58:44-04:00 Add testcases for already fixed #16432 They were fixed by 40c7daed0. Fixes #16432 - - - - - c686ba9b by Hécate Moonlight at 2023-06-01T14:03:02-06:00 Port the remains of Hi-Haddock - - - - - 9d8a85fd by Hécate Moonlight at 2023-06-01T14:03:06-06:00 Stdout for tests - - - - - 36331d07 by Finley McIlwaine at 2023-06-01T14:06:02-06:00 Formatting, organize imports - - - - - a06059b1 by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix empty context confusion in Convert module - - - - - 379346ae by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix associated type families in Hoogle output - - - - - fc6ea7ed by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix test refs Accept several changes in Hoogle tests: Pretty printing logic no longer prints the `(Proxy (Proxy (...))` chain in Bug806 with parentheses. Since this test was only meant to verify that line breaks do not occur, accept the change. `tyThingToLHsDecl` is called for class and data declarations, which ends up "synifying" the type variables and giving unlifted types kind signatures. As a result, type variables of kind `Type -> Type` are now printed with kind signatures in Hoogle output. This could be changed by manually drop kind signatures from class variables in the Hoogle backend if the behavior is deemed unacceptable. Sometimes subordinate declarations are exported separate from their parent declarations (e.g. record selectors). In this case, a type signature is cobbled together for the export item in `extractDecl`. Since this type signature is very manually constructed, it may lack kind signatures of decls constructed from `tyThingToLHsDecl`. An example of this is the `type-sigs` Hoogle test. Change `*` to `Type` in Hoogle test refs. I don't think this will break Hoogle behavior, since it appears to not consider type signatures in search. I have not fully verified this. - - - - - e14b7e58 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix LaTeX backend test refs Changes to GHC pretty printing code have resulted in some differences to Haddock's LaTeX output. - Type variables are printed explicitly quantified in the LinearTypes test - Wildcard types in type family equations are now printed numbered, e.g. _1 _2, in the TypeFamilies3 test - Combined signatures in DefaultSignatures test are now documented as separate signatures - - - - - 41b5b296 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and test source updates - Stop using kind `*` in html test sources - Add TypeOperators where necessary to avoid warnings and future errors - Rename some test modules to match their module names - - - - - c640e2a2 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix missing deprecation warnings on record fields `lookupOccEnv` was used to resolve `OccNames` with warnings attached, but it doesn't look in the record field namespace. Thus, if a record field had a warning attached, it would not resolve and the warning map would not include it. This commit fixes by using `lookupOccEnv_WithFields` instead. - - - - - fad0c462 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and some comments - - - - - 751fd023 by Finley McIlwaine at 2023-06-01T14:11:41-06:00 Accept HTML test diffs All diffs now boil down to the expected differences resulting from declarations being reified from TyThings in hi-haddock. Surface syntax now has much less control over the syntax used in the documentation. - - - - - d835c845 by Finley McIlwaine at 2023-06-01T14:11:45-06:00 Adapt to new load' type - - - - - dcf776c4 by Finley McIlwaine at 2023-06-01T14:13:13-06:00 Update mkWarningMap and moduleWarning - - - - - 8e8432fd by Finley McIlwaine at 2023-06-01T14:28:54-06:00 Revert load' changes - - - - - aeb2982c by Finley McIlwaine at 2023-06-01T14:40:24-06:00 Accept change to Instances test in html-test Link to Data.Tuple instead of GHC.Tuple.Prim - - - - - 8adfdbac by Finley McIlwaine at 2023-06-01T15:53:17-06:00 Reset ghc dep to ^>= 9.6 - - - - - f6e060cc by Krzysztof Gogolewski at 2023-06-02T09:07:25-04:00 cleanup: Remove unused field from SelfBoot It is no longer needed since Note [Extra dependencies from .hs-boot files] was deleted in 6998772043. I've also added tildes to Note headers, otherwise they're not detected by the linter. - - - - - 82eacab6 by sheaf at 2023-06-02T09:08:01-04:00 Delete GHC.Tc.Utils.Zonk This module was split up into GHC.Tc.Zonk.Type and GHC.Tc.Zonk.TcType in commit f62d8195, but I forgot to delete the original module - - - - - 4a4eb761 by Ben Gamari at 2023-06-02T23:53:21-04:00 base: Add build-order import of GHC.Types in GHC.IO.Handle.Types For reasons similar to those described in Note [Depend on GHC.Num.Integer]. Fixes #23411. - - - - - f53ac0ae by Sylvain Henry at 2023-06-02T23:54:01-04:00 JS: fix and enhance non-minimized code generation (#22455) Flag -ddisable-js-minimizer was producing invalid code. Fix that and also a few other things to generate nicer JS code for debugging. The added test checks that we don't regress when using the flag. - - - - - f7744e8e by Andrey Mokhov at 2023-06-03T16:49:44-04:00 [hadrian] Fix multiline synopsis rendering - - - - - b2c745db by Andrew Lelechenko at 2023-06-03T16:50:23-04:00 Elaborate on performance properties of Data.List.++ - - - - - 7cd8a61e by Matthew Pickering at 2023-06-05T11:46:23+01:00 Big TcLclEnv and CtLoc refactoring The overall goal of this refactoring is to reduce the dependency footprint of the parser and syntax tree. Good reasons include: - Better module graph parallelisability - Make it easier to migrate error messages without introducing module loops - Philosophically, there's not reason for the AST to depend on half the compiler. One of the key edges which added this dependency was > GHC.Hs.Expr -> GHC.Tc.Types (TcLclEnv) As this in turn depending on TcM which depends on HscEnv and so on. Therefore the goal of this patch is to move `TcLclEnv` out of `GHC.Tc.Types` so that `GHC.Hs.Expr` can import TcLclEnv without incurring a huge dependency chain. The changes in this patch are: * Move TcLclEnv from GHC.Tc.Types to GHC.Tc.Types.LclEnv * Create new smaller modules for the types used in TcLclEnv New Modules: - GHC.Tc.Types.ErrCtxt - GHC.Tc.Types.BasicTypes - GHC.Tc.Types.TH - GHC.Tc.Types.LclEnv - GHC.Tc.Types.CtLocEnv - GHC.Tc.Errors.Types.PromotionErr Removed Boot File: - {-# SOURCE #-} GHC.Tc.Types * Introduce TcLclCtxt, the part of the TcLclEnv which doesn't participate in restoreLclEnv. * Replace TcLclEnv in CtLoc with specific CtLocEnv which is defined in GHC.Tc.Types.CtLocEnv. Use CtLocEnv in Implic and CtLoc to record the location of the implication and constraint. By splitting up TcLclEnv from GHC.Tc.Types we allow GHC.Hs.Expr to no longer depend on the TcM monad and all that entails. Fixes #23389 #23409 - - - - - 3d8d39d1 by Matthew Pickering at 2023-06-05T11:46:23+01:00 Remove dependency of GHC.Tc.Utils.TcType on GHC.Driver.Session This removes the usage of DynFlags from Tc.Utils.TcType so that it no longer depends on GHC.Driver.Session. In general we don't want anything which is a dependency of Language.Haskell.Syntax to depend on GHC.Driver.Session and removing this edge gets us closer to that goal. - - - - - 18db5ada by Matthew Pickering at 2023-06-05T11:46:23+01:00 Move isIrrefutableHsPat to GHC.Rename.Utils and rename to isIrrefutableHsPatRn This removes edge from GHC.Hs.Pat to GHC.Driver.Session, which makes Language.Haskell.Syntax end up depending on GHC.Driver.Session. - - - - - 12919dd5 by Matthew Pickering at 2023-06-05T11:46:23+01:00 Remove dependency of GHC.Tc.Types.Constraint on GHC.Driver.Session - - - - - eb852371 by Matthew Pickering at 2023-06-05T11:46:24+01:00 hole fit plugins: Split definition into own module The hole fit plugins are defined in terms of TcM, a type we want to avoid depending on from `GHC.Tc.Errors.Types`. By moving it into its own module we can remove this dependency. It also simplifies the necessary boot file. - - - - - 9e5246d7 by Matthew Pickering at 2023-06-05T11:46:24+01:00 Move GHC.Core.Opt.CallerCC Types into separate module This allows `GHC.Driver.DynFlags` to depend on these types without depending on CoreM and hence the entire simplifier pipeline. We can also remove a hs-boot file with this change. - - - - - 52d6a7d7 by Matthew Pickering at 2023-06-05T11:46:24+01:00 Remove unecessary SOURCE import - - - - - 698d160c by Matthew Pickering at 2023-06-05T11:46:24+01:00 testsuite: Accept new output for CountDepsAst and CountDepsParser tests These are in a separate commit as the improvement to these tests is the cumulative effect of the previous set of patches rather than just the responsibility of the last one in the patchset. - - - - - 58ccf02e by sheaf at 2023-06-05T16:00:47-04:00 TTG: only allow VarBind at GhcTc The VarBind constructor of HsBind is only used at the GhcTc stage. This commit makes that explicit by setting the extension field of VarBind to be DataConCantHappen at all other stages. This allows us to delete a dead code path in GHC.HsToCore.Quote.rep_bind, and remove some panics. - - - - - 2b1ce93d by Finley McIlwaine at 2023-06-06T07:50:04-06:00 Update CHANGES.md, user guide, recomp avoidance * Add --trace-args flag for tracing arguments received to standard output * Avoid recompiling due to changes in optimization flags * Update users guide and changes.md - - - - - 54b83253 by Matthew Craven at 2023-06-06T12:59:25-04:00 Generate Addr# access ops programmatically The existing utils/genprimopcode/gen_bytearray_ops.py was relocated and extended for this purpose. Additionally, hadrian now knows about this script and uses it when generating primops.txt - - - - - ecadbc7e by Matthew Pickering at 2023-06-06T13:00:01-04:00 ghcup-metadata: Only add Nightly tag when replacing LatestNightly Previously we were always adding the Nightly tag, but this led to all the previous builds getting an increasing number of nightly tags over time. Now we just add it once, when we remove the LatestNightly tag. - - - - - f3da6676 by Finley McIlwaine at 2023-06-06T14:12:56-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - bf36c467 by Matthew Pickering at 2023-06-07T10:16:09+01:00 Revert back to e16e20d592a6f5d9ed1af17b77fafd6495242345 Neither of these MRs are ready to land yet which causes issues with other MRs which are ready to land and need haddock changes. - - - - - 421510a9 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 atSign has no unicode variant Prior to this change, atSign was defined as follows: atSign unicode = text (if unicode then "@" else "@") Yes, this is the same symbol '\64' and not your font playing tricks on you. Now we define: atSign = char '@' Both the LaTeX and the Xhtml backend are updated accordingly. - - - - - 3785c276 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 LaTeX: fix printing of type variable bindings Consider this type signature: kindOf :: forall {k} (a :: k). Proxy a -> Proxy k Prior to this fix, the LaTeX backend rendered it like this: kindOf :: forall k a. Proxy a -> Proxy k Now we preserve explicit specificity and kind annotations. - - - - - 0febf3a8 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 Add support for invisible binders in type declarations - - - - - 4aea0a72 by Vladislav Zavialov at 2023-06-07T12:06:46+02:00 Invisible binders in type declarations (#22560) This patch implements @k-binders introduced in GHC Proposal #425 and guarded behind the TypeAbstractions extension: type D :: forall k j. k -> j -> Type data D @k @j a b = ... ^^ ^^ To represent the new syntax, we modify LHsQTyVars as follows: - hsq_explicit :: [LHsTyVarBndr () pass] + hsq_explicit :: [LHsTyVarBndr (HsBndrVis pass) pass] HsBndrVis is a new data type that records the distinction between type variable binders written with and without the @ sign: data HsBndrVis pass = HsBndrRequired | HsBndrInvisible (LHsToken "@" pass) The rest of the patch updates GHC, template-haskell, and haddock to handle the new syntax. Parser: The PsErrUnexpectedTypeAppInDecl error message is removed. The syntax it used to reject is now permitted. Renamer: The @ sign does not affect the scope of a binder, so the changes to the renamer are minimal. See rnLHsTyVarBndrVisFlag. Type checker: There are three code paths that were updated to deal with the newly introduced invisible type variable binders: 1. checking SAKS: see kcCheckDeclHeader_sig, matchUpSigWithDecl 2. checking CUSK: see kcCheckDeclHeader_cusk 3. inference: see kcInferDeclHeader, rejectInvisibleBinders Helper functions bindExplicitTKBndrs_Q_Skol and bindExplicitTKBndrs_Q_Tv are generalized to work with HsBndrVis. Updates the haddock submodule. Metric Increase: MultiLayerModulesTH_OneShot Co-authored-by: Simon Peyton Jones <simon.peytonjones at gmail.com> - - - - - b7600997 by Josh Meredith at 2023-06-07T13:10:21-04:00 JS: clean up FFI 'fat arrow' calls in base:System.Posix.Internals (#23481) - - - - - e5d3940d by Sebastian Graf at 2023-06-07T18:01:28-04:00 Update CODEOWNERS - - - - - 960ef111 by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Remove IPE enabled builds from CI" This reverts commit 41b41577c8a28c236fa37e8f73aa1c6dc368d951. - - - - - bad1c8cc by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Update user's guide and release notes, small fixes" This reverts commit 3ded9a1cd22f9083f31bc2f37ee1b37f9d25dab7. - - - - - 12726d90 by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Add IPE-enabled builds to CI" This reverts commit 09d93bd0305b0f73422ce7edb67168c71d32c15f. - - - - - dbdd989d by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Add optional dependencies to ./configure output" This reverts commit a00488665cd890a26a5564a64ba23ff12c9bec58. - - - - - 240483af by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Fix IPE data decompression buffer allocation" This reverts commit 0e85099b9316ee24565084d5586bb7290669b43a. - - - - - 9b8c7dd8 by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Fix byte order of IPE data, fix IPE tests" This reverts commit 7872e2b6f08ea40d19a251c4822a384d0b397327. - - - - - 3364379b by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Add note describing IPE data compression" This reverts commit 69563c97396b8fde91678fae7d2feafb7ab9a8b0. - - - - - fda30670 by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Fix libzstd detection in configure and RTS" This reverts commit 5aef5658ad5fb96bac7719710e0ea008bf7b62e0. - - - - - 1cbcda9a by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "IPE data compression" This reverts commit b7a640acf7adc2880e5600d69bcf2918fee85553. - - - - - fb5e99aa by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Add IPE compression to configure" This reverts commit 5d1f2411f4becea8650d12d168e989241edee186. - - - - - 2cdcb3a5 by Matthew Pickering at 2023-06-07T18:02:04-04:00 Revert "Restructure IPE buffer layout" This reverts commit f3556d6cefd3d923b36bfcda0c8185abb1d11a91. - - - - - 2b0c9f5e by Simon Peyton Jones at 2023-06-08T07:52:34+00:00 Don't report redundant Givens from quantified constraints This fixes #23323 See (RC4) in Note [Tracking redundant constraints] - - - - - 13e33bb3 by Finley McIlwaine at 2023-06-08T07:51:59-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - 3e5340ce by Finley McIlwaine at 2023-06-08T07:54:27-06:00 Add note about stubdir to recompilation docs - - - - - db7e84dc by Finley at 2023-06-08T08:11:03-06:00 Merge pull request haskell/haddock#1597 from haskell/finley/hi-haddock-9.6 hi-haddock for ghc 9.6 - - - - - 567b32e1 by David Binder at 2023-06-08T18:41:29-04:00 Update the outdated instructions in HACKING.md on how to compile GHC - - - - - 2b1a4abe by Ryan Scott at 2023-06-09T07:56:58-04:00 Restore mingwex dependency on Windows This partially reverts some of the changes in !9475 to make `base` and `ghc-prim` depend on the `mingwex` library on Windows. It also restores the RTS's stubs for `mingwex`-specific symbols such as `_lock_file`. This is done because the C runtime provides `libmingwex` nowadays, and moreoever, not linking against `mingwex` requires downstream users to link against it explicitly in difficult-to-predict circumstances. Better to always link against `mingwex` and prevent users from having to do the guesswork themselves. See https://gitlab.haskell.org/ghc/ghc/-/merge_requests/10360#note_495873 for the discussion that led to this. - - - - - 28954758 by Ryan Scott at 2023-06-09T07:56:58-04:00 RtsSymbols.c: Remove mingwex symbol stubs As of !9475, the RTS now links against `ucrt` instead of `msvcrt` on Windows, which means that the RTS no longer needs to declare stubs for the `__mingw_*` family of symbols. Let's remove these stubs to avoid confusion. Fixes #23309. - - - - - 3ab0155b by Ryan Scott at 2023-06-09T07:57:35-04:00 Consistently use validity checks for TH conversion of data constructors We were checking that TH-spliced data declarations do not look like this: ```hs data D :: Type = MkD Int ``` But we were only doing so for `data` declarations' data constructors, not for `newtype`s, `data instance`s, or `newtype instance`s. This patch factors out the necessary validity checks into its own `cvtDataDefnCons` function and uses it in all of the places where it needs to be. Fixes #22559. - - - - - a24b83dd by Matthew Pickering at 2023-06-09T15:19:00-04:00 Fix behaviour of -keep-tmp-files when used in OPTIONS_GHC pragma This fixes the behaviour of -keep-tmp-files when used in an OPTIONS_GHC pragma for files with module level scope. Instead of simple not deleting the files, we also need to remove them from the TmpFs so they are not deleted later on when all the other files are deleted. There are additional complications because you also need to remove the directory where these files live from the TmpFs so we don't try to delete those later either. I added two tests. 1. Tests simply that -keep-tmp-files works at all with a single module and --make mode. 2. The other tests that temporary files are deleted for other modules which don't enable -keep-tmp-files. Fixes #23339 - - - - - dcf32882 by Matthew Pickering at 2023-06-09T15:19:00-04:00 withDeferredDiagnostics: When debugIsOn, write landmine into IORef to catch use-after-free. Ticket #23305 reports an error where we were attempting to use the logger which was created by withDeferredDiagnostics after its scope had ended. This problem would have been caught by this patch and a validate build: ``` +*** Exception: Use after free +CallStack (from HasCallStack): + error, called at compiler/GHC/Driver/Make.hs:<line>:<column> in <package-id>:GHC.Driver.Make ``` This general issue is tracked by #20981 - - - - - 432c736c by Matthew Pickering at 2023-06-09T15:19:00-04:00 Don't return complete HscEnv from upsweep By returning a complete HscEnv from upsweep the logger (as introduced by withDeferredDiagnostics) was escaping the scope of withDeferredDiagnostics and hence we were losing error messages. This is reminiscent of #20981, which also talks about writing errors into messages after their scope has ended. See #23305 for details. - - - - - 26013cdc by Alexander McKenna at 2023-06-09T15:19:41-04:00 Dump `SpecConstr` specialisations separately Introduce a `-ddump-spec-constr` flag which debugs specialisations from `SpecConstr`. These are no longer shown when you use `-ddump-spec`. - - - - - 4639100b by Matthew Pickering at 2023-06-09T18:50:43-04:00 Add role annotations to SNat, SSymbol and SChar Ticket #23454 explained it was possible to implement unsafeCoerce because SNat was lacking a role annotation. As these are supposed to be singleton types but backed by an efficient representation the correct annotation is nominal to ensure these kinds of coerces are forbidden. These annotations were missed from https://github.com/haskell/core-libraries-committee/issues/85 which was implemented in 532de36870ed9e880d5f146a478453701e9db25d. CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/170 Fixes #23454 - - - - - 9c0dcff7 by Matthew Pickering at 2023-06-09T18:51:19-04:00 Remove non-existant bytearray-ops.txt.pp file from ghc.cabal.in This broke the sdist generation. Fixes #23489 - - - - - 273ff0c7 by David Binder at 2023-06-09T18:52:00-04:00 Regression test T13438 is no longer marked as "expect_broken" in the testsuite driver. - - - - - b84a2900 by Andrei Borzenkov at 2023-06-10T08:27:28-04:00 Fix -Wterm-variable-capture scope (#23434) -Wterm-variable-capture wasn't accordant with type variable scoping in associated types, in type classes. For example, this code produced the warning: k = 12 class C k a where type AT a :: k -> Type I solved this issue by reusing machinery of newTyVarNameRn function that is accordand with associated types: it does lookup for each free type variable when we are in the type class context. And in this patch I use result of this work to make sure that -Wterm-variable-capture warns only on implicitly quantified type variables. - - - - - 9d1a8d87 by Jorge Mendes at 2023-06-10T08:28:10-04:00 Remove redundant case statement in rts/js/mem.js. - - - - - a1f350e2 by Oleg Grenrus at 2023-06-13T09:42:16-04:00 Change WarningWithFlag to plural WarningWithFlags Resolves #22825 Now each diagnostic can name multiple different warning flags for its reason. There is currently one use case: missing signatures. Currently we need to check which warning flags are enabled when generating the diagnostic, which is against the declarative nature of the diagnostic framework. This patch allows a warning diagnostic to have multiple warning flags, which makes setup more declarative. The WarningWithFlag pattern synonym is added for backwards compatibility The 'msgEnvReason' field is added to MsgEnvelope to store the `ResolvedDiagnosticReason`, which accounts for the enabled flags, and then that is used for pretty printing the diagnostic. - - - - - ec01f0ec by Matthew Pickering at 2023-06-13T09:42:59-04:00 Add a test Way for running ghci with Core optimizations Tracking ticket: #23059 This runs compile_and_run tests with optimised code with bytecode interpreter Changed submodules: hpc, process Co-authored-by: Torsten Schmits <git at tryp.io> - - - - - c6741e72 by Rodrigo Mesquita at 2023-06-13T09:43:38-04:00 Configure -Qunused-arguments instead of hardcoding it When GHC invokes clang, it currently passes -Qunused-arguments to discard warnings resulting from GHC using multiple options that aren't used. In this commit, we configure -Qunused-arguments into the Cc options instead of checking if the compiler is clang at runtime and hardcoding the flag into GHC. This is part of the effort to centralise toolchain information in toolchain target files at configure time with the end goal of a runtime retargetable GHC. This also means we don't need to call getCompilerInfo ever, which improves performance considerably (see !10589). Metric Decrease: PmSeriesG T10421 T11303b T12150 T12227 T12234 T12425 T13035 T13253-spj T13386 T15703 T16875 T17836b T17977 T17977b T18140 T18282 T18304 T18698a T18698b T18923 T20049 T21839c T3064 T5030 T5321FD T5321Fun T5837 T6048 T9020 T9198 T9872d T9961 - - - - - 0128db87 by Victor Cacciari Miraldo at 2023-06-13T09:44:18-04:00 Improve docs for Data.Fixed; adds 'realToFrac' as an option for conversion between different precisions. - - - - - 95b69cfb by Ryan Scott at 2023-06-13T09:44:55-04:00 Add regression test for #23143 !10541, the fix for #23323, also fixes #23143. Let's add a regression test to ensure that it stays fixed. Fixes #23143. - - - - - ed2dbdca by Emily Martins at 2023-06-13T09:45:37-04:00 delete GHCi.UI.Tags module and remove remaining references Co-authored-by: Tilde Rose <t1lde at protonmail.com> - - - - - c90d96e4 by Victor Cacciari Miraldo at 2023-06-13T09:46:26-04:00 Add regression test for 17328 - - - - - de58080c by Victor Cacciari Miraldo at 2023-06-13T09:46:26-04:00 Skip checking whether constructors are in scope when deriving newtype instances. Fixes #17328 - - - - - 5e3c2b05 by Philip Hazelden at 2023-06-13T09:47:07-04:00 Don't suggest `DeriveAnyClass` when instance can't be derived. Fixes #19692. Prototypical cases: class C1 a where x1 :: a -> Int data G1 = G1 deriving C1 class C2 a where x2 :: a -> Int x2 _ = 0 data G2 = G2 deriving C2 Both of these used to give this suggestion, but for C1 the suggestion would have failed (generated code with undefined methods, which compiles but warns). Now C2 still gives the suggestion but C1 doesn't. - - - - - 80a0b099 by David Binder at 2023-06-13T09:47:49-04:00 Add testcase for error GHC-00711 to testsuite - - - - - e4b33a1d by Oleg Grenrus at 2023-06-14T07:01:21-04:00 Add -Wmissing-poly-kind-signatures Implements #22826 This is a restricted version of -Wmissing-kind-signatures shown only for polykinded types. - - - - - f8395b94 by doyougnu at 2023-06-14T07:02:01-04:00 ci: special case in req_host_target_ghc for JS - - - - - b852a5b6 by Gergő Érdi at 2023-06-14T07:02:42-04:00 When forcing a `ModIface`, force the `MINIMAL` pragmas in class definitions Fixes #23486 - - - - - c29b45ee by Krzysztof Gogolewski at 2023-06-14T07:03:19-04:00 Add a testcase for #20076 Remove 'recursive' in the error message, since the error can arise without recursion. - - - - - b80ef202 by Krzysztof Gogolewski at 2023-06-14T07:03:56-04:00 Use tcInferFRR to prevent bad generalisation Fixes #23176 - - - - - bd8ef37d by Matthew Pickering at 2023-06-14T07:04:31-04:00 ci: Add dependenices on necessary aarch64 jobs for head.hackage ci These need to be added since we started testing aarch64 on head.hackage CI. The jobs will sometimes fail because they will start before the relevant aarch64 job has finished. Fixes #23511 - - - - - a0c27cee by Vladislav Zavialov at 2023-06-14T07:05:08-04:00 Add standalone kind signatures for Code and TExp CodeQ and TExpQ already had standalone kind signatures even before this change: type TExpQ :: TYPE r -> Kind.Type type CodeQ :: TYPE r -> Kind.Type Now Code and TExp have signatures too: type TExp :: TYPE r -> Kind.Type type Code :: (Kind.Type -> Kind.Type) -> TYPE r -> Kind.Type This is a stylistic change. - - - - - e70c1245 by Tom Ellis at 2023-06-14T07:05:48-04:00 Warn that GHC.TypeLits.Internal should not be used - - - - - 100650e3 by Tom Ellis at 2023-06-14T07:05:48-04:00 Warn that GHC.TypeNats.Internal should not be used - - - - - 4e085d17 by Finley McIlwaine at 2023-06-14T13:41:06-06:00 Replace SYB traversals - - - - - 7b39aec5 by Finley McIlwaine at 2023-06-14T14:20:17-06:00 Test ref accepts, remove unused HaddockClsInst - - - - - 078250ef by Jacco Krijnen at 2023-06-14T17:17:53-04:00 Add more flags for dumping core passes (#23491) - - - - - 1b7604af by Jacco Krijnen at 2023-06-14T17:17:53-04:00 Add tests for dumping flags (#23491) - - - - - 42000000 by Sebastian Graf at 2023-06-14T17:18:29-04:00 Provide a demand signature for atomicModifyMutVar.# (#23047) Fixes #23047 - - - - - 8f27023b by Ben Gamari at 2023-06-15T03:10:24-04:00 compiler: Cross-reference Note [StgToJS design] In particular, the numeric representations are quite useful context in a few places. - - - - - a71b60e9 by Andrei Borzenkov at 2023-06-15T03:11:00-04:00 Implement the -Wimplicit-rhs-quantification warning (#23510) GHC Proposal #425 "Invisible binders in type declarations" forbids implicit quantification of type variables that occur free on the right-hand side of a type synonym but are not mentioned on the left-hand side. The users are expected to rewrite this using invisible binders: type T1 :: forall a . Maybe a type T1 = 'Nothing :: Maybe a -- old type T1 @a = 'Nothing :: Maybe a -- new Since the @k-binders are a new feature, we need to wait for three releases before we require the use of the new syntax. In the meantime, we ought to provide users with a new warning, -Wimplicit-rhs-quantification, that would detect when such implicit quantification takes place, and include it in -Wcompat. - - - - - 0078dd00 by Sven Tennie at 2023-06-15T03:11:36-04:00 Minor refactorings to mkSpillInstr and mkLoadInstr Better error messages. And, use the existing `off` constant to reduce duplication. - - - - - 1792b57a by doyougnu at 2023-06-15T03:12:17-04:00 JS: merge util modules Merge Core and StgUtil modules for StgToJS pass. Closes: #23473 - - - - - 469ff08b by Vladislav Zavialov at 2023-06-15T03:12:57-04:00 Check visibility of nested foralls in can_eq_nc (#18863) Prior to this change, `can_eq_nc` checked the visibility of the outermost layer of foralls: forall a. forall b. forall c. phi1 forall x. forall y. forall z. phi2 ^^ up to here Then it delegated the rest of the work to `can_eq_nc_forall`, which split off all foralls: forall a. forall b. forall c. phi1 forall x. forall y. forall z. phi2 ^^ up to here This meant that some visibility flags were completely ignored. We fix this oversight by moving the check to `can_eq_nc_forall`. - - - - - 59c9065b by Luite Stegeman at 2023-06-15T03:13:37-04:00 JS: use regular mask for blocking IO Blocking IO used uninterruptibleMask which should make any thread blocked on IO unreachable by async exceptions (such as those from timeout). This changes it to a regular mask. It's important to note that the nodejs runtime does not actually interrupt the blocking IO when the Haskell thread receives an async exception, and that file positions may be updated and buffers may be written after the Haskell thread has already resumed. Any file descriptor affected by an async exception interruption should therefore be used with caution. - - - - - 907c06c3 by Luite Stegeman at 2023-06-15T03:13:37-04:00 JS: nodejs: do not set 'readable' handler on stdin at startup The Haskell runtime used to install a 'readable' handler on stdin at startup in nodejs. This would cause the nodejs system to start buffering the stream, causing data loss if the stdin file descriptor is passed to another process. This change delays installation of the 'readable' handler until the first read of stdin by Haskell code. - - - - - a54b40a9 by Luite Stegeman at 2023-06-15T03:13:37-04:00 JS: reserve one more virtual (negative) file descriptor This is needed for upcoming support of the process package - - - - - 78cd1132 by Andrei Borzenkov at 2023-06-15T11:16:11+04:00 Report scoped kind variables at the type-checking phase (#16635) This patch modifies the renamer to respect ScopedTypeVariables in kind signatures. This means that kind variables bound by the outermost `forall` now scope over the type: type F = '[Right @a @() () :: forall a. Either a ()] -- ^^^^^^^^^^^^^^^ ^^^ -- in scope here bound here However, any use of such variables is a type error, because we don't have type-level lambdas to bind them in Core. This is described in the new Note [Type variable scoping errors during type check] in GHC.Tc.Types. - - - - - df9c2090 by Finley McIlwaine at 2023-06-15T08:02:51-06:00 Use batchMsg for progress reporting during load With hi-haddock as is, there is an awkward silence during the load operation. This commit makes haddock use the default `batchMsg` Messager for progress reporting, and makes the default GHC verbosity level 1, so the user can see what GHC is doing. - - - - - f23679a8 by Hécate Moonlight at 2023-06-15T20:31:53+02:00 Merge pull request haskell/haddock#1600 from haskell/finley/hi-haddock-optim - - - - - a7982192 by Finley McIlwaine at 2023-06-15T15:02:16-06:00 hi-haddock squashed - - - - - 4a41ba75 by Sylvain Henry at 2023-06-15T18:09:15-04:00 JS: testsuite: use correct ticket number Replace #22356 with #22349 for these tests because #22356 has been fixed but now these tests fail because of #22349. - - - - - 15f150c8 by Sylvain Henry at 2023-06-15T18:09:16-04:00 JS: testsuite: update ticket numbers - - - - - 08d8e9ef by Sylvain Henry at 2023-06-15T18:09:16-04:00 JS: more triage - - - - - e8752e12 by Krzysztof Gogolewski at 2023-06-15T18:09:52-04:00 Fix test T18522-deb-ppr Fixes #23509 - - - - - c34f0c8d by Finley McIlwaine at 2023-06-15T16:22:03-06:00 Merge remote-tracking branch 'origin/ghc-9.6' into finley/hi-haddock-squashed - - - - - 62c56416 by Ben Price at 2023-06-16T05:52:39-04:00 Lint: more details on "Occurrence is GlobalId, but binding is LocalId" This is helpful when debugging a pass which accidentally shadowed a binder. - - - - - d4c10238 by Ryan Hendrickson at 2023-06-16T05:53:22-04:00 Clean a stray bit of text in user guide - - - - - 93647b5c by Vladislav Zavialov at 2023-06-16T05:54:02-04:00 testsuite: Add forall visibility test cases The added tests ensure that the type checker does not confuse visible and invisible foralls. VisFlag1: kind-checking type applications and inferred type variable instantiations VisFlag1_ql: kind-checking Quick Look instantiations VisFlag2: kind-checking type family instances VisFlag3: checking kind annotations on type parameters of associated type families VisFlag4: checking kind annotations on type parameters in type declarations with SAKS VisFlag5: checking the result kind annotation of data family instances - - - - - 40452797 by Bartłomiej Cieślar at 2023-06-16T12:26:04+02:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - e58673bf by Ben Gamari at 2023-06-16T09:33:35-04:00 Bump GHC version to 9.8 - - - - - 74bdf972 by Ben Gamari at 2023-06-16T09:36:18-04:00 Merge commit 'fcaaad06770a26d35d4aafd65772dedadf17669c' into ghc-head - - - - - a5f0c00e by Sylvain Henry at 2023-06-16T12:25:40-04:00 JS: factorize SaneDouble into its own module Follow-up of b159e0e9 whose ticket is #22736 - - - - - 0baf9e7c by Krzysztof Gogolewski at 2023-06-16T12:26:17-04:00 Add tests for #21973 - - - - - 640ea90e by Diego Diverio at 2023-06-16T23:07:55-04:00 Update documentation for `<**>` - - - - - 2469a813 by Diego Diverio at 2023-06-16T23:07:55-04:00 Update text - - - - - 1f515bbb by Diego Diverio at 2023-06-16T23:07:55-04:00 Update examples - - - - - 7af99a0d by Diego Diverio at 2023-06-16T23:07:55-04:00 Update documentation to actually display code correctly - - - - - 800aad7e by Andrei Borzenkov at 2023-06-16T23:08:32-04:00 Type/data instances: require that variables on the RHS are mentioned on the LHS (#23512) GHC Proposal #425 "Invisible binders in type declarations" restricts the scope of type and data family instances as follows: In type family and data family instances, require that every variable mentioned on the RHS must also occur on the LHS. For example, here are three equivalent type instance definitions accepted before this patch: type family F1 a :: k type instance F1 Int = Any :: j -> j type family F2 a :: k type instance F2 @(j -> j) Int = Any :: j -> j type family F3 a :: k type instance forall j. F3 Int = Any :: j -> j - In F1, j is implicitly quantified and it occurs only on the RHS; - In F2, j is implicitly quantified and it occurs both on the LHS and the RHS; - In F3, j is explicitly quantified. Now F1 is rejected with an out-of-scope error, while F2 and F3 continue to be accepted. - - - - - 9132d529 by Sylvain Henry at 2023-06-18T02:50:34-04:00 JS: testsuite: use correct ticket numbers - - - - - c3a1274c by Sylvain Henry at 2023-06-18T02:50:34-04:00 JS: don't dump eventlog to stderr by default Fix T16707 Bump stm submodule - - - - - 89bb8ad8 by Ryan Hendrickson at 2023-06-18T02:51:14-04:00 Fix TH name lookup for symbolic tycons (#23525) - - - - - cb9e1ce4 by Finley McIlwaine at 2023-06-18T21:16:45-06:00 IPE data compression IPE data resulting from the `-finfo-table-map` flag may now be compressed by configuring the GHC build with the `--enable-ipe-data-compression` flag. This results in about a 20% reduction in the size of IPE-enabled build results. The compression library, zstd, may optionally be statically linked by configuring with the `--enabled-static-libzstd` flag (on non-darwin platforms) libzstd version 1.4.0 or greater is required. - - - - - 0cbc3ae0 by Gergő Érdi at 2023-06-19T09:11:38-04:00 Add `IfaceWarnings` to represent the `ModIface`-storable parts of a `Warnings GhcRn`. Fixes #23516 - - - - - 3e80c2b4 by Arnaud Spiwack at 2023-06-20T03:19:41-04:00 Avoid desugaring non-recursive lets into recursive lets This prepares for having linear let expressions in the frontend. When desugaring lets, SPECIALISE statements create more copies of a let binding. Because of the rewrite rules attached to the bindings, there are dependencies between the generated binds. Before this commit, we simply wrapped all these in a mutually recursive let block, and left it to the simplified to sort it out. With this commit: we are careful to generate the bindings in dependency order, so that we can wrap them in consecutive lets (if the source is non-recursive). - - - - - 9fad49e0 by Ben Gamari at 2023-06-20T03:20:19-04:00 rts: Do not call exit() from SIGINT handler Previously `shutdown_handler` would call `stg_exit` if the scheduler was Oalready found to be in `SCHED_INTERRUPTING` state (or higher). However, `stg_exit` is not signal-safe as it calls `exit` (which calls `atexit` handlers). The only safe thing to do in this situation is to call `_exit`, which terminates with minimal cleanup. Fixes #23417. - - - - - 7485f848 by Andrew Lelechenko at 2023-06-20T03:20:57-04:00 Bump Cabal submodule This requires changing the recomp007 test because now cabal passes `this-unit-id` to executable components, and that unit-id contains a hash which includes the ABI of the dependencies. Therefore changing the dependencies means that -this-unit-id changes and recompilation is triggered. The spririt of the test is to test GHC's recompilation logic assuming that `-this-unit-id` is constant, so we explicitly pass `-ipid` to `./configure` rather than letting `Cabal` work it out. - - - - - 1464a2a8 by mangoiv at 2023-06-20T03:21:34-04:00 [feat] add a hint to `HasField` error message - add a hint that indicates that the record that the record dot is used on might just be missing a field - as the intention of the programmer is not entirely clear, it is only shown if the type is known - This addresses in part issue #22382 - - - - - 418ee3dc by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Remove NFData SourceText, IfaceWarnings updates The NFData SourceText instance is now available in GHC Handle mod_iface mi_warns now being IfaceWarnings - - - - - 62f31380 by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Accept Instances.hs test output Due to ghc!10469. - - - - - a8f2fc0e by Ben Gamari at 2023-06-20T15:48:08-04:00 Test fixes for "Fix associated data family doc structure items" Associated data families were being given their own export DocStructureItems, which resulted in them being documented separately from their classes in haddocks. This commit fixes it. - - - - - b65e78dd by Ben Gamari at 2023-06-20T16:56:43-04:00 rts/ipe: Fix unused lock warning - - - - - 6086effd by Ben Gamari at 2023-06-20T16:56:44-04:00 rts/ProfilerReportJson: Fix memory leak - - - - - 1e48c434 by Ben Gamari at 2023-06-20T16:56:44-04:00 rts: Various warnings fixes - - - - - 471486b9 by Ben Gamari at 2023-06-20T16:56:44-04:00 rts: Fix printf format mismatch - - - - - 80603fb3 by Ben Gamari at 2023-06-20T16:56:44-04:00 rts: Fix incorrect #include <sys/poll.h> According to Alpine's warnings and poll(2), <poll.h> should be preferred. - - - - - ff18e6fd by Ben Gamari at 2023-06-20T16:56:44-04:00 nonmoving: Fix unused definition warrnings - - - - - 6e7fe8ee by Ben Gamari at 2023-06-20T16:56:44-04:00 Disable futimens on Darwin. See #22938 - - - - - b7706508 by Ben Gamari at 2023-06-20T16:56:44-04:00 rts: Fix incorrect CPP guard - - - - - 94f00e9b by Ben Gamari at 2023-06-20T16:56:44-04:00 hadrian: Ensure that -Werror is passed when compiling the RTS. Previously the `+werror` transformer would only pass `-Werror` to GHC, which does not ensure that the same is passed to the C compiler when building the RTS. Arguably this is itself a bug but for now we will just work around this by passing `-optc-Werror` to GHC. I tried to enable `-Werror` in all C compilations but the boot libraries are something of a portability nightmare. - - - - - 5fb54bf8 by Ben Gamari at 2023-06-20T16:56:44-04:00 rts: Disable `#pragma GCC`s on clang compilers Otherwise the build fails due to warnings. See #23530. - - - - - cf87f380 by Ben Gamari at 2023-06-20T16:56:44-04:00 rts: Fix capitalization of prototype - - - - - 17f250d7 by Ben Gamari at 2023-06-20T16:56:44-04:00 rts: Fix incorrect format specifier - - - - - 0ff1c501 by Josh Meredith at 2023-06-20T16:57:20-04:00 JS: remove js_broken(22576) in favour of the pre-existing wordsize(32) condition (#22576) - - - - - 3d1d42b7 by Finley McIlwaine at 2023-06-21T12:04:58-04:00 Memory usage fixes for Haddock - Do not include `mi_globals` in the `NoBackend` backend. It was only included for Haddock, but Haddock does not actually need it. This causes a 200MB reduction in max residency when generating haddocks on the Agda codebase (roughly 1GB to 800MB). - Make haddock_{parser,renamer}_perf tests more accurate by forcing docs to be written to interface files using `-fwrite-interface` Bumps haddock submodule. Metric Decrease: haddock.base - - - - - 8185b1c2 by Finley McIlwaine at 2023-06-21T12:04:58-04:00 Fix associated data family doc structure items Associated data families were being given their own export DocStructureItems, which resulted in them being documented separately from their classes in haddocks. This commit fixes it. - - - - - 4d356ea3 by Sylvain Henry at 2023-06-21T12:04:59-04:00 JS: implement TH support - Add ghc-interp.js bootstrap script for the JS interpreter - Interactively link and execute iserv code from the ghci package - Incrementally load and run JS code for splices into the running iserv Co-authored-by: Luite Stegeman <stegeman at gmail.com> - - - - - 3249cf12 by Sylvain Henry at 2023-06-21T12:04:59-04:00 Don't use getKey - - - - - f84ff161 by Sylvain Henry at 2023-06-21T12:04:59-04:00 Stg: return imported FVs This is used to determine what to link when using the interpreter. For now it's only used by the JS interpreter but it could easily be used by the native interpreter too (instead of extracting names from compiled BCOs). - - - - - fab2ad23 by Sylvain Henry at 2023-06-21T12:04:59-04:00 Fix some recompilation avoidance tests - - - - - a897dc13 by Sylvain Henry at 2023-06-21T12:04:59-04:00 TH_import_loop is now broken as expected - - - - - dbb4ad51 by Sylvain Henry at 2023-06-21T12:04:59-04:00 JS: always recompile when TH is enabled (cf #23013) - - - - - cb1ac33e by Bartłomiej Cieślar at 2023-06-21T12:56:02-04:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - 9933e10b by Ben Gamari at 2023-06-21T12:56:02-04:00 Bump GHC version to 9.8 - - - - - 711b1d24 by Bartłomiej Cieślar at 2023-06-21T12:59:27-04:00 Add support for deprecating exported items (proposal #134) This is an implementation of the deprecated exports proposal #134. The proposal introduces an ability to introduce warnings to exports. This allows for deprecating a name only when it is exported from a specific module, rather than always depreacting its usage. In this example: module A ({-# DEPRECATED "do not use" #-} x) where x = undefined --- module B where import A(x) `x` will emit a warning when it is explicitly imported. Like the declaration warnings, export warnings are first accumulated within the `Warnings` struct, then passed into the ModIface, from which they are then looked up and warned about in the importing module in the `lookup_ie` helpers of the `filterImports` function (for the explicitly imported names) and in the `addUsedGRE(s)` functions where they warn about regular usages of the imported name. In terms of the AST information, the custom warning is stored in the extension field of the variants of the `IE` type (see Trees that Grow for more information). The commit includes a bump to the haddock submodule added in MR #28 Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - c1865854 by Ben Gamari at 2023-06-21T12:59:30-04:00 configure: Bump version to 9.8 Bumps Haddock submodule - - - - - fe8c18b6 by Ben Gamari at 2023-06-21T15:36:29-04:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - c61a0d5b by Ben Gamari at 2023-06-21T16:10:51-04:00 Bump GHC version to 9.9 - - - - - 4e1de71c by Ben Gamari at 2023-06-21T21:07:48-04:00 configure: Bump version to 9.9 Bumps haddock submodule. - - - - - 5b6612bc by Ben Gamari at 2023-06-23T03:56:49-04:00 rts: Work around missing prototypes errors Darwin's toolchain inexpliciably claims that `write_barrier` and friends have declarations without prototypes, despite the fact that (a) they are definitions, and (b) the prototypes appear only a few lines above. Work around this by making the definitions proper prototypes. - - - - - 43b66a13 by Matthew Pickering at 2023-06-23T03:57:26-04:00 ghcup-metadata: Fix date modifier (M = minutes, m = month) Fixes #23552 - - - - - 564164ef by Luite Stegeman at 2023-06-24T10:27:29+09:00 Support large stack frames/offsets in GHCi bytecode interpreter Bytecode instructions like PUSH_L (push a local variable) contain an operand that refers to the stack slot. Before this patch, the operand type was SmallOp (Word16), limiting the maximum stack offset to 65535 words. This could cause compiler panics in some cases (See #22888). This patch changes the operand type for stack offsets from SmallOp to Op, removing the stack offset limit. Fixes #22888 - - - - - 8d6574bc by Sylvain Henry at 2023-06-26T13:15:06-04:00 JS: support levity-polymorphic datatypes (#22360,#22291) - thread knowledge about levity into PrimRep instead of panicking - JS: remove assumption that unlifted heap objects are rts objects (TVar#, etc.) Doing this also fixes #22291 (test added). There is a small performance hit (~1% more allocations). Metric Increase: T18698a T18698b - - - - - 5578bbad by Matthew Pickering at 2023-06-26T13:15:43-04:00 MR Review Template: Mention "Blocked on Review" label In order to improve our MR review processes we now have the label "Blocked on Review" which allows people to signal that a MR is waiting on a review to happen. See: https://mail.haskell.org/pipermail/ghc-devs/2023-June/021255.html - - - - - 4427e9cf by Matthew Pickering at 2023-06-26T13:15:43-04:00 Move MR template to Default.md This makes it more obvious what you have to modify to affect the default template rather than looking in the project settings. - - - - - 522bd584 by Arnaud Spiwack at 2023-06-26T13:16:33-04:00 Revert "Avoid desugaring non-recursive lets into recursive lets" This (temporary) reverts commit 3e80c2b40213bebe302b1bd239af48b33f1b30ef. Fixes #23550 - - - - - c59fbb0b by Torsten Schmits at 2023-06-26T19:34:20+02:00 Propagate breakpoint information when inlining across modules Tracking ticket: #23394 MR: !10448 * Add constructor `IfaceBreakpoint` to `IfaceTickish` * Store breakpoint data in interface files * Store `BreakArray` for the breakpoint's module, not the current module, in BCOs * Store module name in BCOs instead of `Unique`, since the `Unique` from an `Iface` doesn't match the modules in GHCi's state * Allocate module name in `ModBreaks`, like `BreakArray` * Lookup breakpoint by module name in GHCi * Skip creating breakpoint instructions when no `ModBreaks` are available, rather than injecting `ModBreaks` in the linker when breakpoints are enabled, and panicking when `ModBreaks` is missing - - - - - 6f904808 by Greg Steuck at 2023-06-27T16:53:07-04:00 Remove undefined FP_PROG_LD_BUILD_ID from configure.ac's - - - - - e89aa072 by Andrei Borzenkov at 2023-06-27T16:53:44-04:00 Remove arity inference in type declarations (#23514) Arity inference in type declarations was introduced as a workaround for the lack of @k-binders. They were added in 4aea0a72040, so I simplified all of this by simply removing arity inference altogether. This is part of GHC Proposal #425 "Invisible binders in type declarations". - - - - - 459dee1b by Torsten Schmits at 2023-06-27T16:54:20-04:00 Relax defaulting of RuntimeRep/Levity when printing Fixes #16468 MR: !10702 Only default RuntimeRep to LiftedRep when variables are bound by the toplevel forall - - - - - 151f8f18 by Torsten Schmits at 2023-06-27T16:54:57-04:00 Remove duplicate link label in linear types docs - - - - - ecdc4353 by Rodrigo Mesquita at 2023-06-28T12:24:57-04:00 Stop configuring unused Ld command in `settings` GHC has no direct dependence on the linker. Rather, we depend upon the C compiler for linking and an object-merging program (which is typically `ld`) for production of GHCi objects and merging of C stubs into final object files. Despite this, for historical reasons we still recorded information about the linker into `settings`. Remove these entries from `settings`, `hadrian/cfg/system.config`, as well as the `configure` logic responsible for this information. Closes #23566. - - - - - bf9ec3e4 by Bryan Richter at 2023-06-28T12:25:33-04:00 Remove extraneous debug output - - - - - 7eb68dd6 by Bryan Richter at 2023-06-28T12:25:33-04:00 Work with unset vars in -e mode - - - - - 49c27936 by Bryan Richter at 2023-06-28T12:25:33-04:00 Pass positional arguments in their positions By quoting $cmd, the default "bash -i" is a single argument to run, and no file named "bash -i" actually exists to be run. - - - - - 887dc4fc by Bryan Richter at 2023-06-28T12:25:33-04:00 Handle unset value in -e context - - - - - 5ffc7d7b by Rodrigo Mesquita at 2023-06-28T21:07:36-04:00 Configure CPP into settings There is a distinction to be made between the Haskell Preprocessor and the C preprocessor. The former is used to preprocess Haskell files, while the latter is used in C preprocessing such as Cmm files. In practice, they are both the same program (usually the C compiler) but invoked with different flags. Previously we would, at configure time, configure the haskell preprocessor and save the configuration in the settings file, but, instead of doing the same for CPP, we had hardcoded in GHC that the CPP program was either `cc -E` or `cpp`. This commit fixes that asymmetry by also configuring CPP at configure time, and tries to make more explicit the difference between HsCpp and Cpp (see Note [Preprocessing invocations]). Note that we don't use the standard CPP and CPPFLAGS to configure Cpp, but instead use the non-standard --with-cpp and --with-cpp-flags. The reason is that autoconf sets CPP to "$CC -E", whereas we expect the CPP command to be configured as a standalone executable rather than a command. These are symmetrical with --with-hs-cpp and --with-hs-cpp-flags. Cleanup: Hadrian no longer needs to pass the CPP configuration for CPP to be C99 compatible through -optP, since we now configure that into settings. Closes #23422 - - - - - 5efa9ca5 by Ben Gamari at 2023-06-28T21:08:13-04:00 hadrian: Always canonicalize topDirectory Hadrian's `topDirectory` is intended to provide an absolute path to the root of the GHC tree. However, if the tree is reached via a symlink this One question here is whether the `canonicalizePath` call is expensive enough to warrant caching. In a quick microbenchmark I observed that `canonicalizePath "."` takes around 10us per call; this seems sufficiently low not to worry. Alternatively, another approach here would have been to rather move the canonicalization into `m4/fp_find_root.m4`. This would have avoided repeated canonicalization but sadly path canonicalization is a hard problem in POSIX shell. Addresses #22451. - - - - - b3e1436f by aadaa_fgtaa at 2023-06-28T21:08:53-04:00 Optimise ELF linker (#23464) - cache last elements of `relTable`, `relaTable` and `symbolTables` in `ocInit_ELF` - cache shndx table in ObjectCode - run `checkProddableBlock` only with debug rts - - - - - 30525b00 by Ben Gamari at 2023-06-28T21:09:30-04:00 compiler: Introduce MO_{ACQUIRE,RELEASE}_FENCE - - - - - b787e259 by Ben Gamari at 2023-06-28T21:09:30-04:00 compiler: Drop MO_WriteBarrier rts: Drop write_barrier - - - - - 7550b4a5 by Ben Gamari at 2023-06-28T21:09:30-04:00 rts: Drop load_store_barrier() This is no longer used. - - - - - d5f2875e by Ben Gamari at 2023-06-28T21:09:31-04:00 rts: Drop last instances of prim_{write,read}_barrier - - - - - 965ac2ba by Ben Gamari at 2023-06-28T21:09:31-04:00 rts: Eliminate remaining uses of load_load_barrier - - - - - 0fc5cb97 by Sven Tennie at 2023-06-28T21:09:31-04:00 compiler: Drop MO_ReadBarrier - - - - - 7a7d326c by Ben Gamari at 2023-06-28T21:09:31-04:00 rts: Drop load_load_barrier This is no longer used. - - - - - 9f63da66 by Sven Tennie at 2023-06-28T21:09:31-04:00 Delete write_barrier function - - - - - bb0ed354 by Ben Gamari at 2023-06-28T21:09:31-04:00 rts: Make collectFreshWeakPtrs definition a prototype x86-64/Darwin's toolchain inexplicably warns that collectFreshWeakPtrs needs to be a prototype. - - - - - ef81a1eb by Sven Tennie at 2023-06-28T21:10:08-04:00 Fix number of free double regs D1..D4 are defined for aarch64 and thus not free. - - - - - c335fb7c by Ryan Scott at 2023-06-28T21:10:44-04:00 Fix typechecking of promoted empty lists The `'[]` case in `tc_infer_hs_type` is smart enough to handle arity-0 uses of `'[]` (see the newly added `T23543` test case for an example), but the `'[]` case in `tc_hs_type` was not. We fix this by changing the `tc_hs_type` case to invoke `tc_infer_hs_type`, as prescribed in `Note [Future-proofing the type checker]`. There are some benign changes to test cases' expected output due to the new code path using `forall a. [a]` as the kind of `'[]` rather than `[k]`. Fixes #23543. - - - - - fcf310e7 by Rodrigo Mesquita at 2023-06-28T21:11:21-04:00 Configure MergeObjs supports response files rather than Ld The previous configuration script to test whether Ld supported response files was * Incorrect (see #23542) * Used, in practice, to check if the *merge objects tool* supported response files. This commit modifies the macro to run the merge objects tool (rather than Ld), using a response file, and checking the result with $NM Fixes #23542 - - - - - 78b2f3cc by Sylvain Henry at 2023-06-28T21:12:02-04:00 JS: fix JS stack printing (#23565) - - - - - 9f01d14b by Matthew Pickering at 2023-06-29T04:13:41-04:00 Add -fpolymorphic-specialisation flag (off by default at all optimisation levels) Polymorphic specialisation has led to a number of hard to diagnose incorrect runtime result bugs (see #23469, #23109, #21229, #23445) so this commit introduces a flag `-fpolymorhphic-specialisation` which allows users to turn on this experimental optimisation if they are willing to buy into things going very wrong. Ticket #23469 - - - - - b1e611d5 by Ben Gamari at 2023-06-29T04:14:17-04:00 Rip out runtime linker/compiler checks We used to choose flags to pass to the toolchain at runtime based on the platform running GHC, and in this commit we drop all of those runtime linker checks Ultimately, this represents a change in policy: We no longer adapt at runtime to the toolchain being used, but rather make final decisions about the toolchain used at /configure time/ (we have deleted Note [Run-time linker info] altogether!). This works towards the goal of having all toolchain configuration logic living in the same place, which facilities the work towards a runtime-retargetable GHC (see #19877). As of this commit, the runtime linker/compiler logic was moved to autoconf, but soon it, and the rest of the existing toolchain configuration logic, will live in the standalone ghc-toolchain program (see !9263) In particular, what used to be done at runtime is now as follows: * The flags -Wl,--no-as-needed for needed shared libs are configured into settings * The flag -fstack-check is configured into settings * The check for broken tables-next-to-code was outdated * We use the configured c compiler by default as the assembler program * We drop `asmOpts` because we already configure -Qunused-arguments flag into settings (see !10589) Fixes #23562 Co-author: Rodrigo Mesquita (@alt-romes) - - - - - 8b35e8ca by Ben Gamari at 2023-06-29T18:46:12-04:00 Define FFI_GO_CLOSURES The libffi shipped with Apple's XCode toolchain does not contain a definition of the FFI_GO_CLOSURES macro, despite containing references to said macro. Work around this by defining the macro, following the model of a similar workaround in OpenJDK [1]. [1] https://github.com/openjdk/jdk17u-dev/pull/741/files - - - - - d7ef1704 by Ben Gamari at 2023-06-29T18:46:12-04:00 base: Fix incorrect CPP guard This was guarded on `darwin_HOST_OS` instead of `defined(darwin_HOST_OS)`. - - - - - 7c7d1f66 by Ben Gamari at 2023-06-29T18:46:48-04:00 rts/Trace: Ensure that debugTrace arguments are used As debugTrace is a macro we must take care to ensure that the fact is clear to the compiler lest we see warnings. - - - - - cb92051e by Ben Gamari at 2023-06-29T18:46:48-04:00 rts: Various warnings fixes - - - - - dec81dd1 by Ben Gamari at 2023-06-29T18:46:48-04:00 hadrian: Ignore warnings in unix and semaphore-compat - - - - - d7f6448a by Matthew Pickering at 2023-06-30T12:38:43-04:00 hadrian: Fix dependencies of docs:* rule For the docs:* rule we need to actually build the package rather than just the haddocks for the dependent packages. Therefore we depend on the .conf files of the packages we are trying to build documentation for as well as the .haddock files. Fixes #23472 - - - - - cec90389 by sheaf at 2023-06-30T12:39:27-04:00 Add tests for #22106 Fixes #22106 - - - - - 083794b1 by Torsten Schmits at 2023-07-03T03:27:27-04:00 Add -fbreak-points to control breakpoint insertion Rather than statically enabling breakpoints only for the interpreter, this adds a new flag. Tracking ticket: #23057 MR: !10466 - - - - - fd8c5769 by Ben Gamari at 2023-07-03T03:28:04-04:00 rts: Ensure that pinned allocations respect block size Previously, it was possible for pinned, aligned allocation requests to allocate beyond the end of the pinned accumulator block. Specifically, we failed to account for the padding needed to achieve the requested alignment in the "large object" check. With large alignment requests, this can result in the allocator using the capability's pinned object accumulator block to service a request which is larger than `PINNED_EMPTY_SIZE`. To fix this we reorganize `allocatePinned` to consistently account for the alignment padding in all large object checks. This is a bit subtle as we must handle the case of a small allocation request filling the accumulator block, as well as large requests. Fixes #23400. - - - - - 98185d52 by Ben Gamari at 2023-07-03T03:28:05-04:00 testsuite: Add test for #23400 - - - - - 4aac0540 by Ben Gamari at 2023-07-03T03:28:42-04:00 ghc-heap: Support for BLOCKING_QUEUE closures - - - - - 03f941f4 by Ben Bellick at 2023-07-03T03:29:29-04:00 Add some structured diagnostics in Tc/Validity.hs This addresses the work of ticket #20118 Created the following constructors for TcRnMessage - TcRnInaccessibleCoAxBranch - TcRnPatersonCondFailure - - - - - 6074cc3c by Moisés Ackerman at 2023-07-03T03:30:13-04:00 Add failing test case for #23492 - - - - - 356a2692 by Moisés Ackerman at 2023-07-03T03:30:13-04:00 Use generated src span for catch-all case of record selector functions This fixes #23492. The problem was that we used the real source span of the field declaration for the generated catch-all case in the selector function, in particular in the generated call to `recSelError`, which meant it was included in the HIE output. Using `generatedSrcSpan` instead means that it is not included. - - - - - 3efe7f39 by Moisés Ackerman at 2023-07-03T03:30:13-04:00 Introduce genLHsApp and genLHsLit helpers in GHC.Rename.Utils - - - - - dd782343 by Moisés Ackerman at 2023-07-03T03:30:13-04:00 Construct catch-all default case using helpers GHC.Rename.Utils concrete helpers instead of wrapGenSpan + HS AST constructors - - - - - 0e09c38e by Ryan Hendrickson at 2023-07-03T03:30:56-04:00 Add regression test for #23549 - - - - - 32741743 by Alexis King at 2023-07-03T03:31:36-04:00 perf tests: Increase default stack size for MultiLayerModules An unhelpfully small stack size appears to have been the real culprit behind the metric fluctuations in #19293. Debugging metric decreases triggered by !10729 helped to finally identify the problem. Metric Decrease: MultiLayerModules MultiLayerModulesTH_Make T13701 T14697 - - - - - 82ac6bf1 by Bryan Richter at 2023-07-03T03:32:15-04:00 Add missing void prototypes to rts functions See #23561. - - - - - 6078b429 by Ben Gamari at 2023-07-03T03:32:51-04:00 gitlab-ci: Refactor compilation of gen_ci Flakify and document it, making it far less sensitive to the build environment. - - - - - aa2db0ae by Ben Gamari at 2023-07-03T03:33:29-04:00 testsuite: Update documentation - - - - - 924a2362 by Gregory Gerasev at 2023-07-03T03:34:10-04:00 Better error for data deriving of type synonym/family. Closes #23522 - - - - - 4457da2a by Dave Barton at 2023-07-03T03:34:51-04:00 Fix some broken links and typos - - - - - de5830d0 by Ben Gamari at 2023-07-04T22:03:59-04:00 configure: Rip out Solaris dyld check Solaris 11 was released over a decade ago and, moreover, I doubt we have any Solaris users - - - - - 59c5fe1d by doyougnu at 2023-07-04T22:04:56-04:00 CI: add JS release and debug builds, regen CI jobs - - - - - 679bbc97 by Vladislav Zavialov at 2023-07-04T22:05:32-04:00 testsuite: Do not require CUSKs Numerous tests make use of CUSKs (complete user-supplied kinds), a legacy feature scheduled for deprecation. In order to proceed with the said deprecation, the tests have been updated to use SAKS instead (standalone kind signatures). This also allows us to remove the Haskell2010 language pragmas that were added in 115cd3c85a8 to work around the lack of CUSKs in GHC2021. - - - - - 945d3599 by Ben Gamari at 2023-07-04T22:06:08-04:00 gitlab: Drop backport-for-8.8 MR template Its usefulness has long passed. - - - - - 66c721d3 by Alan Zimmerman at 2023-07-04T22:06:44-04:00 EPA: Simplify GHC/Parser.y comb2 Use the HasLoc instance from Ast.hs to allow comb2 to work with anything with a SrcSpan This gets rid of the custom comb2A, comb2Al, comb2N functions, and removes various reLoc calls. - - - - - 2be99b7e by Matthew Pickering at 2023-07-04T22:07:21-04:00 Fix deprecation warning when deprecated identifier is from another module A stray 'Just' was being printed in the deprecation message. Fixes #23573 - - - - - 46c9bcd6 by Ben Gamari at 2023-07-04T22:07:58-04:00 rts: Don't rely on initializers for sigaction_t As noted in #23577, CentOS's ancient toolchain throws spurious missing-field-initializer warnings. - - - - - ec55035f by Ben Gamari at 2023-07-04T22:07:58-04:00 hadrian: Don't treat -Winline warnings as fatal Such warnings are highly dependent upon the toolchain, platform, and build configuration. It's simply too fragile to rely on these. - - - - - 3a09b789 by Ben Gamari at 2023-07-04T22:07:58-04:00 hadrian: Only pass -Wno-nonportable-include-path on Darwin This flag, which was introduced due to #17798, is only understood by Clang and consequently throws warnings on platforms using gcc. Sadly, there is no good way to treat such warnings as non-fatal with `-Werror` so for now we simply make this flag specific to platforms known to use Clang and case-insensitive filesystems (Darwin and Windows). See #23577. - - - - - 4af7eac2 by Mario Blažević at 2023-07-04T22:08:38-04:00 Fixed ticket #23571, TH.Ppr.pprLit hanging on large numeric literals - - - - - 2304c697 by Ben Gamari at 2023-07-04T22:09:15-04:00 compiler: Make OccSet opaque - - - - - cf735db8 by Andrei Borzenkov at 2023-07-04T22:09:51-04:00 Add Note about why we need forall in Code to be on the right - - - - - fb140f82 by Hécate Moonlight at 2023-07-04T22:10:34-04:00 Relax the constraint about the foreign function's calling convention of FinalizerPtr to capi as well as ccall. - - - - - 9ce44336 by meooow25 at 2023-07-05T11:42:37-04:00 Improve the situation with the stimes cycle Currently the Semigroup stimes cycle is resolved in GHC.Base by importing stimes implementations from a hs-boot file. Resolve the cycle using hs-boot files for required classes (Num, Integral) instead. Now stimes can be defined directly in GHC.Base, making inlining and specialization possible. This leads to some new boot files for `GHC.Num` and `GHC.Real`, the methods for those are only used to implement `stimes` so it doesn't appear that these boot files will introduce any new performance traps. Metric Decrease: T13386 T8095 Metric Increase: T13253 T13386 T18698a T18698b T19695 T8095 - - - - - 9edcb1fb by Jaro Reinders at 2023-07-05T11:43:24-04:00 Refactor Unique to be represented by Word64 In #22010 we established that Int was not always sufficient to store all the uniques we generate during compilation on 32-bit platforms. This commit addresses that problem by using Word64 instead of Int for uniques. The core of the change is in GHC.Core.Types.Unique and GHC.Core.Types.Unique.Supply. However, the representation of uniques is used in many other places, so those needed changes too. Additionally, the RTS has been extended with an atomic_inc64 operation. One major change from this commit is the introduction of the Word64Set and Word64Map data types. These are adapted versions of IntSet and IntMap from the containers package. These are planned to be upstreamed in the future. As a natural consequence of these changes, the compiler will be a bit slower and take more space on 32-bit platforms. Our CI tests indicate around a 5% residency increase. Metric Increase: CoOpt_Read CoOpt_Singletons LargeRecord ManyAlternatives ManyConstructors MultiComponentModules MultiComponentModulesRecomp MultiLayerModulesTH_OneShot RecordUpdPerf T10421 T10547 T12150 T12227 T12234 T12425 T12707 T13035 T13056 T13253 T13253-spj T13379 T13386 T13719 T14683 T14697 T14766 T15164 T15703 T16577 T16875 T17516 T18140 T18223 T18282 T18304 T18698a T18698b T18923 T1969 T19695 T20049 T21839c T3064 T3294 T4801 T5030 T5321FD T5321Fun T5631 T5642 T5837 T6048 T783 T8095 T9020 T9198 T9233 T9630 T9675 T9872a T9872b T9872b_defer T9872c T9872d T9961 TcPlugin_RewritePerf UniqLoop WWRec hard_hole_fits - - - - - 6b9db7d4 by Brandon Chinn at 2023-07-05T11:44:03-04:00 Fix docs for __GLASGOW_HASKELL_FULL_VERSION__ macro - - - - - 40f4ef7c by Torsten Schmits at 2023-07-05T18:06:19-04:00 Substitute free variables captured by breakpoints in SpecConstr Fixes #23267 - - - - - 2b55cb5f by sheaf at 2023-07-05T18:07:07-04:00 Reinstate untouchable variable error messages This extra bit of information was accidentally being discarded after a refactoring of the way we reported problems when unifying a type variable with another type. This patch rectifies that. - - - - - 53ed21c5 by Rodrigo Mesquita at 2023-07-05T18:07:47-04:00 configure: Drop Clang command from settings Due to 01542cb7227614a93508b97ecad5b16dddeb6486 we no longer use the `runClang` function, and no longer need to configure into settings the Clang command. We used to determine options at runtime to pass clang when it was used as an assembler, but now that we configure at configure time we no longer need to. - - - - - 6fdcf969 by Torsten Schmits at 2023-07-06T12:12:09-04:00 Filter out nontrivial substituted expressions in substTickish Fixes #23272 - - - - - 41968fd6 by Sylvain Henry at 2023-07-06T12:13:02-04:00 JS: testsuite: use req_c predicate instead of js_broken - - - - - 74a4dd2e by Sylvain Henry at 2023-07-06T12:13:02-04:00 JS: implement some file primitives (lstat,rmdir) (#22374) - Implement lstat and rmdir. - Implement base_c_s_is* functions (testing a file type) - Enable passing tests - - - - - 7e759914 by Sylvain Henry at 2023-07-07T02:39:38-04:00 JS: cleanup utils (#23314) - Removed unused code - Don't export unused functions - Move toTypeList to Closure module - - - - - f617655c by Sylvain Henry at 2023-07-07T02:39:38-04:00 JS: rename VarType/Vt into JSRep - - - - - 19216ca5 by Sylvain Henry at 2023-07-07T02:39:38-04:00 JS: remove custom PrimRep conversion (#23314) We use the usual conversion to PrimRep and then we convert these PrimReps to JSReps. - - - - - d3de8668 by Sylvain Henry at 2023-07-07T02:39:38-04:00 JS: don't use isRuntimeRepKindedTy in JS FFI - - - - - 8d1b75cb by Matthew Pickering at 2023-07-07T02:40:15-04:00 ghcup-metadata: Also updates ghcup-nightlies-0.0.7.yaml file Fixes #23600 - - - - - e524fa7f by Matthew Pickering at 2023-07-07T02:40:15-04:00 ghcup-metadata: Use dynamically linked alpine bindists In theory these will work much better on alpine to allow people to build statically linked applications there. We don't need to distribute a statically linked application ourselves in order to allow that. Fixes #23602 - - - - - 0c2a756e by sheaf at 2023-07-07T13:45:12+02:00 Avoid incomplete record update in Haddock Hoogle This commit avoids running into an incomplete record update warning in the Hoogle backend of Haddock. This was only noticed now, because incomplete record updates were broken in GHC 9.6. Now that they are fixed, we have to avoid running into them! - - - - - b9e7beb9 by Ben Gamari at 2023-07-07T11:32:22-04:00 Drop circle-ci-job.sh - - - - - 9955eead by Ben Gamari at 2023-07-07T11:32:22-04:00 testsuite: Allow preservation of unexpected output Here we introduce a new flag to the testsuite driver, --unexpected-output-dir=<dir>, which allows the user to ask the driver to preserve unexpected output from tests. The intent is for this to be used in CI to allow users to more easily fix unexpected platform-dependent output. - - - - - 48f80968 by Ben Gamari at 2023-07-07T11:32:22-04:00 gitlab-ci: Preserve unexpected output Here we enable use of the testsuite driver's `--unexpected-output-dir` flag by CI, preserving the result as an artifact for use by users. - - - - - 76983a0d by Matthew Pickering at 2023-07-07T11:32:58-04:00 driver: Fix -S with .cmm files There was an oversight in the driver which assumed that you would always produce a `.o` file when compiling a .cmm file. Fixes #23610 - - - - - 6df15e93 by Mike Pilgrem at 2023-07-07T11:33:40-04:00 Update Hadrian's stack.yaml - - - - - 1dff43cf by Ben Gamari at 2023-07-08T05:05:37-04:00 compiler: Rework ShowSome Previously the field used to filter the sub-declarations to show was rather ad-hoc and was only able to show at most one sub-declaration. - - - - - 8165404b by Ben Gamari at 2023-07-08T05:05:37-04:00 testsuite: Add test to catch changes in core libraries This adds testing infrastructure to ensure that changes in core libraries (e.g. `base` and `ghc-prim`) are caught in CI. - - - - - ec1c32e2 by Melanie Phoenix at 2023-07-08T05:06:14-04:00 Deprecate Data.List.NonEmpty.unzip - - - - - 5d2442b8 by Ben Gamari at 2023-07-08T05:06:51-04:00 Drop latent mentions of -split-objs Closes #21134. - - - - - a9bc20cb by Oleg Grenrus at 2023-07-08T05:07:31-04:00 Add warn_and_run test kind This is a compile_and_run variant which also captures the GHC's stderr. The warn_and_run name is best I can come up with, as compile_and_run is taken. This is useful specifically for testing warnings. We want to test that when warning triggers, and it's not a false positive, i.e. that the runtime behaviour is indeed "incorrect". As an example a single test is altered to use warn_and_run - - - - - c7026962 by Ben Gamari at 2023-07-08T05:08:11-04:00 configure: Don't use ld.gold on i386 ld.gold appears to produce invalid static constructor tables on i386. While ideally we would add an autoconf check to check for this brokenness, sadly such a check isn't easy to compose. Instead to summarily reject such linkers on i386. Somewhat hackily closes #23579. - - - - - 054261dd by Andrew Lelechenko at 2023-07-08T19:32:47-04:00 Add since annotations for Data.Foldable1 - - - - - 550af505 by Sylvain Henry at 2023-07-08T19:33:28-04:00 JS: support -this-unit-id for programs in the linker (#23613) - - - - - d284470a by Andrew Lelechenko at 2023-07-08T19:34:08-04:00 Bump text submodule - - - - - 8e11630e by jade at 2023-07-10T16:58:40-04:00 Add a hint to enable ExplicitNamespaces for type operator imports (Fixes/Enhances #20007) As suggested in #20007 and implemented in !8895, trying to import type operators will suggest a fix to use the 'type' keyword, without considering whether ExplicitNamespaces is enabled. This patch will query whether ExplicitNamespaces is enabled and add a hint to suggest enabling ExplicitNamespaces if it isn't enabled, alongside the suggestion of adding the 'type' keyword. - - - - - 61b1932e by sheaf at 2023-07-10T16:59:26-04:00 tyThingLocalGREs: include all DataCons for RecFlds The GREInfo for a record field should include the collection of all the data constructors of the parent TyCon that have this record field. This information was being incorrectly computed in the tyThingLocalGREs function for a DataCon, as we were not taking into account other DataCons with the same parent TyCon. Fixes #23546 - - - - - e6627cbd by Alan Zimmerman at 2023-07-10T17:00:05-04:00 EPA: Simplify GHC/Parser.y comb3 A follow up to !10743 - - - - - ee20da34 by Andrew Lelechenko at 2023-07-10T17:01:01-04:00 Document that compareByteArrays# is available since ghc-prim-0.5.2.0 - - - - - 4926af7b by Matthew Pickering at 2023-07-10T17:01:38-04:00 Revert "Bump text submodule" This reverts commit d284470a77042e6bc17bdb0ab0d740011196958a. This commit requires that we bootstrap with ghc-9.4, which we do not require until #23195 has been completed. Subsequently this has broken nighty jobs such as the rocky8 job which in turn has broken nightly releases. - - - - - d1c92bf3 by Ben Gamari at 2023-07-11T08:07:02-04:00 compiler: Fingerprint more code generation flags Previously our recompilation check was quite inconsistent in its coverage of non-optimisation code generation flags. Specifically, we failed to account for most flags that would affect the behavior of generated code in ways that might affect the result of a program's execution (e.g. `-feager-blackholing`, `-fstrict-dicts`) Closes #23369. - - - - - eb623149 by Ben Gamari at 2023-07-11T08:07:02-04:00 compiler: Record original thunk info tables on stack Here we introduce a new code generation option, `-forig-thunk-info`, which ensures that an `stg_orig_thunk_info` frame is pushed before every update frame. This can be invaluable when debugging thunk cycles and similar. See Note [Original thunk info table frames] for details. Closes #23255. - - - - - 4731f44e by Jaro Reinders at 2023-07-11T08:07:40-04:00 Fix wrong MIN_VERSION_GLASGOW_HASKELL macros I forgot to change these after rebasing. - - - - - dd38aca9 by Andreas Schwab at 2023-07-11T13:55:56+00:00 Hadrian: enable GHCi support on riscv64 - - - - - 09a5c6cc by Josh Meredith at 2023-07-12T11:25:13-04:00 JavaScript: support unicode code points > 2^16 in toJSString using String.fromCodePoint (#23628) - - - - - 29fbbd4e by Matthew Pickering at 2023-07-12T11:25:49-04:00 Remove references to make build system in mk/build.mk Fixes #23636 - - - - - 630e3026 by sheaf at 2023-07-12T11:26:43-04:00 Valid hole fits: don't panic on a Given The function GHC.Tc.Errors.validHoleFits would end up panicking when encountering a Given constraint. To fix this, it suffices to filter out the Givens before continuing. Fixes #22684 - - - - - c39f279b by Matthew Pickering at 2023-07-12T23:18:38-04:00 Use deb10 for i386 bindists deb9 is now EOL so it's time to upgrade the i386 bindist to use deb10 Fixes #23585 - - - - - bf9b9de0 by Krzysztof Gogolewski at 2023-07-12T23:19:15-04:00 Fix #23567, a specializer bug Found by Simon in https://gitlab.haskell.org/ghc/ghc/-/issues/23567#note_507834 The testcase isn't ideal because it doesn't detect the bug in master, unless doNotUnbox is removed as in https://gitlab.haskell.org/ghc/ghc/-/issues/23567#note_507692. But I have confirmed that with that modification, it fails before and passes afterwards. - - - - - 84c1a4a2 by Bartłomiej Cieślar at 2023-07-12T23:20:08-04:00 Comments - - - - - b2846cb5 by Bartłomiej Cieślar at 2023-07-12T23:20:08-04:00 updates to comments - - - - - 2af23f0e by Bartłomiej Cieślar at 2023-07-12T23:20:08-04:00 changes - - - - - 6143838a by sheaf at 2023-07-13T08:02:17-04:00 Fix deprecation of record fields Commit 3f374399 inadvertently broke the deprecation/warning mechanism for record fields due to its introduction of record field namespaces. This patch ensures that, when a top-level deprecation is applied to an identifier, it applies to all the record fields as well. This is achieved by refactoring GHC.Rename.Env.lookupLocalTcNames, and GHC.Rename.Env.lookupBindGroupOcc, to not look up a fixed number of NameSpaces but to look up all NameSpaces and filter out the irrelevant ones. - - - - - 6fd8f566 by sheaf at 2023-07-13T08:02:17-04:00 Introduce greInfo, greParent These are simple helper functions that wrap the internal field names gre_info, gre_par. - - - - - 7f0a86ed by sheaf at 2023-07-13T08:02:17-04:00 Refactor lookupGRE_... functions This commit consolidates all the logic for looking up something in the Global Reader Environment into the single function lookupGRE. This allows us to declaratively specify all the different modes of looking up in the GlobalRdrEnv, and avoids manually passing around filtering functions as was the case in e.g. the function GHC.Rename.Env.lookupSubBndrOcc_helper. ------------------------- Metric Decrease: T8095 ------------------------- ------------------------- Metric Increase: T8095 ------------------------- - - - - - 5e951395 by Rodrigo Mesquita at 2023-07-13T08:02:54-04:00 configure: Drop DllWrap command We used to configure into settings a DllWrap command for windows builds and distributions, however, we no longer do, and dllwrap is effectively unused. This simplification is motivated in part by the larger toolchain-selection project (#19877, !9263) - - - - - e10556b6 by Teo Camarasu at 2023-07-14T16:28:46-04:00 base: fix haddock syntax in GHC.Profiling - - - - - 0f3fda81 by Matthew Pickering at 2023-07-14T16:29:23-04:00 Revert "CI: add JS release and debug builds, regen CI jobs" This reverts commit 59c5fe1d4b624423b1c37891710f2757bb58d6af. This commit added two duplicate jobs on all validate pipelines, so we are reverting for now whilst we work out what the best way forward is. Ticket #23618 - - - - - 54bca324 by Alan Zimmerman at 2023-07-15T03:23:26-04:00 EPA: Simplify GHC/Parser.y sLL Follow up to !10743 - - - - - c8863828 by sheaf at 2023-07-15T03:24:06-04:00 Configure: canonicalise PythonCmd on Windows This change makes PythonCmd resolve to a canonical absolute path on Windows, which prevents HLS getting confused (now that we have a build-time dependency on python). fixes #23652 - - - - - ca1e636a by Rodrigo Mesquita at 2023-07-15T03:24:42-04:00 Improve Note [Binder-swap during float-out] - - - - - cf86f3ec by Matthew Craven at 2023-07-16T01:42:09+02:00 Equality of forall-types is visibility aware This patch finally (I hope) nails the question of whether (forall a. ty) and (forall a -> ty) are `eqType`: they aren't! There is a long discussion in #22762, plus useful Notes: * Note [ForAllTy and type equality] in GHC.Core.TyCo.Compare * Note [Comparing visiblities] in GHC.Core.TyCo.Compare * Note [ForAllCo] in GHC.Core.TyCo.Rep It also establishes a helpful new invariant for ForAllCo, and ForAllTy, when the bound variable is a CoVar:in that case the visibility must be coreTyLamForAllTyFlag. All this is well documented in revised Notes. - - - - - 7f13acbf by Vladislav Zavialov at 2023-07-16T01:56:27-04:00 List and Tuple<n>: update documentation Add the missing changelog.md entries and @since-annotations. - - - - - 2afbddb0 by Andrei Borzenkov at 2023-07-16T10:21:24+04:00 Type patterns (#22478, #18986) Improved name resolution and type checking of type patterns in constructors: 1. HsTyPat: a new dedicated data type that represents type patterns in HsConPatDetails instead of reusing HsPatSigType 2. rnHsTyPat: a new function that renames a type pattern and collects its binders into three groups: - explicitly bound type variables, excluding locally bound variables - implicitly bound type variables from kind signatures (only if ScopedTypeVariables are enabled) - named wildcards (only from kind signatures) 2a. rnHsPatSigTypeBindingVars: removed in favour of rnHsTyPat 2b. rnImplcitTvBndrs: removed because no longer needed 3. collect_pat: updated to collect type variable binders from type patterns (this means that types and terms use the same infrastructure to detect conflicting bindings, unused variables and name shadowing) 3a. CollVarTyVarBinders: a new CollectFlag constructor that enables collection of type variables 4. tcHsTyPat: a new function that typechecks type patterns, capable of handling polymorphic kinds. See Note [Type patterns: binders and unifiers] Examples of code that is now accepted: f = \(P @a) -> \(P @a) -> ... -- triggers -Wname-shadowing g :: forall a. Proxy a -> ... g (P @a) = ... -- also triggers -Wname-shadowing h (P @($(TH.varT (TH.mkName "t")))) = ... -- t is bound at splice time j (P @(a :: (x,x))) = ... -- (x,x) is no longer rejected data T where MkT :: forall (f :: forall k. k -> Type). f Int -> f Maybe -> T k :: T -> () k (MkT @f (x :: f Int) (y :: f Maybe)) = () -- f :: forall k. k -> Type Examples of code that is rejected with better error messages: f (Left @a @a _) = ... -- new message: -- • Conflicting definitions for ‘a’ -- Bound at: Test.hs:1:11 -- Test.hs:1:14 Examples of code that is now rejected: {-# OPTIONS_GHC -Werror=unused-matches #-} f (P @a) = () -- Defined but not used: type variable ‘a’ - - - - - eb1a6ab1 by sheaf at 2023-07-16T09:20:45-04:00 Don't use substTyUnchecked in newMetaTyVar There were some comments that explained that we needed to use an unchecked substitution function because of issue #12931, but that has since been fixed, so we should be able to use substTy instead now. - - - - - c7bbad9a by sheaf at 2023-07-17T02:48:19-04:00 rnImports: var shouldn't import NoFldSelectors In an import declaration such as import M ( var ) the import of the variable "var" should **not** bring into scope record fields named "var" which are defined with NoFieldSelectors. Doing so can cause spurious "unused import" warnings, as reported in ticket #23557. Fixes #23557 - - - - - 1af2e773 by sheaf at 2023-07-17T02:48:19-04:00 Suggest similar names in imports This commit adds similar name suggestions when importing. For example module A where { spelling = 'o' } module B where { import B ( speling ) } will give rise to the error message: Module ‘A’ does not export ‘speling’. Suggested fix: Perhaps use ‘spelling’ This also provides hints when users try to import record fields defined with NoFieldSelectors. - - - - - 654fdb98 by Alan Zimmerman at 2023-07-17T02:48:55-04:00 EPA: Store leading AnnSemi for decllist in al_rest This simplifies the markAnnListA implementation in ExactPrint - - - - - 22565506 by sheaf at 2023-07-17T21:12:59-04:00 base: add COMPLETE pragma to BufferCodec PatSyn This implements CLC proposal #178, rectifying an oversight in the implementation of CLC proposal #134 which could lead to spurious pattern match warnings. https://github.com/haskell/core-libraries-committee/issues/178 https://github.com/haskell/core-libraries-committee/issues/134 - - - - - 860f6269 by sheaf at 2023-07-17T21:13:00-04:00 exactprint: silence incomplete record update warnings - - - - - df706de3 by sheaf at 2023-07-17T21:13:00-04:00 Re-instate -Wincomplete-record-updates Commit e74fc066 refactored the handling of record updates to use the HsExpanded mechanism. This meant that the pattern matching inherent to a record update was considered to be "generated code", and thus we stopped emitting "incomplete record update" warnings entirely. This commit changes the "data Origin = Source | Generated" datatype, adding a field to the Generated constructor to indicate whether we still want to perform pattern-match checking. We also have to do a bit of plumbing with HsCase, to record that the HsCase arose from an HsExpansion of a RecUpd, so that the error message continues to mention record updates as opposed to a generic "incomplete pattern matches in case" error. Finally, this patch also changes the way we handle inaccessible code warnings. Commit e74fc066 was also a regression in this regard, as we were emitting "inaccessible code" warnings for case statements spuriously generated when desugaring a record update (remember: the desugaring mechanism happens before typechecking; it thus can't take into account e.g. GADT information in order to decide which constructors to include in the RHS of the desugaring of the record update). We fix this by changing the mechanism through which we disable inaccessible code warnings: we now check whether we are in generated code in GHC.Tc.Utils.TcMType.newImplication in order to determine whether to emit inaccessible code warnings. Fixes #23520 Updates haddock submodule, to avoid incomplete record update warnings - - - - - 1d05971e by sheaf at 2023-07-17T21:13:00-04:00 Propagate long-distance information in do-notation The preceding commit re-enabled pattern-match checking inside record updates. This revealed that #21360 was in fact NOT fixed by e74fc066. This commit makes sure we correctly propagate long-distance information in do blocks, e.g. in ```haskell data T = A { fld :: Int } | B f :: T -> Maybe T f r = do a at A{} <- Just r Just $ case a of { A _ -> A 9 } ``` we need to propagate the fact that "a" is headed by the constructor "A" to see that the case expression "case a of { A _ -> A 9 }" cannot fail. Fixes #21360 - - - - - bea0e323 by sheaf at 2023-07-17T21:13:00-04:00 Skip PMC for boring patterns Some patterns introduce no new information to the pattern-match checker (such as plain variable or wildcard patterns). We can thus skip doing any pattern-match checking on them when the sole purpose for doing so was introducing new long-distance information. See Note [Boring patterns] in GHC.Hs.Pat. Doing this avoids regressing in performance now that we do additional pattern-match checking inside do notation. - - - - - ddcdd88c by Rodrigo Mesquita at 2023-07-17T21:13:36-04:00 Split GHC.Platform.ArchOS from ghc-boot into ghc-platform Split off the `GHC.Platform.ArchOS` module from the `ghc-boot` package into this reinstallable standalone package which abides by the PVP, in part motivated by the ongoing work on `ghc-toolchain` towards runtime retargetability. - - - - - b55a8ea7 by Sylvain Henry at 2023-07-17T21:14:27-04:00 JS: better implementation for plusWord64 (#23597) - - - - - 889c2bbb by sheaf at 2023-07-18T06:37:32-04:00 Do primop rep-poly checks when instantiating This patch changes how we perform representation-polymorphism checking for primops (and other wired-in Ids such as coerce). When instantiating the primop, we check whether each type variable is required to instantiated to a concrete type, and if so we create a new concrete metavariable (a ConcreteTv) instead of a simple MetaTv. (A little subtlety is the need to apply the substitution obtained from instantiating to the ConcreteTvOrigins, see Note [substConcreteTvOrigin] in GHC.Tc.Utils.TcMType.) This allows us to prevent representation-polymorphism in non-argument position, as that is required for some of these primops. We can also remove the logic in tcRemainingValArgs, except for the part concerning representation-polymorphic unlifted newtypes. The function has been renamed rejectRepPolyNewtypes; all it does now is reject unsaturated occurrences of representation-polymorphic newtype constructors when the representation of its argument isn't a concrete RuntimeRep (i.e. still a PHASE 1 FixedRuntimeRep check). The Note [Eta-expanding rep-poly unlifted newtypes] in GHC.Tc.Gen.Head gives more explanation about a possible path to PHASE 2, which would be in line with the treatment for primops taken in this patch. We also update the Core Lint check to handle this new framework. This means Core Lint now checks representation-polymorphism in continuation position like needed for catch#. Fixes #21906 ------------------------- Metric Increase: LargeRecord ------------------------- - - - - - 00648e5d by Krzysztof Gogolewski at 2023-07-18T06:38:10-04:00 Core Lint: distinguish let and letrec in locations Lint messages were saying "in the body of letrec" even for non-recursive let. I've also renamed BodyOfLetRec to BodyOfLet in stg, since there's no separate letrec. - - - - - 787bae96 by Krzysztof Gogolewski at 2023-07-18T06:38:50-04:00 Use extended literals when deriving Show This implements GHC proposal https://github.com/ghc-proposals/ghc-proposals/pull/596 Also add support for Int64# and Word64#; see testcase ShowPrim. - - - - - 257f1567 by Jaro Reinders at 2023-07-18T06:39:29-04:00 Add StgFromCore and StgCodeGen linting - - - - - 34d08a20 by Ben Gamari at 2023-07-19T03:33:22-04:00 Reg.Liveness: Strictness - - - - - c5deaa27 by Ben Gamari at 2023-07-19T03:33:22-04:00 Reg.Liveness: Don't repeatedly construct UniqSets - - - - - b947250b by Ben Gamari at 2023-07-19T03:33:22-04:00 compiler/Types: Ensure that fromList-type operations can fuse In #20740 I noticed that mkUniqSet does not fuse. In practice, allowing it to do so makes a considerable difference in allocations due to the backend. Metric Decrease: T12707 T13379 T3294 T4801 T5321FD T5321Fun T783 - - - - - 6c88c2ba by Sven Tennie at 2023-07-19T03:33:59-04:00 x86 Codegen: Implement MO_S_MulMayOflo for W16 - - - - - 5f1154e0 by Sven Tennie at 2023-07-19T03:33:59-04:00 x86 CodeGen: MO_S_MulMayOflo better error message for rep > W64 It's useful to see which value made the pattern match fail. (If it ever occurs.) - - - - - e8c9a95f by Sven Tennie at 2023-07-19T03:33:59-04:00 x86 CodeGen: Implement MO_S_MulMayOflo for W8 This case wasn't handled before. But, the test-primops test suite showed that it actually might appear. - - - - - a36f9dc9 by Sven Tennie at 2023-07-19T03:33:59-04:00 Add test for %mulmayoflo primop The test expects a perfect implementation with no false positives. - - - - - 38a36248 by Matthew Pickering at 2023-07-19T03:34:36-04:00 lint-ci-config: Generate jobs-metadata.json We also now save the jobs-metadata.json and jobs.yaml file as artifacts as: * It might be useful for someone who is modifying CI to copy jobs.yaml if they are having trouble regenerating locally. * jobs-metadata.json is very useful for downstream pipelines to work out the right job to download. Fixes #23654 - - - - - 1535a671 by Vladislav Zavialov at 2023-07-19T03:35:12-04:00 Initialize 9.10.1-notes.rst Create new release notes for the next GHC release (GHC 9.10) - - - - - 3bd4d5b5 by sheaf at 2023-07-19T03:35:53-04:00 Prioritise Parent when looking up class sub-binder When we look up children GlobalRdrElts of a given Parent, we sometimes would rather prioritise those GlobalRdrElts which have the right Parent, and sometimes prioritise those that have the right NameSpace: - in export lists, we should prioritise NameSpace - for class/instance binders, we should prioritise Parent See Note [childGREPriority] in GHC.Types.Name.Reader. fixes #23664 - - - - - 9c8fdda3 by Alan Zimmerman at 2023-07-19T03:36:29-04:00 EPA: Improve annotation management in getMonoBind Ensure the LHsDecl for a FunBind has the correct leading comments and trailing annotations. See the added note for details. - - - - - ff884b77 by Matthew Pickering at 2023-07-19T11:42:02+01:00 Remove unused files in .gitlab These were left over after 6078b429 - - - - - 29ef590c by Matthew Pickering at 2023-07-19T11:42:52+01:00 gen_ci: Add hie.yaml file This allows you to load `gen_ci.hs` into HLS, and now it is a huge module, that is quite useful. - - - - - 808b55cf by Matthew Pickering at 2023-07-19T12:24:41+01:00 ci: Make "fast-ci" the default validate configuration We are trying out a lighter weight validation pipeline where by default we just test on 5 platforms: * x86_64-deb10-slow-validate * windows * x86_64-fedora33-release * aarch64-darwin * aarch64-linux-deb10 In order to enable the "full" validation pipeline you can apply the `full-ci` label which will enable all the validation pipelines. All the validation jobs are still run on a marge batch. The goal is to reduce the overall CI capacity so that pipelines start faster for MRs and marge bot batches are faster. Fixes #23694 - - - - - 0b23db03 by Alan Zimmerman at 2023-07-20T05:28:47-04:00 EPA: Simplify GHC/Parser.y sL1 This is the next patch in a series simplifying location management in GHC/Parser.y This one simplifies sL1, to use the HasLoc instances introduced in !10743 (closed) - - - - - 3ece9856 by Ben Gamari at 2023-07-21T07:30:45-04:00 nativeGen: Explicitly set flags of text sections on Windows The binutils documentation (for COFF) claims, > If no flags are specified, the default flags depend upon the section > name. If the section name is not recognized, the default will be for the > section to be loaded and writable. We previously assumed that this would do the right thing for split sections (e.g. a section named `.text$foo` would be correctly inferred to be a text section). However, we have observed that this is not the case (at least under the clang toolchain used on Windows): when split-sections is enabled, text sections are treated by the assembler as data (matching the "default" behavior specified by the documentation). Avoid this by setting section flags explicitly. This should fix split sections on Windows. Fixes #22834. - - - - - db7f7240 by Ben Gamari at 2023-07-21T07:30:45-04:00 nativeGen: Set explicit section types on all platforms - - - - - b444c16f by Finley McIlwaine at 2023-07-21T07:31:28-04:00 Insert documentation into parsed signature modules Causes haddock comments in signature modules to be properly inserted into the AST (just as they are for regular modules) if the `-haddock` flag is given. Also adds a test that compares `-ddump-parsed-ast` output for a signature module to prevent further regressions. Fixes #23315 - - - - - f9b952a7 by Ben Gamari at 2023-07-21T11:58:05-04:00 Bump base bound to <4.20 For GHC 9.8. - - - - - c30cea53 by Ben Gamari at 2023-07-21T23:23:49-04:00 primops: Introduce unsafeThawByteArray# This addresses an odd asymmetry in the ByteArray# primops, which previously provided unsafeFreezeByteArray# but no corresponding thaw operation. Closes #22710 - - - - - 87f9bd47 by Ben Gamari at 2023-07-21T23:23:49-04:00 testsuite: Elaborate in interface stability README This discussion didn't make it into the original MR. - - - - - e4350b41 by Matthew Pickering at 2023-07-21T23:24:25-04:00 Allow users to override non-essential haddock options in a Flavour We now supply the non-essential options to haddock using the `extraArgs` field, which can be specified in a Flavour so that if an advanced user wants to change how documentation is generated then they can use something other than the `defaultHaddockExtraArgs`. This does have the potential to regress some packaging if a user has overridden `extraArgs` themselves, because now they also need to add the haddock options to extraArgs. This can easily be done by appending `defaultHaddockExtraArgs` to their extraArgs invocation but someone might not notice this behaviour has changed. In any case, I think passing the non-essential options in this manner is the right thing to do and matches what we do for the "ghc" builder, which by default doesn't pass any optmisation levels, and would likewise be very bad if someone didn't pass suitable `-O` levels for builds. Fixes #23625 - - - - - fc186b0c by Ilias Tsitsimpis at 2023-07-21T23:25:03-04:00 ghc-prim: Link against libatomic Commit b4d39adbb58 made 'hs_cmpxchg64()' available to all architectures. Unfortunately this made GHC to fail to build on armel, since armel needs libatomic to support atomic operations on 64-bit word sizes. Configure libraries/ghc-prim/ghc-prim.cabal to link against libatomic, the same way as we do in rts/rts.cabal. - - - - - 4f5538a8 by Matthew Pickering at 2023-07-21T23:25:39-04:00 simplifier: Correct InScopeSet in rule matching The in-scope set passedto the `exprIsLambda_maybe` call lacked all the in-scope binders. @simonpj suggests this fix where we augment the in-scope set with the free variables of expression which fixes this failure mode in quite a direct way. Fixes #23630 - - - - - 5ad8d597 by Krzysztof Gogolewski at 2023-07-21T23:26:17-04:00 Add a test for #23413 It was fixed by commit e1590ddc661d6: Add the SolverStage monad. - - - - - 7e05f6df by sheaf at 2023-07-21T23:26:56-04:00 Finish migration of diagnostics in GHC.Tc.Validity This patch finishes migrating the error messages in GHC.Tc.Validity to use the new diagnostic infrastructure. It also refactors the error message datatypes for class and family instances, to common them up under a single datatype as much as possible. - - - - - 4876fddc by Matthew Pickering at 2023-07-21T23:27:33-04:00 ci: Enable some more jobs to run in a marge batch In !10907 I made the majority of jobs not run on a validate pipeline but then forgot to renable a select few jobs on the marge batch MR. - - - - - 026991d7 by Jens Petersen at 2023-07-21T23:28:13-04:00 user_guide/flags.py: python-3.12 no longer includes distutils packaging.version seems able to handle this fine - - - - - b91bbc2b by Matthew Pickering at 2023-07-21T23:28:50-04:00 ci: Mention ~full-ci label in MR template We mention that if you need a full validation pipeline then you can apply the ~full-ci label to your MR in order to test against the full validation pipeline (like we do for marge). - - - - - 42b05e9b by sheaf at 2023-07-22T12:36:00-04:00 RTS: declare setKeepCAFs symbol Commit 08ba8720 failed to declare the dependency of keepCAFsForGHCi on the symbol setKeepCAFs in the RTS, which led to undefined symbol errors on Windows, as exhibited by the testcase frontend001. Thanks to Moritz Angermann and Ryan Scott for the diagnosis and fix. Fixes #22961 - - - - - a72015d6 by sheaf at 2023-07-22T12:36:01-04:00 Mark plugins-external as broken on Windows This test is broken on Windows, so we explicitly mark it as such now that we stop skipping plugin tests on Windows. - - - - - cb9c93d7 by sheaf at 2023-07-22T12:36:01-04:00 Stop marking plugin tests as fragile on Windows Now that b2bb3e62 has landed we are in a better situation with regards to plugins on Windows, allowing us to unmark many plugin tests as fragile. Fixes #16405 - - - - - a7349217 by Krzysztof Gogolewski at 2023-07-22T12:36:37-04:00 Misc cleanup - Remove unused RDR names - Fix typos in comments - Deriving: simplify boxConTbl and remove unused litConTbl - chmod -x GHC/Exts.hs, this seems accidental - - - - - 33b6850a by Vladislav Zavialov at 2023-07-23T10:27:37-04:00 Visible forall in types of terms: Part 1 (#22326) This patch implements part 1 of GHC Proposal #281, introducing explicit `type` patterns and `type` arguments. Summary of the changes: 1. New extension flag: RequiredTypeArguments 2. New user-facing syntax: `type p` patterns (represented by EmbTyPat) `type e` expressions (represented by HsEmbTy) 3. Functions with required type arguments (visible forall) can now be defined and applied: idv :: forall a -> a -> a -- signature (relevant change: checkVdqOK in GHC/Tc/Validity.hs) idv (type a) (x :: a) = x -- definition (relevant change: tcPats in GHC/Tc/Gen/Pat.hs) x = idv (type Int) 42 -- usage (relevant change: tcInstFun in GHC/Tc/Gen/App.hs) 4. template-haskell support: TH.TypeE corresponds to HsEmbTy TH.TypeP corresponds to EmbTyPat 5. Test cases and a new User's Guide section Changes *not* included here are the t2t (term-to-type) transformation and term variable capture; those belong to part 2. - - - - - 73b5c7ce by sheaf at 2023-07-23T10:28:18-04:00 Add test for #22424 This is a simple Template Haskell test in which we refer to record selectors by their exact Names, in two different ways. Fixes #22424 - - - - - 83cbc672 by Ben Gamari at 2023-07-24T07:40:49+00:00 ghc-toolchain: Initial commit - - - - - 31dcd26c by Rodrigo Mesquita at 2023-07-24T07:40:49+00:00 ghc-toolchain: Toolchain Selection This commit integrates ghc-toolchain, the brand new way of configuring toolchains for GHC, with the Hadrian build system, with configure, and extends and improves the first iteration of ghc-toolchain. The general overview is * We introduce a program invoked `ghc-toolchain --triple=...` which, when run, produces a file with a `Target`. A `GHC.Toolchain.Target.Target` describes the properties of a target and the toolchain (executables and configured flags) to produce code for that target * Hadrian was modified to read Target files, and will both * Invoke the toolchain configured in the Target file as needed * Produce a `settings` file for GHC based on the Target file for that stage * `./configure` will invoke ghc-toolchain to generate target files, but it will also generate target files based on the flags configure itself configured (through `.in` files that are substituted) * By default, the Targets generated by configure are still (for now) the ones used by Hadrian * But we additionally validate the Target files generated by ghc-toolchain against the ones generated by configure, to get a head start on catching configuration bugs before we transition completely. * When we make that transition, we will want to drop a lot of the toolchain configuration logic from configure, but keep it otherwise. * For each compiler stage we should have 1 target file (up to a stage compiler we can't run in our machine) * We just have a HOST target file, which we use as the target for stage0 * And a TARGET target file, which we use for stage1 (and later stages, if not cross compiling) * Note there is no BUILD target file, because we only support cross compilation where BUILD=HOST * (for more details on cross-compilation see discussion on !9263) See also * Note [How we configure the bundled windows toolchain] * Note [ghc-toolchain consistency checking] * Note [ghc-toolchain overview] Ticket: #19877 MR: !9263 - - - - - a732b6d3 by Rodrigo Mesquita at 2023-07-24T07:40:49+00:00 Add flag to enable/disable ghc-toolchain based configurations This flag is disabled by default, and we'll use the configure-generated-toolchains by default until we remove the toolchain configuration logic from configure. - - - - - 61eea240 by Rodrigo Mesquita at 2023-07-24T07:40:49+00:00 Split ghc-toolchain executable to new packge In light of #23690, we split the ghc-toolchain executable out of the library package to be able to ship it in the bindist using Hadrian. Ideally, we eventually revert this commit. - - - - - 38e795ff by Rodrigo Mesquita at 2023-07-24T07:40:49+00:00 Ship ghc-toolchain in the bindist Add the ghc-toolchain binary to the binary distribution we ship to users, and teach the bindist configure to use the existing ghc-toolchain. - - - - - 32cae784 by Matthew Craven at 2023-07-24T16:48:24-04:00 Kill off gen_bytearray_addr_access_ops.py The relevant primop descriptions are now generated directly by genprimopcode. This makes progress toward fixing #23490, but it is not a complete fix since there is more than one way in which cabal-reinstall (hadrian/build build-cabal) is broken. - - - - - 02e6a6ce by Matthew Pickering at 2023-07-24T16:49:00-04:00 compiler: Remove unused `containers.h` include Fixes #23712 - - - - - 822ef66b by Matthew Pickering at 2023-07-25T08:44:50-04:00 Fix pretty printing of WARNING pragmas There is still something quite unsavoury going on with WARNING pragma printing because the printing relies on the fact that for decl deprecations the SourceText of WarningTxt is empty. However, I let that lion sleep and just fixed things directly. Fixes #23465 - - - - - e7b38ede by Matthew Pickering at 2023-07-25T08:45:28-04:00 ci-images: Bump to commit which has 9.6 image The test-bootstrap job has been failing for 9.6 because we accidentally used a non-master commit. - - - - - bb408936 by Matthew Pickering at 2023-07-25T08:45:28-04:00 Update bootstrap plans for 9.6.2 and 9.4.5 - - - - - 355e1792 by Alan Zimmerman at 2023-07-26T10:17:32-04:00 EPA: Simplify GHC/Parser.y comb4/comb5 Use the HasLoc instance from Ast.hs to allow comb4/comb5 to work with anything with a SrcSpan Also get rid of some more now unnecessary reLoc calls. - - - - - 9393df83 by Gavin Zhao at 2023-07-26T10:18:16-04:00 compiler: make -ddump-asm work with wasm backend NCG Fixes #23503. Now the `-ddump-asm` flag is respected in the wasm backend NCG, so developers can directly view the generated ASM instead of needing to pass `-S` or `-keep-tmp-files` and manually find & open the assembly file. Ideally, we should be able to output the assembly files in smaller chunks like in other NCG backends. This would also make dumping assembly stats easier. However, this would require a large refactoring, so for short-term debugging purposes I think the current approach works fine. Signed-off-by: Gavin Zhao <git at gzgz.dev> - - - - - 79463036 by Krzysztof Gogolewski at 2023-07-26T10:18:54-04:00 llvm: Restore accidentally deleted code in 0fc5cb97 Fixes #23711 - - - - - 20db7e26 by Rodrigo Mesquita at 2023-07-26T10:19:33-04:00 configure: Default missing options to False when preparing ghc-toolchain Targets This commit fixes building ghc with 9.2 as the boostrap compiler. The ghc-toolchain patch assumed all _STAGE0 options were available, and forgot to account for this missing information in 9.2. Ghc 9.2 does not have in settings whether ar supports -l, hence can't report it with --info (unliked 9.4 upwards). The fix is to default the missing information (we default "ar supports -l" and other missing options to False) - - - - - fac9e84e by Naïm Favier at 2023-07-26T10:20:16-04:00 docs: Fix typo - - - - - 503fd647 by Bartłomiej Cieślar at 2023-07-26T17:23:10-04:00 This MR is an implementation of the proposal #516. It adds a warning -Wincomplete-record-selectors for usages of a record field access function (either a record selector or getField @"rec"), while trying to silence the warning whenever it can be sure that a constructor without the record field would not be invoked (which would otherwise cause the program to fail). For example: data T = T1 | T2 {x :: Bool} f a = x a -- this would throw an error g T1 = True g a = x a -- this would not throw an error h :: HasField "x" r Bool => r -> Bool h = getField @"x" j :: T -> Bool j = h -- this would throw an error because of the `HasField` -- constraint being solved See the tests DsIncompleteRecSel* and TcIncompleteRecSel for more examples of the warning. See Note [Detecting incomplete record selectors] in GHC.HsToCore.Expr for implementation details - - - - - af6fdf42 by Arnaud Spiwack at 2023-07-26T17:23:52-04:00 Fix user-facing label in MR template - - - - - 5d45b92a by Matthew Pickering at 2023-07-27T05:46:46-04:00 ci: Test bootstrapping configurations with full-ci and on marge batches There have been two incidents recently where bootstrapping has been broken by removing support for building with 9.2.*. The process for bumping the minimum required version starts with bumping the configure version and then other CI jobs such as the bootstrap jobs have to be updated. We must not silently bump the minimum required version. Now we are running a slimmed down validate pipeline it seems worthwile to test these bootstrap configurations in the full-ci pipeline. - - - - - 25d4fee7 by Matthew Pickering at 2023-07-27T05:46:46-04:00 bootstrap: Remove ghc-9_2_* plans We are anticipating shortly making it necessary to use ghc-9.4 to boot the compiler. - - - - - 2f66da16 by Matthew Pickering at 2023-07-27T05:46:46-04:00 Update bootstrap plans for ghc-platform and ghc-toolchain dependencies Fixes #23735 - - - - - c8c6eab1 by Matthew Pickering at 2023-07-27T05:46:46-04:00 bootstrap: Disable -selftest flag from bootstrap plans This saves on building one dependency (QuickCheck) which is unecessary for bootstrapping. - - - - - a80ca086 by Andrew Lelechenko at 2023-07-27T05:47:26-04:00 Link reference paper and package from System.Mem.{StableName,Weak} - - - - - a5319358 by David Knothe at 2023-07-28T13:13:10-04:00 Update Match Datatype EquationInfo currently contains a list of the equation's patterns together with a CoreExpr that is to be evaluated after a successful match on this equation. All the match-functions only operate on the first pattern of an equation - after successfully matching it, match is called recursively on the tail of the pattern list. We can express this more clearly and make the code a little more elegant by updating the datatype of EquationInfo as follows: data EquationInfo = EqnMatch { eqn_pat = Pat GhcTc, eqn_rest = EquationInfo } | EqnDone { eqn_rhs = MatchResult CoreExpr } An EquationInfo now explicitly exposes its first pattern which most functions operate on, and exposes the equation that remains after processing the first pattern. An EqnDone signifies an empty equation where the CoreExpr can now be evaluated. - - - - - 86ad1af9 by David Binder at 2023-07-28T13:13:53-04:00 Improve documentation for Data.Fixed - - - - - f8fa1d08 by Ben Gamari at 2023-07-28T13:14:31-04:00 ghc-prim: Use C11 atomics Previously `ghc-prim`'s atomic wrappers used the legacy `__sync_*` family of C builtins. Here we refactor these to rather use the appropriate C11 atomic equivalents, allowing us to be more explicit about the expected ordering semantics. - - - - - 0bfc8908 by Finley McIlwaine at 2023-07-28T18:46:26-04:00 Include -haddock in DynFlags fingerprint The -haddock flag determines whether or not the resulting .hi files contain haddock documentation strings. If the existing .hi files do not contain haddock documentation strings and the user requests them, we should recompile. - - - - - 40425c50 by Andreas Klebinger at 2023-07-28T18:47:02-04:00 Aarch64 NCG: Use encoded immediates for literals. Try to generate instr x2, <imm> instead of mov x1, lit instr x2, x1 When possible. This get's rid if quite a few redundant mov instructions. I believe this causes a metric decrease for LargeRecords as we reduce register pressure. ------------------------- Metric Decrease: LargeRecord ------------------------- - - - - - e9a0fa3f by Andrew Lelechenko at 2023-07-28T18:47:42-04:00 Bump filepath submodule to 1.4.100.4 Resolves #23741 Metric Decrease: MultiComponentModules MultiComponentModulesRecomp MultiLayerModules MultiLayerModulesRecomp T10421 T12234 T12425 T13035 T13701 T13719 T16875 T18304 T18698a T18698b T21839c T9198 TcPlugin_RewritePerf hard_hole_fits Metric decrease on Windows can be probably attributed to https://github.com/haskell/filepath/pull/183 - - - - - ee93edfd by Andrew Lelechenko at 2023-07-28T18:48:21-04:00 Add since pragmas to GHC.IO.Handle.FD - - - - - d0369802 by Simon Peyton Jones at 2023-07-30T09:24:48+01:00 Make the occurrence analyser smarter about join points This MR addresses #22404. There is a big Note Note [Occurrence analysis for join points] that explains it all. Significant changes * New field occ_join_points in OccEnv * The NonRec case of occAnalBind splits into two cases: one for existing join points (which does the special magic for Note [Occurrence analysis for join points], and one for other bindings. * mkOneOcc adds in info from occ_join_points. * All "bring into scope" activity is centralised in the new function `addInScope`. * I made a local data type LocalOcc for use inside the occurrence analyser It is like OccInfo, but lacks IAmDead and IAmALoopBreaker, which in turn makes computationns over it simpler and more efficient. * I found quite a bit of allocation in GHC.Core.Rules.getRules so I optimised it a bit. More minor changes * I found I was using (Maybe Arity) a lot, so I defined a new data type JoinPointHood and used it everwhere. This touches a lot of non-occ-anal files, but it makes everything more perspicuous. * Renamed data constructor WithUsageDetails to WUD, and WithTailUsageDetails to WTUD This also fixes #21128, on the way. --------- Compiler perf ----------- I spent quite a time on performance tuning, so even though it does more than before, the occurrence analyser runs slightly faster on average. Here are the compile-time allocation changes over 0.5% CoOpt_Read(normal) ghc/alloc 766,025,520 754,561,992 -1.5% CoOpt_Singletons(normal) ghc/alloc 759,436,840 762,925,512 +0.5% LargeRecord(normal) ghc/alloc 1,814,482,440 1,799,530,456 -0.8% PmSeriesT(normal) ghc/alloc 68,159,272 67,519,720 -0.9% T10858(normal) ghc/alloc 120,805,224 118,746,968 -1.7% T11374(normal) ghc/alloc 164,901,104 164,070,624 -0.5% T11545(normal) ghc/alloc 79,851,808 78,964,704 -1.1% T12150(optasm) ghc/alloc 73,903,664 71,237,544 -3.6% GOOD T12227(normal) ghc/alloc 333,663,200 331,625,864 -0.6% T12234(optasm) ghc/alloc 52,583,224 52,340,344 -0.5% T12425(optasm) ghc/alloc 81,943,216 81,566,720 -0.5% T13056(optasm) ghc/alloc 294,517,928 289,642,512 -1.7% T13253-spj(normal) ghc/alloc 118,271,264 59,859,040 -49.4% GOOD T15164(normal) ghc/alloc 1,102,630,352 1,091,841,296 -1.0% T15304(normal) ghc/alloc 1,196,084,000 1,166,733,000 -2.5% T15630(normal) ghc/alloc 148,729,632 147,261,064 -1.0% T15703(normal) ghc/alloc 379,366,664 377,600,008 -0.5% T16875(normal) ghc/alloc 32,907,120 32,670,976 -0.7% T17516(normal) ghc/alloc 1,658,001,888 1,627,863,848 -1.8% T17836(normal) ghc/alloc 395,329,400 393,080,248 -0.6% T18140(normal) ghc/alloc 71,968,824 73,243,040 +1.8% T18223(normal) ghc/alloc 456,852,568 453,059,088 -0.8% T18282(normal) ghc/alloc 129,105,576 131,397,064 +1.8% T18304(normal) ghc/alloc 71,311,712 70,722,720 -0.8% T18698a(normal) ghc/alloc 208,795,112 210,102,904 +0.6% T18698b(normal) ghc/alloc 230,320,736 232,697,976 +1.0% BAD T19695(normal) ghc/alloc 1,483,648,128 1,504,702,976 +1.4% T20049(normal) ghc/alloc 85,612,024 85,114,376 -0.6% T21839c(normal) ghc/alloc 415,080,992 410,906,216 -1.0% GOOD T4801(normal) ghc/alloc 247,590,920 250,726,272 +1.3% T6048(optasm) ghc/alloc 95,699,416 95,080,680 -0.6% T783(normal) ghc/alloc 335,323,384 332,988,120 -0.7% T9233(normal) ghc/alloc 709,641,224 685,947,008 -3.3% GOOD T9630(normal) ghc/alloc 965,635,712 948,356,120 -1.8% T9675(optasm) ghc/alloc 444,604,152 428,987,216 -3.5% GOOD T9961(normal) ghc/alloc 303,064,592 308,798,800 +1.9% BAD WWRec(normal) ghc/alloc 503,728,832 498,102,272 -1.1% geo. mean -1.0% minimum -49.4% maximum +1.9% In fact these figures seem to vary between platforms; generally worse on i386 for some reason. The Windows numbers vary by 1% espec in benchmarks where the total allocation is low. But the geom mean stays solidly negative, which is good. The "increase/decrease" list below covers all platforms. The big win on T13253-spj comes because it has a big nest of join points, each occurring twice in the next one. The new occ-anal takes only one iteration of the simplifier to do the inlining; the old one took four. Moreover, we get much smaller code with the new one: New: Result size of Tidy Core = {terms: 429, types: 84, coercions: 0, joins: 14/14} Old: Result size of Tidy Core = {terms: 2,437, types: 304, coercions: 0, joins: 10/10} --------- Runtime perf ----------- No significant changes in nofib results, except a 1% reduction in compiler allocation. Metric Decrease: CoOpt_Read T13253-spj T9233 T9630 T9675 T12150 T21839c LargeRecord MultiComponentModulesRecomp T10421 T13701 T10421 T13701 T12425 Metric Increase: T18140 T9961 T18282 T18698a T18698b T19695 - - - - - 42aa7fbd by Julian Ospald at 2023-07-30T17:22:01-04:00 Improve documentation around IOException and ioe_filename See: * https://github.com/haskell/core-libraries-committee/issues/189 * https://github.com/haskell/unix/pull/279 * https://github.com/haskell/unix/pull/289 - - - - - 33598ecb by Sylvain Henry at 2023-08-01T14:45:54-04:00 JS: implement getMonotonicTime (fix #23687) - - - - - d2bedffd by Bartłomiej Cieślar at 2023-08-01T14:46:40-04:00 Implementation of the Deprecated Instances proposal #575 This commit implements the ability to deprecate certain instances, which causes the compiler to emit the desired deprecation message whenever they are instantiated. For example: module A where class C t where instance {-# DEPRECATED "dont use" #-} C Int where module B where import A f :: C t => t f = undefined g :: Int g = f -- "dont use" emitted here The implementation is as follows: - In the parser, we parse deprecations/warnings attached to instances: instance {-# DEPRECATED "msg" #-} Show X deriving instance {-# WARNING "msg2" #-} Eq Y (Note that non-standalone deriving instance declarations do not support this mechanism.) - We store the resulting warning message in `ClsInstDecl` (respectively, `DerivDecl`). In `GHC.Tc.TyCl.Instance.tcClsInstDecl` (respectively, `GHC.Tc.Deriv.Utils.newDerivClsInst`), we pass on that information to `ClsInst` (and eventually store it in `IfaceClsInst` too). - Finally, when we solve a constraint using such an instance, in `GHC.Tc.Instance.Class.matchInstEnv`, we emit the appropriate warning that was stored in `ClsInst`. Note that we only emit a warning when the instance is used in a different module than it is defined, which keeps the behaviour in line with the deprecation of top-level identifiers. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - d5a65af6 by Ben Gamari at 2023-08-01T14:47:18-04:00 compiler: Style fixes - - - - - 7218c80a by Ben Gamari at 2023-08-01T14:47:19-04:00 rts: Fix implicit cast This ensures that Task.h can be built with a C++ compiler. - - - - - d6d5aafc by Ben Gamari at 2023-08-01T14:47:19-04:00 testsuite: Fix warning in hs_try_putmvar001 - - - - - d9eddf7a by Ben Gamari at 2023-08-01T14:47:19-04:00 testsuite: Add AtomicModifyIORef test - - - - - f9eea4ba by Ben Gamari at 2023-08-01T14:47:19-04:00 rts: Introduce NO_WARN macro This allows fine-grained ignoring of warnings. - - - - - 497b24ec by Ben Gamari at 2023-08-01T14:47:19-04:00 rts: Simplify atomicModifyMutVar2# implementation Previously we would perform a redundant load in the non-threaded RTS in atomicModifyMutVar2# implementation for the benefit of the non-moving GC's write barrier. Eliminate this. - - - - - 52ee082b by Ben Gamari at 2023-08-01T14:47:19-04:00 rts: Introduce more principled fence operations - - - - - cd3c0377 by Ben Gamari at 2023-08-01T14:47:19-04:00 rts: Introduce SET_INFO_RELAXED - - - - - 6df2352a by Ben Gamari at 2023-08-01T14:47:19-04:00 rts: Style fixes - - - - - 4ef6f319 by Ben Gamari at 2023-08-01T14:47:19-04:00 codeGen/tsan: Rework handling of spilling - - - - - f9ca7e27 by Ben Gamari at 2023-08-01T14:47:19-04:00 hadrian: More debug information - - - - - df4153ac by Ben Gamari at 2023-08-01T14:47:19-04:00 Improve TSAN documentation - - - - - fecae988 by Ben Gamari at 2023-08-01T14:47:19-04:00 hadrian: More selective TSAN instrumentation - - - - - 465a9a0b by Alan Zimmerman at 2023-08-01T14:47:56-04:00 EPA: Provide correct annotation span for ImportDecl Use the whole declaration, rather than just the span of the 'import' keyword. Metric Decrease: T9961 T5205 Metric Increase: T13035 - - - - - ae63d0fa by Bartłomiej Cieślar at 2023-08-01T14:48:40-04:00 Add cases to T23279: HasField for deprecated record fields This commit adds additional tests from ticket #23279 to ensure that we don't regress on reporting deprecated record fields in conjunction with HasField, either when using overloaded record dot syntax or directly through `getField`. Fixes #23279 - - - - - 00fb6e6b by Andreas Klebinger at 2023-08-01T14:49:17-04:00 AArch NCG: Pure refactor Combine some alternatives. Add some line breaks for overly long lines - - - - - 8f3b3b78 by Andreas Klebinger at 2023-08-01T14:49:54-04:00 Aarch ncg: Optimize immediate use for address calculations When the offset doesn't fit into the immediate we now just reuse the general getRegister' code path which is well optimized to compute the offset into a register instead of a special case for CmmRegOff. This means we generate a lot less code under certain conditions which is why performance metrics for these improve. ------------------------- Metric Decrease: T4801 T5321FD T5321Fun ------------------------- - - - - - 74a882dc by MorrowM at 2023-08-02T06:00:03-04:00 Add a RULE to make lookup fuse See https://github.com/haskell/core-libraries-committee/issues/175 Metric Increase: T18282 - - - - - cca74dab by Ben Gamari at 2023-08-02T06:00:39-04:00 hadrian: Ensure that way-flags are passed to CC Previously the way-specific compilation flags (e.g. `-DDEBUG`, `-DTHREADED_RTS`) would not be passed to the CC invocations. This meant that C dependency files would not correctly reflect dependencies predicated on the way, resulting in the rather painful #23554. Closes #23554. - - - - - 622b483c by Jaro Reinders at 2023-08-02T06:01:20-04:00 Native 32-bit Enum Int64/Word64 instances This commits adds more performant Enum Int64 and Enum Word64 instances for 32-bit platforms, replacing the Integer-based implementation. These instances are a copy of the Enum Int and Enum Word instances with minimal changes to manipulate Int64 and Word64 instead. On i386 this yields a 1.5x performance increase and for the JavaScript back end it even yields a 5.6x speedup. Metric Decrease: T18964 - - - - - c8bd7fa4 by Sylvain Henry at 2023-08-02T06:02:03-04:00 JS: fix typos in constants (#23650) - - - - - b9d5bfe9 by Josh Meredith at 2023-08-02T06:02:40-04:00 JavaScript: update MK_TUP macros to use current tuple constructors (#23659) - - - - - 28211215 by Matthew Pickering at 2023-08-02T06:03:19-04:00 ci: Pass -Werror when building hadrian in hadrian-ghc-in-ghci job Warnings when building Hadrian can end up cluttering the output of HLS, and we've had bug reports in the past about these warnings when building Hadrian. It would be nice to turn on -Werror on at least one build of Hadrian in CI to avoid a patch introducing warnings when building Hadrian. Fixes #23638 - - - - - aca20a5d by Ben Gamari at 2023-08-02T06:03:55-04:00 codeGen: Ensure that TSAN is aware of writeArray# write barriers By using a proper release store instead of a fence. - - - - - 453c0531 by Ben Gamari at 2023-08-02T06:03:55-04:00 codeGen: Ensure that array reads have necessary barriers This was the cause of #23541. - - - - - 93a0d089 by Arnaud Spiwack at 2023-08-02T06:04:37-04:00 Add test for #23550 - - - - - 6a2f4a20 by Arnaud Spiwack at 2023-08-02T06:04:37-04:00 Desugar non-recursive lets to non-recursive lets (take 2) This reverts commit 522bd584f71ddeda21efdf0917606ce3d81ec6cc. And takes care of the case that I missed in my previous attempt. Namely the case of an AbsBinds with no type variables and no dictionary variable. Ironically, the comment explaining why non-recursive lets were desugared to recursive lets were pointing specifically at this case as the reason. I just failed to understand that it was until Simon PJ pointed it out to me. See #23550 for more discussion. - - - - - ff81d53f by jade at 2023-08-02T06:05:20-04:00 Expand documentation of List & Data.List This commit aims to improve the documentation and examples of symbols exported from Data.List - - - - - fa4e5913 by Jade at 2023-08-02T06:06:03-04:00 Improve documentation of Semigroup & Monoid This commit aims to improve the documentation of various symbols exported from Data.Semigroup and Data.Monoid - - - - - 1b27e151 by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Check for puns (see ghc#23368) - - - - - 457341fd by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Remove fake exports for (~), List, and Tuple<n> The old reasoning no longer applies, nowadays those names can be mentioned in export lists. - - - - - bf3dcddf by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Fix pretty-printing of Solo and MkSolo - - - - - e2c91bff by Gergő Érdi at 2023-08-03T02:55:46+01:00 Desugar bindings in the context of their evidence Closes #23172 - - - - - 481f4a46 by Gergő Érdi at 2023-08-03T07:48:43+01:00 Add flag to `-f{no-}specialise-incoherents` to enable/disable specialisation of incoherent instances Fixes #23287 - - - - - d751c583 by Profpatsch at 2023-08-04T12:24:26-04:00 base: Improve String & IsString documentation - - - - - 01db1117 by Ben Gamari at 2023-08-04T12:25:02-04:00 rts/win32: Ensure reliability of IO manager shutdown When the Win32 threaded IO manager shuts down, `ioManagerDie` sends an `IO_MANAGER_DIE` event to the IO manager thread using the `io_manager_event` event object. Finally, it will closes the event object, and invalidate `io_manager_event`. Previously, `readIOManagerEvent` would see that `io_manager_event` is invalid and return `0`, suggesting that everything is right with the world. This meant that if `ioManagerDie` invalidated the handle before the event manager was blocked on the event we would end up in a situation where the event manager would never realize it was asked to shut down. Fix this by ensuring that `readIOManagerEvent` instead returns `IO_MANAGER_DIE` when we detect that the event object has been invalidated by `ioManagerDie`. Fixes #23691. - - - - - fdef003a by Ryan Scott at 2023-08-04T12:25:39-04:00 Look through TH splices in splitHsApps This modifies `splitHsApps` (a key function used in typechecking function applications) to look through untyped TH splices and quasiquotes. Not doing so was the cause of #21077. This builds on !7821 by making `splitHsApps` match on `HsUntypedSpliceTop`, which contains the `ThModFinalizers` that must be run as part of invoking the TH splice. See the new `Note [Looking through Template Haskell splices in splitHsApps]` in `GHC.Tc.Gen.Head`. Along the way, I needed to make the type of `splitHsApps.set` slightly more general to accommodate the fact that the location attached to a quasiquote is a `SrcAnn NoEpAnns` rather than a `SrcSpanAnnA`. Fixes #21077. - - - - - e77a0b41 by Ben Gamari at 2023-08-04T12:26:15-04:00 Bump deepseq submodule to 1.5. And bump bounds (cherry picked from commit 1228d3a4a08d30eaf0138a52d1be25b38339ef0b) - - - - - cebb5819 by Ben Gamari at 2023-08-04T12:26:15-04:00 configure: Bump minimal boot GHC version to 9.4 (cherry picked from commit d3ffdaf9137705894d15ccc3feff569d64163e8e) - - - - - 83766dbf by Ben Gamari at 2023-08-04T12:26:15-04:00 template-haskell: Bump version to 2.21.0.0 Bumps exceptions submodule. (cherry picked from commit bf57fc9aea1196f97f5adb72c8b56434ca4b87cb) - - - - - 1211112a by Ben Gamari at 2023-08-04T12:26:15-04:00 base: Bump version to 4.19 Updates all boot library submodules. (cherry picked from commit 433d99a3c24a55b14ec09099395e9b9641430143) - - - - - 3ab5efd9 by Ben Gamari at 2023-08-04T12:26:15-04:00 testsuite: Normalise versions more aggressively In backpack hashes can contain `+` characters. (cherry picked from commit 024861af51aee807d800e01e122897166a65ea93) - - - - - d52be957 by Ben Gamari at 2023-08-04T12:26:15-04:00 testsuite: Declare bkpcabal08 as fragile Due to spurious output changes described in #23648. (cherry picked from commit c046a2382420f2be2c4a657c56f8d95f914ea47b) - - - - - e75a58d1 by Ben Gamari at 2023-08-04T12:26:15-04:00 gitlab-ci: Only mark linker_unload_native as broken in static jobs This test passes on dynamically-linked Alpine. (cherry picked from commit f356a7e8ec8ec3d6b2b30fd175598b9b80065d87) - - - - - 8b176514 by Ben Gamari at 2023-08-04T12:26:15-04:00 testsuite: Update base-exports - - - - - 4b647936 by Ben Gamari at 2023-08-04T12:26:15-04:00 testsuite/interface-stability: normalise versions This eliminates spurious changes from version bumps. - - - - - 0eb54c05 by Ben Gamari at 2023-08-04T12:26:51-04:00 linker/PEi386: Don't sign-extend symbol section number Previously we incorrectly interpreted PE section numbers as signed values. However, this isn't the case; rather, it's an unsigned 16-bit number with a few special bit-patterns (0xffff and 0xfffe). This resulted in #22941 as the linker would conclude that the sections were invalid. Fixing this required quite a bit of refactoring. Closes #22941. - - - - - fd7ce39c by Ben Gamari at 2023-08-04T12:27:28-04:00 testsuite: Mark MulMayOflo_full as broken rather than skipping To ensure that we don't accidentally fix it. See #23742. - - - - - 824092f2 by Ben Gamari at 2023-08-04T12:27:28-04:00 nativeGen/AArch64: Fix sign extension in MulMayOflo Previously the 32-bit implementations of MulMayOflo would use the a non-sensical sign-extension mode. Rewrite these to reflect what gcc 11 produces. Also similarly rework the 16- and 8-bit cases. This now passes the MulMayOflo tests in ghc/test-primops> in all four widths, including the precision tests. Fixes #23721. - - - - - 1b15dbc4 by Jan Hrček at 2023-08-04T12:28:08-04:00 Fix haddock markup in code example for coerce - - - - - 46fd8ced by Vladislav Zavialov at 2023-08-04T12:28:44-04:00 Fix (~) and (@) infix operators in TH splices (#23748) 8168b42a "Whitespace-sensitive bang patterns" allows GHC to accept the following infix operators: a ~ b = () a @ b = () But not if TH is used to generate those declarations: $([d| a ~ b = () a @ b = () |]) -- Test.hs:5:2: error: [GHC-55017] -- Illegal variable name: ‘~’ -- When splicing a TH declaration: (~_0) a_1 b_2 = GHC.Tuple.Prim.() This is easily fixed by modifying `reservedOps` in GHC.Utils.Lexeme - - - - - a1899d8f by Aaron Allen at 2023-08-04T12:29:24-04:00 [#23663] Show Flag Suggestions in GHCi Makes suggestions when using `:set` in GHCi with a misspelled flag. This mirrors how invalid flags are handled when passed to GHC directly. Logic for producing flag suggestions was moved to GHC.Driver.Sesssion so it can be shared. resolves #23663 - - - - - 03f2debd by Rodrigo Mesquita at 2023-08-04T12:30:00-04:00 Improve ghc-toolchain validation configure warning Fixes the layout of the ghc-toolchain validation warning produced by configure. - - - - - de25487d by Alan Zimmerman at 2023-08-04T12:30:36-04:00 EPA make getLocA a synonym for getHasLoc This is basically a no-op change, but allows us to make future changes that can rely on the HasLoc instances And I presume this means we can use more precise functions based on class resolution, so the Windows CI build reports Metric Decrease: T12234 T13035 - - - - - 3ac423b9 by Ben Gamari at 2023-08-04T12:31:13-04:00 ghc-platform: Add upper bound on base Hackage upload requires this. - - - - - 8ba20b21 by Matthew Craven at 2023-08-04T17:22:59-04:00 Adjust and clarify handling of primop effects Fixes #17900; fixes #20195. The existing "can_fail" and "has_side_effects" primop attributes that previously governed this were used in inconsistent and confusingly-documented ways, especially with regard to raising exceptions. This patch replaces them with a single "effect" attribute, which has four possible values: NoEffect, CanFail, ThrowsException, and ReadWriteEffect. These are described in Note [Classifying primop effects]. A substantial amount of related documentation has been re-drafted for clarity and accuracy. In the process of making this attribute format change for literally every primop, several existing mis-classifications were detected and corrected. One of these mis-classifications was tagToEnum#, which is now considered CanFail; this particular fix is known to cause a regression in performance for derived Enum instances. (See #23782.) Fixing this is left as future work. New primop attributes "cheap" and "work_free" were also added, and used in the corresponding parts of GHC.Core.Utils. In view of their actual meaning and uses, `primOpOkForSideEffects` and `exprOkForSideEffects` have been renamed to `primOpOkToDiscard` and `exprOkToDiscard`, respectively. Metric Increase: T21839c - - - - - 41bf2c09 by sheaf at 2023-08-04T17:23:42-04:00 Update inert_solved_dicts for ImplicitParams When adding an implicit parameter dictionary to the inert set, we must make sure that it replaces any previous implicit parameter dictionaries that overlap, in order to get the appropriate shadowing behaviour, as in let ?x = 1 in let ?x = 2 in ?x We were already doing this for inert_cans, but we weren't doing the same thing for inert_solved_dicts, which lead to the bug reported in #23761. The fix is thus to make sure that, when handling an implicit parameter dictionary in updInertDicts, we update **both** inert_cans and inert_solved_dicts to ensure a new implicit parameter dictionary correctly shadows old ones. Fixes #23761 - - - - - 43578d60 by Matthew Craven at 2023-08-05T01:05:36-04:00 Bump bytestring submodule to 0.11.5.1 - - - - - 91353622 by Ben Gamari at 2023-08-05T01:06:13-04:00 Initial commit of Note [Thunks, blackholes, and indirections] This Note attempts to summarize the treatment of thunks, thunk update, and indirections. This fell out of work on #23185. - - - - - 8d686854 by sheaf at 2023-08-05T01:06:54-04:00 Remove zonk in tcVTA This removes the zonk in GHC.Tc.Gen.App.tc_inst_forall_arg and its accompanying Note [Visible type application zonk]. Indeed, this zonk is no longer necessary, as we no longer maintain the invariant that types are well-kinded without zonking; only that typeKind does not crash; see Note [The Purely Kinded Type Invariant (PKTI)]. This commit removes this zonking step (as well as a secondary zonk), and replaces the aforementioned Note with the explanatory Note [Type application substitution], which justifies why the substitution performed in tc_inst_forall_arg remains valid without this zonking step. Fixes #23661 - - - - - 19dea673 by Ben Gamari at 2023-08-05T01:07:30-04:00 Bump nofib submodule Ensuring that nofib can be build using the same range of bootstrap compilers as GHC itself. - - - - - aa07402e by Luite Stegeman at 2023-08-05T23:15:55+09:00 JS: Improve compatibility with recent emsdk The JavaScript code in libraries/base/jsbits/base.js had some hardcoded offsets for fields in structs, because we expected the layout of the data structures to remain unchanged. Emsdk 3.1.42 changed the layout of the stat struct, breaking this assumption, and causing code in .hsc files accessing the stat struct to fail. This patch improves compatibility with recent emsdk by removing the assumption that data layouts stay unchanged: 1. offsets of fields in structs used by JavaScript code are now computed by the configure script, so both the .js and .hsc files will automatically use the new layout if anything changes. 2. the distrib/configure script checks that the emsdk version on a user's system is the same version that a bindist was booted with, to avoid data layout inconsistencies See #23641 - - - - - b938950d by Luite Stegeman at 2023-08-07T06:27:51-04:00 JS: Fix missing local variable declarations This fixes some missing local variable declarations that were found by running the testsuite in strict mode. Fixes #23775 - - - - - 6c0e2247 by sheaf at 2023-08-07T13:31:21-04:00 Update Haddock submodule to fix #23368 This submodule update adds the following three commits: bbf1c8ae - Check for puns 0550694e - Remove fake exports for (~), List, and Tuple<n> 5877bceb - Fix pretty-printing of Solo and MkSolo These commits fix the issues with Haddock HTML rendering reported in ticket #23368. Fixes #23368 - - - - - 5b5be3ea by Matthew Pickering at 2023-08-07T13:32:00-04:00 Revert "Bump bytestring submodule to 0.11.5.1" This reverts commit 43578d60bfc478e7277dcd892463cec305400025. Fixes #23789 - - - - - 01961be3 by Ben Gamari at 2023-08-08T02:47:14-04:00 configure: Derive library version from ghc-prim.cabal.in Since ghc-prim.cabal is now generated by Hadrian, we cannot depend upon it. Closes #23726. - - - - - 3b373838 by Ryan Scott at 2023-08-08T02:47:49-04:00 tcExpr: Push expected types for untyped TH splices inwards In !10911, I deleted a `tcExpr` case for `HsUntypedSplice` in favor of a much simpler case that simply delegates to `tcApp`. Although this passed the test suite at the time, this was actually an error, as the previous `tcExpr` case was critically pushing the expected type inwards. This actually matters for programs like the one in #23796, which GHC would not accept with type inference alone—we need full-blown type _checking_ to accept these. I have added back the previous `tcExpr` case for `HsUntypedSplice` and now explain why we have two different `HsUntypedSplice` cases (one in `tcExpr` and another in `splitHsApps`) in `Note [Looking through Template Haskell splices in splitHsApps]` in `GHC.Tc.Gen.Head`. Fixes #23796. - - - - - 0ef1d8ae by sheaf at 2023-08-08T21:26:51-04:00 Compute all emitted diagnostic codes This commit introduces in GHC.Types.Error.Codes the function constructorCodes :: forall diag. (...) => Map DiagnosticCode String which computes a collection of all the diagnostic codes that correspond to a particular type. In particular, we can compute the collection of all diagnostic codes emitted by GHC using the invocation constructorCodes @GhcMessage We then make use of this functionality in the new "codes" test which checks consistency and coverage of GHC diagnostic codes. It performs three checks: - check 1: all non-outdated GhcDiagnosticCode equations are statically used. - check 2: all outdated GhcDiagnosticCode equations are statically unused. - check 3: all statically used diagnostic codes are covered by the testsuite (modulo accepted exceptions). - - - - - 4bc7b1e5 by Fraser Tweedale at 2023-08-08T21:27:32-04:00 numberToRangedRational: fix edge cases for exp ≈ (maxBound :: Int) Currently a negative exponent less than `minBound :: Int` results in Infinity, which is very surprising and obviously wrong. ``` λ> read "1e-9223372036854775808" :: Double 0.0 λ> read "1e-9223372036854775809" :: Double Infinity ``` There is a further edge case where the exponent can overflow when increased by the number of tens places in the integer part, or underflow when decreased by the number of leading zeros in the fractional part if the integer part is zero: ``` λ> read "10e9223372036854775807" :: Double 0.0 λ> read "0.01e-9223372036854775808" :: Double Infinity ``` To resolve both of these issues, perform all arithmetic and comparisons involving the exponent in type `Integer`. This approach also eliminates the need to explicitly check the exponent against `maxBound :: Int` and `minBound :: Int`, because the allowed range of the exponent (i.e. the result of `floatRange` for the target floating point type) is certainly within those bounds. This change implements CLC proposal 192: https://github.com/haskell/core-libraries-committee/issues/192 - - - - - 6eab07b2 by Alan Zimmerman at 2023-08-08T21:28:10-04:00 EPA: Remove Location from WarningTxt source This is not needed. - - - - - 1a98d673 by Sebastian Graf at 2023-08-09T16:24:29-04:00 Cleanup a TODO introduced in 1f94e0f7 The change must have slipped through review of !4412 - - - - - 2274abc8 by Sebastian Graf at 2023-08-09T16:24:29-04:00 More explicit strictness in GHC.Real - - - - - ce8aa54c by Sebastian Graf at 2023-08-09T16:24:30-04:00 exprIsTrivial: Factor out shared implementation The duplication between `exprIsTrivial` and `getIdFromTrivialExpr_maybe` has been bugging me for a long time. This patch introduces an inlinable worker function `trivial_expr_fold` acting as the single, shared decision procedure of triviality. It "returns" a Church-encoded `Maybe (Maybe Id)`, so when it is inlined, it fuses to similar code as before. (Better code, even, in the case of `getIdFromTrivialExpr` which presently allocates a `Just` constructor that cancels away after this patch.) - - - - - d004a36d by Sebastian Graf at 2023-08-09T16:24:30-04:00 Simplify: Simplification of arguments in a single function The Simplifier had a function `simplArg` that wasn't called in `rebuildCall`, which seems to be the main way to simplify args. Hence I consolidated the code path to call `simplArg`, too, renaming to `simplLazyArg`. - - - - - 8c73505e by Sebastian Graf at 2023-08-09T16:24:30-04:00 Core.Ppr: Omit case binder for empty case alternatives A minor improvement to pretty-printing - - - - - d8d993f1 by Sebastian Graf at 2023-08-09T16:24:30-04:00 Disable tests RepPolyWrappedVar2 and RepPolyUnsafeCoerce1 in JS backend ... because those coerce between incompatible/unknown PrimReps. - - - - - f06e87e4 by Sebastian Graf at 2023-08-09T16:24:30-04:00 Inlining literals into boring contexts is OK - - - - - 4a6b7c87 by Sebastian Graf at 2023-08-09T16:24:30-04:00 Clarify floating of unsafeEqualityProofs (#23754) - - - - - b0f4752e by Sebastian Graf at 2023-08-09T16:24:30-04:00 Kill SetLevel.notWorthFloating.is_triv (#23270) We have had it since b84ba676034, when it operated on annotated expressions. Nowadays it operates on vanilla `CoreExpr` though, so we should just call `exprIsTrivial`; thus handling empty cases and string literals correctly. - - - - - 7e0c8b3b by Sebastian Graf at 2023-08-09T16:24:30-04:00 ANFise string literal arguments (#23270) This instates the invariant that a trivial CoreExpr translates to an atomic StgExpr. Nice. Annoyingly, in -O0 we sometimes generate ``` foo = case "blah"# of sat { __DEFAULT -> unpackCString# sat } ``` which makes it a bit harder to spot that we can emit a standard `stg_unpack_cstring` thunk. Fixes #23270. - - - - - 357f2738 by Sebastian Graf at 2023-08-09T16:24:30-04:00 Deactivate -fcatch-nonexhaustive-cases in ghc-bignum (#23345) - - - - - 59202c80 by Sebastian Graf at 2023-08-09T16:24:30-04:00 CorePrep: Eliminate EmptyCase and unsafeEqualityProof in CoreToStg instead We eliminate EmptyCase by way of `coreToStg (Case e _ _ []) = coreToStg e` now. The main reason is that it plays far better in conjunction with eta expansion (as we aim to do for arguments in CorePrep, #23083), because we can discard any arguments, `(case e of {}) eta == case e of {}`, whereas in `(e |> co) eta` it's impossible to discard the argument. We do also give the same treatment to unsafeCoerce proofs and treat them as trivial iff their RHS is trivial. It is also both much simpler to describe than the previous mechanism of emitting an unsafe coercion and simpler to implement, removing quite a bit of commentary and `CorePrepProv`. In the ghc/alloc perf test `LargeRecord`, we introduce an additional Simplifier iteration due to #17910. E.g., FloatOut produces a binding ``` lvl_s6uK [Occ=Once1] :: GHC.Types.Int [LclId] lvl_s6uK = GHC.Types.I# 2# lvl_s6uL [Occ=Once1] :: GHC.Types.Any [LclId] lvl_s6uL = case Unsafe.Coerce.unsafeEqualityProof ... of { Unsafe.Coerce.UnsafeRefl v2_i6tr -> lvl_s6uK `cast` (... v2_i6tr ...) } ``` That occurs once and hence is pre-inlined unconditionally in the next Simplifier pass. It's non-trivial to find a way around that, but not really harmful otherwise. Hence we accept a 1.2% increase on some architectures. Metric Increase: LargeRecord - - - - - 00d31188 by Sebastian Graf at 2023-08-09T16:24:30-04:00 CorePrep: Eta expand arguments (#23083) Previously, we'd only eta expand let bindings and lambdas, now we'll also eta expand arguments such as in T23083: ```hs g f h = f (h `seq` (h $)) ``` Unless `-fpedantic-bottoms` is set, we'll now transform to ```hs g f h = f (\eta -> h eta) ``` in CorePrep. See the new `Note [Eta expansion of arguments in CorePrep]` for the details. We only do this optimisation with -O2 because we saw 2-3% ghc/alloc regressions in T4801 and T5321FD. Fixes #23083. - - - - - bf885d7a by Matthew Craven at 2023-08-09T16:25:07-04:00 Bump bytestring submodule to 0.11.5, again Fixes #23789. The bytestring commit used here is unreleased; a release can be made when necessary. - - - - - 7acbf0fd by Sven Tennie at 2023-08-10T19:17:11-04:00 Serialize CmmRetInfo in .rodata The handling of case was missing. - - - - - 0c3136f2 by Sven Tennie at 2023-08-10T19:17:11-04:00 Reference StgRetFun payload by its struct field address This is easier to grasp than relative pointer offsets. - - - - - f68ff313 by Sven Tennie at 2023-08-10T19:17:11-04:00 Better variable name: u -> frame The 'u' was likely introduced by copy'n'paste. - - - - - 0131bb7f by Sven Tennie at 2023-08-10T19:17:11-04:00 Make checkSTACK() public Such that it can also be used in tests. - - - - - 7b6e1e53 by Sven Tennie at 2023-08-10T19:17:11-04:00 Publish stack related fields in DerivedConstants.h These will be used in ghc-heap to decode these parts of the stack. - - - - - 907ed054 by Sven Tennie at 2023-08-10T19:17:11-04:00 ghc-heap: Decode StgStack and its stack frames Previously, ghc-heap could only decode heap closures. The approach is explained in detail in note [Decoding the stack]. - - - - - 6beb6ac2 by Sven Tennie at 2023-08-10T19:17:11-04:00 Remove RetFunType from RetFun stack frame representation It's a technical detail. The single usage is replaced by a predicate. - - - - - 006bb4f3 by Sven Tennie at 2023-08-10T19:17:11-04:00 Better parameter name The call-site uses the term "offset", too. - - - - - d4c2c1af by Sven Tennie at 2023-08-10T19:17:11-04:00 Make closure boxing pure There seems to be no need to do something complicated. However, the strictness of the closure pointer matters, otherwise a thunk gets decoded. - - - - - 8d8426c9 by Sven Tennie at 2023-08-10T19:17:11-04:00 Document entertainGC in test It wasn't obvious why it's there and what its role is. Also, increase the "entertainment level" a bit. I checked in STG and Cmm dumps that this really generates closures (and is not e.g. constant folded away.) - - - - - cc52c358 by Finley McIlwaine at 2023-08-10T19:17:47-04:00 Add -dipe-stats flag This is useful for seeing which info tables have information. - - - - - 261c4acb by Finley McIlwaine at 2023-08-10T19:17:47-04:00 Add -finfo-table-map-with-fallback -finfo-table-map-with-stack The -fno-info-table-map-with-stack flag omits STACK info tables from the info table map, and the -fno-info-table-map-with-fallback flag omits info tables with defaulted source locations from the map. In a test on the Agda codebase the build results were about 7% smaller when both of those types of tables were omitted. Adds a test that verifies that passing each combination of these flags results in the correct output for -dipe-stats, which is disabled for the js backend since profiling is not implemented. This commit also refactors a lot of the logic around extracting info tables from the Cmm results and building the info table map. This commit also fixes some issues in the users guide rst source to fix warnings that were noticed while debugging the documentation for these flags. Fixes #23702 - - - - - d7047e0d by Jaro Reinders at 2023-08-14T04:41:42-04:00 Add changelog entry for specialised Enum Int64/Word64 instances - - - - - 52f5e8fb by cydparser at 2023-08-14T04:42:20-04:00 Fix -ddump-to-file and -ddump-timings interaction (#20316) - - - - - 1274c5d6 by cydparser at 2023-08-14T04:42:20-04:00 Update release notes (#20316) - - - - - 8e699b23 by Matthew Pickering at 2023-08-14T10:44:47-04:00 base: Add changelog entry for CLC #188 This proposal modified the implementations of copyBytes, moveBytes and fillBytes (as detailed in the proposal) https://github.com/haskell/core-libraries-committee/issues/188 - - - - - 026f040a by Matthew Pickering at 2023-08-14T10:45:23-04:00 packaging: Build manpage in separate directory to other documentation We were installing two copies of the manpage: * One useless one in the `share/doc` folder, because we copy the doc/ folder into share/ * The one we deliberately installed into `share/man` etc The solution is to build the manpage into the `manpage` directory when building the bindist, and then just install it separately. Fixes #23707 - - - - - 524c60c8 by Bartłomiej Cieślar at 2023-08-14T13:46:33-04:00 Report deprecated fields bound by record wildcards when used This commit ensures that we emit the appropriate warnings when a deprecated record field bound by a record wildcard is used. For example: module A where data Foo = Foo {x :: Int, y :: Bool, z :: Char} {-# DEPRECATED x "Don't use x" #-} {-# WARNING y "Don't use y" #-} module B where import A foo (Foo {..}) = x This will cause us to emit a "Don't use x" warning, with location the location of the record wildcard. Note that we don't warn about `y`, because it is unused in the RHS of `foo`. Fixes #23382 - - - - - d6130065 by Matthew Pickering at 2023-08-14T13:47:11-04:00 Add zstd suffix to jobs which rely on zstd This was causing some confusion as the job was named simply "x86_64-linux-deb10-validate", which implies a standard configuration rather than any dependency on libzstd. - - - - - e24e44fc by Matthew Pickering at 2023-08-14T13:47:11-04:00 ci: Always run project-version job This is needed for the downstream test-primops pipeline to workout what the version of a bindist produced by a pipeline is. - - - - - f17b9d62 by Matthew Pickering at 2023-08-14T13:47:11-04:00 gen_ci: Rework how jobs-metadata.json is generated * We now represent a job group a triple of Maybes, which makes it easier to work out when jobs are enabled/disabled on certain pipelines. ``` data JobGroup a = StandardTriple { v :: Maybe (NamedJob a) , n :: Maybe (NamedJob a) , r :: Maybe (NamedJob a) } ``` * `jobs-metadata.json` generation is reworked using the following algorithm. - For each pipeline type, find all the platforms we are doing builds for. - Select one build per platform - Zip together the results This way we can choose different pipelines for validate/nightly/release which makes the metadata also useful for validate pipelines. This feature is used by the test-primops downstream CI in order to select the right bindist for testing validate pipelines. This makes it easier to inspect which jobs are going to be enabled on a particular pipeline. - - - - - f9a5563d by Matthew Pickering at 2023-08-14T13:47:11-04:00 gen_ci: Rules rework In particular we now distinguish between whether we are dealing with a Nightly/Release pipeline (which labels don't matter for) and a validate pipeline where labels do matter. The overall goal here is to allow a disjunction of labels for validate pipelines, for example, > Run a job if we have the full-ci label or test-primops label Therefore the "ValidateOnly" rules are treated as a set of disjunctions rather than conjunctions like before. What this means in particular is that if we want to ONLY run a job if a label is set, for example, "FreeBSD" label then we have to override the whole label set. Fixes #23772 - - - - - d54b0c1d by Matthew Pickering at 2023-08-14T13:47:11-04:00 ci: set -e for lint-ci-config scripts - - - - - 994a9b35 by Matthew Pickering at 2023-08-14T13:47:11-04:00 ci: Fix job metadata generation - - - - - e194ed2b by Ben Gamari at 2023-08-15T00:58:09-04:00 users-guide: Note that GHC2021 doesn't include ExplicitNamespaces As noted in #23801. - - - - - d814bda9 by Ben Gamari at 2023-08-15T05:43:53-04:00 users-guide: Support both distutils and packaging As noted in #23818, some old distributions (e.g. Debian 9) only include `distutils` while newer distributions only include `packaging`. Fixes #23818. - - - - - 1726db3f by Ben Gamari at 2023-08-15T05:43:53-04:00 users-guide: Ensure extlinks is compatible with Sphinx <4 The semantics of the `extlinks` attribute annoyingly changed in Sphinx 4. Reflect this in our configuration. See #22690. Fixes #23807. - - - - - 173338cf by Matthew Pickering at 2023-08-15T22:00:24-04:00 ci: Run full-ci on master and release branches Fixes #23737 - - - - - bdab6898 by Andrew Lelechenko at 2023-08-15T22:01:03-04:00 Add @since pragmas for Data.Ord.clamp and GHC.Float.clamp - - - - - 662d351b by Matthew Pickering at 2023-08-16T09:35:04-04:00 ghc-toolchain: Match CPP args with configure script At the moment we need ghc-toolchain to precisely match the output as provided by the normal configure script. The normal configure script (FP_HSCPP_CMD_WITH_ARGS) branches on whether we are using clang or gcc so we match that logic exactly in ghc-toolchain. The old implementation (which checks if certain flags are supported) is better but for now we have to match to catch any potential errors in the configuration. Ticket: #23720 - - - - - 09c6759e by Matthew Pickering at 2023-08-16T09:35:04-04:00 configure: Fix `-Wl,--no-as-needed` check The check was failing because the args supplied by $$1 were quoted which failed because then the C compiler thought they were an input file. Fixes #23720 - - - - - 2129678b by Matthew Pickering at 2023-08-16T09:35:04-04:00 configure: Add flag which turns ghc-toolchain check into error We want to catch these errors in CI, but first we need to a flag which turns this check into an error. - - - - - 6e2aa8e0 by Matthew Pickering at 2023-08-16T09:35:04-04:00 ci: Enable --enable-strict-ghc-toolchain-check for all CI jobs This will cause any CI job to fail if we have a mismatch between what ghc-toolchain reports and what ./configure natively reports. Fixing these kinds of issues is highest priority for 9.10 release. - - - - - 12d39e24 by Rodrigo Mesquita at 2023-08-16T09:35:04-04:00 Pass user-specified options to ghc-toolchain The current user interface to configuring target toolchains is `./configure`. In !9263 we added a new tool to configure target toolchains called `ghc-toolchain`, but the blessed way of creating these toolchains is still through configure. However, we were not passing the user-specified options given with the `./configure` invocation to the ghc-toolchain tool. This commit remedies that by storing the user options and environment variables in USER_* variables, which then get passed to GHC-toolchain. The exception to the rule is the windows bundled toolchain, which overrides the USER_* variables with whatever flags the windows bundled toolchain requires to work. We consider the bundled toolchain to be effectively the user specifying options, since the actual user delegated that configuration work. Closes #23678 - - - - - f7b3c3a0 by Rodrigo Mesquita at 2023-08-16T09:35:04-04:00 ghc-toolchain: Parse javascript and ghcjs as a Arch and OS - - - - - 8a0ae4ee by Rodrigo Mesquita at 2023-08-16T09:35:04-04:00 ghc-toolchain: Fix ranlib option - - - - - 31e9ec96 by Rodrigo Mesquita at 2023-08-16T09:35:04-04:00 Check Link Works with -Werror - - - - - bc1998b3 by Matthew Pickering at 2023-08-16T09:35:04-04:00 Only check for no_compact_unwind support on darwin While writing ghc-toolchain we noticed that the FP_PROG_LD_NO_COMPACT_UNWIND check is subtly wrong. Specifically, we pass -Wl,-no_compact_unwind to cc. However, ld.gold interprets this as -n o_compact_unwind, which is a valid argument. Fixes #23676 - - - - - 0283f36e by Matthew Pickering at 2023-08-16T09:35:04-04:00 Add some javascript special cases to ghc-toolchain On javascript there isn't a choice of toolchain but some of the configure checks were not accurately providing the correct answer. 1. The linker was reported as gnu LD because the --version output mentioned gnu LD. 2. The --target flag makes no sense on javascript but it was just ignored by the linker, so we add a special case to stop ghc-toolchain thinking that emcc supports --target when used as a linker. - - - - - a48ec5f8 by Matthew Pickering at 2023-08-16T09:35:04-04:00 check for emcc in gnu_LD check - - - - - 50df2e69 by Matthew Pickering at 2023-08-16T09:35:04-04:00 Add ldOverrideWhitelist to only default to ldOverride on windows/linux On some platforms - ie darwin, javascript etc we really do not want to allow the user to use any linker other than the default one as this leads to all kinds of bugs. Therefore it is a bit more prudant to add a whitelist which specifies on which platforms it might be possible to use a different linker. - - - - - a669a39c by Matthew Pickering at 2023-08-16T09:35:04-04:00 Fix plaform glob in FPTOOLS_SET_C_LD_FLAGS A normal triple may look like x86_64-unknown-linux but when cross-compiling you get $target set to a quad such as.. aarch64-unknown-linux-gnu Which should also match this check. - - - - - c52b6769 by Matthew Pickering at 2023-08-16T09:35:04-04:00 ghc-toolchain: Pass ld-override onto ghc-toolchain - - - - - 039b484f by Matthew Pickering at 2023-08-16T09:35:04-04:00 ld override: Make whitelist override user given option - - - - - d2b63cbc by Matthew Pickering at 2023-08-16T09:35:05-04:00 ghc-toolchain: Add format mode to normalise differences before diffing. The "format" mode takes an "--input" and "--ouput" target file and formats it. This is intended to be useful on windows where the configure/ghc-toolchain target files can't be diffed very easily because the path separators are different. - - - - - f2b39e4a by Matthew Pickering at 2023-08-16T09:35:05-04:00 ghc-toolchain: Bump ci-images commit to get new ghc-wasm-meta We needed to remove -Wno-unused-command-line-argument from the arguments passed in order for the configure check to report correctly. See https://gitlab.haskell.org/ghc/ghc/-/merge_requests/10976#note_516335 - - - - - 92103830 by Matthew Pickering at 2023-08-16T09:35:05-04:00 configure: MergeObjsCmd - distinguish between empty string and unset variable If `MergeObjsCmd` is explicitly set to the empty string then we should assume that MergeObjs is just not supported. This is especially important for windows where we set MergeObjsCmd to "" in m4/fp_setup_windows_toolchain.m4. - - - - - 3500bb2c by Matthew Pickering at 2023-08-16T09:35:05-04:00 configure: Add proper check to see if object merging works - - - - - 08c9a014 by Matthew Pickering at 2023-08-16T09:35:05-04:00 ghc-toolchain: If MergeObjsCmd is not set, replace setting with Nothing If the user explicitly chooses to not set a MergeObjsCmd then it is correct to use Nothing for tgtMergeObjs field in the Target file. - - - - - c9071d94 by Matthew Pickering at 2023-08-16T09:35:05-04:00 HsCppArgs: Augment the HsCppOptions This is important when we pass -I when setting up the windows toolchain. - - - - - 294a6d80 by Matthew Pickering at 2023-08-16T09:35:05-04:00 Set USER_CPP_ARGS when setting up windows toolchain - - - - - bde4b5d4 by Rodrigo Mesquita at 2023-08-16T09:35:05-04:00 Improve handling of Cc as a fallback - - - - - f4c1c3a3 by Rodrigo Mesquita at 2023-08-16T09:35:05-04:00 ghc-toolchain: Configure Cpp and HsCpp correctly when user specifies flags In ghc-toolchain, we were only /not/ configuring required flags when the user specified any flags at all for the of the HsCpp and Cpp tools. Otherwise, the linker takes into consideration the user specified flags to determine whether to search for a better linker implementation, but already configured the remaining GHC and platform-specific flags regardless of the user options. Other Tools consider the user options as a baseline for further configuration (see `findProgram`), so #23689 is not applicable. Closes #23689 - - - - - bfe4ffac by Matthew Pickering at 2023-08-16T09:35:05-04:00 CPP_ARGS: Put new options after user specified options This matches up with the behaviour of ghc-toolchain, so that the output of both matches. - - - - - a6828173 by Gergő Érdi at 2023-08-16T09:35:41-04:00 If a defaulting plugin made progress, re-zonk wanteds before built-in defaulting Fixes #23821. - - - - - e2b38115 by Sylvain Henry at 2023-08-17T07:54:06-04:00 JS: implement openat(AT_FDCWD...) (#23697) Use `openSync` to implement `openat(AT_FDCWD...)`. - - - - - a975c663 by sheaf at 2023-08-17T07:54:47-04:00 Use unsatisfiable for missing methods w/ defaults When a class instance has an Unsatisfiable constraint in its context and the user has not explicitly provided an implementation of a method, we now always provide a RHS of the form `unsatisfiable @msg`, even if the method has a default definition available. This ensures that, when deferring type errors, users get the appropriate error message instead of a possible runtime loop, if class default methods were defined recursively. Fixes #23816 - - - - - 45ca51e5 by Ben Gamari at 2023-08-17T15:16:41-04:00 ghc-internal: Initial commit of the skeleton - - - - - 88bbf8c5 by Ben Gamari at 2023-08-17T15:16:41-04:00 ghc-experimental: Initial commit - - - - - 664468c0 by Ben Gamari at 2023-08-17T15:17:17-04:00 testsuite/cloneStackLib: Fix incorrect format specifiers - - - - - eaa835bb by Ben Gamari at 2023-08-17T15:17:17-04:00 rts/ipe: Fix const-correctness of IpeBufferListNode Both info tables and the string table should be `const` - - - - - 78f6f6fd by Ben Gamari at 2023-08-17T15:17:17-04:00 nonmoving: Drop dead debugging utilities These are largely superceded by support in the ghc-utils GDB extension. - - - - - 3f6e8f42 by Ben Gamari at 2023-08-17T15:17:17-04:00 nonmoving: Refactor management of mark thread Here we refactor that treatment of the worker thread used by the nonmoving GC for concurrent marking, avoiding creating a new thread with every major GC cycle. As well, the new scheme is considerably easier to reason about, consolidating all state in one place, accessed via a small set of accessors with clear semantics. - - - - - 88c32b7d by Ben Gamari at 2023-08-17T15:17:17-04:00 testsuite: Skip T23221 in nonmoving GC ways This test is very dependent upon GC behavior. - - - - - 381cfaed by Ben Gamari at 2023-08-17T15:17:17-04:00 ghc-heap: Don't expose stack dirty and marking fields These are GC metadata and are not relevant to the end-user. Moreover, they are unstable which makes ghc-heap harder to test than necessary. - - - - - 16828ca5 by Luite Stegeman at 2023-08-21T18:42:53-04:00 bump process submodule to include macOS fix and JS support - - - - - b4d5f6ed by Matthew Pickering at 2023-08-21T18:43:29-04:00 ci: Add support for triggering test-primops pipelines This commit adds 4 ways to trigger testing with test-primops. 1. Applying the ~test-primops label to a validate pipeline. 2. A manually triggered job on a validate pipeline 3. A nightly pipeline job 4. A release pipeline job Fixes #23695 - - - - - 32c50daa by Matthew Pickering at 2023-08-21T18:43:29-04:00 Add test-primops label support The test-primops CI job requires some additional builds in the validation pipeline, so we make sure to enable these jobs when test-primops label is set. - - - - - 73ca8340 by Matthew Pickering at 2023-08-21T18:43:29-04:00 Revert "Aarch ncg: Optimize immediate use for address calculations" This reverts commit 8f3b3b78a8cce3bd463ed175ee933c2aabffc631. See #23793 - - - - - 5546ad9e by Matthew Pickering at 2023-08-21T18:43:29-04:00 Revert "AArch NCG: Pure refactor" This reverts commit 00fb6e6b06598752414a0b9a92840fb6ca61338d. See #23793 - - - - - 02dfcdc2 by Matthew Pickering at 2023-08-21T18:43:29-04:00 Revert "Aarch64 NCG: Use encoded immediates for literals." This reverts commit 40425c5021a9d8eb5e1c1046e2d5fa0a2918f96c. See #23793 ------------------------- Metric Increase: T4801 T5321FD T5321Fun ------------------------- - - - - - 7be4a272 by Matthew Pickering at 2023-08-22T08:55:20+01:00 ci: Remove manually triggered test-ci job This doesn't work on slimmed down pipelines as the needed jobs don't exist. If you want to run test-primops then apply the label. - - - - - 76a4d11b by Jaro Reinders at 2023-08-22T08:08:13-04:00 Remove Ptr example from roles docs - - - - - 069729d3 by Bryan Richter at 2023-08-22T08:08:49-04:00 Guard against duplicate pipelines in forks - - - - - f861423b by Rune K. Svendsen at 2023-08-22T08:09:35-04:00 dump-decls: fix "Ambiguous module name"-error Fixes errors of the following kind, which happen when dump-decls is run on a package that contains a module name that clashes with that of another package. ``` dump-decls: <no location info>: error: Ambiguous module name `System.Console.ANSI.Types': it was found in multiple packages: ansi-terminal-0.11.4 ansi-terminal-types-0.11.5 ``` - - - - - edd8bc43 by Krzysztof Gogolewski at 2023-08-22T12:31:20-04:00 Fix MultiWayIf linearity checking (#23814) Co-authored-by: Thomas BAGREL <thomas.bagrel at tweag.io> - - - - - 4ba088d1 by konsumlamm at 2023-08-22T12:32:02-04:00 Update `Control.Concurrent.*` documentation - - - - - 015886ec by ARATA Mizuki at 2023-08-22T15:13:13-04:00 Support 128-bit SIMD on AArch64 via LLVM backend - - - - - 52a6d868 by Krzysztof Gogolewski at 2023-08-22T15:13:51-04:00 Testsuite cleanup - Remove misleading help text in perf_notes, ways are not metrics - Remove no_print_summary - this was used for Phabricator - In linters tests, run 'git ls-files' just once. Previously, it was called on each has_ls_files() - Add ghc-prim.cabal to gitignore, noticed in #23726 - Remove ghc-prim.cabal, it was accidentally committed in 524c60c8cd - - - - - ab40aa52 by Alan Zimmerman at 2023-08-22T15:14:28-04:00 EPA: Use Introduce [DeclTag] in AnnSortKey The AnnSortKey is used to keep track of the order of declarations for printing when the container has split them apart. This applies to HsValBinds and ClassDecl, ClsInstDecl. When making modifications to the list of declarations, the new order must be captured for when it must be printed. For each list of declarations (binds and sigs for a HsValBind) we can just store the list in order. To recreate the list when printing, we must merge them, and this is what the AnnSortKey records. It used to be indexed by SrcSpan, we now simply index by a marker as to which list to take the next item from. - - - - - e7db36c1 by sheaf at 2023-08-23T08:41:28-04:00 Don't attempt pattern synonym error recovery This commit gets rid of the pattern synonym error recovery mechanism (recoverPSB). The rationale is that the fake pattern synonym binding that the recovery mechanism introduced could lead to undesirable knock-on errors, and it isn't really feasible to conjure up a satisfactory binding as pattern synonyms can be used both in expressions and patterns. See Note [Pattern synonym error recovery] in GHC.Tc.TyCl.PatSyn. It isn't such a big deal to eagerly fail compilation on a pattern synonym that doesn't typecheck anyway. Fixes #23467 - - - - - 6ccd9d65 by Ben Gamari at 2023-08-23T08:42:05-04:00 base: Don't use Data.ByteString.Internals.memcpy This function is now deprecated from `bytestring`. Use `Foreign.Marshal.Utils.copyBytes` instead. Fixes #23880. - - - - - 0bfa0031 by Matthew Pickering at 2023-08-23T13:43:48-04:00 hadrian: Uniformly pass buildOptions to all builders in runBuilder In Builder.hs, runBuilderWith mostly ignores the buildOptions in BuildInfo. This leads to hard to diagnose bugs as any build options you pass with runBuilderWithCmdOptions are ignored for many builders. Solution: Uniformly pass buildOptions to the invocation of cmd. Fixes #23845 - - - - - 9cac8f11 by Matthew Pickering at 2023-08-23T13:43:48-04:00 Abstract windows toolchain setup This commit splits up the windows toolchain setup logic into two functions. * FP_INSTALL_WINDOWS_TOOLCHAIN - deals with downloading the toolchain if it isn't already downloaded * FP_SETUP_WINDOWS_TOOLCHAIN - sets the environment variables to point to the correct place FP_SETUP_WINDOWS_TOOLCHAIN is abstracted from the location of the mingw toolchain and also the eventual location where we will install the toolchain in the installed bindist. This is the first step towards #23608 - - - - - 6c043187 by Matthew Pickering at 2023-08-23T13:43:48-04:00 Generate build.mk for bindists The config.mk.in script was relying on some variables which were supposed to be set by build.mk but therefore never were when used to install a bindist. Specifically * BUILD_PROF_LIBS to determine whether we had profiled libraries or not * DYNAMIC_GHC_PROGRAMS to determine whether we had shared libraries or not Not only were these never set but also not really accurate because you could have shared libaries but still statically linked ghc executable. In addition variables like GhcLibWays were just never used, so those have been deleted from the script. Now instead we generate a build.mk file which just directly specifies which RtsWays we have supplied in the bindist and whether we have DYNAMIC_GHC_PROGRAMS. - - - - - fe23629b by Matthew Pickering at 2023-08-23T13:43:48-04:00 hadrian: Add reloc-binary-dist-* targets This adds a command line option to build a "relocatable" bindist. The bindist is created by first creating a normal bindist and then installing it using the `RelocatableBuild=YES` option. This creates a bindist without any wrapper scripts pointing to the libdir. The motivation for this feature is that we want to ship relocatable bindists on windows and this method is more uniform than the ad-hoc method which lead to bugs such as #23608 and #23476 The relocatable bindist can be built with the "reloc-binary-dist" target and supports the same suffixes as the normal "binary-dist" command to specify the compression style. - - - - - 41cbaf44 by Matthew Pickering at 2023-08-23T13:43:48-04:00 packaging: Fix installation scripts on windows/RelocatableBuild case This includes quite a lot of small fixes which fix the installation makefile to work on windows properly. This also required fixing the RelocatableBuild variable which seemed to have been broken for a long while. Sam helped me a lot writing this patch by providing a windows machine to test the changes. Without him it would have taken ages to tweak everything. Co-authored-by: sheaf <sam.derbyshire at gmail.com> - - - - - 03474456 by Matthew Pickering at 2023-08-23T13:43:48-04:00 ci: Build relocatable bindist on windows We now build the relocatable bindist target on windows, which means we test and distribute the new method of creating a relocatable bindist. - - - - - d0b48113 by Matthew Pickering at 2023-08-23T13:43:48-04:00 hadrian: Add error when trying to build binary-dist target on windows The binary dist produced by `binary-dist` target doesn't work on windows because of the wrapper script the makefile installs. In order to not surprise any packagers we just give an error if someone tries to build the old binary-dist target rather than the reloc-binary-dist target. - - - - - 7cbf9361 by Matthew Pickering at 2023-08-23T13:43:48-04:00 hadrian: Remove query' logic to use tooldir - - - - - 03fad42e by Matthew Pickering at 2023-08-23T13:43:48-04:00 configure: Set WindresCmd directly and removed unused variables For some reason there was an indirection via the Windres variable before setting WindresCmd. That indirection led to #23855. I then also noticed that these other variables were just not used anywhere when trying to work out what the correct condition was for this bit of the configure script. - - - - - c82770f5 by sheaf at 2023-08-23T13:43:48-04:00 Apply shellcheck suggestion to SUBST_TOOLDIR - - - - - 896e35e5 by sheaf at 2023-08-23T13:44:34-04:00 Compute hints from TcSolverReportMsg This commit changes how hints are handled in conjunction with constraint solver report messages. Instead of storing `[GhcHint]` in the TcRnSolverReport error constructor, we compute the hints depending on the underlying TcSolverReportMsg. This disentangles the logic and makes it easier to add new hints for certain errors. - - - - - a05cdaf0 by Alexander Esgen at 2023-08-23T13:45:16-04:00 users-guide: remove note about fatal Haddock parse failures - - - - - 4908d798 by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Introduce Data.Enum - - - - - f59707c7 by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Num.Integer - - - - - b1054053 by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Num - - - - - 6baa481d by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Num.Natural - - - - - 2ac15233 by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Float - - - - - f3c489de by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add export list to GHC.Real - - - - - 94f59eaa by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Eliminate module reexport in GHC.Exception The metric increase here isn't strictly due to this commit but it's a rather small, incidental change. Metric Increase: T8095 T13386 Metric Decrease: T8095 T13386 T18304 - - - - - be1fc7df by Ben Gamari at 2023-08-23T17:36:41-04:00 base: Add disclaimers in internal modules To warn users that these modules are internal and their interfaces may change with little warning. As proposed in Core Libraries Committee #146 [CLC146]. [CLC146]: https://github.com/haskell/core-libraries-committee/issues/146 - - - - - 0326f3f4 by sheaf at 2023-08-23T17:37:29-04:00 Bump Cabal submodule We need to bump the Cabal submodule to include commit ec75950 which fixes an issue with a dodgy import Rep(..) which relied on GHC bug #23570 - - - - - 0504cd08 by Facundo Domínguez at 2023-08-23T17:38:11-04:00 Fix typos in the documentation of Data.OldList.permutations - - - - - 1420b8cb by Antoine Leblanc at 2023-08-24T16:18:17-04:00 Be more eager in TyCon boot validity checking This commit performs boot-file consistency checking for TyCons into checkValidTyCl. This ensures that we eagerly catch any mismatches, which prevents the compiler from seeing these inconsistencies and panicking as a result. See Note [TyCon boot consistency checking] in GHC.Tc.TyCl. Fixes #16127 - - - - - d99c816f by Finley McIlwaine at 2023-08-24T16:18:55-04:00 Refactor estimation of stack info table provenance This commit greatly refactors the way we compute estimated provenance for stack info tables. Previously, this process was done using an entirely separate traversal of the whole Cmm code stream to build the map from info tables to source locations. The separate traversal is now fused with the Cmm code generation pipeline in GHC.Driver.Main. This results in very significant code generation speed ups when -finfo-table-map is enabled. In testing, this patch reduces code generation times by almost 30% with -finfo-table-map and -O0, and 60% with -finfo-table-map and -O1 or -O2 . Fixes #23103 - - - - - d3e0124c by Finley McIlwaine at 2023-08-24T16:18:55-04:00 Add a test checking overhead of -finfo-table-map We want to make sure we don't end up with poor codegen performance resulting from -finfo-table-map again as in #23103. This test adds a performance test tracking total allocations while compiling ExactPrint with -finfo-table-map. - - - - - fcfc1777 by Ben Gamari at 2023-08-25T10:58:16-04:00 llvmGen: Add export list to GHC.Llvm.MetaData - - - - - 5880fff6 by Ben Gamari at 2023-08-25T10:58:16-04:00 llvmGen: Allow LlvmLits in MetaExprs This omission appears to be an oversight. - - - - - 86ce92a2 by Ben Gamari at 2023-08-25T10:58:16-04:00 compiler: Move platform feature predicates to GHC.Driver.DynFlags These are useful in `GHC.Driver.Config.*`. - - - - - a6a38742 by Ben Gamari at 2023-08-25T10:58:16-04:00 llvmGen: Introduce infrastructure for module flag metadata - - - - - e9af2cf3 by Ben Gamari at 2023-08-25T10:58:16-04:00 llvmGen: Don't pass stack alignment via command line As of https://reviews.llvm.org/D103048 LLVM no longer supports the `-stack-alignment=...` flag. Instead this information is passed via a module flag metadata node. This requires dropping support for LLVM 11 and 12. Fixes #23870 - - - - - a936f244 by Alan Zimmerman at 2023-08-25T10:58:56-04:00 EPA: Keep track of "in" token for WarningTxt category A warning can now be written with a category, e.g. {-# WARNInG in "x-c" e "d" #-} Keep track of the location of the 'in' keyword and string, as well as the original SourceText of the label, in case it uses character escapes. - - - - - 3df8a653 by Matthew Pickering at 2023-08-25T17:42:18-04:00 Remove redundant import in InfoTableProv The copyBytes function is provided by the import of Foreign. Fixes #23889 - - - - - d6f807ec by Ben Gamari at 2023-08-25T17:42:54-04:00 gitlab/issue-template: Mention report-a-bug - - - - - 50b9f75d by Artin Ghasivand at 2023-08-26T20:02:50+03:30 Added StandaloneKindSignature examples to replace CUSKs ones - - - - - 2f6309a4 by Vladislav Zavialov at 2023-08-27T03:47:37-04:00 Remove outdated CPP in compiler/* and template-haskell/* The boot compiler was bumped to 9.4 in cebb5819b43. There is no point supporting older GHC versions with CPP. - - - - - 5248fdf7 by Zubin Duggal at 2023-08-28T15:01:09+05:30 testsuite: Add regression test for #23861 Simon says this was fixed by commit 8d68685468d0b6e922332a3ee8c7541efbe46137 Author: sheaf <sam.derbyshire at gmail.com> Date: Fri Aug 4 15:28:45 2023 +0200 Remove zonk in tcVTA - - - - - b6903f4d by Zubin Duggal at 2023-08-28T12:33:58-04:00 testsuite: Add regression test for #23864 Simon says this was fixed by commit 59202c800f2c97c16906120ab2561f6e1556e4af Author: Sebastian Graf <sebastian.graf at kit.edu> Date: Fri Mar 31 17:35:22 2023 +0200 CorePrep: Eliminate EmptyCase and unsafeEqualityProof in CoreToStg instead We eliminate EmptyCase by way of `coreToStg (Case e _ _ []) = coreToStg e` now. The main reason is that it plays far better in conjunction with eta expansion (as we aim to do for arguments in CorePrep, #23083), because we can discard any arguments, `(case e of {}) eta == case e of {}`, whereas in `(e |> co) eta` it's impossible to discard the argument. - - - - - 9eecdf33 by sheaf at 2023-08-28T18:54:06+00:00 Remove ScopedTypeVariables => TypeAbstractions This commit implements [amendment 604](https://github.com/ghc-proposals/ghc-proposals/pull/604/) to [GHC proposal 448](https://github.com/ghc-proposals/ghc-proposals/pull/448) by removing the implication of language extensions ScopedTypeVariables => TypeAbstractions To limit breakage, we now allow type arguments in constructor patterns when both ScopedTypeVariables and TypeApplications are enabled, but we emit a warning notifying the user that this is deprecated behaviour that will go away starting in GHC 9.12. Fixes #23776 - - - - - fadd5b4d by sheaf at 2023-08-28T18:54:06+00:00 .stderr: ScopedTypeVariables =/> TypeAbstractions This commit accepts testsuite changes for the changes in the previous commit, which mean that TypeAbstractions is no longer implied by ScopedTypeVariables. - - - - - 4f5fb500 by Greg Steuck at 2023-08-29T07:55:13-04:00 Repair `codes` test on OpenBSD by explicitly requesting extended RE - - - - - 6bbde581 by Vasily Sterekhov at 2023-08-29T12:06:58-04:00 Add test for #23540 `T23540.hs` makes use of `explainEv` from `HieQueries.hs`, so `explainEv` has been moved to `TestUtils.hs`. - - - - - 257bb3bd by Vasily Sterekhov at 2023-08-29T12:06:58-04:00 Add test for #23120 - - - - - 4f192947 by Vasily Sterekhov at 2023-08-29T12:06:58-04:00 Make some evidence uses reachable by toHie Resolves #23540, #23120 This adds spans to certain expressions in the typechecker and renamer, and lets 'toHie' make use of those spans. Therefore the relevant evidence uses for the following syntax will now show up under the expected nodes in 'HieAst's: - Overloaded literals ('IsString', 'Num', 'Fractional') - Natural patterns and N+k patterns ('Eq', 'Ord', and instances from the overloaded literals being matched on) - Arithmetic sequences ('Enum') - Monadic bind statements ('Monad') - Monadic body statements ('Monad', 'Alternative') - ApplicativeDo ('Applicative', 'Functor') - Overloaded lists ('IsList') Also see Note [Source locations for implicit function calls] In the process of handling overloaded lists I added an extra 'SrcSpan' field to 'VAExpansion' - this allows us to more accurately reconstruct the locations from the renamer in 'rebuildHsApps'. This also happens to fix #23120. See the additions to Note [Looking through HsExpanded] - - - - - fe9fcf9d by Sylvain Henry at 2023-08-29T12:07:50-04:00 ghc-heap: rename C file (fix #23898) - - - - - b60d6576 by Krzysztof Gogolewski at 2023-08-29T12:08:29-04:00 Misc cleanup - Builtin.PrimOps: ReturnsAlg was used only for unboxed tuples. Rename to ReturnsTuple. - Builtin.Utils: use SDoc for a panic message. The comment about <<details unavailable>> was obsoleted by e8d356773b56. - TagCheck: fix wrong logic. It was zipping a list 'args' with its version 'args_cmm' after filtering. - Core.Type: remove an outdated 1999 comment about unlifted polymorphic types - hadrian: remove leftover debugging print - - - - - 3054fd6d by Krzysztof Gogolewski at 2023-08-29T12:09:08-04:00 Add a regression test for #23903 The bug has been fixed by commit bad2f8b8aa8424. - - - - - 21584b12 by Ben Gamari at 2023-08-29T19:52:02-04:00 README: Refer to ghc-hq repository for contributor and governance information - - - - - e542d590 by sheaf at 2023-08-29T19:52:40-04:00 Export setInertSet from GHC.Tc.Solver.Monad We used to export getTcSInerts and setTcSInerts from GHC.Tc.Solver.Monad. These got renamed to getInertSet/setInertSet in e1590ddc. That commit also removed the export of setInertSet, but that function is useful for the GHC API. - - - - - 694ec5b1 by sheaf at 2023-08-30T10:18:32-04:00 Don't bundle children for non-parent Avails We used to bundle all children of the parent Avail with things that aren't the parent, e.g. with class C a where type T a meth :: .. we would bundle the whole Avail (C, T, meth) with all of C, T and meth, instead of only with C. Avoiding this fixes #23570 - - - - - d926380d by Krzysztof Gogolewski at 2023-08-30T10:19:08-04:00 Fix typos - - - - - d07080d2 by Josh Meredith at 2023-08-30T19:42:32-04:00 JS: Implement missing C functions `rename`, `realpath`, and `getcwd` (#23806) - - - - - e2940272 by David Binder at 2023-08-30T19:43:08-04:00 Bump submodules of hpc and hpc-bin to version 0.7.0.0 hpc 0.7.0.0 dropped SafeHaskell safety guarantees in order to simplify compatibility with newer versions of the directory package which dropped all SafeHaskell guarantees. - - - - - 5d56d05c by David Binder at 2023-08-30T19:43:08-04:00 Bump hpc bound in ghc.cabal.in - - - - - 99fff496 by Dominik Schrempf at 2023-08-31T00:04:46-04:00 ghc classes documentation: rm redundant comment - - - - - fe021bab by Dominik Schrempf at 2023-08-31T00:04:46-04:00 prelude documentation: various nits - - - - - 48c84547 by Dominik Schrempf at 2023-08-31T00:04:46-04:00 integer documentation: minor corrections - - - - - 20cd12f4 by Dominik Schrempf at 2023-08-31T00:04:46-04:00 real documentation: nits - - - - - dd39bdc0 by sheaf at 2023-08-31T00:05:27-04:00 Add a test for #21765 This issue (of reporting a constraint as being redundant even though removing it causes typechecking to fail) was fixed in aed1974e. This commit simply adds a regression test. Fixes #21765 - - - - - f1ec3628 by Andrew Lelechenko at 2023-08-31T23:53:30-04:00 Export foldl' from Prelude and bump submodules See https://github.com/haskell/core-libraries-committee/issues/167 for discussion Metric Decrease: T8095 T13386 Metric Increase: T13386 T8095 T8095 ghc/alloc decreased on x86_64, but increased on aarch64. T13386 ghc/alloc decreased on x86_64-windows, but increased on other platforms. Neither has anything to do with `foldl'`, so I conclude that both are flaky. - - - - - 3181b97d by Gergő Érdi at 2023-08-31T23:54:06-04:00 Allow cross-tyvar defaulting proposals from plugins Fixes #23832. - - - - - 495b2241 by Matthew Pickering at 2023-09-01T13:02:07+02:00 Fix issue with duplicate reexported definitions (T23616) When a class method was reexported, it's default methods were also showing up in the generated html page. The simplest and most non-invasive fix is to not look for the default method if we are just exporting the class method.. because the backends are just showing default methods when the whole class is exported. In general it would be worthwhile to rewrite this bit of code I think as the logic and what gets included is split over `lookupDocs` and `availExportDecl` it would be clearer to combine the two. The result of lookupDocs is always just passed to availExportDecl so it seems simpler and more obvious to just write the function directly. - - - - - e4af506e by Sebastian Graf at 2023-09-01T14:29:12-04:00 Clarify Note [GlobalId/LocalId] after CorePrep (#23797) Fixes #23797. - - - - - ac29787c by Sylvain Henry at 2023-09-01T14:30:02-04:00 Fix warning with UNPACK on sum type (#23921) - - - - - 9765ac7b by Zubin Duggal at 2023-09-05T00:37:45-04:00 hadrian: track python dependencies in doc rules - - - - - 1578215f by sheaf at 2023-09-05T00:38:26-04:00 Bump Haddock to fix #23616 This commit updates the Haddock submodule to include the fix to #23616. Fixes #23616 - - - - - 5a2fe35a by David Binder at 2023-09-05T00:39:07-04:00 Fix example in GHC user guide in SafeHaskell section The example given in the SafeHaskell section uses an implementation of Monad which no longer works. This MR removes the non-canonical return instance and adds the necessary instances of Functor and Applicative. - - - - - 291d81ae by Matthew Pickering at 2023-09-05T14:03:10-04:00 driver: Check transitive closure of haskell package dependencies when deciding whether to relink We were previously just checking whether direct package dependencies had been modified. This caused issues when compiling without optimisations as we wouldn't relink the direct dependency if one of its dependenices changed. Fixes #23724 - - - - - 35da0775 by Krzysztof Gogolewski at 2023-09-05T14:03:47-04:00 Re-export GHC.Utils.Panic.Plain from GHC.Utils.Panic Fixes #23930 - - - - - 6551824d by Finley McIlwaine at 2023-09-05T13:06:57-07:00 Remove fake export of `FUN` from Prelude This prevents `data FUN` from being shown at the top of the Prelude docs. Fixes \#23920 on GHC. - - - - - 3930d793 by Jaro Reinders at 2023-09-06T18:42:55-04:00 Make STG rewriter produce updatable closures - - - - - 0104221a by Krzysztof Gogolewski at 2023-09-06T18:43:32-04:00 configure: update message to use hadrian (#22616) - - - - - b34f8586 by Alan Zimmerman at 2023-09-07T10:58:38-04:00 EPA: Incorrect locations for UserTyVar with '@' In T13343.hs, the location for the @ is not within the span of the surrounding UserTyVar. type Bad @v = (forall (v1 :: RuntimeRep) (a1 :: TYPE v). a1) :: TYPE v Widen it so it is captured. Closes #23887 - - - - - 8046f020 by Finley McIlwaine at 2023-09-07T10:59:15-04:00 Bump haddock submodule to fix #23920 Removes the fake export of `FUN` from Prelude. Fixes #23920. Bumps haddock submodule. - - - - - e0aa8c6e by Krzysztof Gogolewski at 2023-09-07T11:00:03-04:00 Fix wrong role in mkSelCo_maybe In the Lint failure in #23938, we start with a coercion Refl :: T a ~R T a, and call mkSelCo (SelTyCon 1 nominal) Refl. The function incorrectly returned Refl :: a ~R a. The returned role should be nominal, according to the SelCo rule: co : (T s1..sn) ~r0 (T t1..tn) r = tyConRole tc r0 i ---------------------------------- SelCo (SelTyCon i r) : si ~r ti In this test case, r is nominal while r0 is representational. - - - - - 1d92f2df by Gergő Érdi at 2023-09-08T04:04:30-04:00 If we have multiple defaulting plugins, then we should zonk in between them after any defaulting has taken place, to avoid a defaulting plugin seeing a metavariable that has already been filled. Fixes #23821. - - - - - eaee4d29 by Gergő Érdi at 2023-09-08T04:04:30-04:00 Improvements to the documentation of defaulting plugins Based on @simonpj's draft and comments in !11117 - - - - - ede3df27 by Alan Zimmerman at 2023-09-08T04:05:06-04:00 EPA: Incorrect span for LWarnDec GhcPs The code (from T23465.hs) {-# WARNInG in "x-c" e "d" #-} e = e gives an incorrect span for the LWarnDecl GhcPs Closes #23892 It also fixes the Test23465/Test23464 mixup - - - - - a0ccef7a by Krzysztof Gogolewski at 2023-09-08T04:05:42-04:00 Valid hole fits: don't suggest unsafeCoerce (#17940) - - - - - 9ab5a448 by Alan Zimmerman at 2023-09-08T18:26:53+01:00 Match changes in wip/az/T23885-unicode-funtycon - - - - - 88b942c4 by Oleg Grenrus at 2023-09-08T19:58:42-04:00 Add warning for badly staged types. Resolves #23829. The stage violation results in out-of-bound names in splices. Technically this is an error, but someone might rely on this!? Internal changes: - we now track stages for TyVars. - thLevel (RunSplice _) = 0, instead of panic, as reifyInstances does in fact rename its argument type, and it can contain variables. - - - - - 9861f787 by Ben Gamari at 2023-09-08T19:59:19-04:00 rts: Fix invalid symbol type I suspect this code is dead since we haven't observed this failing despite the obviously incorrect macro name. - - - - - 03ed6a9a by Ben Gamari at 2023-09-08T19:59:19-04:00 testsuite: Add simple test exercising C11 atomics in GHCi See #22012. - - - - - 1aa5733a by Ben Gamari at 2023-09-08T19:59:19-04:00 rts/RtsSymbols: Add AArch64 outline atomic operations Fixes #22012 by adding the symbols described in https://github.com/llvm/llvm-project/blob/main/llvm/docs/Atomics.rst#libcalls-atomic. Ultimately this would be better addressed by #22011, but this is a first step in the right direction and fixes the immediate symptom. Note that we dropped the `__arch64_cas16` operations as these provided by all platforms's compilers. Also, we don't link directly against the libgcc/compiler-rt definitions but rather provide our own wrappers to work around broken toolchains (e.g. https://bugs.gentoo.org/868018). Generated via https://gitlab.haskell.org/ghc/ghc/-/snippets/5733. - - - - - 8f7d3041 by Matthew Pickering at 2023-09-08T19:59:55-04:00 ci: Build debian12 and fedora38 bindists This adds builds for the latest releases for fedora and debian We build these bindists in nightly and release pipelines. - - - - - a1f0d55c by Felix Leitz at 2023-09-08T20:00:37-04:00 Fix documentation around extension implication for MultiParamTypeClasses/ConstrainedClassMethods. - - - - - 98166389 by Teo Camarasu at 2023-09-12T04:30:54-04:00 docs: move -xn flag beside --nonmoving-gc It makes sense to have these beside each other as they are aliases. - - - - - f367835c by Teo Camarasu at 2023-09-12T04:30:55-04:00 nonmoving: introduce a family of dense allocators Supplement the existing power 2 sized nonmoving allocators with a family of dense allocators up to a configurable threshold. This should reduce waste from rounding up block sizes while keeping the amount of allocator sizes manageable. This patch: - Adds a new configuration option `--nonmoving-dense-allocator-count` to control the amount of these new dense allocators. - Adds some constants to `NonmovingAllocator` in order to keep marking fast with the new allocators. Resolves #23340 - - - - - 2b07bf2e by Teo Camarasu at 2023-09-12T04:30:55-04:00 Add changelog entry for #23340 - - - - - f96fe681 by sheaf at 2023-09-12T04:31:44-04:00 Use printGhciException in run{Stmt, Decls} When evaluating statements in GHCi, we need to use printGhciException instead of the printException function that GHC provides in order to get the appropriate error messages that are customised for ghci use. - - - - - d09b932b by psilospore at 2023-09-12T04:31:44-04:00 T23686: Suggest how to enable Language Extension when in ghci Fixes #23686 - - - - - da30f0be by Matthew Craven at 2023-09-12T04:32:24-04:00 Unarise: Split Rubbish literals in function args Fixes #23914. Also adds a check to STG lint that these args are properly unary or nullary after unarisation - - - - - 261b6747 by Matthew Pickering at 2023-09-12T04:33:04-04:00 darwin: Bump MAXOSX_DEPLOYMENT_TARGET to 10.13 This bumps the minumum supported version to 10.13 (High Sierra) which is 6 years old at this point. Fixes #22938 - - - - - f418f919 by Mario Blažević at 2023-09-12T04:33:45-04:00 Fix TH pretty-printing of nested GADTs, issue #23937 This commit fixes `Language.Haskell.TH.Ppr.pprint` so that it correctly pretty-prints GADTs declarations contained within data family instances. Fixes #23937 - - - - - d7a64753 by John Ericson at 2023-09-12T04:34:20-04:00 Put hadrian non-bootstrap plans through `jq` This makes it possible to review changes with conventional diffing tools. This is picking up where ad8cfed4195b1bbfc15b841f010e75e71f63157d left off. - - - - - ff0a709a by Sylvain Henry at 2023-09-12T08:46:28-04:00 JS: fix some tests - Tests using Setup programs need to pass --with-hc-pkg - Several other fixes See https://gitlab.haskell.org/ghc/ghc/-/wikis/javascript-backend/bug_triage for the current status - - - - - fc86f0e7 by Krzysztof Gogolewski at 2023-09-12T08:47:04-04:00 Fix in-scope set assertion failure (#23918) Patch by Simon - - - - - 21a906c2 by Matthew Pickering at 2023-09-12T17:21:04+02:00 Add -Winconsistent-flags warning The warning fires when inconsistent command line flags are passed. For example: * -dynamic-too and -dynamic * -dynamic-too on windows * -O and --interactive * etc This is on by default and allows users to control whether the warning is displayed and whether it should be an error or not. Fixes #22572 - - - - - dfc4f426 by Krzysztof Gogolewski at 2023-09-12T20:31:35-04:00 Avoid serializing BCOs with the internal interpreter Refs #23919 - - - - - 9217950b by Finley McIlwaine at 2023-09-13T08:06:03-04:00 Fix numa auto configure - - - - - 98e7c1cf by Simon Peyton Jones at 2023-09-13T08:06:40-04:00 Add -fno-cse to T15426 and T18964 This -fno-cse change is to avoid these performance tests depending on flukey CSE stuff. Each contains several independent tests, and we don't want them to interact. See #23925. By killing CSE we expect a 400% increase in T15426, and 100% in T18964. Metric Increase: T15426 T18964 - - - - - 236a134e by Simon Peyton Jones at 2023-09-13T08:06:40-04:00 Tiny refactor canEtaReduceToArity was only called internally, and always with two arguments equal to zero. This patch just specialises the function, and renames it to cantEtaReduceFun. No change in behaviour. - - - - - 56b403c9 by Ben Gamari at 2023-09-13T19:21:36-04:00 spec-constr: Lift argument limit for SPEC-marked functions When the user adds a SPEC argument to a function, they are informing us that they expect the function to be specialised. However, previously this instruction could be preempted by the specialised-argument limit (sc_max_args). Fix this. This fixes #14003. - - - - - 6840012e by Simon Peyton Jones at 2023-09-13T19:22:13-04:00 Fix eta reduction Issue #23922 showed that GHC was bogusly eta-reducing a join point. We should never eta-reduce (\x -> j x) to j, if j is a join point. It is extremly difficult to trigger this bug. It took me 45 mins of trying to make a small tests case, here immortalised as T23922a. - - - - - e5c00092 by Andreas Klebinger at 2023-09-14T08:57:43-04:00 Profiling: Properly escape characters when using `-pj`. There are some ways in which unusual characters like quotes or others can make it into cost centre names. So properly escape these. Fixes #23924 - - - - - ec490578 by Ellie Hermaszewska at 2023-09-14T08:58:24-04:00 Use clearer example variable names for bool eliminator - - - - - 5126a2fe by Sylvain Henry at 2023-09-15T11:18:02-04:00 Add missing int64/word64-to-double/float rules (#23907) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/203 - - - - - 566ef411 by Mario Blažević at 2023-09-15T11:18:43-04:00 Fix and test TH pretty-printing of type operator role declarations This commit fixes and tests `Language.Haskell.TH.Ppr.pprint` so that it correctly pretty-prints `type role` declarations for operator names. Fixes #23954 - - - - - 8e05c54a by Simon Peyton Jones at 2023-09-16T01:42:33-04:00 Use correct FunTyFlag in adjustJoinPointType As the Lint error in #23952 showed, the function adjustJoinPointType was failing to adjust the FunTyFlag when adjusting the type. I don't think this caused the seg-fault reported in the ticket, but it is definitely. This patch fixes it. It is tricky to come up a small test case; Krzysztof came up with this one, but it only triggers a failure in GHC 9.6. - - - - - 778c84b6 by Pierre Le Marre at 2023-09-16T01:43:15-04:00 Update to Unicode 15.1.0 See: https://www.unicode.org/versions/Unicode15.1.0/ - - - - - f9d79a6c by Alan Zimmerman at 2023-09-18T00:00:14-04:00 EPA: track unicode version for unrestrictedFunTyCon Closes #23885 Updates haddock submodule - - - - - 9374f116 by Andrew Lelechenko at 2023-09-18T00:00:54-04:00 Bump parsec submodule to allow text-2.1 and bytestring-0.12 - - - - - 7ca0240e by Ben Gamari at 2023-09-18T15:16:48-04:00 base: Advertise linear time of readFloat As noted in #23538, `readFloat` has runtime that scales nonlinearly in the size of its input. Consequently, its use on untrusted input can be exploited as a denial-of-service vector. Point this out and suggest use of `read` instead. See #23538. - - - - - f3f58f13 by Simon Peyton Jones at 2023-09-18T15:17:24-04:00 Remove dead code GHC.CoreToStg.Prep.canFloat This function never fires, so we can delete it: #23965. - - - - - ccab5b15 by Ben Gamari at 2023-09-18T15:18:02-04:00 base/changelog: Move fix for #23907 to 9.8.1 section Since the fix was backported to 9.8.1 - - - - - 51b57d65 by Matthew Pickering at 2023-09-19T08:44:31-04:00 Add aarch64 alpine bindist This is dynamically linked and makes creating statically linked executables more straightforward. Fixes #23482 - - - - - 02c87213 by Matthew Pickering at 2023-09-19T08:44:31-04:00 Add aarch64-deb11 bindist This adds a debian 11 release job for aarch64. Fixes #22005 - - - - - 8b61dfd6 by Alexis King at 2023-09-19T08:45:13-04:00 Don’t store the async exception masking state in CATCH frames - - - - - 86d2971e by doyougnu at 2023-09-19T19:08:19-04:00 compiler,ghci: error codes link to HF error index closes: #23259 - adds -fprint-error-index-links={auto|always|never} flag - - - - - 5f826c18 by sheaf at 2023-09-19T19:09:03-04:00 Pass quantified tyvars in tcDefaultAssocDecl This commit passes the correct set of quantified type variables written by the user in associated type default declarations for validity checking. This ensures that validity checking of associated type defaults mirrors that of standalone type family instances. Fixes #23768 (see testcase T23734 in subsequent commit) - - - - - aba18424 by sheaf at 2023-09-19T19:09:03-04:00 Avoid panic in mkGADTVars This commit avoids panicking in mkGADTVars when we encounter a type variable as in #23784 that is bound by a user-written forall but not actually used. Fixes #23784 - - - - - a525a92a by sheaf at 2023-09-19T19:09:03-04:00 Adjust reporting of unused tyvars in data FamInsts This commit adjusts the validity checking of data family instances to improve the reporting of unused type variables. See Note [Out of scope tvs in data family instances] in GHC.Tc.Validity. The problem was that, in a situation such as data family D :: Type data instance forall (d :: Type). D = MkD the RHS passed to 'checkFamPatBinders' would be the TyCon app R:D d which mentions the type variable 'd' quantified in the user-written forall. Thus, when computing the set of unused type variables in the RHS of the data family instance, we would find that 'd' is used, and report a strange error message that would say that 'd' is not bound on the LHS. To fix this, we special-case the data-family instance case, manually extracting all the type variables that appear in the arguments of all the data constructores of the data family instance. Fixes #23778 - - - - - 28dd52ee by sheaf at 2023-09-19T19:09:03-04:00 Unused tyvars in FamInst: only report user tyvars This commit changes how we perform some validity checking for coercion axioms to mirror how we handle default declarations for associated type families. This allows us to keep track of whether type variables in type and data family instances were user-written or not, in order to only report the user-written ones in "unused type variable" error messages. Consider for example: {-# LANGUAGE PolyKinds #-} type family F type instance forall a. F = () In this case, we get two quantified type variables, (k :: Type) and (a :: k); the second being user-written, but the first is introduced by the typechecker. We should only report 'a' as being unused, as the user has no idea what 'k' is. Fixes #23734 - - - - - 1eed645c by sheaf at 2023-09-19T19:09:03-04:00 Validity: refactor treatment of data families This commit refactors the reporting of unused type variables in type and data family instances to be more principled. This avoids ad-hoc logic in the treatment of data family instances. - - - - - 35bc506b by John Ericson at 2023-09-19T19:09:40-04:00 Remove `ghc-cabal` It is dead code since the Make build system was removed. I tried to go over every match of `git grep -i ghc-cabal` to find other stray bits. Some of those might be workarounds that can be further removed. - - - - - 665ca116 by John Paul Adrian Glaubitz at 2023-09-19T19:10:39-04:00 Re-add unregisterised build support for sparc and sparc64 Closes #23959 - - - - - 142f8740 by Matthew Pickering at 2023-09-19T19:11:16-04:00 Bump ci-images to use updated version of Alex Fixes #23977 - - - - - fa977034 by John Ericson at 2023-09-21T12:55:25-04:00 Use Cabal 3.10 for Hadrian We need the newer version for `CABAL_FLAG_*` env vars for #17191. - - - - - a5d22cab by John Ericson at 2023-09-21T12:55:25-04:00 hadrian: `need` any `configure` script we will call When the script is changed, we should reconfigure. - - - - - db882b57 by John Ericson at 2023-09-21T12:55:25-04:00 hadrian: Make it easier to debug Cabal configure Right now, output is squashed. This make per-package configure scripts extremely hard to maintain, because we get vague "library is missing" errors when the actually probably is usually completely unrelated except for also involving the C/C++ toolchain. (I can always pass `-VVV` to Hadrian locally, but these errors are subtle and I often cannot reproduce them locally!) `--disable-option-checking` was added back in 75c6e0684dda585c37b4ac254cd7a13537a59a91 but seems to be a bit overkill; if other flags are passed that are not recognized behind the two from Cabal mentioned in the former comment, we *do* want to know about it. - - - - - 7ed65f5a by John Ericson at 2023-09-21T12:55:25-04:00 hadrian: Increase verbosity of certain cabal commands This is a hack to get around the cabal function we're calling *decreasing* the verbosity it passes to another function, which is the stuff we often actually care about. Sigh. Keeping this a separate commit so if this makes things too verbose it is easy to revert. - - - - - a4fde569 by John Ericson at 2023-09-21T12:55:25-04:00 rts: Move most external symbols logic to the configure script This is much more terse because we are programmatically handling the leading underscore. `findPtr` however is still handled in the Cabal file because we need a newer Cabal to pass flags to the configure script automatically. Co-Authored-By: Ben Gamari <ben at well-typed.com> - - - - - 56cc85fb by Andrew Lelechenko at 2023-09-21T12:56:21-04:00 Bump Cabal submodule to allow text-2.1 and bytestring-0.12 - - - - - 0cd6148c by Matthew Pickering at 2023-09-21T12:56:21-04:00 hadrian: Generate Distribution/Fields/Lexer.x before creating a source-dist - - - - - b10ba6a3 by Andrew Lelechenko at 2023-09-21T12:56:21-04:00 Bump hadrian's index-state to upgrade alex at least to 3.2.7.3 - - - - - 11ecc37b by Luite Stegeman at 2023-09-21T12:57:03-04:00 JS: correct file size and times Programs produced by the JavaScript backend were returning incorrect file sizes and modification times, causing cabal related tests to fail. This fixes the problem and adds an additional test that verifies basic file information operations. fixes #23980 - - - - - b35fd2cd by Ben Gamari at 2023-09-21T12:57:39-04:00 gitlab-ci: Drop libiserv from upload_ghc_libs libiserv has been merged into the ghci package. - - - - - 37ad04e8 by Ben Gamari at 2023-09-21T12:58:15-04:00 testsuite: Fix Windows line endings - - - - - 5795b365 by Ben Gamari at 2023-09-21T12:58:15-04:00 testsuite: Use makefile_test - - - - - 15118740 by Ben Gamari at 2023-09-21T12:58:55-04:00 system-cxx-std-lib: Add license and description - - - - - 0208f1d5 by Ben Gamari at 2023-09-21T12:59:33-04:00 gitlab/issue-templates: Rename bug.md -> default.md So that it is visible by default. - - - - - 23cc3f21 by Andrew Lelechenko at 2023-09-21T20:18:11+01:00 Bump submodule text to 2.1 - - - - - b8e4fe23 by Andrew Lelechenko at 2023-09-22T20:05:05-04:00 Bump submodule unix to 2.8.2.1 - - - - - 54b2016e by John Ericson at 2023-09-23T11:40:41-04:00 Move lib{numa,dw} defines to RTS configure Clean up the m4 to handle the auto case always and be more consistent. Also simplify the CPP --- we should always have both headers if we are using libnuma. "side effects" (AC_DEFINE, and AC_SUBST) are removed from the macros to better separate searching from actions taken based on search results. This might seem overkill now, but will make shuffling logic between configure scripts easier later. The macro comments are converted from `dnl` to `#` following the recomendation in https://www.gnu.org/software/autoconf/manual/autoconf-2.71/html_node/Macro-Definitions.html - - - - - d51b601b by John Ericson at 2023-09-23T11:40:50-04:00 Shuffle libzstd configuring between scripts Like the prior commit for libdw and libnuma, `AC_DEFINE` to RTS configure, `AC_SUBST` goes to the top-level configure script, and the documentation of the m4 macro is improved. - - - - - d1425af0 by John Ericson at 2023-09-23T11:41:03-04:00 Move `FP_ARM_OUTLINE_ATOMICS` to RTS configure It is just `AC_DEFINE` it belongs there instead. - - - - - 18de37e4 by John Ericson at 2023-09-23T11:41:03-04:00 Move mmap in the runtime linker check to the RTS configure `AC_DEFINE` should go there instead. - - - - - 74132c2b by Andrew Lelechenko at 2023-09-25T21:56:54-04:00 Elaborate comment on GHC_NO_UNICODE - - - - - de142aa2 by Ben Gamari at 2023-09-26T15:25:03-04:00 gitlab-ci: Mark T22012 as broken on CentOS 7 Due to #23979. - - - - - 6a896ce8 by Teo Camarasu at 2023-09-26T15:25:39-04:00 hadrian: better error for failing to find file's dependencies Resolves #24004 - - - - - d697a6c2 by Stefan Holdermans at 2023-09-26T20:58:37+00:00 Refactor uses of `partitionEithers . map` This patch changes occurences of the idiom `partitionEithers (map f xs)` by the simpler form `partitionWith f xs` where `partitionWith` is the utility function defined in `GHC.Utils.Misc`. Resolves: #23953 - - - - - 8a2968b7 by Stefan Holdermans at 2023-09-26T20:58:37+00:00 Refactor uses of `partitionEithers <$> mapM f xs` This patch changes occurences of the idiom `partitionEithers <$> mapM f xs` by the simpler form `partitionWithM f xs` where `partitionWithM` is a utility function newly added to `GHC.Utils.Misc`. - - - - - 6a27eb97 by Stefan Holdermans at 2023-09-26T20:58:37+00:00 Mark `GHC.Utils.Misc.partitionWithM` as inlineable This patch adds an `INLINEABLE` pragma for `partitionWithM` to ensure that the right-hand side of the definition of this function remains available for specialisation at call sites. - - - - - f1e5245a by David Binder at 2023-09-27T01:19:00-04:00 Add RTS option to supress tix file - - - - - 1f43124f by David Binder at 2023-09-27T01:19:00-04:00 Add expected output to testsuite in test interface-stability/base-exports - - - - - b9d2c354 by David Binder at 2023-09-27T01:19:00-04:00 Expose HpcFlags and getHpcFlags from GHC.RTS.Flags - - - - - 345675c6 by David Binder at 2023-09-27T01:19:00-04:00 Fix expected output of interface-stability test - - - - - 146e1c39 by David Binder at 2023-09-27T01:19:00-04:00 Implement getHpcFlags - - - - - 61ba8e20 by David Binder at 2023-09-27T01:19:00-04:00 Add section in user guide - - - - - ea05f890 by David Binder at 2023-09-27T01:19:01-04:00 Rename --emit-tix-file to --write-tix-file - - - - - cabce2ce by David Binder at 2023-09-27T01:19:01-04:00 Update the golden files for interface stability - - - - - 1dbdb9d0 by Krzysztof Gogolewski at 2023-09-27T01:19:37-04:00 Refactor: introduce stgArgRep The function 'stgArgType' returns the type in STG. But this violates the abstraction: in STG we're supposed to operate on PrimReps. This introduces stgArgRep ty = typePrimRep (stgArgType ty) stgArgRep1 ty = typePrimRep1 (stgArgType ty) stgArgRep_maybe ty = typePrimRep_maybe (stgArgType ty) stgArgType is still directly used for unboxed tuples (should be fixable), FFI and in ticky. - - - - - b02f8042 by Mario Blažević at 2023-09-27T17:33:28-04:00 Fix TH pretty-printer's parenthesization This PR Fixes `Language.Haskell.TH.Ppr.pprint` so it correctly emits parentheses where needed. Fixes #23962, #23968, #23971, and #23986 - - - - - 79104334 by Krzysztof Gogolewski at 2023-09-27T17:34:04-04:00 Add a testcase for #17564 The code in the ticket relied on the behaviour of Derived constraints. Derived constraints were removed in GHC 9.4 and now the code works as expected. - - - - - d7a80143 by sheaf at 2023-09-28T03:25:53-04:00 lint-codes: add new modes of operation This commit adds two new modes of operation to the lint-codes utility: list - list all statically used diagnostic codes outdated - list all outdated diagnostic codes The previous behaviour is now: test - test consistency and coverage of diagnostic codes - - - - - 477d223c by sheaf at 2023-09-28T03:25:53-04:00 lint codes: avoid using git-grep We manually traverse through the filesystem to find the diagnostic codes embedded in .stdout and .stderr files, to avoid any issues with old versions of grep. Fixes #23843 - - - - - a38ae69a by sheaf at 2023-09-28T03:25:53-04:00 lint-codes: add Hadrian targets This commit adds new Hadrian targets: codes, codes:used - list all used diagnostic codes codes:outdated - list outdated diagnostic codes This allows users to easily query GHC for used and outdated diagnostic codes, e.g. hadrian/build -j --flavour=<..> codes will list all used diagnostic codes in the command line by running the lint-codes utility in the "list codes" mode of operation. The diagnostic code consistency and coverage test is still run as usual, through the testsuite: hadrian/build test --only="codes" - - - - - 9cdd629b by Ben Gamari at 2023-09-28T03:26:29-04:00 hadrian: Install LICENSE files in bindists Fixes #23548. - - - - - b8ebf876 by Matthew Craven at 2023-09-28T03:27:05-04:00 Fix visibility when eta-reducing a type lambda Fixes #24014. - - - - - d3874407 by Torsten Schmits at 2023-09-30T16:08:10-04:00 Fix several mistakes around free variables in iface breakpoints Fixes #23612 , #23607, #23998 and #23666. MR: !11026 The fingerprinting logic in `Iface.Recomp` failed lookups when processing decls containing breakpoints for two reasons: * IfaceBreakpoint created binders for free variables instead of expressions * When collecting free names for the dependency analysis for fingerprinting, breakpoint FVs were skipped - - - - - ef5342cd by Simon Peyton Jones at 2023-09-30T16:08:48-04:00 Refactor to combine HsLam and HsLamCase This MR is pure refactoring (#23916): * Combine `HsLam` and `HsLamCase` * Combine `HsCmdLam` and `HsCmdLamCase` This just arranges to treat uniformly \x -> e \case pi -> ei \cases pis -> ie In the exising code base the first is treated differently to the latter two. No change in behaviour. More specifics: * Combine `HsLam` and `HsLamCase` (constructors of `Language.Haskell.Syntax.Expr.HsExpr`) into one data construtor covering * Lambda * `\case` * `\cases` * The new `HsLam` has an argument of type `HsLamVariant` to distinguish the three cases. * Similarly, combine `HsCmdLam` and `HsCmdLamCase` (constructors of `Language.Haskell.Syntax.Expr.HsCmd` ) into one. * Similarly, combine `mkHsLamPV` and `mkHsLamCasePV` (methods of class `DisambECP`) into one. (Thank you Alan Zimmerman.) * Similarly, combine `LambdaExpr` and `LamCaseAlt` (constructors of `Language.Haskell.Syntax.Expr.HsMatchContext`) into one: `LamAlt` with a `HsLamVariant` argument. * Similarly, combine `KappaExpr` and `ArrowLamCaseAlt` (constructors of `Language.Haskell.Syntax.Expr.HsArrowMatchContext`) into one: `ArrowLamAlt` with a `HsLamVariant` argument. * Similarly, combine `PsErrLambdaInPat` and `PsErrLambdaCaseInPat` (constructors of `GHC.Parser.Errors.Ppr.PsError`) into one. * Similarly, combine `PsErrLambdaInPat` and `PsErrLambdaCaseInPat` (constructors of `GHC.Parser.Errors.Ppr.PsError`) into one. * In the same `PsError` data type, combine `PsErrLambdaCmdInFunAppCmd` and `PsErrLambdaCaseCmdInFunAppCmd` into one. * In the same `PsError` data tpye, combine `PsErrLambdaInFunAppExpr` and `PsErrLambdaCaseInFunAppExpr` into one. p* Smilarly combine `ExpectedFunTyLam` and `ExpectedFunTyLamCase` (constructors of `GHC.Tc.Types.Origin.ExpectedFunTyOrigin`) into one. Phew! - - - - - b048bea0 by Andreas Klebinger at 2023-09-30T16:09:24-04:00 Arm: Make ppr methods easier to use by not requiring NCGConfig - - - - - 2adc0508 by Andreas Klebinger at 2023-09-30T16:09:24-04:00 AArch64: Fix broken conditional jumps for offsets >= 1MB Rewrite conditional jump instructions with offsets >= 1MB to use unconditional jumps to avoid overflowing the immediate. Fixes #23746 - - - - - 1424f790 by Alan Zimmerman at 2023-09-30T16:10:00-04:00 EPA: Replace Monoid with NoAnn We currently use the Monoid class as a constraint on Exact Print Annotation functions, so we can use mempty. But this leads to requiring Semigroup instances too, which do not always make sense. Instead, introduce a class NoAnn, with a function noAnn analogous to mempty. Closes #20372 Updates haddock submodule - - - - - c1a3ecde by Ben Gamari at 2023-09-30T16:10:36-04:00 users-guide: Refactor handling of :base-ref: et al. - - - - - bc204783 by Richard Eisenberg at 2023-10-02T14:50:52+02:00 Simplify and correct nasty case in coercion opt This fixes #21062. No test case, because triggering this code seems challenging. - - - - - 9c9ca67e by Andrew Lelechenko at 2023-10-04T05:42:28-04:00 Bump bytestring submodule to 0.12.0.2 - - - - - 4e46dc2b by Andrew Lelechenko at 2023-10-04T05:42:28-04:00 Inline bucket_match - - - - - f6b2751f by Ben Gamari at 2023-10-04T05:43:05-04:00 configure: Fix #21712 again This is a bit of a shot in the dark to fix #24033, which appears to be another instance of #21712. For some reason the ld-override logic *still* appears to be active on Darwin targets (or at least one). Consequently, on misconfigured systems we may choose a non-`ld64` linker. It's a bit unclear exactly what happened in #24033 but ultimately the check added for #21712 was not quite right, checking for the `ghc_host_os` (the value of which depends upon the bootstrap compiler) instead of the target platform. Fix this. Fixes #24033. - - - - - 2f0a101d by Krzysztof Gogolewski at 2023-10-04T05:43:42-04:00 Add a regression test for #24029 - - - - - 8cee3fd7 by sheaf at 2023-10-04T05:44:22-04:00 Fix non-symbolic children lookup of fixity decl The fix for #23664 did not correctly account for non-symbolic names when looking up children of a given parent. This one-line fix changes that. Fixes #24037 - - - - - a4785b33 by Cheng Shao at 2023-10-04T05:44:59-04:00 rts: fix incorrect ticket reference - - - - - e037f459 by Ben Gamari at 2023-10-04T05:45:35-04:00 users-guide: Fix discussion of -Wpartial-fields * fix a few typos * add a new example showing when the warning fires * clarify the existing example * point out -Wincomplete-record-selects Fixes #24049. - - - - - 8ff3134e by Matthew Pickering at 2023-10-05T05:34:58-04:00 Revert "Pass preprocessor options to C compiler when building foreign C files (#16737)" This reverts commit 1c18d3b41f897f34a93669edaebe6069f319f9e2. `-optP` should pass options to the preprocessor, that might be a very different program to the C compiler, so passing the options to the C compiler is likely to result in `-optP` being useless. Fixes #17185 and #21291 - - - - - 8f6010b9 by Ben Gamari at 2023-10-05T05:35:36-04:00 rts/nonmoving: Fix on LLP64 platforms Previously `NONMOVING_SEGMENT_MASK` and friends were defined with the `UL` size suffix. However, this is wrong on LLP64 platforms like Windows, where `long` is 32-bits. Fixes #23003. Fixes #24042. - - - - - f20d02f8 by Andreas Klebinger at 2023-10-05T05:36:14-04:00 Fix isAArch64Bitmask for 32bit immediates. Fixes #23802 - - - - - 63afb701 by Bryan Richter at 2023-10-05T05:36:49-04:00 Work around perf note fetch failure Addresses #24055. - - - - - 242102f4 by Krzysztof Gogolewski at 2023-10-05T05:37:26-04:00 Add a test for #21348 - - - - - 7d390bce by Rewbert at 2023-10-05T05:38:08-04:00 Fixes #24046 - - - - - 69abb171 by Finley McIlwaine at 2023-10-06T14:06:28-07:00 Ensure unconstrained instance dictionaries get IPE info In the `StgRhsCon` case of `GHC.Stg.Debug.collectStgRhs`, we were not coming up with an initial source span based on the span of the binder, which was causing instance dictionaries without dynamic superclass constraints to not have source locations in their IPE info. Now they do. Resolves #24005 - - - - - 390443b7 by Andreas Klebinger at 2023-10-07T10:00:20-04:00 rts: Split up rts/include/stg/MachRegs.h by arch - - - - - 3685942f by Bryan Richter at 2023-10-07T10:00:56-04:00 Actually set hackage index state Or at least, use a version of the cabal command that *claims* to set the index state. Time will tell. - - - - - 46a0e5be by Bryan Richter at 2023-10-07T10:00:56-04:00 Update hackage index state - - - - - d4b037de by Bryan Richter at 2023-10-07T10:00:56-04:00 Ensure hadrian uses CI's hackage index state - - - - - e206be64 by Andrew Lelechenko at 2023-10-08T15:06:14-04:00 Do not use O_NONBLOCK on regular files or block devices CLC proposal https://github.com/haskell/core-libraries-committee/issues/166 - - - - - a06197c4 by David Binder at 2023-10-08T15:06:55-04:00 Update hpc-bin submodule to 0.69 - - - - - ed6785b6 by David Binder at 2023-10-08T15:06:55-04:00 Update Hadrian with correct path to happy file for hpc-bin - - - - - 94066d58 by Alan Zimmerman at 2023-10-09T21:35:53-04:00 EPA: Introduce HasAnnotation class The class is defined as class HasAnnotation e where noAnnSrcSpan :: SrcSpan -> e This generalises noAnnSrcSpan, and allows noLocA :: (HasAnnotation e) => a -> GenLocated e a noLocA = L (noAnnSrcSpan noSrcSpan) - - - - - 8792a1bc by Ben Gamari at 2023-10-09T21:36:29-04:00 Bump unix submodule to v2.8.3.0 - - - - - e96c51cb by Andreas Klebinger at 2023-10-10T16:44:27+01:00 Add a flag -fkeep-auto-rules to optionally keep auto-generated rules around. The motivation for the flag is given in #21917. - - - - - 3ed58cef by Matthew Pickering at 2023-10-10T19:01:22-04:00 hadrian: Add ghcToolchain to tool args list This allows you to load ghc-toolchain and ghc-toolchain-bin into HLS. - - - - - 476c02d4 by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Normalise triple via config.sub We were not normalising the target triple anymore like we did with the old make build system. Fixes #23856 - - - - - 303dd237 by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Add missing vendor normalisation This is copied from m4/ghc_convert_vendor.m4 Towards #23868 - - - - - 838026c9 by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Add loongarch64 to parseArch Towards #23868 - - - - - 1a5bc0b5 by Matthew Pickering at 2023-10-10T19:01:22-04:00 Add same LD hack to ghc-toolchain In the ./configure script, if you pass the `LD` variable then this has the effect of stopping use searching for a linker and hence passing `-fuse-ld=...`. We want to emulate this logic in ghc-toolchain, if a use explicilty specifies `LD` variable then don't add `-fuse-ld=..` with the goal of making ./configure and ghc-toolchain agree on which flags to use when using the C compiler as a linker. This is quite unsavoury as we don't bake the choice of LD into the configuration anywhere but what's important for now is making ghc-toolchain and ./configure agree as much as possible. See #23857 for more discussion - - - - - 42d50b5a by Ben Gamari at 2023-10-10T19:01:22-04:00 ghc-toolchain: Check for C99 support with -std=c99 Previously we failed to try enabling C99 support with `-std=c99`, as `autoconf` attempts. This broke on older compilers (e.g. CentOS 7) which don't enable C99 by default. Fixes #23879. - - - - - da2961af by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Add endianess check using __BYTE_ORDER__ macro In very old toolchains the BYTE_ORDER macro is not set but thankfully the __BYTE_ORDER__ macro can be used instead. - - - - - d8da73cd by Matthew Pickering at 2023-10-10T19:01:22-04:00 configure: AC_PATH_TARGET_TOOL for LD We want to make sure that LD is set to an absolute path in order to be consistent with the `LD=$(command -v ld)` call. The AC_PATH_TARGET_TOOL macro uses the absolute path rather than AC_CHECK_TARGET_TOOL which might use a relative path. - - - - - 171f93cc by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Check whether we need -std=gnu99 for CPP as well In ./configure the C99 flag is passed to the C compiler when used as a C preprocessor. So we also check the same thing in ghc-toolchain. - - - - - 89a0918d by Matthew Pickering at 2023-10-10T19:01:22-04:00 Check for --target linker flag separately to C compiler There are situations where the C compiler doesn't accept `--target` but when used as a linker it does (but doesn't do anything most likely) In particular with old gcc toolchains, the C compiler doesn't support --target but when used as a linker it does. - - - - - 37218329 by Matthew Pickering at 2023-10-10T19:01:22-04:00 Use Cc to compile test file in nopie check We were attempting to use the C compiler, as a linker, to compile a file in the nopie check, but that won't work in general as the flags we pass to the linker might not be compatible with the ones we pass when using the C compiler. - - - - - 9b2dfd21 by Matthew Pickering at 2023-10-10T19:01:22-04:00 configure: Error when ghc-toolchain fails to compile This is a small QOL change as if you are working on ghc-toolchain and it fails to compile then configure will continue and can give you outdated results. - - - - - 1f0de49a by Matthew Pickering at 2023-10-10T19:01:22-04:00 configure: Check whether -no-pie works when the C compiler is used as a linker `-no-pie` is a flag we pass when using the C compiler as a linker (see pieCCLDOpts in GHC.Driver.Session) so we should test whether the C compiler used as a linker supports the flag, rather than just the C compiler. - - - - - 62cd2579 by Matthew Pickering at 2023-10-10T19:01:22-04:00 ghc-toolchain: Remove javascript special case for --target detection emcc when used as a linker seems to ignore the --target flag, and for consistency with configure which now tests for --target, we remove this special case. - - - - - 0720fde7 by Ben Gamari at 2023-10-10T19:01:22-04:00 toolchain: Don't pass --target to emscripten toolchain As noted in `Note [Don't pass --target to emscripten toolchain]`, emscripten's `emcc` is rather inconsistent with respect to its treatment of the `--target` flag. Avoid this by special-casing this toolchain in the `configure` script and `ghc-toolchain`. Fixes on aspect of #23744. - - - - - 6354e1da by Matthew Pickering at 2023-10-10T19:01:22-04:00 hadrian: Don't pass `--gcc-options` as a --configure-arg to cabal configure Stop passing -gcc-options which mixed together linker flags and non-linker flags. There's no guarantee the C compiler will accept both of these in each mode. - - - - - c00a4bd6 by Ben Gamari at 2023-10-10T19:01:22-04:00 configure: Probe stage0 link flags For consistency with later stages and CC. - - - - - 1f11e7c4 by Sebastian Graf at 2023-10-10T19:01:58-04:00 Stricter Binary.get in GHC.Types.Unit (#23964) I noticed some thunking while looking at Core. This change has very modest, but throughout positive ghc/alloc effect: ``` hard_hole_fits(normal) ghc/alloc 283,057,664 281,620,872 -0.5% geo. mean -0.1% minimum -0.5% maximum +0.0% ``` Fixes #23964. - - - - - a4f1a181 by Bryan Richter at 2023-10-10T19:02:37-04:00 rel_eng/upload.sh cleanups - - - - - 80705335 by doyougnu at 2023-10-10T19:03:18-04:00 ci: add javascript label rule This adds a rule which triggers the javascript job when the "javascript" label is assigned to an MR. - - - - - a2c0fff6 by Matthew Craven at 2023-10-10T19:03:54-04:00 Make 'wWarningFlagsDeps' include every WarningFlag Fixes #24071. - - - - - d055f099 by Jan Hrček at 2023-10-10T19:04:33-04:00 Fix pretty printing of overlap pragmas in TH splices (fixes #24074) - - - - - 0746b868 by Andreas Klebinger at 2023-10-10T19:05:09-04:00 Aarch64 NCG: Use encoded immediates for literals. Try to generate instr x2, <imm> instead of mov x1, lit instr x2, x1 When possible. This get's rid if quite a few redundant mov instructions. I believe this causes a metric decrease for LargeRecords as we reduce register pressure. ------------------------- Metric Decrease: LargeRecord ------------------------- - - - - - 739f4e6f by Andreas Klebinger at 2023-10-10T19:05:09-04:00 AArch NCG: Refactor getRegister' Remove some special cases which can be handled just as well by the generic case. This increases code re-use while also fixing #23749. Since some of the special case wasn't upholding Note [Signed arithmetic on AArch64]. - - - - - 1b213d33 by Andreas Klebinger at 2023-10-10T19:05:09-04:00 Aarch ncg: Optimize immediate use for address calculations When the offset doesn't fit into the immediate we now just reuse the general getRegister' code path which is well optimized to compute the offset into a register instead of a special case for CmmRegOff. This means we generate a lot less code under certain conditions which is why performance metrics for these improve. ------------------------- Metric Decrease: T4801 T5321FD T5321Fun ------------------------- - - - - - b7df0732 by John Ericson at 2023-10-11T16:02:11-04:00 RTS configure: Move over mem management checks These are for heap allocation, a strictly RTS concern. All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. The RTS configure one has a new ``` AC_CHECK_SIZEOF([void *]) ``` that the top-level configure version didn't have, so that `ac_cv_sizeof_void_p` is defined. Once more code is moved over in latter commits, that can go away. Progress towards #17191 - - - - - 41130a65 by John Ericson at 2023-10-11T16:02:11-04:00 RTS configure: Move over `__thread` check This used by (@bgamari thinks) the `GCThread` abstraction in the RTS. All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. Progress towards #17191 - - - - - cc5ec2bd by John Ericson at 2023-10-11T16:02:11-04:00 RTS configure: Move over misc function checks These are for general use in the RTS. All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. Progress towards #17191 - - - - - 809e7c2d by John Ericson at 2023-10-11T16:02:11-04:00 RTS configure: Move over `eventfd` check This check is for the RTS part of the event manager and has a corresponding part in `base`. All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. Progress towards #17191 - - - - - 58f3babf by John Ericson at 2023-10-11T16:02:48-04:00 Split `FP_CHECK_PTHREADS` and move part to RTS configure `NEED_PTHREAD_LIB` is unused since 3609340743c1b25fdfd0e18b1670dac54c8d8623 (part of the make build system), and so is no longer defined. Progress towards #17191 - - - - - e99cf237 by Moritz Angermann at 2023-10-11T16:03:24-04:00 nativeGen: section flags for .text$foo only Commit 3ece9856d157c85511d59f9f862ab351bbd9b38b, was supposed to fix #22834 in !9810. It does however add "xr" indiscriminatly to .text sections even if splitSections is disabled. This leads to the assembler saying: ghc_1.s:7849:0: error: Warning: Ignoring changed section attributes for .text | 7849 | .section .text,"xr" | ^ - - - - - f383a242 by Sylvain Henry at 2023-10-11T16:04:04-04:00 Modularity: pass TempDir instead of DynFlags (#17957) - - - - - 34fc28b0 by John Ericson at 2023-10-12T06:48:28-04:00 Test that functions from `mingwex` are available Ryan wrote these two minimizations, but they never got added to the test suite. See #23309, #23378 Co-Authored-By: Ben Gamari <bgamari.foss at gmail.com> Co-Authored-By: Ryan Scott <ryan.gl.scott at gmail.com> - - - - - bdb54a0e by John Ericson at 2023-10-12T06:48:28-04:00 Do not check for the `mingwex` library in `/configure` See the recent discussion in !10360 --- Cabal will itself check for the library for the packages that need it, and while the autoconf check additionally does some other things like define a `HAS_LIBMINGWEX` C Preprocessor macro, those other things are also unused and unneeded. Progress towards #17191, which aims to get rid of `/configure` entirely. - - - - - 43e814e1 by Ben Gamari at 2023-10-12T06:49:40-04:00 base: Introduce move modules into src The only non-move changes here are whitespace changes to pass the `whitespace` test and a few testsuite adaptations. - - - - - df81536f by Moritz Angermann at 2023-10-12T06:50:16-04:00 [PEi386 linker] Bounds check and null-deref guard We should resonably be able to expect that we won't exceed the number of sections if we assume to be dealing with legal object files. We can however not guarantee that we get some negative values, and while we try to special case most, we should exclude negative indexing into the sections array. We also need to ensure that we do not try to derefences targetSection, if it is NULL, due to the switch statement. - - - - - c74c4f00 by John Ericson at 2023-10-12T10:31:13-04:00 Move apple compat check to RTS configure - - - - - c80778ea by John Ericson at 2023-10-12T10:31:13-04:00 Move clock/timer fun checks to RTS configure Actual library check (which will set the Cabal flag) is left in the top-level configure for now. Progress towards #17191 - - - - - 7f9f2686 by John Ericson at 2023-10-12T10:31:13-04:00 Move visibility and "musttail" annotation checks to the RTS configure All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. Progress towards #17191 - - - - - ffb3efe6 by John Ericson at 2023-10-12T10:31:13-04:00 Move leading underscore checks to RTS configure `CabalLeadingUnderscore` is done via Hadrian already, so we can stop `AC_SUBST`ing it completely. - - - - - 25fa4b02 by John Ericson at 2023-10-12T10:31:13-04:00 Move alloca, fork, const, and big endian checks to RTS configure All of this should boil down to `AC_DEFINE` not `AC_SUBST`, so it belongs in the RTS configure and should be safe to move without modification. - - - - - 5170f42a by John Ericson at 2023-10-12T10:31:13-04:00 Move libdl check to RTS configure - - - - - ea7a1447 by John Ericson at 2023-10-12T10:31:13-04:00 Adjust `FP_FIND_LIBFFI` Just set vars, and `AC_SUBST` in top-level configure. Don't define `HAVE_SYSTEM_LIBFFI` because nothing is using it. It hasn't be in used since 3609340743c1b25fdfd0e18b1670dac54c8d8623 (part of the make build system). - - - - - f399812c by John Ericson at 2023-10-12T10:31:13-04:00 Split BFD support to RTS configure The flag is still in the top-level configure, but the other checks (which define various macros --- important) are in the RTS configure. - - - - - f64f44e9 by John Ericson at 2023-10-12T10:31:13-04:00 Split libm check between top level and RTS - - - - - dafc4709 by Moritz Angermann at 2023-10-12T10:31:49-04:00 CgUtils.fixStgRegStmt respect register width This change ensure that the reg + offset computation is always of the same size. Before this we could end up with a 64bit register, and then add a 32bit offset (on 32bit platforms). This not only would fail type sanity checking, but also incorrectly truncate 64bit values into 32bit values silently on 32bit architectures. - - - - - 9e6ef7ba by Matthew Pickering at 2023-10-12T20:35:00-04:00 hadrian: Decrease verbosity of cabal commands In Normal, most tools do not produce output to stdout unless there are error conditions. Reverts 7ed65f5a1bc8e040e318ccff395f53a9bbfd8217 - - - - - 08fc27af by John Ericson at 2023-10-12T20:35:36-04:00 Do not substitute `@...@` for stage-specific values in cabal files `rts` and `ghc-prim` now no longer have a `*.cabal.in` to set Cabal flag defaults; instead manual choices are passed to configure in the usual way. The old way was fundamentally broken, because it meant we were baking these Cabal files for a specific stage. Now we only do stage-agnostic @...@ substitution in cabal files (the GHC version), and so all stage-specific configuration is properly confined to `_build` and the right stage dir. Also `include-ghc-prim` is a flag that no longer exists for `ghc-prim` (it was removed in 835d8ddbbfb11796ea8a03d1806b7cee38ba17a6) so I got rid of it. Co-Authored-By: Matthew Pickering <matthewtpickering at gmail.com> - - - - - a0ac8785 by Sebastian Graf at 2023-10-14T19:17:12-04:00 Fix restarts in .ghcid Using the whole of `hadrian/` restarted in a loop for me. - - - - - fea9ecdb by Sebastian Graf at 2023-10-14T19:17:12-04:00 CorePrep: Refactor FloatingBind (#23442) A drastically improved architecture for local floating in CorePrep that decouples the decision of whether a float is going to be let- or case-bound from how far it can float (out of strict contexts, out of lazy contexts, to top-level). There are a couple of new Notes describing the effort: * `Note [Floating in CorePrep]` for the overview * `Note [BindInfo and FloatInfo]` for the new classification of floats * `Note [Floats and FloatDecision]` for how FloatInfo is used to inform floating decisions This is necessary ground work for proper treatment of Strict fields and unlifted values at top-level. Fixes #23442. NoFib results (omitted = 0.0%): ``` -------------------------------------------------------------------------------- Program Allocs Instrs -------------------------------------------------------------------------------- pretty 0.0% -1.6% scc 0.0% -1.7% -------------------------------------------------------------------------------- Min 0.0% -1.7% Max 0.0% -0.0% Geometric Mean -0.0% -0.0% ``` - - - - - 32523713 by Matthew Pickering at 2023-10-14T19:17:49-04:00 hadrian: Move ghcBinDeps into ghcLibDeps This completes a5227080b57cb51ac34d4c9de1accdf6360b818b, the `ghc-usage.txt` and `ghci-usage.txt` file are also used by the `ghc` library so need to make sure they are present in the libdir even if we are not going to build `ghc-bin`. This also fixes things for cross compilers because the stage2 cross-compiler requires the ghc-usage.txt file, but we are using the stage2 lib folder but not building stage3:exe:ghc-bin so ghc-usage.txt was not being generated. - - - - - ec3c4488 by sheaf at 2023-10-14T19:18:29-04:00 Combine GREs when combining in mkImportOccEnv In `GHC.Rename.Names.mkImportOccEnv`, we sometimes discard one import item in favour of another, as explained in Note [Dealing with imports] in `GHC.Rename.Names`. However, this can cause us to lose track of important parent information. Consider for example #24084: module M1 where { class C a where { type T a } } module M2 ( module M1 ) where { import M1 } module M3 where { import M2 ( C, T ); instance C () where T () = () } When processing the import list of `M3`, we start off (for reasons that are not relevant right now) with two `Avail`s attached to `T`, namely `C(C, T)` and `T(T)`. We combine them in the `combine` function of `mkImportOccEnv`; as described in Note [Dealing with imports] we discard `C(C, T)` in favour of `T(T)`. However, in doing so, we **must not** discard the information want that `C` is the parent of `T`. Indeed, losing track of this information can cause errors when importing, as we could get an error of the form ‘T’ is not a (visible) associated type of class ‘C’ We fix this by combining the two GREs for `T` using `plusGRE`. Fixes #24084 - - - - - 257c2807 by Ilias Tsitsimpis at 2023-10-14T19:19:07-04:00 hadrian: Pass -DNOSMP to C compiler when needed Hadrian passes the -DNOSMP flag to GHC when the target doesn't support SMP, but doesn't pass it to CC as well, leading to the following compilation error on mips64el: | Run Cc (FindCDependencies CDep) Stage1: rts/sm/NonMovingScav.c => _build/stage1/rts/build/c/sm/NonMovingScav.o.d Command line: /usr/bin/mips64el-linux-gnuabi64-gcc -E -MM -MG -MF _build/stage1/rts/build/c/hooks/FlagDefaults.thr_debug_p_o.d -MT _build/stage1/rts/build/c/hooks/FlagDefaults.o -Irts/include -I_build/stage1/rts/build -I_build/stage1/rts/build/include -Irts/include -x c rts/hooks/FlagDefaults.c -Wall -Wextra -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Winline -Wpointer-arith -Wmissing-noreturn -Wnested-externs -Wredundant-decls -Wundef -fno-strict-aliasing -DTHREADED_RTS -DDEBUG -fomit-frame-pointer -O2 -g -Irts -I_build/stage1/rts/build -DDEBUG -fno-omit-frame-pointer -g3 -O0 ===> Command failed with error code: 1 In file included from rts/include/Stg.h:348, from rts/include/Rts.h:38, from rts/hooks/FlagDefaults.c:8: rts/include/stg/SMP.h:416:2: error: #error memory barriers unimplemented on this architecture 416 | #error memory barriers unimplemented on this architecture | ^~~~~ rts/include/stg/SMP.h:440:2: error: #error memory barriers unimplemented on this architecture 440 | #error memory barriers unimplemented on this architecture | ^~~~~ rts/include/stg/SMP.h:464:2: error: #error memory barriers unimplemented on this architecture 464 | #error memory barriers unimplemented on this architecture | ^~~~~ The old make system correctly passed this flag to both GHC and CC [1]. Fix this error by passing -DNOSMP to CC as well. [1] https://gitlab.haskell.org/ghc/ghc/-/blob/00920f176b0235d5bb52a8e054d89a664f8938fe/rts/ghc.mk#L407 Closes #24082 - - - - - 13d3c613 by John Ericson at 2023-10-14T19:19:42-04:00 Users Guide: Drop dead code for Haddock refs to `parallel` I noticed while working on !11451 that `@LIBRARY_parallel_UNIT_ID@` was not substituted. It is dead code -- there is no `parallel-ref` usages and it doesn't look like there ever was (going back to 3e5d0f188d6c8633e55e9ba6c8941c07e459fa4b), so let's delete it. - - - - - fe067577 by Sylvain Henry at 2023-10-18T19:40:25-04:00 Avoid out-of-bound array access in bigNatIsPowerOf2 (fix #24066) bigNatIndex# in the `where` clause wasn't guarded by "bigNatIsZero a". - - - - - cc1625b1 by Sylvain Henry at 2023-10-18T19:40:25-04:00 Bignum: fix right shift of negative BigNat with native backend - - - - - cbe4400d by Sylvain Henry at 2023-10-18T19:40:25-04:00 Rts: expose rtsOutOfBoundsAccess symbol - - - - - 72c7380c by Sylvain Henry at 2023-10-18T19:40:25-04:00 Hadrian: enable `-fcheck-prim-bounds` in validate flavour This allows T24066 to fail when the bug is present. Otherwise the out-of-bound access isn't detected as it happens in ghc-bignum which wasn't compiled with the bounds check. - - - - - f9436990 by John Ericson at 2023-10-18T19:41:01-04:00 Make Hadrian solely responsible for substituting `docs/users_guide/ghc_config.py.in` Fixes #24091 Progress on #23966 Issue #24091 reports that `@ProjectVersion@` is no longer being substituted in the GHC user's guide. I assume this is a recent issue, but I am not sure how it's worked since c1a3ecde720b3bddc2c8616daaa06ee324e602ab; it looks like both Hadrian and configure are trying to substitute the same `.in` file! Now only Hadrian does. That is better anyways; already something that issue #23966 requested. It seems like we were missing some dependencies in Hadrian. (I really, really hate that this is possible!) Hopefully it is fixed now. - - - - - b12df0bb by John Ericson at 2023-10-18T19:41:37-04:00 `ghcversion.h`: No need to cope with undefined `ProjectPatchLevel*` Since 4e6c80197f1cc46dfdef0300de46847c7cfbdcb0, these are guaranteed to be defined. (Guaranteed including a test in the testsuite.) - - - - - 0295375a by John Ericson at 2023-10-18T19:41:37-04:00 Generate `ghcversion.h` from a `.in` file Now that there are no conditional sections (see the previous commit), we can just a do simple substitution rather than pasting it together line by line. Progress on #23966 - - - - - 740a1b85 by Krzysztof Gogolewski at 2023-10-19T11:37:20-04:00 Add a regression test for #24064 - - - - - 921fbf2f by Hécate Moonlight at 2023-10-19T11:37:59-04:00 CLC Proposal #182: Export List from Data.List Proposal link: https://github.com/haskell/core-libraries-committee/issues/182 - - - - - 4f02d3c1 by Sylvain Henry at 2023-10-20T04:01:32-04:00 rts: fix small argument passing on big-endian arch (fix #23387) - - - - - b86243b4 by Sylvain Henry at 2023-10-20T04:02:13-04:00 Interpreter: fix literal alignment on big-endian architectures (fix #19261) Literals weren't correctly aligned on big-endian, despite what the comment said. - - - - - a4b2ec47 by Sylvain Henry at 2023-10-20T04:02:54-04:00 Testsuite: recomp011 and recomp015 are fixed on powerpc These tests have been fixed but not tested and re-enabled on big-endian powerpc (see comments in #11260 and #11323) - - - - - fded7dd4 by Sebastian Graf at 2023-10-20T04:03:30-04:00 CorePrep: Allow floating dictionary applications in -O0 into a Rec (#24102) - - - - - 02efc181 by John Ericson at 2023-10-22T02:48:55-04:00 Move function checks to RTS configure Some of these functions are used in `base` too, but we can copy the checks over to its configure if that's an issue. - - - - - 5f4bccab by John Ericson at 2023-10-22T02:48:55-04:00 Move over a number of C-style checks to RTS configure - - - - - 5cf04f58 by John Ericson at 2023-10-22T02:48:55-04:00 Move/Copy more `AC_DEFINE` to RTS config Only exception is the LLVM version macros, which are used for GHC itself. - - - - - b8ce5dfe by John Ericson at 2023-10-22T02:48:55-04:00 Define `TABLES_NEXT_TO_CODE` in the RTS configure We create a new cabal flag to facilitate this. - - - - - 4a40271e by John Ericson at 2023-10-22T02:48:55-04:00 Configure scripts: `checkOS`: Make a bit more robust `mingw64` and `mingw32` are now both accepted for `OSMinGW32`. This allows us to cope with configs/triples that we haven't normalized extra being what GNU `config.sub` does. - - - - - 16bec0a0 by John Ericson at 2023-10-22T02:48:55-04:00 Generate `ghcplatform.h` from RTS configure We create a new cabal flag to facilitate this. - - - - - 7dfcab2f by John Ericson at 2023-10-22T02:48:55-04:00 Get rid of all mention of `mk/config.h` The RTS configure script is now solely responsible for managing its headers; the top level configure script does not help. - - - - - c1e3719c by Cheng Shao at 2023-10-22T02:49:33-04:00 rts: drop stale mentions of MIN_UPD_SIZE We used to have MIN_UPD_SIZE macro that describes the minimum reserved size for thunks, so that the thunk can be overwritten in place as indirections or blackholes. However, this macro has not been actually defined or used anywhere since a long time ago; StgThunkHeader already reserves a padding word for this purpose. Hence this patch which drops stale mentions of MIN_UPD_SIZE. - - - - - d24b0d85 by Andrew Lelechenko at 2023-10-22T02:50:11-04:00 base changelog: move non-backported entries from 4.19 section to 4.20 Neither !10933 (check https://hackage.haskell.org/package/base-4.19.0.0/docs/src/Text.Read.Lex.html#numberToRangedRational) nor !10189 (check https://hackage.haskell.org/package/base-4.19.0.0/docs/src/Data.List.NonEmpty.html#unzip) were backported to `base-4.19.0.0`. Moving them to `base-4.20.0.0` section. Also minor stylistic changes to other entries, bringing them to a uniform form. - - - - - de78b32a by Alan Zimmerman at 2023-10-23T09:09:41-04:00 EPA Some tweaks to annotations - Fix span for GRHS - Move TrailingAnns from last match to FunBind - Fix GADT 'where' clause span - Capture full range for a CaseAlt Match - - - - - d5a8780d by Simon Hengel at 2023-10-23T09:10:23-04:00 Update primitives.rst - - - - - 4d075924 by Josh Meredith at 2023-10-24T23:04:12+11:00 JS/userguide: add explanation of writing jsbits - - - - - 07ab5cc1 by Cheng Shao at 2023-10-24T15:40:32-04:00 testsuite: increase timeout of ghc-api tests for wasm32 ghc-api tests for wasm32 are more likely to timeout due to the large wasm module sizes, especially when testing with wasm native tail calls, given wasmtime's handling of tail call opcodes are suboptimal at the moment. It makes sense to increase timeout specifically for these tests on wasm32. This doesn't affect other targets, and for wasm32 we don't increase timeout for all tests, so not to risk letting major performance regressions slip through the testsuite. - - - - - 0d6acca5 by Greg Steuck at 2023-10-26T08:44:23-04:00 Explicitly require RLIMIT_AS before use in OSMem.c This is done elsewhere in the source tree. It also suddenly is required on OpenBSD. - - - - - 9408b086 by Sylvain Henry at 2023-10-26T08:45:03-04:00 Modularity: modularize external linker Decouple runLink from DynFlags to allow calling runLink more easily. This is preliminary work for calling Emscripten's linker (emcc) from our JavaScript linker. - - - - - e0f35030 by doyougnu at 2023-10-27T08:41:12-04:00 js: add JStg IR, remove unsaturated constructor - Major step towards #22736 and adding the optimizer in #22261 - - - - - 35587eba by Simon Peyton Jones at 2023-10-27T08:41:48-04:00 Fix a bug in tail calls with ticks See #24078 for the diagnosis. The change affects only the Tick case of occurrence analysis. It's a bit hard to test, so no regression test (yet anyway). - - - - - 9bc5cb92 by Matthew Craven at 2023-10-28T07:06:17-04:00 Teach tag-inference about SeqOp/seq# Fixes the STG/tag-inference analogue of #15226. Co-Authored-By: Simon Peyton Jones <simon.peytonjones at gmail.com> - - - - - 34f06334 by Moritz Angermann at 2023-10-28T07:06:53-04:00 [PEi386] Mask SYM_TYPE_DUP_DISCARD in makeSymbolExtra 48e391952c17ff7eab10b0b1456e3f2a2af28a9b introduced `SYM_TYPE_DUP_DISCARD` to the bitfield. The linker however, failed to mask the `SYM_TYPE_DUP_DISCARD` value. Thus `== SYM_TYPE_CODE` comparisons easily failed. This lead to us relocating DATA lookups (GOT) into E8 (call) and E9 (jump) instructions. - - - - - 5b51b2a2 by Mario Blažević at 2023-10-28T07:07:33-04:00 Fix and test for issue #24111, TH.Ppr output of pattern synonyms - - - - - 723bc352 by Alan Zimmerman at 2023-10-30T20:36:41-04:00 EPA: print doc comments as normal comments And ignore the ones allocated in haddock processing. It does not guarantee that every original haddock-like comment appears in the output, as it discards ones that have no legal attachment point. closes #23459 - - - - - 21b76843 by Simon Peyton Jones at 2023-10-30T20:37:17-04:00 Fix non-termination bug in equality solver constraint left-to-right then right to left, forever. Easily fixed. - - - - - 270867ac by Sebastian Graf at 2023-10-30T20:37:52-04:00 ghc-toolchain: build with `-package-env=-` (#24131) Otherwise globally installed libraries (via `cabal install --lib`) break the build. Fixes #24131. - - - - - 4d08364e by Alan Zimmerman at 2023-10-31T19:46:45+00:00 EPA: match changes in GHC - EPA: Comments in AnchorOperation - EPA: Remove EpaEofComment - - - - - 7a90020f by Krzysztof Gogolewski at 2023-10-31T20:03:37-04:00 docs: fix ScopedTypeVariables example (#24101) The previous example didn't compile. Furthermore, it wasn't demonstrating the point properly. I have changed it to an example which shows that 'a' in the signature must be the same 'a' as in the instance head. - - - - - 49f69f50 by Krzysztof Gogolewski at 2023-10-31T20:04:13-04:00 Fix pretty-printing of type family dependencies "where" should be after the injectivity annotation. - - - - - 73c191c0 by Ben Gamari at 2023-10-31T20:04:49-04:00 gitlab-ci: Bump LLVM bootstrap jobs to Debian 12 As the Debian 10 images have too old an LLVM. Addresses #24056. - - - - - 5b0392e0 by Matthew Pickering at 2023-10-31T20:04:49-04:00 ci: Run aarch64 llvm backend job with "LLVM backend" label This brings it into line with the x86 LLVM backend job. - - - - - 9f9c9227 by Ryan Scott at 2023-11-01T09:19:12-04:00 More robust checking for DataKinds As observed in #22141, GHC was not doing its due diligence in catching code that should require `DataKinds` in order to use. Most notably, it was allowing the use of arbitrary data types in kind contexts without `DataKinds`, e.g., ```hs data Vector :: Nat -> Type -> Type where ``` This patch revamps how GHC tracks `DataKinds`. The full specification is written out in the `DataKinds` section of the GHC User's Guide, and the implementation thereof is described in `Note [Checking for DataKinds]` in `GHC.Tc.Validity`. In brief: * We catch _type_-level `DataKinds` violations in the renamer. See `checkDataKinds` in `GHC.Rename.HsType` and `check_data_kinds` in `GHC.Rename.Pat`. * We catch _kind_-level `DataKinds` violations in the typechecker, as this allows us to catch things that appear beneath type synonyms. (We do *not* want to do this in type-level contexts, as it is perfectly fine for a type synonym to mention something that requires DataKinds while still using the type synonym in a module that doesn't enable DataKinds.) See `checkValidType` in `GHC.Tc.Validity`. * There is now a single `TcRnDataKindsError` that classifies all manner of `DataKinds` violations, both in the renamer and the typechecker. The `NoDataKindsDC` error has been removed, as it has been subsumed by `TcRnDataKindsError`. * I have added `CONSTRAINT` is `isKindTyCon`, which is what checks for illicit uses of data types at the kind level without `DataKinds`. Previously, `isKindTyCon` checked for `Constraint` but not `CONSTRAINT`. This is inconsistent, given that both `Type` and `TYPE` were checked by `isKindTyCon`. Moreover, it thwarted the implementation of the `DataKinds` check in `checkValidType`, since we would expand `Constraint` (which was OK without `DataKinds`) to `CONSTRAINT` (which was _not_ OK without `DataKinds`) and reject it. Now both are allowed. * I have added a flurry of additional test cases that test various corners of `DataKinds` checking. Fixes #22141. - - - - - 575d7690 by Sylvain Henry at 2023-11-01T09:19:53-04:00 JS: fix FFI "wrapper" and "dynamic" Fix codegen and helper functions for "wrapper" and "dynamic" foreign imports. Fix tests: - ffi006 - ffi011 - T2469 - T4038 Related to #22363 - - - - - 81fb8885 by Alan Zimmerman at 2023-11-01T22:23:56-04:00 EPA: Use full range for Anchor This change requires a series of related changes, which must all land at the same time, otherwise all the EPA tests break. * Use the current Anchor end as prior end Use the original anchor location end as the source of truth for calculating print deltas. This allows original spacing to apply in most cases, only changed AST items need initial delta positions. * Add DArrow to TrailingAnn * EPA Introduce HasTrailing in ExactPrint Use [TrailingAnn] in enterAnn and remove it from ExactPrint (LocatedN RdrName) * In HsDo, put TrailingAnns at top of LastStmt * EPA: do not convert comments to deltas when balancing. * EPA: deal with fallout from getMonoBind * EPA fix captureLineSpacing * EPA print any comments in the span before exiting it * EPA: Add comments to AnchorOperation * EPA: remove AnnEofComment, it is no longer used Updates Haddock submodule - - - - - 03e82511 by Rodrigo Mesquita at 2023-11-01T22:24:32-04:00 Fix in docs regarding SSymbol, SNat, SChar (#24119) - - - - - 362cc693 by Matthew Pickering at 2023-11-01T22:25:08-04:00 hadrian: Update bootstrap plans (9.4.6, 9.4.7, 9.6.2, 9.6.3, 9.8.1) Updating the bootstrap plans with more recent GHC versions. - - - - - 00b9b8d3 by Matthew Pickering at 2023-11-01T22:25:08-04:00 ci: Add 9.8.1 bootstrap testing job - - - - - ef3d20f8 by Matthew Pickering at 2023-11-01T22:25:08-04:00 Compatibility with 9.8.1 as boot compiler This fixes several compatability issues when using 9.8.1 as the boot compiler. * An incorrect version guard on the stack decoding logic in ghc-heap * Some ghc-prim bounds need relaxing * ghc is no longer wired in, so we have to remove the -this-unit-id ghc call. Fixes #24077 - - - - - 6755d833 by Jaro Reinders at 2023-11-03T10:54:42+01:00 Add NCG support for common 64bit operations to the x86 backend. These used to be implemented via C calls which was obviously quite bad for performance for operations like simple addition. Co-authored-by: Andreas Klebinger - - - - - 0dfb1fa7 by Vladislav Zavialov at 2023-11-03T14:08:41-04:00 T2T in Expressions (#23738) This patch implements the T2T (term-to-type) transformation in expressions. Given a function with a required type argument vfun :: forall a -> ... the user can now call it as vfun (Maybe Int) instead of vfun (type (Maybe Int)) The Maybe Int argument is parsed and renamed as a term (HsExpr), but then undergoes a conversion to a type (HsType). See the new function expr_to_type in compiler/GHC/Tc/Gen/App.hs and Note [RequiredTypeArguments and the T2T mapping] Left as future work: checking for puns. - - - - - cc1c7c54 by Duncan Coutts at 2023-11-05T00:23:44-04:00 Add a test for I/O managers It tries to cover the cases of multiple threads waiting on the same fd for reading and multiple threads waiting for writing, including wait cancellation by async exceptions. It should work for any I/O manager, in-RTS or in-Haskell. Unfortunately it will not currently work for Windows because it relies on anonymous unix sockets. It could in principle be ported to use Windows named pipes. - - - - - 2e448f98 by Cheng Shao at 2023-11-05T00:23:44-04:00 Skip the IOManager test on wasm32 arch. The test relies on the sockets API which are not (yet) available. - - - - - fe50eb35 by Cheng Shao at 2023-11-05T00:24:20-04:00 compiler: fix eager blackhole symbol in wasm32 NCG - - - - - af771148 by Cheng Shao at 2023-11-05T00:24:20-04:00 testsuite: fix optasm tests for wasm32 - - - - - 1b90735c by Matthew Pickering at 2023-11-05T00:24:20-04:00 testsuite: Add wasm32 to testsuite arches with NCG The compiler --info reports that wasm32 compilers have a NCG, so we should agree with that here. - - - - - db9a6496 by Alan Zimmerman at 2023-11-05T00:24:55-04:00 EPA: make locA a function, not a field name And use it to generalise reLoc The following for the windows pipeline one. 5.5% Metric Increase: T5205 - - - - - 833e250c by Simon Peyton Jones at 2023-11-05T00:25:31-04:00 Update the unification count in wrapUnifierX Omitting this caused type inference to fail in #24146. This was an accidental omision in my refactoring of the equality solver. - - - - - e451139f by Andreas Klebinger at 2023-11-05T00:26:07-04:00 Remove an accidental git conflict marker from a comment. - - - - - e7da0d25 by Alan Zimmerman at 2023-11-05T11:20:31+00:00 EPA: match changes in GHC, l2l cleanup - - - - - 30baac7a by Tobias Haslop at 2023-11-06T10:50:32+00:00 Add laws relating between Foldable/Traversable with their Bi- superclasses See https://github.com/haskell/core-libraries-committee/issues/205 for discussion. This commit also documents that the tuple instances only satisfy the laws up to lazyness, similar to the documentation added in !9512. - - - - - df626f00 by Tobias Haslop at 2023-11-07T02:20:37-05:00 Elaborate on the quantified superclass of Bifunctor This was requested in the comment https://github.com/haskell/core-libraries-committee/issues/93#issuecomment-1597271700 for when Traversable becomes a superclass of Bitraversable, but similarly applies to Functor/Bifunctor, which already are in a superclass relationship. - - - - - 8217acb8 by Alan Zimmerman at 2023-11-07T02:21:12-05:00 EPA: get rid of l2l and friends Replace them with l2l to convert the location la2la to convert a GenLocated thing Updates haddock submodule - - - - - dd88a260 by Luite Stegeman at 2023-11-07T02:21:53-05:00 JS: remove broken newIdents from JStg Monad GHC.JS.JStg.Monad.newIdents was broken, resulting in duplicate identifiers being generated in h$c1, h$c2, ... . This change removes the broken newIdents. - - - - - 455524a2 by Matthew Craven at 2023-11-09T08:41:59-05:00 Create specially-solved DataToTag class Closes #20532. This implements CLC proposal 104: https://github.com/haskell/core-libraries-committee/issues/104 The design is explained in Note [DataToTag overview] in GHC.Tc.Instance.Class. This replaces the existing `dataToTag#` primop. These metric changes are not "real"; they represent Unique-related flukes triggering on a different set of jobs than they did previously. See also #19414. Metric Decrease: T13386 T8095 Metric Increase: T13386 T8095 Co-authored-by: Simon Peyton Jones <simon.peytonjones at gmail.com> - - - - - a05f4554 by Alan Zimmerman at 2023-11-09T08:42:35-05:00 EPA: get rid of glRR and friends in GHC/Parser.y With the HasLoc and HasAnnotation classes, we can replace a number of type-specific helper functions in the parser with polymorphic ones instead Metric Decrease: MultiLayerModulesTH_Make - - - - - 18498538 by Cheng Shao at 2023-11-09T16:58:12+00:00 ci: bump ci-images for wasi-sdk upgrade - - - - - 52c0fc69 by PHO at 2023-11-09T19:16:22-05:00 Don't assume the current locale is *.UTF-8, set the encoding explicitly primops.txt contains Unicode characters: > LC_ALL=C ./genprimopcode --data-decl < ./primops.txt > genprimopcode: <stdin>: hGetContents: invalid argument (cannot decode byte sequence starting from 226) Hadrian must also avoid using readFile' to read primops.txt because it tries to decode the file with a locale-specific encoding. - - - - - 7233b3b1 by PHO at 2023-11-09T19:17:01-05:00 Use '[' instead of '[[' because the latter is a Bash-ism It doesn't work on platforms where /bin/sh is something other than Bash. - - - - - 6dbab180 by Simon Peyton Jones at 2023-11-09T19:17:36-05:00 Add an extra check in kcCheckDeclHeader_sig Fix #24083 by checking for a implicitly-scoped type variable that is not actually bound. See Note [Disconnected type variables] in GHC.Tc.Gen.HsType For some reason, on aarch64-darwin we saw a 2.8% decrease in compiler allocations for MultiLayerModulesTH_Make; but 0.0% on other architectures. Metric Decrease: MultiLayerModulesTH_Make - - - - - 22551364 by Sven Tennie at 2023-11-11T06:35:22-05:00 AArch64: Delete unused LDATA pseudo-instruction Though there were consuming functions for LDATA, there were no producers. Thus, the removed code was "dead". - - - - - 2a0ec8eb by Alan Zimmerman at 2023-11-11T06:35:59-05:00 EPA: harmonise acsa and acsA in GHC/Parser.y With the HasLoc class, we can remove the acsa helper function, using acsA instead. - - - - - 4ceac14d by Alan Zimmerman at 2023-11-11T15:16:41+00:00 EPA: Replace Anchor with EpaLocation Match GHC - - - - - 7ae517a0 by Teo Camarasu at 2023-11-12T08:04:12-05:00 nofib: bump submodule This includes changes that: - fix building a benchmark with HEAD - remove a Makefile-ism that causes errors in bash scripts Resolves #24178 - - - - - 3f0036ec by Alan Zimmerman at 2023-11-12T08:04:47-05:00 EPA: Replace Anchor with EpaLocation An Anchor has a location and an operation, which is either that it is unchanged or that it has moved with a DeltaPos data Anchor = Anchor { anchor :: RealSrcSpan , anchor_op :: AnchorOperation } An EpaLocation also has either a location or a DeltaPos data EpaLocation = EpaSpan !RealSrcSpan !(Strict.Maybe BufSpan) | EpaDelta !DeltaPos ![LEpaComment] Now that we do not care about always having a location in the anchor, we remove Anchor and replace it with EpaLocation We do this with a type alias initially, to ease the transition. The alias will be removed in time. We also have helpers to reconstruct the AnchorOperation from an EpaLocation. This is also temporary. Updates Haddock submodule - - - - - a7492048 by Alan Zimmerman at 2023-11-12T13:43:07+00:00 EPA: get rid of AnchorOperation Now that the Anchor type is an alias for EpaLocation, remove AnchorOperation. Updates haddock submodule - - - - - 0745c34d by Andrew Lelechenko at 2023-11-13T16:25:07-05:00 Add since annotation for showHFloat - - - - - e98051a5 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 Suppress duplicate librares linker warning of new macOS linker Fixes #24167 XCode 15 introduced a new linker which warns on duplicate libraries being linked. To disable this warning, we pass -Wl,-no_warn_duplicate_libraries as suggested by Brad King in CMake issue #25297. This flag isn't necessarily available to other linkers on darwin, so we must only configure it into the CC linker arguments if valid. - - - - - c411c431 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 testsuite: Encoding test witnesses recent iconv bug is fragile A regression in the new iconv() distributed with XCode 15 and MacOS Sonoma causes the test 'encoding004' to fail in the CP936 roundrip. We mark this test as fragile until this is fixed upstream (rather than broken, since previous versions of iconv pass the test) See #24161 - - - - - ce7fe5a9 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 testsuite: Update to LC_ALL=C no longer being ignored in darwin MacOS seems to have fixed an issue where it used to ignore the variable `LC_ALL` in program invocations and default to using Unicode. Since the behaviour seems to be fixed to account for the locale variable, we mark tests that were previously broken in spite of it as fragile (since they now pass in recent macOS distributions) See #24161 - - - - - e6c803f7 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 darwin: Fix single_module is obsolete warning In XCode 15's linker, -single_module is the default and otherwise passing it as a flag results in a warning being raised: ld: warning: -single_module is obsolete This patch fixes this warning by, at configure time, determining whether the linker supports -single_module (which is likely false for all non-darwin linkers, and true for darwin linkers in previous versions of macOS), and using that information at runtime to decide to pass or not the flag in the invocation. Fixes #24168 - - - - - 929ba2f9 by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 testsuite: Skip MultiLayerModulesTH_Make on darwin The recent toolchain upgrade on darwin machines resulted in the MultiLayerModulesTH_Make test metrics varying too much from the baseline, ultimately blocking the CI pipelines. This commit skips the test on darwin to temporarily avoid failures due to the environment change in the runners. However, the metrics divergence is being investigated still (tracked in #24177) - - - - - af261ccd by Rodrigo Mesquita at 2023-11-15T13:18:58-05:00 configure: check target (not build) understands -no_compact_unwind Previously, we were branching on whether the build system was darwin to shortcut this check, but we really want to branch on whether the target system (which is what we are configuring ld_prog for) is darwin. - - - - - 2125c176 by Luite Stegeman at 2023-11-15T13:19:38-05:00 JS: Fix missing variable declarations The JStg IR update was missing some local variable declarations that were present earlier, causing global variables to be used implicitly (or an error in JavaScript strict mode). This adds the local variable declarations again. - - - - - 99ced73b by Krzysztof Gogolewski at 2023-11-15T13:20:14-05:00 Remove loopy superclass solve mechanism Programs with a -Wloopy-superclass-solve warning will now fail with an error. Fixes #23017 - - - - - 2aff2361 by Zubin Duggal at 2023-11-15T13:20:50-05:00 users-guide: Fix links to libraries from the users-guide. The unit-ids generated in c1a3ecde720b3bddc2c8616daaa06ee324e602ab include the package name, so we don't need to explicitly add it to the links. Fixes #24151 - - - - - 27981fac by Alan Zimmerman at 2023-11-15T13:21:25-05:00 EPA: splitLHsForAllTyInvis does not return ann We did not use the annotations returned from splitLHsForAllTyInvis, so do not return them. - - - - - a6467834 by Krzysztof Gogolewski at 2023-11-15T22:22:59-05:00 Document defaulting of RuntimeReps Fixes #24099 - - - - - 2776920e by Simon Peyton Jones at 2023-11-15T22:23:35-05:00 Second fix to #24083 My earlier fix turns out to be too aggressive for data/type families See wrinkle (DTV1) in Note [Disconnected type variables] - - - - - cee81370 by Sylvain Henry at 2023-11-16T09:57:46-05:00 Fix unusable units and module reexport interaction (#21097) This commit fixes an issue with ModUnusable introduced in df0f148feae. In mkUnusableModuleNameProvidersMap we traverse the list of unusable units and generate ModUnusable origin for all the modules they contain: exposed modules, hidden modules, and also re-exported modules. To do this we have a two-level map: ModuleName -> Unit:ModuleName (aka Module) -> ModuleOrigin So for each module name "M" in broken unit "u" we have: "M" -> u:M -> ModUnusable reason However in the case of module reexports we were using the *target* module as a key. E.g. if "u:M" is a reexport for "X" from unit "o": "M" -> o:X -> ModUnusable reason Case 1: suppose a reexport without module renaming (u:M -> o:M) from unusable unit u: "M" -> o:M -> ModUnusable reason Here it's claiming that the import of M is unusable because a reexport from u is unusable. But if unit o isn't unusable we could also have in the map: "M" -> o:M -> ModOrigin ... Issue: the Semigroup instance of ModuleOrigin doesn't handle the case (ModUnusable <> ModOrigin) Case 2: similarly we could have 2 unusable units reexporting the same module without renaming, say (u:M -> o:M) and (v:M -> o:M) with u and v unusable. It gives: "M" -> o:M -> ModUnusable ... (for u) "M" -> o:M -> ModUnusable ... (for v) Issue: the Semigroup instance of ModuleOrigin doesn't handle the case (ModUnusable <> ModUnusable). This led to #21097, #16996, #11050. To fix this, in this commit we make ModUnusable track whether the module used as key is a reexport or not (for better error messages) and we use the re-export module as key. E.g. if "u:M" is a reexport for "o:X" and u is unusable, we now record: "M" -> u:M -> ModUnusable reason reexported=True So now, we have two cases for a reexport u:M -> o:X: - u unusable: "M" -> u:M -> ModUnusable ... reexported=True - u usable: "M" -> o:X -> ModOrigin ... reexportedFrom=u:M The second case is indexed with o:X because in this case the Semigroup instance of ModOrigin is used to combine valid expositions of a module (directly or via reexports). Note that module lookup functions select usable modules first (those who have a ModOrigin value), so it doesn't matter if we add new ModUnusable entries in the map like this: "M" -> { u:M -> ModUnusable ... reexported=True o:M -> ModOrigin ... } The ModOrigin one will be used. Only if there is no ModOrigin or ModHidden entry will the ModUnusable error be printed. See T21097 for an example printing several reasons why an import is unusable. - - - - - 3e606230 by Krzysztof Gogolewski at 2023-11-16T09:58:22-05:00 Fix IPE test A helper function was defined in a different module than used. To reproduce: ./hadrian/build test --test-root-dirs=testsuite/tests/rts/ipe - - - - - 49f5264b by Andreas Klebinger at 2023-11-16T20:52:11-05:00 Properly compute unpacked sizes for -funpack-small-strict-fields. Use rep size rather than rep count to compute the size. Fixes #22309 - - - - - b4f84e4b by James Henri Haydon at 2023-11-16T20:52:53-05:00 Explicit methods for Alternative Compose Explicitly define some and many in Alternative instance for Data.Functor.Compose Implementation of https://github.com/haskell/core-libraries-committee/issues/181 - - - - - 9bc0dd1f by Ignat Insarov at 2023-11-16T20:53:34-05:00 Add permutations for non-empty lists. Approved by CLC in https://github.com/haskell/core-libraries-committee/issues/68#issuecomment-1221409837 - - - - - 5643ecf9 by Andrew Lelechenko at 2023-11-16T20:53:34-05:00 Update changelog and since annotations for Data.List.NonEmpty.permutations Approved by CLC in https://github.com/haskell/core-libraries-committee/issues/68#issuecomment-1221409837 - - - - - 94ff2134 by Oleg Alexander at 2023-11-16T20:54:15-05:00 Update doc string for traceShow Updated doc string for traceShow. - - - - - faff671a by Luite Stegeman at 2023-11-17T14:12:51+01:00 JS: clean up some foreign imports - - - - - 856e0a4e by Sven Tennie at 2023-11-18T06:54:11-05:00 AArch64: Remove unused instructions As these aren't ever emitted, we don't even know if they work or will ever be used. If one of them is needed in future, we may easily re-add it. Deleted instructions are: - CMN - ANDS - BIC - BICS - EON - ORN - ROR - TST - STP - LDP - DMBSY - - - - - 615441ef by Alan Zimmerman at 2023-11-18T06:54:46-05:00 EPA: Replace Monoid with NoAnn Remove the final Monoid instances in the exact print infrastructure. For Windows CI Metric Decrease: T5205 - - - - - 5a6c49d4 by David Feuer at 2023-11-20T18:53:18-05:00 Speed up stimes in instance Semigroup Endo As discussed at https://github.com/haskell/core-libraries-committee/issues/4 - - - - - cf9da4b3 by Andrew Lelechenko at 2023-11-20T18:53:18-05:00 base: reflect latest changes in the changelog - - - - - 48bf364e by Alan Zimmerman at 2023-11-20T18:53:54-05:00 EPA: Use SrcSpan in EpaSpan This is more natural, since we already need to deal with invalid RealSrcSpans, and that is exactly what SrcSpan.UnhelpfulSpan is for. Updates haddock submodule. - - - - - 97ec37cc by Sebastian Graf at 2023-11-20T18:54:31-05:00 Add regression test for #6070 Fixes #6070. - - - - - e9d5ae41 by Owen Shepherd at 2023-11-21T18:32:23-05:00 chore: Correct typo in the gitlab MR template [skip ci] - - - - - f158a8d0 by Rodrigo Mesquita at 2023-11-21T18:32:59-05:00 Improve error message when reading invalid `.target` files A `.target` file generated by ghc-toolchain or by configure can become invalid if the target representation (`Toolchain.Target`) is changed while the files are not re-generated by calling `./configure` or `ghc-toolchain` again. There is also the issue of hadrian caching the dependencies on `.target` files, which makes parsing fail when reading reading the cached value if the representation has been updated. This patch provides a better error message in both situations, moving away from a terrible `Prelude.read: no parse` error that you would get otherwise. Fixes #24199 - - - - - 955520c6 by Ben Gamari at 2023-11-21T18:33:34-05:00 users guide: Note that QuantifiedConstraints implies ExplicitForAll Fixes #24025. - - - - - 17ec3e97 by Owen Shepherd at 2023-11-22T09:37:28+01:00 fix: Change type signatures in NonEmpty export comments to reflect reality This fixes several typos in the comments of Data.List.NonEmpty export list items. - - - - - 2fd78f9f by Samuel Thibault at 2023-11-22T11:49:13-05:00 Fix the platform string for GNU/Hurd As commited in Cargo https://github.com/haskell/cabal/pull/9434 there is confusion between "gnu" and "hurd". This got fixed in Cargo, we need the converse in Hadrian. Fixes #24180 - - - - - a79960fe by Alan Zimmerman at 2023-11-22T11:49:48-05:00 EPA: Tuple Present no longer has annotation The Present constructor for a Tuple argument will never have an exact print annotation. So make this impossible. - - - - - 121c9ab7 by David Binder at 2023-11-22T21:12:29-05:00 Unify the hpc testsuites The hpc testsuite was split between testsuite/tests/hpc and the submodule libraries/hpc/test. This commit unifies the two testsuites in the GHC repository in the directory testsuite/tests/hpc. - - - - - d2733a05 by Alan Zimmerman at 2023-11-22T21:13:05-05:00 EPA: empty tup_tail has noAnn In Parser.y, the tup_tail rule had the following option | {- empty -} %shift { return [Left noAnn] } Once this works through PostProcess.hs, it means we add an extra Missing constructor if the last item was a comma. Change the annotation type to a Bool to indicate this, and use the EpAnn Anchor for the print location for the others. - - - - - fa576eb8 by Andreas Klebinger at 2023-11-24T08:29:13-05:00 Fix FMA primops generating broken assembly on x86. `genFMA3Code` assumed that we had to take extra precations to avoid overwriting the result of `getNonClobberedReg`. One of these special cases caused a bug resulting in broken assembly. I believe we don't need to hadle these cases specially at all, which means this MR simply deletes the special cases to fix the bug. Fixes #24160 - - - - - 34d86315 by Alan Zimmerman at 2023-11-24T08:29:49-05:00 EPA: Remove parenthesizeHsType This is called from PostProcess.hs, and adds spurious parens. With the looser version of exact printing we had before we could tolerate this, as they would be swallowed by the original at the same place. But with the next change (remove EpAnnNotUsed) they result in duplicates in the output. For Darwin build: Metric Increase: MultiLayerModulesTH_OneShot - - - - - 3ede659d by Vladislav Zavialov at 2023-11-26T06:43:32-05:00 Add name for -Wdeprecated-type-abstractions (#24154) This warning had no name or flag and was triggered unconditionally. Now it is part of -Wcompat. - - - - - 7902ebf8 by Alan Zimmerman at 2023-11-26T06:44:08-05:00 EPA: Remove EpAnnNotUsed We no longer need the EpAnnNotUsed constructor for EpAnn, as we can represent an unused annotation with an anchor having a EpaDelta of zero, and empty comments and annotations. This simplifies code handling annotations considerably. Updates haddock submodule Metric Increase: parsing001 - - - - - 471b2672 by Mario Blažević at 2023-11-26T06:44:48-05:00 Bumped the upper bound of text to <2.2 - - - - - d1bf25c7 by Vladislav Zavialov at 2023-11-26T11:45:49-05:00 Term variable capture (#23740) This patch changes type variable lookup rules (lookupTypeOccRn) and implicit quantification rules (filterInScope) so that variables bound in the term namespace can be captured at the type level {-# LANGUAGE RequiredTypeArguments #-} f1 x = g1 @x -- `x` used in a type application f2 x = g2 (undefined :: x) -- `x` used in a type annotation f3 x = g3 (type x) -- `x` used in an embedded type f4 x = ... where g4 :: x -> x -- `x` used in a type signature g4 = ... This change alone does not allow us to accept examples shown above, but at least it gets them past the renamer. - - - - - da863d15 by Vladislav Zavialov at 2023-11-26T11:46:26-05:00 Update Note [hsScopedTvs and visible foralls] The Note was written before GHC gained support for visible forall in types of terms. Rewrite a few sentences and use a better example. - - - - - b5213542 by Matthew Pickering at 2023-11-27T12:53:59-05:00 testsuite: Add mechanism to collect generic metrics * Generalise the metric logic by adding an additional field which allows you to specify how to query for the actual value. Previously the method of querying the baseline value was abstracted (but always set to the same thing). * This requires rejigging how the stat collection works slightly but now it's more uniform and hopefully simpler. * Introduce some new "generic" helper functions for writing generic stats tests. - collect_size ( deviation, path ) Record the size of the file as a metric - stat_from_file ( metric, deviation, path ) Read a value from the given path, and store that as a metric - collect_generic_stat ( metric, deviation, get_stat) Provide your own `get_stat` function, `lambda way: <Int>`, which can be used to establish the current value of the metric. - collect_generic_stats ( metric_info ): Like collect_generic_stat but provide the whole dictionary of metric definitions. { metric: { deviation: <Int> current: lambda way: <Int> } } * Introduce two new "size" metrics for keeping track of build products. - `size_hello_obj` - The size of `hello.o` from compiling hello.hs - `libdir` - The total size of the `libdir` folder. * Track the number of modules in the AST tests - CountDepsAst - CountDepsParser This lays the infrastructure for #24191 #22256 #17129 - - - - - 7d9a2e44 by ARATA Mizuki at 2023-11-27T12:54:39-05:00 x86: Don't require -mavx2 when using 256-bit floating-point SIMD primitives Fixes #24222 - - - - - 4e5ff6a4 by Alan Zimmerman at 2023-11-27T12:55:15-05:00 EPA: Remove SrcSpanAnn Now that we only have a single constructor for EpAnn, And it uses a SrcSpan for its location, we can do away with SrcSpanAnn completely. It only existed to wrap the original SrcSpan in a location, and provide a place for the exact print annotation. For darwin only: Metric Increase: MultiLayerModulesTH_OneShot Updates haddock submodule - - - - - e05bca39 by Krzysztof Gogolewski at 2023-11-28T08:00:55-05:00 testsuite: don't initialize testdir to '.' The test directory is removed during cleanup, if there's an interrupt that could remove the entire repository. Fixes #24219 - - - - - af881674 by Alan Zimmerman at 2023-11-28T08:01:30-05:00 EPA: Clean up mkScope in Ast.hs Now that we have HasLoc we can get rid of all the custom variants of mkScope For deb10-numa Metric Increase: libdir - - - - - 292983c8 by Ben Gamari at 2023-11-28T22:44:28-05:00 distrib: Rediscover otool and install_name_tool on Darwin In the bindist configure script we must rediscover the `otool` and `install_name_tool`s since they may be different from the build environment. Fixes #24211. - - - - - dfe1c354 by Stefan Schulze Frielinghaus at 2023-11-28T22:45:04-05:00 llvmGen: Align objects in the data section Objects in the data section may be referenced via tagged pointers. Thus, align those objects to a 4- or 8-byte boundary for 32- or 64-bit platforms, respectively. Note, this may need to be reconsidered if objects with a greater natural alignment requirement are emitted as e.g. 128-bit atomics. Fixes #24163. - - - - - f6c486c3 by Matthew Pickering at 2023-11-29T11:08:13-05:00 metrics: Widen libdir and size_hello_obj acceptance window af8816740d9b8759be1a22af8adcb5f13edeb61d shows that the libdir size can fluctuate quite significantly even when the change is quite small. Therefore we widen the acceptance window to 10%. - - - - - 99a6a49c by Alan Zimmerman at 2023-11-29T11:08:49-05:00 EPA: Clean up TC Monad Utils We no longer need the alternative variant of addLocM (addLocMA) nor wrapLocAM, wrapLocSndMA. aarch64-darwin Metric Increase: MultiLayerModulesTH_OneShot deb10-numa-slow Metric Decrease: libdir - - - - - 94fb8d47 by Alan Zimmerman at 2023-11-29T18:10:26+00:00 Match GHC, No comments in EpaDelta for comments - - - - - cbc03fa0 by Sebastian Graf at 2023-11-30T12:37:21-05:00 perf tests: Move comments into new `Note [Sensitivity to unique increment]` (#19414) And additionally to T12545, link from T8095, T13386 to this new Note. - - - - - c7623b22 by Alan Zimmerman at 2023-11-30T12:37:56-05:00 EPA: EpaDelta for comment has no comments EpaLocation is used to position things. It has two constructors, EpaSpan holding a SrcSpan, and EpaDelta with a delta position and a possible list of comments. The comment list is needed because the location in EpaDelta has no absolute information to decide which comments should be emitted before them when printing. But it is also used for specifying the position of a comment. To prevent the absurdity of a comment position having a list of comments in it, we make EpaLocation parameterisable, using comments for the normal case and a constant for within comments. Updates haddock submodule. aarch64-darwin Metric Decrease: MultiLayerModulesTH_OneShot - - - - - bd8acc0c by Krzysztof Gogolewski at 2023-11-30T12:38:32-05:00 Kind-check body of a required forall We now require that in 'forall a -> ty', ty has kind TYPE r for some r. Fixes #24176 - - - - - 010fb784 by Owen Shepherd at 2023-12-03T00:10:09-05:00 docs(NonEmpty/group): Remove incorrect haddock link quotes in code block - - - - - cda9c12d by Owen Shepherd at 2023-12-03T00:10:09-05:00 docs(NonEmpty/group): Remove cycle from group haddock example - - - - - 495265b9 by Owen Shepherd at 2023-12-03T00:10:09-05:00 docs(NonEmpty/group): Use repl haddock syntax in group docs - - - - - d134d1de by Owen Shepherd at 2023-12-03T00:10:09-05:00 docs(NonEmpty/group): Use list [] notation in group haddock - - - - - dfcf629c by Owen Shepherd at 2023-12-03T00:10:10-05:00 docs(NonEmpty/group): Specify final property of group function in haddock - - - - - cad3b734 by Owen Shepherd at 2023-12-03T00:10:10-05:00 fix: Add missing property of List.group - - - - - bad37656 by Matthew Pickering at 2023-12-03T00:10:46-05:00 testsuite: Fix T21097b test with make 4.1 (deb9) cee81370cd6ef256f66035e3116878d4cb82e28b recently added a test which failed on deb9 because the version of make was emitting the recipe failure to stdout rather than stderr. One way to fix this is to be more precise in the test about which part of the output we care about inspecting. - - - - - 5efdf421 by Matthew Pickering at 2023-12-03T00:11:21-05:00 testsuite: Track size of libdir in bytes For consistency it's better if we track all size metrics in bytes. Metric Increase: libdir - - - - - f5eb0f29 by Matthew Pickering at 2023-12-03T00:11:22-05:00 testsuite: Remove rogue trace in testsuite I accidentally left a trace in the generics metric patch. - - - - - d5610737 by Claudio Bley at 2023-12-06T16:13:33-05:00 Only exit ghci in -e mode when :add command fails Previously, when running `ghci -e ':add Sample.hs'` the process would exit with exit code 1 if the file exists and could be loaded. Fixes #24115 - - - - - 0f0c53a5 by Vladislav Zavialov at 2023-12-06T16:14:09-05:00 T2T in Patterns (#23739) This patch implements the T2T (term-to-type) transformation in patterns. Patterns that are checked against a visible forall can now be written without the `type` keyword: \(type t) (x :: t) -> ... -- old \t (x :: t) -> ... -- new The `t` binder is parsed and renamed as a term pattern (Pat), but then undergoes a conversion to a type pattern (HsTyPat). See the new function pat_to_type_pat in compiler/GHC/Tc/Gen/Pat.hs - - - - - 10a1a6c6 by Sebastian Graf at 2023-12-06T16:14:45-05:00 Pmc: Fix SrcLoc and warning for incomplete irrefutable pats (#24234) Before, the source location would point at the surrounding function definition, causing the confusion in #24234. I also took the opportunity to introduce a new `LazyPatCtx :: HsMatchContext _` to make the warning message say "irrefutable pattern" instead of "pattern binding". - - - - - 36b9a38c by Matthew Pickering at 2023-12-06T16:15:21-05:00 libraries: Bump filepath to 1.4.200.1 and unix to 2.8.4.0 Updates filepath submodule Updates unix submodule Fixes #24240 - - - - - 91ff0971 by Matthew Pickering at 2023-12-06T16:15:21-05:00 Submodule linter: Allow references to tags We modify the submodule linter so that if the bumped commit is a specific tag then the commit is accepted. Fixes #24241 - - - - - 86f652dc by Zubin Duggal at 2023-12-06T16:15:21-05:00 hadrian: set -Wno-deprecations for directory and Win32 The filepath bump to 1.4.200.1 introduces a deprecation warning. See https://gitlab.haskell.org/ghc/ghc/-/issues/24240 https://github.com/haskell/filepath/pull/206 - - - - - 7ac6006e by Sylvain Henry at 2023-12-06T16:16:02-05:00 Zap OccInfo on case binders during StgCse #14895 #24233 StgCse can revive dead binders: case foo of dead { Foo x y -> Foo x y; ... } ===> case foo of dead { Foo x y -> dead; ... } -- dead is no longer dead So we must zap occurrence information on case binders. Fix #14895 and #24233 - - - - - 57c391c4 by Sebastian Graf at 2023-12-06T16:16:37-05:00 Cpr: Turn an assertion into a check to deal with some dead code (#23862) See the new `Note [Dead code may contain type confusions]`. Fixes #23862. - - - - - c1c8abf8 by Zubin Duggal at 2023-12-08T02:25:07-05:00 testsuite: add test for #23944 - - - - - 6329d308 by Zubin Duggal at 2023-12-08T02:25:07-05:00 driver: Only run a dynamic-too pipeline if object files are going to be generated Otherwise we run into a panic in hscMaybeWriteIface: "Unexpected DT_Dyn state when writing simple interface" when dynamic-too is enabled We could remove the panic and just write the interface even if the state is `DT_Dyn`, but it seems pointless to run the pipeline twice when `hscMaybeWriteIface` is already designed to write both `hi` and `dyn_hi` files if dynamic-too is enabled. Fixes #23944. - - - - - 28811f88 by Simon Peyton Jones at 2023-12-08T05:47:18-05:00 Improve duplicate elimination in SpecConstr This partially fixes #24229. See the new Note [Pattern duplicate elimination] in SpecConstr - - - - - fec7894f by Simon Peyton Jones at 2023-12-08T05:47:18-05:00 Make SpecConstr deal with casts better This patch does two things, to fix #23209: * It improves SpecConstr so that it no longer quantifies over coercion variables. See Note [SpecConstr and casts] * It improves the rule matcher to deal nicely with the case where the rule does not quantify over coercion variables, but the the template has a cast in it. See Note [Casts in the template] - - - - - 8db8d2fd by Zubin Duggal at 2023-12-08T05:47:54-05:00 driver: Don't lose track of nodes when we fail to resolve cycles The nodes that take part in a cycle should include both hs-boot and hs files, but when we fail to resolve a cycle, we were only counting the nodes from the graph without boot files. Fixes #24196 - - - - - c5b4efd3 by Zubin Duggal at 2023-12-08T05:48:30-05:00 testsuite: Skip MultiLayerModulesTH_OneShot on darwin See #24177 - - - - - fae472a9 by Wendao Lee at 2023-12-08T05:49:12-05:00 docs(Data.Char):Add more detailed descriptions for some functions Related changed function's docs: -GHC.Unicode.isAlpha -GHC.Unicode.isPrint -GHC.Unicode.isAlphaNum Add more details for what the function will return. Co-authored-by: Bodigrim <andrew.lelechenko at gmail.com> - - - - - ca7510e4 by Malik Ammar Faisal at 2023-12-08T05:49:55-05:00 Fix float parsing in GHC Cmm Lexer Add test case for bug #24224 - - - - - d8baa1bd by Simon Peyton Jones at 2023-12-08T15:40:37+00:00 Take care when simplifying unfoldings This MR fixes a very subtle bug exposed by #24242. See Note [Environment for simplLetUnfolding]. I also updated a bunch of Notes on shadowing - - - - - 03ca551d by Simon Peyton Jones at 2023-12-08T15:54:50-05:00 Comments only in FloatIn Relevant to #3458 - - - - - 50c78779 by Simon Peyton Jones at 2023-12-08T15:54:50-05:00 Comments only in SpecConstr - - - - - 9431e195 by Simon Peyton Jones at 2023-12-08T15:54:50-05:00 Add test for #22238 - - - - - d9e4c597 by Vladislav Zavialov at 2023-12-11T04:19:34-05:00 Make forall a keyword (#23719) Before this change, GHC used to accept `forall` as a term-level identifier: -- from constraints-0.13 forall :: forall p. (forall a. Dict (p a)) -> Dict (Forall p) forall d = ... Now it is a parse error. The -Wforall-identifier warning has served its purpose and is now a deprecated no-op. - - - - - 58d56644 by Zubin Duggal at 2023-12-11T04:20:10-05:00 driver: Ensure we actually clear the interactive context before reloading Previously we called discardIC, but immediately after set the session back to an old HscEnv that still contained the IC Partially addresses #24107 Fixes #23405 - - - - - 8e5745a0 by Zubin Duggal at 2023-12-11T04:20:10-05:00 driver: Ensure we force the lookup of old build artifacts before returning the build plan This prevents us from retaining all previous build artifacts in memory until a recompile finishes, instead only retaining the exact artifacts we need. Fixes #24118 - - - - - 105c370c by Zubin Duggal at 2023-12-11T04:20:10-05:00 testsuite: add test for #24118 and #24107 MultiLayerModulesDefsGhci was not able to catch the leak because it uses :l which discards the previous environment. Using :r catches both of these leaks - - - - - e822ff88 by Zubin Duggal at 2023-12-11T04:20:10-05:00 compiler: Add some strictness annotations to ImportSpec and related constructors This prevents us from retaining entire HscEnvs. Force these ImportSpecs when forcing the GlobalRdrEltX Adds an NFData instance for Bag Fixes #24107 - - - - - 522c12a4 by Zubin Duggal at 2023-12-11T04:20:10-05:00 compiler: Force IfGlobalRdrEnv in NFData instance. - - - - - 188b280d by Arnaud Spiwack at 2023-12-11T15:33:31+01:00 LinearTypes => MonoLocalBinds - - - - - 8e0446df by Arnaud Spiwack at 2023-12-11T15:44:28+01:00 Linear let and where bindings For expediency, the initial implementation of linear types in GHC made it so that let and where binders would always be considered unrestricted. This was rather unpleasant, and probably a big obstacle to adoption. At any rate, this was not how the proposal was designed. This patch fixes this infelicity. It was surprisingly difficult to build, which explains, in part, why it took so long to materialise. As of this patch, let or where bindings marked with %1 will be linear (respectively %p for an arbitrary multiplicity p). Unmarked let will infer their multiplicity. Here is a prototypical example of program that used to be rejected and is accepted with this patch: ```haskell f :: A %1 -> B g :: B %1 -> C h :: A %1 -> C h x = g y where y = f x ``` Exceptions: - Recursive let are unrestricted, as there isn't a clear semantics of what a linear recursive binding would be. - Destructive lets with lazy bindings are unrestricted, as their desugaring isn't linear (see also #23461). - (Strict) destructive lets with inferred polymorphic type are unrestricted. Because the desugaring isn't linear (See #18461 down-thread). Closes #18461 and #18739 Co-authored-by: @jackohughes - - - - - effa7e2d by Matthew Craven at 2023-12-12T04:37:20-05:00 Introduce `dataToTagSmall#` primop (closes #21710) ...and use it to generate slightly better code when dataToTag# is used at a "small data type" where there is no need to mess with "is_too_big_tag" or potentially look at an info table. Metric Decrease: T18304 - - - - - 35c7aef6 by Matthew Craven at 2023-12-12T04:37:20-05:00 Fix formatting of Note [alg-alt heap check] - - - - - 7397c784 by Oleg Grenrus at 2023-12-12T04:37:56-05:00 Allow untyped brackets in typed splices and vice versa. Resolves #24190 Apparently the check was essentially always (as far as I can trace back: d0d47ba76f8f0501cf3c4966bc83966ab38cac27), and while it does catch some mismatches, the type-checker will catch them too. OTOH, it prevents writing completely reasonable programs. - - - - - 32d208e1 by Vladislav Zavialov at 2023-12-12T20:41:36+03:00 EPA: Match changes to LHsToken removal - - - - - a3ee3b99 by Moritz Angermann at 2023-12-12T19:50:58-05:00 Drop hard Xcode dependency XCODE_VERSION calls out to `xcodebuild`, which is only available when having `Xcode` installed. The CommandLineTools are not sufficient. To install Xcode, you must have an apple id to download the Xcode.xip from apple. We do not use xcodebuild anywhere in our build explicilty. At best it appears to be a proxy for checking the linker or the compiler. These should rather be done with ``` xcrun ld -version ``` or similar, and not by proxy through Xcode. The CLR should be sufficient for building software on macOS. - - - - - 1c9496e0 by Vladislav Zavialov at 2023-12-12T19:51:34-05:00 docs: update information on RequiredTypeArguments Update the User's Guide and Release Notes to account for the recent progress in the implementation of RequiredTypeArguments. - - - - - d0b17576 by Ben Gamari at 2023-12-13T06:33:37-05:00 rts/eventlog: Fix off-by-one in assertion Previously we failed to account for the NULL terminator `postString` asserted that there is enough room in the buffer for the string. - - - - - a10f9b9b by Ben Gamari at 2023-12-13T06:33:37-05:00 rts/eventlog: Honor result of ensureRoomForVariableEvent is Previously we would keep plugging along, even if isn't enough room for the event. - - - - - 0e0f41c0 by Ben Gamari at 2023-12-13T06:33:37-05:00 rts/eventlog: Avoid truncating event sizes Previously ensureRoomForVariableEvent would truncate the desired size to 16-bits, resulting in #24197. Fixes #24197. - - - - - 64e724c8 by Artin Ghasivand at 2023-12-13T06:34:20-05:00 Remove the "Derived Constraint" argument of TcPluginSolver, docs - - - - - fe6d97dd by Vladislav Zavialov at 2023-12-13T06:34:56-05:00 EPA: Move tokens into GhcPs extension fields (#23447) Summary of changes * Remove Language.Haskell.Syntax.Concrete * Move all tokens into GhcPs extension fields (LHsToken -> EpToken) * Create new TTG extension fields as needed * Drop the MultAnn wrapper Updates the haddock submodule. Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - 8106e695 by Zubin Duggal at 2023-12-13T06:35:34-05:00 testsuite: use copy_files in T23405 This prevents the tree from being dirtied when the file is modified. - - - - - ed0e4099 by Bryan Richter at 2023-12-14T04:30:53-05:00 Document ghc package's PVP-noncompliance This changes nothing, it just makes the status quo explicit. - - - - - 8bef8d9f by Luite Stegeman at 2023-12-14T04:31:33-05:00 JS: Mark spurious CI failures js_fragile(24259) This marks the spurious test failures on the JS platform as js_fragile(24259), so we don't hold up merge requests while fixing the underlying issues. See #24259 - - - - - 1c79526a by Finley McIlwaine at 2023-12-15T12:24:40-08:00 Late plugins - - - - - 000c3302 by Finley McIlwaine at 2023-12-15T12:24:40-08:00 withTiming on LateCCs and late plugins - - - - - be4551ac by Finley McIlwaine at 2023-12-15T12:24:40-08:00 add test for late plugins - - - - - 7c29da9f by Finley McIlwaine at 2023-12-15T12:24:40-08:00 Document late plugins - - - - - 9a52ae46 by Ben Gamari at 2023-12-20T07:07:26-05:00 Fix thunk update ordering Previously we attempted to ensure soundness of concurrent thunk update by synchronizing on the access of the thunk's info table pointer field. This was believed to be sufficient since the indirectee (which may expose a closure allocated by another core) would not be examined until the info table pointer update is complete. However, it turns out that this can result in data races in the presence of multiple threads racing a update a single thunk. For instance, consider this interleaving under the old scheme: Thread A Thread B --------- --------- t=0 Enter t 1 Push update frame 2 Begin evaluation 4 Pause thread 5 t.indirectee=tso 6 Release t.info=BLACKHOLE 7 ... (e.g. GC) 8 Resume thread 9 Finish evaluation 10 Relaxed t.indirectee=x 11 Load t.info 12 Acquire fence 13 Inspect t.indirectee 14 Release t.info=BLACKHOLE Here Thread A enters thunk `t` but is soon paused, resulting in `t` being lazily blackholed at t=6. Then, at t=10 Thread A finishes evaluation and updates `t.indirectee` with a relaxed store. Meanwhile, Thread B enters the blackhole. Under the old scheme this would introduce an acquire-fence but this would only synchronize with Thread A at t=6. Consequently, the result of the evaluation, `x`, is not visible to Thread B, introducing a data race. We fix this by treating the `indirectee` field as we do all other mutable fields. This means we must always access this field with acquire-loads and release-stores. See #23185. - - - - - f4b53538 by Vladislav Zavialov at 2023-12-20T07:08:02-05:00 docs: Fix link to 051-ghc-base-libraries.rst The proposal is no longer available at the previous URL. - - - - - f7e21fab by Matthew Pickering at 2023-12-21T14:57:40+00:00 hadrian: Build all executables in bin/ folder In the end the bindist creation logic copies them all into the bin folder. There is no benefit to building a specific few binaries in the lib/bin folder anymore. This also removes the ad-hoc logic to copy the touchy and unlit executables from stage0 into stage1. It takes <1s to build so we might as well just build it. - - - - - 0038d052 by Zubin Duggal at 2023-12-22T23:28:00-05:00 testsuite: mark jspace as fragile on i386. This test has been flaky for some time and has been failing consistently on i386-linux since 8e0446df landed. See #24261 - - - - - dfd670a0 by Ben Bellick at 2023-12-24T10:10:31-05:00 Deprecate -ddump-json and introduce -fdiagnostics-as-json Addresses #19278 This commit deprecates the underspecified -ddump-json flag and introduces a newer, well-specified flag -fdiagnostics-as-json. Also included is a JSON schema as part of the documentation. The -ddump-json flag will be slated for removal shortly after this merge. - - - - - 609e6225 by Ben Bellick at 2023-12-24T10:10:31-05:00 Deprecate -ddump-json and introduce -fdiagnostics-as-json Addresses #19278 This commit deprecates the underspecified -ddump-json flag and introduces a newer, well-specified flag -fdiagnostics-as-json. Also included is a JSON schema as part of the documentation. The -ddump-json flag will be slated for removal shortly after this merge. - - - - - 865513b2 by Ömer Sinan Ağacan at 2023-12-24T10:11:13-05:00 Fix BNF in user manual 6.6.8.2: formal syntax for instance declarations - - - - - c247b6be by Zubin Duggal at 2023-12-25T16:01:23-05:00 docs: document permissibility of -XOverloadedLabels (#24249) Document the permissibility introduced by https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0170-unrestricted-overloadedlabels.rst - - - - - e5b7eb59 by Ömer Sinan Ağacan at 2023-12-25T16:02:03-05:00 Fix a code block syntax in user manual sec. 6.8.8.6 - - - - - 2db11c08 by Ben Gamari at 2023-12-29T15:35:48-05:00 genSym: Reimplement via CAS on 32-bit platforms Previously the remaining use of the C implementation on 32-bit platforms resulted in a subtle bug, #24261. This was due to the C object (which used the RTS's `atomic_inc64` macro) being compiled without `-threaded` yet later being used in a threaded compiler. Side-step this issue by using the pure Haskell `genSym` implementation on all platforms. This required implementing `fetchAddWord64Addr#` in terms of CAS on 64-bit platforms. - - - - - 19328a8c by Xiaoyan Ren at 2023-12-29T15:36:30-05:00 Do not color the diagnostic code in error messages (#24172) - - - - - 685b467c by Krzysztof Gogolewski at 2023-12-29T15:37:06-05:00 Enforce that bindings of implicit parameters are lifted Fixes #24298 - - - - - bc4d67b7 by Matthew Craven at 2023-12-31T06:15:42-05:00 StgToCmm: Detect some no-op case-continuations ...and generate no code for them. Fixes #24264. - - - - - 5b603139 by Krzysztof Gogolewski at 2023-12-31T06:16:18-05:00 Revert "testsuite: mark jspace as fragile on i386." This reverts commit 0038d052c8c80b4b430bb2aa1c66d5280be1aa95. The atomicity bug should be fixed by !11802. - - - - - d55216ad by Krzysztof Gogolewski at 2024-01-01T12:05:49-05:00 Refactor: store [[PrimRep]] rather than [Type] in STG StgConApp stored a list of types. This list was used exclusively during unarisation of unboxed sums (mkUbxSum). However, this is at a wrong level of abstraction: STG shouldn't be concerned with Haskell types, only PrimReps. Update the code to store a [[PrimRep]]. Also, there's no point in storing this list when we're not dealing with an unboxed sum. - - - - - 8b340bc7 by Ömer Sinan Ağacan at 2024-01-01T12:06:29-05:00 Kind signatures docs: mention that they're allowed in newtypes - - - - - 989bf8e5 by Zubin Duggal at 2024-01-03T20:08:47-05:00 ci: Ensure we use the correct bindist name for the test artifact when generating release ghcup metadata Fixes #24268 - - - - - 89299a89 by Krzysztof Gogolewski at 2024-01-03T20:09:23-05:00 Refactor: remove calls to typePrimRepArgs The function typePrimRepArgs is just a thin wrapper around typePrimRep, adding a VoidRep if the list is empty. However, in StgToByteCode, we were discarding that VoidRep anyway, so there's no point in calling it. - - - - - c7be0c68 by mmzk1526 at 2024-01-03T20:10:07-05:00 Use "-V" for alex version check for better backward compatibility Fixes #24302. In recent versions of alex, "-v" is used for "--verbose" instead of "-version". - - - - - 67dbcc0a by Krzysztof Gogolewski at 2024-01-05T02:07:18-05:00 Fix VoidRep handling in ghci debugger 'go' inside extractSubTerms was giving a bad result given a VoidRep, attempting to round towards the next multiple of 0. I don't understand much about the debugger but the code should be better than it was. Fixes #24306 - - - - - 90ea574e by Krzysztof Gogolewski at 2024-01-05T02:07:54-05:00 VoidRep-related refactor * In GHC.StgToByteCode, replace bcIdPrimId with idPrimRep, bcIdArgRep with idArgRep, atomPrimRep with stgArgRep1. All of them were duplicates. * In GHC.Stg.Unarise, we were converting a PrimRep to a Type and back to PrimRep. Remove the calls to primRepToType and typePrimRep1 which cancel out. * In GHC.STG.Lint, GHC.StgToCmm, GHC.Types.RepType we were filtering out VoidRep from the result of typePrimRep. But typePrimRep never returns VoidRep - remove the filtering. - - - - - eaf72479 by brian at 2024-01-06T23:03:09-05:00 Add unaligned Addr# primops Implements CLC proposal #154: https://github.com/haskell/core-libraries-committee/issues/154 * add unaligned addr primops * add tests * accept tests * add documentation * fix js primops * uncomment in access ops * use Word64 in tests * apply suggestions * remove extra file * move docs * remove random options * use setByteArray# primop * better naming * update base-exports test * add base-exports for other architectures - - - - - d471d445 by Krzysztof Gogolewski at 2024-01-06T23:03:47-05:00 Remove VoidRep from PrimRep, introduce PrimOrVoidRep This introduces data PrimOrVoidRep = VoidRep | NVRep PrimRep changes typePrimRep1 to return PrimOrVoidRep, and adds a new function typePrimRepU to be used when the argument is definitely non-void. Details in Note [VoidRep] in GHC.Types.RepType. Fixes #19520 - - - - - 48720a07 by Matthew Craven at 2024-01-08T18:57:36-05:00 Apply Note [Sensitivity to unique increment] to LargeRecord - - - - - 9e2e180f by Sebastian Graf at 2024-01-08T18:58:13-05:00 Debugging: Add diffUFM for convenient diffing between UniqFMs - - - - - 948f3e35 by Sebastian Graf at 2024-01-08T18:58:13-05:00 Rename Opt_D_dump_stranal to Opt_D_dump_dmdanal ... and Opt_D_dump_str_signatures to Opt_D_dump_dmd_signatures - - - - - 4e217e3e by Sebastian Graf at 2024-01-08T18:58:13-05:00 Deprecate -ddump-stranal and -ddump-str-signatures ... and suggest -ddump-dmdanal and -ddump-dmd-signatures instead - - - - - 6c613c90 by Sebastian Graf at 2024-01-08T18:58:13-05:00 Move testsuite/tests/stranal to testsuite/tests/dmdanal A separate commit so that the rename is obvious to Git(Lab) - - - - - c929f02b by Sebastian Graf at 2024-01-08T18:58:13-05:00 CoreSubst: Stricten `substBndr` and `cloneBndr` Doing so reduced allocations of `cloneBndr` by about 25%. ``` T9233(normal) ghc/alloc 672,488,656 663,083,216 -1.4% GOOD T9675(optasm) ghc/alloc 423,029,256 415,812,200 -1.7% geo. mean -0.1% minimum -1.7% maximum +0.1% ``` Metric Decrease: T9233 - - - - - e3ca78f3 by Krzysztof Gogolewski at 2024-01-10T17:35:59-05:00 Deprecate -Wsemigroup This warning was used to prepare for Semigroup becoming a superclass of Monoid, and for (<>) being exported from Prelude. This happened in GHC 8.4 in 8ae263ceb3566 and feac0a3bc69fd3. The leftover logic for (<>) has been removed in GHC 9.8, 4d29ecdfcc79. Now the warning does nothing at all and can be deprecated. - - - - - 08d14925 by amesgen at 2024-01-10T17:36:42-05:00 WASM metadata: use correct GHC version - - - - - 7a808419 by Xiaoyan Ren at 2024-01-10T17:37:24-05:00 Allow SCC declarations in TH (#24081) - - - - - 28827c51 by Xiaoyan Ren at 2024-01-10T17:37:24-05:00 Fix prettyprinting of SCC pragmas - - - - - ae9cc1a8 by Matthew Craven at 2024-01-10T17:38:01-05:00 Fix loopification in the presence of void arguments This also removes Note [Void arguments in self-recursive tail calls], which was just misleading. It's important to count void args both in the function's arity and at the call site. Fixes #24295. - - - - - b718b145 by Zubin Duggal at 2024-01-10T17:38:36-05:00 testsuite: Teach testsuite driver about c++ sources - - - - - 09cb57ad by Zubin Duggal at 2024-01-10T17:38:36-05:00 driver: Set -DPROFILING when compiling C++ sources with profiling Earlier, we used to pass all preprocessor flags to the c++ compiler. This meant that -DPROFILING was passed to the c++ compiler because it was a part of C++ flags However, this was incorrect and the behaviour was changed in 8ff3134ed4aa323b0199ad683f72165e51a59ab6. See #21291. But that commit exposed this bug where -DPROFILING was no longer being passed when compiling c++ sources. The fix is to explicitly include -DPROFILING in `opt_cxx` when profiling is enabled to ensure we pass the correct options for the way to both C and C++ compilers Fixes #24286 - - - - - 2cf9dd96 by Zubin Duggal at 2024-01-10T17:38:36-05:00 testsuite: rename objcpp -> objcxx To avoid confusion with C Pre Processsor - - - - - af6932d6 by Simon Peyton Jones at 2024-01-10T17:39:12-05:00 Make TYPE and CONSTRAINT not-apart Issue #24279 showed up a bug in the logic in GHC.Core.Unify.unify_ty which is supposed to make TYPE and CONSTRAINT be not-apart. Easily fixed. - - - - - 4a39b5ff by Zubin Duggal at 2024-01-10T17:39:48-05:00 ci: Fix typo in mk_ghcup_metadata.py There was a missing colon in the fix to #24268 in 989bf8e53c08eb22de716901b914b3607bc8dd08 - - - - - 13503451 by Zubin Duggal at 2024-01-10T17:40:24-05:00 release-ci: remove release-x86_64-linux-deb11-release+boot_nonmoving_gc job There is no reason to have this release build or distribute this variation. This configuration is for testing purposes only. - - - - - afca46a4 by Sebastian Graf at 2024-01-10T17:41:00-05:00 Parser: Add a Note detailing why we need happy's `error` to implement layout - - - - - eaf8a06d by Krzysztof Gogolewski at 2024-01-11T00:43:17+01:00 Turn -Wtype-equality-out-of-scope on by default Also remove -Wnoncanonical-{monoid,monad}-instances from -Wcompat, since they are enabled by default. Refresh wcompat-warnings/ test with new -Wcompat warnings. Part of #24267 Co-authored-by: sheaf <sam.derbyshire at gmail.com> - - - - - 42bee5aa by Sebastian Graf at 2024-01-12T21:16:21-05:00 Arity: Require called *exactly once* for eta exp with -fpedantic-bottoms (#24296) In #24296, we had a program in which we eta expanded away an error despite the presence of `-fpedantic-bottoms`. This was caused by turning called *at least once* lambdas into one-shot lambdas, while with `-fpedantic-bottoms` it is only sound to eta expand over lambdas that are called *exactly* once. An example can be found in `Note [Combining arity type with demand info]`. Fixes #24296. - - - - - 7e95f738 by Andreas Klebinger at 2024-01-12T21:16:57-05:00 Aarch64: Enable -mfma by default. Fixes #24311 - - - - - e43788d0 by Jason Shipman at 2024-01-14T12:47:38-05:00 Add more instances for Compose: Fractional, RealFrac, Floating, RealFloat CLC proposal #226 https://github.com/haskell/core-libraries-committee/issues/226 - - - - - ae6d8cd2 by Sebastian Graf at 2024-01-14T12:48:15-05:00 Pmc: COMPLETE pragmas associated with Family TyCons should apply to representation TyCons as well (#24326) Fixes #24326. - - - - - c5fc7304 by sheaf at 2024-01-15T14:15:29-05:00 Use lookupOccRn_maybe in TH.lookupName When looking up a value, we want to be able to find both variables and record fields. So we should not use the lookupSameOccRn_maybe function, as we can't know ahead of time which record field namespace a record field with the given textual name will belong to. Fixes #24293 - - - - - da908790 by Krzysztof Gogolewski at 2024-01-15T14:16:05-05:00 Make the build more strict on documentation errors * Detect undefined labels. This can be tested by adding :ref:`nonexistent` to a documentation rst file; attempting to build docs will fail. Fixed the undefined label in `9.8.1-notes.rst`. * Detect errors. While we have plenty of warnings, we can at least enforce that Sphinx does not report errors. Fixed the error in `required_type_arguments.rst`. Unrelated change: I have documented that the `-dlint` enables `-fcatch-nonexhaustive-cases`, as can be verified by checking `enableDLint`. - - - - - 5077416e by Javier Sagredo at 2024-01-16T15:40:06-05:00 Profiling: Adds an option to not start time profiling at startup Using the functionality provided by d89deeba47ce04a5198a71fa4cbc203fe2c90794, this patch creates a new rts flag `--no-automatic-time-samples` which disables the time profiling when starting a program. It is then expected that the user starts it whenever it is needed. Fixes #24337 - - - - - 5776008c by Matthew Pickering at 2024-01-16T15:40:42-05:00 eventlog: Fix off-by-one error in postIPE We were missing the extra_comma from the calculation of the size of the payload of postIPE. This was causing assertion failures when the event would overflow the buffer by one byte, as ensureRoomForVariable event would report there was enough space for `n` bytes but then we would write `n + 1` bytes into the buffer. Fixes #24287 - - - - - 66dc09b1 by Simon Peyton Jones at 2024-01-16T15:41:18-05:00 Improve SpecConstr (esp nofib/spectral/ansi) This MR makes three improvements to SpecConstr: see #24282 * It fixes an outright (and recently-introduced) bug in `betterPat`, which was wrongly forgetting to compare the lengths of the argument lists. * It enhances ConVal to inclue a boolean for work-free-ness, so that the envt can contain non-work-free constructor applications, so that we can do more: see Note [ConVal work-free-ness] * It rejigs `subsumePats` so that it doesn't reverse the list. This can make a difference because, when patterns overlap, we arbitrarily pick the first. There is no "right" way, but this retains the old pre-subsumePats behaviour, thereby "fixing" the regression in #24282. Nofib results +======================================== | spectral/ansi -21.14% | spectral/hartel/comp_lab_zift -0.12% | spectral/hartel/parstof +0.09% | spectral/last-piece -2.32% | spectral/multiplier +6.03% | spectral/para +0.60% | spectral/simple -0.26% +======================================== | geom mean -0.18% +---------------------------------------- The regression in `multiplier` is sad, but it simply replicates GHC's previous behaviour (e.g. GHC 9.6). - - - - - 65da79b3 by Matthew Pickering at 2024-01-16T15:41:54-05:00 hadrian: Reduce Cabal verbosity The comment claims that `simpleUserHooks` decrease verbosity, and it does, but only for the `postConf` phase. The other phases are too verbose with `-V`. At the moment > 5000 lines of the build log are devoted to output from `cabal copy`. So I take the simple approach and just decrease the verbosity level again. If the output of `postConf` is essential then it would be better to implement our own `UserHooks` which doesn't decrease the verbosity for `postConf`. Fixes #24338 - - - - - 16414d7d by Matthew Pickering at 2024-01-17T10:54:59-05:00 Stop retaining old ModGuts throughout subsequent simplifier phases Each phase of the simplifier typically rewrites the majority of ModGuts, so we want to be able to release the old ModGuts as soon as possible. `name_ppr_ctxt` lives throught the whole optimiser phase and it was retaining a reference to `ModGuts`, so we were failing to release the old `ModGuts` until the end of the phase (potentially doubling peak memory usage for that particular phase). This was discovered using eras profiling (#24332) Fixes #24328 - - - - - 7f0879e1 by Matthew Pickering at 2024-01-17T10:55:35-05:00 Update nofib submodule - - - - - 320454d3 by Cheng Shao at 2024-01-17T23:02:40+00:00 ci: bump ci-images for updated wasm image - - - - - 2eca52b4 by Cheng Shao at 2024-01-17T23:06:44+00:00 base: treat all FDs as "nonblocking" on wasm On posix platforms, when performing read/write on FDs, we check the nonblocking flag first. For FDs without this flag (e.g. stdout), we call fdReady() first, which in turn calls poll() to wait for I/O to be available on that FD. This is problematic for wasm32-wasi: although select()/poll() is supported via the poll_oneoff() wasi syscall, that syscall is rather heavyweight and runtime behavior differs in different wasi implementations. The issue is even worse when targeting browsers, given there's no satisfactory way to implement async I/O as a synchronous syscall, so existing JS polyfills for wasi often give up and simply return ENOSYS. Before we have a proper I/O manager that avoids poll_oneoff() for async I/O on wasm, this patch improves the status quo a lot by merely pretending all FDs are "nonblocking". Read/write on FDs will directly invoke read()/write(), which are much more reliably handled in existing wasi implementations, especially those in browsers. Fixes #23275 and the following test cases: T7773 isEOF001 openFile009 T4808 cgrun025 Approved by CLC proposal #234: https://github.com/haskell/core-libraries-committee/issues/234 - - - - - 83c6c710 by Andrew Lelechenko at 2024-01-18T05:21:49-05:00 base: clarify how to disable warnings about partiality of Data.List.{head,tail} - - - - - c4078f2f by Simon Peyton Jones at 2024-01-18T05:22:25-05:00 Fix four bug in handling of (forall cv. body_ty) These bugs are all described in #24335 It's not easy to provoke the bug, hence no test case. - - - - - 119586ea by Alexis King at 2024-01-19T00:08:00-05:00 Always refresh profiling CCSes after running pending initializers Fixes #24171. - - - - - 9718d970 by Oleg Grenrus at 2024-01-19T00:08:36-05:00 Set default-language: GHC2021 in ghc library Go through compiler/ sources, and remove all BangPatterns (and other GHC2021 enabled extensions in these files). - - - - - 3ef71669 by Matthew Pickering at 2024-01-19T21:55:16-05:00 testsuite: Remove unused have_library function Also remove the hence unused testsuite option `--test-package-db`. Fixes #24342 - - - - - 5b7fa20c by Jade at 2024-01-19T21:55:53-05:00 Fix Spelling in the compiler Tracking: #16591 - - - - - 09875f48 by Matthew Pickering at 2024-01-20T12:20:44-05:00 testsuite: Implement `isInTreeCompiler` in a more robust way Just a small refactoring to avoid redundantly specifying the same strings in two different places. - - - - - 0d12b987 by Jade at 2024-01-20T12:21:20-05:00 Change maintainer email from cvs-ghc at haskell.org to ghc-devs at haskell.org. Fixes #22142 - - - - - eebdd316 by Apoorv Ingle at 2024-01-23T13:49:12+00:00 Changes for haskell/haddock#18324 - - - - - 1fa1c00c by Jade at 2024-01-23T19:17:03-05:00 Enhance Documentation of functions exported by Data.Function This patch aims to improve the documentation of functions exported in Data.Function Tracking: #17929 Fixes: #10065 - - - - - ab47a43d by Jade at 2024-01-23T19:17:39-05:00 Improve documentation of hGetLine. - Add explanation for whether a newline is returned - Add examples Fixes #14804 - - - - - dd4af0e5 by Cheng Shao at 2024-01-23T19:18:17-05:00 Fix genapply for cross-compilation by nuking fragile CPP logic This commit fixes incorrectly built genapply when cross compiling (#24347) by nuking all fragile CPP logic in it from the orbit. All target-specific info are now read from DerivedConstants.h at runtime, see added note for details. Also removes a legacy Makefile and adds haskell language server support for genapply. - - - - - 0cda2b8b by Cheng Shao at 2024-01-23T19:18:17-05:00 rts: enable wasm32 register mapping The wasm backend didn't properly make use of all Cmm global registers due to #24347. Now that it is fixed, this patch re-enables full register mapping for wasm32, and we can now generate smaller & faster wasm modules that doesn't always spill arguments onto the stack. Fixes #22460 #24152. - - - - - 0325a6e5 by Greg Steuck at 2024-01-24T01:29:44-05:00 Avoid utf8 in primops.txt.pp comments They don't make it through readFile' without explicitly setting the encoding. See https://gitlab.haskell.org/ghc/ghc/-/issues/17755 - - - - - 1aaf0bd8 by David Binder at 2024-01-24T01:30:20-05:00 Bump hpc and hpc-bin submodule Bump hpc to 0.7.0.1 Bump hpc-bin to commit d1780eb2 - - - - - e693a4e8 by Ben Gamari at 2024-01-24T01:30:56-05:00 testsuite: Ignore stderr in T8089 Otherwise spurious "Killed: 9" messages to stderr may cause the test to fail. Fixes #24361. - - - - - a40f4ab2 by sheaf at 2024-01-24T14:04:33-05:00 Fix FMA instruction on LLVM We were emitting the wrong instructions for fused multiply-add operations on LLVM: - the instruction name is "llvm.fma.f32" or "llvm.fma.f64", not "fmadd" - LLVM does not support other instructions such as "fmsub"; instead we implement these by flipping signs of some arguments - the instruction is an LLVM intrinsic, which requires handling it like a normal function call instead of a machine instruction Fixes #24223 - - - - - 69abc786 by Andrei Borzenkov at 2024-01-24T14:05:09-05:00 Add changelog entry for renaming tuples from (,,...,,) to Tuple<n> (24291) - - - - - 0ac8f385 by Cheng Shao at 2024-01-25T00:27:48-05:00 compiler: remove unused GHC.Linker module The GHC.Linker module is empty and unused, other than as a hack for the make build system. We can remove it now that make is long gone; the note is moved to GHC.Linker.Loader instead. - - - - - 699da01b by Hécate Moonlight at 2024-01-25T00:28:27-05:00 Clarification for newtype constructors when using `coerce` - - - - - b2d8cd85 by Matt Walker at 2024-01-26T09:50:08-05:00 Fix #24308 Add tests for semicolon separated where clauses - - - - - 0da490a1 by Ben Gamari at 2024-01-26T17:34:41-05:00 hsc2hs: Bump submodule - - - - - 3f442fd2 by Ben Gamari at 2024-01-26T17:34:41-05:00 Bump containers submodule to 0.7 - - - - - 82a1c656 by Sebastian Nagel at 2024-01-29T02:32:40-05:00 base: with{Binary}File{Blocking} only annotates own exceptions Fixes #20886 This ensures that inner, unrelated exceptions are not misleadingly annotated with the opened file. - - - - - 9294a086 by Andreas Klebinger at 2024-01-29T02:33:15-05:00 Fix fma warning when using llvm on aarch64. On aarch64 fma is always on so the +fma flag doesn't exist for that target. Hence no need to try and pass +fma to llvm. Fixes #24379 - - - - - ced2e731 by sheaf at 2024-01-29T17:27:12-05:00 No shadowing warnings for NoFieldSelector fields This commit ensures we don't emit shadowing warnings when a user shadows a field defined with NoFieldSelectors. Fixes #24381 - - - - - 8eeadfad by Patrick at 2024-01-29T17:27:51-05:00 Fix bug wrong span of nested_doc_comment #24378 close #24378 1. Update the start position of span in `nested_doc_comment` correctly. and hence the spans of identifiers of haddoc can be computed correctly. 2. add test `HaddockSpanIssueT24378`. - - - - - a557580f by Alexey Radkov at 2024-01-30T19:41:52-05:00 Fix irrelevant dodgy-foreign-imports warning on import f-pointers by value A test *сс018* is attached (not sure about the naming convention though). Note that without the fix, the test fails with the *dodgy-foreign-imports* warning passed to stderr. The warning disappears after the fix. GHC shouldn't warn on imports of natural function pointers from C by value (which is feasible with CApiFFI), such as ```haskell foreign import capi "cc018.h value f" f :: FunPtr (Int -> IO ()) ``` where ```c void (*f)(int); ``` See a related real-world use-case [here](https://gitlab.com/daniel-casanueva/pcre-light/-/merge_requests/17). There, GHC warns on import of C function pointer `pcre_free`. - - - - - ca99efaf by Alexey Radkov at 2024-01-30T19:41:53-05:00 Rename test cc018 -> T24034 - - - - - 88c38dd5 by Ben Gamari at 2024-01-30T19:42:28-05:00 rts/TraverseHeap.c: Ensure that PosixSource.h is included first - - - - - ca2e919e by Simon Peyton Jones at 2024-01-31T09:29:45+00:00 Make decomposeRuleLhs a bit more clever This fixes #24370 by making decomposeRuleLhs undertand dictionary /functions/ as well as plain /dictionaries/ - - - - - 94ce031d by Teo Camarasu at 2024-02-01T05:49:49-05:00 doc: Add -Dn flag to user guide Resolves #24394 - - - - - 31553b11 by Ben Gamari at 2024-02-01T12:21:29-05:00 cmm: Introduce MO_RelaxedRead In hand-written Cmm it can sometimes be necessary to atomically load from memory deep within an expression (e.g. see the `CHECK_GC` macro). This MachOp provides a convenient way to do so without breaking the expression into multiple statements. - - - - - 0785cf81 by Ben Gamari at 2024-02-01T12:21:29-05:00 codeGen: Use relaxed accesses in ticky bumping - - - - - be423dda by Ben Gamari at 2024-02-01T12:21:29-05:00 base: use atomic write when updating timer manager - - - - - 8a310e35 by Ben Gamari at 2024-02-01T12:21:29-05:00 Use relaxed atomics to manipulate TSO status fields - - - - - d6809ee4 by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Add necessary barriers when manipulating TSO owner - - - - - 39e3ac5d by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Use `switch` to branch on why_blocked This is a semantics-preserving refactoring. - - - - - 515eb33d by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Fix synchronization on thread blocking state We now use a release barrier whenever we update a thread's blocking state. This required widening StgTSO.why_blocked as AArch64 does not support atomic writes on 16-bit values. - - - - - eb38812e by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Fix data race in threadPaused This only affects an assertion in the debug RTS and only needs relaxed ordering. - - - - - 26c48dd6 by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Fix data race in threadStatus# - - - - - 6af43ab4 by Ben Gamari at 2024-02-01T12:21:29-05:00 rts: Fix data race in Interpreter's preemption check - - - - - 9502ad3c by Ben Gamari at 2024-02-01T12:21:29-05:00 rts/Messages: Fix data race - - - - - 60802db5 by Ben Gamari at 2024-02-01T12:21:30-05:00 rts/Prof: Fix data race - - - - - ef8ccef5 by Ben Gamari at 2024-02-01T12:21:30-05:00 rts: Use relaxed ordering on dirty/clean info tables updates When changing the dirty/clean state of a mutable object we needn't have any particular ordering. - - - - - 76fe2b75 by Ben Gamari at 2024-02-01T12:21:30-05:00 codeGen: Use relaxed-read in closureInfoPtr - - - - - a6316eb4 by Ben Gamari at 2024-02-01T12:21:30-05:00 STM: Use acquire loads when possible Full sequential consistency is not needed here. - - - - - 6bddfd3d by Ben Gamari at 2024-02-01T12:21:30-05:00 rts: Use fence rather than redundant load Previously we would use an atomic load to ensure acquire ordering. However, we now have `ACQUIRE_FENCE_ON`, which allows us to express this more directly. - - - - - 55c65dbc by Ben Gamari at 2024-02-01T12:21:30-05:00 rts: Fix data races in profiling timer - - - - - 856b5e75 by Ben Gamari at 2024-02-01T12:21:30-05:00 Add Note [C11 memory model] - - - - - 6534da24 by Cheng Shao at 2024-02-01T12:22:07-05:00 compiler: move generic cmm optimization logic in NCG to a standalone module This commit moves GHC.CmmToAsm.cmmToCmm to a standalone module, GHC.Cmm.GenericOpt. The main motivation is enabling this logic to be run in the wasm backend NCG code, which is defined in other modules that's imported by GHC.CmmToAsm, causing a cyclic dependency issue. - - - - - 87e34888 by Cheng Shao at 2024-02-01T12:22:07-05:00 compiler: explicitly disable PIC in wasm32 NCG This commit explicitly disables the ncgPIC flag for the wasm32 target. The wasm backend doesn't support PIC for the time being. - - - - - c6ce242e by Cheng Shao at 2024-02-01T12:22:07-05:00 compiler: enable generic cmm optimizations in wasm backend NCG This commit enables the generic cmm optimizations in other NCGs to be run in the wasm backend as well, followed by a late cmm control-flow optimization pass. The added optimizations do catch some corner cases not handled by the pre-NCG cmm pipeline and are useful in generating smaller CFGs. - - - - - 151dda4e by Andrei Borzenkov at 2024-02-01T12:22:43-05:00 Namespacing for WARNING/DEPRECATED pragmas (#24396) New syntax for WARNING and DEPRECATED pragmas was added, namely namespace specifierss: namespace_spec ::= 'type' | 'data' | {- empty -} warning ::= warning_category namespace_spec namelist strings deprecation ::= namespace_spec namelist strings A new data type was introduced to represent these namespace specifiers: data NamespaceSpecifier = NoSpecifier | TypeNamespaceSpecifier (EpToken "type") | DataNamespaceSpecifier (EpToken "data") Extension field XWarning now contains this NamespaceSpecifier. lookupBindGroupOcc function was changed: it now takes NamespaceSpecifier and checks that the namespace of the found names matches the passed flag. With this change {-# WARNING data D "..." #-} pragma will only affect value namespace and {-# WARNING type D "..." #-} will only affect type namespace. The same logic is applicable to DEPRECATED pragmas. Finding duplicated warnings inside rnSrcWarnDecls now takes into consideration NamespaceSpecifier flag to allow warnings with the same names that refer to different namespaces. - - - - - 38c3afb6 by Bryan Richter at 2024-02-01T12:23:19-05:00 CI: Disable the test-cabal-reinstall job Fixes #24363 - - - - - 27020458 by Matthew Craven at 2024-02-03T01:53:26-05:00 Bump bytestring submodule to something closer to 0.12.1 ...mostly so that 16d6b7e835ffdcf9b894e79f933dd52348dedd0c (which reworks unaligned writes in Builder) and the stuff in https://github.com/haskell/bytestring/pull/631 can see wider testing. The less-terrible code for unaligned writes used in Builder on hosts not known to be ulaigned-friendly also takes less effort for GHC to compile, resulting in a metric decrease for T21839c on some platforms. The metric increase on T21839r is caused by the unrelated commit 750dac33465e7b59100698a330b44de7049a345c. It perhaps warrants further analysis and discussion (see #23822) but is not critical. Metric Decrease: T21839c Metric Increase: T21839r - - - - - cdddeb0f by Rodrigo Mesquita at 2024-02-03T01:54:02-05:00 Work around autotools setting C11 standard in CC/CXX In autoconf >=2.70, C11 is set by default for $CC and $CXX via the -std=...11 flag. In this patch, we split the "-std" flag out of the $CC and $CXX variables, which we traditionally assume to be just the executable name/path, and move it to $CFLAGS/$CXXFLAGS instead. Fixes #24324 - - - - - 5ff7cc26 by Apoorv Ingle at 2024-02-03T13:14:46-06:00 Expand `do` blocks right before typechecking using the `HsExpansion` philosophy. - Fixes #18324 #20020 #23147 #22788 #15598 #22086 #21206 - The change is detailed in - Note [Expanding HsDo with HsExpansion] in `GHC.Tc.Gen.Do` - Note [Doing HsExpansion in the Renamer vs Typechecker] in `GHC.Rename.Expr` expains the rational of doing expansions in type checker as opposed to in the renamer - Adds new datatypes: - `GHC.Hs.Expr.XXExprGhcRn`: new datatype makes this expansion work easier 1. Expansion bits for Expressions, Statements and Patterns in (`ExpandedThingRn`) 2. `PopErrCtxt` a special GhcRn Phase only artifcat to pop the previous error message in the error context stack - `GHC.Basic.Origin` now tracks the reason for expansion in case of Generated This is useful for type checking cf. `GHC.Tc.Gen.Expr.tcExpr` case for `HsLam` - Kills `HsExpansion` and `HsExpanded` as we have inlined them in `XXExprGhcRn` and `XXExprGhcTc` - Ensures warnings such as 1. Pattern match checks 2. Failable patterns 3. non-() return in body statements are preserved - Kill `HsMatchCtxt` in favor of `TcMatchAltChecker` - Testcases: * T18324 T20020 T23147 T22788 T15598 T22086 * T23147b (error message check), * DoubleMatch (match inside a match for pmc check) * pattern-fails (check pattern match with non-refutable pattern, eg. newtype) * Simple-rec (rec statements inside do statment) * T22788 (code snippet from #22788) * DoExpanion1 (Error messages for body statments) * DoExpansion2 (Error messages for bind statements) * DoExpansion3 (Error messages for let statements) Also repoint haddock to the right submodule so that the test (haddockHypsrcTest) pass Metric Increase 'compile_time/bytes allocated': T9020 The testcase is a pathalogical example of a `do`-block with many statements that do nothing. Given that we are expanding the statements into function binds, we will have to bear a (small) 2% cost upfront in the compiler to unroll the statements. - - - - - 0df8ce27 by Vladislav Zavialov at 2024-02-04T03:55:14-05:00 Reduce parser allocations in allocateCommentsP In the most common case, the comment queue is empty, so we can skip the work of processing it. This reduces allocations by about 10% in the parsing001 test. Metric Decrease: MultiLayerModulesRecomp parsing001 - - - - - ae856a82 by Matthew Pickering at 2024-02-05T12:22:39+00:00 ghc-internals fallout - - - - - cfd68290 by Simon Peyton Jones at 2024-02-05T17:58:33-05:00 Stop dropping a case whose binder is demanded This MR fixes #24251. See Note [Case-to-let for strictly-used binders] in GHC.Core.Opt.Simplify.Iteration, plus #24251, for lots of discussion. Final Nofib changes over 0.1%: +----------------------------------------- | imaginary/digits-of-e2 -2.16% | imaginary/rfib -0.15% | real/fluid -0.10% | real/gamteb -1.47% | real/gg -0.20% | real/maillist +0.19% | real/pic -0.23% | real/scs -0.43% | shootout/n-body -0.41% | shootout/spectral-norm -0.12% +======================================== | geom mean -0.05% Pleasingly, overall executable size is down by just over 1%. Compile times (in perf/compiler) wobble around a bit +/- 0.5%, but the geometric mean is -0.1% which seems good. - - - - - e4d137bb by Simon Peyton Jones at 2024-02-05T17:58:33-05:00 Add Note [Bangs in Integer functions] ...to document the bangs in the functions in GHC.Num.Integer - - - - - ce90f12f by Andrei Borzenkov at 2024-02-05T17:59:09-05:00 Hide WARNING/DEPRECATED namespacing under -XExplicitNamespaces (#24396) - - - - - e2ea933f by Simon Peyton Jones at 2024-02-06T10:12:04-05:00 Refactoring in preparation for lazy skolemisation * Make HsMatchContext and HsStmtContext be parameterised over the function name itself, rather than over the pass. See [mc_fun field of FunRhs] in Language.Haskell.Syntax.Expr - Replace types HsMatchContext GhcPs --> HsMatchContextPs HsMatchContext GhcRn --> HsMatchContextRn HsMatchContext GhcTc --> HsMatchContextRn (sic! not Tc) HsStmtContext GhcRn --> HsStmtContextRn - Kill off convertHsMatchCtxt * Split GHC.Tc.Type.BasicTypes.TcSigInfo so that TcCompleteSig (describing a complete user-supplied signature) is its own data type. - Split TcIdSigInfo(CompleteSig, PartialSig) into TcCompleteSig(CSig) TcPartialSig(PSig) - Use TcCompleteSig in tcPolyCheck, CheckGen - Rename types and data constructors: TcIdSigInfo --> TcIdSig TcPatSynInfo(TPSI) --> TcPatSynSig(PatSig) - Shuffle around helper functions: tcSigInfoName (moved to GHC.Tc.Types.BasicTypes) completeSigPolyId_maybe (moved to GHC.Tc.Types.BasicTypes) tcIdSigName (inlined and removed) tcIdSigLoc (introduced) - Rearrange the pattern match in chooseInferredQuantifiers * Rename functions and types: tcMatchesCase --> tcCaseMatches tcMatchesFun --> tcFunBindMatches tcMatchLambda --> tcLambdaMatches tcPats --> tcMatchPats matchActualFunTysRho --> matchActualFunTys matchActualFunTySigma --> matchActualFunTy * Add HasDebugCallStack constraints to: mkBigCoreVarTupTy, mkBigCoreTupTy, boxTy, mkPiTy, mkPiTys, splitAppTys, splitTyConAppNoView_maybe * Use `penv` from the outer context in the inner loop of GHC.Tc.Gen.Pat.tcMultiple * Move tcMkVisFunTy, tcMkInvisFunTy, tcMkScaledFunTys down the file, factor out and export tcMkScaledFunTy. * Move isPatSigCtxt down the file. * Formatting and comments Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - f5d3e03c by Andrei Borzenkov at 2024-02-06T10:12:04-05:00 Lazy skolemisation for @a-binders (#17594) This patch is a preparation for @a-binders implementation. The main changes are: * Skolemisation is now prepared to deal with @binders. See Note [Skolemisation overview] in GHC.Tc.Utils.Unify. Most of the action is in - Utils.Unify.matchExpectedFunTys - Gen.Pat.tcMatchPats - Gen.Expr.tcPolyExprCheck - Gen.Binds.tcPolyCheck Some accompanying refactoring: * I found that funTyConAppTy_maybe was doing a lot of allocation, and rejigged userTypeError_maybe to avoid calling it. - - - - - 532993c8 by Zubin Duggal at 2024-02-06T10:12:41-05:00 driver: Really don't lose track of nodes when we fail to resolve cycles This fixes a bug in 8db8d2fd1c881032b1b360c032b6d9d072c11723, where we could lose track of acyclic components at the start of an unresolved cycle. We now ensure we never loose track of any of these components. As T24275 demonstrates, a "cyclic" SCC might not really be a true SCC: When viewed without boot files, we have a single SCC ``` [REC main:T24275B [main:T24275B {-# SOURCE #-}, main:T24275A {-# SOURCE #-}] main:T24275A [main:T24275A {-# SOURCE #-}]] ``` But with boot files this turns into ``` [NONREC main:T24275B {-# SOURCE #-} [], REC main:T24275B [main:T24275B {-# SOURCE #-}, main:T24275A {-# SOURCE #-}] main:T24275A {-# SOURCE #-} [main:T24275B], NONREC main:T24275A [main:T24275A {-# SOURCE #-}]] ``` Note that this is truly not an SCC, as no nodes are reachable from T24275B.hs-boot. However, we treat this entire group as a single "SCC" because it seems so when we analyse the graph without taking boot files into account. Indeed, we must return a single ResolvedCycle element in the BuildPlan for this as described in Note [Upsweep]. However, since after resolving this is not a true SCC anymore, `findCycle` fails to find a cycle and we have a sub-optimal error message as a result. To handle this, I extended `findCycle` to not assume its input is an SCC, and to try harder to find cycles in its input. Fixes #24275 - - - - - b35dd613 by Zubin Duggal at 2024-02-06T10:13:17-05:00 GHCi: Lookup breakpoint CCs in the correct module We need to look up breakpoint CCs in the module that the breakpoint points to, and not the current module. Fixes #24327 - - - - - b09e6958 by Zubin Duggal at 2024-02-06T10:13:17-05:00 testsuite: Add test for #24327 - - - - - 569b4c10 by doyougnu at 2024-02-07T03:06:26-05:00 ts: add compile_artifact, ignore_extension flag In b521354216f2821e00d75f088d74081d8b236810 the testsuite gained the capability to collect generic metrics. But this assumed that the test was not linking and producing artifacts and we only wanted to track object files, interface files, or build artifacts from the compiler build. However, some backends, such as the JS backend, produce artifacts when compiling, such as the jsexe directory which we want to track. This patch: - tweaks the testsuite to collect generic metrics on any build artifact in the test directory. - expands the exe_extension function to consider windows and adds the ignore_extension flag. - Modifies certain tests to add the ignore_extension flag. Tests such as heaprof002 expect a .ps file, but on windows without ignore_extensions the testsuite will look for foo.exe.ps. Hence the flag. - adds the size_hello_artifact test - - - - - 75a31379 by doyougnu at 2024-02-07T03:06:26-05:00 ts: add wasm_arch, heapprof002 wasm extension - - - - - c9731d6d by Rodrigo Mesquita at 2024-02-07T03:07:03-05:00 Synchronize bindist configure for #24324 In cdddeb0f1280b40cc194028bbaef36e127175c4c, we set up a workaround for #24324 in the in-tree configure script, but forgot to update the bindist configure script accordingly. This updates it. - - - - - d309f4e7 by Matthew Pickering at 2024-02-07T03:07:38-05:00 distrib/configure: Fix typo in CONF_GCC_LINKER_OPTS_STAGE2 variable Instead we were setting CONF_GCC_LINK_OPTS_STAGE2 which meant that we were missing passing `--target` when invoking the linker. Fixes #24414 - - - - - 77db84ab by Ben Gamari at 2024-02-08T00:35:22-05:00 llvmGen: Adapt to allow use of new pass manager. We now must use `-passes` in place of `-O<n>` due to #21936. Closes #21936. - - - - - 3c9ddf97 by Matthew Pickering at 2024-02-08T00:35:59-05:00 testsuite: Mark length001 as fragile on javascript Modifying the timeout multiplier is not a robust way to get this test to reliably fail. Therefore we mark it as fragile until/if javascript ever supports the stack limit. - - - - - 20b702b5 by Matthew Pickering at 2024-02-08T00:35:59-05:00 Javascript: Don't filter out rtsDeps list This logic appears to be incorrect as it would drop any dependency which was not in a direct dependency of the package being linked. In the ghc-internals split this started to cause errors because `ghc-internal` is not a direct dependency of most packages, and hence important symbols to keep which are hard coded into the js runtime were getting dropped. - - - - - 2df96366 by Ben Gamari at 2024-02-08T00:35:59-05:00 base: Cleanup whitespace in cbits - - - - - 44f6557a by Ben Gamari at 2024-02-08T00:35:59-05:00 Move `base` to `ghc-internal` Here we move a good deal of the implementation of `base` into a new package, `ghc-internal` such that it can be evolved independently from the user-visible interfaces of `base`. While we want to isolate implementation from interfaces, naturally, we would like to avoid turning `base` into a mere set of module re-exports. However, this is a non-trivial undertaking for a variety of reasons: * `base` contains numerous known-key and wired-in things, requiring corresponding changes in the compiler * `base` contains a significant amount of C code and corresponding autoconf logic, which is very fragile and difficult to break apart * `base` has numerous import cycles, which are currently dealt with via carefully balanced `hs-boot` files * We must not break existing users To accomplish this migration, I tried the following approaches: * [Split-GHC.Base]: Break apart the GHC.Base knot to allow incremental migration of modules into ghc-internal: this knot is simply too intertwined to be easily pulled apart, especially given the rather tricky import cycles that it contains) * [Move-Core]: Moving the "core" connected component of base (roughly 150 modules) into ghc-internal. While the Haskell side of this seems tractable, the C dependencies are very subtle to break apart. * [Move-Incrementally]: 1. Move all of base into ghc-internal 2. Examine the module structure and begin moving obvious modules (e.g. leaves of the import graph) back into base 3. Examine the modules remaining in ghc-internal, refactor as necessary to facilitate further moves 4. Go to (2) iterate until the cost/benefit of further moves is insufficient to justify continuing 5. Rename the modules moved into ghc-internal to ensure that they don't overlap with those in base 6. For each module moved into ghc-internal, add a shim module to base with the declarations which should be exposed and any requisite Haddocks (thus guaranteeing that base will be insulated from changes in the export lists of modules in ghc-internal Here I am using the [Move-Incrementally] approach, which is empirically the least painful of the unpleasant options above Bumps haddock submodule. Metric Decrease: haddock.Cabal haddock.base Metric Increase: MultiComponentModulesRecomp T16875 size_hello_artifact - - - - - e8fb2451 by Vladislav Zavialov at 2024-02-08T00:36:36-05:00 Haddock comments on infix constructors (#24221) Rewrite the `HasHaddock` instance for `ConDecl GhcPs` to account for infix constructors. This change fixes a Haddock regression (introduced in 19e80b9af252) that affected leading comments on infix data constructor declarations: -- | Docs for infix constructor | Int :* Bool The comment should be associated with the data constructor (:*), not with its left-hand side Int. - - - - - 9060d55b by Ben Gamari at 2024-02-08T00:37:13-05:00 Add os-string as a boot package Introduces `os-string` submodule. This will be necessary for `filepath-1.5`. - - - - - 9d65235a by Ben Gamari at 2024-02-08T00:37:13-05:00 gitignore: Ignore .hadrian_ghci_multi/ - - - - - d7ee12ea by Ben Gamari at 2024-02-08T00:37:13-05:00 hadrian: Set -this-package-name When constructing the GHC flags for a package Hadrian must take care to set `-this-package-name` in addition to `-this-unit-id`. This hasn't broken until now as we have not had any uses of qualified package imports. However, this will change with `filepath-1.5` and the corresponding `unix` bump, breaking `hadrian/multi-ghci`. - - - - - f2dffd2e by Ben Gamari at 2024-02-08T00:37:13-05:00 Bump filepath to 1.5.0.0 Required bumps of the following submodules: * `directory` * `filepath` * `haskeline` * `process` * `unix` * `hsc2hs` * `Win32` * `semaphore-compat` and the addition of `os-string` as a boot package. - - - - - ab533e71 by Matthew Pickering at 2024-02-08T00:37:50-05:00 Use specific clang assembler when compiling with -fllvm There are situations where LLVM will produce assembly which older gcc toolchains can't handle. For example on Deb10, it seems that LLVM >= 13 produces assembly which the default gcc doesn't support. A more robust solution in the long term is to require a specific LLVM compatible assembler when using -fllvm. Fixes #16354 - - - - - c32b6426 by Matthew Pickering at 2024-02-08T00:37:50-05:00 Update CI images with LLVM 15, ghc-9.6.4 and cabal-install-3.10.2.0 - - - - - 5fcd58be by Matthew Pickering at 2024-02-08T00:37:50-05:00 Update bootstrap plans for 9.4.8 and 9.6.4 - - - - - 707a32f5 by Matthew Pickering at 2024-02-08T00:37:50-05:00 Add alpine 3_18 release job This is mainly experimental and future proofing to enable a smooth transition to newer alpine releases once 3_12 is too old. - - - - - c37931b3 by John Ericson at 2024-02-08T06:39:05-05:00 Generate LLVM min/max bound policy via Hadrian Per #23966, I want the top-level configure to only generate configuration data for Hadrian, not do any "real" tasks on its own. This is part of that effort --- one less file generated by it. (It is still done with a `.in` file, so in a future world non-Hadrian also can easily create this file.) Split modules: - GHC.CmmToLlvm.Config - GHC.CmmToLlvm.Version - GHC.CmmToLlvm.Version.Bounds - GHC.CmmToLlvm.Version.Type This also means we can get rid of the silly `unused.h` introduced in !6803 / 7dfcab2f4bcb7206174ea48857df1883d05e97a2 as temporary kludge. Part of #23966 - - - - - 9f987235 by Apoorv Ingle at 2024-02-08T06:39:42-05:00 Enable mdo statements to use HsExpansions Fixes: #24411 Added test T24411 for regression - - - - - f8429266 by Jade at 2024-02-08T14:56:50+01:00 Adjust test for ghc MR !10993 - - - - - 762b2120 by Jade at 2024-02-08T15:17:15+00:00 Improve Monad, Functor & Applicative docs This patch aims to improve the documentation of Functor, Applicative, Monad and related symbols. The main goal is to make it more consistent and make accessible. See also: !10979 (closed) and !10985 (closed) Ticket #17929 Updates haddock submodule - - - - - 151770ca by Josh Meredith at 2024-02-10T14:28:15-05:00 JavaScript codegen: Use GHC's tag inference where JS backend-specific evaluation inference was previously used (#24309) - - - - - 2e880635 by Zubin Duggal at 2024-02-10T14:28:51-05:00 ci: Allow release-hackage-lint to fail Otherwise it blocks the ghcup metadata pipeline from running. - - - - - b0293f78 by Matthew Pickering at 2024-02-10T14:29:28-05:00 rts: eras profiling mode The eras profiling mode is useful for tracking the life-time of closures. When a closure is written, the current era is recorded in the profiling header. This records the era in which the closure was created. * Enable with -he * User mode: Use functions ghc-experimental module GHC.Profiling.Eras to modify the era * Automatically: --automatic-era-increment, increases the user era on major collections * The first era is era 1 * -he<era> can be used with other profiling modes to select a specific era If you just want to record the era but not to perform heap profiling you can use `-he --no-automatic-heap-samples`. https://well-typed.com/blog/2024/01/ghc-eras-profiling/ Fixes #24332 - - - - - be674a2c by Jade at 2024-02-10T14:30:04-05:00 Adjust error message for trailing whitespace in as-pattern. Fixes #22524 - - - - - 53ef83f9 by doyougnu at 2024-02-10T14:30:47-05:00 gitlab: js: add codeowners Fixes: - #24409 Follow on from: - #21078 and MR !9133 - When we added the JS backend this was forgotten. This patch adds the rightful codeowners. - - - - - 8bbe12f2 by Matthew Pickering at 2024-02-10T14:31:23-05:00 Bump CI images so that alpine3_18 image includes clang15 The only changes here are that clang15 is now installed on the alpine-3_18 image. - - - - - df9fd9f7 by Sylvain Henry at 2024-02-12T12:18:42-05:00 JS: handle stored null StablePtr Some Haskell codes unsafely cast StablePtr into ptr to compare against NULL. E.g. in direct-sqlite: if castStablePtrToPtr aggStPtr /= nullPtr then where `aggStPtr` is read (`peek`) from zeroed memory initially. We fix this by giving these StablePtr the same representation as other null pointers. It's safe because StablePtr at offset 0 is unused (for this exact reason). - - - - - 55346ede by Sylvain Henry at 2024-02-12T12:18:42-05:00 JS: disable MergeObjsMode test This isn't implemented for JS backend objects. - - - - - aef587f6 by Sylvain Henry at 2024-02-12T12:18:42-05:00 JS: add support for linking C sources Support linking C sources with JS output of the JavaScript backend. See the added documentation in the users guide. The implementation simply extends the JS linker to use the objects (.o) that were already produced by the emcc compiler and which were filtered out previously. I've also added some options to control the link with C functions (see the documentation about pragmas). With this change I've successfully compiled the direct-sqlite package which embeds the sqlite.c database code. Some wrappers are still required (see the documentation about wrappers) but everything generic enough to be reused for other libraries have been integrated into rts/js/mem.js. - - - - - b71b392f by Sylvain Henry at 2024-02-12T12:18:42-05:00 JS: avoid EMCC logging spurious failure emcc would sometime output messages like: cache:INFO: generating system asset: symbol_lists/424b44514e43d789148e69e4e7d1c7fdc0350b79.json... (this will be cached in "/emsdk/upstream/emscripten/cache/symbol_lists/424b44514e43d789148e69e4e7d1c7fdc0350b79.json" for subsequent builds) cache:INFO: - ok Cf https://github.com/emscripten-core/emscripten/issues/18607 This breaks our tests matching the stderr output. We avoid this by setting EMCC_LOGGING=0 - - - - - ff2c0cc9 by Simon Peyton Jones at 2024-02-12T12:19:17-05:00 Remove a dead comment Just remove an out of date block of commented-out code, and tidy up the relevant Notes. See #8317. - - - - - bedb4f0d by Teo Camarasu at 2024-02-12T18:50:33-05:00 nonmoving: Add support for heap profiling Add support for heap profiling while using the nonmoving collector. We greatly simply the implementation by disabling concurrent collection for GCs when heap profiling is enabled. This entails that the marked objects on the nonmoving heap are exactly the live objects. Note that we match the behaviour for live bytes accounting by taking the size of objects on the nonmoving heap to be that of the segment's block rather than the object itself. Resolves #22221 - - - - - d0d5acb5 by Teo Camarasu at 2024-02-12T18:51:09-05:00 doc: Add requires prof annotation to options that require it Resolves #24421 - - - - - 6d1e2386 by Alan Zimmerman at 2024-02-13T22:00:28+03:00 EPA: Match changes to HsParTy and HsFunTy - - - - - 57bb8c92 by Cheng Shao at 2024-02-13T14:07:49-05:00 deriveConstants: add needed constants for wasm backend This commit adds needed constants to deriveConstants. They are used by RTS code in the wasm backend to support the JSFFI logic. - - - - - 615eb855 by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: make genSym use C-based atomic increment on non-JS 32-bit platforms The pure Haskell implementation causes i386 regression in unrelated work that can be fixed by using C-based atomic increment, see added comment for details. - - - - - a9918891 by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: allow JSFFI for wasm32 This commit allows the javascript calling convention to be used when the target platform is wasm32. - - - - - 8771a53b by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: allow boxed JSVal as a foreign type This commit allows the boxed JSVal type to be used as a foreign argument/result type. - - - - - 053c92b3 by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: ensure ctors have the right priority on wasm32 This commit fixes the priorities of ctors generated by GHC codegen on wasm32, see the referred note for details. - - - - - b7942e0a by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: add JSFFI desugar logic for wasm32 This commit adds JSFFI desugar logic for the wasm backend. - - - - - 2c1dca76 by Cheng Shao at 2024-02-13T14:07:49-05:00 compiler: add JavaScriptFFI to supported extension list on wasm32 This commit adds JavaScriptFFI as a supported extension when the target platform is wasm32. - - - - - 9ad0e2b4 by Cheng Shao at 2024-02-13T14:07:49-05:00 rts/ghc-internal: add JSFFI support logic for wasm32 This commit adds rts/ghc-internal logic to support the wasm backend's JSFFI functionality. - - - - - e9ebea66 by Cheng Shao at 2024-02-13T14:07:49-05:00 ghc-internal: fix threadDelay for wasm in browsers This commit fixes broken threadDelay for wasm when it runs in browsers, see added note for detailed explanation. - - - - - f85f3fdb by Cheng Shao at 2024-02-13T14:07:49-05:00 utils: add JSFFI utility code This commit adds JavaScript util code to utils to support the wasm backend's JSFFI functionality: - jsffi/post-link.mjs, a post-linker to process the linked wasm module and emit a small complement JavaScript ESM module to be used with it at runtime - jsffi/prelude.js, a tiny bit of prelude code as the JavaScript side of runtime logic - jsffi/test-runner.mjs, run the jsffi test cases Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - 77e91500 by Cheng Shao at 2024-02-13T14:07:49-05:00 hadrian: distribute jsbits needed for wasm backend's JSFFI support The post-linker.mjs/prelude.js files are now distributed in the bindist libdir, so when using the wasm backend's JSFFI feature, the user wouldn't need to fetch them from a ghc checkout manually. - - - - - c47ba1c3 by Cheng Shao at 2024-02-13T14:07:49-05:00 testsuite: add opts.target_wrapper This commit adds opts.target_wrapper which allows overriding the target wrapper on a per test case basis when testing a cross target. This is used when testing the wasm backend's JSFFI functionality; the rest of the cases are tested using wasmtime, though the jsffi cases are tested using the node.js based test runner. - - - - - 8e048675 by Cheng Shao at 2024-02-13T14:07:49-05:00 testsuite: T22774 should work for wasm JSFFI T22774 works since the wasm backend now supports the JSFFI feature. - - - - - 1d07f9a6 by Cheng Shao at 2024-02-13T14:07:49-05:00 testsuite: add JSFFI test cases for wasm backend This commit adds a few test cases for the wasm backend's JSFFI functionality, as well as a simple README to instruct future contributors to add new test cases. - - - - - b8997080 by Cheng Shao at 2024-02-13T14:07:49-05:00 docs: add documentation for wasm backend JSFFI This commit adds changelog and user facing documentation for the wasm backend's JSFFI feature. - - - - - ffeb000d by David Binder at 2024-02-13T14:08:30-05:00 Add tests from libraries/process/tests and libraries/Win32/tests to GHC These tests were previously part of the libraries, which themselves are submodules of the GHC repository. This commit moves the tests directly to the GHC repository. - - - - - 5a932cf2 by David Binder at 2024-02-13T14:08:30-05:00 Do not execute win32 tests on non-windows runners - - - - - 500d8cb8 by Jade at 2024-02-13T14:09:07-05:00 prevent GHCi (and runghc) from suggesting other symbols when not finding main Fixes: #23996 - - - - - b19ec331 by Cheng Shao at 2024-02-13T14:09:46-05:00 rts: update xxHash to v0.8.2 - - - - - 4a97bdb8 by Cheng Shao at 2024-02-13T14:09:46-05:00 rts: use XXH3_64bits hash on all 64-bit platforms This commit enables XXH3_64bits hash to be used on all 64-bit platforms. Previously it was only enabled on x86_64, so platforms like aarch64 silently falls back to using XXH32 which degrades the hashing function quality. - - - - - ee01de7d by Cheng Shao at 2024-02-13T14:09:46-05:00 rts: define XXH_INLINE_ALL This commit cleans up how we include the xxhash.h header and only define XXH_INLINE_ALL, which is sufficient to inline the xxHash functions without symbol collision. - - - - - 0e01e1db by Alan Zimmerman at 2024-02-14T02:13:22-05:00 EPA: Move EpAnn out of extension points Leaving a few that are too tricky, maybe some other time. Also - remove some unneeded helpers from Parser.y - reduce allocations with strictness annotations Updates haddock submodule Metric Decrease: parsing001 - - - - - de589554 by Andreas Klebinger at 2024-02-14T02:13:59-05:00 Fix ffi callbacks with >6 args and non-64bit args. Check for ptr/int arguments rather than 64-bit width arguments when counting integer register arguments. The old approach broke when we stopped using exclusively W64-sized types to represent sub-word sized integers. Fixes #24314 - - - - - 9c588f19 by Fendor at 2024-02-14T11:05:36+01:00 Adapt to GHC giving better Name's for linking - - - - - 325b7613 by Ben Gamari at 2024-02-14T14:27:45-05:00 rts/EventLog: Place eliminate duplicate strlens Previously many of the `post*` implementations would first compute the length of the event's strings in order to determine the event length. Later we would then end up computing the length yet again in `postString`. Now we instead pass the string length to `postStringLen`, avoiding the repeated work. - - - - - 8aafa51c by Ben Gamari at 2024-02-14T14:27:46-05:00 rts/eventlog: Place upper bound on IPE string field lengths The strings in IPE events may be of unbounded length. Limit the lengths of these fields to 64k characters to ensure that we don't exceed the maximum event length. - - - - - 0e60d52c by Zubin Duggal at 2024-02-14T14:27:46-05:00 rts: drop unused postString function - - - - - d8d1333a by Cheng Shao at 2024-02-14T14:28:23-05:00 compiler/rts: fix wasm unreg regression This commit fixes two wasm unreg regressions caught by a nightly pipeline: - Unknown stg_scheduler_loopzh symbol when compiling scheduler.cmm - Invalid _hs_constructor(101) function name when handling ctor - - - - - 264a4fa9 by Owen Shepherd at 2024-02-15T09:41:06-05:00 feat: Add sortOn to Data.List.NonEmpty Adds `sortOn` to `Data.List.NonEmpty`, and adds comments describing when to use it, compared to `sortWith` or `sortBy . comparing`. The aim is to smooth out the API between `Data.List`, and `Data.List.NonEmpty`. This change has been discussed in the [clc issue](https://github.com/haskell/core-libraries-committee/issues/227). - - - - - b57200de by Fendor at 2024-02-15T09:41:47-05:00 Prefer RdrName over OccName for looking up locations in doc renaming step Looking up by OccName only does not take into account when functions are only imported in a qualified way. Fixes issue #24294 Bump haddock submodule to include regression test - - - - - 8ad02724 by Luite Stegeman at 2024-02-15T17:33:32-05:00 JS: add simple optimizer The simple optimizer reduces the size of the code generated by the JavaScript backend without the complexity and performance penalty of the optimizer in GHCJS. Also see #22736 Metric Decrease: libdir size_hello_artifact - - - - - 20769b36 by Matthew Pickering at 2024-02-15T17:34:07-05:00 base: Expose `--no-automatic-time-samples` in `GHC.RTS.Flags` API This patch builds on 5077416e12cf480fb2048928aa51fa4c8fc22cf1 and modifies the base API to reflect the new RTS flag. CLC proposal #243 - https://github.com/haskell/core-libraries-committee/issues/243 Fixes #24337 - - - - - 778e1db3 by Andrei Borzenkov at 2024-02-16T16:12:07+03:00 Namespace specifiers for fixity signatures - - - - - 08031ada by Teo Camarasu at 2024-02-16T13:37:00-05:00 base: export System.Mem.performBlockingMajorGC The corresponding C function was introduced in ba73a807edbb444c49e0cf21ab2ce89226a77f2e. As part of #22264. Resolves #24228 The CLC proposal was disccused at: https://github.com/haskell/core-libraries-committee/issues/230 Co-authored-by: Ben Gamari <bgamari.foss at gmail.com> - - - - - 1f534c2e by Florian Weimer at 2024-02-16T13:37:42-05:00 Fix C output for modern C initiative GCC 14 on aarch64 rejects the C code written by GHC with this kind of error: error: assignment to ‘ffi_arg’ {aka ‘long unsigned int’} from ‘HsPtr’ {aka ‘void *’} makes integer from pointer without a cast [-Wint-conversion] 68 | *(ffi_arg*)resp = cret; | ^ Add the correct cast. For more information on this see: https://fedoraproject.org/wiki/Changes/PortingToModernC Tested-by: Richard W.M. Jones <rjones at redhat.com> - - - - - 5d3f7862 by Matthew Craven at 2024-02-16T13:38:18-05:00 Bump bytestring submodule to 0.12.1.0 - - - - - 902ebcc2 by Ian-Woo Kim at 2024-02-17T06:01:01-05:00 Add missing BCO handling in scavenge_one. - - - - - 97d26206 by Sylvain Henry at 2024-02-17T06:01:44-05:00 Make cast between words and floats real primops (#24331) First step towards fixing #24331. Replace foreign prim imports with real primops. - - - - - a40e4781 by Sylvain Henry at 2024-02-17T06:01:44-05:00 Perf: add constant folding for bitcast between float and word (#24331) - - - - - 5fd2c00f by Sylvain Henry at 2024-02-17T06:01:44-05:00 Perf: replace stack checks with assertions in casting primops There are RESERVED_STACK_WORDS free words (currently 21) on the stack, so omit the checks. Suggested by Cheng Shao. - - - - - 401dfe7b by Sylvain Henry at 2024-02-17T06:01:44-05:00 Reexport primops from GHC.Float + add deprecation - - - - - 4ab48edb by Ben Gamari at 2024-02-17T06:02:21-05:00 rts/Hash: Don't iterate over chunks if we don't need to free data When freeing a `HashTable` there is no reason to walk over the hash list before freeing it if the user has not given us a `dataFreeFun`. Noticed while looking at #24410. - - - - - bd5a1f91 by Cheng Shao at 2024-02-17T06:03:00-05:00 compiler: add SEQ_CST fence support In addition to existing Acquire/Release fences, this commit adds SEQ_CST fence support to GHC, allowing Cmm code to explicitly emit a fence that enforces total memory ordering. The following logic is added: - The MO_SeqCstFence callish MachOp - The %prim fence_seq_cst() Cmm syntax and the SEQ_CST_FENCE macro in Cmm.h - MO_SeqCstFence lowering logic in every single GHC codegen backend - - - - - 2ce2a493 by Cheng Shao at 2024-02-17T06:03:38-05:00 testsuite: fix hs_try_putmvar002 for targets without pthread.h hs_try_putmvar002 includes pthread.h and doesn't work on targets without this header (e.g. wasm32). It doesn't need to include this header at all. This was previously unnoticed by wasm CI, though recent toolchain upgrade brought in upstream changes that completely removes pthread.h in the single-threaded wasm32-wasi sysroot, therefore we need to handle that change. - - - - - 1fb3974e by Cheng Shao at 2024-02-17T06:03:38-05:00 ci: bump ci-images to use updated wasm image This commit bumps our ci-images revision to use updated wasm image. - - - - - 56e3f097 by Andrew Lelechenko at 2024-02-17T06:04:13-05:00 Bump submodule text to 2.1.1 T17123 allocates less because of improvements to Data.Text.concat in 1a6a06a. Metric Decrease: T17123 - - - - - a7569495 by Cheng Shao at 2024-02-17T06:04:51-05:00 rts: remove redundant rCCCS initialization This commit removes the redundant logic of initializing each Capability's rCCCS to CCS_SYSTEM in initProfiling(). Before initProfiling() is called during RTS startup, each Capability's rCCCS has already been assigned CCS_SYSTEM when they're first initialized. - - - - - 7a0293cc by Ben Gamari at 2024-02-19T07:11:00-05:00 Drop dependence on `touch` This drops GHC's dependence on the `touch` program, instead implementing it within GHC. This eliminates an external dependency and means that we have one fewer program to keep track of in the `configure` script - - - - - 0dbd729e by Andrei Borzenkov at 2024-02-19T07:11:37-05:00 Parser, renamer, type checker for @a-binders (#17594) GHC Proposal 448 introduces binders for invisible type arguments (@a-binders) in various contexts. This patch implements @-binders in lambda patterns and function equations: {-# LANGUAGE TypeAbstractions #-} id1 :: a -> a id1 @t x = x :: t -- @t-binder on the LHS of a function equation higherRank :: (forall a. (Num a, Bounded a) => a -> a) -> (Int8, Int16) higherRank f = (f 42, f 42) ex :: (Int8, Int16) ex = higherRank (\ @a x -> maxBound @a - x ) -- @a-binder in a lambda pattern in an argument -- to a higher-order function Syntax ------ To represent those @-binders in the AST, the list of patterns in Match now uses ArgPat instead of Pat: data Match p body = Match { ... - m_pats :: [LPat p], + m_pats :: [LArgPat p], ... } + data ArgPat pass + = VisPat (XVisPat pass) (LPat pass) + | InvisPat (XInvisPat pass) (HsTyPat (NoGhcTc pass)) + | XArgPat !(XXArgPat pass) The VisPat constructor represents patterns for visible arguments, which include ordinary value-level arguments and required type arguments (neither is prefixed with a @), while InvisPat represents invisible type arguments (prefixed with a @). Parser ------ In the grammar (Parser.y), the lambda and lambda-cases productions of aexp non-terminal were updated to accept argpats instead of apats: aexp : ... - | '\\' apats '->' exp + | '\\' argpats '->' exp ... - | '\\' 'lcases' altslist(apats) + | '\\' 'lcases' altslist(argpats) ... + argpat : apat + | PREFIX_AT atype Function left-hand sides did not require any changes to the grammar, as they were already parsed with productions capable of parsing @-binders. Those binders were being rejected in post-processing (isFunLhs), and now we accept them. In Parser.PostProcess, patterns are constructed with the help of PatBuilder, which is used as an intermediate data structure when disambiguating between FunBind and PatBind. In this patch we define ArgPatBuilder to accompany PatBuilder. ArgPatBuilder is a short-lived data structure produced in isFunLhs and consumed in checkFunBind. Renamer ------- Renaming of @-binders builds upon prior work on type patterns, implemented in 2afbddb0f24, which guarantees proper scoping and shadowing behavior of bound type variables. This patch merely defines rnLArgPatsAndThen to process a mix of visible and invisible patterns: + rnLArgPatsAndThen :: NameMaker -> [LArgPat GhcPs] -> CpsRn [LArgPat GhcRn] + rnLArgPatsAndThen mk = mapM (wrapSrcSpanCps rnArgPatAndThen) where + rnArgPatAndThen (VisPat x p) = ... rnLPatAndThen ... + rnArgPatAndThen (InvisPat _ tp) = ... rnHsTyPat ... Common logic between rnArgPats and rnPats is factored out into the rn_pats_general helper. Type checker ------------ Type-checking of @-binders builds upon prior work on lazy skolemisation, implemented in f5d3e03c56f. This patch extends tcMatchPats to handle @-binders. Now it takes and returns a list of LArgPat rather than LPat: tcMatchPats :: ... - -> [LPat GhcRn] + -> [LArgPat GhcRn] ... - -> TcM ([LPat GhcTc], a) + -> TcM ([LArgPat GhcTc], a) Invisible binders in the Match are matched up with invisible (Specified) foralls in the type. This is done with a new clause in the `loop` worker of tcMatchPats: loop :: [LArgPat GhcRn] -> [ExpPatType] -> TcM ([LArgPat GhcTc], a) loop (L l apat : pats) (ExpForAllPatTy (Bndr tv vis) : pat_tys) ... -- NEW CLAUSE: | InvisPat _ tp <- apat, isSpecifiedForAllTyFlag vis = ... In addition to that, tcMatchPats no longer discards type patterns. This is done by filterOutErasedPats in the desugarer instead. x86_64-linux-deb10-validate+debug_info Metric Increase: MultiLayerModulesTH_OneShot - - - - - 486979b0 by Jade at 2024-02-19T07:12:13-05:00 Add specialized sconcat implementation for Data.Monoid.First and Data.Semigroup.First Approved CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/246 Fixes: #24346 - - - - - 17e309d2 by John Ericson at 2024-02-19T07:12:49-05:00 Fix reST in users guide It appears that aef587f65de642142c1dcba0335a301711aab951 wasn't valid syntax. - - - - - 35b0ad90 by Brandon Chinn at 2024-02-19T07:13:25-05:00 Fix searching for errors in sphinx build - - - - - 4696b966 by Cheng Shao at 2024-02-19T07:14:02-05:00 hadrian: fix wasm backend post linker script permissions The post-link.mjs script was incorrectly copied and installed as a regular data file without executable permission, this commit fixes it. - - - - - a6142e0c by Cheng Shao at 2024-02-19T07:14:40-05:00 testsuite: mark T23540 as fragile on i386 See #24449 for details. - - - - - 249caf0d by Matthew Craven at 2024-02-19T20:36:09-05:00 Add @since annotation to Data.Data.mkConstrTag - - - - - cdd939e7 by Jade at 2024-02-19T20:36:46-05:00 Enhance documentation of Data.Complex - - - - - d04f384f by Ben Gamari at 2024-02-21T04:59:23-05:00 hadrian/bindist: Ensure that phony rules are marked as such Otherwise make may not run the rule if file with the same name as the rule happens to exist. - - - - - efcbad2d by Ben Gamari at 2024-02-21T04:59:23-05:00 hadrian: Generate HSC2HS_EXTRAS variable in bindist installation We must generate the hsc2hs wrapper at bindist installation time since it must contain `--lflag` and `--cflag` arguments which depend upon the installation path. The solution here is to substitute these variables in the configure script (see mk/hsc2hs.in). This is then copied over a dummy wrapper in the install rules. Fixes #24050. - - - - - c540559c by Matthew Pickering at 2024-02-21T04:59:23-05:00 ci: Show --info for installed compiler - - - - - ab9281a2 by Matthew Pickering at 2024-02-21T04:59:23-05:00 configure: Correctly set --target flag for linker opts Previously we were trying to use the FP_CC_SUPPORTS_TARGET with 4 arguments, when it only takes 3 arguments. Instead we need to use the `FP_PROG_CC_LINKER_TARGET` function in order to set the linker flags. Actually fixes #24414 - - - - - 9460d504 by Rodrigo Mesquita at 2024-02-21T04:59:59-05:00 configure: Do not override existing linker flags in FP_LD_NO_FIXUP_CHAINS - - - - - 77629e76 by Andrei Borzenkov at 2024-02-21T05:00:35-05:00 Namespacing for fixity signatures (#14032) Namespace specifiers were added to syntax of fixity signatures: - sigdecl ::= infix prec ops | ... + sigdecl ::= infix prec namespace_spec ops | ... To preserve namespace during renaming MiniFixityEnv type now has separate FastStringEnv fields for names that should be on the term level and for name that should be on the type level. makeMiniFixityEnv function was changed to fill MiniFixityEnv in the right way: - signatures without namespace specifiers fill both fields - signatures with 'data' specifier fill data field only - signatures with 'type' specifier fill type field only Was added helper function lookupMiniFixityEnv that takes care about looking for a name in an appropriate namespace. Updates haddock submodule. Metric Decrease: MultiLayerModulesTH_OneShot - - - - - 84357d11 by Teo Camarasu at 2024-02-21T05:01:11-05:00 rts: only collect live words in nonmoving census when non-concurrent This avoids segfaults when the mutator modifies closures as we examine them. Resolves #24393 - - - - - 9ca56dd3 by Ian-Woo Kim at 2024-02-21T05:01:53-05:00 mutex wrap in refreshProfilingCCSs - - - - - 1387966a by Cheng Shao at 2024-02-21T05:02:32-05:00 rts: remove unused HAVE_C11_ATOMICS macro This commit removes the unused HAVE_C11_ATOMICS macro. We used to have a few places that have fallback paths when HAVE_C11_ATOMICS is not defined, but that is completely redundant, since the FP_CC_SUPPORTS__ATOMICS configure check will fail when the C compiler doesn't support C11 style atomics. There are also many places (e.g. in unreg backend, SMP.h, library cbits, etc) where we unconditionally use C11 style atomics anyway which work in even CentOS 7 (gcc 4.8), the oldest distro we test in our CI, so there's no value in keeping HAVE_C11_ATOMICS. - - - - - 0f40d68f by Andreas Klebinger at 2024-02-21T05:03:09-05:00 RTS: -Ds - make sure incall is non-zero before dereferencing it. Fixes #24445 - - - - - e5886de5 by Ben Gamari at 2024-02-21T05:03:44-05:00 rts/AdjustorPool: Use ExecPage abstraction This is just a minor cleanup I found while reviewing the implementation. - - - - - 826c5b47 by Torsten Schmits at 2024-02-21T13:17:05+01:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 09941666 by Adam Gundry at 2024-02-21T13:53:12+00:00 Define GHC2024 language edition (#24320) See https://github.com/ghc-proposals/ghc-proposals/pull/613. Also fixes #24343 and improves the documentation of language editions. Co-authored-by: Joachim Breitner <mail at joachim-breitner.de> - - - - - 2cff14d5 by Ben Gamari at 2024-02-22T09:35:56-05:00 Bump bounds - - - - - f49376b3 by Ben Gamari at 2024-02-22T09:35:56-05:00 Allow `@since` annotations in export lists Here we extend Haddock to admit `@since` annotations in export lists. These can be attached to most export list items (although not subordinate lists). These annotations supercede the declaration's `@since` annotation in produced Haddocks. - - - - - b5aa93df by Ben Gamari at 2024-02-22T12:09:06-05:00 Allow package-qualified @since declarations - - - - - 8f5957f2 by Ben Gamari at 2024-02-22T13:55:19-05:00 Documentation changes from ghc-internal restructuring Previously many declarations (e.g. `Int`) were declared to have a "home" in `Prelude`. However, now Haddock instead chooses to put these in more specific homes (e.g. `Data.Int`). Given that the "home" decision is driven by heuristics and in general these changes seem quite reasonable I am accepting them: * `Int` moved from `Prelude` to `Data.Int` * `(~)` moved from `Prelude` to `Data.Type.Equality` * `Type` moved from `GHC.Types` to `Data.Kind` * `Maybe` moved from `Prelude` to `Data.Maybe` * `Bool` moved from `Prelude` to `Data.Bool` * `Ordering` moved from `Prelude` to `Data.Ord` As well, more identifiers are now hyperlinked; it's not immediately clear *why*, but it is an improvement nevertheless. - - - - - ec33fec3 by Ben Gamari at 2024-02-22T20:36:24-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 5121a4ed by Ben Gamari at 2024-02-23T06:40:55-05:00 Allow docstrings after exports Here we extend the parser and AST to preserve docstrings following export items. We then extend Haddock to parse `@since` annotations in such docstrings, allowing changes in export structure to be properly documented. Bumps haddock submodule. - - - - - 30cfd251 by Torsten Schmits at 2024-02-24T13:00:42-05:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 0eb2265d by Hécate Moonlight at 2024-02-24T16:02:16-05:00 Improve the synopsis and description of base - - - - - 2e36f5d2 by Jade at 2024-02-24T16:02:51-05:00 Error Messages: Properly align cyclic module error Fixes: #24476 - - - - - bbfb051c by Ben Gamari at 2024-02-24T19:10:23-05:00 Allow docstrings after exports Here we extend the parser and AST to preserve docstrings following export items. We then extend Haddock to parse `@since` annotations in such docstrings, allowing changes in export structure to be properly documented. - - - - - d8d6ad8c by Ben Gamari at 2024-02-24T19:10:23-05:00 ghc-internal: Move modules into GHC.Internal.* namespace Bumps haddock submodule due to testsuite output changes. - - - - - a82af7cd by Ben Gamari at 2024-02-24T19:10:23-05:00 ghc-internal: Rewrite `@since ` to `@since base-` These will be incrementally moved to the export sites in `base` where possible. - - - - - ca3836e1 by Ben Gamari at 2024-02-24T19:10:23-05:00 base: Migrate Haddock `not-home` pragmas from `ghc-internal` This ensures that we do not use `base` stub modules as declarations' homes when not appropriate. - - - - - c8cf3e26 by Ben Gamari at 2024-02-24T19:10:23-05:00 base: Partially freeze exports of GHC.Base Sadly there are still a few module reexports. However, at least we have decoupled from the exports of `GHC.Internal.Base`. - - - - - 272573c6 by Ben Gamari at 2024-02-24T19:10:23-05:00 Move Haddock named chunks - - - - - 2d8a881d by Ben Gamari at 2024-02-24T19:10:23-05:00 Drop GHC.Internal.Data.Int - - - - - 55c4c385 by Ben Gamari at 2024-02-24T19:10:23-05:00 compiler: Fix mention to `GHC....` modules in wasm desugaring Really, these references should be via known-key names anyways. I have fixed the proximate issue here but have opened #24472 to track the additional needed refactoring. - - - - - 64150911 by Ben Gamari at 2024-02-24T19:10:23-05:00 Accept performance shifts from ghc-internal restructure As expected, Haddock now does more work. Less expected is that some other testcases actually get faster, presumably due to less interface file loading. As well, the size_hello_artifact test regressed a bit when debug information is enabled due to debug information for the new stub symbols. Metric Decrease: T12227 T13056 Metric Increase: haddock.Cabal haddock.base MultiLayerModulesTH_OneShot size_hello_artifact - - - - - 317a915b by Ben Gamari at 2024-02-24T19:10:23-05:00 Expose GHC.Wasm.Prim from ghc-experimental Previously this was only exposed from `ghc-internal` which violates our agreement that users shall not rely on things exposed from that package. Fixes #24479. - - - - - 3bbd2bf2 by Ben Gamari at 2024-02-24T19:10:23-05:00 compiler/tc: Small optimisation of evCallStack Don't lookupIds unless we actually need them. - - - - - 3e5c9e3c by Ben Gamari at 2024-02-24T19:10:23-05:00 compiler/tc: Use toException instead of SomeException - - - - - 125714a6 by Ben Gamari at 2024-02-24T19:10:23-05:00 base: Factor out errorBelch This was useful when debugging - - - - - 3d6aae7c by Ben Gamari at 2024-02-24T19:10:23-05:00 base: Clean up imports of GHC.Stack.CloneStack - - - - - 6900306e by Ben Gamari at 2024-02-24T19:10:24-05:00 base: Move PrimMVar to GHC.Internal.MVar - - - - - 28f8a148 by Ben Gamari at 2024-02-24T19:10:24-05:00 base: Move prettyCallStack to GHC.Internal.Stack - - - - - 4892de47 by Ben Gamari at 2024-02-24T19:10:24-05:00 base: Explicit dependency to workaround #24436 Currently `ghc -M` fails to account for `.hs-boot` files correctly, leading to issues with cross-package one-shot builds failing. This currently manifests in `GHC.Exception` due to the boot file for `GHC.Internal.Stack`. Work around this by adding an explicit `import`, ensuring that `GHC.Internal.Stack` is built before `GHC.Exception`. See #24436. - - - - - 294c93a5 by Ben Gamari at 2024-02-24T19:10:24-05:00 base: Use displayException in top-level exception handler Happily this also allows us to eliminate a special case for Deadlock exceptions. Implements [CLC #198](https://github.com/haskell/core-libraries-committee/issues/198). - - - - - 732db81d by Ben Gamari at 2024-02-24T19:12:18-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - cf756a25 by Ben Gamari at 2024-02-24T22:11:53-05:00 rts: Fix symbol references in Wasm RTS - - - - - 4e4d47a0 by Jade at 2024-02-26T15:17:20-05:00 GHCi: Improve response to unloading, loading and reloading modules Fixes #13869 - - - - - f3de8a3c by Zubin Duggal at 2024-02-26T15:17:57-05:00 rel-eng/fetch-gitlab.py: Fix name of aarch64 alpine 3_18 release job - - - - - c71bfdff by Cheng Shao at 2024-02-26T15:18:35-05:00 hadrian/hie-bios: pass -j to hadrian This commit passes -j to hadrian in the hadrian/hie-bios scripts. When the user starts HLS in a fresh clone that has just been configured, it takes quite a while for hie-bios to pick up the ghc flags and start actual indexing, due to the fact that the hadrian build step defaulted to -j1, so -j speeds things up and improve HLS user experience in GHC. Also add -j flag to .ghcid to speed up ghcid, and sets the Windows build root to .hie-bios which also works and unifies with other platforms, the previous build root _hie-bios was missing from .gitignore anyway. - - - - - 50bfdb46 by Cheng Shao at 2024-02-26T15:18:35-05:00 ci: enable parallelism in hadrian/ghci scripts This commit enables parallelism when the hadrian/ghci scripts are called in CI. The time bottleneck is in the hadrian build step, but previously the build step wasn't parallelized. - - - - - 61a78231 by Felix Yan at 2024-02-26T15:19:14-05:00 m4: Correctly detect GCC version When calling as `cc`, GCC does not outputs lowercased "gcc" at least in 13.2.1 version here. ``` $ cc --version cc (GCC) 13.2.1 20230801 ... ``` This fails the check and outputs the confusing message: `configure: $CC is not gcc; assuming it's a reasonably new C compiler` This patch makes it check for upper-cased "GCC" too so that it works correctly: ``` checking version of gcc... 13.2.1 ``` - - - - - 001aa539 by Teo Camarasu at 2024-02-27T13:26:46-05:00 Fix formatting in whereFrom docstring Previously it used markdown syntax rather than Haddock syntax for code quotes - - - - - e8034d15 by Teo Camarasu at 2024-02-27T13:26:46-05:00 Move ClosureType type to ghc-internal - Use ClosureType for InfoProv.ipDesc. - Use ClosureType for CloneStack.closureType. - Now ghc-heap re-exports this type from ghc-internal. See the accompanying CLC proposal: https://github.com/haskell/core-libraries-committee/issues/210 Resolves #22600 - - - - - 3da0a551 by Matthew Craven at 2024-02-27T13:27:22-05:00 StgToJS: Simplify ExprInline constructor of ExprResult Its payload was used only for a small optimization in genAlts, avoiding a few assignments for programs of this form: case NormalDataCon arg1 arg2 of x { NormalDataCon x1 x2 -> ... ; } But when compiling with optimizations, this sort of code is generally eliminated by case-of-known-constructor in Core-to-Core. So it doesn't seem worth tracking and cleaning up again in StgToJS. - - - - - 61bc92cc by Cheng Shao at 2024-02-27T16:58:42-05:00 rts: add missing ccs_mutex guard to internal_dlopen See added comment for details. Closes #24423. - - - - - dd29d3b2 by doyougnu at 2024-02-27T16:59:23-05:00 cg: Remove GHC.Cmm.DataFlow.Collections In pursuit of #15560 and #17957 and generally removing redundancy. - - - - - d3a050d2 by Cheng Shao at 2024-02-27T17:00:00-05:00 utils: remove unused lndir from tree Ever since the removal of the make build system, the in tree lndir hasn't been actually built, so this patch removes it. - - - - - 86bf7010 by Ben Gamari at 2024-02-27T19:28:10-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 74b24a9b by Teo Camarasu at 2024-02-28T16:32:58+00:00 rts: avoid checking bdescr of value outside of Haskell heap In nonmovingTidyWeaks we want to check if the key of a weak pointer lives in the non-moving heap. We do this by checking the flags of the block the key lives in. But we need to be careful with values that live outside the Haskell heap, since they will lack a block descriptor and looking for one may lead to a segfault. In this case we should just accept that it isn't on the non-moving heap. Resolves #24492 - - - - - b4cae4ec by Simon Peyton Jones at 2024-02-29T02:10:08-05:00 In mkDataConRep, ensure the in-scope set is right A small change that fixes #24489 - - - - - 3836a110 by Cheng Shao at 2024-02-29T21:25:45-05:00 testsuite: fix T23540 fragility on 32-bit platforms T23540 is fragile on 32-bit platforms. The root cause is usage of `getEvidenceTreesAtPoint`, which internally relies on `Name`'s `Ord` instance, which is indeterministic. The solution is adding a deterministic `Ord` instance for `EvidenceInfo` and sorting the evidence trees before pretty printing. Fixes #24449. - - - - - 960c8d47 by Teo Camarasu at 2024-02-29T21:26:20-05:00 Reduce AtomicModifyIORef increment count This test leads to a lot of contention when N>2 and becomes very slow. Let's reduce the amount of work we do to compensate. Resolves #24490 - - - - - 2e46c8ad by Matthew Pickering at 2024-03-01T05:48:06-05:00 hadrian: Improve parallelism in binary-dist-dir rule I noticed that the "docs" target was needed after the libraries and executables were built. We can improve the parallelism by needing everything at once so that documentation can be built immediately after a library is built for example. - - - - - cb6c11fe by Matthew Pickering at 2024-03-01T05:48:07-05:00 ci: Bump windows and freebsd boot compilers to 9.6.4 We have previously bumped the docker images to use 9.6.4, but neglected to bump the windows images until now. - - - - - 30f06996 by Matthew Pickering at 2024-03-01T05:48:07-05:00 ci: darwin: Update to 9.6.2 for boot compiler 9.6.4 is currently broken due to #24050 Also update to use LLVM-15 rather than LLVM-11, which is out of date. - - - - - d9d69e12 by Matthew Pickering at 2024-03-01T05:48:07-05:00 Bump minimum bootstrap version to 9.6 - - - - - 67ace1c5 by Matthew Pickering at 2024-03-01T05:48:07-05:00 ci: Enable more documentation building Here we enable documentation building on 1. Darwin: The sphinx toolchain was already installed so we enable html and manpages. 2. Rocky8: Full documentation (toolchain already installed) 3. Alpine: Full documetnation (toolchain already installed) 4. Windows: HTML and manpages (toolchain already installed) Fixes #24465 - - - - - 39583c39 by Matthew Pickering at 2024-03-01T05:48:42-05:00 ci: Bump ci-images to allow updated aarch64-alpine image with llvm15 and clang15 - - - - - d91d00fc by Torsten Schmits at 2024-03-01T15:01:50-05:00 Introduce ListTuplePuns extension This implements Proposal 0475, introducing the `ListTuplePuns` extension which is enabled by default. Disabling this extension makes it invalid to refer to list, tuple and sum type constructors by using built-in syntax like `[Int]`, `(Int, Int)`, `(# Int#, Int# #)` or `(# Int | Int #)`. Instead, this syntax exclusively denotes data constructors for use with `DataKinds`. The conventional way of referring to these data constructors by prefixing them with a single quote (`'(Int, Int)`) is now a parser error. Tuple declarations have been moved to `GHC.Tuple.Prim` and the `Solo` data constructor has been renamed to `MkSolo` (in a previous commit). Unboxed tuples and sums now have real source declarations in `GHC.Types`. Unit and solo types for tuples are now called `Unit`, `Unit#`, `Solo` and `Solo#`. Constraint tuples now have the unambiguous type constructors `CTuple<n>` as well as `CUnit` and `CSolo`, defined in `GHC.Classes` like before. A new parser construct has been added for the unboxed sum data constructor declarations. The type families `Tuple`, `Sum#` etc. that were intended to provide nicer syntax have been omitted from this change set due to inference problems, to be implemented at a later time. See the MR discussion for more info. Updates the submodule utils/haddock. Updates the cabal submodule due to new language extension. Metric Increase: haddock.base Metric Decrease: MultiLayerModulesTH_OneShot size_hello_artifact Proposal document: https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0475-tuple-syntax.rst Merge request: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/8820 Tracking ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/21294 - - - - - bbdb6286 by Sylvain Henry at 2024-03-01T15:01:50-05:00 JS linker: filter unboxed tuples - - - - - dec6d8d3 by Arnaud Spiwack at 2024-03-01T15:02:30-05:00 Improve error messages coming from non-linear patterns This enriched the `CtOrigin` for non-linear patterns to include data of the pattern that created the constraint (which can be quite useful if it occurs nested in a pattern) as well as an explanation why the pattern is non-restricted in (at least in some cases). - - - - - 6612388e by Arnaud Spiwack at 2024-03-01T15:02:30-05:00 Adjust documentation of linear lets according to committee decision - - - - - 1c064ef1 by Cheng Shao at 2024-03-02T17:11:19-05:00 compiler: start deprecating cmmToRawCmmHook cmmToRawCmmHook was added 4 years ago in d561c8f6244f8280a2483e8753c38e39d34c1f01. Its only user is the Asterius project, which has been archived and deprecated in favor of the ghc wasm backend. This patch starts deprecating cmmToRawCmmHook by placing a DEPRECATED pragma, and actual removal shall happen in a future GHC major release if no issue to oppose the deprecation has been raised in the meantime. - - - - - 9b74845f by Andrew Lelechenko at 2024-03-02T17:11:55-05:00 Data.List.NonEmpty.unzip: use WARNING with category instead of DEPRECATED CLC proposal: https://github.com/haskell/core-libraries-committee/issues/258 - - - - - 61bb5ff6 by Finley McIlwaine at 2024-03-04T09:01:40-08:00 add -fprof-late-overloaded and -fprof-late-overloaded-calls * Refactor late cost centre insertion for extensibility * Add two more late cost centre insertion methods that add SCCs to overloaded top level bindings and call sites with dictionary arguments. * Some tests for the basic functionality of the new insertion methods Resolves: #24500 - - - - - 82ccb801 by Andreas Klebinger at 2024-03-04T19:59:14-05:00 x86-ncg: Fix fma codegen when arguments are globals Fix a bug in the x86 ncg where results would be wrong when the desired output register and one of the input registers were the same global. Also adds a tiny optimization to make use of the memory addressing support when convenient. Fixes #24496 - - - - - 18ad1077 by Matthew Pickering at 2024-03-05T14:22:31-05:00 rel_eng: Update hackage docs upload scripts This adds the upload of ghc-internal and ghc-experimental to our scripts which upload packages to hackage. - - - - - bf47c9ba by Matthew Pickering at 2024-03-05T14:22:31-05:00 docs: Remove stray module comment from GHC.Profiling.Eras - - - - - 37d9b340 by Matthew Pickering at 2024-03-05T14:22:31-05:00 Fix ghc-internal cabal file The file mentioned some artifacts relating to the base library. I have renamed these to the new ghc-internal variants. - - - - - 23f2a478 by Matthew Pickering at 2024-03-05T14:22:31-05:00 Fix haddock source links and hyperlinked source There were a few issues with the hackage links: 1. We were using the package id rather than the package name for the package links. This is fixed by now allowing the template to mention %pkg% or %pkgid% and substituing both appropiatly. 2. The `--haddock-base-url` flag is renamed to `--haddock-for-hackage` as the new base link works on a local or remote hackage server. 3. The "src" path including too much stuff, so cross-package source links were broken as the template was getting double expanded. Fixes #24086 - - - - - 2fa336a9 by Ben Gamari at 2024-03-05T14:23:07-05:00 filepath: Bump submodule to 1.5.2.0 - - - - - 31217944 by Ben Gamari at 2024-03-05T14:23:07-05:00 os-string: Bump submodule to 2.0.2 - - - - - 4074a3f2 by Matthew Pickering at 2024-03-05T21:44:35-05:00 base: Reflect new era profiling RTS flags in GHC.RTS.Flags * -he profiling mode * -he profiling selector * --automatic-era-increment CLC proposal #254 - https://github.com/haskell/core-libraries-committee/issues/254 - - - - - a8c0e31b by Sylvain Henry at 2024-03-05T21:45:14-05:00 JS: faster implementation for some numeric primitives (#23597) Use faster implementations for the following primitives in the JS backend by not using JavaScript's BigInt: - plusInt64 - minusInt64 - minusWord64 - timesWord64 - timesInt64 Co-authored-by: Josh Meredith <joshmeredith2008 at gmail.com> - - - - - 21e3f325 by Cheng Shao at 2024-03-05T21:45:52-05:00 rts: add -xr option to control two step allocator reserved space size This patch adds a -xr RTS option to control the size of virtual memory address space reserved by the two step allocator on a 64-bit platform, see added documentation for explanation. Closes #24498. - - - - - dedcf102 by Cheng Shao at 2024-03-06T13:39:04-05:00 rts: expose HeapAlloc.h as public header This commit exposes HeapAlloc.h as a public header. The intention is to expose HEAP_ALLOCED/HEAP_ALLOCED_GC, so they can be used in assertions in other public headers, and they may also be useful for user code. - - - - - d19441d7 by Cheng Shao at 2024-03-06T13:39:04-05:00 rts: assert pointer is indeed heap allocated in Bdescr() This commit adds an assertion to Bdescr() to assert the pointer is indeed heap allocated. This is useful to rule out RTS bugs that attempt to access non-existent block descriptor of a static closure, #24492 being one such example. - - - - - 9a656a04 by Ben Gamari at 2024-03-06T13:39:39-05:00 ghc-experimental: Add dummy dependencies to work around #23942 This is a temporary measure to improve CI reliability until a proper solution is developed. Works around #23942. - - - - - 1e84b924 by Simon Peyton Jones at 2024-03-06T13:39:39-05:00 Three compile perf improvements with deep nesting These were changes are all triggered by #24471. 1. Make GHC.Core.Opt.SetLevels.lvlMFE behave better when there are many free variables. See Note [Large free-variable sets]. 2. Make GHC.Core.Opt.Arity.floatIn a bit lazier in its Cost argument. This benefits the common case where the ArityType turns out to be nullary. See Note [Care with nested expressions] 3. Make GHC.CoreToStg.Prep.cpeArg behave for deeply-nested expressions. See Note [Eta expansion of arguments in CorePrep] wrinkle (EA2). Compile times go down by up to 4.5%, and much more in artificial cases. (Geo mean of compiler/perf changes is -0.4%.) Metric Decrease: CoOpt_Read T10421 T12425 - - - - - c4b13113 by Hécate Moonlight at 2024-03-06T13:40:17-05:00 Use "module" instead of "library" when applicable in base haddocks - - - - - 9cd9efb4 by Vladislav Zavialov at 2024-03-07T13:01:54+03:00 Rephrase error message to say "visible arguments" (#24318) * Main change: make the error message generated by mkFunTysMsg more accurate by changing "value arguments" to "visible arguments". * Refactor: define a new type synonym VisArity and use it instead of Arity in a few places. It might be the case that there other places in the compiler that should talk about visible arguments rather than value arguments, but I haven't tried to find them all, focusing only on the error message reported in the ticket. - - - - - 4b6e76b5 by Patrick at 2024-03-07T22:09:30+08:00 fix haskell/haddock#24493, with module name introduced in hieAst The accompanies haddoc PR with GHC PR https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12153 Two things have been done: 1. Link is introduced to every `X` in `module X where`, since we introduce the module name to HieAst, 2. `{-# LINE 4 "hypsrc-test/src/PositionPragmas.hs" #-}` is added before the `module PositionPragmas where` in ` hypsrc-test/ref/src/PositionPragmas.html `.It ensures only a single hieAst for file `hypsrc-test/src/PositionPragmas.hs` is generated. - - - - - d523a6a7 by Ben Gamari at 2024-03-07T19:40:45-05:00 Bump array submodule - - - - - 7e55003c by Ben Gamari at 2024-03-07T19:40:45-05:00 Bump stm submodule - - - - - 32d337ef by Ben Gamari at 2024-03-07T19:40:45-05:00 Introduce exception context Here we introduce the `ExceptionContext` type and `ExceptionAnnotation` class, allowing dynamically-typed user-defined annotations to be attached to exceptions. CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/199 GHC Proposal: https://github.com/ghc-proposals/ghc-proposals/pull/330 - - - - - 39f3d922 by Ben Gamari at 2024-03-07T19:40:46-05:00 testsuite/interface-stability: Update documentation - - - - - fdea7ada by Ben Gamari at 2024-03-07T19:40:46-05:00 ghc-internal: comment formatting - - - - - 4fba42ef by Ben Gamari at 2024-03-07T19:40:46-05:00 compiler: Default and warn ExceptionContext constraints - - - - - 3886a205 by Ben Gamari at 2024-03-07T19:40:46-05:00 base: Introduce exception backtraces Here we introduce the `Backtraces` type and associated machinery for attaching these via `ExceptionContext`. These has a few compile-time regressions (`T15703` and `T9872d`) due to the additional dependencies in the exception machinery. As well, there is a surprisingly large regression in the `size_hello_artifact` test. This appears to be due to various `Integer` and `Read` bits now being reachable at link-time. I believe it should be possible to avoid this but I have accepted the change for now to get the feature merged. CLC Proposal: https://github.com/haskell/core-libraries-committee/issues/199 GHC Proposal: https://github.com/ghc-proposals/ghc-proposals/pull/330 Metric Increase: T15703 T9872d size_hello_artifact - - - - - 18c5409f by Ben Gamari at 2024-03-07T19:40:46-05:00 users guide: Release notes for exception backtrace work - - - - - f849c5fc by Ben Gamari at 2024-03-07T19:40:46-05:00 compiler: Don't show ExceptionContext of GhcExceptions Most GhcExceptions are user-facing errors and therefore the ExceptionContext has little value. Ideally we would enable it in the DEBUG compiler but I am leaving this for future work. - - - - - dc646e6f by Sylvain Henry at 2024-03-07T19:40:46-05:00 Disable T9930fail for the JS target (cf #19174) - - - - - bfc09760 by Alan Zimmerman at 2024-03-07T19:41:22-05:00 Update showAstData to honour blanking of AnnParen Also tweak rendering of SrcSpan to remove extra blank line. - - - - - 50454a29 by Ben Gamari at 2024-03-08T03:32:42-05:00 ghc-internal: Eliminate GHC.Internal.Data.Kind This was simply reexporting things from `ghc-prim`. Instead reexport these directly from `Data.Kind`. Also add build ordering dependency to work around #23942. - - - - - 38a4b6ab by Ben Gamari at 2024-03-08T03:33:18-05:00 rts: Fix SET_HDR initialization of retainer set This fixes a regression in retainer set profiling introduced by b0293f78cb6acf2540389e22bdda420d0ab874da. Prior to that commit the heap traversal word would be initialized by `SET_HDR` using `LDV_RECORD_CREATE`. However, the commit added a `doingLDVProfiling` check in `LDV_RECORD_CREATE`, meaning that this initialization no longer happened. Given that this initialization was awkwardly indirectly anyways, I have fixed this by explicitly initializating the heap traversal word to `NULL` in `SET_PROF_HDR`. This is equivalent to the previous behavior, but much more direct. Fixes #24513. - - - - - 635abccc by Ben Gamari at 2024-03-08T17:09:06-05:00 Bump ghc version to 9.10 - - - - - 2859a637 by Ben Gamari at 2024-03-08T18:26:47-05:00 base: Use strerror_r instead of strerror As noted by #24344, `strerror` is not necessarily thread-safe. Thankfully, POSIX.1-2001 has long offered `strerror_r`, which is safe to use. Fixes #24344. CLC discussion: https://github.com/haskell/core-libraries-committee/issues/249 - - - - - 5b934048 by Ben Gamari at 2024-03-08T18:50:12-05:00 Bump base upper bound - - - - - b30d134e by Ben Gamari at 2024-03-08T18:50:44-05:00 Testsuite output update - - - - - edb9bf77 by Jade at 2024-03-09T03:39:38-05:00 Error messages: Improve Error messages for Data constructors in type signatures. This patch improves the error messages from invalid type signatures by trying to guess what the user did and suggesting an appropriate fix. Partially fixes: #17879 - - - - - cfb197e3 by Patrick at 2024-03-09T03:40:15-05:00 HieAst: add module name #24493 The main purpose of this is to tuck the module name `xxx` in `module xxx where` into the hieAst. It should fix #24493. The following have been done: 1. Renamed and update the `tcg_doc_hdr :: Maybe (LHsDoc GhcRn)` to `tcg_hdr_info :: (Maybe (LHsDoc GhcRn), Maybe (XRec GhcRn ModuleName))` To store the located module name information. 2. update the `RenamedSource` and `RenamedStuff` with extra `Maybe (XRec GhcRn ModuleName)` located module name information. 3. add test `testsuite/tests/hiefile/should_compile/T24493.hs` to ensure the module name is added and update several relevent tests. 4. accompanied submodule haddoc test update MR in https://gitlab.haskell.org/ghc/haddock/-/merge_requests/53 - - - - - 2341d81e by Vaibhav Sagar at 2024-03-09T03:40:54-05:00 GHC.Utils.Binary: fix a couple of typos - - - - - 5580e1bd by Ben Gamari at 2024-03-09T03:41:30-05:00 rts: Drop .wasm suffix from .prof file names This replicates the behavior on Windows, where `Hi.exe` will produce profiling output named `Hi.prof` instead of `Hi.exe.prof`. While in the area I also fixed the extension-stripping logic, which incorrectly rewrote `Hi.exefoo` to `Hi.foo`. Closes #24515. - - - - - 259495ee by Cheng Shao at 2024-03-09T03:41:30-05:00 testsuite: drop exe extension from .hp & .prof filenames See #24515 for details. - - - - - c477a8d2 by Ben Gamari at 2024-03-09T03:42:05-05:00 rts/linker: Enable GOT support on all platforms There is nothing platform-dependent about our GOT implementation and GOT support is needed by `T24171` on i386. - - - - - 2e592857 by Vladislav Zavialov at 2024-03-09T03:42:41-05:00 Drop outdated comment on TcRnIllformedTypePattern This should have been done in 0f0c53a501b but I missed it. - - - - - c554b4da by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/CloneStack: Bounds check array write - - - - - 15c590a5 by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/CloneStack: Don't expose helper functions in header - - - - - e831ce31 by Ben Gamari at 2024-03-09T09:39:20-05:00 base: Move internals of GHC.InfoProv into GHC.InfoProv.Types Such that we can add new helpers into GHC.InfoProv.Types without breakage. - - - - - 6948e24d by Ben Gamari at 2024-03-09T09:39:20-05:00 rts: Lazily decode IPE tables Previously we would eagerly allocate `InfoTableEnt`s for each info table registered in the info table provenance map. However, this costs considerable memory and initialization time. Instead we now lazily decode these tables. This allows us to use one-third the memory *and* opens the door to taking advantage of sharing opportunities within a module. This required considerable reworking since lookupIPE now must be passed its result buffer. - - - - - 9204a04e by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/IPE: Don't expose helper in header - - - - - 308926ff by Ben Gamari at 2024-03-09T09:39:20-05:00 rts/IPE: Share module_name within a Node This allows us to shave a 64-bit word off of the packed IPE entry size. - - - - - bebdea05 by Ben Gamari at 2024-03-09T09:39:20-05:00 IPE: Expose unit ID in InfoTableProv Here we add the unit ID to the info table provenance structure. - - - - - 6519c9ad by Ben Gamari at 2024-03-09T09:39:35-05:00 rts: Refactor GHC.Stack.CloneStack.decode Don't allocate a Ptr constructor per frame. - - - - - ed0b69dc by Ben Gamari at 2024-03-09T09:39:35-05:00 base: Do not expose whereFrom# from GHC.Exts - - - - - 2b1faea9 by Vladislav Zavialov at 2024-03-09T17:38:21-05:00 docs: Update info on TypeAbstractions * Mention TypeAbstractions in 9.10.1-notes.rst * Set the status to "Experimental". * Add a "Since: GHC 9.x" comment to each section. - - - - - f8b88918 by Ben Gamari at 2024-03-09T21:21:46-05:00 ci-images: Bump Alpine image to bootstrap with 9.8.2 - - - - - 705e6927 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Mark T24171 as fragile due to #24512 I will fix this but not in time for 9.10.1-alpha1 - - - - - c74196e1 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Mark linker_unload_native as fragile In particular this fails on platforms without `dlinfo`. I plan to address this but not before 9.10.1-alpha1. - - - - - f4d87f7a by Ben Gamari at 2024-03-09T21:21:46-05:00 configure: Bump version to 9.10 - - - - - 88df9a5f by Ben Gamari at 2024-03-09T21:21:46-05:00 Bump transformers submodule to 0.6.1.1 - - - - - 8176d5e8 by Ben Gamari at 2024-03-09T21:21:46-05:00 testsuite: Increase ulimit for T18623 1 MByte was just too tight and failed intermittently on some platforms (e.g. CentOS 7). Bumping the limit to 8 MByte should provide sufficient headroom. Fixes #23139. - - - - - c74b38a3 by Ben Gamari at 2024-03-09T21:21:46-05:00 base: Bump version to 4.20.0.0 - - - - - b2937fc3 by Ben Gamari at 2024-03-09T21:21:46-05:00 ghc-internal: Set initial version at 9.1001.0 This provides PVP compliance while maintaining a clear correspondence between GHC releases and `ghc-internal` versions. - - - - - 4ae7d868 by Ben Gamari at 2024-03-09T21:21:46-05:00 ghc-prim: Bump version to 0.11.0 - - - - - 50798dc6 by Ben Gamari at 2024-03-09T21:21:46-05:00 template-haskell: Bump version to 2.22.0.0 - - - - - 8564f976 by Ben Gamari at 2024-03-09T21:21:46-05:00 base-exports: Accommodate spurious whitespace changes in 32-bit output It appears that this was - - - - - 9d4f0e98 by Ben Gamari at 2024-03-09T21:21:46-05:00 users-guide: Move exception backtrace relnotes to 9.10 This was previously mistakenly added to the GHC 9.8 release notes. - - - - - 145eae60 by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/rel_eng: Fix name of Rocky8 artifact - - - - - 39c2a630 by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/rel_eng: Fix path of generate_jobs_metadata - - - - - aed034de by Ben Gamari at 2024-03-09T21:21:46-05:00 gitlab/upload: Rework recompression The old `combine` approach was quite fragile due to use of filename globbing. Moreover, it didn't parallelize well. This refactoring makes the goal more obvious, parallelizes better, and is more robust. - - - - - 9bdf3586 by Ben Gamari at 2024-03-09T21:37:44-05:00 Merge branch 'ghc-9.10' into ghc-head - - - - - cec76981 by Ben Gamari at 2024-03-09T21:54:00-05:00 Bump GHC version to 9.11 - - - - - 4c59feb7 by Ben Gamari at 2024-03-09T22:15:01-05:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - dc207d06 by Ben Gamari at 2024-03-10T08:56:08-04:00 configure: Bump GHC version to 9.11 Bumps haddock submodule. - - - - - 8b2513e8 by Ben Gamari at 2024-03-11T01:20:03-04:00 rts/linker: Don't unload code when profiling is enabled The heap census may contain references (e.g. `Counter.identity`) to static data which must be available when the census is reported at the end of execution. Fixes #24512. - - - - - 7810b4c3 by Ben Gamari at 2024-03-11T01:20:03-04:00 rts/linker: Don't unload native objects when dlinfo isn't available To do so is unsafe as we have no way of identifying references to symbols provided by the object. Fixes #24513. Fixes #23993. - - - - - 0590764c by Ben Gamari at 2024-03-11T01:20:39-04:00 rel_eng/upload: Purge both $rel_name/ and $ver/ This is necessary for prereleases, where GHCup accesses the release via `$ver/` - - - - - b85a4631 by Brandon Chinn at 2024-03-12T19:25:56-04:00 Remove duplicate code normalising slashes - - - - - c91946f9 by Brandon Chinn at 2024-03-12T19:25:56-04:00 Simplify regexes with raw strings - - - - - 1a5f53c6 by Brandon Chinn at 2024-03-12T19:25:57-04:00 Don't normalize backslashes in characters - - - - - 7ea971d3 by Andrei Borzenkov at 2024-03-12T19:26:32-04:00 Fix compiler crash caused by implicit RHS quantification in type synonyms (#24470) - - - - - 39f3ac3e by Cheng Shao at 2024-03-12T19:27:11-04:00 Revert "compiler: make genSym use C-based atomic increment on non-JS 32-bit platforms" This reverts commit 615eb855416ce536e02ed935ecc5a6f25519ae16. It was originally intended to fix #24449, but it was merely sweeping the bug under the rug. 3836a110577b5c9343915fd96c1b2c64217e0082 has properly fixed the fragile test, and we no longer need the C version of genSym. Furthermore, the C implementation causes trouble when compiling with clang that targets i386 due to alignment warning and libatomic linking issue, so it makes sense to revert it. - - - - - e6bfb85c by Cheng Shao at 2024-03-12T19:27:11-04:00 compiler: fix out-of-bound memory access of genSym on 32-bit This commit fixes an unnoticed out-of-bound memory access of genSym on 32-bit. ghc_unique_inc is 32-bit sized/aligned on 32-bit platforms, but we mistakenly treat it as a Word64 pointer in genSym, and therefore will accidentally load 2 garbage higher bytes, or with a small but non-zero chance, overwrite something else in the data section depends on how the linker places the data segments. This regression was introduced in !11802 and fixed here. - - - - - 77171cd1 by Ben Orchard at 2024-03-14T09:00:40-04:00 Note mutability of array and address access primops Without an understanding of immutable vs. mutable memory, the index primop family have a potentially non-intuitive type signature: indexOffAddr :: Addr# -> Int# -> a readOffAddr :: Addr# -> Int# -> State# d -> (# State# d, a #) indexOffAddr# might seem like a free generality improvement, which it certainly is not! This change adds a brief note on mutability expectations for most index/read/write access primops. - - - - - 7da7f8f6 by Alan Zimmerman at 2024-03-14T09:01:15-04:00 EPA: Fix regression discarding comments in contexts Closes #24533 - - - - - 73be65ab by Fendor at 2024-03-19T01:42:53-04:00 Fix sharing of 'IfaceTyConInfo' during core to iface type translation During heap analysis, we noticed that during generation of 'mi_extra_decls' we have lots of duplicates for the instances: * `IfaceTyConInfo NotPromoted IfaceNormalTyCon` * `IfaceTyConInfo IsPromoted IfaceNormalTyCon` which should be shared instead of duplicated. This duplication increased the number of live bytes by around 200MB while loading the agda codebase into GHCi. These instances are created during `CoreToIface` translation, in particular `toIfaceTyCon`. The generated core looks like: toIfaceTyCon = \ tc_sjJw -> case $wtoIfaceTyCon tc_sjJw of { (# ww_sjJz, ww1_sjNL, ww2_sjNM #) -> IfaceTyCon ww_sjJz (IfaceTyConInfo ww1_sjNL ww2_sjNM) } whichs removes causes the sharing to work propery. Adding explicit sharing, with NOINLINE annotations, changes the core to: toIfaceTyCon = \ tc_sjJq -> case $wtoIfaceTyCon tc_sjJq of { (# ww_sjNB, ww1_sjNC #) -> IfaceTyCon ww_sjNB ww1_sjNC } which looks much more like sharing is happening. We confirmed via ghc-debug that all duplications were eliminated and the number of live bytes are noticeably reduced. - - - - - bd8209eb by Alan Zimmerman at 2024-03-19T01:43:28-04:00 EPA: Address more 9.10.1-alpha1 regressions from recent changes Closes #24533 Hopefully for good this time - - - - - 31bf85ee by Fendor at 2024-03-19T14:48:08-04:00 Escape multiple arguments in the settings file Uses responseFile syntax. The issue arises when GHC is installed on windows into a location that has a space, for example the user name is 'Fake User'. The $topdir will also contain a space, consequentially. When we resolve the top dir in the string `-I$topdir/mingw/include`, then `words` will turn this single argument into `-I/C/Users/Fake` and `User/.../mingw/include` which trips up the flag argument parser of various tools such as gcc or clang. We avoid this by escaping the $topdir before replacing it in `initSettngs`. Additionally, we allow to escape spaces and quotation marks for arguments in `settings` file. Add regression test case to count the number of options after variable expansion and argument escaping took place. Additionally, we check that escaped spaces and double quotation marks are correctly parsed. - - - - - f45f700e by Matthew Pickering at 2024-03-19T14:48:44-04:00 Read global package database from settings file Before this patch, the global package database was always assumed to be in libdir </> package.conf.d. This causes issues in GHC's build system because there are sometimes situations where the package database you need to use is not located in the same place as the settings file. * The stage1 compiler needs to use stage1 libraries, so we should set "Global Package DB" for the stage1 compiler to the stage1 package database. * Stage 2 cross compilers need to use stage2 libraries, so likewise, we should set the package database path to `_build/stage2/lib/` * The normal situation is where the stage2 compiler uses stage1 libraries. Then everything lines up. * When installing we have rearranged everything so that the settings file and package database line up properly, so then everything should continue to work as before. In this case we set the relative package db path to `package.conf.d`, so it resolves the same as before. * ghc-pkg needs to be modified as well to look in the settings file fo the package database rather than assuming the global package database location relative to the lib folder. * Cabal/cabal-install will work correctly because they query the global package database using `--print-global-package-db`. A reasonable question is why not generate the "right" settings files in the right places in GHC's build system. In order to do this you would need to engineer wrappers for all executables to point to a specific libdir. There are also situations where the same package db is used by two different compilers with two different settings files (think stage2 cross compiler and stage3 compiler). In short, this 10 line patch allows for some reasonable simplifications in Hadrian at very little cost to anything else. Fixes #24502 - - - - - 4c8f1794 by Matthew Pickering at 2024-03-19T14:48:44-04:00 hadrian: Remove stage1 testsuite wrappers logic Now instead of producing wrappers which pass the global package database argument to ghc and ghc-pkg, we write the location of the correct package database into the settings file so you can just use the intree compiler directly. - - - - - da0d8ba5 by Matthew Craven at 2024-03-19T14:49:20-04:00 Remove unused ghc-internal module "GHC.Internal.Constants" - - - - - b56d2761 by Matthew Craven at 2024-03-19T14:49:20-04:00 CorePrep: Rework lowering of BigNat# literals Don't use bigNatFromWord#, because that's terrible: * We shouldn't have to traverse a linked list at run-time to build a BigNat# literal. That's just silly! * The static List object we have to create is much larger than the actual BigNat#'s contents, bloating code size. * We have to read the corresponding interface file, which causes un-tracked implicit dependencies. (#23942) Instead, encode them into the appropriate platform-dependent sequence of bytes, and generate code that copies these bytes at run-time from an Addr# literal into a new ByteArray#. A ByteArray# literal would be the correct thing to generate, but these are not yet supported; see also #17747. Somewhat surprisingly, this change results in a slight reduction in compiler allocations, averaging around 0.5% on ghc's compiler performance tests, including when compiling programs that contain no bignum literals to begin with. The specific cause of this has not been investigated. Since this lowering no longer reads the interface file for GHC.Num.BigNat, the reasoning in Note [Depend on GHC.Num.Integer] is obsoleted. But the story of un-tracked built-in dependencies remains complex, and Note [Tracking dependencies on primitives] now exists to explain this complexity. Additionally, many empty imports have been modified to refer to this new note and comply with its guidance. Several empty imports necessary for other reasons have also been given brief explanations. Metric Decrease: MultiLayerModulesTH_OneShot - - - - - 349ea330 by Fendor at 2024-03-19T14:50:00-04:00 Eliminate thunk in 'IfaceTyCon' Heap analysis showed that `IfaceTyCon` retains a thunk to `IfaceTyConInfo`, defeating the sharing of the most common instances of `IfaceTyConInfo`. We make sure the indirection is removed by adding bang patterns to `IfaceTyCon`. Experimental results on the agda code base, where the `mi_extra_decls` were read from disk: Before this change, we observe around 8654045 instances of: `IfaceTyCon[Name,THUNK_1_0]` But these thunks almost exclusively point to a shared value! Forcing the thunk a little bit more, leads to `ghc-debug` reporting: `IfaceTyCon[Name:Name,IfaceTyConInfo]` and a noticeable reduction of live bytes (on agda ~10%). - - - - - 594bee0b by Krzysztof Gogolewski at 2024-03-19T14:50:36-04:00 Minor misc cleanups - GHC.HsToCore.Foreign.JavaScript: remove dropRuntimeRepArgs; boxed tuples don't take RuntimeRep args - GHC.HsToCore.Foreign.Call: avoid partial pattern matching - GHC.Stg.Unarise: strengthen the assertion; we can assert that non-rubbish literals are unary rather than just non-void - GHC.Tc.Gen.HsType: make sure the fsLit "literal" rule fires - users_guide/using-warnings.rst: remove -Wforall-identifier, now deprecated and does nothing - users_guide/using.rst: fix formatting - andy_cherry/test.T: remove expect_broken_for(23272...), 23272 is fixed The rest are simple cleanups. - - - - - cf55a54b by Ben Gamari at 2024-03-19T14:51:12-04:00 mk/relpath: Fix quoting Previously there were two instances in this script which lacked proper quoting. This resulted in `relpath` invocations in the binary distribution Makefile producing incorrect results on Windows, leading to confusing failures from `sed` and the production of empty package registrations. Fixes #24538. - - - - - 5ff88389 by Bryan Richter at 2024-03-19T14:51:48-04:00 testsuite: Disable T21336a on wasm - - - - - 60023351 by Ben Gamari at 2024-03-19T22:33:10-04:00 hadrian/bindist: Eliminate extraneous `dirname` invocation Previously we would call `dirname` twice per installed library file. We now instead reuse this result. This helps appreciably on Windows, where processes are quite expensive. - - - - - 616ac300 by Ben Gamari at 2024-03-19T22:33:10-04:00 hadrian: Package mingw toolchain in expected location This fixes #24525, a regression due to 41cbaf44a6ab5eb9fa676d65d32df8377898dc89. Specifically, GHC expects to find the mingw32 toolchain in the binary distribution root. However, after this patch it was packaged in the `lib/` directory. - - - - - de9daade by Ben Gamari at 2024-03-19T22:33:11-04:00 gitlab/rel_eng: More upload.sh tweaks - - - - - 1dfe12db by Ben Gamari at 2024-03-19T22:33:11-04:00 rel_eng: Drop dead prepare_docs codepath - - - - - dd2d748b by Ben Gamari at 2024-03-19T22:33:11-04:00 rel_env/recompress_all: unxz before recompressing Previously we would rather compress the xz *again*, before in addition compressing it with the desired scheme. Fixes #24545. - - - - - 9d936c57 by Ben Gamari at 2024-03-19T22:33:11-04:00 mk-ghcup-metadata: Fix directory of testsuite tarball As reported in #24546, the `dlTest` artifact should be extracted into the `testsuite` directory. - - - - - 6d398066 by Ben Gamari at 2024-03-19T22:33:11-04:00 ghcup-metadata: Don't populate dlOutput unless necessary ghcup can apparently infer the output name of an artifact from its URL. Consequently, we should only include the `dlOutput` field when it would differ from the filename of `dlUri`. Fixes #24547. - - - - - 576f8b7e by Zubin Duggal at 2024-03-19T22:33:46-04:00 Revert "Apply shellcheck suggestion to SUBST_TOOLDIR" This reverts commit c82770f57977a2b5add6e1378f234f8dd6153392. The shellcheck suggestion is spurious and results in SUBST_TOOLDIR being a no-op. `set` sets positional arguments for bash, but we want to set the variable given as the first autoconf argument. Fixes #24542 Metric decreases because the paths in the settings file are now shorter, so we allocate less when we read the settings file. ------------------------- Metric Decrease: T12425 T13035 T9198 ------------------------- - - - - - cdfe6e01 by Fendor at 2024-03-19T22:34:22-04:00 Compact serialisation of IfaceAppArgs In #24563, we identified that IfaceAppArgs serialisation tags each cons cell element with a discriminator byte. These bytes add up quickly, blowing up interface files considerably when '-fwrite-if-simplified-core' is enabled. We compact the serialisation by writing out the length of 'IfaceAppArgs', followed by serialising the elements directly without any discriminator byte. This improvement can decrease the size of some interface files by up to 35%. - - - - - 97a2bb1c by Simon Peyton Jones at 2024-03-20T17:11:29+00:00 Expand untyped splices in tcPolyExprCheck Fixes #24559 - - - - - 5f275176 by Alan Zimmerman at 2024-03-20T22:44:12-04:00 EPA: Clean up Exactprint helper functions a bit - Introduce a helper lens to compose on `EpAnn a` vs `a` versions - Rename some prime versions of functions back to non-prime They were renamed during the rework - - - - - da2a10ce by Vladislav Zavialov at 2024-03-20T22:44:48-04:00 Type operators in promoteOccName (#24570) Type operators differ from term operators in that they are lexically classified as (type) constructors, not as (type) variables. Prior to this change, promoteOccName did not account for this difference, causing a scoping issue that affected RequiredTypeArguments. type (!@#) = Bool f = idee (!@#) -- Not in scope: ‘!@#’ (BUG) Now we have a special case in promoteOccName to account for this. - - - - - 247fc0fa by Preetham Gujjula at 2024-03-21T10:19:18-04:00 docs: Remove mention of non-existent Ord instance for Complex The documentation for Data.Complex says that the Ord instance for Complex Float is deficient, but there is no Ord instance for Complex a. The Eq instance for Complex Float is similarly deficient, so we use that as an example instead. - - - - - 6fafc51e by Andrei Borzenkov at 2024-03-21T10:19:54-04:00 Fix TH handling in `pat_to_type_pat` function (#24571) There was missing case for `SplicePat` in `pat_to_type_at` function, hence patterns with splicing that checked against `forall->` doesn't work properly because they fall into the "illegal pattern" case. Code example that is now accepted: g :: forall a -> () g $([p| a |]) = () - - - - - 52072f8e by Sylvain Henry at 2024-03-21T21:01:59-04:00 Type-check default declarations before deriving clauses (#24566) See added Note and #24566. Default declarations must be type-checked before deriving clauses. - - - - - 7dfdf3d9 by Sylvain Henry at 2024-03-21T21:02:40-04:00 Lexer: small perf changes - Use unsafeChr because we know our values to be valid - Remove some unnecessary use of `ord` (return Word8 values directly) - - - - - 864922ef by Sylvain Henry at 2024-03-21T21:02:40-04:00 JS: fix some comments - - - - - 3e0b2b1f by Sebastian Graf at 2024-03-21T21:03:16-04:00 Simplifier: Re-do dependency analysis in abstractFloats (#24551) In #24551, we abstracted a string literal binding over a type variable, triggering a CoreLint error when that binding floated to top-level. The solution implemented in this patch fixes this by re-doing dependency analysis on a simplified recursive let binding that is about to be type abstracted, in order to find the minimal set of type variables to abstract over. See wrinkle (AB5) of Note [Floating and type abstraction] for more details. Fixes #24551 - - - - - 8a8ac65a by Matthew Craven at 2024-03-23T00:20:52-04:00 Improve toInteger @Word32 on 64-bit platforms On 64-bit platforms, every Word32 fits in an Int, so we can convert to Int# without having to perform the overflow check integerFromWord# uses internally. - - - - - 0c48f2b9 by Apoorv Ingle at 2024-03-23T00:21:28-04:00 Fix for #24552 (see testcase T24552) Fixes for a bug in desugaring pattern synonyms matches, introduced while working on on expanding `do`-blocks in #18324 The `matchWrapper` unecessarily (and incorrectly) filtered out the default wild patterns in a match. Now the wild pattern alternative is simply ignored by the pm check as its origin is `Generated`. The current code now matches the expected semantics according to the language spec. - - - - - b72705e9 by Simon Peyton Jones at 2024-03-23T00:22:04-04:00 Print more info about kinds in error messages This fixes #24553, where GHC unhelpfully said error: [GHC-83865] • Expected kind ‘* -> * -> *’, but ‘Foo’ has kind ‘* -> * -> *’ See Note [Showing invisible bits of types in error messages] - - - - - 8f7cfc7e by Tristan Cacqueray at 2024-03-23T00:22:44-04:00 docs: remove the don't use float hint This hint is outdated, ``Complex Float`` are now specialised, and the heap space suggestion needs more nuance so it should be explained in the unboxed/storable array documentation. - - - - - 5bd8ed53 by Andreas Klebinger at 2024-03-23T16:18:33-04:00 NCG: Fix a bug in jump shortcutting. When checking if a jump has more than one destination account for the possibility of some jumps not being representable by a BlockId. We do so by having isJumpishInstr return a `Maybe BlockId` where Nothing represents non-BlockId jump destinations. Fixes #24507 - - - - - 8d67f247 by Ben Gamari at 2024-03-23T16:19:09-04:00 docs: Drop old release notes, add for 9.12.1 - - - - - 7db8c992 by Cheng Shao at 2024-03-25T13:45:46-04:00 rts: fix clang compilation on aarch64 This patch fixes function prototypes in ARMOutlineAtomicsSymbols.h which causes "error: address argument to atomic operation must be a pointer to _Atomic type" when compiling with clang on aarch64. - - - - - 237194ce by Sylvain Henry at 2024-03-25T13:46:27-04:00 Lexer: fix imports for Alex 3.5.1 (#24583) - - - - - 810660b7 by Cheng Shao at 2024-03-25T22:19:16-04:00 libffi-tarballs: bump libffi-tarballs submodule to libffi 3.4.6 This commit bumps the libffi-tarballs submodule to libffi 3.4.6, which includes numerous upstream libffi fixes, especially https://github.com/libffi/libffi/issues/760. - - - - - d2ba41e8 by Alan Zimmerman at 2024-03-25T22:19:51-04:00 EPA: do not duplicate comments in signature RHS - - - - - 32a8103f by Rodrigo Mesquita at 2024-03-26T21:16:12-04:00 configure: Use LDFLAGS when trying linkers A user may configure `LDFLAGS` but not `LD`. When choosing a linker, we will prefer `ldd`, then `ld.gold`, then `ld.bfd` -- however, we have to check for a working linker. If either of these fail, we try the next in line. However, we were not considering the `$LDFLAGS` when checking if these linkers worked. So we would pick a linker that does not support the current $LDFLAGS and fail further down the line when we used that linker with those flags. Fixes #24565, where `LDFLAGS=-Wl,-z,pack-relative-relocs` is not supported by `ld.gold` but that was being picked still. - - - - - bf65a7c3 by Rodrigo Mesquita at 2024-03-26T21:16:48-04:00 bindist: Clean xattrs of bin and lib at configure time For issue #21506, we started cleaning the extended attributes of binaries and libraries from the bindist *after* they were installed to workaround notarisation (#17418), as part of `make install`. However, the `ghc-toolchain` binary that is now shipped with the bindist must be run at `./configure` time. Since we only cleaned the xattributes of the binaries and libs after they were installed, in some situations users would be unable to run `ghc-toolchain` from the bindist, failing at configure time (#24554). In this commit we move the xattr cleaning logic to the configure script. Fixes #24554 - - - - - cfeb70d3 by Rodrigo Mesquita at 2024-03-26T21:17:24-04:00 Revert "NCG: Fix a bug in jump shortcutting." This reverts commit 5bd8ed53dcefe10b72acb5729789e19ceb22df66. Fixes #24586 - - - - - 13223f6d by Serge S. Gulin at 2024-03-27T07:28:51-04:00 JS: `h$rts_isProfiled` is removed from `profiling` and left its version at `rts/js/config.js` - - - - - 0acfe391 by Alan Zimmerman at 2024-03-27T07:29:27-04:00 EPA: Do not extend declaration range for trailine zero len semi The lexer inserts virtual semicolons having zero width. Do not use them to extend the list span of items in a list. - - - - - cd0fb82f by Alan Zimmerman at 2024-03-27T19:33:08+00:00 EPA: Fix FamDecl range The span was incorrect if opt_datafam_kind_sig was empty - - - - - f8f384a8 by Ben Gamari at 2024-03-29T01:23:03-04:00 Fix type of _get_osfhandle foreign import Fixes #24601. - - - - - 00d3ecf0 by Alan Zimmerman at 2024-03-29T12:19:10+00:00 EPA: Extend StringLiteral range to include trailing commas This goes slightly against the exact printing philosophy where trailing decorations should be in an annotation, but the practicalities of adding it to the WarningTxt environment, and the problems caused by deviating do not make a more principles approach worthwhile. - - - - - efab3649 by brandon s allbery kf8nh at 2024-03-31T20:04:01-04:00 clarify Note [Preproccesing invocations] - - - - - c8a4c050 by Ben Gamari at 2024-04-02T12:50:35-04:00 rts: Fix TSAN_ENABLED CPP guard This should be `#if defined(TSAN_ENABLED)`, not `#if TSAN_ENABLED`, lest we suffer warnings. - - - - - e91dad93 by Cheng Shao at 2024-04-02T12:50:35-04:00 rts: fix errors when compiling with TSAN This commit fixes rts compilation errors when compiling with TSAN: - xxx_FENCE macros are redefined and trigger CPP warnings. - Use SIZEOF_W. WORD_SIZE_IN_BITS is provided by MachDeps.h which Cmm.h doesn't include by default. - - - - - a9ab9455 by Cheng Shao at 2024-04-02T12:50:35-04:00 rts: fix clang-specific errors when compiling with TSAN This commit fixes clang-specific rts compilation errors when compiling with TSAN: - clang doesn't have -Wtsan flag - Fix prototype of ghc_tsan_* helper functions - __tsan_atomic_* functions aren't clang built-ins and sanitizer/tsan_interface_atomic.h needs to be included - On macOS, TSAN runtime library is libclang_rt.tsan_osx_dynamic.dylib, not libtsan. -fsanitize-thread as a link-time flag will take care of linking the TSAN runtime library anyway so remove tsan as an rts extra library - - - - - 865bd717 by Cheng Shao at 2024-04-02T12:50:35-04:00 compiler: fix github link to __tsan_memory_order in a comment - - - - - 07cb627c by Cheng Shao at 2024-04-02T12:50:35-04:00 ci: improve TSAN CI jobs - Run TSAN jobs with +thread_sanitizer_cmm which enables Cmm instrumentation as well. - Run TSAN jobs in deb12 which ships gcc-12, a reasonably recent gcc that @bgamari confirms he's using in #GHC:matrix.org. Ideally we should be using latest clang release for latest improvements in sanitizers, though that's left as future work. - Mark TSAN jobs as manual+allow_failure in validate pipelines. The purpose is to demonstrate that we have indeed at least fixed building of TSAN mode in CI without blocking the patch to land, and once merged other people can begin playing with TSAN using their own dev setups and feature branches. - - - - - a1c18c7b by Andrei Borzenkov at 2024-04-02T12:51:11-04:00 Merge tc_infer_hs_type and tc_hs_type into one function using ExpType philosophy (#24299, #23639) This patch implements refactoring which is a prerequisite to updating kind checking of type patterns. This is a huge simplification of the main worker that checks kind of HsType. It also fixes the issues caused by previous code duplication, e.g. that we didn't add module finalizers from splices in inference mode. - - - - - 817e8936 by Rodrigo Mesquita at 2024-04-02T20:13:05-04:00 th: Hide the Language.Haskell.TH.Lib.Internal module from haddock Fixes #24562 - - - - - b36ee57b by Sylvain Henry at 2024-04-02T20:13:46-04:00 JS: reenable h$appendToHsString optimization (#24495) The optimization introducing h$appendToHsString wasn't kicking in anymore (while it did in 9.8.1) because of the changes introduced in #23270 (7e0c8b3bab30). This patch reenables the optimization by matching on case-expression, as done in Cmm for unpackCString# standard thunks. The test is also T24495 added in the next commits (two commits for ease of backporting to 9.8). - - - - - 527616e9 by Sylvain Henry at 2024-04-02T20:13:46-04:00 JS: fix h$appendToHsString implementation (#24495) h$appendToHsString needs to wrap its argument in an updatable thunk to behave like unpackAppendCString#. Otherwise if a SingleEntry thunk is passed, it is stored as-is in a CONS cell, making the resulting list impossible to deepseq (forcing the thunk doesn't update the contents of the CONS cell)! The added test checks that the optimization kicks in and that h$appendToHsString works as intended. Fix #24495 - - - - - faa30b41 by Simon Peyton Jones at 2024-04-02T20:14:22-04:00 Deal with duplicate tyvars in type declarations GHC was outright crashing before this fix: #24604 - - - - - e0b0c717 by Simon Peyton Jones at 2024-04-02T20:14:58-04:00 Try using MCoercion in exprIsConApp_maybe This is just a simple refactor that makes exprIsConApp_maybe a little bit more direct, simple, and efficient. Metrics: compile_time/bytes allocated geo. mean -0.1% minimum -2.0% maximum -0.0% Not a big gain, but worthwhile given that the code is, if anything, easier to grok. - - - - - 15f4d867 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Initial ./configure support for selecting I/O managers In this patch we just define new CPP vars, but don't yet use them or replace the existing approach. That will follow. The intention here is that every I/O manager can be enabled/disabled at GHC build time (subject to some constraints). More than one I/O manager can be enabled to be built. At least one I/O manager supporting the non-threaded RTS must be enabled as well as at least one supporting the non-threaded RTS. The I/O managers enabled here will become the choices available at runtime at RTS startup (in later patches). The choice can be made with RTS flags. There are separate sets of choices for the threaded and non-threaded RTS ways, because most I/O managers are specific to these ways. Furthermore we must establish a default I/O manager for the threaded and non-threaded RTS. Most I/O managers are platform-specific so there are checks to ensure each one can be enabled on the platform. Such checks are also where (in future) any system dependencies (e.g. libraries) can be checked. The output is a set of CPP flags (in the mk/config.h file), with one flag per named I/O manager: * IOMGR_BUILD_<name> : which ones should be built (some) * IOMGR_DEFAULT_NON_THREADED_<name> : which one is default (exactly one) * IOMGR_DEFAULT_THREADED_<name> : which one is default (exactly one) and a set of derived flags in IOManager.h * IOMGR_ENABLED_<name> : enabled for the current RTS way Note that IOMGR_BUILD_<name> just says that an I/O manager will be built for _some_ RTS way (i.e. threaded or non-threaded). The derived flags IOMGR_ENABLED_<name> in IOManager.h say if each I/O manager is enabled in the "current" RTS way. These are the ones that can be used for conditional compilation of the I/O manager code. Co-authored-by: Pi Delport <pi at well-typed.com> - - - - - 85b0f87a by Duncan Coutts at 2024-04-03T01:27:17-04:00 Change the handling of the RTS flag --io-manager= Now instead of it being just used on Windows to select between the WinIO vs the MIO or Win32-legacy I/O managers, it is now used on all platforms for selecting the I/O manager to use. Right now it remains the case that there is only an actual choice on Windows, but that will change later. Document the --io-manager flag in the user guide. This change is also reflected in the RTS flags types in the base library. Deprecate the export of IoSubSystem from GHC.RTS.Flags with a message to import it from GHC.IO.Subsystem. The way the 'IoSubSystem' is detected also changes. Instead of looking at the RTS flag, there is now a C bool global var in the RTS which gets set on startup when the I/O manager is selected. This bool var says whether the selected I/O manager classifies as "native" on Windows, which in practice means the WinIO I/O manager has been selected. Similarly, the is_io_mng_native_p RTS helper function is re-implemented in terms of the selected I/O manager, rather than based on the RTS flags. We do however remove the ./configure --native-io-manager flag because we're bringing the WinIO/MIO/Win32-legacy choice under the new general scheme for selecting I/O managers, and that new scheme involves no ./configure time user choices, just runtime RTS flag choices. - - - - - 1a8f020f by Duncan Coutts at 2024-04-03T01:27:17-04:00 Convert {init,stop,exit}IOManager to switch style Rather than ad-hoc cpp conitionals on THREADED_RTS and mingw32_HOST_OS, we use a style where we switch on the I/O manager impl, with cases for each I/O manager impl. - - - - - a5bad3d2 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Split up the CapIOManager content by I/O manager Using the new IOMGR_ENABLED_<name> CPP defines. - - - - - 1d36e609 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Convert initIOManagerAfterFork and wakeupIOManager to switch style - - - - - c2f26f36 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move most of waitRead#/Write# from cmm to C Moves it into the IOManager.c where we can follow the new pattern of switching on the selected I/O manager. - - - - - 457705a8 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move most of the delay# impl from cmm to C Moves it into the IOManager.c where we can follow the new pattern of switching on the selected I/O manager. Uses a new IOManager API: syncDelay, following the naming convention of sync* for thread-synchronous I/O & timer/delay operations. As part of porting from cmm to C, we maintain the rule that the why_blocked gets accessed using load acquire and store release atomic memory operations. There was one exception to this rule: in the delay# primop cmm code on posix (not win32), the why_blocked was being updated using a store relaxed, not a store release. I've no idea why. In this convesion I'm playing it safe here and using store release consistently. - - - - - e93058e0 by Duncan Coutts at 2024-04-03T01:27:18-04:00 insertIntoSleepingQueue is no longer public No longer defined in IOManager.h, just a private function in IOManager.c. Since it is no longer called from cmm code, just from syncDelay. It ought to get moved further into the select() I/O manager impl, rather than living in IOManager.c. On the other hand appendToIOBlockedQueue is still called from cmm code in the win32-legacy I/O manager primops async{Read,Write}#, and it is also used by the select() I/O manager. Update the CPP and comments to reflect this. - - - - - 60ce9910 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move anyPendingTimeoutsOrIO impl from .h to .c The implementation is eventually going to need to use more private things, which will drag in unwanted includes into IOManager.h, so it's better to move the impl out of the header file and into the .c file, at the slight cost of it no longer being inline. At the same time, change to the "switch (iomgr_type)" style. - - - - - f70b8108 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Take a simpler approach to gcc warnings in IOManager.c We have lots of functions with conditional implementations for different I/O managers. Some functions, for some I/O managers, naturally have implementations that do nothing or barf. When only one such I/O manager is enabled then the whole function implementation will have an implementation that does nothing or barfs. This then results in warnings from gcc that parameters are unused, or that the function should be marked with attribute noreturn (since barf does not return). The USED_IF_THREADS trick for fine-grained warning supression is fine for just two cases, but an equivalent here would need USED_IF_THE_ONLY_ENABLED_IOMGR_IS_X_OR_Y which would have combinitorial blowup. So we take a coarse grained approach and simply disable these two warnings for the whole file. So we use a GCC pragma, with its handy push/pop support: #pragma GCC diagnostic push #pragma GCC diagnostic ignored "-Wsuggest-attribute=noreturn" #pragma GCC diagnostic ignored "-Wunused-parameter" ... #pragma GCC diagnostic pop - - - - - b48805b9 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add a new trace class for the iomanager It makes sense now for it to be separate from the scheduler class of tracers. Enabled with +RTS -Do. Document the -Do debug flag in the user guide. - - - - - f0c1f862 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Have the throwTo impl go via (new) IOManager APIs rather than directly operating on the IO manager's data structures. Specifically, when thowing an async exception to a thread that is blocked waiting for I/O or waiting for a timer, then we want to cancel that I/O waiting or cancel the timer. Currently this is done directly in removeFromQueues() in RaiseAsync.c. We want it to go via proper APIs both for modularity but also to let us support multiple I/O managers. So add sync{IO,Delay}Cancel, which is the cancellation for the corresponding sync{IO,Delay}. The implementations of these use the usual "switch (iomgr_type)" style. - - - - - 4f9e9c4e by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move awaitEvent into a proper IOManager API and have the scheduler use it. Previously the scheduler calls awaitEvent directly, and awaitEvent is implemented directly in the RTS I/O managers (select, win32). This relies on the old scheme where there's a single active I/O manager for each platform and RTS way. We want to move that to go via an API in IOManager.{h,c} which can then call out to the active I/O manager. Also take the opportunity to split awaitEvent into two. The existing awaitEvent has a bool wait parameter, to say if the call should be blocking or non-blocking. We split this into two separate functions: pollCompletedTimeoutsOrIO and awaitCompletedTimeoutsOrIO. We split them for a few reasons: they have different post-conditions (specifically the await version is supposed to guarantee that there are threads runnable when it completes). Secondly, it is also anticipated that in future I/O managers the implementations of the two cases will be simpler if they are separated. - - - - - 5ad4b30f by Duncan Coutts at 2024-04-03T01:27:18-04:00 Rename awaitEvent in select and win32 I/O managers These are now just called from IOManager.c and are the per-I/O manager backend impls (whereas previously awaitEvent was the entry point). Follow the new naming convention in the IOManager.{h,c} of awaitCompletedTimeoutsOrIO, with the I/O manager's name as a suffix: so awaitCompletedTimeoutsOrIO{Select,Win32}. - - - - - d30c6bc6 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Tidy up a couple things in Select.{h,c} Use the standard #include {Begin,End}Private.h style rather than RTS_PRIVATE on individual decls. And conditionally build the code for the select I/O manager based on the new CPP IOMGR_ENABLED_SELECT rather than on THREADED_RTS. - - - - - 4161f516 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add an IOManager API for scavenging TSO blocked_info When the GC scavenges a TSO it needs to scavenge the tso->blocked_info but the blocked_info is a big union and what lives there depends on the two->why_blocked, which for I/O-related reasons is something that in principle is the responsibility of the I/O manager and not the GC. So the right thing to do is for the GC to ask the I/O manager to sscavenge the blocked_info if it encounters any I/O-related why_blocked reasons. So we add scavengeTSOIOManager in IOManager.{h,c} with the usual style. Now as it happens, right now, there is no special scavenging to do, so the implementation of scavengeTSOIOManager is a fancy no-op. That's because the select I/O manager uses only the fd and target members, which are not GC pointers, and the win32-legacy I/O manager _ought_ to be using GC-managed heap objects for the StgAsyncIOResult but it is actually usingthe C heap, so again no GC pointers. If the win32-legacy were doing this more sensibly, then scavengeTSOIOManager would be the right place to do the GC magic. Future I/O managers will need GC heap objects in the tso->blocked_info and will make use of this functionality. - - - - - 94a87d21 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add I/O manager API notifyIOManagerCapabilitiesChanged Used in setNumCapabilities. It only does anything for MIO on Posix. Previously it always invoked Haskell code, but that code only did anything on non-Windows (and non-JS), and only threaded. That currently effectively means the MIO I/O manager on Posix. So now it only invokes it for the MIO Posix case. - - - - - 3be6d591 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Select an I/O manager early in RTS startup We need to select the I/O manager to use during startup before the per-cap I/O manager initialisation. - - - - - aaa294d0 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Make struct CapIOManager be fully opaque Provide an opaque (forward) definition in Capability.h (since the cap contains a *CapIOManager) and then only provide a full definition in a new file IOManagerInternals.h. This new file is only supposed to be included by the IOManager implementation, not by its users. So that means IOManager.c and individual I/O manager implementations. The posix/Signals.c still needs direct access, but that should be eliminated. Anything that needs direct access either needs to be clearly part of an I/O manager (e.g. the sleect() one) or go via a proper API. - - - - - 877a2a80 by Duncan Coutts at 2024-04-03T01:27:18-04:00 The select() I/O manager does have some global initialisation It's just to make sure an exception CAF is a GC root. - - - - - 9c51473b by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add tracing for the main I/O manager actions Using the new tracer class. Note: The unconditional definition of showIOManager should be compatible with the debugTrace change in 7c7d1f6. Co-authored-by: Pi Delport <pi at well-typed.com> - - - - - c7d3e3a3 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Include the default I/O manager in the +RTS --info output Document the extra +RTS --info output in the user guide - - - - - 8023bad4 by Duncan Coutts at 2024-04-03T01:27:18-04:00 waitRead# / waitWrite# do not work for win32-legacy I/O manager Previously it was unclear that they did not work because the code path was shared with other I/O managers (in particular select()). Following the code carefully shows that what actually happens is that the calling thread would block forever: the thread will be put into the blocked queue, but no other action is scheduled that will ever result in it getting unblocked. It's better to just fail loudly in case anyone accidentally calls it, also it's less confusing code. - - - - - 83a74d20 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Conditionally ignore some GCC warnings Some GCC versions don't know about some warnings, and they complain that we're ignoring unknown warnings. So we try to ignore the warning based on the GCC version. - - - - - 1adc6fa4 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Accept changes to base-exports All the changes are in fact not changes at all. Previously, the IoSubSystem data type was defined in GHC.RTS.Flags and exported from both GHC.RTS.Flags and GHC.IO.SubSystem. Now, the data type is defined in GHC.IO.SubSystem and still exported from both modules. Therefore, the same exports and same instances are still available from both modules. But the base-exports records only the defining module, and so it looks like a change when it is fully compatible. Related: we do add a deprecation to the export of the type via GHC.RTS.Flags, telling people to use the export from GHC.IO.SubSystem. Also the sort order for some unrelated Show instances changed. No idea why. The same changes apply in the other versions, with a few more changes due to sort order weirdness. - - - - - 8d950968 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Accept metric decrease in T12227 I can't think of any good reason that anything in this MR should have changed the number of allocations, up or down. (Yes this is an empty commit.) Metric Decrease: T12227 - - - - - e869605e by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Several improvements to the handling of coercions * Make `mkSymCo` and `mkInstCo` smarter Fixes #23642 * Fix return role of `SelCo` in the coercion optimiser. Fixes #23617 * Make the coercion optimiser `opt_trans_rule` work better for newtypes Fixes #23619 - - - - - 1efd0714 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 FloatOut: improve floating for join point See the new Note [Floating join point bindings]. * Completely get rid of the complicated join_ceiling nonsense, which I have never understood. * Do not float join points at all, except perhaps to top level. * Some refactoring around wantToFloat, to treat Rec and NonRec more uniformly - - - - - 9c00154d by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Improve eta-expansion through call stacks See Note [Eta expanding through CallStacks] in GHC.Core.Opt.Arity This is a one-line change, that fixes an inconsistency - || isCallStackPredTy ty + || isCallStackPredTy ty || isCallStackTy ty - - - - - 95a9a172 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Spelling, layout, pretty-printing only - - - - - bdf1660f by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Improve exprIsConApp_maybe a little Eliminate a redundant case at birth. This sometimes reduces Simplifier iterations. See Note [Case elim in exprIsConApp_maybe]. - - - - - 609cd32c by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Inline GHC.HsToCore.Pmc.Solver.Types.trvVarInfo When exploring compile-time regressions after meddling with the Simplifier, I discovered that GHC.HsToCore.Pmc.Solver.Types.trvVarInfo was very delicately balanced. It's a small, heavily used, overloaded function and it's important that it inlines. By a fluke it was before, but at various times in my journey it stopped doing so. So I just added an INLINE pragma to it; no sense in depending on a delicately-balanced fluke. - - - - - ae24c9bc by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Slight improvement in WorkWrap Ensure that WorkWrap preserves lambda binders, in case of join points. Sadly I have forgotten why I made this change (it was while I was doing a lot of meddling in the Simplifier, but * it does no harm, * it is slightly more efficient, and * presumably it made something better! Anyway I have kept it in a separate commit. - - - - - e9297181 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Use named record fields for the CastIt { ... } data constructor This is a pure refactor - - - - - b4581e23 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Remove a long-commented-out line Pure refactoring - - - - - e026bdf2 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Simplifier improvements This MR started as: allow the simplifer to do more in one pass, arising from places I could see the simplifier taking two iterations where one would do. But it turned into a larger project, because these changes unexpectedly made inlining blow up, especially join points in deeply-nested cases. The main changes are below. There are also many new or rewritten Notes. Avoiding simplifying repeatedly ~~~~~~~~~~~~~~~ See Note [Avoiding simplifying repeatedly] * The SimplEnv now has a seInlineDepth field, which says how deep in unfoldings we are. See Note [Inline depth] in Simplify.Env. Currently used only for the next point: avoiding repeatedly simplifying coercions. * Avoid repeatedly simplifying coercions. see Note [Avoid re-simplifying coercions] in Simplify.Iteration As you'll see from the Note, this makes use of the seInlineDepth. * Allow Simplify.Iteration.simplAuxBind to inline used-once things. This is another part of Note [Post-inline for single-use things], and is really good for reducing simplifier iterations in situations like case K e of { K x -> blah } wher x is used once in blah. * Make GHC.Core.SimpleOpt.exprIsConApp_maybe do some simple case elimination. Note [Case elim in exprIsConApp_maybe] * Improve the case-merge transformation: - Move the main code to `GHC.Core.Utils.mergeCaseAlts`, to join `filterAlts` and friends. See Note [Merge Nested Cases] in GHC.Core.Utils. - Add a new case for `tagToEnum#`; see wrinkle (MC3). - Add a new case to look through join points: see wrinkle (MC4) postInlineUnconditionally ~~~~~~~~~~~~~~~~~~~~~~~~~ * Allow Simplify.Utils.postInlineUnconditionally to inline variables that are used exactly once. See Note [Post-inline for single-use things]. * Do not postInlineUnconditionally join point, ever. Doing so does not reduce allocation, which is the main point, and with join points that are used a lot it can bloat code. See point (1) of Note [Duplicating join points] in GHC.Core.Opt.Simplify.Iteration. * Do not postInlineUnconditionally a strict (demanded) binding. It will not allocate a thunk (it'll turn into a case instead) so again the main point of inlining it doesn't hold. Better to check per-call-site. * Improve occurrence analyis for bottoming function calls, to help postInlineUnconditionally. See Note [Bottoming function calls] in GHC.Core.Opt.OccurAnal Inlining generally ~~~~~~~~~~~~~~~~~~ * In GHC.Core.Opt.Simplify.Utils.interestingCallContext, use RhsCtxt NonRecursive (not BoringCtxt) for a plain-seq case. See Note [Seq is boring] Also, wrinkle (SB1), inline in that `seq` context only for INLINE functions (UnfWhen guidance). * In GHC.Core.Opt.Simplify.Utils.interestingArg, - return ValueArg for OtherCon [c1,c2, ...], but - return NonTrivArg for OtherCon [] This makes a function a little less likely to inline if all we know is that the argument is evaluated, but nothing else. * isConLikeUnfolding is no longer true for OtherCon {}. This propagates to exprIsConLike. Con-like-ness has /positive/ information. Join points ~~~~~~~~~~~ * Be very careful about inlining join points. See these two long Notes Note [Duplicating join points] in GHC.Core.Opt.Simplify.Iteration Note [Inlining join points] in GHC.Core.Opt.Simplify.Inline * When making join points, don't do so if the join point is so small it will immediately be inlined; check uncondInlineJoin. * In GHC.Core.Opt.Simplify.Inline.tryUnfolding, improve the inlining heuristics for join points. In general we /do not/ want to inline join points /even if they are small/. See Note [Duplicating join points] GHC.Core.Opt.Simplify.Iteration. But sometimes we do: see Note [Inlining join points] in GHC.Core.Opt.Simplify.Inline; and the new `isBetterUnfoldingThan` function. * Do not add an unfolding to a join point at birth. This is a tricky one and has a long Note [Do not add unfoldings to join points at birth] It shows up in two places - In `mkDupableAlt` do not add an inlining - (trickier) In `simplLetUnfolding` don't add an unfolding for a fresh join point I am not fully satisifed with this, but it works and is well documented. * In GHC.Core.Unfold.sizeExpr, make jumps small, so that we don't penalise having a non-inlined join point. Performance changes ~~~~~~~~~~~~~~~~~~~ * Binary sizes fall by around 2.6%, according to nofib. * Compile times improve slightly. Here are the figures over 1%. I investiate the biggest differnce in T18304. It's a very small module, just a few hundred nodes. The large percentage difffence is due to a single function that didn't quite inline before, and does now, making code size a bit bigger. I decided gains outweighed the losses. Metrics: compile_time/bytes allocated (changes over +/- 1%) ------------------------------------------------ CoOpt_Singletons(normal) -9.2% GOOD LargeRecord(normal) -23.5% GOOD MultiComponentModulesRecomp(normal) +1.2% MultiLayerModulesTH_OneShot(normal) +4.1% BAD PmSeriesS(normal) -3.8% PmSeriesV(normal) -1.5% T11195(normal) -1.3% T12227(normal) -20.4% GOOD T12545(normal) -3.2% T12707(normal) -2.1% GOOD T13253(normal) -1.2% T13253-spj(normal) +8.1% BAD T13386(normal) -3.1% GOOD T14766(normal) -2.6% GOOD T15164(normal) -1.4% T15304(normal) +1.2% T15630(normal) -8.2% T15630a(normal) NEW T15703(normal) -14.7% GOOD T16577(normal) -2.3% GOOD T17516(normal) -39.7% GOOD T18140(normal) +1.2% T18223(normal) -17.1% GOOD T18282(normal) -5.0% GOOD T18304(normal) +10.8% BAD T18923(normal) -2.9% GOOD T1969(normal) +1.0% T19695(normal) -1.5% T20049(normal) -12.7% GOOD T21839c(normal) -4.1% GOOD T3064(normal) -1.5% T3294(normal) +1.2% BAD T4801(normal) +1.2% T5030(normal) -15.2% GOOD T5321Fun(normal) -2.2% GOOD T6048(optasm) -16.8% GOOD T783(normal) -1.2% T8095(normal) -6.0% GOOD T9630(normal) -4.7% GOOD T9961(normal) +1.9% BAD WWRec(normal) -1.4% info_table_map_perf(normal) -1.3% parsing001(normal) +1.5% geo. mean -2.0% minimum -39.7% maximum +10.8% * Runtimes generally improve. In the testsuite perf/should_run gives: Metrics: runtime/bytes allocated ------------------------------------------ Conversions(normal) -0.3% T13536a(optasm) -41.7% GOOD T4830(normal) -0.1% haddock.Cabal(normal) -0.1% haddock.base(normal) -0.1% haddock.compiler(normal) -0.1% geo. mean -0.8% minimum -41.7% maximum +0.0% * For runtime, nofib is a better test. The news is mostly good. Here are the number more than +/- 0.1%: # bytes allocated ==========================++========== imaginary/digits-of-e1 || -14.40% imaginary/digits-of-e2 || -4.41% imaginary/paraffins || -0.17% imaginary/rfib || -0.15% imaginary/wheel-sieve2 || -0.10% real/compress || -0.47% real/fluid || -0.10% real/fulsom || +0.14% real/gamteb || -1.47% real/gg || -0.20% real/infer || +0.24% real/pic || -0.23% real/prolog || -0.36% real/scs || -0.46% real/smallpt || +4.03% shootout/k-nucleotide || -20.23% shootout/n-body || -0.42% shootout/spectral-norm || -0.13% spectral/boyer2 || -3.80% spectral/constraints || -0.27% spectral/hartel/ida || -0.82% spectral/mate || -20.34% spectral/para || +0.46% spectral/rewrite || +1.30% spectral/sphere || -0.14% ==========================++========== geom mean || -0.59% real/smallpt has a huge nest of local definitions, and I could not pin down a reason for a regression. But there are three big wins! Metric Decrease: CoOpt_Singletons LargeRecord T12227 T12707 T13386 T13536a T14766 T15703 T16577 T17516 T18223 T18282 T18923 T21839c T20049 T5321Fun T5030 T6048 T8095 T9630 T783 Metric Increase: MultiLayerModulesTH_OneShot T13253-spj T18304 T18698a T9961 T3294 - - - - - 27db3c5e by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Testsuite message changes from simplifier improvements - - - - - 271a7812 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Account for bottoming functions in OccurAnal This fixes #24582, a small but long-standing bug - - - - - 0fde229f by Ben Gamari at 2024-04-04T07:04:58-04:00 testsuite: Introduce template-haskell-exports test - - - - - 0c4a9686 by Luite Stegeman at 2024-04-04T07:05:39-04:00 Update correct counter in bumpTickyAllocd - - - - - 5f085d3a by Fendor at 2024-04-04T14:47:33-04:00 Replace `SizedSeq` with `FlatBag` for flattened structure LinkedLists are notoriously memory ineffiecient when all we do is traversing a structure. As 'UnlinkedBCO' has been identified as a data structure that impacts the overall memory usage of GHCi sessions, we avoid linked lists and prefer flattened structure for storing. We introduce a new memory efficient representation of sequential elements that has special support for the cases: * Empty * Singleton * Tuple Elements This improves sharing in the 'Empty' case and avoids the overhead of 'Array' until its constant overhead is justified. - - - - - 82cfe10c by Fendor at 2024-04-04T14:47:33-04:00 Compact FlatBag array representation `Array` contains three additional `Word`'s we do not need in `FlatBag`. Move `FlatBag` to `SmallArray`. Expand the API of SmallArray by `sizeofSmallArray` and add common traversal functions, such as `mapSmallArray` and `foldMapSmallArray`. Additionally, allow users to force the elements of a `SmallArray` via `rnfSmallArray`. - - - - - 36a75b80 by Andrei Borzenkov at 2024-04-04T14:48:10-04:00 Change how invisible patterns represented in haskell syntax and TH AST (#24557) Before this patch: data ArgPat p = InvisPat (LHsType p) | VisPat (LPat p) With this patch: data Pat p = ... | InvisPat (LHsType p) ... And the same transformation in the TH land. The rest of the changes is just updating code to handle new AST and writing tests to check if it is possible to create invalid states using TH. Metric Increase: MultiLayerModulesTH_OneShot - - - - - 28009fbc by Matthew Pickering at 2024-04-04T14:48:46-04:00 Fix off by one error in seekBinNoExpand and seekBin - - - - - 9b9e031b by Ben Gamari at 2024-04-04T21:30:08-04:00 compiler: Allow more types in GHCForeignImportPrim For many, many years `GHCForeignImportPrim` has suffered from the rather restrictive limitation of not allowing any non-trivial types in arguments or results. This limitation was justified by the code generator allegely barfing in the presence of such types. However, this restriction appears to originate well before the NCG rewrite and the new NCG does not appear to have any trouble with such types (see the added `T24598` test). Lift this restriction. Fixes #24598. - - - - - 1324b862 by Alan Zimmerman at 2024-04-04T21:30:44-04:00 EPA: Use EpaLocation not SrcSpan in ForeignDecls This allows us to update them for makeDeltaAst in ghc-exactprint - - - - - 19883a23 by Alan Zimmerman at 2024-04-05T16:58:17-04:00 EPA: Use EpaLocation for RecFieldsDotDot So we can update it to a delta position in makeDeltaAst if needed. - - - - - e8724327 by Matthew Pickering at 2024-04-05T16:58:53-04:00 Remove accidentally committed test.hs - - - - - 88cb3e10 by Fendor at 2024-04-08T09:03:34-04:00 Avoid UArray when indexing is not required `UnlinkedBCO`'s can occur many times in the heap. Each `UnlinkedBCO` references two `UArray`'s but never indexes them. They are only needed to encode the elements into a `ByteArray#`. The three words for the lower bound, upper bound and number of elements are essentially unused, thus we replace `UArray` with a wrapper around `ByteArray#`. This saves us up to three words for each `UnlinkedBCO`. Further, to avoid re-allocating these words for `ResolvedBCO`, we repeat the procedure for `ResolvedBCO` and add custom `Binary` and `Show` instances. For example, agda's repl session has around 360_000 UnlinkedBCO's, so avoiding these three words is already saving us around 8MB residency. - - - - - f2cc1107 by Fendor at 2024-04-08T09:04:11-04:00 Never UNPACK `FastMutInt` for counting z-encoded `FastString`s In `FastStringTable`, we count the number of z-encoded FastStrings that exist in a GHC session. We used to UNPACK the counters to not waste memory, but live retainer analysis showed that we allocate a lot of `FastMutInt`s, retained by `mkFastZString`. We lazily compute the `FastZString`, only incrementing the counter when the `FastZString` is forced. The function `mkFastStringWith` calls `mkZFastString` and boxes the `FastMutInt`, leading to the following core: mkFastStringWith = \ mk_fs _ -> = case stringTable of { FastStringTable _ n_zencs segments# _ -> ... case ((mk_fs (I# ...) (FastMutInt n_zencs)) `cast` <Co:2> :: ...) ... Marking this field as `NOUNPACK` avoids this reboxing, eliminating the allocation of a fresh `FastMutInt` on every `FastString` allocation. - - - - - c6def949 by Matthew Pickering at 2024-04-08T16:06:51-04:00 Force in_multi to avoid retaining entire hsc_env - - - - - fbb91a63 by Fendor at 2024-04-08T16:06:51-04:00 Eliminate name thunk in declaration fingerprinting Thunk analysis showed that we have about 100_000 thunks (in agda and `-fwrite-simplified-core`) pointing to the name of the name decl. Forcing this thunk fixes this issue. The thunk created here is retained by the thunk created by forkM, it is better to eagerly force this because the result (a `Name`) is already retained indirectly via the `IfaceDecl`. - - - - - 3b7b0c1c by Alan Zimmerman at 2024-04-08T16:07:27-04:00 EPA: Use EpaLocation in WarningTxt This allows us to use an EpDelta if needed when using makeDeltaAst. - - - - - 12b997df by Alan Zimmerman at 2024-04-08T16:07:27-04:00 EPA: Move DeltaPos and EpaLocation' into GHC.Types.SrcLoc This allows us to use a NoCommentsLocation for the possibly trailing comma location in a StringLiteral. This in turn allows us to correctly roundtrip via makeDeltaAst. - - - - - 868c8a78 by Fendor at 2024-04-09T08:51:50-04:00 Prefer packed representation for CompiledByteCode As there are many 'CompiledByteCode' objects alive during a GHCi session, representing its element in a more packed manner improves space behaviour at a minimal cost. When running GHCi on the agda codebase, we find around 380 live 'CompiledByteCode' objects. Packing their respective 'UnlinkedByteCode' can save quite some pointers. - - - - - be3bddde by Alan Zimmerman at 2024-04-09T08:52:26-04:00 EPA: Capture all comments in a ClassDecl Hopefully the final fix needed for #24533 - - - - - 3d0806fc by Jade at 2024-04-10T05:39:53-04:00 Validate -main-is flag using parseIdentifier Fixes #24368 - - - - - dd530bb7 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 rts: free error message before returning Fixes a memory leak in rts/linker/PEi386.c - - - - - e008a19a by Alexis King at 2024-04-10T05:40:29-04:00 linker: Avoid linear search when looking up Haskell symbols via dlsym See the primary Note [Looking up symbols in the relevant objects] for a more in-depth explanation. When dynamically loading a Haskell symbol (typical when running a splice or GHCi expression), before this commit we would search for the symbol in all dynamic libraries that were loaded. However, this could be very inefficient when too many packages are loaded (which can happen if there are many package dependencies) because the time to lookup the would be linear in the number of packages loaded. This commit drastically improves symbol loading performance by introducing a mapping from units to the handles of corresponding loaded dlls. These handles are returned by dlopen when we load a dll, and can then be used to look up in a specific dynamic library. Looking up a given Name is now much more precise because we can get lookup its unit in the mapping and lookup the symbol solely in the handles of the dynamic libraries loaded for that unit. In one measurement, the wait time before the expression was executed went from +-38 seconds down to +-2s. This commit also includes Note [Symbols may not be found in pkgs_loaded], explaining the fallback to the old behaviour in case no dll can be found in the unit mapping for a given Name. Fixes #23415 Co-authored-by: Rodrigo Mesquita (@alt-romes) - - - - - dcfaa190 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 rts: Make addDLL a wrapper around loadNativeObj Rewrite the implementation of `addDLL` as a wrapper around the more principled `loadNativeObj` rts linker function. The latter should be preferred while the former is preserved for backwards compatibility. `loadNativeObj` was previously only available on ELF platforms, so this commit further refactors the rts linker to transform loadNativeObj_ELF into loadNativeObj_POSIX, which is available in ELF and MachO platforms. The refactor made it possible to remove the `dl_mutex` mutex in favour of always using `linker_mutex` (rather than a combination of both). Lastly, we implement `loadNativeObj` for Windows too. - - - - - 12931698 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 Use symbol cache in internal interpreter too This commit makes the symbol cache that was used by the external interpreter available for the internal interpreter too. This follows from the analysis in #23415 that suggests the internal interpreter could benefit from this cache too, and that there is no good reason not to have the cache for it too. It also makes it a bit more uniform to have the symbol cache range over both the internal and external interpreter. This commit also refactors the cache into a function which is used by both `lookupSymbol` and also by `lookupSymbolInDLL`, extending the caching logic to `lookupSymbolInDLL` too. - - - - - dccd3ea1 by Ben Gamari at 2024-04-10T05:40:29-04:00 testsuite: Add test for lookupSymbolInNativeObj - - - - - 1b1a92bd by Alan Zimmerman at 2024-04-10T05:41:05-04:00 EPA: Remove unnecessary XRec in CompleteMatchSig The XRec for [LIdP pass] is not needed for exact printing, remove it. - - - - - 6e18ce2b by Ben Gamari at 2024-04-12T08:16:09-04:00 users-guide: Clarify language extension documentation Over the years the users guide's language extension documentation has gone through quite a few refactorings. In the process some of the descriptions have been rendered non-sensical. For instance, the description of `NoImplicitPrelude` actually describes the semantics of `ImplicitPrelude`. To fix this we: * ensure that all extensions are named in their "positive" sense (e.g. `ImplicitPrelude` rather than `NoImplicitPrelude`). * rework the documentation to avoid flag-oriented wording like "enable" and "disable" * ensure that the polarity of the documentation is consistent with reality. Fixes #23895. - - - - - a933aff3 by Zubin Duggal at 2024-04-12T08:16:45-04:00 driver: Make `checkHomeUnitsClosed` faster The implementation of `checkHomeUnitsClosed` was traversing every single path in the unit dependency graph - this grows exponentially and quickly grows to be infeasible on larger unit dependency graphs. Instead we replace this with a faster implementation which follows from the specificiation of the closure property - there is a closure error if there are units which are both are both (transitively) depended upon by home units and (transitively) depend on home units, but are not themselves home units. To compute the set of units required for closure, we first compute the closure of the unit dependency graph, then the transpose of this closure, and find all units that are reachable from the home units in the transpose of the closure. - - - - - 23c3e624 by Andreas Klebinger at 2024-04-12T08:17:21-04:00 RTS: Emit warning when -M < -H Fixes #24487 - - - - - d23afb8c by Ben Gamari at 2024-04-12T08:17:56-04:00 testsuite: Add broken test for CApiFFI with -fprefer-bytecode See #24634. - - - - - a4bb3a51 by Ben Gamari at 2024-04-12T08:18:32-04:00 base: Deprecate GHC.Pack As proposed in #21461. Closes #21540. - - - - - 55eb8c98 by Ben Gamari at 2024-04-12T08:19:08-04:00 ghc-internal: Fix mentions of ghc-internal in deprecation warnings Closes #24609. - - - - - b0fbd181 by Ben Gamari at 2024-04-12T08:19:44-04:00 rts: Implement set_initial_registers for AArch64 Fixes #23680. - - - - - 14c9ec62 by Ben Gamari at 2024-04-12T08:20:20-04:00 ghcup-metadata: Use Debian 9 binaries on Ubuntu 16, 17 Closes #24646. - - - - - 35a1621e by Ben Gamari at 2024-04-12T08:20:55-04:00 Bump unix submodule to 2.8.5.1 Closes #24640. - - - - - a1c24df0 by Finley McIlwaine at 2024-04-12T08:21:31-04:00 Correct default -funfolding-use-threshold in docs - - - - - 0255d03c by Oleg Grenrus at 2024-04-12T08:22:07-04:00 FastString is a __Modified__ UTF-8 - - - - - c3489547 by Matthew Pickering at 2024-04-12T13:13:44-04:00 rts: Improve tracing message when nursery is resized It is sometimes more useful to know how much bigger or smaller the nursery got when it is resized. In particular I am trying to investigate situations where we end up with fragmentation due to the nursery (#24577) - - - - - 5e4f4ba8 by Simon Peyton Jones at 2024-04-12T13:14:20-04:00 Don't generate wrappers for `type data` constructors with StrictData Previously, the logic for checking if a data constructor needs a wrapper or not would take into account whether the constructor's fields have explicit strictness (e.g., `data T = MkT !Int`), but the logic would _not_ take into account whether `StrictData` was enabled. This meant that something like `type data T = MkT Int` would incorrectly generate a wrapper for `MkT` if `StrictData` was enabled, leading to the horrible errors seen in #24620. To fix this, we disable generating wrappers for `type data` constructors altogether. Fixes #24620. Co-authored-by: Ryan Scott <ryan.gl.scott at gmail.com> - - - - - dbdf1995 by Alex Mason at 2024-04-15T15:28:26+10:00 Implements MO_S_Mul2 and MO_U_Mul2 using the UMULH, UMULL and SMULH instructions for AArch64 Also adds a test for MO_S_Mul2 - - - - - 42bd0407 by Teo Camarasu at 2024-04-16T20:06:39-04:00 Make template-haskell a stage1 package Promoting template-haskell from a stage0 to a stage1 package means that we can much more easily refactor template-haskell. We implement this by duplicating the in-tree `template-haskell`. A new `template-haskell-next` library is autogenerated to mirror `template-haskell` `stage1:ghc` to depend on the new interface of the library including the `Binary` instances without adding an explicit dependency on `template-haskell`. This is controlled by the `bootstrap-th` cabal flag When building `template-haskell` modules as part of this vendoring we do not have access to quote syntax, so we cannot use variable quote notation (`'Just`). So we either replace these with hand-written `Name`s or hide the code behind CPP. We can remove the `th_hack` from hadrian, which was required when building stage0 packages using the in-tree `template-haskell` library. For more details see Note [Bootstrapping Template Haskell]. Resolves #23536 Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> Co-Authored-By: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 3d973e47 by Ben Gamari at 2024-04-16T20:07:15-04:00 Bump parsec submodule to 3.1.17.0 - - - - - 9d38bfa0 by Simon Peyton Jones at 2024-04-16T20:07:51-04:00 Clone CoVars in CorePrep This MR addresses #24463. It's all explained in the new Note [Cloning CoVars and TyVars] - - - - - 0fe2b410 by Andreas Klebinger at 2024-04-16T20:08:27-04:00 NCG: Fix a bug where we errounously removed a required jump instruction. Add a new method to the Instruction class to check if we can eliminate a jump in favour of fallthrough control flow. Fixes #24507 - - - - - 9f99126a by Teo Camarasu at 2024-04-16T20:09:04-04:00 Fix documentation preview from doc-tarball job - Include all the .html files and assets in the job artefacts - Include all the .pdf files in the job artefacts - Mark the artefact as an "exposed" artefact meaning it turns up in the UI. Resolves #24651 - - - - - 3a0642ea by Ben Gamari at 2024-04-16T20:09:39-04:00 rts: Ignore EINTR while polling in timerfd itimer implementation While the RTS does attempt to mask signals, it may be that a foreign library unmasks them. This previously caused benign warnings which we now ignore. See #24610. - - - - - 9a53cd3f by Alan Zimmerman at 2024-04-16T20:10:15-04:00 EPA: Add additional comments field to AnnsModule This is used in exact printing to store comments coming after the `where` keyword but before any comments allocated to imports or decls. It is used in ghc-exactprint, see https://github.com/alanz/ghc-exactprint/commit/44bbed311fd8f0d053053fef195bf47c17d34fa7 - - - - - e5c43259 by Bryan Richter at 2024-04-16T20:10:51-04:00 Remove unrunnable FreeBSD CI jobs FreeBSD runner supply is inelastic. Currently there is only one, and it's unavailable because of a hardware issue. - - - - - 914eb49a by Ben Gamari at 2024-04-16T20:11:27-04:00 rel-eng: Fix mktemp usage in recompress-all We need a temporary directory, not a file. - - - - - f30e4984 by Teo Camarasu at 2024-04-16T20:12:03-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 - - - - - d7a3d6b5 by Ben Gamari at 2024-04-16T20:12:39-04:00 template-haskell: Declare TH.Lib.Internal as not-home Rather than `hide`. Closes #24659. - - - - - 5eaa46e7 by Matthew Pickering at 2024-04-19T02:14:55-04:00 testsuite: Rename isCross() predicate to needsTargetWrapper() isCross() was a misnamed because it assumed that all cross targets would provide a target wrapper, but the two most common cross targets (javascript, wasm) don't need a target wrapper. Therefore we rename this predicate to `needsTargetWrapper()` so situations in the testsuite where we can check whether running executables requires a target wrapper or not. - - - - - 55a9d699 by Simon Peyton Jones at 2024-04-19T02:15:32-04:00 Do not float HNFs out of lambdas This MR adjusts SetLevels so that it is less eager to float a HNF (lambda or constructor application) out of a lambda, unless it gets to top level. Data suggests that this change is a small net win: * nofib bytes-allocated falls by -0.09% (but a couple go up) * perf/should_compile bytes-allocated falls by -0.5% * perf/should_run bytes-allocated falls by -0.1% See !12410 for more detail. When fiddling elsewhere, I also found that this patch had a huge positive effect on the (very delicate) test perf/should_run/T21839r But that improvement doesn't show up in this MR by itself. Metric Decrease: MultiLayerModulesRecomp T15703 parsing001 - - - - - f0701585 by Alan Zimmerman at 2024-04-19T02:16:08-04:00 EPA: Fix comments in mkListSyntaxTy0 Also extend the test to confirm. Addresses #24669, 1 of 4 - - - - - b01c01d4 by Serge S. Gulin at 2024-04-19T02:16:51-04:00 JS: set image `x86_64-linux-deb11-emsdk-closure` for build - - - - - c90c6039 by Alan Zimmerman at 2024-04-19T02:17:27-04:00 EPA: Provide correct span for PatBind And remove unused parameter in checkPatBind Contributes to #24669 - - - - - bee54c24 by Krzysztof Gogolewski at 2024-04-19T11:13:00+02:00 Update quantification order following GHC haskell/haddock#23764 - - - - - 2814eb89 by Ben Gamari at 2024-04-19T18:57:05+02:00 hypsrc-test: Fix output of PositionPragmas.html - - - - - 26036f96 by Alan Zimmerman at 2024-04-19T13:11:08-04:00 EPA: Fix span for PatBuilderAppType Include the location of the prefix @ in the span for InVisPat. Also removes unnecessary annotations from HsTP. Contributes to #24669 - - - - - dba03aab by Matthew Craven at 2024-04-19T13:11:44-04:00 testsuite: Give the pre_cmd for mhu-perf more time - - - - - d31fbf6c by Krzysztof Gogolewski at 2024-04-19T21:04:09-04:00 Fix quantification order for a `op` b and a %m -> b Fixes #23764 Implements https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0640-tyop-quantification-order.rst Updates haddock submodule. - - - - - 385cd1c4 by Sebastian Graf at 2024-04-19T21:04:45-04:00 Make `seq#` a magic Id and inline it in CorePrep (#24124) We can save much code and explanation in Tag Inference and StgToCmm by making `seq#` a known-key Magic Id in `GHC.Internal.IO` and inline this definition in CorePrep. See the updated `Note [seq# magic]`. I also implemented a new `Note [Flatten case-bind]` to get better code for otherwise nested case scrutinees. I renamed the contructors of `ArgInfo` to use an `AI` prefix in order to resolve the clash between `type CpeApp = CoreExpr` and the data constructor of `ArgInfo`, as well as fixed typos in `Note [CorePrep invariants]`. Fixes #24252 and #24124. - - - - - 275e41a9 by Jade at 2024-04-20T11:10:40-04:00 Put the newline after errors instead of before them This mainly has consequences for GHCi but also slightly alters how the output of GHC on the commandline looks. Fixes: #22499 - - - - - dd339c7a by Teo Camarasu at 2024-04-20T11:11:16-04:00 Remove unecessary stage0 packages Historically quite a few packages had to be stage0 as they depended on `template-haskell` and that was stage0. In #23536 we made it so that was no longer the case. This allows us to remove a bunch of packages from this list. A few still remain. A new version of `Win32` is required by `semaphore-compat`. Including `Win32` in the stage0 set requires also including `filepath` because otherwise Hadrian's dependency logic gets confused. Once our boot compiler has a newer version of `Win32` all of these will be able to be dropped. Resolves #24652 - - - - - 2f8e3a25 by Alan Zimmerman at 2024-04-20T11:11:52-04:00 EPA: Avoid duplicated comments in splice decls Contributes to #24669 - - - - - c70b9ddb by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: fix typos and namings (fixes #24602) You may noted that I've also changed term of ``` , global "h$vt_double" ||= toJExpr IntV ``` See "IntV" and ``` WaitReadOp -> \[] [fd] -> pure $ PRPrimCall $ returnS (app "h$waidRead" [fd]) ``` See "h$waidRead" - - - - - 3db54f9b by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: trivial checks for variable presence (fixes #24602) - - - - - 777f108f by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: fs module imported twice (by emscripten and by ghc-internal). ghc-internal import wrapped in a closure to prevent conflict with emscripten (fixes #24602) Better solution is to use some JavaScript module system like AMD, CommonJS or even UMD. It will be investigated at other issues. At first glance we should try UMD (See https://github.com/umdjs/umd) - - - - - a45a5712 by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: thread.js requires h$fds and h$fdReady to be declared for static code analysis, minimal code copied from GHCJS (fixes #24602) I've just copied some old pieces of GHCJS from publicly available sources (See https://github.com/Taneb/shims/blob/a6dd0202dcdb86ad63201495b8b5d9763483eb35/src/io.js#L607). Also I didn't put details to h$fds. I took minimal and left only its object initialization: `var h$fds = {};` - - - - - ad90bf12 by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: heap and stack overflows reporting defined as js hard failure (fixes #24602) These errors were treated as a hard failure for browser application. The fix is trivial: just throw error. - - - - - 5962fa52 by Serge S. Gulin at 2024-04-21T16:33:44+03:00 JS: Stubs for code without actual implementation detected by Google Closure Compiler (fixes #24602) These errors were fixed just by introducing stubbed functions with throw for further implementation. - - - - - a0694298 by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Add externs to linker (fixes #24602) After enabling jsdoc and built-in google closure compiler types I was needed to deal with the following: 1. Define NodeJS-environment types. I've just copied minimal set of externs from semi-official repo (see https://github.com/externs/nodejs/blob/6c6882c73efcdceecf42e7ba11f1e3e5c9c041f0/v8/nodejs.js#L8). 2. Define Emscripten-environment types: `HEAP8`. Emscripten already provides some externs in our code but it supposed to be run in some module system. And its definitions do not work well in plain bundle. 3. We have some functions which purpose is to add to functions some contextual information via function properties. These functions should be marked as `modifies` to let google closure compiler remove calls if these functions are not used actually by call graph. Such functions are: `h$o`, `h$sti`, `h$init_closure`, `h$setObjInfo`. 4. STG primitives such as registries and stuff from `GHC.StgToJS`. `dXX` properties were already present at externs generator function but they are started from `7`, not from `1`. This message is related: `// fixme does closure compiler bite us here?` - - - - - e58bb29f by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: added both tests: for size and for correctness (fixes #24602) By some reason MacOS builds add to stderr messages like: Ignoring unexpected archive entry: __.SYMDEF ... However I left stderr to `/dev/null` for compatibility with linux CI builds. - - - - - 909f3a9c by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Disable js linker warning for empty symbol table to make js tests running consistent across environments - - - - - 83eb10da by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Add special preprocessor for js files due of needing to keep jsdoc comments (fixes #24602) Our js files have defined google closure compiler types at jsdoc entries but these jsdoc entries are removed by cpp preprocessor. I considered that reusing them in javascript-backend would be a nice thing. Right now haskell processor uses `-traditional` option to deal with comments and `//` operators. But now there are following compiler options: `-C` and `-CC`. You can read about them at GCC (see https://gcc.gnu.org/onlinedocs/gcc/Preprocessor-Options.html#index-CC) and CLang (see https://clang.llvm.org/docs/ClangCommandLineReference.html#cmdoption-clang-CC). It seems that `-CC` works better for javascript jsdoc than `-traditional`. At least it leaves `/* ... */` comments w/o changes. - - - - - e1cf8dc2 by brandon s allbery kf8nh at 2024-04-22T03:48:26-04:00 fix link in CODEOWNERS It seems that our local Gitlab no longer has documentation for the `CODEOWNERS` file, but the master documentation still does. Use that instead. - - - - - a27c6a49 by Fendor at 2024-04-22T10:13:03+02:00 Adapt to UserData split - - - - - 1efc5a7a by Fendor at 2024-04-22T10:13:03+02:00 Adapt to BinHandle split - - - - - 593f4e04 by Fendor at 2024-04-23T10:19:14-04:00 Add performance regression test for '-fwrite-simplified-core' - - - - - 1ba39b05 by Fendor at 2024-04-23T10:19:14-04:00 Typecheck corebindings lazily during bytecode generation This delays typechecking the corebindings until the bytecode generation happens. We also avoid allocating a thunk that is retained by `unsafeInterleaveIO`. In general, we shouldn't retain values of the hydrated `Type`, as not evaluating the bytecode object keeps it alive. It is better if we retain the unhydrated `IfaceType`. See Note [Hydrating Modules] - - - - - e916fc92 by Alan Zimmerman at 2024-04-23T10:19:50-04:00 EPA: Keep comments in a CaseAlt match The comments now live in the surrounding location, not inside the Match. Make sure we keep them. Closes #24707 - - - - - d2b17f32 by Cheng Shao at 2024-04-23T15:01:22-04:00 driver: force merge objects when building dynamic objects This patch forces the driver to always merge objects when building dynamic objects even when ar -L is supported. It is an oversight of !8887: original rationale of that patch is favoring the relatively cheap ar -L operation over object merging when ar -L is supported, which makes sense but only if we are building static objects! Omitting check for whether we are building dynamic objects will result in broken .so files with undefined reference errors at executable link time when building GHC with llvm-ar. Fixes #22210. - - - - - 209d09f5 by Julian Ospald at 2024-04-23T15:02:03-04:00 Allow non-absolute values for bootstrap GHC variable Fixes #24682 - - - - - 3fff0977 by Matthew Pickering at 2024-04-23T15:02:38-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. - - - - - c62dc317 by Cheng Shao at 2024-04-25T01:32:02-04:00 ghc-bignum: remove obsolete ln script This commit removes an obsolete ln script in ghc-bignum/gmp. See 060251c24ad160264ae8553efecbb8bed2f06360 for its original intention, but it's been obsolete for a long time, especially since the removal of the make build system. Hence the house cleaning. - - - - - 6399d52b by Cheng Shao at 2024-04-25T01:32:02-04:00 ghc-bignum: update gmp to 6.3.0 This patch bumps the gmp-tarballs submodule and updates gmp to 6.3.0. The tarball format is now xz, and gmpsrc.patch has been patched into the tarball so hadrian no longer needs to deal with patching logic when building in-tree GMP. - - - - - 65b4b92f by Cheng Shao at 2024-04-25T01:32:02-04:00 hadrian: remove obsolete Patch logic This commit removes obsolete Patch logic from hadrian, given we no longer need to patch the gmp tarball when building in-tree GMP. - - - - - 71f28958 by Cheng Shao at 2024-04-25T01:32:02-04:00 autoconf: remove obsolete patch detection This commit removes obsolete deletection logic of the patch command from autoconf scripts, given we no longer need to patch anything in the GHC build process. - - - - - daeda834 by Sylvain Henry at 2024-04-25T01:32:43-04:00 JS: correctly handle RUBBISH literals (#24664) - - - - - 8a06ddf6 by Matthew Pickering at 2024-04-25T11:16:16-04:00 Linearise ghc-internal and base build This is achieved by requesting the final package database for ghc-internal, which mandates it is fully built as a dependency of configuring the `base` package. This is at the expense of cross-package parrallelism between ghc-internal and the base package. Fixes #24436 - - - - - 94da9365 by Andrei Borzenkov at 2024-04-25T11:16:54-04:00 Fix tuple puns renaming (24702) Move tuple renaming short cutter from `isBuiltInOcc_maybe` to `isPunOcc_maybe`, so we consider incoming module. I also fixed some hidden bugs that raised after the change was done. - - - - - fa03b1fb by Fendor at 2024-04-26T18:03:13-04:00 Refactor the Binary serialisation interface The goal is simplifiy adding deduplication tables to `ModIface` interface serialisation. We identify two main points of interest that make this difficult: 1. UserData hardcodes what `Binary` instances can have deduplication tables. Moreover, it heavily uses partial functions. 2. GHC.Iface.Binary hardcodes the deduplication tables for 'Name' and 'FastString', making it difficult to add more deduplication. Instead of having a single `UserData` record with fields for all the types that can have deduplication tables, we allow to provide custom serialisers for any `Typeable`. These are wrapped in existentials and stored in a `Map` indexed by their respective `TypeRep`. The `Binary` instance of the type to deduplicate still needs to explicitly look up the decoder via `findUserDataReader` and `findUserDataWriter`, which is no worse than the status-quo. `Map` was chosen as microbenchmarks indicate it is the fastest for a small number of keys (< 10). To generalise the deduplication table serialisation mechanism, we introduce the types `ReaderTable` and `WriterTable` which provide a simple interface that is sufficient to implement a general purpose deduplication mechanism for `writeBinIface` and `readBinIface`. This allows us to provide a list of deduplication tables for serialisation that can be extended more easily, for example for `IfaceTyCon`, see the issue https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for more motivation. In addition to this refactoring, we split `UserData` into `ReaderUserData` and `WriterUserData`, to avoid partial functions and reduce overall memory usage, as we need fewer mutable variables. Bump haddock submodule to accomodate for `UserData` split. ------------------------- Metric Increase: MultiLayerModulesTH_Make MultiLayerModulesRecomp T21839c ------------------------- - - - - - bac57298 by Fendor at 2024-04-26T18:03:13-04:00 Split `BinHandle` into `ReadBinHandle` and `WriteBinHandle` A `BinHandle` contains too much information for reading data. For example, it needs to keep a `FastMutInt` and a `IORef BinData`, when the non-mutable variants would suffice. Additionally, this change has the benefit that anyone can immediately tell whether the `BinHandle` is used for reading or writing. Bump haddock submodule BinHandle split. - - - - - 4d6394dd by Simon Peyton Jones at 2024-04-26T18:03:49-04:00 Fix missing escaping-kind check in tcPatSynSig Note [Escaping kind in type signatures] explains how we deal with escaping kinds in type signatures, e.g. f :: forall r (a :: TYPE r). a where the kind of the body is (TYPE r), but `r` is not in scope outside the forall-type. I had missed this subtlety in tcPatSynSig, leading to #24686. This MR fixes it; and a similar bug in tc_top_lhs_type. (The latter is tested by T24686a.) - - - - - 981c2c2c by Alan Zimmerman at 2024-04-26T18:04:25-04:00 EPA: check-exact: check that the roundtrip reproduces the source Closes #24670 - - - - - a8616747 by Andrew Lelechenko at 2024-04-26T18:05:01-04:00 Document that setEnv is not thread-safe - - - - - 1e41de83 by Bryan Richter at 2024-04-26T18:05:37-04:00 CI: Work around frequent Signal 9 errors - - - - - a6d5f9da by Naïm Favier at 2024-04-27T17:52:40-04:00 ghc-internal: add MonadFix instance for (,) Closes https://gitlab.haskell.org/ghc/ghc/-/issues/24288, implements CLC proposal https://github.com/haskell/core-libraries-committee/issues/238. Adds a MonadFix instance for tuples, permitting value recursion in the "native" writer monad and bringing consistency with the existing instance for transformers's WriterT (and, to a lesser extent, for Solo). - - - - - 64feadcd by Rodrigo Mesquita at 2024-04-27T17:53:16-04:00 bindist: Fix xattr cleaning The original fix (725343aa) was incorrect because it used the shell bracket syntax which is the quoting syntax in autoconf, making the test for existence be incorrect and therefore `xattr` was never run. Fixes #24554 - - - - - e2094df3 by damhiya at 2024-04-28T23:52:00+09:00 Make read accepts binary integer formats CLC proposal : https://github.com/haskell/core-libraries-committee/issues/177 - - - - - c62239b7 by Sylvain Henry at 2024-04-29T10:35:00+02:00 Fix tests for T22229 - - - - - 1c2fd963 by Alan Zimmerman at 2024-04-29T23:17:00-04:00 EPA: Preserve comments in Match Pats Closes #24708 Closes #24715 Closes #24734 - - - - - 4189d17e by Sylvain Henry at 2024-04-29T23:17:42-04:00 LLVM: better unreachable default destination in Switch (#24717) See added note. Co-authored-by: Siddharth Bhat <siddu.druid at gmail.com> - - - - - a3725c88 by Cheng Shao at 2024-04-29T23:18:20-04:00 ci: enable wasm jobs for MRs with wasm label This patch enables wasm jobs for MRs with wasm label. Previously the wasm label didn't actually have any effect on the CI pipeline, and full-ci needed to be applied to run wasm jobs which was a waste of runners when working on the wasm backend, hence the fix here. - - - - - 702f7964 by Matthew Pickering at 2024-04-29T23:18:56-04:00 Make interface files and object files depend on inplace .conf file A potential fix for #24737 - - - - - 728af21e by Cheng Shao at 2024-04-30T05:30:23-04:00 utils: remove obsolete vagrant scripts Vagrantfile has long been removed in !5288. This commit further removes the obsolete vagrant scripts in the tree. - - - - - 36f2c342 by Cheng Shao at 2024-04-30T05:31:00-04:00 Update autoconf scripts Scripts taken from autoconf 948ae97ca5703224bd3eada06b7a69f40dd15a02 - - - - - ecbf22a6 by Ben Gamari at 2024-04-30T05:31:36-04:00 ghcup-metadata: Drop output_name field This is entirely redundant to the filename of the URL. There is no compelling reason to name the downloaded file differently from its source. - - - - - c56d728e by Zubin Duggal at 2024-04-30T22:45:09-04:00 testsuite: Handle exceptions in framework_fail when testdir is not initialised When `framework_fail` is called before initialising testdir, it would fail with an exception reporting the testdir not being initialised instead of the actual failure. Ensure we report the actual reason for the failure instead of failing in this way. One way this can manifest is when trying to run a test that doesn't exist using `--only` - - - - - d5bea4d6 by Alan Zimmerman at 2024-04-30T22:45:45-04:00 EPA: Fix range for GADT decl with sig only Closes #24714 - - - - - 4d78c53c by Sylvain Henry at 2024-05-01T17:23:06-04:00 Fix TH dependencies (#22229) Add a dependency between Syntax and Internal (via module reexport). - - - - - 37e38db4 by Sylvain Henry at 2024-05-01T17:23:06-04:00 Bump haddock submodule - - - - - ca13075c by Sylvain Henry at 2024-05-01T17:23:47-04:00 JS: cleanup to prepare for #24743 - - - - - 40026ac3 by Alan Zimmerman at 2024-05-01T22:45:07-04:00 EPA: Preserve comments for PrefixCon Preserve comments in fun (Con {- c1 -} a b) = undefined Closes #24736 - - - - - 92134789 by Hécate Moonlight at 2024-05-01T22:45:42-04:00 Correct `@since` metadata in HpcFlags It was introduced in base-4.20, not 4.22. Fix #24721 - - - - - a580722e by Cheng Shao at 2024-05-02T08:18:45-04:00 testsuite: fix req_target_smp predicate - - - - - ac9c5f84 by Andreas Klebinger at 2024-05-02T08:18:45-04:00 STM: Remove (unused)coarse grained locking. The STM code had a coarse grained locking mode guarded by #defines that was unused. This commit removes the code. - - - - - 917ef81b by Andreas Klebinger at 2024-05-02T08:18:45-04:00 STM: Be more optimistic when validating in-flight transactions. * Don't lock tvars when performing non-committal validation. * If we encounter a locked tvar don't consider it a failure. This means in-flight validation will only fail if committing at the moment of validation is *guaranteed* to fail. This prevents in-flight validation from failing spuriously if it happens in parallel on multiple threads or parallel to thread comitting. - - - - - 167a56a0 by Alan Zimmerman at 2024-05-02T08:19:22-04:00 EPA: fix span for empty \case(s) In instance SDecide Nat where SZero %~ (SSucc _) = Disproved (\case) Ensure the span for the HsLam covers the full construct. Closes #24748 - - - - - 9bae34d8 by doyougnu at 2024-05-02T15:41:08-04:00 testsuite: expand size testing infrastructure - closes #24191 - adds windows_skip, wasm_skip, wasm_arch, find_so, _find_so - path_from_ghcPkg, collect_size_ghc_pkg, collect_object_size, find_non_inplace functions to testsuite - adds on_windows and req_dynamic_ghc predicate to testsuite The design is to not make the testsuite too smart and simply offload to ghc-pkg for locations of object files and directories. - - - - - b85b1199 by Sylvain Henry at 2024-05-02T15:41:49-04:00 GHCi: support inlining breakpoints (#24712) When a breakpoint is inlined, its context may change (e.g. tyvars in scope). We must take this into account and not used the breakpoint tick index as its sole identifier. Each instance of a breakpoint (even with the same tick index) now gets a different "info" index. We also need to distinguish modules: - tick module: module with the break array (tick counters, status, etc.) - info module: module having the CgBreakInfo (info at occurrence site) - - - - - 649c24b9 by Oleg Grenrus at 2024-05-03T20:45:42-04:00 Expose constructors of SNat, SChar and SSymbol in ghc-internal - - - - - d603f199 by Mikolaj Konarski at 2024-05-03T20:46:19-04:00 Add DCoVarSet to PluginProv (!12037) - - - - - ba480026 by Serge S. Gulin at 2024-05-03T20:47:01-04:00 JS: Enable more efficient packing of string data (fixes #24706) - - - - - be1e60ee by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Track in-scope variables in ruleCheckProgram This small patch fixes #24726, by tracking in-scope variables properly in -drule-check. Not hard to do! - - - - - 58408c77 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Add a couple more HasCallStack constraints in SimpleOpt Just for debugging, no effect on normal code - - - - - 70e245e8 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Add comments to Prep.hs This documentation patch fixes a TODO left over from !12364 - - - - - e5687186 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Use HasDebugCallStack, rather than HasCallStack - - - - - 631cefec by Cheng Shao at 2024-05-03T20:48:17-04:00 driver: always merge objects when possible This patch makes the driver always merge objects with `ld -r` when possible, and only fall back to calling `ar -L` when merge objects command is unavailable. This completely reverts !8887 and !12313, given more fixes in Cabal seems to be needed to avoid breaking certain configurations and the maintainence cost is exceeding the behefits in this case :/ - - - - - 1dacb506 by Ben Gamari at 2024-05-03T20:48:53-04:00 Bump time submodule to 1.14 As requested in #24528. ------------------------- Metric Decrease: ghc_bignum_so rts_so Metric Increase: cabal_syntax_dir rts_so time_dir time_so ------------------------- - - - - - 4941b90e by Ben Gamari at 2024-05-03T20:48:53-04:00 Bump terminfo submodule to current master - - - - - 43d48b44 by Cheng Shao at 2024-05-03T20:49:30-04:00 wasm: use scheduler.postTask() for context switch when available This patch makes use of scheduler.postTask() for JSFFI context switch when it's available. It's a more principled approach than our MessageChannel based setImmediate() implementation, and it's available in latest version of Chromium based browsers. - - - - - 08207501 by Cheng Shao at 2024-05-03T20:50:08-04:00 testsuite: give pre_cmd for mhu-perf 5x time - - - - - bf3d4db0 by Alan Zimmerman at 2024-05-03T20:50:43-04:00 EPA: Preserve comments for pattern synonym sig Closes #24749 - - - - - c49493f2 by Matthew Pickering at 2024-05-04T06:02:57-04:00 tests: Widen acceptance window for dir and so size tests These are testing things which are sometimes out the control of a GHC developer. Therefore we shouldn't fail CI if something about these dependencies change because we can't do anything about it. It is still useful to have these statistics for visualisation in grafana though. Ticket #24759 - - - - - 9562808d by Matthew Pickering at 2024-05-04T06:02:57-04:00 Disable rts_so test It has already manifested large fluctuations and destabilising CI Fixes #24762 - - - - - fc24c5cf by Ryan Scott at 2024-05-04T06:03:33-04:00 unboxedSum{Type,Data}Name: Use GHC.Types as the module Unboxed sum constructors are now defined in the `GHC.Types` module, so if you manually quote an unboxed sum (e.g., `''Sum2#`), you will get a `Name` like: ```hs GHC.Types.Sum2# ``` The `unboxedSumTypeName` function in `template-haskell`, however, mistakenly believes that unboxed sum constructors are defined in `GHC.Prim`, so `unboxedSumTypeName 2` would return an entirely different `Name`: ```hs GHC.Prim.(#|#) ``` This is a problem for Template Haskell users, as it means that they can't be sure which `Name` is the correct one. (Similarly for `unboxedSumDataName`.) This patch fixes the implementations of `unboxedSum{Type,Data}Name` to use `GHC.Types` as the module. For consistency with `unboxedTupleTypeName`, the `unboxedSumTypeName` function now uses the non-punned syntax for unboxed sums (`Sum<N>#`) as the `OccName`. Fixes #24750. - - - - - 7eab4e01 by Alan Zimmerman at 2024-05-04T16:14:55+01:00 EPA: Widen stmtslist to include last semicolon Closes #24754 - - - - - 06f7db40 by Teo Camarasu at 2024-05-05T00:19:38-04:00 doc: Fix type error in hs_try_putmvar example - - - - - af000532 by Moritz Schuler at 2024-05-05T06:30:58-04:00 Fix parsing of module names in CLI arguments closes issue #24732 - - - - - da74e9c9 by Ben Gamari at 2024-05-05T06:31:34-04:00 ghc-platform: Add Setup.hs The Hadrian bootstrapping script relies upon `Setup.hs` to drive its build. Addresses #24761. - - - - - 35d34fde by Alan Zimmerman at 2024-05-05T12:52:40-04:00 EPA: preserve comments in class and data decls Fix checkTyClHdr which was discarding comments. Closes #24755 - - - - - 03c5dfbf by Simon Peyton Jones at 2024-05-05T12:53:15-04:00 Fix a float-out error Ticket #24768 showed that the Simplifier was accidentally destroying a join point. It turned out to be that we were sending a bottoming join point to the top, accidentally abstracting over /other/ join points. Easily fixed. - - - - - adba68e7 by John Ericson at 2024-05-05T19:35:56-04:00 Substitute bindist files with Hadrian not configure The `ghc-toolchain` overhaul will eventually replace all this stuff with something much more cleaned up, but I think it is still worth making this sort of cleanup in the meantime so other untanglings and dead code cleaning can procede. I was able to delete a fair amount of dead code doing this too. `LLVMTarget_CPP` is renamed to / merged with `LLVMTarget` because it wasn't actually turned into a valid CPP identifier. (Original to 1345c7cc42c45e63ab1726a8fd24a7e4d4222467, actually.) Progress on #23966 Co-Authored-By: Sylvain Henry <hsyl20 at gmail.com> - - - - - 18f4ff84 by Alan Zimmerman at 2024-05-05T19:36:32-04:00 EPA: fix mkHsOpTyPV duplicating comments Closes #24753 - - - - - a19201d4 by Matthew Craven at 2024-05-06T19:54:29-04:00 Add test cases for #24664 ...since none are present in the original MR !12463 fixing this issue. - - - - - 46328a49 by Alan Zimmerman at 2024-05-06T19:55:05-04:00 EPA: preserve comments in data decls Closes #24771 - - - - - 3b51995c by Andrei Borzenkov at 2024-05-07T14:39:40-04:00 Rename Solo# data constructor to MkSolo# (#24673) - data Solo# a = (# a #) + data Solo# a = MkSolo# a And `(# foo #)` syntax now becomes just a syntactic sugar for `MkSolo# a`. - - - - - 4d59abf2 by Arsen Arsenović at 2024-05-07T14:40:24-04:00 Add the cmm_cpp_is_gcc predicate to the testsuite A future C-- test called T24474-cmm-override-g0 relies on the GCC-specific behaviour of -g3 implying -dD, which, in turn, leads to it emitting #defines past the preprocessing stage. Clang, at least, does not do this, so the test would fail if ran on Clang. As the behaviour here being tested is ``-optCmmP-g3'' undoing effects of the workaround we apply as a fix for bug #24474, and the workaround was for GCC-specific behaviour, the test needs to be marked as fragile on other compilers. - - - - - 25b0b404 by Arsen Arsenović at 2024-05-07T14:40:24-04:00 Split out the C-- preprocessor, and make it pass -g0 Previously, C-- was processed with the C preprocessor program. This means that it inherited flags passed via -optc. A flag that is somewhat often passed through -optc is -g. At certain -g levels (>=2), GCC starts emitting defines *after* preprocessing, for the purposes of debug info generation. This is not useful for the C-- compiler, and, in fact, causes lexer errors. We can suppress this effect (safely, if supported) via -g0. As a workaround, in older versions of GCC (<=10), GCC only emitted defines if a certain set of -g*3 flags was passed. Newer versions check the debug level. For the former, we filter out those -g*3 flags and, for the latter, we specify -g0 on top of that. As a compatible and effective solution, this change adds a C-- preprocessor distinct from the C compiler and preprocessor, but that keeps its flags. The command line produced for C-- preprocessing now looks like: $pgmCmmP $optCs_without_g3 $g0_if_supported $optCmmP Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/24474 - - - - - 9b4129a5 by Andreas Klebinger at 2024-05-08T13:24:20-04:00 -fprof-late: Only insert cost centres on functions/non-workfree cafs. They are usually useless and doing so for data values comes with a large compile time/code size overhead. Fixes #24103 - - - - - 259b63d3 by Sebastian Graf at 2024-05-08T13:24:57-04:00 Simplifier: Preserve OccInfo on DataAlt fields when case binder is dead (#24770) See the adjusted `Note [DataAlt occ info]`. This change also has a positive repercussion on `Note [Combine case alts: awkward corner]`. Fixes #24770. We now try not to call `dataConRepStrictness` in `adjustFieldsIdInfo` when all fields are lazy anyway, leading to a 2% ghc/alloc decrease in T9675. Metric Decrease: T9675 - - - - - 31b28cdb by Sebastian Graf at 2024-05-08T13:24:57-04:00 Kill seqRule, discard dead seq# in Prep (#24334) Discarding seq#s in Core land via `seqRule` was problematic; see #24334. So instead we discard certain dead, discardable seq#s in Prep now. See the updated `Note [seq# magic]`. This fixes the symptoms of #24334. - - - - - b2682534 by Rodrigo Mesquita at 2024-05-10T01:47:51-04:00 Document NcgImpl methods Fixes #19914 - - - - - 4d3acbcf by Zejun Wu at 2024-05-10T01:48:28-04:00 Make renamer to be more flexible with parens in the LHS of the rules We used to reject LHS like `(f a) b` in RULES and requires it to be written as `f a b`. It will be handy to allow both as the expression may be more readable with extra parens in some cases when infix operator is involved. Espceially when TemplateHaskell is used, extra parens may be added out of user's control and result in "valid" rules being rejected and there are not always ways to workaround it. Fixes #24621 - - - - - ab840ce6 by Ben Gamari at 2024-05-10T01:49:04-04:00 IPE: Eliminate dependency on Read Instead of encoding the closure type as decimal string we now simply represent it as an integer, eliminating the need for `Read` in `GHC.Internal.InfoProv.Types.peekInfoProv`. Closes #24504. ------------------------- Metric Decrease: T24602_perf_size size_hello_artifact ------------------------- - - - - - a9979f55 by Cheng Shao at 2024-05-10T01:49:43-04:00 testsuite: fix testwsdeque with recent clang This patch fixes compilation of testwsdeque.c with recent versions of clang, which will fail with the error below: ``` testwsdeque.c:95:33: error: warning: format specifies type 'long' but the argument has type 'void *' [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~~ ^ testwsdeque.c:95:39: error: warning: format specifies type 'int' but the argument has type 'StgWord' (aka 'unsigned long') [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~ ^~~ | %lu testwsdeque.c:133:42: error: error: incompatible function pointer types passing 'void (void *)' to parameter of type 'OSThreadProc *' (aka 'void *(*)(void *)') [-Wincompatible-function-pointer-types] 133 | createOSThread(&ids[n], "thief", thief, (void*)(StgWord)n); | ^~~~~ /workspace/ghc/_build/stage1/lib/../lib/x86_64-linux-ghc-9.11.20240502/rts-1.0.2/include/rts/OSThreads.h:193:51: error: note: passing argument to parameter 'startProc' here 193 | OSThreadProc *startProc, void *param); | ^ 2 warnings and 1 error generated. ``` - - - - - c2b33fc9 by Rodrigo Mesquita at 2024-05-10T01:50:20-04:00 Rename pre-processor invocation args Small clean up. Uses proper names for the various groups of arguments that make up the pre-processor invocation. - - - - - 2b1af08b by Cheng Shao at 2024-05-10T01:50:55-04:00 ghc-heap: fix typo in ghc-heap cbits - - - - - fc2d6de1 by Jade at 2024-05-10T21:07:16-04:00 Improve performance of Data.List.sort(By) This patch improves the algorithm to sort lists in base. It does so using two strategies: 1) Use a four-way-merge instead of the 'default' two-way-merge. This is able to save comparisons and allocations. 2) Use `(>) a b` over `compare a b == GT` and allow inlining and specialization. This mainly benefits types with a fast (>). Note that this *may* break instances with a *malformed* Ord instance where `a > b` is *not* equal to `compare a b == GT`. CLC proposal: https://github.com/haskell/core-libraries-committee/issues/236 Fixes #24280 ------------------------- Metric Decrease: MultiLayerModulesTH_Make T10421 T13719 T15164 T18698a T18698b T1969 T9872a T9961 T18730 WWRec T12425 T15703 ------------------------- - - - - - 1012e8aa by Matthew Pickering at 2024-05-10T21:07:52-04:00 Revert "ghcup-metadata: Drop output_name field" This reverts commit ecbf22a6ac397a791204590f94c0afa82e29e79f. This breaks the ghcup metadata generation on the nightly jobs. - - - - - daff1e30 by Jannis at 2024-05-12T13:38:35-04:00 Division by constants optimization - - - - - 413217ba by Andreas Klebinger at 2024-05-12T13:39:11-04:00 Tidy: Add flag to expose unfoldings if they take dictionary arguments. Add the flag `-fexpose-overloaded-unfoldings` to be able to control this behaviour. For ghc's boot libraries file size grew by less than 1% when it was enabled. However I refrained from enabling it by default for now. I've also added a section on specialization more broadly to the users guide. ------------------------- Metric Decrease: MultiLayerModulesTH_OneShot Metric Increase: T12425 T13386 hard_hole_fits ------------------------- - - - - - c5d89412 by Zubin Duggal at 2024-05-13T22:19:53-04:00 Don't store a GlobalRdrEnv in `mi_globals` for GHCi. GHCi only needs the `mi_globals` field for modules imported with :module +*SomeModule. It uses this field to make the top level environment in `SomeModule` available to the repl. By default, only the first target in the command line parameters is "star" loaded into GHCi. Other modules have to be manually "star" loaded into the repl. Storing the top level GlobalRdrEnv for each module is very wasteful, especially given that we will most likely never need most of these environments. Instead we store only the information needed to reconstruct the top level environment in a module, which is the `IfaceTopEnv` data structure, consisting of all import statements as well as all top level symbols defined in the module (not taking export lists into account) When a particular module is "star-loaded" into GHCi (as the first commandline target, or via an explicit `:module +*SomeModule`, we reconstruct the top level environment on demand using the `IfaceTopEnv`. - - - - - d65bf4a2 by Fendor at 2024-05-13T22:20:30-04:00 Add perf regression test for `-fwrite-if-simplified-core` - - - - - 2c0f8ddb by Andrei Borzenkov at 2024-05-13T22:21:07-04:00 Improve pattern to type pattern transformation (23739) `pat_to_type_pat` function now can handle more patterns: - TuplePat - ListPat - LitPat - NPat - ConPat Allowing these new constructors in type patterns significantly increases possible shapes of type patterns without `type` keyword. This patch also changes how lookups in `lookupOccRnConstr` are performed, because we need to fall back into types when we didn't find a constructor on data level to perform `ConPat` to type transformation properly. - - - - - be514bb4 by Cheng Shao at 2024-05-13T22:21:43-04:00 hadrian: fix hadrian building with ghc-9.10.1 - - - - - ad38e954 by Cheng Shao at 2024-05-13T22:21:43-04:00 linters: fix lint-whitespace compilation with ghc-9.10.1 - - - - - a593f284 by Andreas Klebinger at 2024-05-15T07:32:10-04:00 Expand the `inline` rule to look through casts/ticks. Fixes #24808 - - - - - b1e0c313 by Cheng Shao at 2024-05-15T07:32:46-04:00 testsuite: bump PartialDownSweep timeout to 5x on wasm32 - - - - - b2227487 by Fendor at 2024-05-15T17:14:06-04:00 Add Eq and Ord instance to `IfaceType` We add an `Ord` instance so that we can store `IfaceType` in a `Data.Map` container. This is required to deduplicate `IfaceType` while writing `.hi` files to disk. Deduplication has many beneficial consequences to both file size and memory usage, as the deduplication enables implicit sharing of values. See issue #24540 for more motivation. The `Ord` instance would be unnecessary if we used a `TrieMap` instead of `Data.Map` for the deduplication process. While in theory this is clerarly the better option, experiments on the agda code base showed that a `TrieMap` implementation has worse run-time performance characteristics. To the change itself, we mostly derive `Eq` and `Ord`. This requires us to change occurrences of `FastString` with `LexicalFastString`, since `FastString` has no `Ord` instance. We change the definition of `IfLclName` to a newtype of `LexicalFastString`, to make such changes in the future easier. Bump haddock submodule for IfLclName changes - - - - - d368f9a6 by Fendor at 2024-05-15T17:14:06-04:00 Move out LiteralMap to avoid cyclic module dependencies - - - - - 2fcc09fd by Fendor at 2024-05-15T17:14:06-04:00 Add deduplication table for `IfaceType` The type `IfaceType` is a highly redundant, tree-like data structure. While benchmarking, we realised that the high redundancy of `IfaceType` causes high memory consumption in GHCi sessions when byte code is embedded into the `.hi` file via `-fwrite-if-simplified-core` or `-fbyte-code-and-object-code`. Loading such `.hi` files from disk introduces many duplicates of memory expensive values in `IfaceType`, such as `IfaceTyCon`, `IfaceTyConApp`, `IA_Arg` and many more. We improve the memory behaviour of GHCi by adding an additional deduplication table for `IfaceType` to the serialisation of `ModIface`, similar to how we deduplicate `Name`s and `FastString`s. When reading the interface file back, the table allows us to automatically share identical values of `IfaceType`. To provide some numbers, we evaluated this patch on the agda code base. We loaded the full library from the `.hi` files, which contained the embedded core expressions (`-fwrite-if-simplified-core`). Before this patch: * Load time: 11.7 s, 2.5 GB maximum residency. After this patch: * Load time: 7.3 s, 1.7 GB maximum residency. This deduplication has the beneficial side effect to additionally reduce the size of the on-disk interface files tremendously. For example, on agda, we reduce the size of `.hi` files (with `-fwrite-if-simplified-core`): * Before: 101 MB on disk * Now: 24 MB on disk This has even a beneficial side effect on the cabal store. We reduce the size of the store on disk: * Before: 341 MB on disk * Now: 310 MB on disk Note, none of the dependencies have been compiled with `-fwrite-if-simplified-core`, but `IfaceType` occurs in multiple locations in a `ModIface`. We also add IfaceType deduplication table to .hie serialisation and refactor .hie file serialisation to use the same infrastrucutre as `putWithTables`. Bump haddock submodule to accomodate for changes to the deduplication table layout and binary interface. - - - - - 36aa7cf1 by Fendor at 2024-05-15T17:14:06-04:00 Add run-time configurability of `.hi` file compression Introduce the flag `-fwrite-if-compression=<n>` which allows to configure the compression level of writing .hi files. The motivation is that some deduplication operations are too expensive for the average use case. Hence, we introduce multiple compression levels with variable impact on performance, but still reduce the memory residency and `.hi` file size on disk considerably. We introduce three compression levels: * `1`: `Normal` mode. This is the least amount of compression. It deduplicates only `Name` and `FastString`s, and is naturally the fastest compression mode. * `2`: `Safe` mode. It has a noticeable impact on .hi file size and is marginally slower than `Normal` mode. In general, it should be safe to always use `Safe` mode. * `3`: `Full` deduplication mode. Deduplicate as much as we can, resulting in minimal .hi files, but at the cost of additional compilation time. Reading .hi files doesn't need to know the initial compression level, and can always deserialise a `ModIface`, as we write out a byte that indicates the next value has been deduplicated. This allows users to experiment with different compression levels for packages, without recompilation of dependencies. Note, the deduplication also has an additional side effect of reduced memory consumption to implicit sharing of deduplicated elements. See https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for example where that matters. ------------------------- Metric Decrease: MultiLayerModulesDefsGhciWithCore T16875 T21839c T24471 hard_hole_fits libdir ------------------------- - - - - - 1e63a6fb by Matthew Pickering at 2024-05-15T17:14:07-04:00 Introduce regression tests for `.hi` file sizes Add regression tests to track how `-fwrite-if-compression` levels affect the size of `.hi` files. - - - - - 639d742b by M Farkas-Dyck at 2024-05-15T17:14:49-04:00 TTG: ApplicativeStatement exist only in Rn and Tc Co-Authored-By: romes <rodrigo.m.mesquita at gmail.com> - - - - - aa7b336b by Jade at 2024-05-15T23:06:17-04:00 Documentation: Improve documentation for symbols exported from System.IO - - - - - c561de8f by Jade at 2024-05-15T23:06:54-04:00 Improve suggestions for language extensions - When suggesting Language extensions, also suggest Extensions which imply them - Suggest ExplicitForAll and GADTSyntax instead of more specific extensions - Rephrase suggestion to include the term 'Extension' - Also moves some flag specific definitions out of Session.hs into Flags.hs (#24478) Fixes: #24477 Fixes: #24448 Fixes: #10893 - - - - - 4c7ae2a1 by Andreas Klebinger at 2024-05-15T23:07:30-04:00 Testsuite: Check if llvm assembler is available for have_llvm - - - - - bc672166 by Torsten Schmits at 2024-05-15T23:08:06-04:00 refactor quadratic search in warnMissingHomeModules - - - - - 7875e8cb by Torsten Schmits at 2024-05-15T23:08:06-04:00 add test that runs MakeDepend on thousands of modules - - - - - b84b91f5 by Adam Gundry at 2024-05-16T15:32:06-04:00 Representation-polymorphic HasField (fixes #22156) This generalises the HasField class to support representation polymorphism, so that instead of type HasField :: forall {k} . k -> Type -> Type -> Constraint we have type HasField :: forall {k} {r_rep} {a_rep} . k -> TYPE r_rep -> TYPE a_rep -> Constraint - - - - - 05285090 by Matthew Pickering at 2024-05-16T15:32:43-04:00 Bump os-string submodule to 2.0.2.2 Closes #24786 - - - - - 886ab43a by Cheng Shao at 2024-05-17T01:34:50-04:00 rts: do not prefetch mark_closure bdescr in non-moving gc when ASSERTS_ENABLED This commit fixes a small an oversight in !12148: the prefetch logic in non-moving GC may trap in debug RTS because it calls Bdescr() for mark_closure which may be a static one. It's fine in non-debug RTS because even invalid bdescr addresses are prefetched, they will not cause segfaults, so this commit implements the most straightforward fix: don't prefetch mark_closure bdescr when assertions are enabled. - - - - - b38dcf39 by Teo Camarasu at 2024-05-17T01:34:50-04:00 rts: Allocate non-moving segments with megablocks Non-moving segments are 8 blocks long and need to be aligned. Previously we serviced allocations by grabbing 15 blocks, finding an aligned 8 block group in it and returning the rest. This proved to lead to high levels of fragmentation as a de-allocating a segment caused an 8 block gap to form, and this could not be reused for allocation. This patch introduces a segment allocator based around using entire megablocks to service segment allocations in bulk. When there are no free segments, we grab an entire megablock and fill it with aligned segments. As the megablock is free, we can easily guarantee alignment. Any unused segments are placed on a free list. It only makes sense to free segments in bulk when all of the segments in a megablock are freeable. After sweeping, we grab the free list, sort it, and find all groups of segments where they cover the megablock and free them. This introduces a period of time when free segments are not available to the mutator, but the risk that this would lead to excessive allocation is low. Right after sweep, we should have an abundance of partially full segments, and this pruning step is relatively quick. In implementing this we drop the logic that kept NONMOVING_MAX_FREE segments on the free list. We also introduce an eventlog event to log the amount of pruned/retained free segments. See Note [Segment allocation strategy] Resolves #24150 ------------------------- Metric Decrease: T13253 T19695 ------------------------- - - - - - 710665bd by Cheng Shao at 2024-05-17T01:35:30-04:00 rts: fix I/O manager compilation errors for win32 target This patch fixes I/O manager compilation errors for win32 target discovered when cross-compiling to win32 using recent clang: ``` rts/win32/ThrIOManager.c:117:7: error: error: call to undeclared function 'is_io_mng_native_p'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 117 | if (is_io_mng_native_p ()) { | ^ | 117 | if (is_io_mng_native_p ()) { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/fs.c:143:28: error: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] 143 | int setErrNoFromWin32Error () { | ^ | void | 143 | int setErrNoFromWin32Error () { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:227:9: error: error: call to undeclared function 'interruptIOManagerEvent'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 227 | interruptIOManagerEvent (); | ^ | 227 | interruptIOManagerEvent (); | ^ rts/win32/ConsoleHandler.c:227:9: error: note: did you mean 'getIOManagerEvent'? | 227 | interruptIOManagerEvent (); | ^ rts/include/rts/IOInterface.h:27:10: error: note: 'getIOManagerEvent' declared here 27 | void * getIOManagerEvent (void); | ^ | 27 | void * getIOManagerEvent (void); | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:196:9: error: error: call to undeclared function 'setThreadLabel'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/win32/ConsoleHandler.c:196:9: error: note: did you mean 'postThreadLabel'? | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/eventlog/EventLog.h:118:6: error: note: 'postThreadLabel' declared here 118 | void postThreadLabel(Capability *cap, | ^ | 118 | void postThreadLabel(Capability *cap, | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) ``` - - - - - 28b9cee0 by Rodrigo Mesquita at 2024-05-17T01:36:05-04:00 configure: Check C99-compat for Cmm preprocessor Fixes #24815 - - - - - 8927e0c3 by Andreas Klebinger at 2024-05-17T01:36:41-04:00 Ensure `tcHasFixedRuntimeRep (# #)` returns True. - - - - - 04179044 by doyougnu at 2024-05-17T09:00:32-04:00 testsuite: make find_so regex less general Closes #24759 Background. In MR !12372 we began tracking shared object files and directories sizes for dependencies. However, this broke release builds because release builds alter the filenames swapping "in-place" for a hash. This was not considered in the MR and thus broke release pipelines. Furthermore, the rts_so test was found to be wildly varying and was therefore disabled in !12561. This commit fixes both of these issues: - fix the rts_so test by making the regex less general, now the rts_so test and all other foo.so tests must match "libHS<some-lib>-<version>-<hash|'in-place>-<ghc>". This prevents the rts_so test from accidentally matching different rts variants such as rts_threaded, which was the cause of the wild swings after !12372. - add logic to match either a hash or the string in-place. This should make the find_so function build agnostic. - - - - - 0962b50d by Andreas Klebinger at 2024-05-17T09:01:08-04:00 TagAnalysis: Treat all bottom ids as tagged during analysis. Ticket #24806 showed that we also need to treat dead end thunks as tagged during the analysis. - - - - - 7eb9f184 by Ben Gamari at 2024-05-17T11:23:37-04:00 Remove haddock submodule In preparation for merge into the GHC, as proposed in #23178. - - - - - 47b14dcc by Fendor at 2024-05-17T11:28:17-04:00 Adapt to `IfLclName` newtype changes (cherry picked from commit a711607e29b925f3d69e27c5fde4ba655c711ff1) - - - - - 6cc6681d by Fendor at 2024-05-17T11:28:17-04:00 Add IfaceType deduplication table to interface file serialisation Although we do not really need it in the interface file serialisation, as the deserialisation uses `getWithUserData`, we need to mirror the structure `getWithUserData` expects. Thus, we write essentially an empty `IfaceType` table at the end of the file, as the interface file doesn't reference `IfaceType`. (cherry picked from commit c9bc29c6a708483d2abc3d8ec9262510ce87ca61) - - - - - b9721206 by Ben Gamari at 2024-05-17T11:30:22-04:00 ghc-tags.yaml: Initial commit - - - - - 074e7d8f by Ben Gamari at 2024-05-17T11:31:29-04:00 fourmolu: Add configuration - - - - - 151b1736 by Ben Gamari at 2024-05-17T11:32:52-04:00 Makefile: Rework for use by haddock developers Previously the Makefile was present only for GHC's old make-based build system. Now since the make-based build system is gone we can use it for more useful ends. - - - - - a7dcf13b by Ben Gamari at 2024-05-17T11:36:14-04:00 Reformat with fourmolu Using previously-added configuration and `fourmolu -i .` Note that we exclude the test-cases (`./{hoogle,html-hypsrc,latex}-test`) as they are sensitive to formatting. - - - - - 0ea6017b by Ben Gamari at 2024-05-17T11:40:04-04:00 Add 'utils/haddock/' from commit 'a7dcf13bfbb97b20e75cc8ce650e2bb628db4660' git-subtree-dir: utils/haddock git-subtree-mainline: 7eb9f1849b1c72a1c61dee88462b4244550406f3 git-subtree-split: a7dcf13bfbb97b20e75cc8ce650e2bb628db4660 - - - - - aba1d304 by Hécate Moonlight at 2024-05-17T11:40:48-04:00 Add exceptions to the dangling notes list - - - - - 527bfbfb by Hécate Moonlight at 2024-05-17T11:40:52-04:00 Add haddock to the whitespace lint ignore list - - - - - 43274677 by Ben Gamari at 2024-05-17T11:41:20-04:00 git-blame-ignore-revs: Ignore haddock reformatting - - - - - 0e679e37 by Fendor at 2024-05-18T00:27:24-04:00 Pass cpp options to the CC builder in hadrian - - - - - bb40244e by Sylvain Henry at 2024-05-18T00:28:06-04:00 JS: fix allocation constant (fix #24746) - - - - - 646d30ab by Jade at 2024-05-18T19:23:31+02:00 Add highlighting for inline-code snippets in haddock - - - - - 64459a3e by Hécate Moonlight at 2024-05-19T08:42:27-04:00 haddock: Add a .readthedocs.yml file for online documentation - - - - - 7d3d9bbf by Serge S. Gulin at 2024-05-19T18:47:05+00:00 Unicode: General Category size test (related #24789) Added trivial size performance test which involves unicode general category usage via `read`. The `read` itself uses general category to detect spaces. The purpose for this test is to measure outcome of applying improvements at General Category representation in code discussed at #24789. - - - - - 8e04efcf by Alan Zimmerman at 2024-05-19T21:29:34-04:00 EPA: Remove redundant code Remove unused epAnnAnns function various cases for showAstData that no longer exist - - - - - 071d7a1e by Rodrigo Mesquita at 2024-05-20T10:55:16-04:00 Improve docs on closed type families in hs-boots Fixes #24776 - - - - - d9e2c119 by Torsten Schmits at 2024-05-20T10:55:52-04:00 Use default deviation for large-project test This new performance test has the purpose of detecting regressions in complexity in relation to the number of modules in a project, so 1% deviation is way too small to avoid false positives. - - - - - 20b0136a by Ben Gamari at 2024-05-22T00:31:39-04:00 ghcup-metadata: Various fixes from 9.10.1 Use Debian 12/x86-64, Debian 10/aarch64, and Debian 11/aarch64 bindists where possible. - - - - - 6838a7c3 by Sylvain Henry at 2024-05-22T00:32:23-04:00 Reverse arguments to stgCallocBytes (fix #24828) - - - - - f50f46c3 by Fendor at 2024-05-22T00:32:59-04:00 Add log messages for Iface serialisation compression level Fix the label of the number of 'IfaceType' entries in the log message. Add log message for the compression level that is used to serialise a an interface file. Adds `Outputable` instance for 'CompressionIFace'. - - - - - 3bad5d55 by Hécate Moonlight at 2024-05-22T00:33:40-04:00 base: Update doctests outputs ghc-internal: Update doctests outputs - - - - - 9317c6fb by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix the testsuites of the haddock-library - Apply all the metadata revisions from Hackage to the cabal file. - Fix the `ParserSpec.hs` file in the `spec` testsuite of haddock-library. - Make `CHANGES.md` an extra-doc-file instead of an extra-source-file. - - - - - 54073b02 by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix parser of @since pragma The testsuite contained tests for annotations of the form `@since foo-bar-0.5.0`, but the parser was written incorrectly. - - - - - ede6ede3 by Matthew Pickering at 2024-05-22T00:34:57-04:00 Fix nightly pages job It seems likely broken by 9f99126a which moved `index.html` from the root folder into `docs/` folder. Fixes #24840 - - - - - b7bcf729 by Cheng Shao at 2024-05-22T00:35:32-04:00 autoconf: remove unused context diff check This patch removes redundant autoconf check for the context diff program given it isn't actually been used anywhere, especially since make removal. - - - - - ea2fe66e by Hécate Moonlight at 2024-05-22T00:36:13-04:00 haddock: Rework the contributing guide - - - - - 0f302a94 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 haddock: Add module relationships diagrams of haddock-api and haddock-library - - - - - d1a9f34f by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add instructions - - - - - b880ee80 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add SVG outputs - - - - - 6d7e6ad8 by Ben Gamari at 2024-05-22T13:40:05-04:00 rts: Fix size of StgOrigThunkInfo frames Previously the entry code of the `stg_orig_thunk` frame failed to account for the size of the profiling header as it hard-coded the frame size. Fix this. Fixes #24809. - - - - - c645fe40 by Fendor at 2024-05-22T13:40:05-04:00 Add regression test T24809 for stg_orig_thunk_info_frame size - - - - - 4181aa40 by Andreas Klebinger at 2024-05-22T13:40:42-04:00 bindists: Check for existence of share folder before trying to copy it. This folder isn't distributed in windows bindists A lack of doing so resulted us copying loads of files twice. - - - - - d216510e by Matthew Pickering at 2024-05-22T13:40:42-04:00 Remove ad-hoc installation of mingw toolchain in relocatable bindists This reverts 616ac30026e8dd7d2ebb98d92dde071eedf5d951 The choice about whether to install mingw is taken in the installation makefile. This is also broken on non-windows systems. The actual issue was the EnableDistroToolchain variable wasn't declared in mk/config.mk and therefore the check to install mingw was failing. - - - - - 7b4c1998 by Cheng Shao at 2024-05-22T21:52:52-04:00 testsuite: fix T17920 for wasm backend T17920 was marked as fragile on wasm before; it can be trivially fixed by avoiding calling variadic printf() in cmm. - - - - - c739383b by Cheng Shao at 2024-05-22T21:53:29-04:00 testsuite: bump T22744 timeout to 5x - - - - - c4c6d714 by Cheng Shao at 2024-05-22T21:54:06-04:00 testsuite: don't attempt to detect host cpu features when testing cross ghc The testsuite driver CPU feature detection logic only detects host CPU and only makes sense when we are not testing a cross GHC. - - - - - 3d9e4ce6 by Simon Peyton Jones at 2024-05-22T21:54:43-04:00 Better skolemisation As #24810 showed, it is (a little) better to skolemise en-bloc, so that Note [Let-bound skolems] fires more often. See Note [Skolemisation en bloc] in GHC.Tc.Utils.Instantiate. - - - - - a3cd3a1d by Ryan Scott at 2024-05-22T21:55:19-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. - - - - - 4bb2a7cc by Hécate Moonlight at 2024-05-22T21:55:59-04:00 [base] Document the memory overhead of ByteArray Add a diagram that shows the constituent parts of a ByteArray and their memory overhead. - - - - - 8b2a016a by Hécate Moonlight at 2024-05-22T21:56:38-04:00 Haddock: Add MR template for Haddock - - - - - ead75532 by Peter Trommler at 2024-05-23T02:28:05-04:00 PPC: Support ELF v2 on powerpc64 big-endian Detect ELF v2 on PowerPC 64-bit systems. Check for `_CALL_ELF` preprocessor macro. Fixes #21191 - - - - - 9d4c10f2 by Hécate Kleidukos at 2024-05-23T02:28:44-04:00 gitlab: Add @Kleidukos to CODEOWNERS for utils/haddock - - - - - 28e64170 by Preetham Gujjula at 2024-05-23T07:20:48-04:00 haddock: Add cabal-fmt to tools for `make style` - - - - - 00126a89 by Andrei Borzenkov at 2024-05-23T07:21:24-04:00 haddock: fix verbosity option parsing - - - - - a3e0b68b by Ryan Hendrickson at 2024-05-23T15:52:03-04:00 base: specify tie-breaking behavior of min, max, and related list/Foldable functions - - - - - bdcc0f37 by doyougnu at 2024-05-24T07:51:18-04:00 cmm: add word <-> double/float bitcast - closes: #25331 This is the last step in the project plan described in #25331. This commit: - adds bitcast operands for x86_64, LLVM, aarch64 - For PPC and i386 we resort to using the cmm implementations - renames conversion MachOps from Conv to Round|Truncate - - - - - f0d257f7 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor Some functions in StgToByteCode were filtering out void arguments. However, StgToByteCode is called after unarisation: the void arguments should have been removed earlier. Instead of filtering out, we assert that the args are non-void. - - - - - 03137fd2 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor `layoutNativeCall` was always called with a `primRepCmmType platform` callback. Hence we can put it inside of `layoutNativeCall` rather than repeat it. - - - - - 27c430f3 by David Binder at 2024-05-24T07:52:38-04:00 haddock: Remove compatibility shims for GHC < 8.4 from haddock-library - - - - - 8dd8a076 by Cheng Shao at 2024-05-24T07:53:14-04:00 compiler: avoid saving foreign call target to local when there are no caller-save GlobalRegs This patch makes the STG->Cmm backend avoid saving foreign call target to local when there are no caller-save GlobalRegs. Since 321941a8ebe25192cdeece723e1058f2f47809ea, when we lower a foreign call, we unconditionally save the foreign call target to a temporary local first, then rely on cmmSink to clean it up later, which only happens with -fcmm-sink (implied by -O) and not in unoptimized code. And this is troublesome for the wasm backend NCG, which needs to infer a foreign call target symbol's type signature from the Cmm call site. Previously, the NCG has been emitting incorrect type signatures for unoptimized code, which happens to work with `wasm-ld` most of the time, but this is never future-proof against upstream toolchain updates, and it causes horrible breakages when LTO objects are included in linker input. Hence this patch. - - - - - 986df1ab by Cheng Shao at 2024-05-24T07:53:14-04:00 testsuite: add callee-no-local regression test - - - - - 52d62e2a by Sylvain Henry at 2024-05-24T07:53:57-04:00 Fix HasCallStack leftovers from !12514 / #24726 - - - - - c5e00c35 by crumbtoo at 2024-05-24T07:54:38-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 - - - - - bd323b0e by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Ensure that CHANGELOG is included in extra-source-files This was missed in the `ghc-internal` split. Closes #24831. - - - - - 1bfd32e8 by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Fix changelog reference to setBacktraceMechanismState (cherry picked from commit b63f7ba01fdfd98a01d2f0dec8d9262b3e595c5d) - - - - - 43e8e4f3 by Sylvain Henry at 2024-05-24T12:16:43-04:00 Float/double unboxed literal support for HexFloatLiterals (fix #22155) - - - - - 4a7f4713 by Fendor at 2024-05-24T12:17:19-04:00 Improve test labels for binary interface file size tests Test labels for binary interface file sizes are hard to read and overly verbose at the same time. Extend the name for the metric title, but shorten it in the actual comparison table. - - - - - 14e554cf by Zubin Duggal at 2024-05-24T12:17:55-04:00 Revert "Fix haskell/haddock#783 Don't show button if --quickjump not present" This reverts commit 7776566531e72c415f66dd3b13da9041c52076aa. - - - - - f56838c3 by Ben Gamari at 2024-05-24T12:17:55-04:00 Fix default hyperlinked sources pattern Previously this didn't include the `%M` token which manifested as broken links to the hyperlinked sources of reexports of declarations defined in other packages. Fixes haddock#1628. (cherry picked from commit 1432bcc943d41736eca491ecec4eb9a6304dab36) - - - - - 42efa62c by Ben Gamari at 2024-05-24T12:17:55-04:00 Make DocPaths a proper data type (cherry picked from commit 7f3a5c4da0023ae47b4c376c9b1ea2d706c94d8c) - - - - - 53d9ceb3 by Ben Gamari at 2024-05-24T12:17:55-04:00 haddock: Bump version to 2.30 (cherry picked from commit 994989ed3d535177e57b778629726aeabe8c7602) - - - - - e4db1112 by Zubin Duggal at 2024-05-24T12:17:55-04:00 haddock-api: allow base 4.20 and ghc 9.11 - - - - - e294f7a2 by PHO at 2024-05-24T12:17:55-04:00 Add a flag "threaded" for building haddock with the threaded RTS GHC isn't guaranteed to have a threaded RTS. There should be a way to build it with the vanilla one. (cherry picked from commit 75a94e010fb5b0236c670d22b04f5472397dc15d) - - - - - 51165bc9 by Andreas Klebinger at 2024-05-25T10:58:03-04:00 Update ticky counter event docs. Add the info about the info table address and json fields. Fixes #23200 - - - - - 98597ad5 by Sylvain Henry at 2024-05-25T10:58:45-04:00 Export extractPromotedList (#24866) This can be useful in plugins. - - - - - 228dcae6 by Teo Camarasu at 2024-05-28T13:12:24+00:00 template-haskell: Move wired-ins to ghc-internal Thus we make `template-haskell` reinstallable and keep it as the public API for Template Haskell. All of the wired-in identifiers are moved to `ghc-internal`. This necessitates also moving much of `ghc-boot-th` into `ghc-internal`. These modules are then re-exported from `ghc-boot-th` and `template-haskell`. To avoid a dependency on `template-haskell` from `lib:ghc`, we instead depend on the TH ASTs via `ghc-boot-th`. As `template-haskell` no longer has special status, we can drop the logic adding an implicit dependency on `template-haskell` when using TH. We can also drop the `template-haskell-next` package, which was previously used when bootstrapping. When bootstrapping, we need to vendor the TH AST modules from `ghc-internal` into `ghc-boot-th`. This is controlled by the `bootstrap` cabal flag as before. See Note [Bootstrapping Template Haskell]. We split out a GHC.Internal.TH.Lift module resolving #24752. This module is only built when not bootstrapping. Resolves #24703 ------------------------- Metric Increase: ghc_boot_th_dir ghc_boot_th_so ------------------------- - - - - - 62dded28 by Teo Camarasu at 2024-05-28T13:12:24+00:00 testsuite: mark tests broken by #24886 Now that `template-haskell` is no longer wired-in. These tests are triggering #24886, and so need to be marked broken. - - - - - 3ca72ad9 by Cheng Shao at 2024-05-30T02:57:06-04:00 rts: fix missing function prototypes in ClosureMacros.h - - - - - e0029e3d by Andreas Klebinger at 2024-05-30T02:57:43-04:00 UnliftedFFITypes: Allow `(# #)` as argument when it's the only argument. This allows representing functions like: int foo(void); to be imported like this: foreign import ccall "a_number_c" c_number :: (# #) -> Int64# Which can be useful when the imported function isn't implicitly stateful. - - - - - d0401335 by Matthew Pickering at 2024-05-30T02:58:19-04:00 ci: Update ci-images commit for fedora38 image The fedora38 nightly job has been failing for quite a while because `diff` was no longer installed. The ci-images bump explicitly installs `diffutils` into these images so hopefully they now pass again. - - - - - 3c97c74a by Jan Hrček at 2024-05-30T02:58:58-04:00 Update exactprint docs - - - - - 77760cd7 by Jan Hrček at 2024-05-30T02:58:58-04:00 Incorporate review feedback - - - - - 87591368 by Jan Hrček at 2024-05-30T02:58:58-04:00 Remove no longer relevant reference to comments - - - - - 05f4f142 by Jan Hrček at 2024-05-30T02:58:59-04:00 Replace outdated code example - - - - - 45a4a5f3 by Andreas Klebinger at 2024-05-30T02:59:34-04:00 Reword error resulting from missing -XBangPatterns. It can be the result of either a bang pattern or strict binding, so now we say so instead of claiming it must be a bang pattern. Fixes #21032 - - - - - e17f2df9 by Cheng Shao at 2024-05-30T03:00:10-04:00 testsuite: bump MultiLayerModulesDefsGhciReload timeout to 10x - - - - - 7a660042 by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - c77a48af by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 05c4fafb by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 5ff83bfc by Sylvain Henry at 2024-05-30T14:43:10-04:00 JS: remove useless h$CLOCK_REALTIME (#23202) - - - - - 95ef2d58 by Matthew Pickering at 2024-05-30T14:43:47-04:00 ghcup-metadata: Fix metadata generation There were some syntax errors in the generation script which were preventing it from running. I have tested this with: ``` nix shell --extra-experimental-features nix-command -f .gitlab/rel_eng -c ghcup-metadata --metadata ghcup-0.0.7.yaml --date="2024-05-27" --pipeline-id=95534 --version=9.11.20240525 ``` which completed successfully. - - - - - 1bc66ee4 by Jakob Bruenker at 2024-05-30T14:44:22-04:00 Add diagrams to Arrows documentation This adds diagrams to the documentation of Arrows, similar to the ones found on https://www.haskell.org/arrows/. It does not add diagrams for ArrowChoice for the time being, mainly because it's not clear to me how to visually distinguish them from the ones for Arrow. Ideally, you might want to do something like highlight the arrows belonging to the same tuple or same Either in common colors, but that's not really possible with unicode. - - - - - d10a1c65 by Matthew Craven at 2024-05-30T23:35:48-04:00 Make UnsafeSNat et al. into pattern synonyms ...so that they do not cause coerce to bypass the nominal role on the corresponding singleton types when they are imported. See Note [Preventing unsafe coercions for singleton types] and the discussion at #23478. This also introduces unsafeWithSNatCo (and analogues for Char and Symbol) so that users can still access the dangerous coercions that importing the real constructors would allow, but only in a very localized way. - - - - - 0958937e by Cheng Shao at 2024-05-30T23:36:25-04:00 hadrian: build C/C++ with split sections when enabled When split sections is enabled, ensure -fsplit-sections is passed to GHC as well when invoking GHC to compile C/C++; and pass -ffunction-sections -fdata-sections to gcc/clang when compiling C/C++ with the hadrian Cc builder. Fixes #23381. - - - - - 02b1f91e by Cheng Shao at 2024-05-30T23:36:25-04:00 driver: build C/C++ with -ffunction-sections -fdata-sections when split sections is enabled When -fsplit-sections is passed to GHC, pass -ffunction-sections -fdata-sections to gcc/clang when building C/C++. Previously, -fsplit-sections was only respected by the NCG/LLVM backends, but not the unregisterised backend; the GHC driver did not pass -fdata-sections and -ffunction-sections to the C compiler, which resulted in excessive executable sizes. Fixes #23381. ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - fd47e2e3 by Cheng Shao at 2024-05-30T23:37:00-04:00 testsuite: mark process005 as fragile on JS - - - - - 34a04ea1 by Matthew Pickering at 2024-05-31T06:08:36-04:00 Add -Wderiving-typeable to -Wall Deriving `Typeable` does nothing, and it hasn't done for a long while. There has also been a warning for a long while which warns you about uselessly deriving it but it wasn't enabled in -Wall. Fixes #24784 - - - - - 75fa7b0b by Matthew Pickering at 2024-05-31T06:08:36-04:00 docs: Fix formatting of changelog entries - - - - - 303c4b33 by Preetham Gujjula at 2024-05-31T06:09:21-04:00 docs: Fix link to injective type families paper Closes #24863 - - - - - df97e9a6 by Ben Gamari at 2024-05-31T06:09:57-04:00 ghc-internal: Fix package description The previous description was inherited from `base` and was inappropriate for `ghc-internal`. Also fix the maintainer and bug reporting fields. Closes #24906. - - - - - bf0737c0 by Cheng Shao at 2024-05-31T06:10:33-04:00 compiler: remove ArchWasm32 special case in cmmDoCmmSwitchPlans This patch removes special consideration for ArchWasm32 in cmmDoCmmSwitchPlans, which means the compiler will now disable cmmImplementSwitchPlans for wasm unreg backend, just like unreg backend of other targets. We enabled it in the past to workaround some compile-time panic in older versions of LLVM, but those panics are no longer present, hence no need to keep this workaround. - - - - - 7eda4bd2 by Cheng Shao at 2024-05-31T15:52:04-04:00 utils: add hie.yaml config file for ghc-config Add hie.yaml to ghc-config project directory so it can be edited using HLS. - - - - - 1e5752f6 by Cheng Shao at 2024-05-31T15:52:05-04:00 hadrian: handle findExecutable "" gracefully hadrian may invoke findExecutable "" at run-time due to a certain program is not found by configure script. Which is fine and findExecutable is supposed to return Nothing in this case. However, on Windows there's a directory bug that throws an exception (see https://github.com/haskell/directory/issues/180), so we might as well use a wrapper for findExecutable and handle exceptions gracefully. - - - - - 4eb5ad09 by Cheng Shao at 2024-05-31T15:52:05-04:00 configure: do not set LLC/OPT/LLVMAS fallback values when FIND_LLVM_PROG fails When configure fails to find LLC/OPT/LLVMAS within supported version range, it used to set "llc"/"opt"/"clang" as fallback values. This behavior is particularly troublesome when the user has llc/opt/clang with other versions in their PATH and run the testsuite, since hadrian will incorrectly assume have_llvm=True and pass that to the testsuite driver, resulting in annoying optllvm test failures (#23186). If configure determines llc/opt/clang wouldn't work, then we shouldn't pretend it'll work at all, and the bindist configure will invoke FIND_LLVM_PROG check again at install time anyway. - - - - - 5f1afdf7 by Sylvain Henry at 2024-05-31T15:52:52-04:00 Introduce UniqueSet and use it to replace 'UniqSet Unique' 'UniqSet Unique' represents a set of uniques as a 'Map Unique Unique', which is wasting space (associated key/value are always the same). Fix #23572 and #23605 - - - - - e0aa42b9 by crumbtoo at 2024-05-31T15:53:33-04:00 Improve template-haskell haddocks Closes #15822 - - - - - ae170155 by Olivier Benz at 2024-06-01T09:35:17-04:00 Bump max LLVM version to 19 (not inclusive) - - - - - 92aa65ea by Matthew Pickering at 2024-06-01T09:35:17-04:00 ci: Update CI images to test LLVM 18 The debian12 image in this commit has llvm 18 installed. - - - - - adb1fe42 by Serge S. Gulin at 2024-06-01T09:35:53-04:00 Unicode: make ucd2haskell build-able again ucd2haskell tool used streamly library which version in cabal was out of date. It is updated to the latest version at hackage with deprecated parts rewritten. Also following fixes were applied to existing code in suppose that from its last run the code structure was changed and now it was required to be up to date with actual folder structures: 1. Ghc module path environment got a suffix with `src`. 2. Generated code got 2.1 `GHC.Internal` prefix for `Data.*`. 2.2 `GHC.Unicode.Internal` swapped on `GHC.Internal.Unicode` according to actual structure. - - - - - ad56fd84 by Jade at 2024-06-01T09:36:29-04:00 Replace 'NB' with 'Note' in error messages - - - - - 6346c669 by Cheng Shao at 2024-06-01T09:37:04-04:00 compiler: fix -ddump-cmm-raw when compiling .cmm This patch fixes missing -ddump-cmm-raw output when compiling .cmm, which is useful for debugging cmm related codegen issues. - - - - - 1c834ad4 by Ryan Scott at 2024-06-01T09:37:40-04:00 Print namespace specifiers in FixitySig's Outputable instance For whatever reason, the `Outputable` instance for `FixitySig` simply did not print out namespace specifiers, leading to the confusing `-ddump-splices` output seen in #24911. This patch corrects this oversight. Fixes #24911. - - - - - cf49fb5f by Sylvain Henry at 2024-06-01T09:38:19-04:00 Configure: display C++ compiler path - - - - - f9c1ae12 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable PIC for in-tree GMP on wasm32 This patch disables PIC for in-tree GMP on wasm32 target. Enabling PIC unconditionally adds undesired code size and runtime overhead for wasm32. - - - - - 1a32f828 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable in-tree gmp fft code path for wasm32 This patch disables in-tree GMP FFT code paths for wasm32 target in order to give up some performance of multiplying very large operands in exchange for reduced code size. - - - - - 06277d56 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: build in-tree GMP with malloc-notreentrant on wasm32 This patch makes hadrian build in-tree GMP with the --enable-alloca=malloc-notreentrant configure option. We will only need malloc-reentrant when we have threaded RTS and SMP support on wasm32, which will take some time to happen, before which we should use malloc-notreentrant to avoid undesired runtime overhead. - - - - - 9f614270 by ARATA Mizuki at 2024-06-02T14:02:35-04:00 Set package include paths when assembling .S files Fixes #24839. Co-authored-by: Sylvain Henry <hsyl20 at gmail.com> - - - - - 4998a6ed by Alex Mason at 2024-06-03T02:09:29-04:00 Improve performance of genericWordQuotRem2Op (#22966) Implements the algorithm from compiler-rt's udiv128by64to64default. This rewrite results in a roughly 24x improvement in runtime on AArch64 (and likely any other arch that uses it). - - - - - ae50a8eb by Cheng Shao at 2024-06-03T02:10:05-04:00 testsuite: mark T7773 as fragile on wasm - - - - - c8ece0df by Fendor at 2024-06-03T19:43:22-04:00 Migrate `Finder` component to `OsPath`, fixed #24616 For each module in a GHCi session, we keep alive one `ModLocation`. A `ModLocation` is fairly inefficiently packed, as `String`s are expensive in memory usage. While benchmarking the agda codebase, we concluded that we keep alive around 11MB of `FilePath`'s, solely retained by `ModLocation`. We provide a more densely packed encoding of `ModLocation`, by moving from `FilePath` to `OsPath`. Further, we migrate the full `Finder` component to `OsPath` to avoid unnecessary transformations. As the `Finder` component is well-encapsulated, this requires only a minimal amount of changes in other modules. We introduce pattern synonym for 'ModLocation' which maintains backwards compatibility and avoids breaking consumers of 'ModLocation'. - - - - - 0cff083a by Cheng Shao at 2024-06-03T19:43:58-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. - - - - - 18f63970 by Sebastian Graf at 2024-06-04T05:05:27-04:00 Parser: Remove unused `apats` rule - - - - - 38757c30 by David Knothe at 2024-06-04T05:05:27-04:00 Implement Or Patterns (#22596) This commit introduces a new language extension, `-XOrPatterns`, as described in GHC Proposal 522. An or-pattern `pat1; ...; patk` succeeds iff one of the patterns `pat1`, ..., `patk` succeed, in this order. See also the summary `Note [Implmentation of OrPatterns]`. Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> - - - - - 395412e8 by Cheng Shao at 2024-06-04T05:06:04-04:00 compiler/ghci/rts: remove stdcall support completely We have formally dropped i386 windows support (#18487) a long time ago. The stdcall foreign call convention is only used by i386 windows, and the legacy logic around it is a significant maintenance burden for future work that adds arm64 windows support (#24603). Therefore, this patch removes stdcall support completely from the compiler as well as the RTS (#24883): - stdcall is still recognized as a FFI calling convention in Haskell syntax. GHC will now unconditionally emit a warning (-Wunsupported-calling-conventions) and treat it as ccall. - Apart from minimum logic to support the parsing and warning logic, all other code paths related to stdcall has been completely stripped from the compiler. - ghci only supports FFI_DEFAULT_ABI and ccall convention from now on. - FFI foreign export adjustor code on all platforms no longer handles the stdcall case and only handles ccall from now on. - The Win32 specific parts of RTS no longer has special code paths for stdcall. This commit is the final nail on the coffin for i386 windows support. Further commits will perform more housecleaning to strip the legacy code paths and pave way for future arm64 windows support. - - - - - d1fe9ab6 by Cheng Shao at 2024-06-04T05:06:04-04:00 rts: remove legacy i386 windows code paths This commit removes some legacy i386 windows related code paths in the RTS, given this target is no longer supported. - - - - - a605e4b2 by Cheng Shao at 2024-06-04T05:06:04-04:00 autoconf: remove i386 windows related logic This commit removes legacy i386 windows logic in autoconf scripts. - - - - - 91e5ac5e by Cheng Shao at 2024-06-04T05:06:04-04:00 llvm-targets: remove i386 windows support This commit removes i386 windows from llvm-targets and the script to generate it. - - - - - 65fe75a4 by Cheng Shao at 2024-06-04T05:06:04-04:00 libraries/utils: remove stdcall related legacy logic This commit removes stdcall related legacy logic in libraries and utils. ccall should be used uniformly for all supported windows hosts from now on. - - - - - d2a83302 by Cheng Shao at 2024-06-04T05:06:04-04:00 testsuite: adapt the testsuite for stdcall removal This patch adjusts test cases to handle the stdcall removal: - Some stdcall usages are replaced with ccall since stdcall doesn't make sense anymore. - We also preserve some stdcall usages, and check in the expected warning messages to ensure GHC always warn about stdcall usages (-Wunsupported-calling-conventions) as expected. - Error code testsuite coverage is slightly improved, -Wunsupported-calling-conventions is now tested. - Obsolete code paths related to i386 windows are also removed. - - - - - cef8f47a by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: minor adjustments for stdcall removal This commit include minor adjustments of documentation related to stdcall removal. - - - - - 54332437 by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: mention i386 Windows removal in 9.12 changelog This commit mentions removal of i386 Windows support and stdcall related change in the 9.12 changelog. - - - - - 2aaea8a1 by Cheng Shao at 2024-06-04T05:06:40-04:00 hadrian: improve user settings documentation This patch adds minor improvements to hadrian user settings documentation: - Add missing `ghc.cpp.opts` case - Remove non-existent `cxx` case - Clarify `cc.c.opts` also works for C++, while `cc.deps.opts` doesn't - Add example of passing configure argument to autoconf packages - - - - - 71010381 by Alex Mason at 2024-06-04T12:09:07-04:00 Add AArch64 CLZ, CTZ, RBIT primop implementations. Adds support for emitting the clz and rbit instructions, which are used by GHC.Prim.clz*#, GHC.Prim.ctz*# and GHC.Prim.bitReverse*#. - - - - - 44e2abfb by Cheng Shao at 2024-06-04T12:09:43-04:00 hadrian: add +text_simdutf flavour transformer to allow building text with simdutf This patch adds a +text_simdutf flavour transformer to hadrian to allow downstream packagers and users that build from source to opt-in simdutf support for text, in order to benefit from SIMD speedup at run-time. It's still disabled by default for the time being. - - - - - 077cb2e1 by Cheng Shao at 2024-06-04T12:09:43-04:00 ci: enable +text_simdutf flavour transformer for wasm jobs This commit enables +text_simdutf flavour transformer for wasm jobs, so text is now built with simdutf support for wasm. - - - - - b23746ad by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Use TemplateHaskellQuotes in instance Lift ByteArray Resolves #24852 - - - - - 3fd25743 by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Mark addrToByteArray as NOINLINE This function should never be inlined in order to keep code size small. - - - - - 98ad1ea5 by Cheng Shao at 2024-06-04T22:51:26-04:00 compiler: remove unused CompilerInfo/LinkerInfo types This patch removes CompilerInfo/LinkerInfo types from the compiler since they aren't actually used anywhere. - - - - - 11795244 by Cheng Shao at 2024-06-05T06:33:17-04:00 rts: remove unused PowerPC/IA64 native adjustor code This commit removes unused PowerPC/IA64 native adjustor code which is never actually enabled by autoconf/hadrian. Fixes #24920. - - - - - 5132754b by Sylvain Henry at 2024-06-05T06:33:57-04:00 RTS: fix warnings with doing*Profiling (#24918) - - - - - accc8c33 by Cheng Shao at 2024-06-05T11:35:36-04:00 hadrian: don't depend on inplace/mingw when --enable-distro-toolchain on Windows - - - - - 6ffbd678 by Cheng Shao at 2024-06-05T11:35:37-04:00 autoconf: normalize paths of some build-time dependencies on Windows This commit applies path normalization via cygpath -m to some build-time dependencies on Windows. Without this logic, the /clang64/bin prefixed msys2-style paths cause the build to fail with --enable-distro-toolchain. - - - - - 075dc6d4 by Cheng Shao at 2024-06-05T11:36:12-04:00 hadrian: remove OSDarwin mention from speedHack This commit removes mentioning of OSDarwin from speedHack, since speedHack is purely for i386 and we no longer support i386 darwin (#24921). - - - - - 83235c4c by Cheng Shao at 2024-06-05T11:36:12-04:00 compiler: remove 32-bit darwin logic This commit removes all 32-bit darwin logic from the compiler, given we no longer support 32-bit apple systems (#24921). Also contains a bit more cleanup of obsolete i386 windows logic. - - - - - 1eb99bc3 by Cheng Shao at 2024-06-05T11:36:12-04:00 rts: remove 32-bit darwin/ios logic This commit removes 32-bit darwin/ios related logic from the rts, given we no longer support them (#24921). - - - - - 24f65892 by Cheng Shao at 2024-06-05T11:36:12-04:00 llvm-targets: remove 32-bit darwin/ios targets This commit removes 32-bit darwin/ios targets from llvm-targets given we no longer support them (#24921). - - - - - ccdbd689 by Cheng Shao at 2024-06-05T11:36:12-04:00 testsuite: remove 32-bit darwin logic This commit removes 32-bit darwin logic from the testsuite given it's no longer supported (#24921). Also contains more cleanup of obsolete i386 windows logic. - - - - - 11d661c4 by Cheng Shao at 2024-06-05T11:36:13-04:00 docs: mention 32-bit darwin/ios removal in 9.12 changelog This commit mentions removal of 32-bit darwin/ios support (#24921) in the 9.12 changelog. - - - - - 7c173310 by Georgi Lyubenov at 2024-06-05T15:17:22-04:00 Add firstA and secondA to Data.Bitraversable Please see https://github.com/haskell/core-libraries-committee/issues/172 for related discussion - - - - - 3b6f9fd1 by Ben Gamari at 2024-06-05T15:17:59-04:00 base: Fix name of changelog Fixes #24899. Also place it under `extra-doc-files` to better reflect its nature and avoid triggering unnecessary recompilation if it changes. - - - - - 1f4d2ef7 by Sebastian Graf at 2024-06-05T15:18:34-04:00 Announce Or-patterns in the release notes for GHC 9.12 (#22596) Leftover from !9229. - - - - - 8650338d by Jan Hrček at 2024-06-06T10:39:24-04:00 Improve haddocks of Language.Haskell.Syntax.Pat.Pat - - - - - 2eee65e1 by Cheng Shao at 2024-06-06T10:40:00-04:00 testsuite: bump T7653 timeout for wasm - - - - - 990fed60 by Sylvain Henry at 2024-06-07T14:45:23-04:00 StgToCmm: refactor opTranslate and friends - Change arguments order to avoid `\args -> ...` lambdas - Fix documentation - Rename StgToCmm options ("big" doesn't mean anything) - - - - - 1afad514 by Sylvain Henry at 2024-06-07T14:45:23-04:00 NCG x86: remove dead code (#5444) Since 6755d833af8c21bbad6585144b10e20ac4a0a1ab this code is dead. - - - - - 595c0894 by Cheng Shao at 2024-06-07T14:45:58-04:00 testsuite: skip objc-hi/objcxx-hi when cross compiling objc-hi/objcxx-hi should be skipped when cross compiling. The existing opsys('darwin') predicate only asserts the host system is darwin but tells us nothing about the target, hence the oversight. - - - - - edfe6140 by qqwy at 2024-06-08T11:23:54-04:00 Replace '?callStack' implicit param with HasCallStack in GHC.Internal.Exception.throw - - - - - 35a64220 by Cheng Shao at 2024-06-08T11:24:30-04:00 rts: cleanup inlining logic This patch removes pre-C11 legacy code paths related to INLINE_HEADER/STATIC_INLINE/EXTERN_INLINE macros, ensure EXTERN_INLINE is treated as static inline in most cases (fixes #24945), and also corrects the comments accordingly. - - - - - 9ea90ed2 by Andrew Lelechenko at 2024-06-08T11:25:06-04:00 CODEOWNERS: add @core-libraries to track base interface changes A low-tech tactical solution for #24919 - - - - - 580fef7b by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update CHANGELOG to reflect current version - - - - - 391ecff5 by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update prologue.txt to reflect package description - - - - - 3dca3b7d by Ben Gamari at 2024-06-09T01:27:57-04:00 compiler: Clarify comment regarding need for MOVABS The comment wasn't clear in stating that it was only applicable to immediate source and memory target operands. - - - - - 6bd850e8 by doyougnu at 2024-06-09T21:02:14-04:00 JS: establish single source of truth for symbols In pursuit of: #22736. This MR moves ad-hoc symbols used throughout the js backend into a single symbols file. Why? First, this cleans up the code by removing ad-hoc strings created on the fly and therefore makes the code more maintainable. Second, it makes it much easier to eventually type these identifiers. - - - - - f3017dd3 by Cheng Shao at 2024-06-09T21:02:49-04:00 rts: replace ad-hoc MYTASK_USE_TLV with proper CC_SUPPORTS_TLS This patch replaces the ad-hoc `MYTASK_USE_TLV` with the `CC_SUPPORTS_TLS` macro. If TLS support is detected by autoconf, then we should use that for managing `myTask` in the threaded RTS. - - - - - e17d7e8c by Ben Gamari at 2024-06-11T05:25:21-04:00 users-guide: Fix stylistic issues in 9.12 release notes - - - - - 8a8a982a by Hugo Peters at 2024-06-11T05:25:57-04:00 fix typo in the simplifier debug output: baling -> bailing - - - - - 16475bb8 by Hécate Moonlight at 2024-06-12T03:07:55-04:00 haddock: Correct the Makefile to take into account Darwin systems - - - - - a2f60da5 by Hécate Kleidukos at 2024-06-12T03:08:35-04:00 haddock: Remove obsolete links to github.com/haskell/haddock in the docs - - - - - de4395cd by qqwy at 2024-06-12T03:09:12-04:00 Add `__GLASGOW_HASKELL_ASSERTS_IGNORED__` as CPP macro name if `-fasserts-ignored is set. This allows users to create their own Control.Exception.assert-like functionality that does something other than raising an `AssertFailed` exception. Fixes #24967 - - - - - 0e9c4dee by Ryan Hendrickson at 2024-06-12T03:09:53-04:00 compiler: add hint to TcRnBadlyStaged message - - - - - 2747cd34 by Simon Peyton Jones at 2024-06-12T12:51:37-04:00 Fix a QuickLook bug This MR fixes the bug exposed by #24676. The problem was that quickLookArg was trying to avoid calling tcInstFun unnecessarily; but it was in fact necessary. But that in turn forced me into a significant refactoring, putting more fields into EValArgQL. Highlights: see Note [Quick Look overview] in GHC.Tc.Gen.App * Instantiation variables are now distinguishable from ordinary unification variables, by level number = QLInstVar. This is treated like "level infinity". See Note [The QLInstVar TcLevel] in GHC.Tc.Utils.TcType. * In `tcApp`, we don't track the instantiation variables in a set Delta any more; instead, we just tell them apart by their level number. * EValArgQL now much more clearly captures the "half-done" state of typechecking an argument, ready for later resumption. See Note [Quick Look at value arguments] in GHC.Tc.Gen.App * Elminated a bogus (never used) fast-path in GHC.Tc.Utils.Instantiate.instCallConstraints See Note [Possible fast path for equality constraints] Many other small refactorings. - - - - - 1b1523b1 by George Thomas at 2024-06-12T12:52:18-04:00 Fix non-compiling extensible record `HasField` example - - - - - 97b141a3 by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Fix hyperlinker source urls (#24907) This fixes a bug introduced by f56838c36235febb224107fa62334ebfe9941aba Links to external modules in the hyperlinker are uniformly generated using splicing the template given to us instead of attempting to construct the url in an ad-hoc manner. - - - - - 954f864c by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Add name anchor to external source urls from documentation page URLs for external source links from documentation pages were missing a splice location for the name. Fixes #24912 - - - - - b0b64177 by Simon Peyton Jones at 2024-06-12T12:53:31-04:00 Prioritise nominal equalities The main payload of this patch is * Prioritise nominal equalities in the constraint solver. This ameliorates the incompleteness of solving for representational constraints over newtypes: see #24887. See (EX2) in Note [Decomposing newtype equalities] in GHC.Tc.Solver.Equality In doing this patch I tripped over some other things that I refactored: * Move `isCoVarType` from `GHC.Core.Type` to `GHC.Core.Predicate` where it seems more at home. * Clarify the "rewrite role" of a constraint. I was very puzzled about what the role of, say `(Eq a)` might be, but see the new Note [The rewrite-role of a constraint]. In doing so I made predTypeEqRel crash when given a non-equality. Usually it expects an equality; but it was being mis-used for the above rewrite-role stuff. - - - - - cb7c1b83 by Liam Goodacre at 2024-06-12T12:54:09-04:00 compiler: missing-deriving-strategies suggested fix Extends the missing-deriving-strategies warning with a suggested fix that includes which deriving strategies were assumed. For info about the warning, see comments for `TcRnNoDerivStratSpecified`, `TcRnNoDerivingClauseStrategySpecified`, & `TcRnNoStandaloneDerivingStrategySpecified`. For info about the suggested fix, see `SuggestExplicitDerivingClauseStrategies` & `SuggestExplicitStandalanoDerivingStrategy`. docs: Rewords missing-deriving-strategies to mention the suggested fix. Resolves #24955 - - - - - 4e36d3a3 by Jan Hrček at 2024-06-12T12:54:48-04:00 Further haddocks improvements in Language.Haskell.Syntax.Pat.Pat - - - - - 558353f4 by Cheng Shao at 2024-06-12T12:55:24-04:00 rts: use page sized mblocks on wasm This patch changes mblock size to page size on wasm. It allows us to simplify our wasi-libc fork, makes it much easier to test third party libc allocators like emmalloc/mimalloc, as well as experimenting with threaded RTS in wasm. - - - - - b3cc5366 by Matthew Pickering at 2024-06-12T23:06:57-04:00 compiler: Make ghc-experimental not wired in If you need to wire in definitions, then place them in ghc-internal and reexport them from ghc-experimental. Ticket #24903 - - - - - 700eeab9 by Hécate Kleidukos at 2024-06-12T23:07:37-04:00 base: Use a more appropriate unicode arrow for the ByteArray diagram This commit rectifies the usage of a unicode arrow in favour of one that doesn't provoke mis-alignment. - - - - - cca7de25 by Matthew Pickering at 2024-06-12T23:08:14-04:00 ghcup-metadata: Fix debian version ranges This was caught by `ghcup-ci` failing and attempting to install a deb12 bindist on deb11. ``` configure: WARNING: m4/prep_target_file.m4: Expecting YES/NO but got in ArSupportsDashL_STAGE0. Defaulting to False. bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by bin/ghc-toolchain-bin) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) ``` Fixes #24974 - - - - - 7b23ce8b by Pierre Le Marre at 2024-06-13T15:35:04-04:00 ucd2haskell: remove Streamly dependency + misc - Remove dead code. - Remove `streamly` dependency. - Process files with `bytestring`. - Replace Unicode files parsers with the corresponding ones from the package `unicode-data-parser`. - Simplify cabal file and rename module - Regenerate `ghc-internal` Unicode files with new header - - - - - 4570319f by Jacco Krijnen at 2024-06-13T15:35:41-04:00 Document how to run haddocks tests (#24976) Also remove ghc 9.7 requirement - - - - - fb629e24 by amesgen at 2024-06-14T00:28:20-04:00 compiler: refactor lower_CmmExpr_Ptr - - - - - def46c8c by amesgen at 2024-06-14T00:28:20-04:00 compiler: handle CmmRegOff in lower_CmmExpr_Ptr - - - - - ce76bf78 by Simon Peyton Jones at 2024-06-14T00:28:56-04:00 Small documentation update in Quick Look - - - - - 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - a7247841 by Ben Gamari at 2024-09-23T14:35:04-04:00 Bump process submodule to v1.6.24.0 - - - - - 23 changed files: - .ghcid - + .git-blame-ignore-revs - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - − .gitlab/circle-ci-job.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - − .gitlab/gen_ci.hs - + .gitlab/generate-ci/LICENSE - + .gitlab/generate-ci/README.mkd - + .gitlab/generate-ci/flake.lock - + .gitlab/generate-ci/flake.nix - + .gitlab/generate-ci/gen_ci.hs - + .gitlab/generate-ci/generate-ci.cabal - + .gitlab/generate-ci/generate-job-metadata - + .gitlab/generate-ci/generate-jobs - + .gitlab/generate-ci/hie.yaml - − .gitlab/generate_jobs - + .gitlab/hello.hs - − .gitlab/issue_templates/bug.md - + .gitlab/issue_templates/default.md The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e057774a0c5a2f51b99e23ca5b3527bcc0a75b82...a72478410d7d0e4dd20eb07c143e8084a60e343c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e057774a0c5a2f51b99e23ca5b3527bcc0a75b82...a72478410d7d0e4dd20eb07c143e8084a60e343c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 18:43:53 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Mon, 23 Sep 2024 14:43:53 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8] 10 commits: hadrian: Generate HSC2HS_EXTRAS variable in bindist installation Message-ID: <66f1b6e9e88bf_173fba4cfa7c6259@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8 at Glasgow Haskell Compiler / GHC Commits: 2eca9c7a by Ben Gamari at 2024-02-22T18:31:44+05:30 hadrian: Generate HSC2HS_EXTRAS variable in bindist installation We must generate the hsc2hs wrapper at bindist installation time since it must contain `--lflag` and `--cflag` arguments which depend upon the installation path. The solution here is to substitute these variables in the configure script (see mk/hsc2hs.in). This is then copied over a dummy wrapper in the install rules. Fixes #24050. (cherry picked from commit efcbad2dfd242d0bc2c91da5390fe8456a536cc3) - - - - - d5246e19 by Matthew Pickering at 2024-02-22T18:31:58+05:30 ci: Show --info for installed compiler (cherry picked from commit c540559cf188625bec668fa6cd94d4f94413d730) - - - - - f3225ed4 by Zubin Duggal at 2024-02-22T23:42:18+05:30 Accept change in MultiLayerModulesTH_Make This test is flaky on darwin Metric Decrease: MultiLayerModulesTH_Make - - - - - 1c08e245 by Ben Gamari at 2024-02-23T12:31:12+05:30 testsuite: Ignore stderr in T8089 Otherwise spurious "Killed: 9" messages to stderr may cause the test to fail. Fixes #24361. (cherry picked from commit e693a4e8589bad35588c51fccc87f4388e7d5874) - - - - - 9cb7e73a by Sebastian Graf at 2024-05-06T15:11:25+02:00 exprIsTrivial: Factor out shared implementation The duplication between `exprIsTrivial` and `getIdFromTrivialExpr_maybe` has been bugging me for a long time. This patch introduces an inlinable worker function `trivial_expr_fold` acting as the single, shared decision procedure of triviality. It "returns" a Church-encoded `Maybe (Maybe Id)`, so when it is inlined, it fuses to similar code as before. (Better code, even, in the case of `getIdFromTrivialExpr` which presently allocates a `Just` constructor that cancels away after this patch.) - - - - - 78a25354 by Sebastian Graf at 2024-05-07T09:25:04+02:00 Some cherry-picked bits of 59202c8 to fix #24718 As noted in f3225ed4b3f3c4, the test below is flaky on Darwin. Metric Decrease: MultiLayerModulesTH_Make - - - - - 5e5ba257 by Ben Gamari at 2024-09-23T14:33:41-04:00 Bump process submodule to 1.6.24.0 - - - - - 8d7a027a by Ben Gamari at 2024-09-23T14:36:47-04:00 Bump stm submodule to v2.5.3.1 - - - - - 2755511d by Andrew Lelechenko at 2024-09-23T14:42:39-04:00 Bump submodule array to 0.5.8.0 (cherry picked from commit 80769bc9f56541601796366485283a697c52a18b) - - - - - 52e82b4d by Andrew Lelechenko at 2024-09-23T14:43:06-04:00 Bump submodule deepseq to 1.5.1.0 (cherry picked from commit 8e462f4d4bdf2a6c34c249e7be8084565600d300) - - - - - 15 changed files: - .gitlab/ci.sh - compiler/GHC/Core/Utils.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/Unit/Types.hs - distrib/configure.ac.in - hadrian/bindist/Makefile - hadrian/src/Rules/BinaryDist.hs - libraries/array - libraries/base/tests/all.T - libraries/deepseq - libraries/process - libraries/stm - + mk/hsc2hs.in - + testsuite/tests/core-to-stg/T24718.hs - testsuite/tests/core-to-stg/all.T Changes: ===================================== .gitlab/ci.sh ===================================== @@ -575,6 +575,8 @@ function install_bindist() { --prefix="$instdir" \ "${args[@]+"${args[@]}"}" make_install_destdir "$TOP"/destdir "$instdir" + # And check the `--info` of the installed compiler, sometimes useful in CI log. + "$instdir"/bin/ghc --info ;; esac popd ===================================== compiler/GHC/Core/Utils.hs ===================================== @@ -23,9 +23,9 @@ module GHC.Core.Utils ( -- * Properties of expressions exprType, coreAltType, coreAltsType, mkLamType, mkLamTypes, mkFunctionType, - exprIsDupable, exprIsTrivial, getIdFromTrivialExpr, - getIdFromTrivialExpr_maybe, - exprIsCheap, exprIsExpandable, exprIsCheapX, CheapAppFun, + exprIsTrivial, getIdFromTrivialExpr, getIdFromTrivialExpr_maybe, + trivial_expr_fold, + exprIsDupable, exprIsCheap, exprIsExpandable, exprIsCheapX, CheapAppFun, exprIsHNF, exprOkForSpeculation, exprOkForSideEffects, exprOkForSpecEval, exprIsWorkFree, exprIsConLike, isCheapApp, isExpandableApp, isSaturatedConApp, @@ -1046,20 +1046,37 @@ and that confuses the code generator (#11155). So best to kill it off at source. -} +{-# INLINE trivial_expr_fold #-} +trivial_expr_fold :: (Id -> r) -> (Literal -> r) -> r -> r -> CoreExpr -> r +-- ^ The worker function for Note [exprIsTrivial] and Note [getIdFromTrivialExpr] +-- This is meant to have the code of both functions in one place and make it +-- easy to derive custom predicates. +-- +-- (trivial_expr_fold k_id k_triv k_not_triv e) +-- * returns (k_id x) if `e` is a variable `x` (with trivial wrapping) +-- * returns (k_lit x) if `e` is a trivial literal `l` (with trivial wrapping) +-- * returns k_triv if `e` is a literal, type, or coercion (with trivial wrapping) +-- * returns k_not_triv otherwise +-- +-- where "trivial wrapping" is +-- * Type application or abstraction +-- * Ticks other than `tickishIsCode` +-- * `case e of {}` an empty case +trivial_expr_fold k_id k_lit k_triv k_not_triv = go + where + go (Var v) = k_id v -- See Note [Variables are trivial] + go (Lit l) | litIsTrivial l = k_lit l + go (Type _) = k_triv + go (Coercion _) = k_triv + go (App f t) | not (isRuntimeArg t) = go f + go (Lam b e) | not (isRuntimeVar b) = go e + go (Tick t e) | not (tickishIsCode t) = go e -- See Note [Tick trivial] + go (Cast e _) = go e + go (Case e _ _ []) = go e -- See Note [Empty case is trivial] + go _ = k_not_triv + exprIsTrivial :: CoreExpr -> Bool --- If you modify this function, you may also --- need to modify getIdFromTrivialExpr -exprIsTrivial (Var _) = True -- See Note [Variables are trivial] -exprIsTrivial (Type _) = True -exprIsTrivial (Coercion _) = True -exprIsTrivial (Lit lit) = litIsTrivial lit -exprIsTrivial (App e arg) = not (isRuntimeArg arg) && exprIsTrivial e -exprIsTrivial (Lam b e) = not (isRuntimeVar b) && exprIsTrivial e -exprIsTrivial (Tick t e) = not (tickishIsCode t) && exprIsTrivial e - -- See Note [Tick trivial] -exprIsTrivial (Cast e _) = exprIsTrivial e -exprIsTrivial (Case e _ _ []) = exprIsTrivial e -- See Note [Empty case is trivial] -exprIsTrivial _ = False +exprIsTrivial e = trivial_expr_fold (const True) (const True) True False e {- Note [getIdFromTrivialExpr] @@ -1079,24 +1096,13 @@ T12076lit for an example where this matters. -} getIdFromTrivialExpr :: HasDebugCallStack => CoreExpr -> Id -getIdFromTrivialExpr e - = fromMaybe (pprPanic "getIdFromTrivialExpr" (ppr e)) - (getIdFromTrivialExpr_maybe e) - -getIdFromTrivialExpr_maybe :: CoreExpr -> Maybe Id -- See Note [getIdFromTrivialExpr] --- Th equations for this should line up with those for exprIsTrivial -getIdFromTrivialExpr_maybe e - = go e +getIdFromTrivialExpr e = trivial_expr_fold id (const panic) panic panic e where - go (App f t) | not (isRuntimeArg t) = go f - go (Tick t e) | not (tickishIsCode t) = go e - go (Cast e _) = go e - go (Lam b e) | not (isRuntimeVar b) = go e - go (Case e _ _ []) = go e - go (Var v) = Just v - go _ = Nothing + panic = pprPanic "getIdFromTrivialExpr" (ppr e) +getIdFromTrivialExpr_maybe :: CoreExpr -> Maybe Id +getIdFromTrivialExpr_maybe e = trivial_expr_fold Just (const Nothing) Nothing Nothing e {- ********************************************************************* * * ===================================== compiler/GHC/CoreToStg.hs ===================================== @@ -19,8 +19,7 @@ module GHC.CoreToStg ( CoreToStgOpts (..), coreToStg ) where import GHC.Prelude import GHC.Core -import GHC.Core.Utils ( exprType, findDefault, isJoinBind - , exprIsTickedString_maybe ) +import GHC.Core.Utils import GHC.Core.Opt.Arity ( manifestArity ) import GHC.Core.Type import GHC.Core.TyCon @@ -49,7 +48,7 @@ import GHC.Unit.Module import GHC.Data.FastString import GHC.Platform ( Platform ) import GHC.Platform.Ways -import GHC.Builtin.PrimOps ( PrimCall(..), primOpWrapperId ) +import GHC.Builtin.PrimOps import GHC.Utils.Outputable import GHC.Utils.Monad @@ -574,6 +573,19 @@ coreToStgApp f args ticks = do -- This is the guy that turns applications into A-normal form -- --------------------------------------------------------------------------- +getStgArgFromTrivialArg :: HasDebugCallStack => CoreArg -> StgArg +-- A (non-erased) trivial CoreArg corresponds to an atomic StgArg. +-- CoreArgs may not immediately look trivial, e.g., `case e of {}` or +-- `case unsafeequalityProof of UnsafeRefl -> e` might intervene. +-- Good thing we can just call `trivial_expr_fold` here. +getStgArgFromTrivialArg e + | Just s <- exprIsTickedString_maybe e -- This case is just for backport to GHC 9.8, + = StgLitArg (LitString s) -- where we used to treat strings as valid StgArgs + | otherwise + = trivial_expr_fold StgVarArg StgLitArg panic panic e + where + panic = pprPanic "getStgArgFromTrivialArg" (ppr e) + coreToStgArgs :: [CoreArg] -> CtsM ([StgArg], [StgTickish]) coreToStgArgs [] = return ([], []) @@ -586,42 +598,29 @@ coreToStgArgs (Coercion _ : args) -- Coercion argument; See Note [Coercion token = do { (args', ts) <- coreToStgArgs args ; return (StgVarArg coercionTokenId : args', ts) } -coreToStgArgs (Tick t e : args) - = assert (not (tickishIsCode t)) $ - do { (args', ts) <- coreToStgArgs (e : args) - ; let !t' = coreToStgTick (exprType e) t - ; return (args', t':ts) } - coreToStgArgs (arg : args) = do -- Non-type argument (stg_args, ticks) <- coreToStgArgs args - arg' <- coreToStgExpr arg - let - (aticks, arg'') = stripStgTicksTop tickishFloatable arg' - stg_arg = case arg'' of - StgApp v [] -> StgVarArg v - StgConApp con _ [] _ -> StgVarArg (dataConWorkId con) - StgOpApp (StgPrimOp op) [] _ -> StgVarArg (primOpWrapperId op) - StgLit lit -> StgLitArg lit - _ -> pprPanic "coreToStgArgs" (ppr arg $$ pprStgExpr panicStgPprOpts arg' $$ pprStgExpr panicStgPprOpts arg'') - - -- WARNING: what if we have an argument like (v `cast` co) - -- where 'co' changes the representation type? - -- (This really only happens if co is unsafe.) - -- Then all the getArgAmode stuff in CgBindery will set the - -- cg_rep of the CgIdInfo based on the type of v, rather - -- than the type of 'co'. - -- This matters particularly when the function is a primop - -- or foreign call. - -- Wanted: a better solution than this hacky warning - + -- We know that `arg` must be trivial, but it may contain Ticks. + -- Example from test case `decodeMyStack`: + -- $ @... ((src Data.Tuple.snd) @Int @[..]) + -- Note that unfortunately the Tick is not at the top. + -- So we'll traverse the expression twice: + -- * Once with `stripTicksT` (which collects *all* ticks from the expression) + -- * and another time with `getStgArgFromTrivialArg`. + -- Since the argument is trivial, the only place the Tick can occur is + -- somehow wrapping a variable (give or take type args, as above). platform <- getPlatform - let - arg_rep = typePrimRep (exprType arg) - stg_arg_rep = typePrimRep (stgArgType stg_arg) + let arg_ty = exprType arg + ticks' = map (coreToStgTick arg_ty) (stripTicksT (not . tickishIsCode) arg) + arg' = getStgArgFromTrivialArg arg + arg_rep = typePrimRep arg_ty + stg_arg_rep = typePrimRep (stgArgType arg') bad_args = not (primRepsCompatible platform arg_rep stg_arg_rep) - warnPprTrace bad_args "Dangerous-looking argument. Probable cause: bad unsafeCoerce#" (ppr arg) $ - return (stg_arg : stg_args, ticks ++ aticks) + massertPpr (length ticks' <= 1) (text "More than one Tick in trivial arg:" <+> ppr arg) + warnPprTrace bad_args "Dangerous-looking argument. Probable cause: bad unsafeCoerce#" (ppr arg) (return ()) + + return (arg' : stg_args, ticks' ++ ticks) coreToStgTick :: Type -- type of the ticked expression -> CoreTickish ===================================== compiler/GHC/Unit/Types.hs ===================================== @@ -101,9 +101,9 @@ import GHC.Utils.Fingerprint import GHC.Utils.Misc import GHC.Settings.Config (cProjectUnitId) -import Control.DeepSeq +import Control.DeepSeq (NFData(..)) import Data.Data -import Data.List (sortBy ) +import Data.List (sortBy) import Data.Function import Data.Bifunctor import qualified Data.ByteString as BS ===================================== distrib/configure.ac.in ===================================== @@ -298,6 +298,7 @@ AC_SUBST(UseLibdw) FP_SETTINGS AC_CONFIG_FILES([config.mk]) +AC_CONFIG_FILES([mk/hsc2hs]) AC_OUTPUT # We get caught by ===================================== hadrian/bindist/Makefile ===================================== @@ -222,13 +222,19 @@ install_man: fi export SHELL -install_wrappers: install_bin_libdir +.PHONY: install_wrappers +install_wrappers: install_bin_libdir install_hsc2hs_wrapper @echo "Installing wrapper scripts" $(INSTALL_DIR) "$(DESTDIR)$(WrapperBinsDir)" for p in `cd wrappers; $(FIND) . ! -type d`; do \ mk/install_script.sh "$$p" "$(DESTDIR)/$(WrapperBinsDir)/$$p" "$(WrapperBinsDir)" "$(ActualBinsDir)" "$(ActualBinsDir)/$$p" "$(ActualLibsDir)" "$(docdir)" "$(includedir)"; \ done +.PHONY: install_hsc2hs_wrapper +install_hsc2hs_wrapper: + @echo Copying hsc2hs wrapper + cp mk/hsc2hs wrappers/hsc2hs-ghc-$(ProjectVersion) + PKG_CONFS = $(shell find "$(DESTDIR)$(ActualLibsDir)/package.conf.d" -name '*.conf' | sed "s: :\0xxx\0:g") update_package_db: install_bin install_lib @echo "Installing C++ standard library virtual package" ===================================== hadrian/src/Rules/BinaryDist.hs ===================================== @@ -361,6 +361,7 @@ bindistInstallFiles = , "mk" -/- "project.mk" , "mk" -/- "relpath.sh" , "mk" -/- "system-cxx-std-lib-1.0.conf.in" + , "mk" -/- "hsc2hs.in" , "mk" -/- "install_script.sh" , "README", "INSTALL" ] @@ -425,17 +426,8 @@ haddockWrapper = pure $ "exec \"$executablename\" -B\"$libdir\" -l\"$libdir\" ${ commonWrapper :: Action String commonWrapper = pure $ "exec \"$executablename\" ${1+\"$@\"}\n" --- echo 'HSC2HS_EXTRA="$(addprefix --cflag=,$(CONF_CC_OPTS_STAGE1)) $(addprefix --lflag=,$(CONF_GCC_LINKER_OPTS_STAGE1))"' >> "$(WRAPPER)" hsc2hsWrapper :: Action String -hsc2hsWrapper = do - ccArgs <- map ("--cflag=" <>) <$> settingList (ConfCcArgs Stage1) - ldFlags <- map ("--lflag=" <>) <$> settingList (ConfGccLinkerArgs Stage1) - wrapper <- drop 4 . lines <$> liftIO (readFile "utils/hsc2hs/hsc2hs.wrapper") - return $ unlines - ( "HSC2HS_EXTRA=\"" <> unwords (ccArgs ++ ldFlags) <> "\"" - : "tflag=\"--template=$libdir/template-hsc.h\"" - : "Iflag=\"-I$includedir/\"" - : wrapper ) +hsc2hsWrapper = return "Copied from mk/hsc2hs" runGhcWrapper :: Action String runGhcWrapper = pure $ "exec \"$executablename\" -f \"$exedir/ghc\" ${1+\"$@\"}\n" ===================================== libraries/array ===================================== @@ -1 +1 @@ -Subproject commit 0daca5dfa33d6c522e9fb8e94a2b66a5ed658c20 +Subproject commit c9cb2c1e8762aa83b6e77af82c87a55e03e990e4 ===================================== libraries/base/tests/all.T ===================================== @@ -232,8 +232,12 @@ test('T9681', normal, compile_fail, ['']) # Probably something like 1s is already enough, but I don't know enough to # make an educated guess how long it needs to be guaranteed to reach the C # call." +# +# We ignore stderr since the test itself may print "Killed: 9" (see #24361); +# all we care about is that the test timed out, for which the +# exit_code check is sufficient. test('T8089', - [exit_code(99), run_timeout_multiplier(0.01)], + [exit_code(99), ignore_stderr, run_timeout_multiplier(0.01)], compile_and_run, ['']) test('T8684', expect_broken(8684), compile_and_run, ['']) test('hWaitForInput-accurate-stdin', [js_broken(22349), expect_broken_for(16535, threaded_ways), req_process], compile_and_run, ['']) ===================================== libraries/deepseq ===================================== @@ -1 +1 @@ -Subproject commit 045cee4801ce6a66e9992bff648d951d8e5fcd68 +Subproject commit 7ce6e2d3760b23336fd5f9a36f50df6571606947 ===================================== libraries/process ===================================== @@ -1 +1 @@ -Subproject commit 3466b14dacddc4628427c4d787482899dd0b17cd +Subproject commit 7a55e3fa1bc80f82fedea3b2b2467a0af3a81242 ===================================== libraries/stm ===================================== @@ -1 +1 @@ -Subproject commit a1e91f4ea010be61a9c8a94d6a200bfb4dc6d049 +Subproject commit ff8f8ceeceb14ac59accd53dd82a5d32c7e08626 ===================================== mk/hsc2hs.in ===================================== @@ -0,0 +1,41 @@ +HSC2HS_C="@SettingsCCompilerFlags@" + +HSC2HS_L="@SettingsCCompilerLinkFlags@" + +tflag="--template=$libdir/template-hsc.h" +Iflag="-I$includedir/include/" + +for f in ${HSC2HS_C}; do + cflags="${cflags} --cflag=$f" +done + +for f in ${HSC2HS_L}; do + lflags="${lflags} --lflag=$f" +done + +HSC2HS_EXTRA="$cflags $lflags" + +read_response() { + response_file=$1 + if [ -f "$response_file" ]; then + while read -r arg; do + case "$arg" in + -t*) tflag=;; + --template=*) tflag=;; + @*) read_response "${arg#"@"}" ;; + --) break;; + esac + done < "$response_file" + fi +} + +for arg do + case "$arg" in + -t*) tflag=;; + --template=*) tflag=;; + @*) read_response "${arg#"@"}" ;; + --) break;; + esac +done + +exec "$executablename" ${tflag:+"$tflag"} $HSC2HS_EXTRA ${1+"$@"} "$Iflag" ===================================== testsuite/tests/core-to-stg/T24718.hs ===================================== @@ -0,0 +1,12 @@ +module T24718 where + +import GHC.Exts ( Any ) +import Unsafe.Coerce ( unsafeCoerce ) + +data T = MkT (Any -> Any) + +g :: () -> () +g x = x + +f :: T +f = unsafeCoerce MkT g ===================================== testsuite/tests/core-to-stg/all.T ===================================== @@ -2,3 +2,4 @@ test('T19700', normal, compile, ['-O']) test('T23914', normal, compile, ['-O']) +test('T24718', normal, compile, ['-O']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f15dd7bdbed3fc251e39ca1908552ce1490d0d81...52e82b4d710ad7987cae219eab1dec113a457b53 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f15dd7bdbed3fc251e39ca1908552ce1490d0d81...52e82b4d710ad7987cae219eab1dec113a457b53 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 19:02:48 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 23 Sep 2024 15:02:48 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/package-deps-bytecode-2024-09-23] Package deps bytecode linking Message-ID: <66f1bb58dbfff_173fba5adc006432@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/package-deps-bytecode-2024-09-23 at Glasgow Haskell Compiler / GHC Commits: 56f3fbe9 by Torsten Schmits at 2024-09-23T21:02:37+02:00 Package deps bytecode linking - - - - - 20 changed files: - compiler/GHC/Iface/Load.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/Finder.hs - + testsuite/tests/bytecode/T25090/Dep.hs - + testsuite/tests/bytecode/T25090/DepApi.hs - + testsuite/tests/bytecode/T25090/Local.hs - testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/Num.hs - + testsuite/tests/bytecode/T25090/Num.hs-boot - + testsuite/tests/bytecode/T25090/Obj.hs - + testsuite/tests/bytecode/T25090/PkgBytecode.hs - + testsuite/tests/bytecode/T25090/PkgBytecode.stdout - testsuite/tests/bytecode/T25090/all.T - + testsuite/tests/bytecode/T25090/dep.conf - + testsuite/tests/bytecode/T25090/obj.conf - + testsuite/tests/bytecode/T25090/prep.bash - + testsuite/tests/bytecode/T25090/run.bash - + testsuite/tests/bytecode/T25090/unit1 - + testsuite/tests/bytecode/T25090/unit2 Changes: ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -674,7 +674,11 @@ dontLeakTheHUG thing_inside = do in hsc_env { hsc_targets = panic "cleanTopEnv: hsc_targets" , hsc_mod_graph = panic "cleanTopEnv: hsc_mod_graph" - , hsc_IC = panic "cleanTopEnv: hsc_IC" + -- TODO this is needed for bytecode compilation of package deps + -- only. local EPS modules work fine. + -- Also it appears to work when the local modules use multiple + -- home units?!?!? + -- , hsc_IC = panic "cleanTopEnv: hsc_IC" , hsc_type_env_vars = case maybe_type_vars of Just vars -> vars Nothing -> panic "cleanTopEnv: hsc_type_env_vars" ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -6,6 +6,7 @@ {-# LANGUAGE CPP #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE NamedFieldPuns #-} module GHC.Linker.Deps ( LinkDepsOpts (..) @@ -28,6 +29,7 @@ import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM import GHC.Utils.Outputable +import qualified GHC.Utils.Outputable as Outputable import GHC.Utils.Panic import GHC.Utils.Error @@ -47,9 +49,11 @@ import GHC.Utils.Misc import GHC.Unit.Home import GHC.Data.Maybe -import Control.Monad import Control.Applicative +import Control.Monad.IO.Class (MonadIO (liftIO)) +import Control.Monad.Trans.Except (ExceptT, runExceptT, throwE) +import Data.Foldable (traverse_) import qualified Data.Set as Set import qualified Data.Map as M import Data.List (isSuffixOf) @@ -68,15 +72,16 @@ data LinkDepsOpts = LinkDepsOpts , ldWays :: !Ways -- ^ Enabled ways , ldFinderCache :: !FinderCache , ldFinderOpts :: !FinderOpts - , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) - , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation))) + , ldLoadByteCode :: !(Module -> IO (Maybe (IO Linkable))) + , ldDebugTrace :: !(SDoc -> IO ()) } data LinkDeps = LinkDeps { ldNeededLinkables :: [Linkable] , ldAllLinkables :: [Linkable] - , ldUnits :: [UnitId] - , ldNeededUnits :: UniqDSet UnitId + , ldNeededUnits :: [UnitId] + , ldAllUnits :: UniqDSet UnitId } -- | Find all the packages and linkables that a set of modules depends on @@ -102,7 +107,6 @@ getLinkDeps opts interp pls span mods = do get_link_deps opts pls maybe_normal_osuf span mods - get_link_deps :: LinkDepsOpts -> LoaderState @@ -111,47 +115,48 @@ get_link_deps -> [Module] -> IO LinkDeps get_link_deps opts pls maybe_normal_osuf span mods = do - -- 1. Find the dependent home-pkg-modules/packages from each iface - -- (omitting modules from the interactive package, which is already linked) - (mods_s, pkgs_s) <- - -- Why two code paths here? There is a significant amount of repeated work - -- performed calculating transitive dependencies - -- if --make uses the oneShot code path (see MultiLayerModulesTH_* tests) - if ldOneShotMode opts - then follow_deps (filterOut isInteractiveModule mods) - emptyUniqDSet emptyUniqDSet; - else do - (pkgs, mmods) <- unzip <$> mapM get_mod_info all_home_mods - return (catMaybes mmods, unionManyUniqDSets (init_pkg_set : pkgs)) - - let - -- 2. Exclude ones already linked - -- Main reason: avoid findModule calls in get_linkable - (mods_needed, links_got) = partitionWith split_mods mods_s - pkgs_needed = eltsUDFM $ getUniqDSet pkgs_s `minusUDFM` pkgs_loaded pls - - split_mods mod = - let is_linked = lookupModuleEnv (objs_loaded pls) mod - <|> lookupModuleEnv (bcos_loaded pls) mod - in case is_linked of - Just linkable -> Right linkable - Nothing -> Left mod - - -- 3. For each dependent module, find its linkable - -- This will either be in the HPT or (in the case of one-shot - -- compilation) we may need to use maybe_getFileLinkable - lnks_needed <- mapM (get_linkable (ldObjSuffix opts)) mods_needed - - return $ LinkDeps - { ldNeededLinkables = lnks_needed - , ldAllLinkables = links_got ++ lnks_needed - , ldUnits = pkgs_needed - , ldNeededUnits = pkgs_s - } + -- 1. Find the dependent home-pkg-modules/packages from each iface + -- (omitting modules from the interactive package, which is already linked) + -- Why two code paths here? There is a significant amount of repeated work + -- performed calculating transitive dependencies + -- if --make uses the oneShot code path (see MultiLayerModulesTH_* tests) + deps <- if ldOneShotMode opts + then oneshot_deps opts (filterOut isInteractiveModule mods) + else make_deps + + -- 2. Exclude ones already linked + -- Main reason: avoid findModule calls in get_linkable + -- TODO outdated + let (loaded_modules, needed_modules, ldAllUnits, ldNeededUnits) = + classify_deps pls deps + + -- 3. For each dependent module, find its linkable + -- This will either be in the HPT or (in the case of one-shot + -- compilation) we may need to use maybe_getFileLinkable + -- TODO outdated + ldNeededLinkables <- mapM module_linkable needed_modules + + pure LinkDeps { + ldNeededLinkables, + ldAllLinkables = loaded_modules ++ ldNeededLinkables, + ldNeededUnits, + ldAllUnits + } where mod_graph = ldModuleGraph opts unit_env = ldUnitEnv opts + make_deps = do + (pkgs, mmods) <- unzip <$> mapM get_mod_info all_home_mods + let + link_mods = + listToUDFM [(moduleName (mi_module (hm_iface m)), m) | m <- mmods] + link_libs = + uniqDSetToList (unionManyUniqDSets (init_pkg_set : pkgs)) + pure $ + LinkModules (LinkHomeModule <$> link_mods) : + (LinkLibrary <$> link_libs) + -- This code is used in `--make` mode to calculate the home package and unit dependencies -- for a set of modules. -- @@ -183,73 +188,14 @@ get_link_deps opts pls maybe_normal_osuf span mods = do get_mod_info (ModNodeKeyWithUid gwib uid) = case lookupHug (ue_home_unit_graph unit_env) uid (gwib_mod gwib) of - Just hmi -> - let iface = (hm_iface hmi) - mmod = case mi_hsc_src iface of - HsBootFile -> link_boot_mod_error (mi_module iface) - _ -> return $ Just (mi_module iface) - - in (mkUniqDSet $ Set.toList $ dep_direct_pkgs (mi_deps iface),) <$> mmod + Just hmi -> do + let iface = hm_iface hmi + case mi_hsc_src iface of + HsBootFile -> throwProgramError opts $ link_boot_mod_error (mi_module iface) + _ -> pure (mkUniqDSet $ Set.toList $ dep_direct_pkgs (mi_deps iface), hmi) Nothing -> throwProgramError opts $ text "getLinkDeps: Home module not loaded" <+> ppr (gwib_mod gwib) <+> ppr uid - - -- This code is used in one-shot mode to traverse downwards through the HPT - -- to find all link dependencies. - -- The ModIface contains the transitive closure of the module dependencies - -- within the current package, *except* for boot modules: if we encounter - -- a boot module, we have to find its real interface and discover the - -- dependencies of that. Hence we need to traverse the dependency - -- tree recursively. See bug #936, testcase ghci/prog007. - follow_deps :: [Module] -- modules to follow - -> UniqDSet Module -- accum. module dependencies - -> UniqDSet UnitId -- accum. package dependencies - -> IO ([Module], UniqDSet UnitId) -- result - follow_deps [] acc_mods acc_pkgs - = return (uniqDSetToList acc_mods, acc_pkgs) - follow_deps (mod:mods) acc_mods acc_pkgs - = do - mb_iface <- ldLoadIface opts msg mod - iface <- case mb_iface of - Failed err -> throwProgramError opts $ - missingInterfaceErrorDiagnostic (ldMsgOpts opts) err - Succeeded iface -> return iface - - when (mi_boot iface == IsBoot) $ link_boot_mod_error mod - - let - pkg = moduleUnit mod - deps = mi_deps iface - - pkg_deps = dep_direct_pkgs deps - (boot_deps, mod_deps) = flip partitionWith (Set.toList (dep_direct_mods deps)) $ - \case - (_, GWIB m IsBoot) -> Left m - (_, GWIB m NotBoot) -> Right m - - mod_deps' = case ue_homeUnit unit_env of - Nothing -> [] - Just home_unit -> filter (not . (`elementOfUniqDSet` acc_mods)) (map (mkHomeModule home_unit) $ (boot_deps ++ mod_deps)) - acc_mods' = case ue_homeUnit unit_env of - Nothing -> acc_mods - Just home_unit -> addListToUniqDSet acc_mods (mod : map (mkHomeModule home_unit) mod_deps) - acc_pkgs' = addListToUniqDSet acc_pkgs (Set.toList pkg_deps) - - case ue_homeUnit unit_env of - Just home_unit | isHomeUnit home_unit pkg -> follow_deps (mod_deps' ++ mods) - acc_mods' acc_pkgs' - _ -> follow_deps mods acc_mods (addOneToUniqDSet acc_pkgs' (toUnitId pkg)) - where - msg = text "need to link module" <+> ppr mod <+> - text "due to use of Template Haskell" - - - - link_boot_mod_error :: Module -> IO a - link_boot_mod_error mod = throwProgramError opts $ - text "module" <+> ppr mod <+> - text "cannot be linked; it is only available as a boot module" - no_obj :: Outputable a => a -> IO b no_obj mod = dieWith opts span $ text "cannot find object file for module " <> @@ -258,6 +204,18 @@ get_link_deps opts pls maybe_normal_osuf span mods = do while_linking_expr = text "while linking an interpreted expression" + module_linkable = \case + LinkHomeModule hmi -> + adjust_linkable (expectJust "getLinkDeps" (homeModLinkable hmi)) + + LinkObjectModule iface loc -> do + let mod = mi_module iface + findObjectLinkableMaybe mod loc >>= \case + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + + LinkByteCodeModule _ load_bytecode -> + load_bytecode -- See Note [Using Byte Code rather than Object Code for Template Haskell] homeModLinkable :: HomeModInfo -> Maybe Linkable @@ -266,57 +224,256 @@ get_link_deps opts pls maybe_normal_osuf span mods = do then homeModInfoByteCode hmi <|> homeModInfoObject hmi else homeModInfoObject hmi <|> homeModInfoByteCode hmi - get_linkable osuf mod -- A home-package module - | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) - = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) - | otherwise - = do -- It's not in the HPT because we are in one shot mode, - -- so use the Finder to get a ModLocation... - case ue_homeUnit unit_env of - Nothing -> no_obj mod - Just home_unit -> do - from_bc <- ldLoadByteCode opts mod - maybe (fallback_no_bytecode home_unit mod) pure from_bc - where - - fallback_no_bytecode home_unit mod = do - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc _ -> do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - _ -> no_obj (moduleName mod) - - adjust_linkable lnk - | Just new_osuf <- maybe_normal_osuf = do - new_parts <- mapM (adjust_part new_osuf) - (linkableParts lnk) - return lnk{ linkableParts=new_parts } - | otherwise = - return lnk - - adjust_part new_osuf part = case part of - DotO file ModuleObject -> do - massert (osuf `isSuffixOf` file) - let file_base = fromJust (stripExtension osuf file) - new_file = file_base <.> new_osuf - ok <- doesFileExist new_file - if (not ok) - then dieWith opts span $ - text "cannot find object file " - <> quotes (text new_file) $$ while_linking_expr - else return (DotO new_file ModuleObject) - DotO file ForeignObject -> pure (DotO file ForeignObject) - DotA fp -> panic ("adjust_ul DotA " ++ show fp) - DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) - BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) + adjust_linkable lnk + | Just new_osuf <- maybe_normal_osuf = do + new_uls <- mapM (adjust_part (ldObjSuffix opts) new_osuf) + (linkableParts lnk) + return lnk {linkableParts = new_uls} + | otherwise = + return lnk + + adjust_part osuf new_osuf part = case part of + DotO file ModuleObject -> do + massert (osuf `isSuffixOf` file) + let file_base = fromJust (stripExtension osuf file) + new_file = file_base <.> new_osuf + ok <- doesFileExist new_file + if (not ok) + then dieWith opts span $ + text "cannot find object file " + <> quotes (text new_file) $$ while_linking_expr + else return (DotO new_file ModuleObject) + DotO file ForeignObject -> pure (DotO file ForeignObject) + DotA fp -> panic ("adjust_part DotA " ++ show fp) + DotDLL fp -> panic ("adjust_part DotDLL " ++ show fp) + BCOs {} -> pure part + LazyBCOs{} -> pure part + CoreBindings WholeCoreBindings {wcb_module} -> + pprPanic "Unhydrated core bindings" (ppr wcb_module) + + -- get_linkable osuf mod -- A home-package module + -- | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) + -- = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) + -- | otherwise + -- = do -- It's not in the HPT because we are in one shot mode, + -- -- so use the Finder to get a ModLocation... + -- case ue_homeUnit unit_env of + -- Nothing -> no_obj mod + -- Just home_unit -> do + -- from_bc <- ldLoadByteCode opts mod + -- maybe (fallback_no_bytecode home_unit mod) pure from_bc + -- where + -- + -- fallback_no_bytecode home_unit mod = do + -- let fc = ldFinderCache opts + -- let fopts = ldFinderOpts opts + -- mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + -- case mb_stuff of + -- Found loc _ -> do + -- mb_lnk <- findObjectLinkableMaybe mod loc + -- case mb_lnk of + -- Nothing -> no_obj mod + -- Just lnk -> adjust_linkable lnk + -- _ -> no_obj (moduleName mod) + +data LinkModule = + LinkHomeModule HomeModInfo + | + LinkObjectModule ModIface ModLocation + | + LinkByteCodeModule ModIface (IO Linkable) + +link_module_iface :: LinkModule -> ModIface +link_module_iface = \case + LinkHomeModule hmi -> hm_iface hmi + LinkObjectModule iface _ -> iface + LinkByteCodeModule iface _ -> iface + +instance Outputable LinkModule where + ppr = \case + LinkHomeModule hmi -> ppr (mi_module (hm_iface hmi)) <+> brackets (text "HMI") + LinkObjectModule iface _ -> ppr (mi_module iface) + LinkByteCodeModule iface _ -> ppr (mi_module iface) <+> brackets (text "BC") + +data LinkDep = + LinkModules (UniqDFM ModuleName LinkModule) + | + LinkLibrary UnitId + +instance Outputable LinkDep where + ppr = \case + LinkModules mods -> text "modules:" <+> ppr (eltsUDFM mods) + LinkLibrary uid -> text "library:" <+> ppr uid + +data OneshotError = + NoLocation Module + | + NoInterface MissingInterfaceError + | + LinkBootModule Module + +-- This code is used in one-shot mode to traverse downwards through the HPT +-- to find all link dependencies. +-- The ModIface contains the transitive closure of the module dependencies +-- within the current package, *except* for boot modules: if we encounter +-- a boot module, we have to find its real interface and discover the +-- dependencies of that. Hence we need to traverse the dependency +-- tree recursively. See bug #936, testcase ghci/prog007. +oneshot_deps :: + LinkDepsOpts -> + -- | Modules whose imports to follow + [Module] -> + IO [LinkDep] +oneshot_deps opts mods = + runExceptT (oneshot_deps_loop opts mods emptyUDFM) >>= \case + Right a -> pure (eltsUDFM a) + Left err -> throwProgramError opts (message err) + where + message = \case + NoLocation mod -> + pprPanic "found iface but no location" (ppr mod) + NoInterface err -> + missingInterfaceErrorDiagnostic (ldMsgOpts opts) err + LinkBootModule mod -> + link_boot_mod_error mod + +oneshot_deps_loop :: + LinkDepsOpts -> + [Module] -> + UniqDFM UnitId LinkDep -> + ExceptT OneshotError IO (UniqDFM UnitId LinkDep) +oneshot_deps_loop _ [] acc = + pure acc +oneshot_deps_loop opts (mod : mods) acc = do + (new_acc, new_mods, action) <- process_module + traverse_ debug_log action + oneshot_deps_loop opts (new_mods ++ mods) new_acc + where + debug_log action = + liftIO $ ldDebugTrace opts $ + text "TH dep" <+> ppr mod <+> brackets (sep [ + if is_home then text "home" else Outputable.empty, + text action + ]) + + process_module + | already_seen = pure (acc, [], Nothing) + | is_home || bytecode = try_iface + | otherwise = add_library + + already_seen + | Just (LinkModules mods) <- mod_dep + = elemUDFM mod_name mods + | Just (LinkLibrary _) <- mod_dep + = True + | otherwise + = False + + try_iface = + liftIO (ldLoadIface opts load_reason mod) >>= \case + Failed err -> throwE (NoInterface err) + Succeeded (iface, loc) -> do + mb_load_bc <- liftIO (ldLoadByteCode opts (mi_module iface)) + with_iface loc iface mb_load_bc + + with_iface loc iface mb_load_bc + | IsBoot <- mi_boot iface + = throwE (LinkBootModule mod) + + | bytecode + , Just load_bc <- mb_load_bc + = pure (add_module iface (LinkByteCodeModule iface load_bc) "bytecode") + + | is_home + = pure (add_module iface (LinkObjectModule iface loc) "object") + + | otherwise + = add_library + + add_library = + pure (addToUDFM acc mod_unit_id (LinkLibrary mod_unit_id), [], Just "library") + + add_module iface lmod action = + (addListToUDFM with_mod (direct_pkgs iface), new_deps iface, Just action) + where + with_mod = alterUDFM (add_package_module lmod) acc mod_unit_id + + add_package_module lmod = \case + Just (LinkLibrary u) -> Just (LinkLibrary u) + Just (LinkModules old) -> Just (LinkModules (addToUDFM old mod_name lmod)) + Nothing -> Just (LinkModules (unitUDFM mod_name lmod)) + + direct_pkgs iface + | bytecode + = [] + | otherwise + = [(u, LinkLibrary u) | u <- Set.toList (dep_direct_pkgs (mi_deps iface))] + + new_deps iface + | bytecode + -- TODO How can we better determine the external deps? + = [usg_mod | UsagePackageModule {usg_mod} <- mi_usages iface] ++ local + | is_home + = local + | otherwise + = [] + where + local = + [ + mkModule mod_unit m + -- TODO Somehow this just works, no idea what the deal was in the + -- old code with boot modules. + | (_, GWIB m _) <- Set.toList (dep_direct_mods (mi_deps iface)) + ] + + is_home + | Just home <- mb_home + = homeUnitAsUnit home == mod_unit + | otherwise + = False + + mod_dep = lookupUDFM acc mod_unit_id + mod_name = moduleName mod + mod_unit_id = moduleUnitId mod + mod_unit = moduleUnit mod + load_reason = + text "need to link module" <+> ppr mod <+> + text "due to use of Template Haskell" + + bytecode = ldUseByteCode opts + mb_home = ue_homeUnit (ldUnitEnv opts) + +link_boot_mod_error :: Module -> SDoc +link_boot_mod_error mod = + text "module" <+> ppr mod <+> + text "cannot be linked; it is only available as a boot module" + +classify_deps :: + LoaderState -> + [LinkDep] -> + ([Linkable], [LinkModule], UniqDSet UnitId, [UnitId]) +classify_deps pls deps = + (loaded_modules, needed_modules, all_packages, needed_packages) + where + (loaded_modules, needed_modules) = + partitionWith loaded_or_needed (concatMap eltsUDFM modules) + + needed_packages = + eltsUDFM (getUniqDSet all_packages `minusUDFM` pkgs_loaded pls) + + all_packages = mkUniqDSet packages + + (modules, packages) = flip partitionWith deps $ \case + LinkModules mods -> Left mods + LinkLibrary lib -> Right lib + + loaded_or_needed lm = + maybe (Right lm) Left (loaded_linkable (mi_module (link_module_iface lm))) + + loaded_linkable mod = + lookupModuleEnv (objs_loaded pls) mod + <|> + lookupModuleEnv (bcos_loaded pls) mod {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -77,11 +77,13 @@ import GHC.Utils.TmpFs import GHC.Unit.Env import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) +import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.State as Packages import qualified GHC.Data.ShortText as ST import GHC.Data.FastString +import qualified GHC.Data.Maybe as Maybe import GHC.Linker.Deps import GHC.Linker.MacOS @@ -94,6 +96,7 @@ import Control.Monad import qualified Data.Set as Set import Data.Char (isSpace) import qualified Data.Foldable as Foldable +import Data.Functor ((<&>)) import Data.IORef import Data.List (intercalate, isPrefixOf, nub, partition) import Data.Maybe @@ -231,10 +234,10 @@ loadDependencies interp hsc_env pls span needed_mods = do -- Find what packages and linkables are required deps <- getLinkDeps opts interp pls span needed_mods - let this_pkgs_needed = ldNeededUnits deps + let this_pkgs_needed = ldAllUnits deps -- Link the packages and modules required - pls1 <- loadPackages' interp hsc_env (ldUnits deps) pls + pls1 <- loadPackages' interp hsc_env (ldNeededUnits deps) pls (pls2, succ) <- loadModuleLinkables interp hsc_env pls1 (ldNeededLinkables deps) let this_pkgs_loaded = udfmRestrictKeys all_pkgs_loaded $ getUniqDSet trans_pkgs_needed all_pkgs_loaded = pkgs_loaded pls2 @@ -645,19 +648,40 @@ initLinkDepsOpts hsc_env = opts , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags + -- , ldHugFinderOpts = initFinderOpts . homeUnitEnv_dflags <$> hsc_HUG hsc_env , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags , ldLoadIface , ldLoadByteCode + , ldDebugTrace = debugTraceMsg (hsc_logger hsc_env) 3 } dflags = hsc_dflags hsc_env - ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env - $ loadInterface msg mod (ImportByUser NotBoot) + + ldLoadIface msg mod = + initIfaceCheck (text "loader") hsc_env (loadInterface msg mod (ImportByUser NotBoot)) >>= \case + Maybe.Failed err -> pure (Maybe.Failed err) + Maybe.Succeeded iface -> + find_location mod <&> \case + InstalledFound loc _ -> Maybe.Succeeded (iface, loc) + err -> Maybe.Failed $ + cannotFindInterface unit_state home_unit + (targetProfile dflags) (moduleName mod) err + + find_location mod = + liftIO $ + findExactModule (hsc_FC hsc_env) (initFinderOpts dflags) + (initFinderOpts . homeUnitEnv_dflags <$> hsc_HUG hsc_env) + unit_state home_unit + (toUnitId <$> mod) + + unit_state = hsc_units hsc_env + + home_unit = ue_homeUnit (hsc_unit_env hsc_env) ldLoadByteCode mod = do EPS {eps_iface_bytecode} <- hscEPS hsc_env - sequence (lookupModuleEnv eps_iface_bytecode mod) + pure (lookupModuleEnv eps_iface_bytecode mod) @@ -1347,6 +1371,14 @@ locateLib interp hsc_env is_hs lib_dirs gcc_dirs lib0 | loading_dynamic_hs_libs -- search for .so libraries first. = findHSDll `orElse` findDynObject `orElse` + -- TODO Added the next two because static archives would be ignored when + -- linking bytecode. + -- Not sure if this is expected, but the condition for this branch is just + -- that ghc is _capable_ of dynamic linking, so maybe this function was just + -- never used when linking TH before this patch? + -- Add a test to the oneshot base MR that uses archives to compare. + findObject `orElse` + findArchive `orElse` assumeDll | otherwise ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -734,7 +734,7 @@ mkStubPaths fopts mod location stub_basename <.> os "h" -- ----------------------------------------------------------------------------- --- findLinkable isn't related to the other stuff in here, +-- findObjectLinkable isn't related to the other stuff in here, -- but there's no other obvious place for it findObjectLinkableMaybe :: Module -> ModLocation -> IO (Maybe Linkable) ===================================== testsuite/tests/bytecode/T25090/Dep.hs ===================================== @@ -0,0 +1,15 @@ +module Dep where + +data A = A Int + +used :: Int +used = 9681 + +dep :: A +dep = A used + +unused1 :: A +unused1 = A 1 + +unused2 :: A +unused2 = unused1 ===================================== testsuite/tests/bytecode/T25090/DepApi.hs ===================================== @@ -0,0 +1,7 @@ +module DepApi (A (A), dep) where + +import Dep (A (A)) +import qualified Dep + +dep :: A +dep = Dep.dep ===================================== testsuite/tests/bytecode/T25090/Local.hs ===================================== @@ -0,0 +1,16 @@ +{-# language PackageImports #-} + +module Local where + +import GHC.Prim +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +-- just to be sure that the file isn't accidentally picked up locally +import "dep" DepApi (dep, A (A)) +import {-# source #-} Num (num) +import Obj (numo) + +splc :: ExpQ +splc = lift @_ @Int (num + d + numo) + where + A d = dep ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -19,3 +19,37 @@ T25090a: T25090b: $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 ./exe + +# TODO it works even without -package obj, but it should complain about the package not being exposed +DB := -package-db db -package dep +BASIC := $(TEST_HC_OPTS) $(DB) -this-unit-id=pkgdep -v0 +BC := -fprefer-byte-code -fbyte-code-and-object-code +ARGS := $(BASIC) $(BC) + +.PHONY: PkgArchive +PkgArchive: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" 1 + ./run.bash "$(TEST_HC)" "$(ARGS)" + +.PHONY: PkgEmptyArchive +PkgEmptyArchive: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" 2 + ./run.bash "$(TEST_HC)" "$(ARGS)" + +.PHONY: PkgNoArchive +PkgNoArchive: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" 3 + ./run.bash "$(TEST_HC)" "$(ARGS)" + +.PHONY: PkgArchiveObjCode +PkgArchiveObjCode: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" 1 + ./run.bash "$(TEST_HC)" "$(BASIC) -fbyte-code-and-object-code" + +.PHONY: PkgMultiUnit +PkgMultiUnit: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" 1 + mkdir -p unit2-src/ + mv Local.hs Num.hs Num.hs-boot unit2-src/ + "$(TEST_HC)" $(TEST_HC_OPTS) $(ARGS) -unit @unit1 -unit @unit2 + ./PkgBytecode ===================================== testsuite/tests/bytecode/T25090/Num.hs ===================================== @@ -0,0 +1,4 @@ +module Num where + +num :: Int +num = 48332 ===================================== testsuite/tests/bytecode/T25090/Num.hs-boot ===================================== @@ -0,0 +1,3 @@ +module Num where + +num :: Int ===================================== testsuite/tests/bytecode/T25090/Obj.hs ===================================== @@ -0,0 +1,4 @@ +module Obj where + +numo :: Int +numo = 0 ===================================== testsuite/tests/bytecode/T25090/PkgBytecode.hs ===================================== @@ -0,0 +1,12 @@ +{-# language TemplateHaskell #-} + +module Main where + +import GHC.Prim +import Local (splc) + +a :: Int +a = $(splc) + +main :: IO () +main = putStrLn (show a) ===================================== testsuite/tests/bytecode/T25090/PkgBytecode.stdout ===================================== @@ -0,0 +1 @@ +58013 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -16,3 +16,34 @@ def test_T25090(name): test_T25090('T25090a') test_T25090('T25090b') + +def test_pkg(suf, files = []): + name = f'Pkg{suf}' + test( + name, + [ + extra_files([ + 'PkgBytecode.hs', + 'Local.hs', + 'Dep.hs', + 'DepApi.hs', + 'Num.hs', + 'Num.hs-boot', + 'Obj.hs', + 'dep.conf', + 'obj.conf', + 'prep.bash', + 'run.bash', + ] + files), + use_specs({'stdout': 'PkgBytecode.stdout'}), + ], + makefile_test, + [name], + ) + +test_pkg('Archive') +test_pkg('EmptyArchive') +test_pkg('NoArchive') +# TODO broken +# test_pkg('ArchiveObjCode') +test_pkg('MultiUnit', ['unit1', 'unit2']) ===================================== testsuite/tests/bytecode/T25090/dep.conf ===================================== @@ -0,0 +1,8 @@ +name: dep +version: 1.0 +id: dep-1.0 +key: dep-1.0 +exposed: True +exposed-modules: DepApi +import-dirs: ${pkgroot}/dep +library-dirs: ${pkgroot}/dep ===================================== testsuite/tests/bytecode/T25090/obj.conf ===================================== @@ -0,0 +1,8 @@ +name: obj +version: 1.0 +id: obj-1.0 +key: obj-1.0 +exposed: True +exposed-modules: Obj +import-dirs: ${pkgroot}/obj +library-dirs: ${pkgroot}/obj ===================================== testsuite/tests/bytecode/T25090/prep.bash ===================================== @@ -0,0 +1,52 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_opts="$2" +ghc_pkg_cmd="$3" +archive="$4" + +base="$PWD" +db="$base/db" +dep="$base/dep" +conf_dep="${dep}/dep.conf" +obj="$base/obj" +conf_obj="${obj}/obj.conf" + +ghc_pkg() +{ + eval "${ghc_pkg_cmd at Q} --no-user-package-db --package-db=${db at Q} $@" +} + +ghc() +{ + eval "${ghc_cmd at Q} $ghc_opts $@" +} + +mkdir -p "$dep" "$obj" "$db" +mv Dep.hs DepApi.hs "$dep/" +cp dep.conf "$dep/" +mv Obj.hs "$obj/" +cp obj.conf "$obj/" + +ghc_pkg recache + +ghc "-package-db ${db at Q} -hidir ${dep at Q} -O0 -this-unit-id dep-1.0 -fbyte-code-and-object-code -c ${dep at Q}/Dep.hs ${dep at Q}/DepApi.hs" + +ghc "-package-db ${db at Q} -hidir ${obj at Q} -O0 -this-unit-id obj-1.0 -c ${obj at Q}/Obj.hs" +$AR cqs "${obj}/libHSobj-1.0.a" "${obj}/Obj.o" +echo 'hs-libraries: HSobj-1.0' >> "$conf_obj" + +if [[ "$archive" == 1 ]] +then + $AR cqs "${dep}/libHSdep-1.0.a" "${dep}/Dep.o" "${dep}/DepApi.o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$archive" == 2 ]] +then + $AR cqs "${dep}/libHSdep-1.0.a" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +fi + +ghc_pkg -v0 register "${conf_dep at Q}" +ghc_pkg -v0 register "${conf_obj at Q}" ===================================== testsuite/tests/bytecode/T25090/run.bash ===================================== @@ -0,0 +1,16 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_opts="$2" + +ghc() +{ + eval "${ghc_cmd at Q} $ghc_opts $@" +} + +ghc -c Num.hs-boot Num.hs Local.hs +ghc -c PkgBytecode.hs +ghc PkgBytecode.o -o PkgBytecode +./PkgBytecode ===================================== testsuite/tests/bytecode/T25090/unit1 ===================================== @@ -0,0 +1 @@ +-i -i. PkgBytecode -this-unit-id unit1 -package-id unit2 ===================================== testsuite/tests/bytecode/T25090/unit2 ===================================== @@ -0,0 +1 @@ +-i -i./unit2-src Local Num -this-unit-id unit2 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/56f3fbe9f747725dca63fc7420868fcb7fe9ee4a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/56f3fbe9f747725dca63fc7420868fcb7fe9ee4a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 23 22:19:20 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 23 Sep 2024 18:19:20 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25281 Message-ID: <66f1e9682a362_236f4e7a68e078725@gitlab.mail> Simon Peyton Jones pushed new branch wip/T25281 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25281 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 02:56:24 2024 From: gitlab at gitlab.haskell.org (Matthew Craven (@clyring)) Date: Mon, 23 Sep 2024 22:56:24 -0400 Subject: [Git][ghc/ghc][wip/T23490-part2] 61 commits: AArch64: Implement takeRegRegMoveInstr Message-ID: <66f22a5864aca_236f4e11671189219d@gitlab.mail> Matthew Craven pushed to branch wip/T23490-part2 at Glasgow Haskell Compiler / GHC Commits: 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 54c6256e by Matthew Craven at 2024-09-23T22:51:31-04:00 Bump transformers submodule The svg image files mentioned in transformers.cabal were previously not checked in, which broke sdist generation. - - - - - bde588c0 by Matthew Craven at 2024-09-23T22:51:31-04:00 Remove reference to non-existent file in haddock.cabal - - - - - 21fde253 by Matthew Craven at 2024-09-23T22:51:31-04:00 Move tests T11462 and T11525 into tests/tcplugins - - - - - 068c301c by Matthew Craven at 2024-09-23T22:51:32-04:00 Repair the 'build-cabal' hadrian target Fixes #23117. Fixes #23281. Fixes #23490. This required: * Updating the bit-rotted compiler/Setup.hs and its setup-depends * Listing a few recently-added libraries and utilities in cabal.project-reinstall * Setting allow-boot-library-installs to 'True' since Cabal now considers the 'ghc' package itself a boot library for the purposes of this flag Additionally, the allow-newer block in cabal.project-reinstall was removed. This block was probably added because when the libraries/Cabal submodule is too new relative to the cabal-install executable, solving the setup-depends for any package with a custom setup requires building an old Cabal (from Hackage) against the in-tree version of base, and this can fail un-necessarily due to tight version bounds on base. However, the blind allow-newer can also cause the solver to go berserk and choose a stupid build plan that has no business succeeding, and the failures when this happens are dreadfully confusing. (See #23281 and #24363.) Why does setup-depends solving insist on an old version of Cabal? See: https://github.com/haskell/cabal/blob/0a0b33983b0f022b9697f7df3a69358ee9061a89/cabal-install/src/Distribution/Client/ProjectPlanning.hs#L1393-L1410 The right solution here is probably to use the in-tree cabal-install from libraries/Cabal/cabal-install with the build-cabal target rather than whatever the environment happens to provide. But this is left for future work. - - - - - f3b81447 by Matthew Craven at 2024-09-23T22:51:32-04:00 Revert "CI: Disable the test-cabal-reinstall job" This reverts commit 38c3afb64d3ffc42f12163c6f0f0d5c414aa8255. - - - - - 18 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - CODEOWNERS - cabal.project-reinstall - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/25d52bf9c59431642f27a7a9e49e65ea013c4cfb...f3b81447cb71a89ca5c547f1c7e854cd0516b024 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/25d52bf9c59431642f27a7a9e49e65ea013c4cfb...f3b81447cb71a89ca5c547f1c7e854cd0516b024 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 04:24:04 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 24 Sep 2024 00:24:04 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 20 commits: The X86 SIMD patch. Message-ID: <66f23ee47402e_2f8ade45f6b415949@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: b6c8783d by sheaf at 2024-09-24T00:23:37-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 195cc08d by sheaf at 2024-09-24T00:23:37-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - a1b40531 by sheaf at 2024-09-24T00:23:37-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 27fdfbbb by sheaf at 2024-09-24T00:23:38-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 1991af34 by sheaf at 2024-09-24T00:23:38-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 1d1a265e by sheaf at 2024-09-24T00:23:38-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - a11a15fa by sheaf at 2024-09-24T00:23:38-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 032b2e01 by sheaf at 2024-09-24T00:23:38-04:00 Add test for C calls & SIMD vectors - - - - - e2e3c619 by sheaf at 2024-09-24T00:23:38-04:00 Add test for #25169 - - - - - 805cf1bf by sheaf at 2024-09-24T00:23:38-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 5cca7265 by sheaf at 2024-09-24T00:23:38-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 73093d86 by sheaf at 2024-09-24T00:23:38-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - eb482a39 by sheaf at 2024-09-24T00:23:38-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - f37664a5 by sheaf at 2024-09-24T00:23:38-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 1bfe3e2e by sheaf at 2024-09-24T00:23:38-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - c1b8b91a by Brandon Chinn at 2024-09-24T00:23:39-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 82b09fc9 by ARATA Mizuki at 2024-09-24T00:23:42-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 5cc10e20 by Hécate Kleidukos at 2024-09-24T00:23:48-04:00 Add setup for tags generation This commit creates a top-level Makefile and a configuration file for the ghc-tags program. closes #25239 - - - - - 9fbb2087 by Simon Peyton Jones at 2024-09-24T00:23:48-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - ced2b207 by Alan Zimmerman at 2024-09-24T00:23:49-04:00 EPA: Remove unused hsCaseAnnsRest We never populate it, so remove it. - - - - - 30 changed files: - .gitignore - + Makefile - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6985e5669ce660fb4ad315025e0fc305317a6057...ced2b20725ba965f8ea2ae3ff494d9b162fdef6f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/6985e5669ce660fb4ad315025e0fc305317a6057...ced2b20725ba965f8ea2ae3ff494d9b162fdef6f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 06:48:01 2024 From: gitlab at gitlab.haskell.org (Fabricio Nascimento (@fabu)) Date: Tue, 24 Sep 2024 02:48:01 -0400 Subject: [Git][ghc/ghc][wip/fabu/T25014-mistakenly-accepted-parent] Apply 1 suggestion(s) to 1 file(s) Message-ID: <66f260a1378dc_2f8ade914f244065d@gitlab.mail> Fabricio Nascimento pushed to branch wip/fabu/T25014-mistakenly-accepted-parent at Glasgow Haskell Compiler / GHC Commits: 655f9b9f by Fabricio Nascimento at 2024-09-24T06:47:58+00:00 Apply 1 suggestion(s) to 1 file(s) Co-authored-by: sheaf <sam.derbyshire at gmail.com> - - - - - 1 changed file: - compiler/GHC/Rename/Env.hs Changes: ===================================== compiler/GHC/Rename/Env.hs ===================================== @@ -720,6 +720,39 @@ Test cases: - T11970 (both cases) - T25014{a,b,c,d,e,f,g,h} (export lists) - T23664, T24452{a,b,c,d,e,f} (class instances) + +As for reporting an error when renaming fails, we can do better than a simple +"Not in scope" error. For example: + + (1) + + module IncorrectParent (A (b)) where + data A = A { a :: () } + data B = B { b :: () } + + Instead of + Not in scope: ‘b’ + we prefer to emit + The type constructor ‘A’ is not the parent of the record selector ‘b’ + + (2) + + {-# LANGUAGE DuplicateRecordFields #-} + module IncorrectParent (A (other)) where + data A = A { one :: () } + data B = B { other :: () } + data C = C { other :: () } + + Instead of: + Ambiguous occurrence ‘other’. + It could refer to + either the field ‘other’ of record ‘B’ ... + or the field ‘other’ of record ‘C’ ... + we also prefer + The type constructor ‘A’ is not the parent of the record selector ‘other’ (...) + +The work of figuring out which error message to emit is done in +error_no_occurrence_after_disambiguation. -} lookupInstanceDeclarationSubBndr :: DeprecationWarnings -> Name -- ^ Parent View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/655f9b9f89dd51939816eda3bae36c8bf6edb660 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/655f9b9f89dd51939816eda3bae36c8bf6edb660 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 06:58:34 2024 From: gitlab at gitlab.haskell.org (Fabricio Nascimento (@fabu)) Date: Tue, 24 Sep 2024 02:58:34 -0400 Subject: [Git][ghc/ghc][wip/fabu/T25014-mistakenly-accepted-parent] compiler: refactors renamer lookup for sub binder occurences Message-ID: <66f2631a4e64b_2f8adeaced2442413@gitlab.mail> Fabricio Nascimento pushed to branch wip/fabu/T25014-mistakenly-accepted-parent at Glasgow Haskell Compiler / GHC Commits: e6edc967 by Fabricio de Sousa Nascimento at 2024-09-24T15:57:31+09:00 compiler: refactors renamer lookup for sub binder occurences Refactors lookupSubBndrOcc_helper into two functions that separately deal with lookup for type classes and export lists. Removes the Semigroup instance of DisambigInfo in favor of directly filtering the GRE occurences, the refactored logic also fix and issue with a program with multiple references being incorrectly accepted Fix #25014 - - - - - 22 changed files: - compiler/GHC/Rename/Env.hs - compiler/GHC/Tc/Gen/Export.hs - compiler/GHC/Types/Name/Reader.hs - testsuite/tests/rename/T24452/T24452b.hs - testsuite/tests/rename/T24452/T24452b.stderr - testsuite/tests/rename/T24452/all.T - + testsuite/tests/rename/T25014/Ambig1.hs - + testsuite/tests/rename/T25014/Ambig2.hs - + testsuite/tests/rename/T25014/T25014a.hs - + testsuite/tests/rename/T25014/T25014a.stderr - + testsuite/tests/rename/T25014/T25014b.hs - + testsuite/tests/rename/T25014/T25014b.stderr - + testsuite/tests/rename/T25014/T25014c.hs - + testsuite/tests/rename/T25014/T25014c.stderr - + testsuite/tests/rename/T25014/T25014d.hs - + testsuite/tests/rename/T25014/T25014d.stderr - + testsuite/tests/rename/T25014/T25014e.hs - + testsuite/tests/rename/T25014/T25014f.hs - + testsuite/tests/rename/T25014/T25014g.hs - + testsuite/tests/rename/T25014/T25014g.stderr - + testsuite/tests/rename/T25014/T25014h.hs - + testsuite/tests/rename/T25014/all.T Changes: ===================================== compiler/GHC/Rename/Env.hs ===================================== @@ -32,7 +32,8 @@ module GHC.Rename.Env ( getUpdFieldLbls, ChildLookupResult(..), - lookupSubBndrOcc_helper, + lookupChildExportListSubBndr, + lookupInstanceDeclarationSubBndr, HsSigCtxt(..), lookupLocalTcNames, lookupSigOccRn, lookupSigOccRnN, lookupSigCtxtOccRn, @@ -113,7 +114,6 @@ import Data.Either ( partitionEithers ) import Data.Function ( on ) import Data.List ( find, partition, groupBy, sortBy ) import qualified Data.List.NonEmpty as NE -import qualified Data.Semigroup as Semi import System.IO.Unsafe ( unsafePerformIO ) {- @@ -403,7 +403,7 @@ lookupInstDeclBndr cls what rdr -- In an instance decl you aren't allowed -- to use a qualified name for the method -- (Although it'd make perfect sense.) - ; mb_name <- lookupSubBndrOcc + ; mb_name <- lookupInstanceDeclarationSubBndr NoDeprecationWarnings -- we don't give deprecated -- warnings when a deprecated class @@ -679,144 +679,274 @@ disambiguation anyway, because `x` is an original name, and lookupGlobalOccRn will find it. -} --- | Used in export lists to lookup the children. -lookupSubBndrOcc_helper :: Bool -> DeprecationWarnings - -> Name - -> RdrName -- ^ thing we are looking up - -> LookupChild -- ^ how to look it up (e.g. which - -- 'NameSpace's to look in) - -> RnM ChildLookupResult -lookupSubBndrOcc_helper must_have_parent warn_if_deprec parent rdr_name how_lkup - | isUnboundName parent - -- Avoid an error cascade - = return (FoundChild (mkUnboundGRERdr rdr_name)) - - | otherwise = do - gre_env <- getGlobalRdrEnv - let original_gres = lookupGRE gre_env (LookupChildren (rdrNameOcc rdr_name) how_lkup) - picked_gres = pick_gres original_gres - -- The remaining GREs are things that we *could* export here. - -- Note that this includes things which have `NoParent`; - -- those are sorted in `checkPatSynParent`. - traceTc "parent" (ppr parent) - traceTc "lookupExportChild must_have_parent:" (ppr must_have_parent) - traceTc "lookupExportChild original_gres:" (ppr original_gres) - traceTc "lookupExportChild picked_gres:" (ppr picked_gres) +{- +Note [Renaming child GREs] +~~~~~~~~~~~~~~~~~~~~~~~~~~ +When renaming a GRE, we sometimes make use of GRE parent information to +disambiguate or to improve error messages. This happens in two situations: + + - when renaming an export list, e.g. `T`, `fld` in `module M ( A(T, fld) )`, + - when renaming methods of a class instance, e.g. + `instance C a where { type Assoc a = Int; method a = a }` + +In both of these situations, we first look up all matching GREs, but then +further refine by filtering out GREs with incorrect parents. This is done in +pick_matching_gres, using the DisambigInfo datatype. We proceed as follows: + + 1. We first check if there are no matching GRE at all, and return NoOccurence. + 2. Then we check whether there is a single matching GRE with the right parent, + say gre. + If so, return "MatchingParentOccurrence gre" + 2. If there are multiple matching GREs with the right parent, + return those, using AmbiguousOccurrence. + 3. In the absence of GREs with the right parent, we check whether there is + a single matching GRE, say gre. + If so, return "NoParentOccurrence gre". + 5. Finally, there are multiple matching GREs (none with the right parent), + return all matches, using AmbiguousOccurrence. + +We then consume this information slightly differently for the export case and +for the instance method case, because for exports we can accept a GRE which has +no parent (e.g. when bundling a pattern synonym, as per Note [Parents] +in GHC.Types.Name.Reader), whereas for a class instance we definitely need +the class itself to be the parent, as in the example: + + import Control.Applicative ( Alternative ) + import Data.Set ( empty ) + instance Alternative Foo where + empty = ... + +Test cases: + - T11970 (both cases) + - T25014{a,b,c,d,e,f,g,h} (export lists) + - T23664, T24452{a,b,c,d,e,f} (class instances) + +As for reporting an error when renaming fails, we can do better than a simple +"Not in scope" error. For example: + + (1) + + module IncorrectParent (A (b)) where + data A = A { a :: () } + data B = B { b :: () } + + Instead of + Not in scope: ‘b’ + we prefer to emit + The type constructor ‘A’ is not the parent of the record selector ‘b’ + + (2) + + {-# LANGUAGE DuplicateRecordFields #-} + module IncorrectParent (A (other)) where + data A = A { one :: () } + data B = B { other :: () } + data C = C { other :: () } + + Instead of: + Ambiguous occurrence ‘other’. + It could refer to + either the field ‘other’ of record ‘B’ ... + or the field ‘other’ of record ‘C’ ... + we also prefer + The type constructor ‘A’ is not the parent of the record selector ‘other’ (...) + +The work of figuring out which error message to emit is done in +error_no_occurrence_after_disambiguation. +-} +lookupInstanceDeclarationSubBndr :: DeprecationWarnings + -> Name -- ^ Parent + -> SDoc + -> RdrName -- ^ thing we are looking up + -> RnM (Either NotInScopeError Name) +lookupInstanceDeclarationSubBndr warn_if_deprec parent doc rdr_name = + lookupExactOrOrig rdr_name (Right . greName) $ + -- This happens for built-in classes, see mod52 for example + do + let lookup_method = LookupChild { wantedParent = parent + , lookupDataConFirst = False + , prioritiseParent = True -- See T23664. + } + (picked_gres, _) <- pick_matching_gres parent rdr_name lookup_method + traceRn "lookupInstanceDeclarationSubBndr" (ppr picked_gres) + -- See [Mismatched class methods and associated type families] + -- in TcInstDecls. + case picked_gres of + MatchingParentOccurrence g -> do + addUsedGRE warn_if_deprec g + return $ Right (greName g) + NoOccurrence -> + return $ Left (UnknownSubordinate doc) + NoParentOccurrence _ -> + return $ Left (UnknownSubordinate doc) + AmbiguousOccurrence _ -> + return $ Left (UnknownSubordinate doc) + +-- For details, see [Renaming child GREs] +lookupChildExportListSubBndr :: DeprecationWarnings + -> Name -- ^ Parent + -> RdrName -- ^ thing we are looking up + -> LookupChild -- ^ how to look it up (e.g. which + -- 'NameSpace's to look in) + -> RnM ChildLookupResult +lookupChildExportListSubBndr warn_if_deprec parent rdr_name lookup_method = do + (picked_gres, original_gres) <- + pick_matching_gres parent rdr_name lookup_method + traceRn "lookupChildExportListSubBndr" (ppr picked_gres) case picked_gres of + NoParentOccurrence g -> + success_found_child warn_if_deprec g + MatchingParentOccurrence g -> + success_found_child warn_if_deprec g NoOccurrence -> - noMatchingParentErr original_gres - UniqueOccurrence g -> - if must_have_parent - then noMatchingParentErr original_gres - else checkFld g - DisambiguatedOccurrence g -> - checkFld g + error_no_occurrence_after_disambiguation parent rdr_name original_gres AmbiguousOccurrence gres -> - if must_have_parent - -- It is more helpful to tell the user that the ambiguous matches - -- are for a wrong parent, then that there is a name clash, - -- see (#24452). Also since `gres` is NonEmpty and is a sub-list - -- of `original_gres` we are sure the original list is NonEmpty. - then mkIncorrectParentErr (NE.fromList original_gres) - else mkNameClashErr gres - where - checkFld :: GlobalRdrElt -> RnM ChildLookupResult - checkFld g = do - addUsedGRE warn_if_deprec g - return $ FoundChild g - - -- Called when we find no matching GREs after disambiguation but - -- there are three situations where this happens. - -- 1. There were none to begin with. - -- 2. None of the matching ones were the parent but - -- a. They were from an overloaded record field so we can report - -- a better error. - -- b. The original lookup was actually ambiguous. - -- For example, the case where overloading is off and two - -- record fields are in scope from different record - -- constructors, neither of which is the parent. - noMatchingParentErr :: [GlobalRdrElt] -> RnM ChildLookupResult - noMatchingParentErr original_gres = do - traceRn "noMatchingParentErr" (ppr original_gres) - dup_fields_ok <- xoptM LangExt.DuplicateRecordFields - case original_gres of - [] -> return NameNotFound - [g] -> mkIncorrectParentErr (NE.fromList [g]) - gss@(g:gss'@(_:_)) -> - if dup_fields_ok && all isRecFldGRE gss - then mkIncorrectParentErr (NE.fromList gss) - else mkNameClashErr $ g NE.:| gss' - - mkIncorrectParentErr :: NE.NonEmpty GlobalRdrElt -> RnM ChildLookupResult - mkIncorrectParentErr gres = return $ IncorrectParent parent (NE.head gres) - [p | x <- NE.toList gres, ParentIs p <- [greParent x]] - - mkNameClashErr :: NE.NonEmpty GlobalRdrElt -> RnM ChildLookupResult - mkNameClashErr gres = do - addNameClashErrRn rdr_name gres - return (FoundChild (NE.head gres)) - - pick_gres :: [GlobalRdrElt] -> DisambigInfo - -- For Unqual, find GREs that are in scope qualified or unqualified - -- For Qual, find GREs that are in scope with that qualification - pick_gres gres - | isUnqual rdr_name - = mconcat (map right_parent gres) - | otherwise - = mconcat (map right_parent (pickGREs rdr_name gres)) - - right_parent :: GlobalRdrElt -> DisambigInfo - right_parent gre - = case greParent gre of - ParentIs cur_parent - | parent == cur_parent -> DisambiguatedOccurrence gre - | otherwise -> NoOccurrence - NoParent -> UniqueOccurrence gre -{-# INLINEABLE lookupSubBndrOcc_helper #-} + error_name_clash rdr_name gres + where + success_found_child warn_if_deprec g = do + addUsedGRE warn_if_deprec g + return $ FoundChild g + +pick_matching_gres :: Name -- Parent + -> RdrName -- ^ thing we are looking up + -> LookupChild -- ^ how to look it up (e.g. which + -- 'NameSpace's to look in) + -> RnM (DisambigInfo, [GlobalRdrEltX GREInfo]) +pick_matching_gres parent rdr_name lookup_method = do + if isUnboundName parent + -- Avoid an error cascade, see Note [ Unbound vs Ambiguous Names ] + then return (MatchingParentOccurrence (mkUnboundGRERdr rdr_name), []) + else do + gre_env <- getGlobalRdrEnv + let lookup_chidren = LookupChildren (rdrNameOcc rdr_name) lookup_method + original_gres = lookupGRE gre_env lookup_chidren + picked_gres = pick_gres original_gres + -- The remaining GREs are things that we *could* export here. + -- Note that this includes things which have 'NoParent'; + -- those are sorted in 'checkPatSynParent'. + traceTc "pick_matching_gres original_gres:" (ppr original_gres) + return (picked_gres, original_gres) + where + -- See Note [Renaming child GREs] for details of what is happening here. + pick_gres :: [GlobalRdrElt] -> DisambigInfo + pick_gres gres + | null no_parent_gres && null matching_parent_gres = + NoOccurrence + | [gre] <- matching_parent_gres = + MatchingParentOccurrence gre + | [gre] <- no_parent_gres, null matching_parent_gres = + -- Checking `null matching_parent_gres` prevents a program to compile + -- when there is a parent ambiguity. See T24014g + NoParentOccurrence gre + | otherwise = do + let all_gres = matching_parent_gres ++ no_parent_gres + AmbiguousOccurrence (NE.fromList all_gres) + where + resolved_gres = resolve_gres rdr_name gres + (matching_parent_gres, no_parent_gres) = partition_gres resolved_gres + + -- foldr preserves the order of the errors as they appear in the source + partition_gres :: [DisambigInfo] -> ([GlobalRdrElt], [GlobalRdrElt]) + partition_gres = foldr separate_gres_matches ([], []) + where + separate_gres_matches :: DisambigInfo -> ([GlobalRdrElt], [GlobalRdrElt]) -> ([GlobalRdrElt], [GlobalRdrElt]) + separate_gres_matches (MatchingParentOccurrence g) (matching_parent_gres, no_parent_gres) = (g:matching_parent_gres, no_parent_gres) + separate_gres_matches (NoParentOccurrence g) (matching_parent_gres, no_parent_gres) = (matching_parent_gres, g:no_parent_gres) + separate_gres_matches _ acc = acc + + -- For Unqual, find GREs that are in scope qualified or unqualified + -- For Qual, find GREs that are in scope with that qualification + resolve_gres :: RdrName -> [GlobalRdrElt] -> [DisambigInfo] + resolve_gres rdr_name gres + | isUnqual rdr_name = map (match_parent parent) gres + | otherwise = map (match_parent parent) (pickGREs rdr_name gres) + +{- Note [Better errors for no matching GREs] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +When GHC does not find a matching name on GREs after disambiguation +(see [Picking and disambiguating children candidates]) it outputs an error like +`Not in scope: ...` (NoOccurence). In some cases we can offer a better error +by looking at the original GRE matches before disambiguation and attempt to +surface problems that could have caused GHC to not being able to find the +correct identifier. This is what error_no_occurrence_after_disambiguation does. + +1. For example where the name exists for a different parent. + + module IncorrectParent (A (b)) where + data A = A { a :: () } + data B = B { b :: () } + +In this case instead of `Not in scope: ‘b’` we prefer the error. + The type constructor ‘A’ is not the parent of the record selector ‘b’ (...) + +2. Another case is when there is an ambiguity and we have DuplicateRecordFields. + + {-# LANGUAGE DuplicateRecordFields #-} + module IncorrectParent (A (other)) where + data A = A { one :: () } + data B = B { other :: () } + data C = C { other :: () } + +we also prefer + The type constructor ‘A’ is not the parent of the record selector ‘other’ (...) + +instead of: + Ambiguous occurrence ‘other’. + It could refer to + either the field ‘other’ of record ‘B’ ... + or the field ‘other’ of record ‘C’ ... +-} +error_no_occurrence_after_disambiguation :: Name + -> RdrName + -> [GlobalRdrEltX GREInfo] + -> RnM ChildLookupResult +error_no_occurrence_after_disambiguation parent rdr_name original_gres = do + traceRn "error_no_matching_parent" (ppr original_gres) + dup_fields_ok <- xoptM LangExt.DuplicateRecordFields + case original_gres of + [] -> return NameNotFound + [g] -> error_incorrect_parent parent (NE.fromList [g]) + gss@(g:gss'@(_:_)) -> + if dup_fields_ok && all isRecFldGRE gss + then error_incorrect_parent parent (NE.fromList gss) + else error_name_clash rdr_name $ g NE.:| gss' + +error_name_clash :: RdrName -> NE.NonEmpty GlobalRdrElt -> RnM ChildLookupResult +error_name_clash rdr_name gres = do + addNameClashErrRn rdr_name gres + return (FoundChild (NE.head gres)) -- Avoid an error cascade, see Note [ Unbound vs Ambiguous Names ] + +error_incorrect_parent :: Name -> NE.NonEmpty GlobalRdrElt -> RnM ChildLookupResult +error_incorrect_parent parent gres = return $ IncorrectParent parent (NE.head gres) + [p | x <- NE.toList gres, ParentIs p <- [greParent x]] + + +match_parent :: Name -> GlobalRdrElt -> DisambigInfo +match_parent parent gre + = case greParent gre of + ParentIs cur_parent + | parent == cur_parent -> MatchingParentOccurrence gre + | otherwise -> NoOccurrence + NoParent -> NoParentOccurrence gre -- | This domain specific datatype is used to record why we decided it was -- possible that a GRE could be exported with a parent. data DisambigInfo = NoOccurrence -- ^ The GRE could not be found, or it has the wrong parent. - | UniqueOccurrence GlobalRdrElt + | NoParentOccurrence GlobalRdrElt -- ^ The GRE has no parent. It could be a pattern synonym. - | DisambiguatedOccurrence GlobalRdrElt - -- ^ The parent of the GRE is the correct parent. + | MatchingParentOccurrence GlobalRdrElt + -- ^ The parent of the GRE is the correct parent. See match_parent. | AmbiguousOccurrence (NE.NonEmpty GlobalRdrElt) -- ^ The GRE is ambiguous. - -- - -- For example, two normal identifiers with the same name are in - -- scope. They will both be resolved to "UniqueOccurrence" and the - -- monoid will combine them to this failing case. - instance Outputable DisambigInfo where ppr NoOccurrence = text "NoOccurrence" - ppr (UniqueOccurrence gre) = text "UniqueOccurrence:" <+> ppr gre - ppr (DisambiguatedOccurrence gre) = text "DiambiguatedOccurrence:" <+> ppr gre + ppr (NoParentOccurrence gre) = text "UniqueOccurrence:" <+> ppr gre + ppr (MatchingParentOccurrence gre) = text "MatchingParentOccurrence:" + <+> ppr gre ppr (AmbiguousOccurrence gres) = text "Ambiguous:" <+> ppr gres -instance Semi.Semigroup DisambigInfo where - -- These are the key lines: we prefer disambiguated occurrences to other - -- names. - _ <> DisambiguatedOccurrence g' = DisambiguatedOccurrence g' - DisambiguatedOccurrence g' <> _ = DisambiguatedOccurrence g' - - NoOccurrence <> m = m - m <> NoOccurrence = m - UniqueOccurrence g <> UniqueOccurrence g' - = AmbiguousOccurrence $ g NE.:| [g'] - UniqueOccurrence g <> AmbiguousOccurrence gs - = AmbiguousOccurrence (g `NE.cons` gs) - AmbiguousOccurrence gs <> UniqueOccurrence g' - = AmbiguousOccurrence (g' `NE.cons` gs) - AmbiguousOccurrence gs <> AmbiguousOccurrence gs' - = AmbiguousOccurrence (gs Semi.<> gs') - -instance Monoid DisambigInfo where - mempty = NoOccurrence - mappend = (Semi.<>) - -- Lookup SubBndrOcc can never be ambiguous -- -- Records the result of looking up a child. @@ -829,7 +959,6 @@ data ChildLookupResult [Name] -- ^ list of possible parents -- | We resolved to a child | FoundChild GlobalRdrElt - instance Outputable ChildLookupResult where ppr NameNotFound = text "NameNotFound" ppr (FoundChild n) = text "Found:" <+> ppr (greParent n) <+> ppr n @@ -837,28 +966,6 @@ instance Outputable ChildLookupResult where = text "IncorrectParent" <+> hsep [ppr p, ppr $ greName g, ppr ns] -lookupSubBndrOcc :: DeprecationWarnings - -> Name -- Parent - -> SDoc - -> RdrName - -> RnM (Either NotInScopeError Name) --- ^ Find all the things the 'RdrName' maps to, --- and pick the one with the right 'Parent' 'Name'. -lookupSubBndrOcc warn_if_deprec the_parent doc rdr_name = - lookupExactOrOrig rdr_name (Right . greName) $ - -- This happens for built-in classes, see mod052 for example - do { child <- lookupSubBndrOcc_helper True warn_if_deprec the_parent rdr_name what_lkup - ; return $ case child of - FoundChild g -> Right (greName g) - NameNotFound -> Left (UnknownSubordinate doc) - IncorrectParent {} -> Left (UnknownSubordinate doc) } - -- See [Mismatched class methods and associated type families] - -- in TcInstDecls. - where - what_lkup = LookupChild { wantedParent = the_parent - , lookupDataConFirst = False - , prioritiseParent = True -- See T23664. - } {- Note [Family instance binders] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -2169,7 +2276,7 @@ lookupBindGroupOcc ctxt what rdr_name also_try_tycon_ns ns_spec = NE.singleton (Left err) lookup_cls_op cls - = NE.singleton <$> lookupSubBndrOcc AllDeprecationWarnings cls doc rdr_name + = NE.singleton <$> lookupInstanceDeclarationSubBndr AllDeprecationWarnings cls doc rdr_name where doc = text "method of class" <+> quotes (ppr cls) ===================================== compiler/GHC/Tc/Gen/Export.hs ===================================== @@ -654,32 +654,67 @@ If the module has NO main function: The IO action ‘main’ is not defined in module ‘Main’ -} +{- +Note [Renaming children on export lists] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Renaming export lists has many corner cases, and 5 different things can appear +in a children export list under a parent. --- Renaming exports lists is a minefield. Five different things can appear in --- children export lists ( T(A, B, C) ). --- 1. Record selectors --- 2. Type constructors --- 3. Data constructors --- 4. Pattern Synonyms --- 5. Pattern Synonym Selectors --- --- However, things get put into weird name spaces. --- 1. Some type constructors are parsed as variables (-.->) for example. --- 2. All data constructors are parsed as type constructors --- 3. When there is ambiguity, we default type constructors to data --- constructors and require the explicit `type` keyword for type --- constructors. --- --- This function first establishes the possible namespaces that an --- identifier might be in (`choosePossibleNameSpaces`). --- --- Then for each namespace in turn, tries to find the correct identifier --- there returning the first positive result or the first terminating --- error. --- + module M (R (s), D (MkD), Maybe (Empty), Either (Empty), pattern Px) where + + -- 1. Record Selector + data R = R { s :: Int } + + -- 2. Data Constructor + data D a = MkD a + + -- 3. Type Constructor + type S = MkD Int + + -- 4. Pattern Synonyms + class Empty a where + isEmpty :: a -> Bool + instance Empty (Maybe a) where + isEmpty Nothing = True + instance Empty (Either a b) where + isEmpty (Left _) = True + pattern Empty :: Empty a => a + pattern Empty <- (isEmpty -> True) + + -- 5. Record Pattern Synonym selectors + data Point = Point Int Int + + pattern Px :: Int -> Point + pattern Px{x} <- Point x _ + + +To makes matter more complicated: +1. Some type constructors are parsed as variables (-.->) for example. +2. All data constructors are parsed as type constructors +3. When there is ambiguity, we default type constructors to data +constructors and require the explicit `type` keyword for type +constructors. +4. Pattern synonyms are very flexible in which parents they can be exported with +(see [Typing Pattern Synonym Exports]). + +We proceed in two steps: + + 1. We look up GREs, handling the possible NameSpaces to look up in. + See Note [Configurable GRE lookup priority]. + 2. We refine by using the GRE parent information. + See Note [Renaming child GREs]. + +For more details see +[Renaming the LHS on type class Instances], +[Configurable GRE lookup priority] and [Picking and disambiguating children +candidates]. + +Also notice that this logic is similar to +[Renaming the LHS on type class Instances] +-} lookupChildrenExport :: Name -> [LIEWrappedName GhcPs] -> RnM ([(LIEWrappedName GhcRn, GlobalRdrElt)]) lookupChildrenExport spec_parent rdr_items = mapAndReportM doOne rdr_items @@ -698,8 +733,8 @@ lookupChildrenExport spec_parent rdr_items = mapAndReportM doOne rdr_items , prioritiseParent = False -- See T11970. } - -- Do not report export list declaration deprecations - name <- lookupSubBndrOcc_helper False ExportDeprecationWarnings + -- Do not report export list declaration deprecations + name <- lookupChildExportListSubBndr ExportDeprecationWarnings spec_parent bareName what_lkup traceRn "lookupChildrenExport" (ppr name) -- Default to data constructors for slightly better error ===================================== compiler/GHC/Types/Name/Reader.hs ===================================== @@ -1215,7 +1215,7 @@ data LookupChild -- - @True@: prioritise getting the right 'Parent' -- - @False@: prioritise getting the right 'NameSpace' -- - -- See Note [childGREPriority]. + -- See Note [Configurable GRE lookup priority]. } instance Outputable LookupChild where @@ -1262,21 +1262,27 @@ greIsRelevant which_gres ns gre where other_ns = greNameSpace gre -{- Note [childGREPriority] -~~~~~~~~~~~~~~~~~~~~~~~~~~ +{- Note [Configurable GRE lookup priority] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ There are currently two places in the compiler where we look up GlobalRdrElts -which have a given Parent. These are the two calls to lookupSubBndrOcc_helper: +which have a given Parent. - A. Looking up children in an export item, e.g. + A. lookupChildExportListSubBndr looks up children in an export item, e.g. module M ( T(MkT, D) ) where { data T = MkT; data D = D } - B. Looking up binders in a class or instance declaration, e.g. + -- see [Renaming children on export lists] + + B. lookupInstanceDeclarationSubBndr looks up binders in a class or + instance declaration, e.g. + the operator +++ in the fixity declaration: - class C a where { type (+++) :: a -> a ->; infixl 6 +++ } + class C a where { type (+++) :: a -> a -> a; infixl 6 +++ } (+++) :: Int -> Int -> Int; (+++) = (+) + -- see [Renaming the LHS on type class Instances] + In these two situations, there are two competing metrics for finding the "best" 'GlobalRdrElt' that a particular 'OccName' resolves to: @@ -1307,7 +1313,7 @@ Not doing so led to #23664. -- -- We score by 'Parent' and 'NameSpace', with higher priorities having lower -- numbers. Which lexicographic order we use ('Parent' or 'NameSpace' first) --- is determined by the first argument; see Note [childGREPriority]. +-- is determined by the first argument; see Note [Configurable GRE lookup priority]. childGREPriority :: LookupChild -- ^ what kind of child do we want, -- e.g. what should its parent be? -> NameSpace -- ^ what 'NameSpace' are we originally looking in? @@ -1327,7 +1333,7 @@ childGREPriority (LookupChild { wantedParent = wanted_parent in Just $ if par_first then (par_prio, ns_prio) else (ns_prio, par_prio) - -- See Note [childGREPriority]. + -- See Note [Configurable GRE lookup priority]. where -- Pick out the possible 'NameSpace's in order of priority. ===================================== testsuite/tests/rename/T24452/T24452b.hs ===================================== @@ -1,5 +1,4 @@ -- Alternative.empty is not visible -{-# LANGUAGE GeneralizedNewtypeDeriving #-} module BugReproduce where import Control.Applicative (Alternative) ===================================== testsuite/tests/rename/T24452/T24452b.stderr ===================================== @@ -1,2 +1,2 @@ -T24452b.hs:10:3: error: [GHC-54721] +T24452b.hs:9:3: error: [GHC-54721] ‘empty’ is not a (visible) method of class ‘Alternative’ \ No newline at end of file ===================================== testsuite/tests/rename/T24452/all.T ===================================== @@ -3,4 +3,4 @@ test('T24452b', normal, compile_fail, ['']) test('T24452c', normal, compile_fail, ['']) test('T24452d', normal, compile_fail, ['']) test('T24452e', normal, compile_fail, ['']) -test('T24452f', [extra_files(['AmbigPatSynA.hs', 'AmbigPatSynB.hs'])], multimod_compile_fail, ['T24452f','-v0']) \ No newline at end of file +test('T24452f', [extra_files(['AmbigPatSynA.hs', 'AmbigPatSynB.hs'])], multimod_compile_fail, ['T24452f','-v0']) ===================================== testsuite/tests/rename/T25014/Ambig1.hs ===================================== @@ -0,0 +1,5 @@ +-- A module that is ambiguous with Ambig2 +{-# LANGUAGE TypeFamilies #-} +module Ambig1 where + data family T a + data instance T Bool = MkT ===================================== testsuite/tests/rename/T25014/Ambig2.hs ===================================== @@ -0,0 +1,5 @@ +-- A module that is ambiguous with Ambig1 +{-# LANGUAGE TypeFamilies #-} +module Ambig2 where + import Ambig1 (T) + data instance T Int = MkT ===================================== testsuite/tests/rename/T25014/T25014a.hs ===================================== @@ -0,0 +1,5 @@ +-- Should not compile as it is unclear what gets exported +module T25014a (T(MkT)) where + import Ambig1 (T(MkT)) + import Ambig2 (T(MkT)) + data S \ No newline at end of file ===================================== testsuite/tests/rename/T25014/T25014a.stderr ===================================== @@ -0,0 +1,10 @@ +T25014a.hs:2:17: [GHC-87543] + Ambiguous occurrence ‘MkT’. + It could refer to + either ‘Ambig1.MkT’, + imported from ‘Ambig1’ at T25014a.hs:3:18-23 + (and originally defined at Ambig1.hs:5:26-28), + or ‘Ambig2.MkT’, + imported from ‘Ambig2’ at T25014a.hs:4:18-23 + (and originally defined at Ambig2.hs:5:25-27). + In the export: T(MkT) \ No newline at end of file ===================================== testsuite/tests/rename/T25014/T25014b.hs ===================================== @@ -0,0 +1,4 @@ +-- Should not compile trying to export a missing name +module T25014b (A (foo)) where + +data A ===================================== testsuite/tests/rename/T25014/T25014b.stderr ===================================== @@ -0,0 +1,3 @@ +T25014b.hs:2:17: [GHC-76037] + Not in scope: ‘foo’ + In the export: A(foo) \ No newline at end of file ===================================== testsuite/tests/rename/T25014/T25014c.hs ===================================== @@ -0,0 +1,8 @@ +-- Should not compile trying to export a missing name +module T25014c (A (foo)) where + +data A + +data B = B { + foo :: Int +} ===================================== testsuite/tests/rename/T25014/T25014c.stderr ===================================== @@ -0,0 +1,5 @@ +T25014c.hs:2:17: [GHC-88993] + The type constructor ‘A’ is not the parent of the record selector ‘foo’. + Record selectors can only be exported with their parent type constructor. + Parent: B + In the export: A(foo) \ No newline at end of file ===================================== testsuite/tests/rename/T25014/T25014d.hs ===================================== @@ -0,0 +1,6 @@ +-- Should not compile trying to export a name with the wrong parent +module T25014b (A (foo)) where + +data A + +foo = 1 ===================================== testsuite/tests/rename/T25014/T25014d.stderr ===================================== @@ -0,0 +1,4 @@ +T25014d.hs:2:17: [GHC-88993] + The type constructor ‘A’ is not the parent of the identifier ‘foo’. + Identifiers can only be exported with their parent type constructor. + In the export: A(foo) \ No newline at end of file ===================================== testsuite/tests/rename/T25014/T25014e.hs ===================================== @@ -0,0 +1,6 @@ +-- Should compile as A.foo matches parent +module T25014b (A (foo)) where + +data A = A { + foo :: Int +} ===================================== testsuite/tests/rename/T25014/T25014f.hs ===================================== @@ -0,0 +1,4 @@ +-- Should not compile as it is unclear what gets exported +module T25014a (T(MkT)) where + import Ambig1 (T(MkT)) + data S \ No newline at end of file ===================================== testsuite/tests/rename/T25014/T25014g.hs ===================================== @@ -0,0 +1,9 @@ +-- Even though there is a single MkT without a parent, there is an ambiguity +-- between Ambig1.T.MkT and Ambig2.T.MkT so we can't compile +{-# LANGUAGE PatternSynonyms #-} + +module T25014g (T(MkT)) where + import Ambig1 (T(MkT)) + import Ambig2 (T(MkT)) + + pattern MkT = Ambig1.MkT ===================================== testsuite/tests/rename/T25014/T25014g.stderr ===================================== @@ -0,0 +1,11 @@ +T25014g.hs:5:17: [GHC-87543] + Ambiguous occurrence ‘MkT’. + It could refer to + either ‘Ambig1.MkT’, + imported from ‘Ambig1’ at T25014g.hs:6:18-23 + (and originally defined at Ambig1.hs:5:26-28), + or ‘Ambig2.MkT’, + imported from ‘Ambig2’ at T25014g.hs:7:18-23 + (and originally defined at Ambig2.hs:5:25-27), + or ‘T25014g.MkT’, defined at T25014g.hs:9:3. + In the export: T(MkT) ===================================== testsuite/tests/rename/T25014/T25014h.hs ===================================== @@ -0,0 +1,8 @@ +-- A comment +{-# LANGUAGE PatternSynonyms #-} + +module T25014h (T(MkT)) where + import Ambig1 (T(MkT)) + import qualified Ambig1 as Am + + pattern MkT = Am.MkT ===================================== testsuite/tests/rename/T25014/all.T ===================================== @@ -0,0 +1,8 @@ +test('T25014a', [extra_files(['Ambig1.hs', 'Ambig2.hs'])], multimod_compile_fail, ['T25014a','-v0']) +test('T25014b', [], compile_fail, ['']) +test('T25014c', [], compile_fail, ['']) +test('T25014d', [], compile_fail, ['']) +test('T25014e', [], compile, ['']) +test('T25014f', [extra_files(['Ambig1.hs', 'Ambig2.hs'])], multimod_compile, ['T25014f','-v0']) +test('T25014g', [extra_files(['Ambig1.hs', 'Ambig2.hs'])], multimod_compile_fail, ['T25014g','-v0']) +test('T25014h', [extra_files(['Ambig1.hs'])], multimod_compile, ['T25014h','-v0']) \ No newline at end of file View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e6edc9673f534ff498a14cf32166761daf7b6d74 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e6edc9673f534ff498a14cf32166761daf7b6d74 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 08:44:55 2024 From: gitlab at gitlab.haskell.org (Vladislav Zavialov (@int-index)) Date: Tue, 24 Sep 2024 04:44:55 -0400 Subject: [Git][ghc/ghc][wip/int-index/wildcard-binders] 74 commits: Haddock: Add no-compilation flag Message-ID: <66f27c07eab54_35f0ae2641ac629f4@gitlab.mail> Vladislav Zavialov pushed to branch wip/int-index/wildcard-binders at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 8011bc88 by Vladislav Zavialov at 2024-09-24T04:22:34+03:00 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - 22 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fa21f7b5f43d4a51dd122bad02a2f0820284e94b...8011bc880b194219b6203b9f6939b17798678ad3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fa21f7b5f43d4a51dd122bad02a2f0820284e94b...8011bc880b194219b6203b9f6939b17798678ad3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 09:03:51 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Tue, 24 Sep 2024 05:03:51 -0400 Subject: [Git][ghc/ghc][wip/unwire-base] 6 commits: Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import Message-ID: <66f2807785af9_35f0ae540f9c79623@gitlab.mail> Zubin pushed to branch wip/unwire-base at Glasgow Haskell Compiler / GHC Commits: 4aa58551 by Matthew Pickering at 2024-09-24T14:33:41+05:30 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 6f90b915 by Matthew Pickering at 2024-09-24T14:33:41+05:30 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - d7abb6f3 by Matthew Pickering at 2024-09-24T14:33:41+05:30 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - ec4dd266 by Matthew Pickering at 2024-09-24T14:33:41+05:30 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - b9992d1d by Matthew Pickering at 2024-09-24T14:33:41+05:30 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - a7619873 by Matthew Pickering at 2024-09-24T14:33:41+05:30 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 25 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Parser/Header.hs - compiler/GHC/Rename/Names.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Types/Error/Codes.hs - compiler/GHC/Unit/State.hs - compiler/GHC/Unit/Types.hs - docs/users_guide/exts/rebindable_syntax.rst - libraries/base/base.cabal.in - libraries/base/src/Control/Monad/Zip.hs - libraries/base/src/Data/List/NonEmpty.hs - libraries/ghc-internal/ghc-internal.cabal.in - libraries/ghc-internal/src/GHC/Internal/Base.hs - + libraries/ghc-internal/src/GHC/Internal/Control/Monad/Zip.hs - libraries/ghc-internal/src/GHC/Internal/Data/List/NonEmpty.hs - rts/include/RtsAPI.h - testsuite/tests/backpack/should_compile/bkp16.stderr - testsuite/tests/backpack/should_fail/bkpfail17.stderr - testsuite/tests/backpack/should_fail/bkpfail19.stderr - testsuite/tests/iface/IfaceSharingIfaceType.hs - testsuite/tests/iface/IfaceSharingName.hs - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8b91e994f202e4a6dfad139be6d7f841879e4a46...a7619873eee80609e4d1049ecb04eedd0e927c6f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8b91e994f202e4a6dfad139be6d7f841879e4a46...a7619873eee80609e4d1049ecb04eedd0e927c6f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 09:20:25 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Tue, 24 Sep 2024 05:20:25 -0400 Subject: [Git][ghc/ghc][wip/t25150] driver: Fix -working-dir for foreign files Message-ID: <66f28459ea9b6_35f0ae6adcf4876c5@gitlab.mail> Zubin pushed to branch wip/t25150 at Glasgow Haskell Compiler / GHC Commits: 37e03564 by Matthew Pickering at 2024-09-24T14:50:16+05:30 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 5 changed files: - compiler/GHC/Driver/Pipeline.hs - testsuite/tests/driver/all.T - testsuite/tests/driver/multipleHomeUnits/unitCFile - + testsuite/tests/driver/t25150/dir/Main.hs - + testsuite/tests/driver/t25150/dir/a.c Changes: ===================================== compiler/GHC/Driver/Pipeline.hs ===================================== @@ -544,28 +544,28 @@ oneShot orig_hsc_env stop_phase srcs = do compileFile :: HscEnv -> StopPhase -> (FilePath, Maybe Phase) -> IO (Maybe FilePath) compileFile hsc_env stop_phase (src, mb_phase) = do - exists <- doesFileExist src + let offset_file = augmentByWorkingDirectory dflags src + dflags = hsc_dflags hsc_env + mb_o_file = outputFile dflags + ghc_link = ghcLink dflags -- Set by -c or -no-link + notStopPreprocess | StopPreprocess <- stop_phase = False + | _ <- stop_phase = True + -- When linking, the -o argument refers to the linker's output. + -- otherwise, we use it as the name for the pipeline's output. + output + | not (backendGeneratesCode (backend dflags)), notStopPreprocess = NoOutputFile + -- avoid -E -fno-code undesirable interactions. see #20439 + | NoStop <- stop_phase, not (isNoLink ghc_link) = Persistent + -- -o foo applies to linker + | isJust mb_o_file = SpecificFile + -- -o foo applies to the file we are compiling now + | otherwise = Persistent + pipe_env = mkPipeEnv stop_phase offset_file mb_phase output + pipeline = pipelineStart pipe_env (setDumpPrefix pipe_env hsc_env) offset_file mb_phase + + exists <- doesFileExist offset_file when (not exists) $ - throwGhcExceptionIO (CmdLineError ("does not exist: " ++ src)) - - let - dflags = hsc_dflags hsc_env - mb_o_file = outputFile dflags - ghc_link = ghcLink dflags -- Set by -c or -no-link - notStopPreprocess | StopPreprocess <- stop_phase = False - | _ <- stop_phase = True - -- When linking, the -o argument refers to the linker's output. - -- otherwise, we use it as the name for the pipeline's output. - output - | not (backendGeneratesCode (backend dflags)), notStopPreprocess = NoOutputFile - -- avoid -E -fno-code undesirable interactions. see #20439 - | NoStop <- stop_phase, not (isNoLink ghc_link) = Persistent - -- -o foo applies to linker - | isJust mb_o_file = SpecificFile - -- -o foo applies to the file we are compiling now - | otherwise = Persistent - pipe_env = mkPipeEnv stop_phase src mb_phase output - pipeline = pipelineStart pipe_env (setDumpPrefix pipe_env hsc_env) src mb_phase + throwGhcExceptionIO (CmdLineError ("does not exist: " ++ offset_file)) runPipeline (hsc_hooks hsc_env) pipeline ===================================== testsuite/tests/driver/all.T ===================================== @@ -328,3 +328,4 @@ test('T23613', normal, compile_and_run, ['-this-unit-id=foo']) test('T23944', [unless(have_dynamic(), skip), extra_files(['T23944A.hs'])], multimod_compile, ['T23944 T23944A', '-fprefer-byte-code -fbyte-code -fno-code -dynamic-too -fwrite-interface']) test('T24286', [cxx_src, unless(have_profiling(), skip), extra_files(['T24286.cpp'])], compile, ['-prof -no-hs-main']) test('T24839', [unless(arch('x86_64') or arch('aarch64'), skip), extra_files(["t24839_sub.S"])], compile_and_run, ['t24839_sub.S']) +test('t25150', [extra_files(["t25150"])], multimod_compile, ['Main.hs', '-v0 -working-dir t25150/dir a.c']) ===================================== testsuite/tests/driver/multipleHomeUnits/unitCFile ===================================== @@ -1 +1 @@ --working-dir c-file C c-file/c.c -Iinclude +-working-dir c-file C c.c -Iinclude ===================================== testsuite/tests/driver/t25150/dir/Main.hs ===================================== @@ -0,0 +1,6 @@ +{-# LANGUAGE ForeignFunctionInterface #-} +module Main where +foreign import ccall "hello" + hello :: IO () +main :: IO () +main = hello ===================================== testsuite/tests/driver/t25150/dir/a.c ===================================== @@ -0,0 +1,5 @@ +// dir/Aux.c +#include +void hello() { + printf("hi"); +} View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/37e035642a934407f4bafbc9efc722e45061a69f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/37e035642a934407f4bafbc9efc722e45061a69f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 09:50:17 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 05:50:17 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/romes/25283 Message-ID: <66f28b59469f4_28038249fa0814d3@gitlab.mail> Rodrigo Mesquita pushed new branch wip/romes/25283 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/romes/25283 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 09:50:41 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 05:50:41 -0400 Subject: [Git][ghc/ghc][wip/T24807] base: Propagate `error` CallStack to thrown exception Message-ID: <66f28b714e090_280381444988329b@gitlab.mail> Rodrigo Mesquita pushed to branch wip/T24807 at Glasgow Haskell Compiler / GHC Commits: 5e81a005 by Ben Gamari at 2024-09-24T10:47:06+01:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - 30 changed files: - docs/users_guide/9.12.1-notes.rst - libraries/base/changelog.md - libraries/base/tests/T19288.stderr - + libraries/base/tests/T24807.hs - + libraries/base/tests/T24807.stderr - libraries/base/tests/all.T - libraries/ghc-internal/src/GHC/Internal/Exception.hs - testsuite/tests/arityanal/should_run/T21694a.stderr - testsuite/tests/arityanal/should_run/T24296.stderr - testsuite/tests/codeGen/should_run/T16846.stderr - testsuite/tests/codeGen/should_run/T5626.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/codeGen/should_run/cgrun045.stderr - testsuite/tests/codeGen/should_run/cgrun051.stderr - testsuite/tests/codeGen/should_run/cgrun059.stderr - testsuite/tests/concurrent/should_run/T13330.stderr - testsuite/tests/concurrent/should_run/conc021.stderr - testsuite/tests/deSugar/should_run/Or5.stderr - testsuite/tests/deSugar/should_run/T11193.stderr - testsuite/tests/deSugar/should_run/T11572.stderr - testsuite/tests/deSugar/should_run/T11601.stderr - testsuite/tests/dmdanal/should_run/T12368.stderr - testsuite/tests/dmdanal/should_run/T12368a.stderr - testsuite/tests/dmdanal/should_run/T23208.stderr - testsuite/tests/dmdanal/should_run/strun002.stderr - testsuite/tests/driver/multipleHomeUnits/multipleHomeUnits_callstack.stderr - testsuite/tests/ffi/should_run/ffi008.stderr - testsuite/tests/hpc/simple/tixs/T10529a.stderr - testsuite/tests/hpc/simple/tixs/T10529b.stderr - testsuite/tests/hpc/simple/tixs/T10529c.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5e81a005b836fb256414ddea64c439c251f79206 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5e81a005b836fb256414ddea64c439c251f79206 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 10:29:41 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 06:29:41 -0400 Subject: [Git][ghc/ghc][wip/romes/exceptions-layout] Display type and callstack of exception on handler Message-ID: <66f29495afacc_d7d5fe37101548@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exceptions-layout at Glasgow Haskell Compiler / GHC Commits: a971620c by Rodrigo Mesquita at 2024-09-24T11:28:50+01:00 Display type and callstack of exception on handler This commit changes the Exception instance of SomeException to *simply* display the underlying exception in `displayException`. The augmented exception message that included the type and backtrace of the exception are now only printed on a call to `displayExceptionWithInfo`. At a surface level, existing programs should behave the same since the `uncaughtExceptionHandler`, which is responsible for printing out uncaught exceptions to the user, will use `displayExceptionWithInfo` by default. However, unlike the instance's `displayException` method, the `uncaughtExceptionHandler` can be overriden with `setUncaughtExceptionHandler`. This makes the extra information opt-in without fixing it the instance, which can be valuable if your program wants to display uncaught exceptions to users in a user-facing way (ie without backtraces). This is what was originally agreed for CLC#231 or CLC#261 with regard to the type of the exception information. The call stack also becoming part of the default handler rather than the Exception instance is an ammendment to CLC#164. Discussion of the ammendment is part of CLC#285. - - - - - 4 changed files: - libraries/base/tests/T24807.stderr - libraries/ghc-internal/src/GHC/Internal/Conc/Sync.hs - libraries/ghc-internal/src/GHC/Internal/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs Changes: ===================================== libraries/base/tests/T24807.stderr ===================================== @@ -1,11 +1,11 @@ -T24807: Exception: +T24807: Uncaught exception ghc-internal:GHC.Internal.Exception.ErrorCall: + hi CallStack (from HasCallStack): error, called at T24807.hs:1:8 in main:Main -Package: ghc-internal -Module: GHC.Internal.Exception -Type: ErrorCall + HasCallStack backtrace: - collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:92:13 in ghc-internal:GHC.Internal.Exception - toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:127:5 in ghc-internal:GHC.Internal.Exception + collectBacktraces, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:93:13 in ghc-internal:GHC.Internal.Exception + toExceptionWithBacktrace, called at libraries/ghc-internal/src/GHC/Internal/Exception.hs:128:5 in ghc-internal:GHC.Internal.Exception error, called at T24807.hs:1:8 in main:Main + ===================================== libraries/ghc-internal/src/GHC/Internal/Conc/Sync.hs ===================================== @@ -952,7 +952,7 @@ uncaughtExceptionHandler = unsafePerformIO (newIORef defaultHandler) defaultHandler se = do (hFlush stdout) `catchAny` (\ _ -> return ()) - let exMsg = displayException se + let exMsg = displayExceptionWithInfo se msg = "Uncaught exception " ++ exMsg withCString "%s" $ \cfmt -> withCString msg $ \cmsg -> ===================================== libraries/ghc-internal/src/GHC/Internal/Exception.hs ===================================== @@ -35,6 +35,7 @@ module GHC.Internal.Exception -- * 'SomeException' , SomeException(..) + , displayExceptionWithInfo -- * Exception context , someExceptionContext ===================================== libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs ===================================== @@ -29,6 +29,7 @@ module GHC.Internal.Exception.Type ( Exception(..) -- Class , SomeException(..) + , displayExceptionWithInfo , someExceptionContext , addExceptionContext , mapExceptionContext @@ -213,22 +214,35 @@ instance Exception SomeException where in SomeException e fromException = Just backtraceDesired (SomeException e) = backtraceDesired e - displayException (SomeException e) = - case displayContext ?exceptionContext of - "" -> msg - dc -> msg ++ "\n\n" ++ dc - where - msg = - displayExceptionInfo (Typeable.typeOf e) - ++ "\n\n" - ++ displayException e - - displayExceptionInfo :: TypeRep -> String - displayExceptionInfo rep = - tyMsg ++ ":" - where - tyMsg = Typeable.tyConPackage tyCon ++ ":" ++ Typeable.tyConModule tyCon ++ "." ++ Typeable.tyConName tyCon - tyCon = Typeable.typeRepTyCon rep + displayException (SomeException e) = displayException e + +-- | Displays a 'SomeException' with additional information: +-- +-- * The type of the underlying exception +-- * The exception context +-- +-- By default, 'uncaughtExceptionHandler' uses 'displayExceptionWithInfo' to print uncaught exceptions. +-- This default can be overriden with 'setUncaughtExceptionHandler', for +-- instance, to present custom error messages on exceptions to the user. +-- +-- @since base-4.21 +displayExceptionWithInfo :: SomeException -> String +displayExceptionWithInfo (SomeException e) = + case displayContext ?exceptionContext of + "" -> msg + dc -> msg ++ "\n\n" ++ dc + where + msg = + displayExceptionType (Typeable.typeOf e) + ++ "\n\n" + ++ displayException e + + displayExceptionType :: TypeRep -> String + displayExceptionType rep = + tyMsg ++ ":" + where + tyMsg = Typeable.tyConPackage tyCon ++ ":" ++ Typeable.tyConModule tyCon ++ "." ++ Typeable.tyConName tyCon + tyCon = Typeable.typeRepTyCon rep displayContext :: ExceptionContext -> String displayContext (ExceptionContext anns0) = mconcat $ intersperse "\n" $ map go anns0 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a971620c7f026a0f97d883eafe9feae7be7246d6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a971620c7f026a0f97d883eafe9feae7be7246d6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 10:30:21 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 06:30:21 -0400 Subject: [Git][ghc/ghc][wip/romes/exceptions-layout] 3 commits: base: Propagate `error` CallStack to thrown exception Message-ID: <66f294bd47819_d7d5fce2c0198e@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exceptions-layout at Glasgow Haskell Compiler / GHC Commits: 5e81a005 by Ben Gamari at 2024-09-24T10:47:06+01:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - e61bb185 by Rodrigo Mesquita at 2024-09-24T11:29:54+01:00 exceptions: Improve the message layout as per #285 This commit fixes the layout of the additional information included when displaying an exception, namely the type of the exception. It also fixes the default handler's heading message to work well together with the improved display message of SomeException. - - - - - 233d9931 by Rodrigo Mesquita at 2024-09-24T11:29:54+01:00 Display type and callstack of exception on handler This commit changes the Exception instance of SomeException to *simply* display the underlying exception in `displayException`. The augmented exception message that included the type and backtrace of the exception are now only printed on a call to `displayExceptionWithInfo`. At a surface level, existing programs should behave the same since the `uncaughtExceptionHandler`, which is responsible for printing out uncaught exceptions to the user, will use `displayExceptionWithInfo` by default. However, unlike the instance's `displayException` method, the `uncaughtExceptionHandler` can be overriden with `setUncaughtExceptionHandler`. This makes the extra information opt-in without fixing it the instance, which can be valuable if your program wants to display uncaught exceptions to users in a user-facing way (ie without backtraces). This is what was originally agreed for CLC#231 or CLC#261 with regard to the type of the exception information. The call stack also becoming part of the default handler rather than the Exception instance is an ammendment to CLC#164. Discussion of the ammendment is part of CLC#285. - - - - - 30 changed files: - docs/users_guide/9.12.1-notes.rst - libraries/base/changelog.md - libraries/base/tests/IO/T4808.stderr - libraries/base/tests/IO/mkdirExists.stderr - libraries/base/tests/IO/openFile002.stderr - libraries/base/tests/IO/withBinaryFile001.stderr - libraries/base/tests/IO/withBinaryFile002.stderr - libraries/base/tests/IO/withFile001.stderr - libraries/base/tests/IO/withFile002.stderr - libraries/base/tests/IO/withFileBlocking001.stderr - libraries/base/tests/IO/withFileBlocking002.stderr - libraries/base/tests/T15349.stderr - libraries/base/tests/T16111.stderr - libraries/base/tests/T19288.stderr - + libraries/base/tests/T24807.hs - + libraries/base/tests/T24807.stderr - libraries/base/tests/all.T - libraries/base/tests/assert.stderr - libraries/base/tests/readFloat.stderr - libraries/base/tests/topHandler04.stderr - libraries/ghc-compact/tests/compact_function.stderr - libraries/ghc-compact/tests/compact_mutable.stderr - libraries/ghc-compact/tests/compact_pinned.stderr - libraries/ghc-internal/src/GHC/Internal/Conc/Sync.hs - libraries/ghc-internal/src/GHC/Internal/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs - testsuite/tests/arityanal/should_run/T21694a.stderr - testsuite/tests/arityanal/should_run/T24296.stderr - testsuite/tests/array/should_run/arr003.stderr - testsuite/tests/array/should_run/arr004.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a971620c7f026a0f97d883eafe9feae7be7246d6...233d99319958dee0da9c924f69d996b83182eb6b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a971620c7f026a0f97d883eafe9feae7be7246d6...233d99319958dee0da9c924f69d996b83182eb6b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 11:04:09 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Tue, 24 Sep 2024 07:04:09 -0400 Subject: [Git][ghc/ghc][wip/romes/12935] 9 commits: ci: Run abi-test on test-abi label Message-ID: <66f29ca93c4c4_12f733120b2415275@gitlab.mail> Zubin pushed to branch wip/romes/12935 at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/RV64/CodeGen.hs - compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f1782d19565fd4016a3c157e626321c611bfd3c5...7b37afc9f3e79559055488998ee73187886a0e00 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f1782d19565fd4016a3c157e626321c611bfd3c5...7b37afc9f3e79559055488998ee73187886a0e00 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 11:44:03 2024 From: gitlab at gitlab.haskell.org (Sjoerd Visscher (@trac-sjoerd_visscher)) Date: Tue, 24 Sep 2024 07:44:03 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/parallel-depanal-downsweep] Parallelize getRootSummary computations in dep analysis downsweep Message-ID: <66f2a6031fd7_1aaaedf9498351ab@gitlab.mail> Sjoerd Visscher pushed to branch wip/torsten.schmits/parallel-depanal-downsweep at Glasgow Haskell Compiler / GHC Commits: 8237a881 by Torsten Schmits at 2024-09-24T13:43:19+02:00 Parallelize getRootSummary computations in dep analysis downsweep This reuses the upsweep step's infrastructure to process batches of modules in parallel. I benchmarked this by running `ghc -M` on two sets of 10,000 modules; one with a linear dependency chain and the other with a binary tree. Comparing different values for the number of modules per thread suggested an optimum at `length targets `div` (n_cap * 2)`, with results similar to this one (6 cores, 12 threads): ``` Benchmark 1: linear 1 jobs Time (mean ± σ): 1.775 s ± 0.026 s [User: 1.377 s, System: 0.399 s] Range (min … max): 1.757 s … 1.793 s 2 runs Benchmark 2: linear 6 jobs Time (mean ± σ): 876.2 ms ± 20.9 ms [User: 1833.2 ms, System: 518.6 ms] Range (min … max): 856.2 ms … 898.0 ms 3 runs Benchmark 3: linear 12 jobs Time (mean ± σ): 793.5 ms ± 23.2 ms [User: 2318.9 ms, System: 718.6 ms] Range (min … max): 771.9 ms … 818.0 ms 3 runs ``` Results don't differ much when the batch size is reduced to a quarter of that, but there's significant thread scheduling overhead for a size of 1: ``` Benchmark 1: linear 1 jobs Time (mean ± σ): 2.611 s ± 0.029 s [User: 2.851 s, System: 0.783 s] Range (min … max): 2.591 s … 2.632 s 2 runs Benchmark 2: linear 6 jobs Time (mean ± σ): 1.189 s ± 0.007 s [User: 2.707 s, System: 1.103 s] Range (min … max): 1.184 s … 1.194 s 2 runs Benchmark 3: linear 12 jobs Time (mean ± σ): 1.097 s ± 0.006 s [User: 2.938 s, System: 1.300 s] Range (min … max): 1.093 s … 1.101 s 2 runs ``` Larger batches also slightly worsen performance. - - - - - 5 changed files: - compiler/GHC/Driver/Make.hs - compiler/GHC/Unit/Finder.hs - testsuite/tests/ghc-api/downsweep/OldModLocation.hs - testsuite/tests/ghc-api/downsweep/PartialDownsweep.hs - utils/haddock/haddock-api/src/Haddock/Interface.hs Changes: ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -5,6 +5,8 @@ {-# LANGUAGE ApplicativeDo #-} {-# LANGUAGE MultiWayIf #-} {-# LANGUAGE RecordWildCards #-} +{-# LANGUAGE LambdaCase #-} +{-# LANGUAGE BlockArguments #-} -- ----------------------------------------------------------------------------- -- @@ -122,7 +124,7 @@ import qualified Control.Monad.Catch as MC import Data.IORef import Data.Maybe import Data.Time -import Data.List (sortOn) +import Data.List (sortOn, unfoldr) import Data.Bifunctor (first) import System.Directory import System.FilePath @@ -169,7 +171,7 @@ depanal :: GhcMonad m => -> Bool -- ^ allow duplicate roots -> m ModuleGraph depanal excluded_mods allow_dup_roots = do - (errs, mod_graph) <- depanalE excluded_mods allow_dup_roots + (errs, mod_graph) <- depanalE mkUnknownDiagnostic Nothing excluded_mods allow_dup_roots if isEmptyMessages errs then pure mod_graph else throwErrors (fmap GhcDriverMessage errs) @@ -177,12 +179,14 @@ depanal excluded_mods allow_dup_roots = do -- | Perform dependency analysis like in 'depanal'. -- In case of errors, the errors and an empty module graph are returned. depanalE :: GhcMonad m => -- New for #17459 - [ModuleName] -- ^ excluded modules + (GhcMessage -> AnyGhcDiagnostic) + -> Maybe Messager + -> [ModuleName] -- ^ excluded modules -> Bool -- ^ allow duplicate roots -> m (DriverMessages, ModuleGraph) -depanalE excluded_mods allow_dup_roots = do +depanalE diag_wrapper msg excluded_mods allow_dup_roots = do hsc_env <- getSession - (errs, mod_graph) <- depanalPartial excluded_mods allow_dup_roots + (errs, mod_graph) <- depanalPartial diag_wrapper msg excluded_mods allow_dup_roots if isEmptyMessages errs then do hsc_env <- getSession @@ -220,11 +224,13 @@ depanalE excluded_mods allow_dup_roots = do -- new module graph. depanalPartial :: GhcMonad m - => [ModuleName] -- ^ excluded modules + => (GhcMessage -> AnyGhcDiagnostic) + -> Maybe Messager + -> [ModuleName] -- ^ excluded modules -> Bool -- ^ allow duplicate roots -> m (DriverMessages, ModuleGraph) -- ^ possibly empty 'Bag' of errors and a module graph. -depanalPartial excluded_mods allow_dup_roots = do +depanalPartial diag_wrapper msg excluded_mods allow_dup_roots = do hsc_env <- getSession let targets = hsc_targets hsc_env @@ -243,7 +249,7 @@ depanalPartial excluded_mods allow_dup_roots = do liftIO $ flushFinderCaches (hsc_FC hsc_env) (hsc_unit_env hsc_env) (errs, graph_nodes) <- liftIO $ downsweep - hsc_env (mgModSummaries old_graph) + hsc_env diag_wrapper msg (mgModSummaries old_graph) excluded_mods allow_dup_roots let mod_graph = mkModuleGraph graph_nodes @@ -497,8 +503,8 @@ loadWithCache :: GhcMonad m => Maybe ModIfaceCache -- ^ Instructions about how t -> LoadHowMuch -- ^ How much `loadWithCache` should load -> m SuccessFlag loadWithCache cache diag_wrapper how_much = do - (errs, mod_graph) <- depanalE [] False -- #17459 msg <- mkBatchMsg <$> getSession + (errs, mod_graph) <- depanalE diag_wrapper (Just msg) [] False -- #17459 success <- load' cache how_much diag_wrapper (Just msg) mod_graph if isEmptyMessages errs then pure success @@ -506,7 +512,7 @@ loadWithCache cache diag_wrapper how_much = do -- Note [Unused packages] -- ~~~~~~~~~~~~~~~~~~~~~~ --- Cabal passes `--package-id` flag for each direct dependency. But GHC +-- Cabal passes `-package-id` flag for each direct dependency. But GHC -- loads them lazily, so when compilation is done, we have a list of all -- actually loaded packages. All the packages, specified on command line, -- but never loaded, are probably unused dependencies. @@ -1551,6 +1557,8 @@ type DownsweepCache = M.Map (UnitId, PkgQual, ModuleNameWithIsBoot) [Either Driv -- module, plus one for any hs-boot files. The imports of these nodes -- are all there, including the imports of non-home-package modules. downsweep :: HscEnv + -> (GhcMessage -> AnyGhcDiagnostic) + -> Maybe Messager -> [ModSummary] -- ^ Old summaries -> [ModuleName] -- Ignore dependencies on these; treat @@ -1562,17 +1570,38 @@ downsweep :: HscEnv -- The non-error elements of the returned list all have distinct -- (Modules, IsBoot) identifiers, unless the Bool is true in -- which case there can be repeats -downsweep hsc_env old_summaries excl_mods allow_dup_roots +downsweep hsc_env diag_wrapper msg old_summaries excl_mods allow_dup_roots = do + n_jobs <- mkWorkerLimit (hsc_dflags hsc_env) + new <- rootSummariesParallel n_jobs hsc_env diag_wrapper msg summary + downsweep_imports hsc_env old_summary_map excl_mods allow_dup_roots new + where + summary = getRootSummary excl_mods old_summary_map + + -- A cache from file paths to the already summarised modules. The same file + -- can be used in multiple units so the map is also keyed by which unit the + -- file was used in. + -- Reuse these if we can because the most expensive part of downsweep is + -- reading the headers. + old_summary_map :: M.Map (UnitId, FilePath) ModSummary + old_summary_map = + M.fromList [((ms_unitid ms, msHsFilePath ms), ms) | ms <- old_summaries] + +downsweep_imports :: HscEnv + -> M.Map (UnitId, FilePath) ModSummary + -> [ModuleName] + -> Bool + -> ([(UnitId, DriverMessages)], [ModSummary]) + -> IO ([DriverMessages], [ModuleGraphNode]) +downsweep_imports hsc_env old_summaries excl_mods allow_dup_roots (root_errs, rootSummariesOk) = do - (root_errs, rootSummariesOk) <- partitionWithM getRootSummary roots -- #17549 let root_map = mkRootMap rootSummariesOk checkDuplicates root_map (deps, map0) <- loopSummaries rootSummariesOk (M.empty, root_map) - let closure_errs = checkHomeUnitsClosed (hsc_unit_env hsc_env) - let unit_env = hsc_unit_env hsc_env - let tmpfs = hsc_tmpfs hsc_env + let closure_errs = checkHomeUnitsClosed unit_env + unit_env = hsc_unit_env hsc_env + tmpfs = hsc_tmpfs hsc_env - let downsweep_errs = lefts $ concat $ M.elems map0 + downsweep_errs = lefts $ concat $ M.elems map0 downsweep_nodes = M.elems deps (other_errs, unit_nodes) = partitionEithers $ unitEnv_foldWithKey (\nodes uid hue -> nodes ++ unitModuleNodes downsweep_nodes uid hue) [] (hsc_HUG hsc_env) @@ -1604,46 +1633,6 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots [(ms_unitid ms, b, c) | (b, c) <- msDeps ms ] logger = hsc_logger hsc_env - roots = hsc_targets hsc_env - - -- A cache from file paths to the already summarised modules. The same file - -- can be used in multiple units so the map is also keyed by which unit the - -- file was used in. - -- Reuse these if we can because the most expensive part of downsweep is - -- reading the headers. - old_summary_map :: M.Map (UnitId, FilePath) ModSummary - old_summary_map = M.fromList [((ms_unitid ms, msHsFilePath ms), ms) | ms <- old_summaries] - - getRootSummary :: Target -> IO (Either (UnitId, DriverMessages) ModSummary) - getRootSummary Target { targetId = TargetFile file mb_phase - , targetContents = maybe_buf - , targetUnitId = uid - } - = do let offset_file = augmentByWorkingDirectory dflags file - exists <- liftIO $ doesFileExist offset_file - if exists || isJust maybe_buf - then first (uid,) <$> - summariseFile hsc_env home_unit old_summary_map offset_file mb_phase - maybe_buf - else return $ Left $ (uid,) $ singleMessage - $ mkPlainErrorMsgEnvelope noSrcSpan (DriverFileNotFound offset_file) - where - dflags = homeUnitEnv_dflags (ue_findHomeUnitEnv uid (hsc_unit_env hsc_env)) - home_unit = ue_unitHomeUnit uid (hsc_unit_env hsc_env) - getRootSummary Target { targetId = TargetModule modl - , targetContents = maybe_buf - , targetUnitId = uid - } - = do maybe_summary <- summariseModule hsc_env home_unit old_summary_map NotBoot - (L rootLoc modl) (ThisPkg (homeUnitId home_unit)) - maybe_buf excl_mods - case maybe_summary of - FoundHome s -> return (Right s) - FoundHomeWithError err -> return (Left err) - _ -> return $ Left $ (uid, moduleNotFoundErr modl) - where - home_unit = ue_unitHomeUnit uid (hsc_unit_env hsc_env) - rootLoc = mkGeneralSrcSpan (fsLit "") -- In a root module, the filename is allowed to diverge from the module -- name, so we have to check that there aren't multiple root files @@ -1711,7 +1700,7 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots loopImports ss done summarised | otherwise = do - mb_s <- summariseModule hsc_env home_unit old_summary_map + mb_s <- summariseModule hsc_env home_unit old_summaries is_boot wanted_mod mb_pkg Nothing excl_mods case mb_s of @@ -1736,6 +1725,90 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots GWIB { gwib_mod = L loc mod, gwib_isBoot = is_boot } = gwib wanted_mod = L loc mod +getRootSummary :: + [ModuleName] -> + M.Map (UnitId, FilePath) ModSummary -> + HscEnv -> + Target -> + IO (Either (UnitId, DriverMessages) ModSummary) +getRootSummary excl_mods old_summary_map hsc_env target + | TargetFile file mb_phase <- targetId + = do + let offset_file = augmentByWorkingDirectory dflags file + exists <- liftIO $ doesFileExist offset_file + if exists || isJust maybe_buf + then first (uid,) <$> + summariseFile hsc_env home_unit old_summary_map offset_file mb_phase + maybe_buf + else + return $ Left $ (uid,) $ singleMessage $ + mkPlainErrorMsgEnvelope noSrcSpan (DriverFileNotFound offset_file) + | TargetModule modl <- targetId + = do + maybe_summary <- summariseModule hsc_env home_unit old_summary_map NotBoot + (L rootLoc modl) (ThisPkg (homeUnitId home_unit)) + maybe_buf excl_mods + pure case maybe_summary of + FoundHome s -> Right s + FoundHomeWithError err -> Left err + _ -> Left (uid, moduleNotFoundErr modl) + where + Target {targetId, targetContents = maybe_buf, targetUnitId = uid} = target + home_unit = ue_unitHomeUnit uid (hsc_unit_env hsc_env) + rootLoc = mkGeneralSrcSpan (fsLit "") + dflags = homeUnitEnv_dflags (ue_findHomeUnitEnv uid (hsc_unit_env hsc_env)) + +-- | Execute 'getRootSummary' for the 'Target's using the parallelism pipeline +-- system. +-- Create bundles of 'Target's wrapped in a 'MakeAction' that uses +-- 'withAbstractSem' to wait for a free slot, limiting the number of +-- concurrently computed summaries to the value of the @-j@ option or the slots +-- allocated by the job server, if that is used. +-- +-- The 'MakeAction' returns 'Maybe', which is not handled as an error, because +-- 'runLoop' only sets it to 'Nothing' when an exception was thrown, so the +-- result won't be read anyway here. +-- +-- To emulate the current behavior, we funnel exceptions past the concurrency +-- barrier and rethrow the first one afterwards. +rootSummariesParallel :: + WorkerLimit -> + HscEnv -> + (GhcMessage -> AnyGhcDiagnostic) -> + Maybe Messager -> + (HscEnv -> Target -> IO (Either (UnitId, DriverMessages) ModSummary)) -> + IO ([(UnitId, DriverMessages)], [ModSummary]) +rootSummariesParallel n_jobs hsc_env diag_wrapper msg get_summary = do + (actions, get_results) <- unzip <$> mapM action_and_result (zip [1..] bundles) + runPipelines n_jobs hsc_env diag_wrapper msg actions + (sequence . catMaybes <$> sequence get_results) >>= \case + Right results -> pure (partitionEithers (concat results)) + Left exc -> throwIO exc + where + bundles = mk_bundles targets + + mk_bundles = unfoldr \case + [] -> Nothing + ts -> Just (splitAt bundle_size ts) + + bundle_size = 20 + + targets = hsc_targets hsc_env + + action_and_result (log_queue_id, ts) = do + res_var <- liftIO newEmptyMVar + pure $! (MakeAction (action log_queue_id ts) res_var, readMVar res_var) + + action log_queue_id target_bundle = do + env at MakeEnv {compile_sem} <- ask + lift $ lift $ + withAbstractSem compile_sem $ + withLoggerHsc log_queue_id env \ lcl_hsc_env -> + MC.try (mapM (get_summary lcl_hsc_env) target_bundle) >>= \case + Left e | Just (_ :: SomeAsyncException) <- fromException e -> + throwIO e + a -> pure a + -- | This function checks then important property that if both p and q are home units -- then any dependency of p, which transitively depends on q is also a home unit. -- @@ -2453,12 +2526,12 @@ wrapAction msg_wrapper hsc_env k = do let lcl_logger = hsc_logger hsc_env lcl_dynflags = hsc_dflags hsc_env print_config = initPrintConfig lcl_dynflags - let logg err = printMessages lcl_logger print_config (initDiagOpts lcl_dynflags) (msg_wrapper <$> srcErrorMessages err) + logg err = printMessages lcl_logger print_config (initDiagOpts lcl_dynflags) (msg_wrapper <$> srcErrorMessages err) -- MP: It is a bit strange how prettyPrintGhcErrors handles some errors but then we handle -- SourceError and ThreadKilled differently directly below. TODO: Refactor to use `catches` -- directly. MP should probably use safeTry here to not catch async exceptions but that will regress performance due to -- internally using forkIO. - mres <- MC.try $ liftIO $ prettyPrintGhcErrors lcl_logger $ k + mres <- MC.try $ prettyPrintGhcErrors lcl_logger $ k case mres of Right res -> return $ Just res Left exc -> do @@ -2657,7 +2730,7 @@ R.hs: module R where == Why we need to rehydrate A's ModIface before compiling R.hs After compiling A.hs we'll have a TypeEnv in which the Id for `f` has a type -type uses the AbstractTyCon T; and a TyCon for `S` that also mentions that same +that uses the AbstractTyCon T; and a TyCon for `S` that also mentions that same AbstractTyCon. (Abstract because it came from R.hs-boot; we know nothing about it.) ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -86,6 +86,20 @@ type BaseName = OsPath -- Basename of file -- ----------------------------------------------------------------------------- -- The finder's cache +{- +[Note: Monotonic addToFinderCache] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +addToFinderCache is only used by functions that return the cached value +if there is one, or by functions that always write an InstalledFound value. +Without multithreading it is then safe to always directly write the value +without checking the previously cached value. + +However, with multithreading, it is possible that another function has +written a value into cache between the lookup and the addToFinderCache call. +in this case we should check to not overwrite an InstalledFound with an +InstalledNotFound. +-} initFinderCache :: IO FinderCache initFinderCache = do @@ -100,7 +114,12 @@ initFinderCache = do addToFinderCache :: InstalledModuleWithIsBoot -> InstalledFindResult -> IO () addToFinderCache key val = - atomicModifyIORef' mod_cache $ \c -> (extendInstalledModuleWithIsBootEnv c key val, ()) + atomicModifyIORef' mod_cache $ \c -> + case (lookupInstalledModuleWithIsBootEnv c key, val) of + -- Don't overwrite an InstalledFound with an InstalledNotFound + -- See [Note Monotonic addToFinderCache] + (Just InstalledFound{}, InstalledNotFound{}) -> (c, ()) + _ -> (extendInstalledModuleWithIsBootEnv c key val, ()) lookupFinderCache :: InstalledModuleWithIsBoot -> IO (Maybe InstalledFindResult) lookupFinderCache key = do ===================================== testsuite/tests/ghc-api/downsweep/OldModLocation.hs ===================================== @@ -6,6 +6,7 @@ import GHC import GHC.Driver.Make import GHC.Driver.Session import GHC.Driver.Env +import GHC.Types.Error (mkUnknownDiagnostic) import GHC.Unit.Module.Graph import GHC.Unit.Finder @@ -47,13 +48,13 @@ main = do liftIO $ do - _emss <- downsweep hsc_env [] [] False + _emss <- downsweep hsc_env mkUnknownDiagnostic Nothing [] [] False flushFinderCaches (hsc_FC hsc_env) (hsc_unit_env hsc_env) createDirectoryIfMissing False "mydir" renameFile "B.hs" "mydir/B.hs" - (_, nodes) <- downsweep hsc_env [] [] False + (_, nodes) <- downsweep hsc_env mkUnknownDiagnostic Nothing [] [] False -- If 'checkSummaryTimestamp' were to call 'addHomeModuleToFinder' with -- (ms_location old_summary) like summariseFile used to instead of ===================================== testsuite/tests/ghc-api/downsweep/PartialDownsweep.hs ===================================== @@ -6,6 +6,7 @@ import GHC import GHC.Driver.Make import GHC.Driver.Session +import GHC.Types.Error (mkUnknownDiagnostic) import GHC.Utils.Outputable import GHC.Utils.Exception (ExceptionMonad) import GHC.Data.Bag @@ -168,7 +169,7 @@ go label mods cnd = setTargets [tgt] hsc_env <- getSession - (_, nodes) <- liftIO $ downsweep hsc_env [] [] False + (_, nodes) <- liftIO $ downsweep hsc_env mkUnknownDiagnostic Nothing [] [] False it label $ cnd (mapMaybe moduleGraphNodeModSum nodes) ===================================== utils/haddock/haddock-api/src/Haddock/Interface.hs ===================================== @@ -170,7 +170,7 @@ createIfaces verbosity modules flags instIfaceMap = do _ <- setSessionDynFlags dflags'' targets <- mapM (\(filePath, _) -> guessTarget filePath Nothing Nothing) hs_srcs setTargets targets - (_errs, modGraph) <- depanalE [] False + (_errs, modGraph) <- depanalE mkUnknownDiagnostic (Just batchMsg) [] False -- Create (if necessary) and load .hi-files. With --no-compilation this happens later. when (Flag_NoCompilation `notElem` flags) $ do View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8237a881600064e45fdcdda263405c95e4f78ea9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8237a881600064e45fdcdda263405c95e4f78ea9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 12:13:37 2024 From: gitlab at gitlab.haskell.org (Ryan Scott (@RyanGlScott)) Date: Tue, 24 Sep 2024 08:13:37 -0400 Subject: [Git][ghc/ghc][wip/T25148] Yet more review comments Message-ID: <66f2acf128716_1aaaed398c94586c5@gitlab.mail> Ryan Scott pushed to branch wip/T25148 at Glasgow Haskell Compiler / GHC Commits: 6c892383 by Ryan Scott at 2024-09-24T08:13:29-04:00 Yet more review comments - - - - - 1 changed file: - compiler/GHC/Tc/Deriv/Generate.hs Changes: ===================================== compiler/GHC/Tc/Deriv/Generate.hs ===================================== @@ -1699,12 +1699,16 @@ coercing from. So from, say, instance C a => C a (T x) where op @c = coerce @(a -> [] -> c -> Int) @(a -> [T x] -> c -> Int) - op + (op @c) In addition to the type applications, we also use a type abstraction to bring -the method-bound variable `c` into scope over the two type applications. -See Note [GND and QuantifiedConstraints] for more information on why this -is important. +the method-bound variable `c` into scope. We do this for two reasons: + +* We need to bring `c` into scope over the two type applications to `coerce`. + See Note [GND and QuantifiedConstraints] for more information on why this + is important. +* We need to bring `c` into scope over the type application to `op`. See + Note [GND and ambiguity] for more information on why this is important. (In the surface syntax, only specified type variables can be used in type abstractions. Since a method signature could contain both specified and @@ -1970,26 +1974,33 @@ ambiguous type variables. Here are a couple of examples to illustrate this: newtype T a = MkT a deriving newtype Facts - If we only generate the following code for the derived `Facts` instance: + When generating code for the derived `Facts` instance, we must use a type + abstraction to bring `n` into scope over the type applications to `coerce` + (see Note [Newtype-deriving instances] for more why this is needed). A first + attempt at generating the instance would be: instance Facts a => Facts (T a) where - fact1 = coerce @(Proxy a -> Dict (0 <= n)) - @(Proxy (T a) -> Dict (0 <= n)) - (fact1 @a) + fact1 @n = coerce @(Proxy a -> Dict (0 <= n)) + @(Proxy (T a) -> Dict (0 <= n)) + (fact1 @a) - Then it won't typecheck, as GHC won't know how to instantiate `n` in the call - to `fact1 @a`. To compensate for the possibility of ambiguity here, we bind - `n` on the LHS of `fact1`'s equation using a type abstraction and explicitly - instantiate `n` on the RHS: + This still won't typecheck, however, as GHC doesn't know how to instantiate + `n` in the call to `fact1 @a`. To compensate for the possibility of ambiguity + here, we also visibly apply `n` in the call to `fact1` on the RHS: instance Facts a => Facts (T a) where fact1 @n = coerce @(Proxy a -> Dict (0 <= n)) @(Proxy (T a) -> Dict (0 <= n)) - (fact1 @a @n) + (fact1 @a @n) -- Note the @n here! + + This takes advantage of the fact that we *already* need to bring `n` into + scope using a type abstraction, and so we are able to use it both for + instantiating the call to `coerce` and instantiating the call to `fact1`. - Note that we use this same approach for resolving ambiguity in default - methods, as described in Note [Default methods in instances] (Wrinkle: - Ambiguous types from vanilla method type signatures) in GHC.Tc.TyCl.Instance. + Note that we use this same type abstractions-based approach for resolving + ambiguity in default methods, as described in Note [Default methods in + instances] (Wrinkle: Ambiguous types from vanilla method type signatures) in + GHC.Tc.TyCl.Instance. -} gen_Newtype_binds :: SrcSpan View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6c892383d1edfa7a21cb194a60893df0e2720973 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6c892383d1edfa7a21cb194a60893df0e2720973 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 12:21:32 2024 From: gitlab at gitlab.haskell.org (Ryan Scott (@RyanGlScott)) Date: Tue, 24 Sep 2024 08:21:32 -0400 Subject: [Git][ghc/ghc][wip/T25148] Resolve ambiguous method-bound type variables in vanilla defaults and GND Message-ID: <66f2aecc9032d_1aaaed3990906563c@gitlab.mail> Ryan Scott pushed to branch wip/T25148 at Glasgow Haskell Compiler / GHC Commits: b0f71f0a by Ryan Scott at 2024-09-24T08:20:55-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 15 changed files: - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/TyCl/Instance.hs - docs/users_guide/exts/default_signatures.rst - testsuite/tests/deriving/should_compile/T14578.stderr - testsuite/tests/deriving/should_compile/T14682.stderr - testsuite/tests/deriving/should_compile/T17240.stderr - + testsuite/tests/deriving/should_compile/T25148c.hs - testsuite/tests/deriving/should_compile/all.T - testsuite/tests/deriving/should_compile/deriving-inferred-ty-arg.stderr - testsuite/tests/generics/T10604/T10604_deriving.stderr - + testsuite/tests/generics/T14266.hs - + testsuite/tests/generics/T25148a.hs - + testsuite/tests/generics/T25148b.hs - testsuite/tests/generics/all.T - testsuite/tests/unsatisfiable/UnsatDefault.stderr Changes: ===================================== compiler/GHC/Tc/Deriv/Generate.hs ===================================== @@ -1699,12 +1699,16 @@ coercing from. So from, say, instance C a => C a (T x) where op @c = coerce @(a -> [] -> c -> Int) @(a -> [T x] -> c -> Int) - op + (op @c) In addition to the type applications, we also use a type abstraction to bring -the method-bound variable `c` into scope over the two type applications. -See Note [GND and QuantifiedConstraints] for more information on why this -is important. +the method-bound variable `c` into scope. We do this for two reasons: + +* We need to bring `c` into scope over the two type applications to `coerce`. + See Note [GND and QuantifiedConstraints] for more information on why this + is important. +* We need to bring `c` into scope over the type application to `op`. See + Note [GND and ambiguity] for more information on why this is important. (In the surface syntax, only specified type variables can be used in type abstractions. Since a method signature could contain both specified and @@ -1737,6 +1741,27 @@ However, to allow VTA with polytypes we must switch on -XImpredicativeTypes locally in GHC.Tc.Deriv.genInst. See #8503 for more discussion. +The following Notes describe further nuances of GeneralizedNewtypeDeriving: + +----- +-- In GHC.Tc.Deriv +----- + +* Note [Newtype deriving] +* Note [Newtype representation] +* Note [Recursive newtypes] +* Note [Determining whether newtype-deriving is appropriate] +* Note [GND and associated type families] +* Note [Bindings for Generalised Newtype Deriving] + +----- +-- In GHC.Tc.Deriv.Generate +----- + +* Note [Newtype-deriving trickiness] +* Note [GND and QuantifiedConstraints] +* Note [GND and ambiguity] + Note [Inferred invisible patterns] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Consider the following: @@ -1915,27 +1940,68 @@ coerce with a polytype, and we can only do that with VTA or QuickLook. Note [GND and ambiguity] ~~~~~~~~~~~~~~~~~~~~~~~~ We make an effort to make the code generated through GND be robust w.r.t. -ambiguous type variables. As one example, consider the following example -(from #15637): +ambiguous type variables. Here are a couple of examples to illustrate this: + +* In this example (from #15637), the class-bound type variable `a` is ambiguous + in the type of `f`: - class C a where f :: String - instance C () where f = "foo" - newtype T = T () deriving C + class C a where + f :: String -- f :: forall a. C a => String + instance C () + where f = "foo" + newtype T = T () + deriving C -A naïve attempt and generating a C T instance would be: + A naïve attempt and generating a C T instance would be: - instance C T where - f = coerce @String @String f + instance C T where + f = coerce @String @String f -This isn't going to typecheck, however, since GHC doesn't know what to -instantiate the type variable `a` with in the call to `f` in the method body. -(Note that `f :: forall a. String`!) To compensate for the possibility of -ambiguity here, we explicitly instantiate `a` like so: + This isn't going to typecheck, however, since GHC doesn't know what to + instantiate the type variable `a` with in the call to `f` in the method body. + (Note that `f :: forall a. String`!) To compensate for the possibility of + ambiguity here, we explicitly instantiate `a` like so: - instance C T where - f = coerce @String @String (f @()) + instance C T where + f = coerce @String @String (f @()) -All better now. + All better now. + +* In this example (adapted from #25148), the ambiguity arises from the `n` + type variable bound by the type signature for `fact1`: + + class Facts a where + fact1 :: forall n. Proxy a -> Dict (0 <= n) + newtype T a = MkT a + deriving newtype Facts + + When generating code for the derived `Facts` instance, we must use a type + abstraction to bring `n` into scope over the type applications to `coerce` + (see Note [Newtype-deriving instances] for more why this is needed). A first + attempt at generating the instance would be: + + instance Facts a => Facts (T a) where + fact1 @n = coerce @(Proxy a -> Dict (0 <= n)) + @(Proxy (T a) -> Dict (0 <= n)) + (fact1 @a) + + This still won't typecheck, however, as GHC doesn't know how to instantiate + `n` in the call to `fact1 @a`. To compensate for the possibility of ambiguity + here, we also visibly apply `n` in the call to `fact1` on the RHS: + + instance Facts a => Facts (T a) where + fact1 @n = coerce @(Proxy a -> Dict (0 <= n)) + @(Proxy (T a) -> Dict (0 <= n)) + (fact1 @a @n) -- Note the @n here! + + This takes advantage of the fact that we *already* need to bring `n` into + scope using a type abstraction, and so we are able to use it both for + instantiating the call to `coerce` and instantiating the call to `fact1`. + + Note that we use this same type abstractions-based approach for resolving + ambiguity in default methods, as described in Note [Default methods in + instances] (Wrinkle: Ambiguous types from vanilla method type signatures) in + GHC.Tc.TyCl.Instance. -} gen_Newtype_binds :: SrcSpan @@ -2002,11 +2068,16 @@ gen_Newtype_binds loc' cls inst_tvs inst_tys rhs_ty `nlHsAppType` to_tau `nlHsApp` meth_app - -- The class method, applied to all of the class instance types - -- (including the representation type) to avoid potential ambiguity. - -- See Note [GND and ambiguity] + -- The class method, applied to the following types to avoid potential + -- ambiguity: + -- + -- 1. All of the class instance types (including the representation type) + -- 2. All of `to_tvbs` + -- + -- See Note [GND and ambiguity]. meth_app = foldl' nlHsAppType (nlHsVar meth_RDR) $ - filterOutInferredTypes (classTyCon cls) underlying_inst_tys + filterOutInferredTypes (classTyCon cls) underlying_inst_tys ++ -- (1) + [mkTyVarTy tv | Bndr tv spec <- to_tvbs, spec /= InferredSpec] -- (2) -- Filter out any inferred arguments, since they can't be -- applied with visible type application. ===================================== compiler/GHC/Tc/TyCl/Instance.hs ===================================== @@ -1846,8 +1846,8 @@ tcMethods skol_info dfun_id clas tyvars dfun_ev_vars inst_tys meth_bind = mkVarBind meth_id $ mkLHsWrap lam_wrapper meth_rhs ; return (meth_id, meth_bind, Nothing) } - Just (dm_name, _) -> - do { (meth_bind, inline_prags) <- mkDefMethBind inst_loc dfun_id clas sel_id dm_name + Just (dm_name, dm_spec) -> + do { (meth_bind, inline_prags) <- mkDefMethBind inst_loc dfun_id clas sel_id dm_name dm_spec ; tcMethodBody skol_info clas tyvars dfun_ev_vars inst_tys dfun_ev_binds is_derived hs_sig_fn spec_inst_prags inline_prags @@ -2194,14 +2194,15 @@ mk_meth_spec_prags meth_id spec_inst_prags spec_prags_for_me mkDefMethBind :: SrcSpan -> DFunId -> Class -> Id -> Name + -> DefMethSpec Type -> TcM (LHsBind GhcRn, [LSig GhcRn]) -- The is a default method (vanailla or generic) defined in the class --- So make a binding op = $dmop @t1 @t2 --- where $dmop is the name of the default method in the class, --- and t1,t2 are the instance types. --- See Note [Default methods in instances] for why we use --- visible type application here -mkDefMethBind loc dfun_id clas sel_id dm_name +-- So make a binding op @m1 @m2 @m3 = $dmop @i1 @i2 @m1 @m2 @m3 +-- where $dmop is the name of the default method in the class; +-- i1 and t2 are the instance types; and m1, m2, and m3 are the type variables +-- from the method's type signature. See Note [Default methods in instances] for +-- why we use visible type application here. +mkDefMethBind loc dfun_id clas sel_id dm_name dm_spec = do { logger <- getLogger ; dm_id <- tcLookupId dm_name ; let inline_prag = idInlinePragma dm_id @@ -2212,29 +2213,64 @@ mkDefMethBind loc dfun_id clas sel_id dm_name -- Copy the inline pragma (if any) from the default method -- to this version. Note [INLINE and default methods] - fn = noLocA (idName sel_id) - visible_inst_tys = [ ty | (tcb, ty) <- tyConBinders (classTyCon clas) `zip` inst_tys - , tyConBinderForAllTyFlag tcb /= Inferred ] - rhs = foldl' mk_vta (nlHsVar dm_name) visible_inst_tys - bind = L (noAnnSrcSpan loc) - $ mkTopFunBind (Generated OtherExpansion SkipPmc) fn - [mkSimpleMatch (mkPrefixFunRhs fn) (noLocA []) rhs] - ; liftIO (putDumpFileMaybe logger Opt_D_dump_deriv "Filling in method body" FormatHaskell (vcat [ppr clas <+> ppr inst_tys, - nest 2 (ppr sel_id <+> equals <+> ppr rhs)])) + nest 2 (ppr bind)])) ; return (bind, inline_prags) } where (_, _, _, inst_tys) = tcSplitDFunTy (idType dfun_id) + (_, _, sel_tau) = tcSplitMethodTy (idType sel_id) + (sel_tvbs, _) = tcSplitForAllInvisTVBinders sel_tau + + -- Compute the instance types to use in the visible type application. See + -- Note [Default methods in instances]. + visible_inst_tys = + [ ty | (tcb, ty) <- tyConBinders (classTyCon clas) `zip` inst_tys + , tyConBinderForAllTyFlag tcb /= Inferred ] + + visible_sel_tvbs = + case dm_spec of + -- When dealing with a vanilla default method, compute the type + -- variables from the method's type signature. That way, we can bind + -- them with TypeAbstractions (visible_sel_pats) and use them in the + -- visible type application (visible_sel_tys). See Note [Default methods + -- in instances] (Wrinkle: Ambiguous types from vanilla method type + -- signatures). + VanillaDM -> filter (\tvb -> binderFlag tvb /= InferredSpec) sel_tvbs + -- If we are dealing with a generic default method, on the other hand, + -- don't bother doing any of this. See Note [Default methods + -- in instances] (Wrinkle: Ambiguous types from generic default method + -- type signatures). + GenericDM {} -> [] + visible_sel_pats = map mk_ty_pat visible_sel_tvbs + visible_sel_tys = map (mkTyVarTy . binderVar) visible_sel_tvbs + + fn = noLocA (idName sel_id) + rhs = foldl' mk_vta (nlHsVar dm_name) $ + visible_inst_tys ++ visible_sel_tys + bind = L (noAnnSrcSpan loc) + $ mkTopFunBind (Generated OtherExpansion SkipPmc) fn + [mkSimpleMatch (mkPrefixFunRhs fn) (noLocA visible_sel_pats) rhs] + + mk_ty_pat :: VarBndr TyVar Specificity -> LPat GhcRn + mk_ty_pat (Bndr tv spec) = + noLocA $ + InvisPat spec $ + HsTP (HsTPRn [] [tyVarName tv] []) $ + nlHsTyVar NotPromoted $ + tyVarName tv mk_vta :: LHsExpr GhcRn -> Type -> LHsExpr GhcRn mk_vta fun ty = noLocA (HsAppType noExtField fun - (mkEmptyWildCardBndrs $ nlHsParTy $ noLocA $ XHsType ty)) + (mkEmptyWildCardBndrs $ type_to_hs_type ty)) -- NB: use visible type application -- See Note [Default methods in instances] + type_to_hs_type :: Type -> LHsType GhcRn + type_to_hs_type = parenthesizeHsType appPrec . noLocA . XHsType + ---------------------- derivBindCtxt :: Id -> Class -> [Type ] -> SDoc derivBindCtxt sel_id clas tys @@ -2277,8 +2313,8 @@ From the class decl we get $dmfoo :: forall v x. Baz v x => x -> x $dmfoo y = -Notice that the type is ambiguous. So we use Visible Type Application -to disambiguate: +Notice that the type of `v` is ambiguous. So we use Visible Type Application +(VTA) to disambiguate: $dBazIntInt = MkBaz fooIntInt fooIntInt = $dmfoo @Int @Int @@ -2291,6 +2327,151 @@ Historical note: before we had VTA we had to generate post-type-checked code, which took a lot more code, and didn't work for generic default methods. +----- +-- Wrinkle: Ambiguous types from vanilla method type signatures +----- + +In the Bar example above, the ambiguity arises from `v`, a type variable +arising from the class header. It is also possible for the ambiguity to arise +from a type variable bound by the method's type signature itself (see #14266 +and #25148). For example: + + class A t where + f :: forall x m. Monoid x => t m -> m + f = + + instance A [] + +The class declaration gives rise to the following default function: + + $dmf :: forall t. A t => forall x m. Monoid x => t m -> m + $dmf = + +And the instance declaration gives rise to generated code that looks roughly +like this: + + instance A [] where + f = $dmf @[] ... + +In this example, it is not enough to use VTA to specify the type of `t`, since +the type of `x` (bound by `f`'s type signature) is also ambiguous. We need to +generate code that looks more like this: + + instance A [] where + f = $dmf @[] @x @m + +But where should `x` and `m` be bound? It's tempting to use ScopedTypeVariables +and InstanceSigs to accomplish this: + + instance A [] where + f :: forall x m. Monoid x => [m] -> m + f = $dmf @[] @x @m + +GHC will reject this code, however, as the type signature for `f` will fail the +subtype check for InstanceSigs: + + • Could not deduce (Monoid x0) + from the context: Monoid x + bound by the type signature for: + f :: forall x m. Monoid x => [m] -> m + The type variable ‘x0’ is ambiguous + • When checking that instance signature for ‘f’ + is more general than its signature in the class + Instance sig: forall x m. Monoid x => [m] -> m + Class sig: forall x m. Monoid x => [m] -> m + In the instance declaration for ‘A []’ + +See #17898. To avoid this problem, we instead bind `x` and `m` using +TypeAbstractions: + + instance A [] where + f @x @m = $dmf @[] @x @m + +This resolves the ambiguity and avoids the need for a subtype check. (We also +use a similar trick for resolving ambiguity in GeneralizedNewtypeDeriving: see +also Note [GND and ambiguity] in GHC.Tc.Deriv.Generate.) + +----- +-- Wrinkle: Ambiguous types from generic default method type signatures +----- + +Note that the approach described above (in Wrinkle: Ambiguous types from +vanilla method type signatures) will only work for vanilla default methods and +/not/ for generic default methods (i.e., methods using DefaultSignatures). This +is because for vanilla default methods, the type of the generated $dm* function +will always quantify the same type variables as the method's original type +signature, in the same order and with the same specificities. For example, the +type of the $dmf function will be: + + $dmf :: forall t. A t => forall x m. Monoid x => t m -> m + +As such, it is guaranteed that the type variables from the method's original +type signature will line up exactly with the type variables from the $dm* +function (after instantiating all of the class variables): + + instance A [] where + f @x @m = $dmf @[] @x @m + +We cannot guarantee this property for generic default methods, however. As +such, we must be more conservative and generate code without instantiating any +of the type variables bound by the method's type signature (only the type +variables bound by the class header): + + instance A [] where + f = $dmf @[] + +There are a number of reasons why we cannot reliably instantiate the type +variables bound by a generic default method's type signature: + +* Default methods can quantify type variables in a different order, e.g., + + class A t where + f :: forall x m. Monoid x => t m -> m + default f :: forall m x. Monoid x => t m -> m + f = + + Note that the default signature quantifies the type variables in the opposite + order from the method's original type signature. As such, the type of $dmf + will be: + + $dmf :: forall t. A t => forall m x. Monoid x => t m -> m + + Therefore, `f @x @m = $dmf @[] @x @m` would be incorrect. Nor would it be + straightforward to infer what the correct order of type variables should be. + +* Default methods can quantify a different number of type variables, e.g., + + class A t where + f :: forall x m. Monoid x => t m -> m + default f :: forall p q r m. C a t p q r => t m -> m + f = + + This gives rise to: + + $dmf :: forall t. A t => forall p q r m. C a t p q r => t m -> m + + And thus generating `f @x @m = $dmf @[] @x @m` would be incorrect, for + similar reasons as in the example above. + +* Default methods can use different type variable specificities, e.g., + + class A t where + f :: forall x m. Monoid x => t m -> m + default f :: forall {x} m. Monoid x => t m -> m + f = + + This gives rise to: + + $dmf :: forall t. A t => forall {x} m. Monoid x => t m -> m + + Therefore, generating `f @x @m = $dmf @[] @x @m` would be incorrect because + the `x` in the type of $dmf is inferred, so it is not eligible for visible + type application. + +As such, we do not bother trying to resolve the ambiguity of any method-bound +type variables when dealing with generic defaults. This means that GHC won't be +able to typecheck the default method examples above, but so be it. + Note [INLINE and default methods] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Default methods need special case. They are supposed to behave rather like ===================================== docs/users_guide/exts/default_signatures.rst ===================================== @@ -189,3 +189,15 @@ default type signatures. This works, but at the expense of changing ``p``'s behavior with respect to :ref:`visible-type-application`. + +- The default signature cannot use any ambiguous type variables. For example, + GHC will not permit instances of the following class: :: + + class A t where + f :: forall x m. Monoid x => t m -> m + default :: forall x m. Monoid x => t m -> m + f = ... + + Note that ``x`` is ambiguous, as none of the argument or result types + determine what type ``x`` should be when calling ``f``. GHC will reject + instances of ``A``, even if :extension:`AllowAmbiguousTypes` is enabled. ===================================== testsuite/tests/deriving/should_compile/T14578.stderr ===================================== @@ -28,7 +28,7 @@ Derived class instances: -> T14578.App (Data.Functor.Compose.Compose f g) a) @(b -> T14578.Wat f g a -> T14578.Wat f g a) (GHC.Internal.Base.stimes - @(T14578.App (Data.Functor.Compose.Compose f g) a)) + @(T14578.App (Data.Functor.Compose.Compose f g) a) @b) instance GHC.Internal.Base.Functor f => GHC.Internal.Base.Functor (T14578.App f) where @@ -36,38 +36,38 @@ Derived class instances: = GHC.Prim.coerce @((a -> b) -> f a -> f b) @((a -> b) -> T14578.App f a -> T14578.App f b) - (GHC.Internal.Base.fmap @f) + (GHC.Internal.Base.fmap @f @a @b) (GHC.Internal.Base.<$) @a @b = GHC.Prim.coerce @(a -> f b -> f a) @(a -> T14578.App f b -> T14578.App f a) - ((GHC.Internal.Base.<$) @f) + ((GHC.Internal.Base.<$) @f @a @b) instance GHC.Internal.Base.Applicative f => GHC.Internal.Base.Applicative (T14578.App f) where GHC.Internal.Base.pure @a = GHC.Prim.coerce - @(a -> f a) @(a -> T14578.App f a) (GHC.Internal.Base.pure @f) + @(a -> f a) @(a -> T14578.App f a) (GHC.Internal.Base.pure @f @a) (GHC.Internal.Base.<*>) @a @b = GHC.Prim.coerce @(f (a -> b) -> f a -> f b) @(T14578.App f (a -> b) -> T14578.App f a -> T14578.App f b) - ((GHC.Internal.Base.<*>) @f) + ((GHC.Internal.Base.<*>) @f @a @b) GHC.Internal.Base.liftA2 @a @b @c = GHC.Prim.coerce @((a -> b -> c) -> f a -> f b -> f c) @((a -> b -> c) -> T14578.App f a -> T14578.App f b -> T14578.App f c) - (GHC.Internal.Base.liftA2 @f) + (GHC.Internal.Base.liftA2 @f @a @b @c) (GHC.Internal.Base.*>) @a @b = GHC.Prim.coerce @(f a -> f b -> f b) @(T14578.App f a -> T14578.App f b -> T14578.App f b) - ((GHC.Internal.Base.*>) @f) + ((GHC.Internal.Base.*>) @f @a @b) (GHC.Internal.Base.<*) @a @b = GHC.Prim.coerce @(f a -> f b -> f a) @(T14578.App f a -> T14578.App f b -> T14578.App f a) - ((GHC.Internal.Base.<*) @f) + ((GHC.Internal.Base.<*) @f @a @b) Derived type family instances: @@ -76,14 +76,14 @@ Derived type family instances: ==================== Filling in method body ==================== GHC.Internal.Base.Semigroup [T14578.App f a] - GHC.Internal.Base.sconcat = GHC.Internal.Base.$dmsconcat - @(T14578.App f a) + GHC.Internal.Base.sconcat + = GHC.Internal.Base.$dmsconcat @(T14578.App f a) ==================== Filling in method body ==================== GHC.Internal.Base.Semigroup [T14578.App f a] - GHC.Internal.Base.stimes = GHC.Internal.Base.$dmstimes - @(T14578.App f a) + GHC.Internal.Base.stimes @b + = GHC.Internal.Base.$dmstimes @(T14578.App f a) @b ===================================== testsuite/tests/deriving/should_compile/T14682.stderr ===================================== @@ -90,121 +90,121 @@ Derived type family instances: ==================== Filling in method body ==================== GHC.Internal.Show.Show [T14682.Foo] - GHC.Internal.Show.show = GHC.Internal.Show.$dmshow @(T14682.Foo) + GHC.Internal.Show.show = GHC.Internal.Show.$dmshow @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Show.Show [T14682.Foo] - GHC.Internal.Show.showList = GHC.Internal.Show.$dmshowList - @(T14682.Foo) + GHC.Internal.Show.showList + = GHC.Internal.Show.$dmshowList @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.dataCast1 = GHC.Internal.Data.Data.$dmdataCast1 - @(T14682.Foo) + GHC.Internal.Data.Data.dataCast1 @t @c + = GHC.Internal.Data.Data.$dmdataCast1 @T14682.Foo @t @c ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.dataCast2 = GHC.Internal.Data.Data.$dmdataCast2 - @(T14682.Foo) + GHC.Internal.Data.Data.dataCast2 @t @c + = GHC.Internal.Data.Data.$dmdataCast2 @T14682.Foo @t @c ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapT = GHC.Internal.Data.Data.$dmgmapT - @(T14682.Foo) + GHC.Internal.Data.Data.gmapT + = GHC.Internal.Data.Data.$dmgmapT @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapQl = GHC.Internal.Data.Data.$dmgmapQl - @(T14682.Foo) + GHC.Internal.Data.Data.gmapQl @r @r' + = GHC.Internal.Data.Data.$dmgmapQl @T14682.Foo @r @r' ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapQr = GHC.Internal.Data.Data.$dmgmapQr - @(T14682.Foo) + GHC.Internal.Data.Data.gmapQr @r @r' + = GHC.Internal.Data.Data.$dmgmapQr @T14682.Foo @r @r' ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapQ = GHC.Internal.Data.Data.$dmgmapQ - @(T14682.Foo) + GHC.Internal.Data.Data.gmapQ @u + = GHC.Internal.Data.Data.$dmgmapQ @T14682.Foo @u ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapQi = GHC.Internal.Data.Data.$dmgmapQi - @(T14682.Foo) + GHC.Internal.Data.Data.gmapQi @u + = GHC.Internal.Data.Data.$dmgmapQi @T14682.Foo @u ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapM = GHC.Internal.Data.Data.$dmgmapM - @(T14682.Foo) + GHC.Internal.Data.Data.gmapM @m + = GHC.Internal.Data.Data.$dmgmapM @T14682.Foo @m ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapMp = GHC.Internal.Data.Data.$dmgmapMp - @(T14682.Foo) + GHC.Internal.Data.Data.gmapMp @m + = GHC.Internal.Data.Data.$dmgmapMp @T14682.Foo @m ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapMo = GHC.Internal.Data.Data.$dmgmapMo - @(T14682.Foo) + GHC.Internal.Data.Data.gmapMo @m + = GHC.Internal.Data.Data.$dmgmapMo @T14682.Foo @m ==================== Filling in method body ==================== GHC.Classes.Eq [T14682.Foo] - GHC.Classes./= = GHC.Classes.$dm/= @(T14682.Foo) + (GHC.Classes./=) = GHC.Classes.$dm/= @T14682.Foo ==================== Filling in method body ==================== GHC.Classes.Ord [T14682.Foo] - GHC.Classes.max = GHC.Classes.$dmmax @(T14682.Foo) + GHC.Classes.max = GHC.Classes.$dmmax @T14682.Foo ==================== Filling in method body ==================== GHC.Classes.Ord [T14682.Foo] - GHC.Classes.min = GHC.Classes.$dmmin @(T14682.Foo) + GHC.Classes.min = GHC.Classes.$dmmin @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Ix.Ix [T14682.Foo] - GHC.Internal.Ix.index = GHC.Internal.Ix.$dmindex @(T14682.Foo) + GHC.Internal.Ix.index = GHC.Internal.Ix.$dmindex @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Ix.Ix [T14682.Foo] - GHC.Internal.Ix.rangeSize = GHC.Internal.Ix.$dmrangeSize - @(T14682.Foo) + GHC.Internal.Ix.rangeSize + = GHC.Internal.Ix.$dmrangeSize @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Ix.Ix [T14682.Foo] - GHC.Internal.Ix.unsafeRangeSize = GHC.Internal.Ix.$dmunsafeRangeSize - @(T14682.Foo) + GHC.Internal.Ix.unsafeRangeSize + = GHC.Internal.Ix.$dmunsafeRangeSize @T14682.Foo ===================================== testsuite/tests/deriving/should_compile/T17240.stderr ===================================== @@ -31,12 +31,12 @@ Derived type family instances: ==================== Filling in method body ==================== GHC.Classes.Eq [T17240.Nullary] - GHC.Classes./= = GHC.Classes.$dm/= @(T17240.Nullary) + (GHC.Classes./=) = GHC.Classes.$dm/= @T17240.Nullary ==================== Filling in method body ==================== GHC.Classes.Eq [T17240.T] - GHC.Classes./= = GHC.Classes.$dm/= @(T17240.T) + (GHC.Classes./=) = GHC.Classes.$dm/= @T17240.T ===================================== testsuite/tests/deriving/should_compile/T25148c.hs ===================================== @@ -0,0 +1,17 @@ +{-# LANGUAGE AllowAmbiguousTypes #-} +{-# LANGUAGE DataKinds #-} +{-# LANGUAGE DerivingStrategies #-} +module T25148c where + +import Data.Kind +import Data.Proxy +import GHC.TypeLits + +data Dict :: Constraint -> Type where + Dict :: c => Dict c + +class Facts a where + fact1 :: forall n. Proxy a -> Dict (0 <= n) + +newtype T a = MkT a + deriving newtype Facts ===================================== testsuite/tests/deriving/should_compile/all.T ===================================== @@ -150,4 +150,5 @@ test('T15798c', normal, compile, ['']) test('T24955a', normal, compile, ['']) test('T24955b', normal, compile, ['']) test('T24955c', normal, compile, ['']) +test('T25148c', normal, compile, ['']) test('deriving-inferred-ty-arg', normal, compile, ['-ddump-deriv -dsuppress-uniques']) ===================================== testsuite/tests/deriving/should_compile/deriving-inferred-ty-arg.stderr ===================================== @@ -8,7 +8,7 @@ Derived class instances: @(GHC.Internal.Data.Functor.Const.Const a b) @(GHC.Internal.Data.Functor.Const.Const (DerivingInferredTyArg.T a) b) - (DerivingInferredTyArg.m @a) + (DerivingInferredTyArg.m @a @b) instance DerivingInferredTyArg.C2 a => DerivingInferredTyArg.C2 (DerivingInferredTyArg.T a) where @@ -26,7 +26,7 @@ Derived class instances: @(GHC.Internal.Data.Functor.Const.Const a p) @(GHC.Internal.Data.Functor.Const.Const (DerivingInferredTyArg.T a) p) - (DerivingInferredTyArg.m3 @a) + (DerivingInferredTyArg.m3 @a @b) instance DerivingInferredTyArg.C4 a => DerivingInferredTyArg.C4 (DerivingInferredTyArg.T a) where @@ -35,7 +35,7 @@ Derived class instances: @(GHC.Internal.Data.Functor.Const.Const a p) @(GHC.Internal.Data.Functor.Const.Const (DerivingInferredTyArg.T a) p) - (DerivingInferredTyArg.m4 @a) + (DerivingInferredTyArg.m4 @a @b) instance DerivingInferredTyArg.C5 a => DerivingInferredTyArg.C5 (DerivingInferredTyArg.T a) where ===================================== testsuite/tests/generics/T10604/T10604_deriving.stderr ===================================== @@ -542,7 +542,7 @@ Derived type family instances: ==================== Filling in method body ==================== GHC.Internal.Base.Functor [T10604_deriving.Proxy @(*)] - GHC.Internal.Base.<$ = GHC.Internal.Base.$dm<$ - @(T10604_deriving.Proxy @(*)) + (GHC.Internal.Base.<$) @a @b + = GHC.Internal.Base.$dm<$ @(T10604_deriving.Proxy @(*)) @a @b ===================================== testsuite/tests/generics/T14266.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE AllowAmbiguousTypes #-} +module T14266 where + +class A t where + f :: forall x m. Monoid x => t m -> m + f = undefined +instance A [] ===================================== testsuite/tests/generics/T25148a.hs ===================================== @@ -0,0 +1,24 @@ +{-# LANGUAGE AllowAmbiguousTypes #-} +{-# LANGUAGE DataKinds #-} +module T25148a where + +import Data.Kind +import Data.Proxy +import GHC.TypeLits +import Unsafe.Coerce + +data Dict :: Constraint -> Type where + Dict :: c => Dict c + +fact0 :: forall n a. Proxy a -> Dict (0 <= n) +fact0 _ = unsafeCoerce (Dict :: Dict (0 <= 0)) + +class Facts a where + fact1 :: forall n. Proxy a -> Dict (0 <= n) + fact1 _ = unsafeCoerce (Dict :: Dict (0 <= 0)) +instance Facts () + +class Facts' a where + fact' :: forall n. Proxy n -> Proxy a -> Dict (0 <= n) + fact' _ _ = unsafeCoerce (Dict :: Dict (0 <= 0)) +instance Facts' () ===================================== testsuite/tests/generics/T25148b.hs ===================================== @@ -0,0 +1,42 @@ +{-# LANGUAGE DataKinds #-} +{-# LANGUAGE DeriveAnyClass #-} +{-# LANGUAGE DerivingStrategies #-} +{-# LANGUAGE TypeFamilies #-} +module T25148b where + +import Data.Functor.Const +import Data.Proxy + +class Monoid a => C a where + m :: forall {k} (b :: k). Const a b + m = Const mempty + +class Monoid a => C2 a where + m2 :: forall {k} {b :: k}. Const a b + m2 = Const mempty + +class Monoid a => C3 a where + m3 :: forall {k} (b :: k) {p :: Proxy b}. Const a p + m3 = Const mempty + +data VisProxy k (a :: k) = VisProxy + +class Monoid a => C4 a where + m4 :: forall {k} (b :: k) {p :: VisProxy k b}. Const a p + m4 = Const mempty + +type family Any :: k + +class Monoid a => C5 a where + m5 :: Proxy Any -> a + m5 _ = mempty + +----- + +data T = MkT + deriving anyclass (C, C2, C3, C4, C5) + +instance Semigroup T where + _ <> _ = MkT +instance Monoid T where + mempty = MkT ===================================== testsuite/tests/generics/all.T ===================================== @@ -45,5 +45,8 @@ test('T10361b', normal, compile, ['']) test('T11358', normal, compile_and_run, ['']) test('T12220', normal, compile, ['']) test('T15012', [extra_files(['T15012.hs', 'T15012a.hs'])], makefile_test, []) +test('T14266', normal, compile, ['']) test('T19819', normal, compile_and_run, ['']) test('T21185', normal, compile, ['']) +test('T25148a', normal, compile, ['']) +test('T25148b', normal, compile, ['']) ===================================== testsuite/tests/unsatisfiable/UnsatDefault.stderr ===================================== @@ -1,6 +1,6 @@ - UnsatDefault.hs:14:10: error: [GHC-22250] • Please define the method manually. You can try... - • In the expression: UnsatDefault.$dmmethod @(Int) - In an equation for ‘method’: method = UnsatDefault.$dmmethod @(Int) + • In the expression: UnsatDefault.$dmmethod @Int + In an equation for ‘method’: method = UnsatDefault.$dmmethod @Int In the instance declaration for ‘C Int’ + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/b0f71f0a36ed41df044de2e2918babfa07560287 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/b0f71f0a36ed41df044de2e2918babfa07560287 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 12:35:17 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Tue, 24 Sep 2024 08:35:17 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 19 commits: The X86 SIMD patch. Message-ID: <66f2b2051764c_1aaaed398cd0743c5@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 52fae8f6 by sheaf at 2024-09-24T08:34:56-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - b5b9fefd by sheaf at 2024-09-24T08:34:56-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - ff565bee by sheaf at 2024-09-24T08:34:56-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 43af2bd9 by sheaf at 2024-09-24T08:34:56-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 131cfad3 by sheaf at 2024-09-24T08:34:56-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - cd0ab6d5 by sheaf at 2024-09-24T08:34:56-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 810d58d2 by sheaf at 2024-09-24T08:34:56-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - a67c0e32 by sheaf at 2024-09-24T08:34:56-04:00 Add test for C calls & SIMD vectors - - - - - f191a1a1 by sheaf at 2024-09-24T08:34:56-04:00 Add test for #25169 - - - - - 0d31bc31 by sheaf at 2024-09-24T08:34:56-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T1969 T9198 ------------------------- - - - - - 3078ea50 by sheaf at 2024-09-24T08:34:56-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 0ab092f2 by sheaf at 2024-09-24T08:34:56-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - b0080e72 by sheaf at 2024-09-24T08:34:56-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 94631163 by sheaf at 2024-09-24T08:34:56-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 33b1abd3 by sheaf at 2024-09-24T08:34:56-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 61cc4c1d by Brandon Chinn at 2024-09-24T08:34:57-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 3f45e68e by ARATA Mizuki at 2024-09-24T08:35:00-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 7f34ebf7 by Zubin Duggal at 2024-09-24T08:35:01-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - ed046c52 by Simon Peyton Jones at 2024-09-24T08:35:01-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/PPC/Regs.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ced2b20725ba965f8ea2ae3ff494d9b162fdef6f...ed046c5272e4b73314512ca7e9a0d8c5d3139480 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ced2b20725ba965f8ea2ae3ff494d9b162fdef6f...ed046c5272e4b73314512ca7e9a0d8c5d3139480 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 12:58:14 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 08:58:14 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/romes/exceptions-propagate Message-ID: <66f2b7661bdeb_1aaaed9c8bc8938ba@gitlab.mail> Rodrigo Mesquita pushed new branch wip/romes/exceptions-propagate at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/romes/exceptions-propagate You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 13:12:52 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Tue, 24 Sep 2024 09:12:52 -0400 Subject: [Git][ghc/ghc][wip/andreask/spec-constr-args] SpecConstr: Introduce a separate argument limit for forced specs. Message-ID: <66f2bad456fec_2971c42c123040729@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/spec-constr-args at Glasgow Haskell Compiler / GHC Commits: 07ef3573 by Andreas Klebinger at 2024-09-24T14:53:31+02:00 SpecConstr: Introduce a separate argument limit for forced specs. We used to put no limit at all on specializations forced via the SPEC argument. This isn't always reasonable so we introduce a very high limit that applies to forced specializations, a flag to control it, and we now emit a warning if we fail a specialization because we exceed the warning. Fixes #25197 - - - - - 8 changed files: - compiler/GHC/Core/Opt/SpecConstr.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Session.hs - docs/users_guide/using-optimisation.rst - + testsuite/tests/simplCore/should_compile/T25197.hs - + testsuite/tests/simplCore/should_compile/T25197.stderr - + testsuite/tests/simplCore/should_compile/T25197_TH.hs - testsuite/tests/simplCore/should_compile/all.T Changes: ===================================== compiler/GHC/Core/Opt/SpecConstr.hs ===================================== @@ -20,7 +20,8 @@ ToDo [Oct 2013] module GHC.Core.Opt.SpecConstr( specConstrProgram, - SpecConstrAnnotation(..) + SpecConstrAnnotation(..), + SpecFailWarning(..) ) where import GHC.Prelude @@ -51,6 +52,7 @@ import GHC.Core.Make ( mkImpossibleExpr ) import GHC.Unit.Module import GHC.Unit.Module.ModGuts +import GHC.Types.Error (MessageClass(..), Severity(..), DiagnosticReason(WarningWithoutFlag), ResolvedDiagnosticReason (..)) import GHC.Types.Literal ( litIsLifted ) import GHC.Types.Id import GHC.Types.Id.Info ( IdDetails(..) ) @@ -526,7 +528,7 @@ sc_force to True when calling specLoop. This flag does four things: (see argToPat; #4448) (FS4) Only specialise on recursive types a finite number of times (see sc_recursive; #5550; Note [Limit recursive specialisation]) -(FS5) Lift the restriction on the maximum number of arguments which +(FS5) Use a different restriction on the maximum number of arguments which the optimisation will specialise. (see `too_many_worker_args` in `callsToNewPats`; #14003) @@ -782,16 +784,25 @@ specConstrProgram :: ModGuts -> CoreM ModGuts specConstrProgram guts = do { env0 <- initScEnv guts ; us <- getUniqueSupplyM - ; let (_usg, binds') = initUs_ us $ + ; let (_usg, binds', warnings) = initUs_ us $ scTopBinds env0 (mg_binds guts) + ; when (not (null warnings)) $ msg specConstr_warn_class (warn_msg warnings) + ; return (guts { mg_binds = binds' }) } -scTopBinds :: ScEnv -> [InBind] -> UniqSM (ScUsage, [OutBind]) -scTopBinds _env [] = return (nullUsage, []) -scTopBinds env (b:bs) = do { (usg, b', bs') <- scBind TopLevel env b $ + where + specConstr_warn_class = MCDiagnostic SevWarning (ResolvedDiagnosticReason WarningWithoutFlag) Nothing + warn_msg :: SpecFailWarnings -> SDoc + warn_msg warnings = text "SpecConstr encountered one or more function(s) with a SPEC argument that resulted in too many arguments," $$ + text "which resulted in no specialization being generated for these functions:" $$ + nest 2 (vcat (map ppr warnings)) $$ + (text "If this is expected you might want to increase -fmax-forced-spec-args to force specialization anyway.") +scTopBinds :: ScEnv -> [InBind] -> UniqSM (ScUsage, [OutBind], [SpecFailWarning]) +scTopBinds _env [] = return (nullUsage, [], []) +scTopBinds env (b:bs) = do { (usg, b', bs', warnings) <- scBind TopLevel env b $ (\env -> scTopBinds env bs) - ; return (usg, b' ++ bs') } + ; return (usg, b' ++ bs', warnings) } {- ************************************************************************ @@ -905,6 +916,11 @@ data SpecConstrOpts = SpecConstrOpts -- ^ The threshold at which a worker-wrapper transformation used as part of -- this pass will no longer happen, measured in the number of arguments. + , sc_max_forced_args :: !Int + -- ^ The threshold at which a worker-wrapper transformation used as part of + -- this pass will no longer happen even if a SPEC arg was used to force + -- specialization. Measured in the number of arguments. + , sc_debug :: !Bool -- ^ Whether to print debug information @@ -975,6 +991,7 @@ instance Outputable Value where initScOpts :: DynFlags -> Module -> SpecConstrOpts initScOpts dflags this_mod = SpecConstrOpts { sc_max_args = maxWorkerArgs dflags, + sc_max_forced_args = maxForcedSpecArgs dflags, sc_debug = hasPprDebug dflags, sc_uf_opts = unfoldingOpts dflags, sc_module = this_mod, @@ -1388,29 +1405,29 @@ creates specialised versions of functions. -} scBind :: TopLevelFlag -> ScEnv -> InBind - -> (ScEnv -> UniqSM (ScUsage, a)) -- Specialise the scope of the binding - -> UniqSM (ScUsage, [OutBind], a) + -> (ScEnv -> UniqSM (ScUsage, a, [SpecFailWarning])) -- Specialise the scope of the binding + -> UniqSM (ScUsage, [OutBind], a, [SpecFailWarning]) scBind top_lvl env (NonRec bndr rhs) do_body | isTyVar bndr -- Type-lets may be created by doBeta - = do { (final_usage, body') <- do_body (extendScSubst env bndr rhs) - ; return (final_usage, [], body') } + = do { (final_usage, body', warnings) <- do_body (extendScSubst env bndr rhs) + ; return (final_usage, [], body', warnings) } | not (isTopLevel top_lvl) -- Nested non-recursive value binding -- See Note [Specialising local let bindings] = do { let (body_env, bndr') = extendBndr env bndr -- Not necessary at top level; but here we are nested - ; rhs_info <- scRecRhs env (bndr',rhs) + ; (rhs_info, rhs_ws) <- scRecRhs env (bndr',rhs) ; let body_env2 = extendHowBound body_env [bndr'] RecFun rhs' = ri_new_rhs rhs_info body_env3 = extendValEnv body_env2 bndr' (isValue (sc_vals env) rhs') - ; (body_usg, body') <- do_body body_env3 + ; (body_usg, body', warnings_body) <- do_body body_env3 -- Now make specialised copies of the binding, -- based on calls in body_usg - ; (spec_usg, specs) <- specNonRec env (scu_calls body_usg) rhs_info + ; (spec_usg, specs, warnings_bnd) <- specNonRec env (scu_calls body_usg) rhs_info -- NB: For non-recursive bindings we inherit sc_force flag from -- the parent function (see Note [Forcing specialisation]) @@ -1419,7 +1436,7 @@ scBind top_lvl env (NonRec bndr rhs) do_body bind_usage = (body_usg `delCallsFor` [bndr']) `combineUsage` spec_usg -- Note [spec_usg includes rhs_usg] - ; return (bind_usage, spec_bnds, body') + ; return (bind_usage, spec_bnds, body', mconcat [warnings_bnd, warnings_body, rhs_ws]) } | otherwise -- Top-level, non-recursive value binding @@ -1431,15 +1448,15 @@ scBind top_lvl env (NonRec bndr rhs) do_body -- -- I tried always specialising non-recursive top-level bindings too, -- but found some regressions (see !8135). So I backed off. - = do { (rhs_usage, rhs') <- scExpr env rhs + = do { (rhs_usage, rhs', ws_rhs) <- scExpr env rhs -- At top level, we've already put all binders into scope; see initScEnv -- Hence no need to call `extendBndr`. But we still want to -- extend the `ValueEnv` to record the value of this binder. ; let body_env = extendValEnv env bndr (isValue (sc_vals env) rhs') - ; (body_usage, body') <- do_body body_env + ; (body_usage, body', body_warnings) <- do_body body_env - ; return (rhs_usage `combineUsage` body_usage, [NonRec bndr rhs'], body') } + ; return (rhs_usage `combineUsage` body_usage, [NonRec bndr rhs'], body', body_warnings ++ ws_rhs) } scBind top_lvl env (Rec prs) do_body | isTopLevel top_lvl @@ -1450,19 +1467,20 @@ scBind top_lvl env (Rec prs) do_body -- ToDo: I'm honestly not sure of the rationale of this size-testing, nor -- why it only applies at top level. But that's the way it has been -- for a while. See #21456. - do { (body_usg, body') <- do_body rhs_env2 - ; (rhs_usgs, rhss') <- mapAndUnzipM (scExpr env) rhss + do { (body_usg, body', warnings_body) <- do_body rhs_env2 + ; (rhs_usgs, rhss', rhs_ws) <- mapAndUnzip3M (scExpr env) rhss ; let all_usg = (combineUsages rhs_usgs `combineUsage` body_usg) `delCallsFor` bndrs' bind' = Rec (bndrs' `zip` rhss') - ; return (all_usg, [bind'], body') } + ; return (all_usg, [bind'], body', warnings_body ++ concat rhs_ws) } | otherwise - = do { rhs_infos <- mapM (scRecRhs rhs_env2) (bndrs' `zip` rhss) - ; (body_usg, body') <- do_body rhs_env2 + = do { (rhs_infos, rhs_wss) <- mapAndUnzipM (scRecRhs rhs_env2) (bndrs' `zip` rhss) + ; let rhs_ws = mconcat rhs_wss + ; (body_usg, body', warnings_body) <- do_body rhs_env2 - ; (spec_usg, specs) <- specRec (scForce rhs_env2 force_spec) - (scu_calls body_usg) rhs_infos + ; (spec_usg, specs, spec_ws) <- specRec (scForce rhs_env2 force_spec) + (scu_calls body_usg) rhs_infos -- Do not unconditionally generate specialisations from rhs_usgs -- Instead use them only if we find an unspecialised call -- See Note [Seeding recursive groups] @@ -1473,7 +1491,7 @@ scBind top_lvl env (Rec prs) do_body -- zipWithEqual: length of returned [SpecInfo] -- should be the same as incoming [RhsInfo] - ; return (all_usg, [bind'], body') } + ; return (all_usg, [bind'], body', mconcat [warnings_body,rhs_ws,spec_ws]) } where (bndrs,rhss) = unzip prs force_spec = any (forceSpecBndr env) bndrs -- Note [Forcing specialisation] @@ -1501,59 +1519,63 @@ recursive function, but that's not essential and might even be harmful. I'm not sure. -} +withWarnings :: SpecFailWarnings -> (ScUsage, CoreExpr, SpecFailWarnings) -> (ScUsage, CoreExpr, SpecFailWarnings) +withWarnings ws (use,expr,ws2) = (use,expr,ws ++ ws2) + ------------------------ -scExpr, scExpr' :: ScEnv -> CoreExpr -> UniqSM (ScUsage, CoreExpr) +scExpr, scExpr' :: ScEnv -> CoreExpr -> UniqSM (ScUsage, CoreExpr, SpecFailWarnings) -- The unique supply is needed when we invent -- a new name for the specialised function and its args scExpr env e = scExpr' env e scExpr' env (Var v) = case scSubstId env v of - Var v' -> return (mkVarUsage env v' [], Var v') + Var v' -> return (mkVarUsage env v' [], Var v', []) e' -> scExpr (zapScSubst env) e' scExpr' env (Type t) = let !(MkSolo ty') = scSubstTy env t - in return (nullUsage, Type ty') -scExpr' env (Coercion c) = return (nullUsage, Coercion (scSubstCo env c)) -scExpr' _ e@(Lit {}) = return (nullUsage, e) -scExpr' env (Tick t e) = do (usg, e') <- scExpr env e - return (usg, Tick (scTickish env t) e') -scExpr' env (Cast e co) = do (usg, e') <- scExpr env e - return (usg, mkCast e' (scSubstCo env co)) + in return (nullUsage, Type ty', []) +scExpr' env (Coercion c) = return (nullUsage, Coercion (scSubstCo env c), []) +scExpr' _ e@(Lit {}) = return (nullUsage, e, []) +scExpr' env (Tick t e) = do (usg, e', ws) <- scExpr env e + return (usg, Tick (scTickish env t) e', ws) +scExpr' env (Cast e co) = do (usg, e', ws) <- scExpr env e + return (usg, mkCast e' (scSubstCo env co), ws) -- Important to use mkCast here -- See Note [SpecConstr call patterns] scExpr' env e@(App _ _) = scApp env (collectArgs e) scExpr' env (Lam b e) = do let (env', b') = extendBndr env b - (usg, e') <- scExpr env' e - return (usg, Lam b' e') + (usg, e', ws) <- scExpr env' e + return (usg, Lam b' e', ws) scExpr' env (Let bind body) - = do { (final_usage, binds', body') <- scBind NotTopLevel env bind $ + = do { (final_usage, binds', body', ws) <- scBind NotTopLevel env bind $ (\env -> scExpr env body) - ; return (final_usage, mkLets binds' body') } + ; return (final_usage, mkLets binds' body', ws) } scExpr' env (Case scrut b ty alts) - = do { (scrut_usg, scrut') <- scExpr env scrut + = do { (scrut_usg, scrut', ws) <- scExpr env scrut ; case isValue (sc_vals env) scrut' of Just (ConVal args_are_work_free con args) - | args_are_work_free -> sc_con_app con args scrut' + | args_are_work_free -> sc_con_app con args scrut' ws -- Don't duplicate work!! #7865 -- See Note [ConVal work-free-ness] (1) - _other -> sc_vanilla scrut_usg scrut' + _other -> sc_vanilla scrut_usg scrut' ws } where - sc_con_app con args scrut' -- Known constructor; simplify + sc_con_app con args scrut' ws -- Known constructor; simplify = do { let Alt _ bs rhs = findAlt con alts `orElse` Alt DEFAULT [] (mkImpossibleExpr ty "SpecConstr") alt_env' = extendScSubstList env ((b,scrut') : bs `zip` trimConArgs con args) - ; scExpr alt_env' rhs } + ; (use',expr',ws_new) <- scExpr alt_env' rhs + ; return (use',expr',ws ++ ws_new) } - sc_vanilla scrut_usg scrut' -- Normal case + sc_vanilla scrut_usg scrut' ws -- Normal case = do { let (alt_env,b') = extendBndrWith RecArg env b -- Record RecArg for the components - ; (alt_usgs, alt_occs, alts') <- mapAndUnzip3M (sc_alt alt_env scrut' b') alts + ; (alt_usgs, alt_occs, alts', ws_alts) <- mapAndUnzip4M (sc_alt alt_env scrut' b') alts ; let scrut_occ = foldr combineOcc NoOcc alt_occs scrut_usg' = setScrutOcc env scrut_usg scrut' scrut_occ @@ -1563,21 +1585,21 @@ scExpr' env (Case scrut b ty alts) ; let !(MkSolo ty') = scSubstTy env ty ; return (foldr combineUsage scrut_usg' alt_usgs, - Case scrut' b' ty' alts') } + Case scrut' b' ty' alts', ws ++ concat ws_alts) } single_alt = isSingleton alts sc_alt env scrut' b' (Alt con bs rhs) = do { let (env1, bs1) = extendBndrsWith RecArg env bs (env2, bs2) = extendCaseBndrs env1 scrut' b' con bs1 - ; (usg, rhs') <- scExpr env2 rhs + ; (usg, rhs', ws) <- scExpr env2 rhs ; let (usg', b_occ:arg_occs) = lookupOccs usg (b':bs2) scrut_occ = case con of DataAlt dc -- See Note [Do not specialise evals] | not (single_alt && all deadArgOcc arg_occs) -> ScrutOcc (unitUFM dc arg_occs) _ -> UnkOcc - ; return (usg', b_occ `combineOcc` scrut_occ, Alt con bs2 rhs') } + ; return (usg', b_occ `combineOcc` scrut_occ, Alt con bs2 rhs', ws) } -- | Substitute the free variables captured by a breakpoint. @@ -1626,19 +1648,20 @@ follows. still worth specialising on x. Hence the /single-alternative/ guard. -} -scApp :: ScEnv -> (InExpr, [InExpr]) -> UniqSM (ScUsage, CoreExpr) +scApp :: ScEnv -> (InExpr, [InExpr]) -> UniqSM (ScUsage, CoreExpr, SpecFailWarnings) scApp env (Var fn, args) -- Function is a variable = assert (not (null args)) $ do { args_w_usgs <- mapM (scExpr env) args - ; let (arg_usgs, args') = unzip args_w_usgs + ; let (arg_usgs, args', arg_ws) = unzip3 args_w_usgs arg_usg = combineUsages arg_usgs + arg_w = concat arg_ws ; case scSubstId env fn of - fn'@(Lam {}) -> scExpr (zapScSubst env) (doBeta fn' args') + fn'@(Lam {}) -> withWarnings arg_w <$> scExpr (zapScSubst env) (doBeta fn' args') -- Do beta-reduction and try again Var fn' -> return (arg_usg' `combineUsage` mkVarUsage env fn' args', - mkApps (Var fn') args') + mkApps (Var fn') args', arg_w ) where -- arg_usg': see Note [Specialising on dictionaries] arg_usg' | Just cls <- isClassOpId_maybe fn' @@ -1647,7 +1670,7 @@ scApp env (Var fn, args) -- Function is a variable | otherwise = arg_usg - other_fn' -> return (arg_usg, mkApps other_fn' args') } + other_fn' -> return (arg_usg, mkApps other_fn' args', arg_w) } -- NB: doing this ignores any usage info from the substituted -- function, but I don't think that matters. If it does -- we can fix it. @@ -1661,9 +1684,9 @@ scApp env (Var fn, args) -- Function is a variable -- which it may, we can get -- (let f = ...f... in f) arg1 arg2 scApp env (other_fn, args) - = do { (fn_usg, fn') <- scExpr env other_fn - ; (arg_usgs, args') <- mapAndUnzipM (scExpr env) args - ; return (combineUsages arg_usgs `combineUsage` fn_usg, mkApps fn' args') } + = do { (fn_usg, fn', fn_ws) <- scExpr env other_fn + ; (arg_usgs, args', arg_ws) <- mapAndUnzip3M (scExpr env) args + ; return (combineUsages arg_usgs `combineUsage` fn_usg, mkApps fn' args', combineSpecWarning fn_ws (concat arg_ws)) } ---------------------- mkVarUsage :: ScEnv -> Id -> [CoreExpr] -> ScUsage @@ -1679,16 +1702,16 @@ mkVarUsage env fn args | otherwise = evalScrutOcc ---------------------- -scRecRhs :: ScEnv -> (OutId, InExpr) -> UniqSM RhsInfo +scRecRhs :: ScEnv -> (OutId, InExpr) -> UniqSM (RhsInfo, SpecFailWarnings) scRecRhs env (bndr,rhs) = do { let (arg_bndrs,body) = collectBinders rhs (body_env, arg_bndrs') = extendBndrsWith RecArg env arg_bndrs - ; (body_usg, body') <- scExpr body_env body + ; (body_usg, body', body_ws) <- scExpr body_env body ; let (rhs_usg, arg_occs) = lookupOccs body_usg arg_bndrs' ; return (RI { ri_rhs_usg = rhs_usg , ri_fn = bndr, ri_new_rhs = mkLams arg_bndrs' body' , ri_lam_bndrs = arg_bndrs, ri_lam_body = body - , ri_arg_occs = arg_occs }) } + , ri_arg_occs = arg_occs }, body_ws) } -- The arg_occs says how the visible, -- lambda-bound binders of the RHS are used -- (including the TyVar binders) @@ -1757,7 +1780,7 @@ initSpecInfo (RI { ri_rhs_usg = rhs_usg }) specNonRec :: ScEnv -> CallEnv -- Calls in body -> RhsInfo -- Structure info usage info for un-specialised RHS - -> UniqSM (ScUsage, SpecInfo) -- Usage from RHSs (specialised and not) + -> UniqSM (ScUsage, SpecInfo, [SpecFailWarning]) -- Usage from RHSs (specialised and not) -- plus details of specialisations specNonRec env body_calls rhs_info @@ -1767,11 +1790,12 @@ specNonRec env body_calls rhs_info specRec :: ScEnv -> CallEnv -- Calls in body -> [RhsInfo] -- Structure info and usage info for un-specialised RHSs - -> UniqSM (ScUsage, [SpecInfo]) -- Usage from all RHSs (specialised and not) + -> UniqSM (ScUsage, [SpecInfo], SpecFailWarnings) + -- Usage from all RHSs (specialised and not) -- plus details of specialisations specRec env body_calls rhs_infos - = go 1 body_calls nullUsage (map initSpecInfo rhs_infos) + = go 1 body_calls nullUsage (map initSpecInfo rhs_infos) [] -- body_calls: see Note [Seeding recursive groups] -- NB: 'go' always calls 'specialise' once, which in turn unleashes -- si_mb_unspec if there are any boring calls in body_calls, @@ -1786,23 +1810,25 @@ specRec env body_calls rhs_infos -- Two accumulating parameters: -> ScUsage -- Usage from earlier specialisations -> [SpecInfo] -- Details of specialisations so far - -> UniqSM (ScUsage, [SpecInfo]) - go n_iter seed_calls usg_so_far spec_infos + -> SpecFailWarnings -- Warnings so far + -> UniqSM (ScUsage, [SpecInfo], SpecFailWarnings) + go n_iter seed_calls usg_so_far spec_infos ws_so_far = -- pprTrace "specRec3" (vcat [ text "bndrs" <+> ppr (map ri_fn rhs_infos) -- , text "iteration" <+> int n_iter -- , text "spec_infos" <+> ppr (map (map os_pat . si_specs) spec_infos) -- ]) $ do { specs_w_usg <- zipWithM (specialise env seed_calls) rhs_infos spec_infos - ; let (extra_usg_s, all_spec_infos) = unzip specs_w_usg + + ; let (extra_usg_s, all_spec_infos, extra_ws ) = unzip3 specs_w_usg extra_usg = combineUsages extra_usg_s all_usg = usg_so_far `combineUsage` extra_usg new_calls = scu_calls extra_usg - ; go_again n_iter new_calls all_usg all_spec_infos } + ; go_again n_iter new_calls all_usg all_spec_infos (ws_so_far ++ concat extra_ws) } -- go_again deals with termination - go_again n_iter seed_calls usg_so_far spec_infos + go_again n_iter seed_calls usg_so_far spec_infos ws_so_far | isEmptyVarEnv seed_calls - = return (usg_so_far, spec_infos) + = return (usg_so_far, spec_infos, ws_so_far) -- Limit recursive specialisation -- See Note [Limit recursive specialisation] @@ -1816,10 +1842,10 @@ specRec env body_calls rhs_infos -- for the unspecialised function, since it may now be called -- pprTrace "specRec2" (ppr (map (map os_pat . si_specs) spec_infos)) $ let rhs_usgs = combineUsages (mapMaybe si_mb_unspec spec_infos) - in return (usg_so_far `combineUsage` rhs_usgs, spec_infos) + in return (usg_so_far `combineUsage` rhs_usgs, spec_infos, ws_so_far) | otherwise - = go (n_iter + 1) seed_calls usg_so_far spec_infos + = go (n_iter + 1) seed_calls usg_so_far spec_infos ws_so_far -- See Note [Limit recursive specialisation] the_limit = case sc_count opts of @@ -1832,7 +1858,7 @@ specialise -> CallEnv -- Info on newly-discovered calls to this function -> RhsInfo -> SpecInfo -- Original RHS plus patterns dealt with - -> UniqSM (ScUsage, SpecInfo) -- New specialised versions and their usage + -> UniqSM (ScUsage, SpecInfo, [SpecFailWarning]) -- New specialised versions and their usage -- See Note [spec_usg includes rhs_usg] @@ -1850,7 +1876,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs | isDeadEndId fn -- Note [Do not specialise diverging functions] -- /and/ do not generate specialisation seeds from its RHS = -- pprTrace "specialise bot" (ppr fn) $ - return (nullUsage, spec_info) + return (nullUsage, spec_info, []) | not (isNeverActive (idInlineActivation fn)) -- See Note [Transfer activation] @@ -1861,7 +1887,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs , not (null arg_bndrs) -- Only specialise functions , Just all_calls <- lookupVarEnv bind_calls fn -- Some calls to it = -- pprTrace "specialise entry {" (ppr fn <+> ppr all_calls) $ - do { (boring_call, pats_discarded, new_pats) + do { (boring_call, pats_discarded, new_pats, warnings) <- callsToNewPats env fn spec_info arg_occs all_calls ; let n_pats = length new_pats @@ -1876,7 +1902,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs -- , text "new_pats" <+> ppr new_pats]) ; let spec_env = decreaseSpecCount env n_pats - ; (spec_usgs, new_specs) <- mapAndUnzipM (spec_one spec_env fn arg_bndrs body) + ; (spec_usgs, new_specs, new_wss) <- mapAndUnzip3M (spec_one spec_env fn arg_bndrs body) (new_pats `zip` [spec_count..]) -- See Note [Specialise original body] @@ -1900,15 +1926,16 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs ; return (new_usg, SI { si_specs = new_specs ++ specs , si_n_specs = spec_count + n_pats - , si_mb_unspec = mb_unspec' }) } + , si_mb_unspec = mb_unspec' } + ,warnings ++ concat new_wss) } | otherwise -- No calls, inactive, or not a function -- Behave as if there was a single, boring call = -- pprTrace "specialise inactive" (ppr fn $$ ppr mb_unspec) $ case mb_unspec of -- Behave as if there was a single, boring call - Just rhs_usg -> return (rhs_usg, spec_info { si_mb_unspec = Nothing }) + Just rhs_usg -> return (rhs_usg, spec_info { si_mb_unspec = Nothing }, []) -- See Note [spec_usg includes rhs_usg] - Nothing -> return (nullUsage, spec_info) + Nothing -> return (nullUsage, spec_info, []) --------------------- @@ -1917,7 +1944,7 @@ spec_one :: ScEnv -> [InVar] -- Lambda-binders of RHS; should match patterns -> InExpr -- Body of the original function -> (CallPat, Int) - -> UniqSM (ScUsage, OneSpec) -- Rule and binding + -> UniqSM (ScUsage, OneSpec, SpecFailWarnings) -- Rule and binding, warnings if any -- spec_one creates a specialised copy of the function, together -- with a rule for using it. I'm very proud of how short this @@ -1969,7 +1996,7 @@ spec_one env fn arg_bndrs body (call_pat, rule_number) -- Specialise the body -- ; pprTraceM "body_subst_for" $ ppr (spec_occ) $$ ppr (sc_subst body_env) - ; (spec_usg, spec_body) <- scExpr body_env body + ; (spec_usg, spec_body, body_warnings) <- scExpr body_env body -- And build the results ; (qvars', pats') <- generaliseDictPats qvars pats @@ -2018,7 +2045,7 @@ spec_one env fn arg_bndrs body (call_pat, rule_number) -- ] ; return (spec_usg, OS { os_pat = call_pat, os_rule = rule , os_id = spec_id - , os_rhs = spec_rhs }) } + , os_rhs = spec_rhs }, body_warnings) } generaliseDictPats :: [Var] -> [CoreExpr] -- Quantified vars and pats -> UniqSM ([Var], [CoreExpr]) -- New quantified vars and pats @@ -2402,12 +2429,26 @@ instance Outputable CallPat where , text "cp_args =" <+> ppr args , text "cp_strict_args = " <> ppr strict ]) +newtype SpecFailWarning = SpecFailForcedArgCount { spec_failed_fun_name :: Name } + +type SpecFailWarnings = [SpecFailWarning] + +instance Outputable SpecFailWarning where + ppr (SpecFailForcedArgCount name) = ppr name <+> pprDefinedAt name + +combineSpecWarning :: SpecFailWarnings -> SpecFailWarnings -> SpecFailWarnings +combineSpecWarning = (++) + +data ArgCountResult = WorkerSmallEnough | WorkerTooLarge | WorkerTooLargeForced Name + callsToNewPats :: ScEnv -> Id -> SpecInfo -> [ArgOcc] -> [Call] -> UniqSM ( Bool -- At least one boring call , Bool -- Patterns were discarded - , [CallPat] ) -- Patterns to specialise + , [CallPat] -- Patterns to specialise + , [SpecFailWarning] -- Things that didn't specialise we want to warn the user about) + ) -- Result has no duplicate patterns, -- nor ones mentioned in si_specs (hence "new" patterns) -- Bool indicates that there was at least one boring pattern @@ -2433,12 +2474,18 @@ callsToNewPats env fn spec_info@(SI { si_specs = done_specs }) bndr_occs calls non_dups = subsumePats in_scope new_pats -- Remove ones that have too many worker variables - small_pats = filterOut too_many_worker_args non_dups + (small_pats, arg_count_warnings) = partitionByWorkerSize too_many_worker_args non_dups - too_many_worker_args _ - | sc_force env = False -- See (FS5) of Note [Forcing specialisation] + -- too_many_worker_args :: CallPat -> Either SpecFailWarning Bool too_many_worker_args (CP { cp_qvars = vars, cp_args = args }) - = not (isWorkerSmallEnough (sc_max_args $ sc_opts env) (valArgCount args) vars) + | sc_force env + -- See (FS5) of Note [Forcing specialisation] + = if (isWorkerSmallEnough (sc_max_forced_args $ sc_opts env) (valArgCount args) vars) + then WorkerSmallEnough + else WorkerTooLargeForced (idName fn) + | (isWorkerSmallEnough (sc_max_args $ sc_opts env) (valArgCount args) vars) + = WorkerSmallEnough + | otherwise = WorkerTooLarge -- We are about to construct w/w pair in 'spec_one'. -- Omit specialisation leading to high arity workers. -- See Note [Limit w/w arity] in GHC.Core.Opt.WorkWrap.Utils @@ -2454,10 +2501,21 @@ callsToNewPats env fn spec_info@(SI { si_specs = done_specs }) bndr_occs calls -- , text "done_specs:" <+> ppr (map os_pat done_specs) -- , text "trimmed_pats:" <+> ppr trimmed_pats ]) - ; return (have_boring_call, pats_were_discarded, trimmed_pats) } + ; return (have_boring_call, pats_were_discarded, trimmed_pats, arg_count_warnings) } -- If any of the calls does not give rise to a specialisation, either -- because it is boring, or because there are too many specialisations, -- return a flag to say so, so that we know to keep the original function. + where + partitionByWorkerSize worker_size pats = go pats [] [] + where + go [] small warnings = (small, warnings) + go (p:ps) small warnings + | WorkerSmallEnough <- worker_size p + = go ps (p:small) warnings + | WorkerTooLarge <- worker_size p + = go ps small warnings + | WorkerTooLargeForced name <- worker_size p + = go ps small (SpecFailForcedArgCount name : warnings) trim_pats :: ScEnv -> Id -> SpecInfo -> [CallPat] -> (Bool, [CallPat]) ===================================== compiler/GHC/Driver/DynFlags.hs ===================================== @@ -394,6 +394,7 @@ data DynFlags = DynFlags { unfoldingOpts :: !UnfoldingOpts, maxWorkerArgs :: Int, + maxForcedSpecArgs :: Int, ghciHistSize :: Int, @@ -676,6 +677,7 @@ defaultDynFlags mySettings = unfoldingOpts = defaultUnfoldingOpts, maxWorkerArgs = 10, + maxForcedSpecArgs = 333, ghciHistSize = 50, -- keep a log of length 50 by default ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -1820,6 +1820,8 @@ dynamic_flags_deps = [ , make_ord_flag defFlag "fmax-worker-args" (intSuffix (\n d -> d {maxWorkerArgs = n})) + , make_ord_flag defFlag "fmax-forced-spec-args" + (intSuffix (\n d -> d {maxForcedSpecArgs = n})) , make_ord_flag defGhciFlag "fghci-hist-size" (intSuffix (\n d -> d {ghciHistSize = n})) , make_ord_flag defGhcFlag "fmax-inline-alloc-size" ===================================== docs/users_guide/using-optimisation.rst ===================================== @@ -870,6 +870,21 @@ as such you shouldn't need to set any of them explicitly. A flag value arguments of the resulting worker exceeds both that of the original function and this setting. +.. ghc-flag:: -fmax-forced-spec-args=⟨n⟩ + :shortdesc: *default: 333.* Maximum number of value arguments for forced SpecConstr specializations. + :type: dynamic + :category: + + :default: 512 + + When using ``SPEC`` from ``GHC.Types`` to force SpecConstr to fire on a function + sometimes this can result in functions taking a ridicolously large number of arguments + resulting a very large compile time hits for minor performance benefits. + + Since this is usually unintended we prevent SpecConstr from firing and generate + a warning if the number of arguments in the resulting function would exceed + the value given by ``-fmax-forced-spec-args``. + .. ghc-flag:: -fno-opt-coercion :shortdesc: Turn off the coercion optimiser :type: dynamic ===================================== testsuite/tests/simplCore/should_compile/T25197.hs ===================================== @@ -0,0 +1,30 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25197 where + +import T25197_TH +import GHC.Exts + +{- +This test applies a large statically known data structure to a function with +a SPEC argument, forcing the function to be specialised for the argument. +However when the complete structure of the argument is not statically known, +or as here the leaves of the structures are primitive literals for which we do +not specialize this results in a specialized function that can take hundreds of +arguments. + +Typically this is not intended, therefore we use a limit on the number of +arguments for specializations. As at some point this sort of specialization +comes with a heavy compile time cost. However we allow users to specify this +limit just in case they really depend on this sort of specialization. +-} + +foo :: [a] -> Int +foo = go SPEC + where + go :: SPEC -> [a] -> Int + go s [] = s `seq` 0 + go s (_:xs) = 1 + go s xs + +main :: IO () +main = print $ foo $(gen 1000) ===================================== testsuite/tests/simplCore/should_compile/T25197.stderr ===================================== @@ -0,0 +1,8 @@ +[1 of 2] Compiling T25197_TH ( T25197_TH.hs, T25197_TH.o ) +[2 of 2] Compiling T25197 ( T25197.hs, T25197.o ) +T25197.hs: warning: + SpecConstr encountered one or more function(s) with a SPEC argument that resulted in too many arguments, + which resulted in no specialization being generated for these functions: + $wgo Defined at T25197.hs:26:5 + If this is expected you might want to increase -fmax-forced-spec-args to force specialization anyway. + ===================================== testsuite/tests/simplCore/should_compile/T25197_TH.hs ===================================== @@ -0,0 +1,9 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25197_TH where + +import Language.Haskell.TH.Syntax + +gen :: Int -> Q Exp +gen 0 = [| [] |] +gen n = [| $(lift (show n)) : $(gen (n-1)) |] ===================================== testsuite/tests/simplCore/should_compile/all.T ===================================== @@ -530,3 +530,4 @@ test('T24625', [ grep_errmsg(r'case lazy') ], compile, ['-O -fno-ignore-asserts test('T24725a', [ grep_errmsg(r'testedRule')], compile, ['-O -ddump-rule-firings']) test('T25033', normal, compile, ['-O']) test('T25160', normal, compile, ['-O -ddump-rules']) +test('T25197', [req_th, extra_files(["T25197_TH.hs"]), only_ways(['optasm'])], multimod_compile, ['T25197', '-O2']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/07ef357348716cc7a9e02d1ab190cd81e7e40fa2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/07ef357348716cc7a9e02d1ab190cd81e7e40fa2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 14:49:55 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 10:49:55 -0400 Subject: [Git][ghc/ghc][wip/romes/exceptions-propagate] fixup! Freeze call stack in error throwing functions Message-ID: <66f2d193d4d66_2eb0994a89a4968f@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exceptions-propagate at Glasgow Haskell Compiler / GHC Commits: fa66a31a by Rodrigo Mesquita at 2024-09-24T15:34:39+01:00 fixup! Freeze call stack in error throwing functions - - - - - 30 changed files: - libraries/base/tests/T19288.stderr - libraries/base/tests/T24807.stderr - testsuite/tests/arityanal/should_run/T21694a.stderr - testsuite/tests/arityanal/should_run/T24296.stderr - testsuite/tests/codeGen/should_run/T16846.stderr - testsuite/tests/codeGen/should_run/T5626.stderr - testsuite/tests/codeGen/should_run/cgrun016.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/codeGen/should_run/cgrun045.stderr - testsuite/tests/codeGen/should_run/cgrun051.stderr - testsuite/tests/codeGen/should_run/cgrun059.stderr - testsuite/tests/concurrent/should_run/T13330.stderr - testsuite/tests/concurrent/should_run/conc021.stderr - testsuite/tests/deSugar/should_run/Or5.stderr - testsuite/tests/deSugar/should_run/T11193.stderr - testsuite/tests/deSugar/should_run/T11572.stderr - testsuite/tests/deSugar/should_run/T11601.stderr - testsuite/tests/dmdanal/should_run/T12368.stderr - testsuite/tests/dmdanal/should_run/T12368a.stderr - testsuite/tests/dmdanal/should_run/T23208.stderr - testsuite/tests/dmdanal/should_run/strun002.stderr - testsuite/tests/driver/multipleHomeUnits/multipleHomeUnits_callstack.stderr - testsuite/tests/ffi/should_run/ffi008.stderr - testsuite/tests/hpc/simple/tixs/T10529a.stderr - testsuite/tests/hpc/simple/tixs/T10529b.stderr - testsuite/tests/hpc/simple/tixs/T10529c.stderr - testsuite/tests/rebindable/RebindableFailA.stderr - testsuite/tests/safeHaskell/safeLanguage/SafeLang09.stderr - testsuite/tests/simplCore/should_fail/T7411.stderr - testsuite/tests/simplCore/should_run/T16066.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fa66a31a747a8d71c8369124a8756ec9f4900fd0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fa66a31a747a8d71c8369124a8756ec9f4900fd0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 15:19:40 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 11:19:40 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] 4 commits: Fix toException method for ExceptionWithContext Message-ID: <66f2d88c22b32_2eb0996a94741053ee@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: a0c2dadd by Matthew Pickering at 2024-09-24T16:19:25+01:00 Fix toException method for ExceptionWithContext Fixes #25235 - - - - - c8317f80 by Matthew Pickering at 2024-09-24T16:19:25+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. - - - - - a1a52283 by Rodrigo Mesquita at 2024-09-24T16:19:25+01:00 Add unicode drawing lines - - - - - 3d75a03f by Rodrigo Mesquita at 2024-09-24T16:19:26+01:00 fixup! Add unicode drawing lines - - - - - 10 changed files: - libraries/base/src/Control/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Control/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Control/Exception/Base.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs - libraries/ghc-internal/src/GHC/Internal/IO.hs - libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Text.hs - libraries/ghc-internal/src/GHC/Internal/System/IO/Error.hs Changes: ===================================== libraries/base/src/Control/Exception.hs ===================================== @@ -38,6 +38,8 @@ module Control.Exception someExceptionContext, annotateIO, ExceptionWithContext(..), + WhileHandling(..), + -- * Concrete exception types IOException, ArithException(..), @@ -65,6 +67,7 @@ module Control.Exception -- * Throwing exceptions throw, throwIO, + rethrowIO, ioError, throwTo, -- * Catching Exceptions @@ -73,6 +76,7 @@ module Control.Exception -- $catchall -- ** The @catch@ functions catch, + catchNoPropagate, catches, Handler(..), catchJust, @@ -81,6 +85,7 @@ module Control.Exception handleJust, -- ** The @try@ functions try, + tryWithContext, tryJust, -- ** The @evaluate@ function evaluate, @@ -111,11 +116,11 @@ module Control.Exception bracketOnError, finally, onException + ) where import GHC.Internal.Control.Exception import GHC.Internal.Exception.Type -import GHC.Internal.IO (annotateIO) {- $catching ===================================== libraries/ghc-internal/src/GHC/Internal/Control/Exception.hs ===================================== @@ -62,20 +62,25 @@ module GHC.Internal.Control.Exception ( -- * Throwing exceptions throw, throwIO, + rethrowIO, ioError, throwTo, -- ** The @catch@ functions catch, + catchNoPropagate, catches, Handler(..), catchJust, + -- ** Exception annotation + -- ** The @handle@ functions handle, handleJust, -- ** The @try@ functions try, + tryWithContext, tryJust, -- ** The @evaluate@ function @@ -105,6 +110,12 @@ module GHC.Internal.Control.Exception ( finally, onException, + -- * Annotating exceptions + + ExceptionContext(..), + annotateIO, + WhileHandling(..), + ) where import GHC.Internal.Control.Exception.Base ===================================== libraries/ghc-internal/src/GHC/Internal/Control/Exception/Base.hs ===================================== @@ -46,6 +46,7 @@ module GHC.Internal.Control.Exception.Base ( -- * Throwing exceptions throwIO, + rethrowIO, throw, ioError, throwTo, @@ -54,6 +55,7 @@ module GHC.Internal.Control.Exception.Base ( -- ** The @catch@ functions catch, + catchNoPropagate, catchJust, -- ** The @handle@ functions @@ -62,6 +64,7 @@ module GHC.Internal.Control.Exception.Base ( -- ** The @try@ functions try, + tryWithContext, tryJust, onException, @@ -85,6 +88,13 @@ module GHC.Internal.Control.Exception.Base ( assert, + -- * Annotating exceptions + + ExceptionContext(..), + annotateIO, + WhileHandling(..), + + -- * Utilities bracket, @@ -105,6 +115,7 @@ import GHC.Internal.Base import GHC.Internal.Exception import GHC.Internal.IO hiding (bracket, finally, onException) import GHC.Internal.IO.Exception +import GHC.Internal.Exception.Type import GHC.Internal.Show -- import GHC.Internal.Exception hiding ( Exception ) import GHC.Internal.Conc.Sync @@ -132,10 +143,11 @@ catchJust -> IO a -- ^ Computation to run -> (b -> IO a) -- ^ Handler -> IO a -catchJust p a handler = catch a handler' - where handler' e = case p e of - Nothing -> throwIO e - Just b -> handler b +catchJust p a handler = catchNoPropagate a handler' + where handler' ec@(ExceptionWithContext _ e) = + case p e of + Nothing -> rethrowIO ec + Just b -> annotateIO (whileHandling ec) (handler b) -- | A version of 'catch' with the arguments swapped around; useful in -- situations where the code for the handler is shorter. For example: @@ -176,23 +188,23 @@ mapException f v = unsafePerformIO (catch (evaluate v) try :: Exception e => IO a -> IO (Either e a) try a = catch (a >>= \ v -> return (Right v)) (\e -> return (Left e)) +-- | Like 'try' but also returns the exception context, which is useful if you intend +-- to rethrow the exception later. +tryWithContext :: Exception e => IO a -> IO (Either (ExceptionWithContext e) a) +tryWithContext a = catchNoPropagate (a >>= \v -> return (Right v)) (\e -> return (Left e)) + -- | A variant of 'try' that takes an exception predicate to select -- which exceptions are caught (c.f. 'catchJust'). If the exception -- does not match the predicate, it is re-thrown. tryJust :: Exception e => (e -> Maybe b) -> IO a -> IO (Either b a) -tryJust p a = do - r <- try a - case r of - Right v -> return (Right v) - Left e -> case p e of - Nothing -> throwIO e - Just b -> return (Left b) +tryJust p a = catchJust p (Right `fmap` a) (return . Left) -- | Like 'finally', but only performs the final action if there was an -- exception raised by the computation. onException :: IO a -> IO b -> IO a -onException io what = io `catch` \e -> do _ <- what - throwIO (e :: SomeException) +onException io what = io `catchNoPropagate` \e -> do + _ <- what + rethrowIO (e :: ExceptionWithContext SomeException) ----------------------------------------------------------------------------- -- Some Useful Functions ===================================== libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs ===================================== @@ -39,13 +39,16 @@ module GHC.Internal.Exception.Type , emptyExceptionContext , mergeExceptionContext , ExceptionWithContext(..) + -- * Exception propagation + , WhileHandling(..) + , whileHandling -- * Arithmetic exceptions , ArithException(..) , divZeroException, overflowException, ratioZeroDenomException , underflowException ) where -import GHC.Internal.Data.OldList (intersperse) +import GHC.Internal.Data.OldList (intersperse, lines, unlines, null) import GHC.Internal.Data.Maybe import GHC.Internal.Data.Typeable (Typeable, TypeRep, cast) import qualified GHC.Internal.Data.Typeable as Typeable @@ -69,6 +72,27 @@ will be subject to defaulting, as described above. -} type HasExceptionContext = (?exceptionContext :: ExceptionContext) +{- | @WhileHandling@ is used to annotate rethrow exceptions. By inspecting + the @WhileHandling@ annotation, all the places the exception has been rethrow + can be recovered. +-} + +data WhileHandling = WhileHandling SomeException deriving Show + +instance ExceptionAnnotation WhileHandling where + displayExceptionAnnotation (WhileHandling e) = + "While handling " ++ case lines $ displayException e of + [] -> "" + (l1:ls) -> + -- Indent lines forward. + -- displayException may be ill prepared for this?... + unlines $ l1:[if null l then " |" else " | " ++ l | l <- ls] + + +-- | Create 'SomeException' from an explicit context and exception. +whileHandling :: Exception e => ExceptionWithContext e -> WhileHandling +whileHandling e = WhileHandling (toException e) + {- | The @SomeException@ type is the root of the exception type hierarchy. When an exception of type @e@ is thrown, behind the scenes it is @@ -258,8 +282,10 @@ instance Show a => Show (ExceptionWithContext a) where instance Exception a => Exception (ExceptionWithContext a) where toException (ExceptionWithContext ctxt e) = - SomeException e - where ?exceptionContext = ctxt + case toException e of + SomeException c -> + let ?exceptionContext = ctxt + in SomeException c fromException se = do e <- fromException se return (ExceptionWithContext (someExceptionContext se) e) ===================================== libraries/ghc-internal/src/GHC/Internal/IO.hs ===================================== @@ -40,7 +40,8 @@ module GHC.Internal.IO ( FilePath, - catch, catchException, catchAny, throwIO, + catch, catchNoPropagate, catchException, catchExceptionNoPropagate, catchAny, throwIO, + rethrowIO, mask, mask_, uninterruptibleMask, uninterruptibleMask_, MaskingState(..), getMaskingState, unsafeUnmask, interruptible, @@ -51,7 +52,7 @@ module GHC.Internal.IO ( import GHC.Internal.Base import GHC.Internal.ST import GHC.Internal.Exception -import GHC.Internal.Exception.Type (NoBacktrace(..)) +import GHC.Internal.Exception.Type (NoBacktrace(..), WhileHandling(..), HasExceptionContext, ExceptionWithContext(..)) import GHC.Internal.Show import GHC.Internal.IO.Unsafe import GHC.Internal.Unsafe.Coerce ( unsafeCoerce ) @@ -152,6 +153,10 @@ have to work around that in the definition of catch below). catchException :: Exception e => IO a -> (e -> IO a) -> IO a catchException !io handler = catch io handler +-- | A variant of 'catchException' which does not annotate the handler with 'WhileHandling' +catchExceptionNoPropagate :: Exception e => IO a -> (ExceptionWithContext e -> IO a) -> IO a +catchExceptionNoPropagate !io handler = catchNoPropagate io handler + -- | This is the simplest of the exception-catching functions. It -- takes a single argument, runs it, and if an exception is raised -- the \"handler\" is executed, with the value of the exception passed as an @@ -191,6 +196,23 @@ catch :: Exception e -> IO a -- See #exceptions_and_strictness#. catch (IO io) handler = IO $ catch# io handler' + where + handler' e = + case fromException e of + Just e' -> unIO (annotateIO (WhileHandling e) (handler e')) + Nothing -> raiseIO# e + +-- | A variant of 'catch' which doesn't annotate the handler with the exception +-- which was caught. This function should be used when you are implementing your own +-- error handling functions which may rethrow the exceptions. +-- +-- In the case where you rethrow an exception without modifying it, you should +-- rethrow the exception with the old exception context. +catchNoPropagate :: Exception e + => IO a + -> (ExceptionWithContext e -> IO a) + -> IO a +catchNoPropagate (IO io) handler = IO $ catch# io handler' where handler' e = case fromException e of @@ -202,10 +224,12 @@ catch (IO io) handler = IO $ catch# io handler' -- Note that this function is /strict/ in the action. That is, -- @catchAny undefined b == _|_ at . See #exceptions_and_strictness# for -- details. -catchAny :: IO a -> (forall e . Exception e => e -> IO a) -> IO a +-- +-- If you rethrow an exception, you should reuse the supplied ExceptionContext. +catchAny :: IO a -> (forall e . (HasExceptionContext, Exception e) => e -> IO a) -> IO a catchAny !(IO io) handler = IO $ catch# io handler' where - handler' (SomeException e) = unIO (handler e) + handler' se@(SomeException e) = unIO (annotateIO (WhileHandling se) (handler e)) -- | Execute an 'IO' action, adding the given 'ExceptionContext' -- to any thrown synchronous exceptions. @@ -260,6 +284,11 @@ throwIO e = do se <- toExceptionWithBacktrace e IO (raiseIO# se) +-- | A utility to use when rethrowing exceptions, no new backtrace will be attached +-- when rethrowing an exception but you must supply the existing context. +rethrowIO :: Exception e => ExceptionWithContext e -> IO a +rethrowIO e = throwIO (NoBacktrace e) + -- ----------------------------------------------------------------------------- -- Controlling asynchronous exception delivery @@ -332,9 +361,9 @@ getMaskingState = IO $ \s -> _ -> MaskedInterruptible #) onException :: IO a -> IO b -> IO a -onException io what = io `catchException` \e -> do +onException io what = io `catchExceptionNoPropagate` \e -> do _ <- what - throwIO $ NoBacktrace (e :: SomeException) + rethrowIO (e :: ExceptionWithContext SomeException) -- | Executes an IO computation with asynchronous -- exceptions /masked/. That is, any thread which attempts to raise ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs ===================================== @@ -315,7 +315,7 @@ ioException :: HasCallStack => IOException -> IO a ioException err = throwIO err -- | Raise an 'IOError' in the 'IO' monad. -ioError :: IOError -> IO a +ioError :: HasCallStack => IOError -> IO a ioError = ioException -- --------------------------------------------------------------------------- ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs ===================================== @@ -1,4 +1,5 @@ {-# LANGUAGE Trustworthy #-} +{-# LANGUAGE TypeApplications #-} {-# LANGUAGE CPP, NoImplicitPrelude #-} ----------------------------------------------------------------------------- @@ -27,12 +28,13 @@ module GHC.Internal.IO.Handle.FD ( import GHC.Internal.Base import GHC.Internal.Show -import GHC.Internal.Control.Exception (try) +import GHC.Internal.Control.Exception (tryWithContext) import GHC.Internal.Data.Maybe import GHC.Internal.Data.Either (either) import GHC.Internal.Data.Typeable import GHC.Internal.Foreign.C.Types import GHC.Internal.MVar +import GHC.Internal.Exception.Type import GHC.Internal.IO import GHC.Internal.IO.Encoding import GHC.Internal.IO.Device as IODevice @@ -116,6 +118,18 @@ addFilePathToIOError :: String -> FilePath -> IOException -> IOException addFilePathToIOError fun fp ioe = ioe{ ioe_location = fun, ioe_filename = Just fp } +-- It could be good in future to use native exception annotation here rather than +-- modifying IOException +catchAndAnnotate :: FilePath -> String -> IO a -> IO a +catchAndAnnotate fp s a = + catchExceptionNoPropagate @IOError a + (\(ExceptionWithContext c e) -> rethrowIO (ExceptionWithContext c (addFilePathToIOError s fp e))) + +-- | Specialised 'rethrowIO' to 'IOError', helpful for type inference +rethrowError :: ExceptionWithContext IOError -> IO a +rethrowError = rethrowIO + + -- | Computation 'openFile' @file mode@ allocates and returns a new, open -- handle to manage the file @file at . It manages input if @mode@ -- is 'ReadMode', output if @mode@ is 'WriteMode' or 'AppendMode', @@ -151,9 +165,9 @@ addFilePathToIOError fun fp ioe -- be using 'openBinaryFile'. openFile :: FilePath -> IOMode -> IO Handle openFile fp im = - catchException + catchAndAnnotate + fp "openFile" (openFile' fp im dEFAULT_OPEN_IN_BINARY_MODE True) - (\e -> ioError (addFilePathToIOError "openFile" fp e)) -- | @'withFile' name mode act@ opens a file like 'openFile' and passes -- the resulting handle to the computation @act at . The handle will be @@ -166,10 +180,9 @@ openFile fp im = withFile :: FilePath -> IOMode -> (Handle -> IO r) -> IO r withFile fp im act = do -- Only annotate when setup or teardown of withFile' raised the exception - catchException - (withFile' fp im dEFAULT_OPEN_IN_BINARY_MODE True (try . act)) - (\e -> ioError (addFilePathToIOError "withFile" fp e)) - >>= either ioError pure + catchAndAnnotate fp "withFile" + (withFile' fp im dEFAULT_OPEN_IN_BINARY_MODE True (tryWithContext . act)) + >>= either rethrowError pure -- | Like 'openFile', but opens the file in ordinary blocking mode. -- This can be useful for opening a FIFO for writing: if we open in @@ -187,9 +200,8 @@ withFile fp im act = do -- @since base-4.4.0.0 openFileBlocking :: FilePath -> IOMode -> IO Handle openFileBlocking fp im = - catchException + catchAndAnnotate fp "openFileBlocking" (openFile' fp im dEFAULT_OPEN_IN_BINARY_MODE False) - (\e -> ioError (addFilePathToIOError "openFileBlocking" fp e)) -- | @'withFileBlocking' name mode act@ opens a file like 'openFileBlocking' -- and passes the resulting handle to the computation @act at . The handle will @@ -202,10 +214,9 @@ openFileBlocking fp im = withFileBlocking :: FilePath -> IOMode -> (Handle -> IO r) -> IO r withFileBlocking fp im act = do -- Only annotate when setup or teardown of withFile' raised the exception - catchException - (withFile' fp im dEFAULT_OPEN_IN_BINARY_MODE False (try . act)) - (\e -> ioError (addFilePathToIOError "withFileBlocking" fp e)) - >>= either ioError pure + catchAndAnnotate fp "withFileBlocking" + (withFile' fp im dEFAULT_OPEN_IN_BINARY_MODE False (tryWithContext . act)) + >>= either rethrowError pure -- | Like 'openFile', but open the file in binary mode. -- On Windows, reading a file in text mode (which is the default) @@ -220,9 +231,8 @@ withFileBlocking fp im act = do -- described in "Control.Exception". openBinaryFile :: FilePath -> IOMode -> IO Handle openBinaryFile fp m = - catchException + catchAndAnnotate fp "openBinaryFile" (openFile' fp m True True) - (\e -> ioError (addFilePathToIOError "openBinaryFile" fp e)) -- | A version of `openBinaryFile` that takes an action to perform -- with the handle. If an exception occurs in the action, then @@ -234,10 +244,9 @@ openBinaryFile fp m = withBinaryFile :: FilePath -> IOMode -> (Handle -> IO r) -> IO r withBinaryFile fp im act = -- Only annotate when setup or teardown of withFile' raised the exception - catchException - (withFile' fp im True True (try . act)) - (\e -> ioError (addFilePathToIOError "withBinaryFile" fp e)) - >>= either ioError pure + catchAndAnnotate fp "withBinaryFile" + (withFile' fp im True True (tryWithContext . act)) + >>= either rethrowError pure -- | Open a file and perform an action with it. If the action throws an -- exception, then the file will be closed. If the last argument is 'True', ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs ===================================== @@ -79,6 +79,7 @@ import GHC.Internal.Real import GHC.Internal.Word import GHC.Internal.Base import GHC.Internal.Exception +import GHC.Internal.Exception.Type import GHC.Internal.Num ( Num(..) ) import GHC.Internal.Show import GHC.Internal.IORef @@ -178,13 +179,13 @@ do_operation :: String -> Handle -> (Handle__ -> IO a) -> MVar Handle__ -> IO a do_operation fun h act m = do h_ <- takeMVar m checkHandleInvariants h_ - act h_ `catchException` handler h_ + act h_ `catchExceptionNoPropagate` handler h_ where - handler h_ e = do + handler h_ (ExceptionWithContext c e) = do putMVar m h_ case () of _ | Just ioe <- fromException e -> - ioError (augmentIOError ioe fun h) + rethrowIO (ExceptionWithContext c $ augmentIOError ioe fun h) _ | Just async_ex <- fromException e -> do -- see Note [async] let _ = async_ex :: SomeAsyncException t <- myThreadId ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Handle/Text.hs ===================================== @@ -41,6 +41,7 @@ import GHC.Internal.IO.Buffer import qualified GHC.Internal.IO.BufferedIO as Buffered import GHC.Internal.IO.Exception import GHC.Internal.Exception +import GHC.Internal.Exception.Type import GHC.Internal.IO.Handle.Types import GHC.Internal.IO.Handle.Internals import qualified GHC.Internal.IO.Device as IODevice @@ -495,7 +496,7 @@ hGetContents' handle = do Left e -> case fromException e of Just ioe -> throwIO (augmentIOError ioe "hGetContents'" handle) - Nothing -> throwIO e + Nothing -> throwIO (NoBacktrace e) strictRead :: Handle -> Handle__ -> IO (Handle__, Either SomeException String) strictRead h handle_ at Handle__{..} = do ===================================== libraries/ghc-internal/src/GHC/Internal/System/IO/Error.hs ===================================== @@ -90,6 +90,7 @@ import GHC.Internal.Data.Either import GHC.Internal.Data.Maybe import GHC.Internal.Base +import GHC.Internal.Exception.Type import GHC.Internal.IO import GHC.Internal.IO.Exception import GHC.Internal.IO.Handle.Types @@ -320,7 +321,7 @@ ioeSetFileName ioe filename = ioe{ ioe_filename = Just filename } -- | Catch any 'IOError' that occurs in the computation and throw a -- modified version. modifyIOError :: (IOError -> IOError) -> IO a -> IO a -modifyIOError f io = catch io (\e -> ioError (f e)) +modifyIOError f io = catch io (\e -> throwIO (NoBacktrace $ f e)) -- ----------------------------------------------------------------------------- -- annotating an IOError View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/52a6fde3eae77cb98be9e2569cf8c2e7669f50fe...3d75a03f1a456987f922d1c0b7770528c812915b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/52a6fde3eae77cb98be9e2569cf8c2e7669f50fe...3d75a03f1a456987f922d1c0b7770528c812915b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 15:20:49 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 11:20:49 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] Exception rethrowing Message-ID: <66f2d8d1ccb58_2eb0996a8b781068e7@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: 41d8b803 by Matthew Pickering at 2024-09-24T16:20:33+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. - - - - - 10 changed files: - libraries/base/src/Control/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Control/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Control/Exception/Base.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs - libraries/ghc-internal/src/GHC/Internal/IO.hs - libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Text.hs - libraries/ghc-internal/src/GHC/Internal/System/IO/Error.hs Changes: ===================================== libraries/base/src/Control/Exception.hs ===================================== @@ -38,6 +38,8 @@ module Control.Exception someExceptionContext, annotateIO, ExceptionWithContext(..), + WhileHandling(..), + -- * Concrete exception types IOException, ArithException(..), @@ -65,6 +67,7 @@ module Control.Exception -- * Throwing exceptions throw, throwIO, + rethrowIO, ioError, throwTo, -- * Catching Exceptions @@ -73,6 +76,7 @@ module Control.Exception -- $catchall -- ** The @catch@ functions catch, + catchNoPropagate, catches, Handler(..), catchJust, @@ -81,6 +85,7 @@ module Control.Exception handleJust, -- ** The @try@ functions try, + tryWithContext, tryJust, -- ** The @evaluate@ function evaluate, @@ -111,11 +116,11 @@ module Control.Exception bracketOnError, finally, onException + ) where import GHC.Internal.Control.Exception import GHC.Internal.Exception.Type -import GHC.Internal.IO (annotateIO) {- $catching ===================================== libraries/ghc-internal/src/GHC/Internal/Control/Exception.hs ===================================== @@ -62,20 +62,25 @@ module GHC.Internal.Control.Exception ( -- * Throwing exceptions throw, throwIO, + rethrowIO, ioError, throwTo, -- ** The @catch@ functions catch, + catchNoPropagate, catches, Handler(..), catchJust, + -- ** Exception annotation + -- ** The @handle@ functions handle, handleJust, -- ** The @try@ functions try, + tryWithContext, tryJust, -- ** The @evaluate@ function @@ -105,6 +110,12 @@ module GHC.Internal.Control.Exception ( finally, onException, + -- * Annotating exceptions + + ExceptionContext(..), + annotateIO, + WhileHandling(..), + ) where import GHC.Internal.Control.Exception.Base ===================================== libraries/ghc-internal/src/GHC/Internal/Control/Exception/Base.hs ===================================== @@ -46,6 +46,7 @@ module GHC.Internal.Control.Exception.Base ( -- * Throwing exceptions throwIO, + rethrowIO, throw, ioError, throwTo, @@ -54,6 +55,7 @@ module GHC.Internal.Control.Exception.Base ( -- ** The @catch@ functions catch, + catchNoPropagate, catchJust, -- ** The @handle@ functions @@ -62,6 +64,7 @@ module GHC.Internal.Control.Exception.Base ( -- ** The @try@ functions try, + tryWithContext, tryJust, onException, @@ -85,6 +88,13 @@ module GHC.Internal.Control.Exception.Base ( assert, + -- * Annotating exceptions + + ExceptionContext(..), + annotateIO, + WhileHandling(..), + + -- * Utilities bracket, @@ -105,6 +115,7 @@ import GHC.Internal.Base import GHC.Internal.Exception import GHC.Internal.IO hiding (bracket, finally, onException) import GHC.Internal.IO.Exception +import GHC.Internal.Exception.Type import GHC.Internal.Show -- import GHC.Internal.Exception hiding ( Exception ) import GHC.Internal.Conc.Sync @@ -132,10 +143,11 @@ catchJust -> IO a -- ^ Computation to run -> (b -> IO a) -- ^ Handler -> IO a -catchJust p a handler = catch a handler' - where handler' e = case p e of - Nothing -> throwIO e - Just b -> handler b +catchJust p a handler = catchNoPropagate a handler' + where handler' ec@(ExceptionWithContext _ e) = + case p e of + Nothing -> rethrowIO ec + Just b -> annotateIO (whileHandling ec) (handler b) -- | A version of 'catch' with the arguments swapped around; useful in -- situations where the code for the handler is shorter. For example: @@ -176,23 +188,23 @@ mapException f v = unsafePerformIO (catch (evaluate v) try :: Exception e => IO a -> IO (Either e a) try a = catch (a >>= \ v -> return (Right v)) (\e -> return (Left e)) +-- | Like 'try' but also returns the exception context, which is useful if you intend +-- to rethrow the exception later. +tryWithContext :: Exception e => IO a -> IO (Either (ExceptionWithContext e) a) +tryWithContext a = catchNoPropagate (a >>= \v -> return (Right v)) (\e -> return (Left e)) + -- | A variant of 'try' that takes an exception predicate to select -- which exceptions are caught (c.f. 'catchJust'). If the exception -- does not match the predicate, it is re-thrown. tryJust :: Exception e => (e -> Maybe b) -> IO a -> IO (Either b a) -tryJust p a = do - r <- try a - case r of - Right v -> return (Right v) - Left e -> case p e of - Nothing -> throwIO e - Just b -> return (Left b) +tryJust p a = catchJust p (Right `fmap` a) (return . Left) -- | Like 'finally', but only performs the final action if there was an -- exception raised by the computation. onException :: IO a -> IO b -> IO a -onException io what = io `catch` \e -> do _ <- what - throwIO (e :: SomeException) +onException io what = io `catchNoPropagate` \e -> do + _ <- what + rethrowIO (e :: ExceptionWithContext SomeException) ----------------------------------------------------------------------------- -- Some Useful Functions ===================================== libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs ===================================== @@ -39,13 +39,16 @@ module GHC.Internal.Exception.Type , emptyExceptionContext , mergeExceptionContext , ExceptionWithContext(..) + -- * Exception propagation + , WhileHandling(..) + , whileHandling -- * Arithmetic exceptions , ArithException(..) , divZeroException, overflowException, ratioZeroDenomException , underflowException ) where -import GHC.Internal.Data.OldList (intersperse) +import GHC.Internal.Data.OldList (intersperse, lines, unlines, null) import GHC.Internal.Data.Maybe import GHC.Internal.Data.Typeable (Typeable, TypeRep, cast) import qualified GHC.Internal.Data.Typeable as Typeable @@ -69,6 +72,27 @@ will be subject to defaulting, as described above. -} type HasExceptionContext = (?exceptionContext :: ExceptionContext) +{- | @WhileHandling@ is used to annotate rethrow exceptions. By inspecting + the @WhileHandling@ annotation, all the places the exception has been rethrow + can be recovered. +-} + +data WhileHandling = WhileHandling SomeException deriving Show + +instance ExceptionAnnotation WhileHandling where + displayExceptionAnnotation (WhileHandling e) = + "While handling " ++ case lines $ displayException e of + [] -> "" + (l1:ls) -> + -- Indent lines forward. + -- displayException may be ill prepared for this?... + unlines $ l1:[if null l then " |" else " | " ++ l | l <- ls] + + +-- | Create 'SomeException' from an explicit context and exception. +whileHandling :: Exception e => ExceptionWithContext e -> WhileHandling +whileHandling e = WhileHandling (toException e) + {- | The @SomeException@ type is the root of the exception type hierarchy. When an exception of type @e@ is thrown, behind the scenes it is ===================================== libraries/ghc-internal/src/GHC/Internal/IO.hs ===================================== @@ -40,7 +40,8 @@ module GHC.Internal.IO ( FilePath, - catch, catchException, catchAny, throwIO, + catch, catchNoPropagate, catchException, catchExceptionNoPropagate, catchAny, throwIO, + rethrowIO, mask, mask_, uninterruptibleMask, uninterruptibleMask_, MaskingState(..), getMaskingState, unsafeUnmask, interruptible, @@ -51,7 +52,7 @@ module GHC.Internal.IO ( import GHC.Internal.Base import GHC.Internal.ST import GHC.Internal.Exception -import GHC.Internal.Exception.Type (NoBacktrace(..)) +import GHC.Internal.Exception.Type (NoBacktrace(..), WhileHandling(..), HasExceptionContext, ExceptionWithContext(..)) import GHC.Internal.Show import GHC.Internal.IO.Unsafe import GHC.Internal.Unsafe.Coerce ( unsafeCoerce ) @@ -152,6 +153,10 @@ have to work around that in the definition of catch below). catchException :: Exception e => IO a -> (e -> IO a) -> IO a catchException !io handler = catch io handler +-- | A variant of 'catchException' which does not annotate the handler with 'WhileHandling' +catchExceptionNoPropagate :: Exception e => IO a -> (ExceptionWithContext e -> IO a) -> IO a +catchExceptionNoPropagate !io handler = catchNoPropagate io handler + -- | This is the simplest of the exception-catching functions. It -- takes a single argument, runs it, and if an exception is raised -- the \"handler\" is executed, with the value of the exception passed as an @@ -191,6 +196,23 @@ catch :: Exception e -> IO a -- See #exceptions_and_strictness#. catch (IO io) handler = IO $ catch# io handler' + where + handler' e = + case fromException e of + Just e' -> unIO (annotateIO (WhileHandling e) (handler e')) + Nothing -> raiseIO# e + +-- | A variant of 'catch' which doesn't annotate the handler with the exception +-- which was caught. This function should be used when you are implementing your own +-- error handling functions which may rethrow the exceptions. +-- +-- In the case where you rethrow an exception without modifying it, you should +-- rethrow the exception with the old exception context. +catchNoPropagate :: Exception e + => IO a + -> (ExceptionWithContext e -> IO a) + -> IO a +catchNoPropagate (IO io) handler = IO $ catch# io handler' where handler' e = case fromException e of @@ -202,10 +224,12 @@ catch (IO io) handler = IO $ catch# io handler' -- Note that this function is /strict/ in the action. That is, -- @catchAny undefined b == _|_ at . See #exceptions_and_strictness# for -- details. -catchAny :: IO a -> (forall e . Exception e => e -> IO a) -> IO a +-- +-- If you rethrow an exception, you should reuse the supplied ExceptionContext. +catchAny :: IO a -> (forall e . (HasExceptionContext, Exception e) => e -> IO a) -> IO a catchAny !(IO io) handler = IO $ catch# io handler' where - handler' (SomeException e) = unIO (handler e) + handler' se@(SomeException e) = unIO (annotateIO (WhileHandling se) (handler e)) -- | Execute an 'IO' action, adding the given 'ExceptionContext' -- to any thrown synchronous exceptions. @@ -260,6 +284,11 @@ throwIO e = do se <- toExceptionWithBacktrace e IO (raiseIO# se) +-- | A utility to use when rethrowing exceptions, no new backtrace will be attached +-- when rethrowing an exception but you must supply the existing context. +rethrowIO :: Exception e => ExceptionWithContext e -> IO a +rethrowIO e = throwIO (NoBacktrace e) + -- ----------------------------------------------------------------------------- -- Controlling asynchronous exception delivery @@ -332,9 +361,9 @@ getMaskingState = IO $ \s -> _ -> MaskedInterruptible #) onException :: IO a -> IO b -> IO a -onException io what = io `catchException` \e -> do +onException io what = io `catchExceptionNoPropagate` \e -> do _ <- what - throwIO $ NoBacktrace (e :: SomeException) + rethrowIO (e :: ExceptionWithContext SomeException) -- | Executes an IO computation with asynchronous -- exceptions /masked/. That is, any thread which attempts to raise ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs ===================================== @@ -315,7 +315,7 @@ ioException :: HasCallStack => IOException -> IO a ioException err = throwIO err -- | Raise an 'IOError' in the 'IO' monad. -ioError :: IOError -> IO a +ioError :: HasCallStack => IOError -> IO a ioError = ioException -- --------------------------------------------------------------------------- ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs ===================================== @@ -1,4 +1,5 @@ {-# LANGUAGE Trustworthy #-} +{-# LANGUAGE TypeApplications #-} {-# LANGUAGE CPP, NoImplicitPrelude #-} ----------------------------------------------------------------------------- @@ -27,12 +28,13 @@ module GHC.Internal.IO.Handle.FD ( import GHC.Internal.Base import GHC.Internal.Show -import GHC.Internal.Control.Exception (try) +import GHC.Internal.Control.Exception (tryWithContext) import GHC.Internal.Data.Maybe import GHC.Internal.Data.Either (either) import GHC.Internal.Data.Typeable import GHC.Internal.Foreign.C.Types import GHC.Internal.MVar +import GHC.Internal.Exception.Type import GHC.Internal.IO import GHC.Internal.IO.Encoding import GHC.Internal.IO.Device as IODevice @@ -116,6 +118,18 @@ addFilePathToIOError :: String -> FilePath -> IOException -> IOException addFilePathToIOError fun fp ioe = ioe{ ioe_location = fun, ioe_filename = Just fp } +-- It could be good in future to use native exception annotation here rather than +-- modifying IOException +catchAndAnnotate :: FilePath -> String -> IO a -> IO a +catchAndAnnotate fp s a = + catchExceptionNoPropagate @IOError a + (\(ExceptionWithContext c e) -> rethrowIO (ExceptionWithContext c (addFilePathToIOError s fp e))) + +-- | Specialised 'rethrowIO' to 'IOError', helpful for type inference +rethrowError :: ExceptionWithContext IOError -> IO a +rethrowError = rethrowIO + + -- | Computation 'openFile' @file mode@ allocates and returns a new, open -- handle to manage the file @file at . It manages input if @mode@ -- is 'ReadMode', output if @mode@ is 'WriteMode' or 'AppendMode', @@ -151,9 +165,9 @@ addFilePathToIOError fun fp ioe -- be using 'openBinaryFile'. openFile :: FilePath -> IOMode -> IO Handle openFile fp im = - catchException + catchAndAnnotate + fp "openFile" (openFile' fp im dEFAULT_OPEN_IN_BINARY_MODE True) - (\e -> ioError (addFilePathToIOError "openFile" fp e)) -- | @'withFile' name mode act@ opens a file like 'openFile' and passes -- the resulting handle to the computation @act at . The handle will be @@ -166,10 +180,9 @@ openFile fp im = withFile :: FilePath -> IOMode -> (Handle -> IO r) -> IO r withFile fp im act = do -- Only annotate when setup or teardown of withFile' raised the exception - catchException - (withFile' fp im dEFAULT_OPEN_IN_BINARY_MODE True (try . act)) - (\e -> ioError (addFilePathToIOError "withFile" fp e)) - >>= either ioError pure + catchAndAnnotate fp "withFile" + (withFile' fp im dEFAULT_OPEN_IN_BINARY_MODE True (tryWithContext . act)) + >>= either rethrowError pure -- | Like 'openFile', but opens the file in ordinary blocking mode. -- This can be useful for opening a FIFO for writing: if we open in @@ -187,9 +200,8 @@ withFile fp im act = do -- @since base-4.4.0.0 openFileBlocking :: FilePath -> IOMode -> IO Handle openFileBlocking fp im = - catchException + catchAndAnnotate fp "openFileBlocking" (openFile' fp im dEFAULT_OPEN_IN_BINARY_MODE False) - (\e -> ioError (addFilePathToIOError "openFileBlocking" fp e)) -- | @'withFileBlocking' name mode act@ opens a file like 'openFileBlocking' -- and passes the resulting handle to the computation @act at . The handle will @@ -202,10 +214,9 @@ openFileBlocking fp im = withFileBlocking :: FilePath -> IOMode -> (Handle -> IO r) -> IO r withFileBlocking fp im act = do -- Only annotate when setup or teardown of withFile' raised the exception - catchException - (withFile' fp im dEFAULT_OPEN_IN_BINARY_MODE False (try . act)) - (\e -> ioError (addFilePathToIOError "withFileBlocking" fp e)) - >>= either ioError pure + catchAndAnnotate fp "withFileBlocking" + (withFile' fp im dEFAULT_OPEN_IN_BINARY_MODE False (tryWithContext . act)) + >>= either rethrowError pure -- | Like 'openFile', but open the file in binary mode. -- On Windows, reading a file in text mode (which is the default) @@ -220,9 +231,8 @@ withFileBlocking fp im act = do -- described in "Control.Exception". openBinaryFile :: FilePath -> IOMode -> IO Handle openBinaryFile fp m = - catchException + catchAndAnnotate fp "openBinaryFile" (openFile' fp m True True) - (\e -> ioError (addFilePathToIOError "openBinaryFile" fp e)) -- | A version of `openBinaryFile` that takes an action to perform -- with the handle. If an exception occurs in the action, then @@ -234,10 +244,9 @@ openBinaryFile fp m = withBinaryFile :: FilePath -> IOMode -> (Handle -> IO r) -> IO r withBinaryFile fp im act = -- Only annotate when setup or teardown of withFile' raised the exception - catchException - (withFile' fp im True True (try . act)) - (\e -> ioError (addFilePathToIOError "withBinaryFile" fp e)) - >>= either ioError pure + catchAndAnnotate fp "withBinaryFile" + (withFile' fp im True True (tryWithContext . act)) + >>= either rethrowError pure -- | Open a file and perform an action with it. If the action throws an -- exception, then the file will be closed. If the last argument is 'True', ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs ===================================== @@ -79,6 +79,7 @@ import GHC.Internal.Real import GHC.Internal.Word import GHC.Internal.Base import GHC.Internal.Exception +import GHC.Internal.Exception.Type import GHC.Internal.Num ( Num(..) ) import GHC.Internal.Show import GHC.Internal.IORef @@ -178,13 +179,13 @@ do_operation :: String -> Handle -> (Handle__ -> IO a) -> MVar Handle__ -> IO a do_operation fun h act m = do h_ <- takeMVar m checkHandleInvariants h_ - act h_ `catchException` handler h_ + act h_ `catchExceptionNoPropagate` handler h_ where - handler h_ e = do + handler h_ (ExceptionWithContext c e) = do putMVar m h_ case () of _ | Just ioe <- fromException e -> - ioError (augmentIOError ioe fun h) + rethrowIO (ExceptionWithContext c $ augmentIOError ioe fun h) _ | Just async_ex <- fromException e -> do -- see Note [async] let _ = async_ex :: SomeAsyncException t <- myThreadId ===================================== libraries/ghc-internal/src/GHC/Internal/IO/Handle/Text.hs ===================================== @@ -41,6 +41,7 @@ import GHC.Internal.IO.Buffer import qualified GHC.Internal.IO.BufferedIO as Buffered import GHC.Internal.IO.Exception import GHC.Internal.Exception +import GHC.Internal.Exception.Type import GHC.Internal.IO.Handle.Types import GHC.Internal.IO.Handle.Internals import qualified GHC.Internal.IO.Device as IODevice @@ -495,7 +496,7 @@ hGetContents' handle = do Left e -> case fromException e of Just ioe -> throwIO (augmentIOError ioe "hGetContents'" handle) - Nothing -> throwIO e + Nothing -> throwIO (NoBacktrace e) strictRead :: Handle -> Handle__ -> IO (Handle__, Either SomeException String) strictRead h handle_ at Handle__{..} = do ===================================== libraries/ghc-internal/src/GHC/Internal/System/IO/Error.hs ===================================== @@ -90,6 +90,7 @@ import GHC.Internal.Data.Either import GHC.Internal.Data.Maybe import GHC.Internal.Base +import GHC.Internal.Exception.Type import GHC.Internal.IO import GHC.Internal.IO.Exception import GHC.Internal.IO.Handle.Types @@ -320,7 +321,7 @@ ioeSetFileName ioe filename = ioe{ ioe_filename = Just filename } -- | Catch any 'IOError' that occurs in the computation and throw a -- modified version. modifyIOError :: (IOError -> IOError) -> IO a -> IO a -modifyIOError f io = catch io (\e -> ioError (f e)) +modifyIOError f io = catch io (\e -> throwIO (NoBacktrace $ f e)) -- ----------------------------------------------------------------------------- -- annotating an IOError View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/41d8b8034624aecea1d38213df165cb3f4b874d0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/41d8b8034624aecea1d38213df165cb3f4b874d0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 16:16:12 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Tue, 24 Sep 2024 12:16:12 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] Exception rethrowing Message-ID: <66f2e5cc2352c_7522b1282ac47486@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: 723e4471 by Matthew Pickering at 2024-09-24T17:16:04+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. Implements CLC proposal#202 <https://github.com/haskell/core-libraries-committee/issues/202> - - - - - 30 changed files: - docs/users_guide/9.12.1-notes.rst - libraries/base/changelog.md - libraries/base/src/Control/Exception.hs - libraries/base/tests/IO/T4808.stderr - libraries/base/tests/IO/mkdirExists.stderr - libraries/base/tests/IO/openFile002.stderr - libraries/base/tests/IO/withBinaryFile001.stderr - libraries/base/tests/IO/withBinaryFile002.stderr - libraries/base/tests/IO/withFile001.stderr - libraries/base/tests/IO/withFile002.stderr - libraries/base/tests/IO/withFileBlocking001.stderr - libraries/base/tests/IO/withFileBlocking002.stderr - libraries/base/tests/T15349.stderr - libraries/ghc-internal/src/GHC/Internal/Control/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Control/Exception/Base.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs - libraries/ghc-internal/src/GHC/Internal/IO.hs - libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Text.hs - libraries/ghc-internal/src/GHC/Internal/System/IO/Error.hs - testsuite/tests/codeGen/should_run/cgrun016.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/ffi/should_run/T7170.stderr - testsuite/tests/ghc-e/should_fail/T18441fail2.stderr - testsuite/tests/ghc-e/should_fail/T18441fail7.stderr - testsuite/tests/ghc-e/should_fail/T18441fail8.stderr - testsuite/tests/ghc-e/should_fail/T23663.stderr - testsuite/tests/ghc-e/should_fail/T9930fail.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/723e4471c356de06271f48bfb777db141a7ae675 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/723e4471c356de06271f48bfb777db141a7ae675 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 18:54:26 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Tue, 24 Sep 2024 14:54:26 -0400 Subject: [Git][ghc/ghc][wip/enable-late-ccs-release] 78 commits: simplifier: Fix space leak during demand analysis Message-ID: <66f30ae2e2890_1d89831561ac85661@gitlab.mail> Matthew Pickering pushed to branch wip/enable-late-ccs-release at Glasgow Haskell Compiler / GHC Commits: 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - eca31282 by Matthew Pickering at 2024-09-24T19:52:51+01:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - f1079022 by Matthew Pickering at 2024-09-24T19:52:51+01:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - ea8b2e7d by Matthew Pickering at 2024-09-24T19:53:02+01:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - 22 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Dwarf/Constants.hs - compiler/GHC/CmmToAsm/PIC.hs - + compiler/GHC/CmmToAsm/RV64.hs - + compiler/GHC/CmmToAsm/RV64/CodeGen.hs - + compiler/GHC/CmmToAsm/RV64/Cond.hs - + compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5c6fda94a7b801f05f2af6045996e980dc582e7b...ea8b2e7db800a80e2e1da30b4f7445dbd9b5b71d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5c6fda94a7b801f05f2af6045996e980dc582e7b...ea8b2e7db800a80e2e1da30b4f7445dbd9b5b71d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 20:18:12 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Tue, 24 Sep 2024 16:18:12 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8] 4 commits: rts: ensure gc_thread/gen_workspace is allocated with proper alignment Message-ID: <66f31e84b7726_2a9a86bc624-1b1@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8 at Glasgow Haskell Compiler / GHC Commits: 05b7544d by Cheng Shao at 2024-09-23T15:30:53-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - de722d34 by Cheng Shao at 2024-09-23T15:32:08-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - d93fd1bd by Cheng Shao at 2024-09-23T15:32:12-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 2050caf9 by Cheng Shao at 2024-09-23T15:33:15-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - 4 changed files: - rts/include/Stg.h - rts/sm/GC.c - rts/sm/NonMovingMark.c - rts/sm/Sanity.c Changes: ===================================== rts/include/Stg.h ===================================== @@ -108,7 +108,7 @@ /* Compute offsets of struct fields */ -#define STG_FIELD_OFFSET(s_type, field) ((StgWord)&(((s_type*)0)->field)) +#define STG_FIELD_OFFSET(s_type, field) __builtin_offsetof(s_type, field) /* * 'Portable' inlining: ===================================== rts/sm/GC.c ===================================== @@ -55,6 +55,7 @@ #include "NonMoving.h" #include "Ticky.h" +#include #include // for memset() #include @@ -1240,8 +1241,9 @@ initGcThreads (uint32_t from USED_IF_THREADS, uint32_t to USED_IF_THREADS) for (i = from; i < to; i++) { gc_threads[i] = - stgMallocBytes(sizeof(gc_thread) + + stgMallocAlignedBytes(sizeof(gc_thread) + RtsFlags.GcFlags.generations * sizeof(gen_workspace), + alignof(gc_thread), "alloc_gc_threads"); new_gc_thread(i, gc_threads[i]); @@ -1266,7 +1268,7 @@ freeGcThreads (void) { freeWSDeque(gc_threads[i]->gens[g].todo_q); } - stgFree (gc_threads[i]); + stgFreeAligned (gc_threads[i]); } closeCondition(&gc_running_cv); closeMutex(&gc_running_mutex); ===================================== rts/sm/NonMovingMark.c ===================================== @@ -918,7 +918,7 @@ static MarkQueueEnt markQueuePop (MarkQueue *q) // The entry may not be a MARK_CLOSURE but it doesn't matter, our // MarkQueueEnt encoding always places the pointer to the object to be // marked first. - prefetchForRead(&new.mark_closure.p->header.info); + prefetchForRead(&(UNTAG_CLOSURE(new.mark_closure.p)->header.info)); prefetchForRead(Bdescr((StgPtr) new.mark_closure.p)); q->prefetch_queue[i] = new; i = (i + 1) % MARK_PREFETCH_QUEUE_DEPTH; ===================================== rts/sm/Sanity.c ===================================== @@ -359,7 +359,8 @@ checkClosure( const StgClosure* p ) load_load_barrier(); if (IS_FORWARDING_PTR(info)) { - barf("checkClosure: found EVACUATED closure %d", info->type); + ASSERT(LOOKS_LIKE_CLOSURE_PTR(info)); + barf("checkClosure: found EVACUATED closure %u", GET_INFO((StgClosure*)UN_FORWARDING_PTR(info))->type); } #if defined(PROFILING) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/52e82b4d710ad7987cae219eab1dec113a457b53...2050caf9ebc81c4fcba05bd394315e739ac18d36 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/52e82b4d710ad7987cae219eab1dec113a457b53...2050caf9ebc81c4fcba05bd394315e739ac18d36 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Tue Sep 24 20:23:33 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Tue, 24 Sep 2024 16:23:33 -0400 Subject: [Git][ghc/ghc][wip/T25066] 2 commits: base: Fix #25066 Message-ID: <66f31fc54165d_2a9a86170a3416c@gitlab.mail> Ben Gamari pushed to branch wip/T25066 at Glasgow Haskell Compiler / GHC Commits: 10a1a6df by Ben Gamari at 2024-09-24T12:43:53-04:00 base: Fix #25066 As noted in #25066, the exception backtrace proposal introduced a rather subtle performance regression due to simplification producing Core which the demand analyser concludes may diverge with a precise exception. The nature of the problem is more completely described in the new Note [Hiding precise exception signature in throw]. The (rather hacky) solution we use here hides the problematic optimisation through judicious use of `noinline`. Ultimately however we will want a more principled solution (e.g. #23847). - - - - - 208d8a8c by Ben Gamari at 2024-09-24T12:43:53-04:00 base: Improve documentation of Control.Exception.Backtrace - - - - - 4 changed files: - libraries/base/src/Control/Exception/Backtrace.hs - libraries/base/tests/T25066.stderr - libraries/ghc-internal/src/GHC/Internal/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Backtrace.hs Changes: ===================================== libraries/base/src/Control/Exception/Backtrace.hs ===================================== @@ -7,8 +7,43 @@ -- Stability : internal -- Portability : non-portable (GHC Extensions) -- --- Mechanisms for collecting diagnostic backtraces and their representation. +-- This module provides the 'Backtrace'\ s type, which provides a +-- common representation for backtrace information which can be, e.g., attached +-- to exceptions (via the 'Control.Exception.Context.ExceptionContext' facility). +-- These backtraces preserve useful context about the execution state of the program +-- using a variety of means; we call these means *backtrace mechanisms*. -- +-- We currently support four backtrace mechanisms: +-- +-- - 'CostCentreBacktrace' captures the current cost-centre stack +-- using 'GHC.Stack.CCS.getCurrentCCS'. +-- - 'HasCallStackBacktrace' captures the 'HasCallStack' 'CallStack'. +-- - 'ExecutionBacktrace' captures the execution stack, unwound and resolved +-- to symbols via DWARF debug information. +-- - 'IPEBacktrace' captures the execution stack, resolved to names via info-table +-- provenance information. +-- +-- Each of these are useful in different situations. While 'CostCentreBacktrace's are +-- readily mapped back to the source program, they require that the program be instrumented +-- with cost-centres, incurring runtime cost. Similarly, 'HasCallStackBacktrace's require that +-- the program be manually annotated with 'HasCallStack' constraints. +-- +-- By contrast, 'IPEBacktrace's incur no runtime instrumentation but require that (at least +-- some subset of) the program be built with GHC\'s @-finfo-table-map@ flag. Moreover, because +-- info-table provenance information is derived after optimisation, it may be harder to relate +-- back to the structure of the source program. +-- +-- 'ExecutionBacktrace's are similar to 'IPEBacktrace's but use DWARF stack unwinding +-- and symbol resolution; this allows for useful backtraces even in the presence +-- of foreign calls, both into and out of Haskell. However, for robust stack unwinding +-- the entirety of the program (and its dependencies, both Haskell and native) must +-- be compiled with debugging information (e.g. using GHC\'s @-g@ flag). + + +-- Note [Backtrace mechanisms] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- See module docstring above. + module Control.Exception.Backtrace ( -- * Backtrace mechanisms ===================================== libraries/base/tests/T25066.stderr ===================================== @@ -5,7 +5,7 @@ T25066.$fShowMyException: T25066.$tc'MyException: T25066.$tcMyException: T25066.$trModule: -T25066.g: x +T25066.g: b @@ -15,6 +15,6 @@ T25066.$fShowMyException: T25066.$tc'MyException: T25066.$tcMyException: T25066.$trModule: -T25066.g: x +T25066.g: b ===================================== libraries/ghc-internal/src/GHC/Internal/Exception.hs ===================================== @@ -81,9 +81,86 @@ import GHC.Internal.Exception.Type throw :: forall (r :: RuntimeRep). forall (a :: TYPE r). forall e. (HasCallStack, Exception e) => e -> a throw e = - let !se = unsafePerformIO (toExceptionWithBacktrace e) + -- Note the absolutely crucial bang "!" on this binding! + -- See Note [Capturing the backtrace in throw] + -- Note also the absolutely crucial `noinine` in the RHS! + -- See Note [Hiding precise exception signature in throw] + let se :: SomeException + !se = noinline (unsafePerformIO (toExceptionWithBacktrace e)) in raise# se +-- Note [Capturing the backtrace in throw] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- When `throw` captures a backtrace, it must be the backtrace *at the moment +-- that `throw` is called*. That is why the binding of `se` is marked strict, +-- via the `!`: +-- +-- !se = +-- +-- GHC can capture /four/ different sorts of backtraces (See Note [Backtrace +-- mechanisms] in "Control.Exception.Backtrace" for details). One of them +-- (`CallStack` constraints) does not need this strict-binding treatment, +-- because the `CallStack` constraint is captured in the thunk. However, the +-- other two (DWARF stack unwinding, and native Haskell stack unwinding) are +-- much more fragile, and can only be captured right at the call of `throw`. +-- +-- However, making `se` strict has downsides: see +-- Note [Hiding precise exception signature in throw] below. +-- +-- +-- Note [Hiding precise exception signature in throw] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- In 'throw' we use `unsafePerformIO . toExceptionWithBacktrace' to collect +-- the backtraces which will be attached as the exception's 'ExceptionContext'. +-- We must ensure that this is evaluated immediately in `throw` since +-- `toExceptionWithBacktrace` must capture the execution state at the moment +-- that the exception is thrown (see Note [Capturing the backtrace in throw]). +-- Unfortunately, unless we take particular care this can lead to a +-- catastrophic regression in 'throw's demand signature which will infect +-- all callers (#25066) +-- +-- Specifically, GHC's demand analysis has an approximate heuristic for tracking +-- whether divergent functions diverge with precise or imprecise exceptions (namely +-- the 'ExnOrDiv' and 'Diverges' constructors of 'GHC.Types.Demand.Divergence', +-- respectively). This is because we can take considerably more liberties in +-- optimising around functions which are known not to diverge via precise +-- exception (see Note [Precise exceptions and strictness analysis]). +-- For this reason, it is important that 'throw' have a 'Diverges' divergence +-- type. +-- +-- Unfortunately, this is broken if we allow `unsafePerformIO` to inline. Specifically, +-- if we allow this inlining we will end up with Core of the form: +-- +-- throw = \e -> +-- case runRW# (\s -> ... toExceptionWithBacktrace e s ...) of +-- se -> raise# se +-- +-- so far this is fine; the demand analyzer's divergence heuristic +-- will give 'throw' the expected 'Diverges' divergence. +-- +-- However, the simplifier will subsequently notice that `raise#` can be fruitfully +-- floated into the body of the `runRW#`: +-- +-- throw = \e -> +-- runRW# (\s -> case toExceptionWithBacktrace e s of +-- (# s', se #) -> raise# se) +-- +-- This is problematic as one of the demand analyser's heuristics examines +-- `case` scrutinees, looking for those that result in a `RealWorld#` token +-- (see Note [Which scrutinees may throw precise exceptions], test (1)). The +-- `case toExceptionWithBacktrace e of ...` here fails this check, causing the +-- heuristic to conclude that `throw` may indeed diverge with a precise +-- exception. This resulted in the significant performance regression noted in +-- #25066. +-- +-- To avoid this, we use `noinline` to ensure that `unsafePerformIO` does not unfold, +-- meaning that the `raise#` cannot be floated under the `toExceptionWithBacktrace` +-- case analysis. +-- +-- Ultimately this is a bit of a horrible hack; the right solution would be to have +-- primops which allow more precise guidance of the demand analyser's heuristic +-- (e.g. #23847). + -- | @since base-4.20.0.0 toExceptionWithBacktrace :: (HasCallStack, Exception e) => e -> IO SomeException ===================================== libraries/ghc-internal/src/GHC/Internal/Exception/Backtrace.hs ===================================== @@ -33,7 +33,7 @@ data BacktraceMechanism = CostCentreBacktrace -- | collect 'HasCallStack' backtraces | HasCallStackBacktrace - -- | collect backtraces from native execution stack unwinding + -- | collect backtraces via native execution stack unwinding (e.g. using DWARF debug information) | ExecutionBacktrace -- | collect backtraces from Info Table Provenance Entries | IPEBacktrace View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8945337e0d4c400040c3257cfc4a70889c7b05a5...208d8a8c192a55a3884c4f210da20a1c0cea3448 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8945337e0d4c400040c3257cfc4a70889c7b05a5...208d8a8c192a55a3884c4f210da20a1c0cea3448 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 04:34:58 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Wed, 25 Sep 2024 00:34:58 -0400 Subject: [Git][ghc/ghc][wip/interpolated-strings] 5 commits: Merge branch 'wip/strings' into HEAD Message-ID: <66f392f2f92f_152f2f64c56c450ce@gitlab.mail> Brandon Chinn pushed to branch wip/interpolated-strings at Glasgow Haskell Compiler / GHC Commits: 0b20a4a8 by Brandon Chinn at 2024-09-24T21:04:45-07:00 Merge branch 'wip/strings' into HEAD - - - - - f46ed78e by Brandon Chinn at 2024-09-24T21:04:45-07:00 Unify ITstring + ITstringMulti - - - - - 34df2e36 by Brandon Chinn at 2024-09-24T21:04:46-07:00 Unify HsString + HsMultilineString - - - - - 5851b8bf by Brandon Chinn at 2024-09-24T21:07:12-07:00 Move multiline string processing functions to top-level - - - - - cc7d07b3 by Brandon Chinn at 2024-09-24T21:34:40-07:00 Implement interpolated strings - - - - - 15 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Lit.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/Language/Haskell/Syntax/Expr.hs - compiler/Language/Haskell/Syntax/Extension.hs - compiler/Language/Haskell/Syntax/Lit.hs - utils/check-exact/ExactPrint.hs - utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -242,6 +242,9 @@ type instance XIPVar GhcRn = NoExtField type instance XIPVar GhcTc = DataConCantHappen type instance XOverLitE (GhcPass _) = NoExtField type instance XLitE (GhcPass _) = NoExtField +type instance XInterString (GhcPass _) = NoExtField +type instance XInterStringRaw (GhcPass _) = NoExtField +type instance XInterStringExp (GhcPass _) = NoExtField type instance XLam (GhcPass _) = [AddEpAnn] type instance XApp (GhcPass _) = NoExtField ===================================== compiler/GHC/Hs/Lit.hs ===================================== @@ -49,7 +49,6 @@ import Language.Haskell.Syntax.Lit type instance XHsChar (GhcPass _) = SourceText type instance XHsCharPrim (GhcPass _) = SourceText type instance XHsString (GhcPass _) = SourceText -type instance XHsMultilineString (GhcPass _) = SourceText type instance XHsStringPrim (GhcPass _) = SourceText type instance XHsInt (GhcPass _) = NoExtField type instance XHsIntPrim (GhcPass _) = SourceText @@ -136,7 +135,6 @@ hsLitNeedsParens p = go go (HsChar {}) = False go (HsCharPrim {}) = False go (HsString {}) = False - go (HsMultilineString {}) = False go (HsStringPrim {}) = False go (HsInt _ x) = p > topPrec && il_neg x go (HsInteger _ x _) = p > topPrec && x < 0 @@ -160,7 +158,6 @@ convertLit :: HsLit (GhcPass p1) -> HsLit (GhcPass p2) convertLit (HsChar a x) = HsChar a x convertLit (HsCharPrim a x) = HsCharPrim a x convertLit (HsString a x) = HsString a x -convertLit (HsMultilineString a x) = HsMultilineString a x convertLit (HsStringPrim a x) = HsStringPrim a x convertLit (HsInt a x) = HsInt a x convertLit (HsIntPrim a x) = HsIntPrim a x @@ -197,11 +194,11 @@ Equivalently it's True if instance Outputable (HsLit (GhcPass p)) where ppr (HsChar st c) = pprWithSourceText st (pprHsChar c) ppr (HsCharPrim st c) = pprWithSourceText st (pprPrimChar c) - ppr (HsString st s) = pprWithSourceText st (pprHsString s) - ppr (HsMultilineString st s) = - case st of - NoSourceText -> pprHsString s - SourceText src -> vcat $ map text $ split '\n' (unpackFS src) + ppr (HsString st ty s) = + case (ty, st) of + (HsStringTypeSingle, _) -> pprWithSourceText st (pprHsString s) + (HsStringTypeMulti, NoSourceText) -> pprHsString s + (HsStringTypeMulti, SourceText src) -> vcat $ map text $ split '\n' (unpackFS src) ppr (HsStringPrim st s) = pprWithSourceText st (pprHsBytes s) ppr (HsInt _ i) = pprWithSourceText (il_text i) (integer (il_value i)) @@ -241,7 +238,6 @@ pmPprHsLit :: HsLit (GhcPass x) -> SDoc pmPprHsLit (HsChar _ c) = pprHsChar c pmPprHsLit (HsCharPrim _ c) = pprHsChar c pmPprHsLit (HsString st s) = pprWithSourceText st (pprHsString s) -pmPprHsLit (HsMultilineString st s) = pprWithSourceText st (pprHsString s) pmPprHsLit (HsStringPrim _ s) = pprHsBytes s pmPprHsLit (HsInt _ i) = integer (il_value i) pmPprHsLit (HsIntPrim _ i) = integer i ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -75,8 +75,7 @@ hsPatType (SplicePat v _) = dataConCantHappen v hsLitType :: HsLit (GhcPass p) -> Type hsLitType (HsChar _ _) = charTy hsLitType (HsCharPrim _ _) = charPrimTy -hsLitType (HsString _ _) = stringTy -hsLitType (HsMultilineString _ _) = stringTy +hsLitType (HsString _ _ _) = stringTy hsLitType (HsStringPrim _ _) = addrPrimTy hsLitType (HsInt _ _) = intTy hsLitType (HsIntPrim _ _) = intPrimTy ===================================== compiler/GHC/HsToCore/Match/Literal.hs ===================================== @@ -121,7 +121,6 @@ dsLit l = do HsDoublePrim _ fl -> return (Lit (LitDouble (rationalFromFractionalLit fl))) HsChar _ c -> return (mkCharExpr c) HsString _ str -> mkStringExprFS str - HsMultilineString _ str -> mkStringExprFS str HsInteger _ i _ -> return (mkIntegerExpr platform i) HsInt _ i -> return (mkIntExpr platform (il_value i)) HsRat _ fl ty -> dsFractionalLitToRational fl ty @@ -475,7 +474,6 @@ getSimpleIntegralLit (HsInteger _ i ty) = Just (i, ty) getSimpleIntegralLit HsChar{} = Nothing getSimpleIntegralLit HsCharPrim{} = Nothing getSimpleIntegralLit HsString{} = Nothing -getSimpleIntegralLit HsMultilineString{} = Nothing getSimpleIntegralLit HsStringPrim{} = Nothing getSimpleIntegralLit HsRat{} = Nothing getSimpleIntegralLit HsFloatPrim{} = Nothing ===================================== compiler/GHC/HsToCore/Quote.hs ===================================== @@ -3025,8 +3025,7 @@ repLiteral lit HsDoublePrim _ _ -> Just doublePrimLName HsChar _ _ -> Just charLName HsCharPrim _ _ -> Just charPrimLName - HsString _ _ -> Just stringLName - HsMultilineString _ _ -> Just stringLName + HsString _ _ _ -> Just stringLName HsRat _ _ _ -> Just rationalLName _ -> Nothing ===================================== compiler/GHC/Parser.y ===================================== @@ -727,8 +727,13 @@ are the most common patterns, rewritten as regular expressions for clarity: LABELVARID { L _ (ITlabelvarid _ _) } CHAR { L _ (ITchar _ _) } - STRING { L _ (ITstring _ _) } - STRING_MULTI { L _ (ITstringMulti _ _) } + STRING { L _ (ITstring _ _ StringTypeSingle) } + STRING_MULTI { L _ (ITstring _ _ StringTypeMulti) } + STRING_INTER_BEGIN { L _ (ITstringInterBegin _) } + STRING_INTER_RAW { L _ (ITstringInterRaw _ _) } + STRING_INTER_EXP_OPEN { L _ ITstringInterExpOpen } + STRING_INTER_EXP_CLOSE { L _ ITstringInterExpClose } + STRING_INTER_END { L _ (ITstringInterEnd _) } INTEGER { L _ (ITinteger _) } RATIONAL { L _ (ITrational _) } @@ -3080,6 +3085,8 @@ aexp2 :: { ECP } -- into HsOverLit when -XOverloadedStrings is on. -- | STRING { sL (getLoc $1) (HsOverLit $! mkHsIsString (getSTRINGs $1) -- (getSTRING $1) noExtField) } + | stringInter {% fmap ecpFromExp + (ams1 $1 (HsInterString NoExtField $! unLoc $1)) } | INTEGER { ECP $ mkHsOverLitPV (sL1a $1 $ mkHsIntegral (getINTEGER $1)) } | RATIONAL { ECP $ mkHsOverLitPV (sL1a $1 $ mkHsFractional (getRATIONAL $1)) } @@ -3677,6 +3684,14 @@ ipvar :: { Located HsIPName } overloaded_label :: { Located (SourceText, FastString) } : LABELVARID { sL1 $1 (getLABELVARIDs $1, getLABELVARID $1) } +----------------------------------------------------------------------------- +-- Interpolated strings +-- See Note [Interpolated strings] in GHC.Parser.String + +stringInter :: { [Either FastString (LHsExpr GhcPs)] } + -- TODO(bchinn): break out recursive stringInterPart rule + : STRING_INTER_BEGIN (STRING_INTER_RAW | STRING_INTER_EXP_OPEN exp STRING_INTER_EXP_CLOSE)* STRING_INTER_END { undefined } + ----------------------------------------------------------------------------- -- Warnings and deprecations @@ -4153,8 +4168,8 @@ getQCONSYM (L _ (ITqconsym x)) = x getIPDUPVARID (L _ (ITdupipvarid x)) = x getLABELVARID (L _ (ITlabelvarid _ x)) = x getCHAR (L _ (ITchar _ x)) = x -getSTRING (L _ (ITstring _ x)) = x -getSTRINGMULTI (L _ (ITstringMulti _ x)) = x +getSTRING (L _ (ITstring _ x StringTypeSingle)) = x +getSTRINGMULTI (L _ (ITstring _ x StringTypeMulti)) = x getINTEGER (L _ (ITinteger x)) = x getRATIONAL (L _ (ITrational x)) = x getPRIMCHAR (L _ (ITprimchar _ x)) = x @@ -4179,8 +4194,8 @@ getVOCURLY (L (RealSrcSpan l _) ITvocurly) = srcSpanStartCol l getINTEGERs (L _ (ITinteger (IL src _ _))) = src getCHARs (L _ (ITchar src _)) = src -getSTRINGs (L _ (ITstring src _)) = src -getSTRINGMULTIs (L _ (ITstringMulti src _)) = src +getSTRINGs (L _ (ITstring src _ StringTypeSingle)) = src +getSTRINGMULTIs (L _ (ITstring src _ StringTypeMulti)) = src getPRIMCHARs (L _ (ITprimchar src _)) = src getPRIMSTRINGs (L _ (ITprimstring src _)) = src getPRIMINTEGERs (L _ (ITprimint src _)) = src ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -168,7 +168,7 @@ $idchar = [$small $large $digit $uniidchar \'] $unigraphic = \x06 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $graphic = [$small $large $symbol $digit $idchar $special $unigraphic \"\'] -$charesc = [a b f n r t v \\ \" \' \&] +$charesc = [a b f n r t v \\ \" \' \& \$] $binit = 0-1 $octit = 0-7 @@ -226,8 +226,9 @@ $docsym = [\| \^ \* \$] -- character sets can be subtracted, not strings @escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) @escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) - at stringchar = ($graphic # [\\ \"]) | $space | @escape | @gap - at char = ($graphic # [\\ \']) | $space | @escapechar + at stringchar = ($graphic # [\\ \"]) | $space | @escape | @gap + at char = ($graphic # [\\ \']) | $space | @escapechar + at stringinterchar = ($graphic # [\\ \" \$]) | $space | @escape | @gap -- normal signed numerical literals can only be explicitly negative, -- not explicitly positive (contrast @exponent) @@ -700,6 +701,21 @@ $unigraphic / { isSmartQuote } { smart_quote_error } (\" | \"\") / ([\n .] # \") { tok_string_multi_content } } +-- See Note [Interpolated strings] +<0> { + s \" { \span _ _ _ -> pushLexState string_inter_content >> pure (L span (ITstringInterBegin StringTypeSingle)) } + -- TODO(bchinn): interpolated multiline strings +} + +-- TODO(bchinn): add string_inter state to all <0> states that can be in an interpolated string + { + @stringinterchar* { tok_string_inter_raw } + \$ \{ { \span _ _ _ -> pushLexState string_inter >> pure (L span ITstringInterExpOpen) } + \" { \span _ _ _ -> popLexState >> pure (L span (ITstringInterEnd StringTypeSingle)) } + + -- TODO(bchinn): check for smart quotes +} + <0> { \'\' { token ITtyQuote } @@ -991,9 +1007,16 @@ data Token -- have a string literal as a label -- Note [Literal source text] in "GHC.Types.SourceText" - | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" - | ITstring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" - | ITstringMulti SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstring SourceText FastString StringType -- Note [Literal source text] in "GHC.Types.SourceText" + + -- See Note [Interpolated strings] + | ITstringInterBegin StringType + | ITstringInterRaw SourceText RawLexedString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringInterExpOpen + | ITstringInterExpClose + | ITstringInterEnd StringType + | ITinteger IntegralLit -- Note [Literal source text] in "GHC.Types.SourceText" | ITrational FractionalLit @@ -1743,8 +1766,11 @@ open_brace span _str _len _buf2 = do setContext (NoLayout:ctx) return (L span ITocurly) close_brace span _str _len _buf2 = do - popContext - return (L span ITccurly) + ctx <- getContext + sc <- getLexState + if null ctx && sc == string_inter + then popLexState >> pure (L span ITstringInterExpClose) + else popContext >> pure (L span ITccurly) qvarid, qconid :: StringBuffer -> Int -> Token qvarid buf len = ITqvarid $! splitQualName buf len False @@ -2234,7 +2260,7 @@ tok_string span buf len _buf2 = do addError err pure $ L span (ITprimstring src (unsafeMkByteString s)) else - pure $ L span (ITstring src (mkFastString s)) + pure $ L span (ITstring src (mkFastString s) StringTypeSingle) where src = SourceText $ lexemeToFastString buf len endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' @@ -2263,10 +2289,10 @@ tok_string_multi startSpan startBuf _len _buf2 = do let contentLen = byteDiff contentStartBuf contentEndBuf s <- either (throwStringLexError (AI startLoc startBuf)) pure $ - lexMultilineString contentLen contentStartBuf + lexString StringTypeMulti contentLen contentStartBuf setInput i' - pure $ L span $ ITstringMulti src (mkFastString s) + pure $ L span $ ITstring src (mkFastString s) StringTypeMulti where goContent i0 = case alexScan i0 string_multi_content of @@ -2310,7 +2336,7 @@ tok_string_multi_content = panic "tok_string_multi_content unexpectedly invoked" lex_chars :: (String, String) -> PsSpan -> StringBuffer -> Int -> P String lex_chars (startDelim, endDelim) span buf len = either (throwStringLexError i0) pure $ - lexString contentLen contentBuf + lexString StringTypeSingle contentLen contentBuf where i0@(AI _ contentBuf) = advanceInputBytes (length startDelim) $ AI (psSpanStart span) buf ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -5,8 +5,12 @@ module GHC.Parser.String ( StringLexError (..), + StringType (..), lexString, - lexMultilineString, + + -- * Raw strings + RawLexedString, + lexStringRaw, -- * Unicode smart quote helpers isDoubleSmartQuote, @@ -17,6 +21,7 @@ import GHC.Prelude hiding (getChar) import Control.Arrow ((>>>)) import Control.Monad (when) +import Data.Bifunctor (first) import Data.Char (chr, ord) import qualified Data.Foldable1 as Foldable1 import qualified Data.List.NonEmpty as NonEmpty @@ -37,13 +42,16 @@ import GHC.Utils.Panic (panic) type BufPos = Int data StringLexError = StringLexError LexErr BufPos -lexString :: Int -> StringBuffer -> Either StringLexError String -lexString = lexStringWith processChars processChars +data StringType = StringTypeSingle | StringTypeMulti deriving (Show) + +lexString :: StringType -> Int -> StringBuffer -> Either StringLexError String +lexString strType = lexStringWith processChars processChars where processChars :: HasChar c => [c] -> Either (c, LexErr) [c] processChars = - collapseGaps - >>> resolveEscapes + case strType of + StringTypeSingle -> processCharsSingle + StringTypeMulti -> processCharsMulti -- ----------------------------------------------------------------------------- -- Lexing interface @@ -122,6 +130,11 @@ bufferLocatedChars initialBuf len = go initialBuf -- ----------------------------------------------------------------------------- -- Lexing phases +processCharsSingle :: HasChar c => [c] -> Either (c, LexErr) [c] +processCharsSingle = + collapseGaps + >>> resolveEscapes + collapseGaps :: HasChar c => [c] -> [c] collapseGaps = go where @@ -255,75 +268,71 @@ isSingleSmartQuote = \case -- Assumes string is lexically valid. Skips the steps about splitting -- and rejoining lines, and instead manually find newline characters, -- for performance. -lexMultilineString :: Int -> StringBuffer -> Either StringLexError String -lexMultilineString = lexStringWith processChars processChars +processCharsMulti :: HasChar c => [c] -> Either (c, LexErr) [c] +processCharsMulti = fmap from . processCharsMulti' . to where - processChars :: HasChar c => [c] -> Either (c, LexErr) [c] - processChars = - collapseGaps -- Step 1 - >>> expandLeadingTabs -- Step 3 - >>> rmCommonWhitespacePrefix -- Step 4 - >>> collapseOnlyWsLines -- Step 5 - >>> rmFirstNewline -- Step 7a - >>> rmLastNewline -- Step 7b - >>> resolveEscapes -- Step 8 - - -- expands all tabs, since the lexer will verify that tabs can only appear - -- as leading indentation - expandLeadingTabs :: HasChar c => [c] -> [c] - expandLeadingTabs = - let go !col = \case - c@(Char '\t') : cs -> - let fill = 8 - (col `mod` 8) - in replicate fill (setChar ' ' c) ++ go (col + fill) cs - c : cs -> c : go (if getChar c == '\n' then 0 else col + 1) cs - [] -> [] - in go 0 + -- Convert a normal multiline string to/from an interpolated multiline string + -- with no interpolated expressions. + to s = [Left s] + from = \case + [Left s] -> s + _ -> panic "Got unexpected result when processing characters in multiline string" + +-- | Process multiline characters generally, for both normal multiline strings and interpolated +-- multiline strings. +processCharsMulti' :: HasChar c => [Either [c] x] -> Either (c, LexErr) [Either [c] x] +processCharsMulti' = + overRaw collapseGaps -- Step 1 + >>> expandLeadingTabs -- Step 3 + >>> rmCommonWhitespacePrefix -- Step 4 + >>> collapseOnlyWsLines -- Step 5 + >>> rmFirstNewline -- Step 7a + >>> rmLastNewline -- Step 7b + >>> overRaw resolveEscapes -- Step 8 + +-- Run the given function over all raw strings, ignoring expressions +overRaw :: (s -> s) -> [Either s x] -> [Either s x] +overRaw f = map $ \case + Left s -> Left $ f s + Right x -> Right x + +-- | Expands all tabs blindly, since the lexer will verify that tabs can only appear +-- as leading indentation +expandLeadingTabs :: HasChar c => [Either [c] x] -> [Either [c] x] +expandLeadingTabs = + -- we can expand each raw string part independently, because leading + -- indentation will never contain an interpolated expression + overRaw $ go 0 + where + go !col = \case + c@(Char '\t') : cs -> + let fill = 8 - (col `mod` 8) + in replicate fill (setChar ' ' c) ++ go (col + fill) cs + c : cs -> c : go (if getChar c == '\n' then 0 else col + 1) cs + [] -> [] - rmCommonWhitespacePrefix :: HasChar c => [c] -> [c] - rmCommonWhitespacePrefix cs0 = - let commonWSPrefix = getCommonWsPrefix (map getChar cs0) - go = \case - c@(Char '\n') : cs -> c : go (dropLine commonWSPrefix cs) - c : cs -> c : go cs - [] -> [] - -- drop x characters from the string, or up to a newline, whichever - -- comes first - dropLine !x = \case - cs | x <= 0 -> cs - cs@(Char '\n' : _) -> cs - _ : cs -> dropLine (x - 1) cs - [] -> [] - in go cs0 - - collapseOnlyWsLines :: HasChar c => [c] -> [c] - collapseOnlyWsLines = - let go = \case - c@(Char '\n') : cs | Just cs' <- checkAllWs cs -> c : go cs' - c : cs -> c : go cs - [] -> [] - checkAllWs = \case - -- got all the way to a newline or the end of the string, return - cs@(Char '\n' : _) -> Just cs - cs@[] -> Just cs - -- found whitespace, continue - Char c : cs | is_space c -> checkAllWs cs - -- anything else, stop - _ -> Nothing - in go - - rmFirstNewline :: HasChar c => [c] -> [c] - rmFirstNewline = \case - Char '\n' : cs -> cs - cs -> cs - - rmLastNewline :: HasChar c => [c] -> [c] - rmLastNewline = - let go = \case - [] -> [] - [Char '\n'] -> [] - c : cs -> c : go cs - in go +rmCommonWhitespacePrefix :: HasChar c => [Either [c] x] -> [Either [c] x] +rmCommonWhitespacePrefix parts = + -- Whitespace prefix, by definition, only comes after newline characters, and there can + -- never be an interpolated expr within a whitespace prefix (since the expr would end + -- the prefix). So we can use a plain `map` to just process the string parts, because + -- the "drop prefix" logic will never span over multiple parts. TODO(bchinn): what about two Strings next to each other? + map (first go) parts + where + -- treat interpolated exprs as a single, non-space character string + commonWSPrefix = getCommonWsPrefix $ concatMap (either (map getChar) (const "x")) parts + + go = \case + c@(Char '\n') : cs -> c : go (dropPrefix commonWSPrefix cs) + c : cs -> c : go cs + [] -> [] + + -- drop x characters from the string, or up to a newline, whichever comes first + dropPrefix !x = \case + cs | x <= 0 -> cs + cs@(Char '\n' : _) -> cs + _ : cs -> dropPrefix (x - 1) cs + [] -> [] -- | See step 4 in Note [Multiline string literals] -- @@ -339,6 +348,57 @@ getCommonWsPrefix s = . drop 1 -- ignore first line in calculation $ lines s +-- TODO(bchinn): treat interpolated exprs as one character non-space string +collapseOnlyWsLines :: HasChar c => [Either [c] x] -> [Either [c] x] +collapseOnlyWsLines = goParts + where + goParts = \case + Right x : parts -> Right x : goParts parts + [Left s] -> _ + Left s : _ -> _ + [] -> [] + + go = \case + c@(Char '\n') : cs | Just cs' <- checkAllWs cs -> c : go cs' + c : cs -> c : go cs + [] -> [] + + -- Return (String, Bool, String), where: + -- * String - The line that was just parsed, ending with a newline character, unless + -- we reached the end of the string + -- * Bool - Whether the line that was just parsed was all whitespace characters + -- * String - The rest of the string + takeLine = + let go acc allWS = \case + -- reached a newline or the end of the string + c@(Char '\n') : cs -> (dlistToList (acc `dlistSnoc` c), allWS, cs) + [] -> (dlistToList acc, allWS, []) + -- found whitespace + c : cs | is_space (getChar c) -> go (acc `dlistSnoc` c) allWS + -- found non-whitespace + c : cs -> go (acc `dlistSnoc` c) False + in go dlistEmpty True + +rmFirstNewline :: HasChar c => [Either [c] x] -> [Either [c] x] +rmFirstNewline = \case + Left (Char '\n' : cs) : parts -> Left cs : parts + parts -> parts + +rmLastNewline :: HasChar c => [Either [c] x] -> [Either [c] x] +rmLastNewline = goParts + where + goParts = \case + [] -> [] + [Left s] -> + case goLastLine s of + [] -> [] + s' -> [Left s'] + c : cs -> c : goParts cs + goLastLine = \case + [] -> [] + [Char '\n'] -> [] + c : cs -> c : goLastLine cs + {- Note [Multiline string literals] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -348,8 +408,8 @@ proposal: https://github.com/ghc-proposals/ghc-proposals/pull/569 Multiline string literals are syntax sugar for normal string literals, with an extra post processing step. This all happens in the Lexer; that -is, HsMultilineString will contain the post-processed string. This matches -the same behavior as HsString, which contains the normalized string +is, HsString will contain the post-processed string. This matches the same +behavior as single-line HsString, which contains the normalized string (see Note [Literal source text]). The canonical steps for post processing a multiline string are: @@ -378,6 +438,76 @@ It's more precisely defined with the following algorithm: 3. Calculate the longest prefix of whitespace shared by all lines in the remaining list -} +-- ----------------------------------------------------------------------------- +-- Interpolated strings + +-- | A string that's been validated to be lexically correct, but still +-- contains the raw string lexed, without anything resolved. +newtype RawLexedString = RawLexedString {unRawLexedString :: String} + +-- | Load and validate the string in the given StringBuffer. +-- +-- e.g. Lexing "a\nb" will return RawLexedString ['a', '\\', 'n', 'b']. +lexStringRaw :: Int -> StringBuffer -> Either StringLexError RawLexedString +lexStringRaw len buf = RawLexedString (bufferChars len buf) <$ lexString strType len buf + where + -- Always lex as a single-line string, because single-line and multi-line strings + -- have the same validation logic, so we can skip the multi-line processing steps + strType = StringTypeSingle + +fromRawLexedStringSingle :: RawLexedString -> String +fromRawLexedStringSingle (RawLexedString s) = + case processCharsSingle s of + Right s' -> s' + Left _ -> panic "Unexpectedly got an error when re-lexing the string" + +fromRawLexedStringMulti :: [Either RawLexedString x] -> [Either String x] +fromRawLexedStringMulti parts = + case processCharsMulti' $ map (first unRawLexedString) parts of + Right parts' -> parts' + Left _ -> panic "Unexpectedly got an error when re-lexing the string" + +{- +Note [Interpolated strings] +~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Interpolated string syntax was accepted in this proposal: +https://github.com/ghc-proposals/ghc-proposals/pull/570 + +Interpolated strings are syntax sugar for + +Interpolated strings are implemented in the following manner: + +1. Lexer takes the string as input: + + s"Hello ${Text.toUpper name}!" + + and outputs the following tokens: + + [ ITstringInterBegin src StringTypeSingle + , ITstringInterRaw src "Hello " + , ITstringInterExpOpen src + , ITqvarid ("Text.toUpper", "name") + , ITvarid "name" + , ITstringInterExpClose src + , ITstringInterRaw src "!" + , ITstringInterEnd src StringTypeSingle + ] + +2. The parser will then parse the tokens into the following HsExpr: + + HsInterString ext + [ HsInterRaw ext "Hello " + , HsInterExp ext $ + HsApp ext + (HsVar ext 'Text.toUpper) + (HsVar ext 'name) + , HsInterRaw ext "!" + ] + +3. This will then be desugared into +-} + -- ----------------------------------------------------------------------------- -- DList ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -377,8 +377,7 @@ rnExpr (HsLit x lit) | Just (src, s) <- stringLike lit ; return (HsLit x (convertLit lit), emptyFVs) } } where stringLike = \case - HsString src s -> Just (src, s) - HsMultilineString src s -> Just (src, s) + HsString src _ s -> Just (src, s) _ -> Nothing rnExpr (HsLit x lit) @@ -393,6 +392,9 @@ rnExpr (HsOverLit x lit) return (HsApp noExtField (noLocA neg) (noLocA (HsOverLit x lit')) , fvs ) } +rnExpr (HsInterString x ty parts) + = undefined -- TODO(bchinn) + rnExpr (HsApp x fun arg) = do { (fun',fvFun) <- rnLExpr fun ; (arg',fvArg) <- rnLExpr arg ===================================== compiler/GHC/Tc/Gen/HsType.hs ===================================== @@ -4725,8 +4725,7 @@ addTyConFlavCtxt name flav tyLitFromLit :: HsLit GhcRn -> Maybe (HsTyLit GhcRn) -tyLitFromLit (HsString x str) = Just (HsStrTy x str) -tyLitFromLit (HsMultilineString x str) = Just (HsStrTy x str) +tyLitFromLit (HsString x _ str) = Just (HsStrTy x str) tyLitFromLit (HsChar x char) = Just (HsCharTy x char) tyLitFromLit _ = Nothing ===================================== compiler/Language/Haskell/Syntax/Expr.hs ===================================== @@ -350,7 +350,6 @@ data HsExpr p -- erroring expression will be written after -- solving. See Note [Holes] in GHC.Tc.Types.Constraint. - | HsRecSel (XRecSel p) (FieldOcc p) -- ^ Variable pointing to record selector -- See Note [Non-overloaded record field selectors] and @@ -361,12 +360,19 @@ data HsExpr p | HsIPVar (XIPVar p) HsIPName -- ^ Implicit parameter (not in use after typechecking) + | HsOverLit (XOverLitE p) (HsOverLit p) -- ^ Overloaded literals | HsLit (XLitE p) (HsLit p) -- ^ Simple (non-overloaded) literals + | -- | See Note [Interpolated strings] + HsInterString + (XInterString p) + HsStringType + [Either FastString (LHsExpr p)] + -- | Lambda, Lambda-case, and Lambda-cases -- -- - 'GHC.Parser.Annotation.AnnKeywordId' : 'GHC.Parser.Annotation.AnnLam', ===================================== compiler/Language/Haskell/Syntax/Extension.hs ===================================== @@ -413,6 +413,9 @@ type family XOverLabel x type family XIPVar x type family XOverLitE x type family XLitE x +type family XInterString x +type family XInterStringRaw x +type family XInterStringExp x type family XLam x type family XLamCase x type family XApp x @@ -556,7 +559,6 @@ type family XXParStmtBlock x x' type family XHsChar x type family XHsCharPrim x type family XHsString x -type family XHsMultilineString x type family XHsStringPrim x type family XHsInt x type family XHsIntPrim x ===================================== compiler/Language/Haskell/Syntax/Lit.hs ===================================== @@ -51,9 +51,7 @@ data HsLit x -- ^ Character | HsCharPrim (XHsCharPrim x) {- SourceText -} Char -- ^ Unboxed character - | HsString (XHsString x) {- SourceText -} FastString - -- ^ String - | HsMultilineString (XHsMultilineString x) {- SourceText -} FastString + | HsString (XHsString x) {- SourceText -} HsStringType FastString -- ^ String | HsStringPrim (XHsStringPrim x) {- SourceText -} !ByteString -- ^ Packed bytes @@ -145,3 +143,5 @@ instance Ord OverLitVal where compare (HsIsString _ s1) (HsIsString _ s2) = s1 `lexicalCompareFS` s2 compare (HsIsString _ _) (HsIntegral _) = GT compare (HsIsString _ _) (HsFractional _) = GT + +data HsStringType = HsStringTypeSingle | HsStringTypeMulti ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -4923,8 +4923,7 @@ hsLit2String lit = case lit of HsChar src v -> toSourceTextWithSuffix src v "" HsCharPrim src p -> toSourceTextWithSuffix src p "" - HsString src v -> toSourceTextWithSuffix src v "" - HsMultilineString src v -> toSourceTextWithSuffix src v "" + HsString src _ v -> toSourceTextWithSuffix src v "" HsStringPrim src v -> toSourceTextWithSuffix src v "" HsInt _ (IL src _ v) -> toSourceTextWithSuffix src v "" HsIntPrim src v -> toSourceTextWithSuffix src v "" ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs ===================================== @@ -320,7 +320,6 @@ classify tok = ITlabelvarid{} -> TkUnknown ITchar{} -> TkChar ITstring{} -> TkString - ITstringMulti{} -> TkString ITinteger{} -> TkNumber ITrational{} -> TkNumber ITprimchar{} -> TkChar View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/65d1c307e0dbf367ccc3555a1e6411b45c8efd09...cc7d07b3b8a0eb678af4eeaed2726c1cc8c79ab3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/65d1c307e0dbf367ccc3555a1e6411b45c8efd09...cc7d07b3b8a0eb678af4eeaed2726c1cc8c79ab3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 05:34:19 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Wed, 25 Sep 2024 01:34:19 -0400 Subject: [Git][ghc/ghc][wip/interpolated-strings] 6 commits: Merge branch 'wip/strings' into HEAD Message-ID: <66f3a0db36035_152f2f81d0a846016@gitlab.mail> Brandon Chinn pushed to branch wip/interpolated-strings at Glasgow Haskell Compiler / GHC Commits: 03079a7d by Brandon Chinn at 2024-09-24T22:26:16-07:00 Merge branch 'wip/strings' into HEAD - - - - - 8cede8ef by Brandon Chinn at 2024-09-24T22:26:16-07:00 Unify ITstring + ITstringMulti - - - - - 3975ce84 by Brandon Chinn at 2024-09-24T22:26:16-07:00 Unify HsString + HsMultilineString - - - - - 21c589bd by Brandon Chinn at 2024-09-24T22:26:16-07:00 Move multiline string processing functions to top-level - - - - - 284b92d1 by Brandon Chinn at 2024-09-24T22:28:01-07:00 Always use processCharsSingle to get StringLexError - - - - - 8e7eae73 by Brandon Chinn at 2024-09-24T22:34:03-07:00 Implement interpolated strings - - - - - 15 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Lit.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/Language/Haskell/Syntax/Expr.hs - compiler/Language/Haskell/Syntax/Extension.hs - compiler/Language/Haskell/Syntax/Lit.hs - utils/check-exact/ExactPrint.hs - utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -242,6 +242,9 @@ type instance XIPVar GhcRn = NoExtField type instance XIPVar GhcTc = DataConCantHappen type instance XOverLitE (GhcPass _) = NoExtField type instance XLitE (GhcPass _) = NoExtField +type instance XInterString (GhcPass _) = NoExtField +type instance XInterStringRaw (GhcPass _) = NoExtField +type instance XInterStringExp (GhcPass _) = NoExtField type instance XLam (GhcPass _) = [AddEpAnn] type instance XApp (GhcPass _) = NoExtField ===================================== compiler/GHC/Hs/Lit.hs ===================================== @@ -49,7 +49,6 @@ import Language.Haskell.Syntax.Lit type instance XHsChar (GhcPass _) = SourceText type instance XHsCharPrim (GhcPass _) = SourceText type instance XHsString (GhcPass _) = SourceText -type instance XHsMultilineString (GhcPass _) = SourceText type instance XHsStringPrim (GhcPass _) = SourceText type instance XHsInt (GhcPass _) = NoExtField type instance XHsIntPrim (GhcPass _) = SourceText @@ -136,7 +135,6 @@ hsLitNeedsParens p = go go (HsChar {}) = False go (HsCharPrim {}) = False go (HsString {}) = False - go (HsMultilineString {}) = False go (HsStringPrim {}) = False go (HsInt _ x) = p > topPrec && il_neg x go (HsInteger _ x _) = p > topPrec && x < 0 @@ -160,7 +158,6 @@ convertLit :: HsLit (GhcPass p1) -> HsLit (GhcPass p2) convertLit (HsChar a x) = HsChar a x convertLit (HsCharPrim a x) = HsCharPrim a x convertLit (HsString a x) = HsString a x -convertLit (HsMultilineString a x) = HsMultilineString a x convertLit (HsStringPrim a x) = HsStringPrim a x convertLit (HsInt a x) = HsInt a x convertLit (HsIntPrim a x) = HsIntPrim a x @@ -197,11 +194,11 @@ Equivalently it's True if instance Outputable (HsLit (GhcPass p)) where ppr (HsChar st c) = pprWithSourceText st (pprHsChar c) ppr (HsCharPrim st c) = pprWithSourceText st (pprPrimChar c) - ppr (HsString st s) = pprWithSourceText st (pprHsString s) - ppr (HsMultilineString st s) = - case st of - NoSourceText -> pprHsString s - SourceText src -> vcat $ map text $ split '\n' (unpackFS src) + ppr (HsString st ty s) = + case (ty, st) of + (HsStringTypeSingle, _) -> pprWithSourceText st (pprHsString s) + (HsStringTypeMulti, NoSourceText) -> pprHsString s + (HsStringTypeMulti, SourceText src) -> vcat $ map text $ split '\n' (unpackFS src) ppr (HsStringPrim st s) = pprWithSourceText st (pprHsBytes s) ppr (HsInt _ i) = pprWithSourceText (il_text i) (integer (il_value i)) @@ -241,7 +238,6 @@ pmPprHsLit :: HsLit (GhcPass x) -> SDoc pmPprHsLit (HsChar _ c) = pprHsChar c pmPprHsLit (HsCharPrim _ c) = pprHsChar c pmPprHsLit (HsString st s) = pprWithSourceText st (pprHsString s) -pmPprHsLit (HsMultilineString st s) = pprWithSourceText st (pprHsString s) pmPprHsLit (HsStringPrim _ s) = pprHsBytes s pmPprHsLit (HsInt _ i) = integer (il_value i) pmPprHsLit (HsIntPrim _ i) = integer i ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -75,8 +75,7 @@ hsPatType (SplicePat v _) = dataConCantHappen v hsLitType :: HsLit (GhcPass p) -> Type hsLitType (HsChar _ _) = charTy hsLitType (HsCharPrim _ _) = charPrimTy -hsLitType (HsString _ _) = stringTy -hsLitType (HsMultilineString _ _) = stringTy +hsLitType (HsString _ _ _) = stringTy hsLitType (HsStringPrim _ _) = addrPrimTy hsLitType (HsInt _ _) = intTy hsLitType (HsIntPrim _ _) = intPrimTy ===================================== compiler/GHC/HsToCore/Match/Literal.hs ===================================== @@ -121,7 +121,6 @@ dsLit l = do HsDoublePrim _ fl -> return (Lit (LitDouble (rationalFromFractionalLit fl))) HsChar _ c -> return (mkCharExpr c) HsString _ str -> mkStringExprFS str - HsMultilineString _ str -> mkStringExprFS str HsInteger _ i _ -> return (mkIntegerExpr platform i) HsInt _ i -> return (mkIntExpr platform (il_value i)) HsRat _ fl ty -> dsFractionalLitToRational fl ty @@ -475,7 +474,6 @@ getSimpleIntegralLit (HsInteger _ i ty) = Just (i, ty) getSimpleIntegralLit HsChar{} = Nothing getSimpleIntegralLit HsCharPrim{} = Nothing getSimpleIntegralLit HsString{} = Nothing -getSimpleIntegralLit HsMultilineString{} = Nothing getSimpleIntegralLit HsStringPrim{} = Nothing getSimpleIntegralLit HsRat{} = Nothing getSimpleIntegralLit HsFloatPrim{} = Nothing ===================================== compiler/GHC/HsToCore/Quote.hs ===================================== @@ -3025,8 +3025,7 @@ repLiteral lit HsDoublePrim _ _ -> Just doublePrimLName HsChar _ _ -> Just charLName HsCharPrim _ _ -> Just charPrimLName - HsString _ _ -> Just stringLName - HsMultilineString _ _ -> Just stringLName + HsString _ _ _ -> Just stringLName HsRat _ _ _ -> Just rationalLName _ -> Nothing ===================================== compiler/GHC/Parser.y ===================================== @@ -727,8 +727,13 @@ are the most common patterns, rewritten as regular expressions for clarity: LABELVARID { L _ (ITlabelvarid _ _) } CHAR { L _ (ITchar _ _) } - STRING { L _ (ITstring _ _) } - STRING_MULTI { L _ (ITstringMulti _ _) } + STRING { L _ (ITstring _ _ StringTypeSingle) } + STRING_MULTI { L _ (ITstring _ _ StringTypeMulti) } + STRING_INTER_BEGIN { L _ (ITstringInterBegin _) } + STRING_INTER_RAW { L _ (ITstringInterRaw _ _) } + STRING_INTER_EXP_OPEN { L _ ITstringInterExpOpen } + STRING_INTER_EXP_CLOSE { L _ ITstringInterExpClose } + STRING_INTER_END { L _ (ITstringInterEnd _) } INTEGER { L _ (ITinteger _) } RATIONAL { L _ (ITrational _) } @@ -3080,6 +3085,8 @@ aexp2 :: { ECP } -- into HsOverLit when -XOverloadedStrings is on. -- | STRING { sL (getLoc $1) (HsOverLit $! mkHsIsString (getSTRINGs $1) -- (getSTRING $1) noExtField) } + | stringInter {% fmap ecpFromExp + (ams1 $1 (HsInterString NoExtField $! unLoc $1)) } | INTEGER { ECP $ mkHsOverLitPV (sL1a $1 $ mkHsIntegral (getINTEGER $1)) } | RATIONAL { ECP $ mkHsOverLitPV (sL1a $1 $ mkHsFractional (getRATIONAL $1)) } @@ -3677,6 +3684,14 @@ ipvar :: { Located HsIPName } overloaded_label :: { Located (SourceText, FastString) } : LABELVARID { sL1 $1 (getLABELVARIDs $1, getLABELVARID $1) } +----------------------------------------------------------------------------- +-- Interpolated strings +-- See Note [Interpolated strings] in GHC.Parser.String + +stringInter :: { [Either FastString (LHsExpr GhcPs)] } + -- TODO(bchinn): break out recursive stringInterPart rule + : STRING_INTER_BEGIN (STRING_INTER_RAW | STRING_INTER_EXP_OPEN exp STRING_INTER_EXP_CLOSE)* STRING_INTER_END { undefined } + ----------------------------------------------------------------------------- -- Warnings and deprecations @@ -4153,8 +4168,8 @@ getQCONSYM (L _ (ITqconsym x)) = x getIPDUPVARID (L _ (ITdupipvarid x)) = x getLABELVARID (L _ (ITlabelvarid _ x)) = x getCHAR (L _ (ITchar _ x)) = x -getSTRING (L _ (ITstring _ x)) = x -getSTRINGMULTI (L _ (ITstringMulti _ x)) = x +getSTRING (L _ (ITstring _ x StringTypeSingle)) = x +getSTRINGMULTI (L _ (ITstring _ x StringTypeMulti)) = x getINTEGER (L _ (ITinteger x)) = x getRATIONAL (L _ (ITrational x)) = x getPRIMCHAR (L _ (ITprimchar _ x)) = x @@ -4179,8 +4194,8 @@ getVOCURLY (L (RealSrcSpan l _) ITvocurly) = srcSpanStartCol l getINTEGERs (L _ (ITinteger (IL src _ _))) = src getCHARs (L _ (ITchar src _)) = src -getSTRINGs (L _ (ITstring src _)) = src -getSTRINGMULTIs (L _ (ITstringMulti src _)) = src +getSTRINGs (L _ (ITstring src _ StringTypeSingle)) = src +getSTRINGMULTIs (L _ (ITstring src _ StringTypeMulti)) = src getPRIMCHARs (L _ (ITprimchar src _)) = src getPRIMSTRINGs (L _ (ITprimstring src _)) = src getPRIMINTEGERs (L _ (ITprimint src _)) = src ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -168,7 +168,7 @@ $idchar = [$small $large $digit $uniidchar \'] $unigraphic = \x06 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $graphic = [$small $large $symbol $digit $idchar $special $unigraphic \"\'] -$charesc = [a b f n r t v \\ \" \' \&] +$charesc = [a b f n r t v \\ \" \' \& \$] $binit = 0-1 $octit = 0-7 @@ -226,8 +226,9 @@ $docsym = [\| \^ \* \$] -- character sets can be subtracted, not strings @escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) @escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) - at stringchar = ($graphic # [\\ \"]) | $space | @escape | @gap - at char = ($graphic # [\\ \']) | $space | @escapechar + at stringchar = ($graphic # [\\ \"]) | $space | @escape | @gap + at char = ($graphic # [\\ \']) | $space | @escapechar + at stringinterchar = ($graphic # [\\ \" \$]) | $space | @escape | @gap -- normal signed numerical literals can only be explicitly negative, -- not explicitly positive (contrast @exponent) @@ -700,6 +701,21 @@ $unigraphic / { isSmartQuote } { smart_quote_error } (\" | \"\") / ([\n .] # \") { tok_string_multi_content } } +-- See Note [Interpolated strings] +<0> { + s \" { \span _ _ _ -> pushLexState string_inter_content >> pure (L span (ITstringInterBegin StringTypeSingle)) } + -- TODO(bchinn): interpolated multiline strings +} + +-- TODO(bchinn): add string_inter state to all <0> states that can be in an interpolated string + { + @stringinterchar* { tok_string_inter_raw } + \$ \{ { \span _ _ _ -> pushLexState string_inter >> pure (L span ITstringInterExpOpen) } + \" { \span _ _ _ -> popLexState >> pure (L span (ITstringInterEnd StringTypeSingle)) } + + -- TODO(bchinn): check for smart quotes +} + <0> { \'\' { token ITtyQuote } @@ -991,9 +1007,16 @@ data Token -- have a string literal as a label -- Note [Literal source text] in "GHC.Types.SourceText" - | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" - | ITstring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" - | ITstringMulti SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstring SourceText FastString StringType -- Note [Literal source text] in "GHC.Types.SourceText" + + -- See Note [Interpolated strings] + | ITstringInterBegin StringType + | ITstringInterRaw SourceText RawLexedString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringInterExpOpen + | ITstringInterExpClose + | ITstringInterEnd StringType + | ITinteger IntegralLit -- Note [Literal source text] in "GHC.Types.SourceText" | ITrational FractionalLit @@ -1743,8 +1766,11 @@ open_brace span _str _len _buf2 = do setContext (NoLayout:ctx) return (L span ITocurly) close_brace span _str _len _buf2 = do - popContext - return (L span ITccurly) + ctx <- getContext + sc <- getLexState + if null ctx && sc == string_inter + then popLexState >> pure (L span ITstringInterExpClose) + else popContext >> pure (L span ITccurly) qvarid, qconid :: StringBuffer -> Int -> Token qvarid buf len = ITqvarid $! splitQualName buf len False @@ -2234,7 +2260,7 @@ tok_string span buf len _buf2 = do addError err pure $ L span (ITprimstring src (unsafeMkByteString s)) else - pure $ L span (ITstring src (mkFastString s)) + pure $ L span (ITstring src (mkFastString s) StringTypeSingle) where src = SourceText $ lexemeToFastString buf len endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' @@ -2263,10 +2289,10 @@ tok_string_multi startSpan startBuf _len _buf2 = do let contentLen = byteDiff contentStartBuf contentEndBuf s <- either (throwStringLexError (AI startLoc startBuf)) pure $ - lexMultilineString contentLen contentStartBuf + lexString StringTypeMulti contentLen contentStartBuf setInput i' - pure $ L span $ ITstringMulti src (mkFastString s) + pure $ L span $ ITstring src (mkFastString s) StringTypeMulti where goContent i0 = case alexScan i0 string_multi_content of @@ -2310,7 +2336,7 @@ tok_string_multi_content = panic "tok_string_multi_content unexpectedly invoked" lex_chars :: (String, String) -> PsSpan -> StringBuffer -> Int -> P String lex_chars (startDelim, endDelim) span buf len = either (throwStringLexError i0) pure $ - lexString contentLen contentBuf + lexString StringTypeSingle contentLen contentBuf where i0@(AI _ contentBuf) = advanceInputBytes (length startDelim) $ AI (psSpanStart span) buf ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -5,8 +5,12 @@ module GHC.Parser.String ( StringLexError (..), + StringType (..), lexString, - lexMultilineString, + + -- * Raw strings + RawLexedString, + lexStringRaw, -- * Unicode smart quote helpers isDoubleSmartQuote, @@ -17,8 +21,11 @@ import GHC.Prelude hiding (getChar) import Control.Arrow ((>>>)) import Control.Monad (when) +import Data.Bifunctor (first) import Data.Char (chr, ord) import qualified Data.Foldable1 as Foldable1 +import Data.Functor.Identity (Identity (..)) +import Data.List (unsnoc) import qualified Data.List.NonEmpty as NonEmpty import Data.Maybe (listToMaybe, mapMaybe) import GHC.Data.StringBuffer (StringBuffer) @@ -37,13 +44,16 @@ import GHC.Utils.Panic (panic) type BufPos = Int data StringLexError = StringLexError LexErr BufPos -lexString :: Int -> StringBuffer -> Either StringLexError String -lexString = lexStringWith processChars processChars +data StringType = StringTypeSingle | StringTypeMulti deriving (Show) + +lexString :: StringType -> Int -> StringBuffer -> Either StringLexError String +lexString strType = lexStringWith processChars where - processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars :: HasChar c => [c] -> Maybe [c] processChars = - collapseGaps - >>> resolveEscapes + case strType of + StringTypeSingle -> fromRight . processCharsSingle + StringTypeMulti -> processCharsMulti -- ----------------------------------------------------------------------------- -- Lexing interface @@ -66,26 +76,27 @@ So what we'll do is do two passes. The first pass is optimistic; just convert to a plain String and process it. If this results in an error, we do a second pass, this time where each character is annotated with its position. Now, the error has all the information it needs. - -Ideally, lexStringWith would take a single (forall c. HasChar c => ...) function, -but to help the specializer, we pass it in twice to concretize it for the two -types we actually use. -} -- | See Note [Lexing strings] -lexStringWith :: - ([Char] -> Either (Char, LexErr) [Char]) - -> ([CharPos] -> Either (CharPos, LexErr) [CharPos]) - -> Int - -> StringBuffer - -> Either StringLexError String -lexStringWith processChars processCharsPos len buf = +lexStringWith :: (String -> Maybe String) -> Int -> StringBuffer -> Either StringLexError String +lexStringWith processChars len buf = case processChars $ bufferChars buf len of - Right s -> Right s - Left _ -> - case processCharsPos $ bufferLocatedChars buf len of - Right _ -> panic "expected lex error on second pass" - Left ((_, pos), e) -> Left $ StringLexError e pos + Just s -> Right s + Nothing -> do + validateString len buf -- should return Left + panic "expected lex error on second pass" + +-- | Find any lexical errors in the string. +-- +-- Can validate both single- and multi-line strings, since multi-line strings +-- have the same validation logic as single-line strings, and none of the +-- multi-line string processing steps affect the validity of the string. +validateString :: Int -> StringBuffer -> Either StringLexError () +validateString len buf = + case processCharsSingle $ bufferLocatedChars buf len of + Right _ -> Right () + Left ((_, pos), e) -> Left $ StringLexError e pos class HasChar c where getChar :: c -> Char @@ -122,6 +133,11 @@ bufferLocatedChars initialBuf len = go initialBuf -- ----------------------------------------------------------------------------- -- Lexing phases +processCharsSingle :: HasChar c => [c] -> Either (c, LexErr) [c] +processCharsSingle = + collapseGaps + >>> resolveEscapes + collapseGaps :: HasChar c => [c] -> [c] collapseGaps = go where @@ -149,6 +165,9 @@ resolveEscapes = go dlistEmpty Left (c, e) -> Left (c, e) c : cs -> go (acc `dlistSnoc` c) cs +resolveEscapesMaybe :: HasChar c => [c] -> Maybe [c] +resolveEscapesMaybe = fromRight . resolveEscapes + -- ----------------------------------------------------------------------------- -- Escape characters @@ -247,6 +266,75 @@ isSingleSmartQuote = \case '’' -> True _ -> False +-- ----------------------------------------------------------------------------- +-- Interpolated strings + +-- | A string that's been validated to be lexically correct, but still +-- contains the raw string lexed, without anything resolved. +newtype RawLexedString = RawLexedString {unRawLexedString :: String} + +-- | Load and validate the string in the given StringBuffer. +-- +-- e.g. Lexing "a\nb" will return RawLexedString ['a', '\\', 'n', 'b']. +lexStringRaw :: Int -> StringBuffer -> Either StringLexError RawLexedString +lexStringRaw len buf = RawLexedString (bufferChars len buf) <$ validateString len buf + +fromRawLexedStringSingle :: RawLexedString -> String +fromRawLexedStringSingle (RawLexedString s) = + case processCharsSingle s of + Right s' -> s' + Left _ -> panic "Unexpectedly got an error when re-lexing the string" + +fromRawLexedStringMulti :: (RawLexedString, [(x, RawLexedString)]) -> (String, [(x, String)]) +fromRawLexedStringMulti s = + case processCharsMulti (to s) of + Just s' -> from s' + Nothing -> panic "Unexpectedly got an error when re-lexing the string" + where + to (pre, parts) = InterMultiString pre parts + from (InterMultiString pre parts) = (pre, parts) + +{- +Note [Interpolated strings] +~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Interpolated string syntax was accepted in this proposal: +https://github.com/ghc-proposals/ghc-proposals/pull/570 + +Interpolated strings are syntax sugar for + +Interpolated strings are implemented in the following manner: + +1. Lexer takes the string as input: + + s"Hello ${Text.toUpper name}!" + + and outputs the following tokens: + + [ ITstringInterBegin src StringTypeSingle + , ITstringInterRaw src "Hello " + , ITstringInterExpOpen src + , ITqvarid ("Text.toUpper", "name") + , ITvarid "name" + , ITstringInterExpClose src + , ITstringInterRaw src "!" + , ITstringInterEnd src StringTypeSingle + ] + +2. The parser will then parse the tokens into the following HsExpr: + + HsInterString ext + [ HsInterRaw ext "Hello " + , HsInterExp ext $ + HsApp ext + (HsVar ext 'Text.toUpper) + (HsVar ext 'name) + , HsInterRaw ext "!" + ] + +3. This will then be desugared into +-} + -- ----------------------------------------------------------------------------- -- Multiline strings @@ -255,75 +343,84 @@ isSingleSmartQuote = \case -- Assumes string is lexically valid. Skips the steps about splitting -- and rejoining lines, and instead manually find newline characters, -- for performance. -lexMultilineString :: Int -> StringBuffer -> Either StringLexError String -lexMultilineString = lexStringWith processChars processChars +processCharsMulti :: String -> Maybe String +processCharsMulti = fmap from . processCharsMulti' . to where - processChars :: HasChar c => [c] -> Either (c, LexErr) [c] - processChars = - collapseGaps -- Step 1 - >>> expandLeadingTabs -- Step 3 - >>> rmCommonWhitespacePrefix -- Step 4 - >>> collapseOnlyWsLines -- Step 5 - >>> rmFirstNewline -- Step 7a - >>> rmLastNewline -- Step 7b - >>> resolveEscapes -- Step 8 - - -- expands all tabs, since the lexer will verify that tabs can only appear - -- as leading indentation - expandLeadingTabs :: HasChar c => [c] -> [c] - expandLeadingTabs = - let go !col = \case - c@(Char '\t') : cs -> - let fill = 8 - (col `mod` 8) - in replicate fill (setChar ' ' c) ++ go (col + fill) cs - c : cs -> c : go (if getChar c == '\n' then 0 else col + 1) cs - [] -> [] - in go 0 + -- Convert a normal multiline string to/from an interpolated multiline string + -- with no interpolated expressions. + to s = InterMultiString s [] + from = \case + InterMultiString s [] -> s + _ -> panic "Got unexpected result when processing characters in multiline string" + +-- | An interpolated, multiline string to be processed. +-- +-- `x` here will only ever be instantiated as `HsExpr`, but we'll leave it general to ensure +-- we never modify it, we only ever propagate it. +-- +-- We represent this as a list of (x, String) tuples instead of [Either x String] to guarantee +-- that we don't have to handle two raw Strings next to each other. +data InterMultiString x = + InterMultiString + String -- ^ beginning of the string before the first interpolated expr + [(x, String)] -- ^ (expr, raw string) interleaved groups + +-- Run the given function over all raw strings, ignoring expressions +overRaw :: (String -> String) -> InterMultiString x -> InterMultiString x +overRaw f = runIdentity . overRawM (Identity . f) + +overRawM :: Monad m => (String -> m String) -> InterMultiString x -> m (InterMultiString x) +overRawM f (InterMultiString pre parts) = InterMultiString <$> f pre <*> (traverse . traverse) f parts + +-- | Process multiline characters generally, for both normal multiline strings and interpolated +-- multiline strings. +processCharsMulti' :: InterMultiString x -> Maybe (InterMultiString x) +processCharsMulti' = + overRaw collapseGaps -- Step 1 + >>> expandLeadingTabs -- Step 3 + >>> rmCommonWhitespacePrefix -- Step 4 + >>> collapseOnlyWsLines -- Step 5 + >>> rmFirstNewline -- Step 7a + >>> rmLastNewline -- Step 7b + >>> overRawM resolveEscapesMaybe -- Step 8 + +-- | Expands all tabs blindly, since the lexer will verify that tabs can only appear +-- as leading indentation +expandLeadingTabs :: InterMultiString x -> InterMultiString x +expandLeadingTabs = + -- we can expand each raw string part independently, because leading + -- indentation will never contain an interpolated expression + overRaw $ go 0 + where + go !col = \case + c@'\t' : cs -> + let fill = 8 - (col `mod` 8) + in replicate fill ' ' ++ go (col + fill) cs + c : cs -> c : go (if c == '\n' then 0 else col + 1) cs + [] -> [] + +rmCommonWhitespacePrefix :: InterMultiString x -> InterMultiString x +rmCommonWhitespacePrefix s0 = + -- Whitespace prefix, by definition, only comes after newline characters, and there can + -- never be an interpolated expr within a whitespace prefix (since the expr would end + -- the prefix). So we can use a plain `map` to just process the string parts, because + -- the "drop prefix" logic will never span over multiple parts. + map (first go) parts + where + -- treat interpolated exprs as a single, non-space character string + commonWSPrefix = getCommonWsPrefix $ case s0 of InterMultiString pre parts -> pre ++ concatMap snd parts - rmCommonWhitespacePrefix :: HasChar c => [c] -> [c] - rmCommonWhitespacePrefix cs0 = - let commonWSPrefix = getCommonWsPrefix (map getChar cs0) - go = \case - c@(Char '\n') : cs -> c : go (dropLine commonWSPrefix cs) - c : cs -> c : go cs - [] -> [] - -- drop x characters from the string, or up to a newline, whichever - -- comes first - dropLine !x = \case - cs | x <= 0 -> cs - cs@(Char '\n' : _) -> cs - _ : cs -> dropLine (x - 1) cs - [] -> [] - in go cs0 - - collapseOnlyWsLines :: HasChar c => [c] -> [c] - collapseOnlyWsLines = - let go = \case - c@(Char '\n') : cs | Just cs' <- checkAllWs cs -> c : go cs' - c : cs -> c : go cs - [] -> [] - checkAllWs = \case - -- got all the way to a newline or the end of the string, return - cs@(Char '\n' : _) -> Just cs - cs@[] -> Just cs - -- found whitespace, continue - Char c : cs | is_space c -> checkAllWs cs - -- anything else, stop - _ -> Nothing - in go - - rmFirstNewline :: HasChar c => [c] -> [c] - rmFirstNewline = \case - Char '\n' : cs -> cs - cs -> cs - - rmLastNewline :: HasChar c => [c] -> [c] - rmLastNewline = - let go = \case - [] -> [] - [Char '\n'] -> [] - c : cs -> c : go cs - in go + go = \case + c@'\n' : cs -> c : go (dropPrefix commonWSPrefix cs) + c : cs -> c : go cs + [] -> [] + + -- drop x characters from the string, or up to a newline, whichever comes first + dropPrefix !x = \case + cs | x <= 0 -> cs + cs@('\n' : _) -> cs + _ : cs -> dropPrefix (x - 1) cs + [] -> [] -- | See step 4 in Note [Multiline string literals] -- @@ -339,6 +436,55 @@ getCommonWsPrefix s = . drop 1 -- ignore first line in calculation $ lines s +-- TODO(bchinn): treat interpolated exprs as one character non-space string +collapseOnlyWsLines :: InterMultiString x -> InterMultiString x +collapseOnlyWsLines = goParts + where + goParts = \case + Right x : parts -> Right x : goParts parts + [Left s] -> _ + Left s : _ -> _ + [] -> [] + + go = \case + c@'\n' : cs | Just cs' <- checkAllWs cs -> c : go cs' + c : cs -> c : go cs + [] -> [] + + -- Return (String, Bool, String), where: + -- * String - The line that was just parsed, ending with a newline character, unless + -- we reached the end of the string + -- * Bool - Whether the line that was just parsed was all whitespace characters + -- * String - The rest of the string + takeLine = + let go acc allWS = \case + -- reached a newline or the end of the string + c@'\n' : cs -> (dlistToList (acc `dlistSnoc` c), allWS, cs) + [] -> (dlistToList acc, allWS, []) + -- found whitespace + c : cs | is_space c -> go (acc `dlistSnoc` c) allWS + -- found non-whitespace + c : cs -> go (acc `dlistSnoc` c) False + in go dlistEmpty True + +rmFirstNewline :: InterMultiString x -> InterMultiString x +rmFirstNewline = \case + InterMultiString ('\n' : pre) parts -> InterMultiString pre parts + s -> s + +rmLastNewline :: InterMultiString x -> InterMultiString x +rmLastNewline (InterMultiString pre parts) = + case unsnoc parts of + Nothing -> + InterMultiString (go pre) parts + Just (parts0, (x, lastLine)) -> + InterMultiString pre (parts0 ++ [(x, go lastLine)]) + where + go = \case + [] -> [] + ['\n'] -> [] + c : cs -> c : go cs + {- Note [Multiline string literals] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -348,8 +494,8 @@ proposal: https://github.com/ghc-proposals/ghc-proposals/pull/569 Multiline string literals are syntax sugar for normal string literals, with an extra post processing step. This all happens in the Lexer; that -is, HsMultilineString will contain the post-processed string. This matches -the same behavior as HsString, which contains the normalized string +is, HsString will contain the post-processed string. This matches the same +behavior as single-line HsString, which contains the normalized string (see Note [Literal source text]). The canonical steps for post processing a multiline string are: @@ -391,3 +537,9 @@ dlistToList (DList f) = f [] dlistSnoc :: DList a -> a -> DList a dlistSnoc (DList f) x = DList (f . (x :)) + +-- ----------------------------------------------------------------------------- +-- Other utilities + +fromRight :: Either e a -> Maybe a +fromRight = either (const Nothing) Just ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -377,8 +377,7 @@ rnExpr (HsLit x lit) | Just (src, s) <- stringLike lit ; return (HsLit x (convertLit lit), emptyFVs) } } where stringLike = \case - HsString src s -> Just (src, s) - HsMultilineString src s -> Just (src, s) + HsString src _ s -> Just (src, s) _ -> Nothing rnExpr (HsLit x lit) @@ -393,6 +392,9 @@ rnExpr (HsOverLit x lit) return (HsApp noExtField (noLocA neg) (noLocA (HsOverLit x lit')) , fvs ) } +rnExpr (HsInterString x ty parts) + = undefined -- TODO(bchinn) + rnExpr (HsApp x fun arg) = do { (fun',fvFun) <- rnLExpr fun ; (arg',fvArg) <- rnLExpr arg ===================================== compiler/GHC/Tc/Gen/HsType.hs ===================================== @@ -4725,8 +4725,7 @@ addTyConFlavCtxt name flav tyLitFromLit :: HsLit GhcRn -> Maybe (HsTyLit GhcRn) -tyLitFromLit (HsString x str) = Just (HsStrTy x str) -tyLitFromLit (HsMultilineString x str) = Just (HsStrTy x str) +tyLitFromLit (HsString x _ str) = Just (HsStrTy x str) tyLitFromLit (HsChar x char) = Just (HsCharTy x char) tyLitFromLit _ = Nothing ===================================== compiler/Language/Haskell/Syntax/Expr.hs ===================================== @@ -350,7 +350,6 @@ data HsExpr p -- erroring expression will be written after -- solving. See Note [Holes] in GHC.Tc.Types.Constraint. - | HsRecSel (XRecSel p) (FieldOcc p) -- ^ Variable pointing to record selector -- See Note [Non-overloaded record field selectors] and @@ -361,12 +360,19 @@ data HsExpr p | HsIPVar (XIPVar p) HsIPName -- ^ Implicit parameter (not in use after typechecking) + | HsOverLit (XOverLitE p) (HsOverLit p) -- ^ Overloaded literals | HsLit (XLitE p) (HsLit p) -- ^ Simple (non-overloaded) literals + | -- | See Note [Interpolated strings] + HsInterString + (XInterString p) + HsStringType + [Either FastString (LHsExpr p)] + -- | Lambda, Lambda-case, and Lambda-cases -- -- - 'GHC.Parser.Annotation.AnnKeywordId' : 'GHC.Parser.Annotation.AnnLam', ===================================== compiler/Language/Haskell/Syntax/Extension.hs ===================================== @@ -413,6 +413,9 @@ type family XOverLabel x type family XIPVar x type family XOverLitE x type family XLitE x +type family XInterString x +type family XInterStringRaw x +type family XInterStringExp x type family XLam x type family XLamCase x type family XApp x @@ -556,7 +559,6 @@ type family XXParStmtBlock x x' type family XHsChar x type family XHsCharPrim x type family XHsString x -type family XHsMultilineString x type family XHsStringPrim x type family XHsInt x type family XHsIntPrim x ===================================== compiler/Language/Haskell/Syntax/Lit.hs ===================================== @@ -51,9 +51,7 @@ data HsLit x -- ^ Character | HsCharPrim (XHsCharPrim x) {- SourceText -} Char -- ^ Unboxed character - | HsString (XHsString x) {- SourceText -} FastString - -- ^ String - | HsMultilineString (XHsMultilineString x) {- SourceText -} FastString + | HsString (XHsString x) {- SourceText -} HsStringType FastString -- ^ String | HsStringPrim (XHsStringPrim x) {- SourceText -} !ByteString -- ^ Packed bytes @@ -145,3 +143,5 @@ instance Ord OverLitVal where compare (HsIsString _ s1) (HsIsString _ s2) = s1 `lexicalCompareFS` s2 compare (HsIsString _ _) (HsIntegral _) = GT compare (HsIsString _ _) (HsFractional _) = GT + +data HsStringType = HsStringTypeSingle | HsStringTypeMulti ===================================== utils/check-exact/ExactPrint.hs ===================================== @@ -4923,8 +4923,7 @@ hsLit2String lit = case lit of HsChar src v -> toSourceTextWithSuffix src v "" HsCharPrim src p -> toSourceTextWithSuffix src p "" - HsString src v -> toSourceTextWithSuffix src v "" - HsMultilineString src v -> toSourceTextWithSuffix src v "" + HsString src _ v -> toSourceTextWithSuffix src v "" HsStringPrim src v -> toSourceTextWithSuffix src v "" HsInt _ (IL src _ v) -> toSourceTextWithSuffix src v "" HsIntPrim src v -> toSourceTextWithSuffix src v "" ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs ===================================== @@ -320,7 +320,6 @@ classify tok = ITlabelvarid{} -> TkUnknown ITchar{} -> TkChar ITstring{} -> TkString - ITstringMulti{} -> TkString ITinteger{} -> TkNumber ITrational{} -> TkNumber ITprimchar{} -> TkChar View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/cc7d07b3b8a0eb678af4eeaed2726c1cc8c79ab3...8e7eae73deb6ddf26024d1fa1d1570c321fb8164 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/cc7d07b3b8a0eb678af4eeaed2726c1cc8c79ab3...8e7eae73deb6ddf26024d1fa1d1570c321fb8164 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 07:56:47 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 03:56:47 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 16 commits: ci: Run abi-test on test-abi label Message-ID: <66f3c23fbd97f_152f2fe2e0f0958cd@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - 9dca0dfa by Simon Peyton Jones at 2024-09-25T03:56:17-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 40878c2d by Brandon Chinn at 2024-09-25T03:56:18-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf - - - - - 5169a025 by Ryan Scott at 2024-09-25T03:56:19-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 2b6ba52f by ARATA Mizuki at 2024-09-25T03:56:22-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 45da2432 by Zubin Duggal at 2024-09-25T03:56:23-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - d4604efa by Sylvain Henry at 2024-09-25T03:56:37-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 3b086110 by Simon Peyton Jones at 2024-09-25T03:56:38-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 27 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ed046c5272e4b73314512ca7e9a0d8c5d3139480...3b086110f4e28cec1189eefa54fa250c579e3eb7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ed046c5272e4b73314512ca7e9a0d8c5d3139480...3b086110f4e28cec1189eefa54fa250c579e3eb7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 08:30:33 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Wed, 25 Sep 2024 04:30:33 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 34 commits: GNU/Hurd: Add getExecutablePath support Message-ID: <66f3ca298156_1cacf514c738900e3@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 61a060eb by sheaf at 2024-09-25T10:29:39+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - b05410b3 by sheaf at 2024-09-25T10:29:39+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 37d16c80 by sheaf at 2024-09-25T10:29:39+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 699457cf by sheaf at 2024-09-25T10:29:40+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - b386e1ca by sheaf at 2024-09-25T10:29:40+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - fac72d54 by sheaf at 2024-09-25T10:29:40+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 73657caf by sheaf at 2024-09-25T10:29:41+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 8a80f932 by sheaf at 2024-09-25T10:29:41+02:00 Add test for C calls & SIMD vectors - - - - - 6798e3b8 by sheaf at 2024-09-25T10:29:41+02:00 Add test for #25169 - - - - - 646c5b63 by sheaf at 2024-09-25T10:30:00+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - 3b41ae5e by sheaf at 2024-09-25T10:30:11+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f0c2ff0a by sheaf at 2024-09-25T10:30:11+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - e8d68fd1 by sheaf at 2024-09-25T10:30:11+02:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 629b174a by sheaf at 2024-09-25T10:30:11+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 3fc52e65 by sheaf at 2024-09-25T10:30:11+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9d0f68ea956702d09115ff8c8353b9d4eee8bc7d...3fc52e65de1dfdb7e57df8494614e2d3d48b01f4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9d0f68ea956702d09115ff8c8353b9d4eee8bc7d...3fc52e65de1dfdb7e57df8494614e2d3d48b01f4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 10:27:41 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 06:27:41 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 7 commits: Add Given injectivity for built-in type families Message-ID: <66f3e59dc7e40_2a75f33cab544186@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: a948ae4f by Simon Peyton Jones at 2024-09-25T06:27:10-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - c615193c by Ryan Scott at 2024-09-25T06:27:11-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 8f238895 by ARATA Mizuki at 2024-09-25T06:27:14-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 2ec0e30e by Zubin Duggal at 2024-09-25T06:27:15-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 3e379148 by Sylvain Henry at 2024-09-25T06:27:28-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - fff2dcdd by Simon Peyton Jones at 2024-09-25T06:27:29-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 48700abd by Sylvain Henry at 2024-09-25T06:27:33-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 10 changed files: - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3b086110f4e28cec1189eefa54fa250c579e3eb7...48700abd54735d941ed4dc130eca15b6b91d610c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3b086110f4e28cec1189eefa54fa250c579e3eb7...48700abd54735d941ed4dc130eca15b6b91d610c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 10:42:44 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 25 Sep 2024 06:42:44 -0400 Subject: [Git][ghc/ghc][wip/enable-late-ccs-release] 2 commits: testsuite: Fix normalisation of prof_files removing newlines Message-ID: <66f3e92473992_30a221bd060204d5@gitlab.mail> Matthew Pickering pushed to branch wip/enable-late-ccs-release at Glasgow Haskell Compiler / GHC Commits: 92e8b1aa by Matthew Pickering at 2024-09-25T11:42:32+01:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 980548bd by Matthew Pickering at 2024-09-25T11:42:32+01:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - 12 changed files: - hadrian/doc/flavours.md - hadrian/src/Flavour.hs - hadrian/src/Settings/Flavours/Release.hs - testsuite/driver/testglobals.py - testsuite/driver/testlib.py - testsuite/tests/codeGen/should_run/cgrun057.stderr - testsuite/tests/profiling/should_run/T12962.prof.sample - testsuite/tests/profiling/should_run/T2552.prof.sample - testsuite/tests/profiling/should_run/T5559.prof.sample - testsuite/tests/profiling/should_run/T5654-O0.prof.sample - testsuite/tests/profiling/should_run/T5654-O1.prof.sample - testsuite/tests/profiling/should_run/T5654b-O0.prof.sample The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ea8b2e7db800a80e2e1da30b4f7445dbd9b5b71d...980548bda4a0a59be38dc9a32957368ee4ef9177 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ea8b2e7db800a80e2e1da30b4f7445dbd9b5b71d...980548bda4a0a59be38dc9a32957368ee4ef9177 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 10:46:58 2024 From: gitlab at gitlab.haskell.org (Teo Camarasu (@teo)) Date: Wed, 25 Sep 2024 06:46:58 -0400 Subject: [Git][ghc/ghc][wip/T25282] Expose list of wired-in units in --info Message-ID: <66f3ea226374_30a221eccfc215bd@gitlab.mail> Teo Camarasu pushed to branch wip/T25282 at Glasgow Haskell Compiler / GHC Commits: 4288440b by Teo Camarasu at 2024-09-25T11:45:32+01:00 Expose list of wired-in units in --info This lets Cabal know which packages are wired-in. In turn Cabal can use this information to determine which packages should be re-installable. Previously this was a fixed list that was hard-coded into Cabal. See: <https://github.com/haskell/cabal/issues/10087> Resolves #25282 - - - - - 1 changed file: - compiler/GHC/Driver/Session.hs Changes: ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -3423,7 +3423,8 @@ compilerInfo dflags ("LibDir", topDir dflags), -- This is always an absolute path, unlike "Relative Global Package DB" which is -- in the settings file. - ("Global Package DB", globalPackageDatabasePath dflags) + ("Global Package DB", globalPackageDatabasePath dflags), + ("Wired-in units", unwords . map unitIdString $ wiredInUnitIds) ] where showBool True = "YES" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4288440b9f8d8c107f469df8d77e36407e13ed1d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4288440b9f8d8c107f469df8d77e36407e13ed1d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 12:19:07 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 25 Sep 2024 08:19:07 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/trigger-ghcup-ci Message-ID: <66f3ffbbdf998_3b75311a7cdc6207b@gitlab.mail> Matthew Pickering pushed new branch wip/trigger-ghcup-ci at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/trigger-ghcup-ci You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 12:20:01 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Wed, 25 Sep 2024 08:20:01 -0400 Subject: [Git][ghc/ghc][wip/andreask/spec-constr-args] SpecConstr: Introduce a separate argument limit for forced specs. Message-ID: <66f3fff14f28a_3b75311a764c639a3@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/spec-constr-args at Glasgow Haskell Compiler / GHC Commits: f71122d2 by Andreas Klebinger at 2024-09-25T14:00:29+02:00 SpecConstr: Introduce a separate argument limit for forced specs. We used to put no limit at all on specializations forced via the SPEC argument. This isn't always reasonable so we introduce a very high limit that applies to forced specializations, a flag to control it, and we now emit a warning if we fail a specialization because we exceed the warning. Fixes #25197 - - - - - 8 changed files: - compiler/GHC/Core/Opt/SpecConstr.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Session.hs - docs/users_guide/using-optimisation.rst - + testsuite/tests/simplCore/should_compile/T25197.hs - + testsuite/tests/simplCore/should_compile/T25197.stderr - + testsuite/tests/simplCore/should_compile/T25197_TH.hs - testsuite/tests/simplCore/should_compile/all.T Changes: ===================================== compiler/GHC/Core/Opt/SpecConstr.hs ===================================== @@ -20,7 +20,8 @@ ToDo [Oct 2013] module GHC.Core.Opt.SpecConstr( specConstrProgram, - SpecConstrAnnotation(..) + SpecConstrAnnotation(..), + SpecFailWarning(..) ) where import GHC.Prelude @@ -51,6 +52,7 @@ import GHC.Core.Make ( mkImpossibleExpr ) import GHC.Unit.Module import GHC.Unit.Module.ModGuts +import GHC.Types.Error (MessageClass(..), Severity(..), DiagnosticReason(WarningWithoutFlag), ResolvedDiagnosticReason (..)) import GHC.Types.Literal ( litIsLifted ) import GHC.Types.Id import GHC.Types.Id.Info ( IdDetails(..) ) @@ -526,9 +528,11 @@ sc_force to True when calling specLoop. This flag does four things: (see argToPat; #4448) (FS4) Only specialise on recursive types a finite number of times (see sc_recursive; #5550; Note [Limit recursive specialisation]) -(FS5) Lift the restriction on the maximum number of arguments which - the optimisation will specialise. - (see `too_many_worker_args` in `callsToNewPats`; #14003) +(FS5) Use a different restriction on the maximum number of arguments which + the optimisation will specialise. We tried removing the limit on worker + args for forced specs (#14003) but this caused issues when specializing + code for large data structures (#25197). + This is handled by `too_many_worker_args` in `callsToNewPats` The flag holds only for specialising a single binding group, and NOT for nested bindings. (So really it should be passed around explicitly @@ -782,16 +786,25 @@ specConstrProgram :: ModGuts -> CoreM ModGuts specConstrProgram guts = do { env0 <- initScEnv guts ; us <- getUniqueSupplyM - ; let (_usg, binds') = initUs_ us $ + ; let (_usg, binds', warnings) = initUs_ us $ scTopBinds env0 (mg_binds guts) + ; when (not (null warnings)) $ msg specConstr_warn_class (warn_msg warnings) + ; return (guts { mg_binds = binds' }) } -scTopBinds :: ScEnv -> [InBind] -> UniqSM (ScUsage, [OutBind]) -scTopBinds _env [] = return (nullUsage, []) -scTopBinds env (b:bs) = do { (usg, b', bs') <- scBind TopLevel env b $ + where + specConstr_warn_class = MCDiagnostic SevWarning (ResolvedDiagnosticReason WarningWithoutFlag) Nothing + warn_msg :: SpecFailWarnings -> SDoc + warn_msg warnings = text "SpecConstr encountered one or more function(s) with a SPEC argument that resulted in too many arguments," $$ + text "which resulted in no specialization being generated for these functions:" $$ + nest 2 (vcat (map ppr warnings)) $$ + (text "If this is expected you might want to increase -fmax-forced-spec-args to force specialization anyway.") +scTopBinds :: ScEnv -> [InBind] -> UniqSM (ScUsage, [OutBind], [SpecFailWarning]) +scTopBinds _env [] = return (nullUsage, [], []) +scTopBinds env (b:bs) = do { (usg, b', bs', warnings) <- scBind TopLevel env b $ (\env -> scTopBinds env bs) - ; return (usg, b' ++ bs') } + ; return (usg, b' ++ bs', warnings) } {- ************************************************************************ @@ -905,6 +918,12 @@ data SpecConstrOpts = SpecConstrOpts -- ^ The threshold at which a worker-wrapper transformation used as part of -- this pass will no longer happen, measured in the number of arguments. + , sc_max_forced_args :: !Int + -- ^ The threshold at which a worker-wrapper transformation used as part of + -- this pass will no longer happen even if a SPEC arg was used to force + -- specialization. Measured in the number of arguments. + -- See Note [Forcing specialisation] + , sc_debug :: !Bool -- ^ Whether to print debug information @@ -975,6 +994,7 @@ instance Outputable Value where initScOpts :: DynFlags -> Module -> SpecConstrOpts initScOpts dflags this_mod = SpecConstrOpts { sc_max_args = maxWorkerArgs dflags, + sc_max_forced_args = maxForcedSpecArgs dflags, sc_debug = hasPprDebug dflags, sc_uf_opts = unfoldingOpts dflags, sc_module = this_mod, @@ -1388,29 +1408,29 @@ creates specialised versions of functions. -} scBind :: TopLevelFlag -> ScEnv -> InBind - -> (ScEnv -> UniqSM (ScUsage, a)) -- Specialise the scope of the binding - -> UniqSM (ScUsage, [OutBind], a) + -> (ScEnv -> UniqSM (ScUsage, a, [SpecFailWarning])) -- Specialise the scope of the binding + -> UniqSM (ScUsage, [OutBind], a, [SpecFailWarning]) scBind top_lvl env (NonRec bndr rhs) do_body | isTyVar bndr -- Type-lets may be created by doBeta - = do { (final_usage, body') <- do_body (extendScSubst env bndr rhs) - ; return (final_usage, [], body') } + = do { (final_usage, body', warnings) <- do_body (extendScSubst env bndr rhs) + ; return (final_usage, [], body', warnings) } | not (isTopLevel top_lvl) -- Nested non-recursive value binding -- See Note [Specialising local let bindings] = do { let (body_env, bndr') = extendBndr env bndr -- Not necessary at top level; but here we are nested - ; rhs_info <- scRecRhs env (bndr',rhs) + ; (rhs_info, rhs_ws) <- scRecRhs env (bndr',rhs) ; let body_env2 = extendHowBound body_env [bndr'] RecFun rhs' = ri_new_rhs rhs_info body_env3 = extendValEnv body_env2 bndr' (isValue (sc_vals env) rhs') - ; (body_usg, body') <- do_body body_env3 + ; (body_usg, body', warnings_body) <- do_body body_env3 -- Now make specialised copies of the binding, -- based on calls in body_usg - ; (spec_usg, specs) <- specNonRec env (scu_calls body_usg) rhs_info + ; (spec_usg, specs, warnings_bnd) <- specNonRec env (scu_calls body_usg) rhs_info -- NB: For non-recursive bindings we inherit sc_force flag from -- the parent function (see Note [Forcing specialisation]) @@ -1419,7 +1439,7 @@ scBind top_lvl env (NonRec bndr rhs) do_body bind_usage = (body_usg `delCallsFor` [bndr']) `combineUsage` spec_usg -- Note [spec_usg includes rhs_usg] - ; return (bind_usage, spec_bnds, body') + ; return (bind_usage, spec_bnds, body', mconcat [warnings_bnd, warnings_body, rhs_ws]) } | otherwise -- Top-level, non-recursive value binding @@ -1431,15 +1451,15 @@ scBind top_lvl env (NonRec bndr rhs) do_body -- -- I tried always specialising non-recursive top-level bindings too, -- but found some regressions (see !8135). So I backed off. - = do { (rhs_usage, rhs') <- scExpr env rhs + = do { (rhs_usage, rhs', ws_rhs) <- scExpr env rhs -- At top level, we've already put all binders into scope; see initScEnv -- Hence no need to call `extendBndr`. But we still want to -- extend the `ValueEnv` to record the value of this binder. ; let body_env = extendValEnv env bndr (isValue (sc_vals env) rhs') - ; (body_usage, body') <- do_body body_env + ; (body_usage, body', body_warnings) <- do_body body_env - ; return (rhs_usage `combineUsage` body_usage, [NonRec bndr rhs'], body') } + ; return (rhs_usage `combineUsage` body_usage, [NonRec bndr rhs'], body', body_warnings ++ ws_rhs) } scBind top_lvl env (Rec prs) do_body | isTopLevel top_lvl @@ -1450,19 +1470,20 @@ scBind top_lvl env (Rec prs) do_body -- ToDo: I'm honestly not sure of the rationale of this size-testing, nor -- why it only applies at top level. But that's the way it has been -- for a while. See #21456. - do { (body_usg, body') <- do_body rhs_env2 - ; (rhs_usgs, rhss') <- mapAndUnzipM (scExpr env) rhss + do { (body_usg, body', warnings_body) <- do_body rhs_env2 + ; (rhs_usgs, rhss', rhs_ws) <- mapAndUnzip3M (scExpr env) rhss ; let all_usg = (combineUsages rhs_usgs `combineUsage` body_usg) `delCallsFor` bndrs' bind' = Rec (bndrs' `zip` rhss') - ; return (all_usg, [bind'], body') } + ; return (all_usg, [bind'], body', warnings_body ++ concat rhs_ws) } | otherwise - = do { rhs_infos <- mapM (scRecRhs rhs_env2) (bndrs' `zip` rhss) - ; (body_usg, body') <- do_body rhs_env2 + = do { (rhs_infos, rhs_wss) <- mapAndUnzipM (scRecRhs rhs_env2) (bndrs' `zip` rhss) + ; let rhs_ws = mconcat rhs_wss + ; (body_usg, body', warnings_body) <- do_body rhs_env2 - ; (spec_usg, specs) <- specRec (scForce rhs_env2 force_spec) - (scu_calls body_usg) rhs_infos + ; (spec_usg, specs, spec_ws) <- specRec (scForce rhs_env2 force_spec) + (scu_calls body_usg) rhs_infos -- Do not unconditionally generate specialisations from rhs_usgs -- Instead use them only if we find an unspecialised call -- See Note [Seeding recursive groups] @@ -1473,7 +1494,7 @@ scBind top_lvl env (Rec prs) do_body -- zipWithEqual: length of returned [SpecInfo] -- should be the same as incoming [RhsInfo] - ; return (all_usg, [bind'], body') } + ; return (all_usg, [bind'], body', mconcat [warnings_body,rhs_ws,spec_ws]) } where (bndrs,rhss) = unzip prs force_spec = any (forceSpecBndr env) bndrs -- Note [Forcing specialisation] @@ -1501,59 +1522,63 @@ recursive function, but that's not essential and might even be harmful. I'm not sure. -} +withWarnings :: SpecFailWarnings -> (ScUsage, CoreExpr, SpecFailWarnings) -> (ScUsage, CoreExpr, SpecFailWarnings) +withWarnings ws (use,expr,ws2) = (use,expr,ws ++ ws2) + ------------------------ -scExpr, scExpr' :: ScEnv -> CoreExpr -> UniqSM (ScUsage, CoreExpr) +scExpr, scExpr' :: ScEnv -> CoreExpr -> UniqSM (ScUsage, CoreExpr, SpecFailWarnings) -- The unique supply is needed when we invent -- a new name for the specialised function and its args scExpr env e = scExpr' env e scExpr' env (Var v) = case scSubstId env v of - Var v' -> return (mkVarUsage env v' [], Var v') + Var v' -> return (mkVarUsage env v' [], Var v', []) e' -> scExpr (zapScSubst env) e' scExpr' env (Type t) = let !(MkSolo ty') = scSubstTy env t - in return (nullUsage, Type ty') -scExpr' env (Coercion c) = return (nullUsage, Coercion (scSubstCo env c)) -scExpr' _ e@(Lit {}) = return (nullUsage, e) -scExpr' env (Tick t e) = do (usg, e') <- scExpr env e - return (usg, Tick (scTickish env t) e') -scExpr' env (Cast e co) = do (usg, e') <- scExpr env e - return (usg, mkCast e' (scSubstCo env co)) + in return (nullUsage, Type ty', []) +scExpr' env (Coercion c) = return (nullUsage, Coercion (scSubstCo env c), []) +scExpr' _ e@(Lit {}) = return (nullUsage, e, []) +scExpr' env (Tick t e) = do (usg, e', ws) <- scExpr env e + return (usg, Tick (scTickish env t) e', ws) +scExpr' env (Cast e co) = do (usg, e', ws) <- scExpr env e + return (usg, mkCast e' (scSubstCo env co), ws) -- Important to use mkCast here -- See Note [SpecConstr call patterns] scExpr' env e@(App _ _) = scApp env (collectArgs e) scExpr' env (Lam b e) = do let (env', b') = extendBndr env b - (usg, e') <- scExpr env' e - return (usg, Lam b' e') + (usg, e', ws) <- scExpr env' e + return (usg, Lam b' e', ws) scExpr' env (Let bind body) - = do { (final_usage, binds', body') <- scBind NotTopLevel env bind $ + = do { (final_usage, binds', body', ws) <- scBind NotTopLevel env bind $ (\env -> scExpr env body) - ; return (final_usage, mkLets binds' body') } + ; return (final_usage, mkLets binds' body', ws) } scExpr' env (Case scrut b ty alts) - = do { (scrut_usg, scrut') <- scExpr env scrut + = do { (scrut_usg, scrut', ws) <- scExpr env scrut ; case isValue (sc_vals env) scrut' of Just (ConVal args_are_work_free con args) - | args_are_work_free -> sc_con_app con args scrut' + | args_are_work_free -> sc_con_app con args scrut' ws -- Don't duplicate work!! #7865 -- See Note [ConVal work-free-ness] (1) - _other -> sc_vanilla scrut_usg scrut' + _other -> sc_vanilla scrut_usg scrut' ws } where - sc_con_app con args scrut' -- Known constructor; simplify + sc_con_app con args scrut' ws -- Known constructor; simplify = do { let Alt _ bs rhs = findAlt con alts `orElse` Alt DEFAULT [] (mkImpossibleExpr ty "SpecConstr") alt_env' = extendScSubstList env ((b,scrut') : bs `zip` trimConArgs con args) - ; scExpr alt_env' rhs } + ; (use',expr',ws_new) <- scExpr alt_env' rhs + ; return (use',expr',ws ++ ws_new) } - sc_vanilla scrut_usg scrut' -- Normal case + sc_vanilla scrut_usg scrut' ws -- Normal case = do { let (alt_env,b') = extendBndrWith RecArg env b -- Record RecArg for the components - ; (alt_usgs, alt_occs, alts') <- mapAndUnzip3M (sc_alt alt_env scrut' b') alts + ; (alt_usgs, alt_occs, alts', ws_alts) <- mapAndUnzip4M (sc_alt alt_env scrut' b') alts ; let scrut_occ = foldr combineOcc NoOcc alt_occs scrut_usg' = setScrutOcc env scrut_usg scrut' scrut_occ @@ -1563,21 +1588,21 @@ scExpr' env (Case scrut b ty alts) ; let !(MkSolo ty') = scSubstTy env ty ; return (foldr combineUsage scrut_usg' alt_usgs, - Case scrut' b' ty' alts') } + Case scrut' b' ty' alts', ws ++ concat ws_alts) } single_alt = isSingleton alts sc_alt env scrut' b' (Alt con bs rhs) = do { let (env1, bs1) = extendBndrsWith RecArg env bs (env2, bs2) = extendCaseBndrs env1 scrut' b' con bs1 - ; (usg, rhs') <- scExpr env2 rhs + ; (usg, rhs', ws) <- scExpr env2 rhs ; let (usg', b_occ:arg_occs) = lookupOccs usg (b':bs2) scrut_occ = case con of DataAlt dc -- See Note [Do not specialise evals] | not (single_alt && all deadArgOcc arg_occs) -> ScrutOcc (unitUFM dc arg_occs) _ -> UnkOcc - ; return (usg', b_occ `combineOcc` scrut_occ, Alt con bs2 rhs') } + ; return (usg', b_occ `combineOcc` scrut_occ, Alt con bs2 rhs', ws) } -- | Substitute the free variables captured by a breakpoint. @@ -1626,19 +1651,20 @@ follows. still worth specialising on x. Hence the /single-alternative/ guard. -} -scApp :: ScEnv -> (InExpr, [InExpr]) -> UniqSM (ScUsage, CoreExpr) +scApp :: ScEnv -> (InExpr, [InExpr]) -> UniqSM (ScUsage, CoreExpr, SpecFailWarnings) scApp env (Var fn, args) -- Function is a variable = assert (not (null args)) $ do { args_w_usgs <- mapM (scExpr env) args - ; let (arg_usgs, args') = unzip args_w_usgs + ; let (arg_usgs, args', arg_ws) = unzip3 args_w_usgs arg_usg = combineUsages arg_usgs + arg_w = concat arg_ws ; case scSubstId env fn of - fn'@(Lam {}) -> scExpr (zapScSubst env) (doBeta fn' args') + fn'@(Lam {}) -> withWarnings arg_w <$> scExpr (zapScSubst env) (doBeta fn' args') -- Do beta-reduction and try again Var fn' -> return (arg_usg' `combineUsage` mkVarUsage env fn' args', - mkApps (Var fn') args') + mkApps (Var fn') args', arg_w ) where -- arg_usg': see Note [Specialising on dictionaries] arg_usg' | Just cls <- isClassOpId_maybe fn' @@ -1647,7 +1673,7 @@ scApp env (Var fn, args) -- Function is a variable | otherwise = arg_usg - other_fn' -> return (arg_usg, mkApps other_fn' args') } + other_fn' -> return (arg_usg, mkApps other_fn' args', arg_w) } -- NB: doing this ignores any usage info from the substituted -- function, but I don't think that matters. If it does -- we can fix it. @@ -1661,9 +1687,9 @@ scApp env (Var fn, args) -- Function is a variable -- which it may, we can get -- (let f = ...f... in f) arg1 arg2 scApp env (other_fn, args) - = do { (fn_usg, fn') <- scExpr env other_fn - ; (arg_usgs, args') <- mapAndUnzipM (scExpr env) args - ; return (combineUsages arg_usgs `combineUsage` fn_usg, mkApps fn' args') } + = do { (fn_usg, fn', fn_ws) <- scExpr env other_fn + ; (arg_usgs, args', arg_ws) <- mapAndUnzip3M (scExpr env) args + ; return (combineUsages arg_usgs `combineUsage` fn_usg, mkApps fn' args', combineSpecWarning fn_ws (concat arg_ws)) } ---------------------- mkVarUsage :: ScEnv -> Id -> [CoreExpr] -> ScUsage @@ -1679,16 +1705,16 @@ mkVarUsage env fn args | otherwise = evalScrutOcc ---------------------- -scRecRhs :: ScEnv -> (OutId, InExpr) -> UniqSM RhsInfo +scRecRhs :: ScEnv -> (OutId, InExpr) -> UniqSM (RhsInfo, SpecFailWarnings) scRecRhs env (bndr,rhs) = do { let (arg_bndrs,body) = collectBinders rhs (body_env, arg_bndrs') = extendBndrsWith RecArg env arg_bndrs - ; (body_usg, body') <- scExpr body_env body + ; (body_usg, body', body_ws) <- scExpr body_env body ; let (rhs_usg, arg_occs) = lookupOccs body_usg arg_bndrs' ; return (RI { ri_rhs_usg = rhs_usg , ri_fn = bndr, ri_new_rhs = mkLams arg_bndrs' body' , ri_lam_bndrs = arg_bndrs, ri_lam_body = body - , ri_arg_occs = arg_occs }) } + , ri_arg_occs = arg_occs }, body_ws) } -- The arg_occs says how the visible, -- lambda-bound binders of the RHS are used -- (including the TyVar binders) @@ -1757,7 +1783,7 @@ initSpecInfo (RI { ri_rhs_usg = rhs_usg }) specNonRec :: ScEnv -> CallEnv -- Calls in body -> RhsInfo -- Structure info usage info for un-specialised RHS - -> UniqSM (ScUsage, SpecInfo) -- Usage from RHSs (specialised and not) + -> UniqSM (ScUsage, SpecInfo, [SpecFailWarning]) -- Usage from RHSs (specialised and not) -- plus details of specialisations specNonRec env body_calls rhs_info @@ -1767,11 +1793,12 @@ specNonRec env body_calls rhs_info specRec :: ScEnv -> CallEnv -- Calls in body -> [RhsInfo] -- Structure info and usage info for un-specialised RHSs - -> UniqSM (ScUsage, [SpecInfo]) -- Usage from all RHSs (specialised and not) + -> UniqSM (ScUsage, [SpecInfo], SpecFailWarnings) + -- Usage from all RHSs (specialised and not) -- plus details of specialisations specRec env body_calls rhs_infos - = go 1 body_calls nullUsage (map initSpecInfo rhs_infos) + = go 1 body_calls nullUsage (map initSpecInfo rhs_infos) [] -- body_calls: see Note [Seeding recursive groups] -- NB: 'go' always calls 'specialise' once, which in turn unleashes -- si_mb_unspec if there are any boring calls in body_calls, @@ -1786,23 +1813,25 @@ specRec env body_calls rhs_infos -- Two accumulating parameters: -> ScUsage -- Usage from earlier specialisations -> [SpecInfo] -- Details of specialisations so far - -> UniqSM (ScUsage, [SpecInfo]) - go n_iter seed_calls usg_so_far spec_infos + -> SpecFailWarnings -- Warnings so far + -> UniqSM (ScUsage, [SpecInfo], SpecFailWarnings) + go n_iter seed_calls usg_so_far spec_infos ws_so_far = -- pprTrace "specRec3" (vcat [ text "bndrs" <+> ppr (map ri_fn rhs_infos) -- , text "iteration" <+> int n_iter -- , text "spec_infos" <+> ppr (map (map os_pat . si_specs) spec_infos) -- ]) $ do { specs_w_usg <- zipWithM (specialise env seed_calls) rhs_infos spec_infos - ; let (extra_usg_s, all_spec_infos) = unzip specs_w_usg + + ; let (extra_usg_s, all_spec_infos, extra_ws ) = unzip3 specs_w_usg extra_usg = combineUsages extra_usg_s all_usg = usg_so_far `combineUsage` extra_usg new_calls = scu_calls extra_usg - ; go_again n_iter new_calls all_usg all_spec_infos } + ; go_again n_iter new_calls all_usg all_spec_infos (ws_so_far ++ concat extra_ws) } -- go_again deals with termination - go_again n_iter seed_calls usg_so_far spec_infos + go_again n_iter seed_calls usg_so_far spec_infos ws_so_far | isEmptyVarEnv seed_calls - = return (usg_so_far, spec_infos) + = return (usg_so_far, spec_infos, ws_so_far) -- Limit recursive specialisation -- See Note [Limit recursive specialisation] @@ -1816,10 +1845,10 @@ specRec env body_calls rhs_infos -- for the unspecialised function, since it may now be called -- pprTrace "specRec2" (ppr (map (map os_pat . si_specs) spec_infos)) $ let rhs_usgs = combineUsages (mapMaybe si_mb_unspec spec_infos) - in return (usg_so_far `combineUsage` rhs_usgs, spec_infos) + in return (usg_so_far `combineUsage` rhs_usgs, spec_infos, ws_so_far) | otherwise - = go (n_iter + 1) seed_calls usg_so_far spec_infos + = go (n_iter + 1) seed_calls usg_so_far spec_infos ws_so_far -- See Note [Limit recursive specialisation] the_limit = case sc_count opts of @@ -1832,7 +1861,7 @@ specialise -> CallEnv -- Info on newly-discovered calls to this function -> RhsInfo -> SpecInfo -- Original RHS plus patterns dealt with - -> UniqSM (ScUsage, SpecInfo) -- New specialised versions and their usage + -> UniqSM (ScUsage, SpecInfo, [SpecFailWarning]) -- New specialised versions and their usage -- See Note [spec_usg includes rhs_usg] @@ -1850,7 +1879,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs | isDeadEndId fn -- Note [Do not specialise diverging functions] -- /and/ do not generate specialisation seeds from its RHS = -- pprTrace "specialise bot" (ppr fn) $ - return (nullUsage, spec_info) + return (nullUsage, spec_info, []) | not (isNeverActive (idInlineActivation fn)) -- See Note [Transfer activation] @@ -1861,7 +1890,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs , not (null arg_bndrs) -- Only specialise functions , Just all_calls <- lookupVarEnv bind_calls fn -- Some calls to it = -- pprTrace "specialise entry {" (ppr fn <+> ppr all_calls) $ - do { (boring_call, pats_discarded, new_pats) + do { (boring_call, pats_discarded, new_pats, warnings) <- callsToNewPats env fn spec_info arg_occs all_calls ; let n_pats = length new_pats @@ -1876,7 +1905,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs -- , text "new_pats" <+> ppr new_pats]) ; let spec_env = decreaseSpecCount env n_pats - ; (spec_usgs, new_specs) <- mapAndUnzipM (spec_one spec_env fn arg_bndrs body) + ; (spec_usgs, new_specs, new_wss) <- mapAndUnzip3M (spec_one spec_env fn arg_bndrs body) (new_pats `zip` [spec_count..]) -- See Note [Specialise original body] @@ -1900,15 +1929,16 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs ; return (new_usg, SI { si_specs = new_specs ++ specs , si_n_specs = spec_count + n_pats - , si_mb_unspec = mb_unspec' }) } + , si_mb_unspec = mb_unspec' } + ,warnings ++ concat new_wss) } | otherwise -- No calls, inactive, or not a function -- Behave as if there was a single, boring call = -- pprTrace "specialise inactive" (ppr fn $$ ppr mb_unspec) $ case mb_unspec of -- Behave as if there was a single, boring call - Just rhs_usg -> return (rhs_usg, spec_info { si_mb_unspec = Nothing }) + Just rhs_usg -> return (rhs_usg, spec_info { si_mb_unspec = Nothing }, []) -- See Note [spec_usg includes rhs_usg] - Nothing -> return (nullUsage, spec_info) + Nothing -> return (nullUsage, spec_info, []) --------------------- @@ -1917,7 +1947,7 @@ spec_one :: ScEnv -> [InVar] -- Lambda-binders of RHS; should match patterns -> InExpr -- Body of the original function -> (CallPat, Int) - -> UniqSM (ScUsage, OneSpec) -- Rule and binding + -> UniqSM (ScUsage, OneSpec, SpecFailWarnings) -- Rule and binding, warnings if any -- spec_one creates a specialised copy of the function, together -- with a rule for using it. I'm very proud of how short this @@ -1969,7 +1999,7 @@ spec_one env fn arg_bndrs body (call_pat, rule_number) -- Specialise the body -- ; pprTraceM "body_subst_for" $ ppr (spec_occ) $$ ppr (sc_subst body_env) - ; (spec_usg, spec_body) <- scExpr body_env body + ; (spec_usg, spec_body, body_warnings) <- scExpr body_env body -- And build the results ; (qvars', pats') <- generaliseDictPats qvars pats @@ -2018,7 +2048,7 @@ spec_one env fn arg_bndrs body (call_pat, rule_number) -- ] ; return (spec_usg, OS { os_pat = call_pat, os_rule = rule , os_id = spec_id - , os_rhs = spec_rhs }) } + , os_rhs = spec_rhs }, body_warnings) } generaliseDictPats :: [Var] -> [CoreExpr] -- Quantified vars and pats -> UniqSM ([Var], [CoreExpr]) -- New quantified vars and pats @@ -2402,12 +2432,26 @@ instance Outputable CallPat where , text "cp_args =" <+> ppr args , text "cp_strict_args = " <> ppr strict ]) +newtype SpecFailWarning = SpecFailForcedArgCount { spec_failed_fun_name :: Name } + +type SpecFailWarnings = [SpecFailWarning] + +instance Outputable SpecFailWarning where + ppr (SpecFailForcedArgCount name) = ppr name <+> pprDefinedAt name + +combineSpecWarning :: SpecFailWarnings -> SpecFailWarnings -> SpecFailWarnings +combineSpecWarning = (++) + +data ArgCountResult = WorkerSmallEnough | WorkerTooLarge | WorkerTooLargeForced Name + callsToNewPats :: ScEnv -> Id -> SpecInfo -> [ArgOcc] -> [Call] -> UniqSM ( Bool -- At least one boring call , Bool -- Patterns were discarded - , [CallPat] ) -- Patterns to specialise + , [CallPat] -- Patterns to specialise + , [SpecFailWarning] -- Things that didn't specialise we want to warn the user about) + ) -- Result has no duplicate patterns, -- nor ones mentioned in si_specs (hence "new" patterns) -- Bool indicates that there was at least one boring pattern @@ -2433,12 +2477,18 @@ callsToNewPats env fn spec_info@(SI { si_specs = done_specs }) bndr_occs calls non_dups = subsumePats in_scope new_pats -- Remove ones that have too many worker variables - small_pats = filterOut too_many_worker_args non_dups + (small_pats, arg_count_warnings) = partitionByWorkerSize too_many_worker_args non_dups - too_many_worker_args _ - | sc_force env = False -- See (FS5) of Note [Forcing specialisation] + -- too_many_worker_args :: CallPat -> Either SpecFailWarning Bool too_many_worker_args (CP { cp_qvars = vars, cp_args = args }) - = not (isWorkerSmallEnough (sc_max_args $ sc_opts env) (valArgCount args) vars) + | sc_force env + -- See (FS5) of Note [Forcing specialisation] + = if (isWorkerSmallEnough (sc_max_forced_args $ sc_opts env) (valArgCount args) vars) + then WorkerSmallEnough + else WorkerTooLargeForced (idName fn) + | (isWorkerSmallEnough (sc_max_args $ sc_opts env) (valArgCount args) vars) + = WorkerSmallEnough + | otherwise = WorkerTooLarge -- We are about to construct w/w pair in 'spec_one'. -- Omit specialisation leading to high arity workers. -- See Note [Limit w/w arity] in GHC.Core.Opt.WorkWrap.Utils @@ -2454,10 +2504,21 @@ callsToNewPats env fn spec_info@(SI { si_specs = done_specs }) bndr_occs calls -- , text "done_specs:" <+> ppr (map os_pat done_specs) -- , text "trimmed_pats:" <+> ppr trimmed_pats ]) - ; return (have_boring_call, pats_were_discarded, trimmed_pats) } + ; return (have_boring_call, pats_were_discarded, trimmed_pats, arg_count_warnings) } -- If any of the calls does not give rise to a specialisation, either -- because it is boring, or because there are too many specialisations, -- return a flag to say so, so that we know to keep the original function. + where + partitionByWorkerSize worker_size pats = go pats [] [] + where + go [] small warnings = (small, warnings) + go (p:ps) small warnings + | WorkerSmallEnough <- worker_size p + = go ps (p:small) warnings + | WorkerTooLarge <- worker_size p + = go ps small warnings + | WorkerTooLargeForced name <- worker_size p + = go ps small (SpecFailForcedArgCount name : warnings) trim_pats :: ScEnv -> Id -> SpecInfo -> [CallPat] -> (Bool, [CallPat]) ===================================== compiler/GHC/Driver/DynFlags.hs ===================================== @@ -394,6 +394,7 @@ data DynFlags = DynFlags { unfoldingOpts :: !UnfoldingOpts, maxWorkerArgs :: Int, + maxForcedSpecArgs :: Int, ghciHistSize :: Int, @@ -676,6 +677,8 @@ defaultDynFlags mySettings = unfoldingOpts = defaultUnfoldingOpts, maxWorkerArgs = 10, + maxForcedSpecArgs = 333, + -- 333 is fairly arbitrary, see Note [Forcing specialisation]:FS5 ghciHistSize = 50, -- keep a log of length 50 by default ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -1820,6 +1820,8 @@ dynamic_flags_deps = [ , make_ord_flag defFlag "fmax-worker-args" (intSuffix (\n d -> d {maxWorkerArgs = n})) + , make_ord_flag defFlag "fmax-forced-spec-args" + (intSuffix (\n d -> d {maxForcedSpecArgs = n})) , make_ord_flag defGhciFlag "fghci-hist-size" (intSuffix (\n d -> d {ghciHistSize = n})) , make_ord_flag defGhcFlag "fmax-inline-alloc-size" ===================================== docs/users_guide/using-optimisation.rst ===================================== @@ -870,6 +870,21 @@ as such you shouldn't need to set any of them explicitly. A flag value arguments of the resulting worker exceeds both that of the original function and this setting. +.. ghc-flag:: -fmax-forced-spec-args=⟨n⟩ + :shortdesc: *default: 333.* Maximum number of value arguments for forced SpecConstr specializations. + :type: dynamic + :category: + + :default: 512 + + When using ``SPEC`` from ``GHC.Types`` to force SpecConstr to fire on a function + sometimes this can result in functions taking a ridicolously large number of arguments + resulting a very large compile time hits for minor performance benefits. + + Since this is usually unintended we prevent SpecConstr from firing and generate + a warning if the number of arguments in the resulting function would exceed + the value given by ``-fmax-forced-spec-args``. + .. ghc-flag:: -fno-opt-coercion :shortdesc: Turn off the coercion optimiser :type: dynamic ===================================== testsuite/tests/simplCore/should_compile/T25197.hs ===================================== @@ -0,0 +1,30 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25197 where + +import T25197_TH +import GHC.Exts + +{- +This test applies a large statically known data structure to a function with +a SPEC argument, forcing the function to be specialised for the argument. +However when the complete structure of the argument is not statically known, +or as here the leaves of the structures are primitive literals for which we do +not specialize this results in a specialized function that can take hundreds of +arguments. + +Typically this is not intended, therefore we use a limit on the number of +arguments for specializations. As at some point this sort of specialization +comes with a heavy compile time cost. However we allow users to specify this +limit just in case they really depend on this sort of specialization. +-} + +foo :: [a] -> Int +foo = go SPEC + where + go :: SPEC -> [a] -> Int + go s [] = s `seq` 0 + go s (_:xs) = 1 + go s xs + +main :: IO () +main = print $ foo $(gen 1000) ===================================== testsuite/tests/simplCore/should_compile/T25197.stderr ===================================== @@ -0,0 +1,8 @@ +[1 of 2] Compiling T25197_TH ( T25197_TH.hs, T25197_TH.o ) +[2 of 2] Compiling T25197 ( T25197.hs, T25197.o ) +T25197.hs: warning: + SpecConstr encountered one or more function(s) with a SPEC argument that resulted in too many arguments, + which resulted in no specialization being generated for these functions: + $wgo Defined at T25197.hs:26:5 + If this is expected you might want to increase -fmax-forced-spec-args to force specialization anyway. + ===================================== testsuite/tests/simplCore/should_compile/T25197_TH.hs ===================================== @@ -0,0 +1,9 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25197_TH where + +import Language.Haskell.TH.Syntax + +gen :: Int -> Q Exp +gen 0 = [| [] |] +gen n = [| $(lift (show n)) : $(gen (n-1)) |] ===================================== testsuite/tests/simplCore/should_compile/all.T ===================================== @@ -530,3 +530,4 @@ test('T24625', [ grep_errmsg(r'case lazy') ], compile, ['-O -fno-ignore-asserts test('T24725a', [ grep_errmsg(r'testedRule')], compile, ['-O -ddump-rule-firings']) test('T25033', normal, compile, ['-O']) test('T25160', normal, compile, ['-O -ddump-rules']) +test('T25197', [req_th, extra_files(["T25197_TH.hs"]), only_ways(['optasm'])], multimod_compile, ['T25197', '-O2']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f71122d20841e4a1c536771e433abaa8d2602914 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f71122d20841e4a1c536771e433abaa8d2602914 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 12:31:52 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Wed, 25 Sep 2024 08:31:52 -0400 Subject: [Git][ghc/ghc][wip/T24824] 2 commits: Pmc: Improve implementation of -Wincomplete-record-selectors (#24824, #24891) Message-ID: <66f402b8c2e00_3b75314acbf872726@gitlab.mail> Sebastian Graf pushed to branch wip/T24824 at Glasgow Haskell Compiler / GHC Commits: 2c9a685c by Sebastian Graf at 2024-09-25T14:31:36+02:00 Pmc: Improve implementation of -Wincomplete-record-selectors (#24824, #24891) We now incorporate the result type of unsaturated record selector applications as well as consider long-distance information in getField applications. See the updated Note [Detecting incomplete record selectors]. Fixes #24824 and #24891. - - - - - 7cf5a9a2 by Sebastian Graf at 2024-09-25T14:31:36+02:00 Pmc: Improve warning messages of -Wincomplete-record-selectors ... as suggested by Adam Gundry in !12685. - - - - - 21 changed files: - compiler/GHC/HsToCore/Errors/Ppr.hs - compiler/GHC/HsToCore/Errors/Types.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Types.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Tc/Gen/App.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Types/CtLocEnv.hs - compiler/GHC/Tc/Types/LclEnv.hs - compiler/GHC/Tc/Utils/Monad.hs - testsuite/tests/pmcheck/should_compile/DsIncompleteRecSel1.stderr - testsuite/tests/pmcheck/should_compile/DsIncompleteRecSel2.stderr - testsuite/tests/pmcheck/should_compile/DsIncompleteRecSel3.stderr - + testsuite/tests/pmcheck/should_compile/T24824.hs - + testsuite/tests/pmcheck/should_compile/T24891.hs - + testsuite/tests/pmcheck/should_compile/T24891.stderr - testsuite/tests/pmcheck/should_compile/all.T - testsuite/tests/typecheck/should_compile/TcIncompleteRecSel.stderr Changes: ===================================== compiler/GHC/HsToCore/Errors/Ppr.hs ===================================== @@ -207,10 +207,12 @@ instance Diagnostic DsMessage where <+> text "for"<+> quotes (ppr lhs_id) <+> text "might fire first") ] - DsIncompleteRecordSelector name cons_wo_field not_full_examples -> mkSimpleDecorated $ - text "The application of the record field" <+> quotes (ppr name) - <+> text "may fail for the following constructors:" - <+> vcat (map ppr cons_wo_field ++ [text "..." | not_full_examples]) + DsIncompleteRecordSelector name cons maxCons -> mkSimpleDecorated $ + hang (text "Selecting the record field" <+> quotes (ppr name) + <+> text "may fail for the following constructors:") + 2 + (hsep $ punctuate comma $ + map ppr (take maxCons cons) ++ [ text "..." | lengthExceeds cons maxCons ]) diagnosticReason = \case DsUnknownMessage m -> diagnosticReason m ===================================== compiler/GHC/HsToCore/Errors/Types.hs ===================================== @@ -163,7 +163,9 @@ data DsMessage DsIncompleteRecSel2 DsIncompleteRecSel3 -} - | DsIncompleteRecordSelector !Name ![ConLike] !Bool + | DsIncompleteRecordSelector !Name -- ^ The selector + ![ConLike] -- ^ The partial constructors + !Int -- ^ The max number of constructors reported deriving Generic ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -1,6 +1,6 @@ - {-# LANGUAGE TypeFamilies #-} {-# LANGUAGE ViewPatterns #-} +{-# LANGUAGE MultiWayIf #-} {-# OPTIONS_GHC -Wno-incomplete-uni-patterns #-} {-# LANGUAGE LambdaCase #-} @@ -33,6 +33,7 @@ import GHC.HsToCore.Pmc import GHC.HsToCore.Errors.Types import GHC.Types.SourceText import GHC.Types.Name hiding (varName) +import GHC.Types.Name.Reader import GHC.Core.FamInstEnv( topNormaliseType ) import GHC.HsToCore.Quote import GHC.HsToCore.Ticks (stripTicksTopHsExpr) @@ -43,6 +44,7 @@ import GHC.Hs import GHC.Tc.Utils.TcType import GHC.Tc.Types.Evidence import GHC.Tc.Utils.Monad +import GHC.Tc.Instance.Class (lookupHasFieldLabel) import GHC.Core.Type import GHC.Core.TyCo.Rep import GHC.Core @@ -67,7 +69,6 @@ import GHC.Utils.Outputable as Outputable import GHC.Utils.Panic import GHC.Core.PatSyn import Control.Monad -import GHC.Types.Error {- ************************************************************************ @@ -261,35 +262,18 @@ dsLExpr (L loc e) = putSrcSpanDsA loc $ dsExpr e dsExpr :: HsExpr GhcTc -> DsM CoreExpr dsExpr (HsVar _ (L _ id)) = dsHsVar id -{- Record selectors are warned about if they are not -present in all of the parent data type's constructor, -or always in case of pattern synonym record selectors -(regulated by a flag). However, this only produces -a warning if it's not a part of a record selector -application. For example: - - data T = T1 | T2 {s :: Bool} - f x = s x -- the warning from this case will be supressed - -See the `HsApp` case for where it is filtered out --} +-- See of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc dsExpr (HsRecSel _ (FieldOcc id _)) = do { let name = getName id RecSelId {sel_cons = (_, cons_wo_field)} = idDetails id - ; cons_trimmed <- trim_cons cons_wo_field - ; unless (null cons_wo_field) $ diagnosticDs - $ DsIncompleteRecordSelector name cons_trimmed (cons_trimmed /= cons_wo_field) - -- This only produces a warning if it's not a part of a - -- record selector application (e.g. `s a` where `s` is a selector) - -- See the `HsApp` case for where it is filtered out + ; suppress_here <- getSuppressIncompleteRecSelsDs + -- See (4) and (5) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + ; unless (suppress_here || null cons_wo_field) $ do + dflags <- getDynFlags + let maxCons = maxUncoveredPatterns dflags + diagnosticDs $ DsIncompleteRecordSelector name cons_wo_field maxCons ; dsHsVar id } - where - trim_cons :: [ConLike] -> DsM [ConLike] - trim_cons cons_wo_field = do - dflags <- getDynFlags - let maxConstructors = maxUncoveredPatterns dflags - return $ take maxConstructors cons_wo_field dsExpr (HsUnboundVar (HER ref _ _) _) = dsEvTerm =<< readMutVar ref @@ -358,28 +342,17 @@ dsExpr (HsLam _ variant a_Match) = uncurry mkCoreLams <$> matchWrapper (LamAlt variant) Nothing a_Match dsExpr e@(HsApp _ fun arg) - -- We want to have a special case that uses the PMC information to filter - -- out some of the incomplete record selectors warnings and not trigger - -- the warning emitted during the desugaring of dsExpr(HsRecSel) - -- See Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc - = do { (msgs, fun') <- captureMessagesDs $ dsLExpr fun - -- Make sure to filter out the generic incomplete record selector warning - -- if it's a raw record selector + = do { fun' <- suppressIncompleteRecSelsDs $ dsLExpr fun + -- See (5) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc ; arg' <- dsLExpr arg - ; case getIdFromTrivialExpr_maybe fun' of - Just fun_id | isRecordSelector fun_id - -> do { let msgs' = filterMessages is_incomplete_rec_sel_msg msgs - ; addMessagesDs msgs' - ; pmcRecSel fun_id arg' } - _ -> addMessagesDs msgs + ; mb_rec_sel <- decomposeRecSelHead fun' + ; case mb_rec_sel of + -- See (2) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + Just sel_id + -> pmcRecSel sel_id arg' + _ -> return () ; warnUnusedBindValue fun arg (exprType arg') ; return $ mkCoreAppDs (text "HsApp" <+> ppr e) fun' arg' } - where - is_incomplete_rec_sel_msg :: MsgEnvelope DsMessage -> Bool - is_incomplete_rec_sel_msg (MsgEnvelope {errMsgDiagnostic = DsIncompleteRecordSelector{}}) - = False - is_incomplete_rec_sel_msg _ = True - dsExpr e@(HsAppType {}) = dsHsWrapped e @@ -1015,7 +988,19 @@ warnDiscardedDoBindings rhs rhs_ty ------------------------------ dsHsWrapped :: HsExpr GhcTc -> DsM CoreExpr dsHsWrapped orig_hs_expr - = go idHsWrapper orig_hs_expr + = do { res <- suppressIncompleteRecSelsDs $ go idHsWrapper orig_hs_expr + ; suppress_here <- getSuppressIncompleteRecSelsDs + -- See (5) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + ; if | not suppress_here + , Just fun_id <- getIdFromTrivialExpr_maybe res + , isRecordSelector fun_id + , Just (FTF_T_T, _, arg_ty,_res_ty) <- splitFunTy_maybe (exprType res) + -> do { dummy <- newSysLocalDs (Scaled ManyTy arg_ty) + -- See (3) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + ; pmcRecSel fun_id (Var dummy) } + | otherwise + -> return () + ; return res } where go wrap (HsPar _ (L _ hs_e)) = go wrap hs_e @@ -1037,3 +1022,48 @@ dsHsWrapped orig_hs_expr { addTyCs FromSource (hsWrapDictBinders wrap) $ do { e <- dsExpr hs_e ; return (wrap' e) } } } + +decomposeRecSelHead :: CoreExpr -> DsM (Maybe Id) +-- ^ Detect whether the given CoreExpr is +-- * a record selector, or +-- * a resolved getField application listing the record selector +-- See (6) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc. +decomposeRecSelHead fun + -- First plain record selectors; `sel |> co`. Easy: + | Just sel_id <- getIdFromTrivialExpr_maybe fun + , isRecordSelector sel_id + = pure (Just sel_id) + + -- Now resolved getField applications. General form: + -- getField + -- @GHC.Types.Symbol {k} + -- @"sel" x + -- @T r + -- @Int a + -- ($dHasField :: HasField "sel" T Int) + -- :: T -> Int + -- where + -- $dHasField = sel |> (co :: T -> Int ~R# HasField "sel" T Int) + -- Alas, we cannot simply look at the unfolding of $dHasField below because it + -- has not been set yet, so we have to reconstruct the selector from the types. + | App fun2 dict <- fun + -- cheap test first + , Just _dict_id <- getIdFromTrivialExpr_maybe dict + -- looks good so far. Now match deeper + , get_field `App` _k `App` Type x_ty `App` Type r_ty `App` _a_ty <- fun2 + , Just get_field_id <- getIdFromTrivialExpr_maybe get_field + , get_field_id `hasKey` getFieldClassOpKey + -- Checks out! Now get a hold of the record selector. + = do fam_inst_envs <- dsGetFamInstEnvs + rdr_env <- dsGetGlobalRdrEnv + -- Look up the field named x/"sel" in the type r/T + case lookupHasFieldLabel fam_inst_envs x_ty r_ty of + Just fl + | Just _ <- lookupGRE_FieldLabel rdr_env fl + -- Make sure the field is actually visible in this module; + -- otherwise this might not be the implicit HasField instance + -> Just <$> dsLookupGlobalId (flSelector fl) + _ -> pure Nothing + + | otherwise + = pure Nothing ===================================== compiler/GHC/HsToCore/Monad.hs ===================================== @@ -26,7 +26,7 @@ module GHC.HsToCore.Monad ( mkNamePprCtxDs, newUnique, UniqSupply, newUniqueSupply, - getGhcModeDs, dsGetFamInstEnvs, + getGhcModeDs, dsGetFamInstEnvs, dsGetGlobalRdrEnv, dsLookupGlobal, dsLookupGlobalId, dsLookupTyCon, dsLookupDataCon, dsLookupConLike, getCCIndexDsM, @@ -36,6 +36,9 @@ module GHC.HsToCore.Monad ( -- Getting and setting pattern match oracle states getPmNablas, updPmNablas, + -- Locally suppress -Wincomplete-record-selectors warnings + getSuppressIncompleteRecSelsDs, suppressIncompleteRecSelsDs, + -- Tracking evidence variable coherence addUnspecables, getUnspecables, @@ -45,7 +48,6 @@ module GHC.HsToCore.Monad ( -- Warnings and errors DsWarning, diagnosticDs, errDsCoreExpr, failWithDs, failDs, discardWarningsDs, - addMessagesDs, captureMessagesDs, -- Data types DsMatchContext(..), @@ -407,6 +409,7 @@ mkDsEnvs unit_env mod rdr_env type_env fam_inst_env ptc msg_var cc_st_var , dsl_loc = real_span , dsl_nablas = initNablas , dsl_unspecables = mempty + , dsl_suppress_incomplete_rec_sel = False } in (gbl_env, lcl_env) @@ -475,6 +478,13 @@ addUnspecables unspecables = updLclEnv (\env -> env{ dsl_unspecables = unspecabl getUnspecables :: DsM (S.Set EvId) getUnspecables = dsl_unspecables <$> getLclEnv +suppressIncompleteRecSelsDs :: DsM a -> DsM a +suppressIncompleteRecSelsDs = updLclEnv (\dsl -> dsl { dsl_suppress_incomplete_rec_sel = True }) + +-- | Get the current pattern match oracle state. See 'dsl_nablas'. +getSuppressIncompleteRecSelsDs :: DsM Bool +getSuppressIncompleteRecSelsDs = do { env <- getLclEnv; return (dsl_suppress_incomplete_rec_sel env) } + getSrcSpanDs :: DsM SrcSpan getSrcSpanDs = do { env <- getLclEnv ; return (RealSrcSpan (dsl_loc env) Strict.Nothing) } @@ -499,12 +509,6 @@ diagnosticDs dsMessage ; let msg = mkMsgEnvelope diag_opts loc (ds_name_ppr_ctx env) dsMessage ; updMutVar (ds_msgs env) (\ msgs -> msg `addMessage` msgs) } -addMessagesDs :: Messages DsMessage -> DsM () -addMessagesDs msgs1 - = do { msg_var <- ds_msgs <$> getGblEnv - ; msgs0 <- liftIO $ readIORef msg_var - ; liftIO $ writeIORef msg_var (msgs0 `unionMessages` msgs1) } - -- | Issue an error, but return the expression for (), so that we can continue -- reporting errors. errDsCoreExpr :: DsMessage -> DsM CoreExpr @@ -520,13 +524,6 @@ failWithDs msg failDs :: DsM a failDs = failM -captureMessagesDs :: DsM a -> DsM (Messages DsMessage, a) -captureMessagesDs thing_inside - = do { msg_var <- liftIO $ newIORef emptyMessages - ; res <- updGblEnv (\gbl -> gbl {ds_msgs = msg_var}) thing_inside - ; msgs <- liftIO $ readIORef msg_var - ; return (msgs, res) } - mkNamePprCtxDs :: DsM NamePprCtx mkNamePprCtxDs = ds_name_ppr_ctx <$> getGblEnv @@ -567,6 +564,9 @@ dsGetFamInstEnvs dsGetMetaEnv :: DsM (NameEnv DsMetaVal) dsGetMetaEnv = do { env <- getLclEnv; return (dsl_meta env) } +dsGetGlobalRdrEnv :: DsM GlobalRdrEnv +dsGetGlobalRdrEnv = ds_gbl_rdr_env <$> getGblEnv + -- | The @COMPLETE@ pragmas that are in scope. dsGetCompleteMatches :: DsM DsCompleteMatches dsGetCompleteMatches = ds_complete_matches <$> getGblEnv ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -201,55 +201,180 @@ pmcMatches origin ctxt vars matches = {-# SCC "pmcMatches" #-} do {- Note [Detecting incomplete record selectors] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -A record selector occurrence is incomplete iff. it could fail due to -being applied to a data type constructor not present for this record field. - -e.g. - data T = T1 | T2 {x :: Int} - d = x someComputation -- `d` may fail - -There are 4 parts to detecting and warning about -incomplete record selectors to consider: - - - Computing which constructors a general application of a record field will succeed on, - and which ones it will fail on. This is stored in the `sel_cons` field of - `IdDetails` datatype, which is a part of an `Id` and calculated when renaming a - record selector in `mkOneRecordSelector` - - - Emitting a warning whenever a `HasField` constraint is solved. - This is checked in `matchHasField` and emitted only for when - the constraint is resolved with an implicit instance rather than a - custom one (since otherwise the warning will be emitted in - the custom implementation anyways) - - e.g. - g :: HasField "x" t Int => t -> Int - g = getField @"x" - - f :: T -> Int - f = g -- warning will be emitted here - - - Emitting a warning for a general occurrence of the record selector - This is done during the renaming of a `HsRecSel` expression in `dsExpr` - and simply pulls the information about incompleteness from the `Id` - - e.g. - l :: T -> Int - l a = x a -- warning will be emitted here - - - Emitting a warning for a record selector `sel` applied to a variable `y`. - In that case we want to use the long-distance information from the - pattern match checker to rule out impossible constructors - (See Note [Long-distance information]). We first add constraints to - the long-distance `Nablas` that `y` cannot be one of the constructors that - contain `sel` (function `checkRecSel` in GHC.HsToCore.Pmc.Check). If the - `Nablas` are still inhabited, we emit a warning with the inhabiting constructors - as examples of where `sel` may fail. - - e.g. - z :: T -> Int - z T1 = 0 - z a = x a -- warning will not be emitted here since `a` can only be `T2` +This Note describes the implementation of +GHC proposal 516 "Add warning for incomplete record selectors". + +A **partial field** is a field that does not belong to every constructor of the +corresponding datatype. +A **partial selector occurrence** is a use of a record selector for a partial +field, either as a selector function in an expression, or as the solution to a +HasField constraint. + +Partial selector occurrences desugar to case expressions which may crash at +runtime: + + data T a where + T1 :: T Int + T2 {sel :: Int} :: T Bool + + urgh :: T a -> Int + urgh x = sel x + ===> + urgh x = case x of + T1 -> error "no record field sel" + T2 f -> f + +As such, it makes sense to warn about such potential crashes. +We do so whenever -Wincomplete-record-selectors is present, and we utilise +the pattern-match coverage checker for precise results, because there are many +uses of selectors for partial fields which are in fact dynamically safe. + +Pmc can detect two very common safe uses for which we will not warn: + + (LDI) Ambient pattern-matches unleash Note [Long-distance information] that + render a naively flagged partial selector occurrence safe, as in + ldi :: T a -> Int + ldi T1 = 0 + ldi arg = sel arg + We should not warn here, because `arg` cannot be `T1`. + + (RES) Constraining the result type of a GADT such as T might render + naively flagged partial selector occurrences safe, as in + resTy :: T Bool -> Int + resTy = sel + Here, `T1 :: T Int` is ruled out because it has the wrong result type. + +Additionally, we want to support incomplete -XOverloadedRecordDot access as +well, in either the (LDI) use case or the (RES) use case: + + data Dot = No | Yes { sel2 :: Int } + dot d = d.sel2 -- should warn + ldiDot No = 0 + ldiDot d = d.sel2 -- should not warn + resTyDot :: T Bool -> Int + resTyDot x = x.sel -- should not warn + +From a user's point of view, function `ldiDot` looks very like example `ldi` and +`resTyDot` looks very like `resTy`. But from an /implementation/ point of view +they are very different: both `ldiDot` and `resTyDot` simply emit `HasField` +constraints, and it is those constraints that the implementation must use to +determine incompleteness. + +Furthermore, HasField constraints allow to delay the completeness check from +the field access site to a caller, as in test cases TcIncompleteRecSel and T24891: + + accessDot :: HasField "sel2" t Int => t -> Int + accessDot x = x.sel2 + solveDot :: Dot -> Int + solveDot = accessDot + +We should warn in `solveDot`, but not in `accessDot`. + +Here is how we achieve all this in the implementation: + + 1. When renaming a record selector in `mkOneRecordSelector`, + we precompute the constructors the selector succeeds on. + That would be `T2` for `sel` because `sel (T2 42)` succeeds, + and `Yes` for `sel2` because `sel2 (Yes 13)` succeeds. + We store this information in the `sel_cons` field of `RecSelId`. + (Remember, the same field may occur in several constructors of the data + type; hence the selector may succeed on more than one constructor.) + +The next three items describe mechanisms for producing warnings on record +selectors and situations in which they trigger. +They are ordered by specificity, so we prefer (2) over (3) over (4). +Item (5) below describes how we resolve the overlap. +(-XOverloadedRecordDot is discussed separately in Item (6) and (7).) + + 2. In function `ldi`, we have a record selector application `sel arg`. + This situation is detected in the `HsApp` case of `dsExpr`. + We call out to the pattern-match checker to determine whether use of the + selector is safe, by calling GHC.HsToCore.Pmc.pmcRecSel, passing the + `RecSelId` `sel` as well as `arg`. + + The pattern-match checker reduces the partial-selector-occurrence problem to + a complete-match problem by adding a negative constructor constraint such as + `arg /~ T2` for every constructor in the precomputed `sel_cons` of `sel`. + Recall that these were exactly the constructors which define a field `sel`. + `pmcRecSel` then tests + + case arg of {} + + for completeness. Any incomplete match, such as in the original `urgh`, must + reference a constructor that does not have field `sel`, such as `T1`. + In case of `urgh`, `T1` is indeed the case that we report as inexhaustive. + + However, in function `ldi`, we have *both* the result type of + `arg::T a` (boring, but see (3)) as well as Note [Long-distance information] + about `arg` from the ambient match, and the latter lists the constraint + `arg /~ T1`. Consequently, since `arg` is neither `T1` nor `T2` in the + reduced problem, the match is exhaustive and the use of the record selector + safe. + + 3. In function `resTy`, the record selector is unsaturated, but the result type + ensures a safe use of the selector. + This situation is detected in `dsHsWrapped`, where the record selector + is elaborated with its type arguments; we simply match on desugared Core + `sel @Bool :: T Bool -> Int` to learn the result type `T Bool`. + We again call `pmcRecSel`, but this time with a fresh dummy Id `ds::T Bool`. + + 4. In case of an unsaturated record selector that is *not* applied to any type + argument after elaboration (e.g. in `urgh2 = sel2 :: Dot -> Int`), we simply + produce a warning about all `sel_cons`; no need to call `pmcRecSel`. + This happens in the `HsRecSel` case of `dsExpr`. + +We resolve the overlap between situations (2)-(4) by preferring (2) over (3) +over (4) as follows: + + 5. (4) produces warnings in the `HsRecSel` case of `dsExpr`. + (3) produces warnings in a potentially surrounding call to `dsHsWrapped`. + (2) produces warnings in a potentially surrounding `HsApp` case in `dsExpr`. + Since (2) surrounds (3) surrounds (4), this is simply implemented via the + flag `dsl_suppress_incomplete_rec_sel` of the `DsLclEnv`, which is set + in (2) and (3) before desugaring subexpressions. + +Finally, there are 2 more items addressing -XOverloadedRecordDot: + + 6. -XOverloadedRecordDot such as in function `ldiDot` desugars as follows: + getField + @GHC.Types.Symbol + @"sel2" + @Dot + @Int + ($dHasField :: HasField "sel2" Dot Int) + d + where + $dHasField = sel2 |> (co :: Dot -> Int ~R# HasField "sel2" Dot Int) + We want to catch these applications in the saturated (2) case. + (The unsaturated case is handled implicitly by (7).) + For example, we do not want to generate a warning for function `ldiDot`! + For that, we need to be smart in `decomposeRecSelHead`, which matches out + the record selector. It must treat the above expression similar to a vanilla + RecSel app `sel2 d`. + This is a bit nasty since we cannot look at the unfolding of `$dHasField`. + Tested in T24891. + + 7. For `accessDot` above, `decomposeRecSelHead` will fail to find a record + selector, because type `t` is not obviously a record type. + That's good, because it means we won't emit a warning for `accessDot`. + But we should really emit a warning for `solveDot`! + There, the compiler solves a `HasField` constraint and without an immediate + `getField`, roughly `solveDot = accessDot @Dot $d`. + It is the job of the solver to warn about incompleteness here, + in `GHC.Tc.Instance.Class.matchHasField`. + + What makes this complicated is that we do not *also* want to warn in the + example `dot d = d.sel2` above, which is covered by more precise case (6)! + We suppress the warning in this case as follows: + 1. The type-checker (`GHC.Tc.Gen.tcApp`) produces `getField @.. $d e` + (Remember that (6) will detect `getField @.. $d e` as well.) + 2. Through `tcl_suppress_incomplete_rec_sel`, we suppress warnings when + solving `$d`. + 3. ... but not when checking `e`, because `e` might itself be a field + access that would need to be checked individually. + 4. What complicates matters is that the solver runs *after* type-checking, + so we must persist `tcl_suppress_incomplete_rec_sel` in the `CtLocEnv`. + What a hassle. This is all tested in T24891. -} pmcRecSel :: Id -- ^ Id of the selector @@ -270,13 +395,11 @@ pmcRecSel sel_id arg sel_name = varName sel_id warn_incomplete arg_id uncov_nablas = do dflags <- getDynFlags - let maxConstructors = maxUncoveredPatterns dflags - unc_examples <- getNFirstUncovered MinimalCover [arg_id] (maxConstructors + 1) uncov_nablas + let maxPatterns = maxUncoveredPatterns dflags + unc_examples <- getNFirstUncovered MinimalCover [arg_id] (maxPatterns + 1) uncov_nablas let cons = [con | unc_example <- unc_examples , Just (PACA (PmAltConLike con) _ _) <- [lookupSolution unc_example arg_id]] - not_full_examples = length cons == (maxConstructors + 1) - cons' = take maxConstructors cons - diagnosticDs $ DsIncompleteRecordSelector sel_name cons' not_full_examples + diagnosticDs $ DsIncompleteRecordSelector sel_name cons maxPatterns pmcRecSel _ _ = return () ===================================== compiler/GHC/HsToCore/Types.hs ===================================== @@ -11,7 +11,7 @@ module GHC.HsToCore.Types ( DsMetaEnv, DsMetaVal(..), CompleteMatches ) where -import GHC.Prelude (Int) +import GHC.Prelude (Int, Bool) import Data.IORef import qualified Data.Set as S @@ -80,6 +80,9 @@ data DsLclEnv -- ^ See Note [Long-distance information] in "GHC.HsToCore.Pmc". -- The set of reaching values Nablas is augmented as we walk inwards, refined -- through each pattern match in turn + , dsl_suppress_incomplete_rec_sel :: Bool + -- ^ Whether to suppress -Wincomplete-record-selectors warnings. + -- See (5) of Note [Detecting incomplete record selectors] , dsl_unspecables :: S.Set EvVar -- ^ See Note [Desugaring non-canonical evidence]: this field collects -- all un-specialisable evidence variables in scope. ===================================== compiler/GHC/Tc/Errors/Ppr.hs ===================================== @@ -1536,9 +1536,12 @@ instance Diagnostic TcRnMessage where vcat [ sep [ text "Definition of partial record field" <> colon , nest 2 $ quotes (ppr (occName fld)) ] , text "Record selection and update using this field will be partial." ] - TcRnHasFieldResolvedIncomplete name -> mkSimpleDecorated $ - text "The invocation of `getField` on the record field" <+> quotes (ppr name) - <+> text "may produce an error since it is not defined for all data constructors" + TcRnHasFieldResolvedIncomplete name cons maxCons -> mkSimpleDecorated $ + hang (text "Selecting the record field" <+> quotes (ppr name) + <+> text "may fail for the following constructors:") + 2 + (hsep $ punctuate comma $ + map ppr (take maxCons cons) ++ [ text "..." | lengthExceeds cons maxCons ]) TcRnBadFieldAnnotation n con reason -> mkSimpleDecorated $ hang (pprBadFieldAnnotationReason reason) 2 (text "on the" <+> speakNth n ===================================== compiler/GHC/Tc/Errors/Types.hs ===================================== @@ -3573,7 +3573,10 @@ data TcRnMessage where Test cases: TcIncompleteRecSel -} - TcRnHasFieldResolvedIncomplete :: !Name -> TcRnMessage + TcRnHasFieldResolvedIncomplete :: !Name -- ^ The selector + -> ![ConLike] -- ^ The partial constructors + -> !Int -- ^ The max number of constructors reported + -> TcRnMessage {-| TcRnBadFieldAnnotation is an error/warning group indicating that a strictness/unpack related data type field annotation is invalid. ===================================== compiler/GHC/Tc/Gen/App.hs ===================================== @@ -403,10 +403,18 @@ tcApp rn_expr exp_res_ty ; (tc_fun, fun_sigma) <- tcInferAppHead fun ; let tc_head = (tc_fun, fun_ctxt) + ; let supp_incomplete_rec_sel + | XExpr (ExpandedThingRn (OrigExpr HsGetField{}) _) <- rn_expr + -- See (7) of Note [Detecting incomplete record selectors] + = setSuppressIncompleteRecSelsTc True + | otherwise + = id + -- Step 3: Instantiate the function type (taking a quick look at args) ; do_ql <- wantQuickLook rn_fun ; (inst_args, app_res_rho) - <- setQLInstLevel do_ql $ -- See (TCAPP1) and (TCAPP2) in + <- supp_incomplete_rec_sel $ + setQLInstLevel do_ql $ -- See (TCAPP1) and (TCAPP2) in -- Note [tcApp: typechecking applications] tcInstFun do_ql True tc_head fun_sigma rn_args @@ -764,7 +772,9 @@ tcInstFun do_ql inst_final (tc_fun, fun_ctxt) fun_sigma rn_args ; go (pos+1) (addArgWrap wrap acc) inst_body rest_args } go1 pos acc fun_ty (EWrap w : args) - = go1 pos (EWrap w : acc) fun_ty args + = setSuppressIncompleteRecSelsTc False $ + -- See (7) of Note [Detecting incomplete record selectors] + go1 pos (EWrap w : acc) fun_ty args go1 pos acc fun_ty (EPrag sp prag : args) = go1 pos (EPrag sp prag : acc) fun_ty args @@ -2269,4 +2279,3 @@ rejectRepPolyNewtypes (fun,_) app_res_rho = case fun of tcExprPrag :: HsPragE GhcRn -> HsPragE GhcTc tcExprPrag (HsPragSCC x1 ann) = HsPragSCC x1 ann - ===================================== compiler/GHC/Tc/Instance/Class.hs ===================================== @@ -5,7 +5,8 @@ module GHC.Tc.Instance.Class ( matchGlobalInst, matchEqualityInst, ClsInstResult(..), InstanceWhat(..), safeOverlap, instanceReturnsDictCon, - AssocInstInfo(..), isNotAssociated + AssocInstInfo(..), isNotAssociated, + lookupHasFieldLabel ) where import GHC.Prelude @@ -22,7 +23,7 @@ import GHC.Tc.Instance.Typeable import GHC.Tc.Utils.TcMType import GHC.Tc.Types.Evidence import GHC.Tc.Types.Origin (InstanceWhat (..), SafeOverlapping) -import GHC.Tc.Instance.Family( tcGetFamInstEnvs, tcInstNewTyCon_maybe, tcLookupDataFamInst ) +import GHC.Tc.Instance.Family( tcGetFamInstEnvs, tcInstNewTyCon_maybe, tcLookupDataFamInst, FamInstEnvs ) import GHC.Rename.Env( addUsedGRE, addUsedDataCons, DeprecationWarnings (..) ) import GHC.Builtin.Types @@ -1252,15 +1253,9 @@ matchHasField dflags short_cut clas tys ; case tys of -- We are matching HasField {k} {r_rep} {a_rep} x r a... [_k_ty, _r_rep, _a_rep, x_ty, r_ty, a_ty] - -- x should be a literal string - | Just x <- isStrLitTy x_ty - -- r should be an applied type constructor - , Just (tc, args) <- tcSplitTyConApp_maybe r_ty - -- use representation tycon (if data family); it has the fields - , let r_tc = fstOf3 (tcLookupDataFamInst fam_inst_envs tc args) - -- x should be a field of r - , Just fl <- lookupTyConFieldLabel (FieldLabelString x) r_tc - -- the field selector should be in scope + -- Look up the field named x in the type r + | Just fl <- lookupHasFieldLabel fam_inst_envs x_ty r_ty + -- and ensure the field selector is in scope , Just gre <- lookupGRE_FieldLabel rdr_env fl -> do { let name = flSelector fl @@ -1295,10 +1290,12 @@ matchHasField dflags short_cut clas tys then do { -- See Note [Unused name reporting and HasField] addUsedGRE AllDeprecationWarnings gre ; keepAlive name - ; unless (null $ snd $ sel_cons $ idDetails sel_id) - $ addDiagnostic $ TcRnHasFieldResolvedIncomplete name - -- Only emit an incomplete selector warning if it's an implicit instance - -- See Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + ; let maxCons = maxUncoveredPatterns dflags + ; let (_, fallible_cons) = sel_cons (idDetails sel_id) + ; suppress <- getSuppressIncompleteRecSelsTc + -- See (7) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + ; unless (null fallible_cons || suppress) $ do + addDiagnostic $ TcRnHasFieldResolvedIncomplete name fallible_cons maxCons ; return OneInst { cir_new_theta = theta , cir_mk_ev = mk_ev , cir_canonical = EvCanonical @@ -1306,3 +1303,21 @@ matchHasField dflags short_cut clas tys else matchInstEnv dflags short_cut clas tys } _ -> matchInstEnv dflags short_cut clas tys } + +lookupHasFieldLabel :: FamInstEnvs -> Type -> Type -> Maybe FieldLabel +-- The call (lookupHasFieldLabel fam_envs (LitTy "fld") (T t1..tn)) +-- returns the `FieldLabel` of field "fld" in the data type T. +-- A complication is that `T` might be a data family, so we need to +-- look it up in the `fam_envs` to find its representation tycon. +lookupHasFieldLabel fam_inst_envs x_ty r_ty + -- x should be a literal string + | Just x <- isStrLitTy x_ty + -- r should be an applied type constructor + , Just (tc, args) <- tcSplitTyConApp_maybe r_ty + -- use representation tycon (if data family); it has the fields + , let r_tc = fstOf3 (tcLookupDataFamInst fam_inst_envs tc args) + -- x should be a field of r + = lookupTyConFieldLabel (FieldLabelString x) r_tc + + | otherwise + = Nothing ===================================== compiler/GHC/Tc/Types/CtLocEnv.hs ===================================== @@ -29,7 +29,8 @@ data CtLocEnv = CtLocEnv { ctl_ctxt :: ![ErrCtxt] , ctl_bndrs :: !TcBinderStack , ctl_tclvl :: !TcLevel , ctl_in_gen_code :: !Bool - , ctl_rdr :: !LocalRdrEnv } + , ctl_rdr :: !LocalRdrEnv + , ctl_suppress_incomplete_rec_sels :: !Bool } getCtLocEnvLoc :: CtLocEnv -> RealSrcSpan @@ -57,4 +58,4 @@ setCtLocEnvLoc env loc@(UnhelpfulSpan _) = env ctLocEnvInGeneratedCode :: CtLocEnv -> Bool -ctLocEnvInGeneratedCode = ctl_in_gen_code \ No newline at end of file +ctLocEnvInGeneratedCode = ctl_in_gen_code ===================================== compiler/GHC/Tc/Types/LclEnv.hs ===================================== @@ -12,6 +12,7 @@ module GHC.Tc.Types.LclEnv ( , getLclEnvRdrEnv , getLclEnvTcLevel , getLclEnvThStage + , getLclEnvSuppressIncompleteRecSels , setLclEnvTcLevel , setLclEnvLoc , setLclEnvRdrEnv @@ -19,6 +20,7 @@ module GHC.Tc.Types.LclEnv ( , setLclEnvErrCtxt , setLclEnvThStage , setLclEnvTypeEnv + , setLclEnvSuppressIncompleteRecSels , modifyLclEnvTcLevel , lclEnvInGeneratedCode @@ -117,9 +119,11 @@ data TcLclCtxt tcl_arrow_ctxt :: ArrowCtxt, -- Arrow-notation context - tcl_env :: TcTypeEnv -- The local type environment: + tcl_env :: TcTypeEnv, -- The local type environment: -- Ids and TyVars defined in this module + tcl_suppress_incomplete_rec_sel :: Bool -- True <=> Suppress warnings about incomplete record selectors + -- See (7) of Note [Detecting incomplete record selectors] } getLclEnvThStage :: TcLclEnv -> ThStage @@ -179,6 +183,12 @@ getLclEnvRdrEnv = tcl_rdr . tcl_lcl_ctxt setLclEnvRdrEnv :: LocalRdrEnv -> TcLclEnv -> TcLclEnv setLclEnvRdrEnv rdr_env = modifyLclCtxt (\env -> env { tcl_rdr = rdr_env }) +getLclEnvSuppressIncompleteRecSels :: TcLclEnv -> Bool +getLclEnvSuppressIncompleteRecSels = tcl_suppress_incomplete_rec_sel . tcl_lcl_ctxt + +setLclEnvSuppressIncompleteRecSels :: Bool -> TcLclEnv -> TcLclEnv +setLclEnvSuppressIncompleteRecSels suppress = modifyLclCtxt (\env -> env { tcl_suppress_incomplete_rec_sel = suppress }) + modifyLclCtxt :: (TcLclCtxt -> TcLclCtxt) -> TcLclEnv -> TcLclEnv modifyLclCtxt upd env = let !res = upd (tcl_lcl_ctxt env) ===================================== compiler/GHC/Tc/Utils/Monad.hs ===================================== @@ -94,6 +94,7 @@ module GHC.Tc.Utils.Monad( mkTcRnMessage, reportDiagnostic, reportDiagnostics, warnIf, diagnosticTc, diagnosticTcM, addDiagnosticTc, addDiagnosticTcM, addDiagnostic, addDiagnosticAt, + getSuppressIncompleteRecSelsTc, setSuppressIncompleteRecSelsTc, -- * Type constraints newTcEvBinds, newNoTcEvBinds, cloneEvBindsVar, @@ -398,7 +399,8 @@ initTcWithGbl hsc_env gbl_env loc do_this tcl_arrow_ctxt = NoArrowCtxt, tcl_env = emptyNameEnv, tcl_bndrs = [], - tcl_tclvl = topTcLevel + tcl_tclvl = topTcLevel, + tcl_suppress_incomplete_rec_sel = False }, tcl_usage = usage_var, tcl_lie = lie_var, @@ -1273,6 +1275,7 @@ mkCtLocEnv lcl_env = , ctl_tclvl = getLclEnvTcLevel lcl_env , ctl_in_gen_code = lclEnvInGeneratedCode lcl_env , ctl_rdr = getLclEnvRdrEnv lcl_env + , ctl_suppress_incomplete_rec_sels = getLclEnvSuppressIncompleteRecSels lcl_env } setCtLocM :: CtLoc -> TcM a -> TcM a @@ -1281,6 +1284,7 @@ setCtLocM (CtLoc { ctl_env = lcl }) thing_inside = updLclEnv (\env -> setLclEnvLoc (ctl_loc lcl) $ setLclEnvErrCtxt (ctl_ctxt lcl) $ setLclEnvBinderStack (ctl_bndrs lcl) + $ setLclEnvSuppressIncompleteRecSels (ctl_suppress_incomplete_rec_sels lcl) $ env) thing_inside {- ********************************************************************* @@ -1662,6 +1666,12 @@ add_diagnostic msg ; mkTcRnMessage loc (TcRnMessageWithInfo unit_state msg) >>= reportDiagnostic } +getSuppressIncompleteRecSelsTc :: TcRn Bool +getSuppressIncompleteRecSelsTc = getLclEnvSuppressIncompleteRecSels <$> getLclEnv + +setSuppressIncompleteRecSelsTc :: Bool -> TcRn a -> TcRn a +setSuppressIncompleteRecSelsTc b = updLclEnv (setLclEnvSuppressIncompleteRecSels b) + {- ----------------------------------- ===================================== testsuite/tests/pmcheck/should_compile/DsIncompleteRecSel1.stderr ===================================== @@ -1,3 +1,4 @@ - DsIncompleteRecSel1.hs:8:5: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘x’ may fail for the following constructors: T2 + Selecting the record field ‘x’ may fail for the following constructors: + T2 + ===================================== testsuite/tests/pmcheck/should_compile/DsIncompleteRecSel2.stderr ===================================== @@ -1,6 +1,8 @@ - DsIncompleteRecSel2.hs:22:8: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘x’ may fail for the following constructors: T4 + Selecting the record field ‘x’ may fail for the following constructors: + T4 DsIncompleteRecSel2.hs:28:19: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘x’ may fail for the following constructors: P + Selecting the record field ‘x’ may fail for the following constructors: + P + ===================================== testsuite/tests/pmcheck/should_compile/DsIncompleteRecSel3.stderr ===================================== @@ -1,80 +1,48 @@ - DsIncompleteRecSel3.hs:29:7: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘x’ may fail for the following constructors: T2 + Selecting the record field ‘x’ may fail for the following constructors: + T2 DsIncompleteRecSel3.hs:34:7: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:13: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:19: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:25: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:31: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:37: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:43: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:49: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:55: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:34:61: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... DsIncompleteRecSel3.hs:37:5: warning: [GHC-17335] [-Wincomplete-record-selectors] - The application of the record field ‘y’ may fail for the following constructors: G2 - G3 - G4 - G5 - ... + Selecting the record field ‘y’ may fail for the following constructors: + G2, G3, G4, G5, ... + ===================================== testsuite/tests/pmcheck/should_compile/T24824.hs ===================================== @@ -0,0 +1,36 @@ +{-# LANGUAGE TypeFamilies #-} + +module T24824 where + +import GHC.Hs hiding (DataConCantHappen) + +main :: IO () +main = do + let hsModule = undefined :: HsModule GhcPs + let _ = hsmodImports $ hsModule -- warns + let _ = hsmodImports hsModule -- does not warn + pure () + +data S a where + S1 :: S Int + S2 :: { x::Int } -> S a + S3 :: { x::Int } -> S a + +-- x :: forall a. S a -> Int +-- A partial function + +g :: S Bool -> Int +g s = (x @Bool) $ s + +data W a where + W1 :: !(F a) -> W a + W2 :: { y::Int } -> W a + W3 :: { y::Int } -> W a + +data DataConCantHappen + +type family F a +type instance F Bool = DataConCantHappen + +h :: W Bool -> Int +h w = y @Bool $ w ===================================== testsuite/tests/pmcheck/should_compile/T24891.hs ===================================== @@ -0,0 +1,30 @@ +{-# LANGUAGE GADTs, OverloadedRecordDot, DataKinds #-} + +module T24891 where + +import GHC.Records + +data T a where + T1 :: T Int + T2 :: {sel :: Int} -> T Bool + T3 :: T Bool + +f :: T Bool -> Int +f x = x.sel -- warn, but only once, suggesting to match on T3 + +data Dot = No | Yes {sel2 :: Int} + +ldiDot :: Dot -> Int +ldiDot No = 0 +ldiDot d = d.sel2 -- do not warn + +accessDot :: HasField "sel2" t Int => t -> Int +accessDot x = x.sel2 -- do not warn + +solveDot :: Dot -> Int +solveDot = accessDot -- warn + +data Dot2 t = No2 | Yes2 {sel3 :: t} + +accessDot2 :: HasField "sel2" t Int => Dot2 t -> Int +accessDot2 x = x.sel3.sel2 -- warn about x.sel3 ===================================== testsuite/tests/pmcheck/should_compile/T24891.stderr ===================================== @@ -0,0 +1,12 @@ +T24891.hs:13:7: warning: [GHC-17335] [-Wincomplete-record-selectors] + Selecting the record field ‘sel’ may fail for the following constructors: + T3 + +T24891.hs:25:12: warning: [GHC-86894] [-Wincomplete-record-selectors] + Selecting the record field ‘sel2’ may fail for the following constructors: + No + +T24891.hs:30:16: warning: [GHC-17335] [-Wincomplete-record-selectors] + Selecting the record field ‘sel3’ may fail for the following constructors: + No2 + ===================================== testsuite/tests/pmcheck/should_compile/all.T ===================================== @@ -172,4 +172,6 @@ test('DsIncompleteRecSel2', normal, compile, ['-Wincomplete-record-selectors']) test('DsIncompleteRecSel3', [collect_compiler_stats('bytes allocated', 10)], compile, ['-Wincomplete-record-selectors']) test('DoubleMatch', normal, compile, [overlapping_incomplete]) test('T24817', normal, compile, [overlapping_incomplete]) +test('T24824', normal, compile, ['-package ghc -Wincomplete-record-selectors']) +test('T24891', normal, compile, ['-Wincomplete-record-selectors']) test('T25257', normal, compile, [overlapping_incomplete]) ===================================== testsuite/tests/typecheck/should_compile/TcIncompleteRecSel.stderr ===================================== @@ -1,3 +1,4 @@ - TcIncompleteRecSel.hs:16:5: warning: [GHC-86894] [-Wincomplete-record-selectors] - The invocation of `getField` on the record field ‘x’ may produce an error since it is not defined for all data constructors + Selecting the record field ‘x’ may fail for the following constructors: + T2 + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/024addc0420e57cfc882e6af0a4af2ec7409c413...7cf5a9a24d0a9ed90e855fd259bf788a14e387c0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/024addc0420e57cfc882e6af0a4af2ec7409c413...7cf5a9a24d0a9ed90e855fd259bf788a14e387c0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 12:32:26 2024 From: gitlab at gitlab.haskell.org (Sebastian Graf (@sgraf812)) Date: Wed, 25 Sep 2024 08:32:26 -0400 Subject: [Git][ghc/ghc][wip/T24824] 14 commits: Demand: Combine examples into Note (#25107) Message-ID: <66f402da862ee_3b75314a8d3c731ec@gitlab.mail> Sebastian Graf pushed to branch wip/T24824 at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - f1fba383 by Sebastian Graf at 2024-09-25T14:32:18+02:00 Pmc: Improve implementation of -Wincomplete-record-selectors (#24824, #24891) We now incorporate the result type of unsaturated record selector applications as well as consider long-distance information in getField applications. See the updated Note [Detecting incomplete record selectors]. Fixes #24824 and #24891. - - - - - d358a313 by Sebastian Graf at 2024-09-25T14:32:18+02:00 Pmc: Improve warning messages of -Wincomplete-record-selectors ... as suggested by Adam Gundry in !12685. - - - - - 30 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/HsToCore/Errors/Ppr.hs - compiler/GHC/HsToCore/Errors/Types.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Types.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Tc/Gen/App.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Types/CtLocEnv.hs - compiler/GHC/Tc/Types/LclEnv.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7cf5a9a24d0a9ed90e855fd259bf788a14e387c0...d358a313f0c32d769ed5f612bbf58eaaffd242ac -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7cf5a9a24d0a9ed90e855fd259bf788a14e387c0...d358a313f0c32d769ed5f612bbf58eaaffd242ac You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 14:37:53 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Wed, 25 Sep 2024 10:37:53 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] fixed ambiguity regression Message-ID: <66f4204111dd_1231021240e44120@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 4775c0fa by Hassan Al-Awwadi at 2024-09-25T16:37:16+02:00 fixed ambiguity regression - - - - - 17 changed files: - compiler/GHC/Hs/Instances.hs - compiler/GHC/Hs/Type.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Rename/Fixity.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Tc/Gen/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - compiler/Language/Haskell/Syntax/Extension.hs - compiler/Language/Haskell/Syntax/Pat.hs - compiler/Language/Haskell/Syntax/Type.hs Changes: ===================================== compiler/GHC/Hs/Instances.hs ===================================== @@ -549,6 +549,10 @@ deriving instance Data (FieldOcc GhcRn) deriving instance Data (FieldOcc GhcTc) deriving instance Data AmbiguousFieldOcc +deriving instance Data (UpdFieldOcc GhcPs) +deriving instance Data (UpdFieldOcc GhcRn) +deriving instance Data (UpdFieldOcc GhcTc) + -- deriving instance (DataId name) => Data (ImportDecl name) deriving instance Data (ImportDecl GhcPs) deriving instance Data (ImportDecl GhcRn) ===================================== compiler/GHC/Hs/Type.hs ===================================== @@ -56,7 +56,8 @@ module GHC.Hs.Type ( ConDeclField(..), LConDeclField, pprConDeclFields, HsConDetails(..), noTypeArgs, - + UpdFieldOcc(..), LUpdFieldOcc, mkUpdFieldOcc, + updFieldOccRdrName, updFieldOccLRdrName, FieldOcc(..), LFieldOcc, mkFieldOcc, fieldOccRdrName, fieldOccLRdrName, AmbiguousFieldOcc(..), @@ -1087,14 +1088,38 @@ type instance XCFieldOcc GhcRn = RdrName type instance XCFieldOcc GhcTc = RdrName type instance XXFieldOcc GhcPs = DataConCantHappen -type instance XXFieldOcc GhcRn = AmbiguousFieldOcc +type instance XXFieldOcc GhcRn = DataConCantHappen type instance XXFieldOcc GhcTc = DataConCantHappen +type instance XCUpdFieldOcc GhcPs = NoExtField +type instance XCUpdFieldOcc GhcRn = NoExtField +type instance XCUpdFieldOcc GhcTc = NoExtField + +type instance XXUpdFieldOcc GhcPs = DataConCantHappen +type instance XXUpdFieldOcc GhcRn = AmbiguousFieldOcc +type instance XXUpdFieldOcc GhcTc = DataConCantHappen + -------------------------------------------------------------------------------- +mkUpdFieldOcc :: LocatedN RdrName -> UpdFieldOcc GhcPs +mkUpdFieldOcc rdr@(L l _) = UpdFieldOcc noExtField (L (l2l l) $ mkFieldOcc rdr) + mkFieldOcc :: LocatedN RdrName -> FieldOcc GhcPs mkFieldOcc rdr = FieldOcc noExtField rdr +updFieldOccRdrName :: forall p. IsPass p => UpdFieldOcc (GhcPass p) -> RdrName +updFieldOccRdrName = unLoc . updFieldOccLRdrName + +updFieldOccLRdrName :: forall p. IsPass p => UpdFieldOcc (GhcPass p) -> LocatedN RdrName +updFieldOccLRdrName (UpdFieldOcc _ (L _ fo)) = fieldOccLRdrName fo +updFieldOccLRdrName (XUpdFieldOcc xfo) = case ghcPass @p of + GhcRn -> case xfo of + Ambiguous l -> l + -- Are these cases required? I don't get a hint that they aren't so + -- they are included for now. + GhcPs -> case xfo of + GhcTc -> case xfo of + fieldOccRdrName :: forall p. IsPass p => FieldOcc (GhcPass p) -> RdrName fieldOccRdrName fo = case ghcPass @p of GhcPs -> unLoc $ foLabel fo @@ -1108,7 +1133,6 @@ fieldOccLRdrName fo = case ghcPass @p of FieldOcc rdr sel -> let (L l _) = sel in L l rdr - XFieldOcc (Ambiguous l) -> l GhcTc -> let (L l _) = foLabel fo in L l (foExt fo) @@ -1254,6 +1278,13 @@ instance (OutputableBndrId pass) => OutputableBndr (GenLocated SrcSpan (FieldOcc pprInfixOcc = pprInfixOcc . unLoc pprPrefixOcc = pprPrefixOcc . unLoc +instance (IsPass p) => Outputable (UpdFieldOcc (GhcPass p))where + ppr = ppr . updFieldOccRdrName + +instance (IsPass p) => OutputableBndr (UpdFieldOcc (GhcPass p)) where + pprInfixOcc = pprInfixOcc . updFieldOccRdrName + pprPrefixOcc = pprPrefixOcc . updFieldOccRdrName + ppr_tylit :: (HsTyLit (GhcPass p)) -> SDoc ppr_tylit (HsNumTy source i) = pprWithSourceText source (integer i) @@ -1565,3 +1596,4 @@ type instance Anno (ConDeclField (GhcPass p)) = SrcSpanAnnA type instance Anno (FieldOcc (GhcPass p)) = SrcSpanAnnA type instance Anno AmbiguousFieldOcc = SrcSpanAnnA +type instance Anno (UpdFieldOcc (GhcPass p)) = SrcSpanAnnA ===================================== compiler/GHC/HsToCore/Quote.hs ===================================== @@ -1539,7 +1539,6 @@ repE (HsIPVar _ n) = rep_implicit_param_name n >>= repImplicitParamVar repE (HsOverLabel _ s) = repOverLabel s repE (HsRecSel _ (FieldOcc _ (L _ x))) = repE (HsVar noExtField (noLocA x)) -repE r@(HsRecSel _ (XFieldOcc _)) = notHandled (ThAmbiguousRecordSelectors r) -- Remember, we're desugaring renamer output here, so -- HsOverlit can definitely occur @@ -1817,11 +1816,12 @@ repUpdFields = repListM fieldExpTyConName rep_fld where rep_fld :: LHsRecUpdField GhcRn GhcRn -> MetaM (Core (M TH.FieldExp)) rep_fld (L l fld) = case unLoc (hfbLHS fld) of - FieldOcc _ (L _ sel_name) -> do { fn <- lookupLOcc (L l sel_name) - ; e <- repLE (hfbRHS fld) - ; repFieldExp fn e - } - (XFieldOcc _) -> notHandled (ThAmbiguousRecordUpdates fld) + UpdFieldOcc _ (L _ (FieldOcc _ (L _ sel_name))) -> + do { fn <- lookupLOcc (L l sel_name) + ; e <- repLE (hfbRHS fld) + ; repFieldExp fn e + } + (XUpdFieldOcc _) -> notHandled (ThAmbiguousRecordUpdates fld) ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1257,7 +1257,9 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where RecordUpd { rupd_expr = expr , rupd_flds = RegularRecUpdFields { recUpdFields = upds } }-> [ toHie expr - , toHie $ map (RC RecFieldAssign) upds + , case hiePass @p of + HieRn -> toHie $ map (RC RecFieldAssign) upds + HieTc -> toHie $ map (RC RecFieldAssign) upds ] RecordUpd { rupd_expr = expr , rupd_flds = OverloadedRecUpdFields {} }-> @@ -1476,6 +1478,10 @@ instance ( ToHie (RFContext label) , toHie expr ] +instance HiePass p => ToHie (RFContext (LocatedA (UpdFieldOcc (GhcPass p)))) where + toHie (RFC c rhs (L nspan (UpdFieldOcc _ (L _ fo)))) = concatM + [toHie (RFC c rhs (L nspan fo))] + toHie (RFC _ _ (L _ (XUpdFieldOcc _))) = concatM [] instance HiePass p => ToHie (RFContext (LocatedA (FieldOcc (GhcPass p)))) where toHie (RFC c rhs (L nspan f)) = concatM $ case hiePass @p of @@ -1483,7 +1489,6 @@ instance HiePass p => ToHie (RFContext (LocatedA (FieldOcc (GhcPass p)))) where case f of FieldOcc _ fld -> [toHie $ C (RecField c rhs) (L (locA nspan) $ unLoc fld)] - XFieldOcc (Ambiguous _) -> [] HieTc -> case f of FieldOcc _ fld -> @@ -2073,6 +2078,7 @@ instance ToHie (LocatedA (DataFamInstDecl GhcRn)) where instance ToHie (LocatedA (TyFamInstDecl GhcRn)) where toHie (L sp (TyFamInstDecl _ d)) = toHie $ TS (ResolvedScopes [mkScope sp]) d + instance HiePass p => ToHie (Context (FieldOcc (GhcPass p))) where toHie (C c (FieldOcc _ l)) = toHie (C c l) toHie (C _ (XFieldOcc _)) = concatM [] ===================================== compiler/GHC/Parser/PostProcess.hs ===================================== @@ -2907,7 +2907,7 @@ mkRdrRecordUpd overloaded_on exp@(L loc _) fbinds anns = do True -> do let qualifiedFields = [ L l lbl | L _ (HsFieldBind _ (L l lbl) _ _) <- fs' - , isQual . fieldOccRdrName $ lbl + , isQual . updFieldOccRdrName $ lbl ] case qualifiedFields of qf:_ -> addFatalError $ mkPlainErrorMsgEnvelope (getLocA qf) $ @@ -2953,7 +2953,7 @@ mk_rec_fields fs (Just s) = HsRecFields { rec_ext = noExtField, rec_flds = fs mk_rec_upd_field :: HsRecField GhcPs (LHsExpr GhcPs) -> HsRecUpdField GhcPs GhcPs mk_rec_upd_field (HsFieldBind noAnn (L loc (FieldOcc _ rdr)) arg pun) - = HsFieldBind noAnn (L loc (FieldOcc noExtField rdr)) arg pun + = HsFieldBind noAnn (L loc $ UpdFieldOcc noExtField (L loc (FieldOcc noExtField rdr))) arg pun mkInlinePragma :: SourceText -> (InlineSpec, RuleMatchInfo) -> Maybe Activation -> InlinePragma ===================================== compiler/GHC/Rename/Env.hs ===================================== @@ -1526,7 +1526,7 @@ lookupGlobalOccRn_overloaded rdr_name = return (Just gre) } getFieldUpdLbl :: IsPass p => LHsRecUpdField (GhcPass p) q -> LocatedN RdrName -getFieldUpdLbl = fieldOccLRdrName . unLoc . hfbLHS . unLoc +getFieldUpdLbl = updFieldOccLRdrName . unLoc . hfbLHS . unLoc -- | Returns all possible collections of field labels for the given -- record update. @@ -1623,7 +1623,7 @@ lookupRecUpdFields flds getUpdFieldLbls :: forall p q. IsPass p => [LHsRecUpdField (GhcPass p) q] -> [RdrName] getUpdFieldLbls - = map $ fieldOccRdrName + = map $ updFieldOccRdrName . unXRec @(GhcPass p) . hfbLHS . unXRec @(GhcPass p) ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -417,7 +417,7 @@ rnExpr (OpApp _ e1 op e2) -- should prevent bad things happening. ; fixity <- case op' of L _ (HsVar _ (L _ n)) -> lookupFixityRn n - L _ (HsRecSel _ f) -> fromJust <$> lookupFieldFixityRn f + L _ (HsRecSel _ f) -> lookupFieldFixityRn f _ -> return (Fixity minPrecedence InfixL) -- c.f. lookupFixity for unbound ===================================== compiler/GHC/Rename/Fixity.hs ===================================== @@ -201,6 +201,5 @@ lookupFixityRn_help name lookupTyFixityRn :: LocatedN Name -> RnM Fixity lookupTyFixityRn = lookupFixityRn . unLoc -lookupFieldFixityRn :: FieldOcc GhcRn -> RnM (Maybe Fixity) -lookupFieldFixityRn (FieldOcc _ n) = Just <$> lookupFixityRn (unLoc n) -lookupFieldFixityRn _ = pure Nothing +lookupFieldFixityRn :: FieldOcc GhcRn -> RnM Fixity +lookupFieldFixityRn (FieldOcc _ n) = lookupFixityRn (unLoc n) ===================================== compiler/GHC/Rename/HsType.hs ===================================== @@ -1574,13 +1574,7 @@ lookupFixityOp :: OpName -> RnM Fixity lookupFixityOp (NormalOp n) = lookupFixityRn n lookupFixityOp NegateOp = lookupFixityRn negateName lookupFixityOp (UnboundOp u) = lookupFixityRn (mkUnboundName (occName u)) -lookupFixityOp (RecFldOp f) = - -- We could get rid of this panic by parametrising FieldOcc with (k :: - -- MaybeAmbiguous), say 'PossiblyAmbiguous | 'Unambiguous, and using - -- 'PossiblyAmbiguous only for record field updates... but seems overkill - fromMaybe (panic "lookupFixityOp: RecFldOp should not be ambiguous!") <$> - lookupFieldFixityRn f - +lookupFixityOp (RecFldOp f) = lookupFieldFixityRn f -- Precedence-related error messages ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -1054,7 +1054,7 @@ rnHsRecUpdFields flds -> RnM ([LHsRecUpdField GhcRn GhcRn], FreeVars) rn_flds _ _ [] = return ([], emptyFVs) rn_flds pun_ok mb_unambig_lbls - ((L l (HsFieldBind { hfbLHS = L loc (FieldOcc _ f) + ((L l (HsFieldBind { hfbLHS = L loc (UpdFieldOcc _ (L _ (FieldOcc _ f))) , hfbRHS = arg , hfbPun = pun })):flds) = do { let lbl = unLoc f @@ -1066,19 +1066,23 @@ rnHsRecUpdFields flds ; return (L (l2l loc) (HsVar noExtField (L (l2l loc) arg_rdr))) } else return arg ; (arg'', fvs) <- rnLExpr arg' - ; let lbl' :: FieldOcc GhcRn + ; let lbl' :: UpdFieldOcc GhcRn lbl' = case mb_unambig_lbls of { Just (fl:_) -> let sel_name = flSelector fl - in FieldOcc lbl (L (l2l loc) sel_name) - ; _ -> XFieldOcc (Ambiguous (L (l2l loc) lbl)) } + in UpdFieldOcc noExtField (L (l2l loc) (FieldOcc lbl (L (l2l loc) sel_name))) + ; _ -> XUpdFieldOcc (Ambiguous (L (l2l loc) lbl)) } fld' :: LHsRecUpdField GhcRn GhcRn fld' = L l (HsFieldBind { hfbAnn = noAnn - , hfbLHS = L loc lbl' + , hfbLHS = L (l2l loc) lbl' , hfbRHS = arg'' , hfbPun = pun }) ; (flds', fvs') <- rn_flds pun_ok (tail <$> mb_unambig_lbls) flds ; return (fld' : flds', fvs `plusFV` fvs') } + -- I have been told this datacon isn't necessary and yet my ghc gets mad at me + -- when I don't add it. + rn_flds _ _ ((L _ (HsFieldBind { hfbLHS = L _ (XUpdFieldOcc impossible ) })):_) + = case impossible of getFieldIds :: [LHsRecField GhcRn arg] -> [Name] getFieldIds flds = map (hsRecFieldSel . unLoc) flds ===================================== compiler/GHC/Tc/Gen/Expr.hs ===================================== @@ -1300,8 +1300,9 @@ expandRecordUpd record_expr possible_parents rbnds res_ty -- See Note [Disambiguating record updates] in GHC.Rename.Pat. ; (cons, rbinds) <- disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty - ; let upd_flds = map (unLoc . hfbLHS . unLoc) rbinds - sel_ids = map (unLoc . foLabel) upd_flds + ; let upd_flds = map (unLoc . hfbLHS . unLoc) rbinds + -- upd_flds has type 'UpdFieldOcc GhcTc' so ufoField is not partial here. + sel_ids = map (unLoc . foLabel . unLoc . ufoField) upd_flds upd_fld_names = map idName sel_ids relevant_cons = nonDetEltsUniqSet cons relevant_con = head relevant_cons @@ -1584,7 +1585,7 @@ disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty -> TcM (LHsRecUpdField GhcTc GhcRn) lookupField fld_gre (L l upd) = do { let L loc af = hfbLHS upd - lbl = fieldOccRdrName af + lbl = updFieldOccRdrName af mb_gre = pickGREs lbl [fld_gre] -- NB: this GRE can be 'Nothing' when in GHCi. -- See test T10439. @@ -1596,7 +1597,7 @@ disambiguateRecordBinds record_expr record_rho possible_parents rbnds res_ty ; sel <- tcLookupId (greName fld_gre) ; return $ L l HsFieldBind { hfbAnn = hfbAnn upd - , hfbLHS = L (l2l loc) $ FieldOcc lbl (L (l2l loc) sel) + , hfbLHS = L (l2l loc) (UpdFieldOcc noExtField (L (l2l loc) $ FieldOcc lbl (L (l2l loc) sel))) , hfbRHS = hfbRHS upd , hfbPun = hfbPun upd } } @@ -1688,7 +1689,6 @@ tcRecordField con_like flds_w_tys (L loc (FieldOcc rdr (L l sel_name))) rhs ; return Nothing } where field_lbl = FieldLabelString $ occNameFS $ rdrNameOcc rdr -tcRecordField _ _ (L _ (XFieldOcc (Ambiguous _))) _ = pure Nothing checkMissingFields :: ConLike -> HsRecordBinds GhcRn -> [Scaled TcType] -> TcM () ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -567,7 +567,7 @@ tcInferAppHead_maybe :: HsExpr GhcRn tcInferAppHead_maybe fun = case fun of HsVar _ (L _ nm) -> Just <$> tcInferId nm - HsRecSel _ f -> tcInferRecSelId f + HsRecSel _ f -> Just <$> tcInferRecSelId f ExprWithTySig _ e hs_ty -> Just <$> tcExprWithSig e hs_ty HsOverLit _ lit -> Just <$> tcInferOverLit lit _ -> return Nothing @@ -596,11 +596,11 @@ addHeadCtxt fun_ctxt thing_inside ********************************************************************* -} tcInferRecSelId :: FieldOcc GhcRn - -> TcM (Maybe (HsExpr GhcTc, TcSigmaType)) + -> TcM ( (HsExpr GhcTc, TcSigmaType)) tcInferRecSelId (FieldOcc sel_name (L l n)) = do { sel_id <- tc_rec_sel_id ; let expr = HsRecSel noExtField (FieldOcc sel_name (L l sel_id)) - ; return $ Just(expr, idType sel_id) + ; return $ (expr, idType sel_id) } where occ :: OccName @@ -624,7 +624,6 @@ tcInferRecSelId (FieldOcc sel_name (L l n)) -- hence no checkTh stuff here _ -> failWithTc $ TcRnExpectedValueId thing } -tcInferRecSelId (XFieldOcc _) = pure Nothing ------------------------ ===================================== compiler/GHC/Tc/Gen/Pat.hs ===================================== @@ -1654,14 +1654,6 @@ tcConValArgs con_like arg_tys penv con_args thing_inside = case con_args of ; (pat', res) <- tcConArg penv (pat, pat_ty) thing_inside ; return (L l (HsFieldBind ann (L loc (FieldOcc rdr (L lr sel'))) pat' pun), res) } - tc_field _ - (L _ (HsFieldBind _ (L _ (XFieldOcc (Ambiguous (L _ _)))) _ _)) - _ - -- I don't like leaving things undefined, and I don't like leaving - -- leaving pattern not matched. I think I should just throw in this - -- case, but I don't know exactly how to do that. - = undefined - -- See Note [Omitted record fields and linearity] check_omitted_fields_multiplicity :: TcM MultiplicityCheckCoercions check_omitted_fields_multiplicity = do ===================================== compiler/GHC/ThToHs.hs ===================================== @@ -1159,7 +1159,7 @@ cvtl e = wrapLA (cvt e) ; return $ mkRdrRecordCon c' (HsRecFields noExtField flds' Nothing) noAnn } cvt (RecUpdE e flds) = do { e' <- cvtl e ; flds' - <- mapM (cvtFld (wrapParLA mkFieldOcc)) + <- mapM (cvtFld (wrapParLA mkUpdFieldOcc)) flds ; return $ RecordUpd noAnn e' $ RegularRecUpdFields ===================================== compiler/Language/Haskell/Syntax/Extension.hs ===================================== @@ -686,6 +686,9 @@ type family XXConDeclField x -- --------------------------------------------------------------------- -- FieldOcc type families +type family XCUpdFieldOcc x +type family XXUpdFieldOcc x + type family XCFieldOcc x type family XXFieldOcc x ===================================== compiler/Language/Haskell/Syntax/Pat.hs ===================================== @@ -326,7 +326,7 @@ type LHsRecUpdField p q = XRec p (HsRecUpdField p q) type HsRecField p arg = HsFieldBind (LFieldOcc p) arg -- | Haskell Record Update Field -type HsRecUpdField p q = HsFieldBind (LFieldOcc p) (LHsExpr q) +type HsRecUpdField p q = HsFieldBind (LUpdFieldOcc p) (LHsExpr q) -- | Haskell Field Binding -- ===================================== compiler/Language/Haskell/Syntax/Type.hs ===================================== @@ -50,6 +50,7 @@ module Language.Haskell.Syntax.Type ( HsConDetails(..), noTypeArgs, + UpdFieldOcc(..), LUpdFieldOcc, FieldOcc(..), LFieldOcc, mapHsOuterImplicit, @@ -1265,6 +1266,14 @@ The SrcSpan is the span of the original HsPar * * ************************************************************************ -} +type LUpdFieldOcc pass = XRec pass (UpdFieldOcc pass) + +data UpdFieldOcc pass + = UpdFieldOcc { + ufoExt :: XCUpdFieldOcc pass, + ufoField :: LFieldOcc pass + } + | XUpdFieldOcc (XXUpdFieldOcc pass) -- | Located Field Occurrence type LFieldOcc pass = XRec pass (FieldOcc pass) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4775c0facf18503d21a5544fb61b51f0129ef946 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/4775c0facf18503d21a5544fb61b51f0129ef946 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 14:39:51 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Wed, 25 Sep 2024 10:39:51 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] 15 commits: Demand: Combine examples into Note (#25107) Message-ID: <66f420b77652_123102d7cf86216@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - f4cb5db9 by Jade at 2024-09-25T16:39:36+02:00 Refactor FieldOcc vs AmbiguousFieldOcc with TTG Improves the design of FieldOcc vs AmbiguousFieldOcc, and removes a dependency on `RdrName` from the Language.Haskell.* namespace (#21592). The design: * The FieldOcc constructor of FieldOcc always refers to an unambiguous field occurrence. * During renaming, a FieldOcc may be ambiguous and only be resolvable during Typechecking * Therefore, we extend (with TTG) `FieldOcc GhcRn` with a constructor `AmbiguousFieldOcc` that constructs a definitely ambiguous `FieldOcc`. * During typechecking, all ambiguous field occurrences must be resolved, so the `AmbiguousFieldOcc` constructor no longer exists See Note [Lifecycle of a FieldOcc] Co-authored-by: Rodrigo Mesquita <rodrigo.m.mesquita at gmail.com> - - - - - 20bc6759 by Hassan Al-Awwadi at 2024-09-25T16:39:36+02:00 Wrangled until everything compiled. There are two issues: - FieldOcc used to only have one constructor and always be unambigious, this is no longer the case. Calls to foLabel are thus partial now. Don't know how much we care about this, since the partial calls are mostly inside functions that used to operate on the operate on the unambigious FieldOcc - Lots of functions that take in a FieldOcc, or a HsExpr (with the HsRecSel constructor) now have a new case. It was not always clear to me what the correct implementation was for these. I have filled them in as far as I could and left one undefined... - - - - - 02c78124 by Hassan Al-Awwadi at 2024-09-25T16:39:36+02:00 fixed ambiguity regression - - - - - 30 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs.hs - compiler/GHC/Hs/Binds.hs - compiler/GHC/Hs/Extension.hs - compiler/GHC/Hs/Instances.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/Hs/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Docs.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Bind.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Rename/Fixity.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Rename/Pat.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4775c0facf18503d21a5544fb61b51f0129ef946...02c781249e01800380ddab8d8bb3ff9df9b03ad2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4775c0facf18503d21a5544fb61b51f0129ef946...02c781249e01800380ddab8d8bb3ff9df9b03ad2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 15:15:44 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 25 Sep 2024 11:15:44 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/romes/det-llvm Message-ID: <66f42920b03d7_123102503458237cd@gitlab.mail> Rodrigo Mesquita pushed new branch wip/romes/det-llvm at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/romes/det-llvm You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 15:19:11 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Wed, 25 Sep 2024 11:19:11 -0400 Subject: [Git][ghc/ghc][wip/romes/det-llvm] determinism: Deterministic MonadGetUnique LlvmM Message-ID: <66f429efa83bf_1231026527f02691a@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/det-llvm at Glasgow Haskell Compiler / GHC Commits: 165f0b4b by Rodrigo Mesquita at 2024-09-25T16:18:08+01:00 determinism: Deterministic MonadGetUnique LlvmM Update LlvmM to thread a unique deterministic supply (using UniqDSMT), and use it in the MonadGetUnique instance. This makes uniques sampled from LlvmM deterministic, which guarantees object determinism with -fllvm. Fixes #25274 - - - - - 2 changed files: - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs Changes: ===================================== compiler/GHC/CmmToLlvm.hs ===================================== @@ -85,25 +85,23 @@ llvmCodeGen logger cfg h dus cmm_stream llvm_ver = fromMaybe supportedLlvmVersionLowerBound mb_ver -- run code generation - a <- runLlvm logger cfg llvm_ver bufh $ - llvmCodeGen' cfg dus cmm_stream + (a, _) <- runLlvm logger cfg llvm_ver bufh dus $ + llvmCodeGen' cfg cmm_stream bFlush bufh return a llvmCodeGen' :: LlvmCgConfig - -> DUniqSupply -- ^ The deterministic uniq supply to run the CgStream. - -- See Note [Deterministic Uniques in the CG] -> CgStream RawCmmGroup a -> LlvmM a -llvmCodeGen' cfg dus cmm_stream +llvmCodeGen' cfg cmm_stream = do -- Preamble renderLlvm (llvmHeader cfg) (llvmHeader cfg) ghcInternalFunctions cmmMetaLlvmPrelude -- Procedures - (a, _) <- runUDSMT dus $ Stream.consume cmm_stream (hoistUDSMT liftIO) (liftUDSMT . llvmGroupLlvmGens) + a <- Stream.consume cmm_stream (GHC.CmmToLlvm.Base.liftUDSMT) (llvmGroupLlvmGens) -- Declare aliases for forward references decls <- generateExternDecls ===================================== compiler/GHC/CmmToLlvm/Base.hs ===================================== @@ -23,7 +23,7 @@ module GHC.CmmToLlvm.Base ( ghcInternalFunctions, getPlatform, getConfig, getMetaUniqueId, - setUniqMeta, getUniqMeta, liftIO, + setUniqMeta, getUniqMeta, liftIO, liftUDSMT, cmmToLlvmType, widthToLlvmFloat, widthToLlvmInt, llvmFunTy, llvmFunSig, llvmFunArgs, llvmStdFunAttrs, llvmFunAlign, llvmInfAlign, @@ -55,7 +55,6 @@ import GHC.Types.Unique.FM import GHC.Types.Unique import GHC.Utils.BufHandle ( BufHandle ) import GHC.Types.Unique.Set -import GHC.Types.Unique.Supply import qualified GHC.Types.Unique.DSM as DSM import GHC.Utils.Logger @@ -64,6 +63,7 @@ import Control.Monad.Trans.State (StateT (..)) import Data.List (isPrefixOf) import qualified Data.List.NonEmpty as NE import Data.Ord (comparing) +import qualified Control.Monad.IO.Class as IO -- ---------------------------------------------------------------------------- -- * Some Data Types @@ -277,14 +277,13 @@ data LlvmEnv = LlvmEnv type LlvmEnvMap = UniqFM Unique LlvmType -- | The Llvm monad. Wraps @LlvmEnv@ state as well as the @IO@ monad -newtype LlvmM a = LlvmM { runLlvmM :: LlvmEnv -> IO (a, LlvmEnv) } +newtype LlvmM a = LlvmM { runLlvmM :: LlvmEnv -> DSM.UniqDSMT IO (a, LlvmEnv) } deriving stock (Functor) - deriving (Applicative, Monad) via StateT LlvmEnv IO + deriving (Applicative, Monad) via StateT LlvmEnv (DSM.UniqDSMT IO) instance HasLogger LlvmM where getLogger = LlvmM $ \env -> return (envLogger env, env) - -- | Get target platform getPlatform :: LlvmM Platform getPlatform = llvmCgPlatform <$> getConfig @@ -293,23 +292,30 @@ getConfig :: LlvmM LlvmCgConfig getConfig = LlvmM $ \env -> return (envConfig env, env) --- TODO(#25274): If you want Llvm code to be deterministic, this instance should use a --- deterministic unique supply to produce uniques, rather than using 'uniqFromTag'. +-- This instance uses a deterministic unique supply from UniqDSMT, so new +-- uniques within LlvmM will be sampled deterministically. instance DSM.MonadGetUnique LlvmM where getUniqueM = do tag <- getEnv envTag - liftIO $! uniqFromTag tag + liftUDSMT $! do + uq <- DSM.getUniqueM + return (newTagUnique uq tag) -- | Lifting of IO actions. Not exported, as we want to encapsulate IO. liftIO :: IO a -> LlvmM a -liftIO m = LlvmM $ \env -> do x <- m +liftIO m = LlvmM $ \env -> do x <- IO.liftIO m return (x, env) +-- | Lifting of UniqDSMT actions. Gives access to the deterministic unique supply being threaded through by LlvmM. +liftUDSMT :: DSM.UniqDSMT IO a -> LlvmM a +liftUDSMT m = LlvmM $ \env -> do x <- m + return (x, env) + -- | Get initial Llvm environment. -runLlvm :: Logger -> LlvmCgConfig -> LlvmVersion -> BufHandle -> LlvmM a -> IO a -runLlvm logger cfg ver out m = do - (a, _) <- runLlvmM m env - return a +runLlvm :: Logger -> LlvmCgConfig -> LlvmVersion -> BufHandle -> DSM.DUniqSupply -> LlvmM a -> IO (a, DSM.DUniqSupply) +runLlvm logger cfg ver out us m = do + ((a, _), us') <- DSM.runUDSMT us $ runLlvmM m env + return (a, us') where env = LlvmEnv { envFunMap = emptyUFM , envVarMap = emptyUFM , envStackRegs = [] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/165f0b4b9f8548d9cba99a8c8fd4bd70a36f178c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/165f0b4b9f8548d9cba99a8c8fd4bd70a36f178c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 15:53:53 2024 From: gitlab at gitlab.haskell.org (Brandon Chinn (@brandonchinn178)) Date: Wed, 25 Sep 2024 11:53:53 -0400 Subject: [Git][ghc/ghc][wip/strings] Replace manual string lexing (#25158) Message-ID: <66f43210e4417_1231029523104788c@gitlab.mail> Brandon Chinn pushed to branch wip/strings at Glasgow Haskell Compiler / GHC Commits: 92f9a4df by Brandon Chinn at 2024-09-25T08:48:02-07:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 29 changed files: - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - testsuite/driver/testlib.py - testsuite/tests/ghci/prog013/prog013.stderr - testsuite/tests/ghci/scripts/ghci022.stderr - testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr - testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr - testsuite/tests/parser/should_fail/T21843c.stderr - testsuite/tests/parser/should_fail/T21843e.stderr - testsuite/tests/parser/should_fail/T21843f.stderr - testsuite/tests/parser/should_fail/T3751.stderr - testsuite/tests/parser/should_fail/T5425.stderr - testsuite/tests/parser/should_fail/readFail002.stderr - testsuite/tests/parser/should_fail/readFail004.stderr - testsuite/tests/parser/should_fail/readFail005.stderr - testsuite/tests/parser/should_fail/readFail033.stderr - testsuite/tests/parser/unicode/all.T - + testsuite/tests/parser/unicode/lex_unicode_ids.hs - + testsuite/tests/parser/unicode/lex_unispace.hs - testsuite/tests/parser/unicode/utf8_010.stderr - testsuite/tests/parser/unicode/utf8_011.stderr - testsuite/tests/parser/unicode/utf8_020.stderr - testsuite/tests/parser/unicode/utf8_021.stderr - testsuite/tests/parser/unicode/utf8_022.stderr - utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs Changes: ===================================== compiler/GHC/Parser.y ===================================== @@ -728,7 +728,7 @@ are the most common patterns, rewritten as regular expressions for clarity: CHAR { L _ (ITchar _ _) } STRING { L _ (ITstring _ _) } - MULTILINESTRING { L _ (ITmultilinestring _ _) } + STRING_MULTI { L _ (ITstringMulti _ _) } INTEGER { L _ (ITinteger _) } RATIONAL { L _ (ITrational _) } @@ -2357,8 +2357,8 @@ atype :: { LHsType GhcPs } (getCHAR $1) } | STRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGs $1) (getSTRING $1) } - | MULTILINESTRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getMULTILINESTRINGs $1) - (getMULTILINESTRING $1) } + | STRING_MULTI { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGMULTIs $1) + (getSTRINGMULTI $1) } | '_' { sL1a $1 $ mkAnonWildCardTy } -- Type variables are never exported, so `M.tyvar` will be rejected by the renamer. -- We let it pass the parser because the renamer can generate a better error message. @@ -4047,8 +4047,8 @@ literal :: { Located (HsLit GhcPs) } : CHAR { sL1 $1 $ HsChar (getCHARs $1) $ getCHAR $1 } | STRING { sL1 $1 $ HsString (getSTRINGs $1) $ getSTRING $1 } - | MULTILINESTRING { sL1 $1 $ HsMultilineString (getMULTILINESTRINGs $1) - $ getMULTILINESTRING $1 } + | STRING_MULTI { sL1 $1 $ HsMultilineString (getSTRINGMULTIs $1) + $ getSTRINGMULTI $1 } | PRIMINTEGER { sL1 $1 $ HsIntPrim (getPRIMINTEGERs $1) $ getPRIMINTEGER $1 } | PRIMWORD { sL1 $1 $ HsWordPrim (getPRIMWORDs $1) @@ -4154,7 +4154,7 @@ getIPDUPVARID (L _ (ITdupipvarid x)) = x getLABELVARID (L _ (ITlabelvarid _ x)) = x getCHAR (L _ (ITchar _ x)) = x getSTRING (L _ (ITstring _ x)) = x -getMULTILINESTRING (L _ (ITmultilinestring _ x)) = x +getSTRINGMULTI (L _ (ITstringMulti _ x)) = x getINTEGER (L _ (ITinteger x)) = x getRATIONAL (L _ (ITrational x)) = x getPRIMCHAR (L _ (ITprimchar _ x)) = x @@ -4180,7 +4180,7 @@ getVOCURLY (L (RealSrcSpan l _) ITvocurly) = srcSpanStartCol l getINTEGERs (L _ (ITinteger (IL src _ _))) = src getCHARs (L _ (ITchar src _)) = src getSTRINGs (L _ (ITstring src _)) = src -getMULTILINESTRINGs (L _ (ITmultilinestring src _)) = src +getSTRINGMULTIs (L _ (ITstringMulti src _)) = src getPRIMCHARs (L _ (ITprimchar src _)) = src getPRIMSTRINGs (L _ (ITprimstring src _)) = src getPRIMINTEGERs (L _ (ITprimint src _)) = src ===================================== compiler/GHC/Parser/CharClass.hs ===================================== @@ -36,14 +36,14 @@ cDigit = 64 {-# INLINABLE is_ctype #-} is_ctype :: Word8 -> Char -> Bool -is_ctype mask c = (charType c .&. mask) /= 0 +is_ctype mask c = c <= '\127' && (charType c .&. mask) /= 0 is_ident, is_symbol, is_any, is_space, is_lower, is_upper, is_digit, is_alphanum :: Char -> Bool is_ident = is_ctype cIdent is_symbol = is_ctype cSymbol is_any = is_ctype cAny -is_space = \c -> c <= '\x7f' && is_ctype cSpace c -- is_space only works for <= '\x7f' (#3751, #5425) +is_space = is_ctype cSpace is_lower = is_ctype cLower is_upper = is_ctype cUpper is_digit = is_ctype cDigit ===================================== compiler/GHC/Parser/Errors/Ppr.hs ===================================== @@ -157,8 +157,6 @@ instance Diagnostic PsMessage where LexUnknownPragma -> text "unknown pragma" LexErrorInPragma -> text "lexical error in pragma" LexNumEscapeRange -> text "numeric escape sequence out of range" - LexStringCharLit -> text "lexical error in string/character literal" - LexStringCharLitEOF -> text "unexpected end-of-file in string/character literal" LexUnterminatedComment -> text "unterminated `{-'" LexUnterminatedOptions -> text "unterminated OPTIONS pragma" LexUnterminatedQQ -> text "unterminated quasiquotation" ===================================== compiler/GHC/Parser/Errors/Types.hs ===================================== @@ -591,8 +591,6 @@ data LexErr | LexUnknownPragma -- ^ Unknown pragma | LexErrorInPragma -- ^ Lexical error in pragma | LexNumEscapeRange -- ^ Numeric escape sequence out of range - | LexStringCharLit -- ^ Lexical error in string/character literal - | LexStringCharLitEOF -- ^ Unexpected end-of-file in string/character literal | LexUnterminatedComment -- ^ Unterminated `{-' | LexUnterminatedOptions -- ^ Unterminated OPTIONS pragma | LexUnterminatedQQ -- ^ Unterminated quasiquotation ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -140,7 +140,8 @@ import GHC.Parser.String -- Any changes here should likely be reflected there. $unispace = \x05 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $nl = [\n\r\f] -$whitechar = [$nl\v\ $unispace] +$space = [\ $unispace] +$whitechar = [$nl \v $space] $white_no_nl = $whitechar # \n -- TODO #8424 $tab = \t @@ -167,6 +168,7 @@ $idchar = [$small $large $digit $uniidchar \'] $unigraphic = \x06 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $graphic = [$small $large $symbol $digit $idchar $special $unigraphic \"\'] +$charesc = [a b f n r t v \\ \" \' \&] $binit = 0-1 $octit = 0-7 @@ -213,6 +215,20 @@ $docsym = [\| \^ \* \$] @floating_point = @numspc @decimal \. @decimal @exponent? | @numspc @decimal @exponent @hex_floating_point = @numspc @hexadecimal \. @hexadecimal @bin_exponent? | @numspc @hexadecimal @bin_exponent + at gap = \\ $whitechar+ \\ + at cntrl = $asclarge | \@ | \[ | \\ | \] | \^ | \_ + at ascii = \^ @cntrl | "NUL" | "SOH" | "STX" | "ETX" | "EOT" | "ENQ" | "ACK" + | "BEL" | "BS" | "HT" | "LF" | "VT" | "FF" | "CR" | "SO" | "SI" | "DLE" + | "DC1" | "DC2" | "DC3" | "DC4" | "NAK" | "SYN" | "ETB" | "CAN" + | "EM" | "SUB" | "ESC" | "FS" | "GS" | "RS" | "US" | "SP" | "DEL" +-- N.B. ideally, we would do `@escape # \\ \&` instead of duplicating in @escapechar, +-- which is what the Haskell Report says, but this isn't valid Alex syntax, as only +-- character sets can be subtracted, not strings + at escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) + at escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) + at stringchar = ($graphic # [\\ \"]) | $space | @escape | @gap + at char = ($graphic # [\\ \']) | $space | @escapechar + -- normal signed numerical literals can only be explicitly negative, -- not explicitly positive (contrast @exponent) @negative = \- @@ -460,7 +476,7 @@ $unigraphic / { isSmartQuote } { smart_quote_error } <0> { "#" $idchar+ / { ifExtension OverloadedLabelsBit } { skip_one_varid_src ITlabelvarid } - "#" \" / { ifExtension OverloadedLabelsBit } { lex_quoted_label } + "#" \" @stringchar* \" / { ifExtension OverloadedLabelsBit } { tok_quoted_label } } <0> { @@ -660,14 +676,38 @@ $unigraphic / { isSmartQuote } { smart_quote_error } } --- Strings and chars are lexed by hand-written code. The reason is --- that even if we recognise the string or char here in the regex --- lexer, we would still have to parse the string afterward in order --- to convert it to a String. <0> { - \' { lex_char_tok } - \"\"\" / { ifExtension MultilineStringsBit} { lex_string_tok StringTypeMulti } - \" { lex_string_tok StringTypeSingle } + \"\"\" / { ifExtension MultilineStringsBit } { tok_string_multi } + \" @stringchar* \" { tok_string } + \" @stringchar* \" \# / { ifExtension MagicHashBit } { tok_string } + \' @char \' { tok_char } + \' @char \' \# / { ifExtension MagicHashBit } { tok_char } + + -- Check for smart quotes and throw better errors than a plain lexical error (#21843) + \' \\ $unigraphic / { isSmartQuote } { smart_quote_error } + \" @stringchar* \\ $unigraphic / { isSmartQuote } { smart_quote_error } + -- See Note [Bare smart quote error] + -- The valid string rule will take precedence because it'll match more + -- characters than this rule, so this rule will only fire if the string + -- could not be lexed correctly + \" @stringchar* $unigraphic / { isSmartQuote } { smart_quote_error } +} + + { + -- Parse as much of the multiline string as possible, except for quotes + @stringchar* ($nl ([\ $tab] | @gap)* @stringchar*)* { tok_string_multi_content } + -- Allow bare quotes if it's not a triple quote + (\" | \"\") / ([\n .] # \") { tok_string_multi_content } +} + +<0> { + \'\' { token ITtyQuote } + + -- The normal character match takes precedence over this because it matches + -- more characters. However, if that pattern didn't match, then this quote + -- could be a quoted identifier, like 'x. Here, just return ITsimpleQuote, + -- as the parser will lex the varid separately. + \' / ($graphic # \\ | " ") { token ITsimpleQuote } } -- Note [Whitespace-sensitive operator parsing] @@ -953,7 +993,7 @@ data Token | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" | ITstring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" - | ITmultilinestring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringMulti SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" | ITinteger IntegralLit -- Note [Literal source text] in "GHC.Types.SourceText" | ITrational FractionalLit @@ -2181,156 +2221,128 @@ lex_string_prag_comment mkTok span _buf _len _buf2 -- ----------------------------------------------------------------------------- -- Strings & Chars --- This stuff is horrible. I hates it. - -lex_string_tok :: LexStringType -> Action -lex_string_tok strType span buf _len _buf2 = do - s <- lex_string strType - - i <- getInput - case strType of - StringTypeSingle -> - lex_magic_hash i >>= \case - Just i' -> do - when (any (> '\xFF') s) $ do - pState <- getPState - let msg = PsErrPrimStringInvalidChar - let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg - addError err - - setInput i' - let (psSpan, src) = getStringLoc (buf, locStart) i' - pure $ L psSpan (ITprimstring src (unsafeMkByteString s)) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITstring src (mkFastString s)) - StringTypeMulti -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITmultilinestring src (mkFastString s)) +tok_string :: Action +tok_string span buf len _buf2 = do + s <- lex_chars ("\"", "\"") span buf (if endsInHash then len - 1 else len) + + if endsInHash + then do + when (any (> '\xFF') s) $ do + pState <- getPState + let msg = PsErrPrimStringInvalidChar + let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg + addError err + pure $ L span (ITprimstring src (unsafeMkByteString s)) + else + pure $ L span (ITstring src (mkFastString s)) where - locStart = psSpanStart span + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + +-- | Ideally, we would define this completely with Alex syntax, like normal strings. +-- Instead, this is defined as a hybrid solution by manually invoking lex states, which +-- we're doing for two reasons: +-- 1. The multiline string should all be one lexical token, not multiple +-- 2. We need to allow bare quotes, which can't be done with one regex +tok_string_multi :: Action +tok_string_multi startSpan startBuf _len _buf2 = do + -- advance to the end of the multiline string + let startLoc = psSpanStart startSpan + let i@(AI _ contentStartBuf) = + case lexDelim $ AI startLoc startBuf of + Just i -> i + Nothing -> panic "tok_string_multi did not start with a delimiter" + (AI _ contentEndBuf, i'@(AI endLoc endBuf)) <- goContent i + + -- build the values pertaining to the entire multiline string, including delimiters + let span = mkPsSpan startLoc endLoc + let len = byteDiff startBuf endBuf + let src = SourceText $ lexemeToFastString startBuf len + + -- load the content of the multiline string + let contentLen = byteDiff contentStartBuf contentEndBuf + s <- + either (throwStringLexError (AI startLoc startBuf)) pure $ + lexMultilineString contentLen contentStartBuf + + setInput i' + pure $ L span $ ITstringMulti src (mkFastString s) + where + goContent i0 = + case alexScan i0 string_multi_content of + AlexToken i1 len _ + | Just i2 <- lexDelim i1 -> pure (i1, i2) + | -- is the next token a tab character? + -- need this explicitly because there's a global rule matching $tab + Just ('\t', _) <- alexGetChar' i1 -> setInput i1 >> lexError LexError + | isEOF i1 -> checkSmartQuotes >> lexError LexError + | len == 0 -> panic $ "parsing multiline string got into infinite loop at: " ++ show i0 + | otherwise -> goContent i1 + AlexSkip i1 _ -> goContent i1 + _ -> lexError LexError + + lexDelim = + let go 0 i = Just i + go n i = + case alexGetChar' i of + Just ('"', i') -> go (n - 1) i' + _ -> Nothing + in go (3 :: Int) + + -- See Note [Bare smart quote error] + checkSmartQuotes = do + let findSmartQuote i0@(AI loc _) = + case alexGetChar' i0 of + Just ('\\', i1) | Just (_, i2) <- alexGetChar' i1 -> findSmartQuote i2 + Just (c, i1) + | isDoubleSmartQuote c -> Just (c, loc) + | otherwise -> findSmartQuote i1 + _ -> Nothing + case findSmartQuote (AI (psSpanStart startSpan) startBuf) of + Just (c, loc) -> throwSmartQuoteError c loc + Nothing -> pure () + +-- | Dummy action that should never be called. Should only be used in lex states +-- that are manually lexed in tok_string_multi. +tok_string_multi_content :: Action +tok_string_multi_content = panic "tok_string_multi_content unexpectedly invoked" + +lex_chars :: (String, String) -> PsSpan -> StringBuffer -> Int -> P String +lex_chars (startDelim, endDelim) span buf len = + either (throwStringLexError i0) pure $ + lexString contentLen contentBuf + where + i0@(AI _ contentBuf) = advanceInputBytes (length startDelim) $ AI (psSpanStart span) buf + -- assumes delimiters are ASCII, with 1 byte per Char + contentLen = len - length startDelim - length endDelim -lex_quoted_label :: Action -lex_quoted_label span buf _len _buf2 = do - s <- lex_string StringTypeSingle - (AI end bufEnd) <- getInput - let - token = ITlabelvarid (SourceText src) (mkFastString s) - src = lexemeToFastString (stepOn buf) (cur bufEnd - cur buf - 1) - start = psSpanStart span - - return $ L (mkPsSpan start end) token - - -lex_string :: LexStringType -> P String -lex_string strType = do - start <- getInput - (str, next) <- either fromStringLexError pure $ lexString strType alexGetChar' start - setInput next - pure str - - -lex_char_tok :: Action --- Here we are basically parsing character literals, such as 'x' or '\n' --- but we additionally spot 'x and ''T, returning ITsimpleQuote and --- ITtyQuote respectively, but WITHOUT CONSUMING the x or T part --- (the parser does that). --- So we have to do two characters of lookahead: when we see 'x we need to --- see if there's a trailing quote -lex_char_tok span buf _len _buf2 = do -- We've seen ' - i1 <- getInput -- Look ahead to first character - let loc = psSpanStart span - case alexGetChar' i1 of - Nothing -> lit_error i1 - - Just ('\'', i2@(AI end2 _)) -> do -- We've seen '' - setInput i2 - return (L (mkPsSpan loc end2) ITtyQuote) - - Just ('\\', i2@(AI end2 _)) -> do -- We've seen 'backslash - (lit_ch, i3) <- - either fromStringLexError pure $ - resolveEscapeCharacter alexGetChar' i2 - case alexGetChar' i3 of - Just ('\'', i4) -> do - setInput i4 - finish_char_tok buf loc lit_ch - Just (mc, _) | isSingleSmartQuote mc -> add_smart_quote_error mc end2 - _ -> lit_error i3 - - Just (c, i2@(AI end2 _)) - | not (isAnyChar c) -> lit_error i1 - | otherwise -> - - -- We've seen 'x, where x is a valid character - -- (i.e. not newline etc) but not a quote or backslash - case alexGetChar' i2 of -- Look ahead one more character - Just ('\'', i3) -> do -- We've seen 'x' - setInput i3 - finish_char_tok buf loc c - Just (c, _) | isSingleSmartQuote c -> add_smart_quote_error c end2 - _other -> do -- We've seen 'x not followed by quote - -- (including the possibility of EOF) - -- Just parse the quote only - let (AI end _) = i1 - return (L (mkPsSpan loc end) ITsimpleQuote) - --- We've already seen the closing quote --- Just need to check for trailing # -finish_char_tok :: StringBuffer -> PsLoc -> Char -> P (PsLocated Token) -finish_char_tok buf loc ch = do - i <- getInput - lex_magic_hash i >>= \case - Just i' -> do - setInput i' - -- Include the trailing # in SourceText - let (psSpan, src) = getStringLoc (buf, loc) i' - pure $ L psSpan (ITprimchar src ch) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, loc) i - pure $ L psSpan (ITchar src ch) +throwStringLexError :: AlexInput -> StringLexError -> P a +throwStringLexError i (StringLexError e pos) = setInput (advanceInputTo pos i) >> lexError e --- | Get the span and source text for a string from the given start to the given end. -getStringLoc :: (StringBuffer, PsLoc) -> AlexInput -> (PsSpan, SourceText) -getStringLoc (bufStart, locStart) (AI locEnd bufEnd) = (psSpan, SourceText src) +tok_quoted_label :: Action +tok_quoted_label span buf len _buf2 = do + s <- lex_chars ("#\"", "\"") span buf len + pure $ L span (ITlabelvarid src (mkFastString s)) where - psSpan = mkPsSpan locStart locEnd - src = lexemeToFastString bufStart (cur bufEnd - cur bufStart) - - --- Return Just if we found the magic hash, with the next input. -lex_magic_hash :: AlexInput -> P (Maybe AlexInput) -lex_magic_hash i = do - magicHash <- getBit MagicHashBit - if magicHash - then - case alexGetChar' i of - Just ('#', i') -> pure (Just i') - _other -> pure Nothing - else pure Nothing - -fromStringLexError :: StringLexError AlexInput -> P a -fromStringLexError = \case - UnexpectedEOF i squote -> checkSQuote squote >> throw i LexStringCharLitEOF - BadCharInitialLex i squote -> checkSQuote squote >> throw i LexStringCharLit - EscapeBadChar i -> throw i LexStringCharLit - EscapeUnexpectedEOF i -> throw i LexStringCharLitEOF - EscapeNumRangeError i -> throw i LexNumEscapeRange - EscapeSmartQuoteError c (AI loc _) -> add_smart_quote_error c loc + -- skip leading '#' + src = SourceText . mkFastString . drop 1 $ lexemeToString buf len + + +tok_char :: Action +tok_char span buf len _buf2 = do + c <- lex_chars ("'", "'") span buf (if endsInHash then len - 1 else len) >>= \case + [c] -> pure c + s -> panic $ "tok_char expected exactly one character, got: " ++ show s + pure . L span $ + if endsInHash + then ITprimchar src c + else ITchar src c where - throw i e = setInput i >> lexError e - checkSQuote = \case - NoSmartQuote -> pure () - SmartQuote c (AI loc _) -> add_nonfatal_smart_quote_error c loc - --- before calling lit_error, ensure that the current input is pointing to --- the position of the error in the buffer. This is so that we can report --- a correct location to the user, but also so we can detect UTF-8 decoding --- errors if they occur. -lit_error :: AlexInput -> P a -lit_error i = do setInput i; lexError LexStringCharLit + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + -- ----------------------------------------------------------------------------- -- QuasiQuote @@ -2389,32 +2401,28 @@ quasiquote_error start = do isSmartQuote :: AlexAccPred ExtsBitmap isSmartQuote _ _ _ (AI _ buf) = let c = prevChar buf ' ' in isSingleSmartQuote c || isDoubleSmartQuote c -smart_quote_error_message :: Char -> PsLoc -> MsgEnvelope PsMessage -smart_quote_error_message c loc = - let (correct_char, correct_char_name) = - if isSingleSmartQuote c then ('\'', "Single Quote") else ('"', "Quotation Mark") - err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ - PsErrUnicodeCharLooksLike c correct_char correct_char_name in - err - +throwSmartQuoteError :: Char -> PsLoc -> P a +throwSmartQuoteError c loc = addFatalError err + where + err = + mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ + PsErrUnicodeCharLooksLike c correct_char correct_char_name + (correct_char, correct_char_name) = + if isSingleSmartQuote c + then ('\'', "Single Quote") + else ('"', "Quotation Mark") + +-- | Throw a smart quote error, where the smart quote was the last character lexed smart_quote_error :: Action -smart_quote_error span buf _len _buf2 = do - let c = currentChar buf - addFatalError (smart_quote_error_message c (psSpanStart span)) - -add_smart_quote_error :: Char -> PsLoc -> P a -add_smart_quote_error c loc = addFatalError (smart_quote_error_message c loc) - -add_nonfatal_smart_quote_error :: Char -> PsLoc -> P () -add_nonfatal_smart_quote_error c loc = addError (smart_quote_error_message c loc) +smart_quote_error span _ _ buf2 = do + let c = prevChar buf2 (panic "smart_quote_error unexpectedly called on beginning of input") + throwSmartQuoteError c (psSpanStart span) -advance_to_smart_quote_character :: P () -advance_to_smart_quote_character = do - i <- getInput - case alexGetChar' i of - Just (c, _) | isDoubleSmartQuote c -> return () - Just (_, i2) -> do setInput i2; advance_to_smart_quote_character - Nothing -> return () -- should never get here +-- Note [Bare smart quote error] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- A smart quote inside of a string is allowed, but if a complete valid string +-- couldn't be lexed, we want to see if there's a smart quote that the user +-- thought ended the string, but in fact didn't. -- ----------------------------------------------------------------------------- -- Warnings @@ -2652,7 +2660,7 @@ getLastLocIncludingComments = P $ \s@(PState { prev_loc = prev_loc }) -> POk s p getLastLoc :: P PsSpan getLastLoc = P $ \s@(PState { last_loc = last_loc }) -> POk s last_loc -data AlexInput = AI !PsLoc !StringBuffer +data AlexInput = AI !PsLoc !StringBuffer deriving (Show) {- Note [Unicode in Alex] @@ -2763,6 +2771,19 @@ alexGetChar' (AI loc s) where (c,s') = nextChar s loc' = advancePsLoc loc c +-- | Advance the given input N bytes. +advanceInputBytes :: Int -> AlexInput -> AlexInput +advanceInputBytes n i0@(AI _ buf0) = advanceInputTo (cur buf0 + n) i0 + +-- | Advance the given input to the given position. +advanceInputTo :: Int -> AlexInput -> AlexInput +advanceInputTo pos = go + where + go i@(AI _ buf) + | cur buf >= pos = i + | Just (_, i') <- alexGetChar' i = go i' + | otherwise = i -- reached the end, just return the last input + getInput :: P AlexInput getInput = P $ \s at PState{ loc=l, buffer=b } -> POk s (AI l b) @@ -2770,9 +2791,10 @@ setInput :: AlexInput -> P () setInput (AI l b) = P $ \s -> POk s{ loc=l, buffer=b } () nextIsEOF :: P Bool -nextIsEOF = do - AI _ s <- getInput - return $ atEnd s +nextIsEOF = isEOF <$> getInput + +isEOF :: AlexInput -> Bool +isEOF (AI _ buf) = atEnd buf pushLexState :: Int -> P () pushLexState ls = P $ \s at PState{ lex_state=l } -> POk s{lex_state=ls:l} () @@ -3516,6 +3538,11 @@ topNoLayoutContainsCommas [] = False topNoLayoutContainsCommas (ALRLayout _ _ : ls) = topNoLayoutContainsCommas ls topNoLayoutContainsCommas (ALRNoLayout b _ : _) = b +-- If the generated alexScan/alexScanUser functions are called multiple times +-- in this file, alexScanUser gets broken out into a separate function and +-- increases memory usage. Make sure GHC inlines this function and optimizes it. +{-# INLINE alexScanUser #-} + lexToken :: P (PsLocated Token) lexToken = do inp@(AI loc1 buf) <- getInput ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -1,284 +1,199 @@ {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE PatternSynonyms #-} +{-# LANGUAGE TupleSections #-} +{-# LANGUAGE ViewPatterns #-} module GHC.Parser.String ( StringLexError (..), - ContainsSmartQuote (..), - LexStringType (..), lexString, + lexMultilineString, -- * Unicode smart quote helpers isDoubleSmartQuote, isSingleSmartQuote, - - -- * Other helpers - isAnyChar, - resolveEscapeCharacter, ) where -import GHC.Prelude +import GHC.Prelude hiding (getChar) import Control.Arrow ((>>>)) -import Control.Monad (guard, unless, when) -import Data.Char (chr, isPrint, ord) -import Data.List (unfoldr) +import Control.Monad (when) +import Data.Char (chr, ord) +import qualified Data.Foldable1 as Foldable1 +import qualified Data.List.NonEmpty as NonEmpty import Data.Maybe (listToMaybe, mapMaybe) +import GHC.Data.StringBuffer (StringBuffer) +import qualified GHC.Data.StringBuffer as StringBuffer import GHC.Parser.CharClass ( hexDigit, - is_any, is_decdigit, is_hexdigit, is_octdigit, is_space, octDecDigit, ) +import GHC.Parser.Errors.Types (LexErr (..)) import GHC.Utils.Panic (panic) -data LexStringType = StringTypeSingle | StringTypeMulti +type BufPos = Int +data StringLexError = StringLexError LexErr BufPos --- | State to accumulate while iterating through string literal. --- --- Fields are strict here to avoid space leak when iterating through large string literal --- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 -data LexStringState loc = LexStringState - { stringAcc :: !String - -- ^ The string seen so far, reversed - , multilineCommonWsPrefix :: !Int - -- ^ The common prefix for multiline strings. See Note [Multiline string literals] - , initialLoc :: !loc - -- ^ The location of the beginning of the string literal - } - --- | Get the character at the given location, with the location --- of the next character. Returns Nothing if at the end of the --- input. -type GetChar loc = loc -> Maybe (Char, loc) - -lexString :: LexStringType -> GetChar loc -> loc -> Either (StringLexError loc) (String, loc) -lexString strType getChar initialLoc = go initialState initialLoc +lexString :: Int -> StringBuffer -> Either StringLexError String +lexString = lexStringWith processChars processChars where - initialState = - LexStringState - { stringAcc = "" - , multilineCommonWsPrefix = - case strType of - StringTypeMulti -> maxBound - _ -> 0 - , initialLoc = initialLoc - } - - -- 's' is strict here to avoid space leak when iterating through large string literal - -- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 - go !s loc0 = - case getChar loc0 of - -- found closing delimiter - Just ('"', _) | Just loc1 <- checkDelimiter strType getChar loc0 -> do - let postprocess = - case strType of - StringTypeSingle -> id - StringTypeMulti -> postprocessMultiline (multilineCommonWsPrefix s) - Right (postprocess . reverse $ stringAcc s, loc1) - - -- found backslash - Just (c0@'\\', loc1) -> do - case getChar loc1 of - -- found '\&' character, which should be elided - Just ('&', loc2) -> go s loc2 - -- found start of a string gap - Just (c1, loc2) | is_space c1 -> collapseStringGap getChar s loc2 >>= go s - -- some other escape character - Just (c1, loc2) -> - case strType of - StringTypeSingle -> do - (c', loc') <- resolveEscapeCharacter getChar loc1 - go (addChar c' s) loc' - StringTypeMulti -> do - -- keep escape characters unresolved until after post-processing, - -- to distinguish between a user-newline and the user writing "\n". - -- but still process the characters here, to find any errors - _ <- resolveEscapeCharacter getChar loc1 - go (addChar c1 . addChar c0 $ s) loc2 - -- backslash at end of input - Nothing -> Left $ BadCharInitialLex loc1 (hasSQuote getChar s) - - -- found newline character in multiline string - Just (c0@'\n', loc1) | StringTypeMulti <- strType -> - uncurry go $ parseLeadingWS getChar (addChar c0 s) loc1 - - -- found some other character - Just (c0, loc1) | isAnyChar c0 -> go (addChar c0 s) loc1 - - -- found some unknown character - Just (_, _) -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - - -- reached EOF before finding end of string - Nothing -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) -{-# INLINE lexString #-} - -checkDelimiter :: LexStringType -> GetChar loc -> loc -> Maybe loc -checkDelimiter strType getChar loc0 = - case strType of - StringTypeSingle -> do - ('"', loc1) <- getChar loc0 - Just loc1 - StringTypeMulti -> do - ('"', loc1) <- getChar loc0 - ('"', loc2) <- getChar loc1 - ('"', loc3) <- getChar loc2 - Just loc3 -{-# INLINE checkDelimiter #-} - --- | A helper for adding the given character to the lexed string. -addChar :: Char -> LexStringState loc -> LexStringState loc -addChar c s = s{stringAcc = c : stringAcc s} -{-# INLINE addChar #-} - --- | Return whether the string we've parsed so far contains any smart quotes. -hasSQuote :: GetChar loc -> LexStringState loc -> ContainsSmartQuote loc -hasSQuote getChar s - | any isDoubleSmartQuote (stringAcc s) - , (c, loc) : _ <- filter (isDoubleSmartQuote . fst) allChars = - SmartQuote c loc - | otherwise = - NoSmartQuote + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps + >>> resolveEscapes + +-- ----------------------------------------------------------------------------- +-- Lexing interface + +{- +Note [Lexing strings] +~~~~~~~~~~~~~~~~~~~~~ + +After verifying if a string is lexically valid with Alex, we still need to do +some post processing of the string, namely: +1. Collapse string gaps +2. Resolve escape characters + +The problem: 'lexemeToString' is more performant than manually reading +characters from the StringBuffer. However, that completely erases the position +of each character, which we need in order to report the correct position for +error messages (e.g. when resolving escape characters). + +So what we'll do is do two passes. The first pass is optimistic; just convert +to a plain String and process it. If this results in an error, we do a second +pass, this time where each character is annotated with its position. Now, the +error has all the information it needs. + +Ideally, lexStringWith would take a single (forall c. HasChar c => ...) function, +but to help the specializer, we pass it in twice to concretize it for the two +types we actually use. +-} + +-- | See Note [Lexing strings] +lexStringWith :: + ([Char] -> Either (Char, LexErr) [Char]) + -> ([CharPos] -> Either (CharPos, LexErr) [CharPos]) + -> Int + -> StringBuffer + -> Either StringLexError String +lexStringWith processChars processCharsPos len buf = + case processChars $ bufferChars buf len of + Right s -> Right s + Left _ -> + case processCharsPos $ bufferLocatedChars buf len of + Right _ -> panic "expected lex error on second pass" + Left ((_, pos), e) -> Left $ StringLexError e pos + +class HasChar c where + getChar :: c -> Char + setChar :: Char -> c -> c + +instance HasChar Char where + getChar = id + setChar = const + +instance HasChar (Char, x) where + getChar = fst + setChar c (_, x) = (c, x) + +pattern Char :: HasChar c => Char -> c +pattern Char c <- (getChar -> c) +{-# COMPLETE Char #-} + +bufferChars :: StringBuffer -> Int -> [Char] +bufferChars = StringBuffer.lexemeToString + +type CharPos = (Char, BufPos) + +bufferLocatedChars :: StringBuffer -> Int -> [CharPos] +bufferLocatedChars initialBuf len = go initialBuf where - allChars = unfoldr getCharWithLoc (initialLoc s) - getCharWithLoc loc = - case getChar loc of - Just (c, loc') -> Just ((c, loc), loc') - Nothing -> Nothing -{-# INLINE hasSQuote #-} - --- | After parsing a backslash and a space character, consume the rest of --- the string gap and return the next location. -collapseStringGap :: GetChar loc -> LexStringState loc -> loc -> Either (StringLexError loc) loc -collapseStringGap getChar s = go + go buf + | atEnd buf = [] + | otherwise = + let (c, buf') = StringBuffer.nextChar buf + in (c, StringBuffer.cur buf) : go buf' + + atEnd buf = StringBuffer.byteDiff initialBuf buf >= len + +-- ----------------------------------------------------------------------------- +-- Lexing phases + +collapseGaps :: HasChar c => [c] -> [c] +collapseGaps = go where - go loc0 = - case getChar loc0 of - Just ('\\', loc1) -> pure loc1 - Just (c0, loc1) | is_space c0 -> go loc1 - Just _ -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - Nothing -> Left $ UnexpectedEOF loc0 (hasSQuote getChar s) -{-# INLINE collapseStringGap #-} + go = \case + c1@(Char '\\') : c2@(Char c) : cs + | is_space c -> go $ dropGap cs + | otherwise -> c1 : c2 : go cs + c : cs -> c : go cs + [] -> [] --- | See Note [Multiline string literals] -parseLeadingWS :: GetChar loc -> LexStringState loc -> loc -> (LexStringState loc, loc) -parseLeadingWS getChar = go 0 + dropGap = \case + Char '\\' : cs -> cs + _ : cs -> dropGap cs + [] -> panic "gap unexpectedly ended" + +resolveEscapes :: HasChar c => [c] -> Either (c, LexErr) [c] +resolveEscapes = go dlistEmpty where - go !col s loc = - case getChar loc of - Just (c@' ', loc') -> go (col + 1) (addChar c s) loc' - -- expand tabs - Just ('\t', loc') -> - let fill = 8 - (col `mod` 8) - s' = applyN fill (addChar ' ') s - in go (col + fill) s' loc' - -- if we see a newline or string delimiter, then this line only contained whitespace, so - -- don't include it in the common whitespace prefix - Just ('\n', _) -> (s, loc) - Just ('"', _) | Just _ <- checkDelimiter StringTypeMulti getChar loc -> (s, loc) - -- found some other character, so we're done parsing leading whitespace - _ -> - let s' = s{multilineCommonWsPrefix = min col (multilineCommonWsPrefix s)} - in (s', loc) - - applyN :: Int -> (a -> a) -> a -> a - applyN n f x0 = iterate f x0 !! n -{-# INLINE parseLeadingWS #-} - -data StringLexError loc - = UnexpectedEOF !loc !(ContainsSmartQuote loc) - -- ^ Unexpectedly hit EOF when lexing string - | BadCharInitialLex !loc !(ContainsSmartQuote loc) - -- ^ Found invalid character when initially lexing string - | EscapeBadChar !loc - -- ^ Found invalid character when parsing an escaped character - | EscapeUnexpectedEOF !loc - -- ^ Unexpectedly hit EOF when parsing an escaped character - | EscapeNumRangeError !loc - -- ^ Escaped number exceeds range - | EscapeSmartQuoteError !Char !loc - -- ^ Found escaped smart unicode chars as `\’` or `\”` - deriving (Show) - --- | When initially lexing the string, we want to track if we've --- seen a smart quote, to show a helpful "you might be accidentally --- using a smart quote" error. -data ContainsSmartQuote loc - = NoSmartQuote - | SmartQuote !Char !loc - deriving (Show) + go !acc = \case + [] -> pure $ dlistToList acc + Char '\\' : Char '&' : cs -> go acc cs + backslash@(Char '\\') : cs -> + case resolveEscapeChar cs of + Right (esc, cs') -> go (acc `dlistSnoc` setChar esc backslash) cs' + Left (c, e) -> Left (c, e) + c : cs -> go (acc `dlistSnoc` c) cs -- ----------------------------------------------------------------------------- -- Escape characters --- | After finding a backslash, parse the rest of the escape character, starting --- at the given location. -resolveEscapeCharacter :: GetChar loc -> loc -> Either (StringLexError loc) (Char, loc) -resolveEscapeCharacter getChar loc0 = do - (c0, loc1) <- expectChar loc0 - case c0 of - 'a' -> pure ('\a', loc1) - 'b' -> pure ('\b', loc1) - 'f' -> pure ('\f', loc1) - 'n' -> pure ('\n', loc1) - 'r' -> pure ('\r', loc1) - 't' -> pure ('\t', loc1) - 'v' -> pure ('\v', loc1) - '\\' -> pure ('\\', loc1) - '"' -> pure ('\"', loc1) - '\'' -> pure ('\'', loc1) - -- escape codes - 'x' -> expectNum is_hexdigit 16 hexDigit loc1 - 'o' -> expectNum is_octdigit 8 octDecDigit loc1 - _ | is_decdigit c0 -> expectNum is_decdigit 10 octDecDigit loc0 - -- control characters (e.g. '\^M') - '^' -> do - (c1, loc2) <- expectChar loc1 - unless ('@' <= c1 && c1 <= '_') $ Left $ EscapeBadChar loc1 - pure (chr $ ord c1 - ord '@', loc2) - -- long form escapes (e.g. '\NUL') - _ | Just (c1, loc2) <- parseLongEscape getChar c0 loc1 -> pure (c1, loc2) - -- check unicode smart quotes (#21843) - _ | isDoubleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - _ | isSingleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - -- unknown escape - _ -> Left $ EscapeBadChar loc0 +-- | Resolve a escape character, after having just lexed a backslash. +-- Assumes escape character is valid. +resolveEscapeChar :: HasChar c => [c] -> Either (c, LexErr) (Char, [c]) +resolveEscapeChar = \case + Char 'a' : cs -> pure ('\a', cs) + Char 'b' : cs -> pure ('\b', cs) + Char 'f' : cs -> pure ('\f', cs) + Char 'n' : cs -> pure ('\n', cs) + Char 'r' : cs -> pure ('\r', cs) + Char 't' : cs -> pure ('\t', cs) + Char 'v' : cs -> pure ('\v', cs) + Char '\\' : cs -> pure ('\\', cs) + Char '"' : cs -> pure ('\"', cs) + Char '\'' : cs -> pure ('\'', cs) + -- escape codes + Char 'x' : cs -> parseNum is_hexdigit 16 hexDigit cs + Char 'o' : cs -> parseNum is_octdigit 8 octDecDigit cs + cs@(Char c : _) | is_decdigit c -> parseNum is_decdigit 10 octDecDigit cs + -- control characters (e.g. '\^M') + Char '^' : Char c : cs -> pure (chr $ ord c - ord '@', cs) + -- long form escapes (e.g. '\NUL') + cs | Just (esc, cs') <- parseLongEscape cs -> pure (esc, cs') + -- shouldn't happen + Char c : _ -> panic $ "found unexpected escape character: " ++ show c + [] -> panic "escape character unexpectedly ended" where - expectChar loc = - case getChar loc of - Just x -> pure x - Nothing -> Left $ EscapeUnexpectedEOF loc - - expectNum isDigit base toDigit loc1 = do - (c1, loc2) <- expectChar loc1 - unless (isDigit c1) $ Left $ EscapeBadChar loc1 - let parseNum x loc = - case getChar loc of - Just (c, loc') | isDigit c -> do - let x' = x * base + toDigit c - when (x' > 0x10ffff) $ Left $ EscapeNumRangeError loc - parseNum x' loc' - _ -> - pure (chr x, loc) - parseNum (toDigit c1) loc2 -{-# INLINE resolveEscapeCharacter #-} - -parseLongEscape :: GetChar loc -> Char -> loc -> Maybe (Char, loc) -parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCodes + parseNum isDigit base toDigit = + let go x = \case + ch@(Char c) : cs | isDigit c -> do + let x' = x * base + toDigit c + when (x' > 0x10ffff) $ Left (ch, LexNumEscapeRange) + go x' cs + cs -> pure (chr x, cs) + in go 0 + +parseLongEscape :: HasChar c => [c] -> Maybe (Char, [c]) +parseLongEscape cs = listToMaybe (mapMaybe tryParse longEscapeCodes) where - tryParse (prefix, c) = do - p0 : p <- pure prefix - guard (p0 == c0) -- see if the first character matches - loc <- parsePrefix loc1 p -- see if the rest of the prefix matches - pure (c, loc) - - parsePrefix loc = \case - [] -> pure loc - p : ps -> do - (c, loc') <- getChar loc - guard (p == c) - parsePrefix loc' ps + tryParse (code, esc) = + case splitAt (length code) cs of + (pre, cs') | map getChar pre == code -> Just (esc, cs') + _ -> Nothing longEscapeCodes = [ ("NUL", '\NUL') @@ -289,14 +204,14 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("ENQ", '\ENQ') , ("ACK", '\ACK') , ("BEL", '\BEL') - , ("BS", '\BS') - , ("HT", '\HT') - , ("LF", '\LF') - , ("VT", '\VT') - , ("FF", '\FF') - , ("CR", '\CR') - , ("SO", '\SO') - , ("SI", '\SI') + , ("BS" , '\BS' ) + , ("HT" , '\HT' ) + , ("LF" , '\LF' ) + , ("VT" , '\VT' ) + , ("FF" , '\FF' ) + , ("CR" , '\CR' ) + , ("SO" , '\SO' ) + , ("SI" , '\SI' ) , ("DLE", '\DLE') , ("DC1", '\DC1') , ("DC2", '\DC2') @@ -306,17 +221,16 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("SYN", '\SYN') , ("ETB", '\ETB') , ("CAN", '\CAN') - , ("EM", '\EM') + , ("EM" , '\EM' ) , ("SUB", '\SUB') , ("ESC", '\ESC') - , ("FS", '\FS') - , ("GS", '\GS') - , ("RS", '\RS') - , ("US", '\US') - , ("SP", '\SP') + , ("FS" , '\FS' ) + , ("GS" , '\GS' ) + , ("RS" , '\RS' ) + , ("US" , '\US' ) + , ("SP" , '\SP' ) , ("DEL", '\DEL') ] -{-# INLINE parseLongEscape #-} -- ----------------------------------------------------------------------------- -- Unicode Smart Quote detection (#21843) @@ -333,6 +247,98 @@ isSingleSmartQuote = \case '’' -> True _ -> False +-- ----------------------------------------------------------------------------- +-- Multiline strings + +-- | See Note [Multiline string literals] +-- +-- Assumes string is lexically valid. Skips the steps about splitting +-- and rejoining lines, and instead manually find newline characters, +-- for performance. +lexMultilineString :: Int -> StringBuffer -> Either StringLexError String +lexMultilineString = lexStringWith processChars processChars + where + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps -- Step 1 + >>> expandLeadingTabs -- Step 3 + >>> rmCommonWhitespacePrefix -- Step 4 + >>> collapseOnlyWsLines -- Step 5 + >>> rmFirstNewline -- Step 7a + >>> rmLastNewline -- Step 7b + >>> resolveEscapes -- Step 8 + + -- expands all tabs, since the lexer will verify that tabs can only appear + -- as leading indentation + expandLeadingTabs :: HasChar c => [c] -> [c] + expandLeadingTabs = + let go !col = \case + c@(Char '\t') : cs -> + let fill = 8 - (col `mod` 8) + in replicate fill (setChar ' ' c) ++ go (col + fill) cs + c : cs -> c : go (if getChar c == '\n' then 0 else col + 1) cs + [] -> [] + in go 0 + + rmCommonWhitespacePrefix :: HasChar c => [c] -> [c] + rmCommonWhitespacePrefix cs0 = + let commonWSPrefix = getCommonWsPrefix (map getChar cs0) + go = \case + c@(Char '\n') : cs -> c : go (dropLine commonWSPrefix cs) + c : cs -> c : go cs + [] -> [] + -- drop x characters from the string, or up to a newline, whichever + -- comes first + dropLine !x = \case + cs | x <= 0 -> cs + cs@(Char '\n' : _) -> cs + _ : cs -> dropLine (x - 1) cs + [] -> [] + in go cs0 + + collapseOnlyWsLines :: HasChar c => [c] -> [c] + collapseOnlyWsLines = + let go = \case + c@(Char '\n') : cs | Just cs' <- checkAllWs cs -> c : go cs' + c : cs -> c : go cs + [] -> [] + checkAllWs = \case + -- got all the way to a newline or the end of the string, return + cs@(Char '\n' : _) -> Just cs + cs@[] -> Just cs + -- found whitespace, continue + Char c : cs | is_space c -> checkAllWs cs + -- anything else, stop + _ -> Nothing + in go + + rmFirstNewline :: HasChar c => [c] -> [c] + rmFirstNewline = \case + Char '\n' : cs -> cs + cs -> cs + + rmLastNewline :: HasChar c => [c] -> [c] + rmLastNewline = + let go = \case + [] -> [] + [Char '\n'] -> [] + c : cs -> c : go cs + in go + +-- | See step 4 in Note [Multiline string literals] +-- +-- Assumes tabs have already been expanded. +getCommonWsPrefix :: String -> Int +getCommonWsPrefix s = + case NonEmpty.nonEmpty includedLines of + Nothing -> 0 + Just ls -> Foldable1.minimum $ NonEmpty.map (length . takeWhile is_space) ls + where + includedLines = + filter (not . all is_space) -- ignore whitespace-only lines + . drop 1 -- ignore first line in calculation + $ lines s + {- Note [Multiline string literals] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -351,23 +357,13 @@ The canonical steps for post processing a multiline string are: 2. Split the string by newlines 3. Convert leading tabs into spaces * In each line, any tabs preceding non-whitespace characters are replaced with spaces up to the next tab stop -4. Remove common whitespace prefix in every line (see below) +4. Remove common whitespace prefix in every line except the first (see below) 5. If a line contains only whitespace, remove all of the whitespace 6. Join the string back with `\n` delimiters -7. If the first character of the string is a newline, remove it +7a. If the first character of the string is a newline, remove it +7b. If the last character of the string is a newline, remove it 8. Interpret escaped characters -However, for performance reasons, we do as much of this in one pass as possible: -1. As we lex the string, do the following steps as they appear: - a. Collapse string gaps - b. Keep track of the common whitespace prefix so far - c. Validate escaped characters -2. At the very end, post process the lexed string: - a. Remove the common whitespace prefix from every line - b. Remove all whitespace from all-whitespace lines - c. Remove initial newline character - d. Resolve escaped characters - The common whitespace prefix can be informally defined as "The longest prefix of whitespace shared by all lines in the string, excluding the first line and any whitespace-only lines". @@ -382,75 +378,16 @@ It's more precisely defined with the following algorithm: 3. Calculate the longest prefix of whitespace shared by all lines in the remaining list -} --- | See Note [Multiline string literals] -postprocessMultiline :: Int -> String -> String -postprocessMultiline commonWSPrefix = - rmCommonWhitespacePrefix - >>> collapseOnlyWsLines - >>> rmFirstNewline - >>> rmLastNewline - >>> resolveEscapeChars - where - rmCommonWhitespacePrefix = - let go = \case - '\n' : s -> '\n' : go (dropLine commonWSPrefix s) - c : s -> c : go s - [] -> [] - -- drop x characters from the string, or up to a newline, whichever - -- comes first - dropLine !x = \case - s | x <= 0 -> s - s@('\n' : _) -> s - _ : s -> dropLine (x - 1) s - [] -> [] - in go - - collapseOnlyWsLines = - let go = \case - '\n' : s | Just s' <- checkAllWs s -> '\n' : go s' - c : s -> c : go s - [] -> [] - checkAllWs = \case - -- got all the way to a newline or the end of the string, return - s@('\n' : _) -> Just s - s@[] -> Just s - -- found whitespace, continue - c : s | is_space c -> checkAllWs s - -- anything else, stop - _ -> Nothing - in go - - rmFirstNewline = \case - '\n' : s -> s - s -> s +-- ----------------------------------------------------------------------------- +-- DList - rmLastNewline = - let go = \case - [] -> [] - ['\n'] -> [] - c : cs -> c : go cs - in go +newtype DList a = DList ([a] -> [a]) - -- resolve escape characters, deferred from lexString. guaranteed - -- to not throw any errors, since we already checked them in lexString - resolveEscapeChars = \case - [] -> [] - '\\' : s -> - -- concretizing 'loc' to String: - -- resolveEscapeCharacter :: (String -> Maybe (Char, String)) -> String -> Either _ (Char, String) - case resolveEscapeCharacter uncons s of - Left e -> panic $ "resolving escape characters in multiline string unexpectedly found errors: " ++ show e - Right (c, s') -> c : resolveEscapeChars s' - c : s -> c : resolveEscapeChars s - - uncons = \case - c : cs -> Just (c, cs) - [] -> Nothing +dlistEmpty :: DList a +dlistEmpty = DList id --- ----------------------------------------------------------------------------- --- Helpers +dlistToList :: DList a -> [a] +dlistToList (DList f) = f [] -isAnyChar :: Char -> Bool -isAnyChar c - | c > '\x7f' = isPrint c - | otherwise = is_any c +dlistSnoc :: DList a -> a -> DList a +dlistSnoc (DList f) x = DList (f . (x :)) ===================================== testsuite/driver/testlib.py ===================================== @@ -2868,7 +2868,7 @@ def normalise_errmsg(s: str) -> str: # normalise slashes to minimise Windows/Unix filename differences, # but don't normalize backslashes in chars - s = re.sub(r"(?!')\\", '/', s) + s = re.sub(r"(?:9:1: error: [GHC-58481] parse error on input ‘+’ -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' + +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/ghci/scripts/ghci022.stderr ===================================== @@ -1,3 +1,2 @@ +ghci022.hs:2:5: error: [GHC-21231] lexical error at character '\n' -ghci022.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr ===================================== @@ -1,3 +1,3 @@ - MultilineStringsInnerTab.hs:8:5: error: [GHC-21231] - lexical error in string/character literal at character '\t' + lexical error at character '\t' + ===================================== testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr ===================================== @@ -1,6 +1,2 @@ - MultilineStringsSmartQuotes.hs:9:3: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -MultilineStringsSmartQuotes.hs:10:1: [GHC-21231] - lexical error in string/character literal at end of input ===================================== testsuite/tests/parser/should_fail/T21843c.stderr ===================================== @@ -1,6 +1,2 @@ - -T21843c.hs:3:19: [GHC-31623] +T21843c.hs:3:13: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -T21843c.hs:3:20: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/T21843e.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843e.hs:3:15: [GHC-31623] +T21843e.hs:3:13: error: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not + ===================================== testsuite/tests/parser/should_fail/T21843f.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843f.hs:3:13: [GHC-31623] +T21843f.hs:3:11: error: [GHC-31623] Unicode character '‘' ('\8216') looks like ''' (Single Quote), but it is not + ===================================== testsuite/tests/parser/should_fail/T3751.stderr ===================================== @@ -1,3 +1,2 @@ +T3751.hs:3:5: error: [GHC-21231] lexical error at character '\167' -T3751.hs:3:7: error: [GHC-21231] - lexical error in string/character literal at character '\167' ===================================== testsuite/tests/parser/should_fail/T5425.stderr ===================================== @@ -1,3 +1,2 @@ +T5425.hs:4:1: error: [GHC-21231] lexical error at character '\955' -T5425.hs:5:2: error: [GHC-21231] - lexical error in string/character literal at character '\955' ===================================== testsuite/tests/parser/should_fail/readFail002.stderr ===================================== @@ -1,3 +1,3 @@ +readFail002.hs:5:5: error: [GHC-21231] + lexical error at character '\n' -readFail002.hs:5:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/readFail004.stderr ===================================== @@ -1,3 +1,3 @@ +readFail004.hs:17:16: error: [GHC-21231] + lexical error at character '.' -readFail004.hs:19:1: error: [GHC-21231] - lexical error in string/character literal at character '.' ===================================== testsuite/tests/parser/should_fail/readFail005.stderr ===================================== @@ -1,3 +1,2 @@ - -readFail005.hs:4:7: error: [GHC-21231] - lexical error in string/character literal at character '&' +readFail005.hs:4:5: [GHC-21231] + lexical error at character '&' ===================================== testsuite/tests/parser/should_fail/readFail033.stderr ===================================== @@ -1,3 +1,3 @@ +readFail033.hs:2:5: error: [GHC-21231] + lexical error at character '\t' -readFail033.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\t' ===================================== testsuite/tests/parser/unicode/all.T ===================================== @@ -33,3 +33,6 @@ test('T18225B', normal, compile_fail, ['']) test('T18158', normal, compile, ['']) test('T18158b', normal, compile_fail, ['']) + +test('lex_unispace', normal, compile, ['']) +test('lex_unicode_ids', normal, compile, ['']) ===================================== testsuite/tests/parser/unicode/lex_unicode_ids.hs ===================================== @@ -0,0 +1,11 @@ +-- Regression tests for unicode identifiers + +{-# LANGUAGE TemplateHaskellQuotes #-} + +module ShouldCompile where + +ƞ = 1 +eta = 'ƞ + +data Ʊ +upsilon = ''Ʊ ===================================== testsuite/tests/parser/unicode/lex_unispace.hs ===================================== @@ -0,0 +1,9 @@ +-- Regression tests for unicode whitespace + +module ShouldCompile where + +-- https://github.com/fosskers/aura/blob/cdebca1f48254ebb8286d8e38591bf644282866f/haskell/aura/lib/Aura/Languages.hs#L107 +x1 = '　' -- \12288 + +-- https://github.com/jgm/pandoc/blob/98e77e02f6436e4b74a164762d0f3149ae7ecefa/src/Text/Pandoc/Writers/FB2.hs#L295C3-L295C32 +x2 = " " -- \xa0 ===================================== testsuite/tests/parser/unicode/utf8_010.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_010.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_011.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_011.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_020.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_020.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_021.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_021.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_022.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_022.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs ===================================== @@ -320,7 +320,7 @@ classify tok = ITlabelvarid{} -> TkUnknown ITchar{} -> TkChar ITstring{} -> TkString - ITmultilinestring{} -> TkString + ITstringMulti{} -> TkString ITinteger{} -> TkNumber ITrational{} -> TkNumber ITprimchar{} -> TkChar View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/92f9a4dfbd862eb0b80e55d4ac388af356c32069 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/92f9a4dfbd862eb0b80e55d4ac388af356c32069 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 16:03:29 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 25 Sep 2024 12:03:29 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 4656 commits: [haddock @ 2002-04-04 16:23:43 by simonmar] Message-ID: <66f43451b7ce6_123102adf9a854620@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: 2b39cd94 by Simon Marlow at 2002-04-04T16:23:43+00:00 [haddock @ 2002-04-04 16:23:43 by simonmar] This is Haddock, my stab at a Haskell documentation tool. It's not quite ready for release yet, but I'm putting it in the repository so others can take a look. It uses a locally modified version of the hssource parser, extended with support for GHC extensions and documentation annotations. - - - - - 99ede94f by Simon Marlow at 2002-04-04T16:24:10+00:00 [haddock @ 2002-04-04 16:24:10 by simonmar] forgot one file - - - - - 8363294c by Simon Marlow at 2002-04-05T13:58:15+00:00 [haddock @ 2002-04-05 13:58:15 by simonmar] Remap names in the exported declarations to be "closer" to the current module. eg. if an exported declaration mentions a type 'T' which is imported from module A then re-exported from the current module, then links from the type or indeed the documentation will point to the current module rather than module A. This is to support better hiding: module A won't be referred to in the generated output. - - - - - 1570cbc1 by Simon Marlow at 2002-04-05T13:58:23+00:00 [haddock @ 2002-04-05 13:58:23 by simonmar] update the TODO list - - - - - 3a62f96b by Simon Marlow at 2002-04-05T14:11:51+00:00 [haddock @ 2002-04-05 14:11:51 by simonmar] Fix the anchor for a class declaration - - - - - c5d9a471 by Simon Marlow at 2002-04-05T14:18:41+00:00 [haddock @ 2002-04-05 14:18:41 by simonmar] remove underlines on visited links - - - - - 97280525 by Simon Marlow at 2002-04-05T16:11:47+00:00 [haddock @ 2002-04-05 16:11:47 by simonmar] - Update to generate more correct HTML. - Use our own non-overloaded table combinators, as the overloaded versions were giving me a headache. The improved type safety caught several errors in the HTML generation. - - - - - 9acd3a4d by Simon Marlow at 2002-04-05T16:32:19+00:00 [haddock @ 2002-04-05 16:32:19 by simonmar] Add width property to the title, and add TD.children for the module contents page. - - - - - ec9a0847 by Simon Marlow at 2002-04-08T16:39:56+00:00 [haddock @ 2002-04-08 16:39:56 by simonmar] Fix a problem with exports of the form T(..). - - - - - e4627dc8 by Simon Marlow at 2002-04-08T16:41:38+00:00 [haddock @ 2002-04-08 16:41:37 by simonmar] - Add our own versions of Html & BlockTable for the time being. - Add support for generating an index to the HTML backend - - - - - 2d73fd75 by Simon Marlow at 2002-04-09T11:23:24+00:00 [haddock @ 2002-04-09 11:23:24 by simonmar] Add '-- /' as a synonym for '-- |', for compatibility with IDoc. - - - - - 3675464e by Simon Marlow at 2002-04-09T11:33:55+00:00 [haddock @ 2002-04-09 11:33:54 by simonmar] - add the <...> syntax for marking up URLs in documentation - Make the output for data & class declarations more compact when there aren't any documentation annotations on the individual methods or constructors respectively. - - - - - 5077f5b1 by Simon Marlow at 2002-04-09T11:36:04+00:00 [haddock @ 2002-04-09 11:36:04 by simonmar] Update the TODO list - - - - - 9e83c54d by Simon Marlow at 2002-04-10T10:50:06+00:00 [haddock @ 2002-04-10 10:50:06 by simonmar] Use explicit 'px' suffix on pixel sizes; IE seems to prefer them - - - - - 052de51c by Simon Marlow at 2002-04-10T13:23:13+00:00 [haddock @ 2002-04-10 13:23:13 by simonmar] Lex URLs as a single token to avoid having to escape special characters inside the URL string. - - - - - 47187edb by Simon Marlow at 2002-04-10T13:23:55+00:00 [haddock @ 2002-04-10 13:23:55 by simonmar] Not sure why I made the constructor name for a record declaration into a TyCls name, but change it back into a Var name anyhow. - - - - - 3dc6aa81 by Simon Marlow at 2002-04-10T13:26:10+00:00 [haddock @ 2002-04-10 13:26:09 by simonmar] Lots of changes, including: - add index support to the HTML backend - clean up the renamer, put it into a monad - propogate unresolved names to the top level and report them in a nicer way - various bugfixes - - - - - c2a70a72 by Simon Marlow at 2002-04-10T13:32:39+00:00 [haddock @ 2002-04-10 13:32:39 by simonmar] Skeleton documentation - - - - - 50c98d17 by Simon Marlow at 2002-04-10T13:37:23+00:00 [haddock @ 2002-04-10 13:37:23 by simonmar] Update the TODO list, separate into pre-1.0 and post-1.0 items - - - - - f3778be6 by Simon Marlow at 2002-04-10T14:30:58+00:00 [haddock @ 2002-04-10 14:30:58 by simonmar] Add an introduction - - - - - cfbaf9f7 by Simon Marlow at 2002-04-10T14:59:51+00:00 [haddock @ 2002-04-10 14:59:51 by simonmar] Sort the module tree - - - - - 76bd7b34 by Simon Marlow at 2002-04-10T15:50:11+00:00 [haddock @ 2002-04-10 15:50:10 by simonmar] Generate a little table of contents at the top of the module doc (only if the module actually contains some section headings, though). - - - - - bb8560a1 by Simon Marlow at 2002-04-10T16:10:26+00:00 [haddock @ 2002-04-10 16:10:26 by simonmar] Now we understand (or at least don't barf on) type signatures in patterns such as you might find when scoped type variables are in use. - - - - - 86c2a026 by Simon Marlow at 2002-04-10T16:10:49+00:00 [haddock @ 2002-04-10 16:10:49 by simonmar] more updates - - - - - 1c052b0e by Simon Marlow at 2002-04-10T16:28:05+00:00 [haddock @ 2002-04-10 16:28:05 by simonmar] Parse errors in doc strings are now reported as warnings rather that causing the whole thing to fall over. It still needs cleaning up (the warning is emitted with trace) but this will do for the time being. - - - - - ace03e8f by Simon Marlow at 2002-04-10T16:38:03+00:00 [haddock @ 2002-04-10 16:38:03 by simonmar] update again - - - - - 69006c3e by Simon Marlow at 2002-04-11T13:38:02+00:00 [haddock @ 2002-04-11 13:38:02 by simonmar] mention Opera - - - - - fe9b10f8 by Simon Marlow at 2002-04-11T13:40:31+00:00 [haddock @ 2002-04-11 13:40:30 by simonmar] - copy haddock.css into the same place as the generated HTML - new option: --css <file> specifies the style sheet to use - new option: -o <dir> specifies the directory in which to generate the output. - because Haddock now needs to know where to find its default stylesheet, we have to have a wrapper script and do the haddock-inplace thing (Makefile code copied largely from fptools/happy). - - - - - 106adbbe by Simon Marlow at 2002-04-24T15:12:41+00:00 [haddock @ 2002-04-24 15:12:41 by simonmar] Stop slurping comment lines when we see a row of dashes longer than length 2: these are useful as separators. - - - - - 995d3f9e by Simon Marlow at 2002-04-24T15:14:12+00:00 [haddock @ 2002-04-24 15:14:11 by simonmar] Grok the kind of module headers we use in fptools/libraries, and pass the "portability", "stability", and "maintainer" strings through into the generated HTML. If the module header doesn't match the pattern, then we don't include the info in the HTML. - - - - - e14da136 by Simon Marlow at 2002-04-24T15:16:57+00:00 [haddock @ 2002-04-24 15:16:57 by simonmar] Done module headers now. - - - - - 2ca8dfd4 by Simon Marlow at 2002-04-24T15:57:48+00:00 [haddock @ 2002-04-24 15:57:47 by simonmar] Handle gcons in export lists (a common extension). - - - - - 044cea81 by Simon Marlow at 2002-04-25T14:20:12+00:00 [haddock @ 2002-04-25 14:20:12 by simonmar] Add the little lambda icon - - - - - 63955027 by Simon Marlow at 2002-04-25T14:40:05+00:00 [haddock @ 2002-04-25 14:40:05 by simonmar] - Add support for named chunks of documentation which can be referenced from the export list. - Copy the icon from $libdir to the destination in HTML mode. - - - - - 36e3f913 by Simon Marlow at 2002-04-25T16:48:36+00:00 [haddock @ 2002-04-25 16:48:36 by simonmar] More keyboard bashing - - - - - 7ae18dd0 by Simon Marlow at 2002-04-26T08:43:33+00:00 [haddock @ 2002-04-26 08:43:33 by simonmar] Package util reqd. to compile with 4.08.2 - - - - - bbd5fbab by Simon Marlow at 2002-04-26T10:13:00+00:00 [haddock @ 2002-04-26 10:13:00 by simonmar] Include $(GHC_HAPPY_OPTS) when compiling HsParser - - - - - 31c53d79 by Simon Marlow at 2002-04-26T11:18:57+00:00 [haddock @ 2002-04-26 11:18:56 by simonmar] - support for fundeps (partially contributed by Brett Letner - thanks Brett). - make it build with GHC 4.08.2 - - - - - c415ce76 by Simon Marlow at 2002-04-26T13:15:02+00:00 [haddock @ 2002-04-26 13:15:02 by simonmar] Move the explicit formatting of the little table for the stability/portability/maintainer info from the HTML into the CSS, and remove the explicit table size (just right-align it). - - - - - 520ee21a by Simon Marlow at 2002-04-26T16:01:44+00:00 [haddock @ 2002-04-26 16:01:44 by simonmar] Yet more keyboard bashing - this is pretty much complete now. - - - - - 2ae37179 by Simon Marlow at 2002-04-26T16:02:14+00:00 [haddock @ 2002-04-26 16:02:14 by simonmar] Add a couple of things I forgot about - - - - - b7211e04 by Simon Marlow at 2002-04-29T15:28:12+00:00 [haddock @ 2002-04-29 15:28:12 by simonmar] bugfix for declBinders on a NewTypeDecl - - - - - 640c154a by Simon Marlow at 2002-04-29T15:28:54+00:00 [haddock @ 2002-04-29 15:28:54 by simonmar] Allow '-- |' style annotations on constructors and record fields. - - - - - 393f258a by Simon Marlow at 2002-04-29T15:37:32+00:00 [haddock @ 2002-04-29 15:37:32 by simonmar] syntax fix - - - - - 8a2c2549 by Simon Marlow at 2002-04-29T15:37:48+00:00 [haddock @ 2002-04-29 15:37:48 by simonmar] Add an example - - - - - db88f8a2 by Simon Marlow at 2002-04-29T15:55:46+00:00 [haddock @ 2002-04-29 15:55:46 by simonmar] remove a trace - - - - - 2b0248e0 by Simon Marlow at 2002-04-29T15:56:19+00:00 [haddock @ 2002-04-29 15:56:19 by simonmar] Fix for 'make install' - - - - - 120453a0 by Simon Marlow at 2002-04-29T15:56:39+00:00 [haddock @ 2002-04-29 15:56:39 by simonmar] Install the auxilliary bits - - - - - 950e6dbb by Simon Marlow at 2002-04-29T15:57:30+00:00 [haddock @ 2002-04-29 15:57:30 by simonmar] Add BinDist bits - - - - - 154b9d71 by Simon Marlow at 2002-05-01T11:02:52+00:00 [haddock @ 2002-05-01 11:02:52 by simonmar] update - - - - - ba6c39fa by Simon Marlow at 2002-05-01T11:03:26+00:00 [haddock @ 2002-05-01 11:03:26 by simonmar] Add another item - - - - - bacb5e33 by Simon Marlow at 2002-05-03T08:50:00+00:00 [haddock @ 2002-05-03 08:50:00 by simonmar] Fix some typos. - - - - - 54c87895 by Sven Panne at 2002-05-05T19:40:51+00:00 [haddock @ 2002-05-05 19:40:51 by panne] As a temporary hack/workaround for a bug in GHC's simplifier, don't pass Happy the -c option for generating the parsers in this subdir. Furthermore, disable -O for HaddocParse, too. - - - - - e6c08703 by Simon Marlow at 2002-05-06T09:51:10+00:00 [haddock @ 2002-05-06 09:51:10 by simonmar] Add RPM spec file (thanks to Tom Moertel <tom-rpms at moertel.com>) - - - - - 7b8fa8e7 by Simon Marlow at 2002-05-06T12:29:26+00:00 [haddock @ 2002-05-06 12:29:26 by simonmar] Add missing type signature (a different workaround for the bug in GHC's simplifier). - - - - - cd0e300d by Simon Marlow at 2002-05-06T12:30:09+00:00 [haddock @ 2002-05-06 12:30:09 by simonmar] Remove workaround for simplifier bug in previous revision. - - - - - 687e68fa by Simon Marlow at 2002-05-06T12:32:32+00:00 [haddock @ 2002-05-06 12:32:32 by simonmar] Allow empty data declarations (another GHC extension). - - - - - 8f29f696 by Simon Marlow at 2002-05-06T12:49:21+00:00 [haddock @ 2002-05-06 12:49:21 by simonmar] Fix silly bug in named documentation block lookup. - - - - - 8e0059af by Simon Marlow at 2002-05-06T13:02:42+00:00 [haddock @ 2002-05-06 13:02:42 by simonmar] Add another named chunk with a different name - - - - - 68f8a896 by Simon Marlow at 2002-05-06T13:32:32+00:00 [haddock @ 2002-05-06 13:32:32 by simonmar] Be more lenient about extra paragraph breaks - - - - - 65fc31db by Simon Marlow at 2002-05-07T15:36:36+00:00 [haddock @ 2002-05-07 15:36:36 by simonmar] DocEmpty is a right and left-unit of DocAppend (remove it in the smart constructor). - - - - - adc81078 by Simon Marlow at 2002-05-07T15:37:15+00:00 [haddock @ 2002-05-07 15:37:15 by simonmar] Allow code blocks to be denoted with bird-tracks in addition to [...]. - - - - - 1283a3c1 by Simon Marlow at 2002-05-08T11:21:56+00:00 [haddock @ 2002-05-08 11:21:56 by simonmar] Add a facility for specifying options that affect Haddock's treatment of the module. Options are given at the top of the module in a comma-separated list, beginning with '-- #'. eg. -- # prune, hide, ignore-exports Options currently available, with their meanings: prune: ignore declarations which have no documentation annotations ignore-exports: act as if the export list were not specified (i.e. export everything local to the module). hide: do not include this module in the generated documentation, but propagate any exported definitions to modules which re-export them. There's a slight change in the semantics for re-exporting a full module by giving 'module M' in the export list: if module M does not have the 'hide' option, then the documentation will now just contain a reference to module M rather than the full inlined contents of that module. These features, and some other changes in the pipeline, are the result of discussions between myself and Manuel Chakravarty <chak at cse.unsw.edu.au> (author of IDoc) yesterday. Also: some cleanups, use a Writer monad to collect error messages in some places instead of just printing them with trace. - - - - - a2239cf5 by Simon Marlow at 2002-05-08T11:22:30+00:00 [haddock @ 2002-05-08 11:22:30 by simonmar] Update to test new features. - - - - - 6add955f by Simon Marlow at 2002-05-08T13:37:25+00:00 [haddock @ 2002-05-08 13:37:25 by simonmar] Change the markup for typewriter-font from [...] to @... at . The reasoning is that the '@' symbol is much less likely to be needed than square brackets, and we don't want to have to escape square brackets in code fragments. This will be mildly painful in the short term, but it's better to get the change out of the way as early as possible. - - - - - cda06447 by Simon Marlow at 2002-05-08T13:39:56+00:00 [haddock @ 2002-05-08 13:39:56 by simonmar] Allow nested-style comments to be used as documentation annotations too. eg. {-| ... -} is equivalent to -- | ... An extra space can also be left after the comment opener: {- | ... -}. The only version that isn't allowed is {-# ... -}, because this syntax overlaps with Haskell pragmas; use {- # ... -} instead. - - - - - db23f65e by Simon Marlow at 2002-05-08T14:48:41+00:00 [haddock @ 2002-05-08 14:48:39 by simonmar] Add support for existential quantifiers on constructors. - - - - - adce3794 by Simon Marlow at 2002-05-08T15:43:25+00:00 [haddock @ 2002-05-08 15:43:25 by simonmar] update - - - - - 62a1f436 by Simon Marlow at 2002-05-08T15:44:10+00:00 [haddock @ 2002-05-08 15:44:10 by simonmar] Update to version 0.2 - - - - - f6a24ba3 by Simon Marlow at 2002-05-09T08:48:29+00:00 [haddock @ 2002-05-09 08:48:29 by simonmar] typo - - - - - 9f9522a4 by Simon Marlow at 2002-05-09T10:33:14+00:00 [haddock @ 2002-05-09 10:33:14 by simonmar] oops, left out '/' from the special characters in the last change. - - - - - 14abcb39 by Simon Marlow at 2002-05-09T10:34:44+00:00 [haddock @ 2002-05-09 10:34:44 by simonmar] Fix buglet - - - - - b8d878be by Simon Marlow at 2002-05-09T10:35:00+00:00 [haddock @ 2002-05-09 10:35:00 by simonmar] Give a more useful instance of Show for Module. - - - - - f7bfd626 by Simon Marlow at 2002-05-09T10:37:07+00:00 [haddock @ 2002-05-09 10:37:07 by simonmar] The last commit to Main.lhs broke the delicate balance of laziness which was being used to avoid computing the dependency graph of modules. So I finally bit the bullet and did a proper topological sort of the module graph, which turned out to be easy (stealing the Digraph module from GHC - this really ought to be in the libraries somewhere). - - - - - b481c1d0 by Simon Marlow at 2002-05-09T10:37:25+00:00 [haddock @ 2002-05-09 10:37:25 by simonmar] another item done - - - - - 032e2b42 by Simon Marlow at 2002-05-09T10:44:15+00:00 [haddock @ 2002-05-09 10:44:15 by simonmar] Don't consider a module re-export as having documentation, for the purposes of deciding whether we need a Synopsis section or not. - - - - - 5fb45e92 by Simon Marlow at 2002-05-09T11:10:55+00:00 [haddock @ 2002-05-09 11:10:55 by simonmar] Add a special case for list types in ppHsAType - - - - - 1937e428 by Simon Marlow at 2002-05-09T12:43:06+00:00 [haddock @ 2002-05-09 12:43:06 by simonmar] Type synonyms can accept a ctype on the RHS, to match GHC. - - - - - 0f16ce56 by Simon Marlow at 2002-05-09T12:45:19+00:00 [haddock @ 2002-05-09 12:45:19 by simonmar] Add 'stdcall' keyword - - - - - 29b0d7d2 by Simon Marlow at 2002-05-09T13:35:45+00:00 [haddock @ 2002-05-09 13:35:45 by simonmar] Add System Requirements section - - - - - bf14dddd by Simon Marlow at 2002-05-09T13:36:11+00:00 [haddock @ 2002-05-09 13:36:11 by simonmar] Test existential types, amongst other things - - - - - 502f8f6f by Simon Marlow at 2002-05-09T13:37:35+00:00 [haddock @ 2002-05-09 13:37:35 by simonmar] Print the module name in a doc-string parse error - - - - - ca1f8d49 by Simon Marlow at 2002-05-09T13:38:04+00:00 [haddock @ 2002-05-09 13:38:04 by simonmar] Add dependency - - - - - 8d3d91ff by Simon Marlow at 2002-05-09T15:37:57+00:00 [haddock @ 2002-05-09 15:37:57 by simonmar] Add the changelog/release notes - - - - - f3960959 by Simon Marlow at 2002-05-09T15:47:47+00:00 [haddock @ 2002-05-09 15:47:47 by simonmar] mention the backquote-style of markup - - - - - 089fb6e6 by Simon Marlow at 2002-05-09T15:59:45+00:00 [haddock @ 2002-05-09 15:59:45 by simonmar] update - - - - - bdd3be0b by Simon Marlow at 2002-05-09T15:59:56+00:00 [haddock @ 2002-05-09 15:59:56 by simonmar] Document changes since 0.1 - - - - - 00fc4af8 by Simon Marlow at 2002-05-10T08:22:48+00:00 [haddock @ 2002-05-10 08:22:48 by simonmar] oops, update to version 0.2 - - - - - a8a79041 by Simon Marlow at 2002-05-10T16:05:08+00:00 [haddock @ 2002-05-10 16:05:08 by simonmar] Only include a mini-contents if there are 2 or more sections - - - - - 06653319 by Simon Marlow at 2002-05-13T09:13:12+00:00 [haddock @ 2002-05-13 09:13:12 by simonmar] fix typos - - - - - 1402b19b by Simon Marlow at 2002-05-13T10:14:22+00:00 [haddock @ 2002-05-13 10:14:22 by simonmar] Allow backquote as the right-hand quote as well as the left-hand quote, as suggested by Dean Herrington. Clean up the grammar a litte. - - - - - dcd5320d by Simon Marlow at 2002-05-13T10:44:10+00:00 [haddock @ 2002-05-13 10:44:10 by simonmar] a couple more things, prioritise a bit - - - - - a90130c4 by Simon Marlow at 2002-05-13T15:19:03+00:00 [haddock @ 2002-05-13 15:19:03 by simonmar] Cope with datatypes which have documentation on the constructor but not the type itself, and records which have documentation on the fields but not the constructor. (Thanks to Ross Paterson for pointing out the bugs). - - - - - a774d432 by Simon Marlow at 2002-05-13T15:20:54+00:00 [haddock @ 2002-05-13 15:20:54 by simonmar] Fix one of the record examples - - - - - 2d1d5218 by Simon Marlow at 2002-05-15T12:44:35+00:00 [haddock @ 2002-05-15 12:44:35 by simonmar] Preserve the newline before a bird-track, but only within a paragraph. - - - - - 1554c09a by Simon Marlow at 2002-05-15T13:03:02+00:00 [haddock @ 2002-05-15 13:03:01 by simonmar] Reworking of the internals to support documenting function arguments (the Most Wanted new feature by the punters). The old method of keeping parsed documentation in a Name -> Doc mapping wasn't going to cut it for anntations on type components, where there's no name to attach the documentation to, so I've moved to storing all the documentation in the abstract syntax. Previously some of the documentation was left in the abstract syntax by the parser, but was later extracted into the mapping. In order to avoid having to parameterise the abstract syntax over the type of documentation stored in it, we have to parse the documentation at the same time as we parse the Haskell source (well, I suppose we could store 'Either String Doc' in the HsSyn, but that's clunky). One upshot is that documentation is now parsed eagerly, and documentation parse errors are fatal (but have better line numbers in the error message). The new story simplifies matters for the code that processes the source modules, because we don't have to maintain the extra Name->Doc mapping, and it should improve efficiency a little too. New features: - Function arguments and return values can now have doc annotations. - If you refer to a qualified name in a doc string, eg. 'IO.putStr', then Haddock will emit a hyperlink even if the identifier is not in scope, so you don't have to make sure everything referred to from the documentation is imported. - several bugs & minor infelicities fixed. - - - - - 57344dc3 by Simon Marlow at 2002-05-15T13:03:19+00:00 [haddock @ 2002-05-15 13:03:19 by simonmar] Bump to version 0.3 - - - - - b2791812 by Simon Marlow at 2002-05-15T13:03:41+00:00 [haddock @ 2002-05-15 13:03:41 by simonmar] update - - - - - fead183e by Simon Marlow at 2002-05-15T13:10:15+00:00 [haddock @ 2002-05-15 13:10:15 by simonmar] Rename Foo.hs to Test.hs, and add a Makefile - - - - - b0b1f89f by Simon Marlow at 2002-05-15T13:16:07+00:00 [haddock @ 2002-05-15 13:16:07 by simonmar] - Remove the note about function argument docs not being implemented - Note that qualified identifiers can be used to point to entities that aren't in scope. - - - - - 5665f31a by Simon Marlow at 2002-05-15T13:28:46+00:00 [haddock @ 2002-05-15 13:28:46 by simonmar] Patch to add support for GHC-style primitive strings ".."#, from Ross Paterson. - - - - - 0564505d by Simon Marlow at 2002-05-17T10:51:57+00:00 [haddock @ 2002-05-17 10:51:57 by simonmar] Fix bugs in qualified name handling (A.B.f was returned as B.f) - - - - - 10e7311c by Simon Marlow at 2002-05-21T10:24:52+00:00 [haddock @ 2002-05-21 10:24:52 by simonmar] - Use an alternate tabular layout for datatypes, which is more compact - Fix some problems with the function argument documentation - - - - - 2f91c2a6 by Simon Marlow at 2002-05-21T10:27:40+00:00 [haddock @ 2002-05-21 10:27:40 by simonmar] add a few more test cases - - - - - 01c2ddd2 by Simon Marlow at 2002-05-21T10:28:33+00:00 [haddock @ 2002-05-21 10:28:33 by simonmar] Rearrange a bit, and add support for tabular datatype rendering - - - - - a4e4c5f8 by Simon Marlow at 2002-05-27T09:03:52+00:00 [haddock @ 2002-05-27 09:03:51 by simonmar] Lots of changes: - instances of a class are listed with the class, and instances involving a datatype are listed with that type. Derived instances aren't included at the moment: the calculation to find the instance head for a derived instance is non-trivial. - some formatting changes; use rows with specified height rather than cellspacing in some places. - various fixes (source file links were wrong, amongst others) - - - - - 48722e68 by Simon Marlow at 2002-05-27T12:30:38+00:00 [haddock @ 2002-05-27 12:30:37 by simonmar] - Put function arguments *before* the doc for the function, as suggested by Sven Panne. This looks nicer when the function documentation is long. - Switch to using bold for binders at the definition site, and use underline for keywords. This makes the binder stand out more. - - - - - 657204d2 by Simon Marlow at 2002-05-27T13:19:49+00:00 [haddock @ 2002-05-27 13:19:49 by simonmar] Fix bug: we weren't renaming HsDocCommentNamed in renameDecl - - - - - 592aae66 by Simon Marlow at 2002-05-27T14:10:27+00:00 [haddock @ 2002-05-27 14:10:27 by simonmar] Fix some bugs in the rendering of qualified type signatures. - - - - - 69c8f763 by Simon Marlow at 2002-05-27T14:36:45+00:00 [haddock @ 2002-05-27 14:36:45 by simonmar] warning message tweak - - - - - 16e64e21 by Simon Marlow at 2002-05-27T14:53:53+00:00 [haddock @ 2002-05-27 14:53:53 by simonmar] hyperlinked identifiers should be in <tt> - - - - - 8d5e4783 by Simon Marlow at 2002-05-27T15:56:45+00:00 [haddock @ 2002-05-27 15:56:45 by simonmar] Do something sensible for modules which don't export anything (except instances). - - - - - 9d3ef811 by Simon Marlow at 2002-05-28T10:12:50+00:00 [haddock @ 2002-05-28 10:12:50 by simonmar] Rename the module documentation properly (bug reported by Sven Panne). - - - - - ef03a1cc by Simon Marlow at 2002-05-28T10:13:04+00:00 [haddock @ 2002-05-28 10:13:04 by simonmar] Add some more test cases - - - - - 92baa0e8 by Simon Marlow at 2002-05-28T11:17:55+00:00 [haddock @ 2002-05-28 11:17:55 by simonmar] If an identifier doesn't lex, then just replace it by a DocString. - - - - - a3156213 by Simon Marlow at 2002-05-28T16:16:19+00:00 [haddock @ 2002-05-28 16:16:19 by simonmar] Only link to names in the current module which are actually listed in the documentation. A name may be exported but not present in the documentation if it is exported as part of a 'module M' export specifier. - - - - - 31acf941 by Simon Marlow at 2002-05-28T16:17:11+00:00 [haddock @ 2002-05-28 16:17:11 by simonmar] update - - - - - 7e474ebf by Sigbjorn Finne at 2002-05-28T22:42:08+00:00 [haddock @ 2002-05-28 22:42:08 by sof] Handle lone occurrences of '/', e.g., -- | This/that. [did this in the lexer rather than in the parser, as I couldn't see a way not to introduce an S/R conflict that way.] - - - - - 093f7e53 by Simon Marlow at 2002-05-29T09:09:49+00:00 [haddock @ 2002-05-29 09:09:49 by simonmar] Back out previous change until we can find a better way to do this. - - - - - 9234389c by Simon Marlow at 2002-05-29T13:19:06+00:00 [haddock @ 2002-05-29 13:19:06 by simonmar] Make the markup syntax a little more friendly: - single quotes are now interpreted literally unless they surround a valid Haskell identifier. So for example now there's no need to escape a single quote used as an apostrophe. - text to the right of a bird track is now literal (if you want marked-up text in a code block, use @...@). - - - - - b3333526 by Simon Marlow at 2002-05-29T13:38:51+00:00 [haddock @ 2002-05-29 13:38:51 by simonmar] Document recent changes to markup syntax - - - - - f93641d6 by Simon Marlow at 2002-05-29T15:27:18+00:00 [haddock @ 2002-05-29 15:27:18 by simonmar] Include the instances in abstract data types too - - - - - 613f21e3 by Simon Marlow at 2002-06-03T13:05:58+00:00 [haddock @ 2002-06-03 13:05:57 by simonmar] Allow exporting of individual class methods and record selectors. For these we have to invent the correct type signature, which we do in the simplest possible way (i.e. no context reduction nonsense in the class case). - - - - - 14b36807 by Simon Marlow at 2002-06-03T13:20:00+00:00 [haddock @ 2002-06-03 13:20:00 by simonmar] Fix linking to qualified names again (thanks to Sven Panne for pointing out the bug). - - - - - 95b10eac by Simon Marlow at 2002-06-03T13:46:48+00:00 [haddock @ 2002-06-03 13:46:48 by simonmar] Fix for exporting record selectors from a newtype declaration - - - - - 272f932e by Simon Marlow at 2002-06-03T13:56:38+00:00 [haddock @ 2002-06-03 13:56:38 by simonmar] update to version 0.3 - - - - - 1c0a3bed by Simon Marlow at 2002-06-03T14:05:07+00:00 [haddock @ 2002-06-03 14:05:07 by simonmar] Add changes in version 0.3 - - - - - 145b4626 by Simon Marlow at 2002-06-03T14:12:38+00:00 [haddock @ 2002-06-03 14:12:38 by simonmar] Render class names as proper binders - - - - - 052106b3 by Simon Marlow at 2002-06-03T14:15:10+00:00 [haddock @ 2002-06-03 14:15:10 by simonmar] update, and separate into bugs, features, and cosmetic items. - - - - - 854f4914 by Simon Marlow at 2002-06-03T14:16:13+00:00 [haddock @ 2002-06-03 14:16:13 by simonmar] More test cases - - - - - 466922c8 by Simon Marlow at 2002-06-03T14:16:56+00:00 [haddock @ 2002-06-03 14:16:56 by simonmar] Example from the paper - - - - - 9962a045 by Simon Marlow at 2002-06-03T14:17:49+00:00 [haddock @ 2002-06-03 14:17:49 by simonmar] A debugging version of the style-sheet, which gives some tables coloured backgrounds so we can see what's going on. - - - - - f16b79db by Simon Marlow at 2002-06-03T14:19:46+00:00 [haddock @ 2002-06-03 14:19:46 by simonmar] typo - - - - - 620db27b by Simon Marlow at 2002-06-03T14:48:32+00:00 [haddock @ 2002-06-03 14:48:32 by simonmar] oops, fix markup bugs - - - - - 53fd105c by Simon Marlow at 2002-06-05T09:05:07+00:00 [haddock @ 2002-06-05 09:05:07 by simonmar] Keep foreign imports when there is no export list (bug reported by Sven Panne). - - - - - 6d98989c by Simon Marlow at 2002-06-05T09:12:02+00:00 [haddock @ 2002-06-05 09:12:02 by simonmar] Identifiers in single quotes can be symbol names too (bug reported by Hal Daume). - - - - - 001811e5 by Sven Panne at 2002-06-08T14:03:36+00:00 [haddock @ 2002-06-08 14:03:36 by panne] Tiny workaround for the fact that Haddock currently ignores HsImportSpecs: Let the local_orig_env take precedence. This is no real solution at all, but improves things sometimes, e.g. in my GLUT documentation. :-) - - - - - 504d19c9 by Simon Marlow at 2002-06-11T09:23:25+00:00 [haddock @ 2002-06-11 09:23:25 by simonmar] portability nit - - - - - e13b5af4 by Simon Marlow at 2002-06-20T12:38:07+00:00 [haddock @ 2002-06-20 12:38:07 by simonmar] Empty declaration fixes. - - - - - f467a9b6 by Simon Marlow at 2002-06-20T12:39:02+00:00 [haddock @ 2002-06-20 12:39:01 by simonmar] Add support for a "prologue" - a description for the whole library, placed on the contents page before the module list. - - - - - b8dbfe20 by Simon Marlow at 2002-06-21T12:43:06+00:00 [haddock @ 2002-06-21 12:43:06 by simonmar] When we have a single code block paragraph, don't place it in <pre>..</pre>, just use <tt>..</tt> to avoid generating extra vertical white space in some browsers. - - - - - 4831dbbd by Simon Marlow at 2002-06-21T15:50:42+00:00 [haddock @ 2002-06-21 15:50:42 by simonmar] Add support for reading and writing interface files(!) This turned out to be quite easy, and necessary to get decent hyperlinks between the documentation for separate packages in the libraries. The functionality isn't quite complete yet: for a given package of modules, you'd like to say "the HTML for these modules lives in directory <dir>" (currently they are assumed to be all in the same place). Two new flags: --dump-interface=FILE dump an interface file in FILE --read-interface=FILE read interface from FILE an interface file describes *all* the modules being processed. Only the exported names are kept in the interface: if you re-export a name from a module in another interface the signature won't be copied. This is a compromise to keep the size of the interfaces sensible. Also, I added another useful option: --no-implicit-prelude avoids trying to import the Prelude. Previously this was the default, but now importing the Prelude from elsewhere makes sense if you also read in an interface containing the Prelude module, so Haddock imports the Prelude implicitly according to the Haskell spec. - - - - - d3640a19 by Sven Panne at 2002-06-23T14:54:00+00:00 [haddock @ 2002-06-23 14:54:00 by panne] Make it compile with newer GHCs - - - - - 780c506b by Sven Panne at 2002-06-23T15:44:31+00:00 [haddock @ 2002-06-23 15:44:31 by panne] Cleaned up build root handling and added more docs - - - - - 45290d2e by Simon Marlow at 2002-06-24T14:37:43+00:00 [haddock @ 2002-06-24 14:37:42 by simonmar] When reading an interface, allow a file path offset to be specified which represents the path to the HTML files for the modules specified by that interface. The path may be either relative (to the location of the HTML for this package), or absolute. The syntax is --read-interface=PATH,FILE where PATH is the path to the HTML, and FILE is the filename containing the interface. - - - - - 4e2b9ae6 by Simon Marlow at 2002-07-03T16:01:08+00:00 [haddock @ 2002-07-03 16:01:07 by simonmar] Handle import specs properly, include 'hiding'. Haddock now has a complete implementation of the Haskell module system (more or less; I won't claim it's 100% correct). - - - - - 9a9aa1a8 by Simon Marlow at 2002-07-03T16:18:16+00:00 [haddock @ 2002-07-03 16:18:16 by simonmar] Update - - - - - 560c3026 by Simon Marlow at 2002-07-04T14:56:10+00:00 [haddock @ 2002-07-04 14:56:10 by simonmar] Clean up the code that constructs the exported declarations, and fix a couple of bugs along the way. Now if you import a class hiding one of the methods, then re-export the class, the version in the documentation will correctly have the appropriate method removed. - - - - - 2c26e77d by Simon Marlow at 2002-07-04T15:26:13+00:00 [haddock @ 2002-07-04 15:26:13 by simonmar] More bugfixes to the export handling - - - - - 03e0710d by Simon Marlow at 2002-07-09T10:12:10+00:00 [haddock @ 2002-07-09 10:12:10 by simonmar] Don't require that the list type comes from "Prelude" for it to be treated as special syntax (sometimes it comes from Data.List or maybe even GHC.Base). - - - - - 44f3891a by Simon Marlow at 2002-07-09T10:12:51+00:00 [haddock @ 2002-07-09 10:12:51 by simonmar] commented-out debugging code - - - - - 97280873 by Krasimir Angelov at 2002-07-09T16:33:33+00:00 [haddock @ 2002-07-09 16:33:31 by krasimir] 'Microsoft HTML Help' support - - - - - 3dc04655 by Simon Marlow at 2002-07-10T09:40:56+00:00 [haddock @ 2002-07-10 09:40:56 by simonmar] Fix for rendering of the (->) type constructor, from Ross Paterson. - - - - - c9f149c6 by Simon Marlow at 2002-07-10T10:26:11+00:00 [haddock @ 2002-07-10 10:26:11 by simonmar] Tweaks to the MS Help support: the extra files are now only generated if you ask for them (--ms-help). - - - - - e8acc1e6 by Simon Marlow at 2002-07-10T10:57:10+00:00 [haddock @ 2002-07-10 10:57:10 by simonmar] Document all the new options since 0.3 - - - - - 8bb85544 by Simon Marlow at 2002-07-10T10:58:31+00:00 [haddock @ 2002-07-10 10:58:31 by simonmar] Sort the options a bit - - - - - abc0dd59 by Simon Marlow at 2002-07-15T09:19:38+00:00 [haddock @ 2002-07-15 09:19:38 by simonmar] Fix a bug in mkExportItems when processing a module without an explicit export list. We were placing one copy of a declaration for each binder in the declaration, which for a data type would mean one copy of the whole declaration per constructor or record selector. - - - - - dde65bb9 by Simon Marlow at 2002-07-15T09:54:16+00:00 [haddock @ 2002-07-15 09:54:16 by simonmar] merge rev. 1.35 - - - - - bd7eb8c4 by Simon Marlow at 2002-07-15T10:14:31+00:00 [haddock @ 2002-07-15 10:14:30 by simonmar] Be a bit more liberal in the kind of commenting styles we allow, as suggested by Malcolm Wallace. Mostly this consists of allowing doc comments either side of a separator token. In an export list, a section heading is now allowed before the comma, as well as after it. eg. module M where ( T(..) -- * a section heading , f -- * another section heading , g ) In record fields, doc comments are allowed anywhere (previously a doc-next was allowed only after the comma, and a doc-before was allowed only before the comma). eg. data R = C { -- | describes 'f' f :: Int -- | describes 'g' , g :: Int } - - - - - 8f6dfe34 by Simon Marlow at 2002-07-15T10:21:56+00:00 [haddock @ 2002-07-15 10:21:56 by simonmar] Mention alternative commenting styles. - - - - - fc515bb7 by Simon Marlow at 2002-07-15T16:16:50+00:00 [haddock @ 2002-07-15 16:16:50 by simonmar] Allow multiple sections/subsections before and after a comma in the export list. Also at the same time I made the syntax a little stricter (multiple commas now aren't allowed between export specs). - - - - - 80a97e74 by Simon Marlow at 2002-07-19T09:13:10+00:00 [haddock @ 2002-07-19 09:13:10 by simonmar] Allow special id's ([], (), etc.) to be used in an import declaration. - - - - - a69d7378 by Simon Marlow at 2002-07-19T09:59:02+00:00 [haddock @ 2002-07-19 09:59:02 by simonmar] Allow special id's ([], (), etc.) to be used in an import declarations. - - - - - d205fa60 by Simon Marlow at 2002-07-19T10:00:16+00:00 [haddock @ 2002-07-19 10:00:16 by simonmar] Relax the restrictions which require doc comments to be followed by semi colons - in some cases this isn't necessary. Now you can write module M where { -- | some doc class C where {} } without needing to put a semicolon before the class declaration. - - - - - e9301e14 by Simon Marlow at 2002-07-23T08:24:09+00:00 [haddock @ 2002-07-23 08:24:09 by simonmar] A new TODO list item - - - - - e5d77586 by Simon Marlow at 2002-07-23T08:40:56+00:00 [haddock @ 2002-07-23 08:40:56 by simonmar] - update the acknowledgements - remove the paragraph that described how to use explicit layout with doc comments; it isn't relevant any more. - - - - - 78a94137 by Simon Marlow at 2002-07-23T08:43:02+00:00 [haddock @ 2002-07-23 08:43:02 by simonmar] more tests - - - - - 5c320927 by Simon Marlow at 2002-07-23T08:43:26+00:00 [haddock @ 2002-07-23 08:43:26 by simonmar] Updates for version 0.4 - - - - - 488e99ae by Simon Marlow at 2002-07-23T09:10:46+00:00 [haddock @ 2002-07-23 09:10:46 by simonmar] Fix the %changelog (rpm complained that it wasn't in the right order) - - - - - a77bb373 by Simon Marlow at 2002-07-23T09:12:38+00:00 [haddock @ 2002-07-23 09:12:38 by simonmar] Another item for the TODO list - - - - - f1ec1813 by Simon Marlow at 2002-07-23T10:18:46+00:00 [haddock @ 2002-07-23 10:18:46 by simonmar] Add a version banner when invoked with -v - - - - - 1d44cadf by Simon Marlow at 2002-07-24T09:28:19+00:00 [haddock @ 2002-07-24 09:28:19 by simonmar] Remove ^Ms - - - - - 4d8d5e94 by Simon Marlow at 2002-07-24T09:42:18+00:00 [haddock @ 2002-07-24 09:42:17 by simonmar] Patches to quieten ghc -Wall, from those nice folks at Galois. - - - - - d6edc43e by Simon Marlow at 2002-07-25T14:37:29+00:00 [haddock @ 2002-07-25 14:37:28 by simonmar] Patch to allow simple hyperlinking to an arbitrary location in another module's documentation, from Volker Stolz. Now in a doc comment: #foo# creates <a name="foo"></a> And you can use the form "M\#foo" to hyperlink to the label 'foo' in module 'M'. Note that the backslash is necessary for now. - - - - - b34d18fa by Simon Marlow at 2002-08-02T09:08:22+00:00 [haddock @ 2002-08-02 09:08:22 by simonmar] The <TT> and <PRE> environments seem to use a font that is a little too small in IE. Compensate. (suggestion from Daan Leijen). - - - - - 8106b086 by Simon Marlow at 2002-08-02T09:25:23+00:00 [haddock @ 2002-08-02 09:25:20 by simonmar] Remove <P>..</P> from around list items, to reduce excess whitespace between the items of bulleted and ordered lists. (Suggestion from Daan Leijen). - - - - - c1acff8f by Simon Marlow at 2002-08-05T09:03:49+00:00 [haddock @ 2002-08-05 09:03:49 by simonmar] update - - - - - f968661c by Simon Marlow at 2002-11-11T09:32:57+00:00 [haddock @ 2002-11-11 09:32:57 by simonmar] Fix cut-n-pasto - - - - - 12d02619 by Simon Marlow at 2002-11-13T09:49:46+00:00 [haddock @ 2002-11-13 09:49:46 by simonmar] Small bugfix in the --read-interface option parsing from Brett Letner. - - - - - 30e32d5e by Ross Paterson at 2003-01-16T15:07:57+00:00 [haddock @ 2003-01-16 15:07:57 by ross] Adjust for the new exception libraries (as well as the old ones). - - - - - 871f65df by Sven Panne at 2003-02-20T21:31:40+00:00 [haddock @ 2003-02-20 21:31:40 by panne] * Add varsyms and consyms to index * Exclude empty entries from index - - - - - bc42cc87 by Sven Panne at 2003-02-24T21:26:29+00:00 [haddock @ 2003-02-24 21:26:29 by panne] Don't convert a "newtype" to a single-constructor "data" for non-abstractly exported types, they are quite different regarding strictness/pattern matching. Now a "data" without any constructors is only emitted for an abstractly exported type, regardless if it is actually a "newtype" or a "data". - - - - - 0c2a1d99 by Sven Panne at 2003-03-08T19:02:38+00:00 [haddock @ 2003-03-08 19:02:38 by panne] Fixed some broken/redirected/canonicalized links found by a very picky link checker. - - - - - 25459269 by Sven Panne at 2003-03-09T21:13:43+00:00 [haddock @ 2003-03-09 21:13:43 by panne] Don't append a fragment to non-defining index entries, only documents with a defining occurrence have a name anchor. - - - - - 6be4db86 by Sven Panne at 2003-03-10T21:34:25+00:00 [haddock @ 2003-03-10 21:34:24 by panne] Escape fragments. This fixes e.g. links to operators. - - - - - eb12972c by Ross Paterson at 2003-04-25T10:50:06+00:00 [haddock @ 2003-04-25 10:50:05 by ross] An 80% solution to generating derived instances. A complete solution would duplicate the instance inference logic, but if a type variable occurs as a constructor argument, then we can just propagate the derived class to the variable. But we know nothing of the constraints on any type variables that occur elsewhere. For example, the declarations data Either a b = Left a | Right b deriving (Eq, Ord) data Ptr a = Ptr Addr# deriving (Eq, Ord) newtype IORef a = IORef (STRef RealWorld a) deriving Eq yield the instances (Eq a, Eq b) => Eq (Either a b) (Ord a, Ord b) => Ord (Either a b) Eq (Ptr a) Ord (Ptr a) (??? a) => Eq (IORef a) The last example shows the limits of this local analysis. Note that a type variable may be in both categories: then we know a constraint, but there may be more, or a stronger constraint, e.g. data Tree a = Node a [Tree a] deriving Eq yields (Eq a, ??? a) => Eq (Tree a) - - - - - de886f78 by Simon Marlow at 2003-04-25T11:17:55+00:00 [haddock @ 2003-04-25 11:17:55 by simonmar] Some updates, including moving the derived instance item down to the bottom of the list now that Ross has contributed some code that does the job for common cases. - - - - - 1b52cffd by Simon Marlow at 2003-04-30T14:02:32+00:00 [haddock @ 2003-04-30 14:02:32 by simonmar] When installing on Windows, run cygpath over $(HADDOCKLIB) so that haddock (a mingw program, built by GHC) can understand it. You still need to be in a cygwin environment to run Haddock, because of the shell script wrapper. - - - - - d4f638de by Simon Marlow at 2003-05-06T10:04:47+00:00 [haddock @ 2003-05-06 10:04:47 by simonmar] Catch another case of a paragraph containing just a DocMonospaced that should turn into a DocCodeBlock. - - - - - 4162b2b9 by Simon Marlow at 2003-05-06T10:11:44+00:00 [haddock @ 2003-05-06 10:11:44 by simonmar] Add some more code-block tests. - - - - - 4f5802c8 by Simon Marlow at 2003-05-06T10:14:52+00:00 [haddock @ 2003-05-06 10:14:52 by simonmar] Don't turn a single DocCodeBlock into a DocMonospaced, because that tends to remove the line breaks in the code. - - - - - ef8c45f7 by Simon Marlow at 2003-05-21T15:07:21+00:00 [haddock @ 2003-05-21 15:07:21 by simonmar] Only omit the module contents when there are no section headings at all. - - - - - bcee1e75 by Sigbjorn Finne at 2003-05-30T16:50:45+00:00 [haddock @ 2003-05-30 16:50:45 by sof] cygpath: for now, steer clear of --mixed - - - - - 30567af3 by Sigbjorn Finne at 2003-05-30T17:59:28+00:00 [haddock @ 2003-05-30 17:59:28 by sof] oops, drop test defn from prev commit - - - - - b0856e7d by Simon Marlow at 2003-06-03T09:55:26+00:00 [haddock @ 2003-06-03 09:55:26 by simonmar] Two small fixes to make the output valid HTML 4.01 (transitional). Thanks to Malcolm Wallace for pointing out the problems. - - - - - 70e137ea by Simon Marlow at 2003-07-28T13:30:35+00:00 [haddock @ 2003-07-28 13:30:35 by simonmar] Add tests for a couple of bugs. - - - - - 122bd578 by Simon Marlow at 2003-07-28T13:31:25+00:00 [haddock @ 2003-07-28 13:31:25 by simonmar] Add documentation for anchors. - - - - - 0bd27cb2 by Simon Marlow at 2003-07-28T13:31:46+00:00 [haddock @ 2003-07-28 13:31:46 by simonmar] Update - - - - - 08052d42 by Simon Marlow at 2003-07-28T13:32:12+00:00 [haddock @ 2003-07-28 13:32:12 by simonmar] layout tweak. - - - - - 13942749 by Simon Marlow at 2003-07-28T13:33:03+00:00 [haddock @ 2003-07-28 13:33:03 by simonmar] Differentiate links to types/classes from links to variables/constructors with a prefix ("t:" and "v:" respectively). - - - - - d7f493b9 by Simon Marlow at 2003-07-28T13:35:17+00:00 [haddock @ 2003-07-28 13:35:16 by simonmar] When a module A exports another module's contents via 'module B', then modules which import entities from B re-exported by A should link to B.foo rather than A.foo. See examples/Bug2.hs. - - - - - d94cf705 by Simon Marlow at 2003-07-28T13:36:14+00:00 [haddock @ 2003-07-28 13:36:14 by simonmar] Update to version 0.5 - - - - - dbb776cd by Sven Panne at 2003-07-28T14:02:43+00:00 [haddock @ 2003-07-28 14:02:43 by panne] * Updated to version 0.5 * Automagically generate configure if it is not there - - - - - 6cfeee53 by Simon Marlow at 2003-07-28T14:32:43+00:00 [haddock @ 2003-07-28 14:32:42 by simonmar] Update to avoid using hslibs with GHC >= 5.04 - - - - - a1ce838f by Simon Marlow at 2003-07-28T14:33:37+00:00 [haddock @ 2003-07-28 14:33:37 by simonmar] Update for 0.5 - - - - - c0fe6493 by Simon Marlow at 2003-07-28T14:53:22+00:00 [haddock @ 2003-07-28 14:53:22 by simonmar] Markup fix - - - - - 6ea31596 by Sven Panne at 2003-07-28T16:40:45+00:00 [haddock @ 2003-07-28 16:40:45 by panne] Make it compile with GHC >= 6.01 - - - - - afcd30fc by Simon Marlow at 2003-07-30T15:04:52+00:00 [haddock @ 2003-07-30 15:04:52 by simonmar] Pay attention to import specs when building the the import env, as well as the orig env. This may fix some wrong links in documentation when import specs are being used. - - - - - 17c3137f by Simon Marlow at 2003-07-30T16:05:41+00:00 [haddock @ 2003-07-30 16:05:40 by simonmar] Rename instances based on the import_env for the module in which they are to be displayed. This should give, in many cases, better links for the types and classes mentioned in the instance head. This involves keeping around the import_env in the iface until the end, because instances are not collected up until all the modules have been processed. Fortunately it doesn't seem to affect performance much. Instance heads are now attached to ExportDecls, rather than the HTML backend passing around a separate mapping for instances. This is a cleanup. - - - - - 3d3b5c87 by Sven Panne at 2003-08-04T10:18:24+00:00 [haddock @ 2003-08-04 10:18:24 by panne] Don't print parentheses around one-element contexts - - - - - 9e3f3f2d by Simon Marlow at 2003-08-04T12:59:47+00:00 [haddock @ 2003-08-04 12:59:47 by simonmar] A couple of TODOs. - - - - - e9d8085c by Simon Marlow at 2003-08-05T14:10:31+00:00 [haddock @ 2003-08-05 14:10:31 by simonmar] I'm not sure why, but it seems that the index entries for non-defining occurrences of entities did not have an anchor - the link just pointed to the module. This fixes it. - - - - - ff5c7d6d by Simon Marlow at 2003-08-15T14:42:59+00:00 [haddock @ 2003-08-15 14:42:59 by simonmar] Convert the lexer to Alex, and fix a bug in the process. - - - - - 1aa077bf by Simon Marlow at 2003-08-15T15:00:18+00:00 [haddock @ 2003-08-15 15:00:18 by simonmar] Update - - - - - d3de1e38 by Simon Marlow at 2003-08-15T15:01:03+00:00 [haddock @ 2003-08-15 15:01:03 by simonmar] wibbles - - - - - b40ece3b by Simon Marlow at 2003-08-18T10:04:47+00:00 [haddock @ 2003-08-18 10:04:47 by simonmar] Lex the 'mdo' keyword as 'do'. - - - - - 8f9a1146 by Simon Marlow at 2003-08-18T11:48:24+00:00 [haddock @ 2003-08-18 11:48:24 by simonmar] Two bugs from Sven. - - - - - ea54ebc0 by Simon Marlow at 2003-08-18T11:48:46+00:00 [haddock @ 2003-08-18 11:48:46 by simonmar] Fixes to the new lexer. - - - - - d5f6a4b5 by Simon Marlow at 2003-08-19T09:09:03+00:00 [haddock @ 2003-08-19 09:09:03 by simonmar] Further wibbles to the syntax. - - - - - 6bbdadb7 by Sven Panne at 2003-08-26T18:45:35+00:00 [haddock @ 2003-08-26 18:45:35 by panne] Use autoreconf instead of autoconf - - - - - 32e889cb by Sven Panne at 2003-08-26T19:01:19+00:00 [haddock @ 2003-08-26 19:01:18 by panne] Made option handling a bit more consistent with other tools, in particular: Every program in fptools should output * version info on stdout and terminate successfully when -V or --version * usage info on stdout and terminate successfully when -? or --help * usage info on stderr and terminate unsuccessfully when an unknown option is given. - - - - - 5d156a91 by Sven Panne at 2003-08-26T19:20:55+00:00 [haddock @ 2003-08-26 19:20:55 by panne] Make it *very* clear that we terminate when given a -V/--version flag - - - - - e6577265 by Sven Panne at 2003-08-27T07:50:03+00:00 [haddock @ 2003-08-27 07:50:02 by panne] * Made -D a short option for --dump-interface. * Made -m a short option for --ms-help. * Made -n a short option for --no-implicit-prelude. * Made -c a short option for --css. * Removed DocBook options from executable (they didn't do anything), but mark them as reserved in the docs. Note that the short option for DocBook output is now -S (from SGML) instead of -d. The latter is now a short option for --debug. * The order of the Options in the documentation now matches the order printed by Haddock itself. Note: Although changing the names of options is often a bad idea, I'd really like to make the options for the programs in fptools more consistent and compatible to the ones used in common GNU programs. - - - - - d303ff98 by Simon Marlow at 2003-09-10T08:23:48+00:00 [haddock @ 2003-09-10 08:23:48 by simonmar] Add doc subdir. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 9a70e46a by Simon Marlow at 2003-09-10T08:24:32+00:00 [haddock @ 2003-09-10 08:24:32 by simonmar] Install these files in $(datadir), not $(libdir), since they're architecture independent. Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - bbb87e7a by Simon Marlow at 2003-09-10T08:25:31+00:00 [haddock @ 2003-09-10 08:25:31 by simonmar] Haddock's supplementary HTML bits now live in $(datadir), not $(libdir). Patch contributed by: Ian Lynagh <igloo at earth.li>. - - - - - 3587c24b by Simon Marlow at 2003-09-22T10:34:38+00:00 [haddock @ 2003-09-22 10:34:38 by simonmar] Allow installing of docs. - - - - - d510b517 by Sven Panne at 2003-10-11T08:10:44+00:00 [haddock @ 2003-10-11 08:10:44 by panne] Include architecture-independent files in file list - - - - - 187d7618 by Sigbjorn Finne at 2003-10-20T17:19:24+00:00 [haddock @ 2003-10-20 17:19:22 by sof] support for i-parameters + zip comprehensions - - - - - b6c7a273 by Simon Marlow at 2003-11-03T14:24:24+00:00 [haddock @ 2003-11-03 14:24:24 by simonmar] Update TODO file. - - - - - 58513e33 by Simon Marlow at 2003-11-05T11:22:04+00:00 [haddock @ 2003-11-05 11:22:04 by simonmar] Remove the last of the uses of 'trace' to emit warnings, and tidy up a couple of places where duplicate warnings were being emitted. - - - - - 33a78846 by Simon Marlow at 2003-11-05T11:30:53+00:00 [haddock @ 2003-11-05 11:30:52 by simonmar] - Suppress warnings about unknown imported modules by default. - Add a -v/--verbose flag to re-enable these warnings. The general idea is to suppress the "Warning: unknown module: Prelude" warnings which most Haddock users will see every time, and which aren't terribly useful. - - - - - a969de7f by Simon Marlow at 2003-11-05T12:30:28+00:00 [haddock @ 2003-11-05 12:30:28 by simonmar] - Remove the emboldening of index entries for defining locations. This isn't useful, and breaks abstractions. - If an entity is re-exported by a module but the module doesn't include documentation for that entity (perhaps because it is re-exported by 'module M'), then don't attempt to hyperlink to the documentation from the index. Instead, just list that module in the index, to indicate that the entity is exported from there. - - - - - f14ea82a by Simon Marlow at 2003-11-05T15:15:59+00:00 [haddock @ 2003-11-05 15:15:59 by simonmar] Index overhaul: - no more separate type/class and variable/function indices - the index now makes a distinction between different entities with the same name. One example is a type constructor with the same name as a data constructor, but another example is simply a function with the same name exported by two different modules. For example, the index entry for 'catch' now looks like this: catch 1 (Function) Control.Exception 2 (Function) GHC.Exception, Prelude, System.IO, System.IO.Error making it clear that there are two different 'catch'es, but one of them is exported by several modules. - Each index page now has the index contents (A B C ...) at the top. Please let me know if you really hate any of this. - - - - - 01a25ca6 by Simon Marlow at 2003-11-05T15:16:38+00:00 [haddock @ 2003-11-05 15:16:38 by simonmar] Update - - - - - 1a7ccb86 by Simon Marlow at 2003-11-05T17:16:05+00:00 [haddock @ 2003-11-05 17:16:04 by simonmar] Support for generating a single unified index for several packages. --use-index=URL turns off normal index generation, causes Index links to point to URL. --gen-index generates an combined index from the specified interfaces. Currently doesn't work exactly right, because the interfaces don't contain the iface_reexported info. I'll need to fix that up. - - - - - a2bca16d by Simon Marlow at 2003-11-06T10:44:52+00:00 [haddock @ 2003-11-06 10:44:52 by simonmar] Include iface_reexported in the .haddock file. This unfortunately bloats the file (40% for base). If this gets to be a problem we can always apply the dictionary trick that GHC uses for squashing .hi files. - - - - - 0a09c293 by Simon Marlow at 2003-11-06T12:39:47+00:00 [haddock @ 2003-11-06 12:39:46 by simonmar] - Add definition lists, marked up like this: -- | This is a definition list: -- -- [@foo@] The description of @foo at . -- -- [@bar@] The description of @bar at . Cunningly, the [] characters are not treated specially unless a [ is found at the beginning of a paragraph, in which case the ] becomes special in the following text. - Add --use-contents and --gen-contents, along the lines of --use-index and --gen-index added yesterday. Now we can generate a combined index and contents for the whole of the hierarchical libraries, and in theory the index/contents on the system could be updated as new packages are added. - - - - - fe1b3460 by Simon Marlow at 2003-11-06T14:47:36+00:00 [haddock @ 2003-11-06 14:47:36 by simonmar] Remove the 'Parent' button - it is of dubious use, and often points into thin air. - - - - - db6d762f by Simon Marlow at 2003-11-06T16:48:14+00:00 [haddock @ 2003-11-06 16:48:11 by simonmar] - Include the OptHide setting in the interface, so we don't include hidden modules in the combined index/contents. - Add a -k/--package flag to set the package name for the current set of modules. The package name for each module is now shown in the right-hand column of the contents, in a combined contents page. - - - - - 7d71718b by Simon Marlow at 2003-11-06T16:50:28+00:00 [haddock @ 2003-11-06 16:50:28 by simonmar] Add -k/--package docs - - - - - ef43949d by Simon Marlow at 2003-11-06T16:51:23+00:00 [haddock @ 2003-11-06 16:51:23 by simonmar] Bump to 0.6 - - - - - 1c419e06 by Simon Marlow at 2003-11-06T16:51:50+00:00 [haddock @ 2003-11-06 16:51:50 by simonmar] update - - - - - 69422327 by Simon Marlow at 2003-11-10T14:41:06+00:00 [haddock @ 2003-11-10 14:41:05 by simonmar] Re-exporting names from a different package is problematic, because we don't have access to the full documentation for the entity. Currently Haddock just ignores entities with no documentation, but this results in bogus-looking empty documentation for many of the modules in the haskell98 package. So: - the documentation will now just list the name, as a link pointing to the location of the actual documentation. - now we don't attempt to link to these re-exported entities if they are referred to by the current module. Additionally: - If there is no documentation in the current module, include just the Synopsis section (rather than just the documentation section, as it was before). This just looks nicer and was on the TODO list. - - - - - 3c3fc433 by Simon Marlow at 2003-11-10T14:51:59+00:00 [haddock @ 2003-11-10 14:51:59 by simonmar] Fix for getReExports: take into account names which are not visible because they are re-exported from a different package. - - - - - 31c8437b by Simon Marlow at 2003-11-10T15:10:53+00:00 [haddock @ 2003-11-10 15:10:53 by simonmar] Version 0.6 changes - - - - - a7c2430b by Simon Marlow at 2003-11-10T15:15:58+00:00 [haddock @ 2003-11-10 15:15:58 by simonmar] getReExports: one error case that isn't - - - - - 00cc459c by Simon Marlow at 2003-11-10T16:15:19+00:00 [haddock @ 2003-11-10 16:15:18 by simonmar] copyright update - - - - - ca62408d by Simon Marlow at 2003-11-11T09:57:25+00:00 [haddock @ 2003-11-11 09:57:25 by simonmar] Version 0.6 - - - - - 3acbf818 by Simon Marlow at 2003-11-11T12:10:44+00:00 [haddock @ 2003-11-11 12:10:44 by simonmar] Go back to producing just the documentation section, rather than just the synopsis section, for a module with no documentation annotations. One reason is that the synopsis section tries to link each entity to its documentation on the same page. Also, the doc section anchors each entity, and it lists instances which the synopsis doesn't. - - - - - 6c90abc2 by Simon Marlow at 2003-11-12T10:03:39+00:00 [haddock @ 2003-11-12 10:03:39 by simonmar] 2002 -> 2003 - - - - - 090bbc4c by Simon Marlow at 2003-11-28T12:08:00+00:00 [haddock @ 2003-11-28 12:08:00 by simonmar] update - - - - - 8096a832 by Simon Marlow at 2003-11-28T12:09:58+00:00 [haddock @ 2003-11-28 12:09:58 by simonmar] Fix some of the problems with Haddock generating pages that are too wide. Now we only specify 'nowrap' when it is necessary to avoid a code box getting squashed up by the text to the right of it. - - - - - 35294929 by Sven Panne at 2003-12-29T17:16:31+00:00 [haddock @ 2003-12-29 17:16:31 by panne] Updated my email address - - - - - cdb697bf by Simon Marlow at 2004-01-08T10:14:24+00:00 [haddock @ 2004-01-08 10:14:24 by simonmar] Add instructions for using GHC to pre-process source for feeding to Haddock. - - - - - 8dfc491f by Simon Marlow at 2004-01-09T12:45:46+00:00 [haddock @ 2004-01-09 12:45:46 by simonmar] Add -optP-P to example ghc command line. - - - - - ac41b820 by Simon Marlow at 2004-02-03T11:02:03+00:00 [haddock @ 2004-02-03 11:02:03 by simonmar] Fix bug in index generation - - - - - f4e7edcb by Simon Marlow at 2004-02-10T11:51:16+00:00 [haddock @ 2004-02-10 11:51:16 by simonmar] Don't throw away whitespace at the beginning of a line (experimental fix). - - - - - 68e212d2 by Simon Marlow at 2004-02-10T12:10:08+00:00 [haddock @ 2004-02-10 12:10:08 by simonmar] Fix for previous commit: I now realise why the whitespace was stripped from the beginning of the line. Work around it. - - - - - e7d7f2df by Sven Panne at 2004-02-10T18:38:45+00:00 [haddock @ 2004-02-10 18:38:45 by panne] Make Haddock link with the latest relocated monad transformer package - - - - - 992d4225 by Simon Marlow at 2004-02-16T10:21:35+00:00 [haddock @ 2004-02-16 10:21:35 by simonmar] Add a TODO - - - - - 1ac55326 by Simon Marlow at 2004-03-12T11:33:39+00:00 [haddock @ 2004-03-12 11:33:39 by simonmar] Add an item. - - - - - 0478e903 by Simon Marlow at 2004-03-15T12:24:05+00:00 [haddock @ 2004-03-15 12:24:05 by simonmar] Add an item. - - - - - 6f26d21a by Simon Marlow at 2004-03-18T14:21:29+00:00 [haddock @ 2004-03-18 14:21:29 by simonmar] Fix URL - - - - - 19b6bb99 by Simon Marlow at 2004-03-22T14:09:03+00:00 [haddock @ 2004-03-22 14:09:03 by simonmar] getReExports was bogus: we should really look in the import_env to find the documentation for an entity which we are re-exporting without documentation. Suggested by: Ross Paterson (patch modified by me). - - - - - 5c756031 by Simon Marlow at 2004-03-24T09:42:11+00:00 [haddock @ 2004-03-24 09:42:10 by simonmar] hiding bug from Ross Paterson (fixed in rev 1.59 of Main.hs) - - - - - 1b692e6c by Simon Marlow at 2004-03-24T10:10:50+00:00 [haddock @ 2004-03-24 10:10:50 by simonmar] mkExportItems fix & simplification: we should be looking at the actual exported names (calculated earlier) to figure out which subordinates of a declaration are exported. This means that if you export a record, and name its fields separately in the export list, the fields will still be visible in the documentation for the constructor. - - - - - 90e5e294 by Simon Marlow at 2004-03-24T10:12:08+00:00 [haddock @ 2004-03-24 10:12:08 by simonmar] Make restrictCons take into account record field names too (removing a ToDo). - - - - - 2600efa4 by Simon Marlow at 2004-03-24T10:16:17+00:00 [haddock @ 2004-03-24 10:16:17 by simonmar] Record export tests. - - - - - 6a8575c7 by Simon Marlow at 2004-03-25T09:35:14+00:00 [haddock @ 2004-03-25 09:35:14 by simonmar] restrictTo: fix for restricting a newtype with a record field. - - - - - dcf55a8d by Simon Marlow at 2004-03-25T10:01:42+00:00 [haddock @ 2004-03-25 10:01:42 by simonmar] Fix duplicate instance bug - - - - - f49aa758 by Simon Marlow at 2004-03-25T10:02:41+00:00 [haddock @ 2004-03-25 10:02:41 by simonmar] Duplicate instance bug. - - - - - 7b87344c by Simon Marlow at 2004-03-25T10:29:56+00:00 [haddock @ 2004-03-25 10:29:56 by simonmar] If a name is imported from two places, one hidden and one not, choose the unhidden one to link to. Also, when there's only a hidden module to link to, don't try linking to it. - - - - - 40f44d7b by Simon Marlow at 2004-03-25T15:17:24+00:00 [haddock @ 2004-03-25 15:17:23 by simonmar] Add support for collaspible parts of the page, with a +/- button and a bit of JavaScript. Make the instances collapsible, and collapse them by default. This makes documentation with long lists of instances (eg. the Prelude) much easier to read. Maybe we should give other documentation sections the same treatment. - - - - - 9b64dc0f by Simon Marlow at 2004-03-25T15:20:55+00:00 [haddock @ 2004-03-25 15:20:55 by simonmar] Update - - - - - c2fff7f2 by Simon Marlow at 2004-03-25T15:45:10+00:00 [haddock @ 2004-03-25 15:45:10 by simonmar] Eliminate some unnecessary spaces in the HTML rendering - - - - - b7948ff0 by Simon Marlow at 2004-03-25T16:00:37+00:00 [haddock @ 2004-03-25 16:00:36 by simonmar] Remove all that indentation in the generated HTML to keep the file sizes down. - - - - - da2bb4ca by Sven Panne at 2004-03-27T09:57:58+00:00 [haddock @ 2004-03-27 09:57:57 by panne] Added the new-born haddock.js to the build process and the documentation. - - - - - b99e6f8c by Sven Panne at 2004-03-27T10:32:20+00:00 [haddock @ 2004-03-27 10:32:20 by panne] "type" is a required attribute of the "script" element - - - - - 562b185a by Sven Panne at 2004-03-27T12:52:34+00:00 [haddock @ 2004-03-27 12:52:34 by panne] Add a doctype for the contents page, too. - - - - - f6a99c2d by Simon Marlow at 2004-04-14T10:03:25+00:00 [haddock @ 2004-04-14 10:03:25 by simonmar] fix for single-line comment syntax - - - - - de366303 by Simon Marlow at 2004-04-20T13:08:04+00:00 [haddock @ 2004-04-20 13:08:04 by simonmar] Allow a 'type' declaration to include documentation comments. These will be ignored by Haddock, but at least one user (Johannes Waldmann) finds this feature useful, and it's easy to add. - - - - - fd78f51e by Simon Marlow at 2004-05-07T15:14:56+00:00 [haddock @ 2004-05-07 15:14:56 by simonmar] - update copyright - add version to abstract - - - - - 59f53e32 by Sven Panne at 2004-05-09T14:39:53+00:00 [haddock @ 2004-05-09 14:39:53 by panne] Fix the fix for single-line comment syntax, ------------------------------------------- is now a valid comment line again. - - - - - 8b18f2fe by Simon Marlow at 2004-05-10T10:11:51+00:00 [haddock @ 2004-05-10 10:11:51 by simonmar] Update - - - - - 225a491d by Ross Paterson at 2004-05-19T13:10:23+00:00 [haddock @ 2004-05-19 13:10:23 by ross] Make the handling of "deriving" slightly smarter, by ignoring data constructor arguments that are identical to the lhs. Now handles things like data Tree a = Leaf a | Branch (Tree a) (Tree a) deriving ... - - - - - 37588686 by Mike Thomas at 2004-05-21T06:38:14+00:00 [haddock @ 2004-05-21 06:38:14 by mthomas] Windows exe extensions (bin remains for Unix). - - - - - cf2b9152 by Simon Marlow at 2004-05-25T09:34:54+00:00 [haddock @ 2004-05-25 09:34:54 by simonmar] Add some TODO items - - - - - 4d29cdfc by Simon Marlow at 2004-05-25T10:41:46+00:00 [haddock @ 2004-05-25 10:41:46 by simonmar] Complain if -h is used with --gen-index or --gen-contents, because it'll overwrite the new index/contents. - - - - - 2e0771e0 by Mike Thomas at 2004-05-28T20:17:55+00:00 [haddock @ 2004-05-28 20:17:55 by mthomas] Windows: search for templates in executable directory. Unix: Haddock tries cwd first rather than error if no -l arg. - - - - - 8d10bde1 by Sven Panne at 2004-06-05T16:53:34+00:00 [haddock @ 2004-06-05 16:53:34 by panne] Misc. rpm spec file cleanup, including: * make BuildRoot handling more consistent * added default file attributes * consistent defines and tags - - - - - 59974349 by Sven Panne at 2004-06-05T18:01:00+00:00 [haddock @ 2004-06-05 18:01:00 by panne] More rpm spec file cleanup, including: * added some BuildRequires * changed packager to me, so people can complain at the right place :-] * consistently refer to haskell.org instead of www.haskell.org - - - - - b94d4903 by Simon Marlow at 2004-07-01T11:08:58+00:00 [haddock @ 2004-07-01 11:08:57 by simonmar] Update to the +/- buttons: use a resized image rather than a <button>. Still seeing some strange effects in Konqueror, so might need to use a fixed-size image instead. - - - - - d5278f67 by Sven Panne at 2004-07-04T15:15:55+00:00 [haddock @ 2004-07-04 15:15:55 by panne] Install pictures for +/- pictures, too (JPEG is a strange format for graphics like this, I would have expected GIF or PNG here.) Things look fine with Konqueror and Netscape on Linux now, the only downside is that the cursor doesn't change when positioned above the "button". - - - - - 46dec6c5 by Sven Panne at 2004-07-13T17:59:28+00:00 [haddock @ 2004-07-13 17:59:28 by panne] A quote is a valid part of a Haskell identifier, but it would interfere with an ECMA script string delimiter, so escape it there. - - - - - 1d7bc432 by Simon Marlow at 2004-07-22T08:54:06+00:00 [haddock @ 2004-07-22 08:54:06 by simonmar] Add single quote to $ident, so you can say eg. 'foldl'' to refer to foldl' (the longest match rule is our friend). Bug reported by Adrian Hey <ahey at iee.org> - - - - - f183618b by Krasimir Angelov at 2004-07-27T22:59:35+00:00 [haddock @ 2004-07-27 22:58:23 by krasimir] Add basic support for Microsoft HTML Help 2.0 - - - - - d515d0c2 by Krasimir Angelov at 2004-07-27T23:02:36+00:00 [haddock @ 2004-07-27 23:02:36 by krasimir] escape names in the index - - - - - a5f1be23 by Krasimir Angelov at 2004-07-27T23:05:21+00:00 [haddock @ 2004-07-27 23:05:21 by krasimir] Add jsFile, plusFile and minusFile to the file list - - - - - c4fb4881 by Krasimir Angelov at 2004-07-28T22:12:10+00:00 [haddock @ 2004-07-28 22:12:09 by krasimir] bugfix. Move contentsHtmlFile, indexHtmlFile and subIndexHtmlFile functions to HaddockUtil.hs module to make them accessible from HaddockHH2.hs - - - - - 64d30b1d by Krasimir Angelov at 2004-07-30T22:15:47+00:00 [haddock @ 2004-07-30 22:15:45 by krasimir] more stuffs - support for separated compilation of packages - the contents page now uses DHTML TreeView - fixed copyFile bug - - - - - 133c8c5c by Krasimir Angelov at 2004-07-31T12:04:38+00:00 [haddock @ 2004-07-31 12:04:37 by krasimir] make the DHtmlTree in contents page more portable. The +/- buttons are replaced with new images which looks more beatiful. - - - - - 79040963 by Krasimir Angelov at 2004-07-31T13:10:20+00:00 [haddock @ 2004-07-31 13:10:20 by krasimir] Make DHtmlTree compatible with Mozila browser - - - - - 1a55dc90 by Krasimir Angelov at 2004-07-31T14:52:55+00:00 [haddock @ 2004-07-31 14:52:55 by krasimir] fix - - - - - 85ce0237 by Krasimir Angelov at 2004-07-31T14:53:28+00:00 [haddock @ 2004-07-31 14:53:28 by krasimir] HtmlHelp 1.x - - - - - 3c0c53ba by Krasimir Angelov at 2004-07-31T20:35:21+00:00 [haddock @ 2004-07-31 20:35:21 by krasimir] Added support for DevHelp - - - - - d42b5af1 by Krasimir Angelov at 2004-07-31T21:17:51+00:00 [haddock @ 2004-07-31 21:17:51 by krasimir] Document new features in HtmlHelp - - - - - 790fe21e by Krasimir Angelov at 2004-08-01T15:14:02+00:00 [haddock @ 2004-08-01 15:14:02 by krasimir] add missing imports - - - - - fd7cc6bc by Krasimir Angelov at 2004-08-01T19:52:08+00:00 [haddock @ 2004-08-01 19:52:06 by krasimir] fix some bugs. Now I have got the entire libraries documentation in HtmlHelp 2.0 format. - - - - - 94ad7ac8 by Krasimir Angelov at 2004-08-01T19:53:50+00:00 [haddock @ 2004-08-01 19:53:50 by krasimir] I forgot to add the new +/- images - - - - - f0c65388 by Krasimir Angelov at 2004-08-02T16:25:53+00:00 [haddock @ 2004-08-02 16:25:53 by krasimir] Add root node to the table of contents. All modules in tree are not children of the root - - - - - f50bd85d by Sven Panne at 2004-08-02T18:17:46+00:00 [haddock @ 2004-08-02 18:17:46 by panne] Mainly DocBook fixes - - - - - 09527ce3 by Sven Panne at 2004-08-02T20:02:29+00:00 [haddock @ 2004-08-02 20:02:29 by panne] Fixed -o/--odir handling. Generating the output, especially the directory handling, is getting a bit convoluted nowadays... - - - - - c8fbacfa by Sven Panne at 2004-08-02T20:31:13+00:00 [haddock @ 2004-08-02 20:31:13 by panne] Warning police - - - - - 37830bff by Sven Panne at 2004-08-02T20:32:29+00:00 [haddock @ 2004-08-02 20:32:28 by panne] Nuked dead code - - - - - 13847171 by Sven Panne at 2004-08-02T21:12:27+00:00 [haddock @ 2004-08-02 21:12:25 by panne] Use pathJoin instead of low-level list-based manipulation for FilePaths - - - - - c711d61e by Sven Panne at 2004-08-02T21:16:02+00:00 [haddock @ 2004-08-02 21:16:02 by panne] Removed WinDoze CRs - - - - - b1f7dc88 by Sven Panne at 2004-08-03T19:35:59+00:00 [haddock @ 2004-08-03 19:35:59 by panne] Fixed spelling of "http-equiv" attribute - - - - - dd5f394e by Sven Panne at 2004-08-03T19:44:03+00:00 [haddock @ 2004-08-03 19:44:03 by panne] Pacify W3C validator: * Added document encoding (currently UTF-8, not sure if this is completely correct) * Fixed syntax of `id' attributes * Added necessary `alt' attribute for +/- images Small layout improvement: * Added space after +/- images (still not perfect, but better than before) - - - - - 919c47c6 by Sigbjorn Finne at 2004-08-03T19:45:11+00:00 [haddock @ 2004-08-03 19:45:11 by sof] make it compile with <= ghc-6.1 - - - - - 4d6f01d8 by Sigbjorn Finne at 2004-08-03T19:45:30+00:00 [haddock @ 2004-08-03 19:45:30 by sof] ffi wibble - - - - - 4770643a by Sven Panne at 2004-08-03T20:47:46+00:00 [haddock @ 2004-08-03 20:47:46 by panne] Fixed CSS for button style. Note that only "0" is a valid measure without a unit! - - - - - 14aaf2e5 by Sven Panne at 2004-08-03T21:07:59+00:00 [haddock @ 2004-08-03 21:07:58 by panne] Improved spacing of dynamic module tree - - - - - 97c3579a by Simon Marlow at 2004-08-09T11:03:04+00:00 [haddock @ 2004-08-09 11:03:04 by simonmar] Add FormatVersion Patch submitted by: George Russell <ger at informatik.uni-bremen.de> - - - - - af7f8c03 by Simon Marlow at 2004-08-09T11:55:07+00:00 [haddock @ 2004-08-09 11:55:05 by simonmar] Add support for a short description for each module, which is included in the contents. The short description should be given in a "Description: " field of the header. Included in this patch are changes that make the format of the header a little more flexible. From the comments: -- all fields in the header are optional and have the form -- -- [spaces1][field name][spaces] ":" -- [text]"\n" ([spaces2][space][text]"\n" | [spaces]"\n")* -- where each [spaces2] should have [spaces1] as a prefix. -- -- Thus for the key "Description", -- -- > Description : this is a -- > rather long -- > -- > description -- > -- > The module comment starts here -- -- the value will be "this is a .. description" and the rest will begin -- at "The module comment". The header fields must be in the following order: Module, Description, Copyright, License, Maintainer, Stability, Portability. Patches submitted by: George Russell <ger at informatik.uni-bremen.de>, with a few small changes be me, mostly to merge with other recent changes. ToDo: document the module header. - - - - - 7b865ad3 by Simon Marlow at 2004-08-10T14:09:57+00:00 [haddock @ 2004-08-10 14:09:57 by simonmar] Fixes for DevHelp/HtmlHelp following introduction of short module description. - - - - - 814766cd by Simon Marlow at 2004-08-10T14:33:46+00:00 [haddock @ 2004-08-10 14:33:45 by simonmar] Fixes to installation under Windows. - - - - - 39cf9ede by Simon Marlow at 2004-08-12T12:08:23+00:00 [haddock @ 2004-08-12 12:08:23 by simonmar] Avoid using string-gap tricks. - - - - - b6d78551 by Simon Marlow at 2004-08-13T10:53:21+00:00 [haddock @ 2004-08-13 10:53:21 by simonmar] Update - - - - - eaae7417 by Simon Marlow at 2004-08-13T10:53:50+00:00 [haddock @ 2004-08-13 10:53:50 by simonmar] Test for primes in quoted links - - - - - 68c34f06 by Sven Panne at 2004-08-16T19:59:38+00:00 [haddock @ 2004-08-16 19:59:36 by panne] XMLification - - - - - 7f45a6f9 by Sven Panne at 2004-08-18T16:42:54+00:00 [haddock @ 2004-08-18 16:42:54 by panne] Re-added indices + minor fixes - - - - - 8a5dd97c by Sigbjorn Finne at 2004-08-25T17:15:42+00:00 [haddock @ 2004-08-25 17:15:42 by sof] backquote HADDOCK_VERSION defn for <= ghc-6.0.x; believe this is only needed under mingw - - - - - 4b1b42ea by Sven Panne at 2004-08-26T20:08:50+00:00 [haddock @ 2004-08-26 20:08:49 by panne] SGML is dead, long live DocBook XML! Note: The BuildRequires tags in the spec files are still incomplete and the documentation about the DocBook tools needs to be updated, too. Stay tuned... - - - - - 8d52cedb by Sven Panne at 2004-08-26T21:03:19+00:00 [haddock @ 2004-08-26 21:03:19 by panne] Updated BuildRequires tags. Alas, there seems to be no real standard here, so your mileage may vary... At least the current specs should work on SuSE Linux. - - - - - e6982912 by Sigbjorn Finne at 2004-08-30T15:44:59+00:00 [haddock @ 2004-08-30 15:44:59 by sof] escape HADDOCK_VERSION double quotes on all platforms when compiling with <=6.0.x - - - - - b3fbc867 by Simon Marlow at 2004-08-31T13:09:42+00:00 [haddock @ 2004-08-31 13:09:42 by simonmar] Avoid GHC/shell versionitis and create Version.hs - - - - - c359e16a by Sven Panne at 2004-09-05T19:12:33+00:00 [haddock @ 2004-09-05 19:12:32 by panne] * HTML documentation for "foo.xml" goes into directory "foo" again, not "foo-html". This is nicer and consistent with the behaviour for building the docs from SGML. * Disabled building PostScript documentation in the spec files for now, there are some strange issues with the FO->PS conversion for some files which have to be clarified first. - - - - - c68b1eba by Sven Panne at 2004-09-24T07:04:38+00:00 [haddock @ 2004-09-24 07:04:38 by panne] Switched the default state for instances and the module hierarchy to non-collapsed. This can be reversed when we finally use cookies from JavaScript to have a more persistent state. Previously going back and forth in the documentation was simply too annoying because everything was collapsed again and therefore the documentation was not easily navigatable. - - - - - dfb32615 by Simon Marlow at 2004-09-30T08:21:29+00:00 [haddock @ 2004-09-30 08:21:29 by simonmar] Add a feature request - - - - - 45ff783c by Sven Panne at 2004-10-23T19:54:00+00:00 [haddock @ 2004-10-23 19:54:00 by panne] Improved the Cygwin/MinGW chaos a little bit. There is still confusion about host platform vs. target platform... - - - - - 5f644714 by Krasimir Angelov at 2004-10-28T16:01:51+00:00 [haddock @ 2004-10-28 16:01:51 by krasimir] update for ghc-6.3+ - - - - - 92d9753e by Sven Panne at 2004-11-01T16:39:01+00:00 [haddock @ 2004-11-01 16:39:01 by panne] Revert previous commit: It's Network.URI which should be changed, not Haddock. - - - - - 05f70f6e by Simon Marlow at 2005-01-04T16:15:51+00:00 [haddock @ 2005-01-04 16:15:51 by simonmar] parser fix: allow qualified specialids. - - - - - 47870837 by Simon Marlow at 2005-01-04T16:16:54+00:00 [haddock @ 2005-01-04 16:16:54 by simonmar] Add a test - - - - - ff11fc2c by Ross Paterson at 2005-01-10T19:18:22+00:00 [haddock @ 2005-01-10 19:18:22 by ross] Render non-ASCII characters using numeric character references, to simplify charset issues. There's a META tag saying the charset is UTF-8, but GHC outputs characters as raw bytes. Ideally we need an encoding on the input side too, primarily in comments, because source files containing non-ASCII characters aren't portable between locales. - - - - - eba2fc4e by Simon Marlow at 2005-01-11T10:44:37+00:00 [haddock @ 2005-01-11 10:44:37 by simonmar] Remove string gap - - - - - b899a381 by Ross Paterson at 2005-01-13T11:41:33+00:00 [haddock @ 2005-01-13 11:41:33 by ross] recognize SGML-style numeric character references &#ddd; or &#xhhhh; and translate them into Chars. - - - - - 106e3cf0 by Ross Paterson at 2005-01-13T14:43:41+00:00 [haddock @ 2005-01-13 14:43:41 by ross] also allow uppercase X in hexadecimal character references (like SGML) - - - - - e8f54f25 by Ross Paterson at 2005-01-13T14:44:24+00:00 [haddock @ 2005-01-13 14:44:24 by ross] Describe numeric character references. - - - - - 914ccdce by Sven Panne at 2005-01-15T18:44:48+00:00 [haddock @ 2005-01-15 18:44:45 by panne] Make Haddock compile again after the recent base package changed. The Map/Set legacy hell has been factored out, so that all modules can simply use the new non-deprecated interfaces. Probably a lot of things can be improved by a little bit of Map/Set/List algebra, this can be done later if needed. Small note: Currently the list of instances in HTML code is reversed. This will hopefully be fixed later. - - - - - 6ab20e84 by Sven Panne at 2005-01-16T12:18:26+00:00 [haddock @ 2005-01-16 12:18:26 by panne] Trim imports - - - - - efb81da9 by Sven Panne at 2005-01-16T12:58:08+00:00 [haddock @ 2005-01-16 12:58:03 by panne] Correctly handle the new order of arguments for the combining function given to fromListWith. - - - - - e27b5834 by Sven Panne at 2005-01-16T14:14:41+00:00 [haddock @ 2005-01-16 14:14:39 by panne] Data.Map.unions is left-biased. - - - - - dae3cc3e by Sven Panne at 2005-01-16T14:22:44+00:00 [haddock @ 2005-01-16 14:22:44 by panne] Added the last missing "flip" to get identical HTML output as previous versions. - - - - - 951d8408 by Sven Panne at 2005-01-16T14:37:10+00:00 [haddock @ 2005-01-16 14:37:10 by panne] Refactored Text.PrettyPrint legacy hell into a separate module. - - - - - f1c4b892 by Sven Panne at 2005-01-16T15:41:25+00:00 [haddock @ 2005-01-16 15:41:21 by panne] Cleaned up imports and dropped support for GHC < 5.03, it never worked, anyway. - - - - - 60824c6e by Simon Marlow at 2005-01-18T10:02:48+00:00 [haddock @ 2005-01-18 10:02:48 by simonmar] Add a TODO - - - - - a8c82f23 by Krasimir Angelov at 2005-01-28T23:19:39+00:00 [haddock @ 2005-01-28 23:19:39 by krasimir] import Foreign/Foreign.C are required for Windows - - - - - d8450a23 by Simon Marlow at 2005-02-02T16:23:04+00:00 [haddock @ 2005-02-02 16:23:00 by simonmar] Revamp the linking strategy in Haddock. Now name resolution is done in two phases: - first resolve everything to original names, like a Haskell compiler would. - then, figure out the "home" location for every entity, and point all the links to there. The home location is the lowest non-hidden module in the import hierarchy that documents the entity. If there are multiple candidates, one is chosen at random. Also: - Haddock should not generate any HTML with dangling links any more. Unlinked references are just rendered as plain text. - Error reporting is better: if we can't find a link destination for an entity reference, we now emit a warning. - - - - - 1cce71d0 by Simon Marlow at 2005-02-03T13:42:19+00:00 [haddock @ 2005-02-03 13:42:19 by simonmar] - add --ignore-all-exports flag, which behaves as if every module has the ignore-exports attribute (requested by Chris Ryder). - add --hide option to hide a module on the command line. - add --use-package option to get Haddock info for a package from ghc-pkg (largely untested). - remove reexports from the .haddock file, they aren't used any more. - - - - - 767123ef by Ross Paterson at 2005-02-03T16:17:37+00:00 [haddock @ 2005-02-03 16:17:37 by ross] fix typo for < 6.3 - - - - - 0c680c04 by Simon Marlow at 2005-02-04T12:03:31+00:00 [haddock @ 2005-02-04 12:03:31 by simonmar] Fix bug in renameExportItems that meant links in instances weren't being renamed properly. - - - - - ff7abe5f by Simon Marlow at 2005-02-04T12:15:53+00:00 [haddock @ 2005-02-04 12:15:52 by simonmar] Add attribute #not-home, to indicate that the current module should not be considered to be a home module for the each entity it exports, unless there is no other module that exports the entity. - - - - - fc2cfd27 by Simon Marlow at 2005-02-04T12:40:02+00:00 [haddock @ 2005-02-04 12:40:02 by simonmar] Update the documentation w.r.t. home modules and the not-home attribute. - - - - - 26b8ddf7 by Ross Paterson at 2005-02-04T13:36:06+00:00 [haddock @ 2005-02-04 13:36:05 by ross] sort lists of instances by - arity of the type constructors (so higher-kinded instances come first) - name of the class - argument types - - - - - 26bfb19c by Simon Marlow at 2005-02-23T15:57:12+00:00 [haddock @ 2005-02-23 15:57:12 by simonmar] Fix documentation regarding the module attributes. - - - - - 9c3afd02 by Simon Marlow at 2005-02-28T16:18:17+00:00 [haddock @ 2005-02-28 16:18:17 by simonmar] version 0.7 - - - - - a95fd63f by Simon Marlow at 2005-02-28T16:22:08+00:00 [haddock @ 2005-02-28 16:22:08 by simonmar] Attempt to fix the layout of the package names in the contents. Having tried just about everything, the only thing I can get to work reliably is to make the package names line up on a fixed offset from the left margin. This obviously isn't ideal, so anyone else that would like to have a go at improving it is welcome. One option is to remove the +/- buttons from the contents list and go back to a plain table. The contents page now uses CSS for layout rather than tables. It seems that most browsers have different interpretations of CSS layout, so only the simplest things lead to consistent results. - - - - - 905d42f7 by Simon Marlow at 2005-03-01T17:16:42+00:00 [haddock @ 2005-03-01 17:16:40 by simonmar] Another attempt at lining up the package names on the contents page. Now, they line up with Konqueror, and almost line up with Firefox & IE (different layout in each case). - - - - - a0e1d178 by Wolfgang Thaller at 2005-03-09T08:28:39+00:00 [haddock @ 2005-03-09 08:28:39 by wolfgang] Hack haddock's lexer to accept the output from Apple's broken version of cpp (Apple's cpp leaves #pragma set_debug_pwd directives in it's output). - - - - - 9e1eb784 by Simon Marlow at 2005-04-22T14:27:15+00:00 [haddock @ 2005-04-22 14:27:15 by simonmar] Add a TODO item - - - - - 23281f78 by Ross Paterson at 2005-05-18T12:41:59+00:00 [haddock @ 2005-05-18 12:41:59 by ross] fix 3 bugs in --use-package, and document it. - - - - - 00074a68 by Sven Panne at 2005-05-21T12:35:29+00:00 [haddock @ 2005-05-21 12:35:29 by panne] Warning/versionitis police - - - - - 341fa822 by Simon Marlow at 2005-06-15T15:43:21+00:00 [haddock @ 2005-06-15 15:43:21 by simonmar] Allow "licence" as an alternate spelling of "license" - - - - - 3b953f8b by Simon Marlow at 2005-06-16T08:14:12+00:00 [haddock @ 2005-06-16 08:14:12 by simonmar] wibble - - - - - abfd9826 by Simon Marlow at 2005-06-27T14:46:40+00:00 [haddock @ 2005-06-27 14:46:40 by simonmar] name hierarchical HTML files as A-B-C.html instead of A.B.C.html. The old way confused Apache because the extensions are sometimes interpreted as having special meanings. - - - - - a01eea00 by Simon Marlow at 2005-08-04T13:59:40+00:00 [haddock @ 2005-08-04 13:59:40 by simonmar] 0.7 changes - - - - - 170ef87e by Simon Marlow at 2005-08-04T15:08:03+00:00 [haddock @ 2005-08-04 15:08:03 by simonmar] spec file from Jens Peterson - - - - - 7621fde4 by Simon Marlow at 2005-08-04T15:59:30+00:00 [haddock @ 2005-08-04 15:59:30 by simonmar] replace mingw tests with $(Windows) - - - - - a20739bb by Sven Panne at 2005-08-05T07:01:12+00:00 [haddock @ 2005-08-05 07:01:12 by panne] Reverted to previous version (but with bumped version number), the last commit broke RPM building on SuSE systems due to differently named dependencies. As a clarification: All .spec files in the repository have to work at least on SuSE, because that's the system I'm using. And as "Mr. Building Police", I reserve me the right to keep them that way... >:-) It might very well be the case that we need different .spec files for different platforms, so packagers which are unhappy with the current .spec files should contact me, stating the actual problems. - - - - - 4afb15cf by Simon Marlow at 2005-10-05T10:51:45+00:00 [haddock @ 2005-10-05 10:51:45 by simonmar] Add a bug - - - - - 60f69f82 by Simon Marlow at 2005-10-05T12:52:03+00:00 [haddock @ 2005-10-05 12:52:03 by simonmar] Document new behaviour of -s option - - - - - f7e520ca by Simon Marlow at 2005-10-10T15:02:55+00:00 [haddock @ 2005-10-10 15:02:55 by simonmar] extractRecSel: ignore non-record constructors (fixes a crash when using datatypes with a mixture of record and non-record style constructors). - - - - - b2edbedb by Simon Marlow at 2005-10-14T09:44:21+00:00 Start CHANGES for 0.8 - - - - - 21c7ac8d by Simon Marlow at 2005-10-14T23:11:19+00:00 First cut of Cabal build system - - - - - 766cecdd by Simon Marlow at 2005-10-29T08:14:43+00:00 Add configure script and Makefile for the docs Add a separate configure script and build system for building the documentation. The configure and Makefile code is stolen from fptools. This is left as a separate build system so that the main Cabal setup doesn't require a Unix build environment or DocBook XML tools. - - - - - aa36c783 by Duncan Coutts at 2006-01-17T19:29:55+00:00 Add a --wiki=URL flag to add a per-module link to a correspondng wiki page. So each html page gets an extra link (placed next to the source code and contents links) to a corresponding wiki page. The idea is to let readers contribute their own notes, examples etc to the documentation. Also slightly tidy up the code for the --source option. - - - - - e06e2da2 by Simon Marlow at 2006-01-18T09:28:15+00:00 TODO: documnet --wiki - - - - - 17adfda9 by Duncan Coutts at 2006-01-19T20:17:59+00:00 Add an optional wiki link for each top level exported name. In each module, for each "top level" exported entity we add a hyper link to a corresponding wiki page. The link url gets the name of the exported entity as a '#'-style anchor, so if there is an anchor in the page with that name then the users browser should jump directly to it. By "top level" we mean functions, classes, class members and data types (data, type, newtype), but not data constructors, class instances or data type class membership. The link is added at the right of the page and in a small font. Hopefully this is the right balance of visibility/distraction. We also include a link to the wiki base url in the contents and index pages. - - - - - f52324bb by Duncan Coutts at 2006-01-19T20:28:27+00:00 Rewrite pathJoin to only add a path separator when necessary. When the path ends in a file seperator there is no need to add another. Now using "--wiki=http://blah.com/foo/" should do the right thing. (Code snippet adapted from Isaac's FilePath package.) - - - - - 43bb89fa by Duncan Coutts at 2006-01-21T17:15:27+00:00 Teach haddock about line pragmas and add accurate source code links Teach haddock about C and Haskell style line pragmas. Extend the lexer/parser's source location tracking to include the file name as well as line/column. This way each AST item that is tagged with a SrcLoc gets the original file name too. Use this original file name to add source links to each exported item, in the same visual style as the wiki links. Note that the per-export source links are to the defining module rather than whichever module haddock pretends it is exported from. This is what we want for source code links. The source code link URL can also contain the name of the export so one could implement jumping to the actual location of the function in the file if it were linked to an html version of the source rather than just plain text. The name can be selected with the %N wild card. So for linking to the raw source code one might use: --source=http://darcs/haskell.org/foo/%F Or for linking to html syntax highlighted code: --source=http://darcs/haskell.org/foo/%M.html#%N - - - - - edd9f229 by Duncan Coutts at 2006-01-22T00:02:00+00:00 Extend URL variable expansion syntax and add source links to the contents page Like the wiki link on the contents and index page, add a source code link too. Extend the wiki & source URL variable expansion syntax. The original syntax was: %F for the source file name (the .hs version only, not the .lhs or .hs.pp one) %M for the module name (with '.' replaced by '/') The new syntax is: %F or %{FILE} for the original source file name %M or %{MODULE} for the module name (no replacements) %N or %{NAME} for the function/type export name %K or %{KIND} for a type/value flag "t" or "v" with these extensions: %{MODULE/./c} to replace the '.' module seperator with any other char c %{VAR|some text with the % char in it} which means if the VAR is not in use in this URL context then "" else replace the given text with the '%' char replaced by the string value of the VAR. This extension allows us to construct URLs wit optional parts, since the module/file name is not available for the URL in the contents/index pages and the value/type name is not available for the URL at the top level of each module. - - - - - eb3c6ada by Duncan Coutts at 2006-01-23T13:42:34+00:00 Remove the complex substitutions and add more command line flags instead. Instead of incomprehensable URL substitutions like ${MODULE/./-|?m=%} we now use three seperate command line flags for the top level, per-module and per-entity source and wiki links. They are: --source-base, --source-module, --source-entity --comments-base, --comments-module, --comments-entity We leave -s, --source as an alias for --source-module which is how that option behaved previously. The long forms of the substitutions are still available, ${FILE} ${MODULE} etc and the only non-trivial substitution is ${MODULE/./c} to replace the '.' characters in the module name with any other character c. eg ${MODULE/./-} Seperating the source and wiki url flags has the added bonus that they can be turned on or off individually. So users can have per-module links for example without having to also have per-entity links.` - - - - - a2f0f2af by Duncan Coutts at 2006-01-23T13:54:52+00:00 Make the --help output fit in 80 columns. This is a purely cosmetic patch, feel free to ignore it. The only trickery going on is that we don't display the deprecated -s, --source flags in the help message, but we do still accept them. - - - - - 2d3a4b0c by Duncan Coutts at 2006-01-23T14:12:16+00:00 Add documentation for the new --source-* and --comments-* command line options - - - - - 1a82a297 by Simon Marlow at 2006-01-23T17:03:27+00:00 fix markup - - - - - 100d464a by Duncan Coutts at 2006-01-23T18:31:13+00:00 remove a couple TODO items that have been done The --wiki, or rather the --comment-* options are now documented. There is probably no need to have haddock invoke unlit or cpp itself since it can now pick up the line pragmas to get the source locations right. Tools like Cabal will arrange for preprocessors to be run so there is less of a need for tools like haddock to do it themselves. - - - - - 3162fa91 by Simon Marlow at 2006-01-24T14:21:56+00:00 add a test I had lying around - - - - - 98947063 by Simon Marlow at 2006-01-31T13:52:54+00:00 add scabal-version field - - - - - c41876e6 by Neil Mitchell at 2006-02-26T17:48:21+00:00 Add Hoogle output option - - - - - f86fb9c0 by Simon Marlow at 2006-03-08T09:15:20+00:00 add haskell.vim Contributed by Brad Bowman <bsb at bereft.net>, thanks! - - - - - 35d3c511 by benjamin.franksen at 2006-03-03T22:39:54+00:00 fixed libdir (/html was missing) - - - - - 4d08fd7d by Simon Marlow at 2006-03-10T11:13:31+00:00 add PatternGuards extension - - - - - 3f095e70 by Simon Marlow at 2006-03-13T11:40:42+00:00 bug fixes from Brad Bowman - - - - - 8610849d by Sven Panne at 2006-03-19T17:02:56+00:00 Fixed Cabal/RPM build - - - - - 34a994d6 by sven.panne at 2006-04-20T12:39:23+00:00 Avoid pattern guards Due to the use of pattern guards in Haddock, GHC was called with -fglasgow-exts. This in turn enables bang patterns, too, which broke the Haddock build. Removing some unnecessary pattern guards seemed to be the better way of fixing this instead of using a pragma to disable pattern guards. - - - - - bb523f51 by Ross Paterson at 2006-04-24T09:03:25+00:00 extend 'deriving' heuristic a little If an argument of a data constructor has a type variable head, it is irreducible and the same type class can be copied into the constraint. (Formerly we just did this for type variable arguments.) - - - - - dab9fe7a by Simon Marlow at 2006-04-26T10:02:31+00:00 record an idea - - - - - 748b7078 by Simon Marlow at 2006-05-08T08:28:53+00:00 add section about deriving - - - - - 11252ea1 by Simon Marlow at 2006-05-24T15:43:10+00:00 replace a fatal error in lexChar with a parseError - - - - - 382c9411 by Simon Marlow at 2006-05-24T15:45:47+00:00 add a bug - - - - - b79272f5 by Simon Marlow at 2006-05-24T15:46:29+00:00 add a bug report - - - - - 912edf65 by David Waern at 2006-07-10T19:09:23+00:00 Initial modifications -- doesn't compile - - - - - a3c7ba99 by David Waern at 2006-07-11T00:54:19+00:00 More porting work -- doesn't compile - - - - - 0a173d19 by David Waern at 2006-07-11T11:30:03+00:00 Make the repos temporarily compile and illustrate a problem - - - - - bad316de by David Waern at 2006-07-11T15:43:47+00:00 Progress on the porting process - - - - - bbf12d02 by David Waern at 2006-07-11T23:07:44+00:00 More progress on the porting -- first pass starting to shape up - - - - - de580ba2 by David Waern at 2006-07-20T17:48:30+00:00 More progress -- still on phase1 - - - - - 75a917a2 by David Waern at 2006-07-23T18:22:43+00:00 More work on pass1 -- mostly done - - - - - 6697b3f7 by David Waern at 2006-07-23T22:17:40+00:00 More work, started working on the renaming phase -- this code will need a cleanup soon :) - - - - - 82a5bcbb by David Waern at 2006-07-29T16:16:43+00:00 Add instances, build renaming environment, start on the renamer - - - - - c3f8f4f1 by David Waern at 2006-07-29T21:37:48+00:00 Complete the renamer - - - - - 7e00d464 by David Waern at 2006-07-30T21:01:57+00:00 Start porting the Html renderer - - - - - f04ce121 by David Waern at 2006-08-09T20:04:56+00:00 More Html rendering progress - - - - - 20c21b53 by David Waern at 2006-08-10T17:37:47+00:00 More progress - - - - - d7097e0d by David Waern at 2006-08-11T20:31:51+00:00 Cleanup - - - - - a7351e86 by David Waern at 2006-08-12T11:44:47+00:00 Render H98 Data declarations - - - - - 3fb2208e by David Waern at 2006-08-12T17:15:34+00:00 Perfect rendering of Test.hs - - - - - 454fd062 by David Waern at 2006-08-13T21:57:08+00:00 Misc fixes and interface load/save - - - - - 7ef7e7be by David Waern at 2006-08-14T00:56:07+00:00 Some refactoring - - - - - a7d3efef by David Waern at 2006-08-19T20:07:55+00:00 Adapt to latest GHC - - - - - 5fc3c0d7 by David Waern at 2006-08-20T21:28:11+00:00 Move interface read/write to its own module + some cleanup - - - - - 037e011c by David Waern at 2006-08-20T21:38:24+00:00 Small cleanup - - - - - da3a1023 by David Waern at 2006-09-03T16:05:22+00:00 Change mode to BatchCompile to avoid GHC API bug - - - - - 3cc9be3b by David Waern at 2006-09-03T16:06:59+00:00 Starting work on GADT rendering - - - - - 94506037 by David Waern at 2006-09-03T20:02:48+00:00 Compensate for change of export list order in GHC - - - - - c2cec4eb by David Waern at 2006-09-04T20:53:01+00:00 Rename a function - - - - - 9a9735ba by David Waern at 2006-09-05T15:51:21+00:00 Change version number to 2.0 - - - - - 3758a714 by David Waern at 2006-09-05T15:51:49+00:00 Align comment properly - - - - - 68478d9e by David Waern at 2006-09-15T18:03:00+00:00 Remove interface reading/writing code and use the GHC api for creating package environments instead - - - - - d2eedd95 by David Waern at 2006-09-15T18:05:29+00:00 Change the executable name to haddock-ghc-nolib - - - - - fcfbcf66 by David Waern at 2006-09-15T18:05:45+00:00 Small source code cleanup - - - - - d08eb017 by David Waern at 2006-09-15T18:06:21+00:00 Remove handling of --package flag - - - - - b8a4cf53 by David Waern at 2006-09-15T18:07:16+00:00 Remove commented-out code - - - - - bef0a684 by David Waern at 2006-09-15T18:37:57+00:00 Don't warn about missing links to () - - - - - e7d25fd7 by David Waern at 2006-09-15T19:50:49+00:00 Remove Interface and Binary2 modules - - - - - 9894f2a1 by David Waern at 2006-09-15T19:53:43+00:00 Remove debug printing from HaddockHtml - - - - - a0e7455d by David Waern at 2006-09-16T00:16:29+00:00 Comments only - - - - - d5b26fa7 by David Waern at 2006-09-16T00:16:57+00:00 Refactor PackageData creation code and start on building the doc env propery (unfinished) - - - - - 06aaa779 by David Waern at 2006-09-16T00:19:25+00:00 Better comments in Main.hs - - - - - 1a52d1b4 by David Waern at 2006-09-18T22:17:11+00:00 Comments and spacing change - - - - - e5a97767 by David Waern at 2006-09-21T17:02:45+00:00 Remove unnecessary fmapM import in Main - - - - - 9d0f9d3a by David Waern at 2006-09-22T18:07:07+00:00 Make import list in HaddockHtml prettier - - - - - 3452f662 by David Waern at 2006-09-22T18:08:47+00:00 Refactor context rendering - - - - - 12d0a6d0 by David Waern at 2006-09-22T18:09:52+00:00 Do proper HsType rendering (inser parentheses correctly) - - - - - 2c20c2f9 by David Waern at 2006-09-22T18:10:45+00:00 Fix a bug in Main.toHsType - - - - - c5396443 by David Waern at 2006-09-22T18:11:16+00:00 Skip external package modules sort for now - - - - - 3fb95547 by David Waern at 2006-09-22T20:35:40+00:00 Take away trailin "2" on all previously clashing type names - - - - - 2174755f by David Waern at 2006-09-22T20:51:43+00:00 Remove unused imports in Main - - - - - 1e9f7a39 by David Waern at 2006-09-22T20:52:11+00:00 Fix a comment in Main - - - - - 32d9e028 by David Waern at 2006-10-05T16:40:11+00:00 Merge with changes to ghc HEAD - - - - - 3058c8f5 by David Waern at 2006-10-05T16:41:02+00:00 Comment fixes - - - - - b9c217ec by David Waern at 2006-10-05T16:49:59+00:00 Filter out more builtin type constructors from warning messages - - - - - 67e7d252 by David Waern at 2006-10-05T19:38:22+00:00 Refactoring -- better structured pass1 - - - - - cd21c0c1 by David Waern at 2006-10-05T19:44:42+00:00 Remove read/dump interface flags - - - - - 313f9e69 by David Waern at 2006-10-05T19:49:26+00:00 Remove unused pretty printing - - - - - 480f09d1 by David Waern at 2006-12-28T13:22:24+00:00 Update to build with latest GHC HEAD - - - - - 63dccfcb by David Waern at 2007-01-05T01:38:45+00:00 Fixed a bug so that --ghc-flag works correctly - - - - - 3117dadc by David Waern at 2006-12-29T18:53:39+00:00 Automatically get the GHC lib dir - - - - - 9dc84a5c by David Waern at 2006-12-29T19:58:53+00:00 Comments - - - - - 0b0237cc by David Waern at 2007-01-05T16:48:30+00:00 Collect docs based on SrcLoc, syncing with removal of DeclEntity from GHC - - - - - a962c256 by David Waern at 2007-01-05T17:02:47+00:00 Add tabs in haddock.cabal - - - - - 0ca30c97 by David Waern at 2007-01-05T17:04:11+00:00 Add GHCUtils.hs - - - - - c0ab9abe by David Waern at 2007-01-10T11:43:08+00:00 Change package name to haddock-ghc, version 0.1 - - - - - 38e18b27 by David Waern at 2007-01-12T12:03:52+00:00 No binder name for foreign exports - - - - - d18587ab by David Waern at 2007-01-12T12:08:15+00:00 Temp record - - - - - ba6251a0 by David Waern at 2007-01-12T18:27:55+00:00 Remove read/dump-interface (again) - - - - - f4ba2b39 by David Waern at 2007-01-12T18:31:36+00:00 Remove DocOption, use the GHC type - - - - - 511be8bd by David Waern at 2007-01-12T18:32:41+00:00 Use exceptions instead of Either when loading package info - - - - - 0f2144d8 by David Waern at 2007-01-12T18:33:23+00:00 Small type change - - - - - 77507eb7 by David Waern at 2007-01-12T18:33:59+00:00 Remove interface file read/write - - - - - 0ea1e14f by David Waern at 2007-01-17T21:40:26+00:00 Add trace_ppr to GHCUtils - - - - - 3878b493 by David Waern at 2007-01-17T21:40:53+00:00 Sort external package modules and build a doc env - - - - - 8dc323fc by David Waern at 2007-01-17T21:42:41+00:00 Remove comment - - - - - f4c5b097 by David Waern at 2007-01-18T23:22:18+00:00 Add haddock-ghc.cabal and remove ghc option pragma in source file - - - - - da242b2c by David Waern at 2007-01-18T23:22:46+00:00 Remove some tabs - - - - - 288ed096 by David Waern at 2007-01-18T23:39:28+00:00 Moved the defaultErrorHandler to scope only over sortAndCheckModules for now - - - - - 4dd150fe by David Waern at 2007-02-03T21:23:56+00:00 Let restrictCons handle infix constructors - - - - - 97893442 by David Waern at 2007-02-04T16:26:00+00:00 Render infix data constructors - - - - - da89db72 by David Waern at 2007-02-04T16:26:33+00:00 CHange project name to Haddock-GHC - - - - - e93d48af by David Waern at 2007-02-04T16:59:08+00:00 Render infix type constructors properly - - - - - 357bc99b by David Waern at 2007-02-04T17:37:08+00:00 Insert spaces around infix function names - - - - - ab6cfc49 by David Waern at 2007-02-04T17:59:54+00:00 Do not list entities without documentation - - - - - 04249c7e by David Waern at 2007-02-04T19:16:25+00:00 Add GADT support (quite untested) - - - - - 2c223f8d by David Waern at 2007-02-04T19:25:10+00:00 Add package file write/save again! - - - - - b07ed218 by David Waern at 2007-02-04T19:33:02+00:00 Comment out minf_iface based stuff - - - - - 953d1fa7 by David Waern at 2007-02-05T00:12:23+00:00 Solve conflicts - - - - - 593247fc by David Waern at 2007-02-06T19:48:48+00:00 Remove -package flag, GHC's can be used instead - - - - - f658ded2 by David Waern at 2007-02-06T20:50:44+00:00 Start for support of ATs - - - - - 97f9e913 by David Waern at 2007-02-06T20:52:27+00:00 Wibble - - - - - 2ce8e4cf by David Waern at 2007-02-16T12:09:49+00:00 Add the DocOptions change - - - - - dee4a9b5 by David Waern at 2007-03-06T01:24:48+00:00 Wibble - - - - - 7cb99d18 by David Waern at 2007-03-06T01:24:58+00:00 Change version to 2.0 and executable name to haddock - - - - - c5aa02bc by David Waern at 2007-03-08T15:59:49+00:00 Go back to -B flag - - - - - 3a349201 by David Waern at 2007-03-09T13:31:59+00:00 Better exception handling and parsing of GHC flags - - - - - 05a69b71 by David Waern at 2007-03-09T17:45:44+00:00 Remove commented-out DocEntity printing - - - - - 755032cb by davve at dtek.chalmers.se at 2007-03-23T23:30:20+00:00 Remove a file that shouldn't be here - - - - - a7077e5f by davve at dtek.chalmers.se at 2007-03-24T03:58:48+00:00 Remove an import - - - - - 6f55aa8b by davve at dtek.chalmers.se at 2007-03-25T00:46:48+00:00 Start work on Haddock API - - - - - f0199480 by davve at dtek.chalmers.se at 2007-03-25T00:56:36+00:00 Prettify some comments - - - - - f952f9d1 by davve at dtek.chalmers.se at 2007-03-25T00:56:53+00:00 Remove ppr in HaddockTypes - - - - - bc594904 by davve at dtek.chalmers.se at 2007-03-25T00:57:53+00:00 Remove commented out doc env inference - - - - - 11ebf08d by davve at dtek.chalmers.se at 2007-03-25T01:23:25+00:00 De-flatten the namespace - - - - - f696b4bc by davve at dtek.chalmers.se at 2007-03-25T03:21:48+00:00 Add missing stuff to API - - - - - 9a2a04c3 by davve at dtek.chalmers.se at 2007-03-25T03:22:02+00:00 Wibble - - - - - 7d04a6d5 by davve at dtek.chalmers.se at 2007-03-25T03:22:08+00:00 Avoid a GHC bug with parseStaticFlags [] - - - - - 4d2820ba by davve at dtek.chalmers.se at 2007-03-26T04:57:01+00:00 Add fall-through case to mkExportItem - - - - - 6ebc8950 by Stefan O'Rear at 2007-03-26T04:14:53+00:00 Add shebang line to Setup.lhs - - - - - 80966ec5 by davve at dtek.chalmers.se at 2007-03-26T05:24:26+00:00 Fix stupid compile error - - - - - 1ea1385d by davve at dtek.chalmers.se at 2007-04-05T17:19:56+00:00 Do save/read of interface files properly - - - - - 0e4f6541 by David Waern at 2007-04-10T21:08:36+00:00 Add version to ghc dependency - - - - - b0499b63 by David Waern at 2007-04-10T21:37:08+00:00 Change package name to haddock - - - - - 9d50d27e by David Waern at 2007-04-24T00:22:14+00:00 Use filepath package instead of FilePath - - - - - 87c7fcdf by David Waern at 2007-07-10T21:03:04+00:00 Add new package dependencies - - - - - 4768709c by David Waern at 2007-07-11T20:37:11+00:00 Follow changes to record constructor representation - - - - - b9a02fee by Simon Marlow at 2007-05-30T14:00:48+00:00 update to compile with the latest GHC & Cabal - - - - - c0ebdc01 by David Waern at 2007-07-11T21:35:45+00:00 Fix conflicts - - - - - 97f7afd4 by David Waern at 2007-07-11T21:52:38+00:00 Follow changes to the GHC API - - - - - a5b7b58f by David Waern at 2007-07-12T20:36:48+00:00 Call parseStaticFlags before newSession - - - - - f7f50dbc by David Waern at 2007-08-01T21:52:58+00:00 Better indentation in haddock.cabal - - - - - d84e52ad by David Waern at 2007-08-02T00:08:18+00:00 Wibble - - - - - a23f494a by David Waern at 2007-08-02T00:08:24+00:00 Be better at trying to load all module dependencies (debugging) - - - - - ee917f13 by David Waern at 2007-08-03T18:48:08+00:00 Load all targets explicitly (checkModule doesn't chase dependencies anymore) - - - - - 5182d631 by David Waern at 2007-08-16T16:48:55+00:00 Finalize support for links to other packages - - - - - dfd1e3da by David Waern at 2007-08-16T16:51:11+00:00 Fix haddock comment errors in Haddock.Types - - - - - 50c0d83e by David Waern at 2007-08-16T16:51:37+00:00 Remove a debug import - - - - - d84b7c2b by David Waern at 2007-08-16T17:06:30+00:00 Rename PackageData to HaddockPackage - - - - - 3b52cb9f by David Waern at 2007-08-16T17:09:42+00:00 Simplify some comments - - - - - 66fa68d9 by David Waern at 2007-08-16T17:11:38+00:00 Comment the HaddockPackage definition - - - - - 8674c761 by David Waern at 2007-08-16T17:25:54+00:00 Improve code layout in Main - - - - - 571a3a0b by David Waern at 2007-08-16T17:32:13+00:00 Remove explict module imports in Main - - - - - d31b3cb0 by David Waern at 2007-08-16T17:36:23+00:00 Correct comments - - - - - 7f8a9f2b by David Waern at 2007-08-16T17:39:50+00:00 Fix layout problems in Haddock.Types - - - - - 9f421d7f by David Waern at 2007-08-17T11:16:48+00:00 Move options out of Main into Haddock.Options - - - - - 80042b63 by David Waern at 2007-08-17T11:26:59+00:00 Small comment/layout fixes - - - - - b141b982 by David Waern at 2007-08-17T11:28:28+00:00 Change project name from Haddock-GHC to Haddock - - - - - dbeb4a81 by David Waern at 2007-08-17T11:41:05+00:00 Add top module comment to all files - - - - - ce99cc9e by David Waern at 2007-08-17T14:53:04+00:00 Factor out typechecking phase into Haddock.Typecheck - - - - - 6bf75d9e by David Waern at 2007-08-17T16:55:35+00:00 Factor out package code to Haddock.Packages - - - - - b396db37 by David Waern at 2007-08-29T22:40:23+00:00 Major refactoring - - - - - 3d4f95ee by David Waern at 2007-08-29T23:26:24+00:00 Rename HaddockModule to Interface and a few more refactorings - - - - - c55326db by David Waern at 2007-08-29T23:48:03+00:00 Some comment cleanup - - - - - 9a84fc46 by David Waern at 2007-08-29T23:49:29+00:00 Add some modules that I forgot to add earlier - - - - - 4536dce2 by David Waern at 2007-08-29T23:55:24+00:00 Wibble - - - - - 9b7f0206 by David Waern at 2007-08-30T16:03:29+00:00 Wibble - - - - - c52c050a by David Waern at 2007-08-30T16:30:37+00:00 Rename HaddockModule to Interface - - - - - eae2995f by David Waern at 2007-08-30T16:42:59+00:00 Simplify createInterfaces - - - - - 53f99caa by David Waern at 2007-09-29T00:04:31+00:00 Add build-type: Simple to the cabal file - - - - - 0d3103a8 by David Waern at 2007-09-29T00:04:58+00:00 Add containers and array dependency - - - - - 6acf5f30 by David Waern at 2007-09-29T00:13:36+00:00 Prettify the cabal file - - - - - 87c1e378 by David Waern at 2007-09-29T13:16:39+00:00 FIX: consym data headers with more than two variables - - - - - b67fc16a by David Waern at 2007-09-29T14:01:32+00:00 FIX: prefix types used as operators should be quoted - - - - - a8f925bc by David Waern at 2007-09-29T14:02:26+00:00 Use isSymOcc from OccName instead of isConSym - - - - - fc330701 by David Waern at 2007-09-29T14:15:37+00:00 Use isLexConSym/isLexVarSym from OccName - - - - - e4f3dbad by David Waern at 2007-09-29T15:01:08+00:00 FIX: do not quote varsym type operators - - - - - 402207d2 by David Waern at 2007-09-29T15:01:50+00:00 Wibble - - - - - f9d89ef0 by David Waern at 2007-09-29T15:17:40+00:00 Take care when pp tyvars - add parens on syms - - - - - 849e2a77 by David Waern at 2007-10-01T21:56:39+00:00 Go back to using a ModuleMap instead of LookupMod - fixes a bug - - - - - 549dbac6 by David Waern at 2007-10-02T01:05:19+00:00 Improve parsing of doc options - - - - - a36021b8 by David Waern at 2007-10-02T23:05:00+00:00 FIX: double arrows in constructor contexts - - - - - d03bf347 by David Waern at 2007-10-09T16:14:05+00:00 Add a simple test suite - - - - - c252c140 by David Waern at 2007-10-17T16:02:28+00:00 Add --optghc=.. style flag passing to GHC - - - - - cce6c1b3 by David Waern at 2007-10-18T22:03:20+00:00 Add support for --read-interface again - - - - - 33d059c0 by David Waern at 2007-10-18T22:30:18+00:00 Refactoring -- get rid of Haddock.Packages - - - - - f9ed0a4c by David Waern at 2007-10-18T22:34:36+00:00 Name changes - - - - - 8a1c816f by David Waern at 2007-10-20T14:24:23+00:00 Add --ghc-version option - - - - - 4925aaa1 by David Waern at 2007-10-21T14:34:26+00:00 Add some Outputable utils - - - - - 69e7e47f by David Waern at 2007-10-21T14:35:49+00:00 FIX: Ord for OrdName was not comparing modules - - - - - 5a4ae535 by David Waern at 2007-10-21T21:18:48+00:00 Wibble - - - - - 03d48e20 by David Waern at 2007-10-24T15:52:56+00:00 Remove Main from "other modules" - - - - - c66f6d82 by David Waern at 2007-10-24T16:37:18+00:00 Make it possible to run haddock on itself - - - - - 21d156d8 by David Waern at 2007-10-25T14:02:14+00:00 Don't set boot modules as targets - - - - - f8bcf91c by David Waern at 2007-10-31T22:11:17+00:00 Add optimisation flags - - - - - 7ac758f2 by David Waern at 2007-11-04T09:48:28+00:00 Go back to loading only targets (seems to work now) - - - - - 4862aae1 by David Waern at 2007-11-05T22:24:57+00:00 Do full compilation of modules -- temporary fix for GHC API problem - - - - - 697e1517 by David Waern at 2007-11-05T22:25:50+00:00 Don't warn about not being able to link to wired/system/builtin-names - - - - - 892186da by David Waern at 2007-11-06T00:49:21+00:00 Filter out instances with TyCons that are not exported - - - - - 9548314c by David Waern at 2007-11-06T09:37:14+00:00 Wibble - - - - - 5cafd627 by David Waern at 2007-11-08T01:43:07+00:00 Filter out all non-vanilla type sigs - - - - - 04621830 by David Waern at 2007-11-08T01:45:13+00:00 Synch loading of names from .haddock files with GHC's name cache - - - - - 88d37f77 by David Waern at 2007-11-08T01:46:21+00:00 Remove commented-out code - - - - - 6409c911 by David Waern at 2007-11-08T01:56:00+00:00 Small bugfix and cleanup in getDeclFromTyCls - - - - - af59d9c2 by David Waern at 2007-11-08T02:08:44+00:00 Remove OrdName stuff - - - - - 3a615e2e by David Waern at 2007-11-08T02:13:41+00:00 Update runtests.hs following changes to haddock - - - - - 01f3314e by David Waern at 2007-11-08T02:33:01+00:00 Complain if we can't link to wired-in names - - - - - fcafb5d1 by David Waern at 2007-11-09T02:40:16+00:00 Don't exit when there are no file arguments - - - - - 194bc332 by David Waern at 2007-11-09T02:55:37+00:00 Wibble - - - - - dbe4cb55 by David Waern at 2007-11-09T02:56:14+00:00 Wibble - - - - - 82869fda by David Waern at 2007-11-10T17:01:43+00:00 Introduce InstalledInterface structure and add more stuff to the .haddock files We introduce InstalledInterface capturing the part of Interface that is stored in the interface files. We change the ppHtmlContents and ppHtmllIndex to take this structure instead of a partial Interface. We add stuff like the doc map and exported names to the .haddock file (via InstalledInterface). - - - - - d6bb57bf by David Waern at 2007-11-10T17:19:48+00:00 FIX: contents and index should include external package modules when --gen-contents/--gen-index - - - - - e8814716 by David Waern at 2007-11-11T00:29:27+00:00 Remove lDocLinkName and its use in Html backend - - - - - 6f9bd702 by David Waern at 2007-11-11T00:50:57+00:00 Do some refactoring in the html backend This also merges an old patch by Augustsson: Wed Jul 12 19:54:36 CEST 2006 lennart.augustsson at credit-suisse.com * Print type definitions like signatures if given arrows. - - - - - 09d0ce24 by Malcolm.Wallace at 2006-07-20T13:13:57+00:00 mention HsColour in the docs, next to option flags for linking to source code - - - - - 24da6c34 by Malcolm.Wallace at 2006-07-20T13:14:50+00:00 change doc references to CVS to give darcs repository location instead - - - - - 74d52cd6 by David Waern at 2007-11-11T00:55:33+00:00 Update copyright - - - - - fcaa3b4f by Duncan Coutts at 2006-09-08T13:41:00+00:00 Eliminate dep on network by doing a little cut'n'paste haddock depending on the network causes a circular dependency at least if you want to build the network lib with haddock docs. - - - - - 10cc9bda by David Waern at 2007-11-11T02:09:41+00:00 Fix conflicts - - - - - 4e3acd39 by David Waern at 2007-11-11T02:21:19+00:00 Manual merge of a patch from Duncan Coutts that removes the dependency on mtl - - - - - fa9070da by Neil Mitchell at 2006-09-29T15:52:03+00:00 Do not generate an empty table if there are no exports, this fixes a <table></table> tag being generated, which is not valid HTML 4.01 - - - - - d7431c85 by David Waern at 2007-11-11T02:28:50+00:00 Fix conflicts - - - - - f87e8f98 by Simon Marlow at 2006-10-10T11:37:16+00:00 changes for 0.8 - - - - - db929565 by Simon Marlow at 2006-10-10T12:07:12+00:00 fix the name of the source file - - - - - 8220aa4b by Simon Marlow at 2006-10-11T14:17:37+00:00 Rename haddock.js to haddock-util.js haddock.js will be run automatically by Windows when you type 'haddock' if it is found on the PATH, so rename to avoid confusion. Spotted by Adrian Hey. - - - - - 6bccdaa1 by sven.panne at 2006-10-12T15:28:23+00:00 Cabal's sdist does not generate "-src.tar.gz" files, but ".tar.gz" ones - - - - - d3f3fc19 by Simon Marlow at 2006-12-06T16:05:07+00:00 add todo item for --maintainer - - - - - 2da7e269 by Simon Marlow at 2006-12-15T15:52:00+00:00 TODO: do something better about re-exported symbols from another package - - - - - 42d85549 by David Waern at 2007-11-11T02:30:59+00:00 Fix conflicts - - - - - 5e7ef6e5 by Neil Mitchell at 2007-01-11T15:41:15+00:00 Never do spliting index files into many - - - - - f3d4aebe by Neil Mitchell at 2007-01-11T17:07:09+00:00 Add searching on the index page - - - - - bad3ab66 by Neil Mitchell at 2007-01-11T18:17:46+00:00 Delete dead code, now there is only one index page - - - - - cd09eedb by Neil Mitchell at 2007-01-11T18:21:19+00:00 Delete more stuff that is no longer required - - - - - e2806646 by David Waern at 2007-11-11T02:41:53+00:00 Fix conflicts - - - - - a872a823 by Neil Mitchell at 2007-01-11T18:51:43+00:00 Make the index be in case-insensitive alphabetic order - - - - - 8bddd9d7 by Neil Mitchell at 2007-02-06T17:49:12+00:00 Do not create empty tables for data declarations which don't have any constructors, instances or comments. Gets better HTML 4.01 compliance - - - - - 036b8120 by David Waern at 2007-11-11T02:56:58+00:00 Fix conflicts - - - - - f50c1639 by Conal Elliott at 2007-02-14T21:54:00+00:00 added substitution %{FILE///c} - - - - - 402e166a by David Waern at 2007-11-11T03:35:46+00:00 Manual merge of old patch: Sat Apr 21 04:36:43 CEST 2007 Roberto Zunino <zunrob at users.sf.net> * URL expansion for %%, %L, %{LINE} - - - - - 2f264fbd by David Waern at 2007-11-11T03:40:33+00:00 Manual merge of an old patch: Thu Apr 19 20:23:40 CEST 2007 Wolfgang Jeltsch <g9ks157k at acme.softbase.org> * bug fix When Haddock was invoked with the --ignore-all-exports flag but the ignore-exports module attribute wasn't used, hyperlinks weren't created for non-exported names. This fix might not be as clean as one would wish (since --ignore-all-exports now results in ignore_all_exports = True *and* an additional OptIgnoreExports option for every module) but at least the bug seems to be resolved now. - - - - - 7d7ae106 by sven.panne at 2007-09-02T12:18:02+00:00 Install LICENSE in the correct place - - - - - 66eaa924 by David Waern at 2007-11-11T19:02:46+00:00 Fix a bug that made haddock loop - - - - - 4ed47b58 by David Waern at 2007-11-11T19:03:09+00:00 Rename java-script file (this wasn't merge correctly) - - - - - d569534a by David Waern at 2007-11-11T19:06:44+00:00 Don't require -B <ghc-libdir> when no argument files Change readInterfaceFile to take a Maybe Session, to avoid having to pass -B <ghc-libdir> to Haddock when there're no source files to process. This is nice when computing contents/index for external packages. - - - - - 373368bc by Neil Mitchell at 2007-01-11T18:22:44+00:00 Change from tabs to spaces in the ppHtmlIndex function - - - - - 6b063a77 by Neil Mitchell at 2007-01-12T12:17:46+00:00 Rewrite much of the index searching code, previously was too slow to execute on the base library with IE, the new version guarantees less than O(log n) operations be performed, where n is the number in the list (before was always O(n)) - - - - - bfad00b7 by David Waern at 2007-11-11T23:33:53+00:00 Fix conflicts - - - - - cd2dcc09 by Neil Mitchell at 2007-01-12T12:25:01+00:00 Make the max number of results 75 instead of 50, to allow map searching in the base library to work - - - - - 3ae74764 by Neil Mitchell at 2007-01-12T12:58:17+00:00 Make the search box in a form so that enter does the default search - - - - - 142103e5 by David Waern at 2007-11-12T00:03:18+00:00 Merge patch from the old branch: Fri Aug 31 13:21:45 CEST 2007 Duncan Coutts <duncan at haskell.org> * Add category: Development to .cabal file Otherwise it appears on the hackage website in the "Unclassified" category. - - - - - 22ec2ddb by David Waern at 2007-11-25T01:55:29+00:00 A a list of small improvements to the TODO file - - - - - eb0129f4 by Wolfgang Jeltsch at 2007-12-03T23:47:55+00:00 addition of type equality support (at least for HTML generation) - - - - - 816a7e22 by David Waern at 2007-12-08T15:46:26+00:00 Handle class operators correctly when rendering predicates - - - - - 68baaad2 by David Waern at 2007-12-08T16:15:54+00:00 Code layout changes - - - - - 09b77fb4 by David Waern at 2007-12-08T16:16:03+00:00 Handle infix operators correctly in the Type -> HsType translation - - - - - 31c36da2 by David Waern at 2007-12-08T16:24:27+00:00 Add ppLParendTypes/ppLParendType - - - - - b17cc818 by David Waern at 2007-12-08T16:26:12+00:00 Use ppParendType when printing types args in predicates - - - - - ffd1f2cf by David Waern at 2007-12-08T16:45:06+00:00 Fix rendering of instance heads to handle infix operators This is also a refactoring to share this code for rendering predicates. - - - - - ff886d45 by David Waern at 2007-12-08T17:27:46+00:00 Fix rendering of class operators - - - - - e2fcbb9e by David Waern at 2007-12-08T17:59:28+00:00 Fix a bug (use ppTyName instead of ppName to print names in type apps) - - - - - 79a1056e by David Waern at 2007-12-08T21:25:18+00:00 Update tests - - - - - 867741ac by David Waern at 2007-12-08T21:25:49+00:00 Give a diff on test failure - - - - - 7e5eb274 by David Waern at 2008-01-05T14:33:45+00:00 Add DrIFT commands - - - - - 3656454d by David Waern at 2008-01-05T20:26:00+00:00 Add "cabal-version: >= 1.2" to the cabal file - - - - - 77974efc by Simon Marlow at 2007-12-20T09:52:44+00:00 add an item - - - - - f6ac1708 by Simon Marlow at 2007-12-06T14:00:10+00:00 Source links must point to the original module, not the referring module - - - - - eda1d5c9 by David Waern at 2008-01-06T14:40:52+00:00 Manual merge of a patch to the 0.8 branch Thu Dec 6 15:00:10 CET 2007 Simon Marlow <simonmar at microsoft.com> * Source links must point to the original module, not the referring module - - - - - 378f4085 by David Waern at 2008-01-06T16:03:45+00:00 Change stability from stable to experimental - - - - - 8bdafe44 by David Waern at 2008-01-06T16:14:22+00:00 Add haskell.vim (it had been removed somehow) - - - - - ea34d02e by David Waern at 2008-01-06T16:36:57+00:00 Change version to 2.0.0.0 - - - - - 34631ac0 by David Waern at 2008-01-06T16:44:57+00:00 Add missing modules to the cabal file - - - - - 9e142935 by David Waern at 2008-01-06T17:25:42+00:00 Depend on ghc >= 6.8.2 && < 6.9 - - - - - 59f9eeaa by Simon Marlow at 2007-12-20T10:43:04+00:00 add build scripts - - - - - 1c29ae30 by Simon Marlow at 2007-12-20T10:47:07+00:00 update version number - - - - - fe16a3e4 by Simon Marlow at 2007-12-20T10:48:03+00:00 update version - - - - - f688530f by Simon Marlow at 2007-12-20T10:48:29+00:00 doc updates - - - - - ce71b611 by David Waern at 2008-01-07T13:46:32+00:00 Change version in docs and spec - - - - - 03ab8d6f by David Waern at 2008-01-07T13:47:38+00:00 Manually merge over changes to CHANGES for 0.9 - - - - - 39f1b042 by David Waern at 2008-01-07T15:17:41+00:00 Remove the -use-package flag, we don't support it anyway - - - - - 7274a544 by David Waern at 2008-01-07T15:33:05+00:00 Update CHANGES for 2.0.0.0 - - - - - 96594f5d by David Waern at 2008-01-07T15:46:49+00:00 Wibble - - - - - f4c5a4c4 by David Waern at 2008-01-07T15:55:36+00:00 Change url to repo in documentation - - - - - 8a4c77f0 by David Waern at 2008-01-07T16:00:54+00:00 Update CHANGES - - - - - cb3a9288 by David Waern at 2008-01-07T16:02:55+00:00 Documentation fix - - - - - d8e45539 by David Waern at 2008-01-07T16:12:00+00:00 Update docs to say that Haddock accets .lhs files and module names - - - - - 4b5ce824 by David Waern at 2008-01-07T16:12:25+00:00 Document -B option - - - - - 47274262 by David Waern at 2008-01-07T16:23:07+00:00 Update CHANGES - - - - - 7ff314a9 by David Waern at 2008-01-07T16:23:20+00:00 Remove --use-package, --package & --no-implicit.. flags from docs - - - - - 6c3819c0 by David Waern at 2008-01-07T16:23:52+00:00 Remove --no-implicit-prelide flag - - - - - 1b14ae40 by David Waern at 2008-01-07T16:32:26+00:00 Update the "Using literate or pre-processed source" section - - - - - 0117f620 by David Waern at 2008-01-07T16:41:55+00:00 Document the --optghc flag - - - - - 087ab1cf by David Waern at 2008-01-07T16:42:10+00:00 Remove the documenation section on derived instances The problem mentioned there doesn't exist in Haddock 2.0.0.0 - - - - - 7253951e by David Waern at 2008-01-07T16:48:40+00:00 Document OPTIONS_HADDOCK - - - - - 3b6bdcf6 by David Waern at 2008-01-07T16:56:54+00:00 Wibble - - - - - 3025adf9 by David Waern at 2008-01-07T17:08:14+00:00 Wibble - - - - - 5f30f1a0 by David Waern at 2008-01-07T17:15:44+00:00 Change synopsis field to description - - - - - 1673f54b by David Waern at 2008-01-07T17:18:21+00:00 Change my email address in the cabal file - - - - - 55aa9808 by David Waern at 2008-01-07T18:18:02+00:00 Add documentation for readInterfaceFile - - - - - eaea417f by David Waern at 2008-01-07T18:21:30+00:00 Export necessary stuff from Distribution.Haddock - - - - - 7ea18759 by David Waern at 2008-01-07T18:31:49+00:00 Remove dep on Cabal - - - - - 7b79c74e by David Waern at 2008-01-07T18:33:49+00:00 Remove dep on process - - - - - ce3054e6 by David Waern at 2008-01-16T23:01:21+00:00 Add feature-requsts from Henning Thielemann to TODO - - - - - 0c08f1ec by David Waern at 2008-01-16T23:03:02+00:00 Record a bug in TODO - - - - - b04605f3 by David Waern at 2008-01-23T16:59:06+00:00 Add a bug reported by Ross to TODO - - - - - 5b17c030 by David Waern at 2008-01-23T18:05:53+00:00 A a bug report to TODO - - - - - 1c993b0d by David Waern at 2008-01-25T16:30:25+00:00 Accept test output - - - - - c22fc0d0 by David Waern at 2008-01-25T16:34:49+00:00 Accept test output - - - - - 4b795811 by David Waern at 2008-01-25T16:38:37+00:00 Change Hidden.hs (test) to use OPTIONS_HADDOCK - - - - - c124dbd9 by David Waern at 2008-01-25T16:39:23+00:00 Accept test output - - - - - ec6f6eea by David Waern at 2008-01-25T16:42:08+00:00 Add Hidden.html.ref to tests - - - - - 1dc9610c by David Waern at 2008-02-02T20:50:51+00:00 Add a comment about UNPACK bug in TODO - - - - - 2d3f7081 by David Waern at 2008-02-09T22:33:24+00:00 Change the representation of DocNames Ross Paterson reported a bug where links would point to the defining module instead of the "best" module for an identifier (e.g Int pointing to GHC.Base instead of Data.Int). This patch fixes this problem by refactoring the way renamed names are represented. Instead of representing them by: > data DocName = Link Name | NoLink Name they are now represented as such: > data DocName = Documented Name Module | Undocumented Name and the the link-env looks like this: > type LinkEnv = Map Name Module There are several reasons for this. First of all, the bug was caused by changing the module part of Names during the renaming process, without changing the Unique field. This caused names to be overwritten during the loading of .haddock files (which caches names using the NameCache of the GHC session). So we might create new Uniques during renaming to fix this (but I'm not sure that would be problem-free). Instead, we just keep the Name and add the Module where the name is best documented, since it can be useful to keep the original Name around (for e.g. source-code location info and for users of the Haddock API). Also, the names Link/NoLink don't really make sense, since wether to use links or not is entirely up to the users of DocName. In the process of following this change into H.Backends.Html I removed the assumption that binder names are Undocumented (which was just an unnecessary assumption, the OccName is the only thing needed to render these). This will probably make it possible to get rid of the renamer and replace it with a traversal from SYB or Uniplate. Since DocName has changed, InterfaceFile has changed so this patch also increments the file-format version. No backwards-compatibility is implemented. - - - - - 0f28c921 by David Waern at 2008-02-09T23:00:36+00:00 H.GHC.Utils: remove unused imports/exports - - - - - 0c44cad5 by David Waern at 2008-02-10T00:28:13+00:00 H.GHC.Utils: add some functions that were removed by mistake - - - - - e3452f49 by David Waern at 2008-02-10T00:28:48+00:00 Fix some trivial warnings in H.InterfaceFile - - - - - a6d74644 by David Waern at 2008-02-10T00:48:06+00:00 Update the version message to fit in small terminals - - - - - 76c9cd3e by David Waern at 2008-02-10T14:47:39+00:00 Remove bugs from TODO that don't apply anymore since the port - - - - - 5e10e090 by David Waern at 2008-02-10T15:22:47+00:00 Remove bugs from TODO that weren't actual bugs - - - - - fef70878 by David Waern at 2008-02-10T15:23:44+00:00 Remove yet another item from TODO that was not an actual bug - - - - - e1af47b8 by David Waern at 2008-02-11T10:25:57+00:00 Bump the version number to 2.1.0 Since the exported datatype DocName has changed, we need to bump the major version number. Let's also drop the fourth version component, it's not that useful. - - - - - e3be7825 by David Waern at 2008-04-11T14:29:04+00:00 Add a bug to TODO - - - - - cb6574be by David Waern at 2008-04-11T16:00:45+00:00 Use the in-place haddock when running tests - - - - - c6d7af0d by David Waern at 2008-04-11T16:09:16+00:00 Turn off GHC warnings when running tests - - - - - 7f61b546 by David Waern at 2008-04-11T17:24:00+00:00 Add a flag for turning off all warnings - - - - - 883b8422 by David Waern at 2008-04-12T14:02:18+00:00 Fix printing of data binders - - - - - 2a0db8fc by David Waern at 2008-04-12T18:52:46+00:00 Fix missing parenthesis in constructor args bug - - - - - 1b3ac3f9 by David Waern at 2008-04-12T18:57:23+00:00 Simplify test suite and add tests I move all tests into one single directory to simplify things, and add a test for the last bug that was fixed. - - - - - 8f178376 by David Waern at 2008-04-12T19:00:15+00:00 Add a script for copying test output to "expected" output - - - - - 193e3a03 by David Waern at 2008-04-12T19:16:37+00:00 Remove two fixed bugs from TODO - - - - - ddc9130c by David Waern at 2008-04-12T19:37:06+00:00 Update test README - - - - - 956069c0 by David Waern at 2008-05-01T12:16:14+00:00 Update version number in spec and docs - - - - - 5478621c by David Waern at 2008-05-01T12:28:12+00:00 Remove claim of backwards compatibility from docs for readInterfaceFile - - - - - 4a16dea9 by David Waern at 2008-05-01T12:33:04+00:00 Update CHANGES - - - - - 804216fb by David Waern at 2008-05-01T12:43:16+00:00 Add a synopsis - - - - - fd0c84d5 by David Waern at 2008-05-01T12:44:44+00:00 Add Haddock.DocName to the cabal file - - - - - 9f4a7439 by David Waern at 2008-05-01T12:45:53+00:00 Remove -fglasgow-exts and -fasm - - - - - aee7c145 by David Waern at 2008-05-01T12:54:01+00:00 Add LANGUAGE pragmas to source files - - - - - 9a58428b by David Waern at 2008-05-01T12:54:19+00:00 Add extensions to cabal file - - - - - 494f1bee by David Waern at 2008-05-01T13:12:09+00:00 Export DocName in the API - - - - - c938196b by David Waern at 2008-05-01T13:12:19+00:00 Add hide options to some source files - - - - - 236e86af by Neil Mitchell at 2008-06-07T20:45:10+00:00 Rewrite the --hoogle flag support - - - - - 6d910950 by Neil Mitchell at 2008-06-14T10:56:50+00:00 Simplify the newtype/data outputting in Hoogle, as haddock does it automatically - - - - - f87a95a8 by Neil Mitchell at 2008-06-14T12:10:18+00:00 Add initial structure for outputting documentation as well, but does not yet output anything - - - - - 7c3bce54 by Neil Mitchell at 2008-06-14T12:27:07+00:00 Remove <document comment> from the Hoogle output - - - - - 9504a325 by Neil Mitchell at 2008-06-16T06:33:21+00:00 Default to "main" if there is no package, otherwise will clobber hoogle's hoogle info - - - - - 4a794a79 by Neil Mitchell at 2008-06-16T06:53:29+00:00 Change packageName to packageStr, as it better reflects the information stored in it - - - - - 7abc9baf by Neil Mitchell at 2008-06-16T07:09:49+00:00 Add modulePkgInfo to Haddock.GHC.Utils, which gives back package name and version info - - - - - 8ca11514 by Neil Mitchell at 2008-06-16T07:13:48+00:00 Change Hoogle to take the package name and package version separately - - - - - a6da452d by Neil Mitchell at 2008-06-18T11:29:46+00:00 In Hoogle do not list things that are not local to this module - - - - - 974b76b7 by David Waern at 2008-06-19T18:40:13+00:00 Be more consistent with GHC API naming in H.GHC.Utils - - - - - 2facb4eb by David Waern at 2008-06-19T19:03:03+00:00 Update test output - - - - - c501de72 by David Waern at 2008-06-26T20:26:49+00:00 Use ghc-paths to get the lib dir The path can still be overridden using the -B flag. It's not longer required to pass the lib dir to the program that runs the test suite. - - - - - ac4c6836 by David Waern at 2008-06-26T20:33:08+00:00 Update CHANGES - - - - - 9d21c60a by David Waern at 2008-06-26T20:34:53+00:00 Update README - - - - - 741448f0 by David Waern at 2008-06-26T21:12:57+00:00 Improve wording in the help message - - - - - b1b42b11 by David Waern at 2008-06-30T10:16:17+00:00 Rename ForeignType - - - - - 6d6c2b34 by David Waern at 2008-06-30T10:25:09+00:00 Rename TyFamily - - - - - 8d1125ed by David Waern at 2008-06-30T10:37:21+00:00 Rename type patterns - - - - - 7610a4cb by David Waern at 2008-06-30T10:45:07+00:00 Rename associated types - - - - - 8eeba14c by David Waern at 2008-06-30T10:47:41+00:00 Remove the TODO file now that we have a trac - - - - - 1af5b25b by David Waern at 2008-07-02T18:19:28+00:00 Render type family declarations (untested) - - - - - ceb99797 by David Waern at 2008-07-02T18:24:06+00:00 Remove redundant check for summary when rendering data types - - - - - b36a58e0 by David Waern at 2008-07-02T22:01:38+00:00 More support for type families and associated types Now we just need to render the instances - - - - - 78784879 by David Waern at 2008-07-07T22:13:58+00:00 Remove filtering of instances We were filtering out all instances for types with unknown names. This was probably an attempt to filter out instances for internal types. I am removing the filtering for the moment, and will try to fix this properly later. - - - - - 3e758dad by David Waern at 2008-06-30T18:50:30+00:00 Run haddock in-place during testing - - - - - d9dab0ce by David Waern at 2008-07-08T21:04:32+00:00 Remove index.html and doc-index.html from output, they should not be versioned - - - - - 3e6c4681 by David Waern at 2008-07-08T21:06:42+00:00 Update test output following change to instance filtering - - - - - e34a3f14 by David Waern at 2008-07-12T16:48:28+00:00 Stop using the map from exported names to declarations During creation of the interface, we were using two maps: one from exported names to declarations, and one from all defined names in the module to declarations. The first contained subordinate names while the second one didn't. The first map was never used to look up names not defined in the associated module, so if we add subordinate names to the second map, we could use it everywhere. That's that this patch does. This simplifies code because we don't have to pass around two maps everywhere. We now store the map from locally defined things in the interface structure instead of the one from exported names. - - - - - 2e1d2766 by David Waern at 2008-07-12T16:55:21+00:00 Get the all locally defined names from GHC API We previously had some code to compute all locally defined names in a module including subordinate names. We don't need it since we can get the names from modInfoTyThings in the GHC API. - - - - - bf637994 by David Waern at 2008-07-13T13:09:16+00:00 Refactoring in H.Interface.Create We were creating a doc map, a declaration map and a list of entities separately by going through the HsGroup. These structures were all used to build the interface of a module. Instead of doing this, we can start by creating a list of declarations from the HsGroup, then collect the docs directly from this list (instead of using the list of entities), creating a documentation map. We no longer need the Entity data type, and we can store a single map from names to declarations and docs in the interface, instead of the declaration map and the doc map. This way, there is only one place where we filter out the declarations that we don't want, and we can remove a lot of code. Another advantage of this is that we can create the exports directly out of the list of declarations when we export the full module contents. (Previously we did a look up for each name to find the declarations). This is faster and removes another point where we depend on names to identify exported declarations, which is good because it eliminates problems with instances (which don't have names). - - - - - 547e410e by David Waern at 2008-07-13T13:34:51+00:00 Remove FastString import and FSLIT macro in H.I.Create -- they were unused - - - - - 693759d1 by David Waern at 2008-07-13T13:36:23+00:00 Remove unused import from H.I.Create - - - - - cde6e7fb by David Waern at 2008-07-13T13:51:54+00:00 Small touches - - - - - 96de8f1d by David Waern at 2008-07-20T11:21:46+00:00 Preparation for rendering instances as separate declarations We want to be able to render instances as separate declarations. So we remove the Name argument of ExportDecl, since instances are nameless. This patch also contains the first steps needed to gather type family instances and display them in the backend, but the implementation is far from complete. Because of this, we don't actually show the instances yet. - - - - - b0f824fb by David Waern at 2008-07-20T15:53:08+00:00 Follow changes to ExportDecl in Hoogle - - - - - 1192eff3 by Neil Mitchell at 2008-06-26T00:28:10+00:00 Change how the Hoogle backend outputs classes, adding the context in - - - - - 7a0d1464 by Neil Mitchell at 2008-06-26T00:28:46+00:00 Remove the indent utility function from Hoogle backend - - - - - 3361241b by Neil Mitchell at 2008-06-26T09:45:09+00:00 Add support for Hoogle writing ForeignImport/ForeignExport properly - - - - - 795ad3bf by Neil Mitchell at 2008-06-26T12:15:25+00:00 Flesh out the Hoogle code to render documentation - - - - - 23277995 by Neil Mitchell at 2008-06-26T14:56:41+00:00 Fix a bug in the Hoogle backend, unordered lists were being written out <ul>...</u> - - - - - db739b27 by Neil Mitchell at 2008-06-26T15:09:54+00:00 Remove any white space around a <li> element - - - - - f2e6bb8c by Neil Mitchell at 2008-07-10T15:30:47+00:00 Remove the TODO in the Hoogle HTML generation, was already done - - - - - 693ec9a3 by Neil Mitchell at 2008-07-10T15:53:00+00:00 Put brackets round operators in more places in the Hoogle output - - - - - 842313aa by Neil Mitchell at 2008-07-10T16:01:25+00:00 Print type signatures with brackets around the name - - - - - cf93deb0 by David Waern at 2008-07-20T17:04:22+00:00 Bump version number to 2.2.0 - - - - - 30e6a8d1 by David Waern at 2008-07-20T17:04:41+00:00 Resolve conflicts in H.B.Hoogle - - - - - 1f0071c9 by David Waern at 2008-07-23T23:05:01+00:00 Add "all" command to runtests.hs that runs all tests despite failures - - - - - f2723023 by David Waern at 2008-07-23T23:08:39+00:00 Update tests/README - - - - - c0304a11 by David Waern at 2008-07-23T23:21:15+00:00 Be compatible with GHC 6.8.3 The cabal file is converted to use the "new" syntax with explicit Library and Executable sections. We define the __GHC_PATCHLEVEL__ symbol using a conditinal cpp-options field in the cabal file. (Ideally, Cabal would define the symbol for us, like it does for __GLASGOW_HASKELL__). We use these symbols to #ifdef around a small difference between 6.8.2 and 6.8.3. Previously, we only supported GHC 6.8.2 officially but the dependencies field said "ghc <= 6.9". This was just for convenience when testing against the (then compatible) HEAD version of GHC, and was left in the release by mistake. Now, we support both GHC 6.8.2 and 6.8.3 and the dependencies field correctly reflects this. - - - - - 88a5fe71 by David Waern at 2008-07-23T23:54:16+00:00 Depend on the currently available ghc-paths versions only - - - - - 8738d97b by David Waern at 2008-07-24T10:50:44+00:00 FIX haskell/haddock#44: Propagate parenthesis level when printing documented types - - - - - 05339119 by David Waern at 2008-07-24T16:06:18+00:00 Drop unnecessary parenthesis in types, put in by the user We were putting in parenthesis were the user did. Let's remove this since it just clutters up the types. The types are readable anyway since we print parens around infix operators and do not rely on fixity levels. When doing this I discovered that we were relying on user parenthesis when printin types like (a `O` b) c. This patchs fixes this problem so that parenthesis are always inserted around an infix op application in case it is applied to further arguments, or if it's an arguments to a type constructor. Tests are updated. - - - - - b3a99828 by David Waern at 2008-07-24T10:19:43+00:00 Print parenthesis around non-atomic banged types Fixes half of haskell/haddock#44 - - - - - ab5238e0 by David Waern at 2008-07-24T22:07:49+00:00 Add a reference file for the TypeFamilies test - - - - - 1941cc11 by David Waern at 2008-07-25T17:15:53+00:00 Simplify definition of pretty and trace_ppr - - - - - e3bfa33c by David Waern at 2008-07-25T17:18:27+00:00 Warning messages Output a warning when filtering out data/type instances and associated types in instances. We don't show these in the documentation yet, and we need to let the user know. - - - - - 9b85fc89 by David Waern at 2008-07-25T17:45:40+00:00 Doc: Mention Hoogle in the Introduction - - - - - afb2dd60 by David Waern at 2008-07-25T17:49:00+00:00 Doc: update -B description - - - - - 584c0c91 by David Waern at 2008-07-25T18:11:38+00:00 Doc: describe -w flag - - - - - 77619c24 by David Waern at 2008-07-28T12:29:07+00:00 Remove TODO from cabal file - - - - - 96717d5f by David Waern at 2008-07-28T12:29:27+00:00 Support type equality predicates - - - - - c2fd2330 by David Waern at 2008-07-29T19:45:14+00:00 Move unL from H.B.Hoogle to H.GHC.Utils I like Neil's shorter unL better than unLoc from the GHC API. - - - - - c4c3bf6a by David Waern at 2008-07-29T19:47:36+00:00 Do not export ATs when not in list of subitems - - - - - bf9a7b85 by David Waern at 2008-08-03T11:42:59+00:00 Filter out ForeignExports - - - - - df59fcb0 by David Waern at 2008-08-03T14:02:51+00:00 Filter out more declarations The previous refactorings in H.I.Create introduced a few bugs. Filtering of some types of declarations that we don't handle was removed. This patch fixes this. - - - - - 2f8a958b by David Waern at 2008-08-03T15:24:07+00:00 Move reL to H.GHC.Utils so we can use it everywhere - - - - - 8ec15efd by David Waern at 2008-08-03T15:25:00+00:00 Use isVanillaLSig from GHC API instead of home brewn function - - - - - 300f93a2 by David Waern at 2008-08-03T15:25:27+00:00 Filter out separately exported ATs This is a quick and dirty hack to get rid of separately exported ATs. We haven't decided how to handle them yet. No warning message is given. - - - - - 8776d1ec by David Waern at 2008-08-03T16:21:21+00:00 Filter out more declarations and keep only vanilla type sigs in classes - - - - - ea07eada by David Waern at 2008-08-03T16:48:00+00:00 Fix layout - - - - - dd5e8199 by David Waern at 2008-08-03T16:50:52+00:00 Move some utility functions from H.I.Create to H.GHC.Utils - - - - - 4a1dbd72 by David Waern at 2008-08-03T17:39:55+00:00 Do not filter out doc declarations - - - - - 0bc8dca4 by David Waern at 2008-08-03T17:47:26+00:00 Filter out separately exported ATs (take two) - - - - - af970fe8 by David Waern at 2008-08-03T22:39:17+00:00 Update CHANGES - - - - - 5436ad24 by David Waern at 2008-08-03T22:40:20+00:00 Bump version number to 2.2.1 - - - - - d66de448 by David Waern at 2008-08-05T19:00:32+00:00 Remove version restriction on ghc-paths - - - - - 534b1364 by David Waern at 2008-08-05T19:04:35+00:00 Bump version to 2.2.2 and update CHANGES - - - - - 549188ff by David Waern at 2008-08-05T19:16:49+00:00 Fix CHANGES - - - - - 0d156bb4 by Luke Plant at 2008-08-11T15:20:59+00:00 invoking haddock clarification and help - - - - - 748295cc by David Waern at 2008-08-11T18:56:37+00:00 Doc: say that the --hoogle option is functional - - - - - 43301db4 by David Waern at 2008-08-05T19:26:08+00:00 Change ghc version dependency to >= 6.8.2 - - - - - 3e5a53b6 by David Waern at 2008-08-10T22:42:05+00:00 Make H.GHC.Utils build with GHC HEAD - - - - - 7568ace0 by David Waern at 2008-08-11T19:41:54+00:00 Import Control.OldException instead of C.Exception when using ghc >= 6.9 We should really test for base version instead, but I don't currently know which version to test for. - - - - - b71ae991 by David Waern at 2008-08-12T22:40:39+00:00 Make our .haddock file version number depend on the GHC version We need to do this, since our .haddock format can potentially change whenever GHC's version changes (even when only the patchlevel changes). - - - - - 6307ce3f by David Waern at 2008-08-12T22:49:57+00:00 Remove matching on NoteTy in AttachInstances, it has been removed - - - - - 2dbcfd5f by David Waern at 2008-08-12T23:02:02+00:00 Comment out H.GHC.loadPackages - it is unused and doesn't build with ghc >= 6.9 - - - - - c74db5c2 by David Waern at 2008-08-12T23:03:58+00:00 Hide <.> from GHC import in Hoogle only for ghc <= 6.8.3 - - - - - 69a44ebb by David Waern at 2008-08-12T23:11:12+00:00 Follow changes to parseDynamic/StaticFlags - - - - - 5881f3f0 by David Waern at 2008-08-13T21:43:58+00:00 Add __GHC_PATCHLEVEL__ symbol also when building the library - - - - - 8574dc11 by David Waern at 2008-08-13T21:44:17+00:00 Follow move of package string functions from PackageConfig to Module - - - - - c9baa77f by David Waern at 2008-08-13T21:45:29+00:00 Follow extensible exceptions changes - - - - - 9092de15 by David Waern at 2008-08-13T21:46:20+00:00 Update test following Haddock version change - - - - - ebe569a4 by David Waern at 2008-08-13T21:46:54+00:00 Follow changes to parseDynamic- parseStaticFlags in GHC - - - - - b8a5ffd3 by David Waern at 2008-08-13T21:47:36+00:00 Follow changes to Binary in GHC 6.9 - - - - - edfda1cc by David Waern at 2008-08-13T21:50:17+00:00 Change ghc version dependency to >= 6.8.2 && <= 6.9 - - - - - d59be1cf by Neil Mitchell at 2008-08-12T16:02:53+00:00 Output all items, even if they are not defined in this module - ensures map comes from Prelude, not just GHC.Base - - - - - dda93b9f by Neil Mitchell at 2008-08-12T21:37:32+00:00 Add support for type synonyms to Hoogle, was accidentally missing before (woops!) - - - - - b6ee795c by Neil Mitchell at 2008-08-13T14:03:24+00:00 Generalise Hoogle.doc and add a docWith - - - - - 415e1bb2 by Neil Mitchell at 2008-08-13T14:03:46+00:00 Make Hoogle add documentation to a package - - - - - 790a1202 by Neil Mitchell at 2008-08-18T12:52:43+00:00 Use the same method to put out signatures as class methods in the Hoogle backend - - - - - ded37eba by Neil Mitchell at 2008-08-18T12:53:04+00:00 Remove Explicit top-level forall's when pretty-printing signatures - - - - - 6468c722 by Neil Mitchell at 2008-08-20T07:59:13+00:00 Simplify the code by removing not-to-important use of <.> in the Hoogle back end - - - - - 788c3a8b by Neil Mitchell at 2008-08-21T18:20:24+00:00 In the hoogle back end, markup definition lists using <i>, not <b> - - - - - 77d4b000 by Ian Lynagh at 2008-08-14T10:49:14+00:00 Add a Makefile for GHC's build system. Still won't work yet, but we're closer - - - - - 920440d7 by Ian Lynagh at 2008-08-27T18:06:46+00:00 Add haddock.wrapper - - - - - bcda925f by Ian Lynagh at 2008-08-27T18:07:02+00:00 Add a manual Cabal flag to control the ghc-paths dependency - - - - - 04d194e2 by Ian Lynagh at 2008-08-27T20:41:27+00:00 Update extensions in Cabal file Use ScopedTypeVariables instead of PatternSignatures - - - - - 12480043 by Ian Lynagh at 2008-08-27T20:41:55+00:00 Increase the upper bound on the GHC version number - - - - - b1f809a5 by Ian Lynagh at 2008-08-27T21:32:22+00:00 Fix some warnings - - - - - aea0453d by Ian Lynagh at 2008-08-28T14:22:29+00:00 Fixes for using haddock in a GHC build tree - - - - - ad23bf86 by Ian Lynagh at 2008-08-28T21:14:27+00:00 Don't use Cabal wrappers on Windows - - - - - 35858e4c by Ian Lynagh at 2008-08-29T00:07:42+00:00 Fix in-tree haddock on Windows - - - - - c2642066 by Ian Lynagh at 2008-09-03T22:35:53+00:00 follow library changes - - - - - 2eb55d50 by Ian Lynagh at 2008-09-07T18:52:51+00:00 bindist fixes - - - - - 3daa5b59 by Ian Lynagh at 2008-09-10T16:58:18+00:00 We need to tell haddock that its datasubdir is . or it can't find package.conf - - - - - 388fd8c2 by Ian Lynagh at 2008-09-10T19:47:44+00:00 Fix haddock inplace on Windows - - - - - 70a641c1 by Ian Lynagh at 2008-09-10T22:15:44+00:00 Fix installed haddock on Windows - - - - - 83c1e997 by Neil Mitchell at 2008-09-11T10:48:55+00:00 Import GHC.Paths if not IN_GHC_TREE, seems to match the use of GHC.Paths functions much better - - - - - b452519b by Ian Lynagh at 2008-09-12T12:58:24+00:00 Add a LANGUAGE ForeignFunctionInterface pragma - - - - - afbd592c by Ian Lynagh at 2008-09-12T12:59:13+00:00 Wibble imports - - - - - 547ac4ad by Ian Lynagh at 2008-09-14T15:34:22+00:00 Add a "#!/bin/sh" to haddock.wrapper - - - - - f207a807 by Ian Lynagh at 2008-09-15T10:02:32+00:00 Use "exec" when calling haddock in the wrapper - - - - - 2ee68509 by Thomas Schilling at 2008-09-15T09:09:16+00:00 Port Haddock.Interface to new GHC API. This required one bigger change: 'readInterfaceFile' used to take an optional 'Session' argument. This was used to optionally update the name cache of an existing GHC session. This does not work with the new GHC API, because an active session requires the function to return a 'GhcMonad' action, but this is not possible if no session is provided. The solution is to use an argument of functions for reading and updating the name cache and to make the function work for any monad that embeds IO, so it's result type can adapt to the calling context. While refactoring, I tried to make the code a little more self-documenting, mostly turning comments into function names. - - - - - 3bb96431 by Thomas Schilling at 2008-09-15T09:09:37+00:00 Reflect GHC API changes. - - - - - 2e60f714 by Thomas Schilling at 2008-09-15T09:10:37+00:00 Port Haddock.GHC.Typecheck to new GHC API. - - - - - 9cfd4cff by Thomas Schilling at 2008-09-15T09:11:00+00:00 Port Haddock.GHC to new GHC API. - - - - - caffa003 by Thomas Schilling at 2008-09-15T09:11:25+00:00 Port Main to new GHC API. - - - - - 069a4608 by Ian Lynagh at 2008-09-21T11:19:00+00:00 Fix paths used on Windows frmo a GHC tree: There is no whare directory - - - - - 7ceee1f7 by Ian Lynagh at 2008-09-21T12:20:16+00:00 Fix the in-tree haddock on Windows - - - - - 0d486514 by Ian Lynagh at 2008-09-23T18:06:58+00:00 Increase the GHC upper bound from 6.11 to 6.13 - - - - - f092c414 by Neil Mitchell at 2008-09-11T14:56:07+00:00 Do not wrap __ in brackets - - - - - 036bdd13 by Ian Lynagh at 2008-09-28T01:42:35+00:00 Fix building haddock when GhcProfiled=YES - - - - - 01434a89 by David Waern at 2008-09-24T20:24:21+00:00 Add PatternSignatures LANGUAGE pragma to Main and Utils - - - - - 1671a750 by David Waern at 2008-10-02T22:57:25+00:00 For source links, get original module from declaration name instead of environment. Getting it from the environment must have been a remnant from the times when we were using unqualified names (versions 0.x). - - - - - a25dde99 by David Waern at 2008-10-02T22:59:57+00:00 Remove ifaceEnv from Interface - it's no longer used - - - - - 610993da by David Waern at 2008-10-02T23:04:58+00:00 Write a comment about source links for type instance declarations - - - - - 5a96b5d5 by Thomas Schilling at 2008-10-03T10:45:08+00:00 Follow GHC API change of parseModule. - - - - - 5a943ae5 by Ian Lynagh at 2008-10-03T15:56:58+00:00 TAG 2008-10-03 - - - - - 76cdd6ae by Thomas Schilling at 2008-10-08T12:29:50+00:00 Only load modules once when typechecking with GHC. This still doesn't fix the memory leak since the typechecked source is retained and then processed separately. To fix the leak, modules must be processed directly after typechecking. - - - - - 7074d251 by David Waern at 2008-10-09T23:53:54+00:00 Interleave typechecking with interface creation At the same time, we fix a bug where the list of interfaces were processed in the wrong order, when building the links and renaming the interfaces. - - - - - 4b9b2b2d by David Waern at 2008-10-09T23:54:49+00:00 Add some strictness annotations in Interface We add some strictness annotations to the fields of Interface, so that less GHC data is hold on to during processing. - - - - - 22035628 by David Waern at 2008-10-10T20:02:31+00:00 Remove typecheckFiles and MonadUtils import from H.GHC.Typeccheck - - - - - be637ad3 by David Waern at 2008-10-10T20:33:38+00:00 Make Haddock build with GHC 6.8.2 - - - - - 523b3404 by David Waern at 2008-10-10T21:08:09+00:00 Fix documentation for createInterfaces - - - - - e1556702 by David Waern at 2008-10-10T21:26:19+00:00 Hide H.Utils in library - - - - - a8e751c3 by David Waern at 2008-10-10T21:34:59+00:00 Add back .haddock file versioning based on GHC version It was accidentally removed in the patch for GHC 6.8.2 compatibility - - - - - 06fb3c01 by David Waern at 2008-10-10T21:47:15+00:00 Bump version number to 2.3.0 - - - - - ff087fce by David Waern at 2008-10-10T22:35:49+00:00 Add support for DocPic The support for DocPic was merged into the GHC source long ago, but the support in Haddock was forgotten. Thanks Peter Gavin for submitting this fix! - - - - - 3af85bf6 by David Waern at 2008-10-10T23:34:05+00:00 Update tests - - - - - 0966873c by Simon Marlow at 2008-10-10T14:43:04+00:00 no need for handleErrMsg now, we don't throw any ErrMsgs - - - - - f1870de3 by Clemens Fruhwirth at 2008-10-10T13:29:36+00:00 Compile with wrapper but remove it for dist-install - - - - - 7b440dc2 by David Waern at 2008-10-11T14:02:25+00:00 Remove interface from LinksInfo It was there to know the documentation home module when creating a wiki link, but we already know this since we have the DocName. - - - - - e5729e6a by David Waern at 2008-10-15T20:49:18+00:00 Wibble - - - - - b2a8e01a by David Waern at 2008-10-15T21:03:36+00:00 Use type synonyms for declarations and docs in H.I.Create - - - - - be71a15b by David Waern at 2008-10-15T21:12:17+00:00 Comment out unused type family stuff completely - - - - - 91aaf075 by David Waern at 2008-10-15T21:49:04+00:00 Wibble - - - - - 42ba4eb4 by David Waern at 2008-10-15T21:53:53+00:00 Move convenient type synonym to H.Types - - - - - db11b723 by David Waern at 2008-10-15T22:14:07+00:00 Add DeclInfo to H.Types - - - - - 193552b6 by David Waern at 2008-10-15T22:15:01+00:00 Add subordinates with docs to the declaration map The only place in the code where we want the subordinates for a declaration is right after having looked up the declaration in the map. And since we include subordinates in the map, we might as well take the opportunity to store those subordinates that belong to a particular declaration together with that declaration. We also store the documentation for each subordinate. - - - - - 31e6eebc by David Waern at 2008-10-16T17:18:47+00:00 Wibble - - - - - 0dcbd79f by David Waern at 2008-10-16T20:58:42+00:00 Fix haskell/haddock#61 We were not getting docs for re-exported class methods. This was because we were looking up the docs in a map made from the declarations in the current module being rendered. Obviously, re-exported class methods come from another module. Class methods and ATs were the only thing we were looking up using the doc map, everything else we found in the ExporItems. So now I've put subordinate docs in the ExportItem's directly, to make things a bit more consistent. To do this, I added subordinates to the the declarations in the declaration map. This was easy since we were computing subordinates anyway, to store stand-alone in the map. I added a new type synonym 'DeclInfo', which is what we call what is now stored in the map. This little refactoring removes duplicate code to retrieve subordinates and documentation from the HsGroup. - - - - - de47f20a by David Waern at 2008-10-16T22:06:35+00:00 Document function and improve its layout - - - - - e74e625a by Thomas Schilling at 2008-10-20T11:12:57+00:00 Force interface more aggressively. For running Haddock on GHC this reduces memory usage by about 50 MB on a 32 bit system. A heap profile shows total memory usage peak at about 100 MB, but actual usage is at around 300 MB even with compacting GC (+RTS -c). - - - - - b63ac9a1 by David Waern at 2008-10-20T20:25:50+00:00 Make renamer consistent Instead of explicitly making some binders Undocumented, treat all names the same way (that is, try to find a Documented name). - - - - - f6de0bb0 by Ian Lynagh at 2008-09-19T00:54:43+00:00 TAG GHC 6.10 fork - - - - - 74599cd0 by David Waern at 2008-10-20T21:13:24+00:00 Do not save hidden modules in the .haddock file We were saving interfaces of all processed modules including those hidden using {-# OPTIONS_HADDOCK hide #-} in the .haddock file. This caused broken links when generating the index for the libraries that come with GHC. This patch excludes modules with hidden documentation when writing .haddock files. It should fix the above problem. - - - - - 7b6742e9 by David Waern at 2008-10-21T19:54:52+00:00 Do not save hidden modules in the .haddock file (also for ghc >= 6.9) When writing the first patch, I forgot to do the fix in both branches of an #if macro. - - - - - b99b1951 by David Waern at 2008-10-22T20:04:18+00:00 Remove subordinate map and its usage It is not needed now that we store subordinate names in the DeclInfo map. - - - - - da97cddc by David Waern at 2008-10-22T20:11:46+00:00 Tidy up code in H.I.Create a little Remove commented out half-done type instance support, and remove DeclWithDoc synonym. - - - - - 6afa76f3 by David Waern at 2008-10-22T21:17:29+00:00 Fix warnings in H.GHC.Utils - - - - - 171ea1e8 by David Waern at 2008-10-22T21:35:04+00:00 Fix warnings in H.Utils - - - - - c8cb3b91 by David Waern at 2008-10-22T21:36:49+00:00 Wibble - - - - - 767fa06a by David Waern at 2008-10-27T19:59:04+00:00 Make named doc comments into ExportDoc instead of ExportDecl Fixes a crash when processing modules without export lists containing named docs. - - - - - e638bbc6 by David Waern at 2008-11-02T22:21:10+00:00 Add HCAR entry - - - - - 92b4ffcf by David Waern at 2008-11-02T22:44:19+00:00 Update CHANGES - - - - - 84d4da6e by David Waern at 2008-11-03T11:25:04+00:00 Add failing test for template haskell crash - - - - - 2a9cd2b1 by David Waern at 2008-11-04T21:13:44+00:00 Add tests/TH.hs - - - - - 8a59348e by David Waern at 2008-11-04T21:30:26+00:00 TAG 2.3.0 - - - - - 54f70d31 by Thomas Schilling at 2008-10-24T17:04:08+00:00 Enable framed view of the HTML documentation. This patch introduces: - A page that displays the documentation in a framed view. The left side will show a full module index. Clicking a module name will show it in the right frame. If Javascript is enabled, the left side is split again to show the modules at the top and a very short synopsis for the module currently displayed on the right. - Code to generate the mini-synopsis for each module and the mini module index ("index-frames.html"). - CSS rules for the mini-synopsis. - A very small amount of javascript to update the mini-synopsis (but only if inside a frame.) Some perhaps controversial things: - Sharing code was very difficult, so there is a small amount of code duplication. - The amount of generated pages has been doubled, since every module now also gets a mini-synopsis. The overhead should not be too much, but I haven't checked. Alternatively, the mini-synopsis could also be generated using Javascript if we properly annotate the actual synopsis. - - - - - 5d7ea5a6 by David Waern at 2008-11-04T23:20:17+00:00 Follow change to ExportDecl in frames code - - - - - 60e16308 by David Waern at 2008-11-04T23:35:26+00:00 Update CHANGES - - - - - d63fd26d by David Waern at 2008-11-04T23:37:43+00:00 Bump version number - - - - - c1660c39 by David Waern at 2008-11-04T23:44:46+00:00 Update CHANGES - - - - - 995ab384 by David Waern at 2008-11-04T23:55:21+00:00 Remove .ref files from tests/output/ - - - - - 1abbbe75 by David Waern at 2008-11-04T23:57:41+00:00 Output version info before running tests - - - - - 649b182f by David Waern at 2008-11-05T22:45:37+00:00 Add ANNOUNCE message - - - - - c36ae0bb by David Waern at 2008-11-05T23:15:35+00:00 Update ANNOUNCE - - - - - 9c4f3d40 by David Waern at 2008-11-05T23:18:30+00:00 Wibble - - - - - 5aac87ce by David Waern at 2008-11-06T21:07:48+00:00 Depend on base 4.* when using GHC >= 6.9, otherwise 3.* - - - - - b9796a74 by David Waern at 2008-11-06T21:13:40+00:00 Bump version to 2.4.1 and update CHANGES - - - - - d4b26baa by David Waern at 2008-11-06T21:26:33+00:00 Depend on base 4.0.* instead of 4.* - - - - - 2cb0903c by David Waern at 2008-11-06T21:46:53+00:00 Fix warnings in H.B.HH and H.B.HH2 - - - - - e568e89a by David Waern at 2008-11-06T21:47:12+00:00 Fix warnings in Haddock.ModuleTree - - - - - 9dc14fbd by David Waern at 2008-11-06T21:47:52+00:00 Fix warnings in Haddock.Version - - - - - 02ac197c by David Waern at 2008-11-06T21:51:31+00:00 Fix warnings in H.InterfaceFile and H.Options - - - - - 63e7439a by David Waern at 2008-11-06T21:59:45+00:00 Fix warnings in H.GHC.Typecheck - - - - - 4bca5b68 by David Waern at 2008-11-08T13:43:42+00:00 Set HscTarget to HscNothing instead of HscAsm There used to be a bug in the GHC API that prevented us from setting this value. - - - - - 07357aec by David Waern at 2008-11-09T22:27:00+00:00 Re-export NameCache and friends from Distribution.Haddock - - - - - ea554b5a by David Waern at 2008-11-09T23:14:10+00:00 Add Haddock.GHC.Utils to other-modules in library - - - - - 74aecfd7 by David Waern at 2008-11-10T01:18:57+00:00 Export DocName in the library - - - - - 241a58b3 by David Waern at 2008-11-10T01:19:18+00:00 Document the functions in H.DocName - - - - - edc2ef1b by David Waern at 2008-11-10T01:20:52+00:00 Export H.DocName in the library - - - - - 4f588d55 by David Waern at 2008-11-10T01:29:14+00:00 Make DocName an instance of NamedThing - - - - - b4647244 by David Waern at 2008-11-15T22:58:18+00:00 Reflect version bump in test suite - - - - - 4bee8ce2 by David Waern at 2008-11-15T22:58:45+00:00 Update tests For unknown reasons, test output for Bug1 and Test has changed for the better. - - - - - 1690e2f9 by David Waern at 2008-11-15T22:59:33+00:00 Store hidden modules in .haddock files We store documentation for an entity in the 'InstalledInterface' of the definition site module, and never in the same structure for a module which re-exports the entity. So when a client of the Haddock library wants to look up some documentation, he/she might need to access a hidden module. But we currently don't store hidden modules in the .haddock files. So we add the hidden modules and the Haddock options to the .haddock files. The options will be used to filter the module list to obtain the visible modules only, which is necessary for generating the contents and index for installed packages. - - - - - 8add6435 by David Waern at 2008-11-16T14:35:50+00:00 Bump major version number due to .haddock file format change - - - - - 48bfcf82 by David Waern at 2008-11-23T14:32:52+00:00 Update tests to account for version number bump - - - - - 0bbd1738 by David Waern at 2008-11-23T14:33:31+00:00 HADDOCK_DATA_DIR changed to haddock_datadir - - - - - 5088b78c by David Waern at 2008-11-23T17:13:21+00:00 FIX haskell/haddock#45: generate two anchors for each name We generate two anchor tags for each name, one where we don't escape the name and one where we URI-encode it. This is for compatibility between IE and Opera. Test output is updated. - - - - - 5ee5ca3b by Neil Mitchell at 2008-11-27T14:38:11+00:00 Drop HsDocTy annotations, they mess up pretty printing and also have a bracketing bug (#2584) - - - - - 51c014e9 by Roman Cheplyaka at 2008-11-27T22:27:36+00:00 Allow referring to a specific section within a module in a module link Fixes haskell/haddock#65 - - - - - 4094bdc5 by David Waern at 2008-11-28T21:13:33+00:00 Update tests following anchor change - - - - - f89552dd by Thomas Schilling at 2008-11-29T16:16:20+00:00 Haddock really shouldn't try to overwrite files. - - - - - 98127499 by David Waern at 2008-12-07T14:09:15+00:00 Solve conflict - - - - - 319356c5 by David Waern at 2008-10-22T21:16:55+00:00 Add -Wall -Werror to ghc-options - - - - - 3c4968c9 by David Waern at 2008-11-04T23:38:56+00:00 TAG 2.4.0 - - - - - 4b21e003 by David Waern at 2008-11-06T21:14:04+00:00 TAG 2.4.1 - - - - - 8e0cad5c by David Waern at 2008-12-07T14:12:54+00:00 Remove -Werror - - - - - 299d6deb by David Waern at 2008-12-07T14:25:18+00:00 Remove -Wall, we'll focus on warnings after 6.10.2 is out - - - - - 5f4216b6 by David Waern at 2008-12-07T20:58:05+00:00 Resolve conflict properly - - - - - 67d774e7 by Neil Mitchell at 2008-12-15T11:44:26+00:00 Make forall's in constructors explicit, i.e. data Foo = Foo {foo :: Eq a => a} - - - - - 61851792 by Neil Mitchell at 2008-12-18T15:39:39+00:00 Try and find a better package name than "main" for Hoogle, goes wrong when working on an executable rather than a library - - - - - 2fab8554 by David Waern at 2008-12-08T23:19:48+00:00 Make visible names from ExportItems Instead of a complicated calculation of visible names out of GHC's export items, we can get them straight out of the already calculated ExportItems. The ExportItems should represent exactly those items that are visible in an interface. If store all the exported sub-names in ExportDecl instead of only those with documentation, the calculation becomes very simple. So we do this change as well (should perhaps have been a separate patch). This should fix the problem with names from ghc-prim not appearing in the link environment. - - - - - 7caadd8c by Ian Lynagh at 2008-12-11T17:01:04+00:00 Wrap the GHC usage with defaultCleanupHandler This fixes a bug where haddock leaves /tmp/ghc* directories uncleaned. - - - - - 7c9fc9a5 by David Waern at 2009-01-02T21:38:27+00:00 Show re-exported names from external packages again This fixes GHC ticket 2746. In order to also link to the exported subordinate names of a declaration, we need to re-introduce the sub map in the .haddock files. - - - - - 119e4e05 by David Waern at 2009-01-06T23:34:17+00:00 Do not process boot modules We should of course not try to produce documentation for boot modules! The reason this has worked in the past is that the output of "real" modules overwrites the output of boot modules later in the process. However, this causes a subtle link environment problem. So let's get rid of this stupid behaviour. We avoid processing boot modules, but we continue to typecheck them. - - - - - c285b9d2 by David Waern at 2009-01-08T18:03:36+00:00 Export modules also when coming from external packages This seems to have regressed since a refactoring that was part of the 2.3.0 release. - - - - - 24031c17 by David Waern at 2009-01-10T15:26:26+00:00 Change version to 2.4.2 - no need to go to 2.5.0 - - - - - 864d1c3f by David Waern at 2009-01-10T15:35:20+00:00 Update tests to account for version number change - - - - - 524ba886 by David Waern at 2009-01-10T18:29:17+00:00 Add test for Template Haskell splicing - - - - - 05e6e003 by David Waern at 2009-01-10T19:35:42+00:00 Fix Trac haskell/haddock#68: Turn on compilation via C for Template Haskell packages We can't use HscNothing if we need to run code coming from modules inside the processed package during typechecking, which is the case for some packages using Template Haskell. This could be improved, to e.g. use HscInterpreted and HscNothing where possible, instead of using HscC for all modules in the package. - - - - - 2b2bafa1 by David Waern at 2009-01-10T20:22:25+00:00 Only use needsTemplateHaskell when compiling with GHC 6.10.2 or above - - - - - bedc3a93 by Ian Lynagh at 2009-01-11T14:58:41+00:00 Fix the location of INPLACE_PKG_CONF; fixes the build Spotted by Conal Elliott - - - - - 943107c8 by David Waern at 2009-01-20T19:27:39+00:00 Document H.I.Create.collectDocs better - - - - - c6252e37 by David Waern at 2009-01-20T19:29:51+00:00 Fix Trac haskell/haddock#59: TH-generated declarations disappearing This patch was contributed by Joachim Breitner (nomeata). - - - - - 3568a6af by David Waern at 2009-01-21T21:41:48+00:00 Do not indicate that a constructor argument is unboxed We only show the strictness annotation for an unboxed constructor argument. The fact that it is unboxed is an implementation detail and should not be part of the module interface. - - - - - 562a4523 by David Waern at 2009-01-22T18:53:49+00:00 Fix Trac haskell/haddock#50: do not attach docs to pragmas or other kinds of non-declarations We now filter out everything that is not a proper Haskell declaration before collecting the docs and attaching them to declarations. - - - - - 6fdf21c2 by David Waern at 2009-01-22T19:48:09+00:00 Add test for quasi quotation. No reference output yet. - - - - - dc4100fd by David Waern at 2009-01-22T19:57:47+00:00 Improve quasi-quotation test and add reference output - - - - - 908b74bb by David Waern at 2009-01-23T23:22:03+00:00 Filter out separately exported associated types in a smarter way - - - - - f6b42ecb by David Waern at 2009-01-24T16:54:39+00:00 Correct spelling mistake in error message - - - - - 24e4245d by David Waern at 2009-01-24T17:48:03+00:00 Correct comment - - - - - b5e8462f by David Waern at 2009-02-07T13:22:29+00:00 Do not show a subordinate at the top level if its parent is also exported See note in the source code for more info. - - - - - 4b09de57 by David Waern at 2009-02-07T13:53:53+00:00 Update test following change to top level subordinates - - - - - 76379896 by David Waern at 2009-02-07T13:58:04+00:00 Remove html files in the tests/output/ directory which have been accidentally added - - - - - 1a6d8b10 by Joachim Breitner at 2009-02-20T10:29:43+00:00 Typo in comment - - - - - fec367d0 by David Waern at 2009-02-24T20:21:17+00:00 Fix small bug The rule is to prefer type constructors to other things when an identifier in a doc string can refer to multiple things. This stopped working with newer GHC versions (due to a tiny change in the GHC renamer). We implement this rule in the HTML backend for now, instead of fixing it in GHC, since we will move renaming of doc strings to Haddock in the future anyway. - - - - - 9b4172eb by David Waern at 2009-02-25T20:04:38+00:00 Fix bad error handling with newer GHCs When support for GHC 6.10 was added, an error handler was installed only around the typechecking phase. This had the effect that errors thrown during dependency chasing were caught in the top-level exception handler and not printed with enough detail. With this patch we wrap the error handler around all our usage of the Ghc monad. - - - - - de2df363 by Simon Peyton Jones at 2009-02-02T16:47:42+00:00 Hide funTyConName, now exported by TypeRep - - - - - 4d40a29f by Ian Lynagh at 2009-02-12T18:57:49+00:00 Don't build the library when building in the GHC tree - - - - - 1cd0abe4 by Ian Lynagh at 2009-02-13T13:58:53+00:00 Add a ghc.mk - - - - - 3d814eeb by Ian Lynagh at 2009-02-13T18:50:28+00:00 do .depend generation for haddock with the stage1 compiler This is a bit of a hack. We mkdepend with stage1 as if .depend depends on the stage2 compiler then make goes wrong: haddock's .depend gets included, which means that make won't reload until it's built, but we can't build it without the stage2 compiler. We therefore build the stage2 compiler before its .depend file is available, and so compilation fails. - - - - - b55036a4 by Ian Lynagh at 2009-02-25T01:38:13+00:00 Give haddock a wrapper on unix in the new GHC build system - - - - - 9eabfe68 by Ian Lynagh at 2009-02-25T19:21:32+00:00 Create inplace/lib/html in the new GHC build system - - - - - 93af30c7 by Ian Lynagh at 2008-11-07T19:18:23+00:00 TAG GHC 6.10.1 release - - - - - 06e6e34a by Thomas Schilling at 2009-02-24T18:11:00+00:00 Define __GHC_PATCHLEVEL__ for recent version of GHC (stable). - - - - - 680e6ed8 by Thomas Schilling at 2009-02-24T18:12:26+00:00 'needsTemplateHaskell' is not defined in current stable GHC. - - - - - 6c5619df by David Waern at 2009-02-25T22:15:23+00:00 Hide fynTyConName only for recent GHC versions - - - - - 6b2344f1 by Ian Lynagh at 2009-02-26T00:49:56+00:00 Add the module to one of haddocks warnings - - - - - e5d11c70 by David Waern at 2009-02-27T21:37:20+00:00 Bug fix We tried to filter out subordinates that were already exported through their parent. This didn't work properly since we were in some cases looking at the grand-parent and not the parent. We now properly compute all the parent-child relations of a declaration, and use this information to get the parent of a subordinate. We also didn't consider record fields with multiple parents. This is now handled correctly. We don't currently support separately exported associated types. But when we do, they should be handled correctly by this process too. Also slightly improved the warning message that we give when filtering out subordinates. - - - - - 10a79a60 by David Waern at 2009-02-27T22:08:08+00:00 Fix error message conflict The module name is already written in the beginning of the message, as seems to be the convention in Haddock. Perhaps not so clear, but we should change it everywhere in that case. Leaving it as it is for now. - - - - - c5055c7f by David Waern at 2009-02-27T22:15:17+00:00 Shorten warning message - - - - - a72fed3a by David Waern at 2009-02-28T00:53:55+00:00 Do not show package name in warning message - - - - - a5daccb2 by Ian Lynagh at 2009-03-01T14:59:35+00:00 Install haddock in the new GHC build system - - - - - dfdb025c by Ian Lynagh at 2009-03-07T23:56:29+00:00 Relax base dependency to < 4.2, not < 4.1 - - - - - 5769c8b4 by David Waern at 2009-03-21T14:58:52+00:00 Bump .haddock file version number (due to change of format) - - - - - f1b8f67b by David Waern at 2009-03-21T14:59:26+00:00 Define __GHC_PATCHLEVEL__=1 when using ghc-6.10.1 - - - - - 23f78831 by David Waern at 2009-03-21T16:40:52+00:00 Update CHANGES - - - - - 7d2735e9 by David Waern at 2009-03-21T16:50:33+00:00 Update ANNOUNCE - - - - - 0771e00a by David Waern at 2009-03-21T16:54:40+00:00 Update ANNOUNCE, again - - - - - 81a6942a by David Waern at 2009-03-21T17:50:06+00:00 Don't be too verbose in CHANGES - - - - - 29861dcf by David Waern at 2009-03-21T18:03:31+00:00 TAG 2.4.2 - - - - - a585f285 by David Waern at 2009-03-21T19:20:29+00:00 Require Cabal >= 1.2.3 - - - - - 7c611662 by David Waern at 2009-03-21T19:21:48+00:00 TAG 2.4.2 with cabal-version >= 1.2.3 - - - - - 23b7deff by Simon Marlow at 2009-03-20T15:43:42+00:00 new GHC build system: use shell-wrappers macro - - - - - 25f8afe7 by Ian Lynagh at 2009-03-21T19:13:53+00:00 Fix (with a hack?) haddock in teh new build system - - - - - 6a29a37e by David Waern at 2009-03-24T22:10:15+00:00 Remove unnecessary LANGUAGE pragma - - - - - 954da57d by David Waern at 2009-03-24T22:21:23+00:00 Fix warnings in H.B.DevHelp - - - - - 1619f1df by David Waern at 2009-03-26T23:20:44+00:00 -Wall police in H.B.Html - - - - - b211e13b by Simon Marlow at 2009-03-24T13:00:56+00:00 install Haddock's html stuff - - - - - 78e0b107 by David Waern at 2008-12-07T19:58:53+00:00 Add verbosity flag and utils, remove "verbose" flag - - - - - 913dae06 by David Waern at 2008-12-07T20:01:05+00:00 Add some basic "verbose" mode logging in H.Interface - - - - - 1cbff3bf by David Waern at 2009-03-27T00:07:26+00:00 Fix conflicts - - - - - 22f82032 by David Waern at 2009-03-27T21:15:11+00:00 Remove H.GHC.Typecheck - - - - - 81557804 by David Waern at 2009-03-27T21:19:22+00:00 Remove docNameOrig and use getName everywhere instead - - - - - d8267213 by David Waern at 2009-03-27T21:21:46+00:00 Use docNameOcc instead of nameOccName . getName - - - - - 5d55deab by David Waern at 2009-03-27T21:33:04+00:00 Remove H.DocName and put DocName in H.Types - - - - - 8ba72611 by David Waern at 2009-03-27T22:06:26+00:00 Document DocName - - - - - 605f8ca5 by David Waern at 2009-03-27T22:45:21+00:00 -Wall police - - - - - e4da93ae by David Waern at 2009-03-27T23:12:53+00:00 -Wall police in H.B.Hoogle - - - - - bb255519 by David Waern at 2009-03-27T23:41:28+00:00 Define Foldable and Traversable instances for Located - - - - - f1195cfe by David Waern at 2009-03-27T23:51:34+00:00 Wibble - - - - - 23818d7c by David Waern at 2009-03-28T00:03:55+00:00 -Wall police in H.I.Rename - - - - - 0f050d67 by David Waern at 2009-03-28T00:15:15+00:00 -Wall police in H.I.AttachInstances - - - - - 0f3fe038 by David Waern at 2009-03-28T21:09:41+00:00 Wibble - - - - - 275d4865 by David Waern at 2009-03-28T21:27:06+00:00 Layout fix - - - - - 54ff0ef8 by David Waern at 2009-03-28T21:59:07+00:00 -Wall police in H.I.Create - - - - - 7f58b117 by David Waern at 2009-03-28T22:10:19+00:00 -Wall police in H.Interface - - - - - f0c03b44 by David Waern at 2009-03-28T22:22:59+00:00 -Wall police in Main - - - - - 29da355c by David Waern at 2009-03-28T22:23:39+00:00 Turn on -Wall -Werror - - - - - 446d3060 by David Waern at 2009-04-01T20:40:30+00:00 hlint police - - - - - 3867c9fc by David Waern at 2009-04-01T20:48:42+00:00 hlint police - - - - - bd1f1600 by David Waern at 2009-04-01T20:58:02+00:00 hlint police - - - - - e0e90866 by David Waern at 2009-04-05T12:42:53+00:00 Move H.GHC.Utils to H.GhcUtils - - - - - 9cbd426b by David Waern at 2009-04-05T12:57:21+00:00 Remove Haddock.GHC and move its (small) contents to Main - - - - - b5c2cbfd by David Waern at 2009-04-05T13:07:04+00:00 Fix whitespace and stylistic issues in Main - - - - - 3c04aa56 by porges at 2008-12-07T08:22:19+00:00 add unicode output - - - - - 607918da by David Waern at 2009-04-26T15:09:43+00:00 Resolve conflict - - - - - 4bec6b6b by Simon Marlow at 2009-05-13T10:00:31+00:00 fix markup - - - - - 436ad6f4 by Simon Marlow at 2009-03-23T11:54:45+00:00 clean up - - - - - bdcd1398 by Simon Marlow at 2009-03-24T10:36:45+00:00 new GHC build system: add $(exeext) - - - - - 9c0972f3 by Simon Marlow at 2009-03-24T11:04:31+00:00 update for new GHC build system layout - - - - - d0f3f83a by Ian Lynagh at 2009-03-29T15:31:43+00:00 GHC new build system fixes - - - - - 5a8245c2 by Ian Lynagh at 2009-04-04T20:44:23+00:00 Tweak new build system - - - - - 9c6f2d7b by Simon Marlow at 2009-05-13T10:01:27+00:00 add build instructions for GHC - - - - - 66d07c76 by Ian Lynagh at 2009-05-31T00:37:53+00:00 Quote program paths in ghc.mk - - - - - bb7de2cd by Ian Lynagh at 2009-06-03T22:57:55+00:00 Use a bang pattern on an unlifted binding - - - - - 3ad283fc by Ian Lynagh at 2009-06-13T16:17:50+00:00 Include haddock in GHC bindists - - - - - ac447ff4 by David Waern at 2009-06-24T21:07:50+00:00 Delete Haddock.Exception and move contents to Haddock.Types Only a few lines of code that mainly declares a type - why not just put it in Haddock.Types. - - - - - 4464fb9b by David Waern at 2009-06-24T22:23:23+00:00 Add Haddock module headers Add a proper Haddock module header to each module, with a more finegrained copyright. If you feel mis-accreditted, please correct any copyright notice! The maintainer field is set to haddock at projects.haskell.org. Next step is to add a brief description to each module. - - - - - 5f4c95dd by David Waern at 2009-06-24T22:39:44+00:00 Fix spelling error - - - - - 6d074cdb by David Waern at 2009-06-25T21:53:56+00:00 Document Interface and InstalledInterface better - - - - - d0cbd183 by David Waern at 2009-06-27T12:46:46+00:00 Remove misplaced whitespace in H.I.Rename - - - - - fa381c49 by David Waern at 2009-06-27T13:26:03+00:00 Fix haskell/haddock#104 - create output directory if missing - - - - - 91fb77ae by Ian Lynagh at 2009-06-25T15:59:50+00:00 TAG 2009-06-25 - - - - - 0d853f40 by Simon Peyton Jones at 2009-07-02T15:35:22+00:00 Follow extra field in ConDecl - - - - - b201735d by Ian Lynagh at 2009-07-05T16:50:35+00:00 Update Makefile for the new GHC build system - - - - - df6c0092 by Ian Lynagh at 2009-07-05T17:01:13+00:00 Resolve conflicts - - - - - 1066870a by Ian Lynagh at 2009-07-05T17:01:48+00:00 Remove the -Wwarn hack in the GHC build system - - - - - 7e856076 by Ian Lynagh at 2009-07-05T17:17:59+00:00 Fix warnings - - - - - 5d4cd958 by Ian Lynagh at 2009-07-05T19:35:40+00:00 Bump version number Cabal needs to distinguish between haddocks having a --verbose and --verbosity flag - - - - - 6ee07c99 by David Waern at 2009-07-06T20:14:57+00:00 Wibble - - - - - 2308b66f by David Waern at 2009-07-06T20:24:20+00:00 Clearer printing of versions by runtests.hs - - - - - d4b5d9ab by David Waern at 2009-07-06T21:22:42+00:00 Fix (invisible) bug introduced by unicode patch - - - - - 2caca8d8 by David Waern at 2009-07-06T21:44:10+00:00 Use HscAsm instead of HscC when using TH - - - - - 18f3b755 by David Waern at 2009-07-06T22:10:22+00:00 Update HCAR entry (by Janis) - - - - - a72ac9db by David Waern at 2009-07-06T23:01:35+00:00 Follow HsRecTy change with an #if __GLASGOW_HASKEL__ >= 611 - - - - - 549135d2 by David Waern at 2009-07-06T23:11:41+00:00 Remove unused functions from Haddock.Utils - - - - - b450134a by Isaac Dupree at 2009-07-11T14:59:00+00:00 revert to split-index for large indices - remove the search-box, because browsers have search-for-text abilities anyway. - pick 150 items in index as the arbitrary time at which to split it - notice the bug that identifiers starting with non-ASCII characters won't be listed in split-index, but don't bother to fix it yet (see ticket haskell/haddock#116, http://trac.haskell.org/haddock/ticket/116 ) - - - - - 78a5661e by Isaac Dupree at 2009-07-20T15:37:18+00:00 Implement GADT records in HTML backend - - - - - 4e163555 by Isaac Dupree at 2009-07-21T22:03:25+00:00 add test for GADT records - - - - - 79aa4d6e by David Waern at 2009-07-23T20:40:37+00:00 Update test suite following version bump - - - - - 5932c011 by David Waern at 2009-08-02T10:25:39+00:00 Fix documentation bug - - - - - a6970fca by David Waern at 2009-08-12T23:08:53+00:00 Remove support for ghc 6.8.* from .cabal file - - - - - c1695902 by Ian Lynagh at 2009-07-07T13:35:45+00:00 Fix unused import warnings - - - - - fb6df7f9 by Ian Lynagh at 2009-07-16T00:20:31+00:00 Use cProjectVersion directly rather than going through compilerInfo Fixes the build after changes in GHC - - - - - 548cdd66 by Simon Marlow at 2009-07-28T14:27:04+00:00 follow changes in GHC's ForeignType - - - - - 9395aaa0 by David Waern at 2009-08-13T22:17:33+00:00 Switch from PatternSignatures to ScopedTypeVariables in Main - - - - - eebf39bd by David Waern at 2009-08-14T17:14:28+00:00 Version .haddock files made with GHC 6.10.3/4 correclty - - - - - 58f3e735 by David Waern at 2009-08-14T17:19:37+00:00 Support GHC 6.10.* and 6.11.* only - - - - - 5f63cecc by David Waern at 2009-08-14T22:03:20+00:00 Do not version .haddock file based on GHC patchlevel version We require that the instances of Binary that we use from GHC will not change between patchlevel versions. - - - - - d519de9f by David Waern at 2009-08-14T23:50:00+00:00 Update CHANGES - - - - - 35dccf5c by David Waern at 2009-08-14T23:51:38+00:00 Update version number everywhere - - - - - 6d363fea by David Waern at 2009-08-15T09:46:49+00:00 Update ANNOUNCE - - - - - c7ee6bc2 by David Waern at 2009-08-15T09:47:13+00:00 Remove -Werror Forgot that Hackage doesn't like it. - - - - - a125c12b by David Waern at 2009-08-15T09:49:50+00:00 Require Cabal >= 1.6 - - - - - adb2f560 by Isaac Dupree at 2009-08-12T03:47:14+00:00 Cross-Package Documentation version 4 - - - - - 3d6dc04d by David Waern at 2009-08-15T23:42:57+00:00 Put all the IN_GHC_TREE stuff inside getGhcLibDir - - - - - 56624097 by David Waern at 2009-08-15T23:52:03+00:00 Add --print-ghc-libdir - - - - - f15d3ccb by David Waern at 2009-08-16T00:37:52+00:00 Read base.haddock when running tests We can now test cross-package docs. - - - - - 283f0fb9 by David Waern at 2009-08-16T00:50:59+00:00 Update test output - we now have more links - - - - - 673d1004 by David Waern at 2009-08-16T01:26:08+00:00 Read process.haddock when running tests - - - - - 0d127f82 by David Waern at 2009-08-16T01:43:04+00:00 Add a test for cross-package documentation - - - - - f94db967 by Ian Lynagh at 2009-08-16T18:42:44+00:00 Follow GHC build system changes - - - - - 5151278a by Isaac Dupree at 2009-08-16T19:58:05+00:00 make cross-package list types look nicer - - - - - c41e8228 by Isaac Dupree at 2009-08-18T01:47:47+00:00 Haddock.Convert: export more functions This lets us remove some code in Haddock.Interface.AttachInstances - - - - - 2e5fa398 by Isaac Dupree at 2009-08-18T02:11:05+00:00 switch AttachInstances to use synify code It changed an instance from showing ((,) a b) to (a, b) because my synify code is more sophisticated; I hope the latter is a good thing rather than a bad thing aesthetically, here. But this definitely reduces code duplication! - - - - - b8b07123 by Isaac Dupree at 2009-08-18T02:23:31+00:00 Find instances using GHC, which is more complete. In particular, it works cross-package. An intermediate patch also moved the instance-finding into createInterface, but that move turned out not to be necessary, so if we want to do that, it'd go in a separate patch. (Is that possible? Or will we need GHC to have loaded all the modules first, before we can go searching for the instances (e.g. if the modules are recursive or something)?) - - - - - 6959b451 by Isaac Dupree at 2009-08-17T00:37:18+00:00 fix preprocessor conditional sense - - - - - 942823af by Isaac Dupree at 2009-08-16T22:46:48+00:00 remove ghc 6.8 conditionals from Haddock.Interface - - - - - 4b3ad888 by Isaac Dupree at 2009-08-18T20:24:38+00:00 Fix GHC 6.11 build in Haddock.Convert - - - - - 0a89c5ab by Isaac Dupree at 2009-08-23T00:08:58+00:00 hacks to make it compile without fnArgDocsn - - - - - 7b3bed43 by Isaac Dupree at 2009-08-23T03:01:28+00:00 less big-Map-based proper extraction of constructor subdocs - - - - - b21c279a by Isaac Dupree at 2009-08-23T03:02:06+00:00 Html: remove unnecessary+troublesome GHC. qualifications - - - - - 96c97115 by Isaac Dupree at 2009-08-23T03:08:03+00:00 Move doc parsing/lexing into Haddock for ghc>=6.11 - - - - - e1cec02d by Isaac Dupree at 2009-08-23T05:08:14+00:00 get rid of unused DocMap parameter in Html - - - - - 66960c59 by Isaac Dupree at 2009-08-23T05:54:20+00:00 fix horrible named-docs-disappearing bug :-) - - - - - a9d7eff3 by Isaac Dupree at 2009-08-23T06:26:36+00:00 re-implement function-argument docs ..on top of the lexParseRn work. This patch doesn't change the InstalledInterface format, and thus, it does not work cross-package, but that will be easy to add subsequently. - - - - - 8bf6852c by Isaac Dupree at 2009-08-23T07:26:05+00:00 cross-package fnArgDocs. WARNING: changes .haddock binary format While breaking the format, I took the opportunity to unrename the DocMap that's saved to disk, because there's really no reason that we want to know what *another* package's favorite place to link a Name to was. (Is that true? Or might we want to know, someday?) Also, I added instance Binary Map in InterfaceFile. It makes the code a little simpler without changing anything of substance. Also it lets us add another Map hidden inside another Map (fnArgsDocs in instDocMap) without having really-convoluted serialization code. Instances are neat! I don't understand why this change to InterfaceFile seemed to subtly break binary compatibility all by itself, but no matter, I'll just roll it into the greater format-changing patch. Done! - - - - - 30115a64 by Isaac Dupree at 2009-08-23T18:22:47+00:00 Improve behavior for unfindable .haddock - - - - - aa364bda by Isaac Dupree at 2009-08-23T18:28:16+00:00 add comment for FnArgsDoc type - - - - - 49b23a99 by Isaac Dupree at 2009-08-23T21:52:48+00:00 bugfix: restore fnArgDocs for type-synonyms - - - - - f65f9467 by Isaac Dupree at 2009-08-23T22:06:55+00:00 Backends.Hoogle: eliminate warnings - - - - - a292d216 by Isaac Dupree at 2009-08-23T22:10:24+00:00 Haddock.Convert: eliminate warnings - - - - - 5546cd20 by Isaac Dupree at 2009-08-23T22:12:31+00:00 Haddock.Interface.Rename: eliminate warnings - - - - - 0a9798b6 by Isaac Dupree at 2009-08-23T22:18:47+00:00 Main.hs: remove ghc<6.9 conditionals - - - - - e8f9867f by Isaac Dupree at 2009-08-23T22:27:46+00:00 Main.hs: eliminate warnings (except for OldException) - - - - - 61c64247 by Isaac Dupree at 2009-08-23T22:41:01+00:00 move get*LibDir code in Main.hs, to +consistent code, -duplication - - - - - 948f1e69 by Isaac Dupree at 2009-08-23T23:14:26+00:00 Main.hs: OldException->Exception: which eliminates warnings - - - - - 3d5d5e03 by Isaac Dupree at 2009-08-23T23:20:11+00:00 GhcUtils: ghc >= 6.10 - - - - - 2771d657 by Isaac Dupree at 2009-08-23T23:21:55+00:00 InterfaceFile: ghc >= 6.10 - - - - - d9f2b9d1 by Isaac Dupree at 2009-08-23T23:22:58+00:00 Types: ghc >= 6.10 - - - - - ca39210e by Isaac Dupree at 2009-08-23T23:23:26+00:00 ModuleTree: ghc >= 6.10 - - - - - 883c4e59 by Isaac Dupree at 2009-08-23T23:24:04+00:00 Backends.DevHelp: ghc >= 6.10 - - - - - 04667df5 by Isaac Dupree at 2009-08-23T23:24:37+00:00 Backends.Html: ghc >= 6.10 - - - - - a9f7f25f by Isaac Dupree at 2009-08-23T23:25:24+00:00 Utils: ghc >= 6.10 - - - - - b7105022 by Isaac Dupree at 2009-08-23T23:37:47+00:00 eliminate haskell98 dependency, following GHC's example It turns out I/we already had, and it was only a matter of deleting it from the cabal file. - - - - - 292e0911 by Isaac Dupree at 2009-08-24T01:22:44+00:00 refactor out subordinatesWithNoDocs dep of inferenced-decls fix - - - - - c2ed46a2 by Isaac Dupree at 2009-08-24T01:24:03+00:00 Eradicate wrong runtime warning for type-inferenced exported-functions see the long comment in the patch for why I did it this way :-) - - - - - 4ac0b57c by David Waern at 2009-09-04T22:56:20+00:00 Clean up tyThingToHsSynSig a little Factor out noLoc and use the case construct. Also rename the function to tyThingToLHsDecl, since it doesn't just create type signatures. - - - - - 28ab9201 by David Waern at 2009-09-04T22:58:50+00:00 Wibble - - - - - 0d9fe6d0 by David Waern at 2009-09-06T18:39:30+00:00 Add more copyright owners to H.I.AttachInstances - - - - - 122441b1 by David Waern at 2009-09-06T18:44:12+00:00 Style police - - - - - 1fa79463 by David Waern at 2009-09-06T18:57:45+00:00 Move toHsInstHead to Haddock.Convert and call it synifyInstHead - - - - - 0d42a8aa by David Waern at 2009-09-06T21:11:38+00:00 Use colordiff to display test results if available - - - - - ea9d8e03 by Simon Marlow at 2009-08-24T08:46:14+00:00 Follow changes in GHC's interface file format Word32 instead of Int for FastString and Name offsets - - - - - 537e051e by Simon Marlow at 2009-07-29T14:16:53+00:00 define unpackPackageId (it was removed from GHC) - - - - - 50c63aa7 by David Waern at 2009-09-09T23:18:03+00:00 Remove commented-out code - - - - - 511631fe by David Waern at 2009-09-09T23:19:05+00:00 Correct copyright in H.I.ParseModuleHeader - - - - - 898ec768 by David Waern at 2009-09-11T11:22:29+00:00 Use Map.fromList/toList intead of fromAscList/toAscList when serializing Maps This fixes the missing docs problem. The Eq and Ord instances for Name uses the unique number in Name. This number is created at deserialization time by GHC's magic Binary instance for Name, and it is random. Thus, fromAscList can't be used at deserialization time, even though toAscList was used at serialization time. - - - - - 37bec0d5 by Simon Peyton Jones at 2009-09-11T08:28:04+00:00 Track change in HsType - - - - - eb3a97c3 by Ian Lynagh at 2009-09-11T16:07:09+00:00 Allow building with base 4.2 - - - - - bb4205ed by Ian Lynagh at 2009-09-22T13:50:02+00:00 Loosen the GHC dependency - - - - - 5c75deb2 by Ian Lynagh at 2009-09-22T14:08:39+00:00 Fix building with GHC >= 6.12 - - - - - fb131481 by David Waern at 2009-09-11T11:24:48+00:00 Update runtests.hs to work with GHC 6.11 - - - - - ac3a419d by David Waern at 2009-09-11T11:25:14+00:00 Update CrossPackageDocs test - - - - - ec65c3c6 by David Waern at 2009-09-11T11:25:40+00:00 Add reference output for CrossPackageDocs - - - - - 520c2758 by Ian Lynagh at 2009-10-25T17:26:40+00:00 Fix installation in the GHC build system - - - - - 28b3d7df by Ian Lynagh at 2009-11-05T15:57:27+00:00 GHC build system: Make *nix installation work in paths containing spaces - - - - - 5c9bb541 by David Waern at 2009-11-14T11:56:39+00:00 Track change in HsType for the right compiler version - - - - - 905097ce by David Waern at 2009-11-14T12:10:47+00:00 hlint police - - - - - 04920630 by Ian Lynagh at 2009-11-20T13:46:30+00:00 Use defaultObjectTarget rather than HscAsm This fixes haddock when we don't have a native code generator - - - - - 966eb079 by David Waern at 2009-11-15T12:32:21+00:00 Remove commented-out code - - - - - 37f00fc4 by David Waern at 2009-11-22T13:58:48+00:00 Make runtests.hs strip links before diffing Generates easier to read diffs when tests fail. The content of the links is not important anyway since it is not taken into account by the tests. - - - - - 3a9bb8ef by David Waern at 2009-11-22T14:05:06+00:00 Follow findProgramOnPath signature change in runtests.hs - - - - - b26b9e5a by David Waern at 2009-11-22T14:08:40+00:00 Follow removal of GHC.MVar from base in CrossPackageDocs - - - - - f4d90ae4 by David Waern at 2009-11-22T14:48:47+00:00 Make copy.hs strip link contents before copying No more updating of reference files when URLs in links changes. - - - - - 4c9c420d by David Waern at 2009-11-22T15:26:41+00:00 Update test reference output * More links (Int, Float etc) * Stripped link contents - - - - - a62b80e3 by David Waern at 2009-11-23T23:19:39+00:00 Update CrossPackageDocs reference output - Remove GHC.MVar import (removed from base) - Strip link contents - - - - - 43491394 by David Waern at 2009-11-23T23:20:00+00:00 Update test reference files with comments on instances - - - - - 0d370a0b by David Waern at 2009-11-23T23:25:16+00:00 Bump version number - - - - - 2293113e by David Waern at 2009-11-24T20:55:49+00:00 Comments on instances Implementing this was a little trickier than I thought, since we need to match up instances from the renamed syntax with instances represented by InstEnv.Instance. This is due to the current design of Haddock, which matches comments with declarations from the renamed syntax, while getting the list of instances of a class/family directly using the GHC API. - Works for class instances only (Haddock has no support for type family instances yet) - The comments are rendered to the right of the instance head in the HTML output - No change to the .haddock file format - Works for normal user-written instances only. No comments are added on derived or TH-generated instances - - - - - bf586f29 by David Waern at 2009-11-27T22:05:15+00:00 Whitespace police - - - - - b8f03afa by David Waern at 2009-11-27T22:11:46+00:00 Remove bad whitespace and commented-out pieces - - - - - 90b8ee90 by David Waern at 2009-11-27T22:15:04+00:00 Whitespace police - - - - - b5ede900 by David Waern at 2009-11-27T22:15:50+00:00 Whitespace police - - - - - e3fddbfe by David Waern at 2009-11-28T13:37:59+00:00 Remove Name from DocInstance It's not used. - - - - - 9502786c by David Waern at 2009-11-28T13:56:54+00:00 Require at least GHC 6.12 While regression testing Haddock, I found a bug that happens with GHC 6.10.3, but not with GHC 6.12-rc2 (haven't tried 6.10.4). I don't have time to track it down. I think we should just always require the latest major GHC version. The time spent on making Haddock work with older versions is too high compared to the time spent on bugfixing, refactoring and features. - - - - - 8fa688d8 by David Waern at 2009-11-28T15:05:03+00:00 Remove cruft due to compatibility with older GHCs - - - - - 46fbbe9d by David Waern at 2009-11-28T15:07:50+00:00 Add a documentation header to Haddock.Convert - - - - - c3d2cc4a by David Waern at 2009-11-28T15:10:14+00:00 Remove unused H.Utils.FastMutInt2 - - - - - 490aba80 by David Waern at 2009-11-28T15:36:36+00:00 Rename Distribution.Haddock into Documentation.Haddock - - - - - 33ee2397 by David Waern at 2009-11-28T15:36:47+00:00 Fix error message - - - - - a5a3b950 by David Waern at 2009-11-28T16:58:39+00:00 Add a test flag that brings in QuickCheck - - - - - fa049e13 by David Waern at 2009-11-28T19:32:18+00:00 Say that we want quickcheck 2 - - - - - f32b0d9b by David Waern at 2009-11-28T19:32:40+00:00 Add an Arbitrary instance for HsDoc - - - - - da9a8bd7 by David Waern at 2009-11-28T20:15:30+00:00 Rename HsDoc back into Doc - - - - - edb60101 by David Waern at 2009-11-28T22:16:16+00:00 Move H.Interface.Parse/Lex to H.Parse/Lex These are not just used to build Interfaces. - - - - - 0656a9b8 by David Waern at 2009-11-28T23:12:14+00:00 Update version number in test suite - - - - - 5e8c6f4a by David Waern at 2009-12-21T14:12:41+00:00 Improve doc of DocName - - - - - 7868e551 by Ian Lynagh at 2009-09-22T10:43:03+00:00 TAG GHC 6.12-branch created - - - - - 0452a3ea by Ian Lynagh at 2009-12-15T12:46:07+00:00 TAG GHC 6.12.1 release - - - - - 65e9be62 by David Waern at 2009-12-21T16:58:58+00:00 Update CHANGES - - - - - 145cee32 by David Waern at 2009-12-21T16:59:09+00:00 TAG 2.6.0 - - - - - 3c552008 by David Waern at 2009-12-22T17:11:14+00:00 Update ANNOUNCE - - - - - 931f9db4 by David Waern at 2010-01-22T19:57:17+00:00 Convert haddock.vim to use unix newlines - - - - - 4e56588f by David Waern at 2010-01-22T22:11:17+00:00 Remove unnecessary (and inexplicable) uses of nub - - - - - 744bb4d1 by David Waern at 2010-01-22T22:12:14+00:00 Follow move of parser and lexer - - - - - e34bab14 by David Waern at 2010-01-22T22:49:13+00:00 Use findProgramLocation instead of findProgramOnPath in runtests.hs - - - - - 8d39891b by Isaac Dupree at 2010-01-14T18:53:18+00:00 fix html arg-doc off-by-one and silliness - - - - - 9401f2e9 by David Waern at 2010-01-22T22:57:03+00:00 Create a test for function argument docs - - - - - 507a82d7 by David Waern at 2010-01-22T23:24:47+00:00 Put parenthesis around type signature arguments of function type - - - - - 8a305c28 by David Waern at 2010-01-23T17:26:59+00:00 Add reference file for the FunArgs test - - - - - 1309d5e1 by David Waern at 2010-01-24T16:05:08+00:00 Improve FunArg test and update Test.html.ref - - - - - 2990f055 by Yitzchak Gale at 2010-02-14T16:03:46+00:00 Do not generate illegal character in HTML ID attribute. - - - - - c5bcab7a by David Waern at 2010-02-22T22:10:30+00:00 Fix Haddock markup error in comment - - - - - c6416a73 by David Waern at 2010-02-24T22:55:08+00:00 Large additions to the Haddock API Also improved and added more doc comments. - - - - - 57d289d7 by David Waern at 2010-02-24T22:58:02+00:00 Remove unused ifaceLocals - - - - - 80528d93 by David Waern at 2010-02-25T21:05:09+00:00 Add HaddockModInfo to the API - - - - - 82806848 by David Waern at 2010-02-25T21:05:27+00:00 Wibble - - - - - 744cad4c by David Waern at 2010-02-25T23:30:59+00:00 Make it possible to run a single test - - - - - 6a806e4c by David Waern at 2010-03-14T14:19:39+00:00 Bump version number - - - - - a5a8e4a7 by David Waern at 2010-03-14T14:36:35+00:00 Update ANNOUNCE - - - - - 6f05435e by Simon Hengel at 2010-03-15T20:52:42+00:00 Add missing dependencies for 'library' in haddock.cabal - - - - - faefe2bd by David Waern at 2010-03-15T22:29:37+00:00 Solve conflicts - - - - - 9808ad52 by David Waern at 2010-03-15T22:51:21+00:00 Bump version number - - - - - eb0bf60b by David Waern at 2010-03-15T22:52:32+00:00 Update CHANGES - - - - - f95cd891 by David Waern at 2010-03-15T23:01:06+00:00 Add Paths_haddock to other-modules of library - - - - - 65997b0a by David Waern at 2010-03-15T23:14:59+00:00 Update CHANGES - - - - - 7e251731 by David Waern at 2010-03-15T23:15:30+00:00 Bump version number - - - - - c9cd0ddc by David Waern at 2010-03-16T00:28:34+00:00 Fix warning - - - - - 1cac2d93 by Simon Peyton Jones at 2010-01-04T15:22:16+00:00 Fix imports for new location of splitKindFunTys - - - - - 474f26f6 by Simon Peyton Jones at 2010-02-10T14:36:06+00:00 Update Haddock for quasiquotes - - - - - 0dcc06c0 by Simon Peyton Jones at 2010-02-10T10:59:45+00:00 Track changes in HsTyVarBndr - - - - - 2d84733a by Simon Peyton Jones at 2010-02-10T14:52:44+00:00 Track HsSyn chnages - - - - - 9e3adb8b by Ian Lynagh at 2010-02-20T17:09:42+00:00 Resolve conflicts - - - - - a3e72ff8 by Simon Peyton Jones at 2010-03-04T13:05:16+00:00 Track change in HsUtils; and use a nicer function not an internal one - - - - - 27994854 by David Waern at 2010-03-18T22:22:27+00:00 Fix build with GHC 6.12.1 - - - - - 11f6e488 by David Waern at 2010-03-18T22:24:09+00:00 Bump version in test reference files - - - - - 0ef2f11b by David Waern at 2010-03-20T00:56:30+00:00 Fix library part of cabal file when in ghc tree - - - - - 3f6146ff by Mark Lentczner at 2010-03-20T22:30:11+00:00 First, experimental XHTML rendering switch to using the xhtml package copied Html.hs to Xhtml.hs and split into sub-modules under Haddock/Backends/Xhtml and detabify moved footer into div, got ready for iface change headers converted to semantic markup contents in semantic markup summary as semantic markup description in semantic markup, info block in header fixed factored out rendering so during debug it can be readable (see renderToString) - - - - - b8ab329b by Mark Lentczner at 2010-03-20T22:54:01+00:00 apply changes to Html.hs to Xhtml/*.hs incorporate changes that were made between the time Html.hs was copied and split into Xhtml.hs and Xhtml/*.hs includes patchs after "Wibble" (!) through "Fix build with GHC 6.12.1" - - - - - 73df2433 by Ian Lynagh at 2010-03-20T21:56:37+00:00 Follow LazyUniqFM->UniqFM in GHC - - - - - db4f602b by David Waern at 2010-03-29T22:00:01+00:00 Fix build with GHC 6.12 - - - - - d8dca088 by Simon Hengel at 2010-04-02T16:39:55+00:00 Add missing dependencies to cabal file - - - - - e2adc437 by Simon Hengel at 2010-04-02T14:08:40+00:00 Add markup support for interactive examples - - - - - e882ac05 by Simon Hengel at 2010-04-02T14:11:53+00:00 Add tests for interactive examples - - - - - 5a07a6d3 by David Waern at 2010-04-07T17:05:20+00:00 Propagate source positions from Lex.x to Parse.y - - - - - 6493b46f by David Waern at 2010-04-07T21:48:57+00:00 Let runtests.hs die when haddock has not been built - - - - - 5e34423e by David Waern at 2010-04-07T22:01:13+00:00 Make runtests.hs slightly more readable - - - - - 321d59b3 by David Waern at 2010-04-07T22:13:27+00:00 Fix haskell/haddock#75 Add colons to the $ident character set. - - - - - 37b08b8d by David Waern at 2010-04-08T00:32:52+00:00 Fix haskell/haddock#118 Avoid being too greedy when lexing URL markup (<..>), in order to allow multiple URLs on the same line. Do the same thing with <<..>> and #..#. - - - - - df8feac9 by David Waern at 2010-04-08T00:57:33+00:00 Make it easier to add new package deps to test suite This is a hack - we should use Cabal to get the package details instead. - - - - - 1ca6f84b by David Waern at 2010-04-08T01:03:06+00:00 Add ghc-prim to test suite deps - - - - - 27371e3a by Simon Hengel at 2010-04-08T19:26:34+00:00 Let parsing fails on paragraphs that are immediately followed by an example This is more consistent with the way we treat code blocks. - - - - - 83096e4a by David Waern at 2010-04-08T21:20:00+00:00 Improve function name - - - - - 439983ce by David Waern at 2010-04-10T10:46:14+00:00 Fix haskell/haddock#112 No link was generated for 'Addr#' in a doc comment. The reason was simply that the identifier didn't parse. We were using parseIdentifier from the GHC API, with a parser state built from 'defaultDynFlags'. If we pass the dynflags of the module instead, the right options are turned on on while parsing the identifer (in this case -XMagicHash), and the parse succeeds. - - - - - 5c0d35d7 by David Waern at 2010-04-10T10:54:06+00:00 Rename startGhc into withGhc - - - - - dca081fa by Simon Hengel at 2010-04-12T19:09:16+00:00 Add documentation for interactive examples - - - - - c7f26bfa by David Waern at 2010-04-13T00:51:51+00:00 Slight fix to the documentation of examples - - - - - 06eb7c4c by David Waern at 2010-04-13T00:57:05+00:00 Rename Interactive Examples into Examples (and simplify explanation) - - - - - 264830cb by David Waern at 2010-05-10T20:07:27+00:00 Update CHANGES with info about 2.6.1 - - - - - 8e5d4514 by Simon Hengel at 2010-04-18T18:16:54+00:00 Add unit tests for parser - - - - - 68297f40 by David Waern at 2010-05-10T21:53:37+00:00 Improve testsuite README - - - - - f04eb6e4 by David Waern at 2010-05-11T19:14:31+00:00 Re-organise the testsuite structure - - - - - a360f710 by David Waern at 2010-05-11T19:18:03+00:00 Shorten function name - - - - - 1d5dd359 by David Waern at 2010-05-11T21:40:02+00:00 Update runtests.hs following testsuite re-organisation - - - - - ffebe217 by David Waern at 2010-05-11T21:40:10+00:00 Update runtests.hs to use base-4.2.0.1 - - - - - 635de402 by David Waern at 2010-05-11T21:41:11+00:00 Update runparsetests.hs following testsuite reorganisation - - - - - 72137910 by Ian Lynagh at 2010-05-06T20:43:06+00:00 Fix build - - - - - 1a80b76e by Ian Lynagh at 2010-05-06T22:25:29+00:00 Remove redundant import - - - - - 1031a80c by Simon Peyton Jones at 2010-05-07T13:21:09+00:00 Minor wibbles to HsBang stuff - - - - - dd8e7fe5 by Ian Lynagh at 2010-05-08T15:22:00+00:00 GHC build system: Follow "rm" variable changes - - - - - 7f5e6748 by David Waern at 2010-05-13T11:53:02+00:00 Fix build with GHC 6.12.2 - - - - - 7953d4d8 by David Waern at 2010-05-13T18:45:01+00:00 Fixes to comments only - - - - - 8ae8eb64 by David Waern at 2010-05-13T18:57:26+00:00 ModuleMap -> IfaceMap - - - - - 1c3eadc6 by David Waern at 2010-05-13T19:03:13+00:00 Fix whitespace style issues - - - - - e96783c0 by David Waern at 2010-05-13T19:08:53+00:00 Fix comment - - - - - c998a78b by David Waern at 2010-05-13T19:39:00+00:00 Position the module header the same way everywhere Silly, but nice with some consistency :-) - - - - - b48a714e by David Waern at 2010-05-13T19:41:32+00:00 Position of module header, this time in the HTML backends - - - - - f9bfb12e by David Waern at 2010-05-13T19:43:05+00:00 Two newlines between declarations in Main - - - - - 071d44c7 by David Waern at 2010-05-13T19:44:21+00:00 Newlines in Convert - - - - - 036346db by David Waern at 2010-05-13T19:46:47+00:00 Fix a few stylistic issues in H.InterfaceFile - - - - - f0b8379e by David Waern at 2010-05-13T19:47:53+00:00 Add newlines to H.ModuleTree - - - - - 27409f8e by David Waern at 2010-05-13T19:51:10+00:00 Fix stylistic issues in H.Utils - - - - - 24774a11 by David Waern at 2010-05-13T20:00:43+00:00 Structure H.Types better - - - - - 7b6f5e40 by David Waern at 2010-05-13T20:01:04+00:00 Remove bad Arbitrary instance - - - - - fac9f1f6 by David Waern at 2010-05-13T20:05:50+00:00 Get rid of H.Utils.pathJoin and use System.FilePath.joinPath instead - - - - - fe6d00c4 by David Waern at 2010-05-13T20:51:55+00:00 Export a couple of more types from the API - - - - - b2e33a5f by David Waern at 2010-05-13T21:27:51+00:00 Improve doc comment for Interface - - - - - c585f2ce by David Waern at 2010-05-13T21:30:14+00:00 Improve documentation of Haddock.Interface - - - - - e6791db2 by David Waern at 2010-05-13T22:07:35+00:00 Remove meaningless comments - - - - - 7801b390 by David Waern at 2010-05-14T17:53:33+00:00 Remove unused modules - - - - - f813e937 by David Waern at 2010-05-14T17:55:17+00:00 Re-direct compilation output to a temporary directory Also add a flag --no-tmp-comp-dir that can be used to get the old behaviour of writing compilation files to GHC's output directory (default "."). - - - - - e56737ec by David Waern at 2010-05-14T18:06:11+00:00 Wibble - - - - - e40b0447 by David Waern at 2010-05-14T19:01:52+00:00 Move flag evaluation code from Main to Haddock.Options Determining the value of "singular" flags (by e.g. taking the last occurrence of the flag) and other flag evaluation should done in Haddock.Options which is the module that is supposed to define the command line interface. This makes Main a bit easier on the eyes as well. - - - - - 27091f57 by David Waern at 2010-05-14T19:05:10+00:00 Wibble - - - - - c658cf61 by David Waern at 2010-05-14T19:06:49+00:00 Re-order things in Haddock.Options a bit - - - - - 8cfdd342 by David Waern at 2010-05-14T19:20:29+00:00 De-tabify Haddock.Options and fix other whitespace issues - - - - - 0df16b62 by David Waern at 2010-05-14T19:25:07+00:00 Improve comments - - - - - 80b38e2b by David Waern at 2010-05-14T19:26:42+00:00 Whitespace police - - - - - fe580255 by David Waern at 2010-05-14T19:31:23+00:00 Wibbles to comments - - - - - a2b43fad by David Waern at 2010-05-14T20:24:32+00:00 Move some more flag functions to Haddock.Options - - - - - 3f895547 by David Waern at 2010-05-14T20:37:12+00:00 Make renderStep a top-level function in Main - - - - - 5cdca11d by David Waern at 2010-05-14T20:39:27+00:00 Spelling in comment - - - - - ad98d14c by David Waern at 2010-05-14T20:40:26+00:00 Comment fixes - - - - - 0bb9218f by David Waern at 2010-05-14T20:49:01+00:00 Whitespace police - - - - - 0f0a533f by David Waern at 2010-05-15T16:42:29+00:00 Improve description of --dump-interface - - - - - 5b2833ac by David Waern at 2010-05-15T17:16:53+00:00 Document --no-tmp-comp-dir - - - - - 8160b170 by David Waern at 2010-05-15T17:18:59+00:00 Wibble - - - - - 570dbe33 by David Waern at 2010-05-18T21:15:38+00:00 HLint police - - - - - 204e425f by David Waern at 2010-05-18T21:16:30+00:00 HLint police - - - - - 6db657ac by David Waern at 2010-05-18T21:16:37+00:00 Wibble - - - - - b942ccd7 by Simon Marlow at 2010-06-02T08:27:30+00:00 Interrupted disappeared in GHC 6.13 (GHC ticket haskell/haddock#4100) - - - - - 3b94a819 by Simon Marlow at 2010-06-02T08:45:08+00:00 Allow base-4.3 - - - - - c5a1fb7c by Simon Marlow at 2010-06-02T09:03:04+00:00 Fix compilation with GHC 6.13 - - - - - 6181296c by David Waern at 2010-06-08T21:09:05+00:00 Display name of prologue file when parsing it fails - - - - - 7cbc6f60 by Ian Lynagh at 2010-06-13T16:20:25+00:00 Remove redundant imports - - - - - 980c804b by Simon Marlow at 2010-06-22T08:41:50+00:00 isLocalAndTypeInferenced: fix for local module names overlapping package modules - - - - - d74d4a12 by Simon Marlow at 2010-06-23T12:03:27+00:00 Unresolved identifiers in Doc get replaced with DocMonospaced rather than plain strings - - - - - d8546783 by Simon Marlow at 2010-06-30T12:45:17+00:00 LaTeX backend (new options: --latex, --latex-style=<style>) - - - - - 437afa9e by David Waern at 2010-07-01T12:02:44+00:00 Fix a few stylistic whitespace issues in LaTeX backend - - - - - 85bc1fae by David Waern at 2010-07-01T15:42:45+00:00 Make runtest.hs work with GHC 6.12.3 (we should really stop hard coding this) - - - - - 7d2eb86f by David Waern at 2010-07-01T15:43:33+00:00 Update test following Simon's patch to render unresolved names in monospaced font - - - - - 08fcbcd2 by David Waern at 2010-07-01T16:12:18+00:00 Warning police - - - - - d04a8d7a by David Waern at 2010-07-04T14:53:39+00:00 Fix a bug in attachInstances We didn't look for instance docs in all the interfaces of the package. This had the effect of instance docs not always showing up under a declaration. I took the opportunity to clean up the code in H.I.AttachInstances a bit as well. More cleanup is needed, however. - - - - - d10344eb by Simon Hengel at 2010-07-10T09:19:04+00:00 Add missing dependencies to cabal file - - - - - 24090531 by Mark Lentczner at 2010-03-21T04:51:16+00:00 add exports to Xhtml modules - - - - - 84f9a333 by Mark Lentczner at 2010-04-03T19:14:22+00:00 clean up Doc formatting code - add CSS for lists - renderToString now uses showHtml since prettyHtml messes up <pre> sections - - - - - bebccf52 by Mark Lentczner at 2010-04-04T04:51:08+00:00 tweak list css - - - - - 0c2aeb5e by Mark Lentczner at 2010-04-04T06:24:14+00:00 all decls now generate Html not HtmlTable - ppDecl return Html, and so now do all of the functions it calls - added some internal tables to some decls, which is wrong, and will have to be fixed - decl "Box" functions became "Elem" functions to make clear they aren't in a table anymore (see Layout.hs) - docBox went away, as only used in one place (and its days are numbered) - cleaned up logic in a number of places, removed dead code - added maybeDocToHtml which simplified a number of places in the code - - - - - dbf73e6e by Mark Lentczner at 2010-04-05T05:02:43+00:00 clean up processExport and place a div around each decl - - - - - e25b7e9f by Mark Lentczner at 2010-04-10T21:23:21+00:00 data decls are now a sequence of paragraphs, not a table - - - - - 89ee0294 by Mark Lentczner at 2010-04-10T21:29:16+00:00 removed commented out code that can't be maintained - - - - - d466f536 by Mark Lentczner at 2010-04-12T04:56:27+00:00 removed declWithDoc and cleaned up data decls in summary - - - - - ed755832 by Mark Lentczner at 2010-04-12T05:07:53+00:00 merge in markupExample changes - - - - - c36f51fd by Mark Lentczner at 2010-04-25T04:56:37+00:00 made record fields be an unordList, not a table - - - - - ed3a28d6 by Mark Lentczner at 2010-04-25T05:23:28+00:00 fixed surround of instance and constructor tables - - - - - 0e35bbc4 by Mark Lentczner at 2010-04-25T05:36:59+00:00 fix class member boxes in summary - - - - - 5041749b by Mark Lentczner at 2010-04-25T05:38:35+00:00 remove unused bodyBox - - - - - e91724db by Mark Lentczner at 2010-04-25T06:26:10+00:00 fixed javascript quoting/escpaing issue - - - - - f4abbb73 by Mark Lentczner at 2010-05-03T23:04:31+00:00 adjust css for current markup - - - - - e75fec4c by Mark Lentczner at 2010-05-04T06:14:34+00:00 added assoicated types and methods back into class decls - - - - - 84169323 by Mark Lentczner at 2010-05-24T13:13:42+00:00 merge in changes from the big-whitespace cleanup - - - - - 3c1c872e by Mark Lentczner at 2010-06-11T21:03:58+00:00 adjust synopsis and bottom bar spacing - - - - - 3c1f9ef7 by Mark Lentczner at 2010-06-11T21:14:44+00:00 fix missing space in "module" lines in synoposis - - - - - 9a137e6d by Mark Lentczner at 2010-06-11T21:34:08+00:00 changed tt elements to code elements - - - - - 50f71ef1 by Mark Lentczner at 2010-06-11T23:27:46+00:00 factored out ppInstances - - - - - 3b9a9de5 by Mark Lentczner at 2010-06-17T17:36:01+00:00 push single constructors (newtype) onto line with decl - - - - - e0f8f2ec by Mark Lentczner at 2010-06-17T22:20:56+00:00 remove <++> connector - - - - - 56c075dd by Mark Lentczner at 2010-07-13T05:26:21+00:00 change to new page structure - - - - - 04be6ca7 by Mark Lentczner at 2010-07-14T04:21:55+00:00 constructors and args as dl lists, built in Layout.hs - - - - - 65aeafc2 by Mark Lentczner at 2010-07-14T05:38:32+00:00 better interface to subDecls - - - - - 72032189 by Mark Lentczner at 2010-07-14T07:04:10+00:00 made subDecl tables looks just so - - - - - b782eca2 by Mark Lentczner at 2010-07-14T16:00:54+00:00 convert args to SubDecl format - - - - - cc75e98f by Mark Lentczner at 2010-07-14T16:28:53+00:00 convert instances to SubDecl - - - - - 34e2aa5a by Mark Lentczner at 2010-07-14T21:07:32+00:00 removing old table cruft from Layout.hs - - - - - d5810d95 by Mark Lentczner at 2010-07-14T21:54:58+00:00 methods and associated types in new layout scheme - - - - - 65ef9579 by Mark Lentczner at 2010-07-14T23:43:42+00:00 clean up synopsis lists - - - - - e523318f by Mark Lentczner at 2010-07-15T05:02:26+00:00 clean up of anchors - - - - - 1215dfc5 by Mark Lentczner at 2010-07-15T23:53:01+00:00 added two new themes and rough css switcher - - - - - 7f0fd36f by Mark Lentczner at 2010-07-16T04:57:38+00:00 fixed package catpion, added style menu - - - - - 0dd4999c by Mark Lentczner at 2010-07-16T20:12:39+00:00 new output for mini_ pages - - - - - 64b2810b by Mark Lentczner at 2010-07-16T20:58:41+00:00 reformat index-frames - - - - - 3173f555 by Mark Lentczner at 2010-07-16T22:41:53+00:00 convert index to new markup - - - - - b0a4b7c9 by Mark Lentczner at 2010-07-17T04:07:22+00:00 convert index.html to new markup, adjust module markup - - - - - 8261ae1e by Mark Lentczner at 2010-07-17T05:07:29+00:00 classing styling of ancillary pages - - - - - 2a4fb025 by Mark Lentczner at 2010-07-17T05:11:45+00:00 clean up Layout.hs: no more vanillaTable - - - - - 87eec685 by Mark Lentczner at 2010-07-17T05:35:16+00:00 clean up Util.hs - - - - - d304e9b0 by Mark Lentczner at 2010-07-17T05:38:50+00:00 qualify import of XHtml as XHtml - - - - - 7dc05807 by Mark Lentczner at 2010-07-17T06:17:53+00:00 factored out head element generation - - - - - 9cdaec9e by Mark Lentczner at 2010-07-17T06:44:54+00:00 refactored out main page body generation - - - - - 8a51019e by Mark Lentczner at 2010-07-17T06:48:20+00:00 moved footer into only place that used it - - - - - efa479da by Mark Lentczner at 2010-07-17T18:48:30+00:00 styling auxillary pages for tibbe and snappy themes - - - - - 81de5509 by Mark Lentczner at 2010-07-18T04:41:38+00:00 fixed alphabet on index page, and styling of it and packages in module lists - - - - - 20718c1a by Mark Lentczner at 2010-07-18T05:34:29+00:00 cleaned up div functions in Layout.hs - - - - - 60d50453 by Mark Lentczner at 2010-07-18T05:48:39+00:00 added content div to main pages - - - - - ed16561c by Mark Lentczner at 2010-07-18T06:12:22+00:00 add .doc class to documentation blocks - - - - - f5c781b0 by Mark Lentczner at 2010-07-19T05:20:53+00:00 refactoring of anchor ID and fragment handling - - - - - a69a93bf by Mark Lentczner at 2010-07-19T05:35:55+00:00 remove an explicit bold tag - replace with .def class - - - - - d76c7225 by Mark Lentczner at 2010-07-19T06:56:15+00:00 rename Haddock.Backends.Xhtml.Util to Utils - - - - - 5a58c0da by David Waern at 2010-07-21T13:30:54+00:00 Remove trailing whitespace in Haddock.Backends.Xhtml - - - - - 0652aa17 by David Waern at 2010-07-21T13:33:21+00:00 Align a few comments - - - - - 785776c3 by David Waern at 2010-07-21T13:39:04+00:00 Remove trailing whitespace in H.B.X.Decl - - - - - 71a30710 by David Waern at 2010-07-21T13:44:27+00:00 Remove more trailing whitespace - - - - - 38750394 by David Waern at 2010-07-21T13:50:43+00:00 Style police - - - - - 3023d940 by David Waern at 2010-07-21T14:01:22+00:00 Style police in H.B.X.Decl - - - - - df16e9e6 by David Waern at 2010-07-21T14:14:45+00:00 Style police in H.B.X.DocMarkup - - - - - 6020e321 by David Waern at 2010-07-21T14:17:32+00:00 More style police - - - - - 86ad8bf5 by David Waern at 2010-07-21T14:21:02+00:00 Style police in H.B.Xhtml - - - - - aea27d03 by David Waern at 2010-07-21T14:42:03+00:00 Fix warnings in LaTeX backend - - - - - 2aff34a9 by David Waern at 2010-07-21T14:50:46+00:00 Style police in LaTeX backend (mainly more newlines) - - - - - e517162d by David Waern at 2010-07-21T15:05:47+00:00 Doc sections in Main - - - - - b971aa0c by David Waern at 2010-07-21T15:06:17+00:00 Trailing whitespace in Documentation.Haddock - - - - - f11628fb by David Waern at 2010-07-21T15:07:06+00:00 Trailing whitespace in Haddock.Convert - - - - - cbaf284c by David Waern at 2010-07-21T15:08:11+00:00 Style police in Haddock.GhcUtils - - - - - 71feb77b by David Waern at 2010-07-21T15:09:06+00:00 Style police in Haddock.InterfaceFile - - - - - 0a9c80e6 by David Waern at 2010-07-21T15:11:33+00:00 Whitespace police - - - - - 6168376c by David Waern at 2010-07-21T15:16:35+00:00 Style police in Haddock.Utils - - - - - 9fe4dd90 by David Waern at 2010-07-21T15:19:31+00:00 Add -fwarn-tabs - - - - - a000d752 by Mark Lentczner at 2010-07-20T17:25:52+00:00 move CSS Theme functions into Themes.hs - - - - - b52b440f by Mark Lentczner at 2010-07-20T17:29:35+00:00 add Thomas Schilling's theme - - - - - e43fa7e8 by Mark Lentczner at 2010-07-21T04:49:34+00:00 correct icon used with Snappy theme - - - - - ba5092d3 by Mark Lentczner at 2010-07-21T04:56:47+00:00 apply Tibbe's updates to his theme - - - - - 7804eef6 by Mark Lentczner at 2010-07-21T05:15:49+00:00 space between "Style" and the downward triangle - - - - - 7131d4c6 by Mark Lentczner at 2010-07-21T17:43:35+00:00 merge with David's source cleanups - - - - - ee65f1cb by David Waern at 2010-07-22T16:50:46+00:00 Fix a bug where we allowed --hoogle, --latex, etc without input files - - - - - e413ff7a by David Waern at 2010-07-22T17:21:58+00:00 Improve function name - - - - - a0fd14f3 by Simon Marlow at 2010-06-30T15:34:32+00:00 fix warnings - - - - - 31f73d2a by David Waern at 2010-07-22T19:29:41+00:00 Solve conflicts - - - - - d563b4a5 by Simon Marlow at 2010-06-30T15:34:37+00:00 fix warning - - - - - 412b6469 by David Waern at 2010-07-22T19:31:28+00:00 Solve conflict - - - - - 35174b94 by Ian Lynagh at 2010-07-06T17:27:16+00:00 Follow mkPState argument order change - - - - - b5c3585c by Simon Marlow at 2010-07-14T08:49:21+00:00 common up code for instance rendering - - - - - d8009560 by Simon Marlow at 2010-07-14T12:37:11+00:00 fix warnings - - - - - a6d88695 by David Waern at 2010-07-24T15:33:33+00:00 Fix build with ghc < 6.13 - - - - - 94cf9de1 by David Waern at 2010-07-24T15:34:37+00:00 Remove conflict left-over - - - - - 313b15c0 by Mark Lentczner at 2010-07-21T22:09:04+00:00 reorganization of nhaddock.css with tibbe - - - - - 9defed80 by Mark Lentczner at 2010-07-21T22:42:14+00:00 further cleanup of nhaddock.css, float TOC, support aux. pages - - - - - 6d944c1b by Mark Lentczner at 2010-07-22T06:22:23+00:00 remove old HTML backend - - - - - b3e8cba5 by Mark Lentczner at 2010-07-22T06:43:32+00:00 remove --html-help support - it was old, out-of-date, and mostly missing - - - - - d2654a08 by Mark Lentczner at 2010-07-22T21:45:34+00:00 tweaks to nhaddock.css - - - - - f73b285c by Mark Lentczner at 2010-07-23T06:19:35+00:00 command like processing for theme selection The bulk of the change is threadnig the selected theme set through functions in Xhtml.hs so that the selected themes can be used when generating the page output. There isn't much going on in most of these changes, just passing it along. The real work is all done in Themes.hs. - - - - - 8bddc90d by Mark Lentczner at 2010-07-23T06:58:31+00:00 drop --themes support, add named theme support decided that --themes was silly - no one would do that, just use multiple --theme arguments made --theme a synonym for --css and -c made those arguments, if no file is found, look up the argument as the name of a built in theme all of this let's haddock be invoked with "--theme=classic" for example. - - - - - 20cafd4f by Mark Lentczner at 2010-07-23T17:44:29+00:00 rename --default-themes to --built-in-themes - - - - - 0fe41307 by Mark Lentczner at 2010-07-23T18:33:02+00:00 tweaks to theme for info table, headings, and tables - - - - - cba4fee0 by Mark Lentczner at 2010-07-23T19:13:59+00:00 tweaks for dl layout, though still not used - - - - - 463fa294 by Mark Lentczner at 2010-07-23T21:07:19+00:00 tweak look of mini pages, keywords, and preblocks - - - - - 5472fc02 by Mark Lentczner at 2010-07-24T05:36:15+00:00 slide out Synopsis drawer - - - - - 9d5d5de5 by Mark Lentczner at 2010-07-24T06:02:42+00:00 extend package header and footer to edges of page - - - - - a47c91a2 by Mark Lentczner at 2010-07-24T06:28:44+00:00 fields are def lists, tweak css for style menu, mini pages, arguments - - - - - ca20f23b by Mark Lentczner at 2010-07-24T16:55:22+00:00 excisting last vestiges of the --xhtml flag - - - - - 71fb012e by Mark Lentczner at 2010-07-25T18:47:49+00:00 change how collapsing sections are done make whole .caption be the target improve javascript for class toggling have plus/minus images come from .css, not img tags - - - - - c168c8d3 by Mark Lentczner at 2010-07-26T00:32:05+00:00 reorganize files in the html lib data dir - - - - - 93324301 by Mark Lentczner at 2010-07-26T01:27:42+00:00 cleaned up Themes.hs - - - - - ad3b5dd4 by Mark Lentczner at 2010-07-26T02:39:15+00:00 make module list use new collapsers - - - - - 1df9bfc6 by Mark Lentczner at 2010-07-27T19:09:25+00:00 remove Tibbe theme - - - - - 8b9b01b3 by Mark Lentczner at 2010-07-27T20:04:03+00:00 move themes into html dir with .theme and .std-theme extensions - - - - - a7beb965 by Mark Lentczner at 2010-07-27T21:06:34+00:00 give a class to empty dd elements so they can be hidden - - - - - a258c117 by Mark Lentczner at 2010-07-27T21:23:58+00:00 remove custom version of copyFile in Xhtml.hs - - - - - b70dba6e by Mark Lentczner at 2010-07-27T22:12:45+00:00 apply margin changes to pre and headings as per group decision, and small cleanups - - - - - e6f722a2 by Mark Lentczner at 2010-07-28T00:03:12+00:00 make info block and package bar links be floatable by placing them first in the dom tree - - - - - c8278867 by Mark Lentczner at 2010-07-28T19:01:18+00:00 styling source links on declarations - - - - - 88fdc399 by Mark Lentczner at 2010-07-29T01:12:46+00:00 styling tweaks don't generate an empty li for absent style menu in links area update css for Classic and Snappy to handle: dl lists links in package header and in declarations floating of links and info block in package and module headers - - - - - 8a75b213 by Ian Lynagh at 2010-07-30T20:21:46+00:00 Fix build in GHC tree - - - - - ce8e18b3 by Simon Hengel at 2010-08-03T18:37:26+00:00 Adapt paths to data files in cabal file - - - - - 9701a455 by Simon Hengel at 2010-08-07T13:20:27+00:00 Add missing dependency to cabal file - - - - - 01b838d1 by Mark Lentczner at 2010-07-30T20:19:40+00:00 improved synopsis drawer: on click, not hover - - - - - 7b6f3e59 by Mark Lentczner at 2010-07-30T23:38:55+00:00 put the synopsis back in the other themes - - - - - 7b2904c9 by Mark Lentczner at 2010-08-11T11:11:26+00:00 close arrows on expanded synopsis drawer - - - - - ea19e177 by Mark Lentczner at 2010-08-12T21:16:45+00:00 width and font changes removed the max width restrictions on the page as a whole and the synopsis made the main font size smaller (nominally 14pt) and then tweaked most font sizes (relative) to be more consistent - - - - - 5ced00c0 by Mark Lentczner at 2010-08-13T15:09:55+00:00 implemented YUI's CSS font approach - - - - - 2799c548 by Mark Lentczner at 2010-08-13T15:11:59+00:00 adjusted margin to 2em, 1 wasn't enough - - - - - 58f06893 by Mark Lentczner at 2010-08-13T15:48:44+00:00 removed underlining on hover for named anchors headings in interface lost thier a element, no need, just put id on heading css for a elements now only applies to those with href attribute - - - - - 7aced4c4 by Mark Lentczner at 2010-08-13T15:50:22+00:00 more space between elements - - - - - 5a3c1cce by Mark Lentczner at 2010-08-13T16:43:43+00:00 adjusted font sizes of auxilary pages per new scheme - - - - - 487539ef by Mark Lentczner at 2010-08-13T21:43:41+00:00 add Frames button and clean up frames.html - - - - - c1a140b6 by Mark Lentczner at 2010-08-13T22:17:48+00:00 move frames button to js - - - - - b0bdb68e by Mark Lentczner at 2010-08-14T03:44:46+00:00 build style menu in javascript moved to javascript, so as to not polute the content with the style menu removed menu building code in Themes.hs removed onclick in Utils.hs changed text of button in header from "Source code" to "Source" more consistent with links in rest of page - - - - - 43ab7120 by Mark Lentczner at 2010-08-16T15:15:37+00:00 font size and margin tweaks - - - - - c0b68652 by Mark Lentczner at 2010-08-17T18:19:52+00:00 clean up collapser logics javascript code for collapasble sections cleaned up rewrote class utilities in javascript to be more robust refactored utilities for generating collapsable sections made toc be same color as synopsis module list has needed clear attribute in CSS - - - - - 5d573427 by Mark Lentczner at 2010-08-17T23:06:02+00:00 don't collapse entries in module list when clicking on links - - - - - 8c307c4a by Mark Lentczner at 2010-08-17T23:21:43+00:00 add missing data file to .cabal - - - - - 414bcfcf by Mark Lentczner at 2010-08-17T23:28:47+00:00 remove synopsis when in frames - - - - - ba0fa98a by Mark Lentczner at 2010-08-18T16:16:11+00:00 layout tweeks - mini page font size, toc color, etc. - - - - - 63c1bed1 by Mark Lentczner at 2010-08-18T19:50:02+00:00 margin fiddling - - - - - c311c094 by Mark Lentczner at 2010-08-20T01:37:55+00:00 better synopsis handling logic - no flashing - - - - - f1fe5fa8 by Mark Lentczner at 2010-08-20T01:41:06+00:00 fix small layout issues mini frames should have same size top heading give info block dts some padding so they don't collide in some browsers - - - - - 0de84d77 by Mark Lentczner at 2010-08-20T02:13:09+00:00 made style changing and cookies storage robust - - - - - 1ef064f9 by Thomas Schilling at 2010-08-04T13:12:22+00:00 Make synopsis frame behave properly in Firefox. In Firefox, pressing the back button first reverted the synopsis frame, and only clicking the back button a second time would update the main frame. - - - - - dd1c9a94 by Mark Lentczner at 2010-08-21T01:46:19+00:00 remove Snappy theme - - - - - 2353a90d by Mark Lentczner at 2010-08-25T05:16:19+00:00 fix occasional v.scroll bars on pre blocks (I think) - - - - - 459b8bf1 by Simon Hengel at 2010-08-08T10:12:45+00:00 Add createInterfaces' (a more high-level alternative to createInterfaces) to Haddock API - - - - - b1b68675 by David Waern at 2010-08-26T20:31:58+00:00 Follow recent API additions with some refactorings Simon Hegel's patch prompted me to do some refactorings in Main, Haddock.Documentation and Haddock.Interface. - - - - - 264d4d67 by David Waern at 2010-08-26T21:40:59+00:00 Get rid of GhcModule and related cruft We can get everything we need directly from TypecheckedModule. - - - - - 0feacec2 by Mark Lentczner at 2010-08-26T23:44:13+00:00 fixed CSS for ordered lists and def lists in doc blocks - - - - - 2997e0c2 by Mark Lentczner at 2010-08-26T23:45:03+00:00 support both kinds of enumerated lists in doc markup The documentation for Haddock says enumerated lists can use either of (1) first item 2. second item The second form wasn't actually supported - - - - - 5d4ddeec by Mark Lentczner at 2010-08-27T21:29:48+00:00 fix broken header link margins - - - - - 614456ba by Mark Lentczner at 2010-08-27T22:16:19+00:00 fix table of contents CSS - - - - - 03f329a2 by David Waern at 2010-08-28T16:36:09+00:00 Update tests following switch to the Xhtml backend - - - - - ca689fa2 by Mark Lentczner at 2010-08-28T18:25:16+00:00 fix def lists - - - - - 18e1d3d2 by Mark Lentczner at 2010-08-28T18:26:18+00:00 push footer to bottom of window - - - - - b0ab8d82 by David Waern at 2010-08-28T22:04:32+00:00 Whitespace police - - - - - 2d217977 by David Waern at 2010-08-29T12:44:45+00:00 Remove Snappy data files - - - - - 01e27d5f by David Waern at 2010-08-29T13:03:28+00:00 Add source entity path to --read-interface You can now use this flag like this: --read-interface=<html path>,<source entity path>,<.haddock file> By "source entity path" I mean the same thing that is specified with the --source-entity flag. The purpose of this is to be able to specify the source entity path per package, to allow source links to work in the presence of cross-package documentation. When given two arguments or less the --read-interface flag behaves as before. - - - - - 20bf4aaa by David Waern at 2010-08-29T13:11:03+00:00 Naming wibbles - - - - - ad22463f by Mark Lentczner at 2010-08-29T15:14:54+00:00 make portability block be a table - solves layout issues - - - - - 97bd1ae6 by Mark Lentczner at 2010-08-29T15:17:42+00:00 update golden test for Test due to portability box change - - - - - d37e139e by Mark Lentczner at 2010-08-29T17:07:17+00:00 move TOC and Info blocks down 0.5em to improve layout issue w/Test.hs - - - - - acf52501 by David Waern at 2010-08-29T17:32:36+00:00 Allow building with ghc < 6.16 - - - - - 1cb34ed8 by Ian Lynagh at 2010-07-24T23:18:49+00:00 Flatten the dynflags before parsing - - - - - b36845b4 by Ian Lynagh at 2010-07-24T23:26:49+00:00 Follow flattenLanguageFlags -> flattenExtensionFlags rename - - - - - 7f7fcc7e by David Waern at 2010-08-29T17:46:23+00:00 Use flattenExtensionFlags with ghc >= 6.13 only - - - - - 13cf9411 by Ian Lynagh at 2010-08-01T18:09:54+00:00 Make the main haddock script versioned, and make plain "haddock" a symlink - - - - - 495cbff2 by Ian Lynagh at 2010-08-18T18:57:24+00:00 Fix installation in the GHC build system Data-files are now in subdirectories, so we need to handle that - - - - - 88ebab0a by Ian Lynagh at 2010-08-18T19:43:53+00:00 GHC build system: Add all the data files to BINDIST_EXTRAS - - - - - 65837172 by David Waern at 2010-08-29T20:12:34+00:00 Update Test - - - - - 094bbaa2 by David Waern at 2010-08-29T20:55:14+00:00 Revert update to Test - - - - - a881cfb3 by David Waern at 2010-08-31T18:24:15+00:00 Bump version number - - - - - 1fc8a3eb by David Waern at 2010-08-31T22:32:27+00:00 Update ANNOUNCE - - - - - ee1df9d0 by David Waern at 2010-08-31T22:33:11+00:00 Update CHANGES - - - - - 394cc854 by David Waern at 2010-08-31T22:33:23+00:00 Update interface file versioning to work with ghc 6.14/15 - - - - - 7d03b79b by David Waern at 2010-08-31T22:36:00+00:00 Update test output following version change - - - - - a48d82d1 by Mark Lentczner at 2010-09-01T04:29:35+00:00 sort options in doc to match --help output removed --html-help option, as it is no longer supported - - - - - 06561aeb by Mark Lentczner at 2010-09-01T05:29:32+00:00 update options documentation rewrote doc for --html added doc for --theme and --built-in-themes added --use-contents and --gen-contents - - - - - 57dea832 by Mark Lentczner at 2010-09-01T05:31:27+00:00 slight wording change about Frames mode - - - - - fa1f6da3 by David Waern at 2010-09-01T10:57:44+00:00 Update doc configure script to find docbook stylesheets on arch linux - - - - - addff770 by David Waern at 2010-09-01T11:02:29+00:00 Wibble - - - - - 8399006d by David Waern at 2010-09-01T11:19:21+00:00 Replace ghci> with >>> in example syntax - - - - - 35074cf8 by David Waern at 2010-09-01T19:03:27+00:00 Improve docs for --no-tmp-comp-dir - - - - - 0f8f8cfd by David Waern at 2010-09-02T11:22:27+00:00 Add a list of contributors to the user guide Break out everyone thanked in the `Acknowledgements` chapter into a separate contributor list and add everyone from `darcs show authors`. We consider everyone who is thanked to be a contributor as a conservative estimation :-) I have added some more contributors that I know about, who were not in the darcs history, but others may be missing. So please add anyone that you think is missing from the list. - - - - - 42ccf099 by David Waern at 2010-09-02T11:29:22+00:00 Update copyright years in license - - - - - 0d560479 by David Waern at 2010-09-02T11:38:52+00:00 Update release instructions - - - - - 72ab7796 by David Waern at 2010-09-02T19:27:08+00:00 Add a note to ANNOUNCE - - - - - bf9d9c5d by David Waern at 2010-09-02T19:27:48+00:00 H.Utils needs FFI on Win+MinGW - - - - - 048ae44a by Mark Lentczner at 2010-09-04T23:19:47+00:00 make TOC group header identifiers validate - - - - - 8c6faf36 by Simon Michael at 2010-09-22T07:12:34+00:00 add hints for cleaner darcs show authors output - - - - - 9909bd17 by Simon Michael at 2010-09-22T17:58:06+00:00 print haddock coverage info on stdout when generating docs A module's haddockable items are its exports and the module itself. The output is lightly formatted so you can align the :'s and sort for readability. - - - - - 6da72171 by David Waern at 2010-10-03T21:31:24+00:00 Style wibble - - - - - 2f8d8e4d by Tobias Brandt at 2010-08-27T07:01:21+00:00 adding the option to fully qualify identifiers - - - - - 833be6c6 by Tobias Brandt at 2010-08-27T15:50:28+00:00 adding support for local and relative name qualification - - - - - df15c4e9 by Tobias Brandt at 2010-08-27T15:56:37+00:00 corrected qualification help message - - - - - 449e9ce1 by David Waern at 2010-10-16T17:34:30+00:00 Solve conflicts - - - - - 3469bda5 by David Waern at 2010-10-16T18:42:40+00:00 Use "qual" as an abbreviation for qualification instead of "quali" for consistency - - - - - 97c2d728 by David Waern at 2010-10-16T18:47:07+00:00 Style police - - - - - ce14fbea by David Waern at 2010-10-16T21:15:25+00:00 Style police - - - - - fdf29e9d by David Waern at 2010-10-17T00:30:44+00:00 Add a pointer to the style guide - - - - - 8e6b44e8 by rrnewton at 2010-10-24T03:19:28+00:00 Change to index pages: include an 'All' option even when subdividing A-Z. - - - - - 755b131c by David Waern at 2010-11-14T19:39:36+00:00 Bump version - - - - - d0345a04 by David Waern at 2010-11-14T19:41:59+00:00 TAG 2.8.1 - - - - - f6221508 by Simon Peyton Jones at 2010-09-13T09:53:00+00:00 Adapt to minor changes in internal GHC functions - - - - - 1290713d by Ian Lynagh at 2010-09-15T10:37:18+00:00 Remove duplicate Outputable instance for Data.Map.Map - - - - - 87f69eef by Ian Lynagh at 2010-09-21T15:01:10+00:00 Bump GHC dep upper bound - - - - - af36e087 by Ian Lynagh at 2010-09-21T15:12:02+00:00 Fix up __GLASGOW_HASKELL__ tests - - - - - ad67716c by Ian Lynagh at 2010-09-21T20:31:35+00:00 Don't build haddock is HADDOCK_DOCS is NO - - - - - 63b3f1f5 by Ian Lynagh at 2010-09-21T21:39:51+00:00 Fixes for when HADDOCK_DOCS=NO - - - - - e92bfa42 by Ian Lynagh at 2010-09-29T21:15:38+00:00 Fix URL creation on Windows: Use / not \ in URLs. Fixes haskell/haddock#4353 - - - - - 66c55e05 by Ian Lynagh at 2010-09-30T17:03:34+00:00 Tidy up haddock symlink installation In particular, it now doesn't get created if we aren't installing haddock. - - - - - 549b5556 by Ian Lynagh at 2010-10-23T21:17:14+00:00 Follow extension-flattening change in GHC - - - - - d7c2f72b by David Waern at 2010-11-14T20:17:55+00:00 Bump version to 2.8.2 - - - - - 6989a3a9 by David Waern at 2010-11-14T20:26:01+00:00 Solve conflict - - - - - 055c6910 by Ian Lynagh at 2010-09-22T15:36:20+00:00 Bump GHC dep - - - - - c96c0763 by Simon Marlow at 2010-10-27T11:09:44+00:00 follow changes in the GHC API - - - - - 45907129 by David Waern at 2010-11-07T14:00:58+00:00 Update the HCAR entry - - - - - 61940b95 by David Waern at 2010-11-07T14:07:34+00:00 Make the HCAR entry smaller - - - - - aa590b7d by David Waern at 2010-11-14T21:30:59+00:00 Update HCAR entry with November 2010 version - - - - - 587f9847 by David Waern at 2010-11-14T23:48:17+00:00 Require ghc >= 7.0 - - - - - ff5c647c by David Waern at 2010-11-14T23:49:09+00:00 TAG 2.8.2 - - - - - 937fcb4f by David Waern at 2010-11-14T23:49:45+00:00 Solve conflict - - - - - 8e5d0c1a by David Waern at 2010-11-15T21:09:50+00:00 Remove code for ghc < 7 - - - - - 3d47b70a by David Waern at 2010-11-15T21:11:06+00:00 Fix bad merge - - - - - 7f4a0d8a by David Waern at 2010-11-15T21:13:57+00:00 Remove more ghc < 7 code - - - - - 9ee34b50 by David Waern at 2010-11-15T21:31:25+00:00 Match all AsyncExceptions in exception handler - - - - - 42849c70 by David Waern at 2010-11-15T21:35:31+00:00 Just say "internal error" instead of "internal Haddock or GHC error" - - - - - c88c809b by David Waern at 2010-11-15T21:44:19+00:00 Remove docNameOcc under the motto "don't name compositions" - - - - - b798fc7c by David Waern at 2010-11-15T23:27:13+00:00 Wibble - - - - - 2228197e by David Waern at 2010-11-15T23:28:24+00:00 Rename the HCAR entry file - - - - - 8a3f9090 by David Waern at 2010-11-16T00:05:29+00:00 Remove Haskell 2010 extensions from .cabal file - - - - - c7a0c597 by David Waern at 2010-11-16T00:10:28+00:00 Style wibbles - - - - - cde707a5 by David Waern at 2010-11-16T00:12:00+00:00 Remove LANGUAGE ForeignFunctionInterface pragmas - - - - - 1dbda8ed by David Waern at 2010-11-16T00:17:21+00:00 Make a little more use of DoAndIfThenElse - - - - - 4c45ff6e by David Waern at 2010-11-16T00:59:41+00:00 hlint police - - - - - d2feaf09 by David Waern at 2010-11-16T01:14:15+00:00 hlint police - - - - - 99876e97 by David Waern at 2010-11-20T19:06:00+00:00 Haddock documentation updates - - - - - 65ce6987 by David Waern at 2010-11-20T19:42:51+00:00 Follow the style guide closer in Haddock.Types and improve docs - - - - - 28ca304a by tob.brandt at 2010-11-20T17:04:40+00:00 add full qualification for undocumented names - - - - - d61341e3 by David Waern at 2010-11-20T20:04:15+00:00 Re-structure qualification code a little - - - - - 0057e4d6 by David Waern at 2010-11-20T20:07:55+00:00 Re-order functions - - - - - d7279afd by David Waern at 2010-11-21T03:39:54+00:00 Add BangPatterns to alex and happy source files - - - - - 629fe60e by tob.brandt at 2010-11-23T23:35:11+00:00 documentation for qualification - - - - - 37031cee by David Waern at 2010-11-23T21:06:44+00:00 Update CHANGES - don't mention 2.8.2, we won't release it - - - - - f2489e19 by David Waern at 2010-12-01T21:57:11+00:00 Update deps of runtests.hs to work with ghc 7.0.1 - - - - - d3657e9a by David Waern at 2010-12-01T22:04:57+00:00 Make tests compile with ghc 7.0.1 - - - - - a2f09d9b by David Waern at 2010-12-01T22:06:59+00:00 Update tests following version bump - - - - - 50883ebb by David Waern at 2010-12-06T14:09:18+00:00 Update tests following recent changes - - - - - fc2fadeb by David Waern at 2010-12-06T14:17:29+00:00 Add a flag --pretty-html for rendering indented html with newlines - - - - - 30832ef2 by David Waern at 2010-12-06T14:17:35+00:00 Use --pretty-html when running the test suite. Makes it easier to compare output - - - - - a0b81b31 by David Waern at 2010-12-06T14:18:27+00:00 Wibble - - - - - 3aaa23fe by David Waern at 2010-12-06T14:19:29+00:00 Haddockify ppHtml comments - - - - - 24bb24f0 by David Waern at 2010-12-06T14:23:15+00:00 Remove --debug. It was't used, and --verbosity should take its place - - - - - 6bc076e5 by David Waern at 2010-12-06T14:25:37+00:00 Rename golden-tests into html-tests. "golden tests" sounds strange - - - - - 53301e55 by David Waern at 2010-12-06T14:26:26+00:00 QUALI -> QUAL in the description --qual for consistency - - - - - 98b6affb by David Waern at 2010-12-06T21:54:02+00:00 Bump version - - - - - 371bf1b3 by David Waern at 2010-12-06T22:08:55+00:00 Update tests following version bump - - - - - 25be762d by David Waern at 2010-12-06T22:21:03+00:00 Update CHANGES - - - - - 7c7dac71 by David Waern at 2010-12-06T22:33:43+00:00 Update ANNOUNCE - - - - - 30d7a5f2 by Simon Peyton Jones at 2010-11-15T08:38:38+00:00 Alex generates BangPatterns, so make Lex.x accept them (It'd be better for Alex to generate this pragma.) - - - - - 605e8018 by Simon Marlow at 2010-11-17T11:37:24+00:00 Add {-# LANGUAGE BangPatterns #-} to mollify GHC - - - - - a46607ba by David Waern at 2010-12-07T14:08:10+00:00 Solve conflicts - - - - - b28cda66 by David Waern at 2010-12-09T20:41:35+00:00 Docs: Mention that \ is a special character in markup - - - - - a435bfdd by Ian Lynagh at 2010-11-17T14:01:19+00:00 TAG GHC 7.0.1 release - - - - - 5a15a05a by David Waern at 2010-12-11T17:51:19+00:00 Fix indentation problem - - - - - 4232289a by Lennart Kolmodin at 2010-12-17T18:32:03+00:00 Revise haddock.cabal given that we now require ghc-7 default-language should be Haskell2010, slight new semantics for extensions. Rewrite into clearer dependencies of base and Cabal. - - - - - a36302dc by David Waern at 2010-12-19T17:12:37+00:00 Update CHANGES - - - - - 7c8b85b3 by David Waern at 2010-12-19T17:14:24+00:00 Bump version - - - - - cff22813 by Ian Lynagh at 2011-01-05T18:24:27+00:00 Write hoogle output in utf8; fixes GHC build on Windows - - - - - c7e762ea by David Waern at 2011-01-22T00:00:35+00:00 Put title outside doc div when HTML:fying title+prologue Avoids indenting the title, and makes more sense since the title is not a doc string anyway. - - - - - 5f639054 by David Waern at 2011-01-22T16:09:44+00:00 Fix spelling error - contributed by Marco Silva - - - - - c11dce78 by Ian Lynagh at 2011-01-07T02:33:11+00:00 Follow GHC build system changes - - - - - 101cfaf5 by David Waern at 2011-01-08T14:06:44+00:00 Bump version - - - - - af62348b by David Waern at 2011-01-08T14:07:07+00:00 TAG 2.9.2 - - - - - 4d1f6461 by Ian Lynagh at 2011-01-07T23:06:57+00:00 Name the haddock script haddock-ghc-7.0.2 instead of haddock-7.0.2; haskell/haddock#4882 "7.0.2" looked like a haddock version number before - - - - - 8ee4d5d3 by Simon Peyton Jones at 2011-01-10T17:31:12+00:00 Update Haddock to reflect change in hs_tyclds field of HsGroup - - - - - 06f3e3db by Ian Lynagh at 2011-03-03T15:02:37+00:00 TAG GHC 7.0.2 release - - - - - 7de0667d by David Waern at 2011-03-10T22:47:13+00:00 Update CHANGES - - - - - 33a9f1c8 by David Waern at 2011-03-10T22:47:31+00:00 Fix build with ghc 7.0.1 - - - - - 4616f861 by David Waern at 2011-03-10T22:47:50+00:00 TAG 2.9.2-actual - - - - - 0dab5e3c by Simon Hengel at 2011-04-08T15:53:01+00:00 Set shell script for unit tests back to work - - - - - 85c54dee by Simon Hengel at 2011-04-08T16:01:24+00:00 Set unit tests back to work Here "ghci>" was still used instead of ">>>". - - - - - 1cea9b78 by Simon Hengel at 2011-04-08T16:25:36+00:00 Update runtests.hs for GHC 7.0.2 - - - - - 8e5b3bbb by Simon Hengel at 2011-04-08T16:28:49+00:00 Update Haddock version in *.html.ref - - - - - 2545e955 by Simon Hengel at 2011-04-08T17:09:28+00:00 Add support for blank lines in the result of examples Result lines that only contain the string "<BLANKLINE>" are treated as a blank line. - - - - - adf64d2e by Simon Hengel at 2011-04-08T17:36:50+00:00 Add documentation for "support for blank lines in the result of examples" - - - - - c51352ca by David Waern at 2011-05-21T23:57:56+00:00 Improve a haddock comment - - - - - 7419cf2c by David Waern at 2011-05-22T15:41:52+00:00 Use cabal's test suite support to run the test suite This gives up proper dependency tracking of the test script. - - - - - 7770070c by David Waern at 2011-05-22T01:45:44+00:00 We don't need to send DocOptions nor a flag to mkExportItems - - - - - 9d95b7b6 by David Waern at 2011-05-22T21:39:03+00:00 Fix a bug - - - - - 1f93699b by David Waern at 2011-05-22T21:40:21+00:00 Break out fullContentsOf, give it a better name and some documentation The documentation describes how we want this function to eventually behave, once we have fixed a few problems with the current implementation. - - - - - 9a86432f by David Waern at 2011-05-22T21:53:52+00:00 Fix some stylistic issues in mkExportItems - - - - - c271ff0c by David Waern at 2011-05-22T22:09:11+00:00 Indentation - - - - - 93e602b1 by David Waern at 2011-06-10T01:35:31+00:00 Add git commits since switchover: darcs format (followed by a conflict resolution): commit 6f92cdd12d1354dfbd80f8323ca333bea700896a Merge: f420cc4 28df3a1 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Thu May 19 17:54:34 2011 +0100 Merge remote branch 'origin/master' into ghc-generics commit 28df3a119f770fdfe85c687dd73d5f6712b8e7d0 Author: Max Bolingbroke <batterseapower at hotmail.com> Date: Sat May 14 22:37:02 2011 +0100 Unicode fix for getExecDir on Windows commit 89813e729be8bce26765b95419a171a7826f6d70 Merge: 6df3a04 797ab27 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 9 11:55:17 2011 +0100 Merge branch 'ghc-new-co' commit 6df3a040da3dbddee67c6e30a892f87e6b164383 Author: Ian Lynagh <igloo at earth.li> Date: Sun May 8 17:05:50 2011 +0100 Follow changes in SDoc commit f420cc48b9259f0b1afd2438b12f9a2bde57053d Author: Jose Pedro Magalhaes <jpm at cs.uu.nl> Date: Wed May 4 17:31:52 2011 +0200 Adapt haddock to the removal of HsNumTy and TypePat. commit 797ab27bdccf39c73ccad374fea265f124cb52ea Merge: 1d81436 5a91450 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:05:03 2011 +0100 Merge remote branch 'origin/master' into ghc-new-co commit 1d8143659a81cf9611668348e33fd0775c7ab1d2 Author: Simon Peyton Jones <simonpj at microsoft.com> Date: Mon May 2 12:03:46 2011 +0100 Wibbles for ghc-new-co branch commit 5a91450e2ea5a93c70bd3904b022445c9cc82488 Author: Ian Lynagh <igloo at earth.li> Date: Fri Apr 22 00:51:56 2011 +0100 Follow defaultDynFlags change in GHC - - - - - 498da5ae by David Waern at 2011-06-11T00:33:33+00:00 * Merge in git patch from Michal Terepeta >From 6fc71d067738ef4b7de159327bb6dc3d0596be29 Mon Sep 17 00:00:00 2001 From: Michal Terepeta <michal.terepeta at gmail.com> Date: Sat, 14 May 2011 19:18:22 +0200 Subject: [PATCH] Follow the change of TypeSig in GHC. This follows the change in GHC to make TypeSig take a list of names (instead of just one); GHC ticket haskell/haddock#1595. This should also improve the Haddock output in case the user writes a type signature that refers to many names: -- | Some comment.. foo, bar :: ... will now generate the expected output with one signature for both names. - - - - - 094607fe by Ian Lynagh at 2011-06-17T19:10:29+01:00 Fix build - - - - - 8fa35740 by Ian Lynagh at 2011-06-26T21:06:40+01:00 Bump GHC dep to allow 7.2 - - - - - e4d2ca3c by Ian Lynagh at 2011-07-07T23:06:28+01:00 Relax base dep - - - - - b948fde9 by Ian Lynagh at 2011-07-28T16:39:45+01:00 GHC build system: Don't install the datafiles twice - - - - - f82f6d70 by Simon Marlow at 2011-08-11T12:08:15+01:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - b341cc12 by Max Bolingbroke at 2011-08-22T20:25:27+01:00 Fix compilation with no-pred-ty GHC - - - - - 30494581 by Max Bolingbroke at 2011-08-23T10:20:54+01:00 Remaining fixes for PredTy removal - - - - - 0b197138 by Max Bolingbroke at 2011-08-26T08:27:45+01:00 Rename factKind to constraintKind - - - - - a379bec5 by Max Bolingbroke at 2011-09-04T12:54:47+01:00 Deal with change to IParam handling in GHC - - - - - f94e421b by Max Bolingbroke at 2011-09-06T17:34:31+01:00 Adapt Haddock for the ConstraintKind extension changes - - - - - 8821e5cc by Max Bolingbroke at 2011-09-09T08:24:59+01:00 Ignore associated type defaults (just as we ignore default methods) - - - - - 31a0afd4 by Max Bolingbroke at 2011-09-09T09:06:00+01:00 Merge branch 'no-pred-ty' of ssh://darcs.haskell.org/srv/darcs/haddock into no-pred-ty - - - - - dd3b530a by Max Bolingbroke at 2011-09-09T14:10:25+01:00 Merge branch 'no-pred-ty' Conflicts: src/Haddock/Convert.hs - - - - - 5f25ec96 by Max Bolingbroke at 2011-09-09T14:10:40+01:00 Replace FactTuple with ConstraintTuple - - - - - cd30b9cc by David Waern at 2011-09-26T02:17:55+02:00 Bump to version 2.9.3 - - - - - 4fbfd397 by Max Bolingbroke at 2011-09-27T14:55:21+01:00 Follow changes to BinIface Name serialization - - - - - 92257d90 by David Waern at 2011-09-30T23:45:07+02:00 Fix problem with test files not added to distribution tarball - - - - - 00255bda by David Waern at 2011-09-30T23:48:24+02:00 Merge branch 'development' - - - - - 5421264f by David Waern at 2011-10-01T01:25:39+02:00 Merge in darcs patch from Simon Meier: Wed Jun 1 19:41:16 CEST 2011 iridcode at gmail.com * prettier haddock coverage info The new coverage info rendering uses less horizontal space. This reduces the number of unnecessary line-wrappings. Moreover, the most important information, how much has been documented already, is now put up front. Hopefully, this makes it more likely that a library author is bothered by the low coverage of his modules and fixes that issue ;-) - - - - - 07d318ef by David Waern at 2011-10-01T01:34:10+02:00 Use printException instead of deprecated printExceptionAndWarnings - - - - - 40d52ee4 by David Waern at 2011-10-01T01:41:13+02:00 Merge in darcs pach: Mon Apr 11 18:09:54 JST 2011 Liyang HU <haddock at liyang.hu> * Remember collapsed sections in index.html / haddock-util.js - - - - - 279d6dd4 by David Waern at 2011-10-01T01:55:45+02:00 Merge in darcs patch: Joachim Breitner <mail at joachim-breitner.de>**20110619201645 Ignore-this: f6c51228205b0902ad5bfad5040b989a As reported on http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=578301, generating the global index takes much too long if type-level (with lots of auto-generated types) is installed. The patch avoids a quadratic runtime in the subfunction getIfaceIndex of ppHtmlIndex by using a temporary set. Runtime improvement observed here from 25.36s to 2.86s. - - - - - d1612383 by David Waern at 2011-10-01T01:56:48+02:00 Merge branch 'development' - - - - - 347520c1 by David Waern at 2011-10-01T01:56:54+02:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9a0c95e8 by David Waern at 2011-10-01T02:19:10+02:00 Improve .cabal file - - - - - 6967dc64 by Ian Lynagh at 2011-10-01T01:34:06+01:00 Follow changes to ForeignImport/ForeignExport in GHC - - - - - 565cb26b by Simon Marlow at 2011-10-04T00:15:04+02:00 Hack this to make it work with both Alex 2.x and Alex 3.x. Unicode in documentation strings is (still) mangled. I don't think it's possible to make it so that we get the current behaviour with Alex 2.x but magic Unicode support if you use Alex 3.x. At some point we have to decide that Alex 3.x is a requirement, then we can do Unicode. - - - - - 8b74f512 by David Waern at 2011-10-04T00:18:17+02:00 Requre ghc >= 7.2 - - - - - 271d360c by David Waern at 2011-10-04T00:22:50+02:00 Bump version to 2.9.4 - - - - - 37f3edb0 by David Waern at 2011-10-06T02:30:21+02:00 Add alex and happy to build-tools. - - - - - 7ac2bb6e by David Terei at 2011-10-12T14:02:55-07:00 Add safe haskell indication to haddock output - - - - - 42c91a47 by David Terei at 2011-10-12T14:06:03-07:00 Fix CSS issue with info table not being contained in module header - - - - - 0eddab6c by David Terei at 2011-10-12T14:06:58-07:00 Add safe haskell indication to haddock output - - - - - 3df058eb by David Terei at 2011-10-12T14:07:07-07:00 Fix CSS issue with info table not being contained in module header - - - - - a40a6c3f by David Waern at 2011-10-22T11:29:06+02:00 Bump .haddock file version since the format has changed recently - - - - - 8a6254be by David Waern at 2011-10-22T11:30:42+02:00 Merge branch 'development' - - - - - 642e3e02 by David Waern at 2011-10-23T21:23:39+02:00 Sort import list - - - - - 36371cf8 by David Waern at 2011-10-23T22:48:18+02:00 Remove NEW_GHC_LAYOUT conditional. - - - - - 5604b499 by David Waern at 2011-10-27T00:15:03+02:00 Add --print-ghc-path. - - - - - 463499fa by David Waern at 2011-10-27T00:16:22+02:00 Make testsuite able to find its dependencies automatically. - - - - - a3506172 by Ryan Newton at 2011-11-05T05:59:58-04:00 Improved declNames internal error. Added a case to handle DocD. - - - - - 001b8baf by David Waern at 2011-11-05T20:37:29+01:00 Rename copy.hs -> accept.hs. - - - - - 55d808d3 by David Waern at 2011-11-05T23:30:02+01:00 Fix build. - - - - - deb5c3be by David Waern at 2011-11-06T00:01:47+01:00 Merge branch 'master' of http://darcs.haskell.org/haddock - - - - - 9b663554 by David Waern at 2011-11-06T00:03:45+01:00 Merge https://github.com/rrnewton/haddock - - - - - 1abb0ff6 by David Waern at 2011-11-06T01:20:37+01:00 Use getDeclMainBinder instead of declNames. - - - - - 4b005c01 by David Waern at 2011-11-06T19:09:53+01:00 Fix build. - - - - - c2c51bc7 by Ian Lynagh at 2011-11-06T23:01:33+00:00 Remove -DNEW_GHC_LAYOUT in ghc.mk - - - - - f847d703 by Jose Pedro Magalhaes at 2011-11-11T09:07:39+00:00 New kind-polymorphic core This big patch implements a kind-polymorphic core for GHC. The current implementation focuses on making sure that all kind-monomorphic programs still work in the new core; it is not yet guaranteed that kind-polymorphic programs (using the new -XPolyKinds flag) will work. For more information, see http://haskell.org/haskellwiki/GHC/Kinds - - - - - 7d7c3b09 by Jose Pedro Magalhaes at 2011-11-16T21:42:22+01:00 Follow changes to tuple sorts in master - - - - - 8430e03e by Simon Peyton Jones at 2011-11-17T10:20:27+00:00 Remove redundant imports - - - - - d1b06832 by Ian Lynagh at 2011-11-19T01:33:21+00:00 Follow GHC build system change to the way we call rm - - - - - 9e2230ed by David Waern at 2011-11-24T15:00:24+01:00 Fix a bug in test runner and get rid of regex-compat dependency. - - - - - 52039b21 by David Waern at 2011-11-24T23:55:36+01:00 Avoid haskell98 dependency in test - - - - - 92e1220d by David Waern at 2011-11-25T00:03:33+01:00 Avoid depency on regex-compat also in accept.hs. - - - - - ddac6b6f by David Waern at 2011-11-25T02:13:38+01:00 Accept test output. - - - - - 5a720455 by David Waern at 2011-11-25T02:16:20+01:00 Some more changes to test scripts. - - - - - 170a9004 by David Waern at 2011-11-25T02:30:41+01:00 Add flag --interface-version. - - - - - d225576c by David Waern at 2011-11-25T02:39:26+01:00 Remove #ifs for older compiler versions. - - - - - f0d0a4f5 by David Waern at 2011-11-26T04:20:12+01:00 Give preference to type over data constructors for doc comment links at renaming time. Previously this was done in the backends. Also, warn when a doc comment refers to something that is in scope but which we don't have the .haddock file for. These changes mean we can make DocIdentifier [a] into DocIdentifier a. - - - - - eef0e776 by David Waern at 2011-11-26T17:01:06+01:00 Allow doc comments to link to out-of-scope things (#78). (A bug that should have been fixed long ago.) - - - - - 565ad529 by David Waern at 2011-11-26T19:56:21+01:00 Update tests. - - - - - fb3ce7b9 by David Waern at 2011-11-26T21:44:28+01:00 Cleanup. - - - - - d0328126 by David Waern at 2011-11-26T22:10:28+01:00 Fix module reference bug. - - - - - c03765f8 by David Waern at 2011-12-03T05:20:20+01:00 Slightly better behaviour on top-levels without type signatures. - Docs don't get attached to the next top-level with signature by mistake. - If there's an export list and the top-level is part of it, its doc comment shows up in the documentation. - - - - - 48461d31 by David Waern at 2011-12-03T05:38:10+01:00 Add a test for Unicode doc comments. - - - - - 549c4b4e by David Waern at 2011-12-03T19:07:55+01:00 Cleanup. - - - - - 7bfecf91 by David Waern at 2011-12-03T20:13:08+01:00 More cleanup. - - - - - 14fab722 by Ian Lynagh at 2011-12-12T21:21:35+00:00 Update dependencies and binaryInterfaceVersion - - - - - 469e6568 by Ian Lynagh at 2011-12-18T12:56:16+00:00 Fix (untested) building from source tarball without alex/happy haddock's .cabal file was declaring that it needed alex and happy to build, but in the GHC source tarballs it doesn't. - - - - - 895c9a8c by David Waern at 2011-12-27T12:57:43+01:00 Go back to having a doc, sub and decl map instead of one big decl map. This setup makes more sense since when we add value bindings to the processed declarations (for type inference), we will have multiple declarations which should share documentation. Also, we already have a separate doc map for instances which we can now merge into the main doc map. Another benefit is that we don't need the DeclInfo type any longer. - - - - - 736767d9 by David Waern at 2011-12-27T13:33:41+01:00 Merge ../../../haddock Conflicts: src/Haddock/InterfaceFile.hs - - - - - 20016f79 by David Waern at 2011-12-27T13:57:23+01:00 Bump version. - - - - - 31f276fb by David Waern at 2011-12-27T13:57:32+01:00 Merge ../ghc/utils/haddock - - - - - 95b367cd by David Waern at 2011-12-27T14:57:29+01:00 Update tests following version bump. - - - - - fa3c94cd by David Waern at 2011-12-27T14:57:51+01:00 Get rid of quite unnecessary use of different lists. - - - - - 9c4d3c54 by David Waern at 2011-12-27T15:26:42+01:00 Cleanup. - - - - - 2caf9f90 by David Waern at 2011-12-27T16:18:05+01:00 Wibbles. - - - - - 3757d09b by David Waern at 2011-12-27T20:50:26+01:00 Complete support for inferring types for top-level bindings. - - - - - 53418734 by David Waern at 2011-12-28T15:02:13+01:00 Minor fixes and cleanup. - - - - - 0c9d0385 by Ian Lynagh at 2012-01-03T18:31:29+00:00 Follow rename of Instance to ClsInst in GHC - - - - - c9bc969a by Simon Hengel at 2012-01-12T21:28:14+01:00 Make sure that generated xhtml is valid (close haskell/haddock#186) Thanks to Phyx. - - - - - 836a0b9a by David Waern at 2012-02-01T02:30:05+01:00 Fix bug introduced in my recent refactoring. - - - - - c7d733eb by David Waern at 2012-02-01T02:30:26+01:00 Cleanup mkMaps and avoid quadratic behaviour. - - - - - da3cda8f by David Waern at 2012-02-01T02:56:56+01:00 Require ghc >= 7.4. - - - - - 83a3287e by David Waern at 2012-02-01T02:57:36+01:00 Update CHANGES. - - - - - 93408f0b by Simon Hengel at 2012-02-04T00:48:04+01:00 Add reference renderings - - - - - 49d00d2c by Simon Hengel at 2012-02-04T00:48:25+01:00 Set unit tests for parser back to work - - - - - eb450980 by Simon Hengel at 2012-02-04T00:49:07+01:00 Add .gitignore - - - - - a841602c by Simon Hengel at 2012-02-04T00:49:16+01:00 Add .ghci file - - - - - 8861199d by Simon Hengel at 2012-02-04T00:49:29+01:00 tests/html-tests/copy.hs: Use mapM_ instead of mapM So we do net get a list of () on stdout when running with runhaskell. - - - - - b477d9b5 by Simon Hengel at 2012-02-04T00:49:46+01:00 Remove index files from golden tests - - - - - 9dbda34e by Simon Hengel at 2012-02-04T00:49:57+01:00 Add /tests/html-tests/tests/*index*.ref to .gitignore - - - - - a9434817 by Simon Hengel at 2012-02-04T00:50:04+01:00 Add DocWarning to Doc The Xhtml backend has special markup for that, Hoogle and LaTeX reuse what we have for DocEmphasis. - - - - - de2fb6fa by Simon Hengel at 2012-02-04T00:50:13+01:00 Add support for module warnings - - - - - 0640920e by Simon Hengel at 2012-02-04T00:50:21+01:00 Add tests for module warnings - - - - - 30ce0d77 by Simon Hengel at 2012-02-04T00:50:29+01:00 Add support for warnings - - - - - bb367960 by Simon Hengel at 2012-02-04T00:50:37+01:00 Add tests for warnings - - - - - 6af1dc2d by Simon Hengel at 2012-02-04T00:50:50+01:00 Expand type signatures in export list (fixes haskell/haddock#192) - - - - - a06cbf25 by Simon Hengel at 2012-02-04T00:51:04+01:00 Expand type signatures for modules without explicit export list - - - - - 57dda796 by Simon Hengel at 2012-02-04T00:51:15+01:00 Remove obsolete TODO - - - - - 270c3253 by David Waern at 2012-02-04T00:51:24+01:00 Fix issues in support for warnings. * Match against local names only. * Simplify (it's OK to map over the warnings). - - - - - 683634bd by David Waern at 2012-02-04T00:55:11+01:00 Some cleanup and make sure we filter warnings through exports. - - - - - 210cb4ca by David Waern at 2012-02-04T03:01:30+01:00 Merge branch 'fix-for-186' of https://github.com/sol/haddock into ghc-7.4 - - - - - e8db9031 by David Waern at 2012-02-04T03:07:51+01:00 Style police. - - - - - 261f9462 by David Waern at 2012-02-04T03:20:16+01:00 Update tests. - - - - - 823cfc7c by David Waern at 2012-02-04T03:21:12+01:00 Use mapM_ in accept.hs as well. - - - - - 873dd619 by David Waern at 2012-02-04T03:21:33+01:00 Remove copy.hs - use accept.hs instead. - - - - - 0e31a14a by David Waern at 2012-02-04T03:47:33+01:00 Use <> instead of mappend. - - - - - 2ff7544f by David Waern at 2012-02-04T03:48:55+01:00 Remove code for older ghc versions. - - - - - dacf2786 by David Waern at 2012-02-04T15:52:51+01:00 Clean up some code from last SoC project. - - - - - 00cbb117 by David Waern at 2012-02-04T21:43:49+01:00 Mostly hlint-inspired cleanup. - - - - - 7dc86cc2 by Simon Peyton Jones at 2012-02-06T09:14:41+00:00 Track changes in HsDecls - - - - - f91f82fe by Ian Lynagh at 2012-02-16T13:40:11+00:00 Follow changes in GHC caused by the CAPI CTYPE pragma - - - - - a0ea6b0b by Ian Lynagh at 2012-02-22T02:26:12+00:00 Follow changes in GHC - - - - - b23b07d1 by Simon Peyton Jones at 2012-03-02T16:36:41+00:00 Follow changes in data representation from the big PolyKinds commit - - - - - 43406022 by Simon Hengel at 2012-03-05T11:18:34+01:00 Save/restore global state for static flags when running GHC actions This is necessary if we want to run createInterfaces (from Documentation.Haddock) multiple times in the same process. - - - - - 9fba16fe by Paolo Capriotti at 2012-03-06T10:57:33+00:00 Update .gitignore. - - - - - a9325044 by Simon Peyton Jones at 2012-03-14T17:35:42+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - fd48065a by Iavor Diatchki at 2012-03-15T22:43:35-07:00 Add support for type-level literals. - - - - - 2e8206dd by Simon Peyton Jones at 2012-03-16T14:18:22+00:00 Follow changes to tcdKindSig (Trac haskell/haddock#5937) - - - - - 93e13319 by Simon Peyton Jones at 2012-03-17T01:04:05+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock Conflicts: src/Haddock/Convert.hs - - - - - d253fa71 by Iavor Diatchki at 2012-03-19T20:12:18-07:00 Merge remote-tracking branch 'origin/master' into type-nats - - - - - fc40acc8 by Iavor Diatchki at 2012-03-19T20:31:27-07:00 Add a missing case for type literals. - - - - - fd2ad699 by Iavor Diatchki at 2012-03-24T13:28:29-07:00 Rename variable to avoid shadowing warning. - - - - - 9369dd3c by Simon Peyton Jones at 2012-03-26T09:14:23+01:00 Follow refactoring of TyClDecl/HsTyDefn - - - - - 38825ca5 by Simon Peyton Jones at 2012-03-26T09:14:37+01:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - 4324ac0f by David Waern at 2012-04-01T01:51:19+02:00 Disable unicode test. - - - - - 3165b750 by David Waern at 2012-04-01T01:51:34+02:00 Take reader environment directly from TypecheckedSource. - - - - - 213b644c by David Waern at 2012-04-01T01:55:20+02:00 Cleanup. - - - - - 3118b4ba by David Waern at 2012-04-01T02:16:15+02:00 Don't filter out unexported names from the four maps - fixes a regression. - - - - - d6524e17 by David Waern at 2012-04-01T02:40:34+02:00 Fix crash when using --qual. Naughty GHC API! - - - - - ea3c43d8 by Henning Thielemann at 2012-04-01T13:03:07+02:00 add QualOption type for distinction between qualification argument given by the user and the actual qualification for a concrete module - - - - - 5422ff05 by Henning Thielemann at 2012-04-01T16:25:02+02:00 emit an error message when the --qual option is used incorrectly - - - - - 026e3404 by David Waern at 2012-04-01T18:10:30+02:00 Don't crash on unicode strings in doc comments. - - - - - ce006632 by David Waern at 2012-04-01T20:13:35+02:00 Add test for --ignore-all-exports flag/ignore-exports pragma. - - - - - 6e4dd33c by David Waern at 2012-04-01T20:21:03+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.4 - - - - - 734ae124 by Henning Thielemann at 2012-04-01T20:22:10+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - 622f9ba5 by David Waern at 2012-04-01T21:26:13+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 55ce17cb by Henning Thielemann at 2012-04-01T22:03:25+02:00 'abbreviate' qualification style - basic support Currently we ignore the package a module is imported from. This means that a module import would shadow another one with the same module name from a different package. - - - - - c85314ef by David Waern at 2012-04-01T22:05:12+02:00 Check qualification option before processing modules. - - - - - ae4b626c by Henning Thielemann at 2012-04-02T00:19:36+02:00 abbreviated qualification: use Packages.lookupModuleInAllPackages for finding the package that a module belongs to - - - - - 60bdbcf5 by Henning Thielemann at 2012-04-02T00:25:31+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - df44301d by Henning Thielemann at 2012-04-02T00:29:05+02:00 qualification style 'abbreviated' -> 'aliased' - - - - - f4192a64 by David Waern at 2012-04-02T01:05:47+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - 7ba09067 by David Terei at 2012-04-04T15:08:21-07:00 Fix reporting of modules safe haskell mode (#5989) - - - - - d0cc33d0 by David Terei at 2012-04-06T15:50:41+01:00 Fix reporting of modules safe haskell mode (#5989) - - - - - 6e3434c5 by Simon Peyton Jones at 2012-04-20T18:37:46+01:00 Track changes in HsSyn - - - - - 22014ed0 by Simon Peyton Jones at 2012-05-11T22:45:15+01:00 Follow changes to LHsTyVarBndrs - - - - - d9a07b24 by David Waern at 2012-05-15T01:46:35+02:00 Merge branch 'ghc-7.4' of http://darcs.haskell.org/haddock into ghc-7.4 - - - - - a6c4ebc6 by David Waern at 2012-05-16T02:18:32+02:00 Update CHANGES. - - - - - 8e181d29 by David Waern at 2012-05-16T02:27:56+02:00 Merge http://code.haskell.org/~thielema/haddock/ into ghc-7.4 - - - - - e358210d by David Waern at 2012-05-16T02:35:33+02:00 Mention the new aliased --qual mode in CHANGES. - - - - - efd36a28 by David Waern at 2012-05-16T21:33:13+02:00 Bump version number. - - - - - d6b3af14 by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for deprecated record field - - - - - 927f800e by Simon Hengel at 2012-05-17T19:08:20+02:00 Use >>= instead of fmap and join - - - - - 048b41d5 by Simon Hengel at 2012-05-17T19:08:20+02:00 newtype-wrap Doc nodes for things that may have warnings attached - - - - - e3a89fc3 by Simon Hengel at 2012-05-17T19:08:20+02:00 Attach warnings to `Documentation` type - - - - - 5d4cc43d by Simon Hengel at 2012-05-17T19:08:20+02:00 Simplify lookupWarning - - - - - cf8ae69d by Simon Hengel at 2012-05-17T19:08:20+02:00 Add test for haskell/haddock#205 - - - - - cb409b19 by Simon Peyton Jones at 2012-05-25T08:30:11+01:00 Follow changes in LHsTyVarBndrs - - - - - 2d5f4179 by Simon Hengel at 2012-05-26T19:21:29+02:00 Add Applicative instance for (GenRnM a) - - - - - e4373060 by Simon Hengel at 2012-05-26T19:21:33+02:00 Use a map for warnings, as suggested by @waern - - - - - 597a68c7 by Simon Hengel at 2012-05-27T08:48:24+02:00 Add an optional label to URLs - - - - - ef1ac7fe by Simon Hengel at 2012-05-27T08:48:24+02:00 Add support for hyperlink labels to parser - - - - - 41f2adce by Simon Hengel at 2012-05-27T08:48:24+02:00 Add golden test for hyperlinks - - - - - 83d5e764 by Simon Hengel at 2012-05-27T08:50:02+02:00 Use LANGUAGE pragmas instead of default-extensions in cabal file - - - - - ddb755e5 by Simon Hengel at 2012-05-27T08:50:02+02:00 Fix typo in comment - - - - - 110676b4 by Simon Hengel at 2012-05-27T08:50:02+02:00 Add a type signature for a where-binding - - - - - 7d9ba2a0 by Ian Lynagh at 2012-06-12T14:38:01+01:00 Follow changes in GHC - - - - - 47c704f2 by Ian Lynagh at 2012-06-12T18:52:16+01:00 Follow changes in GHC - - - - - e1efe1ab by Simon Peyton Jones at 2012-06-13T17:25:29+01:00 Follow changes for the implementation of implicit parameters - - - - - 69abc81c by Ian Lynagh at 2012-06-19T22:52:58+01:00 Follow changes in base - - - - - 9d074a21 by Paolo Capriotti at 2012-06-22T18:26:47+01:00 Use right docMap to get decl documentation. - - - - - e3292ef6 by Ian Lynagh at 2012-07-15T01:31:19+01:00 Follow changes in GHC - - - - - ceae56b0 by Ian Lynagh at 2012-07-16T21:22:48+01:00 Fix haddock following some GHC changes Passing _|_ as the Settings for defaultDynFlags no longer works well enough - - - - - 9df72735 by Paolo Capriotti at 2012-07-19T16:49:32+01:00 Forward port changes from stable. - - - - - 572f5fcf by Ian Lynagh at 2012-07-19T20:38:26+01:00 Merge branch 'master' of darcs.haskell.org:/srv/darcs//haddock - - - - - 9195aca4 by Paolo Capriotti at 2012-07-20T10:27:28+01:00 Update dependencies. - - - - - 33db3923 by Ian Lynagh at 2012-07-20T17:54:43+01:00 Build with GHC 7.7 - - - - - 925a2cea by David Waern at 2012-07-23T16:50:40+02:00 Merge branch 'dev' of https://github.com/sol/haddock into ghc-7.6 Conflicts: src/Haddock/InterfaceFile.hs - - - - - d710ef97 by David Waern at 2012-07-23T16:52:07+02:00 Bump version number. - - - - - eb0c2f83 by David Waern at 2012-07-23T16:57:58+02:00 Update CHANGES. - - - - - b3f56943 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Hide "internal" instances This fixes haskell/haddock#37 (http://trac.haskell.org/haddock/ticket/37) Precisely, we show an instance iff its class and all the types are exported by non-hidden modules. - - - - - a70aa412 by Roman Cheplyaka at 2012-07-27T13:00:13+03:00 Tests for hiding instances (#37) - - - - - c0f4aa58 by Simon Hengel at 2012-07-27T13:00:13+03:00 Add an other test for hiding instances (#37) - - - - - a7ed6268 by Ian Lynagh at 2012-08-07T14:48:13+01:00 Follow changes in GHC - - - - - 0ab30d38 by Ian Lynagh at 2012-08-13T22:12:27+01:00 Improve haddock memory usage - - - - - 0eaa4e30 by Ian Lynagh at 2012-08-13T23:58:46+01:00 Improve haddock memory usage - - - - - 659d26cf by Ian Lynagh at 2012-08-14T13:16:48+01:00 Remove some temporary pragmas I accidentally recorded - - - - - d97fceb6 by Simon Hengel at 2012-08-25T13:19:34+02:00 Add missing dependency to library - - - - - 4c910697 by Simon Hengel at 2012-08-28T07:39:14+02:00 Move .ghci to project root - - - - - fc3c601a by Simon Hengel at 2012-08-28T07:39:14+02:00 accept.hs: Ignore some files - - - - - 1af9b984 by Simon Hengel at 2012-08-28T07:40:04+02:00 Update reference renderings (bump version) - - - - - 980dc253 by Simon Hengel at 2012-08-28T07:40:32+02:00 Update reference renderings (remove links for ()) - - - - - 33651dbf by Simon Hengel at 2012-08-28T07:41:50+02:00 Update documentation of `runInteractiveProcess` in reference rendering - - - - - 7ab25078 by David Waern at 2012-09-07T10:38:50+02:00 Merge branch 'hiddenInstances2' of http://github.com/feuerbach/haddock into ghc-7.6 - - - - - c3de3a4b by David Waern at 2012-09-07T14:29:27+02:00 Follow changes in GHC. - - - - - 298c43ac by David Waern at 2012-09-07T14:59:24+02:00 Update CHANGES. - - - - - e797993a by David Waern at 2012-09-07T15:21:30+02:00 Update ANNOUNCE. - - - - - d0b44790 by David Waern at 2012-09-07T15:22:43+02:00 Merge branch 'hidden-instances' into ghc-7.6 - - - - - 41a4adc8 by Simon Hengel at 2012-09-08T12:08:37+02:00 Update doc/README - - - - - 71ad1040 by Simon Hengel at 2012-09-08T12:17:17+02:00 Add documentation for URL labels - - - - - 9bb41afd by Simon Peyton Jones at 2012-09-20T18:14:26+01:00 Follow data type changes in the tc-untouchables branch Relating entirely to SynTyConRhs - - - - - b8139bfa by Simon Hengel at 2012-09-21T14:24:16+02:00 Disable Unicode test for now - - - - - a5fafdd7 by Simon Hengel at 2012-09-21T14:35:45+02:00 Update TypeOperators test for GHC 7.6.1 Type operators can't be used as type variables anymore! - - - - - 6ccf0025 by Simon Hengel at 2012-09-21T16:02:24+02:00 Remove (Monad (Either e)) instance from ref. rendering of CrossPackageDocs I do not really understand why the behavior changed, so I'll open a ticket, so that we can further investigate. - - - - - b5c6c138 by Ian Lynagh at 2012-09-27T02:00:57+01:00 Follow changes in GHC build system - - - - - b98eded0 by David Waern at 2012-09-27T15:37:02+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 76cc2051 by David Waern at 2012-09-27T15:48:19+02:00 Update hidden instances tests. - - - - - aeaa1c59 by David Waern at 2012-09-28T10:21:32+02:00 Make API buildable with GHC 7.6. - - - - - d76be1b0 by Simon Peyton Jones at 2012-09-28T15:57:05+01:00 Merge remote-tracking branch 'origin/master' into tc-untouchables - - - - - a1922af8 by David Waern at 2012-09-28T19:50:20+02:00 Fix spurious superclass constraints bug. - - - - - bc41bdbb by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove old examples - - - - - bed7d3dd by Simon Hengel at 2012-10-01T11:30:51+02:00 Adapt parsetests for GHC 7.6.1 - - - - - dcdb22bb by Simon Hengel at 2012-10-01T11:30:51+02:00 Add test-suite section for parsetests to cabal file + get rid of HUnit dependency - - - - - 1e5263c9 by Simon Hengel at 2012-10-01T11:30:51+02:00 Remove test flag from cabal file This was not really used. - - - - - 4beee98b by David Waern at 2012-09-28T23:42:28+02:00 Merge branch 'ghc-7.6' of http://darcs.haskell.org/haddock into ghc-7.6 - - - - - 11dd2256 by Ian Lynagh at 2012-10-03T16:17:35+01:00 Follow change in GHC build system - - - - - fbd77962 by Simon Hengel at 2012-10-03T18:49:40+02:00 Remove redundant dependency from cabal file - - - - - 09218989 by Simon Hengel at 2012-10-04T16:03:05+02:00 Fix typo - - - - - 93a2d5f9 by Simon Hengel at 2012-10-04T16:11:41+02:00 Remove trailing whitespace from cabal file - - - - - c8b46cd3 by Simon Hengel at 2012-10-04T16:12:17+02:00 Export Haddock's main entry point from library - - - - - b411e77b by Simon Hengel at 2012-10-04T16:29:46+02:00 Depend on library for executable The main motivation for this is to increase build speed. In GHC's source tree the library is not build, but all modules are now required for the executable, so that GHC's validate will now detect build failures for the library. - - - - - f8f0979f by Simon Hengel at 2012-10-05T00:32:57+02:00 Set executable flag for Setup.lhs - - - - - dd045998 by Simon Hengel at 2012-10-07T16:44:06+02:00 Extend rather than set environment when running HTML tests On some platforms (e.g. ppc64) GHC requires gcc in the path. - - - - - 7b39c3ae by Simon Hengel at 2012-10-07T17:05:45+02:00 cross-package test: re-export IsString instead of Monad There is a monad instance for Q, which is not available on platforms that do not have GHCi support. This caused CrossPackageDocs to fail on those platforms. Re-exporting IsString should test the same thing, but it works on all platforms. - - - - - 0700c605 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Fix some warnings - - - - - f78eca79 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Make -Wall proof - - - - - 6beec041 by Simon Hengel at 2012-10-07T19:06:34+02:00 runtests.hs: Use listToMaybe/fromMaybe instead of safeHead/maybe - - - - - 44b8ce86 by Ian Lynagh at 2012-10-08T21:59:46+01:00 Follow changes in GHC - - - - - 6da5f702 by Simon Hengel at 2012-10-09T11:16:19+02:00 Update .ghci - - - - - 9ac1a1b9 by Kazu Yamamoto at 2012-10-09T12:45:31+02:00 Add markup support for properties - - - - - 1944cb42 by Simon Hengel at 2012-10-09T12:45:31+02:00 Simplify lexing/parsing of properties In contrast to what we do for examples, we do not really need to capture the "prompt" here. - - - - - bffd8e62 by Simon Hengel at 2012-10-09T13:40:14+02:00 Add HTML test for properties - - - - - 2fe9c5cb by Simon Hengel at 2012-10-09T13:40:21+02:00 Add unit tests for properties - - - - - 874e361b by Simon Hengel at 2012-10-09T13:40:33+02:00 Bump interface version - - - - - 2506cc37 by Simon Hengel at 2012-10-09T15:15:04+02:00 Fix parser bug - - - - - 743d2b7d by Simon Hengel at 2012-10-09T15:31:06+02:00 Allow to load interface files with compatible versions - - - - - 981a1660 by Simon Hengel at 2012-10-10T10:32:05+02:00 Export more types from Documentation.Haddock (fixes haskell/haddock#216) - - - - - dff7dc76 by Simon Hengel at 2012-10-10T11:15:19+02:00 Update ANNOUNCE and CHANGES - - - - - edd2bb01 by Simon Hengel at 2012-10-10T11:22:50+02:00 Bump version - - - - - 5039163b by Simon Hengel at 2012-10-10T13:56:04+02:00 Fix typo in documentation - - - - - e4ce34da by Simon Hengel at 2012-10-10T14:28:35+02:00 Add documentation for properties - - - - - 9555ebca by Simon Hengel at 2012-10-11T10:49:04+02:00 Remove redundant if-defs, more source documentation - - - - - 87aa67e1 by Simon Hengel at 2012-10-11T12:32:51+02:00 Adapt cabal file - - - - - c44c1dee by Simon Hengel at 2012-10-11T12:41:58+02:00 Require ghc 7.6 - - - - - 8383bc34 by Simon Hengel at 2012-10-11T12:50:24+02:00 Bump version - - - - - 1030eb38 by Simon Hengel at 2012-10-11T12:55:44+02:00 Update ANNOUNCE and CHANGES - - - - - 74955088 by Simon Hengel at 2012-10-12T09:49:31+02:00 Improve note about `binaryInterfaceVersion` (thanks David) - - - - - ee30f6b7 by Simon Hengel at 2012-10-13T13:40:59+02:00 Update version in html tests, rpm spec file, and user manual - - - - - f2861f18 by Simon Hengel at 2012-10-13T14:40:33+02:00 Remove unused MonadFix constraint - - - - - dfdf1a74 by Simon Hengel at 2012-10-13T15:15:38+02:00 Minor code simplification - - - - - 4ecd1e70 by Simon Hengel at 2012-10-13T15:33:43+02:00 Increase code locality - - - - - f7df5cc9 by Simon Hengel at 2012-10-13T16:03:12+02:00 Minor code simplification - - - - - e737eb6e by Simon Hengel at 2012-10-13T19:03:04+02:00 Handle HsExplicitListTy in renameer (fixes haskell/haddock#213) - - - - - c2dc8f17 by Simon Hengel at 2012-10-13T20:46:31+02:00 Better error messages - - - - - 14d48b4c by Simon Hengel at 2012-10-14T00:21:07+02:00 Simplify RnM type - - - - - 6c2cc547 by Simon Hengel at 2012-10-14T00:23:35+02:00 Simplify lookupRn - - - - - bc77ce85 by Simon Hengel at 2012-10-14T01:51:32+02:00 Organize unite tests hierarchically - - - - - 2306d117 by Simon Hengel at 2012-10-14T10:34:58+02:00 Handle more cases in renameType - - - - - 8a864203 by Simon Hengel at 2012-10-14T11:47:59+02:00 Add mini_HiddenInstances.html.ref and mini_HiddenInstancesB.html.ref - - - - - 3a978eca by Simon Hengel at 2012-10-14T11:49:28+02:00 Add /tests/html-tests/output/ to .gitignore - - - - - db18888a by Simon Hengel at 2012-10-14T13:38:21+02:00 Allow haddock markup in deprecation messages - - - - - e7cfee9f by Simon Hengel at 2012-10-14T14:00:23+02:00 If parsing of deprecation message fails, include it verbatim - - - - - 242a85be by Simon Hengel at 2012-10-14T14:13:24+02:00 Add description for PruneWithWarning test - - - - - 43d33df1 by Simon Hengel at 2012-10-14T15:40:53+02:00 Minor formatting change - - - - - 22768c44 by Simon Hengel at 2012-10-14T16:03:43+02:00 Properly handle deprecation messages for re-exported things (fixes haskell/haddock#220) - - - - - cb4b9111 by Simon Hengel at 2012-10-14T17:30:28+02:00 Add build artifacts for documentation to .gitignore - - - - - 854cd8de by Simon Hengel at 2012-10-14T23:34:51+02:00 unit-tests: Improve readability Add IsString instance for (Doc RdrName) + use <> instead of DocAppend. - - - - - c4446d54 by Simon Hengel at 2012-10-14T23:37:21+02:00 unit-tests: Minor refactoring Rename parse to parseParas. - - - - - 04f2703c by Simon Hengel at 2012-10-15T00:36:42+02:00 Fix typo - - - - - 3d109e44 by Simon Hengel at 2012-10-15T10:30:07+02:00 Add description for DeprecatedReExport test - - - - - 84f0985c by Simon Hengel at 2012-10-15T14:54:19+02:00 Move resources to /resources directory - - - - - a5de7ca6 by Simon Hengel at 2012-10-15T15:46:18+02:00 Move HTML tests to directory /html-test/ - - - - - e21f727d by Simon Hengel at 2012-10-15T19:32:42+02:00 Move HTML reference renderings to /html-test/ref/ - - - - - 3a3c6c75 by Simon Hengel at 2012-10-15T19:32:42+02:00 Copy css, images, etc. on accept - - - - - 40ead6dc by Simon Hengel at 2012-10-15T19:32:42+02:00 Move unit tests to /test directory - - - - - 99a28231 by Simon Hengel at 2012-10-15T19:32:42+02:00 Fix Setup.lhs /usr/bin/runhaskell is not installed on all systems. - - - - - 95faf45e by Simon Hengel at 2012-10-15T19:32:42+02:00 Make test management scripts more robust * They are now independent from the current directory, and hence can be called from everywhere * On UNIX/Linux they can now be run as scripts - - - - - 027aaa2d by Simon Hengel at 2012-10-15T19:53:40+02:00 Add 'dev' flag to cabal file, that builds without -O2 That way --disable-optimization can be used, which decreases build time considerably. - - - - - e0266ede by Simon Hengel at 2012-10-15T20:03:43+02:00 Add test case for "spurious superclass constraints bug" - - - - - 52a2aa92 by Simon Hengel at 2012-10-15T20:28:55+02:00 Adapt accept.lhs, so that it ignores more index files - - - - - 53530781 by Simon Hengel at 2012-10-15T20:49:39+02:00 Rename html-test/runtests.lhs to html-test/run.lhs - - - - - 84518797 by Simon Hengel at 2012-10-15T20:49:39+02:00 Move source files for HTML tests to html-test/src - - - - - a911dc6c by Simon Hengel at 2012-10-15T20:49:39+02:00 Adapt output directory for HTML tests - - - - - d3c15857 by Ian Lynagh at 2012-10-16T16:54:43+01:00 Follow dopt->gopt rename - - - - - 956665a5 by Simon Hengel at 2012-10-18T08:42:48+02:00 Update html-test/README - - - - - 903b1029 by Simon Hengel at 2012-10-18T08:50:26+02:00 Use markdown for html-test/README - - - - - 150b4d63 by Ian Lynagh at 2012-10-18T16:36:00+01:00 Follow changes in GHC: 'flags' has been renamed 'generalFlags' - - - - - 41e04ff9 by Simon Hengel at 2012-11-28T09:54:35+01:00 Export missing types from Documentation.Haddock - - - - - 9be59237 by Ian Lynagh at 2012-11-30T23:20:47+00:00 Update dependencies - - - - - e06842f5 by Simon Hengel at 2012-12-07T20:58:05+01:00 Bump version - - - - - e3dbede0 by Simon Hengel at 2012-12-07T20:58:05+01:00 Add missing test files to cabal file (fixes haskell/haddock#230) - - - - - ee0dcca7 by Simon Hengel at 2012-12-07T20:58:05+01:00 Update CHANGES - - - - - 51601bdb by Simon Peyton Jones at 2012-12-19T17:28:35+00:00 Track changes in UNPACK pragma stuff - - - - - f2573bc1 by Richard Eisenberg at 2012-12-21T20:56:25-05:00 Implement overlapping type family instances. An ordered, overlapping type family instance is introduced by 'type instance where', followed by equations. See the new section in the user manual (7.7.2.2) for details. The canonical example is Boolean equality at the type level: type family Equals (a :: k) (b :: k) :: Bool type instance where Equals a a = True Equals a b = False A branched family instance, such as this one, checks its equations in order and applies only the first the matches. As explained in the note [Instance checking within groups] in FamInstEnv.lhs, we must be careful not to simplify, say, (Equals Int b) to False, because b might later unify with Int. This commit includes all of the commits on the overlapping-tyfams branch. SPJ requested that I combine all my commits over the past several months into one monolithic commit. The following GHC repos are affected: ghc, testsuite, utils/haddock, libraries/template-haskell, and libraries/dph. Here are some details for the interested: - The definition of CoAxiom has been moved from TyCon.lhs to a new file CoAxiom.lhs. I made this decision because of the number of definitions necessary to support BranchList. - BranchList is a GADT whose type tracks whether it is a singleton list or not-necessarily-a-singleton-list. The reason I introduced this type is to increase static checking of places where GHC code assumes that a FamInst or CoAxiom is indeed a singleton. This assumption takes place roughly 10 times throughout the code. I was worried that a future change to GHC would invalidate the assumption, and GHC might subtly fail to do the right thing. By explicitly labeling CoAxioms and FamInsts as being Unbranched (singleton) or Branched (not-necessarily-singleton), we make this assumption explicit and checkable. Furthermore, to enforce the accuracy of this label, the list of branches of a CoAxiom or FamInst is stored using a BranchList, whose constructors constrain its type index appropriately. I think that the decision to use BranchList is probably the most controversial decision I made from a code design point of view. Although I provide conversions to/from ordinary lists, it is more efficient to use the brList... functions provided in CoAxiom than always to convert. The use of these functions does not wander far from the core CoAxiom/FamInst logic. BranchLists are motivated and explained in the note [Branched axioms] in CoAxiom.lhs. - The CoAxiom type has changed significantly. You can see the new type in CoAxiom.lhs. It uses a CoAxBranch type to track branches of the CoAxiom. Correspondingly various functions producing and consuming CoAxioms had to change, including the binary layout of interface files. - To get branched axioms to work correctly, it is important to have a notion of type "apartness": two types are apart if they cannot unify, and no substitution of variables can ever get them to unify, even after type family simplification. (This is different than the normal failure to unify because of the type family bit.) This notion in encoded in tcApartTys, in Unify.lhs. Because apartness is finer-grained than unification, the tcUnifyTys now calls tcApartTys. - CoreLinting axioms has been updated, both to reflect the new form of CoAxiom and to enforce the apartness rules of branch application. The formalization of the new rules is in docs/core-spec/core-spec.pdf. - The FamInst type (in types/FamInstEnv.lhs) has changed significantly, paralleling the changes to CoAxiom. Of course, this forced minor changes in many files. - There are several new Notes in FamInstEnv.lhs, including one discussing confluent overlap and why we're not doing it. - lookupFamInstEnv, lookupFamInstEnvConflicts, and lookup_fam_inst_env' (the function that actually does the work) have all been more-or-less completely rewritten. There is a Note [lookup_fam_inst_env' implementation] describing the implementation. One of the changes that affects other files is to change the type of matches from a pair of (FamInst, [Type]) to a new datatype (which now includes the index of the matching branch). This seemed a better design. - The TySynInstD constructor in Template Haskell was updated to use the new datatype TySynEqn. I also bumped the TH version number, requiring changes to DPH cabal files. (That's why the DPH repo has an overlapping-tyfams branch.) - As SPJ requested, I refactored some of the code in HsDecls: * splitting up TyDecl into SynDecl and DataDecl, correspondingly changing HsTyDefn to HsDataDefn (with only one constructor) * splitting FamInstD into TyFamInstD and DataFamInstD and splitting FamInstDecl into DataFamInstDecl and TyFamInstDecl * making the ClsInstD take a ClsInstDecl, for parallelism with InstDecl's other constructors * changing constructor TyFamily into FamDecl * creating a FamilyDecl type that stores the details for a family declaration; this is useful because FamilyDecls can appear in classes but other decls cannot * restricting the associated types and associated type defaults for a * class to be the new, more restrictive types * splitting cid_fam_insts into cid_tyfam_insts and cid_datafam_insts, according to the new types * perhaps one or two more that I'm overlooking None of these changes has far-reaching implications. - The user manual, section 7.7.2.2, is updated to describe the new type family instances. - - - - - f788d0fb by Simon Peyton Jones at 2012-12-23T15:49:58+00:00 Track changes in HsBang - - - - - ca460a0c by Simon Peyton Jones at 2012-12-23T15:50:28+00:00 Merge branch 'master' of http://darcs.haskell.org//haddock - - - - - f078fea6 by Simon Peyton Jones at 2013-01-02T08:33:13+00:00 Use InstEnv.instanceSig rather than instanceHead (name change) - - - - - 88e41305 by Simon Peyton Jones at 2013-01-14T17:10:27+00:00 Track change to HsBang type - - - - - e1ad4e19 by Kazu Yamamoto at 2013-02-01T11:59:24+09:00 Merge branch 'ghc-7.6' into ghc-7.6-merge-2 Conflicts: haddock.cabal src/Haddock/Interface/AttachInstances.hs src/Haddock/Interface/Create.hs src/Haddock/Interface/LexParseRn.hs src/Haddock/InterfaceFile.hs src/Haddock/Types.hs Only GHC HEAD can compile this. GHC 7.6.x cannot compile this. Some test fail. - - - - - 62bec012 by Kazu Yamamoto at 2013-02-06T11:12:28+09:00 Using tcSplitSigmaTy in instanceHead' (FIXME is resolved.) - - - - - 013fd2e4 by Kazu Yamamoto at 2013-02-06T17:56:21+09:00 Refactoring instanceHead'. - - - - - 3148ce0e by Kazu Yamamoto at 2013-02-07T17:45:10+09:00 Using new syntax in html-test/src/GADTRecords.hs. - - - - - 626dabe7 by Gabor Greif at 2013-02-15T22:42:01+01:00 Typo - - - - - 1eb667ae by Ian Lynagh at 2013-02-16T17:02:07+00:00 Follow changes in base - - - - - 3ef8253a by Ian Lynagh at 2013-03-01T23:23:57+00:00 Follow changes in GHC's build system - - - - - 1a265a3c by Ian Lynagh at 2013-03-03T23:12:07+00:00 Follow changes in GHC build system - - - - - 69941c79 by Max Bolingbroke at 2013-03-10T09:38:28-07:00 Use Alex 3's Unicode support to properly lex source files as UTF-8 Signed-off-by: David Waern <david.waern at gmail.com> - - - - - ea687dad by Simon Peyton Jones at 2013-03-15T14:16:10+00:00 Adapt to tcRnGetInfo returning family instances too This API change was part of the fix to Trac haskell/haddock#4175. But it offers new information to Haddock: the type-family instances, as well as the class instances, of this type. This patch just drops the new information on the floor, but there's an open opportunity to use it in the information that Haddock displays. - - - - - 971a30b0 by Andreas Voellmy at 2013-05-19T20:47:39+01:00 Fix for haskell/haddock#7879. Changed copy of utils/haddock/html/resources/html to use "cp -RL" rather than "cp -R". This allows users to run validate in a build tree, where the build tree was setup using lndir with a relative path to the source directory. - - - - - 31fb7694 by Ian Lynagh at 2013-05-19T20:47:49+01:00 Use "cp -L" when making $(INPLACE_LIB)/latex too - - - - - e9952233 by Simon Hengel at 2013-06-01T18:06:50+02:00 Add -itest to .ghci - - - - - b06873b3 by Mateusz Kowalczyk at 2013-06-01T18:06:50+02:00 Workaround for a failing build with --enable-tests. - - - - - e7858d16 by Simon Hengel at 2013-06-01T19:29:28+02:00 Fix broken test - - - - - 0690acb1 by Richard Eisenberg at 2013-06-21T14:08:25+01:00 Updates to reflect changes in HsDecls to support closed type families. - - - - - 7fd347ec by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 53ed81b6 by Simon Hengel at 2013-07-08T10:28:48+02:00 Fix failing test - - - - - 931c4f4f by Richard Eisenberg at 2013-07-24T13:15:59+01:00 Remove (error "synifyKind") to use WithinType, to allow haddock to process base. - - - - - 55a9c804 by Richard Eisenberg at 2013-08-02T15:54:55+01:00 Changes to reflect changes in GHC's type HsTyVarBndr - - - - - b6e9226c by Mathieu Boespflug at 2013-08-04T10:39:43-07:00 Output Copright and License keys in Xhtml backend. This information is as relevant in the documentation as it is in the source files themselves. Signed-off-by: David Waern <david.waern at gmail.com> - - - - - 4c66028a by David Waern at 2013-08-04T15:27:36-07:00 Bump interface file version. - - - - - 67340163 by David Waern at 2013-08-09T16:12:51-07:00 Update tests. - - - - - 2087569b by Mateusz Kowalczyk at 2013-08-25T09:24:13+02:00 Add spec tests. This adds tests for all elements we can create during regular parsing. This also adds tests for text with unicode in it. - - - - - 97f36a11 by Mateusz Kowalczyk at 2013-08-27T06:59:12+01:00 Fix ticket haskell/haddock#247. I do the same thing that the XHTML backend does: give these no special treatment and just act as if they are regular functions. - - - - - 60681b4f by Mateusz Kowalczyk at 2013-08-27T21:22:48+02:00 LaTeX tests setup - - - - - fa4c27b2 by Mateusz Kowalczyk at 2013-09-02T23:21:43+01:00 Fixes haskell/haddock#253 - - - - - 1a202490 by Mateusz Kowalczyk at 2013-09-03T01:12:50+01:00 Use Hspec instead of nanospec This is motivated by the fact that Haddock tests are not ran by the GHC's ‘validate’ script so we're pretty liberal on dependencies in that area. Full Hspec gives us some nice features such as Quickcheck integration. - - - - - 8cde3b20 by David Luposchainsky at 2013-09-08T07:27:28-05:00 Fix AMP warnings Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - d10661f2 by Herbert Valerio Riedel at 2013-09-11T15:15:01+02:00 Update Git repo URL in `.cabal` file - - - - - 16a44eb5 by Richard Eisenberg at 2013-09-17T09:34:26-04:00 Revision to reflect new role annotation syntax in GHC. - - - - - 4b9833b9 by Herbert Valerio Riedel at 2013-09-18T10:15:28+02:00 Add missing `traverse` method for `GenLocated` As `Traversable` needs at least one of `traverse` or `sequenceA` to be overridden. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - b71fed5d by Simon Hengel at 2013-09-18T22:43:34+02:00 Add test helper - - - - - 4fc1ea86 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#231 - - - - - 435872f6 by Mateusz Kowalczyk at 2013-09-18T22:43:34+02:00 Fixes haskell/haddock#256 We inject -dynamic-too into flags before we run all our actions in the GHC monad. - - - - - b8b24abb by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Add new field to DynFlags - - - - - 49558795 by Simon Hengel at 2013-09-18T22:43:35+02:00 Fallback to ./resources when Cabal data is not found (so that themes are found during development) - - - - - bf79d05c by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Fixes haskell/haddock#5 - - - - - e1baebc2 by Mateusz Kowalczyk at 2013-09-18T22:43:35+02:00 Print missing documentation. Fixes haskell/haddock#258. - - - - - 02ea74de by Austin Seipp at 2013-10-09T10:52:22-05:00 Don't consider StaticFlags when parsing arguments. Instead, discard any static flags before parsing the command line using GHC's DynFlags parser. See http://ghc.haskell.org/trac/ghc/ticket/8276 Based off a patch from Simon Hengel. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 704fd5bb by Simon Hengel at 2013-11-09T00:15:13+01:00 Update HTML tests - - - - - f9fed49e by Simon Hengel at 2013-11-10T18:43:58+01:00 Bump version - - - - - 97ae1999 by Simon Peyton Jones at 2013-11-25T17:25:14+00:00 Track changes in HsSpliceTy data constructor - - - - - 59ad8268 by Simon Peyton Jones at 2014-01-10T18:17:43+00:00 Adapt to small change in Pretty's exports - - - - - 8b12e6aa by Simon Hengel at 2014-01-12T14:48:35-06:00 Some code simplification by using traverse - - - - - fc5ea9a2 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix warnings in test helper - - - - - 6dbb3ba5 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Add ByteString version of Attoparsec - - - - - 968d7774 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 One pass parser and tests. We remove the HTML test as it is no longer necessary. We cover the test case in spec tests and other HTML tests but keeping this around fails: this is because the new parser has different semantics there. In fact, I suspect the original behaviour was a bug that wasn't caught/fixed but simply included as-is during the testing. - - - - - 37a07c9c by Simon Hengel at 2014-01-12T14:48:35-06:00 Rename Haddock.ParseSpec to Haddock.ParserSpec - - - - - f0f68fe9 by Simon Hengel at 2014-01-12T14:48:35-06:00 Don't append newline to parseString input We also check that we have parsed everything with endOfInput. - - - - - 95d60093 by Simon Hengel at 2014-01-12T14:48:35-06:00 Fix totality, unicode, examples, paragraph parsing Also simplify specs and parsers while we're at it. Some parsers were made more generic. This commit is a part of GHC pre-merge squash, email fuuzetsu at fuuzetsu.co.uk if you need the full commit history. - - - - - 7d99108c by Simon Hengel at 2014-01-12T14:48:35-06:00 Update acceptance tests - - - - - d1b59640 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Support for bold. Conflicts: src/Haddock/Backends/Hoogle.hs src/Haddock/Interface/Rename.hs src/Haddock/Parser.hs - - - - - 4b412b39 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Allow for headings inside function documentation. LaTeX will treat the h3-h6 headings the same as we'd have to hack the style file heavily otherwise and it would make the headings tiny anyway. Hoogle upstream said they will put in the functionality on their end. Conflicts: src/Haddock/Interface/Rename.hs src/Haddock/Types.hs test/Haddock/ParserSpec.hs - - - - - fdcca428 by Mateusz Kowalczyk at 2014-01-12T14:48:35-06:00 Per-module extension flags and language listing. Any extensions that are not enabled by a used language (Haskell2010 &c) will be shown. Furthermore, any implicitly enabled are also going to be shown. While we could eliminate this either by using the GHC API or a dirty hack, I opted not to: if a user doesn't want the implied flags to show, they are recommended to use enable extensions more carefully or individually. Perhaps this will encourage users to not enable the most powerful flags needlessly. Enabled with show-extensions. Conflicts: src/Haddock/InterfaceFile.hs - - - - - 368942a2 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Bump interface version There were some breaking changes over the last few patches so we bump the interface version. This causes a big problem with testing: 1. To generate cross package docs, we first need to generate docs for the package used. 2. To generate package docs with new interface version, we need to use Haddock which has the version bumped. 3. To get Haddock with the version bump, we first need to test cross package docs 4. GOTO 1 So the problem is the chicken and the egg problem. It seems that the only solution would be to generate some interface files on the fly but it is non-trivial. To run this test, you'll have to: * build Haddock without the test (make sure everything else passes) * rebuild the packages used in the test with your shiny new binary making sure they are visible to Haddock * remove the ‘_hidden’ suffix and re-run the tests Note: because the packages currently used for this test are those provided by GHC, it's probably non-trivial to just re-build them. Preferably something less tedious to rebuild should be used and something that is not subject to change. - - - - - 124ae7a9 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow for nesting of paragraphs under lists. The nesting rules are similar to Markdown's with the exception that we can not simply indent the first line of a hard wrapped indented paragraph and have it treated as if it was fully indented. The reason is differences in markup as some of our constructs care about whitespace while others just swallow everything up so it's just a lot easier to not bother with it rather than making arbitrary rules. Note that we now drop trailing for string entities inside of lists. They weren't needed and it makes the output look uniform whether we use a single or double newline between list elements. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - c7913535 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Allow escaping in URLs and pictures. Some tests were moved under parseString as they weren't about paragraph level markup. Conflicts: src/Haddock/Parser.hs test/Haddock/ParserSpec.hs - - - - - 32326680 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update documentation. - - - - - fbef6406 by Mateusz Kowalczyk at 2014-01-12T14:48:36-06:00 Update maintainer - - - - - b40e82f4 by Mateusz Kowalczyk at 2014-01-13T02:39:25-06:00 Fixes haskell/haddock#271 Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - f4eafbf8 by Gergő Érdi at 2014-01-19T15:35:16-06:00 Support for -XPatternSynonyms Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - a8939591 by Austin Seipp at 2014-01-29T08:09:04-06:00 Update CPP check for __GLASGOW_HASKELL__ Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 30d7e9d5 by Gergő Érdi at 2014-01-31T00:15:01+08:00 <+>: Don't insert a space when concatenating empty nodes - - - - - a25ccd4d by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Fix @ code blocks In cases where we had some horizontal space before the closing ‘@’, the parser would not accept the block as a code block and we'd get ugly output. - - - - - 0f67305a by Mateusz Kowalczyk at 2014-01-30T17:22:34+01:00 Update tests This updates tests due to Haddock Trac haskell/haddock#271 fix and due to removal of TypeHoles as an extension from GHC. - - - - - 157322a7 by Gergő Érdi at 2014-01-31T01:03:17+08:00 Handle infix vs prefix names correctly everywhere, by explicitly specifying the context The basic idea is that "a" and "+" are either pretty-printed as "a" and "(+)" or "`a`" and "+" - - - - - aa6d9685 by Mateusz Kowalczyk at 2014-01-30T17:21:50+00:00 Correct whitespace in ‘hidden’ test for <+> change - - - - - 121872f0 by Mateusz Kowalczyk at 2014-02-09T17:59:12+00:00 Document module header. Fixes Haddock Trac haskell/haddock#270. - - - - - e3253746 by Mateusz Kowalczyk at 2014-02-10T21:37:48+00:00 Insert a space between module link and description Fixes Haddock Trac haskell/haddock#277. - - - - - 771d2384 by Mateusz Kowalczyk at 2014-02-10T23:27:21+00:00 Ensure a space between type signature and ‘Source’ This is briefly related to Haddock Trac haskell/haddock#249 and employs effectively the suggested fix _but_ it doesn't actually fix the reported issue. This commit simply makes copying the full line a bit less of a pain. - - - - - 8cda9eff by nand at 2014-02-11T15:48:30+00:00 Add support for type/data families This adds support for type/data families with their respective instances, as well as closed type families and associated type/data families. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 3f22c510 by nand at 2014-02-11T15:53:50+00:00 Improve display of poly-kinded type operators This now displays them as (==) k a b c ... to mirror GHC's behavior, instead of the old (k == a) b c ... which was just wrong. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - effb2d6b by nand at 2014-02-11T15:56:50+00:00 Add test case for PatternSynonyms This just tests various stuff including poly-kinded patterns and operator patterns to make sure the rendering isn't broken. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - b38faf0d by Niklas Haas at 2014-02-13T21:53:32+00:00 Get rid of re-implementation of sortBy I have no idea what this was doing lying around here, and due to the usage of tuples it's actually slower, too. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - ac1e0413 by Mateusz Kowalczyk at 2014-02-13T23:57:16+00:00 Only warn about missing docs when docs are missing This fixes the ‘Missing documentation for…’ message for modules with 100% coverage. - - - - - cae2e36a by Niklas Haas at 2014-02-15T21:56:18+00:00 Add test case for inter-module type/data family instances These should show up in every place where the class is visible, and indeed they do right now. Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - 8bea5c3a by Mateusz Kowalczyk at 2014-02-19T05:11:34+00:00 Use a bespoke data type to indicate fixity This deals with what I imagine was an ancient TODO and makes it much clearer what the argument actually does rather than having the user chase down the comment. - - - - - 5b52d57c by Niklas Haas at 2014-02-22T21:31:03+01:00 Strip a single leading space from bird tracks (#201) This makes bird tracks in the form > foo > bar > bat parse as if they had been written as >foo >bar >bat ie. without the leading whitespace in front of every line. Ideally we also want to look into how leading whitespace affects code blocks written using the @ @ syntax, which are currently unaffected by this patch. - - - - - 5a1315a5 by Simon Hengel at 2014-02-22T21:55:35+01:00 Turn a source code comment into specs - - - - - 784cfe58 by Mateusz Kowalczyk at 2014-02-23T05:02:22+00:00 Update test case for lifted GADT type rendering The parsing of these seems to have been fixed by GHC folk and it now renders differently. IMHO it now renders in a better way so I'm updating the test to reflect this. - - - - - c3c88c2f by Mateusz Kowalczyk at 2014-02-23T06:37:14+00:00 Don't shadow ‘strip’. -Wall complains - - - - - 293031d8 by Niklas Haas at 2014-02-23T15:21:52+01:00 Make ImplicitParams render correctly (#260) This introduces a new precedence level for single contexts (because implicit param contexts always need parens around them, but other types of contexts don't necessarily, even when alone) - - - - - 4200842d by Niklas Haas at 2014-02-23T15:37:13+01:00 Lower precedence of equality constraints This drops them to the new precedence pREC_CTX, which makes single eqaulity constraints show up as (a ~ b) => ty, in line with GHC's rendering. Additional tests added to make sure other type operators render as intended. Current behavior matches GHC - - - - - b59e3227 by Niklas Haas at 2014-02-23T16:11:22+01:00 Add RankNTypes test case to ImplicitParams.hs This test actually tests what haskell/haddock#260 originally reported - I omitted the RankNTypes scenario from the original fix because I realized it's not relevant to the underlying issue and indeed, this renders as intended now. Still good to have more tests. - - - - - c373dbf7 by Mateusz Kowalczyk at 2014-02-24T06:09:54+00:00 Fix rendering of Contents when links are present Fixes Haddock Trac haskell/haddock#267. - - - - - 9ecb0e56 by Mateusz Kowalczyk at 2014-02-24T06:26:50+00:00 Fix wording in the docs - - - - - 4f4dcd8e by Mateusz Kowalczyk at 2014-02-27T03:00:33+00:00 Change rendering of duplicate record field docs See Haddock Trac haskell/haddock#195. We now change this behaviour to only rendering the documentation attached to the first instance of a duplicate field. Perhaps we could improve this by rendering the first instance that has documentation attached to it but for now, we'll stick with this. - - - - - ad8aa609 by Niklas Haas at 2014-03-08T09:43:26+01:00 Render fixity information Affects functions, type synonyms, type families, class names, data type names, constructors, data families, associated TFs/DFs, type synonyms, pattern synonyms and everything else I could think of. - - - - - 6a39c917 by Niklas Haas at 2014-03-09T07:43:39+01:00 Reorder topDeclElem to move the source/wiki links to the top They appear in the same position due to the float: right attribute but now they're always at the top of the box instead of at the bottom. - - - - - 2d34b3b4 by Niklas Haas at 2014-03-09T07:53:46+01:00 Use optLast instead of listToMaybe for sourceUrls/wikiUrls This lets you override them using eg. cabal haddock --haddock-options, which can come in handy if you want to use a different layout or URL for your source code links than cabal-install generates. - - - - - 0eff4624 by Niklas Haas at 2014-03-09T07:53:46+01:00 Differentiate between TH splices (line-links) and regular names This adds a new type of source code link, to a specific line rather than a specific declaration/name - this is used to link to the location of a TH splice that defines a certain name. Rather hefty changes throughout and still one unresolved issue (the line URLs aren't parsed from the third form of --read-interface which means they're currently restricted to same-interface links). Not sure if this issue is really worth all the hassle, especially since we could just use line links in general. This commit also contains some cleanup/clarification of the types in Haddock.Backends.Xhtml.Decl and shortens some overlong lines in the process. Notably, the Bool parameter was replaced by a Unicode type synonym to help clarify its presence in type signatures. - - - - - 66d6f77b by Niklas Haas at 2014-03-09T20:02:43+01:00 Group similar fixities together Identical fixities declared for the same line should now render using syntax like: infix 4 <, >=, >, <= - - - - - 6587f9f5 by Mateusz Kowalczyk at 2014-03-10T04:24:18+00:00 Update changelog - - - - - 7387ddad by Niklas Haas at 2014-03-11T10:26:04+01:00 Include fixity information in the Interface file This resolves fixity information not appearing across package borders. The binary file version has been increased accordingly. - - - - - ab46ef44 by Niklas Haas at 2014-03-11T10:26:04+01:00 Update changelog - - - - - 565cab6f by Niklas Haas at 2014-03-11T10:26:04+01:00 Update appearance of fixity annotations This moves them in-line with their corresponding lines, similar to a presentation envision by @hvr and described in #ghc. Redundant operator names are also omitted when no ambiguity is present. - - - - - 5d7afd67 by Niklas Haas at 2014-03-11T10:26:05+01:00 Filter family instances of hidden types Currently, this check does not extend to hidden right hand sides, although it probably should hide them in that case. - - - - - ec291b0c by Niklas Haas at 2014-03-11T10:26:05+01:00 Add documentation for --source-entity-line - - - - - 0922e581 by Niklas Haas at 2014-03-11T10:37:32+01:00 Revert "Reorder topDeclElem to move the source/wiki links to the top" This reverts commit 843c42c4179526a2ad3526e4c7d38cbf4d50001d. This change is no longer needed with the new rendering style, and it messes with copy/pasting lines. - - - - - 30618e8b by Mateusz Kowalczyk at 2014-03-11T09:41:07+00:00 Bump version to 2.15.0 - - - - - adf3f1bb by Mateusz Kowalczyk at 2014-03-11T09:41:09+00:00 Fix up some whitespace - - - - - 8905f57d by Niklas Haas at 2014-03-13T19:18:06+00:00 Hide RHS of TFs with non-exported right hand sides Not sure what to do about data families yet, since technically it would not make a lot of sense to display constructors that cannot be used by the user. - - - - - 5c44d5c2 by Niklas Haas at 2014-03-13T19:18:08+00:00 Add UnicodeSyntax alternatives for * and -> I could not find a cleaner way to do this other than checking for string equality with the given built-in types. But seeing as it's actually equivalent to string rewriting in GHC's implementation of UnicodeSyntax, it's probably fitting. - - - - - b04a63e6 by Niklas Haas at 2014-03-13T19:18:10+00:00 Display minimal complete definitions for type classes This corresponds to the new {-# MINIMAL #-} pragma present in GHC 7.8+. I also cleaned up some of the places in which ExportDecl is used to make adding fields easier in the future. Lots of test cases have been updated since they now render with minimality information. - - - - - a4a20b16 by Niklas Haas at 2014-03-13T19:18:12+00:00 Strip links from recently added html tests These were accidentally left there when the tests were originally added - - - - - d624f315 by Mateusz Kowalczyk at 2014-03-13T19:19:31+00:00 Update changelog - - - - - d27a21ac by Mateusz Kowalczyk at 2014-03-13T21:19:07+00:00 Always read in prologue files as UTF8 (#286). - - - - - 54b2fd78 by Mateusz Kowalczyk at 2014-03-13T21:28:09+00:00 Style only - - - - - fa4fe650 by Simon Hengel at 2014-03-15T09:04:18+01:00 Add Fuuzetsu maintainers field in cabal file - - - - - f83484b7 by Niklas Haas at 2014-03-15T18:20:24+00:00 Hide minimal definition for only-method classes Previously this was not covered by the All xs check since here it is not actually an All, rather a single Var n. This also adds the previously missing html-test/src/Minimal.hs. - - - - - 0099d276 by Niklas Haas at 2014-03-15T18:20:26+00:00 Fix issue haskell/haddock#281 This is a regression from the data family instances change. Data instances are now distinguished from regular lists by usage of the new class "inst", and the style has been updated to only apply to those. I've also updated the appropriate test case to test this a bit better, including GADT instances with GADT-style records. - - - - - 1f9687bd by Mateusz Kowalczyk at 2014-03-21T17:48:37+00:00 Please cabal sdist - - - - - 75542693 by Mateusz Kowalczyk at 2014-03-22T16:36:16+00:00 Drop needless --split-objs which slows us down. Involves tiny cleanup of all the dynflag bindings. Fixes haskell/haddock#292. - - - - - 31214dc3 by Herbert Valerio Riedel at 2014-03-23T18:01:01+01:00 Fix a few typos Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 0b73e638 by Mateusz Kowalczyk at 2014-03-31T05:34:36+01:00 Print kind signatures on GADTs - - - - - 2bab42f3 by Mateusz Kowalczyk at 2014-03-31T16:53:25+01:00 Add default for new PlatformConstraints field - - - - - 42647c5f by Mateusz Kowalczyk at 2014-03-31T18:29:04+01:00 Drop leading whitespace in @-style blocks. Fixes haskell/haddock#201. - - - - - 98208294 by Niklas Haas at 2014-03-31T20:09:58+02:00 Crash when exporting record selectors of data family instances This fixes bug haskell/haddock#294. This also fixes a related but never-before-mentioned bug about the display of GADT record selectors with non-polymorphic type signatures. Note: Associated data type constructors fail to show up if nothing is exported that they could be attached to. Exporting any of the data types in the instance head, or the class + data family itself, causes them to show up, but in the absence of either of these, exporting just the associated data type with the constructor itself will result in it being hidden. The only scenario I can come up that would involve this kind of situation involved OverlappingInstances, and even then it can be mitigated by just exporting the class itself, so I'm not going to solve it since the logic would most likely be very complicated. - - - - - 3832d171 by Mateusz Kowalczyk at 2014-04-01T19:07:33+01:00 Make CHANGES consistent with what's now in 2.14.2 - - - - - c386ae89 by Mateusz Kowalczyk at 2014-04-01T19:18:36+01:00 Actually bundle extra spec tests in sdist - - - - - bd57a6d3 by Mateusz Kowalczyk at 2014-04-03T21:13:48+01:00 Update test cases for GHC bug haskell/haddock#8945, Haddock haskell/haddock#188 The order of signature groups has been corrected upstream. Here we add a test case and update some existing test-cases to reflect this change. We remove grouped signature in test cases that we can (Minimal, BugDeprecated &c) so that the test is as self-contained as possible. - - - - - 708b88b1 by Mateusz Kowalczyk at 2014-04-03T21:16:07+01:00 Enforce strict GHC version in cabal file This stops people with 7.6.3 trying to install 2.15.x which clearly won't work. Unfortunately we shipped 2.14.x without realising this. - - - - - 60334f7c by Mateusz Kowalczyk at 2014-04-03T21:19:24+01:00 Initialise some new PlatformConstants fields - - - - - ea77f668 by Mateusz Kowalczyk at 2014-04-11T16:52:23+01:00 We don't actually want unicode here - - - - - 0b651cae by Mateusz Kowalczyk at 2014-04-11T18:13:30+01:00 Parse identifiers with ^ and ⋆ in them. Fixes haskell/haddock#298. - - - - - e8ad0f5f by Mateusz Kowalczyk at 2014-04-11T18:47:41+01:00 Ignore version string during HTML tests. - - - - - de489089 by Mateusz Kowalczyk at 2014-04-11T18:59:30+01:00 Update CHANGES to follow 2.14.3 - - - - - beb464a9 by Gergő Érdi at 2014-04-13T16:31:10+08:00 remove Origin flag from LHsBindsLR - - - - - cb16f07c by Herbert Valerio Riedel at 2014-04-21T17:16:50+02:00 Replace local `die` by new `System.Exit.die` Starting with GHC 7.10, System.Exit exports the new `die` which is essentially the same as Haddock.Util.die, so this commit changes Haddock.Util.die to be a simple re-export of System.Exit.die. See also https://ghc.haskell.org/trac/ghc/ticket/9016 for more details. Signed-off-by: Herbert Valerio Riedel <hvr at gnu.org> - - - - - 9b9b23c7 by Mateusz Kowalczyk at 2014-05-03T15:40:11+02:00 Disambiguate ‘die’ in test runners. - - - - - 5d28a2b8 by Mateusz Kowalczyk at 2014-05-05T09:19:49+02:00 Prepare modules for parser split. We have to generalise the Doc (now DocH) slightly to remove the dependency on GHC-supplied type. - - - - - d3967ff3 by Mateusz Kowalczyk at 2014-05-05T11:00:41+02:00 Move parser + parser tests out to own package. We move some types out that are necessary as well and then re-export and specialise them in the core Haddock. Reason for moving out spec tests is that if we're working on the parser, we can simply work on that and we can ignore the rest of Haddock. The downside is that it's a little inconvenient if at the end of the day we want to see that everything passes. - - - - - 522a448d by Mateusz Kowalczyk at 2014-05-05T11:14:47+02:00 Move out Show and Eq instances to Types They are much more useful to the users here. - - - - - 11a6f0f2 by Mateusz Kowalczyk at 2014-05-06T13:50:31+02:00 Remove no longer necessary parser error handling. We can now drop some Maybe tests and even lets us strip an error handling monad away in a few places. - - - - - 6992c924 by Mateusz Kowalczyk at 2014-05-14T02:23:55+02:00 Please the GHC build-system. As I can not figure out how to do this properly, if we're in GHC tree, we treat the library as being the same package. If we're not in the tree, we require that the library be installed separately. - - - - - 7a8ad763 by Mateusz Kowalczyk at 2014-05-14T14:50:25+02:00 Update issue tracker URL - - - - - f616c521 by Mateusz Kowalczyk at 2014-05-14T14:53:32+02:00 Update issue tracker URL for haddock-library - - - - - 66580ded by Gergő Érdi at 2014-05-25T14:24:16+08:00 Accomodate change in PatSyn representation - - - - - 0e43b988 by Mateusz Kowalczyk at 2014-05-29T03:15:29+02:00 Revert "Accomodate change in PatSyn representation" This reverts commit 57aa591362d7c8ba21285fccd6a958629a422091. I am reverting this because I pushed it to master when it was meant to stay on a wip-branch. Sorry Gergo and everyone who had trouble due to this. - - - - - e10d7ec8 by Mateusz Kowalczyk at 2014-05-29T03:24:11+02:00 Revert "Revert "Accomodate change in PatSyn representation"" This reverts commit e110e6e70e40eed06c06676fd2e62578da01d295. Apparently as per GHC commit ac2796e6ddbd54c5762c53e2fcf29f20ea162fd5 this was actually intended. Embarrasing for me. - - - - - 5861aca9 by Mateusz Kowalczyk at 2014-06-05T19:49:27+02:00 Clear up highlighting of identifiers with ‘'’s. - - - - - d7cc420f by Simon Peyton Jones at 2014-06-06T12:41:09+01:00 Follow change in patSynSig - - - - - 938b4fd8 by Mateusz Kowalczyk at 2014-06-12T07:24:29+02:00 Slightly update the readme. Style-sheets are no longer a recent thing, dead links, old maintainers, different formats. - - - - - c7799dea by Mateusz Kowalczyk at 2014-06-18T00:05:56+02:00 Update cabal files Update repository urls, use subdir property for haddock-library and use a separate versioning scheme for haddock-library in preparation for release. - - - - - a2750b6a by Simon Hengel at 2014-06-18T11:01:18+08:00 Compatibility with older versions of base and bytestring - - - - - 009b4b03 by Simon Hengel at 2014-06-18T11:14:01+08:00 Enable travis-ci for haddock-library - - - - - 9b5862eb by Simon Hengel at 2014-06-18T11:14:01+08:00 haddock-library: Do not depend on haddock-library in test suite I think you either add src to hs-source-dirs or the library to build-depends. But doing both does not make sense (AFAICT). - - - - - fb1f3279 by Simon Hengel at 2014-06-18T11:49:05+08:00 haddock-library: Use -Wall for specs - - - - - 649340e1 by Mateusz Kowalczyk at 2014-06-18T06:58:54+02:00 Use Travis with multiple GHC versions When using HEAD, we build haddock-library directly from repository as a dependency (and thanks to --enable-tests, the tests get ran anyway). In all other cases, we manually run the tests on haddock-library only and don't test the main project. - - - - - d7eeeec2 by Mateusz Kowalczyk at 2014-06-18T07:49:04+02:00 Comment improvements + few words in cabal file - - - - - 0f8db914 by Simon Hengel at 2014-06-18T13:52:23+08:00 Use doctest to check examples in documentation - - - - - 2888a8dc by Simon Hengel at 2014-06-18T14:16:48+08:00 Remove doctest dependency (so that we can use haddock-library with doctest) - - - - - 626d5e85 by Mateusz Kowalczyk at 2014-06-18T08:41:25+02:00 Travis tweaks - - - - - 41d4f9cc by Mateusz Kowalczyk at 2014-06-18T08:58:43+02:00 Don't actually forget to install specified GHC. - - - - - c6aa512a by John MacFarlane at 2014-06-18T10:43:57-07:00 Removed reliance on LambdaCase (which breaks build with ghc 7.4). - - - - - b9b93b6f by John MacFarlane at 2014-06-18T10:54:56-07:00 Fixed haddock warnings. - - - - - a41b0ab5 by Mateusz Kowalczyk at 2014-06-19T01:20:10+02:00 Update Travis, bump version - - - - - 864bf62a by Mateusz Kowalczyk at 2014-06-25T10:36:54+02:00 Fix anchors. Closes haskell/haddock#308. - - - - - 53df91bb by Mateusz Kowalczyk at 2014-06-25T15:04:49+02:00 Drop DocParagraph from front of headers I can not remember why they were wrapped in paragraphs to begin with and it seems unnecessary now that I test it. Closes haskell/haddock#307. - - - - - 29b5f2fa by Mateusz Kowalczyk at 2014-06-25T15:17:20+02:00 Don't mangle append order for nested lists. The benefit of this is that the ‘top-level’ element of such lists is properly wrapped in <p> tags so any CSS working with these will be applied properly. It also just makes more sense. Pointed out at jgm/pandoc#1346. - - - - - 05cb6e9c by Mateusz Kowalczyk at 2014-06-25T15:19:45+02:00 Bump haddock-library to 1.1.0 for release - - - - - 70feab15 by Iavor Diatchki at 2014-07-01T03:37:07-07:00 Propagate overloading-mode for instance declarations in haddock (#9242) - - - - - d4ca34a7 by Simon Peyton Jones at 2014-07-14T16:23:15+01:00 Adapt to new definition of HsDecls.TyFamEqn This is a knock-on from the refactoring from Trac haskell/haddock#9063. I'll push the corresponding changes to GHC shortly. - - - - - f91e2276 by Edward Z. Yang at 2014-07-21T08:14:19-07:00 Track GHC PackageId to PackageKey renaming. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - b010f9ef by Edward Z. Yang at 2014-07-25T16:28:46-07:00 Track changes for module reexports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: src/Haddock/Interface/Create.hs - - - - - 8b85f9f9 by Mateusz Kowalczyk at 2014-07-28T13:25:43+02:00 Catch mid-line URLs. Fixes haskell/haddock#314. - - - - - 4c613a78 by Edward Z. Yang at 2014-08-05T03:11:00-07:00 Track type signature change of lookupModuleInAllPackages Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - e80b051c by Edward Z. Yang at 2014-08-05T17:34:26+01:00 If GhcProfiled, also build Haddock profiled. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f9cccd29 by Edward Z. Yang at 2014-08-07T14:23:35+01:00 Ignore TAGS files. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 00b3af52 by Mateusz Kowalczyk at 2014-08-08T04:58:19+02:00 Update to attoparsec-0.12.1.1 There seems to be memory and speed improvement. - - - - - 5457dc71 by Mateusz Kowalczyk at 2014-08-08T18:24:02+02:00 Fix forgotten src - - - - - 3520cb04 by Mateusz Kowalczyk at 2014-08-14T20:19:07+01:00 Bump down the version for master to 2.14.4 - - - - - dc98c21b by Mateusz Kowalczyk at 2014-08-14T20:23:27+01:00 Revert "Track type signature change of lookupModuleInAllPackages" This reverts commit d59fec2c9551b5662a3507c0011e32a09a9c118f. - - - - - 3f2038c0 by Mateusz Kowalczyk at 2014-08-14T20:23:31+01:00 Revert "Track changes for module reexports." This reverts commit b99b57c0df072d12b67816b45eca2a03cb1da96d. - - - - - 56d4e49e by Mateusz Kowalczyk at 2014-08-14T20:23:42+01:00 Revert "Track GHC PackageId to PackageKey renaming." This reverts commit 8ac42d3327473939c013551750425cac191ff0fd. - - - - - 726ea3cb by Mateusz Kowalczyk at 2014-08-14T20:23:47+01:00 Revert "Adapt to new definition of HsDecls.TyFamEqn" This reverts commit cb96b4f1ed0462b4a394b9fda6612c3bea9886bd. - - - - - 61a88ff0 by Mateusz Kowalczyk at 2014-08-14T20:23:52+01:00 Revert "Propagate overloading-mode for instance declarations in haddock (#9242)" This reverts commit 8d20ca8d5a9bee73252ff2035ec45f9c03d0820c. - - - - - a32ba674 by Mateusz Kowalczyk at 2014-08-14T20:26:03+01:00 Revert "Disambiguate ‘die’ in test runners." This reverts commit dba02d6df32534aac5d257f2d28596238d248942. - - - - - f335820f by Mateusz Kowalczyk at 2014-08-14T20:26:09+01:00 Revert "Replace local `die` by new `System.Exit.die`" This reverts commit 08aa509ebac58bfb202ea79c7c41291ec280a1c5. - - - - - 107078e4 by Mateusz Kowalczyk at 2014-08-14T20:27:34+01:00 Merge branch 'reverts' This reverts any changes that were made to have Haddock compile with 7.9. When 7.10 release comes, we can simply re-apply all the patches and any patches that occur on ghc-head branch from now on. This allows us to build master with 7.8.3 - - - - - b44b3871 by Mateusz Kowalczyk at 2014-08-15T02:47:40+01:00 Fix haskell/haddock#313 by doing some list munging. I get rid of the Monoid instance because we weren't satisfying the laws. Convenience of having <> didn't outweigh the shock-factor of having it behave badly. - - - - - e1a62cde by Mateusz Kowalczyk at 2014-08-15T02:52:56+01:00 Stop testing haskell/haddock#188. Because the change is in GHC 7.9 and we now work against 7.8.3, this test no longer makes sense. We revert it until 7.10 becomes the standard version. If anything, there should be a test for this in GHC itself. - - - - - 54e8286d by Mateusz Kowalczyk at 2014-08-15T05:31:57+01:00 Add haskell/haddock#313 to CHANGES - - - - - 9df7ad5d by Simon Hengel at 2014-08-20T11:25:32+08:00 Fix warning - - - - - ee2574d6 by Simon Hengel at 2014-08-20T12:07:01+08:00 Fix travis builds - - - - - 384cf2e6 by Simon Hengel at 2014-08-20T12:14:31+08:00 Require GHC 7.8.3 - - - - - d4779863 by Simon Hengel at 2014-08-22T12:14:16+08:00 Move Haddock API to a separate package - - - - - 80f3e0e1 by Simon Hengel at 2014-08-22T14:57:38+08:00 Bump version to 2.15.0 and add version constraints - - - - - 309a94ce by Simon Hengel at 2014-08-22T15:18:06+08:00 Add deprecated compatibility module - - - - - 4d1e4e3f by Luite Stegeman at 2014-08-22T20:46:45+02:00 export things to allow customizing how the Ghc session is run - - - - - 47884591 by Luite Stegeman at 2014-08-22T20:46:51+02:00 ghc 7.8.2 compatibility - - - - - 5ea94e2c by Luite Stegeman at 2014-08-22T22:08:58+02:00 install dependencies for haddock-api on travis - - - - - 9fb845b2 by Mateusz Kowalczyk at 2014-08-23T10:09:34+01:00 Move sources under haddock-api/src - - - - - 85817dc4 by Mateusz Kowalczyk at 2014-08-23T10:10:48+01:00 Remove compat stuff - - - - - 151c6169 by Niklas Haas at 2014-08-24T08:14:10+02:00 Fix extra whitespace on signatures and update all test cases This was long overdue, now running ./accept.lhs on a clean test from master will not generate a bunch of changes. - - - - - d320e0d2 by Niklas Haas at 2014-08-24T08:14:35+02:00 Omit unnecessary foralls and fix haskell/haddock#315 This also fixes haskell/haddock#86. - - - - - bdafe108 by Mateusz Kowalczyk at 2014-08-24T15:06:46+01:00 Update CHANGES - - - - - fafa6d6e by Mateusz Kowalczyk at 2014-08-24T15:14:23+01:00 Delete few unused/irrelevant/badly-place files. - - - - - 3634923d by Duncan Coutts at 2014-08-27T13:49:31+01:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. - - - - - 539a7e70 by Herbert Valerio Riedel at 2014-08-31T11:36:32+02:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 9e3a0e5b by Mateusz Kowalczyk at 2014-08-31T12:54:43+01:00 Bump version in doc - - - - - 4a177525 by Mateusz Kowalczyk at 2014-08-31T13:01:23+01:00 Bump haddock-library version - - - - - f99c1384 by Mateusz Kowalczyk at 2014-08-31T13:05:25+01:00 Remove references to deleted files - - - - - 5e51a247 by Mateusz Kowalczyk at 2014-08-31T14:18:44+01:00 Make the doc parser not complain - - - - - 2cedb49a by Mateusz Kowalczyk at 2014-09-03T03:33:15+01:00 CONTRIBUTING file for issues - - - - - 88027143 by Mateusz Kowalczyk at 2014-09-04T00:46:59+01:00 Mention --print-missing-docs - - - - - 42f6754f by Alan Zimmerman at 2014-09-05T18:13:24-05:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e712719e by Austin Seipp at 2014-09-09T01:03:27-05:00 Fix import of 'empty' due to AMP. Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 71c29755 by Herbert Valerio Riedel at 2014-09-09T17:35:20+02:00 Bump `base` constraint for AMP - - - - - 0bf9f3ed by Mateusz Kowalczyk at 2014-09-12T19:18:32+01:00 Delete stale ANNOUNCE - - - - - cac89ee6 by Krzysztof Gogolewski at 2014-09-14T17:17:09+02:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 4d683426 by Edward Z. Yang at 2014-09-18T13:38:11-07:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 80697fd5 by Herbert Valerio Riedel at 2014-09-19T00:07:52+02:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. - - - - - c015eb70 by Herbert Valerio Riedel at 2014-09-19T00:10:36+02:00 Revert "Followup changes to addition of -fwarn-context-quantification" This reverts commit 4023817d7c0e46db012ba2eea28022626841ca9b temporarily as the respective feature hasn't landed in GHC HEAD yet, but this commit blocks later commits from being referenced in GHC HEAD. - - - - - 38ded784 by Edward Z. Yang at 2014-09-18T15:32:15-07:00 Revert "Revert "Followup changes to addition of -fwarn-context-quantification"" This reverts commit db14fd8ab4fab43694139bc203808b814eafb2dc. It's in HEAD now. - - - - - f55d59c9 by Herbert Valerio Riedel at 2014-09-26T19:18:28+02:00 Revert "Fix import of 'empty' due to AMP." This reverts commit 0cc5bc85e9fca92ab712b68a2ba2c0dd9d3d79f4 since it turns out we don't need to re-export `empty` from Control.Monad after all. - - - - - 467050f1 by David Feuer at 2014-10-09T20:07:36-04:00 Fix improper lazy IO use Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. - - - - - cc47b699 by Edward Z. Yang at 2014-10-09T21:38:19-07:00 Fix use-after-close lazy IO bug Make `getPrologue` force `parseParas dflags str` before returning. Without this, it will attempt to read from the file after it is closed, with unspecified and generally bad results. Signed-off-by: David Feuer <David.Feuer at gmail.com> Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 87babcbe by Austin Seipp at 2014-10-20T20:05:27-05:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - ab259516 by Austin Seipp at 2014-10-20T20:07:01-05:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - b918093c by Mateusz Kowalczyk at 2014-10-29T03:59:39+00:00 Experimental support for collapsable headers Closes haskell/haddock#335 - - - - - 849db129 by Mateusz Kowalczyk at 2014-10-29T10:07:26+01:00 Experimental support for collapsable headers (cherry picked from commit e2ed3b9d8dfab09f1b1861dbc8e74f08e137ebcc) - - - - - a4cc4789 by Herbert Valerio Riedel at 2014-10-31T11:08:26+01:00 Collapse user-defined section by default (re haskell/haddock#335) - - - - - 9da1b33e by Yuras Shumovich at 2014-10-31T16:11:04-05:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - c625aefc by Austin Seipp at 2014-10-31T19:34:10-05:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - c7738e5e by Simon Hengel at 2014-11-02T07:25:30+08:00 Remove -fobject-code from .ghci (this slows down reloads on modifications) - - - - - d4a86e95 by Simon Hengel at 2014-11-03T09:26:11+08:00 Get rid of StandaloneDeriving - - - - - a974e311 by Simon Hengel at 2014-11-03T09:26:11+08:00 Derive more instances - - - - - 8aa0c4d7 by Simon Hengel at 2014-11-03T09:27:08+08:00 Remove unused language extensions - - - - - 3052d46a by Simon Hengel at 2014-11-03T09:30:46+08:00 Minor refactoring - - - - - 4281d3cb by Simon Hengel at 2014-11-03T09:30:46+08:00 parser: Try to parse definition lists right before text paragraphs - - - - - 8ba12bf9 by Simon Hengel at 2014-11-03T09:34:19+08:00 Add support for markdown links (closes haskell/haddock#336) - - - - - a2f8d747 by Simon Hengel at 2014-11-03T09:34:19+08:00 Allow markdown links at the beginning of a paragraph - - - - - 53b11207 by Simon Hengel at 2014-11-03T09:34:20+08:00 Update documentation - - - - - 652267c6 by Simon Hengel at 2014-11-03T09:34:20+08:00 Add support for markdown images - - - - - 9d667502 by Simon Hengel at 2014-11-03T09:34:20+08:00 Allow an optional colon after the closing bracket of definition lists This is to disambiguate them from markdown links and will be require with a future release. - - - - - 8167fc32 by Mateusz Kowalczyk at 2014-11-04T01:16:51+00:00 whitespace only - - - - - 3da62981 by Mateusz Kowalczyk at 2014-11-04T01:17:31+00:00 Fix re-exports of built-in type families Fixes haskell/haddock#310 - - - - - edc76b34 by Mateusz Kowalczyk at 2014-11-04T02:54:28+00:00 Turn some uses of error into recoverable warnings This should at the very least not abort when something weird happens. It does feel like we should have a type that carries these errors until the end however as the user might not see them unless they are printed at the end. - - - - - 0a137400 by Mateusz Kowalczyk at 2014-11-04T04:09:44+00:00 Fix warnings - - - - - d068fc21 by Mateusz Kowalczyk at 2014-11-04T21:04:07+00:00 Fix parsing of identifiers written in infix way - - - - - 1a9f2f3d by Simon Hengel at 2014-11-08T11:32:42+08:00 Minor code simplification - - - - - 6475e9b1 by Simon Hengel at 2014-11-08T17:28:33+08:00 newtype-wrap parser monad - - - - - dc1ea105 by Herbert Valerio Riedel at 2014-11-15T11:55:43+01:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - fbb1aca4 by Simon Hengel at 2014-11-16T08:51:38+08:00 State intention rather than implementation details in Haddock comment - - - - - 97851ab2 by Simon Hengel at 2014-11-16T10:20:19+08:00 (wip) Add support for @since (closes haskell/haddock#26) - - - - - 34bcd18e by Gergő Érdi at 2014-11-20T22:35:38+08:00 Update Haddock to new pattern synonym type signature syntax - - - - - 304b7dc3 by Jan Stolarek at 2014-11-20T17:48:43+01:00 Follow changes from haskell/haddock#9812 - - - - - 920f9b03 by Richard Eisenberg at 2014-11-20T16:52:50-05:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 0bfe4e78 by Alan Zimmerman at 2014-11-21T11:23:09-06:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 356ed45a by Thomas Winant at 2014-11-28T16:11:22-06:00 Support for PartialTypeSignatures - - - - - 5dc8f3b1 by Gergő Érdi at 2014-11-29T15:39:09+08:00 For pattern synonyms, render "pattern" as a keyword - - - - - fe704480 by Mateusz Kowalczyk at 2014-12-09T03:38:32+00:00 List new module in cabal file - - - - - b9ad5a29 by Mateusz Kowalczyk at 2014-12-10T00:58:24+00:00 Allow the parser to spit out meta-info Currently we only use it only for ‘since’ annotations but with these patches it should be fairly simple to add new attributes if we wish to. Closes haskell/haddock#26. It seems to work fine but due to 7.10 rush I don't have the chance to do more exhaustive testing right now. The way the meta is output (emphasis at the end of the whole comment) is fairly arbitrary and subject to bikeshedding. Note that this makes test for Bug310 fail due to interface version bump: it can't find the docs for base with this interface version so it fails. There is not much we can do to help this because it tests for ’built-in’ identifier, not something we can provide ourselves. - - - - - 765af0e3 by Mateusz Kowalczyk at 2014-12-10T01:17:19+00:00 Update doctest parts of comments - - - - - 8670272b by jpmoresmau at 2014-12-10T01:35:31+00:00 header could contain several lines Closes haskell/haddock#348 - - - - - 4f9ae4f3 by Mateusz Kowalczyk at 2014-12-12T06:22:31+00:00 Revert "Merge branch 'reverts'" This reverts commit 5c93cc347773c7634321edd5f808d5b55b46301f, reversing changes made to 5b81a9e53894d2ae591ca0c6c96199632d39eb06. Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - e974ac94 by Duncan Coutts at 2014-12-12T06:26:11+00:00 Changes due to ghc api changes in package representation Also fix a bug with finding the package name and version given a module. This had become wrong due to the package key changes (it was very hacky in the first place). We now look up the package key in the package db to get the package info properly. Conflicts: haddock-api/src/Haddock.hs - - - - - 2f3a2365 by Herbert Valerio Riedel at 2014-12-12T06:26:51+00:00 Import Data.Word w/o import-list This is needed to keep the compilation warning free (and thus pass GHC's ./validate) regardless of whether Word is re-exported from Prelude or not See https://ghc.haskell.org/trac/ghc/ticket/9531 for more details - - - - - 1dbd6390 by Alan Zimmerman at 2014-12-12T06:32:07+00:00 Follow changes to TypeAnnot in GHC HEAD Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - bb6ff1f4 by Mateusz Kowalczyk at 2014-12-12T06:35:07+00:00 Bump ‘base’ constraint Follows the similar commit made on ghc-head branch - - - - - 466fe4ab by Krzysztof Gogolewski at 2014-12-12T06:37:42+00:00 Followup changes to addition of -fwarn-context-quantification (GHC Trac haskell/haddock#4426) - - - - - 97e080c9 by Edward Z. Yang at 2014-12-12T06:39:35+00:00 Properly render package ID (not package key) in index, fixes haskell/haddock#329. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> Conflicts: haddock-api/src/Haddock/ModuleTree.hs - - - - - 20b2af56 by Herbert Valerio Riedel at 2014-12-12T06:42:50+00:00 Disambiguate string-literals GHC fails type-inference with `OverloadedStrings` + `Data.Foldable.elem` otherwise. Conflicts: haddock-library/src/Documentation/Haddock/Parser.hs - - - - - b3ad269d by Austin Seipp at 2014-12-12T06:44:14+00:00 Add an .arcconfig file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - 072df0dd by Austin Seipp at 2014-12-12T06:45:01+00:00 Add .arclint file. Signed-off-by: Austin Seipp <austin at well-typed.com> - - - - - dbb9294a by Herbert Valerio Riedel at 2014-12-12T06:46:17+00:00 Collapse user-defined section by default (re haskell/haddock#335) Conflicts: haddock-api/src/Haddock/Backends/Xhtml/DocMarkup.hs - - - - - f23ab545 by Yuras Shumovich at 2014-12-12T06:46:41+00:00 reflect ForeignType constructore removal Reviewers: austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D358 - - - - - 753a4b67 by Austin Seipp at 2014-12-12T06:46:51+00:00 Remove overlapping pattern match Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8954e8f5 by Herbert Valerio Riedel at 2014-12-12T06:50:53+00:00 Make compatible with `deepseq-1.4.0.0` ...by not relying on the default method implementation of `rnf` - - - - - d2b06d61 by Gergő Érdi at 2014-12-12T07:07:30+00:00 Update Haddock to new pattern synonym type signature syntax Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 1ff02426 by Jan Stolarek at 2014-12-12T07:13:24+00:00 Follow changes from haskell/haddock#9812 Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - 06ad7600 by Richard Eisenberg at 2014-12-12T07:13:43+00:00 Changes to reflect refactoring in GHC as part of haskell/haddock#7484 - - - - - 8fd2aa8b by Alan Zimmerman at 2014-12-12T07:22:25+00:00 Follow API changes in D426 Signed-off-by: Austin Seipp <aseipp at pobox.com> Conflicts: haddock-api/src/Haddock/Backends/LaTeX.hs haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs - - - - - 95c3db98 by Thomas Winant at 2014-12-12T07:35:49+00:00 Support for PartialTypeSignatures Conflicts: haddock-api/src/Haddock/Backends/Xhtml/Decl.hs haddock-api/src/Haddock/Convert.hs haddock-api/src/Haddock/Interface/Create.hs - - - - - 45494428 by Gergő Érdi at 2014-12-12T07:36:18+00:00 For pattern synonyms, render "pattern" as a keyword - - - - - a237e3eb by Mateusz Kowalczyk at 2014-12-12T12:27:13+00:00 Various fixups and bumps for next release - - - - - 22918bcd by Herbert Valerio Riedel at 2014-12-14T10:11:47+01:00 Remove redundant wild-card pattern match (this would otherwise cause a build-failure with `-Werror`) - - - - - 1d6ce947 by Herbert Valerio Riedel at 2014-12-14T10:17:06+01:00 Treat GHC 7.10 the same as GHC 7.9 ...since the current GHC 7.9 is going to become GHC 7.10 real-soon-now anyway - - - - - f434ea89 by Herbert Valerio Riedel at 2014-12-14T18:26:50+01:00 Fixup ghc.mk (follow-up to 1739375eb23342) This makes the GHC build-system aware of the data-files to be copied into the bindist (as haddock.cabal doesn't list those anymore) - - - - - 6fb839eb by Mateusz Kowalczyk at 2014-12-17T09:28:59+00:00 Only keep one Version instead of blindly appending - - - - - 40645489 by Mateusz Kowalczyk at 2014-12-18T07:09:44+00:00 Fix dependency version - - - - - 8b3b927b by Mateusz Kowalczyk at 2014-12-18T07:14:23+00:00 Print missing docs by default Adds --no-print-missing-docs - - - - - 59666694 by Mateusz Kowalczyk at 2014-12-18T07:21:37+00:00 update changelog - - - - - aa6d168e by Mateusz Kowalczyk at 2014-12-18T07:30:58+00:00 Update docs for @since - - - - - 2d7043ee by Luite Stegeman at 2014-12-19T18:29:35-06:00 hide projectVersion from DynFlags since it clashes with Haddock.Version.projectVersion - - - - - aaa70fc0 by Luite Stegeman at 2014-12-22T15:58:43+01:00 Add missing import for standalone haddock-api package - - - - - 9ce01269 by Herbert Valerio Riedel at 2014-12-22T17:48:45+01:00 Reset ghc-head with master's tree (this is an overwriting git merge of master into ghc-head) - - - - - fcd6fec1 by Herbert Valerio Riedel at 2014-12-22T17:51:52+01:00 Bump versions for ghc-7.11 - - - - - 525ec900 by Mateusz Kowalczyk at 2014-12-23T13:36:24+00:00 travis-ci: test with HEAD - - - - - cbf494b5 by Simon Peyton Jones at 2014-12-23T15:22:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 50e01c99 by Mateusz Kowalczyk at 2014-12-29T15:28:47+00:00 Make travis use 7.10.x - - - - - 475e60b0 by Njagi Mwaniki at 2014-12-29T15:30:44+00:00 Turn the README into GitHub Markdown format. Closes haskell/haddock#354 - - - - - 8cacf48e by Luite Stegeman at 2015-01-05T16:25:37+01:00 bump haddock-api ghc dependency to allow release candidate and first release - - - - - 6ed6cf1f by Simon Peyton Jones at 2015-01-06T16:37:47+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 8b484f33 by Simon Peyton Jones at 2015-01-08T15:50:22+00:00 Track naming change in DataCon - - - - - 23c5c0b5 by Alan Zimmerman at 2015-01-16T10:15:11-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - e7a5532c by JP Moresmau at 2015-01-22T17:19:03+00:00 Ignore warnings, install Cabal 1.22 - - - - - 86942c84 by jpmoresmau at 2015-01-22T17:19:04+00:00 solve dataDir ambiguity - - - - - 5ceb743e by jpmoresmau at 2015-01-22T19:17:32+00:00 support GHC 7.10: no Safe-Inferred, Foldable instance - - - - - 6a3b3fb5 by Mateusz Kowalczyk at 2015-01-22T19:32:10+00:00 Update test files Test: a correct behaviour for fields comma-separating values. I'm surprised we had no bug open for this. Maybe it affects how haskell/haddock#301 renders now but I doubt. Operators: Seems GHC is giving us a new order for operators, something must have changed on their side again. cc @haasn , this makes the fixity to the side not match the order on the LHS which is a bit unpleasant. Maybe the fixity can be made to match the GHC order? Bug335: We expand examples by default now. Bug310: Now inferred safe. - - - - - 708f8b2f by jpmoresmau at 2015-01-22T19:36:59+00:00 Links to source location of class instance definitions - - - - - 5cf8a6da by Vincent Berthoux at 2015-01-22T19:59:58+00:00 Filter '\r' from comments due to Windows problems. On Windows this was causing newline to be rendered twice in code blocks. Closes haskell/haddock#359, fixes haskell/haddock#356. - - - - - 1749e6f0 by Mateusz Kowalczyk at 2015-01-22T20:31:27+00:00 Changelog only - - - - - c8145f90 by Mateusz Kowalczyk at 2015-01-22T23:34:05+00:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. - - - - - 14248254 by Mateusz Kowalczyk at 2015-01-22T23:43:18+00:00 Sort out some module import warnings - - - - - d8a38989 by Simon Peyton Jones at 2015-01-23T07:10:16-06:00 Track naming change in DataCon (cherry picked from commit 04cf63d0195837ed52075ed7d2676e71831e8a0b) - - - - - d3ac6ae4 by Alan Zimmerman at 2015-01-23T07:17:19-06:00 Follow API changes in D538 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit d61bbc75890e4eb0ad508b9c2a27b91f691213e6) - - - - - 4c1ffeb0 by Simon Peyton Jones at 2015-02-10T12:10:33+00:00 Track changes in HsSyn for quasi-quotes - - - - - 775d20f7 by Mateusz Kowalczyk at 2015-03-15T08:11:48+01:00 --package-name and --package-version flags Used for --hoogle amongst other things. Now we need to teach cabal to use it. The situation is still a bit sub-par because if the flags aren't passed in, the crash will occur. Closes haskell/haddock#353. (cherry picked from commit 8e06728afb0784128ab2df0be7a5d7a191d30ff4) - - - - - f9245e72 by Phil Ruffwind at 2015-03-16T04:32:01-04:00 Prevent Synopsis from using up too much horizontal space When long type signatures occur in the Synopsis, the element is stretched beyond the width of the window. Scrollbars don't appear, so it's impossible to read anything when this happens. - - - - - cd8fa415 by Mateusz Kowalczyk at 2015-03-17T21:59:39+00:00 Update changelog Closes haskell/haddock#151 due to 71170fc77962f10d7d001e3b8bc8b92bfeda99bc - - - - - b5248b47 by Ben Gamari at 2015-03-25T17:12:17+00:00 Make the error encountered when a package can't be found more user-friendly Closes haskell/haddock#369 - - - - - b756b772 by Mateusz Kowalczyk at 2015-03-26T16:31:40+00:00 Remove now redundant imports - - - - - 5ea5e8dd by Mateusz Kowalczyk at 2015-03-26T16:45:52+00:00 Update test to account for \r filtering - - - - - 6539bfb3 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Test for anchor defaulting I delete the old tests because it turns out that: * test runner would never put them in scope of each other even with imports so just one would suffice * test runner actually needed some hacking to keep links so in the end we would end up with no anchors making them useless - - - - - 1a01d950 by Mateusz Kowalczyk at 2015-03-27T00:20:09+00:00 Clearly default to variables in out of scope case - - - - - 7943abe8 by Mateusz Kowalczyk at 2015-03-27T01:14:11+00:00 Fix Hoogle display of constructors Fixes haskell/haddock#361 - - - - - 6d6e587e by Mateusz Kowalczyk at 2015-03-27T01:45:18+00:00 Fully qualify names in Hoogle instances output Closes haskell/haddock#263 - - - - - 52dac365 by Mateusz Kowalczyk at 2015-03-27T01:55:01+00:00 Update changelog - - - - - ca5af9a8 by Mateusz Kowalczyk at 2015-03-27T02:43:55+00:00 Output method documentation in Hoogle backend One thing of note is that we no longer preserve grouping of methods and print each method on its own line. We could preserve it if no documentation is present for any methods in the group if someone asks for it though. Fixes haskell/haddock#259 - - - - - a33f0c10 by Mateusz Kowalczyk at 2015-03-27T03:04:21+00:00 Don't print instance safety information in Hoogle Fixes haskell/haddock#168 - - - - - df6c935a by Mateusz Kowalczyk at 2015-03-28T00:11:47+00:00 Post-release version bumps and changelog - - - - - dde8f7c0 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Loosen bounds on haddock-* - - - - - de93bf89 by Mateusz Kowalczyk at 2015-03-28T20:39:10+00:00 Expand response files in arguments Closes haskell/haddock#285 - - - - - 1f0b0856 by Zejun Wu at 2015-04-26T16:35:35-07:00 Do not insert anchor for section headings in contents box - - - - - 860439d7 by Simon Peyton Jones at 2015-05-01T09:36:47+01:00 Track change in API of TyCon - - - - - a32f3e5f by Adam Gundry at 2015-05-04T15:32:59+01:00 Track API changes to support empty closed type familes - - - - - 77e98bee by Ben Gamari at 2015-05-06T20:17:08+01:00 Ignore doc/haddock.{ps,pdf} - - - - - 663d0204 by Murray Campbell at 2015-05-11T04:47:37-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> - - - - - 8bb0dcf5 by Murray Campbell at 2015-05-11T06:35:06-05:00 Change ModuleTree Node to carry PackageKey and SourcePackageId to resolve haskell/haddock#385 Signed-off-by: Austin Seipp <aseipp at pobox.com> (cherry picked from commit 2380f07c430c525b205ce2eae6dab23c8388d899) - - - - - bad900ea by Adam Bergmark at 2015-05-11T15:29:39+01:00 haddock-library: require GHC >= 7.4 `Data.Monoid.<>` was added in base-4.5/GHC-7.4 Closes haskell/haddock#394 Signed-off-by: Mateusz Kowalczyk <fuuzetsu at fuuzetsu.co.uk> - - - - - daceff85 by Simon Peyton Jones at 2015-05-13T12:04:21+01:00 Track the new location of setRdrNameSpace - - - - - 1937d1c4 by Alan Zimmerman at 2015-05-25T21:27:15+02:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - ee0fb6c2 by Łukasz Hanuszczak at 2015-05-27T11:51:31+02:00 Create simple method for indentation parsing. - - - - - 7d6fcad5 by Łukasz Hanuszczak at 2015-05-27T21:36:13+02:00 Make nested lists count indentation according to first item. - - - - - d6819398 by Łukasz Hanuszczak at 2015-05-27T22:46:13+02:00 Add simple test case for arbitrary-depth list nesting. - - - - - 2929c54d by Łukasz Hanuszczak at 2015-06-03T02:11:31+02:00 Add arbitrary-indent spec test for parser. - - - - - 9a0a9bb0 by Mateusz Kowalczyk at 2015-06-03T05:25:29+01:00 Update docs with info on new list nesting rule Fixes haskell/haddock#278 through commits from PR haskell/haddock#401 - - - - - 12efc92c by Mateusz Kowalczyk at 2015-06-03T05:29:26+01:00 Update some meta data at the top of the docs - - - - - 765ee49f by Bartosz Nitka at 2015-06-07T08:40:59-07:00 Add some Hacking docs for getting started - - - - - 19aaf851 by Bartosz Nitka at 2015-06-07T08:44:30-07:00 Fix markdown - - - - - 2a90cb70 by Mateusz Kowalczyk at 2015-06-08T15:08:36+01:00 Refine hacking instructions slightly - - - - - 0894da6e by Thomas Winant at 2015-06-08T23:47:28-05:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 10a9bb76 by Emanuel Borsboom at 2015-06-12T02:46:23+01:00 Build executable with '-threaded' (fixes haskell/haddock#399) - - - - - 7696b94f by Mateusz Kowalczyk at 2015-06-12T02:59:19+01:00 Update changelog for -threaded Closes haskell/haddock#400 - - - - - d3c118ec by Bartosz Nitka at 2015-06-12T03:00:58+01:00 Fix haddock: internal error: spliceURL UnhelpfulSpan (#207) Inferred type signatures don't have SrcSpans, so let's use the one from the declaration. I've tested this manually on the test-case from haskell/haddock#207, but I got stuck at trying to run the test-suite. - - - - - b67e843b by Mateusz Kowalczyk at 2015-06-12T03:01:50+01:00 Changelog for haskell/haddock#207 Fixes haskell/haddock#207, closes haskell/haddock#402 - - - - - 841d785e by jpmoresmau at 2015-06-12T16:03:16+01:00 Attach to instance location the name that has the same location file Fixes haskell/haddock#383 - - - - - 98791cae by Mateusz Kowalczyk at 2015-06-12T16:08:27+01:00 Update changelog Closes haskell/haddock#398 - - - - - 7c0b5a87 by Phil Ruffwind at 2015-06-12T13:07:25-04:00 Fix alignment of Source links in instance table in Firefox Due to a Firefox bug [1], a combination of 'whitespace: nowrap' on the parent element with 'float: right' on the inner element can cause the floated element to be displaced downwards for no apparent reason. To work around this, the left side is wrapped in its own <span> and set to 'float: left'. As a precautionary measure to prevent the parent element from collapsing entirely, we also add the classic "clearfix" hack. The latter is not strictly needed but it helps prevent bugs if the layout is altered again in the future. Fixes haskell/haddock#384. Remark: line 159 of src/Haddock/Backends/Xhtml/Layout.hs was indented to prevent confusion over the operator precedence of (<+>) vs (<<). [1]: https://bugzilla.mozilla.org/show_bug.cgi?id=488725 - - - - - cfe86e73 by Mateusz Kowalczyk at 2015-06-14T10:49:01+01:00 Update tests for the CSS changes - - - - - 2d4983c1 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding for Haskell source parser module. - - - - - 29548785 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement function for tagging parsed chunks with source spans. - - - - - 6a5e4074 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement simple string chunking based on HsColour library. - - - - - 6e52291f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create basic token classification method. - - - - - da971a27 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Adapt source span tagging to work with current whitespace handling. - - - - - 4feb5a22 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add record accessors to exports of hyperlinker parser module. - - - - - a8cc4e39 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Make parser module export all types and associated accessors. - - - - - fb8d468f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple HTML renderer for parsed source file. - - - - - 80747822 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for specifying the CSS file path in HTML source renderer. - - - - - 994dc1f5 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix identifier recognition in Haskell source parser. - - - - - b1bd0430 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix comment recognition in Haskell source parser. - - - - - 11db85ae by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for recognizing compiler pragmas in source parser. - - - - - 736c7bd3 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create scaffolding of module for associating tokens with AST names. - - - - - 7e149bc2 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement utility method for extracting variable identifiers from AST. - - - - - 32eb640a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Create simple mechanism for associating tokens with AST names. - - - - - d4eba5bc by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add dummy support for hyperlinking named tokens. - - - - - 2b76141f by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matcher bug causing wrong items being hyperlinked. - - - - - 2d48002e by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Constrain elements exported by hyperlinker modules. - - - - - 9715eec6 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for type token recognition. - - - - - 8fa401cb by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Add support for binding token recognition. - - - - - d062400b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement go-to-definition mechanism for local bindings. - - - - - f4dc229b by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Implement module export- and import-list item hyperlinking. - - - - - c9a46d58 by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix span matching to allow parenthesized operators hyperlinking. - - - - - 03aad95a by Łukasz Hanuszczak at 2015-06-30T22:37:48+02:00 Fix weird hyperlinking of parenthesized operators. - - - - - b4694a7d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for type declaration anchors. - - - - - 7358d2d2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for top-level function declaration anchors. - - - - - dfc24b24 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix external anchors to contain HTML file extension. - - - - - a045926c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Refactor the way AST names are handled within detailed tokens. - - - - - c76049b4 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement hyperlinking of imported module names. - - - - - 2d2a1572 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix parsing of single line comments with broken up newlines. - - - - - 11afdcf2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix bug with improper newline handling. - - - - - 8137f104 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issues with escaped newlines in comments. - - - - - 34759b19 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for parsing C preprocessor macros. - - - - - 09f0f847 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for parser module of source hyperlinker. - - - - - 709a8389 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add some documentation for AST module of source hyperlinker. - - - - - 4df5c227 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add command line option for generating hyperlinked source. - - - - - 7a755ea2 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Extend module interface with rich source token stream field. - - - - - 494f4ab1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement source tokenization during interface creation process. - - - - - 5f21c953 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create hyperlinker module and plug it into the Haddock pipeline. - - - - - 0cc8a216 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for providing custom CSS files for hyperlinked source. - - - - - a32bbdc1 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add support for fancy highlighting upon hovering over identifier. - - - - - d16d642a by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make source hyperlinker generate output in apropriate directory. - - - - - ae12953d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Create module with hyperlinker utility functions. - - - - - 6d4952c5 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make external hyperlinks point to locations specified by source URLs. - - - - - 8417555d by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Rewrite source generation to fixed links and directory structure. - - - - - ce9cec01 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Add basic support for cross-package hyperlink generation. - - - - - 7eaf025c by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Disable generating hyperlinks for module references. - - - - - a50bf92e by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make Haddock generate source for all interfaces (also hidden ones). - - - - - f5ae2838 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Prevent source parser from throwing exception when lexing fails. - - - - - db9ffbe0 by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Implement workaround for Chrome highlighting issues. - - - - - 0b6b453b by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate correct anchors for data constructors. - - - - - c86d38bc by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Make hyperlinker generate anchors for record field declarations. - - - - - 063abf7f by Łukasz Hanuszczak at 2015-06-30T22:37:49+02:00 Fix issue with hyperlink highlight styling in Chrome browser. - - - - - 880fc611 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking constructor names in patters. - - - - - c9e89b95 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record patterns. - - - - - 17a11996 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add support for hyperlinking field names in record expressions. - - - - - 0eef932d by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Make hyperlinker respect pretty-printer flag and add documentation. - - - - - f87c1776 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Unexpose hyperlinker modules in Cabal configuration. - - - - - 4c9e2b06 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Setup HSpec framework for Haddock API package. - - - - - 4b20cb30 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add basic tests related to comment parsing. - - - - - 6842e919 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add tests related to parsing basic language constructs. - - - - - 87bffb35 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add simple tests for do-notation parsing. - - - - - e7af1841 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add very simple QuickCheck properties for source parser spec. - - - - - c84efcf1 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Create simple test runner for hyperlinker tests. - - - - - 76b90447 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for basic identifier hyperlinking. - - - - - 0fbf4df6 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for operator hyperlinking. - - - - - 731aa039 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for constructor hyperlinking. - - - - - 995a78a2 by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for record expressions and patterns hyperlinking. - - - - - 3566875a by Łukasz Hanuszczak at 2015-06-30T22:37:50+02:00 Add test case for literal syntax highlighting. - - - - - 68469a35 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add hyperlinker test runner to .cabal and .gitignore files. - - - - - aa946c93 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Adapt hyperlinker test runner to have the same interface as HTML one. - - - - - ce34da16 by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Fix hyperlinker test runner file paths and add pretty-printing option. - - - - - 0d7dd65e by Łukasz Hanuszczak at 2015-06-30T22:38:33+02:00 Add reference files for hyperlinker test cases. - - - - - efa4a1e0 by Łukasz Hanuszczak at 2015-07-01T00:47:32+02:00 Make hyperlinker test runner strip local links from generated source. - - - - - 3e96e584 by Łukasz Hanuszczak at 2015-07-01T01:14:59+02:00 Create simple script for accepting hyperlinker test case references. - - - - - 526fe610 by Łukasz Hanuszczak at 2015-07-01T01:16:41+02:00 Re-accept hyperlinker test cases with local references stripped out. - - - - - 892e2cb3 by Łukasz Hanuszczak at 2015-07-01T01:22:09+02:00 Fix bug with diffing wrong files in hyperlinker test runner. - - - - - 9ff46039 by Łukasz Hanuszczak at 2015-07-01T18:04:46+02:00 Remove unused dependencies in Haddock API spec configuration. - - - - - 47969c07 by Łukasz Hanuszczak at 2015-07-01T18:32:19+02:00 Add support for hyperlinking synonyms in patterns. - - - - - a73449e0 by Łukasz Hanuszczak at 2015-07-01T18:33:44+02:00 Create test case for hyperlinking @-patterns. - - - - - c2077ed8 by Łukasz Hanuszczak at 2015-07-01T19:06:04+02:00 Add support for hyperlinking universally quantified type variables. - - - - - 68017342 by Łukasz Hanuszczak at 2015-07-01T19:28:32+02:00 Create hyperlinker test case with quantified type variables. - - - - - 51c01a78 by Łukasz Hanuszczak at 2015-07-01T19:34:22+02:00 Add scoped type variables test for polymorphism test case. - - - - - 13181ae2 by Łukasz Hanuszczak at 2015-07-01T19:56:27+02:00 Add record wildcards test for records hyperlinking test case. - - - - - 991b81dd by Łukasz Hanuszczak at 2015-07-01T21:01:42+02:00 Document some functions in XHTML utlity module. - - - - - 98c8dfe5 by Łukasz Hanuszczak at 2015-07-01T22:25:21+02:00 Make hyperlinker render qualified names as one entity. - - - - - 75e13b9b by Łukasz Hanuszczak at 2015-07-01T22:27:38+02:00 Add qualified name test for identifiers hyperlinking test case. - - - - - de1e143f by Łukasz Hanuszczak at 2015-07-02T12:32:59+02:00 Fix crash happening when hyperlinking type family declarations. - - - - - 7a8fb175 by Łukasz Hanuszczak at 2015-07-02T12:47:03+02:00 Add support for anchoring data family constructor declarations. - - - - - 3b404e49 by Łukasz Hanuszczak at 2015-07-02T13:31:05+02:00 Improve support for hyperlinking type families. - - - - - 59eb7143 by Łukasz Hanuszczak at 2015-07-02T13:33:34+02:00 Add hyperlinker test case for checking type and type family declarations. - - - - - d1cda0c0 by Łukasz Hanuszczak at 2015-07-02T13:41:38+02:00 Fix issue with operators being recognized as preprocessor directives. - - - - - da206c9d by Łukasz Hanuszczak at 2015-07-02T17:18:12+02:00 Fix broken tests for parsing and hyperlinking hash operators. - - - - - 53750d1b by Łukasz Hanuszczak at 2015-07-02T18:53:28+02:00 Add support for anchoring signatures in type class declarations. - - - - - 1fa5bb10 by Łukasz Hanuszczak at 2015-07-02T19:04:47+02:00 Make hyperlinker generate anchors only to top-level value bindings. - - - - - a542305c by Łukasz Hanuszczak at 2015-07-02T19:05:58+02:00 Create hyperlinker test case for type classes. - - - - - b0dd4581 by Łukasz Hanuszczak at 2015-07-04T16:28:26+02:00 Update docs with information about source hyperlinking. - - - - - 9795302a by Łukasz Hanuszczak at 2015-07-04T16:52:15+02:00 Update docs on using `--read-interface` option. - - - - - 9acdc002 by Łukasz Hanuszczak at 2015-07-04T17:15:26+02:00 Remove potentially dangerous record access in hyperlinker AST module. - - - - - fb3ab7be by Łukasz Hanuszczak at 2015-07-04T17:40:10+02:00 Make Haddock generate warnings about potential misuse of hyperlinker. - - - - - a324c504 by Łukasz Hanuszczak at 2015-07-04T17:43:22+02:00 Fix incorrect specification of source style option in doc file. - - - - - 3f01a8e4 by Łukasz Hanuszczak at 2015-07-05T17:06:36+02:00 Refactor source path mapping to use modules as indices. - - - - - ac70f5b1 by Łukasz Hanuszczak at 2015-07-05T17:47:34+02:00 Fix bug where not all module interfaces were added to source mapping. - - - - - f5e57da9 by Łukasz Hanuszczak at 2015-07-06T16:39:57+02:00 Extract main hyperlinker types to separate module. - - - - - 43974905 by Łukasz Hanuszczak at 2015-07-06T16:52:13+02:00 Move source paths types to hyperlinker types module. - - - - - 3e236055 by Łukasz Hanuszczak at 2015-07-06T17:06:19+02:00 Add support for hyperlinking modules in import lists. - - - - - 58233d9f by Łukasz Hanuszczak at 2015-07-06T17:26:49+02:00 Add short documentation for hyperlinker source map type. - - - - - 14da016d by Łukasz Hanuszczak at 2015-07-06T18:07:20+02:00 Fix bug with module name being hyperlinked to `Prelude`. - - - - - 8f79db52 by Łukasz Hanuszczak at 2015-07-06T18:23:47+02:00 Fix problem with spec build in Haddock API configuration. - - - - - e7cc056c by Adam Sandberg Eriksson at 2015-07-07T23:22:21+01:00 StrictData: print correct strictness marks - - - - - e8253ca8 by Mateusz Kowalczyk at 2015-07-07T23:58:28+01:00 Update changelog - - - - - 0aba676b by Mateusz Kowalczyk at 2015-07-07T23:58:33+01:00 Relax upper bound on GHC a bit - - - - - 7a595381 by Mateusz Kowalczyk at 2015-07-07T23:58:52+01:00 Delete trailing whitespace - - - - - 50976d5e by Adam Sandberg Eriksson at 2015-07-08T15:03:04+02:00 StrictData: changes in HsBang type - - - - - 83b045fa by Mateusz Kowalczyk at 2015-07-11T14:35:18+01:00 Fix expansion icon for user-collapsible sections Closes haskell/haddock#412 - - - - - b2a3b0d1 by Mateusz Kowalczyk at 2015-07-22T22:03:21+01:00 Make some version changes after 2.16.1 release - - - - - a8294423 by Ben Gamari at 2015-07-27T13:16:07+02:00 Merge pull request haskell/haddock#422 from adamse/adamse-D1033 Merge for GHC D1033 - - - - - c0173f17 by randen at 2015-07-30T14:49:08-07:00 Break the response file by line termination rather than spaces, since spaces may be within the parameters. This simple approach avoids having the need for any quoting and/or escaping (although a newline char will not be possible in a parameter and has no escape mechanism to allow it). - - - - - 47c0ca14 by Alan Zimmerman at 2015-07-31T10:41:52+02:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - 45a9d770 by Mateusz Kowalczyk at 2015-07-31T09:47:43+01:00 Update changelog - - - - - 347a20a3 by Phil Ruffwind at 2015-08-02T23:15:26+01:00 Avoid JavaScript error during page load in non-frame mode In non-frame mode, parent.window.synopsis refers to the synopsis div rather than the nonexistent frame. Unfortunately, the script wrongly assumes that if it exists it must be a frame, leading to an error where it tries to access the nonexistent attribute 'replace' of an undefined value (synopsis.location). Closes haskell/haddock#406 - - - - - 54ebd519 by Phil Ruffwind at 2015-08-02T23:27:10+01:00 Link to the definitions to themselves Currently, the definitions already have an anchor tag that allows URLs with fragment identifiers to locate them, but it is rather inconvenient to obtain such a URL (so-called "permalink") as it would require finding the a link to the corresponding item in the Synopsis or elsewhere. This commit adds hyperlinks to the definitions themselves, allowing users to obtain links to them easily. To preserve the original aesthetics of the definitions, we alter the color of the link so as to be identical to what it was, except it now has a hover effect indicating that it is clickable. Additionally, the anchor now uses the 'id' attribute instead of the (obsolete) 'name' attribute. Closes haskell/haddock#407 - - - - - 02cc8bb7 by Phil Ruffwind at 2015-08-02T23:28:02+01:00 Fix typo in Haddock.Backends.Xhtml.Layout: divSynposis -> divSynopsis Closes haskell/haddock#408 - - - - - 2eb0a458 by Phil Ruffwind at 2015-08-02T23:30:07+01:00 Fix record field alignment when name is too long Change <dl> to <ul> and use display:table rather than floats to layout the record fields. This avoids bug haskell/haddock#301 that occurs whenever the field name gets too long. Slight aesthetic change: the entire cell of the field's source code is now shaded gray rather than just the area where text exists. Fixes haskell/haddock#301. Closes haskell/haddock#421 - - - - - 7abb3402 by Łukasz Hanuszczak at 2015-08-02T23:32:14+01:00 Add some utility definitions for generating line anchors. - - - - - e0b1d79b by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Make hyperlinked source renderer generate line anchors. - - - - - 24dd4c9f by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Re-accept test cases after adding line anchors for each of them. - - - - - 0372cfcb by Łukasz Hanuszczak at 2015-08-02T23:32:15+01:00 Override source line flags when source hyperlinker is enabled. - - - - - a81bcd07 by Mateusz Kowalczyk at 2015-08-02T23:58:25+01:00 Update tests to follow HTML changes - - - - - d2d7426f by Łukasz Hanuszczak at 2015-08-06T20:54:59+02:00 Fix quote syntax for promoted types. - - - - - 668cf029 by Łukasz Hanuszczak at 2015-08-06T21:12:00+02:00 Apply promoted type quoting to type-level consing. - - - - - 89f8e7c6 by Łukasz Hanuszczak at 2015-08-06T21:17:10+02:00 Extend advanced types test case with other examples. - - - - - 86494bca by Łukasz Hanuszczak at 2015-08-06T21:22:06+02:00 Rename advanced types test case and accept new output. - - - - - dbb7c7c0 by Adam Sandberg Eriksson at 2015-08-09T23:01:05+02:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 2a7704fa by Ben Gamari at 2015-08-10T13:18:05+02:00 Merge pull request haskell/haddock#433 from adamse/split-hsbang HsBang is split into HsSrcBang and HsImplBang - - - - - 891954bc by Thomas Miedema at 2015-08-15T14:51:18+02:00 Follow changes in GHC build system - - - - - b55d32ab by Mateusz Kowalczyk at 2015-08-21T18:06:09+01:00 Make Travis use 7.10.2 - - - - - 97348b51 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Move SYB utilities to standalone module. - - - - - 748ec081 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement `everywhere` transformation in SYB module. - - - - - 011cc543 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Implement generic transformation constructor. - - - - - b9510db2 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Create simple utility module for type specialization. - - - - - 43229fa6 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Make type of type specialization function more general. - - - - - fd844e90 by Łukasz Hanuszczak at 2015-08-21T18:22:29+01:00 Add basic HTML test case for checking instance specialization. - - - - - 6ea0ad04 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Make HTML class instance printer take optional signature argument. - - - - - 65aa41b6 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Refactor instance head type to record instead of a meaningless tuple. - - - - - 3fc3bede by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add expandable method section for each class instance declaration. - - - - - 99ceb107 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Move dummy post-family instances for `DocName` to `Types` module. - - - - - e98f4708 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create convenience functions for type specialization module. - - - - - b947552f by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Hook type specialization logic with HTML pretty-printer. - - - - - dcaa8030 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create stub functions for sugaring specialized types. - - - - - fa84bc65 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement list syntax sugaring logic for specialized types. - - - - - e8b05b07 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement tuple syntax sugaring logic for specialized types. - - - - - 68a2e5bc by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Get rid of code duplication in type specialization module. - - - - - 4721c336 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create scaffolding of a framework for renaming specialized types. - - - - - 271b488d by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in missing cases in specialized type renaming function. - - - - - bfa5f2a4 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Remove code duplication in specialized type renamer. - - - - - ea6bd0e8 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Change state of the type renaming monad. - - - - - 77c5496e by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Implement simple mechanism for generating new type names. - - - - - 91bfb48b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fill in stub behaviour with actual environment renaming. - - - - - d244517b by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix logic behind binder type renaming. - - - - - f3c5e360 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Add SYB-like utility function for performing stateful queries. - - - - - eb3f9154 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Create function for retrieving free variables from given type. - - - - - a94561d3 by Łukasz Hanuszczak at 2015-08-21T18:22:30+01:00 Fix compilation error caused by incorrect type signature. - - - - - 8bb707cf by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Move `SetName` class definition to types module. - - - - - 5800b13b by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Hook type renamer with instance method HTML pretty-printer. - - - - - 6a480164 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some test cases for type renamer. - - - - - 839842f7 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized signatures refer to original signature declaration. - - - - - 4880f7c9 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make specialized methods be nicely formatted again. - - - - - ab5a6a2e by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Attach source locations to the specialized class methods. - - - - - 43f8a559 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Extend instances test case to also test multi-name type signatures. - - - - - 59bc751c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix tab-based indentation in instances test case. - - - - - c2126815 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Improve placement of instance methods expander button. - - - - - 0a32e287 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add new data type declaration to instance specialization test case. - - - - - 5281af1f by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Make type renamer first try single-letter names as alternatives. - - - - - 7d509475 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix type renamer bug with incorrect names being generated. - - - - - 0f35bf7c by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Add some documentation and refactor type specialization module. - - - - - da1d0803 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix another bug where type renamer was generating incorrect names. - - - - - cd39b5cb by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Refactor type renamer to rebinding and pure renaming phases. - - - - - 850251f4 by Łukasz Hanuszczak at 2015-08-21T18:22:31+01:00 Fix unwitting compilation bug. - - - - - e5e9fc01 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Integrate instance specification type into class instance definition. - - - - - 825b0ea0 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer neccessary instance specification type. - - - - - cdba44eb by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix declaration converter to use more appropriate mode for methods. - - - - - bc45c309 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug with types not being specialized at all. - - - - - 5d8e5d89 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix bug where instance expander was opening wrong section. - - - - - 6001ee41 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix another type renamer bug where not all names were rebound. - - - - - 5f58ce2a by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Fix yet another renamer bug where some names were not unique. - - - - - 8265e521 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split instance subsection layout method to top-level declarations. - - - - - e5e66298 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Rearrange layout of instance methods in generated documentation. - - - - - a50b4eea by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Get rid of no longer used layout method. - - - - - 2ff36ec2 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach section title to the instance methods block. - - - - - 7ac15300 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Add basic tests for associated types in instances test case. - - - - - db0ea2f9 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Attach associated types information to instance header. - - - - - 71cad4d5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details section contain associated types information. - - - - - deee2809 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Improve look of rendered associated families in instance details. - - - - - 839d13a5 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Introduce alternative type for family declarations. - - - - - d397f03f by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Make instance details record use new type for family declarations. - - - - - 2b23fe97 by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Split printer of type family header to separate functions. - - - - - c3498cdc by Łukasz Hanuszczak at 2015-08-21T18:22:32+01:00 Implement HTML renderer for pseudo-family declarations. - - - - - c12bbb04 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Apply type specializer to associated type family declarations. - - - - - 2fd69ff2 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Create helper method for specializing type signatures. - - - - - 475826e7 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Refactor specializer module to be independent from XHTML backend. - - - - - f00b431c by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add some documentation for instance head specializer. - - - - - a9fef2dc by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix bug with missing space in documentation for associated types. - - - - - 50e29056 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with incorrect instance details sections being expanded. - - - - - e6dfdd03 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by adding instance details section. - - - - - 75565b2a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make section identifier of instance details more GHC-independent. - - - - - add0c23e by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Re-accept tests after applying deterministic section identifiers. - - - - - 878f2534 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Make identifier generation also architecture-independent. - - - - - 48be69f8 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Fix issue with instance expander hijacking type hyperlink click. - - - - - 47830c1f by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Get rid of dreadful hashing function for generating identifiers. - - - - - 956cd5af by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Move `InstOrigin` type declaration to more appropriate module. - - - - - bf672ed3 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Accept tests affected by changes related to instance expander. - - - - - 8f2a949a by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add examples with type operators to the instances test case. - - - - - 64600a84 by Łukasz Hanuszczak at 2015-08-21T18:22:33+01:00 Add basic support for sugaring infix type operators. - - - - - 747d71b8 by Łukasz Hanuszczak at 2015-08-21T18:22:34+01:00 Add support for sugaring built-in function syntax. - - - - - d4696ffb by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Remove default methods from Hoogle class output. - - - - - bf0e09d7 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add fixity declarations in Hoogle backend output. - - - - - 90e91a51 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix bug with incorrect fixities being generated in Hoogle backend. - - - - - 48f11d35 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve class type family declarations output in Hoogle backend. - - - - - 661e8e8f by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Add missing default family equations in Hoogle output. - - - - - e2d64103 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Improve formatting of class details output in Hoogle backend. - - - - - 490fc377 by Łukasz Hanuszczak at 2015-08-21T18:31:31+01:00 Fix weird-looking Hoogle output for familyless classes. - - - - - ea115b64 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create script file for new HTML test runner. - - - - - 609913d3 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Set default behaviour if no arguments given. - - - - - dc115f67 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for providing optional arguments for test runner. - - - - - d93ec867 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve output of test runner error messages. - - - - - 0be9fe12 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add support for executing Haddock process in test runner. - - - - - 4e4d00d9 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Add GHC path to test runner configuration. - - - - - d67a2086 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make GHC path a test runner command-line argument. - - - - - c810079a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Extend test runner configuration with Haddock arguments. - - - - - fee18845 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor test runner and create stub functions. - - - - - ff7c161f by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make test runner actually run Haddock executable. - - - - - 391f73e6 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with test runner not producing any output files. - - - - - 81a74e2d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Setup skeleton of framework for running tests. - - - - - f8a79ec4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Fix bug with modules not being found in global search mode. - - - - - 7e700b4d by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make Haddock standard output redirection be more configurable. - - - - - 53b4c17a by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Incorporate old, ugly functions for comparing output files. - - - - - 8277c8aa by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Refactor architecture of test runner output checking functions. - - - - - 587bb414 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement actual diffing mechanism. - - - - - 9ed2b5e4 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Improve code style to match popular guidelines. - - - - - 14bffaf8 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Make it possible to choose alternative diff tool. - - - - - 5cdfb005 by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Create stub methods for processing test output as XML documents. - - - - - 7ef8e12e by Łukasz Hanuszczak at 2015-08-22T23:40:26+02:00 Implement link-stripping logic as simple SYB transformation. - - - - - 8a1fcd4f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Incorporate link stripping to output diffing mechanism. - - - - - 37dba2bc by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement footer-stripping logic. - - - - - 9cd52120 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Add missing dependencies in Cabal configuration file. - - - - - e0f83c6e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix issue with output being printed in incorrect order. - - - - - 0a94fbb0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to run tests without generating diff. - - - - - 76a58c6f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor HTML test suite boilerplate to external package. - - - - - af41e6b0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utilities for storing directory configuration. - - - - - d8f0698f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move IO-dependent config of HTML test suite to test package. - - - - - 17369fa0 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Enable all compiler warnings in Haddock test package configuration. - - - - - 9d03b47a by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Move Haddock runner of HTML test suite to Haddock test package. - - - - - 4b3483c5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 03754194 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience wrappers to simplify in test entry points. - - - - - 27476ab7 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adjust module visibility and items they export. - - - - - c40002ba by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Remove no longer useful test option. - - - - - 55ab2541 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Change extension of test files used for diffing. - - - - - 136bf4e4 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Refactor and simplify XHTML helper module of test package. - - - - - 69f7e3df by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix typo in link stripper of HTML test suite runner. - - - - - 0c3c1c6b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create convenience script for running specific HTML tests. - - - - - 489e1b05 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement utility functions for conditional link stripping. - - - - - 0f985dc3 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt `hypsrc-test` module to work with new testing framework. - - - - - 927406f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Implement output accepting mechanism in test package. - - - - - 8545715e by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Create utility function for recursive obtaining directory contents. - - - - - cb70381f by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make Haddock test package more generic. - - - - - 019599b5 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix path handling in test runner. - - - - - 399b985b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Make it possible to specify ignored files for test output. - - - - - 41b3d93d by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Adapt HTML test runner to use new ignoring functionality. - - - - - e2091c8b by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Fix bug with not all test output files being checked. - - - - - b22134f9 by Łukasz Hanuszczak at 2015-08-22T23:40:27+02:00 Specify ignored files for hyperlinker source test runner. - - - - - 3301dfa1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Copy test runner script for hyperlinked source case. - - - - - d39a6dfa by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner invoking Haddock in incorrect mode. - - - - - f32c8ff3 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix path handling in test module loader. - - - - - 10f94ee9 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make test runner ignore test packages with no modules. - - - - - 5dc4239c by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create test runner entry points for LaTeX test suite. - - - - - 58d1f7cf by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with unnecessary checking old test output. - - - - - c7ce76e1 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Re-implement test acceptance functionality. - - - - - 13bbabe8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix warning about no longer needed definition. - - - - - 958a99b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Adapt Cabal configuration to execute LaTeX suite with new runner. - - - - - 550ff663 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Setup test suite for Hoogle backend. - - - - - 3aa969c4 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Make Hoogle backend create output directory if needed. - - - - - eb085b02 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Add appropriate .gitignore entry and configure Hoogle test suite. - - - - - a50bf915 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Fix bug with test runner failing when run on multiple test packages. - - - - - bf5368b8 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create simple test cases for Hoogle backend. - - - - - 6121ba4b by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Create helper function for conversion between XML and XHTML. - - - - - cb516061 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Refactor existing code to use XHTML printer instead of XML one. - - - - - e2de8c82 by Łukasz Hanuszczak at 2015-08-22T23:40:28+02:00 Improve portability of test runner scripts. - - - - - 9563e774 by Łukasz Hanuszczak at 2015-08-22T23:43:16+02:00 Remove redundant import statement. - - - - - 55353df1 by Łukasz Hanuszczak at 2015-08-24T23:09:20+02:00 Fix bug with accepting to non-existing directory. - - - - - 00a334ca by Łukasz Hanuszczak at 2015-08-24T23:09:47+02:00 Accept output for Hoogle and LaTeX backends. - - - - - 29191d8b by Łukasz Hanuszczak at 2015-08-24T23:14:18+02:00 Get rid of obsolete testing utilities. - - - - - bbb25db3 by Łukasz Hanuszczak at 2015-08-24T23:18:50+02:00 Update sandbox setup guide to work with Haddock test package. - - - - - cfd45248 by Łukasz Hanuszczak at 2015-08-24T23:51:30+02:00 Make Travis aware of Haddock test package. - - - - - 74185b7a by Łukasz Hanuszczak at 2015-08-25T17:41:59+02:00 Fix test suite failure when used with Stack. - - - - - 18769697 by Łukasz Hanuszczak at 2015-08-25T18:02:09+02:00 Add sample Stack setup to the hacking guide. - - - - - 22715eeb by Łukasz Hanuszczak at 2015-08-25T18:04:47+02:00 Fix Markdown formatting of README file. - - - - - b49ec386 by Łukasz Hanuszczak at 2015-08-25T18:13:36+02:00 Setup Haddock executable path in Travis configuration. - - - - - 5d29eb03 by Eric Seidel at 2015-08-30T09:55:58-07:00 account for changes to ipClass - - - - - f111740a by Ben Gamari at 2015-09-02T13:20:37+02:00 Merge pull request haskell/haddock#443 from bgamari/ghc-head account for changes to ipClass - - - - - a2654bf6 by Jan Stolarek at 2015-09-03T01:32:57+02:00 Follow changes from haskell/haddock#6018 - - - - - 2678bafe by Richard Eisenberg at 2015-09-21T12:00:47-04:00 React to refactoring CoAxiom branch lists. - - - - - ebc56e24 by Edward Z. Yang at 2015-09-21T11:53:46-07:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4a8c4198 by Tamar Christina at 2015-09-27T13:59:08+02:00 Create Process: removed PhaseFailed - - - - - 7e99b790 by Oleg Grenrus at 2015-09-27T20:52:10+03:00 Generate docs for orphan instances - - - - - 32e932e2 by Oleg Grenrus at 2015-09-28T07:21:11+03:00 Have source links for orphan instances - - - - - c2eb9f4f by Oleg Grenrus at 2015-09-28T07:24:58+03:00 Print orphan instances header only if required - - - - - ff96f978 by Oleg Grenrus at 2015-09-28T07:40:54+03:00 Add orphan instances link to contents box - - - - - d72490a6 by Oleg Grenrus at 2015-09-28T16:37:44+03:00 Fix orphan instance collapsing - - - - - 25d3dfe5 by Ben Gamari at 2015-10-03T12:38:09+02:00 Merge pull request haskell/haddock#448 from Mistuke/fix-silent-death-of-runInteractive Remove PhaseFailed - - - - - 1e45e43b by Edward Z. Yang at 2015-10-11T13:10:10-07:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b1370ac1 by Adam Gundry at 2015-10-16T16:26:42+01:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 60bef421 by Simon Peyton Jones at 2015-10-26T12:52:36+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 4c1898ca by Simon Peyton Jones at 2015-10-27T14:24:56+00:00 Track change to PatSyn.patSynSig - - - - - 25108e85 by Simon Peyton Jones at 2015-10-27T17:34:18+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - 693643ac by Ben Gamari at 2015-10-28T14:33:06+01:00 Account for Typeable changes The treatment of type families changed. - - - - - cd7c2221 by Simon Peyton Jones at 2015-10-30T13:03:51+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - 712032cb by Herbert Valerio Riedel at 2015-10-31T11:01:45+01:00 Relax upper bound on `base` to allow base-4.9 - - - - - 0bfa0475 by Simon Peyton Jones at 2015-10-31T19:08:13+00:00 More adaption to wildcard-refactor - - - - - 0a3c0cb7 by Simon Peyton Jones at 2015-10-31T22:14:43+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor Conflicts: haddock-api/src/Haddock/Convert.hs - - - - - c4fd4ec9 by Alan Zimmerman at 2015-11-01T11:16:34+01:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 42cdd882 by Matthew Pickering at 2015-11-06T20:02:16+00:00 Change for IEThingWith - - - - - f368b7be by Ben Gamari at 2015-11-11T11:35:51+01:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - e32965b8 by Simon Peyton Jones at 2015-11-13T12:18:17+00:00 Merge with origin/head - - - - - ebcf795a by Edward Z. Yang at 2015-11-13T21:56:27-08:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4e23989f by Simon Peyton Jones at 2015-11-18T11:32:54+00:00 Wibbles to Haddock - - - - - 2289cd4a by Simon Peyton Jones at 2015-11-20T23:12:49+00:00 Merge remote-tracking branch 'origin/ghc-head' into wip/spj-wildcard-refactor - - - - - 695975a6 by Alan Zimmerman at 2015-11-21T21:16:12+02:00 Update to match GHC wip/T11019 - - - - - bbba21e7 by Simon Peyton Jones at 2015-11-23T13:54:31+00:00 merge with origin/ghc-head - - - - - 3d664258 by Simon Peyton Jones at 2015-11-23T17:17:18+00:00 Wibble - - - - - e64cf586 by Herbert Valerio Riedel at 2015-12-05T00:29:55+01:00 Canonicalise Monad instances - - - - - a2de15a7 by Alan Zimmerman at 2015-12-05T17:33:52+02:00 Matching changes for haskell/haddock#11028 - - - - - cc29a3e4 by Alan Zimmerman at 2015-12-05T19:45:33+02:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 95dd15d1 by Richard Eisenberg at 2015-12-11T17:33:39-06:00 Update for type=kinds - - - - - cb5fd9ed by Herbert Valerio Riedel at 2015-12-14T15:07:30+00:00 Bump versions for ghc-7.11 - - - - - 4f286d96 by Simon Peyton Jones at 2015-12-14T15:10:56+00:00 Eliminate instanceHead' in favour of GHC's instanceSig This is made possible by the elimination of "silent superclass parameters" in GHC - - - - - 13ea2733 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Remove redundant constraints from haddock, discovered by -fwarn-redundant-constraints - - - - - 098df8b8 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track changes in HsSyn for quasi-quotes - - - - - 716a64de by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track change in API of TyCon - - - - - 77a66bca by Adam Gundry at 2015-12-14T15:10:58+00:00 Track API changes to support empty closed type familes - - - - - f2808305 by Simon Peyton Jones at 2015-12-14T15:10:58+00:00 Track the new location of setRdrNameSpace - - - - - ba8b08a4 by Alan Zimmerman at 2015-12-14T15:10:59+00:00 ApiAnnotations : strings in warnings do not return SourceText The strings used in a WARNING pragma are captured via strings :: { Located ([AddAnn],[Located FastString]) } : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) } .. The STRING token has a method getSTRINGs that returns the original source text for a string. A warning of the form {-# WARNING Logic , mkSolver , mkSimpleSolver , mkSolverForLogic , solverSetParams , solverPush , solverPop , solverReset , solverGetNumScopes , solverAssertCnstr , solverAssertAndTrack , solverCheck , solverCheckAndGetModel , solverGetReasonUnknown "New Z3 API support is still incomplete and fragile: \ \you may experience segmentation faults!" #-} returns the concatenated warning string rather than the original source. - - - - - a4ded87e by Thomas Winant at 2015-12-14T15:14:05+00:00 Update after wild card renaming refactoring in D613 Summary: * Move `Post*` type instances to `Haddock.Types` as other modules than `Haddock.Interface.Rename` will rely on these type instances. * Update after wild card renaming refactoring in D613. Reviewers: simonpj, austin Reviewed By: austin Differential Revision: https://phabricator.haskell.org/D954 GHC Trac Issues: haskell/haddock#10098 - - - - - 25c78107 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: print correct strictness marks - - - - - 6cbc41c4 by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 StrictData: changes in HsBang type - - - - - ad46821a by Alan Zimmerman at 2015-12-14T15:14:06+00:00 Replace (SourceText,FastString) with WithSourceText data type Phab:D907 introduced SourceText for a number of data types, by replacing FastString with (SourceText,FastString). Since this has an Outputable instance, no warnings are generated when ppr is called on it, but unexpected output is generated. See Phab:D1096 for an example of this. Replace the (SourceText,FastString) tuples with a new data type data WithSourceText = WithSourceText SourceText FastString Trac ticket: haskell/haddock#10692 - - - - - abc0ae5b by Adam Sandberg Eriksson at 2015-12-14T15:14:06+00:00 HsBang is split into HsSrcBang and HsImplBang With recent changes in GHC handling of strictness annotations in Haddock is simplified. - - - - - 3308d06c by Thomas Miedema at 2015-12-14T15:14:07+00:00 Follow changes in GHC build system - - - - - 6c763deb by Eric Seidel at 2015-12-14T15:14:07+00:00 account for changes to ipClass - - - - - ae5b4eac by Jan Stolarek at 2015-12-14T15:17:00+00:00 Follow changes from haskell/haddock#6018 - - - - - ffbc40e0 by Richard Eisenberg at 2015-12-14T15:17:02+00:00 React to refactoring CoAxiom branch lists. - - - - - d1f531e9 by Edward Z. Yang at 2015-12-14T15:17:02+00:00 Track msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 79f73754 by Tamar Christina at 2015-12-14T15:17:02+00:00 Create Process: removed PhaseFailed - - - - - 3d37bebb by Edward Z. Yang at 2015-12-14T15:20:46+00:00 s/PackageKey/UnitId/g and s/packageKey/unitId/g Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 5f8a9e44 by Adam Gundry at 2015-12-14T15:20:48+00:00 Roughly fix up haddock for DuplicateRecordFields changes This compiles, but will probably need more work to produce good documentation when the DuplicateRecordFields extension is used. - - - - - 79dda70f by Simon Peyton Jones at 2015-12-14T15:26:02+00:00 Track wip/spj-wildcard-refactor on main repo - - - - - 959930fb by Simon Peyton Jones at 2015-12-14T15:37:50+00:00 Follow changes to HsTYpe Not yet complete (but on a wip/ branch) - - - - - e18a8df5 by Simon Peyton Jones at 2015-12-14T15:37:52+00:00 Work on updating Haddock to wip/spj-wildard-recactor Still incomplete - - - - - aa35ab52 by Simon Peyton Jones at 2015-12-14T15:40:18+00:00 More adaption to wildcard-refactor - - - - - 8ceef94b by Simon Peyton Jones at 2015-12-14T15:46:04+00:00 Track change to PatSyn.patSynSig - - - - - cd81e83d by Ben Gamari at 2015-12-14T15:46:06+00:00 Account for Typeable changes The treatment of type families changed. - - - - - 63c9117c by Herbert Valerio Riedel at 2015-12-14T15:46:34+00:00 Relax upper bound on `base` to allow base-4.9 - - - - - a484c613 by Alan Zimmerman at 2015-12-14T15:47:46+00:00 Matching change GHC haskell/haddock#11017 BooleanFormula located - - - - - 2c26fa51 by Matthew Pickering at 2015-12-14T15:47:47+00:00 Change for IEThingWith - - - - - 593baa0f by Ben Gamari at 2015-12-14T15:49:21+00:00 Eliminate support for deprecated GADT syntax Follows from GHC D1460. - - - - - b6b5ca78 by Edward Z. Yang at 2015-12-14T15:49:54+00:00 Undo msHsFilePath change. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b5b0e072 by Alan Zimmerman at 2015-12-14T15:54:20+00:00 Update to match GHC wip/T11019 - - - - - 14ddeb68 by Simon Peyton Jones at 2015-12-14T15:54:22+00:00 Wibble - - - - - 10a90ad8 by Herbert Valerio Riedel at 2015-12-14T15:54:22+00:00 Canonicalise Monad instances - - - - - ed68ac50 by Alan Zimmerman at 2015-12-14T15:55:48+00:00 Matching changes for haskell/haddock#11028 - - - - - 3f7e5a2d by Alan Zimmerman at 2015-12-14T15:55:49+00:00 Placeholder for record style GADT declaration A GADT Declaration is now presented as CmmCondBranch :: {..} -> CmmNode O C cml_pred :: CmmExpr cml_true, cml_false :: !Label cml_likely :: Maybe Bool for CmmCondBranch :: { -- conditional branch cml_pred :: CmmExpr, cml_true, cml_false :: ULabel, cml_likely :: Maybe Bool -- likely result of the conditional, -- if known } -> CmmNode O C - - - - - 6543a73f by Richard Eisenberg at 2015-12-14T15:59:55+00:00 Update for type=kinds - - - - - 193a5c48 by Matthew Pickering at 2015-12-14T18:17:00+00:00 Changes to compile with 8.0 - - - - - add669ec by Matthew Pickering at 2015-12-14T18:47:12+00:00 Warnings - - - - - 223f3fb4 by Ben Gamari at 2015-12-15T23:45:05+01:00 Update for D1200 - - - - - d058388f by Ben Gamari at 2015-12-16T05:40:17-05:00 Types: Add Outputable[Bndr] DocName instances - - - - - 62ecd7fb by Ben Gamari at 2015-12-16T09:23:09-05:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ddbc187a by Ben Gamari at 2015-12-16T17:54:55+01:00 Update for D1200 - - - - - cec83b52 by Ben Gamari at 2015-12-16T17:54:55+01:00 Types: Add Outputable[Bndr] DocName instances - - - - - d12ecc98 by Ben Gamari at 2015-12-16T17:54:55+01:00 Fix fallout from wildcards refactoring The wildcard refactoring was introduced a new type of signature, `ClassOpSig`, which is carried by typeclasses. The original patch adapting Haddock for this change missed a few places where this constructor needed to be handled, resulting in no class methods in documentation produced by Haddock. Additionally, this moves and renames the `isVanillaLSig` helper from GHC's HsBinds module into GhcUtils, since it is only used by Haddock. - - - - - ada1616f by Ben Gamari at 2015-12-16T17:54:58+01:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a4f0383d by Ben Gamari at 2015-12-16T23:32:38+01:00 Fix Hyperlinker GHC.con_names is now GHC.getConNames - - - - - a10e6849 by Ben Gamari at 2015-12-20T00:54:11+01:00 Merge remote-tracking branch 'mrhania/testing-framework-improvements' into ghc-head - - - - - f078b4fd by Ben Gamari at 2015-12-20T00:59:51+01:00 test: Compatibility with Cabal 1.23 - - - - - 88a511a9 by Ben Gamari at 2015-12-20T01:14:35+01:00 Merge remote-tracking branch 'phadej/orphans' into ghc-head - - - - - 4e250f36 by Ben Gamari at 2015-12-20T01:14:52+01:00 Add html-test for orphan instances output - - - - - 87fffbad by Alan Zimmerman at 2015-12-20T09:50:42+02:00 Update for GHC trac#11258 Adding locations to RdrName in FieldOcc and AmbiguousFieldOcc - - - - - 6b7e51c9 by idontgetoutmuch at 2015-12-20T21:01:47+00:00 Merge pull request haskell/haddock#1 from haskell/ghc-head Ghc head - - - - - 229c1fb5 by Dominic Steinitz at 2015-12-21T07:19:16+00:00 Handle inline math with mathjax. - - - - - 57902d66 by Dominic Steinitz at 2015-12-21T08:07:11+00:00 Fix the documentation for haddock itself. Change notation and add support for inline math. Allow newlines in display math. Add a command line option for the mathjax url (you might want to use a locally installed version). Rebase tests because of extra url and version change. Respond to (some of the) comments. Fix warnings in InterfaceFile.hs - - - - - 0e69f236 by Herbert Valerio Riedel at 2015-12-21T18:30:43+01:00 Fix-up left-over assumptions of GHC 7.12 into GHC 8.0 - - - - - c67f8444 by Simon Peyton Jones at 2015-12-22T16:26:56+00:00 Follow removal of NamedWildCard from HsType - - - - - da40327a by Ben Gamari at 2015-12-23T14:15:28+01:00 html-test/Operators: Clear up ambiguous types For reasons that aren't entirely clear a class with ambiguous types was accepted by GHC <8.0. I've added a functional dependency to clear up this ambiguity. - - - - - 541b7fa4 by Ben Gamari at 2015-12-23T14:18:51+01:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 0febc947 by Ben Gamari at 2015-12-24T00:30:20+01:00 hoogle-test/AssocTypes: Allow AmbiguousTypes GHC 8.0 complains otherwise - - - - - 25810841 by Ben Gamari at 2015-12-24T00:33:18+01:00 OrphanInstances: Accept test output - - - - - 841987f3 by Ben Gamari at 2015-12-25T11:03:11+01:00 Merge remote-tracking branch 'idontgetoutmuch/ghc-head' into ghc-head - - - - - 358391f0 by Ben Gamari at 2015-12-26T10:44:50+01:00 Add missing import - - - - - a8896885 by Ben Gamari at 2015-12-26T10:45:27+01:00 travis: Use Travis containers - - - - - 85e82134 by Herbert Valerio Riedel at 2015-12-30T17:25:39+01:00 tweak version bounds for GHC-8.1 - - - - - 672a5f75 by randen at 2016-01-01T23:45:25-08:00 The Haddock part for fully gcc-like response files " driver/Main.hs * Moved the response file handling into ResponseFile.hs, updating import section as appropriate. * driver/ResponseFile.hs * New file. In anticipation that maybe some day this could be provided by another library, and to make it possible to unit test, this functionality is pulled out of the Main.hs module, and expanded to support the style/format of response files which gcc uses. * The specification for the format of response files which gcc generates and consumes, seems to be best derived from the gcc code itself (libiberty/argv.c), so that is what has been done here. * This is intended to fix haskell/haddock#379 * driver-test/Main.hs * New file for testing code in the driver source tree * driver-test/ResponseFileSpec.hs * Tests, adapted/adopted from the same gcc code where the escaping/unescaping is from, in the hspec style of unit tests * haddock.cabal * Add the driver-test test-suite. Introduces a new library dependency (upon hspec) for the haddock driver target in the haddock.cabal file, but practically, this should not be a problem as the haddock-api tests already depend on hspec. - - - - - 498781df by Ben Gamari at 2016-01-06T13:41:04+01:00 Version bumps and changelog - - - - - 8451e46a by Ben Gamari at 2016-01-06T13:47:17+01:00 Merge remote-tracking branch 'randen/bug468' - - - - - fb2d9181 by Ben Gamari at 2016-01-06T08:14:42-05:00 Add ResponseFile to OtherModules - - - - - 2cb2d2e3 by Ben Gamari at 2016-01-06T14:35:00+01:00 Merge branch 'master' into ghc-head - - - - - 913477d4 by Eric Seidel at 2016-01-11T14:57:57-08:00 deal with un-wiring of IP class - - - - - c557a4b3 by Alan Zimmerman at 2016-01-15T11:14:35+02:00 Update to match wip/T11430 in GHC - - - - - 3e135093 by Alan Zimmerman at 2016-01-16T18:21:59+01:00 Update to match wip/T11430 in GHC - - - - - c48ef2f9 by Ben Gamari at 2016-01-18T09:50:06+01:00 Merge remote-tracking branch 'gridaphobe/ghc-head' into ghc-head - - - - - 9138a1b0 by Eric Seidel at 2016-01-18T12:50:15+01:00 deal with un-wiring of IP class (cherry picked from commit 17388b0f0029d969d79353be7737eb01c7b8dc5f) - - - - - b48c172e by Joachim Breitner at 2016-01-19T00:11:38+01:00 Make sure --mathjax affects all written HTML files This fixes haskell/haddock#475. - - - - - af61fe63 by Ryan Scott at 2016-02-07T23:25:57+01:00 Render */# instead of TYPE 'Lifted/TYPE 'Unlifted (fixes haskell/haddock#473) - - - - - b6458693 by Ben Gamari at 2016-02-07T23:29:27+01:00 Merge pull request haskell/haddock#477 from haskell/issue-475 Make sure --mathjax affects all written HTML files - - - - - adcc0071 by Ben Gamari at 2016-02-07T23:34:52+01:00 Merge branch 'master' into ghc-head - - - - - d0404e61 by Ben Gamari at 2016-02-08T12:46:49+01:00 doc: Switch to Sphinx - - - - - acb153b3 by Ben Gamari at 2016-02-08T12:46:56+01:00 Document --use-unicode flag - - - - - c20bdf1d by Ben Gamari at 2016-02-08T13:41:24+01:00 Fix GHC and haddock-library dependency bounds - - - - - 8d946801 by Ben Gamari at 2016-02-08T14:54:56+01:00 testsuite: Rework handling of output sanitization Previously un-cleaned artifacts were kept as reference output, making it difficult to tell what has changed and causing spurious changes in the version control history. Here we rework this, cleaning the output during acceptance. To accomplish this it was necessary to move to strict I/O to ensure the reference handle was closed before accept attempts to open the reference file. - - - - - c465705d by Ben Gamari at 2016-02-08T15:36:05+01:00 test: Compare on dump For reasons I don't understand the Xml representations differ despite their textual representations being identical. - - - - - 1ec0227a by Ben Gamari at 2016-02-08T15:36:05+01:00 html-test: Accept test output - - - - - eefbd63a by Ben Gamari at 2016-02-08T15:36:08+01:00 hypsrc-test: Accept test output And fix impredicative Polymorphism testcase. - - - - - d1df4372 by Ben Gamari at 2016-02-08T15:40:44+01:00 Merge branch 'fix-up-testsuite' - - - - - 206a3859 by Phil Ruffwind at 2016-02-08T17:51:21+01:00 Move the permalinks to "#" on the right side Since pull request haskell/haddock#407, the identifiers have been permalinked to themselves, but this makes it difficult to copy the identifier by double-clicking. To work around this usability problem, the permalinks are now placed on the far right adjacent to "Source", indicated by "#". Also, 'namedAnchor' now uses 'id' instead of 'name' (which is obsolete). - - - - - 6c89fa03 by Phil Ruffwind at 2016-02-08T17:54:44+01:00 Update tests for previous commit - - - - - effaa832 by Ben Gamari at 2016-02-08T17:56:17+01:00 Merge branch 'anchors-redux' - - - - - 9a2bec90 by Ben Gamari at 2016-02-08T17:58:40+01:00 Use -fprint-unicode-syntax when --use-unicode is enabled This allows GHC to render `*` as its Unicode representation, among other things. - - - - - 28ecac5b by Ben Gamari at 2016-02-11T18:53:03+01:00 Merge pull request haskell/haddock#480 from bgamari/sphinx Move documentation to ReStructuredText - - - - - 222e5920 by Ryan Scott at 2016-02-11T15:42:42-05:00 Collapse type/data family instances by default - - - - - a80ac03b by Ryan Scott at 2016-02-11T20:17:09-05:00 Ensure expanded family instances render correctly - - - - - 7f985231 by Ben Gamari at 2016-02-12T10:04:22+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - d4eda086 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Various cleanups - - - - - 79bee48d by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show kind signatures for type family variables Addresses GHC haskell/haddock#11588. - - - - - b2981d98 by Ben Gamari at 2016-02-18T00:05:56+01:00 Xhtml.Decl: Show 'where ...' after closed type family Seems like we should ideally show the actual equations as well but that seems like it would be a fair amount of work - - - - - cfc0e621 by Ben Gamari at 2016-02-18T22:48:12+01:00 Merge pull request haskell/haddock#483 from bgamari/T11588 Fix GHC haskell/haddock#11588 This fixes GHC haskell/haddock#11588: * Show where ... after closed type families * Show kind signatures on type family type variables - - - - - 256e8a0d by Ben Gamari at 2016-02-18T23:15:39+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 32402036 by Richard Eisenberg at 2016-02-24T13:21:44-05:00 Follow-on changes to support RuntimeRep - - - - - 2b1c572d by Matthew Pickering at 2016-03-04T21:04:02+00:00 Remove unused functions - - - - - eb906f50 by Richard Eisenberg at 2016-03-13T21:17:20+01:00 Follow-on changes to support RuntimeRep (cherry picked from commit ab954263a793d8ced734459d6194a5d89214b66c) - - - - - 8c34ef34 by Richard Eisenberg at 2016-03-14T23:47:23-04:00 Changes due to fix for GHC#11648. - - - - - 0e022014 by Richard Eisenberg at 2016-03-15T14:06:45+01:00 Changes due to fix for GHC#11648. (cherry picked from commit bb994de1ab0c76d1aaf1e39c54158db2526d31f1) - - - - - ed3f78ab by Rik Steenkamp at 2016-04-02T22:20:36+01:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 - - - - - d3210042 by Rik Steenkamp at 2016-04-04T15:43:32+02:00 Fix printing of pattern synonym types Removes the call to `patSynType :: PatSyn -> Type` in `Convert.hs` as this function will be removed from GHC. Instead, we use the function `patSynSig` and build the `HsDecl` manually. This also fixes the printing of the two contexts and the quantified type variables in a pattern synonym type. Reviewers: goldfire, bgamari, mpickering Differential Revision: https://phabricator.haskell.org/D2048 (cherry picked from commit 3ddcbd6b8e6884bd95028381176eb33bee6896fb) - - - - - 236eec90 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 692ee7e0 by Ben Gamari at 2016-04-10T23:40:15+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. - - - - - 79619f57 by Ben Gamari at 2016-04-10T23:46:22+02:00 doc: Only install if BUILD_SPHINX_HTML==YES Fixes GHC haskell/haddock#11818. (cherry picked from commit c6d6a18d85e5e2d9bb5904e6919e8a8d7e31c4c5) - - - - - 3358ccb4 by Ben Gamari at 2016-04-10T23:47:27+02:00 doc: Fix option references (cherry picked from commit f915fb3c74328fb994235bbbd42092a691539197) - - - - - 264949b1 by Ben Gamari at 2016-04-16T17:50:23+02:00 Merge pull request haskell/haddock#482 from RyanGlScott/ghc-head Collapse type/data family instances by default - - - - - 478c483a by Ben Gamari at 2016-04-16T17:51:09+02:00 Merge pull request haskell/haddock#489 from mpickering/unused-functions Remove some unused functions - - - - - c94e55f0 by Ryan Scott at 2016-04-16T17:57:54+02:00 Collapse type/data family instances by default (cherry picked from commit 2da130a8db8f995c119b544fad807533236cf088) - - - - - 31e633d3 by Ryan Scott at 2016-04-16T17:58:06+02:00 Ensure expanded family instances render correctly (cherry picked from commit 1338b5d7c32939de6bbc31af0049477e4f847103) - - - - - 03e4d197 by Matthew Pickering at 2016-04-16T17:58:21+02:00 Remove unused functions (cherry picked from commit b89d1c2456bdb2d4208d94ded56155f7088a37d0) - - - - - ed4116f6 by Ben Gamari at 2016-04-20T10:46:57+02:00 ghc: Install files for needed --hyperlinked-source - - - - - 0be999c4 by Ben Gamari at 2016-04-20T11:37:54+02:00 ghc: Install files for needed --hyperlinked-source (cherry picked from commit 5c82c9fc2d21ddaae4a2470f1c375426968f19c6) - - - - - 4d17544c by Simon Peyton Jones at 2016-04-20T12:42:28+01:00 Track change to HsGroup This relates to a big GHC patch for Trac haskell/haddock#11348 - - - - - 1700a50d by Ben Gamari at 2016-05-01T13:19:27+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. - - - - - 0b7c8125 by Ben Gamari at 2016-05-01T13:21:43+02:00 doc: At long last fix ghc.mk The variable reference was incorrectly escaped, meaning that Sphinx documentation was never installed. (cherry picked from commit 609018dd09c4ffe27f9248b2d8b50f6196cd42b9) - - - - - af115ce0 by Ryan Scott at 2016-05-04T22:15:50-04:00 Render Haddocks for derived instances Currently, one can document top-level instance declarations, but derived instances (both those in `deriving` clauses and standalone `deriving` instances) do not enjoy the same privilege. This makes the necessary changes to the Haddock API to enable rendering Haddock comments for derived instances. This is part of a fix for Trac haskell/haddock#11768. - - - - - 76fa1edc by Ben Gamari at 2016-05-10T18:13:25+02:00 haddock-test: A bit of refactoring for debuggability - - - - - 7d4c4b20 by Ben Gamari at 2016-05-10T18:13:25+02:00 Create: Mark a comment as TODO - - - - - 2a6d0c90 by Ben Gamari at 2016-05-10T18:13:25+02:00 html-test: Update reference output - - - - - bd60913d by Ben Gamari at 2016-05-10T18:13:25+02:00 hypsrc-test: Fix reference file path in cabal file It appears the haddock insists on prefixing --hyperlinked-sourcer output with directory which the source appeared in. - - - - - c1548057 by Ben Gamari at 2016-05-10T18:22:12+02:00 doc: Update extra-source-files in Cabal file - - - - - 41d5bae3 by Ben Gamari at 2016-05-10T18:29:21+02:00 Bump versions - - - - - ca75b779 by Ben Gamari at 2016-05-11T16:03:44+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 4e3cfd62 by Ben Gamari at 2016-05-11T16:06:45+02:00 Merge remote-tracking branch 'RyanGlScott/ghc-head' into ghc-head - - - - - a2379970 by Ben Gamari at 2016-05-11T23:15:11+02:00 doc: Add clean targets - - - - - f275212e by Ben Gamari at 2016-05-11T23:15:14+02:00 doc: Add html as an all-target for ghc Otherwise the html documentation won't be installed for binary-dist. - - - - - 388fc0af by Ben Gamari at 2016-05-12T09:49:12+02:00 Update CHANGES - - - - - bad81ad5 by Ben Gamari at 2016-05-12T09:49:38+02:00 Version bump - - - - - c01688a7 by Ben Gamari at 2016-05-12T10:04:58+02:00 Revert "Version bump" This bump was a bit premature. This reverts commit 7b238d9c5be9b07aa2d10df323b5c7b8d1634dc8. - - - - - 7ed05724 by Ben Gamari at 2016-05-12T10:05:33+02:00 doc: Fix GHC clean rule Apparently GHC's build system doesn't permit wildcards in clean paths. - - - - - 5d9611f4 by Ben Gamari at 2016-05-12T17:43:50+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 653566b2 by Ben Gamari at 2016-05-14T09:57:31+02:00 Version bump to 2.17.2 - - - - - b355c439 by Ben Gamari at 2016-05-14T09:57:51+02:00 doc: Use `$(MAKE)` instead of `make` This is necessary to ensure we use gmake. - - - - - 8a18537d by Ben Gamari at 2016-05-14T10:15:45+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - b3290ef1 by Sebastian Meric de Bellefon at 2016-05-14T11:29:47-04:00 Fix haskell/haddock#303. Hide footer when printing The "Produced by Haddock" footer was overlapping the page's body when printing. This patch hides the footer with a css media rule. - - - - - b4a76f89 by Sebastian Meric de Bellefon at 2016-05-15T02:12:46-04:00 Fix haskell/haddock#280. Parsing of module header The initial newlines were counted as indentation spaces, thus disturbing the parsing of next lines - - - - - ba797c9e by Ben Gamari at 2016-05-16T14:53:46+02:00 doc: Vendorize alabaster Sphinx theme Alabaster is now the default sphinx theme and is a significant improvement over the previous default that it's worthproviding it when unavailable (e.g. Sphinx <1.3). - - - - - c9283e44 by Ben Gamari at 2016-05-16T14:55:17+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 1c9ea198 by Sebastian Méric de Bellefon at 2016-05-16T12:30:40-04:00 Merge pull request haskell/haddock#502 from Helkafen/master Fix haskell/haddock#303. Hide footer when printing - - - - - 33631016 by Ben Gamari at 2016-05-16T19:56:11+02:00 Revert "doc: Vendorize alabaster Sphinx theme" This ended up causes far too many issues to be worthwhile. We'll just have to live with inconsistent haddock documentation. This reverts commit cec21957001143794e71bcd9420283df18e7de40. - - - - - 93317d26 by Ben Gamari at 2016-05-16T19:56:11+02:00 cabal: Fix README path - - - - - c8695b22 by Ben Gamari at 2016-05-16T19:58:51+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 0b50eaaa by Ben Gamari at 2016-05-16T21:02:08+02:00 doc: Use whichever theme sphinx deems appropriate - - - - - 857c1c9c by Ben Gamari at 2016-05-16T21:07:08+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 15fc5637 by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Remove redundant imports - - - - - 132ddc6a by Ben Gamari at 2016-05-22T12:43:59+02:00 Create: Better debug output For tracking down haskell/haddock#505 - - - - - 2252a149 by Ben Gamari at 2016-05-22T12:43:59+02:00 Don't consider default class ops when looking for decls When we are looking for an operation within a class we don't care about `default`-type declarations. This was the cause of haskell/haddock#505. - - - - - 4886b2ec by Oleg Grenrus at 2016-05-24T16:19:48+03:00 UnfelpfulSpan line number omitted Kind of resolves https://github.com/haskell/haddock/issues/508 - - - - - a4befd36 by Oleg Grenrus at 2016-05-24T16:53:35+03:00 Change Hyperlinked lexer to know about DataKinds ticks - - - - - f45cb52e by David Feuer at 2016-05-24T18:48:53-04:00 Make parser state a newtype Previously, it was `data` wrapping a `Maybe`, which seems a bit silly. Obviously, this can be changed back if anyone wants to add more fields some day. - - - - - 05013dd7 by Sebastian Meric de Bellefon at 2016-05-24T22:03:55-04:00 remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) Frames are a bit broken, ignored by Hackage, and considered obsolete in general. This patch disables frames generation. The mini_*.html files are still used in the synopsis. - - - - - b8163a88 by Ben Gamari at 2016-05-25T14:44:15+02:00 Merge pull request haskell/haddock#507 from bgamari/T505 Fix haskell/haddock#505 - - - - - ea1b30c6 by Sebastian Meric de Bellefon at 2016-05-25T14:17:00-04:00 Update CHANGES - - - - - eddfc258 by Sebastian Méric de Bellefon at 2016-05-25T15:17:40-04:00 Merge pull request haskell/haddock#514 from Helkafen/frames remove framed view of the HTML documentation (see haskell/haddock#114 and haskell/haddock#274) - - - - - 0e506818 by Alex Biehl at 2016-05-26T12:43:09+02:00 Remove misplaced haddock comment - - - - - a07d28c0 by Ben Gamari at 2016-05-27T11:34:59+02:00 Merge pull request haskell/haddock#515 from alexbiehl/master Remove misplaced haddock comment - - - - - 9001d267 by Ben Gamari at 2016-05-27T11:35:46+02:00 Merge pull request haskell/haddock#513 from treeowl/newtype-since Make parser state a newtype - - - - - 74e1a018 by Sebastian Méric de Bellefon at 2016-05-28T17:28:15-04:00 Merge pull request haskell/haddock#504 from Helkafen/issue-280 Fix haskell/haddock#280. Parsing of module header - - - - - 37557f4f by Alan Zimmerman at 2016-05-29T23:36:50+02:00 Matching changes for haskell/haddock#12105 - - - - - 7d09e5d6 by Sebastian Meric de Bellefon at 2016-06-03T18:07:48-04:00 Version bumps (2.17.3, 1.4.2) - - - - - 85b4bc15 by Sebastian Méric de Bellefon at 2016-06-06T18:35:13-04:00 Merge pull request haskell/haddock#521 from Helkafen/master Version bumps (2.17.3, 1.4.2) - - - - - e95f0dee by Sebastian Meric de Bellefon at 2016-06-06T19:11:35-04:00 publish haddock-test library - - - - - 4de40586 by Sebastian Méric de Bellefon at 2016-06-06T20:26:30-04:00 Merge pull request haskell/haddock#512 from phadej/oleg-fixes Fixes for haskell/haddock#508 and haskell/haddock#510 - - - - - ddfd0789 by Dominic Steinitz at 2016-06-09T09:27:28+01:00 Documentation for LaTeX markup. - - - - - 697a503a by Dominic Steinitz at 2016-06-09T09:33:59+01:00 Fix spelling mistake. - - - - - 246f6fff by Dominic Steinitz at 2016-06-09T09:37:15+01:00 Camel case MathJax. - - - - - 4684bd23 by Dominic Steinitz at 2016-06-09T09:44:53+01:00 Fix math typo and add link. - - - - - f20c037c by Simon Peyton Jones at 2016-06-13T18:26:03+01:00 Follow changes to LHsSigWcType - - - - - 0c58996d by Simon Peyton Jones at 2016-06-15T12:56:01+01:00 Follow GHC re-adding FunTy - - - - - 401b5ca7 by Sebastian Méric de Bellefon at 2016-06-15T12:16:47-04:00 Merge pull request haskell/haddock#525 from idontgetoutmuch/master Documentation for LaTeX markup. - - - - - 92d263b7 by Sebastian Méric de Bellefon at 2016-06-15T12:17:29-04:00 Merge pull request haskell/haddock#522 from Helkafen/master publish haddock-test library - - - - - 0953a2ca by Sebastian Meric de Bellefon at 2016-06-16T00:46:46-04:00 Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 65453e14 by Ben Gamari at 2016-06-16T11:16:32+02:00 ocean: Ensure that synopsis fully covers other content Previously MathJax content was being rendered on top of the synopsis due to ambiguous z-ordering. Here we explicitly give the synopsis block a higher z-index to ensure it is rendered on top. Fixes haskell/haddock#531. - - - - - 68e411a1 by Sebastian Méric de Bellefon at 2016-06-16T23:34:39-04:00 Merge pull request haskell/haddock#534 from bgamari/T531 ocean: Ensure that synopsis fully covers other content - - - - - fad6491b by Sebastian Méric de Bellefon at 2016-06-18T23:57:20-04:00 Merge pull request haskell/haddock#533 from Helkafen/master Copyright holders shown on several lines. Fix haskell/haddock#279 - - - - - 6108e21b by Sebastian Meric de Bellefon at 2016-06-22T23:08:28-04:00 do not create empty src directory Fix haskell/haddock#536. - - - - - 1ef23823 by Sebastian Méric de Bellefon at 2016-06-24T00:04:48-04:00 Merge pull request haskell/haddock#537 from Helkafen/master do not create empty src directory - - - - - 966baa96 by Omari Norman at 2016-06-29T21:59:34-04:00 Add $ as a special character If this character is not escaped, documentation built with Haddock 2.17.2 will fail. This was not an issue with 2.16 series, which causes builds to fail and there is nothing in the docs or error message giving a clue about why builds that used to succeed now don't. - - - - - 324adb60 by Ben Gamari at 2016-07-01T12:18:51+02:00 GhcUtils: Changes for multi-pattern signatures - - - - - d7571675 by Ömer Sinan Ağacan at 2016-07-21T13:30:47+02:00 Add support for unboxed sums - - - - - 29d0907b by Simon Marlow at 2016-07-22T13:55:48+01:00 Disable NFData instances for GHC types when GHC >= 8.2 - - - - - 702d95f3 by Simon Marlow at 2016-08-02T15:57:30+02:00 Disable NFData instances for GHC types when GHC >= 8.0.2 (cherry picked from commit a3309e797c42dae9bccdeb17ce52fcababbaff8a) - - - - - f4fa79c3 by Ben Gamari at 2016-08-07T13:51:18+02:00 ghc.mk: Don't attempt to install html/frames.html The frames business has been removed. - - - - - 9cd63daf by Ben Gamari at 2016-08-07T13:51:40+02:00 Haddock.Types: More precise version guard This allows haddock to be built with GHC 8.0.2 pre-releases. - - - - - f3d7e03f by Mateusz Kowalczyk at 2016-08-29T20:47:45+01:00 Merge pull request haskell/haddock#538 from massysett/master Add $ as a special character - - - - - 16dbf7fd by Bartosz Nitka at 2016-09-20T19:44:04+01:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 - - - - - 7c31c1ff by Bartosz Nitka at 2016-09-27T17:32:22-04:00 Fix rendering of class methods for Eq and Ord See haskell/haddock#549 and GHC issue haskell/haddock#12519 (cherry picked from commit 073d899a8f94ddec698f617a38d3420160a7fd0b) - - - - - 33a90dce by Ryan Scott at 2016-09-30T20:53:41-04:00 Haddock changes for T10598 See https://ghc.haskell.org/trac/ghc/ticket/10598 - - - - - 1f32f7cb by Ben Gamari at 2016-10-13T20:01:26-04:00 Update for refactoring of NameCache - - - - - 1678ff2e by Ben Gamari at 2016-11-15T17:42:48-05:00 Bump upper bound on base - - - - - 9262a7c5 by Alan Zimmerman at 2016-12-07T21:14:28+02:00 Match changes in GHC wip/T3384 branch - - - - - ac0eaf1a by Ben Gamari at 2016-12-09T09:48:41-05:00 haddock-api: Don't use stdcall calling convention on 64-bit Windows See GHC haskell/haddock#12890. - - - - - 04afe4f7 by Alan Zimmerman at 2016-12-12T20:07:21+02:00 Matching changes for GHC wip/T12942 - - - - - e1d1701d by Ben Gamari at 2016-12-13T16:50:41-05:00 Bump base upper bound - - - - - 3d3eacd1 by Alan Zimmerman at 2017-01-10T16:59:38+02:00 HsIParamTy now has a Located name - - - - - 7dbceefd by Kyrill Briantsev at 2017-01-12T13:23:50+03:00 Prevent GHC API from doing optimization passes. - - - - - d48d1e33 by Richard Eisenberg at 2017-01-19T08:41:41-05:00 Upstream changes re levity polymorphism - - - - - 40c25ed6 by Alan Zimmerman at 2017-01-26T15:16:18+02:00 Changes to match haskell/haddock#13163 in GHC - - - - - 504f586d by Ben Gamari at 2017-02-02T17:19:37-05:00 Kill remaining static flags - - - - - 49147ea0 by Justus Adam at 2017-03-02T15:33:34+01:00 Adding MDoc to exports of Documentation.Haddock - - - - - 1cfba9b4 by Justus Adam at 2017-03-09T11:41:44+01:00 Also exposing toInstalledIface - - - - - 53f0c0dd by Ben Gamari at 2017-03-09T13:10:08-05:00 Bump for GHC 8.3 - - - - - c7902d2e by Ben Gamari at 2017-03-09T23:46:02-05:00 Bump for GHC 8.2 - - - - - 4f3a74f8 by Ben Gamari at 2017-03-10T10:21:55-05:00 Merge branch 'ghc-head' - - - - - e273b72f by Richard Eisenberg at 2017-03-14T13:34:04-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - 6ec3d436 by Richard Eisenberg at 2017-03-14T15:15:52-04:00 Update Haddock w.r.t. new HsImplicitBndrs - - - - - eee3cda1 by Ben Gamari at 2017-03-15T15:19:59-04:00 Adapt to EnumSet - - - - - 017cf58e by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 736d6773 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 475f84a0 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 13240b53 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - cd16d529 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 3bea97ae by Edward Z. Yang at 2017-03-15T22:50:46-07:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - b2b051ce by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 0f082795 by Edward Z. Yang at 2017-03-15T22:50:46-07:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 20ef63c9 by Edward Z. Yang at 2017-03-22T13:48:12-07:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 45692dcb by Edward Z. Yang at 2017-03-22T14:11:25-07:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 4eae8caf by Ben Gamari at 2017-03-23T09:25:33-04:00 Merge commit '240bc38b94ed2d0af27333b23392d03eeb615e82' into HEAD - - - - - 0bbe03f5 by Ben Gamari at 2017-03-23T09:27:28-04:00 haddock-api: Bump bound on GHC - - - - - 65f3ac9d by Alex Biehl at 2017-03-23T17:36:11+01:00 Merge pull request haskell/haddock#581 from JustusAdam/master Adding more exports to Documentation.Haddock - - - - - 37d49a47 by Alex Biehl at 2017-03-23T17:39:14+01:00 Merge pull request haskell/haddock#568 from awson/ghc-head Prevent GHC API from doing optimization passes. - - - - - 1ed047e4 by Brian Huffman at 2017-03-23T17:45:58+01:00 Print any user-supplied kind signatures on type parameters. This applies to type parameters on data, newtype, type, and class declarations, and also to forall-bound type vars in type signatures. - - - - - 1b78ca5c by Brian Huffman at 2017-03-23T17:45:58+01:00 Update test suite to expect kind annotations on type parameters. - - - - - a856b162 by Alex Biehl at 2017-03-23T17:49:32+01:00 Include travis build indication badge - - - - - 8e2e2c56 by Ben Gamari at 2017-03-23T17:20:08-04:00 haddock-api: Bump bound on GHC - - - - - 4d2d9995 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Correctly handle Backpack identity/semantic modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 26d6c150b31bc4580ab17cfd07b6e7f9afe10737) - - - - - a650e20f by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Add a field marking if interface is a signature or not. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 930cfbe58e2e87f5a4d431d89a3c204934e6e858) - - - - - caa282c2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render signature module tree separately from modules. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 2067a2d0afa9cef381d26fb7140b67c62f433fc0) - - - - - 49684884 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Documentation. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 0671abfe7e8ceae2269467a30b77ed9d9656e2cc) - - - - - 4dcfeb1a by Edward Z. Yang at 2017-03-23T17:20:08-04:00 More docs. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 3d77b373dd5807d5d956719dd7c849a11534fa6a) - - - - - 74dd19d2 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 TODO on moduleExports. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 94610e9b446324f4231fa6ad4c6ac51e4eba8c0e) - - - - - a9b19a23 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Better Backpack support with signature merging. When we merge signatures, we gain exports that don't necessarily have a source-level declaration corresponding to them. This meant Haddock dropped them. There are two big limitations: * If there's no export list, we won't report inherited signatures. * If the type has a subordinate, the current hiDecl implementation doesn't reconstitute them. These are probably worth fixing eventually, but this gets us to minimum viable functionality. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 6cc832dfb1de6088a4abcaae62b25a7e944d55c3) - - - - - d3631064 by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Fix haddock-test to work with latest version of Cabal. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit bf3c4d72a0fda38561376eac7eda216158783267) - - - - - ef2148fc by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Annotate signature docs with (signature) Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 07b88c5d4e79b87a319fbb08f8ea01dbb41063c1) - - - - - 2f29518b by Edward Z. Yang at 2017-03-23T17:20:08-04:00 Render help documentation link next to (signature) in title. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit 4eb765ca4205c79539d60b7afa9b7e261a4a49fe) - - - - - 37de047d by Phil Ruffwind at 2017-04-03T11:57:14+02:00 Update MathJax URL MathJax is shutting down their CDN: https://www.mathjax.org/cdn-shutting-down/ They recommend migrating to cdnjs. - - - - - e9d24ba8 by David C. Turner at 2017-04-03T14:58:01+02:00 Add highlight for :target to ocean.css - - - - - 4819a202 by Alex Biehl at 2017-04-11T19:36:48+02:00 Allow base-4.10 for haddock-test - - - - - 44cec69c by Alex Biehl at 2017-04-11T19:39:22+02:00 cabal.project for haddock-api, haddock-library and haddock-test - - - - - 935d0f6a by Alex Biehl at 2017-04-11T19:46:29+02:00 Move dist scripts to scripts/ - - - - - 128e150c by Alex Biehl at 2017-04-11T20:34:46+02:00 Add haddock to cabal.project - - - - - cc8e08ea by Alex Biehl at 2017-04-11T20:35:08+02:00 Read files for hyperlinker eagerly This also exposes Documentation.Haddock.Utf8 - - - - - 152dda78 by Alex Biehl at 2017-04-11T20:37:06+02:00 Explicit import list ofr Control.DeepSeq in Haddock.Interface.Create - - - - - 501b33c4 by Kyrill Briantsev at 2017-04-11T21:01:42+02:00 Prevent GHC API from doing optimization passes. - - - - - c9f3f5ff by Alexander Biehl at 2017-04-12T16:36:53+02:00 Add @alexbiehl as maintaner - - - - - 76f214cc by Alex Biehl at 2017-04-13T07:27:18+02:00 Disable doctest with ghc-8.3 Currently doctest doesn't support ghc-head - - - - - 46b4f5fc by Edward Z. Yang at 2017-04-22T20:38:26-07:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - f0555235 by Alex Biehl at 2017-04-25T10:08:48+02:00 Travis: Use ghc-8.2.1 on master - - - - - 966ea348 by Alex Biehl at 2017-04-25T10:32:01+02:00 Travis: Verbose cabal output cf. https://travis-ci.org/haskell/haddock/jobs/225512194#L377 - - - - - 36972bcd by Alex Biehl at 2017-04-25T10:40:43+02:00 Use travis_retry for cabal invocations - - - - - b3a09d2c by Alex Biehl at 2017-04-25T17:02:20+02:00 Use new MathJax URL in html-test 18ed871afb82560d5433b2f53e31b4db9353a74e switched to a new MathJax URL but didn't update the tests. - - - - - ae331e5f by Alexander Biehl at 2017-04-25T17:02:20+02:00 Expand signatures for class declarations - - - - - e573c65a by Alexander Biehl at 2017-04-25T17:02:20+02:00 Hoogle: Correctly print classes with associated data types - - - - - 3fc6be9b by Edward Z. Yang at 2017-04-25T17:02:20+02:00 Render (signature) only if it actually is a signature! I forgot a conditional, oops! Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> (cherry picked from commit a0c4790e15a2d3fab8d830eee8fcd639fe6d39c9) - - - - - 6725c060 by Herbert Valerio Riedel at 2017-04-25T17:02:20+02:00 `html-test --accept` deltas to reference samples - - - - - 7d444d61 by Alex Biehl at 2017-04-26T07:13:50+02:00 Remove anything related to obsolete frames mode - - - - - b888972c by Alex Biehl at 2017-04-26T07:49:10+02:00 Cherry-picked remaining commits from haddock-2.17.4-release (#603) * Release haddock/haddock-api 2.17.4 and haddock-library 1.4.3 * Set version bounds for haddock-library NB: This allows GHC 8.2.1's base * Set version bounds for haddock & haddock-api The version bounds support GHC 8.2 * Merge (temporary) v2.17.3 branch into v2.17 This allows us to delete the v2.17.3 branch * Fixup changelog * Pin down haddock-api to a single version as otherwise `haddock`'s package version has no proper meaning * fix source-repo spec for haddock-api - - - - - 4161099b by Alex Biehl at 2017-04-26T11:11:20+02:00 Update changelog to reflect news in HEAD - - - - - eed72cb8 by Alex Biehl at 2017-04-26T11:11:20+02:00 Markdownify changelog - - - - - 5815cea1 by Alex Biehl at 2017-04-26T11:32:33+02:00 Bump to 2.18.0 (#605) - - - - - a551d558 by Alex Biehl at 2017-04-29T22:00:25+02:00 Update attoparsec-0.12.1.1 to attoparsec-0.13.1.0 - - - - - ea164a8d by Sergey Vinokurov at 2017-04-29T22:42:36+02:00 Improve error message - - - - - 2e10122f by Alex Biehl at 2017-04-30T10:07:46+02:00 Correctly remember collapsed sections (#608) Now the "collapsed" cookie stores which sections have changed state instead of which are collapsed. - - - - - f9b24d99 by Alex Biehl at 2017-05-01T17:40:36+02:00 Lazily decode docMap and argMap (#610) These are only used in case of a doc reexport so most of the time decoding these is wasted work. - - - - - 2372af62 by Alex Biehl at 2017-05-01T21:59:23+02:00 Fix Binary instance for InstalledInterface (#611) (#610) introduced lazy decoding for docs from InstalledInterface but forgot to remove the original calls to get and put_ - - - - - 6c633c13 by Nathan Collins at 2017-05-11T11:47:55+02:00 Improve documenation of Haddock markup (#614) * Improve documentation of Haddock markup. - document that Haddock supports inferring types top-level functions with without type signatures, but also explain why using this feature is discouraged. Looks like this feature has been around since version 2.0.0.0 in 2008! - rework the "Module description" section: - move the general discussion of field formatting to the section intro and add examples illustrating the prose for multiline fields. - mention that newlines are preserved in some multiline fields, but not in others (I also noticed that commas in the `Copyright` field are not preserved; I'll look into this bug later). - add a subsection for the module description fields documentation, and put the field keywords in code formatting (double back ticks) instead of double quotes, to be consistent with the typesetting of keywords in other parts of the documentation. - mention that "Named chunks" are not supported in the long-form "Module description" documentation. - fix formatting of keywords in the "Module attributes" section. Perhaps these errors were left over from an automatic translation to ReST from some other format as part of the transition to using Sphinx for Haddock documentation? Also, add a missing reference here; it just said "See ?"! - update footnote about special treatment for re-exporting partially imported modules not being implemented. In my tests it's not implemented at all -- I tried re-exporting both `import B hiding (f)` and `import B (a, b)` style partial imports, and in both cases got the same result as with full imports `import B`: I only get a module reference. * Rework the `Controlling the documentation structure` section. My main goal was to better explain how to use Haddock without an export list, since that's my most common use case, but I hope I improved the section overall: - remove the incomplete `Omitting the export list` section and fold it into the other sections. In particular, summarize the differences between using and not using an export list -- i.e. control over what and in what order is documented -- in the section lead. - add "realistic" examples that use the structure markup, both with and without an export list. I wanted a realistic example here to capture how it can be useful to explain the relationship between a group of functions in a section, in addition to documenting their individual APIs. - make it clear that you can associate documentation chunks with documentation sections when you aren't using an export list, and that doing it in the most obvious way -- i.e. with `-- |`, as you can in the export list -- doesn't work without an export list. It took me a while to figure this out the first time, since the docs didn't explain it at all before. - add a "no export list" example to the section header section. - add more cross references. * Add examples of gotchas for markup in `@...@`. I'm not sure this will help anyone, since I think most people first learn about `@...@` by reading other people's Haddocks, but I've documented the mistakes which I've made and then gotten confused by. * Use consistent Capitalization of Titles. Some titles were in usual title caps, and others only had the first word capitalized. I chose making them all use title caps because that seems to make the cross references look better. - - - - - d4734f45 by Ben Gamari at 2017-05-12T20:36:08+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 93883f37 by Alex Biehl at 2017-05-12T21:02:33+02:00 Haddock: Fix broken lazy IO in prologue reading (#615) We previously used withFile in conjunction with hGetContents. The list returned by the latter wasn't completely forced by the time we left the withFile block, meaning that we would try to read from a closed handle. - - - - - 5b8f179c by Alex Biehl at 2017-05-13T12:48:10+02:00 Consequently use inClass and notInClass in haddock-library (#617) These allow attoparsec to do some clever lookup optimization - - - - - 77984b82 by Doug Wilson at 2017-05-27T17:37:38+02:00 Don't enable compilation for template haskell (#624) This is no longer necessary after ghc commit 53c78be0aab76a3107c4dacbb1d177afacdd37fa - - - - - 5a3de2b4 by Doug Wilson at 2017-05-27T19:54:53+02:00 Improve Syb code (#621) Specialize.hs and Ast.hs are modified to have their Syb code not recurse into Name or Id in HsSyn types. Specialize.hs is refactored to have fewer calls to Syb functions. Syb.hs has some foldl calls replaced with foldl' calls. There is still a lot of performance on the floor of Ast.hs. The RenamedSource is traversed many times, and lookupBySpan is very inefficient. everywhereBut and lookupBySpan dominate the runtime whenever --hyperlinked-source is passed. - - - - - 3d35a949 by Alex Biehl at 2017-05-30T19:01:37+02:00 Clear fixme comment (#625) - - - - - 2a44bd0c by Alex Biehl at 2017-05-30T19:02:12+02:00 Make haddock-library and haddock-api warning free (#626) - - - - - bd1a0e42 by Alex Biehl at 2017-06-01T10:40:33+02:00 Include `driver-test/*.hs` sdist (#630) This lead to haskell/haddock#629. - - - - - 184a3ab6 by Doug Wilson at 2017-06-03T12:02:08+02:00 Disable pattern match warnings (#628) This disables the pattern match checker which can be very expensive in some cases. The disabled warnings include: * Opt_WarnIncompletePatterns * Opt_WarnIncompleteUniPatterns * Opt_WarnIncompletePatternsRecUpd * Opt_WarnOverlappingPatterns - - - - - 0cf68004 by Alex Biehl at 2017-06-03T20:37:28+02:00 Allow user defined signatures for pattern synonyms (#631) - - - - - 7f51a58a by Alex Biehl at 2017-06-04T11:56:38+02:00 Use NameSet for isExported check (#632) - - - - - d8f044a9 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Match new AST as per GHC wip/new-tree-one-param See https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow - - - - - da1254e3 by Alan Zimmerman at 2017-06-05T22:26:55+02:00 Rename extension index tags - - - - - 538c7514 by Christiaan Baaij at 2017-06-09T08:26:43+02:00 Haddock support for bundled pattern synonyms (#627) * Haddock support for bundled pattern synonyms * Add fixities to bundled pattern synonyms * Add bundled pattern synonyms to the synopsis * Store bundled pattern fixities in expItemFixities * Add test for bundled pattern synonyms * Stop threading fixities * Include bundled pattern synonyms for re-exported data types Sadly, fixity information isn't found for re-exported data types * Support for pattern synonyms * Modify tests after haskell/haddock#631 * Test some reexport variations * Also lookup bundled pattern synonyms from `InstalledInterface`s * Check isExported for bundled pattern synonyms * Pattern synonym is exported check * Always look for pattern synonyms in the current module Another overlooked cornercase * Account for types named twice in export lists Also introduce a fast function for nubbing on a `Name` and use it throughout the code base. * correct fixities for reexported pattern synonyms * Fuse concatMap and map * Remove obsolete import * Add pattern synonyms to visible exports * Fix test * Remove corner case - - - - - a050bffd by Doug Wilson at 2017-06-21T09:27:33+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#636) There is some performance improvement. GHC compiler: | version | bytes allocated | cpu_seconds --------------------------------- | before | 56057108648 | 41.0 | after | 51592019560 | 35.1 base: | version | bytes allocated | cpu_seconds --------------------------------- | before | 25174011784 | 14.6 | after | 23712637272 | 13.1 Cabal: | version | bytes allocated | cpu_seconds --------------------------------- | before | 18754966920 | 12.6 | after | 18198208864 | 11.6 - - - - - 5d06b871 by Doug Wilson at 2017-06-22T20:23:29+02:00 Use new function getNameToInstancesIndex instead of tcRnGetInfo (#639) * Use new function getNameToInstancesIndex instead of tcRnGetInfo There is some significant performance improvement in the ghc testsuite. haddock.base: -23.3% haddock.Cabal: -16.7% haddock.compiler: -19.8% * Remove unused imports - - - - - b11bb73a by Alex Biehl at 2017-06-23T14:44:41+02:00 Lookup fixities for reexports without subordinates (#642) So we agree that reexported declarations which do not have subordinates (for example top-level functions) shouldn't have gotten fixities reexported according to the current logic. I wondered why for example Prelude.($) which is obviously reexported from GHC.Base has fixities attached (c.f. http://hackage.haskell.org/package/base-4.9.1.0/docs/Prelude.html#v:-36-). The reason is this: In mkMaps we lookup all the subordinates of top-level declarations, of course top-level functions don't have subordinates so for them the resulting list is empty. In haskell/haddock#644 I established the invariant that there won't be any empty lists in the subordinate map. Without the patch from haskell/haddock#642 top-level functions now started to fail reexporting their fixities. - - - - - d2a6dad6 by Alex Biehl at 2017-06-23T18:30:45+02:00 Don't include names with empty subordinates in maps (#644) These are unecessary anyway and just blow up interface size - - - - - 69c2aac4 by Alex Biehl at 2017-06-29T19:54:49+02:00 Make per-argument docs for class methods work again (#648) * Make per-argument docs for class methods work again * Test case - - - - - c9448d54 by Bartosz Nitka at 2017-07-02T12:12:01+02:00 Fix haddock: internal error: links: UnhelpfulSpan (#561) * Fix haddock: internal error: links: UnhelpfulSpan This fixes haskell/haddock#554 for me. I believe this is another fall out of `wildcard-refactor`, like haskell/haddock#549. * Comment to clarify why we take the methods name location - - - - - d4f29eb7 by Alex Biehl at 2017-07-03T19:43:04+02:00 Document record fields when DuplicateRecordFields is enabled (#649) - - - - - 9d6e3423 by Yuji Yamamoto at 2017-07-03T22:37:58+02:00 Fix test failures on Windows (#564) * Ignore .stack-work * Fix for windows: use nul instead of /dev/null * Fix for windows: canonicalize line separator * Also normalize osx line endings - - - - - 7d81e8b3 by Yuji Yamamoto at 2017-07-04T16:13:12+02:00 Avoid errors on non UTF-8 Windows (#566) * Avoid errors on non UTF-8 Windows Problem ==== haddock exits with errors like below: `(1)` ``` haddock: internal error: <stderr>: hPutChar: invalid argument (invalid character) ``` `(2)` ``` haddock: internal error: Language\Haskell\HsColour\Anchors.hs: hGetContents: invalid argument (invalid byte sequence) ``` `(1)` is caused by printing [the "bullet" character](http://www.fileformat.info/info/unicode/char/2022/index.htm) onto stderr. For example, this warning contains it: ``` Language\Haskell\HsColour\ANSI.hs:62:10: warning: [-Wmissing-methods] • No explicit implementation for ‘toEnum’ • In the instance declaration for ‘Enum Highlight’ ``` `(2)` is caused when the input file of `readFile` contains some Unicode characters. In the case above, '⇒' is the cause. Environment ---- OS: Windows 10 haddock: 2.17.3 GHC: 8.0.1 Solution ==== Add `hSetEncoding handle utf8` to avoid the errors. Note ==== - I found the detailed causes by these changes for debugging: - https://github.com/haskell/haddock/commit/8f29edb6b02691c1cf4c479f6c6f3f922b35a55b - https://github.com/haskell/haddock/commit/1dd23bf2065a1e1f2c14d0f4abd847c906b4ecb4 - These errors happen even after executing `chcp 65001` on the console. According to the debug code, `hGetEncoding stderr` returns `CP932` regardless of the console encoding. * Avoid 'internal error: <stderr>: hPutChar: invalid argument (invalid character)' non UTF-8 Windows Better solution for 59411754a6db41d17820733c076e6a72bcdbd82b's (1) - - - - - eded67d2 by Alex Biehl at 2017-07-07T19:17:15+02:00 Remove redudant import warning (#651) - - - - - 05114757 by Alex Biehl at 2017-07-08T00:33:12+02:00 Avoid missing home module warning (#652) * Avoid missing home module warning * Update haddock-library.cabal - - - - - e9cfc902 by Bryn Edwards at 2017-07-17T07:51:20+02:00 Fix haskell/haddock#249 (#655) - - - - - eb02792b by Herbert Valerio Riedel at 2017-07-20T09:09:15+02:00 Fix compilation of lib:haddock-library w/ GHC < 8 - - - - - 9200bfbc by Alex Biehl at 2017-07-20T09:20:38+02:00 Prepare 2.18.1 release (#657) - - - - - 46ddd22c by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Tweak haddock-api.cabal for pending release - - - - - 85e33d29 by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Avoid trivial use of LambdaCase otherwise we can't test w/ e.g. GHC 7.4.2 - - - - - 3afb4bfe by Herbert Valerio Riedel at 2017-07-20T10:05:14+02:00 Refactor .cabal to use sub-lib for vendored lib A practical benefit is that we can control the build-depends and also avoid some recompilation between library and test-suite. - - - - - e56a552e by Herbert Valerio Riedel at 2017-07-20T10:17:48+02:00 haddock-api: add changelog pointing to haddock's changelog This addresses https://github.com/haskell/haddock/issues/638#issuecomment-309283297 - - - - - 2222ff0d by Herbert Valerio Riedel at 2017-07-20T10:19:56+02:00 Drop obsolete/misleading `stability: experimental` This .cabal property has long been considered obsolete - - - - - 9b882905 by Alex Biehl at 2017-07-20T11:25:54+02:00 Beef up haddock description (#658) * Beef up haddock description * Handle empty lines - - - - - bb60e95c by Herbert Valerio Riedel at 2017-07-20T12:08:53+02:00 Import @aisamanra's Haddock cheatsheet from https://github.com/aisamanra/haddock-cheatsheet - - - - - 0761e456 by Herbert Valerio Riedel at 2017-07-20T12:12:55+02:00 Add cheatsheet to haddock.cabal - - - - - 2ece0f0f by Herbert Valerio Riedel at 2017-07-20T12:18:38+02:00 Mention new-build in README - - - - - 947b7865 by Herbert Valerio Riedel at 2017-07-20T12:32:16+02:00 Update README Also improves markup and removes/fixes redundant/obsolete parts [skip ci] - - - - - 785e09ad by Alex Biehl at 2017-07-27T07:28:57+02:00 Bump haddock to 2.18.2, haddock-library to 1.4.5 - - - - - e3ff1ca3 by Alex Biehl at 2017-07-31T20:15:32+02:00 Move `DocMarkup` from haddock-api to haddock-library (#659) * Move `DocMarkup` from haddock-api to haddock-library * Move more markup related functions * Markup module * CHANGELOG - - - - - cda7c20c by Alex Biehl at 2017-07-31T20:35:49+02:00 Fixup haddock - - - - - 583b6812 by Alex Biehl at 2017-07-31T21:20:45+02:00 Changelog for haddock-library - - - - - bac6a0eb by Alex Biehl at 2017-07-31T21:50:24+02:00 Prepare haddock-library-1.4.5 release - - - - - 58ce6877 by Moritz Drexl at 2017-08-05T16:44:40+02:00 Fix renaming after instance signature specializing (#660) * rework rename * Add regression test for Bug 613 * update tests * update changelog - - - - - b8137ec8 by Tim Baumann at 2017-08-06T11:33:38+02:00 Fix: Generate pattern signatures for constructors exported as patterns (#663) * Fix pretty-printing of pattern signatures Pattern synonyms can have up to two contexts, both having a different semantic meaning: The first holds the constraints required to perform the matching, the second contains the constraints provided by a successful pattern match. When the first context is empty but the second is not it is necessary to render the first, empty context. * Generate pattern synonym signatures for ctors exported as patterns This fixes haskell/haddock#653. * Simplify extractPatternSyn It is not necessary to generate the simplest type signature since it will be simplified when pretty-printed. * Add changelog entries for PR haskell/haddock#663 * Fix extractPatternSyn error message - - - - - d037086b by Alex Biehl at 2017-08-06T12:43:25+02:00 Bump haddock-library - - - - - 99d7e792 by Alex Biehl at 2017-08-06T12:44:07+02:00 Bump haddock-library in haddock-api - - - - - 94802a5b by Alex Biehl at 2017-08-06T13:18:02+02:00 Provide --show-interface option to dump interfaces (#645) * WIP: Provide --show-interface option to dump interfaces Like ghcs own --show-iface this flag dumps a binary interface file to stdout in a human (and machine) readable fashion. Currently it uses json as output format. * Fill all the jsonNull stubs * Rework Bifunctor instance of DocH, update changelog and documentation * replace changelog, bring DocMarkupH doc back * Update CHANGES.md * Update CHANGES.md * Move Control.Arrow up It would result in unused import if the Bifunctor instance is not generated. - - - - - c662e476 by Ryan Scott at 2017-08-14T21:00:21-04:00 Adapt to haskell/haddock#14060 - - - - - b891eb73 by Alex Biehl at 2017-08-16T08:24:48+02:00 Bifoldable and Bitraversable for DocH and MetaDoc - - - - - 021bb56c by Alex Biehl at 2017-08-16T09:06:40+02:00 Refactoring: Make doc renaming monadic This allows us to later throw warnings if can't find an identifier - - - - - 39fbf022 by Alex Biehl at 2017-08-19T20:35:27+02:00 Hyperlinker: Avoid linear lookup in enrichToken (#669) * Make Span strict in Position * Hyperlinker: Use a proper map to enrich tokens - - - - - e13baedd by Alex Biehl at 2017-08-21T20:05:42+02:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 27dd6e87 by Alex Biehl at 2017-08-21T22:06:35+02:00 Drop Avails from export list - - - - - 86b247e2 by Alex Biehl at 2017-08-22T08:44:22+02:00 Bump ghc version for haddock-api tests - - - - - d4607ca0 by Alex Biehl at 2017-08-22T08:45:17+02:00 Revert "Drop Avails from export list" This reverts commit a850ba86d88a4fb9c0bd175453a2580e544e3def. - - - - - c9c54c30 by Alex Biehl at 2017-08-22T09:26:01+02:00 IntefaceFile version - - - - - a85b7c02 by Ben Gamari at 2017-08-22T09:29:52-04:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 34e976f5 by Ben Gamari at 2017-08-22T17:40:06+02:00 haddock: Add Documentation.Haddock.Markup to other-modules - - - - - 577abf06 by Ryan Scott at 2017-08-23T14:47:29-04:00 Update for haskell/haddock#14131 - - - - - da68fc55 by Florian Eggenhofer at 2017-08-27T18:21:56+02:00 Generate an index for package content search (#662) Generate an index for package content search - - - - - 39e62302 by Alex Biehl at 2017-08-27T18:50:16+02:00 Content search for haddock html doc - - - - - 91fd6fb2 by Alex Biehl at 2017-08-28T18:39:58+02:00 Fix tests for content search - - - - - b4a3798a by Alex Biehl at 2017-08-28T18:44:08+02:00 Add search button to #page-menu - - - - - 25a7ca65 by Alex Biehl at 2017-08-28T18:47:43+02:00 Load javascript below the fold - - - - - 8d323c1a by Alex Biehl at 2017-08-28T18:49:22+02:00 Accept tests - - - - - c5dac557 by Alex Biehl at 2017-08-28T19:14:55+02:00 Content search css - - - - - 89a5af57 by Paolo Veronelli at 2017-08-29T07:42:13+02:00 Removed `nowrap` for interface method sigs (#674) with nowrap the interfaces method sigs would expand at libitum - - - - - a505f6f7 by Alex Biehl at 2017-08-29T08:05:33+02:00 Include subordinates in content index - - - - - 4bb698c4 by Alexander Biehl at 2017-08-29T11:40:19+02:00 QuickNav: Make docbase configurable - - - - - c783bf44 by Alexander Biehl at 2017-08-29T11:48:36+02:00 QuickNav: Also use baseUrl for doc-index.json request - - - - - 47017510 by Alex Biehl at 2017-08-29T17:56:47+02:00 Fix test fallout (again) - - - - - 924fc318 by Alex Biehl at 2017-08-30T09:24:56+02:00 Write meta.json when generating html output (#676) - - - - - 717dea52 by Alex Biehl at 2017-09-01T09:20:34+02:00 Use relative URL when no docBaseUrl given - - - - - e5d85f3b by Alex Biehl at 2017-09-01T09:35:19+02:00 Add missing js files to data-files (#677) - - - - - 95b9231a by Alex Biehl at 2017-09-01T11:01:36+02:00 Rename "Search" tab to "Quick Jump" - - - - - da0ead0b by Alex Biehl at 2017-09-01T13:03:49+02:00 Make trigger link configurable (#678) QuickNav: Configurable show/hide trigger - - - - - de7da594 by Ben Gamari at 2017-09-05T06:49:55-04:00 Account for "Remember the AvailInfo for each IE" As of GHC commit f609374a55bdcf3b79f3a299104767aae2ffbf21 GHC retains the AvailInfo associated with each IE. @alexbiehl has a patch making proper use of this change, but this is just to keep things building. - - - - - b05cd3b3 by Ben Gamari at 2017-09-14T07:55:07-04:00 Bump upper bound on base - - - - - 79db899e by Herbert Valerio Riedel at 2017-09-21T23:27:52+02:00 Make compatible with Prelude.<> export in GHC 8.4/base-4.11 - - - - - 3405dd52 by Tim Baumann at 2017-09-23T22:02:01+02:00 Add compile step that bundles and compresses JS files (#684) * Add compile step that bundles and compresses JS files Also, manage dependencies on third-party JS libraries using NPM. * Compile JS from TypeScript * Enable 'noImplicitAny' in TypeScript * QuickJump: use JSX syntax * Generate source maps from TypeScript for easier debugging * TypeScript: more accurate type * Separate quick jump css file from ocean theme - - - - - df0b5742 by Alex Biehl at 2017-09-29T21:15:40+02:00 Bump base for haddock-library and haddock-test - - - - - 62b12ea0 by Merijn Verstraaten at 2017-10-04T16:03:13+02:00 Inhibit output of coverage information for hidden modules. (#687) * Inhibit output of coverage information for hidden modules. * Add changelog entry. - - - - - 8daf8bc1 by Alexander Biehl at 2017-10-05T11:27:05+02:00 Don't use subMap in attachInstances - - - - - ad75114e by Alexander Biehl at 2017-10-05T11:27:58+02:00 Revert "Don't use subMap in attachInstances" This reverts commit 3adf5bcb1a6c5326ab33dc77b4aa229a91d91ce9. - - - - - 7d4aa02f by Alex Biehl at 2017-10-08T15:32:28+02:00 Precise Haddock: Use Avails for export resolution (#688) * Use Avails for export resolution * Support reexported modules * Factor out availExportItem * Use avails for fullModuleExports * Don't use subMap in attachInstances * lookupDocs without subMap * Completely remove subMap * Only calculate unqualified modules when explicit export list is given * Refactor * Refine comment * return * Fix * Refactoring * Split avail if declaration is not exported itself * Move avail splitting - - - - - b9b4faa8 by Alex Biehl at 2017-10-08T19:38:21+02:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 43325295 by Alex Biehl at 2017-10-08T20:18:46+02:00 Fix merge fallout - - - - - c6423cc0 by Alex Biehl at 2017-10-08T20:36:12+02:00 Copy QuickJump files over - - - - - 1db587c3 by Tim Baumann at 2017-10-09T18:33:09+02:00 Use <details> element for collapsibles (#690) * Remove unnecessary call to 'collapseSection' The call is unnecessary since there is no corresponding toggle for hiding the section of orphan instances. * Use <details> for collapsibles This makes them work even when JS is disabled. Closes haskell/haddock#560. - - - - - 1b54c64b by Tim Baumann at 2017-10-10T09:50:59+02:00 Quick Jump: Show error when loading 'doc-index.json' failed (#691) - - - - - 910f716d by Veronika Romashkina at 2017-10-24T07:36:20+02:00 Fix tiny typo in docs (#693) - - - - - b21de7e5 by Ryan Scott at 2017-10-24T13:07:15+02:00 Overhaul Haddock's rendering of kind signatures (#681) * Overhaul Haddock's rendering of kind signatures * Strip off kind signatures when specializing As an added bonus, this lets us remove an ugly hack specifically for `(->)`. Yay! * Update due to 0390e4a0f61e37bd1dcc24a36d499e92f2561b67 * @alexbiehl's suggestions * Import injectiveVarsOfBinder from GHC - - - - - 6704405c by Ryan Scott at 2017-10-28T07:10:27+02:00 Fix Haddock rendering of kind-indexed data family instances (#694) - - - - - 470f6b9c by Alex Biehl at 2017-10-30T08:45:51+01:00 Add QuickJump version to meta.json (#696) - - - - - b89eccdf by Alex Biehl at 2017-10-30T10:15:49+01:00 Put Quickjump behind --quickjump flag (#697) - - - - - 3095fb58 by Alex Biehl at 2017-10-30T19:09:06+01:00 Add build command to package.json - - - - - f223fda9 by Alex Biehl at 2017-10-30T19:10:39+01:00 Decrease threshold for fuzzy matching - - - - - 80245dda by Edward Z. Yang at 2017-10-31T20:35:05+01:00 Supported reexported-modules via --reexport flag. Signed-off-by: Edward Z. Yang <ezyang at cs.stanford.edu> - - - - - 7e389742 by Alex Biehl at 2017-10-31T20:37:56+01:00 Correct missing title in changelog - - - - - 1a2a1c03 by Alex Biehl at 2017-10-31T20:59:07+01:00 Copy quickjump.css for nicer error messages - - - - - db234bb9 by Alex Biehl at 2017-10-31T21:31:18+01:00 Reexported modules: Report warnings if argument cannot be parsed or ... module cannot be found - - - - - eea8a205 by Carlo Hamalainen at 2017-10-31T21:43:14+01:00 More general type for nameCacheFromGhc. (#539) - - - - - 580eb42a by Alex Biehl at 2017-10-31T21:46:52+01:00 Remote tab - - - - - 0e599498 by Alex Biehl at 2017-10-31T21:48:55+01:00 Merge remote-tracking branch 'origin/master' into ghc-head - - - - - 7b8539bb by Alex Biehl at 2017-10-31T22:28:34+01:00 fullModuleContents traverses exports in declaration order - - - - - 0c91fbf2 by Alex Biehl at 2017-10-31T22:32:31+01:00 Remove excessive use of list comprehensions - - - - - f7356e02 by Alex Biehl at 2017-11-01T19:11:03+01:00 Make better use of AvailInfo - - - - - f3e512d5 by Alex Biehl at 2017-11-02T12:16:22+01:00 Always return documentation for exported subordinates ... event if they have no documentation (e.g. noDocForDecl) By using the information in the AvailInfo we don't need additional export checks. - - - - - 7cf58898 by Alan Zimmerman at 2017-11-07T08:28:03+02:00 Match changes for Trees that Grow in GHC - - - - - e5105a41 by Alan Zimmerman at 2017-11-08T17:21:58+02:00 Match Trees That Grow - - - - - 55178266 by Alan Zimmerman at 2017-11-11T22:20:31+02:00 Match Trees that Grow in GHC for HsExpr - - - - - 2082ab02 by Ryan Scott at 2017-11-14T15:27:03+01:00 Actually render infix type operators as infix (#703) * Actually render infix type operators as infix * Account for things like `(f :*: g) p`, too - - - - - c52ab7d0 by Alan Zimmerman at 2017-11-14T23:14:26+02:00 Clean up use of PlaceHolder, to match TTG - - - - - 81cc9851 by Moritz Angermann at 2017-11-20T07:52:49+01:00 Declare use of `Paths_haddock` module in other-modules (#705) This was detected by `-Wmissing-home-modules` - - - - - f9d27598 by Moritz Angermann at 2017-11-20T12:47:34+01:00 Drop Paths_haddock from ghc.mk (#707) With haskell/haddock#705 and haskell/haddock#706, the custom addition should not be necessary any more. # Conflicts: # ghc.mk - - - - - f34818dc by Moritz Angermann at 2017-11-20T12:47:59+01:00 Add autogen-modules (#706) > Packages using 'cabal-version: >= 1.25' and the autogenerated module Paths_* must include it also on the 'autogen-modules' field besides 'exposed-modules' and 'other-modules'. This specifies that the module does not come with the package and is generated on setup. Modules built with a custom Setup.hs script also go here to ensure that commands like sdist don't fail. # Conflicts: # haddock.cabal - - - - - bb43a0aa by Ben Gamari at 2017-11-21T15:50:12-05:00 Revert "Clean up use of PlaceHolder, to match TTG" This reverts commit 134a7bb054ea730b13c8629a76232d73e3ace049. - - - - - af9ebb2b by Ben Gamari at 2017-11-21T15:50:14-05:00 Revert "Match Trees that Grow in GHC for HsExpr" This reverts commit 9f054dc365379c66668de6719840918190ae6e44. - - - - - 5d35c3af by Ben Gamari at 2017-11-21T15:50:15-05:00 Revert "Match Trees That Grow" This reverts commit 73a26af844ac50b8bec39de11d64452a6286b00c. - - - - - 99a8e43b by Ben Gamari at 2017-11-21T16:36:06-05:00 Revert "Match changes for Trees that Grow in GHC" This reverts commit 01eeeb048acd2dd05ff6471ae148a97cf0720547. - - - - - c4d650c2 by Ben Gamari at 2017-12-04T15:06:07-05:00 Bump GHC version - - - - - 027b2274 by Ben Gamari at 2017-12-04T17:06:31-05:00 Bump GHC bound to 8.4.* - - - - - 58eaf755 by Alex Biehl at 2017-12-06T15:44:24+01:00 Update changelog - - - - - d68f5584 by Simon Peyton Jones at 2017-12-07T14:39:56+00:00 Track changes to follow Trac haskell/haddock#14529 This tracks the refactoring of HsDecl.ConDecl. - - - - - dc519d6b by Alec Theriault at 2018-01-06T08:20:43-08:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - 8285118c by Alec Theriault at 2018-01-13T12:12:37+01:00 Constructor and pattern synonym argument docs (#709) * Support Haddocks on constructor arguments This is in conjunction with https://phabricator.haskell.org/D4094. Adds support for rendering Haddock's on (non-record) constructor arguments, both for regular and GADT constructors. * Support haddocks on pattern synonym arguments It appears that GHC already parsed these - we just weren't using them. In the process of doing this, I tried to deduplicate some code around handling patterns. * Update the markup guide Add some information about the new support for commenting constructor arguments, and mention pattern synonyms and GADT-style constructors. * Overhaul LaTeX support for data/pattern decls This includes at least * fixing several bugs that resulted in invalid LaTeX * fixing GADT data declaration headers * overhaul handling of record fields * overhaul handling of GADT constructors * overhaul handling of bundled patterns * add support for constructor argument docs * Support GADT record constructors This means changes what existing HTML docs look like. As for LaTeX, looks like GADT records were never even supported. Now they are. * Clean up code/comments Made code/comments consistent between the LaTeX and XHTML backend when possible. * Update changelog * Patch post-rebase regressions * Another post-rebase change We want return values to be documentable on record GADT constructors. - - - - - ca4fabb4 by Alec Theriault at 2018-01-15T17:12:18-08:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 4c472fea by Ryan Scott at 2018-01-19T10:44:02+01:00 Fix haskell/haddock#732 (#733) - - - - - bff14dbd by Alex Biehl at 2018-01-19T15:33:30+01:00 extractDecl: Extract associated types correctly (#736) - - - - - a2a94a73 by Alex Biehl at 2018-01-19T15:34:40+01:00 extractDecl: Extract associated types correctly (#736) - - - - - 26df93dc by Alex Biehl at 2018-01-20T10:18:22+01:00 haddock-api: bump ghc to ^>= 8.4 - - - - - f65aeb1d by Alex Biehl at 2018-01-20T19:18:20+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - 0e721b97 by Alex Biehl at 2018-01-20T19:20:19+01:00 Fix duplicate declarations and TypeFamilies specifics - - - - - cb6234f6 by Ben Gamari at 2018-01-26T13:40:55-05:00 Merge remote-tracking branch 'harpocrates/fix/missing-orphan-instances' into ghc-head - - - - - 0fc28554 by Alec Theriault at 2018-02-01T14:58:18+01:00 Pass to GHC visible modules for instance filtering The GHC-side `getNameToInstancesIndex` filters out incorrectly some instances because it is not aware of what modules are visible. On the Haddock side, we need to pass in the modules we are processing. On the GHC side, we need to check against _those_ modules when checking if an instance is visible. - - - - - b9123772 by Alec Theriault at 2018-02-01T14:58:18+01:00 Update the GblRdrEnv when processing modules Without a complete environment, we will miss some instances that were encountered during typechecking. - - - - - 0c12e274 by Ryan Scott at 2018-02-01T14:58:18+01:00 Fix haskell/haddock#548 by rendering datatype kinds more carefully (#702) - - - - - 8876d20b by Alec Theriault at 2018-02-01T14:58:18+01:00 Use the GHC lexer for the Hyperlinker backend (#714) * Start changing to use GHC lexer * better cpp * Change SrcSpan to RealSrcSpan * Remove error * Try to stop too many open files * wip * wip * Revert "wip" This reverts commit b605510a195f26315e3d8ca90e6d95a6737553e1. Conflicts: haddock-api/haddock-api.cabal haddock-api/src/Haddock/Interface.hs * Remove pointless 'caching' * Use dlist rather than lists when finding vars * Use a map rather than list * Delete bogus comment * Rebase followup Things now run using the GHC lexer. There are still - stray debug statements - unnecessary changes w.r.t. master * Cleaned up differences w.r.t. current Haddock HEAD Things are looking good. quasiquotes in particular look beautiful: the TH ones (with Haskell source inside) colour/link their contents too! Haven't yet begun to check for possible performance problems. * Support CPP and top-level pragmas The support for these is hackier - but no more hacky than the existing support. * Tests pass, CPP is better recognized The tests were in some cases altered: I consider the new output to be more correct than the old one.... * Fix shrinking of source without tabs in test * Replace 'Position'/'Span' with GHC counterparts Replaces 'Position' -> 'GHC.RealSrcLoc' and 'Span' -> 'GHC.RealSrcSpan'. * Nits * Forgot entry in .cabal * Update changelog - - - - - 95c6a771 by Alec Theriault at 2018-02-01T14:58:18+01:00 Clickable anchors for headings (#716) See haskell/haddock#579. This just adds an <a> tag around the heading, pointing to the heading itself. - - - - - 21463d28 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump: Matches on function names weight more than matches in ... module names. - - - - - 8023af39 by Alex Biehl at 2018-02-01T14:58:18+01:00 Treat escaped \] better in definition lists (#717) This fixes haskell/haddock#546. - - - - - e4866dc1 by Alex Biehl at 2018-02-01T14:58:18+01:00 Remove scanner, takeWhile1_ already takes care of escaping - - - - - 9bcaa49d by Alex Biehl at 2018-02-01T14:58:18+01:00 Take until line feed - - - - - 01d2af93 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Add simple framework for running parser fixtures (#668) * Add simple framework for running parser fixtures * Compatible with tree-diff-0.0.0.1 * Use parseParas to parse fixtures This allows to test all syntactic constructs available in haddock markup. - - - - - 31128417 by Alec Theriault at 2018-02-01T14:58:18+01:00 Patch flaky parser test (#720) * Patch flaky parser test This test was a great idea, but it doesn't port over too well to using the GHC lexer. GHC rewrites its input a bit - nothing surprising, but we need to guard against those cases for the test. * Change instance head * Change use site - - - - - 9704f214 by Herbert Valerio Riedel at 2018-02-01T14:58:18+01:00 Include secondary LICENSE file in source dist - - - - - 51f25074 by Oleg Grenrus at 2018-02-01T14:58:18+01:00 Grid Tables (#718) * Add table examples * Add table types and adopt simple parser Simple parser is done by Giovanni Cappellotto (@potomak) in https://github.com/haskell/haddock/pull/577 It seems to support single fine full tables, so far from full RST-grid tables, but it's good start. Table type support row- and colspans, but obviously parser is lacking. Still TODO: - Latex backend. Should we use multirow package https://ctan.org/pkg/multirow?lang=en? - Hoogle backend: ? * Implement grid-tables * Refactor table parser * Add two ill-examples * Update CHANGES.md * Basic documentation for tables * Fix documentation example - - - - - 670d6200 by Alex Biehl at 2018-02-01T14:58:18+01:00 Add grid table example to cheatsheet (pdf and svg need to be regenerated thought) - - - - - 4262dec9 by Alec Theriault at 2018-02-01T14:58:18+01:00 Fix infinite loop when specializing instance heads (#723) * Fix infinite loop when specializing instance heads The bug can only be triggered from TH, hence why it went un-noticed for so long. * Add test for haskell/haddock#679 and haskell/haddock#710 - - - - - 67ecd803 by Alec Theriault at 2018-02-01T14:58:18+01:00 Filter RTS arguments from 'ghc-options' arguments (#725) This fixes haskell/haddock#666. - - - - - 7db26992 by Alex Biehl at 2018-02-01T14:58:18+01:00 Quickjump Scrollable overlay - - - - - da9ff634 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Hyperlinker: Adjust parser to new PFailed constructor - - - - - 7b7cf8cb by Alexander Biehl at 2018-02-01T14:58:18+01:00 Specialize: Add missing IdP annotations - - - - - 78cd7231 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Convert: Correct pass type - - - - - a2d0f590 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Warning free compilation - - - - - cd861cf3 by Alexander Biehl at 2018-02-01T14:58:18+01:00 hadock-2.19.0 / haddock-api-2.19.0 / haddock-library-1.5.0 - - - - - c6651b72 by Alexander Biehl at 2018-02-01T14:58:18+01:00 Adjust changelogs - - - - - 1e93da0b by Alexander Biehl at 2018-02-01T14:58:18+01:00 haddock-library: Info about breaking changes - - - - - f9b11db8 by Alec Theriault at 2018-02-02T12:36:02+01:00 Properly color pragma contents in hyperlinker The hyperlinker backend now classifies the content of pragmas as 'TkPragma'. That means that in something like '{-# INLINE foo #-}', 'foo' still gets classified as a pragma token. - - - - - c40b0043 by Alec Theriault at 2018-02-02T12:36:02+01:00 Support the new 'ITcolumn_prag' token - - - - - 4a2a4d39 by Alex Biehl at 2018-02-03T12:11:55+01:00 QuickJump: Mitigate encoding problems on Windows - - - - - bb34503a by Alex Biehl at 2018-02-04T18:39:31+01:00 Use withBinaryFile - - - - - 637605bf by Herbert Valerio Riedel at 2018-02-05T09:48:32+01:00 Try GHC 8.4.1 for Travis CI job - - - - - 7abb67e4 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 try harder to build w/ GHC 8.4.1 - - - - - 8255cc98 by Herbert Valerio Riedel at 2018-02-05T10:05:42+01:00 Add `SPDX-License-Identifier` as alised for "license" module header tokens C.f. SPDX 2.1 - Appendix V https://spdx.org/spdx-specification-21-web-version#h.twlc0ztnng3b The tag should appear on its own line in the source file, generally as part of a comment. SPDX-License-Identifier: <SPDX License Expression> Cherry-picked from haskell/haddock#743 - - - - - 267cd23d by Herbert Valerio Riedel at 2018-02-05T10:24:34+01:00 Make test-suite SMP compatible - - - - - 95d4bf40 by Alec Theriault at 2018-02-05T22:01:04+01:00 Hyperlink pattern synonyms and 'module' imports (#744) Links to pattern synonyms are now generated, as well as links from modules in import lists. Fixes haskell/haddock#731. - - - - - 67838dcd by Alec Theriault at 2018-02-06T08:23:36+01:00 Don't warn about missing '~' (#746) This manually filters out '~' from the list of things to warn about. It truly makes no sense to warn on this since '~' has nothing it could link to - it is magical. This fixes haskell/haddock#532. - - - - - ab6c3f9f by Alec Theriault at 2018-02-06T08:24:47+01:00 Don't barf on 'HsSpliceTy' (#745) This handles 'HsSpliceTy's by replacing them with what they expand to. IIUC everything that is happening, 'renameHsSpliceTy' should not be able to fail for the inputs we feed it from GHC. This fixes haskell/haddock#574. - - - - - 92bf95ad by Alex Biehl at 2018-02-06T08:28:23+01:00 Rename: renameHsSpliceTy ttg - - - - - 3130b1e1 by Alex Biehl at 2018-02-06T09:02:14+01:00 Expand SigDs - - - - - c72adae5 by Alex Biehl at 2018-02-06T09:20:51+01:00 fullModuleContents: support named docs - - - - - de2e4dbf by Alex Biehl at 2018-02-06T13:56:17+01:00 Hyperlinker: Also link pattern synonym arguments - - - - - b7c98237 by Alex Biehl at 2018-02-09T18:44:23+01:00 Expand SigD in a better place In https://github.com/haskell/haddock/issues/287 we found that haddock-2.19.0 would miss documentation on class methods with multiples names. This patch uses expandSigDecls in a more sensible place. - - - - - 8f598b27 by Alec Theriault at 2018-02-11T12:29:56+01:00 Add module tooltips to linked identifiers (#753) No more clicking to figure out whether your bytestring is strict or lazy! - - - - - d812e65d by Alec Theriault at 2018-02-11T12:31:44+01:00 Add 'show' option to complement 'hide' (#752) * Add 'show' option to complement 'hide' The behaviour is for flags passed in the command line to override flags in file headers. In the command line, later flags override earlier ones. Fixes haskell/haddock#751 and haskell/haddock#266. * Add a '--show-all' option - - - - - 6676cecb by Alex Biehl at 2018-02-18T11:07:15-05:00 QuickJump: Mitigate encoding problems on Windows (cherry picked from commit 86292c54bfee2343aee84559ec01f1fc68f52231) - - - - - e753dd88 by Alex Biehl at 2018-02-18T17:59:54+01:00 Use withBinaryFile - - - - - 724dc881 by Tamar Christina at 2018-02-19T05:34:49+01:00 Haddock: support splitted include paths. (#689) - - - - - 9b6d6f50 by Alex Biehl at 2018-02-19T05:57:02+01:00 Teach the HTML backend how to render methods with multiple names - - - - - a74aa754 by Alexander Biehl at 2018-02-19T10:04:34+01:00 Hoogle/Latex: Remove use of partial function - - - - - 66d8bb0e by Alec Theriault at 2018-02-25T16:04:01+01:00 Fix file handle leak (#763) (#764) Brought back some mistakenly deleted code for handling encoding and eager reading of files from e0ada1743cb722d2f82498a95b201f3ffb303137. - - - - - bb92d03d by Alex Biehl at 2018-03-02T14:21:23+01:00 Enable running test suite with stock haddock and ghc using ``` $ cabal new-run -- html-test --haddock-path=$(which haddock) --ghc-path=$(which ghc) ``` - - - - - dddb3cb2 by Alex Biehl at 2018-03-02T15:43:21+01:00 Make testsuite work with haddock-1.19.0 release (#766) - - - - - f38636ed by Alec Theriault at 2018-03-02T15:48:36+01:00 Support unicode operators, proper modules Unicode operators are a pretty big thing in Haskell, so supporting linking them seems like it outweighs the cost of the extra machinery to force Attoparsec to look for unicode. Fixes haskell/haddock#458. - - - - - 09d89f7c by Alec Theriault at 2018-03-02T15:48:43+01:00 Remove bang pattern - - - - - d150a687 by Alex Biehl at 2018-03-02T15:48:48+01:00 fix test - - - - - d6fd71a5 by Alex Biehl at 2018-03-02T16:22:38+01:00 haddock-test: Be more explicit which packages to pass We now pass `-hide-all-packages` to haddock when invoking the testsuite. This ensures we don't accidentally pick up any dependencies up through ghc.env files. - - - - - 0932c78c by Alex Biehl at 2018-03-02T17:50:38+01:00 Revert "fix test" This reverts commit 1ac2f9569242f6cb074ba6e577285a4c33ae1197. - - - - - 52516029 by Alex Biehl at 2018-03-02T18:16:50+01:00 Fix Bug548 for real - - - - - 89df9eb5 by Alex Biehl at 2018-03-05T18:28:19+01:00 Hyperlinker: Links for TyOps, class methods and associated types - - - - - d019a4cb by Ryan Scott at 2018-03-06T13:43:56-05:00 Updates for haskell/haddock#13324 - - - - - 6d5a42ce by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump haddock-2.19.0.1, haddock-api-2.19.0.1, haddock-library-1.5.0.1 - - - - - c0e6f380 by Alex Biehl at 2018-03-10T18:25:57+01:00 Update changelogs for haddock-2.19.0.1 and haddock-library-1.5.0.1 - - - - - 500da489 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Update to QC 2.11 - - - - - ce8362e9 by Herbert Valerio Riedel at 2018-03-10T18:25:57+01:00 Restore backward-compat with base-4.5 through base-4.8 - - - - - baae4435 by Alex Biehl at 2018-03-10T18:25:57+01:00 Bump lower bound for haddock-library - - - - - 10b7a73e by Alex Biehl at 2018-03-10T18:25:57+01:00 Haddock: Straighten out base bound - - - - - a6096f7b by Alex Biehl at 2018-03-13T08:45:06+01:00 extractDecl: Extract constructor patterns from data family instances (#776) * extractDecl: Allow extraction of data family instance constructors * extractDecl: extract data family instance constructors - - - - - ba4a0744 by Simon Jakobi at 2018-03-14T08:26:42+01:00 Readme: Update GHC version (#778) - - - - - 8de157d4 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for definition lists - - - - - 425b46f9 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for links - - - - - d53945d8 by Simon Jakobi at 2018-03-14T20:39:29+01:00 Add fixture test for inline links - - - - - f1dc7c99 by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Slightly unmangle output - - - - - 0879d31c by Simon Jakobi at 2018-03-14T20:39:29+01:00 fixtures: Prevent stdout buffering - - - - - 1f9e5f1b by Simon Jakobi at 2018-03-14T20:39:29+01:00 haddock-library.cabal: Clean up GHC options - - - - - 066b891a by Simon Jakobi at 2018-03-14T20:39:29+01:00 Make a proper definition for the <link> parser - - - - - 573d6ba7 by Alec Theriault at 2018-03-21T09:16:57+01:00 Show where instances are defined (#748) * Indicate source module of instances Above instance, we now also display a link to the module where the instance was defined. This is sometimes helpful in figuring out what to import. * Source module for type/data families too * Remove parens * Accept tests - - - - - 99b5d28b by Alex Biehl at 2018-03-21T09:20:36+01:00 Prepare changelog for next release - - - - - 482d3a93 by Alex Biehl at 2018-03-23T15:57:36+01:00 Useful cost centres, timers and allocation counters (#785) * Add some useful cost-centres for profiling * Add withTiming for each haddock phase Invoking haddock with `--optghc=-ddump-timings` now shows the amount of time spent and the number of allocated bytes for each phase. - - - - - 773b41bb by Alec Theriault at 2018-03-27T08:35:59+02:00 @since includes package name (#749) * Metadoc stores a package name This means that '@since' annotations can be package aware. * Get the package name the right way This should extract the package name for `@since` annotations the right way. I had to move `modulePackageInfo` around to do this and, in the process, I took the liberty to update it. Since it appears that finding the package name is something that can fail, I added a warning for this case. * Silence warnings * Hide package for local 'since' annotations As discussed, this is still the usual case (and we should avoid being noisy for it). Although this commit is large, it is basically only about threading a 'Maybe Package' from 'Haddock.render' all the way to 'Haddock.Backends.Xhtml.DocMarkup.renderMeta'. * Bump binary interface version * Add a '--since-qual' option This controls when to qualify since annotations with the package they come from. The default is always, but I've left an 'external' variant where only those annotations coming from outside of the current package are qualified. * Make ParserSpec work * Make Fixtures work * Use package name even if package version is not available The @since stuff needs only the package name passed in, so it makes sense to not be forced to pass in a version too. - - - - - e42c57bc by Alex Biehl at 2018-03-27T08:42:50+02:00 haddock-2.19.1, haddock-api-2.19.1, haddock-library-1.6.0 - - - - - 8373a529 by Alex Biehl at 2018-03-28T10:17:11+02:00 Bump haddock and haddock-api to 2.20.0 - - - - - 5038eddd by Jack Henahan at 2018-04-03T13:28:12+02:00 Clear search string on hide for haskell/haddock#781 (#789) - - - - - 920ca1eb by Alex Biehl at 2018-04-03T16:35:50+02:00 Travis: Build with ghc-8.4.2 (#793) - - - - - a232f0eb by Alan Zimmerman at 2018-04-07T14:14:32+02:00 Match changes in GHC for D4199 Removing HasSourceText and SourceTextX classes. - - - - - ab85060b by Alan Zimmerman at 2018-04-09T21:20:24+02:00 Match GHC changes for TTG - - - - - 739302b6 by Alan Zimmerman at 2018-04-13T13:31:44+02:00 Match GHC for TTG implemented on HsBinds, D4581 - - - - - 2f56d3cb by Ryan Scott at 2018-04-19T11:42:58-04:00 Bump upper bound on base to < 4.13 See https://ghc.haskell.org/trac/ghc/ticket/15018. - - - - - a49df92a by Alex Biehl at 2018-04-20T07:31:44+02:00 Don't treat fixity signatures like declarations - - - - - d02c103b by Ryan Scott at 2018-04-24T11:20:11-04:00 Add regression test for haskell/haddock#413 Fixes haskell/haddock#413. - - - - - c7577f52 by Ryan Scott at 2018-04-24T13:51:06-07:00 Improve the Hoogle backend's treatment of type families (#808) Fixes parts 1 and 2 of haskell/haddock#806. - - - - - d88f85b1 by Alec Theriault at 2018-04-25T11:24:07-07:00 Replace 'attoparsec' with 'parsec' (#799) * Remove attoparsec with parsec and start fixing failed parses * Make tests pass * Fix encoding issues The Haddock parser no longer needs to worry about bytestrings. All the internal parsing work in haddock-library happens over 'Text'. * Remove attoparsec vendor * Fix stuff broken in 'attoparsec' -> 'parsec' * hyperlinks * codeblocks * examples Pretty much all issues are due to attoparsec's backtracking failure behaviour vs. parsec's non-backtracking failure behaviour. * Fix small TODOs * Missing quote + Haddocks * Better handle spaces before/after paragraphs * Address review comments - - - - - fc25e2fe by Alan Zimmerman at 2018-04-27T15:36:53+02:00 Match changes in GHC for TTG - - - - - 06175f91 by Herbert Valerio Riedel at 2018-05-01T18:11:09+02:00 Merge branch 'ghc-head' with 'ghc-8.4' - - - - - 879caaa8 by Alec Theriault at 2018-05-07T18:53:15-07:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 3e0120cb by Simon Jakobi at 2018-05-07T19:00:18-07:00 Add docs for some DocH constructors (#814) - - - - - 0a32c6db by Alec Theriault at 2018-05-08T02:15:45-07:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 8816e783 by Simon Jakobi at 2018-05-08T10:48:11-07:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - ad60366f by Ryan Scott at 2018-05-10T11:19:47-04:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - 03b7cc3b by Ryan Scott at 2018-05-10T11:24:38-04:00 Wibbles - - - - - b03dd563 by Chaitanya Koparkar at 2018-05-10T11:44:58-04:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 9f298a40 by Ben Gamari at 2018-05-13T17:36:04-04:00 Account for refactoring of LitString - - - - - ea3dabe7 by Ryan Scott at 2018-05-16T09:21:43-04:00 Merge pull request haskell/haddock#826 from haskell/T825 Remove Hoogle backend hack that butchers infix datatype names - - - - - 0d234f7c by Alec Theriault at 2018-05-23T11:29:05+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 15fc9712 by Simon Jakobi at 2018-05-31T04:17:47+02:00 Adjust to new HsDocString internals - - - - - 6f1e19a8 by Ben Gamari at 2018-06-02T16:18:58-04:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 0d0355d9 by Ryan Scott at 2018-06-04T21:26:59-04:00 DerivingVia changes - - - - - 0d93475a by Simon Jakobi at 2018-06-05T19:47:05+02:00 Bump a few dependency bounds (#845) - - - - - 5cbef804 by Alec Theriault at 2018-06-05T19:47:16+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 9094c56f by Alec Theriault at 2018-06-05T22:53:25+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 70188719 by Simon Jakobi at 2018-06-08T22:20:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - 495cd1fc by Chaitanya Koparkar at 2018-06-13T23:01:34+02:00 Use the response file utilities defined in `base` (#821) Summary: The response file related modules were recently copied from `haddock` into `base`. This patch removes them from `haddock`. GHC Trac Issues: haskell/haddock#13896 - - - - - 81088732 by Ben Gamari at 2018-06-13T23:01:34+02:00 Account for refactoring of LitString - - - - - 7baf6587 by Simon Jakobi at 2018-06-13T23:05:08+02:00 Adjust to new HsDocString internals - - - - - bb61464d by Ben Gamari at 2018-06-13T23:05:22+02:00 Remove ParallelArrays and Data Parallel Haskell - - - - - 5d8cb87f by Ryan Scott at 2018-06-13T23:39:30+02:00 DerivingVia changes - - - - - 73d373a3 by Alec Theriault at 2018-06-13T23:39:30+02:00 Extract docs from strict/unpacked constructor args (#839) This fixes haskell/haddock#836. - - - - - 4865e254 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Remove `ITtildehsh` token - - - - - b867db54 by Alec Theriault at 2018-06-13T23:39:30+02:00 Filter out CRLFs in hyperlinker backend (#813) This prevents spurious lines from appearing in the final output. - - - - - 9598e392 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Add docs for some DocH constructors (#814) - - - - - 8a59035b by Alec Theriault at 2018-06-13T23:39:30+02:00 Remove 'TokenGroup' from Hyperlinker (#818) Since the hyperlinker backend now relies on the GHC tokenizer, something like 'Bar.Baz.foo' already gets bunched together into one token (as opposed to being spread across 'Bar', '.', 'Baz', '.', and 'foo'). - - - - - 29350fc8 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about out of scope identifiers. (#819) - - - - - 2590bbd9 by Ryan Scott at 2018-06-13T23:39:30+02:00 Remove Hoogle backend hack that butchers infix datatype names - - - - - a9939fdc by Ryan Scott at 2018-06-13T23:39:30+02:00 Wibbles - - - - - a22f7df4 by Alec Theriault at 2018-06-13T23:39:30+02:00 Use `ClassOpSig` instead of `TypeSig` for class methods (#835) * Fix minimal pragma handling Class declarations contain 'ClassOpSig' not 'Typesig'. This should fix haskell/haddock#834. * Accept html-test output - - - - - 8741015d by Simon Jakobi at 2018-06-13T23:39:30+02:00 Bump a few dependency bounds (#845) - - - - - 4791e1cc by Alec Theriault at 2018-06-13T23:39:30+02:00 Improve hyperlinker's 'spanToNewline' (#846) 'spanToNewline' is used to help break apart the source into lines which can then be partioned into CPP and non-CPP chunks. It is important that 'spanToNewline' not break apart tokens, so it needs to properly handle things like * block comments, possibly nested * string literals, possibly multi-line * CPP macros, possibly multi-line String literals in particular were not being properly handled. The fix is to to fall back in 'Text.Read.lex' to help lex things that are not comments. Fixes haskell/haddock#837. - - - - - 311d3216 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Renamer: Warn about ambiguous identifiers (#831) * Renamer: Warn about ambiguous identifiers Example: Warning: 'elem' is ambiguous. It is defined * in ‘Data.Foldable’ * at /home/simon/tmp/hdk/src/Lib.hs:7:1 You may be able to disambiguate the identifier by qualifying it or by hiding some imports. Defaulting to 'elem' defined at /home/simon/tmp/hdk/src/Lib.hs:7:1 Fixes haskell/haddock#830. * Deduplicate warnings Fixes haskell/haddock#832. - - - - - d0577817 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Complete FixitySig and FamilyDecl pattern matches - - - - - 055b3aa7 by Simon Jakobi at 2018-06-13T23:39:30+02:00 Fix redundant import warnings - - - - - f9ce19b1 by Simon Jakobi at 2018-06-13T23:49:52+02:00 html-test: Accept output - - - - - 04604ea7 by Simon Jakobi at 2018-06-13T23:54:37+02:00 Bump bounds on Cabal - - - - - 0713b692 by Simon Jakobi at 2018-06-14T00:00:12+02:00 Merge branch 'ghc-head' into ghc-head-update-3 - - - - - c6a56bfd by Simon Jakobi at 2018-06-14T02:33:27+02:00 Bump ghc bound for haddock-api spec test-suite - - - - - 119d04b2 by Simon Jakobi at 2018-06-14T12:37:48+02:00 Travis: `--allow-newer` for all packages - - - - - 0e876e2c by Alex Biehl at 2018-06-14T15:28:52+02:00 Merge pull request haskell/haddock#857 from sjakobi/ghc-head-update-3 Update ghc-head - - - - - 5be46454 by Alec Theriault at 2018-06-14T21:42:45+02:00 Improved handling of interfaces in 'haddock-test' (#851) This should now work with an inplace GHC where (for instance) HTML directories may not be properly recorded in the package DB. - - - - - 96ab1387 by Vladislav Zavialov at 2018-06-14T17:06:21-04:00 Handle -XStarIsType - - - - - e518f8c4 by Ben Gamari at 2018-06-14T17:48:00-04:00 Revert unintentional reversion of fix of haskell/haddock#548 - - - - - 01b9f96d by Alan Zimmerman at 2018-06-19T11:52:22+02:00 Match changes in GHC for haskell/haddock#14259 - - - - - 7f8c8298 by Ben Gamari at 2018-06-19T18:14:27-04:00 Bump GHC version to 8.6 - - - - - 11c6b5d2 by Ryan Scott at 2018-06-19T23:17:31-04:00 Remove HsEqTy and XEqTy - - - - - b33347c2 by Herbert Valerio Riedel at 2018-06-20T23:14:52+02:00 Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. - - - - - f0d2460e by Herbert Valerio Riedel at 2018-06-20T23:28:46+02:00 Update Travis CI job - - - - - ef239223 by Herbert Valerio Riedel at 2018-06-20T23:32:41+02:00 Drop GHC HEAD from CI and update GHC to 8.4.3 It's a waste of resource to even try to build this branch w/ ghc-head; so let's not do that... - - - - - 41c4a9fa by Ben Gamari at 2018-06-20T18:26:20-04:00 Bump GHC version to 8.7 - - - - - 8be593dc by Herbert Valerio Riedel at 2018-06-21T22:32:15+02:00 Update CI job to use GHC 8.7.* - - - - - b91d334a by Simon Jakobi at 2018-06-30T13:41:38+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section - - - - - f707d848 by Alec Theriault at 2018-07-05T10:43:35-04:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. - - - - - a6d2b8dc by Alec Theriault at 2018-07-06T10:06:32-04:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case - - - - - 13819f71 by Alan Zimmerman at 2018-07-15T19:33:51+02:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 - - - - - c346aa78 by Simon Jakobi at 2018-07-19T12:29:32+02:00 haddock-library: Bump bounds for containers - - - - - 722e733c by Simon Jakobi at 2018-07-19T13:36:45+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] - - - - - f0bd83fd by Alec Theriault at 2018-07-19T14:39:57+02:00 Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. - - - - - 394053a8 by Simon Jakobi at 2018-07-19T14:58:07+02:00 haddock-library: Bump bounds for containers - - - - - 1bda11a2 by Alec Theriault at 2018-07-20T09:04:03+02:00 Add HEAD.hackage overlay (#887) * Add HEAD.hackage overlay * Add HCPKG variable - - - - - c7b4ab45 by Alec Theriault at 2018-07-20T12:01:16+02:00 Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test - - - - - c05d32ad by Alec Theriault at 2018-07-20T12:01:49+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output - - - - - 24b39ee4 by Alec Theriault at 2018-07-20T12:02:16+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. - - - - - cb9d2099 by Simon Jakobi at 2018-07-20T13:39:29+02:00 README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) - - - - - 133f24f5 by Alec Theriault at 2018-07-20T13:39:29+02:00 Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) - - - - - 11024149 by Alec Theriault at 2018-07-20T13:39:29+02:00 Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) - - - - - de0c139e by Simon Jakobi at 2018-07-20T13:39:29+02:00 tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) - - - - - 6435e952 by Alec Theriault at 2018-07-20T13:39:29+02:00 Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) - - - - - 1461af39 by Alec Theriault at 2018-07-20T13:39:29+02:00 Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) - - - - - 69d3bde1 by Alec Theriault at 2018-07-20T13:49:47+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) - - - - - 6a5c73c7 by Alec Theriault at 2018-07-20T13:50:00+02:00 Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests - - - - - 92ca94c6 by Alec Theriault at 2018-07-20T13:55:36+02:00 Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) - - - - - 981bc7fa by Simon Jakobi at 2018-07-20T15:06:06+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers - - - - - 27e7c0c5 by Simon Jakobi at 2018-07-20T15:09:05+02:00 Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 49e1a415 by Simon Jakobi at 2018-07-20T16:02:02+02:00 Update the ghc-8.6 branch (#889) * Revert "Bump GHC version to 8.6" This was applied to the wrong branch; there's now a `ghc-8.6` branch; ghc-head is always supposed to point to GHC HEAD, i.e. an odd major version. The next version bump to `ghc-head` is supposed to go from e.g. 8.5 to 8.7 This reverts commit 5e3cf5d8868323079ff5494a8225b0467404a5d1. * README updates (#856) * README: Remove mentions of master branch * README: Add instructions for using html-test * README: Change command to run _all_ the testsuites * README: Add project overview section (cherry picked from commit 61d6f935da97eb96685f07bf385102c2dbc2a33c) * Export more fixities for Hoogle (#871) This exports fixities for more things, including class methods and type-level operators. (cherry picked from commit 88316b972e3d47197b1019111bae0f7f87275fce) * Avoid line breaks due to line length in Hoogle (#868) * Avoid line breaks due to line length in Hoogle Hoogle operates in a line-oriented fashion, so we should avoid ever breaking due to long lines. One way of doing this non-intrusively is to modify the 'DynFlags' that are threaded through the 'Hoogle' module (note this is anyways only passed through for use in the various 'showSDoc' functions). * Amend test case (cherry picked from commit 657b1b3d519545f8d4ca048c06210d6cbf0f0da0) * tyThingToLHsDecl: Preserve type synonyms that contain a forall (#880) * tyThingToLHsDecls: Preserve type synonyms that contain a forall Fixes haskell/haddock#879. * Add Note [Invariant: Never expand type synonyms] * Clarify Note [Invariant: Never expand type synonyms] (cherry picked from commit c3eb3f0581f69e816f9453b1747a9f2a3ba02bb9) * Fix HEAD html-test (#860) * Update tests for 'StarIsType' * Accept tests * Revert "Update tests for 'StarIsType'" This reverts commit 7f0c01383bbba6dc5af554ee82988d2cf44e407a. * Refactor handling of parens in types (#874) * Fix type parenthesization in Hoogle backend Ported the logic in the HTML and LaTeX backends for adding in parens into something top-level in 'GhcUtil'. Calling that from the Hoogle backend fixes haskell/haddock#873. * Remove parenthesizing logic from LaTeX and XHTML backends Now, the only times that parenthesis in types are added in any backend is through the explicit 'HsParTy' constructor. Precedence is also represented as its own datatype. * List out cases explicitly vs. catch-all * Fix printing of parens for QuantifiedConstraints The priority of printing 'forall' types was just one too high. Fixes haskell/haddock#877. * Accept HTML output for quantified contexts test * Preserve docs on type family instances (#867) * Preserve docs on type family instances The only problem was that the instance location was slightly off for type family instances. * Accept output (cherry picked from commit 133e9c2c168db19c1135479f7ab144c4e33af2a4) * Fix broken instance source links (#869) The problem manifests itself in instances that are defined in modules other than the module where the class is defined. The fix is just to thread through the 'Module' of the instance further along. Since orphan instances appear to already have been working, I didn't do anything there. (cherry picked from commit 2de7c2acf9b1ec85b09027a8bb58bf8512e91c05) * Add some more unicode related tests (#872) This has been fixed for sure ever since we switched from attoparsec to parsec. Parts of it may have been working before that, but there was a point where this would have failed (see haskell/haddock#191). A regression test never hurt anyone. :) (cherry picked from commit 5ec7715d418bfac0f26aec6039792a99a6e89370) * Misc tests (#858) * More tests * spliced types * constructor/pattern argument docs * strictness marks on fields with argument docs * latex test cases need seperate directory * Accept tests * Additional tests for the identifier parser (#816) * Add tests for the identifier parser * docs: Clarify how to delimit identifiers (cherry picked from commit 0861affeca4d72938f05a2eceddfae2c19199071) - - - - - 5ca14bed by Simon Jakobi at 2018-07-20T16:05:47+02:00 Revert "Revert "Bump GHC version to 8.6"" That commit didn't belong onto the ghc-8.6 branch. This reverts commit acbaef3b9daf1d2dea10017964bf886e77a8e967. - - - - - 2dd600dd by Simon Jakobi at 2018-07-20T16:18:21+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. - - - - - fceb2422 by Simon Jakobi at 2018-07-20T16:18:21+02:00 outOfScope: Recommend qualifying the identifier - - - - - acea5d23 by Simon Jakobi at 2018-07-20T16:19:35+02:00 outOfScope: Recommend qualifying the identifier (cherry picked from commit 73707ed58d879cc04cb644c5dab88c39ca1465b7) - - - - - 1a83ca55 by Simon Jakobi at 2018-07-20T16:19:35+02:00 Don't warn about ambiguous identifiers when the candidate names belong to the same type This also changes the defaulting heuristic for ambiguous identifiers. We now prefer local names primarily, and type constructors or class names secondarily. Partially fixes haskell/haddock#854. (cherry picked from commit d504a2864a4e1982e142cf88c023e7caeea3b76f) - - - - - 48374451 by Masahiro Sakai at 2018-07-20T17:06:42+02:00 Add # as a special character (#884) '#' has special meaning used for anchors and can be escaped using backslash. Therefore it would be nice to be listed as special characters. - - - - - 5e1a5275 by Alec Theriault at 2018-07-20T23:37:24+02:00 Let `haddock-test` bypass interface version check (#890) This means `haddock-test` might * crash during deserialization * deserialize incorrectly Still - it means things _might_ work where they were previously sure not to. - - - - - 27286754 by Yuji Yamamoto at 2018-07-23T08:16:01+02:00 Avoid "invalid argument (invalid character)" on non-unicode Windows (#892) Steps to reproduce and the error message ==== ``` > stack haddock basement ... snip ... Warning: 'A' is out of scope. Warning: 'haddock: internal error: <stdout>: commitBuffer: invalid argument (invalid character) ``` Environment ==== OS: Windows 10 ver. 1709 haddock: [HEAD of ghc-8.4 when I reproduce the error](https://github.com/haskell/haddock/commit/532b209d127e4cecdbf7e9e3dcf4f653a5605b5a). (I had to use this version to avoid another probrem already fixed in HEAD) GHC: 8.4.3 stack: Version 1.7.1, Git revision 681c800873816c022739ca7ed14755e85a579565 (5807 commits) x86_64 hpack-0.28.2 Related pull request ==== https://github.com/haskell/haddock/pull/566 - - - - - 6729d361 by Alec Theriault at 2018-07-23T13:52:56-07:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). - - - - - 8cf4e6b5 by Ryan Scott at 2018-07-27T11:28:03-04:00 eqTyCon_RDR now lives in TysWiredIn After GHC commit http://git.haskell.org/ghc.git/commit/f265008fb6f70830e7e92ce563f6d83833cef071 - - - - - 1ad251a6 by Alan Zimmerman at 2018-07-30T13:28:09-04:00 Match XFieldOcc rename in GHC Trac haskell/haddock#15386 (cherry picked from commit e3926b50ab8a7269fd6904b06e881745f08bc5d6) - - - - - 8aea2492 by Richard Eisenberg at 2018-08-02T10:54:17-04:00 Update against new HsImplicitBndrs - - - - - e42cada9 by Alec Theriault at 2018-08-04T17:51:30+02:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 0e852512 by Alex Biehl at 2018-08-06T13:04:02+02:00 Make --package-version optional for --hoogle generation (#899) * Make --package-version optional for --hoogle generation * Import mkVersion * It's makeVersion not mkVersion - - - - - d2abd684 by Noel Bourke at 2018-08-21T09:34:18+02:00 Remove unnecessary backslashes from docs (#908) On https://haskell-haddock.readthedocs.io/en/latest/markup.html#special-characters the backslash and backtick special characters showed up with an extra backslash before them – I think the escaping is not (or no longer) needed for those characters in rst. - - - - - 7a578a9e by Matthew Pickering at 2018-08-21T09:34:50+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - aa3d4db3 by Matthew Pickering at 2018-08-21T09:37:34+02:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 - - - - - ede91744 by Alec Theriault at 2018-08-21T09:42:52+02:00 Better test output when Haddock crashes on a test (#902) In particular: we report the tests that crashed seperately from the tests that produced incorrect output. In order for tests to pass (and exit 0), they must not crash and must produce the right output. - - - - - 4a872b84 by Guillaume Bouchard at 2018-08-21T09:45:57+02:00 Fix a typo (#878) - - - - - 4dbf7595 by Ben Sklaroff at 2018-08-21T12:04:09-04:00 Add ITcomment_line_prag token to Hyperlinker Parser This token is necessary for parsing #line pragmas inside nested comments. Reviewers: bgamari Reviewed By: bgamari Differential Revision: https://phabricator.haskell.org/D4935 - - - - - 9170b2a9 by Ben Gamari at 2018-08-21T17:55:15-04:00 Merge pull request haskell/haddock#893 from harpocrates/get-name-to-instances Accumulate explicitly which modules to load for 'attachInstances' - - - - - d57b57cc by Ben Gamari at 2018-08-21T17:59:13-04:00 Merge branch 'ghc-head' of github.com:haskell/haddock into ghc-head - - - - - 14601ca2 by Alec Theriault at 2018-08-21T19:09:37-04:00 Accumulate explicitly which modules to load for 'attachInstances' The old approach to fixing haskell/haddock#469, while correct, consumes a lot of memory. We ended up with a HUGE 'GblRdrEnv' in 'ic_rn_gbl_env'. However, 'getNameToInstancesIndex' takes that environment and compresses it down to a much smaller 'ModuleSet'. Now, we compute that 'ModuleSet' explicitly as we process modules. That way we can just tell 'getNameToInstancesIndex' what modules to load (instead of it trying to compute that information from the interactive context). (cherry picked from commit 5c7c596c51d69b92164e9ba920157b36ce2b2ec1) - - - - - 438c645e by Matthew Pickering at 2018-08-21T19:12:39-04:00 Load plugins when starting a GHC session (#905) Fixes haskell/haddock#900 (cherry picked from commit e6aa8fb47b9477cc5ef5e46097524fe83e080f6d) - - - - - a80c5161 by Alec Theriault at 2018-08-21T22:06:40-07:00 Better rendering of unboxed sums/tuples * adds space after/before the '#' marks * properly reify 'HsSumTy' in 'synifyType' - - - - - 88456cc1 by Alec Theriault at 2018-08-21T22:06:40-07:00 Handle promoted tuples in 'synifyType' When we have a fully applied promoted tuple, we can expand it out properly. - - - - - fd1c1094 by Alec Theriault at 2018-08-21T22:19:34-07:00 Accept test cases - - - - - 6e80d9e0 by Alec Theriault at 2018-08-21T22:24:03-07:00 Merge pull request haskell/haddock#914 from harpocrates/feature/unboxed-stuff Better rendering of unboxed sums, unboxed tuples, promoted tuples. - - - - - 181a23f1 by Ben Gamari at 2018-08-23T15:53:48-04:00 Merge remote-tracking branch 'origin/ghc-8.6' into ghc-8.6 - - - - - 3a18c1d8 by Alec Theriault at 2018-08-27T14:15:25-07:00 Properly synify promoted list types We reconstruct promoted list literals whenever possible. That means that 'synifyType' produces '[Int, Bool, ()] instead of (Int ': (() ': (Bool ': ([] :: [Type])))) - - - - - b4794946 by Alec Theriault at 2018-09-03T07:19:55-07:00 Only look at visible types when synifying a 'HsListTy' The other types are still looked at when considering whether to make a kind signature or not. - - - - - a231fce2 by Alec Theriault at 2018-09-03T07:38:10-07:00 Merge pull request haskell/haddock#922 from harpocrates/promoted-lists Properly synify promoted list types - - - - - 0fdf044e by Ningning Xie at 2018-09-15T10:25:58-04:00 Update according to GHC Core changes - - - - - 7379b115 by Ningning Xie at 2018-09-15T15:40:18-04:00 update dataFullSig to work with Co Quantification This should have been in the previous patch, but wasn't. - - - - - cf84a046 by Alec Theriault at 2018-09-17T20:12:18-07:00 Fix/add to various docs * Add documentation for a bunch of previously undocumented options (fixes haskell/haddock#870) * Extend the documentation of `--hoogle` considerably (see haskell/haddock#807) * Describe how to add docs to `deriving` clauses (fixes haskell/haddock#912) * Fix inaccurate docs about hyperlinking infix identifiers (fixes haskell/haddock#780) - - - - - ae017935 by Alec Theriault at 2018-09-22T08:32:16-07:00 Update Travis - - - - - d95ae753 by Alec Theriault at 2018-09-22T09:34:10-07:00 Accept failing tests Also silence orphan warnings. - - - - - f3e67024 by Alec Theriault at 2018-09-22T09:41:23-07:00 Bump haddock-api-2.21.0, haddock-library-1.7.0 * Update CHANGELOGS * Update new versions in Cabal files * Purge references to ghc-8.4/master branches in README - - - - - 3f136d4a by Alec Theriault at 2018-09-22T10:53:31-07:00 Turn haddock-library into a minor release Fix some version bounds in haddock-library too. - - - - - b9def006 by Alec Theriault at 2018-09-22T13:07:35-07:00 keep cabal.project file - - - - - 4909aca7 by Alec Theriault at 2018-10-16T09:36:30-07:00 Build on 7.4 and 7.8 - - - - - 99d20a28 by Herbert Valerio Riedel at 2018-10-16T18:45:52+02:00 Minor tweak to package description - - - - - a8059618 by Herbert Valerio Riedel at 2018-10-16T18:47:24+02:00 Merge pull request haskell/haddock#945 haddock-api 2.21.0 and haddock-library 1.6.1 release - - - - - 2d9bdfc1 by Alec Theriault at 2018-10-16T10:54:21-07:00 Bump haddock-library to 1.7.0 The 1.6.1 release should've been a major bump, since types in the `Documentation.Haddock.Parser.Monad` module changed. This version makes that module internal (as it morally should be). - - - - - ed340cef by Alec Theriault at 2018-10-16T14:59:13-07:00 Merge branch 'ghc-8.4' into ghc-8.6 - - - - - 2821a8df by Alec Theriault at 2018-10-16T15:14:48-07:00 Merge branch 'ghc-8.6' into ghc-head - - - - - a722dc84 by Alec Theriault at 2018-10-16T16:28:55-07:00 Latex type families (#734) * Support for type families in LaTeX The code is ported over from the XHTML backend. * Refactor XHTML and LaTeX family handling This is mostly a consolidation effort: stripping extra exports, inlining some short definitions, and trying to make the backends match. The LaTeX backend now has preliminary support for data families, although the only the data instance head is printed (not the actual constructors). Both backends also now use "newtype" for newtype data family instances. * Add some tests - - - - - 63377496 by Alec Theriault at 2018-10-16T16:39:07-07:00 Update changelog - - - - - 099a0110 by Alec Theriault at 2018-10-16T16:49:28-07:00 Merge pull request haskell/haddock#942 from harpocrates/update-docs Fix & add to documentation - - - - - 0927416f by Alec Theriault at 2018-10-16T16:50:14-07:00 Set UTF-8 encoding before writing files (#934) This should fix haskell/haddock#929, as well as guard against future problems of this sort in other places. Basically replaces 'writeFile' (which selects the users default locale) with 'writeUtf8File' (which always uses utf8). - - - - - 83b7b017 by Alec Theriault at 2018-10-16T17:42:05-07:00 Output pattern synonyms in Hoogle backend (#947) * Output pattern synonyms in Hoogle backend We were previously weren't outputting _any_ pattern synonyms, bundled or not. Now, we output both. Fixes haskell/haddock#946. * Update changelog - - - - - 81e5033d by Alec Theriault at 2018-10-16T18:04:40-07:00 Release `haddock{,-api}-2.22.0` This version will accompany ghc-8.6.2 - - - - - 9661744e by Alex Biehl at 2018-10-18T08:14:32-07:00 Add NewOcean theme And make it the default theme. - - - - - 7ae6d722 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Improve appearance and readability These changes include: - use latest Haskell's logo colors - decrease #content width to improve readability - use nicer font - improve sizes and distances - - - - - 37f8703d by NunoAlexandre at 2018-10-18T08:14:32-07:00 Include custom font in the html head - - - - - 1d5e1d79 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Update html test reference files - - - - - 53b7651f by NunoAlexandre at 2018-10-18T08:14:32-07:00 Make it responsive - It makes small screens taking more space than larger ones - fixes a few issues present in small screens currently - make it look good across different screen sizes. - - - - - 6aa1aeb1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make the style consistent with hackage Several things are addressed here: - better responsive behaviour on the header - better space usage - consistent colors overall - other nit PR comments - - - - - 3a250c5c by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Place the package name before the menu links This supports the expected responsive menu design, where the package name appears above the menu links. - - - - - cae699b3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update html-test reference files The package name element in the package-header is now a div instead of a paragraph, and it is now above the menu ul.links instead of below. - - - - - 2ec7fd2d by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve synopsis style and code - Use CSS3 instead of loading pictures to show "+" and "-" symbols - Drop redundant code - - - - - 0c874c01 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Decrease space between code blocks There was too much space between code blocks as pointed out by reviewers. - - - - - 85568ce2 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Add an initial-scale property to all haddock pages This solves an issue reported about the content looking incredibly small on mobile devices. - - - - - c1538926 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Address several PR comments - Darken text color like done for hackage - Move synopsis to left side - Make table of contents stick to the left on wide screens - Wrap links to avoid page overflow - Improve expand/collapse buttons - Fix issue with content size on mobile devices - Fix issue with font-size on landscape mode - Increase width of the content - Change colors of table of contents and synopsis - Etc - - - - - e6639e5f by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make synopsis scrollable on wide screens When the synopsis is longer than the screen, you can’t see its end and you can't scroll down either, making the content unreachable. - - - - - 1f0591ff by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve information density - Reduce font size - Improve space between and within code blocks - Improve alignments - Improve spacing within sub-blocks - - - - - bf083097 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Minor adjustments Bring in some adjustments made to hackage: - link colors - page header show everything when package title is too long - - - - - 10375fc7 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix responsive triggers overlap issue The min and max width triggers have the same values, which caused the style resolution to take an intersection of both style declarations when the screen resolution had the size of the limts (say 1280px), causing an odd behaviour and look. - - - - - 95ff2f95 by NunoAlexandre at 2018-10-18T08:14:32-07:00 Fix issue with menu alignment on firefox Reported and described here: https://github.com/haskell/haddock/pull/721#issuecomment-374668869 - - - - - dc86587e by Alex Biehl at 2018-10-18T08:14:32-07:00 Changelog entry for NewOcean - - - - - 27195e47 by Herbert Valerio Riedel at 2018-10-18T08:14:32-07:00 html-test --accept - - - - - 83f4f9c0 by Alex Biehl at 2018-10-18T08:14:32-07:00 Avoid name shadowing - - - - - 231487f1 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font to PT Sans Also migrate some general text related changes from hackage. - - - - - 313db81a by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Use 'flex' to fix header alignment - - - - - 5087367b by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Misc of tweaks - Update link colors to hackage scheme - Tune spacing between content elements - Update footer style - Fix and improve code blocks identation - - - - - b08020df by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update font in Xhtml.hs to PT Sans - - - - - 78ce06e3 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Improve code blocks styling - Fix and improve spacing - Improve colors and borders - - - - - 81262d20 by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Make package-header caption backward-compatible The current html generator of this branch wraps the package-header caption as a div, which does not work (without style adjustments) with the old themes. Changing it from div to span does the trick, without needing to adjust the old stylesheets. - - - - - dc4475cb by Nuno Alexandre at 2018-10-18T08:14:32-07:00 Update test-suite reference html pages - - - - - 393d35d8 by Alec Theriault at 2018-10-18T08:25:36-07:00 Accept tests - - - - - a94484ba by Alec Theriault at 2018-10-21T10:29:29-07:00 Fix CHANGELOG - - - - - 8797eca3 by Alec Theriault at 2018-10-21T10:36:19-07:00 Update 'data-files' to include NewOcean stuff - - - - - 1ae51e4a by Simon Jakobi at 2018-10-23T11:29:14+02:00 Fix typo in a warning - - - - - 009ad8e8 by Alec Theriault at 2018-10-24T12:47:47-07:00 Update JS dependencies This was done via `npm audit fix`. I think this fixes haskell/haddock#903 along with some more serious vulnerabilities that nobody seems to have noticed. - - - - - 051994db by Alec Theriault at 2018-10-24T17:31:09-07:00 Resurrect the style-switcher This fixes haskell/haddock#810. Looks like things were broken during the quickjump refactor of the JS. For the (git) record: I do not think the style switcher is a good idea. I'm fixing it for the same reason @mzero added it; as an answer to "rumblings from some that they didn't want their pixels changed on bit" - - - - - 2a1d620f by Alec Theriault at 2018-10-24T17:38:07-07:00 Fix copy-pasta error in data-files - - - - - ed5bfb7f by Alec Theriault at 2018-10-24T20:42:14-07:00 Fix the synopsis button Here's these changes are supposed to do: * put the synopsis back on the right side * properly have it on the edge of the screen on wide screens * adjust the background of the synopsis to match the button (otherwise the grey blends in with what is underneath) * get rid of the dotted purple line * the synopsis contents are now scrollable even when in wide screens (this has been a long-standing bug) - - - - - 883fd74b by Alec Theriault at 2018-10-25T20:16:46-07:00 Avoid more conflicts in generated ids (#954) This fixes haskell/haddock#953 by passing more names into the generated ids. - - - - - ea54e331 by Alec Theriault at 2018-10-25T21:07:12-07:00 Don't hide bullets in method docs I think thst CSS was meant only to deal with fields and the effect on bullets was accidental. Fixes haskell/haddock#926. - - - - - 9a14ef4a by Alec Theriault at 2018-10-25T22:02:07-07:00 Indent more things + slightly smaller font - - - - - b9f17e29 by Alec Theriault at 2018-10-25T22:10:01-07:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 096a3cfa by Alec Theriault at 2018-10-25T22:24:38-07:00 Accept HTML output - - - - - 2669517d by Alec Theriault at 2018-10-26T09:02:35-07:00 User manual + stuff for building GHC docs - - - - - 46b27687 by Alec Theriault at 2018-10-26T09:10:59-07:00 Make 'Contents' in NewOcean scrollable This only happens if the contents block on the left is so big that it doesn't fit (vertically) on the page. If that happens, we want it to be scrollable. - - - - - 3443dd94 by Alec Theriault at 2018-10-26T09:36:46-07:00 Revert "Make 'Contents' in NewOcean scrollable" This reverts commit f909ffd8353d6463fd5dd184998a32aa98d5c922. I missed the fact this also forces the 'Contents' to always go down to the bottom of the page. - - - - - ed081424 by Alec Theriault at 2018-10-26T14:22:23-07:00 Avoid some partiality AFAICT this wasn't causing any crashes, but that's mostly because we happen not to be forcing `pkgStr` when it would diverge. We come dangerously close to doing that in `ppHtmlIndex`. Fixes haskell/haddock#569. - - - - - 6a5bec41 by Alec Theriault at 2018-10-27T10:05:04-07:00 Fix documentation in `haddock-api` (#957) * Fix misplaced Haddocks in Haddock itself Haddock should be able to generate documentation for 'haddock-api' again. * Make CI check that documentation can be built. * Add back a doc that is OK - - - - - 5100450a by Matthew Yacavone at 2018-10-27T14:51:38-04:00 More explicit foralls (GHC Proposal 0007) - - - - - 8771a6b0 by Alec Theriault at 2018-11-05T13:58:11-08:00 Only run MathJax on entities with "mathjax" class (#960) Correspondingly, we wrap all inline/diplay math in <span class="mathjax"> ... the math .... </span> This fixes haskell/haddock#959. - - - - - bd7ff5c5 by Alec Theriault at 2018-11-05T15:54:22-08:00 Deduplicate some work in 'AttachInstances' Perf only change: * avoid needlessly union-ing maps * avoid synify-ing instances twice Took this opportunity to add some docs too - - - - - cf99fd8f by Alec Theriault at 2018-11-05T15:54:22-08:00 Specialize some SYB functions Perf only change: * Add a 'SPECIALIZE' pragma to help GHC optimize a 'Data a =>' constraint * Manually specialize the needlessly general type of 'specializeTyVarBndrs' - - - - - 4f91c473 by Alec Theriault at 2018-11-05T15:54:22-08:00 Improve perf of renaming Perf only change: * don't look up type variable names (they're never in the environment) * use a difference list for accumulating missing names * more efficient 'Functor'/'Applicative' instances for 'RnM' - - - - - 4bbab0d4 by Alec Theriault at 2018-11-05T15:54:22-08:00 Faster 'Text' driven parser combinators Perf only change: * use 'getParserState'/'setParserState' to make 'Text'-optimized parser combinators * minimize uses of 'Data.Text.{pack,unpack,cons,snoc}' - - - - - fa430c02 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support hyperlink labels with inline markup The parser for pictures hasn't been properly adjusted yet. - - - - - c1431035 by Alec Theriault at 2018-11-06T12:03:24-08:00 Support (and flatten) inline markup in image links Inline markup is supported in image links but, as per the [commonmark recommendation][0], it is stripped back to a plain text representation. [0]: https://spec.commonmark.org/0.28/#example-547 - - - - - d4ee1ba5 by Alec Theriault at 2018-11-06T12:03:24-08:00 Accept test case - - - - - 8088aeb1 by Alec Theriault at 2018-11-06T12:03:24-08:00 Fix/add to haddock-library test suite - - - - - e78f644d by Alec Theriault at 2018-11-06T13:26:31-08:00 Bump version bounds - - - - - 644335eb by Alec Theriault at 2018-11-06T13:53:30-08:00 Merge pull request haskell/haddock#875 from harpocrates/feature/markup-in-hyperlinks Inline markup in markdown-style links and images - - - - - e173ed0d by Alec Theriault at 2018-11-07T12:37:18-08:00 Fix issues around plus/minus * swap the minimize unicode to something more intuitive * use new unicode expander/collapser for instance lists * address some alignment issues in the "index" page - - - - - b2d92df7 by Alec Theriault at 2018-11-07T13:41:57-08:00 Allow "Contents" summary to scroll in a fixed div In the unfortunate event that the "Contents" summary doesn't fit vertically (like in the "Prelude"), it will be scrollable. - - - - - ca704c23 by Alec Theriault at 2018-11-07T13:45:15-08:00 Accept HTML output changes - - - - - 82c0ec6d by Alec Theriault at 2018-11-07T18:12:54-08:00 overflow-y 'scroll' -> 'auto' - - - - - 571d7657 by Alec Theriault at 2018-11-08T19:44:12-08:00 Clicking on "Contents" navigates to top of page - - - - - 8065a012 by Alec Theriault at 2018-11-08T19:44:17-08:00 Space out functions more Also, functions and data decls now have the same space before and after them. - - - - - cc650ede by Alec Theriault at 2018-11-09T08:13:35-08:00 Merge branch 'ghc-8.6' into wip/new-ocean - - - - - 65f8c17f by Alec Theriault at 2018-11-10T14:04:06-08:00 Update changelog - - - - - 20473847 by Alec Theriault at 2018-11-10T14:21:40-08:00 Replace oplus/ominus expander/collapser icons with triangles - - - - - 16592957 by Alec Theriault at 2018-11-10T14:35:10-08:00 Merge pull request haskell/haddock#949 from haskell/wip/new-ocean Introduce NewOcean theme. - - - - - 357cefe1 by Alec Theriault at 2018-11-10T16:02:13-08:00 Merge branch 'ghc-8.6' into ghc-head - - - - - de612267 by Alec Theriault at 2018-11-11T20:01:21-08:00 Rename 'NewOcean' theme to 'Linuwial' - - - - - 954b5baa by Alec Theriault at 2018-11-12T08:33:18-08:00 Add blockquote styling Matches b71da1feabf33efbbc517ac376bb690b5a604c2f from hackage-server. Fixes haskell/haddock#967. - - - - - d32c0b0b by Fangyi Zhou at 2018-11-12T10:24:13-08:00 Fix some broken links (#15733) Summary: For links in subpackages as well. https://phabricator.haskell.org/D5257 Test Plan: Manually verify links Reviewers: mpickering, bgamari, osa1 Reviewed By: osa1 GHC Trac Issues: haskell/haddock#15733 Differential Revision: https://phabricator.haskell.org/D5262 - - - - - 41098b1f by Alp Mestanogullari at 2018-11-15T22:40:09+01:00 Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change It got introduced in ghc/ghc at ae2c9b40f5b6bf272251d1f4107c60003f541b62. - - - - - c5c1c7e0 by Alec Theriault at 2018-11-15T13:48:13-08:00 Merge pull request haskell/haddock#970 from alpmestan/alp/fix-promotionflag Follow GHC HEAD's HsTypes.Promoted -> BasicTypes.PromotionFlag change - - - - - 6473d3a4 by Shayan-Najd at 2018-11-23T01:38:49+01:00 [TTG: Handling Source Locations] Foundation and Pat Trac Issues haskell/haddock#15495 This patch removes the ping-pong style from HsPat (only, for now), using the plan laid out at https://ghc.haskell.org/trac/ghc/wiki/ImplementingTreesThatGrow/HandlingSourceLocations (solution A). - the class `HasSrcSpan`, and its functions (e.g., `cL` and `dL`), are introduced - some instances of `HasSrcSpan` are introduced - some constructors `L` are replaced with `cL` - some patterns `L` are replaced with `dL->L` view pattern - some type annotation are necessarily updated (e.g., `Pat p` --> `Pat (GhcPass p)`) - - - - - 7a088dfe by Alec Theriault at 2018-11-26T11:11:28-08:00 More uniform handling of `forall`'s in HTML/LaTeX * don't forget to print explicit `forall`'s when there are arg docs * when printing an explicit `forall`, print all tyvars Fixes haskell/haddock#973 - - - - - d735e570 by Alec Theriault at 2018-12-12T08:42:09-08:00 Fix warnings, accept output * remove redundant imports (only brought to light due to recent work for improving redundant import detection) * fix a bug that was casuing exports to appear in reverse order * fix something in haddock-library that prevented compilation on old GHC's - - - - - a3852f8a by Zejun Wu at 2018-12-14T09:37:47-05:00 Output better debug infromation on internal error in extractDecl This will make investigation of haskell/haddock#979 easier - - - - - 2eccb5b9 by Alec Theriault at 2018-12-17T09:25:10-05:00 Refactor names + unused functions (#982) This commit should not introduce any change in functionality! * consistently use `getOccString` to convert `Name`s to strings * compare names directly when possible (instead of comparing strings) * get rid of unused utility functions - - - - - e82e4df8 by Alec Theriault at 2018-12-20T16:16:30-05:00 Load plugins when compiling each module (#983) * WIP: Load (typechecker) plugins from language pragmas * Revert "Load plugins when starting a GHC session (#905)" This reverts commit 72d82e52f2a6225686d9668790ac33c1d1743193. * Simplify plugin initialization code - - - - - 96e86f38 by Alec Theriault at 2018-12-23T10:23:20-05:00 Properly synify and render promoted type variables (#985) * Synify and render properly promoted type variables Fixes haskell/haddock#923. * Accept output - - - - - 23343345 by Alec Theriault at 2018-12-27T16:39:38-05:00 Remove `haddock-test`'s dep. on `syb` (#987) The functionality is easily inlined into one short function: `gmapEverywhere`. This doesn't warrant pulling in another package. - - - - - d0734f21 by Alec Theriault at 2018-12-27T16:39:52-05:00 Address deprecation warnings in `haddock-test` (#988) Fixes haskell/haddock#885. - - - - - 4d9f144e by mynguyen at 2018-12-30T23:42:26-05:00 Visible kind application haddock update - - - - - ffe0e9ed by Alec Theriault at 2019-01-07T13:55:22-08:00 Print kinded tyvars in constructors for Hoogle (#993) Fixes haskell/haddock#992 - - - - - 2e18b55d by Alec Theriault at 2019-01-10T16:42:45-08:00 Accept new output `GHC.Maybe` -> `Data.Maybe` (#996) Since 53874834b779ad0dfbcde6650069c37926da1b79 in GHC, "GHC.Maybe" is marked as `not-home`. That changes around some test output. - - - - - 055da666 by Gabor Greif at 2019-01-22T14:41:51+01:00 Lone typofix - - - - - 01bb71c9 by Alec Theriault at 2019-01-23T11:46:46-08:00 Keep forall on H98 existential data constructors (#1003) The information about whether or not there is a source-level `forall` is already available on a `ConDecl` (as `con_forall`), so we should use it instead of always assuming `False`! Fixes haskell/haddock#1002. - - - - - f9b9bc0e by Ryan Scott at 2019-01-27T09:28:12-08:00 Fix haskell/haddock#1004 with a pinch of dropForAlls - - - - - 5cfcdd0a by Alec Theriault at 2019-01-28T16:49:57-08:00 Loosen 'QuickCheck' and 'hspec' bounds It looks like the new versions don't cause any breakage and loosening the bounds helps deps fit in one stack resolver. - - - - - 3545d3dd by Alec Theriault at 2019-01-31T01:37:25-08:00 Use `.hie` files for the Hyperlinker backend (#977) # Summary This is a large architectural change to the Hyperlinker. * extract link (and now also type) information from `.hie` instead of doing ad-hoc SYB traversals of the `RenamedSource`. Also adds a superb type-on-hover feature (#715). * re-engineer the lexer to avoid needless string conversions. By going directly through GHC's `P` monad and taking bytestring slices, we avoid a ton of allocation and have better handling of position pragmas and CPP. In terms of performance, the Haddock side of things has gotten _much_ more efficient. Unfortunately, much of this is cancelled out by the increased GHC workload for generating `.hie` files. For the full set of boot libs (including `ghc`-the-library) * the sum of total time went down by 9-10% overall * the sum of total allocations went down by 6-7% # Motivation Haddock is moving towards working entirely over `.hi` and `.hie` files. This change means we no longer need the `RenamedSource` from `TypecheckedModule` (something which is _not_ in `.hi` files). # Details Along the way a bunch of things were fixed: * Cross package (and other) links are now more reliable (#496) * The lexer tries to recover from errors on every line (instead of at CPP boundaries) * `LINE`/`COLUMN` pragmas are taken into account * filter out zero length tokens before rendering * avoid recomputing the `ModuleName`-based `SrcMap` * remove the last use of `Documentation.Haddock.Utf8` (see haskell/haddock#998) * restructure temporary folder logic for `.hi`/`.hie` model - - - - - 2ded3359 by Herbert Valerio Riedel at 2019-02-02T12:06:12+01:00 Update/modernise haddock-library.cabal file - - - - - 62b93451 by Herbert Valerio Riedel at 2019-02-02T12:19:31+01:00 Tentatively declare support for unreleased base-4.13/ghc-8.8 - - - - - 6041e767 by Herbert Valerio Riedel at 2019-02-02T16:04:32+01:00 Normalise LICENSE text w/ cabal's BSD2 template Also, correct the `.cabal` files to advertise `BSD2` instead of the incorrect `BSD3` license. - - - - - 0b459d7f by Alec Theriault at 2019-02-02T18:06:12-08:00 CI: fetch GHC from validate artifact Should help make CI be less broken - - - - - 6b5c07cf by Alec Theriault at 2019-02-02T18:06:12-08:00 Fix some Hyperlinker test suite fallout * Amend `ParserSpec` to match new Hyperlinker API - pass in compiler info - strip out null tokens * Make `hypsrc-test` pass reliably - strip out `local-*` ids - strip out `line-*` ids from the `ClangCppBug` test - re-accept output - - - - - ded34791 by Nathan Collins at 2019-02-02T18:31:23-08:00 Update README instructions for Stack No need to `stack install` Haddock to test it. Indeed, `stack install` changes the `haddock` on user's `PATH` if `~/.local/bin` is on user's `PATH` which may not be desirable when hacking on Haddock. - - - - - 723298c9 by Alec Theriault at 2019-02-03T09:11:05-08:00 Remove `Documentation.Haddock.Utf8` The circumstances under which this module appeared are completely gone. The Hyperlinker backend no longer needs this module (it uses the more efficient `Encoding` module from `ghc`). Why no deprecation? Because this module really shouldn't exist! - It isn't used in `haddock-library`/`haddock-api` anymore - It was copy pasted directly from `utf8-string` - Folks seeking a boot-lib only solution can use `ghc`'s `Encoding` - - - - - 51050006 by Alec Theriault at 2019-02-03T22:58:58-08:00 Miscellaneous improvements to `Convert` (#1020) Now that Haddock is moving towards working entirely over `.hi` and `.hie` files, all declarations and types are going to be synthesized via the `Convert` module. In preparation for this change, here are a bunch of fixes to this module: * Add kind annotations to type variables in `forall`'s whose kind is not `Type`, unless the kind can be inferred from some later use of the variable. See `implicitForAll` and `noKindTyVars` in particular if you wish to dive into this. * Properly detect `HsQualTy` in `synifyType`. This is done by following suit with what GHC's `toIfaceTypeX` does and checking the first argument of `FunTy{} :: Type` to see if it classified as a given/wanted in the typechecker (see `isPredTy`). * Beef up the logic around figuring out when an explicit `forall` is needed. This includes: observing if any of the type variables will need kind signatures, if the inferred type variable order _without_ a forall will still match the one GHC claims, and some other small things. * Add some (not yet used) functionality for default levity polymorphic type signatures. This functionality similar to `fprint-explicit-runtime-reps`. Couple other smaller fixes only worth mentioning: * Show the family result signature only when it isn't `Type` * Fix rendering of implicit parameters in the LaTeX and Hoogle backends * Better handling of the return kind of polykinded H98 data declarations * Class decls produced by `tyThingToLHsDecl` now contain associated type defaults and default method signatures when appropriate * Filter out more `forall`'s in pattern synonyms - - - - - 841980c4 by Oleg Grenrus at 2019-02-04T08:44:25-08:00 Make a fixture of weird parsing of lists (#997) The second example is interesting. If there's a list directly after the header, and that list has deeper structure, the parser is confused: It finds two lists: - One with the first nested element, - everything after it I'm not trying to fix this, as I'm not even sure this is a bug, and not a feature. - - - - - 7315c0c8 by Ryan Scott at 2019-02-04T12:17:56-08:00 Fix haskell/haddock#1015 with dataConUserTyVars (#1022) The central trick in this patch is to use `dataConUserTyVars` instead of `univ_tvs ++ ex_tvs`, which displays the foralls in a GADT constructor in a way that's more faithful to how the user originally wrote it. Fixes haskell/haddock#1015. - - - - - ee0b49a3 by Ryan Scott at 2019-02-04T15:25:17-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. - - - - - 1c850dc8 by Alan Zimmerman at 2019-02-05T21:54:18+02:00 Matching changes in GHC for haskell/haddock#16236 - - - - - ab03c38e by Simon Marlow at 2019-02-06T08:07:33+00:00 Merge pull request haskell/haddock#1014 from hvr/pr/bsd2-normalise Normalise LICENSE text w/ cabal's BSD2 template - - - - - 5a92ccae by Alec Theriault at 2019-02-10T06:21:55-05:00 Merge remote-tracking branch 'gitlab/wip/T16236-2' into ghc-head - - - - - c0485a1d by Alec Theriault at 2019-02-10T03:32:52-08:00 Removes `haddock-test`s dependency on `xml`/`xhtml` (#1027) This means that `html-test`, `latex-test`, `hoogle-test`, and `hypsrc-test` now only depend on GHC boot libs. So we should now be able to build and run these as part of GHC's testsuite. \o/ The reference output has changed very slightly, in three ways: * we don't convert quotes back into `&quot;` as the `xml` lib did * we don't add extra `&nbsp;` as the `xml` lib did * we now remove the entire footer `div` (instead of just emptying it) - - - - - 65a448e3 by Alec Theriault at 2019-02-11T12:27:41-05:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - 360ca937 by Alec Theriault at 2019-02-13T11:36:11-05:00 Clean up logic for guessing `-B` and `--lib` (#1026) Haddock built with the `in-ghc-tree` flag tries harder to find the GHC lib folder and its own resources. This should make it possible to use `in-ghc-tree`-built Haddock without having to specify the `-B` and `--lib` options (just how you can use in-tree GHC without always specifying the `-B` option). The logic to do this relies on `getExecutablePath`, so we only get this auto-detection on platforms where this function works. - - - - - d583e364 by Alec Theriault at 2019-02-16T10:41:22-05:00 Fix tests broken by GHC Changes in 19626218566ea709b5f6f287d3c296b0c4021de2 affected some of the hyperlinker output. Accepted the new output (hovering over a `..` now shows you what that wildcard binds). Also fixed some stray deprecation warnings. - - - - - da0c42cc by Vladislav Zavialov at 2019-02-17T11:39:19+03:00 Parser changes to match !380 - - - - - ab96bed7 by Ryan Scott at 2019-02-18T04:44:08-05:00 Bump ghc version to 8.9 - - - - - 44b7c714 by Alec Theriault at 2019-02-22T05:49:43-08:00 Match GHC changes for T16185 `FunTy` now has an `AnonArgFlag` that indicates whether the arrow is a `t1 => t2` or `t1 -> t2`. This commit shouldn't change any functionality in Haddock. - - - - - 2ee653b1 by Alec Theriault at 2019-02-24T18:53:33-08:00 Update .travis.yml Points to the new GHC CI artifact. - - - - - 90939d71 by Alec Theriault at 2019-02-25T00:42:41-08:00 Support value/type namespaces on identifier links Identifier links can be prefixed with a 'v' or 't' to indicate the value or type namespace of the desired identifier. For example: -- | Some link to a value: v'Data.Functor.Identity' -- -- Some link to a type: t'Data.Functor.Identity' The default is still the type (with a warning about the ambiguity) - - - - - d6ed496c by Alec Theriault at 2019-02-25T00:42:46-08:00 Better identifier parsing * '(<|>)' and '`elem`' now get parsed and rendered properly as links * 'DbModule'/'DbUnitId' now properly get split apart into two links * tuple names now get parsed properly * some more small niceties... The identifier parsing code is more precise and more efficient (although to be fair: it is also longer and in its own module). On the rendering side, we need to pipe through information about backticks/parens/neither all the way through from renaming to the backends. In terms of impact: a total of 35 modules in the entirety of the bootlib + ghc lib docs change. The only "regression" is things like '\0'. These should be changed to @\\0@ (the path by which this previously worked seems accidental). - - - - - 3c3b404c by Alec Theriault at 2019-02-25T22:12:11-08:00 Fix standalone deriving docs Docs on standalone deriving decls for classes with associated types should be associated with the class instance, not the associated type instance. Fixes haskell/haddock#1033 - - - - - d51ef69e by Alec Theriault at 2019-02-26T19:14:59-08:00 Fix bogus identifier defaulting This avoids a situation in which an identifier would get defaulted to a completely different identifier. Prior to this commit, the 'Bug1035' test case would hyperlink 'Foo' into 'Bar'! Fixes haskell/haddock#1035. - - - - - 88cbbdc7 by Ryan Scott at 2019-02-27T10:14:03-05:00 Visible dependent quantification (#16326) changes - - - - - 0dcf6cee by Xia Li-yao at 2019-02-27T21:53:27-05:00 Menu item controlling which instances are expanded/collapsed (#1007) Adds a menu item (like "Quick Jump") for options related to displaying instances. This provides functionality for: * expanding/collapsing all instances on the currently opened page * controlling whether instances are expanded/collapsed by default * controlling whether the state of instances should be "remembered" This new functionality is implemented in Typescript in `details-helper`. The built-in-themes style switcher also got a revamp so that all three of QuickJump, the style switcher, and instance preferences now have the same style and implementation structure. See also: https://mail.haskell.org/pipermail/haskell-cafe/2019-January/130495.html Fixes haskell/haddock#698. Co-authored-by: Lysxia <lysxia at gmail.com> Co-authored-by: Nathan Collins <conathan at galois.com> - - - - - 3828c0fb by Alec Theriault at 2019-02-28T12:42:49-05:00 `--show-interface` should output to stdout. (#1040) Fixes haskell/haddock#864. - - - - - a50f4cda by gbaz at 2019-03-01T07:43:16-08:00 Increase contrast of Linuwal theme (#1037) This is to address the concern that, on less nice and older screens, some of the shades of grey blend in too easily with the white background. * darken the font slightly * darken slightly the grey behind type signatures and such * add a border and round the corners on code blocks * knock the font down by one point - - - - - ab4d41de by Alec Theriault at 2019-03-03T09:23:26-08:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 12f509eb by Ben Gamari at 2019-03-04T22:13:20-05:00 Remove reference to Opt_SplitObjs flag Split-objects has been removed. - - - - - 5b3e4c9a by Ryan Scott at 2019-03-06T19:16:24-05:00 Update html-test output to reflect haskell/haddock#16391 changes - - - - - fc228af1 by Alec Theriault at 2019-03-09T08:29:23-08:00 Match changes for "Stop inferring over-polymorphic kinds" The `hsq_ext` field of `HsQTvs` is now just the implicit variables (instead of also including information about which of these variables are dependent). This commit shouldn't change any functionality in Haddock. - - - - - 6ac109eb by Alec Theriault at 2019-03-09T11:22:55-08:00 Add .hi, .dyn_hi, etc files to .gitignore Fixes haskell/haddock#1030. - - - - - b55f0c05 by Alec Theriault at 2019-03-09T11:22:55-08:00 Better support for default methods in classes * default methods now get rendered differently * default associated types get rendered * fix a forgotten `s/TypeSig/ClassOpSig/` refactor in LaTeX backend * LaTeX backend now renders default method signatures NB: there is still no way to document default class members and the NB: LaTeX backend still crashes on associated types - - - - - 10aea0cf by Alec Theriault at 2019-03-09T11:22:55-08:00 Avoid multi-line `emph` in LaTeX backend `markupWarning` often processes inputs which span across paragraphs. Unfortunately, LaTeX's `emph` is not made to handle this (and will crash). Fixes haskell/haddock#936. - - - - - d22dc2c9 by Alec Theriault at 2019-03-09T11:22:55-08:00 Many LaTeX backend fixes After this commit, we can run with `--latex` on all boot libraries without crashing (although the generated LaTeX still fails to compile in a handful of larger packages like `ghc` and `base`). * Add newlines after all block elements in LaTeX. This is important to prevent the final output from being more an more indented. See the `latext-test/src/Example` test case for a sample of this. * Support associated types in class declarations (but not yet defaults) * Several small issues for producing compiling LaTeX; - avoid empy `\haddockbeginargs` lists (ex: `type family Any`) - properly escape identifiers depending on context (ex: `Int#`) - add `vbox` around `itemize`/`enumerate` (so they can be in tables) * Several spacing fixes: - limit the width of `Pretty`-arranged monospaced code - cut out extra space characters in export lists - only escape spaces if there are _multiple_ spaces - allow type signatures to be multiline (even without docs) * Remove uninteresting and repetitive `main.tex`/`haddock.sty` files from `latex-test` test reference output. Fixes haskell/haddock#935, haskell/haddock#929 (LaTeX docs for `text` build & compile) Fixes haskell/haddock#727, haskell/haddock#930 (I think both are really about type families...) - - - - - 0e6cee00 by Alec Theriault at 2019-03-29T12:11:56-07:00 Remove workaround for now-fixed Clang CPP bug (#1028) Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where lines that started with an octothorpe but turned out not to lex like pragmas would have an extra line added after them. Since this bug has been fixed upstream and that it doesn't have dire consequences anyways, the workaround is not really worth it anymore - we can just tell people to update their clang version (or re-structure their pragma code). - - - - - ce05434d by Alan Zimmerman at 2019-03-29T12:12:11-07:00 Matching changes in GHC for haskell/haddock#16236 (cherry picked from commit 3ee6526d4ae7bf4deb7cd1caf24b3d7355573576) - - - - - d85766b2 by Ben Gamari at 2019-03-29T12:14:04-07:00 Bump GHC to 8.8 - - - - - 5a82cbaf by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Redo ParseModuleHeader - - - - - b9033348 by Oleg Grenrus at 2019-05-05T13:02:00-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - bb55c8f4 by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove outdated `.ghci` files and `scripts` The `.ghci` files are actively annoying when trying to `cabal v2-repl`. As for the `scripts`, the distribution workflow is completely different. - - - - - 5ee244dc by Alec Theriault at 2019-05-13T16:10:07-07:00 Remove obsolete arcanist files + STYLE Now that GHC is hosted on Gitlab, the arcanist files don't make sense anymore. The STYLE file contains nothing more than a dead link too. - - - - - d07c1928 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Redo ParseModuleHeader - - - - - 492762d2 by Oleg Grenrus at 2019-05-13T16:41:43-07:00 Comment C, which clarifies why e.g. ReadP is not enough - - - - - af2ac773 by Ryan Scott at 2019-05-14T17:22:13-04:00 Changes for haskell/haddock#16110/#16356 - - - - - 6820ed0d by Alec Theriault at 2019-05-17T08:51:27-07:00 Unbreak haskell/haddock#1004 test case `fail` is no longer part of `Monad`. - - - - - 6bf7be98 by Alec Theriault at 2019-05-17T08:51:27-07:00 Fix haskell/haddock#1063 with better parenthesization logic for contexts The only other change in html/hoogle/hyperlinker output for the boot libraries that this caused is a fix to some Hoogle output for implicit params. ``` $ diff -r _build/docs/ old_docs diff -r _build/docs/html/libraries/base/base.txt old_docs/html/libraries/base/base.txt 13296c13296 < assertError :: (?callStack :: CallStack) => Bool -> a -> a --- > assertError :: ?callStack :: CallStack => Bool -> a -> a ``` - - - - - b5716b61 by Ryan Scott at 2019-05-22T17:24:32-04:00 Match changes with haskell/haddock#14332 - - - - - c115abf6 by Alec Theriault at 2019-05-26T16:01:58-04:00 Remove Haddock's dependency on `Cabal` At this point, Haddock depended on Cabal-the-library solely for a verbosity parser (which misleadingly accepts all sorts of verbosity options that Haddock never uses). Now, the only dependency on Cabal is for `haddock-test` (which uses Cabal to locate the Haddock interface files of a couple boot libraries). - - - - - e5b2d4a3 by Alec Theriault at 2019-05-26T16:16:25-04:00 Regression test: promoted lists in associated types When possible, associated types with promoted lists should use the promoted list literal syntax (instead of repeated applications of ': and '[]). This was fixed in 2122de5473fd5b434af690ff9ccb1a2e58491f8c. Closes haskell/haddock#466, - - - - - cc5ad5d3 by Alec Theriault at 2019-05-26T17:55:54-04:00 Merge branch 'ghc-8.6' into ghc-8.8 - - - - - 4b3301a6 by Alec Theriault at 2019-05-26T17:57:52-04:00 Release haddock-2.23, haddock-library-1.8.0 Tentatively adjust bounds and changelogs for the release to be bundled with GHC 8.8.1. - - - - - 69c7cfce by Matthew Pickering at 2019-05-30T10:54:27+01:00 Update hyperlinker tests for new types in .hie files - - - - - 29b7e738 by Zubin Duggal at 2019-05-30T10:57:51+01:00 update for new way to store hiefile headers - - - - - aeca5d5f by Zubin Duggal at 2019-06-04T18:57:42-04:00 update for new way to store hiefile headers - - - - - ba2ca518 by Ben Gamari at 2019-06-07T23:11:14+00:00 Update test output for introduction of Safe-Inferred - - - - - 3a975a6c by Ryan Scott at 2019-07-03T12:06:27-04:00 Changes for haskell/haddock#15247 - - - - - 0df46555 by Zubin Duggal at 2019-07-22T10:52:50+01:00 Fix haddockHypsrcTest - - - - - 2688686b by Sylvain Henry at 2019-09-12T23:19:39+02:00 Fix for GHC module renaming - - - - - 9ec0f3fc by Alec Theriault at 2019-09-20T03:21:00-04:00 Fix Travis CI, loosen .cabal bounds (#1089) Tentatively for the 2.23 release: * updated Travis CI to work again * tweaked bounds in the `.cabal` files * adjusted `extra-source-files` to properly identify test files - - - - - ca559beb by Matthías Páll Gissurarson at 2019-09-28T12:14:40-04:00 Small change in to facilitate extended typed-holes (#1090) This change has no functional effect on haddock itself, it just changes one pattern to use `_ (` rather than `_(`, so that we may use `_(` as a token for extended typed-holes later. - - - - - 02e28976 by Vladislav Zavialov at 2019-09-28T12:17:45-04:00 Remove spaces around @-patterns (#1093) This is needed to compile `haddock` when [GHC Proposal haskell/haddock#229](https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0229-whitespace-bang-patterns.rst) is implemented. - - - - - 83cbbf55 by Alexis King at 2019-09-30T21:12:42-04:00 Fix the ignore-exports option (#1082) The `ignore-exports` option has been broken since haskell/haddock#688, as mentioned in https://github.com/haskell/haddock/pull/766#issue-172505043. This PR fixes it. - - - - - e127e0ab by Ben Gamari at 2019-10-06T15:12:06-04:00 Fix a few haddock issues - - - - - 3a0f5c89 by Zubin Duggal at 2019-10-07T17:56:13-04:00 Fix crash when there are no srcspans in the file due to CPP - - - - - 339c5ff8 by Alec Theriault at 2019-10-07T17:56:13-04:00 Prefer un-hyperlinked sources to no sources It is possible to fail to extract an HIE ast. This is however not a reason to produce _no_ output - we should still make a colorized HTML page. - - - - - d47ef478 by Alec Theriault at 2019-10-07T17:56:13-04:00 Add a regression test for haskell/haddock#1091 Previously, this input would crash Haddock. - - - - - ed7c8b0f by Alec Theriault at 2019-10-07T20:56:48-04:00 Add Hyperlinker test cases for TH-related stuff Hopefully this will guard against regressions around quasiquotes, TH quotes, and TH splices. - - - - - d00436ab by Andreas Klebinger at 2019-10-21T15:53:03+02:00 Refactor for withTiming changes. - - - - - 4230e712 by Ben Gamari at 2019-10-22T09:36:37-04:00 Merge pull request haskell/haddock#1101 from AndreasPK/withTimingRefactor Refactor for withTiming changes. - - - - - d155c5f4 by Ryan Scott at 2019-10-23T10:37:17-04:00 Reify oversaturated data family instances correctly (#1103) This fixes haskell/haddock#1103 by adapting the corresponding patch for GHC (see https://gitlab.haskell.org/ghc/ghc/issues/17296 and https://gitlab.haskell.org/ghc/ghc/merge_requests/1877). - - - - - 331a5adf by Sebastian Graf at 2019-10-25T17:14:40+02:00 Refactor for OutputableBndrId changes - - - - - 48a490e0 by Ben Gamari at 2019-10-27T10:16:16-04:00 Merge pull request haskell/haddock#1105 from sgraf812/wip/flexible-outputable Refactor for OutputableBndrId changes - - - - - f62a7dfc by Sebastian Graf at 2019-11-01T11:54:16+00:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - d9b242ed by Ryan Scott at 2019-11-03T13:20:03-05:00 Changes from haskell/haddock#14579 We now have a top-level `tyConAppNeedsKindSig` function, which means that we can delete lots of code in `Convert`. (cherry picked from commit cfd682c5fd03b099a3d78c44f9279faf56a0ac70) - - - - - dfd42406 by Sebastian Graf at 2019-11-04T07:02:14-05:00 Define `XRec` for location information and get rid of `HasSrcSpan` In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a simpler way to encode location information into the GHC and Haddock AST while incurring no cost for e.g. TH which doesn't need location information. These are just changes that have to happen in lock step. - - - - - 0b15be7c by Ben Gamari at 2019-11-09T13:21:33-05:00 Import isRuntimeRepVar from Type rather than TyCoRep isRuntimeRepVar is not longer exported from TyCoRep due to ghc#17441. - - - - - 091f7283 by Ben Gamari at 2019-11-10T12:47:06-05:00 Bump to GHC 8.10 - - - - - e88c71f2 by Ben Gamari at 2019-11-14T00:22:24-05:00 Merge pull request haskell/haddock#1110 from haskell/wip/T17441 Import isRuntimeRepVar from Type rather than TyCoRep - - - - - 4e0bbc17 by Ben Gamari at 2019-11-14T00:22:45-05:00 Version bumps for GHC 8.11 - - - - - 0e85ceb4 by Ben Gamari at 2019-11-15T11:59:45-05:00 Bump to GHC 8.10 - - - - - 00d6d68b by Ben Gamari at 2019-11-16T18:35:58-05:00 Bump ghc version to 8.11 - - - - - dde1fc3f by Ben Gamari at 2019-11-16T20:40:37-05:00 Drop support for base 4.13 - - - - - f52e331d by Vladislav Zavialov at 2019-11-24T13:02:28+03:00 Update Hyperlinker.Parser.classify to use ITdollar - - - - - 1ad96198 by Vladislav Zavialov at 2019-11-28T16:12:33+03:00 Remove HasSrcSpan (#17494) - - - - - 651afd70 by Herbert Valerio Riedel at 2019-12-08T12:08:16+01:00 Document error-prone conditional definition of instances This can easily trip up people if one isn't aware of it. Usually it's better to avoid this kind of conditionality especially for typeclasses for which there's an compat-package as conditional instances like these tend to fragment the ecosystem into those packages that go the extra mile to provide backward compat via those compat-packages and those that fail to do so. - - - - - b521af56 by Herbert Valerio Riedel at 2019-12-08T12:09:54+01:00 Fix build-failure regression for base < 4.7 The `$>` operator definition is available only since base-4.7 which unfortunately wasn't caught before release to Hackage (but has been fixed up by a metadata-revision) This commit introduces a `CompatPrelude` module which allows to reduce the amount of CPP by ousting it to a central location, i.e. the new `CompatPrelude` module. This pattern also tends to reduce the tricks needed to silence unused import warnings. Addresses haskell/haddock#1119 - - - - - 556c375d by Sylvain Henry at 2020-01-02T19:01:55+01:00 Fix after Iface modules renaming - - - - - bd6c53e5 by Sylvain Henry at 2020-01-07T00:48:48+01:00 hsyl20-modules-renamer - - - - - fb23713b by Ryan Scott at 2020-01-08T07:41:13-05:00 Changes for GHC#17608 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2372 - - - - - 4a4dd382 by Ryan Scott at 2020-01-25T08:08:26-05:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - e782a44d by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename PackageConfig into UnitInfo - - - - - ba3c9f05 by Sylvain Henry at 2020-01-26T02:12:37+01:00 Rename lookupPackage - - - - - ab37f9b3 by Ben Gamari at 2020-01-29T13:00:44-05:00 Merge pull request haskell/haddock#1125 from haskell/wip/T17566-take-two Changes for GHC#17566 - - - - - 3ebd5ae0 by Ryan Scott at 2020-01-31T05:56:50-05:00 Merge branch 'wip-hsyl20-package-refactor' into ghc-head - - - - - 602a747e by Richard Eisenberg at 2020-02-04T09:05:43+00:00 Echo GHC's removal of PlaceHolder module This goes with GHC's !2083. - - - - - ccfe5679 by Sylvain Henry at 2020-02-10T10:13:56+01:00 Module hierarchy: runtime (cf haskell/haddock#13009) - - - - - 554914ce by Cale Gibbard at 2020-02-10T16:10:39-05:00 Fix build of haddock in stage1 We have to use the correct version of the GHC API, but the version of the compiler itself doesn't matter. - - - - - 5b6fa2a7 by John Ericson at 2020-02-10T16:18:07-05:00 Noramlize `tested-with` fields in cabal files - - - - - e6eb3ebe by Vladislav Zavialov at 2020-02-16T13:25:26+03:00 No MonadFail/Alternative for P - - - - - 90e181f7 by Ben Gamari at 2020-02-18T14:13:47-05:00 Merge pull request haskell/haddock#1129 from obsidiansystems/wip/fix-stage1-build Fix build of haddock in stage1 - - - - - 93b64636 by Sylvain Henry at 2020-02-19T11:20:27+01:00 Modules: Driver (#13009) - - - - - da4f6c7b by Vladislav Zavialov at 2020-02-22T15:33:02+03:00 Use RealSrcSpan in InstMap - - - - - 479b1b50 by Ben Gamari at 2020-02-23T10:28:13-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 55ecacf0 by Sylvain Henry at 2020-02-25T15:18:27+01:00 Modules: Core (#13009) - - - - - 60867b3b by Vladislav Zavialov at 2020-02-28T15:53:52+03:00 Ignore the BufLoc/BufSpan added in GHC's !2516 - - - - - 1e5506d3 by Sylvain Henry at 2020-03-02T12:32:43+01:00 Modules: Core (#13009) - - - - - 6fb53177 by Richard Eisenberg at 2020-03-09T14:49:40+00:00 Changes in GHC's !1913. - - - - - 30b792ea by Ben Gamari at 2020-03-16T12:45:02-04:00 Merge pull request haskell/haddock#1130 from hsyl20/wip/hsyl20-modules-core2 Modules: Core (#13009) - - - - - cd761ffa by Sylvain Henry at 2020-03-18T15:24:00+01:00 Modules: Types - - - - - b6646486 by Ben Gamari at 2020-03-18T14:42:43-04:00 Merge pull request haskell/haddock#1133 from hsyl20/wip/hsyl20/modules/types Modules: Types - - - - - 9325d734 by Kleidukos at 2020-03-19T12:38:31-04:00 Replace the 'caption' class so that the collapsible sections are shown - - - - - 5e2bb555 by Kleidukos at 2020-03-19T12:38:31-04:00 Force ghc-8.8.3 - - - - - c6fcd0aa by Kleidukos at 2020-03-19T12:38:31-04:00 Update test fixtures - - - - - 5c849cb1 by Sylvain Henry at 2020-03-20T09:34:39+01:00 Modules: Types - - - - - 7f439155 by Alec Theriault at 2020-03-20T20:17:01-04:00 Merge branch 'ghc-8.8' into ghc-8.10 - - - - - b7904e5c by Alina Banerjee at 2020-03-20T20:24:17-04:00 Update parsing to strip whitespace from table cells (#1074) * Update parsing to strip leading & trailing whitespace from table cells * Update fixture data to disallow whitespaces at both ends in table cells * Add test case for whitespaces stripped from both ends of table cells * Update table reference test data for html tests - - - - - b9d60a59 by Alec Theriault at 2020-03-22T11:46:42-04:00 Clean up warnings * unused imports * imports of `Data.List` without import lists * missing `CompatPrelude` file in `.cabal` - - - - - 0c317dbe by Alec Theriault at 2020-03-22T18:46:54-04:00 Fix NPM security warnings This was done by calling `npm audit fix`. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. - - - - - 6e306242 by Alec Theriault at 2020-03-22T20:10:52-04:00 Tentative 2.24 release Adjusted changelogs and versions in `.cabal` files in preparation for the upcoming release bundled with GHC 8.10. - - - - - 1bfb4645 by Ben Gamari at 2020-03-23T16:40:54-04:00 Merge commit '3c2944c037263b426c4fe60a3424c27b852ea71c' into HEAD More changes from the GHC types module refactoring. - - - - - be8c6f3d by Alec Theriault at 2020-03-26T20:10:53-04:00 Update `.travis.yml` to work with GHC 8.10.1 * Regenerated the Travis file with `haskell-ci` * Beef up `.cabal` files with more `tested-with` information - - - - - b025a9c6 by Alec Theriault at 2020-03-26T20:10:53-04:00 Update README Removed some out of date links/info, added some more useful links. * badge to Hackage * update old trac link * `ghc-head` => `ghc-8.10` * `cabal new-*` is now `cabal v2-*` and it should Just Work * `--test-option='--accept'` is the way to accept testsuite output - - - - - 564d889a by Alec Theriault at 2020-03-27T20:34:33-04:00 Fix crash in `haddock-library` on unicode space Our quickcheck tests for `haddock-library` stumbled across an edge case input that was causing Haddock to crash: it was a unicode space character. The root cause of the crash is that we were implicitly assuming that if a space character was not " \t\f\v\r", it would have to be "\n". We fix this by instead defining horizontal space as: any space character that is not '\n'. Fixes haskell/haddock#1142 - - - - - 2d360ba1 by Alec Theriault at 2020-03-27T21:57:32-04:00 Disallow qualified uses of reserved identifiers This a GHC bug (https://gitlab.haskell.org/ghc/ghc/issues/14109) too, but it is a relatively easy fix in Haddock. Note that the fix must live in `haddock-api` instead of `haddock-library` because we can only really decide if an identifier is a reserved one by asking the GHC lexer. Fixes haskell/haddock#952 - - - - - 47ae22ed by Alec Theriault at 2020-03-28T13:36:25-04:00 Remove unused `Haddock.Utils` functions * removed functions in `Haddock.Utils` that were not used anywhere (or exported from the `haddock-api` package) * moved GHC-specific utils from `Haddock.Utils` to `Haddock.GhcUtils` - - - - - c0291245 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use TTG empty extensions to remove some `error`'s None of these error cases should ever have been reachable, so this is just a matter of leveraging the type system to assert this. * Use the `NoExtCon` and `noExtCon` to handle case matches for no extension constructors, instead of throwing an `error`. * Use the extension field of `HsSpliceTy` to ensure that this variant of `HsType` cannot exist in an `HsType DocNameI`. - - - - - 0aff8dc4 by Alec Theriault at 2020-03-28T13:36:25-04:00 Use `unLoc`/`noLoc` from GHC instead of `unL`/`reL` * `unL` is already defined by GHC as `unLoc` * `reL` is already defined by GHC as `noLoc` (in a safer way too!) * Condense `setOutputDir` and add a about exporting from GHC Fixes haskell/haddock#978 - - - - - bf6f2fb7 by Alec Theriault at 2020-03-28T13:36:25-04:00 Cleanup up GHC flags in `.cabal` files * enable more useful warning flags in `haddock-api`, handle the new warnings generated * remove `-fwarn-tabs` (now we'd use `-Wtabs`, but this has been in `-Wall` for a while now) - - - - - c576fbf1 by Alec Theriault at 2020-03-28T13:36:25-04:00 `haddock-library` document header level Document the fact the header level is going to always be between 1 and 6 inclusive. Along the way, I also optimized the parsing code a bit. - - - - - 71bce0ee by Alec Theriault at 2020-03-28T14:26:27-04:00 Disallow links in section headers This is quite straightforward to implement, since we already had a function `docToHtmlNoAnchors` (which we used to generate the link in the sidebar "Contents"). This breaks test `Bug387`, but that test case has aged badly: we now automatically generate anchors for all headings, so manually adding an anchor in a section makes no sense. Nested anchors are, as pointed out in haskell/haddock#1054, disallowed by the HTML standard. Fixes haskell/haddock#1054 - - - - - b461b0ed by Sylvain Henry at 2020-03-30T10:34:23+02:00 Modules: type checker - - - - - cd8cd1ee by Ben Gamari at 2020-03-31T12:45:02-04:00 Merge pull request haskell/haddock#1152 from hsyl20/wip/hsyl20/modules/tc Module renaming - - - - - 5e8f8ea7 by Felix Yan at 2020-04-01T17:58:06-07:00 Allow QuickCheck 2.14 Builds fine and all tests pass. - - - - - dc6b1633 by Sylvain Henry at 2020-04-05T16:43:44+02:00 Module renaming: amend previous patch - - - - - eee2f4ae by Ryan Scott at 2020-04-05T09:04:43-07:00 Fix haskell/haddock#1050 by filtering out invisible AppTy arguments This makes the `synifyType` case for `AppTy` more intelligent by taking into consideration the visibilities of each `AppTy` argument and filtering out any invisible arguments, as they aren't intended to be displayed in the source code. (See haskell/haddock#1050 for an example of what can happen if you fail to filter these out.) Along the way, I noticed that a special `synifyType` case for `AppTy t1 (CoercionTy {})` could be consolidated with the case below it, so I took the opportunity to tidy this up. - - - - - 23eb99e8 by Ben Gamari at 2020-04-07T11:19:58-04:00 Merge pull request haskell/haddock#1154 from hsyl20/wip/hsyl20/modules/tc Module renaming: amend previous patch - - - - - 072d994d by Ryan Scott at 2020-04-07T19:32:47-04:00 Make NoExtCon fields strict These changes are a part of a fix for [GHC#17992](https://gitlab.haskell.org/ghc/ghc/issues/17992). - - - - - d8ebf6c8 by Ignat Insarov at 2020-04-09T21:15:01-04:00 Recode Doc to Json. (#1159) * Recode Doc to Json. * More descriptive field labels. - - - - - 52df4b4e by Sylvain Henry at 2020-04-10T12:39:18+02:00 Module renaming - - - - - d9ab8ec8 by Cale Gibbard at 2020-04-14T11:43:34-04:00 Add instance of XCollectPat for DocNameI - - - - - 323d221d by Cale Gibbard at 2020-04-14T11:43:34-04:00 Rename XCollectPat -> CollectPass - - - - - 2df80867 by Alec Theriault at 2020-04-15T07:30:51-07:00 Prune docstrings that are never rendered When first creating a Haddock interface, trim `ifaceDocMap` and `ifaceArgMap` to not include docstrings that can never appear in the final output. Besides checking with GHC which names are exported, we also need to keep all the docs attached to instance declarations (it is much tougher to detect when an instance is fully private). This change means: * slightly smaller interface files (7% reduction on boot libs) * slightly less work to do processing docstrings that aren't used * no warnings in Haddock's output about private docstrings (see haskell/haddock#1070) I've tested manually that this does not affect any of the boot library generated docs (the only change in output was some small re-ordering in a handful of instance lists). This should mean no docstrings have been incorrectly dropped. - - - - - f49c90cc by Alec Theriault at 2020-04-15T07:30:51-07:00 Don't warn about missing links in miminal sigs When renaming the Haddock interface, never emit warnings when renaming a minimal signature. Also added some documention around `renameInterface`. Minimal signatures intentionally include references to potentially un-exported methods (see the discussion in haskell/haddock#330), so it is expected that they will not always have a link destination. On the principle that warnings should always be resolvable, this shouldn't produce a warning. See haskell/haddock#1070. - - - - - a9eda64d by Ben Gamari at 2020-04-17T09:27:35-04:00 Merge pull request haskell/haddock#1160 from hsyl20/wip/hsyl20/modules/systools Module renaming - - - - - f40d7879 by Cale Gibbard at 2020-04-20T11:30:38-04:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ttg-con-pat - - - - - a50e7753 by Ben Gamari at 2020-04-20T11:36:10-04:00 Merge pull request haskell/haddock#1165 from obsidiansystems/wip/ttg-con-pat Trees that Grow refactor (GHC !2553) - - - - - 6a24795c by Alec Theriault at 2020-04-21T08:06:45-07:00 Fallback to `hiDecl` when `extractDecl` fails Sometimes, the declaration being exported is a subdecl (for instance, a record accessor getting exported at the top-level). For these cases, Haddock has to find a way to produce some synthetic sensible top-level declaration. This is done with `extractDecl`. As is shown by haskell/haddock#1067, this is sometimes impossible to do just at a syntactic level (for instance when the subdecl is re-exported). In these cases, the only sensible thing to do is to try to reify a declaration based on a GHC `TyThing` via `hiDecl`. - - - - - eee1a8b7 by Sylvain Henry at 2020-04-24T15:46:05+02:00 Module structure - - - - - 50b9259c by Iñaki at 2020-04-25T18:38:11-04:00 Add support for custom section anchors (#1179) This allows to have stable anchors for groups, even if the set of groups in the documentation is altered. The syntax for setting the anchor of a group is -- * Group name #desiredAnchor# Which will produce an html anchor of the form '#g:desiredAnchor' Co-authored-by: Iñaki García Etxebarria <git at inaki.blueleaf.cc> - - - - - 4003c97a by Ben Gamari at 2020-04-26T09:35:15-04:00 Merge pull request haskell/haddock#1166 from hsyl20/wip/hsyl20/modules/utils Module structure - - - - - 5206ab60 by Sylvain Henry at 2020-04-27T16:47:39+02:00 Renamed UnitInfo fields - - - - - c32c333b by Sylvain Henry at 2020-04-27T17:32:58+02:00 UnitId has been renamed into Unit - - - - - 3e87db64 by Sylvain Henry at 2020-04-27T17:36:00+02:00 Fix for GHC.Unit.* modules - - - - - ae3323a7 by Ben Gamari at 2020-04-29T12:36:37-04:00 Merge pull request haskell/haddock#1183 from hsyl20/wip/hsyl20/unitid Refactoring of Unit code - - - - - b105564a by Artem Pelenitsyn at 2020-05-03T08:14:10+01:00 add dependency on exceptions because GHC.Exception was boiled down (ghc haskell/haddock#18075) - - - - - 9857eff3 by Zubin Duggal at 2020-05-04T18:48:25+01:00 Atomic update of NameCache in readHieFile - - - - - 86bbb226 by Sylvain Henry at 2020-05-14T16:36:27+02:00 Fix after Config module renaming - - - - - a4bbdbc2 by Gert-Jan Bottu at 2020-05-15T22:09:44+02:00 Explicit Specificity Support for Haddock - - - - - 46199daf by Ben Gamari at 2020-05-19T09:59:56-04:00 Merge pull request haskell/haddock#1192 from hsyl20/hsyl20/modules-config Fix after Config module renaming - - - - - f9a9d2ba by Gert-Jan Bottu at 2020-05-20T16:48:38-04:00 Explicit Specificity Support for Haddock - - - - - 55c5b7ea by Ben Gamari at 2020-05-21T00:32:02-04:00 Merge commit 'a8d7e66da4dcc3b242103271875261604be42d6e' into ghc-head - - - - - a566557f by Cale Gibbard at 2020-05-21T16:02:06-04:00 isBootSummary now produces a result of type IsBootInterface - - - - - ea52f905 by Zubin Duggal at 2020-05-24T17:55:48+01:00 update for hiefile-typeclass-info - - - - - 49ba7a67 by Willem Van Onsem at 2020-05-25T12:23:01-04:00 Use floor over round to calculate the percentage (#1195) If we compile documentation where only a small fraction is undocumented, it is misleading to see 100% coverage - 99% is more intuitive. Fixes haskell/haddock#1194 - - - - - c025ebf1 by Ben Gamari at 2020-05-29T14:32:42-04:00 Merge pull request haskell/haddock#1185 from obsidiansystems/boot-disambig isBootSummary now produces a result of type IsBootInterface - - - - - 74ab9415 by Ben Gamari at 2020-05-29T20:23:39-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - b40be944 by Ben Gamari at 2020-06-03T17:02:31-04:00 testsuite: Update expected output for simplified subsumption - - - - - 624be71c by Ryan Scott at 2020-06-05T12:43:23-04:00 Changes for GHC#18191 See https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3337. - - - - - fbd8f7ce by Sylvain Henry at 2020-06-08T15:31:47+02:00 Fix after unit refactoring - - - - - 743fda4d by Ben Gamari at 2020-06-09T12:09:58-04:00 Merge pull request haskell/haddock#1202 from hsyl20/wip/hsyl20/unitid-ii Fix after unit refactoring - - - - - d07a06a9 by Ryan Scott at 2020-06-13T07:16:55-04:00 Use HsForAllTelescope (GHC#18235) - - - - - 389bb60d by Ben Gamari at 2020-06-13T15:30:52-04:00 haddock: Bounds bumps for GHC 8.12 - - - - - 7a377f5f by Ben Gamari at 2020-06-17T14:53:16-04:00 Merge pull request haskell/haddock#1199 from bgamari/wip/ghc-8.12 haddock: Bounds bumps for GHC 8.12 - - - - - 9fd9e586 by Krzysztof Gogolewski at 2020-06-17T16:09:07-04:00 Adapt Haddock to LinearTypes See ghc/ghc!852. - - - - - 46fe7636 by Ben Gamari at 2020-06-18T14:20:02-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 35a3c9e2 by Zubin Duggal at 2020-06-21T21:19:18+05:30 Use functions exported from HsToCore - - - - - 8abe3928 by Ben Gamari at 2020-06-24T13:53:39-04:00 Merge pull request haskell/haddock#1204 from wz1000/wip/haddock-hstocore Use functions exported from GHC.HsToCore.Docs - - - - - 22f2c937 by Matthías Páll Gissurarson at 2020-06-26T19:07:03+02:00 Adapt Haddock for QualifiedDo - - - - - 3f6208d7 by Vladislav Zavialov at 2020-06-28T14:28:16+03:00 Handle LexicalNegation's ITprefixminus - - - - - 03a19f41 by Sylvain Henry at 2020-07-02T09:37:38+02:00 Rename hsctarget into backend - - - - - ea17ff23 by Andreas Klebinger at 2020-07-02T17:44:18+02:00 Update for UniqFM changes. - - - - - 9872f2f3 by Ben Gamari at 2020-07-09T10:39:19-04:00 Merge pull request haskell/haddock#1209 from AndreasPK/wip/typed_uniqfm Update for UniqFM changes. - - - - - 68f7b668 by Krzysztof Gogolewski at 2020-07-12T18:16:57+02:00 Sync with GHC removing {-# CORE #-} pragma See ghc ticket haskell/haddock#18048 - - - - - eb372681 by Sylvain Henry at 2020-07-20T11:41:30+02:00 Rename hscTarget into backend - - - - - fb7f78bf by Ben Gamari at 2020-07-21T12:15:25-04:00 Merge pull request haskell/haddock#1214 from hsyl20/wip/hsyl20/hadrian/ncg Rename hscTarget into backend - - - - - 1e8f5b56 by Ben Gamari at 2020-07-23T09:11:50-04:00 Merge commit '904dce0cafe0a241dd3ef355775db47fc12f434d' into ghc-head - - - - - d8fd1775 by Zubin Duggal at 2020-07-23T18:46:40+05:30 Update for modular ping pong - - - - - 8416f872 by Ben Gamari at 2020-07-23T09:35:03-04:00 Merge pull request haskell/haddock#1200 from wz1000/wip/wz1000-modular-ping-pong Modular ping pong - - - - - a24a8577 by Ben Gamari at 2020-07-28T15:23:36-04:00 Bump GHC version to 9.0 - - - - - 6a51c9dd by Sylvain Henry at 2020-08-05T18:47:05+02:00 Fix after Outputable refactoring - - - - - c05e1c99 by Ben Gamari at 2020-08-10T14:41:41-04:00 Merge pull request haskell/haddock#1223 from hsyl20/wip/hsyl20/dynflags/exception Fix after Outputable refactoring - - - - - d964f15b by Sylvain Henry at 2020-08-12T11:58:49+02:00 Fix after HomeUnit - - - - - 8e6d5b23 by Ben Gamari at 2020-08-12T14:25:30-04:00 Merge pull request haskell/haddock#1225 from hsyl20/wip/hsyl20/plugins/homeunit Fix after HomeUnit - - - - - 8c7880fe by Sylvain Henry at 2020-08-17T14:13:29+02:00 Remove Ord FastString instance - - - - - 8ea410db by Alex Biehl at 2020-08-19T10:56:32+02:00 Another round of `npm audit fix` (#1228) This should shut down the warnings on Github. Note that the security issues seem to have been entirely in the build dependencies, since the output JS has not changed. Last NPM dependency audit happend in d576b2327e2bc117f912fe0a9d595e9ae62614e0 Co-authored-by: Alex Biehl <alex.biehl at target.com> - - - - - 7af6e2a8 by Ben Gamari at 2020-08-31T13:59:34-04:00 Merge pull request haskell/haddock#1226 from hsyl20/wip/hsyl20/fs_ord Remove Ord FastString instance - - - - - ffbc8702 by Alan Zimmerman at 2020-09-07T21:47:41+01:00 Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - a93f1268 by Alan Zimmerman at 2020-09-07T23:11:38+01:00 Merge pull request haskell/haddock#1232 from haskell/wip/T18639-remove-generated-pragma, Match GHC for haskell/haddock#18639, remove GENERATED pragma - - - - - 1f605d50 by Ben Gamari at 2020-09-14T18:30:01-04:00 Bump GHC version to 9.1 - - - - - 6599df62 by Vladislav Zavialov at 2020-09-18T14:05:15+03:00 Bump base upper bound to 4.16 - - - - - a01b3c43 by Ben Gamari at 2020-09-22T15:41:48-04:00 Update hypsrc-test for QuickLook This appears to be a spurious change. - - - - - e9cc6cac by Vladislav Zavialov at 2020-09-26T21:00:12+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 30e3ca7c by Sylvain Henry at 2020-09-29T11:18:32-04:00 Update for parser (#1234) - - - - - b172f3e3 by Vladislav Zavialov at 2020-09-30T01:01:30+03:00 Updates for the new linear types syntax: a %p -> b - - - - - 0b9c08d3 by Sylvain Henry at 2020-09-30T11:02:33+02:00 Adapt to GHC parser changes - - - - - b9540b7a by Sylvain Henry at 2020-10-12T09:13:38-04:00 Don't pass the HomeUnitId (#1239) - - - - - 34762e80 by HaskellMouse at 2020-10-13T12:58:04+03:00 Changed tests due to unification of `Nat` and `Natural` in the follwing merge request: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/3583 - - - - - 256f86b6 by Vladislav Zavialov at 2020-10-15T10:48:03+03:00 Add whitespace in: map ($ v) - - - - - 4a3f711b by Alan Zimmerman at 2020-10-19T08:57:27+01:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled - - - - - 072cdd21 by Alan Zimmerman at 2020-10-21T14:48:28-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 9e09a445 by Alan Zimmerman at 2020-10-21T23:53:34-04:00 Match GHC, adding IsUnicodeSyntax field to HsFunTy and HsScaled (cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1) - - - - - 636d7de3 by Sylvain Henry at 2020-10-26T14:31:54-04:00 GHC.Driver.Types refactoring (#1242) - - - - - a597f000 by Ryan Scott at 2020-10-29T04:18:05-04:00 Adapt to the removal of Hs{Boxed,Constraint}Tuple See ghc/ghc!4097 and GHC#18723. - - - - - b96660fb by Ryan Scott at 2020-10-30T04:53:05-04:00 Adapt to HsConDecl{H98,GADT}Details split Needed for GHC#18844. - - - - - c287d82c by Ryan Scott at 2020-10-30T19:35:59-04:00 Adapt to HsOuterTyVarBndrs These changes accompany ghc/ghc!4107, which aims to be a fix for haskell/haddock#16762. - - - - - a34c31a1 by Ryan Scott at 2020-11-13T13:38:34-05:00 Adapt to splitPiTysInvisible being renamed to splitInvisPiTys This is a part of !4434, a fix for GHC#18939. - - - - - 66ea459d by Sylvain Henry at 2020-11-16T10:59:30+01:00 Fix after Plugins moved into HscEnv - - - - - 508556d8 by Ben Gamari at 2020-11-18T15:47:40-05:00 Merge pull request haskell/haddock#1253 from hsyl20/wip/hsyl20/plugins/hscenv Fix after Plugins moved into HscEnv - - - - - 620fec1a by Andreas Klebinger at 2020-11-24T20:51:59+01:00 Update for changes in GHC's Pretty - - - - - 01cc13ab by Richard Eisenberg at 2020-11-25T23:18:35-05:00 Avoid GHC#18932. - - - - - 8d29ba21 by Cale Gibbard at 2020-11-25T23:18:35-05:00 Add type arguments to PrefixCon - - - - - 414d5f87 by Sylvain Henry at 2020-11-30T17:06:04+01:00 DynFlags's unit fields moved to HscEnv - - - - - e356668c by Ben Gamari at 2020-11-30T11:11:37-05:00 Merge pull request haskell/haddock#1258 from hsyl20/wip/hsyl20/hscenv/unitstate Unit fields moved from DynFlags to HscEnv - - - - - 7cf552f1 by Ben Gamari at 2020-12-03T10:31:27-05:00 Merge pull request haskell/haddock#1257 from AndreasPK/wip/andreask/opt_dumps Update for changes in GHC's Pretty - - - - - fc0871c3 by Veronika Romashkina at 2020-12-08T16:35:33+01:00 Fix docs links from Darcs to GitHub in intro (#1262) - - - - - 7059e808 by Veronika Romashkina at 2020-12-08T16:36:16+01:00 Use gender neutral word in docs (#1260) - - - - - 1b16e5ee by Maximilian Tagher at 2020-12-08T16:40:03+01:00 Allow scrolling search results (#1235) Closes https://github.com/haskell/haddock/issues/1231 - - - - - 8a118c01 by dependabot[bot] at 2020-12-08T16:40:25+01:00 Bump bl from 1.2.2 to 1.2.3 in /haddock-api/resources/html (#1255) Bumps [bl](https://github.com/rvagg/bl) from 1.2.2 to 1.2.3. - [Release notes](https://github.com/rvagg/bl/releases) - [Commits](https://github.com/rvagg/bl/compare/v1.2.2...v1.2.3) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - c89ff587 by Xia Li-yao at 2020-12-08T16:42:17+01:00 Allow more characters in anchor following module reference (#1220) - - - - - 14af7d64 by Xia Li-yao at 2020-12-08T16:43:05+01:00 Add dangling changes from branches ghc-8.6 and ghc-8.8 (#1243) * Fix multiple typos and inconsistencies in doc/markup.rst Note: I noticed some overlap with haskell/haddock#1112 from @wygulmage and haskell/haddock#1081 from @parsonsmatt after creating these proposed changes - mea culpa for not looking at the open PRs sooner. * Fix haskell/haddock#1113 If no Signatures, no section of index.html * Change the formatting of missing link destinations The current formatting of the missing link destination does not really help user to understand the reasons of the missing link. To address this, I've changed the formatting in two ways: - the missing link symbol name is now fully qualified. This way you immediately know which haskell module cannot be linked. It is then easier to understand why this module does not have documentation (hidden module or broken documentation). - one line per missing link, that's more readable now that symbol name can be longer due to qualification. For example, before haddock was listing missing symbol such as: ``` could not find link destinations for: Word8 Word16 mapMaybe ``` Now it is listed as: ``` could not find link destinations for: - Data.Word.Word8 - Data.Word.Word16 - Data.Maybe.mapMaybe ``` * Add `--ignore-link-symbol` command line argument This argument can be used multiples time. A missing link to a symbol listed by `--ignore-link-symbol` won't trigger "missing link" warning. * Forbid spaces in anchors (#1148) * Improve error messages with context information (#1060) Co-authored-by: Matt Audesse <matt at mattaudesse.com> Co-authored-by: Mike Pilgrem <mpilgrem at users.noreply.github.com> Co-authored-by: Guillaume Bouchard <guillaume.bouchard at tweag.io> Co-authored-by: Pepe Iborra <pepeiborra at gmail.com> - - - - - 89e3af13 by tomjaguarpaw at 2020-12-08T18:00:04+01:00 Enable two warnings (#1245) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - c3320f8d by Willem Van Onsem at 2020-12-08T18:26:55+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 685df308 by Alex Biehl at 2020-12-08T20:06:26+01:00 Changes for GHC#17566 See https://gitlab.haskell.org/ghc/ghc/merge_requests/2469 - - - - - be3ec3c0 by Alex Biehl at 2020-12-08T20:06:26+01:00 Import intercalate - - - - - 32c33912 by Matthías Páll Gissurarson at 2020-12-08T21:15:30+01:00 Adapt Haddock for QualifiedDo - - - - - 31696088 by Alex Biehl at 2020-12-08T22:06:02+01:00 Fix haddock-library tests - - - - - fbc0998a by Alex Biehl at 2020-12-08T23:08:23+01:00 Move to GitHub CI (#1266) * Initial version of ci.yml This is a straight copy from Dmitrii Kovanikov's blog post at https://kodimensional.dev/github-actions. Will adapt to haddock in successive commits. * Delete .travis.yml * Modify to only test on ghc-8.10.{1,2} * Use actions/setup-haskell at v1.1.4 * Relax QuickCheck bound on haddock-api * Remove stack matrix for now * Nail down to ghc-8.10 branch for now * Pin index state to 2020-12-08T20:13:44Z for now * Disable macOS and Windows tests for now for speed up - - - - - 5b946b9a by tomjaguarpaw at 2020-12-10T19:01:41+01:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - bc5a408f by dependabot[bot] at 2020-12-10T19:02:16+01:00 Bump ini from 1.3.5 to 1.3.7 in /haddock-api/resources/html (#1269) Bumps [ini](https://github.com/isaacs/ini) from 1.3.5 to 1.3.7. - [Release notes](https://github.com/isaacs/ini/releases) - [Commits](https://github.com/isaacs/ini/compare/v1.3.5...v1.3.7) Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - d02995f1 by Andrew Martin at 2020-12-14T16:48:40-05:00 Update for boxed rep - - - - - a381aeff by Ben Gamari at 2020-12-15T15:13:30-05:00 Revert "Enable two warnings (#1245) (#1268)" As this does not build on GHC `master`. This reverts commit 7936692badfe38f23ae95b51fb7bd7c2ff7e9bce. - - - - - a63c0a9e by Ben Gamari at 2020-12-15T15:17:59-05:00 Revert "Update for boxed rep" This reverts commit 4ffb30d8b637ccebecc81ce610f0af451ac8088d. - - - - - 53bfbb29 by Ben Gamari at 2020-12-15T15:37:24-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - bae76a30 by Ben Gamari at 2020-12-16T02:44:42+00:00 Update output for nullary TyConApp optimisation (ghc/ghc!2952) - - - - - 4b733b57 by Krzysztof Gogolewski at 2020-12-16T20:03:14+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. - - - - - ee463bd3 by Ryan Scott at 2020-12-16T16:55:23-05:00 Adapt to HsCoreTy (formerly NewHsTypeX) becoming a type synonym Needed for !4417, the fix for GHC#15706 and GHC#18914. - - - - - ed0b02f8 by tomjaguarpaw at 2020-12-19T10:17:19+00:00 Enable two warnings (#1245) (#1268) because they will be soon be added to -Wall. See https://gitlab.haskell.org/ghc/ghc/-/issues/15656 - - - - - d80bf8f5 by Sylvain Henry at 2020-12-21T10:09:25+01:00 Fix after binder collect changes - - - - - bf4c9d32 by Adam Gundry at 2020-12-23T21:35:01+00:00 Adapt to changes to GlobalRdrElt and AvailInfo Needed for ghc/ghc!4467 - - - - - 37736c4c by John Ericson at 2020-12-28T12:27:02-05:00 Support a new ghc --make node type for parallel backpack upsweep - - - - - 717bdeac by Vladislav Zavialov at 2020-12-29T10:50:02+03:00 Inline and fix getGADTConTypeG The getGADTConTypeG used HsRecTy, which is at odds with GHC issue haskell/haddock#18782. I noticed that getGADTConTypeG was only used in the Hoogle backend. Interestingly, when handling H98 constructors, Hoogle converts RecCon to PrefixCon (see Haddock.Backends.Hoogle.ppCtor). So I changed getGADTConTypeG to handle RecConGADT in the same manner as PrefixConGADT, and after this simplification moved it into the 'where' clause of ppCtor, to the only place where it is used. The practical effect of this change is as follows. Consider this example: data TestH98 = T98 { bar::Int } data TestGADT where TG :: { foo :: Int } -> TestGADT Before this patch, haddock --hoogle used to produce: T98 :: Int -> TestH98 [TG] :: {foo :: Int} -> TestGADT Notice how the record syntax was discarded in T98 but not TG. With this patch, we always produce signatures without record syntax: T98 :: Int -> TestH98 [TG] :: Int -> TestGADT I suspect this might also be a bugfix, as currently Hoogle doesn't seem to render GADT record constructors properly. - - - - - cb1b8c56 by Andreas Abel at 2020-12-30T21:12:37+01:00 Build instructions: haddock-library and -api first! - - - - - b947f6ad by Ben Gamari at 2020-12-31T13:04:19-05:00 Merge pull request haskell/haddock#1281 from obsidiansystems/wip/backpack-j Changes to support -j with backpack - - - - - 120e1cfd by Hécate Moonlight at 2021-01-04T19:54:58+01:00 Merge pull request haskell/haddock#1282 from andreasabel/master Build instructions: haddock-library and -api first! - - - - - fd45e41a by Ben Gamari at 2021-01-05T16:14:31-05:00 Merge remote-tracking branch 'origin/ghc-8.10' into ghc-9.0 - - - - - b471bdec by Ben Gamari at 2021-01-05T16:23:02-05:00 Merge commit '1e56f63c3197e7ca1c1e506e083c2bad25d08793' into ghc-9.0 - - - - - 81cdbc41 by Alex Biehl at 2021-01-09T12:14:41+01:00 Prepare Haddock for being a GHC Plugin - - - - - b646d952 by Alex Biehl at 2021-01-09T12:14:41+01:00 Make Haddock a GHC Plugin - - - - - cc044674 by Alex Biehl at 2021-01-09T12:14:41+01:00 Add -j[n] CLI param to Haddock executable It translates to `--ghcopt=-j[n]` - - - - - 84a04073 by Alex Biehl at 2021-01-09T12:14:41+01:00 Abstract Monad for interface creation I found that when running as a plugin the lookupName function (which runs in Ghc monad) does not work correctly from the typeCheckResultAction hook. Instead, we abstracted the monad used when creating interfaces, so that access to GHC session specific parts is explicit and so that the TcM can provide their (correct) implementation of lookupName. - - - - - 5be2c4f7 by Alex Biehl at 2021-01-09T12:14:41+01:00 Accept tests - - - - - 8cefee9d by Alex Biehl at 2021-01-09T16:10:47+01:00 Add missing dependency for mtl - - - - - 3681f919 by Ben Gamari at 2021-01-13T18:39:25-05:00 Merge remote-tracking branch 'origin/ghc-9.0' into ghc-head - - - - - 33c6b152 by Hécate Moonlight at 2021-01-14T16:04:20+01:00 Merge pull request haskell/haddock#1273 from hsyl20/wip/hsyl20/arrows Fix after binder collect changes - - - - - 70d13e8e by Joachim Breitner at 2021-01-22T19:03:45+01:00 Make haddock more robust to changes to the `Language` data type With the introduction of GHC2021, the `Languages` data type in GHC will grow. In preparation of that (and to avoid changing haddock with each new language), this change makes the code handle extensions to that data type gracefully. (cherry picked from commit c341dd7c9c3fc5ebc83a2d577c5a726f3eb152a5) - - - - - 7d6dd57a by John Ericson at 2021-01-22T22:02:02+00:00 Add `NoGhcTc` instance now that it's not closed - - - - - e5fdaf0a by Alan Zimmerman at 2021-01-23T22:57:44+00:00 Merge pull request haskell/haddock#1293 from obsidiansystems/wip/fix-18936 Add `NoGhcTc` instance now that it's not closed - - - - - 989a1e05 by Oleg Grenrus at 2021-01-24T16:11:46+03:00 Add import list to Data.List - - - - - 368e144a by Ben Gamari at 2021-01-28T22:15:48+01:00 Adapt to "Make PatSyn immutable" - - - - - abe66c21 by Alfredo Di Napoli at 2021-02-01T08:05:35+01:00 Rename pprLogErrMsg to new name - - - - - e600e75c by Hécate Moonlight at 2021-02-05T14:53:00+01:00 Move CI to ghc-9.0 - - - - - dd492961 by Vladislav Zavialov at 2021-02-05T14:53:00+01:00 Update cabal.project and README build instructions - - - - - 31bd292a by Hécate Moonlight at 2021-02-05T15:03:56+01:00 Merge pull request haskell/haddock#1296 from Kleidukos/ghc-9.0 Merge the late additions to ghc-8.10 into ghc-9.0 - - - - - 6388989e by Vladislav Zavialov at 2021-02-05T17:41:57+03:00 Cleanup: fix build warnings - - - - - f99407ef by Daniel Rogozin at 2021-02-05T18:11:48+03:00 type level characters support for haddock (required for haskell/haddock#11342) - - - - - d8c6b26f by Hécate Moonlight at 2021-02-05T17:44:50+01:00 Add a CONTRIBUTING.md file - - - - - 6a01ad98 by Hécate Moonlight at 2021-02-05T17:58:16+01:00 Merge pull request haskell/haddock#1312 from Kleidukos/proper-branch-etiquette Add a CONTRIBUTING.md file - - - - - 955eecc4 by Vladislav Zavialov at 2021-02-05T20:29:00+03:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into ghc-head - - - - - 47b3d6ab by Hécate Moonlight at 2021-02-05T19:09:38+01:00 Amend the CONTRIBUTING.md file - - - - - 23de6137 by Hécate Moonlight at 2021-02-05T19:16:49+01:00 Merge pull request haskell/haddock#1313 from Kleidukos/amend-contributing Amend the CONTRIBUTING.md file - - - - - 69026b59 by Krzysztof Gogolewski at 2021-02-05T23:05:56+01:00 Display linear/multiplicity arrows correctly (#1238) Previously we were ignoring multiplicity and displayed a %1 -> b as a -> b. (cherry picked from commit b4b4d896d2d68d6c48e7db7bfe95c185ca0709cb) - - - - - ea026b78 by Oleg Grenrus at 2021-02-06T17:14:45+01:00 Add import list to Data.List - - - - - 5204326f by Hécate Moonlight at 2021-02-06T17:15:44+01:00 Merge pull request haskell/haddock#1316 from Kleidukos/explicit-imports-to-data-list Add import list to Data.List - - - - - 1f4d2136 by Ben Gamari at 2021-02-06T11:53:31-05:00 Merge remote-tracking branch 'origin/ghc-head' into wip/ghc-head-merge - - - - - 13f0d09a by Ben Gamari at 2021-02-06T11:53:45-05:00 Fix partial record selector warning - - - - - 5c115f7e by Ben Gamari at 2021-02-06T11:55:52-05:00 Merge commit 'a917dfd29f3103b69378138477514cbfa38558a9' into wip/ghc-head-merge - - - - - b6fd8b75 by Ben Gamari at 2021-02-06T12:01:31-05:00 Merge commit '41964cb2fd54b5a10f8c0f28147015b7d5ad2c02' into wip/ghc-head-merge - - - - - a967194c by Ben Gamari at 2021-02-06T18:30:35-05:00 Merge branch 'wip/ghc-head-merge' into ghc-head - - - - - 1f4c3a91 by MorrowM at 2021-02-07T01:52:33+02:00 Fix search div not scrolling - - - - - 684b1287 by Iñaki García Etxebarria at 2021-02-07T16:13:04+01:00 Add support for labeled module references Support a markdown-style way of annotating module references. For instance -- | [label]("Module.Name#anchor") will create a link that points to the same place as the module reference "Module.Name#anchor" but the text displayed on the link will be "label". - - - - - bdb55a5d by Hécate Moonlight at 2021-02-07T16:18:10+01:00 Merge pull request haskell/haddock#1319 from alexbiehl/alex/compat Backward compat: Add support for labeled module references - - - - - 6ca70991 by Hécate Moonlight at 2021-02-07T16:21:29+01:00 Merge pull request haskell/haddock#1314 from tweag/show-linear-backport Backport haskell/haddock#1238 (linear types) to ghc-9.0 - - - - - d9d73298 by Alex Biehl at 2021-02-07T17:46:25+01:00 Remove dubious parseModLink Instead construct the ModLink value directly when parsing. - - - - - 33b4d020 by Hécate Moonlight at 2021-02-07T17:52:05+01:00 Merge pull request haskell/haddock#1320 from haskell/alex/fix Remove dubious parseModLink - - - - - 54211316 by Hécate Moonlight at 2021-02-07T18:12:07+01:00 Merge pull request haskell/haddock#1318 from MorrowM/ghc-9.0 Fix search div not scrolling - - - - - 19db679e by alexbiehl-gc at 2021-02-07T18:14:46+01:00 Merge pull request haskell/haddock#1317 from bgamari/wip/ghc-head-merge Merge ghc-8.10 into ghc-head - - - - - 6bc1e9e4 by Willem Van Onsem at 2021-02-07T18:25:30+01:00 simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - c8537cf8 by alexbiehl-gc at 2021-02-07T18:30:40+01:00 Merge pull request haskell/haddock#1322 from haskell/alex/forward-port simplify calculating percentages fixing haskell/haddock#1194 (#1236) - - - - - 2d47ae4e by alexbiehl-gc at 2021-02-07T18:39:59+01:00 Merge branch 'ghc-head' into ghc-9.0 - - - - - 849e4733 by Hécate Moonlight at 2021-02-07T18:43:19+01:00 Merge pull request haskell/haddock#1321 from Kleidukos/ghc-9.0 Merge ghc-9.0 into ghc-head - - - - - ee6095d7 by Sylvain Henry at 2021-02-08T11:36:38+01:00 Update for Logger - - - - - 4ad688c9 by Alex Biehl at 2021-02-08T18:11:24+01:00 Merge pull request haskell/haddock#1310 from hsyl20/wip/hsyl20/logger2 Logger refactoring - - - - - 922a9e0e by Ben Gamari at 2021-02-08T12:54:33-05:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - 991649d2 by Sylvain Henry at 2021-02-09T10:55:17+01:00 Fix to build with HEAD - - - - - a8348dc2 by Hécate Moonlight at 2021-02-09T10:58:51+01:00 Merge pull request haskell/haddock#1327 from hsyl20/wip/hsyl20/logger2 Fix to build with HEAD - - - - - 0abdbca6 by Fendor at 2021-02-09T20:06:15+01:00 Add UnitId to Target record - - - - - d5790a0e by Alex Biehl at 2021-02-11T10:32:32+01:00 Stable sort for (data/newtype) instances - - - - - 8e6036f5 by Alex Biehl at 2021-02-11T10:32:32+01:00 Also make TyLit deterministic - - - - - f76d2945 by Hécate Moonlight at 2021-02-11T11:00:31+01:00 Merge pull request haskell/haddock#1329 from hsyl20/hsyl20/stabe_iface Stable sort for instances - - - - - 5e0469ea by Oleg Grenrus at 2021-02-14T15:28:15+02:00 Add import list to Data.List in Haddock.Interface.Create - - - - - fa57cd24 by Hécate Moonlight at 2021-02-14T17:19:27+01:00 Merge pull request haskell/haddock#1331 from phadej/more-explicit-data-list-imports Add import list to Data.List in Haddock.Interface.Create - - - - - f0cd629c by Hécate Moonlight at 2021-02-21T00:22:01+01:00 Merge pull request haskell/haddock#1311 from fendor/wip/add-targetUnitId-to-target Add UnitId to Target record - - - - - 674ef723 by Joachim Breitner at 2021-02-22T10:39:18+01:00 html-test: Always set language from ghc-9.2 on, the “default” langauge of GHC is expected to change more wildly. To prepare for that (and unblock https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4853), this sets the language for all the test files to `Haskell2010`. This should insolate this test suite against changes to the default. Cherry-picked from https://github.com/haskell/haddock/pull/1341 - - - - - f072d623 by Hécate Moonlight at 2021-02-22T10:56:51+01:00 Merge pull request haskell/haddock#1342 from nomeata/joachim/haskell2010-in-tests-ghc-head html-test: Always set language - - - - - caebbfca by Hécate Moonlight at 2021-02-22T11:53:07+01:00 Clean-up of Interface and Interface.Create's imports and pragmata - - - - - f6caa89d by Hécate Moonlight at 2021-02-22T11:54:58+01:00 Merge pull request haskell/haddock#1345 from Kleidukos/head/fix-interface-imports [ghc-head] Clean-up of Interface and Interface.Create's imports and pragmata - - - - - 7395c9cb by Hécate Moonlight at 2021-02-22T18:44:57+01:00 Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 6e9fb5d5 by Hécate Moonlight at 2021-02-22T18:45:28+01:00 Merge pull request haskell/haddock#1348 from Kleidukos/head/explicit-imports-interface Explicit imports for Haddock.Interface and Haddock.Interface.Create - - - - - 9198b118 by Alan Zimmerman at 2021-02-22T20:04:24+00:00 Context becomes a Maybe in the GHC AST This prevents noLoc's appearing in the ParsedSource. Match the change in GHC. - - - - - 0af20f64 by Hécate Moonlight at 2021-02-23T12:36:12+01:00 Fix the call-site of guessTarget in Interface.hs Explicit the imports from GHC.HsToCore.Docs - - - - - b7886885 by Hécate Moonlight at 2021-02-23T12:37:54+01:00 Merge pull request haskell/haddock#1349 from Kleidukos/fix-interface-guesstarget-call Fix the call-site of guessTarget in Interface.hs - - - - - 9cf041ba by Sylvain Henry at 2021-02-24T11:08:20+01:00 Fix haddockHypsrcTest output in ghc-head - - - - - b194182a by Hécate Moonlight at 2021-02-24T11:12:36+01:00 Merge pull request haskell/haddock#1351 from hsyl20/wip/hsyl20/fix-head Fix haddockHypsrcTest output in ghc-head - - - - - 3ce8b375 by Shayne Fletcher at 2021-03-06T09:55:03-05:00 Add ITproj to parser - - - - - d2abf762 by Ben Gamari at 2021-03-06T19:26:49-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - a0f6047d by Andrew Martin at 2021-03-07T11:25:23-05:00 Update for boxed rep - - - - - 6f63c99e by Ben Gamari at 2021-03-10T13:20:21-05:00 Update for "FastString: Use FastMutInt instead of IORef Int" - - - - - e13f01df by Luke Lau at 2021-03-10T15:38:40-05:00 Implement template-haskell's putDoc This catches up to GHC using the new extractTHDocs function, which returns documentation added via the putDoc function (provided it was compiled with Opt_Haddock). Since it's already a map from names -> docs, there's no need to do traversal etc. It also matches the change from the argument map being made an IntMap rather than a Map Int - - - - - 89263d94 by Alan Zimmerman at 2021-03-15T17:15:26+00:00 Match changes in GHC AST for in-tree API Annotations As landed via https://gitlab.haskell.org/ghc/ghc/-/merge_requests/2418 - - - - - 28db1934 by Alan Zimmerman at 2021-03-15T20:40:09+00:00 Change some type family test results. It is not clear to me whether the original was printing incorrectly (since we did not have the TopLevel flag before now), or if this behaviour is expected. For the time being I am assuming the former. - - - - - 7c11c989 by Sylvain Henry at 2021-03-22T10:05:19+01:00 Fix after NameCache changes - - - - - addbde15 by Sylvain Henry at 2021-03-22T10:05:19+01:00 NameCache doesn't store a UniqSupply anymore - - - - - 15ec6cec by Ben Gamari at 2021-03-22T17:53:44-04:00 Bump GHC version to 9.2 - - - - - dbd6aa63 by Hécate Moonlight at 2021-03-24T14:28:36+01:00 Merge pull request haskell/haddock#1365 from hsyl20/wip/hsyl20/iface1 NameCache refactoring - - - - - 2d32da7e by Oleg Grenrus at 2021-03-27T01:12:00+02:00 Specialization of Data.List - - - - - 32b84fa6 by Fendor at 2021-03-27T10:50:17+01:00 Add UnitId to Target record This way we always know to which home-unit a given target belongs to. So far, there only exists a single home-unit at a time, but it enables having multiple home-units at the same time. - - - - - 54bf9f0e by Hécate Moonlight at 2021-03-28T14:08:35+02:00 Merge pull request haskell/haddock#1368 from fendor/target-unit-id-revert Add UnitId to Target record - - - - - 7dea168a by Alan Zimmerman at 2021-03-29T08:45:52+01:00 EPA : Rename ApiAnn to EpAnn - - - - - 72967f65 by Alfredo Di Napoli at 2021-03-29T09:47:01+02:00 pprError changed name in GHC - - - - - 4bc61035 by Alan Zimmerman at 2021-03-29T16:16:27-04:00 EPA : Rename ApiAnn to EpAnn - - - - - 108d031d by Ben Gamari at 2021-03-29T18:49:36-04:00 Merge commit '36418c4f70d7d2b179a77925b3ad5caedb08c9b5' into HEAD - - - - - 1444f700 by Ben Gamari at 2021-03-31T09:18:39-04:00 Merge pull request haskell/haddock#1370 from adinapoli/wip/adinapoli-diag-reason-severity Rename pprError to mkParserErr - - - - - d3087b79 by Ben Gamari at 2021-03-31T11:34:17-04:00 Merge commit 'd8d8024ad6796549a8d3b5512dabf3288d14e30f' into ghc-head - - - - - 170b79e9 by Ben Gamari at 2021-03-31T12:24:56-04:00 Merge remote-tracking branch 'upstream/ghc-head' into ghc-head - - - - - db0d6bae by Ben Gamari at 2021-04-10T09:34:35-04:00 Bump GHC version to 9.3 - - - - - a9f2c421 by Alan Zimmerman at 2021-04-19T18:26:46-04:00 Update for EPA changes in GHC (cherry picked from commit cafb48118f7c111020663776845897e225607b41) - - - - - 1ee4b7c7 by Sylvain Henry at 2021-05-11T10:00:06+02:00 Removal of HsVersions.h (#1388) * Update for EPA changes in GHC * Account for HsVersions.h removal Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - 79e819e9 by Hécate Moonlight at 2021-05-11T10:14:47+02:00 Revert "Removal of HsVersions.h (#1388)" This reverts commit 72118896464f94d81f10c52f5d9261efcacc57a6. - - - - - 3dbd3f8b by Alan Zimmerman at 2021-05-11T10:15:17+02:00 Update for EPA changes in GHC - - - - - 2ce80c17 by Sylvain Henry at 2021-05-11T10:15:19+02:00 Account for HsVersions.h removal - - - - - 00e4c918 by Christiaan Baaij at 2021-05-13T08:21:56+02:00 Add Haddock support for the OPAQUE pragma (#1380) - - - - - 8f9049b2 by Hécate Moonlight at 2021-05-13T08:40:22+02:00 fixup! Use GHC 9.2 in CI runner - - - - - 27ddec38 by Alan Zimmerman at 2021-05-13T22:51:20+01:00 EPA: match changes from GHC T19834 - - - - - f8a1d714 by Felix Yan at 2021-05-14T17:10:04+02:00 Allow hspec 2.8 (#1389) All tests are passing. - - - - - df44453b by Divam Narula at 2021-05-20T15:42:42+02:00 Update ref, the variables got renamed. (#1391) This is due to ghc/ghc!5555 which caused a change in ModDetails in case of NoBackend. Now the initModDetails is used to recreate the ModDetails from interface and in-memory ModDetails is not used. - - - - - e46bfc87 by Alan Zimmerman at 2021-05-20T19:05:09+01:00 Remove Maybe from HsQualTy Match changes in GHC for haskell/haddock#19845 - - - - - 79bd7b62 by Shayne Fletcher at 2021-05-22T08:20:39+10:00 FieldOcc: rename extFieldOcc to foExt - - - - - 6ed68c74 by Ben Gamari at 2021-05-21T22:29:30-04:00 Merge commit '3b6a8774bdb543dad59b2618458b07feab8a55e9' into ghc-head - - - - - f9a02d34 by Alfredo Di Napoli at 2021-05-24T13:53:00+02:00 New Parser diagnostics interface - - - - - 392807d0 by Ben Gamari at 2021-05-24T09:57:40-04:00 Merge pull request haskell/haddock#1394 from adinapoli/wip/adinapoli-align-ps-messages Align Haddock to use the new Parser diagnostics interface - - - - - 33023cd8 by Ben Gamari at 2021-05-24T11:19:16-04:00 Revert "Add Haddock support for the OPAQUE pragma (#1380)" This reverts commit a1337c599ef7720b0482a25c55f11794112496dc. The GHC patch associated with this change is not yet ready to be merged. - - - - - 8c005af7 by Ben Simms at 2021-05-28T07:56:20+02:00 CI configuration for ghc-head (#1395) - - - - - 1e947612 by Hécate Moonlight at 2021-05-28T12:27:35+02:00 Use GHC 9.2 in CI runner (#1378) - - - - - e6fa10ab by CGenie at 2021-05-31T09:02:13+02:00 Add page about common errors (#1396) * Update index.rst Common errors page * Create common-errors.rst * Update common-errors.rst * Use GHC 9.2 in CI runner (#1378) * [haddock-api] remove .hspec-failures Co-authored-by: Hécate Moonlight <Kleidukos at users.noreply.github.com> - - - - - abc72a8d by Sylvain Henry at 2021-06-01T10:02:06+02:00 Adapt Haddock to Logger and Parser changes (#1399) - - - - - 91373656 by Zubin Duggal at 2021-06-01T20:45:10+02:00 Update haddockHypSrc tests since we now compute slighly more type info (#1397) - - - - - ed712822 by Marcin Szamotulski at 2021-06-02T08:54:33+02:00 Added myself to contributors - - - - - 49fdbcb7 by Marcin Szamotulski at 2021-06-02T08:57:24+02:00 Document multi component support - - - - - 9ddc8d7d by Hécate Moonlight at 2021-06-02T09:35:55+02:00 Merge pull request haskell/haddock#1379 from coot/coot/document-multi-component-support Document multi component support - - - - - 585b5c5e by Ben Simms at 2021-06-02T19:46:54+02:00 Update CONTRIBUTING.md (#1402) - - - - - 1df4a605 by Ben Simms at 2021-06-02T19:47:14+02:00 Update CONTRIBUTING.md (#1403) - - - - - 58ea43d2 by sheaf at 2021-06-02T22:09:06+02:00 Update Haddock Bug873 to account for renaming - - - - - c5d0ab23 by Vladislav Zavialov at 2021-06-10T13:35:42+03:00 HsToken in FunTy, RecConGADT - - - - - 1ae2f40c by Hécate Moonlight at 2021-06-11T11:19:09+02:00 Update the CI badges - - - - - 6fdc4de2 by Sylvain Henry at 2021-06-28T19:21:17+02:00 Fix mkParserOpts (#1411) - - - - - 18201670 by Alfredo Di Napoli at 2021-07-05T07:55:12+02:00 Rename getErrorMessages Lexer import This commit renames the Lexer import in `Hyperlinker.Parser` from `getErrorMessages` to `getPsErrorMessages` to eliminate the ambiguity with the `getErrorMessages` function defined in `GHC.Types.Error`. - - - - - 23173ca3 by Ben Gamari at 2021-07-07T11:31:44-04:00 Merge pull request haskell/haddock#1413 from adinapoli/wip/adinapoli-issue-19920 Rename getErrorMessages Lexer import - - - - - b3dc4ed8 by Alan Zimmerman at 2021-07-28T22:30:59+01:00 EPA: match changes from GHC T19834 (cherry picked from commit 2fec1b44e0ee7e263286709aa528b4ecb99ac6c2) - - - - - 5f177278 by Ben Gamari at 2021-08-06T01:17:37-04:00 Merge commit '2a966c8ca37' into HEAD - - - - - cdd81d08 by Marcin Szamotulski at 2021-08-08T17:19:06+02:00 coot/multiple packages (ghc-9.2) (#1418) - - - - - be0d71f1 by Marcin Szamotulski at 2021-08-16T08:46:03+02:00 coot/multiple package (ghc-head) (#1419) * FromJSON class Aeson style FromJSON class with Parsec based json parser. * doc-index.json file for multiple packages When creating haddock summary page for multiple packages render doc-index.json file using contents of all found 'doc-index.json' files. * Render doc-index.json When rendering html, render doc-index.json file independently of maybe_index_url option. doc-index.json file is useful now even if maybe_index_url is not `Nothing`. * base url option New `Flag_BaseURL` which configures from where static files are loaded (--base-url). If given and not equal "." static files are not coppied, as this indicates that they are not read from the the directory where we'd copy them. The default value is ".". - - - - - 3b09dbdf by Hécate Moonlight at 2021-10-07T23:26:03+02:00 Update GHC 9.2 to latest pre-release in CI - - - - - 7ac55417 by Zubin Duggal at 2021-10-11T12:10:19+02:00 Enable Haddock tests in GHC windows CI (#1428) * testsuite: strip windows line endings for haddock * hyperlinker: Work around double escaping (#19236) * deterministic SCC - - - - - 1cb81f25 by Andrew Lelechenko at 2021-10-12T15:23:19+02:00 haddock-library does not depend on bytestring or transformers (#1426) - - - - - a890b9aa by sheaf at 2021-10-15T22:19:42+02:00 update haddockHypsrcTest for GHC MR !6705 (#1430) - - - - - 42a55c6c by Sylvain Henry at 2021-10-15T22:20:10+02:00 Fix after PkgQual refactoring (#1429) - - - - - 91659238 by Alan Zimmerman at 2021-10-28T18:57:10+01:00 Update for changes in GHC for branch wip/az/no-srcspan-anno-instances - - - - - acf23e60 by Vladislav Zavialov at 2021-11-05T02:09:47+03:00 Do not use forall as an identifier See GHC ticket haskell/haddock#20609 - - - - - c565db0e by Krzysztof Gogolewski at 2021-11-27T02:42:35+01:00 Update after NoExtCon -> DataConCantHappen rename - - - - - b5f55590 by Artem Pelenitsyn at 2021-11-27T11:14:17+01:00 fix CI for 9.2 (#1436) - - - - - 25cd621e by Matthew Pickering at 2021-12-02T11:46:54+00:00 Update html-test for Data.List revert - - - - - 1d5ff85f by malteneuss at 2021-12-15T07:56:55+01:00 Add hint about inline link issue (#1444) - - - - - 791fde81 by Sylvain Henry at 2021-12-16T09:29:51+01:00 Bump ghc-head (#1445) * Update after NoExtCon -> DataConCantHappen rename * Update html-test for Data.List revert * Fix for new Plugins datatype Co-authored-by: Krzysztof Gogolewski <krzysztof.gogolewski at tweag.io> Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 44236317 by Sylvain Henry at 2021-12-17T09:39:00+01:00 Fix for new Plugins datatype - - - - - 80ada0fa by Hécate Moonlight at 2021-12-17T17:28:48+01:00 Remove ghc-head workflow (#1446) Contributions of GHC glue code are now done on the GHC gitlab, not in the GitHub repo anymore. - - - - - 49e171cd by Matthew Pickering at 2021-12-28T09:47:09+00:00 Remove use of ExtendedModSummary - - - - - 0e91b5ea by askeblad at 2022-01-04T09:18:35+01:00 update URLs - - - - - 9f13c212 by Hécate Moonlight at 2022-02-25T10:19:46+01:00 Fix solver for GHC 9.2 - - - - - 386751a1 by Meng Weng Wong at 2022-02-25T19:19:11+01:00 IDoc link has bitrotted; replaced with web.archive.org cache. (#1454) - - - - - d877cbe6 by Hécate Moonlight at 2022-02-25T19:21:58+01:00 Fix haddock user guide (#1456) - - - - - cc47f036 by Andrew Lelechenko at 2022-03-04T17:29:36+01:00 Allow text-2.0 in haddock-library (#1459) - - - - - 7b3685a3 by malteneuss at 2022-03-07T19:27:24+01:00 Add multi-line style hint to style section (#1460) - - - - - c51088b8 by John Ericson at 2022-03-11T16:46:26+01:00 Fix CollectPass instance to match TTG refactor Companion to GHC !7614 (https://gitlab.haskell.org/ghc/ghc/-/merge_requests/7614) - - - - - b882195b by Vladislav Zavialov at 2022-03-14T20:32:30+01:00 Link to (~) - - - - - 877349b8 by Christiaan Baaij at 2022-03-16T09:20:43+01:00 Add Haddock support for the OPAQUE pragma - - - - - 0ea22721 by askeblad at 2022-03-16T09:44:27+01:00 typos (#1464) - - - - - a6d13da1 by Matthew Pickering at 2022-03-22T13:41:17+00:00 Minimum changes needed for compilation with hi-haddock With hi-haddock, of course there is a much large refactoring of haddock which could be achieved but that is left for a future patch which can implemented at any time independently of GHC. - - - - - e7ac9129 by Matthew Pickering at 2022-03-22T21:17:50+00:00 Update test output - - - - - 6d916214 by Matthew Pickering at 2022-03-24T15:06:26+00:00 Merge branch 'wip/opaque_pragma' into 'ghc-head' Add Haddock support for the OPAQUE pragma See merge request ghc/haddock!2 - - - - - 42208183 by Steve Hart at 2022-03-25T20:43:50+01:00 Fix CI (#1467) * CI: Reinstall GHC with docs CI tests were failing because the GHC preinstalled to the CI environment does not include documentation, which is required for running the Haddock tests. This commit causes the CI workflow to reinstall GHC with docs so that tests can succeed. - - - - - 9676fd79 by Steve Hart at 2022-03-25T21:33:34+01:00 Make links in Synopsis functional again (#1458) Commit e41c1cbe9f0476997eac7b4a3f17cbc6b2262faf added a call to e.preventDefault() when handling click events that reach a toggle element. This prevents the browser from following hyperlinks within the Synopsis section when they are clicked by a user. This commit restores functioning hyperlinks within the Synopsis section by removing the call to e.preventDefault(), as it does not appear to be necessary, and removing it increases the flexibility of the details-helper code. - - - - - d1edd637 by sheaf at 2022-04-01T12:02:02+02:00 Keep track of promotion ticks in HsOpTy Keeping track of promotion ticks in HsOpTy allows us to properly pretty-print promoted constructors such as lists. - - - - - 9dcb2dfc by Jakob Brünker at 2022-04-01T15:46:22+00:00 Add support for \cases See merge request ghc/ghc!7873 - - - - - b0412ee5 by askeblad at 2022-04-06T17:47:57+02:00 spelling errors (#1471) - - - - - 6b18829b by Vladislav Zavialov at 2022-04-06T18:53:58+02:00 Rename [] to List - - - - - 2d046691 by Vladislav Zavialov at 2022-04-07T20:25:54+03:00 HsToken ConDeclGADT con_dcolon - - - - - 90b43da4 by Steve Hart at 2022-04-12T13:29:46+02:00 Parse Markdown links at beginning of line within a paragraph (#1470) * Catch Markdown links at beginning of line within paragraph Per Issue haskell/haddock#774, Markdown links were being parsed as ordinary text when they occurred at the beginning of a line other than the first line of the paragraph. This occurred because the parser was not interpreting a left square bracket as a special character that could delimit special markup. A space character was considered a special character, so, if a space occurred at the beginning of the new line, then the parser would interpret the space by itself and then continue parsing, thereby catching the Markdown link. '\n' was not treated as a special character, so the parser did not catch a Markdown link that may have followed. Note that this will allow for Markdown links that are not surrounded by spaces. For example, the following text includes a Markdown link that will be parsed: Hello, world[label](url) This is consistent with how the parser handles other types of markup. * Remove obsolete documentation hint Commit 6b9aeafddf20efc65d3725c16e3fc43a20aac343 should eliminate the need for the workaround suggested in the documentation. - - - - - 5b08312d by Hécate Moonlight at 2022-04-12T13:36:38+02:00 Force ghc-9.2 in the cabal.project - - - - - 0d0ea349 by dependabot[bot] at 2022-04-12T13:57:41+02:00 Bump path-parse from 1.0.5 to 1.0.7 in /haddock-api/resources/html (#1469) Bumps [path-parse](https://github.com/jbgutierrez/path-parse) from 1.0.5 to 1.0.7. - [Release notes](https://github.com/jbgutierrez/path-parse/releases) - [Commits](https://github.com/jbgutierrez/path-parse/commits/v1.0.7) --- updated-dependencies: - dependency-name: path-parse dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 2b9fc65e by dependabot[bot] at 2022-04-12T13:57:54+02:00 Bump copy-props from 2.0.4 to 2.0.5 in /haddock-api/resources/html (#1468) Bumps [copy-props](https://github.com/gulpjs/copy-props) from 2.0.4 to 2.0.5. - [Release notes](https://github.com/gulpjs/copy-props/releases) - [Changelog](https://github.com/gulpjs/copy-props/blob/master/CHANGELOG.md) - [Commits](https://github.com/gulpjs/copy-props/compare/2.0.4...2.0.5) --- updated-dependencies: - dependency-name: copy-props dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - ea98a6fb by Ben Gamari at 2022-04-23T22:54:37-04:00 Update for GHC 9.4 - - - - - 9e11864f by Ben Gamari at 2022-04-25T16:24:31-04:00 Merge remote-tracking branch 'upstream/ghc-9.2' into ghc-head - - - - - f83cc506 by Ben Gamari at 2022-04-25T17:00:25-04:00 Bump ghc version to 9.5 - - - - - e01c2e7d by Ben Gamari at 2022-04-28T16:19:04-04:00 Revert "Bump ghc-head (#1445)" This reverts commit b29a78ef6926101338f62e84f456dac8659dc9d2. This should not have been merged. - - - - - a2b5ee8c by Ben Gamari at 2022-04-28T16:19:24-04:00 Merge commit '2627a86c' into ghc-head - - - - - 0c6fe4f9 by Ben Gamari at 2022-04-29T10:05:54-04:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-9.4 - - - - - b6e5cb0a by Ben Gamari at 2022-04-29T11:46:06-04:00 Revert "HsToken ConDeclGADT con_dcolon" This reverts commit 24208496649a02d5f87373052c430ea4a97842c5. - - - - - 15a62888 by Ben Gamari at 2022-04-29T15:12:55-04:00 Bump base upper bound - - - - - 165b9031 by Ben Gamari at 2022-04-29T23:58:38-04:00 Update test output - - - - - e0c3e5da by Phil de Joux at 2022-05-02T14:46:38+02:00 Add hlint action .hlint.yaml with ignores & CPP. (#1475) - - - - - ead1158d by Raphael Das Gupta at 2022-05-02T14:46:48+02:00 fix grammar in docs: "can the" → "can be" (#1477) - - - - - cff97944 by Ben Gamari at 2022-05-02T18:38:56-04:00 Allow base-4.17 - - - - - e4ecb201 by Phil de Joux at 2022-05-03T13:14:55+02:00 Remove unused imports that GHC warned about. (#1480) - - - - - 222890b1 by Phil de Joux at 2022-05-03T13:15:46+02:00 Follow hlint suggestion to remove redundant bang. (#1479) - - - - - 058b671f by Phil de Joux at 2022-05-03T13:34:04+02:00 Follow hlint, remove language pragmas in libs. (#1478) - - - - - 0a645049 by Ben Simms at 2022-05-03T14:19:24+02:00 Keep track of ordered list indexes and render them (#1407) * Keep track of ordered list indexes and render them * Rename some identifiers to clarify - - - - - f0433304 by Norman Ramsey at 2022-05-04T15:13:34-04:00 update for changes in GHC API - - - - - 3740cf71 by Emily Martins at 2022-05-06T18:23:48+02:00 Add link to the readthedocs in cabal description to show on hackage. (cherry picked from commit 52e2d40d47295c02d3181aac0c53028e730f1e3b) - - - - - 5d754f1e by Hécate Moonlight at 2022-05-06T18:44:57+02:00 remove Bug873 - - - - - 968fc267 by Hécate Moonlight at 2022-05-06T18:48:28+02:00 Ignore "Use second" HLint suggestion. It increases laziness. - - - - - 02d14e97 by Jade Lovelace at 2022-05-07T17:42:08+02:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` - - - - - b22b87ed by Artem Pelenitsyn at 2022-05-08T16:19:47+02:00 fix parsing trailing quotes in backticked identifiers (#1408) (#1483) - - - - - 80ae107b by Alex Biehl at 2022-05-08T16:37:16+02:00 Fix "Defined by not used" error (cherry picked from commit 6e02a620a26c3a44f98675dd1b93b08070c36c0a) - - - - - 4c838e84 by Hécate Moonlight at 2022-05-08T16:37:16+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - fc9827b4 by Hécate Moonlight at 2022-05-08T16:40:40+02:00 Fix the changelog and bump the version of haddock-library on ghc-9.2 - - - - - b153b555 by Xia Li-yao at 2022-05-20T17:52:42+02:00 Hide synopsis from search when hidden (#1486) Fix haskell/haddock#1451 - - - - - f3e38b85 by Marcin Szamotulski at 2022-05-21T23:32:31+02:00 Allow to hide interfaces when rendering multiple components (#1487) This is useful when one wishes to `--gen-contents` when rendering multiple components, but one does not want to render all modules. This is in particular useful when adding base package. - - - - - f942863b by Marcin Szamotulski at 2022-05-24T08:29:59+02:00 Check if doc-index.json exists before reading it (#1488) - - - - - 31e92982 by Marcin Szamotulski at 2022-05-25T16:22:13+02:00 Version bump 2.26.1 (#1489) * Version bump 2.26.1 We extended format accepted by `--read-interface` option, which requires updating the minor version. * Update documentation of --read-interface option - - - - - 7cc873e0 by sheaf at 2022-05-25T16:42:31+02:00 Updated HaddockHypsrcTest output for record update changes (MR !7981) - - - - - cd196942 by Marcin Szamotulski at 2022-05-25T20:28:47+02:00 Use visibility to decide which interfaces are included in quickjump (#1490) This is also consistent with how html index is build. See haskell/cabal#7669 for rationale behind this decision. - - - - - 00c713c5 by Hécate Moonlight at 2022-05-26T17:09:15+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 2f3039f1 by Hécate Moonlight at 2022-05-26T17:10:59+02:00 Add code of conduct and hspec failure files in .gitignore - - - - - 63a5650c by romes at 2022-05-31T12:43:22+01:00 TTG: Match new GHC AST - - - - - dd7d1617 by romes at 2022-06-02T16:11:00+01:00 Update for IE changes in !8228 - - - - - c23aaab7 by cydparser at 2022-06-06T08:48:14+02:00 Fix and improve CI (#1495) * Pin GHC version before creating the freeze file * Use newest action versions * Improve caching * Avoid unnecessarily reinstalling GHC * Use GHC 9.2.2 for CI Co-authored-by: Cyd Wise <cwise at tripshot.com> - - - - - c156fa77 by Hécate Moonlight at 2022-06-06T11:59:35+02:00 Add Mergify configuration (#1496) - - - - - 2dba4188 by Hécate Moonlight at 2022-06-06T16:12:50+02:00 Bump haddock's version in cabal file to 2.26.1 (#1497) - - - - - d7d4b8b9 by Marcin Szamotulski at 2022-06-07T06:09:40+00:00 Render module tree per package in the content page (#1492) * Render module tree per package in the content page When rendering content page for multiple packages it is useful to split the module tree per package. Package names in this patch are inferred from haddock's interface file names. * Write PackageInfo into interface file To keep interface file format backward compatible, instead of using `Binary` instance for `InterfaceFile` we introduce functions to serialise and deserialise, which depends on the interface file version. - - - - - 77765665 by Mike Pilgrem at 2022-06-12T21:57:19+01:00 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - b0e079b0 by mergify[bot] at 2022-06-13T11:49:37+00:00 Merge pull request haskell/haddock#1108 from mpilgrem/fix783 Fix haskell/haddock#783 Don't show button if --quickjump not present - - - - - 6c0292b1 by Hécate Moonlight at 2022-06-21T17:21:08+02:00 Update the contribution guide - - - - - e413b9fa by dependabot[bot] at 2022-06-21T23:38:19+02:00 Bump shell-quote from 1.6.1 to 1.7.3 in /haddock-api/resources/html (#1500) Bumps [shell-quote](https://github.com/substack/node-shell-quote) from 1.6.1 to 1.7.3. - [Release notes](https://github.com/substack/node-shell-quote/releases) - [Changelog](https://github.com/substack/node-shell-quote/blob/master/CHANGELOG.md) - [Commits](https://github.com/substack/node-shell-quote/compare/1.6.1...1.7.3) --- updated-dependencies: - dependency-name: shell-quote dependency-type: indirect ... Signed-off-by: dependabot[bot] <support at github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> - - - - - 29d0ef70 by romes at 2022-07-06T11:29:39+02:00 TTG: AST Updates for !8308 - - - - - 1bae7c87 by Alan Zimmerman at 2022-07-06T22:50:43+01:00 Match GHC changes for T21805 This brings in a newtype for FieldLabelString - - - - - 6fe8b988 by Phil de Joux at 2022-07-16T20:54:26+00:00 Bump hlint version to 3.4.1, the version with counts. (#1503) Redo the counts available with the --default option. - - - - - 48fb43af by Phil de Joux at 2022-07-19T09:32:55+02:00 Follow hlint suggestion: unused LANGUAGE pragma. (#1504) * Follow hlint suggestion: unused LANGUAGE pragma. * Ignore within modules to pass linting and pass tests. - - - - - c1cf1fa7 by Phil de Joux at 2022-07-24T13:45:59+02:00 Follow hlint suggestion: redundant $. (#1505) * Follow hlint suggestion: redundant $. * Remove $ and surplus blank lines in Operators. - - - - - 74777eb2 by Jade Lovelace at 2022-07-29T11:02:41+01:00 Fix hyperlinks to external items and modules (#1482) Fixes haskell/haddock#1481. There were two bugs in this: * We were assuming that we were always getting a relative path to the module in question, while Nix gives us file:// URLs sometimes. This change checks for those and stops prepending `..` to them. * We were not linking to the file under the module. This seems to have been a regression introduced by haskell/haddock#977. That is, the URLs were going to something like file:///nix/store/3bwbsy0llxxn1pixx3ll02alln56ivxy-ghc-9.0.2-doc/share/doc/ghc/html/libraries/base-4.15.1.0/src which does not have the appropriate HTML file or fragment for the item in question at the end. There is a remaining instance of the latter bug, but not in the hyperlinker: the source links to items reexported from other modules are also not including the correct file name. e.g. the reexport of Entity in esqueleto, from persistent. NOTE: This needs to get tested with relative-path located modules. It seems correct for Nix based on my testing. Testing strategy: ``` nix-shell '<nixpkgs>' --pure -A haskell.packages.ghc922.aeson mkdir /tmp/aesonbuild && cd /tmp/aesonbuild export out=/tmp/aesonbuild/out genericBuild ln -sf $HOME/co/haddock/haddock-api/resources . ./Setup haddock --with-haddock=$HOME/path/to/haddock/exec --hyperlink-source ``` (cherry picked from commit ab53ccf089ea703b767581ac14be0f6c78a7678a) - - - - - faa4cfcf by Hécate Moonlight at 2022-07-29T20:31:20+02:00 Merge pull request haskell/haddock#1516 from duog/9-4-backport-fix-hyperlinks Backport 9-4: Fix hyperlinks to external items and modules (#1482) - - - - - 5d2450f3 by Ben Gamari at 2022-08-05T17:41:15-04:00 Merge remote-tracking branch 'origin/ghc-9.4' - - - - - 63954f73 by Ben Gamari at 2022-08-05T19:08:36-04:00 Clean up build and testsuite for GHC 9.4 - - - - - d4568cb8 by Hécate Moonlight at 2022-08-05T19:10:49-04:00 Bump the versions - - - - - 505583a4 by Ben Gamari at 2022-08-06T13:58:27-04:00 Merge pull request haskell/haddock#1518 from bgamari/wip/ghc-9.4-merge Merge GHC 9.4 into `main` - - - - - 5706f6a4 by Ben Gamari at 2022-08-06T22:57:21-04:00 html-test: Testsuite changes for GHC 9.4.1 - - - - - 5f2a45a2 by Ben Gamari at 2022-08-15T14:33:05-04:00 doc: Fix a few minor ReST issues Sphinx was complaining about too-short title underlines. - - - - - 220e6410 by Ben Gamari at 2022-08-15T14:41:24-04:00 Merge branch 'main' into ghc-head - - - - - fbeb1b02 by Ben Gamari at 2022-08-15T14:45:16-04:00 Updates for GHC 9.5 - - - - - eee562eb by Vladislav Zavialov at 2022-08-15T14:46:13-04:00 HsToken ConDeclGADT con_dcolon - - - - - c5f073db by Ben Gamari at 2022-08-15T16:55:35-04:00 Updates for GHC 9.5 - - - - - 3f7ab242 by Vladislav Zavialov at 2022-08-15T16:55:35-04:00 HsToken ConDeclGADT con_dcolon - - - - - a18e473d by Ben Gamari at 2022-08-16T08:35:19-04:00 Merge branch 'wip/ghc-head-bump' into ghc-head - - - - - af0ff3a4 by M Farkas-Dyck at 2022-09-15T21:16:05+00:00 Disuse `mapLoc`. - - - - - a748fc38 by Matthew Farkas-Dyck at 2022-09-17T10:44:18+00:00 Scrub partiality about `NewOrData`. - - - - - 2758fb6c by John Ericson at 2022-09-18T03:27:37+02:00 Test output changed because of change to `base` Spooky, but I guess that is intended? - - - - - a7eec128 by Torsten Schmits at 2022-09-21T11:06:55+02:00 update tests for the move of tuples to GHC.Tuple.Prim - - - - - 461e7b9d by Ross Paterson at 2022-09-24T22:01:25+00:00 match implementation of GHC proposal haskell/haddock#106 (Define Kinds Without Promotion) - - - - - f7fd77ef by sheaf at 2022-10-17T14:53:01+02:00 Update Haddock for GHC MR !8563 (configuration of diagnostics) - - - - - 3d3e85ab by Vladislav Zavialov at 2022-10-22T23:04:06+03:00 Class layout info - - - - - cbde4cb0 by Simon Peyton Jones at 2022-10-25T23:19:18+01:00 Adapt to Constraint-vs-Type See haskell/haddock#21623 and !8750 - - - - - 7108ba96 by Tom Smeding at 2022-11-01T22:33:23+01:00 Remove outdated footnote about module re-exports The footnote is invalid with GHC 9.2.4 (and possibly earlier): the described behaviour in the main text works fine. - - - - - 206c6bc7 by Hécate Moonlight at 2022-11-01T23:00:46+01:00 Merge pull request haskell/haddock#1534 from tomsmeding/patch-1 - - - - - a57b4c4b by Andrew Lelechenko at 2022-11-21T00:39:52+00:00 Support mtl-2.3 - - - - - e9d62453 by Simon Peyton Jones at 2022-11-25T13:49:12+01:00 Track small API change in TyCon.hs - - - - - eb1c73f7 by Ben Gamari at 2022-12-07T08:46:21-05:00 Update for GhC 9.6 - - - - - 063268dd by Ben Gamari at 2022-12-07T11:26:32-05:00 Merge remote-tracking branch 'upstream/ghc-head' into HEAD - - - - - 4ca722fe by Ben Gamari at 2022-12-08T14:43:26-05:00 Bump bounds to accomodate base-4.18 - - - - - 340b7511 by Vladislav Zavialov at 2022-12-10T12:31:28+00:00 HsToken in HsAppKindTy - - - - - 946226ec by Ben Gamari at 2022-12-13T20:12:56-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - fd8faa66 by Ben Gamari at 2022-12-22T13:44:28-05:00 Bump GHC version to 9.7 - - - - - 2958aa9c by Ben Gamari at 2022-12-22T14:49:16-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 9e0fefd8 by Andrei Borzenkov at 2023-01-30T14:02:04+04:00 Rename () to Unit, Rename (,,...,,) to Tuple<n> - - - - - eb3968b5 by Ben Gamari at 2023-03-10T02:32:43-05:00 Bump versions for ghc-9.6 release - - - - - 4aeead36 by Adam Gundry at 2023-03-23T13:53:47+01:00 Adapt to warning categories changes - - - - - 642d8d60 by sheaf at 2023-03-29T13:35:56+02:00 Adapt to record field refactor This commit adapts to the changes in GHC MR !8686, which overhauls the treatment of record fields in the renamer, adding separate record field namespaces and entirely removing the GreName datatype. - - - - - ac8d4333 by doyougnu at 2023-03-29T11:11:44-04:00 Update UniqMap API - - - - - 7866fc86 by Ben Orchard at 2023-04-20T11:29:33+02:00 update classify with new tokens - - - - - ffcdd683 by Finley McIlwaine at 2023-04-24T09:36:18-06:00 Remove index-state - - - - - 05b70982 by Finley McIlwaine at 2023-04-26T08:16:31-06:00 `renameInterface` space leak fixes - Change logic for accumulation of names for which link warnings will be generated - Change handling of `--ignore-link-symbol` to allow qualified and unqualified names. Added to CHANGES.md - Some formatting changes and comments here and there - - - - - e5697d7c by Finley McIlwaine at 2023-04-27T18:46:36-06:00 Messy things - ghc-debug dependency and instrumentation - cabal.project custom with-compiler - hie.yaml files - traces and such - - - - - 0b8ef80b by Finley McIlwaine at 2023-05-02T18:08:52-06:00 Stop retaining GRE closures GRE closures should never be necessary to Haddock, so we never want to keep them on the heap. Despite that, they are retained by a lot of the data structures that Haddock makes use of. - Attempt to fix that situation by adding strictness to various computations and pruning the `ifaceInstances` field of `Interface` to a much thinner data type. - Removes the `ifaceFamInstances` field, as it was never used. - Move some of the attach instances types (e.g. `SimpleType`) to the types module - - - - - 8bda991b by Finley McIlwaine at 2023-05-08T16:07:51-06:00 Memory usage fixes - Refactor `ifaceDeclMap` to drastically reduce memory footprint. We no longer store all declarations associated with a given name, since we only cared to determine if the only declaration associated with a name was a value declaration. Change the `DeclMap` type to better reflect this. - Drop pre-renaming export items after the renaming step. Since the Hoogle backend used the pre-renamed export items, this isn't trivial. We now generate Hoogle output for exported declarations during the renaming step (if Hoogle output /should/ be generated), and store that with the renamed export item. - Slightly refactor Hoogle backend to handle the above change and allow for early generation of Hoogle output. - Remove the `ifaceRnDocMap` and `ifaceRnArgMap` fields of the `Interface` type, as they were never used. - Remove some unnecessary strictness - Remove a lot of dead code from `Syb` module - - - - - 1611ac0c by Finley McIlwaine at 2023-05-09T11:51:57-06:00 Unify ErrMsgM and IfM - Delete ErrMsgM, stop accumulating warnings in a writer - Make IfM a state monad, print warnings directly to stdout, move IfM type into types module - Drop ErrMsg = String synonym - Unset IORefs from plugin after they are read, preventing unnecessary retention of interfaces - - - - - 42d696ab by Finley McIlwaine at 2023-05-11T15:52:07-06:00 Thunk leak fixes The strictness introduced in this commit was motivated by observing thunk leaks in the eventlog2html output. - Refactor attach instances list comprehension to avoid large intermediate thunks - Refactor some HTML backend list comprehensions to avoid large intermediate thunks - Avoid thunks accumulating in documentation types or documentation parser - A lot of orphan NFData instances to allow us to force documentation values - - - - - 68561cf6 by Finley McIlwaine at 2023-05-11T17:02:10-06:00 Remove GHC debug dep - - - - - 10519e3d by Finley McIlwaine at 2023-05-15T12:40:48-06:00 Force HIE file path Removes a potential retainer of `ModSummary`s - - - - - 1e4a6ec6 by Finley McIlwaine at 2023-05-15T14:20:34-06:00 Re-add index-state, with-compiler, delete hie.yamls - - - - - a2363fe9 by Hécate Moonlight at 2023-05-15T22:45:16+02:00 Merge pull request haskell/haddock#1594 from FinleyMcIlwaine/finley/ghc-9.6-mem-fixes Reduce memory usage - - - - - e8a78383 by Finley McIlwaine at 2023-05-17T12:19:16-06:00 Merge branch ghc-9.6 into ghc-head - - - - - 22e25581 by Finley McIlwaine at 2023-05-17T12:20:23-06:00 Merge branch 'ghc-head' of gitlab.haskell.org:ghc/haddock into ghc-head - - - - - 41bbf0df by Bartłomiej Cieślar at 2023-05-24T08:57:58+02:00 changes to the WarningTxt cases Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - c686ba9b by Hécate Moonlight at 2023-06-01T14:03:02-06:00 Port the remains of Hi-Haddock - - - - - 9d8a85fd by Hécate Moonlight at 2023-06-01T14:03:06-06:00 Stdout for tests - - - - - 36331d07 by Finley McIlwaine at 2023-06-01T14:06:02-06:00 Formatting, organize imports - - - - - a06059b1 by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix empty context confusion in Convert module - - - - - 379346ae by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix associated type families in Hoogle output - - - - - fc6ea7ed by Finley McIlwaine at 2023-06-01T14:06:04-06:00 Fix test refs Accept several changes in Hoogle tests: Pretty printing logic no longer prints the `(Proxy (Proxy (...))` chain in Bug806 with parentheses. Since this test was only meant to verify that line breaks do not occur, accept the change. `tyThingToLHsDecl` is called for class and data declarations, which ends up "synifying" the type variables and giving unlifted types kind signatures. As a result, type variables of kind `Type -> Type` are now printed with kind signatures in Hoogle output. This could be changed by manually drop kind signatures from class variables in the Hoogle backend if the behavior is deemed unacceptable. Sometimes subordinate declarations are exported separate from their parent declarations (e.g. record selectors). In this case, a type signature is cobbled together for the export item in `extractDecl`. Since this type signature is very manually constructed, it may lack kind signatures of decls constructed from `tyThingToLHsDecl`. An example of this is the `type-sigs` Hoogle test. Change `*` to `Type` in Hoogle test refs. I don't think this will break Hoogle behavior, since it appears to not consider type signatures in search. I have not fully verified this. - - - - - e14b7e58 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix LaTeX backend test refs Changes to GHC pretty printing code have resulted in some differences to Haddock's LaTeX output. - Type variables are printed explicitly quantified in the LinearTypes test - Wildcard types in type family equations are now printed numbered, e.g. _1 _2, in the TypeFamilies3 test - Combined signatures in DefaultSignatures test are now documented as separate signatures - - - - - 41b5b296 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and test source updates - Stop using kind `*` in html test sources - Add TypeOperators where necessary to avoid warnings and future errors - Rename some test modules to match their module names - - - - - c640e2a2 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Fix missing deprecation warnings on record fields `lookupOccEnv` was used to resolve `OccNames` with warnings attached, but it doesn't look in the record field namespace. Thus, if a record field had a warning attached, it would not resolve and the warning map would not include it. This commit fixes by using `lookupOccEnv_WithFields` instead. - - - - - fad0c462 by Finley McIlwaine at 2023-06-01T14:06:05-06:00 Formatting and some comments - - - - - 751fd023 by Finley McIlwaine at 2023-06-01T14:11:41-06:00 Accept HTML test diffs All diffs now boil down to the expected differences resulting from declarations being reified from TyThings in hi-haddock. Surface syntax now has much less control over the syntax used in the documentation. - - - - - d835c845 by Finley McIlwaine at 2023-06-01T14:11:45-06:00 Adapt to new load' type - - - - - dcf776c4 by Finley McIlwaine at 2023-06-01T14:13:13-06:00 Update mkWarningMap and moduleWarning - - - - - 8e8432fd by Finley McIlwaine at 2023-06-01T14:28:54-06:00 Revert load' changes - - - - - aeb2982c by Finley McIlwaine at 2023-06-01T14:40:24-06:00 Accept change to Instances test in html-test Link to Data.Tuple instead of GHC.Tuple.Prim - - - - - 8adfdbac by Finley McIlwaine at 2023-06-01T15:53:17-06:00 Reset ghc dep to ^>= 9.6 - - - - - 2b1ce93d by Finley McIlwaine at 2023-06-06T07:50:04-06:00 Update CHANGES.md, user guide, recomp avoidance * Add --trace-args flag for tracing arguments received to standard output * Avoid recompiling due to changes in optimization flags * Update users guide and changes.md - - - - - f3da6676 by Finley McIlwaine at 2023-06-06T14:12:56-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - bf36c467 by Matthew Pickering at 2023-06-07T10:16:09+01:00 Revert back to e16e20d592a6f5d9ed1af17b77fafd6495242345 Neither of these MRs are ready to land yet which causes issues with other MRs which are ready to land and need haddock changes. - - - - - 421510a9 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 atSign has no unicode variant Prior to this change, atSign was defined as follows: atSign unicode = text (if unicode then "@" else "@") Yes, this is the same symbol '\64' and not your font playing tricks on you. Now we define: atSign = char '@' Both the LaTeX and the Xhtml backend are updated accordingly. - - - - - 3785c276 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 LaTeX: fix printing of type variable bindings Consider this type signature: kindOf :: forall {k} (a :: k). Proxy a -> Proxy k Prior to this fix, the LaTeX backend rendered it like this: kindOf :: forall k a. Proxy a -> Proxy k Now we preserve explicit specificity and kind annotations. - - - - - 0febf3a8 by Vladislav Zavialov at 2023-06-07T09:36:30+00:00 Add support for invisible binders in type declarations - - - - - 13e33bb3 by Finley McIlwaine at 2023-06-08T07:51:59-06:00 Add "Avoiding Recompilation" section to docs This section is a bit of a WIP due to the unstable nature of hi-haddock and the lack of tooling supporting it, but its a good start. - - - - - 3e5340ce by Finley McIlwaine at 2023-06-08T07:54:27-06:00 Add note about stubdir to recompilation docs - - - - - db7e84dc by Finley at 2023-06-08T08:11:03-06:00 Merge pull request haskell/haddock#1597 from haskell/finley/hi-haddock-9.6 hi-haddock for ghc 9.6 - - - - - 4e085d17 by Finley McIlwaine at 2023-06-14T13:41:06-06:00 Replace SYB traversals - - - - - 7b39aec5 by Finley McIlwaine at 2023-06-14T14:20:17-06:00 Test ref accepts, remove unused HaddockClsInst - - - - - df9c2090 by Finley McIlwaine at 2023-06-15T08:02:51-06:00 Use batchMsg for progress reporting during load With hi-haddock as is, there is an awkward silence during the load operation. This commit makes haddock use the default `batchMsg` Messager for progress reporting, and makes the default GHC verbosity level 1, so the user can see what GHC is doing. - - - - - f23679a8 by Hécate Moonlight at 2023-06-15T20:31:53+02:00 Merge pull request haskell/haddock#1600 from haskell/finley/hi-haddock-optim - - - - - a7982192 by Finley McIlwaine at 2023-06-15T15:02:16-06:00 hi-haddock squashed - - - - - c34f0c8d by Finley McIlwaine at 2023-06-15T16:22:03-06:00 Merge remote-tracking branch 'origin/ghc-9.6' into finley/hi-haddock-squashed - - - - - 40452797 by Bartłomiej Cieślar at 2023-06-16T12:26:04+02:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - e58673bf by Ben Gamari at 2023-06-16T09:33:35-04:00 Bump GHC version to 9.8 - - - - - 74bdf972 by Ben Gamari at 2023-06-16T09:36:18-04:00 Merge commit 'fcaaad06770a26d35d4aafd65772dedadf17669c' into ghc-head - - - - - 418ee3dc by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Remove NFData SourceText, IfaceWarnings updates The NFData SourceText instance is now available in GHC Handle mod_iface mi_warns now being IfaceWarnings - - - - - 62f31380 by Finley McIlwaine at 2023-06-20T15:39:05-04:00 Accept Instances.hs test output Due to ghc!10469. - - - - - a8f2fc0e by Ben Gamari at 2023-06-20T15:48:08-04:00 Test fixes for "Fix associated data family doc structure items" Associated data families were being given their own export DocStructureItems, which resulted in them being documented separately from their classes in haddocks. This commit fixes it. - - - - - cb1ac33e by Bartłomiej Cieślar at 2023-06-21T12:56:02-04:00 Changes related to MR !10283 MR !10283 changes the alternatives for WarningTxt pass. This MR reflects those changes in the haddock codebase. Signed-off-by: Bartłomiej Cieślar <bcieslar2001 at gmail.com> - - - - - 9933e10b by Ben Gamari at 2023-06-21T12:56:02-04:00 Bump GHC version to 9.8 - - - - - fe8c18b6 by Ben Gamari at 2023-06-21T15:36:29-04:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - c61a0d5b by Ben Gamari at 2023-06-21T16:10:51-04:00 Bump GHC version to 9.9 - - - - - 0c2a756e by sheaf at 2023-07-07T13:45:12+02:00 Avoid incomplete record update in Haddock Hoogle This commit avoids running into an incomplete record update warning in the Hoogle backend of Haddock. This was only noticed now, because incomplete record updates were broken in GHC 9.6. Now that they are fixed, we have to avoid running into them! - - - - - f9b952a7 by Ben Gamari at 2023-07-21T11:58:05-04:00 Bump base bound to <4.20 For GHC 9.8. - - - - - 1b27e151 by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Check for puns (see ghc#23368) - - - - - 457341fd by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Remove fake exports for (~), List, and Tuple<n> The old reasoning no longer applies, nowadays those names can be mentioned in export lists. - - - - - bf3dcddf by Vladislav Zavialov at 2023-08-02T10:42:11+00:00 Fix pretty-printing of Solo and MkSolo - - - - - 495b2241 by Matthew Pickering at 2023-09-01T13:02:07+02:00 Fix issue with duplicate reexported definitions (T23616) When a class method was reexported, it's default methods were also showing up in the generated html page. The simplest and most non-invasive fix is to not look for the default method if we are just exporting the class method.. because the backends are just showing default methods when the whole class is exported. In general it would be worthwhile to rewrite this bit of code I think as the logic and what gets included is split over `lookupDocs` and `availExportDecl` it would be clearer to combine the two. The result of lookupDocs is always just passed to availExportDecl so it seems simpler and more obvious to just write the function directly. - - - - - 6551824d by Finley McIlwaine at 2023-09-05T13:06:57-07:00 Remove fake export of `FUN` from Prelude This prevents `data FUN` from being shown at the top of the Prelude docs. Fixes \#23920 on GHC. - - - - - 9ab5a448 by Alan Zimmerman at 2023-09-08T18:26:53+01:00 Match changes in wip/az/T23885-unicode-funtycon - - - - - 4d08364e by Alan Zimmerman at 2023-10-31T19:46:45+00:00 EPA: match changes in GHC - EPA: Comments in AnchorOperation - EPA: Remove EpaEofComment - - - - - e7da0d25 by Alan Zimmerman at 2023-11-05T11:20:31+00:00 EPA: match changes in GHC, l2l cleanup - - - - - 4ceac14d by Alan Zimmerman at 2023-11-11T15:16:41+00:00 EPA: Replace Anchor with EpaLocation Match GHC - - - - - 94fb8d47 by Alan Zimmerman at 2023-11-29T18:10:26+00:00 Match GHC, No comments in EpaDelta for comments - - - - - 32d208e1 by Vladislav Zavialov at 2023-12-12T20:41:36+03:00 EPA: Match changes to LHsToken removal - - - - - eebdd316 by Apoorv Ingle at 2024-01-23T13:49:12+00:00 Changes for haskell/haddock#18324 - - - - - ae856a82 by Matthew Pickering at 2024-02-05T12:22:39+00:00 ghc-internals fallout - - - - - f8429266 by Jade at 2024-02-08T14:56:50+01:00 Adjust test for ghc MR !10993 - - - - - 6d1e2386 by Alan Zimmerman at 2024-02-13T22:00:28+03:00 EPA: Match changes to HsParTy and HsFunTy - - - - - 9c588f19 by Fendor at 2024-02-14T11:05:36+01:00 Adapt to GHC giving better Name's for linking - - - - - 778e1db3 by Andrei Borzenkov at 2024-02-16T16:12:07+03:00 Namespace specifiers for fixity signatures - - - - - 826c5b47 by Torsten Schmits at 2024-02-21T13:17:05+01:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 2cff14d5 by Ben Gamari at 2024-02-22T09:35:56-05:00 Bump bounds - - - - - f49376b3 by Ben Gamari at 2024-02-22T09:35:56-05:00 Allow `@since` annotations in export lists Here we extend Haddock to admit `@since` annotations in export lists. These can be attached to most export list items (although not subordinate lists). These annotations supercede the declaration's `@since` annotation in produced Haddocks. - - - - - b5aa93df by Ben Gamari at 2024-02-22T12:09:06-05:00 Allow package-qualified @since declarations - - - - - 8f5957f2 by Ben Gamari at 2024-02-22T13:55:19-05:00 Documentation changes from ghc-internal restructuring Previously many declarations (e.g. `Int`) were declared to have a "home" in `Prelude`. However, now Haddock instead chooses to put these in more specific homes (e.g. `Data.Int`). Given that the "home" decision is driven by heuristics and in general these changes seem quite reasonable I am accepting them: * `Int` moved from `Prelude` to `Data.Int` * `(~)` moved from `Prelude` to `Data.Type.Equality` * `Type` moved from `GHC.Types` to `Data.Kind` * `Maybe` moved from `Prelude` to `Data.Maybe` * `Bool` moved from `Prelude` to `Data.Bool` * `Ordering` moved from `Prelude` to `Data.Ord` As well, more identifiers are now hyperlinked; it's not immediately clear *why*, but it is an improvement nevertheless. - - - - - ec33fec3 by Ben Gamari at 2024-02-22T20:36:24-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 30cfd251 by Torsten Schmits at 2024-02-24T13:00:42-05:00 rename GHC.Tuple.Prim to GHC.Tuple - - - - - 732db81d by Ben Gamari at 2024-02-24T19:12:18-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 86bf7010 by Ben Gamari at 2024-02-27T19:28:10-05:00 Merge remote-tracking branch 'origin/ghc-head' into HEAD - - - - - 4b6e76b5 by Patrick at 2024-03-07T22:09:30+08:00 fix haskell/haddock#24493, with module name introduced in hieAst The accompanies haddoc PR with GHC PR https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12153 Two things have been done: 1. Link is introduced to every `X` in `module X where`, since we introduce the module name to HieAst, 2. `{-# LINE 4 "hypsrc-test/src/PositionPragmas.hs" #-}` is added before the `module PositionPragmas where` in ` hypsrc-test/ref/src/PositionPragmas.html `.It ensures only a single hieAst for file `hypsrc-test/src/PositionPragmas.hs` is generated. - - - - - 635abccc by Ben Gamari at 2024-03-08T17:09:06-05:00 Bump ghc version to 9.10 - - - - - 5b934048 by Ben Gamari at 2024-03-08T18:50:12-05:00 Bump base upper bound - - - - - b30d134e by Ben Gamari at 2024-03-08T18:50:44-05:00 Testsuite output update - - - - - 9bdf3586 by Ben Gamari at 2024-03-09T21:37:44-05:00 Merge branch 'ghc-9.10' into ghc-head - - - - - cec76981 by Ben Gamari at 2024-03-09T21:54:00-05:00 Bump GHC version to 9.11 - - - - - 4c59feb7 by Ben Gamari at 2024-03-09T22:15:01-05:00 Merge remote-tracking branch 'origin/ghc-head' into ghc-head - - - - - 97a2bb1c by Simon Peyton Jones at 2024-03-20T17:11:29+00:00 Expand untyped splices in tcPolyExprCheck Fixes #24559 - - - - - 5f275176 by Alan Zimmerman at 2024-03-20T22:44:12-04:00 EPA: Clean up Exactprint helper functions a bit - Introduce a helper lens to compose on `EpAnn a` vs `a` versions - Rename some prime versions of functions back to non-prime They were renamed during the rework - - - - - da2a10ce by Vladislav Zavialov at 2024-03-20T22:44:48-04:00 Type operators in promoteOccName (#24570) Type operators differ from term operators in that they are lexically classified as (type) constructors, not as (type) variables. Prior to this change, promoteOccName did not account for this difference, causing a scoping issue that affected RequiredTypeArguments. type (!@#) = Bool f = idee (!@#) -- Not in scope: ‘!@#’ (BUG) Now we have a special case in promoteOccName to account for this. - - - - - 247fc0fa by Preetham Gujjula at 2024-03-21T10:19:18-04:00 docs: Remove mention of non-existent Ord instance for Complex The documentation for Data.Complex says that the Ord instance for Complex Float is deficient, but there is no Ord instance for Complex a. The Eq instance for Complex Float is similarly deficient, so we use that as an example instead. - - - - - 6fafc51e by Andrei Borzenkov at 2024-03-21T10:19:54-04:00 Fix TH handling in `pat_to_type_pat` function (#24571) There was missing case for `SplicePat` in `pat_to_type_at` function, hence patterns with splicing that checked against `forall->` doesn't work properly because they fall into the "illegal pattern" case. Code example that is now accepted: g :: forall a -> () g $([p| a |]) = () - - - - - 52072f8e by Sylvain Henry at 2024-03-21T21:01:59-04:00 Type-check default declarations before deriving clauses (#24566) See added Note and #24566. Default declarations must be type-checked before deriving clauses. - - - - - 7dfdf3d9 by Sylvain Henry at 2024-03-21T21:02:40-04:00 Lexer: small perf changes - Use unsafeChr because we know our values to be valid - Remove some unnecessary use of `ord` (return Word8 values directly) - - - - - 864922ef by Sylvain Henry at 2024-03-21T21:02:40-04:00 JS: fix some comments - - - - - 3e0b2b1f by Sebastian Graf at 2024-03-21T21:03:16-04:00 Simplifier: Re-do dependency analysis in abstractFloats (#24551) In #24551, we abstracted a string literal binding over a type variable, triggering a CoreLint error when that binding floated to top-level. The solution implemented in this patch fixes this by re-doing dependency analysis on a simplified recursive let binding that is about to be type abstracted, in order to find the minimal set of type variables to abstract over. See wrinkle (AB5) of Note [Floating and type abstraction] for more details. Fixes #24551 - - - - - 8a8ac65a by Matthew Craven at 2024-03-23T00:20:52-04:00 Improve toInteger @Word32 on 64-bit platforms On 64-bit platforms, every Word32 fits in an Int, so we can convert to Int# without having to perform the overflow check integerFromWord# uses internally. - - - - - 0c48f2b9 by Apoorv Ingle at 2024-03-23T00:21:28-04:00 Fix for #24552 (see testcase T24552) Fixes for a bug in desugaring pattern synonyms matches, introduced while working on on expanding `do`-blocks in #18324 The `matchWrapper` unecessarily (and incorrectly) filtered out the default wild patterns in a match. Now the wild pattern alternative is simply ignored by the pm check as its origin is `Generated`. The current code now matches the expected semantics according to the language spec. - - - - - b72705e9 by Simon Peyton Jones at 2024-03-23T00:22:04-04:00 Print more info about kinds in error messages This fixes #24553, where GHC unhelpfully said error: [GHC-83865] • Expected kind ‘* -> * -> *’, but ‘Foo’ has kind ‘* -> * -> *’ See Note [Showing invisible bits of types in error messages] - - - - - 8f7cfc7e by Tristan Cacqueray at 2024-03-23T00:22:44-04:00 docs: remove the don't use float hint This hint is outdated, ``Complex Float`` are now specialised, and the heap space suggestion needs more nuance so it should be explained in the unboxed/storable array documentation. - - - - - 5bd8ed53 by Andreas Klebinger at 2024-03-23T16:18:33-04:00 NCG: Fix a bug in jump shortcutting. When checking if a jump has more than one destination account for the possibility of some jumps not being representable by a BlockId. We do so by having isJumpishInstr return a `Maybe BlockId` where Nothing represents non-BlockId jump destinations. Fixes #24507 - - - - - 8d67f247 by Ben Gamari at 2024-03-23T16:19:09-04:00 docs: Drop old release notes, add for 9.12.1 - - - - - 7db8c992 by Cheng Shao at 2024-03-25T13:45:46-04:00 rts: fix clang compilation on aarch64 This patch fixes function prototypes in ARMOutlineAtomicsSymbols.h which causes "error: address argument to atomic operation must be a pointer to _Atomic type" when compiling with clang on aarch64. - - - - - 237194ce by Sylvain Henry at 2024-03-25T13:46:27-04:00 Lexer: fix imports for Alex 3.5.1 (#24583) - - - - - 810660b7 by Cheng Shao at 2024-03-25T22:19:16-04:00 libffi-tarballs: bump libffi-tarballs submodule to libffi 3.4.6 This commit bumps the libffi-tarballs submodule to libffi 3.4.6, which includes numerous upstream libffi fixes, especially https://github.com/libffi/libffi/issues/760. - - - - - d2ba41e8 by Alan Zimmerman at 2024-03-25T22:19:51-04:00 EPA: do not duplicate comments in signature RHS - - - - - 32a8103f by Rodrigo Mesquita at 2024-03-26T21:16:12-04:00 configure: Use LDFLAGS when trying linkers A user may configure `LDFLAGS` but not `LD`. When choosing a linker, we will prefer `ldd`, then `ld.gold`, then `ld.bfd` -- however, we have to check for a working linker. If either of these fail, we try the next in line. However, we were not considering the `$LDFLAGS` when checking if these linkers worked. So we would pick a linker that does not support the current $LDFLAGS and fail further down the line when we used that linker with those flags. Fixes #24565, where `LDFLAGS=-Wl,-z,pack-relative-relocs` is not supported by `ld.gold` but that was being picked still. - - - - - bf65a7c3 by Rodrigo Mesquita at 2024-03-26T21:16:48-04:00 bindist: Clean xattrs of bin and lib at configure time For issue #21506, we started cleaning the extended attributes of binaries and libraries from the bindist *after* they were installed to workaround notarisation (#17418), as part of `make install`. However, the `ghc-toolchain` binary that is now shipped with the bindist must be run at `./configure` time. Since we only cleaned the xattributes of the binaries and libs after they were installed, in some situations users would be unable to run `ghc-toolchain` from the bindist, failing at configure time (#24554). In this commit we move the xattr cleaning logic to the configure script. Fixes #24554 - - - - - cfeb70d3 by Rodrigo Mesquita at 2024-03-26T21:17:24-04:00 Revert "NCG: Fix a bug in jump shortcutting." This reverts commit 5bd8ed53dcefe10b72acb5729789e19ceb22df66. Fixes #24586 - - - - - 13223f6d by Serge S. Gulin at 2024-03-27T07:28:51-04:00 JS: `h$rts_isProfiled` is removed from `profiling` and left its version at `rts/js/config.js` - - - - - 0acfe391 by Alan Zimmerman at 2024-03-27T07:29:27-04:00 EPA: Do not extend declaration range for trailine zero len semi The lexer inserts virtual semicolons having zero width. Do not use them to extend the list span of items in a list. - - - - - cd0fb82f by Alan Zimmerman at 2024-03-27T19:33:08+00:00 EPA: Fix FamDecl range The span was incorrect if opt_datafam_kind_sig was empty - - - - - f8f384a8 by Ben Gamari at 2024-03-29T01:23:03-04:00 Fix type of _get_osfhandle foreign import Fixes #24601. - - - - - 00d3ecf0 by Alan Zimmerman at 2024-03-29T12:19:10+00:00 EPA: Extend StringLiteral range to include trailing commas This goes slightly against the exact printing philosophy where trailing decorations should be in an annotation, but the practicalities of adding it to the WarningTxt environment, and the problems caused by deviating do not make a more principles approach worthwhile. - - - - - efab3649 by brandon s allbery kf8nh at 2024-03-31T20:04:01-04:00 clarify Note [Preproccesing invocations] - - - - - c8a4c050 by Ben Gamari at 2024-04-02T12:50:35-04:00 rts: Fix TSAN_ENABLED CPP guard This should be `#if defined(TSAN_ENABLED)`, not `#if TSAN_ENABLED`, lest we suffer warnings. - - - - - e91dad93 by Cheng Shao at 2024-04-02T12:50:35-04:00 rts: fix errors when compiling with TSAN This commit fixes rts compilation errors when compiling with TSAN: - xxx_FENCE macros are redefined and trigger CPP warnings. - Use SIZEOF_W. WORD_SIZE_IN_BITS is provided by MachDeps.h which Cmm.h doesn't include by default. - - - - - a9ab9455 by Cheng Shao at 2024-04-02T12:50:35-04:00 rts: fix clang-specific errors when compiling with TSAN This commit fixes clang-specific rts compilation errors when compiling with TSAN: - clang doesn't have -Wtsan flag - Fix prototype of ghc_tsan_* helper functions - __tsan_atomic_* functions aren't clang built-ins and sanitizer/tsan_interface_atomic.h needs to be included - On macOS, TSAN runtime library is libclang_rt.tsan_osx_dynamic.dylib, not libtsan. -fsanitize-thread as a link-time flag will take care of linking the TSAN runtime library anyway so remove tsan as an rts extra library - - - - - 865bd717 by Cheng Shao at 2024-04-02T12:50:35-04:00 compiler: fix github link to __tsan_memory_order in a comment - - - - - 07cb627c by Cheng Shao at 2024-04-02T12:50:35-04:00 ci: improve TSAN CI jobs - Run TSAN jobs with +thread_sanitizer_cmm which enables Cmm instrumentation as well. - Run TSAN jobs in deb12 which ships gcc-12, a reasonably recent gcc that @bgamari confirms he's using in #GHC:matrix.org. Ideally we should be using latest clang release for latest improvements in sanitizers, though that's left as future work. - Mark TSAN jobs as manual+allow_failure in validate pipelines. The purpose is to demonstrate that we have indeed at least fixed building of TSAN mode in CI without blocking the patch to land, and once merged other people can begin playing with TSAN using their own dev setups and feature branches. - - - - - a1c18c7b by Andrei Borzenkov at 2024-04-02T12:51:11-04:00 Merge tc_infer_hs_type and tc_hs_type into one function using ExpType philosophy (#24299, #23639) This patch implements refactoring which is a prerequisite to updating kind checking of type patterns. This is a huge simplification of the main worker that checks kind of HsType. It also fixes the issues caused by previous code duplication, e.g. that we didn't add module finalizers from splices in inference mode. - - - - - 817e8936 by Rodrigo Mesquita at 2024-04-02T20:13:05-04:00 th: Hide the Language.Haskell.TH.Lib.Internal module from haddock Fixes #24562 - - - - - b36ee57b by Sylvain Henry at 2024-04-02T20:13:46-04:00 JS: reenable h$appendToHsString optimization (#24495) The optimization introducing h$appendToHsString wasn't kicking in anymore (while it did in 9.8.1) because of the changes introduced in #23270 (7e0c8b3bab30). This patch reenables the optimization by matching on case-expression, as done in Cmm for unpackCString# standard thunks. The test is also T24495 added in the next commits (two commits for ease of backporting to 9.8). - - - - - 527616e9 by Sylvain Henry at 2024-04-02T20:13:46-04:00 JS: fix h$appendToHsString implementation (#24495) h$appendToHsString needs to wrap its argument in an updatable thunk to behave like unpackAppendCString#. Otherwise if a SingleEntry thunk is passed, it is stored as-is in a CONS cell, making the resulting list impossible to deepseq (forcing the thunk doesn't update the contents of the CONS cell)! The added test checks that the optimization kicks in and that h$appendToHsString works as intended. Fix #24495 - - - - - faa30b41 by Simon Peyton Jones at 2024-04-02T20:14:22-04:00 Deal with duplicate tyvars in type declarations GHC was outright crashing before this fix: #24604 - - - - - e0b0c717 by Simon Peyton Jones at 2024-04-02T20:14:58-04:00 Try using MCoercion in exprIsConApp_maybe This is just a simple refactor that makes exprIsConApp_maybe a little bit more direct, simple, and efficient. Metrics: compile_time/bytes allocated geo. mean -0.1% minimum -2.0% maximum -0.0% Not a big gain, but worthwhile given that the code is, if anything, easier to grok. - - - - - 15f4d867 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Initial ./configure support for selecting I/O managers In this patch we just define new CPP vars, but don't yet use them or replace the existing approach. That will follow. The intention here is that every I/O manager can be enabled/disabled at GHC build time (subject to some constraints). More than one I/O manager can be enabled to be built. At least one I/O manager supporting the non-threaded RTS must be enabled as well as at least one supporting the non-threaded RTS. The I/O managers enabled here will become the choices available at runtime at RTS startup (in later patches). The choice can be made with RTS flags. There are separate sets of choices for the threaded and non-threaded RTS ways, because most I/O managers are specific to these ways. Furthermore we must establish a default I/O manager for the threaded and non-threaded RTS. Most I/O managers are platform-specific so there are checks to ensure each one can be enabled on the platform. Such checks are also where (in future) any system dependencies (e.g. libraries) can be checked. The output is a set of CPP flags (in the mk/config.h file), with one flag per named I/O manager: * IOMGR_BUILD_<name> : which ones should be built (some) * IOMGR_DEFAULT_NON_THREADED_<name> : which one is default (exactly one) * IOMGR_DEFAULT_THREADED_<name> : which one is default (exactly one) and a set of derived flags in IOManager.h * IOMGR_ENABLED_<name> : enabled for the current RTS way Note that IOMGR_BUILD_<name> just says that an I/O manager will be built for _some_ RTS way (i.e. threaded or non-threaded). The derived flags IOMGR_ENABLED_<name> in IOManager.h say if each I/O manager is enabled in the "current" RTS way. These are the ones that can be used for conditional compilation of the I/O manager code. Co-authored-by: Pi Delport <pi at well-typed.com> - - - - - 85b0f87a by Duncan Coutts at 2024-04-03T01:27:17-04:00 Change the handling of the RTS flag --io-manager= Now instead of it being just used on Windows to select between the WinIO vs the MIO or Win32-legacy I/O managers, it is now used on all platforms for selecting the I/O manager to use. Right now it remains the case that there is only an actual choice on Windows, but that will change later. Document the --io-manager flag in the user guide. This change is also reflected in the RTS flags types in the base library. Deprecate the export of IoSubSystem from GHC.RTS.Flags with a message to import it from GHC.IO.Subsystem. The way the 'IoSubSystem' is detected also changes. Instead of looking at the RTS flag, there is now a C bool global var in the RTS which gets set on startup when the I/O manager is selected. This bool var says whether the selected I/O manager classifies as "native" on Windows, which in practice means the WinIO I/O manager has been selected. Similarly, the is_io_mng_native_p RTS helper function is re-implemented in terms of the selected I/O manager, rather than based on the RTS flags. We do however remove the ./configure --native-io-manager flag because we're bringing the WinIO/MIO/Win32-legacy choice under the new general scheme for selecting I/O managers, and that new scheme involves no ./configure time user choices, just runtime RTS flag choices. - - - - - 1a8f020f by Duncan Coutts at 2024-04-03T01:27:17-04:00 Convert {init,stop,exit}IOManager to switch style Rather than ad-hoc cpp conitionals on THREADED_RTS and mingw32_HOST_OS, we use a style where we switch on the I/O manager impl, with cases for each I/O manager impl. - - - - - a5bad3d2 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Split up the CapIOManager content by I/O manager Using the new IOMGR_ENABLED_<name> CPP defines. - - - - - 1d36e609 by Duncan Coutts at 2024-04-03T01:27:17-04:00 Convert initIOManagerAfterFork and wakeupIOManager to switch style - - - - - c2f26f36 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move most of waitRead#/Write# from cmm to C Moves it into the IOManager.c where we can follow the new pattern of switching on the selected I/O manager. - - - - - 457705a8 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move most of the delay# impl from cmm to C Moves it into the IOManager.c where we can follow the new pattern of switching on the selected I/O manager. Uses a new IOManager API: syncDelay, following the naming convention of sync* for thread-synchronous I/O & timer/delay operations. As part of porting from cmm to C, we maintain the rule that the why_blocked gets accessed using load acquire and store release atomic memory operations. There was one exception to this rule: in the delay# primop cmm code on posix (not win32), the why_blocked was being updated using a store relaxed, not a store release. I've no idea why. In this convesion I'm playing it safe here and using store release consistently. - - - - - e93058e0 by Duncan Coutts at 2024-04-03T01:27:18-04:00 insertIntoSleepingQueue is no longer public No longer defined in IOManager.h, just a private function in IOManager.c. Since it is no longer called from cmm code, just from syncDelay. It ought to get moved further into the select() I/O manager impl, rather than living in IOManager.c. On the other hand appendToIOBlockedQueue is still called from cmm code in the win32-legacy I/O manager primops async{Read,Write}#, and it is also used by the select() I/O manager. Update the CPP and comments to reflect this. - - - - - 60ce9910 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move anyPendingTimeoutsOrIO impl from .h to .c The implementation is eventually going to need to use more private things, which will drag in unwanted includes into IOManager.h, so it's better to move the impl out of the header file and into the .c file, at the slight cost of it no longer being inline. At the same time, change to the "switch (iomgr_type)" style. - - - - - f70b8108 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Take a simpler approach to gcc warnings in IOManager.c We have lots of functions with conditional implementations for different I/O managers. Some functions, for some I/O managers, naturally have implementations that do nothing or barf. When only one such I/O manager is enabled then the whole function implementation will have an implementation that does nothing or barfs. This then results in warnings from gcc that parameters are unused, or that the function should be marked with attribute noreturn (since barf does not return). The USED_IF_THREADS trick for fine-grained warning supression is fine for just two cases, but an equivalent here would need USED_IF_THE_ONLY_ENABLED_IOMGR_IS_X_OR_Y which would have combinitorial blowup. So we take a coarse grained approach and simply disable these two warnings for the whole file. So we use a GCC pragma, with its handy push/pop support: #pragma GCC diagnostic push #pragma GCC diagnostic ignored "-Wsuggest-attribute=noreturn" #pragma GCC diagnostic ignored "-Wunused-parameter" ... #pragma GCC diagnostic pop - - - - - b48805b9 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add a new trace class for the iomanager It makes sense now for it to be separate from the scheduler class of tracers. Enabled with +RTS -Do. Document the -Do debug flag in the user guide. - - - - - f0c1f862 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Have the throwTo impl go via (new) IOManager APIs rather than directly operating on the IO manager's data structures. Specifically, when thowing an async exception to a thread that is blocked waiting for I/O or waiting for a timer, then we want to cancel that I/O waiting or cancel the timer. Currently this is done directly in removeFromQueues() in RaiseAsync.c. We want it to go via proper APIs both for modularity but also to let us support multiple I/O managers. So add sync{IO,Delay}Cancel, which is the cancellation for the corresponding sync{IO,Delay}. The implementations of these use the usual "switch (iomgr_type)" style. - - - - - 4f9e9c4e by Duncan Coutts at 2024-04-03T01:27:18-04:00 Move awaitEvent into a proper IOManager API and have the scheduler use it. Previously the scheduler calls awaitEvent directly, and awaitEvent is implemented directly in the RTS I/O managers (select, win32). This relies on the old scheme where there's a single active I/O manager for each platform and RTS way. We want to move that to go via an API in IOManager.{h,c} which can then call out to the active I/O manager. Also take the opportunity to split awaitEvent into two. The existing awaitEvent has a bool wait parameter, to say if the call should be blocking or non-blocking. We split this into two separate functions: pollCompletedTimeoutsOrIO and awaitCompletedTimeoutsOrIO. We split them for a few reasons: they have different post-conditions (specifically the await version is supposed to guarantee that there are threads runnable when it completes). Secondly, it is also anticipated that in future I/O managers the implementations of the two cases will be simpler if they are separated. - - - - - 5ad4b30f by Duncan Coutts at 2024-04-03T01:27:18-04:00 Rename awaitEvent in select and win32 I/O managers These are now just called from IOManager.c and are the per-I/O manager backend impls (whereas previously awaitEvent was the entry point). Follow the new naming convention in the IOManager.{h,c} of awaitCompletedTimeoutsOrIO, with the I/O manager's name as a suffix: so awaitCompletedTimeoutsOrIO{Select,Win32}. - - - - - d30c6bc6 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Tidy up a couple things in Select.{h,c} Use the standard #include {Begin,End}Private.h style rather than RTS_PRIVATE on individual decls. And conditionally build the code for the select I/O manager based on the new CPP IOMGR_ENABLED_SELECT rather than on THREADED_RTS. - - - - - 4161f516 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add an IOManager API for scavenging TSO blocked_info When the GC scavenges a TSO it needs to scavenge the tso->blocked_info but the blocked_info is a big union and what lives there depends on the two->why_blocked, which for I/O-related reasons is something that in principle is the responsibility of the I/O manager and not the GC. So the right thing to do is for the GC to ask the I/O manager to sscavenge the blocked_info if it encounters any I/O-related why_blocked reasons. So we add scavengeTSOIOManager in IOManager.{h,c} with the usual style. Now as it happens, right now, there is no special scavenging to do, so the implementation of scavengeTSOIOManager is a fancy no-op. That's because the select I/O manager uses only the fd and target members, which are not GC pointers, and the win32-legacy I/O manager _ought_ to be using GC-managed heap objects for the StgAsyncIOResult but it is actually usingthe C heap, so again no GC pointers. If the win32-legacy were doing this more sensibly, then scavengeTSOIOManager would be the right place to do the GC magic. Future I/O managers will need GC heap objects in the tso->blocked_info and will make use of this functionality. - - - - - 94a87d21 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add I/O manager API notifyIOManagerCapabilitiesChanged Used in setNumCapabilities. It only does anything for MIO on Posix. Previously it always invoked Haskell code, but that code only did anything on non-Windows (and non-JS), and only threaded. That currently effectively means the MIO I/O manager on Posix. So now it only invokes it for the MIO Posix case. - - - - - 3be6d591 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Select an I/O manager early in RTS startup We need to select the I/O manager to use during startup before the per-cap I/O manager initialisation. - - - - - aaa294d0 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Make struct CapIOManager be fully opaque Provide an opaque (forward) definition in Capability.h (since the cap contains a *CapIOManager) and then only provide a full definition in a new file IOManagerInternals.h. This new file is only supposed to be included by the IOManager implementation, not by its users. So that means IOManager.c and individual I/O manager implementations. The posix/Signals.c still needs direct access, but that should be eliminated. Anything that needs direct access either needs to be clearly part of an I/O manager (e.g. the sleect() one) or go via a proper API. - - - - - 877a2a80 by Duncan Coutts at 2024-04-03T01:27:18-04:00 The select() I/O manager does have some global initialisation It's just to make sure an exception CAF is a GC root. - - - - - 9c51473b by Duncan Coutts at 2024-04-03T01:27:18-04:00 Add tracing for the main I/O manager actions Using the new tracer class. Note: The unconditional definition of showIOManager should be compatible with the debugTrace change in 7c7d1f6. Co-authored-by: Pi Delport <pi at well-typed.com> - - - - - c7d3e3a3 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Include the default I/O manager in the +RTS --info output Document the extra +RTS --info output in the user guide - - - - - 8023bad4 by Duncan Coutts at 2024-04-03T01:27:18-04:00 waitRead# / waitWrite# do not work for win32-legacy I/O manager Previously it was unclear that they did not work because the code path was shared with other I/O managers (in particular select()). Following the code carefully shows that what actually happens is that the calling thread would block forever: the thread will be put into the blocked queue, but no other action is scheduled that will ever result in it getting unblocked. It's better to just fail loudly in case anyone accidentally calls it, also it's less confusing code. - - - - - 83a74d20 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Conditionally ignore some GCC warnings Some GCC versions don't know about some warnings, and they complain that we're ignoring unknown warnings. So we try to ignore the warning based on the GCC version. - - - - - 1adc6fa4 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Accept changes to base-exports All the changes are in fact not changes at all. Previously, the IoSubSystem data type was defined in GHC.RTS.Flags and exported from both GHC.RTS.Flags and GHC.IO.SubSystem. Now, the data type is defined in GHC.IO.SubSystem and still exported from both modules. Therefore, the same exports and same instances are still available from both modules. But the base-exports records only the defining module, and so it looks like a change when it is fully compatible. Related: we do add a deprecation to the export of the type via GHC.RTS.Flags, telling people to use the export from GHC.IO.SubSystem. Also the sort order for some unrelated Show instances changed. No idea why. The same changes apply in the other versions, with a few more changes due to sort order weirdness. - - - - - 8d950968 by Duncan Coutts at 2024-04-03T01:27:18-04:00 Accept metric decrease in T12227 I can't think of any good reason that anything in this MR should have changed the number of allocations, up or down. (Yes this is an empty commit.) Metric Decrease: T12227 - - - - - e869605e by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Several improvements to the handling of coercions * Make `mkSymCo` and `mkInstCo` smarter Fixes #23642 * Fix return role of `SelCo` in the coercion optimiser. Fixes #23617 * Make the coercion optimiser `opt_trans_rule` work better for newtypes Fixes #23619 - - - - - 1efd0714 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 FloatOut: improve floating for join point See the new Note [Floating join point bindings]. * Completely get rid of the complicated join_ceiling nonsense, which I have never understood. * Do not float join points at all, except perhaps to top level. * Some refactoring around wantToFloat, to treat Rec and NonRec more uniformly - - - - - 9c00154d by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Improve eta-expansion through call stacks See Note [Eta expanding through CallStacks] in GHC.Core.Opt.Arity This is a one-line change, that fixes an inconsistency - || isCallStackPredTy ty + || isCallStackPredTy ty || isCallStackTy ty - - - - - 95a9a172 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Spelling, layout, pretty-printing only - - - - - bdf1660f by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Improve exprIsConApp_maybe a little Eliminate a redundant case at birth. This sometimes reduces Simplifier iterations. See Note [Case elim in exprIsConApp_maybe]. - - - - - 609cd32c by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Inline GHC.HsToCore.Pmc.Solver.Types.trvVarInfo When exploring compile-time regressions after meddling with the Simplifier, I discovered that GHC.HsToCore.Pmc.Solver.Types.trvVarInfo was very delicately balanced. It's a small, heavily used, overloaded function and it's important that it inlines. By a fluke it was before, but at various times in my journey it stopped doing so. So I just added an INLINE pragma to it; no sense in depending on a delicately-balanced fluke. - - - - - ae24c9bc by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Slight improvement in WorkWrap Ensure that WorkWrap preserves lambda binders, in case of join points. Sadly I have forgotten why I made this change (it was while I was doing a lot of meddling in the Simplifier, but * it does no harm, * it is slightly more efficient, and * presumably it made something better! Anyway I have kept it in a separate commit. - - - - - e9297181 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Use named record fields for the CastIt { ... } data constructor This is a pure refactor - - - - - b4581e23 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Remove a long-commented-out line Pure refactoring - - - - - e026bdf2 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Simplifier improvements This MR started as: allow the simplifer to do more in one pass, arising from places I could see the simplifier taking two iterations where one would do. But it turned into a larger project, because these changes unexpectedly made inlining blow up, especially join points in deeply-nested cases. The main changes are below. There are also many new or rewritten Notes. Avoiding simplifying repeatedly ~~~~~~~~~~~~~~~ See Note [Avoiding simplifying repeatedly] * The SimplEnv now has a seInlineDepth field, which says how deep in unfoldings we are. See Note [Inline depth] in Simplify.Env. Currently used only for the next point: avoiding repeatedly simplifying coercions. * Avoid repeatedly simplifying coercions. see Note [Avoid re-simplifying coercions] in Simplify.Iteration As you'll see from the Note, this makes use of the seInlineDepth. * Allow Simplify.Iteration.simplAuxBind to inline used-once things. This is another part of Note [Post-inline for single-use things], and is really good for reducing simplifier iterations in situations like case K e of { K x -> blah } wher x is used once in blah. * Make GHC.Core.SimpleOpt.exprIsConApp_maybe do some simple case elimination. Note [Case elim in exprIsConApp_maybe] * Improve the case-merge transformation: - Move the main code to `GHC.Core.Utils.mergeCaseAlts`, to join `filterAlts` and friends. See Note [Merge Nested Cases] in GHC.Core.Utils. - Add a new case for `tagToEnum#`; see wrinkle (MC3). - Add a new case to look through join points: see wrinkle (MC4) postInlineUnconditionally ~~~~~~~~~~~~~~~~~~~~~~~~~ * Allow Simplify.Utils.postInlineUnconditionally to inline variables that are used exactly once. See Note [Post-inline for single-use things]. * Do not postInlineUnconditionally join point, ever. Doing so does not reduce allocation, which is the main point, and with join points that are used a lot it can bloat code. See point (1) of Note [Duplicating join points] in GHC.Core.Opt.Simplify.Iteration. * Do not postInlineUnconditionally a strict (demanded) binding. It will not allocate a thunk (it'll turn into a case instead) so again the main point of inlining it doesn't hold. Better to check per-call-site. * Improve occurrence analyis for bottoming function calls, to help postInlineUnconditionally. See Note [Bottoming function calls] in GHC.Core.Opt.OccurAnal Inlining generally ~~~~~~~~~~~~~~~~~~ * In GHC.Core.Opt.Simplify.Utils.interestingCallContext, use RhsCtxt NonRecursive (not BoringCtxt) for a plain-seq case. See Note [Seq is boring] Also, wrinkle (SB1), inline in that `seq` context only for INLINE functions (UnfWhen guidance). * In GHC.Core.Opt.Simplify.Utils.interestingArg, - return ValueArg for OtherCon [c1,c2, ...], but - return NonTrivArg for OtherCon [] This makes a function a little less likely to inline if all we know is that the argument is evaluated, but nothing else. * isConLikeUnfolding is no longer true for OtherCon {}. This propagates to exprIsConLike. Con-like-ness has /positive/ information. Join points ~~~~~~~~~~~ * Be very careful about inlining join points. See these two long Notes Note [Duplicating join points] in GHC.Core.Opt.Simplify.Iteration Note [Inlining join points] in GHC.Core.Opt.Simplify.Inline * When making join points, don't do so if the join point is so small it will immediately be inlined; check uncondInlineJoin. * In GHC.Core.Opt.Simplify.Inline.tryUnfolding, improve the inlining heuristics for join points. In general we /do not/ want to inline join points /even if they are small/. See Note [Duplicating join points] GHC.Core.Opt.Simplify.Iteration. But sometimes we do: see Note [Inlining join points] in GHC.Core.Opt.Simplify.Inline; and the new `isBetterUnfoldingThan` function. * Do not add an unfolding to a join point at birth. This is a tricky one and has a long Note [Do not add unfoldings to join points at birth] It shows up in two places - In `mkDupableAlt` do not add an inlining - (trickier) In `simplLetUnfolding` don't add an unfolding for a fresh join point I am not fully satisifed with this, but it works and is well documented. * In GHC.Core.Unfold.sizeExpr, make jumps small, so that we don't penalise having a non-inlined join point. Performance changes ~~~~~~~~~~~~~~~~~~~ * Binary sizes fall by around 2.6%, according to nofib. * Compile times improve slightly. Here are the figures over 1%. I investiate the biggest differnce in T18304. It's a very small module, just a few hundred nodes. The large percentage difffence is due to a single function that didn't quite inline before, and does now, making code size a bit bigger. I decided gains outweighed the losses. Metrics: compile_time/bytes allocated (changes over +/- 1%) ------------------------------------------------ CoOpt_Singletons(normal) -9.2% GOOD LargeRecord(normal) -23.5% GOOD MultiComponentModulesRecomp(normal) +1.2% MultiLayerModulesTH_OneShot(normal) +4.1% BAD PmSeriesS(normal) -3.8% PmSeriesV(normal) -1.5% T11195(normal) -1.3% T12227(normal) -20.4% GOOD T12545(normal) -3.2% T12707(normal) -2.1% GOOD T13253(normal) -1.2% T13253-spj(normal) +8.1% BAD T13386(normal) -3.1% GOOD T14766(normal) -2.6% GOOD T15164(normal) -1.4% T15304(normal) +1.2% T15630(normal) -8.2% T15630a(normal) NEW T15703(normal) -14.7% GOOD T16577(normal) -2.3% GOOD T17516(normal) -39.7% GOOD T18140(normal) +1.2% T18223(normal) -17.1% GOOD T18282(normal) -5.0% GOOD T18304(normal) +10.8% BAD T18923(normal) -2.9% GOOD T1969(normal) +1.0% T19695(normal) -1.5% T20049(normal) -12.7% GOOD T21839c(normal) -4.1% GOOD T3064(normal) -1.5% T3294(normal) +1.2% BAD T4801(normal) +1.2% T5030(normal) -15.2% GOOD T5321Fun(normal) -2.2% GOOD T6048(optasm) -16.8% GOOD T783(normal) -1.2% T8095(normal) -6.0% GOOD T9630(normal) -4.7% GOOD T9961(normal) +1.9% BAD WWRec(normal) -1.4% info_table_map_perf(normal) -1.3% parsing001(normal) +1.5% geo. mean -2.0% minimum -39.7% maximum +10.8% * Runtimes generally improve. In the testsuite perf/should_run gives: Metrics: runtime/bytes allocated ------------------------------------------ Conversions(normal) -0.3% T13536a(optasm) -41.7% GOOD T4830(normal) -0.1% haddock.Cabal(normal) -0.1% haddock.base(normal) -0.1% haddock.compiler(normal) -0.1% geo. mean -0.8% minimum -41.7% maximum +0.0% * For runtime, nofib is a better test. The news is mostly good. Here are the number more than +/- 0.1%: # bytes allocated ==========================++========== imaginary/digits-of-e1 || -14.40% imaginary/digits-of-e2 || -4.41% imaginary/paraffins || -0.17% imaginary/rfib || -0.15% imaginary/wheel-sieve2 || -0.10% real/compress || -0.47% real/fluid || -0.10% real/fulsom || +0.14% real/gamteb || -1.47% real/gg || -0.20% real/infer || +0.24% real/pic || -0.23% real/prolog || -0.36% real/scs || -0.46% real/smallpt || +4.03% shootout/k-nucleotide || -20.23% shootout/n-body || -0.42% shootout/spectral-norm || -0.13% spectral/boyer2 || -3.80% spectral/constraints || -0.27% spectral/hartel/ida || -0.82% spectral/mate || -20.34% spectral/para || +0.46% spectral/rewrite || +1.30% spectral/sphere || -0.14% ==========================++========== geom mean || -0.59% real/smallpt has a huge nest of local definitions, and I could not pin down a reason for a regression. But there are three big wins! Metric Decrease: CoOpt_Singletons LargeRecord T12227 T12707 T13386 T13536a T14766 T15703 T16577 T17516 T18223 T18282 T18923 T21839c T20049 T5321Fun T5030 T6048 T8095 T9630 T783 Metric Increase: MultiLayerModulesTH_OneShot T13253-spj T18304 T18698a T9961 T3294 - - - - - 27db3c5e by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Testsuite message changes from simplifier improvements - - - - - 271a7812 by Simon Peyton Jones at 2024-04-03T01:27:55-04:00 Account for bottoming functions in OccurAnal This fixes #24582, a small but long-standing bug - - - - - 0fde229f by Ben Gamari at 2024-04-04T07:04:58-04:00 testsuite: Introduce template-haskell-exports test - - - - - 0c4a9686 by Luite Stegeman at 2024-04-04T07:05:39-04:00 Update correct counter in bumpTickyAllocd - - - - - 5f085d3a by Fendor at 2024-04-04T14:47:33-04:00 Replace `SizedSeq` with `FlatBag` for flattened structure LinkedLists are notoriously memory ineffiecient when all we do is traversing a structure. As 'UnlinkedBCO' has been identified as a data structure that impacts the overall memory usage of GHCi sessions, we avoid linked lists and prefer flattened structure for storing. We introduce a new memory efficient representation of sequential elements that has special support for the cases: * Empty * Singleton * Tuple Elements This improves sharing in the 'Empty' case and avoids the overhead of 'Array' until its constant overhead is justified. - - - - - 82cfe10c by Fendor at 2024-04-04T14:47:33-04:00 Compact FlatBag array representation `Array` contains three additional `Word`'s we do not need in `FlatBag`. Move `FlatBag` to `SmallArray`. Expand the API of SmallArray by `sizeofSmallArray` and add common traversal functions, such as `mapSmallArray` and `foldMapSmallArray`. Additionally, allow users to force the elements of a `SmallArray` via `rnfSmallArray`. - - - - - 36a75b80 by Andrei Borzenkov at 2024-04-04T14:48:10-04:00 Change how invisible patterns represented in haskell syntax and TH AST (#24557) Before this patch: data ArgPat p = InvisPat (LHsType p) | VisPat (LPat p) With this patch: data Pat p = ... | InvisPat (LHsType p) ... And the same transformation in the TH land. The rest of the changes is just updating code to handle new AST and writing tests to check if it is possible to create invalid states using TH. Metric Increase: MultiLayerModulesTH_OneShot - - - - - 28009fbc by Matthew Pickering at 2024-04-04T14:48:46-04:00 Fix off by one error in seekBinNoExpand and seekBin - - - - - 9b9e031b by Ben Gamari at 2024-04-04T21:30:08-04:00 compiler: Allow more types in GHCForeignImportPrim For many, many years `GHCForeignImportPrim` has suffered from the rather restrictive limitation of not allowing any non-trivial types in arguments or results. This limitation was justified by the code generator allegely barfing in the presence of such types. However, this restriction appears to originate well before the NCG rewrite and the new NCG does not appear to have any trouble with such types (see the added `T24598` test). Lift this restriction. Fixes #24598. - - - - - 1324b862 by Alan Zimmerman at 2024-04-04T21:30:44-04:00 EPA: Use EpaLocation not SrcSpan in ForeignDecls This allows us to update them for makeDeltaAst in ghc-exactprint - - - - - 19883a23 by Alan Zimmerman at 2024-04-05T16:58:17-04:00 EPA: Use EpaLocation for RecFieldsDotDot So we can update it to a delta position in makeDeltaAst if needed. - - - - - e8724327 by Matthew Pickering at 2024-04-05T16:58:53-04:00 Remove accidentally committed test.hs - - - - - 88cb3e10 by Fendor at 2024-04-08T09:03:34-04:00 Avoid UArray when indexing is not required `UnlinkedBCO`'s can occur many times in the heap. Each `UnlinkedBCO` references two `UArray`'s but never indexes them. They are only needed to encode the elements into a `ByteArray#`. The three words for the lower bound, upper bound and number of elements are essentially unused, thus we replace `UArray` with a wrapper around `ByteArray#`. This saves us up to three words for each `UnlinkedBCO`. Further, to avoid re-allocating these words for `ResolvedBCO`, we repeat the procedure for `ResolvedBCO` and add custom `Binary` and `Show` instances. For example, agda's repl session has around 360_000 UnlinkedBCO's, so avoiding these three words is already saving us around 8MB residency. - - - - - f2cc1107 by Fendor at 2024-04-08T09:04:11-04:00 Never UNPACK `FastMutInt` for counting z-encoded `FastString`s In `FastStringTable`, we count the number of z-encoded FastStrings that exist in a GHC session. We used to UNPACK the counters to not waste memory, but live retainer analysis showed that we allocate a lot of `FastMutInt`s, retained by `mkFastZString`. We lazily compute the `FastZString`, only incrementing the counter when the `FastZString` is forced. The function `mkFastStringWith` calls `mkZFastString` and boxes the `FastMutInt`, leading to the following core: mkFastStringWith = \ mk_fs _ -> = case stringTable of { FastStringTable _ n_zencs segments# _ -> ... case ((mk_fs (I# ...) (FastMutInt n_zencs)) `cast` <Co:2> :: ...) ... Marking this field as `NOUNPACK` avoids this reboxing, eliminating the allocation of a fresh `FastMutInt` on every `FastString` allocation. - - - - - c6def949 by Matthew Pickering at 2024-04-08T16:06:51-04:00 Force in_multi to avoid retaining entire hsc_env - - - - - fbb91a63 by Fendor at 2024-04-08T16:06:51-04:00 Eliminate name thunk in declaration fingerprinting Thunk analysis showed that we have about 100_000 thunks (in agda and `-fwrite-simplified-core`) pointing to the name of the name decl. Forcing this thunk fixes this issue. The thunk created here is retained by the thunk created by forkM, it is better to eagerly force this because the result (a `Name`) is already retained indirectly via the `IfaceDecl`. - - - - - 3b7b0c1c by Alan Zimmerman at 2024-04-08T16:07:27-04:00 EPA: Use EpaLocation in WarningTxt This allows us to use an EpDelta if needed when using makeDeltaAst. - - - - - 12b997df by Alan Zimmerman at 2024-04-08T16:07:27-04:00 EPA: Move DeltaPos and EpaLocation' into GHC.Types.SrcLoc This allows us to use a NoCommentsLocation for the possibly trailing comma location in a StringLiteral. This in turn allows us to correctly roundtrip via makeDeltaAst. - - - - - 868c8a78 by Fendor at 2024-04-09T08:51:50-04:00 Prefer packed representation for CompiledByteCode As there are many 'CompiledByteCode' objects alive during a GHCi session, representing its element in a more packed manner improves space behaviour at a minimal cost. When running GHCi on the agda codebase, we find around 380 live 'CompiledByteCode' objects. Packing their respective 'UnlinkedByteCode' can save quite some pointers. - - - - - be3bddde by Alan Zimmerman at 2024-04-09T08:52:26-04:00 EPA: Capture all comments in a ClassDecl Hopefully the final fix needed for #24533 - - - - - 3d0806fc by Jade at 2024-04-10T05:39:53-04:00 Validate -main-is flag using parseIdentifier Fixes #24368 - - - - - dd530bb7 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 rts: free error message before returning Fixes a memory leak in rts/linker/PEi386.c - - - - - e008a19a by Alexis King at 2024-04-10T05:40:29-04:00 linker: Avoid linear search when looking up Haskell symbols via dlsym See the primary Note [Looking up symbols in the relevant objects] for a more in-depth explanation. When dynamically loading a Haskell symbol (typical when running a splice or GHCi expression), before this commit we would search for the symbol in all dynamic libraries that were loaded. However, this could be very inefficient when too many packages are loaded (which can happen if there are many package dependencies) because the time to lookup the would be linear in the number of packages loaded. This commit drastically improves symbol loading performance by introducing a mapping from units to the handles of corresponding loaded dlls. These handles are returned by dlopen when we load a dll, and can then be used to look up in a specific dynamic library. Looking up a given Name is now much more precise because we can get lookup its unit in the mapping and lookup the symbol solely in the handles of the dynamic libraries loaded for that unit. In one measurement, the wait time before the expression was executed went from +-38 seconds down to +-2s. This commit also includes Note [Symbols may not be found in pkgs_loaded], explaining the fallback to the old behaviour in case no dll can be found in the unit mapping for a given Name. Fixes #23415 Co-authored-by: Rodrigo Mesquita (@alt-romes) - - - - - dcfaa190 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 rts: Make addDLL a wrapper around loadNativeObj Rewrite the implementation of `addDLL` as a wrapper around the more principled `loadNativeObj` rts linker function. The latter should be preferred while the former is preserved for backwards compatibility. `loadNativeObj` was previously only available on ELF platforms, so this commit further refactors the rts linker to transform loadNativeObj_ELF into loadNativeObj_POSIX, which is available in ELF and MachO platforms. The refactor made it possible to remove the `dl_mutex` mutex in favour of always using `linker_mutex` (rather than a combination of both). Lastly, we implement `loadNativeObj` for Windows too. - - - - - 12931698 by Rodrigo Mesquita at 2024-04-10T05:40:29-04:00 Use symbol cache in internal interpreter too This commit makes the symbol cache that was used by the external interpreter available for the internal interpreter too. This follows from the analysis in #23415 that suggests the internal interpreter could benefit from this cache too, and that there is no good reason not to have the cache for it too. It also makes it a bit more uniform to have the symbol cache range over both the internal and external interpreter. This commit also refactors the cache into a function which is used by both `lookupSymbol` and also by `lookupSymbolInDLL`, extending the caching logic to `lookupSymbolInDLL` too. - - - - - dccd3ea1 by Ben Gamari at 2024-04-10T05:40:29-04:00 testsuite: Add test for lookupSymbolInNativeObj - - - - - 1b1a92bd by Alan Zimmerman at 2024-04-10T05:41:05-04:00 EPA: Remove unnecessary XRec in CompleteMatchSig The XRec for [LIdP pass] is not needed for exact printing, remove it. - - - - - 6e18ce2b by Ben Gamari at 2024-04-12T08:16:09-04:00 users-guide: Clarify language extension documentation Over the years the users guide's language extension documentation has gone through quite a few refactorings. In the process some of the descriptions have been rendered non-sensical. For instance, the description of `NoImplicitPrelude` actually describes the semantics of `ImplicitPrelude`. To fix this we: * ensure that all extensions are named in their "positive" sense (e.g. `ImplicitPrelude` rather than `NoImplicitPrelude`). * rework the documentation to avoid flag-oriented wording like "enable" and "disable" * ensure that the polarity of the documentation is consistent with reality. Fixes #23895. - - - - - a933aff3 by Zubin Duggal at 2024-04-12T08:16:45-04:00 driver: Make `checkHomeUnitsClosed` faster The implementation of `checkHomeUnitsClosed` was traversing every single path in the unit dependency graph - this grows exponentially and quickly grows to be infeasible on larger unit dependency graphs. Instead we replace this with a faster implementation which follows from the specificiation of the closure property - there is a closure error if there are units which are both are both (transitively) depended upon by home units and (transitively) depend on home units, but are not themselves home units. To compute the set of units required for closure, we first compute the closure of the unit dependency graph, then the transpose of this closure, and find all units that are reachable from the home units in the transpose of the closure. - - - - - 23c3e624 by Andreas Klebinger at 2024-04-12T08:17:21-04:00 RTS: Emit warning when -M < -H Fixes #24487 - - - - - d23afb8c by Ben Gamari at 2024-04-12T08:17:56-04:00 testsuite: Add broken test for CApiFFI with -fprefer-bytecode See #24634. - - - - - a4bb3a51 by Ben Gamari at 2024-04-12T08:18:32-04:00 base: Deprecate GHC.Pack As proposed in #21461. Closes #21540. - - - - - 55eb8c98 by Ben Gamari at 2024-04-12T08:19:08-04:00 ghc-internal: Fix mentions of ghc-internal in deprecation warnings Closes #24609. - - - - - b0fbd181 by Ben Gamari at 2024-04-12T08:19:44-04:00 rts: Implement set_initial_registers for AArch64 Fixes #23680. - - - - - 14c9ec62 by Ben Gamari at 2024-04-12T08:20:20-04:00 ghcup-metadata: Use Debian 9 binaries on Ubuntu 16, 17 Closes #24646. - - - - - 35a1621e by Ben Gamari at 2024-04-12T08:20:55-04:00 Bump unix submodule to 2.8.5.1 Closes #24640. - - - - - a1c24df0 by Finley McIlwaine at 2024-04-12T08:21:31-04:00 Correct default -funfolding-use-threshold in docs - - - - - 0255d03c by Oleg Grenrus at 2024-04-12T08:22:07-04:00 FastString is a __Modified__ UTF-8 - - - - - c3489547 by Matthew Pickering at 2024-04-12T13:13:44-04:00 rts: Improve tracing message when nursery is resized It is sometimes more useful to know how much bigger or smaller the nursery got when it is resized. In particular I am trying to investigate situations where we end up with fragmentation due to the nursery (#24577) - - - - - 5e4f4ba8 by Simon Peyton Jones at 2024-04-12T13:14:20-04:00 Don't generate wrappers for `type data` constructors with StrictData Previously, the logic for checking if a data constructor needs a wrapper or not would take into account whether the constructor's fields have explicit strictness (e.g., `data T = MkT !Int`), but the logic would _not_ take into account whether `StrictData` was enabled. This meant that something like `type data T = MkT Int` would incorrectly generate a wrapper for `MkT` if `StrictData` was enabled, leading to the horrible errors seen in #24620. To fix this, we disable generating wrappers for `type data` constructors altogether. Fixes #24620. Co-authored-by: Ryan Scott <ryan.gl.scott at gmail.com> - - - - - dbdf1995 by Alex Mason at 2024-04-15T15:28:26+10:00 Implements MO_S_Mul2 and MO_U_Mul2 using the UMULH, UMULL and SMULH instructions for AArch64 Also adds a test for MO_S_Mul2 - - - - - 42bd0407 by Teo Camarasu at 2024-04-16T20:06:39-04:00 Make template-haskell a stage1 package Promoting template-haskell from a stage0 to a stage1 package means that we can much more easily refactor template-haskell. We implement this by duplicating the in-tree `template-haskell`. A new `template-haskell-next` library is autogenerated to mirror `template-haskell` `stage1:ghc` to depend on the new interface of the library including the `Binary` instances without adding an explicit dependency on `template-haskell`. This is controlled by the `bootstrap-th` cabal flag When building `template-haskell` modules as part of this vendoring we do not have access to quote syntax, so we cannot use variable quote notation (`'Just`). So we either replace these with hand-written `Name`s or hide the code behind CPP. We can remove the `th_hack` from hadrian, which was required when building stage0 packages using the in-tree `template-haskell` library. For more details see Note [Bootstrapping Template Haskell]. Resolves #23536 Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> Co-Authored-By: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 3d973e47 by Ben Gamari at 2024-04-16T20:07:15-04:00 Bump parsec submodule to 3.1.17.0 - - - - - 9d38bfa0 by Simon Peyton Jones at 2024-04-16T20:07:51-04:00 Clone CoVars in CorePrep This MR addresses #24463. It's all explained in the new Note [Cloning CoVars and TyVars] - - - - - 0fe2b410 by Andreas Klebinger at 2024-04-16T20:08:27-04:00 NCG: Fix a bug where we errounously removed a required jump instruction. Add a new method to the Instruction class to check if we can eliminate a jump in favour of fallthrough control flow. Fixes #24507 - - - - - 9f99126a by Teo Camarasu at 2024-04-16T20:09:04-04:00 Fix documentation preview from doc-tarball job - Include all the .html files and assets in the job artefacts - Include all the .pdf files in the job artefacts - Mark the artefact as an "exposed" artefact meaning it turns up in the UI. Resolves #24651 - - - - - 3a0642ea by Ben Gamari at 2024-04-16T20:09:39-04:00 rts: Ignore EINTR while polling in timerfd itimer implementation While the RTS does attempt to mask signals, it may be that a foreign library unmasks them. This previously caused benign warnings which we now ignore. See #24610. - - - - - 9a53cd3f by Alan Zimmerman at 2024-04-16T20:10:15-04:00 EPA: Add additional comments field to AnnsModule This is used in exact printing to store comments coming after the `where` keyword but before any comments allocated to imports or decls. It is used in ghc-exactprint, see https://github.com/alanz/ghc-exactprint/commit/44bbed311fd8f0d053053fef195bf47c17d34fa7 - - - - - e5c43259 by Bryan Richter at 2024-04-16T20:10:51-04:00 Remove unrunnable FreeBSD CI jobs FreeBSD runner supply is inelastic. Currently there is only one, and it's unavailable because of a hardware issue. - - - - - 914eb49a by Ben Gamari at 2024-04-16T20:11:27-04:00 rel-eng: Fix mktemp usage in recompress-all We need a temporary directory, not a file. - - - - - f30e4984 by Teo Camarasu at 2024-04-16T20:12:03-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 - - - - - d7a3d6b5 by Ben Gamari at 2024-04-16T20:12:39-04:00 template-haskell: Declare TH.Lib.Internal as not-home Rather than `hide`. Closes #24659. - - - - - 5eaa46e7 by Matthew Pickering at 2024-04-19T02:14:55-04:00 testsuite: Rename isCross() predicate to needsTargetWrapper() isCross() was a misnamed because it assumed that all cross targets would provide a target wrapper, but the two most common cross targets (javascript, wasm) don't need a target wrapper. Therefore we rename this predicate to `needsTargetWrapper()` so situations in the testsuite where we can check whether running executables requires a target wrapper or not. - - - - - 55a9d699 by Simon Peyton Jones at 2024-04-19T02:15:32-04:00 Do not float HNFs out of lambdas This MR adjusts SetLevels so that it is less eager to float a HNF (lambda or constructor application) out of a lambda, unless it gets to top level. Data suggests that this change is a small net win: * nofib bytes-allocated falls by -0.09% (but a couple go up) * perf/should_compile bytes-allocated falls by -0.5% * perf/should_run bytes-allocated falls by -0.1% See !12410 for more detail. When fiddling elsewhere, I also found that this patch had a huge positive effect on the (very delicate) test perf/should_run/T21839r But that improvement doesn't show up in this MR by itself. Metric Decrease: MultiLayerModulesRecomp T15703 parsing001 - - - - - f0701585 by Alan Zimmerman at 2024-04-19T02:16:08-04:00 EPA: Fix comments in mkListSyntaxTy0 Also extend the test to confirm. Addresses #24669, 1 of 4 - - - - - b01c01d4 by Serge S. Gulin at 2024-04-19T02:16:51-04:00 JS: set image `x86_64-linux-deb11-emsdk-closure` for build - - - - - c90c6039 by Alan Zimmerman at 2024-04-19T02:17:27-04:00 EPA: Provide correct span for PatBind And remove unused parameter in checkPatBind Contributes to #24669 - - - - - bee54c24 by Krzysztof Gogolewski at 2024-04-19T11:13:00+02:00 Update quantification order following GHC haskell/haddock#23764 - - - - - 2814eb89 by Ben Gamari at 2024-04-19T18:57:05+02:00 hypsrc-test: Fix output of PositionPragmas.html - - - - - 26036f96 by Alan Zimmerman at 2024-04-19T13:11:08-04:00 EPA: Fix span for PatBuilderAppType Include the location of the prefix @ in the span for InVisPat. Also removes unnecessary annotations from HsTP. Contributes to #24669 - - - - - dba03aab by Matthew Craven at 2024-04-19T13:11:44-04:00 testsuite: Give the pre_cmd for mhu-perf more time - - - - - d31fbf6c by Krzysztof Gogolewski at 2024-04-19T21:04:09-04:00 Fix quantification order for a `op` b and a %m -> b Fixes #23764 Implements https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0640-tyop-quantification-order.rst Updates haddock submodule. - - - - - 385cd1c4 by Sebastian Graf at 2024-04-19T21:04:45-04:00 Make `seq#` a magic Id and inline it in CorePrep (#24124) We can save much code and explanation in Tag Inference and StgToCmm by making `seq#` a known-key Magic Id in `GHC.Internal.IO` and inline this definition in CorePrep. See the updated `Note [seq# magic]`. I also implemented a new `Note [Flatten case-bind]` to get better code for otherwise nested case scrutinees. I renamed the contructors of `ArgInfo` to use an `AI` prefix in order to resolve the clash between `type CpeApp = CoreExpr` and the data constructor of `ArgInfo`, as well as fixed typos in `Note [CorePrep invariants]`. Fixes #24252 and #24124. - - - - - 275e41a9 by Jade at 2024-04-20T11:10:40-04:00 Put the newline after errors instead of before them This mainly has consequences for GHCi but also slightly alters how the output of GHC on the commandline looks. Fixes: #22499 - - - - - dd339c7a by Teo Camarasu at 2024-04-20T11:11:16-04:00 Remove unecessary stage0 packages Historically quite a few packages had to be stage0 as they depended on `template-haskell` and that was stage0. In #23536 we made it so that was no longer the case. This allows us to remove a bunch of packages from this list. A few still remain. A new version of `Win32` is required by `semaphore-compat`. Including `Win32` in the stage0 set requires also including `filepath` because otherwise Hadrian's dependency logic gets confused. Once our boot compiler has a newer version of `Win32` all of these will be able to be dropped. Resolves #24652 - - - - - 2f8e3a25 by Alan Zimmerman at 2024-04-20T11:11:52-04:00 EPA: Avoid duplicated comments in splice decls Contributes to #24669 - - - - - c70b9ddb by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: fix typos and namings (fixes #24602) You may noted that I've also changed term of ``` , global "h$vt_double" ||= toJExpr IntV ``` See "IntV" and ``` WaitReadOp -> \[] [fd] -> pure $ PRPrimCall $ returnS (app "h$waidRead" [fd]) ``` See "h$waidRead" - - - - - 3db54f9b by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: trivial checks for variable presence (fixes #24602) - - - - - 777f108f by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: fs module imported twice (by emscripten and by ghc-internal). ghc-internal import wrapped in a closure to prevent conflict with emscripten (fixes #24602) Better solution is to use some JavaScript module system like AMD, CommonJS or even UMD. It will be investigated at other issues. At first glance we should try UMD (See https://github.com/umdjs/umd) - - - - - a45a5712 by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: thread.js requires h$fds and h$fdReady to be declared for static code analysis, minimal code copied from GHCJS (fixes #24602) I've just copied some old pieces of GHCJS from publicly available sources (See https://github.com/Taneb/shims/blob/a6dd0202dcdb86ad63201495b8b5d9763483eb35/src/io.js#L607). Also I didn't put details to h$fds. I took minimal and left only its object initialization: `var h$fds = {};` - - - - - ad90bf12 by Serge S. Gulin at 2024-04-21T16:33:43+03:00 JS: heap and stack overflows reporting defined as js hard failure (fixes #24602) These errors were treated as a hard failure for browser application. The fix is trivial: just throw error. - - - - - 5962fa52 by Serge S. Gulin at 2024-04-21T16:33:44+03:00 JS: Stubs for code without actual implementation detected by Google Closure Compiler (fixes #24602) These errors were fixed just by introducing stubbed functions with throw for further implementation. - - - - - a0694298 by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Add externs to linker (fixes #24602) After enabling jsdoc and built-in google closure compiler types I was needed to deal with the following: 1. Define NodeJS-environment types. I've just copied minimal set of externs from semi-official repo (see https://github.com/externs/nodejs/blob/6c6882c73efcdceecf42e7ba11f1e3e5c9c041f0/v8/nodejs.js#L8). 2. Define Emscripten-environment types: `HEAP8`. Emscripten already provides some externs in our code but it supposed to be run in some module system. And its definitions do not work well in plain bundle. 3. We have some functions which purpose is to add to functions some contextual information via function properties. These functions should be marked as `modifies` to let google closure compiler remove calls if these functions are not used actually by call graph. Such functions are: `h$o`, `h$sti`, `h$init_closure`, `h$setObjInfo`. 4. STG primitives such as registries and stuff from `GHC.StgToJS`. `dXX` properties were already present at externs generator function but they are started from `7`, not from `1`. This message is related: `// fixme does closure compiler bite us here?` - - - - - e58bb29f by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: added both tests: for size and for correctness (fixes #24602) By some reason MacOS builds add to stderr messages like: Ignoring unexpected archive entry: __.SYMDEF ... However I left stderr to `/dev/null` for compatibility with linux CI builds. - - - - - 909f3a9c by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Disable js linker warning for empty symbol table to make js tests running consistent across environments - - - - - 83eb10da by Serge S. Gulin at 2024-04-21T16:34:07+03:00 JS: Add special preprocessor for js files due of needing to keep jsdoc comments (fixes #24602) Our js files have defined google closure compiler types at jsdoc entries but these jsdoc entries are removed by cpp preprocessor. I considered that reusing them in javascript-backend would be a nice thing. Right now haskell processor uses `-traditional` option to deal with comments and `//` operators. But now there are following compiler options: `-C` and `-CC`. You can read about them at GCC (see https://gcc.gnu.org/onlinedocs/gcc/Preprocessor-Options.html#index-CC) and CLang (see https://clang.llvm.org/docs/ClangCommandLineReference.html#cmdoption-clang-CC). It seems that `-CC` works better for javascript jsdoc than `-traditional`. At least it leaves `/* ... */` comments w/o changes. - - - - - e1cf8dc2 by brandon s allbery kf8nh at 2024-04-22T03:48:26-04:00 fix link in CODEOWNERS It seems that our local Gitlab no longer has documentation for the `CODEOWNERS` file, but the master documentation still does. Use that instead. - - - - - a27c6a49 by Fendor at 2024-04-22T10:13:03+02:00 Adapt to UserData split - - - - - 1efc5a7a by Fendor at 2024-04-22T10:13:03+02:00 Adapt to BinHandle split - - - - - 593f4e04 by Fendor at 2024-04-23T10:19:14-04:00 Add performance regression test for '-fwrite-simplified-core' - - - - - 1ba39b05 by Fendor at 2024-04-23T10:19:14-04:00 Typecheck corebindings lazily during bytecode generation This delays typechecking the corebindings until the bytecode generation happens. We also avoid allocating a thunk that is retained by `unsafeInterleaveIO`. In general, we shouldn't retain values of the hydrated `Type`, as not evaluating the bytecode object keeps it alive. It is better if we retain the unhydrated `IfaceType`. See Note [Hydrating Modules] - - - - - e916fc92 by Alan Zimmerman at 2024-04-23T10:19:50-04:00 EPA: Keep comments in a CaseAlt match The comments now live in the surrounding location, not inside the Match. Make sure we keep them. Closes #24707 - - - - - d2b17f32 by Cheng Shao at 2024-04-23T15:01:22-04:00 driver: force merge objects when building dynamic objects This patch forces the driver to always merge objects when building dynamic objects even when ar -L is supported. It is an oversight of !8887: original rationale of that patch is favoring the relatively cheap ar -L operation over object merging when ar -L is supported, which makes sense but only if we are building static objects! Omitting check for whether we are building dynamic objects will result in broken .so files with undefined reference errors at executable link time when building GHC with llvm-ar. Fixes #22210. - - - - - 209d09f5 by Julian Ospald at 2024-04-23T15:02:03-04:00 Allow non-absolute values for bootstrap GHC variable Fixes #24682 - - - - - 3fff0977 by Matthew Pickering at 2024-04-23T15:02:38-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. - - - - - c62dc317 by Cheng Shao at 2024-04-25T01:32:02-04:00 ghc-bignum: remove obsolete ln script This commit removes an obsolete ln script in ghc-bignum/gmp. See 060251c24ad160264ae8553efecbb8bed2f06360 for its original intention, but it's been obsolete for a long time, especially since the removal of the make build system. Hence the house cleaning. - - - - - 6399d52b by Cheng Shao at 2024-04-25T01:32:02-04:00 ghc-bignum: update gmp to 6.3.0 This patch bumps the gmp-tarballs submodule and updates gmp to 6.3.0. The tarball format is now xz, and gmpsrc.patch has been patched into the tarball so hadrian no longer needs to deal with patching logic when building in-tree GMP. - - - - - 65b4b92f by Cheng Shao at 2024-04-25T01:32:02-04:00 hadrian: remove obsolete Patch logic This commit removes obsolete Patch logic from hadrian, given we no longer need to patch the gmp tarball when building in-tree GMP. - - - - - 71f28958 by Cheng Shao at 2024-04-25T01:32:02-04:00 autoconf: remove obsolete patch detection This commit removes obsolete deletection logic of the patch command from autoconf scripts, given we no longer need to patch anything in the GHC build process. - - - - - daeda834 by Sylvain Henry at 2024-04-25T01:32:43-04:00 JS: correctly handle RUBBISH literals (#24664) - - - - - 8a06ddf6 by Matthew Pickering at 2024-04-25T11:16:16-04:00 Linearise ghc-internal and base build This is achieved by requesting the final package database for ghc-internal, which mandates it is fully built as a dependency of configuring the `base` package. This is at the expense of cross-package parrallelism between ghc-internal and the base package. Fixes #24436 - - - - - 94da9365 by Andrei Borzenkov at 2024-04-25T11:16:54-04:00 Fix tuple puns renaming (24702) Move tuple renaming short cutter from `isBuiltInOcc_maybe` to `isPunOcc_maybe`, so we consider incoming module. I also fixed some hidden bugs that raised after the change was done. - - - - - fa03b1fb by Fendor at 2024-04-26T18:03:13-04:00 Refactor the Binary serialisation interface The goal is simplifiy adding deduplication tables to `ModIface` interface serialisation. We identify two main points of interest that make this difficult: 1. UserData hardcodes what `Binary` instances can have deduplication tables. Moreover, it heavily uses partial functions. 2. GHC.Iface.Binary hardcodes the deduplication tables for 'Name' and 'FastString', making it difficult to add more deduplication. Instead of having a single `UserData` record with fields for all the types that can have deduplication tables, we allow to provide custom serialisers for any `Typeable`. These are wrapped in existentials and stored in a `Map` indexed by their respective `TypeRep`. The `Binary` instance of the type to deduplicate still needs to explicitly look up the decoder via `findUserDataReader` and `findUserDataWriter`, which is no worse than the status-quo. `Map` was chosen as microbenchmarks indicate it is the fastest for a small number of keys (< 10). To generalise the deduplication table serialisation mechanism, we introduce the types `ReaderTable` and `WriterTable` which provide a simple interface that is sufficient to implement a general purpose deduplication mechanism for `writeBinIface` and `readBinIface`. This allows us to provide a list of deduplication tables for serialisation that can be extended more easily, for example for `IfaceTyCon`, see the issue https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for more motivation. In addition to this refactoring, we split `UserData` into `ReaderUserData` and `WriterUserData`, to avoid partial functions and reduce overall memory usage, as we need fewer mutable variables. Bump haddock submodule to accomodate for `UserData` split. ------------------------- Metric Increase: MultiLayerModulesTH_Make MultiLayerModulesRecomp T21839c ------------------------- - - - - - bac57298 by Fendor at 2024-04-26T18:03:13-04:00 Split `BinHandle` into `ReadBinHandle` and `WriteBinHandle` A `BinHandle` contains too much information for reading data. For example, it needs to keep a `FastMutInt` and a `IORef BinData`, when the non-mutable variants would suffice. Additionally, this change has the benefit that anyone can immediately tell whether the `BinHandle` is used for reading or writing. Bump haddock submodule BinHandle split. - - - - - 4d6394dd by Simon Peyton Jones at 2024-04-26T18:03:49-04:00 Fix missing escaping-kind check in tcPatSynSig Note [Escaping kind in type signatures] explains how we deal with escaping kinds in type signatures, e.g. f :: forall r (a :: TYPE r). a where the kind of the body is (TYPE r), but `r` is not in scope outside the forall-type. I had missed this subtlety in tcPatSynSig, leading to #24686. This MR fixes it; and a similar bug in tc_top_lhs_type. (The latter is tested by T24686a.) - - - - - 981c2c2c by Alan Zimmerman at 2024-04-26T18:04:25-04:00 EPA: check-exact: check that the roundtrip reproduces the source Closes #24670 - - - - - a8616747 by Andrew Lelechenko at 2024-04-26T18:05:01-04:00 Document that setEnv is not thread-safe - - - - - 1e41de83 by Bryan Richter at 2024-04-26T18:05:37-04:00 CI: Work around frequent Signal 9 errors - - - - - a6d5f9da by Naïm Favier at 2024-04-27T17:52:40-04:00 ghc-internal: add MonadFix instance for (,) Closes https://gitlab.haskell.org/ghc/ghc/-/issues/24288, implements CLC proposal https://github.com/haskell/core-libraries-committee/issues/238. Adds a MonadFix instance for tuples, permitting value recursion in the "native" writer monad and bringing consistency with the existing instance for transformers's WriterT (and, to a lesser extent, for Solo). - - - - - 64feadcd by Rodrigo Mesquita at 2024-04-27T17:53:16-04:00 bindist: Fix xattr cleaning The original fix (725343aa) was incorrect because it used the shell bracket syntax which is the quoting syntax in autoconf, making the test for existence be incorrect and therefore `xattr` was never run. Fixes #24554 - - - - - e2094df3 by damhiya at 2024-04-28T23:52:00+09:00 Make read accepts binary integer formats CLC proposal : https://github.com/haskell/core-libraries-committee/issues/177 - - - - - c62239b7 by Sylvain Henry at 2024-04-29T10:35:00+02:00 Fix tests for T22229 - - - - - 1c2fd963 by Alan Zimmerman at 2024-04-29T23:17:00-04:00 EPA: Preserve comments in Match Pats Closes #24708 Closes #24715 Closes #24734 - - - - - 4189d17e by Sylvain Henry at 2024-04-29T23:17:42-04:00 LLVM: better unreachable default destination in Switch (#24717) See added note. Co-authored-by: Siddharth Bhat <siddu.druid at gmail.com> - - - - - a3725c88 by Cheng Shao at 2024-04-29T23:18:20-04:00 ci: enable wasm jobs for MRs with wasm label This patch enables wasm jobs for MRs with wasm label. Previously the wasm label didn't actually have any effect on the CI pipeline, and full-ci needed to be applied to run wasm jobs which was a waste of runners when working on the wasm backend, hence the fix here. - - - - - 702f7964 by Matthew Pickering at 2024-04-29T23:18:56-04:00 Make interface files and object files depend on inplace .conf file A potential fix for #24737 - - - - - 728af21e by Cheng Shao at 2024-04-30T05:30:23-04:00 utils: remove obsolete vagrant scripts Vagrantfile has long been removed in !5288. This commit further removes the obsolete vagrant scripts in the tree. - - - - - 36f2c342 by Cheng Shao at 2024-04-30T05:31:00-04:00 Update autoconf scripts Scripts taken from autoconf 948ae97ca5703224bd3eada06b7a69f40dd15a02 - - - - - ecbf22a6 by Ben Gamari at 2024-04-30T05:31:36-04:00 ghcup-metadata: Drop output_name field This is entirely redundant to the filename of the URL. There is no compelling reason to name the downloaded file differently from its source. - - - - - c56d728e by Zubin Duggal at 2024-04-30T22:45:09-04:00 testsuite: Handle exceptions in framework_fail when testdir is not initialised When `framework_fail` is called before initialising testdir, it would fail with an exception reporting the testdir not being initialised instead of the actual failure. Ensure we report the actual reason for the failure instead of failing in this way. One way this can manifest is when trying to run a test that doesn't exist using `--only` - - - - - d5bea4d6 by Alan Zimmerman at 2024-04-30T22:45:45-04:00 EPA: Fix range for GADT decl with sig only Closes #24714 - - - - - 4d78c53c by Sylvain Henry at 2024-05-01T17:23:06-04:00 Fix TH dependencies (#22229) Add a dependency between Syntax and Internal (via module reexport). - - - - - 37e38db4 by Sylvain Henry at 2024-05-01T17:23:06-04:00 Bump haddock submodule - - - - - ca13075c by Sylvain Henry at 2024-05-01T17:23:47-04:00 JS: cleanup to prepare for #24743 - - - - - 40026ac3 by Alan Zimmerman at 2024-05-01T22:45:07-04:00 EPA: Preserve comments for PrefixCon Preserve comments in fun (Con {- c1 -} a b) = undefined Closes #24736 - - - - - 92134789 by Hécate Moonlight at 2024-05-01T22:45:42-04:00 Correct `@since` metadata in HpcFlags It was introduced in base-4.20, not 4.22. Fix #24721 - - - - - a580722e by Cheng Shao at 2024-05-02T08:18:45-04:00 testsuite: fix req_target_smp predicate - - - - - ac9c5f84 by Andreas Klebinger at 2024-05-02T08:18:45-04:00 STM: Remove (unused)coarse grained locking. The STM code had a coarse grained locking mode guarded by #defines that was unused. This commit removes the code. - - - - - 917ef81b by Andreas Klebinger at 2024-05-02T08:18:45-04:00 STM: Be more optimistic when validating in-flight transactions. * Don't lock tvars when performing non-committal validation. * If we encounter a locked tvar don't consider it a failure. This means in-flight validation will only fail if committing at the moment of validation is *guaranteed* to fail. This prevents in-flight validation from failing spuriously if it happens in parallel on multiple threads or parallel to thread comitting. - - - - - 167a56a0 by Alan Zimmerman at 2024-05-02T08:19:22-04:00 EPA: fix span for empty \case(s) In instance SDecide Nat where SZero %~ (SSucc _) = Disproved (\case) Ensure the span for the HsLam covers the full construct. Closes #24748 - - - - - 9bae34d8 by doyougnu at 2024-05-02T15:41:08-04:00 testsuite: expand size testing infrastructure - closes #24191 - adds windows_skip, wasm_skip, wasm_arch, find_so, _find_so - path_from_ghcPkg, collect_size_ghc_pkg, collect_object_size, find_non_inplace functions to testsuite - adds on_windows and req_dynamic_ghc predicate to testsuite The design is to not make the testsuite too smart and simply offload to ghc-pkg for locations of object files and directories. - - - - - b85b1199 by Sylvain Henry at 2024-05-02T15:41:49-04:00 GHCi: support inlining breakpoints (#24712) When a breakpoint is inlined, its context may change (e.g. tyvars in scope). We must take this into account and not used the breakpoint tick index as its sole identifier. Each instance of a breakpoint (even with the same tick index) now gets a different "info" index. We also need to distinguish modules: - tick module: module with the break array (tick counters, status, etc.) - info module: module having the CgBreakInfo (info at occurrence site) - - - - - 649c24b9 by Oleg Grenrus at 2024-05-03T20:45:42-04:00 Expose constructors of SNat, SChar and SSymbol in ghc-internal - - - - - d603f199 by Mikolaj Konarski at 2024-05-03T20:46:19-04:00 Add DCoVarSet to PluginProv (!12037) - - - - - ba480026 by Serge S. Gulin at 2024-05-03T20:47:01-04:00 JS: Enable more efficient packing of string data (fixes #24706) - - - - - be1e60ee by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Track in-scope variables in ruleCheckProgram This small patch fixes #24726, by tracking in-scope variables properly in -drule-check. Not hard to do! - - - - - 58408c77 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Add a couple more HasCallStack constraints in SimpleOpt Just for debugging, no effect on normal code - - - - - 70e245e8 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Add comments to Prep.hs This documentation patch fixes a TODO left over from !12364 - - - - - e5687186 by Simon Peyton Jones at 2024-05-03T20:47:37-04:00 Use HasDebugCallStack, rather than HasCallStack - - - - - 631cefec by Cheng Shao at 2024-05-03T20:48:17-04:00 driver: always merge objects when possible This patch makes the driver always merge objects with `ld -r` when possible, and only fall back to calling `ar -L` when merge objects command is unavailable. This completely reverts !8887 and !12313, given more fixes in Cabal seems to be needed to avoid breaking certain configurations and the maintainence cost is exceeding the behefits in this case :/ - - - - - 1dacb506 by Ben Gamari at 2024-05-03T20:48:53-04:00 Bump time submodule to 1.14 As requested in #24528. ------------------------- Metric Decrease: ghc_bignum_so rts_so Metric Increase: cabal_syntax_dir rts_so time_dir time_so ------------------------- - - - - - 4941b90e by Ben Gamari at 2024-05-03T20:48:53-04:00 Bump terminfo submodule to current master - - - - - 43d48b44 by Cheng Shao at 2024-05-03T20:49:30-04:00 wasm: use scheduler.postTask() for context switch when available This patch makes use of scheduler.postTask() for JSFFI context switch when it's available. It's a more principled approach than our MessageChannel based setImmediate() implementation, and it's available in latest version of Chromium based browsers. - - - - - 08207501 by Cheng Shao at 2024-05-03T20:50:08-04:00 testsuite: give pre_cmd for mhu-perf 5x time - - - - - bf3d4db0 by Alan Zimmerman at 2024-05-03T20:50:43-04:00 EPA: Preserve comments for pattern synonym sig Closes #24749 - - - - - c49493f2 by Matthew Pickering at 2024-05-04T06:02:57-04:00 tests: Widen acceptance window for dir and so size tests These are testing things which are sometimes out the control of a GHC developer. Therefore we shouldn't fail CI if something about these dependencies change because we can't do anything about it. It is still useful to have these statistics for visualisation in grafana though. Ticket #24759 - - - - - 9562808d by Matthew Pickering at 2024-05-04T06:02:57-04:00 Disable rts_so test It has already manifested large fluctuations and destabilising CI Fixes #24762 - - - - - fc24c5cf by Ryan Scott at 2024-05-04T06:03:33-04:00 unboxedSum{Type,Data}Name: Use GHC.Types as the module Unboxed sum constructors are now defined in the `GHC.Types` module, so if you manually quote an unboxed sum (e.g., `''Sum2#`), you will get a `Name` like: ```hs GHC.Types.Sum2# ``` The `unboxedSumTypeName` function in `template-haskell`, however, mistakenly believes that unboxed sum constructors are defined in `GHC.Prim`, so `unboxedSumTypeName 2` would return an entirely different `Name`: ```hs GHC.Prim.(#|#) ``` This is a problem for Template Haskell users, as it means that they can't be sure which `Name` is the correct one. (Similarly for `unboxedSumDataName`.) This patch fixes the implementations of `unboxedSum{Type,Data}Name` to use `GHC.Types` as the module. For consistency with `unboxedTupleTypeName`, the `unboxedSumTypeName` function now uses the non-punned syntax for unboxed sums (`Sum<N>#`) as the `OccName`. Fixes #24750. - - - - - 7eab4e01 by Alan Zimmerman at 2024-05-04T16:14:55+01:00 EPA: Widen stmtslist to include last semicolon Closes #24754 - - - - - 06f7db40 by Teo Camarasu at 2024-05-05T00:19:38-04:00 doc: Fix type error in hs_try_putmvar example - - - - - af000532 by Moritz Schuler at 2024-05-05T06:30:58-04:00 Fix parsing of module names in CLI arguments closes issue #24732 - - - - - da74e9c9 by Ben Gamari at 2024-05-05T06:31:34-04:00 ghc-platform: Add Setup.hs The Hadrian bootstrapping script relies upon `Setup.hs` to drive its build. Addresses #24761. - - - - - 35d34fde by Alan Zimmerman at 2024-05-05T12:52:40-04:00 EPA: preserve comments in class and data decls Fix checkTyClHdr which was discarding comments. Closes #24755 - - - - - 03c5dfbf by Simon Peyton Jones at 2024-05-05T12:53:15-04:00 Fix a float-out error Ticket #24768 showed that the Simplifier was accidentally destroying a join point. It turned out to be that we were sending a bottoming join point to the top, accidentally abstracting over /other/ join points. Easily fixed. - - - - - adba68e7 by John Ericson at 2024-05-05T19:35:56-04:00 Substitute bindist files with Hadrian not configure The `ghc-toolchain` overhaul will eventually replace all this stuff with something much more cleaned up, but I think it is still worth making this sort of cleanup in the meantime so other untanglings and dead code cleaning can procede. I was able to delete a fair amount of dead code doing this too. `LLVMTarget_CPP` is renamed to / merged with `LLVMTarget` because it wasn't actually turned into a valid CPP identifier. (Original to 1345c7cc42c45e63ab1726a8fd24a7e4d4222467, actually.) Progress on #23966 Co-Authored-By: Sylvain Henry <hsyl20 at gmail.com> - - - - - 18f4ff84 by Alan Zimmerman at 2024-05-05T19:36:32-04:00 EPA: fix mkHsOpTyPV duplicating comments Closes #24753 - - - - - a19201d4 by Matthew Craven at 2024-05-06T19:54:29-04:00 Add test cases for #24664 ...since none are present in the original MR !12463 fixing this issue. - - - - - 46328a49 by Alan Zimmerman at 2024-05-06T19:55:05-04:00 EPA: preserve comments in data decls Closes #24771 - - - - - 3b51995c by Andrei Borzenkov at 2024-05-07T14:39:40-04:00 Rename Solo# data constructor to MkSolo# (#24673) - data Solo# a = (# a #) + data Solo# a = MkSolo# a And `(# foo #)` syntax now becomes just a syntactic sugar for `MkSolo# a`. - - - - - 4d59abf2 by Arsen Arsenović at 2024-05-07T14:40:24-04:00 Add the cmm_cpp_is_gcc predicate to the testsuite A future C-- test called T24474-cmm-override-g0 relies on the GCC-specific behaviour of -g3 implying -dD, which, in turn, leads to it emitting #defines past the preprocessing stage. Clang, at least, does not do this, so the test would fail if ran on Clang. As the behaviour here being tested is ``-optCmmP-g3'' undoing effects of the workaround we apply as a fix for bug #24474, and the workaround was for GCC-specific behaviour, the test needs to be marked as fragile on other compilers. - - - - - 25b0b404 by Arsen Arsenović at 2024-05-07T14:40:24-04:00 Split out the C-- preprocessor, and make it pass -g0 Previously, C-- was processed with the C preprocessor program. This means that it inherited flags passed via -optc. A flag that is somewhat often passed through -optc is -g. At certain -g levels (>=2), GCC starts emitting defines *after* preprocessing, for the purposes of debug info generation. This is not useful for the C-- compiler, and, in fact, causes lexer errors. We can suppress this effect (safely, if supported) via -g0. As a workaround, in older versions of GCC (<=10), GCC only emitted defines if a certain set of -g*3 flags was passed. Newer versions check the debug level. For the former, we filter out those -g*3 flags and, for the latter, we specify -g0 on top of that. As a compatible and effective solution, this change adds a C-- preprocessor distinct from the C compiler and preprocessor, but that keeps its flags. The command line produced for C-- preprocessing now looks like: $pgmCmmP $optCs_without_g3 $g0_if_supported $optCmmP Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/24474 - - - - - 9b4129a5 by Andreas Klebinger at 2024-05-08T13:24:20-04:00 -fprof-late: Only insert cost centres on functions/non-workfree cafs. They are usually useless and doing so for data values comes with a large compile time/code size overhead. Fixes #24103 - - - - - 259b63d3 by Sebastian Graf at 2024-05-08T13:24:57-04:00 Simplifier: Preserve OccInfo on DataAlt fields when case binder is dead (#24770) See the adjusted `Note [DataAlt occ info]`. This change also has a positive repercussion on `Note [Combine case alts: awkward corner]`. Fixes #24770. We now try not to call `dataConRepStrictness` in `adjustFieldsIdInfo` when all fields are lazy anyway, leading to a 2% ghc/alloc decrease in T9675. Metric Decrease: T9675 - - - - - 31b28cdb by Sebastian Graf at 2024-05-08T13:24:57-04:00 Kill seqRule, discard dead seq# in Prep (#24334) Discarding seq#s in Core land via `seqRule` was problematic; see #24334. So instead we discard certain dead, discardable seq#s in Prep now. See the updated `Note [seq# magic]`. This fixes the symptoms of #24334. - - - - - b2682534 by Rodrigo Mesquita at 2024-05-10T01:47:51-04:00 Document NcgImpl methods Fixes #19914 - - - - - 4d3acbcf by Zejun Wu at 2024-05-10T01:48:28-04:00 Make renamer to be more flexible with parens in the LHS of the rules We used to reject LHS like `(f a) b` in RULES and requires it to be written as `f a b`. It will be handy to allow both as the expression may be more readable with extra parens in some cases when infix operator is involved. Espceially when TemplateHaskell is used, extra parens may be added out of user's control and result in "valid" rules being rejected and there are not always ways to workaround it. Fixes #24621 - - - - - ab840ce6 by Ben Gamari at 2024-05-10T01:49:04-04:00 IPE: Eliminate dependency on Read Instead of encoding the closure type as decimal string we now simply represent it as an integer, eliminating the need for `Read` in `GHC.Internal.InfoProv.Types.peekInfoProv`. Closes #24504. ------------------------- Metric Decrease: T24602_perf_size size_hello_artifact ------------------------- - - - - - a9979f55 by Cheng Shao at 2024-05-10T01:49:43-04:00 testsuite: fix testwsdeque with recent clang This patch fixes compilation of testwsdeque.c with recent versions of clang, which will fail with the error below: ``` testwsdeque.c:95:33: error: warning: format specifies type 'long' but the argument has type 'void *' [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~~ ^ testwsdeque.c:95:39: error: warning: format specifies type 'int' but the argument has type 'StgWord' (aka 'unsigned long') [-Wformat] 95 | barf("FAIL: %ld %d %d", p, n, val); | ~~ ^~~ | %lu testwsdeque.c:133:42: error: error: incompatible function pointer types passing 'void (void *)' to parameter of type 'OSThreadProc *' (aka 'void *(*)(void *)') [-Wincompatible-function-pointer-types] 133 | createOSThread(&ids[n], "thief", thief, (void*)(StgWord)n); | ^~~~~ /workspace/ghc/_build/stage1/lib/../lib/x86_64-linux-ghc-9.11.20240502/rts-1.0.2/include/rts/OSThreads.h:193:51: error: note: passing argument to parameter 'startProc' here 193 | OSThreadProc *startProc, void *param); | ^ 2 warnings and 1 error generated. ``` - - - - - c2b33fc9 by Rodrigo Mesquita at 2024-05-10T01:50:20-04:00 Rename pre-processor invocation args Small clean up. Uses proper names for the various groups of arguments that make up the pre-processor invocation. - - - - - 2b1af08b by Cheng Shao at 2024-05-10T01:50:55-04:00 ghc-heap: fix typo in ghc-heap cbits - - - - - fc2d6de1 by Jade at 2024-05-10T21:07:16-04:00 Improve performance of Data.List.sort(By) This patch improves the algorithm to sort lists in base. It does so using two strategies: 1) Use a four-way-merge instead of the 'default' two-way-merge. This is able to save comparisons and allocations. 2) Use `(>) a b` over `compare a b == GT` and allow inlining and specialization. This mainly benefits types with a fast (>). Note that this *may* break instances with a *malformed* Ord instance where `a > b` is *not* equal to `compare a b == GT`. CLC proposal: https://github.com/haskell/core-libraries-committee/issues/236 Fixes #24280 ------------------------- Metric Decrease: MultiLayerModulesTH_Make T10421 T13719 T15164 T18698a T18698b T1969 T9872a T9961 T18730 WWRec T12425 T15703 ------------------------- - - - - - 1012e8aa by Matthew Pickering at 2024-05-10T21:07:52-04:00 Revert "ghcup-metadata: Drop output_name field" This reverts commit ecbf22a6ac397a791204590f94c0afa82e29e79f. This breaks the ghcup metadata generation on the nightly jobs. - - - - - daff1e30 by Jannis at 2024-05-12T13:38:35-04:00 Division by constants optimization - - - - - 413217ba by Andreas Klebinger at 2024-05-12T13:39:11-04:00 Tidy: Add flag to expose unfoldings if they take dictionary arguments. Add the flag `-fexpose-overloaded-unfoldings` to be able to control this behaviour. For ghc's boot libraries file size grew by less than 1% when it was enabled. However I refrained from enabling it by default for now. I've also added a section on specialization more broadly to the users guide. ------------------------- Metric Decrease: MultiLayerModulesTH_OneShot Metric Increase: T12425 T13386 hard_hole_fits ------------------------- - - - - - c5d89412 by Zubin Duggal at 2024-05-13T22:19:53-04:00 Don't store a GlobalRdrEnv in `mi_globals` for GHCi. GHCi only needs the `mi_globals` field for modules imported with :module +*SomeModule. It uses this field to make the top level environment in `SomeModule` available to the repl. By default, only the first target in the command line parameters is "star" loaded into GHCi. Other modules have to be manually "star" loaded into the repl. Storing the top level GlobalRdrEnv for each module is very wasteful, especially given that we will most likely never need most of these environments. Instead we store only the information needed to reconstruct the top level environment in a module, which is the `IfaceTopEnv` data structure, consisting of all import statements as well as all top level symbols defined in the module (not taking export lists into account) When a particular module is "star-loaded" into GHCi (as the first commandline target, or via an explicit `:module +*SomeModule`, we reconstruct the top level environment on demand using the `IfaceTopEnv`. - - - - - d65bf4a2 by Fendor at 2024-05-13T22:20:30-04:00 Add perf regression test for `-fwrite-if-simplified-core` - - - - - 2c0f8ddb by Andrei Borzenkov at 2024-05-13T22:21:07-04:00 Improve pattern to type pattern transformation (23739) `pat_to_type_pat` function now can handle more patterns: - TuplePat - ListPat - LitPat - NPat - ConPat Allowing these new constructors in type patterns significantly increases possible shapes of type patterns without `type` keyword. This patch also changes how lookups in `lookupOccRnConstr` are performed, because we need to fall back into types when we didn't find a constructor on data level to perform `ConPat` to type transformation properly. - - - - - be514bb4 by Cheng Shao at 2024-05-13T22:21:43-04:00 hadrian: fix hadrian building with ghc-9.10.1 - - - - - ad38e954 by Cheng Shao at 2024-05-13T22:21:43-04:00 linters: fix lint-whitespace compilation with ghc-9.10.1 - - - - - a593f284 by Andreas Klebinger at 2024-05-15T07:32:10-04:00 Expand the `inline` rule to look through casts/ticks. Fixes #24808 - - - - - b1e0c313 by Cheng Shao at 2024-05-15T07:32:46-04:00 testsuite: bump PartialDownSweep timeout to 5x on wasm32 - - - - - b2227487 by Fendor at 2024-05-15T17:14:06-04:00 Add Eq and Ord instance to `IfaceType` We add an `Ord` instance so that we can store `IfaceType` in a `Data.Map` container. This is required to deduplicate `IfaceType` while writing `.hi` files to disk. Deduplication has many beneficial consequences to both file size and memory usage, as the deduplication enables implicit sharing of values. See issue #24540 for more motivation. The `Ord` instance would be unnecessary if we used a `TrieMap` instead of `Data.Map` for the deduplication process. While in theory this is clerarly the better option, experiments on the agda code base showed that a `TrieMap` implementation has worse run-time performance characteristics. To the change itself, we mostly derive `Eq` and `Ord`. This requires us to change occurrences of `FastString` with `LexicalFastString`, since `FastString` has no `Ord` instance. We change the definition of `IfLclName` to a newtype of `LexicalFastString`, to make such changes in the future easier. Bump haddock submodule for IfLclName changes - - - - - d368f9a6 by Fendor at 2024-05-15T17:14:06-04:00 Move out LiteralMap to avoid cyclic module dependencies - - - - - 2fcc09fd by Fendor at 2024-05-15T17:14:06-04:00 Add deduplication table for `IfaceType` The type `IfaceType` is a highly redundant, tree-like data structure. While benchmarking, we realised that the high redundancy of `IfaceType` causes high memory consumption in GHCi sessions when byte code is embedded into the `.hi` file via `-fwrite-if-simplified-core` or `-fbyte-code-and-object-code`. Loading such `.hi` files from disk introduces many duplicates of memory expensive values in `IfaceType`, such as `IfaceTyCon`, `IfaceTyConApp`, `IA_Arg` and many more. We improve the memory behaviour of GHCi by adding an additional deduplication table for `IfaceType` to the serialisation of `ModIface`, similar to how we deduplicate `Name`s and `FastString`s. When reading the interface file back, the table allows us to automatically share identical values of `IfaceType`. To provide some numbers, we evaluated this patch on the agda code base. We loaded the full library from the `.hi` files, which contained the embedded core expressions (`-fwrite-if-simplified-core`). Before this patch: * Load time: 11.7 s, 2.5 GB maximum residency. After this patch: * Load time: 7.3 s, 1.7 GB maximum residency. This deduplication has the beneficial side effect to additionally reduce the size of the on-disk interface files tremendously. For example, on agda, we reduce the size of `.hi` files (with `-fwrite-if-simplified-core`): * Before: 101 MB on disk * Now: 24 MB on disk This has even a beneficial side effect on the cabal store. We reduce the size of the store on disk: * Before: 341 MB on disk * Now: 310 MB on disk Note, none of the dependencies have been compiled with `-fwrite-if-simplified-core`, but `IfaceType` occurs in multiple locations in a `ModIface`. We also add IfaceType deduplication table to .hie serialisation and refactor .hie file serialisation to use the same infrastrucutre as `putWithTables`. Bump haddock submodule to accomodate for changes to the deduplication table layout and binary interface. - - - - - 36aa7cf1 by Fendor at 2024-05-15T17:14:06-04:00 Add run-time configurability of `.hi` file compression Introduce the flag `-fwrite-if-compression=<n>` which allows to configure the compression level of writing .hi files. The motivation is that some deduplication operations are too expensive for the average use case. Hence, we introduce multiple compression levels with variable impact on performance, but still reduce the memory residency and `.hi` file size on disk considerably. We introduce three compression levels: * `1`: `Normal` mode. This is the least amount of compression. It deduplicates only `Name` and `FastString`s, and is naturally the fastest compression mode. * `2`: `Safe` mode. It has a noticeable impact on .hi file size and is marginally slower than `Normal` mode. In general, it should be safe to always use `Safe` mode. * `3`: `Full` deduplication mode. Deduplicate as much as we can, resulting in minimal .hi files, but at the cost of additional compilation time. Reading .hi files doesn't need to know the initial compression level, and can always deserialise a `ModIface`, as we write out a byte that indicates the next value has been deduplicated. This allows users to experiment with different compression levels for packages, without recompilation of dependencies. Note, the deduplication also has an additional side effect of reduced memory consumption to implicit sharing of deduplicated elements. See https://gitlab.haskell.org/ghc/ghc/-/issues/24540 for example where that matters. ------------------------- Metric Decrease: MultiLayerModulesDefsGhciWithCore T16875 T21839c T24471 hard_hole_fits libdir ------------------------- - - - - - 1e63a6fb by Matthew Pickering at 2024-05-15T17:14:07-04:00 Introduce regression tests for `.hi` file sizes Add regression tests to track how `-fwrite-if-compression` levels affect the size of `.hi` files. - - - - - 639d742b by M Farkas-Dyck at 2024-05-15T17:14:49-04:00 TTG: ApplicativeStatement exist only in Rn and Tc Co-Authored-By: romes <rodrigo.m.mesquita at gmail.com> - - - - - aa7b336b by Jade at 2024-05-15T23:06:17-04:00 Documentation: Improve documentation for symbols exported from System.IO - - - - - c561de8f by Jade at 2024-05-15T23:06:54-04:00 Improve suggestions for language extensions - When suggesting Language extensions, also suggest Extensions which imply them - Suggest ExplicitForAll and GADTSyntax instead of more specific extensions - Rephrase suggestion to include the term 'Extension' - Also moves some flag specific definitions out of Session.hs into Flags.hs (#24478) Fixes: #24477 Fixes: #24448 Fixes: #10893 - - - - - 4c7ae2a1 by Andreas Klebinger at 2024-05-15T23:07:30-04:00 Testsuite: Check if llvm assembler is available for have_llvm - - - - - bc672166 by Torsten Schmits at 2024-05-15T23:08:06-04:00 refactor quadratic search in warnMissingHomeModules - - - - - 7875e8cb by Torsten Schmits at 2024-05-15T23:08:06-04:00 add test that runs MakeDepend on thousands of modules - - - - - b84b91f5 by Adam Gundry at 2024-05-16T15:32:06-04:00 Representation-polymorphic HasField (fixes #22156) This generalises the HasField class to support representation polymorphism, so that instead of type HasField :: forall {k} . k -> Type -> Type -> Constraint we have type HasField :: forall {k} {r_rep} {a_rep} . k -> TYPE r_rep -> TYPE a_rep -> Constraint - - - - - 05285090 by Matthew Pickering at 2024-05-16T15:32:43-04:00 Bump os-string submodule to 2.0.2.2 Closes #24786 - - - - - 886ab43a by Cheng Shao at 2024-05-17T01:34:50-04:00 rts: do not prefetch mark_closure bdescr in non-moving gc when ASSERTS_ENABLED This commit fixes a small an oversight in !12148: the prefetch logic in non-moving GC may trap in debug RTS because it calls Bdescr() for mark_closure which may be a static one. It's fine in non-debug RTS because even invalid bdescr addresses are prefetched, they will not cause segfaults, so this commit implements the most straightforward fix: don't prefetch mark_closure bdescr when assertions are enabled. - - - - - b38dcf39 by Teo Camarasu at 2024-05-17T01:34:50-04:00 rts: Allocate non-moving segments with megablocks Non-moving segments are 8 blocks long and need to be aligned. Previously we serviced allocations by grabbing 15 blocks, finding an aligned 8 block group in it and returning the rest. This proved to lead to high levels of fragmentation as a de-allocating a segment caused an 8 block gap to form, and this could not be reused for allocation. This patch introduces a segment allocator based around using entire megablocks to service segment allocations in bulk. When there are no free segments, we grab an entire megablock and fill it with aligned segments. As the megablock is free, we can easily guarantee alignment. Any unused segments are placed on a free list. It only makes sense to free segments in bulk when all of the segments in a megablock are freeable. After sweeping, we grab the free list, sort it, and find all groups of segments where they cover the megablock and free them. This introduces a period of time when free segments are not available to the mutator, but the risk that this would lead to excessive allocation is low. Right after sweep, we should have an abundance of partially full segments, and this pruning step is relatively quick. In implementing this we drop the logic that kept NONMOVING_MAX_FREE segments on the free list. We also introduce an eventlog event to log the amount of pruned/retained free segments. See Note [Segment allocation strategy] Resolves #24150 ------------------------- Metric Decrease: T13253 T19695 ------------------------- - - - - - 710665bd by Cheng Shao at 2024-05-17T01:35:30-04:00 rts: fix I/O manager compilation errors for win32 target This patch fixes I/O manager compilation errors for win32 target discovered when cross-compiling to win32 using recent clang: ``` rts/win32/ThrIOManager.c:117:7: error: error: call to undeclared function 'is_io_mng_native_p'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 117 | if (is_io_mng_native_p ()) { | ^ | 117 | if (is_io_mng_native_p ()) { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/fs.c:143:28: error: error: a function declaration without a prototype is deprecated in all versions of C [-Werror,-Wstrict-prototypes] 143 | int setErrNoFromWin32Error () { | ^ | void | 143 | int setErrNoFromWin32Error () { | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:227:9: error: error: call to undeclared function 'interruptIOManagerEvent'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 227 | interruptIOManagerEvent (); | ^ | 227 | interruptIOManagerEvent (); | ^ rts/win32/ConsoleHandler.c:227:9: error: note: did you mean 'getIOManagerEvent'? | 227 | interruptIOManagerEvent (); | ^ rts/include/rts/IOInterface.h:27:10: error: note: 'getIOManagerEvent' declared here 27 | void * getIOManagerEvent (void); | ^ | 27 | void * getIOManagerEvent (void); | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) rts/win32/ConsoleHandler.c:196:9: error: error: call to undeclared function 'setThreadLabel'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration] 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/win32/ConsoleHandler.c:196:9: error: note: did you mean 'postThreadLabel'? | 196 | setThreadLabel(cap, t, "signal handler thread"); | ^ rts/eventlog/EventLog.h:118:6: error: note: 'postThreadLabel' declared here 118 | void postThreadLabel(Capability *cap, | ^ | 118 | void postThreadLabel(Capability *cap, | ^ 1 error generated. `x86_64-w64-mingw32-clang' failed in phase `C Compiler'. (Exit code: 1) ``` - - - - - 28b9cee0 by Rodrigo Mesquita at 2024-05-17T01:36:05-04:00 configure: Check C99-compat for Cmm preprocessor Fixes #24815 - - - - - 8927e0c3 by Andreas Klebinger at 2024-05-17T01:36:41-04:00 Ensure `tcHasFixedRuntimeRep (# #)` returns True. - - - - - 04179044 by doyougnu at 2024-05-17T09:00:32-04:00 testsuite: make find_so regex less general Closes #24759 Background. In MR !12372 we began tracking shared object files and directories sizes for dependencies. However, this broke release builds because release builds alter the filenames swapping "in-place" for a hash. This was not considered in the MR and thus broke release pipelines. Furthermore, the rts_so test was found to be wildly varying and was therefore disabled in !12561. This commit fixes both of these issues: - fix the rts_so test by making the regex less general, now the rts_so test and all other foo.so tests must match "libHS<some-lib>-<version>-<hash|'in-place>-<ghc>". This prevents the rts_so test from accidentally matching different rts variants such as rts_threaded, which was the cause of the wild swings after !12372. - add logic to match either a hash or the string in-place. This should make the find_so function build agnostic. - - - - - 0962b50d by Andreas Klebinger at 2024-05-17T09:01:08-04:00 TagAnalysis: Treat all bottom ids as tagged during analysis. Ticket #24806 showed that we also need to treat dead end thunks as tagged during the analysis. - - - - - 7eb9f184 by Ben Gamari at 2024-05-17T11:23:37-04:00 Remove haddock submodule In preparation for merge into the GHC, as proposed in #23178. - - - - - 47b14dcc by Fendor at 2024-05-17T11:28:17-04:00 Adapt to `IfLclName` newtype changes (cherry picked from commit a711607e29b925f3d69e27c5fde4ba655c711ff1) - - - - - 6cc6681d by Fendor at 2024-05-17T11:28:17-04:00 Add IfaceType deduplication table to interface file serialisation Although we do not really need it in the interface file serialisation, as the deserialisation uses `getWithUserData`, we need to mirror the structure `getWithUserData` expects. Thus, we write essentially an empty `IfaceType` table at the end of the file, as the interface file doesn't reference `IfaceType`. (cherry picked from commit c9bc29c6a708483d2abc3d8ec9262510ce87ca61) - - - - - b9721206 by Ben Gamari at 2024-05-17T11:30:22-04:00 ghc-tags.yaml: Initial commit - - - - - 074e7d8f by Ben Gamari at 2024-05-17T11:31:29-04:00 fourmolu: Add configuration - - - - - 151b1736 by Ben Gamari at 2024-05-17T11:32:52-04:00 Makefile: Rework for use by haddock developers Previously the Makefile was present only for GHC's old make-based build system. Now since the make-based build system is gone we can use it for more useful ends. - - - - - a7dcf13b by Ben Gamari at 2024-05-17T11:36:14-04:00 Reformat with fourmolu Using previously-added configuration and `fourmolu -i .` Note that we exclude the test-cases (`./{hoogle,html-hypsrc,latex}-test`) as they are sensitive to formatting. - - - - - 0ea6017b by Ben Gamari at 2024-05-17T11:40:04-04:00 Add 'utils/haddock/' from commit 'a7dcf13bfbb97b20e75cc8ce650e2bb628db4660' git-subtree-dir: utils/haddock git-subtree-mainline: 7eb9f1849b1c72a1c61dee88462b4244550406f3 git-subtree-split: a7dcf13bfbb97b20e75cc8ce650e2bb628db4660 - - - - - aba1d304 by Hécate Moonlight at 2024-05-17T11:40:48-04:00 Add exceptions to the dangling notes list - - - - - 527bfbfb by Hécate Moonlight at 2024-05-17T11:40:52-04:00 Add haddock to the whitespace lint ignore list - - - - - 43274677 by Ben Gamari at 2024-05-17T11:41:20-04:00 git-blame-ignore-revs: Ignore haddock reformatting - - - - - 0e679e37 by Fendor at 2024-05-18T00:27:24-04:00 Pass cpp options to the CC builder in hadrian - - - - - bb40244e by Sylvain Henry at 2024-05-18T00:28:06-04:00 JS: fix allocation constant (fix #24746) - - - - - 646d30ab by Jade at 2024-05-18T19:23:31+02:00 Add highlighting for inline-code snippets in haddock - - - - - 64459a3e by Hécate Moonlight at 2024-05-19T08:42:27-04:00 haddock: Add a .readthedocs.yml file for online documentation - - - - - 7d3d9bbf by Serge S. Gulin at 2024-05-19T18:47:05+00:00 Unicode: General Category size test (related #24789) Added trivial size performance test which involves unicode general category usage via `read`. The `read` itself uses general category to detect spaces. The purpose for this test is to measure outcome of applying improvements at General Category representation in code discussed at #24789. - - - - - 8e04efcf by Alan Zimmerman at 2024-05-19T21:29:34-04:00 EPA: Remove redundant code Remove unused epAnnAnns function various cases for showAstData that no longer exist - - - - - 071d7a1e by Rodrigo Mesquita at 2024-05-20T10:55:16-04:00 Improve docs on closed type families in hs-boots Fixes #24776 - - - - - d9e2c119 by Torsten Schmits at 2024-05-20T10:55:52-04:00 Use default deviation for large-project test This new performance test has the purpose of detecting regressions in complexity in relation to the number of modules in a project, so 1% deviation is way too small to avoid false positives. - - - - - 20b0136a by Ben Gamari at 2024-05-22T00:31:39-04:00 ghcup-metadata: Various fixes from 9.10.1 Use Debian 12/x86-64, Debian 10/aarch64, and Debian 11/aarch64 bindists where possible. - - - - - 6838a7c3 by Sylvain Henry at 2024-05-22T00:32:23-04:00 Reverse arguments to stgCallocBytes (fix #24828) - - - - - f50f46c3 by Fendor at 2024-05-22T00:32:59-04:00 Add log messages for Iface serialisation compression level Fix the label of the number of 'IfaceType' entries in the log message. Add log message for the compression level that is used to serialise a an interface file. Adds `Outputable` instance for 'CompressionIFace'. - - - - - 3bad5d55 by Hécate Moonlight at 2024-05-22T00:33:40-04:00 base: Update doctests outputs ghc-internal: Update doctests outputs - - - - - 9317c6fb by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix the testsuites of the haddock-library - Apply all the metadata revisions from Hackage to the cabal file. - Fix the `ParserSpec.hs` file in the `spec` testsuite of haddock-library. - Make `CHANGES.md` an extra-doc-file instead of an extra-source-file. - - - - - 54073b02 by David Binder at 2024-05-22T00:34:21-04:00 haddock: Fix parser of @since pragma The testsuite contained tests for annotations of the form `@since foo-bar-0.5.0`, but the parser was written incorrectly. - - - - - ede6ede3 by Matthew Pickering at 2024-05-22T00:34:57-04:00 Fix nightly pages job It seems likely broken by 9f99126a which moved `index.html` from the root folder into `docs/` folder. Fixes #24840 - - - - - b7bcf729 by Cheng Shao at 2024-05-22T00:35:32-04:00 autoconf: remove unused context diff check This patch removes redundant autoconf check for the context diff program given it isn't actually been used anywhere, especially since make removal. - - - - - ea2fe66e by Hécate Moonlight at 2024-05-22T00:36:13-04:00 haddock: Rework the contributing guide - - - - - 0f302a94 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 haddock: Add module relationships diagrams of haddock-api and haddock-library - - - - - d1a9f34f by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add instructions - - - - - b880ee80 by Hécate Moonlight at 2024-05-22T00:36:52-04:00 Add SVG outputs - - - - - 6d7e6ad8 by Ben Gamari at 2024-05-22T13:40:05-04:00 rts: Fix size of StgOrigThunkInfo frames Previously the entry code of the `stg_orig_thunk` frame failed to account for the size of the profiling header as it hard-coded the frame size. Fix this. Fixes #24809. - - - - - c645fe40 by Fendor at 2024-05-22T13:40:05-04:00 Add regression test T24809 for stg_orig_thunk_info_frame size - - - - - 4181aa40 by Andreas Klebinger at 2024-05-22T13:40:42-04:00 bindists: Check for existence of share folder before trying to copy it. This folder isn't distributed in windows bindists A lack of doing so resulted us copying loads of files twice. - - - - - d216510e by Matthew Pickering at 2024-05-22T13:40:42-04:00 Remove ad-hoc installation of mingw toolchain in relocatable bindists This reverts 616ac30026e8dd7d2ebb98d92dde071eedf5d951 The choice about whether to install mingw is taken in the installation makefile. This is also broken on non-windows systems. The actual issue was the EnableDistroToolchain variable wasn't declared in mk/config.mk and therefore the check to install mingw was failing. - - - - - 7b4c1998 by Cheng Shao at 2024-05-22T21:52:52-04:00 testsuite: fix T17920 for wasm backend T17920 was marked as fragile on wasm before; it can be trivially fixed by avoiding calling variadic printf() in cmm. - - - - - c739383b by Cheng Shao at 2024-05-22T21:53:29-04:00 testsuite: bump T22744 timeout to 5x - - - - - c4c6d714 by Cheng Shao at 2024-05-22T21:54:06-04:00 testsuite: don't attempt to detect host cpu features when testing cross ghc The testsuite driver CPU feature detection logic only detects host CPU and only makes sense when we are not testing a cross GHC. - - - - - 3d9e4ce6 by Simon Peyton Jones at 2024-05-22T21:54:43-04:00 Better skolemisation As #24810 showed, it is (a little) better to skolemise en-bloc, so that Note [Let-bound skolems] fires more often. See Note [Skolemisation en bloc] in GHC.Tc.Utils.Instantiate. - - - - - a3cd3a1d by Ryan Scott at 2024-05-22T21:55:19-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. - - - - - 4bb2a7cc by Hécate Moonlight at 2024-05-22T21:55:59-04:00 [base] Document the memory overhead of ByteArray Add a diagram that shows the constituent parts of a ByteArray and their memory overhead. - - - - - 8b2a016a by Hécate Moonlight at 2024-05-22T21:56:38-04:00 Haddock: Add MR template for Haddock - - - - - ead75532 by Peter Trommler at 2024-05-23T02:28:05-04:00 PPC: Support ELF v2 on powerpc64 big-endian Detect ELF v2 on PowerPC 64-bit systems. Check for `_CALL_ELF` preprocessor macro. Fixes #21191 - - - - - 9d4c10f2 by Hécate Kleidukos at 2024-05-23T02:28:44-04:00 gitlab: Add @Kleidukos to CODEOWNERS for utils/haddock - - - - - 28e64170 by Preetham Gujjula at 2024-05-23T07:20:48-04:00 haddock: Add cabal-fmt to tools for `make style` - - - - - 00126a89 by Andrei Borzenkov at 2024-05-23T07:21:24-04:00 haddock: fix verbosity option parsing - - - - - a3e0b68b by Ryan Hendrickson at 2024-05-23T15:52:03-04:00 base: specify tie-breaking behavior of min, max, and related list/Foldable functions - - - - - bdcc0f37 by doyougnu at 2024-05-24T07:51:18-04:00 cmm: add word <-> double/float bitcast - closes: #25331 This is the last step in the project plan described in #25331. This commit: - adds bitcast operands for x86_64, LLVM, aarch64 - For PPC and i386 we resort to using the cmm implementations - renames conversion MachOps from Conv to Round|Truncate - - - - - f0d257f7 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor Some functions in StgToByteCode were filtering out void arguments. However, StgToByteCode is called after unarisation: the void arguments should have been removed earlier. Instead of filtering out, we assert that the args are non-void. - - - - - 03137fd2 by Krzysztof Gogolewski at 2024-05-24T07:51:55-04:00 StgToByteCode: minor refactor `layoutNativeCall` was always called with a `primRepCmmType platform` callback. Hence we can put it inside of `layoutNativeCall` rather than repeat it. - - - - - 27c430f3 by David Binder at 2024-05-24T07:52:38-04:00 haddock: Remove compatibility shims for GHC < 8.4 from haddock-library - - - - - 8dd8a076 by Cheng Shao at 2024-05-24T07:53:14-04:00 compiler: avoid saving foreign call target to local when there are no caller-save GlobalRegs This patch makes the STG->Cmm backend avoid saving foreign call target to local when there are no caller-save GlobalRegs. Since 321941a8ebe25192cdeece723e1058f2f47809ea, when we lower a foreign call, we unconditionally save the foreign call target to a temporary local first, then rely on cmmSink to clean it up later, which only happens with -fcmm-sink (implied by -O) and not in unoptimized code. And this is troublesome for the wasm backend NCG, which needs to infer a foreign call target symbol's type signature from the Cmm call site. Previously, the NCG has been emitting incorrect type signatures for unoptimized code, which happens to work with `wasm-ld` most of the time, but this is never future-proof against upstream toolchain updates, and it causes horrible breakages when LTO objects are included in linker input. Hence this patch. - - - - - 986df1ab by Cheng Shao at 2024-05-24T07:53:14-04:00 testsuite: add callee-no-local regression test - - - - - 52d62e2a by Sylvain Henry at 2024-05-24T07:53:57-04:00 Fix HasCallStack leftovers from !12514 / #24726 - - - - - c5e00c35 by crumbtoo at 2024-05-24T07:54:38-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 - - - - - bd323b0e by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Ensure that CHANGELOG is included in extra-source-files This was missed in the `ghc-internal` split. Closes #24831. - - - - - 1bfd32e8 by Ben Gamari at 2024-05-24T07:55:15-04:00 base: Fix changelog reference to setBacktraceMechanismState (cherry picked from commit b63f7ba01fdfd98a01d2f0dec8d9262b3e595c5d) - - - - - 43e8e4f3 by Sylvain Henry at 2024-05-24T12:16:43-04:00 Float/double unboxed literal support for HexFloatLiterals (fix #22155) - - - - - 4a7f4713 by Fendor at 2024-05-24T12:17:19-04:00 Improve test labels for binary interface file size tests Test labels for binary interface file sizes are hard to read and overly verbose at the same time. Extend the name for the metric title, but shorten it in the actual comparison table. - - - - - 14e554cf by Zubin Duggal at 2024-05-24T12:17:55-04:00 Revert "Fix haskell/haddock#783 Don't show button if --quickjump not present" This reverts commit 7776566531e72c415f66dd3b13da9041c52076aa. - - - - - f56838c3 by Ben Gamari at 2024-05-24T12:17:55-04:00 Fix default hyperlinked sources pattern Previously this didn't include the `%M` token which manifested as broken links to the hyperlinked sources of reexports of declarations defined in other packages. Fixes haddock#1628. (cherry picked from commit 1432bcc943d41736eca491ecec4eb9a6304dab36) - - - - - 42efa62c by Ben Gamari at 2024-05-24T12:17:55-04:00 Make DocPaths a proper data type (cherry picked from commit 7f3a5c4da0023ae47b4c376c9b1ea2d706c94d8c) - - - - - 53d9ceb3 by Ben Gamari at 2024-05-24T12:17:55-04:00 haddock: Bump version to 2.30 (cherry picked from commit 994989ed3d535177e57b778629726aeabe8c7602) - - - - - e4db1112 by Zubin Duggal at 2024-05-24T12:17:55-04:00 haddock-api: allow base 4.20 and ghc 9.11 - - - - - e294f7a2 by PHO at 2024-05-24T12:17:55-04:00 Add a flag "threaded" for building haddock with the threaded RTS GHC isn't guaranteed to have a threaded RTS. There should be a way to build it with the vanilla one. (cherry picked from commit 75a94e010fb5b0236c670d22b04f5472397dc15d) - - - - - 51165bc9 by Andreas Klebinger at 2024-05-25T10:58:03-04:00 Update ticky counter event docs. Add the info about the info table address and json fields. Fixes #23200 - - - - - 98597ad5 by Sylvain Henry at 2024-05-25T10:58:45-04:00 Export extractPromotedList (#24866) This can be useful in plugins. - - - - - 228dcae6 by Teo Camarasu at 2024-05-28T13:12:24+00:00 template-haskell: Move wired-ins to ghc-internal Thus we make `template-haskell` reinstallable and keep it as the public API for Template Haskell. All of the wired-in identifiers are moved to `ghc-internal`. This necessitates also moving much of `ghc-boot-th` into `ghc-internal`. These modules are then re-exported from `ghc-boot-th` and `template-haskell`. To avoid a dependency on `template-haskell` from `lib:ghc`, we instead depend on the TH ASTs via `ghc-boot-th`. As `template-haskell` no longer has special status, we can drop the logic adding an implicit dependency on `template-haskell` when using TH. We can also drop the `template-haskell-next` package, which was previously used when bootstrapping. When bootstrapping, we need to vendor the TH AST modules from `ghc-internal` into `ghc-boot-th`. This is controlled by the `bootstrap` cabal flag as before. See Note [Bootstrapping Template Haskell]. We split out a GHC.Internal.TH.Lift module resolving #24752. This module is only built when not bootstrapping. Resolves #24703 ------------------------- Metric Increase: ghc_boot_th_dir ghc_boot_th_so ------------------------- - - - - - 62dded28 by Teo Camarasu at 2024-05-28T13:12:24+00:00 testsuite: mark tests broken by #24886 Now that `template-haskell` is no longer wired-in. These tests are triggering #24886, and so need to be marked broken. - - - - - 3ca72ad9 by Cheng Shao at 2024-05-30T02:57:06-04:00 rts: fix missing function prototypes in ClosureMacros.h - - - - - e0029e3d by Andreas Klebinger at 2024-05-30T02:57:43-04:00 UnliftedFFITypes: Allow `(# #)` as argument when it's the only argument. This allows representing functions like: int foo(void); to be imported like this: foreign import ccall "a_number_c" c_number :: (# #) -> Int64# Which can be useful when the imported function isn't implicitly stateful. - - - - - d0401335 by Matthew Pickering at 2024-05-30T02:58:19-04:00 ci: Update ci-images commit for fedora38 image The fedora38 nightly job has been failing for quite a while because `diff` was no longer installed. The ci-images bump explicitly installs `diffutils` into these images so hopefully they now pass again. - - - - - 3c97c74a by Jan Hrček at 2024-05-30T02:58:58-04:00 Update exactprint docs - - - - - 77760cd7 by Jan Hrček at 2024-05-30T02:58:58-04:00 Incorporate review feedback - - - - - 87591368 by Jan Hrček at 2024-05-30T02:58:58-04:00 Remove no longer relevant reference to comments - - - - - 05f4f142 by Jan Hrček at 2024-05-30T02:58:59-04:00 Replace outdated code example - - - - - 45a4a5f3 by Andreas Klebinger at 2024-05-30T02:59:34-04:00 Reword error resulting from missing -XBangPatterns. It can be the result of either a bang pattern or strict binding, so now we say so instead of claiming it must be a bang pattern. Fixes #21032 - - - - - e17f2df9 by Cheng Shao at 2024-05-30T03:00:10-04:00 testsuite: bump MultiLayerModulesDefsGhciReload timeout to 10x - - - - - 7a660042 by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - c77a48af by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 05c4fafb by Cheng Shao at 2024-05-30T14:42:29-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 5ff83bfc by Sylvain Henry at 2024-05-30T14:43:10-04:00 JS: remove useless h$CLOCK_REALTIME (#23202) - - - - - 95ef2d58 by Matthew Pickering at 2024-05-30T14:43:47-04:00 ghcup-metadata: Fix metadata generation There were some syntax errors in the generation script which were preventing it from running. I have tested this with: ``` nix shell --extra-experimental-features nix-command -f .gitlab/rel_eng -c ghcup-metadata --metadata ghcup-0.0.7.yaml --date="2024-05-27" --pipeline-id=95534 --version=9.11.20240525 ``` which completed successfully. - - - - - 1bc66ee4 by Jakob Bruenker at 2024-05-30T14:44:22-04:00 Add diagrams to Arrows documentation This adds diagrams to the documentation of Arrows, similar to the ones found on https://www.haskell.org/arrows/. It does not add diagrams for ArrowChoice for the time being, mainly because it's not clear to me how to visually distinguish them from the ones for Arrow. Ideally, you might want to do something like highlight the arrows belonging to the same tuple or same Either in common colors, but that's not really possible with unicode. - - - - - d10a1c65 by Matthew Craven at 2024-05-30T23:35:48-04:00 Make UnsafeSNat et al. into pattern synonyms ...so that they do not cause coerce to bypass the nominal role on the corresponding singleton types when they are imported. See Note [Preventing unsafe coercions for singleton types] and the discussion at #23478. This also introduces unsafeWithSNatCo (and analogues for Char and Symbol) so that users can still access the dangerous coercions that importing the real constructors would allow, but only in a very localized way. - - - - - 0958937e by Cheng Shao at 2024-05-30T23:36:25-04:00 hadrian: build C/C++ with split sections when enabled When split sections is enabled, ensure -fsplit-sections is passed to GHC as well when invoking GHC to compile C/C++; and pass -ffunction-sections -fdata-sections to gcc/clang when compiling C/C++ with the hadrian Cc builder. Fixes #23381. - - - - - 02b1f91e by Cheng Shao at 2024-05-30T23:36:25-04:00 driver: build C/C++ with -ffunction-sections -fdata-sections when split sections is enabled When -fsplit-sections is passed to GHC, pass -ffunction-sections -fdata-sections to gcc/clang when building C/C++. Previously, -fsplit-sections was only respected by the NCG/LLVM backends, but not the unregisterised backend; the GHC driver did not pass -fdata-sections and -ffunction-sections to the C compiler, which resulted in excessive executable sizes. Fixes #23381. ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - fd47e2e3 by Cheng Shao at 2024-05-30T23:37:00-04:00 testsuite: mark process005 as fragile on JS - - - - - 34a04ea1 by Matthew Pickering at 2024-05-31T06:08:36-04:00 Add -Wderiving-typeable to -Wall Deriving `Typeable` does nothing, and it hasn't done for a long while. There has also been a warning for a long while which warns you about uselessly deriving it but it wasn't enabled in -Wall. Fixes #24784 - - - - - 75fa7b0b by Matthew Pickering at 2024-05-31T06:08:36-04:00 docs: Fix formatting of changelog entries - - - - - 303c4b33 by Preetham Gujjula at 2024-05-31T06:09:21-04:00 docs: Fix link to injective type families paper Closes #24863 - - - - - df97e9a6 by Ben Gamari at 2024-05-31T06:09:57-04:00 ghc-internal: Fix package description The previous description was inherited from `base` and was inappropriate for `ghc-internal`. Also fix the maintainer and bug reporting fields. Closes #24906. - - - - - bf0737c0 by Cheng Shao at 2024-05-31T06:10:33-04:00 compiler: remove ArchWasm32 special case in cmmDoCmmSwitchPlans This patch removes special consideration for ArchWasm32 in cmmDoCmmSwitchPlans, which means the compiler will now disable cmmImplementSwitchPlans for wasm unreg backend, just like unreg backend of other targets. We enabled it in the past to workaround some compile-time panic in older versions of LLVM, but those panics are no longer present, hence no need to keep this workaround. - - - - - 7eda4bd2 by Cheng Shao at 2024-05-31T15:52:04-04:00 utils: add hie.yaml config file for ghc-config Add hie.yaml to ghc-config project directory so it can be edited using HLS. - - - - - 1e5752f6 by Cheng Shao at 2024-05-31T15:52:05-04:00 hadrian: handle findExecutable "" gracefully hadrian may invoke findExecutable "" at run-time due to a certain program is not found by configure script. Which is fine and findExecutable is supposed to return Nothing in this case. However, on Windows there's a directory bug that throws an exception (see https://github.com/haskell/directory/issues/180), so we might as well use a wrapper for findExecutable and handle exceptions gracefully. - - - - - 4eb5ad09 by Cheng Shao at 2024-05-31T15:52:05-04:00 configure: do not set LLC/OPT/LLVMAS fallback values when FIND_LLVM_PROG fails When configure fails to find LLC/OPT/LLVMAS within supported version range, it used to set "llc"/"opt"/"clang" as fallback values. This behavior is particularly troublesome when the user has llc/opt/clang with other versions in their PATH and run the testsuite, since hadrian will incorrectly assume have_llvm=True and pass that to the testsuite driver, resulting in annoying optllvm test failures (#23186). If configure determines llc/opt/clang wouldn't work, then we shouldn't pretend it'll work at all, and the bindist configure will invoke FIND_LLVM_PROG check again at install time anyway. - - - - - 5f1afdf7 by Sylvain Henry at 2024-05-31T15:52:52-04:00 Introduce UniqueSet and use it to replace 'UniqSet Unique' 'UniqSet Unique' represents a set of uniques as a 'Map Unique Unique', which is wasting space (associated key/value are always the same). Fix #23572 and #23605 - - - - - e0aa42b9 by crumbtoo at 2024-05-31T15:53:33-04:00 Improve template-haskell haddocks Closes #15822 - - - - - ae170155 by Olivier Benz at 2024-06-01T09:35:17-04:00 Bump max LLVM version to 19 (not inclusive) - - - - - 92aa65ea by Matthew Pickering at 2024-06-01T09:35:17-04:00 ci: Update CI images to test LLVM 18 The debian12 image in this commit has llvm 18 installed. - - - - - adb1fe42 by Serge S. Gulin at 2024-06-01T09:35:53-04:00 Unicode: make ucd2haskell build-able again ucd2haskell tool used streamly library which version in cabal was out of date. It is updated to the latest version at hackage with deprecated parts rewritten. Also following fixes were applied to existing code in suppose that from its last run the code structure was changed and now it was required to be up to date with actual folder structures: 1. Ghc module path environment got a suffix with `src`. 2. Generated code got 2.1 `GHC.Internal` prefix for `Data.*`. 2.2 `GHC.Unicode.Internal` swapped on `GHC.Internal.Unicode` according to actual structure. - - - - - ad56fd84 by Jade at 2024-06-01T09:36:29-04:00 Replace 'NB' with 'Note' in error messages - - - - - 6346c669 by Cheng Shao at 2024-06-01T09:37:04-04:00 compiler: fix -ddump-cmm-raw when compiling .cmm This patch fixes missing -ddump-cmm-raw output when compiling .cmm, which is useful for debugging cmm related codegen issues. - - - - - 1c834ad4 by Ryan Scott at 2024-06-01T09:37:40-04:00 Print namespace specifiers in FixitySig's Outputable instance For whatever reason, the `Outputable` instance for `FixitySig` simply did not print out namespace specifiers, leading to the confusing `-ddump-splices` output seen in #24911. This patch corrects this oversight. Fixes #24911. - - - - - cf49fb5f by Sylvain Henry at 2024-06-01T09:38:19-04:00 Configure: display C++ compiler path - - - - - f9c1ae12 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable PIC for in-tree GMP on wasm32 This patch disables PIC for in-tree GMP on wasm32 target. Enabling PIC unconditionally adds undesired code size and runtime overhead for wasm32. - - - - - 1a32f828 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: disable in-tree gmp fft code path for wasm32 This patch disables in-tree GMP FFT code paths for wasm32 target in order to give up some performance of multiplying very large operands in exchange for reduced code size. - - - - - 06277d56 by Cheng Shao at 2024-06-02T14:01:55-04:00 hadrian: build in-tree GMP with malloc-notreentrant on wasm32 This patch makes hadrian build in-tree GMP with the --enable-alloca=malloc-notreentrant configure option. We will only need malloc-reentrant when we have threaded RTS and SMP support on wasm32, which will take some time to happen, before which we should use malloc-notreentrant to avoid undesired runtime overhead. - - - - - 9f614270 by ARATA Mizuki at 2024-06-02T14:02:35-04:00 Set package include paths when assembling .S files Fixes #24839. Co-authored-by: Sylvain Henry <hsyl20 at gmail.com> - - - - - 4998a6ed by Alex Mason at 2024-06-03T02:09:29-04:00 Improve performance of genericWordQuotRem2Op (#22966) Implements the algorithm from compiler-rt's udiv128by64to64default. This rewrite results in a roughly 24x improvement in runtime on AArch64 (and likely any other arch that uses it). - - - - - ae50a8eb by Cheng Shao at 2024-06-03T02:10:05-04:00 testsuite: mark T7773 as fragile on wasm - - - - - c8ece0df by Fendor at 2024-06-03T19:43:22-04:00 Migrate `Finder` component to `OsPath`, fixed #24616 For each module in a GHCi session, we keep alive one `ModLocation`. A `ModLocation` is fairly inefficiently packed, as `String`s are expensive in memory usage. While benchmarking the agda codebase, we concluded that we keep alive around 11MB of `FilePath`'s, solely retained by `ModLocation`. We provide a more densely packed encoding of `ModLocation`, by moving from `FilePath` to `OsPath`. Further, we migrate the full `Finder` component to `OsPath` to avoid unnecessary transformations. As the `Finder` component is well-encapsulated, this requires only a minimal amount of changes in other modules. We introduce pattern synonym for 'ModLocation' which maintains backwards compatibility and avoids breaking consumers of 'ModLocation'. - - - - - 0cff083a by Cheng Shao at 2024-06-03T19:43:58-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. - - - - - 18f63970 by Sebastian Graf at 2024-06-04T05:05:27-04:00 Parser: Remove unused `apats` rule - - - - - 38757c30 by David Knothe at 2024-06-04T05:05:27-04:00 Implement Or Patterns (#22596) This commit introduces a new language extension, `-XOrPatterns`, as described in GHC Proposal 522. An or-pattern `pat1; ...; patk` succeeds iff one of the patterns `pat1`, ..., `patk` succeed, in this order. See also the summary `Note [Implmentation of OrPatterns]`. Co-Authored-By: Sebastian Graf <sgraf1337 at gmail.com> - - - - - 395412e8 by Cheng Shao at 2024-06-04T05:06:04-04:00 compiler/ghci/rts: remove stdcall support completely We have formally dropped i386 windows support (#18487) a long time ago. The stdcall foreign call convention is only used by i386 windows, and the legacy logic around it is a significant maintenance burden for future work that adds arm64 windows support (#24603). Therefore, this patch removes stdcall support completely from the compiler as well as the RTS (#24883): - stdcall is still recognized as a FFI calling convention in Haskell syntax. GHC will now unconditionally emit a warning (-Wunsupported-calling-conventions) and treat it as ccall. - Apart from minimum logic to support the parsing and warning logic, all other code paths related to stdcall has been completely stripped from the compiler. - ghci only supports FFI_DEFAULT_ABI and ccall convention from now on. - FFI foreign export adjustor code on all platforms no longer handles the stdcall case and only handles ccall from now on. - The Win32 specific parts of RTS no longer has special code paths for stdcall. This commit is the final nail on the coffin for i386 windows support. Further commits will perform more housecleaning to strip the legacy code paths and pave way for future arm64 windows support. - - - - - d1fe9ab6 by Cheng Shao at 2024-06-04T05:06:04-04:00 rts: remove legacy i386 windows code paths This commit removes some legacy i386 windows related code paths in the RTS, given this target is no longer supported. - - - - - a605e4b2 by Cheng Shao at 2024-06-04T05:06:04-04:00 autoconf: remove i386 windows related logic This commit removes legacy i386 windows logic in autoconf scripts. - - - - - 91e5ac5e by Cheng Shao at 2024-06-04T05:06:04-04:00 llvm-targets: remove i386 windows support This commit removes i386 windows from llvm-targets and the script to generate it. - - - - - 65fe75a4 by Cheng Shao at 2024-06-04T05:06:04-04:00 libraries/utils: remove stdcall related legacy logic This commit removes stdcall related legacy logic in libraries and utils. ccall should be used uniformly for all supported windows hosts from now on. - - - - - d2a83302 by Cheng Shao at 2024-06-04T05:06:04-04:00 testsuite: adapt the testsuite for stdcall removal This patch adjusts test cases to handle the stdcall removal: - Some stdcall usages are replaced with ccall since stdcall doesn't make sense anymore. - We also preserve some stdcall usages, and check in the expected warning messages to ensure GHC always warn about stdcall usages (-Wunsupported-calling-conventions) as expected. - Error code testsuite coverage is slightly improved, -Wunsupported-calling-conventions is now tested. - Obsolete code paths related to i386 windows are also removed. - - - - - cef8f47a by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: minor adjustments for stdcall removal This commit include minor adjustments of documentation related to stdcall removal. - - - - - 54332437 by Cheng Shao at 2024-06-04T05:06:04-04:00 docs: mention i386 Windows removal in 9.12 changelog This commit mentions removal of i386 Windows support and stdcall related change in the 9.12 changelog. - - - - - 2aaea8a1 by Cheng Shao at 2024-06-04T05:06:40-04:00 hadrian: improve user settings documentation This patch adds minor improvements to hadrian user settings documentation: - Add missing `ghc.cpp.opts` case - Remove non-existent `cxx` case - Clarify `cc.c.opts` also works for C++, while `cc.deps.opts` doesn't - Add example of passing configure argument to autoconf packages - - - - - 71010381 by Alex Mason at 2024-06-04T12:09:07-04:00 Add AArch64 CLZ, CTZ, RBIT primop implementations. Adds support for emitting the clz and rbit instructions, which are used by GHC.Prim.clz*#, GHC.Prim.ctz*# and GHC.Prim.bitReverse*#. - - - - - 44e2abfb by Cheng Shao at 2024-06-04T12:09:43-04:00 hadrian: add +text_simdutf flavour transformer to allow building text with simdutf This patch adds a +text_simdutf flavour transformer to hadrian to allow downstream packagers and users that build from source to opt-in simdutf support for text, in order to benefit from SIMD speedup at run-time. It's still disabled by default for the time being. - - - - - 077cb2e1 by Cheng Shao at 2024-06-04T12:09:43-04:00 ci: enable +text_simdutf flavour transformer for wasm jobs This commit enables +text_simdutf flavour transformer for wasm jobs, so text is now built with simdutf support for wasm. - - - - - b23746ad by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Use TemplateHaskellQuotes in instance Lift ByteArray Resolves #24852 - - - - - 3fd25743 by Teo Camarasu at 2024-06-04T22:50:50-04:00 base: Mark addrToByteArray as NOINLINE This function should never be inlined in order to keep code size small. - - - - - 98ad1ea5 by Cheng Shao at 2024-06-04T22:51:26-04:00 compiler: remove unused CompilerInfo/LinkerInfo types This patch removes CompilerInfo/LinkerInfo types from the compiler since they aren't actually used anywhere. - - - - - 11795244 by Cheng Shao at 2024-06-05T06:33:17-04:00 rts: remove unused PowerPC/IA64 native adjustor code This commit removes unused PowerPC/IA64 native adjustor code which is never actually enabled by autoconf/hadrian. Fixes #24920. - - - - - 5132754b by Sylvain Henry at 2024-06-05T06:33:57-04:00 RTS: fix warnings with doing*Profiling (#24918) - - - - - accc8c33 by Cheng Shao at 2024-06-05T11:35:36-04:00 hadrian: don't depend on inplace/mingw when --enable-distro-toolchain on Windows - - - - - 6ffbd678 by Cheng Shao at 2024-06-05T11:35:37-04:00 autoconf: normalize paths of some build-time dependencies on Windows This commit applies path normalization via cygpath -m to some build-time dependencies on Windows. Without this logic, the /clang64/bin prefixed msys2-style paths cause the build to fail with --enable-distro-toolchain. - - - - - 075dc6d4 by Cheng Shao at 2024-06-05T11:36:12-04:00 hadrian: remove OSDarwin mention from speedHack This commit removes mentioning of OSDarwin from speedHack, since speedHack is purely for i386 and we no longer support i386 darwin (#24921). - - - - - 83235c4c by Cheng Shao at 2024-06-05T11:36:12-04:00 compiler: remove 32-bit darwin logic This commit removes all 32-bit darwin logic from the compiler, given we no longer support 32-bit apple systems (#24921). Also contains a bit more cleanup of obsolete i386 windows logic. - - - - - 1eb99bc3 by Cheng Shao at 2024-06-05T11:36:12-04:00 rts: remove 32-bit darwin/ios logic This commit removes 32-bit darwin/ios related logic from the rts, given we no longer support them (#24921). - - - - - 24f65892 by Cheng Shao at 2024-06-05T11:36:12-04:00 llvm-targets: remove 32-bit darwin/ios targets This commit removes 32-bit darwin/ios targets from llvm-targets given we no longer support them (#24921). - - - - - ccdbd689 by Cheng Shao at 2024-06-05T11:36:12-04:00 testsuite: remove 32-bit darwin logic This commit removes 32-bit darwin logic from the testsuite given it's no longer supported (#24921). Also contains more cleanup of obsolete i386 windows logic. - - - - - 11d661c4 by Cheng Shao at 2024-06-05T11:36:13-04:00 docs: mention 32-bit darwin/ios removal in 9.12 changelog This commit mentions removal of 32-bit darwin/ios support (#24921) in the 9.12 changelog. - - - - - 7c173310 by Georgi Lyubenov at 2024-06-05T15:17:22-04:00 Add firstA and secondA to Data.Bitraversable Please see https://github.com/haskell/core-libraries-committee/issues/172 for related discussion - - - - - 3b6f9fd1 by Ben Gamari at 2024-06-05T15:17:59-04:00 base: Fix name of changelog Fixes #24899. Also place it under `extra-doc-files` to better reflect its nature and avoid triggering unnecessary recompilation if it changes. - - - - - 1f4d2ef7 by Sebastian Graf at 2024-06-05T15:18:34-04:00 Announce Or-patterns in the release notes for GHC 9.12 (#22596) Leftover from !9229. - - - - - 8650338d by Jan Hrček at 2024-06-06T10:39:24-04:00 Improve haddocks of Language.Haskell.Syntax.Pat.Pat - - - - - 2eee65e1 by Cheng Shao at 2024-06-06T10:40:00-04:00 testsuite: bump T7653 timeout for wasm - - - - - 990fed60 by Sylvain Henry at 2024-06-07T14:45:23-04:00 StgToCmm: refactor opTranslate and friends - Change arguments order to avoid `\args -> ...` lambdas - Fix documentation - Rename StgToCmm options ("big" doesn't mean anything) - - - - - 1afad514 by Sylvain Henry at 2024-06-07T14:45:23-04:00 NCG x86: remove dead code (#5444) Since 6755d833af8c21bbad6585144b10e20ac4a0a1ab this code is dead. - - - - - 595c0894 by Cheng Shao at 2024-06-07T14:45:58-04:00 testsuite: skip objc-hi/objcxx-hi when cross compiling objc-hi/objcxx-hi should be skipped when cross compiling. The existing opsys('darwin') predicate only asserts the host system is darwin but tells us nothing about the target, hence the oversight. - - - - - edfe6140 by qqwy at 2024-06-08T11:23:54-04:00 Replace '?callStack' implicit param with HasCallStack in GHC.Internal.Exception.throw - - - - - 35a64220 by Cheng Shao at 2024-06-08T11:24:30-04:00 rts: cleanup inlining logic This patch removes pre-C11 legacy code paths related to INLINE_HEADER/STATIC_INLINE/EXTERN_INLINE macros, ensure EXTERN_INLINE is treated as static inline in most cases (fixes #24945), and also corrects the comments accordingly. - - - - - 9ea90ed2 by Andrew Lelechenko at 2024-06-08T11:25:06-04:00 CODEOWNERS: add @core-libraries to track base interface changes A low-tech tactical solution for #24919 - - - - - 580fef7b by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update CHANGELOG to reflect current version - - - - - 391ecff5 by Ben Gamari at 2024-06-09T01:27:21-04:00 ghc-internal: Update prologue.txt to reflect package description - - - - - 3dca3b7d by Ben Gamari at 2024-06-09T01:27:57-04:00 compiler: Clarify comment regarding need for MOVABS The comment wasn't clear in stating that it was only applicable to immediate source and memory target operands. - - - - - 6bd850e8 by doyougnu at 2024-06-09T21:02:14-04:00 JS: establish single source of truth for symbols In pursuit of: #22736. This MR moves ad-hoc symbols used throughout the js backend into a single symbols file. Why? First, this cleans up the code by removing ad-hoc strings created on the fly and therefore makes the code more maintainable. Second, it makes it much easier to eventually type these identifiers. - - - - - f3017dd3 by Cheng Shao at 2024-06-09T21:02:49-04:00 rts: replace ad-hoc MYTASK_USE_TLV with proper CC_SUPPORTS_TLS This patch replaces the ad-hoc `MYTASK_USE_TLV` with the `CC_SUPPORTS_TLS` macro. If TLS support is detected by autoconf, then we should use that for managing `myTask` in the threaded RTS. - - - - - e17d7e8c by Ben Gamari at 2024-06-11T05:25:21-04:00 users-guide: Fix stylistic issues in 9.12 release notes - - - - - 8a8a982a by Hugo Peters at 2024-06-11T05:25:57-04:00 fix typo in the simplifier debug output: baling -> bailing - - - - - 16475bb8 by Hécate Moonlight at 2024-06-12T03:07:55-04:00 haddock: Correct the Makefile to take into account Darwin systems - - - - - a2f60da5 by Hécate Kleidukos at 2024-06-12T03:08:35-04:00 haddock: Remove obsolete links to github.com/haskell/haddock in the docs - - - - - de4395cd by qqwy at 2024-06-12T03:09:12-04:00 Add `__GLASGOW_HASKELL_ASSERTS_IGNORED__` as CPP macro name if `-fasserts-ignored is set. This allows users to create their own Control.Exception.assert-like functionality that does something other than raising an `AssertFailed` exception. Fixes #24967 - - - - - 0e9c4dee by Ryan Hendrickson at 2024-06-12T03:09:53-04:00 compiler: add hint to TcRnBadlyStaged message - - - - - 2747cd34 by Simon Peyton Jones at 2024-06-12T12:51:37-04:00 Fix a QuickLook bug This MR fixes the bug exposed by #24676. The problem was that quickLookArg was trying to avoid calling tcInstFun unnecessarily; but it was in fact necessary. But that in turn forced me into a significant refactoring, putting more fields into EValArgQL. Highlights: see Note [Quick Look overview] in GHC.Tc.Gen.App * Instantiation variables are now distinguishable from ordinary unification variables, by level number = QLInstVar. This is treated like "level infinity". See Note [The QLInstVar TcLevel] in GHC.Tc.Utils.TcType. * In `tcApp`, we don't track the instantiation variables in a set Delta any more; instead, we just tell them apart by their level number. * EValArgQL now much more clearly captures the "half-done" state of typechecking an argument, ready for later resumption. See Note [Quick Look at value arguments] in GHC.Tc.Gen.App * Elminated a bogus (never used) fast-path in GHC.Tc.Utils.Instantiate.instCallConstraints See Note [Possible fast path for equality constraints] Many other small refactorings. - - - - - 1b1523b1 by George Thomas at 2024-06-12T12:52:18-04:00 Fix non-compiling extensible record `HasField` example - - - - - 97b141a3 by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Fix hyperlinker source urls (#24907) This fixes a bug introduced by f56838c36235febb224107fa62334ebfe9941aba Links to external modules in the hyperlinker are uniformly generated using splicing the template given to us instead of attempting to construct the url in an ad-hoc manner. - - - - - 954f864c by Zubin Duggal at 2024-06-12T12:52:55-04:00 haddock: Add name anchor to external source urls from documentation page URLs for external source links from documentation pages were missing a splice location for the name. Fixes #24912 - - - - - b0b64177 by Simon Peyton Jones at 2024-06-12T12:53:31-04:00 Prioritise nominal equalities The main payload of this patch is * Prioritise nominal equalities in the constraint solver. This ameliorates the incompleteness of solving for representational constraints over newtypes: see #24887. See (EX2) in Note [Decomposing newtype equalities] in GHC.Tc.Solver.Equality In doing this patch I tripped over some other things that I refactored: * Move `isCoVarType` from `GHC.Core.Type` to `GHC.Core.Predicate` where it seems more at home. * Clarify the "rewrite role" of a constraint. I was very puzzled about what the role of, say `(Eq a)` might be, but see the new Note [The rewrite-role of a constraint]. In doing so I made predTypeEqRel crash when given a non-equality. Usually it expects an equality; but it was being mis-used for the above rewrite-role stuff. - - - - - cb7c1b83 by Liam Goodacre at 2024-06-12T12:54:09-04:00 compiler: missing-deriving-strategies suggested fix Extends the missing-deriving-strategies warning with a suggested fix that includes which deriving strategies were assumed. For info about the warning, see comments for `TcRnNoDerivStratSpecified`, `TcRnNoDerivingClauseStrategySpecified`, & `TcRnNoStandaloneDerivingStrategySpecified`. For info about the suggested fix, see `SuggestExplicitDerivingClauseStrategies` & `SuggestExplicitStandalanoDerivingStrategy`. docs: Rewords missing-deriving-strategies to mention the suggested fix. Resolves #24955 - - - - - 4e36d3a3 by Jan Hrček at 2024-06-12T12:54:48-04:00 Further haddocks improvements in Language.Haskell.Syntax.Pat.Pat - - - - - 558353f4 by Cheng Shao at 2024-06-12T12:55:24-04:00 rts: use page sized mblocks on wasm This patch changes mblock size to page size on wasm. It allows us to simplify our wasi-libc fork, makes it much easier to test third party libc allocators like emmalloc/mimalloc, as well as experimenting with threaded RTS in wasm. - - - - - b3cc5366 by Matthew Pickering at 2024-06-12T23:06:57-04:00 compiler: Make ghc-experimental not wired in If you need to wire in definitions, then place them in ghc-internal and reexport them from ghc-experimental. Ticket #24903 - - - - - 700eeab9 by Hécate Kleidukos at 2024-06-12T23:07:37-04:00 base: Use a more appropriate unicode arrow for the ByteArray diagram This commit rectifies the usage of a unicode arrow in favour of one that doesn't provoke mis-alignment. - - - - - cca7de25 by Matthew Pickering at 2024-06-12T23:08:14-04:00 ghcup-metadata: Fix debian version ranges This was caught by `ghcup-ci` failing and attempting to install a deb12 bindist on deb11. ``` configure: WARNING: m4/prep_target_file.m4: Expecting YES/NO but got in ArSupportsDashL_STAGE0. Defaulting to False. bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by bin/ghc-toolchain-bin) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) bin/ghc-toolchain-bin: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /tmp/tmp.LBhwvFbVoy/foobarbaz/.ghcup/tmp/ghcup-708d9668d5d82287/ghc-9.11.20240609-x86_64-unknown-linux/bin/../lib/x86_64-linux-ghc-9.11.20240609/libHSunix-2.8.5.1-inplace-ghc9.11.20240609.so) ``` Fixes #24974 - - - - - 7b23ce8b by Pierre Le Marre at 2024-06-13T15:35:04-04:00 ucd2haskell: remove Streamly dependency + misc - Remove dead code. - Remove `streamly` dependency. - Process files with `bytestring`. - Replace Unicode files parsers with the corresponding ones from the package `unicode-data-parser`. - Simplify cabal file and rename module - Regenerate `ghc-internal` Unicode files with new header - - - - - 4570319f by Jacco Krijnen at 2024-06-13T15:35:41-04:00 Document how to run haddocks tests (#24976) Also remove ghc 9.7 requirement - - - - - fb629e24 by amesgen at 2024-06-14T00:28:20-04:00 compiler: refactor lower_CmmExpr_Ptr - - - - - def46c8c by amesgen at 2024-06-14T00:28:20-04:00 compiler: handle CmmRegOff in lower_CmmExpr_Ptr - - - - - ce76bf78 by Simon Peyton Jones at 2024-06-14T00:28:56-04:00 Small documentation update in Quick Look - - - - - 19bcfc9b by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Add hack for #24623 ..Th bug in #24623 is randomly triggered by this MR!.. - - - - - 7a08a025 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Various fixes to type-tidying This MR was triggered by #24868, but I found a number of bugs and infelicities in type-tidying as I went along. Highlights: * Fix to #24868 is in GHC.Tc.Errors.report_unsolved: avoid using the OccNames of /bound/ variables when tidying /free/ variables; see the call to `tidyAvoiding`. That avoid the gratuitous renaming which was the cause of #24868. See Note [tidyAvoiding] in GHC.Core.TyCo.Tidy * Refactor and document the tidying of open types. See GHC.Core.TyCo.Tidy Note [Tidying open types] Note [Tidying is idempotent] * Tidy the coercion variable in HoleCo. That's important so that tidied types have tidied kinds. * Some small renaming to make things consistent. In particular the "X" forms return a new TidyEnv. E.g. tidyOpenType :: TidyEnv -> Type -> Type tidyOpenTypeX :: TidyEnv -> Type -> (TidyEnv, Type) - - - - - 2eac0288 by Simon Peyton Jones at 2024-06-14T14:44:19-04:00 Wibble - - - - - e5d24cc2 by Simon Peyton Jones at 2024-06-14T14:44:20-04:00 Wibbles - - - - - 246bc3a4 by Simon Peyton Jones at 2024-06-14T14:44:56-04:00 Localise a case-binder in SpecConstr.mkSeqs This small change fixes #24944 See (SCF1) in Note [SpecConstr and strict fields] - - - - - a5994380 by Sylvain Henry at 2024-06-15T03:20:29-04:00 PPC: display foreign label in panic message (cf #23969) - - - - - bd95553a by Rodrigo Mesquita at 2024-06-15T03:21:06-04:00 cmm: Parse MO_BSwap primitive operation Parsing this operation allows it to be tested using `test-primops` in a subsequent MR. - - - - - e0099721 by Andrew Lelechenko at 2024-06-16T17:57:38-04:00 Make flip representation polymorphic, similar to ($) and (&) CLC proposal: https://github.com/haskell/core-libraries-committee/issues/245 - - - - - 118a1292 by Alan Zimmerman at 2024-06-16T17:58:15-04:00 EPA: Add location to Match Pats list So we can freely modify the pats and the following item spacing will still be valid when exact printing. Closes #24862 - - - - - db343324 by Fabricio de Sousa Nascimento at 2024-06-17T10:01:51-04:00 compiler: Rejects RULES whose LHS immediately fails to type-check Fixes GHC crashing on `decomposeRuleLhs` due to ignoring coercion values. This happens when we have a RULE that does not type check, and enable `-fdefer-type-errors`. We prevent this to happen by rejecting RULES with an immediately LHS type error. Fixes #24026 - - - - - e7a95662 by Dylan Thinnes at 2024-06-17T10:02:35-04:00 Add hscTypecheckRenameWithDiagnostics, for HLS (#24996) Use runHsc' in runHsc so that both functions can't fall out of sync We're currently copying parts of GHC code to get structured warnings in HLS, so that we can recreate `hscTypecheckRenameWithDiagnostics` locally. Once we get this function into GHC we can drop the copied code in future versions of HLS. - - - - - d70abb49 by sheaf at 2024-06-18T18:47:20-04:00 Clarify -XGADTs enables existential quantification Even though -XGADTs does not turn on -XExistentialQuantification, it does allow the user of existential quantification syntax, without needing to use GADT-style syntax. Fixes #20865 - - - - - 13fdf788 by David Binder at 2024-06-18T18:48:02-04:00 Add RTS flag --read-tix-file (GHC Proposal 612) This commit introduces the RTS flag `--read-tix-file=<yes|no>` which controls whether a preexisting .tix file is read in at the beginning of a program run. The default is currently `--read-tix-file=yes` but will change to `--read-tix-file=no` in a future release of GHC. For this reason, whenever a .tix file is read in a warning is emitted to stderr. This warning can be silenced by explicitly passing the `--read-tix-file=yes` option. Details can be found in the GHC proposal cited below. Users can query whether this flag has been used with the help of the module `GHC.RTS.Flags`. A new field `readTixFile` was added to the record `HpcFlags`. These changes have been discussed and approved in - GHC proposal 612: https://github.com/ghc-proposals/ghc-proposals/pull/612 - CLC proposal 276: https://github.com/haskell/core-libraries-committee/issues/276 - - - - - f0e3cb6a by Fendor at 2024-06-18T18:48:38-04:00 Improve sharing of duplicated values in `ModIface`, fixes #24723 As a `ModIface` often contains duplicated values that are not necessarily shared, we improve sharing by serialising the `ModIface` to an in-memory byte array. Serialisation uses deduplication tables, and deserialisation implicitly shares duplicated values. This helps reducing the peak memory usage while compiling in `--make` mode. The peak memory usage is especially smaller when generating interface files with core expressions (`-fwrite-if-simplified-core`). On agda, this reduces the peak memory usage: * `2.2 GB` to `1.9 GB` for a ghci session. On `lib:Cabal`, we report: * `570 MB` to `500 MB` for a ghci session * `790 MB` to `667 MB` for compiling `lib:Cabal` with ghc There is a small impact on execution time, around 2% on the agda code base. - - - - - 1bab7dde by Fendor at 2024-06-18T18:48:38-04:00 Avoid unneccessarily re-serialising the `ModIface` To reduce memory usage of `ModIface`, we serialise `ModIface` to an in-memory byte array, which implicitly shares duplicated values. This serialised byte array can be reused to avoid work when we actually write the `ModIface` to disk. We introduce a new field to `ModIface` which allows us to save the byte array, and write it direclty to disk if the `ModIface` wasn't changed after the initial serialisation. This requires us to change absolute offsets, for example to jump to the deduplication table for `Name` or `FastString` with relative offsets, as the deduplication byte array doesn't contain header information, such as fingerprints. To allow us to dump the binary blob to disk, we need to replace all absolute offsets with relative ones. We introduce additional helpers for `ModIface` binary serialisation, which construct relocatable binary blobs. We say the binary blob is relocatable, if the binary representation can be moved and does not contain any absolute offsets. Further, we introduce new primitives for `Binary` that allow to create relocatable binaries, such as `forwardGetRel` and `forwardPutRel`. ------------------------- Metric Decrease: MultiLayerModulesDefsGhcWithCore Metric Increase: MultiComponentModules MultiLayerModules T10421 T12150 T12234 T12425 T13035 T13253-spj T13701 T13719 T14697 T15703 T16875 T18698b T18140 T18304 T18698a T18730 T18923 T20049 T24582 T5837 T6048 T9198 T9961 mhu-perf ------------------------- These metric increases may look bad, but they are all completely benign, we simply allocate 1 MB per module for `shareIface`. As this allocation is quite quick, it has a negligible impact on run-time performance. In fact, the performance difference wasn't measurable on my local machine. Reducing the size of the pre-allocated 1 MB buffer avoids these test failures, but also requires us to reallocate the buffer if the interface file is too big. These reallocations *did* have an impact on performance, which is why I have opted to accept all these metric increases, as the number of allocated bytes is merely a guidance. This 1MB allocation increase causes a lot of tests to fail that generally have a low allocation number. E.g., increasing from 40MB to 41MB is a 2.5% increase. In particular, the tests T12150, T13253-spj, T18140, T18304, T18698a, T18923, T20049, T24582, T5837, T6048, and T9961 only fail on i386-darwin job, where the number of allocated bytes seems to be lower than in other jobs. The tests T16875 and T18698b fail on i386-linux for the same reason. - - - - - 099992df by Andreas Klebinger at 2024-06-18T18:49:14-04:00 Improve documentation of @Any@ type. In particular mention possible uses for non-lifted types. Fixes #23100. - - - - - 5e75412b by Jakob Bruenker at 2024-06-18T18:49:51-04:00 Update user guide to indicate support for 64-tuples - - - - - 4f5da595 by Andreas Klebinger at 2024-06-18T18:50:28-04:00 lint notes: Add more info to notes.stdout When fixing a note reference CI fails with a somewhat confusing diff. See #21123. This commit adds a line to the output file being compared which hopefully makes it clear this is the list of broken refs, not all refs. Fixes #21123 - - - - - 1eb15c61 by Jakob Bruenker at 2024-06-18T18:51:04-04:00 docs: Update mention of ($) type in user guide Fixes #24909 - - - - - 1d66c9e3 by Jan Hrček at 2024-06-18T18:51:47-04:00 Remove duplicate Anno instances - - - - - 8ea0ba95 by Sven Tennie at 2024-06-18T18:52:23-04:00 AArch64: Delete unused RegNos This has the additional benefit of getting rid of the -1 encoding (real registers start at 0.) - - - - - 325422e0 by Sjoerd Visscher at 2024-06-18T18:53:04-04:00 Bump stm submodule to current master - - - - - 64fba310 by Cheng Shao at 2024-06-18T18:53:40-04:00 testsuite: bump T17572 timeout on wasm32 - - - - - eb612fbc by Sven Tennie at 2024-06-19T06:46:00-04:00 AArch64: Simplify BL instruction The BL constructor carried unused data in its third argument. - - - - - b0300503 by Alan Zimmerman at 2024-06-19T06:46:36-04:00 TTG: Move SourceText from `Fixity` to `FixitySig` It is only used there, simplifies the use of `Fixity` in the rest of the code, and is moved into a TTG extension point. Precedes !12842, to simplify it - - - - - 842e119b by Rodrigo Mesquita at 2024-06-19T06:47:13-04:00 base: Deprecate some .Internal modules Deprecates the following modules according to clc-proposal #217: https://github.com/haskell/core-libraries-committee/issues/217 * GHC.TypeNats.Internal * GHC.TypeLits.Internal * GHC.ExecutionStack.Internal Closes #24998 - - - - - 24e89c40 by Jacco Krijnen at 2024-06-20T07:21:27-04:00 ttg: Use List instead of Bag in AST for LHsBindsLR Considering that the parser used to create a Bag of binds using a cons-based approach, it can be also done using lists. The operations in the compiler don't really require Bag. By using lists, there is no dependency on GHC.Data.Bag anymore from the AST. Progress towards #21592 - - - - - 04f5bb85 by Simon Peyton Jones at 2024-06-20T07:22:03-04:00 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. See Note [Tracking Given equalities] and Note [Let-bound skolems] both in GHC.Tc.Solver.InertSet. Then * Test LocalGivenEqs succeeds for a different reason than before; see (LBS2) in Note [Let-bound skolems] * New test T24938a succeeds because of (LBS2), whereas it failed before. * Test LocalGivenEqs2 now fails, as it should. * Test T224938, the repro from the ticket, fails, as it should. - - - - - 9a757a27 by Simon Peyton Jones at 2024-06-20T07:22:40-04:00 Fix demand signatures for join points This MR tackles #24623 and #23113 The main change is to give a clearer notion of "worker/wrapper arity", esp for join points. See GHC.Core.Opt.DmdAnal Note [Worker/wrapper arity and join points] This Note is a good summary of what this MR does: (1) The "worker/wrapper arity" of an Id is * For non-join-points: idArity * The join points: the join arity (Id part only of course) This is the number of args we will use in worker/wrapper. See `ww_arity` in `dmdAnalRhsSig`, and the function `workWrapArity`. (2) A join point's demand-signature arity may exceed the Id's worker/wrapper arity. See the `arity_ok` assertion in `mkWwBodies`. (3) In `finaliseArgBoxities`, do trimBoxity on any argument demands beyond the worker/wrapper arity. (4) In WorkWrap.splitFun, make sure we split based on the worker/wrapper arity (re)-computed by workWrapArity. - - - - - 5e8faaf1 by Jan Hrček at 2024-06-20T07:23:20-04:00 Update haddocks of Import/Export AST types - - - - - cd512234 by Hécate Kleidukos at 2024-06-20T07:24:02-04:00 haddock: Update bounds in cabal files and remove allow-newer stanza in cabal.project - - - - - 8a8ff8f2 by Rodrigo Mesquita at 2024-06-20T07:24:38-04:00 cmm: Don't parse MO_BSwap for W8 Don't support parsing bswap8, since bswap8 is not really an operation and would have to be implemented as a no-op (and currently is not implemented at all). Fixes #25002 - - - - - 5cc472f5 by sheaf at 2024-06-20T07:25:14-04:00 Delete unused testsuite files These files were committed by mistake in !11902. This commit simply removes them. - - - - - 7b079378 by Matthew Pickering at 2024-06-20T07:25:50-04:00 Remove left over debugging pragma from 2016 This pragma was accidentally introduced in 648fd73a7b8fbb7955edc83330e2910428e76147 The top-level cost centres lead to a lack of optimisation when compiling with profiling. - - - - - c872e09b by Hécate Kleidukos at 2024-06-20T19:28:36-04:00 haddock: Remove unused pragmata, qualify usages of Data.List functions, add more sanity checking flags by default This commit enables some extensions and GHC flags in the cabal file in a way that allows us to reduce the amount of prologuing on top of each file. We also prefix the usage of some List functions that removes ambiguity when they are also exported from the Prelude, like foldl'. In general, this has the effect of pointing out more explicitly that a linked list is used. Metric Increase: haddock.Cabal haddock.base haddock.compiler - - - - - 8c87d4e1 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 Add test case for #23586 - - - - - 568de8a5 by Arnaud Spiwack at 2024-06-20T19:29:12-04:00 When matching functions in rewrite rules: ignore multiplicity When matching a template variable to an expression, we check that it has the same type as the matched expression. But if the variable `f` has type `A -> B` while the expression `e` has type `A %1 -> B`, the match was previously rejected. A principled solution would have `f` substituted by `\(%Many x) -> e x` or some other appropriate coercion. But since linearity is not properly checked in Core, we can be cheeky and simply ignore multiplicity while matching. Much easier. This has forced a change in the linter which, when `-dlinear-core-lint` is off, must consider that `a -> b` and `a %1 -> b` are equal. This is achieved by adding an argument to configure the behaviour of `nonDetCmpTypeX` and modify `ensureEqTys` to call to the new behaviour which ignores multiplicities when comparing two `FunTy`. Fixes #24725. - - - - - c8a8727e by Simon Peyton Jones at 2024-06-20T19:29:12-04:00 Faster type equality This MR speeds up type equality, triggered by perf regressions that showed up when fixing #24725 by parameterising type equality over whether to ignore multiplicity. The changes are: * Do not use `nonDetCmpType` for type /equality/. Instead use a specialised type-equality function, which we have always had! `nonDetCmpType` remains, but I did not invest effort in refactoring or optimising it. * Type equality is parameterised by - whether to expand synonyms - whether to respect multiplicities - whether it has a RnEnv2 environment In this MR I systematically specialise it for static values of these parameters. Much more direct and predictable than before. See Note [Specialising type equality] * We want to avoid comparing kinds if possible. I refactored how this happens, at least for `eqType`. See Note [Casts and coercions in type comparison] * To make Lint fast, we want to avoid allocating a thunk for <msg> in ensureEqTypes ty1 ty2 <msg> because the test almost always succeeds, and <msg> isn't needed. See Note [INLINE ensureEqTys] Metric Decrease: T13386 T5030 - - - - - 21fc180b by Ryan Hendrickson at 2024-06-22T10:40:55-04:00 base: Add inits1 and tails1 to Data.List - - - - - d640a3b6 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Derive previously hand-written `Lift` instances (#14030) This is possible now that #22229 is fixed. - - - - - 33fee6a2 by Sebastian Graf at 2024-06-22T10:41:32-04:00 Implement the "Derive Lift instances for data types in template-haskell" proposal (#14030) After #22229 had been fixed, we can finally derive the `Lift` instance for the TH AST, as proposed by Ryan Scott in https://mail.haskell.org/pipermail/libraries/2015-September/026117.html. Fixes #14030, #14296, #21759 and #24560. The residency of T24471 increases by 13% because we now load `AnnLookup` from its interface file, which transitively loads the whole TH AST. Unavoidable and not terrible, I think. Metric Increase: T24471 - - - - - 383c01a8 by Matthew Pickering at 2024-06-22T10:42:08-04:00 bindist: Use complete relative paths when cding to directories If a user has configured CDPATH on their system then `cd lib` may change into an unexpected directory during the installation process. If you write `cd ./lib` then it will not consult `CDPATH` to determine what you mean. I have added a check on ghcup-ci to verify that the bindist installation works in this situation. Fixes #24951 - - - - - 5759133f by Hécate Kleidukos at 2024-06-22T10:42:49-04:00 haddock: Use the more precise SDocContext instead of DynFlags The pervasive usage of DynFlags (the parsed command-line options passed to ghc) blurs the border between different components of Haddock, and especially those that focus solely on printing text on the screen. In order to improve the understanding of the real dependencies of a function, the pretty-printer options are made concrete earlier in the pipeline instead of late when pretty-printing happens. This also has the advantage of clarifying which functions actually require DynFlags for purposes other than pretty-printing, thus making the interactions between Haddock and GHC more understandable when exploring the code base. See Henry, Ericson, Young. "Modularizing GHC". https://hsyl20.fr/home/files/papers/2022-ghc-modularity.pdf. 2022 - - - - - 749e089b by Alexander McKenna at 2024-06-22T10:43:24-04:00 Add INLINE [1] pragma to compareInt / compareWord To allow rules to be written on the concrete implementation of `compare` for `Int` and `Word`, we need to have an `INLINE [1]` pragma on these functions, following the `matching_overloaded_methods_in_rules` note in `GHC.Classes`. CLC proposal https://github.com/haskell/core-libraries-committee/issues/179 Fixes https://gitlab.haskell.org/ghc/ghc/-/issues/22643 - - - - - db033639 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ci: Enable strict ghc-toolchain setting for bindists - - - - - 14308a8f by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Improve parse failure error Improves the error message for when `ghc-toolchain` fails to read a valid `Target` value from a file (in doFormat mode). - - - - - 6e7cfff1 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: ghc-toolchain related options in configure - - - - - 958d6931 by Matthew Pickering at 2024-06-24T17:21:15-04:00 ci: Fail when bindist configure fails when installing bindist It is better to fail earlier if the configure step fails rather than carrying on for a more obscure error message. - - - - - f48d157d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 ghc-toolchain: Fix error logging indentation - - - - - f1397104 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 bindist: Correct default.target substitution The substitution on `default.target.in` must be done after `PREP_TARGET_FILE` is called -- that macro is responsible for setting the variables that will be effectively substituted in the target file. Otherwise, the target file is invalid. Fixes #24792 #24574 - - - - - 665e653e by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 configure: Prefer tool name over tool path It is non-obvious whether the toolchain configuration should use full-paths to tools or simply their names. In addressing #24574, we've decided to prefer executable names over paths, ultimately, because the bindist configure script already does this, thus is the default in ghcs out there. Updates the in-tree configure script to prefer tool names (`AC_CHECK_TOOL` rather than `AC_PATH_TOOL`) and `ghc-toolchain` to ignore the full-path-result of `findExecutable`, which it previously used over the program name. This change doesn't undo the fix in bd92182cd56140ffb2f68ec01492e5aa6333a8fc because `AC_CHECK_TOOL` still takes into account the target triples, unlike `AC_CHECK_PROG/AC_PATH_PROG`. - - - - - 463716c2 by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 dist: Don't forget to configure JavascriptCPP We introduced a configuration step for the javascript preprocessor, but only did so for the in-tree configure script. This commit makes it so that we also configure the javascript preprocessor in the configure shipped in the compiler bindist. - - - - - e99cd73d by Rodrigo Mesquita at 2024-06-24T17:21:15-04:00 distrib: LlvmTarget in distrib/configure LlvmTarget was being set and substituted in the in-tree configure, but not in the configure shipped in the bindist. We want to set the LlvmTarget to the canonical LLVM name of the platform that GHC is targetting. Currently, that is going to be the boostrapped llvm target (hence the code which sets LlvmTarget=bootstrap_llvm_target). - - - - - 4199aafe by Matthew Pickering at 2024-06-24T17:21:51-04:00 Update bootstrap plans for recent GHC versions (9.6.5, 9.8.2, 9.10.10) - - - - - f599d816 by Matthew Pickering at 2024-06-24T17:21:51-04:00 ci: Add 9_10 bootstrap testing job - - - - - 8f4b799d by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Move the usage of mkParserOpts directly to ppHyperlinkedModuleSource in order to avoid passing a whole DynFlags Follow up to !12931 - - - - - 210cf1cd by Hécate Kleidukos at 2024-06-24T17:22:30-04:00 haddock: Remove cabal file linting rule This will be reintroduced with a properly ignored commit when the cabal files are themselves formatted for good. - - - - - 7fe85b13 by Peter Trommler at 2024-06-24T22:03:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 - - - - - 949a0e0b by Andrew Lelechenko at 2024-06-24T22:04:17-04:00 base: fix missing changelog entries - - - - - 1bfa9111 by Andreas Klebinger at 2024-06-26T21:49:53-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. - - - - - caf44a2d by Andrew Lelechenko at 2024-06-26T21:50:30-04:00 Implement Data.List.compareLength and Data.List.NonEmpty.compareLength `compareLength xs n` is a safer and faster alternative to `compare (length xs) n`. The latter would force and traverse the entire spine (potentially diverging), while the former traverses as few elements as possible. The implementation is carefully designed to maintain as much laziness as possible. As per https://github.com/haskell/core-libraries-committee/issues/257 - - - - - f4606ae0 by Serge S. Gulin at 2024-06-26T21:51:05-04:00 Unicode: adding compact version of GeneralCategory (resolves #24789) The following features are applied: 1. Lookup code like Cmm-switches (draft implementation proposed by Sylvain Henry @hsyl20) 2. Nested ifs (logarithmic search vs linear search) (the idea proposed by Sylvain Henry @hsyl20) ------------------------- Metric Decrease: size_hello_artifact size_hello_unicode ------------------------- - - - - - 0e424304 by Hécate Kleidukos at 2024-06-26T21:51:44-04:00 haddock: Restructure import statements This commit removes idiosyncrasies that have accumulated with the years in how import statements were laid out, and defines clear but simple guidelines in the CONTRIBUTING.md file. - - - - - 9b8ddaaf by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Rename test for #24725 I must have fumbled my tabs when I copy/pasted the issue number in 8c87d4e1136ae6d28e92b8af31d78ed66224ee16. - - - - - b0944623 by Arnaud Spiwack at 2024-06-26T21:52:23-04:00 Add original reproducer for #24725 - - - - - 77ce65a5 by Matthew Pickering at 2024-06-27T07:57:14-04:00 Expand LLVM version matching regex for compability with bsd systems sed on BSD systems (such as darwin) does not support the + operation. Therefore we take the simple minded approach of manually expanding group+ to groupgroup*. Fixes #24999 - - - - - bdfe4a9e by Matthew Pickering at 2024-06-27T07:57:14-04:00 ci: On darwin configure LLVMAS linker to match LLC and OPT toolchain The version check was previously broken so the toolchain was not detected at all. - - - - - 07e03a69 by Matthew Pickering at 2024-06-27T07:57:15-04:00 Update nixpkgs commit for darwin toolchain One dependency (c-ares) changed where it hosted the releases which breaks the build with the old nixpkgs commit. - - - - - 144afed7 by Rodrigo Mesquita at 2024-06-27T07:57:50-04:00 base: Add changelog entry for #24998 - - - - - eebe1658 by Sylvain Henry at 2024-06-28T07:13:26-04:00 X86/DWARF: support no tables-next-to-code and asm-shortcutting (#22792) - Without TNTC (tables-next-to-code), we must be careful to not duplicate labels in pprNatCmmDecl. Especially, as a CmmProc is identified by the label of its entry block (and not of its info table), we can't reuse the same label to delimit the block end and the proc end. - We generate debug infos from Cmm blocks. However, when asm-shortcutting is enabled, some blocks are dropped at the asm codegen stage and some labels in the DebugBlocks become missing. We fix this by filtering the generated debug-info after the asm codegen to only keep valid infos. Also add some related documentation. - - - - - 6e86d82b by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: handle JMP to ForeignLabels (#23969) - - - - - 9e4b4b0a by Sylvain Henry at 2024-06-28T07:14:06-04:00 PPC NCG: support loading 64-bit value on 32-bit arch (#23969) - - - - - 50caef3e by Sylvain Henry at 2024-06-28T07:14:46-04:00 Fix warnings in genapply - - - - - 37139b17 by Matthew Pickering at 2024-06-28T07:15:21-04:00 libraries: Update os-string to 2.0.4 This updates the os-string submodule to 2.0.4 which removes the usage of `TemplateHaskell` pragma. - - - - - 0f3d3bd6 by Sylvain Henry at 2024-06-30T00:47:40-04:00 Bump array submodule - - - - - 354c350c by Sylvain Henry at 2024-06-30T00:47:40-04:00 GHCi: Don't use deprecated sizeofMutableByteArray# - - - - - 35d65098 by Ben Gamari at 2024-06-30T00:47:40-04:00 primops: Undeprecate addr2Int# and int2Addr# addr2Int# and int2Addr# were marked as deprecated with the introduction of the OCaml code generator (1dfaee318171836b32f6b33a14231c69adfdef2f) due to its use of tagged integers. However, this backend has long vanished and `base` has all along been using `addr2Int#` in the Show instance for Ptr. While it's unlikely that we will have another backend which has tagged integers, we may indeed support platforms which have tagged pointers. Consequently we undeprecate the operations but warn the user that the operations may not be portable. - - - - - 3157d817 by Sylvain Henry at 2024-06-30T00:47:41-04:00 primops: Undeprecate par# par# is still used in base and it's not clear how to replace it with spark# (see #24825) - - - - - c8d5b959 by Ben Gamari at 2024-06-30T00:47:41-04:00 Primops: Make documentation generation more efficient Previously we would do a linear search through all primop names, doing a String comparison on the name of each when preparing the HsDocStringMap. Fix this. - - - - - 65165fe4 by Ben Gamari at 2024-06-30T00:47:41-04:00 primops: Ensure that deprecations are properly tracked We previously failed to insert DEPRECATION pragmas into GHC.Prim's ModIface, meaning that they would appear in the Haddock documentation but not issue warnings. Fix this. See #19629. Haddock also needs to be fixed: https://github.com/haskell/haddock/issues/223 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - bc1d435e by Mario Blažević at 2024-06-30T00:48:20-04:00 Improved pretty-printing of unboxed TH sums and tuples, fixes #24997 - - - - - 0d170eaf by Zubin Duggal at 2024-07-04T11:08:41-04:00 compiler: Turn `FinderCache` into a record of operations so that GHC API clients can have full control over how its state is managed by overriding `hsc_FC`. Also removes the `uncacheModule` function as this wasn't being used by anything since 1893ba12fe1fa2ade35a62c336594afcd569736e Fixes #23604 - - - - - 4664997d by Teo Camarasu at 2024-07-04T11:09:18-04:00 Add HasCallStack to T23221 This makes the test a bit easier to debug - - - - - 66919dcc by Teo Camarasu at 2024-07-04T11:09:18-04:00 rts: use live words to estimate heap size We use live words rather than live blocks to determine the size of the heap for determining memory retention. Most of the time these two metrics align, but they can come apart in normal usage when using the nonmoving collector. The nonmoving collector leads to a lot of partially occupied blocks. So, using live words is more accurate. They can also come apart when the heap is suffering from high levels fragmentation caused by small pinned objects, but in this case, the block size is the more accurate metric. Since this case is best avoided anyway. It is ok to accept the trade-off that we might try (and probably) fail to return more memory in this case. See also the Note [Statistics for retaining memory] Resolves #23397 - - - - - 8dfca66a by Oleg Grenrus at 2024-07-04T11:09:55-04:00 Add reflections of GHC.TypeLits/Nats type families ------------------------- Metric Increase: ghc_experimental_dir ghc_experimental_so ------------------------- - - - - - 6c469bd2 by Adam Gundry at 2024-07-04T11:10:33-04:00 Correct -Wpartial-fields warning to say "Definition" rather than "Use" Fixes #24710. The message and documentation for `-Wpartial-fields` were misleading as (a) the warning occurs at definition sites rather than use sites, and (b) the warning relates to the definition of a field independently of the selector function (e.g. because record updates are also partial). - - - - - 977b6b64 by Max Ulidtko at 2024-07-04T11:11:11-04:00 GHCi: Support local Prelude Fixes #10920, an issue where GHCi bails out when started alongside a file named Prelude.hs or Prelude.lhs (even empty file suffices). The in-source Note [GHCi and local Preludes] documents core reasoning. Supplementary changes: * add debug traces for module lookups under -ddump-if-trace; * drop stale comment in GHC.Iface.Load; * reduce noise in -v3 traces from GHC.Utils.TmpFs; * new test, which also exercizes HomeModError. - - - - - 87cf4111 by Ryan Scott at 2024-07-04T11:11:47-04:00 Add missing gParPat in cvtp's ViewP case When converting a `ViewP` using `cvtp`, we need to ensure that the view pattern is parenthesized so that the resulting code will parse correctly when roundtripped back through GHC's parser. Fixes #24894. - - - - - b05613c5 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation for module cycle errors (see #18516) This removes the re-export of cyclicModuleErr from the top-level GHC module. - - - - - 70389749 by Adam Gundry at 2024-07-04T11:12:23-04:00 Use structured error representation when reloading a nonexistent module - - - - - 680ade3d by sheaf at 2024-07-04T11:12:23-04:00 Use structured errors for a Backpack instantiation error - - - - - 97c6d6de by sheaf at 2024-07-04T11:12:23-04:00 Move mkFileSrcSpan to GHC.Unit.Module.Location - - - - - f9e7bd9b by Adriaan Leijnse at 2024-07-04T11:12:59-04:00 ttg: Remove SourceText from OverloadedLabel Progress towards #21592 - - - - - 00d63245 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: GHC.Prelude -> Prelude Refactor occurrences to GHC.Prelude with Prelude within Language/Haskell. Progress towards #21592 - - - - - cc846ea5 by Alexander Foremny at 2024-07-04T11:12:59-04:00 AST: remove occurrences of GHC.Unit.Module.ModuleName `GHC.Unit.Module` re-exports `ModuleName` from `Language.Haskell.Syntax.Module.Name`. Progress towards #21592 - - - - - 24c7d287 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move Data instance definition for ModuleName to GHC.Unit.Types To remove the dependency on GHC.Utils.Misc inside Language.Haskell.Syntax.Module.Name, the instance definition is moved from there into GHC.Unit.Types. Progress towards #21592 - - - - - 6cbba381 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move negateOverLitVal into GHC.Hs.Lit The function negateOverLitVal is not used within Language.Haskell and therefore can be moved to the respective module inside GHC.Hs. Progress towards #21592 - - - - - 611aa7c6 by Fabian Kirchner at 2024-07-04T11:12:59-04:00 AST: move conDetailsArity into GHC.Rename.Module The function conDetailsArity is only used inside GHC.Rename.Module. We therefore move it there from Language.Haskell.Syntax.Lit. Progress towards #21592 - - - - - 1b968d16 by Mauricio at 2024-07-04T11:12:59-04:00 AST: Remove GHC.Utils.Assert from GHC Simple cleanup. Progress towards #21592 - - - - - 3d192e5d by Fabian Kirchner at 2024-07-04T11:12:59-04:00 ttg: extract Specificity, ForAllTyFlag and helper functions from GHC.Types.Var Progress towards #21592 Specificity, ForAllTyFlag and its' helper functions are extracted from GHC.Types.Var and moved into a new module Language.Haskell.Syntax.Specificity. Note: Eventually (i.e. after Language.Haskell.Syntax.Decls does not depend on GHC.* anymore) these should be moved into Language.Haskell.Syntax.Decls. At this point, this would cause cyclic dependencies. - - - - - 257d1adc by Adowrath at 2024-07-04T11:12:59-04:00 ttg: Split HsSrcBang, remove ref to DataCon from Syntax.Type Progress towards #21592 This splits HsSrcBang up, creating the new HsBang within `Language.Haskell.Syntax.Basic`. `HsBang` holds the unpackedness and strictness information, while `HsSrcBang` only adds the SourceText for usage within the compiler directly. Inside the AST, to preserve the SourceText, it is hidden behind the pre-existing extension point `XBindTy`. All other occurrences of `HsSrcBang` were adapted to deconstruct the inner `HsBang`, and when interacting with the `BindTy` constructor, the hidden `SourceText` is extracted/inserted into the `XBindTy` extension point. `GHC.Core.DataCon` exports both `HsSrcBang` and `HsBang` for convenience. A constructor function `mkHsSrcBang` that takes all individual components has been added. Two exceptions has been made though: - The `Outputable HsSrcBang` instance is replaced by `Outputable HsBang`. While being only GHC-internal, the only place it's used is in outputting `HsBangTy` constructors -- which already have `HsBang`. It wouldn't make sense to reconstruct a `HsSrcBang` just to ignore the `SourceText` anyway. - The error `TcRnUnexpectedAnnotation` did not use the `SourceText`, so it too now only holds a `HsBang`. - - - - - 24757fec by Mauricio at 2024-07-04T11:12:59-04:00 AST: Moved definitions that use GHC.Utils.Panic to GHC namespace Progress towards #21592 - - - - - 9be49379 by Mike Pilgrem at 2024-07-04T11:13:41-04:00 Fix #25032 Refer to Cabal's `includes` field, not `include-files` - - - - - 9e2ecf14 by Andrew Lelechenko at 2024-07-04T11:14:17-04:00 base: fix more missing changelog entries - - - - - a82121b3 by Peter Trommler at 2024-07-04T11:14:53-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 - - - - - fab13100 by Bryan Richter at 2024-07-04T11:15:29-04:00 Add .gitlab/README.md with creds instructions - - - - - 564981bd by Matthew Pickering at 2024-07-05T07:35:29-04:00 configure: Set LD_STAGE0 appropiately when 9.10.1 is used as a boot compiler In 9.10.1 the "ld command" has been removed, so we fall back to using the more precise "merge objects command" when it's available as LD_STAGE0 is only used to set the object merging command in hadrian. Fixes #24949 - - - - - a949c792 by Matthew Pickering at 2024-07-05T07:35:29-04:00 hadrian: Don't build ghci object files for ./hadrian/ghci target There is some convoluted logic which determines whether we build ghci object files are not. In any case, if you set `ghcDynPrograms = pure False` then it forces them to be built. Given we aren't ever building executables with this flavour it's fine to leave `ghcDynPrograms` as the default and it should be a bit faster to build less. Also fixes #24949 - - - - - 48bd8f8e by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Remove STG dump from ticky_ghc flavour transformer This adds 10-15 minutes to build time, it is a better strategy to precisely enable dumps for the modules which show up prominently in a ticky profile. Given I am one of the only people regularly building ticky compilers I think it's worthwhile to remove these. Fixes #23635 - - - - - 5b1aefb7 by Matthew Pickering at 2024-07-05T07:36:06-04:00 hadrian: Add dump_stg flavour transformer This allows you to write `--flavour=default+ticky_ghc+dump_stg` if you really want STG for all modules. - - - - - ab2b60b6 by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtToInstrs type There's no need to hand `Nothing`s around... (there was no case with a `BlockId`.) - - - - - 71a7fa8c by Sven Tennie at 2024-07-08T15:03:41-04:00 AArch64: Simplify stmtsToInstrs type The `BlockId` parameter (`bid`) is never used, only handed around. Deleting it simplifies the surrounding code. - - - - - 8bf6fd68 by Simon Peyton Jones at 2024-07-08T15:04:17-04:00 Fix eta-expansion in Prep As #25033 showed, we were eta-expanding in a way that broke a join point, which messed up Note [CorePrep invariants]. The fix is rather easy. See Wrinkle (EA1) of Note [Eta expansion of arguments in CorePrep] - - - - - 96acf823 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 One-shot Haddock - - - - - 74ec4c06 by Sjoerd Visscher at 2024-07-09T06:16:14-04:00 Remove haddock-stdout test option Superseded by output handling of Hadrian - - - - - ed8a8f0b by Rodrigo Mesquita at 2024-07-09T06:16:51-04:00 ghc-boot: Relax Cabal bound Fixes #25013 - - - - - 3f9548fe by Matthew Pickering at 2024-07-09T06:17:36-04:00 ci: Unset ALEX/HAPPY variables when testing bootstrap jobs Ticket #24826 reports a regression in 9.10.1 when building from a source distribution. This patch is an attempt to reproduce the issue on CI by more aggressively removing `alex` and `happy` from the environment. - - - - - aba2c9d4 by Andrea Bedini at 2024-07-09T06:17:36-04:00 hadrian: Ignore build-tool-depends fields in cabal files hadrian does not utilise the build-tool-depends fields in cabal files and their presence can cause issues when building source distribution (see #24826) Ideally Cabal would support building "full" source distributions which would remove the need for workarounds in hadrian but for now we can patch the build-tool-depends out of the cabal files. Fixes #24826 - - - - - 12bb9e7b by Matthew Pickering at 2024-07-09T06:18:12-04:00 testsuite: Don't attempt to link when checking whether a way is supported It is sufficient to check that the simple test file compiles as it will fail if there are not the relevant library files for the requested way. If you break a way so badly that even a simple executable fails to link (as I did for profiled dynamic way), it will just mean the tests for that way are skipped on CI rather than displayed. - - - - - 46ec0a8e by Torsten Schmits at 2024-07-09T13:37:02+02:00 Improve docs for NondecreasingIndentation The text stated that this affects indentation of layouts nested in do expressions, while it actually affects that of do layouts nested in any other. - - - - - dddc9dff by Zubin Duggal at 2024-07-12T11:41:24-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 - - - - - 145a6477 by Matthew Pickering at 2024-07-12T11:42:00-04:00 Add support for building profiled dynamic way The main payload of this change is to hadrian. * Default settings will produced dynamic profiled objects * `-fexternal-interpreter` is turned on in some situations when there is an incompatibility between host GHC and the way attempting to be built. * Very few changes actually needed to GHC There are also necessary changes to the bootstrap plans to work with the vendored Cabal dependency. These changes should ideally be reverted by the next GHC release. In hadrian support is added for building profiled dynamic libraries (nothing too exciting to see there) Updates hadrian to use a vendored Cabal submodule, it is important that we replace this usage with a released version of Cabal library before the 9.12 release. Fixes #21594 ------------------------- Metric Increase: libdir ------------------------- - - - - - 414a6950 by Matthew Pickering at 2024-07-12T11:42:00-04:00 testsuite: Make find_so regex more precise The hash contains lowercase [a-z0-9] and crucially not _p which meant we sometimes matched on `libHS.._p` profiled shared libraries rather than the normal shared library. - - - - - dee035bf by Alex Mason at 2024-07-12T11:42:41-04:00 ncg(aarch64): Add fsqrt instruction, byteSwap primitives [#24956] Implements the FSQRT machop using native assembly rather than a C call. Implements MO_BSwap by producing assembly to do the byte swapping instead of producing a foreign call a C function. In `tar`, the hot loop for `deserialise` got almost 4x faster by avoiding the foreign call which caused spilling live variables to the stack -- this means the loop did 4x more memory read/writing than necessary in that particular case! - - - - - 5104ee61 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). - - - - - 52d66984 by Sylvain Henry at 2024-07-12T11:43:23-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC - - - - - c34fef56 by Sylvain Henry at 2024-07-12T11:43:23-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). - - - - - 280e4bf5 by Simon Peyton Jones at 2024-07-12T11:43:59-04:00 Make type-equality on synonyms a bit faster This MR make equality fast for (S tys1 `eqType` S tys2), where S is a non-forgetful type synonym. It doesn't affect compile-time allocation much, but then comparison doesn't allocate anyway. But it seems like a Good Thing anyway. See Note [Comparing type synonyms] in GHC.Core.TyCo.Compare and Note [Forgetful type synonyms] in GHC.Core.TyCon Addresses #25009. - - - - - cb83c347 by Alan Zimmerman at 2024-07-12T11:44:35-04:00 EPA: Bring back SrcSpan in EpaDelta When processing files in ghc-exactprint, the usual workflow is to first normalise it with makeDeltaAst, and then operate on it. But we need the original locations to operate on it, in terms of finding things. So restore the original SrcSpan for reference in EpaDelta - - - - - 7bcda869 by Matthew Pickering at 2024-07-12T11:45:11-04:00 Update alpine release job to 3.20 alpine 3.20 was recently released and uses a new python and sphinx toolchain which could be useful to test. - - - - - 43aa99b8 by Matthew Pickering at 2024-07-12T11:45:11-04:00 testsuite: workaround bug in python-3.12 There is some unexplained change to binding behaviour in python-3.12 which requires moving this import from the top-level into the scope of the function. I didn't feel any particular desire to do a deep investigation as to why this changed as the code works when modified like this. No one in the python IRC channel seemed to know what the problem was. - - - - - e3914028 by Adam Sandberg Ericsson at 2024-07-12T11:45:47-04:00 initialise mmap_32bit_base during RTS startup #24847 - - - - - 86b8ecee by Hécate Kleidukos at 2024-07-12T11:46:27-04:00 haddock: Only fetch supported languages and extensions once per Interface list This reduces the number of operations done on each Interface, because supported languages and extensions are determined from architecture and operating system of the build host. This information remains stable across Interfaces, and as such doesn not need to be recovered for each Interface. - - - - - 4f85366f by sheaf at 2024-07-13T05:58:14-04:00 Testsuite: use py-cpuinfo to compute CPU features This replaces the rather hacky logic we had in place for checking CPU features. In particular, this means that feature availability now works properly on Windows. - - - - - 41f1354d by Matthew Pickering at 2024-07-13T05:58:51-04:00 testsuite: Replace $CC with $TEST_CC The TEST_CC variable should be set based on the test compiler, which may be different to the compiler which is set to CC on your system (for example when cross compiling). Fixes #24946 - - - - - 572fbc44 by sheaf at 2024-07-15T08:30:32-04:00 isIrrefutableHsPat: consider COMPLETE pragmas This patch ensures we taken into account COMPLETE pragmas when we compute whether a pattern is irrefutable. In particular, if a pattern synonym is the sole member of a COMPLETE pragma (without a result TyCon), then we consider a pattern match on that pattern synonym to be irrefutable. This affects the desugaring of do blocks, as it ensures we don't use a "fail" operation. Fixes #15681 #16618 #22004 - - - - - 84dadea9 by Zubin Duggal at 2024-07-15T08:31:09-04:00 haddock: Handle non-hs files, so that haddock can generate documentation for modules with foreign imports and template haskell. Fixes #24964 - - - - - 0b4ff9fa by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of warnings/deprecations from dependent packages in `InstalledInterface` and use this to propagate these on items re-exported from dependent packages. Fixes #25037 - - - - - b8b4b212 by Zubin Duggal at 2024-07-15T12:12:30-04:00 haddock: Keep track of instance source locations in `InstalledInterface` and use this to add source locations on out of package instances Fixes #24929 - - - - - 559a7a7c by Matthew Pickering at 2024-07-15T12:13:05-04:00 ci: Refactor job_groups definition, split up by platform The groups are now split up so it's easier to see which jobs are generated for each platform No change in behaviour, just refactoring. - - - - - 20383006 by Matthew Pickering at 2024-07-16T11:48:25+01:00 ci: Replace debian 10 with debian 12 on validation jobs Since debian 10 is now EOL we migrate onwards to debian 12 as the basis for most platform independent validation jobs. - - - - - 12d3b66c by Matthew Pickering at 2024-07-17T13:22:37-04:00 ghcup-metadata: Fix use of arch argument The arch argument was ignored when making the jobname, which lead to failures when generating metadata for the alpine_3_18-aarch64 bindist. Fixes #25089 - - - - - bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 8b626036 by Matthew Pickering at 2024-09-25T16:04:02+01:00 Add missing req_interp modifier to T18441fail3 and T18441fail19 These tests require the interpreter but they were failing in a different way with the javascript backend because the interpreter was disabled and stderr is ignored by the test. - - - - - 2f402ff2 by Matthew Pickering at 2024-09-25T16:04:03+01:00 Use explicit syntax rather than pure - - - - - fa2959f2 by Matthew Pickering at 2024-09-25T16:06:11+01:00 packaging: correctly propagate build/host/target to bindist configure script At the moment the host and target which we will produce a compiler for is fixed at the initial configure time. Therefore we need to persist the choice made at this time into the installation bindist as well so we look for the right tools, with the right prefixes at install time. In the future, we want to provide a bit more control about what kind of bindist we produce so the logic about what the host/target will have to be written by hadrian rather than persisted by the configure script. In particular with cross compilers we want to either build a normal stage 2 cross bindist or a stage 3 bindist, which creates a bindist which has a native compiler for the target platform. Fixes #21970 - - - - - 87d87b57 by Matthew Pickering at 2024-09-25T16:06:18+01:00 hadrian: Fill in more of the default.host toolchain file When you are building a cross compiler this file will be used to build stage1 and it's libraries, so we need enough information here to work accurately. There is still more work to be done (see for example, word size is still fixed). - - - - - b9610af1 by Matthew Pickering at 2024-09-25T16:06:18+01:00 hadrian: Disable docs when cross compiling Before there were a variety of ad-hoc places where doc building was disabled when cross compiling. * Some CI jobs sets --docs=none in gen_ci.hs * Some CI jobs set --docs=none in .gitlab/ci.sh * There was some logic in hadrian to not need the ["docs"] target when making a bindist. Now the situation is simple: * If you are cross compiling then defaultDocsTargets is empty by default. In theory, there is no reason why we can't build documentation for cross compiler bindists, but this is left to future work to generalise the documentation building rules to allow this (#24289) - - - - - 6e0757b3 by Matthew Pickering at 2024-09-25T16:40:01+01:00 hadrian: Build stage 2 cross compilers * Most of hadrian is abstracted over the stage in order to remove the assumption that the target of all stages is the same platform. This allows the RTS to be built for two different targets for example. * Abstracts the bindist creation logic to allow building either normal or cross bindists. Normal bindists use stage 1 libraries and a stage 2 compiler. Cross bindists use stage 2 libararies and a stage 2 compiler. * hadrian: Make binary-dist-dir the default build target. This allows us to have the logic in one place about which libraries/stages to build with cross compilers. Fixes #24192 New hadrian target: * `binary-dist-dir-cross`: Build a cross compiler bindist (compiler = stage 1, libraries = stage 2) ------------------------- Metric Decrease: T10421a T10858 T11195 T11276 T11374 T11822 T15630 T17096 T18478 T20261 Metric Increase: parsing001 ------------------------- - - - - - 613d1fa1 by Matthew Pickering at 2024-09-25T16:40:06+01:00 ci: Test cross bindists We remove the special logic for testing in-tree cross compilers and instead test cross compiler bindists, like we do for all other platforms. - - - - - c1f8c52b by Matthew Pickering at 2024-09-25T16:40:37+01:00 ci: Javascript don't set CROSS_EMULATOR There is no CROSS_EMULATOR needed to run javascript binaries, so we don't set the CROSS_EMULATOR to some dummy value. - - - - - 05a64696 by Matthew Pickering at 2024-09-25T16:40:40+01:00 ci: Introduce CROSS_STAGE variable In preparation for building and testing stage3 bindists we introduce the CROSS_STAGE variable which is used by a CI job to determine what kind of bindist the CI job should produce. At the moment we are only using CROSS_STAGE=2 but in the future we will have some jobs which set CROSS_STAGE=3 to produce native bindists for a target, but produced by a cross compiler, which can be tested on by another CI job on the native platform. CROSS_STAGE=2: Build a normal cross compiler bindist CROSS_STAGE=3: Build a stage 3 bindist, one which is a native compiler and library for the target - - - - - aebfe458 by Matthew Pickering at 2024-09-25T16:51:40+01:00 Split up system.config into host/target config files There were a number of settings which were not applied per-stage, for example if you specified `--ffi-include-dir` then that was applied to both host and target. Now this will just be passed when building the crosscompiler. The solution for now is to separate these two files into host/target and the host file contains very bare-bones . There isn't currently a way to specify with configure anything in the host file, so if you are building a cross-compiler and you need to do that, you have to modify the file yourself. - - - - - 23b0f6ab by Matthew Pickering at 2024-09-25T17:02:26+01:00 Fix unix commit - - - - - 15 changed files: - + .git-blame-ignore-revs - .gitignore - .gitlab-ci.yml - + .gitlab/README.md - .gitlab/ci.sh - .gitlab/darwin/nix/sources.json - .gitlab/darwin/toolchain.nix - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - + .gitlab/merge_request_templates/Haddock.md - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitlab/rel_eng/recompress-all - .gitmodules The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/645bd490cb3c367aae48fecc0929bbf51299097b...23b0f6ab4df1f0ea36bb3d105fc404e1ea842cd0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/645bd490cb3c367aae48fecc0929bbf51299097b...23b0f6ab4df1f0ea36bb3d105fc404e1ea842cd0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 16:03:52 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 25 Sep 2024 12:03:52 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 8 commits: packaging: correctly propagate build/host/target to bindist configure script Message-ID: <66f4346813d7e_123102adf9a855083@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: 318d496e by Matthew Pickering at 2024-09-25T17:03:44+01:00 packaging: correctly propagate build/host/target to bindist configure script At the moment the host and target which we will produce a compiler for is fixed at the initial configure time. Therefore we need to persist the choice made at this time into the installation bindist as well so we look for the right tools, with the right prefixes at install time. In the future, we want to provide a bit more control about what kind of bindist we produce so the logic about what the host/target will have to be written by hadrian rather than persisted by the configure script. In particular with cross compilers we want to either build a normal stage 2 cross bindist or a stage 3 bindist, which creates a bindist which has a native compiler for the target platform. Fixes #21970 - - - - - df6961b2 by Matthew Pickering at 2024-09-25T17:03:44+01:00 hadrian: Fill in more of the default.host toolchain file When you are building a cross compiler this file will be used to build stage1 and it's libraries, so we need enough information here to work accurately. There is still more work to be done (see for example, word size is still fixed). - - - - - 699b9425 by Matthew Pickering at 2024-09-25T17:03:44+01:00 hadrian: Disable docs when cross compiling Before there were a variety of ad-hoc places where doc building was disabled when cross compiling. * Some CI jobs sets --docs=none in gen_ci.hs * Some CI jobs set --docs=none in .gitlab/ci.sh * There was some logic in hadrian to not need the ["docs"] target when making a bindist. Now the situation is simple: * If you are cross compiling then defaultDocsTargets is empty by default. In theory, there is no reason why we can't build documentation for cross compiler bindists, but this is left to future work to generalise the documentation building rules to allow this (#24289) - - - - - d378d7a3 by Matthew Pickering at 2024-09-25T17:03:45+01:00 hadrian: Build stage 2 cross compilers * Most of hadrian is abstracted over the stage in order to remove the assumption that the target of all stages is the same platform. This allows the RTS to be built for two different targets for example. * Abstracts the bindist creation logic to allow building either normal or cross bindists. Normal bindists use stage 1 libraries and a stage 2 compiler. Cross bindists use stage 2 libararies and a stage 2 compiler. * hadrian: Make binary-dist-dir the default build target. This allows us to have the logic in one place about which libraries/stages to build with cross compilers. Fixes #24192 New hadrian target: * `binary-dist-dir-cross`: Build a cross compiler bindist (compiler = stage 1, libraries = stage 2) ------------------------- Metric Decrease: T10421a T10858 T11195 T11276 T11374 T11822 T15630 T17096 T18478 T20261 Metric Increase: parsing001 ------------------------- - - - - - ffe5f00a by Matthew Pickering at 2024-09-25T17:03:45+01:00 ci: Test cross bindists We remove the special logic for testing in-tree cross compilers and instead test cross compiler bindists, like we do for all other platforms. - - - - - 69123eba by Matthew Pickering at 2024-09-25T17:03:45+01:00 ci: Javascript don't set CROSS_EMULATOR There is no CROSS_EMULATOR needed to run javascript binaries, so we don't set the CROSS_EMULATOR to some dummy value. - - - - - 191998d1 by Matthew Pickering at 2024-09-25T17:03:45+01:00 ci: Introduce CROSS_STAGE variable In preparation for building and testing stage3 bindists we introduce the CROSS_STAGE variable which is used by a CI job to determine what kind of bindist the CI job should produce. At the moment we are only using CROSS_STAGE=2 but in the future we will have some jobs which set CROSS_STAGE=3 to produce native bindists for a target, but produced by a cross compiler, which can be tested on by another CI job on the native platform. CROSS_STAGE=2: Build a normal cross compiler bindist CROSS_STAGE=3: Build a stage 3 bindist, one which is a native compiler and library for the target - - - - - 7dec8d6d by Matthew Pickering at 2024-09-25T17:03:45+01:00 Split up system.config into host/target config files There were a number of settings which were not applied per-stage, for example if you specified `--ffi-include-dir` then that was applied to both host and target. Now this will just be passed when building the crosscompiler. The solution for now is to separate these two files into host/target and the host file contains very bare-bones . There isn't currently a way to specify with configure anything in the host file, so if you are building a cross-compiler and you need to do that, you have to modify the file yourself. - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - configure.ac - distrib/configure.ac.in - hadrian/README.md - hadrian/bindist/config.mk.in - hadrian/cfg/default.host.target.in - + hadrian/cfg/system.config.host.in - hadrian/cfg/system.config.in - + hadrian/cfg/system.config.target.in - hadrian/hadrian.cabal - hadrian/src/Base.hs - + hadrian/src/BindistConfig.hs - hadrian/src/Builder.hs - hadrian/src/Context.hs - hadrian/src/Expression.hs - hadrian/src/Flavour.hs - hadrian/src/Flavour/Type.hs - hadrian/src/Hadrian/Expression.hs - hadrian/src/Hadrian/Haskell/Hash.hs - hadrian/src/Hadrian/Oracles/TextFile.hs - hadrian/src/Oracles/Flag.hs - hadrian/src/Oracles/Flavour.hs - hadrian/src/Oracles/Setting.hs - hadrian/src/Oracles/TestSettings.hs - hadrian/src/Packages.hs - hadrian/src/Rules.hs - hadrian/src/Rules/BinaryDist.hs - hadrian/src/Rules/CabalReinstall.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/23b0f6ab4df1f0ea36bb3d105fc404e1ea842cd0...7dec8d6d0196ac509bff76039df2a6227439e668 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/23b0f6ab4df1f0ea36bb3d105fc404e1ea842cd0...7dec8d6d0196ac509bff76039df2a6227439e668 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 16:19:14 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Wed, 25 Sep 2024 12:19:14 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] Apply 2 suggestion(s) to 2 file(s) Message-ID: <66f4380248cf0_20b32110abbc46828@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 52065624 by Hassan Al-Awwadi at 2024-09-25T16:19:11+00:00 Apply 2 suggestion(s) to 2 file(s) Co-authored-by: Rodrigo Mesquita <rodrigo.m.mesquita at gmail.com> - - - - - 2 changed files: - compiler/GHC/Rename/Pat.hs - compiler/Language/Haskell/Syntax/Type.hs Changes: ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -259,42 +259,6 @@ newPatName (LetMk is_top fix_env) rdr_name NotTopLevel -> newLocalBndrRn rdr_name TopLevel -> newTopSrcBinder rdr_name ; bindLocalNames [name] $ - -- Do *not* use bindLocalNameFV here; - -- Do *not* use bindLocalNameFV here; - -- Do *not* use bindLocalNameFV here; - -- Do *not* use bindLocalNameFV here; - -- see Note [View pattern usage] - -- see Note [View pattern usage] - -- see Note [View pattern usage] - -- see Note [View pattern usage] - -- For the TopLevel case - -- For the TopLevel case - -- For the TopLevel case - -- For the TopLevel case - -- see Note [bindLocalNames for an External name] - -- see Note [bindLocalNames for an External name] - -- see Note [bindLocalNames for an External name] - -- see Note [bindLocalNames for an External name] - - -- Do *not* use bindLocalNameFV here; - - -- Do *not* use bindLocalNameFV here; - -- see Note [View pattern usage] - -- see Note [View pattern usage] - -- For the TopLevel case - -- For the TopLevel case - -- see Note [bindLocalNames for an External name] - -- see Note [bindLocalNames for an External nam - - -- Do *not* use bindLocalNameFV here; - -- Do *not* use bindLocalNameFV here; - -- see Note [View pattern usage] - -- see Note [View pattern usage] - -- For the TopLevel case - -- For the TopLevel case - -- see Note [bindLocalNames for an External name] - -- see Note [bindLocalNames for an External name] - -- Do *not* use bindLocalNameFV here; -- see Note [View pattern usage] -- For the TopLevel case ===================================== compiler/Language/Haskell/Syntax/Type.hs ===================================== @@ -1266,8 +1266,14 @@ The SrcSpan is the span of the original HsPar * * ************************************************************************ -} +-- | Located Update Field Occurrence type LUpdFieldOcc pass = XRec pass (UpdFieldOcc pass) +-- | An update field occurrence is a field occurrence that +-- occurs in an update position (such as @x{field=...}@). +-- +-- We differentiate between the two since there may be additional information concerning the update field. +-- In particular, in GHC, an update field occurrence *may* be ambiguous, unlike other field occurrences. data UpdFieldOcc pass = UpdFieldOcc { ufoExt :: XCUpdFieldOcc pass, View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/52065624c9d98a23bb5500ca361413b109aae9a3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/52065624c9d98a23bb5500ca361413b109aae9a3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 16:23:51 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Wed, 25 Sep 2024 12:23:51 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] remove resolved comments Message-ID: <66f439177627_20b321c5ef44749c@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 5d80a324 by Hassan Al-Awwadi at 2024-09-25T18:23:27+02:00 remove resolved comments - - - - - 3 changed files: - compiler/GHC/Hs/Type.hs - compiler/GHC/Rename/Bind.hs - compiler/GHC/Rename/Pat.hs Changes: ===================================== compiler/GHC/Hs/Type.hs ===================================== @@ -1115,10 +1115,8 @@ updFieldOccLRdrName (UpdFieldOcc _ (L _ fo)) = fieldOccLRdrName fo updFieldOccLRdrName (XUpdFieldOcc xfo) = case ghcPass @p of GhcRn -> case xfo of Ambiguous l -> l - -- Are these cases required? I don't get a hint that they aren't so - -- they are included for now. - GhcPs -> case xfo of - GhcTc -> case xfo of + GhcPs -> dataConCantHappen xfo + GhcTc -> dataConCantHappen xfo fieldOccRdrName :: forall p. IsPass p => FieldOcc (GhcPass p) -> RdrName fieldOccRdrName fo = case ghcPass @p of ===================================== compiler/GHC/Rename/Bind.hs ===================================== @@ -823,10 +823,8 @@ rnPatSynBind sig_fn bind@(PSB { psb_id = L l name , psb_dir = dir' , psb_ext = fvs' } selector_names = case details' of - RecCon names -> - -- I don't actually know if its fine or not - -- that foLabel is partial. - map (unLoc . foLabel . recordPatSynField) names + RecCon names -> + map (unLoc . foLabel . recordPatSynField) names _ -> [] ; fvs' `seq` -- See Note [Free-variable space leak] ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -1043,10 +1043,8 @@ rnHsRecUpdFields flds , hfbPun = pun }) ; (flds', fvs') <- rn_flds pun_ok (tail <$> mb_unambig_lbls) flds ; return (fld' : flds', fvs `plusFV` fvs') } - -- I have been told this datacon isn't necessary and yet my ghc gets mad at me - -- when I don't add it. rn_flds _ _ ((L _ (HsFieldBind { hfbLHS = L _ (XUpdFieldOcc impossible ) })):_) - = case impossible of + = dataConCantHappen impossible getFieldIds :: [LHsRecField GhcRn arg] -> [Name] getFieldIds flds = map (hsRecFieldSel . unLoc) flds View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5d80a32415fc4397aa39e701bb08a77731a6e46c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5d80a32415fc4397aa39e701bb08a77731a6e46c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 16:38:33 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Wed, 25 Sep 2024 12:38:33 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] eleminated outdated comments and use dataConCantHappen Message-ID: <66f43c89467c2_20b3213a6c0449563@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 65eb22cb by Hassan Al-Awwadi at 2024-09-25T18:37:45+02:00 eleminated outdated comments and use dataConCantHappen - - - - - 3 changed files: - compiler/GHC/Hs/Type.hs - compiler/GHC/Rename/Bind.hs - compiler/GHC/Rename/Pat.hs Changes: ===================================== compiler/GHC/Hs/Type.hs ===================================== @@ -1115,10 +1115,8 @@ updFieldOccLRdrName (UpdFieldOcc _ (L _ fo)) = fieldOccLRdrName fo updFieldOccLRdrName (XUpdFieldOcc xfo) = case ghcPass @p of GhcRn -> case xfo of Ambiguous l -> l - -- Are these cases required? I don't get a hint that they aren't so - -- they are included for now. - GhcPs -> case xfo of - GhcTc -> case xfo of + GhcPs -> dataConCantHappen xfo + GhcTc -> dataConCantHappen xfo fieldOccRdrName :: forall p. IsPass p => FieldOcc (GhcPass p) -> RdrName fieldOccRdrName fo = case ghcPass @p of ===================================== compiler/GHC/Rename/Bind.hs ===================================== @@ -824,9 +824,7 @@ rnPatSynBind sig_fn bind@(PSB { psb_id = L l name , psb_ext = fvs' } selector_names = case details' of RecCon names -> - -- I don't actually know if its fine or not - -- that foLabel is partial. - map (unLoc . foLabel . recordPatSynField) names + map (unLoc . foLabel . recordPatSynField) names _ -> [] ; fvs' `seq` -- See Note [Free-variable space leak] ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -1043,10 +1043,8 @@ rnHsRecUpdFields flds , hfbPun = pun }) ; (flds', fvs') <- rn_flds pun_ok (tail <$> mb_unambig_lbls) flds ; return (fld' : flds', fvs `plusFV` fvs') } - -- I have been told this datacon isn't necessary and yet my ghc gets mad at me - -- when I don't add it. rn_flds _ _ ((L _ (HsFieldBind { hfbLHS = L _ (XUpdFieldOcc impossible ) })):_) - = case impossible of + = dataConCantHappen impossible getFieldIds :: [LHsRecField GhcRn arg] -> [Name] getFieldIds flds = map (hsRecFieldSel . unLoc) flds View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/65eb22cb743abeb365893b9f871d12ce60512db7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/65eb22cb743abeb365893b9f871d12ce60512db7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 16:44:09 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Wed, 25 Sep 2024 12:44:09 -0400 Subject: [Git][ghc/ghc][wip/T25281] Wibbles Message-ID: <66f43dd9c99d4_20b3214a3c10505fa@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: 20b07472 by Simon Peyton Jones at 2024-09-25T17:43:55+01:00 Wibbles - - - - - 1 changed file: - compiler/GHC/HsToCore/Expr.hs Changes: ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -259,38 +259,11 @@ dsLExpr (L loc e) = putSrcSpanDsA loc $ dsExpr e -- | Desugar a typechecked expression. dsExpr :: HsExpr GhcTc -> DsM CoreExpr -dsExpr (HsVar _ (L _ id)) = dsHsVar id - -{- Record selectors are warned about if they are not -present in all of the parent data type's constructor, -or always in case of pattern synonym record selectors -(regulated by a flag). However, this only produces -a warning if it's not a part of a record selector -application. For example: - - data T = T1 | T2 {s :: Bool} - f x = s x -- the warning from this case will be supressed - -See the `HsApp` case for where it is filtered out --} -dsExpr (HsRecSel _ (FieldOcc id _)) - = do { let name = getName id - RecSelId {sel_cons = (_, cons_wo_field)} - = idDetails id - ; cons_trimmed <- trim_cons cons_wo_field - ; unless (null cons_wo_field) $ diagnosticDs - $ DsIncompleteRecordSelector name cons_trimmed (cons_trimmed /= cons_wo_field) - -- This only produces a warning if it's not a part of a - -- record selector application (e.g. `s a` where `s` is a selector) - -- See the `HsApp` case for where it is filtered out - ; dsHsVar id } - where - trim_cons :: [ConLike] -> DsM [ConLike] - trim_cons cons_wo_field = do - dflags <- getDynFlags - let maxConstructors = maxUncoveredPatterns dflags - return $ take maxConstructors cons_wo_field +dsExpr e@(HsVar {}) = dsApp e +dsExpr e@(HsApp {}) = dsApp e +dsExpr e@(HsAppType {}) = dsApp e +dsExpr e@(HsRecSel {}) = dsApp e dsExpr (HsUnboundVar (HER ref _ _) _) = dsEvTerm =<< readMutVar ref -- See Note [Holes] in GHC.Tc.Types.Constraint @@ -298,11 +271,6 @@ dsExpr (HsUnboundVar (HER ref _ _) _) = dsEvTerm =<< readMutVar ref dsExpr (HsPar _ e) = dsLExpr e dsExpr (ExprWithTySig _ e _) = dsLExpr e -dsExpr (HsIPVar x _) = dataConCantHappen x - -dsExpr (HsGetField x _ _) = dataConCantHappen x -dsExpr (HsProjection x _) = dataConCantHappen x - dsExpr (HsLit _ lit) = do { warnAboutOverflowedLit lit ; dsLit (convertLit lit) } @@ -313,12 +281,13 @@ dsExpr (HsOverLit _ lit) dsExpr e@(XExpr ext_expr_tc) = case ext_expr_tc of + WrapExpr {} -> dsApp e + ConLikeTc {} -> dsApp e + ExpandedThingTc o e | OrigStmt (L loc _) <- o -> putSrcSpanDsA loc $ dsExpr e | otherwise -> dsExpr e - WrapExpr {} -> dsHsWrapped e - ConLikeTc con tvs tys -> dsConLike con tvs tys -- Hpc Support HsTick tickish e -> do e' <- dsLExpr e @@ -347,7 +316,6 @@ dsExpr (NegApp _ (L loc -- See Note [Checking "negative literals"] (lit { ol_val = HsIntegral (negateIntegralLit i) }) ; dsOverLit lit } - ; ; dsSyntaxExpr neg_expr [mkTicks ts expr'] } dsExpr (NegApp _ expr neg_expr) @@ -357,69 +325,6 @@ dsExpr (NegApp _ expr neg_expr) dsExpr (HsLam _ variant a_Match) = uncurry mkCoreLams <$> matchWrapper (LamAlt variant) Nothing a_Match -dsExpr e@(HsApp _ fun arg) - = dsApp e [] - -dsLApp :: LHsExpr GhcTc -> [CoreExpr] -> DsM CoreExpr -dsLApp (L loc e) core_args = dsApp e core_args - -dsApp :: HsExpr GhcTc -> [CoreExpr] -> DsM CoreExpr - -dsApp (HsPar _ e) core_args = dsLApp e core_args - -dsApp (HsApp _ fun arg) core_args - = do { core_arg <- dsLExpr arg - ; dsLApp fun (core_arg : core_args) } - -dsApp (HsAppType arg_ty fun _) core_args - = dsLApp fun (Type ty : core_args) - -dsApp (XExpr (WrapExpr hs_wrap fun)) core_args - = do { (fun_wrap, all_args) <- split_wrap hs_wrap core_args - ; if isIdHsWrapper fun_wrap - then dsLApp fun all_args - else ... } - where - split_wrap :: HsWrapper -> [CoreArg] -> DsM (HsWrapper, [CoreArg]) - split_wrap (WpTyApp ty) args - = return (WpHole, Type ty : args) - split_wrap (WpEvApp tm) args - = do { core_tm <- dsEvTerm tm - ; (WpHole, Type ty : args) - split_wrap (WpCompose w1 w2) args - = do { (w2', args') <- split_wrap w2 args - ; if isIdHsWrapper w2' - then split_wrap w1 args' - else return (w1 <.> w2', args') } - split_wrap wrap args - = return (wrap, args) - - -dsApp (HsVar _ v) core_args - -- We want to have a special case that uses the PMC information to filter - -- out some of the incomplete record selectors warnings and not trigger - -- the warning emitted during the desugaring of dsExpr(HsRecSel) - -- See Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc - = do { (msgs, fun') <- captureMessagesDs $ dsLExpr fun - -- Make sure to filter out the generic incomplete record selector warning - -- if it's a raw record selector - ; arg' <- dsLExpr arg - ; case getIdFromTrivialExpr_maybe fun' of - Just fun_id | isRecordSelector fun_id - -> do { let msgs' = filterMessages is_incomplete_rec_sel_msg msgs - ; addMessagesDs msgs' - ; pmcRecSel fun_id arg' } - _ -> addMessagesDs msgs - ; warnUnusedBindValue fun arg (exprType arg') - ; return $ mkCoreAppDs (text "HsApp" <+> ppr e) fun' arg' } - where - is_incomplete_rec_sel_msg :: MsgEnvelope DsMessage -> Bool - is_incomplete_rec_sel_msg (MsgEnvelope {errMsgDiagnostic = DsIncompleteRecordSelector{}}) - = False - is_incomplete_rec_sel_msg _ = True - - -dsExpr e@(HsAppType {}) = dsHsWrapped e {- Note [Checking "negative literals"] @@ -480,13 +385,17 @@ dsExpr (ExplicitTuple _ tup_args boxity) dsExpr (ExplicitSum types alt arity expr) = mkCoreUnboxedSum arity alt types <$> dsLExpr expr -dsExpr (HsPragE _ prag expr) = - ds_prag_expr prag expr - -dsExpr (HsEmbTy x _) = dataConCantHappen x -dsExpr (HsQual x _ _) = dataConCantHappen x -dsExpr (HsForAll x _ _) = dataConCantHappen x -dsExpr (HsFunArr x _ _ _) = dataConCantHappen x +dsExpr (HsPragE _ (HsPragSCC _ cc)) expr + = do { dflags <- getDynFlags + ; if sccProfilingEnabled dflags && gopt Opt_ProfManualCcs dflags + then do + mod_name <- getModule + count <- goptM Opt_ProfCountEntries + let nm = sl_fs cc + flavour <- mkExprCCFlavour <$> getCCIndexDsM nm + Tick (ProfNote (mkUserCC nm mod_name (getLocA expr) flavour) count True) + <$> dsLExpr expr + else dsLExpr expr } dsExpr (HsCase ctxt discrim matches) = do { core_discrim <- dsLExpr discrim @@ -528,11 +437,6 @@ dsExpr (HsMultiIf res_ty alts) mkErrorExpr = mkErrorAppDs nON_EXHAUSTIVE_GUARDS_ERROR_ID res_ty (text "multi-way if") -{- -\noindent -\underline{\bf Various data construction things} - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --} dsExpr (ExplicitList elt_ty xs) = dsExplicitList elt_ty xs @@ -620,7 +524,6 @@ dsExpr (RecordCon { rcon_con = L _ con_like ; return (mkCoreApps con_expr' con_args) } -dsExpr (RecordUpd x _ _) = dataConCantHappen x -- Here is where we desugar the Template Haskell brackets and escapes @@ -635,27 +538,158 @@ dsExpr (HsUntypedSplice ext _) = dataConCantHappen ext -- Arrow notation extension dsExpr (HsProc _ pat cmd) = dsProcExpr pat cmd - -- HsSyn constructs that just shouldn't be here, because --- the renamer removed them. See GHC.Rename.Expr. +-- the renamer or typechecker removed them. See GHC.Rename.Expr. -- Note [Handling overloaded and rebindable constructs] -dsExpr (HsOverLabel x _) = dataConCantHappen x -dsExpr (OpApp x _ _ _) = dataConCantHappen x -dsExpr (SectionL x _ _) = dataConCantHappen x -dsExpr (SectionR x _ _) = dataConCantHappen x +dsExpr (HsIPVar x _) = dataConCantHappen x +dsExpr (HsGetField x _ _) = dataConCantHappen x +dsExpr (HsProjection x _) = dataConCantHappen x +dsExpr (RecordUpd x _ _) = dataConCantHappen x +dsExpr (HsEmbTy x _) = dataConCantHappen x +dsExpr (HsQual x _ _) = dataConCantHappen x +dsExpr (HsForAll x _ _) = dataConCantHappen x +dsExpr (HsFunArr x _ _ _) = dataConCantHappen x +dsExpr (HsOverLabel x _) = dataConCantHappen x +dsExpr (OpApp x _ _ _) = dataConCantHappen x +dsExpr (SectionL x _ _) = dataConCantHappen x +dsExpr (SectionR x _ _) = dataConCantHappen x + + +{- ********************************************************************* +* * +* Desugaring applications +* * +********************************************************************* -} + +dsApp :: HsExpr -> DsM CoreExpr +dsApp e = ds_app e [] [] + +---------------------- +ds_lapp :: LHsExpr GhcTc -> [LHsExpr GhcTc] -> [CoreExpr] -> DsM CoreExpr +-- The [LHsExpr] args correspond to the [CoreExpr] args, +-- but there may be more of the latter because they include +-- type and dictionary arguments +ds_lapp (L loc e) hs_args core_args + = ds_app e hs_args core_args + +ds_app :: HsExpr GhcTc -> [LHsExpr] -> [CoreExpr] -> DsM CoreExpr +-- The work-horse +ds_app (HsPar _ e) hs_args core_args = dsLApp e hs_args core_args + +ds_app (HsApp _ fun arg) hs_args core_args + = do { core_arg <- dsLExpr arg + ; dsLApp fun (arg : hs_args) (core_arg : core_args) } + +ds_app (HsAppType arg_ty fun _) hs_args core_args + = dsLApp fun hs_args (Type ty : core_args) + +ds_app (XExpr (WrapExpr hs_wrap fun)) hs_args core_args + = do { (fun_wrap, all_args) <- splitHsWrapperArgs hs_wrap core_args + ; if isIdHsWrapper fun_wrap + then ds_lapp fun hs_args all_args + else do { core_fun <- dsHsWrapper fun_wrap $ \core_wrap -> + do { core_fun <- dsLExpr fun + ; return (fun_wrap core_fun) } + ; return (mkCoreApps core_fun all_args) } } + +ds_app (XExpr (ConLikeTc con tvs tys)) _hs_args core_args +-- Desugar desugars 'ConLikeTc': it eta-expands +-- data constructors to make linear types work. +-- See Note [Typechecking data constructors] in GHC.Tc.Gen.Head + = do { ds_con <- dsHsConLike con + ; ids <- newSysLocalsDs tys + -- NB: these 'Id's may be representation-polymorphic; + -- see Wrinkle [Representation-polymorphic lambda] in + -- Note [Typechecking data constructors] in GHC.Tc.Gen.Head. + ; let core_fun = mkLams tvs $ mkLams ids $ + ds_con `mkTyApps` mkTyVarTys tvs + `mkVarApps` ids + ; return (mkApps core_fun core_args) } + +ds_app (HsRecSel _ fld_occ@(FieldOcc { foExt = fun_id })) hs_args core_args + | RecSelId {sel_cons = (_, cons_wo_field)} <- idDetails id + = do { dflags <- getDynFlags + + -- Record selectors are warned about if they are not present in all of the + -- parent data type's constructor, or always in case of pattern synonym record + -- selectors (regulated by a flag). However, this only produces a warning if + -- it's not a part of a record selector application. For example: + -- data T = T1 | T2 {s :: Bool} + -- g y = map s y -- Warn here + -- f x = s x -- No warning here + ; let maxConstructors = maxUncoveredPatterns dflags + cons_trimmed = take maxConstructors cons_wo_field + ; when (null hs_args && not (null cons_wo_field)) $ + diagnosticDs $ DsIncompleteRecordSelector (idName fun_id) cons_trimmed + (cons_trimmed /= cons_wo_field) + + -- Type-based check + -- ToDo: explain + ; case filterOut isTypeArg core_args of + (arg:_) -> pmcRecSel fun_id arg + _ -> return () + + -- Warn about unused return value in + -- do { ...; e; ... } when e returns (say) an Int + ; warnUnusedMonadBindValue fld_occ hs_args core_args + + ; return (mkCoreApps (Var fun_id) core_args) } + +-- We want to have a special case that uses the PMC information to filter +-- out some of the incomplete record selectors warnings and not trigger +-- the warning emitted during the desugaring of dsExpr(HsRecSel) +-- See Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + +ds_app (HsVar _ (L _ fun_id)) core_args + | fun_id `hasKey` noinlineIdKey -- See Note [noinlineId magic] in GHC.Types.Id.Make + , Type _ : arg1 : rest_args <- core_args + , (inner_fun, inner_args) <- collectArgs arg1 + = return (Var fun_id `App` Type (exprType inner_fun) `App` inner_fun) + `mkCoreApps` inner_args `mkCoreApps` rest_args + + | fun_id `hasKey` seqIdKey -- Note [Desugaring seq], points (1) and (2) + , Type _r : Type ty1 : Type ty2 : arg1 : arg2 : rest_args + = reuturn (Case arg1 case_bndr ty2 [Alt DEFAULT [] (mkCoreApps arg2 rest_args)]) + where + case_bndr = case arg1 of + Var v1 | isInternalName (idName v1) + -> v1 -- Note [Desugaring seq], points (2) and (3) + _ -> mkWildValBinder ManyTy ty1 + + | otherwise + = return (mkCoreApps (Var fun_id) core_args) + +------------------------------ +splitHsWrapArgs :: HsWrapper -> [CoreArg] -> DsM (HsWrapper, [CoreArg]) +-- Splits the wrapper into the trailing arguments, and leftover bit +splitHsWrapArg w args = go w args + where + go (WpTyApp ty) args = return (WpHole, Type ty : args) + go (WpEvApp tm) args = do { core_tm <- dsEvTerm tm + ; (WpHole, Type ty : args) + go (WpCompose w1 w2) args + = do { (w2', args') <- split_wrap w2 args + ; if isIdHsWrapper w2' + then go w1 args' + else return (w1 <.> w2', args') } + go wrap wrap args = return (wrap, args) + + +------------------------------ +dsHsConLike :: ConLike -> DsM CoreExpr +dsHsConLike (RealDataCon dc) + = return (varToCoreExpr (dataConWrapId dc)) +dsHsConLike (PatSynCon ps) + | Just (builder_name, _, add_void) <- patSynBuilder ps + = do { builder_id <- dsLookupGlobalId builder_name + ; return (if add_void + then mkCoreApp (text "dsConLike" <+> ppr ps) + (Var builder_id) unboxedUnitExpr + else Var builder_id) } + | otherwise + = pprPanic "dsConLike" (ppr ps) + -ds_prag_expr :: HsPragE GhcTc -> LHsExpr GhcTc -> DsM CoreExpr -ds_prag_expr (HsPragSCC _ cc) expr = do - dflags <- getDynFlags - if sccProfilingEnabled dflags && gopt Opt_ProfManualCcs dflags - then do - mod_name <- getModule - count <- goptM Opt_ProfCountEntries - let nm = sl_fs cc - flavour <- mkExprCCFlavour <$> getCCIndexDsM nm - Tick (ProfNote (mkUserCC nm mod_name (getLocA expr) flavour) count True) - <$> dsLExpr expr - else dsLExpr expr ------------------------------ dsSyntaxExpr :: SyntaxExpr GhcTc -> [CoreExpr] -> DsM CoreExpr @@ -943,49 +977,7 @@ Other places that requires from the same treatment: because we already know 'y' is of the form "Just ...". See test case T21360b. -************************************************************************ -* * - Desugaring Variables -* * -************************************************************************ --} - -dsHsVar :: Id -> DsM CoreExpr --- We could just call dsHsUnwrapped; but this is a short-cut --- for the very common case of a variable with no wrapper. -dsHsVar var - = return (varToCoreExpr var) -- See Note [Desugaring vars] -dsHsConLike :: ConLike -> DsM CoreExpr -dsHsConLike (RealDataCon dc) - = return (varToCoreExpr (dataConWrapId dc)) -dsHsConLike (PatSynCon ps) - | Just (builder_name, _, add_void) <- patSynBuilder ps - = do { builder_id <- dsLookupGlobalId builder_name - ; return (if add_void - then mkCoreApp (text "dsConLike" <+> ppr ps) - (Var builder_id) unboxedUnitExpr - else Var builder_id) } - | otherwise - = pprPanic "dsConLike" (ppr ps) - --- | This function desugars 'ConLikeTc': it eta-expands --- data constructors to make linear types work. --- --- See Note [Typechecking data constructors] in GHC.Tc.Gen.Head -dsConLike :: ConLike -> [TcTyVar] -> [Scaled Type] -> DsM CoreExpr -dsConLike con tvs tys - = do { ds_con <- dsHsConLike con - ; ids <- newSysLocalsDs tys - -- NB: these 'Id's may be representation-polymorphic; - -- see Wrinkle [Representation-polymorphic lambda] in - -- Note [Typechecking data constructors] in GHC.Tc.Gen.Head. - ; return (mkLams tvs $ - mkLams ids $ - ds_con `mkTyApps` mkTyVarTys tvs - `mkVarApps` ids) } - -{- ************************************************************************ * * \subsection{Errors and contexts} @@ -994,26 +986,16 @@ dsConLike con tvs tys -} -- Warn about certain types of values discarded in monadic bindings (#3263) -warnUnusedBindValue :: LHsExpr GhcTc -> LHsExpr GhcTc -> Type -> DsM () -warnUnusedBindValue fun arg@(L loc _) arg_ty - | Just (l, f) <- fish_var fun - , f `hasKey` thenMClassOpKey -- it is a (>>) - = when (isGeneratedSrcSpan l) $ -- it is compiler generated (>>) - putSrcSpanDs (locA loc) $ warnDiscardedDoBindings arg arg_ty - where - -- Retrieve the location info and the head of the application - -- It is important that we /do not/ look through HsApp to avoid - -- generating duplicate warnings - -- See Part 2. of Note [Expanding HsDo with XXExprGhcRn] - fish_var :: LHsExpr GhcTc -> Maybe (SrcSpan , Id) - fish_var (L l (HsVar _ id)) = return (locA l, unLoc id) - fish_var (L _ (HsAppType _ e _)) = fish_var e - fish_var (L l (XExpr (WrapExpr _ e))) = do (l, e') <- fish_var (L l e) - return (l, e') - fish_var (L l (XExpr (ExpandedThingTc _ e))) = fish_var (L l e) - fish_var _ = Nothing - -warnUnusedBindValue _ _ _ = return () +warnUnusedMonadBindValue :: FieldOcc GhcTc -> [LHsExpr GhcTc] -> [CoreExpr] -> Type -> DsM () +warnUnusedMonadBindValue (FieldOcc { foExt = fun_id, foLabel = L loc _ }) + hs_args core_args + | fun_id `hasKey` thenMClassOpKey -- it is a (>>) + , isGeneratedSrcSpan loc -- it is compiler generated (>>) + , Type arg_ty : _ <- core_args + , hs_arg : _ <- hs_args + = putSrcSpanDs (locA loc) $ warnDiscardedDoBindings hs_arg arg_ty + | otherwise + = return () -- Warn about certain types of values discarded in monadic bindings (#3263) warnDiscardedDoBindings :: LHsExpr GhcTc -> Type -> DsM () View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/20b07472ec530556a4e16a5e625a9f0652a09a6e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/20b07472ec530556a4e16a5e625a9f0652a09a6e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 17:12:38 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Wed, 25 Sep 2024 13:12:38 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] Link bytecode from interface-stored core bindings in oneshot mode Message-ID: <66f44486b4d6_20b3216466e457462@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 6b43020b by Cheng Shao at 2024-09-25T19:11:11+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 20 changed files: - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/A.hs - + testsuite/tests/bytecode/T25090/B.hs - + testsuite/tests/bytecode/T25090/C.hs - + testsuite/tests/bytecode/T25090/C.hs-boot - + testsuite/tests/bytecode/T25090/D.hs - + testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/T25090-debug.stderr - + testsuite/tests/bytecode/T25090/T25090.stdout - + testsuite/tests/bytecode/T25090/all.T Changes: ===================================== compiler/GHC/CoreToIface.hs ===================================== @@ -602,7 +602,7 @@ toIfaceTopBind b = in (top_bndr, rhs') -- The sharing behaviour is currently disabled due to #22807, and relies on - -- finished #220056 to be re-enabled. + -- finished #20056 to be re-enabled. disabledDueTo22807 = True already_has_unfolding b = not disabledDueTo22807 @@ -774,8 +774,8 @@ outside of the hs-boot loop. Note [Interface File with Core: Sharing RHSs] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -IMPORTANT: This optimisation is currently disabled due to #22027, it can be - re-enabled once #220056 is implemented. +IMPORTANT: This optimisation is currently disabled due to #22807, it can be + re-enabled once #22056 is implemented. In order to avoid duplicating definitions for bindings which already have unfoldings we do some minor headstands to avoid serialising the RHS of a definition if it has ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -260,7 +260,6 @@ outputForeignStubs Maybe FilePath) -- C file created outputForeignStubs logger tmpfs dflags unit_state mod location stubs = do - let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location stub_c <- newTempName logger tmpfs (tmpDir dflags) TFL_CurrentModule "c" case stubs of @@ -276,8 +275,6 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs stub_h_output_d = pprCode h_code stub_h_output_w = showSDoc dflags stub_h_output_d - createDirectoryIfMissing True (takeDirectory stub_h) - putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export header file" FormatC @@ -299,9 +296,20 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs | platformMisc_libFFI $ platformMisc dflags = "#include \"rts/ghc_ffi.h\"\n" | otherwise = "" - stub_h_file_exists - <- outputForeignStubs_help stub_h stub_h_output_w - ("#include \n" ++ cplusplus_hdr) cplusplus_ftr + stub_h_file_exists <- case ml_hs_file_ospath location of + Nothing -> pure False + Just _ -> do + -- The header path is computed from the module source path, which + -- does not exist when loading interface core bindings for Template + -- Haskell for non-home modules (e.g. when compiling in separate + -- invocations of oneshot mode). + -- The header is only generated for foreign exports. + -- Since those aren't supported for TH with bytecode, we can skip + -- this here for now. + let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location + createDirectoryIfMissing True (takeDirectory stub_h) + outputForeignStubs_help stub_h stub_h_output_w + ("#include \n" ++ cplusplus_hdr) cplusplus_ftr putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export stubs" FormatC stub_c_output_d ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,6 +49,7 @@ module GHC.Driver.Main , HscBackendAction (..), HscRecompStatus (..) , initModDetails , initWholeCoreBindings + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -105,6 +106,7 @@ module GHC.Driver.Main , showModuleIndex , hscAddSptEntries , writeInterfaceOnlyMode + , loadByteCode ) where import GHC.Prelude @@ -274,7 +276,8 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.List ( nub, isPrefixOf, partition ) +import Data.Functor ((<&>)) +import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad import Data.IORef @@ -970,19 +973,23 @@ loadByteCode iface mod_sum = do (mi_foreign iface) return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) _ -> return $ outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- +add_iface_to_hpt :: ModIface -> ModDetails -> HscEnv -> HscEnv +add_iface_to_hpt iface details = + hscUpdateHPT $ \ hpt -> + addToHpt hpt (moduleName (mi_module iface)) + (HomeModInfo iface details emptyHomeModInfoLinkable) -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] initModDetails :: HscEnv -> ModIface -> IO ModDetails initModDetails hsc_env iface = fixIO $ \details' -> do - let act hpt = addToHpt hpt (moduleName $ mi_module iface) - (HomeModInfo iface details' emptyHomeModInfoLinkable) - let !hsc_env' = hscUpdateHPT act hsc_env + let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful @@ -1010,8 +1017,52 @@ compile_for_interpreter hsc_env use = adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Return an 'IO' that hydrates Core bindings and compiles them to bytecode if +-- the interface contains any, using the supplied type env for typechecking. +-- +-- Unlike 'initWholeCoreBindings', this does not use lazy IO. +-- Instead, the 'IO' is only evaluated (in @get_link_deps@) when it is clear +-- that it will be used immediately (because we're linking TH with +-- @-fprefer-byte-code@ in oneshot mode), and the result is cached in +-- 'LoaderState'. +-- +-- 'initWholeCoreBindings' needs the laziness because it is used to populate +-- 'HomeModInfo', which is done preemptively, in anticipation of downstream +-- modules using the bytecode for TH in make mode, which might never happen. +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + compile <$> iface_core_bindings iface location + where + compile decls = do + (bcos, fos) <- compileWholeCoreBindings hsc_env type_env decls + linkable $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time (mi_module iface) parts + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects. +-- them with a lazy IO thunk that compiles them to bytecode and foreign objects, +-- using the supplied environment for type checking. -- -- The laziness is necessary because this value is stored purely in a -- 'HomeModLinkable' in the home package table, rather than some dedicated @@ -1025,29 +1076,71 @@ compile_for_interpreter hsc_env use = -- -- This is sound because generateByteCode just depends on things already loaded -- in the interface file. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env mod_iface details (Linkable utc_time this_mod uls) = +initWholeCoreBindings :: + HscEnv -> + ModIface -> + ModDetails -> + Linkable -> + IO Linkable +initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = Linkable utc_time this_mod <$> mapM go uls where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef (md_types details) - let act hpt = addToHpt hpt (moduleName $ mi_module mod_iface) - (HomeModInfo mod_iface details emptyHomeModInfoLinkable) - kv = knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - hsc_env' = hscUpdateHPT act hsc_env { hsc_type_env_vars = kv } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") hsc_env' $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons (md_types details)) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + go = \case + CoreBindings wcb -> do + ~(bco, fos) <- unsafeInterleaveIO $ + compileWholeCoreBindings hsc_env' type_env wcb + pure (LazyBCOs bco fos) + l -> pure l + + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written or were unavailable due to boot import +-- cycles, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +compileWholeCoreBindings :: + HscEnv -> + TypeEnv -> + WholeCoreBindings -> + IO (CompiledByteCode, [FilePath]) +compileWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + gen_bytecode core_binds stubs foreign_files + where + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + generateByteCode hsc_env cgi_guts wcb_mod_location + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -0,0 +1,15 @@ +module GHC.Driver.Main where + +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) + +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1310,8 +1310,10 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- MP: This is a bit janky, because before you add the entries you have to extend the HPT with the module -- you just compiled. Another option, would be delay adding anything until after upsweep has finished, but I -- am unsure if this is sound (wrt running TH splices for example). - -- This function only does anything if the linkable produced is a BCO, which only happens with the - -- bytecode backend, no need to guard against the backend type additionally. + -- This function only does anything if the linkable produced is a BCO, which + -- used to only happen with the bytecode backend, but with + -- @-fprefer-byte-code@, @HomeModInfo@ has bytecode even when generating + -- object code, see #25230. addSptEntries (hscUpdateHPT (\hpt -> addToHpt hpt (ms_mod_name summary) hmi) hsc_env) (homeModInfoByteCode hmi) @@ -3007,7 +3009,7 @@ which can be checked easily using ghc-debug. a reference to the entire HscEnv, if we are not careful the HscEnv will contain the HomePackageTable at the time the interface was loaded and it will never be released. - Where? dontLeakTheHPT in GHC.Iface.Load + Where? dontLeakTheHUG in GHC.Iface.Load 2. No KnotVars are live at the end of upsweep (#20491) Why? KnotVars contains an old stale reference to the TypeEnv for modules ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -511,6 +512,7 @@ loadInterface doc_str mod from ; new_eps_rules <- tcIfaceRules ignore_prags (mi_rules iface) ; new_eps_anns <- tcIfaceAnnotations (mi_anns iface) ; new_eps_complete_matches <- tcIfaceCompleteMatches (mi_complete_matches iface) + ; purged_hsc_env <- getTopEnv ; let final_iface = iface & set_mi_decls (panic "No mi_decls in PIT") @@ -518,13 +520,26 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + | Just action <- loadIfaceByteCode purged_hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have 'extra_decls' + -- so 'get_link_deps' knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,6 +551,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -569,7 +585,7 @@ loadInterface doc_str mod from {- Note [Loading your own hi-boot file] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Generally speaking, when compiling module M, we should not -load M.hi boot into the EPS. After all, we are very shortly +load M.hi-boot into the EPS. After all, we are very shortly going to have full information about M. Moreover, see Note [Do not update EPS with your own hi-boot] in GHC.Iface.Recomp. @@ -698,7 +714,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -845,7 +861,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -875,7 +891,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -900,7 +916,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -905,11 +905,11 @@ tcTopIfaceBindings :: IORef TypeEnv -> [IfaceBindingX IfaceMaybeRhs IfaceTopBndr -> IfL [CoreBind] tcTopIfaceBindings ty_var ver_decls = do - int <- mapM tcTopBinders ver_decls + int <- mapM tcTopBinders ver_decls let all_ids :: [Id] = concatMap toList int liftIO $ modifyIORef ty_var (flip extendTypeEnvList (map AnId all_ids)) - extendIfaceIdEnv all_ids $ mapM (tc_iface_bindings) int + extendIfaceIdEnv all_ids $ mapM tc_iface_bindings int tcTopBinders :: IfaceBindingX a IfaceTopBndrInfo -> IfL (IfaceBindingX a Id) tcTopBinders = traverse mk_top_id ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -60,16 +60,16 @@ import System.Directory data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -275,21 +275,21 @@ get_link_deps opts pls maybe_normal_osuf span mods = do case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod = do { - -- ...and then find the linkable for it - mb_lnk <- findObjectLinkableMaybe mod loc ; - case mb_lnk of { - Nothing -> no_obj mod ; - Just lnk -> adjust_linkable lnk - }} + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,18 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -68,6 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -139,6 +142,12 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), + eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules eps_fam_inst_env :: !PackageFamInstEnv,-- ^ The total 'FamInstEnv' accumulated ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -28,12 +28,12 @@ import System.FilePath (takeExtension) {- Note [Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A interface file can optionally contain the definitions of all core bindings, this is enabled by the flag `-fwrite-if-simplified-core`. This provides everything needed in addition to the normal ModIface and ModDetails -to restart compilation after typechecking to generate bytecode. The `fi_bindings` field +to restart compilation after typechecking to generate bytecode. The `wcb_bindings` field is stored in the normal interface file and the other fields populated whilst loading the interface file. @@ -62,8 +62,55 @@ after whatever simplification the user requested has been performed. So the simp of the interface file agree with the optimisation level as reported by the interface file. +The lifecycle differs beyond laziness depending on the provenance of a module. +In all cases, the main consumer for interface bytecode is 'get_link_deps', which +traverses a splice's or GHCi expression's dependencies and collects the needed +build artifacts, which can be objects or bytecode, depending on the build +settings. + +1. In make mode, all eligible modules are part of the dependency graph. + Their interfaces are loaded unconditionally and in dependency order by the + compilation manager, and each module's bytecode is prepared before its + dependents are compiled, in one of two ways: + + - If the interface file for a module is missing or out of sync with its + source, it is recompiled and bytecode is generated directly and + immediately, not involving 'WholeCoreBindings' (in 'runHscBackendPhase'). + + - If the interface file is up to date, no compilation is performed, and a + lazy thunk generating bytecode from interface Core bindings is created in + 'compileOne'', which will only be compiled if a downstream module contains + a splice that depends on it, as described above. + + In both cases, the bytecode 'Linkable' is stored in a 'HomeModLinkable' in + the Home Unit Graph, lazy or not. + +2. In oneshot mode, which compiles individual modules without a shared home unit + graph, a previously compiled module is not reprocessed as described for make + mode above. + When 'get_link_deps' encounters a dependency on a local module, it requests + its bytecode from the External Package State, who loads the interface + on-demand. + + Since the EPS stores interfaces for all package dependencies in addition to + local modules in oneshot mode, it has a substantial memory footprint. + We try to curtail that by extracting important data into specialized fields + in the EPS, and retaining only a few fields of 'ModIface' by overwriting the + others with bottom values. + + In order to avoid keeping around all of the interface's components needed for + compiling bytecode, we instead store an IO action in 'eps_iface_bytecode'. + When 'get_link_deps' evaluates this action, the result is not retained in the + EPS, but stored in 'LoaderState', where it may eventually get evicted to free + up the memory. + This IO action retains the dehydrated Core bindings from the interface in its + closure. + Like the bytecode 'Linkable' stored in 'LoaderState', this is preferable to + storing the intermediate representation as rehydrated Core bindings, since + the latter have a significantly greater memory footprint. + Note [Size of Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ How much overhead does `-fwrite-if-simplified-core` add to a typical interface file? As an experiment I compiled the `Cabal` library and `ghc` library (Aug 22) with ===================================== testsuite/tests/bytecode/T25090/A.hs ===================================== @@ -0,0 +1,7 @@ +{-# language TemplateHaskell #-} +module Main where + +import D + +main :: IO () +main = putStrLn (show ($splc :: Int)) ===================================== testsuite/tests/bytecode/T25090/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} C (C) + +data B = B C ===================================== testsuite/tests/bytecode/T25090/C.hs ===================================== @@ -0,0 +1,8 @@ +module C where + +import B + +data C = C Int + +b :: B +b = B (C 2024) ===================================== testsuite/tests/bytecode/T25090/C.hs-boot ===================================== @@ -0,0 +1,3 @@ +module C where + +data C ===================================== testsuite/tests/bytecode/T25090/D.hs ===================================== @@ -0,0 +1,12 @@ +module D where + +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +import B +import C + +splc :: ExpQ +splc = + lift @_ @Int num + where + B (C num) = b ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -0,0 +1,21 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +# Verify that the object files aren't linked by clobbering them. +T25090a: + $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot + $(TEST_HC) -c -fbyte-code-and-object-code B.hs + $(TEST_HC) -c -fbyte-code-and-object-code C.hs + echo 'corrupt' > B.o + echo 'corrupt' > C.o + echo 'corrupt' > C.o-boot + $(TEST_HC) -c -fbyte-code-and-object-code D.hs + echo 'corrupt' > D.o + $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A.o -o exe + ./exe + +T25090b: + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 + ./exe ===================================== testsuite/tests/bytecode/T25090/T25090-debug.stderr ===================================== @@ -0,0 +1,6 @@ +WARNING: + loadInterface + C + Call stack: + CallStack (from HasCallStack): + warnPprTrace, called at compiler/GHC/Iface/Load.hs:: in :GHC.Iface.Load ===================================== testsuite/tests/bytecode/T25090/T25090.stdout ===================================== @@ -0,0 +1 @@ +2024 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -0,0 +1,18 @@ +# This test compiles the boot file separately from its source file, which causes +# a debug assertion warning. +# Since this appears to be intentional according to the Note [Loading your own hi-boot file], +# the warning is added to the expected stderr for debugged builds. +def test_T25090(name): + assert_warn_spec = {'stderr': 'T25090-debug.stderr'} + extra_specs = assert_warn_spec if name == 'T25090a' and compiler_debugged() else {} + return test(name, + [extra_files(['A.hs', 'B.hs', 'C.hs-boot', 'C.hs', 'D.hs']), + req_th, + js_skip, + use_specs(dict(stdout = 'T25090.stdout', **extra_specs)), + ], + makefile_test, + []) + +test_T25090('T25090a') +test_T25090('T25090b') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6b43020bcb54fa3d4d8dc892c0f22618af170cc8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6b43020bcb54fa3d4d8dc892c0f22618af170cc8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 18:35:43 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Wed, 25 Sep 2024 14:35:43 -0400 Subject: [Git][ghc/ghc][wip/T24634-oneshot-bytecode] Link bytecode from interface-stored core bindings in oneshot mode Message-ID: <66f457ffb2a99_20b32198042c6739f@gitlab.mail> Torsten Schmits pushed to branch wip/T24634-oneshot-bytecode at Glasgow Haskell Compiler / GHC Commits: 7205003c by Cheng Shao at 2024-09-25T20:35:27+02:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 21 changed files: - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/A.hs - + testsuite/tests/bytecode/T25090/B.hs - + testsuite/tests/bytecode/T25090/C.hs - + testsuite/tests/bytecode/T25090/C.hs-boot - + testsuite/tests/bytecode/T25090/D.hs - + testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/T25090-debug.stderr - + testsuite/tests/bytecode/T25090/T25090.stdout - + testsuite/tests/bytecode/T25090/all.T Changes: ===================================== compiler/GHC/CoreToIface.hs ===================================== @@ -602,7 +602,7 @@ toIfaceTopBind b = in (top_bndr, rhs') -- The sharing behaviour is currently disabled due to #22807, and relies on - -- finished #220056 to be re-enabled. + -- finished #20056 to be re-enabled. disabledDueTo22807 = True already_has_unfolding b = not disabledDueTo22807 @@ -774,8 +774,8 @@ outside of the hs-boot loop. Note [Interface File with Core: Sharing RHSs] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -IMPORTANT: This optimisation is currently disabled due to #22027, it can be - re-enabled once #220056 is implemented. +IMPORTANT: This optimisation is currently disabled due to #22807, it can be + re-enabled once #22056 is implemented. In order to avoid duplicating definitions for bindings which already have unfoldings we do some minor headstands to avoid serialising the RHS of a definition if it has ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -260,7 +260,6 @@ outputForeignStubs Maybe FilePath) -- C file created outputForeignStubs logger tmpfs dflags unit_state mod location stubs = do - let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location stub_c <- newTempName logger tmpfs (tmpDir dflags) TFL_CurrentModule "c" case stubs of @@ -276,8 +275,6 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs stub_h_output_d = pprCode h_code stub_h_output_w = showSDoc dflags stub_h_output_d - createDirectoryIfMissing True (takeDirectory stub_h) - putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export header file" FormatC @@ -299,9 +296,23 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs | platformMisc_libFFI $ platformMisc dflags = "#include \"rts/ghc_ffi.h\"\n" | otherwise = "" - stub_h_file_exists - <- outputForeignStubs_help stub_h stub_h_output_w - ("#include \n" ++ cplusplus_hdr) cplusplus_ftr + -- The header path is computed from the module source path, which + -- does not exist when loading interface core bindings for Template + -- Haskell for non-home modules (e.g. when compiling in separate + -- invocations of oneshot mode). + -- Stub headers are only generated for foreign exports. + -- Since those aren't supported for TH with bytecode at the moment, + -- it doesn't make much of a difference. + -- In any case, if a stub dir was specified explicitly by the user, it + -- would be used nonetheless. + stub_h_file_exists <- + case mkStubPaths (initFinderOpts dflags) (moduleName mod) location of + Nothing -> pure False + Just path -> do + let stub_h = unsafeDecodeUtf path + createDirectoryIfMissing True (takeDirectory stub_h) + outputForeignStubs_help stub_h stub_h_output_w + ("#include \n" ++ cplusplus_hdr) cplusplus_ftr putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export stubs" FormatC stub_c_output_d ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,6 +49,7 @@ module GHC.Driver.Main , HscBackendAction (..), HscRecompStatus (..) , initModDetails , initWholeCoreBindings + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -105,6 +106,7 @@ module GHC.Driver.Main , showModuleIndex , hscAddSptEntries , writeInterfaceOnlyMode + , loadByteCode ) where import GHC.Prelude @@ -274,7 +276,8 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.List ( nub, isPrefixOf, partition ) +import Data.Functor ((<&>)) +import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad import Data.IORef @@ -970,19 +973,23 @@ loadByteCode iface mod_sum = do (mi_foreign iface) return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) _ -> return $ outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- +add_iface_to_hpt :: ModIface -> ModDetails -> HscEnv -> HscEnv +add_iface_to_hpt iface details = + hscUpdateHPT $ \ hpt -> + addToHpt hpt (moduleName (mi_module iface)) + (HomeModInfo iface details emptyHomeModInfoLinkable) -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] initModDetails :: HscEnv -> ModIface -> IO ModDetails initModDetails hsc_env iface = fixIO $ \details' -> do - let act hpt = addToHpt hpt (moduleName $ mi_module iface) - (HomeModInfo iface details' emptyHomeModInfoLinkable) - let !hsc_env' = hscUpdateHPT act hsc_env + let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful @@ -1010,8 +1017,52 @@ compile_for_interpreter hsc_env use = adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Return an 'IO' that hydrates Core bindings and compiles them to bytecode if +-- the interface contains any, using the supplied type env for typechecking. +-- +-- Unlike 'initWholeCoreBindings', this does not use lazy IO. +-- Instead, the 'IO' is only evaluated (in @get_link_deps@) when it is clear +-- that it will be used immediately (because we're linking TH with +-- @-fprefer-byte-code@ in oneshot mode), and the result is cached in +-- 'LoaderState'. +-- +-- 'initWholeCoreBindings' needs the laziness because it is used to populate +-- 'HomeModInfo', which is done preemptively, in anticipation of downstream +-- modules using the bytecode for TH in make mode, which might never happen. +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + compile <$> iface_core_bindings iface location + where + compile decls = do + (bcos, fos) <- compileWholeCoreBindings hsc_env type_env decls + linkable $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time (mi_module iface) parts + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects. +-- them with a lazy IO thunk that compiles them to bytecode and foreign objects, +-- using the supplied environment for type checking. -- -- The laziness is necessary because this value is stored purely in a -- 'HomeModLinkable' in the home package table, rather than some dedicated @@ -1025,29 +1076,71 @@ compile_for_interpreter hsc_env use = -- -- This is sound because generateByteCode just depends on things already loaded -- in the interface file. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env mod_iface details (Linkable utc_time this_mod uls) = +initWholeCoreBindings :: + HscEnv -> + ModIface -> + ModDetails -> + Linkable -> + IO Linkable +initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = Linkable utc_time this_mod <$> mapM go uls where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef (md_types details) - let act hpt = addToHpt hpt (moduleName $ mi_module mod_iface) - (HomeModInfo mod_iface details emptyHomeModInfoLinkable) - kv = knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - hsc_env' = hscUpdateHPT act hsc_env { hsc_type_env_vars = kv } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") hsc_env' $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons (md_types details)) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + go = \case + CoreBindings wcb -> do + ~(bco, fos) <- unsafeInterleaveIO $ + compileWholeCoreBindings hsc_env' type_env wcb + pure (LazyBCOs bco fos) + l -> pure l + + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written or were unavailable due to boot import +-- cycles, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +compileWholeCoreBindings :: + HscEnv -> + TypeEnv -> + WholeCoreBindings -> + IO (CompiledByteCode, [FilePath]) +compileWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + gen_bytecode core_binds stubs foreign_files + where + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + generateByteCode hsc_env cgi_guts wcb_mod_location + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -0,0 +1,15 @@ +module GHC.Driver.Main where + +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) + +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1310,8 +1310,10 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- MP: This is a bit janky, because before you add the entries you have to extend the HPT with the module -- you just compiled. Another option, would be delay adding anything until after upsweep has finished, but I -- am unsure if this is sound (wrt running TH splices for example). - -- This function only does anything if the linkable produced is a BCO, which only happens with the - -- bytecode backend, no need to guard against the backend type additionally. + -- This function only does anything if the linkable produced is a BCO, which + -- used to only happen with the bytecode backend, but with + -- @-fprefer-byte-code@, @HomeModInfo@ has bytecode even when generating + -- object code, see #25230. addSptEntries (hscUpdateHPT (\hpt -> addToHpt hpt (ms_mod_name summary) hmi) hsc_env) (homeModInfoByteCode hmi) @@ -3007,7 +3009,7 @@ which can be checked easily using ghc-debug. a reference to the entire HscEnv, if we are not careful the HscEnv will contain the HomePackageTable at the time the interface was loaded and it will never be released. - Where? dontLeakTheHPT in GHC.Iface.Load + Where? dontLeakTheHUG in GHC.Iface.Load 2. No KnotVars are live at the end of upsweep (#20491) Why? KnotVars contains an old stale reference to the TypeEnv for modules ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -511,6 +512,7 @@ loadInterface doc_str mod from ; new_eps_rules <- tcIfaceRules ignore_prags (mi_rules iface) ; new_eps_anns <- tcIfaceAnnotations (mi_anns iface) ; new_eps_complete_matches <- tcIfaceCompleteMatches (mi_complete_matches iface) + ; purged_hsc_env <- getTopEnv ; let final_iface = iface & set_mi_decls (panic "No mi_decls in PIT") @@ -518,13 +520,26 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + | Just action <- loadIfaceByteCode purged_hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have 'extra_decls' + -- so 'get_link_deps' knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,6 +551,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -569,7 +585,7 @@ loadInterface doc_str mod from {- Note [Loading your own hi-boot file] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Generally speaking, when compiling module M, we should not -load M.hi boot into the EPS. After all, we are very shortly +load M.hi-boot into the EPS. After all, we are very shortly going to have full information about M. Moreover, see Note [Do not update EPS with your own hi-boot] in GHC.Iface.Recomp. @@ -698,7 +714,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -845,7 +861,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -875,7 +891,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -900,7 +916,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -905,11 +905,11 @@ tcTopIfaceBindings :: IORef TypeEnv -> [IfaceBindingX IfaceMaybeRhs IfaceTopBndr -> IfL [CoreBind] tcTopIfaceBindings ty_var ver_decls = do - int <- mapM tcTopBinders ver_decls + int <- mapM tcTopBinders ver_decls let all_ids :: [Id] = concatMap toList int liftIO $ modifyIORef ty_var (flip extendTypeEnvList (map AnId all_ids)) - extendIfaceIdEnv all_ids $ mapM (tc_iface_bindings) int + extendIfaceIdEnv all_ids $ mapM tc_iface_bindings int tcTopBinders :: IfaceBindingX a IfaceTopBndrInfo -> IfL (IfaceBindingX a Id) tcTopBinders = traverse mk_top_id ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -60,16 +60,16 @@ import System.Directory data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -275,21 +275,21 @@ get_link_deps opts pls maybe_normal_osuf span mods = do case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod = do { - -- ...and then find the linkable for it - mb_lnk <- findObjectLinkableMaybe mod loc ; - case mb_lnk of { - Nothing -> no_obj mod ; - Just lnk -> adjust_linkable lnk - }} + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,18 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -68,6 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -139,6 +142,12 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), + eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules eps_fam_inst_env :: !PackageFamInstEnv,-- ^ The total 'FamInstEnv' accumulated ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -38,7 +38,6 @@ import GHC.Platform.Ways import GHC.Builtin.Names ( gHC_PRIM ) -import GHC.Data.Maybe ( expectJust ) import GHC.Data.OsPath import GHC.Unit.Env @@ -60,6 +59,7 @@ import GHC.Types.PkgQual import GHC.Fingerprint import Data.IORef import System.Directory.OsPath +import Control.Applicative ((<|>)) import Control.Monad import Data.Time import qualified Data.Map as M @@ -711,27 +711,27 @@ mkHiePath fopts basename mod_basename = hie_basename <.> hiesuf -- We don't have to store these in ModLocations, because they can be derived -- from other available information, and they're only rarely needed. +-- | Compute the file name of a header file for foreign stubs, using either the +-- directory explicitly specified in the command line option @-stubdir@, or the +-- directory of the module's source file. +-- +-- When compiling bytecode from interface Core bindings, @ModLocation@ does not +-- contain a source file path, so the header isn't written. +-- This doesn't have an impact, since we cannot support headers importing +-- Haskell symbols defined in bytecode for TH whatsoever at the moment. mkStubPaths :: FinderOpts -> ModuleName -> ModLocation - -> OsPath - -mkStubPaths fopts mod location - = let - stubdir = finder_stubDir fopts - - mod_basename = unsafeEncodeUtf $ moduleNameSlashes mod - src_basename = OsPath.dropExtension $ expectJust "mkStubPaths" - (ml_hs_file_ospath location) - - stub_basename0 - | Just dir <- stubdir = dir mod_basename - | otherwise = src_basename + -> Maybe OsPath +mkStubPaths fopts mod location = do + stub_basename <- in_stub_dir <|> src_basename + pure (stub_basename `mappend` os "_stub" <.> os "h") + where + in_stub_dir = ( mod_basename) <$> (finder_stubDir fopts) - stub_basename = stub_basename0 `mappend` os "_stub" - in - stub_basename <.> os "h" + mod_basename = unsafeEncodeUtf $ moduleNameSlashes mod + src_basename = OsPath.dropExtension <$> ml_hs_file_ospath location -- ----------------------------------------------------------------------------- -- findLinkable isn't related to the other stuff in here, ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -28,12 +28,12 @@ import System.FilePath (takeExtension) {- Note [Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A interface file can optionally contain the definitions of all core bindings, this is enabled by the flag `-fwrite-if-simplified-core`. This provides everything needed in addition to the normal ModIface and ModDetails -to restart compilation after typechecking to generate bytecode. The `fi_bindings` field +to restart compilation after typechecking to generate bytecode. The `wcb_bindings` field is stored in the normal interface file and the other fields populated whilst loading the interface file. @@ -62,8 +62,55 @@ after whatever simplification the user requested has been performed. So the simp of the interface file agree with the optimisation level as reported by the interface file. +The lifecycle differs beyond laziness depending on the provenance of a module. +In all cases, the main consumer for interface bytecode is 'get_link_deps', which +traverses a splice's or GHCi expression's dependencies and collects the needed +build artifacts, which can be objects or bytecode, depending on the build +settings. + +1. In make mode, all eligible modules are part of the dependency graph. + Their interfaces are loaded unconditionally and in dependency order by the + compilation manager, and each module's bytecode is prepared before its + dependents are compiled, in one of two ways: + + - If the interface file for a module is missing or out of sync with its + source, it is recompiled and bytecode is generated directly and + immediately, not involving 'WholeCoreBindings' (in 'runHscBackendPhase'). + + - If the interface file is up to date, no compilation is performed, and a + lazy thunk generating bytecode from interface Core bindings is created in + 'compileOne'', which will only be compiled if a downstream module contains + a splice that depends on it, as described above. + + In both cases, the bytecode 'Linkable' is stored in a 'HomeModLinkable' in + the Home Unit Graph, lazy or not. + +2. In oneshot mode, which compiles individual modules without a shared home unit + graph, a previously compiled module is not reprocessed as described for make + mode above. + When 'get_link_deps' encounters a dependency on a local module, it requests + its bytecode from the External Package State, who loads the interface + on-demand. + + Since the EPS stores interfaces for all package dependencies in addition to + local modules in oneshot mode, it has a substantial memory footprint. + We try to curtail that by extracting important data into specialized fields + in the EPS, and retaining only a few fields of 'ModIface' by overwriting the + others with bottom values. + + In order to avoid keeping around all of the interface's components needed for + compiling bytecode, we instead store an IO action in 'eps_iface_bytecode'. + When 'get_link_deps' evaluates this action, the result is not retained in the + EPS, but stored in 'LoaderState', where it may eventually get evicted to free + up the memory. + This IO action retains the dehydrated Core bindings from the interface in its + closure. + Like the bytecode 'Linkable' stored in 'LoaderState', this is preferable to + storing the intermediate representation as rehydrated Core bindings, since + the latter have a significantly greater memory footprint. + Note [Size of Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ How much overhead does `-fwrite-if-simplified-core` add to a typical interface file? As an experiment I compiled the `Cabal` library and `ghc` library (Aug 22) with ===================================== testsuite/tests/bytecode/T25090/A.hs ===================================== @@ -0,0 +1,7 @@ +{-# language TemplateHaskell #-} +module Main where + +import D + +main :: IO () +main = putStrLn (show ($splc :: Int)) ===================================== testsuite/tests/bytecode/T25090/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} C (C) + +data B = B C ===================================== testsuite/tests/bytecode/T25090/C.hs ===================================== @@ -0,0 +1,8 @@ +module C where + +import B + +data C = C Int + +b :: B +b = B (C 2024) ===================================== testsuite/tests/bytecode/T25090/C.hs-boot ===================================== @@ -0,0 +1,3 @@ +module C where + +data C ===================================== testsuite/tests/bytecode/T25090/D.hs ===================================== @@ -0,0 +1,12 @@ +module D where + +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +import B +import C + +splc :: ExpQ +splc = + lift @_ @Int num + where + B (C num) = b ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -0,0 +1,21 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +# Verify that the object files aren't linked by clobbering them. +T25090a: + $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot + $(TEST_HC) -c -fbyte-code-and-object-code B.hs + $(TEST_HC) -c -fbyte-code-and-object-code C.hs + echo 'corrupt' > B.o + echo 'corrupt' > C.o + echo 'corrupt' > C.o-boot + $(TEST_HC) -c -fbyte-code-and-object-code D.hs + echo 'corrupt' > D.o + $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A.o -o exe + ./exe + +T25090b: + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 + ./exe ===================================== testsuite/tests/bytecode/T25090/T25090-debug.stderr ===================================== @@ -0,0 +1,6 @@ +WARNING: + loadInterface + C + Call stack: + CallStack (from HasCallStack): + warnPprTrace, called at compiler/GHC/Iface/Load.hs:: in :GHC.Iface.Load ===================================== testsuite/tests/bytecode/T25090/T25090.stdout ===================================== @@ -0,0 +1 @@ +2024 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -0,0 +1,18 @@ +# This test compiles the boot file separately from its source file, which causes +# a debug assertion warning. +# Since this appears to be intentional according to the Note [Loading your own hi-boot file], +# the warning is added to the expected stderr for debugged builds. +def test_T25090(name): + assert_warn_spec = {'stderr': 'T25090-debug.stderr'} + extra_specs = assert_warn_spec if name == 'T25090a' and compiler_debugged() else {} + return test(name, + [extra_files(['A.hs', 'B.hs', 'C.hs-boot', 'C.hs', 'D.hs']), + req_th, + js_skip, + use_specs(dict(stdout = 'T25090.stdout', **extra_specs)), + ], + makefile_test, + []) + +test_T25090('T25090a') +test_T25090('T25090b') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7205003c2ba50c2c5af106ed94773730ea327d53 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7205003c2ba50c2c5af106ed94773730ea327d53 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 19:48:22 2024 From: gitlab at gitlab.haskell.org (Jade (@Jade)) Date: Wed, 25 Sep 2024 15:48:22 -0400 Subject: [Git][ghc/ghc][wip/structured-ghci-errors] Add structured errors to GHCi (#23338) Message-ID: <66f469067cb6e_373ff527f29094444@gitlab.mail> Jade pushed to branch wip/structured-ghci-errors at Glasgow Haskell Compiler / GHC Commits: be47d68a by Jade at 2024-09-25T21:47:59+02:00 Add structured errors to GHCi (#23338) This patch creates the 'GhciCommandErrorMessage' data type which implents the 'Diagnostic' class and also provides error code for these error conditions. - - - - - 30 changed files: - compiler/GHC/Driver/Errors.hs - compiler/GHC/Driver/Errors/Ppr.hs - compiler/GHC/Driver/Errors/Types.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/Driver/Ppr.hs - compiler/GHC/HsToCore/Errors/Ppr.hs - compiler/GHC/HsToCore/Errors/Types.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Types/Error.hs - compiler/GHC/Types/Error/Codes.hs - docs/users_guide/9.12.1-notes.rst - ghc/GHCi/UI.hs - ghc/GHCi/UI/Exception.hs - ghc/GHCi/UI/Info.hs - ghc/GHCi/UI/Monad.hs - + ghc/GHCi/UI/Print.hs - ghc/ghc-bin.cabal.in - testsuite/tests/driver/multipleHomeUnits/multiGHCi.stderr - testsuite/tests/ghc-e/should_fail/T18441fail0.stderr - testsuite/tests/ghc-e/should_fail/T18441fail1.stderr - testsuite/tests/ghc-e/should_fail/T18441fail10.stderr - testsuite/tests/ghc-e/should_fail/T18441fail11.stderr - testsuite/tests/ghc-e/should_fail/T18441fail14.stderr - testsuite/tests/ghc-e/should_fail/T18441fail15.stderr - testsuite/tests/ghc-e/should_fail/T18441fail2.stderr - testsuite/tests/ghc-e/should_fail/T18441fail6.stderr - testsuite/tests/ghc-e/should_fail/T18441fail8.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/be47d68a3c0b0a81da08b57541a34256b2056857 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/be47d68a3c0b0a81da08b57541a34256b2056857 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:08:12 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:08:12 -0400 Subject: [Git][ghc/ghc][master] 9 commits: ci: Run abi-test on test-abi label Message-ID: <66f47bbcfdb0_373ff57d77ec1126f3@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/RV64/CodeGen.hs - compiler/GHC/CmmToAsm/RV64/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c0df5aa9999fcb7a7890973c319b56731b77aa7d...7b37afc9f3e79559055488998ee73187886a0e00 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c0df5aa9999fcb7a7890973c319b56731b77aa7d...7b37afc9f3e79559055488998ee73187886a0e00 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:08:51 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:08:51 -0400 Subject: [Git][ghc/ghc][master] Add Given injectivity for built-in type families Message-ID: <66f47be35ec3b_373ff5afc42c118226@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - 10 changed files: - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a63ee33a116599c88e703d9793e025e17c59f7f9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a63ee33a116599c88e703d9793e025e17c59f7f9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:09:39 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:09:39 -0400 Subject: [Git][ghc/ghc][master] Resolve ambiguous method-bound type variables in vanilla defaults and GND Message-ID: <66f47c135964f_373ff5b0bda0124841@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 15 changed files: - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/TyCl/Instance.hs - docs/users_guide/exts/default_signatures.rst - testsuite/tests/deriving/should_compile/T14578.stderr - testsuite/tests/deriving/should_compile/T14682.stderr - testsuite/tests/deriving/should_compile/T17240.stderr - + testsuite/tests/deriving/should_compile/T25148c.hs - testsuite/tests/deriving/should_compile/all.T - testsuite/tests/deriving/should_compile/deriving-inferred-ty-arg.stderr - testsuite/tests/generics/T10604/T10604_deriving.stderr - + testsuite/tests/generics/T14266.hs - + testsuite/tests/generics/T25148a.hs - + testsuite/tests/generics/T25148b.hs - testsuite/tests/generics/all.T - testsuite/tests/unsatisfiable/UnsatDefault.stderr Changes: ===================================== compiler/GHC/Tc/Deriv/Generate.hs ===================================== @@ -1699,12 +1699,16 @@ coercing from. So from, say, instance C a => C a (T x) where op @c = coerce @(a -> [] -> c -> Int) @(a -> [T x] -> c -> Int) - op + (op @c) In addition to the type applications, we also use a type abstraction to bring -the method-bound variable `c` into scope over the two type applications. -See Note [GND and QuantifiedConstraints] for more information on why this -is important. +the method-bound variable `c` into scope. We do this for two reasons: + +* We need to bring `c` into scope over the two type applications to `coerce`. + See Note [GND and QuantifiedConstraints] for more information on why this + is important. +* We need to bring `c` into scope over the type application to `op`. See + Note [GND and ambiguity] for more information on why this is important. (In the surface syntax, only specified type variables can be used in type abstractions. Since a method signature could contain both specified and @@ -1737,6 +1741,27 @@ However, to allow VTA with polytypes we must switch on -XImpredicativeTypes locally in GHC.Tc.Deriv.genInst. See #8503 for more discussion. +The following Notes describe further nuances of GeneralizedNewtypeDeriving: + +----- +-- In GHC.Tc.Deriv +----- + +* Note [Newtype deriving] +* Note [Newtype representation] +* Note [Recursive newtypes] +* Note [Determining whether newtype-deriving is appropriate] +* Note [GND and associated type families] +* Note [Bindings for Generalised Newtype Deriving] + +----- +-- In GHC.Tc.Deriv.Generate +----- + +* Note [Newtype-deriving trickiness] +* Note [GND and QuantifiedConstraints] +* Note [GND and ambiguity] + Note [Inferred invisible patterns] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Consider the following: @@ -1915,27 +1940,68 @@ coerce with a polytype, and we can only do that with VTA or QuickLook. Note [GND and ambiguity] ~~~~~~~~~~~~~~~~~~~~~~~~ We make an effort to make the code generated through GND be robust w.r.t. -ambiguous type variables. As one example, consider the following example -(from #15637): +ambiguous type variables. Here are a couple of examples to illustrate this: + +* In this example (from #15637), the class-bound type variable `a` is ambiguous + in the type of `f`: - class C a where f :: String - instance C () where f = "foo" - newtype T = T () deriving C + class C a where + f :: String -- f :: forall a. C a => String + instance C () + where f = "foo" + newtype T = T () + deriving C -A naïve attempt and generating a C T instance would be: + A naïve attempt and generating a C T instance would be: - instance C T where - f = coerce @String @String f + instance C T where + f = coerce @String @String f -This isn't going to typecheck, however, since GHC doesn't know what to -instantiate the type variable `a` with in the call to `f` in the method body. -(Note that `f :: forall a. String`!) To compensate for the possibility of -ambiguity here, we explicitly instantiate `a` like so: + This isn't going to typecheck, however, since GHC doesn't know what to + instantiate the type variable `a` with in the call to `f` in the method body. + (Note that `f :: forall a. String`!) To compensate for the possibility of + ambiguity here, we explicitly instantiate `a` like so: - instance C T where - f = coerce @String @String (f @()) + instance C T where + f = coerce @String @String (f @()) -All better now. + All better now. + +* In this example (adapted from #25148), the ambiguity arises from the `n` + type variable bound by the type signature for `fact1`: + + class Facts a where + fact1 :: forall n. Proxy a -> Dict (0 <= n) + newtype T a = MkT a + deriving newtype Facts + + When generating code for the derived `Facts` instance, we must use a type + abstraction to bring `n` into scope over the type applications to `coerce` + (see Note [Newtype-deriving instances] for more why this is needed). A first + attempt at generating the instance would be: + + instance Facts a => Facts (T a) where + fact1 @n = coerce @(Proxy a -> Dict (0 <= n)) + @(Proxy (T a) -> Dict (0 <= n)) + (fact1 @a) + + This still won't typecheck, however, as GHC doesn't know how to instantiate + `n` in the call to `fact1 @a`. To compensate for the possibility of ambiguity + here, we also visibly apply `n` in the call to `fact1` on the RHS: + + instance Facts a => Facts (T a) where + fact1 @n = coerce @(Proxy a -> Dict (0 <= n)) + @(Proxy (T a) -> Dict (0 <= n)) + (fact1 @a @n) -- Note the @n here! + + This takes advantage of the fact that we *already* need to bring `n` into + scope using a type abstraction, and so we are able to use it both for + instantiating the call to `coerce` and instantiating the call to `fact1`. + + Note that we use this same type abstractions-based approach for resolving + ambiguity in default methods, as described in Note [Default methods in + instances] (Wrinkle: Ambiguous types from vanilla method type signatures) in + GHC.Tc.TyCl.Instance. -} gen_Newtype_binds :: SrcSpan @@ -2002,11 +2068,16 @@ gen_Newtype_binds loc' cls inst_tvs inst_tys rhs_ty `nlHsAppType` to_tau `nlHsApp` meth_app - -- The class method, applied to all of the class instance types - -- (including the representation type) to avoid potential ambiguity. - -- See Note [GND and ambiguity] + -- The class method, applied to the following types to avoid potential + -- ambiguity: + -- + -- 1. All of the class instance types (including the representation type) + -- 2. All of `to_tvbs` + -- + -- See Note [GND and ambiguity]. meth_app = foldl' nlHsAppType (nlHsVar meth_RDR) $ - filterOutInferredTypes (classTyCon cls) underlying_inst_tys + filterOutInferredTypes (classTyCon cls) underlying_inst_tys ++ -- (1) + [mkTyVarTy tv | Bndr tv spec <- to_tvbs, spec /= InferredSpec] -- (2) -- Filter out any inferred arguments, since they can't be -- applied with visible type application. ===================================== compiler/GHC/Tc/TyCl/Instance.hs ===================================== @@ -1846,8 +1846,8 @@ tcMethods skol_info dfun_id clas tyvars dfun_ev_vars inst_tys meth_bind = mkVarBind meth_id $ mkLHsWrap lam_wrapper meth_rhs ; return (meth_id, meth_bind, Nothing) } - Just (dm_name, _) -> - do { (meth_bind, inline_prags) <- mkDefMethBind inst_loc dfun_id clas sel_id dm_name + Just (dm_name, dm_spec) -> + do { (meth_bind, inline_prags) <- mkDefMethBind inst_loc dfun_id clas sel_id dm_name dm_spec ; tcMethodBody skol_info clas tyvars dfun_ev_vars inst_tys dfun_ev_binds is_derived hs_sig_fn spec_inst_prags inline_prags @@ -2194,14 +2194,15 @@ mk_meth_spec_prags meth_id spec_inst_prags spec_prags_for_me mkDefMethBind :: SrcSpan -> DFunId -> Class -> Id -> Name + -> DefMethSpec Type -> TcM (LHsBind GhcRn, [LSig GhcRn]) -- The is a default method (vanailla or generic) defined in the class --- So make a binding op = $dmop @t1 @t2 --- where $dmop is the name of the default method in the class, --- and t1,t2 are the instance types. --- See Note [Default methods in instances] for why we use --- visible type application here -mkDefMethBind loc dfun_id clas sel_id dm_name +-- So make a binding op @m1 @m2 @m3 = $dmop @i1 @i2 @m1 @m2 @m3 +-- where $dmop is the name of the default method in the class; +-- i1 and t2 are the instance types; and m1, m2, and m3 are the type variables +-- from the method's type signature. See Note [Default methods in instances] for +-- why we use visible type application here. +mkDefMethBind loc dfun_id clas sel_id dm_name dm_spec = do { logger <- getLogger ; dm_id <- tcLookupId dm_name ; let inline_prag = idInlinePragma dm_id @@ -2212,29 +2213,64 @@ mkDefMethBind loc dfun_id clas sel_id dm_name -- Copy the inline pragma (if any) from the default method -- to this version. Note [INLINE and default methods] - fn = noLocA (idName sel_id) - visible_inst_tys = [ ty | (tcb, ty) <- tyConBinders (classTyCon clas) `zip` inst_tys - , tyConBinderForAllTyFlag tcb /= Inferred ] - rhs = foldl' mk_vta (nlHsVar dm_name) visible_inst_tys - bind = L (noAnnSrcSpan loc) - $ mkTopFunBind (Generated OtherExpansion SkipPmc) fn - [mkSimpleMatch (mkPrefixFunRhs fn) (noLocA []) rhs] - ; liftIO (putDumpFileMaybe logger Opt_D_dump_deriv "Filling in method body" FormatHaskell (vcat [ppr clas <+> ppr inst_tys, - nest 2 (ppr sel_id <+> equals <+> ppr rhs)])) + nest 2 (ppr bind)])) ; return (bind, inline_prags) } where (_, _, _, inst_tys) = tcSplitDFunTy (idType dfun_id) + (_, _, sel_tau) = tcSplitMethodTy (idType sel_id) + (sel_tvbs, _) = tcSplitForAllInvisTVBinders sel_tau + + -- Compute the instance types to use in the visible type application. See + -- Note [Default methods in instances]. + visible_inst_tys = + [ ty | (tcb, ty) <- tyConBinders (classTyCon clas) `zip` inst_tys + , tyConBinderForAllTyFlag tcb /= Inferred ] + + visible_sel_tvbs = + case dm_spec of + -- When dealing with a vanilla default method, compute the type + -- variables from the method's type signature. That way, we can bind + -- them with TypeAbstractions (visible_sel_pats) and use them in the + -- visible type application (visible_sel_tys). See Note [Default methods + -- in instances] (Wrinkle: Ambiguous types from vanilla method type + -- signatures). + VanillaDM -> filter (\tvb -> binderFlag tvb /= InferredSpec) sel_tvbs + -- If we are dealing with a generic default method, on the other hand, + -- don't bother doing any of this. See Note [Default methods + -- in instances] (Wrinkle: Ambiguous types from generic default method + -- type signatures). + GenericDM {} -> [] + visible_sel_pats = map mk_ty_pat visible_sel_tvbs + visible_sel_tys = map (mkTyVarTy . binderVar) visible_sel_tvbs + + fn = noLocA (idName sel_id) + rhs = foldl' mk_vta (nlHsVar dm_name) $ + visible_inst_tys ++ visible_sel_tys + bind = L (noAnnSrcSpan loc) + $ mkTopFunBind (Generated OtherExpansion SkipPmc) fn + [mkSimpleMatch (mkPrefixFunRhs fn) (noLocA visible_sel_pats) rhs] + + mk_ty_pat :: VarBndr TyVar Specificity -> LPat GhcRn + mk_ty_pat (Bndr tv spec) = + noLocA $ + InvisPat spec $ + HsTP (HsTPRn [] [tyVarName tv] []) $ + nlHsTyVar NotPromoted $ + tyVarName tv mk_vta :: LHsExpr GhcRn -> Type -> LHsExpr GhcRn mk_vta fun ty = noLocA (HsAppType noExtField fun - (mkEmptyWildCardBndrs $ nlHsParTy $ noLocA $ XHsType ty)) + (mkEmptyWildCardBndrs $ type_to_hs_type ty)) -- NB: use visible type application -- See Note [Default methods in instances] + type_to_hs_type :: Type -> LHsType GhcRn + type_to_hs_type = parenthesizeHsType appPrec . noLocA . XHsType + ---------------------- derivBindCtxt :: Id -> Class -> [Type ] -> SDoc derivBindCtxt sel_id clas tys @@ -2277,8 +2313,8 @@ From the class decl we get $dmfoo :: forall v x. Baz v x => x -> x $dmfoo y = -Notice that the type is ambiguous. So we use Visible Type Application -to disambiguate: +Notice that the type of `v` is ambiguous. So we use Visible Type Application +(VTA) to disambiguate: $dBazIntInt = MkBaz fooIntInt fooIntInt = $dmfoo @Int @Int @@ -2291,6 +2327,151 @@ Historical note: before we had VTA we had to generate post-type-checked code, which took a lot more code, and didn't work for generic default methods. +----- +-- Wrinkle: Ambiguous types from vanilla method type signatures +----- + +In the Bar example above, the ambiguity arises from `v`, a type variable +arising from the class header. It is also possible for the ambiguity to arise +from a type variable bound by the method's type signature itself (see #14266 +and #25148). For example: + + class A t where + f :: forall x m. Monoid x => t m -> m + f = + + instance A [] + +The class declaration gives rise to the following default function: + + $dmf :: forall t. A t => forall x m. Monoid x => t m -> m + $dmf = + +And the instance declaration gives rise to generated code that looks roughly +like this: + + instance A [] where + f = $dmf @[] ... + +In this example, it is not enough to use VTA to specify the type of `t`, since +the type of `x` (bound by `f`'s type signature) is also ambiguous. We need to +generate code that looks more like this: + + instance A [] where + f = $dmf @[] @x @m + +But where should `x` and `m` be bound? It's tempting to use ScopedTypeVariables +and InstanceSigs to accomplish this: + + instance A [] where + f :: forall x m. Monoid x => [m] -> m + f = $dmf @[] @x @m + +GHC will reject this code, however, as the type signature for `f` will fail the +subtype check for InstanceSigs: + + • Could not deduce (Monoid x0) + from the context: Monoid x + bound by the type signature for: + f :: forall x m. Monoid x => [m] -> m + The type variable ‘x0’ is ambiguous + • When checking that instance signature for ‘f’ + is more general than its signature in the class + Instance sig: forall x m. Monoid x => [m] -> m + Class sig: forall x m. Monoid x => [m] -> m + In the instance declaration for ‘A []’ + +See #17898. To avoid this problem, we instead bind `x` and `m` using +TypeAbstractions: + + instance A [] where + f @x @m = $dmf @[] @x @m + +This resolves the ambiguity and avoids the need for a subtype check. (We also +use a similar trick for resolving ambiguity in GeneralizedNewtypeDeriving: see +also Note [GND and ambiguity] in GHC.Tc.Deriv.Generate.) + +----- +-- Wrinkle: Ambiguous types from generic default method type signatures +----- + +Note that the approach described above (in Wrinkle: Ambiguous types from +vanilla method type signatures) will only work for vanilla default methods and +/not/ for generic default methods (i.e., methods using DefaultSignatures). This +is because for vanilla default methods, the type of the generated $dm* function +will always quantify the same type variables as the method's original type +signature, in the same order and with the same specificities. For example, the +type of the $dmf function will be: + + $dmf :: forall t. A t => forall x m. Monoid x => t m -> m + +As such, it is guaranteed that the type variables from the method's original +type signature will line up exactly with the type variables from the $dm* +function (after instantiating all of the class variables): + + instance A [] where + f @x @m = $dmf @[] @x @m + +We cannot guarantee this property for generic default methods, however. As +such, we must be more conservative and generate code without instantiating any +of the type variables bound by the method's type signature (only the type +variables bound by the class header): + + instance A [] where + f = $dmf @[] + +There are a number of reasons why we cannot reliably instantiate the type +variables bound by a generic default method's type signature: + +* Default methods can quantify type variables in a different order, e.g., + + class A t where + f :: forall x m. Monoid x => t m -> m + default f :: forall m x. Monoid x => t m -> m + f = + + Note that the default signature quantifies the type variables in the opposite + order from the method's original type signature. As such, the type of $dmf + will be: + + $dmf :: forall t. A t => forall m x. Monoid x => t m -> m + + Therefore, `f @x @m = $dmf @[] @x @m` would be incorrect. Nor would it be + straightforward to infer what the correct order of type variables should be. + +* Default methods can quantify a different number of type variables, e.g., + + class A t where + f :: forall x m. Monoid x => t m -> m + default f :: forall p q r m. C a t p q r => t m -> m + f = + + This gives rise to: + + $dmf :: forall t. A t => forall p q r m. C a t p q r => t m -> m + + And thus generating `f @x @m = $dmf @[] @x @m` would be incorrect, for + similar reasons as in the example above. + +* Default methods can use different type variable specificities, e.g., + + class A t where + f :: forall x m. Monoid x => t m -> m + default f :: forall {x} m. Monoid x => t m -> m + f = + + This gives rise to: + + $dmf :: forall t. A t => forall {x} m. Monoid x => t m -> m + + Therefore, generating `f @x @m = $dmf @[] @x @m` would be incorrect because + the `x` in the type of $dmf is inferred, so it is not eligible for visible + type application. + +As such, we do not bother trying to resolve the ambiguity of any method-bound +type variables when dealing with generic defaults. This means that GHC won't be +able to typecheck the default method examples above, but so be it. + Note [INLINE and default methods] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Default methods need special case. They are supposed to behave rather like ===================================== docs/users_guide/exts/default_signatures.rst ===================================== @@ -189,3 +189,15 @@ default type signatures. This works, but at the expense of changing ``p``'s behavior with respect to :ref:`visible-type-application`. + +- The default signature cannot use any ambiguous type variables. For example, + GHC will not permit instances of the following class: :: + + class A t where + f :: forall x m. Monoid x => t m -> m + default :: forall x m. Monoid x => t m -> m + f = ... + + Note that ``x`` is ambiguous, as none of the argument or result types + determine what type ``x`` should be when calling ``f``. GHC will reject + instances of ``A``, even if :extension:`AllowAmbiguousTypes` is enabled. ===================================== testsuite/tests/deriving/should_compile/T14578.stderr ===================================== @@ -28,7 +28,7 @@ Derived class instances: -> T14578.App (Data.Functor.Compose.Compose f g) a) @(b -> T14578.Wat f g a -> T14578.Wat f g a) (GHC.Internal.Base.stimes - @(T14578.App (Data.Functor.Compose.Compose f g) a)) + @(T14578.App (Data.Functor.Compose.Compose f g) a) @b) instance GHC.Internal.Base.Functor f => GHC.Internal.Base.Functor (T14578.App f) where @@ -36,38 +36,38 @@ Derived class instances: = GHC.Prim.coerce @((a -> b) -> f a -> f b) @((a -> b) -> T14578.App f a -> T14578.App f b) - (GHC.Internal.Base.fmap @f) + (GHC.Internal.Base.fmap @f @a @b) (GHC.Internal.Base.<$) @a @b = GHC.Prim.coerce @(a -> f b -> f a) @(a -> T14578.App f b -> T14578.App f a) - ((GHC.Internal.Base.<$) @f) + ((GHC.Internal.Base.<$) @f @a @b) instance GHC.Internal.Base.Applicative f => GHC.Internal.Base.Applicative (T14578.App f) where GHC.Internal.Base.pure @a = GHC.Prim.coerce - @(a -> f a) @(a -> T14578.App f a) (GHC.Internal.Base.pure @f) + @(a -> f a) @(a -> T14578.App f a) (GHC.Internal.Base.pure @f @a) (GHC.Internal.Base.<*>) @a @b = GHC.Prim.coerce @(f (a -> b) -> f a -> f b) @(T14578.App f (a -> b) -> T14578.App f a -> T14578.App f b) - ((GHC.Internal.Base.<*>) @f) + ((GHC.Internal.Base.<*>) @f @a @b) GHC.Internal.Base.liftA2 @a @b @c = GHC.Prim.coerce @((a -> b -> c) -> f a -> f b -> f c) @((a -> b -> c) -> T14578.App f a -> T14578.App f b -> T14578.App f c) - (GHC.Internal.Base.liftA2 @f) + (GHC.Internal.Base.liftA2 @f @a @b @c) (GHC.Internal.Base.*>) @a @b = GHC.Prim.coerce @(f a -> f b -> f b) @(T14578.App f a -> T14578.App f b -> T14578.App f b) - ((GHC.Internal.Base.*>) @f) + ((GHC.Internal.Base.*>) @f @a @b) (GHC.Internal.Base.<*) @a @b = GHC.Prim.coerce @(f a -> f b -> f a) @(T14578.App f a -> T14578.App f b -> T14578.App f a) - ((GHC.Internal.Base.<*) @f) + ((GHC.Internal.Base.<*) @f @a @b) Derived type family instances: @@ -76,14 +76,14 @@ Derived type family instances: ==================== Filling in method body ==================== GHC.Internal.Base.Semigroup [T14578.App f a] - GHC.Internal.Base.sconcat = GHC.Internal.Base.$dmsconcat - @(T14578.App f a) + GHC.Internal.Base.sconcat + = GHC.Internal.Base.$dmsconcat @(T14578.App f a) ==================== Filling in method body ==================== GHC.Internal.Base.Semigroup [T14578.App f a] - GHC.Internal.Base.stimes = GHC.Internal.Base.$dmstimes - @(T14578.App f a) + GHC.Internal.Base.stimes @b + = GHC.Internal.Base.$dmstimes @(T14578.App f a) @b ===================================== testsuite/tests/deriving/should_compile/T14682.stderr ===================================== @@ -90,121 +90,121 @@ Derived type family instances: ==================== Filling in method body ==================== GHC.Internal.Show.Show [T14682.Foo] - GHC.Internal.Show.show = GHC.Internal.Show.$dmshow @(T14682.Foo) + GHC.Internal.Show.show = GHC.Internal.Show.$dmshow @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Show.Show [T14682.Foo] - GHC.Internal.Show.showList = GHC.Internal.Show.$dmshowList - @(T14682.Foo) + GHC.Internal.Show.showList + = GHC.Internal.Show.$dmshowList @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.dataCast1 = GHC.Internal.Data.Data.$dmdataCast1 - @(T14682.Foo) + GHC.Internal.Data.Data.dataCast1 @t @c + = GHC.Internal.Data.Data.$dmdataCast1 @T14682.Foo @t @c ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.dataCast2 = GHC.Internal.Data.Data.$dmdataCast2 - @(T14682.Foo) + GHC.Internal.Data.Data.dataCast2 @t @c + = GHC.Internal.Data.Data.$dmdataCast2 @T14682.Foo @t @c ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapT = GHC.Internal.Data.Data.$dmgmapT - @(T14682.Foo) + GHC.Internal.Data.Data.gmapT + = GHC.Internal.Data.Data.$dmgmapT @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapQl = GHC.Internal.Data.Data.$dmgmapQl - @(T14682.Foo) + GHC.Internal.Data.Data.gmapQl @r @r' + = GHC.Internal.Data.Data.$dmgmapQl @T14682.Foo @r @r' ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapQr = GHC.Internal.Data.Data.$dmgmapQr - @(T14682.Foo) + GHC.Internal.Data.Data.gmapQr @r @r' + = GHC.Internal.Data.Data.$dmgmapQr @T14682.Foo @r @r' ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapQ = GHC.Internal.Data.Data.$dmgmapQ - @(T14682.Foo) + GHC.Internal.Data.Data.gmapQ @u + = GHC.Internal.Data.Data.$dmgmapQ @T14682.Foo @u ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapQi = GHC.Internal.Data.Data.$dmgmapQi - @(T14682.Foo) + GHC.Internal.Data.Data.gmapQi @u + = GHC.Internal.Data.Data.$dmgmapQi @T14682.Foo @u ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapM = GHC.Internal.Data.Data.$dmgmapM - @(T14682.Foo) + GHC.Internal.Data.Data.gmapM @m + = GHC.Internal.Data.Data.$dmgmapM @T14682.Foo @m ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapMp = GHC.Internal.Data.Data.$dmgmapMp - @(T14682.Foo) + GHC.Internal.Data.Data.gmapMp @m + = GHC.Internal.Data.Data.$dmgmapMp @T14682.Foo @m ==================== Filling in method body ==================== GHC.Internal.Data.Data.Data [T14682.Foo] - GHC.Internal.Data.Data.gmapMo = GHC.Internal.Data.Data.$dmgmapMo - @(T14682.Foo) + GHC.Internal.Data.Data.gmapMo @m + = GHC.Internal.Data.Data.$dmgmapMo @T14682.Foo @m ==================== Filling in method body ==================== GHC.Classes.Eq [T14682.Foo] - GHC.Classes./= = GHC.Classes.$dm/= @(T14682.Foo) + (GHC.Classes./=) = GHC.Classes.$dm/= @T14682.Foo ==================== Filling in method body ==================== GHC.Classes.Ord [T14682.Foo] - GHC.Classes.max = GHC.Classes.$dmmax @(T14682.Foo) + GHC.Classes.max = GHC.Classes.$dmmax @T14682.Foo ==================== Filling in method body ==================== GHC.Classes.Ord [T14682.Foo] - GHC.Classes.min = GHC.Classes.$dmmin @(T14682.Foo) + GHC.Classes.min = GHC.Classes.$dmmin @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Ix.Ix [T14682.Foo] - GHC.Internal.Ix.index = GHC.Internal.Ix.$dmindex @(T14682.Foo) + GHC.Internal.Ix.index = GHC.Internal.Ix.$dmindex @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Ix.Ix [T14682.Foo] - GHC.Internal.Ix.rangeSize = GHC.Internal.Ix.$dmrangeSize - @(T14682.Foo) + GHC.Internal.Ix.rangeSize + = GHC.Internal.Ix.$dmrangeSize @T14682.Foo ==================== Filling in method body ==================== GHC.Internal.Ix.Ix [T14682.Foo] - GHC.Internal.Ix.unsafeRangeSize = GHC.Internal.Ix.$dmunsafeRangeSize - @(T14682.Foo) + GHC.Internal.Ix.unsafeRangeSize + = GHC.Internal.Ix.$dmunsafeRangeSize @T14682.Foo ===================================== testsuite/tests/deriving/should_compile/T17240.stderr ===================================== @@ -31,12 +31,12 @@ Derived type family instances: ==================== Filling in method body ==================== GHC.Classes.Eq [T17240.Nullary] - GHC.Classes./= = GHC.Classes.$dm/= @(T17240.Nullary) + (GHC.Classes./=) = GHC.Classes.$dm/= @T17240.Nullary ==================== Filling in method body ==================== GHC.Classes.Eq [T17240.T] - GHC.Classes./= = GHC.Classes.$dm/= @(T17240.T) + (GHC.Classes./=) = GHC.Classes.$dm/= @T17240.T ===================================== testsuite/tests/deriving/should_compile/T25148c.hs ===================================== @@ -0,0 +1,17 @@ +{-# LANGUAGE AllowAmbiguousTypes #-} +{-# LANGUAGE DataKinds #-} +{-# LANGUAGE DerivingStrategies #-} +module T25148c where + +import Data.Kind +import Data.Proxy +import GHC.TypeLits + +data Dict :: Constraint -> Type where + Dict :: c => Dict c + +class Facts a where + fact1 :: forall n. Proxy a -> Dict (0 <= n) + +newtype T a = MkT a + deriving newtype Facts ===================================== testsuite/tests/deriving/should_compile/all.T ===================================== @@ -150,4 +150,5 @@ test('T15798c', normal, compile, ['']) test('T24955a', normal, compile, ['']) test('T24955b', normal, compile, ['']) test('T24955c', normal, compile, ['']) +test('T25148c', normal, compile, ['']) test('deriving-inferred-ty-arg', normal, compile, ['-ddump-deriv -dsuppress-uniques']) ===================================== testsuite/tests/deriving/should_compile/deriving-inferred-ty-arg.stderr ===================================== @@ -8,7 +8,7 @@ Derived class instances: @(GHC.Internal.Data.Functor.Const.Const a b) @(GHC.Internal.Data.Functor.Const.Const (DerivingInferredTyArg.T a) b) - (DerivingInferredTyArg.m @a) + (DerivingInferredTyArg.m @a @b) instance DerivingInferredTyArg.C2 a => DerivingInferredTyArg.C2 (DerivingInferredTyArg.T a) where @@ -26,7 +26,7 @@ Derived class instances: @(GHC.Internal.Data.Functor.Const.Const a p) @(GHC.Internal.Data.Functor.Const.Const (DerivingInferredTyArg.T a) p) - (DerivingInferredTyArg.m3 @a) + (DerivingInferredTyArg.m3 @a @b) instance DerivingInferredTyArg.C4 a => DerivingInferredTyArg.C4 (DerivingInferredTyArg.T a) where @@ -35,7 +35,7 @@ Derived class instances: @(GHC.Internal.Data.Functor.Const.Const a p) @(GHC.Internal.Data.Functor.Const.Const (DerivingInferredTyArg.T a) p) - (DerivingInferredTyArg.m4 @a) + (DerivingInferredTyArg.m4 @a @b) instance DerivingInferredTyArg.C5 a => DerivingInferredTyArg.C5 (DerivingInferredTyArg.T a) where ===================================== testsuite/tests/generics/T10604/T10604_deriving.stderr ===================================== @@ -542,7 +542,7 @@ Derived type family instances: ==================== Filling in method body ==================== GHC.Internal.Base.Functor [T10604_deriving.Proxy @(*)] - GHC.Internal.Base.<$ = GHC.Internal.Base.$dm<$ - @(T10604_deriving.Proxy @(*)) + (GHC.Internal.Base.<$) @a @b + = GHC.Internal.Base.$dm<$ @(T10604_deriving.Proxy @(*)) @a @b ===================================== testsuite/tests/generics/T14266.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE AllowAmbiguousTypes #-} +module T14266 where + +class A t where + f :: forall x m. Monoid x => t m -> m + f = undefined +instance A [] ===================================== testsuite/tests/generics/T25148a.hs ===================================== @@ -0,0 +1,24 @@ +{-# LANGUAGE AllowAmbiguousTypes #-} +{-# LANGUAGE DataKinds #-} +module T25148a where + +import Data.Kind +import Data.Proxy +import GHC.TypeLits +import Unsafe.Coerce + +data Dict :: Constraint -> Type where + Dict :: c => Dict c + +fact0 :: forall n a. Proxy a -> Dict (0 <= n) +fact0 _ = unsafeCoerce (Dict :: Dict (0 <= 0)) + +class Facts a where + fact1 :: forall n. Proxy a -> Dict (0 <= n) + fact1 _ = unsafeCoerce (Dict :: Dict (0 <= 0)) +instance Facts () + +class Facts' a where + fact' :: forall n. Proxy n -> Proxy a -> Dict (0 <= n) + fact' _ _ = unsafeCoerce (Dict :: Dict (0 <= 0)) +instance Facts' () ===================================== testsuite/tests/generics/T25148b.hs ===================================== @@ -0,0 +1,42 @@ +{-# LANGUAGE DataKinds #-} +{-# LANGUAGE DeriveAnyClass #-} +{-# LANGUAGE DerivingStrategies #-} +{-# LANGUAGE TypeFamilies #-} +module T25148b where + +import Data.Functor.Const +import Data.Proxy + +class Monoid a => C a where + m :: forall {k} (b :: k). Const a b + m = Const mempty + +class Monoid a => C2 a where + m2 :: forall {k} {b :: k}. Const a b + m2 = Const mempty + +class Monoid a => C3 a where + m3 :: forall {k} (b :: k) {p :: Proxy b}. Const a p + m3 = Const mempty + +data VisProxy k (a :: k) = VisProxy + +class Monoid a => C4 a where + m4 :: forall {k} (b :: k) {p :: VisProxy k b}. Const a p + m4 = Const mempty + +type family Any :: k + +class Monoid a => C5 a where + m5 :: Proxy Any -> a + m5 _ = mempty + +----- + +data T = MkT + deriving anyclass (C, C2, C3, C4, C5) + +instance Semigroup T where + _ <> _ = MkT +instance Monoid T where + mempty = MkT ===================================== testsuite/tests/generics/all.T ===================================== @@ -45,5 +45,8 @@ test('T10361b', normal, compile, ['']) test('T11358', normal, compile_and_run, ['']) test('T12220', normal, compile, ['']) test('T15012', [extra_files(['T15012.hs', 'T15012a.hs'])], makefile_test, []) +test('T14266', normal, compile, ['']) test('T19819', normal, compile_and_run, ['']) test('T21185', normal, compile, ['']) +test('T25148a', normal, compile, ['']) +test('T25148b', normal, compile, ['']) ===================================== testsuite/tests/unsatisfiable/UnsatDefault.stderr ===================================== @@ -1,6 +1,6 @@ - UnsatDefault.hs:14:10: error: [GHC-22250] • Please define the method manually. You can try... - • In the expression: UnsatDefault.$dmmethod @(Int) - In an equation for ‘method’: method = UnsatDefault.$dmmethod @(Int) + • In the expression: UnsatDefault.$dmmethod @Int + In an equation for ‘method’: method = UnsatDefault.$dmmethod @Int In the instance declaration for ‘C Int’ + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/dd8ef342bb5f050309545e89cf735d6f31c17398 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/dd8ef342bb5f050309545e89cf735d6f31c17398 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:10:57 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:10:57 -0400 Subject: [Git][ghc/ghc][master] Document primitive string literals and desugaring of string literals Message-ID: <66f47c61b10f_373ff57b87ac1351d6@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - 2 changed files: - docs/users_guide/exts/magic_hash.rst - docs/users_guide/exts/primitives.rst Changes: ===================================== docs/users_guide/exts/magic_hash.rst ===================================== @@ -30,7 +30,7 @@ The :extension:`MagicHash` also enables some new forms of literals (see - ``'x'#`` has type ``Char#`` -- ``"foo"#`` has type ``Addr#`` +- ``"foo"#`` has type ``Addr#`` (see :ref:`primitive-string-literals`). - ``3#`` has type ``Int#``. In general, any Haskell integer lexeme followed by a ``#`` is an ``Int#`` literal, e.g. ``-0x3A#`` as well as ===================================== docs/users_guide/exts/primitives.rst ===================================== @@ -478,3 +478,46 @@ It is worth noting that :extension:`UnliftedDatatypes` is *not* required to give the data families themselves return kinds involving ``TYPE``, such as the ``G`` example above. The extension is only required for ``data instance`` declarations, such as ``FInt`` and ``GBool`` above. + +.. _primitive-string-literals: + +Primitive string literals +------------------------- + +Primitive string literals, also called unboxed string literals or ``Addr#`` literals, +are string literals with a postfix hash sign ``#``. +They are enabled by the :extension:`MagicHash` extension and have type ``Addr#``. + +A primitive string literal represents a sequence of bytes. +You can embed a NUL byte like ``"foo\0bar"#``, which corresponds to the bytes +``[102,111,111,0,98,97,114] :: [Word8]`` (with an implicit NUL at the end). +Further, you can only embed a character whose code is less than 256; ``"\xFF"#`` is valid +and represents ``[255] :: [Word8]``, but ``"\x3B1"#`` is invalid. +In other words, it is encoded in Latin-1. + +Like C strings, primitive string literals are implicitly terminated by a NUL byte. + +Primitive string literals are immutable and attempting to write to their address results in undefined behavior. + +.. _desugaring-of-string-literals: + +Desugaring of string literals +----------------------------- + +An ASCII string literal without a NUL byte is desugared into a call to +``GHC.CString.unpackCString#`` with the content as a primitive string literal. +For example, ``"some string"`` is translated into ``GHC.CString.unpackCString# "some string"#``. + +If the string contains NUL bytes or non-ASCII characters (characters with code points >= 128), +the string is encoded in Modified UTF-8 and then passed to ``GHC.CString.unpackCStringUtf8#``. +Modified UTF-8 differs from standard UTF-8 in two ways: + +- The NUL byte is encoded as ``"\xC0\x80"#`` (i.e. an over-long encoding of U+0000). +- Surrogate code points are allowed and encoded like other code points. + +Examples: + +- ``"Hello\0world!"`` is desugared to ``GHC.CString.unpackCStringUtf8# "Hello\xC0\x80world!"#``. +- ``"café"`` is desugared to ``GHC.CString.unpackCStringUtf8# "caf\xC3\xA9"#``. +- ``"\x732B\x1F431"`` is desugared to ``GHC.CString.unpackCStringUtf8# "\xE7\x8C\xAB\xF0\x9F\x90\xB1"#``. +- ``"\xDFFF\xD800"`` is desugared to ``GHC.CString.unpackCStringUtf8# "\xED\xBF\xBF\xED\xA0\x80"#``. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0a4da5d2194be981a582a805b0b01ad42ba1c76c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0a4da5d2194be981a582a805b0b01ad42ba1c76c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:11:39 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:11:39 -0400 Subject: [Git][ghc/ghc][master] rts: Fix segfault when using non-moving GC with profiling Message-ID: <66f47c8b8e59e_373ff5b0bda0142242@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - 5 changed files: - rts/sm/GC.c - testsuite/config/ghc - testsuite/driver/testlib.py - + testsuite/tests/rts/T25232.hs - testsuite/tests/rts/all.T Changes: ===================================== rts/sm/GC.c ===================================== @@ -854,6 +854,18 @@ GarbageCollect (struct GcConfig config, } } +#if defined(PROFILING) + // resetStaticObjectForProfiling() must be called before + // zeroing below. + // It must also be called before nonMovingCollect() as that + // swaps out the value of static_flag used as a sentinel + // in gct->scavenged_static_objects + // Not doing this lead to #25232 and #23958 + + // ToDo: fix the gct->scavenged_static_objects below + resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); +#endif + // Mark and sweep the oldest generation. // N.B. This can only happen after we've moved // oldest_gen->scavenged_large_objects back to oldest_gen->large_objects. @@ -946,14 +958,6 @@ GarbageCollect (struct GcConfig config, checkUnload(); } -#if defined(PROFILING) - // resetStaticObjectForProfiling() must be called before - // zeroing below. - - // ToDo: fix the gct->scavenged_static_objects below - resetStaticObjectForProfiling(&g_retainerTraverseState, gct->scavenged_static_objects); -#endif - // Start any pending finalizers. Must be after // updateStableTables() and stableUnlock() (see #4221). RELEASE_SM_LOCK; ===================================== testsuite/config/ghc ===================================== @@ -125,6 +125,8 @@ config.way_flags = { 'ext-interp' : ['-fexternal-interpreter'], 'nonmoving' : [], 'nonmoving_thr': ['-threaded'], + 'nonmoving_prof': ['-prof'], + 'nonmoving_thr_prof': ['-prof', '-threaded'], 'nonmoving_thr_sanity': ['-threaded', '-debug'], 'nonmoving_thr_ghc': ['+RTS', '-xn', '-N2', '-RTS', '-threaded'], 'compacting_gc': [], ===================================== testsuite/driver/testlib.py ===================================== @@ -848,15 +848,12 @@ def _collect_stats(name: TestName, opts, metrics, deviation: Optional[int], if config.compiler_debugged and is_compiler_stats_test: opts.skip = True - # If there are any residency testing metrics then turn on RESIDENCY_OPTS and - # omit nonmoving GC ways, which don't support profiling. + # If there are any residency testing metrics then turn on RESIDENCY_OPTS if residency_testing_metrics() & metrics: if is_compiler_stats_test: _extra_hc_opts(name, opts, RESIDENCY_OPTS) else: _extra_run_opts(name, opts, RESIDENCY_OPTS) - # The nonmoving collector does not support -G1 - _omit_ways(name, opts, [WayName(name) for name in ['nonmoving', 'nonmoving_thr', 'nonmoving_thr_ghc']]) # How to read the result of the performance test def read_stats_file(way, metric_name): @@ -1026,6 +1023,9 @@ def have_dynamic_prof( ) -> bool: def have_profiling( ) -> bool: return config.have_profiling +def have_threaded( ) -> bool: + return config.ghc_with_threaded_rts + def in_tree_compiler( ) -> bool: return config.in_tree_compiler ===================================== testsuite/tests/rts/T25232.hs ===================================== @@ -0,0 +1,14 @@ +module Main where + +import Control.Monad +import Control.Monad.Trans.State.Strict +import Data.IORef + +main :: IO () +main = do + ref <- newIORef () + replicateM_ 1000 $ withIORef ref $ runStateT (pure ()) + +withIORef :: IORef a -> (a -> IO (b, a)) -> IO b +withIORef ref f = + readIORef ref >>= f >>= \(b, a) -> writeIORef ref a >> pure b ===================================== testsuite/tests/rts/all.T ===================================== @@ -616,3 +616,5 @@ test('IOManager', [js_skip, when(arch('wasm32'), skip), when(opsys('mingw32'), s compile_and_run, ['']) test('T24142', [req_target_smp], compile_and_run, ['-threaded -with-rtsopts "-N2"']) + +test('T25232', [unless(have_profiling(), skip), only_ways(['normal','nonmoving','nonmoving_prof','nonmoving_thr_prof']), extra_ways(['nonmoving', 'nonmoving_prof'] + (['nonmoving_thr_prof'] if have_threaded() else []))], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ad0731ada00982ae66e0b5bb88ca7b97ece1313f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ad0731ada00982ae66e0b5bb88ca7b97ece1313f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:12:11 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:12:11 -0400 Subject: [Git][ghc/ghc][master] Fix interaction between fork and kqueue (#24672) Message-ID: <66f47cabd28c7_373ff5f1c8541457e@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 4 changed files: - libraries/ghc-internal/src/GHC/Internal/Event/KQueue.hsc - + testsuite/tests/lib/base/T24672.hs - + testsuite/tests/lib/base/T24672.stdout - testsuite/tests/lib/base/all.T Changes: ===================================== libraries/ghc-internal/src/GHC/Internal/Event/KQueue.hsc ===================================== @@ -44,8 +44,8 @@ import GHC.Internal.Num (Num(..)) import GHC.Internal.Real (quotRem, fromIntegral) import GHC.Internal.Show (Show(show)) import GHC.Internal.Event.Internal (Timeout(..)) -import GHC.Internal.System.Posix.Internals (c_close) -import GHC.Internal.System.Posix.Types (Fd(..)) +import GHC.Internal.System.Posix.Internals (c_close,c_getpid) +import GHC.Internal.System.Posix.Types (Fd(..), CPid) import qualified GHC.Internal.Event.Array as A #if defined(netbsd_HOST_OS) @@ -73,19 +73,26 @@ available = True data KQueue = KQueue { kqueueFd :: {-# UNPACK #-} !KQueueFd , kqueueEvents :: {-# UNPACK #-} !(A.Array Event) + , kqueuePid :: {-# UNPACK #-} !CPid -- ^ pid, used to detect forks } new :: IO E.Backend new = do kqfd <- kqueue events <- A.new 64 - let !be = E.backend poll modifyFd modifyFdOnce delete (KQueue kqfd events) + pid <- c_getpid + let !be = E.backend poll modifyFd modifyFdOnce delete (KQueue kqfd events pid) return be delete :: KQueue -> IO () delete kq = do - _ <- c_close . fromKQueueFd . kqueueFd $ kq - return () + -- detect forks: the queue isn't inherited by a child process created with + -- fork. Hence we mustn't try to close the old fd or we might close a random + -- one (e.g. the one used by timerfd, cf #24672). + pid <- c_getpid + when (pid == kqueuePid kq) $ do + _ <- c_close . fromKQueueFd . kqueueFd $ kq + return () modifyFd :: KQueue -> Fd -> E.Event -> E.Event -> IO Bool modifyFd kq fd oevt nevt = do ===================================== testsuite/tests/lib/base/T24672.hs ===================================== @@ -0,0 +1,13 @@ +module Main where + +import Control.Concurrent +import qualified System.Posix.Process as SPP +import System.IO + +main = do + print "before SPP.forkProcess" + hFlush stdout + threadDelay (2*1000*1000) + SPP.forkProcess $ pure () + threadDelay (2*1000*1000) + print "after SPP.forkProcess" ===================================== testsuite/tests/lib/base/T24672.stdout ===================================== @@ -0,0 +1,2 @@ +"before SPP.forkProcess" +"after SPP.forkProcess" ===================================== testsuite/tests/lib/base/all.T ===================================== @@ -2,6 +2,7 @@ test('DataTypeOrd', normal, compile_and_run, ['']) test('T16586', normal, compile_and_run, ['-O2']) # Event-manager not supported on Windows test('T16916', [when(opsys('mingw32'), skip), js_broken(22261), fragile(16966), req_ghc_with_threaded_rts], compile_and_run, ['-O2 -threaded -with-rtsopts="-I0" -rtsopts']) +test('T24672', [when(opsys('mingw32'), skip), js_broken(22261),req_process], compile_and_run, ['']) test('T17310', normal, compile, ['']) test('T19691', normal, compile, ['']) test('executablePath', [extra_run_opts(config.os), js_broken(22261), when(arch('wasm32'), fragile(23248)), omit_ghci], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e7a26d7a6faf1ea534e036c5085a0a027dbb6f5f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e7a26d7a6faf1ea534e036c5085a0a027dbb6f5f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:12:57 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:12:57 -0400 Subject: [Git][ghc/ghc][master] Improve GHC.Tc.Solver.defaultEquality Message-ID: <66f47cd9de8c5_373ff5f6199015223f@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 3 changed files: - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Utils/Unify.hs Changes: ===================================== compiler/GHC/Tc/Solver.hs ===================================== @@ -29,30 +29,13 @@ module GHC.Tc.Solver( import GHC.Prelude -import GHC.Data.Bag -import GHC.Core.Class -import GHC.Core -import GHC.Core.DataCon -import GHC.Core.Make -import GHC.Core.Coercion( mkNomReflCo ) -import GHC.Driver.DynFlags -import GHC.Data.FastString -import GHC.Data.List.SetOps -import GHC.Types.Name -import GHC.Types.DefaultEnv ( ClassDefaults (..), defaultList ) -import GHC.Types.Unique.Set -import GHC.Types.Id -import GHC.Utils.Outputable -import GHC.Builtin.Utils -import GHC.Builtin.Names import GHC.Tc.Errors import GHC.Tc.Errors.Types import GHC.Tc.Types.Evidence import GHC.Tc.Solver.Solve ( solveSimpleGivens, solveSimpleWanteds ) import GHC.Tc.Solver.Dict ( makeSuperClasses, solveCallStack ) import GHC.Tc.Solver.Rewrite ( rewriteType ) -import GHC.Tc.Utils.Unify ( buildTvImplication, touchabilityAndShapeTest - , simpleUnifyCheck, UnifyCheckCaller(..) ) +import GHC.Tc.Utils.Unify import GHC.Tc.Utils.TcMType as TcM import GHC.Tc.Utils.Monad as TcM import GHC.Tc.Zonk.TcType as TcM @@ -61,17 +44,30 @@ import GHC.Tc.Solver.Monad as TcS import GHC.Tc.Types.Constraint import GHC.Tc.Types.CtLoc( mkGivenLoc ) import GHC.Tc.Instance.FunDeps -import GHC.Core.Predicate import GHC.Tc.Types.Origin import GHC.Tc.Utils.TcType + +import GHC.Core.Class +import GHC.Core.Reduction( Reduction, reductionCoercion ) +import GHC.Core +import GHC.Core.DataCon +import GHC.Core.Make +import GHC.Core.Coercion( mkNomReflCo, isReflCo ) +import GHC.Core.Unify ( tcMatchTyKis ) +import GHC.Core.Predicate import GHC.Core.Type import GHC.Core.Ppr import GHC.Core.TyCon ( TyCon, TyConBinder, isTypeFamilyTyCon ) + +import GHC.Types.Name +import GHC.Types.DefaultEnv ( ClassDefaults (..), defaultList ) +import GHC.Types.Unique.Set +import GHC.Types.Id + +import GHC.Builtin.Utils +import GHC.Builtin.Names import GHC.Builtin.Types -import GHC.Core.Unify ( tcMatchTyKis ) -import GHC.Unit.Module ( getModule ) -import GHC.Utils.Misc -import GHC.Utils.Panic + import GHC.Types.TyThing ( MonadThings(lookupId) ) import GHC.Types.Var import GHC.Types.Var.Env @@ -79,6 +75,18 @@ import GHC.Types.Var.Set import GHC.Types.Basic import GHC.Types.Id.Make ( unboxedUnitExpr ) import GHC.Types.Error + +import GHC.Driver.DynFlags +import GHC.Unit.Module ( getModule ) + +import GHC.Utils.Misc +import GHC.Utils.Panic +import GHC.Utils.Outputable + +import GHC.Data.FastString +import GHC.Data.List.SetOps +import GHC.Data.Bag + import qualified GHC.LanguageExtensions as LangExt import Control.Monad @@ -515,7 +523,7 @@ tryDefaulting wc ; wc2 <- tryConstraintDefaulting wc1 ; wc3 <- tryTypeClassDefaulting wc2 ; wc4 <- tryUnsatisfiableGivens wc3 - ; traceTcS "tryDefaulting:after" (ppr wc) + ; traceTcS "tryDefaulting:after" (ppr wc4) ; return wc4 } solveAgainIf :: Bool -> WantedConstraints -> TcS WantedConstraints @@ -771,35 +779,66 @@ defaultEquality :: CtDefaultingStrategy -- See Note [Defaulting equalities] defaultEquality ct | EqPred NomEq ty1 ty2 <- classifyPredType (ctPred ct) - , Just tv1 <- getTyVar_maybe ty1 = do { -- Remember: `ct` may not be zonked; -- see (DE3) in Note [Defaulting equalities] - z_ty1 <- TcS.zonkTcTyVar tv1 - ; z_ty2 <- TcS.zonkTcType ty2 - ; case getTyVar_maybe z_ty1 of - Just z_tv1 | defaultable z_tv1 z_ty2 - -> do { default_tv z_tv1 z_ty2 - ; return True } - _ -> return False } - | otherwise - = return False + z_ty1 <- TcS.zonkTcType ty1 + ; z_ty2 <- TcS.zonkTcType ty2 + + -- Now see if either LHS or RHS is a bare type variable + -- You might think the type variable will only be on the LHS + -- but with a type function we might get F t1 ~ alpha + ; case (getTyVar_maybe z_ty1, getTyVar_maybe z_ty2) of + (Just z_tv1, _) -> try_default_tv z_tv1 z_ty2 + (_, Just z_tv2) -> try_default_tv z_tv2 z_ty1 + _ -> return False } + | otherwise + = return False + where - defaultable tv1 ty2 - = -- Do the standard unification checks; - -- c.f. uUnfilledVar2 in GHC.Tc.Utils.Unify - -- EXCEPT drop the untouchability test - tyVarKind tv1 `tcEqType` typeKind ty2 - && touchabilityAndShapeTest topTcLevel tv1 ty2 - -- topTcLevel makes the untoucability test vacuous, - -- which is the Whole Point of `defaultEquality` - -- See (DE2) in Note [Defaulting equalities] - && simpleUnifyCheck UC_Defaulting tv1 ty2 - - default_tv tv1 ty2 - = do { unifyTyVar tv1 ty2 -- NB: unifyTyVar adds to the - -- TcS unification counter - ; setEvBindIfWanted (ctEvidence ct) EvCanonical $ - evCoercion (mkNomReflCo ty2) } + try_default_tv lhs_tv rhs_ty + | MetaTv { mtv_info = info, mtv_tclvl = lvl } <- tcTyVarDetails lhs_tv + , tyVarKind lhs_tv `tcEqType` typeKind rhs_ty + , checkTopShape info rhs_ty + -- Do not test for touchability of lhs_tv; that is the whole point! + -- See (DE2) in Note [Defaulting equalities] + = do { traceTcS "defaultEquality 1" (ppr lhs_tv $$ ppr rhs_ty) + + -- checkTyEqRhs: check that we can in fact unify lhs_tv := rhs_ty + -- See Note [Defaulting equalities] + -- LC_Promote: promote deeper unification variables (DE4) + -- LC_Promote True: ...including under type families (DE5) + ; let flags :: TyEqFlags () + flags = TEF { tef_foralls = False + , tef_fam_app = TEFA_Recurse + , tef_lhs = TyVarLHS lhs_tv + , tef_unifying = Unifying info lvl (LC_Promote True) + , tef_occurs = cteInsolubleOccurs } + ; res :: PuResult () Reduction <- wrapTcS (checkTyEqRhs flags rhs_ty) + + ; case res of + PuFail {} -> cant_default_tv "checkTyEqRhs" + PuOK _ redn -> assertPpr (isReflCo (reductionCoercion redn)) (ppr redn) $ + -- With TEFA_Recurse we never get any reductions + default_tv } + | otherwise + = cant_default_tv "fall through" + + where + cant_default_tv msg + = do { traceTcS ("defaultEquality fails: " ++ msg) $ + vcat [ ppr lhs_tv <+> char '~' <+> ppr rhs_ty + , ppr (tyVarKind lhs_tv) + , ppr (typeKind rhs_ty) ] + ; return False } + + -- All tests passed: do the unification + default_tv + = do { traceTcS "defaultEquality success:" (ppr rhs_ty) + ; unifyTyVar lhs_tv rhs_ty -- NB: unifyTyVar adds to the + -- TcS unification counter + ; setEvBindIfWanted (ctEvidence ct) EvCanonical $ + evCoercion (mkNomReflCo rhs_ty) + ; return True } combineStrategies :: CtDefaultingStrategy -> CtDefaultingStrategy -> CtDefaultingStrategy combineStrategies default1 default2 ct @@ -884,22 +923,20 @@ Wrinkles: given equality is at top level. (DE3) The contraint we are looking at may not be fully zonked; for example, - an earlier deafaulting might have affected it. So we zonk-on-the fly in + an earlier defaulting might have affected it. So we zonk-on-the fly in `defaultEquality`. -Note [Don't default in syntactic equalities] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -When there are unsolved syntactic equalities such as +(DE4) Promotion. Suppose we see alpha[2] := Maybe beta[4]. We want to promote + beta[4] to level 2 and unify alpha[2] := Maybe beta'[2]. This is done by + checkTyEqRhs. - rr[sk] ~S# alpha[conc] +(DE5) Promotion. Suppose we see alpha[2] := F beta[4], where F is a type + family. Then we still want to promote beta to beta'[2], and unify. This is + unusual: more commonly, we don't promote unification variables under a + type family. But here we want to. (This mattered in #25251.) -we should not default alpha, lest we obtain a poor error message such as - - Couldn't match kind `rr' with `LiftedRep' - -We would rather preserve the original syntactic equality to be -reported to the user, especially as the concrete metavariable alpha -might store an informative origin for the user. + Hence the Bool flag on LC_Promote, and its use in `tef_unifying` in + `defaultEquality`. Note [Must simplify after defaulting] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Tc/Solver/Monad.hs ===================================== @@ -2102,7 +2102,7 @@ checkTouchableTyVarEq -- with extra wanteds 'cts' -- If it returns (PuFail reason) we can't unify, and the reason explains why. checkTouchableTyVarEq ev lhs_tv rhs - | simpleUnifyCheck UC_Solver lhs_tv rhs + | simpleUnifyCheck UC_Solver lhs_tv rhs -- An (optional) short-cut = do { traceTcS "checkTouchableTyVarEq: simple-check wins" (ppr lhs_tv $$ ppr rhs) ; return (pure (mkReflRedn Nominal rhs)) } @@ -2135,12 +2135,13 @@ checkTouchableTyVarEq ev lhs_tv rhs flags = TEF { tef_foralls = False -- isRuntimeUnkSkol lhs_tv , tef_fam_app = mkTEFA_Break ev NomEq break_wanted - , tef_unifying = Unifying lhs_tv_info lhs_tv_lvl LC_Promote + , tef_unifying = Unifying lhs_tv_info lhs_tv_lvl (LC_Promote False) , tef_lhs = TyVarLHS lhs_tv , tef_occurs = cteInsolubleOccurs } arg_flags = famAppArgFlags flags + break_wanted :: FamAppBreaker Ct break_wanted fam_app -- Occurs check or skolem escape; so flatten = do { let fam_app_kind = typeKind fam_app ===================================== compiler/GHC/Tc/Utils/Unify.hs ===================================== @@ -25,7 +25,7 @@ module GHC.Tc.Utils.Unify ( -- Various unifications unifyType, unifyKind, unifyInvisibleType, unifyExpectedType, unifyExprType, unifyTypeAndEmit, promoteTcType, - swapOverTyVars, touchabilityAndShapeTest, lhsPriority, + swapOverTyVars, touchabilityAndShapeTest, checkTopShape, lhsPriority, UnifyEnv(..), updUEnvLoc, setUEnvRole, uType, @@ -3134,6 +3134,9 @@ data LevelCheck | LC_Promote -- Do a level check between the LHS tyvar and the occurrence tyvar -- If the level check fails, and the occurrence is a unification -- variable, promote it + Bool -- False <=> don't promote under type families (the common case) + -- True <=> promote even under type families + -- see Note [Defaulting equalities] in GHC.Tc.Solver instance Outputable (TyEqFlags a) where ppr (TEF { .. }) = text "TEF" <> braces ( @@ -3145,7 +3148,7 @@ instance Outputable (TyEqFlags a) where instance Outputable (TyEqFamApp a) where ppr TEFA_Fail = text "TEFA_Fail" - ppr TEFA_Recurse = text "TEFA_Fail" + ppr TEFA_Recurse = text "TEFA_Recurse" ppr (TEFA_Break {}) = text "TEFA_Break" instance Outputable AreUnifying where @@ -3154,9 +3157,9 @@ instance Outputable AreUnifying where braces (ppr mi <> comma <+> ppr lvl <> comma <+> ppr lc) instance Outputable LevelCheck where - ppr LC_None = text "LC_None" - ppr LC_Check = text "LC_Check" - ppr LC_Promote = text "LC_Promote" + ppr LC_None = text "LC_None" + ppr LC_Check = text "LC_Check" + ppr (LC_Promote b) = text "LC_Promote" <> ppWhen b (text "(deep)") famAppArgFlags :: TyEqFlags a -> TyEqFlags a -- Adjust the flags when going undter a type family @@ -3170,15 +3173,18 @@ famAppArgFlags flags@(TEF { tef_unifying = unifying }) , tef_occurs = cteSolubleOccurs } -- tef_occurs: under a type family, an occurs check is not definitely-insoluble where - zap_promotion (Unifying info lvl LC_Promote) = Unifying info lvl LC_Check - zap_promotion unifying = unifying + zap_promotion (Unifying info lvl (LC_Promote deeply)) + | not deeply = Unifying info lvl LC_Check + zap_promotion unifying = unifying type FamAppBreaker a = TcType -> TcM (PuResult a Reduction) -- Given a family-application ty, return a Reduction :: ty ~ cvb -- where 'cbv' is a fresh loop-breaker tyvar (for Given), or -- just a fresh TauTv (for Wanted) -checkTyEqRhs :: forall a. TyEqFlags a -> TcType -> TcM (PuResult a Reduction) +checkTyEqRhs :: forall a. TyEqFlags a + -> TcType -- Already zonked + -> TcM (PuResult a Reduction) checkTyEqRhs flags ty = case ty of LitTy {} -> okCheckRefl ty @@ -3229,7 +3235,7 @@ checkCo (TEF { tef_lhs = TyVarLHS lhs_tv = failCheckWith (cteProblem cteCoercionHole) -- Occurs check (can promote) - | Unifying _ lhs_tv_lvl LC_Promote <- unifying + | Unifying _ lhs_tv_lvl (LC_Promote {}) <- unifying = do { reason <- checkPromoteFreeVars occ_prob lhs_tv lhs_tv_lvl (tyCoVarsOfCo co) ; if cterHasNoProblem reason then return (pure co) @@ -3404,8 +3410,9 @@ checkFamApp flags@(TEF { tef_unifying = unifying, tef_occurs = occ_prob = case fam_app_flag of TEFA_Fail -> failCheckWith (cteProblem cteTypeFamily) + -- Occurs check: F ty ~ ...(F ty)... _ | TyFamLHS lhs_tc lhs_tys <- lhs - , tcEqTyConApps lhs_tc lhs_tys tc tys -- F ty ~ ...(F ty)... + , tcEqTyConApps lhs_tc lhs_tys tc tys -> case fam_app_flag of TEFA_Recurse -> failCheckWith (cteProblem occ_prob) TEFA_Break breaker -> breaker fam_app @@ -3421,7 +3428,11 @@ checkFamApp flags@(TEF { tef_unifying = unifying, tef_occurs = occ_prob ; traceTc "under" (ppr tc $$ pprPur tys_res $$ ppr flags) ; return (mkTyConAppRedn Nominal tc <$> tys_res) } - TEFA_Break breaker -- Recurse; and break if there is a problem + -- For TEFA_Break, try recursion; and break if there is a problem + -- e.g. alpha[2] ~ Maybe (F beta[2]) No problem: just unify + -- alpha[2] ~ Maybe (F beta[4]) Level-check problem: break + -- NB: in the latter case, don't promote beta[4]; hence arg_flags! + TEFA_Break breaker -> do { tys_res <- mapCheck (checkTyEqRhs arg_flags) tys ; case tys_res of PuOK cts redns -> return (PuOK cts (mkTyConAppRedn Nominal tc redns)) @@ -3458,7 +3469,7 @@ checkTyVar (TEF { tef_lhs = lhs, tef_unifying = unifying, tef_occurs = occ_prob Nothing -> check_unif info lvl prom lhs_tv } --------------------- - -- We are in the Unifying branch of AreUnifing + -- We are in the Unifying branch of AreUnifying; and occ_tv is unfilled check_unif :: MetaInfo -> TcLevel -> LevelCheck -> TcTyVar -> TcM (PuResult a Reduction) check_unif lhs_tv_info lhs_tv_lvl prom lhs_tv @@ -3472,7 +3483,7 @@ checkTyVar (TEF { tef_lhs = lhs, tef_unifying = unifying, tef_occurs = occ_prob = case prom of LC_None -> pprPanic "check_unif" (ppr lhs_tv $$ ppr occ_tv) LC_Check -> failCheckWith (cteProblem cteSkolemEscape) - LC_Promote + LC_Promote {} | isSkolemTyVar occ_tv -> failCheckWith (cteProblem cteSkolemEscape) | otherwise -> promote lhs_tv lhs_tv_info lhs_tv_lvl View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6863503cf3240f78549436819f20f9e1adc578d3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6863503cf3240f78549436819f20f9e1adc578d3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:13:28 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:13:28 -0400 Subject: [Git][ghc/ghc][master] JS: fix h$withCStringOnHeap helper (#25288) Message-ID: <66f47cf8c7ad5_373ff517b5d1c153855@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 6 changed files: - rts/js/mem.js - + testsuite/tests/javascript/js-c-sources/all.T - + testsuite/tests/javascript/js-c-sources/js-c-sources02.hs - + testsuite/tests/javascript/js-c-sources/js-c-sources02.stdout - + testsuite/tests/javascript/js-c-sources/js-c-sources02_c.c - + testsuite/tests/javascript/js-c-sources/js-c-sources02_js.js Changes: ===================================== rts/js/mem.js ===================================== @@ -1577,7 +1577,8 @@ function h$withCBufferOnHeap(str_d, str_o, len, cont) { // Temporarily allocate a CString on the heap and pass it to the continuation. // The string is freed from the heap when the continuation returns. function h$withCStringOnHeap(str_d, str_o, cont) { - return h$withCBufferOnHeap(str_d, str_o, str_d === null ? 0 : h$strlen(str_d,str_o), cont); + // strlen + 1 for the null terminating byte (#25288) + return h$withCBufferOnHeap(str_d, str_o, str_d === null ? 0 : h$strlen(str_d,str_o)+1, cont); } // Dereference a heap pointer to a heap pointer (a 32-bit offset in the heap) ===================================== testsuite/tests/javascript/js-c-sources/all.T ===================================== @@ -0,0 +1,5 @@ +# These are JavaScript-specific tests +setTestOpts(when(not(js_arch()),skip)) + +test('js-c-sources01', extra_files(['js-c-sources01_c.c','js-c-sources01_js.js']), compile_and_run, ['js-c-sources01_c.c js-c-sources01_js.js']) +test('js-c-sources02', extra_files(['js-c-sources02_c.c','js-c-sources02_js.js']), compile_and_run, ['js-c-sources02_c.c js-c-sources02_js.js']) ===================================== testsuite/tests/javascript/js-c-sources/js-c-sources02.hs ===================================== @@ -0,0 +1,6 @@ +import Foreign.C.String + +main :: IO () +main = withCString "Some string" foo + +foreign import javascript "foo" foo :: CString -> IO () ===================================== testsuite/tests/javascript/js-c-sources/js-c-sources02.stdout ===================================== @@ -0,0 +1 @@ +Arg: Some string ===================================== testsuite/tests/javascript/js-c-sources/js-c-sources02_c.c ===================================== @@ -0,0 +1,6 @@ +#include +#include + +void foo_c(char * s) { + printf("Arg: %s\n", s); +} ===================================== testsuite/tests/javascript/js-c-sources/js-c-sources02_js.js ===================================== @@ -0,0 +1,7 @@ +//#OPTIONS:CPP +//#OPTIONS:EMCC:EXPORTED_FUNCTIONS=_foo_c +//#OPTIONS:EMCC:EXPORTED_FUNCTIONS=_malloc,_free,_strlen + +function foo(str_d,str_o) { + h$withCStringOnHeap(str_d,str_o, (ptr) => _foo_c(ptr)); +} View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/97a6c6c3ded91af838f4376e4ad0781f0e971daf -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/97a6c6c3ded91af838f4376e4ad0781f0e971daf You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:27:54 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 25 Sep 2024 17:27:54 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 5 commits: hadrian: Build stage 2 cross compilers Message-ID: <66f4805a423cb_373ff51934e541704a3@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: 00e38633 by Matthew Pickering at 2024-09-25T22:27:46+01:00 hadrian: Build stage 2 cross compilers * Most of hadrian is abstracted over the stage in order to remove the assumption that the target of all stages is the same platform. This allows the RTS to be built for two different targets for example. * Abstracts the bindist creation logic to allow building either normal or cross bindists. Normal bindists use stage 1 libraries and a stage 2 compiler. Cross bindists use stage 2 libararies and a stage 2 compiler. * hadrian: Make binary-dist-dir the default build target. This allows us to have the logic in one place about which libraries/stages to build with cross compilers. Fixes #24192 New hadrian target: * `binary-dist-dir-cross`: Build a cross compiler bindist (compiler = stage 1, libraries = stage 2) ------------------------- Metric Decrease: T10421a T10858 T11195 T11276 T11374 T11822 T15630 T17096 T18478 T20261 Metric Increase: parsing001 ------------------------- - - - - - 174780d0 by Matthew Pickering at 2024-09-25T22:27:46+01:00 ci: Test cross bindists We remove the special logic for testing in-tree cross compilers and instead test cross compiler bindists, like we do for all other platforms. - - - - - 09f6e8f0 by Matthew Pickering at 2024-09-25T22:27:46+01:00 ci: Javascript don't set CROSS_EMULATOR There is no CROSS_EMULATOR needed to run javascript binaries, so we don't set the CROSS_EMULATOR to some dummy value. - - - - - b6e30bd2 by Matthew Pickering at 2024-09-25T22:27:46+01:00 ci: Introduce CROSS_STAGE variable In preparation for building and testing stage3 bindists we introduce the CROSS_STAGE variable which is used by a CI job to determine what kind of bindist the CI job should produce. At the moment we are only using CROSS_STAGE=2 but in the future we will have some jobs which set CROSS_STAGE=3 to produce native bindists for a target, but produced by a cross compiler, which can be tested on by another CI job on the native platform. CROSS_STAGE=2: Build a normal cross compiler bindist CROSS_STAGE=3: Build a stage 3 bindist, one which is a native compiler and library for the target - - - - - 67ba9efe by Matthew Pickering at 2024-09-25T22:27:46+01:00 Split up system.config into host/target config files There were a number of settings which were not applied per-stage, for example if you specified `--ffi-include-dir` then that was applied to both host and target. Now this will just be passed when building the crosscompiler. The solution for now is to separate these two files into host/target and the host file contains very bare-bones . There isn't currently a way to specify with configure anything in the host file, so if you are building a cross-compiler and you need to do that, you have to modify the file yourself. - - - - - 30 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - configure.ac - distrib/configure.ac.in - hadrian/README.md - hadrian/bindist/config.mk.in - + hadrian/cfg/system.config.host.in - hadrian/cfg/system.config.in - + hadrian/cfg/system.config.target.in - hadrian/hadrian.cabal - hadrian/src/Base.hs - + hadrian/src/BindistConfig.hs - hadrian/src/Builder.hs - hadrian/src/Context.hs - hadrian/src/Expression.hs - hadrian/src/Flavour.hs - hadrian/src/Flavour/Type.hs - hadrian/src/Hadrian/Expression.hs - hadrian/src/Hadrian/Haskell/Hash.hs - hadrian/src/Hadrian/Oracles/TextFile.hs - hadrian/src/Oracles/Flag.hs - hadrian/src/Oracles/Flavour.hs - hadrian/src/Oracles/Setting.hs - hadrian/src/Oracles/TestSettings.hs - hadrian/src/Packages.hs - hadrian/src/Rules.hs - hadrian/src/Rules/BinaryDist.hs - hadrian/src/Rules/CabalReinstall.hs - hadrian/src/Rules/Compile.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7dec8d6d0196ac509bff76039df2a6227439e668...67ba9efee683a93121c453e9e1cc8f43f77a25a4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7dec8d6d0196ac509bff76039df2a6227439e668...67ba9efee683a93121c453e9e1cc8f43f77a25a4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:31:38 2024 From: gitlab at gitlab.haskell.org (Alan Zimmerman (@alanz)) Date: Wed, 25 Sep 2024 17:31:38 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/az/epa-remove-addepann Message-ID: <66f4813a8d800_373ff519345e41719cd@gitlab.mail> Alan Zimmerman pushed new branch wip/az/epa-remove-addepann at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/az/epa-remove-addepann You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:34:13 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Wed, 25 Sep 2024 17:34:13 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] 13 commits: Demand: Combine examples into Note (#25107) Message-ID: <66f481d51a08e_106946bc4bc92493@gitlab.mail> Matthew Pickering pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 5e366df9 by Sven Tennie at 2024-09-25T17:25:10+01:00 WIP: RISCV64 cross-compile testing - - - - - 27 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - compiler/GHC/Core/Opt/DmdAnal.hs - compiler/GHC/Data/StringBuffer.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Driver/Plugins.hs - compiler/GHC/Hs.hs - compiler/GHC/Hs/Pat.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Runtime/Loader.hs - compiler/GHC/Tc/Deriv.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Gen/Pat.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Demand.hs - docs/users_guide/exts/extended_literals.rst - hadrian/bootstrap/bootstrap.py - hadrian/bootstrap/generate_bootstrap_plans - hadrian/bootstrap/plan-9_10_1.json - − hadrian/bootstrap/plan-9_4_1.json The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8279e7667fbb6133128f3eb74461608424533c08...5e366df9137e69feffd70d7a244ac779928daf3e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8279e7667fbb6133128f3eb74461608424533c08...5e366df9137e69feffd70d7a244ac779928daf3e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:38:12 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Wed, 25 Sep 2024 17:38:12 -0400 Subject: [Git][ghc/ghc][wip/T25281] Compiles now Message-ID: <66f482c49e8c8_106946bf11c939e1@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: fd8fb94e by Simon Peyton Jones at 2024-09-25T22:37:40+01:00 Compiles now - - - - - 1 changed file: - compiler/GHC/HsToCore/Expr.hs Changes: ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -31,13 +31,11 @@ import GHC.HsToCore.Arrows import GHC.HsToCore.Monad import GHC.HsToCore.Pmc import GHC.HsToCore.Errors.Types -import GHC.Types.SourceText -import GHC.Types.Name hiding (varName) -import GHC.Core.FamInstEnv( topNormaliseType ) import GHC.HsToCore.Quote import GHC.HsToCore.Ticks (stripTicksTopHsExpr) import GHC.Hs + -- NB: The desugarer, which straddles the source and Core worlds, sometimes -- needs to see source types import GHC.Tc.Utils.TcType @@ -48,26 +46,31 @@ import GHC.Core.TyCo.Rep import GHC.Core import GHC.Core.Utils import GHC.Core.Make +import GHC.Core.PatSyn import GHC.Driver.Session + +import GHC.Types.SourceText +import GHC.Types.Name hiding (varName) +import GHC.Core.FamInstEnv( topNormaliseType ) import GHC.Types.CostCentre import GHC.Types.Id import GHC.Types.Id.Info import GHC.Types.Id.Make +import GHC.Types.Basic +import GHC.Types.SrcLoc +import GHC.Types.Tickish + import GHC.Unit.Module import GHC.Core.ConLike import GHC.Core.DataCon import GHC.Builtin.Types import GHC.Builtin.Names -import GHC.Types.Basic -import GHC.Types.SrcLoc -import GHC.Types.Tickish + import GHC.Utils.Misc import GHC.Utils.Outputable as Outputable import GHC.Utils.Panic -import GHC.Core.PatSyn import Control.Monad -import GHC.Types.Error {- ************************************************************************ @@ -385,7 +388,7 @@ dsExpr (ExplicitTuple _ tup_args boxity) dsExpr (ExplicitSum types alt arity expr) = mkCoreUnboxedSum arity alt types <$> dsLExpr expr -dsExpr (HsPragE _ (HsPragSCC _ cc)) expr +dsExpr (HsPragE _ (HsPragSCC _ cc) expr) = do { dflags <- getDynFlags ; if sccProfilingEnabled dflags && gopt Opt_ProfManualCcs dflags then do @@ -411,11 +414,11 @@ dsExpr (HsLet _ binds body) = do -- We need the `ListComp' form to use `deListComp' (rather than the "do" form) -- because the interpretation of `stmts' depends on what sort of thing it is. -- -dsExpr (HsDo res_ty ListComp (L _ stmts)) = dsListComp stmts res_ty +dsExpr (HsDo res_ty ListComp (L _ stmts)) = dsListComp stmts res_ty +dsExpr (HsDo _ MonadComp (L _ stmts)) = dsMonadComp stmts dsExpr (HsDo res_ty ctx at DoExpr{} (L _ stmts)) = dsDo ctx stmts res_ty dsExpr (HsDo res_ty ctx at GhciStmtCtxt (L _ stmts)) = dsDo ctx stmts res_ty dsExpr (HsDo res_ty ctx at MDoExpr{} (L _ stmts)) = dsDo ctx stmts res_ty -dsExpr (HsDo _ MonadComp (L _ stmts)) = dsMonadComp stmts dsExpr (HsIf _ guard_expr then_expr else_expr) = do { pred <- dsLExpr guard_expr @@ -437,7 +440,6 @@ dsExpr (HsMultiIf res_ty alts) mkErrorExpr = mkErrorAppDs nON_EXHAUSTIVE_GUARDS_ERROR_ID res_ty (text "multi-way if") - dsExpr (ExplicitList elt_ty xs) = dsExplicitList elt_ty xs dsExpr (ArithSeq expr witness seq) @@ -446,41 +448,35 @@ dsExpr (ArithSeq expr witness seq) Just fl -> do { newArithSeq <- dsArithSeq expr seq ; dsSyntaxExpr fl [newArithSeq] } -{- -Static Pointers -~~~~~~~~~~~~~~~ - -See Note [Grand plan for static forms] in GHC.Iface.Tidy.StaticPtrTable for an overview. - +{- Note [Desugaring static pointers] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +See Note [Grand plan for static forms] in GHC.Iface.Tidy.StaticPtrTable +for an overview. g = ... static f ... ==> g = ... makeStatic loc f ... -} -dsExpr (HsStatic (_, whole_ty) expr@(L loc _)) = do - expr_ds <- dsLExpr expr - let (_, [ty]) = splitTyConApp whole_ty - makeStaticId <- dsLookupGlobalId makeStaticName - - dflags <- getDynFlags - let platform = targetPlatform dflags - let (line, col) = case locA loc of - RealSrcSpan r _ -> - ( srcLocLine $ realSrcSpanStart r - , srcLocCol $ realSrcSpanStart r - ) - _ -> (0, 0) - srcLoc = mkCoreTup [ mkIntExprInt platform line - , mkIntExprInt platform col - ] - - putSrcSpanDsA loc $ return $ - mkCoreApps (Var makeStaticId) [ Type ty, srcLoc, expr_ds ] - -{- -\noindent -\underline{\bf Record construction and update} - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +dsExpr (HsStatic (_, whole_ty) expr@(L loc _)) + = do { expr_ds <- dsLExpr expr + ; let (_, [ty]) = splitTyConApp whole_ty + ; makeStaticId <- dsLookupGlobalId makeStaticName + + ; dflags <- getDynFlags + ; let platform = targetPlatform dflags + (line, col) = case locA loc of + RealSrcSpan r _ -> ( srcLocLine $ realSrcSpanStart r + , srcLocCol $ realSrcSpanStart r ) + _ -> (0, 0) + srcLoc = mkCoreTup [ mkIntExprInt platform line + , mkIntExprInt platform col + ] + + ; putSrcSpanDsA loc $ return $ + mkCoreApps (Var makeStaticId) [ Type ty, srcLoc, expr_ds ] } + +{- Note [Desugaring record construction] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For record construction we do this (assuming T has three arguments) \begin{verbatim} T { op2 = e } @@ -503,6 +499,7 @@ constructor @C@, setting all of @C@'s fields to bottom. dsExpr (RecordCon { rcon_con = L _ con_like , rcon_flds = rbinds , rcon_ext = con_expr }) +-- See Note [Desugaring record construction] = do { con_expr' <- dsExpr con_expr ; let (arg_tys, _) = tcSplitFunTys (exprType con_expr') @@ -561,7 +558,7 @@ dsExpr (SectionR x _ _) = dataConCantHappen x * * ********************************************************************* -} -dsApp :: HsExpr -> DsM CoreExpr +dsApp :: HsExpr GhcTc -> DsM CoreExpr dsApp e = ds_app e [] [] ---------------------- @@ -570,26 +567,27 @@ ds_lapp :: LHsExpr GhcTc -> [LHsExpr GhcTc] -> [CoreExpr] -> DsM CoreExpr -- but there may be more of the latter because they include -- type and dictionary arguments ds_lapp (L loc e) hs_args core_args - = ds_app e hs_args core_args + = putSrcSpanDsA loc $ + ds_app e hs_args core_args -ds_app :: HsExpr GhcTc -> [LHsExpr] -> [CoreExpr] -> DsM CoreExpr +ds_app :: HsExpr GhcTc -> [LHsExpr GhcTc] -> [CoreExpr] -> DsM CoreExpr -- The work-horse -ds_app (HsPar _ e) hs_args core_args = dsLApp e hs_args core_args +ds_app (HsPar _ e) hs_args core_args = ds_lapp e hs_args core_args ds_app (HsApp _ fun arg) hs_args core_args = do { core_arg <- dsLExpr arg - ; dsLApp fun (arg : hs_args) (core_arg : core_args) } + ; ds_lapp fun (arg : hs_args) (core_arg : core_args) } ds_app (HsAppType arg_ty fun _) hs_args core_args - = dsLApp fun hs_args (Type ty : core_args) + = ds_lapp fun hs_args (Type arg_ty : core_args) ds_app (XExpr (WrapExpr hs_wrap fun)) hs_args core_args = do { (fun_wrap, all_args) <- splitHsWrapperArgs hs_wrap core_args ; if isIdHsWrapper fun_wrap - then ds_lapp fun hs_args all_args + then ds_app fun hs_args all_args else do { core_fun <- dsHsWrapper fun_wrap $ \core_wrap -> - do { core_fun <- dsLExpr fun - ; return (fun_wrap core_fun) } + do { core_fun <- dsExpr fun + ; return (core_wrap core_fun) } ; return (mkCoreApps core_fun all_args) } } ds_app (XExpr (ConLikeTc con tvs tys)) _hs_args core_args @@ -607,7 +605,7 @@ ds_app (XExpr (ConLikeTc con tvs tys)) _hs_args core_args ; return (mkApps core_fun core_args) } ds_app (HsRecSel _ fld_occ@(FieldOcc { foExt = fun_id })) hs_args core_args - | RecSelId {sel_cons = (_, cons_wo_field)} <- idDetails id + | RecSelId {sel_cons = (_, cons_wo_field)} <- idDetails fun_id = do { dflags <- getDynFlags -- Record selectors are warned about if they are not present in all of the @@ -640,40 +638,42 @@ ds_app (HsRecSel _ fld_occ@(FieldOcc { foExt = fun_id })) hs_args core_args -- the warning emitted during the desugaring of dsExpr(HsRecSel) -- See Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc -ds_app (HsVar _ (L _ fun_id)) core_args +ds_app (HsVar _ (L _ fun_id)) _hs_args core_args | fun_id `hasKey` noinlineIdKey -- See Note [noinlineId magic] in GHC.Types.Id.Make , Type _ : arg1 : rest_args <- core_args , (inner_fun, inner_args) <- collectArgs arg1 - = return (Var fun_id `App` Type (exprType inner_fun) `App` inner_fun) - `mkCoreApps` inner_args `mkCoreApps` rest_args + = return (Var fun_id `App` Type (exprType inner_fun) `App` inner_fun + `mkCoreApps` inner_args `mkCoreApps` rest_args) | fun_id `hasKey` seqIdKey -- Note [Desugaring seq], points (1) and (2) - , Type _r : Type ty1 : Type ty2 : arg1 : arg2 : rest_args - = reuturn (Case arg1 case_bndr ty2 [Alt DEFAULT [] (mkCoreApps arg2 rest_args)]) - where - case_bndr = case arg1 of - Var v1 | isInternalName (idName v1) - -> v1 -- Note [Desugaring seq], points (2) and (3) - _ -> mkWildValBinder ManyTy ty1 + , Type _r : Type ty1 : Type ty2 : arg1 : arg2 : rest_args <- core_args + , let case_bndr = case arg1 of + Var v1 | isInternalName (idName v1) + -> v1 -- Note [Desugaring seq], points (2) and (3) + _ -> mkWildValBinder ManyTy ty1 + = return (Case arg1 case_bndr ty2 [Alt DEFAULT [] (mkCoreApps arg2 rest_args)]) | otherwise = return (mkCoreApps (Var fun_id) core_args) +ds_app e _hs_args core_args + = do { core_e <- dsExpr e + ; return (mkCoreApps core_e core_args) } + ------------------------------ -splitHsWrapArgs :: HsWrapper -> [CoreArg] -> DsM (HsWrapper, [CoreArg]) +splitHsWrapperArgs :: HsWrapper -> [CoreArg] -> DsM (HsWrapper, [CoreArg]) -- Splits the wrapper into the trailing arguments, and leftover bit -splitHsWrapArg w args = go w args +splitHsWrapperArgs wrap args = go wrap args where go (WpTyApp ty) args = return (WpHole, Type ty : args) go (WpEvApp tm) args = do { core_tm <- dsEvTerm tm - ; (WpHole, Type ty : args) + ; return (WpHole, core_tm : args)} go (WpCompose w1 w2) args - = do { (w2', args') <- split_wrap w2 args - ; if isIdHsWrapper w2' - then go w1 args' - else return (w1 <.> w2', args') } - go wrap wrap args = return (wrap, args) - + = do { (w1', args') <- go w1 args + ; if isIdHsWrapper w1' + then go w2 args' + else return (w1' <.> w2, args') } + go wrap args = return (wrap, args) ------------------------------ dsHsConLike :: ConLike -> DsM CoreExpr @@ -689,8 +689,6 @@ dsHsConLike (PatSynCon ps) | otherwise = pprPanic "dsConLike" (ppr ps) - - ------------------------------ dsSyntaxExpr :: SyntaxExpr GhcTc -> [CoreExpr] -> DsM CoreExpr dsSyntaxExpr (SyntaxExprTc { syn_expr = expr @@ -818,10 +816,12 @@ Haskell 98 report: -} dsDo :: HsDoFlavour -> [ExprLStmt GhcTc] -> Type -> DsM CoreExpr --- SG: Surprisingly, this code path seems inactive for regular Do, +-- This code path seems inactive for regular Do, -- which is expanded in GHC.Tc.Gen.Do. --- It's all used for ApplicativeDo (even the BindStmt case), which is *very* +-- It is used only for ApplicativeDo (even the BindStmt case), which is *very* -- annoying because it is a lot of duplicated code that is seldomly tested. +-- But we are on course to expane Applicative in GHC.Tc.Gen.Do, at which +-- point all this will go away dsDo ctx stmts res_ty = goL stmts where @@ -986,11 +986,11 @@ Other places that requires from the same treatment: -} -- Warn about certain types of values discarded in monadic bindings (#3263) -warnUnusedMonadBindValue :: FieldOcc GhcTc -> [LHsExpr GhcTc] -> [CoreExpr] -> Type -> DsM () +warnUnusedMonadBindValue :: FieldOcc GhcTc -> [LHsExpr GhcTc] -> [CoreExpr] -> DsM () warnUnusedMonadBindValue (FieldOcc { foExt = fun_id, foLabel = L loc _ }) hs_args core_args | fun_id `hasKey` thenMClassOpKey -- it is a (>>) - , isGeneratedSrcSpan loc -- it is compiler generated (>>) + , isGeneratedSrcSpan (locA loc) -- it is compiler generated (>>) , Type arg_ty : _ <- core_args , hs_arg : _ <- hs_args = putSrcSpanDs (locA loc) $ warnDiscardedDoBindings hs_arg arg_ty @@ -1023,37 +1023,3 @@ warnDiscardedDoBindings rhs rhs_ty | otherwise -- RHS does have type of form (m ty), which is weird = return () -- but at least this warning is irrelevant - -{- -************************************************************************ -* * - dsHsWrapped -* * -************************************************************************ --} - ------------------------------- -dsHsWrapped :: HsExpr GhcTc -> DsM CoreExpr -dsHsWrapped orig_hs_expr - = go idHsWrapper orig_hs_expr - where - go wrap (HsPar _ (L _ hs_e)) - = go wrap hs_e - go wrap1 (XExpr (WrapExpr wrap2 hs_e)) - = go (wrap1 <.> wrap2) hs_e - go wrap (HsAppType ty (L _ hs_e) _) - = go (wrap <.> WpTyApp ty) hs_e - - go wrap (HsVar _ (L _ var)) - = do { dsHsWrapper wrap $ \wrap' -> do - { let expr = wrap' (varToCoreExpr var) - ty = exprType expr - ; dflags <- getDynFlags - ; warnAboutIdentities dflags var ty - ; return expr } } - - go wrap hs_e - = do { dsHsWrapper wrap $ \wrap' -> do - { addTyCs FromSource (hsWrapDictBinders wrap) $ - do { e <- dsExpr hs_e - ; return (wrap' e) } } } View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fd8fb94e0a7ee13c962102433912c5bf27579f29 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fd8fb94e0a7ee13c962102433912c5bf27579f29 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 21:44:00 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Wed, 25 Sep 2024 17:44:00 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 12 commits: Add Given injectivity for built-in type families Message-ID: <66f484201b3c9_1069464105149739b@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - ed5883ed by Ben Gamari at 2024-09-25T17:43:38-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c782291c by Matthew Pickering at 2024-09-25T17:43:39-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - b1d5b19f by Sylvain Henry at 2024-09-25T17:43:45-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - 106baea4 by Sylvain Henry at 2024-09-25T17:43:45-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 86bdd428 by Sylvain Henry at 2024-09-25T17:43:48-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 12 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Opt/ConstantFold.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/48700abd54735d941ed4dc130eca15b6b91d610c...86bdd4288093f9ed4a1ad4684b92c5740208c4f3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/48700abd54735d941ed4dc130eca15b6b91d610c...86bdd4288093f9ed4a1ad4684b92c5740208c4f3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 22:35:44 2024 From: gitlab at gitlab.haskell.org (Vladislav Zavialov (@int-index)) Date: Wed, 25 Sep 2024 18:35:44 -0400 Subject: [Git][ghc/ghc][wip/int-index/wildcard-binders] 17 commits: ci: Run abi-test on test-abi label Message-ID: <66f4904065bd4_1662441437c81000d@gitlab.mail> Vladislav Zavialov pushed to branch wip/int-index/wildcard-binders at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 4dceb0ff by Vladislav Zavialov at 2024-09-26T01:35:22+03:00 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - 27 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Dwarf.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8011bc880b194219b6203b9f6939b17798678ad3...4dceb0ff71afe56484c46d6e52456650a6d643dc -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8011bc880b194219b6203b9f6939b17798678ad3...4dceb0ff71afe56484c46d6e52456650a6d643dc You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 23:03:58 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Wed, 25 Sep 2024 19:03:58 -0400 Subject: [Git][ghc/ghc][wip/unwire-base] 95 commits: Haddock: Add no-compilation flag Message-ID: <66f496de78000_166244386a08110826@gitlab.mail> Zubin pushed to branch wip/unwire-base at Glasgow Haskell Compiler / GHC Commits: 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 63585cf2 by Matthew Pickering at 2024-09-26T04:29:11+05:30 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 17ef0934 by Matthew Pickering at 2024-09-26T04:29:13+05:30 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - d1240409 by Matthew Pickering at 2024-09-26T04:29:13+05:30 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - b5850af2 by Matthew Pickering at 2024-09-26T04:29:13+05:30 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - eeaf80e4 by Matthew Pickering at 2024-09-26T04:32:49+05:30 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - 1f4b609a by Matthew Pickering at 2024-09-26T04:33:45+05:30 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 30 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a7619873eee80609e4d1049ecb04eedd0e927c6f...1f4b609a8bce5e4ce915a968cf0c50db98c63a60 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a7619873eee80609e4d1049ecb04eedd0e927c6f...1f4b609a8bce5e4ce915a968cf0c50db98c63a60 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 23:43:09 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Wed, 25 Sep 2024 19:43:09 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 32 commits: ci: Run abi-test on test-abi label Message-ID: <66f4a00d20426_166244512f20111712@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - d6ec6f2a by sheaf at 2024-09-26T00:50:17+02:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 8a64f210 by sheaf at 2024-09-26T00:50:21+02:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 5b58a3f1 by sheaf at 2024-09-26T00:50:21+02:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - d3e3c31c by sheaf at 2024-09-26T00:50:22+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 408ede55 by sheaf at 2024-09-26T00:50:51+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 2150879d by sheaf at 2024-09-26T00:50:53+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 1b439775 by sheaf at 2024-09-26T00:50:54+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - a06e9dd0 by sheaf at 2024-09-26T00:50:54+02:00 Add test for C calls & SIMD vectors - - - - - 5358ce5b by sheaf at 2024-09-26T00:50:54+02:00 Add test for #25169 - - - - - 83f08b00 by sheaf at 2024-09-26T00:50:54+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - 26f68f5b by sheaf at 2024-09-26T00:51:08+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 2f123d5f by sheaf at 2024-09-26T00:51:08+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - f624595d by sheaf at 2024-09-26T00:51:08+02:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 2a552087 by sheaf at 2024-09-26T00:51:08+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 72ef4f97 by sheaf at 2024-09-26T00:51:08+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 1df6d224 by sheaf at 2024-09-26T01:01:25+02:00 shuffle fixup - - - - - 21 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3fc52e65de1dfdb7e57df8494614e2d3d48b01f4...1df6d2247e43ff87049c3822512b40c61b4f4076 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3fc52e65de1dfdb7e57df8494614e2d3d48b01f4...1df6d2247e43ff87049c3822512b40c61b4f4076 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Wed Sep 25 23:43:33 2024 From: gitlab at gitlab.haskell.org (sheaf (@sheaf)) Date: Wed, 25 Sep 2024 19:43:33 -0400 Subject: [Git][ghc/ghc][wip/ncg-simd] 12 commits: Add vector shuffle primops Message-ID: <66f4a0254d1eb_1662445afa2811316d@gitlab.mail> sheaf pushed to branch wip/ncg-simd at Glasgow Haskell Compiler / GHC Commits: 42491d52 by sheaf at 2024-09-26T01:43:24+02:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 172a8e02 by sheaf at 2024-09-26T01:43:24+02:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - 056971c0 by sheaf at 2024-09-26T01:43:24+02:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - efae0744 by sheaf at 2024-09-26T01:43:24+02:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - cc2f957e by sheaf at 2024-09-26T01:43:24+02:00 Add test for C calls & SIMD vectors - - - - - ab9269a0 by sheaf at 2024-09-26T01:43:25+02:00 Add test for #25169 - - - - - 558941ea by sheaf at 2024-09-26T01:43:25+02:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - 9cd13d1c by sheaf at 2024-09-26T01:43:25+02:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 151773db by sheaf at 2024-09-26T01:43:25+02:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - db5ec5e5 by sheaf at 2024-09-26T01:43:25+02:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - ace43c6f by sheaf at 2024-09-26T01:43:25+02:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 20026067 by sheaf at 2024-09-26T01:43:25+02:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Reg.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs - compiler/GHC/CmmToAsm/Wasm/Asm.hs - compiler/GHC/CmmToAsm/Wasm/FromCmm.hs - compiler/GHC/CmmToAsm/Wasm/Types.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/CmmToAsm/X86/Instr.hs - compiler/GHC/CmmToAsm/X86/Ppr.hs - compiler/GHC/CmmToC.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/CmmToLlvm/CodeGen.hs - compiler/GHC/CmmToLlvm/Ppr.hs - compiler/GHC/CmmToLlvm/Regs.hs - compiler/GHC/JS/Make.hs - compiler/GHC/Llvm/Ppr.hs - compiler/GHC/Llvm/Syntax.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1df6d2247e43ff87049c3822512b40c61b4f4076...2002606738f327260632402561189a308098da1f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1df6d2247e43ff87049c3822512b40c61b4f4076...2002606738f327260632402561189a308098da1f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 01:14:14 2024 From: gitlab at gitlab.haskell.org (Cheng Shao (@TerrorJack)) Date: Wed, 25 Sep 2024 21:14:14 -0400 Subject: [Git][ghc/ghc] Deleted branch wip/wasm-bump Message-ID: <66f4b5666c69c_166244c1b600147187@gitlab.mail> Cheng Shao deleted branch wip/wasm-bump at Glasgow Haskell Compiler / GHC -- You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 08:14:31 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 04:14:31 -0400 Subject: [Git][ghc/ghc][master] base: Propagate `error` CallStack to thrown exception Message-ID: <66f517e7e1d8e_3f975ae2554975f@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - 30 changed files: - docs/users_guide/9.12.1-notes.rst - libraries/base/changelog.md - libraries/base/tests/T19288.stderr - + libraries/base/tests/T24807.hs - + libraries/base/tests/T24807.stderr - libraries/base/tests/all.T - libraries/ghc-internal/src/GHC/Internal/Exception.hs - testsuite/tests/arityanal/should_run/T21694a.stderr - testsuite/tests/arityanal/should_run/T24296.stderr - testsuite/tests/codeGen/should_run/T16846.stderr - testsuite/tests/codeGen/should_run/T5626.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/codeGen/should_run/cgrun045.stderr - testsuite/tests/codeGen/should_run/cgrun051.stderr - testsuite/tests/codeGen/should_run/cgrun059.stderr - testsuite/tests/concurrent/should_run/T13330.stderr - testsuite/tests/concurrent/should_run/conc021.stderr - testsuite/tests/deSugar/should_run/Or5.stderr - testsuite/tests/deSugar/should_run/T11193.stderr - testsuite/tests/deSugar/should_run/T11572.stderr - testsuite/tests/deSugar/should_run/T11601.stderr - testsuite/tests/dmdanal/should_run/T12368.stderr - testsuite/tests/dmdanal/should_run/T12368a.stderr - testsuite/tests/dmdanal/should_run/T23208.stderr - testsuite/tests/dmdanal/should_run/strun002.stderr - testsuite/tests/driver/multipleHomeUnits/multipleHomeUnits_callstack.stderr - testsuite/tests/ffi/should_run/ffi008.stderr - testsuite/tests/hpc/simple/tixs/T10529a.stderr - testsuite/tests/hpc/simple/tixs/T10529b.stderr - testsuite/tests/hpc/simple/tixs/T10529c.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5f7c20bc6b9deacabb1fb099781f00371cdb4369 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5f7c20bc6b9deacabb1fb099781f00371cdb4369 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 08:15:27 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 04:15:27 -0400 Subject: [Git][ghc/ghc][master] driver: Fix -working-dir for foreign files Message-ID: <66f5181faaf08_3f975ae2c341026ae@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 5 changed files: - compiler/GHC/Driver/Pipeline.hs - testsuite/tests/driver/all.T - testsuite/tests/driver/multipleHomeUnits/unitCFile - + testsuite/tests/driver/t25150/dir/Main.hs - + testsuite/tests/driver/t25150/dir/a.c Changes: ===================================== compiler/GHC/Driver/Pipeline.hs ===================================== @@ -544,28 +544,28 @@ oneShot orig_hsc_env stop_phase srcs = do compileFile :: HscEnv -> StopPhase -> (FilePath, Maybe Phase) -> IO (Maybe FilePath) compileFile hsc_env stop_phase (src, mb_phase) = do - exists <- doesFileExist src + let offset_file = augmentByWorkingDirectory dflags src + dflags = hsc_dflags hsc_env + mb_o_file = outputFile dflags + ghc_link = ghcLink dflags -- Set by -c or -no-link + notStopPreprocess | StopPreprocess <- stop_phase = False + | _ <- stop_phase = True + -- When linking, the -o argument refers to the linker's output. + -- otherwise, we use it as the name for the pipeline's output. + output + | not (backendGeneratesCode (backend dflags)), notStopPreprocess = NoOutputFile + -- avoid -E -fno-code undesirable interactions. see #20439 + | NoStop <- stop_phase, not (isNoLink ghc_link) = Persistent + -- -o foo applies to linker + | isJust mb_o_file = SpecificFile + -- -o foo applies to the file we are compiling now + | otherwise = Persistent + pipe_env = mkPipeEnv stop_phase offset_file mb_phase output + pipeline = pipelineStart pipe_env (setDumpPrefix pipe_env hsc_env) offset_file mb_phase + + exists <- doesFileExist offset_file when (not exists) $ - throwGhcExceptionIO (CmdLineError ("does not exist: " ++ src)) - - let - dflags = hsc_dflags hsc_env - mb_o_file = outputFile dflags - ghc_link = ghcLink dflags -- Set by -c or -no-link - notStopPreprocess | StopPreprocess <- stop_phase = False - | _ <- stop_phase = True - -- When linking, the -o argument refers to the linker's output. - -- otherwise, we use it as the name for the pipeline's output. - output - | not (backendGeneratesCode (backend dflags)), notStopPreprocess = NoOutputFile - -- avoid -E -fno-code undesirable interactions. see #20439 - | NoStop <- stop_phase, not (isNoLink ghc_link) = Persistent - -- -o foo applies to linker - | isJust mb_o_file = SpecificFile - -- -o foo applies to the file we are compiling now - | otherwise = Persistent - pipe_env = mkPipeEnv stop_phase src mb_phase output - pipeline = pipelineStart pipe_env (setDumpPrefix pipe_env hsc_env) src mb_phase + throwGhcExceptionIO (CmdLineError ("does not exist: " ++ offset_file)) runPipeline (hsc_hooks hsc_env) pipeline ===================================== testsuite/tests/driver/all.T ===================================== @@ -328,3 +328,4 @@ test('T23613', normal, compile_and_run, ['-this-unit-id=foo']) test('T23944', [unless(have_dynamic(), skip), extra_files(['T23944A.hs'])], multimod_compile, ['T23944 T23944A', '-fprefer-byte-code -fbyte-code -fno-code -dynamic-too -fwrite-interface']) test('T24286', [cxx_src, unless(have_profiling(), skip), extra_files(['T24286.cpp'])], compile, ['-prof -no-hs-main']) test('T24839', [unless(arch('x86_64') or arch('aarch64'), skip), extra_files(["t24839_sub.S"])], compile_and_run, ['t24839_sub.S']) +test('t25150', [extra_files(["t25150"])], multimod_compile, ['Main.hs', '-v0 -working-dir t25150/dir a.c']) ===================================== testsuite/tests/driver/multipleHomeUnits/unitCFile ===================================== @@ -1 +1 @@ --working-dir c-file C c-file/c.c -Iinclude +-working-dir c-file C c.c -Iinclude ===================================== testsuite/tests/driver/t25150/dir/Main.hs ===================================== @@ -0,0 +1,6 @@ +{-# LANGUAGE ForeignFunctionInterface #-} +module Main where +foreign import ccall "hello" + hello :: IO () +main :: IO () +main = hello ===================================== testsuite/tests/driver/t25150/dir/a.c ===================================== @@ -0,0 +1,5 @@ +// dir/Aux.c +#include +void hello() { + printf("hi"); +} View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c20d51867c824e32c61bd1e002680bef268e4f51 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c20d51867c824e32c61bd1e002680bef268e4f51 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 08:16:53 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 04:16:53 -0400 Subject: [Git][ghc/ghc][master] 2 commits: Enum deriving: reuse predError, succError, toEnumError Message-ID: <66f5187525ed0_3f975ae259010867b@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 18 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/Tc/Deriv/Generate.hs - libraries/base/tests/enum01.stdout - libraries/base/tests/enum01.stdout-alpha-dec-osf3 - libraries/base/tests/enum01.stdout-ws-64 - libraries/base/tests/enum02.stdout - libraries/base/tests/enum02.stdout-alpha-dec-osf3 - libraries/base/tests/enum02.stdout-mips-sgi-irix - libraries/base/tests/enum02.stdout-ws-64 - libraries/base/tests/enum02.stdout-x86_64-unknown-openbsd - libraries/base/tests/enum03.stdout - libraries/base/tests/enum03.stdout-alpha-dec-osf3 - libraries/base/tests/enum03.stdout-mips-sgi-irix - libraries/base/tests/enum03.stdout-ws-64 - libraries/base/tests/enum03.stdout-x86_64-unknown-openbsd - libraries/ghc-internal/codepages/MakeTable.hs - libraries/ghc-internal/src/GHC/Internal/Enum.hs Changes: ===================================== compiler/GHC/Builtin/Names.hs ===================================== @@ -751,9 +751,13 @@ left_RDR, right_RDR :: RdrName left_RDR = nameRdrName leftDataConName right_RDR = nameRdrName rightDataConName -fromEnum_RDR, toEnum_RDR :: RdrName +fromEnum_RDR, toEnum_RDR, toEnumError_RDR, succError_RDR, predError_RDR, enumIntToWord_RDR :: RdrName fromEnum_RDR = varQual_RDR gHC_INTERNAL_ENUM (fsLit "fromEnum") toEnum_RDR = varQual_RDR gHC_INTERNAL_ENUM (fsLit "toEnum") +toEnumError_RDR = varQual_RDR gHC_INTERNAL_ENUM (fsLit "toEnumError") +succError_RDR = varQual_RDR gHC_INTERNAL_ENUM (fsLit "succError") +predError_RDR = varQual_RDR gHC_INTERNAL_ENUM (fsLit "predError") +enumIntToWord_RDR = varQual_RDR gHC_INTERNAL_ENUM (fsLit "enumIntToWord") enumFrom_RDR, enumFromTo_RDR, enumFromThen_RDR, enumFromThenTo_RDR :: RdrName enumFrom_RDR = nameRdrName enumFromName @@ -1300,6 +1304,7 @@ fromIntegralName = varQual gHC_INTERNAL_REAL (fsLit "fromIntegral")fromInteg realToFracName = varQual gHC_INTERNAL_REAL (fsLit "realToFrac") realToFracIdKey mkRationalBase2Name = varQual gHC_INTERNAL_REAL (fsLit "mkRationalBase2") mkRationalBase2IdKey mkRationalBase10Name = varQual gHC_INTERNAL_REAL (fsLit "mkRationalBase10") mkRationalBase10IdKey + -- GHC.Internal.Float classes floatingClassName, realFloatClassName :: Name floatingClassName = clsQual gHC_INTERNAL_FLOAT (fsLit "Floating") floatingClassKey ===================================== compiler/GHC/Hs/Utils.hs ===================================== @@ -58,7 +58,7 @@ module GHC.Hs.Utils( nlHsIntLit, nlHsVarApps, nlHsDo, nlHsOpApp, nlHsLam, nlHsPar, nlHsIf, nlHsCase, nlList, mkLHsTupleExpr, mkLHsVarTuple, missingTupArg, - mkLocatedList, + mkLocatedList, nlAscribe, -- * Bindings mkFunBind, mkVarBind, mkHsVarBind, mkSimpleGeneratedFunBind, mkTopFunBind, @@ -769,6 +769,13 @@ mkClassOpSigs sigs = L loc (ClassOpSig anns False nms (dropWildCards ty)) fiddle sig = sig + +-- | Type ascription: (e :: ty) +nlAscribe :: RdrName -> LHsExpr GhcPs -> LHsExpr GhcPs +nlAscribe ty e = noLocA $ ExprWithTySig noAnn e + $ mkHsWildCardBndrs $ noLocA $ mkHsImplicitSigType + $ nlHsTyVar NotPromoted ty + {- ********************************************************************* * * --------- HsWrappers: type args, dict args, casts --------- ===================================== compiler/GHC/Tc/Deriv/Generate.hs ===================================== @@ -594,9 +594,7 @@ unliftedCompare lt_op eq_op a_expr b_expr lt eq gt -- mean more tests (dynamically) nlHsIf (ascribeBool $ genPrimOpApp a_expr eq_op b_expr) eq gt where - ascribeBool e = noLocA $ ExprWithTySig noAnn e - $ mkHsWildCardBndrs $ noLocA $ mkHsImplicitSigType - $ nlHsTyVar NotPromoted boolTyCon_RDR + ascribeBool = nlAscribe boolTyCon_RDR nlConWildPat :: DataCon -> LPat GhcPs -- The pattern (K {}) @@ -681,7 +679,7 @@ gen_Enum_binds loc (DerivInstTys{dit_rep_tc = tycon}) = do untag_Expr [(a_RDR, ah_RDR)] $ nlHsIf (nlHsApps eq_RDR [nlHsVar maxtag_RDR, nlHsVarApps intDataCon_RDR [ah_RDR]]) - (illegal_Expr "succ" occ_nm "tried to take `succ' of last tag in enumeration") + (nlHsApp (nlHsVar succError_RDR) (nlHsLit (mkHsString occ_nm))) (nlHsApp (nlHsVar tag2con_RDR) (nlHsApps plus_RDR [nlHsVarApps intDataCon_RDR [ah_RDR], nlHsIntLit 1])) @@ -691,7 +689,7 @@ gen_Enum_binds loc (DerivInstTys{dit_rep_tc = tycon}) = do untag_Expr [(a_RDR, ah_RDR)] $ nlHsIf (nlHsApps eq_RDR [nlHsIntLit 0, nlHsVarApps intDataCon_RDR [ah_RDR]]) - (illegal_Expr "pred" occ_nm "tried to take `pred' of first tag in enumeration") + (nlHsApp (nlHsVar predError_RDR) (nlHsLit (mkHsString occ_nm))) (nlHsApp (nlHsVar tag2con_RDR) (nlHsApps plus_RDR [ nlHsVarApps intDataCon_RDR [ah_RDR] @@ -700,12 +698,16 @@ gen_Enum_binds loc (DerivInstTys{dit_rep_tc = tycon}) = do to_enum tag2con_RDR maxtag_RDR = mkSimpleGeneratedFunBind loc toEnum_RDR (noLocA [a_Pat]) $ - nlHsIf (nlHsApps and_RDR - [nlHsApps ge_RDR [nlHsVar a_RDR, nlHsIntLit 0], - nlHsApps le_RDR [ nlHsVar a_RDR - , nlHsVar maxtag_RDR]]) + let to_word = nlHsApp (nlHsVar enumIntToWord_RDR) + -- cast to Word to check both bounds (0,maxtag) with one comparison + in nlHsIf (nlHsApps le_RDR [ to_word (nlHsVar a_RDR), to_word (nlHsVar maxtag_RDR)]) (nlHsVarApps tag2con_RDR [a_RDR]) - (illegal_toEnum_tag occ_nm maxtag_RDR) + (nlHsApps toEnumError_RDR + [ nlHsLit (mkHsString occ_nm) + , nlHsVar a_RDR + , mkLHsTupleExpr [nlHsIntLit 0, nlHsVar maxtag_RDR] noAnn + ]) + enum_from tag2con_RDR maxtag_RDR = mkSimpleGeneratedFunBind loc enumFrom_RDR (noLocA [a_Pat]) $ @@ -2537,32 +2539,6 @@ nested_compose_Expr (e:es) error_Expr :: FastString -> LHsExpr GhcPs error_Expr string = nlHsApp (nlHsVar error_RDR) (nlHsLit (mkHsStringFS string)) --- illegal_Expr is used when signalling error conditions in the RHS of a derived --- method. It is currently only used by Enum.{succ,pred} -illegal_Expr :: String -> String -> String -> LHsExpr GhcPs -illegal_Expr meth tp msg = - nlHsApp (nlHsVar error_RDR) (nlHsLit (mkHsString (meth ++ '{':tp ++ "}: " ++ msg))) - --- illegal_toEnum_tag is an extended version of illegal_Expr, which also allows you --- to include the value of a_RDR in the error string. -illegal_toEnum_tag :: String -> RdrName -> LHsExpr GhcPs -illegal_toEnum_tag tp maxtag = - nlHsApp (nlHsVar error_RDR) - (nlHsApp (nlHsApp (nlHsVar append_RDR) - (nlHsLit (mkHsString ("toEnum{" ++ tp ++ "}: tag (")))) - (nlHsApp (nlHsApp (nlHsApp - (nlHsVar showsPrec_RDR) - (nlHsIntLit 0)) - (nlHsVar a_RDR)) - (nlHsApp (nlHsApp - (nlHsVar append_RDR) - (nlHsLit (mkHsString ") is outside of enumeration's range (0,"))) - (nlHsApp (nlHsApp (nlHsApp - (nlHsVar showsPrec_RDR) - (nlHsIntLit 0)) - (nlHsVar maxtag)) - (nlHsLit (mkHsString ")")))))) - parenify :: LHsExpr GhcPs -> LHsExpr GhcPs parenify e@(L _ (HsVar _ _)) = e parenify e = mkHsPar e ===================================== libraries/base/tests/enum01.stdout ===================================== @@ -1,10 +1,10 @@ Testing Enum Int: (succ (0::Int)) = 1 (succ (minBound::Int)) = -2147483647 - (succ (maxBound::Int)) = error "Prelude.Enum.succ{Int}: tried to take `succ' of maxBound" + (succ (maxBound::Int)) = error "Enum.succ{Int}: tried to take `succ' of last tag in enumeration" pred (1::Int) = 0 pred (maxBound::Int) = 2147483646 - pred (minBound::Int) = error "Prelude.Enum.pred{Int}: tried to take `pred' of minBound" + pred (minBound::Int) = error "Enum.pred{Int}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int) [1,minBound,maxBound]) = [1,-2147483648,2147483647] (map fromEnum [(1::Int),minBound,maxBound]) = [1,-2147483648,2147483647] (take 7 [(1::Int)..]) = [1,2,3,4,5,6,7] ===================================== libraries/base/tests/enum01.stdout-alpha-dec-osf3 ===================================== @@ -1,10 +1,10 @@ Testing Enum Int: (succ (0::Int)) = 1 (succ (minBound::Int)) = -9223372036854775807 - (succ (maxBound::Int)) = error "Prelude.Enum.succ{Int}: tried to take `succ' of maxBound" + (succ (maxBound::Int)) = error "Enum.succ{Int}: tried to take `succ' of last tag in enumeration" pred (1::Int) = 0 pred (maxBound::Int) = 9223372036854775806 - pred (minBound::Int) = error "Prelude.Enum.pred{Int}: tried to take `pred' of minBound" + pred (minBound::Int) = error "Enum.pred{Int}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int) [1,minBound,maxBound]) = [1,-9223372036854775808,9223372036854775807] (map fromEnum [(1::Int),minBound,maxBound]) = [1,-9223372036854775808,9223372036854775807] (take 7 [(1::Int)..]) = [1,2,3,4,5,6,7] ===================================== libraries/base/tests/enum01.stdout-ws-64 ===================================== @@ -1,10 +1,10 @@ Testing Enum Int: (succ (0::Int)) = 1 (succ (minBound::Int)) = -9223372036854775807 - (succ (maxBound::Int)) = error "Prelude.Enum.succ{Int}: tried to take `succ' of maxBound" + (succ (maxBound::Int)) = error "Enum.succ{Int}: tried to take `succ' of last tag in enumeration" pred (1::Int) = 0 pred (maxBound::Int) = 9223372036854775806 - pred (minBound::Int) = error "Prelude.Enum.pred{Int}: tried to take `pred' of minBound" + pred (minBound::Int) = error "Enum.pred{Int}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int) [1,minBound,maxBound]) = [1,-9223372036854775808,9223372036854775807] (map fromEnum [(1::Int),minBound,maxBound]) = [1,-9223372036854775808,9223372036854775807] (take 7 [(1::Int)..]) = [1,2,3,4,5,6,7] ===================================== libraries/base/tests/enum02.stdout ===================================== @@ -1,12 +1,12 @@ Testing Enum Int8: (succ (0::Int8)) = 1 (succ (minBound::Int8)) = -127 - (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of maxBound" + (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of last tag in enumeration" pred (1::Int8) = 0 pred (maxBound::Int8) = 126 - pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of minBound" + pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int8) [1, fromIntegral (minBound::Int8), fromIntegral (maxBound::Int8)]) = [1,-128,127] - (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (2147483647) is outside of bounds (-128,127)" + (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (2147483647) is outside of enumeration's range (-128,127)" (map fromEnum [(1::Int8),minBound,maxBound]) = [1,-128,127] (take 7 [(1::Int8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int8)-5)..]) = [122,123,124,125,126,127] @@ -36,12 +36,12 @@ Testing Enum Int8: Testing Enum Int16: (succ (0::Int16)) = 1 (succ (minBound::Int16)) = -32767 - (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of maxBound" + (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of last tag in enumeration" pred (1::Int16) = 0 pred (maxBound::Int16) = 32766 - pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of minBound" + pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int16) [1, fromIntegral (minBound::Int16), fromIntegral (maxBound::Int16)]) = [1,-32768,32767] - (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (2147483647) is outside of bounds (-32768,32767)" + (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (2147483647) is outside of enumeration's range (-32768,32767)" (map fromEnum [(1::Int16),minBound,maxBound]) = [1,-32768,32767] (take 7 [(1::Int16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int16)-5)..]) = [32762,32763,32764,32765,32766,32767] @@ -71,10 +71,10 @@ Testing Enum Int16: Testing Enum Int32: (succ (0::Int32)) = 1 (succ (minBound::Int32)) = -2147483647 - (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of maxBound" + (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of last tag in enumeration" pred (1::Int32) = 0 pred (maxBound::Int32) = 2147483646 - pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of minBound" + pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int32) [1, fromIntegral (minBound::Int32), fromIntegral (maxBound::Int32)]) = [1,-2147483648,2147483647] (toEnum (maxBound::Int))::Int32 = 2147483647 (map fromEnum [(1::Int32),minBound,maxBound]) = [1,-2147483648,2147483647] @@ -106,10 +106,10 @@ Testing Enum Int32: Testing Enum Int64: (succ (0::Int64)) = 1 (succ (minBound::Int64)) = -9223372036854775807 - (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of maxBound" + (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of last tag in enumeration" pred (1::Int64) = 0 pred (maxBound::Int64) = 9223372036854775806 - pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of minBound" + pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int64) [1, fromIntegral (minBound::Int64), fromIntegral (maxBound::Int64)]) = [1,0,-1] (toEnum (maxBound::Int))::Int64 = 2147483647 (map fromEnum [(1::Int64),fromIntegral (minBound::Int) ,fromIntegral (maxBound::Int)]) = [1,-2147483648,2147483647] ===================================== libraries/base/tests/enum02.stdout-alpha-dec-osf3 ===================================== @@ -1,12 +1,12 @@ Testing Enum Int8: (succ (0::Int8)) = 1 (succ (minBound::Int8)) = -127 - (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of maxBound" + (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of last tag in enumeration" pred (1::Int8) = 0 pred (maxBound::Int8) = 126 - pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of minBound" + pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int8) [1, toInt (minBound::Int8), toInt (maxBound::Int8)]) = [1,-128,127] - (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (9223372036854775807) is outside of bounds (-128,127)" + (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (9223372036854775807) is outside of enumeration's range (-128,127)" (map fromEnum [(1::Int8),minBound,maxBound]) = [1,-128,127] (take 7 [(1::Int8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int8)-5)..]) = [122,123,124,125,126,127] @@ -36,12 +36,12 @@ Testing Enum Int8: Testing Enum Int16: (succ (0::Int16)) = 1 (succ (minBound::Int16)) = -32767 - (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of maxBound" + (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of last tag in enumeration" pred (1::Int16) = 0 pred (maxBound::Int16) = 32766 - pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of minBound" + pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int16) [1, toInt (minBound::Int16), toInt (maxBound::Int16)]) = [1,-32768,32767] - (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (9223372036854775807) is outside of bounds (-32768,32767)" + (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (9223372036854775807) is outside of enumeration's range (-32768,32767)" (map fromEnum [(1::Int16),minBound,maxBound]) = [1,-32768,32767] (take 7 [(1::Int16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int16)-5)..]) = [32762,32763,32764,32765,32766,32767] @@ -71,12 +71,12 @@ Testing Enum Int16: Testing Enum Int32: (succ (0::Int32)) = 1 (succ (minBound::Int32)) = -2147483647 - (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of maxBound" + (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of last tag in enumeration" pred (1::Int32) = 0 pred (maxBound::Int32) = 2147483646 - pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of minBound" + pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int32) [1, toInt (minBound::Int32), toInt (maxBound::Int32)]) = [1,-2147483648,2147483647] - (toEnum (maxBound::Int))::Int32 = error "Enum.toEnum{Int32}: tag (9223372036854775807) is outside of bounds (-2147483648,2147483647)" + (toEnum (maxBound::Int))::Int32 = error "Enum.toEnum{Int32}: tag (9223372036854775807) is outside of enumeration's range (-2147483648,2147483647)" (map fromEnum [(1::Int32),minBound,maxBound]) = [1,-2147483648,2147483647] (take 7 [(1::Int32)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int32)-5)..]) = [2147483642,2147483643,2147483644,2147483645,2147483646,2147483647] @@ -106,10 +106,10 @@ Testing Enum Int32: Testing Enum Int64: (succ (0::Int64)) = 1 (succ (minBound::Int64)) = -9223372036854775807 - (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of maxBound" + (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of last tag in enumeration" pred (1::Int64) = 0 pred (maxBound::Int64) = 9223372036854775806 - pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of minBound" + pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int64) [1, toInt (minBound::Int64), toInt (maxBound::Int64)]) = [1,-9223372036854775808,9223372036854775807] (toEnum (maxBound::Int))::Int64 = 9223372036854775807 (map fromEnum [(1::Int64),fromInt (minBound::Int) ,fromInt (maxBound::Int)]) = [1,-9223372036854775808,9223372036854775807] ===================================== libraries/base/tests/enum02.stdout-mips-sgi-irix ===================================== @@ -1,12 +1,12 @@ Testing Enum Int8: (succ (0::Int8)) = 1 (succ (minBound::Int8)) = -127 - (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of maxBound" + (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of last tag in enumeration" pred (1::Int8) = 0 pred (maxBound::Int8) = 126 - pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of minBound" + pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int8) [1, fromIntegral (minBound::Int8), fromIntegral (maxBound::Int8)]) = [1,-128,127] - (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (9223372036854775807) is outside of bounds (-128,127)" + (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (9223372036854775807) is outside of enumeration's range (-128,127)" (map fromEnum [(1::Int8),minBound,maxBound]) = [1,-128,127] (take 7 [(1::Int8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int8)-5)..]) = [122,123,124,125,126,127] @@ -36,12 +36,12 @@ Testing Enum Int8: Testing Enum Int16: (succ (0::Int16)) = 1 (succ (minBound::Int16)) = -32767 - (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of maxBound" + (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of last tag in enumeration" pred (1::Int16) = 0 pred (maxBound::Int16) = 32766 - pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of minBound" + pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int16) [1, fromIntegral (minBound::Int16), fromIntegral (maxBound::Int16)]) = [1,-32768,32767] - (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (9223372036854775807) is outside of bounds (-32768,32767)" + (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (9223372036854775807) is outside of enumeration's range (-32768,32767)" (map fromEnum [(1::Int16),minBound,maxBound]) = [1,-32768,32767] (take 7 [(1::Int16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int16)-5)..]) = [32762,32763,32764,32765,32766,32767] @@ -71,12 +71,12 @@ Testing Enum Int16: Testing Enum Int32: (succ (0::Int32)) = 1 (succ (minBound::Int32)) = -2147483647 - (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of maxBound" + (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of last tag in enumeration" pred (1::Int32) = 0 pred (maxBound::Int32) = 2147483646 - pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of minBound" + pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int32) [1, fromIntegral (minBound::Int32), fromIntegral (maxBound::Int32)]) = [1,-2147483648,2147483647] - (toEnum (maxBound::Int))::Int32 = error "Enum.toEnum{Int32}: tag (9223372036854775807) is outside of bounds (-2147483648,2147483647)" + (toEnum (maxBound::Int))::Int32 = error "Enum.toEnum{Int32}: tag (9223372036854775807) is outside of enumeration's range (-2147483648,2147483647)" (map fromEnum [(1::Int32),minBound,maxBound]) = [1,-2147483648,2147483647] (take 7 [(1::Int32)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int32)-5)..]) = [2147483642,2147483643,2147483644,2147483645,2147483646,2147483647] @@ -106,10 +106,10 @@ Testing Enum Int32: Testing Enum Int64: (succ (0::Int64)) = 1 (succ (minBound::Int64)) = -9223372036854775807 - (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of maxBound" + (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of last tag in enumeration" pred (1::Int64) = 0 pred (maxBound::Int64) = 9223372036854775806 - pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of minBound" + pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int64) [1, fromIntegral (minBound::Int64), fromIntegral (maxBound::Int64)]) = [1,-9223372036854775808,9223372036854775807] (toEnum (maxBound::Int))::Int64 = 9223372036854775807 (map fromEnum [(1::Int64),fromIntegral (minBound::Int) ,fromIntegral (maxBound::Int)]) = [1,-9223372036854775808,9223372036854775807] ===================================== libraries/base/tests/enum02.stdout-ws-64 ===================================== @@ -1,12 +1,12 @@ Testing Enum Int8: (succ (0::Int8)) = 1 (succ (minBound::Int8)) = -127 - (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of maxBound" + (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of last tag in enumeration" pred (1::Int8) = 0 pred (maxBound::Int8) = 126 - pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of minBound" + pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int8) [1, fromIntegral (minBound::Int8), fromIntegral (maxBound::Int8)]) = [1,-128,127] - (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (9223372036854775807) is outside of bounds (-128,127)" + (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (9223372036854775807) is outside of enumeration's range (-128,127)" (map fromEnum [(1::Int8),minBound,maxBound]) = [1,-128,127] (take 7 [(1::Int8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int8)-5)..]) = [122,123,124,125,126,127] @@ -36,12 +36,12 @@ Testing Enum Int8: Testing Enum Int16: (succ (0::Int16)) = 1 (succ (minBound::Int16)) = -32767 - (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of maxBound" + (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of last tag in enumeration" pred (1::Int16) = 0 pred (maxBound::Int16) = 32766 - pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of minBound" + pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int16) [1, fromIntegral (minBound::Int16), fromIntegral (maxBound::Int16)]) = [1,-32768,32767] - (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (9223372036854775807) is outside of bounds (-32768,32767)" + (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (9223372036854775807) is outside of enumeration's range (-32768,32767)" (map fromEnum [(1::Int16),minBound,maxBound]) = [1,-32768,32767] (take 7 [(1::Int16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int16)-5)..]) = [32762,32763,32764,32765,32766,32767] @@ -71,12 +71,12 @@ Testing Enum Int16: Testing Enum Int32: (succ (0::Int32)) = 1 (succ (minBound::Int32)) = -2147483647 - (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of maxBound" + (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of last tag in enumeration" pred (1::Int32) = 0 pred (maxBound::Int32) = 2147483646 - pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of minBound" + pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int32) [1, fromIntegral (minBound::Int32), fromIntegral (maxBound::Int32)]) = [1,-2147483648,2147483647] - (toEnum (maxBound::Int))::Int32 = error "Enum.toEnum{Int32}: tag (9223372036854775807) is outside of bounds (-2147483648,2147483647)" + (toEnum (maxBound::Int))::Int32 = error "Enum.toEnum{Int32}: tag (9223372036854775807) is outside of enumeration's range (-2147483648,2147483647)" (map fromEnum [(1::Int32),minBound,maxBound]) = [1,-2147483648,2147483647] (take 7 [(1::Int32)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int32)-5)..]) = [2147483642,2147483643,2147483644,2147483645,2147483646,2147483647] @@ -106,10 +106,10 @@ Testing Enum Int32: Testing Enum Int64: (succ (0::Int64)) = 1 (succ (minBound::Int64)) = -9223372036854775807 - (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of maxBound" + (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of last tag in enumeration" pred (1::Int64) = 0 pred (maxBound::Int64) = 9223372036854775806 - pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of minBound" + pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int64) [1, fromIntegral (minBound::Int64), fromIntegral (maxBound::Int64)]) = [1,-9223372036854775808,9223372036854775807] (toEnum (maxBound::Int))::Int64 = 9223372036854775807 (map fromEnum [(1::Int64),fromIntegral (minBound::Int) ,fromIntegral (maxBound::Int)]) = [1,-9223372036854775808,9223372036854775807] ===================================== libraries/base/tests/enum02.stdout-x86_64-unknown-openbsd ===================================== @@ -1,12 +1,12 @@ Testing Enum Int8: (succ (0::Int8)) = 1 (succ (minBound::Int8)) = -127 - (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of maxBound" + (succ (maxBound::Int8)) = error "Enum.succ{Int8}: tried to take `succ' of last tag in enumeration" pred (1::Int8) = 0 pred (maxBound::Int8) = 126 - pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of minBound" + pred (minBound::Int8) = error "Enum.pred{Int8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int8) [1, fromIntegral (minBound::Int8), fromIntegral (maxBound::Int8)]) = [1,-128,127] - (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (9223372036854775807) is outside of bounds (-128,127)" + (toEnum (maxBound::Int))::Int8 = error "Enum.toEnum{Int8}: tag (9223372036854775807) is outside of enumeration's range (-128,127)" (map fromEnum [(1::Int8),minBound,maxBound]) = [1,-128,127] (take 7 [(1::Int8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int8)-5)..]) = [122,123,124,125,126,127] @@ -36,12 +36,12 @@ Testing Enum Int8: Testing Enum Int16: (succ (0::Int16)) = 1 (succ (minBound::Int16)) = -32767 - (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of maxBound" + (succ (maxBound::Int16)) = error "Enum.succ{Int16}: tried to take `succ' of last tag in enumeration" pred (1::Int16) = 0 pred (maxBound::Int16) = 32766 - pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of minBound" + pred (minBound::Int16) = error "Enum.pred{Int16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int16) [1, fromIntegral (minBound::Int16), fromIntegral (maxBound::Int16)]) = [1,-32768,32767] - (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (9223372036854775807) is outside of bounds (-32768,32767)" + (toEnum (maxBound::Int))::Int16 = error "Enum.toEnum{Int16}: tag (9223372036854775807) is outside of enumeration's range (-32768,32767)" (map fromEnum [(1::Int16),minBound,maxBound]) = [1,-32768,32767] (take 7 [(1::Int16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int16)-5)..]) = [32762,32763,32764,32765,32766,32767] @@ -71,12 +71,12 @@ Testing Enum Int16: Testing Enum Int32: (succ (0::Int32)) = 1 (succ (minBound::Int32)) = -2147483647 - (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of maxBound" + (succ (maxBound::Int32)) = error "Enum.succ{Int32}: tried to take `succ' of last tag in enumeration" pred (1::Int32) = 0 pred (maxBound::Int32) = 2147483646 - pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of minBound" + pred (minBound::Int32) = error "Enum.pred{Int32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int32) [1, fromIntegral (minBound::Int32), fromIntegral (maxBound::Int32)]) = [1,-2147483648,2147483647] - (toEnum (maxBound::Int))::Int32 = error "Enum.toEnum{Int32}: tag (9223372036854775807) is outside of bounds (-2147483648,2147483647)" + (toEnum (maxBound::Int))::Int32 = error "Enum.toEnum{Int32}: tag (9223372036854775807) is outside of enumeration's range (-2147483648,2147483647)" (map fromEnum [(1::Int32),minBound,maxBound]) = [1,-2147483648,2147483647] (take 7 [(1::Int32)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Int32)-5)..]) = [2147483642,2147483643,2147483644,2147483645,2147483646,2147483647] @@ -106,10 +106,10 @@ Testing Enum Int32: Testing Enum Int64: (succ (0::Int64)) = 1 (succ (minBound::Int64)) = -9223372036854775807 - (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of maxBound" + (succ (maxBound::Int64)) = error "Enum.succ{Int64}: tried to take `succ' of last tag in enumeration" pred (1::Int64) = 0 pred (maxBound::Int64) = 9223372036854775806 - pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of minBound" + pred (minBound::Int64) = error "Enum.pred{Int64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Int64) [1, fromIntegral (minBound::Int64), fromIntegral (maxBound::Int64)]) = [1,-9223372036854775808,9223372036854775807] (toEnum (maxBound::Int))::Int64 = 9223372036854775807 (map fromEnum [(1::Int64),fromIntegral (minBound::Int) ,fromIntegral (maxBound::Int)]) = [1,-9223372036854775808,9223372036854775807] ===================================== libraries/base/tests/enum03.stdout ===================================== @@ -1,12 +1,12 @@ Testing Enum Word8: (succ (0::Word8)) = 1 (succ (minBound::Word8)) = 1 - (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of maxBound" + (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of last tag in enumeration" pred (1::Word8) = 0 pred (maxBound::Word8) = 254 - pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of minBound" + pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word8) [1, fromIntegral (minBound::Word8)::Int, fromIntegral (maxBound::Word8)::Int]) = [1,0,255] - (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (2147483647) is outside of bounds (0,255)" + (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (2147483647) is outside of enumeration's range (0,255)" (map fromEnum [(1::Word8),minBound,maxBound]) = [1,0,255] (take 7 [(1::Word8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word8)-5)..]) = [250,251,252,253,254,255] @@ -36,12 +36,12 @@ Testing Enum Word8: Testing Enum Word16: (succ (0::Word16)) = 1 (succ (minBound::Word16)) = 1 - (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of maxBound" + (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of last tag in enumeration" pred (1::Word16) = 0 pred (maxBound::Word16) = 65534 - pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of minBound" + pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word16) [1, fromIntegral (minBound::Word16)::Int, fromIntegral (maxBound::Word16)::Int]) = [1,0,65535] - (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (2147483647) is outside of bounds (0,65535)" + (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (2147483647) is outside of enumeration's range (0,65535)" (map fromEnum [(1::Word16),minBound,maxBound]) = [1,0,65535] (take 7 [(1::Word16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word16)-5)..]) = [65530,65531,65532,65533,65534,65535] @@ -71,10 +71,10 @@ Testing Enum Word16: Testing Enum Word32: (succ (0::Word32)) = 1 (succ (minBound::Word32)) = 1 - (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of maxBound" + (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of last tag in enumeration" pred (1::Word32) = 0 pred (maxBound::Word32) = 4294967294 - pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of minBound" + pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word32) [1, fromIntegral (minBound::Word32)::Int, fromIntegral (maxBound::Int32)::Int]) = [1,0,2147483647] (toEnum (maxBound::Int))::Word32 = 2147483647 (map fromEnum [(1::Word32),minBound,fromIntegral (maxBound::Int)]) = [1,0,2147483647] @@ -107,10 +107,10 @@ Testing Enum Word32: Testing Enum Word64: (succ (0::Word64)) = 1 (succ (minBound::Word64)) = 1 - (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of maxBound" + (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of last tag in enumeration" pred (1::Word64) = 0 pred (maxBound::Word64) = 18446744073709551614 - pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of minBound" + pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word64) [1, fromIntegral (minBound::Word64)::Int, maxBound::Int]) = [1,0,2147483647] (toEnum (maxBound::Int))::Word64 = 2147483647 (map fromEnum [(1::Word64),minBound,fromIntegral (maxBound::Int)]) = [1,0,2147483647] ===================================== libraries/base/tests/enum03.stdout-alpha-dec-osf3 ===================================== @@ -1,12 +1,12 @@ Testing Enum Word8: (succ (0::Word8)) = 1 (succ (minBound::Word8)) = 1 - (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of maxBound" + (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of last tag in enumeration" pred (1::Word8) = 0 pred (maxBound::Word8) = 254 - pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of minBound" + pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word8) [1, fromIntegral (minBound::Word8)::Int, fromIntegral (maxBound::Word8)::Int]) = [1,0,255] - (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (9223372036854775807) is outside of bounds (0,255)" + (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (9223372036854775807) is outside of enumeration's range (0,255)" (map fromEnum [(1::Word8),minBound,maxBound]) = [1,0,255] (take 7 [(1::Word8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word8)-5)..]) = [250,251,252,253,254,255] @@ -36,12 +36,12 @@ Testing Enum Word8: Testing Enum Word16: (succ (0::Word16)) = 1 (succ (minBound::Word16)) = 1 - (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of maxBound" + (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of last tag in enumeration" pred (1::Word16) = 0 pred (maxBound::Word16) = 65534 - pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of minBound" + pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word16) [1, fromIntegral (minBound::Word16)::Int, fromIntegral (maxBound::Word16)::Int]) = [1,0,65535] - (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (9223372036854775807) is outside of bounds (0,65535)" + (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (9223372036854775807) is outside of enumeration's range (0,65535)" (map fromEnum [(1::Word16),minBound,maxBound]) = [1,0,65535] (take 7 [(1::Word16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word16)-5)..]) = [65530,65531,65532,65533,65534,65535] @@ -71,12 +71,12 @@ Testing Enum Word16: Testing Enum Word32: (succ (0::Word32)) = 1 (succ (minBound::Word32)) = 1 - (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of maxBound" + (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of last tag in enumeration" pred (1::Word32) = 0 pred (maxBound::Word32) = 4294967294 - pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of minBound" + pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word32) [1, fromIntegral (minBound::Word32)::Int, fromIntegral (maxBound::Int32)::Int]) = [1,0,2147483647] - (toEnum (maxBound::Int))::Word32 = error "Enum.toEnum{Word32}: tag (9223372036854775807) is outside of bounds (0,4294967295)" + (toEnum (maxBound::Int))::Word32 = error "Enum.toEnum{Word32}: tag (9223372036854775807) is outside of enumeration's range (0,4294967295)" (map fromEnum [(1::Word32),minBound,fromIntegral (maxBound::Int)]) = [1,0,4294967295] fromEnum (maxBound::Word32) = 4294967295 (take 7 [(1::Word32)..]) = [1,2,3,4,5,6,7] @@ -107,10 +107,10 @@ Testing Enum Word32: Testing Enum Word64: (succ (0::Word64)) = 1 (succ (minBound::Word64)) = 1 - (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of maxBound" + (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of last tag in enumeration" pred (1::Word64) = 0 pred (maxBound::Word64) = 18446744073709551614 - pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of minBound" + pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word64) [1, fromIntegral (minBound::Word64)::Int, maxBound::Int]) = [1,0,9223372036854775807] (toEnum (maxBound::Int))::Word64 = 9223372036854775807 (map fromEnum [(1::Word64),minBound,fromIntegral (maxBound::Int)]) = [1,0,9223372036854775807] ===================================== libraries/base/tests/enum03.stdout-mips-sgi-irix ===================================== @@ -1,12 +1,12 @@ Testing Enum Word8: (succ (0::Word8)) = 1 (succ (minBound::Word8)) = 1 - (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of maxBound" + (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of last tag in enumeration" pred (1::Word8) = 0 pred (maxBound::Word8) = 254 - pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of minBound" + pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word8) [1, fromIntegral (minBound::Word8)::Int, fromIntegral (maxBound::Word8)::Int]) = [1,0,255] - (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (9223372036854775807) is outside of bounds (0,255)" + (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (9223372036854775807) is outside of enumeration's range (0,255)" (map fromEnum [(1::Word8),minBound,maxBound]) = [1,0,255] (take 7 [(1::Word8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word8)-5)..]) = [250,251,252,253,254,255] @@ -36,12 +36,12 @@ Testing Enum Word8: Testing Enum Word16: (succ (0::Word16)) = 1 (succ (minBound::Word16)) = 1 - (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of maxBound" + (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of last tag in enumeration" pred (1::Word16) = 0 pred (maxBound::Word16) = 65534 - pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of minBound" + pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word16) [1, fromIntegral (minBound::Word16)::Int, fromIntegral (maxBound::Word16)::Int]) = [1,0,65535] - (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (9223372036854775807) is outside of bounds (0,65535)" + (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (9223372036854775807) is outside of enumeration's range (0,65535)" (map fromEnum [(1::Word16),minBound,maxBound]) = [1,0,65535] (take 7 [(1::Word16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word16)-5)..]) = [65530,65531,65532,65533,65534,65535] @@ -71,12 +71,12 @@ Testing Enum Word16: Testing Enum Word32: (succ (0::Word32)) = 1 (succ (minBound::Word32)) = 1 - (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of maxBound" + (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of last tag in enumeration" pred (1::Word32) = 0 pred (maxBound::Word32) = 4294967294 - pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of minBound" + pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word32) [1, fromIntegral (minBound::Word32)::Int, fromIntegral (maxBound::Int32)::Int]) = [1,0,2147483647] - (toEnum (maxBound::Int))::Word32 = error "Enum.toEnum{Word32}: tag (9223372036854775807) is outside of bounds (0,4294967295)" + (toEnum (maxBound::Int))::Word32 = error "Enum.toEnum{Word32}: tag (9223372036854775807) is outside of enumeration's range (0,4294967295)" (map fromEnum [(1::Word32),minBound,fromIntegral (maxBound::Int)]) = [1,0,4294967295] fromEnum (maxBound::Word32) = 4294967295 (take 7 [(1::Word32)..]) = [1,2,3,4,5,6,7] @@ -107,10 +107,10 @@ Testing Enum Word32: Testing Enum Word64: (succ (0::Word64)) = 1 (succ (minBound::Word64)) = 1 - (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of maxBound" + (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of last tag in enumeration" pred (1::Word64) = 0 pred (maxBound::Word64) = 18446744073709551614 - pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of minBound" + pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word64) [1, fromIntegral (minBound::Word64)::Int, maxBound::Int]) = [1,0,9223372036854775807] (toEnum (maxBound::Int))::Word64 = 9223372036854775807 (map fromEnum [(1::Word64),minBound,fromIntegral (maxBound::Int)]) = [1,0,9223372036854775807] ===================================== libraries/base/tests/enum03.stdout-ws-64 ===================================== @@ -1,12 +1,12 @@ Testing Enum Word8: (succ (0::Word8)) = 1 (succ (minBound::Word8)) = 1 - (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of maxBound" + (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of last tag in enumeration" pred (1::Word8) = 0 pred (maxBound::Word8) = 254 - pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of minBound" + pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word8) [1, fromIntegral (minBound::Word8)::Int, fromIntegral (maxBound::Word8)::Int]) = [1,0,255] - (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (9223372036854775807) is outside of bounds (0,255)" + (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (9223372036854775807) is outside of enumeration's range (0,255)" (map fromEnum [(1::Word8),minBound,maxBound]) = [1,0,255] (take 7 [(1::Word8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word8)-5)..]) = [250,251,252,253,254,255] @@ -36,12 +36,12 @@ Testing Enum Word8: Testing Enum Word16: (succ (0::Word16)) = 1 (succ (minBound::Word16)) = 1 - (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of maxBound" + (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of last tag in enumeration" pred (1::Word16) = 0 pred (maxBound::Word16) = 65534 - pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of minBound" + pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word16) [1, fromIntegral (minBound::Word16)::Int, fromIntegral (maxBound::Word16)::Int]) = [1,0,65535] - (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (9223372036854775807) is outside of bounds (0,65535)" + (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (9223372036854775807) is outside of enumeration's range (0,65535)" (map fromEnum [(1::Word16),minBound,maxBound]) = [1,0,65535] (take 7 [(1::Word16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word16)-5)..]) = [65530,65531,65532,65533,65534,65535] @@ -71,12 +71,12 @@ Testing Enum Word16: Testing Enum Word32: (succ (0::Word32)) = 1 (succ (minBound::Word32)) = 1 - (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of maxBound" + (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of last tag in enumeration" pred (1::Word32) = 0 pred (maxBound::Word32) = 4294967294 - pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of minBound" + pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word32) [1, fromIntegral (minBound::Word32)::Int, fromIntegral (maxBound::Int32)::Int]) = [1,0,2147483647] - (toEnum (maxBound::Int))::Word32 = error "Enum.toEnum{Word32}: tag (9223372036854775807) is outside of bounds (0,4294967295)" + (toEnum (maxBound::Int))::Word32 = error "Enum.toEnum{Word32}: tag (9223372036854775807) is outside of enumeration's range (0,4294967295)" (map fromEnum [(1::Word32),minBound,fromIntegral (maxBound::Int)]) = [1,0,4294967295] fromEnum (maxBound::Word32) = 4294967295 (take 7 [(1::Word32)..]) = [1,2,3,4,5,6,7] @@ -107,10 +107,10 @@ Testing Enum Word32: Testing Enum Word64: (succ (0::Word64)) = 1 (succ (minBound::Word64)) = 1 - (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of maxBound" + (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of last tag in enumeration" pred (1::Word64) = 0 pred (maxBound::Word64) = 18446744073709551614 - pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of minBound" + pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word64) [1, fromIntegral (minBound::Word64)::Int, maxBound::Int]) = [1,0,9223372036854775807] (toEnum (maxBound::Int))::Word64 = 9223372036854775807 (map fromEnum [(1::Word64),minBound,fromIntegral (maxBound::Int)]) = [1,0,9223372036854775807] ===================================== libraries/base/tests/enum03.stdout-x86_64-unknown-openbsd ===================================== @@ -1,12 +1,12 @@ Testing Enum Word8: (succ (0::Word8)) = 1 (succ (minBound::Word8)) = 1 - (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of maxBound" + (succ (maxBound::Word8)) = error "Enum.succ{Word8}: tried to take `succ' of last tag in enumeration" pred (1::Word8) = 0 pred (maxBound::Word8) = 254 - pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of minBound" + pred (minBound::Word8) = error "Enum.pred{Word8}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word8) [1, fromIntegral (minBound::Word8)::Int, fromIntegral (maxBound::Word8)::Int]) = [1,0,255] - (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (9223372036854775807) is outside of bounds (0,255)" + (toEnum (maxBound::Int))::Word8 = error "Enum.toEnum{Word8}: tag (9223372036854775807) is outside of enumeration's range (0,255)" (map fromEnum [(1::Word8),minBound,maxBound]) = [1,0,255] (take 7 [(1::Word8)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word8)-5)..]) = [250,251,252,253,254,255] @@ -36,12 +36,12 @@ Testing Enum Word8: Testing Enum Word16: (succ (0::Word16)) = 1 (succ (minBound::Word16)) = 1 - (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of maxBound" + (succ (maxBound::Word16)) = error "Enum.succ{Word16}: tried to take `succ' of last tag in enumeration" pred (1::Word16) = 0 pred (maxBound::Word16) = 65534 - pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of minBound" + pred (minBound::Word16) = error "Enum.pred{Word16}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word16) [1, fromIntegral (minBound::Word16)::Int, fromIntegral (maxBound::Word16)::Int]) = [1,0,65535] - (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (9223372036854775807) is outside of bounds (0,65535)" + (toEnum (maxBound::Int))::Word16 = error "Enum.toEnum{Word16}: tag (9223372036854775807) is outside of enumeration's range (0,65535)" (map fromEnum [(1::Word16),minBound,maxBound]) = [1,0,65535] (take 7 [(1::Word16)..]) = [1,2,3,4,5,6,7] (take 7 [((maxBound::Word16)-5)..]) = [65530,65531,65532,65533,65534,65535] @@ -71,12 +71,12 @@ Testing Enum Word16: Testing Enum Word32: (succ (0::Word32)) = 1 (succ (minBound::Word32)) = 1 - (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of maxBound" + (succ (maxBound::Word32)) = error "Enum.succ{Word32}: tried to take `succ' of last tag in enumeration" pred (1::Word32) = 0 pred (maxBound::Word32) = 4294967294 - pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of minBound" + pred (minBound::Word32) = error "Enum.pred{Word32}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word32) [1, fromIntegral (minBound::Word32)::Int, fromIntegral (maxBound::Int32)::Int]) = [1,0,2147483647] - (toEnum (maxBound::Int))::Word32 = error "Enum.toEnum{Word32}: tag (9223372036854775807) is outside of bounds (0,4294967295)" + (toEnum (maxBound::Int))::Word32 = error "Enum.toEnum{Word32}: tag (9223372036854775807) is outside of enumeration's range (0,4294967295)" (map fromEnum [(1::Word32),minBound,fromIntegral (maxBound::Int)]) = [1,0,4294967295] fromEnum (maxBound::Word32) = 4294967295 (take 7 [(1::Word32)..]) = [1,2,3,4,5,6,7] @@ -107,10 +107,10 @@ Testing Enum Word32: Testing Enum Word64: (succ (0::Word64)) = 1 (succ (minBound::Word64)) = 1 - (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of maxBound" + (succ (maxBound::Word64)) = error "Enum.succ{Word64}: tried to take `succ' of last tag in enumeration" pred (1::Word64) = 0 pred (maxBound::Word64) = 18446744073709551614 - pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of minBound" + pred (minBound::Word64) = error "Enum.pred{Word64}: tried to take `pred' of first tag in enumeration" (map (toEnum::Int->Word64) [1, fromIntegral (minBound::Word64)::Int, maxBound::Int]) = [1,0,9223372036854775807] (toEnum (maxBound::Int))::Word64 = 9223372036854775807 (map fromEnum [(1::Word64),minBound,fromIntegral (maxBound::Int)]) = [1,0,9223372036854775807] ===================================== libraries/ghc-internal/codepages/MakeTable.hs ===================================== @@ -33,7 +33,7 @@ main :: IO () main = do moduleName:outFile:files <- getArgs let badFiles = -- These fail with an error like - -- MakeTable: Enum.toEnum{Word8}: tag (33088) is outside of bounds (0,255) + -- MakeTable: Enum.toEnum{Word8}: tag (33088) is outside of enumeration's range (0,255) -- I have no idea what's going on, so for now we just -- skip them. ["CPs/CP932.TXT", ===================================== libraries/ghc-internal/src/GHC/Internal/Enum.hs ===================================== @@ -22,14 +22,20 @@ #include "MachDeps.h" -module GHC.Internal.Enum( - Bounded(..), Enum(..), - boundedEnumFrom, boundedEnumFromThen, - toEnumError, fromEnumError, succError, predError, - - -- Instances for Bounded and Enum: (), Char, Int - - ) where +module GHC.Internal.Enum + ( Bounded(..) + , Enum(..) + , boundedEnumFrom + , boundedEnumFromThen + , toEnumError + , fromEnumError + , succError + , predError + , enumIntToWord + + -- Instances for Bounded and Enum: (), Char, Int + ) +where import GHC.Internal.Base hiding ( many ) import GHC.Internal.Char @@ -246,11 +252,11 @@ Alternatives might be {-# NOINLINE toEnumError #-} toEnumError :: (Show a) => String -> Int -> (a,a) -> b -toEnumError inst_ty i bnds = +toEnumError inst_ty i (mi,ma) = errorWithoutStackTrace $ "Enum.toEnum{" ++ inst_ty ++ "}: tag (" ++ show i ++ - ") is outside of bounds " ++ - show bnds + ") is outside of enumeration's range (" ++ + show mi ++ "," ++ show ma ++ ")" {-# NOINLINE fromEnumError #-} fromEnumError :: (Show a) => String -> a -> b @@ -263,12 +269,12 @@ fromEnumError inst_ty x = {-# NOINLINE succError #-} succError :: String -> a succError inst_ty = - errorWithoutStackTrace $ "Enum.succ{" ++ inst_ty ++ "}: tried to take `succ' of maxBound" + errorWithoutStackTrace $ "Enum.succ{" ++ inst_ty ++ "}: tried to take `succ' of last tag in enumeration" {-# NOINLINE predError #-} predError :: String -> a predError inst_ty = - errorWithoutStackTrace $ "Enum.pred{" ++ inst_ty ++ "}: tried to take `pred' of minBound" + errorWithoutStackTrace $ "Enum.pred{" ++ inst_ty ++ "}: tried to take `pred' of first tag in enumeration" ------------------------------------------------------------------------ -- Tuples @@ -565,10 +571,10 @@ instance Bounded Int where -- | @since base-2.01 instance Enum Int where succ x - | x == maxBound = errorWithoutStackTrace "Prelude.Enum.succ{Int}: tried to take `succ' of maxBound" + | x == maxBound = succError "Int" | otherwise = x + 1 pred x - | x == minBound = errorWithoutStackTrace "Prelude.Enum.pred{Int}: tried to take `pred' of minBound" + | x == minBound = predError "Int" | otherwise = x - 1 toEnum x = x @@ -1088,6 +1094,9 @@ enumNegDeltaToNatural x0 ndelta lim = go x0 | x >= ndelta = x : go (x-ndelta) | otherwise = [x] +-- | Convert an Int into a Word (used in derived Enum instances) +enumIntToWord :: Int -> Word +enumIntToWord (I# i) = W# (int2Word# i) -- Instances from GHC.Types View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c20d51867c824e32c61bd1e002680bef268e4f51...e9fa116326c56859e26c6a713788a756c1318cda -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c20d51867c824e32c61bd1e002680bef268e4f51...e9fa116326c56859e26c6a713788a756c1318cda You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 08:17:58 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 04:17:58 -0400 Subject: [Git][ghc/ghc][master] Core: add absorb rules for binary or/and (#16351) Message-ID: <66f518b64fbe1_3f975ae2c4810889b@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 4 changed files: - compiler/GHC/Core/Opt/ConstantFold.hs - + testsuite/tests/codeGen/should_compile/T16351.hs - + testsuite/tests/codeGen/should_compile/T16351.stderr - testsuite/tests/codeGen/should_compile/all.T Changes: ===================================== compiler/GHC/Core/Opt/ConstantFold.hs ===================================== @@ -3056,6 +3056,11 @@ andFoldingRules' platform arg1 arg2 num_ops = case (arg1, arg2) of -- (l1 or x) and (l2 or y) ==> (l1 and l2) or (x and l2) or (l1 and y) or (x and y) -- increase operation numbers + -- x and (y or ... or x or ... or z) ==> x + (x, is_or_list num_ops -> Just xs) + | any (cheapEqExpr x) xs + -> Just x + _ -> Nothing where mkL = Lit . mkNumLiteral platform num_ops @@ -3079,6 +3084,11 @@ orFoldingRules' platform arg1 arg2 num_ops = case (arg1, arg2) of -- (l1 and x) or (l2 and y) ==> (l1 and l2) or (x and l2) or (l1 and y) or (x and y) -- increase operation numbers + -- x or (y and ... and x and ... and z) ==> x + (x, is_and_list num_ops -> Just xs) + | any (cheapEqExpr x) xs + -> Just x + _ -> Nothing where mkL = Lit . mkNumLiteral platform num_ops @@ -3117,7 +3127,7 @@ is_op op e = case e of App (OpVal op') x | op == op' -> Just x _ -> Nothing -is_add, is_sub, is_mul, is_and, is_or, is_div :: NumOps -> CoreExpr -> Maybe (Arg CoreBndr, Arg CoreBndr) +is_add, is_sub, is_mul, is_and, is_or, is_div :: NumOps -> CoreExpr -> Maybe (CoreArg, CoreArg) is_add num_ops e = is_binop (numAdd num_ops) e is_sub num_ops e = is_binop (numSub num_ops) e is_mul num_ops e = is_binop (numMul num_ops) e @@ -3128,6 +3138,25 @@ is_div num_ops e = numDiv num_ops >>= \op -> is_binop op e is_neg :: NumOps -> CoreExpr -> Maybe (Arg CoreBndr) is_neg num_ops e = numNeg num_ops >>= \op -> is_op op e +-- Return a list of operands for a given operation. +-- E.e. is_and_list (a and ... and z) => [a,...,z] for any nesting of the and +-- operation +is_list :: (CoreExpr -> Maybe (CoreArg,CoreArg)) -> CoreExpr -> Maybe [CoreArg] +is_list f e_org = case f e_org of -- do we have the operator at all? + Just (a,b) -> Just (go [a,b]) + Nothing -> Nothing + where + go = \case + [] -> [] + (e:es) -> case f e of + -- we can't split any more: add to the result list + Nothing -> e : go es + Just (a,b) -> go (a:b:es) + +is_and_list, is_or_list :: NumOps -> CoreExpr -> Maybe [CoreArg] +is_and_list ops = is_list (is_and ops) +is_or_list ops = is_list (is_or ops) + -- match operation with a literal (handles commutativity) is_lit_add, is_lit_mul, is_lit_and, is_lit_or :: NumOps -> CoreExpr -> Maybe (Integer, Arg CoreBndr) is_lit_add num_ops e = is_lit' is_add num_ops e ===================================== testsuite/tests/codeGen/should_compile/T16351.hs ===================================== @@ -0,0 +1,23 @@ +module T16351 where + +import Data.Bits + +x1,x2,x3,x4,x5,x6,x7,x8 :: Int -> Int -> Int + +x1 a b = a .&. (a .|. b) +x2 a b = a .|. (a .&. b) +x3 a b = a .&. (b .|. a) +x4 a b = a .|. (b .&. a) +x5 a b = (a .|. b) .&. a +x6 a b = (a .&. b) .|. a +x7 a b = (b .|. a) .&. a +x8 a b = (b .&. a) .|. a + +-- add more nesting +x10,x11,x12,x13,x14 :: Int -> Int -> Int -> Int + +x10 a b c = a .&. ((a .|. b) .|. c) +x11 a b c = a .&. (c .|. (a .|. b)) +x12 a b c = a .&. (c .|. (b .|. a)) +x13 a b c = a .&. ((b .|. a) .|. c) +x14 a b c = a .&. ((c .|. (b .|. a)) .|. c) ===================================== testsuite/tests/codeGen/should_compile/T16351.stderr ===================================== @@ -0,0 +1,42 @@ + +==================== Tidy Core ==================== +Result size of Tidy Core + = {terms: 63, types: 72, coercions: 0, joins: 0/0} + +x14 + = \ a b c -> + case a of wild { I# x# -> + case c of { I# x#1 -> case b of { I# x#2 -> wild } } + } + +x13 + = \ a b c -> + case a of wild { I# x# -> + case b of { I# x#1 -> case c of { I# y# -> wild } } + } + +x12 = x14 + +x11 = x14 + +x10 = x13 + +x8 = \ a b -> case b of { I# x# -> a } + +x7 = x8 + +x6 + = \ a b -> case a of wild { I# x# -> case b of { I# y# -> wild } } + +x5 = x6 + +x4 = x6 + +x3 = x6 + +x2 = x6 + +x1 = x6 + + + ===================================== testsuite/tests/codeGen/should_compile/all.T ===================================== @@ -143,3 +143,5 @@ test('T25166', [req_cmm], makefile_test, []) # dump Core to ensure that d is defined as: d = D 10## RUBBISH(IntRep) test('T25177', normal, compile, ['-O2 -dno-typeable-binds -ddump-simpl -dsuppress-all -dsuppress-uniques -v0']) + +test('T16351', normal, compile, ['-O2 -ddump-simpl -dno-typeable-binds -dsuppress-all -dsuppress-uniques']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/383af0743a46e0961c342261f13a4527ae838873 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/383af0743a46e0961c342261f13a4527ae838873 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 08:39:38 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 04:39:38 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] wip fixup Message-ID: <66f51dcadf963_3f975a9a96d811389d@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: f3e44e88 by Matthew Pickering at 2024-09-26T09:39:30+01:00 wip fixup - - - - - 1 changed file: - hadrian/src/Settings/Builders/RunTest.hs Changes: ===================================== hadrian/src/Settings/Builders/RunTest.hs ===================================== @@ -114,7 +114,7 @@ inTreeCompilerArgs stg = do wordsize <- show @Int . (*8) <$> queryTargetTarget ghcStage (wordSize2Bytes . tgtWordSize) llc_cmd <- settingsFileSetting ToolchainSetting_LlcCommand ghcStage - llvm_as_cmd <- settingsFileSetting ToolchainSetting_LlvmAsCommand ghcStage + _llvm_as_cmd <- settingsFileSetting ToolchainSetting_LlvmAsCommand ghcStage have_llvm <- liftIO (isJust <$> findExecutable llc_cmd) top <- topDirectory View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f3e44e88a241706a07b9ad363f409eba5f3f34d3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f3e44e88a241706a07b9ad363f409eba5f3f34d3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 08:42:03 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 04:42:03 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] supersven Message-ID: <66f51e5ba7e2e_560fcbc4a845089@gitlab.mail> Matthew Pickering pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 74332393 by Matthew Pickering at 2024-09-26T09:41:43+01:00 supersven - - - - - 1 changed file: - .gitlab-ci.yml Changes: ===================================== .gitlab-ci.yml ===================================== @@ -1,7 +1,7 @@ variables: GIT_SSL_NO_VERIFY: "1" - # Commit of ghc/ci-images repository from which to pull Docker images + # Commit of supersven/ci-images repository from which to pull Docker images DOCKER_REV: bd0e387ad7f67e16b79a7d940ca65452a6d20c04 # Sequential version number of all cached things. @@ -102,9 +102,9 @@ workflow: .bootstrap_matrix : &bootstrap_matrix matrix: - GHC_VERSION: 9.8.1 - DOCKER_IMAGE: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12-ghc9_8:$DOCKER_REV" + DOCKER_IMAGE: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12-ghc9_8:$DOCKER_REV" - GHC_VERSION: 9.10.1 - DOCKER_IMAGE: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12-ghc9_10:$DOCKER_REV" + DOCKER_IMAGE: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12-ghc9_10:$DOCKER_REV" # Allow linters to fail on draft MRs. # This must be explicitly transcluded in lint jobs which @@ -216,7 +216,7 @@ include: '.gitlab/jobs.yaml' ghc-linters: stage: tool-lint - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" extends: .lint-params variables: BUILD_FLAVOUR: default @@ -230,7 +230,7 @@ ghc-linters: # Run mypy Python typechecker on linter scripts. lint-linters: - image: "registry.gitlab.haskell.org/ghc/ci-images/linters:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/linters:$DOCKER_REV" extends: .lint script: - mypy testsuite/tests/linters/regex-linters/*.py @@ -238,7 +238,7 @@ lint-linters: # Check that .T files all parse by listing broken tests. lint-testsuite: - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb9:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb9:$DOCKER_REV" extends: .lint script: - make -Ctestsuite list_broken TEST_HC=$GHC @@ -246,7 +246,7 @@ lint-testsuite: # Run mypy Python typechecker on testsuite driver typecheck-testsuite: - image: "registry.gitlab.haskell.org/ghc/ci-images/linters:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/linters:$DOCKER_REV" extends: .lint script: - mypy testsuite/driver/runtests.py @@ -256,7 +256,7 @@ typecheck-testsuite: # accommodate, e.g., haddock changes not yet upstream) but not on `master` or # Marge jobs. .lint-submods: - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" extends: .lint-params variables: BUILD_FLAVOUR: default @@ -274,7 +274,7 @@ typecheck-testsuite: # accommodate, e.g., haddock changes not yet upstream) but not on `master` or # Marge jobs. lint-author: - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" extends: .lint-params variables: BUILD_FLAVOUR: default @@ -349,7 +349,7 @@ lint-submods-branch: .lint-params: needs: [] - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" extends: .lint before_script: - export PATH="/opt/toolchain/bin:$PATH" @@ -369,7 +369,7 @@ lint-submods-branch: # Disabled due to #22830 .hlint-ghc-and-base: extends: .lint-params - image: "registry.gitlab.haskell.org/ghc/ci-images/linters:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/linters:$DOCKER_REV" variables: BUILD_FLAVOUR: default script: @@ -390,7 +390,7 @@ hadrian-ghc-in-ghci: - job: lint-linters - job: lint-submods optional: true - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" before_script: # workaround for docker permissions - sudo chown ghc:ghc -R . @@ -430,7 +430,7 @@ hadrian-multi: - job: release-x86_64-linux-fedora33-release optional: true dependencies: null - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV" before_script: # workaround for docker permissions - sudo chown ghc:ghc -R . @@ -516,7 +516,7 @@ abi-test-nightly: - job: nightly-x86_64-linux-fedora33-release tags: - x86_64-linux - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV" dependencies: null before_script: - mkdir -p normal @@ -557,7 +557,7 @@ doc-tarball: tags: - x86_64-linux - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" dependencies: null variables: LINUX_BINDIST: "ghc-x86_64-linux-deb12.tar.xz" @@ -606,7 +606,7 @@ hackage-doc-tarball: - job: source-tarball tags: - x86_64-linux - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" dependencies: null variables: # Don't clone the git repo.. @@ -634,7 +634,7 @@ source-tarball: stage: full-build tags: - x86_64-linux - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" dependencies: [] artifacts: paths: @@ -676,7 +676,7 @@ package-hadrian-bootstrap-sources: tags: - x86_64-linux needs: ["generate-hadrian-bootstrap-sources"] - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" artifacts: paths: - hadrian-bootstrap-sources-all.tar.gz @@ -934,7 +934,7 @@ perf-nofib: optional: true - job: release-x86_64-linux-fedora33-release optional: true - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV" rules: - when: never - *full-ci @@ -978,7 +978,7 @@ perf: - job: release-x86_64-linux-fedora33-release optional: true dependencies: null - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV" tags: - x86_64-linux-perf script: @@ -1016,7 +1016,7 @@ abi-test: - job: release-x86_64-linux-fedora33-release optional: true dependencies: null - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV" rules: - if: $CI_MERGE_REQUEST_ID - if: '$CI_COMMIT_BRANCH == "master"' @@ -1086,7 +1086,7 @@ pages: project-version: stage: packaging - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV" tags: - x86_64-linux variables: @@ -1186,7 +1186,7 @@ ghcup-metadata-nightly: # Update the ghcup metadata with information about this nightly pipeline ghcup-metadata-nightly-push: stage: deploy - image: "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV" + image: "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV" dependencies: null tags: - x86_64-linux View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7433239356bafe88d650ec08395c41bdf17d39f7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7433239356bafe88d650ec08395c41bdf17d39f7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 08:47:45 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 04:47:45 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 16 commits: base: Propagate `error` CallStack to thrown exception Message-ID: <66f51fb0f265c_560fcbc96c47269@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - e9be7e0e by Matthew Pickering at 2024-09-26T04:47:27-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 56f82e38 by Matthew Pickering at 2024-09-26T04:47:27-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - eb77ca40 by Matthew Pickering at 2024-09-26T04:47:27-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 56d9c035 by Matthew Pickering at 2024-09-26T04:47:27-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - de860cdc by Brandon Chinn at 2024-09-26T04:47:28-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 116ba727 by Matthew Pickering at 2024-09-26T04:47:28-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - a17b05bf by Matthew Pickering at 2024-09-26T04:47:28-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 0b81ff07 by Matthew Pickering at 2024-09-26T04:47:28-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - e15e02ba by Matthew Pickering at 2024-09-26T04:47:28-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - 36ca323e by Matthew Pickering at 2024-09-26T04:47:29-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - 8b6bc6c5 by Matthew Pickering at 2024-09-26T04:47:29-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 30 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Core/Opt/ConstantFold.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Header.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - compiler/GHC/Rename/Names.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Types/Error/Codes.hs - compiler/GHC/Unit/State.hs - compiler/GHC/Unit/Types.hs - docs/users_guide/9.12.1-notes.rst - docs/users_guide/exts/rebindable_syntax.rst - hadrian/doc/flavours.md - hadrian/src/Flavour.hs - hadrian/src/Settings/Flavours/Release.hs - libraries/base/base.cabal.in - libraries/base/changelog.md - libraries/base/src/Control/Monad/Zip.hs - libraries/base/src/Data/List/NonEmpty.hs - libraries/base/tests/T19288.stderr - + libraries/base/tests/T24807.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/86bdd4288093f9ed4a1ad4684b92c5740208c4f3...8b6bc6c53e1858e59bd99f12eefc5ebfd3cfa330 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/86bdd4288093f9ed4a1ad4684b92c5740208c4f3...8b6bc6c53e1858e59bd99f12eefc5ebfd3cfa330 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 09:23:34 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 05:23:34 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Run job on label Message-ID: <66f52816315c4_bfb19cd1544913f@gitlab.mail> Matthew Pickering pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: fe79abc2 by Matthew Pickering at 2024-09-26T10:23:26+01:00 Run job on label - - - - - 2 changed files: - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml Changes: ===================================== .gitlab/generate-ci/gen_ci.hs ===================================== @@ -626,6 +626,7 @@ data ValidateRule = FullCI -- ^ Run this job when the "full-ci" label is present. | LLVMBackend -- ^ Run this job when the "LLVM backend" label is present | JSBackend -- ^ Run this job when the "javascript" label is present + | RiscV -- ^ Run this job when the "RISC-V" label is present | WasmBackend -- ^ Run this job when the "wasm" label is present | FreeBSDLabel -- ^ Run this job when the "FreeBSD" label is set. | NonmovingGc -- ^ Run this job when the "non-moving GC" label is set. @@ -673,6 +674,7 @@ validateRuleString FullCI = or_all ([ labelString "full-ci" validateRuleString LLVMBackend = labelString "LLVM backend" validateRuleString JSBackend = labelString "javascript" +validateRuleString RiscV = labelString "RISC-V" validateRuleString WasmBackend = labelString "wasm" validateRuleString FreeBSDLabel = labelString "FreeBSD" validateRuleString NonmovingGc = labelString "non-moving GC" @@ -1123,7 +1125,7 @@ cross_jobs = [ validateBuilds Amd64 (Linux Debian11) (crossConfig "aarch64-linux-gnu" (Emulator "qemu-aarch64 -L /usr/aarch64-linux-gnu") Nothing) -- x86_64 -> riscv - , validateBuilds Amd64 (Linux Debian12Riscv) (crossConfig "riscv64-linux-gnu" (Emulator "qemu-riscv64 -L /usr/riscv64-linux-gnu") Nothing) + , addValidateRule RiscV (validateBuilds Amd64 (Linux Debian12Riscv) (crossConfig "riscv64-linux-gnu" (Emulator "qemu-riscv64 -L /usr/riscv64-linux-gnu") Nothing)) -- Javascript , addValidateRule JSBackend (validateBuilds Amd64 (Linux Debian11Js) javascriptConfig) ===================================== .gitlab/jobs.yaml ===================================== @@ -5322,7 +5322,7 @@ ], "rules": [ { - "if": "((($CI_MERGE_REQUEST_LABELS =~ /.*full-ci.*/) || ($CI_MERGE_REQUEST_LABELS =~ /.*marge_bot_batch_merge_job.*/) || ($CI_COMMIT_BRANCH == \"master\") || ($CI_COMMIT_BRANCH =~ /ghc-[0-9]+\\.[0-9]+/))) && ($RELEASE_JOB != \"yes\") && ($NIGHTLY == null)", + "if": "((($CI_MERGE_REQUEST_LABELS =~ /.*full-ci.*/) || ($CI_MERGE_REQUEST_LABELS =~ /.*marge_bot_batch_merge_job.*/) || ($CI_COMMIT_BRANCH == \"master\") || ($CI_COMMIT_BRANCH =~ /ghc-[0-9]+\\.[0-9]+/)) || ($CI_MERGE_REQUEST_LABELS =~ /.*RISC-V.*/)) && ($RELEASE_JOB != \"yes\") && ($NIGHTLY == null)", "when": "on_success" } ], View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fe79abc22446151a132a34b5f9068f119750bb48 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/fe79abc22446151a132a34b5f9068f119750bb48 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 10:20:57 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 06:20:57 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] supersven Message-ID: <66f53589645fe_11bcf41ebf40107773@gitlab.mail> Matthew Pickering pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 233bfa78 by Matthew Pickering at 2024-09-26T11:20:43+01:00 supersven - - - - - 2 changed files: - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml Changes: ===================================== .gitlab/generate-ci/gen_ci.hs ===================================== @@ -368,7 +368,7 @@ dockerImage arch (Linux distro) = Just image where image = mconcat - [ "registry.gitlab.haskell.org/ghc/ci-images/" + [ "registry.gitlab.haskell.org/supersven/ci-images/" , archName arch , "-linux-" , distroName distro ===================================== .gitlab/jobs.yaml ===================================== @@ -94,7 +94,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -156,7 +156,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -218,7 +218,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/i386-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/i386-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -347,7 +347,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-alpine3_18:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-alpine3_18:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -411,7 +411,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-deb10:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-deb10:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -474,7 +474,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -537,7 +537,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -600,7 +600,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/i386-linux-deb10:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/i386-linux-deb10:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -663,7 +663,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/i386-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/i386-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -796,7 +796,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -860,7 +860,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -924,7 +924,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -988,7 +988,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1052,7 +1052,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1116,7 +1116,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1180,7 +1180,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_20:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_20:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1244,7 +1244,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-centos7:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-centos7:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1309,7 +1309,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb10:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb10:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1372,7 +1372,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb10:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb10:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1435,7 +1435,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb11:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb11:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1500,7 +1500,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb11-emsdk-closure:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb11-emsdk-closure:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1566,7 +1566,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb11:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb11:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1629,7 +1629,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1692,7 +1692,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1755,7 +1755,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1819,7 +1819,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1882,7 +1882,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12-riscv:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12-riscv:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -1947,7 +1947,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2010,7 +2010,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2073,7 +2073,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2136,7 +2136,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2199,7 +2199,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2264,7 +2264,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2327,7 +2327,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb9:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb9:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2390,7 +2390,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2455,7 +2455,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2521,7 +2521,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2586,7 +2586,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora38:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora38:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2649,7 +2649,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-rocky8:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-rocky8:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2712,7 +2712,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-ubuntu18_04:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-ubuntu18_04:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -2775,7 +2775,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-ubuntu20_04:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-ubuntu20_04:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3032,7 +3032,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-alpine3_18:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-alpine3_18:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3098,7 +3098,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-deb10:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-deb10:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3163,7 +3163,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/aarch64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/aarch64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3228,7 +3228,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/i386-linux-deb10:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/i386-linux-deb10:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3293,7 +3293,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/i386-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/i386-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3429,7 +3429,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3495,7 +3495,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3561,7 +3561,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3627,7 +3627,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_20:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_20:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3693,7 +3693,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-centos7:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-centos7:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3759,7 +3759,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb10:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb10:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3824,7 +3824,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb10:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb10:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3889,7 +3889,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb11:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb11:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -3954,7 +3954,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4019,7 +4019,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb9:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb9:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4084,7 +4084,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4151,7 +4151,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4218,7 +4218,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4285,7 +4285,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora38:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora38:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4350,7 +4350,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-rocky8:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-rocky8:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4415,7 +4415,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-ubuntu18_04:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-ubuntu18_04:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4480,7 +4480,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-ubuntu20_04:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-ubuntu20_04:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4742,7 +4742,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4805,7 +4805,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4868,7 +4868,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4932,7 +4932,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-alpine3_18-wasm:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -4996,7 +4996,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb11:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb11:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5060,7 +5060,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb11-emsdk-closure:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb11-emsdk-closure:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5125,7 +5125,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5187,7 +5187,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5250,7 +5250,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5313,7 +5313,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12-riscv:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12-riscv:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5377,7 +5377,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5439,7 +5439,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5501,7 +5501,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5563,7 +5563,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5625,7 +5625,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5690,7 +5690,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-deb12:$DOCKER_REV", "needs": [ { "artifacts": false, @@ -5752,7 +5752,7 @@ ] }, "dependencies": [], - "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-fedora33:$DOCKER_REV", + "image": "registry.gitlab.haskell.org/supersven/ci-images/x86_64-linux-fedora33:$DOCKER_REV", "needs": [ { "artifacts": false, View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/233bfa78a4b97562faa166167dbfb82001a0e2d5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/233bfa78a4b97562faa166167dbfb82001a0e2d5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 10:36:45 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 06:36:45 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 5 commits: ci: Javascript don't set CROSS_EMULATOR Message-ID: <66f5393d662be_11bcf4386b5c1143b8@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: 9f1f924f by Matthew Pickering at 2024-09-26T10:48:04+01:00 ci: Javascript don't set CROSS_EMULATOR There is no CROSS_EMULATOR needed to run javascript binaries, so we don't set the CROSS_EMULATOR to some dummy value. - - - - - 5019b35d by Matthew Pickering at 2024-09-26T10:48:04+01:00 ci: Introduce CROSS_STAGE variable In preparation for building and testing stage3 bindists we introduce the CROSS_STAGE variable which is used by a CI job to determine what kind of bindist the CI job should produce. At the moment we are only using CROSS_STAGE=2 but in the future we will have some jobs which set CROSS_STAGE=3 to produce native bindists for a target, but produced by a cross compiler, which can be tested on by another CI job on the native platform. CROSS_STAGE=2: Build a normal cross compiler bindist CROSS_STAGE=3: Build a stage 3 bindist, one which is a native compiler and library for the target - - - - - 61f3fa32 by Matthew Pickering at 2024-09-26T10:48:04+01:00 Split up system.config into host/target config files There were a number of settings which were not applied per-stage, for example if you specified `--ffi-include-dir` then that was applied to both host and target. Now this will just be passed when building the crosscompiler. The solution for now is to separate these two files into host/target and the host file contains very bare-bones . There isn't currently a way to specify with configure anything in the host file, so if you are building a cross-compiler and you need to do that, you have to modify the file yourself. - - - - - 676a4b04 by Matthew Pickering at 2024-09-26T10:48:04+01:00 wip fixup - - - - - 4a8b0a9a by Matthew Pickering at 2024-09-26T11:33:06+01:00 Fix location of emsdk-version - - - - - 26 changed files: - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - configure.ac - hadrian/README.md - + hadrian/cfg/system.config.host.in - hadrian/cfg/system.config.in - + hadrian/cfg/system.config.target.in - hadrian/src/Base.hs - hadrian/src/Builder.hs - hadrian/src/Expression.hs - hadrian/src/Hadrian/Oracles/TextFile.hs - hadrian/src/Oracles/Flag.hs - hadrian/src/Oracles/Setting.hs - hadrian/src/Oracles/TestSettings.hs - hadrian/src/Rules/Generate.hs - hadrian/src/Rules/Gmp.hs - hadrian/src/Rules/Libffi.hs - hadrian/src/Rules/Rts.hs - hadrian/src/Settings/Builders/Cabal.hs - hadrian/src/Settings/Builders/Common.hs - hadrian/src/Settings/Builders/Ghc.hs - hadrian/src/Settings/Builders/Hsc2Hs.hs - hadrian/src/Settings/Builders/RunTest.hs - hadrian/src/Settings/Packages.hs - hadrian/src/Settings/Warnings.hs Changes: ===================================== .gitlab/ci.sh ===================================== @@ -58,6 +58,9 @@ Hadrian build system Environment variables affecting both build systems: CROSS_TARGET Triple of cross-compilation target. + CROSS_STAGE The stage of the cross-compiler to build either + * 2: Build a normal cross-compiler bindist + * 3: Build a target executable bindist (with the stage2 cross-compiler) VERBOSE Set to non-empty for verbose build output RUNTEST_ARGS Arguments passed to runtest.py MSYSTEM (Windows-only) Which platform to build from (CLANG64). @@ -496,6 +499,12 @@ function build_hadrian() { export XZ_OPT="${XZ_OPT:-} -T$cores" fi + case "${CROSS_STAGE:-2}" in + 2) BINDIST_TARGET="binary-dist";; + 3) BINDIST_TARGET="binary-dist-stage3";; + *) fail "Unknown CROSS_STAGE, must be 2 or 3";; + esac + if [[ -n "${REINSTALL_GHC:-}" ]]; then run_hadrian build-cabal -V else @@ -505,7 +514,7 @@ function build_hadrian() { mv _build/reloc-bindist/ghc*.tar.xz "$BIN_DIST_NAME.tar.xz" ;; *) - run_hadrian test:all_deps binary-dist -V + run_hadrian test:all_deps $BINDIST_TARGET mv _build/bindist/ghc*.tar.xz "$BIN_DIST_NAME.tar.xz" ;; esac ===================================== .gitlab/generate-ci/gen_ci.hs ===================================== @@ -147,6 +147,7 @@ data BuildConfig , withNuma :: Bool , withZstd :: Bool , crossTarget :: Maybe String + , crossStage :: Maybe Int , crossEmulator :: CrossEmulator , configureWrapper :: Maybe String , fullyStatic :: Bool @@ -209,6 +210,7 @@ vanilla = BuildConfig , withNuma = False , withZstd = False , crossTarget = Nothing + , crossStage = Nothing , crossEmulator = NoEmulator , configureWrapper = Nothing , fullyStatic = False @@ -257,6 +259,7 @@ crossConfig :: String -- ^ target triple -> BuildConfig crossConfig triple emulator configure_wrapper = vanilla { crossTarget = Just triple + , crossStage = Just 2 , crossEmulator = emulator , configureWrapper = configure_wrapper } @@ -771,6 +774,7 @@ job arch opsys buildConfig = NamedJob { name = jobName, jobInfo = Job {..} } , "INSTALL_CONFIGURE_ARGS" =: "--enable-strict-ghc-toolchain-check" , maybe mempty ("CONFIGURE_WRAPPER" =:) (configureWrapper buildConfig) , maybe mempty ("CROSS_TARGET" =:) (crossTarget buildConfig) + , maybe mempty (("CROSS_STAGE" =:) . show) (crossStage buildConfig) , case crossEmulator buildConfig of NoEmulator -> case crossTarget buildConfig of Nothing -> mempty @@ -1026,7 +1030,6 @@ debian_x86 = ] where validate_debian = Debian12 - perfProfilingJob arch sys buildConfig = -- Rename the job to avoid conflicts rename (<> "-perf") @@ -1131,7 +1134,7 @@ cross_jobs = [ make_wasm_jobs wasm_build_config {unregisterised = True} ] where - javascriptConfig = (crossConfig "javascript-unknown-ghcjs" (Emulator "js-emulator") (Just "emconfigure")) + javascriptConfig = (crossConfig "javascript-unknown-ghcjs" NoEmulatorNeeded (Just "emconfigure")) { bignumBackend = Native } make_wasm_jobs cfg = ===================================== .gitlab/jobs.yaml ===================================== @@ -1017,6 +1017,7 @@ "BIN_DIST_NAME": "ghc-x86_64-linux-alpine3_18-wasm-cross_wasm32-wasi-release+fully_static+text_simdutf", "BUILD_FLAVOUR": "release+fully_static+text_simdutf", "CONFIGURE_ARGS": "--with-intree-gmp --with-system-libffi --enable-strict-ghc-toolchain-check", + "CROSS_STAGE": "2", "CROSS_TARGET": "wasm32-wasi", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-alpine3_18-wasm-cross_wasm32-wasi-release+fully_static+text_simdutf", @@ -1080,6 +1081,7 @@ "BIN_DIST_NAME": "ghc-x86_64-linux-alpine3_18-wasm-int_native-cross_wasm32-wasi-release+fully_static+text_simdutf", "BUILD_FLAVOUR": "release+fully_static+text_simdutf", "CONFIGURE_ARGS": "--with-intree-gmp --with-system-libffi --enable-strict-ghc-toolchain-check", + "CROSS_STAGE": "2", "CROSS_TARGET": "wasm32-wasi", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-alpine3_18-wasm-int_native-cross_wasm32-wasi-release+fully_static+text_simdutf", @@ -1143,6 +1145,7 @@ "BIN_DIST_NAME": "ghc-x86_64-linux-alpine3_18-wasm-unreg-cross_wasm32-wasi-release+fully_static+text_simdutf", "BUILD_FLAVOUR": "release+fully_static+text_simdutf", "CONFIGURE_ARGS": "--enable-unregisterised --with-intree-gmp --with-system-libffi --enable-strict-ghc-toolchain-check", + "CROSS_STAGE": "2", "CROSS_TARGET": "wasm32-wasi", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-alpine3_18-wasm-unreg-cross_wasm32-wasi-release+fully_static+text_simdutf", @@ -1462,6 +1465,7 @@ "BUILD_FLAVOUR": "validate", "CONFIGURE_ARGS": "--with-intree-gmp --enable-strict-ghc-toolchain-check", "CROSS_EMULATOR": "qemu-aarch64 -L /usr/aarch64-linux-gnu", + "CROSS_STAGE": "2", "CROSS_TARGET": "aarch64-linux-gnu", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", @@ -1527,7 +1531,7 @@ "BUILD_FLAVOUR": "validate", "CONFIGURE_ARGS": "--with-intree-gmp --enable-strict-ghc-toolchain-check", "CONFIGURE_WRAPPER": "emconfigure", - "CROSS_EMULATOR": "js-emulator", + "CROSS_STAGE": "2", "CROSS_TARGET": "javascript-unknown-ghcjs", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", @@ -4766,6 +4770,7 @@ "BIN_DIST_NAME": "ghc-x86_64-linux-alpine3_18-wasm-cross_wasm32-wasi-release+fully_static+text_simdutf", "BUILD_FLAVOUR": "release+fully_static+text_simdutf", "CONFIGURE_ARGS": "--with-intree-gmp --with-system-libffi --enable-strict-ghc-toolchain-check", + "CROSS_STAGE": "2", "CROSS_TARGET": "wasm32-wasi", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-alpine3_18-wasm-cross_wasm32-wasi-release+fully_static+text_simdutf" @@ -4829,6 +4834,7 @@ "BIN_DIST_NAME": "ghc-x86_64-linux-alpine3_18-wasm-int_native-cross_wasm32-wasi-release+fully_static+text_simdutf", "BUILD_FLAVOUR": "release+fully_static+text_simdutf", "CONFIGURE_ARGS": "--with-intree-gmp --with-system-libffi --enable-strict-ghc-toolchain-check", + "CROSS_STAGE": "2", "CROSS_TARGET": "wasm32-wasi", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-alpine3_18-wasm-int_native-cross_wasm32-wasi-release+fully_static+text_simdutf" @@ -4892,6 +4898,7 @@ "BIN_DIST_NAME": "ghc-x86_64-linux-alpine3_18-wasm-unreg-cross_wasm32-wasi-release+fully_static+text_simdutf", "BUILD_FLAVOUR": "release+fully_static+text_simdutf", "CONFIGURE_ARGS": "--enable-unregisterised --with-intree-gmp --with-system-libffi --enable-strict-ghc-toolchain-check", + "CROSS_STAGE": "2", "CROSS_TARGET": "wasm32-wasi", "RUNTEST_ARGS": "", "TEST_ENV": "x86_64-linux-alpine3_18-wasm-unreg-cross_wasm32-wasi-release+fully_static+text_simdutf" @@ -4955,6 +4962,7 @@ "BUILD_FLAVOUR": "validate", "CONFIGURE_ARGS": "--with-intree-gmp --enable-strict-ghc-toolchain-check", "CROSS_EMULATOR": "qemu-aarch64 -L /usr/aarch64-linux-gnu", + "CROSS_STAGE": "2", "CROSS_TARGET": "aarch64-linux-gnu", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", @@ -5019,7 +5027,7 @@ "BUILD_FLAVOUR": "validate", "CONFIGURE_ARGS": "--with-intree-gmp --enable-strict-ghc-toolchain-check", "CONFIGURE_WRAPPER": "emconfigure", - "CROSS_EMULATOR": "js-emulator", + "CROSS_STAGE": "2", "CROSS_TARGET": "javascript-unknown-ghcjs", "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", "RUNTEST_ARGS": "", ===================================== configure.ac ===================================== @@ -1010,6 +1010,8 @@ FIND_GHC_TOOLCHAIN([hadrian/cfg]) AC_CONFIG_FILES( [ hadrian/cfg/system.config + hadrian/cfg/system.config.host + hadrian/cfg/system.config.target hadrian/ghci-cabal hadrian/ghci-multi-cabal hadrian/ghci-stack ===================================== hadrian/README.md ===================================== @@ -327,6 +327,16 @@ workflow, for now. Note: On windows you need to use the `reloc-binary-dist` target. +#### Cross Compiling + +If you are cross compiling then all the settings specified to ./configure are +for the target system. For example, if you specify `--with-ffi-includes` then +this is specifically for the target. + +If your host system needs additional configuration in order to build a stage1 compiler, +then at the moment you need to manually edit the "cfg/system.config.host.in" file to +specify these options in the right place. + #### Relocatable Binary Distribution If you require a relocatable binary distribution (for example on Windows), then you ===================================== hadrian/cfg/system.config.host.in ===================================== @@ -0,0 +1,78 @@ +# This file is processed by the configure script. +# See hadrian/src/UserSettings.hs for user-defined settings. +#=========================================================== + +# Paths to builders: +#=================== + +# Information about builders: +#============================ + +cc-llvm-backend = NO + + +# Information about build, host and target systems: +#================================================== + +dynamic-extension = @soext_host@ + +bootstrap-threaded-rts = YES + +# Settings: +#========== + +# We are in the process of moving the settings file from being entirely +# generated by configure, to generated being by the build system. Many of these +# might become redundant. +# See Note [tooldir: How GHC finds mingw on Windows] + +settings-otool-command = otool +settings-install_name_tool-command = install_name_tool +settings-llc-command = llc +settings-opt-command = opt +settings-llvm-as-command = clang +settings-use-distro-mingw = NO + +target-has-libm = YES + +# Include and library directories: +#================================= + +curses-lib-dir = +curses-include-dir = + +iconv-include-dir = +iconv-lib-dir = + +intree-gmp = NO +gmp-framework-preferred = NO +gmp-include-dir = +gmp-lib-dir = + +use-system-ffi = NO +ffi-include-dir = +ffi-lib-dir = + +libdw-include-dir = +libdw-lib-dir = + +libnuma-include-dir = +libnuma-lib-dir = + +libzstd-include-dir = +libzstd-lib-dir = + +# Optional Dependencies: +#======================= + +use-lib-dw = NO +use-lib-zstd = NO +static-lib-zstd = NO +use-lib-numa = NO +use-lib-m = YES +use-lib-rt = YES +use-lib-dl = YES +use-lib-bfd = NO +use-lib-pthread = NO +need-libatomic = NO + ===================================== hadrian/cfg/system.config.in ===================================== @@ -31,11 +31,11 @@ python = @PythonCmd@ # Information about builders: #============================ -cc-llvm-backend = @CcLlvmBackend@ - llvm-min-version = @LlvmMinVersion@ llvm-max-version = @LlvmMaxVersion@ +emsdk-version = @ConfiguredEmsdkVersion@ + # Build options: #=============== @@ -53,8 +53,6 @@ target-platform-full = @TargetPlatformFull@ cross-compiling = @CrossCompiling@ -dynamic-extension = @soext_target@ - ghc-version = @GhcVersion@ ghc-major-version = @GhcMajVersion@ ghc-minor-version = @GhcMinVersion@ @@ -72,62 +70,3 @@ project-patch-level1 = @ProjectPatchLevel1@ project-patch-level2 = @ProjectPatchLevel2@ project-git-commit-id = @ProjectGitCommitId@ -# Settings: -#========== - -# We are in the process of moving the settings file from being entirely -# generated by configure, to generated being by the build system. Many of these -# might become redundant. -# See Note [tooldir: How GHC finds mingw on Windows] - -settings-otool-command = @SettingsOtoolCommand@ -settings-install_name_tool-command = @SettingsInstallNameToolCommand@ -settings-llc-command = @SettingsLlcCommand@ -settings-opt-command = @SettingsOptCommand@ -settings-llvm-as-command = @SettingsLlvmAsCommand@ -settings-use-distro-mingw = @SettingsUseDistroMINGW@ - -target-has-libm = @TargetHasLibm@ - -# Include and library directories: -#================================= - -curses-lib-dir = @CURSES_LIB_DIRS@ -curses-include-dir = @CURSES_INCLUDE_DIRS@ - -iconv-include-dir = @ICONV_INCLUDE_DIRS@ -iconv-lib-dir = @ICONV_LIB_DIRS@ - -intree-gmp = @GMP_FORCE_INTREE@ -gmp-framework-preferred = @GMP_PREFER_FRAMEWORK@ -gmp-include-dir = @GMP_INCLUDE_DIRS@ -gmp-lib-dir = @GMP_LIB_DIRS@ - -use-system-ffi = @UseSystemLibFFI@ -ffi-include-dir = @FFIIncludeDir@ -ffi-lib-dir = @FFILibDir@ - -libdw-include-dir = @LibdwIncludeDir@ -libdw-lib-dir = @LibdwLibDir@ - -libnuma-include-dir = @LibNumaIncludeDir@ -libnuma-lib-dir = @LibNumaLibDir@ - -libzstd-include-dir = @LibZstdIncludeDir@ -libzstd-lib-dir = @LibZstdLibDir@ - -# Optional Dependencies: -#======================= - -use-lib-dw = @UseLibdw@ -use-lib-zstd = @UseLibZstd@ -static-lib-zstd = @UseStaticLibZstd@ -use-lib-numa = @UseLibNuma@ -use-lib-m = @UseLibm@ -use-lib-rt = @UseLibrt@ -use-lib-dl = @UseLibdl@ -use-lib-bfd = @UseLibbfd@ -use-lib-pthread = @UseLibpthread@ -need-libatomic = @NeedLibatomic@ - -emsdk-version = @ConfiguredEmsdkVersion@ ===================================== hadrian/cfg/system.config.target.in ===================================== @@ -0,0 +1,84 @@ +# This file is processed by the configure script. +# See hadrian/src/UserSettings.hs for user-defined settings. +#=========================================================== + +# Paths to builders: +#=================== + +# Information about builders: +#============================ + +cc-llvm-backend = @CcLlvmBackend@ + + +# Information about build, host and target systems: +#================================================== + +# Q: Is the *-platform information available in the target? +# A: Yes, it is. We pass @BuildPlatform@ and @HostPlatform@ and @TargetPlatform@ to ghc-toolchain using --target=that +# And we can reconstruct the platform info using targetPlatformTriple +# Q: What is TargetPlatformFull? +target-platform-full = @TargetPlatformFull@ + +dynamic-extension = @soext_target@ + +bootstrap-threaded-rts = @GhcThreadedRts@ + +# Settings: +#========== + +# We are in the process of moving the settings file from being entirely +# generated by configure, to generated being by the build system. Many of these +# might become redundant. +# See Note [tooldir: How GHC finds mingw on Windows] + +settings-otool-command = @SettingsOtoolCommand@ +settings-install_name_tool-command = @SettingsInstallNameToolCommand@ +settings-llc-command = @SettingsLlcCommand@ +settings-opt-command = @SettingsOptCommand@ +settings-llvm-as-command = @SettingsLlvmAsCommand@ +settings-use-distro-mingw = @SettingsUseDistroMINGW@ + +target-has-libm = @TargetHasLibm@ + +# Include and library directories: +#================================= + +curses-lib-dir = @CURSES_LIB_DIRS@ +curses-include-dir = @CURSES_INCLUDE_DIRS@ + +iconv-include-dir = @ICONV_INCLUDE_DIRS@ +iconv-lib-dir = @ICONV_LIB_DIRS@ + +intree-gmp = @GMP_FORCE_INTREE@ +gmp-framework-preferred = @GMP_PREFER_FRAMEWORK@ +gmp-include-dir = @GMP_INCLUDE_DIRS@ +gmp-lib-dir = @GMP_LIB_DIRS@ + +use-system-ffi = @UseSystemLibFFI@ +ffi-include-dir = @FFIIncludeDir@ +ffi-lib-dir = @FFILibDir@ + +libdw-include-dir = @LibdwIncludeDir@ +libdw-lib-dir = @LibdwLibDir@ + +libnuma-include-dir = @LibNumaIncludeDir@ +libnuma-lib-dir = @LibNumaLibDir@ + +libzstd-include-dir = @LibZstdIncludeDir@ +libzstd-lib-dir = @LibZstdLibDir@ + +# Optional Dependencies: +#======================= + +use-lib-dw = @UseLibdw@ +use-lib-zstd = @UseLibZstd@ +static-lib-zstd = @UseStaticLibZstd@ +use-lib-numa = @UseLibNuma@ +use-lib-m = @UseLibm@ +use-lib-rt = @UseLibrt@ +use-lib-dl = @UseLibdl@ +use-lib-bfd = @UseLibbfd@ +use-lib-pthread = @UseLibpthread@ +need-libatomic = @NeedLibatomic@ + ===================================== hadrian/src/Base.hs ===================================== @@ -29,7 +29,8 @@ module Base ( module Way, -- * Paths - hadrianPath, configPath, configFile, sourcePath, shakeFilesDir, + hadrianPath, configPath, configFile, buildConfigFileHost, buildConfigFileTarget, + sourcePath, shakeFilesDir, stageBinPath, stageLibPath, templateHscPath, buildTargetFile, hostTargetFile, targetTargetFile, ghcLibDeps, haddockDeps, @@ -80,6 +81,13 @@ configPath = hadrianPath -/- "cfg" configFile :: FilePath configFile = configPath -/- "system.config" +buildConfigFileHost :: FilePath +buildConfigFileHost = configPath -/- "system.config.host" + +buildConfigFileTarget :: FilePath +buildConfigFileTarget = configPath -/- "system.config.target" + + -- | The target configuration file generated by ghc-toolchain for the -- compilation build platform buildTargetFile :: FilePath ===================================== hadrian/src/Builder.hs ===================================== @@ -26,15 +26,13 @@ import Hadrian.Builder.Tar import Hadrian.Oracles.Path import Hadrian.Oracles.TextFile import Hadrian.Utilities -import Oracles.Setting (bashPath, targetStage) import System.Exit import System.IO (stderr) import Base import Context import Oracles.Flag -import Oracles.Setting (setting, Setting(..)) -import Oracles.Setting (settingsFileSetting, ToolchainSetting(..)) +import Oracles.Setting import Packages import GHC.IO.Encoding (getFileSystemEncoding) @@ -240,10 +238,10 @@ instance H.Builder Builder where Ghc _ st -> do root <- buildRoot unlitPath <- builderPath Unlit - distro_mingw <- settingsFileSetting ToolchainSetting_DistroMinGW + distro_mingw <- settingsFileSetting ToolchainSetting_DistroMinGW st -- TODO: Check this is the right stage libffi_adjustors <- targetUseLibffiForAdjustors st - use_system_ffi <- flag UseSystemFfi + use_system_ffi <- buildFlag UseSystemFfi st return $ [ unlitPath ] ++ [ root -/- mingwStamp | windowsHost, distro_mingw == "NO" ] ===================================== hadrian/src/Expression.hs ===================================== @@ -146,7 +146,7 @@ buildingCompilerStage' f = f . succStage <$> getStage -- compiler's RTS ways. See Note [Linking ghc-bin against threaded stage0 RTS] -- in Settings.Packages for details. threadedBootstrapper :: Predicate -threadedBootstrapper = expr (flag BootstrapThreadedRts) +threadedBootstrapper = staged (buildFlag BootstrapThreadedRts) -- | Is a certain package /not/ built right now? notPackage :: Package -> Predicate ===================================== hadrian/src/Hadrian/Oracles/TextFile.hs ===================================== @@ -13,7 +13,7 @@ -- to read configuration or package metadata files and cache the parsing. ----------------------------------------------------------------------------- module Hadrian.Oracles.TextFile ( - lookupValue, lookupValueOrEmpty, lookupValueOrError, lookupSystemConfig, lookupValues, + lookupValue, lookupValueOrEmpty, lookupValueOrError, lookupSystemConfig, lookupHostBuildConfig, lookupTargetBuildConfig, lookupValues, lookupValuesOrEmpty, lookupValuesOrError, lookupDependencies, textFileOracle, getBuildTarget, getHostTarget, getTargetTarget, queryBuildTarget, queryHostTarget @@ -52,6 +52,23 @@ lookupSystemConfig = lookupValueOrError (Just configError) configFile where configError = "Perhaps you need to rerun ./configure" +lookupHostBuildConfig :: String -> Action String +lookupHostBuildConfig key = do + cross <- (== "YES") <$> lookupSystemConfig "cross-compiling" + -- If we are not cross compiling, the build the host compiler like the target. + let cfgFile = if cross then buildConfigFileHost else buildConfigFileTarget + lookupValueOrError (Just configError) cfgFile key + where + configError = "Perhaps you need to rerun ./configure" + +lookupTargetBuildConfig :: String -> Action String +lookupTargetBuildConfig key = + lookupValueOrError (Just configError) buildConfigFileTarget key + where + configError = "Perhaps you need to rerun ./configure" + + + -- | Lookup a list of values in a text file, tracking the result. Each line of -- the file is expected to have @key value1 value2 ...@ format. lookupValues :: FilePath -> String -> Action (Maybe [String]) @@ -104,6 +121,7 @@ getHostTarget = do -- MP: If we are not cross compiling then we should use the target file in order to -- build things for the host, in particular we want to use the configured values for the -- target for building the RTS (ie are we using Libffi for adjustors, and the wordsize) + -- TODO: Use "flag CrossCompiling" ht <- getTargetConfig hostTargetFile tt <- getTargetConfig targetTargetFile if (Toolchain.targetPlatformTriple ht) == (Toolchain.targetPlatformTriple tt) ===================================== hadrian/src/Oracles/Flag.hs ===================================== @@ -2,6 +2,7 @@ module Oracles.Flag ( Flag (..), flag, getFlag, + BuildFlag(..), buildFlag, targetSupportsSharedLibs, targetSupportsGhciObjects, targetSupportsThreadedRts, @@ -22,7 +23,8 @@ import qualified GHC.Toolchain as Toolchain import GHC.Platform.ArchOS data Flag = CrossCompiling - | CcLlvmBackend + | UseGhcToolchain +data BuildFlag = CcLlvmBackend | GmpInTree | GmpFrameworkPref | UseSystemFfi @@ -38,7 +40,15 @@ data Flag = CrossCompiling | UseLibbfd | UseLibpthread | NeedLibatomic - | UseGhcToolchain + | TargetHasLibm + +parseFlagResult :: String -> String -> Bool +parseFlagResult key value = + if (value `notElem` ["YES", "NO", ""]) + then error $ "Configuration flag " + ++ quote (key ++ " = " ++ value) ++ " cannot be parsed." + else value == "YES" + -- Note, if a flag is set to empty string we treat it as set to NO. This seems -- fragile, but some flags do behave like this. @@ -46,6 +56,12 @@ flag :: Flag -> Action Bool flag f = do let key = case f of CrossCompiling -> "cross-compiling" + UseGhcToolchain -> "use-ghc-toolchain" + parseFlagResult key <$> lookupSystemConfig key + +buildFlag :: BuildFlag -> Stage -> Action Bool +buildFlag f st = + let key = case f of CcLlvmBackend -> "cc-llvm-backend" GmpInTree -> "intree-gmp" GmpFrameworkPref -> "gmp-framework-preferred" @@ -62,11 +78,13 @@ flag f = do UseLibbfd -> "use-lib-bfd" UseLibpthread -> "use-lib-pthread" NeedLibatomic -> "need-libatomic" - UseGhcToolchain -> "use-ghc-toolchain" - value <- lookupSystemConfig key - when (value `notElem` ["YES", "NO", ""]) . error $ "Configuration flag " - ++ quote (key ++ " = " ++ value) ++ " cannot be parsed." - return $ value == "YES" + TargetHasLibm -> "target-has-libm" + in parseFlagResult key <$> (tgtConfig st key) + where + tgtConfig Stage0 {} = lookupHostBuildConfig + tgtConfig Stage1 = lookupHostBuildConfig + tgtConfig Stage2 = lookupTargetBuildConfig + tgtConfig Stage3 = lookupTargetBuildConfig -- | Get a configuration setting. getFlag :: Flag -> Expr c b Bool ===================================== hadrian/src/Oracles/Setting.hs ===================================== @@ -1,7 +1,8 @@ module Oracles.Setting ( configFile, -- * Settings - Setting (..), setting, getSetting, + ProjectSetting (..), setting, getSetting, + BuildSetting(..), buildSetting, ToolchainSetting (..), settingsFileSetting, -- * Helpers @@ -39,28 +40,14 @@ import GHC.Platform.ArchOS -- tracking the result in the Shake database. -- -- * ROMES:TODO: How to handle target-platform-full? -data Setting = CursesIncludeDir - | CursesLibDir - | DynamicExtension - | FfiIncludeDir - | FfiLibDir - | GhcMajorVersion +data ProjectSetting = + GhcMajorVersion | GhcMinorVersion | GhcPatchLevel | GhcVersion | GhcSourcePath | LlvmMinVersion | LlvmMaxVersion - | GmpIncludeDir - | GmpLibDir - | IconvIncludeDir - | IconvLibDir - | LibdwIncludeDir - | LibdwLibDir - | LibnumaIncludeDir - | LibnumaLibDir - | LibZstdIncludeDir - | LibZstdLibDir | ProjectGitCommitId | ProjectName | ProjectVersion @@ -75,6 +62,25 @@ data Setting = CursesIncludeDir | BourneShell | EmsdkVersion + +-- Things which configure how a specific stage is built +data BuildSetting = + CursesIncludeDir + | CursesLibDir + | DynamicExtension + | FfiIncludeDir + | FfiLibDir + | GmpIncludeDir + | GmpLibDir + | IconvIncludeDir + | IconvLibDir + | LibdwIncludeDir + | LibdwLibDir + | LibnumaIncludeDir + | LibnumaLibDir + | LibZstdIncludeDir + | LibZstdLibDir + -- TODO compute solely in Hadrian, removing these variables' definitions -- from aclocal.m4 whenever they can be calculated from other variables -- already fed into Hadrian. @@ -95,13 +101,8 @@ data ToolchainSetting -- | Look up the value of a 'Setting' in @cfg/system.config@, tracking the -- result. -setting :: Setting -> Action String +setting :: ProjectSetting -> Action String setting key = lookupSystemConfig $ case key of - CursesIncludeDir -> "curses-include-dir" - CursesLibDir -> "curses-lib-dir" - DynamicExtension -> "dynamic-extension" - FfiIncludeDir -> "ffi-include-dir" - FfiLibDir -> "ffi-lib-dir" GhcMajorVersion -> "ghc-major-version" GhcMinorVersion -> "ghc-minor-version" GhcPatchLevel -> "ghc-patch-level" @@ -109,16 +110,6 @@ setting key = lookupSystemConfig $ case key of GhcSourcePath -> "ghc-source-path" LlvmMinVersion -> "llvm-min-version" LlvmMaxVersion -> "llvm-max-version" - GmpIncludeDir -> "gmp-include-dir" - GmpLibDir -> "gmp-lib-dir" - IconvIncludeDir -> "iconv-include-dir" - IconvLibDir -> "iconv-lib-dir" - LibdwIncludeDir -> "libdw-include-dir" - LibdwLibDir -> "libdw-lib-dir" - LibnumaIncludeDir -> "libnuma-include-dir" - LibnumaLibDir -> "libnuma-lib-dir" - LibZstdIncludeDir -> "libzstd-include-dir" - LibZstdLibDir -> "libzstd-lib-dir" ProjectGitCommitId -> "project-git-commit-id" ProjectName -> "project-name" ProjectVersion -> "project-version" @@ -133,22 +124,51 @@ setting key = lookupSystemConfig $ case key of BourneShell -> "bourne-shell" EmsdkVersion -> "emsdk-version" +buildSetting :: BuildSetting -> Stage -> Action String +buildSetting key stage = tgtConfig stage $ case key of + CursesIncludeDir -> "curses-include-dir" + CursesLibDir -> "curses-lib-dir" + DynamicExtension -> "dynamic-extension" + FfiIncludeDir -> "ffi-include-dir" + FfiLibDir -> "ffi-lib-dir" + GmpIncludeDir -> "gmp-include-dir" + GmpLibDir -> "gmp-lib-dir" + IconvIncludeDir -> "iconv-include-dir" + IconvLibDir -> "iconv-lib-dir" + LibdwIncludeDir -> "libdw-include-dir" + LibdwLibDir -> "libdw-lib-dir" + LibnumaIncludeDir -> "libnuma-include-dir" + LibnumaLibDir -> "libnuma-lib-dir" + LibZstdIncludeDir -> "libzstd-include-dir" + LibZstdLibDir -> "libzstd-lib-dir" + where + tgtConfig Stage0 {} = lookupHostBuildConfig + tgtConfig Stage1 = lookupHostBuildConfig + tgtConfig Stage2 = lookupTargetBuildConfig + tgtConfig Stage3 = lookupTargetBuildConfig + + -- | Look up the value of a 'SettingList' in @cfg/system.config@, tracking the -- result. -- See Note [tooldir: How GHC finds mingw on Windows] -- ROMES:TODO: This should be queryTargetTargetConfig -settingsFileSetting :: ToolchainSetting -> Action String -settingsFileSetting key = lookupSystemConfig $ case key of +settingsFileSetting :: ToolchainSetting -> Stage -> Action String +settingsFileSetting key stage = tgtConfig stage $ case key of ToolchainSetting_OtoolCommand -> "settings-otool-command" ToolchainSetting_InstallNameToolCommand -> "settings-install_name_tool-command" ToolchainSetting_LlcCommand -> "settings-llc-command" ToolchainSetting_OptCommand -> "settings-opt-command" ToolchainSetting_LlvmAsCommand -> "settings-llvm-as-command" ToolchainSetting_DistroMinGW -> "settings-use-distro-mingw" -- ROMES:TODO: This option doesn't seem to be in ghc-toolchain yet. It corresponds to EnableDistroToolchain + where + tgtConfig Stage0 {} = lookupHostBuildConfig + tgtConfig Stage1 = lookupHostBuildConfig + tgtConfig Stage2 = lookupTargetBuildConfig + tgtConfig Stage3 = lookupTargetBuildConfig -- | An expression that looks up the value of a 'Setting' in @cfg/system.config@, -- tracking the result. -getSetting :: Setting -> Expr c b String +getSetting :: ProjectSetting -> Expr c b String getSetting = expr . setting -- | The path to a Bourne shell interpreter. @@ -251,7 +271,7 @@ libsuf :: Stage -> Way -> Action String libsuf st way | not (wayUnit Dynamic way) = return (waySuffix way ++ ".a") -- e.g., _p.a | otherwise = do - extension <- setting DynamicExtension -- e.g., .dll or .so + extension <- buildSetting DynamicExtension st-- e.g., .dll or .so version <- ghcVersionStage st -- e.g. 8.4.4 or 8.9.xxxx let suffix = waySuffix (removeWayUnit Dynamic way) return (suffix ++ "-ghc" ++ version ++ extension) ===================================== hadrian/src/Oracles/TestSettings.hs ===================================== @@ -10,7 +10,7 @@ module Oracles.TestSettings import Base import Hadrian.Oracles.TextFile -import Oracles.Setting (topDirectory, setting, Setting(..), crossStage) +import Oracles.Setting (topDirectory, setting, ProjectSetting(..), crossStage) import Packages import Settings.Program (programContext) import Hadrian.Oracles.Path ===================================== hadrian/src/Rules/Generate.hs ===================================== @@ -11,7 +11,6 @@ import qualified Data.Set as Set import Base import qualified Context import Expression -import Hadrian.Oracles.TextFile (lookupSystemConfig) import Oracles.Flag hiding (arSupportsAtFile, arSupportsDashL) import Oracles.ModuleFiles import Oracles.Setting @@ -50,10 +49,9 @@ ghcPrimDependencies = do rtsDependencies :: Expr [FilePath] rtsDependencies = do - stage <- getStage - rtsPath <- expr (rtsBuildPath stage) - jsTarget <- expr (isJsTarget stage) - useSystemFfi <- expr (flag UseSystemFfi) + rtsPath <- staged rtsBuildPath + jsTarget <- staged isJsTarget + useSystemFfi <- staged (buildFlag UseSystemFfi) let -- headers common to native and JS RTS common_headers = @@ -278,7 +276,7 @@ runInterpolations (Interpolations mk_substs) input = do return (subst input) -- | Interpolate the given variable with the value of the given 'Setting'. -interpolateSetting :: String -> Setting -> Interpolations +interpolateSetting :: String -> ProjectSetting -> Interpolations interpolateSetting name settng = interpolateVar name $ setting settng -- | Interpolate the @ProjectVersion@, @ProjectVersionMunged@, and @ProjectVersionForLib@ variables. @@ -401,14 +399,14 @@ bindistRules = do , interpolateSetting "LlvmMinVersion" LlvmMinVersion , interpolateVar "LlvmTarget" $ getTarget tgtLlvmTarget , interpolateSetting "ProjectVersion" ProjectVersion - , interpolateVar "SettingsUseDistroMINGW" $ settingsFileSetting ToolchainSetting_DistroMinGW + , interpolateVar "SettingsUseDistroMINGW" $ interp (staged (settingsFileSetting ToolchainSetting_DistroMinGW)) , interpolateVar "TablesNextToCode" $ yesNo <$> getTarget tgtTablesNextToCode - , interpolateVar "TargetHasLibm" $ lookupSystemConfig "target-has-libm" + , interpolateVar "TargetHasLibm" $ yesNo <$> interp (staged (buildFlag TargetHasLibm)) , interpolateVar "TargetPlatform" $ getTarget targetPlatformTriple , interpolateVar "TargetWordBigEndian" $ getTarget isBigEndian , interpolateVar "TargetWordSize" $ getTarget wordSize , interpolateVar "Unregisterised" $ yesNo <$> getTarget tgtUnregisterised - , interpolateVar "UseLibdw" $ fmap yesNo $ interp $ getFlag UseLibdw + , interpolateVar "UseLibdw" $ fmap yesNo $ interp $ staged (buildFlag UseLibdw) , interpolateVar "UseLibffiForAdjustors" $ yesNo <$> getTarget tgtUseLibffiForAdjustors , interpolateVar "GhcWithSMP" $ yesNo <$> targetSupportsSMP Stage2 ] @@ -479,8 +477,8 @@ generateSettings settingsFile = do , ("ar supports at file", queryTarget stage arSupportsAtFile') , ("ar supports -L", queryTarget stage arSupportsDashL') , ("ranlib command", queryTarget stage ranlibPath) - , ("otool command", expr $ settingsFileSetting ToolchainSetting_OtoolCommand) - , ("install_name_tool command", expr $ settingsFileSetting ToolchainSetting_InstallNameToolCommand) + , ("otool command", staged $ settingsFileSetting ToolchainSetting_OtoolCommand) + , ("install_name_tool command", staged $ settingsFileSetting ToolchainSetting_InstallNameToolCommand) , ("windres command", queryTarget stage (maybe "/bin/false" prgPath . tgtWindres)) -- TODO: /bin/false is not available on many distributions by default, but we keep it as it were before the ghc-toolchain patch. Fix-me. , ("unlit command", ("$topdir/../bin/" <>) <$> expr (programName (ctx { Context.package = unlit, Context.stage = predStage stage }))) , ("cross compiling", expr $ yesNo <$> crossStage (predStage stage)) @@ -492,13 +490,13 @@ generateSettings settingsFile = do , ("target has GNU nonexec stack", queryTarget stage (yesNo . Toolchain.tgtSupportsGnuNonexecStack)) , ("target has .ident directive", queryTarget stage (yesNo . Toolchain.tgtSupportsIdentDirective)) , ("target has subsections via symbols", queryTarget stage (yesNo . Toolchain.tgtSupportsSubsectionsViaSymbols)) - , ("target has libm", expr $ lookupSystemConfig "target-has-libm") + , ("target has libm", yesNo <$> staged (buildFlag TargetHasLibm)) , ("Unregisterised", queryTarget stage (yesNo . tgtUnregisterised)) , ("LLVM target", queryTarget stage tgtLlvmTarget) - , ("LLVM llc command", expr $ settingsFileSetting ToolchainSetting_LlcCommand) - , ("LLVM opt command", expr $ settingsFileSetting ToolchainSetting_OptCommand) - , ("LLVM llvm-as command", expr $ settingsFileSetting ToolchainSetting_LlvmAsCommand) - , ("Use inplace MinGW toolchain", expr $ settingsFileSetting ToolchainSetting_DistroMinGW) + , ("LLVM llc command", staged $ settingsFileSetting ToolchainSetting_LlcCommand) + , ("LLVM opt command", staged $ settingsFileSetting ToolchainSetting_OptCommand) + , ("LLVM llvm-as command", staged $ settingsFileSetting ToolchainSetting_LlvmAsCommand) + , ("Use inplace MinGW toolchain", staged $ settingsFileSetting ToolchainSetting_DistroMinGW) , ("Use interpreter", expr $ yesNo <$> ghcWithInterpreter stage) , ("Support SMP", expr $ yesNo <$> targetSupportsSMP stage) @@ -506,7 +504,7 @@ generateSettings settingsFile = do , ("Tables next to code", queryTarget stage (yesNo . tgtTablesNextToCode)) , ("Leading underscore", queryTarget stage (yesNo . tgtSymbolsHaveLeadingUnderscore)) , ("Use LibFFI", expr $ yesNo <$> targetUseLibffiForAdjustors stage) - , ("RTS expects libdw", yesNo <$> getFlag UseLibdw) + , ("RTS expects libdw", yesNo <$> staged (buildFlag UseLibdw)) , ("Relative Global Package DB", pure rel_pkg_db) ] let showTuple (k, v) = "(" ++ show k ++ ", " ++ show v ++ ")" ===================================== hadrian/src/Rules/Gmp.hs ===================================== @@ -17,7 +17,7 @@ import Settings.Builders.Common (cArgs, getStagedCCFlags) -- their paths. gmpObjects :: Stage -> Action [FilePath] gmpObjects s = do - isInTree <- flag GmpInTree + isInTree <- buildFlag GmpInTree s if not isInTree then return [] else do @@ -65,8 +65,9 @@ gmpRules = do packageP = takeDirectory buildP librariesP = takeDirectory packageP stageP = takeDirectory librariesP + stage <- parsePath parseStage "" (takeFileName stageP) - isInTree <- flag GmpInTree + isInTree <- buildFlag GmpInTree stage if isInTree then do ===================================== hadrian/src/Rules/Libffi.hs ===================================== @@ -87,7 +87,7 @@ libffiContext stage = do -- | The name of the library libffiName :: Expr String libffiName = do - useSystemFfi <- expr (flag UseSystemFfi) + useSystemFfi <- staged (buildFlag UseSystemFfi) if useSystemFfi then pure "ffi" else libffiLocalName Nothing @@ -118,8 +118,8 @@ libffiHeaderDir stage = do path <- libffiBuildPath stage return $ path -/- "inst/include" -libffiSystemHeaderDir :: Action FilePath -libffiSystemHeaderDir = setting FfiIncludeDir +libffiSystemHeaderDir :: Stage -> Action FilePath +libffiSystemHeaderDir = buildSetting FfiIncludeDir fixLibffiMakefile :: FilePath -> String -> String fixLibffiMakefile top = ===================================== hadrian/src/Rules/Rts.hs ===================================== @@ -54,9 +54,9 @@ withLibffi stage action = needLibffi stage -- See Note [Packaging libffi headers] in GHC.Driver.CodeOutput. copyLibffiHeader :: Stage -> FilePath -> Action () copyLibffiHeader stage header = do - useSystemFfi <- flag UseSystemFfi + useSystemFfi <- buildFlag UseSystemFfi stage (fromStr, headerDir) <- if useSystemFfi - then ("system",) <$> libffiSystemHeaderDir + then ("system",) <$> libffiSystemHeaderDir stage else needLibffi stage >> ("custom",) <$> libffiHeaderDir stage copyFile @@ -115,7 +115,7 @@ rtsLibffiLibrary stage way = do name <- interpretInContext ((rtsContext stage) { way = way }) libffiName suf <- if wayUnit Dynamic way then do - extension <- setting DynamicExtension -- e.g., .dll or .so + extension <- buildSetting DynamicExtension stage -- e.g., .dll or .so let suffix = waySuffix (removeWayUnit Dynamic way) return (suffix ++ extension) -- Static suffix @@ -129,7 +129,7 @@ rtsLibffiLibrary stage way = do needRtsLibffiTargets :: Stage -> Action [FilePath] needRtsLibffiTargets stage = do rtsPath <- rtsBuildPath stage - useSystemFfi <- flag UseSystemFfi + useSystemFfi <- buildFlag UseSystemFfi stage jsTarget <- isJsTarget stage -- Header files (in the rts build dir). ===================================== hadrian/src/Settings/Builders/Cabal.hs ===================================== @@ -203,11 +203,11 @@ configureArgs cFlags' ldFlags' = do mconcat [ conf "CFLAGS" cFlags , conf "LDFLAGS" ldFlags - , conf "--with-iconv-includes" $ arg =<< getSetting IconvIncludeDir - , conf "--with-iconv-libraries" $ arg =<< getSetting IconvLibDir - , conf "--with-gmp-includes" $ arg =<< getSetting GmpIncludeDir - , conf "--with-gmp-libraries" $ arg =<< getSetting GmpLibDir - , conf "--with-curses-libraries" $ arg =<< getSetting CursesLibDir + , conf "--with-iconv-includes" $ arg =<< staged (buildSetting IconvIncludeDir) + , conf "--with-iconv-libraries" $ arg =<< staged (buildSetting IconvLibDir) + , conf "--with-gmp-includes" $ arg =<< staged (buildSetting GmpIncludeDir) + , conf "--with-gmp-libraries" $ arg =<< staged (buildSetting GmpLibDir) + , conf "--with-curses-libraries" $ arg =<< staged (buildSetting CursesLibDir) , conf "--host" $ arg =<< flip queryTarget targetPlatformTriple . predStage' =<< getStage , conf "--with-cc" $ arg =<< getBuilderPath . (Cc CompileC) =<< getStage , ghcVersionH ===================================== hadrian/src/Settings/Builders/Common.hs ===================================== @@ -51,9 +51,9 @@ cppArgs = mempty cWarnings :: Args cWarnings = mconcat [ arg "-Wall" - , flag CcLlvmBackend ? arg "-Wno-unknown-pragmas" - , notM (flag CcLlvmBackend) ? not windowsHost ? arg "-Werror=unused-but-set-variable" - , notM (flag CcLlvmBackend) ? arg "-Wno-error=inline" ] + , staged (buildFlag CcLlvmBackend) ? arg "-Wno-unknown-pragmas" + , notM (staged (buildFlag CcLlvmBackend)) ? not windowsHost ? arg "-Werror=unused-but-set-variable" + , notM (staged (buildFlag CcLlvmBackend)) ? arg "-Wno-error=inline" ] packageDatabaseArgs :: Args packageDatabaseArgs = do ===================================== hadrian/src/Settings/Builders/Ghc.hs ===================================== @@ -102,7 +102,7 @@ ghcLinkArgs = builder (Ghc LinkHs) ? do context <- getContext distPath <- expr (Context.distDynDir context) - useSystemFfi <- expr (flag UseSystemFfi) + useSystemFfi <- staged (buildFlag UseSystemFfi) buildPath <- getBuildPath libffiName' <- libffiName debugged <- buildingCompilerStage' . ghcDebugged =<< expr flavour ===================================== hadrian/src/Settings/Builders/Hsc2Hs.hs ===================================== @@ -11,7 +11,7 @@ hsc2hsBuilderArgs :: Args hsc2hsBuilderArgs = builder Hsc2Hs ? do stage <- getStage ccPath <- getBuilderPath $ Cc CompileC stage - gmpDir <- getSetting GmpIncludeDir + gmpDir <- staged (buildSetting GmpIncludeDir) top <- expr topDirectory hArch <- queryHost queryArch hOs <- queryHost queryOS ===================================== hadrian/src/Settings/Builders/RunTest.hs ===================================== @@ -113,9 +113,9 @@ inTreeCompilerArgs stg = do platform <- queryTargetTarget ghcStage targetPlatformTriple wordsize <- show @Int . (*8) <$> queryTargetTarget ghcStage (wordSize2Bytes . tgtWordSize) - llc_cmd <- settingsFileSetting ToolchainSetting_LlcCommand - llvm_as_cmd <- settingsFileSetting ToolchainSetting_LlvmAsCommand - have_llvm <- liftIO (all isJust <$> mapM findExecutable [llc_cmd, llvm_as_cmd]) + llc_cmd <- settingsFileSetting ToolchainSetting_LlcCommand ghcStage + _llvm_as_cmd <- settingsFileSetting ToolchainSetting_LlvmAsCommand ghcStage + have_llvm <- liftIO (isJust <$> findExecutable llc_cmd) top <- topDirectory ===================================== hadrian/src/Settings/Packages.hs ===================================== @@ -30,10 +30,10 @@ packageArgs = do compilerStageOption f = buildingCompilerStage' . f =<< expr flavour - cursesIncludeDir <- getSetting CursesIncludeDir - cursesLibraryDir <- getSetting CursesLibDir - ffiIncludeDir <- getSetting FfiIncludeDir - ffiLibraryDir <- getSetting FfiLibDir + cursesIncludeDir <- staged (buildSetting CursesIncludeDir) + cursesLibraryDir <- staged (buildSetting CursesLibDir) + ffiIncludeDir <- staged (buildSetting FfiIncludeDir) + ffiLibraryDir <- staged (buildSetting FfiLibDir) stageVersion <- readVersion <$> (expr $ ghcVersionStage stage) mconcat @@ -80,14 +80,14 @@ packageArgs = do [ andM [expr (ghcWithInterpreter stage), notStage0] `cabalFlag` "internal-interpreter" , notM cross `cabalFlag` "terminfo" , arg "-build-tool-depends" - , flag UseLibzstd `cabalFlag` "with-libzstd" + , staged (buildFlag UseLibzstd) `cabalFlag` "with-libzstd" -- ROMES: While the boot compiler is not updated wrt -this-unit-id -- not being fixed to `ghc`, when building stage0, we must set -- -this-unit-id to `ghc` because the boot compiler expects that. -- We do it through a cabal flag in ghc.cabal , stageVersion < makeVersion [9,8,1] ? arg "+hadrian-stage0" - , flag StaticLibzstd `cabalFlag` "static-libzstd" , stage0 `cabalFlag` "bootstrap" + , staged (buildFlag StaticLibzstd) `cabalFlag` "static-libzstd" ] , builder (Haddock BuildPackage) ? arg ("--optghc=-I" ++ path) ] @@ -117,9 +117,9 @@ packageArgs = do -------------------------------- ghcPrim ------------------------------- , package ghcPrim ? mconcat - [ builder (Cabal Flags) ? flag NeedLibatomic `cabalFlag` "need-atomic" + [ builder (Cabal Flags) ? staged (buildFlag NeedLibatomic) `cabalFlag` "need-atomic" - , builder (Cc CompileC) ? (not <$> flag CcLlvmBackend) ? + , builder (Cc CompileC) ? (not <$> staged (buildFlag CcLlvmBackend)) ? input "**/cbits/atomic.c" ? arg "-Wno-sync-nand" ] -------------------------------- ghcBoot ------------------------------ @@ -239,8 +239,8 @@ packageArgs = do ghcBignumArgs :: Args ghcBignumArgs = package ghcBignum ? do -- These are only used for non-in-tree builds. - librariesGmp <- getSetting GmpLibDir - includesGmp <- getSetting GmpIncludeDir + librariesGmp <- staged (buildSetting GmpLibDir) + includesGmp <- staged (buildSetting GmpIncludeDir) backend <- getBignumBackend check <- getBignumCheck @@ -263,15 +263,15 @@ ghcBignumArgs = package ghcBignum ? do -- enable in-tree support: don't depend on external "gmp" -- library - , flag GmpInTree ? arg "--configure-option=--with-intree-gmp" + , staged (buildFlag GmpInTree) ? arg "--configure-option=--with-intree-gmp" -- prefer framework over library (on Darwin) - , flag GmpFrameworkPref ? + , staged (buildFlag GmpFrameworkPref) ? arg "--configure-option=--with-gmp-framework-preferred" -- Ensure that the ghc-bignum package registration includes -- knowledge of the system gmp's library and include directories. - , notM (flag GmpInTree) ? cabalExtraDirs includesGmp librariesGmp + , notM (staged (buildFlag GmpInTree)) ? cabalExtraDirs includesGmp librariesGmp ] ] _ -> mempty @@ -300,15 +300,15 @@ rtsPackageArgs = package rts ? do way <- getWay path <- getBuildPath top <- expr topDirectory - useSystemFfi <- getFlag UseSystemFfi - ffiIncludeDir <- getSetting FfiIncludeDir - ffiLibraryDir <- getSetting FfiLibDir - libdwIncludeDir <- getSetting LibdwIncludeDir - libdwLibraryDir <- getSetting LibdwLibDir - libnumaIncludeDir <- getSetting LibnumaIncludeDir - libnumaLibraryDir <- getSetting LibnumaLibDir - libzstdIncludeDir <- getSetting LibZstdIncludeDir - libzstdLibraryDir <- getSetting LibZstdLibDir + useSystemFfi <- staged (buildFlag UseSystemFfi) + ffiIncludeDir <- staged (buildSetting FfiIncludeDir) + ffiLibraryDir <- staged (buildSetting FfiLibDir) + libdwIncludeDir <- staged (buildSetting LibdwIncludeDir) + libdwLibraryDir <- staged (buildSetting LibdwLibDir) + libnumaIncludeDir <- staged (buildSetting LibnumaIncludeDir) + libnumaLibraryDir <- staged (buildSetting LibnumaLibDir) + libzstdIncludeDir <- staged (buildSetting LibZstdIncludeDir) + libzstdLibraryDir <- staged (buildSetting LibZstdLibDir) -- Arguments passed to GHC when compiling C and .cmm sources. @@ -402,10 +402,10 @@ rtsPackageArgs = package rts ? do -- any warnings in the module. See: -- https://gitlab.haskell.org/ghc/ghc/wikis/working-conventions#Warnings - , (not <$> flag CcLlvmBackend) ? + , (not <$> staged (buildFlag CcLlvmBackend)) ? inputs ["**/Compact.c"] ? arg "-finline-limit=2500" - , input "**/RetainerProfile.c" ? flag CcLlvmBackend ? + , input "**/RetainerProfile.c" ? staged (buildFlag CcLlvmBackend) ? arg "-Wno-incompatible-pointer-types" ] @@ -415,18 +415,18 @@ rtsPackageArgs = package rts ? do , any (wayUnit Debug) rtsWays `cabalFlag` "debug" , any (wayUnit Dynamic) rtsWays `cabalFlag` "dynamic" , any (wayUnit Threaded) rtsWays `cabalFlag` "threaded" - , flag UseLibm `cabalFlag` "libm" - , flag UseLibrt `cabalFlag` "librt" - , flag UseLibdl `cabalFlag` "libdl" + , staged (buildFlag UseLibm) `cabalFlag` "libm" + , staged (buildFlag UseLibrt) `cabalFlag` "librt" + , staged (buildFlag UseLibdl) `cabalFlag` "libdl" , useSystemFfi `cabalFlag` "use-system-libffi" , targetUseLibffiForAdjustors stage `cabalFlag` "libffi-adjustors" - , flag UseLibpthread `cabalFlag` "need-pthread" - , flag UseLibbfd `cabalFlag` "libbfd" - , flag NeedLibatomic `cabalFlag` "need-atomic" - , flag UseLibdw `cabalFlag` "libdw" - , flag UseLibnuma `cabalFlag` "libnuma" - , flag UseLibzstd `cabalFlag` "libzstd" - , flag StaticLibzstd `cabalFlag` "static-libzstd" + , staged (buildFlag UseLibpthread) `cabalFlag` "need-pthread" + , staged (buildFlag UseLibbfd ) `cabalFlag` "libbfd" + , staged (buildFlag NeedLibatomic) `cabalFlag` "need-atomic" + , staged (buildFlag UseLibdw ) `cabalFlag` "libdw" + , staged (buildFlag UseLibnuma ) `cabalFlag` "libnuma" + , staged (buildFlag UseLibzstd ) `cabalFlag` "libzstd" + , staged (buildFlag StaticLibzstd) `cabalFlag` "static-libzstd" , queryTargetTarget stage tgtSymbolsHaveLeadingUnderscore `cabalFlag` "leading-underscore" , ghcUnreg `cabalFlag` "unregisterised" , ghcEnableTNC `cabalFlag` "tables-next-to-code" @@ -447,7 +447,7 @@ rtsPackageArgs = package rts ? do , builder HsCpp ? pure [ "-DTOP=" ++ show top ] - , builder HsCpp ? flag UseLibdw ? arg "-DUSE_LIBDW" ] + , builder HsCpp ? staged (buildFlag UseLibdw) ? arg "-DUSE_LIBDW" ] -- Compile various performance-critical pieces *without* -fPIC -dynamic -- even when building a shared library. If we don't do this, then the ===================================== hadrian/src/Settings/Warnings.hs ===================================== @@ -10,16 +10,15 @@ import Packages -- | Default Haskell warning-related arguments. defaultGhcWarningsArgs :: Args defaultGhcWarningsArgs = do - stage <- getStage mconcat [ notStage0 ? arg "-Wnoncanonical-monad-instances" - , notM (flag CcLlvmBackend) ? arg "-optc-Wno-error=inline" - , flag CcLlvmBackend ? arg "-optc-Wno-unknown-pragmas" + , notM (staged (buildFlag CcLlvmBackend)) ? arg "-optc-Wno-error=inline" + , staged (buildFlag CcLlvmBackend) ? arg "-optc-Wno-unknown-pragmas" -- Cabal can seemingly produce filepaths with incorrect case on filesystems -- with case-insensitive names. Ignore such issues for now as they seem benign. -- See #17798. - , isOsxTarget stage ? arg "-optP-Wno-nonportable-include-path" - , isWinTarget stage ? arg "-optP-Wno-nonportable-include-path" + , staged isOsxTarget ? arg "-optP-Wno-nonportable-include-path" + , staged isWinTarget ? arg "-optP-Wno-nonportable-include-path" ] -- | Package-specific warnings-related arguments, mostly suppressing various warnings. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f3e44e88a241706a07b9ad363f409eba5f3f34d3...4a8b0a9ac64a3e6afb701ea809684230a444f8d1 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f3e44e88a241706a07b9ad363f409eba5f3f34d3...4a8b0a9ac64a3e6afb701ea809684230a444f8d1 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 12:34:57 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 08:34:57 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] Disable testsuite part for now Message-ID: <66f554f1a7066_23519725349c484f9@gitlab.mail> Matthew Pickering pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 5fc49999 by Matthew Pickering at 2024-09-26T13:34:46+01:00 Disable testsuite part for now - - - - - 1 changed file: - .gitlab/ci.sh Changes: ===================================== .gitlab/ci.sh ===================================== @@ -614,13 +614,6 @@ function test_hadrian() { "runtest.opts+=${RUNTEST_ARGS:-}" \ "runtest.opts+=--unexpected-output-dir=$TOP/unexpected-test-output" \ || fail "hadrian main testsuite targetting $CROSS_TARGET" - elif [[ -n "${CROSS_TARGET:-}" ]] && [[ "${CROSS_TARGET:-}" == *"riscv64"* ]]; then - run_hadrian \ - test \ - --summary-junit=./junit.xml \ - "runtest.opts+=${RUNTEST_ARGS:-}" \ - "runtest.opts+=--unexpected-output-dir=$TOP/unexpected-test-output" \ - || fail "hadrian main testsuite targetting $CROSS_TARGET" elif [ -n "${CROSS_TARGET:-}" ]; then local instdir="$TOP/_build/install" local test_compiler="$instdir/bin/${cross_prefix}ghc$exe" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5fc49999c79e8193160090b7f680a204e0d8a5d5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5fc49999c79e8193160090b7f680a204e0d8a5d5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 12:49:16 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 26 Sep 2024 08:49:16 -0400 Subject: [Git][ghc/ghc][wip/romes/exceptions-propagate] 28 commits: ci: Run abi-test on test-abi label Message-ID: <66f5584cbde6e_235197493af45561@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exceptions-propagate at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 7d6a86f0 by Rodrigo Mesquita at 2024-09-26T13:43:37+01:00 exceptions: Improve the message layout as per #285 This commit fixes the layout of the additional information included when displaying an exception, namely the type of the exception. It also fixes the default handler's heading message to work well together with the improved display message of SomeException. - - - - - 70088cab by Rodrigo Mesquita at 2024-09-26T13:43:38+01:00 Display type and callstack of exception on handler This commit changes the Exception instance of SomeException to *simply* display the underlying exception in `displayException`. The augmented exception message that included the type and backtrace of the exception are now only printed on a call to `displayExceptionWithInfo`. At a surface level, existing programs should behave the same since the `uncaughtExceptionHandler`, which is responsible for printing out uncaught exceptions to the user, will use `displayExceptionWithInfo` by default. However, unlike the instance's `displayException` method, the `uncaughtExceptionHandler` can be overriden with `setUncaughtExceptionHandler`. This makes the extra information opt-in without fixing it the instance, which can be valuable if your program wants to display uncaught exceptions to users in a user-facing way (ie without backtraces). This is what was originally agreed for CLC#231 or CLC#261 with regard to the type of the exception information. The call stack also becoming part of the default handler rather than the Exception instance is an ammendment to CLC#164. Discussion of the ammendment is part of CLC#285. - - - - - f83e832a by Matthew Pickering at 2024-09-26T13:43:38+01:00 Fix toException method for ExceptionWithContext Fixes #25235 - - - - - 57ad3910 by Matthew Pickering at 2024-09-26T13:43:38+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. - - - - - fc19fbab by Rodrigo Mesquita at 2024-09-26T13:43:38+01:00 Add unicode drawing lines - - - - - c0a2cb44 by Rodrigo Mesquita at 2024-09-26T13:44:41+01:00 Do #25283 - - - - - 9eecc912 by Rodrigo Mesquita at 2024-09-26T13:48:59+01:00 Freeze call stack in error throwing functions - - - - - 27 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fa66a31a747a8d71c8369124a8756ec9f4900fd0...9eecc912251828a4290e9f71e525bbae662d2e0a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fa66a31a747a8d71c8369124a8756ec9f4900fd0...9eecc912251828a4290e9f71e525bbae662d2e0a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 13:19:28 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 09:19:28 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] fix distrib/configure file Message-ID: <66f55f60523f4_2351977196d4607fc@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: e608df5c by Matthew Pickering at 2024-09-26T14:19:15+01:00 fix distrib/configure file - - - - - 3 changed files: - hadrian/cfg/system.config.in - hadrian/src/Oracles/Setting.hs - hadrian/src/Rules/Generate.hs Changes: ===================================== hadrian/cfg/system.config.in ===================================== @@ -50,6 +50,8 @@ use-ghc-toolchain = @EnableGhcToolchain@ # And we can reconstruct the platform info using targetPlatformTriple # Q: What is TargetPlatformFull? target-platform-full = @TargetPlatformFull@ +build-platform-full = @BuildPlatformFull@ +host-platform-full = @HostPlatformFull@ cross-compiling = @CrossCompiling@ ===================================== hadrian/src/Oracles/Setting.hs ===================================== @@ -59,6 +59,8 @@ data ProjectSetting = | ProjectPatchLevel2 | SystemGhc | TargetPlatformFull + | BuildPlatformFull + | HostPlatformFull | BourneShell | EmsdkVersion @@ -121,6 +123,8 @@ setting key = lookupSystemConfig $ case key of ProjectPatchLevel2 -> "project-patch-level2" SystemGhc -> "system-ghc" TargetPlatformFull -> "target-platform-full" + BuildPlatformFull -> "build-platform-full" + HostPlatformFull -> "host-platform-full" BourneShell -> "bourne-shell" EmsdkVersion -> "emsdk-version" ===================================== hadrian/src/Rules/Generate.hs ===================================== @@ -403,12 +403,17 @@ bindistRules = do , interpolateVar "TablesNextToCode" $ yesNo <$> getTarget tgtTablesNextToCode , interpolateVar "TargetHasLibm" $ yesNo <$> interp (staged (buildFlag TargetHasLibm)) , interpolateVar "TargetPlatform" $ getTarget targetPlatformTriple + , interpolateVar "BuildPlatform" $ interp $ queryBuild targetPlatformTriple + , interpolateVar "HostPlatform" $ interp $ queryHost targetPlatformTriple , interpolateVar "TargetWordBigEndian" $ getTarget isBigEndian , interpolateVar "TargetWordSize" $ getTarget wordSize , interpolateVar "Unregisterised" $ yesNo <$> getTarget tgtUnregisterised , interpolateVar "UseLibdw" $ fmap yesNo $ interp $ staged (buildFlag UseLibdw) , interpolateVar "UseLibffiForAdjustors" $ yesNo <$> getTarget tgtUseLibffiForAdjustors , interpolateVar "GhcWithSMP" $ yesNo <$> targetSupportsSMP Stage2 + , interpolateVar "TargetPlatformFull" (setting TargetPlatformFull) + , interpolateVar "BuildPlatformFull" (setting BuildPlatformFull) + , interpolateVar "HostPlatformFull" (setting HostPlatformFull) ] where interp = interpretInContext (semiEmptyTarget Stage2) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e608df5ced6b207e9422f1401b139bf4f505e3f7 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e608df5ced6b207e9422f1401b139bf4f505e3f7 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 13:22:30 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 09:22:30 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/llvm_19 Message-ID: <66f560165c2ce_2351976bce5c620c6@gitlab.mail> Matthew Pickering pushed new branch wip/llvm_19 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/llvm_19 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 13:58:52 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Thu, 26 Sep 2024 09:58:52 -0400 Subject: [Git][ghc/ghc][wip/T25281] 25 commits: ci: Run abi-test on test-abi label Message-ID: <66f5689c1073a_235197ac6098791e0@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - caf1993b by Simon Peyton Jones at 2024-09-26T10:09:21+01:00 Just a start on #25281 - - - - - 6dd9798a by Simon Peyton Jones at 2024-09-26T10:09:21+01:00 Wibbles - - - - - e943aafb by Simon Peyton Jones at 2024-09-26T10:09:21+01:00 Compiles now - - - - - e10f9395 by Simon Peyton Jones at 2024-09-26T14:58:40+01:00 Wibbles - - - - - 27 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fd8fb94e0a7ee13c962102433912c5bf27579f29...e10f9395359caf4a6b781eb552a6775c298cc6ee -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fd8fb94e0a7ee13c962102433912c5bf27579f29...e10f9395359caf4a6b781eb552a6775c298cc6ee You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 14:11:45 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Thu, 26 Sep 2024 10:11:45 -0400 Subject: [Git][ghc/ghc][wip/T25281] Comments Message-ID: <66f56ba1ab2ae_235197ab77a0861b1@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: 8156fe53 by Simon Peyton Jones at 2024-09-26T15:11:27+01:00 Comments - - - - - 3 changed files: - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/Tc/Instance/Class.hs Changes: ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -558,6 +558,20 @@ dsExpr (SectionR x _ _) = dataConCantHappen x * * ********************************************************************* -} +{- Note [Desugaring applications] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +When we come across an application (f e1 .. en) we collect up +all the desugared arguments, and then dispatch on the function f. +(Including the nullary case where n=0.) + +There are several special cases to handle + +* HsRecSel: a record selector gets warnings if it might fail. +* HsVar: special magic for `noinline` +* HsVar: special magic for `seq` + +-} + dsApp :: HsExpr GhcTc -> DsM CoreExpr dsApp e = ds_app e [] [] @@ -621,8 +635,8 @@ ds_app (HsRecSel _ fld_occ@(FieldOcc { foExt = fun_id })) hs_args core_args diagnosticDs $ DsIncompleteRecordSelector (idName fun_id) cons_trimmed (cons_trimmed /= cons_wo_field) - -- Type-based check - -- ToDo: explain + -- Type-based check. See GHC.HsToCore.Pmc (IRS4) + -- in Note [Detecting incomplete record selectors] ; case filterOut isTypeArg core_args of (arg:_) -> pmcRecSel fun_id arg _ -> return () ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -198,9 +198,8 @@ pmcMatches origin ctxt vars matches = {-# SCC "pmcMatches" #-} do formatReportWarnings ReportMatchGroup ctxt vars result) return (NE.toList (ldiMatchGroup (cr_ret result))) -{- -Note [Detecting incomplete record selectors] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +{- Note [Detecting incomplete record selectors] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A record selector occurrence is incomplete iff. it could fail due to being applied to a data type constructor not present for this record field. @@ -211,16 +210,16 @@ e.g. There are 4 parts to detecting and warning about incomplete record selectors to consider: - - Computing which constructors a general application of a record field will succeed on, - and which ones it will fail on. This is stored in the `sel_cons` field of - `IdDetails` datatype, which is a part of an `Id` and calculated when renaming a - record selector in `mkOneRecordSelector` +(IRS1) Computing which constructors a general application of a record field will + succeed on, and which ones it will fail on. This is stored in the `sel_cons` + field of `IdDetails` datatype, which is a part of an `Id` and calculated + when renaming a record selector in `mkOneRecordSelector` - - Emitting a warning whenever a `HasField` constraint is solved. +(IRS2) Emitting a warning whenever a `HasField` constraint is solved. This is checked in `matchHasField` and emitted only for when the constraint is resolved with an implicit instance rather than a custom one (since otherwise the warning will be emitted in - the custom implementation anyways) + the custom implementation anyways) e.g. g :: HasField "x" t Int => t -> Int @@ -229,7 +228,7 @@ incomplete record selectors to consider: f :: T -> Int f = g -- warning will be emitted here - - Emitting a warning for a general occurrence of the record selector +(IRS3) Emitting a warning for a general occurrence of the record selector This is done during the renaming of a `HsRecSel` expression in `dsExpr` and simply pulls the information about incompleteness from the `Id` @@ -237,7 +236,7 @@ incomplete record selectors to consider: l :: T -> Int l a = x a -- warning will be emitted here - - Emitting a warning for a record selector `sel` applied to a variable `y`. +(IRS4) Emitting a warning for a record selector `sel` applied to a variable `y`. In that case we want to use the long-distance information from the pattern match checker to rule out impossible constructors (See Note [Long-distance information]). We first add constraints to @@ -255,6 +254,7 @@ incomplete record selectors to consider: pmcRecSel :: Id -- ^ Id of the selector -> CoreExpr -- ^ Core expression of the argument to the selector -> DsM () +-- See (IRS4) in Note [Detecting incomplete record selectors] pmcRecSel sel_id arg | RecSelId{ sel_cons = (cons_w_field, _ : _) } <- idDetails sel_id = do !missing <- getLdiNablas ===================================== compiler/GHC/Tc/Instance/Class.hs ===================================== @@ -1298,7 +1298,8 @@ matchHasField dflags short_cut clas tys ; unless (null $ snd $ sel_cons $ idDetails sel_id) $ addDiagnostic $ TcRnHasFieldResolvedIncomplete name -- Only emit an incomplete selector warning if it's an implicit instance - -- See Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + -- See (IRS2) in Note [Detecting incomplete record selectors] + -- in GHC.HsToCore.Pmc ; return OneInst { cir_new_theta = theta , cir_mk_ev = mk_ev , cir_canonical = EvCanonical View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8156fe535b2318c6720ed9260b209756369309f5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8156fe535b2318c6720ed9260b209756369309f5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 14:20:20 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 10:20:20 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/wasm-perf-notes Message-ID: <66f56da49165e_235197d349ec87493@gitlab.mail> Matthew Pickering pushed new branch wip/wasm-perf-notes at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/wasm-perf-notes You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 14:34:26 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 26 Sep 2024 10:34:26 -0400 Subject: [Git][ghc/ghc][wip/romes/det-llvm] 13 commits: Add Given injectivity for built-in type families Message-ID: <66f570f29bb93_235197e8a73889447@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/det-llvm at Glasgow Haskell Compiler / GHC Commits: a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 51ef077e by Rodrigo Mesquita at 2024-09-26T13:32:47+01:00 determinism: Deterministic MonadGetUnique LlvmM Update LlvmM to thread a unique deterministic supply (using UniqDSMT), and use it in the MonadGetUnique instance. This makes uniques sampled from LlvmM deterministic, which guarantees object determinism with -fllvm. Fixes #25274 - - - - - 14 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/CmmToLlvm.hs - compiler/GHC/CmmToLlvm/Base.hs - compiler/GHC/Core/Coercion.hs - compiler/GHC/Core/Coercion.hs-boot - compiler/GHC/Core/Coercion/Axiom.hs - compiler/GHC/Core/Coercion/Opt.hs - compiler/GHC/Core/DataCon.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/FamInstEnv.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Opt/ConstantFold.hs - compiler/GHC/Core/Rules.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/165f0b4b9f8548d9cba99a8c8fd4bd70a36f178c...51ef077ed8405971e5082d5a52c29f889e9dc9c4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/165f0b4b9f8548d9cba99a8c8fd4bd70a36f178c...51ef077ed8405971e5082d5a52c29f889e9dc9c4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 15:00:13 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 26 Sep 2024 11:00:13 -0400 Subject: [Git][ghc/ghc][wip/romes/exceptions-propagate] 2 commits: Do #25283 Message-ID: <66f576fd86ac2_5052bc45874588@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exceptions-propagate at Glasgow Haskell Compiler / GHC Commits: 08995502 by Rodrigo Mesquita at 2024-09-26T15:58:41+01:00 Do #25283 - - - - - eedb64d9 by Rodrigo Mesquita at 2024-09-26T15:58:42+01:00 Freeze call stack in error throwing functions - - - - - 30 changed files: - compiler/GHC/Utils/Panic/Plain.hs - libraries/base/src/GHC/Exception.hs - libraries/base/src/GHC/Stack.hs - libraries/base/tests/IO/T4808.stderr - libraries/base/tests/IO/mkdirExists.stderr - libraries/base/tests/IO/openFile002.stderr - libraries/base/tests/IO/withBinaryFile001.stderr - libraries/base/tests/IO/withFile001.stderr - libraries/base/tests/IO/withFileBlocking001.stderr - libraries/base/tests/T15349.stderr - libraries/base/tests/T19288.stderr - libraries/base/tests/T24807.stderr - libraries/ghc-internal/src/GHC/Internal/Exception.hs - libraries/ghc-internal/src/GHC/Internal/IO.hs - libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Stack.hs - libraries/ghc-internal/src/GHC/Internal/Stack.hs-boot - − libraries/ghc-internal/src/GHC/Internal/Stack/CCS.hs-boot - testsuite/tests/arityanal/should_run/T21694a.stderr - testsuite/tests/arityanal/should_run/T24296.stderr - testsuite/tests/codeGen/should_run/T16846.stderr - testsuite/tests/codeGen/should_run/T5626.stderr - testsuite/tests/codeGen/should_run/cgrun016.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/codeGen/should_run/cgrun045.stderr - testsuite/tests/codeGen/should_run/cgrun051.stderr - testsuite/tests/codeGen/should_run/cgrun059.stderr - testsuite/tests/concurrent/should_run/T13330.stderr - testsuite/tests/concurrent/should_run/conc021.stderr - testsuite/tests/deSugar/should_run/Or5.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9eecc912251828a4290e9f71e525bbae662d2e0a...eedb64d9cde72f0da66961945507fd6a364cd28a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9eecc912251828a4290e9f71e525bbae662d2e0a...eedb64d9cde72f0da66961945507fd6a364cd28a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 15:13:45 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Thu, 26 Sep 2024 11:13:45 -0400 Subject: [Git][ghc/ghc][wip/romes/exceptions-propagate] 3 commits: Do #25283 Message-ID: <66f57a29931e_50521b436074875@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exceptions-propagate at Glasgow Haskell Compiler / GHC Commits: 90ce8cba by Rodrigo Mesquita at 2024-09-26T16:13:36+01:00 Do #25283 - - - - - 0cb315ed by Rodrigo Mesquita at 2024-09-26T16:13:36+01:00 Freeze call stack in error throwing functions - - - - - b4bfc7d5 by Rodrigo Mesquita at 2024-09-26T16:13:36+01:00 Accept tests - - - - - 30 changed files: - compiler/GHC/Utils/Panic/Plain.hs - libraries/base/src/GHC/Exception.hs - libraries/base/src/GHC/Stack.hs - libraries/base/tests/IO/T4808.stderr - libraries/base/tests/IO/mkdirExists.stderr - libraries/base/tests/IO/openFile002.stderr - libraries/base/tests/IO/withBinaryFile001.stderr - libraries/base/tests/IO/withBinaryFile002.stderr - libraries/base/tests/IO/withFile001.stderr - libraries/base/tests/IO/withFile002.stderr - libraries/base/tests/IO/withFileBlocking001.stderr - libraries/base/tests/IO/withFileBlocking002.stderr - libraries/base/tests/T15349.stderr - libraries/base/tests/T19288.stderr - libraries/base/tests/T24807.stderr - libraries/ghc-internal/src/GHC/Internal/Exception.hs - libraries/ghc-internal/src/GHC/Internal/IO.hs - libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Stack.hs - libraries/ghc-internal/src/GHC/Internal/Stack.hs-boot - − libraries/ghc-internal/src/GHC/Internal/Stack/CCS.hs-boot - testsuite/tests/arityanal/should_run/T21694a.stderr - testsuite/tests/arityanal/should_run/T24296.stderr - testsuite/tests/codeGen/should_run/T16846.stderr - testsuite/tests/codeGen/should_run/T5626.stderr - testsuite/tests/codeGen/should_run/cgrun016.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/codeGen/should_run/cgrun045.stderr - testsuite/tests/codeGen/should_run/cgrun051.stderr - testsuite/tests/codeGen/should_run/cgrun059.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/eedb64d9cde72f0da66961945507fd6a364cd28a...b4bfc7d578ea3a580b88fcabd20871afaa7eea43 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/eedb64d9cde72f0da66961945507fd6a364cd28a...b4bfc7d578ea3a580b88fcabd20871afaa7eea43 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 15:38:57 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 11:38:57 -0400 Subject: [Git][ghc/ghc][wip/wasm-perf-notes] ci: Push perf notes from wasm jobs Message-ID: <66f58011f1117_5052371694767bd@gitlab.mail> Matthew Pickering pushed to branch wip/wasm-perf-notes at Glasgow Haskell Compiler / GHC Commits: 7e732e3b by Matthew Pickering at 2024-09-26T16:09:18+01:00 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 1 changed file: - .gitlab/ci.sh Changes: ===================================== .gitlab/ci.sh ===================================== @@ -431,7 +431,9 @@ function push_perf_notes() { return fi - if [ -n "${CROSS_TARGET:-}" ] && [ "${CROSS_EMULATOR:-}" != "js-emulator" ]; then + # TODO: Remove this check, see #25299 + # It is easy to forget to update this when testing a new cross platform + if [[ -n "${CROSS_TARGET:-}" ]] && [[ "${CROSS_TARGET:-}" != "*javascript*" ]] && [[ "${CROSS_TARGET:-}" != *"wasm"* ]]; then info "Can't test cross-compiled build." return fi View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7e732e3bc3806f93fa85acc65b79ee09b6ab3ab3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7e732e3bc3806f93fa85acc65b79ee09b6ab3ab3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 15:51:46 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Thu, 26 Sep 2024 11:51:46 -0400 Subject: [Git][ghc/ghc][wip/T25281] More improvements Message-ID: <66f583123823d_50524de54082257@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: 13ba3beb by Simon Peyton Jones at 2024-09-26T16:51:22+01:00 More improvements - - - - - 8 changed files: - compiler/GHC/Core/ConLike.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/TyCl/Utils.hs - compiler/GHC/Types/Id/Info.hs - compiler/GHC/Types/TyThing.hs Changes: ===================================== compiler/GHC/Core/ConLike.hs ===================================== @@ -23,7 +23,6 @@ module GHC.Core.ConLike ( , conLikeFullSig , conLikeResTy , conLikeFieldType - , conLikesWithFields , conLikeIsInfix , conLikeHasBuilder ) where @@ -231,15 +230,6 @@ conLikeFieldType :: ConLike -> FieldLabelString -> Type conLikeFieldType (PatSynCon ps) label = patSynFieldType ps label conLikeFieldType (RealDataCon dc) label = dataConFieldType dc label - --- | The ConLikes that have *all* the given fields -conLikesWithFields :: [ConLike] -> [FieldLabelString] - -> ( [ConLike] -- ConLikes containing the fields - , [ConLike] ) -- ConLikes not containing the fields -conLikesWithFields con_likes lbls = List.partition has_flds con_likes - where has_flds dc = all (has_fld dc) lbls - has_fld dc lbl = any (\ fl -> flLabel fl == lbl) (conLikeFieldLabels dc) - conLikeIsInfix :: ConLike -> Bool conLikeIsInfix (RealDataCon dc) = dataConIsInfix dc conLikeIsInfix (PatSynCon ps) = patSynIsInfix ps ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -619,7 +619,7 @@ ds_app (XExpr (ConLikeTc con tvs tys)) _hs_args core_args ; return (mkApps core_fun core_args) } ds_app (HsRecSel _ fld_occ@(FieldOcc { foExt = fun_id })) hs_args core_args - | RecSelId {sel_cons = (_, cons_wo_field)} <- idDetails fun_id + | RecSelId {sel_cons = RSI { rsi_undef = cons_wo_field}} <- idDetails fun_id = do { dflags <- getDynFlags -- Record selectors are warned about if they are not present in all of the ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -216,27 +216,28 @@ incomplete record selectors to consider: when renaming a record selector in `mkOneRecordSelector` (IRS2) Emitting a warning whenever a `HasField` constraint is solved. - This is checked in `matchHasField` and emitted only for when - the constraint is resolved with an implicit instance rather than a - custom one (since otherwise the warning will be emitted in - the custom implementation anyways) + Implemented in GHC.Tc.Instance.Class.matchHasField` + The warning is emitted only when the constraint is resolved with an + /implicit/ instance rather than a user-supplied one (since otherwise the + warning will be emitted in the custom implementation anyways) e.g. g :: HasField "x" t Int => t -> Int g = getField @"x" f :: T -> Int - f = g -- warning will be emitted here - -(IRS3) Emitting a warning for a general occurrence of the record selector - This is done during the renaming of a `HsRecSel` expression in `dsExpr` - and simply pulls the information about incompleteness from the `Id` + f = g -- Warning will be emitted here when solving the + -- (HasField "x" T Int) constraint arising from + -- the call of g, because `x` is only partial +(IRS3) Emitting a warning for a general occurrence of the record selector. + Implemented in the `HsRecSel` case of `GHC.HsToCore.Expr.ds_app`. e.g. l :: T -> Int l a = x a -- warning will be emitted here (IRS4) Emitting a warning for a record selector `sel` applied to a variable `y`. + Implemented in `pmcRecSel` in this module. In that case we want to use the long-distance information from the pattern match checker to rule out impossible constructors (See Note [Long-distance information]). We first add constraints to @@ -256,27 +257,29 @@ pmcRecSel :: Id -- ^ Id of the selector -> DsM () -- See (IRS4) in Note [Detecting incomplete record selectors] pmcRecSel sel_id arg - | RecSelId{ sel_cons = (cons_w_field, _ : _) } <- idDetails sel_id = do - !missing <- getLdiNablas - - tracePm "pmcRecSel {" (ppr sel_id) - CheckResult{ cr_ret = PmRecSel{ pr_arg_var = arg_id }, cr_uncov = uncov_nablas } - <- unCA (checkRecSel (PmRecSel () arg cons_w_field)) missing - tracePm "}: " $ ppr uncov_nablas - - inhabited <- isInhabited uncov_nablas - when inhabited $ warn_incomplete arg_id uncov_nablas - where - sel_name = varName sel_id - warn_incomplete arg_id uncov_nablas = do - dflags <- getDynFlags - let maxConstructors = maxUncoveredPatterns dflags - unc_examples <- getNFirstUncovered MinimalCover [arg_id] (maxConstructors + 1) uncov_nablas - let cons = [con | unc_example <- unc_examples - , Just (PACA (PmAltConLike con) _ _) <- [lookupSolution unc_example arg_id]] - not_full_examples = length cons == (maxConstructors + 1) - cons' = take maxConstructors cons - diagnosticDs $ DsIncompleteRecordSelector sel_name cons' not_full_examples + | RecSelId{ sel_cons = RSI { rsi_def = cons_w_field, rsi_undef = cons_wo_field } } + <- idDetails sel_id + , not (null cons_wo_field) + = do { !missing <- getLdiNablas + + ; tracePm "pmcRecSel {" (ppr sel_id) + ; CheckResult{ cr_ret = PmRecSel{ pr_arg_var = arg_id }, cr_uncov = uncov_nablas } + <- unCA (checkRecSel (PmRecSel () arg cons_w_field)) missing + ; tracePm "}: " $ ppr uncov_nablas + + ; inhabited <- isInhabited uncov_nablas + ; when inhabited $ warn_incomplete arg_id uncov_nablas } + where + sel_name = varName sel_id + warn_incomplete arg_id uncov_nablas = do + dflags <- getDynFlags + let maxConstructors = maxUncoveredPatterns dflags + unc_examples <- getNFirstUncovered MinimalCover [arg_id] (maxConstructors + 1) uncov_nablas + let cons = [con | unc_example <- unc_examples + , Just (PACA (PmAltConLike con) _ _) <- [lookupSolution unc_example arg_id]] + not_full_examples = length cons == (maxConstructors + 1) + cons' = take maxConstructors cons + diagnosticDs $ DsIncompleteRecordSelector sel_name cons' not_full_examples pmcRecSel _ _ = return () ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -1713,14 +1713,13 @@ tcIdDetails nm _ (IfRecSelId tc _first_con naughty fl) = do { tc' <- either (fmap RecSelData . tcIfaceTyCon) (fmap (RecSelPatSyn . tyThingPatSyn) . tcIfaceDecl False) tc - ; let all_cons = recSelParentCons tc' - cons_partitioned - = conLikesWithFields all_cons [flLabel fl] + ; let all_cons = recSelParentCons tc' + cons_partitioned = conLikesRecSelInfo all_cons [flLabel fl] ; return (RecSelId - { sel_tycon = tc' - , sel_naughty = naughty + { sel_tycon = tc' + , sel_naughty = naughty , sel_fieldLabel = fl { flSelector = nm } - , sel_cons = cons_partitioned } + , sel_cons = cons_partitioned } -- Reconstructed here since we don't want Uniques in the Iface file ) } where ===================================== compiler/GHC/Tc/Instance/Class.hs ===================================== @@ -1295,7 +1295,7 @@ matchHasField dflags short_cut clas tys then do { -- See Note [Unused name reporting and HasField] addUsedGRE AllDeprecationWarnings gre ; keepAlive name - ; unless (null $ snd $ sel_cons $ idDetails sel_id) + ; unless (null $ rsi_undef $ sel_cons $ idDetails sel_id) $ addDiagnostic $ TcRnHasFieldResolvedIncomplete name -- Only emit an incomplete selector warning if it's an implicit instance -- See (IRS2) in Note [Detecting incomplete record selectors] ===================================== compiler/GHC/Tc/TyCl/Utils.hs ===================================== @@ -881,14 +881,15 @@ mkOneRecordSelector all_cons idDetails fl has_sel sel_id = mkExportedLocalId rec_details sel_name sel_ty -- Find a representative constructor, con1 - cons_partitioned@(cons_w_field, _) = conLikesWithFields all_cons [lbl] + rec_sel_info@(RSI { rsi_def = cons_w_field }) + = conLikesRecSelInfo all_cons [lbl] con1 = assert (not (null cons_w_field)) $ head cons_w_field -- Construct the IdDetails rec_details = RecSelId { sel_tycon = idDetails , sel_naughty = is_naughty , sel_fieldLabel = fl - , sel_cons = cons_partitioned } + , sel_cons = rec_sel_info } -- See Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc ===================================== compiler/GHC/Types/Id/Info.hs ===================================== @@ -21,8 +21,10 @@ module GHC.Types.Id.Info ( -- * The IdDetails type IdDetails(..), pprIdDetails, coVarDetails, isCoVarDetails, JoinArity, isJoinIdDetails_maybe, + RecSelParent(..), recSelParentName, recSelFirstConName, recSelParentCons, idDetailsConcreteTvs, + RecSelInfo(..), conLikesRecSelInfo, -- * The IdInfo type IdInfo, -- Abstract @@ -116,6 +118,7 @@ import GHC.StgToCmm.Types (LambdaFormInfo) import Data.Data ( Data ) import Data.Word +import Data.List as List( partition ) -- infixl so you can say (id `set` a `set` b) infixl 1 `setRuleInfo`, @@ -151,12 +154,9 @@ data IdDetails , sel_naughty :: Bool -- True <=> a "naughty" selector which can't actually exist, for example @x@ in: -- data T = forall a. MkT { x :: a } -- See Note [Naughty record selectors] in GHC.Tc.TyCl - , sel_cons :: ([ConLike], [ConLike]) - -- If record selector is not defined for all constructors - -- of a parent type, this is the pair of lists of constructors that - -- it is and is not defined for. Otherwise, it's Nothing. - -- Cached here based on the RecSelParent. - } -- See Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + , sel_cons :: RecSelInfo + -- Partiality info, cached here based on the RecSelParent. + } | DataConWorkId DataCon -- ^ The 'Id' is for a data constructor /worker/ | DataConWrapId DataCon -- ^ The 'Id' is for a data constructor /wrapper/ @@ -224,6 +224,11 @@ data IdDetails -- The [CbvMark] is always empty (and ignored) until after Tidy for ids from the current -- module. +data RecSelInfo + = RSI { rsi_def :: [ConLike] -- Record selector defined for these + , rsi_undef :: [ConLike] -- Record selector not defined for these + } + idDetailsConcreteTvs :: IdDetails -> ConcreteTyVars idDetailsConcreteTvs = \ case PrimOpId _ conc_tvs -> conc_tvs @@ -232,6 +237,16 @@ idDetailsConcreteTvs = \ case DataConWrapId dc -> dataConConcreteTyVars dc _ -> noConcreteTyVars +-- | The ConLikes that have *all* the given fields +conLikesRecSelInfo :: [ConLike] -> [FieldLabelString] -> RecSelInfo +conLikesRecSelInfo con_likes lbls + = RSI { rsi_def = defs, rsi_undef = undefs } + where + !(defs,undefs) = List.partition has_flds con_likes + + has_flds dc = all (has_fld dc) lbls + has_fld dc lbl = any (\ fl -> flLabel fl == lbl) (conLikeFieldLabels dc) + {- Note [CBV Function Ids] ~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===================================== compiler/GHC/Types/TyThing.hs ===================================== @@ -355,12 +355,12 @@ tyThingGREInfo = \case RecSelPatSyn ps -> unitUniqSet $ PatSynName (patSynName ps) RecSelData tc -> let dcs = map RealDataCon $ tyConDataCons tc in - case conLikesWithFields dcs [flLabel fl] of - ([], _) -> pprPanic "tyThingGREInfo: no DataCons with this FieldLabel" $ + case rsi_def (conLikesRecSelInfo dcs [flLabel fl]) of + [] -> pprPanic "tyThingGREInfo: no DataCons with this FieldLabel" $ vcat [ text "id:" <+> ppr id , text "fl:" <+> ppr fl , text "dcs:" <+> ppr dcs ] - (cons, _) -> mkUniqSet $ map conLikeConLikeName cons + cons -> mkUniqSet $ map conLikeConLikeName cons in IAmRecField $ RecFieldInfo { recFieldLabel = fl View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/13ba3beb17d7fbfea216f32006436a352e6fd07a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/13ba3beb17d7fbfea216f32006436a352e6fd07a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 16:08:12 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 12:08:12 -0400 Subject: [Git][ghc/ghc][master] 4 commits: Don't compile `asBox` with -fprof-late Message-ID: <66f586ec31d1f_50526b4cd4957c7@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - 13 changed files: - hadrian/doc/flavours.md - hadrian/src/Flavour.hs - hadrian/src/Settings/Flavours/Release.hs - libraries/ghc-heap/GHC/Exts/Heap/Closures.hs - testsuite/driver/testglobals.py - testsuite/driver/testlib.py - testsuite/tests/codeGen/should_run/cgrun057.stderr - testsuite/tests/profiling/should_run/T12962.prof.sample - testsuite/tests/profiling/should_run/T2552.prof.sample - testsuite/tests/profiling/should_run/T5559.prof.sample - testsuite/tests/profiling/should_run/T5654-O0.prof.sample - testsuite/tests/profiling/should_run/T5654-O1.prof.sample - testsuite/tests/profiling/should_run/T5654b-O0.prof.sample The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/383af0743a46e0961c342261f13a4527ae838873...2b25f9e28ad763739966f63664864b42766da4a2 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/383af0743a46e0961c342261f13a4527ae838873...2b25f9e28ad763739966f63664864b42766da4a2 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 16:08:58 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 12:08:58 -0400 Subject: [Git][ghc/ghc][master] Replace manual string lexing (#25158) Message-ID: <66f5871a1e800_50524d56d4996b2@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 29 changed files: - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - testsuite/driver/testlib.py - testsuite/tests/ghci/prog013/prog013.stderr - testsuite/tests/ghci/scripts/ghci022.stderr - testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr - testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr - testsuite/tests/parser/should_fail/T21843c.stderr - testsuite/tests/parser/should_fail/T21843e.stderr - testsuite/tests/parser/should_fail/T21843f.stderr - testsuite/tests/parser/should_fail/T3751.stderr - testsuite/tests/parser/should_fail/T5425.stderr - testsuite/tests/parser/should_fail/readFail002.stderr - testsuite/tests/parser/should_fail/readFail004.stderr - testsuite/tests/parser/should_fail/readFail005.stderr - testsuite/tests/parser/should_fail/readFail033.stderr - testsuite/tests/parser/unicode/all.T - + testsuite/tests/parser/unicode/lex_unicode_ids.hs - + testsuite/tests/parser/unicode/lex_unispace.hs - testsuite/tests/parser/unicode/utf8_010.stderr - testsuite/tests/parser/unicode/utf8_011.stderr - testsuite/tests/parser/unicode/utf8_020.stderr - testsuite/tests/parser/unicode/utf8_021.stderr - testsuite/tests/parser/unicode/utf8_022.stderr - utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs Changes: ===================================== compiler/GHC/Parser.y ===================================== @@ -728,7 +728,7 @@ are the most common patterns, rewritten as regular expressions for clarity: CHAR { L _ (ITchar _ _) } STRING { L _ (ITstring _ _) } - MULTILINESTRING { L _ (ITmultilinestring _ _) } + STRING_MULTI { L _ (ITstringMulti _ _) } INTEGER { L _ (ITinteger _) } RATIONAL { L _ (ITrational _) } @@ -2357,8 +2357,8 @@ atype :: { LHsType GhcPs } (getCHAR $1) } | STRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGs $1) (getSTRING $1) } - | MULTILINESTRING { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getMULTILINESTRINGs $1) - (getMULTILINESTRING $1) } + | STRING_MULTI { sLLa $1 $> $ HsTyLit noExtField $ HsStrTy (getSTRINGMULTIs $1) + (getSTRINGMULTI $1) } | '_' { sL1a $1 $ mkAnonWildCardTy } -- Type variables are never exported, so `M.tyvar` will be rejected by the renamer. -- We let it pass the parser because the renamer can generate a better error message. @@ -4047,8 +4047,8 @@ literal :: { Located (HsLit GhcPs) } : CHAR { sL1 $1 $ HsChar (getCHARs $1) $ getCHAR $1 } | STRING { sL1 $1 $ HsString (getSTRINGs $1) $ getSTRING $1 } - | MULTILINESTRING { sL1 $1 $ HsMultilineString (getMULTILINESTRINGs $1) - $ getMULTILINESTRING $1 } + | STRING_MULTI { sL1 $1 $ HsMultilineString (getSTRINGMULTIs $1) + $ getSTRINGMULTI $1 } | PRIMINTEGER { sL1 $1 $ HsIntPrim (getPRIMINTEGERs $1) $ getPRIMINTEGER $1 } | PRIMWORD { sL1 $1 $ HsWordPrim (getPRIMWORDs $1) @@ -4154,7 +4154,7 @@ getIPDUPVARID (L _ (ITdupipvarid x)) = x getLABELVARID (L _ (ITlabelvarid _ x)) = x getCHAR (L _ (ITchar _ x)) = x getSTRING (L _ (ITstring _ x)) = x -getMULTILINESTRING (L _ (ITmultilinestring _ x)) = x +getSTRINGMULTI (L _ (ITstringMulti _ x)) = x getINTEGER (L _ (ITinteger x)) = x getRATIONAL (L _ (ITrational x)) = x getPRIMCHAR (L _ (ITprimchar _ x)) = x @@ -4180,7 +4180,7 @@ getVOCURLY (L (RealSrcSpan l _) ITvocurly) = srcSpanStartCol l getINTEGERs (L _ (ITinteger (IL src _ _))) = src getCHARs (L _ (ITchar src _)) = src getSTRINGs (L _ (ITstring src _)) = src -getMULTILINESTRINGs (L _ (ITmultilinestring src _)) = src +getSTRINGMULTIs (L _ (ITstringMulti src _)) = src getPRIMCHARs (L _ (ITprimchar src _)) = src getPRIMSTRINGs (L _ (ITprimstring src _)) = src getPRIMINTEGERs (L _ (ITprimint src _)) = src ===================================== compiler/GHC/Parser/CharClass.hs ===================================== @@ -36,14 +36,14 @@ cDigit = 64 {-# INLINABLE is_ctype #-} is_ctype :: Word8 -> Char -> Bool -is_ctype mask c = (charType c .&. mask) /= 0 +is_ctype mask c = c <= '\127' && (charType c .&. mask) /= 0 is_ident, is_symbol, is_any, is_space, is_lower, is_upper, is_digit, is_alphanum :: Char -> Bool is_ident = is_ctype cIdent is_symbol = is_ctype cSymbol is_any = is_ctype cAny -is_space = \c -> c <= '\x7f' && is_ctype cSpace c -- is_space only works for <= '\x7f' (#3751, #5425) +is_space = is_ctype cSpace is_lower = is_ctype cLower is_upper = is_ctype cUpper is_digit = is_ctype cDigit ===================================== compiler/GHC/Parser/Errors/Ppr.hs ===================================== @@ -157,8 +157,6 @@ instance Diagnostic PsMessage where LexUnknownPragma -> text "unknown pragma" LexErrorInPragma -> text "lexical error in pragma" LexNumEscapeRange -> text "numeric escape sequence out of range" - LexStringCharLit -> text "lexical error in string/character literal" - LexStringCharLitEOF -> text "unexpected end-of-file in string/character literal" LexUnterminatedComment -> text "unterminated `{-'" LexUnterminatedOptions -> text "unterminated OPTIONS pragma" LexUnterminatedQQ -> text "unterminated quasiquotation" ===================================== compiler/GHC/Parser/Errors/Types.hs ===================================== @@ -591,8 +591,6 @@ data LexErr | LexUnknownPragma -- ^ Unknown pragma | LexErrorInPragma -- ^ Lexical error in pragma | LexNumEscapeRange -- ^ Numeric escape sequence out of range - | LexStringCharLit -- ^ Lexical error in string/character literal - | LexStringCharLitEOF -- ^ Unexpected end-of-file in string/character literal | LexUnterminatedComment -- ^ Unterminated `{-' | LexUnterminatedOptions -- ^ Unterminated OPTIONS pragma | LexUnterminatedQQ -- ^ Unterminated quasiquotation ===================================== compiler/GHC/Parser/Lexer.x ===================================== @@ -140,7 +140,8 @@ import GHC.Parser.String -- Any changes here should likely be reflected there. $unispace = \x05 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $nl = [\n\r\f] -$whitechar = [$nl\v\ $unispace] +$space = [\ $unispace] +$whitechar = [$nl \v $space] $white_no_nl = $whitechar # \n -- TODO #8424 $tab = \t @@ -167,6 +168,7 @@ $idchar = [$small $large $digit $uniidchar \'] $unigraphic = \x06 -- Trick Alex into handling Unicode. See Note [Unicode in Alex]. $graphic = [$small $large $symbol $digit $idchar $special $unigraphic \"\'] +$charesc = [a b f n r t v \\ \" \' \&] $binit = 0-1 $octit = 0-7 @@ -213,6 +215,20 @@ $docsym = [\| \^ \* \$] @floating_point = @numspc @decimal \. @decimal @exponent? | @numspc @decimal @exponent @hex_floating_point = @numspc @hexadecimal \. @hexadecimal @bin_exponent? | @numspc @hexadecimal @bin_exponent + at gap = \\ $whitechar+ \\ + at cntrl = $asclarge | \@ | \[ | \\ | \] | \^ | \_ + at ascii = \^ @cntrl | "NUL" | "SOH" | "STX" | "ETX" | "EOT" | "ENQ" | "ACK" + | "BEL" | "BS" | "HT" | "LF" | "VT" | "FF" | "CR" | "SO" | "SI" | "DLE" + | "DC1" | "DC2" | "DC3" | "DC4" | "NAK" | "SYN" | "ETB" | "CAN" + | "EM" | "SUB" | "ESC" | "FS" | "GS" | "RS" | "US" | "SP" | "DEL" +-- N.B. ideally, we would do `@escape # \\ \&` instead of duplicating in @escapechar, +-- which is what the Haskell Report says, but this isn't valid Alex syntax, as only +-- character sets can be subtracted, not strings + at escape = \\ ( $charesc | @ascii | @decimal | o @octal | x @hexadecimal ) + at escapechar = \\ ( $charesc # \& | @ascii | @decimal | o @octal | x @hexadecimal ) + at stringchar = ($graphic # [\\ \"]) | $space | @escape | @gap + at char = ($graphic # [\\ \']) | $space | @escapechar + -- normal signed numerical literals can only be explicitly negative, -- not explicitly positive (contrast @exponent) @negative = \- @@ -460,7 +476,7 @@ $unigraphic / { isSmartQuote } { smart_quote_error } <0> { "#" $idchar+ / { ifExtension OverloadedLabelsBit } { skip_one_varid_src ITlabelvarid } - "#" \" / { ifExtension OverloadedLabelsBit } { lex_quoted_label } + "#" \" @stringchar* \" / { ifExtension OverloadedLabelsBit } { tok_quoted_label } } <0> { @@ -587,14 +603,38 @@ $unigraphic / { isSmartQuote } { smart_quote_error } } --- Strings and chars are lexed by hand-written code. The reason is --- that even if we recognise the string or char here in the regex --- lexer, we would still have to parse the string afterward in order --- to convert it to a String. <0> { - \' { lex_char_tok } - \"\"\" / { ifExtension MultilineStringsBit} { lex_string_tok StringTypeMulti } - \" { lex_string_tok StringTypeSingle } + \"\"\" / { ifExtension MultilineStringsBit } { tok_string_multi } + \" @stringchar* \" { tok_string } + \" @stringchar* \" \# / { ifExtension MagicHashBit } { tok_string } + \' @char \' { tok_char } + \' @char \' \# / { ifExtension MagicHashBit } { tok_char } + + -- Check for smart quotes and throw better errors than a plain lexical error (#21843) + \' \\ $unigraphic / { isSmartQuote } { smart_quote_error } + \" @stringchar* \\ $unigraphic / { isSmartQuote } { smart_quote_error } + -- See Note [Bare smart quote error] + -- The valid string rule will take precedence because it'll match more + -- characters than this rule, so this rule will only fire if the string + -- could not be lexed correctly + \" @stringchar* $unigraphic / { isSmartQuote } { smart_quote_error } +} + + { + -- Parse as much of the multiline string as possible, except for quotes + @stringchar* ($nl ([\ $tab] | @gap)* @stringchar*)* { tok_string_multi_content } + -- Allow bare quotes if it's not a triple quote + (\" | \"\") / ([\n .] # \") { tok_string_multi_content } +} + +<0> { + \'\' { token ITtyQuote } + + -- The normal character match takes precedence over this because it matches + -- more characters. However, if that pattern didn't match, then this quote + -- could be a quoted identifier, like 'x. Here, just return ITsimpleQuote, + -- as the parser will lex the varid separately. + \' / ($graphic # \\ | " ") { token ITsimpleQuote } } -- Note [Whitespace-sensitive operator parsing] @@ -880,7 +920,7 @@ data Token | ITchar SourceText Char -- Note [Literal source text] in "GHC.Types.SourceText" | ITstring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" - | ITmultilinestring SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" + | ITstringMulti SourceText FastString -- Note [Literal source text] in "GHC.Types.SourceText" | ITinteger IntegralLit -- Note [Literal source text] in "GHC.Types.SourceText" | ITrational FractionalLit @@ -2111,156 +2151,128 @@ lex_string_prag_comment mkTok span _buf _len _buf2 -- ----------------------------------------------------------------------------- -- Strings & Chars --- This stuff is horrible. I hates it. - -lex_string_tok :: LexStringType -> Action -lex_string_tok strType span buf _len _buf2 = do - s <- lex_string strType - - i <- getInput - case strType of - StringTypeSingle -> - lex_magic_hash i >>= \case - Just i' -> do - when (any (> '\xFF') s) $ do - pState <- getPState - let msg = PsErrPrimStringInvalidChar - let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg - addError err - - setInput i' - let (psSpan, src) = getStringLoc (buf, locStart) i' - pure $ L psSpan (ITprimstring src (unsafeMkByteString s)) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITstring src (mkFastString s)) - StringTypeMulti -> do - let (psSpan, src) = getStringLoc (buf, locStart) i - pure $ L psSpan (ITmultilinestring src (mkFastString s)) +tok_string :: Action +tok_string span buf len _buf2 = do + s <- lex_chars ("\"", "\"") span buf (if endsInHash then len - 1 else len) + + if endsInHash + then do + when (any (> '\xFF') s) $ do + pState <- getPState + let msg = PsErrPrimStringInvalidChar + let err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (last_loc pState)) msg + addError err + pure $ L span (ITprimstring src (unsafeMkByteString s)) + else + pure $ L span (ITstring src (mkFastString s)) where - locStart = psSpanStart span + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + +-- | Ideally, we would define this completely with Alex syntax, like normal strings. +-- Instead, this is defined as a hybrid solution by manually invoking lex states, which +-- we're doing for two reasons: +-- 1. The multiline string should all be one lexical token, not multiple +-- 2. We need to allow bare quotes, which can't be done with one regex +tok_string_multi :: Action +tok_string_multi startSpan startBuf _len _buf2 = do + -- advance to the end of the multiline string + let startLoc = psSpanStart startSpan + let i@(AI _ contentStartBuf) = + case lexDelim $ AI startLoc startBuf of + Just i -> i + Nothing -> panic "tok_string_multi did not start with a delimiter" + (AI _ contentEndBuf, i'@(AI endLoc endBuf)) <- goContent i + + -- build the values pertaining to the entire multiline string, including delimiters + let span = mkPsSpan startLoc endLoc + let len = byteDiff startBuf endBuf + let src = SourceText $ lexemeToFastString startBuf len + + -- load the content of the multiline string + let contentLen = byteDiff contentStartBuf contentEndBuf + s <- + either (throwStringLexError (AI startLoc startBuf)) pure $ + lexMultilineString contentLen contentStartBuf + + setInput i' + pure $ L span $ ITstringMulti src (mkFastString s) + where + goContent i0 = + case alexScan i0 string_multi_content of + AlexToken i1 len _ + | Just i2 <- lexDelim i1 -> pure (i1, i2) + | -- is the next token a tab character? + -- need this explicitly because there's a global rule matching $tab + Just ('\t', _) <- alexGetChar' i1 -> setInput i1 >> lexError LexError + | isEOF i1 -> checkSmartQuotes >> lexError LexError + | len == 0 -> panic $ "parsing multiline string got into infinite loop at: " ++ show i0 + | otherwise -> goContent i1 + AlexSkip i1 _ -> goContent i1 + _ -> lexError LexError + + lexDelim = + let go 0 i = Just i + go n i = + case alexGetChar' i of + Just ('"', i') -> go (n - 1) i' + _ -> Nothing + in go (3 :: Int) + + -- See Note [Bare smart quote error] + checkSmartQuotes = do + let findSmartQuote i0@(AI loc _) = + case alexGetChar' i0 of + Just ('\\', i1) | Just (_, i2) <- alexGetChar' i1 -> findSmartQuote i2 + Just (c, i1) + | isDoubleSmartQuote c -> Just (c, loc) + | otherwise -> findSmartQuote i1 + _ -> Nothing + case findSmartQuote (AI (psSpanStart startSpan) startBuf) of + Just (c, loc) -> throwSmartQuoteError c loc + Nothing -> pure () + +-- | Dummy action that should never be called. Should only be used in lex states +-- that are manually lexed in tok_string_multi. +tok_string_multi_content :: Action +tok_string_multi_content = panic "tok_string_multi_content unexpectedly invoked" + +lex_chars :: (String, String) -> PsSpan -> StringBuffer -> Int -> P String +lex_chars (startDelim, endDelim) span buf len = + either (throwStringLexError i0) pure $ + lexString contentLen contentBuf + where + i0@(AI _ contentBuf) = advanceInputBytes (length startDelim) $ AI (psSpanStart span) buf + -- assumes delimiters are ASCII, with 1 byte per Char + contentLen = len - length startDelim - length endDelim -lex_quoted_label :: Action -lex_quoted_label span buf _len _buf2 = do - s <- lex_string StringTypeSingle - (AI end bufEnd) <- getInput - let - token = ITlabelvarid (SourceText src) (mkFastString s) - src = lexemeToFastString (stepOn buf) (cur bufEnd - cur buf - 1) - start = psSpanStart span - - return $ L (mkPsSpan start end) token - - -lex_string :: LexStringType -> P String -lex_string strType = do - start <- getInput - (str, next) <- either fromStringLexError pure $ lexString strType alexGetChar' start - setInput next - pure str - - -lex_char_tok :: Action --- Here we are basically parsing character literals, such as 'x' or '\n' --- but we additionally spot 'x and ''T, returning ITsimpleQuote and --- ITtyQuote respectively, but WITHOUT CONSUMING the x or T part --- (the parser does that). --- So we have to do two characters of lookahead: when we see 'x we need to --- see if there's a trailing quote -lex_char_tok span buf _len _buf2 = do -- We've seen ' - i1 <- getInput -- Look ahead to first character - let loc = psSpanStart span - case alexGetChar' i1 of - Nothing -> lit_error i1 - - Just ('\'', i2@(AI end2 _)) -> do -- We've seen '' - setInput i2 - return (L (mkPsSpan loc end2) ITtyQuote) - - Just ('\\', i2@(AI end2 _)) -> do -- We've seen 'backslash - (lit_ch, i3) <- - either fromStringLexError pure $ - resolveEscapeCharacter alexGetChar' i2 - case alexGetChar' i3 of - Just ('\'', i4) -> do - setInput i4 - finish_char_tok buf loc lit_ch - Just (mc, _) | isSingleSmartQuote mc -> add_smart_quote_error mc end2 - _ -> lit_error i3 - - Just (c, i2@(AI end2 _)) - | not (isAnyChar c) -> lit_error i1 - | otherwise -> - - -- We've seen 'x, where x is a valid character - -- (i.e. not newline etc) but not a quote or backslash - case alexGetChar' i2 of -- Look ahead one more character - Just ('\'', i3) -> do -- We've seen 'x' - setInput i3 - finish_char_tok buf loc c - Just (c, _) | isSingleSmartQuote c -> add_smart_quote_error c end2 - _other -> do -- We've seen 'x not followed by quote - -- (including the possibility of EOF) - -- Just parse the quote only - let (AI end _) = i1 - return (L (mkPsSpan loc end) ITsimpleQuote) - --- We've already seen the closing quote --- Just need to check for trailing # -finish_char_tok :: StringBuffer -> PsLoc -> Char -> P (PsLocated Token) -finish_char_tok buf loc ch = do - i <- getInput - lex_magic_hash i >>= \case - Just i' -> do - setInput i' - -- Include the trailing # in SourceText - let (psSpan, src) = getStringLoc (buf, loc) i' - pure $ L psSpan (ITprimchar src ch) - Nothing -> do - let (psSpan, src) = getStringLoc (buf, loc) i - pure $ L psSpan (ITchar src ch) +throwStringLexError :: AlexInput -> StringLexError -> P a +throwStringLexError i (StringLexError e pos) = setInput (advanceInputTo pos i) >> lexError e --- | Get the span and source text for a string from the given start to the given end. -getStringLoc :: (StringBuffer, PsLoc) -> AlexInput -> (PsSpan, SourceText) -getStringLoc (bufStart, locStart) (AI locEnd bufEnd) = (psSpan, SourceText src) +tok_quoted_label :: Action +tok_quoted_label span buf len _buf2 = do + s <- lex_chars ("#\"", "\"") span buf len + pure $ L span (ITlabelvarid src (mkFastString s)) where - psSpan = mkPsSpan locStart locEnd - src = lexemeToFastString bufStart (cur bufEnd - cur bufStart) - - --- Return Just if we found the magic hash, with the next input. -lex_magic_hash :: AlexInput -> P (Maybe AlexInput) -lex_magic_hash i = do - magicHash <- getBit MagicHashBit - if magicHash - then - case alexGetChar' i of - Just ('#', i') -> pure (Just i') - _other -> pure Nothing - else pure Nothing - -fromStringLexError :: StringLexError AlexInput -> P a -fromStringLexError = \case - UnexpectedEOF i squote -> checkSQuote squote >> throw i LexStringCharLitEOF - BadCharInitialLex i squote -> checkSQuote squote >> throw i LexStringCharLit - EscapeBadChar i -> throw i LexStringCharLit - EscapeUnexpectedEOF i -> throw i LexStringCharLitEOF - EscapeNumRangeError i -> throw i LexNumEscapeRange - EscapeSmartQuoteError c (AI loc _) -> add_smart_quote_error c loc + -- skip leading '#' + src = SourceText . mkFastString . drop 1 $ lexemeToString buf len + + +tok_char :: Action +tok_char span buf len _buf2 = do + c <- lex_chars ("'", "'") span buf (if endsInHash then len - 1 else len) >>= \case + [c] -> pure c + s -> panic $ "tok_char expected exactly one character, got: " ++ show s + pure . L span $ + if endsInHash + then ITprimchar src c + else ITchar src c where - throw i e = setInput i >> lexError e - checkSQuote = \case - NoSmartQuote -> pure () - SmartQuote c (AI loc _) -> add_nonfatal_smart_quote_error c loc - --- before calling lit_error, ensure that the current input is pointing to --- the position of the error in the buffer. This is so that we can report --- a correct location to the user, but also so we can detect UTF-8 decoding --- errors if they occur. -lit_error :: AlexInput -> P a -lit_error i = do setInput i; lexError LexStringCharLit + src = SourceText $ lexemeToFastString buf len + endsInHash = currentChar (offsetBytes (len - 1) buf) == '#' + -- ----------------------------------------------------------------------------- -- QuasiQuote @@ -2319,32 +2331,28 @@ quasiquote_error start = do isSmartQuote :: AlexAccPred ExtsBitmap isSmartQuote _ _ _ (AI _ buf) = let c = prevChar buf ' ' in isSingleSmartQuote c || isDoubleSmartQuote c -smart_quote_error_message :: Char -> PsLoc -> MsgEnvelope PsMessage -smart_quote_error_message c loc = - let (correct_char, correct_char_name) = - if isSingleSmartQuote c then ('\'', "Single Quote") else ('"', "Quotation Mark") - err = mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ - PsErrUnicodeCharLooksLike c correct_char correct_char_name in - err - +throwSmartQuoteError :: Char -> PsLoc -> P a +throwSmartQuoteError c loc = addFatalError err + where + err = + mkPlainErrorMsgEnvelope (mkSrcSpanPs (mkPsSpan loc loc)) $ + PsErrUnicodeCharLooksLike c correct_char correct_char_name + (correct_char, correct_char_name) = + if isSingleSmartQuote c + then ('\'', "Single Quote") + else ('"', "Quotation Mark") + +-- | Throw a smart quote error, where the smart quote was the last character lexed smart_quote_error :: Action -smart_quote_error span buf _len _buf2 = do - let c = currentChar buf - addFatalError (smart_quote_error_message c (psSpanStart span)) - -add_smart_quote_error :: Char -> PsLoc -> P a -add_smart_quote_error c loc = addFatalError (smart_quote_error_message c loc) - -add_nonfatal_smart_quote_error :: Char -> PsLoc -> P () -add_nonfatal_smart_quote_error c loc = addError (smart_quote_error_message c loc) +smart_quote_error span _ _ buf2 = do + let c = prevChar buf2 (panic "smart_quote_error unexpectedly called on beginning of input") + throwSmartQuoteError c (psSpanStart span) -advance_to_smart_quote_character :: P () -advance_to_smart_quote_character = do - i <- getInput - case alexGetChar' i of - Just (c, _) | isDoubleSmartQuote c -> return () - Just (_, i2) -> do setInput i2; advance_to_smart_quote_character - Nothing -> return () -- should never get here +-- Note [Bare smart quote error] +-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +-- A smart quote inside of a string is allowed, but if a complete valid string +-- couldn't be lexed, we want to see if there's a smart quote that the user +-- thought ended the string, but in fact didn't. -- ----------------------------------------------------------------------------- -- Warnings @@ -2582,7 +2590,7 @@ getLastLocIncludingComments = P $ \s@(PState { prev_loc = prev_loc }) -> POk s p getLastLoc :: P PsSpan getLastLoc = P $ \s@(PState { last_loc = last_loc }) -> POk s last_loc -data AlexInput = AI !PsLoc !StringBuffer +data AlexInput = AI !PsLoc !StringBuffer deriving (Show) {- Note [Unicode in Alex] @@ -2693,6 +2701,19 @@ alexGetChar' (AI loc s) where (c,s') = nextChar s loc' = advancePsLoc loc c +-- | Advance the given input N bytes. +advanceInputBytes :: Int -> AlexInput -> AlexInput +advanceInputBytes n i0@(AI _ buf0) = advanceInputTo (cur buf0 + n) i0 + +-- | Advance the given input to the given position. +advanceInputTo :: Int -> AlexInput -> AlexInput +advanceInputTo pos = go + where + go i@(AI _ buf) + | cur buf >= pos = i + | Just (_, i') <- alexGetChar' i = go i' + | otherwise = i -- reached the end, just return the last input + getInput :: P AlexInput getInput = P $ \s at PState{ loc=l, buffer=b } -> POk s (AI l b) @@ -2700,9 +2721,10 @@ setInput :: AlexInput -> P () setInput (AI l b) = P $ \s -> POk s{ loc=l, buffer=b } () nextIsEOF :: P Bool -nextIsEOF = do - AI _ s <- getInput - return $ atEnd s +nextIsEOF = isEOF <$> getInput + +isEOF :: AlexInput -> Bool +isEOF (AI _ buf) = atEnd buf pushLexState :: Int -> P () pushLexState ls = P $ \s at PState{ lex_state=l } -> POk s{lex_state=ls:l} () @@ -3446,6 +3468,11 @@ topNoLayoutContainsCommas [] = False topNoLayoutContainsCommas (ALRLayout _ _ : ls) = topNoLayoutContainsCommas ls topNoLayoutContainsCommas (ALRNoLayout b _ : _) = b +-- If the generated alexScan/alexScanUser functions are called multiple times +-- in this file, alexScanUser gets broken out into a separate function and +-- increases memory usage. Make sure GHC inlines this function and optimizes it. +{-# INLINE alexScanUser #-} + lexToken :: P (PsLocated Token) lexToken = do inp@(AI loc1 buf) <- getInput ===================================== compiler/GHC/Parser/String.hs ===================================== @@ -1,284 +1,199 @@ {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE PatternSynonyms #-} +{-# LANGUAGE TupleSections #-} +{-# LANGUAGE ViewPatterns #-} module GHC.Parser.String ( StringLexError (..), - ContainsSmartQuote (..), - LexStringType (..), lexString, + lexMultilineString, -- * Unicode smart quote helpers isDoubleSmartQuote, isSingleSmartQuote, - - -- * Other helpers - isAnyChar, - resolveEscapeCharacter, ) where -import GHC.Prelude +import GHC.Prelude hiding (getChar) import Control.Arrow ((>>>)) -import Control.Monad (guard, unless, when) -import Data.Char (chr, isPrint, ord) -import Data.List (unfoldr) +import Control.Monad (when) +import Data.Char (chr, ord) +import qualified Data.Foldable1 as Foldable1 +import qualified Data.List.NonEmpty as NonEmpty import Data.Maybe (listToMaybe, mapMaybe) +import GHC.Data.StringBuffer (StringBuffer) +import qualified GHC.Data.StringBuffer as StringBuffer import GHC.Parser.CharClass ( hexDigit, - is_any, is_decdigit, is_hexdigit, is_octdigit, is_space, octDecDigit, ) +import GHC.Parser.Errors.Types (LexErr (..)) import GHC.Utils.Panic (panic) -data LexStringType = StringTypeSingle | StringTypeMulti +type BufPos = Int +data StringLexError = StringLexError LexErr BufPos --- | State to accumulate while iterating through string literal. --- --- Fields are strict here to avoid space leak when iterating through large string literal --- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 -data LexStringState loc = LexStringState - { stringAcc :: !String - -- ^ The string seen so far, reversed - , multilineCommonWsPrefix :: !Int - -- ^ The common prefix for multiline strings. See Note [Multiline string literals] - , initialLoc :: !loc - -- ^ The location of the beginning of the string literal - } - --- | Get the character at the given location, with the location --- of the next character. Returns Nothing if at the end of the --- input. -type GetChar loc = loc -> Maybe (Char, loc) - -lexString :: LexStringType -> GetChar loc -> loc -> Either (StringLexError loc) (String, loc) -lexString strType getChar initialLoc = go initialState initialLoc +lexString :: Int -> StringBuffer -> Either StringLexError String +lexString = lexStringWith processChars processChars where - initialState = - LexStringState - { stringAcc = "" - , multilineCommonWsPrefix = - case strType of - StringTypeMulti -> maxBound - _ -> 0 - , initialLoc = initialLoc - } - - -- 's' is strict here to avoid space leak when iterating through large string literal - -- https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12089#note_576175 - go !s loc0 = - case getChar loc0 of - -- found closing delimiter - Just ('"', _) | Just loc1 <- checkDelimiter strType getChar loc0 -> do - let postprocess = - case strType of - StringTypeSingle -> id - StringTypeMulti -> postprocessMultiline (multilineCommonWsPrefix s) - Right (postprocess . reverse $ stringAcc s, loc1) - - -- found backslash - Just (c0@'\\', loc1) -> do - case getChar loc1 of - -- found '\&' character, which should be elided - Just ('&', loc2) -> go s loc2 - -- found start of a string gap - Just (c1, loc2) | is_space c1 -> collapseStringGap getChar s loc2 >>= go s - -- some other escape character - Just (c1, loc2) -> - case strType of - StringTypeSingle -> do - (c', loc') <- resolveEscapeCharacter getChar loc1 - go (addChar c' s) loc' - StringTypeMulti -> do - -- keep escape characters unresolved until after post-processing, - -- to distinguish between a user-newline and the user writing "\n". - -- but still process the characters here, to find any errors - _ <- resolveEscapeCharacter getChar loc1 - go (addChar c1 . addChar c0 $ s) loc2 - -- backslash at end of input - Nothing -> Left $ BadCharInitialLex loc1 (hasSQuote getChar s) - - -- found newline character in multiline string - Just (c0@'\n', loc1) | StringTypeMulti <- strType -> - uncurry go $ parseLeadingWS getChar (addChar c0 s) loc1 - - -- found some other character - Just (c0, loc1) | isAnyChar c0 -> go (addChar c0 s) loc1 - - -- found some unknown character - Just (_, _) -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - - -- reached EOF before finding end of string - Nothing -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) -{-# INLINE lexString #-} - -checkDelimiter :: LexStringType -> GetChar loc -> loc -> Maybe loc -checkDelimiter strType getChar loc0 = - case strType of - StringTypeSingle -> do - ('"', loc1) <- getChar loc0 - Just loc1 - StringTypeMulti -> do - ('"', loc1) <- getChar loc0 - ('"', loc2) <- getChar loc1 - ('"', loc3) <- getChar loc2 - Just loc3 -{-# INLINE checkDelimiter #-} - --- | A helper for adding the given character to the lexed string. -addChar :: Char -> LexStringState loc -> LexStringState loc -addChar c s = s{stringAcc = c : stringAcc s} -{-# INLINE addChar #-} - --- | Return whether the string we've parsed so far contains any smart quotes. -hasSQuote :: GetChar loc -> LexStringState loc -> ContainsSmartQuote loc -hasSQuote getChar s - | any isDoubleSmartQuote (stringAcc s) - , (c, loc) : _ <- filter (isDoubleSmartQuote . fst) allChars = - SmartQuote c loc - | otherwise = - NoSmartQuote + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps + >>> resolveEscapes + +-- ----------------------------------------------------------------------------- +-- Lexing interface + +{- +Note [Lexing strings] +~~~~~~~~~~~~~~~~~~~~~ + +After verifying if a string is lexically valid with Alex, we still need to do +some post processing of the string, namely: +1. Collapse string gaps +2. Resolve escape characters + +The problem: 'lexemeToString' is more performant than manually reading +characters from the StringBuffer. However, that completely erases the position +of each character, which we need in order to report the correct position for +error messages (e.g. when resolving escape characters). + +So what we'll do is do two passes. The first pass is optimistic; just convert +to a plain String and process it. If this results in an error, we do a second +pass, this time where each character is annotated with its position. Now, the +error has all the information it needs. + +Ideally, lexStringWith would take a single (forall c. HasChar c => ...) function, +but to help the specializer, we pass it in twice to concretize it for the two +types we actually use. +-} + +-- | See Note [Lexing strings] +lexStringWith :: + ([Char] -> Either (Char, LexErr) [Char]) + -> ([CharPos] -> Either (CharPos, LexErr) [CharPos]) + -> Int + -> StringBuffer + -> Either StringLexError String +lexStringWith processChars processCharsPos len buf = + case processChars $ bufferChars buf len of + Right s -> Right s + Left _ -> + case processCharsPos $ bufferLocatedChars buf len of + Right _ -> panic "expected lex error on second pass" + Left ((_, pos), e) -> Left $ StringLexError e pos + +class HasChar c where + getChar :: c -> Char + setChar :: Char -> c -> c + +instance HasChar Char where + getChar = id + setChar = const + +instance HasChar (Char, x) where + getChar = fst + setChar c (_, x) = (c, x) + +pattern Char :: HasChar c => Char -> c +pattern Char c <- (getChar -> c) +{-# COMPLETE Char #-} + +bufferChars :: StringBuffer -> Int -> [Char] +bufferChars = StringBuffer.lexemeToString + +type CharPos = (Char, BufPos) + +bufferLocatedChars :: StringBuffer -> Int -> [CharPos] +bufferLocatedChars initialBuf len = go initialBuf where - allChars = unfoldr getCharWithLoc (initialLoc s) - getCharWithLoc loc = - case getChar loc of - Just (c, loc') -> Just ((c, loc), loc') - Nothing -> Nothing -{-# INLINE hasSQuote #-} - --- | After parsing a backslash and a space character, consume the rest of --- the string gap and return the next location. -collapseStringGap :: GetChar loc -> LexStringState loc -> loc -> Either (StringLexError loc) loc -collapseStringGap getChar s = go + go buf + | atEnd buf = [] + | otherwise = + let (c, buf') = StringBuffer.nextChar buf + in (c, StringBuffer.cur buf) : go buf' + + atEnd buf = StringBuffer.byteDiff initialBuf buf >= len + +-- ----------------------------------------------------------------------------- +-- Lexing phases + +collapseGaps :: HasChar c => [c] -> [c] +collapseGaps = go where - go loc0 = - case getChar loc0 of - Just ('\\', loc1) -> pure loc1 - Just (c0, loc1) | is_space c0 -> go loc1 - Just _ -> Left $ BadCharInitialLex loc0 (hasSQuote getChar s) - Nothing -> Left $ UnexpectedEOF loc0 (hasSQuote getChar s) -{-# INLINE collapseStringGap #-} + go = \case + c1@(Char '\\') : c2@(Char c) : cs + | is_space c -> go $ dropGap cs + | otherwise -> c1 : c2 : go cs + c : cs -> c : go cs + [] -> [] --- | See Note [Multiline string literals] -parseLeadingWS :: GetChar loc -> LexStringState loc -> loc -> (LexStringState loc, loc) -parseLeadingWS getChar = go 0 + dropGap = \case + Char '\\' : cs -> cs + _ : cs -> dropGap cs + [] -> panic "gap unexpectedly ended" + +resolveEscapes :: HasChar c => [c] -> Either (c, LexErr) [c] +resolveEscapes = go dlistEmpty where - go !col s loc = - case getChar loc of - Just (c@' ', loc') -> go (col + 1) (addChar c s) loc' - -- expand tabs - Just ('\t', loc') -> - let fill = 8 - (col `mod` 8) - s' = applyN fill (addChar ' ') s - in go (col + fill) s' loc' - -- if we see a newline or string delimiter, then this line only contained whitespace, so - -- don't include it in the common whitespace prefix - Just ('\n', _) -> (s, loc) - Just ('"', _) | Just _ <- checkDelimiter StringTypeMulti getChar loc -> (s, loc) - -- found some other character, so we're done parsing leading whitespace - _ -> - let s' = s{multilineCommonWsPrefix = min col (multilineCommonWsPrefix s)} - in (s', loc) - - applyN :: Int -> (a -> a) -> a -> a - applyN n f x0 = iterate f x0 !! n -{-# INLINE parseLeadingWS #-} - -data StringLexError loc - = UnexpectedEOF !loc !(ContainsSmartQuote loc) - -- ^ Unexpectedly hit EOF when lexing string - | BadCharInitialLex !loc !(ContainsSmartQuote loc) - -- ^ Found invalid character when initially lexing string - | EscapeBadChar !loc - -- ^ Found invalid character when parsing an escaped character - | EscapeUnexpectedEOF !loc - -- ^ Unexpectedly hit EOF when parsing an escaped character - | EscapeNumRangeError !loc - -- ^ Escaped number exceeds range - | EscapeSmartQuoteError !Char !loc - -- ^ Found escaped smart unicode chars as `\’` or `\”` - deriving (Show) - --- | When initially lexing the string, we want to track if we've --- seen a smart quote, to show a helpful "you might be accidentally --- using a smart quote" error. -data ContainsSmartQuote loc - = NoSmartQuote - | SmartQuote !Char !loc - deriving (Show) + go !acc = \case + [] -> pure $ dlistToList acc + Char '\\' : Char '&' : cs -> go acc cs + backslash@(Char '\\') : cs -> + case resolveEscapeChar cs of + Right (esc, cs') -> go (acc `dlistSnoc` setChar esc backslash) cs' + Left (c, e) -> Left (c, e) + c : cs -> go (acc `dlistSnoc` c) cs -- ----------------------------------------------------------------------------- -- Escape characters --- | After finding a backslash, parse the rest of the escape character, starting --- at the given location. -resolveEscapeCharacter :: GetChar loc -> loc -> Either (StringLexError loc) (Char, loc) -resolveEscapeCharacter getChar loc0 = do - (c0, loc1) <- expectChar loc0 - case c0 of - 'a' -> pure ('\a', loc1) - 'b' -> pure ('\b', loc1) - 'f' -> pure ('\f', loc1) - 'n' -> pure ('\n', loc1) - 'r' -> pure ('\r', loc1) - 't' -> pure ('\t', loc1) - 'v' -> pure ('\v', loc1) - '\\' -> pure ('\\', loc1) - '"' -> pure ('\"', loc1) - '\'' -> pure ('\'', loc1) - -- escape codes - 'x' -> expectNum is_hexdigit 16 hexDigit loc1 - 'o' -> expectNum is_octdigit 8 octDecDigit loc1 - _ | is_decdigit c0 -> expectNum is_decdigit 10 octDecDigit loc0 - -- control characters (e.g. '\^M') - '^' -> do - (c1, loc2) <- expectChar loc1 - unless ('@' <= c1 && c1 <= '_') $ Left $ EscapeBadChar loc1 - pure (chr $ ord c1 - ord '@', loc2) - -- long form escapes (e.g. '\NUL') - _ | Just (c1, loc2) <- parseLongEscape getChar c0 loc1 -> pure (c1, loc2) - -- check unicode smart quotes (#21843) - _ | isDoubleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - _ | isSingleSmartQuote c0 -> Left $ EscapeSmartQuoteError c0 loc0 - -- unknown escape - _ -> Left $ EscapeBadChar loc0 +-- | Resolve a escape character, after having just lexed a backslash. +-- Assumes escape character is valid. +resolveEscapeChar :: HasChar c => [c] -> Either (c, LexErr) (Char, [c]) +resolveEscapeChar = \case + Char 'a' : cs -> pure ('\a', cs) + Char 'b' : cs -> pure ('\b', cs) + Char 'f' : cs -> pure ('\f', cs) + Char 'n' : cs -> pure ('\n', cs) + Char 'r' : cs -> pure ('\r', cs) + Char 't' : cs -> pure ('\t', cs) + Char 'v' : cs -> pure ('\v', cs) + Char '\\' : cs -> pure ('\\', cs) + Char '"' : cs -> pure ('\"', cs) + Char '\'' : cs -> pure ('\'', cs) + -- escape codes + Char 'x' : cs -> parseNum is_hexdigit 16 hexDigit cs + Char 'o' : cs -> parseNum is_octdigit 8 octDecDigit cs + cs@(Char c : _) | is_decdigit c -> parseNum is_decdigit 10 octDecDigit cs + -- control characters (e.g. '\^M') + Char '^' : Char c : cs -> pure (chr $ ord c - ord '@', cs) + -- long form escapes (e.g. '\NUL') + cs | Just (esc, cs') <- parseLongEscape cs -> pure (esc, cs') + -- shouldn't happen + Char c : _ -> panic $ "found unexpected escape character: " ++ show c + [] -> panic "escape character unexpectedly ended" where - expectChar loc = - case getChar loc of - Just x -> pure x - Nothing -> Left $ EscapeUnexpectedEOF loc - - expectNum isDigit base toDigit loc1 = do - (c1, loc2) <- expectChar loc1 - unless (isDigit c1) $ Left $ EscapeBadChar loc1 - let parseNum x loc = - case getChar loc of - Just (c, loc') | isDigit c -> do - let x' = x * base + toDigit c - when (x' > 0x10ffff) $ Left $ EscapeNumRangeError loc - parseNum x' loc' - _ -> - pure (chr x, loc) - parseNum (toDigit c1) loc2 -{-# INLINE resolveEscapeCharacter #-} - -parseLongEscape :: GetChar loc -> Char -> loc -> Maybe (Char, loc) -parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCodes + parseNum isDigit base toDigit = + let go x = \case + ch@(Char c) : cs | isDigit c -> do + let x' = x * base + toDigit c + when (x' > 0x10ffff) $ Left (ch, LexNumEscapeRange) + go x' cs + cs -> pure (chr x, cs) + in go 0 + +parseLongEscape :: HasChar c => [c] -> Maybe (Char, [c]) +parseLongEscape cs = listToMaybe (mapMaybe tryParse longEscapeCodes) where - tryParse (prefix, c) = do - p0 : p <- pure prefix - guard (p0 == c0) -- see if the first character matches - loc <- parsePrefix loc1 p -- see if the rest of the prefix matches - pure (c, loc) - - parsePrefix loc = \case - [] -> pure loc - p : ps -> do - (c, loc') <- getChar loc - guard (p == c) - parsePrefix loc' ps + tryParse (code, esc) = + case splitAt (length code) cs of + (pre, cs') | map getChar pre == code -> Just (esc, cs') + _ -> Nothing longEscapeCodes = [ ("NUL", '\NUL') @@ -289,14 +204,14 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("ENQ", '\ENQ') , ("ACK", '\ACK') , ("BEL", '\BEL') - , ("BS", '\BS') - , ("HT", '\HT') - , ("LF", '\LF') - , ("VT", '\VT') - , ("FF", '\FF') - , ("CR", '\CR') - , ("SO", '\SO') - , ("SI", '\SI') + , ("BS" , '\BS' ) + , ("HT" , '\HT' ) + , ("LF" , '\LF' ) + , ("VT" , '\VT' ) + , ("FF" , '\FF' ) + , ("CR" , '\CR' ) + , ("SO" , '\SO' ) + , ("SI" , '\SI' ) , ("DLE", '\DLE') , ("DC1", '\DC1') , ("DC2", '\DC2') @@ -306,17 +221,16 @@ parseLongEscape getChar c0 loc1 = listToMaybe $ mapMaybe tryParse longEscapeCode , ("SYN", '\SYN') , ("ETB", '\ETB') , ("CAN", '\CAN') - , ("EM", '\EM') + , ("EM" , '\EM' ) , ("SUB", '\SUB') , ("ESC", '\ESC') - , ("FS", '\FS') - , ("GS", '\GS') - , ("RS", '\RS') - , ("US", '\US') - , ("SP", '\SP') + , ("FS" , '\FS' ) + , ("GS" , '\GS' ) + , ("RS" , '\RS' ) + , ("US" , '\US' ) + , ("SP" , '\SP' ) , ("DEL", '\DEL') ] -{-# INLINE parseLongEscape #-} -- ----------------------------------------------------------------------------- -- Unicode Smart Quote detection (#21843) @@ -333,6 +247,98 @@ isSingleSmartQuote = \case '’' -> True _ -> False +-- ----------------------------------------------------------------------------- +-- Multiline strings + +-- | See Note [Multiline string literals] +-- +-- Assumes string is lexically valid. Skips the steps about splitting +-- and rejoining lines, and instead manually find newline characters, +-- for performance. +lexMultilineString :: Int -> StringBuffer -> Either StringLexError String +lexMultilineString = lexStringWith processChars processChars + where + processChars :: HasChar c => [c] -> Either (c, LexErr) [c] + processChars = + collapseGaps -- Step 1 + >>> expandLeadingTabs -- Step 3 + >>> rmCommonWhitespacePrefix -- Step 4 + >>> collapseOnlyWsLines -- Step 5 + >>> rmFirstNewline -- Step 7a + >>> rmLastNewline -- Step 7b + >>> resolveEscapes -- Step 8 + + -- expands all tabs, since the lexer will verify that tabs can only appear + -- as leading indentation + expandLeadingTabs :: HasChar c => [c] -> [c] + expandLeadingTabs = + let go !col = \case + c@(Char '\t') : cs -> + let fill = 8 - (col `mod` 8) + in replicate fill (setChar ' ' c) ++ go (col + fill) cs + c : cs -> c : go (if getChar c == '\n' then 0 else col + 1) cs + [] -> [] + in go 0 + + rmCommonWhitespacePrefix :: HasChar c => [c] -> [c] + rmCommonWhitespacePrefix cs0 = + let commonWSPrefix = getCommonWsPrefix (map getChar cs0) + go = \case + c@(Char '\n') : cs -> c : go (dropLine commonWSPrefix cs) + c : cs -> c : go cs + [] -> [] + -- drop x characters from the string, or up to a newline, whichever + -- comes first + dropLine !x = \case + cs | x <= 0 -> cs + cs@(Char '\n' : _) -> cs + _ : cs -> dropLine (x - 1) cs + [] -> [] + in go cs0 + + collapseOnlyWsLines :: HasChar c => [c] -> [c] + collapseOnlyWsLines = + let go = \case + c@(Char '\n') : cs | Just cs' <- checkAllWs cs -> c : go cs' + c : cs -> c : go cs + [] -> [] + checkAllWs = \case + -- got all the way to a newline or the end of the string, return + cs@(Char '\n' : _) -> Just cs + cs@[] -> Just cs + -- found whitespace, continue + Char c : cs | is_space c -> checkAllWs cs + -- anything else, stop + _ -> Nothing + in go + + rmFirstNewline :: HasChar c => [c] -> [c] + rmFirstNewline = \case + Char '\n' : cs -> cs + cs -> cs + + rmLastNewline :: HasChar c => [c] -> [c] + rmLastNewline = + let go = \case + [] -> [] + [Char '\n'] -> [] + c : cs -> c : go cs + in go + +-- | See step 4 in Note [Multiline string literals] +-- +-- Assumes tabs have already been expanded. +getCommonWsPrefix :: String -> Int +getCommonWsPrefix s = + case NonEmpty.nonEmpty includedLines of + Nothing -> 0 + Just ls -> Foldable1.minimum $ NonEmpty.map (length . takeWhile is_space) ls + where + includedLines = + filter (not . all is_space) -- ignore whitespace-only lines + . drop 1 -- ignore first line in calculation + $ lines s + {- Note [Multiline string literals] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -351,23 +357,13 @@ The canonical steps for post processing a multiline string are: 2. Split the string by newlines 3. Convert leading tabs into spaces * In each line, any tabs preceding non-whitespace characters are replaced with spaces up to the next tab stop -4. Remove common whitespace prefix in every line (see below) +4. Remove common whitespace prefix in every line except the first (see below) 5. If a line contains only whitespace, remove all of the whitespace 6. Join the string back with `\n` delimiters -7. If the first character of the string is a newline, remove it +7a. If the first character of the string is a newline, remove it +7b. If the last character of the string is a newline, remove it 8. Interpret escaped characters -However, for performance reasons, we do as much of this in one pass as possible: -1. As we lex the string, do the following steps as they appear: - a. Collapse string gaps - b. Keep track of the common whitespace prefix so far - c. Validate escaped characters -2. At the very end, post process the lexed string: - a. Remove the common whitespace prefix from every line - b. Remove all whitespace from all-whitespace lines - c. Remove initial newline character - d. Resolve escaped characters - The common whitespace prefix can be informally defined as "The longest prefix of whitespace shared by all lines in the string, excluding the first line and any whitespace-only lines". @@ -382,75 +378,16 @@ It's more precisely defined with the following algorithm: 3. Calculate the longest prefix of whitespace shared by all lines in the remaining list -} --- | See Note [Multiline string literals] -postprocessMultiline :: Int -> String -> String -postprocessMultiline commonWSPrefix = - rmCommonWhitespacePrefix - >>> collapseOnlyWsLines - >>> rmFirstNewline - >>> rmLastNewline - >>> resolveEscapeChars - where - rmCommonWhitespacePrefix = - let go = \case - '\n' : s -> '\n' : go (dropLine commonWSPrefix s) - c : s -> c : go s - [] -> [] - -- drop x characters from the string, or up to a newline, whichever - -- comes first - dropLine !x = \case - s | x <= 0 -> s - s@('\n' : _) -> s - _ : s -> dropLine (x - 1) s - [] -> [] - in go - - collapseOnlyWsLines = - let go = \case - '\n' : s | Just s' <- checkAllWs s -> '\n' : go s' - c : s -> c : go s - [] -> [] - checkAllWs = \case - -- got all the way to a newline or the end of the string, return - s@('\n' : _) -> Just s - s@[] -> Just s - -- found whitespace, continue - c : s | is_space c -> checkAllWs s - -- anything else, stop - _ -> Nothing - in go - - rmFirstNewline = \case - '\n' : s -> s - s -> s +-- ----------------------------------------------------------------------------- +-- DList - rmLastNewline = - let go = \case - [] -> [] - ['\n'] -> [] - c : cs -> c : go cs - in go +newtype DList a = DList ([a] -> [a]) - -- resolve escape characters, deferred from lexString. guaranteed - -- to not throw any errors, since we already checked them in lexString - resolveEscapeChars = \case - [] -> [] - '\\' : s -> - -- concretizing 'loc' to String: - -- resolveEscapeCharacter :: (String -> Maybe (Char, String)) -> String -> Either _ (Char, String) - case resolveEscapeCharacter uncons s of - Left e -> panic $ "resolving escape characters in multiline string unexpectedly found errors: " ++ show e - Right (c, s') -> c : resolveEscapeChars s' - c : s -> c : resolveEscapeChars s - - uncons = \case - c : cs -> Just (c, cs) - [] -> Nothing +dlistEmpty :: DList a +dlistEmpty = DList id --- ----------------------------------------------------------------------------- --- Helpers +dlistToList :: DList a -> [a] +dlistToList (DList f) = f [] -isAnyChar :: Char -> Bool -isAnyChar c - | c > '\x7f' = isPrint c - | otherwise = is_any c +dlistSnoc :: DList a -> a -> DList a +dlistSnoc (DList f) x = DList (f . (x :)) ===================================== testsuite/driver/testlib.py ===================================== @@ -2882,7 +2882,7 @@ def normalise_errmsg(s: str) -> str: # normalise slashes to minimise Windows/Unix filename differences, # but don't normalize backslashes in chars - s = re.sub(r"(?!')\\", '/', s) + s = re.sub(r"(?:9:1: error: [GHC-58481] parse error on input ‘+’ -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' + +Bad.hs:3:7: error: [GHC-21231] lexical error at character '\n' -Bad.hs:3:8: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/ghci/scripts/ghci022.stderr ===================================== @@ -1,3 +1,2 @@ +ghci022.hs:2:5: error: [GHC-21231] lexical error at character '\n' -ghci022.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/MultilineStringsInnerTab.stderr ===================================== @@ -1,3 +1,3 @@ - MultilineStringsInnerTab.hs:8:5: error: [GHC-21231] - lexical error in string/character literal at character '\t' + lexical error at character '\t' + ===================================== testsuite/tests/parser/should_fail/MultilineStringsSmartQuotes.stderr ===================================== @@ -1,6 +1,2 @@ - MultilineStringsSmartQuotes.hs:9:3: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -MultilineStringsSmartQuotes.hs:10:1: [GHC-21231] - lexical error in string/character literal at end of input ===================================== testsuite/tests/parser/should_fail/T21843c.stderr ===================================== @@ -1,6 +1,2 @@ - -T21843c.hs:3:19: [GHC-31623] +T21843c.hs:3:13: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not - -T21843c.hs:3:20: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/T21843e.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843e.hs:3:15: [GHC-31623] +T21843e.hs:3:13: error: [GHC-31623] Unicode character '”' ('\8221') looks like '"' (Quotation Mark), but it is not + ===================================== testsuite/tests/parser/should_fail/T21843f.stderr ===================================== @@ -1,3 +1,3 @@ - -T21843f.hs:3:13: [GHC-31623] +T21843f.hs:3:11: error: [GHC-31623] Unicode character '‘' ('\8216') looks like ''' (Single Quote), but it is not + ===================================== testsuite/tests/parser/should_fail/T3751.stderr ===================================== @@ -1,3 +1,2 @@ +T3751.hs:3:5: error: [GHC-21231] lexical error at character '\167' -T3751.hs:3:7: error: [GHC-21231] - lexical error in string/character literal at character '\167' ===================================== testsuite/tests/parser/should_fail/T5425.stderr ===================================== @@ -1,3 +1,2 @@ +T5425.hs:4:1: error: [GHC-21231] lexical error at character '\955' -T5425.hs:5:2: error: [GHC-21231] - lexical error in string/character literal at character '\955' ===================================== testsuite/tests/parser/should_fail/readFail002.stderr ===================================== @@ -1,3 +1,3 @@ +readFail002.hs:5:5: error: [GHC-21231] + lexical error at character '\n' -readFail002.hs:5:6: error: [GHC-21231] - lexical error in string/character literal at character '\n' ===================================== testsuite/tests/parser/should_fail/readFail004.stderr ===================================== @@ -1,3 +1,3 @@ +readFail004.hs:17:16: error: [GHC-21231] + lexical error at character '.' -readFail004.hs:19:1: error: [GHC-21231] - lexical error in string/character literal at character '.' ===================================== testsuite/tests/parser/should_fail/readFail005.stderr ===================================== @@ -1,3 +1,2 @@ - -readFail005.hs:4:7: error: [GHC-21231] - lexical error in string/character literal at character '&' +readFail005.hs:4:5: [GHC-21231] + lexical error at character '&' ===================================== testsuite/tests/parser/should_fail/readFail033.stderr ===================================== @@ -1,3 +1,3 @@ +readFail033.hs:2:5: error: [GHC-21231] + lexical error at character '\t' -readFail033.hs:2:6: error: [GHC-21231] - lexical error in string/character literal at character '\t' ===================================== testsuite/tests/parser/unicode/all.T ===================================== @@ -33,3 +33,6 @@ test('T18225B', normal, compile_fail, ['']) test('T18158', normal, compile, ['']) test('T18158b', normal, compile_fail, ['']) + +test('lex_unispace', normal, compile, ['']) +test('lex_unicode_ids', normal, compile, ['']) ===================================== testsuite/tests/parser/unicode/lex_unicode_ids.hs ===================================== @@ -0,0 +1,11 @@ +-- Regression tests for unicode identifiers + +{-# LANGUAGE TemplateHaskellQuotes #-} + +module ShouldCompile where + +ƞ = 1 +eta = 'ƞ + +data Ʊ +upsilon = ''Ʊ ===================================== testsuite/tests/parser/unicode/lex_unispace.hs ===================================== @@ -0,0 +1,9 @@ +-- Regression tests for unicode whitespace + +module ShouldCompile where + +-- https://github.com/fosskers/aura/blob/cdebca1f48254ebb8286d8e38591bf644282866f/haskell/aura/lib/Aura/Languages.hs#L107 +x1 = '　' -- \12288 + +-- https://github.com/jgm/pandoc/blob/98e77e02f6436e4b74a164762d0f3149ae7ecefa/src/Text/Pandoc/Writers/FB2.hs#L295C3-L295C32 +x2 = " " -- \xa0 ===================================== testsuite/tests/parser/unicode/utf8_010.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_010.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_011.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_011.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_020.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_020.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_021.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_021.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== testsuite/tests/parser/unicode/utf8_022.stderr ===================================== @@ -1,3 +1,3 @@ - utf8_022.hs:2:8: error: [GHC-21231] - lexical error in string/character literal (UTF-8 decoding error) + lexical error (UTF-8 decoding error) + ===================================== utils/haddock/haddock-api/src/Haddock/Backends/Hyperlinker/Parser.hs ===================================== @@ -320,7 +320,7 @@ classify tok = ITlabelvarid{} -> TkUnknown ITchar{} -> TkChar ITstring{} -> TkString - ITmultilinestring{} -> TkString + ITstringMulti{} -> TkString ITinteger{} -> TkNumber ITrational{} -> TkNumber ITprimchar{} -> TkChar View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/bb030d0d4374f6a30432e821fda7d0ef699425f5 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/bb030d0d4374f6a30432e821fda7d0ef699425f5 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 16:09:54 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 12:09:54 -0400 Subject: [Git][ghc/ghc][master] 6 commits: Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import Message-ID: <66f58752d8fc8_50525b78e0115153@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 26 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Parser/Header.hs - compiler/GHC/Rename/Names.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Types/Error/Codes.hs - compiler/GHC/Unit/State.hs - compiler/GHC/Unit/Types.hs - docs/users_guide/exts/rebindable_syntax.rst - libraries/base/base.cabal.in - libraries/base/src/Control/Monad/Zip.hs - libraries/base/src/Data/List/NonEmpty.hs - libraries/ghc-internal/ghc-internal.cabal.in - libraries/ghc-internal/src/GHC/Internal/Base.hs - + libraries/ghc-internal/src/GHC/Internal/Control/Monad/Zip.hs - libraries/ghc-internal/src/GHC/Internal/Data/List/NonEmpty.hs - rts/include/RtsAPI.h - testsuite/tests/backpack/should_compile/bkp16.stderr - testsuite/tests/backpack/should_fail/bkpfail17.stderr - testsuite/tests/backpack/should_fail/bkpfail19.stderr - testsuite/tests/iface/IfaceSharingIfaceType.hs - testsuite/tests/iface/IfaceSharingName.hs - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/bb030d0d4374f6a30432e821fda7d0ef699425f5...d3dacdfb9f739541ca6ed2659d35723677557955 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/bb030d0d4374f6a30432e821fda7d0ef699425f5...d3dacdfb9f739541ca6ed2659d35723677557955 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 16:32:25 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Thu, 26 Sep 2024 12:32:25 -0400 Subject: [Git][ghc/ghc][wip/T25281] Wibbles Message-ID: <66f58c998c4c4_b6b73113fc8918c5@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: 740c3d27 by Simon Peyton Jones at 2024-09-26T17:32:11+01:00 Wibbles - - - - - 1 changed file: - compiler/GHC/Core/ConLike.hs Changes: ===================================== compiler/GHC/Core/ConLike.hs ===================================== @@ -47,7 +47,6 @@ import GHC.Utils.Outputable import Data.Maybe( isJust ) import qualified Data.Data as Data -import qualified Data.List as List {- ************************************************************************ View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/740c3d27332b9673175d4dfe4e1040d65ac5ba2a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/740c3d27332b9673175d4dfe4e1040d65ac5ba2a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 16:40:07 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Thu, 26 Sep 2024 12:40:07 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 29 commits: Don't compile `asBox` with -fprof-late Message-ID: <66f58e67bfbe8_b6b73124f4495460@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - c43e8c98 by Patrick at 2024-09-26T12:39:50-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 1e12471e by sheaf at 2024-09-26T12:39:51-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - c6025e58 by sheaf at 2024-09-26T12:39:51-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 1380f545 by sheaf at 2024-09-26T12:39:51-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 175e851e by sheaf at 2024-09-26T12:39:52-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - f7eee0db by sheaf at 2024-09-26T12:39:52-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - b029c577 by sheaf at 2024-09-26T12:39:52-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - 0bfb71ca by sheaf at 2024-09-26T12:39:52-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - fd77ced8 by sheaf at 2024-09-26T12:39:52-04:00 Add test for C calls & SIMD vectors - - - - - 5130c505 by sheaf at 2024-09-26T12:39:52-04:00 Add test for #25169 - - - - - e74e6a94 by sheaf at 2024-09-26T12:39:52-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - 3c4d31aa by sheaf at 2024-09-26T12:39:52-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - 66ee8976 by sheaf at 2024-09-26T12:39:52-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - 8105366b by sheaf at 2024-09-26T12:39:52-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 980a37aa by sheaf at 2024-09-26T12:39:52-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 51f67e06 by sheaf at 2024-09-26T12:39:52-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 9647f458 by Cheng Shao at 2024-09-26T12:39:53-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 97e0a693 by Matthew Pickering at 2024-09-26T12:39:59-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 30 changed files: - .gitlab-ci.yml - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8b6bc6c53e1858e59bd99f12eefc5ebfd3cfa330...97e0a69372e897cb8978982f5958d1a3e847f379 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8b6bc6c53e1858e59bd99f12eefc5ebfd3cfa330...97e0a69372e897cb8978982f5958d1a3e847f379 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 17:23:51 2024 From: gitlab at gitlab.haskell.org (Vladislav Zavialov (@int-index)) Date: Thu, 26 Sep 2024 13:23:51 -0400 Subject: [Git][ghc/ghc][wip/int-index/wildcard-binders] 17 commits: base: Propagate `error` CallStack to thrown exception Message-ID: <66f598a7c618c_f364a2f721851532@gitlab.mail> Vladislav Zavialov pushed to branch wip/int-index/wildcard-binders at Glasgow Haskell Compiler / GHC Commits: 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 0065d7ec by Vladislav Zavialov at 2024-09-26T19:14:45+03:00 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - 30 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Core/Opt/ConstantFold.hs - compiler/GHC/Driver/Pipeline.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/Hs/Decls.hs - compiler/GHC/Hs/Instances.hs - compiler/GHC/Hs/Type.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Header.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Parser/String.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Rename/Module.hs - compiler/GHC/Rename/Names.hs - compiler/GHC/Tc/Deriv/Generate.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/GHC/Tc/Gen/Sig.hs - compiler/GHC/Tc/TyCl.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Error/Codes.hs - compiler/GHC/Unit/State.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4dceb0ff71afe56484c46d6e52456650a6d643dc...0065d7ec7bda6a2820f49e063d28906858e66360 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/4dceb0ff71afe56484c46d6e52456650a6d643dc...0065d7ec7bda6a2820f49e063d28906858e66360 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 19:58:53 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 15:58:53 -0400 Subject: [Git][ghc/ghc][wip/wasm-perf-notes] ci: Push perf notes from wasm jobs Message-ID: <66f5bcfdd34d6_f364aa628a4727e@gitlab.mail> Matthew Pickering pushed to branch wip/wasm-perf-notes at Glasgow Haskell Compiler / GHC Commits: c63f910e by Matthew Pickering at 2024-09-26T20:58:45+01:00 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 1 changed file: - .gitlab/ci.sh Changes: ===================================== .gitlab/ci.sh ===================================== @@ -431,7 +431,9 @@ function push_perf_notes() { return fi - if [ -n "${CROSS_TARGET:-}" ] && [ "${CROSS_EMULATOR:-}" != "js-emulator" ]; then + # TODO: Remove this check, see #25299 + # It is easy to forget to update this when testing a new cross platform + if [[ -n "${CROSS_TARGET:-}" ]] && [[ "${CROSS_TARGET:-}" != *"javascript"* ]] && [[ "${CROSS_TARGET:-}" != *"wasm"* ]]; then info "Can't test cross-compiled build." return fi View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c63f910e966a4b5211bcb3a5f59b00379ba4c415 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c63f910e966a4b5211bcb3a5f59b00379ba4c415 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 20:33:59 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 16:33:59 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 2 commits: Fix hardcoded stage1 Message-ID: <66f5c537e9050_f364ac9e9b074719@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: 31aac127 by Matthew Pickering at 2024-09-26T21:32:17+01:00 Fix hardcoded stage1 - - - - - c2f63c36 by Matthew Pickering at 2024-09-26T21:33:48+01:00 Don't recache - - - - - 1 changed file: - hadrian/src/Rules/BinaryDist.hs Changes: ===================================== hadrian/src/Rules/BinaryDist.hs ===================================== @@ -153,9 +153,9 @@ buildBinDistDir root conf at BindistConfig{..} = do version <- setting ProjectVersion targetPlatform <- setting TargetPlatformFull - distDir <- Context.distDir (vanillaContext Stage1 rts) + distDir <- Context.distDir (vanillaContext library_stage rts) - let ghcBuildDir = root -/- stageString Stage1 + let ghcBuildDir = root -/- stageString library_stage bindistFilesDir = root -/- "bindist" -/- ghcVersionPretty ghcVersionPretty = "ghc-" ++ version ++ "-" ++ targetPlatform rtsIncludeDir = distDir -/- "include" @@ -227,8 +227,8 @@ buildBinDistDir root conf at BindistConfig{..} = do -- (c.f. #20267). -- Not going to work for cross - ghcPkgName <- programName (vanillaContext Stage1 ghcPkg) - cmd_ (bindistFilesDir -/- "bin" -/- ghcPkgName) ["recache"] + --ghcPkgName <- programName (vanillaContext Stage1 ghcPkg) + --cmd_ (bindistFilesDir -/- "bin" -/- ghcPkgName) ["recache"] need ["docs"] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e608df5ced6b207e9422f1401b139bf4f505e3f7...c2f63c368b6400753b80bef1da4a0e587757f2fe -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e608df5ced6b207e9422f1401b139bf4f505e3f7...c2f63c368b6400753b80bef1da4a0e587757f2fe You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 20:41:24 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Thu, 26 Sep 2024 16:41:24 -0400 Subject: [Git][ghc/ghc][wip/llvm_19] Bump LLVM upper bound to allow LLVM 19 Message-ID: <66f5c6f4cf1b_f364ad9429876093@gitlab.mail> Matthew Pickering pushed to branch wip/llvm_19 at Glasgow Haskell Compiler / GHC Commits: 381458e8 by Matthew Pickering at 2024-09-26T20:55:37+01:00 Bump LLVM upper bound to allow LLVM 19 Also bumps the ci-images commit so that the deb12 images uses LLVM 19 for testing. ------------------------- Metric Decrease: size_hello_artifact_gzip size_hello_unicode_gzip ------------------------- Fixes #25295 - - - - - 2 changed files: - .gitlab-ci.yml - configure.ac Changes: ===================================== .gitlab-ci.yml ===================================== @@ -2,7 +2,7 @@ variables: GIT_SSL_NO_VERIFY: "1" # Commit of ghc/ci-images repository from which to pull Docker images - DOCKER_REV: 486541129a8e7bf77c2cf7cd76ca998f690d5685 + DOCKER_REV: 575c404d23035c6e9b5815962e5c0e7d4f72ec88 # Sequential version number of all cached things. # Bump to invalidate GitLab CI cache. ===================================== configure.ac ===================================== @@ -528,7 +528,7 @@ AC_SUBST(InstallNameToolCmd) # versions of LLVM simultaneously, but that stopped working around # 3.5/3.6 release of LLVM. LlvmMinVersion=13 # inclusive -LlvmMaxVersion=19 # not inclusive +LlvmMaxVersion=20 # not inclusive AC_SUBST([LlvmMinVersion]) AC_SUBST([LlvmMaxVersion]) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/381458e8b2eae8f53f09d991b70d043a02c4dc66 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/381458e8b2eae8f53f09d991b70d043a02c4dc66 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Thu Sep 26 23:37:24 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Thu, 26 Sep 2024 19:37:24 -0400 Subject: [Git][ghc/ghc][wip/T25281] 15 commits: Don't compile `asBox` with -fprof-late Message-ID: <66f5f034b5397_3def162f9ff482428@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 8819aef9 by Sebastian Graf at 2024-09-26T23:08:47+01:00 Pmc: Improve implementation of -Wincomplete-record-selectors (#24824, #24891) We now incorporate the result type of unsaturated record selector applications as well as consider long-distance information in getField applications. See the updated Note [Detecting incomplete record selectors]. Fixes #24824 and #24891. - - - - - 10865673 by Sebastian Graf at 2024-09-26T23:08:47+01:00 Pmc: Improve warning messages of -Wincomplete-record-selectors ... as suggested by Adam Gundry in !12685. - - - - - 58cb0ecd by Simon Peyton Jones at 2024-09-26T23:08:47+01:00 Just a start on #25281 - - - - - c7463270 by Simon Peyton Jones at 2024-09-27T00:37:03+01:00 Wibbles - - - - - 30 changed files: - compiler/GHC/Builtin/Names.hs - compiler/GHC/Core/ConLike.hs - compiler/GHC/Driver/Pipeline/Execute.hs - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Errors/Ppr.hs - compiler/GHC/HsToCore/Errors/Types.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Types.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/CharClass.hs - compiler/GHC/Parser/Errors/Ppr.hs - compiler/GHC/Parser/Errors/Types.hs - compiler/GHC/Parser/Header.hs - compiler/GHC/Parser/Lexer.x - compiler/GHC/Parser/String.hs - compiler/GHC/Rename/Names.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Tc/Gen/App.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver.hs - compiler/GHC/Tc/TyCl/Utils.hs - compiler/GHC/Tc/Types/CtLoc.hs - compiler/GHC/Tc/Types/LclEnv.hs - compiler/GHC/Tc/Utils/Monad.hs - compiler/GHC/Types/Error/Codes.hs - compiler/GHC/Types/Id/Info.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/740c3d27332b9673175d4dfe4e1040d65ac5ba2a...c7463270bdb2fc0ce6cedf235d05c6573599a3cc -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/740c3d27332b9673175d4dfe4e1040d65ac5ba2a...c7463270bdb2fc0ce6cedf235d05c6573599a3cc You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 10:10:48 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 27 Sep 2024 06:10:48 -0400 Subject: [Git][ghc/ghc][master] Add entity information to HieFile #24544 Message-ID: <66f684a86674d_3ed54d291e044554a@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 5 changed files: - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Iface/Ext/Types.hs - + testsuite/tests/hiefile/should_run/T24544.hs - + testsuite/tests/hiefile/should_run/T24544.stdout - testsuite/tests/hiefile/should_run/all.T Changes: ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -33,7 +33,6 @@ import GHC.Types.Basic import GHC.Data.BooleanFormula import GHC.Core.Class ( className, classSCSelIds ) import GHC.Core.ConLike ( conLikeName ) -import GHC.Core.TyCon ( TyCon, tyConClass_maybe ) import GHC.Core.FVs import GHC.Core.DataCon ( dataConNonlinearType ) import GHC.Types.FieldLabel @@ -41,11 +40,12 @@ import GHC.Hs import GHC.Hs.Syn.Type import GHC.Utils.Monad ( concatMapM, MonadIO(liftIO) ) import GHC.Types.Id ( isDataConId_maybe ) -import GHC.Types.Name ( Name, nameSrcSpan, nameUnique ) +import GHC.Types.Name ( Name, nameSrcSpan, nameUnique, wiredInNameTyThing_maybe ) import GHC.Types.Name.Env ( NameEnv, emptyNameEnv, extendNameEnv, lookupNameEnv ) import GHC.Types.Name.Reader ( RecFieldInfo(..) ) import GHC.Types.SrcLoc import GHC.Core.Type ( Type ) +import GHC.Core.TyCon ( TyCon, tyConClass_maybe ) import GHC.Core.Predicate import GHC.Core.InstEnv import GHC.Tc.Types @@ -81,6 +81,8 @@ import Control.Monad.Trans.State.Strict import Control.Monad.Trans.Reader import Control.Monad.Trans.Class ( lift ) import Control.Applicative ( (<|>) ) +import GHC.Types.TypeEnv ( TypeEnv ) +import Control.Arrow ( second ) {- Note [Updating HieAst for changes in the GHC AST] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -229,6 +231,10 @@ data HieState = HieState -- These are placed at the top level Node in the HieAST after everything -- else has been generated -- This includes things like top level evidence bindings. + , type_env :: TypeEnv + -- tcg_type_env from TcGblEnv contains the type environment for the module + , entity_infos :: NameEntityInfo + -- ^ Information about entities in the module } addUnlocatedEvBind :: Var -> ContextInfo -> HieM () @@ -260,8 +266,20 @@ getUnlocatedEvBinds file = do pure $ (M.fromList nis, asts) +lookupAndInsertEntityName :: Name -> HieM () +lookupAndInsertEntityName name = do + m <- lift $ gets type_env + let tyThing = lookupNameEnv m name <|> wiredInNameTyThing_maybe name + insertEntityInfo name $ maybe (nameEntityInfo name) tyThingEntityInfo tyThing + +-- | Insert entity information for an identifier +insertEntityInfo :: Name -> S.Set EntityInfo -> HieM () +insertEntityInfo ident info = do + lift $ modify' $ \s -> + s { entity_infos = M.insertWith S.union ident info (entity_infos s) } + initState :: HieState -initState = HieState emptyNameEnv emptyDVarEnv +initState = HieState emptyNameEnv emptyDVarEnv mempty mempty class ModifyState a where -- See Note [Name Remapping] addSubstitution :: a -> a -> HieState -> HieState @@ -302,8 +320,9 @@ mkHieFileWithSource src_file src ms ts rs = let tc_binds = tcg_binds ts top_ev_binds = tcg_ev_binds ts insts = tcg_insts ts + tte = tcg_type_env ts tcs = tcg_tcs ts - (asts',arr) = getCompressedAsts tc_binds rs top_ev_binds insts tcs in + (asts',arr,entityInfos) = getCompressedAsts tc_binds rs top_ev_binds insts tcs tte in HieFile { hie_hs_file = src_file , hie_module = ms_mod ms @@ -312,18 +331,20 @@ mkHieFileWithSource src_file src ms ts rs = -- mkIfaceExports sorts the AvailInfos for stability , hie_exports = mkIfaceExports (tcg_exports ts) , hie_hs_src = src + , hie_entity_infos = entityInfos } -getCompressedAsts :: TypecheckedSource -> RenamedSource -> Bag EvBind -> [ClsInst] -> [TyCon] - -> (HieASTs TypeIndex, A.Array TypeIndex HieTypeFlat) -getCompressedAsts ts rs top_ev_binds insts tcs = - let asts = enrichHie ts rs top_ev_binds insts tcs in - compressTypes asts - -enrichHie :: TypecheckedSource -> RenamedSource -> Bag EvBind -> [ClsInst] -> [TyCon] - -> HieASTs Type -enrichHie ts (hsGrp, imports, exports, docs, modName) ev_bs insts tcs = - runIdentity $ flip evalStateT initState $ flip runReaderT SourceInfo $ do +getCompressedAsts :: TypecheckedSource -> RenamedSource -> Bag EvBind -> [ClsInst] -> [TyCon] -> TypeEnv + -> (HieASTs TypeIndex, A.Array TypeIndex HieTypeFlat, NameEntityInfo) +getCompressedAsts ts rs top_ev_binds insts tcs tte = + let (asts, infos) = enrichHie ts rs top_ev_binds insts tcs tte + add c (a, b) = (a,b,c) + in add infos $ compressTypes asts + +enrichHie :: TypecheckedSource -> RenamedSource -> Bag EvBind -> [ClsInst] -> [TyCon] -> TypeEnv + -> (HieASTs Type, NameEntityInfo) +enrichHie ts (hsGrp, imports, exports, docs, modName) ev_bs insts tcs tte = + second entity_infos $ runIdentity $ flip runStateT initState{type_env=tte} $ flip runReaderT SourceInfo $ do modName <- toHie (IEC Export <$> modName) tasts <- toHie $ fmap (BC RegularBind ModuleScope) ts rasts <- processGrp hsGrp @@ -418,6 +439,7 @@ bindingsOnly [] = pure [] bindingsOnly (C c n : xs) = do org <- ask rest <- bindingsOnly xs + lookupAndInsertEntityName n pure $ case nameSrcSpan n of RealSrcSpan span _ -> Node (mkSourcedNodeInfo org nodeinfo) span [] : rest where nodeinfo = NodeInfo S.empty [] (M.singleton (Right n) info) @@ -599,7 +621,7 @@ instance (ToHie a) => ToHie (Maybe a) where instance ToHie (IEContext (LocatedA ModuleName)) where toHie (IEC c (L (EpAnn (EpaSpan (RealSrcSpan span _)) _ _) mname)) = do org <- ask - pure $ [Node (mkSourcedNodeInfo org $ NodeInfo S.empty [] idents) span []] + pure [Node (mkSourcedNodeInfo org $ NodeInfo S.empty [] idents) span []] where details = mempty{identInfo = S.singleton (IEThing c)} idents = M.singleton (Left mname) details toHie _ = pure [] @@ -624,6 +646,9 @@ instance ToHie (Context (Located Var)) where ty = case isDataConId_maybe name' of Nothing -> varType name' Just dc -> dataConNonlinearType dc + -- insert the entity info for the name into the entity_infos map + insertEntityInfo (varName name) $ idEntityInfo name + insertEntityInfo (varName name') $ idEntityInfo name' pure [Node (mkSourcedNodeInfo org $ NodeInfo S.empty [] $ @@ -648,6 +673,9 @@ instance ToHie (Context (Located Name)) where let name = case lookupNameEnv m name' of Just var -> varName var Nothing -> name' + -- insert the entity info for the name into the entity_infos map + lookupAndInsertEntityName name + lookupAndInsertEntityName name' pure [Node (mkSourcedNodeInfo org $ NodeInfo S.empty [] $ ===================================== compiler/GHC/Iface/Ext/Types.hs ===================================== @@ -29,6 +29,13 @@ import GHC.Types.Avail import GHC.Types.Unique import qualified GHC.Utils.Outputable as O ( (<>) ) import GHC.Utils.Panic +import GHC.Core.ConLike ( ConLike(..) ) +import GHC.Core.TyCo.Rep ( Type(..) ) +import GHC.Core.Type ( coreFullView, isFunTy, Var (..) ) +import GHC.Core.TyCon ( isTypeSynonymTyCon, isClassTyCon, isFamilyTyCon ) +import GHC.Types.Id ( Id, isRecordSelector, isClassOpId ) +import GHC.Types.TyThing ( TyThing (..) ) +import GHC.Types.Var ( isTyVar, isFUNArg ) import qualified Data.Array as A import qualified Data.Map as M @@ -84,7 +91,17 @@ data HieFile = HieFile , hie_hs_src :: ByteString -- ^ Raw bytes of the initial Haskell source + + , hie_entity_infos :: NameEntityInfo + -- ^ Entity information for each `Name` in the `hie_asts` } + +type NameEntityInfo = M.Map Name (S.Set EntityInfo) + +instance Binary NameEntityInfo where + put_ bh m = put_ bh $ M.toList m + get bh = fmap M.fromList (get bh) + instance Binary HieFile where put_ bh hf = do put_ bh $ hie_hs_file hf @@ -93,6 +110,7 @@ instance Binary HieFile where put_ bh $ hie_asts hf put_ bh $ hie_exports hf put_ bh $ hie_hs_src hf + put_ bh $ hie_entity_infos hf get bh = HieFile <$> get bh @@ -101,6 +119,7 @@ instance Binary HieFile where <*> get bh <*> get bh <*> get bh + <*> get bh {- @@ -783,3 +802,84 @@ toHieName name (nameOccName name) (removeBufSpan $ nameSrcSpan name) | otherwise = LocalName (nameOccName name) (removeBufSpan $ nameSrcSpan name) + + +{- Note [Capture Entity Information] + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +We need to capture the entity information for the identifier in HieAst, so that +language tools and protocols can take advantage making use of it. + +Capture `EntityInfo` for a `Name` or `Id` in `renamedSource` or `typecheckedSource` +if it is a name, we ask the env for the `TyThing` then compute the `EntityInfo` from tyThing +if it is an Id, we compute the `EntityInfo` directly from Id + +see issue #24544 for more details +-} + + +-- | Entity information +-- `EntityInfo` is a simplified version of `TyThing` and richer version than `Namespace` in `OccName`. +-- It state the kind of the entity, such as `Variable`, `TypeVariable`, `DataConstructor`, etc.. +data EntityInfo + = EntityVariable + | EntityFunction + | EntityDataConstructor + | EntityTypeVariable + | EntityClassMethod + | EntityPatternSynonym + | EntityTypeConstructor + | EntityTypeClass + | EntityTypeSynonym + | EntityTypeFamily + | EntityRecordField + deriving (Eq, Ord, Enum, Show) + + +instance Outputable EntityInfo where + ppr EntityVariable = text "variable" + ppr EntityFunction = text "function" + ppr EntityDataConstructor = text "data constructor" + ppr EntityTypeVariable = text "type variable" + ppr EntityClassMethod = text "class method" + ppr EntityPatternSynonym = text "pattern synonym" + ppr EntityTypeConstructor = text "type constructor" + ppr EntityTypeClass = text "type class" + ppr EntityTypeSynonym = text "type synonym" + ppr EntityTypeFamily = text "type family" + ppr EntityRecordField = text "record field" + + +instance Binary EntityInfo where + put_ bh b = putByte bh (fromIntegral (fromEnum b)) + get bh = do x <- getByte bh; pure $! toEnum (fromIntegral x) + + +-- | Get the `EntityInfo` for an `Id` +idEntityInfo :: Id -> S.Set EntityInfo +idEntityInfo vid = S.fromList $ [EntityTypeVariable | isTyVar vid] <> [EntityFunction | isFunType $ varType vid] + <> [EntityRecordField | isRecordSelector vid] <> [EntityClassMethod | isClassOpId vid] <> [EntityVariable] + where + isFunType a = case coreFullView a of + ForAllTy _ t -> isFunType t + FunTy { ft_af = flg, ft_res = rhs } -> isFUNArg flg || isFunType rhs + _x -> isFunTy a + +-- | Get the `EntityInfo` for a `TyThing` +tyThingEntityInfo :: TyThing -> S.Set EntityInfo +tyThingEntityInfo ty = case ty of + AnId vid -> idEntityInfo vid + AConLike con -> case con of + RealDataCon _ -> S.singleton EntityDataConstructor + PatSynCon _ -> S.singleton EntityPatternSynonym + ATyCon tyCon -> S.fromList $ [EntityTypeSynonym | isTypeSynonymTyCon tyCon] <> [EntityTypeFamily | isFamilyTyCon tyCon] + <> [EntityTypeClass | isClassTyCon tyCon] <> [EntityTypeConstructor] + ACoAxiom _ -> S.empty + +nameEntityInfo :: Name -> S.Set EntityInfo +nameEntityInfo name + | isTyVarName name = S.fromList [EntityVariable, EntityTypeVariable] + | isDataConName name = S.singleton EntityDataConstructor + | isTcClsNameSpace (occNameSpace $ occName name) = S.singleton EntityTypeConstructor + | isFieldName name = S.fromList [EntityVariable, EntityRecordField] + | isVarName name = S.fromList [EntityVariable] + | otherwise = S.empty ===================================== testsuite/tests/hiefile/should_run/T24544.hs ===================================== @@ -0,0 +1,54 @@ +{-# LANGUAGE PatternSynonyms #-} +{-# LANGUAGE TypeFamilies #-} +module Main where + +import TestUtils +import qualified Data.Map.Strict as M +import qualified Data.Set as S +import Data.Either +import Data.Maybe +import Data.Bifunctor (first) +import GHC.Plugins (moduleNameString, nameStableString, nameOccName, occNameString, isDerivedOccName) +import GHC.Iface.Ext.Types +-- data EntityInfo +-- = +-- EntityVariable +boo = 1 +-- | EntityFunction +foo :: a -> a +-- | EntityTypeVariable +foo a = a +-- | EntityTypeConstructor +data DataFoo +-- | EntityDataConstructor + = DFoo +-- | EntityTypeClass +class ClassBoo a where +-- | EntityClassMethod + cboo :: a +-- | EntityPatternSynonym +pattern PatternFoo = 1 +-- | EntityTypeFamily +type family FamilyFoo +data family DataFamilyBar +-- | EntityTypeSynonym +type TypeSynonymFoo = Int +-- | EntityRecordField +data RecordFoo = RecordFoo { recordFoo :: Int } + +points :: [(Int,Int)] +points = [(16,1), (18,9), (20,1), (22,6), (24,6), (26,7), (28,2), (30,9), (32,13), (33,13), (35,6), (37,30)] + +getIdentifierEntityInfo :: HieFile -> Identifier -> S.Set EntityInfo +getIdentifierEntityInfo hf (Right ident) = M.findWithDefault S.empty ident (hie_entity_infos hf) +getIdentifierEntityInfo hf (Left _) = S.empty + +isNotDerived :: (Identifier, a) -> Bool +isNotDerived ((Right name), _) = not $ isDerivedOccName (nameOccName name) +isNotDerived ((Left _), _) = True +main = do + (df, hf) <- readTestHie "T24544.hie" + let asts = fmap (fromMaybe (error "nothing") . selectPoint hf) points + idents = concatMap (M.toList . sourcedNodeIdents . sourcedNodeInfo) asts + names = map (\(x, _) -> (either moduleNameString (occNameString . nameOccName) x, getIdentifierEntityInfo hf x)) $ filter isNotDerived idents + mapM_ (print) names ===================================== testsuite/tests/hiefile/should_run/T24544.stdout ===================================== @@ -0,0 +1,12 @@ +("boo",fromList [EntityVariable]) +("a",fromList [EntityVariable,EntityTypeVariable]) +("foo",fromList [EntityVariable,EntityFunction]) +("DataFoo",fromList [EntityTypeConstructor]) +("DFoo",fromList [EntityDataConstructor]) +("ClassBoo",fromList [EntityTypeConstructor,EntityTypeClass]) +("PatternFoo",fromList [EntityPatternSynonym]) +("FamilyFoo",fromList [EntityTypeConstructor,EntityTypeFamily]) +("DataFamilyBar",fromList [EntityTypeConstructor,EntityTypeFamily]) +("TypeSynonymFoo",fromList [EntityTypeConstructor,EntityTypeSynonym]) +("RecordFoo",fromList [EntityDataConstructor]) +("recordFoo",fromList [EntityVariable,EntityFunction,EntityRecordField]) \ No newline at end of file ===================================== testsuite/tests/hiefile/should_run/all.T ===================================== @@ -7,3 +7,4 @@ test('SpliceTypes', [req_th, extra_run_opts('"' + config.libdir + '"'), extra_fi test('HieVdq', [extra_run_opts('"' + config.libdir + '"'), extra_files(['TestUtils.hs'])], compile_and_run, ['-package ghc -fwrite-ide-info']) test('T23540', [extra_run_opts('"' + config.libdir + '"'), extra_files(['TestUtils.hs'])], compile_and_run, ['-package ghc -fwrite-ide-info']) test('T23120', [extra_run_opts('"' + config.libdir + '"'), extra_files(['TestUtils.hs'])], compile_and_run, ['-package ghc -fwrite-ide-info']) +test('T24544', [extra_run_opts('"' + config.libdir + '"'), extra_files(['TestUtils.hs'])], compile_and_run, ['-package ghc -fwrite-ide-info']) \ No newline at end of file View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1b39363bd673314df76b74f5f9c65af6fe84f1c8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1b39363bd673314df76b74f5f9c65af6fe84f1c8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 10:11:55 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 27 Sep 2024 06:11:55 -0400 Subject: [Git][ghc/ghc][master] 15 commits: The X86 SIMD patch. Message-ID: <66f684ebb6b44_3ed54d2a689061618@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 30 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/PPC/Ppr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1b39363bd673314df76b74f5f9c65af6fe84f1c8...12504a9ff2a40649c4eb0c3f696876b9bfe50a3d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1b39363bd673314df76b74f5f9c65af6fe84f1c8...12504a9ff2a40649c4eb0c3f696876b9bfe50a3d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 10:12:45 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 27 Sep 2024 06:12:45 -0400 Subject: [Git][ghc/ghc][master] Link bytecode from interface-stored core bindings in oneshot mode Message-ID: <66f6851d115ca_3ed54d2a689065983@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 21 changed files: - compiler/GHC/CoreToIface.hs - compiler/GHC/Driver/CodeOutput.hs - compiler/GHC/Driver/Main.hs - + compiler/GHC/Driver/Main.hs-boot - compiler/GHC/Driver/Make.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/External.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/A.hs - + testsuite/tests/bytecode/T25090/B.hs - + testsuite/tests/bytecode/T25090/C.hs - + testsuite/tests/bytecode/T25090/C.hs-boot - + testsuite/tests/bytecode/T25090/D.hs - + testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/T25090-debug.stderr - + testsuite/tests/bytecode/T25090/T25090.stdout - + testsuite/tests/bytecode/T25090/all.T Changes: ===================================== compiler/GHC/CoreToIface.hs ===================================== @@ -599,7 +599,7 @@ toIfaceTopBind b = in (top_bndr, rhs') -- The sharing behaviour is currently disabled due to #22807, and relies on - -- finished #220056 to be re-enabled. + -- finished #20056 to be re-enabled. disabledDueTo22807 = True already_has_unfolding b = not disabledDueTo22807 @@ -771,8 +771,8 @@ outside of the hs-boot loop. Note [Interface File with Core: Sharing RHSs] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -IMPORTANT: This optimisation is currently disabled due to #22027, it can be - re-enabled once #220056 is implemented. +IMPORTANT: This optimisation is currently disabled due to #22807, it can be + re-enabled once #22056 is implemented. In order to avoid duplicating definitions for bindings which already have unfoldings we do some minor headstands to avoid serialising the RHS of a definition if it has ===================================== compiler/GHC/Driver/CodeOutput.hs ===================================== @@ -275,7 +275,6 @@ outputForeignStubs Maybe FilePath) -- C file created outputForeignStubs logger tmpfs dflags unit_state mod location stubs = do - let stub_h = unsafeDecodeUtf $ mkStubPaths (initFinderOpts dflags) (moduleName mod) location stub_c <- newTempName logger tmpfs (tmpDir dflags) TFL_CurrentModule "c" case stubs of @@ -291,8 +290,6 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs stub_h_output_d = pprCode h_code stub_h_output_w = showSDoc dflags stub_h_output_d - createDirectoryIfMissing True (takeDirectory stub_h) - putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export header file" FormatC @@ -314,9 +311,23 @@ outputForeignStubs logger tmpfs dflags unit_state mod location stubs | platformMisc_libFFI $ platformMisc dflags = "#include \"rts/ghc_ffi.h\"\n" | otherwise = "" - stub_h_file_exists - <- outputForeignStubs_help stub_h stub_h_output_w - ("#include \n" ++ cplusplus_hdr) cplusplus_ftr + -- The header path is computed from the module source path, which + -- does not exist when loading interface core bindings for Template + -- Haskell for non-home modules (e.g. when compiling in separate + -- invocations of oneshot mode). + -- Stub headers are only generated for foreign exports. + -- Since those aren't supported for TH with bytecode at the moment, + -- it doesn't make much of a difference. + -- In any case, if a stub dir was specified explicitly by the user, it + -- would be used nonetheless. + stub_h_file_exists <- + case mkStubPaths (initFinderOpts dflags) (moduleName mod) location of + Nothing -> pure False + Just path -> do + let stub_h = unsafeDecodeUtf path + createDirectoryIfMissing True (takeDirectory stub_h) + outputForeignStubs_help stub_h stub_h_output_w + ("#include \n" ++ cplusplus_hdr) cplusplus_ftr putDumpFileMaybe logger Opt_D_dump_foreign "Foreign export stubs" FormatC stub_c_output_d ===================================== compiler/GHC/Driver/Main.hs ===================================== @@ -49,6 +49,7 @@ module GHC.Driver.Main , HscBackendAction (..), HscRecompStatus (..) , initModDetails , initWholeCoreBindings + , loadIfaceByteCode , hscMaybeWriteIface , hscCompileCmmFile @@ -105,6 +106,7 @@ module GHC.Driver.Main , showModuleIndex , hscAddSptEntries , writeInterfaceOnlyMode + , loadByteCode ) where import GHC.Prelude @@ -275,7 +277,8 @@ import GHC.SysTools (initSysTools) import GHC.SysTools.BaseDir (findTopDir) import Data.Data hiding (Fixity, TyCon) -import Data.List ( nub, isPrefixOf, partition ) +import Data.Functor ((<&>)) +import Data.List ( nub, isPrefixOf, partition ) import qualified Data.List.NonEmpty as NE import Control.Monad import Data.IORef @@ -972,19 +975,23 @@ loadByteCode iface mod_sum = do (mi_foreign iface) return (UpToDateItem (Linkable if_date this_mod (NE.singleton (CoreBindings fi)))) _ -> return $ outOfDateItemBecause MissingBytecode Nothing + -------------------------------------------------------------- -- Compilers -------------------------------------------------------------- +add_iface_to_hpt :: ModIface -> ModDetails -> HscEnv -> HscEnv +add_iface_to_hpt iface details = + hscUpdateHPT $ \ hpt -> + addToHpt hpt (moduleName (mi_module iface)) + (HomeModInfo iface details emptyHomeModInfoLinkable) -- Knot tying! See Note [Knot-tying typecheckIface] -- See Note [ModDetails and --make mode] initModDetails :: HscEnv -> ModIface -> IO ModDetails initModDetails hsc_env iface = fixIO $ \details' -> do - let act hpt = addToHpt hpt (moduleName $ mi_module iface) - (HomeModInfo iface details' emptyHomeModInfoLinkable) - let !hsc_env' = hscUpdateHPT act hsc_env + let !hsc_env' = add_iface_to_hpt iface details' hsc_env -- NB: This result is actually not that useful -- in one-shot mode, since we're not going to do -- any further typechecking. It's much more useful @@ -1012,8 +1019,52 @@ compile_for_interpreter hsc_env use = adapt_way want = if want (hscInterp hsc_env) then addWay else removeWay +-- | Assemble 'WholeCoreBindings' if the interface contains Core bindings. +iface_core_bindings :: ModIface -> ModLocation -> Maybe WholeCoreBindings +iface_core_bindings iface wcb_mod_location = + mi_extra_decls <&> \ wcb_bindings -> + WholeCoreBindings { + wcb_bindings, + wcb_module = mi_module, + wcb_mod_location, + wcb_foreign = mi_foreign + } + where + ModIface {mi_module, mi_extra_decls, mi_foreign} = iface + +-- | Return an 'IO' that hydrates Core bindings and compiles them to bytecode if +-- the interface contains any, using the supplied type env for typechecking. +-- +-- Unlike 'initWholeCoreBindings', this does not use lazy IO. +-- Instead, the 'IO' is only evaluated (in @get_link_deps@) when it is clear +-- that it will be used immediately (because we're linking TH with +-- @-fprefer-byte-code@ in oneshot mode), and the result is cached in +-- 'LoaderState'. +-- +-- 'initWholeCoreBindings' needs the laziness because it is used to populate +-- 'HomeModInfo', which is done preemptively, in anticipation of downstream +-- modules using the bytecode for TH in make mode, which might never happen. +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) +loadIfaceByteCode hsc_env iface location type_env = + compile <$> iface_core_bindings iface location + where + compile decls = do + (bcos, fos) <- compileWholeCoreBindings hsc_env type_env decls + linkable $ BCOs bcos :| [DotO fo ForeignObject | fo <- fos] + + linkable parts = do + if_time <- modificationTimeIfExists (ml_hi_file location) + time <- maybe getCurrentTime pure if_time + return $! Linkable time (mi_module iface) parts + -- | If the 'Linkable' contains Core bindings loaded from an interface, replace --- them with a lazy IO thunk that compiles them to bytecode and foreign objects. +-- them with a lazy IO thunk that compiles them to bytecode and foreign objects, +-- using the supplied environment for type checking. -- -- The laziness is necessary because this value is stored purely in a -- 'HomeModLinkable' in the home package table, rather than some dedicated @@ -1027,29 +1078,71 @@ compile_for_interpreter hsc_env use = -- -- This is sound because generateByteCode just depends on things already loaded -- in the interface file. -initWholeCoreBindings :: HscEnv -> ModIface -> ModDetails -> Linkable -> IO Linkable -initWholeCoreBindings hsc_env mod_iface details (Linkable utc_time this_mod uls) = +initWholeCoreBindings :: + HscEnv -> + ModIface -> + ModDetails -> + Linkable -> + IO Linkable +initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = Linkable utc_time this_mod <$> mapM go uls where - go (CoreBindings wcb at WholeCoreBindings {wcb_foreign, wcb_mod_location}) = do - types_var <- newIORef (md_types details) - let act hpt = addToHpt hpt (moduleName $ mi_module mod_iface) - (HomeModInfo mod_iface details emptyHomeModInfoLinkable) - kv = knotVarsFromModuleEnv (mkModuleEnv [(this_mod, types_var)]) - hsc_env' = hscUpdateHPT act hsc_env { hsc_type_env_vars = kv } - ~(bcos, fos) <- unsafeInterleaveIO $ do - core_binds <- initIfaceCheck (text "l") hsc_env' $ - typecheckWholeCoreBindings types_var wcb - (stubs, foreign_files) <- - decodeIfaceForeign logger (hsc_tmpfs hsc_env) - (tmpDir (hsc_dflags hsc_env)) wcb_foreign - let cgi_guts = CgInteractiveGuts this_mod core_binds - (typeEnvTyCons (md_types details)) stubs foreign_files - Nothing [] - trace_if logger (text "Generating ByteCode for" <+> ppr this_mod) - generateByteCode hsc_env cgi_guts wcb_mod_location - pure (LazyBCOs bcos fos) - go ul = return ul + go = \case + CoreBindings wcb -> do + ~(bco, fos) <- unsafeInterleaveIO $ + compileWholeCoreBindings hsc_env' type_env wcb + pure (LazyBCOs bco fos) + l -> pure l + + hsc_env' = add_iface_to_hpt iface details hsc_env + type_env = md_types details + +-- | Hydrate interface Core bindings and compile them to bytecode. +-- +-- This consists of: +-- +-- 1. Running a typechecking step to insert the global names that were removed +-- when the interface was written or were unavailable due to boot import +-- cycles, converting the bindings to 'CoreBind'. +-- +-- 2. Restoring the foreign build inputs from their serialized format, resulting +-- in a set of foreign import stubs and source files added via +-- 'qAddForeignFilePath'. +-- +-- 3. Generating bytecode and foreign objects from the results of the previous +-- steps using the usual pipeline actions. +compileWholeCoreBindings :: + HscEnv -> + TypeEnv -> + WholeCoreBindings -> + IO (CompiledByteCode, [FilePath]) +compileWholeCoreBindings hsc_env type_env wcb = do + core_binds <- typecheck + (stubs, foreign_files) <- decode_foreign + gen_bytecode core_binds stubs foreign_files + where + typecheck = do + types_var <- newIORef type_env + let + tc_env = hsc_env { + hsc_type_env_vars = + knotVarsFromModuleEnv (mkModuleEnv [(wcb_module, types_var)]) + } + initIfaceCheck (text "l") tc_env $ + typecheckWholeCoreBindings types_var wcb + + decode_foreign = + decodeIfaceForeign logger (hsc_tmpfs hsc_env) + (tmpDir (hsc_dflags hsc_env)) wcb_foreign + + gen_bytecode core_binds stubs foreign_files = do + let cgi_guts = CgInteractiveGuts wcb_module core_binds + (typeEnvTyCons type_env) stubs foreign_files + Nothing [] + trace_if logger (text "Generating ByteCode for" <+> ppr wcb_module) + generateByteCode hsc_env cgi_guts wcb_mod_location + + WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb logger = hsc_logger hsc_env ===================================== compiler/GHC/Driver/Main.hs-boot ===================================== @@ -0,0 +1,15 @@ +module GHC.Driver.Main where + +import GHC.Driver.Env.Types (HscEnv) +import GHC.Linker.Types (Linkable) +import GHC.Prelude.Basic +import GHC.Types.TypeEnv (TypeEnv) +import GHC.Unit.Module.Location (ModLocation) +import GHC.Unit.Module.ModIface (ModIface) + +loadIfaceByteCode :: + HscEnv -> + ModIface -> + ModLocation -> + TypeEnv -> + Maybe (IO Linkable) ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1310,8 +1310,10 @@ upsweep_mod hsc_env mHscMessage old_hmi summary mod_index nmods = do -- MP: This is a bit janky, because before you add the entries you have to extend the HPT with the module -- you just compiled. Another option, would be delay adding anything until after upsweep has finished, but I -- am unsure if this is sound (wrt running TH splices for example). - -- This function only does anything if the linkable produced is a BCO, which only happens with the - -- bytecode backend, no need to guard against the backend type additionally. + -- This function only does anything if the linkable produced is a BCO, which + -- used to only happen with the bytecode backend, but with + -- @-fprefer-byte-code@, @HomeModInfo@ has bytecode even when generating + -- object code, see #25230. addSptEntries (hscUpdateHPT (\hpt -> addToHpt hpt (ms_mod_name summary) hmi) hsc_env) (homeModInfoByteCode hmi) @@ -3007,7 +3009,7 @@ which can be checked easily using ghc-debug. a reference to the entire HscEnv, if we are not careful the HscEnv will contain the HomePackageTable at the time the interface was loaded and it will never be released. - Where? dontLeakTheHPT in GHC.Iface.Load + Where? dontLeakTheHUG in GHC.Iface.Load 2. No KnotVars are live at the end of upsweep (#20491) Why? KnotVars contains an old stale reference to the TypeEnv for modules ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -115,6 +115,7 @@ import Data.Map ( toList ) import System.FilePath import System.Directory import GHC.Driver.Env.KnotVars +import {-# source #-} GHC.Driver.Main (loadIfaceByteCode) import GHC.Iface.Errors.Types import Data.Function ((&)) @@ -474,7 +475,7 @@ loadInterface doc_str mod from -- Template Haskell original-name). Succeeded (iface, loc) -> let - loc_doc = text loc + loc_doc = text (ml_hi_file loc) in initIfaceLcl (mi_semantic_module iface) loc_doc (mi_boot iface) $ @@ -511,6 +512,7 @@ loadInterface doc_str mod from ; new_eps_rules <- tcIfaceRules ignore_prags (mi_rules iface) ; new_eps_anns <- tcIfaceAnnotations (mi_anns iface) ; new_eps_complete_matches <- tcIfaceCompleteMatches (mi_complete_matches iface) + ; purged_hsc_env <- getTopEnv ; let final_iface = iface & set_mi_decls (panic "No mi_decls in PIT") @@ -518,13 +520,26 @@ loadInterface doc_str mod from & set_mi_fam_insts (panic "No mi_fam_insts in PIT") & set_mi_rules (panic "No mi_rules in PIT") & set_mi_anns (panic "No mi_anns in PIT") + & set_mi_extra_decls (panic "No mi_extra_decls in PIT") - ; let bad_boot = mi_boot iface == IsBoot + bad_boot = mi_boot iface == IsBoot && isJust (lookupKnotVars (if_rec_types gbl_env) mod) -- Warn against an EPS-updating import -- of one's own boot file! (one-shot only) -- See Note [Loading your own hi-boot file] + -- Create an IO action that loads and compiles bytecode from Core + -- bindings. + -- + -- See Note [Interface Files with Core Definitions] + add_bytecode old + | Just action <- loadIfaceByteCode purged_hsc_env iface loc (mkNameEnv new_eps_decls) + = extendModuleEnv old mod action + -- Don't add an entry if the iface doesn't have 'extra_decls' + -- so 'get_link_deps' knows that it should load object code. + | otherwise + = old + ; warnPprTrace bad_boot "loadInterface" (ppr mod) $ updateEps_ $ \ eps -> if elemModuleEnv mod (eps_PIT eps) || is_external_sig mhome_unit iface @@ -536,6 +551,7 @@ loadInterface doc_str mod from eps { eps_PIT = extendModuleEnv (eps_PIT eps) mod final_iface, eps_PTE = addDeclsToPTE (eps_PTE eps) new_eps_decls, + eps_iface_bytecode = add_bytecode (eps_iface_bytecode eps), eps_rule_base = extendRuleBaseList (eps_rule_base eps) new_eps_rules, eps_complete_matches @@ -569,7 +585,7 @@ loadInterface doc_str mod from {- Note [Loading your own hi-boot file] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Generally speaking, when compiling module M, we should not -load M.hi boot into the EPS. After all, we are very shortly +load M.hi-boot into the EPS. After all, we are very shortly going to have full information about M. Moreover, see Note [Do not update EPS with your own hi-boot] in GHC.Iface.Recomp. @@ -698,7 +714,7 @@ computeInterface -> SDoc -> IsBootInterface -> Module - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) computeInterface hsc_env doc_str hi_boot_file mod0 = do massert (not (isHoleModule mod0)) let mhome_unit = hsc_home_unit_maybe hsc_env @@ -845,7 +861,7 @@ findAndReadIface -- this to check the consistency of the requirements of the -- module we read out. -> IsBootInterface -- ^ Looking for .hi-boot or .hi file - -> IO (MaybeErr MissingInterfaceError (ModIface, FilePath)) + -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation)) findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let profile = targetProfile dflags @@ -875,7 +891,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do let iface = case ghcPrimIfaceHook hooks of Nothing -> ghcPrimIface Just h -> h - return (Succeeded (iface, "")) + return (Succeeded (iface, panic "GHC.Prim ModLocation (findAndReadIface)")) else do let fopts = initFinderOpts dflags -- Look for the file @@ -900,7 +916,7 @@ findAndReadIface hsc_env doc_str mod wanted_mod hi_boot_file = do iface loc case r2 of Failed sdoc -> return (Failed sdoc) - Succeeded {} -> return $ Succeeded (iface,_fp) + Succeeded {} -> return $ Succeeded (iface, loc) err -> do trace_if logger (text "...not found") return $ Failed $ cannotFindInterface ===================================== compiler/GHC/IfaceToCore.hs ===================================== @@ -904,11 +904,11 @@ tcTopIfaceBindings :: IORef TypeEnv -> [IfaceBindingX IfaceMaybeRhs IfaceTopBndr -> IfL [CoreBind] tcTopIfaceBindings ty_var ver_decls = do - int <- mapM tcTopBinders ver_decls + int <- mapM tcTopBinders ver_decls let all_ids :: [Id] = concatMap toList int liftIO $ modifyIORef ty_var (flip extendTypeEnvList (map AnId all_ids)) - extendIfaceIdEnv all_ids $ mapM (tc_iface_bindings) int + extendIfaceIdEnv all_ids $ mapM tc_iface_bindings int tcTopBinders :: IfaceBindingX a IfaceTopBndrInfo -> IfL (IfaceBindingX a Id) tcTopBinders = traverse mk_top_id ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -60,16 +60,16 @@ import System.Directory data LinkDepsOpts = LinkDepsOpts { ldObjSuffix :: !String -- ^ Suffix of .o files , ldOneShotMode :: !Bool -- ^ Is the driver in one-shot mode? - , ldModuleGraph :: !ModuleGraph -- ^ Module graph - , ldUnitEnv :: !UnitEnv -- ^ Unit environment + , ldModuleGraph :: !ModuleGraph + , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages - , ldFinderCache :: !FinderCache -- ^ Finder cache - , ldFinderOpts :: !FinderOpts -- ^ Finder options , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways - , ldLoadIface :: SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface) - -- ^ Interface loader function + , ldFinderCache :: !FinderCache + , ldFinderOpts :: !FinderOpts + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) + , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) } data LinkDeps = LinkDeps @@ -275,21 +275,21 @@ get_link_deps opts pls maybe_normal_osuf span mods = do case ue_homeUnit unit_env of Nothing -> no_obj mod Just home_unit -> do - - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc mod -> found loc mod - _ -> no_obj (moduleName mod) + from_bc <- ldLoadByteCode opts mod + maybe (fallback_no_bytecode home_unit mod) pure from_bc where - found loc mod = do { - -- ...and then find the linkable for it - mb_lnk <- findObjectLinkableMaybe mod loc ; - case mb_lnk of { - Nothing -> no_obj mod ; - Just lnk -> adjust_linkable lnk - }} + + fallback_no_bytecode home_unit mod = do + let fc = ldFinderCache opts + let fopts = ldFinderOpts opts + mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) + case mb_stuff of + Found loc _ -> do + mb_lnk <- findObjectLinkableMaybe mod loc + case mb_lnk of + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + _ -> no_obj (moduleName mod) adjust_linkable lnk | Just new_osuf <- maybe_normal_osuf = do ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -76,6 +76,7 @@ import GHC.Utils.Logger import GHC.Utils.TmpFs import GHC.Unit.Env +import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) import GHC.Unit.Module import GHC.Unit.State as Packages @@ -641,18 +642,23 @@ initLinkDepsOpts hsc_env = opts , ldOneShotMode = isOneShot (ghcMode dflags) , ldModuleGraph = hsc_mod_graph hsc_env , ldUnitEnv = hsc_unit_env hsc_env - , ldLoadIface = load_iface , ldPprOpts = initSDocContext dflags defaultUserStyle , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags + , ldLoadIface + , ldLoadByteCode } dflags = hsc_dflags hsc_env - load_iface msg mod = initIfaceCheck (text "loader") hsc_env + ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env $ loadInterface msg mod (ImportByUser NotBoot) + ldLoadByteCode mod = do + EPS {eps_iface_bytecode} <- hscEPS hsc_env + sequence (lookupModuleEnv eps_iface_bytecode mod) + {- ********************************************************************** ===================================== compiler/GHC/Unit/External.hs ===================================== @@ -31,6 +31,8 @@ import GHC.Types.CompleteMatch import GHC.Types.TypeEnv import GHC.Types.Unique.DSet +import GHC.Linker.Types (Linkable) + import Data.IORef @@ -68,6 +70,7 @@ initExternalPackageState = EPS , eps_PIT = emptyPackageIfaceTable , eps_free_holes = emptyInstalledModuleEnv , eps_PTE = emptyTypeEnv + , eps_iface_bytecode = emptyModuleEnv , eps_inst_env = emptyInstEnv , eps_fam_inst_env = emptyFamInstEnv , eps_rule_base = mkRuleBase builtinRules @@ -139,6 +142,12 @@ data ExternalPackageState -- interface files we have sucked in. The domain of -- the mapping is external-package modules + -- | If an interface was written with @-fwrite-if-simplified-core@, this + -- will contain an IO action that compiles bytecode from core bindings. + -- + -- See Note [Interface Files with Core Definitions] + eps_iface_bytecode :: !(ModuleEnv (IO Linkable)), + eps_inst_env :: !PackageInstEnv, -- ^ The total 'InstEnv' accumulated -- from all the external-package modules eps_fam_inst_env :: !PackageFamInstEnv,-- ^ The total 'FamInstEnv' accumulated ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -38,7 +38,6 @@ import GHC.Platform.Ways import GHC.Builtin.Names ( gHC_PRIM ) -import GHC.Data.Maybe ( expectJust ) import GHC.Data.OsPath import GHC.Unit.Env @@ -60,6 +59,7 @@ import GHC.Types.PkgQual import GHC.Fingerprint import Data.IORef import System.Directory.OsPath +import Control.Applicative ((<|>)) import Control.Monad import Data.Time import qualified Data.Map as M @@ -711,27 +711,27 @@ mkHiePath fopts basename mod_basename = hie_basename <.> hiesuf -- We don't have to store these in ModLocations, because they can be derived -- from other available information, and they're only rarely needed. +-- | Compute the file name of a header file for foreign stubs, using either the +-- directory explicitly specified in the command line option @-stubdir@, or the +-- directory of the module's source file. +-- +-- When compiling bytecode from interface Core bindings, @ModLocation@ does not +-- contain a source file path, so the header isn't written. +-- This doesn't have an impact, since we cannot support headers importing +-- Haskell symbols defined in bytecode for TH whatsoever at the moment. mkStubPaths :: FinderOpts -> ModuleName -> ModLocation - -> OsPath - -mkStubPaths fopts mod location - = let - stubdir = finder_stubDir fopts - - mod_basename = unsafeEncodeUtf $ moduleNameSlashes mod - src_basename = OsPath.dropExtension $ expectJust "mkStubPaths" - (ml_hs_file_ospath location) - - stub_basename0 - | Just dir <- stubdir = dir mod_basename - | otherwise = src_basename + -> Maybe OsPath +mkStubPaths fopts mod location = do + stub_basename <- in_stub_dir <|> src_basename + pure (stub_basename `mappend` os "_stub" <.> os "h") + where + in_stub_dir = ( mod_basename) <$> (finder_stubDir fopts) - stub_basename = stub_basename0 `mappend` os "_stub" - in - stub_basename <.> os "h" + mod_basename = unsafeEncodeUtf $ moduleNameSlashes mod + src_basename = OsPath.dropExtension <$> ml_hs_file_ospath location -- ----------------------------------------------------------------------------- -- findLinkable isn't related to the other stuff in here, ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -28,12 +28,12 @@ import System.FilePath (takeExtension) {- Note [Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A interface file can optionally contain the definitions of all core bindings, this is enabled by the flag `-fwrite-if-simplified-core`. This provides everything needed in addition to the normal ModIface and ModDetails -to restart compilation after typechecking to generate bytecode. The `fi_bindings` field +to restart compilation after typechecking to generate bytecode. The `wcb_bindings` field is stored in the normal interface file and the other fields populated whilst loading the interface file. @@ -62,8 +62,55 @@ after whatever simplification the user requested has been performed. So the simp of the interface file agree with the optimisation level as reported by the interface file. +The lifecycle differs beyond laziness depending on the provenance of a module. +In all cases, the main consumer for interface bytecode is 'get_link_deps', which +traverses a splice's or GHCi expression's dependencies and collects the needed +build artifacts, which can be objects or bytecode, depending on the build +settings. + +1. In make mode, all eligible modules are part of the dependency graph. + Their interfaces are loaded unconditionally and in dependency order by the + compilation manager, and each module's bytecode is prepared before its + dependents are compiled, in one of two ways: + + - If the interface file for a module is missing or out of sync with its + source, it is recompiled and bytecode is generated directly and + immediately, not involving 'WholeCoreBindings' (in 'runHscBackendPhase'). + + - If the interface file is up to date, no compilation is performed, and a + lazy thunk generating bytecode from interface Core bindings is created in + 'compileOne'', which will only be compiled if a downstream module contains + a splice that depends on it, as described above. + + In both cases, the bytecode 'Linkable' is stored in a 'HomeModLinkable' in + the Home Unit Graph, lazy or not. + +2. In oneshot mode, which compiles individual modules without a shared home unit + graph, a previously compiled module is not reprocessed as described for make + mode above. + When 'get_link_deps' encounters a dependency on a local module, it requests + its bytecode from the External Package State, who loads the interface + on-demand. + + Since the EPS stores interfaces for all package dependencies in addition to + local modules in oneshot mode, it has a substantial memory footprint. + We try to curtail that by extracting important data into specialized fields + in the EPS, and retaining only a few fields of 'ModIface' by overwriting the + others with bottom values. + + In order to avoid keeping around all of the interface's components needed for + compiling bytecode, we instead store an IO action in 'eps_iface_bytecode'. + When 'get_link_deps' evaluates this action, the result is not retained in the + EPS, but stored in 'LoaderState', where it may eventually get evicted to free + up the memory. + This IO action retains the dehydrated Core bindings from the interface in its + closure. + Like the bytecode 'Linkable' stored in 'LoaderState', this is preferable to + storing the intermediate representation as rehydrated Core bindings, since + the latter have a significantly greater memory footprint. + Note [Size of Interface Files with Core Definitions] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ How much overhead does `-fwrite-if-simplified-core` add to a typical interface file? As an experiment I compiled the `Cabal` library and `ghc` library (Aug 22) with ===================================== testsuite/tests/bytecode/T25090/A.hs ===================================== @@ -0,0 +1,7 @@ +{-# language TemplateHaskell #-} +module Main where + +import D + +main :: IO () +main = putStrLn (show ($splc :: Int)) ===================================== testsuite/tests/bytecode/T25090/B.hs ===================================== @@ -0,0 +1,5 @@ +module B where + +import {-# source #-} C (C) + +data B = B C ===================================== testsuite/tests/bytecode/T25090/C.hs ===================================== @@ -0,0 +1,8 @@ +module C where + +import B + +data C = C Int + +b :: B +b = B (C 2024) ===================================== testsuite/tests/bytecode/T25090/C.hs-boot ===================================== @@ -0,0 +1,3 @@ +module C where + +data C ===================================== testsuite/tests/bytecode/T25090/D.hs ===================================== @@ -0,0 +1,12 @@ +module D where + +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +import B +import C + +splc :: ExpQ +splc = + lift @_ @Int num + where + B (C num) = b ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -0,0 +1,21 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +# Verify that the object files aren't linked by clobbering them. +T25090a: + $(TEST_HC) -c -fbyte-code-and-object-code C.hs-boot + $(TEST_HC) -c -fbyte-code-and-object-code B.hs + $(TEST_HC) -c -fbyte-code-and-object-code C.hs + echo 'corrupt' > B.o + echo 'corrupt' > C.o + echo 'corrupt' > C.o-boot + $(TEST_HC) -c -fbyte-code-and-object-code D.hs + echo 'corrupt' > D.o + $(TEST_HC) -c -fbyte-code-and-object-code -fprefer-byte-code A.hs + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A.o -o exe + ./exe + +T25090b: + $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 + ./exe ===================================== testsuite/tests/bytecode/T25090/T25090-debug.stderr ===================================== @@ -0,0 +1,6 @@ +WARNING: + loadInterface + C + Call stack: + CallStack (from HasCallStack): + warnPprTrace, called at compiler/GHC/Iface/Load.hs:: in :GHC.Iface.Load ===================================== testsuite/tests/bytecode/T25090/T25090.stdout ===================================== @@ -0,0 +1 @@ +2024 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -0,0 +1,18 @@ +# This test compiles the boot file separately from its source file, which causes +# a debug assertion warning. +# Since this appears to be intentional according to the Note [Loading your own hi-boot file], +# the warning is added to the expected stderr for debugged builds. +def test_T25090(name): + assert_warn_spec = {'stderr': 'T25090-debug.stderr'} + extra_specs = assert_warn_spec if name == 'T25090a' and compiler_debugged() else {} + return test(name, + [extra_files(['A.hs', 'B.hs', 'C.hs-boot', 'C.hs', 'D.hs']), + req_th, + js_skip, + use_specs(dict(stdout = 'T25090.stdout', **extra_specs)), + ], + makefile_test, + []) + +test_T25090('T25090a') +test_T25090('T25090b') View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2bb1e8df8be1d31094b3160114a38a3e8d5ec963 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2bb1e8df8be1d31094b3160114a38a3e8d5ec963 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 10:13:18 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 27 Sep 2024 06:13:18 -0400 Subject: [Git][ghc/ghc][master] ci: Fix variable inheritence for ghcup-metadata testing job Message-ID: <66f6853e6f484_3ed54d291e5467574@gitlab.mail> Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 1 changed file: - .gitlab-ci.yml Changes: ===================================== .gitlab-ci.yml ===================================== @@ -1231,10 +1231,17 @@ ghcup-metadata-release: UPSTREAM_PROJECT_ID: "$CI_PROJECT_ID" UPSTREAM_PIPELINE_ID: "$CI_PIPELINE_ID" RELEASE_JOB: "$RELEASE_JOB" + # Do not inherit global variables (such as CONFIGURE_ARGS) as these take + # precedence over the variables defined in the downstream job. + inherit: + variables: false trigger: project: "ghc/ghcup-ci" branch: "upstream-testing" strategy: "depend" + forward: + yaml_variables: true + pipeline_variables: false ghcup-metadata-testing-nightly: needs: View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 10:43:52 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Fri, 27 Sep 2024 06:43:52 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 21 commits: Add entity information to HieFile #24544 Message-ID: <66f68c68302a9_1f6287ecff484ca@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - a0e5c3d3 by Vladislav Zavialov at 2024-09-27T06:43:32-04:00 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - 6f5ba343 by Leo at 2024-09-27T06:43:33-04:00 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 18778847 by Matthew Pickering at 2024-09-27T06:43:35-04:00 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/97e0a69372e897cb8978982f5958d1a3e847f379...1877884729ffd43bf02a7936c4ead50f1a9ba2dc -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/97e0a69372e897cb8978982f5958d1a3e847f379...1877884729ffd43bf02a7936c4ead50f1a9ba2dc You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 11:10:51 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Fri, 27 Sep 2024 07:10:51 -0400 Subject: [Git][ghc/ghc][wip/T25281] More wibbles Message-ID: <66f692bbb5dd_3f1cfabc57041e5@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: e2c28824 by Simon Peyton Jones at 2024-09-27T12:10:27+01:00 More wibbles - - - - - 16 changed files: - compiler/GHC/Core/Type.hs - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match/Literal.hs - compiler/GHC/Rename/Env.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Rename/Pat.hs - compiler/GHC/Rename/Utils.hs - compiler/GHC/Tc/Gen/Do.hs - compiler/GHC/Tc/Gen/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Module.hs - compiler/GHC/Tc/Utils/Instantiate.hs - testsuite/tests/linear/should_fail/LinearTHFail2.stderr - testsuite/tests/linear/should_fail/LinearTHFail3.stderr Changes: ===================================== compiler/GHC/Core/Type.hs ===================================== @@ -33,7 +33,7 @@ module GHC.Core.Type ( mkScaledFunTys, mkInvisFunTy, mkInvisFunTys, tcMkVisFunTy, tcMkScaledFunTys, tcMkInvisFunTy, - splitFunTy, splitFunTy_maybe, + splitFunTy, splitFunTy_maybe, splitVisibleFunTy_maybe, splitFunTys, funResultTy, funArgTy, funTyConAppTy_maybe, funTyFlagTyCon, tyConAppFunTy_maybe, tyConAppFunCo_maybe, @@ -1443,6 +1443,15 @@ splitFunTy_maybe ty | FunTy af w arg res <- coreFullView ty = Just (af, w, arg, res) | otherwise = Nothing +{-# INLINE splitVisibleFunTy_maybe #-} +splitVisibleFunTy_maybe :: Type -> Maybe (Type, Type) +-- ^ Works on visible function types only (t1 -> t2), and +-- returns t1 and t2, but not the multiplicity +splitVisibleFunTy_maybe ty + | FunTy af _ arg res <- coreFullView ty + , isVisibleFunArg af = Just (arg, res) + | otherwise = Nothing + splitFunTys :: Type -> ([Scaled Type], Type) splitFunTys ty = split [] ty ty where ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -146,11 +146,6 @@ noSyntaxExpr = case ghcPass @p of mkSyntaxExpr :: HsExpr GhcRn -> SyntaxExprRn mkSyntaxExpr = SyntaxExprRn --- | Make a 'SyntaxExpr' from a 'Name' (the "rn" is because this is used in the --- renamer). -mkRnSyntaxExpr :: Name -> SyntaxExprRn -mkRnSyntaxExpr name = SyntaxExprRn $ HsVar noExtField $ noLocA name - instance Outputable SyntaxExprRn where ppr (SyntaxExprRn expr) = ppr expr ppr NoSyntaxExprRn = text "" @@ -653,7 +648,7 @@ ppr_lexpr e = ppr_expr (unLoc e) ppr_expr :: forall p. (OutputableBndrId p) => HsExpr (GhcPass p) -> SDoc -ppr_expr (HsVar _ (L _ v)) = pprPrefixOcc v +ppr_expr (HsVar _ (L loc v)) = pprPrefixOcc v ppr_expr (HsUnboundVar _ uv) = pprPrefixOcc uv ppr_expr (HsRecSel _ f) = pprPrefixOcc f ppr_expr (HsIPVar _ v) = ppr v ===================================== compiler/GHC/Hs/Utils.hs ===================================== @@ -53,10 +53,10 @@ module GHC.Hs.Utils( mkLHsPar, mkHsCmdWrap, mkLHsCmdWrap, mkHsCmdIf, mkConLikeTc, - nlHsTyApp, nlHsTyApps, nlHsVar, nl_HsVar, nlHsDataCon, + nlHsTyApp, nlHsTyApps, nlHsVar, nlHsDataCon, nlHsLit, nlHsApp, nlHsApps, nlHsSyntaxApps, nlHsIntLit, nlHsVarApps, - nlHsDo, nlHsOpApp, nlHsLam, nlHsPar, nlHsIf, nlHsCase, nlList, + nlHsDo, nlHsOpApp, nlHsPar, nlHsIf, nlHsCase, nlList, mkLHsTupleExpr, mkLHsVarTuple, missingTupArg, mkLocatedList, nlAscribe, @@ -504,10 +504,6 @@ nlHsVar :: IsSrcSpanAnn p a => IdP (GhcPass p) -> LHsExpr (GhcPass p) nlHsVar n = noLocA (HsVar noExtField (noLocA n)) -nl_HsVar :: IsSrcSpanAnn p a - => IdP (GhcPass p) -> HsExpr (GhcPass p) -nl_HsVar n = HsVar noExtField (noLocA n) - -- | NB: Only for 'LHsExpr' 'Id'. nlHsDataCon :: DataCon -> LHsExpr GhcTc nlHsDataCon con = noLocA (mkConLikeTc (RealDataCon con)) @@ -602,16 +598,11 @@ nlHsDo ctxt stmts = noLocA (mkHsDo ctxt (noLocA stmts)) nlHsOpApp :: LHsExpr GhcPs -> IdP GhcPs -> LHsExpr GhcPs -> LHsExpr GhcPs nlHsOpApp e1 op e2 = noLocA (mkHsOpApp e1 op e2) -nlHsLam :: LMatch GhcPs (LHsExpr GhcPs) -> LHsExpr GhcPs nlHsPar :: IsPass p => LHsExpr (GhcPass p) -> LHsExpr (GhcPass p) nlHsCase :: LHsExpr GhcPs -> [LMatch GhcPs (LHsExpr GhcPs)] -> LHsExpr GhcPs nlList :: [LHsExpr GhcPs] -> LHsExpr GhcPs --- AZ:Is this used? -nlHsLam match = noLocA $ HsLam noAnn LamSingle - $ mkMatchGroup (Generated OtherExpansion SkipPmc) (noLocA [match]) - nlHsPar e = noLocA (gHsPar e) -- nlHsIf should generate if-expressions which are NOT subject to ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -30,6 +30,7 @@ import GHC.HsToCore.Utils import GHC.HsToCore.Arrows import GHC.HsToCore.Monad import GHC.HsToCore.Pmc +import GHC.HsToCore.Pmc.Utils import GHC.HsToCore.Errors.Types import GHC.HsToCore.Quote import GHC.HsToCore.Ticks (stripTicksTopHsExpr) @@ -58,7 +59,6 @@ import GHC.Types.Name hiding (varName) import GHC.Types.Name.Reader( lookupGRE_FieldLabel ) import GHC.Types.CostCentre import GHC.Types.Id -import GHC.Types.Var( isVisibleFunArg ) import GHC.Types.Id.Info import GHC.Types.Id.Make import GHC.Types.Basic @@ -74,6 +74,7 @@ import GHC.Builtin.Names import GHC.Utils.Misc import GHC.Utils.Outputable as Outputable import GHC.Utils.Panic +import Control.Arrow ( first ) import Control.Monad import Data.Maybe( isJust ) @@ -626,20 +627,20 @@ ds_app (XExpr (ConLikeTc con tvs tys)) _hs_args core_args ds_app (HsRecSel _ (FieldOcc { foExt = sel_id })) _hs_args core_args = ds_app_rec_sel sel_id core_args -ds_app (HsVar _ (L _ fun_id)) hs_args core_args - = ds_app_var fun_id hs_args core_args +ds_app (HsVar _ lfun) hs_args core_args + = ds_app_var lfun hs_args core_args ds_app e _hs_args core_args = do { core_e <- dsExpr e ; return (mkCoreApps core_e core_args) } --------------- -ds_app_var :: Id -> [LHsExpr GhcTc] -> [CoreExpr] -> DsM CoreExpr +ds_app_var :: LocatedN Id -> [LHsExpr GhcTc] -> [CoreExpr] -> DsM CoreExpr -- Desugar an application with HsVar at the head -ds_app_var fun_id hs_args core_args - | fun_id `hasKey` getFieldClassOpKey - , (_k : Type x_ty : Type r_ty : _a_ty : _dict : rest_args) <- core_args - -- A getField applications. General form: +ds_app_var (L loc fun_id) hs_args core_args + + ----------------------- + -- Deal with getField applications. General form: -- getField -- @GHC.Types.Symbol {k} -- @"sel" x_ty @@ -651,6 +652,8 @@ ds_app_var fun_id hs_args core_args -- $dHasField = sel |> (co :: T -> Int ~R# HasField "sel" T Int) -- Alas, we cannot simply look at the unfolding of $dHasField below because it -- has not been set yet, so we have to reconstruct the selector Id from the types. + | fun_id `hasKey` getFieldClassOpKey + , (_k : Type x_ty : Type r_ty : _a_ty : _dict : rest_args) <- core_args = do { fam_inst_envs <- dsGetFamInstEnvs ; rdr_env <- dsGetGlobalRdrEnv -- Look up the field named x/"sel" in the type r/T @@ -662,23 +665,46 @@ ds_app_var fun_id hs_args core_args ; ds_app_rec_sel sel_id rest_args } _ -> ds_app_done fun_id core_args } - | fun_id `hasKey` thenMClassOpKey -- it is a (>>) - , Type arg_ty : _ <- core_args - , hs_arg : _ <- hs_args + ----------------------- + -- Warn about identities for (fromInteger :: Integer -> Integer) etc + -- They all have a type like: forall . => arg_ty -> res_ty + | idName fun_id `elem` numericConversionNames + , let (inst_fun_ty, _) = apply_type_args fun_id core_args + , (_, conv_ty) <- splitInvisPiTys inst_fun_ty + , Just (arg_ty, res_ty) <- splitVisibleFunTy_maybe conv_ty + = do { dflags <- getDynFlags + ; when (wopt Opt_WarnIdentities dflags + && arg_ty `eqType` res_ty) $ + -- So we are converting ty -> ty + diagnosticDs (DsIdentitiesFound fun_id conv_ty) + + ; ds_app_done fun_id core_args } + + ----------------------- -- Warn about unused return value in -- do { ...; e; ... } when e returns (say) an Int - = do { loc <- getSrcSpanDs - ; when (isGeneratedSrcSpan loc) $ -- it is compiler generated (>>) - warnDiscardedDoBindings hs_arg arg_ty + | fun_id `hasKey` thenMClassOpKey -- It is the built-in Prelude.(>>) + -- (>>) :: forall m. Monad m => forall a b. m a -> (b->m b) -> m b + , Type m_ty : _dict : Type arg_ty : _ <- core_args + , hs_arg : _ <- hs_args + = do { tracePm ">>" (ppr loc $$ ppr arg_ty $$ ppr (isGeneratedSrcSpan (locA loc))) + ; when (isGeneratedSrcSpan (locA loc)) $ -- It is a compiler-generated (>>) + warnDiscardedDoBindings hs_arg m_ty arg_ty ; ds_app_done fun_id core_args } - | fun_id `hasKey` noinlineIdKey -- See Note [noinlineId magic] in GHC.Types.Id.Make + ----------------------- + -- Deal with `noinline` + -- See Note [noinlineId magic] in GHC.Types.Id.Make + | fun_id `hasKey` noinlineIdKey , Type _ : arg1 : rest_args <- core_args , (inner_fun, inner_args) <- collectArgs arg1 = return (Var fun_id `App` Type (exprType inner_fun) `App` inner_fun `mkCoreApps` inner_args `mkCoreApps` rest_args) - | fun_id `hasKey` seqIdKey -- Note [Desugaring seq], points (1) and (2) + ----------------------- + -- Deal with `seq` + -- See Note [Desugaring seq], points (1) and (2) + | fun_id `hasKey` seqIdKey , Type _r : Type ty1 : Type ty2 : arg1 : arg2 : rest_args <- core_args , let case_bndr = case arg1 of Var v1 | isInternalName (idName v1) @@ -687,6 +713,8 @@ ds_app_var fun_id hs_args core_args = return (Case arg1 case_bndr ty2 [Alt DEFAULT [] arg2] `mkCoreApps` rest_args) + ----------------------- + -- Phew! No more special cases. Just build an applications | otherwise = ds_app_done fun_id core_args @@ -707,7 +735,7 @@ ds_app_rec_sel sel_id core_args -- data T = T1 | T2 {s :: Bool} -- g y = map s y -- Warn here -- f x = s x -- No warning here - ; let (fun, val_args) = apply_type_args (Var sel_id) core_args + ; let (fun_ty, val_args) = apply_type_args sel_id core_args -- Type-based check. -- See (3) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc @@ -718,8 +746,7 @@ ds_app_rec_sel sel_id core_args -- No value argument, but the selector is -- applied to all its type arguments - [] | Just (af, _, val_arg_ty, _) <- splitFunTy_maybe (exprType fun) - , isVisibleFunArg af + [] | Just (val_arg_ty, _) <- splitVisibleFunTy_maybe fun_ty -> do { dummy <- newSysLocalDs (Scaled ManyTy val_arg_ty) ; pmcRecSel sel_id (Var dummy) } @@ -735,11 +762,23 @@ ds_app_rec_sel sel_id core_args | otherwise = pprPanic "ds_app_rec_sel" (ppr sel_id $$ ppr (idDetails sel_id)) where - -- apply_type_args applies the record selector to its - -- initial type args, returning the remaining args, if any - apply_type_args fun (arg : args) - | isTypeArg arg = apply_type_args (App fun arg) args - apply_type_args fun args = (fun,args) + +-- apply_type_args applies the record selector to its +-- initial type args, returning the remaining args, if any +apply_type_args :: Id -> [CoreExpr] -> (Type, [CoreExpr]) +-- Apply function to the initial /type/ args; +-- return the type of the instantiated function, +-- and the remaining args +-- e.g. apply_type_args (++) [Type Int, Var xs] +-- = ([Int] -> [Int] -> [Int], [Var xs]) +apply_type_args fun args + = (piResultTys (idType fun) arg_tys, rest_args) + where + (arg_tys, rest_args) = go args + + go :: [CoreExpr] -> ([Type], [CoreExpr]) + go (Type ty : args) = first (ty :) (go args) + go args = ([], args) ------------------------------ splitHsWrapperArgs :: HsWrapper -> [CoreArg] -> DsM (HsWrapper, [CoreArg]) @@ -915,7 +954,9 @@ dsDo ctx stmts res_ty go _ (BodyStmt _ rhs then_expr _) stmts = do { rhs2 <- dsLExpr rhs - ; warnDiscardedDoBindings rhs (exprType rhs2) + ; case tcSplitAppTy_maybe (exprType rhs2) of + Just (m_ty, elt_ty) -> warnDiscardedDoBindings rhs m_ty elt_ty + Nothing -> return () -- Odd, but not warning ; rest <- goL stmts ; dsSyntaxExpr then_expr [rhs2, rest] } @@ -1067,9 +1108,8 @@ Other places that requires from the same treatment: -} -- Warn about certain types of values discarded in monadic bindings (#3263) -warnDiscardedDoBindings :: LHsExpr GhcTc -> Type -> DsM () -warnDiscardedDoBindings rhs rhs_ty - | Just (m_ty, elt_ty) <- tcSplitAppTy_maybe rhs_ty +warnDiscardedDoBindings :: LHsExpr GhcTc -> Type -> Type -> DsM () +warnDiscardedDoBindings rhs m_ty elt_ty = do { warn_unused <- woptM Opt_WarnUnusedDoBind ; warn_wrong <- woptM Opt_WarnWrongDoBind ; when (warn_unused || warn_wrong) $ @@ -1083,12 +1123,11 @@ warnDiscardedDoBindings rhs rhs_ty -- Warn about discarding m a things in 'monadic' binding of the same type, -- but only if we didn't already warn due to Opt_WarnUnusedDoBind + -- Example: do { return 3; blah } + -- We get (>>) @m d @(m Int) (return 3) blah when warn_wrong $ - case tcSplitAppTy_maybe norm_elt_ty of - Just (elt_m_ty, _) - | m_ty `eqType` topNormaliseType fam_inst_envs elt_m_ty - -> diagnosticDs (DsWrongDoBind rhs elt_ty) - _ -> return () } } - - | otherwise -- RHS does have type of form (m ty), which is weird - = return () -- but at least this warning is irrelevant + case tcSplitAppTy_maybe norm_elt_ty of + Just (elt_m_ty, _) + | m_ty `eqType` topNormaliseType fam_inst_envs elt_m_ty + -> diagnosticDs (DsWrongDoBind rhs elt_ty) + _ -> return () } } ===================================== compiler/GHC/HsToCore/Match/Literal.hs ===================================== @@ -18,7 +18,7 @@ module GHC.HsToCore.Match.Literal ( dsLit, dsOverLit, hsLitKey , tidyLitPat, tidyNPat , matchLiterals, matchNPlusKPats, matchNPats - , warnAboutIdentities + , numericConversionNames , warnAboutOverflowedOverLit, warnAboutOverflowedLit , warnAboutEmptyEnumerations ) @@ -275,23 +275,13 @@ between one type and another when the to- and from- types are the same. Then it's probably (albeit not definitely) the identity -} -warnAboutIdentities :: DynFlags -> Id -> Type -> DsM () -warnAboutIdentities dflags conv_fn type_of_conv - | wopt Opt_WarnIdentities dflags - , idName conv_fn `elem` conversionNames - , Just (_, _, arg_ty, res_ty) <- splitFunTy_maybe type_of_conv - , arg_ty `eqType` res_ty -- So we are converting ty -> ty - = diagnosticDs (DsIdentitiesFound conv_fn type_of_conv) -warnAboutIdentities _ _ _ = return () - -conversionNames :: [Name] -conversionNames +numericConversionNames :: [Name] +numericConversionNames = [ toIntegerName, toRationalName , fromIntegralName, realToFracName ] -- We can't easily add fromIntegerName, fromRationalName, -- because they are generated by literals - -- | Emit warnings on overloaded integral literals which overflow the bounds -- implied by their type. warnAboutOverflowedOverLit :: HsOverLit GhcTc -> DsM () ===================================== compiler/GHC/Rename/Env.hs ===================================== @@ -51,8 +51,7 @@ module GHC.Rename.Env ( lookupIfThenElse, -- QualifiedDo - lookupQualifiedDoExpr, lookupQualifiedDo, - lookupQualifiedDoName, lookupNameWithQualifier, + lookupQualifiedDo, lookupQualifiedDoName, lookupNameWithQualifier, -- Constructing usage information DeprecationWarnings(..), @@ -2350,14 +2349,14 @@ lookupSyntaxExpr :: Name -- ^ The standard name -> RnM (HsExpr GhcRn, FreeVars) -- ^ Possibly a non-standard name lookupSyntaxExpr std_name = do { (name, fvs) <- lookupSyntaxName std_name - ; return (nl_HsVar name, fvs) } + ; return (genHsVar name, fvs) } lookupSyntax :: Name -- The standard name -> RnM (SyntaxExpr GhcRn, FreeVars) -- Possibly a non-standard -- name lookupSyntax std_name - = do { (expr, fvs) <- lookupSyntaxExpr std_name - ; return (mkSyntaxExpr expr, fvs) } + = do { (name, fvs) <- lookupSyntaxName std_name + ; return (mkRnSyntaxExpr name, fvs) } lookupSyntaxNames :: [Name] -- Standard names -> RnM ([HsExpr GhcRn], FreeVars) -- See comments with HsExpr.ReboundNames @@ -2387,15 +2386,9 @@ by the Opt_QualifiedDo dynamic flag. -- Lookup operations for a qualified do. If the context is not a qualified -- do, then use lookupSyntaxExpr. See Note [QualifiedDo]. -lookupQualifiedDoExpr :: HsStmtContext fn -> Name -> RnM (HsExpr GhcRn, FreeVars) -lookupQualifiedDoExpr ctxt std_name - = first nl_HsVar <$> lookupQualifiedDoName ctxt std_name - --- Like lookupQualifiedDoExpr but for producing SyntaxExpr. --- See Note [QualifiedDo]. lookupQualifiedDo :: HsStmtContext fn -> Name -> RnM (SyntaxExpr GhcRn, FreeVars) lookupQualifiedDo ctxt std_name - = first mkSyntaxExpr <$> lookupQualifiedDoExpr ctxt std_name + = first mkRnSyntaxExpr <$> lookupQualifiedDoName ctxt std_name lookupNameWithQualifier :: Name -> ModuleName -> RnM (Name, FreeVars) lookupNameWithQualifier std_name modName @@ -2406,7 +2399,7 @@ lookupNameWithQualifier std_name modName lookupQualifiedDoName :: HsStmtContext fn -> Name -> RnM (Name, FreeVars) lookupQualifiedDoName ctxt std_name = case qualifiedDoModuleName_maybe ctxt of - Nothing -> lookupSyntaxName std_name + Nothing -> lookupSyntaxName std_name Just modName -> lookupNameWithQualifier std_name modName -------------------------------------------------------------------------------- ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -30,32 +30,27 @@ module GHC.Rename.Expr ( ) where import GHC.Prelude -import GHC.Data.FastString - -import GHC.Rename.Bind ( rnLocalBindsAndThen, rnLocalValBindsLHS, rnLocalValBindsRHS - , rnMatchGroup, rnGRHS, makeMiniFixityEnv) import GHC.Hs + import GHC.Tc.Errors.Types import GHC.Tc.Utils.Env ( isBrackStage ) import GHC.Tc.Utils.Monad -import GHC.Unit.Module ( getModule, isInteractiveModule ) + +import GHC.Rename.Bind ( rnLocalBindsAndThen, rnLocalValBindsLHS, rnLocalValBindsRHS + , rnMatchGroup, rnGRHS, makeMiniFixityEnv) import GHC.Rename.Env import GHC.Rename.Fixity -import GHC.Rename.Utils ( bindLocalNamesFV, checkDupNames - , bindLocalNames - , mapMaybeFvRn, mapFvRn - , warnUnusedLocalBinds, typeAppErr - , checkUnusedRecordWildcard - , wrapGenSpan, genHsIntegralLit, genHsTyLit - , genHsVar, genLHsVar, genHsApp, genHsApps, genHsApps' - , genAppType ) +import GHC.Rename.Utils import GHC.Rename.Unbound ( reportUnboundName ) -import GHC.Rename.Splice ( rnTypedBracket, rnUntypedBracket, rnTypedSplice, rnUntypedSpliceExpr, checkThLocalName ) +import GHC.Rename.Splice ( rnTypedBracket, rnUntypedBracket, rnTypedSplice + , rnUntypedSpliceExpr, checkThLocalName ) import GHC.Rename.HsType import GHC.Rename.Pat + import GHC.Driver.DynFlags import GHC.Builtin.Names import GHC.Builtin.Types ( nilDataConName ) +import GHC.Unit.Module ( getModule, isInteractiveModule ) import GHC.Types.Basic (TypeOrKind (TypeLevel)) import GHC.Types.FieldLabel @@ -67,13 +62,16 @@ import GHC.Types.Name.Reader import GHC.Types.Unique.Set import GHC.Types.SourceText import GHC.Types.SrcLoc + import GHC.Utils.Misc -import GHC.Data.List.SetOps ( removeDupsOn ) -import GHC.Data.Maybe import GHC.Utils.Error import GHC.Utils.Panic import GHC.Utils.Outputable as Outputable +import GHC.Data.FastString +import GHC.Data.List.SetOps ( removeDupsOn ) +import GHC.Data.Maybe + import qualified GHC.LanguageExtensions as LangExt import Language.Haskell.Syntax.Basic (FieldLabelString(..)) @@ -1415,7 +1413,7 @@ lookupQualifiedDoStmtName ctxt n = case qualifiedDoModuleName_maybe ctxt of Nothing -> lookupStmtName ctxt n Just modName -> - first (mkSyntaxExpr . nl_HsVar) <$> lookupNameWithQualifier n modName + first mkRnSyntaxExpr <$> lookupNameWithQualifier n modName lookupStmtName :: HsStmtContextRn -> Name -> RnM (SyntaxExpr GhcRn, FreeVars) -- Like lookupSyntax, but respects contexts @@ -2032,10 +2030,9 @@ rearrangeForApplicativeDo _ [] = return ([], emptyNameSet) rearrangeForApplicativeDo ctxt [(one,_)] = do (return_name, _) <- lookupQualifiedDoName (HsDoStmt ctxt) returnMName (pure_name, _) <- lookupQualifiedDoName (HsDoStmt ctxt) pureAName - let pure_expr = nl_HsVar pure_name let monad_names = MonadNames { return_name = return_name , pure_name = pure_name } - return $ case needJoin monad_names [one] (Just pure_expr) of + return $ case needJoin monad_names [one] (Just pure_name) of (False, one') -> (one', emptyNameSet) (True, _) -> ([one], emptyNameSet) rearrangeForApplicativeDo ctxt stmts0 = do @@ -2197,8 +2194,8 @@ stmtTreeToStmts monad_names ctxt (StmtTreeOne (L _ (BodyStmt _ rhs _ _),_)) }] False tail' stmtTreeToStmts monad_names ctxt (StmtTreeOne (let_stmt@(L _ LetStmt{}),_)) tail _tail_fvs = do - (pure_expr, _) <- lookupQualifiedDoExpr (HsDoStmt ctxt) pureAName - return $ case needJoin monad_names tail (Just pure_expr) of + (pure_name, _) <- lookupQualifiedDoName (HsDoStmt ctxt) pureAName + return $ case needJoin monad_names tail (Just pure_name) of (False, tail') -> (let_stmt : tail', emptyNameSet) (True, _) -> (let_stmt : tail, emptyNameSet) @@ -2256,8 +2253,8 @@ stmtTreeToStmts monad_names ctxt (StmtTreeApplicative trees) tail tail_fvs = do | otherwise -> do -- Need 'pureAName' and not 'returnMName' here, so that it requires -- 'Applicative' and not 'Monad' whenever possible (until #20540 is fixed). - (ret, _) <- lookupQualifiedDoExpr (HsDoStmt ctxt) pureAName - let expr = HsApp noExtField (noLocA ret) tup + (pure_name, _) <- lookupQualifiedDoName (HsDoStmt ctxt) pureAName + let expr = HsApp noExtField (noLocA (genHsVar pure_name)) tup return (expr, emptyFVs) return ( ApplicativeArgMany { xarg_app_arg_many = noExtField @@ -2523,7 +2520,7 @@ needJoin :: MonadNames -> [ExprLStmt GhcRn] -- If this is @Just pure@, replace return by pure -- If this is @Nothing@, strip the return/pure - -> Maybe (HsExpr GhcRn) + -> Maybe Name -> (Bool, [ExprLStmt GhcRn]) needJoin _monad_names [] _mb_pure = (False, []) -- we're in an ApplicativeArg needJoin monad_names [L loc (LastStmt _ e _ t)] mb_pure @@ -2544,18 +2541,18 @@ isReturnApp :: MonadNames -> LHsExpr GhcRn -- If this is @Just pure@, replace return by pure -- If this is @Nothing@, strip the return/pure - -> Maybe (HsExpr GhcRn) + -> Maybe Name -> Maybe (LHsExpr GhcRn, Maybe Bool) isReturnApp monad_names (L _ (HsPar _ expr)) mb_pure = isReturnApp monad_names expr mb_pure isReturnApp monad_names (L loc e) mb_pure = case e of OpApp x l op r - | Just pure_expr <- mb_pure, is_return l, is_dollar op -> - Just (L loc (OpApp x (to_pure l pure_expr) op r), Nothing) + | Just pure_name <- mb_pure, is_return l, is_dollar op -> + Just (L loc (OpApp x (to_pure l pure_name) op r), Nothing) | is_return l, is_dollar op -> Just (r, Just True) HsApp x f arg - | Just pure_expr <- mb_pure, is_return f -> - Just (L loc (HsApp x (to_pure f pure_expr) arg), Nothing) + | Just pure_name <- mb_pure, is_return f -> + Just (L loc (HsApp x (to_pure f pure_name) arg), Nothing) | is_return f -> Just (arg, Just False) _otherwise -> Nothing where @@ -2567,7 +2564,7 @@ isReturnApp monad_names (L loc e) mb_pure = case e of is_return = is_var (\n -> n == return_name monad_names || n == pure_name monad_names) - to_pure (L loc _) pure_expr = L loc pure_expr + to_pure (L loc _) pure_name = L loc (genHsVar pure_name) is_dollar = is_var (`hasKey` dollarIdKey) {- @@ -2793,13 +2790,13 @@ getMonadFailOp ctxt reallyGetMonadFailOp rebindableSyntax overloadedStrings | (isQualifiedDo || rebindableSyntax) && overloadedStrings = do - (failExpr, failFvs) <- lookupQualifiedDoExpr ctxt failMName + (failName, failFvs) <- lookupQualifiedDoName ctxt failMName (fromStringExpr, fromStringFvs) <- lookupSyntaxExpr fromStringName let arg_lit = mkVarOccFS (fsLit "arg") arg_name <- newSysName arg_lit let arg_syn_expr = nlHsVar arg_name body :: LHsExpr GhcRn = - nlHsApp (noLocA failExpr) + nlHsApp (noLocA (genHsVar failName)) (nlHsApp (noLocA $ fromStringExpr) arg_syn_expr) let failAfterFromStringExpr :: HsExpr GhcRn = unLoc $ mkHsLam (noLocA [noLocA $ VarPat noExtField $ noLocA arg_name]) body ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -1176,9 +1176,9 @@ rnOverLit origLit lit' = lit { ol_ext = OverLitRn { ol_rebindable = rebindable , ol_from_fun = L (noAnnSrcSpan loc) from_thing_name } } ; if isNegativeZeroOverLit lit' - then do { (negate_name, fvs2) <- lookupSyntaxExpr negateName - ; return ((lit' { ol_val = negateOverLitVal val }, Just negate_name) - , fvs1 `plusFV` fvs2) } + then do { (negate_expr, fvs2) <- lookupSyntaxExpr negateName + ; return ((lit' { ol_val = negateOverLitVal val }, Just negate_expr) + , fvs1 `plusFV` fvs2) } else return ((lit', Nothing), fvs1) } ===================================== compiler/GHC/Rename/Utils.hs ===================================== @@ -25,10 +25,9 @@ module GHC.Rename.Utils ( genLHsLit, genHsIntegralLit, genHsTyLit, genSimpleConPat, genVarPat, genWildPat, genSimpleFunBind, genFunBind, + genHsLamDoExp, genHsCaseAltDoExp, genSimpleMatch, genHsLet, - genHsLamDoExp, genHsCaseAltDoExp, genSimpleMatch, - - genHsLet, + mkRnSyntaxExpr, newLocalBndrRn, newLocalBndrsRn, @@ -713,6 +712,14 @@ wrapGenSpan :: (HasAnnotation an) => a -> GenLocated an a -- See Note [Rebindable syntax and XXExprGhcRn] wrapGenSpan x = L (noAnnSrcSpan generatedSrcSpan) x +-- | Make a 'SyntaxExpr' from a 'Name' (the "rn" is because this is used in the +-- renamer). +mkRnSyntaxExpr :: Name -> SyntaxExprRn +mkRnSyntaxExpr = SyntaxExprRn . genHsVar + +genHsVar :: Name -> HsExpr GhcRn +genHsVar n = HsVar noExtField (wrapGenSpan n) + genHsApps :: Name -> [LHsExpr GhcRn] -> HsExpr GhcRn genHsApps fun args = foldl genHsApp (genHsVar fun) args @@ -733,9 +740,6 @@ genLHsApp fun arg = wrapGenSpan (genHsApp fun arg) genLHsVar :: Name -> LHsExpr GhcRn genLHsVar nm = wrapGenSpan $ genHsVar nm -genHsVar :: Name -> HsExpr GhcRn -genHsVar nm = HsVar noExtField $ wrapGenSpan nm - genAppType :: HsExpr GhcRn -> HsType (NoGhcTc GhcRn) -> HsExpr GhcRn genAppType expr ty = HsAppType noExtField (wrapGenSpan expr) (mkEmptyWildCardBndrs (wrapGenSpan ty)) ===================================== compiler/GHC/Tc/Gen/Do.hs ===================================== @@ -23,7 +23,8 @@ import GHC.Prelude import GHC.Rename.Utils ( wrapGenSpan, genHsExpApps, genHsApp, genHsLet, genHsLamDoExp, genHsCaseAltDoExp, genWildPat ) -import GHC.Rename.Env ( irrefutableConLikeRn ) +import GHC.Rename.Env ( irrefutableConLikeRn ) + import GHC.Tc.Utils.Monad import GHC.Tc.Utils.TcMType ===================================== compiler/GHC/Tc/Gen/Expr.hs ===================================== @@ -568,7 +568,7 @@ tcExpr expr@(RecordCon { rcon_con = L loc con_name , rcon_flds = rbinds }) res_ty = do { con_like <- tcLookupConLike con_name - ; (con_expr, con_sigma) <- tcInferId con_name + ; (con_expr, con_sigma) <- tcInferConLike con_like ; (con_wrap, con_tau) <- topInstantiate orig con_sigma -- a shallow instantiation should really be enough for -- a data constructor. ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -23,7 +23,7 @@ module GHC.Tc.Gen.Head , leadingValArgs, isVisibleArg , tcInferAppHead, tcInferAppHead_maybe - , tcInferId, tcCheckId, obviousSig + , tcInferId, tcCheckId, tcInferConLike, obviousSig , tyConOf, tyConOfET, fieldNotInType , nonBidirectionalErr @@ -58,7 +58,7 @@ import GHC.Tc.Zonk.TcType import GHC.Core.FamInstEnv ( FamInstEnvs ) import GHC.Core.UsageEnv ( singleUsageUE ) -import GHC.Core.PatSyn( PatSyn ) +import GHC.Core.PatSyn( PatSyn, patSynName ) import GHC.Core.ConLike( ConLike(..) ) import GHC.Core.DataCon import GHC.Core.TyCon @@ -566,7 +566,7 @@ tcInferAppHead_maybe :: HsExpr GhcRn -- Returns Nothing for a complicated head tcInferAppHead_maybe fun = case fun of - HsVar _ (L _ nm) -> Just <$> tcInferId nm + HsVar _ nm -> Just <$> tcInferId nm HsRecSel _ f -> Just <$> tcInferRecSelId f ExprWithTySig _ e hs_ty -> Just <$> tcExprWithSig e hs_ty HsOverLit _ lit -> Just <$> tcInferOverLit lit @@ -793,7 +793,7 @@ tcInferOverLit lit@(OverLit { ol_val = val tcCheckId :: Name -> ExpRhoType -> TcM (HsExpr GhcTc) tcCheckId name res_ty - = do { (expr, actual_res_ty) <- tcInferId name + = do { (expr, actual_res_ty) <- tcInferId (noLocA name) ; traceTc "tcCheckId" (vcat [ppr name, ppr actual_res_ty, ppr res_ty]) ; addFunResCtxt expr [] actual_res_ty res_ty $ tcWrapResultO (OccurrenceOf name) rn_fun expr actual_res_ty res_ty } @@ -801,33 +801,33 @@ tcCheckId name res_ty rn_fun = HsVar noExtField (noLocA name) ------------------------ -tcInferId :: Name -> TcM (HsExpr GhcTc, TcSigmaType) +tcInferId :: LocatedN Name -> TcM (HsExpr GhcTc, TcSigmaType) -- Look up an occurrence of an Id -- Do not instantiate its type -tcInferId id_name +tcInferId lname@(L _ id_name) | id_name `hasKey` assertIdKey = do { dflags <- getDynFlags ; if gopt Opt_IgnoreAsserts dflags - then tc_infer_id id_name - else tc_infer_assert id_name } + then tc_infer_id lname + else tc_infer_assert lname } | otherwise - = do { (expr, ty) <- tc_infer_id id_name + = do { (expr, ty) <- tc_infer_id lname ; traceTc "tcInferId" (ppr id_name <+> dcolon <+> ppr ty) ; return (expr, ty) } -tc_infer_assert :: Name -> TcM (HsExpr GhcTc, TcSigmaType) +tc_infer_assert :: LocatedN Name -> TcM (HsExpr GhcTc, TcSigmaType) -- Deal with an occurrence of 'assert' -- See Note [Adding the implicit parameter to 'assert'] -tc_infer_assert assert_name +tc_infer_assert (L loc assert_name) = do { assert_error_id <- tcLookupId assertErrorName ; (wrap, id_rho) <- topInstantiate (OccurrenceOf assert_name) (idType assert_error_id) - ; return (mkHsWrap wrap (HsVar noExtField (noLocA assert_error_id)), id_rho) + ; return (mkHsWrap wrap (HsVar noExtField (L loc assert_error_id)), id_rho) } -tc_infer_id :: Name -> TcM (HsExpr GhcTc, TcSigmaType) -tc_infer_id id_name +tc_infer_id :: LocatedN Name -> TcM (HsExpr GhcTc, TcSigmaType) +tc_infer_id (L loc id_name) = do { thing <- tcLookup id_name ; case thing of ATcId { tct_id = id } @@ -839,14 +839,14 @@ tc_infer_id id_name -- nor does it need the 'lifting' treatment -- Hence no checkTh stuff here - AGlobal (AConLike (RealDataCon con)) -> tcInferDataCon con - AGlobal (AConLike (PatSynCon ps)) -> tcInferPatSyn id_name ps + AGlobal (AConLike cl) -> tcInferConLike cl + (tcTyThingTyCon_maybe -> Just tc) -> failIllegalTyCon WL_Anything (tyConName tc) ATyVar name _ -> failIllegalTyVal name _ -> failWithTc $ TcRnExpectedValueId thing } where - return_id id = return (HsVar noExtField (noLocA id), idType id) + return_id id = return (HsVar noExtField (L loc id), idType id) {- Note [Suppress hints with RequiredTypeArguments] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -903,6 +903,10 @@ check_naughty lbl id | isNaughtyRecordSelector id = failWithTc (TcRnRecSelectorEscapedTyVar lbl) | otherwise = return () +tcInferConLike :: ConLike -> TcM (HsExpr GhcTc, TcSigmaType) +tcInferConLike (RealDataCon con) = tcInferDataCon con +tcInferConLike (PatSynCon ps) = tcInferPatSyn ps + tcInferDataCon :: DataCon -> TcM (HsExpr GhcTc, TcSigmaType) -- See Note [Typechecking data constructors] tcInferDataCon con @@ -934,11 +938,11 @@ tcInferDataCon con ; return (Scaled mul_var ty) } linear_to_poly scaled_ty = return scaled_ty -tcInferPatSyn :: Name -> PatSyn -> TcM (HsExpr GhcTc, TcSigmaType) -tcInferPatSyn id_name ps +tcInferPatSyn :: PatSyn -> TcM (HsExpr GhcTc, TcSigmaType) +tcInferPatSyn ps = case patSynBuilderOcc ps of Just (expr,ty) -> return (expr,ty) - Nothing -> failWithTc (nonBidirectionalErr id_name) + Nothing -> failWithTc (nonBidirectionalErr (patSynName ps)) nonBidirectionalErr :: Name -> TcRnMessage nonBidirectionalErr = TcRnPatSynNotBidirectional ===================================== compiler/GHC/Tc/Module.hs ===================================== @@ -91,7 +91,7 @@ import GHC.Rename.Names import GHC.Rename.Env import GHC.Rename.Module import GHC.Rename.Doc -import GHC.Rename.Utils ( mkNameClashErr ) +import GHC.Rename.Utils ( mkNameClashErr, mkRnSyntaxExpr ) import GHC.Iface.Decl ( coAxiomToIfaceDecl ) import GHC.Iface.Env ( externaliseName ) ===================================== compiler/GHC/Tc/Utils/Instantiate.hs ===================================== @@ -71,6 +71,8 @@ import GHC.Tc.Utils.TcType import GHC.Tc.Errors.Types import GHC.Tc.Zonk.Monad ( ZonkM ) +import GHC.Rename.Utils( mkRnSyntaxExpr ) + import GHC.Types.Id.Make( mkDictFunId ) import GHC.Types.Basic ( TypeOrKind(..), Arity, VisArity ) import GHC.Types.Error ===================================== testsuite/tests/linear/should_fail/LinearTHFail2.stderr ===================================== @@ -1,3 +1,3 @@ -LinearTHFail2.hs:7:2: error: [GHC-65904] +LinearTHFail2.hs:7:3: error: [GHC-65904] Non-linear fields in data constructors not (yet) handled by Template Haskell ===================================== testsuite/tests/linear/should_fail/LinearTHFail3.stderr ===================================== @@ -1,3 +1,3 @@ -LinearTHFail3.hs:7:2: error: [GHC-65904] +LinearTHFail3.hs:7:3: error: [GHC-65904] Non-linear fields in data constructors not (yet) handled by Template Haskell View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e2c288248a4be2a427b08659248742f79eb41612 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e2c288248a4be2a427b08659248742f79eb41612 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 11:17:53 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Fri, 27 Sep 2024 07:17:53 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/9.13-start Message-ID: <66f6946197810_3f1cfabc5707877@gitlab.mail> Zubin pushed new branch wip/9.13-start at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/9.13-start You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 11:18:34 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Fri, 27 Sep 2024 07:18:34 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] 35 commits: ci: Run abi-test on test-abi label Message-ID: <66f6948ad70c8_3f1cfa21d3d88016@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - aa6d7374 by Matthew Pickering at 2024-09-27T11:09:01+01:00 Fix toException method for ExceptionWithContext Fixes #25235 - - - - - 55a3c7aa by Matthew Pickering at 2024-09-27T12:15:42+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. Implements CLC proposal#202 <https://github.com/haskell/core-libraries-committee/issues/202> - - - - - b23f019c by Rodrigo Mesquita at 2024-09-27T12:18:04+01:00 Add test for #25300 This reverts commit f781b76635390b9f8c04362638193dc1b9c2541a. - - - - - 27 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/Pipeline.hs - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reducibility.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Switch.hs - compiler/GHC/Cmm/Switch/Implement.hs - compiler/GHC/Cmm/ThreadSanitizer.hs - + compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/BlockLayout.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/723e4471c356de06271f48bfb777db141a7ae675...b23f019cf18224f937eacf3cde3e8deece4ec8eb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/723e4471c356de06271f48bfb777db141a7ae675...b23f019cf18224f937eacf3cde3e8deece4ec8eb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 11:25:26 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 07:25:26 -0400 Subject: [Git][ghc/ghc][wip/llvm_19] 51 commits: ci: Run abi-test on test-abi label Message-ID: <66f696266ffbf_3f1cfa50587014865@gitlab.mail> Matthew Pickering pushed to branch wip/llvm_19 at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 2a26423d by Matthew Pickering at 2024-09-27T11:25:21+00:00 Bump LLVM upper bound to allow LLVM 19 Also bumps the ci-images commit so that the deb12 images uses LLVM 19 for testing. ------------------------- Metric Decrease: size_hello_artifact_gzip size_hello_unicode_gzip ------------------------- Fixes #25295 - - - - - 21 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/381458e8b2eae8f53f09d991b70d043a02c4dc66...2a26423d5decd302b8d722933c53ca3b5507b28f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/381458e8b2eae8f53f09d991b70d043a02c4dc66...2a26423d5decd302b8d722933c53ca3b5507b28f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 11:41:11 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Fri, 27 Sep 2024 07:41:11 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/romes/25304 Message-ID: <66f699d74606e_3f1cfa4a3c4c215e0@gitlab.mail> Rodrigo Mesquita pushed new branch wip/romes/25304 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/romes/25304 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 12:24:32 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 08:24:32 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/simd-system-cxx Message-ID: <66f6a400b86f7_3f64061e639c147b0@gitlab.mail> Matthew Pickering pushed new branch wip/simd-system-cxx at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/simd-system-cxx You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 12:36:24 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 08:36:24 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 2 commits: hadrian: Make text_simdutf flavour transformer configurable per-stage Message-ID: <66f6a6c8237de_3f640632e27c205a4@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: cbb14af8 by Matthew Pickering at 2024-09-27T12:16:37+01:00 hadrian: Make text_simdutf flavour transformer configurable per-stage Before it was globally enabled, which was probably not what you want as you don't need text-simd for your boot compiler nor your boot compiler if you're building a cross-compiler. This brings it into line with the other modifiers.. such as ghcProfiled etc Fixes #25302 - - - - - dc59add2 by Matthew Pickering at 2024-09-27T12:20:53+01:00 hadrian: Refactor system-cxx-std-lib rules0 I noticed a few things wrong with the hadrian rules for `system-cxx-std-lib` rules. * For `text` there is an ad-hoc check to depend on `system-cxx-std-lib` outside of `configurePackage`. * The `system-cxx-std-lib` dependency is not read from cabal files. * Recache is not called on the packge database after the `.conf` file is generated, a more natural place for this rule is `registerRules`. Treating this uniformly like other packages is complicated by it not having any source code or a cabal file. However we can do a bit better by reporting the dependency firstly in `PackageData` and then needing the `.conf` file in the same place as every other package in `configurePackage`. Fixes #25303 - - - - - 8 changed files: - hadrian/src/Flavour.hs - hadrian/src/Flavour/Type.hs - hadrian/src/Hadrian/Haskell/Cabal/Parse.hs - hadrian/src/Hadrian/Haskell/Cabal/Type.hs - hadrian/src/Rules/Generate.hs - hadrian/src/Rules/Register.hs - hadrian/src/Settings/Default.hs - hadrian/src/Settings/Packages.hs Changes: ===================================== hadrian/src/Flavour.hs ===================================== @@ -300,7 +300,7 @@ useNativeBignum flavour = -- | Enable building the @text@ package with @simdutf@ support. enableTextWithSIMDUTF :: Flavour -> Flavour enableTextWithSIMDUTF flavour = flavour { - textWithSIMDUTF = True + textWithSIMDUTF = (>= Stage2) } -- | Build stage2 compiler with -fomit-interface-pragmas to reduce ===================================== hadrian/src/Flavour/Type.hs ===================================== @@ -25,7 +25,8 @@ data Flavour = Flavour { bignumCheck :: Bool, -- | Build the @text@ package with @simdutf@ support. Disabled by -- default due to packaging difficulties described in #20724. - textWithSIMDUTF :: Bool, + textWithSIMDUTF :: Stage -- ^ stage of the /built/ compiler + -> Bool, -- | Build libraries these ways. libraryWays :: Ways, -- | Build RTS these ways. ===================================== hadrian/src/Hadrian/Haskell/Cabal/Parse.hs ===================================== @@ -94,10 +94,11 @@ parsePackageData pkg = do sorted = sort [ C.unPackageName p | C.Dependency p _ _ <- allDeps ] deps = nubOrd sorted \\ [name] depPkgs = mapMaybe findPackageByName deps + cxxStdLib = elem "system-cxx-std-lib" deps return $ PackageData name version (C.fromShortText (C.synopsis pd)) (C.fromShortText (C.description pd)) - depPkgs gpd + depPkgs cxxStdLib gpd where -- Collect an overapproximation of dependencies by ignoring conditionals collectDeps :: Maybe (C.CondTree v [C.Dependency] a) -> [C.Dependency] @@ -151,7 +152,9 @@ configurePackage :: Context -> Action () configurePackage context at Context {..} = do putProgressInfo $ "| Configure package " ++ quote (pkgName package) gpd <- pkgGenericDescription package - depPkgs <- packageDependencies <$> readPackageData package + pd <- readPackageData package + let depPkgs = packageDependencies pd + needSystemCxxStdLib = dependsOnSystemCxxStdLib pd -- Stage packages are those we have in this stage. stagePkgs <- stagePackages stage @@ -170,7 +173,12 @@ configurePackage context at Context {..} = do -- We'll need those packages in our package database. deps <- sequence [ pkgConfFile (context { package = pkg, iplace = forceBaseAfterGhcInternal pkg }) | pkg <- depPkgs, pkg `elem` stagePkgs ] - need $ extraPreConfigureDeps ++ deps + -- system-cxx-std-lib is magic.. it doesn't have a cabal file or source code, so we have + -- to treat it specially as `pkgConfFile` uses `readPackageData` to compute the version. + systemCxxStdLib <- sequence [ systemCxxStdLibConfPath (PackageDbLoc stage iplace) | needSystemCxxStdLib ] + need $ extraPreConfigureDeps + ++ deps + ++ systemCxxStdLib -- Figure out what hooks we need. let configureFile = replaceFileName (pkgCabalFile package) "configure" ===================================== hadrian/src/Hadrian/Haskell/Cabal/Type.hs ===================================== @@ -30,6 +30,7 @@ data PackageData = PackageData , synopsis :: String , description :: String , packageDependencies :: [Package] + , dependsOnSystemCxxStdLib :: Bool , genericPackageDescription :: GenericPackageDescription } deriving (Eq, Generic, Show, Typeable) ===================================== hadrian/src/Rules/Generate.hs ===================================== @@ -221,9 +221,6 @@ copyRules = do prefix -/- "html/**" <~ return "utils/haddock/haddock-api/resources" prefix -/- "latex/**" <~ return "utils/haddock/haddock-api/resources" - forM_ [Inplace, Final] $ \iplace -> - root -/- relativePackageDbPath (PackageDbLoc stage iplace) -/- systemCxxStdLibConf %> \file -> do - copyFile ("mk" -/- "system-cxx-std-lib-1.0.conf") file generateRules :: Rules () generateRules = do ===================================== hadrian/src/Rules/Register.hs ===================================== @@ -52,14 +52,6 @@ configurePackageRules = do isGmp <- (== "gmp") <$> interpretInContext ctx getBignumBackend when isGmp $ need [buildP -/- "include/ghc-gmp.h"] - when (pkg == text) $ do - simdutf <- textWithSIMDUTF <$> flavour - when simdutf $ do - -- This is required, otherwise you get Error: hadrian: - -- Encountered missing or private dependencies: - -- system-cxx-std-lib ==1.0 - cxxStdLib <- systemCxxStdLibConfPath $ PackageDbLoc stage Inplace - need [cxxStdLib] Cabal.configurePackage ctx root -/- "**/autogen/cabal_macros.h" %> \out -> do @@ -114,6 +106,12 @@ registerPackageRules rs stage iplace = do target (Context stage compiler vanilla iplace) (GhcPkg Recache stage) [] [] writeFileLines stamp [] + -- Special rule for registering system-cxx-std-lib + root -/- relativePackageDbPath (PackageDbLoc stage iplace) -/- systemCxxStdLibConf %> \file -> do + copyFile ("mk" -/- "system-cxx-std-lib-1.0.conf") file + buildWithResources rs $ + target (Context stage compiler vanilla iplace) (GhcPkg Recache stage) [] [] + -- Register a package. root -/- relativePackageDbPath (PackageDbLoc stage iplace) -/- "*.conf" %> \conf -> do historyDisable ===================================== hadrian/src/Settings/Default.hs ===================================== @@ -275,7 +275,7 @@ defaultFlavour = Flavour , packages = defaultPackages , bignumBackend = defaultBignumBackend , bignumCheck = False - , textWithSIMDUTF = False + , textWithSIMDUTF = const False , libraryWays = defaultLibraryWays , rtsWays = defaultRtsWays , dynamicGhcPrograms = defaultDynamicGhcPrograms ===================================== hadrian/src/Settings/Packages.hs ===================================== @@ -192,7 +192,7 @@ packageArgs = do ---------------------------------- text -------------------------------- , package text ? - ifM (textWithSIMDUTF <$> expr flavour) + ifM (compilerStageOption textWithSIMDUTF) (builder (Cabal Flags) ? arg "+simdutf") (builder (Cabal Flags) ? arg "-simdutf") View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c2f63c368b6400753b80bef1da4a0e587757f2fe...dc59add291ee428398c323ce6a5a59078849b911 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/c2f63c368b6400753b80bef1da4a0e587757f2fe...dc59add291ee428398c323ce6a5a59078849b911 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 13:03:07 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Fri, 27 Sep 2024 09:03:07 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/torsten.schmits/persistent-worker-test-2024-09-27 Message-ID: <66f6ad0b2281f_3f6406580fc0253cc@gitlab.mail> Torsten Schmits pushed new branch wip/torsten.schmits/persistent-worker-test-2024-09-27 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/torsten.schmits/persistent-worker-test-2024-09-27 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 13:42:34 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Fri, 27 Sep 2024 09:42:34 -0400 Subject: [Git][ghc/ghc][wip/9.13-start] Bump GHC version to 9.13 Message-ID: <66f6b64a13bed_3f64066e12702771c@gitlab.mail> Zubin pushed to branch wip/9.13-start at Glasgow Haskell Compiler / GHC Commits: e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - 5 changed files: - configure.ac - − docs/users_guide/9.12.1-notes.rst - + docs/users_guide/9.14.1-notes.rst - utils/haddock/haddock-api/haddock-api.cabal - utils/haddock/haddock-api/src/Haddock/InterfaceFile.hs Changes: ===================================== configure.ac ===================================== @@ -13,7 +13,7 @@ dnl # see what flags are available. (Better yet, read the documentation!) # -AC_INIT([The Glorious Glasgow Haskell Compilation System], [9.12], [glasgow-haskell-bugs at haskell.org], [ghc-AC_PACKAGE_VERSION]) +AC_INIT([The Glorious Glasgow Haskell Compilation System], [9.13], [glasgow-haskell-bugs at haskell.org], [ghc-AC_PACKAGE_VERSION]) # Version on master must be X.Y (not X.Y.Z) for ProjectVersionMunged variable # to be useful (cf #19058). However, the version must have three components # (X.Y.Z) on stable branches (e.g. ghc-9.2) to ensure that pre-releases are @@ -22,7 +22,7 @@ AC_INIT([The Glorious Glasgow Haskell Compilation System], [9.12], [glasgow-hask AC_CONFIG_MACRO_DIRS([m4]) # Set this to YES for a released version, otherwise NO -: ${RELEASE=YES} +: ${RELEASE=NO} # The primary version (e.g. 7.5, 7.4.1) is set in the AC_INIT line # above. If this is not a released version, then we will append the ===================================== docs/users_guide/9.12.1-notes.rst deleted ===================================== @@ -1,283 +0,0 @@ -.. _release-9-11-1: - -Version 9.12.1 -============== - -The significant changes to the various parts of the compiler are listed in the -following sections. See the `migration guide -`_ on the GHC Wiki -for specific guidance on migrating programs to this release. - -Language -~~~~~~~~ - -- New language extension: :extension:`OrPatterns` implements `GHC Proposal #522 - `_). -- GHC Proposal `#569 `_ - "Multiline string literals" has been implemented. - The following code is now accepted by GHC:: - - {-# LANGUAGE MultilineStrings #-} - - x :: String - x = - """ - This is a - multiline - - string - - literal - """ - - This feature is guarded behind :extension:`MultilineStrings`. - -- The ordering of variables used for visible type application has been changed in two cases. - It is supposed to be left-to-right, but due to an oversight, it was wrong: - - - in an infix application ``f :: a `op` b``, it is now ``forall a op b.`` rather than - ``forall op a b.`` - - in a linear type ``f :: a %m -> b``, it is now ``forall a m b.`` rather than - ``forall a b m.``. - - This change is backwards-incompatible, although in practice we don't expect it - to cause significant disruption. - -- The built-in ``HasField`` class, used by :extension:`OverloadedRecordDot`, now - supports representation polymorphism (implementing part of `GHC Proposal #583 - `_). - This means that code using :extension:`UnliftedDatatypes` or - :extension:`UnliftedNewtypes` can now use :extension:`OverloadedRecordDot`. - -- Unboxed ``Float#``/``Double#`` literals now support the HexFloatLiterals extension - (`#22155 `_). - -- :extension:`UnliftedFFITypes`: GHC will now accept FFI types like: ``(# #) -> T`` where ``(# #)`` - is used as the one and only function argument. - -- The venerable ``default`` declarations have been generalized. They can now name a class - other than ``Num`` and the class defaults can be exported. The functionality is guarded - by the new ``NamedDefaults`` language extension. See the `GHC proposal - `__ - for details. - -- GHC now takes COMPLETE pragmas into consideration when deciding whether - pattern matches in do notation are fallible. - -- As part of `GHC Proposal #281 `_ - GHC now accepts type syntax in expressions, namely function type arrow ``a -> b``, - constraint arrow ``a => b``, and ``forall`` telescopes: :: - - g = f (forall a. Show a => a -> String) - where f :: forall a -> ... - -- In accordance with `GHC Proposal #425 `_, - GHC now permits wildcard binders in type declarations: :: - - type Const a b = a -- before: the `b` had to be named even if unused on the RHS - type Const a _ = a -- now: the compiler accepts a wildcard binder `_` - -Compiler -~~~~~~~~ - -- Constructor ``PluginProv`` of type ``UnivCoProvenance``, relevant - for typing plugins, gets an extra ``DCoVarSet`` argument. - The argument is intended to contain the in-scope coercion variables - that the the proof represented by the coercion makes use of. - See ``Note [The importance of tracking UnivCo dependencies]`` - in ``GHC.Core.TyCo.Rep``, :ref:`constraint-solving-with-plugins` - and the migration guide. - -- The flag :ghc-flag:`-fprof-late` will no longer prevent top level constructors from being statically allocated. - - It used to be the case that we would add a cost centre for bindings like ``foo = Just bar``. - This turned the binding into a CAF that would allocate the constructor on first evaluation. - - However without the cost centre ``foo`` can be allocated at compile time. This reduces code-bloat and - reduces overhead for short-running applications. - - The tradeoff is that calling ``whoCreated`` on top level value definitions like ``foo`` will be less informative. - -- A new flag :ghc-flag:`-fexpose-overloaded-unfoldings` has been added providing a lightweight alternative to :ghc-flag:`-fexpose-all-unfoldings`. - -- :ghc-flag:`-Wderiving-typeable` has been added to :ghc-flag:`-Wall`. - -- SIMD support has been added to the X86 native code generator. - For the time being, only 128 bit wide vectors are supported, with most - floating-point operations implemented, together with a few integer vector - operations. Other operations still require the LLVM backend. Contributors - welcome! - -- i386 Windows support is now completely removed amid massive cleanup - of legacy code to pave way for Arm64 Windows support (`#24883 - `_). Rest - assured, this does not impact existing support for x86_64 Windows or - i386 Linux. For end users, the ``stdcall`` C calling convention is - now fully deprecated and GHC will unconditionally produce a warning - and treat it as ``ccall``. All C import/export declarations on - Windows should now use ``ccall``. - -- 32-bit macOS/iOS support has also been completely removed (`#24921 - `_). This does - not affect existing support of apple systems on x86_64/aarch64. - -- The flag :ghc-flag:`-fignore-asserts` will now also enable the - :extension:`CPP` macro ``__GLASGOW_HASKELL_ASSERTS_IGNORED__`` (`#24967 - `_). - This enables people to write their own custom assertion functions. - See :ref:`assertions`. - -- The flag :ghc-flag:`-fkeep-auto-rules` that forces GHC to keep auto generated - specialization rules was added. It was actually added ghc-9.10.1 already but - mistakenly not mentioned in the 9.10.1 changelog. - -- Fixed a bug that caused GHC to panic when using the aarch64 ncg and -fregs-graph - on certain programs. (#24941) - -- A new warning :ghc-flag:`-Wview-pattern-signatures` is introduced to notify users about - future changes in parsing of view patterns in combination with pattern signatures - -- GHC now includes a new experimental flag, :ghc-flag:`-fobject-determinism`, - which enables deterministic object code generation, with a minor trade-off in - compilation speed (typically a 1-2% regression). Note, however, that this - flag doesn't yet ensure determinism across all compilation configurations; we - anticipate these cases will be addressed in future updates (:ghc-ticket:`12935`). - -GHCi -~~~~ - -- Fix a bug where GHCi would not start alongside a local file called ``Prelude.hs`` - or ``Prelude.lhs`` (:ghc-ticket:`10920`). - - -Runtime system -~~~~~~~~~~~~~~ - -- Reduce fragmentation incurred by the nonmoving GC's segment allocator. In one application this reduced resident set size by 26%. See :ghc-ticket:`24150`. - -- Memory return logic now uses live bytes rather than live blocks to measure the size of the heap. - This primarily affects the non-moving GC, which should now be more willing to return memory to the OS. - Users who have fine-tuned the :rts-flag:`-F ⟨factor⟩`, :rts-flag:`-Fd ⟨factor⟩`, or :rts-flag:`-O ⟨size⟩` flags, - and use the non-moving GC, should see if adjustments are needed in light of this change. - -- The new runtime flag :rts-flag:`--read-tix-file=\` allows to modify whether a preexisting .tix file is read in at the beginning of a program run. - The default is currently ``--read-tix-file=yes`` but will change to ``--read-tix-file=no`` in a future version of GHC. - For this reason, a warning is emitted if a .tix file is read in implicitly. You can silence this warning by explicitly passing ``--read-tix-file=yes``. - Details can be found in `GHC proposal 612 `__. - -Cmm -~~~ - -- The ``bits128``, ``bits256`` and ``bits512`` types have been removed, in - favour of ``vec128``, ``vec256`` and ``vec512``. - -- The ``[*]`` jump annotation ("all registers live") has been removed, in favour - of more specific annotations ``GP_ARG_REGS`` (all general-purpose registers - live), ``SCALAR_ARG_REGS`` (all scalar registers live), and ``V16_ARG_REGS``, - ``V32_ARG_REGS`` and ``V64_ARG_REGS`` (meaning: all scalar registers plus - all vector registers up to the given vector width in bytes). - -``base`` library -~~~~~~~~~~~~~~~~ - -- Propagate HasCallStack from `errorCallWithCallStackException` to exception backtraces, fixing a bug in the implementation of `CLC proposal 164 `. - -- Add exception type metadata to SomeException's displayException and - "Exception:" header to the default handler - (i.e. ``GHC.Conc.Sync.uncaughtExceptionHandler``): - - https://github.com/haskell/core-libraries-committee/issues/231 - https://github.com/haskell/core-libraries-committee/issues/261 - -- The `deprecation process of GHC.Pack ` has come its term. The module has now been removed from ``base``. - -``ghc-prim`` library -~~~~~~~~~~~~~~~~~~~~ - -- Usage of deprecated primops is now correctly reported (#19629). -- New primops `isMutableByteArrayWeaklyPinned#` and `isByteArrayWeaklyPinned#` - to allow users to avoid copying large arrays safely when dealing with ffi. - See the users guide for more details on the different kinds of - pinned arrays in 9.12. - - This need for this distinction originally surfaced in https://gitlab.haskell.org/ghc/ghc/-/issues/22255 - -- New fused multiply-add instructions for vectors of floating-point values, - such as ``fmaddFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# -> FloatX4#`` and - ``fnmsubDoubleX2# :: DoubleX2# -> DoubleX2# -> DoubleX2# -> DoubleX2#``. - These follow the same semantics as ``fmadd``/``fmsub``/``fnmadd``/``fnmsub``, - operating in parallel on vectors of floating-point values. - -- New vector shuffle instructions, such as ``shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4#``. - These instructions take two input vectors and a collection of indices (which must - be compile-time literal integers), and constructs a result vector by extracting - out the values at those indices. For instance, ``shuffleFloatX4#`` on input vectors with - components ``(# 0.1#, 11.1#, 22.1#, 33.1# #)`` and ``(# 44.1#, 55.1#, 66.1#, 77.1# #)``, - and indices ``(# 4#, 3#, 6#, 1# #)``, will return a vector with components - ``(# 44.1#, 33.1#, 66.1#, 11.1# #)``. - -- New instructions for minimum/maximum, such as `minDouble#` and - `minFloatX4#`. These instructions compute the minimum/maximum of their inputs, - working component-wise for SIMD vectors. Supported argument types are vector - integer values (e.g. `Word16X8#`, `Int32X4#` etc) and both scalar and vector - floating point values (e.g. `Float#`, `DoubleX2#`, `FloatX8#` etc). - -``ghc`` library -~~~~~~~~~~~~~~~ - -``ghc-heap`` library -~~~~~~~~~~~~~~~~~~~~ - -``ghc-experimental`` library -~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - -``template-haskell`` library -~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - -Included libraries -~~~~~~~~~~~~~~~~~~ - -The package database provided with this distribution also contains a number of -packages other than GHC itself. See the changelogs provided with these packages -for further change information. - -.. ghc-package-list:: - - libraries/array/array.cabal: Dependency of ``ghc`` library - libraries/base/base.cabal: Core library - libraries/binary/binary.cabal: Dependency of ``ghc`` library - libraries/bytestring/bytestring.cabal: Dependency of ``ghc`` library - libraries/Cabal/Cabal/Cabal.cabal: Dependency of ``ghc-pkg`` utility - libraries/Cabal/Cabal-syntax/Cabal-syntax.cabal: Dependency of ``ghc-pkg`` utility - libraries/containers/containers/containers.cabal: Dependency of ``ghc`` library - libraries/deepseq/deepseq.cabal: Dependency of ``ghc`` library - libraries/directory/directory.cabal: Dependency of ``ghc`` library - libraries/exceptions/exceptions.cabal: Dependency of ``ghc`` and ``haskeline`` library - libraries/filepath/filepath.cabal: Dependency of ``ghc`` library - compiler/ghc.cabal: The compiler itself - libraries/ghci/ghci.cabal: The REPL interface - libraries/ghc-boot/ghc-boot.cabal: Internal compiler library - libraries/ghc-boot-th/ghc-boot-th.cabal: Internal compiler library - libraries/ghc-compact/ghc-compact.cabal: Core library - libraries/ghc-heap/ghc-heap.cabal: GHC heap-walking library - libraries/ghc-prim/ghc-prim.cabal: Core library - utils/haddock/haddock-api/haddock-api.cabal: Dependency of ``haddock`` executable - utils/haddock/haddock-library/haddock-library.cabal: Dependency of ``haddock`` executable - libraries/haskeline/haskeline.cabal: Dependency of ``ghci`` executable - libraries/hpc/hpc.cabal: Dependency of ``hpc`` executable - libraries/integer-gmp/integer-gmp.cabal: Core library - libraries/mtl/mtl.cabal: Dependency of ``Cabal`` library - libraries/parsec/parsec.cabal: Dependency of ``Cabal`` library - libraries/pretty/pretty.cabal: Dependency of ``ghc`` library - libraries/process/process.cabal: Dependency of ``ghc`` library - libraries/stm/stm.cabal: Dependency of ``haskeline`` library - libraries/template-haskell/template-haskell.cabal: Core library - libraries/terminfo/terminfo.cabal: Dependency of ``haskeline`` library - libraries/text/text.cabal: Dependency of ``Cabal`` library - libraries/time/time.cabal: Dependency of ``ghc`` library - libraries/transformers/transformers.cabal: Dependency of ``ghc`` library - libraries/unix/unix.cabal: Dependency of ``ghc`` library - libraries/Win32/Win32.cabal: Dependency of ``ghc`` library - libraries/xhtml/xhtml.cabal: Dependency of ``haddock`` executable - libraries/os-string/os-string.cabal: Dependency of ``filepath`` library - libraries/file-io/file-io.cabal: Dependency of ``directory`` library ===================================== docs/users_guide/9.14.1-notes.rst ===================================== @@ -0,0 +1,90 @@ +.. _release-9-14-1: + +Version 9.14.1 +============== + +The significant changes to the various parts of the compiler are listed in the +following sections. See the `migration guide +`_ on the GHC Wiki +for specific guidance on migrating programs to this release. + +Language +~~~~~~~~ + +Compiler +~~~~~~~~ + +GHCi +~~~~ + +Runtime system +~~~~~~~~~~~~~~ + +Cmm +~~~ + +``base`` library +~~~~~~~~~~~~~~~~ + +``ghc-prim`` library +~~~~~~~~~~~~~~~~~~~~ + +``ghc`` library +~~~~~~~~~~~~~~~ + +``ghc-heap`` library +~~~~~~~~~~~~~~~~~~~~ + +``ghc-experimental`` library +~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +``template-haskell`` library +~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Included libraries +~~~~~~~~~~~~~~~~~~ + +The package database provided with this distribution also contains a number of +packages other than GHC itself. See the changelogs provided with these packages +for further change information. + +.. ghc-package-list:: + + libraries/array/array.cabal: Dependency of ``ghc`` library + libraries/base/base.cabal: Core library + libraries/binary/binary.cabal: Dependency of ``ghc`` library + libraries/bytestring/bytestring.cabal: Dependency of ``ghc`` library + libraries/Cabal/Cabal/Cabal.cabal: Dependency of ``ghc-pkg`` utility + libraries/Cabal/Cabal-syntax/Cabal-syntax.cabal: Dependency of ``ghc-pkg`` utility + libraries/containers/containers/containers.cabal: Dependency of ``ghc`` library + libraries/deepseq/deepseq.cabal: Dependency of ``ghc`` library + libraries/directory/directory.cabal: Dependency of ``ghc`` library + libraries/exceptions/exceptions.cabal: Dependency of ``ghc`` and ``haskeline`` library + libraries/filepath/filepath.cabal: Dependency of ``ghc`` library + compiler/ghc.cabal: The compiler itself + libraries/ghci/ghci.cabal: The REPL interface + libraries/ghc-boot/ghc-boot.cabal: Internal compiler library + libraries/ghc-boot-th/ghc-boot-th.cabal: Internal compiler library + libraries/ghc-compact/ghc-compact.cabal: Core library + libraries/ghc-heap/ghc-heap.cabal: GHC heap-walking library + libraries/ghc-prim/ghc-prim.cabal: Core library + utils/haddock/haddock-api/haddock-api.cabal: Dependency of ``haddock`` executable + utils/haddock/haddock-library/haddock-library.cabal: Dependency of ``haddock`` executable + libraries/haskeline/haskeline.cabal: Dependency of ``ghci`` executable + libraries/hpc/hpc.cabal: Dependency of ``hpc`` executable + libraries/integer-gmp/integer-gmp.cabal: Core library + libraries/mtl/mtl.cabal: Dependency of ``Cabal`` library + libraries/parsec/parsec.cabal: Dependency of ``Cabal`` library + libraries/pretty/pretty.cabal: Dependency of ``ghc`` library + libraries/process/process.cabal: Dependency of ``ghc`` library + libraries/stm/stm.cabal: Dependency of ``haskeline`` library + libraries/template-haskell/template-haskell.cabal: Core library + libraries/terminfo/terminfo.cabal: Dependency of ``haskeline`` library + libraries/text/text.cabal: Dependency of ``Cabal`` library + libraries/time/time.cabal: Dependency of ``ghc`` library + libraries/transformers/transformers.cabal: Dependency of ``ghc`` library + libraries/unix/unix.cabal: Dependency of ``ghc`` library + libraries/Win32/Win32.cabal: Dependency of ``ghc`` library + libraries/xhtml/xhtml.cabal: Dependency of ``haddock`` executable + libraries/os-string/os-string.cabal: Dependency of ``filepath`` library + libraries/file-io/file-io.cabal: Dependency of ``directory`` library ===================================== utils/haddock/haddock-api/haddock-api.cabal ===================================== @@ -79,7 +79,7 @@ library -- this package typically supports only single major versions build-depends: base >= 4.16 && < 4.21 - , ghc ^>= 9.12 + , ghc ^>= 9.13 , haddock-library ^>= 1.11 , xhtml ^>= 3000.2.2 , parsec ^>= 3.1.13.0 ===================================== utils/haddock/haddock-api/src/Haddock/InterfaceFile.hs ===================================== @@ -139,7 +139,7 @@ binaryInterfaceMagic = 0xD0Cface -- (2) set `binaryInterfaceVersionCompatibility` to [binaryInterfaceVersion] -- binaryInterfaceVersion :: Word16 -#if MIN_VERSION_ghc(9,9,0) && !MIN_VERSION_ghc(9,12,0) +#if MIN_VERSION_ghc(9,11,0) && !MIN_VERSION_ghc(9,14,0) binaryInterfaceVersion = 44 binaryInterfaceVersionCompatibility :: [Word16] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e4ac1b0d281b85a0144d1ef6f84a1df00e236052 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e4ac1b0d281b85a0144d1ef6f84a1df00e236052 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 13:55:06 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Fri, 27 Sep 2024 09:55:06 -0400 Subject: [Git][ghc/ghc][wip/andreask/spec-constr-args] SpecConstr: Introduce a separate argument limit for forced specs. Message-ID: <66f6b93a22e46_3f64066ede3028491@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/spec-constr-args at Glasgow Haskell Compiler / GHC Commits: 62d1cabf by Andreas Klebinger at 2024-09-27T15:35:47+02:00 SpecConstr: Introduce a separate argument limit for forced specs. We used to put no limit at all on specializations forced via the SPEC argument. This isn't always reasonable so we introduce a very high limit that applies to forced specializations, a flag to control it, and we now emit a warning if we fail a specialization because we exceed the warning. Fixes #25197 - - - - - 8 changed files: - compiler/GHC/Core/Opt/SpecConstr.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Session.hs - docs/users_guide/using-optimisation.rst - + testsuite/tests/simplCore/should_compile/T25197.hs - + testsuite/tests/simplCore/should_compile/T25197.stderr - + testsuite/tests/simplCore/should_compile/T25197_TH.hs - testsuite/tests/simplCore/should_compile/all.T Changes: ===================================== compiler/GHC/Core/Opt/SpecConstr.hs ===================================== @@ -20,7 +20,8 @@ ToDo [Oct 2013] module GHC.Core.Opt.SpecConstr( specConstrProgram, - SpecConstrAnnotation(..) + SpecConstrAnnotation(..), + SpecFailWarning(..) ) where import GHC.Prelude @@ -51,6 +52,7 @@ import GHC.Core.Make ( mkImpossibleExpr ) import GHC.Unit.Module import GHC.Unit.Module.ModGuts +import GHC.Types.Error (MessageClass(..), Severity(..), DiagnosticReason(WarningWithoutFlag), ResolvedDiagnosticReason (..)) import GHC.Types.Literal ( litIsLifted ) import GHC.Types.Id import GHC.Types.Id.Info ( IdDetails(..) ) @@ -526,9 +528,11 @@ sc_force to True when calling specLoop. This flag does four things: (see argToPat; #4448) (FS4) Only specialise on recursive types a finite number of times (see sc_recursive; #5550; Note [Limit recursive specialisation]) -(FS5) Lift the restriction on the maximum number of arguments which - the optimisation will specialise. - (see `too_many_worker_args` in `callsToNewPats`; #14003) +(FS5) Use a different restriction on the maximum number of arguments which + the optimisation will specialise. We tried removing the limit on worker + args for forced specs (#14003) but this caused issues when specializing + code for large data structures (#25197). + This is handled by `too_many_worker_args` in `callsToNewPats` The flag holds only for specialising a single binding group, and NOT for nested bindings. (So really it should be passed around explicitly @@ -782,16 +786,25 @@ specConstrProgram :: ModGuts -> CoreM ModGuts specConstrProgram guts = do { env0 <- initScEnv guts ; us <- getUniqueSupplyM - ; let (_usg, binds') = initUs_ us $ + ; let (_usg, binds', warnings) = initUs_ us $ scTopBinds env0 (mg_binds guts) + ; when (not (null warnings)) $ msg specConstr_warn_class (warn_msg warnings) + ; return (guts { mg_binds = binds' }) } -scTopBinds :: ScEnv -> [InBind] -> UniqSM (ScUsage, [OutBind]) -scTopBinds _env [] = return (nullUsage, []) -scTopBinds env (b:bs) = do { (usg, b', bs') <- scBind TopLevel env b $ + where + specConstr_warn_class = MCDiagnostic SevWarning (ResolvedDiagnosticReason WarningWithoutFlag) Nothing + warn_msg :: SpecFailWarnings -> SDoc + warn_msg warnings = text "SpecConstr encountered one or more function(s) with a SPEC argument that resulted in too many arguments," $$ + text "which resulted in no specialization being generated for these functions:" $$ + nest 2 (vcat (map ppr warnings)) $$ + (text "If this is expected you might want to increase -fmax-forced-spec-args to force specialization anyway.") +scTopBinds :: ScEnv -> [InBind] -> UniqSM (ScUsage, [OutBind], [SpecFailWarning]) +scTopBinds _env [] = return (nullUsage, [], []) +scTopBinds env (b:bs) = do { (usg, b', bs', warnings) <- scBind TopLevel env b $ (\env -> scTopBinds env bs) - ; return (usg, b' ++ bs') } + ; return (usg, b' ++ bs', warnings) } {- ************************************************************************ @@ -905,6 +918,12 @@ data SpecConstrOpts = SpecConstrOpts -- ^ The threshold at which a worker-wrapper transformation used as part of -- this pass will no longer happen, measured in the number of arguments. + , sc_max_forced_args :: !Int + -- ^ The threshold at which a worker-wrapper transformation used as part of + -- this pass will no longer happen even if a SPEC arg was used to force + -- specialization. Measured in the number of arguments. + -- See Note [Forcing specialisation] + , sc_debug :: !Bool -- ^ Whether to print debug information @@ -975,6 +994,7 @@ instance Outputable Value where initScOpts :: DynFlags -> Module -> SpecConstrOpts initScOpts dflags this_mod = SpecConstrOpts { sc_max_args = maxWorkerArgs dflags, + sc_max_forced_args = maxForcedSpecArgs dflags, sc_debug = hasPprDebug dflags, sc_uf_opts = unfoldingOpts dflags, sc_module = this_mod, @@ -1388,29 +1408,29 @@ creates specialised versions of functions. -} scBind :: TopLevelFlag -> ScEnv -> InBind - -> (ScEnv -> UniqSM (ScUsage, a)) -- Specialise the scope of the binding - -> UniqSM (ScUsage, [OutBind], a) + -> (ScEnv -> UniqSM (ScUsage, a, [SpecFailWarning])) -- Specialise the scope of the binding + -> UniqSM (ScUsage, [OutBind], a, [SpecFailWarning]) scBind top_lvl env (NonRec bndr rhs) do_body | isTyVar bndr -- Type-lets may be created by doBeta - = do { (final_usage, body') <- do_body (extendScSubst env bndr rhs) - ; return (final_usage, [], body') } + = do { (final_usage, body', warnings) <- do_body (extendScSubst env bndr rhs) + ; return (final_usage, [], body', warnings) } | not (isTopLevel top_lvl) -- Nested non-recursive value binding -- See Note [Specialising local let bindings] = do { let (body_env, bndr') = extendBndr env bndr -- Not necessary at top level; but here we are nested - ; rhs_info <- scRecRhs env (bndr',rhs) + ; (rhs_info, rhs_ws) <- scRecRhs env (bndr',rhs) ; let body_env2 = extendHowBound body_env [bndr'] RecFun rhs' = ri_new_rhs rhs_info body_env3 = extendValEnv body_env2 bndr' (isValue (sc_vals env) rhs') - ; (body_usg, body') <- do_body body_env3 + ; (body_usg, body', warnings_body) <- do_body body_env3 -- Now make specialised copies of the binding, -- based on calls in body_usg - ; (spec_usg, specs) <- specNonRec env (scu_calls body_usg) rhs_info + ; (spec_usg, specs, warnings_bnd) <- specNonRec env (scu_calls body_usg) rhs_info -- NB: For non-recursive bindings we inherit sc_force flag from -- the parent function (see Note [Forcing specialisation]) @@ -1419,7 +1439,7 @@ scBind top_lvl env (NonRec bndr rhs) do_body bind_usage = (body_usg `delCallsFor` [bndr']) `combineUsage` spec_usg -- Note [spec_usg includes rhs_usg] - ; return (bind_usage, spec_bnds, body') + ; return (bind_usage, spec_bnds, body', mconcat [warnings_bnd, warnings_body, rhs_ws]) } | otherwise -- Top-level, non-recursive value binding @@ -1431,15 +1451,15 @@ scBind top_lvl env (NonRec bndr rhs) do_body -- -- I tried always specialising non-recursive top-level bindings too, -- but found some regressions (see !8135). So I backed off. - = do { (rhs_usage, rhs') <- scExpr env rhs + = do { (rhs_usage, rhs', ws_rhs) <- scExpr env rhs -- At top level, we've already put all binders into scope; see initScEnv -- Hence no need to call `extendBndr`. But we still want to -- extend the `ValueEnv` to record the value of this binder. ; let body_env = extendValEnv env bndr (isValue (sc_vals env) rhs') - ; (body_usage, body') <- do_body body_env + ; (body_usage, body', body_warnings) <- do_body body_env - ; return (rhs_usage `combineUsage` body_usage, [NonRec bndr rhs'], body') } + ; return (rhs_usage `combineUsage` body_usage, [NonRec bndr rhs'], body', body_warnings ++ ws_rhs) } scBind top_lvl env (Rec prs) do_body | isTopLevel top_lvl @@ -1450,19 +1470,20 @@ scBind top_lvl env (Rec prs) do_body -- ToDo: I'm honestly not sure of the rationale of this size-testing, nor -- why it only applies at top level. But that's the way it has been -- for a while. See #21456. - do { (body_usg, body') <- do_body rhs_env2 - ; (rhs_usgs, rhss') <- mapAndUnzipM (scExpr env) rhss + do { (body_usg, body', warnings_body) <- do_body rhs_env2 + ; (rhs_usgs, rhss', rhs_ws) <- mapAndUnzip3M (scExpr env) rhss ; let all_usg = (combineUsages rhs_usgs `combineUsage` body_usg) `delCallsFor` bndrs' bind' = Rec (bndrs' `zip` rhss') - ; return (all_usg, [bind'], body') } + ; return (all_usg, [bind'], body', warnings_body ++ concat rhs_ws) } | otherwise - = do { rhs_infos <- mapM (scRecRhs rhs_env2) (bndrs' `zip` rhss) - ; (body_usg, body') <- do_body rhs_env2 + = do { (rhs_infos, rhs_wss) <- mapAndUnzipM (scRecRhs rhs_env2) (bndrs' `zip` rhss) + ; let rhs_ws = mconcat rhs_wss + ; (body_usg, body', warnings_body) <- do_body rhs_env2 - ; (spec_usg, specs) <- specRec (scForce rhs_env2 force_spec) - (scu_calls body_usg) rhs_infos + ; (spec_usg, specs, spec_ws) <- specRec (scForce rhs_env2 force_spec) + (scu_calls body_usg) rhs_infos -- Do not unconditionally generate specialisations from rhs_usgs -- Instead use them only if we find an unspecialised call -- See Note [Seeding recursive groups] @@ -1473,7 +1494,7 @@ scBind top_lvl env (Rec prs) do_body -- zipWithEqual: length of returned [SpecInfo] -- should be the same as incoming [RhsInfo] - ; return (all_usg, [bind'], body') } + ; return (all_usg, [bind'], body', mconcat [warnings_body,rhs_ws,spec_ws]) } where (bndrs,rhss) = unzip prs force_spec = any (forceSpecBndr env) bndrs -- Note [Forcing specialisation] @@ -1501,59 +1522,63 @@ recursive function, but that's not essential and might even be harmful. I'm not sure. -} +withWarnings :: SpecFailWarnings -> (ScUsage, CoreExpr, SpecFailWarnings) -> (ScUsage, CoreExpr, SpecFailWarnings) +withWarnings ws (use,expr,ws2) = (use,expr,ws ++ ws2) + ------------------------ -scExpr, scExpr' :: ScEnv -> CoreExpr -> UniqSM (ScUsage, CoreExpr) +scExpr, scExpr' :: ScEnv -> CoreExpr -> UniqSM (ScUsage, CoreExpr, SpecFailWarnings) -- The unique supply is needed when we invent -- a new name for the specialised function and its args scExpr env e = scExpr' env e scExpr' env (Var v) = case scSubstId env v of - Var v' -> return (mkVarUsage env v' [], Var v') + Var v' -> return (mkVarUsage env v' [], Var v', []) e' -> scExpr (zapScSubst env) e' scExpr' env (Type t) = let !(MkSolo ty') = scSubstTy env t - in return (nullUsage, Type ty') -scExpr' env (Coercion c) = return (nullUsage, Coercion (scSubstCo env c)) -scExpr' _ e@(Lit {}) = return (nullUsage, e) -scExpr' env (Tick t e) = do (usg, e') <- scExpr env e - return (usg, Tick (scTickish env t) e') -scExpr' env (Cast e co) = do (usg, e') <- scExpr env e - return (usg, mkCast e' (scSubstCo env co)) + in return (nullUsage, Type ty', []) +scExpr' env (Coercion c) = return (nullUsage, Coercion (scSubstCo env c), []) +scExpr' _ e@(Lit {}) = return (nullUsage, e, []) +scExpr' env (Tick t e) = do (usg, e', ws) <- scExpr env e + return (usg, Tick (scTickish env t) e', ws) +scExpr' env (Cast e co) = do (usg, e', ws) <- scExpr env e + return (usg, mkCast e' (scSubstCo env co), ws) -- Important to use mkCast here -- See Note [SpecConstr call patterns] scExpr' env e@(App _ _) = scApp env (collectArgs e) scExpr' env (Lam b e) = do let (env', b') = extendBndr env b - (usg, e') <- scExpr env' e - return (usg, Lam b' e') + (usg, e', ws) <- scExpr env' e + return (usg, Lam b' e', ws) scExpr' env (Let bind body) - = do { (final_usage, binds', body') <- scBind NotTopLevel env bind $ + = do { (final_usage, binds', body', ws) <- scBind NotTopLevel env bind $ (\env -> scExpr env body) - ; return (final_usage, mkLets binds' body') } + ; return (final_usage, mkLets binds' body', ws) } scExpr' env (Case scrut b ty alts) - = do { (scrut_usg, scrut') <- scExpr env scrut + = do { (scrut_usg, scrut', ws) <- scExpr env scrut ; case isValue (sc_vals env) scrut' of Just (ConVal args_are_work_free con args) - | args_are_work_free -> sc_con_app con args scrut' + | args_are_work_free -> sc_con_app con args scrut' ws -- Don't duplicate work!! #7865 -- See Note [ConVal work-free-ness] (1) - _other -> sc_vanilla scrut_usg scrut' + _other -> sc_vanilla scrut_usg scrut' ws } where - sc_con_app con args scrut' -- Known constructor; simplify + sc_con_app con args scrut' ws -- Known constructor; simplify = do { let Alt _ bs rhs = findAlt con alts `orElse` Alt DEFAULT [] (mkImpossibleExpr ty "SpecConstr") alt_env' = extendScSubstList env ((b,scrut') : bs `zip` trimConArgs con args) - ; scExpr alt_env' rhs } + ; (use',expr',ws_new) <- scExpr alt_env' rhs + ; return (use',expr',ws ++ ws_new) } - sc_vanilla scrut_usg scrut' -- Normal case + sc_vanilla scrut_usg scrut' ws -- Normal case = do { let (alt_env,b') = extendBndrWith RecArg env b -- Record RecArg for the components - ; (alt_usgs, alt_occs, alts') <- mapAndUnzip3M (sc_alt alt_env scrut' b') alts + ; (alt_usgs, alt_occs, alts', ws_alts) <- mapAndUnzip4M (sc_alt alt_env scrut' b') alts ; let scrut_occ = foldr combineOcc NoOcc alt_occs scrut_usg' = setScrutOcc env scrut_usg scrut' scrut_occ @@ -1563,21 +1588,21 @@ scExpr' env (Case scrut b ty alts) ; let !(MkSolo ty') = scSubstTy env ty ; return (foldr combineUsage scrut_usg' alt_usgs, - Case scrut' b' ty' alts') } + Case scrut' b' ty' alts', ws ++ concat ws_alts) } single_alt = isSingleton alts sc_alt env scrut' b' (Alt con bs rhs) = do { let (env1, bs1) = extendBndrsWith RecArg env bs (env2, bs2) = extendCaseBndrs env1 scrut' b' con bs1 - ; (usg, rhs') <- scExpr env2 rhs + ; (usg, rhs', ws) <- scExpr env2 rhs ; let (usg', b_occ:arg_occs) = lookupOccs usg (b':bs2) scrut_occ = case con of DataAlt dc -- See Note [Do not specialise evals] | not (single_alt && all deadArgOcc arg_occs) -> ScrutOcc (unitUFM dc arg_occs) _ -> UnkOcc - ; return (usg', b_occ `combineOcc` scrut_occ, Alt con bs2 rhs') } + ; return (usg', b_occ `combineOcc` scrut_occ, Alt con bs2 rhs', ws) } -- | Substitute the free variables captured by a breakpoint. @@ -1626,19 +1651,20 @@ follows. still worth specialising on x. Hence the /single-alternative/ guard. -} -scApp :: ScEnv -> (InExpr, [InExpr]) -> UniqSM (ScUsage, CoreExpr) +scApp :: ScEnv -> (InExpr, [InExpr]) -> UniqSM (ScUsage, CoreExpr, SpecFailWarnings) scApp env (Var fn, args) -- Function is a variable = assert (not (null args)) $ do { args_w_usgs <- mapM (scExpr env) args - ; let (arg_usgs, args') = unzip args_w_usgs + ; let (arg_usgs, args', arg_ws) = unzip3 args_w_usgs arg_usg = combineUsages arg_usgs + arg_w = concat arg_ws ; case scSubstId env fn of - fn'@(Lam {}) -> scExpr (zapScSubst env) (doBeta fn' args') + fn'@(Lam {}) -> withWarnings arg_w <$> scExpr (zapScSubst env) (doBeta fn' args') -- Do beta-reduction and try again Var fn' -> return (arg_usg' `combineUsage` mkVarUsage env fn' args', - mkApps (Var fn') args') + mkApps (Var fn') args', arg_w ) where -- arg_usg': see Note [Specialising on dictionaries] arg_usg' | Just cls <- isClassOpId_maybe fn' @@ -1647,7 +1673,7 @@ scApp env (Var fn, args) -- Function is a variable | otherwise = arg_usg - other_fn' -> return (arg_usg, mkApps other_fn' args') } + other_fn' -> return (arg_usg, mkApps other_fn' args', arg_w) } -- NB: doing this ignores any usage info from the substituted -- function, but I don't think that matters. If it does -- we can fix it. @@ -1661,9 +1687,9 @@ scApp env (Var fn, args) -- Function is a variable -- which it may, we can get -- (let f = ...f... in f) arg1 arg2 scApp env (other_fn, args) - = do { (fn_usg, fn') <- scExpr env other_fn - ; (arg_usgs, args') <- mapAndUnzipM (scExpr env) args - ; return (combineUsages arg_usgs `combineUsage` fn_usg, mkApps fn' args') } + = do { (fn_usg, fn', fn_ws) <- scExpr env other_fn + ; (arg_usgs, args', arg_ws) <- mapAndUnzip3M (scExpr env) args + ; return (combineUsages arg_usgs `combineUsage` fn_usg, mkApps fn' args', combineSpecWarning fn_ws (concat arg_ws)) } ---------------------- mkVarUsage :: ScEnv -> Id -> [CoreExpr] -> ScUsage @@ -1679,16 +1705,16 @@ mkVarUsage env fn args | otherwise = evalScrutOcc ---------------------- -scRecRhs :: ScEnv -> (OutId, InExpr) -> UniqSM RhsInfo +scRecRhs :: ScEnv -> (OutId, InExpr) -> UniqSM (RhsInfo, SpecFailWarnings) scRecRhs env (bndr,rhs) = do { let (arg_bndrs,body) = collectBinders rhs (body_env, arg_bndrs') = extendBndrsWith RecArg env arg_bndrs - ; (body_usg, body') <- scExpr body_env body + ; (body_usg, body', body_ws) <- scExpr body_env body ; let (rhs_usg, arg_occs) = lookupOccs body_usg arg_bndrs' ; return (RI { ri_rhs_usg = rhs_usg , ri_fn = bndr, ri_new_rhs = mkLams arg_bndrs' body' , ri_lam_bndrs = arg_bndrs, ri_lam_body = body - , ri_arg_occs = arg_occs }) } + , ri_arg_occs = arg_occs }, body_ws) } -- The arg_occs says how the visible, -- lambda-bound binders of the RHS are used -- (including the TyVar binders) @@ -1757,7 +1783,7 @@ initSpecInfo (RI { ri_rhs_usg = rhs_usg }) specNonRec :: ScEnv -> CallEnv -- Calls in body -> RhsInfo -- Structure info usage info for un-specialised RHS - -> UniqSM (ScUsage, SpecInfo) -- Usage from RHSs (specialised and not) + -> UniqSM (ScUsage, SpecInfo, [SpecFailWarning]) -- Usage from RHSs (specialised and not) -- plus details of specialisations specNonRec env body_calls rhs_info @@ -1767,11 +1793,12 @@ specNonRec env body_calls rhs_info specRec :: ScEnv -> CallEnv -- Calls in body -> [RhsInfo] -- Structure info and usage info for un-specialised RHSs - -> UniqSM (ScUsage, [SpecInfo]) -- Usage from all RHSs (specialised and not) + -> UniqSM (ScUsage, [SpecInfo], SpecFailWarnings) + -- Usage from all RHSs (specialised and not) -- plus details of specialisations specRec env body_calls rhs_infos - = go 1 body_calls nullUsage (map initSpecInfo rhs_infos) + = go 1 body_calls nullUsage (map initSpecInfo rhs_infos) [] -- body_calls: see Note [Seeding recursive groups] -- NB: 'go' always calls 'specialise' once, which in turn unleashes -- si_mb_unspec if there are any boring calls in body_calls, @@ -1786,23 +1813,25 @@ specRec env body_calls rhs_infos -- Two accumulating parameters: -> ScUsage -- Usage from earlier specialisations -> [SpecInfo] -- Details of specialisations so far - -> UniqSM (ScUsage, [SpecInfo]) - go n_iter seed_calls usg_so_far spec_infos + -> SpecFailWarnings -- Warnings so far + -> UniqSM (ScUsage, [SpecInfo], SpecFailWarnings) + go n_iter seed_calls usg_so_far spec_infos ws_so_far = -- pprTrace "specRec3" (vcat [ text "bndrs" <+> ppr (map ri_fn rhs_infos) -- , text "iteration" <+> int n_iter -- , text "spec_infos" <+> ppr (map (map os_pat . si_specs) spec_infos) -- ]) $ do { specs_w_usg <- zipWithM (specialise env seed_calls) rhs_infos spec_infos - ; let (extra_usg_s, all_spec_infos) = unzip specs_w_usg + + ; let (extra_usg_s, all_spec_infos, extra_ws ) = unzip3 specs_w_usg extra_usg = combineUsages extra_usg_s all_usg = usg_so_far `combineUsage` extra_usg new_calls = scu_calls extra_usg - ; go_again n_iter new_calls all_usg all_spec_infos } + ; go_again n_iter new_calls all_usg all_spec_infos (ws_so_far ++ concat extra_ws) } -- go_again deals with termination - go_again n_iter seed_calls usg_so_far spec_infos + go_again n_iter seed_calls usg_so_far spec_infos ws_so_far | isEmptyVarEnv seed_calls - = return (usg_so_far, spec_infos) + = return (usg_so_far, spec_infos, ws_so_far) -- Limit recursive specialisation -- See Note [Limit recursive specialisation] @@ -1816,10 +1845,10 @@ specRec env body_calls rhs_infos -- for the unspecialised function, since it may now be called -- pprTrace "specRec2" (ppr (map (map os_pat . si_specs) spec_infos)) $ let rhs_usgs = combineUsages (mapMaybe si_mb_unspec spec_infos) - in return (usg_so_far `combineUsage` rhs_usgs, spec_infos) + in return (usg_so_far `combineUsage` rhs_usgs, spec_infos, ws_so_far) | otherwise - = go (n_iter + 1) seed_calls usg_so_far spec_infos + = go (n_iter + 1) seed_calls usg_so_far spec_infos ws_so_far -- See Note [Limit recursive specialisation] the_limit = case sc_count opts of @@ -1832,7 +1861,7 @@ specialise -> CallEnv -- Info on newly-discovered calls to this function -> RhsInfo -> SpecInfo -- Original RHS plus patterns dealt with - -> UniqSM (ScUsage, SpecInfo) -- New specialised versions and their usage + -> UniqSM (ScUsage, SpecInfo, [SpecFailWarning]) -- New specialised versions and their usage -- See Note [spec_usg includes rhs_usg] @@ -1850,7 +1879,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs | isDeadEndId fn -- Note [Do not specialise diverging functions] -- /and/ do not generate specialisation seeds from its RHS = -- pprTrace "specialise bot" (ppr fn) $ - return (nullUsage, spec_info) + return (nullUsage, spec_info, []) | not (isNeverActive (idInlineActivation fn)) -- See Note [Transfer activation] @@ -1861,7 +1890,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs , not (null arg_bndrs) -- Only specialise functions , Just all_calls <- lookupVarEnv bind_calls fn -- Some calls to it = -- pprTrace "specialise entry {" (ppr fn <+> ppr all_calls) $ - do { (boring_call, pats_discarded, new_pats) + do { (boring_call, pats_discarded, new_pats, warnings) <- callsToNewPats env fn spec_info arg_occs all_calls ; let n_pats = length new_pats @@ -1876,7 +1905,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs -- , text "new_pats" <+> ppr new_pats]) ; let spec_env = decreaseSpecCount env n_pats - ; (spec_usgs, new_specs) <- mapAndUnzipM (spec_one spec_env fn arg_bndrs body) + ; (spec_usgs, new_specs, new_wss) <- mapAndUnzip3M (spec_one spec_env fn arg_bndrs body) (new_pats `zip` [spec_count..]) -- See Note [Specialise original body] @@ -1900,15 +1929,16 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs ; return (new_usg, SI { si_specs = new_specs ++ specs , si_n_specs = spec_count + n_pats - , si_mb_unspec = mb_unspec' }) } + , si_mb_unspec = mb_unspec' } + ,warnings ++ concat new_wss) } | otherwise -- No calls, inactive, or not a function -- Behave as if there was a single, boring call = -- pprTrace "specialise inactive" (ppr fn $$ ppr mb_unspec) $ case mb_unspec of -- Behave as if there was a single, boring call - Just rhs_usg -> return (rhs_usg, spec_info { si_mb_unspec = Nothing }) + Just rhs_usg -> return (rhs_usg, spec_info { si_mb_unspec = Nothing }, []) -- See Note [spec_usg includes rhs_usg] - Nothing -> return (nullUsage, spec_info) + Nothing -> return (nullUsage, spec_info, []) --------------------- @@ -1917,7 +1947,7 @@ spec_one :: ScEnv -> [InVar] -- Lambda-binders of RHS; should match patterns -> InExpr -- Body of the original function -> (CallPat, Int) - -> UniqSM (ScUsage, OneSpec) -- Rule and binding + -> UniqSM (ScUsage, OneSpec, SpecFailWarnings) -- Rule and binding, warnings if any -- spec_one creates a specialised copy of the function, together -- with a rule for using it. I'm very proud of how short this @@ -1969,7 +1999,7 @@ spec_one env fn arg_bndrs body (call_pat, rule_number) -- Specialise the body -- ; pprTraceM "body_subst_for" $ ppr (spec_occ) $$ ppr (sc_subst body_env) - ; (spec_usg, spec_body) <- scExpr body_env body + ; (spec_usg, spec_body, body_warnings) <- scExpr body_env body -- And build the results ; (qvars', pats') <- generaliseDictPats qvars pats @@ -2018,7 +2048,7 @@ spec_one env fn arg_bndrs body (call_pat, rule_number) -- ] ; return (spec_usg, OS { os_pat = call_pat, os_rule = rule , os_id = spec_id - , os_rhs = spec_rhs }) } + , os_rhs = spec_rhs }, body_warnings) } generaliseDictPats :: [Var] -> [CoreExpr] -- Quantified vars and pats -> UniqSM ([Var], [CoreExpr]) -- New quantified vars and pats @@ -2402,12 +2432,26 @@ instance Outputable CallPat where , text "cp_args =" <+> ppr args , text "cp_strict_args = " <> ppr strict ]) +newtype SpecFailWarning = SpecFailForcedArgCount { spec_failed_fun_name :: Name } + +type SpecFailWarnings = [SpecFailWarning] + +instance Outputable SpecFailWarning where + ppr (SpecFailForcedArgCount name) = ppr name <+> pprDefinedAt name + +combineSpecWarning :: SpecFailWarnings -> SpecFailWarnings -> SpecFailWarnings +combineSpecWarning = (++) + +data ArgCountResult = WorkerSmallEnough | WorkerTooLarge | WorkerTooLargeForced Name + callsToNewPats :: ScEnv -> Id -> SpecInfo -> [ArgOcc] -> [Call] -> UniqSM ( Bool -- At least one boring call , Bool -- Patterns were discarded - , [CallPat] ) -- Patterns to specialise + , [CallPat] -- Patterns to specialise + , [SpecFailWarning] -- Things that didn't specialise we want to warn the user about) + ) -- Result has no duplicate patterns, -- nor ones mentioned in si_specs (hence "new" patterns) -- Bool indicates that there was at least one boring pattern @@ -2433,12 +2477,18 @@ callsToNewPats env fn spec_info@(SI { si_specs = done_specs }) bndr_occs calls non_dups = subsumePats in_scope new_pats -- Remove ones that have too many worker variables - small_pats = filterOut too_many_worker_args non_dups + (small_pats, arg_count_warnings) = partitionByWorkerSize too_many_worker_args non_dups - too_many_worker_args _ - | sc_force env = False -- See (FS5) of Note [Forcing specialisation] + -- too_many_worker_args :: CallPat -> Either SpecFailWarning Bool too_many_worker_args (CP { cp_qvars = vars, cp_args = args }) - = not (isWorkerSmallEnough (sc_max_args $ sc_opts env) (valArgCount args) vars) + | sc_force env + -- See (FS5) of Note [Forcing specialisation] + = if (isWorkerSmallEnough (sc_max_forced_args $ sc_opts env) (valArgCount args) vars) + then WorkerSmallEnough + else WorkerTooLargeForced (idName fn) + | (isWorkerSmallEnough (sc_max_args $ sc_opts env) (valArgCount args) vars) + = WorkerSmallEnough + | otherwise = WorkerTooLarge -- We are about to construct w/w pair in 'spec_one'. -- Omit specialisation leading to high arity workers. -- See Note [Limit w/w arity] in GHC.Core.Opt.WorkWrap.Utils @@ -2454,10 +2504,21 @@ callsToNewPats env fn spec_info@(SI { si_specs = done_specs }) bndr_occs calls -- , text "done_specs:" <+> ppr (map os_pat done_specs) -- , text "trimmed_pats:" <+> ppr trimmed_pats ]) - ; return (have_boring_call, pats_were_discarded, trimmed_pats) } + ; return (have_boring_call, pats_were_discarded, trimmed_pats, arg_count_warnings) } -- If any of the calls does not give rise to a specialisation, either -- because it is boring, or because there are too many specialisations, -- return a flag to say so, so that we know to keep the original function. + where + partitionByWorkerSize worker_size pats = go pats [] [] + where + go [] small warnings = (small, warnings) + go (p:ps) small warnings + | WorkerSmallEnough <- worker_size p + = go ps (p:small) warnings + | WorkerTooLarge <- worker_size p + = go ps small warnings + | WorkerTooLargeForced name <- worker_size p + = go ps small (SpecFailForcedArgCount name : warnings) trim_pats :: ScEnv -> Id -> SpecInfo -> [CallPat] -> (Bool, [CallPat]) ===================================== compiler/GHC/Driver/DynFlags.hs ===================================== @@ -394,6 +394,7 @@ data DynFlags = DynFlags { unfoldingOpts :: !UnfoldingOpts, maxWorkerArgs :: Int, + maxForcedSpecArgs :: Int, ghciHistSize :: Int, @@ -676,6 +677,8 @@ defaultDynFlags mySettings = unfoldingOpts = defaultUnfoldingOpts, maxWorkerArgs = 10, + maxForcedSpecArgs = 333, + -- 333 is fairly arbitrary, see Note [Forcing specialisation]:FS5 ghciHistSize = 50, -- keep a log of length 50 by default ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -1820,6 +1820,8 @@ dynamic_flags_deps = [ , make_ord_flag defFlag "fmax-worker-args" (intSuffix (\n d -> d {maxWorkerArgs = n})) + , make_ord_flag defFlag "fmax-forced-spec-args" + (intSuffix (\n d -> d {maxForcedSpecArgs = n})) , make_ord_flag defGhciFlag "fghci-hist-size" (intSuffix (\n d -> d {ghciHistSize = n})) , make_ord_flag defGhcFlag "fmax-inline-alloc-size" ===================================== docs/users_guide/using-optimisation.rst ===================================== @@ -870,6 +870,21 @@ as such you shouldn't need to set any of them explicitly. A flag value arguments of the resulting worker exceeds both that of the original function and this setting. +.. ghc-flag:: -fmax-forced-spec-args=⟨n⟩ + :shortdesc: *default: 333.* Maximum number of value arguments for forced SpecConstr specializations. + :type: dynamic + :category: + + :default: 512 + + When using ``SPEC`` from ``GHC.Types`` to force SpecConstr to fire on a function + sometimes this can result in functions taking a ridicolously large number of arguments + resulting a very large compile time hits for minor performance benefits. + + Since this is usually unintended we prevent SpecConstr from firing and generate + a warning if the number of arguments in the resulting function would exceed + the value given by ``-fmax-forced-spec-args``. + .. ghc-flag:: -fno-opt-coercion :shortdesc: Turn off the coercion optimiser :type: dynamic ===================================== testsuite/tests/simplCore/should_compile/T25197.hs ===================================== @@ -0,0 +1,30 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25197 where + +import T25197_TH +import GHC.Exts + +{- +This test applies a large statically known data structure to a function with +a SPEC argument, forcing the function to be specialised for the argument. +However when the complete structure of the argument is not statically known, +or as here the leaves of the structures are primitive literals for which we do +not specialize this results in a specialized function that can take hundreds of +arguments. + +Typically this is not intended, therefore we use a limit on the number of +arguments for specializations. As at some point this sort of specialization +comes with a heavy compile time cost. However we allow users to specify this +limit just in case they really depend on this sort of specialization. +-} + +foo :: [a] -> Int +foo = go SPEC + where + go :: SPEC -> [a] -> Int + go s [] = s `seq` 0 + go s (_:xs) = 1 + go s xs + +main :: IO () +main = print $ foo $(gen 1000) ===================================== testsuite/tests/simplCore/should_compile/T25197.stderr ===================================== @@ -0,0 +1,8 @@ +[1 of 2] Compiling T25197_TH ( T25197_TH.hs, T25197_TH.o ) +[2 of 2] Compiling T25197 ( T25197.hs, T25197.o ) +T25197.hs: warning: + SpecConstr encountered one or more function(s) with a SPEC argument that resulted in too many arguments, + which resulted in no specialization being generated for these functions: + $wgo Defined at T25197.hs:26:5 + If this is expected you might want to increase -fmax-forced-spec-args to force specialization anyway. + ===================================== testsuite/tests/simplCore/should_compile/T25197_TH.hs ===================================== @@ -0,0 +1,9 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25197_TH where + +import Language.Haskell.TH.Syntax + +gen :: Int -> Q Exp +gen 0 = [| [] |] +gen n = [| $(lift (show n)) : $(gen (n-1)) |] ===================================== testsuite/tests/simplCore/should_compile/all.T ===================================== @@ -530,3 +530,4 @@ test('T24625', [ grep_errmsg(r'case lazy') ], compile, ['-O -fno-ignore-asserts test('T24725a', [ grep_errmsg(r'testedRule')], compile, ['-O -ddump-rule-firings']) test('T25033', normal, compile, ['-O']) test('T25160', normal, compile, ['-O -ddump-rules']) +test('T25197', [req_th, extra_files(["T25197_TH.hs"]), only_ways(['optasm'])], multimod_compile, ['T25197', '-O2 -v0']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/62d1cabf6de59476bfb2d031d51bdd5fc0263006 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/62d1cabf6de59476bfb2d031d51bdd5fc0263006 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 14:42:28 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Fri, 27 Sep 2024 10:42:28 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] 21 commits: Add entity information to HieFile #24544 Message-ID: <66f6c4547e3db_f7e921267cc6768@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 6df57436 by Matthew Pickering at 2024-09-27T15:42:18+01:00 Fix toException method for ExceptionWithContext Fixes #25235 - - - - - c351c148 by Matthew Pickering at 2024-09-27T15:42:18+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. Implements CLC proposal#202 <https://github.com/haskell/core-libraries-committee/issues/202> - - - - - fa92c2f4 by Rodrigo Mesquita at 2024-09-27T15:42:18+01:00 Add test for #25300 This reverts commit f781b76635390b9f8c04362638193dc1b9c2541a. - - - - - 30 changed files: - .gitlab-ci.yml - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b23f019cf18224f937eacf3cde3e8deece4ec8eb...fa92c2f414dcfa6172e35efb75e4da649a2ebe7c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b23f019cf18224f937eacf3cde3e8deece4ec8eb...fa92c2f414dcfa6172e35efb75e4da649a2ebe7c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 15:09:43 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Fri, 27 Sep 2024 11:09:43 -0400 Subject: [Git][ghc/ghc][wip/romes/25304] determinism: Interface re-export list det Message-ID: <66f6cab7c0990_f7e922fda3c703ba@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/25304 at Glasgow Haskell Compiler / GHC Commits: ee4a870d by Rodrigo Mesquita at 2024-09-27T16:04:54+01:00 determinism: Interface re-export list det In 'DocStructureItem' we want to make sure the 'Avails' are sorted, for interface file determinism. This commit introduces 'SortedAvails', an abstract newtype that can only be constructed by sorting Avails with 'sortAvails'. This newtype is used by 'DocStructureItem' where 'Avails' was previously used to ensure the list of avails is deterministically sorted by construction. Fixes #25304 - - - - - 11 changed files: - compiler/GHC/Hs/Doc.hs - compiler/GHC/HsToCore/Docs.hs - compiler/GHC/Iface/Make.hs - compiler/GHC/Types/Avail.hs - + testsuite/tests/determinism/T25304/A.hs - + testsuite/tests/determinism/T25304/B.hs - + testsuite/tests/determinism/T25304/Makefile - + testsuite/tests/determinism/T25304/all.T - testsuite/tests/showIface/DocsInHiFileTH.stdout - testsuite/tests/showIface/NoExportList.stdout - utils/haddock/haddock-api/src/Haddock/Interface/Create.hs Changes: ===================================== compiler/GHC/Hs/Doc.hs ===================================== @@ -124,7 +124,7 @@ data DocStructureItem = DsiSectionHeading !Int !(HsDoc GhcRn) | DsiDocChunk !(HsDoc GhcRn) | DsiNamedChunkRef !String - | DsiExports !Avails + | DsiExports !SortedAvails | DsiModExport !(NonEmpty ModuleName) -- ^ We might re-export avails from multiple -- modules with a single export declaration. E.g. @@ -136,7 +136,7 @@ data DocStructureItem -- -- Invariant: This list of ModuleNames must be -- sorted to guarantee interface file determinism. - !Avails + !SortedAvails -- ^ Invariant: This list of Avails must be sorted -- to guarantee interface file determinism. ===================================== compiler/GHC/HsToCore/Docs.hs ===================================== @@ -160,7 +160,7 @@ mkDocStructureFromExportList mdl import_avails export_list = (IEGroup _ level doc, _) -> DsiSectionHeading level (unLoc doc) (IEDoc _ doc, _) -> DsiDocChunk (unLoc doc) (IEDocNamed _ name, _) -> DsiNamedChunkRef name - (_, avails) -> DsiExports (nubAvails avails) + (_, avails) -> DsiExports (sortAvails (nubAvails avails)) moduleExport :: ModuleName -- Alias -> Avails @@ -199,12 +199,12 @@ mkDocStructureFromDecls env all_exports decls = map unLoc (sortLocated (docs ++ avails)) where avails :: [Located DocStructureItem] - avails = flip fmap all_exports $ \avail -> + avails = flip fmap ({- sortAvails ?-} all_exports) $ \avail -> case M.lookup (availName avail) name_locs of - Just loc -> L loc (DsiExports [avail]) + Just loc -> L loc (DsiExports (UnsafeSortedAvails [avail])) -- FIXME: This is just a workaround that we use when handling e.g. -- associated data families like in the html-test Instances.hs. - Nothing -> noLoc (DsiExports []) + Nothing -> noLoc (DsiExports (UnsafeSortedAvails [])) -- This causes the associated data family to be incorrectly documented -- separately from its class: ===================================== compiler/GHC/Iface/Make.hs ===================================== @@ -518,8 +518,8 @@ mkIfaceImports = map go go (ImpUserSpec decl (ImpUserExplicit env)) = IfaceImport decl (ImpIfaceExplicit (forceGlobalRdrEnv env)) go (ImpUserSpec decl (ImpUserEverythingBut ns)) = IfaceImport decl (ImpIfaceEverythingBut ns) -mkIfaceExports :: [AvailInfo] -> [IfaceExport] -- Sort to make canonical -mkIfaceExports = sortAvails +mkIfaceExports :: [AvailInfo] -> [IfaceExport] -- Sort to make canonical +mkIfaceExports as = case sortAvails as of SortedAvails sas -> sas {- Note [Original module] ===================================== compiler/GHC/Types/Avail.hs ===================================== @@ -1,5 +1,7 @@ {-# LANGUAGE DeriveDataTypeable #-} +{-# LANGUAGE DerivingStrategies #-} +{-# LANGUAGE PatternSynonyms #-} -- -- (c) The University of Glasgow -- @@ -20,6 +22,7 @@ module GHC.Types.Avail ( filterAvails, nubAvails, sortAvails, + SortedAvails(SortedAvails, UnsafeSortedAvails) ) where import GHC.Prelude @@ -65,6 +68,20 @@ data AvailInfo -- | A collection of 'AvailInfo' - several things that are \"available\" type Avails = [AvailInfo] +-- | Occurrences of Avails in interface files must be sorted to guarantee +-- interface file determinism. +-- +-- To construct 'SortedAvails' using 'UnsafeSortedAvails' you must be sure the +-- 'Avails' are already sorted. Otherwise, you should use 'sortAvails'. +newtype SortedAvails = UnsafeSortedAvails Avails + deriving newtype (Binary, Outputable, NFData) + +-- | Safe matching on 'SortedAvails' +-- To construct 'SortedAvails' use 'sortAvails'. +pattern SortedAvails :: Avails -> SortedAvails +pattern SortedAvails x <- UnsafeSortedAvails x +{-# COMPLETE SortedAvails #-} + {- Note [Representing pattern synonym fields in AvailInfo] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Record pattern synonym fields cannot be represented using AvailTC like fields of @@ -133,8 +150,8 @@ availSubordinateNames avail@(AvailTC _ ns) | otherwise = ns -- | Sort 'Avails'/'AvailInfo's -sortAvails :: Avails -> Avails -sortAvails = sortBy stableAvailCmp . map sort_subs +sortAvails :: Avails -> SortedAvails +sortAvails = UnsafeSortedAvails . sortBy stableAvailCmp . map sort_subs where sort_subs :: AvailInfo -> AvailInfo sort_subs (Avail n) = Avail n ===================================== testsuite/tests/determinism/T25304/A.hs ===================================== @@ -0,0 +1,84 @@ +module A + ( MyType(..) + ) where + +data MyType + = A + | B + | C + | D + | E + | F + | G + | H + | I + | J + | K + | L + | M + | N + | O + | P + | Q + | R + | S + | T + | U + | V + | W + | X + | Y + | Z + | AA + | AB + | AC + | AD + | AE + | AF + | AG + | AH + | AI + | AJ + | AK + | AL + | AM + | AN + | AO + | AP + | AQ + | AR + | AS + | AT + | AU + | AV + | AW + | AX + | AY + | AZ + | BA + | BB + | BC + | BD + | BE + | BF + | BG + | BH + | BI + | BJ + | BK + | BL + | BM + | BN + | BO + | BP + | BQ + | BR + | BS + | BT + | BU + | BV + | BW + | BX + | BY + | BZ + | CA ===================================== testsuite/tests/determinism/T25304/B.hs ===================================== @@ -0,0 +1,86 @@ +module B +( MyType + ( AA + , AB + , AC + , AD + , AE + , AF + , AG + , AH + , AI + , AJ + , AK + , AL + , AM + , AN + , AO + , AP + , AQ + , AR + , AS + , AT + , AU + , AV + , AW + , AX + , AY + , AZ + , BA + , BB + , BC + , BD + , BE + , BF + , BG + , BH + , BI + , BJ + , BK + , BL + , BM + , BN + , BO + , BP + , BQ + , BR + , BS + , BT + , BU + , BV + , BW + , BX + , BY + , BZ + , CA + , A + , B + , C + , D + , E + , F + , G + , H + , I + , J + , K + , L + , M + , N + , O + , P + , Q + , R + , S + , T + , U + , V + , W + , X + , Y + , Z + ) +) where + +import A + ===================================== testsuite/tests/determinism/T25304/Makefile ===================================== @@ -0,0 +1,14 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +T25304: + $(RM) A.hi A.o B.hi B.o + '$(TEST_HC)' $(TEST_HC_OPTS) -v0 B.hs + '$(TEST_HC)' --show-iface A.hi > A_clean_iface + '$(TEST_HC)' --show-iface B.hi > B_clean_iface + '$(TEST_HC)' $(TEST_HC_OPTS) -dunique-increment=-1 -v0 A.hs B.hs -fforce-recomp + '$(TEST_HC)' --show-iface A.hi > A_dirty_iface + '$(TEST_HC)' --show-iface B.hi > B_dirty_iface + diff A_clean_iface A_dirty_iface + diff B_clean_iface B_dirty_iface ===================================== testsuite/tests/determinism/T25304/all.T ===================================== @@ -0,0 +1 @@ +test('T25304', [extra_files(['A.hs', 'B.hs'])], makefile_test, ['T25304']) ===================================== testsuite/tests/showIface/DocsInHiFileTH.stdout ===================================== @@ -187,7 +187,7 @@ docs: avails: [i] avails: - [WD11{WD11, WD11Bool, WD11Int, WD11Foo}] + [WD11{WD11, WD11Bool, WD11Foo, WD11Int}] avails: [WD13{WD13}] avails: @@ -219,17 +219,17 @@ docs: avails: [wd20] avails: - [Pretty{Pretty, prettyPrint}] + [Tup2] avails: - [Corge{Corge, runCorge, Corge}] + [qux] avails: - [Quuz{Quuz, quuz1_a, Quuz}] + [Corge{Corge, Corge, runCorge}] avails: - [Quux{Quux, Quux2, Quux1}] + [Pretty{Pretty, prettyPrint}] avails: - [Tup2] + [Quux{Quux, Quux1, Quux2}] avails: - [qux] + [Quuz{Quuz, Quuz, quuz1_a}] named chunks: haddock options: language: ===================================== testsuite/tests/showIface/NoExportList.stdout ===================================== @@ -32,7 +32,7 @@ docs: -- Actually we have only one type. identifiers: avails: - [R{R, fβ, fα, R}] + [R{R, R, fα, fβ}] section heading, level 1: text: -- * Functions ===================================== utils/haddock/haddock-api/src/Haddock/Interface/Create.hs ===================================== @@ -201,7 +201,7 @@ createInterface1' flags unit_state dflags hie_file mod_iface ifaces inst_ifaces -- See Note [Exporting built-in items] let builtinTys = DsiSectionHeading 1 (WithHsDocIdentifiers (mkGeneratedHsDocString "Builtin syntax") []) bonus_ds mods - | mdl == gHC_PRIM = [builtinTys, DsiExports funAvail] <> mods + | mdl == gHC_PRIM = [builtinTys, DsiExports (UnsafeSortedAvails funAvail)] <> mods | otherwise = mods let @@ -461,11 +461,11 @@ mkExportItems Just hsDoc' -> do doc <- processDocStringParas parserOpts sDocContext pkgName hsDoc' pure [ExportDoc doc] - DsiExports avails -> + DsiExports (SortedAvails avails) -> -- TODO: We probably don't need nubAvails here. -- mkDocStructureFromExportList already uses it. concat <$> traverse availExport (nubAvails avails) - DsiModExport mod_names avails -> do + DsiModExport mod_names (SortedAvails avails) -> do -- only consider exporting a module if we are sure we are really -- exporting the whole module and not some subset. (unrestricted_mods, remaining_avails) <- unrestrictedModExports sDocContext thisMod modMap instIfaceMap avails (NE.toList mod_names) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ee4a870d76e2bd931d96250531d3284454f07296 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ee4a870d76e2bd931d96250531d3284454f07296 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 15:27:54 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 11:27:54 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 71 commits: ci: Run abi-test on test-abi label Message-ID: <66f6cefa3f695_f7e924258607484b@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 1c03caf3 by Matthew Pickering at 2024-09-27T16:21:28+01:00 Add missing req_interp modifier to T18441fail3 and T18441fail19 These tests require the interpreter but they were failing in a different way with the javascript backend because the interpreter was disabled and stderr is ignored by the test. - - - - - f947a955 by Matthew Pickering at 2024-09-27T16:21:28+01:00 Use explicit syntax rather than pure - - - - - 6d025bff by Matthew Pickering at 2024-09-27T16:21:28+01:00 packaging: correctly propagate build/host/target to bindist configure script At the moment the host and target which we will produce a compiler for is fixed at the initial configure time. Therefore we need to persist the choice made at this time into the installation bindist as well so we look for the right tools, with the right prefixes at install time. In the future, we want to provide a bit more control about what kind of bindist we produce so the logic about what the host/target will have to be written by hadrian rather than persisted by the configure script. In particular with cross compilers we want to either build a normal stage 2 cross bindist or a stage 3 bindist, which creates a bindist which has a native compiler for the target platform. Fixes #21970 - - - - - 3b510384 by Matthew Pickering at 2024-09-27T16:21:28+01:00 hadrian: Fill in more of the default.host toolchain file When you are building a cross compiler this file will be used to build stage1 and it's libraries, so we need enough information here to work accurately. There is still more work to be done (see for example, word size is still fixed). - - - - - 74d0c3a2 by Matthew Pickering at 2024-09-27T16:21:28+01:00 hadrian: Disable docs when cross compiling Before there were a variety of ad-hoc places where doc building was disabled when cross compiling. * Some CI jobs sets --docs=none in gen_ci.hs * Some CI jobs set --docs=none in .gitlab/ci.sh * There was some logic in hadrian to not need the ["docs"] target when making a bindist. Now the situation is simple: * If you are cross compiling then defaultDocsTargets is empty by default. In theory, there is no reason why we can't build documentation for cross compiler bindists, but this is left to future work to generalise the documentation building rules to allow this (#24289) - - - - - ee7e5ba2 by Matthew Pickering at 2024-09-27T16:27:39+01:00 hadrian: Build stage 2 cross compilers * Most of hadrian is abstracted over the stage in order to remove the assumption that the target of all stages is the same platform. This allows the RTS to be built for two different targets for example. * Abstracts the bindist creation logic to allow building either normal or cross bindists. Normal bindists use stage 1 libraries and a stage 2 compiler. Cross bindists use stage 2 libararies and a stage 2 compiler. * hadrian: Make binary-dist-dir the default build target. This allows us to have the logic in one place about which libraries/stages to build with cross compilers. Fixes #24192 New hadrian target: * `binary-dist-dir-cross`: Build a cross compiler bindist (compiler = stage 1, libraries = stage 2) ------------------------- Metric Decrease: T10421a T10858 T11195 T11276 T11374 T11822 T15630 T17096 T18478 T20261 Metric Increase: parsing001 ------------------------- - - - - - c4351d8c by Matthew Pickering at 2024-09-27T16:27:40+01:00 ci: Test cross bindists We remove the special logic for testing in-tree cross compilers and instead test cross compiler bindists, like we do for all other platforms. - - - - - 57668fb8 by Matthew Pickering at 2024-09-27T16:27:40+01:00 ci: Javascript don't set CROSS_EMULATOR There is no CROSS_EMULATOR needed to run javascript binaries, so we don't set the CROSS_EMULATOR to some dummy value. - - - - - 2f74e234 by Matthew Pickering at 2024-09-27T16:27:40+01:00 ci: Introduce CROSS_STAGE variable In preparation for building and testing stage3 bindists we introduce the CROSS_STAGE variable which is used by a CI job to determine what kind of bindist the CI job should produce. At the moment we are only using CROSS_STAGE=2 but in the future we will have some jobs which set CROSS_STAGE=3 to produce native bindists for a target, but produced by a cross compiler, which can be tested on by another CI job on the native platform. CROSS_STAGE=2: Build a normal cross compiler bindist CROSS_STAGE=3: Build a stage 3 bindist, one which is a native compiler and library for the target - - - - - b162f9bc by Matthew Pickering at 2024-09-27T16:27:40+01:00 Split up system.config into host/target config files There were a number of settings which were not applied per-stage, for example if you specified `--ffi-include-dir` then that was applied to both host and target. Now this will just be passed when building the crosscompiler. The solution for now is to separate these two files into host/target and the host file contains very bare-bones . There isn't currently a way to specify with configure anything in the host file, so if you are building a cross-compiler and you need to do that, you have to modify the file yourself. - - - - - 29016e79 by Matthew Pickering at 2024-09-27T16:27:40+01:00 wip fixup - - - - - 3252c35a by Matthew Pickering at 2024-09-27T16:27:40+01:00 Fix location of emsdk-version - - - - - f1f91cf0 by Matthew Pickering at 2024-09-27T16:27:40+01:00 fix distrib/configure file - - - - - 7b1300cd by Matthew Pickering at 2024-09-27T16:27:40+01:00 Fix hardcoded stage1 - - - - - 07441a69 by Matthew Pickering at 2024-09-27T16:27:40+01:00 Don't recache - - - - - 8b9bee49 by Matthew Pickering at 2024-09-27T16:27:40+01:00 hadrian: Make text_simdutf flavour transformer configurable per-stage Before it was globally enabled, which was probably not what you want as you don't need text-simd for your boot compiler nor your boot compiler if you're building a cross-compiler. This brings it into line with the other modifiers.. such as ghcProfiled etc Fixes #25302 - - - - - 5ea0d26e by Matthew Pickering at 2024-09-27T16:27:40+01:00 hadrian: Refactor system-cxx-std-lib rules0 I noticed a few things wrong with the hadrian rules for `system-cxx-std-lib` rules. * For `text` there is an ad-hoc check to depend on `system-cxx-std-lib` outside of `configurePackage`. * The `system-cxx-std-lib` dependency is not read from cabal files. * Recache is not called on the packge database after the `.conf` file is generated, a more natural place for this rule is `registerRules`. Treating this uniformly like other packages is complicated by it not having any source code or a cabal file. However we can do a bit better by reporting the dependency firstly in `PackageData` and then needing the `.conf` file in the same place as every other package in `configurePackage`. Fixes #25303 - - - - - 77ecb41b by Matthew Pickering at 2024-09-27T16:27:40+01:00 fixes for simdutf8 - - - - - 12db4aed by Matthew Pickering at 2024-09-27T16:27:40+01:00 use building for target in llvm flavour transformer - - - - - 30bbc6a9 by Matthew Pickering at 2024-09-27T16:27:40+01:00 bindist: Pass path to package database we want to recache This fixes recaching on cross compilers - - - - - ed47556d by Matthew Pickering at 2024-09-27T16:27:40+01:00 testsuite: T9930fail now passes on javascript I didn't investigate why, but the comment says it should be fixed by building a stage2 cross compiler (and it is). - - - - - 21 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/dc59add291ee428398c323ce6a5a59078849b911...ed47556df729efe9608853c99bcee7da72e8a557 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/dc59add291ee428398c323ce6a5a59078849b911...ed47556df729efe9608853c99bcee7da72e8a557 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 15:37:06 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Fri, 27 Sep 2024 11:37:06 -0400 Subject: [Git][ghc/ghc][wip/T25281] Tiny wibble Message-ID: <66f6d12253d2a_f7e92564af0801b@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: 276f6d1e by Simon Peyton Jones at 2024-09-27T16:36:48+01:00 Tiny wibble - - - - - 1 changed file: - compiler/GHC/Hs/Expr.hs Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -648,7 +648,7 @@ ppr_lexpr e = ppr_expr (unLoc e) ppr_expr :: forall p. (OutputableBndrId p) => HsExpr (GhcPass p) -> SDoc -ppr_expr (HsVar _ (L loc v)) = pprPrefixOcc v +ppr_expr (HsVar _ (L _ v)) = pprPrefixOcc v ppr_expr (HsUnboundVar _ uv) = pprPrefixOcc uv ppr_expr (HsRecSel _ f) = pprPrefixOcc f ppr_expr (HsIPVar _ v) = ppr v View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/276f6d1e853e46eb5788699389cadfb84eb8ce9b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/276f6d1e853e46eb5788699389cadfb84eb8ce9b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 15:38:36 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Fri, 27 Sep 2024 11:38:36 -0400 Subject: [Git][ghc/ghc][wip/T23479] 84 commits: Linker: add support for extra built-in symbols (#25155) Message-ID: <66f6d17ca576b_f7e9257cc40805ed@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - fdf5b0f4 by Serge S. Gulin at 2024-09-27T18:38:20+03:00 JS: Re-add optimization for literal strings in genApp (fixes 23479 (muted temporary)) Based on https://gitlab.haskell.org/ghc/ghc/-/merge_requests/10588/ - - - - - eb0291f3 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Use name defined at `GHC.Builtin.Names` - - - - - e12b9e79 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Apply 1 suggestion(s) to 1 file(s) Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - d79089b0 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Attempt to take 805 for id - - - - - 6cb26f49 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Attempt to add to basicKnownKeyNames Co-authored-by: Andrei Borzenkov <root at sandwitch.dev> Co-authored-by: Danil Berestov <goosedb at yandex.ru> - - - - - a3b19ec3 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Naive attempt to add `StgLitArg (LitString bs)` - - - - - ce64ac65 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 WIP add logging - - - - - 6a354e4e by Serge S. Gulin at 2024-09-27T18:38:21+03:00 WIP add logging - - - - - 9cb15cfa by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Add STG debug from JS Sinker - - - - - 5efea55e by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Add eager Sinker's strings unfloater - - - - - f2670904 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Add STG debug from JS Sinker - - - - - 4b02440e by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Add limitations to unfloat string lits - - - - - 643f27f7 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 Fix build - - - - - cf2331b9 by Serge S. Gulin at 2024-09-27T18:38:21+03:00 String lits were removed too early, need do it at Linker instead - - - - - 21 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/91765d8ecc525c1bf17e0be6b008d6b9af184501...cf2331b9db92ade4500906f59966a1b0ddab0874 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/91765d8ecc525c1bf17e0be6b008d6b9af184501...cf2331b9db92ade4500906f59966a1b0ddab0874 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 15:41:15 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 11:41:15 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/hadrian-cross-stage2-riscv Message-ID: <66f6d21b2a39a_f7e92640e2484415@gitlab.mail> Matthew Pickering pushed new branch wip/hadrian-cross-stage2-riscv at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/hadrian-cross-stage2-riscv You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 15:43:11 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 11:43:11 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2-riscv] Bump LLVM upper bound to allow LLVM 19 Message-ID: <66f6d28f2e75c_f7e9258185884663@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2-riscv at Glasgow Haskell Compiler / GHC Commits: 5b51dc5f by Matthew Pickering at 2024-09-27T16:42:58+01:00 Bump LLVM upper bound to allow LLVM 19 Also bumps the ci-images commit so that the deb12 images uses LLVM 19 for testing. ------------------------- Metric Decrease: size_hello_artifact_gzip size_hello_unicode_gzip ------------------------- Fixes #25295 - - - - - 2 changed files: - .gitlab-ci.yml - configure.ac Changes: ===================================== .gitlab-ci.yml ===================================== @@ -2,7 +2,7 @@ variables: GIT_SSL_NO_VERIFY: "1" # Commit of ghc/ci-images repository from which to pull Docker images - DOCKER_REV: bd0e387ad7f67e16b79a7d940ca65452a6d20c04 + DOCKER_REV: 575c404d23035c6e9b5815962e5c0e7d4f72ec88 # Sequential version number of all cached things. # Bump to invalidate GitLab CI cache. ===================================== configure.ac ===================================== @@ -561,7 +561,7 @@ AC_SUBST(InstallNameToolCmd) # versions of LLVM simultaneously, but that stopped working around # 3.5/3.6 release of LLVM. LlvmMinVersion=13 # inclusive -LlvmMaxVersion=19 # not inclusive +LlvmMaxVersion=20 # not inclusive AC_SUBST([LlvmMinVersion]) AC_SUBST([LlvmMaxVersion]) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5b51dc5fc7bd7e99bc16b16c2b93e1f77eb4e29a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5b51dc5fc7bd7e99bc16b16c2b93e1f77eb4e29a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 15:44:11 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 11:44:11 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2-riscv] docker rev Message-ID: <66f6d2cb4858f_f7e927eb8008484@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2-riscv at Glasgow Haskell Compiler / GHC Commits: 2fa9807a by Matthew Pickering at 2024-09-27T16:44:05+01:00 docker rev - - - - - 1 changed file: - .gitlab-ci.yml Changes: ===================================== .gitlab-ci.yml ===================================== @@ -2,7 +2,7 @@ variables: GIT_SSL_NO_VERIFY: "1" # Commit of ghc/ci-images repository from which to pull Docker images - DOCKER_REV: 575c404d23035c6e9b5815962e5c0e7d4f72ec88 + DOCKER_REV: 6efac743853f9c2172777e934d7aea44434415ec # Sequential version number of all cached things. # Bump to invalidate GitLab CI cache. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2fa9807a6247263ef4bd7c21bc2fa255e45df971 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2fa9807a6247263ef4bd7c21bc2fa255e45df971 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 15:45:51 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 11:45:51 -0400 Subject: [Git][ghc/ghc][wip/llvm_19] Bump LLVM upper bound to allow LLVM 19 Message-ID: <66f6d32ff116c_f7e92826770850cb@gitlab.mail> Matthew Pickering pushed to branch wip/llvm_19 at Glasgow Haskell Compiler / GHC Commits: 9c71310f by Matthew Pickering at 2024-09-27T16:45:20+01:00 Bump LLVM upper bound to allow LLVM 19 Also bumps the ci-images commit so that the deb12 images uses LLVM 19 for testing. ------------------------- Metric Decrease: size_hello_artifact_gzip size_hello_unicode_gzip ------------------------- Fixes #25295 - - - - - 2 changed files: - .gitlab-ci.yml - configure.ac Changes: ===================================== .gitlab-ci.yml ===================================== @@ -2,7 +2,7 @@ variables: GIT_SSL_NO_VERIFY: "1" # Commit of ghc/ci-images repository from which to pull Docker images - DOCKER_REV: 486541129a8e7bf77c2cf7cd76ca998f690d5685 + DOCKER_REV: 6efac743853f9c2172777e934d7aea44434415ec # Sequential version number of all cached things. # Bump to invalidate GitLab CI cache. ===================================== configure.ac ===================================== @@ -528,7 +528,7 @@ AC_SUBST(InstallNameToolCmd) # versions of LLVM simultaneously, but that stopped working around # 3.5/3.6 release of LLVM. LlvmMinVersion=13 # inclusive -LlvmMaxVersion=19 # not inclusive +LlvmMaxVersion=20 # not inclusive AC_SUBST([LlvmMinVersion]) AC_SUBST([LlvmMaxVersion]) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9c71310fc41cb47586f5554bf72ec3527590f4ed -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9c71310fc41cb47586f5554bf72ec3527590f4ed You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 16:13:29 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Fri, 27 Sep 2024 12:13:29 -0400 Subject: [Git][ghc/ghc][wip/romes/25304] determinism: Interface re-export list det Message-ID: <66f6d9a9719d9_f7e92a6d7cc8598a@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/25304 at Glasgow Haskell Compiler / GHC Commits: cefebb13 by Rodrigo Mesquita at 2024-09-27T17:13:15+01:00 determinism: Interface re-export list det In 'DocStructureItem' we want to make sure the 'Avails' are sorted, for interface file determinism. This commit introduces 'SortedAvails', an abstract newtype that can only be constructed by sorting Avails with 'sortAvails'. This newtype is used by 'DocStructureItem' where 'Avails' was previously used to ensure the list of avails is deterministically sorted by construction. Fixes #25304 - - - - - 12 changed files: - compiler/GHC/Hs/Doc.hs - compiler/GHC/HsToCore/Docs.hs - compiler/GHC/Iface/Make.hs - compiler/GHC/Types/Avail.hs - + testsuite/tests/determinism/T25304/A.hs - + testsuite/tests/determinism/T25304/B.hs - + testsuite/tests/determinism/T25304/Makefile - + testsuite/tests/determinism/T25304/T25304a.stdout - + testsuite/tests/determinism/T25304/all.T - testsuite/tests/showIface/DocsInHiFileTH.stdout - testsuite/tests/showIface/NoExportList.stdout - utils/haddock/haddock-api/src/Haddock/Interface/Create.hs Changes: ===================================== compiler/GHC/Hs/Doc.hs ===================================== @@ -124,7 +124,7 @@ data DocStructureItem = DsiSectionHeading !Int !(HsDoc GhcRn) | DsiDocChunk !(HsDoc GhcRn) | DsiNamedChunkRef !String - | DsiExports !Avails + | DsiExports !SortedAvails | DsiModExport !(NonEmpty ModuleName) -- ^ We might re-export avails from multiple -- modules with a single export declaration. E.g. @@ -136,7 +136,7 @@ data DocStructureItem -- -- Invariant: This list of ModuleNames must be -- sorted to guarantee interface file determinism. - !Avails + !SortedAvails -- ^ Invariant: This list of Avails must be sorted -- to guarantee interface file determinism. ===================================== compiler/GHC/HsToCore/Docs.hs ===================================== @@ -160,7 +160,11 @@ mkDocStructureFromExportList mdl import_avails export_list = (IEGroup _ level doc, _) -> DsiSectionHeading level (unLoc doc) (IEDoc _ doc, _) -> DsiDocChunk (unLoc doc) (IEDocNamed _ name, _) -> DsiNamedChunkRef name - (_, avails) -> DsiExports (nubAvails avails) + (IEThingWith{}, avails) -> + DsiExports $ + {- For explicit export lists, use the explicit order. It is deterministic by construction -} + UnsafeSortedAvails (nubAvails avails) + (_, avails) -> DsiExports (sortAvails (nubAvails avails)) moduleExport :: ModuleName -- Alias -> Avails @@ -199,12 +203,12 @@ mkDocStructureFromDecls env all_exports decls = map unLoc (sortLocated (docs ++ avails)) where avails :: [Located DocStructureItem] - avails = flip fmap all_exports $ \avail -> + avails = flip fmap ({- sortAvails ?-} all_exports) $ \avail -> case M.lookup (availName avail) name_locs of - Just loc -> L loc (DsiExports [avail]) + Just loc -> L loc (DsiExports (UnsafeSortedAvails [avail])) -- FIXME: This is just a workaround that we use when handling e.g. -- associated data families like in the html-test Instances.hs. - Nothing -> noLoc (DsiExports []) + Nothing -> noLoc (DsiExports (UnsafeSortedAvails [])) -- This causes the associated data family to be incorrectly documented -- separately from its class: ===================================== compiler/GHC/Iface/Make.hs ===================================== @@ -518,8 +518,8 @@ mkIfaceImports = map go go (ImpUserSpec decl (ImpUserExplicit env)) = IfaceImport decl (ImpIfaceExplicit (forceGlobalRdrEnv env)) go (ImpUserSpec decl (ImpUserEverythingBut ns)) = IfaceImport decl (ImpIfaceEverythingBut ns) -mkIfaceExports :: [AvailInfo] -> [IfaceExport] -- Sort to make canonical -mkIfaceExports = sortAvails +mkIfaceExports :: [AvailInfo] -> [IfaceExport] -- Sort to make canonical +mkIfaceExports as = case sortAvails as of SortedAvails sas -> sas {- Note [Original module] ===================================== compiler/GHC/Types/Avail.hs ===================================== @@ -1,5 +1,7 @@ {-# LANGUAGE DeriveDataTypeable #-} +{-# LANGUAGE DerivingStrategies #-} +{-# LANGUAGE PatternSynonyms #-} -- -- (c) The University of Glasgow -- @@ -20,6 +22,7 @@ module GHC.Types.Avail ( filterAvails, nubAvails, sortAvails, + SortedAvails(SortedAvails, UnsafeSortedAvails) ) where import GHC.Prelude @@ -65,6 +68,20 @@ data AvailInfo -- | A collection of 'AvailInfo' - several things that are \"available\" type Avails = [AvailInfo] +-- | Occurrences of Avails in interface files must be sorted to guarantee +-- interface file determinism. +-- +-- To construct 'SortedAvails' using 'UnsafeSortedAvails' you must be sure the +-- 'Avails' are already sorted. Otherwise, you should use 'sortAvails'. +newtype SortedAvails = UnsafeSortedAvails Avails + deriving newtype (Binary, Outputable, NFData) + +-- | Safe matching on 'SortedAvails' +-- To construct 'SortedAvails' use 'sortAvails'. +pattern SortedAvails :: Avails -> SortedAvails +pattern SortedAvails x <- UnsafeSortedAvails x +{-# COMPLETE SortedAvails #-} + {- Note [Representing pattern synonym fields in AvailInfo] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Record pattern synonym fields cannot be represented using AvailTC like fields of @@ -133,8 +150,8 @@ availSubordinateNames avail@(AvailTC _ ns) | otherwise = ns -- | Sort 'Avails'/'AvailInfo's -sortAvails :: Avails -> Avails -sortAvails = sortBy stableAvailCmp . map sort_subs +sortAvails :: Avails -> SortedAvails +sortAvails = UnsafeSortedAvails . sortBy stableAvailCmp . map sort_subs where sort_subs :: AvailInfo -> AvailInfo sort_subs (Avail n) = Avail n ===================================== testsuite/tests/determinism/T25304/A.hs ===================================== @@ -0,0 +1,84 @@ +module A + ( MyType(..) + ) where + +data MyType + = A + | B + | C + | D + | E + | F + | G + | H + | I + | J + | K + | L + | M + | N + | O + | P + | Q + | R + | S + | T + | U + | V + | W + | X + | Y + | Z + | AA + | AB + | AC + | AD + | AE + | AF + | AG + | AH + | AI + | AJ + | AK + | AL + | AM + | AN + | AO + | AP + | AQ + | AR + | AS + | AT + | AU + | AV + | AW + | AX + | AY + | AZ + | BA + | BB + | BC + | BD + | BE + | BF + | BG + | BH + | BI + | BJ + | BK + | BL + | BM + | BN + | BO + | BP + | BQ + | BR + | BS + | BT + | BU + | BV + | BW + | BX + | BY + | BZ + | CA ===================================== testsuite/tests/determinism/T25304/B.hs ===================================== @@ -0,0 +1,86 @@ +module B +( MyType + ( BA + , BB + , BC + , BD + , BE + , BF + , BG + , BH + , BI + , BJ + , BK + , BL + , BM + , BN + , BO + , BP + , BQ + , BR + , BS + , BT + , BU + , BV + , BW + , BX + , BY + , BZ + , CA + , AA + , AB + , AC + , AD + , AE + , AF + , AG + , AH + , AI + , AJ + , AK + , AL + , AM + , AN + , AO + , AP + , AQ + , AR + , AS + , AT + , AU + , AV + , AW + , AX + , AY + , AZ + , A + , B + , C + , D + , E + , F + , G + , H + , I + , J + , K + , L + , M + , N + , O + , P + , Q + , R + , S + , T + , U + , V + , W + , X + , Y + , Z + ) +) where + +import A + ===================================== testsuite/tests/determinism/T25304/Makefile ===================================== @@ -0,0 +1,25 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +T25304: + $(RM) A.hi A.o B.hi B.o + # Use -haddock to get docs: output in the interface file + '$(TEST_HC)' $(TEST_HC_OPTS) -v0 -haddock B.hs + '$(TEST_HC)' --show-iface A.hi > A_clean_iface + '$(TEST_HC)' --show-iface B.hi > B_clean_iface + '$(TEST_HC)' $(TEST_HC_OPTS) -dunique-increment=-1 -v0 -haddock A.hs B.hs -fforce-recomp + '$(TEST_HC)' --show-iface A.hi > A_dirty_iface + '$(TEST_HC)' --show-iface B.hi > B_dirty_iface + diff A_clean_iface A_dirty_iface + diff B_clean_iface B_dirty_iface + +T25304a: + $(RM) A.hi A.o B.hi B.o + # Use -haddock to get docs: output in the interface file + '$(TEST_HC)' $(TEST_HC_OPTS) -v0 -haddock B.hs + '$(TEST_HC)' --show-iface B.hi > B_clean_iface + # The goal is to see the export list in the documentation structure of the + # interface file preserves the order used in the source + cat B_clean_iface | grep -A7 "documentation structure" + ===================================== testsuite/tests/determinism/T25304/T25304a.stdout ===================================== @@ -0,0 +1,8 @@ + documentation structure: + avails: + [A.MyType{A.MyType, A.BA, A.BB, A.BC, A.BD, A.BE, A.BF, A.BG, A.BH, + A.BI, A.BJ, A.BK, A.BL, A.BM, A.BN, A.BO, A.BP, A.BQ, A.BR, A.BS, + A.BT, A.BU, A.BV, A.BW, A.BX, A.BY, A.BZ, A.CA, A.AA, A.AB, A.AC, + A.AD, A.AE, A.AF, A.AG, A.AH, A.AI, A.AJ, A.AK, A.AL, A.AM, A.AN, + A.AO, A.AP, A.AQ, A.AR, A.AS, A.AT, A.AU, A.AV, A.AW, A.AX, A.AY, + A.AZ, A.A, A.B, A.C, A.D, A.E, A.F, A.G, A.H, A.I, A.J, A.K, A.L, ===================================== testsuite/tests/determinism/T25304/all.T ===================================== @@ -0,0 +1,2 @@ +test('T25304', [extra_files(['A.hs', 'B.hs'])], makefile_test, ['T25304']) +test('T25304a', [extra_files(['A.hs', 'B.hs'])], makefile_test, ['T25304a']) ===================================== testsuite/tests/showIface/DocsInHiFileTH.stdout ===================================== @@ -187,7 +187,7 @@ docs: avails: [i] avails: - [WD11{WD11, WD11Bool, WD11Int, WD11Foo}] + [WD11{WD11, WD11Bool, WD11Foo, WD11Int}] avails: [WD13{WD13}] avails: @@ -219,17 +219,17 @@ docs: avails: [wd20] avails: - [Pretty{Pretty, prettyPrint}] + [Tup2] avails: - [Corge{Corge, runCorge, Corge}] + [qux] avails: - [Quuz{Quuz, quuz1_a, Quuz}] + [Corge{Corge, Corge, runCorge}] avails: - [Quux{Quux, Quux2, Quux1}] + [Pretty{Pretty, prettyPrint}] avails: - [Tup2] + [Quux{Quux, Quux1, Quux2}] avails: - [qux] + [Quuz{Quuz, Quuz, quuz1_a}] named chunks: haddock options: language: ===================================== testsuite/tests/showIface/NoExportList.stdout ===================================== @@ -32,7 +32,7 @@ docs: -- Actually we have only one type. identifiers: avails: - [R{R, fβ, fα, R}] + [R{R, R, fα, fβ}] section heading, level 1: text: -- * Functions ===================================== utils/haddock/haddock-api/src/Haddock/Interface/Create.hs ===================================== @@ -201,7 +201,7 @@ createInterface1' flags unit_state dflags hie_file mod_iface ifaces inst_ifaces -- See Note [Exporting built-in items] let builtinTys = DsiSectionHeading 1 (WithHsDocIdentifiers (mkGeneratedHsDocString "Builtin syntax") []) bonus_ds mods - | mdl == gHC_PRIM = [builtinTys, DsiExports funAvail] <> mods + | mdl == gHC_PRIM = [builtinTys, DsiExports (UnsafeSortedAvails funAvail)] <> mods | otherwise = mods let @@ -461,11 +461,11 @@ mkExportItems Just hsDoc' -> do doc <- processDocStringParas parserOpts sDocContext pkgName hsDoc' pure [ExportDoc doc] - DsiExports avails -> + DsiExports (SortedAvails avails) -> -- TODO: We probably don't need nubAvails here. -- mkDocStructureFromExportList already uses it. concat <$> traverse availExport (nubAvails avails) - DsiModExport mod_names avails -> do + DsiModExport mod_names (SortedAvails avails) -> do -- only consider exporting a module if we are sure we are really -- exporting the whole module and not some subset. (unrestricted_mods, remaining_avails) <- unrestrictedModExports sDocContext thisMod modMap instIfaceMap avails (NE.toList mod_names) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/cefebb13c7f8ac9821037c703be3b8624e612c79 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/cefebb13c7f8ac9821037c703be3b8624e612c79 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 16:23:08 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Fri, 27 Sep 2024 12:23:08 -0400 Subject: [Git][ghc/ghc][wip/andreask/spec-constr-args] SpecConstr: Introduce a separate argument limit for forced specs. Message-ID: <66f6dbec32cf_f7e92a79144862f7@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/spec-constr-args at Glasgow Haskell Compiler / GHC Commits: 5e228031 by Andreas Klebinger at 2024-09-27T18:03:50+02:00 SpecConstr: Introduce a separate argument limit for forced specs. We used to put no limit at all on specializations forced via the SPEC argument. This isn't always reasonable so we introduce a very high limit that applies to forced specializations, a flag to control it, and we now emit a warning if we fail a specialization because we exceed the warning. Fixes #25197 - - - - - 8 changed files: - compiler/GHC/Core/Opt/SpecConstr.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Session.hs - docs/users_guide/using-optimisation.rst - + testsuite/tests/simplCore/should_compile/T25197.hs - + testsuite/tests/simplCore/should_compile/T25197.stderr - + testsuite/tests/simplCore/should_compile/T25197_TH.hs - testsuite/tests/simplCore/should_compile/all.T Changes: ===================================== compiler/GHC/Core/Opt/SpecConstr.hs ===================================== @@ -20,7 +20,8 @@ ToDo [Oct 2013] module GHC.Core.Opt.SpecConstr( specConstrProgram, - SpecConstrAnnotation(..) + SpecConstrAnnotation(..), + SpecFailWarning(..) ) where import GHC.Prelude @@ -51,6 +52,7 @@ import GHC.Core.Make ( mkImpossibleExpr ) import GHC.Unit.Module import GHC.Unit.Module.ModGuts +import GHC.Types.Error (MessageClass(..), Severity(..), DiagnosticReason(WarningWithoutFlag), ResolvedDiagnosticReason (..)) import GHC.Types.Literal ( litIsLifted ) import GHC.Types.Id import GHC.Types.Id.Info ( IdDetails(..) ) @@ -526,9 +528,11 @@ sc_force to True when calling specLoop. This flag does four things: (see argToPat; #4448) (FS4) Only specialise on recursive types a finite number of times (see sc_recursive; #5550; Note [Limit recursive specialisation]) -(FS5) Lift the restriction on the maximum number of arguments which - the optimisation will specialise. - (see `too_many_worker_args` in `callsToNewPats`; #14003) +(FS5) Use a different restriction on the maximum number of arguments which + the optimisation will specialise. We tried removing the limit on worker + args for forced specs (#14003) but this caused issues when specializing + code for large data structures (#25197). + This is handled by `too_many_worker_args` in `callsToNewPats` The flag holds only for specialising a single binding group, and NOT for nested bindings. (So really it should be passed around explicitly @@ -782,16 +786,25 @@ specConstrProgram :: ModGuts -> CoreM ModGuts specConstrProgram guts = do { env0 <- initScEnv guts ; us <- getUniqueSupplyM - ; let (_usg, binds') = initUs_ us $ + ; let (_usg, binds', warnings) = initUs_ us $ scTopBinds env0 (mg_binds guts) + ; when (not (null warnings)) $ msg specConstr_warn_class (warn_msg warnings) + ; return (guts { mg_binds = binds' }) } -scTopBinds :: ScEnv -> [InBind] -> UniqSM (ScUsage, [OutBind]) -scTopBinds _env [] = return (nullUsage, []) -scTopBinds env (b:bs) = do { (usg, b', bs') <- scBind TopLevel env b $ + where + specConstr_warn_class = MCDiagnostic SevWarning (ResolvedDiagnosticReason WarningWithoutFlag) Nothing + warn_msg :: SpecFailWarnings -> SDoc + warn_msg warnings = text "SpecConstr encountered one or more function(s) with a SPEC argument that resulted in too many arguments," $$ + text "which resulted in no specialization being generated for these functions:" $$ + nest 2 (vcat (map ppr warnings)) $$ + (text "If this is expected you might want to increase -fmax-forced-spec-args to force specialization anyway.") +scTopBinds :: ScEnv -> [InBind] -> UniqSM (ScUsage, [OutBind], [SpecFailWarning]) +scTopBinds _env [] = return (nullUsage, [], []) +scTopBinds env (b:bs) = do { (usg, b', bs', warnings) <- scBind TopLevel env b $ (\env -> scTopBinds env bs) - ; return (usg, b' ++ bs') } + ; return (usg, b' ++ bs', warnings) } {- ************************************************************************ @@ -905,6 +918,12 @@ data SpecConstrOpts = SpecConstrOpts -- ^ The threshold at which a worker-wrapper transformation used as part of -- this pass will no longer happen, measured in the number of arguments. + , sc_max_forced_args :: !Int + -- ^ The threshold at which a worker-wrapper transformation used as part of + -- this pass will no longer happen even if a SPEC arg was used to force + -- specialization. Measured in the number of arguments. + -- See Note [Forcing specialisation] + , sc_debug :: !Bool -- ^ Whether to print debug information @@ -975,6 +994,7 @@ instance Outputable Value where initScOpts :: DynFlags -> Module -> SpecConstrOpts initScOpts dflags this_mod = SpecConstrOpts { sc_max_args = maxWorkerArgs dflags, + sc_max_forced_args = maxForcedSpecArgs dflags, sc_debug = hasPprDebug dflags, sc_uf_opts = unfoldingOpts dflags, sc_module = this_mod, @@ -1388,29 +1408,29 @@ creates specialised versions of functions. -} scBind :: TopLevelFlag -> ScEnv -> InBind - -> (ScEnv -> UniqSM (ScUsage, a)) -- Specialise the scope of the binding - -> UniqSM (ScUsage, [OutBind], a) + -> (ScEnv -> UniqSM (ScUsage, a, [SpecFailWarning])) -- Specialise the scope of the binding + -> UniqSM (ScUsage, [OutBind], a, [SpecFailWarning]) scBind top_lvl env (NonRec bndr rhs) do_body | isTyVar bndr -- Type-lets may be created by doBeta - = do { (final_usage, body') <- do_body (extendScSubst env bndr rhs) - ; return (final_usage, [], body') } + = do { (final_usage, body', warnings) <- do_body (extendScSubst env bndr rhs) + ; return (final_usage, [], body', warnings) } | not (isTopLevel top_lvl) -- Nested non-recursive value binding -- See Note [Specialising local let bindings] = do { let (body_env, bndr') = extendBndr env bndr -- Not necessary at top level; but here we are nested - ; rhs_info <- scRecRhs env (bndr',rhs) + ; (rhs_info, rhs_ws) <- scRecRhs env (bndr',rhs) ; let body_env2 = extendHowBound body_env [bndr'] RecFun rhs' = ri_new_rhs rhs_info body_env3 = extendValEnv body_env2 bndr' (isValue (sc_vals env) rhs') - ; (body_usg, body') <- do_body body_env3 + ; (body_usg, body', warnings_body) <- do_body body_env3 -- Now make specialised copies of the binding, -- based on calls in body_usg - ; (spec_usg, specs) <- specNonRec env (scu_calls body_usg) rhs_info + ; (spec_usg, specs, warnings_bnd) <- specNonRec env (scu_calls body_usg) rhs_info -- NB: For non-recursive bindings we inherit sc_force flag from -- the parent function (see Note [Forcing specialisation]) @@ -1419,7 +1439,7 @@ scBind top_lvl env (NonRec bndr rhs) do_body bind_usage = (body_usg `delCallsFor` [bndr']) `combineUsage` spec_usg -- Note [spec_usg includes rhs_usg] - ; return (bind_usage, spec_bnds, body') + ; return (bind_usage, spec_bnds, body', mconcat [warnings_bnd, warnings_body, rhs_ws]) } | otherwise -- Top-level, non-recursive value binding @@ -1431,15 +1451,15 @@ scBind top_lvl env (NonRec bndr rhs) do_body -- -- I tried always specialising non-recursive top-level bindings too, -- but found some regressions (see !8135). So I backed off. - = do { (rhs_usage, rhs') <- scExpr env rhs + = do { (rhs_usage, rhs', ws_rhs) <- scExpr env rhs -- At top level, we've already put all binders into scope; see initScEnv -- Hence no need to call `extendBndr`. But we still want to -- extend the `ValueEnv` to record the value of this binder. ; let body_env = extendValEnv env bndr (isValue (sc_vals env) rhs') - ; (body_usage, body') <- do_body body_env + ; (body_usage, body', body_warnings) <- do_body body_env - ; return (rhs_usage `combineUsage` body_usage, [NonRec bndr rhs'], body') } + ; return (rhs_usage `combineUsage` body_usage, [NonRec bndr rhs'], body', body_warnings ++ ws_rhs) } scBind top_lvl env (Rec prs) do_body | isTopLevel top_lvl @@ -1450,19 +1470,20 @@ scBind top_lvl env (Rec prs) do_body -- ToDo: I'm honestly not sure of the rationale of this size-testing, nor -- why it only applies at top level. But that's the way it has been -- for a while. See #21456. - do { (body_usg, body') <- do_body rhs_env2 - ; (rhs_usgs, rhss') <- mapAndUnzipM (scExpr env) rhss + do { (body_usg, body', warnings_body) <- do_body rhs_env2 + ; (rhs_usgs, rhss', rhs_ws) <- mapAndUnzip3M (scExpr env) rhss ; let all_usg = (combineUsages rhs_usgs `combineUsage` body_usg) `delCallsFor` bndrs' bind' = Rec (bndrs' `zip` rhss') - ; return (all_usg, [bind'], body') } + ; return (all_usg, [bind'], body', warnings_body ++ concat rhs_ws) } | otherwise - = do { rhs_infos <- mapM (scRecRhs rhs_env2) (bndrs' `zip` rhss) - ; (body_usg, body') <- do_body rhs_env2 + = do { (rhs_infos, rhs_wss) <- mapAndUnzipM (scRecRhs rhs_env2) (bndrs' `zip` rhss) + ; let rhs_ws = mconcat rhs_wss + ; (body_usg, body', warnings_body) <- do_body rhs_env2 - ; (spec_usg, specs) <- specRec (scForce rhs_env2 force_spec) - (scu_calls body_usg) rhs_infos + ; (spec_usg, specs, spec_ws) <- specRec (scForce rhs_env2 force_spec) + (scu_calls body_usg) rhs_infos -- Do not unconditionally generate specialisations from rhs_usgs -- Instead use them only if we find an unspecialised call -- See Note [Seeding recursive groups] @@ -1473,7 +1494,7 @@ scBind top_lvl env (Rec prs) do_body -- zipWithEqual: length of returned [SpecInfo] -- should be the same as incoming [RhsInfo] - ; return (all_usg, [bind'], body') } + ; return (all_usg, [bind'], body', mconcat [warnings_body,rhs_ws,spec_ws]) } where (bndrs,rhss) = unzip prs force_spec = any (forceSpecBndr env) bndrs -- Note [Forcing specialisation] @@ -1501,59 +1522,63 @@ recursive function, but that's not essential and might even be harmful. I'm not sure. -} +withWarnings :: SpecFailWarnings -> (ScUsage, CoreExpr, SpecFailWarnings) -> (ScUsage, CoreExpr, SpecFailWarnings) +withWarnings ws (use,expr,ws2) = (use,expr,ws ++ ws2) + ------------------------ -scExpr, scExpr' :: ScEnv -> CoreExpr -> UniqSM (ScUsage, CoreExpr) +scExpr, scExpr' :: ScEnv -> CoreExpr -> UniqSM (ScUsage, CoreExpr, SpecFailWarnings) -- The unique supply is needed when we invent -- a new name for the specialised function and its args scExpr env e = scExpr' env e scExpr' env (Var v) = case scSubstId env v of - Var v' -> return (mkVarUsage env v' [], Var v') + Var v' -> return (mkVarUsage env v' [], Var v', []) e' -> scExpr (zapScSubst env) e' scExpr' env (Type t) = let !(MkSolo ty') = scSubstTy env t - in return (nullUsage, Type ty') -scExpr' env (Coercion c) = return (nullUsage, Coercion (scSubstCo env c)) -scExpr' _ e@(Lit {}) = return (nullUsage, e) -scExpr' env (Tick t e) = do (usg, e') <- scExpr env e - return (usg, Tick (scTickish env t) e') -scExpr' env (Cast e co) = do (usg, e') <- scExpr env e - return (usg, mkCast e' (scSubstCo env co)) + in return (nullUsage, Type ty', []) +scExpr' env (Coercion c) = return (nullUsage, Coercion (scSubstCo env c), []) +scExpr' _ e@(Lit {}) = return (nullUsage, e, []) +scExpr' env (Tick t e) = do (usg, e', ws) <- scExpr env e + return (usg, Tick (scTickish env t) e', ws) +scExpr' env (Cast e co) = do (usg, e', ws) <- scExpr env e + return (usg, mkCast e' (scSubstCo env co), ws) -- Important to use mkCast here -- See Note [SpecConstr call patterns] scExpr' env e@(App _ _) = scApp env (collectArgs e) scExpr' env (Lam b e) = do let (env', b') = extendBndr env b - (usg, e') <- scExpr env' e - return (usg, Lam b' e') + (usg, e', ws) <- scExpr env' e + return (usg, Lam b' e', ws) scExpr' env (Let bind body) - = do { (final_usage, binds', body') <- scBind NotTopLevel env bind $ + = do { (final_usage, binds', body', ws) <- scBind NotTopLevel env bind $ (\env -> scExpr env body) - ; return (final_usage, mkLets binds' body') } + ; return (final_usage, mkLets binds' body', ws) } scExpr' env (Case scrut b ty alts) - = do { (scrut_usg, scrut') <- scExpr env scrut + = do { (scrut_usg, scrut', ws) <- scExpr env scrut ; case isValue (sc_vals env) scrut' of Just (ConVal args_are_work_free con args) - | args_are_work_free -> sc_con_app con args scrut' + | args_are_work_free -> sc_con_app con args scrut' ws -- Don't duplicate work!! #7865 -- See Note [ConVal work-free-ness] (1) - _other -> sc_vanilla scrut_usg scrut' + _other -> sc_vanilla scrut_usg scrut' ws } where - sc_con_app con args scrut' -- Known constructor; simplify + sc_con_app con args scrut' ws -- Known constructor; simplify = do { let Alt _ bs rhs = findAlt con alts `orElse` Alt DEFAULT [] (mkImpossibleExpr ty "SpecConstr") alt_env' = extendScSubstList env ((b,scrut') : bs `zip` trimConArgs con args) - ; scExpr alt_env' rhs } + ; (use',expr',ws_new) <- scExpr alt_env' rhs + ; return (use',expr',ws ++ ws_new) } - sc_vanilla scrut_usg scrut' -- Normal case + sc_vanilla scrut_usg scrut' ws -- Normal case = do { let (alt_env,b') = extendBndrWith RecArg env b -- Record RecArg for the components - ; (alt_usgs, alt_occs, alts') <- mapAndUnzip3M (sc_alt alt_env scrut' b') alts + ; (alt_usgs, alt_occs, alts', ws_alts) <- mapAndUnzip4M (sc_alt alt_env scrut' b') alts ; let scrut_occ = foldr combineOcc NoOcc alt_occs scrut_usg' = setScrutOcc env scrut_usg scrut' scrut_occ @@ -1563,21 +1588,21 @@ scExpr' env (Case scrut b ty alts) ; let !(MkSolo ty') = scSubstTy env ty ; return (foldr combineUsage scrut_usg' alt_usgs, - Case scrut' b' ty' alts') } + Case scrut' b' ty' alts', ws ++ concat ws_alts) } single_alt = isSingleton alts sc_alt env scrut' b' (Alt con bs rhs) = do { let (env1, bs1) = extendBndrsWith RecArg env bs (env2, bs2) = extendCaseBndrs env1 scrut' b' con bs1 - ; (usg, rhs') <- scExpr env2 rhs + ; (usg, rhs', ws) <- scExpr env2 rhs ; let (usg', b_occ:arg_occs) = lookupOccs usg (b':bs2) scrut_occ = case con of DataAlt dc -- See Note [Do not specialise evals] | not (single_alt && all deadArgOcc arg_occs) -> ScrutOcc (unitUFM dc arg_occs) _ -> UnkOcc - ; return (usg', b_occ `combineOcc` scrut_occ, Alt con bs2 rhs') } + ; return (usg', b_occ `combineOcc` scrut_occ, Alt con bs2 rhs', ws) } -- | Substitute the free variables captured by a breakpoint. @@ -1626,19 +1651,20 @@ follows. still worth specialising on x. Hence the /single-alternative/ guard. -} -scApp :: ScEnv -> (InExpr, [InExpr]) -> UniqSM (ScUsage, CoreExpr) +scApp :: ScEnv -> (InExpr, [InExpr]) -> UniqSM (ScUsage, CoreExpr, SpecFailWarnings) scApp env (Var fn, args) -- Function is a variable = assert (not (null args)) $ do { args_w_usgs <- mapM (scExpr env) args - ; let (arg_usgs, args') = unzip args_w_usgs + ; let (arg_usgs, args', arg_ws) = unzip3 args_w_usgs arg_usg = combineUsages arg_usgs + arg_w = concat arg_ws ; case scSubstId env fn of - fn'@(Lam {}) -> scExpr (zapScSubst env) (doBeta fn' args') + fn'@(Lam {}) -> withWarnings arg_w <$> scExpr (zapScSubst env) (doBeta fn' args') -- Do beta-reduction and try again Var fn' -> return (arg_usg' `combineUsage` mkVarUsage env fn' args', - mkApps (Var fn') args') + mkApps (Var fn') args', arg_w ) where -- arg_usg': see Note [Specialising on dictionaries] arg_usg' | Just cls <- isClassOpId_maybe fn' @@ -1647,7 +1673,7 @@ scApp env (Var fn, args) -- Function is a variable | otherwise = arg_usg - other_fn' -> return (arg_usg, mkApps other_fn' args') } + other_fn' -> return (arg_usg, mkApps other_fn' args', arg_w) } -- NB: doing this ignores any usage info from the substituted -- function, but I don't think that matters. If it does -- we can fix it. @@ -1661,9 +1687,9 @@ scApp env (Var fn, args) -- Function is a variable -- which it may, we can get -- (let f = ...f... in f) arg1 arg2 scApp env (other_fn, args) - = do { (fn_usg, fn') <- scExpr env other_fn - ; (arg_usgs, args') <- mapAndUnzipM (scExpr env) args - ; return (combineUsages arg_usgs `combineUsage` fn_usg, mkApps fn' args') } + = do { (fn_usg, fn', fn_ws) <- scExpr env other_fn + ; (arg_usgs, args', arg_ws) <- mapAndUnzip3M (scExpr env) args + ; return (combineUsages arg_usgs `combineUsage` fn_usg, mkApps fn' args', combineSpecWarning fn_ws (concat arg_ws)) } ---------------------- mkVarUsage :: ScEnv -> Id -> [CoreExpr] -> ScUsage @@ -1679,16 +1705,16 @@ mkVarUsage env fn args | otherwise = evalScrutOcc ---------------------- -scRecRhs :: ScEnv -> (OutId, InExpr) -> UniqSM RhsInfo +scRecRhs :: ScEnv -> (OutId, InExpr) -> UniqSM (RhsInfo, SpecFailWarnings) scRecRhs env (bndr,rhs) = do { let (arg_bndrs,body) = collectBinders rhs (body_env, arg_bndrs') = extendBndrsWith RecArg env arg_bndrs - ; (body_usg, body') <- scExpr body_env body + ; (body_usg, body', body_ws) <- scExpr body_env body ; let (rhs_usg, arg_occs) = lookupOccs body_usg arg_bndrs' ; return (RI { ri_rhs_usg = rhs_usg , ri_fn = bndr, ri_new_rhs = mkLams arg_bndrs' body' , ri_lam_bndrs = arg_bndrs, ri_lam_body = body - , ri_arg_occs = arg_occs }) } + , ri_arg_occs = arg_occs }, body_ws) } -- The arg_occs says how the visible, -- lambda-bound binders of the RHS are used -- (including the TyVar binders) @@ -1757,7 +1783,7 @@ initSpecInfo (RI { ri_rhs_usg = rhs_usg }) specNonRec :: ScEnv -> CallEnv -- Calls in body -> RhsInfo -- Structure info usage info for un-specialised RHS - -> UniqSM (ScUsage, SpecInfo) -- Usage from RHSs (specialised and not) + -> UniqSM (ScUsage, SpecInfo, [SpecFailWarning]) -- Usage from RHSs (specialised and not) -- plus details of specialisations specNonRec env body_calls rhs_info @@ -1767,11 +1793,12 @@ specNonRec env body_calls rhs_info specRec :: ScEnv -> CallEnv -- Calls in body -> [RhsInfo] -- Structure info and usage info for un-specialised RHSs - -> UniqSM (ScUsage, [SpecInfo]) -- Usage from all RHSs (specialised and not) + -> UniqSM (ScUsage, [SpecInfo], SpecFailWarnings) + -- Usage from all RHSs (specialised and not) -- plus details of specialisations specRec env body_calls rhs_infos - = go 1 body_calls nullUsage (map initSpecInfo rhs_infos) + = go 1 body_calls nullUsage (map initSpecInfo rhs_infos) [] -- body_calls: see Note [Seeding recursive groups] -- NB: 'go' always calls 'specialise' once, which in turn unleashes -- si_mb_unspec if there are any boring calls in body_calls, @@ -1786,23 +1813,25 @@ specRec env body_calls rhs_infos -- Two accumulating parameters: -> ScUsage -- Usage from earlier specialisations -> [SpecInfo] -- Details of specialisations so far - -> UniqSM (ScUsage, [SpecInfo]) - go n_iter seed_calls usg_so_far spec_infos + -> SpecFailWarnings -- Warnings so far + -> UniqSM (ScUsage, [SpecInfo], SpecFailWarnings) + go n_iter seed_calls usg_so_far spec_infos ws_so_far = -- pprTrace "specRec3" (vcat [ text "bndrs" <+> ppr (map ri_fn rhs_infos) -- , text "iteration" <+> int n_iter -- , text "spec_infos" <+> ppr (map (map os_pat . si_specs) spec_infos) -- ]) $ do { specs_w_usg <- zipWithM (specialise env seed_calls) rhs_infos spec_infos - ; let (extra_usg_s, all_spec_infos) = unzip specs_w_usg + + ; let (extra_usg_s, all_spec_infos, extra_ws ) = unzip3 specs_w_usg extra_usg = combineUsages extra_usg_s all_usg = usg_so_far `combineUsage` extra_usg new_calls = scu_calls extra_usg - ; go_again n_iter new_calls all_usg all_spec_infos } + ; go_again n_iter new_calls all_usg all_spec_infos (ws_so_far ++ concat extra_ws) } -- go_again deals with termination - go_again n_iter seed_calls usg_so_far spec_infos + go_again n_iter seed_calls usg_so_far spec_infos ws_so_far | isEmptyVarEnv seed_calls - = return (usg_so_far, spec_infos) + = return (usg_so_far, spec_infos, ws_so_far) -- Limit recursive specialisation -- See Note [Limit recursive specialisation] @@ -1816,10 +1845,10 @@ specRec env body_calls rhs_infos -- for the unspecialised function, since it may now be called -- pprTrace "specRec2" (ppr (map (map os_pat . si_specs) spec_infos)) $ let rhs_usgs = combineUsages (mapMaybe si_mb_unspec spec_infos) - in return (usg_so_far `combineUsage` rhs_usgs, spec_infos) + in return (usg_so_far `combineUsage` rhs_usgs, spec_infos, ws_so_far) | otherwise - = go (n_iter + 1) seed_calls usg_so_far spec_infos + = go (n_iter + 1) seed_calls usg_so_far spec_infos ws_so_far -- See Note [Limit recursive specialisation] the_limit = case sc_count opts of @@ -1832,7 +1861,7 @@ specialise -> CallEnv -- Info on newly-discovered calls to this function -> RhsInfo -> SpecInfo -- Original RHS plus patterns dealt with - -> UniqSM (ScUsage, SpecInfo) -- New specialised versions and their usage + -> UniqSM (ScUsage, SpecInfo, [SpecFailWarning]) -- New specialised versions and their usage -- See Note [spec_usg includes rhs_usg] @@ -1850,7 +1879,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs | isDeadEndId fn -- Note [Do not specialise diverging functions] -- /and/ do not generate specialisation seeds from its RHS = -- pprTrace "specialise bot" (ppr fn) $ - return (nullUsage, spec_info) + return (nullUsage, spec_info, []) | not (isNeverActive (idInlineActivation fn)) -- See Note [Transfer activation] @@ -1861,7 +1890,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs , not (null arg_bndrs) -- Only specialise functions , Just all_calls <- lookupVarEnv bind_calls fn -- Some calls to it = -- pprTrace "specialise entry {" (ppr fn <+> ppr all_calls) $ - do { (boring_call, pats_discarded, new_pats) + do { (boring_call, pats_discarded, new_pats, warnings) <- callsToNewPats env fn spec_info arg_occs all_calls ; let n_pats = length new_pats @@ -1876,7 +1905,7 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs -- , text "new_pats" <+> ppr new_pats]) ; let spec_env = decreaseSpecCount env n_pats - ; (spec_usgs, new_specs) <- mapAndUnzipM (spec_one spec_env fn arg_bndrs body) + ; (spec_usgs, new_specs, new_wss) <- mapAndUnzip3M (spec_one spec_env fn arg_bndrs body) (new_pats `zip` [spec_count..]) -- See Note [Specialise original body] @@ -1900,15 +1929,16 @@ specialise env bind_calls (RI { ri_fn = fn, ri_lam_bndrs = arg_bndrs ; return (new_usg, SI { si_specs = new_specs ++ specs , si_n_specs = spec_count + n_pats - , si_mb_unspec = mb_unspec' }) } + , si_mb_unspec = mb_unspec' } + ,warnings ++ concat new_wss) } | otherwise -- No calls, inactive, or not a function -- Behave as if there was a single, boring call = -- pprTrace "specialise inactive" (ppr fn $$ ppr mb_unspec) $ case mb_unspec of -- Behave as if there was a single, boring call - Just rhs_usg -> return (rhs_usg, spec_info { si_mb_unspec = Nothing }) + Just rhs_usg -> return (rhs_usg, spec_info { si_mb_unspec = Nothing }, []) -- See Note [spec_usg includes rhs_usg] - Nothing -> return (nullUsage, spec_info) + Nothing -> return (nullUsage, spec_info, []) --------------------- @@ -1917,7 +1947,7 @@ spec_one :: ScEnv -> [InVar] -- Lambda-binders of RHS; should match patterns -> InExpr -- Body of the original function -> (CallPat, Int) - -> UniqSM (ScUsage, OneSpec) -- Rule and binding + -> UniqSM (ScUsage, OneSpec, SpecFailWarnings) -- Rule and binding, warnings if any -- spec_one creates a specialised copy of the function, together -- with a rule for using it. I'm very proud of how short this @@ -1969,7 +1999,7 @@ spec_one env fn arg_bndrs body (call_pat, rule_number) -- Specialise the body -- ; pprTraceM "body_subst_for" $ ppr (spec_occ) $$ ppr (sc_subst body_env) - ; (spec_usg, spec_body) <- scExpr body_env body + ; (spec_usg, spec_body, body_warnings) <- scExpr body_env body -- And build the results ; (qvars', pats') <- generaliseDictPats qvars pats @@ -2018,7 +2048,7 @@ spec_one env fn arg_bndrs body (call_pat, rule_number) -- ] ; return (spec_usg, OS { os_pat = call_pat, os_rule = rule , os_id = spec_id - , os_rhs = spec_rhs }) } + , os_rhs = spec_rhs }, body_warnings) } generaliseDictPats :: [Var] -> [CoreExpr] -- Quantified vars and pats -> UniqSM ([Var], [CoreExpr]) -- New quantified vars and pats @@ -2402,12 +2432,26 @@ instance Outputable CallPat where , text "cp_args =" <+> ppr args , text "cp_strict_args = " <> ppr strict ]) +newtype SpecFailWarning = SpecFailForcedArgCount { spec_failed_fun_name :: Name } + +type SpecFailWarnings = [SpecFailWarning] + +instance Outputable SpecFailWarning where + ppr (SpecFailForcedArgCount name) = ppr name <+> pprDefinedAt name + +combineSpecWarning :: SpecFailWarnings -> SpecFailWarnings -> SpecFailWarnings +combineSpecWarning = (++) + +data ArgCountResult = WorkerSmallEnough | WorkerTooLarge | WorkerTooLargeForced Name + callsToNewPats :: ScEnv -> Id -> SpecInfo -> [ArgOcc] -> [Call] -> UniqSM ( Bool -- At least one boring call , Bool -- Patterns were discarded - , [CallPat] ) -- Patterns to specialise + , [CallPat] -- Patterns to specialise + , [SpecFailWarning] -- Things that didn't specialise we want to warn the user about) + ) -- Result has no duplicate patterns, -- nor ones mentioned in si_specs (hence "new" patterns) -- Bool indicates that there was at least one boring pattern @@ -2433,12 +2477,18 @@ callsToNewPats env fn spec_info@(SI { si_specs = done_specs }) bndr_occs calls non_dups = subsumePats in_scope new_pats -- Remove ones that have too many worker variables - small_pats = filterOut too_many_worker_args non_dups + (small_pats, arg_count_warnings) = partitionByWorkerSize too_many_worker_args non_dups - too_many_worker_args _ - | sc_force env = False -- See (FS5) of Note [Forcing specialisation] + -- too_many_worker_args :: CallPat -> Either SpecFailWarning Bool too_many_worker_args (CP { cp_qvars = vars, cp_args = args }) - = not (isWorkerSmallEnough (sc_max_args $ sc_opts env) (valArgCount args) vars) + | sc_force env + -- See (FS5) of Note [Forcing specialisation] + = if (isWorkerSmallEnough (sc_max_forced_args $ sc_opts env) (valArgCount args) vars) + then WorkerSmallEnough + else WorkerTooLargeForced (idName fn) + | (isWorkerSmallEnough (sc_max_args $ sc_opts env) (valArgCount args) vars) + = WorkerSmallEnough + | otherwise = WorkerTooLarge -- We are about to construct w/w pair in 'spec_one'. -- Omit specialisation leading to high arity workers. -- See Note [Limit w/w arity] in GHC.Core.Opt.WorkWrap.Utils @@ -2454,10 +2504,21 @@ callsToNewPats env fn spec_info@(SI { si_specs = done_specs }) bndr_occs calls -- , text "done_specs:" <+> ppr (map os_pat done_specs) -- , text "trimmed_pats:" <+> ppr trimmed_pats ]) - ; return (have_boring_call, pats_were_discarded, trimmed_pats) } + ; return (have_boring_call, pats_were_discarded, trimmed_pats, arg_count_warnings) } -- If any of the calls does not give rise to a specialisation, either -- because it is boring, or because there are too many specialisations, -- return a flag to say so, so that we know to keep the original function. + where + partitionByWorkerSize worker_size pats = go pats [] [] + where + go [] small warnings = (small, warnings) + go (p:ps) small warnings + | WorkerSmallEnough <- worker_size p + = go ps (p:small) warnings + | WorkerTooLarge <- worker_size p + = go ps small warnings + | WorkerTooLargeForced name <- worker_size p + = go ps small (SpecFailForcedArgCount name : warnings) trim_pats :: ScEnv -> Id -> SpecInfo -> [CallPat] -> (Bool, [CallPat]) ===================================== compiler/GHC/Driver/DynFlags.hs ===================================== @@ -394,6 +394,7 @@ data DynFlags = DynFlags { unfoldingOpts :: !UnfoldingOpts, maxWorkerArgs :: Int, + maxForcedSpecArgs :: Int, ghciHistSize :: Int, @@ -676,6 +677,8 @@ defaultDynFlags mySettings = unfoldingOpts = defaultUnfoldingOpts, maxWorkerArgs = 10, + maxForcedSpecArgs = 333, + -- 333 is fairly arbitrary, see Note [Forcing specialisation]:FS5 ghciHistSize = 50, -- keep a log of length 50 by default ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -1820,6 +1820,8 @@ dynamic_flags_deps = [ , make_ord_flag defFlag "fmax-worker-args" (intSuffix (\n d -> d {maxWorkerArgs = n})) + , make_ord_flag defFlag "fmax-forced-spec-args" + (intSuffix (\n d -> d {maxForcedSpecArgs = n})) , make_ord_flag defGhciFlag "fghci-hist-size" (intSuffix (\n d -> d {ghciHistSize = n})) , make_ord_flag defGhcFlag "fmax-inline-alloc-size" ===================================== docs/users_guide/using-optimisation.rst ===================================== @@ -870,6 +870,21 @@ as such you shouldn't need to set any of them explicitly. A flag value arguments of the resulting worker exceeds both that of the original function and this setting. +.. ghc-flag:: -fmax-forced-spec-args=⟨n⟩ + :shortdesc: *default: 333.* Maximum number of value arguments for forced SpecConstr specializations. + :type: dynamic + :category: + + :default: 512 + + When using ``SPEC`` from ``GHC.Types`` to force SpecConstr to fire on a function + sometimes this can result in functions taking a ridicolously large number of arguments + resulting a very large compile time hits for minor performance benefits. + + Since this is usually unintended we prevent SpecConstr from firing and generate + a warning if the number of arguments in the resulting function would exceed + the value given by ``-fmax-forced-spec-args``. + .. ghc-flag:: -fno-opt-coercion :shortdesc: Turn off the coercion optimiser :type: dynamic ===================================== testsuite/tests/simplCore/should_compile/T25197.hs ===================================== @@ -0,0 +1,30 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25197 where + +import T25197_TH +import GHC.Exts + +{- +This test applies a large statically known data structure to a function with +a SPEC argument, forcing the function to be specialised for the argument. +However when the complete structure of the argument is not statically known, +or as here the leaves of the structures are primitive literals for which we do +not specialize this results in a specialized function that can take hundreds of +arguments. + +Typically this is not intended, therefore we use a limit on the number of +arguments for specializations. As at some point this sort of specialization +comes with a heavy compile time cost. However we allow users to specify this +limit just in case they really depend on this sort of specialization. +-} + +foo :: [a] -> Int +foo = go SPEC + where + go :: SPEC -> [a] -> Int + go s [] = s `seq` 0 + go s (_:xs) = 1 + go s xs + +main :: IO () +main = print $ foo $(gen 1000) ===================================== testsuite/tests/simplCore/should_compile/T25197.stderr ===================================== @@ -0,0 +1,6 @@ +T25197.hs: warning: + SpecConstr encountered one or more function(s) with a SPEC argument that resulted in too many arguments, + which resulted in no specialization being generated for these functions: + $wgo Defined at T25197.hs:26:5 + If this is expected you might want to increase -fmax-forced-spec-args to force specialization anyway. + ===================================== testsuite/tests/simplCore/should_compile/T25197_TH.hs ===================================== @@ -0,0 +1,9 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T25197_TH where + +import Language.Haskell.TH.Syntax + +gen :: Int -> Q Exp +gen 0 = [| [] |] +gen n = [| $(lift (show n)) : $(gen (n-1)) |] ===================================== testsuite/tests/simplCore/should_compile/all.T ===================================== @@ -530,3 +530,4 @@ test('T24625', [ grep_errmsg(r'case lazy') ], compile, ['-O -fno-ignore-asserts test('T24725a', [ grep_errmsg(r'testedRule')], compile, ['-O -ddump-rule-firings']) test('T25033', normal, compile, ['-O']) test('T25160', normal, compile, ['-O -ddump-rules']) +test('T25197', [req_th, extra_files(["T25197_TH.hs"]), only_ways(['optasm'])], multimod_compile, ['T25197', '-O2 -v0']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5e228031cbb0292adcea66ecf5ca3c0b59f9c2be -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5e228031cbb0292adcea66ecf5ca3c0b59f9c2be You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 16:24:19 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Fri, 27 Sep 2024 12:24:19 -0400 Subject: [Git][ghc/ghc][wip/T23479] Add tracing capabilities for JS Modules Message-ID: <66f6dc332b4e3_f7e92c1629086540@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: 5f13cd11 by Serge S. Gulin at 2024-09-27T19:23:48+03:00 Add tracing capabilities for JS Modules - - - - - 3 changed files: - compiler/GHC/StgToJS/Linker/Linker.hs - compiler/GHC/StgToJS/Types.hs - compiler/GHC/Types/ForeignCall.hs Changes: ===================================== compiler/GHC/StgToJS/Linker/Linker.hs ===================================== @@ -548,6 +548,16 @@ data ModuleCode = ModuleCode , mc_frefs :: ![ForeignJSRef] } +instance Outputable ModuleCode where + ppr m = hang (text "ModuleCode") 2 $ vcat + [ hcat [text "Module: ", ppr (mc_module m)] + , hcat [text "JS Code:", pretty True (mc_js_code m)] + , hcat [text "JS Exports:", pprHsBytes (mc_exports m)] + , hang (text "JS Closures::") 2 (vcat (fmap (text . show) (mc_closures m))) + , hang (text "JS Statics::") 2 (vcat (fmap (text . show) (mc_statics m))) + , hang (text "JS ForeignRefs::") 2 (vcat (fmap (text . show) (mc_frefs m))) + ] + -- | ModuleCode after link with other modules. -- -- It contains less information than ModuleCode because they have been commoned ===================================== compiler/GHC/StgToJS/Types.hs ===================================== @@ -297,6 +297,7 @@ data ForeignJSRef = ForeignJSRef , foreignRefArgs :: ![FastString] , foreignRefResult :: !FastString } + deriving (Show) -- | data used to generate one ObjBlock in our object file data LinkableUnit = LinkableUnit ===================================== compiler/GHC/Types/ForeignCall.hs ===================================== @@ -156,7 +156,7 @@ data CCallConv | StdCallConv | PrimCallConv | JavaScriptCallConv - deriving (Eq, Data, Enum) + deriving (Show, Eq, Data, Enum) instance Outputable CCallConv where ppr StdCallConv = text "stdcall" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5f13cd110033aabb88a3d7fa3a195e560c4a2aaa -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5f13cd110033aabb88a3d7fa3a195e560c4a2aaa You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 16:32:36 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 12:32:36 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/happy-version Message-ID: <66f6de245bbdc_f7e92c91b988856a@gitlab.mail> Matthew Pickering pushed new branch wip/happy-version at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/happy-version You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 16:36:59 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 12:36:59 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 8 commits: hadrian: Refactor system-cxx-std-lib rules0 Message-ID: <66f6df2b60660_f7e92e9dfa492269@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: e146b97f by Matthew Pickering at 2024-09-27T17:36:41+01:00 hadrian: Refactor system-cxx-std-lib rules0 I noticed a few things wrong with the hadrian rules for `system-cxx-std-lib` rules. * For `text` there is an ad-hoc check to depend on `system-cxx-std-lib` outside of `configurePackage`. * The `system-cxx-std-lib` dependency is not read from cabal files. * Recache is not called on the packge database after the `.conf` file is generated, a more natural place for this rule is `registerRules`. Treating this uniformly like other packages is complicated by it not having any source code or a cabal file. However we can do a bit better by reporting the dependency firstly in `PackageData` and then needing the `.conf` file in the same place as every other package in `configurePackage`. Fixes #25303 - - - - - 9908da66 by Matthew Pickering at 2024-09-27T17:36:41+01:00 fixes for simdutf8 - - - - - f4ec9b83 by Matthew Pickering at 2024-09-27T17:36:41+01:00 use building for target in llvm flavour transformer - - - - - b4e3a77b by Matthew Pickering at 2024-09-27T17:36:41+01:00 bindist: Pass path to package database we want to recache This fixes recaching on cross compilers - - - - - a4177c90 by Matthew Pickering at 2024-09-27T17:36:41+01:00 testsuite: T9930fail now passes on javascript I didn't investigate why, but the comment says it should be fixed by building a stage2 cross compiler (and it is). - - - - - 73f911cf by Sven Tennie at 2024-09-27T17:36:41+01:00 ci: RISCV64 cross-compile testing This adds a validation job which tests that we can build a riscv64 cross compiler and build a simple program using it. We do not currently run the whole testsuite. Towards #25254 Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - af47a83a by Matthew Pickering at 2024-09-27T17:36:41+01:00 Bump LLVM upper bound to allow LLVM 19 Also bumps the ci-images commit so that the deb12 images uses LLVM 19 for testing. ------------------------- Metric Decrease: size_hello_artifact_gzip size_hello_unicode_gzip ------------------------- Fixes #25295 - - - - - a22ee5c4 by Matthew Pickering at 2024-09-27T17:36:42+01:00 docker rev - - - - - 15 changed files: - .gitlab-ci.yml - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - configure.ac - hadrian/src/BindistConfig.hs - hadrian/src/Flavour.hs - hadrian/src/Flavour/Type.hs - hadrian/src/Hadrian/Haskell/Cabal/Parse.hs - hadrian/src/Hadrian/Haskell/Cabal/Type.hs - hadrian/src/Rules/BinaryDist.hs - hadrian/src/Rules/Generate.hs - hadrian/src/Rules/Register.hs - hadrian/src/Settings/Default.hs - hadrian/src/Settings/Packages.hs - testsuite/tests/ghc-e/should_fail/all.T Changes: ===================================== .gitlab-ci.yml ===================================== @@ -2,7 +2,7 @@ variables: GIT_SSL_NO_VERIFY: "1" # Commit of ghc/ci-images repository from which to pull Docker images - DOCKER_REV: 486541129a8e7bf77c2cf7cd76ca998f690d5685 + DOCKER_REV: 6efac743853f9c2172777e934d7aea44434415ec # Sequential version number of all cached things. # Bump to invalidate GitLab CI cache. ===================================== .gitlab/generate-ci/gen_ci.hs ===================================== @@ -106,6 +106,7 @@ data Opsys data LinuxDistro = Debian12 + | Debian12Riscv | Debian11 | Debian11Js | Debian10 @@ -305,6 +306,7 @@ distroName :: LinuxDistro -> String distroName Debian12 = "deb12" distroName Debian11 = "deb11" distroName Debian11Js = "deb11-emsdk-closure" +distroName Debian12Riscv = "deb12-riscv" distroName Debian10 = "deb10" distroName Debian9 = "deb9" distroName Fedora33 = "fedora33" @@ -627,6 +629,7 @@ data ValidateRule = FullCI -- ^ Run this job when the "full-ci" label is present. | LLVMBackend -- ^ Run this job when the "LLVM backend" label is present | JSBackend -- ^ Run this job when the "javascript" label is present + | RiscV -- ^ Run this job when the "RISC-V" label is present | WasmBackend -- ^ Run this job when the "wasm" label is present | FreeBSDLabel -- ^ Run this job when the "FreeBSD" label is set. | NonmovingGc -- ^ Run this job when the "non-moving GC" label is set. @@ -674,6 +677,7 @@ validateRuleString FullCI = or_all ([ labelString "full-ci" validateRuleString LLVMBackend = labelString "LLVM backend" validateRuleString JSBackend = labelString "javascript" +validateRuleString RiscV = labelString "RISC-V" validateRuleString WasmBackend = labelString "wasm" validateRuleString FreeBSDLabel = labelString "FreeBSD" validateRuleString NonmovingGc = labelString "non-moving GC" @@ -1123,6 +1127,9 @@ cross_jobs = [ -- x86 -> aarch64 validateBuilds Amd64 (Linux Debian11) (crossConfig "aarch64-linux-gnu" (Emulator "qemu-aarch64 -L /usr/aarch64-linux-gnu") Nothing) + -- x86_64 -> riscv + , addValidateRule RiscV (validateBuilds Amd64 (Linux Debian12Riscv) (crossConfig "riscv64-linux-gnu" (Emulator "qemu-riscv64 -L /usr/riscv64-linux-gnu") Nothing)) + -- Javascript , addValidateRule JSBackend (validateBuilds Amd64 (Linux Debian11Js) javascriptConfig) ===================================== .gitlab/jobs.yaml ===================================== @@ -1855,6 +1855,71 @@ "XZ_OPT": "-9" } }, + "nightly-x86_64-linux-deb12-riscv-cross_riscv64-linux-gnu-validate": { + "after_script": [ + ".gitlab/ci.sh save_cache", + ".gitlab/ci.sh save_test_output", + ".gitlab/ci.sh clean", + "cat ci_timings" + ], + "allow_failure": false, + "artifacts": { + "expire_in": "8 weeks", + "paths": [ + "ghc-x86_64-linux-deb12-riscv-cross_riscv64-linux-gnu-validate.tar.xz", + "junit.xml", + "unexpected-test-output.tar.gz" + ], + "reports": { + "junit": "junit.xml" + }, + "when": "always" + }, + "cache": { + "key": "x86_64-linux-deb12-riscv-$CACHE_REV", + "paths": [ + "cabal-cache", + "toolchain" + ] + }, + "dependencies": [], + "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12-riscv:$DOCKER_REV", + "needs": [ + { + "artifacts": false, + "job": "hadrian-ghc-in-ghci" + } + ], + "rules": [ + { + "if": "(\"true\" == \"true\") && ($RELEASE_JOB != \"yes\") && ($NIGHTLY)", + "when": "on_success" + } + ], + "script": [ + "sudo chown ghc:ghc -R .", + ".gitlab/ci.sh setup", + ".gitlab/ci.sh configure", + ".gitlab/ci.sh build_hadrian", + ".gitlab/ci.sh test_hadrian" + ], + "stage": "full-build", + "tags": [ + "x86_64-linux" + ], + "variables": { + "BIGNUM_BACKEND": "gmp", + "BIN_DIST_NAME": "ghc-x86_64-linux-deb12-riscv-cross_riscv64-linux-gnu-validate", + "BUILD_FLAVOUR": "validate", + "CONFIGURE_ARGS": "--with-intree-gmp --enable-strict-ghc-toolchain-check", + "CROSS_EMULATOR": "qemu-riscv64 -L /usr/riscv64-linux-gnu", + "CROSS_TARGET": "riscv64-linux-gnu", + "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", + "RUNTEST_ARGS": "", + "TEST_ENV": "x86_64-linux-deb12-riscv-cross_riscv64-linux-gnu-validate", + "XZ_OPT": "-9" + } + }, "nightly-x86_64-linux-deb12-unreg-validate": { "after_script": [ ".gitlab/ci.sh save_cache", @@ -5222,6 +5287,70 @@ "TEST_ENV": "x86_64-linux-deb12-numa-slow-validate" } }, + "x86_64-linux-deb12-riscv-cross_riscv64-linux-gnu-validate": { + "after_script": [ + ".gitlab/ci.sh save_cache", + ".gitlab/ci.sh save_test_output", + ".gitlab/ci.sh clean", + "cat ci_timings" + ], + "allow_failure": false, + "artifacts": { + "expire_in": "2 weeks", + "paths": [ + "ghc-x86_64-linux-deb12-riscv-cross_riscv64-linux-gnu-validate.tar.xz", + "junit.xml", + "unexpected-test-output.tar.gz" + ], + "reports": { + "junit": "junit.xml" + }, + "when": "always" + }, + "cache": { + "key": "x86_64-linux-deb12-riscv-$CACHE_REV", + "paths": [ + "cabal-cache", + "toolchain" + ] + }, + "dependencies": [], + "image": "registry.gitlab.haskell.org/ghc/ci-images/x86_64-linux-deb12-riscv:$DOCKER_REV", + "needs": [ + { + "artifacts": false, + "job": "hadrian-ghc-in-ghci" + } + ], + "rules": [ + { + "if": "((($CI_MERGE_REQUEST_LABELS =~ /.*full-ci.*/) || ($CI_MERGE_REQUEST_LABELS =~ /.*marge_bot_batch_merge_job.*/) || ($CI_COMMIT_BRANCH == \"master\") || ($CI_COMMIT_BRANCH =~ /ghc-[0-9]+\\.[0-9]+/)) || ($CI_MERGE_REQUEST_LABELS =~ /.*RISC-V.*/)) && ($RELEASE_JOB != \"yes\") && ($NIGHTLY == null)", + "when": "on_success" + } + ], + "script": [ + "sudo chown ghc:ghc -R .", + ".gitlab/ci.sh setup", + ".gitlab/ci.sh configure", + ".gitlab/ci.sh build_hadrian", + ".gitlab/ci.sh test_hadrian" + ], + "stage": "full-build", + "tags": [ + "x86_64-linux" + ], + "variables": { + "BIGNUM_BACKEND": "gmp", + "BIN_DIST_NAME": "ghc-x86_64-linux-deb12-riscv-cross_riscv64-linux-gnu-validate", + "BUILD_FLAVOUR": "validate", + "CONFIGURE_ARGS": "--with-intree-gmp --enable-strict-ghc-toolchain-check", + "CROSS_EMULATOR": "qemu-riscv64 -L /usr/riscv64-linux-gnu", + "CROSS_TARGET": "riscv64-linux-gnu", + "INSTALL_CONFIGURE_ARGS": "--enable-strict-ghc-toolchain-check", + "RUNTEST_ARGS": "", + "TEST_ENV": "x86_64-linux-deb12-riscv-cross_riscv64-linux-gnu-validate" + } + }, "x86_64-linux-deb12-unreg-validate": { "after_script": [ ".gitlab/ci.sh save_cache", ===================================== configure.ac ===================================== @@ -561,7 +561,7 @@ AC_SUBST(InstallNameToolCmd) # versions of LLVM simultaneously, but that stopped working around # 3.5/3.6 release of LLVM. LlvmMinVersion=13 # inclusive -LlvmMaxVersion=19 # not inclusive +LlvmMaxVersion=20 # not inclusive AC_SUBST([LlvmMinVersion]) AC_SUBST([LlvmMaxVersion]) ===================================== hadrian/src/BindistConfig.hs ===================================== @@ -33,3 +33,9 @@ implicitBindistConfig = do -- the target. cross <- flag CrossCompiling return $ if cross then crossBindist else normalBindist + +-- | Are we building things in this stage for the final target? +buildingForTarget :: Stage -> Action Bool +buildingForTarget st = do + cfg <- implicitBindistConfig + return $ st >= (library_stage cfg) ===================================== hadrian/src/Flavour.hs ===================================== @@ -39,6 +39,7 @@ import Text.Parsec.Combinator as P import Text.Parsec.Char as P import Control.Monad.Except import UserSettings +import BindistConfig flavourTransformers :: Map String (Flavour -> Flavour) @@ -239,9 +240,9 @@ enableThreadSanitizer instrumentCmm = addArgs $ notStage0 ? mconcat ] ] --- | Use the LLVM backend in stages 1 and later. +-- | Use the LLVM backend in target stages viaLlvmBackend :: Flavour -> Flavour -viaLlvmBackend = addArgs $ notStage0 ? builder Ghc ? arg "-fllvm" +viaLlvmBackend = addArgs $ staged buildingForTarget ? builder Ghc ? arg "-fllvm" -- | Build the GHC executable with profiling enabled in stages 2 and later. It -- is also recommended that you use this with @'dynamicGhcPrograms' = False@ @@ -301,7 +302,7 @@ useNativeBignum flavour = -- | Enable building the @text@ package with @simdutf@ support. enableTextWithSIMDUTF :: Flavour -> Flavour enableTextWithSIMDUTF flavour = flavour { - textWithSIMDUTF = (>= Stage2) + textWithSIMDUTF = buildingForTarget } -- | Build stage2 compiler with -fomit-interface-pragmas to reduce ===================================== hadrian/src/Flavour/Type.hs ===================================== @@ -26,7 +26,7 @@ data Flavour = Flavour { -- | Build the @text@ package with @simdutf@ support. Disabled by -- default due to packaging difficulties described in #20724. textWithSIMDUTF :: Stage -- ^ stage of the /built/ compiler - -> Bool, + -> Action Bool, -- | Build libraries these ways. libraryWays :: Ways, -- | Build RTS these ways. ===================================== hadrian/src/Hadrian/Haskell/Cabal/Parse.hs ===================================== @@ -94,10 +94,11 @@ parsePackageData pkg = do sorted = sort [ C.unPackageName p | C.Dependency p _ _ <- allDeps ] deps = nubOrd sorted \\ [name] depPkgs = mapMaybe findPackageByName deps + cxxStdLib = elem "system-cxx-std-lib" deps return $ PackageData name version (C.fromShortText (C.synopsis pd)) (C.fromShortText (C.description pd)) - depPkgs gpd + depPkgs cxxStdLib gpd where -- Collect an overapproximation of dependencies by ignoring conditionals collectDeps :: Maybe (C.CondTree v [C.Dependency] a) -> [C.Dependency] @@ -151,7 +152,9 @@ configurePackage :: Context -> Action () configurePackage context at Context {..} = do putProgressInfo $ "| Configure package " ++ quote (pkgName package) gpd <- pkgGenericDescription package - depPkgs <- packageDependencies <$> readPackageData package + pd <- readPackageData package + let depPkgs = packageDependencies pd + needSystemCxxStdLib = dependsOnSystemCxxStdLib pd -- Stage packages are those we have in this stage. stagePkgs <- stagePackages stage @@ -170,7 +173,12 @@ configurePackage context at Context {..} = do -- We'll need those packages in our package database. deps <- sequence [ pkgConfFile (context { package = pkg, iplace = forceBaseAfterGhcInternal pkg }) | pkg <- depPkgs, pkg `elem` stagePkgs ] - need $ extraPreConfigureDeps ++ deps + -- system-cxx-std-lib is magic.. it doesn't have a cabal file or source code, so we have + -- to treat it specially as `pkgConfFile` uses `readPackageData` to compute the version. + systemCxxStdLib <- sequence [ systemCxxStdLibConfPath (PackageDbLoc stage iplace) | needSystemCxxStdLib ] + need $ extraPreConfigureDeps + ++ deps + ++ systemCxxStdLib -- Figure out what hooks we need. let configureFile = replaceFileName (pkgCabalFile package) "configure" ===================================== hadrian/src/Hadrian/Haskell/Cabal/Type.hs ===================================== @@ -30,6 +30,7 @@ data PackageData = PackageData , synopsis :: String , description :: String , packageDependencies :: [Package] + , dependsOnSystemCxxStdLib :: Bool , genericPackageDescription :: GenericPackageDescription } deriving (Eq, Generic, Show, Typeable) ===================================== hadrian/src/Rules/BinaryDist.hs ===================================== @@ -227,8 +227,8 @@ buildBinDistDir root conf at BindistConfig{..} = do -- (c.f. #20267). -- Not going to work for cross - --ghcPkgName <- programName (vanillaContext Stage1 ghcPkg) - --cmd_ (bindistFilesDir -/- "bin" -/- ghcPkgName) ["recache"] + ghcPkgName <- programName (vanillaContext Stage1 ghcPkg) + cmd_ (bindistFilesDir -/- "bin" -/- ghcPkgName) ["recache", "--package-db", bindistFilesDir -/- "lib" -/- "package.conf.d" ] need ["docs"] ===================================== hadrian/src/Rules/Generate.hs ===================================== @@ -242,9 +242,6 @@ copyRules = do prefix -/- "html/**" <~ return "utils/haddock/haddock-api/resources" prefix -/- "latex/**" <~ return "utils/haddock/haddock-api/resources" - forM_ [Inplace, Final] $ \iplace -> - root -/- relativePackageDbPath (PackageDbLoc stage iplace) -/- systemCxxStdLibConf %> \file -> do - copyFile ("mk" -/- "system-cxx-std-lib-1.0.conf") file generateRules :: Rules () generateRules = do ===================================== hadrian/src/Rules/Register.hs ===================================== @@ -7,7 +7,6 @@ module Rules.Register ( import Base import Context import Expression ( getContextData ) -import Flavour import Oracles.Setting import Hadrian.BuildPath import Hadrian.Expression @@ -52,14 +51,6 @@ configurePackageRules = do isGmp <- (== "gmp") <$> interpretInContext ctx getBignumBackend when isGmp $ need [buildP -/- "include/ghc-gmp.h"] - when (pkg == text) $ do - simdutf <- textWithSIMDUTF <$> flavour - when simdutf $ do - -- This is required, otherwise you get Error: hadrian: - -- Encountered missing or private dependencies: - -- system-cxx-std-lib ==1.0 - cxxStdLib <- systemCxxStdLibConfPath $ PackageDbLoc stage Inplace - need [cxxStdLib] Cabal.configurePackage ctx root -/- "**/autogen/cabal_macros.h" %> \out -> do @@ -114,6 +105,12 @@ registerPackageRules rs stage iplace = do target (Context stage compiler vanilla iplace) (GhcPkg Recache stage) [] [] writeFileLines stamp [] + -- Special rule for registering system-cxx-std-lib + root -/- relativePackageDbPath (PackageDbLoc stage iplace) -/- systemCxxStdLibConf %> \file -> do + copyFile ("mk" -/- "system-cxx-std-lib-1.0.conf") file + buildWithResources rs $ + target (Context stage compiler vanilla iplace) (GhcPkg Recache stage) [] [] + -- Register a package. root -/- relativePackageDbPath (PackageDbLoc stage iplace) -/- "*.conf" %> \conf -> do historyDisable ===================================== hadrian/src/Settings/Default.hs ===================================== @@ -275,7 +275,7 @@ defaultFlavour = Flavour , packages = defaultPackages , bignumBackend = defaultBignumBackend , bignumCheck = False - , textWithSIMDUTF = const False + , textWithSIMDUTF = const (return False) , libraryWays = defaultLibraryWays , rtsWays = defaultRtsWays , dynamicGhcPrograms = defaultDynamicGhcPrograms ===================================== hadrian/src/Settings/Packages.hs ===================================== @@ -192,7 +192,7 @@ packageArgs = do ---------------------------------- text -------------------------------- , package text ? - ifM (compilerStageOption textWithSIMDUTF) + ifM (staged =<< expr (textWithSIMDUTF <$> flavour)) (builder (Cabal Flags) ? arg "+simdutf") (builder (Cabal Flags) ? arg "-simdutf") ===================================== testsuite/tests/ghc-e/should_fail/all.T ===================================== @@ -14,10 +14,8 @@ test('ghc-e-fail2', req_interp, makefile_test, ['ghc-e-fail2']) # and no failure is induced. test('T9930fail', [extra_files(['T9930']), - when(opsys('mingw32'), skip), - # broken for JS until cross-compilers become stage2 compilers (#19174) - # or until we bootstrap with a 9.10 compiler - js_broken(19174)], + when(opsys('mingw32'), skip) + ], makefile_test, ['T9930fail']) test('T18441fail0', req_interp, makefile_test, ['T18441fail0']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ed47556df729efe9608853c99bcee7da72e8a557...a22ee5c462ca1e2c9d08404b4d62181eb60e6dac -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/ed47556df729efe9608853c99bcee7da72e8a557...a22ee5c462ca1e2c9d08404b4d62181eb60e6dac You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 16:47:54 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 12:47:54 -0400 Subject: [Git][ghc/ghc][wip/happy-version] configure: Allow happy-2.0.2 Message-ID: <66f6e1bab45f9_f7e921035560101075@gitlab.mail> Matthew Pickering pushed to branch wip/happy-version at Glasgow Haskell Compiler / GHC Commits: 438c5b49 by Matthew Pickering at 2024-09-27T17:47:45+01:00 configure: Allow happy-2.0.2 happy-2.0.2 can be used to compile GHC. happy-2.0 and 2.0.1 have bugs which make it unsuitable to use. The version bound is now == 1.20.* || >= 2.0.2 && < 2.1 Fixes #25276 - - - - - 1 changed file: - m4/fptools_happy.m4 Changes: ===================================== m4/fptools_happy.m4 ===================================== @@ -24,10 +24,15 @@ changequote([, ])dnl ]) if test ! -f compiler/GHC/Parser.hs || test ! -f compiler/GHC/Cmm/Parser.hs then + failure_msg="Happy version == 1.20.* || >= 2.0.2 && < 2.1 is required to compile GHC" FP_COMPARE_VERSIONS([$fptools_cv_happy_version],[-lt],[1.20.0], - [AC_MSG_ERROR([Happy version 1.20 or later is required to compile GHC.])])[] + [AC_MSG_ERROR([$failure_msg])])[] FP_COMPARE_VERSIONS([$fptools_cv_happy_version],[-ge],[1.21.0], - [AC_MSG_ERROR([Happy version 1.20 or earlier is required to compile GHC.])])[] + FP_COMPARE_VERSIONS([$fptools_cv_happy_version], [-le], [2.0.1], + [AC_MSG_ERROR([$failure_msg])])[])[] + FP_COMPARE_VERSIONS([$fptools_cv_happy_version],[-ge],[2.1.0], + [AC_MSG_ERROR([$failure_msg])])[] + fi HappyVersion=$fptools_cv_happy_version; AC_SUBST(HappyVersion) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/438c5b49763578f36b017b9a20abd7e8e19a9b84 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/438c5b49763578f36b017b9a20abd7e8e19a9b84 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 17:03:43 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Fri, 27 Sep 2024 13:03:43 -0400 Subject: [Git][ghc/ghc][wip/supersven/riscv64-cross-ci] 51 commits: ci: Run abi-test on test-abi label Message-ID: <66f6e56fb8d4f_f7e9211b6df81030cd@gitlab.mail> Matthew Pickering pushed to branch wip/supersven/riscv64-cross-ci at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 6dfe9ba5 by Sven Tennie at 2024-09-27T18:03:29+01:00 ci: RISCV64 cross-compile testing This adds a validation job which tests that we can build a riscv64 cross compiler and build a simple program using it. We do not currently run the whole testsuite. Towards #25254 Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 21 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5fc49999c79e8193160090b7f680a204e0d8a5d5...6dfe9ba50d5097348c2d3c7ffdb1d9ad903c57a3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5fc49999c79e8193160090b7f680a204e0d8a5d5...6dfe9ba50d5097348c2d3c7ffdb1d9ad903c57a3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 17:05:39 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Fri, 27 Sep 2024 13:05:39 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/torsten.schmits/reexport-internal-test Message-ID: <66f6e5e3e7ad2_f7e9211b6dd0103567@gitlab.mail> Torsten Schmits pushed new branch wip/torsten.schmits/reexport-internal-test at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/torsten.schmits/reexport-internal-test You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 17:07:13 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Fri, 27 Sep 2024 13:07:13 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/reexport-internal-test] test case for a bug with reexported modules from internal libs Message-ID: <66f6e641e385c_f7e9211b6df810378b@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/reexport-internal-test at Glasgow Haskell Compiler / GHC Commits: 939af875 by Torsten Schmits at 2024-09-27T19:07:07+02:00 test case for a bug with reexported modules from internal libs - - - - - 3 changed files: - testsuite/tests/driver/T24384/Makefile - testsuite/tests/driver/T24384/all.T - + testsuite/tests/driver/T24384/setup-b.sh Changes: ===================================== testsuite/tests/driver/T24384/Makefile ===================================== @@ -8,3 +8,7 @@ T24384: mv B.hs lib/ '$(TEST_HC)' A.hs lib/B.hs C.hs D.hs E.hs F.hs F.hs-boot -M -dep-json dep.json -include-pkg-deps -include-cpp-deps -package-db ./db -hide-all-packages -package base -package containers -package template-haskell -package dep -package dep:pub cat dep.json + +T24384b: + ./setup-b.sh "$(TEST_HC)" "$(GHC_PKG)" + '$(TEST_HC)' InternalReexport.hs -v0 -M -dep-json dep.json -package-db ./db -hide-all-packages -package base -package dep ===================================== testsuite/tests/driver/T24384/all.T ===================================== @@ -16,3 +16,11 @@ test('T24384', ], makefile_test, []) + +test('T24384b', + [extra_files([ + 'setup-b.sh', + ]), + ], + makefile_test, + []) ===================================== testsuite/tests/driver/T24384/setup-b.sh ===================================== @@ -0,0 +1,76 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_pkg_cmd="$2" + +base="$PWD" +db="$base/db" +dep="$base/dep" +conf_dep="${dep}/dep.conf" +int="$base/int" +conf_int="${int}/int.conf" +lib_tag="ghc$($ghc_cmd --numeric-version)" + +ghc_pkg() +{ + eval "${ghc_pkg_cmd at Q} --no-user-package-db --package-db=${db at Q} $@" +} + +ghc() +{ + eval "${ghc_cmd at Q} $@" +} + +mkdir -p "$dep" "$int" "$db" + +cat > "$dep/DepA.hs" < "$int/DepB.hs" < "$conf_dep" <<'EOF' +name: dep +package-name: dep +version: 1.0 +id: dep-1.0-z +key: dep-1.0-z +depends: dep-1.0-int +exposed: True +exposed-modules: + DepA, + DepB from dep-1.0-int:DepB +import-dirs: ${pkgroot}/dep +library-dirs: ${pkgroot}/dep +hs-libraries: HSdep-1.0-z +EOF + +cat > "$conf_int" <<'EOF' +name: z-dep-z-int +package-name: dep +lib-name: int +version: 1.0 +id: dep-1.0-int +key: dep-1.0-int +exposed-modules: DepB +import-dirs: ${pkgroot}/int +library-dirs: ${pkgroot}/int +hs-libraries: HSdep-1.0-int +EOF + +ghc_pkg recache + +ghc "-v0 -package-db ${db at Q} -hidir ${int at Q} -O0 -this-unit-id dep-1.0-int -dynamic -shared ${int at Q}/DepB.hs -o ${int at Q}/libHSdep-1.0-int-${lib_tag}.so" +ghc "-v0 -package-db ${db at Q} -hidir ${dep at Q} -O0 -this-unit-id dep-1.0-z -dynamic -shared ${dep at Q}/DepA.hs -o ${dep at Q}/libHSdep-1.0-z-${lib_tag}.so" + +ghc_pkg -v0 register "${conf_int at Q}" +ghc_pkg -v0 register "${conf_dep at Q}" + +cat > InternalReexport.hs < From gitlab at gitlab.haskell.org Fri Sep 27 18:21:37 2024 From: gitlab at gitlab.haskell.org (Serge S. Gulin (@gulin.serge)) Date: Fri, 27 Sep 2024 14:21:37 -0400 Subject: [Git][ghc/ghc][wip/T23479] Add tracing capabilities for GlobalOcc's Message-ID: <66f6f7b1a276a_f7e921694ac8121441@gitlab.mail> Serge S. Gulin pushed to branch wip/T23479 at Glasgow Haskell Compiler / GHC Commits: 285413a0 by Serge S. Gulin at 2024-09-27T21:21:27+03:00 Add tracing capabilities for GlobalOcc's - - - - - 1 changed file: - compiler/GHC/StgToJS/Monad.hs Changes: ===================================== compiler/GHC/StgToJS/Monad.hs ===================================== @@ -31,6 +31,7 @@ import GHC.JS.Transform import GHC.StgToJS.Types import GHC.Unit.Module +import GHC.Utils.Outputable import GHC.Stg.Syntax import GHC.Types.SrcLoc @@ -159,6 +160,13 @@ data GlobalOcc = GlobalOcc , global_count :: !Word } +instance Outputable GlobalOcc where + ppr g = hang (text "GlobalOcc") 2 $ vcat + [ hcat [text "Ident: ", ppr (global_ident g)] + , hcat [text "Id:", ppr (global_id g)] + , hcat [text "Count:", ppr (global_count g)] + ] + -- | Return number of occurrences of every global id used in the given JStgStat. -- Sort by increasing occurrence count. globalOccs :: JStgStat -> G [GlobalOcc] View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/285413a01582a9621cc10a4d4d9dea4a276e6598 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/285413a01582a9621cc10a4d4d9dea4a276e6598 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 21:03:10 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Fri, 27 Sep 2024 17:03:10 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8] 15 commits: Bump submodule deepseq to 1.5.1.0 Message-ID: <66f71d8ea1fec_f7e921bbbf381321bd@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8 at Glasgow Haskell Compiler / GHC Commits: a2a44e47 by Andrew Lelechenko at 2024-09-27T13:31:13-04:00 Bump submodule deepseq to 1.5.1.0 (cherry picked from commit 8e462f4d4bdf2a6c34c249e7be8084565600d300) - - - - - d747384b by Cheng Shao at 2024-09-27T13:31:13-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - 28dd05e8 by Cheng Shao at 2024-09-27T13:31:13-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 43950801 by Cheng Shao at 2024-09-27T13:31:13-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 230a1fc0 by Cheng Shao at 2024-09-27T13:31:13-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fefd77f6 by Andrew Lelechenko at 2024-09-27T14:57:23-04:00 Bump submodule array to 0.5.8.0 (cherry picked from commit 80769bc9f56541601796366485283a697c52a18b) - - - - - 97cad96c by Matthew Pickering at 2024-09-27T15:46:45-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 (cherry picked from commit 7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6) - - - - - 8f3d1a6a by Sylvain Henry at 2024-09-27T15:48:36-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 (cherry picked from commit e7a26d7a6faf1ea534e036c5085a0a027dbb6f5f) - - - - - 9f2e0f50 by Sylvain Henry at 2024-09-27T17:02:50-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 (cherry picked from commit 5092dbff750ee5b6fd082b7eed8574922a2b0bf4) - - - - - 2101a57a by Sylvain Henry at 2024-09-27T17:02:50-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. (cherry picked from commit c749bdfd3e21d712dc2b966482eb010165bdeebe) - - - - - 08963d26 by Sylvain Henry at 2024-09-27T17:02:50-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> (cherry picked from commit 7446a09a2d5b04b95cd43c03659b5647853124ce) - - - - - 257e836d by sheaf at 2024-09-27T17:02:50-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 (cherry picked from commit c29b2b5a77611b2bd6c3089765079bc43aec3e22) - - - - - b0947c95 by Sylvain Henry at 2024-09-27T17:02:50-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 (cherry picked from commit f954f42823f6ca3588425a0d543d93ace86d89e4) - - - - - 12460fb2 by Simon Peyton Jones at 2024-09-27T17:02:50-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify (cherry picked from commit de5d9852dbdd367611bf9e45e69c723d26351992) - - - - - b2b87083 by doyougnu at 2024-09-27T17:02:50-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches (cherry picked from commit bfe4b3d3bbb98b39169fad063c6c32f06d167756) - - - - - 30 changed files: - .gitlab-ci.yml - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Driver/Config/Core/Opt/Simplify.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/StgToJS/Literal.hs - compiler/GHC/SysTools/Cpp.hs - compiler/GHC/Types/Id.hs - compiler/GHC/Unit/Types.hs - libraries/array - libraries/base/GHC/Event/KQueue.hsc - libraries/deepseq - rts/include/Stg.h - rts/linker/PEi386.c - rts/linker/elf_reloc_aarch64.c - rts/sm/GC.c - rts/sm/NonMovingMark.c - rts/sm/Sanity.c - + testsuite/tests/codeGen/should_compile/T25177.hs - + testsuite/tests/codeGen/should_compile/T25177.stderr - testsuite/tests/codeGen/should_compile/all.T - + testsuite/tests/codeGen/should_run/T24700.hs - + testsuite/tests/codeGen/should_run/T24700.stdin - + testsuite/tests/codeGen/should_run/T24700.stdout - + testsuite/tests/codeGen/should_run/T24893.hs - + testsuite/tests/codeGen/should_run/T24893.stdout - testsuite/tests/codeGen/should_run/all.T - testsuite/tests/driver/T20604/T20604.stdout - + testsuite/tests/ghci.debugger/scripts/T25109.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2050caf9ebc81c4fcba05bd394315e739ac18d36...b2b8708398948276d96b1bbd07589431669aa51c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/2050caf9ebc81c4fcba05bd394315e739ac18d36...b2b8708398948276d96b1bbd07589431669aa51c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 22:14:56 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Fri, 27 Sep 2024 18:14:56 -0400 Subject: [Git][ghc/ghc][wip/T25281] More wibbles Message-ID: <66f72e60fc6b_34c29bccec0926e8@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: d7def41c by Simon Peyton Jones at 2024-09-27T23:14:39+01:00 More wibbles - - - - - 8 changed files: - compiler/GHC/HsToCore/Binds.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Match.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Utils.hs - compiler/GHC/Types/Id/Make.hs - compiler/GHC/Types/Var.hs - testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr Changes: ===================================== compiler/GHC/HsToCore/Binds.hs ===================================== @@ -63,7 +63,6 @@ import GHC.Builtin.Types ( naturalTy, typeSymbolKind, charTy ) import GHC.Tc.Types.Evidence import GHC.Types.Id -import GHC.Types.Id.Make ( nospecId ) import GHC.Types.Name import GHC.Types.Var.Set import GHC.Types.Var.Env @@ -832,7 +831,7 @@ dsSpec mb_poly_rhs (L loc (SpecPrag poly_id spec_co spec_inl)) -- perhaps with the body of the lambda wrapped in some WpLets -- E.g. /\a \(d:Eq a). let d2 = $df d in [] (Maybe a) d2 - ; dsHsWrapperForRuleLHS spec_app $ \core_app -> do + ; dsHsWrapper spec_app $ \core_app -> do { let ds_lhs = core_app (Var poly_id) spec_ty = mkLamTypes spec_bndrs (exprType ds_lhs) @@ -1355,7 +1354,7 @@ dsHsWrappers [] k = k [] dsHsWrapper :: HsWrapper -> ((CoreExpr -> CoreExpr) -> DsM a) -> DsM a dsHsWrapper hs_wrap thing_inside - = ds_hs_wrapper False hs_wrap $ \ core_wrap -> + = ds_hs_wrapper hs_wrap $ \ core_wrap -> addTyCs FromSource (hsWrapDictBinders hs_wrap) $ -- addTyCs: Add type evidence to the refinement type -- predicate of the coverage checker @@ -1366,15 +1365,10 @@ dsHsWrapper hs_wrap thing_inside -- oracle. thing_inside core_wrap -dsHsWrapperForRuleLHS :: HsWrapper -> ((CoreExpr -> CoreExpr) -> DsM a) -> DsM a -dsHsWrapperForRuleLHS = ds_hs_wrapper True - -ds_hs_wrapper :: Bool -- True <=> LHS of a RULE - -- See (NC1) in Note [Desugaring non-canonical evidence] - -> HsWrapper +ds_hs_wrapper :: HsWrapper -> ((CoreExpr -> CoreExpr) -> DsM a) -> DsM a -ds_hs_wrapper is_rule_lhs wrap = go wrap +ds_hs_wrapper wrap = go wrap where go WpHole k = k $ \e -> e go (WpTyApp ty) k = k $ \e -> App e (Type ty) @@ -1382,6 +1376,8 @@ ds_hs_wrapper is_rule_lhs wrap = go wrap go (WpTyLam tv) k = k $ Lam tv go (WpCast co) k = assert (coercionRole co == Representational) $ k $ \e -> mkCastDs e co + go (WpEvApp tm) k = do { core_tm <- dsEvTerm tm + ; k $ \e -> e `App` core_tm } go (WpLet ev_binds) k = dsTcEvBinds ev_binds $ \bs -> k (mkCoreLets bs) go (WpCompose c1 c2) k = go c1 $ \w1 -> @@ -1391,37 +1387,15 @@ ds_hs_wrapper is_rule_lhs wrap = go wrap do { x <- newSysLocalDs st ; go c1 $ \w1 -> go c2 $ \w2 -> - let app f a = mkCoreAppDs (text "dsHsWrapper") f a + let app f a = mkCoreApp (text "dsHsWrapper") f a arg = w1 (Var x) in k (\e -> (Lam x (w2 (app e arg)))) } - go (WpEvApp tm) k = do { core_tm <- dsEvTerm tm - - -- See Note [Desugaring non-canonical evidence] - ; unspecables <- getUnspecables - ; let vs = exprFreeVarsList core_tm - is_unspecable_var v = v `S.member` unspecables - is_specable - | is_rule_lhs = True - | otherwise = not $ any (is_unspecable_var) vs - - ; k (\e -> app_ev is_specable e core_tm) } -- See Note [Coercions returned from tcSubMult] in GHC.Tc.Utils.Unify. go (WpMultCoercion co) k = do { unless (isReflexiveCo co) $ diagnosticDs DsMultiplicityCoercionsNotSupported ; k $ \e -> e } --- We are about to construct an evidence application `f dict`. If the dictionary is --- non-specialisable, instead construct --- nospec f dict --- See Note [nospecId magic] in GHC.Types.Id.Make for what `nospec` does. -app_ev :: Bool -> CoreExpr -> CoreExpr -> CoreExpr -app_ev is_specable k core_tm - | not is_specable - = Var nospecId `App` Type (exprType k) `App` k `App` core_tm - - | otherwise - = k `App` core_tm -------------------------------------- dsTcEvBinds_s :: [TcEvBinds] -> ([CoreBind] -> DsM a) -> DsM a ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -44,10 +44,11 @@ import GHC.Tc.Types.Evidence import GHC.Tc.Utils.Monad import GHC.Tc.Instance.Class (lookupHasFieldLabel) +import GHC.Core +import GHC.Core.FVs( exprsFreeVarsList ) import GHC.Core.FamInstEnv( topNormaliseType ) import GHC.Core.Type import GHC.Core.TyCo.Rep -import GHC.Core import GHC.Core.Utils import GHC.Core.Make import GHC.Core.PatSyn @@ -61,6 +62,7 @@ import GHC.Types.CostCentre import GHC.Types.Id import GHC.Types.Id.Info import GHC.Types.Id.Make +import GHC.Types.Var( isInvisibleAnonPiTyBinder ) import GHC.Types.Basic import GHC.Types.SrcLoc import GHC.Types.Tickish @@ -74,9 +76,10 @@ import GHC.Builtin.Names import GHC.Utils.Misc import GHC.Utils.Outputable as Outputable import GHC.Utils.Panic -import Control.Arrow ( first ) +import Control.Arrow( first ) import Control.Monad import Data.Maybe( isJust ) +import qualified Data.Set as S {- ************************************************************************ @@ -576,6 +579,73 @@ There are several special cases to handle * HsVar: special magic for `noinline` * HsVar: special magic for `seq` +Note [Desugaring seq] +~~~~~~~~~~~~~~~~~~~~~ +There are a few subtleties in the desugaring of `seq`, all +implemented in the `seqId` case of `ds_app_var`: + + 1. (as described in #1031) + + Consider, + f x y = x `seq` (y `seq` (# x,y #)) + + Because the argument to the outer 'seq' has an unlifted type, we'll use + call-by-value, and compile it as if we had + + f x y = case (y `seq` (# x,y #)) of v -> x `seq` v + + But that is bad, because we now evaluate y before x! + + Seq is very, very special! So we recognise it right here, and desugar to + case x of _ -> case y of _ -> (# x,y #) + + 2. (as described in #2273) + + Consider + let chp = case b of { True -> fst x; False -> 0 } + in chp `seq` ...chp... + Here the seq is designed to plug the space leak of retaining (snd x) + for too long. + + If we rely on the ordinary inlining of seq, we'll get + let chp = case b of { True -> fst x; False -> 0 } + case chp of _ { I# -> ...chp... } + + But since chp is cheap, and the case is an alluring context, we'll + inline chp into the case scrutinee. Now there is only one use of chp, + so we'll inline a second copy. Alas, we've now ruined the purpose of + the seq, by re-introducing the space leak: + case (case b of {True -> fst x; False -> 0}) of + I# _ -> ...case b of {True -> fst x; False -> 0}... + + We can try to avoid doing this by ensuring that the binder-swap in the + case happens, so we get this at an early stage: + case chp of chp2 { I# -> ...chp2... } + But this is fragile. The real culprit is the source program. Perhaps we + should have said explicitly + let !chp2 = chp in ...chp2... + + But that's painful. So the code here does a little hack to make seq + more robust: a saturated application of 'seq' is turned *directly* into + the case expression, thus: + x `seq` e2 ==> case x of x -> e2 -- Note shadowing! + e1 `seq` e2 ==> case x of _ -> e2 + + So we desugar our example to: + let chp = case b of { True -> fst x; False -> 0 } + case chp of chp { I# -> ...chp... } + And now all is well. + + The reason it's a hack is because if you define mySeq=seq, the hack + won't work on mySeq. + + 3. (as described in #2409) + + The isInternalName ensures that we don't turn + True `seq` e + into + case True of True { ... } + which stupidly tries to bind the datacon 'True'. -} dsApp :: HsExpr GhcTc -> DsM CoreExpr @@ -663,7 +733,7 @@ ds_app_var (L loc fun_id) hs_args core_args -- otherwise this might not be the implicit HasField instance -> do { sel_id <- dsLookupGlobalId (flSelector fl) ; ds_app_rec_sel sel_id rest_args } - _ -> ds_app_done fun_id core_args } + _ -> ds_app_finish fun_id core_args } ----------------------- -- Warn about identities for (fromInteger :: Integer -> Integer) etc @@ -678,7 +748,7 @@ ds_app_var (L loc fun_id) hs_args core_args -- So we are converting ty -> ty diagnosticDs (DsIdentitiesFound fun_id conv_ty) - ; ds_app_done fun_id core_args } + ; ds_app_finish fun_id core_args } ----------------------- -- Warn about unused return value in @@ -690,7 +760,7 @@ ds_app_var (L loc fun_id) hs_args core_args = do { tracePm ">>" (ppr loc $$ ppr arg_ty $$ ppr (isGeneratedSrcSpan (locA loc))) ; when (isGeneratedSrcSpan (locA loc)) $ -- It is a compiler-generated (>>) warnDiscardedDoBindings hs_arg m_ty arg_ty - ; ds_app_done fun_id core_args } + ; ds_app_finish fun_id core_args } ----------------------- -- Deal with `noinline` @@ -716,11 +786,30 @@ ds_app_var (L loc fun_id) hs_args core_args ----------------------- -- Phew! No more special cases. Just build an applications | otherwise - = ds_app_done fun_id core_args - -ds_app_done :: Id -> [CoreExpr] -> DsM CoreExpr -ds_app_done fun_id core_args = return (mkCoreApps (Var fun_id) core_args) + = ds_app_finish fun_id core_args +--------------- +ds_app_finish :: Id -> [CoreExpr] -> DsM CoreExpr +-- We are about to construct an application that may include evidence applications +-- `f dict`. If the dictionary is non-specialisable, instead construct +-- nospec f dict +-- See Note [nospecId magic] in GHC.Types.Id.Make for what `nospec` does. +-- See Note [Desugaring non-canonical evidence] +ds_app_finish fun_id core_args + = do { unspecables <- getUnspecables + ; let fun_ty = idType fun_id + free_dicts = exprsFreeVarsList + [ e | (e,pi_bndr) <- core_args `zip` fst (splitPiTys fun_ty) + , isInvisibleAnonPiTyBinder pi_bndr ] + is_unspecable_var v = v `S.member` unspecables + + fun | not (S.null unspecables) -- Fast path + , any (is_unspecable_var) free_dicts + = Var nospecId `App` Type fun_ty `App` Var fun_id + | otherwise + = Var fun_id + + ; return (mkCoreApps fun core_args) } --------------- ds_app_rec_sel :: Id -> [CoreExpr] -> DsM CoreExpr @@ -757,14 +846,12 @@ ds_app_rec_sel sel_id core_args ; let maxCons = maxUncoveredPatterns dflags ; diagnosticDs $ DsIncompleteRecordSelector (idName sel_id) cons_wo_field maxCons } - ; return (mkCoreApps (Var sel_id) core_args) } + ; ds_app_finish sel_id core_args } | otherwise = pprPanic "ds_app_rec_sel" (ppr sel_id $$ ppr (idDetails sel_id)) where --- apply_type_args applies the record selector to its --- initial type args, returning the remaining args, if any apply_type_args :: Id -> [CoreExpr] -> (Type, [CoreExpr]) -- Apply function to the initial /type/ args; -- return the type of the instantiated function, ===================================== compiler/GHC/HsToCore/Match.hs ===================================== @@ -303,7 +303,7 @@ matchView (var :| vars) ty eqns@(eqn1 :| _) -- compile the view expressions ; viewExpr' <- dsExpr viewExpr ; return (mkViewMatchResult var' - (mkCoreAppDs (text "matchView") viewExpr' (Var var)) + (mkCoreApp (text "matchView") viewExpr' (Var var)) match_result) } -- decompose the first pattern and leave the rest alone ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -264,7 +264,7 @@ Furthermore, HasField constraints allow to delay the completeness check from the field access site to a caller, as in test cases TcIncompleteRecSel and T24891: accessDot :: HasField "sel2" t Int => t -> Int - accessDot x = x.sel2 + accessDot x = x.sel2 -- getField @Symbol @"sel2" @t @Int x solveDot :: Dot -> Int solveDot = accessDot @@ -348,20 +348,22 @@ Finally, there are 2 more items addressing -XOverloadedRecordDot: We want to catch these applications in the saturated (2) case. (The unsaturated case is handled implicitly by (7).) For example, we do not want to generate a warning for function `ldiDot`! - For that, we need to be smart in `decomposeRecSelHead`, which matches out - the record selector. It must treat the above expression similar to a vanilla - RecSel app `sel2 d`. - This is a bit nasty since we cannot look at the unfolding of `$dHasField`. - Tested in T24891. - - 7. For `accessDot` above, `decomposeRecSelHead` will fail to find a record - selector, because type `t` is not obviously a record type. + + Function `GHC.HsToCore.Expr.ds_app_var` spots the `getField` application, + and then treats the above expression similar to a vanilla (RecSel app sel2 d). + This is a bit nasty (it has to do instance lookup) since we cannot look at + the unfolding of `$dHasField`. Tested in T24891. + + 7. For `accessDot` above, `ds_app_var` will fail to find a record selector, + because type `t` is not obviously a record type. + That's good, because it means we won't emit a warning for `accessDot`. - But we should really emit a warning for `solveDot`! - There, the compiler solves a `HasField` constraint and without an immediate - `getField`, roughly `solveDot = accessDot @Dot $d`. - It is the job of the solver to warn about incompleteness here, - in `GHC.Tc.Instance.Class.matchHasField`. + + But we really should emit a warning for `solveDot`! There, the + compiler solves a `HasField` constraint and without an immediate + `getField`, roughly `solveDot = accessDot @Dot $d`. It must be the job + of the solver to warn about incompleteness here, in + `GHC.Tc.Instance.Class.matchHasField`. What makes this complicated is that we do not *also* want to warn in the example `dot d = d.sel2` above, which is covered by more precise case (6)! ===================================== compiler/GHC/HsToCore/Utils.hs ===================================== @@ -28,8 +28,7 @@ module GHC.HsToCore.Utils ( mkCoPrimCaseMatchResult, mkCoAlgCaseMatchResult, mkCoSynCaseMatchResult, wrapBind, wrapBinds, - mkErrorAppDs, mkCoreAppDs, mkCoreAppsDs, mkCastDs, - mkFailExpr, + mkErrorAppDs, mkCastDs, mkFailExpr, seqVar, @@ -77,7 +76,6 @@ import GHC.Types.Unique.Set import GHC.Types.Unique.Supply import GHC.Unit.Module import GHC.Builtin.Names -import GHC.Types.Name( isInternalName ) import GHC.Utils.Outputable import GHC.Utils.Panic import GHC.Types.SrcLoc @@ -337,7 +335,7 @@ mkPatSynCase var ty alt fail = do matcher <- dsLExpr $ mkLHsWrap wrapper $ nlHsTyApp matcher_id [getRuntimeRep ty, ty] cont <- mkCoreLams bndrs <$> runMatchResult fail match_result - return $ mkCoreAppsDs (text "patsyn" <+> ppr var) matcher [Var var, ensure_unstrict cont, Lam voidArgId fail] + return $ mkCoreApps matcher [Var var, ensure_unstrict cont, Lam voidArgId fail] where MkCaseAlt{ alt_pat = psyn, alt_bndrs = bndrs, @@ -469,102 +467,6 @@ mkFailExpr :: HsMatchContextRn -> Type -> DsM CoreExpr mkFailExpr ctxt ty = mkErrorAppDs pAT_ERROR_ID ty (matchContextErrString ctxt) -{- -'mkCoreAppDs' and 'mkCoreAppsDs' handle the special-case desugaring of 'seq'. - -Note [Desugaring seq] -~~~~~~~~~~~~~~~~~~~~~ - -There are a few subtleties in the desugaring of `seq`: - - 1. (as described in #1031) - - Consider, - f x y = x `seq` (y `seq` (# x,y #)) - - Because the argument to the outer 'seq' has an unlifted type, we'll use - call-by-value, and compile it as if we had - - f x y = case (y `seq` (# x,y #)) of v -> x `seq` v - - But that is bad, because we now evaluate y before x! - - Seq is very, very special! So we recognise it right here, and desugar to - case x of _ -> case y of _ -> (# x,y #) - - 2. (as described in #2273) - - Consider - let chp = case b of { True -> fst x; False -> 0 } - in chp `seq` ...chp... - Here the seq is designed to plug the space leak of retaining (snd x) - for too long. - - If we rely on the ordinary inlining of seq, we'll get - let chp = case b of { True -> fst x; False -> 0 } - case chp of _ { I# -> ...chp... } - - But since chp is cheap, and the case is an alluring context, we'll - inline chp into the case scrutinee. Now there is only one use of chp, - so we'll inline a second copy. Alas, we've now ruined the purpose of - the seq, by re-introducing the space leak: - case (case b of {True -> fst x; False -> 0}) of - I# _ -> ...case b of {True -> fst x; False -> 0}... - - We can try to avoid doing this by ensuring that the binder-swap in the - case happens, so we get this at an early stage: - case chp of chp2 { I# -> ...chp2... } - But this is fragile. The real culprit is the source program. Perhaps we - should have said explicitly - let !chp2 = chp in ...chp2... - - But that's painful. So the code here does a little hack to make seq - more robust: a saturated application of 'seq' is turned *directly* into - the case expression, thus: - x `seq` e2 ==> case x of x -> e2 -- Note shadowing! - e1 `seq` e2 ==> case x of _ -> e2 - - So we desugar our example to: - let chp = case b of { True -> fst x; False -> 0 } - case chp of chp { I# -> ...chp... } - And now all is well. - - The reason it's a hack is because if you define mySeq=seq, the hack - won't work on mySeq. - - 3. (as described in #2409) - - The isInternalName ensures that we don't turn - True `seq` e - into - case True of True { ... } - which stupidly tries to bind the datacon 'True'. --} - --- NB: Make sure the argument is not representation-polymorphic -mkCoreAppDs :: SDoc -> CoreExpr -> CoreExpr -> CoreExpr -mkCoreAppDs _ (Var f `App` Type _r `App` Type ty1 `App` Type ty2 `App` arg1) arg2 - | f `hasKey` seqIdKey -- Note [Desugaring seq], points (1) and (2) - = Case arg1 case_bndr ty2 [Alt DEFAULT [] arg2] - where - case_bndr = case arg1 of - Var v1 | isInternalName (idName v1) - -> v1 -- Note [Desugaring seq], points (2) and (3) - _ -> mkWildValBinder ManyTy ty1 - -mkCoreAppDs _ (Var f `App` Type _r) arg - | f `hasKey` noinlineIdKey -- See Note [noinlineId magic] in GHC.Types.Id.Make - , (fun, args) <- collectArgs arg - , not (null args) - = (Var f `App` Type (exprType fun) `App` fun) - `mkCoreApps` args - -mkCoreAppDs s fun arg = mkCoreApp s fun arg -- The rest is done in GHC.Core.Make - --- NB: No argument can be representation-polymorphic -mkCoreAppsDs :: SDoc -> CoreExpr -> [CoreExpr] -> CoreExpr -mkCoreAppsDs s fun args = foldl' (mkCoreAppDs s) fun args - mkCastDs :: CoreExpr -> Coercion -> CoreExpr -- We define a desugarer-specific version of GHC.Core.Utils.mkCast, -- because in the immediate output of the desugarer, we can have ===================================== compiler/GHC/Types/Id/Make.hs ===================================== @@ -2199,7 +2199,7 @@ But actually we give 'noinline' a wired-in name for three distinct reasons: Solution: in the desugarer, rewrite noinline (f x y) ==> noinline f x y - This is done in GHC.HsToCore.Utils.mkCoreAppDs. + This is done in the `noinlineId` case of `GHC.HsToCore.Expr.ds_app_var` This is only needed for noinlineId, not noInlineConstraintId (wrinkle (W1) below), because the latter never shows up in user code. ===================================== compiler/GHC/Types/Var.hs ===================================== @@ -82,7 +82,8 @@ module GHC.Types.Var ( -- * PiTyBinder PiTyBinder(..), PiTyVarBinder, - isInvisiblePiTyBinder, isVisiblePiTyBinder, + isInvisiblePiTyBinder, isInvisibleAnonPiTyBinder, + isVisiblePiTyBinder, isTyBinder, isNamedPiTyBinder, isAnonPiTyBinder, namedPiTyBinder_maybe, anonPiTyBinderType_maybe, piTyBinderType, @@ -751,7 +752,6 @@ instance Outputable PiTyBinder where ppr (Named (Bndr v Specified)) = char '@' <> ppr v ppr (Named (Bndr v Inferred)) = braces (ppr v) - -- | 'PiTyVarBinder' is like 'PiTyBinder', but there can only be 'TyVar' -- in the 'Named' field. type PiTyVarBinder = PiTyBinder @@ -761,6 +761,10 @@ isInvisiblePiTyBinder :: PiTyBinder -> Bool isInvisiblePiTyBinder (Named (Bndr _ vis)) = isInvisibleForAllTyFlag vis isInvisiblePiTyBinder (Anon _ af) = isInvisibleFunArg af +isInvisibleAnonPiTyBinder :: PiTyBinder -> Bool +isInvisibleAnonPiTyBinder (Named {}) = False +isInvisibleAnonPiTyBinder (Anon _ af) = isInvisibleFunArg af + -- | Does this binder bind a visible argument? isVisiblePiTyBinder :: PiTyBinder -> Bool isVisiblePiTyBinder = not . isInvisiblePiTyBinder ===================================== testsuite/tests/parser/should_compile/DumpTypecheckedAst.stderr ===================================== @@ -1997,7 +1997,7 @@ (NoExtField) (L (EpAnn - (EpaSpan { }) + (EpaSpan { DumpTypecheckedAst.hs:20:8-15 }) (NameAnnTrailing []) (EpaComments View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d7def41c2c041f1332031466f2eb6f83f46c99c4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d7def41c2c041f1332031466f2eb6f83f46c99c4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 22:45:36 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Fri, 27 Sep 2024 18:45:36 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8-2] 187 commits: Bump process submodule to 1.6.18.0 Message-ID: <66f7359080079_34c29b2b70509452d@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8-2 at Glasgow Haskell Compiler / GHC Commits: 68e111c8 by Ben Gamari at 2023-09-20T09:19:33-04:00 Bump process submodule to 1.6.18.0 - - - - - 2616921f by Ben Gamari at 2023-09-20T09:19:33-04:00 Bump hsc2hs submodule to 0.68.10 - - - - - a591fde5 by Ben Gamari at 2023-09-20T09:19:33-04:00 Bump deepseq submodule to 1.5.0.0 - - - - - f1b28bc7 by Ben Gamari at 2023-09-20T09:19:33-04:00 Bump parsec submodule to 3.1.17.0 - - - - - e9c4eb47 by Ben Gamari at 2023-09-20T09:19:33-04:00 Bump nofib submodule - - - - - 31a2dd67 by Ben Gamari at 2023-09-20T09:19:33-04:00 base: Update changelog - - - - - 09233267 by Ben Gamari at 2023-09-20T09:19:33-04:00 template-haskell: Update changelog - - - - - f23eda82 by Ben Gamari at 2023-09-25T10:09:27-04:00 base: Fix changelog formatting - - - - - 6a6fb5b5 by Ben Gamari at 2023-09-25T10:09:31-04:00 relnotes: Add a few notable points to the `base` section - - - - - bf9b8f09 by Ben Gamari at 2023-09-25T10:09:31-04:00 users-guide: Fix relnotes wibbles Partially fixes #23988. - - - - - b34ccb2e by Ben Gamari at 2023-09-27T09:27:06-04:00 hadrian: Install LICENSE files in bindists Fixes #23548. - - - - - 63b6ef7d by Ben Gamari at 2023-09-27T09:29:57-04:00 Bump containers submodule To 0.7. - - - - - 3359e6b4 by Ben Gamari at 2023-09-27T11:59:45-04:00 Bump haddock submodule Applying fix from #21984. - - - - - e23147f4 by Ben Gamari at 2023-09-27T11:59:45-04:00 users-guide: Refactor handling of :base-ref: et al. (cherry picked from commit 8f82e99fda693326e55ae798e11f3896c875c966) - - - - - 9c680ee5 by Ben Gamari at 2023-09-27T11:59:45-04:00 Bump nofib submodule - - - - - f7b3fcb3 by Ben Gamari at 2023-09-27T14:25:14-04:00 users-guide: Amend discussion of incoherent specialisation Closing #23988. - - - - - 5a602b33 by Ben Gamari at 2023-09-28T12:00:12-04:00 Bump containers submodule to 0.6.8 - - - - - c83b7688 by Ben Gamari at 2023-09-28T14:27:50-04:00 Bump Cabal submodule to 3.10.2.0 final - - - - - 86c1bdc4 by Ben Gamari at 2023-09-28T23:14:08-04:00 gitlab-ci: Allow release-hackage-lint to fail Head.hackage is not robust in the presence of patches to boot packages, as see in ghc/head.hackage#93. Allow it to fail for now. - - - - - 2a51d113 by Ben Gamari at 2023-10-05T16:13:45-04:00 configure: Fix #21712 again This is a bit of a shot in the dark to fix #24033, which appears to be another instance of #21712. For some reason the ld-override logic *still* appears to be active on Darwin targets (or at least one). Consequently, on misconfigured systems we may choose a non-`ld64` linker. It's a bit unclear exactly what happened in #24033 but ultimately the check added for #21712 was not quite right, checking for the `ghc_host_os` (the value of which depends upon the bootstrap compiler) instead of the target platform. Fix this. Fixes #24033. (cherry picked from commit f6b2751f58df5f4f83caa7a7ca56e66659d02b09) - - - - - 8ed7ffb6 by Ben Gamari at 2023-10-05T16:13:45-04:00 rts/nonmoving: Fix on LLP64 platforms Previously `NONMOVING_SEGMENT_MASK` and friends were defined with the `UL` size suffix. However, this is wrong on LLP64 platforms like Windows, where `long` is 32-bits. Fixes #23003. Fixes #24042. (cherry picked from commit 8f6010b98f560200997a9d84a4e07bfd0ad6e496) - - - - - 894b6027 by sheaf at 2023-10-05T16:13:45-04:00 Fix non-symbolic children lookup of fixity decl The fix for #23664 did not correctly account for non-symbolic names when looking up children of a given parent. This one-line fix changes that. Fixes #24037 (cherry picked from commit 8cee3fd7febdd97a9b4bcae7dddafe69b166149c) - - - - - 9fb11d4a by Andreas Klebinger at 2023-10-05T16:15:25-04:00 Arm: Make ppr methods easier to use by not requiring NCGConfig (cherry picked from commit b048bea095d48751e7c182f2837c4a31fcad8fd7) - - - - - b3ab756f by Andreas Klebinger at 2023-10-05T17:41:42-04:00 AArch64: Fix broken conditional jumps for offsets >= 1MB Rewrite conditional jump instructions with offsets >= 1MB to use unconditional jumps to avoid overflowing the immediate. Fixes #23746 (cherry picked from commit 2adc050857a9c1b992040fbfd55fbe65b2851b19) - - - - - 857967ae by Luite Stegeman at 2023-10-05T17:41:42-04:00 JS: correct file size and times Programs produced by the JavaScript backend were returning incorrect file sizes and modification times, causing cabal related tests to fail. This fixes the problem and adds an additional test that verifies basic file information operations. fixes #23980 (cherry picked from commit 11ecc37bc27ffa1cf31358e21e09e140befa940c) - - - - - 81b1179d by Ben Gamari at 2023-10-05T17:41:42-04:00 system-cxx-std-lib: Add license and description (cherry picked from commit 151187407b7b4bdda5b80bd7b8bdf96d05e278dd) - - - - - 0fc4ce51 by Torsten Schmits at 2023-10-05T17:41:42-04:00 Add -fbreak-points to control breakpoint insertion Rather than statically enabling breakpoints only for the interpreter, this adds a new flag. Tracking ticket: #23057 MR: !10466 (cherry picked from commit 083794b10dc27e9d97b62cc8b8eb4e1da162bf66) - - - - - e538003c by Alexis King at 2023-10-05T17:41:42-04:00 Don’t store the async exception masking state in CATCH frames (cherry picked from commit 8b61dfd6dfc78bfa6bb9449dac9a336e5d668b5e) - - - - - 46896580 by Ben Gamari at 2023-10-05T17:41:42-04:00 gitlab-ci: Drop libiserv from upload_ghc_libs libiserv has been merged into the ghci package. (cherry picked from commit b35fd2cd7a12f3354a7fd2301bdf610c5d435017) - - - - - 34bd605d by Ben Gamari at 2023-10-05T17:41:42-04:00 Bump unix submodule to 2.8.3.0 - - - - - 47b202a9 by Ben Gamari at 2023-10-05T23:32:49-04:00 Bump stm submodule to 2.5.2.0 - - - - - 31d4e495 by Ben Gamari at 2023-10-05T23:32:49-04:00 Bump hpc submodule to 0.7.0.0 - - - - - 7c268df5 by Ben Gamari at 2023-10-05T23:32:54-04:00 Allow perf changes These appear to be spurious. Metric Decrease: MultiLayerModules T13701 T14697 - - - - - b238204e by Bryan Richter at 2023-10-06T04:53:24-04:00 Work around perf note fetch failure Addresses #24055. (cherry picked from commit 63afb701a1638d7bd32c34fb24a9fd3ff897b634) - - - - - bda60b3a by Ben Gamari at 2023-10-06T21:45:17-04:00 Bump haddock submodule to 2.30 - - - - - b2d15ab1 by Ben Gamari at 2023-10-06T21:45:48-04:00 hadrian: Fix hpc parser path - - - - - 443e870d by Ben Gamari at 2023-10-06T21:45:48-04:00 Release GHC 9.8.1 - - - - - 0cbb9e10 by Moritz Angermann at 2024-01-21T17:46:58-05:00 CgUtils.fixStgRegStmt respect register width This change ensure that the reg + offset computation is always of the same size. Before this we could end up with a 64bit register, and then add a 32bit offset (on 32bit platforms). This not only would fail type sanity checking, but also incorrectly truncate 64bit values into 32bit values silently on 32bit architectures. (cherry picked from commit dafc47091c9107dcf81e1e80a105f59211927c89) - - - - - da8d2a29 by Matthew Pickering at 2024-01-21T22:06:34-05:00 Add aarch64 alpine bindist This is dynamically linked and makes creating statically linked executables more straightforward. Fixes #23482 (cherry picked from commit 51b57d6587b75020045c1a1edf9eb02990ca63cc) - - - - - 53b308e9 by sheaf at 2024-01-21T22:06:34-05:00 Combine GREs when combining in mkImportOccEnv In `GHC.Rename.Names.mkImportOccEnv`, we sometimes discard one import item in favour of another, as explained in Note [Dealing with imports] in `GHC.Rename.Names`. However, this can cause us to lose track of important parent information. Consider for example #24084: module M1 where { class C a where { type T a } } module M2 ( module M1 ) where { import M1 } module M3 where { import M2 ( C, T ); instance C () where T () = () } When processing the import list of `M3`, we start off (for reasons that are not relevant right now) with two `Avail`s attached to `T`, namely `C(C, T)` and `T(T)`. We combine them in the `combine` function of `mkImportOccEnv`; as described in Note [Dealing with imports] we discard `C(C, T)` in favour of `T(T)`. However, in doing so, we **must not** discard the information want that `C` is the parent of `T`. Indeed, losing track of this information can cause errors when importing, as we could get an error of the form ‘T’ is not a (visible) associated type of class ‘C’ We fix this by combining the two GREs for `T` using `plusGRE`. Fixes #24084 (cherry picked from commit ec3c4488f456f6f9bdd28a09f0b1e87fd3782db9) - - - - - c4d1edb1 by Cheng Shao at 2024-01-21T22:06:34-05:00 testsuite: increase timeout of ghc-api tests for wasm32 ghc-api tests for wasm32 are more likely to timeout due to the large wasm module sizes, especially when testing with wasm native tail calls, given wasmtime's handling of tail call opcodes are suboptimal at the moment. It makes sense to increase timeout specifically for these tests on wasm32. This doesn't affect other targets, and for wasm32 we don't increase timeout for all tests, so not to risk letting major performance regressions slip through the testsuite. (cherry picked from commit 07ab5cc10d4ce8db0c3b099d8bd34da66db51b9e) - - - - - a2d7130e by Simon Peyton Jones at 2024-01-21T22:06:34-05:00 Fix non-termination bug in equality solver constraint left-to-right then right to left, forever. Easily fixed. (cherry picked from commit 21b76843e9b51cd27be32b8c595f29a784276229) - - - - - d9a34b85 by Cheng Shao at 2024-01-21T22:06:34-05:00 compiler: fix eager blackhole symbol in wasm32 NCG (cherry picked from commit fe50eb3510ca730c4d2fc57778d380fdc493abd8) - - - - - a0631c4a by Cheng Shao at 2024-01-21T22:06:34-05:00 testsuite: fix optasm tests for wasm32 (cherry picked from commit af77114815ad9052261898c830171e49fbed5160) - - - - - 186956a6 by Matthew Pickering at 2024-01-21T22:06:34-05:00 testsuite: Add wasm32 to testsuite arches with NCG The compiler --info reports that wasm32 compilers have a NCG, so we should agree with that here. (cherry picked from commit 1b90735c99f1179328f6dd67dbcc81f964901a19) - - - - - 2bb6c062 by Josh Meredith at 2024-01-21T22:06:34-05:00 JavaScript: support unicode code points > 2^16 in toJSString using String.fromCodePoint (#23628) (cherry picked from commit 09a5c6cccf8f1b517bc01e8cc924e151d9cbae49) - - - - - 7c419b07 by Josh Meredith at 2024-01-21T22:06:34-05:00 JavaScript: update MK_TUP macros to use current tuple constructors (#23659) (cherry picked from commit b9d5bfe9e76e18c78dbcb67e1c5c33a15a54dcd0) - - - - - 60f3a468 by Sylvain Henry at 2024-01-21T22:06:34-05:00 JS: testsuite: use req_c predicate instead of js_broken (cherry picked from commit 41968fd699a285e1a1b43535fda28a61bd1202f1) - - - - - e110b68f by Sylvain Henry at 2024-01-21T22:06:34-05:00 JS: implement some file primitives (lstat,rmdir) (#22374) - Implement lstat and rmdir. - Implement base_c_s_is* functions (testing a file type) - Enable passing tests (cherry picked from commit 74a4dd2ec6e200b11a56b6f82907feb66e94c90b) - - - - - 6c31021e by Andreas Klebinger at 2024-01-21T22:06:34-05:00 Fix FMA primops generating broken assembly on x86. `genFMA3Code` assumed that we had to take extra precations to avoid overwriting the result of `getNonClobberedReg`. One of these special cases caused a bug resulting in broken assembly. I believe we don't need to hadle these cases specially at all, which means this MR simply deletes the special cases to fix the bug. Fixes #24160 (cherry picked from commit fa576eb8a55db8e216eed3e7ea8807121b3496f5) - - - - - c52d657c by Zubin Duggal at 2024-01-21T22:06:34-05:00 driver: Don't lose track of nodes when we fail to resolve cycles The nodes that take part in a cycle should include both hs-boot and hs files, but when we fail to resolve a cycle, we were only counting the nodes from the graph without boot files. Fixes #24196 (cherry picked from commit 717c9b59ecfb919cba1b8cec6597d0c2fb3ef1a1) - - - - - 6a78bcc1 by Sebastian Graf at 2024-01-30T17:14:16+05:30 Cpr: Turn an assertion into a check to deal with some dead code (#23862) See the new `Note [Dead code may contain type confusions]`. Fixes #23862. (cherry picked from commit 57c391c463f26b7025df9b340ad98416cff1d2b2) - - - - - 724102f6 by Zubin Duggal at 2024-01-30T17:14:16+05:30 hadrian: set -Wno-deprecations for directory and Win32 The filepath bump to 1.4.200.1 introduces a deprecation warning. See https://gitlab.haskell.org/ghc/ghc/-/issues/24240 https://github.com/haskell/filepath/pull/206 (cherry picked from commit 86f652dc9a649e59e643609c287a510a565f5408) - - - - - 38cb7fb4 by Moritz Angermann at 2024-01-30T17:14:16+05:30 Drop hard Xcode dependency XCODE_VERSION calls out to `xcodebuild`, which is only available when having `Xcode` installed. The CommandLineTools are not sufficient. To install Xcode, you must have an apple id to download the Xcode.xip from apple. We do not use xcodebuild anywhere in our build explicilty. At best it appears to be a proxy for checking the linker or the compiler. These should rather be done with ``` xcrun ld -version ``` or similar, and not by proxy through Xcode. The CLR should be sufficient for building software on macOS. (cherry picked from commit a3ee3b99e6889fd68da75c6ea7a14d101f71da56) - - - - - 3ac10c2c by Sylvain Henry at 2024-01-30T17:14:16+05:30 Fix unusable units and module reexport interaction (#21097) This commit fixes an issue with ModUnusable introduced in df0f148feae. In mkUnusableModuleNameProvidersMap we traverse the list of unusable units and generate ModUnusable origin for all the modules they contain: exposed modules, hidden modules, and also re-exported modules. To do this we have a two-level map: ModuleName -> Unit:ModuleName (aka Module) -> ModuleOrigin So for each module name "M" in broken unit "u" we have: "M" -> u:M -> ModUnusable reason However in the case of module reexports we were using the *target* module as a key. E.g. if "u:M" is a reexport for "X" from unit "o": "M" -> o:X -> ModUnusable reason Case 1: suppose a reexport without module renaming (u:M -> o:M) from unusable unit u: "M" -> o:M -> ModUnusable reason Here it's claiming that the import of M is unusable because a reexport from u is unusable. But if unit o isn't unusable we could also have in the map: "M" -> o:M -> ModOrigin ... Issue: the Semigroup instance of ModuleOrigin doesn't handle the case (ModUnusable <> ModOrigin) Case 2: similarly we could have 2 unusable units reexporting the same module without renaming, say (u:M -> o:M) and (v:M -> o:M) with u and v unusable. It gives: "M" -> o:M -> ModUnusable ... (for u) "M" -> o:M -> ModUnusable ... (for v) Issue: the Semigroup instance of ModuleOrigin doesn't handle the case (ModUnusable <> ModUnusable). This led to #21097, #16996, #11050. To fix this, in this commit we make ModUnusable track whether the module used as key is a reexport or not (for better error messages) and we use the re-export module as key. E.g. if "u:M" is a reexport for "o:X" and u is unusable, we now record: "M" -> u:M -> ModUnusable reason reexported=True So now, we have two cases for a reexport u:M -> o:X: - u unusable: "M" -> u:M -> ModUnusable ... reexported=True - u usable: "M" -> o:X -> ModOrigin ... reexportedFrom=u:M The second case is indexed with o:X because in this case the Semigroup instance of ModOrigin is used to combine valid expositions of a module (directly or via reexports). Note that module lookup functions select usable modules first (those who have a ModOrigin value), so it doesn't matter if we add new ModUnusable entries in the map like this: "M" -> { u:M -> ModUnusable ... reexported=True o:M -> ModOrigin ... } The ModOrigin one will be used. Only if there is no ModOrigin or ModHidden entry will the ModUnusable error be printed. See T21097 for an example printing several reasons why an import is unusable. (cherry picked from commit cee81370cd6ef256f66035e3116878d4cb82e28b) - - - - - 2af5c8d1 by Moritz Angermann at 2024-01-30T17:14:16+05:30 [PEi386] Mask SYM_TYPE_DUP_DISCARD in makeSymbolExtra 48e391952c17ff7eab10b0b1456e3f2a2af28a9b introduced `SYM_TYPE_DUP_DISCARD` to the bitfield. The linker however, failed to mask the `SYM_TYPE_DUP_DISCARD` value. Thus `== SYM_TYPE_CODE` comparisons easily failed. This lead to us relocating DATA lookups (GOT) into E8 (call) and E9 (jump) instructions. (cherry picked from commit 34f06334025521c2440ebedb0237697fbcc3c6de) - - - - - 97679725 by Ilias Tsitsimpis at 2024-01-30T17:14:16+05:30 hadrian: Pass -DNOSMP to C compiler when needed Hadrian passes the -DNOSMP flag to GHC when the target doesn't support SMP, but doesn't pass it to CC as well, leading to the following compilation error on mips64el: | Run Cc (FindCDependencies CDep) Stage1: rts/sm/NonMovingScav.c => _build/stage1/rts/build/c/sm/NonMovingScav.o.d Command line: /usr/bin/mips64el-linux-gnuabi64-gcc -E -MM -MG -MF _build/stage1/rts/build/c/hooks/FlagDefaults.thr_debug_p_o.d -MT _build/stage1/rts/build/c/hooks/FlagDefaults.o -Irts/include -I_build/stage1/rts/build -I_build/stage1/rts/build/include -Irts/include -x c rts/hooks/FlagDefaults.c -Wall -Wextra -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Winline -Wpointer-arith -Wmissing-noreturn -Wnested-externs -Wredundant-decls -Wundef -fno-strict-aliasing -DTHREADED_RTS -DDEBUG -fomit-frame-pointer -O2 -g -Irts -I_build/stage1/rts/build -DDEBUG -fno-omit-frame-pointer -g3 -O0 ===> Command failed with error code: 1 In file included from rts/include/Stg.h:348, from rts/include/Rts.h:38, from rts/hooks/FlagDefaults.c:8: rts/include/stg/SMP.h:416:2: error: #error memory barriers unimplemented on this architecture 416 | #error memory barriers unimplemented on this architecture | ^~~~~ rts/include/stg/SMP.h:440:2: error: #error memory barriers unimplemented on this architecture 440 | #error memory barriers unimplemented on this architecture | ^~~~~ rts/include/stg/SMP.h:464:2: error: #error memory barriers unimplemented on this architecture 464 | #error memory barriers unimplemented on this architecture | ^~~~~ The old make system correctly passed this flag to both GHC and CC [1]. Fix this error by passing -DNOSMP to CC as well. [1] https://gitlab.haskell.org/ghc/ghc/-/blob/00920f176b0235d5bb52a8e054d89a664f8938fe/rts/ghc.mk#L407 Closes #24082 (cherry picked from commit 257c2807587624592813a42e06a05c5fc34cb38c) - - - - - 0ae27729 by Moritz Angermann at 2024-01-30T17:14:16+05:30 nativeGen: section flags for .text$foo only Commit 3ece9856d157c85511d59f9f862ab351bbd9b38b, was supposed to fix #22834 in !9810. It does however add "xr" indiscriminatly to .text sections even if splitSections is disabled. This leads to the assembler saying: ghc_1.s:7849:0: error: Warning: Ignoring changed section attributes for .text | 7849 | .section .text,"xr" | ^ (cherry picked from commit e99cf237f84db34be0468a893b10394d6b364bce) - - - - - 336c0c39 by Moritz Angermann at 2024-01-30T17:14:16+05:30 [PEi386 linker] Bounds check and null-deref guard We should resonably be able to expect that we won't exceed the number of sections if we assume to be dealing with legal object files. We can however not guarantee that we get some negative values, and while we try to special case most, we should exclude negative indexing into the sections array. We also need to ensure that we do not try to derefences targetSection, if it is NULL, due to the switch statement. (cherry picked from commit df81536f2e53abf521a05eb1e482a076f5849c21) - - - - - c036cf16 by Sylvain Henry at 2024-01-30T17:14:16+05:30 Avoid out-of-bound array access in bigNatIsPowerOf2 (fix #24066) bigNatIndex# in the `where` clause wasn't guarded by "bigNatIsZero a". (cherry picked from commit fe0675770b66a9ed393884d96e276b8d116fa2a2) - - - - - 14d23c31 by Sylvain Henry at 2024-01-30T17:14:16+05:30 Bignum: fix right shift of negative BigNat with native backend (cherry picked from commit cc1625b1ffbdf086b8380bacd35abc8d85861637) - - - - - cead9699 by Sylvain Henry at 2024-01-30T17:14:17+05:30 Rts: expose rtsOutOfBoundsAccess symbol (cherry picked from commit cbe4400d2690104053ec544cf7d0a9a13ee914ee) - - - - - 0ffbcb24 by Sylvain Henry at 2024-01-30T17:14:17+05:30 Hadrian: enable `-fcheck-prim-bounds` in validate flavour This allows T24066 to fail when the bug is present. Otherwise the out-of-bound access isn't detected as it happens in ghc-bignum which wasn't compiled with the bounds check. (cherry picked from commit 72c7380cb780933825bc84924908e01ce0495dc4) - - - - - ef2f3e02 by Profpatsch at 2024-01-30T17:14:17+05:30 base: Improve String & IsString documentation (cherry picked from commit d751c583d29460f033fefb45e685fa40fb3487ad) - - - - - b8af87e8 by Ben Gamari at 2024-01-30T17:14:17+05:30 rts/eventlog: Fix off-by-one in assertion Previously we failed to account for the NULL terminator `postString` asserted that there is enough room in the buffer for the string. (cherry picked from commit d0b17576148d336b67c7d65bcf742f83001413cb) - - - - - 86ac4f65 by Ben Gamari at 2024-01-30T17:14:17+05:30 rts/eventlog: Honor result of ensureRoomForVariableEvent is Previously we would keep plugging along, even if isn't enough room for the event. (cherry picked from commit a10f9b9bc510051a5b47d31238aad1174f7a1966) - - - - - 64de5b77 by Ben Gamari at 2024-01-30T17:14:17+05:30 rts/eventlog: Avoid truncating event sizes Previously ensureRoomForVariableEvent would truncate the desired size to 16-bits, resulting in #24197. Fixes #24197. (cherry picked from commit 0e0f41c0e3d9c67fc669e975060e88bccdc7d823) - - - - - efc34462 by Zubin Duggal at 2024-01-30T17:14:17+05:30 driver: Ensure we actually clear the interactive context before reloading Previously we called discardIC, but immediately after set the session back to an old HscEnv that still contained the IC Partially addresses #24107 Fixes #23405 (cherry picked from commit 58d56644c54b221f265b739829e53f9f5e1216d3) - - - - - 459dc2db by Zubin Duggal at 2024-01-30T17:14:17+05:30 driver: Ensure we force the lookup of old build artifacts before returning the build plan This prevents us from retaining all previous build artifacts in memory until a recompile finishes, instead only retaining the exact artifacts we need. Fixes #24118 (cherry picked from commit 8e5745a0cddf36afb680e1e99675279f0065ad7b) - - - - - 87d262af by Zubin Duggal at 2024-01-30T17:14:17+05:30 testsuite: add test for #24118 and #24107 MultiLayerModulesDefsGhci was not able to catch the leak because it uses :l which discards the previous environment. Using :r catches both of these leaks (cherry picked from commit 105c370ce8d2efe0295c3a50a5866f169b59b5e2) - - - - - a8e0c9ab by Zubin Duggal at 2024-01-30T17:14:17+05:30 compiler: Add some strictness annotations to ImportSpec and related constructors This prevents us from retaining entire HscEnvs. Force these ImportSpecs when forcing the GlobalRdrEltX Adds an NFData instance for Bag Fixes #24107 (cherry picked from commit e822ff88f7b139191163d0b2dafe43b318b779f9) - - - - - 67536e12 by Zubin Duggal at 2024-01-30T17:14:17+05:30 compiler: Force IfGlobalRdrEnv in NFData instance. (cherry picked from commit 522c12a43b34ad4ca7f3f916fa630d33a4fe6efb) - - - - - 057f4e24 by Claudio Bley at 2024-01-30T17:14:17+05:30 Only exit ghci in -e mode when :add command fails Previously, when running `ghci -e ':add Sample.hs'` the process would exit with exit code 1 if the file exists and could be loaded. Fixes #24115 (cherry picked from commit d561073727186c7b456c9ef113ccb7fc0df4560e) - - - - - 28fff112 by mmzk1526 at 2024-01-30T17:14:17+05:30 Use "-V" for alex version check for better backward compatibility Fixes #24302. In recent versions of alex, "-v" is used for "--verbose" instead of "-version". (cherry picked from commit c7be0c680d96ba6209c86e509ab2682e5041a53d) - - - - - 1b7b1345 by Ben Gamari at 2024-01-30T17:14:17+05:30 distrib: Rediscover otool and install_name_tool on Darwin In the bindist configure script we must rediscover the `otool` and `install_name_tool`s since they may be different from the build environment. Fixes #24211. (cherry picked from commit 292983c841b4facd5c48fcec9689448d66bcb90e) - - - - - f8a889b3 by Zubin Duggal at 2024-01-30T17:14:17+05:30 ci: Ensure we use the correct bindist name for the test artifact when generating release ghcup metadata Fixes #24268 (cherry picked from commit 989bf8e53c08eb22de716901b914b3607bc8dd08) - - - - - 9f8fd045 by Stefan Schulze Frielinghaus at 2024-01-30T17:14:17+05:30 llvmGen: Align objects in the data section Objects in the data section may be referenced via tagged pointers. Thus, align those objects to a 4- or 8-byte boundary for 32- or 64-bit platforms, respectively. Note, this may need to be reconsidered if objects with a greater natural alignment requirement are emitted as e.g. 128-bit atomics. Fixes #24163. (cherry picked from commit dfe1c3540e4b519b62b862b5966dfec5cae9ece1) - - - - - ff1c1cfb by Zubin Duggal at 2024-01-30T17:14:17+05:30 docs: document permissibility of -XOverloadedLabels (#24249) Document the permissibility introduced by https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0170-unrestricted-overloadedlabels.rst (cherry picked from commit c247b6befe6a599688bad0a3383424f7ea12d5f2) - - - - - 57a663f1 by Simon Peyton Jones at 2024-01-30T17:14:17+05:30 Take care when simplifying unfoldings This MR fixes a very subtle bug exposed by #24242. See Note [Environment for simplLetUnfolding]. I also updated a bunch of Notes on shadowing (cherry picked from commit d8baa1bdeea1753afc939a20119d3ce555301167) - - - - - e3f862f7 by Simon Peyton Jones at 2024-01-30T17:14:17+05:30 Make TYPE and CONSTRAINT not-apart Issue #24279 showed up a bug in the logic in GHC.Core.Unify.unify_ty which is supposed to make TYPE and CONSTRAINT be not-apart. Easily fixed. (cherry picked from commit af6932d6c068361c6ae300d52e72fbe13f8e1f18) - - - - - a9681fe1 by Zubin Duggal at 2024-01-30T17:14:17+05:30 ci: Fix typo in mk_ghcup_metadata.py There was a missing colon in the fix to #24268 in 989bf8e53c08eb22de716901b914b3607bc8dd08 (cherry picked from commit 4a39b5ffb85246d0dc491e36a80449afdbcc74cc) - - - - - ba4edca6 by sheaf at 2024-01-30T17:14:17+05:30 Use lookupOccRn_maybe in TH.lookupName When looking up a value, we want to be able to find both variables and record fields. So we should not use the lookupSameOccRn_maybe function, as we can't know ahead of time which record field namespace a record field with the given textual name will belong to. Fixes #24293 (cherry picked from commit c5fc7304d56c7a1e0a7bc6e53e23b976772fc10e) - - - - - 5ac5d8d2 by Zubin Duggal at 2024-02-07T15:22:43+05:30 #22362 and 22349 are fixed on the js backend - - - - - 661e2a46 by Greg Steuck at 2024-02-08T15:28:41+05:30 Remove undefined FP_PROG_LD_BUILD_ID from configure.ac's (cherry picked from commit 6f904808c925991bbaf4068c9a12b584675c6209) - - - - - d1a6ad12 by Sylvain Henry at 2024-02-08T15:37:59+05:30 JS: implement some file primitives (lstat,rmdir) (#22374) - Implement lstat and rmdir. - Implement base_c_s_is* functions (testing a file type) - Enable passing tests (cherry picked from commit 74a4dd2ec6e200b11a56b6f82907feb66e94c90b) - - - - - 8986d86d by jade at 2024-02-08T15:40:03+05:30 Expand documentation of List & Data.List This commit aims to improve the documentation and examples of symbols exported from Data.List (cherry picked from commit ff81d53f6404867c7cdd9dde5bb6bf3776912048) - - - - - 2fde790f by Jade at 2024-02-08T15:41:33+05:30 Improve documentation of Semigroup & Monoid This commit aims to improve the documentation of various symbols exported from Data.Semigroup and Data.Monoid (cherry picked from commit fa4e5913251786f2b535b31abd3fad39da8b3602) - - - - - f36b8ba7 by Finley McIlwaine at 2024-02-08T15:44:05+05:30 Add -dipe-stats flag This is useful for seeing which info tables have information. (cherry picked from commit cc52c358316ac8210f80da80db6b0c620dd5bdc3) - - - - - c237fe04 by Finley McIlwaine at 2024-02-08T19:53:20+05:30 Add -finfo-table-map-with-fallback -finfo-table-map-with-stack The -fno-info-table-map-with-stack flag omits STACK info tables from the info table map, and the -fno-info-table-map-with-fallback flag omits info tables with defaulted source locations from the map. In a test on the Agda codebase the build results were about 7% smaller when both of those types of tables were omitted. Adds a test that verifies that passing each combination of these flags results in the correct output for -dipe-stats, which is disabled for the js backend since profiling is not implemented. This commit also refactors a lot of the logic around extracting info tables from the Cmm results and building the info table map. This commit also fixes some issues in the users guide rst source to fix warnings that were noticed while debugging the documentation for these flags. Fixes #23702 (cherry picked from commit 261c4acbfdaf5babfc57ab0cef211edb66153fb1) - - - - - c5e5e6da by Finley McIlwaine at 2024-02-08T19:53:20+05:30 Refactor estimation of stack info table provenance This commit greatly refactors the way we compute estimated provenance for stack info tables. Previously, this process was done using an entirely separate traversal of the whole Cmm code stream to build the map from info tables to source locations. The separate traversal is now fused with the Cmm code generation pipeline in GHC.Driver.Main. This results in very significant code generation speed ups when -finfo-table-map is enabled. In testing, this patch reduces code generation times by almost 30% with -finfo-table-map and -O0, and 60% with -finfo-table-map and -O1 or -O2 . Fixes #23103 (cherry picked from commit d99c816f7b5727a3f344960e02a1932187ea093f) - - - - - 882f3b32 by Finley McIlwaine at 2024-02-08T19:53:20+05:30 Add a test checking overhead of -finfo-table-map We want to make sure we don't end up with poor codegen performance resulting from -finfo-table-map again as in #23103. This test adds a performance test tracking total allocations while compiling ExactPrint with -finfo-table-map. (cherry picked from commit d3e0124c1157a4a423d86a1dc1d7e82c6d32ef06) - - - - - a36e4c46 by Matthew Pickering at 2024-02-08T19:53:20+05:30 Add aarch64-deb11 bindist This adds a debian 11 release job for aarch64. Fixes #22005 (cherry picked from commit 02c87213e1215520d5496130a3082143f27035ae) - - - - - eda05046 by sheaf at 2024-02-08T19:53:20+05:30 Pass quantified tyvars in tcDefaultAssocDecl This commit passes the correct set of quantified type variables written by the user in associated type default declarations for validity checking. This ensures that validity checking of associated type defaults mirrors that of standalone type family instances. Fixes #23768 (see testcase T23734 in subsequent commit) (cherry picked from commit 5f826c180aa9b42ccb0af5763d87a48bee5e4083) - - - - - f8484716 by sheaf at 2024-02-08T19:53:20+05:30 Avoid panic in mkGADTVars This commit avoids panicking in mkGADTVars when we encounter a type variable as in #23784 that is bound by a user-written forall but not actually used. Fixes #23784 (cherry picked from commit aba184241a80e5abc804c4f22db91a16be215c11) - - - - - 5181eda8 by sheaf at 2024-02-08T19:53:20+05:30 Adjust reporting of unused tyvars in data FamInsts This commit adjusts the validity checking of data family instances to improve the reporting of unused type variables. See Note [Out of scope tvs in data family instances] in GHC.Tc.Validity. The problem was that, in a situation such as data family D :: Type data instance forall (d :: Type). D = MkD the RHS passed to 'checkFamPatBinders' would be the TyCon app R:D d which mentions the type variable 'd' quantified in the user-written forall. Thus, when computing the set of unused type variables in the RHS of the data family instance, we would find that 'd' is used, and report a strange error message that would say that 'd' is not bound on the LHS. To fix this, we special-case the data-family instance case, manually extracting all the type variables that appear in the arguments of all the data constructores of the data family instance. Fixes #23778 (cherry picked from commit a525a92a37cc202b6db09a827ba9df9eb8c9fa79) - - - - - 931a0c34 by Antoine Leblanc at 2024-02-08T23:01:53+05:30 Be more eager in TyCon boot validity checking This commit performs boot-file consistency checking for TyCons into checkValidTyCl. This ensures that we eagerly catch any mismatches, which prevents the compiler from seeing these inconsistencies and panicking as a result. See Note [TyCon boot consistency checking] in GHC.Tc.TyCl. Fixes #16127 (cherry picked from commit 1420b8cb8a7d6196eec80dc5293864c780379560) - - - - - e7cf6437 by sheaf at 2024-02-08T23:01:53+05:30 Unused tyvars in FamInst: only report user tyvars This commit changes how we perform some validity checking for coercion axioms to mirror how we handle default declarations for associated type families. This allows us to keep track of whether type variables in type and data family instances were user-written or not, in order to only report the user-written ones in "unused type variable" error messages. Consider for example: {-# LANGUAGE PolyKinds #-} type family F type instance forall a. F = () In this case, we get two quantified type variables, (k :: Type) and (a :: k); the second being user-written, but the first is introduced by the typechecker. We should only report 'a' as being unused, as the user has no idea what 'k' is. Fixes #23734 (cherry picked from commit 28dd52eec98e50c711cd00df22f6ab9e054c8b75) - - - - - ebc725b2 by sheaf at 2024-02-08T23:01:53+05:30 Validity: refactor treatment of data families This commit refactors the reporting of unused type variables in type and data family instances to be more principled. This avoids ad-hoc logic in the treatment of data family instances. (cherry picked from commit 1eed645c8b03b19a14cf58d9be5317cb81cbd30a) - - - - - dcd8ca74 by Krzysztof Gogolewski at 2024-02-08T23:01:53+05:30 docs: fix ScopedTypeVariables example (#24101) The previous example didn't compile. Furthermore, it wasn't demonstrating the point properly. I have changed it to an example which shows that 'a' in the signature must be the same 'a' as in the instance head. (cherry picked from commit 7a90020f167ab016cbfa95decafaa1a54a974bc6) - - - - - f47a95f2 by PHO at 2024-02-08T23:01:53+05:30 Don't assume the current locale is *.UTF-8, set the encoding explicitly primops.txt contains Unicode characters: > LC_ALL=C ./genprimopcode --data-decl < ./primops.txt > genprimopcode: <stdin>: hGetContents: invalid argument (cannot decode byte sequence starting from 226) Hadrian must also avoid using readFile' to read primops.txt because it tries to decode the file with a locale-specific encoding. (cherry picked from commit 52c0fc691e6501e99a96693ec1fc02e3c93a4fbc) - - - - - 73600e14 by Simon Peyton Jones at 2024-02-08T23:01:53+05:30 Add an extra check in kcCheckDeclHeader_sig Fix #24083 by checking for a implicitly-scoped type variable that is not actually bound. See Note [Disconnected type variables] in GHC.Tc.Gen.HsType For some reason, on aarch64-darwin we saw a 2.8% decrease in compiler allocations for MultiLayerModulesTH_Make; but 0.0% on other architectures. Metric Decrease: MultiLayerModulesTH_Make (cherry picked from commit 6dbab1808bfbe484b3fb396aab1d105314f918d8) - - - - - 39179586 by Simon Peyton Jones at 2024-02-08T23:01:53+05:30 Second fix to #24083 My earlier fix turns out to be too aggressive for data/type families See wrinkle (DTV1) in Note [Disconnected type variables] (cherry picked from commit 2776920e642544477a38d0ed9205d4f0b48a782e) (cherry picked from commit 110efc9813ba5f68668ab7fc3f95b9febca52880) - - - - - aa339700 by Matthew Pickering at 2024-02-08T23:01:53+05:30 libraries: Bump filepath to 1.4.200.1 and unix to 2.8.4.0 Updates filepath submodule Updates unix submodule Fixes #24240 (cherry picked from commit 36b9a38cc45a26865c4e45f4949e519a5dede76d) - - - - - 4058ecba by Matthew Pickering at 2024-02-08T23:01:54+05:30 Submodule linter: Allow references to tags We modify the submodule linter so that if the bumped commit is a specific tag then the commit is accepted. Fixes #24241 (cherry picked from commit 91ff0971df64b04938d011fe1562320c5d90849a) - - - - - ab835c3f by Matthew Craven at 2024-02-08T23:01:54+05:30 Fix loopification in the presence of void arguments This also removes Note [Void arguments in self-recursive tail calls], which was just misleading. It's important to count void args both in the function's arity and at the call site. Fixes #24295. (cherry picked from commit ae9cc1a84c9f470b77d98423400e6dfa95b2449b) - - - - - 275997f7 by Andreas Klebinger at 2024-02-08T23:01:54+05:30 Aarch64: Enable -mfma by default. Fixes #24311 (cherry picked from commit 7e95f738620dc805868d198f980e1bdd53e27a2d) - - - - - bb4f3878 by Matthew Pickering at 2024-02-08T23:01:54+05:30 eventlog: Fix off-by-one error in postIPE We were missing the extra_comma from the calculation of the size of the payload of postIPE. This was causing assertion failures when the event would overflow the buffer by one byte, as ensureRoomForVariable event would report there was enough space for `n` bytes but then we would write `n + 1` bytes into the buffer. Fixes #24287 (cherry picked from commit 5776008c7a5581193c3e72e59451ad49abac9d81) - - - - - 8fac10a2 by Jade at 2024-02-08T23:01:54+05:30 Enhance Documentation of functions exported by Data.Function This patch aims to improve the documentation of functions exported in Data.Function Tracking: #17929 Fixes: #10065 (cherry picked from commit 1fa1c00c95325761a5aa914af53d71ba5e7072b3) - - - - - 758ff906 by Jade at 2024-02-08T23:01:54+05:30 Improve documentation of hGetLine. - Add explanation for whether a newline is returned - Add examples Fixes #14804 (cherry picked from commit ab47a43d64f6b7d4fc181645171c31ba2db1eebe) - - - - - 5205b93d by sheaf at 2024-02-08T23:01:54+05:30 Fix FMA instruction on LLVM We were emitting the wrong instructions for fused multiply-add operations on LLVM: - the instruction name is "llvm.fma.f32" or "llvm.fma.f64", not "fmadd" - LLVM does not support other instructions such as "fmsub"; instead we implement these by flipping signs of some arguments - the instruction is an LLVM intrinsic, which requires handling it like a normal function call instead of a machine instruction Fixes #24223 (cherry picked from commit a40f4ab21bcc088e63892cd5e85edbec20d3fc69) - - - - - 7ddb1418 by Hécate Moonlight at 2024-02-08T23:01:54+05:30 Clarification for newtype constructors when using `coerce` (cherry picked from commit 699da01bbbf3e42c7d38b9cfe443dd0e8e256342) - - - - - 4593f490 by Andreas Klebinger at 2024-02-08T23:01:54+05:30 Fix fma warning when using llvm on aarch64. On aarch64 fma is always on so the +fma flag doesn't exist for that target. Hence no need to try and pass +fma to llvm. Fixes #24379 (cherry picked from commit 9294a08643b89509a0e0957cb73c186a39d4f3db) - - - - - 10de6db4 by Patrick at 2024-02-08T23:01:54+05:30 Fix bug wrong span of nested_doc_comment #24378 close #24378 1. Update the start position of span in `nested_doc_comment` correctly. and hence the spans of identifiers of haddoc can be computed correctly. 2. add test `HaddockSpanIssueT24378`. (cherry picked from commit 8eeadfad3a0035f8c5b339782676ff23572e0e5e) - - - - - 85ef42ce by sheaf at 2024-02-08T23:01:54+05:30 No shadowing warnings for NoFieldSelector fields This commit ensures we don't emit shadowing warnings when a user shadows a field defined with NoFieldSelectors. Fixes #24381 (cherry picked from commit ced2e7312b692e3f5402e4db6cfec390653a6a06) - - - - - e59bae2a by Teo Camarasu at 2024-02-08T23:01:54+05:30 doc: Add -Dn flag to user guide Resolves #24394 (cherry picked from commit 94ce031ddc84ee702c12a11793028ef21e65fa00) - - - - - 81cb68aa by Rodrigo Mesquita at 2024-02-08T23:01:54+05:30 Work around autotools setting C11 standard in CC/CXX In autoconf >=2.70, C11 is set by default for $CC and $CXX via the -std=...11 flag. In this patch, we split the "-std" flag out of the $CC and $CXX variables, which we traditionally assume to be just the executable name/path, and move it to $CFLAGS/$CXXFLAGS instead. Fixes #24324 (cherry picked from commit cdddeb0f1280b40cc194028bbaef36e127175c4c) - - - - - a7421910 by Zubin Duggal at 2024-02-08T23:01:54+05:30 driver: Really don't lose track of nodes when we fail to resolve cycles This fixes a bug in 8db8d2fd1c881032b1b360c032b6d9d072c11723, where we could lose track of acyclic components at the start of an unresolved cycle. We now ensure we never loose track of any of these components. As T24275 demonstrates, a "cyclic" SCC might not really be a true SCC: When viewed without boot files, we have a single SCC ``` [REC main:T24275B [main:T24275B {-# SOURCE #-}, main:T24275A {-# SOURCE #-}] main:T24275A [main:T24275A {-# SOURCE #-}]] ``` But with boot files this turns into ``` [NONREC main:T24275B {-# SOURCE #-} [], REC main:T24275B [main:T24275B {-# SOURCE #-}, main:T24275A {-# SOURCE #-}] main:T24275A {-# SOURCE #-} [main:T24275B], NONREC main:T24275A [main:T24275A {-# SOURCE #-}]] ``` Note that this is truly not an SCC, as no nodes are reachable from T24275B.hs-boot. However, we treat this entire group as a single "SCC" because it seems so when we analyse the graph without taking boot files into account. Indeed, we must return a single ResolvedCycle element in the BuildPlan for this as described in Note [Upsweep]. However, since after resolving this is not a true SCC anymore, `findCycle` fails to find a cycle and we have a sub-optimal error message as a result. To handle this, I extended `findCycle` to not assume its input is an SCC, and to try harder to find cycles in its input. Fixes #24275 (cherry picked from commit 532993c8160d960f848e7abd401774b6879e3ee8) - - - - - ba0a5d95 by Matthew Pickering at 2024-02-08T23:01:54+05:30 distrib/configure: Fix typo in CONF_GCC_LINKER_OPTS_STAGE2 variable Instead we were setting CONF_GCC_LINK_OPTS_STAGE2 which meant that we were missing passing `--target` when invoking the linker. Fixes #24414 (cherry picked from commit d309f4e7e37f3795c6d6b150c407d5a9b332854e) - - - - - b3300cb5 by Zubin Duggal at 2024-02-09T01:34:17+05:30 ci: Regenerate jobs.yaml - - - - - 042d45dd by Zubin Duggal at 2024-02-09T13:28:50+05:30 ci: Mark PartialDownsweep and OldModLocation tests as fragile on js backend - - - - - 55bbea03 by Rodrigo Mesquita at 2024-02-09T16:01:30+05:30 Suppress duplicate librares linker warning of new macOS linker Fixes #24167 XCode 15 introduced a new linker which warns on duplicate libraries being linked. To disable this warning, we pass -Wl,-no_warn_duplicate_libraries as suggested by Brad King in CMake issue #25297. This flag isn't necessarily available to other linkers on darwin, so we must only configure it into the CC linker arguments if valid. (cherry picked from commit e98051a5e7251390799f9fdead988c61d72e82e3) - - - - - 74d6260a by Rodrigo Mesquita at 2024-02-09T16:02:03+05:30 testsuite: Encoding test witnesses recent iconv bug is fragile A regression in the new iconv() distributed with XCode 15 and MacOS Sonoma causes the test 'encoding004' to fail in the CP936 roundrip. We mark this test as fragile until this is fixed upstream (rather than broken, since previous versions of iconv pass the test) See #24161 (cherry picked from commit c411c431e7ae1f0effbe9f9a624c7f9171d50f0a) - - - - - 0f6116ca by Rodrigo Mesquita at 2024-02-09T16:02:12+05:30 testsuite: Update to LC_ALL=C no longer being ignored in darwin MacOS seems to have fixed an issue where it used to ignore the variable `LC_ALL` in program invocations and default to using Unicode. Since the behaviour seems to be fixed to account for the locale variable, we mark tests that were previously broken in spite of it as fragile (since they now pass in recent macOS distributions) See #24161 (cherry picked from commit ce7fe5a916d50f471812f4714615e13f557fe57a) - - - - - b746f8a8 by Rodrigo Mesquita at 2024-02-09T16:05:28+05:30 darwin: Fix single_module is obsolete warning In XCode 15's linker, -single_module is the default and otherwise passing it as a flag results in a warning being raised: ld: warning: -single_module is obsolete This patch fixes this warning by, at configure time, determining whether the linker supports -single_module (which is likely false for all non-darwin linkers, and true for darwin linkers in previous versions of macOS), and using that information at runtime to decide to pass or not the flag in the invocation. Fixes #24168 (cherry picked from commit e6c803f702e8b09dfd0073b973b8afcd7071db50) (cherry picked from commit 273f5a3fd392d528664d5661508cc0094e37ec0d) - - - - - 0127a041 by Matthew Craven at 2024-02-19T13:53:39+05:30 Make 'wWarningFlagsDeps' include every WarningFlag Fixes #24071. (cherry picked from commit a2c0fff61afdb14b5f2624374aa5767e7b238ff4) - - - - - 34e38b38 by Torsten Schmits at 2024-02-19T13:53:39+05:30 Fix several mistakes around free variables in iface breakpoints Fixes #23612 , #23607, #23998 and #23666. MR: !11026 The fingerprinting logic in `Iface.Recomp` failed lookups when processing decls containing breakpoints for two reasons: * IfaceBreakpoint created binders for free variables instead of expressions * When collecting free names for the dependency analysis for fingerprinting, breakpoint FVs were skipped (cherry picked from commit d3874407df4223a5e14a43571f4cc344349a537d) - - - - - d07caf0e by Andrei Borzenkov at 2024-02-19T13:53:39+05:30 Add changelog entry for renaming tuples from (,,...,,) to Tuple<n> (24291) (cherry picked from commit 69abc7869bc504631e445083704115fc8a5d29c8) - - - - - fe4a1d24 by Zubin Duggal at 2024-02-19T13:53:39+05:30 ci: Allow release-hackage-lint to fail Otherwise it blocks the ghcup metadata pipeline from running. (cherry picked from commit 2e88063500d7ef33c83bd2de8ca5c7818ffbb026) - - - - - b8b00b53 by Ben Gamari at 2024-02-19T13:53:39+05:30 rts/EventLog: Place eliminate duplicate strlens Previously many of the `post*` implementations would first compute the length of the event's strings in order to determine the event length. Later we would then end up computing the length yet again in `postString`. Now we instead pass the string length to `postStringLen`, avoiding the repeated work. (cherry picked from commit 325b7613ebb2ca012a8969e20d35e95bfccc2bba) - - - - - b5b22004 by Ben Gamari at 2024-02-19T13:53:39+05:30 rts/eventlog: Place upper bound on IPE string field lengths The strings in IPE events may be of unbounded length. Limit the lengths of these fields to 64k characters to ensure that we don't exceed the maximum event length. (cherry picked from commit 8aafa51cb714fb16989089d4bc1ea7e7eb50124c) - - - - - 8bf88fa3 by Zubin Duggal at 2024-02-19T13:53:39+05:30 rts: drop unused postString function (cherry picked from commit 0e60d52cc7e261da11c37bd649511584d92a688b) - - - - - 2dc75024 by Simon Peyton Jones at 2024-02-19T13:53:39+05:30 Make decomposeRuleLhs a bit more clever This fixes #24370 by making decomposeRuleLhs undertand dictionary /functions/ as well as plain /dictionaries/ (cherry picked from commit ca2e919ecca35db412e772d7eadd6a7c4fb20e4b) - - - - - b1bd2eed by ARATA Mizuki at 2024-02-19T13:53:39+05:30 Support 128-bit SIMD on AArch64 via LLVM backend (cherry picked from commit 015886ec78e598f850c4202efdee239bac63b8c7) - - - - - 5396d8ce by ARATA Mizuki at 2024-02-19T13:53:39+05:30 x86: Don't require -mavx2 when using 256-bit floating-point SIMD primitives Fixes #24222 (cherry picked from commit 7d9a2e44e8cce00e24671325aebe47d9e529aee5) - - - - - 2b9b7e77 by Ben Gamari at 2024-02-19T13:53:39+05:30 Fix thunk update ordering Previously we attempted to ensure soundness of concurrent thunk update by synchronizing on the access of the thunk's info table pointer field. This was believed to be sufficient since the indirectee (which may expose a closure allocated by another core) would not be examined until the info table pointer update is complete. However, it turns out that this can result in data races in the presence of multiple threads racing a update a single thunk. For instance, consider this interleaving under the old scheme: Thread A Thread B --------- --------- t=0 Enter t 1 Push update frame 2 Begin evaluation 4 Pause thread 5 t.indirectee=tso 6 Release t.info=BLACKHOLE 7 ... (e.g. GC) 8 Resume thread 9 Finish evaluation 10 Relaxed t.indirectee=x 11 Load t.info 12 Acquire fence 13 Inspect t.indirectee 14 Release t.info=BLACKHOLE Here Thread A enters thunk `t` but is soon paused, resulting in `t` being lazily blackholed at t=6. Then, at t=10 Thread A finishes evaluation and updates `t.indirectee` with a relaxed store. Meanwhile, Thread B enters the blackhole. Under the old scheme this would introduce an acquire-fence but this would only synchronize with Thread A at t=6. Consequently, the result of the evaluation, `x`, is not visible to Thread B, introducing a data race. We fix this by treating the `indirectee` field as we do all other mutable fields. This means we must always access this field with acquire-loads and release-stores. See #23185. (cherry picked from commit 9a52ae46a33b490161e1e3e1cc70caa46c60488a) (cherry picked from commit 88afc6ea885d54523efbbb764f3435a147b799a5) - - - - - 92239ede by Florian Weimer at 2024-02-20T13:38:04+05:30 Fix C output for modern C initiative GCC 14 on aarch64 rejects the C code written by GHC with this kind of error: error: assignment to ‘ffi_arg’ {aka ‘long unsigned int’} from ‘HsPtr’ {aka ‘void *’} makes integer from pointer without a cast [-Wint-conversion] 68 | *(ffi_arg*)resp = cret; | ^ Add the correct cast. For more information on this see: https://fedoraproject.org/wiki/Changes/PortingToModernC Tested-by: Richard W.M. Jones <rjones at redhat.com> (cherry picked from commit 1f534c2e7388273e70534680212c1357614c11ed) - - - - - d13e2e8f by Fendor at 2024-02-20T19:36:34+05:30 Prefer RdrName over OccName for looking up locations in doc renaming step Looking up by OccName only does not take into account when functions are only imported in a qualified way. Fixes issue #24294 Bump haddock submodule to include regression test (cherry picked from commit b57200de601e4ef6827727176611d7192016b8b2) - - - - - e0e4fb95 by Andrew Lelechenko at 2024-02-20T19:36:34+05:30 Bump submodule text to 2.1.1 T17123 allocates less because of improvements to Data.Text.concat in 1a6a06a. Metric Decrease: T17123 (cherry picked from commit 56e3f097fa7205f77e7903af345ed3798ecb039e) - - - - - ba1064a1 by Matthew Craven at 2024-02-20T19:36:34+05:30 Bump bytestring submodule to something closer to 0.12.1 ...mostly so that 16d6b7e835ffdcf9b894e79f933dd52348dedd0c (which reworks unaligned writes in Builder) and the stuff in https://github.com/haskell/bytestring/pull/631 can see wider testing. The less-terrible code for unaligned writes used in Builder on hosts not known to be ulaigned-friendly also takes less effort for GHC to compile, resulting in a metric decrease for T21839c on some platforms. The metric increase on T21839r is caused by the unrelated commit 750dac33465e7b59100698a330b44de7049a345c. It perhaps warrants further analysis and discussion (see #23822) but is not critical. Metric Decrease: T21839c Metric Increase: T21839r (cherry picked from commit 27020458220da55f52f1f94c42a6ae7b4f321387) - - - - - 049d5239 by Matthew Craven at 2024-02-20T19:36:34+05:30 Bump bytestring submodule to 0.12.1.0 (cherry picked from commit 5d3f786264db88a758ae65277c1b8d7f37f2e460) - - - - - 9a192bc1 by Matthew Pickering at 2024-02-20T19:36:34+05:30 driver: Check transitive closure of haskell package dependencies when deciding whether to relink We were previously just checking whether direct package dependencies had been modified. This caused issues when compiling without optimisations as we wouldn't relink the direct dependency if one of its dependenices changed. Fixes #23724 (cherry picked from commit 291d81aef8083290da0d2ce430fbc5e5a33bdb6e) - - - - - 87790f48 by Cheng Shao at 2024-02-20T19:36:34+05:30 Fix genapply for cross-compilation by nuking fragile CPP logic This commit fixes incorrectly built genapply when cross compiling (#24347) by nuking all fragile CPP logic in it from the orbit. All target-specific info are now read from DerivedConstants.h at runtime, see added note for details. Also removes a legacy Makefile and adds haskell language server support for genapply. (cherry picked from commit dd4af0e5a4a3af208bdf57e8237d85261eef24f7) (cherry picked from commit ee800873f62fac8c67cb7034f942a1ed6b72c032) - - - - - 1e695750 by Cheng Shao at 2024-02-20T19:36:34+05:30 rts: enable wasm32 register mapping The wasm backend didn't properly make use of all Cmm global registers due to #24347. Now that it is fixed, this patch re-enables full register mapping for wasm32, and we can now generate smaller & faster wasm modules that doesn't always spill arguments onto the stack. Fixes #22460 #24152. (cherry picked from commit 0cda2b8b15cdbc44c45ffa36a37ed8c2fe8b8b9c) (cherry picked from commit f1f5068b398b1effb837add38ecc5303dc9a381f) - - - - - f60cf966 by Matthew Pickering at 2024-02-20T23:50:08+05:30 testsuite: Fix T21097b test with make 4.1 (deb9) cee81370cd6ef256f66035e3116878d4cb82e28b recently added a test which failed on deb9 because the version of make was emitting the recipe failure to stdout rather than stderr. One way to fix this is to be more precise in the test about which part of the output we care about inspecting. (cherry picked from commit bad3765668cc5badf5d0a19100fac95125985473) - - - - - bfe59b88 by Zubin Duggal at 2024-02-20T23:50:08+05:30 Prepare release 9.8.2 Metric Decrease: T13386 - - - - - f15dd7bd by Zubin Duggal at 2024-02-21T14:10:08+05:30 testsuite: Mark linker_unload_native as fragile See #23993. This test is fragile on Alpine (dynamic) but we don't have a way to mark it as fragile on only that platform, so marking it as fragile on all platforms. (cherry picked from commit 6819b70a7739205a75f0b4fefcfcc9fdab39cab9) - - - - - 2eca9c7a by Ben Gamari at 2024-02-22T18:31:44+05:30 hadrian: Generate HSC2HS_EXTRAS variable in bindist installation We must generate the hsc2hs wrapper at bindist installation time since it must contain `--lflag` and `--cflag` arguments which depend upon the installation path. The solution here is to substitute these variables in the configure script (see mk/hsc2hs.in). This is then copied over a dummy wrapper in the install rules. Fixes #24050. (cherry picked from commit efcbad2dfd242d0bc2c91da5390fe8456a536cc3) - - - - - d5246e19 by Matthew Pickering at 2024-02-22T18:31:58+05:30 ci: Show --info for installed compiler (cherry picked from commit c540559cf188625bec668fa6cd94d4f94413d730) - - - - - f3225ed4 by Zubin Duggal at 2024-02-22T23:42:18+05:30 Accept change in MultiLayerModulesTH_Make This test is flaky on darwin Metric Decrease: MultiLayerModulesTH_Make - - - - - 1c08e245 by Ben Gamari at 2024-02-23T12:31:12+05:30 testsuite: Ignore stderr in T8089 Otherwise spurious "Killed: 9" messages to stderr may cause the test to fail. Fixes #24361. (cherry picked from commit e693a4e8589bad35588c51fccc87f4388e7d5874) - - - - - 9cb7e73a by Sebastian Graf at 2024-05-06T15:11:25+02:00 exprIsTrivial: Factor out shared implementation The duplication between `exprIsTrivial` and `getIdFromTrivialExpr_maybe` has been bugging me for a long time. This patch introduces an inlinable worker function `trivial_expr_fold` acting as the single, shared decision procedure of triviality. It "returns" a Church-encoded `Maybe (Maybe Id)`, so when it is inlined, it fuses to similar code as before. (Better code, even, in the case of `getIdFromTrivialExpr` which presently allocates a `Just` constructor that cancels away after this patch.) - - - - - 78a25354 by Sebastian Graf at 2024-05-07T09:25:04+02:00 Some cherry-picked bits of 59202c8 to fix #24718 As noted in f3225ed4b3f3c4, the test below is flaky on Darwin. Metric Decrease: MultiLayerModulesTH_Make - - - - - 5e5ba257 by Ben Gamari at 2024-09-23T14:33:41-04:00 Bump process submodule to 1.6.24.0 - - - - - 8d7a027a by Ben Gamari at 2024-09-23T14:36:47-04:00 Bump stm submodule to v2.5.3.1 - - - - - a2a44e47 by Andrew Lelechenko at 2024-09-27T13:31:13-04:00 Bump submodule deepseq to 1.5.1.0 (cherry picked from commit 8e462f4d4bdf2a6c34c249e7be8084565600d300) - - - - - d747384b by Cheng Shao at 2024-09-27T13:31:13-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - 28dd05e8 by Cheng Shao at 2024-09-27T13:31:13-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 43950801 by Cheng Shao at 2024-09-27T13:31:13-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 230a1fc0 by Cheng Shao at 2024-09-27T13:31:13-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fefd77f6 by Andrew Lelechenko at 2024-09-27T14:57:23-04:00 Bump submodule array to 0.5.8.0 (cherry picked from commit 80769bc9f56541601796366485283a697c52a18b) - - - - - 97cad96c by Matthew Pickering at 2024-09-27T15:46:45-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 (cherry picked from commit 7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6) - - - - - 8f3d1a6a by Sylvain Henry at 2024-09-27T15:48:36-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 (cherry picked from commit e7a26d7a6faf1ea534e036c5085a0a027dbb6f5f) - - - - - 9f2e0f50 by Sylvain Henry at 2024-09-27T17:02:50-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 (cherry picked from commit 5092dbff750ee5b6fd082b7eed8574922a2b0bf4) - - - - - 2101a57a by Sylvain Henry at 2024-09-27T17:02:50-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. (cherry picked from commit c749bdfd3e21d712dc2b966482eb010165bdeebe) - - - - - 08963d26 by Sylvain Henry at 2024-09-27T17:02:50-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> (cherry picked from commit 7446a09a2d5b04b95cd43c03659b5647853124ce) - - - - - 257e836d by sheaf at 2024-09-27T17:02:50-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 (cherry picked from commit c29b2b5a77611b2bd6c3089765079bc43aec3e22) - - - - - b0947c95 by Sylvain Henry at 2024-09-27T17:02:50-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 (cherry picked from commit f954f42823f6ca3588425a0d543d93ace86d89e4) - - - - - 12460fb2 by Simon Peyton Jones at 2024-09-27T17:02:50-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify (cherry picked from commit de5d9852dbdd367611bf9e45e69c723d26351992) - - - - - b2b87083 by doyougnu at 2024-09-27T17:02:50-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches (cherry picked from commit bfe4b3d3bbb98b39169fad063c6c32f06d167756) - - - - - f77fa175 by Ben Gamari at 2024-09-27T17:12:21-04:00 docs: Add release notes for 9.8.3 - - - - - fff36b5f by Andreas Klebinger at 2024-09-27T17:14:06-04:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. (cherry picked from commit 3f89ab92da74c4ed45da68fe92ff81e7b9caa53d) - - - - - 34ff1d29 by Peter Trommler at 2024-09-27T17:51:50-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 (cherry picked from commit 7fe85b1354a13749f14d588e3cc742b8ae2d8da9) - - - - - 21fb1084 by Ben Gamari at 2024-09-27T17:52:08-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 (cherry picked from commit a82121b3b6fdc2ac47211f71871b3ab21e5f6276) - - - - - 9de7ea1d by Sylvain Henry at 2024-09-27T18:38:10-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). (cherry picked from commit 5104ee615503617a1c124fe1d92f6aa2d263b7d0) - - - - - a215f205 by Sylvain Henry at 2024-09-27T18:38:10-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC (cherry picked from commit 52d6698479f951e07def237b0474ee22d27e621a) - - - - - 736b7a38 by Sylvain Henry at 2024-09-27T18:38:10-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). (cherry picked from commit c34fef56367142fa55e9861092f64cc7b9946fa1) - - - - - 7ec14266 by Cheng Shao at 2024-09-27T18:38:31-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. (cherry picked from commit 0cff083abb24701530974872b21cf897c9955a9a) - - - - - 075032b8 by Andreas Klebinger at 2024-09-27T18:38:41-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. (cherry picked from commit 1bfa91115b8320ed99a5e946147528e21ca4f3e1) - - - - - 750d0b1c by crumbtoo at 2024-09-27T18:39:00-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 (cherry picked from commit c5e00c35927d574f71bf77449817b131d1749750) - - - - - f064b0f0 by Ryan Scott at 2024-09-27T18:39:34-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. (cherry picked from commit a3cd3a1d0d186f2aa4d0273c6b3e74a442de2ef0) - - - - - af9d21c1 by Sylvain Henry at 2024-09-27T18:39:48-04:00 Reverse arguments to stgCallocBytes (fix #24828) (cherry picked from commit 6838a7c32ca29b5d44adc9d6280d3a960f31be7c) - - - - - 5cdb0927 by Teo Camarasu at 2024-09-27T18:40:28-04:00 doc: Fix type error in hs_try_putmvar example (cherry picked from commit 06f7db4001e4eee0f3076d949876f8f4af0eb6fb) - - - - - b05463d3 by Matthew Craven at 2024-09-27T18:40:57-04:00 Add test cases for #24664 ...since none are present in the original MR !12463 fixing this issue. (cherry picked from commit a19201d42cfd3aa54faeb1b5a95b715b9a67a01a) - - - - - a41e8c45 by Zubin Duggal at 2024-09-27T18:42:34-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 (cherry picked from commit dddc9dff0547733a10e7f505612ab9df3a7c21b6) - - - - - f1fe0c4f by Sylvain Henry at 2024-09-27T18:43:07-04:00 JS: correctly handle RUBBISH literals (#24664) (cherry picked from commit daeda83478d5b800d29661408dd67cc4b23df374) - - - - - 9280ee6d by Matthew Pickering at 2024-09-27T18:43:32-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. (cherry picked from commit 3fff09779d5830549ae455a15907b7bb9fe7859a) - - - - - f74bf352 by Teo Camarasu at 2024-09-27T18:45:01-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 (cherry picked from commit f30e4984fb048818051465698ef8e4e20dacb577) - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitlab/rel_eng/upload_ghc_libs.py - .gitlab/test-metrics.sh - compiler/CodeGen.Platform.h - compiler/GHC.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Cond.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/BlockLayout.hs - compiler/GHC/CmmToAsm/Monad.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/PPC/Instr.hs - compiler/GHC/CmmToAsm/Ppr.hs - compiler/GHC/CmmToAsm/Reg/Graph/TrivColorable.hs - compiler/GHC/CmmToAsm/Wasm/FromCmm.hs - compiler/GHC/CmmToAsm/X86.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e4b5cdbdee243e50cc417e1da9507a78222bfb19...f74bf35293e5054c21fdc5d5d7e123bc2489382e -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e4b5cdbdee243e50cc417e1da9507a78222bfb19...f74bf35293e5054c21fdc5d5d7e123bc2489382e You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Fri Sep 27 23:29:00 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Fri, 27 Sep 2024 19:29:00 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8-2] 3 commits: JS: correctly handle RUBBISH literals (#24664) Message-ID: <66f73fbc8f747_34c29b4ef41c100147@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8-2 at Glasgow Haskell Compiler / GHC Commits: 35cb1c86 by Sylvain Henry at 2024-09-27T19:28:47-04:00 JS: correctly handle RUBBISH literals (#24664) (cherry picked from commit daeda83478d5b800d29661408dd67cc4b23df374) - - - - - c35622b2 by Matthew Pickering at 2024-09-27T19:28:47-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. (cherry picked from commit 3fff09779d5830549ae455a15907b7bb9fe7859a) - - - - - 31012d41 by Teo Camarasu at 2024-09-27T19:28:48-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 (cherry picked from commit f30e4984fb048818051465698ef8e4e20dacb577) - - - - - 4 changed files: - compiler/GHC/StgToJS/Literal.hs - docs/index.html.in - hadrian/src/Hadrian/Haskell/Cabal/Parse.hs - hadrian/src/Rules/Generate.hs Changes: ===================================== compiler/GHC/StgToJS/Literal.hs ===================================== @@ -67,7 +67,28 @@ genLit = \case | otherwise -> return [ toJExpr (TxtI (mkRawSymbol True name)) , ValExpr (JInt 0) ] - LitRubbish {} -> return [ null_ ] + LitRubbish _ rr_ty -> + -- Generate appropriate rubbish literals, otherwise it might trip up the + -- code generator when a primop is applied to a rubbish literal (see #24664) + let reps = runtimeRepPrimRep (text "GHC.StgToJS.Literal.genLit") rr_ty + rub = \case + BoxedRep _ -> [ null_ ] + AddrRep -> [ null_, ValExpr (JInt 0) ] + WordRep -> [ ValExpr (JInt 0) ] + Word8Rep -> [ ValExpr (JInt 0) ] + Word16Rep -> [ ValExpr (JInt 0) ] + Word32Rep -> [ ValExpr (JInt 0) ] + Word64Rep -> [ ValExpr (JInt 0), ValExpr (JInt 0) ] + IntRep -> [ ValExpr (JInt 0) ] + Int8Rep -> [ ValExpr (JInt 0) ] + Int16Rep -> [ ValExpr (JInt 0) ] + Int32Rep -> [ ValExpr (JInt 0) ] + Int64Rep -> [ ValExpr (JInt 0), ValExpr (JInt 0) ] + DoubleRep -> [ ValExpr (JInt 0) ] + FloatRep -> [ ValExpr (JInt 0) ] + VoidRep -> panic "GHC.StgToJS.Literal.genLit: LitRubbish(VoidRep)" + VecRep _ _ -> panic "GHC.StgToJS.Literal.genLit: VecRep unsupported" + in return (concatMap rub reps) -- | generate a literal for the static init tables genStaticLit :: Literal -> G [StaticLit] ===================================== docs/index.html.in ===================================== @@ -39,7 +39,7 @@

  • - GHC API + GHC API

    Documentation for the GHC API. ===================================== hadrian/src/Hadrian/Haskell/Cabal/Parse.hs ===================================== @@ -31,6 +31,7 @@ import qualified Distribution.PackageDescription.Parsec as C import qualified Distribution.Simple.Compiler as C import qualified Distribution.Simple.Program.Db as C import qualified Distribution.Simple as C +import qualified Distribution.Simple.GHC as GHC import qualified Distribution.Simple.Program.Builtin as C import qualified Distribution.Simple.Utils as C import qualified Distribution.Simple.Program.Types as C @@ -342,12 +343,11 @@ registerPackage rs context = do need [setupConfig] -- This triggers 'configurePackage' pd <- packageDescription <$> readContextData context db_path <- packageDbPath (PackageDbLoc (stage context) (iplace context)) - dist_dir <- Context.buildPath context pid <- pkgUnitId (stage context) (package context) -- Note: the @cPath@ is ignored. The path that's used is the 'buildDir' path -- from the local build info @lbi at . lbi <- liftIO $ C.getPersistBuildConfig cPath - liftIO $ register db_path pid dist_dir pd lbi + liftIO $ register db_path pid pd lbi -- Then after the register, which just writes the .conf file, do the recache step. buildWithResources rs $ target context (GhcPkg Recache (stage context)) [] [] @@ -356,25 +356,23 @@ registerPackage rs context = do -- into a different package database to the one it was configured against. register :: FilePath -> String -- ^ Package Identifier - -> FilePath -> C.PackageDescription -> LocalBuildInfo -> IO () -register pkg_db pid build_dir pd lbi +register pkg_db pid pd lbi = withLibLBI pd lbi $ \lib clbi -> do - absPackageDBs <- C.absolutePackageDBPaths packageDbs - installedPkgInfo <- C.generateRegistrationInfo - C.silent pd lib lbi clbi False reloc build_dir - (C.registrationPackageDB absPackageDBs) - + when reloc $ error "register does not support reloc" + installedPkgInfo <- generateRegistrationInfo pd lbi lib clbi writeRegistrationFile installedPkgInfo where regFile = pkg_db pid <.> "conf" reloc = relocatable lbi - -- Using a specific package db here is why we have to copy the function from Cabal. - packageDbs = [C.SpecificPackageDB pkg_db] + + generateRegistrationInfo pkg lbi lib clbi = do + abi_hash <- C.mkAbiHash <$> GHC.libAbiHash C.silent pkg lbi lib clbi + return (C.absoluteInstalledPackageInfo pkg abi_hash lib lbi clbi) writeRegistrationFile installedPkgInfo = do writeUTF8File regFile (CP.showInstalledPackageInfo installedPkgInfo) ===================================== hadrian/src/Rules/Generate.hs ===================================== @@ -341,7 +341,7 @@ templateRules = do templateRule "utils/ghc-pkg/ghc-pkg.cabal" $ projectVersion templateRule "libraries/template-haskell/template-haskell.cabal" $ projectVersion templateRule "libraries/prologue.txt" $ packageVersions - templateRule "docs/index.html" $ packageVersions + templateRule "docs/index.html" $ packageUnitIds templateRule "doc/users_guide/ghc_config.py" $ packageUnitIds View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f74bf35293e5054c21fdc5d5d7e123bc2489382e...31012d411b9f307debd3f0f19daa1fcd07326ee8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f74bf35293e5054c21fdc5d5d7e123bc2489382e...31012d411b9f307debd3f0f19daa1fcd07326ee8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 28 07:51:52 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Sat, 28 Sep 2024 03:51:52 -0400 Subject: [Git][ghc/ghc] Deleted branch wip/9.13-start Message-ID: <66f7b59893358_10a61729a06831929@gitlab.mail> Matthew Pickering deleted branch wip/9.13-start at Glasgow Haskell Compiler / GHC -- You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 28 07:52:05 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Sat, 28 Sep 2024 03:52:05 -0400 Subject: [Git][ghc/ghc][master] 5 commits: Fix typo in Prelude doc for (>>=) Message-ID: <66f7b5a5b661c_10a617221d703219b@gitlab.mail> Matthew Pickering pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - 30 changed files: - .gitlab/ci.sh - compiler/GHC/Hs/Decls.hs - compiler/GHC/Hs/Instances.hs - compiler/GHC/Hs/Type.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Rename/Module.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/GHC/Tc/Gen/Sig.hs - compiler/GHC/Tc/TyCl.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Error/Codes.hs - compiler/Language/Haskell/Syntax/Extension.hs - compiler/Language/Haskell/Syntax/Type.hs - configure.ac - − docs/users_guide/9.12.1-notes.rst - + docs/users_guide/9.14.1-notes.rst - docs/users_guide/exts/type_abstractions.rst - libraries/ghc-internal/src/GHC/Internal/Base.hs - testsuite/tests/haddock/should_compile_flag_haddock/T17544.stderr - testsuite/tests/haddock/should_compile_flag_haddock/T17544_kw.stderr - testsuite/tests/parser/should_compile/DumpParsedAst.stderr - testsuite/tests/parser/should_compile/DumpRenamedAst.stderr - testsuite/tests/parser/should_compile/DumpSemis.stderr - testsuite/tests/parser/should_compile/KindSigs.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6...e4ac1b0d281b85a0144d1ef6f84a1df00e236052 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6...e4ac1b0d281b85a0144d1ef6f84a1df00e236052 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 28 08:00:48 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Sat, 28 Sep 2024 04:00:48 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 7 commits: Fix typo in Prelude doc for (>>=) Message-ID: <66f7b7b05cbaf_10a617255c10396e@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - 5a103a4b by Andreas Klebinger at 2024-09-28T04:00:24-04:00 ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps This will be the new place for functions that would have gone into GHC.Exts in the past but are not stable enough to do so now. Addresses #25242 - - - - - f678355c by Sylvain Henry at 2024-09-28T04:00:35-04:00 RTS: cleanup timerfd file descriptors after a fork (#25280) When we init a timerfd-based ticker, we should be careful to cleanup the old file descriptors (e.g. after a fork). - - - - - 30 changed files: - .gitlab/ci.sh - compiler/GHC/Hs/Decls.hs - compiler/GHC/Hs/Instances.hs - compiler/GHC/Hs/Type.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Rename/Module.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/GHC/Tc/Gen/Sig.hs - compiler/GHC/Tc/TyCl.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Error/Codes.hs - compiler/Language/Haskell/Syntax/Extension.hs - compiler/Language/Haskell/Syntax/Type.hs - configure.ac - − docs/users_guide/9.12.1-notes.rst - + docs/users_guide/9.14.1-notes.rst - docs/users_guide/exts/type_abstractions.rst - libraries/base/src/GHC/Exts.hs - libraries/ghc-experimental/ghc-experimental.cabal.in - + libraries/ghc-experimental/src/GHC/PrimOps.hs - libraries/ghc-internal/src/GHC/Internal/Base.hs - libraries/ghc-internal/src/GHC/Internal/Exts.hs - rts/posix/ticker/TimerFd.c - testsuite/tests/haddock/should_compile_flag_haddock/T17544.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1877884729ffd43bf02a7936c4ead50f1a9ba2dc...f678355cebdc1432e52514ecf8f27dd5066ccead -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1877884729ffd43bf02a7936c4ead50f1a9ba2dc...f678355cebdc1432e52514ecf8f27dd5066ccead You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 28 09:50:55 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Sat, 28 Sep 2024 05:50:55 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] 60 commits: ci: Run abi-test on test-abi label Message-ID: <66f7d17f8b2bc_3dd6b73ebebc188e6@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - d5248714 by Jade at 2024-09-28T11:44:28+02:00 Refactor FieldOcc vs AmbiguousFieldOcc with TTG Improves the design of FieldOcc vs AmbiguousFieldOcc, and removes a dependency on `RdrName` from the Language.Haskell.* namespace (#21592). The design: * The FieldOcc constructor of FieldOcc always refers to an unambiguous field occurrence. * During renaming, a FieldOcc may be ambiguous and only be resolvable during Typechecking * Therefore, we extend (with TTG) `FieldOcc GhcRn` with a constructor `AmbiguousFieldOcc` that constructs a definitely ambiguous `FieldOcc`. * During typechecking, all ambiguous field occurrences must be resolved, so the `AmbiguousFieldOcc` constructor no longer exists See Note [Lifecycle of a FieldOcc] Co-authored-by: Rodrigo Mesquita <rodrigo.m.mesquita at gmail.com> - - - - - f0aac3b6 by Hassan Al-Awwadi at 2024-09-28T11:49:50+02:00 Wrangled until everything compiled. There are two issues: - FieldOcc used to only have one constructor and always be unambigious, this is no longer the case. Calls to foLabel are thus partial now. Don't know how much we care about this, since the partial calls are mostly inside functions that used to operate on the operate on the unambigious FieldOcc - Lots of functions that take in a FieldOcc, or a HsExpr (with the HsRecSel constructor) now have a new case. It was not always clear to me what the correct implementation was for these. I have filled them in as far as I could and left one undefined... - - - - - 8f3ff8c5 by Hassan Al-Awwadi at 2024-09-28T11:49:50+02:00 fixed ambiguity regression - - - - - ad44abb0 by Hassan Al-Awwadi at 2024-09-28T11:49:50+02:00 Apply 2 suggestion(s) to 2 file(s) Co-authored-by: Rodrigo Mesquita <rodrigo.m.mesquita at gmail.com> - - - - - de30a81e by Hassan Al-Awwadi at 2024-09-28T11:49:50+02:00 eleminated outdated comments and use dataConCantHappen - - - - - 21 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/65eb22cb743abeb365893b9f871d12ce60512db7...de30a81e86a11b6a05ec4d2bd237c10730fb904d -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/65eb22cb743abeb365893b9f871d12ce60512db7...de30a81e86a11b6a05ec4d2bd237c10730fb904d You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 28 11:59:12 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Sat, 28 Sep 2024 07:59:12 -0400 Subject: [Git][ghc/ghc][wip/ttg-booleanformula] parameterize booleanformula over the pass and try to get everything to compile. Message-ID: <66f7ef9026fd2_4515125d5b4215b1@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/ttg-booleanformula at Glasgow Haskell Compiler / GHC Commits: 1cb56a31 by Hassan Al-Awwadi at 2024-09-28T13:58:38+02:00 parameterize booleanformula over the pass and try to get everything to compile. Started as BooleanFormula a now BooleanFormula p a next step if possible: BooleanFormula p. blocker to figur eout: BooeleanFormula p IfLclName in GHC/Iface/Type.hs. IfLclName wraps faststring, seemingly, and can't figure out how to convert this. - - - - - 17 changed files: - compiler/GHC/Core/Class.hs - compiler/GHC/Data/BooleanFormula.hs - compiler/GHC/Hs/Binds.hs - compiler/GHC/Hs/Decls.hs - compiler/GHC/Hs/Instances.hs - compiler/GHC/Iface/Decl.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Iface/Syntax.hs - compiler/GHC/Parser.y - compiler/GHC/Rename/Bind.hs - compiler/GHC/Tc/TyCl/Class.hs - compiler/Language/Haskell/Syntax/Binds.hs - + compiler/Language/Haskell/Syntax/BooleanFormula.hs - compiler/ghc.cabal.in - utils/haddock/haddock-api/src/Haddock/Convert.hs - utils/haddock/haddock-api/src/Haddock/Interface/Rename.hs - utils/haddock/haddock-api/src/Haddock/Types.hs Changes: ===================================== compiler/GHC/Core/Class.hs ===================================== @@ -26,6 +26,7 @@ import GHC.Prelude import {-# SOURCE #-} GHC.Core.TyCon ( TyCon ) import {-# SOURCE #-} GHC.Core.TyCo.Rep ( Type, PredType ) import {-# SOURCE #-} GHC.Core.TyCo.Ppr ( pprType ) +import GHC.Hs.Extension (GhcRn) import GHC.Types.Var import GHC.Types.Name import GHC.Types.Basic @@ -35,7 +36,7 @@ import GHC.Utils.Panic import GHC.Types.SrcLoc import GHC.Types.Var.Set import GHC.Utils.Outputable -import GHC.Data.BooleanFormula (BooleanFormula, mkTrue) +import Language.Haskell.Syntax.BooleanFormula ( BooleanFormula, mkTrue ) import qualified Data.Data as Data @@ -135,7 +136,7 @@ data TyFamEqnValidityInfo -- Note [Type-checking default assoc decls] in GHC.Tc.TyCl. } -type ClassMinimalDef = BooleanFormula Name -- Required methods +type ClassMinimalDef = BooleanFormula GhcRn Name -- Required methods data ClassBody = AbstractClass ===================================== compiler/GHC/Data/BooleanFormula.hs ===================================== @@ -1,5 +1,5 @@ -{-# LANGUAGE DeriveDataTypeable #-} -{-# LANGUAGE DeriveTraversable #-} +{-# OPTIONS_GHC -Wno-orphans #-} +{-# LANGUAGE TypeFamilies #-} -------------------------------------------------------------------------------- -- | Boolean formulas without quantifiers and without negation. @@ -8,73 +8,78 @@ -- This module is used to represent minimal complete definitions for classes. -- module GHC.Data.BooleanFormula ( - BooleanFormula(..), LBooleanFormula, - mkFalse, mkTrue, mkAnd, mkOr, mkVar, + module Language.Haskell.Syntax.BooleanFormula, + bfSwitchPass, lbfSwitchPass, isFalse, isTrue, eval, simplify, isUnsatisfied, implies, impliesAtom, pprBooleanFormula, pprBooleanFormulaNice ) where -import GHC.Prelude hiding ( init, last ) - -import Data.List ( nub, intersperse ) +import Data.List ( intersperse ) import Data.List.NonEmpty ( NonEmpty (..), init, last ) -import Data.Data -import GHC.Utils.Monad -import GHC.Utils.Outputable -import GHC.Parser.Annotation ( LocatedL ) -import GHC.Types.SrcLoc +import GHC.Prelude hiding ( init, last ) import GHC.Types.Unique import GHC.Types.Unique.Set +import GHC.Types.SrcLoc (GenLocated(L), unLoc) +import GHC.Utils.Outputable +import GHC.Parser.Annotation ( SrcSpanAnnL ) +import GHC.Hs.Extension (GhcPass) +import Language.Haskell.Syntax.Extension (Anno) +import Language.Haskell.Syntax.BooleanFormula + ---------------------------------------------------------------------- -- Boolean formula type and smart constructors ---------------------------------------------------------------------- -type LBooleanFormula a = LocatedL (BooleanFormula a) - -data BooleanFormula a = Var a | And [LBooleanFormula a] | Or [LBooleanFormula a] - | Parens (LBooleanFormula a) - deriving (Eq, Data, Functor, Foldable, Traversable) - -mkVar :: a -> BooleanFormula a -mkVar = Var - -mkFalse, mkTrue :: BooleanFormula a -mkFalse = Or [] -mkTrue = And [] - --- Convert a Bool to a BooleanFormula -mkBool :: Bool -> BooleanFormula a -mkBool False = mkFalse -mkBool True = mkTrue - --- Make a conjunction, and try to simplify -mkAnd :: Eq a => [LBooleanFormula a] -> BooleanFormula a -mkAnd = maybe mkFalse (mkAnd' . nub) . concatMapM fromAnd - where - -- See Note [Simplification of BooleanFormulas] - fromAnd :: LBooleanFormula a -> Maybe [LBooleanFormula a] - fromAnd (L _ (And xs)) = Just xs - -- assume that xs are already simplified - -- otherwise we would need: fromAnd (And xs) = concat <$> traverse fromAnd xs - fromAnd (L _ (Or [])) = Nothing - -- in case of False we bail out, And [..,mkFalse,..] == mkFalse - fromAnd x = Just [x] - mkAnd' [x] = unLoc x - mkAnd' xs = And xs - -mkOr :: Eq a => [LBooleanFormula a] -> BooleanFormula a -mkOr = maybe mkTrue (mkOr' . nub) . concatMapM fromOr - where - -- See Note [Simplification of BooleanFormulas] - fromOr (L _ (Or xs)) = Just xs - fromOr (L _ (And [])) = Nothing - fromOr x = Just [x] - mkOr' [x] = unLoc x - mkOr' xs = Or xs +-- Actually should this be moved to GHC.Hs.Decls? +-- That's where most of the anno instances are, anyway. +-- But I don't know how strictly that is just an implementation detail +-- I'm allowed to ignore? +type instance Anno (BooleanFormula (GhcPass p) a) = SrcSpanAnnL + +-- In a perfect world I could do something like +-- instance (Functor (LBooleanFormula p) => Functor (BooleanFormula p)). +-- But type synonyms need to be fully applied, and I could not figure out the +-- correct way to hack my way around wrapping and unwrapping and ~ to make it work +-- less cleanly. My last hope for using mapXRec also doe not work, because it has +-- the Anno p a ~ Anno p b constraint which seems impossible to get in the class +-- constraint header. +instance Functor (BooleanFormula (GhcPass p)) where + fmap :: (a -> b) -> BooleanFormula (GhcPass p) a -> BooleanFormula (GhcPass p) b + fmap f (Var a ) = Var (f a) + fmap f (And bfs) = And $ fmap (fmap (fmap f)) bfs + fmap f (Or bfs) = Or $ fmap (fmap (fmap f)) bfs + fmap f (Parens bf ) = Parens $ fmap (fmap f) bf + +-- See comment above Functor instance. +instance Traversable (BooleanFormula (GhcPass p)) where + sequenceA :: Applicative f => BooleanFormula (GhcPass p) (f a) -> f (BooleanFormula (GhcPass p) a) + sequenceA (Var a) = Var <$> a + sequenceA (And bfs) = And <$> traverse (traverse sequenceA) bfs + sequenceA (Or bfs) = Or <$> traverse (traverse sequenceA) bfs + sequenceA (Parens bf ) = Parens <$> traverse sequenceA bf + +-- Just putting this comment here to say I have no idea if this is the right +-- design choice. The alterantive is to just coerce", somehow. +-- When I tried that I got hit with the "Couldn't match type ‘Parsed’ with ‘Renamed’"... +bfSwitchPass :: forall p p' l a + . (LBooleanFormula p a ~ GenLocated l (BooleanFormula p a) + , LBooleanFormula p' a ~ GenLocated l (BooleanFormula p' a)) + => BooleanFormula p a -> BooleanFormula p' a +bfSwitchPass (Var a ) = Var a +bfSwitchPass (And bfs) = And $ fmap lbfSwitchPass bfs +bfSwitchPass (Or bfs) = Or $ fmap lbfSwitchPass bfs +bfSwitchPass (Parens bf ) = Parens $ lbfSwitchPass bf + +lbfSwitchPass :: forall p p' l a + . (LBooleanFormula p a ~ GenLocated l (BooleanFormula p a) + , LBooleanFormula p' a ~ GenLocated l (BooleanFormula p' a)) + => LBooleanFormula p a + -> LBooleanFormula p' a +lbfSwitchPass (L loc bf) = L loc (bfSwitchPass bf) {- @@ -115,15 +120,15 @@ We don't show a ridiculous error message like -- Evaluation and simplification ---------------------------------------------------------------------- -isFalse :: BooleanFormula a -> Bool +isFalse :: BooleanFormula (GhcPass p) a -> Bool isFalse (Or []) = True isFalse _ = False -isTrue :: BooleanFormula a -> Bool +isTrue :: BooleanFormula (GhcPass p) a -> Bool isTrue (And []) = True isTrue _ = False -eval :: (a -> Bool) -> BooleanFormula a -> Bool +eval :: (a -> Bool) -> BooleanFormula (GhcPass p) a -> Bool eval f (Var x) = f x eval f (And xs) = all (eval f . unLoc) xs eval f (Or xs) = any (eval f . unLoc) xs @@ -131,18 +136,18 @@ eval f (Parens x) = eval f (unLoc x) -- Simplify a boolean formula. -- The argument function should give the truth of the atoms, or Nothing if undecided. -simplify :: Eq a => (a -> Maybe Bool) -> BooleanFormula a -> BooleanFormula a +simplify :: Eq a => (a -> Maybe Bool) -> BooleanFormula (GhcPass p) a -> BooleanFormula (GhcPass p) a simplify f (Var a) = case f a of Nothing -> Var a Just b -> mkBool b -simplify f (And xs) = mkAnd (map (\(L l x) -> L l (simplify f x)) xs) -simplify f (Or xs) = mkOr (map (\(L l x) -> L l (simplify f x)) xs) +simplify f (And xs) = mkAnd (map (fmap (simplify f)) xs) +simplify f (Or xs) = mkOr (map (fmap (simplify f)) xs) simplify f (Parens x) = simplify f (unLoc x) -- Test if a boolean formula is satisfied when the given values are assigned to the atoms -- if it is, returns Nothing -- if it is not, return (Just remainder) -isUnsatisfied :: Eq a => (a -> Bool) -> BooleanFormula a -> Maybe (BooleanFormula a) +isUnsatisfied :: Eq a => (a -> Bool) -> BooleanFormula (GhcPass p) a -> Maybe (BooleanFormula (GhcPass p) a) isUnsatisfied f bf | isTrue bf' = Nothing | otherwise = Just bf' @@ -155,17 +160,17 @@ isUnsatisfied f bf -- eval f x == False <==> isFalse (simplify (Just . f) x) -- If the boolean formula holds, does that mean that the given atom is always true? -impliesAtom :: Eq a => BooleanFormula a -> a -> Bool +impliesAtom :: Eq a => BooleanFormula (GhcPass p) a -> a -> Bool Var x `impliesAtom` y = x == y -And xs `impliesAtom` y = any (\x -> (unLoc x) `impliesAtom` y) xs +And xs `impliesAtom` y = any (\x -> unLoc x `impliesAtom` y) xs -- we have all of xs, so one of them implying y is enough -Or xs `impliesAtom` y = all (\x -> (unLoc x) `impliesAtom` y) xs -Parens x `impliesAtom` y = (unLoc x) `impliesAtom` y +Or xs `impliesAtom` y = all (\x -> unLoc x `impliesAtom` y) xs +Parens x `impliesAtom` y = unLoc x `impliesAtom` y -implies :: Uniquable a => BooleanFormula a -> BooleanFormula a -> Bool +implies :: Uniquable a => BooleanFormula (GhcPass p) a -> BooleanFormula (GhcPass p) a -> Bool implies e1 e2 = go (Clause emptyUniqSet [e1]) (Clause emptyUniqSet [e2]) where - go :: Uniquable a => Clause a -> Clause a -> Bool + go :: Uniquable a => Clause (GhcPass p) a -> Clause (GhcPass p) a -> Bool go l at Clause{ clauseExprs = hyp:hyps } r = case hyp of Var x | memberClauseAtoms x r -> True @@ -183,14 +188,14 @@ implies e1 e2 = go (Clause emptyUniqSet [e1]) (Clause emptyUniqSet [e2]) go _ _ = False -- A small sequent calculus proof engine. -data Clause a = Clause { +data Clause p a = Clause { clauseAtoms :: UniqSet a, - clauseExprs :: [BooleanFormula a] + clauseExprs :: [BooleanFormula p a] } -extendClauseAtoms :: Uniquable a => Clause a -> a -> Clause a +extendClauseAtoms :: Uniquable a => Clause p a -> a -> Clause p a extendClauseAtoms c x = c { clauseAtoms = addOneToUniqSet (clauseAtoms c) x } -memberClauseAtoms :: Uniquable a => a -> Clause a -> Bool +memberClauseAtoms :: Uniquable a => a -> Clause p a -> Bool memberClauseAtoms x c = x `elementOfUniqSet` clauseAtoms c ---------------------------------------------------------------------- @@ -199,28 +204,28 @@ memberClauseAtoms x c = x `elementOfUniqSet` clauseAtoms c -- Pretty print a BooleanFormula, -- using the arguments as pretty printers for Var, And and Or respectively -pprBooleanFormula' :: (Rational -> a -> SDoc) - -> (Rational -> [SDoc] -> SDoc) - -> (Rational -> [SDoc] -> SDoc) - -> Rational -> BooleanFormula a -> SDoc +pprBooleanFormula' :: (Rational -> a -> SDoc) + -> (Rational -> [SDoc] -> SDoc) + -> (Rational -> [SDoc] -> SDoc) + -> Rational -> BooleanFormula (GhcPass p) a -> SDoc pprBooleanFormula' pprVar pprAnd pprOr = go where go p (Var x) = pprVar p x - go p (And []) = cparen (p > 0) $ empty + go p (And []) = cparen (p > 0) empty go p (And xs) = pprAnd p (map (go 3 . unLoc) xs) go _ (Or []) = keyword $ text "FALSE" go p (Or xs) = pprOr p (map (go 2 . unLoc) xs) go p (Parens x) = go p (unLoc x) -- Pretty print in source syntax, "a | b | c,d,e" -pprBooleanFormula :: (Rational -> a -> SDoc) -> Rational -> BooleanFormula a -> SDoc +pprBooleanFormula :: (Rational -> a -> SDoc) -> Rational -> BooleanFormula (GhcPass p) a -> SDoc pprBooleanFormula pprVar = pprBooleanFormula' pprVar pprAnd pprOr where pprAnd p = cparen (p > 3) . fsep . punctuate comma pprOr p = cparen (p > 2) . fsep . intersperse vbar -- Pretty print human in readable format, "either `a' or `b' or (`c', `d' and `e')"? -pprBooleanFormulaNice :: Outputable a => BooleanFormula a -> SDoc +pprBooleanFormulaNice :: Outputable a => BooleanFormula (GhcPass p) a -> SDoc pprBooleanFormulaNice = pprBooleanFormula' pprVar pprAnd pprOr 0 where pprVar _ = quotes . ppr @@ -230,11 +235,10 @@ pprBooleanFormulaNice = pprBooleanFormula' pprVar pprAnd pprOr 0 pprAnd' (x:xs) = fsep (punctuate comma (init (x:|xs))) <> text ", and" <+> last (x:|xs) pprOr p xs = cparen (p > 1) $ text "either" <+> sep (intersperse (text "or") xs) -instance (OutputableBndr a) => Outputable (BooleanFormula a) where +instance (OutputableBndr a) => Outputable (BooleanFormula (GhcPass p) a) where ppr = pprBooleanFormulaNormal -pprBooleanFormulaNormal :: (OutputableBndr a) - => BooleanFormula a -> SDoc +pprBooleanFormulaNormal :: (OutputableBndr a) => BooleanFormula (GhcPass p) a -> SDoc pprBooleanFormulaNormal = go where go (Var x) = pprPrefixOcc x ===================================== compiler/GHC/Hs/Binds.hs ===================================== @@ -935,7 +935,7 @@ instance Outputable TcSpecPrag where = text (extractSpecPragName $ inl_src inl) <+> pprSpec var (text "") inl pprMinimalSig :: (OutputableBndr name) - => LBooleanFormula (GenLocated l name) -> SDoc + => LBooleanFormula (GhcPass p) (GenLocated l name) -> SDoc pprMinimalSig (L _ bf) = ppr (fmap unLoc bf) {- ===================================== compiler/GHC/Hs/Decls.hs ===================================== @@ -103,6 +103,8 @@ module GHC.Hs.Decls ( import GHC.Prelude import Language.Haskell.Syntax.Decls +import Language.Haskell.Syntax.Extension +import Language.Haskell.Syntax.BooleanFormula (BooleanFormula) import {-# SOURCE #-} GHC.Hs.Expr ( pprExpr, pprUntypedSplice ) -- Because Expr imports Decls via HsBracket @@ -112,7 +114,7 @@ import GHC.Hs.Type import GHC.Hs.Doc import GHC.Types.Basic import GHC.Core.Coercion -import Language.Haskell.Syntax.Extension + import GHC.Hs.Extension import GHC.Parser.Annotation import GHC.Types.Name ===================================== compiler/GHC/Hs/Instances.hs ===================================== @@ -33,6 +33,7 @@ import GHC.Hs.Type import GHC.Hs.Pat import GHC.Hs.ImpExp import GHC.Parser.Annotation +import GHC.Data.BooleanFormula (BooleanFormula(..)) -- --------------------------------------------------------------------- -- Data derivations from GHC.Hs----------------------------------------- @@ -584,3 +585,9 @@ deriving instance Data XXPatGhcTc deriving instance Data XViaStrategyPs -- --------------------------------------------------------------------- + +deriving instance Data a => Data (BooleanFormula GhcPs a) +deriving instance Data a => Data (BooleanFormula GhcRn a) +deriving instance Data a => Data (BooleanFormula GhcTc a) + +--------------------------------------------------------------------- \ No newline at end of file ===================================== compiler/GHC/Iface/Decl.hs ===================================== @@ -33,7 +33,7 @@ import GHC.Core.DataCon import GHC.Core.Type import GHC.Core.Multiplicity - +import GHC.Hs.Extension ( GhcPass ) import GHC.Types.Id import GHC.Types.Var.Env import GHC.Types.Var @@ -336,7 +336,7 @@ tidyTyConBinders = mapAccumL tidyTyConBinder tidyTyVar :: TidyEnv -> TyVar -> IfLclName tidyTyVar (_, subst) tv = toIfaceTyVar (lookupVarEnv subst tv `orElse` tv) -toIfaceBooleanFormula :: BooleanFormula IfLclName -> IfaceBooleanFormula +toIfaceBooleanFormula :: BooleanFormula (GhcPass p) IfLclName -> IfaceBooleanFormula toIfaceBooleanFormula = \case Var nm -> IfVar nm And bfs -> IfAnd (map (toIfaceBooleanFormula . unLoc) bfs) ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -2014,7 +2014,7 @@ instance ToHie PendingRnSplice where instance ToHie PendingTcSplice where toHie (PendingTcSplice _ e) = toHie e -instance ToHie (LBooleanFormula (LocatedN Name)) where +instance Typeable p => ToHie (GenLocated SrcSpanAnnL (BooleanFormula (GhcPass p) (LocatedN Name))) where toHie (L span form) = concatM $ makeNode form (locA span) : case form of Var a -> [ toHie $ C Use a ===================================== compiler/GHC/Iface/Syntax.hs ===================================== @@ -82,7 +82,7 @@ import GHC.Core.DataCon (SrcStrictness(..), SrcUnpackedness(..)) import GHC.Builtin.Types ( constraintKindTyConName ) import GHC.Stg.InferTags.TagSig import GHC.Parser.Annotation (noLocA) -import GHC.Hs.Extension ( GhcRn ) +import GHC.Hs.Extension ( GhcRn, GhcPass ) import GHC.Hs.Doc ( WithHsDocIdentifiers(..) ) import GHC.Utils.Lexeme (isLexSym) @@ -219,7 +219,7 @@ data IfaceBooleanFormula | IfOr [IfaceBooleanFormula] | IfParens IfaceBooleanFormula -fromIfaceBooleanFormula :: IfaceBooleanFormula -> BooleanFormula IfLclName +fromIfaceBooleanFormula :: IfaceBooleanFormula -> BooleanFormula (GhcPass p) IfLclName fromIfaceBooleanFormula = \case IfVar nm -> Var nm IfAnd ibfs -> And (map (noLocA . fromIfaceBooleanFormula) ibfs) @@ -1039,7 +1039,7 @@ pprIfaceDecl ss (IfaceClass { ifName = clas | showSub ss sg = Just $ pprIfaceClassOp ss sg | otherwise = Nothing - pprMinDef :: BooleanFormula IfLclName -> SDoc + pprMinDef :: BooleanFormula (GhcPass p) IfLclName -> SDoc pprMinDef minDef = ppUnless (isTrue minDef) $ -- hide empty definitions text "{-# MINIMAL" <+> pprBooleanFormula ===================================== compiler/GHC/Parser.y ===================================== @@ -39,14 +39,15 @@ module GHC.Parser where -- base -import Control.Monad ( unless, liftM, when, (<=<) ) +import Control.Monad ( unless, liftM, when, (<=<) ) import GHC.Exts -import Data.Maybe ( maybeToList ) +import Data.Maybe ( maybeToList ) import Data.List.NonEmpty ( NonEmpty(..) ) import qualified Data.List.NonEmpty as NE import qualified Prelude -- for happy-generated code import GHC.Hs +import GHC.Hs.Extension (GhcPass, Pass(..)) import GHC.Driver.Backpack.Syntax @@ -3680,27 +3681,27 @@ overloaded_label :: { Located (SourceText, FastString) } ----------------------------------------------------------------------------- -- Warnings and deprecations -name_boolformula_opt :: { LBooleanFormula (LocatedN RdrName) } +name_boolformula_opt :: { LBooleanFormula GhcPs (LocatedN RdrName) } : name_boolformula { $1 } | {- empty -} { noLocA mkTrue } -name_boolformula :: { LBooleanFormula (LocatedN RdrName) } +name_boolformula :: { LBooleanFormula GhcPs (LocatedN RdrName) } : name_boolformula_and { $1 } | name_boolformula_and '|' name_boolformula {% do { h <- addTrailingVbarL $1 (gl $2) ; return (sLLa $1 $> (Or [h,$3])) } } -name_boolformula_and :: { LBooleanFormula (LocatedN RdrName) } +name_boolformula_and :: { LBooleanFormula GhcPs (LocatedN RdrName) } : name_boolformula_and_list { sLLa (head $1) (last $1) (And ($1)) } -name_boolformula_and_list :: { [LBooleanFormula (LocatedN RdrName)] } +name_boolformula_and_list :: { [LBooleanFormula GhcPs (LocatedN RdrName)] } : name_boolformula_atom { [$1] } | name_boolformula_atom ',' name_boolformula_and_list {% do { h <- addTrailingCommaL $1 (gl $2) ; return (h : $3) } } -name_boolformula_atom :: { LBooleanFormula (LocatedN RdrName) } +name_boolformula_atom :: { LBooleanFormula GhcPs (LocatedN RdrName) } : '(' name_boolformula ')' {% amsr (sLL $1 $> (Parens $2)) (AnnList Nothing (Just (mop $1)) (Just (mcp $3)) [] []) } | name_var { sL1a $1 (Var $1) } ===================================== compiler/GHC/Rename/Bind.hs ===================================== @@ -80,6 +80,7 @@ import Control.Monad import Data.List ( partition ) import Data.List.NonEmpty ( NonEmpty(..) ) import GHC.Types.Unique.DSet (mkUniqDSet) +import GHC.Data.BooleanFormula (bfSwitchPass) {- -- ToDo: Put the annotations into the monad, so that they arrive in the proper @@ -1138,7 +1139,7 @@ renameSig ctxt (FixSig _ fsig) renameSig ctxt sig@(MinimalSig (_, s) (L l bf)) = do new_bf <- traverse (lookupSigOccRnN ctxt sig) bf - return (MinimalSig (noAnn, s) (L l new_bf), emptyFVs) + return (MinimalSig (noAnn, s) (L l (bfSwitchPass new_bf)), emptyFVs) renameSig ctxt sig@(PatSynSig _ vs ty) = do { new_vs <- mapM (lookupSigOccRnN ctxt sig) vs ===================================== compiler/GHC/Tc/TyCl/Class.hs ===================================== @@ -402,7 +402,7 @@ findMinimalDef :: [LSig GhcRn] -> Maybe ClassMinimalDef findMinimalDef = firstJusts . map toMinimalDef where toMinimalDef :: LSig GhcRn -> Maybe ClassMinimalDef - toMinimalDef (L _ (MinimalSig _ (L _ bf))) = Just (fmap unLoc bf) + toMinimalDef (L _ (MinimalSig _ (L _ bf))) = Just (fmap unLoc (bfSwitchPass bf)) toMinimalDef _ = Nothing {- ===================================== compiler/Language/Haskell/Syntax/Binds.hs ===================================== @@ -28,6 +28,7 @@ import {-# SOURCE #-} Language.Haskell.Syntax.Expr , GRHSs ) import {-# SOURCE #-} Language.Haskell.Syntax.Pat ( LPat ) +import Language.Haskell.Syntax.BooleanFormula (LBooleanFormula) import Language.Haskell.Syntax.Extension import Language.Haskell.Syntax.Type @@ -35,7 +36,6 @@ import Language.Haskell.Syntax.Type import GHC.Types.Fixity (Fixity) import GHC.Types.Basic (InlinePragma) -import GHC.Data.BooleanFormula (LBooleanFormula) import GHC.Types.SourceText (StringLiteral) import Data.Void @@ -465,7 +465,7 @@ data Sig pass -- 'GHC.Parser.Annotation.AnnClose' -- For details on above see Note [exact print annotations] in GHC.Parser.Annotation - | MinimalSig (XMinimalSig pass) (LBooleanFormula (LIdP pass)) + | MinimalSig (XMinimalSig pass) (LBooleanFormula pass (LIdP pass)) -- | A "set cost centre" pragma for declarations -- ===================================== compiler/Language/Haskell/Syntax/BooleanFormula.hs ===================================== @@ -0,0 +1,68 @@ +{-# LANGUAGE AllowAmbiguousTypes #-} +{-# LANGUAGE UndecidableInstances #-} +{-# LANGUAGE QuantifiedConstraints #-} + +module Language.Haskell.Syntax.BooleanFormula( + BooleanFormula(..), LBooleanFormula, + mkVar, mkFalse, mkTrue, mkBool, mkAnd, mkOr) where + +import Prelude hiding ( init, last ) +import Data.List ( nub ) +import Language.Haskell.Syntax.Extension (XRec, UnXRec (..)) + + + +type LBooleanFormula p a = XRec p (BooleanFormula p a) + +--type role BooleanFormula phantom nominal +data BooleanFormula p a = Var a | And [LBooleanFormula p a] | Or [LBooleanFormula p a] + | Parens (LBooleanFormula p a) + +-- instances +deriving instance (Eq a, Eq (LBooleanFormula p a)) => Eq (BooleanFormula p a) +instance UnXRec p => Foldable (BooleanFormula p) where + foldMap :: Monoid m => (a -> m) -> BooleanFormula p a -> m + foldMap f (Var a ) = f a + foldMap f (And bfs) = foldMap (foldMap f . unXRec @p) bfs + foldMap f (Or bfs) = foldMap (foldMap f . unXRec @p) bfs + foldMap f (Parens bf ) = foldMap f $ unXRec @p bf + + +mkVar :: a -> BooleanFormula p a +mkVar = Var + +mkFalse, mkTrue :: BooleanFormula p a +mkFalse = Or [] +mkTrue = And [] + +-- Convert a Bool to a BooleanFormula +mkBool :: Bool -> BooleanFormula p a +mkBool False = mkFalse +mkBool True = mkTrue + +-- Make a conjunction, and try to simplify +mkAnd :: forall p a. (UnXRec p, Eq a, Eq (LBooleanFormula p a)) => [LBooleanFormula p a] -> BooleanFormula p a +mkAnd = maybe mkFalse (mkAnd' . nub . concat) . mapM fromAnd + where + -- See Note [Simplification of BooleanFormulas] + fromAnd :: LBooleanFormula p a -> Maybe [LBooleanFormula p a] + fromAnd bf = case unXRec @p bf of + (And xs) -> Just xs + -- assume that xs are already simplified + -- otherwise we would need: fromAnd (And xs) = concat <$> traverse fromAnd xs + (Or []) -> Nothing + -- in case of False we bail out, And [..,mkFalse,..] == mkFalse + _ -> Just [bf] + mkAnd' [x] = unXRec @p x + mkAnd' xs = And xs + +mkOr :: forall p a. (UnXRec p, Eq a, Eq (LBooleanFormula p a)) => [LBooleanFormula p a] -> BooleanFormula p a +mkOr = maybe mkTrue (mkOr' . nub . concat) . mapM fromOr + where + -- See Note [Simplification of BooleanFormulas] + fromOr bf = case unXRec @p bf of + (Or xs) -> Just xs + (And []) -> Nothing + _ -> Just [bf] + mkOr' [x] = unXRec @p x + mkOr' xs = Or xs ===================================== compiler/ghc.cabal.in ===================================== @@ -984,6 +984,7 @@ Library Language.Haskell.Syntax Language.Haskell.Syntax.Basic Language.Haskell.Syntax.Binds + Language.Haskell.Syntax.BooleanFormula Language.Haskell.Syntax.Decls Language.Haskell.Syntax.Expr Language.Haskell.Syntax.Extension ===================================== utils/haddock/haddock-api/src/Haddock/Convert.hs ===================================== @@ -45,6 +45,8 @@ import GHC.Builtin.Types , promotedNilDataCon , unitTy ) + +import GHC.Data.BooleanFormula(bfSwitchPass) import GHC.Builtin.Types.Prim (alphaTyVars) import GHC.Core.Class import GHC.Core.Coercion.Axiom @@ -174,7 +176,7 @@ tyThingToLHsDecl prr t = case t of $ snd $ classTvsFds cl , tcdSigs = - noLocA (MinimalSig (noAnn, NoSourceText) . noLocA . fmap noLocA $ classMinimalDef cl) + noLocA (MinimalSig (noAnn, NoSourceText) . noLocA . bfSwitchPass . fmap noLocA $ classMinimalDef cl) : [ noLocA tcdSig | clsOp <- classOpItems cl , tcdSig <- synifyTcIdSig vs clsOp ===================================== utils/haddock/haddock-api/src/Haddock/Interface/Rename.hs ===================================== @@ -30,6 +30,7 @@ import qualified Data.Set as Set import Data.Traversable (mapM) import GHC hiding (NoLink) import GHC.Builtin.Types (eqTyCon_RDR, tupleDataConName, tupleTyConName) +import GHC.Data.BooleanFormula (bfSwitchPass) import GHC.Types.Basic (Boxity (..), TopLevelFlag (..), TupleSort (..)) import GHC.Types.Name import GHC.Types.Name.Reader (RdrName (Exact)) @@ -768,7 +769,7 @@ renameSig sig = case sig of return $ FixSig noExtField (FixitySig noExtField lnames' fixity) MinimalSig _ (L l s) -> do s' <- traverse (traverse lookupRn) s - return $ MinimalSig noExtField (L l s') + return $ MinimalSig noExtField (L l (bfSwitchPass s')) -- we have filtered out all other kinds of signatures in Interface.Create _ -> error "expected TypeSig" ===================================== utils/haddock/haddock-api/src/Haddock/Types.hs ===================================== @@ -52,6 +52,7 @@ import qualified Data.Map as Map import qualified Data.Set as Set import GHC import qualified GHC.Data.Strict as Strict +import GHC.Data.BooleanFormula (BooleanFormula) import GHC.Driver.Session (Language) import qualified GHC.LanguageExtensions as LangExt import GHC.Core.InstEnv (is_dfun_name) @@ -818,6 +819,7 @@ type instance Anno (HsDecl DocNameI) = SrcSpanAnnA type instance Anno (FamilyResultSig DocNameI) = EpAnn NoEpAnns type instance Anno (HsOuterTyVarBndrs Specificity DocNameI) = SrcSpanAnnA type instance Anno (HsSigType DocNameI) = SrcSpanAnnA +type instance Anno (BooleanFormula DocNameI a) = SrcSpanAnnL type XRecCond a = ( XParTy a ~ AnnParen View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1cb56a311d9d1bfb6b53780e11745911ebf635e4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1cb56a311d9d1bfb6b53780e11745911ebf635e4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sat Sep 28 12:21:45 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Sat, 28 Sep 2024 08:21:45 -0400 Subject: [Git][ghc/ghc][wip/ttg-booleanformula] 76 commits: Linker: add support for extra built-in symbols (#25155) Message-ID: <66f7f4d9c5aff_451513dec94260f5@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/ttg-booleanformula at Glasgow Haskell Compiler / GHC Commits: 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - 20a8610b by Hassan Al-Awwadi at 2024-09-28T14:21:07+02:00 Parameterize booleanformula over the pass and try to get everything to compile. Started as BooleanFormula a now BooleanFormula p a next step if possible: BooleanFormula p. blocker to figur eout: BooeleanFormula p IfLclName in GHC/Iface/Type.hs. IfLclName wraps faststring, seemingly, and can't figure out how to convert this. - - - - - 22 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitmodules - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1cb56a311d9d1bfb6b53780e11745911ebf635e4...20a8610b821a16a9796c252bbee82a843ad6f7e8 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1cb56a311d9d1bfb6b53780e11745911ebf635e4...20a8610b821a16a9796c252bbee82a843ad6f7e8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 16:05:22 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Sun, 29 Sep 2024 12:05:22 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8-2] 35 commits: Bump stm submodule to v2.5.3.1 Message-ID: <66f97ac283acf_28e2bb18719455521@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8-2 at Glasgow Haskell Compiler / GHC Commits: 5aafe259 by Ben Gamari at 2024-09-29T12:04:51-04:00 Bump stm submodule to v2.5.3.1 - - - - - fce51010 by Andrew Lelechenko at 2024-09-29T12:04:51-04:00 Bump submodule deepseq to 1.5.1.0 (cherry picked from commit 8e462f4d4bdf2a6c34c249e7be8084565600d300) - - - - - d01f6cb6 by Cheng Shao at 2024-09-29T12:04:51-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - 8965b192 by Cheng Shao at 2024-09-29T12:04:51-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 11e0c2ad by Cheng Shao at 2024-09-29T12:04:51-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - 98582e30 by Cheng Shao at 2024-09-29T12:04:51-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - b1fcd271 by Andrew Lelechenko at 2024-09-29T12:04:51-04:00 Bump submodule array to 0.5.8.0 (cherry picked from commit 80769bc9f56541601796366485283a697c52a18b) - - - - - 4aa51df6 by Matthew Pickering at 2024-09-29T12:04:51-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 (cherry picked from commit 7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6) - - - - - 486b2702 by Sylvain Henry at 2024-09-29T12:04:51-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 (cherry picked from commit e7a26d7a6faf1ea534e036c5085a0a027dbb6f5f) - - - - - 54fd5ea5 by Sylvain Henry at 2024-09-29T12:04:51-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 (cherry picked from commit 5092dbff750ee5b6fd082b7eed8574922a2b0bf4) - - - - - 766dca00 by Sylvain Henry at 2024-09-29T12:04:51-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. (cherry picked from commit c749bdfd3e21d712dc2b966482eb010165bdeebe) - - - - - e0a3038e by Sylvain Henry at 2024-09-29T12:04:52-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> (cherry picked from commit 7446a09a2d5b04b95cd43c03659b5647853124ce) - - - - - 57fde2ad by sheaf at 2024-09-29T12:04:52-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 (cherry picked from commit c29b2b5a77611b2bd6c3089765079bc43aec3e22) - - - - - f3b2ce65 by Sylvain Henry at 2024-09-29T12:04:52-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 (cherry picked from commit f954f42823f6ca3588425a0d543d93ace86d89e4) - - - - - 0a119703 by Simon Peyton Jones at 2024-09-29T12:04:52-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify (cherry picked from commit de5d9852dbdd367611bf9e45e69c723d26351992) - - - - - 0b6d13fc by doyougnu at 2024-09-29T12:04:52-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches (cherry picked from commit bfe4b3d3bbb98b39169fad063c6c32f06d167756) - - - - - d77c6ffc by Ben Gamari at 2024-09-29T12:04:52-04:00 docs: Add release notes for 9.8.3 - - - - - df964112 by Andreas Klebinger at 2024-09-29T12:04:52-04:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. (cherry picked from commit 3f89ab92da74c4ed45da68fe92ff81e7b9caa53d) - - - - - 0b3ebb79 by Peter Trommler at 2024-09-29T12:04:52-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 (cherry picked from commit 7fe85b1354a13749f14d588e3cc742b8ae2d8da9) - - - - - 139c6568 by Ben Gamari at 2024-09-29T12:04:52-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 (cherry picked from commit a82121b3b6fdc2ac47211f71871b3ab21e5f6276) - - - - - b4a161af by Sylvain Henry at 2024-09-29T12:04:52-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). (cherry picked from commit 5104ee615503617a1c124fe1d92f6aa2d263b7d0) - - - - - b58b689a by Sylvain Henry at 2024-09-29T12:04:52-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC (cherry picked from commit 52d6698479f951e07def237b0474ee22d27e621a) - - - - - 854508d5 by Sylvain Henry at 2024-09-29T12:04:52-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). (cherry picked from commit c34fef56367142fa55e9861092f64cc7b9946fa1) - - - - - ba347fd6 by Cheng Shao at 2024-09-29T12:04:52-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. (cherry picked from commit 0cff083abb24701530974872b21cf897c9955a9a) - - - - - e4b99959 by Andreas Klebinger at 2024-09-29T12:04:52-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. (cherry picked from commit 1bfa91115b8320ed99a5e946147528e21ca4f3e1) - - - - - 19019bdc by crumbtoo at 2024-09-29T12:04:52-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 (cherry picked from commit c5e00c35927d574f71bf77449817b131d1749750) - - - - - e273ef6b by Ryan Scott at 2024-09-29T12:04:52-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. (cherry picked from commit a3cd3a1d0d186f2aa4d0273c6b3e74a442de2ef0) - - - - - 96a3e7e6 by Sylvain Henry at 2024-09-29T12:04:52-04:00 Reverse arguments to stgCallocBytes (fix #24828) (cherry picked from commit 6838a7c32ca29b5d44adc9d6280d3a960f31be7c) - - - - - cb8e49a6 by Teo Camarasu at 2024-09-29T12:04:52-04:00 doc: Fix type error in hs_try_putmvar example (cherry picked from commit 06f7db4001e4eee0f3076d949876f8f4af0eb6fb) - - - - - 601468a3 by Matthew Craven at 2024-09-29T12:04:52-04:00 Add test cases for #24664 ...since none are present in the original MR !12463 fixing this issue. (cherry picked from commit a19201d42cfd3aa54faeb1b5a95b715b9a67a01a) - - - - - 2d5c6f80 by Zubin Duggal at 2024-09-29T12:04:52-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 (cherry picked from commit dddc9dff0547733a10e7f505612ab9df3a7c21b6) - - - - - 44173c39 by Sylvain Henry at 2024-09-29T12:04:52-04:00 JS: correctly handle RUBBISH literals (#24664) (cherry picked from commit daeda83478d5b800d29661408dd67cc4b23df374) - - - - - cbf434d3 by Matthew Pickering at 2024-09-29T12:04:52-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. (cherry picked from commit 3fff09779d5830549ae455a15907b7bb9fe7859a) - - - - - 3196bfb3 by Teo Camarasu at 2024-09-29T12:04:52-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 (cherry picked from commit f30e4984fb048818051465698ef8e4e20dacb577) - - - - - aa93076c by Ben Gamari at 2024-09-29T12:04:52-04:00 Accept performance shifts Metric Decrease: MultiLayerModulesTH_Make - - - - - 30 changed files: - .gitlab-ci.yml - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/Reg/Graph/TrivColorable.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Driver/Config/Core/Opt/Simplify.hs - compiler/GHC/Iface/Recomp/Flags.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Literal.hs - compiler/GHC/SysTools/Cpp.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Id.hs - compiler/GHC/Unit/Types.hs - docs/index.html.in - + docs/users_guide/9.8.3-notes.rst - docs/users_guide/exts/ffi.rst - docs/users_guide/exts/multiway_if.rst - docs/users_guide/release-notes.rst - hadrian/src/Hadrian/Haskell/Cabal/Parse.hs - hadrian/src/Rules/Generate.hs - libraries/array - libraries/base/GHC/Event/KQueue.hsc - libraries/deepseq The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/31012d411b9f307debd3f0f19daa1fcd07326ee8...aa93076c5157a813699649230272749fb6511ce9 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/31012d411b9f307debd3f0f19daa1fcd07326ee8...aa93076c5157a813699649230272749fb6511ce9 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 16:05:36 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Sun, 29 Sep 2024 12:05:36 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8] 17 commits: Bump stm submodule to v2.5.3.1 Message-ID: <66f97ad06d903_28e2bb2a938855773@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8 at Glasgow Haskell Compiler / GHC Commits: 83126193 by Ben Gamari at 2024-09-29T12:05:28-04:00 Bump stm submodule to v2.5.3.1 - - - - - c2c85f36 by Andrew Lelechenko at 2024-09-29T12:05:28-04:00 Bump submodule deepseq to 1.5.1.0 (cherry picked from commit 8e462f4d4bdf2a6c34c249e7be8084565600d300) - - - - - af73ed1e by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - 5b2b8321 by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 83341bbc by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - dea8a144 by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - 96a15414 by Andrew Lelechenko at 2024-09-29T12:05:28-04:00 Bump submodule array to 0.5.8.0 (cherry picked from commit 80769bc9f56541601796366485283a697c52a18b) - - - - - d5e89136 by Matthew Pickering at 2024-09-29T12:05:28-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 (cherry picked from commit 7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6) - - - - - bcfa1a51 by Sylvain Henry at 2024-09-29T12:05:28-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 (cherry picked from commit e7a26d7a6faf1ea534e036c5085a0a027dbb6f5f) - - - - - e34c4192 by Sylvain Henry at 2024-09-29T12:05:28-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 (cherry picked from commit 5092dbff750ee5b6fd082b7eed8574922a2b0bf4) - - - - - 1d3bd636 by Sylvain Henry at 2024-09-29T12:05:28-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. (cherry picked from commit c749bdfd3e21d712dc2b966482eb010165bdeebe) - - - - - b060a7a7 by Sylvain Henry at 2024-09-29T12:05:28-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> (cherry picked from commit 7446a09a2d5b04b95cd43c03659b5647853124ce) - - - - - 532c5abf by sheaf at 2024-09-29T12:05:28-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 (cherry picked from commit c29b2b5a77611b2bd6c3089765079bc43aec3e22) - - - - - eb0e402c by Sylvain Henry at 2024-09-29T12:05:28-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 (cherry picked from commit f954f42823f6ca3588425a0d543d93ace86d89e4) - - - - - 98c7e1bb by Simon Peyton Jones at 2024-09-29T12:05:28-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify (cherry picked from commit de5d9852dbdd367611bf9e45e69c723d26351992) - - - - - 9524086e by doyougnu at 2024-09-29T12:05:28-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches (cherry picked from commit bfe4b3d3bbb98b39169fad063c6c32f06d167756) - - - - - a7b9f6c8 by Ben Gamari at 2024-09-29T12:05:28-04:00 Accept performance shifts Metric Decrease: MultiLayerModulesTH_Make - - - - - 30 changed files: - .gitlab-ci.yml - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Driver/Config/Core/Opt/Simplify.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/StgToJS/Literal.hs - compiler/GHC/SysTools/Cpp.hs - compiler/GHC/Types/Id.hs - compiler/GHC/Unit/Types.hs - libraries/array - libraries/base/GHC/Event/KQueue.hsc - libraries/deepseq - libraries/stm - rts/include/Stg.h - rts/linker/PEi386.c - rts/linker/elf_reloc_aarch64.c - rts/sm/GC.c - rts/sm/NonMovingMark.c - rts/sm/Sanity.c - + testsuite/tests/codeGen/should_compile/T25177.hs - + testsuite/tests/codeGen/should_compile/T25177.stderr - testsuite/tests/codeGen/should_compile/all.T - + testsuite/tests/codeGen/should_run/T24700.hs - + testsuite/tests/codeGen/should_run/T24700.stdin - + testsuite/tests/codeGen/should_run/T24700.stdout - + testsuite/tests/codeGen/should_run/T24893.hs - + testsuite/tests/codeGen/should_run/T24893.stdout - testsuite/tests/codeGen/should_run/all.T - testsuite/tests/driver/T20604/T20604.stdout The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b2b8708398948276d96b1bbd07589431669aa51c...a7b9f6c8b6b1629f6f9b9d47ffcf68d9e27aef8b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b2b8708398948276d96b1bbd07589431669aa51c...a7b9f6c8b6b1629f6f9b9d47ffcf68d9e27aef8b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 16:05:59 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Sun, 29 Sep 2024 12:05:59 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8-2] 35 commits: Bump stm submodule to v2.5.3.1 Message-ID: <66f97ae785884_28e2bb42e5dc56184@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8-2 at Glasgow Haskell Compiler / GHC Commits: 83126193 by Ben Gamari at 2024-09-29T12:05:28-04:00 Bump stm submodule to v2.5.3.1 - - - - - c2c85f36 by Andrew Lelechenko at 2024-09-29T12:05:28-04:00 Bump submodule deepseq to 1.5.1.0 (cherry picked from commit 8e462f4d4bdf2a6c34c249e7be8084565600d300) - - - - - af73ed1e by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - 5b2b8321 by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 83341bbc by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - dea8a144 by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - 96a15414 by Andrew Lelechenko at 2024-09-29T12:05:28-04:00 Bump submodule array to 0.5.8.0 (cherry picked from commit 80769bc9f56541601796366485283a697c52a18b) - - - - - d5e89136 by Matthew Pickering at 2024-09-29T12:05:28-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 (cherry picked from commit 7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6) - - - - - bcfa1a51 by Sylvain Henry at 2024-09-29T12:05:28-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 (cherry picked from commit e7a26d7a6faf1ea534e036c5085a0a027dbb6f5f) - - - - - e34c4192 by Sylvain Henry at 2024-09-29T12:05:28-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 (cherry picked from commit 5092dbff750ee5b6fd082b7eed8574922a2b0bf4) - - - - - 1d3bd636 by Sylvain Henry at 2024-09-29T12:05:28-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. (cherry picked from commit c749bdfd3e21d712dc2b966482eb010165bdeebe) - - - - - b060a7a7 by Sylvain Henry at 2024-09-29T12:05:28-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> (cherry picked from commit 7446a09a2d5b04b95cd43c03659b5647853124ce) - - - - - 532c5abf by sheaf at 2024-09-29T12:05:28-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 (cherry picked from commit c29b2b5a77611b2bd6c3089765079bc43aec3e22) - - - - - eb0e402c by Sylvain Henry at 2024-09-29T12:05:28-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 (cherry picked from commit f954f42823f6ca3588425a0d543d93ace86d89e4) - - - - - 98c7e1bb by Simon Peyton Jones at 2024-09-29T12:05:28-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify (cherry picked from commit de5d9852dbdd367611bf9e45e69c723d26351992) - - - - - 9524086e by doyougnu at 2024-09-29T12:05:28-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches (cherry picked from commit bfe4b3d3bbb98b39169fad063c6c32f06d167756) - - - - - a7b9f6c8 by Ben Gamari at 2024-09-29T12:05:28-04:00 Accept performance shifts Metric Decrease: MultiLayerModulesTH_Make - - - - - ed8f2c53 by Ben Gamari at 2024-09-29T12:05:41-04:00 docs: Add release notes for 9.8.3 - - - - - 4cbf2581 by Andreas Klebinger at 2024-09-29T12:05:41-04:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. (cherry picked from commit 3f89ab92da74c4ed45da68fe92ff81e7b9caa53d) - - - - - a8804baf by Peter Trommler at 2024-09-29T12:05:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 (cherry picked from commit 7fe85b1354a13749f14d588e3cc742b8ae2d8da9) - - - - - b41a5f1b by Ben Gamari at 2024-09-29T12:05:41-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 (cherry picked from commit a82121b3b6fdc2ac47211f71871b3ab21e5f6276) - - - - - 8efa9643 by Sylvain Henry at 2024-09-29T12:05:41-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). (cherry picked from commit 5104ee615503617a1c124fe1d92f6aa2d263b7d0) - - - - - 0b4a9b9c by Sylvain Henry at 2024-09-29T12:05:41-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC (cherry picked from commit 52d6698479f951e07def237b0474ee22d27e621a) - - - - - 615e6061 by Sylvain Henry at 2024-09-29T12:05:41-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). (cherry picked from commit c34fef56367142fa55e9861092f64cc7b9946fa1) - - - - - 54c5f12f by Cheng Shao at 2024-09-29T12:05:41-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. (cherry picked from commit 0cff083abb24701530974872b21cf897c9955a9a) - - - - - 43e73dfb by Andreas Klebinger at 2024-09-29T12:05:41-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. (cherry picked from commit 1bfa91115b8320ed99a5e946147528e21ca4f3e1) - - - - - 8e4198ad by crumbtoo at 2024-09-29T12:05:41-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 (cherry picked from commit c5e00c35927d574f71bf77449817b131d1749750) - - - - - e5ae07d4 by Ryan Scott at 2024-09-29T12:05:41-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. (cherry picked from commit a3cd3a1d0d186f2aa4d0273c6b3e74a442de2ef0) - - - - - 80cc3668 by Sylvain Henry at 2024-09-29T12:05:41-04:00 Reverse arguments to stgCallocBytes (fix #24828) (cherry picked from commit 6838a7c32ca29b5d44adc9d6280d3a960f31be7c) - - - - - 76d4aaa3 by Teo Camarasu at 2024-09-29T12:05:41-04:00 doc: Fix type error in hs_try_putmvar example (cherry picked from commit 06f7db4001e4eee0f3076d949876f8f4af0eb6fb) - - - - - 986f7b48 by Matthew Craven at 2024-09-29T12:05:41-04:00 Add test cases for #24664 ...since none are present in the original MR !12463 fixing this issue. (cherry picked from commit a19201d42cfd3aa54faeb1b5a95b715b9a67a01a) - - - - - a1635569 by Zubin Duggal at 2024-09-29T12:05:41-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 (cherry picked from commit dddc9dff0547733a10e7f505612ab9df3a7c21b6) - - - - - e6063183 by Sylvain Henry at 2024-09-29T12:05:41-04:00 JS: correctly handle RUBBISH literals (#24664) (cherry picked from commit daeda83478d5b800d29661408dd67cc4b23df374) - - - - - 0a1b9687 by Matthew Pickering at 2024-09-29T12:05:41-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. (cherry picked from commit 3fff09779d5830549ae455a15907b7bb9fe7859a) - - - - - e34ba65f by Teo Camarasu at 2024-09-29T12:05:41-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 (cherry picked from commit f30e4984fb048818051465698ef8e4e20dacb577) - - - - - 30 changed files: - .gitlab-ci.yml - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/Reg/Graph/TrivColorable.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Driver/Config/Core/Opt/Simplify.hs - compiler/GHC/Iface/Recomp/Flags.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Literal.hs - compiler/GHC/SysTools/Cpp.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Id.hs - compiler/GHC/Unit/Types.hs - docs/index.html.in - + docs/users_guide/9.8.3-notes.rst - docs/users_guide/exts/ffi.rst - docs/users_guide/exts/multiway_if.rst - docs/users_guide/release-notes.rst - hadrian/src/Hadrian/Haskell/Cabal/Parse.hs - hadrian/src/Rules/Generate.hs - libraries/array - libraries/base/GHC/Event/KQueue.hsc - libraries/deepseq The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/aa93076c5157a813699649230272749fb6511ce9...e34ba65f6b3c0b58dcf28c1d17efe5211871e5a3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/aa93076c5157a813699649230272749fb6511ce9...e34ba65f6b3c0b58dcf28c1d17efe5211871e5a3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 17:56:33 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Sun, 29 Sep 2024 13:56:33 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] Moved HsRecSel to the XConstructors. Message-ID: <66f994d13377f_28e2bb8813a0621aa@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 7c89f357 by Hassan Al-Awwadi at 2024-09-29T19:55:38+02:00 Moved HsRecSel to the XConstructors. FieldOcc types kept in Language.Haskell though because they are also used by HsRecUpdField in L.H.S.Pat.hs - - - - - 13 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/Hs/Syn/Type.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/Expr.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Tc/Gen/Expr.hs - compiler/GHC/Tc/Gen/Head.hs - compiler/GHC/Tc/Types/Origin.hs - compiler/GHC/Tc/Zonk/Type.hs - compiler/Language/Haskell/Syntax/Expr.hs Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -514,6 +514,10 @@ data XXExprGhcRn -- Does not presist post renaming phase -- See Part 3. of Note [Expanding HsDo with XXExprGhcRn] -- in `GHC.Tc.Gen.Do` + | HsRecSelRn (FieldOcc GhcRn) -- ^ Variable pointing to record selector + -- See Note [Non-overloaded record field selectors] and + -- Note [Record selectors in the AST] + -- | Wrap a located expression with a `PopErrCtxt` @@ -594,6 +598,11 @@ data XXExprGhcTc Int -- module-local tick number for True Int -- module-local tick number for False (LHsExpr GhcTc) -- sub-expression + + | HsRecSelTc (FieldOcc GhcTc) -- ^ Variable pointing to record selector + -- See Note [Non-overloaded record field selectors] and + -- Note [Record selectors in the AST] + -- | Build a 'XXExprGhcRn' out of an extension constructor, -- and the two components of the expansion: original and @@ -655,7 +664,6 @@ ppr_expr :: forall p. (OutputableBndrId p) => HsExpr (GhcPass p) -> SDoc ppr_expr (HsVar _ (L _ v)) = pprPrefixOcc v ppr_expr (HsUnboundVar _ uv) = pprPrefixOcc uv -ppr_expr (HsRecSel _ f) = pprPrefixOcc f ppr_expr (HsIPVar _ v) = ppr v ppr_expr (HsOverLabel s l) = case ghcPass @p of GhcPs -> helper s @@ -874,11 +882,13 @@ instance Outputable HsThingRn where OrigExpr x -> ppr_builder ":" x OrigStmt x -> ppr_builder ":" x OrigPat x -> ppr_builder ":" x + where ppr_builder prefix x = ifPprDebug (braces (text prefix <+> parens (ppr x))) (ppr x) instance Outputable XXExprGhcRn where ppr (ExpandedThingRn o e) = ifPprDebug (braces $ vcat [ppr o, ppr e]) (ppr o) ppr (PopErrCtxt e) = ifPprDebug (braces (text "" <+> ppr e)) (ppr e) + ppr (HsRecSelRn f) = pprPrefixOcc f instance Outputable XXExprGhcTc where ppr (WrapExpr co_fn e) @@ -907,10 +917,11 @@ instance Outputable XXExprGhcTc where ppr tickIdFalse, text ">(", ppr exp, text ")"] + ppr (HsRecSelTc f) = pprPrefixOcc f + ppr_infix_expr :: forall p. (OutputableBndrId p) => HsExpr (GhcPass p) -> Maybe SDoc ppr_infix_expr (HsVar _ (L _ v)) = Just (pprInfixOcc v) -ppr_infix_expr (HsRecSel _ f) = Just (pprInfixOcc f) ppr_infix_expr (HsUnboundVar _ occ) = Just (pprInfixOcc occ) ppr_infix_expr (XExpr x) = case ghcPass @p of GhcRn -> ppr_infix_expr_rn x @@ -919,7 +930,8 @@ ppr_infix_expr _ = Nothing ppr_infix_expr_rn :: XXExprGhcRn -> Maybe SDoc ppr_infix_expr_rn (ExpandedThingRn thing _) = ppr_infix_hs_expansion thing -ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a +ppr_infix_expr_rn (PopErrCtxt (L _ a)) = ppr_infix_expr a +ppr_infix_expr_rn (HsRecSelRn f) = Just (pprInfixOcc f) ppr_infix_expr_tc :: XXExprGhcTc -> Maybe SDoc ppr_infix_expr_tc (WrapExpr _ e) = ppr_infix_expr e @@ -927,6 +939,8 @@ ppr_infix_expr_tc (ExpandedThingTc thing _) = ppr_infix_hs_expansion thing ppr_infix_expr_tc (ConLikeTc {}) = Nothing ppr_infix_expr_tc (HsTick {}) = Nothing ppr_infix_expr_tc (HsBinTick {}) = Nothing +ppr_infix_expr_tc (HsRecSelTc f) = Just (pprInfixOcc f) + ppr_infix_hs_expansion :: HsThingRn -> Maybe SDoc ppr_infix_hs_expansion (OrigExpr e) = ppr_infix_expr e @@ -1013,7 +1027,6 @@ hsExprNeedsParens prec = go go (HsProc{}) = prec > topPrec go (HsStatic{}) = prec >= appPrec go (RecordCon{}) = False - go (HsRecSel{}) = False go (HsProjection{}) = True go (HsGetField{}) = False go (HsEmbTy{}) = prec > topPrec @@ -1030,10 +1043,12 @@ hsExprNeedsParens prec = go go_x_tc (ConLikeTc {}) = False go_x_tc (HsTick _ (L _ e)) = hsExprNeedsParens prec e go_x_tc (HsBinTick _ _ (L _ e)) = hsExprNeedsParens prec e + go_x_tc (HsRecSelTc{}) = False go_x_rn :: XXExprGhcRn -> Bool go_x_rn (ExpandedThingRn thing _) = hsExpandedNeedsParens thing go_x_rn (PopErrCtxt (L _ a)) = hsExprNeedsParens prec a + go_x_rn (HsRecSelRn{}) = False hsExpandedNeedsParens :: HsThingRn -> Bool hsExpandedNeedsParens (OrigExpr e) = hsExprNeedsParens prec e @@ -1071,21 +1086,22 @@ isAtomicHsExpr (HsOverLit {}) = True isAtomicHsExpr (HsIPVar {}) = True isAtomicHsExpr (HsOverLabel {}) = True isAtomicHsExpr (HsUnboundVar {}) = True -isAtomicHsExpr (HsRecSel{}) = True isAtomicHsExpr (XExpr x) | GhcTc <- ghcPass @p = go_x_tc x | GhcRn <- ghcPass @p = go_x_rn x where go_x_tc :: XXExprGhcTc -> Bool - go_x_tc (WrapExpr _ e) = isAtomicHsExpr e - go_x_tc (ExpandedThingTc thing _) = isAtomicExpandedThingRn thing - go_x_tc (ConLikeTc {}) = True - go_x_tc (HsTick {}) = False - go_x_tc (HsBinTick {}) = False + go_x_tc (WrapExpr _ e) = isAtomicHsExpr e + go_x_tc (ExpandedThingTc thing _) = isAtomicExpandedThingRn thing + go_x_tc (ConLikeTc {}) = True + go_x_tc (HsTick {}) = False + go_x_tc (HsBinTick {}) = False + go_x_tc (HsRecSelTc{}) = True go_x_rn :: XXExprGhcRn -> Bool - go_x_rn (ExpandedThingRn thing _) = isAtomicExpandedThingRn thing - go_x_rn (PopErrCtxt (L _ a)) = isAtomicHsExpr a + go_x_rn (ExpandedThingRn thing _) = isAtomicExpandedThingRn thing + go_x_rn (PopErrCtxt (L _ a)) = isAtomicHsExpr a + go_x_rn (HsRecSelRn{}) = True isAtomicExpandedThingRn :: HsThingRn -> Bool isAtomicExpandedThingRn (OrigExpr e) = isAtomicHsExpr e ===================================== compiler/GHC/Hs/Syn/Type.hs ===================================== @@ -103,7 +103,6 @@ lhsExprType (L _ e) = hsExprType e hsExprType :: HsExpr GhcTc -> Type hsExprType (HsVar _ (L _ id)) = idType id hsExprType (HsUnboundVar (HER _ ty _) _) = ty -hsExprType (HsRecSel _ (FieldOcc _ id)) = idType (unLoc id) hsExprType (HsOverLabel v _) = dataConCantHappen v hsExprType (HsIPVar v _) = dataConCantHappen v hsExprType (HsOverLit _ lit) = overLitType lit @@ -154,6 +153,7 @@ hsExprType (XExpr (ExpandedThingTc _ e)) = hsExprType e hsExprType (XExpr (ConLikeTc con _ _)) = conLikeType con hsExprType (XExpr (HsTick _ e)) = lhsExprType e hsExprType (XExpr (HsBinTick _ _ e)) = lhsExprType e +hsExprType (XExpr (HsRecSelTc (FieldOcc _ id))) = idType (unLoc id) arithSeqInfoType :: ArithSeqInfo GhcTc -> Type arithSeqInfoType asi = mkListTy $ case asi of ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -261,36 +261,6 @@ dsLExpr (L loc e) = putSrcSpanDsA loc $ dsExpr e dsExpr :: HsExpr GhcTc -> DsM CoreExpr dsExpr (HsVar _ (L _ id)) = dsHsVar id -{- Record selectors are warned about if they are not -present in all of the parent data type's constructor, -or always in case of pattern synonym record selectors -(regulated by a flag). However, this only produces -a warning if it's not a part of a record selector -application. For example: - - data T = T1 | T2 {s :: Bool} - f x = s x -- the warning from this case will be supressed - -See the `HsApp` case for where it is filtered out --} -dsExpr (HsRecSel _ (FieldOcc _ (L _ id))) - = do { let name = getName id - RecSelId {sel_cons = (_, cons_wo_field)} - = idDetails id - ; cons_trimmed <- trim_cons cons_wo_field - ; unless (null cons_wo_field) $ diagnosticDs - $ DsIncompleteRecordSelector name cons_trimmed (cons_trimmed /= cons_wo_field) - -- This only produces a warning if it's not a part of a - -- record selector application (e.g. `s a` where `s` is a selector) - -- See the `HsApp` case for where it is filtered out - ; dsHsVar id } - where - trim_cons :: [ConLike] -> DsM [ConLike] - trim_cons cons_wo_field = do - dflags <- getDynFlags - let maxConstructors = maxUncoveredPatterns dflags - return $ take maxConstructors cons_wo_field - dsExpr (HsUnboundVar (HER ref _ _) _) = dsEvTerm =<< readMutVar ref -- See Note [Holes] in GHC.Tc.Types.Constraint @@ -336,6 +306,35 @@ dsExpr e@(XExpr ext_expr_tc) do { assert (exprType e2 `eqType` boolTy) mkBinaryTickBox ixT ixF e2 } + {- Record selectors are warned about if they are not + present in all of the parent data type's constructor, + or always in case of pattern synonym record selectors + (regulated by a flag). However, this only produces + a warning if it's not a part of a record selector + application. For example: + + data T = T1 | T2 {s :: Bool} + f x = s x -- the warning from this case will be supressed + + See the `HsApp` case for where it is filtered out + -} + (HsRecSelTc (FieldOcc _ (L _ id))) -> + do { let name = getName id + RecSelId {sel_cons = (_, cons_wo_field)} = idDetails id + ; cons_trimmed <- trim_cons cons_wo_field + ; unless (null cons_wo_field) $ diagnosticDs + $ DsIncompleteRecordSelector name cons_trimmed (cons_trimmed /= cons_wo_field) + -- This only produces a warning if it's not a part of a + -- record selector application (e.g. `s a` where `s` is a selector) + -- See the `HsApp` case for where it is filtered out + ; dsHsVar id } + where + trim_cons :: [ConLike] -> DsM [ConLike] + trim_cons cons_wo_field = do + dflags <- getDynFlags + let maxConstructors = maxUncoveredPatterns dflags + return $ take maxConstructors cons_wo_field + -- Strip ticks due to #21701, need to be invariant about warnings we produce whether -- this is enabled or not. ===================================== compiler/GHC/HsToCore/Quote.hs ===================================== @@ -1544,7 +1544,6 @@ repE (HsVar _ (L _ x)) = repE (HsIPVar _ n) = rep_implicit_param_name n >>= repImplicitParamVar repE (HsOverLabel _ s) = repOverLabel s -repE (HsRecSel _ (FieldOcc _ (L _ x))) = repE (HsVar noExtField (noLocA x)) -- Remember, we're desugaring renamer output here, so -- HsOverlit can definitely occur @@ -1719,6 +1718,8 @@ repE e@(XExpr (ExpandedThingRn o x)) = notHandled (ThExpressionForm e) repE (XExpr (PopErrCtxt (L _ e))) = repE e +repE (XExpr (HsRecSelRn (FieldOcc _ (L _ x)))) = repE (HsVar noExtField (noLocA x)) + repE e@(HsPragE _ (HsPragSCC {}) _) = notHandled (ThCostCentres e) repE e@(HsTypedBracket{}) = notHandled (ThExpressionForm e) repE e@(HsUntypedBracket{}) = notHandled (ThExpressionForm e) ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -476,7 +476,6 @@ addBinTickLHsExpr boxLabel e@(L pos e0) addTickHsExpr :: HsExpr GhcTc -> TM (HsExpr GhcTc) addTickHsExpr e@(HsVar _ (L _ id)) = do freeVar id; return e addTickHsExpr e@(HsUnboundVar {}) = return e -addTickHsExpr e@(HsRecSel _ (FieldOcc _ id)) = do freeVar (unLoc id); return e addTickHsExpr e@(HsIPVar {}) = return e addTickHsExpr e@(HsOverLit {}) = return e @@ -603,6 +602,8 @@ addTickHsExpr (XExpr (HsTick t e)) = liftM (XExpr . HsTick t) (addTickLHsExprNever e) addTickHsExpr (XExpr (HsBinTick t0 t1 e)) = liftM (XExpr . HsBinTick t0 t1) (addTickLHsExprNever e) + +addTickHsExpr e@(XExpr (HsRecSelTc (FieldOcc _ id))) = do freeVar (unLoc id); return e addTickHsExpr (HsDo srcloc cxt (L l stmts)) = do { (stmts', _) <- addTickLStmts' forQual stmts (return ()) ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1205,9 +1205,6 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where -- Patch up var location since typechecker removes it ] HsUnboundVar _ _ -> [] -- there is an unbound name here, but that causes trouble - HsRecSel _ fld -> - [ toHie $ RFC RecFieldOcc Nothing (L mspan fld) - ] HsOverLabel {} -> [] HsIPVar _ _ -> [] HsOverLit _ o -> @@ -1355,23 +1352,29 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where ] HsGetField {} -> [] HsProjection {} -> [] - XExpr x - | HieTc <- hiePass @p - -> case x of - WrapExpr w a - -> [ toHie $ L mspan a - , toHie (L mspan w) ] - ExpandedThingTc _ e - -> [ toHie (L mspan e) ] - ConLikeTc con _ _ - -> [ toHie $ C Use $ L mspan $ conLikeName con ] - HsTick _ expr - -> [ toHie expr - ] - HsBinTick _ _ expr - -> [ toHie expr - ] - | otherwise -> [] + XExpr x -> case hiePass @p of + HieTc -> case x of + WrapExpr w a + -> [ toHie $ L mspan a + , toHie (L mspan w) ] + ExpandedThingTc _ e + -> [ toHie (L mspan e) ] + ConLikeTc con _ _ + -> [ toHie $ C Use $ L mspan $ conLikeName con ] + HsTick _ expr + -> [ toHie expr + ] + HsBinTick _ _ expr + -> [ toHie expr + ] + HsRecSelTc fld + -> [ toHie $ RFC RecFieldOcc Nothing (L mspan fld) + ] + HieRn -> case x of + HsRecSelRn fld + -> [ toHie $ RFC RecFieldOcc Nothing (L mspan fld) + ] + _ -> [] -- NOTE: no longer have the location instance HiePass p => ToHie (HsTupArg (GhcPass p)) where ===================================== compiler/GHC/Rename/Expr.hs ===================================== @@ -339,7 +339,7 @@ rnExpr (HsVar _ (L l v)) ; this_mod <- getModule ; when (nameIsLocalOrFrom this_mod sel_name) $ checkThLocalName sel_name - ; return (HsRecSel noExtField (FieldOcc v (L l sel_name)), unitFV sel_name) + ; return (XExpr (HsRecSelRn (FieldOcc v (L l sel_name))), unitFV sel_name) } | nm == nilDataConName -- Treat [] as an ExplicitList, so that @@ -416,8 +416,8 @@ rnExpr (OpApp _ e1 op e2) -- more, so I've removed the test. Adding HsPars in GHC.Tc.Deriv.Generate -- should prevent bad things happening. ; fixity <- case op' of - L _ (HsVar _ (L _ n)) -> lookupFixityRn n - L _ (HsRecSel _ f) -> lookupFieldFixityRn f + L _ (HsVar _ (L _ n)) -> lookupFixityRn n + L _ (XExpr (HsRecSelRn f)) -> lookupFieldFixityRn f _ -> return (Fixity minPrecedence InfixL) -- c.f. lookupFixity for unbound @@ -588,7 +588,6 @@ rnExpr (RecordUpd { rupd_expr = L l expr, rupd_flds = rbinds }) (mkRecordDotUpd getField setField (L l e) us) , plusFVs [fv_getField, fv_setField, fv_e, fv_us] ) } -rnExpr (HsRecSel x _) = dataConCantHappen x rnExpr (ExprWithTySig _ expr pty) = do { (pty', fvTy) <- rnHsSigWcType ExprWithTySigCtx pty ===================================== compiler/GHC/Rename/HsType.hs ===================================== @@ -1467,10 +1467,10 @@ data NegationHandling = ReassociateNegation | KeepNegationIntact get_op :: LHsExpr GhcRn -> OpName -- An unbound name could be either HsVar or HsUnboundVar -- See GHC.Rename.Expr.rnUnboundVar -get_op (L _ (HsVar _ n)) = NormalOp (unLoc n) -get_op (L _ (HsUnboundVar _ uv)) = UnboundOp uv -get_op (L _ (HsRecSel _ fld)) = RecFldOp fld -get_op other = pprPanic "get_op" (ppr other) +get_op (L _ (HsVar _ n)) = NormalOp (unLoc n) +get_op (L _ (HsUnboundVar _ uv)) = UnboundOp uv +get_op (L _ (XExpr (HsRecSelRn fld))) = RecFldOp fld +get_op other = pprPanic "get_op" (ppr other) -- Parser left-associates everything, but -- derived instances may have correctly-associated things to ===================================== compiler/GHC/Tc/Gen/Expr.hs ===================================== @@ -293,7 +293,6 @@ tcExpr e@(HsApp {}) res_ty = tcApp e res_ty tcExpr e@(OpApp {}) res_ty = tcApp e res_ty tcExpr e@(HsAppType {}) res_ty = tcApp e res_ty tcExpr e@(ExprWithTySig {}) res_ty = tcApp e res_ty -tcExpr e@(HsRecSel {}) res_ty = tcApp e res_ty tcExpr (XExpr e) res_ty = tcXExpr e res_ty @@ -738,7 +737,6 @@ tcXExpr xe@(ExpandedThingRn o e') res_ty | OrigStmt ls@(L loc _) <- o = setSrcSpanA loc $ mkExpandedStmtTc ls <$> tcApp (XExpr xe) res_ty - tcXExpr xe res_ty = tcApp (XExpr xe) res_ty {- ===================================== compiler/GHC/Tc/Gen/Head.hs ===================================== @@ -567,7 +567,7 @@ tcInferAppHead_maybe :: HsExpr GhcRn tcInferAppHead_maybe fun = case fun of HsVar _ (L _ nm) -> Just <$> tcInferId nm - HsRecSel _ f -> Just <$> tcInferRecSelId f + XExpr (HsRecSelRn f) -> Just <$> tcInferRecSelId f ExprWithTySig _ e hs_ty -> Just <$> tcExprWithSig e hs_ty HsOverLit _ lit -> Just <$> tcInferOverLit lit _ -> return Nothing @@ -599,7 +599,7 @@ tcInferRecSelId :: FieldOcc GhcRn -> TcM ( (HsExpr GhcTc, TcSigmaType)) tcInferRecSelId (FieldOcc sel_name (L l n)) = do { sel_id <- tc_rec_sel_id - ; let expr = HsRecSel noExtField (FieldOcc sel_name (L l sel_id)) + ; let expr = XExpr (HsRecSelTc (FieldOcc sel_name (L l sel_id))) ; return $ (expr, idType sel_id) } where ===================================== compiler/GHC/Tc/Types/Origin.hs ===================================== @@ -718,7 +718,6 @@ exprCtOrigin :: HsExpr GhcRn -> CtOrigin exprCtOrigin (HsVar _ (L _ name)) = OccurrenceOf name exprCtOrigin (HsGetField _ _ (L _ f)) = HasFieldOrigin (field_label $ unLoc $ dfoLabel f) exprCtOrigin (HsUnboundVar {}) = Shouldn'tHappenOrigin "unbound variable" -exprCtOrigin (HsRecSel _ f) = OccurrenceOfRecSel (foExt f) exprCtOrigin (HsOverLabel _ l) = OverLabelOrigin l exprCtOrigin (ExplicitList {}) = ListOrigin exprCtOrigin (HsIPVar _ ip) = IPOccOrigin ip @@ -759,6 +758,7 @@ exprCtOrigin (XExpr (ExpandedThingRn thing _)) | OrigExpr a <- thing = exprCtOri | OrigStmt _ <- thing = DoOrigin | OrigPat p <- thing = DoPatOrigin p exprCtOrigin (XExpr (PopErrCtxt {})) = Shouldn'tHappenOrigin "PopErrCtxt" +exprCtOrigin (XExpr (HsRecSelRn f)) = OccurrenceOfRecSel (foExt f) -- | Extract a suitable CtOrigin from a MatchGroup matchesCtOrigin :: MatchGroup GhcRn (LHsExpr GhcRn) -> CtOrigin ===================================== compiler/GHC/Tc/Zonk/Type.hs ===================================== @@ -935,9 +935,6 @@ zonkExpr (HsUnboundVar her occ) ty' <- zonkTcTypeToTypeX ty return (HER ref ty' u) -zonkExpr (HsRecSel _ (FieldOcc occ (L l v))) - = do { v' <- zonkIdOcc v - ; return (HsRecSel noExtField (FieldOcc occ (L l v'))) } zonkExpr (HsIPVar x _) = dataConCantHappen x @@ -1098,6 +1095,10 @@ zonkExpr (XExpr (ConLikeTc con tvs tys)) -- The tvs come straight from the data-con, and so are strictly redundant -- See Wrinkles of Note [Typechecking data constructors] in GHC.Tc.Gen.Head +zonkExpr (XExpr (HsRecSelTc (FieldOcc occ (L l v)))) + = do { v' <- zonkIdOcc v + ; return (XExpr (HsRecSelTc (FieldOcc occ (L l v')))) } + zonkExpr (RecordUpd x _ _) = dataConCantHappen x zonkExpr (HsGetField x _ _) = dataConCantHappen x zonkExpr (HsProjection x _) = dataConCantHappen x ===================================== compiler/Language/Haskell/Syntax/Expr.hs ===================================== @@ -351,10 +351,6 @@ data HsExpr p -- solving. See Note [Holes] in GHC.Tc.Types.Constraint. - | HsRecSel (XRecSel p) - (FieldOcc p) -- ^ Variable pointing to record selector - -- See Note [Non-overloaded record field selectors] and - -- Note [Record selectors in the AST] | HsOverLabel (XOverLabel p) FastString -- ^ Overloaded label (Note [Overloaded labels] in GHC.OverloadedLabels) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7c89f35761a851f7a8be9f03d5b72f0bbfd8d17f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/7c89f35761a851f7a8be9f03d5b72f0bbfd8d17f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 22:06:58 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Sun, 29 Sep 2024 18:06:58 -0400 Subject: [Git][ghc/ghc] Deleted branch wip/backports-9.8 Message-ID: <66f9cf82cb678_28e2bb110a0e4699a9@gitlab.mail> Ben Gamari deleted branch wip/backports-9.8 at Glasgow Haskell Compiler / GHC -- You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 22:07:01 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Sun, 29 Sep 2024 18:07:01 -0400 Subject: [Git][ghc/ghc][ghc-9.8] 17 commits: Bump stm submodule to v2.5.3.1 Message-ID: <66f9cf851b760_28e2bb10c627c7014c@gitlab.mail> Ben Gamari pushed to branch ghc-9.8 at Glasgow Haskell Compiler / GHC Commits: 83126193 by Ben Gamari at 2024-09-29T12:05:28-04:00 Bump stm submodule to v2.5.3.1 - - - - - c2c85f36 by Andrew Lelechenko at 2024-09-29T12:05:28-04:00 Bump submodule deepseq to 1.5.1.0 (cherry picked from commit 8e462f4d4bdf2a6c34c249e7be8084565600d300) - - - - - af73ed1e by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: ensure gc_thread/gen_workspace is allocated with proper alignment gc_thread/gen_workspace are required to be aligned by 64 bytes. However, this property has not been properly enforced before, and numerous alignment violations at runtime has been caught by UndefinedBehaviorSanitizer that look like: ``` rts/sm/GC.c:1167:8: runtime error: member access within misaligned address 0x0000027a3390 for type 'gc_thread' (aka 'struct gc_thread_'), which requires 64 byte alignment 0x0000027a3390: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1167:8 rts/sm/GC.c:1184:13: runtime error: member access within misaligned address 0x0000027a3450 for type 'gen_workspace' (aka 'struct gen_workspace_'), which requires 64 byte alignment 0x0000027a3450: note: pointer points here 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/GC.c:1184:13 ``` This patch fixes the gc_thread/gen_workspace misalignment issue by explicitly allocating them with alignment constraint. - - - - - 5b2b8321 by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: fix an unaligned load in nonmoving gc This patch fixes an unaligned load in nonmoving gc by ensuring the closure address is properly untagged first before attempting to prefetch its header. The unaligned load is reported by UndefinedBehaviorSanitizer: ``` rts/sm/NonMovingMark.c:921:9: runtime error: member access within misaligned address 0x0042005f3a71 for type 'StgClosure' (aka 'struct StgClosure_'), which requires 8 byte alignment 0x0042005f3a71: note: pointer points here 00 00 00 98 43 13 8e 12 7f 00 00 50 3c 5f 00 42 00 00 00 58 17 b7 92 12 7f 00 00 89 cb 5e 00 42 ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/sm/NonMovingMark.c:921:9 ``` This issue had previously gone unnoticed since it didn't really harm runtime correctness, the invalid header address directly loaded from a tagged pointer is only used as prefetch address and will not cause segfaults. However, it still should be corrected because the prefetch would be rendered useless by this issue, and untagging only involves a single bitwise operation without memory access so it's cheap enough to add. - - - - - 83341bbc by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: use __builtin_offsetof to implement STG_FIELD_OFFSET This patch fixes the STG_FIELD_OFFSET macro definition by using __builtin_offsetof, which is what gcc/clang uses to implement offsetof in standard C. The previous definition that uses NULL pointer involves subtle undefined behavior in C and thus reported by UndefinedBehaviorSanitizer as well: ``` rts/Capability.h:243:58: runtime error: member access within null pointer of type 'Capability' (aka 'struct Capability_') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior rts/Capability.h:243:58 ``` - - - - - dea8a144 by Cheng Shao at 2024-09-29T12:05:28-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - 96a15414 by Andrew Lelechenko at 2024-09-29T12:05:28-04:00 Bump submodule array to 0.5.8.0 (cherry picked from commit 80769bc9f56541601796366485283a697c52a18b) - - - - - d5e89136 by Matthew Pickering at 2024-09-29T12:05:28-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 (cherry picked from commit 7cb7172eff9e18ed1c98f65a98018f76c4fcc9f6) - - - - - bcfa1a51 by Sylvain Henry at 2024-09-29T12:05:28-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 (cherry picked from commit e7a26d7a6faf1ea534e036c5085a0a027dbb6f5f) - - - - - e34c4192 by Sylvain Henry at 2024-09-29T12:05:28-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 (cherry picked from commit 5092dbff750ee5b6fd082b7eed8574922a2b0bf4) - - - - - 1d3bd636 by Sylvain Henry at 2024-09-29T12:05:28-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. (cherry picked from commit c749bdfd3e21d712dc2b966482eb010165bdeebe) - - - - - b060a7a7 by Sylvain Henry at 2024-09-29T12:05:28-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> (cherry picked from commit 7446a09a2d5b04b95cd43c03659b5647853124ce) - - - - - 532c5abf by sheaf at 2024-09-29T12:05:28-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 (cherry picked from commit c29b2b5a77611b2bd6c3089765079bc43aec3e22) - - - - - eb0e402c by Sylvain Henry at 2024-09-29T12:05:28-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 (cherry picked from commit f954f42823f6ca3588425a0d543d93ace86d89e4) - - - - - 98c7e1bb by Simon Peyton Jones at 2024-09-29T12:05:28-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify (cherry picked from commit de5d9852dbdd367611bf9e45e69c723d26351992) - - - - - 9524086e by doyougnu at 2024-09-29T12:05:28-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches (cherry picked from commit bfe4b3d3bbb98b39169fad063c6c32f06d167756) - - - - - a7b9f6c8 by Ben Gamari at 2024-09-29T12:05:28-04:00 Accept performance shifts Metric Decrease: MultiLayerModulesTH_Make - - - - - 30 changed files: - .gitlab-ci.yml - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Driver/Config/Core/Opt/Simplify.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/StgToJS/Literal.hs - compiler/GHC/SysTools/Cpp.hs - compiler/GHC/Types/Id.hs - compiler/GHC/Unit/Types.hs - libraries/array - libraries/base/GHC/Event/KQueue.hsc - libraries/deepseq - libraries/stm - rts/include/Stg.h - rts/linker/PEi386.c - rts/linker/elf_reloc_aarch64.c - rts/sm/GC.c - rts/sm/NonMovingMark.c - rts/sm/Sanity.c - + testsuite/tests/codeGen/should_compile/T25177.hs - + testsuite/tests/codeGen/should_compile/T25177.stderr - testsuite/tests/codeGen/should_compile/all.T - + testsuite/tests/codeGen/should_run/T24700.hs - + testsuite/tests/codeGen/should_run/T24700.stdin - + testsuite/tests/codeGen/should_run/T24700.stdout - + testsuite/tests/codeGen/should_run/T24893.hs - + testsuite/tests/codeGen/should_run/T24893.stdout - testsuite/tests/codeGen/should_run/all.T - testsuite/tests/driver/T20604/T20604.stdout The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/78a253543d466ac511a1664a3e6aff032ca684d5...a7b9f6c8b6b1629f6f9b9d47ffcf68d9e27aef8b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/78a253543d466ac511a1664a3e6aff032ca684d5...a7b9f6c8b6b1629f6f9b9d47ffcf68d9e27aef8b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 22:07:24 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Sun, 29 Sep 2024 18:07:24 -0400 Subject: [Git][ghc/ghc] Deleted branch wip/backports-9.8-2 Message-ID: <66f9cf9c1fdd1_28e2bb110ae04732ac@gitlab.mail> Ben Gamari deleted branch wip/backports-9.8-2 at Glasgow Haskell Compiler / GHC -- You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 22:07:26 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Sun, 29 Sep 2024 18:07:26 -0400 Subject: [Git][ghc/ghc][ghc-9.8] 18 commits: docs: Add release notes for 9.8.3 Message-ID: <66f9cf9ee49a5_28e2bb10d6834734dc@gitlab.mail> Ben Gamari pushed to branch ghc-9.8 at Glasgow Haskell Compiler / GHC Commits: ed8f2c53 by Ben Gamari at 2024-09-29T12:05:41-04:00 docs: Add release notes for 9.8.3 - - - - - 4cbf2581 by Andreas Klebinger at 2024-09-29T12:05:41-04:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. (cherry picked from commit 3f89ab92da74c4ed45da68fe92ff81e7b9caa53d) - - - - - a8804baf by Peter Trommler at 2024-09-29T12:05:41-04:00 PPC NCG: Fix sign hints in C calls Sign hints for parameters are in the second component of the pair. Fixes #23034 (cherry picked from commit 7fe85b1354a13749f14d588e3cc742b8ae2d8da9) - - - - - b41a5f1b by Ben Gamari at 2024-09-29T12:05:41-04:00 X86 NCG: Fix argument promotion in foreign C calls Promote 8 bit and 16 bit signed arguments by sign extension. Fixes #25018 (cherry picked from commit a82121b3b6fdc2ac47211f71871b3ab21e5f6276) - - - - - 8efa9643 by Sylvain Henry at 2024-09-29T12:05:41-04:00 Linker: use m32 allocator for sections when NEED_PLT (#24432) Use M32 allocator to avoid fragmentation when allocating ELF sections. We already did this when NEED_PLT was undefined. Failing to do this led to relocations impossible to fulfil (#24432). (cherry picked from commit 5104ee615503617a1c124fe1d92f6aa2d263b7d0) - - - - - 0b4a9b9c by Sylvain Henry at 2024-09-29T12:05:41-04:00 RTS: allow M32 allocation outside of 4GB range when assuming -fPIC (cherry picked from commit 52d6698479f951e07def237b0474ee22d27e621a) - - - - - 615e6061 by Sylvain Henry at 2024-09-29T12:05:41-04:00 Linker: fix stub offset Remove unjustified +8 offset that leads to memory corruption (cf discussion in #24432). (cherry picked from commit c34fef56367142fa55e9861092f64cc7b9946fa1) - - - - - 54c5f12f by Cheng Shao at 2024-09-29T12:05:41-04:00 compiler: emit NaturallyAligned when element type & index type are the same width This commit fixes a subtle mistake in alignmentFromTypes that used to generate Unaligned when element type & index type are the same width. Fixes #24930. (cherry picked from commit 0cff083abb24701530974872b21cf897c9955a9a) - - - - - 43e73dfb by Andreas Klebinger at 2024-09-29T12:05:41-04:00 GHCi interpreter: Tag constructor closures when possible. When evaluating PUSH_G try to tag the reference we are pushing if it's a constructor. This is potentially helpful for performance and required to fix #24870. (cherry picked from commit 1bfa91115b8320ed99a5e946147528e21ca4f3e1) - - - - - 8e4198ad by crumbtoo at 2024-09-29T12:05:41-04:00 user_guide: Fix typo in MultiWayIf chapter Close #24829 (cherry picked from commit c5e00c35927d574f71bf77449817b131d1749750) - - - - - e5ae07d4 by Ryan Scott at 2024-09-29T12:05:41-04:00 Add missing parenthesizePat in cvtp We need to ensure that the output of `cvtp` is parenthesized (at precedence `sigPrec`) so that any pattern signatures with a surrounding pattern signature can parse correctly. Fixes #24837. (cherry picked from commit a3cd3a1d0d186f2aa4d0273c6b3e74a442de2ef0) - - - - - 80cc3668 by Sylvain Henry at 2024-09-29T12:05:41-04:00 Reverse arguments to stgCallocBytes (fix #24828) (cherry picked from commit 6838a7c32ca29b5d44adc9d6280d3a960f31be7c) - - - - - 76d4aaa3 by Teo Camarasu at 2024-09-29T12:05:41-04:00 doc: Fix type error in hs_try_putmvar example (cherry picked from commit 06f7db4001e4eee0f3076d949876f8f4af0eb6fb) - - - - - 986f7b48 by Matthew Craven at 2024-09-29T12:05:41-04:00 Add test cases for #24664 ...since none are present in the original MR !12463 fixing this issue. (cherry picked from commit a19201d42cfd3aa54faeb1b5a95b715b9a67a01a) - - - - - a1635569 by Zubin Duggal at 2024-09-29T12:05:41-04:00 compiler: Fingerprint -fwrite-if-simplified-core We need to recompile if this flag is changed because later modules might depend on the simplified core for this module if -fprefer-bytecode is enabled. Fixes #24656 (cherry picked from commit dddc9dff0547733a10e7f505612ab9df3a7c21b6) - - - - - e6063183 by Sylvain Henry at 2024-09-29T12:05:41-04:00 JS: correctly handle RUBBISH literals (#24664) (cherry picked from commit daeda83478d5b800d29661408dd67cc4b23df374) - - - - - 0a1b9687 by Matthew Pickering at 2024-09-29T12:05:41-04:00 Don't depend on registerPackage function in Cabal More recent versions of Cabal modify the behaviour of libAbiHash which breaks our usage of registerPackage. It is simpler to inline the part of registerPackage that we need and avoid any additional dependency and complication using the higher-level function introduces. (cherry picked from commit 3fff09779d5830549ae455a15907b7bb9fe7859a) - - - - - e34ba65f by Teo Camarasu at 2024-09-29T12:05:41-04:00 Fix ghc API link in docs/index.html This was missing part of the unit ID meaning it would 404. Resolves #24674 (cherry picked from commit f30e4984fb048818051465698ef8e4e20dacb577) - - - - - 30 changed files: - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/PPC/CodeGen.hs - compiler/GHC/CmmToAsm/Reg/Graph/TrivColorable.hs - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/Iface/Recomp/Flags.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Literal.hs - compiler/GHC/ThToHs.hs - docs/index.html.in - + docs/users_guide/9.8.3-notes.rst - docs/users_guide/exts/ffi.rst - docs/users_guide/exts/multiway_if.rst - docs/users_guide/release-notes.rst - hadrian/src/Hadrian/Haskell/Cabal/Parse.hs - hadrian/src/Rules/Generate.hs - rts/Interpreter.c - rts/linker/Elf.c - rts/linker/M32Alloc.c - rts/linker/elf_plt.c - rts/linker/macho/plt.c - + testsuite/tests/codeGen/should_gen_asm/T24941.hs - testsuite/tests/codeGen/should_gen_asm/all.T - + testsuite/tests/codeGen/should_run/T23034.h - + testsuite/tests/codeGen/should_run/T23034.hs - + testsuite/tests/codeGen/should_run/T23034.stdout - + testsuite/tests/codeGen/should_run/T23034_c.c - + testsuite/tests/codeGen/should_run/T24664a.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a7b9f6c8b6b1629f6f9b9d47ffcf68d9e27aef8b...e34ba65f6b3c0b58dcf28c1d17efe5211871e5a3 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a7b9f6c8b6b1629f6f9b9d47ffcf68d9e27aef8b...e34ba65f6b3c0b58dcf28c1d17efe5211871e5a3 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Sun Sep 29 23:11:52 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Sun, 29 Sep 2024 19:11:52 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/T25266 Message-ID: <66f9deb8dec99_28e2bb143ec3876297@gitlab.mail> Simon Peyton Jones pushed new branch wip/T25266 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/T25266 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 00:03:10 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Sun, 29 Sep 2024 20:03:10 -0400 Subject: [Git][ghc/ghc][wip/bump-process] Bump process submodule to v1.6.25.0 Message-ID: <66f9eabee068d_28e2bb15c1164766c@gitlab.mail> Ben Gamari pushed to branch wip/bump-process at Glasgow Haskell Compiler / GHC Commits: 69ae6b2e by Ben Gamari at 2024-09-29T20:02:56-04:00 Bump process submodule to v1.6.25.0 - - - - - 1 changed file: - libraries/process Changes: ===================================== libraries/process ===================================== @@ -1 +1 @@ -Subproject commit a53f925e3ee246e2429418b7a088ecaa0976007b +Subproject commit b8c88fb5bbdebbcbb3e7c734f0c7515dd3cef84e View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/69ae6b2ef9e11552786693246f7374d104a8331c -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/69ae6b2ef9e11552786693246f7374d104a8331c You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 07:58:44 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Mon, 30 Sep 2024 03:58:44 -0400 Subject: [Git][ghc/ghc][wip/llvm_19] 56 commits: ci: Run abi-test on test-abi label Message-ID: <66fa5a34ef0af_258c83915a8c5726d@gitlab.mail> Matthew Pickering pushed to branch wip/llvm_19 at Glasgow Haskell Compiler / GHC Commits: 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - 3f828d14 by Matthew Pickering at 2024-09-30T07:58:38+00:00 Bump LLVM upper bound to allow LLVM 19 Also bumps the ci-images commit so that the deb12 images uses LLVM 19 for testing. ------------------------- Metric Decrease: size_hello_artifact_gzip size_hello_unicode_gzip ------------------------- Fixes #25295 - - - - - 21 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9c71310fc41cb47586f5554bf72ec3527590f4ed...3f828d14eef0b17c0a54c53281ea87c93aa26cbc -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9c71310fc41cb47586f5554bf72ec3527590f4ed...3f828d14eef0b17c0a54c53281ea87c93aa26cbc You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 08:17:15 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Mon, 30 Sep 2024 04:17:15 -0400 Subject: [Git][ghc/ghc] Pushed new tag ghc-9.13-start Message-ID: <66fa5e8b3ae37_258c83aaac1c588aa@gitlab.mail> Zubin pushed new tag ghc-9.13-start at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/ghc-9.13-start You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 08:17:44 2024 From: gitlab at gitlab.haskell.org (Zubin (@wz1000)) Date: Mon, 30 Sep 2024 04:17:44 -0400 Subject: [Git][ghc/ghc] Pushed new branch ghc-9.12 Message-ID: <66fa5ea8e9880_258c83a809085907b@gitlab.mail> Zubin pushed new branch ghc-9.12 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/ghc-9.12 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 09:21:10 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 30 Sep 2024 05:21:10 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] 2 commits: Exception rethrowing Message-ID: <66fa6d865d4da_24360919bef0854cd@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: 8a62bdcc by Matthew Pickering at 2024-09-30T10:21:02+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. Implements CLC proposal#202 <https://github.com/haskell/core-libraries-committee/issues/202> - - - - - bc17852d by Rodrigo Mesquita at 2024-09-30T10:21:02+01:00 Add test for #25300 - - - - - 30 changed files: - docs/users_guide/9.12.1-notes.rst - libraries/base/changelog.md - libraries/base/src/Control/Exception.hs - libraries/base/tests/IO/T21336/T21336b.stderr - libraries/base/tests/IO/T4808.stderr - libraries/base/tests/IO/mkdirExists.stderr - libraries/base/tests/IO/openFile002.stderr - libraries/base/tests/IO/withBinaryFile001.stderr - libraries/base/tests/IO/withBinaryFile002.stderr - libraries/base/tests/IO/withFile001.stderr - libraries/base/tests/IO/withFile002.stderr - libraries/base/tests/IO/withFileBlocking001.stderr - libraries/base/tests/IO/withFileBlocking002.stderr - libraries/base/tests/T15349.stderr - libraries/ghc-internal/src/GHC/Internal/Control/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Control/Exception/Base.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs - libraries/ghc-internal/src/GHC/Internal/IO.hs - libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Text.hs - libraries/ghc-internal/src/GHC/Internal/System/IO/Error.hs - testsuite/tests/codeGen/should_run/cgrun016.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/concurrent/should_run/T3279.hs - + testsuite/tests/exceptions/T25300.hs - + testsuite/tests/exceptions/T25300a.stdout - testsuite/tests/exceptions/all.T - testsuite/tests/ffi/should_run/T7170.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fa92c2f414dcfa6172e35efb75e4da649a2ebe7c...bc17852db59b8fc4f2f505610e9a0aaf92234e29 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/fa92c2f414dcfa6172e35efb75e4da649a2ebe7c...bc17852db59b8fc4f2f505610e9a0aaf92234e29 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 09:44:14 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 30 Sep 2024 05:44:14 -0400 Subject: [Git][ghc/ghc][wip/exception-propagate] 8 commits: Fix typo in Prelude doc for (>>=) Message-ID: <66fa72ee11f19_2436093e051c9055c@gitlab.mail> Rodrigo Mesquita pushed to branch wip/exception-propagate at Glasgow Haskell Compiler / GHC Commits: 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - 7d6945cd by Matthew Pickering at 2024-09-30T10:31:29+01:00 Fix toException method for ExceptionWithContext Fixes #25235 - - - - - 1b121165 by Matthew Pickering at 2024-09-30T10:43:34+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. Implements CLC proposal#202 <https://github.com/haskell/core-libraries-committee/issues/202> - - - - - f9591422 by Rodrigo Mesquita at 2024-09-30T10:43:34+01:00 Add test for #25300 - - - - - 30 changed files: - .gitlab/ci.sh - compiler/GHC/Hs/Decls.hs - compiler/GHC/Hs/Instances.hs - compiler/GHC/Hs/Type.hs - compiler/GHC/HsToCore/Quote.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Parser.y - compiler/GHC/Parser/PostProcess.hs - compiler/GHC/Rename/HsType.hs - compiler/GHC/Rename/Module.hs - compiler/GHC/Tc/Errors/Ppr.hs - compiler/GHC/Tc/Errors/Types.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/GHC/Tc/Gen/Sig.hs - compiler/GHC/Tc/TyCl.hs - compiler/GHC/ThToHs.hs - compiler/GHC/Types/Error/Codes.hs - compiler/Language/Haskell/Syntax/Extension.hs - compiler/Language/Haskell/Syntax/Type.hs - configure.ac - − docs/users_guide/9.12.1-notes.rst - + docs/users_guide/9.14.1-notes.rst - docs/users_guide/exts/type_abstractions.rst - libraries/base/changelog.md - libraries/base/src/Control/Exception.hs - libraries/base/tests/IO/T21336/T21336b.stderr - libraries/base/tests/IO/T4808.stderr - libraries/base/tests/IO/mkdirExists.stderr - libraries/base/tests/IO/openFile002.stderr - libraries/base/tests/IO/withBinaryFile001.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/bc17852db59b8fc4f2f505610e9a0aaf92234e29...f959142263a0fccbc0b7b9328104a60283de9e49 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/bc17852db59b8fc4f2f505610e9a0aaf92234e29...f959142263a0fccbc0b7b9328104a60283de9e49 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 09:47:04 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 30 Sep 2024 05:47:04 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/romes/exception-propagation-backport Message-ID: <66fa73981b45e_2436093d79d091017@gitlab.mail> Rodrigo Mesquita pushed new branch wip/romes/exception-propagation-backport at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/romes/exception-propagation-backport You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 09:54:34 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 30 Sep 2024 05:54:34 -0400 Subject: [Git][ghc/ghc][wip/romes/exception-propagation-backport] 4 commits: Fix toException method for ExceptionWithContext Message-ID: <66fa755aa9799_2436094cefb410045a@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exception-propagation-backport at Glasgow Haskell Compiler / GHC Commits: 217efe2a by Matthew Pickering at 2024-09-30T10:54:25+01:00 Fix toException method for ExceptionWithContext Fixes #25235 - - - - - 8416eed0 by Matthew Pickering at 2024-09-30T10:54:25+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. Implements CLC proposal#202 <https://github.com/haskell/core-libraries-committee/issues/202> - - - - - 0e5cff66 by Rodrigo Mesquita at 2024-09-30T10:54:25+01:00 Add test for #25300 - - - - - e44e448e by Rodrigo Mesquita at 2024-09-30T10:54:25+01:00 Backport !13302 docs to users guide - - - - - 30 changed files: - docs/users_guide/9.12.1-notes.rst - libraries/base/changelog.md - libraries/base/src/Control/Exception.hs - libraries/base/tests/IO/T21336/T21336b.stderr - libraries/base/tests/IO/T4808.stderr - libraries/base/tests/IO/mkdirExists.stderr - libraries/base/tests/IO/openFile002.stderr - libraries/base/tests/IO/withBinaryFile001.stderr - libraries/base/tests/IO/withBinaryFile002.stderr - libraries/base/tests/IO/withFile001.stderr - libraries/base/tests/IO/withFile002.stderr - libraries/base/tests/IO/withFileBlocking001.stderr - libraries/base/tests/IO/withFileBlocking002.stderr - libraries/base/tests/T15349.stderr - libraries/ghc-internal/src/GHC/Internal/Control/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Control/Exception/Base.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Type.hs - libraries/ghc-internal/src/GHC/Internal/IO.hs - libraries/ghc-internal/src/GHC/Internal/IO/Exception.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/FD.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Internals.hs - libraries/ghc-internal/src/GHC/Internal/IO/Handle/Text.hs - libraries/ghc-internal/src/GHC/Internal/System/IO/Error.hs - testsuite/tests/codeGen/should_run/cgrun016.stderr - testsuite/tests/codeGen/should_run/cgrun025.stderr - testsuite/tests/concurrent/should_run/T3279.hs - + testsuite/tests/exceptions/T25300.hs - + testsuite/tests/exceptions/T25300a.stdout - testsuite/tests/exceptions/all.T - testsuite/tests/ffi/should_run/T7170.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9064b6b67c74622a363b60ff1bf0e526e08865a4...e44e448ea8745a04724420edfa6ab4d24252a53f -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/9064b6b67c74622a363b60ff1bf0e526e08865a4...e44e448ea8745a04724420edfa6ab4d24252a53f You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 10:19:42 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 30 Sep 2024 06:19:42 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/andreask/riscv_minmax Message-ID: <66fa7b3e1a36b_24360983f310103159@gitlab.mail> Andreas Klebinger pushed new branch wip/andreask/riscv_minmax at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/andreask/riscv_minmax You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 10:21:36 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 30 Sep 2024 06:21:36 -0400 Subject: [Git][ghc/ghc][wip/andreask/riscv_minmax] Add support for fp min/max to riscv Message-ID: <66fa7bb01d9e6_2436097fcd44103356@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/riscv_minmax at Glasgow Haskell Compiler / GHC Commits: 238d5325 by Andreas Klebinger at 2024-09-30T12:02:21+02:00 Add support for fp min/max to riscv - - - - - 3 changed files: - compiler/GHC/CmmToAsm/RV64/CodeGen.hs - compiler/GHC/CmmToAsm/RV64/Instr.hs - compiler/GHC/CmmToAsm/RV64/Ppr.hs Changes: ===================================== compiler/GHC/CmmToAsm/RV64/CodeGen.hs ===================================== @@ -1109,6 +1109,8 @@ getRegister' config plat expr = MO_F_Mul w -> floatOp w (\d x y -> unitOL $ annExpr expr (MUL d x y)) MO_F_Quot w -> floatOp w (\d x y -> unitOL $ annExpr expr (DIV d x y)) -- Floating point comparison + MO_F_Min w -> floatOp w (\d x y -> unitOL $ annExpr expr (FMIN d x y)) + MO_F_Max w -> floatOp w (\d x y -> unitOL $ annExpr expr (FMAX d x y)) MO_F_Eq w -> floatCond w (\d x y -> unitOL $ annExpr expr (CSET d x y EQ)) MO_F_Ne w -> floatCond w (\d x y -> unitOL $ annExpr expr (CSET d x y NE)) MO_F_Ge w -> floatCond w (\d x y -> unitOL $ annExpr expr (CSET d x y FGE)) @@ -2208,6 +2210,8 @@ makeFarBranches {- only used when debugging -} _platform statics basic_blocks = FENCE {} -> 1 FCVT {} -> 1 FABS {} -> 1 + FMIN {} -> 1 + FMAX {} -> 1 FMA {} -> 1 -- estimate the subsituted size for jumps to lables -- jumps to registers have size 1 ===================================== compiler/GHC/CmmToAsm/RV64/Instr.hs ===================================== @@ -107,6 +107,8 @@ regUsageOfInstr platform instr = case instr of FENCE _ _ -> usage ([], []) FCVT _variant dst src -> usage (regOp src, regOp dst) FABS dst src -> usage (regOp src, regOp dst) + FMIN dst src1 src2 -> usage (regOp src1 ++ regOp src2, regOp dst) + FMAX dst src1 src2 -> usage (regOp src1 ++ regOp src2, regOp dst) FMA _ dst src1 src2 src3 -> usage (regOp src1 ++ regOp src2 ++ regOp src3, regOp dst) _ -> panic $ "regUsageOfInstr: " ++ instrCon instr @@ -203,6 +205,8 @@ patchRegsOfInstr instr env = case instr of FENCE o1 o2 -> FENCE o1 o2 FCVT variant o1 o2 -> FCVT variant (patchOp o1) (patchOp o2) FABS o1 o2 -> FABS (patchOp o1) (patchOp o2) + FMIN o1 o2 o3 -> FMIN (patchOp o1) (patchOp o2) (patchOp o3) + FMAX o1 o2 o3 -> FMAX (patchOp o1) (patchOp o2) (patchOp o3) FMA s o1 o2 o3 o4 -> FMA s (patchOp o1) (patchOp o2) (patchOp o3) (patchOp o4) _ -> panic $ "patchRegsOfInstr: " ++ instrCon instr @@ -603,6 +607,13 @@ data Instr FCVT FcvtVariant Operand Operand | -- | Floating point ABSolute value FABS Operand Operand + + | -- | Min + -- dest = min(r1) + FMIN Operand Operand Operand + | -- | Max + FMAX Operand Operand Operand + | -- | Floating-point fused multiply-add instructions -- -- - fmadd : d = r1 * r2 + r3 @@ -658,6 +669,8 @@ instrCon i = FENCE {} -> "FENCE" FCVT {} -> "FCVT" FABS {} -> "FABS" + FMIN {} -> "FMIN" + FMAX {} -> "FMAX" FMA variant _ _ _ _ -> case variant of FMAdd -> "FMADD" ===================================== compiler/GHC/CmmToAsm/RV64/Ppr.hs ===================================== @@ -666,6 +666,10 @@ pprInstr platform instr = case instr of $ line (pprOp platform o1 <> text "->" <> pprOp platform o2) FABS o1 o2 | isSingleOp o2 -> op2 (text "\tfabs.s") o1 o2 FABS o1 o2 | isDoubleOp o2 -> op2 (text "\tfabs.d") o1 o2 + FMIN o1 o2 o3 | isSingleOp o1 -> op3 (text "\tfmin.s") o1 o2 o3 + | isDoubleOp o2 -> op3 (text "\tfmin.d") o1 o2 o3 + FMAX o1 o2 o3 | isSingleOp o1 -> op3 (text "\tfmax.s") o1 o2 o3 + | isDoubleOp o2 -> op3 (text "\tfmax.d") o1 o2 o3 FMA variant d r1 r2 r3 -> let fma = case variant of FMAdd -> text "\tfmadd" <> dot <> floatPrecission d View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/238d5325c561dd276e1217b8edbbec017aef49ae -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/238d5325c561dd276e1217b8edbbec017aef49ae You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 12:31:11 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Mon, 30 Sep 2024 08:31:11 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] hopefully fixes the assert error* and clears up some whitespace Message-ID: <66fa9a0fe4812_243609db932c1357a0@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: f3656886 by Hassan Al-Awwadi at 2024-09-30T14:30:43+02:00 hopefully fixes the assert error* and clears up some whitespace *in GHC.Rename.HsType. I'm not certain but I also don't know what else it could be... - - - - - 5 changed files: - compiler/GHC/Hs/Expr.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Ext/Ast.hs - compiler/GHC/Rename/HsType.hs Changes: ===================================== compiler/GHC/Hs/Expr.hs ===================================== @@ -598,7 +598,7 @@ data XXExprGhcTc Int -- module-local tick number for True Int -- module-local tick number for False (LHsExpr GhcTc) -- sub-expression - + | HsRecSelTc (FieldOcc GhcTc) -- ^ Variable pointing to record selector -- See Note [Non-overloaded record field selectors] and -- Note [Record selectors in the AST] @@ -882,7 +882,7 @@ instance Outputable HsThingRn where OrigExpr x -> ppr_builder ":" x OrigStmt x -> ppr_builder ":" x OrigPat x -> ppr_builder ":" x - + where ppr_builder prefix x = ifPprDebug (braces (text prefix <+> parens (ppr x))) (ppr x) instance Outputable XXExprGhcRn where ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -318,7 +318,7 @@ dsExpr e@(XExpr ext_expr_tc) See the `HsApp` case for where it is filtered out -} - (HsRecSelTc (FieldOcc _ (L _ id))) -> + (HsRecSelTc (FieldOcc _ (L _ id))) -> do { let name = getName id RecSelId {sel_cons = (_, cons_wo_field)} = idDetails id ; cons_trimmed <- trim_cons cons_wo_field ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -602,7 +602,7 @@ addTickHsExpr (XExpr (HsTick t e)) = liftM (XExpr . HsTick t) (addTickLHsExprNever e) addTickHsExpr (XExpr (HsBinTick t0 t1 e)) = liftM (XExpr . HsBinTick t0 t1) (addTickLHsExprNever e) - + addTickHsExpr e@(XExpr (HsRecSelTc (FieldOcc _ id))) = do freeVar (unLoc id); return e addTickHsExpr (HsDo srcloc cxt (L l stmts)) ===================================== compiler/GHC/Iface/Ext/Ast.hs ===================================== @@ -1352,7 +1352,7 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where ] HsGetField {} -> [] HsProjection {} -> [] - XExpr x -> case hiePass @p of + XExpr x -> case hiePass @p of HieTc -> case x of WrapExpr w a -> [ toHie $ L mspan a @@ -1367,11 +1367,11 @@ instance HiePass p => ToHie (LocatedA (HsExpr (GhcPass p))) where HsBinTick _ _ expr -> [ toHie expr ] - HsRecSelTc fld + HsRecSelTc fld -> [ toHie $ RFC RecFieldOcc Nothing (L mspan fld) ] - HieRn -> case x of - HsRecSelRn fld + HieRn -> case x of + HsRecSelRn fld -> [ toHie $ RFC RecFieldOcc Nothing (L mspan fld) ] _ -> [] ===================================== compiler/GHC/Rename/HsType.hs ===================================== @@ -1341,7 +1341,7 @@ rnField fl_env env (L l (ConDeclField _ names ty haddock_doc)) lookupField :: FastStringEnv FieldLabel -> FieldOcc GhcPs -> FieldOcc GhcRn lookupField fl_env (FieldOcc _ (L lr rdr)) = - FieldOcc rdr (L lr sel) + FieldOcc (mkRdrUnqual $ occName sel) (L lr sel) where lbl = occNameFS $ rdrNameOcc rdr sel = flSelector View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f3656886de845f36ecd9be6427cd9f715efbd127 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f3656886de845f36ecd9be6427cd9f715efbd127 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 13:42:44 2024 From: gitlab at gitlab.haskell.org (Hassan Al-Awwadi (@hassan.awwadi)) Date: Mon, 30 Sep 2024 09:42:44 -0400 Subject: [Git][ghc/ghc][wip/jade/ast] attempt 2 at fixing the asssert error Message-ID: <66faaad46c841_27938237faf053968@gitlab.mail> Hassan Al-Awwadi pushed to branch wip/jade/ast at Glasgow Haskell Compiler / GHC Commits: 484fba49 by Hassan Al-Awwadi at 2024-09-30T15:42:18+02:00 attempt 2 at fixing the asssert error if I could manage to get the full build working on my machine I'd test it locally but alas, I wil have to make the gitlab runners do the work... - - - - - 1 changed file: - compiler/GHC/Rename/Pat.hs Changes: ===================================== compiler/GHC/Rename/Pat.hs ===================================== @@ -902,7 +902,7 @@ rnHsRecFields ctxt mk_arg (HsRecFields { rec_flds = flds, rec_dotdot = dotdot }) ; (rdr_env, lcl_env) <- getRdrEnvs ; conInfo <- lookupConstructorInfo con ; when (conFieldInfo conInfo == ConHasPositionalArgs) (addErr (TcRnIllegalWildcardsInConstructor con)) - ; let present_flds = mkOccSet $ map nameOccName (getFieldLbls flds) + ; let present_flds = mkOccSet $ map rdrNameOcc (getFieldRdrs flds) -- For constructor uses (but not patterns) -- the arg should be in scope locally; @@ -1049,8 +1049,9 @@ rnHsRecUpdFields flds getFieldIds :: [LHsRecField GhcRn arg] -> [Name] getFieldIds flds = map (hsRecFieldSel . unLoc) flds --- The call to foLabel might be partial now. Don't know enough about --- the rest of the function chain to say if this is an issue. +getFieldRdrs :: [LHsRecField GhcRn arg] -> [RdrName] +getFieldRdrs flds = map (foExt . unXRec @GhcRn . hfbLHS . unLoc) flds + getFieldLbls :: forall p arg . UnXRec p => [LHsRecField p arg] -> [IdP p] getFieldLbls = map (unXRec @p . foLabel . unXRec @p . hfbLHS . unXRec @p) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/484fba49b0698a1b7c8de90d98a786d5f55b7814 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/484fba49b0698a1b7c8de90d98a786d5f55b7814 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 14:22:26 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Mon, 30 Sep 2024 10:22:26 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/riscv-fixes Message-ID: <66fab422e7080_334fc81d35e44139e@gitlab.mail> Matthew Pickering pushed new branch wip/riscv-fixes at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/riscv-fixes You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 14:57:14 2024 From: gitlab at gitlab.haskell.org (Matthew Pickering (@mpickering)) Date: Mon, 30 Sep 2024 10:57:14 -0400 Subject: [Git][ghc/ghc][wip/hadrian-cross-stage2] 3 commits: Fix regArch for riscv64 Message-ID: <66fabc4a2f28b_334fc84be70453257@gitlab.mail> Matthew Pickering pushed to branch wip/hadrian-cross-stage2 at Glasgow Haskell Compiler / GHC Commits: 2cf41cd5 by Matthew Pickering at 2024-09-30T14:56:57+01:00 Fix regArch for riscv64 The register allocator doesn't support vector registers on riscv64, therefore advertise as NoVectors. - - - - - 6e0385a2 by Matthew Pickering at 2024-09-30T15:13:13+01:00 riscv: Avoid using csrr instruction to test for vector registers The csrr instruction isn't allowed in qemu user-mode, and raises an illegal instruction error when it is encountered. Therefore for now, we just hard-code that there is no support for vector registers since the rest of the compiler doesn't support vector registers for riscv. Fixes #25312 - - - - - c6cdccbe by Andreas Klebinger at 2024-09-30T15:20:20+01:00 Add support for fp min/max to riscv Fixes #25313 - - - - - 5 changed files: - compiler/GHC/CmmToAsm/RV64/CodeGen.hs - compiler/GHC/CmmToAsm/RV64/Instr.hs - compiler/GHC/CmmToAsm/RV64/Ppr.hs - compiler/GHC/Platform/Reg/Class.hs - rts/CheckVectorSupport.c Changes: ===================================== compiler/GHC/CmmToAsm/RV64/CodeGen.hs ===================================== @@ -1109,6 +1109,8 @@ getRegister' config plat expr = MO_F_Mul w -> floatOp w (\d x y -> unitOL $ annExpr expr (MUL d x y)) MO_F_Quot w -> floatOp w (\d x y -> unitOL $ annExpr expr (DIV d x y)) -- Floating point comparison + MO_F_Min w -> floatOp w (\d x y -> unitOL $ annExpr expr (FMIN d x y)) + MO_F_Max w -> floatOp w (\d x y -> unitOL $ annExpr expr (FMAX d x y)) MO_F_Eq w -> floatCond w (\d x y -> unitOL $ annExpr expr (CSET d x y EQ)) MO_F_Ne w -> floatCond w (\d x y -> unitOL $ annExpr expr (CSET d x y NE)) MO_F_Ge w -> floatCond w (\d x y -> unitOL $ annExpr expr (CSET d x y FGE)) @@ -2208,6 +2210,8 @@ makeFarBranches {- only used when debugging -} _platform statics basic_blocks = FENCE {} -> 1 FCVT {} -> 1 FABS {} -> 1 + FMIN {} -> 1 + FMAX {} -> 1 FMA {} -> 1 -- estimate the subsituted size for jumps to lables -- jumps to registers have size 1 ===================================== compiler/GHC/CmmToAsm/RV64/Instr.hs ===================================== @@ -107,6 +107,8 @@ regUsageOfInstr platform instr = case instr of FENCE _ _ -> usage ([], []) FCVT _variant dst src -> usage (regOp src, regOp dst) FABS dst src -> usage (regOp src, regOp dst) + FMIN dst src1 src2 -> usage (regOp src1 ++ regOp src2, regOp dst) + FMAX dst src1 src2 -> usage (regOp src1 ++ regOp src2, regOp dst) FMA _ dst src1 src2 src3 -> usage (regOp src1 ++ regOp src2 ++ regOp src3, regOp dst) _ -> panic $ "regUsageOfInstr: " ++ instrCon instr @@ -203,6 +205,8 @@ patchRegsOfInstr instr env = case instr of FENCE o1 o2 -> FENCE o1 o2 FCVT variant o1 o2 -> FCVT variant (patchOp o1) (patchOp o2) FABS o1 o2 -> FABS (patchOp o1) (patchOp o2) + FMIN o1 o2 o3 -> FMIN (patchOp o1) (patchOp o2) (patchOp o3) + FMAX o1 o2 o3 -> FMAX (patchOp o1) (patchOp o2) (patchOp o3) FMA s o1 o2 o3 o4 -> FMA s (patchOp o1) (patchOp o2) (patchOp o3) (patchOp o4) _ -> panic $ "patchRegsOfInstr: " ++ instrCon instr @@ -603,6 +607,13 @@ data Instr FCVT FcvtVariant Operand Operand | -- | Floating point ABSolute value FABS Operand Operand + + | -- | Min + -- dest = min(r1) + FMIN Operand Operand Operand + | -- | Max + FMAX Operand Operand Operand + | -- | Floating-point fused multiply-add instructions -- -- - fmadd : d = r1 * r2 + r3 @@ -658,6 +669,8 @@ instrCon i = FENCE {} -> "FENCE" FCVT {} -> "FCVT" FABS {} -> "FABS" + FMIN {} -> "FMIN" + FMAX {} -> "FMAX" FMA variant _ _ _ _ -> case variant of FMAdd -> "FMADD" ===================================== compiler/GHC/CmmToAsm/RV64/Ppr.hs ===================================== @@ -666,6 +666,10 @@ pprInstr platform instr = case instr of $ line (pprOp platform o1 <> text "->" <> pprOp platform o2) FABS o1 o2 | isSingleOp o2 -> op2 (text "\tfabs.s") o1 o2 FABS o1 o2 | isDoubleOp o2 -> op2 (text "\tfabs.d") o1 o2 + FMIN o1 o2 o3 | isSingleOp o1 -> op3 (text "\tfmin.s") o1 o2 o3 + | isDoubleOp o2 -> op3 (text "\tfmin.d") o1 o2 o3 + FMAX o1 o2 o3 | isSingleOp o1 -> op3 (text "\tfmax.s") o1 o2 o3 + | isDoubleOp o2 -> op3 (text "\tfmax.d") o1 o2 o3 FMA variant d r1 r2 r3 -> let fma = case variant of FMAdd -> text "\tfmadd" <> dot <> floatPrecission d ===================================== compiler/GHC/Platform/Reg/Class.hs ===================================== @@ -49,5 +49,8 @@ registerArch arch = ArchPPC -> Unified ArchPPC_64 {} -> Unified ArchAArch64 -> Unified - ArchRISCV64 -> Separate + -- Support for vector registers not yet implemented for RISC-V + -- see panic in `getFreeRegs`. + --ArchRISCV64 -> Separate + ArchRISCV64 -> NoVectors _ -> NoVectors ===================================== rts/CheckVectorSupport.c ===================================== @@ -65,12 +65,16 @@ int checkVectorSupport(void) { */ #elif defined(__riscv) - unsigned long vlenb; - asm volatile ("csrr %0, vlenb" : "=r" (vlenb)); +// csrr instruction nott allowed in user-mode qemu emulation of riscv +// Backend doesn't yet support vector registers, so hard-coded to no vector support +// for now. +// +// unsigned long vlenb; +// asm volatile ("csrr %0, vlenb" : "=r" (vlenb)); // VLENB gives the length in bytes - supports_V16 = vlenb >= 16; - supports_V32 = vlenb >= 32; - supports_V64 = vlenb >= 64; + supports_V16 = 0; + supports_V32 = 0; + supports_V64 = 0; #else // On other platforms, we conservatively return no vector support. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a22ee5c462ca1e2c9d08404b4d62181eb60e6dac...c6cdccbe9843350f35a55169cb2d97f3e48dd1bf -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a22ee5c462ca1e2c9d08404b4d62181eb60e6dac...c6cdccbe9843350f35a55169cb2d97f3e48dd1bf You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 15:12:23 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 30 Sep 2024 11:12:23 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/torsten.schmits/if-bc-pkg-1s Message-ID: <66fabfd74dcd3_334fc85f8df4600da@gitlab.mail> Torsten Schmits pushed new branch wip/torsten.schmits/if-bc-pkg-1s at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/torsten.schmits/if-bc-pkg-1s You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 15:16:00 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 30 Sep 2024 11:16:00 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/package-deps-bytecode-squashed] 206 commits: testsuite: Delay querying ghc-pkg to find .so dirs until test is run Message-ID: <66fac0b0c82f9_334fc84cc16060338@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/package-deps-bytecode-squashed at Glasgow Haskell Compiler / GHC Commits: bace981e by Matthew Pickering at 2024-07-19T10:14:02-04:00 testsuite: Delay querying ghc-pkg to find .so dirs until test is run The tests which relied on find_so would fail when `test` was run before the tree was built. This was because `find_so` was evaluated too eagerly. We can fix this by waiting to query the location of the libraries until after the compiler has built them. - - - - - 478de1ab by Torsten Schmits at 2024-07-19T10:14:37-04:00 Add `complete` pragmas for backwards compat patsyns `ModLocation` and `ModIface` !12347 and !12582 introduced breaking changes to these two constructors and mitigated that with pattern synonyms. - - - - - b57792a8 by Matthew Pickering at 2024-07-19T10:15:13-04:00 ci: Fix ghcup-metadata generation (again) I made some mistakes in 203830065b81fe29003c1640a354f11661ffc604 * Syntax error * The aarch-deb11 bindist doesn't exist I tested against the latest nightly pipeline locally: ``` nix run .gitlab/generate-ci#generate-job-metadata nix shell -f .gitlab/rel_eng/ -c ghcup-metadata --pipeline-id 98286 --version 9.11.20240715 --fragment --date 2024-07-17 --metadata=/tmp/meta ``` - - - - - 1fa35b64 by Andreas Klebinger at 2024-07-19T17:35:20+02:00 Revert "Allow non-absolute values for bootstrap GHC variable" This broke configure in subtle ways resulting in #25076 where hadrian didn't end up the boot compiler it was configured to use. This reverts commit 209d09f52363b261b900cf042934ae1e81e2caa7. - - - - - 55117e13 by Simon Peyton Jones at 2024-07-24T02:41:12-04:00 Fix bad bug in mkSynonymTyCon, re forgetfulness As #25094 showed, the previous tests for forgetfulness was plain wrong, when there was a forgetful synonym in the RHS of a synonym. - - - - - a8362630 by Sergey Vinokurov at 2024-07-24T12:22:45-04:00 Define Eq1, Ord1, Show1 and Read1 instances for basic Generic representation types This way the Generically1 newtype could be used to derive Eq1 and Ord1 for user types with DerivingVia. The CLC proposal is https://github.com/haskell/core-libraries-committee/issues/273. The GHC issue is https://gitlab.haskell.org/ghc/ghc/-/issues/24312. - - - - - de5d9852 by Simon Peyton Jones at 2024-07-24T12:23:22-04:00 Address #25055, by disabling case-of-runRW# in Gentle phase See Note [Case-of-case and full laziness] in GHC.Driver.Config.Core.Opt.Simplify - - - - - 3f89ab92 by Andreas Klebinger at 2024-07-25T14:12:54+02:00 Fix -freg-graphs for FP and AARch64 NCG (#24941). It seems we reserve 8 registers instead of four for global regs based on the layout in Note [AArch64 Register assignments]. I'm not sure it's neccesary, but for now we just accept this state of affairs and simple update -fregs-graph to account for this. - - - - - f6b4c1c9 by Simon Peyton Jones at 2024-07-27T09:45:44-04:00 Fix nasty bug in occurrence analyser As #25096 showed, the occurrence analyser was getting one-shot info flat out wrong. This commit does two things: * It fixes the bug and actually makes the code a bit tidier too. The work is done in the new function GHC.Core.Opt.OccurAnal.mkRhsOccEnv, especially the bit that prepares the `occ_one_shots` for the RHS. See Note [The OccEnv for a right hand side] * When floating out a binding we must be conservative about one-shot info. But we were zapping the entire demand info, whereas we only really need zap the /top level/ cardinality. See Note [Floatifying demand info when floating] in GHC.Core.Opt.SetLevels For some reason there is a 2.2% improvement in compile-time allocation for CoOpt_Read. Otherwise nickels and dimes. Metric Decrease: CoOpt_Read - - - - - 646ee207 by Torsten Schmits at 2024-07-27T09:46:20-04:00 add missing cell in flavours table - - - - - ec2eafdb by Ben Gamari at 2024-07-28T20:51:12+02:00 users-guide: Drop mention of dead __PARALLEL_HASKELL__ macro This has not existed for over a decade. - - - - - e2f2a56e by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Add tests for 25081 - - - - - 23f50640 by Arnaud Spiwack at 2024-07-28T22:21:07-04:00 Scale multiplicity in list comprehension Fixes #25081 - - - - - d2648289 by romes at 2024-07-30T01:38:12-04:00 TTG HsCmdArrForm: use Fixity via extension point Also migrate Fixity from GHC.Hs to Language.Haskell.Syntax since it no longer uses any GHC-specific data types. Fixed arrow desugaring bug. (This was dead code before.) Remove mkOpFormRn, it is also dead code, only used in the arrow desugaring now removed. Co-authored-by: Fabian Kirchner <kirchner at posteo.de> Co-authored-by: Alan Zimmerman <alan.zimm at gmail.com> - - - - - e258ad54 by Matthew Pickering at 2024-07-30T01:38:48-04:00 ghcup-metadata: More metadata fixes * Incorrect version range on the alpine bindists * Missing underscore in "unknown_versioning" Fixes #25119 - - - - - 72b54c07 by Rodrigo Mesquita at 2024-08-01T00:47:29-04:00 Deriving-via one-shot strict state Monad instances A small refactor to use deriving via GHC.Utils.Monad.State.Strict Monad instances for state Monads with unboxed/strict results which all re-implemented the one-shot trick in the instance and used unboxed tuples: * CmmOptM in GHC.Cmm.GenericOpt * RegM in GHC.CmmToAsm.Reg.Linear.State * UniqSM in GHC.Types.Unique.Supply - - - - - bfe4b3d3 by doyougnu at 2024-08-01T00:48:06-04:00 Rts linker: add case for pc-rel 64 relocation part of the upstream haskell.nix patches - - - - - 5843c7e3 by doyougnu at 2024-08-01T00:48:42-04:00 RTS linker: aarch64: better debug information Dump better debugging information when a symbol address is null. Part of the haskell.nix patches upstream project Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - c2e9c581 by Rodrigo Mesquita at 2024-08-01T00:49:18-04:00 base: Add haddocks to HasExceptionContext Fixes #25091 - - - - - f954f428 by Sylvain Henry at 2024-08-01T00:49:59-04:00 Only lookup ghcversion.h file in the RTS include-dirs by default. The code was introduced in 3549c952b535803270872adaf87262f2df0295a4. It used `getPackageIncludePath` which name doesn't convey that it looks into all include paths of the preload units too. So this behavior is probably unintentional and it should be ok to change it. Fix #25106 - - - - - 951ce3d5 by Matthew Pickering at 2024-08-01T00:50:35-04:00 driver: Fix -Wmissing-home-modules when multiple units have the same module name It was assumed that module names were unique but that isn't true with multiple units. The fix is quite simple, maintain a set of `(ModuleName, UnitId)` and query that to see whether the module has been specified. Fixes #25122 - - - - - bae1fea4 by sheaf at 2024-08-01T00:51:15-04:00 PMC: suggest in-scope COMPLETE sets when possible This commit modifies GHC.HsToCore.Pmc.Solver.generateInhabitingPatterns to prioritise reporting COMPLETE sets in which all of the ConLikes are in scope. This avoids suggesting out of scope constructors when displaying an incomplete pattern match warning, e.g. in baz :: Ordering -> Int baz = \case EQ -> 5 we prefer: Patterns of type 'Ordering' not matched: LT GT over: Patterns of type 'Ordering' not matched: OutOfScope Fixes #25115 - - - - - ff158fcd by Tommy Bidne at 2024-08-02T01:14:32+12:00 Print exception metadata in default handler CLC proposals 231 and 261: - Add exception type metadata to SomeException's displayException. - Add "Exception" header to default exception handler. See: https://github.com/haskell/core-libraries-committee/issues/231 https://github.com/haskell/core-libraries-committee/issues/261 Update stm submodule for test fixes. - - - - - 8b2f70a2 by Andrei Borzenkov at 2024-08-01T23:00:46-04:00 Type syntax in expressions (#24159, #24572, #24226) This patch extends the grammar of expressions with syntax that is typically found only in types: * function types (a -> b), (a ->. b), (a %m -> b) * constrained types (ctx => t) * forall-quantification (forall tvs. t) The new forms are guarded behind the RequiredTypeArguments extension, as specified in GHC Proposal #281. Examples: {-# LANGUAGE RequiredTypeArguments #-} e1 = f (Int -> String) -- function type e2 = f (Int %1 -> String) -- linear function type e3 = f (forall a. Bounded a => a) -- forall type, constraint The GHC AST and the TH AST have been extended as follows: syntax | HsExpr | TH.Exp ---------------+----------+-------------- a -> b | HsFunArr | ConE (->) a %m -> b | HsFunArr | ConE FUN ctx => t | HsQual | ConstrainedE forall a. t | HsForAll | ForallE forall a -> t | HsForAll | ForallVisE Additionally, a new warning flag -Wview-pattern-signatures has been introduced to aid with migration to the new precedence of (e -> p :: t). Co-authored-by: Vladislav Zavialov <vlad.z.4096 at gmail.com> - - - - - 66e7f57d by Brandon Chinn at 2024-08-01T21:50:58-07:00 Implement MultilineStrings (#24390) This commit adds support for multiline strings, proposed at https://github.com/ghc-proposals/ghc-proposals/pull/569. Multiline strings can now be written as: myString = """ this is a multiline string """ The multiline string will have leading indentation stripped away. Full details of this post-processing may be found at the new GHC.Parser.String module. In order to cleanly implement this and maximize reusability, I broke out the lexing logic for strings out of Lexer.x into a new GHC.Parser.String module, which lexes strings with any provided "get next character" function. This also gave us the opportunity to clean up this logic, and even optimize it a bit. With this change, parsing string literals now takes 25% less time and 25% less space. - - - - - cf47b96f by Rodrigo Mesquita at 2024-08-03T05:59:40-04:00 hi: Stable sort avails Sorting the Avails in DocStructures is required to produce fully deterministic interface files in presence of re-exported modules. Fixes #25104 - - - - - af2ae742 by M. Taimoor Zaeem at 2024-08-03T18:52:50+05:00 haddock: decrease margin on top of small headings - - - - - a1e42e7a by Rodrigo Mesquita at 2024-08-05T21:03:04-04:00 hi: Deterministic ImportedMods in Usages The `mi_usages` field of the interface files must use a deterministic list of `Usage`s to guarantee a deterministic interface. However, this list was, in its origins, constructed from a `ModuleEnv` which uses a non-deterministic ordering that was leaking into the interface. Specifically, ImportedMods = ModuleEnv ... would get converted to a list and then passed to `mkUsageInfo` to construct the Usages. The solution is simple. Back `ImportedMods` with a deterministic map. `Map Module ...` is enough, since the Ord instance for `Module` already uses a stable, deterministic, comparison. Fixes #25131 - - - - - eb1cb536 by Serge S. Gulin at 2024-08-06T08:54:55+00:00 testsuite: extend size performance tests with gzip (fixes #25046) The main purpose is to create tests for minimal app (hello world and its variations, i.e. unicode used) distribution size metric. Many platforms support distribution in compressed form via gzip. It would be nice to collect information on how much size is taken by the executional bundle for each platform at minimal edge case. 2 groups of tests are added: 1. We extend javascript backend size tests with gzip-enabled versions for all cases where an optimizing compiler is used (for now it is google closure compiler). 2. We add trivial hello world tests with gzip-enabled versions for all other platforms at CI pipeline where no external optimizing compiler is used. - - - - - d94410f8 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: @since for backtraceDesired Fixes point 1 in #25052 - - - - - bfe600f5 by Rodrigo Mesquita at 2024-08-07T11:49:19-04:00 ghc-internal: No trailing whitespace in exceptions Fixes #25052 - - - - - 62650d9f by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Add since annotation for -fkeep-auto-rules. This partially addresses #25082. - - - - - 5f0e23fd by Andreas Klebinger at 2024-08-07T11:49:54-04:00 Mention `-fkeep-auto-rules` in release notes. It was added earlier but hadn't appeared in any release notes yet. Partially addresses #25082. - - - - - 7446a09a by Sylvain Henry at 2024-08-07T11:50:35-04:00 Cmm: don't perform unsound optimizations on 32-bit compiler hosts - beef61351b240967b49169d27a9a19565cf3c4af enabled the use of MO_Add/MO_Sub for 64-bit operations in the C and LLVM backends - 6755d833af8c21bbad6585144b10e20ac4a0a1ab did the same for the x86 NCG backend However we store some literal values as `Int` in the compiler. As a result, some Cmm optimizations transformed target 64-bit literals into compiler `Int`. If the compiler is 32-bit, this leads to computing with wrong literals (see #24893 and #24700). This patch disables these Cmm optimizations for 32-bit compilers. This is unsatisfying (optimizations shouldn't be compiler-word-size dependent) but it fixes the bug and it makes the patch easy to backport. A proper fix would be much more invasive but it shall be implemented in the future. Co-authored-by: amesgen <amesgen at amesgen.de> - - - - - d59faaf2 by Vladislav Zavialov at 2024-08-07T11:51:11-04:00 docs: Update info on RequiredTypeArguments Add a section on "types in terms" that were implemented in 8b2f70a202 and remove the now outdated suggestion of using `type` for them. - - - - - 39fd6714 by Sylvain Henry at 2024-08-07T11:51:52-04:00 JS: fix minor typo in base's jsbits - - - - - e7764575 by Sylvain Henry at 2024-08-07T11:51:52-04:00 RTS: remove hack to force old cabal to build a library with only JS sources Need to extend JSC externs with Emscripten RTS definitions to avoid JSC_UNDEFINED_VARIABLE errors when linking without the emcc rts. Fix #25138 Some recompilation avoidance tests now fail. This is tracked with the other instances of this failure in #23013. My hunch is that they were working by chance when we used the emcc linker. Metric Decrease: T24602_perf_size - - - - - d1a40233 by Brandon Chinn at 2024-08-07T11:53:08-04:00 Support multiline strings in type literals (#25132) - - - - - 610840eb by Sylvain Henry at 2024-08-07T11:53:50-04:00 JS: fix callback documentation (#24377) Fix #24377 - - - - - 6ae4b76a by Zubin Duggal at 2024-08-13T13:36:57-04:00 haddock: Build haddock-api and haddock-library using hadrian We build these two packages as regular boot library dependencies rather than using the `in-ghc-tree` flag to include the source files into the haddock executable. The `in-ghc-tree` flag is moved into haddock-api to ensure that haddock built from hackage can still find the location of the GHC bindist using `ghc-paths`. Addresses #24834 This causes a metric decrease under non-release flavours because under these flavours libraries are compiled with optimisation but executables are not. Since we move the bulk of the code from the haddock executable to the haddock-api library, we see a metric decrease on the validate flavours. Metric Decrease: haddock.Cabal haddock.base haddock.compiler - - - - - 51ffba5d by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add an extension field to HsRecFields This is the Right Thing to Do™. And it prepares for storing a multiplicity coercion there. First step of the plan outlined here and below https://gitlab.haskell.org/ghc/ghc/-/merge_requests/12947#note_573091 - - - - - 4d2faeeb by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Add test for #24961 - - - - - 623b4337 by Arnaud Spiwack at 2024-08-13T13:37:50-04:00 Ensures that omitted record fields in pattern have multiplicity Many Omitted fields were simply ignored in the type checker and produced incorrect Core code. Fixes #24961 Metric Increase: RecordUpdPerf - - - - - c749bdfd by Sylvain Henry at 2024-08-13T13:38:41-04:00 AARCH64 linker: skip NONE relocations This patch is part of the patches upstreamed from haskell.nix. See https://github.com/input-output-hk/haskell.nix/pull/1960 for the original report/patch. - - - - - 682a6a41 by Brandon Chinn at 2024-08-13T13:39:17-04:00 Support multiline strings in TH - - - - - ee0a9c18 by Matthew Pickering at 2024-08-14T14:27:39-04:00 Extend -reexported-module flag to support module renaming The -reexported-module flag now supports renaming -rexported-modules. ``` -rexported-module "A as B" ``` This feature is only relevant to multi-component sessions. Fixes #25139 - - - - - e9496000 by Arnaud Spiwack at 2024-08-14T14:28:20-04:00 Don't restrict eta-reduction of linear functions This commit simply removes code. All the supporting implementation has been done as part of !12883. Closes #25129 - - - - - 2bb4156e by sheaf at 2024-08-14T14:28:56-04:00 Allow @ character in C labels Generated symbol names can include the '@' character, for example when using `__attribute__((vectorcall))`. - - - - - 7602ca23 by Sylvain Henry at 2024-08-14T14:29:36-04:00 Linker: replace blind tuple with a datatype + docs - - - - - bdd77b9e by sheaf at 2024-08-16T12:47:11-04:00 isIrrefutableHsPat: look up ConLikes in the HscEnv At GhcRn stage, in isIrrefutableHsPat we only looked up data constructors in the RdrEnv, which meant that we lacked fallibility information for out-of-scope constructors (which can arise from Template Haskell splices). Instead, we use 'lookupGREInfo', which looks up the information in the type environment. This was the correct function to call all along, but was not used in 572fbc44 due to import cycle reasons. The appropriate functions, 'irrefutableConLike{Rn,Tc}' have been moved to 'GHC.Rename.Env', which avoids import cycles. Fixes #25164 - - - - - 4bee377c by Sylvain Henry at 2024-08-16T12:47:53-04:00 Linker: some refactoring to prepare for #24886 - Rename LoadedBCOs into LazyBCOs - Bundle SptEntries with CompiledByteCode and removed [SptEntry] field from the BCOs constructor - Rename Linkable's LM constructor into Linkable: in the past we had LM and LP for Module and Package, now we only have the former. - Rename Unlinked into LinkablePart (and linkableUnlinked into linkableParts) - Use NonEmpty to encode invariant in Linkable's linkableParts type - Add helpers: linkableLibs, linkableBCOs, etc. - Add documentation - Remove partial nameOfObject - Rename nameOfObject_maybe into linkablePartPath - Rename byteCodeOfObject into linkablePartAllBCOs. - Refactor linkablePartAllBCOs to avoid a panic if a LazyBCO has a C stub. Document the fact that LazyBCOs are returned in this case (contrary to linkableBCOs which only returns non-lazy ones) Refactoring done while trying to understand how to adapt the linker code to support the JS backend too (cf #24886). - - - - - fa0dbaca by Mario Blažević at 2024-08-17T03:31:32+00:00 Implements the Exportable Named Default proposal (#24305) This squashed commit adds support for exportable named defaults, the accepted GHC proposal at https://github.com/ghc-proposals/ghc-proposals/pull/409 The proposal extends the Haskell '98 declarations default (Int, Double) which were implicitly always applying to Num class alone, to allow specifying an arbitrary single-parameter class: default IsString (Text, String) The effect of this declaration would be to eliminate the ambiguous type errors around string literals when OverloadedStrings extension is active. The declaration by itself has effect only in its module, so the proposal also adds the ability to export class defaults: module MyModule (default IsIstring) Once the language extension is published and established, we can consider using it in base and other libraries. See Note [Named default declarations] in GHC.Tc.Gen.Default for implementation details. - - - - - 1deba6b2 by Simon Peyton Jones at 2024-08-17T13:58:13-04:00 Make kick-out more selective This MR revised the crucial kick-out criteria in the constraint solver. Ticket #24984 showed an example in which * We were kicking out unnecessarily * That gave rise to extra work, of course * But it /also/ led to exponentially-sized coercions due to lack of sharing in coercions (something we want to fix separately #20264) This MR sharpens up the kick-out criteria; specifially in (KK2) we look only under type family applications if (fs>=fw). This forced me to understand the existing kick-out story, and I ended up rewriting many of the careful Notes in GHC.Tc.Solver.InertSet. Especially look at the new `Note [The KickOut Criteria]` The proof of termination is not air-tight, but it is better than before, and both Richard and I think it's correct :-). - - - - - 88488847 by Cheng Shao at 2024-08-18T04:44:01+02:00 testsuite: remove undesired -fasm flag from test ways This patch removes the -fasm flag from test ways, except ways like optasm that explicitly state they are meant to be compiled with NCG backend. Most test ways should use the default codegen backend, and the precense of -fasm can cause stderr mismatches like this when GHC is configured with the unregisterised backend: ``` --- /dev/null +++ /tmp/ghctest-3hydwldj/test spaces/testsuite/tests/profiling/should_compile/prof-late-cc.run/prof-late-cc.comp.stderr.normalised @@ -0,0 +1,2 @@ +when making flags consistent: warning: [GHC-74335] [-Winconsistent-flags (in -Wdefault)] + Target platform uses unregisterised ABI, so compiling via C *** unexpected failure for prof-late-cc(prof_no_auto) ``` This has been breaking the wasm unreg nightly job since !12595 landed. - - - - - 3a145315 by Cheng Shao at 2024-08-18T13:05:45-04:00 ghci: fix isMinTTY.h casing for Windows targets This commit fixes isMinTTY.h casing in isMinTTY.c that's compiled for Windows targets. While this looks harmless given Windows filesystems are case-insensitive by default, it does cause a compilation warning with recent versions of clang, so we might as well fix the casing: ``` driver\ghci\isMinTTY.c:10:10: error: warning: non-portable path to file '"isMinTTY.h"'; specified path differs in case from file name on disk [-Wnonportable-include-path] | 10 | #include "isMINTTY.h" | ^ #include "isMINTTY.h" ^~~~~~~~~~~~ "isMinTTY.h" 1 warning generated. ``` - - - - - 5f972bfb by Zubin Duggal at 2024-08-21T03:18:15-04:00 compiler: Fix pretty printing of ticked prefix constructors (#24237) - - - - - ef0a08e7 by Mike Pilgrem at 2024-08-21T03:18:57-04:00 Fix #15773 Clarify further -rtsopts 'defaults' in docs - - - - - 05a4be58 by Sebastian Graf at 2024-08-21T03:19:33-04:00 Improve efficiency of `assertError` (#24625) ... by moving `lazy` to the exception-throwing branch. It's all documented in `Note [Strictness of assertError]`. - - - - - c29b2b5a by sheaf at 2024-08-21T13:11:30-04:00 GHCi debugger: drop record name spaces for Ids When binding new local variables at a breakpoint, we should create Ids with variable namespace, and not record field namespace. Otherwise the rest of the compiler falls over because the IdDetails are wrong. Fixes #25109 - - - - - bd82ac9f by Hécate Kleidukos at 2024-08-21T13:12:12-04:00 base: Final deprecation of GHC.Pack The timeline mandated by #21461 has come to its term and after two years and four minor releases, we are finally removing GHC.Pack from base. Closes #21536 - - - - - 5092dbff by Sylvain Henry at 2024-08-21T13:12:54-04:00 JS: support rubbish static literals (#25177) Support for rubbish dynamic literals was added in #24664. This patch does the same for static literals. Fix #25177 - - - - - b5a2c061 by Phil de Joux at 2024-08-21T13:13:33-04:00 haddock docs: prefix comes before, postfix comes after - - - - - 6fde3685 by Marcin Szamotulski at 2024-08-21T23:15:39-04:00 haddock: include package info with --show-interface - - - - - 7e02111b by Andreas Klebinger at 2024-08-21T23:16:15-04:00 Document the (x86) SIMD macros. Fixes #25021. - - - - - 05116c83 by Rodrigo Mesquita at 2024-08-22T10:37:44-04:00 ghc-internal: Derive version from ghc's version Fixes #25005 - - - - - 73f5897d by Ben Gamari at 2024-08-22T10:37:44-04:00 base: Deprecate GHC.Desugar See https://github.com/haskell/core-libraries-committee/issues/216. This will be removed in GHC 9.14. - - - - - 821d0a9a by Cheng Shao at 2024-08-22T10:38:22-04:00 compiler: Store ForeignStubs and foreign C files in interfaces This data is used alongside Core bindings to reconstruct intermediate build products when linking Template Haskell splices with bytecode. Since foreign stubs and files are generated in the pipeline, they were lost with only Core bindings stored in interfaces. The interface codec type `IfaceForeign` contains a simplified representation of `ForeignStubs` and the set of foreign sources that were manually added by the user. When the backend phase writes an interface, `mkFullIface` calls `encodeIfaceForeign` to read foreign source file contents and assemble `IfaceForeign`. After the recompilation status check of an upstream module, `initWholeCoreBindings` calls `decodeIfaceForeign` to restore `ForeignStubs` and write the contents of foreign sources to the file system as temporary files. The restored foreign inputs are then processed by `hscInteractive` in the same manner as in a regular pipeline. When linking the stub objects for splices, they are excluded from suffix adjustment for the interpreter way through a new flag in `Unlinked`. For details about these processes, please consult Note [Foreign stubs and TH bytecode linking]. Metric Decrease: T13701 - - - - - f0408eeb by Cheng Shao at 2024-08-23T10:37:10-04:00 git: remove a.out and include it in .gitignore a.out is a configure script byproduct. It was mistakenly checked into the tree in !13118. This patch removes it, and include it in .gitignore to prevent a similar error in the future. - - - - - 1f95c5e4 by Matthew Pickering at 2024-08-23T10:37:46-04:00 docs: Fix code-block syntax on old sphinx version This code-block directive breaks the deb9 sphinx build. Fixes #25201 - - - - - 27dceb42 by Sylvain Henry at 2024-08-26T11:05:11-04:00 JS: add basic support for POSIX *at functions (#25190) openat/fstatat/unlinkat/dup are now used in the recent release of the `directory` and `file-io` packages. As such, these functions are (indirectly) used in the following tests one we'll bump the `directory` submodule (see !13122): - openFile008 - jsOptimizer - T20509 - bkpcabal02 - bkpcabal03 - bkpcabal04 - - - - - c68be356 by Matthew Pickering at 2024-08-26T11:05:11-04:00 Update directory submodule to latest master The primary reason for this bump is to fix the warning from `ghc-pkg check`: ``` Warning: include-dirs: /data/home/ubuntu/.ghcup/ghc/9.6.2/lib/ghc-9.6.2/lib/../lib/aarch64-linux-ghc-9.6.2/directory-1.3.8.1/include doesn't exist or isn't a directory ``` This also requires adding the `file-io` package as a boot library (which is discussed in #25145) Fixes #23594 #25145 - - - - - 4ee094d4 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Fix aarch64-alpine target platform description We are producing bindists where the target triple is aarch64-alpine-linux when it should be aarch64-unknown-linux This is because the bootstrapped compiler originally set the target triple to `aarch64-alpine-linux` which is when propagated forwards by setting `bootstrap_target` from the bootstrap compiler target. In order to break this chain we explicitly specify build/host/target for aarch64-alpine. This requires a new configure flag `--enable-ignore-` which just switches off a validation check that the target platform of the bootstrap compiler is the same as the build platform. It is the same, but the name is just wrong. These commits can be removed when the bootstrap compiler has the correct target triple (I looked into patching this on ci-images, but it looked hard to do correctly as the build/host platform is not in the settings file). Fixes #25200 - - - - - e0e0f2b2 by Matthew Pickering at 2024-08-26T11:05:47-04:00 Bump nixpkgs commit for gen_ci script - - - - - 63a27091 by doyougnu at 2024-08-26T20:39:30-04:00 rts: win32: emit additional debugging information -- migration from haskell.nix - - - - - aaab3d10 by Vladislav Zavialov at 2024-08-26T20:40:06-04:00 Only export defaults when NamedDefaults are enabled (#25206) This is a reinterpretation of GHC Proposal #409 that avoids a breaking change introduced in fa0dbaca6c "Implements the Exportable Named Default proposal" Consider a module M that has no explicit export list: module M where default (Rational) Should it export the default (Rational)? The proposal says "yes", and there's a test case for that: default/DefaultImport04.hs However, as it turns out, this change in behavior breaks existing programs, e.g. the colour-2.3.6 package can no longer be compiled, as reported in #25206. In this patch, we make implicit exports of defaults conditional on the NamedDefaults extension. This fix is unintrusive and compliant with the existing proposal text (i.e. it does not require a proposal amendment). Should the proposal be amended, we can go for a simpler solution, such as requiring all defaults to be exported explicitly. Test case: testsuite/tests/default/T25206.hs - - - - - 3a5bebf8 by Matthew Pickering at 2024-08-28T14:16:42-04:00 simplifier: Fix space leak during demand analysis The lazy structure (a list) in a strict field in `DmdType` is not fully forced which leads to a very large thunk build-up. It seems there is likely still more work to be done here as it seems we may be trading space usage for work done. For now, this is the right choice as rather than using all the memory on my computer, compilation just takes a little bit longer. See #25196 - - - - - c2525e9e by Ryan Scott at 2024-08-28T14:17:17-04:00 Add missing parenthesizeHsType in cvtp's InvisP case We need to ensure that when we convert an `InvisP` (invisible type pattern) to a `Pat`, we parenthesize it (at precedence `appPrec`) so that patterns such as `@(a :: k)` will parse correctly when roundtripped back through the parser. Fixes #25209. - - - - - 1499764f by Sjoerd Visscher at 2024-08-29T16:52:56+02:00 Haddock: Add no-compilation flag This flag makes sure to avoid recompilation of the code when generating documentation by only reading the .hi and .hie files, and throw an error if it can't find them. - - - - - 768fe644 by Andreas Klebinger at 2024-09-03T13:15:20-04:00 Add functions to check for weakly pinned arrays. This commit adds `isByteArrayWeaklyPinned#` and `isMutableByteArrayWeaklyPinned#` primops. These check if a bytearray is *weakly* pinned. Which means it can still be explicitly moved by the user via compaction but won't be moved by the RTS. This moves us one more stop closer to nailing down #22255. - - - - - b16605e7 by Arsen Arsenović at 2024-09-03T13:16:05-04:00 ghc-toolchain: Don't leave stranded a.outs when testing for -g0 This happened because, when ghc-toolchain tests for -g0, it does so by compiling an empty program. This compilation creates an a.out. Since we create a temporary directory, lets place the test program compilation in it also, so that it gets cleaned up. Fixes: 25b0b40467d0a12601497117c0ad14e1fcab0b74 Closes: https://gitlab.haskell.org/ghc/ghc/-/issues/25203 - - - - - 83e70b14 by Torsten Schmits at 2024-09-03T13:16:41-04:00 Build foreign objects for TH with interpreter's way when loading from iface Fixes #25211 When linking bytecode for TH from interface core bindings with `-fprefer-byte-code`, foreign sources are loaded from the interface as well and compiled to object code in an ad-hoc manner. The results are then loaded by the interpreter, whose way may differ from the current build's target way. This patch ensures that foreign objects are compiled with the interpreter's way. - - - - - 0d3bc2fa by Cheng Shao at 2024-09-04T07:20:06-04:00 rts: fix checkClosure error message This patch fixes an error message in checkClosure() when the closure has already been evacuated. The previous logic was meant to print the evacuated closure's type in the error message, but it was completely wrong, given info was not really an info table, but a tagged pointer that points to the closure's new address. - - - - - fb0a4e5c by Sven Tennie at 2024-09-04T07:20:43-04:00 MO_AcquireFence: Less restrictive barrier GCC and CLang translate the built-in `atomic_thread_fence(memory_order_acquire)` to `dmb ishld`, which is a bit less restrictive than `dmb ish` (which also implies stores.) - - - - - a45f1488 by Fendor at 2024-09-04T20:22:00-04:00 testsuite: Add support to capture performance metrics via 'perf' Performance metrics collected via 'perf' can be more accurate for run-time performance than GHC's rts, due to the usage of hardware counters. We allow performance tests to also record PMU events according to 'perf list'. - - - - - ce61fca5 by Fendor at 2024-09-04T20:22:00-04:00 gitlab-ci: Add nightly job for running the testsuite with perf profiling support - - - - - 6dfb9471 by Fendor at 2024-09-04T20:22:00-04:00 Enable perf profiling for compiler performance tests - - - - - da306610 by sheaf at 2024-09-04T20:22:41-04:00 RecordCon lookup: don't allow a TyCon This commit adds extra logic when looking up a record constructor. If GHC.Rename.Env.lookupOccRnConstr returns a TyCon (as it may, due to the logic explained in Note [Pattern to type (P2T) conversion]), we emit an error saying that the data constructor is not in scope. This avoids the compiler falling over shortly thereafter, in the call to 'lookupConstructorInfo' inside 'GHC.Rename.Env.lookupRecFieldOcc', because the record constructor would not have been a ConLike. Fixes #25056 - - - - - 9c354beb by Matthew Pickering at 2024-09-04T20:23:16-04:00 Use deterministic names for temporary files When there are multiple threads they can race to create a temporary file, in some situations the thread will create ghc_1.c and in some it will create ghc_2.c. This filename ends up in the debug info for object files after compiling a C file, therefore contributes to object nondeterminism. In order to fix this we store a prefix in `TmpFs` which serves to namespace temporary files. The prefix is populated from the counter in TmpFs when the TmpFs is forked. Therefore the TmpFs must be forked outside the thread which consumes it, in a deterministic order, so each thread always receives a TmpFs with the same prefix. This assumes that after the initial TmpFs is created, all other TmpFs are created from forking the original TmpFs. Which should have been try anyway as otherwise there would be file collisions and non-determinism. Fixes #25224 - - - - - 59906975 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 Silence x-partial in Haddock.Backends.Xhtml This is an unfortunate consequence of two mechanisms: * GHC provides (possibly-empty) lists of names * The functions that retrieve those names are not equipped to do error reporting, and thus accept these lists at face value. They will have to be attached an effect for error reporting in a later refactoring - - - - - 8afbab62 by Hécate Kleidukos at 2024-09-05T10:57:15-04:00 hadrian: Support loading haddock in ghci There is one tricky aspect with wired-in packages where the boot package is built with `-this-unit-id ghc` but the dependency is reported as `-package-id ghc-9.6...`. This has never been fixed in GHC as the situation of loading wired-in packages into the multi-repl seems like quite a niche feature that is always just easier to workaround. - - - - - 6cac9eb8 by Matthew Pickering at 2024-09-05T10:57:15-04:00 hadrian/multi: Load all targets when ./hadrian/ghci-multi is called This seems to make a bit more sense than just loading `ghc` component (and dependencies). - - - - - 7d84df86 by Matthew Pickering at 2024-09-05T10:57:51-04:00 ci: Beef up determinism interface test There have recently been some determinism issues with the simplifier and documentation. We enable more things to test in the ABI test to check that we produce interface files deterministically. - - - - - 5456e02e by Sylvain Henry at 2024-09-06T11:57:01+02:00 Transform some StgRhsClosure into StgRhsCon after unarisation (#25166) Before unarisation we may have code like: Test.foo :: Test.D [GblId, Unf=OtherCon []] = \u [] case (# |_| #) [GHC.Types.(##)] of sat_sAw [Occ=Once1] { __DEFAULT -> Test.D [GHC.Types.True sat_sAw]; }; After unarisation we get: Test.foo :: Test.D [GblId, Unf=OtherCon []] = {} \u [] Test.D [GHC.Types.True 2#]; Notice that it's still an Updatable closure for no reason anymore. This patch transforms appropriate StgRhsClosures into StgRhsCons after unarisation, allowing these closures to be statically allocated. Now we get the expected: Test.foo :: Test.D [GblId, Unf=OtherCon []] = Test.D! [GHC.Types.True 2#]; Fix #25166 To avoid duplicating code, this patch refactors the mk(Top)StgRhs functions and put them in a GHC.Stg.Make module alongside the new mk(Top)StgRhsCon_maybe functions. - - - - - 958b4518 by Hécate Kleidukos at 2024-09-06T16:40:56-04:00 haddock: Add missing requirements.txt for the online manual - - - - - 573f9833 by Sven Tennie at 2024-09-08T09:58:21+00:00 AArch64: Implement takeRegRegMoveInstr This has likely been forgotten. - - - - - 20b0de7d by Hécate Kleidukos at 2024-09-08T14:19:28-04:00 haddock: Configuration fix for ReadTheDocs - - - - - 03055c71 by Sylvain Henry at 2024-09-09T14:58:15-04:00 JS: fake support for native adjustors (#25159) The JS backend doesn't support adjustors (I believe) and in any case if it ever supports them it will be a native support, not one via libffi. - - - - - 5bf0e6bc by Sylvain Henry at 2024-09-09T14:58:56-04:00 JS: remove redundant h$lstat It was introduced a second time by mistake in 27dceb42376c34b99a38e36a33b2abc346ed390f (cf #25190) - - - - - ffbc2ab0 by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Refactor only newSysLocalDs * Change newSysLocalDs to take a scaled type * Add newSysLocalMDs that takes a type and makes a ManyTy local Lots of files touched, nothing deep. - - - - - 7124e4ad by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Don't introduce 'nospec' on the LHS of a RULE This patch address #25160. The main payload is: * When desugaring the LHS of a RULE, do not introduce the `nospec` call for non-canonical evidence. See GHC.Core.InstEnv Note [Coherence and specialisation: overview] The `nospec` call usually introdued in `dsHsWrapper`, but we don't want it on the LHS of a RULE (that's what caused #25160). So now `dsHsWrapper` takes a flag to say if it's on the LHS of a RULE. See wrinkle (NC1) in `Note [Desugaring non-canonical evidence]` in GHC.HsToCore.Binds. But I think this flag will go away again when I have finished with my (entirely separate) speciaise-on-values patch (#24359). All this meant I had to re-understand the `nospec` stuff and coherence, and that in turn made me do some refactoring, and add a lot of new documentation The big change is that in GHC.Core.InstEnv, I changed the /type synonym/ `Canonical` into a /data type/ `CanonicalEvidence` and documented it a lot better. That in turn made me realise that CalLStacks were being treated with a bit of a hack, which I documented in `Note [CallStack and ExecptionContext hack]`. - - - - - 663daf8d by Simon Peyton Jones at 2024-09-10T00:40:37-04:00 Add defaulting of equalities This MR adds one new defaulting strategy to the top-level defaulting story: see Note [Defaulting equalities] in GHC.Tc.Solver. This resolves #25029 and #25125, which showed that users were accidentally relying on a GHC bug, which was fixed by commit 04f5bb85c8109843b9ac2af2a3e26544d05e02f4 Author: Simon Peyton Jones <simon.peytonjones at gmail.com> Date: Wed Jun 12 17:44:59 2024 +0100 Fix untouchability test This MR fixes #24938. The underlying problem was tha the test for "does this implication bring in scope any equalities" was plain wrong. This fix gave rise to a number of user complaints; but the improved defaulting story of this MR largely resolves them. On the way I did a bit of refactoring, of course * Completely restructure the extremely messy top-level defaulting code. The new code is in GHC.Tc.Solver.tryDefaulting, and is much, much, much esaier to grok. - - - - - e28cd021 by Andrzej Rybczak at 2024-09-10T00:41:18-04:00 Don't name a binding pattern It's a keyword when PatternSynonyms are set. - - - - - b09571e2 by Simon Peyton Jones at 2024-09-10T00:41:54-04:00 Do not use an error thunk for an absent dictionary In worker/wrapper we were using an error thunk for an absent dictionary, but that works very badly for -XDictsStrict, or even (as #24934 showed) in some complicated cases involving strictness analysis and unfoldings. This MR just uses RubbishLit for dictionaries. Simple. No test case, sadly because our only repro case is rather complicated. - - - - - 8bc9f5f6 by Hécate Kleidukos at 2024-09-10T00:42:34-04:00 haddock: Remove support for applehelp format in the Manual - - - - - 9ca15506 by doyougnu at 2024-09-10T10:46:38-04:00 RTS linker: add support for hidden symbols (#25191) Add linker support for hidden symbols. We basically treat them as weak symbols. Patch upstreamed from haskell.nix Co-authored-by: Sylvain Henry <sylvain at haskus.fr> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3b2dc826 by Sven Tennie at 2024-09-10T10:47:14-04:00 Fix C warnings (#25237) GCC 14 treats the fixed warnings as errors by default. I.e. we're gaining GCC 14 compatibility with these fixes. - - - - - 05715994 by Sylvain Henry at 2024-09-10T10:47:55-04:00 JS: fix codegen of static string data Before this patch, when string literals are made trivial, we would generate `h$("foo")` instead of `h$str("foo")`. This was introduced by mistake in 6bd850e887b82c5a28bdacf5870d3dc2fc0f5091. - - - - - 949ebced by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Re-organise cross-OS compatibility layer - - - - - 84ac9a99 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Remove CPP for obsolete GHC and Cabal versions - - - - - 370d1599 by Hécate Kleidukos at 2024-09-10T19:19:40-04:00 haddock: Move the changelog file to the 'extra-doc-files' section in the cabal file - - - - - cfbff65a by Simon Peyton Jones at 2024-09-10T19:20:16-04:00 Add ZonkAny and document it This MR fixed #24817 by adding ZonkAny, which takes a Nat argument. See Note [Any types] in GHC.Builtin.Types, especially wrinkle (Any4). - - - - - 0167e472 by Matthew Pickering at 2024-09-11T02:41:42-04:00 hadrian: Make sure ffi headers are built before using a compiler When we are using ffi adjustors then we rely on `ffi.h` and `ffitarget.h` files during code generation when compiling stubs. Therefore we need to add this dependency to the build system (which this patch does). Reproducer, configure with `--enable-libffi-adjustors` and then build "_build/stage1/libraries/ghc-prim/build/GHC/Types.p_o". Observe that this fails before this patch and works afterwards. Fixes #24864 Co-authored-by: Sylvain Henry <sylvain at haskus.fr> - - - - - 0f696958 by Rodrigo Mesquita at 2024-09-11T02:42:18-04:00 base: Deprecate BCO primops exports from GHC.Exts See https://github.com/haskell/core-libraries-committee/issues/212. These reexports will be removed in GHC 9.14. - - - - - cf0e7729 by Alan Zimmerman at 2024-09-11T02:42:54-04:00 EPA: Remove Anchor = EpaLocation synonym This just causes confusion. - - - - - 8e462f4d by Andrew Lelechenko at 2024-09-11T22:20:37-04:00 Bump submodule deepseq to 1.5.1.0 - - - - - aa4500ae by Sebastian Graf at 2024-09-11T22:21:13-04:00 User's guide: Fix the "no-backtracking" example of -XOrPatterns (#25250) Fixes #25250. - - - - - 1c479c01 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add Native Code Generator (NCG) This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 51b678e1 by Sven Tennie at 2024-09-12T10:39:38+00:00 Adjust test timings for slower computers Increase the delays a bit to be able to run these tests on slower computers. The reference was a Lichee Pi 4a RISCV64 machine. - - - - - a0e41741 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add RTS linker This architecture wasn't supported before. Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - d365b1d4 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Ignore divbyzero test The architecture's behaviour differs from the test's expectations. See comment in code why this is okay. - - - - - abf3d699 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Enable MulMayOflo_full test It works and thus can be tested. - - - - - 38c7ea8c by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: LibffiAdjustor: Ensure code caches are flushed RISCV64 needs a specific code flushing sequence (involving fence.i) when new code is created/loaded. - - - - - 7edc6965 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add additional linker symbols for builtins We're relying on some GCC/Clang builtins. These need to be visible to the linker (and not be stripped away.) - - - - - 92ad3d42 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Add GHCi support As we got a RTS linker for this architecture now, we can enable GHCi for it. - - - - - a145f701 by Sven Tennie at 2024-09-12T10:39:38+00:00 RISCV64: Set codeowners of the NCG - - - - - 8e6d58cf by Sven Tennie at 2024-09-12T10:39:38+00:00 Add test for C calling convention Ensure that parameters and return values are correctly processed. A dedicated test (like this) helps to get the subtleties of calling conventions easily right. The test is failing for WASM32 and marked as fragile to not forget to investigate this (#25249). - - - - - fff55592 by Torsten Schmits at 2024-09-12T21:50:34-04:00 finder: Add `IsBootInterface` to finder cache keys - - - - - cdf530df by Alan Zimmerman at 2024-09-12T21:51:10-04:00 EPA: Sync ghc-exactprint to GHC - - - - - 1374349b by Sebastian Graf at 2024-09-13T07:52:11-04:00 DmdAnal: Fast path for `multDmdType` (#25196) This is in order to counter a regression exposed by SpecConstr. Fixes #25196. - - - - - 80769bc9 by Andrew Lelechenko at 2024-09-13T07:52:47-04:00 Bump submodule array to 0.5.8.0 - - - - - 49ac3fb8 by Sylvain Henry at 2024-09-16T10:33:01-04:00 Linker: add support for extra built-in symbols (#25155) See added Note [Extra RTS symbols] and new user guide entry. Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 3939a8bf by Samuel Thibault at 2024-09-16T10:33:44-04:00 GNU/Hurd: Add getExecutablePath support GNU/Hurd exposes it as /proc/self/exe just like on Linux. - - - - - d3b19851 by Sylvain Henry at 2024-09-17T11:03:28-04:00 RTS: expose closure_sizeW_ (#25252) C code using the closure_sizeW macro can't be linked with the RTS linker without this patch. It fails with: ghc-9.11.20240911: Failed to lookup symbol: closure_sizeW_ Fix #25252 Co-authored-by: Hamish Mackenzie <Hamish.K.Mackenzie at gmail.com> Co-authored-by: Moritz Angermann <moritz.angermann at gmail.com> - - - - - 137bf74d by Sebastian Graf at 2024-09-17T11:04:05-04:00 HsExpr: Inline `HsWrap` into `WrapExpr` This nice refactoring was suggested by Simon during review: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/13261#note_583374 Fixes #25264. - - - - - 7fd9e5e2 by Sebastian Graf at 2024-09-17T11:04:05-04:00 Pmc: Improve Desugaring of overloaded list patterns (#25257) This actually makes things simpler. Fixes #25257. - - - - - e4169ba9 by Ben Gamari at 2024-09-18T07:55:28-04:00 configure: Correctly report when subsections-via-symbols is disabled As noted in #24962, currently subsections-via-symbols is disabled on AArch64/Darwin due to alleged breakage. However, `configure` reports to the user that it is enabled. Fix this. - - - - - 9d20a787 by Mario Blažević at 2024-09-18T07:56:08-04:00 Modified the default export implementation to match the amended spec - - - - - 35eb4f42 by Sylvain Henry at 2024-09-18T07:57:00-04:00 FFI: don't ppr Id/Var symbols with debug info (#25255) Even if `-dpp-debug` is enabled we should still generate valid C code. So we disable debug info printing when rendering with Code style. - - - - - 9e96dad8 by Sebastian Graf at 2024-09-21T17:47:59-04:00 Demand: Combine examples into Note (#25107) Just a leftover from !13060. Fixes #25107. - - - - - 21aaa34b by sheaf at 2024-09-21T17:48:36-04:00 Use x86_64-unknown-windows-gnu target for LLVM on Windows - - - - - 992a7624 by sheaf at 2024-09-21T17:48:36-04:00 LLVM: use -relocation-model=pic on Windows This is necessary to avoid the segfaults reported in #22487. Fixes #22487 - - - - - c50d29be by Ryan Hendrickson at 2024-09-21T17:49:15-04:00 compiler: Use type abstractions when deriving For deriving newtype and deriving via, in order to bring type variables needed for the coercions into scope, GHC generates type signatures for derived class methods. As a simplification, drop the type signatures and instead use type abstractions to bring method type variables into scope. - - - - - f04fd0ae by Zubin Duggal at 2024-09-21T17:49:51-04:00 driver: Ensure we run driverPlugin for staticPlugins (#25217) driverPlugins are only run when the plugin state changes. This meant they were never run for static plugins, as their state never changes. We need to keep track of whether a static plugin has been initialised to ensure we run static driver plugins at least once. This necessitates an additional field in the `StaticPlugin` constructor as this state has to be bundled with the plugin itself, as static plugins have no name/identifier we can use to otherwise reference them - - - - - 620becd7 by Andreas Klebinger at 2024-09-21T17:50:27-04:00 Allow unknown fd device types for setNonBlockingMode. This allows fds with a unknown device type to have blocking mode set. This happens for example for fds from the inotify subsystem. Fixes #25199. - - - - - c76e25b3 by Hécate Kleidukos at 2024-09-21T17:51:07-04:00 Use Hackage version of Cabal 3.14.0.0 for Hadrian. We remove the vendored Cabal submodule. Also update the bootstrap plans Fixes #25086 - - - - - 6c83fd7f by Zubin Duggal at 2024-09-21T17:51:07-04:00 ci: Ensure we source ci.sh in any jobs that run commands outside of ci.sh ci.sh sets up the toolchain environment, including paths for the cabal directory, the toolchain binaries etc. If we run any commands outside of ci.sh, unless we source ci.sh we will use the wrong values for these environment variables. In particular, I ran into an issue where the cabal invocation `hadrian/ghci` was using an old index state despite `ci.sh setup` updating and setting the correct index state. This is because `ci.sh` sets the `CABAL_DIR` to a different place, which is where the index was downloaded to, but we were using the default cabal directory outside ci.sh The solution is to source the correct environment `ci.sh` using `. ci.sh setup` - - - - - 9586998d by Sven Tennie at 2024-09-21T17:51:43-04:00 ghc-toolchain: Set -fuse-ld even for ld.bfd This reflects the behaviour of the autoconf scripts. - - - - - d7016e0d by Sylvain Henry at 2024-09-21T17:52:24-04:00 Parser: be more careful when lexing extended literals (#25258) Previously we would lex invalid prefixes like "8#Int3" as [8#Int, 3]. A side-effect of this patch is that we now allow negative unsigned extended literals. They trigger an overflow warning later anyway. - - - - - ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - f3493553 by Torsten Schmits at 2024-09-30T17:11:37+02:00 Link interface bytecode from package DBs if possible Part of #25090. MR !????? When splices are executed with `-fprefer-byte-code`, the loader will compile Core bindings to bytecode if those are present in interfaces of module dependencies. So far, this only applied to local modules (i.e. home modules in make mode and non-package deps in oneshot mode). This patch extends support to interfaces loaded from a package DB. In `getLinkDeps`, the dependencies chosen for recursion were restricted to `dep_direct_mods`, which has been changed to include `mi_usages`. In order to unify treatment of the different link variants across make/oneshot mode, the algorithm's intermediate results have been abstracted into the data types `LinkDep` and `LinkModule`. At the moment, following package deps is only implemented for oneshot mode. - - - - - 25 changed files: - .gitignore - .gitlab-ci.yml - .gitlab/ci.sh - .gitlab/generate-ci/flake.lock - .gitlab/generate-ci/gen_ci.hs - .gitlab/jobs.yaml - .gitlab/rel_eng/mk-ghcup-metadata/mk_ghcup_metadata.py - .gitmodules - CODEOWNERS - compiler/CodeGen.Platform.h - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Names/TH.hs - compiler/GHC/Builtin/Types.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/GenericOpt.hs - compiler/GHC/Cmm/Graph.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/908ed174cfab250ba373ba39c37d7b077235d3fb...f34935537d538e4ce27afd1e9b01a600c2adad13 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/908ed174cfab250ba373ba39c37d7b077235d3fb...f34935537d538e4ce27afd1e9b01a600c2adad13 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 15:42:46 2024 From: gitlab at gitlab.haskell.org (Andreas Klebinger (@AndreasK)) Date: Mon, 30 Sep 2024 11:42:46 -0400 Subject: [Git][ghc/ghc][wip/andreask/ghc-primops] 58 commits: rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. Message-ID: <66fac6f62aff5_334fc89034a46072f@gitlab.mail> Andreas Klebinger pushed to branch wip/andreask/ghc-primops at Glasgow Haskell Compiler / GHC Commits: ca67d7cb by Zubin Duggal at 2024-09-22T02:34:06-04:00 rts: Ensure we dump new Cost Centres added by freshly loaded objects to the eventlog. To do this, we keep track of the ID of the last cost centre we dumped in DUMPED_CC_ID, and call dumpCostCentresToEventLog from refreshProfilingCCSs, which will dump all the new cost centres up to the one we already dumped in DUMPED_CC_ID. Fixes #24148 - - - - - c0df5aa9 by Alan Zimmerman at 2024-09-22T02:34:42-04:00 EPA: Replace AnnsModule am_main with EpTokens Working towards removing `AddEpAnn` - - - - - 2a551cd5 by Matthew Pickering at 2024-09-24T16:33:50+05:30 ci: Run abi-test on test-abi label - - - - - ab4039ac by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 testsuite: Add a test for object determinism Extends the abi_test with an object determinism check Also includes a standalone test to be run by developers manually when debugging issues with determinism. - - - - - d62c18d8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Sampling uniques in the CG To achieve object determinism, the passes processing Cmm and the rest of the code generation pipeline musn't create new uniques which are non-deterministic. This commit changes occurrences of non-deterministic unique sampling within these code generation passes by a deterministic unique sampling strategy by propagating and threading through a deterministic incrementing counter in them. The threading is done implicitly with `UniqDSM` and `UniqDSMT`. Secondly, the `DUniqSupply` used to run a `UniqDSM` must be threaded through all passes to guarantee uniques in different passes are unique amongst them altogether. Specifically, the same `DUniqSupply` must be threaded through the CG Streaming pipeline, starting with Driver.Main calling `StgToCmm.codeGen`, `cmmPipeline`, `cmmToRawCmm`, and `codeOutput` in sequence. To thread resources through the `Stream` abstraction, we use the `UniqDSMT` transformer on top of `IO` as the Monad underlying the Stream. `UniqDSMT` will thread the `DUniqSupply` through every pass applied to the `Stream`, for every element. We use @type CgStream = Stream (UniqDSMT IO)@ for the Stream used in code generation which that carries through the deterministic unique supply. See Note [Deterministic Uniques in the CG] - - - - - 3bbe4af4 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Cmm unique renaming pass To achieve object determinism, we need to prevent the non-deterministic uniques from leaking into the object code. We can do this by deterministically renaming the non-external uniques in the Cmm groups that are yielded right after StgToCmm. The key to deterministic renaming is observing that the order of declarations, instructions, and data in the Cmm groups are already deterministic (modulo other determinism bugs), regardless of the uniques. We traverse the Cmm AST in this deterministic order and rename the uniques, incrementally, in the order they are found, thus making them deterministic. This renaming is guarded by -fobject-determinism which is disabled by default for now. This is one of the key passes for object determinism. Read about the overview of object determinism and a more detailed explanation of this pass in: * Note [Object determinism] * Note [Renaming uniques deterministically] Significantly closes the gap to #12935 - - - - - 8357ed50 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: DCmmGroup vs CmmGroup Part of our strategy in producing deterministic objects, namely, renaming all Cmm uniques in order, depend on the object code produced having a deterministic order (say, A_closure always comes before B_closure). However, the use of LabelMaps in the Cmm representation invalidated this requirement because the LabelMaps elements would already be in a non-deterministic order (due to the original uniques), and the renaming in sequence wouldn't work because of that non-deterministic order. Therefore, we now start off with lists in CmmGroup (which preserve the original order), and convert them into LabelMaps (for performance in the code generator) after the uniques of the list elements have been renamed. See Note [DCmmGroup vs CmmGroup or: Deterministic Info Tables] and #12935. Co-authored-by: Matthew Pickering <matthewtpickering at gmail.com> - - - - - 0e675fb8 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: Don't print unique in pprFullName This unique was leaking as part of the profiling description in info tables when profiling was enabled, despite not providing information relevant to the profile. - - - - - 340f58b0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: UDFM for distinct-constructor-tables In order to produce deterministic objects when compiling with -distinct-constructor-tables, we also have to update the data constructor map to be backed by a deterministic unique map (UDFM) rather than a non-deterministic one (UniqMap). - - - - - 282f37a0 by Rodrigo Mesquita at 2024-09-24T16:33:50+05:30 determinism: InfoTableMap uniques in generateCgIPEStub Fixes object determinism when using -finfo-table-map Make sure to also deterministically rename the IPE map (as per Note [Renaming uniques deterministically]), and to use a deterministic unique supply when creating new labels for the IPE information to guarantee deterministic objects when IPE information is requested. Note that the Cmm group produced in generateCgIPEStub must /not/ be renamed because renaming uniques is not idempotent, and the references to the previously renamed code in the IPE Cmm group would be renamed twice and become invalid references to non-existent symbols. We do need to det-rename the InfoTableMap that is created in the conversion from Core to Stg. This is not a problem since that map won't refer any already renamed names (since it was created before the renaming). - - - - - 7b37afc9 by Zubin Duggal at 2024-09-24T16:33:50+05:30 ci: Allow abi-test to fail. We are not fully deterministic yet, see #12935 for work that remains to be done. - - - - - a63ee33a by Simon Peyton Jones at 2024-09-25T17:08:24-04:00 Add Given injectivity for built-in type families Ticket #24845 asks (reasonably enough) that if we have [G] a+b ~ 0 then we also know [G] a ~ 0, b ~ 0 and similar injectivity-like facts for other built-in type families. The status quo was that we never generate evidence for injectivity among Givens -- but it is quite reasonnable to do so. All we need is to have /evidence/ for the new constraints This MR implements that goal. I also took the opportunity to * Address #24978: refactoring UnivCo * Fix #25248, which was a consequences of the previous formulation of UnivCo As a result this MR touches a lot of code. The big things are: * Coercion constructor UnivCo now takes a [Coercion] as argument to express the coercions on which the UnivCo depends. A nice consequence is that UnivCoProvenance now has no free variables, simpler in a number of places. * Coercion constructors AxiomInstCo and AxiomRuleCo are combined into AxiomCo. The new AxiomCo, carries a (slightly oddly named) CoAxiomRule, which itself is a sum type of the various forms of built-in axiom. See Note [CoAxiomRule] in GHC.Core.Coercion.Axiom A merit of this is that we can separate the case of open and closed type families, and eliminate the redundant `BranchIndex` in the former case. * Much better representation for data BuiltInSynFamily, which means we no longer need to enumerate built-in axioms as well as built-in tycons. * There is a massive refactor in GHC.Builtin.Types.Literals, which contains all the built-in axioms for type-level operations (arithmetic, append, cons etc). A big change is that instead of redundantly having (a) a hand-written matcher, and (b) a template-based "proves" function, which were hard to keep in sync, the two are derive from one set of human-supplied info. See GHC.Builtin.Types.Literals.mkRewriteAxiom, and friends. * Significant changes in GHC.Tc.Solver.Equality to account for the new opportunity for Given/Given equalities. Smaller things * Improve pretty-printing to avoid parens around atomic coercions. * Do proper eqType in findMatchingIrreds, not `eqTypeNoKindCheck`. Looks like a bug, Richard agrees. * coercionLKind and coercionRKind are hot functions. I refactored the implementation (which I had to change anyway) to increase sharing. See Note [coercionKind performance] in GHC.Core.Coercion * I wrote a new Note [Finding orphan names] in GHC.Core.FVs about orphan names * I improved the `is_concrete` flag in GHC.Core.Type.buildSynTyCon, to avoid calling tyConsOfType. I forget exactly why I did this, but it's definitely better now. * I moved some code from GHC.Tc.Types.Constraint into GHC.Tc.Types.CtLocEnv and I renamed the module GHC.Tc.Types.CtLocEnv to GHC.Tc.Types.CtLoc - - - - - dd8ef342 by Ryan Scott at 2024-09-25T17:09:01-04:00 Resolve ambiguous method-bound type variables in vanilla defaults and GND When defining an instance of a class with a "vanilla" default, such as in the following example (from #14266): ```hs class A t where f :: forall x m. Monoid x => t m -> m f = <blah> instance A [] ``` We have to reckon with the fact that the type of `x` (bound by the type signature for the `f` method) is ambiguous. If we don't deal with the ambiguity somehow, then when we generate the following code: ```hs instance A [] where f = $dmf @[] -- NB: the type of `x` is still ambiguous ``` Then the generated code will not typecheck. (Issue #25148 is a more recent example of the same problem.) To fix this, we bind the type variables from the method's original type signature using `TypeAbstractions` and instantiate `$dmf` with them using `TypeApplications`: ```hs instance A [] where f @x @m = $dmf @[] @x @m -- `x` is no longer ambiguous ``` Note that we only do this for vanilla defaults and not for generic defaults (i.e., defaults using `DefaultSignatures`). For the full details, see `Note [Default methods in instances] (Wrinkle: Ambiguous types from vanilla method type signatures)`. The same problem arose in the code generated by `GeneralizedNewtypeDeriving`, as we also fix it here using the same technique. This time, we can take advantage of the fact that `GeneralizedNewtypeDeriving`-generated code _already_ brings method-bound type variables into scope via `TypeAbstractions` (after !13190), so it is very straightforward to visibly apply the type variables on the right-hand sides of equations. See `Note [GND and ambiguity]`. Fixes #14266. Fixes #25148. - - - - - 0a4da5d2 by ARATA Mizuki at 2024-09-25T17:09:41-04:00 Document primitive string literals and desugaring of string literals Fixes #17474 and #17974 Co-authored-by: Matthew Craven <5086-clyring at users.noreply.gitlab.haskell.org> - - - - - ad0731ad by Zubin Duggal at 2024-09-25T17:10:18-04:00 rts: Fix segfault when using non-moving GC with profiling `nonMovingCollect()` swaps out the `static_flag` value used as a sentinel for `gct->scavenged_static_objects`, but the subsequent call `resetStaticObjectForProfiling()` sees the old value of `static_flag` used as the sentinel and segfaults. So we must call `resetStaticObjectForProfiling()` before calling `nonMovingCollect()` as otherwise it looks for the incorrect sentinel value Fixes #25232 and #23958 Also teach the testsuite driver about nonmoving profiling ways and stop disabling metric collection when nonmoving GC is enabled. - - - - - e7a26d7a by Sylvain Henry at 2024-09-25T17:11:00-04:00 Fix interaction between fork and kqueue (#24672) A kqueue file descriptor isn't inherited by a child created with fork. As such we mustn't try to close this file descriptor as we would close a random one, e.g. the one used by timerfd. Fix #24672 - - - - - 6863503c by Simon Peyton Jones at 2024-09-25T17:11:37-04:00 Improve GHC.Tc.Solver.defaultEquality This MR improves GHC.Tc.Solver.defaultEquality to solve #25251. The main change is to use checkTyEqRhs to check the equality, so that we do promotion properly. But within that we needed a small enhancement to LC_Promote. See Note [Defaulting equalites] (DE4) and (DE5) The tricky case is (alas) hard to trigger, so I have not added a regression test. - - - - - 97a6c6c3 by Sylvain Henry at 2024-09-25T17:12:18-04:00 JS: fix h$withCStringOnHeap helper (#25288) strlen returns the length of the string without the \0 terminating byte, hence CString weren't properly allocated on the heap (ending \0 byte was missing). - - - - - 5f7c20bc by Ben Gamari at 2024-09-26T04:14:05-04:00 base: Propagate `error` CallStack to thrown exception Previously `errorCallWithCallStackException` failed to propagate its `CallStack` argument, which represents the call-chain of the preceding `error` call, to the exception that it returned. Consequently, the call-stack of `error` calls were quite useless. Unfortunately, this is the second time that I have fixed this but it seems the first must have been lost in rebasing. Fixes a bug in the implementation of CLC proposal 164 <https://github.com/haskell/core-libraries-committee/issues/164> Fixes #24807. - - - - - c20d5186 by Matthew Pickering at 2024-09-26T04:14:42-04:00 driver: Fix -working-dir for foreign files -working-dir definitely needs more serious testing, there are some easy ways to test this. * Modify Cabal to call ghc using -working-dir rather than changing directory. * Modify the testsuite to run ghc using `-working-dir` rather than running GHC with cwd = temporary directory. However this will have to wait until after 9.12. Fixes #25150 - - - - - 88eaa7ac by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: reuse predError, succError, toEnumError Reuse predError, succError, and toEnumError when deriving Enum instances to avoid generating different error strings per instance. E.g. before this patch for every instance for a type FOO we would generate a string: "pred{FOO}: tried to take `pred' of first tag in enumeration"# - - - - - e9fa1163 by Sylvain Henry at 2024-09-26T04:15:24-04:00 Enum deriving: generate better code (#16364) Generate better code for Enum.toEnum: check both the lower and the upper bounds at once with an unsigned comparison. Initially I've used a type ascription with a call to 'fromIntegral', hence the slight refactoring of nlAscribe. Using 'fromIntegral' was problematic (too low in the module hierarchy) so 'enumIntToWord' was introduced instead. Combined with the previous commit, T21839c ghc/alloc decrease by 5% Metric Decrease: T21839c - - - - - 383af074 by Sylvain Henry at 2024-09-26T04:16:06-04:00 Core: add absorb rules for binary or/and (#16351) Rules: x or (x and y) ==> x x and (x or y) ==> x - - - - - 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - 8509f699 by Andreas Klebinger at 2024-09-30T17:23:33+02:00 ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps This will be the new place for functions that would have gone into GHC.Exts in the past but are not stable enough to do so now. Addresses #25242 - - - - - 21 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/Types/Literals.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/BlockId.hs - compiler/GHC/Cmm/CLabel.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Dataflow.hs - compiler/GHC/Cmm/Dataflow/Graph.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Info.hs - compiler/GHC/Cmm/Info/Build.hs - compiler/GHC/Cmm/LayoutStack.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3c32c0602eac46602fb8c86088ba7e430d5f2fc6...8509f6999a99f1d57e5eb3ac95aab67a8f69b1d1 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/3c32c0602eac46602fb8c86088ba7e430d5f2fc6...8509f6999a99f1d57e5eb3ac95aab67a8f69b1d1 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 15:52:19 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 30 Sep 2024 11:52:19 -0400 Subject: [Git][ghc/ghc][wip/romes/exceptions-propagate] 41 commits: Don't compile `asBox` with -fprof-late Message-ID: <66fac93388917_15ec30d337451956@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/exceptions-propagate at Glasgow Haskell Compiler / GHC Commits: 783c8b29 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Don't compile `asBox` with -fprof-late The `asBox` function is intended to store exactly the closure which the user passes to it. Placing a cost centre on asBox introduces a thunk, which violates this expectation and can change the result of using asBox when profiling is enabled. See #25212 for more details and ample opportunity to discuss if this is a bug or not. - - - - - 0967dcc7 by Matthew Pickering at 2024-09-26T12:07:44-04:00 Fix normalisation of .prof files Fix 1: If a cost centre contained CAF then the normalisation was corrupted, now only check if CAF is at the start of a line. Fix 2: "no location info" contain a space, which messed up the next normalisation logic which assumed that columns didn't have spaced in. - - - - - 9eda1cb9 by Matthew Pickering at 2024-09-26T12:07:44-04:00 testsuite: Fix normalisation of prof_files removing newlines These normalisation steps were collapsing lines together, which made subsequent normalisation steps fail. ``` foo x y z CAF x y z qux x y z ``` was getting normalised to ``` foo x y z qux x y z ``` which means that subsequent line based filters would not work correctly. - - - - - 2b25f9e2 by Matthew Pickering at 2024-09-26T12:07:44-04:00 packaging: Enable late-ccs for release flavour This enables late cost centres when building profiled libraries and subsequently greatly improves the resolution of cost centre stacks when profiling. This patch also introduces the `grep_prof` test modifier which is used to apply a further filter to the .prof file before they are compared. Fixes #21732 ------------------------- Metric Increase: libdir ------------------------- - - - - - bb030d0d by Brandon Chinn at 2024-09-26T12:08:21-04:00 Replace manual string lexing (#25158) Metric Increase: MultilineStringsPerf This commit replaces the manual string lexing logic with native Alex lexing syntax. This aligns the lexer much closer to the Haskell Report, making it easier to see how the implementation and spec relate. This slightly increases memory usage when compiling multiline strings because we now have two distinct phases: lexing the multiline string with Alex and post-processing the string afterwards. Before, these were done at the same time, but separating them allows us to push as much logic into normal Alex lexing as possible. Since multiline strings are a new feature, this regression shouldn't be too noticeable. We can optimize this over time. - - - - - 16742987 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Revert !4655: Stop 'import "base" Prelude' removing implicit Prelude import This behaviour is problematic for the principle reason that `import Prelude` may not refer to the `base` package, and in which case importing an entirely unrelated module causing your implicit prelude to leave the scope is extremely surprising. See the added test for this example. Discussion on #17045. The secondary reason for reverting this patch is that "base" can't be a wired in package any more (see #24903), so we have to remove special logic which singles out base from the compiler. The rule for implicit shadowing is now simply: * If you write import Prelude (..) then you don't get an implicit prelude import * If you write import "foobar" Prelude (..) for all pkgs foobar, you get an implicit import of prelude. If you want to write a package import of Prelude, then you can enable `NoImplicitPrelude` for the module in question to recover the behaviour of ghc-9.2-9.10. Fixes #17045 - - - - - 57c50f41 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Rename COMPILING_BASE_PACKAGE to COMPILING_GHC_INTERNAL_PACKAGE The COMPILING_BASE_PACKAGE macro is concerned with issues defining symbols and using symbols in the same compilation unit. However, these symbols now exist in ghc-internal rather than base, so we should rename the macro accordingly. The code is guards is likely never used as we never produce windows DLLs but it is simpler to just perform the renaming for now. These days there is little doubt that this macro defined in this ad-hoc manner would be permitted to exist, but these days are not those days. Fixes #25221 - - - - - 70764243 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Preload ghc-internal rather than base This occurence of baseUnitId was missed when moving the bulk of internal definitions into `ghc-internal`. We need to remove this preloading of `base` now because `base` should not be wired in. Towards #24903 - - - - - 12915609 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Remove Data.List compat warning There is currently a warning implemented in -Wcompat which warns you when importing Data.List in a non-qualified manner. ``` A.hs:3:8: warning: [-Wcompat-unqualified-imports] To ensure compatibility with future core libraries changes imports to Data.List should be either qualified or have an explicit import list. | 3 | import Data.List | ^^^^^^^^^ Ok, one module loaded. ``` GHC ticket: https://gitlab.haskell.org/ghc/ghc/-/issues/17244 CLC discussion: https://groups.google.com/g/haskell-core-libraries/c/q3zHLmzBa5E This warning was implemented as part of the migration to making Data.List monomorphic again (and to be used like Data.Set, Data.Map etc). That doesn't seem like it happened, and I imagine that the current CLC would require a new proposal anyway in order to do that now. It's not clear in any case what "future core libraries changes" we are waiting to happen before this warning can be removed. Given the first phase of the proposal has lasted 5 years it doesn't seem that anyone is motivated to carry the proposal to completion. It does seem a bit unnecessary to include a warning in the compiler about "future changes to the module" when there's no timeline or volunteer to implement these changes. The removal of this warning was discussed again at: https://github.com/haskell/core-libraries-committee/issues/269 During the discussion there was no new enthusiasm to move onto the next stages of the proposal so we are removing the warning to unblock the reinstallable "base" project (#24903) Fixes #24904 - - - - - d4e4d498 by Matthew Pickering at 2024-09-26T12:08:57-04:00 Move Control.Monad.Zip into ghc-internal mzip is wired in and therefore needs to be in ghc-internal. Fixes #25222 Towards #24903 - - - - - d3dacdfb by Matthew Pickering at 2024-09-26T12:08:57-04:00 Unwire the base package This patch just removes all the functions related to wiring-in the base package and the `-this-unit-id=base` flag from the cabal file. After this commit "base" becomes just like any other package and the door is opened to moving base into an external repo and releasing base on a separate schedule to the rest of ghc. Closes #24903 - - - - - 1b39363b by Patrick at 2024-09-27T06:10:19-04:00 Add entity information to HieFile #24544 Enhanced HieFile to capture entity information for identifiers, enabling better support for language tools and protocols. See issue #24544 for more details. Work have been done: * Introduction of new data type `EntityInfo` in `GHC.Iface.Ext.Types`. * Add extra field `hie_entity_infos :: NameEntityInfo` to `HieFile` to store the mapping from entity name to corresponding entity infos in `GHC.Iface.Ext.Types`. * Compute `EntityInfo` for each entity name in the HieAst from `TyThing, Id, OccName` when generating the `HieFile` in `GHC.Iface.Ext.Ast`. * Add test T24544 to test the generation of `EntityInfo`. - - - - - 4f3618d8 by sheaf at 2024-09-27T06:10:57-04:00 The X86 SIMD patch. This commit adds support for 128 bit wide SIMD vectors and vector operations to GHC's X86 native code generator. Main changes: - Introduction of vector formats (`GHC.CmmToAsm.Format`) - Introduction of 128-bit virtual register (`GHC.Platform.Reg`), and removal of unused Float virtual register. - Refactor of `GHC.Platform.Reg.Class.RegClass`: it now only contains two classes, `RcInteger` (for general purpose registers) and `RcFloatOrVector` (for registers that can be used for scalar floating point values as well as vectors). - Modify `GHC.CmmToAsm.X86.Instr.regUsageOfInstr` to keep track of which format each register is used at, so that the register allocator can know if it needs to spill the entire vector register or just the lower 64 bits. - Modify spill/load/reg-2-reg code to account for vector registers (`GHC.CmmToAsm.X86.Instr.{mkSpillInstr, mkLoadInstr, mkRegRegMoveInstr, takeRegRegMoveInstr}`). - Modify the register allocator code (`GHC.CmmToAsm.Reg.*`) to propagate the format we are storing in any given register, for instance changing `Reg` to `RegFormat` or `GlobalReg` to `GlobalRegUse`. - Add logic to lower vector `MachOp`s to X86 assembly (see `GHC.CmmToAsm.X86.CodeGen`) - Minor cleanups to genprimopcode, to remove the llvm_only attribute which is no longer applicable. Tests for this feature are provided in the "testsuite/tests/simd" directory. Fixes #7741 Keeping track of register formats adds a small memory overhead to the register allocator (in particular, regUsageOfInstr now allocates more to keep track of the `Format` each register is used at). This explains the following metric increases. ------------------------- Metric Increase: T12707 T13035 T13379 T3294 T4801 T5321FD T5321Fun T783 ------------------------- - - - - - 10e431ef by sheaf at 2024-09-27T06:10:57-04:00 Use xmm registers in genapply This commit updates genapply to use xmm, ymm and zmm registers, for stg_ap_v16/stg_ap_v32/stg_ap_v64, respectively. It also updates the Cmm lexer and parser to produce Cmm vectors rather than 128/256/512 bit wide scalars for V16/V32/V64, removing bits128, bits256 and bits512 in favour of vectors. The Cmm Lint check is weakened for vectors, as (in practice, e.g. on X86) it is okay to use a single vector register to hold multiple different types of data, and we don't know just from seeing e.g. "XMM1" how to interpret the 128 bits of data within. Fixes #25062 - - - - - 8238fb2d by sheaf at 2024-09-27T06:10:57-04:00 Add vector fused multiply-add operations This commit adds fused multiply add operations such as `fmaddDoubleX2#`. These are handled both in the X86 NCG and the LLVM backends. - - - - - 2cb7b748 by sheaf at 2024-09-27T06:10:57-04:00 Add vector shuffle primops This adds vector shuffle primops, such as ``` shuffleFloatX4# :: FloatX4# -> FloatX4# -> (# Int#, Int#, Int#, Int# #) -> FloatX4# ``` which shuffle the components of the input two vectors into the output vector. NB: the indices must be compile time literals, to match the X86 SHUFPD instruction immediate and the LLVM shufflevector instruction. These are handled in the X86 NCG and the LLVM backend. Tested in simd009. - - - - - 0d2428d6 by sheaf at 2024-09-27T06:10:57-04:00 Add Broadcast MachOps This adds proper MachOps for broadcast instructions, allowing us to produce better code for broadcasting a value than simply packing that value (doing many vector insertions in a row). These are lowered in the X86 NCG and LLVM backends. In the LLVM backend, it uses the previously introduced shuffle instructions. - - - - - e6c19a41 by sheaf at 2024-09-27T06:10:57-04:00 Fix treatment of signed zero in vector negation This commit fixes the handling of signed zero in floating-point vector negation. A slight hack was introduced to work around the fact that Cmm doesn't currently have a notion of signed floating point literals (see get_float_broadcast_value_reg). This can be removed once CmmFloat can express the value -0.0. The simd006 test has been updated to use a stricter notion of equality of floating-point values, which ensure the validity of this change. - - - - - f496ff7f by sheaf at 2024-09-27T06:10:57-04:00 Add min/max primops This commit adds min/max primops, such as minDouble# :: Double# -> Double# -> Double# minFloatX4# :: FloatX4# -> FloatX4# -> FloatX4# minWord16X8# :: Word16X8# -> Word16X8# -> Word16X8# These are supported in: - the X86, AArch64 and PowerPC NCGs, - the LLVM backend, - the WebAssembly and JavaScript backends. Fixes #25120 - - - - - 5dd2a423 by sheaf at 2024-09-27T06:10:57-04:00 Add test for C calls & SIMD vectors - - - - - f824e1ee by sheaf at 2024-09-27T06:10:58-04:00 Add test for #25169 - - - - - d54db7f3 by sheaf at 2024-09-27T06:10:58-04:00 Fix #25169 using Plan A from the ticket We now compile certain low-level Cmm functions in the RTS multiple times, with different levels of vector support. We then dispatch at runtime in the RTS, based on what instructions are supported. See Note [realArgRegsCover] in GHC.Cmm.CallConv. Fixes #25169 ------------------------- Metric Increase: T10421 T12425 T18730 T1969 T9198 ------------------------- - - - - - d5f8778a by sheaf at 2024-09-27T06:10:58-04:00 Fix C calls with SIMD vectors This commit fixes the code generation for C calls, to take into account the calling convention. This is particularly tricky on Windows, where all vectors are expected to be passed by reference. See Note [The Windows X64 C calling convention] in GHC.CmmToAsm.X86.CodeGen. - - - - - f64bd564 by sheaf at 2024-09-27T06:10:58-04:00 X86 CodeGen: refactor getRegister CmmLit This refactors the code dealing with loading literals into registers, removing duplication and putting all the code in a single place. It also changes which XOR instruction is used to place a zero value into a register, so that we use VPXOR for a 128-bit integer vector when AVX is supported. - - - - - ab12de6b by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall: promote arg before calling evalArgs The job of evalArgs is to ensure each argument is put into a temporary register, so that it can then be loaded directly into one of the argument registers for the C call, without the generated code clobbering any other register used for argument passing. However, if we promote arguments after calling evalArgs, there is the possibility that the code used for the promotion will clobber a register, defeating the work of evalArgs. To avoid this, we first promote arguments, and only then call evalArgs. - - - - - 8fd12429 by sheaf at 2024-09-27T06:10:58-04:00 X86 genCCall64: simplify loadArg code This commit simplifies the argument loading code by making the assumption that it is safe to directly load the argument into register, because doing so will not clobber any previous assignments. This assumption is borne from the use of 'evalArgs', which evaluates any arguments which might necessitate non-trivial code generation into separate temporary registers. - - - - - 12504a9f by sheaf at 2024-09-27T06:10:58-04:00 LLVM: propagate GlobalRegUse information This commit ensures we keep track of how any particular global register is being used in the LLVM backend. This informs the LLVM type annotations, and avoids type mismatches of the following form: argument is not of expected type '<2 x double>' call ccc <2 x double> (<2 x double>) (<4 x i32> arg) - - - - - 2bb1e8df by Cheng Shao at 2024-09-27T06:11:35-04:00 Link bytecode from interface-stored core bindings in oneshot mode !13042 Part of #T25090 If the flag `-fprefer-byte-code` is given when compiling a module containing TH, GHC will use Core bindings stored in interfaces to compile and link bytecode for splices. This was only implemented for `--make` mode initially, so this commit adds the same mechanism to oneshot mode (`-c`). When an interface is loaded into the EPS in `loadInterface` that has dehydrated Core bindings, an entry is added to the new field `eps_iface_bytecode`, containing an IO action that produces a bytecode `Linkable`, lazily processing the `mi_extra_decls` by calling `loadIfaceByteCode`. When Template Haskell dependencies are resolved in `getLinkDeps`, this action is looked up after loading a module's interface. If it exists, the action is evaluated and the bytecode is added to the set of `Linkable`s used for execution of the splice; otherwise it falls back on the traditional object file. Metric Decrease: MultiLayerModules T13701 - - - - - 7cb7172e by Matthew Pickering at 2024-09-27T06:12:12-04:00 ci: Fix variable inheritence for ghcup-metadata testing job Downstream in ghcup-ci we use the CONFIGURE_ARGS variable to determine how to setup all the different jobs. On the downstream trigger this was being inherited from the default setting in .gitlab.yml file. Therefore this led to job failures as the necessary CONFIGURE_ARGS were not being passed to the configure script when installing the bindist. See docs: * https://docs.gitlab.com/ee/ci/yaml/#inherit * https://docs.gitlab.com/ee/ci/yaml/#triggerforward 1. inherit:variables:fals - This stops the global variables being inherited into the job and hence forwarded onto the downstream job. 2. trigger:forward:* - yaml_variables: true (default) pass yaml variables to downstream, this is important to pass the upstream pipeline id to downstream. - pipeline_variables: false (default) but don't pass pipeline variables (normal environment variables). Fixes #25294 - - - - - 9ffd6163 by Leo at 2024-09-27T16:26:01+05:30 Fix typo in Prelude doc for (>>=) Fix a minor typo ("equivialent" instead of "equivalent") in the documentation for (>>=) in the prelude. - - - - - 5745dbd3 by Vladislav Zavialov at 2024-09-27T16:26:52+05:30 Wildcard binders in type declarations (#23501) Add support for wildcard binders in type declarations: type Const a b = a -- BEFORE: the `b` had to be named -- even if unused on the RHS type Const a _ = a -- AFTER: the compiler accepts -- a wildcard binder `_` The new feature is part of GHC Proposal #425 "Invisible binders in type declarations", and more specifically its amendment #641. Just like a named binder, a wildcard binder `_` may be: * plain: _ * kinded: (_ :: k -> Type) * invisible, plain: @_ * invisible, kinded: @(_ :: k -> Type) Those new forms of binders are allowed to occur on the LHSs of data, newtype, type, class, and type/data family declarations: data D _ = ... newtype N _ = ... type T _ = ... class C _ where ... type family F _ data family DF _ (Test case: testsuite/tests/typecheck/should_compile/T23501a.hs) However, we choose to reject them in forall telescopes and type family result variable binders (the latter being part of the TypeFamilyDependencies extension): type family Fd a = _ -- disallowed (WildcardBndrInTyFamResultVar) fn :: forall _. Int -- disallowed (WildcardBndrInForallTelescope) (Test case: testsuite/tests/rename/should_fail/T23501_fail.hs) See the new Notes: * Note [Type variable binders] * Note [Wildcard binders in disallowed contexts] To accommodate the new forms of binders, HsTyVarBndr was changed as follows (demonstrated without x-fields for clarity) -- BEFORE (ignoring x-fields and locations) data HsTyVarBndr flag = UserTyVar flag Name | KindedTyVar flag Name HsKind -- AFTER (ignoring x-fields and locations) data HsTyVarBndr flag = HsTvb flag HsBndrVar HsBndrKind data HsBndrVar = HsBndrVar Name | HsBndrWildCard data HsBndrKind = HsBndrNoKind | HsBndrKind LHsKind The rest of the patch is downstream from this change. To avoid a breaking change to the TH AST, we generate fresh names to replace wildcard binders instead of adding a dedicated representation for them (as discussed in #641). And to put a cherry on top of the cake, we now allow wildcards in kind-polymorphic type variable binders in constructor patterns, see Note [Type patterns: binders and unifiers] and the tyPatToBndr function in GHC.Tc.Gen.HsType; example: fn (MkT @(_ :: forall k. k -> Type) _ _) = ... (Test case: testsuite/tests/typecheck/should_compile/T23501b.hs) - - - - - ff2bdca2 by Matthew Pickering at 2024-09-27T16:27:08+05:30 ci: Push perf notes from wasm jobs It was observed in #25299 that we were failing to push performance numbers from the wasm jobs. In future we might want to remove this ad-hoc check but for now it's easier to add another special case. Towards #25299 - - - - - 4c76f75c by Zubin Duggal at 2024-09-27T16:44:00+05:30 Bump GHC version to 9.12 - - - - - e4ac1b0d by Zubin Duggal at 2024-09-27T19:12:24+05:30 Bump GHC version to 9.13 - - - - - 7d6945cd by Matthew Pickering at 2024-09-30T10:31:29+01:00 Fix toException method for ExceptionWithContext Fixes #25235 - - - - - 1b121165 by Matthew Pickering at 2024-09-30T10:43:34+01:00 Exception rethrowing Basic changes: * Change `catch` function to propagate exceptions using the WhileHandling mechanism. * Introduce `catchNoPropagate`, which does the same as before, but passes an exception which can be rethrown. * Introduce `rethrowIO` combinator, which rethrows an exception with a context and doesn't add a new backtrace. * Introduce `tryWithContext` for a variant of `try` which can rethrow the exception with it's original context. * onException is modified to rethrow the original error rather than creating a new callstack. * Functions which rethrow in GHC.Internal.IO.Handle.FD, GHC.Internal.IO.Handle.Internals, GHC.Internal.IO.Handle.Text, and GHC.Internal.System.IO.Error are modified to not add a new callstack. Implements CLC proposal#202 <https://github.com/haskell/core-libraries-committee/issues/202> - - - - - f9591422 by Rodrigo Mesquita at 2024-09-30T10:43:34+01:00 Add test for #25300 - - - - - e3e17f21 by Rodrigo Mesquita at 2024-09-30T12:10:46+01:00 exceptions: Improve the message layout as per #285 This commit fixes the layout of the additional information included when displaying an exception, namely the type of the exception. It also fixes the default handler's heading message to work well together with the improved display message of SomeException. CLC proposal#285 - - - - - 6b5162c7 by Rodrigo Mesquita at 2024-09-30T12:11:02+01:00 Display type and callstack of exception on handler This commit changes the Exception instance of SomeException to *simply* display the underlying exception in `displayException`. The augmented exception message that included the type and backtrace of the exception are now only printed on a call to `displayExceptionWithInfo`. At a surface level, existing programs should behave the same since the `uncaughtExceptionHandler`, which is responsible for printing out uncaught exceptions to the user, will use `displayExceptionWithInfo` by default. However, unlike the instance's `displayException` method, the `uncaughtExceptionHandler` can be overriden with `setUncaughtExceptionHandler`. This makes the extra information opt-in without fixing it the instance, which can be valuable if your program wants to display uncaught exceptions to users in a user-facing way (ie without backtraces). This is what was originally agreed for CLC#231 or CLC#261 with regard to the type of the exception information. The call stack also becoming part of the default handler rather than the Exception instance is an ammendment to CLC#164. Discussion of the ammendment is part of CLC#285. - - - - - 5ee369e5 by Rodrigo Mesquita at 2024-09-30T13:47:47+01:00 Do #25283 - - - - - 571aa8f3 by Rodrigo Mesquita at 2024-09-30T14:59:27+01:00 Freeze call stack in error throwing functions - - - - - 30 changed files: - .gitlab-ci.yml - .gitlab/ci.sh - compiler/GHC/Builtin/Names.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - compiler/GHC/Cmm.hs - compiler/GHC/Cmm/CallConv.hs - compiler/GHC/Cmm/Graph.hs - compiler/GHC/Cmm/Lexer.x - compiler/GHC/Cmm/Lint.hs - compiler/GHC/Cmm/Liveness.hs - compiler/GHC/Cmm/MachOp.hs - compiler/GHC/Cmm/Node.hs - compiler/GHC/Cmm/Opt.hs - compiler/GHC/Cmm/Parser.y - compiler/GHC/Cmm/ProcPoint.hs - compiler/GHC/Cmm/Reg.hs - compiler/GHC/Cmm/Sink.hs - compiler/GHC/Cmm/Type.hs - compiler/GHC/Cmm/UniqueRenamer.hs - compiler/GHC/CmmToAsm.hs - compiler/GHC/CmmToAsm/AArch64.hs - compiler/GHC/CmmToAsm/AArch64/CodeGen.hs - compiler/GHC/CmmToAsm/AArch64/Instr.hs - compiler/GHC/CmmToAsm/AArch64/Ppr.hs - compiler/GHC/CmmToAsm/AArch64/Regs.hs - compiler/GHC/CmmToAsm/Config.hs - compiler/GHC/CmmToAsm/Format.hs - compiler/GHC/CmmToAsm/Instr.hs - compiler/GHC/CmmToAsm/PPC.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b4bfc7d578ea3a580b88fcabd20871afaa7eea43...571aa8f3433cbc59cd43c4d9d072c1b9a6182e6b -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b4bfc7d578ea3a580b88fcabd20871afaa7eea43...571aa8f3433cbc59cd43c4d9d072c1b9a6182e6b You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 16:07:03 2024 From: gitlab at gitlab.haskell.org (Rodrigo Mesquita (@alt-romes)) Date: Mon, 30 Sep 2024 12:07:03 -0400 Subject: [Git][ghc/ghc][wip/romes/25304] determinism: Interface re-export list det Message-ID: <66facca76fa1e_15ec3026f96c56413@gitlab.mail> Rodrigo Mesquita pushed to branch wip/romes/25304 at Glasgow Haskell Compiler / GHC Commits: c18e7a48 by Rodrigo Mesquita at 2024-09-30T17:05:16+01:00 determinism: Interface re-export list det In 'DocStructureItem' we want to make sure the 'Avails' are sorted, for interface file determinism. This commit introduces 'SortedAvails', an abstract newtype that can only be constructed by sorting Avails with 'sortAvails'. This newtype is used by 'DocStructureItem' where 'Avails' was previously used to ensure the list of avails is deterministically sorted by construction. Fixes #25304 - - - - - 12 changed files: - compiler/GHC/Hs/Doc.hs - compiler/GHC/HsToCore/Docs.hs - compiler/GHC/Iface/Make.hs - compiler/GHC/Types/Avail.hs - + testsuite/tests/determinism/T25304/A.hs - + testsuite/tests/determinism/T25304/B.hs - + testsuite/tests/determinism/T25304/Makefile - + testsuite/tests/determinism/T25304/T25304a.stdout - + testsuite/tests/determinism/T25304/all.T - testsuite/tests/showIface/DocsInHiFileTH.stdout - testsuite/tests/showIface/NoExportList.stdout - utils/haddock/haddock-api/src/Haddock/Interface/Create.hs Changes: ===================================== compiler/GHC/Hs/Doc.hs ===================================== @@ -124,7 +124,7 @@ data DocStructureItem = DsiSectionHeading !Int !(HsDoc GhcRn) | DsiDocChunk !(HsDoc GhcRn) | DsiNamedChunkRef !String - | DsiExports !Avails + | DsiExports !SortedAvails | DsiModExport !(NonEmpty ModuleName) -- ^ We might re-export avails from multiple -- modules with a single export declaration. E.g. @@ -136,7 +136,7 @@ data DocStructureItem -- -- Invariant: This list of ModuleNames must be -- sorted to guarantee interface file determinism. - !Avails + !SortedAvails -- ^ Invariant: This list of Avails must be sorted -- to guarantee interface file determinism. ===================================== compiler/GHC/HsToCore/Docs.hs ===================================== @@ -160,7 +160,11 @@ mkDocStructureFromExportList mdl import_avails export_list = (IEGroup _ level doc, _) -> DsiSectionHeading level (unLoc doc) (IEDoc _ doc, _) -> DsiDocChunk (unLoc doc) (IEDocNamed _ name, _) -> DsiNamedChunkRef name - (_, avails) -> DsiExports (nubAvails avails) + (IEThingWith{}, avails) -> + DsiExports $ + {- For explicit export lists, use the explicit order. It is deterministic by construction -} + UnsafeSortedAvails (nubAvails avails) + (_, avails) -> DsiExports (sortAvails (nubAvails avails)) moduleExport :: ModuleName -- Alias -> Avails @@ -201,10 +205,10 @@ mkDocStructureFromDecls env all_exports decls = avails :: [Located DocStructureItem] avails = flip fmap all_exports $ \avail -> case M.lookup (availName avail) name_locs of - Just loc -> L loc (DsiExports [avail]) + Just loc -> L loc (DsiExports (sortAvails [avail])) -- FIXME: This is just a workaround that we use when handling e.g. -- associated data families like in the html-test Instances.hs. - Nothing -> noLoc (DsiExports []) + Nothing -> noLoc (DsiExports (sortAvails [])) -- This causes the associated data family to be incorrectly documented -- separately from its class: ===================================== compiler/GHC/Iface/Make.hs ===================================== @@ -518,8 +518,8 @@ mkIfaceImports = map go go (ImpUserSpec decl (ImpUserExplicit env)) = IfaceImport decl (ImpIfaceExplicit (forceGlobalRdrEnv env)) go (ImpUserSpec decl (ImpUserEverythingBut ns)) = IfaceImport decl (ImpIfaceEverythingBut ns) -mkIfaceExports :: [AvailInfo] -> [IfaceExport] -- Sort to make canonical -mkIfaceExports = sortAvails +mkIfaceExports :: [AvailInfo] -> [IfaceExport] -- Sort to make canonical +mkIfaceExports as = case sortAvails as of SortedAvails sas -> sas {- Note [Original module] ===================================== compiler/GHC/Types/Avail.hs ===================================== @@ -1,5 +1,7 @@ {-# LANGUAGE DeriveDataTypeable #-} +{-# LANGUAGE DerivingStrategies #-} +{-# LANGUAGE PatternSynonyms #-} -- -- (c) The University of Glasgow -- @@ -20,6 +22,7 @@ module GHC.Types.Avail ( filterAvails, nubAvails, sortAvails, + SortedAvails(SortedAvails, UnsafeSortedAvails) ) where import GHC.Prelude @@ -65,6 +68,20 @@ data AvailInfo -- | A collection of 'AvailInfo' - several things that are \"available\" type Avails = [AvailInfo] +-- | Occurrences of Avails in interface files must be sorted to guarantee +-- interface file determinism. +-- +-- To construct 'SortedAvails' using 'UnsafeSortedAvails' you must be sure the +-- 'Avails' are already sorted. Otherwise, you should use 'sortAvails'. +newtype SortedAvails = UnsafeSortedAvails Avails + deriving newtype (Binary, Outputable, NFData) + +-- | Safe matching on 'SortedAvails' +-- To construct 'SortedAvails' use 'sortAvails'. +pattern SortedAvails :: Avails -> SortedAvails +pattern SortedAvails x <- UnsafeSortedAvails x +{-# COMPLETE SortedAvails #-} + {- Note [Representing pattern synonym fields in AvailInfo] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Record pattern synonym fields cannot be represented using AvailTC like fields of @@ -133,8 +150,8 @@ availSubordinateNames avail@(AvailTC _ ns) | otherwise = ns -- | Sort 'Avails'/'AvailInfo's -sortAvails :: Avails -> Avails -sortAvails = sortBy stableAvailCmp . map sort_subs +sortAvails :: Avails -> SortedAvails +sortAvails = UnsafeSortedAvails . sortBy stableAvailCmp . map sort_subs where sort_subs :: AvailInfo -> AvailInfo sort_subs (Avail n) = Avail n ===================================== testsuite/tests/determinism/T25304/A.hs ===================================== @@ -0,0 +1,84 @@ +module A + ( MyType(..) + ) where + +data MyType + = A + | B + | C + | D + | E + | F + | G + | H + | I + | J + | K + | L + | M + | N + | O + | P + | Q + | R + | S + | T + | U + | V + | W + | X + | Y + | Z + | AA + | AB + | AC + | AD + | AE + | AF + | AG + | AH + | AI + | AJ + | AK + | AL + | AM + | AN + | AO + | AP + | AQ + | AR + | AS + | AT + | AU + | AV + | AW + | AX + | AY + | AZ + | BA + | BB + | BC + | BD + | BE + | BF + | BG + | BH + | BI + | BJ + | BK + | BL + | BM + | BN + | BO + | BP + | BQ + | BR + | BS + | BT + | BU + | BV + | BW + | BX + | BY + | BZ + | CA ===================================== testsuite/tests/determinism/T25304/B.hs ===================================== @@ -0,0 +1,86 @@ +module B +( MyType + ( BA + , BB + , BC + , BD + , BE + , BF + , BG + , BH + , BI + , BJ + , BK + , BL + , BM + , BN + , BO + , BP + , BQ + , BR + , BS + , BT + , BU + , BV + , BW + , BX + , BY + , BZ + , CA + , AA + , AB + , AC + , AD + , AE + , AF + , AG + , AH + , AI + , AJ + , AK + , AL + , AM + , AN + , AO + , AP + , AQ + , AR + , AS + , AT + , AU + , AV + , AW + , AX + , AY + , AZ + , A + , B + , C + , D + , E + , F + , G + , H + , I + , J + , K + , L + , M + , N + , O + , P + , Q + , R + , S + , T + , U + , V + , W + , X + , Y + , Z + ) +) where + +import A + ===================================== testsuite/tests/determinism/T25304/Makefile ===================================== @@ -0,0 +1,25 @@ +TOP=../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +T25304: + $(RM) A.hi A.o B.hi B.o + # Use -haddock to get docs: output in the interface file + '$(TEST_HC)' $(TEST_HC_OPTS) -v0 -haddock B.hs + '$(TEST_HC)' --show-iface A.hi > A_clean_iface + '$(TEST_HC)' --show-iface B.hi > B_clean_iface + '$(TEST_HC)' $(TEST_HC_OPTS) -dunique-increment=-1 -v0 -haddock A.hs B.hs -fforce-recomp + '$(TEST_HC)' --show-iface A.hi > A_dirty_iface + '$(TEST_HC)' --show-iface B.hi > B_dirty_iface + diff A_clean_iface A_dirty_iface + diff B_clean_iface B_dirty_iface + +T25304a: + $(RM) A.hi A.o B.hi B.o + # Use -haddock to get docs: output in the interface file + '$(TEST_HC)' $(TEST_HC_OPTS) -v0 -haddock B.hs + '$(TEST_HC)' --show-iface B.hi > B_clean_iface + # The goal is to see the export list in the documentation structure of the + # interface file preserves the order used in the source + cat B_clean_iface | grep -A7 "documentation structure" + ===================================== testsuite/tests/determinism/T25304/T25304a.stdout ===================================== @@ -0,0 +1,8 @@ + documentation structure: + avails: + [A.MyType{A.MyType, A.BA, A.BB, A.BC, A.BD, A.BE, A.BF, A.BG, A.BH, + A.BI, A.BJ, A.BK, A.BL, A.BM, A.BN, A.BO, A.BP, A.BQ, A.BR, A.BS, + A.BT, A.BU, A.BV, A.BW, A.BX, A.BY, A.BZ, A.CA, A.AA, A.AB, A.AC, + A.AD, A.AE, A.AF, A.AG, A.AH, A.AI, A.AJ, A.AK, A.AL, A.AM, A.AN, + A.AO, A.AP, A.AQ, A.AR, A.AS, A.AT, A.AU, A.AV, A.AW, A.AX, A.AY, + A.AZ, A.A, A.B, A.C, A.D, A.E, A.F, A.G, A.H, A.I, A.J, A.K, A.L, ===================================== testsuite/tests/determinism/T25304/all.T ===================================== @@ -0,0 +1,2 @@ +test('T25304', [extra_files(['A.hs', 'B.hs'])], makefile_test, ['T25304']) +test('T25304a', [extra_files(['A.hs', 'B.hs'])], makefile_test, ['T25304a']) ===================================== testsuite/tests/showIface/DocsInHiFileTH.stdout ===================================== @@ -187,7 +187,7 @@ docs: avails: [i] avails: - [WD11{WD11, WD11Bool, WD11Int, WD11Foo}] + [WD11{WD11, WD11Bool, WD11Foo, WD11Int}] avails: [WD13{WD13}] avails: @@ -221,11 +221,11 @@ docs: avails: [Pretty{Pretty, prettyPrint}] avails: - [Corge{Corge, runCorge, Corge}] + [Corge{Corge, Corge, runCorge}] avails: - [Quuz{Quuz, quuz1_a, Quuz}] + [Quuz{Quuz, Quuz, quuz1_a}] avails: - [Quux{Quux, Quux2, Quux1}] + [Quux{Quux, Quux1, Quux2}] avails: [Tup2] avails: ===================================== testsuite/tests/showIface/NoExportList.stdout ===================================== @@ -32,7 +32,7 @@ docs: -- Actually we have only one type. identifiers: avails: - [R{R, fβ, fα, R}] + [R{R, R, fα, fβ}] section heading, level 1: text: -- * Functions ===================================== utils/haddock/haddock-api/src/Haddock/Interface/Create.hs ===================================== @@ -201,7 +201,7 @@ createInterface1' flags unit_state dflags hie_file mod_iface ifaces inst_ifaces -- See Note [Exporting built-in items] let builtinTys = DsiSectionHeading 1 (WithHsDocIdentifiers (mkGeneratedHsDocString "Builtin syntax") []) bonus_ds mods - | mdl == gHC_PRIM = [builtinTys, DsiExports funAvail] <> mods + | mdl == gHC_PRIM = [builtinTys, DsiExports (sortAvails funAvail)] <> mods | otherwise = mods let @@ -461,11 +461,11 @@ mkExportItems Just hsDoc' -> do doc <- processDocStringParas parserOpts sDocContext pkgName hsDoc' pure [ExportDoc doc] - DsiExports avails -> + DsiExports (SortedAvails avails) -> -- TODO: We probably don't need nubAvails here. -- mkDocStructureFromExportList already uses it. concat <$> traverse availExport (nubAvails avails) - DsiModExport mod_names avails -> do + DsiModExport mod_names (SortedAvails avails) -> do -- only consider exporting a module if we are sure we are really -- exporting the whole module and not some subset. (unrestricted_mods, remaining_avails) <- unrestrictedModExports sDocContext thisMod modMap instIfaceMap avails (NE.toList mod_names) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c18e7a48af27bdc698db3e604c6a2c521cd718c0 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/c18e7a48af27bdc698db3e604c6a2c521cd718c0 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 16:38:36 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 30 Sep 2024 12:38:36 -0400 Subject: [Git][ghc/ghc][wip/T25281] Make it right Message-ID: <66fad40c272bc_15ec304497ec577a8@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: f82443b9 by Simon Peyton Jones at 2024-09-30T17:38:00+01:00 Make it right - - - - - 12 changed files: - compiler/GHC/Core/LateCC/OverloadedCalls.hs - compiler/GHC/Core/Opt/Arity.hs - compiler/GHC/Core/Type.hs - compiler/GHC/Core/Utils.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Monad.hs - compiler/GHC/HsToCore/Pmc.hs - compiler/GHC/HsToCore/Types.hs - compiler/GHC/Tc/Gen/HsType.hs - compiler/GHC/Tc/TyCl/Instance.hs - compiler/GHC/Tc/Utils/Instantiate.hs - testsuite/tests/pmcheck/should_compile/T24891.hs Changes: ===================================== compiler/GHC/Core/LateCC/OverloadedCalls.hs ===================================== @@ -80,7 +80,7 @@ overloadedCallsCC = -- check if any of the arguments v1 ... vN are dictionaries. let (f, xs) = collectArgs app - resultTy = applyTypeToArgs empty (exprType f) xs + resultTy = applyTypeToArgs (exprType f) xs -- Recursively process the arguments first for no particular reason args <- mapM processExpr xs ===================================== compiler/GHC/Core/Opt/Arity.hs ===================================== @@ -2258,7 +2258,7 @@ etaInfoAppTy :: Type -> EtaInfo -> Type -- If e :: ty -- then etaInfoApp e eis :: etaInfoApp ty eis etaInfoAppTy ty (EI bs mco) - = applyTypeToArgs (text "etaInfoAppTy") ty1 (map varToCoreExpr bs) + = applyTypeToArgs ty1 (map varToCoreExpr bs) where ty1 = case mco of MRefl -> ty ===================================== compiler/GHC/Core/Type.hs ===================================== @@ -82,8 +82,7 @@ module GHC.Core.Type ( coAxNthLHS, stripCoercionTy, - splitInvisPiTys, splitInvisPiTysN, - invisibleTyBndrCount, + splitInvisPiTys, splitInvisPiTysN, invisibleBndrCount, filterOutInvisibleTypes, filterOutInferredTypes, partitionInvisibleTypes, partitionInvisibles, tyConForAllTyFlags, appTyForAllTyFlags, @@ -2097,12 +2096,12 @@ getRuntimeArgTys = go | otherwise = [] -invisibleTyBndrCount :: Type -> Int +invisibleBndrCount :: Type -> Int -- Returns the number of leading invisible forall'd binders in the type -- Includes invisible predicate arguments; e.g. for -- e.g. forall {k}. (k ~ *) => k -> k -- returns 2 not 1 -invisibleTyBndrCount ty = length (fst (splitInvisPiTys ty)) +invisibleBndrCount ty = length (fst (splitInvisPiTys ty)) -- | Like 'splitPiTys', but returns only *invisible* binders, including constraints. -- Stops at the first visible binder. ===================================== compiler/GHC/Core/Utils.hs ===================================== @@ -142,7 +142,7 @@ exprType (Tick _ e) = exprType e exprType (Lam binder expr) = mkLamType binder (exprType expr) exprType e@(App _ _) = case collectArgs e of - (fun, args) -> applyTypeToArgs (pprCoreExpr e) (exprType fun) args + (fun, args) -> applyTypeToArgs (exprType fun) args exprType (Type ty) = pprPanic "exprType" (ppr ty) coreAltType :: CoreAlt -> Type @@ -221,11 +221,10 @@ Various possibilities suggest themselves: Note that there might be existentially quantified coercion variables, too. -} -applyTypeToArgs :: HasDebugCallStack => SDoc -> Type -> [CoreExpr] -> Type +applyTypeToArgs :: HasDebugCallStack => Type -> [CoreExpr] -> Type -- ^ Determines the type resulting from applying an expression with given type --- to given argument expressions. --- The first argument is just for debugging, and gives some context -applyTypeToArgs pp_e op_ty args +applyTypeToArgs op_ty args = go op_ty args where go op_ty [] = op_ty @@ -244,8 +243,7 @@ applyTypeToArgs pp_e op_ty args go_ty_args op_ty rev_tys args = go (piResultTys op_ty (reverse rev_tys)) args - panic_msg as = vcat [ text "Expression:" <+> pp_e - , text "Type:" <+> ppr op_ty + panic_msg as = vcat [ text "Type:" <+> ppr op_ty , text "Args:" <+> ppr args , text "Args':" <+> ppr as ] ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -76,7 +76,6 @@ import GHC.Builtin.Names import GHC.Utils.Misc import GHC.Utils.Outputable as Outputable import GHC.Utils.Panic -import Control.Arrow( first ) import Control.Monad import Data.Maybe( isJust ) import qualified Data.Set as S @@ -695,10 +694,11 @@ ds_app (XExpr (ConLikeTc con tvs tys)) _hs_args core_args ; return (mkApps core_fun core_args) } ds_app (HsRecSel _ (FieldOcc { foExt = sel_id })) _hs_args core_args - = ds_app_rec_sel sel_id core_args + = ds_app_rec_sel sel_id sel_id core_args ds_app (HsVar _ lfun) hs_args core_args - = ds_app_var lfun hs_args core_args + = do { tracePm "ds_app" (ppr lfun <+> ppr core_args) + ; ds_app_var lfun hs_args core_args } ds_app e _hs_args core_args = do { core_e <- dsExpr e @@ -723,24 +723,27 @@ ds_app_var (L loc fun_id) hs_args core_args -- Alas, we cannot simply look at the unfolding of $dHasField below because it -- has not been set yet, so we have to reconstruct the selector Id from the types. | fun_id `hasKey` getFieldClassOpKey - , (_k : Type x_ty : Type r_ty : _a_ty : _dict : rest_args) <- core_args + , (_k : _rrep : _arep : Type x_ty : Type r_ty : _) <- core_args + -- getField :: forall {k} {r_rep} {a_rep} (x::k) (r :: TYPE r_rep) (a :: TYPE a_rep) . + -- HasField x r a => r -> a = do { fam_inst_envs <- dsGetFamInstEnvs ; rdr_env <- dsGetGlobalRdrEnv -- Look up the field named x/"sel" in the type r/T + ; tracePm "getfield" (ppr core_args $$ ppr x_ty $$ ppr r_ty) ; case lookupHasFieldLabel fam_inst_envs x_ty r_ty of Just fl | isJust (lookupGRE_FieldLabel rdr_env fl) -- isJust: Make sure the field is actually visible in this module; -- otherwise this might not be the implicit HasField instance -> do { sel_id <- dsLookupGlobalId (flSelector fl) - ; ds_app_rec_sel sel_id rest_args } + ; tracePm "getfield2" (ppr sel_id) + ; ds_app_rec_sel sel_id fun_id core_args } _ -> ds_app_finish fun_id core_args } ----------------------- -- Warn about identities for (fromInteger :: Integer -> Integer) etc -- They all have a type like: forall . => arg_ty -> res_ty | idName fun_id `elem` numericConversionNames - , let (inst_fun_ty, _) = apply_type_args fun_id core_args - , (_, conv_ty) <- splitInvisPiTys inst_fun_ty + , let (conv_ty, _) = apply_invis_args fun_id core_args , Just (arg_ty, res_ty) <- splitVisibleFunTy_maybe conv_ty = do { dflags <- getDynFlags ; when (wopt Opt_WarnIdentities dflags @@ -793,7 +796,7 @@ ds_app_finish :: Id -> [CoreExpr] -> DsM CoreExpr -- We are about to construct an application that may include evidence applications -- `f dict`. If the dictionary is non-specialisable, instead construct -- nospec f dict --- See Note [nospecId magic] in GHC.Types.Id.Make for what `nospec` does. +-- See Note [nospecId magic] gin GHC.Types.Id.Make for what `nospec` does. -- See Note [Desugaring non-canonical evidence] ds_app_finish fun_id core_args = do { unspecables <- getUnspecables @@ -812,60 +815,61 @@ ds_app_finish fun_id core_args ; return (mkCoreApps fun core_args) } --------------- -ds_app_rec_sel :: Id -> [CoreExpr] -> DsM CoreExpr +ds_app_rec_sel :: Id -- The record selector Id itself + -> Id -- The function at the the head + -> [CoreExpr] -- Its arguments + -> DsM CoreExpr -- Desugar an application with HsRecSelId at the head -ds_app_rec_sel sel_id core_args +ds_app_rec_sel sel_id fun_id core_args | RecSelId{ sel_cons = rec_sel_info } <- idDetails sel_id , RSI { rsi_undef = cons_wo_field } <- rec_sel_info = do { -- Record selectors are warned about if they are not present in all of the - -- parent data type's constructor, or always in case of pattern synonym record + -- parent data type's constructors, or always in case of pattern synonym record -- selectors (regulated by a flag). However, this only produces a warning if -- it's not a part of a record selector application. For example: -- data T = T1 | T2 {s :: Bool} -- g y = map s y -- Warn here -- f x = s x -- No warning here - ; let (fun_ty, val_args) = apply_type_args sel_id core_args + ; let (fun_ty, val_args) = apply_invis_args fun_id core_args - -- Type-based check. - -- See (3) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + ; tracePm "ds_app_rec_sel" (ppr fun_ty $$ ppr val_args) ; case val_args of -- There is a value argument + -- See (IRS2) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc (arg:_) -> pmcRecSel sel_id arg -- No value argument, but the selector is -- applied to all its type arguments + -- See (IRS3) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc [] | Just (val_arg_ty, _) <- splitVisibleFunTy_maybe fun_ty -> do { dummy <- newSysLocalDs (Scaled ManyTy val_arg_ty) ; pmcRecSel sel_id (Var dummy) } -- Not even applied to all its type args - -- See (4) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc + -- See (IRS4) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc _ -> unless (null cons_wo_field) $ do { dflags <- getDynFlags ; let maxCons = maxUncoveredPatterns dflags ; diagnosticDs $ DsIncompleteRecordSelector (idName sel_id) cons_wo_field maxCons } - ; ds_app_finish sel_id core_args } + ; ds_app_finish fun_id core_args } | otherwise = pprPanic "ds_app_rec_sel" (ppr sel_id $$ ppr (idDetails sel_id)) where -apply_type_args :: Id -> [CoreExpr] -> (Type, [CoreExpr]) +apply_invis_args :: Id -> [CoreExpr] -> (Type, [CoreExpr]) -- Apply function to the initial /type/ args; -- return the type of the instantiated function, -- and the remaining args -- e.g. apply_type_args (++) [Type Int, Var xs] -- = ([Int] -> [Int] -> [Int], [Var xs]) -apply_type_args fun args - = (piResultTys (idType fun) arg_tys, rest_args) +apply_invis_args fun_id args + = (applyTypeToArgs fun_ty invis_args, rest_args) where - (arg_tys, rest_args) = go args - - go :: [CoreExpr] -> ([Type], [CoreExpr]) - go (Type ty : args) = first (ty :) (go args) - go args = ([], args) + fun_ty = idType fun_id + (invis_args, rest_args) = splitAt (invisibleBndrCount fun_ty) args ------------------------------ splitHsWrapperArgs :: HsWrapper -> [CoreArg] -> DsM (HsWrapper, [CoreArg]) ===================================== compiler/GHC/HsToCore/Monad.hs ===================================== @@ -36,9 +36,6 @@ module GHC.HsToCore.Monad ( -- Getting and setting pattern match oracle states getPmNablas, updPmNablas, - -- Locally suppress -Wincomplete-record-selectors warnings - getSuppressIncompleteRecSelsDs, suppressIncompleteRecSelsDs, - -- Tracking evidence variable coherence addUnspecables, getUnspecables, @@ -409,7 +406,6 @@ mkDsEnvs unit_env mod rdr_env type_env fam_inst_env ptc msg_var cc_st_var , dsl_loc = real_span , dsl_nablas = initNablas , dsl_unspecables = mempty - , dsl_suppress_incomplete_rec_sel = False } in (gbl_env, lcl_env) @@ -478,13 +474,6 @@ addUnspecables unspecables = updLclEnv (\env -> env{ dsl_unspecables = unspecabl getUnspecables :: DsM (S.Set EvId) getUnspecables = dsl_unspecables <$> getLclEnv -suppressIncompleteRecSelsDs :: DsM a -> DsM a -suppressIncompleteRecSelsDs = updLclEnv (\dsl -> dsl { dsl_suppress_incomplete_rec_sel = True }) - --- | Get the current pattern match oracle state. See 'dsl_nablas'. -getSuppressIncompleteRecSelsDs :: DsM Bool -getSuppressIncompleteRecSelsDs = do { env <- getLclEnv; return (dsl_suppress_incomplete_rec_sel env) } - getSrcSpanDs :: DsM SrcSpan getSrcSpanDs = do { env <- getLclEnv ; return (RealSrcSpan (dsl_loc env) Strict.Nothing) } ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -272,7 +272,7 @@ We should warn in `solveDot`, but not in `accessDot`. Here is how we achieve all this in the implementation: - 1. When renaming a record selector in `mkOneRecordSelector`, +(IRS1) When renaming a record selector in `mkOneRecordSelector`, we precompute the constructors the selector succeeds on. That would be `T2` for `sel` because `sel (T2 42)` succeeds, and `Yes` for `sel2` because `sel2 (Yes 13)` succeeds. @@ -286,7 +286,7 @@ They are ordered by specificity, so we prefer (2) over (3) over (4). Item (5) below describes how we resolve the overlap. (-XOverloadedRecordDot is discussed separately in Item (6) and (7).) - 2. In function `ldi`, we have a record selector application `sel arg`. +(IRS2) In function `ldi`, we have a record selector application `sel arg`. This situation is detected `GHC.HsToCore.Expr.ds_app_rec_sel`, when the record selector is applied to at least one argument. We call out to the pattern-match checker to determine whether use of the selector is safe, @@ -311,14 +311,14 @@ Item (5) below describes how we resolve the overlap. reduced problem, the match is exhaustive and the use of the record selector safe. - 3. In function `resTy`, the record selector is unsaturated, but the result type +(IRS3) In function `resTy`, the record selector is unsaturated, but the result type ensures a safe use of the selector. This situation is also detected in `GHC.HsToCore.Expr.ds_app_rec_sel`. THe selector is elaborated with its type arguments; we simply match on desugared Core `sel @Bool :: T Bool -> Int` to learn the result type `T Bool`. We again call `pmcRecSel`, but this time with a fresh dummy Id `ds::T Bool`. - 4. In case of an unsaturated record selector that is *not* applied to any type +(IRS4) In case of an unsaturated record selector that is *not* applied to any type argument after elaboration (e.g. in `urgh2 = sel2 :: Dot -> Int`), we simply produce a warning about all `sel_cons`; no need to call `pmcRecSel`. This happens in the `HsRecSel` case of `dsExpr`. @@ -382,7 +382,7 @@ Finally, there are 2 more items addressing -XOverloadedRecordDot: pmcRecSel :: Id -- ^ Id of the selector -> CoreExpr -- ^ Core expression of the argument to the selector -> DsM () --- See (IRS4) in Note [Detecting incomplete record selectors] +-- See (IRS2,3) in Note [Detecting incomplete record selectors] pmcRecSel sel_id arg | RecSelId{ sel_cons = rec_sel_info } <- idDetails sel_id , RSI { rsi_def = cons_w_field, rsi_undef = cons_wo_field } <- rec_sel_info ===================================== compiler/GHC/HsToCore/Types.hs ===================================== @@ -80,9 +80,6 @@ data DsLclEnv -- ^ See Note [Long-distance information] in "GHC.HsToCore.Pmc". -- The set of reaching values Nablas is augmented as we walk inwards, refined -- through each pattern match in turn - , dsl_suppress_incomplete_rec_sel :: Bool - -- ^ Whether to suppress -Wincomplete-record-selectors warnings. - -- See (5) of Note [Detecting incomplete record selectors] , dsl_unspecables :: S.Set EvVar -- ^ See Note [Desugaring non-canonical evidence]: this field collects -- all un-specialisable evidence variables in scope. ===================================== compiler/GHC/Tc/Gen/HsType.hs ===================================== @@ -1982,11 +1982,11 @@ checkExpectedKind hs_ty ty act_kind exp_kind ; return (res_ty `mkCastTy` co_k) } } where -- We need to make sure that both kinds have the same number of implicit - -- foralls out front. If the actual kind has more, instantiate accordingly. - -- Otherwise, just pass the type & kind through: the errors are caught - -- in unifyType. - n_exp_invis_bndrs = invisibleTyBndrCount exp_kind - n_act_invis_bndrs = invisibleTyBndrCount act_kind + -- foralls and constraints out front. If the actual kind has more, instantiate + -- accordingly. Otherwise, just pass the type & kind through: the errors + -- are caught in unifyType. + n_exp_invis_bndrs = invisibleBndrCount exp_kind + n_act_invis_bndrs = invisibleBndrCount act_kind n_to_inst = n_act_invis_bndrs - n_exp_invis_bndrs @@ -2707,8 +2707,8 @@ checkExpectedResKind sig_kind res_ki ; let -- Drop invisible binders from sig_kind until they match up -- with res_ki. By analogy with checkExpectedKind. - n_res_invis_bndrs = invisibleTyBndrCount actual_res_ki - n_sig_invis_bndrs = invisibleTyBndrCount sig_kind + n_res_invis_bndrs = invisibleBndrCount actual_res_ki + n_sig_invis_bndrs = invisibleBndrCount sig_kind n_to_inst = n_sig_invis_bndrs - n_res_invis_bndrs (_, sig_kind') = splitInvisPiTysN n_to_inst sig_kind ===================================== compiler/GHC/Tc/TyCl/Instance.hs ===================================== @@ -507,7 +507,7 @@ tcClsInstDecl (L loc (ClsInstDecl { cid_ext = lwarn fst $ splitForAllForAllTyBinders dfun_ty visible_skol_tvs = drop n_inferred skol_tvs - ; traceTc "tcLocalInstDecl 1" (ppr dfun_ty $$ ppr (invisibleTyBndrCount dfun_ty) $$ ppr skol_tvs) + ; traceTc "tcLocalInstDecl 1" (ppr dfun_ty $$ ppr (invisibleBndrCount dfun_ty) $$ ppr skol_tvs) -- Next, process any associated types. ; (datafam_stuff, tyfam_insts) ===================================== compiler/GHC/Tc/Utils/Instantiate.hs ===================================== @@ -489,7 +489,11 @@ tcInstInvisibleTyBinders ty kind = do { (extra_args, kind') <- tcInstInvisibleTyBindersN n_invis kind ; return (mkAppTys ty extra_args, kind') } where - n_invis = invisibleTyBndrCount kind + n_invis = invisibleBndrCount kind + -- We are re-using tcInstInvisibleTyBindersN, which is + -- needed elsewhere; so all that matters is that n_invis + -- is big enough! Does not matter if it is too big. + -- 10,000 would do equally well :-) tcInstInvisibleTyBindersN :: Int -> TcKind -> TcM ([TcType], TcKind) -- Called only to instantiate kinds, in user-written type signatures ===================================== testsuite/tests/pmcheck/should_compile/T24891.hs ===================================== @@ -12,6 +12,7 @@ data T a where f :: T Bool -> Int f x = x.sel -- warn, but only once, suggesting to match on T3 +{- data Dot = No | Yes {sel2 :: Int} ldiDot :: Dot -> Int @@ -28,3 +29,4 @@ data Dot2 t = No2 | Yes2 {sel3 :: t} accessDot2 :: HasField "sel2" t Int => Dot2 t -> Int accessDot2 x = x.sel3.sel2 -- warn about x.sel3 +-} View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f82443b9fec721308f3f46e7beb3a1346c91b57a -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/f82443b9fec721308f3f46e7beb3a1346c91b57a You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 16:45:26 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 30 Sep 2024 12:45:26 -0400 Subject: [Git][ghc/ghc][wip/T25281] Tracing Message-ID: <66fad5a666991_15ec304d0530583a9@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: 40216c67 by Simon Peyton Jones at 2024-09-30T17:45:15+01:00 Tracing - - - - - 1 changed file: - compiler/GHC/HsToCore/Pmc.hs Changes: ===================================== compiler/GHC/HsToCore/Pmc.hs ===================================== @@ -408,6 +408,7 @@ pmcRecSel sel_id arg unc_examples <- getNFirstUncovered MinimalCover [arg_id] (maxPatterns + 1) uncov_nablas let cons = [con | unc_example <- unc_examples , Just (PACA (PmAltConLike con) _ _) <- [lookupSolution unc_example arg_id]] + tracePm "unc-ex" (ppr cons $$ ppr unc_examples) diagnosticDs $ DsIncompleteRecordSelector sel_name cons maxPatterns View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/40216c67a5c816765b9c0f2fe102ee72846427a4 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/40216c67a5c816765b9c0f2fe102ee72846427a4 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 17:57:13 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 30 Sep 2024 13:57:13 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/package-deps-bytecode-squashed] Link interface bytecode from package DBs if possible Message-ID: <66fae67936b2c_15ec30846a70653de@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/package-deps-bytecode-squashed at Glasgow Haskell Compiler / GHC Commits: 8a72a482 by Torsten Schmits at 2024-09-30T19:56:57+02:00 Link interface bytecode from package DBs if possible Part of #25090. MR !????? When splices are executed with `-fprefer-byte-code`, the loader will compile Core bindings to bytecode if those are present in interfaces of module dependencies. So far, this only applied to local modules (i.e. home modules in make mode and non-package deps in oneshot mode). This patch extends support to interfaces loaded from a package DB. In `getLinkDeps`, the dependencies chosen for recursion were restricted to `dep_direct_mods`, which has been changed to include `mi_usages`. In order to unify treatment of the different link variants across make/oneshot mode, the algorithm's intermediate results have been abstracted into the data types `LinkDep` and `LinkModule`. At the moment, following package deps is only implemented for oneshot mode. - - - - - 19 changed files: - compiler/GHC/Iface/Load.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/Dep.hs - + testsuite/tests/bytecode/T25090/DepApi.hs - + testsuite/tests/bytecode/T25090/Local.hs - testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/Num.hs - + testsuite/tests/bytecode/T25090/Num.hs-boot - + testsuite/tests/bytecode/T25090/PkgBytecode.hs - + testsuite/tests/bytecode/T25090/PkgBytecode.stdout - testsuite/tests/bytecode/T25090/all.T - + testsuite/tests/bytecode/T25090/dep.conf - + testsuite/tests/bytecode/T25090/prep.bash - + testsuite/tests/bytecode/T25090/run.bash - + testsuite/tests/bytecode/T25090/unit1 - + testsuite/tests/bytecode/T25090/unit2 Changes: ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -672,7 +672,11 @@ dontLeakTheHUG thing_inside = do in hsc_env { hsc_targets = panic "cleanTopEnv: hsc_targets" , hsc_mod_graph = panic "cleanTopEnv: hsc_mod_graph" - , hsc_IC = panic "cleanTopEnv: hsc_IC" + -- TODO this is needed for bytecode compilation of package deps + -- only. local EPS modules work fine. + -- Also it appears to work when the local modules use multiple + -- home units?!?!? + -- , hsc_IC = panic "cleanTopEnv: hsc_IC" , hsc_type_env_vars = case maybe_type_vars of Just vars -> vars Nothing -> panic "cleanTopEnv: hsc_type_env_vars" ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -6,6 +6,7 @@ {-# LANGUAGE CPP #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE NamedFieldPuns #-} module GHC.Linker.Deps ( LinkDepsOpts (..) @@ -28,6 +29,7 @@ import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM import GHC.Utils.Outputable +import qualified GHC.Utils.Outputable as Outputable import GHC.Utils.Panic import GHC.Utils.Error @@ -47,9 +49,11 @@ import GHC.Utils.Misc import GHC.Unit.Home import GHC.Data.Maybe -import Control.Monad import Control.Applicative +import Control.Monad.IO.Class (MonadIO (liftIO)) +import Control.Monad.Trans.Except (ExceptT, runExceptT, throwE) +import Data.Foldable (traverse_) import qualified Data.Set as Set import qualified Data.Map as M import Data.List (isSuffixOf) @@ -68,15 +72,16 @@ data LinkDepsOpts = LinkDepsOpts , ldWays :: !Ways -- ^ Enabled ways , ldFinderCache :: !FinderCache , ldFinderOpts :: !FinderOpts - , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) - , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation))) + , ldLoadByteCode :: !(Module -> IO (Maybe (IO Linkable))) + , ldDebugTrace :: !(SDoc -> IO ()) } data LinkDeps = LinkDeps { ldNeededLinkables :: [Linkable] , ldAllLinkables :: [Linkable] - , ldUnits :: [UnitId] - , ldNeededUnits :: UniqDSet UnitId + , ldNeededUnits :: [UnitId] + , ldAllUnits :: UniqDSet UnitId } -- | Find all the packages and linkables that a set of modules depends on @@ -102,7 +107,14 @@ getLinkDeps opts interp pls span mods = do get_link_deps opts pls maybe_normal_osuf span mods - +-- | Compute the linkables for the given module set's dependencies. +-- +-- Home modules in make mode are treated separately in a preprocessing step, +-- then all the remaining external deps are processed for both modes. +-- If bytecode is available, transitive external deps are included, otherwise +-- the module's library is linked and processing stops. +-- +-- The results are split into sets of needed/loaded modules/packages. get_link_deps :: LinkDepsOpts -> LoaderState @@ -111,46 +123,41 @@ get_link_deps -> [Module] -> IO LinkDeps get_link_deps opts pls maybe_normal_osuf span mods = do - -- 1. Find the dependent home-pkg-modules/packages from each iface - -- (omitting modules from the interactive package, which is already linked) - (mods_s, pkgs_s) <- - -- Why two code paths here? There is a significant amount of repeated work - -- performed calculating transitive dependencies - -- if --make uses the oneShot code path (see MultiLayerModulesTH_* tests) - if ldOneShotMode opts - then follow_deps (filterOut isInteractiveModule mods) - emptyUniqDSet emptyUniqDSet; - else do - (pkgs, mmods) <- unzip <$> mapM get_mod_info all_home_mods - return (catMaybes mmods, unionManyUniqDSets (init_pkg_set : pkgs)) - - let - -- 2. Exclude ones already linked - -- Main reason: avoid findModule calls in get_linkable - (mods_needed, links_got) = partitionWith split_mods mods_s - pkgs_needed = eltsUDFM $ getUniqDSet pkgs_s `minusUDFM` pkgs_loaded pls - - split_mods mod = - let is_linked = lookupModuleEnv (objs_loaded pls) mod - <|> lookupModuleEnv (bcos_loaded pls) mod - in case is_linked of - Just linkable -> Right linkable - Nothing -> Left mod - - -- 3. For each dependent module, find its linkable - -- This will either be in the HPT or (in the case of one-shot - -- compilation) we may need to use maybe_getFileLinkable - lnks_needed <- mapM (get_linkable (ldObjSuffix opts)) mods_needed - - return $ LinkDeps - { ldNeededLinkables = lnks_needed - , ldAllLinkables = links_got ++ lnks_needed - , ldUnits = pkgs_needed - , ldNeededUnits = pkgs_s - } + (link_deps_home, module_deps_external) <- separate_home_deps + link_deps_external <- external_deps opts module_deps_external + let (loaded_modules, needed_modules, ldAllUnits, ldNeededUnits) = + classify_deps pls (link_deps_home ++ link_deps_external) + ldNeededLinkables <- mapM module_linkable needed_modules + pure LinkDeps { + ldNeededLinkables, + ldAllLinkables = loaded_modules ++ ldNeededLinkables, + ldNeededUnits, + ldAllUnits + } where mod_graph = ldModuleGraph opts unit_env = ldUnitEnv opts + noninteractive = filterOut isInteractiveModule mods + + -- | Preprocess the dependencies in make mode to remove all home modules, + -- since the transitive dependency closure is already cached for those in + -- the HUG (see MultiLayerModulesTH_* tests for the performance impact). + -- + -- Returns the remaining, external, dependencies on the right, which is the + -- entire set for oneshot mode. + separate_home_deps = + if ldOneShotMode opts + then pure ([], noninteractive) + else make_deps + + make_deps = do + (dep_ext, mmods) <- unzip <$> mapM get_mod_info all_home_mods + let + link_mods = + listToUDFM [(moduleName (mi_module (hm_iface m)), m) | m <- mmods] + ext = uniqDSetToList (unionManyUniqDSets (init_ext : dep_ext)) + -- TODO extract HML here already + pure ([LinkModules (LinkHomeModule <$> link_mods)], ext) -- This code is used in `--make` mode to calculate the home package and unit dependencies -- for a set of modules. @@ -160,9 +167,9 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- It is also a matter of correctness to use the module graph so that dependencies between home units -- is resolved correctly. - make_deps_loop :: (UniqDSet UnitId, Set.Set NodeKey) -> [ModNodeKeyWithUid] -> (UniqDSet UnitId, Set.Set NodeKey) + make_deps_loop :: (UniqDSet Module, Set.Set NodeKey) -> [ModNodeKeyWithUid] -> (UniqDSet Module, Set.Set NodeKey) make_deps_loop found [] = found - make_deps_loop found@(found_units, found_mods) (nk:nexts) + make_deps_loop found@(external, found_mods) (nk:nexts) | NodeKey_Module nk `Set.member` found_mods = make_deps_loop found nexts | otherwise = case M.lookup (NodeKey_Module nk) (mgTransDeps mod_graph) of @@ -171,85 +178,26 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- See #936 and the ghci.prog007 test for why we have to continue traversing through -- boot modules. todo_boot_mods = [ModNodeKeyWithUid (GWIB mn NotBoot) uid | NodeKey_Module (ModNodeKeyWithUid (GWIB mn IsBoot) uid) <- Set.toList trans_deps] - in make_deps_loop (found_units, deps `Set.union` found_mods) (todo_boot_mods ++ nexts) + in make_deps_loop (external, deps `Set.union` found_mods) (todo_boot_mods ++ nexts) Nothing -> - let (ModNodeKeyWithUid _ uid) = nk - in make_deps_loop (addOneToUniqDSet found_units uid, found_mods) nexts + let (ModNodeKeyWithUid (GWIB mod_name _) uid) = nk + in make_deps_loop (addOneToUniqDSet external (Module (RealUnit (Definite uid)) mod_name), found_mods) nexts mkNk m = ModNodeKeyWithUid (GWIB (moduleName m) NotBoot) (moduleUnitId m) - (init_pkg_set, all_deps) = make_deps_loop (emptyUniqDSet, Set.empty) $ map mkNk (filterOut isInteractiveModule mods) + (init_ext, all_deps) = make_deps_loop (emptyUniqDSet, Set.empty) $ map mkNk noninteractive all_home_mods = [with_uid | NodeKey_Module with_uid <- Set.toList all_deps] get_mod_info (ModNodeKeyWithUid gwib uid) = case lookupHug (ue_home_unit_graph unit_env) uid (gwib_mod gwib) of - Just hmi -> - let iface = (hm_iface hmi) - mmod = case mi_hsc_src iface of - HsBootFile -> link_boot_mod_error (mi_module iface) - _ -> return $ Just (mi_module iface) - - in (mkUniqDSet $ Set.toList $ dep_direct_pkgs (mi_deps iface),) <$> mmod + Just hmi -> do + let iface = hm_iface hmi + case mi_hsc_src iface of + HsBootFile -> throwProgramError opts $ link_boot_mod_error (mi_module iface) + _ -> pure (mkUniqDSet $ [usg_mod | UsagePackageModule {usg_mod} <- mi_usages iface], hmi) Nothing -> throwProgramError opts $ text "getLinkDeps: Home module not loaded" <+> ppr (gwib_mod gwib) <+> ppr uid - - -- This code is used in one-shot mode to traverse downwards through the HPT - -- to find all link dependencies. - -- The ModIface contains the transitive closure of the module dependencies - -- within the current package, *except* for boot modules: if we encounter - -- a boot module, we have to find its real interface and discover the - -- dependencies of that. Hence we need to traverse the dependency - -- tree recursively. See bug #936, testcase ghci/prog007. - follow_deps :: [Module] -- modules to follow - -> UniqDSet Module -- accum. module dependencies - -> UniqDSet UnitId -- accum. package dependencies - -> IO ([Module], UniqDSet UnitId) -- result - follow_deps [] acc_mods acc_pkgs - = return (uniqDSetToList acc_mods, acc_pkgs) - follow_deps (mod:mods) acc_mods acc_pkgs - = do - mb_iface <- ldLoadIface opts msg mod - iface <- case mb_iface of - Failed err -> throwProgramError opts $ - missingInterfaceErrorDiagnostic (ldMsgOpts opts) err - Succeeded iface -> return iface - - when (mi_boot iface == IsBoot) $ link_boot_mod_error mod - - let - pkg = moduleUnit mod - deps = mi_deps iface - - pkg_deps = dep_direct_pkgs deps - (boot_deps, mod_deps) = flip partitionWith (Set.toList (dep_direct_mods deps)) $ - \case - (_, GWIB m IsBoot) -> Left m - (_, GWIB m NotBoot) -> Right m - - mod_deps' = case ue_homeUnit unit_env of - Nothing -> [] - Just home_unit -> filter (not . (`elementOfUniqDSet` acc_mods)) (map (mkHomeModule home_unit) $ (boot_deps ++ mod_deps)) - acc_mods' = case ue_homeUnit unit_env of - Nothing -> acc_mods - Just home_unit -> addListToUniqDSet acc_mods (mod : map (mkHomeModule home_unit) mod_deps) - acc_pkgs' = addListToUniqDSet acc_pkgs (Set.toList pkg_deps) - - case ue_homeUnit unit_env of - Just home_unit | isHomeUnit home_unit pkg -> follow_deps (mod_deps' ++ mods) - acc_mods' acc_pkgs' - _ -> follow_deps mods acc_mods (addOneToUniqDSet acc_pkgs' (toUnitId pkg)) - where - msg = text "need to link module" <+> ppr mod <+> - text "due to use of Template Haskell" - - - - link_boot_mod_error :: Module -> IO a - link_boot_mod_error mod = throwProgramError opts $ - text "module" <+> ppr mod <+> - text "cannot be linked; it is only available as a boot module" - no_obj :: Outputable a => a -> IO b no_obj mod = dieWith opts span $ text "cannot find object file for module " <> @@ -258,6 +206,20 @@ get_link_deps opts pls maybe_normal_osuf span mods = do while_linking_expr = text "while linking an interpreted expression" + -- | Extract the 'Linkable's for unlinked modules from the intermediate + -- results. + module_linkable = \case + LinkHomeModule hmi -> + adjust_linkable (expectJust "getLinkDeps" (homeModLinkable hmi)) + + LinkObjectModule iface loc -> do + let mod = mi_module iface + findObjectLinkableMaybe mod loc >>= \case + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + + LinkByteCodeModule _ load_bytecode -> + load_bytecode -- See Note [Using Byte Code rather than Object Code for Template Haskell] homeModLinkable :: HomeModInfo -> Maybe Linkable @@ -266,57 +228,257 @@ get_link_deps opts pls maybe_normal_osuf span mods = do then homeModInfoByteCode hmi <|> homeModInfoObject hmi else homeModInfoObject hmi <|> homeModInfoByteCode hmi - get_linkable osuf mod -- A home-package module - | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) - = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) - | otherwise - = do -- It's not in the HPT because we are in one shot mode, - -- so use the Finder to get a ModLocation... - case ue_homeUnit unit_env of - Nothing -> no_obj mod - Just home_unit -> do - from_bc <- ldLoadByteCode opts mod - maybe (fallback_no_bytecode home_unit mod) pure from_bc - where - - fallback_no_bytecode home_unit mod = do - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc _ -> do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - _ -> no_obj (moduleName mod) - - adjust_linkable lnk - | Just new_osuf <- maybe_normal_osuf = do - new_parts <- mapM (adjust_part new_osuf) - (linkableParts lnk) - return lnk{ linkableParts=new_parts } - | otherwise = - return lnk - - adjust_part new_osuf part = case part of - DotO file ModuleObject -> do - massert (osuf `isSuffixOf` file) - let file_base = fromJust (stripExtension osuf file) - new_file = file_base <.> new_osuf - ok <- doesFileExist new_file - if (not ok) - then dieWith opts span $ - text "cannot find object file " - <> quotes (text new_file) $$ while_linking_expr - else return (DotO new_file ModuleObject) - DotO file ForeignObject -> pure (DotO file ForeignObject) - DotA fp -> panic ("adjust_ul DotA " ++ show fp) - DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) - BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) + adjust_linkable lnk + | Just new_osuf <- maybe_normal_osuf = do + new_uls <- mapM (adjust_part (ldObjSuffix opts) new_osuf) + (linkableParts lnk) + return lnk {linkableParts = new_uls} + | otherwise = + return lnk + + adjust_part osuf new_osuf part = case part of + DotO file ModuleObject -> do + massert (osuf `isSuffixOf` file) + let file_base = fromJust (stripExtension osuf file) + new_file = file_base <.> new_osuf + ok <- doesFileExist new_file + if (not ok) + then dieWith opts span $ + text "cannot find object file " + <> quotes (text new_file) $$ while_linking_expr + else return (DotO new_file ModuleObject) + DotO file ForeignObject -> pure (DotO file ForeignObject) + DotA fp -> panic ("adjust_part DotA " ++ show fp) + DotDLL fp -> panic ("adjust_part DotDLL " ++ show fp) + BCOs {} -> pure part + LazyBCOs{} -> pure part + CoreBindings WholeCoreBindings {wcb_module} -> + pprPanic "Unhydrated core bindings" (ppr wcb_module) + +data LinkModule = + LinkHomeModule !HomeModInfo + | + LinkObjectModule !ModIface !ModLocation + | + LinkByteCodeModule !ModIface !(IO Linkable) + +link_module_iface :: LinkModule -> ModIface +link_module_iface = \case + LinkHomeModule hmi -> hm_iface hmi + LinkObjectModule iface _ -> iface + LinkByteCodeModule iface _ -> iface + +instance Outputable LinkModule where + ppr = \case + LinkHomeModule hmi -> ppr (mi_module (hm_iface hmi)) <+> brackets (text "HMI") + LinkObjectModule iface _ -> ppr (mi_module iface) + LinkByteCodeModule iface _ -> ppr (mi_module iface) <+> brackets (text "BC") + +data LinkDep = + LinkModules !(UniqDFM ModuleName LinkModule) + | + LinkLibrary !UnitId + +instance Outputable LinkDep where + ppr = \case + LinkModules mods -> text "modules:" <+> ppr (eltsUDFM mods) + LinkLibrary uid -> text "library:" <+> ppr uid + +data OneshotError = + NoLocation !Module + | + NoInterface !MissingInterfaceError + | + LinkBootModule !Module + +-- Compute the transitive dependency closure of the given modules. +-- +-- Used for all oneshot mode dependencies and for external dependencies of home +-- modules in make mode. +-- +-- TODO is the following still relevant? +-- The ModIface contains the transitive closure of the module dependencies +-- within the current package, *except* for boot modules: if we encounter +-- a boot module, we have to find its real interface and discover the +-- dependencies of that. Hence we need to traverse the dependency +-- tree recursively. See bug #936, testcase ghci/prog007. +external_deps :: + LinkDepsOpts -> + -- | Modules whose imports to follow + [Module] -> + IO [LinkDep] +external_deps opts mods = + runExceptT (external_deps_loop opts mods emptyUDFM) >>= \case + Right a -> pure (eltsUDFM a) + Left err -> throwProgramError opts (message err) + where + message = \case + NoLocation mod -> + pprPanic "found iface but no location" (ppr mod) + NoInterface err -> + missingInterfaceErrorDiagnostic (ldMsgOpts opts) err + LinkBootModule mod -> + link_boot_mod_error mod + +external_deps_loop :: + LinkDepsOpts -> + [Module] -> + UniqDFM UnitId LinkDep -> + ExceptT OneshotError IO (UniqDFM UnitId LinkDep) +external_deps_loop _ [] acc = + pure acc +external_deps_loop opts (mod : mods) acc = do + (new_acc, new_mods, action) <- process_module + traverse_ debug_log action + external_deps_loop opts (new_mods ++ mods) new_acc + where + debug_log action = + liftIO $ ldDebugTrace opts $ + text "TH dep" <+> ppr mod <+> brackets (sep [ + if is_home then text "home" else Outputable.empty, + text action + ]) + + -- | Decide how this module needs to be processed. + -- We only need an interface if we want to load bytecode or if we have to + -- link an object file (which happens for home unit modules, since those + -- have no libraries). + process_module + | already_seen = pure (acc, [], Nothing) + | is_home || prefer_bytecode = try_iface + | otherwise = add_library + + -- | Check whether the current module was processed before. + -- Since the accumulator is keyed by unit ID, we have to perform two + -- lookups. + -- If another module from this module's unit has been determined to be + -- linked as a library previously, we skip this module, assuming that no + -- bytecode is available for the entire package. + already_seen + | Just (LinkModules mods) <- mod_dep + = elemUDFM mod_name mods + | Just (LinkLibrary _) <- mod_dep + = True + | otherwise + = False + + -- | Load the iface and attempt to get bytecode from Core bindings. + try_iface = + liftIO (ldLoadIface opts load_reason mod) >>= \case + Failed err -> throwE (NoInterface err) + Succeeded (iface, loc) -> do + mb_load_bc <- liftIO (ldLoadByteCode opts (mi_module iface)) + with_iface loc iface mb_load_bc + + -- | Decide how to link this module. + -- If bytecode or an object file is available, use those in that order. + -- Otherwise fall back to linking a library. + with_iface loc iface mb_load_bc + | IsBoot <- mi_boot iface + = throwE (LinkBootModule mod) + + | prefer_bytecode + , Just load_bc <- mb_load_bc + = pure (add_module iface (LinkByteCodeModule iface load_bc) "bytecode") + + | is_home + = pure (add_module iface (LinkObjectModule iface loc) "object") + + | otherwise + = add_library + + add_library = + pure (addToUDFM acc mod_unit_id (LinkLibrary mod_unit_id), [], Just "library") + + add_module iface lmod action = + (addListToUDFM with_mod (direct_pkgs iface), new_deps iface, Just action) + where + with_mod = alterUDFM (add_package_module lmod) acc mod_unit_id + + add_package_module lmod = \case + Just (LinkLibrary u) -> Just (LinkLibrary u) + Just (LinkModules old) -> Just (LinkModules (addToUDFM old mod_name lmod)) + Nothing -> Just (LinkModules (unitUDFM mod_name lmod)) + + direct_pkgs iface + | prefer_bytecode + = [] + | otherwise + = [(u, LinkLibrary u) | u <- Set.toList (dep_direct_pkgs (mi_deps iface))] + + new_deps iface + | prefer_bytecode + -- TODO How can we better determine the external deps? + -- OTOH, we probably don't want to link unused dependencies anyway. + = [usg_mod | UsagePackageModule {usg_mod} <- mi_usages iface] ++ local + | is_home + = local + | otherwise + = [] + where + local = + [ + mkModule mod_unit m + -- TODO Somehow this just works, no idea what the deal was in the + -- old code with boot modules. + | (_, GWIB m _) <- Set.toList (dep_direct_mods (mi_deps iface)) + ] + + is_home + | Just home <- mb_home + = homeUnitAsUnit home == mod_unit + | otherwise + = False + + mod_dep = lookupUDFM acc mod_unit_id + mod_name = moduleName mod + mod_unit_id = moduleUnitId mod + mod_unit = moduleUnit mod + load_reason = + text "need to link module" <+> ppr mod <+> + text "due to use of Template Haskell" + + prefer_bytecode = ldUseByteCode opts + mb_home = ue_homeUnit (ldUnitEnv opts) + +link_boot_mod_error :: Module -> SDoc +link_boot_mod_error mod = + text "module" <+> ppr mod <+> + text "cannot be linked; it is only available as a boot module" + +-- | Split link dependencies into the sets of modules and packages that have +-- been linked previously and those that need to be linked now by checking for +-- their presence in the 'LoaderState': +-- +-- - For module dependencies, in the sets of loaded objects and BCOs +-- ('objs_loaded' and 'bcos_loaded') +-- - For package dependencies, in the set of loaded packages ('pkgs_loaded') +classify_deps :: + LoaderState -> + [LinkDep] -> + ([Linkable], [LinkModule], UniqDSet UnitId, [UnitId]) +classify_deps pls deps = + (loaded_modules, needed_modules, all_packages, needed_packages) + where + (loaded_modules, needed_modules) = + partitionWith loaded_or_needed_module (concatMap eltsUDFM modules) + + needed_packages = + eltsUDFM (getUniqDSet all_packages `minusUDFM` pkgs_loaded pls) + + all_packages = mkUniqDSet packages + + (modules, packages) = flip partitionWith deps $ \case + LinkModules mods -> Left mods + LinkLibrary lib -> Right lib + + loaded_or_needed_module lm = + maybe (Right lm) Left (loaded_module (mi_module (link_module_iface lm))) + + loaded_module mod = + lookupModuleEnv (objs_loaded pls) mod + <|> + lookupModuleEnv (bcos_loaded pls) mod {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -77,11 +77,13 @@ import GHC.Utils.TmpFs import GHC.Unit.Env import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) +import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.State as Packages import qualified GHC.Data.ShortText as ST import GHC.Data.FastString +import qualified GHC.Data.Maybe as Maybe import GHC.Linker.Deps import GHC.Linker.MacOS @@ -94,6 +96,7 @@ import Control.Monad import qualified Data.Set as Set import Data.Char (isSpace) import qualified Data.Foldable as Foldable +import Data.Functor ((<&>)) import Data.IORef import Data.List (intercalate, isPrefixOf, nub, partition) import Data.Maybe @@ -231,10 +234,10 @@ loadDependencies interp hsc_env pls span needed_mods = do -- Find what packages and linkables are required deps <- getLinkDeps opts interp pls span needed_mods - let this_pkgs_needed = ldNeededUnits deps + let this_pkgs_needed = ldAllUnits deps -- Link the packages and modules required - pls1 <- loadPackages' interp hsc_env (ldUnits deps) pls + pls1 <- loadPackages' interp hsc_env (ldNeededUnits deps) pls (pls2, succ) <- loadModuleLinkables interp hsc_env pls1 (ldNeededLinkables deps) let this_pkgs_loaded = udfmRestrictKeys all_pkgs_loaded $ getUniqDSet trans_pkgs_needed all_pkgs_loaded = pkgs_loaded pls2 @@ -650,14 +653,34 @@ initLinkDepsOpts hsc_env = opts , ldWays = ways dflags , ldLoadIface , ldLoadByteCode + , ldDebugTrace = debugTraceMsg (hsc_logger hsc_env) 3 } dflags = hsc_dflags hsc_env - ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env - $ loadInterface msg mod (ImportByUser NotBoot) + + ldLoadIface msg mod = + initIfaceCheck (text "loader") hsc_env (loadInterface msg mod (ImportByUser NotBoot)) >>= \case + Maybe.Failed err -> pure (Maybe.Failed err) + Maybe.Succeeded iface -> + find_location mod <&> \case + InstalledFound loc _ -> Maybe.Succeeded (iface, loc) + err -> Maybe.Failed $ + cannotFindInterface unit_state home_unit + (targetProfile dflags) (moduleName mod) err + + find_location mod = + liftIO $ + findExactModule (hsc_FC hsc_env) (initFinderOpts dflags) + (initFinderOpts . homeUnitEnv_dflags <$> hsc_HUG hsc_env) + unit_state home_unit + (toUnitId <$> mod) + + unit_state = hsc_units hsc_env + + home_unit = ue_homeUnit (hsc_unit_env hsc_env) ldLoadByteCode mod = do EPS {eps_iface_bytecode} <- hscEPS hsc_env - sequence (lookupModuleEnv eps_iface_bytecode mod) + pure (lookupModuleEnv eps_iface_bytecode mod) ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -734,7 +734,7 @@ mkStubPaths fopts mod location = do src_basename = OsPath.dropExtension <$> ml_hs_file_ospath location -- ----------------------------------------------------------------------------- --- findLinkable isn't related to the other stuff in here, +-- findObjectLinkable isn't related to the other stuff in here, -- but there's no other obvious place for it findObjectLinkableMaybe :: Module -> ModLocation -> IO (Maybe Linkable) ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -88,9 +88,10 @@ settings. 2. In oneshot mode, which compiles individual modules without a shared home unit graph, a previously compiled module is not reprocessed as described for make mode above. - When 'get_link_deps' encounters a dependency on a local module, it requests - its bytecode from the External Package State, who loads the interface - on-demand. + 'get_link_deps' requests the bytecode of dependencies from the External + Package State, who loads the interface on-demand. + This works for modules in local directories (via @-i@ and @-hidir@) as well + as those exposed from a package DB. Since the EPS stores interfaces for all package dependencies in addition to local modules in oneshot mode, it has a substantial memory footprint. @@ -210,7 +211,9 @@ If the 'HomeModLinkable' already contains bytecode (case 1), this is a no-op. Otherwise, the stub objects from the interface are compiled to objects in 'generateByteCode' and added to the 'HomeModLinkable' as well. -Case 3 is not implemented yet (!13042). +In case 3, Core bindings are loaded from the EPS, where stubs only exist in +their serialized form in the interface, so they must be regenerated like in case +2. Problem 3: ===================================== testsuite/tests/bytecode/T25090/Dep.hs ===================================== @@ -0,0 +1,15 @@ +module Dep where + +data A = A Int + +used :: Int +used = 9681 + +dep :: A +dep = A used + +unused1 :: A +unused1 = A 1 + +unused2 :: A +unused2 = unused1 ===================================== testsuite/tests/bytecode/T25090/DepApi.hs ===================================== @@ -0,0 +1,7 @@ +module DepApi (A (A), dep) where + +import Dep (A (A)) +import qualified Dep + +dep :: A +dep = Dep.dep ===================================== testsuite/tests/bytecode/T25090/Local.hs ===================================== @@ -0,0 +1,15 @@ +{-# language PackageImports #-} + +module Local where + +import GHC.Prim +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +-- just to be sure that the file isn't accidentally picked up locally +import "dep" DepApi (dep, A (A)) +import {-# source #-} Num (num) + +splc :: ExpQ +splc = lift @_ @Int (num + d) + where + A d = dep ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -19,3 +19,35 @@ T25090a: T25090b: $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 ./exe + +DB := -package-db db -package dep +BASIC := $(TEST_HC_OPTS) $(DB) -this-unit-id=pkgdep -v0 +BC := -fprefer-byte-code -fbyte-code-and-object-code +ARGS := $(BASIC) $(BC) + +T25090_pkg: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + ./run.bash "$(TEST_HC)" "$(ARGS) -dynamic" + +T25090_pkg_empty: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared-empty" + ./run.bash "$(TEST_HC)" "$(ARGS) -dynamic" + +T25090_pkg_nolib: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "none" + ./run.bash "$(TEST_HC)" "$(ARGS)" + +T25090_pkg_obj_code: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + ./run.bash "$(TEST_HC)" "$(BASIC) -dynamic -fbyte-code-and-object-code" + +T25090_pkg_multi_unit: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + mkdir -p unit2-src/ + mv Local.hs Num.hs Num.hs-boot unit2-src/ + "$(TEST_HC)" $(TEST_HC_OPTS) $(ARGS) -dynamic -unit @unit1 -unit @unit2 + ./PkgBytecode + +T25090_pkg_archive: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "archive" + ./run.bash "$(TEST_HC)" "$(ARGS)" ===================================== testsuite/tests/bytecode/T25090/Num.hs ===================================== @@ -0,0 +1,4 @@ +module Num where + +num :: Int +num = 48332 ===================================== testsuite/tests/bytecode/T25090/Num.hs-boot ===================================== @@ -0,0 +1,3 @@ +module Num where + +num :: Int ===================================== testsuite/tests/bytecode/T25090/PkgBytecode.hs ===================================== @@ -0,0 +1,12 @@ +{-# language TemplateHaskell #-} + +module Main where + +import GHC.Prim +import Local (splc) + +a :: Int +a = $(splc) + +main :: IO () +main = putStrLn (show a) ===================================== testsuite/tests/bytecode/T25090/PkgBytecode.stdout ===================================== @@ -0,0 +1 @@ +58013 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -16,3 +16,33 @@ def test_T25090(name): test_T25090('T25090a') test_T25090('T25090b') + +def test_pkg(name, files = []): + test( + name, + [ + extra_files([ + 'PkgBytecode.hs', + 'Local.hs', + 'Dep.hs', + 'DepApi.hs', + 'Num.hs', + 'Num.hs-boot', + 'dep.conf', + 'prep.bash', + 'run.bash', + ] + files), + use_specs({'stdout': 'PkgBytecode.stdout'}), + ], + makefile_test, + [], + ) + +test_pkg('T25090_pkg') +test_pkg('T25090_pkg_empty') +test_pkg('T25090_pkg_nolib') +test_pkg('T25090_pkg_obj_code') +test_pkg('T25090_pkg_multi_unit', ['unit1', 'unit2']) +# TODO this doesn't work, because `locateLib` ignores static archives when the interpreter is dynamic, even though a +# comment says "search for .so libraries _first_" (rather than "only"). +# test_pkg('T25090_pkg_archive') ===================================== testsuite/tests/bytecode/T25090/dep.conf ===================================== @@ -0,0 +1,8 @@ +name: dep +version: 1.0 +id: dep-1.0 +key: dep-1.0 +exposed: True +exposed-modules: DepApi +import-dirs: ${pkgroot}/dep +library-dirs: ${pkgroot}/dep ===================================== testsuite/tests/bytecode/T25090/prep.bash ===================================== @@ -0,0 +1,57 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_opts="$2" +ghc_pkg_cmd="$3" +library="$4" + +base="$PWD" +db="$base/db" +dep="$base/dep" +conf_dep="${dep}/dep.conf" + +mkdir -p "$dep" "$db" +mv Dep.hs DepApi.hs "$dep/" +cp dep.conf "$dep/" + +ghc_pkg() +{ + eval "${ghc_pkg_cmd at Q} --no-user-package-db --package-db=${db at Q} $@" +} + +ghc() +{ + eval "${ghc_cmd at Q} ${ghc_opts/-rtsopts/} -package-db ${db at Q} -hidir ${dep at Q} -O0 -this-unit-id dep-1.0 -fbyte-code-and-object-code $@" +} + +version=$(ghc "--numeric-version") + +ghc_pkg recache + +ghc "-dynamic-too -c ${dep at Q}/Dep.hs ${dep at Q}/DepApi.hs" + +if [[ "$library" == 'shared' ]] +then + ghc "-dynamic -shared -o ${dep at Q}/libHSdep-1.0-ghc$version.so ${dep at Q}/Dep.dyn_o ${dep at Q}/DepApi.dyn_o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'shared-empty' ]] +then + echo 'module Dummy where' > Dummy.hs + ghc "-dynamic-too -c Dummy.hs" + ghc "-dynamic -shared -o ${dep at Q}/libHSdep-1.0-ghc$version.so Dummy.dyn_o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'archive' ]] +then + $AR cqs "${dep}/libHSdep-1.0.a" "${dep}/Dep.o" "${dep}/DepApi.o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'none' ]] +then + : +else + echo "Invalid argument for 'library': $library" + exit 1 +fi + +ghc_pkg -v0 register "${conf_dep at Q}" ===================================== testsuite/tests/bytecode/T25090/run.bash ===================================== @@ -0,0 +1,16 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_opts="$2" + +ghc() +{ + eval "${ghc_cmd at Q} $ghc_opts $@" +} + +ghc -c Num.hs-boot Num.hs Local.hs +ghc -c PkgBytecode.hs +ghc PkgBytecode.o -o PkgBytecode +./PkgBytecode ===================================== testsuite/tests/bytecode/T25090/unit1 ===================================== @@ -0,0 +1 @@ +-i -i. PkgBytecode -this-unit-id unit1 -package-id unit2 ===================================== testsuite/tests/bytecode/T25090/unit2 ===================================== @@ -0,0 +1 @@ +-i -i./unit2-src Local Num -this-unit-id unit2 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8a72a4825fdd43d9dec1ca89f7f01105c8539d60 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8a72a4825fdd43d9dec1ca89f7f01105c8539d60 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 17:58:51 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 30 Sep 2024 13:58:51 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/package-deps-bytecode-squashed] Link interface bytecode from package DBs if possible Message-ID: <66fae6db4ab20_15ec3090a588656a3@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/package-deps-bytecode-squashed at Glasgow Haskell Compiler / GHC Commits: aac0c1be by Torsten Schmits at 2024-09-30T19:58:37+02:00 Link interface bytecode from package DBs if possible Part of #25090. MR !????? When splices are executed with `-fprefer-byte-code`, the loader will compile Core bindings to bytecode if those are present in interfaces of module dependencies. So far, this only applied to local modules (i.e. home modules in make mode and non-package deps in oneshot mode). This patch extends support to interfaces loaded from a package DB. In `getLinkDeps`, the dependencies chosen for recursion were restricted to `dep_direct_mods`, which has been changed to include `mi_usages`. In order to unify treatment of the different link variants across make/oneshot mode, the algorithm's intermediate results have been abstracted into the data types `LinkDep` and `LinkModule`. At the moment, following package deps is only implemented for oneshot mode. - - - - - 19 changed files: - compiler/GHC/Iface/Load.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/Dep.hs - + testsuite/tests/bytecode/T25090/DepApi.hs - + testsuite/tests/bytecode/T25090/Local.hs - testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/Num.hs - + testsuite/tests/bytecode/T25090/Num.hs-boot - + testsuite/tests/bytecode/T25090/PkgBytecode.hs - + testsuite/tests/bytecode/T25090/PkgBytecode.stdout - testsuite/tests/bytecode/T25090/all.T - + testsuite/tests/bytecode/T25090/dep.conf - + testsuite/tests/bytecode/T25090/prep.bash - + testsuite/tests/bytecode/T25090/run.bash - + testsuite/tests/bytecode/T25090/unit1 - + testsuite/tests/bytecode/T25090/unit2 Changes: ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -672,7 +672,11 @@ dontLeakTheHUG thing_inside = do in hsc_env { hsc_targets = panic "cleanTopEnv: hsc_targets" , hsc_mod_graph = panic "cleanTopEnv: hsc_mod_graph" - , hsc_IC = panic "cleanTopEnv: hsc_IC" + -- TODO this is needed for bytecode compilation of package deps + -- only. local EPS modules work fine. + -- Also it appears to work when the local modules use multiple + -- home units?!?!? + -- , hsc_IC = panic "cleanTopEnv: hsc_IC" , hsc_type_env_vars = case maybe_type_vars of Just vars -> vars Nothing -> panic "cleanTopEnv: hsc_type_env_vars" ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -6,6 +6,7 @@ {-# LANGUAGE CPP #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE NamedFieldPuns #-} module GHC.Linker.Deps ( LinkDepsOpts (..) @@ -28,6 +29,7 @@ import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM import GHC.Utils.Outputable +import qualified GHC.Utils.Outputable as Outputable import GHC.Utils.Panic import GHC.Utils.Error @@ -47,9 +49,11 @@ import GHC.Utils.Misc import GHC.Unit.Home import GHC.Data.Maybe -import Control.Monad import Control.Applicative +import Control.Monad.IO.Class (MonadIO (liftIO)) +import Control.Monad.Trans.Except (ExceptT, runExceptT, throwE) +import Data.Foldable (traverse_) import qualified Data.Set as Set import qualified Data.Map as M import Data.List (isSuffixOf) @@ -68,15 +72,16 @@ data LinkDepsOpts = LinkDepsOpts , ldWays :: !Ways -- ^ Enabled ways , ldFinderCache :: !FinderCache , ldFinderOpts :: !FinderOpts - , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) - , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation))) + , ldLoadByteCode :: !(Module -> IO (Maybe (IO Linkable))) + , ldDebugTrace :: !(SDoc -> IO ()) } data LinkDeps = LinkDeps { ldNeededLinkables :: [Linkable] , ldAllLinkables :: [Linkable] - , ldUnits :: [UnitId] - , ldNeededUnits :: UniqDSet UnitId + , ldNeededUnits :: [UnitId] + , ldAllUnits :: UniqDSet UnitId } -- | Find all the packages and linkables that a set of modules depends on @@ -102,7 +107,14 @@ getLinkDeps opts interp pls span mods = do get_link_deps opts pls maybe_normal_osuf span mods - +-- | Compute the linkables for the given module set's dependencies. +-- +-- Home modules in make mode are treated separately in a preprocessing step, +-- then all the remaining external deps are processed for both modes. +-- If bytecode is available, transitive external deps are included, otherwise +-- the module's library is linked and processing stops. +-- +-- The results are split into sets of needed/loaded modules/packages. get_link_deps :: LinkDepsOpts -> LoaderState @@ -111,46 +123,40 @@ get_link_deps -> [Module] -> IO LinkDeps get_link_deps opts pls maybe_normal_osuf span mods = do - -- 1. Find the dependent home-pkg-modules/packages from each iface - -- (omitting modules from the interactive package, which is already linked) - (mods_s, pkgs_s) <- - -- Why two code paths here? There is a significant amount of repeated work - -- performed calculating transitive dependencies - -- if --make uses the oneShot code path (see MultiLayerModulesTH_* tests) - if ldOneShotMode opts - then follow_deps (filterOut isInteractiveModule mods) - emptyUniqDSet emptyUniqDSet; - else do - (pkgs, mmods) <- unzip <$> mapM get_mod_info all_home_mods - return (catMaybes mmods, unionManyUniqDSets (init_pkg_set : pkgs)) - - let - -- 2. Exclude ones already linked - -- Main reason: avoid findModule calls in get_linkable - (mods_needed, links_got) = partitionWith split_mods mods_s - pkgs_needed = eltsUDFM $ getUniqDSet pkgs_s `minusUDFM` pkgs_loaded pls - - split_mods mod = - let is_linked = lookupModuleEnv (objs_loaded pls) mod - <|> lookupModuleEnv (bcos_loaded pls) mod - in case is_linked of - Just linkable -> Right linkable - Nothing -> Left mod - - -- 3. For each dependent module, find its linkable - -- This will either be in the HPT or (in the case of one-shot - -- compilation) we may need to use maybe_getFileLinkable - lnks_needed <- mapM (get_linkable (ldObjSuffix opts)) mods_needed - - return $ LinkDeps - { ldNeededLinkables = lnks_needed - , ldAllLinkables = links_got ++ lnks_needed - , ldUnits = pkgs_needed - , ldNeededUnits = pkgs_s - } + (link_deps_home, module_deps_external) <- separate_home_deps + link_deps_external <- external_deps opts module_deps_external + let (loaded_modules, needed_modules, ldAllUnits, ldNeededUnits) = + classify_deps pls (link_deps_home ++ link_deps_external) + ldNeededLinkables <- mapM module_linkable needed_modules + pure LinkDeps { + ldNeededLinkables, + ldAllLinkables = loaded_modules ++ ldNeededLinkables, + ldNeededUnits, + ldAllUnits + } where mod_graph = ldModuleGraph opts unit_env = ldUnitEnv opts + noninteractive = filterOut isInteractiveModule mods + + -- | Preprocess the dependencies in make mode to remove all home modules, + -- since the transitive dependency closure is already cached for those in + -- the HUG (see MultiLayerModulesTH_* tests for the performance impact). + -- + -- Returns the remaining, external, dependencies on the right, which is the + -- entire set for oneshot mode. + separate_home_deps = + if ldOneShotMode opts + then pure ([], noninteractive) + else make_deps + + make_deps = do + (dep_ext, mmods) <- unzip <$> mapM get_mod_info all_home_mods + let + link_mods = + listToUDFM [(moduleName (mi_module (hm_iface m)), m) | m <- mmods] + ext = uniqDSetToList (unionManyUniqDSets (init_ext : dep_ext)) + pure ([LinkModules (LinkHomeModule <$> link_mods)], ext) -- This code is used in `--make` mode to calculate the home package and unit dependencies -- for a set of modules. @@ -160,9 +166,9 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- It is also a matter of correctness to use the module graph so that dependencies between home units -- is resolved correctly. - make_deps_loop :: (UniqDSet UnitId, Set.Set NodeKey) -> [ModNodeKeyWithUid] -> (UniqDSet UnitId, Set.Set NodeKey) + make_deps_loop :: (UniqDSet Module, Set.Set NodeKey) -> [ModNodeKeyWithUid] -> (UniqDSet Module, Set.Set NodeKey) make_deps_loop found [] = found - make_deps_loop found@(found_units, found_mods) (nk:nexts) + make_deps_loop found@(external, found_mods) (nk:nexts) | NodeKey_Module nk `Set.member` found_mods = make_deps_loop found nexts | otherwise = case M.lookup (NodeKey_Module nk) (mgTransDeps mod_graph) of @@ -171,85 +177,26 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- See #936 and the ghci.prog007 test for why we have to continue traversing through -- boot modules. todo_boot_mods = [ModNodeKeyWithUid (GWIB mn NotBoot) uid | NodeKey_Module (ModNodeKeyWithUid (GWIB mn IsBoot) uid) <- Set.toList trans_deps] - in make_deps_loop (found_units, deps `Set.union` found_mods) (todo_boot_mods ++ nexts) + in make_deps_loop (external, deps `Set.union` found_mods) (todo_boot_mods ++ nexts) Nothing -> - let (ModNodeKeyWithUid _ uid) = nk - in make_deps_loop (addOneToUniqDSet found_units uid, found_mods) nexts + let (ModNodeKeyWithUid (GWIB mod_name _) uid) = nk + in make_deps_loop (addOneToUniqDSet external (Module (RealUnit (Definite uid)) mod_name), found_mods) nexts mkNk m = ModNodeKeyWithUid (GWIB (moduleName m) NotBoot) (moduleUnitId m) - (init_pkg_set, all_deps) = make_deps_loop (emptyUniqDSet, Set.empty) $ map mkNk (filterOut isInteractiveModule mods) + (init_ext, all_deps) = make_deps_loop (emptyUniqDSet, Set.empty) $ map mkNk noninteractive all_home_mods = [with_uid | NodeKey_Module with_uid <- Set.toList all_deps] get_mod_info (ModNodeKeyWithUid gwib uid) = case lookupHug (ue_home_unit_graph unit_env) uid (gwib_mod gwib) of - Just hmi -> - let iface = (hm_iface hmi) - mmod = case mi_hsc_src iface of - HsBootFile -> link_boot_mod_error (mi_module iface) - _ -> return $ Just (mi_module iface) - - in (mkUniqDSet $ Set.toList $ dep_direct_pkgs (mi_deps iface),) <$> mmod + Just hmi -> do + let iface = hm_iface hmi + case mi_hsc_src iface of + HsBootFile -> throwProgramError opts $ link_boot_mod_error (mi_module iface) + _ -> pure (mkUniqDSet $ [usg_mod | UsagePackageModule {usg_mod} <- mi_usages iface], hmi) Nothing -> throwProgramError opts $ text "getLinkDeps: Home module not loaded" <+> ppr (gwib_mod gwib) <+> ppr uid - - -- This code is used in one-shot mode to traverse downwards through the HPT - -- to find all link dependencies. - -- The ModIface contains the transitive closure of the module dependencies - -- within the current package, *except* for boot modules: if we encounter - -- a boot module, we have to find its real interface and discover the - -- dependencies of that. Hence we need to traverse the dependency - -- tree recursively. See bug #936, testcase ghci/prog007. - follow_deps :: [Module] -- modules to follow - -> UniqDSet Module -- accum. module dependencies - -> UniqDSet UnitId -- accum. package dependencies - -> IO ([Module], UniqDSet UnitId) -- result - follow_deps [] acc_mods acc_pkgs - = return (uniqDSetToList acc_mods, acc_pkgs) - follow_deps (mod:mods) acc_mods acc_pkgs - = do - mb_iface <- ldLoadIface opts msg mod - iface <- case mb_iface of - Failed err -> throwProgramError opts $ - missingInterfaceErrorDiagnostic (ldMsgOpts opts) err - Succeeded iface -> return iface - - when (mi_boot iface == IsBoot) $ link_boot_mod_error mod - - let - pkg = moduleUnit mod - deps = mi_deps iface - - pkg_deps = dep_direct_pkgs deps - (boot_deps, mod_deps) = flip partitionWith (Set.toList (dep_direct_mods deps)) $ - \case - (_, GWIB m IsBoot) -> Left m - (_, GWIB m NotBoot) -> Right m - - mod_deps' = case ue_homeUnit unit_env of - Nothing -> [] - Just home_unit -> filter (not . (`elementOfUniqDSet` acc_mods)) (map (mkHomeModule home_unit) $ (boot_deps ++ mod_deps)) - acc_mods' = case ue_homeUnit unit_env of - Nothing -> acc_mods - Just home_unit -> addListToUniqDSet acc_mods (mod : map (mkHomeModule home_unit) mod_deps) - acc_pkgs' = addListToUniqDSet acc_pkgs (Set.toList pkg_deps) - - case ue_homeUnit unit_env of - Just home_unit | isHomeUnit home_unit pkg -> follow_deps (mod_deps' ++ mods) - acc_mods' acc_pkgs' - _ -> follow_deps mods acc_mods (addOneToUniqDSet acc_pkgs' (toUnitId pkg)) - where - msg = text "need to link module" <+> ppr mod <+> - text "due to use of Template Haskell" - - - - link_boot_mod_error :: Module -> IO a - link_boot_mod_error mod = throwProgramError opts $ - text "module" <+> ppr mod <+> - text "cannot be linked; it is only available as a boot module" - no_obj :: Outputable a => a -> IO b no_obj mod = dieWith opts span $ text "cannot find object file for module " <> @@ -258,6 +205,20 @@ get_link_deps opts pls maybe_normal_osuf span mods = do while_linking_expr = text "while linking an interpreted expression" + -- | Extract the 'Linkable's for unlinked modules from the intermediate + -- results. + module_linkable = \case + LinkHomeModule hmi -> + adjust_linkable (expectJust "getLinkDeps" (homeModLinkable hmi)) + + LinkObjectModule iface loc -> do + let mod = mi_module iface + findObjectLinkableMaybe mod loc >>= \case + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + + LinkByteCodeModule _ load_bytecode -> + load_bytecode -- See Note [Using Byte Code rather than Object Code for Template Haskell] homeModLinkable :: HomeModInfo -> Maybe Linkable @@ -266,57 +227,257 @@ get_link_deps opts pls maybe_normal_osuf span mods = do then homeModInfoByteCode hmi <|> homeModInfoObject hmi else homeModInfoObject hmi <|> homeModInfoByteCode hmi - get_linkable osuf mod -- A home-package module - | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) - = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) - | otherwise - = do -- It's not in the HPT because we are in one shot mode, - -- so use the Finder to get a ModLocation... - case ue_homeUnit unit_env of - Nothing -> no_obj mod - Just home_unit -> do - from_bc <- ldLoadByteCode opts mod - maybe (fallback_no_bytecode home_unit mod) pure from_bc - where - - fallback_no_bytecode home_unit mod = do - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc _ -> do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - _ -> no_obj (moduleName mod) - - adjust_linkable lnk - | Just new_osuf <- maybe_normal_osuf = do - new_parts <- mapM (adjust_part new_osuf) - (linkableParts lnk) - return lnk{ linkableParts=new_parts } - | otherwise = - return lnk - - adjust_part new_osuf part = case part of - DotO file ModuleObject -> do - massert (osuf `isSuffixOf` file) - let file_base = fromJust (stripExtension osuf file) - new_file = file_base <.> new_osuf - ok <- doesFileExist new_file - if (not ok) - then dieWith opts span $ - text "cannot find object file " - <> quotes (text new_file) $$ while_linking_expr - else return (DotO new_file ModuleObject) - DotO file ForeignObject -> pure (DotO file ForeignObject) - DotA fp -> panic ("adjust_ul DotA " ++ show fp) - DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) - BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) + adjust_linkable lnk + | Just new_osuf <- maybe_normal_osuf = do + new_uls <- mapM (adjust_part (ldObjSuffix opts) new_osuf) + (linkableParts lnk) + return lnk {linkableParts = new_uls} + | otherwise = + return lnk + + adjust_part osuf new_osuf part = case part of + DotO file ModuleObject -> do + massert (osuf `isSuffixOf` file) + let file_base = fromJust (stripExtension osuf file) + new_file = file_base <.> new_osuf + ok <- doesFileExist new_file + if (not ok) + then dieWith opts span $ + text "cannot find object file " + <> quotes (text new_file) $$ while_linking_expr + else return (DotO new_file ModuleObject) + DotO file ForeignObject -> pure (DotO file ForeignObject) + DotA fp -> panic ("adjust_part DotA " ++ show fp) + DotDLL fp -> panic ("adjust_part DotDLL " ++ show fp) + BCOs {} -> pure part + LazyBCOs{} -> pure part + CoreBindings WholeCoreBindings {wcb_module} -> + pprPanic "Unhydrated core bindings" (ppr wcb_module) + +data LinkModule = + LinkHomeModule !HomeModInfo + | + LinkObjectModule !ModIface !ModLocation + | + LinkByteCodeModule !ModIface !(IO Linkable) + +link_module_iface :: LinkModule -> ModIface +link_module_iface = \case + LinkHomeModule hmi -> hm_iface hmi + LinkObjectModule iface _ -> iface + LinkByteCodeModule iface _ -> iface + +instance Outputable LinkModule where + ppr = \case + LinkHomeModule hmi -> ppr (mi_module (hm_iface hmi)) <+> brackets (text "HMI") + LinkObjectModule iface _ -> ppr (mi_module iface) + LinkByteCodeModule iface _ -> ppr (mi_module iface) <+> brackets (text "BC") + +data LinkDep = + LinkModules !(UniqDFM ModuleName LinkModule) + | + LinkLibrary !UnitId + +instance Outputable LinkDep where + ppr = \case + LinkModules mods -> text "modules:" <+> ppr (eltsUDFM mods) + LinkLibrary uid -> text "library:" <+> ppr uid + +data OneshotError = + NoLocation !Module + | + NoInterface !MissingInterfaceError + | + LinkBootModule !Module + +-- Compute the transitive dependency closure of the given modules. +-- +-- Used for all oneshot mode dependencies and for external dependencies of home +-- modules in make mode. +-- +-- TODO is the following still relevant? +-- The ModIface contains the transitive closure of the module dependencies +-- within the current package, *except* for boot modules: if we encounter +-- a boot module, we have to find its real interface and discover the +-- dependencies of that. Hence we need to traverse the dependency +-- tree recursively. See bug #936, testcase ghci/prog007. +external_deps :: + LinkDepsOpts -> + -- | Modules whose imports to follow + [Module] -> + IO [LinkDep] +external_deps opts mods = + runExceptT (external_deps_loop opts mods emptyUDFM) >>= \case + Right a -> pure (eltsUDFM a) + Left err -> throwProgramError opts (message err) + where + message = \case + NoLocation mod -> + pprPanic "found iface but no location" (ppr mod) + NoInterface err -> + missingInterfaceErrorDiagnostic (ldMsgOpts opts) err + LinkBootModule mod -> + link_boot_mod_error mod + +external_deps_loop :: + LinkDepsOpts -> + [Module] -> + UniqDFM UnitId LinkDep -> + ExceptT OneshotError IO (UniqDFM UnitId LinkDep) +external_deps_loop _ [] acc = + pure acc +external_deps_loop opts (mod : mods) acc = do + (new_acc, new_mods, action) <- process_module + traverse_ debug_log action + external_deps_loop opts (new_mods ++ mods) new_acc + where + debug_log action = + liftIO $ ldDebugTrace opts $ + text "TH dep" <+> ppr mod <+> brackets (sep [ + if is_home then text "home" else Outputable.empty, + text action + ]) + + -- | Decide how this module needs to be processed. + -- We only need an interface if we want to load bytecode or if we have to + -- link an object file (which happens for home unit modules, since those + -- have no libraries). + process_module + | already_seen = pure (acc, [], Nothing) + | is_home || prefer_bytecode = try_iface + | otherwise = add_library + + -- | Check whether the current module was processed before. + -- Since the accumulator is keyed by unit ID, we have to perform two + -- lookups. + -- If another module from this module's unit has been determined to be + -- linked as a library previously, we skip this module, assuming that no + -- bytecode is available for the entire package. + already_seen + | Just (LinkModules mods) <- mod_dep + = elemUDFM mod_name mods + | Just (LinkLibrary _) <- mod_dep + = True + | otherwise + = False + + -- | Load the iface and attempt to get bytecode from Core bindings. + try_iface = + liftIO (ldLoadIface opts load_reason mod) >>= \case + Failed err -> throwE (NoInterface err) + Succeeded (iface, loc) -> do + mb_load_bc <- liftIO (ldLoadByteCode opts (mi_module iface)) + with_iface loc iface mb_load_bc + + -- | Decide how to link this module. + -- If bytecode or an object file is available, use those in that order. + -- Otherwise fall back to linking a library. + with_iface loc iface mb_load_bc + | IsBoot <- mi_boot iface + = throwE (LinkBootModule mod) + + | prefer_bytecode + , Just load_bc <- mb_load_bc + = pure (add_module iface (LinkByteCodeModule iface load_bc) "bytecode") + + | is_home + = pure (add_module iface (LinkObjectModule iface loc) "object") + + | otherwise + = add_library + + add_library = + pure (addToUDFM acc mod_unit_id (LinkLibrary mod_unit_id), [], Just "library") + + add_module iface lmod action = + (addListToUDFM with_mod (direct_pkgs iface), new_deps iface, Just action) + where + with_mod = alterUDFM (add_package_module lmod) acc mod_unit_id + + add_package_module lmod = \case + Just (LinkLibrary u) -> Just (LinkLibrary u) + Just (LinkModules old) -> Just (LinkModules (addToUDFM old mod_name lmod)) + Nothing -> Just (LinkModules (unitUDFM mod_name lmod)) + + direct_pkgs iface + | prefer_bytecode + = [] + | otherwise + = [(u, LinkLibrary u) | u <- Set.toList (dep_direct_pkgs (mi_deps iface))] + + new_deps iface + | prefer_bytecode + -- TODO How can we better determine the external deps? + -- OTOH, we probably don't want to link unused dependencies anyway. + = [usg_mod | UsagePackageModule {usg_mod} <- mi_usages iface] ++ local + | is_home + = local + | otherwise + = [] + where + local = + [ + mkModule mod_unit m + -- TODO Somehow this just works, no idea what the deal was in the + -- old code with boot modules. + | (_, GWIB m _) <- Set.toList (dep_direct_mods (mi_deps iface)) + ] + + is_home + | Just home <- mb_home + = homeUnitAsUnit home == mod_unit + | otherwise + = False + + mod_dep = lookupUDFM acc mod_unit_id + mod_name = moduleName mod + mod_unit_id = moduleUnitId mod + mod_unit = moduleUnit mod + load_reason = + text "need to link module" <+> ppr mod <+> + text "due to use of Template Haskell" + + prefer_bytecode = ldUseByteCode opts + mb_home = ue_homeUnit (ldUnitEnv opts) + +link_boot_mod_error :: Module -> SDoc +link_boot_mod_error mod = + text "module" <+> ppr mod <+> + text "cannot be linked; it is only available as a boot module" + +-- | Split link dependencies into the sets of modules and packages that have +-- been linked previously and those that need to be linked now by checking for +-- their presence in the 'LoaderState': +-- +-- - For module dependencies, in the sets of loaded objects and BCOs +-- ('objs_loaded' and 'bcos_loaded') +-- - For package dependencies, in the set of loaded packages ('pkgs_loaded') +classify_deps :: + LoaderState -> + [LinkDep] -> + ([Linkable], [LinkModule], UniqDSet UnitId, [UnitId]) +classify_deps pls deps = + (loaded_modules, needed_modules, all_packages, needed_packages) + where + (loaded_modules, needed_modules) = + partitionWith loaded_or_needed_module (concatMap eltsUDFM modules) + + needed_packages = + eltsUDFM (getUniqDSet all_packages `minusUDFM` pkgs_loaded pls) + + all_packages = mkUniqDSet packages + + (modules, packages) = flip partitionWith deps $ \case + LinkModules mods -> Left mods + LinkLibrary lib -> Right lib + + loaded_or_needed_module lm = + maybe (Right lm) Left (loaded_module (mi_module (link_module_iface lm))) + + loaded_module mod = + lookupModuleEnv (objs_loaded pls) mod + <|> + lookupModuleEnv (bcos_loaded pls) mod {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -77,11 +77,13 @@ import GHC.Utils.TmpFs import GHC.Unit.Env import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) +import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.State as Packages import qualified GHC.Data.ShortText as ST import GHC.Data.FastString +import qualified GHC.Data.Maybe as Maybe import GHC.Linker.Deps import GHC.Linker.MacOS @@ -94,6 +96,7 @@ import Control.Monad import qualified Data.Set as Set import Data.Char (isSpace) import qualified Data.Foldable as Foldable +import Data.Functor ((<&>)) import Data.IORef import Data.List (intercalate, isPrefixOf, nub, partition) import Data.Maybe @@ -231,10 +234,10 @@ loadDependencies interp hsc_env pls span needed_mods = do -- Find what packages and linkables are required deps <- getLinkDeps opts interp pls span needed_mods - let this_pkgs_needed = ldNeededUnits deps + let this_pkgs_needed = ldAllUnits deps -- Link the packages and modules required - pls1 <- loadPackages' interp hsc_env (ldUnits deps) pls + pls1 <- loadPackages' interp hsc_env (ldNeededUnits deps) pls (pls2, succ) <- loadModuleLinkables interp hsc_env pls1 (ldNeededLinkables deps) let this_pkgs_loaded = udfmRestrictKeys all_pkgs_loaded $ getUniqDSet trans_pkgs_needed all_pkgs_loaded = pkgs_loaded pls2 @@ -650,14 +653,34 @@ initLinkDepsOpts hsc_env = opts , ldWays = ways dflags , ldLoadIface , ldLoadByteCode + , ldDebugTrace = debugTraceMsg (hsc_logger hsc_env) 3 } dflags = hsc_dflags hsc_env - ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env - $ loadInterface msg mod (ImportByUser NotBoot) + + ldLoadIface msg mod = + initIfaceCheck (text "loader") hsc_env (loadInterface msg mod (ImportByUser NotBoot)) >>= \case + Maybe.Failed err -> pure (Maybe.Failed err) + Maybe.Succeeded iface -> + find_location mod <&> \case + InstalledFound loc _ -> Maybe.Succeeded (iface, loc) + err -> Maybe.Failed $ + cannotFindInterface unit_state home_unit + (targetProfile dflags) (moduleName mod) err + + find_location mod = + liftIO $ + findExactModule (hsc_FC hsc_env) (initFinderOpts dflags) + (initFinderOpts . homeUnitEnv_dflags <$> hsc_HUG hsc_env) + unit_state home_unit + (toUnitId <$> mod) + + unit_state = hsc_units hsc_env + + home_unit = ue_homeUnit (hsc_unit_env hsc_env) ldLoadByteCode mod = do EPS {eps_iface_bytecode} <- hscEPS hsc_env - sequence (lookupModuleEnv eps_iface_bytecode mod) + pure (lookupModuleEnv eps_iface_bytecode mod) ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -734,7 +734,7 @@ mkStubPaths fopts mod location = do src_basename = OsPath.dropExtension <$> ml_hs_file_ospath location -- ----------------------------------------------------------------------------- --- findLinkable isn't related to the other stuff in here, +-- findObjectLinkable isn't related to the other stuff in here, -- but there's no other obvious place for it findObjectLinkableMaybe :: Module -> ModLocation -> IO (Maybe Linkable) ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -88,9 +88,10 @@ settings. 2. In oneshot mode, which compiles individual modules without a shared home unit graph, a previously compiled module is not reprocessed as described for make mode above. - When 'get_link_deps' encounters a dependency on a local module, it requests - its bytecode from the External Package State, who loads the interface - on-demand. + 'get_link_deps' requests the bytecode of dependencies from the External + Package State, who loads the interface on-demand. + This works for modules in local directories (via @-i@ and @-hidir@) as well + as those exposed from a package DB. Since the EPS stores interfaces for all package dependencies in addition to local modules in oneshot mode, it has a substantial memory footprint. @@ -210,7 +211,9 @@ If the 'HomeModLinkable' already contains bytecode (case 1), this is a no-op. Otherwise, the stub objects from the interface are compiled to objects in 'generateByteCode' and added to the 'HomeModLinkable' as well. -Case 3 is not implemented yet (!13042). +In case 3, Core bindings are loaded from the EPS, where stubs only exist in +their serialized form in the interface, so they must be regenerated like in case +2. Problem 3: ===================================== testsuite/tests/bytecode/T25090/Dep.hs ===================================== @@ -0,0 +1,15 @@ +module Dep where + +data A = A Int + +used :: Int +used = 9681 + +dep :: A +dep = A used + +unused1 :: A +unused1 = A 1 + +unused2 :: A +unused2 = unused1 ===================================== testsuite/tests/bytecode/T25090/DepApi.hs ===================================== @@ -0,0 +1,7 @@ +module DepApi (A (A), dep) where + +import Dep (A (A)) +import qualified Dep + +dep :: A +dep = Dep.dep ===================================== testsuite/tests/bytecode/T25090/Local.hs ===================================== @@ -0,0 +1,15 @@ +{-# language PackageImports #-} + +module Local where + +import GHC.Prim +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +-- just to be sure that the file isn't accidentally picked up locally +import "dep" DepApi (dep, A (A)) +import {-# source #-} Num (num) + +splc :: ExpQ +splc = lift @_ @Int (num + d) + where + A d = dep ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -19,3 +19,35 @@ T25090a: T25090b: $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 ./exe + +DB := -package-db db -package dep +BASIC := $(TEST_HC_OPTS) $(DB) -this-unit-id=pkgdep -v0 +BC := -fprefer-byte-code -fbyte-code-and-object-code +ARGS := $(BASIC) $(BC) + +T25090_pkg: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + ./run.bash "$(TEST_HC)" "$(ARGS) -dynamic" + +T25090_pkg_empty: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared-empty" + ./run.bash "$(TEST_HC)" "$(ARGS) -dynamic" + +T25090_pkg_nolib: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "none" + ./run.bash "$(TEST_HC)" "$(ARGS)" + +T25090_pkg_obj_code: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + ./run.bash "$(TEST_HC)" "$(BASIC) -dynamic -fbyte-code-and-object-code" + +T25090_pkg_multi_unit: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + mkdir -p unit2-src/ + mv Local.hs Num.hs Num.hs-boot unit2-src/ + "$(TEST_HC)" $(TEST_HC_OPTS) $(ARGS) -dynamic -unit @unit1 -unit @unit2 + ./PkgBytecode + +T25090_pkg_archive: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "archive" + ./run.bash "$(TEST_HC)" "$(ARGS)" ===================================== testsuite/tests/bytecode/T25090/Num.hs ===================================== @@ -0,0 +1,4 @@ +module Num where + +num :: Int +num = 48332 ===================================== testsuite/tests/bytecode/T25090/Num.hs-boot ===================================== @@ -0,0 +1,3 @@ +module Num where + +num :: Int ===================================== testsuite/tests/bytecode/T25090/PkgBytecode.hs ===================================== @@ -0,0 +1,12 @@ +{-# language TemplateHaskell #-} + +module Main where + +import GHC.Prim +import Local (splc) + +a :: Int +a = $(splc) + +main :: IO () +main = putStrLn (show a) ===================================== testsuite/tests/bytecode/T25090/PkgBytecode.stdout ===================================== @@ -0,0 +1 @@ +58013 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -16,3 +16,33 @@ def test_T25090(name): test_T25090('T25090a') test_T25090('T25090b') + +def test_pkg(name, files = []): + test( + name, + [ + extra_files([ + 'PkgBytecode.hs', + 'Local.hs', + 'Dep.hs', + 'DepApi.hs', + 'Num.hs', + 'Num.hs-boot', + 'dep.conf', + 'prep.bash', + 'run.bash', + ] + files), + use_specs({'stdout': 'PkgBytecode.stdout'}), + ], + makefile_test, + [], + ) + +test_pkg('T25090_pkg') +test_pkg('T25090_pkg_empty') +test_pkg('T25090_pkg_nolib') +test_pkg('T25090_pkg_obj_code') +test_pkg('T25090_pkg_multi_unit', ['unit1', 'unit2']) +# TODO this doesn't work, because `locateLib` ignores static archives when the interpreter is dynamic, even though a +# comment says "search for .so libraries _first_" (rather than "only"). +# test_pkg('T25090_pkg_archive') ===================================== testsuite/tests/bytecode/T25090/dep.conf ===================================== @@ -0,0 +1,8 @@ +name: dep +version: 1.0 +id: dep-1.0 +key: dep-1.0 +exposed: True +exposed-modules: DepApi +import-dirs: ${pkgroot}/dep +library-dirs: ${pkgroot}/dep ===================================== testsuite/tests/bytecode/T25090/prep.bash ===================================== @@ -0,0 +1,57 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_opts="$2" +ghc_pkg_cmd="$3" +library="$4" + +base="$PWD" +db="$base/db" +dep="$base/dep" +conf_dep="${dep}/dep.conf" + +mkdir -p "$dep" "$db" +mv Dep.hs DepApi.hs "$dep/" +cp dep.conf "$dep/" + +ghc_pkg() +{ + eval "${ghc_pkg_cmd at Q} --no-user-package-db --package-db=${db at Q} $@" +} + +ghc() +{ + eval "${ghc_cmd at Q} ${ghc_opts/-rtsopts/} -package-db ${db at Q} -hidir ${dep at Q} -O0 -this-unit-id dep-1.0 -fbyte-code-and-object-code $@" +} + +version=$(ghc "--numeric-version") + +ghc_pkg recache + +ghc "-dynamic-too -c ${dep at Q}/Dep.hs ${dep at Q}/DepApi.hs" + +if [[ "$library" == 'shared' ]] +then + ghc "-dynamic -shared -o ${dep at Q}/libHSdep-1.0-ghc$version.so ${dep at Q}/Dep.dyn_o ${dep at Q}/DepApi.dyn_o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'shared-empty' ]] +then + echo 'module Dummy where' > Dummy.hs + ghc "-dynamic-too -c Dummy.hs" + ghc "-dynamic -shared -o ${dep at Q}/libHSdep-1.0-ghc$version.so Dummy.dyn_o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'archive' ]] +then + $AR cqs "${dep}/libHSdep-1.0.a" "${dep}/Dep.o" "${dep}/DepApi.o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'none' ]] +then + : +else + echo "Invalid argument for 'library': $library" + exit 1 +fi + +ghc_pkg -v0 register "${conf_dep at Q}" ===================================== testsuite/tests/bytecode/T25090/run.bash ===================================== @@ -0,0 +1,16 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_opts="$2" + +ghc() +{ + eval "${ghc_cmd at Q} $ghc_opts $@" +} + +ghc -c Num.hs-boot Num.hs Local.hs +ghc -c PkgBytecode.hs +ghc PkgBytecode.o -o PkgBytecode +./PkgBytecode ===================================== testsuite/tests/bytecode/T25090/unit1 ===================================== @@ -0,0 +1 @@ +-i -i. PkgBytecode -this-unit-id unit1 -package-id unit2 ===================================== testsuite/tests/bytecode/T25090/unit2 ===================================== @@ -0,0 +1 @@ +-i -i./unit2-src Local Num -this-unit-id unit2 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/aac0c1be7597052551d36529018fd229df5149cd -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/aac0c1be7597052551d36529018fd229df5149cd You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 18:46:15 2024 From: gitlab at gitlab.haskell.org (Torsten Schmits (@torsten.schmits)) Date: Mon, 30 Sep 2024 14:46:15 -0400 Subject: [Git][ghc/ghc][wip/torsten.schmits/package-deps-bytecode-squashed] Link interface bytecode from package DBs if possible Message-ID: <66faf1f6e5b57_15ec30b415f4694a7@gitlab.mail> Torsten Schmits pushed to branch wip/torsten.schmits/package-deps-bytecode-squashed at Glasgow Haskell Compiler / GHC Commits: d9e34209 by Torsten Schmits at 2024-09-30T20:45:59+02:00 Link interface bytecode from package DBs if possible Part of #25090. MR !????? When splices are executed with `-fprefer-byte-code`, the loader will compile Core bindings to bytecode if those are present in interfaces of module dependencies. So far, this only applied to local modules (i.e. home modules in make mode and non-package deps in oneshot mode). This patch extends support to interfaces loaded from a package DB. In `getLinkDeps`, the dependencies chosen for recursion were restricted to `dep_direct_mods`, which has been changed to include `mi_usages`. In order to unify treatment of the different link variants across make/oneshot mode, the algorithm's intermediate results have been abstracted into the data types `LinkDep` and `LinkModule`. At the moment, following package deps is only implemented for oneshot mode. - - - - - 21 changed files: - compiler/GHC/Driver/Flags.hs - compiler/GHC/Driver/Session.hs - compiler/GHC/Iface/Load.hs - compiler/GHC/Linker/Deps.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Unit/Finder.hs - compiler/GHC/Unit/Module/WholeCoreBindings.hs - + testsuite/tests/bytecode/T25090/Dep.hs - + testsuite/tests/bytecode/T25090/DepApi.hs - + testsuite/tests/bytecode/T25090/Local.hs - testsuite/tests/bytecode/T25090/Makefile - + testsuite/tests/bytecode/T25090/Num.hs - + testsuite/tests/bytecode/T25090/Num.hs-boot - + testsuite/tests/bytecode/T25090/PkgBytecode.hs - + testsuite/tests/bytecode/T25090/PkgBytecode.stdout - testsuite/tests/bytecode/T25090/all.T - + testsuite/tests/bytecode/T25090/dep.conf - + testsuite/tests/bytecode/T25090/prep.bash - + testsuite/tests/bytecode/T25090/run.bash - + testsuite/tests/bytecode/T25090/unit1 - + testsuite/tests/bytecode/T25090/unit2 Changes: ===================================== compiler/GHC/Driver/Flags.hs ===================================== @@ -843,6 +843,7 @@ data GeneralFlag | Opt_BuildDynamicToo | Opt_WriteIfSimplifiedCore | Opt_UseBytecodeRatherThanObjects + | Opt_PackageDbBytecode -- safe haskell flags | Opt_DistrustAllPackages ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -2540,6 +2540,7 @@ fFlagsDeps = [ flagSpec "link-rts" Opt_LinkRts, flagSpec "byte-code-and-object-code" Opt_ByteCodeAndObjectCode, flagSpec "prefer-byte-code" Opt_UseBytecodeRatherThanObjects, + flagSpec "pkgdb-byte-code" Opt_PackageDbBytecode, flagSpec' "compact-unwind" Opt_CompactUnwind (\turn_on -> updM (\dflags -> do unless (platformOS (targetPlatform dflags) == OSDarwin && turn_on) ===================================== compiler/GHC/Iface/Load.hs ===================================== @@ -672,7 +672,11 @@ dontLeakTheHUG thing_inside = do in hsc_env { hsc_targets = panic "cleanTopEnv: hsc_targets" , hsc_mod_graph = panic "cleanTopEnv: hsc_mod_graph" - , hsc_IC = panic "cleanTopEnv: hsc_IC" + -- TODO this is needed for bytecode compilation of package deps + -- only. local EPS modules work fine. + -- Also it appears to work when the local modules use multiple + -- home units?!?!? + -- , hsc_IC = panic "cleanTopEnv: hsc_IC" , hsc_type_env_vars = case maybe_type_vars of Just vars -> vars Nothing -> panic "cleanTopEnv: hsc_type_env_vars" ===================================== compiler/GHC/Linker/Deps.hs ===================================== @@ -6,6 +6,7 @@ {-# LANGUAGE CPP #-} {-# LANGUAGE RecordWildCards #-} {-# LANGUAGE LambdaCase #-} +{-# LANGUAGE NamedFieldPuns #-} module GHC.Linker.Deps ( LinkDepsOpts (..) @@ -28,6 +29,7 @@ import GHC.Types.Unique.DSet import GHC.Types.Unique.DFM import GHC.Utils.Outputable +import qualified GHC.Utils.Outputable as Outputable import GHC.Utils.Panic import GHC.Utils.Error @@ -47,9 +49,11 @@ import GHC.Utils.Misc import GHC.Unit.Home import GHC.Data.Maybe -import Control.Monad import Control.Applicative +import Control.Monad.IO.Class (MonadIO (liftIO)) +import Control.Monad.Trans.Except (ExceptT, runExceptT, throwE) +import Data.Foldable (traverse_) import qualified Data.Set as Set import qualified Data.Map as M import Data.List (isSuffixOf) @@ -64,19 +68,21 @@ data LinkDepsOpts = LinkDepsOpts , ldUnitEnv :: !UnitEnv , ldPprOpts :: !SDocContext -- ^ Rendering options for error messages , ldUseByteCode :: !Bool -- ^ Use bytecode rather than objects + , ldPkgByteCode :: !Bool -- ^ Use bytecode for external packages , ldMsgOpts :: !(DiagnosticOpts IfaceMessage) -- ^ Options for diagnostics , ldWays :: !Ways -- ^ Enabled ways , ldFinderCache :: !FinderCache , ldFinderOpts :: !FinderOpts - , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError ModIface)) - , ldLoadByteCode :: !(Module -> IO (Maybe Linkable)) + , ldLoadIface :: !(SDoc -> Module -> IO (MaybeErr MissingInterfaceError (ModIface, ModLocation))) + , ldLoadByteCode :: !(Module -> IO (Maybe (IO Linkable))) + , ldDebugTrace :: !(SDoc -> IO ()) } data LinkDeps = LinkDeps { ldNeededLinkables :: [Linkable] , ldAllLinkables :: [Linkable] - , ldUnits :: [UnitId] - , ldNeededUnits :: UniqDSet UnitId + , ldNeededUnits :: [UnitId] + , ldAllUnits :: UniqDSet UnitId } -- | Find all the packages and linkables that a set of modules depends on @@ -102,7 +108,14 @@ getLinkDeps opts interp pls span mods = do get_link_deps opts pls maybe_normal_osuf span mods - +-- | Compute the linkables for the given module set's dependencies. +-- +-- Home modules in make mode are treated separately in a preprocessing step, +-- then all the remaining external deps are processed for both modes. +-- If bytecode is available, transitive external deps are included, otherwise +-- the module's library is linked and processing stops. +-- +-- The results are split into sets of needed/loaded modules/packages. get_link_deps :: LinkDepsOpts -> LoaderState @@ -111,46 +124,40 @@ get_link_deps -> [Module] -> IO LinkDeps get_link_deps opts pls maybe_normal_osuf span mods = do - -- 1. Find the dependent home-pkg-modules/packages from each iface - -- (omitting modules from the interactive package, which is already linked) - (mods_s, pkgs_s) <- - -- Why two code paths here? There is a significant amount of repeated work - -- performed calculating transitive dependencies - -- if --make uses the oneShot code path (see MultiLayerModulesTH_* tests) - if ldOneShotMode opts - then follow_deps (filterOut isInteractiveModule mods) - emptyUniqDSet emptyUniqDSet; - else do - (pkgs, mmods) <- unzip <$> mapM get_mod_info all_home_mods - return (catMaybes mmods, unionManyUniqDSets (init_pkg_set : pkgs)) - - let - -- 2. Exclude ones already linked - -- Main reason: avoid findModule calls in get_linkable - (mods_needed, links_got) = partitionWith split_mods mods_s - pkgs_needed = eltsUDFM $ getUniqDSet pkgs_s `minusUDFM` pkgs_loaded pls - - split_mods mod = - let is_linked = lookupModuleEnv (objs_loaded pls) mod - <|> lookupModuleEnv (bcos_loaded pls) mod - in case is_linked of - Just linkable -> Right linkable - Nothing -> Left mod - - -- 3. For each dependent module, find its linkable - -- This will either be in the HPT or (in the case of one-shot - -- compilation) we may need to use maybe_getFileLinkable - lnks_needed <- mapM (get_linkable (ldObjSuffix opts)) mods_needed - - return $ LinkDeps - { ldNeededLinkables = lnks_needed - , ldAllLinkables = links_got ++ lnks_needed - , ldUnits = pkgs_needed - , ldNeededUnits = pkgs_s - } + (link_deps_home, module_deps_external) <- separate_home_deps + link_deps_external <- external_deps opts module_deps_external + let (loaded_modules, needed_modules, ldAllUnits, ldNeededUnits) = + classify_deps pls (link_deps_home ++ link_deps_external) + ldNeededLinkables <- mapM module_linkable needed_modules + pure LinkDeps { + ldNeededLinkables, + ldAllLinkables = loaded_modules ++ ldNeededLinkables, + ldNeededUnits, + ldAllUnits + } where mod_graph = ldModuleGraph opts unit_env = ldUnitEnv opts + noninteractive = filterOut isInteractiveModule mods + + -- Preprocess the dependencies in make mode to remove all home modules, + -- since the transitive dependency closure is already cached for those in + -- the HUG (see MultiLayerModulesTH_* tests for the performance impact). + -- + -- Returns the remaining, external, dependencies on the right, which is the + -- entire set for oneshot mode. + separate_home_deps = + if ldOneShotMode opts + then pure ([], noninteractive) + else make_deps + + make_deps = do + (dep_ext, mmods) <- unzip <$> mapM get_mod_info all_home_mods + let + link_mods = + listToUDFM [(moduleName (mi_module (hm_iface m)), m) | m <- mmods] + ext = uniqDSetToList (unionManyUniqDSets (init_ext : dep_ext)) + pure ([LinkModules (LinkHomeModule <$> link_mods)], ext) -- This code is used in `--make` mode to calculate the home package and unit dependencies -- for a set of modules. @@ -160,9 +167,9 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- It is also a matter of correctness to use the module graph so that dependencies between home units -- is resolved correctly. - make_deps_loop :: (UniqDSet UnitId, Set.Set NodeKey) -> [ModNodeKeyWithUid] -> (UniqDSet UnitId, Set.Set NodeKey) + make_deps_loop :: (UniqDSet Module, Set.Set NodeKey) -> [ModNodeKeyWithUid] -> (UniqDSet Module, Set.Set NodeKey) make_deps_loop found [] = found - make_deps_loop found@(found_units, found_mods) (nk:nexts) + make_deps_loop found@(external, found_mods) (nk:nexts) | NodeKey_Module nk `Set.member` found_mods = make_deps_loop found nexts | otherwise = case M.lookup (NodeKey_Module nk) (mgTransDeps mod_graph) of @@ -171,85 +178,26 @@ get_link_deps opts pls maybe_normal_osuf span mods = do -- See #936 and the ghci.prog007 test for why we have to continue traversing through -- boot modules. todo_boot_mods = [ModNodeKeyWithUid (GWIB mn NotBoot) uid | NodeKey_Module (ModNodeKeyWithUid (GWIB mn IsBoot) uid) <- Set.toList trans_deps] - in make_deps_loop (found_units, deps `Set.union` found_mods) (todo_boot_mods ++ nexts) + in make_deps_loop (external, deps `Set.union` found_mods) (todo_boot_mods ++ nexts) Nothing -> - let (ModNodeKeyWithUid _ uid) = nk - in make_deps_loop (addOneToUniqDSet found_units uid, found_mods) nexts + let (ModNodeKeyWithUid (GWIB mod_name _) uid) = nk + in make_deps_loop (addOneToUniqDSet external (Module (RealUnit (Definite uid)) mod_name), found_mods) nexts mkNk m = ModNodeKeyWithUid (GWIB (moduleName m) NotBoot) (moduleUnitId m) - (init_pkg_set, all_deps) = make_deps_loop (emptyUniqDSet, Set.empty) $ map mkNk (filterOut isInteractiveModule mods) + (init_ext, all_deps) = make_deps_loop (emptyUniqDSet, Set.empty) $ map mkNk noninteractive all_home_mods = [with_uid | NodeKey_Module with_uid <- Set.toList all_deps] get_mod_info (ModNodeKeyWithUid gwib uid) = case lookupHug (ue_home_unit_graph unit_env) uid (gwib_mod gwib) of - Just hmi -> - let iface = (hm_iface hmi) - mmod = case mi_hsc_src iface of - HsBootFile -> link_boot_mod_error (mi_module iface) - _ -> return $ Just (mi_module iface) - - in (mkUniqDSet $ Set.toList $ dep_direct_pkgs (mi_deps iface),) <$> mmod + Just hmi -> do + let iface = hm_iface hmi + case mi_hsc_src iface of + HsBootFile -> throwProgramError opts $ link_boot_mod_error (mi_module iface) + _ -> pure (mkUniqDSet $ [usg_mod | UsagePackageModule {usg_mod} <- mi_usages iface], hmi) Nothing -> throwProgramError opts $ text "getLinkDeps: Home module not loaded" <+> ppr (gwib_mod gwib) <+> ppr uid - - -- This code is used in one-shot mode to traverse downwards through the HPT - -- to find all link dependencies. - -- The ModIface contains the transitive closure of the module dependencies - -- within the current package, *except* for boot modules: if we encounter - -- a boot module, we have to find its real interface and discover the - -- dependencies of that. Hence we need to traverse the dependency - -- tree recursively. See bug #936, testcase ghci/prog007. - follow_deps :: [Module] -- modules to follow - -> UniqDSet Module -- accum. module dependencies - -> UniqDSet UnitId -- accum. package dependencies - -> IO ([Module], UniqDSet UnitId) -- result - follow_deps [] acc_mods acc_pkgs - = return (uniqDSetToList acc_mods, acc_pkgs) - follow_deps (mod:mods) acc_mods acc_pkgs - = do - mb_iface <- ldLoadIface opts msg mod - iface <- case mb_iface of - Failed err -> throwProgramError opts $ - missingInterfaceErrorDiagnostic (ldMsgOpts opts) err - Succeeded iface -> return iface - - when (mi_boot iface == IsBoot) $ link_boot_mod_error mod - - let - pkg = moduleUnit mod - deps = mi_deps iface - - pkg_deps = dep_direct_pkgs deps - (boot_deps, mod_deps) = flip partitionWith (Set.toList (dep_direct_mods deps)) $ - \case - (_, GWIB m IsBoot) -> Left m - (_, GWIB m NotBoot) -> Right m - - mod_deps' = case ue_homeUnit unit_env of - Nothing -> [] - Just home_unit -> filter (not . (`elementOfUniqDSet` acc_mods)) (map (mkHomeModule home_unit) $ (boot_deps ++ mod_deps)) - acc_mods' = case ue_homeUnit unit_env of - Nothing -> acc_mods - Just home_unit -> addListToUniqDSet acc_mods (mod : map (mkHomeModule home_unit) mod_deps) - acc_pkgs' = addListToUniqDSet acc_pkgs (Set.toList pkg_deps) - - case ue_homeUnit unit_env of - Just home_unit | isHomeUnit home_unit pkg -> follow_deps (mod_deps' ++ mods) - acc_mods' acc_pkgs' - _ -> follow_deps mods acc_mods (addOneToUniqDSet acc_pkgs' (toUnitId pkg)) - where - msg = text "need to link module" <+> ppr mod <+> - text "due to use of Template Haskell" - - - - link_boot_mod_error :: Module -> IO a - link_boot_mod_error mod = throwProgramError opts $ - text "module" <+> ppr mod <+> - text "cannot be linked; it is only available as a boot module" - no_obj :: Outputable a => a -> IO b no_obj mod = dieWith opts span $ text "cannot find object file for module " <> @@ -258,6 +206,20 @@ get_link_deps opts pls maybe_normal_osuf span mods = do while_linking_expr = text "while linking an interpreted expression" + -- Extract the 'Linkable's for unlinked modules from the intermediate + -- results. + module_linkable = \case + LinkHomeModule hmi -> + adjust_linkable (expectJust "getLinkDeps" (homeModLinkable hmi)) + + LinkObjectModule iface loc -> do + let mod = mi_module iface + findObjectLinkableMaybe mod loc >>= \case + Nothing -> no_obj mod + Just lnk -> adjust_linkable lnk + + LinkByteCodeModule _ load_bytecode -> + load_bytecode -- See Note [Using Byte Code rather than Object Code for Template Haskell] homeModLinkable :: HomeModInfo -> Maybe Linkable @@ -266,57 +228,257 @@ get_link_deps opts pls maybe_normal_osuf span mods = do then homeModInfoByteCode hmi <|> homeModInfoObject hmi else homeModInfoObject hmi <|> homeModInfoByteCode hmi - get_linkable osuf mod -- A home-package module - | Just mod_info <- lookupHugByModule mod (ue_home_unit_graph unit_env) - = adjust_linkable (expectJust "getLinkDeps" (homeModLinkable mod_info)) - | otherwise - = do -- It's not in the HPT because we are in one shot mode, - -- so use the Finder to get a ModLocation... - case ue_homeUnit unit_env of - Nothing -> no_obj mod - Just home_unit -> do - from_bc <- ldLoadByteCode opts mod - maybe (fallback_no_bytecode home_unit mod) pure from_bc - where - - fallback_no_bytecode home_unit mod = do - let fc = ldFinderCache opts - let fopts = ldFinderOpts opts - mb_stuff <- findHomeModule fc fopts home_unit (moduleName mod) - case mb_stuff of - Found loc _ -> do - mb_lnk <- findObjectLinkableMaybe mod loc - case mb_lnk of - Nothing -> no_obj mod - Just lnk -> adjust_linkable lnk - _ -> no_obj (moduleName mod) - - adjust_linkable lnk - | Just new_osuf <- maybe_normal_osuf = do - new_parts <- mapM (adjust_part new_osuf) - (linkableParts lnk) - return lnk{ linkableParts=new_parts } - | otherwise = - return lnk - - adjust_part new_osuf part = case part of - DotO file ModuleObject -> do - massert (osuf `isSuffixOf` file) - let file_base = fromJust (stripExtension osuf file) - new_file = file_base <.> new_osuf - ok <- doesFileExist new_file - if (not ok) - then dieWith opts span $ - text "cannot find object file " - <> quotes (text new_file) $$ while_linking_expr - else return (DotO new_file ModuleObject) - DotO file ForeignObject -> pure (DotO file ForeignObject) - DotA fp -> panic ("adjust_ul DotA " ++ show fp) - DotDLL fp -> panic ("adjust_ul DotDLL " ++ show fp) - BCOs {} -> pure part - LazyBCOs{} -> pure part - CoreBindings WholeCoreBindings {wcb_module} -> - pprPanic "Unhydrated core bindings" (ppr wcb_module) + adjust_linkable lnk + | Just new_osuf <- maybe_normal_osuf = do + new_uls <- mapM (adjust_part (ldObjSuffix opts) new_osuf) + (linkableParts lnk) + return lnk {linkableParts = new_uls} + | otherwise = + return lnk + + adjust_part osuf new_osuf part = case part of + DotO file ModuleObject -> do + massert (osuf `isSuffixOf` file) + let file_base = fromJust (stripExtension osuf file) + new_file = file_base <.> new_osuf + ok <- doesFileExist new_file + if (not ok) + then dieWith opts span $ + text "cannot find object file " + <> quotes (text new_file) $$ while_linking_expr + else return (DotO new_file ModuleObject) + DotO file ForeignObject -> pure (DotO file ForeignObject) + DotA fp -> panic ("adjust_part DotA " ++ show fp) + DotDLL fp -> panic ("adjust_part DotDLL " ++ show fp) + BCOs {} -> pure part + LazyBCOs{} -> pure part + CoreBindings WholeCoreBindings {wcb_module} -> + pprPanic "Unhydrated core bindings" (ppr wcb_module) + +data LinkModule = + LinkHomeModule !HomeModInfo + | + LinkObjectModule !ModIface !ModLocation + | + LinkByteCodeModule !ModIface !(IO Linkable) + +link_module_iface :: LinkModule -> ModIface +link_module_iface = \case + LinkHomeModule hmi -> hm_iface hmi + LinkObjectModule iface _ -> iface + LinkByteCodeModule iface _ -> iface + +instance Outputable LinkModule where + ppr = \case + LinkHomeModule hmi -> ppr (mi_module (hm_iface hmi)) <+> brackets (text "HMI") + LinkObjectModule iface _ -> ppr (mi_module iface) + LinkByteCodeModule iface _ -> ppr (mi_module iface) <+> brackets (text "BC") + +data LinkDep = + LinkModules !(UniqDFM ModuleName LinkModule) + | + LinkLibrary !UnitId + +instance Outputable LinkDep where + ppr = \case + LinkModules mods -> text "modules:" <+> ppr (eltsUDFM mods) + LinkLibrary uid -> text "library:" <+> ppr uid + +data OneshotError = + NoLocation !Module + | + NoInterface !MissingInterfaceError + | + LinkBootModule !Module + +-- Compute the transitive dependency closure of the given modules. +-- +-- Used for all oneshot mode dependencies and for external dependencies of home +-- modules in make mode. +-- +-- TODO is the following still relevant? +-- The ModIface contains the transitive closure of the module dependencies +-- within the current package, *except* for boot modules: if we encounter +-- a boot module, we have to find its real interface and discover the +-- dependencies of that. Hence we need to traverse the dependency +-- tree recursively. See bug #936, testcase ghci/prog007. +external_deps :: + LinkDepsOpts -> + -- | Modules whose imports to follow + [Module] -> + IO [LinkDep] +external_deps opts mods = + runExceptT (external_deps_loop opts mods emptyUDFM) >>= \case + Right a -> pure (eltsUDFM a) + Left err -> throwProgramError opts (message err) + where + message = \case + NoLocation mod -> + pprPanic "found iface but no location" (ppr mod) + NoInterface err -> + missingInterfaceErrorDiagnostic (ldMsgOpts opts) err + LinkBootModule mod -> + link_boot_mod_error mod + +external_deps_loop :: + LinkDepsOpts -> + [Module] -> + UniqDFM UnitId LinkDep -> + ExceptT OneshotError IO (UniqDFM UnitId LinkDep) +external_deps_loop _ [] acc = + pure acc +external_deps_loop opts (mod : mods) acc = do + (new_acc, new_mods, action) <- process_module + traverse_ debug_log action + external_deps_loop opts (new_mods ++ mods) new_acc + where + debug_log action = + liftIO $ ldDebugTrace opts $ + text "TH dep" <+> ppr mod <+> brackets (sep [ + if is_home then text "home" else Outputable.empty, + text action + ]) + + -- Decide how this module needs to be processed. + -- We only need an interface if we want to load bytecode or if we have to + -- link an object file (which happens for home unit modules, since those + -- have no libraries). + process_module + | already_seen = pure (acc, [], Nothing) + | is_home || prefer_bytecode = try_iface + | otherwise = add_library + + -- Check whether the current module was processed before. + -- Since the accumulator is keyed by unit ID, we have to perform two + -- lookups. + -- If another module from this module's unit has been determined to be + -- linked as a library previously, we skip this module, assuming that no + -- bytecode is available for the entire package. + already_seen + | Just (LinkModules mods) <- mod_dep + = elemUDFM mod_name mods + | Just (LinkLibrary _) <- mod_dep + = True + | otherwise + = False + + -- Load the iface and attempt to get bytecode from Core bindings. + try_iface = + liftIO (ldLoadIface opts load_reason mod) >>= \case + Failed err -> throwE (NoInterface err) + Succeeded (iface, loc) -> do + mb_load_bc <- liftIO (ldLoadByteCode opts (mi_module iface)) + with_iface loc iface mb_load_bc + + -- Decide how to link this module. + -- If bytecode or an object file is available, use those in that order. + -- Otherwise fall back to linking a library. + with_iface loc iface mb_load_bc + | IsBoot <- mi_boot iface + = throwE (LinkBootModule mod) + + | prefer_bytecode + , Just load_bc <- mb_load_bc + = pure (add_module iface (LinkByteCodeModule iface load_bc) "bytecode") + + | is_home + = pure (add_module iface (LinkObjectModule iface loc) "object") + + | otherwise + = add_library + + add_library = + pure (addToUDFM acc mod_unit_id (LinkLibrary mod_unit_id), [], Just "library") + + add_module iface lmod action = + (addListToUDFM with_mod (direct_pkgs iface), new_deps iface, Just action) + where + with_mod = alterUDFM (add_package_module lmod) acc mod_unit_id + + add_package_module lmod = \case + Just (LinkLibrary u) -> Just (LinkLibrary u) + Just (LinkModules old) -> Just (LinkModules (addToUDFM old mod_name lmod)) + Nothing -> Just (LinkModules (unitUDFM mod_name lmod)) + + direct_pkgs iface + | prefer_bytecode + = [] + | otherwise + = [(u, LinkLibrary u) | u <- Set.toList (dep_direct_pkgs (mi_deps iface))] + + new_deps iface + | prefer_bytecode + -- TODO How can we better determine the external deps? + -- OTOH, we probably don't want to link unused dependencies anyway. + = [usg_mod | UsagePackageModule {usg_mod} <- mi_usages iface] ++ local + | is_home + = local + | otherwise + = [] + where + local = + [ + mkModule mod_unit m + -- TODO Somehow this just works, no idea what the deal was in the + -- old code with boot modules. + | (_, GWIB m _) <- Set.toList (dep_direct_mods (mi_deps iface)) + ] + + is_home + | Just home <- mb_home + = homeUnitAsUnit home == mod_unit + | otherwise + = False + + mod_dep = lookupUDFM acc mod_unit_id + mod_name = moduleName mod + mod_unit_id = moduleUnitId mod + mod_unit = moduleUnit mod + load_reason = + text "need to link module" <+> ppr mod <+> + text "due to use of Template Haskell" + + prefer_bytecode = ldUseByteCode opts && (is_home || ldPkgByteCode opts) + mb_home = ue_homeUnit (ldUnitEnv opts) + +link_boot_mod_error :: Module -> SDoc +link_boot_mod_error mod = + text "module" <+> ppr mod <+> + text "cannot be linked; it is only available as a boot module" + +-- | Split link dependencies into the sets of modules and packages that have +-- been linked previously and those that need to be linked now by checking for +-- their presence in the 'LoaderState': +-- +-- - For module dependencies, in the sets of loaded objects and BCOs +-- ('objs_loaded' and 'bcos_loaded') +-- - For package dependencies, in the set of loaded packages ('pkgs_loaded') +classify_deps :: + LoaderState -> + [LinkDep] -> + ([Linkable], [LinkModule], UniqDSet UnitId, [UnitId]) +classify_deps pls deps = + (loaded_modules, needed_modules, all_packages, needed_packages) + where + (loaded_modules, needed_modules) = + partitionWith loaded_or_needed_module (concatMap eltsUDFM modules) + + needed_packages = + eltsUDFM (getUniqDSet all_packages `minusUDFM` pkgs_loaded pls) + + all_packages = mkUniqDSet packages + + (modules, packages) = flip partitionWith deps $ \case + LinkModules mods -> Left mods + LinkLibrary lib -> Right lib + + loaded_or_needed_module lm = + maybe (Right lm) Left (loaded_module (mi_module (link_module_iface lm))) + + loaded_module mod = + lookupModuleEnv (objs_loaded pls) mod + <|> + lookupModuleEnv (bcos_loaded pls) mod {- Note [Using Byte Code rather than Object Code for Template Haskell] ===================================== compiler/GHC/Linker/Loader.hs ===================================== @@ -77,11 +77,13 @@ import GHC.Utils.TmpFs import GHC.Unit.Env import GHC.Unit.External (ExternalPackageState (EPS, eps_iface_bytecode)) +import GHC.Unit.Finder import GHC.Unit.Module import GHC.Unit.State as Packages import qualified GHC.Data.ShortText as ST import GHC.Data.FastString +import qualified GHC.Data.Maybe as Maybe import GHC.Linker.Deps import GHC.Linker.MacOS @@ -94,6 +96,7 @@ import Control.Monad import qualified Data.Set as Set import Data.Char (isSpace) import qualified Data.Foldable as Foldable +import Data.Functor ((<&>)) import Data.IORef import Data.List (intercalate, isPrefixOf, nub, partition) import Data.Maybe @@ -231,10 +234,10 @@ loadDependencies interp hsc_env pls span needed_mods = do -- Find what packages and linkables are required deps <- getLinkDeps opts interp pls span needed_mods - let this_pkgs_needed = ldNeededUnits deps + let this_pkgs_needed = ldAllUnits deps -- Link the packages and modules required - pls1 <- loadPackages' interp hsc_env (ldUnits deps) pls + pls1 <- loadPackages' interp hsc_env (ldNeededUnits deps) pls (pls2, succ) <- loadModuleLinkables interp hsc_env pls1 (ldNeededLinkables deps) let this_pkgs_loaded = udfmRestrictKeys all_pkgs_loaded $ getUniqDSet trans_pkgs_needed all_pkgs_loaded = pkgs_loaded pls2 @@ -646,18 +649,39 @@ initLinkDepsOpts hsc_env = opts , ldFinderCache = hsc_FC hsc_env , ldFinderOpts = initFinderOpts dflags , ldUseByteCode = gopt Opt_UseBytecodeRatherThanObjects dflags + , ldPkgByteCode = gopt Opt_PackageDbBytecode dflags , ldMsgOpts = initIfaceMessageOpts dflags , ldWays = ways dflags , ldLoadIface , ldLoadByteCode + , ldDebugTrace = debugTraceMsg (hsc_logger hsc_env) 3 } dflags = hsc_dflags hsc_env - ldLoadIface msg mod = initIfaceCheck (text "loader") hsc_env - $ loadInterface msg mod (ImportByUser NotBoot) + + ldLoadIface msg mod = + initIfaceCheck (text "loader") hsc_env (loadInterface msg mod (ImportByUser NotBoot)) >>= \case + Maybe.Failed err -> pure (Maybe.Failed err) + Maybe.Succeeded iface -> + find_location mod <&> \case + InstalledFound loc _ -> Maybe.Succeeded (iface, loc) + err -> Maybe.Failed $ + cannotFindInterface unit_state home_unit + (targetProfile dflags) (moduleName mod) err + + find_location mod = + liftIO $ + findExactModule (hsc_FC hsc_env) (initFinderOpts dflags) + (initFinderOpts . homeUnitEnv_dflags <$> hsc_HUG hsc_env) + unit_state home_unit + (toUnitId <$> mod) + + unit_state = hsc_units hsc_env + + home_unit = ue_homeUnit (hsc_unit_env hsc_env) ldLoadByteCode mod = do EPS {eps_iface_bytecode} <- hscEPS hsc_env - sequence (lookupModuleEnv eps_iface_bytecode mod) + pure (lookupModuleEnv eps_iface_bytecode mod) ===================================== compiler/GHC/Unit/Finder.hs ===================================== @@ -734,7 +734,7 @@ mkStubPaths fopts mod location = do src_basename = OsPath.dropExtension <$> ml_hs_file_ospath location -- ----------------------------------------------------------------------------- --- findLinkable isn't related to the other stuff in here, +-- findObjectLinkable isn't related to the other stuff in here, -- but there's no other obvious place for it findObjectLinkableMaybe :: Module -> ModLocation -> IO (Maybe Linkable) ===================================== compiler/GHC/Unit/Module/WholeCoreBindings.hs ===================================== @@ -88,9 +88,10 @@ settings. 2. In oneshot mode, which compiles individual modules without a shared home unit graph, a previously compiled module is not reprocessed as described for make mode above. - When 'get_link_deps' encounters a dependency on a local module, it requests - its bytecode from the External Package State, who loads the interface - on-demand. + 'get_link_deps' requests the bytecode of dependencies from the External + Package State, who loads the interface on-demand. + This works for modules in local directories (via @-i@ and @-hidir@) as well + as those exposed from a package DB. Since the EPS stores interfaces for all package dependencies in addition to local modules in oneshot mode, it has a substantial memory footprint. @@ -210,7 +211,9 @@ If the 'HomeModLinkable' already contains bytecode (case 1), this is a no-op. Otherwise, the stub objects from the interface are compiled to objects in 'generateByteCode' and added to the 'HomeModLinkable' as well. -Case 3 is not implemented yet (!13042). +In case 3, Core bindings are loaded from the EPS, where stubs only exist in +their serialized form in the interface, so they must be regenerated like in case +2. Problem 3: ===================================== testsuite/tests/bytecode/T25090/Dep.hs ===================================== @@ -0,0 +1,15 @@ +module Dep where + +data A = A Int + +used :: Int +used = 9681 + +dep :: A +dep = A used + +unused1 :: A +unused1 = A 1 + +unused2 :: A +unused2 = unused1 ===================================== testsuite/tests/bytecode/T25090/DepApi.hs ===================================== @@ -0,0 +1,7 @@ +module DepApi (A (A), dep) where + +import Dep (A (A)) +import qualified Dep + +dep :: A +dep = Dep.dep ===================================== testsuite/tests/bytecode/T25090/Local.hs ===================================== @@ -0,0 +1,15 @@ +{-# language PackageImports #-} + +module Local where + +import GHC.Prim +import Language.Haskell.TH (ExpQ) +import Language.Haskell.TH.Syntax (lift) +-- just to be sure that the file isn't accidentally picked up locally +import "dep" DepApi (dep, A (A)) +import {-# source #-} Num (num) + +splc :: ExpQ +splc = lift @_ @Int (num + d) + where + A d = dep ===================================== testsuite/tests/bytecode/T25090/Makefile ===================================== @@ -19,3 +19,35 @@ T25090a: T25090b: $(TEST_HC) -fbyte-code-and-object-code -fprefer-byte-code A -o exe -v0 ./exe + +DB := -package-db db -package dep +BASIC := $(TEST_HC_OPTS) $(DB) -this-unit-id=pkgdep -v0 +BC := -fprefer-byte-code -fbyte-code-and-object-code +ARGS := $(BASIC) $(BC) + +T25090_pkg: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + ./run.bash "$(TEST_HC)" "$(ARGS) -dynamic" + +T25090_pkg_empty: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared-empty" + ./run.bash "$(TEST_HC)" "$(ARGS) -dynamic" + +T25090_pkg_nolib: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "none" + ./run.bash "$(TEST_HC)" "$(ARGS)" + +T25090_pkg_obj_code: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + ./run.bash "$(TEST_HC)" "$(BASIC) -dynamic -fbyte-code-and-object-code" + +T25090_pkg_multi_unit: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "shared" + mkdir -p unit2-src/ + mv Local.hs Num.hs Num.hs-boot unit2-src/ + "$(TEST_HC)" $(TEST_HC_OPTS) $(ARGS) -dynamic -unit @unit1 -unit @unit2 + ./PkgBytecode + +T25090_pkg_archive: + ./prep.bash "$(TEST_HC)" "$(TEST_HC_OPTS)" "$(GHC_PKG)" "archive" + ./run.bash "$(TEST_HC)" "$(ARGS)" ===================================== testsuite/tests/bytecode/T25090/Num.hs ===================================== @@ -0,0 +1,4 @@ +module Num where + +num :: Int +num = 48332 ===================================== testsuite/tests/bytecode/T25090/Num.hs-boot ===================================== @@ -0,0 +1,3 @@ +module Num where + +num :: Int ===================================== testsuite/tests/bytecode/T25090/PkgBytecode.hs ===================================== @@ -0,0 +1,12 @@ +{-# language TemplateHaskell #-} + +module Main where + +import GHC.Prim +import Local (splc) + +a :: Int +a = $(splc) + +main :: IO () +main = putStrLn (show a) ===================================== testsuite/tests/bytecode/T25090/PkgBytecode.stdout ===================================== @@ -0,0 +1 @@ +58013 ===================================== testsuite/tests/bytecode/T25090/all.T ===================================== @@ -16,3 +16,33 @@ def test_T25090(name): test_T25090('T25090a') test_T25090('T25090b') + +def test_pkg(name, files = []): + test( + name, + [ + extra_files([ + 'PkgBytecode.hs', + 'Local.hs', + 'Dep.hs', + 'DepApi.hs', + 'Num.hs', + 'Num.hs-boot', + 'dep.conf', + 'prep.bash', + 'run.bash', + ] + files), + use_specs({'stdout': 'PkgBytecode.stdout'}), + ], + makefile_test, + [], + ) + +test_pkg('T25090_pkg') +test_pkg('T25090_pkg_empty') +test_pkg('T25090_pkg_nolib') +test_pkg('T25090_pkg_obj_code') +test_pkg('T25090_pkg_multi_unit', ['unit1', 'unit2']) +# TODO this doesn't work, because `locateLib` ignores static archives when the interpreter is dynamic, even though a +# comment says "search for .so libraries _first_" (rather than "only"). +# test_pkg('T25090_pkg_archive') ===================================== testsuite/tests/bytecode/T25090/dep.conf ===================================== @@ -0,0 +1,8 @@ +name: dep +version: 1.0 +id: dep-1.0 +key: dep-1.0 +exposed: True +exposed-modules: DepApi +import-dirs: ${pkgroot}/dep +library-dirs: ${pkgroot}/dep ===================================== testsuite/tests/bytecode/T25090/prep.bash ===================================== @@ -0,0 +1,57 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_opts="$2" +ghc_pkg_cmd="$3" +library="$4" + +base="$PWD" +db="$base/db" +dep="$base/dep" +conf_dep="${dep}/dep.conf" + +mkdir -p "$dep" "$db" +mv Dep.hs DepApi.hs "$dep/" +cp dep.conf "$dep/" + +ghc_pkg() +{ + eval "${ghc_pkg_cmd at Q} --no-user-package-db --package-db=${db at Q} $@" +} + +ghc() +{ + eval "${ghc_cmd at Q} ${ghc_opts/-rtsopts/} -package-db ${db at Q} -hidir ${dep at Q} -O0 -this-unit-id dep-1.0 -fbyte-code-and-object-code $@" +} + +version=$(ghc "--numeric-version") + +ghc_pkg recache + +ghc "-dynamic-too -c ${dep at Q}/Dep.hs ${dep at Q}/DepApi.hs" + +if [[ "$library" == 'shared' ]] +then + ghc "-dynamic -shared -o ${dep at Q}/libHSdep-1.0-ghc$version.so ${dep at Q}/Dep.dyn_o ${dep at Q}/DepApi.dyn_o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'shared-empty' ]] +then + echo 'module Dummy where' > Dummy.hs + ghc "-dynamic-too -c Dummy.hs" + ghc "-dynamic -shared -o ${dep at Q}/libHSdep-1.0-ghc$version.so Dummy.dyn_o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'archive' ]] +then + $AR cqs "${dep}/libHSdep-1.0.a" "${dep}/Dep.o" "${dep}/DepApi.o" + echo 'hs-libraries: HSdep-1.0' >> "$conf_dep" +elif [[ "$library" == 'none' ]] +then + : +else + echo "Invalid argument for 'library': $library" + exit 1 +fi + +ghc_pkg -v0 register "${conf_dep at Q}" ===================================== testsuite/tests/bytecode/T25090/run.bash ===================================== @@ -0,0 +1,16 @@ +#!/usr/bin/env bash + +set -eu + +ghc_cmd="$1" +ghc_opts="$2" + +ghc() +{ + eval "${ghc_cmd at Q} $ghc_opts $@" +} + +ghc -c Num.hs-boot Num.hs Local.hs +ghc -fpkgdb-byte-code -c PkgBytecode.hs +ghc PkgBytecode.o -o PkgBytecode +./PkgBytecode ===================================== testsuite/tests/bytecode/T25090/unit1 ===================================== @@ -0,0 +1 @@ +-i -i. PkgBytecode -this-unit-id unit1 -package-id unit2 ===================================== testsuite/tests/bytecode/T25090/unit2 ===================================== @@ -0,0 +1 @@ +-i -i./unit2-src Local Num -this-unit-id unit2 View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d9e3420952bd885cce867a64d5a8d8e61f9d3dc6 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/d9e3420952bd885cce867a64d5a8d8e61f9d3dc6 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 19:08:31 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Mon, 30 Sep 2024 15:08:31 -0400 Subject: [Git][ghc/ghc] Pushed new branch wip/backports-9.8 Message-ID: <66faf72fa13cf_161302100d24566bf@gitlab.mail> Ben Gamari pushed new branch wip/backports-9.8 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/backports-9.8 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 20:11:10 2024 From: gitlab at gitlab.haskell.org (Ben Gamari (@bgamari)) Date: Mon, 30 Sep 2024 16:11:10 -0400 Subject: [Git][ghc/ghc][wip/backports-9.8] 14 commits: JS: reenable h$appendToHsString optimization (#24495) Message-ID: <66fb05de5cc85_1613024953686607c@gitlab.mail> Ben Gamari pushed to branch wip/backports-9.8 at Glasgow Haskell Compiler / GHC Commits: bcdf3e2c by Sylvain Henry at 2024-09-30T16:10:28-04:00 JS: reenable h$appendToHsString optimization (#24495) The optimization introducing h$appendToHsString wasn't kicking in anymore (while it did in 9.8.1) because of the changes introduced in #23270 (7e0c8b3bab30). This patch reenables the optimization by matching on case-expression, as done in Cmm for unpackCString# standard thunks. The test is also T24495 added in the next commits (two commits for ease of backporting to 9.8). (cherry picked from commit b36ee57bfbecc628b7f0919e1e59b7066495034f) - - - - - e16be40c by Luite Stegeman at 2024-09-30T16:10:28-04:00 Update correct counter in bumpTickyAllocd (cherry picked from commit 0c4a96862081f03e2946a2ed7e80c108f06205a1) - - - - - 10a6aa18 by Rodrigo Mesquita at 2024-09-30T16:10:28-04:00 configure: Use LDFLAGS when trying linkers A user may configure `LDFLAGS` but not `LD`. When choosing a linker, we will prefer `ldd`, then `ld.gold`, then `ld.bfd` -- however, we have to check for a working linker. If either of these fail, we try the next in line. However, we were not considering the `$LDFLAGS` when checking if these linkers worked. So we would pick a linker that does not support the current $LDFLAGS and fail further down the line when we used that linker with those flags. Fixes #24565, where `LDFLAGS=-Wl,-z,pack-relative-relocs` is not supported by `ld.gold` but that was being picked still. (cherry picked from commit 32a8103f3b3e22907fdd67b69c919c5251d8cc20) - - - - - 067aa6e0 by Zubin Duggal at 2024-09-30T16:10:28-04:00 driver: Make `checkHomeUnitsClosed` faster The implementation of `checkHomeUnitsClosed` was traversing every single path in the unit dependency graph - this grows exponentially and quickly grows to be infeasible on larger unit dependency graphs. Instead we replace this with a faster implementation which follows from the specificiation of the closure property - there is a closure error if there are units which are both are both (transitively) depended upon by home units and (transitively) depend on home units, but are not themselves home units. To compute the set of units required for closure, we first compute the closure of the unit dependency graph, then the transpose of this closure, and find all units that are reachable from the home units in the transpose of the closure. (cherry picked from commit a933aff37992ea311a60be878379e7abf650e9fb) - - - - - b0a30ae7 by Cheng Shao at 2024-09-30T16:10:28-04:00 rts: fix clang compilation on aarch64 This patch fixes function prototypes in ARMOutlineAtomicsSymbols.h which causes "error: address argument to atomic operation must be a pointer to _Atomic type" when compiling with clang on aarch64. (cherry picked from commit 7db8c9927fae3369fc4ecff68f80c4cb32eea757) - - - - - ad7b46e7 by Cheng Shao at 2024-09-30T16:10:28-04:00 libffi-tarballs: bump libffi-tarballs submodule to libffi 3.4.6 This commit bumps the libffi-tarballs submodule to libffi 3.4.6, which includes numerous upstream libffi fixes, especially https://github.com/libffi/libffi/issues/760. (cherry picked from commit 810660b780e1111b36c91326bcd0041e1f62706b) - - - - - 66148140 by Teo Camarasu at 2024-09-30T16:10:28-04:00 rts: avoid checking bdescr of value outside of Haskell heap In nonmovingTidyWeaks we want to check if the key of a weak pointer lives in the non-moving heap. We do this by checking the flags of the block the key lives in. But we need to be careful with values that live outside the Haskell heap, since they will lack a block descriptor and looking for one may lead to a segfault. In this case we should just accept that it isn't on the non-moving heap. Resolves #24492 (cherry picked from commit 74b24a9b0084459b8aa426a502956bd332b4d0fb) - - - - - 894710fb by Rodrigo Mesquita at 2024-09-30T16:10:28-04:00 configure: Do not override existing linker flags in FP_LD_NO_FIXUP_CHAINS (cherry picked from commit 9460d5044b8c7239fb7a6ce875d084617f159c7f) - - - - - 5b3dc9c4 by Zubin Duggal at 2024-09-30T16:10:28-04:00 rel-eng/fetch-gitlab.py: Fix name of aarch64 alpine 3_18 release job (cherry picked from commit f3de8a3c6f25b2c7eeb2ec6da95be24eeb496914) - - - - - 98b04cea by Andreas Klebinger at 2024-09-30T16:10:28-04:00 x86-ncg: Fix fma codegen when arguments are globals Fix a bug in the x86 ncg where results would be wrong when the desired output register and one of the input registers were the same global. Also adds a tiny optimization to make use of the memory addressing support when convenient. Fixes #24496 (cherry picked from commit 82ccb8012ba532f0fa06dc6ff96d33217560088a) - - - - - 69ed198f by Ian-Woo Kim at 2024-09-30T16:10:28-04:00 Add missing BCO handling in scavenge_one. (cherry picked from commit 902ebcc2b95707319d37a19d6b23c342cc14b162) - - - - - db376284 by Matthew Craven at 2024-09-30T16:10:28-04:00 Add @since annotation to Data.Data.mkConstrTag (cherry picked from commit 249caf0d2345ac2e6046847522bb8f2fff7f8835) - - - - - 2c18ff79 by Vladislav Zavialov at 2024-09-30T16:10:28-04:00 Fix (~) and (@) infix operators in TH splices (#23748) 8168b42a "Whitespace-sensitive bang patterns" allows GHC to accept the following infix operators: a ~ b = () a @ b = () But not if TH is used to generate those declarations: $([d| a ~ b = () a @ b = () |]) -- Test.hs:5:2: error: [GHC-55017] -- Illegal variable name: ‘~’ -- When splicing a TH declaration: (~_0) a_1 b_2 = GHC.Tuple.Prim.() This is easily fixed by modifying `reservedOps` in GHC.Utils.Lexeme (cherry picked from commit 46fd8ced0cc031f2e50a1a4b348738fd39b4a741) - - - - - c6a95d27 by Jade at 2024-09-30T16:10:29-04:00 Enhance documentation of Data.Complex (cherry picked from commit cdd939e77f2bb20f80310280400b3426598b4eca) - - - - - 25 changed files: - .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py - compiler/GHC/CmmToAsm/X86/CodeGen.hs - compiler/GHC/Driver/Make.hs - compiler/GHC/StgToCmm/Ticky.hs - compiler/GHC/StgToJS/Apply.hs - compiler/GHC/StgToJS/Expr.hs - compiler/GHC/Utils/Lexeme.hs - libffi-tarballs - libraries/base/Data/Complex.hs - libraries/base/Data/Data.hs - m4/fp_cc_linker_flag_try.m4 - m4/fp_ld_no_fixup_chains.m4 - rts/ARMOutlineAtomicsSymbols.h - rts/sm/NonMovingMark.c - rts/sm/Scav.c - + testsuite/tests/driver/multipleHomeUnits/mhu-perf/Makefile - + testsuite/tests/driver/multipleHomeUnits/mhu-perf/Setup.hs - + testsuite/tests/driver/multipleHomeUnits/mhu-perf/all.T - + testsuite/tests/driver/multipleHomeUnits/mhu-perf/genLargeHMU - + testsuite/tests/driver/multipleHomeUnits/mhu-perf/mhu-perf.stderr - + testsuite/tests/primops/should_run/T24496.hs - + testsuite/tests/primops/should_run/T24496.stdout - testsuite/tests/primops/should_run/all.T - + testsuite/tests/th/T23748.hs - testsuite/tests/th/all.T Changes: ===================================== .gitlab/rel_eng/fetch-gitlab-artifacts/fetch_gitlab.py ===================================== @@ -40,7 +40,7 @@ def job_triple(job_name): 'release-armv7-linux-deb10-release': 'armv7-deb10-linux', 'release-aarch64-linux-deb10-release': 'aarch64-deb10-linux', 'release-aarch64-linux-deb11-release': 'aarch64-deb11-linux', - 'release-aarch64-linux-alpine_3_18-release': 'aarch64-alpine3_18-linux', + 'release-aarch64-linux-alpine3_18-release+no_split_sections': 'aarch64-alpine3_18-linux', 'release-aarch64-darwin-release': 'aarch64-apple-darwin', 'release-x86_64-linux-deb11-release+boot_nonmoving_gc': 'unused', ===================================== compiler/GHC/CmmToAsm/X86/CodeGen.hs ===================================== @@ -3209,7 +3209,6 @@ genFMA3Code :: Width -> FMASign -> CmmExpr -> CmmExpr -> CmmExpr -> NatM Register genFMA3Code w signs x y z = do - -- For the FMA instruction, we want to compute x * y + z -- -- There are three possible instructions we could emit: @@ -3230,17 +3229,45 @@ genFMA3Code w signs x y z = do -- -- Currently we follow neither of these optimisations, -- opting to always use fmadd213 for simplicity. + -- + -- We would like to compute the result directly into the requested register. + -- To do so we must first compute `x` into the destination register. This is + -- only possible if the other arguments don't use the destination register. + -- We check for this and if there is a conflict we move the result only after + -- the computation. See #24496 how this went wrong in the past. let rep = floatFormat w (y_reg, y_code) <- getNonClobberedReg y - (z_reg, z_code) <- getNonClobberedReg z + (z_op, z_code) <- getNonClobberedOperand z x_code <- getAnyReg x + x_tmp <- getNewRegNat rep let fma213 = FMA3 rep signs FMA213 - code dst - = y_code `appOL` + + code, code_direct, code_mov :: Reg -> InstrBlock + -- Ideal: Compute the result directly into dst + code_direct dst = x_code dst `snocOL` + fma213 z_op y_reg dst + -- Fallback: Compute the result into a tmp reg and then move it. + code_mov dst = x_code x_tmp `snocOL` + fma213 z_op y_reg x_tmp `snocOL` + MOV rep (OpReg x_tmp) (OpReg dst) + + code dst = + y_code `appOL` z_code `appOL` - x_code dst `snocOL` - fma213 (OpReg z_reg) y_reg dst + ( if arg_regs_conflict then code_mov dst else code_direct dst ) + + where + + arg_regs_conflict = + y_reg == dst || + case z_op of + OpReg z_reg -> z_reg == dst + OpAddr amode -> dst `elem` addrModeRegs amode + OpImm {} -> False + + -- NB: Computing the result into a desired register using Any can be tricky. + -- So for now, we keep it simple. (See #24496). return (Any rep code) ----------- ===================================== compiler/GHC/Driver/Make.hs ===================================== @@ -1578,8 +1578,8 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots let (root_errs, rootSummariesOk) = partitionEithers rootSummaries -- #17549 root_map = mkRootMap rootSummariesOk checkDuplicates root_map - (deps, pkg_deps, map0) <- loopSummaries rootSummariesOk (M.empty, Set.empty, root_map) - let closure_errs = checkHomeUnitsClosed (hsc_unit_env hsc_env) (hsc_all_home_unit_ids hsc_env) (Set.toList pkg_deps) + (deps, map0) <- loopSummaries rootSummariesOk (M.empty, root_map) + let closure_errs = checkHomeUnitsClosed (hsc_unit_env hsc_env) let unit_env = hsc_unit_env hsc_env let tmpfs = hsc_tmpfs hsc_env @@ -1673,19 +1673,19 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots -- This loops over all the mod summaries in the dependency graph, accumulates the actual dependencies for each module/unit loopSummaries :: [ModSummary] - -> (M.Map NodeKey ModuleGraphNode, Set.Set (UnitId, UnitId), + -> (M.Map NodeKey ModuleGraphNode, DownsweepCache) - -> IO ((M.Map NodeKey ModuleGraphNode), Set.Set (UnitId, UnitId), DownsweepCache) + -> IO ((M.Map NodeKey ModuleGraphNode), DownsweepCache) loopSummaries [] done = return done - loopSummaries (ms:next) (done, pkgs, summarised) + loopSummaries (ms:next) (done, summarised) | Just {} <- M.lookup k done - = loopSummaries next (done, pkgs, summarised) + = loopSummaries next (done, summarised) -- Didn't work out what the imports mean yet, now do that. | otherwise = do - (final_deps, pkgs1, done', summarised') <- loopImports (calcDeps ms) done summarised + (final_deps, done', summarised') <- loopImports (calcDeps ms) done summarised -- This has the effect of finding a .hs file if we are looking at the .hs-boot file. - (_, _, done'', summarised'') <- loopImports (maybeToList hs_file_for_boot) done' summarised' - loopSummaries next (M.insert k (ModuleNode final_deps ms) done'', pkgs1 `Set.union` pkgs, summarised'') + (_, done'', summarised'') <- loopImports (maybeToList hs_file_for_boot) done' summarised' + loopSummaries next (M.insert k (ModuleNode final_deps ms) done'', summarised'') where k = NodeKey_Module (msKey ms) @@ -1705,18 +1705,17 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots -- Visited set; the range is a list because -- the roots can have the same module names -- if allow_dup_roots is True - -> IO ([NodeKey], Set.Set (UnitId, UnitId), - + -> IO ([NodeKey], M.Map NodeKey ModuleGraphNode, DownsweepCache) -- The result is the completed NodeMap - loopImports [] done summarised = return ([], Set.empty, done, summarised) + loopImports [] done summarised = return ([], done, summarised) loopImports ((home_uid,mb_pkg, gwib) : ss) done summarised | Just summs <- M.lookup cache_key summarised = case summs of [Right ms] -> do let nk = NodeKey_Module (msKey ms) - (rest, pkgs, summarised', done') <- loopImports ss done summarised - return (nk: rest, pkgs, summarised', done') + (rest, summarised', done') <- loopImports ss done summarised + return (nk: rest, summarised', done') [Left _err] -> loopImports ss done summarised _errs -> do @@ -1728,69 +1727,77 @@ downsweep hsc_env old_summaries excl_mods allow_dup_roots Nothing excl_mods case mb_s of NotThere -> loopImports ss done summarised - External uid -> do - (other_deps, pkgs, done', summarised') <- loopImports ss done summarised - return (other_deps, Set.insert (homeUnitId home_unit, uid) pkgs, done', summarised') + External _ -> do + (other_deps, done', summarised') <- loopImports ss done summarised + return (other_deps, done', summarised') FoundInstantiation iud -> do - (other_deps, pkgs, done', summarised') <- loopImports ss done summarised - return (NodeKey_Unit iud : other_deps, pkgs, done', summarised') + (other_deps, done', summarised') <- loopImports ss done summarised + return (NodeKey_Unit iud : other_deps, done', summarised') FoundHomeWithError (_uid, e) -> loopImports ss done (Map.insert cache_key [(Left e)] summarised) FoundHome s -> do - (done', pkgs1, summarised') <- - loopSummaries [s] (done, Set.empty, Map.insert cache_key [Right s] summarised) - (other_deps, pkgs2, final_done, final_summarised) <- loopImports ss done' summarised' + (done', summarised') <- + loopSummaries [s] (done, Map.insert cache_key [Right s] summarised) + (other_deps, final_done, final_summarised) <- loopImports ss done' summarised' -- MP: This assumes that we can only instantiate non home units, which is probably fair enough for now. - return (NodeKey_Module (msKey s) : other_deps, pkgs1 `Set.union` pkgs2, final_done, final_summarised) + return (NodeKey_Module (msKey s) : other_deps, final_done, final_summarised) where cache_key = (home_uid, mb_pkg, unLoc <$> gwib) home_unit = ue_unitHomeUnit home_uid (hsc_unit_env hsc_env) GWIB { gwib_mod = L loc mod, gwib_isBoot = is_boot } = gwib wanted_mod = L loc mod --- This function checks then important property that if both p and q are home units +-- | This function checks then important property that if both p and q are home units -- then any dependency of p, which transitively depends on q is also a home unit. -checkHomeUnitsClosed :: UnitEnv -> Set.Set UnitId -> [(UnitId, UnitId)] -> [DriverMessages] --- Fast path, trivially closed. -checkHomeUnitsClosed ue home_id_set home_imp_ids - | Set.size home_id_set == 1 = [] - | otherwise = - let res = foldMap loop home_imp_ids - -- Now check whether everything which transitively depends on a home_unit is actually a home_unit - -- These units are the ones which we need to load as home packages but failed to do for some reason, - -- it's a bug in the tool invoking GHC. - bad_unit_ids = Set.difference res home_id_set - in if Set.null bad_unit_ids - then [] - else [singleMessage $ mkPlainErrorMsgEnvelope rootLoc $ DriverHomePackagesNotClosed (Set.toList bad_unit_ids)] - +-- +-- See Note [Multiple Home Units], section 'Closure Property'. +checkHomeUnitsClosed :: UnitEnv -> [DriverMessages] +checkHomeUnitsClosed ue + | Set.null bad_unit_ids = [] + | otherwise = [singleMessage $ mkPlainErrorMsgEnvelope rootLoc $ DriverHomePackagesNotClosed (Set.toList bad_unit_ids)] where + home_id_set = unitEnv_keys $ ue_home_unit_graph ue + bad_unit_ids = upwards_closure Set.\\ home_id_set rootLoc = mkGeneralSrcSpan (fsLit "") - -- TODO: This could repeat quite a bit of work but I struggled to write this function. - -- Which units transitively depend on a home unit - loop :: (UnitId, UnitId) -> Set.Set UnitId -- The units which transitively depend on a home unit - loop (from_uid, uid) = - let us = ue_findHomeUnitEnv from_uid ue in - let um = unitInfoMap (homeUnitEnv_units us) in - case lookupUniqMap um uid of - Nothing -> pprPanic "uid not found" (ppr uid) - Just ui -> - let depends = unitDepends ui - home_depends = Set.fromList depends `Set.intersection` home_id_set - other_depends = Set.fromList depends `Set.difference` home_id_set - in - -- Case 1: The unit directly depends on a home_id - if not (null home_depends) - then - let res = foldMap (loop . (from_uid,)) other_depends - in Set.insert uid res - -- Case 2: Check the rest of the dependencies, and then see if any of them depended on - else - let res = foldMap (loop . (from_uid,)) other_depends - in - if not (Set.null res) - then Set.insert uid res - else res + + graph :: Graph (Node UnitId UnitId) + graph = graphFromEdgedVerticesUniq graphNodes + + -- downwards closure of graph + downwards_closure + = graphFromEdgedVerticesUniq [ DigraphNode uid uid (Set.toList deps) + | (uid, deps) <- M.toList (allReachable graph node_key)] + + inverse_closure = transposeG downwards_closure + + upwards_closure = Set.fromList $ map node_key $ reachablesG inverse_closure [DigraphNode uid uid [] | uid <- Set.toList home_id_set] + + all_unit_direct_deps :: UniqMap UnitId (Set.Set UnitId) + all_unit_direct_deps + = unitEnv_foldWithKey go emptyUniqMap $ ue_home_unit_graph ue + where + go rest this this_uis = + plusUniqMap_C Set.union + (addToUniqMap_C Set.union external_depends this (Set.fromList $ this_deps)) + rest + where + external_depends = mapUniqMap (Set.fromList . unitDepends) (unitInfoMap this_units) + this_units = homeUnitEnv_units this_uis + this_deps = [ toUnitId unit | (unit,Just _) <- explicitUnits this_units] + + graphNodes :: [Node UnitId UnitId] + graphNodes = go Set.empty home_id_set + where + go done todo + = case Set.minView todo of + Nothing -> [] + Just (uid, todo') + | Set.member uid done -> go done todo' + | otherwise -> case lookupUniqMap all_unit_direct_deps uid of + Nothing -> pprPanic "uid not found" (ppr (uid, all_unit_direct_deps)) + Just depends -> + let todo'' = (depends Set.\\ done) `Set.union` todo' + in DigraphNode uid uid (Set.toList depends) : go (Set.insert uid done) todo'' -- | Update the every ModSummary that is depended on -- by a module that needs template haskell. We enable codegen to ===================================== compiler/GHC/StgToCmm/Ticky.hs ===================================== @@ -809,7 +809,7 @@ bumpTickyEntryCount lbl = do bumpTickyAllocd :: CLabel -> Int -> FCode () bumpTickyAllocd lbl bytes = do platform <- getPlatform - bumpTickyLitBy (cmmLabelOffB lbl (pc_OFFSET_StgEntCounter_entry_count (platformConstants platform))) bytes + bumpTickyLitBy (cmmLabelOffB lbl (pc_OFFSET_StgEntCounter_allocd (platformConstants platform))) bytes bumpTickyTagSkip :: CLabel -> FCode () bumpTickyTagSkip lbl = do ===================================== compiler/GHC/StgToJS/Apply.hs ===================================== @@ -44,7 +44,6 @@ import GHC.StgToJS.Rts.Types import GHC.StgToJS.Stack import GHC.StgToJS.Ids -import GHC.Types.Literal import GHC.Types.Id import GHC.Types.Id.Info import GHC.Types.CostCentre @@ -58,7 +57,6 @@ import GHC.Core.TyCon import GHC.Core.DataCon import GHC.Core.Type hiding (typeSize) -import GHC.Utils.Encoding import GHC.Utils.Misc import GHC.Utils.Monad import GHC.Utils.Panic @@ -98,22 +96,6 @@ genApp -> G (JStat, ExprResult) genApp ctx i args - -- Case: unpackCStringAppend# "some string"# str - -- - -- Generates h$appendToHsStringA(str, "some string"), which has a faster - -- decoding loop. - | [StgLitArg (LitString bs), x] <- args - , [top] <- concatMap typex_expr (ctxTarget ctx) - , getUnique i == unpackCStringAppendIdKey - , d <- utf8DecodeByteString bs - = do - prof <- csProf <$> getSettings - let profArg = if prof then [jCafCCS] else [] - a <- genArg x - return ( top |= app "h$appendToHsStringA" (toJExpr d : a ++ profArg) - , ExprInline Nothing - ) - -- let-no-escape | Just n <- ctxLneBindingStackSize ctx i = do ===================================== compiler/GHC/StgToJS/Expr.hs ===================================== @@ -57,11 +57,13 @@ import GHC.Types.Var.Set import GHC.Types.Id import GHC.Types.Unique.FM import GHC.Types.RepType +import GHC.Types.Literal import GHC.Stg.Syntax import GHC.Stg.Utils import GHC.Builtin.PrimOps +import GHC.Builtin.Names import GHC.Core import GHC.Core.TyCon @@ -70,6 +72,7 @@ import GHC.Core.Opt.Arity (isOneShotBndr) import GHC.Core.Type hiding (typeSize) import GHC.Utils.Misc +import GHC.Utils.Encoding import GHC.Utils.Monad import GHC.Utils.Panic import GHC.Utils.Outputable (ppr, renderWithContext, defaultSDocContext) @@ -556,7 +559,37 @@ genCase :: HasDebugCallStack -> LiveVars -> G (JStat, ExprResult) genCase ctx bnd e at alts l - | snd (isInlineExpr (ctxEvaluatedIds ctx) e) = do + -- For: unpackCStringAppend# "some string"# str + -- Generate: h$appendToHsStringA(str, "some string") + -- + -- The latter has a faster decoding loop. + -- + -- Since #23270 and 7e0c8b3bab30, literals strings aren't STG atoms and we + -- need to match the following instead: + -- + -- case "some string"# of b { + -- DEFAULT -> unpackCStringAppend# b str + -- } + -- + -- Wrinkle: it doesn't kick in when literals are floated out to the top level. + -- + | StgLit (LitString bs) <- e + , [GenStgAlt DEFAULT _ rhs] <- alts + , StgApp i args <- rhs + , getUnique i == unpackCStringAppendIdKey + , [StgVarArg b',x] <- args + , bnd == b' + , d <- utf8DecodeByteString bs + , [top] <- concatMap typex_expr (ctxTarget ctx) + = do + prof <- csProf <$> getSettings + let profArg = if prof then [jCafCCS] else [] + a <- genArg x + return ( top |= app "h$appendToHsStringA" (toJExpr d : a ++ profArg) + , ExprInline Nothing + ) + + | snd $ isInlineExpr (ctxEvaluatedIds ctx) e = do bndi <- identsForId bnd let ctx' = ctxSetTop bnd $ ctxSetTarget (assocIdExprs bnd (map toJExpr bndi)) ===================================== compiler/GHC/Utils/Lexeme.hs ===================================== @@ -227,10 +227,11 @@ reservedIds = Set.fromList [ "case", "class", "data", "default", "deriving" , "module", "newtype", "of", "then", "type", "where" , "_" ] --- | All reserved operators. Taken from section 2.4 of the 2010 Report. +-- | All reserved operators. Taken from section 2.4 of the 2010 Report, +-- excluding @\@@ and @~@ that are allowed by GHC (see GHC Proposal #229). reservedOps :: Set.Set String reservedOps = Set.fromList [ "..", ":", "::", "=", "\\", "|", "<-", "->" - , "@", "~", "=>" ] + , "=>" ] -- | Does this string contain only dashes and has at least 2 of them? isDashes :: String -> Bool ===================================== libffi-tarballs ===================================== @@ -1 +1 @@ -Subproject commit 5624fd5c8bbce8432cd3c0b0ea92d152a1bba047 +Subproject commit 89a9b01c5647c8f0d3899435b99df690f582e9f1 ===================================== libraries/base/Data/Complex.hs ===================================== @@ -50,17 +50,41 @@ infix 6 :+ -- ----------------------------------------------------------------------------- -- The Complex type --- | Complex numbers are an algebraic type. +-- | A data type representing complex numbers. -- --- For a complex number @z@, @'abs' z@ is a number with the magnitude of @z@, --- but oriented in the positive real direction, whereas @'signum' z@ --- has the phase of @z@, but unit magnitude. +-- You can read about complex numbers [on wikipedia](https://en.wikipedia.org/wiki/Complex_number). -- --- The 'Foldable' and 'Traversable' instances traverse the real part first. +-- In haskell, complex numbers are represented as @a :+ b@ which can be thought of +-- as representing \(a + bi\). For a complex number @z@, @'abs' z@ is a number with the 'magnitude' of @z@, +-- but oriented in the positive real direction, whereas @'signum' z@ +-- has the 'phase' of @z@, but unit 'magnitude'. +-- Apart from the loss of precision due to IEEE754 floating point numbers, +-- it holds that @z == 'abs' z * 'signum' z at . -- -- Note that `Complex`'s instances inherit the deficiencies from the type -- parameter's. For example, @Complex Float@'s 'Ord' instance has similar -- problems to `Float`'s. +-- +-- As can be seen in the examples, the 'Foldable' +-- and 'Traversable' instances traverse the real part first. +-- +-- ==== __Examples__ +-- +-- >>> (5.0 :+ 2.5) + 6.5 +-- 11.5 :+ 2.5 +-- +-- >>> abs (1.0 :+ 1.0) - sqrt 2.0 +-- 0.0 :+ 0.0 +-- +-- >>> abs (signum (4.0 :+ 3.0)) +-- 1.0 :+ 0.0 +-- +-- >>> foldr (:) [] (1 :+ 2) +-- [1,2] +-- +-- >>> mapM print (1 :+ 2) +-- 1 +-- 2 data Complex a = !a :+ !a -- ^ forms a complex number from its real and imaginary -- rectangular components. @@ -79,38 +103,113 @@ data Complex a -- Functions over Complex -- | Extracts the real part of a complex number. +-- +-- ==== __Examples__ +-- +-- >>> realPart (5.0 :+ 3.0) +-- 5.0 +-- +-- >>> realPart ((5.0 :+ 3.0) * (2.0 :+ 3.0)) +-- 1.0 realPart :: Complex a -> a realPart (x :+ _) = x -- | Extracts the imaginary part of a complex number. +-- +-- ==== __Examples__ +-- +-- >>> imagPart (5.0 :+ 3.0) +-- 3.0 +-- +-- >>> imagPart ((5.0 :+ 3.0) * (2.0 :+ 3.0)) +-- 21.0 imagPart :: Complex a -> a imagPart (_ :+ y) = y --- | The conjugate of a complex number. +-- | The 'conjugate' of a complex number. +-- +-- prop> conjugate (conjugate x) = x +-- +-- ==== __Examples__ +-- +-- >>> conjugate (3.0 :+ 3.0) +-- 3.0 :+ (-3.0) +-- +-- >>> conjugate ((3.0 :+ 3.0) * (2.0 :+ 2.0)) +-- 0.0 :+ (-12.0) {-# SPECIALISE conjugate :: Complex Double -> Complex Double #-} conjugate :: Num a => Complex a -> Complex a conjugate (x:+y) = x :+ (-y) --- | Form a complex number from polar components of magnitude and phase. +-- | Form a complex number from 'polar' components of 'magnitude' and 'phase'. +-- +-- ==== __Examples__ +-- +-- >>> mkPolar 1 (pi / 4) +-- 0.7071067811865476 :+ 0.7071067811865475 +-- +-- >>> mkPolar 1 0 +-- 1.0 :+ 0.0 {-# SPECIALISE mkPolar :: Double -> Double -> Complex Double #-} mkPolar :: Floating a => a -> a -> Complex a mkPolar r theta = r * cos theta :+ r * sin theta --- | @'cis' t@ is a complex value with magnitude @1@ --- and phase @t@ (modulo @2*'pi'@). +-- | @'cis' t@ is a complex value with 'magnitude' @1@ +-- and 'phase' @t@ (modulo @2*'pi'@). +-- +-- @ +-- 'cis' = 'mkPolar' 1 +-- @ +-- +-- ==== __Examples__ +-- +-- >>> cis 0 +-- 1.0 :+ 0.0 +-- +-- The following examples are not perfectly zero due to [IEEE 754](https://en.wikipedia.org/wiki/IEEE_754) +-- +-- >>> cis pi +-- (-1.0) :+ 1.2246467991473532e-16 +-- +-- >>> cis (4 * pi) - cis (2 * pi) +-- 0.0 :+ (-2.4492935982947064e-16) {-# SPECIALISE cis :: Double -> Complex Double #-} cis :: Floating a => a -> Complex a cis theta = cos theta :+ sin theta -- | The function 'polar' takes a complex number and --- returns a (magnitude, phase) pair in canonical form: --- the magnitude is non-negative, and the phase in the range @(-'pi', 'pi']@; --- if the magnitude is zero, then so is the phase. +-- returns a ('magnitude', 'phase') pair in canonical form: +-- the 'magnitude' is non-negative, and the 'phase' in the range @(-'pi', 'pi']@; +-- if the 'magnitude' is zero, then so is the 'phase'. +-- +-- @'polar' z = ('magnitude' z, 'phase' z)@ +-- +-- ==== __Examples__ +-- +-- >>> polar (1.0 :+ 1.0) +-- (1.4142135623730951,0.7853981633974483) +-- +-- >>> polar ((-1.0) :+ 0.0) +-- (1.0,3.141592653589793) +-- +-- >>> polar (0.0 :+ 0.0) +-- (0.0,0.0) {-# SPECIALISE polar :: Complex Double -> (Double,Double) #-} polar :: (RealFloat a) => Complex a -> (a,a) polar z = (magnitude z, phase z) --- | The non-negative magnitude of a complex number. +-- | The non-negative 'magnitude' of a complex number. +-- +-- ==== __Examples__ +-- +-- >>> magnitude (1.0 :+ 1.0) +-- 1.4142135623730951 +-- +-- >>> magnitude (1.0 + 0.0) +-- 1.0 +-- +-- >>> magnitude (0.0 :+ (-5.0)) +-- 5.0 {-# SPECIALISE magnitude :: Complex Double -> Double #-} magnitude :: (RealFloat a) => Complex a -> a magnitude (x:+y) = scaleFloat k @@ -119,8 +218,16 @@ magnitude (x:+y) = scaleFloat k mk = - k sqr z = z * z --- | The phase of a complex number, in the range @(-'pi', 'pi']@. --- If the magnitude is zero, then so is the phase. +-- | The 'phase' of a complex number, in the range @(-'pi', 'pi']@. +-- If the 'magnitude' is zero, then so is the 'phase'. +-- +-- ==== __Examples__ +-- +-- >>> phase (0.5 :+ 0.5) / pi +-- 0.25 +-- +-- >>> phase (0 :+ 4) / pi +-- 0.5 {-# SPECIALISE phase :: Complex Double -> Double #-} phase :: (RealFloat a) => Complex a -> a phase (0 :+ 0) = 0 -- SLPJ July 97 from John Peterson ===================================== libraries/base/Data/Data.hs ===================================== @@ -631,6 +631,8 @@ mkDataType str cs = DataType } -- | Constructs a constructor +-- +-- @since 4.16.0.0 mkConstrTag :: DataType -> String -> Int -> [String] -> Fixity -> Constr mkConstrTag dt str idx fields fix = Constr ===================================== m4/fp_cc_linker_flag_try.m4 ===================================== @@ -9,7 +9,7 @@ AC_DEFUN([FP_CC_LINKER_FLAG_TRY], [ AC_MSG_CHECKING([whether C compiler supports -fuse-ld=$1]) echo 'int main(void) {return 0;}' > conftest.c - if $CC -o conftest.o -fuse-ld=$1 conftest.c > /dev/null 2>&1 + if $CC -o conftest.o -fuse-ld=$1 $LDFLAGS conftest.c > /dev/null 2>&1 then $2="-fuse-ld=$1" AC_MSG_RESULT([yes]) ===================================== m4/fp_ld_no_fixup_chains.m4 ===================================== @@ -12,7 +12,7 @@ AC_DEFUN([FP_LD_NO_FIXUP_CHAINS], [ echo 'int main(void) {return 0;}' > conftest.c if $CC -o conftest.o -Wl,-no_fixup_chains conftest.c > /dev/null 2>&1 then - $2="-Wl,-no_fixup_chains" + $2="$$2 -Wl,-no_fixup_chains" AC_MSG_RESULT([yes]) else AC_MSG_RESULT([no]) ===================================== rts/ARMOutlineAtomicsSymbols.h ===================================== @@ -10,583 +10,583 @@ #include #include -uint8_t ghc___aarch64_cas1_relax(uint8_t old, uint8_t new, uint8_t* p); -uint8_t ghc___aarch64_cas1_relax(uint8_t old, uint8_t new, uint8_t* p) { +uint8_t ghc___aarch64_cas1_relax(uint8_t old, uint8_t new, _Atomic uint8_t* p); +uint8_t ghc___aarch64_cas1_relax(uint8_t old, uint8_t new, _Atomic uint8_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_relaxed, memory_order_relaxed); return old; } -uint8_t ghc___aarch64_cas1_acq(uint8_t old, uint8_t new, uint8_t* p); -uint8_t ghc___aarch64_cas1_acq(uint8_t old, uint8_t new, uint8_t* p) { +uint8_t ghc___aarch64_cas1_acq(uint8_t old, uint8_t new, _Atomic uint8_t* p); +uint8_t ghc___aarch64_cas1_acq(uint8_t old, uint8_t new, _Atomic uint8_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_acquire, memory_order_acquire); return old; } -uint8_t ghc___aarch64_cas1_acq_rel(uint8_t old, uint8_t new, uint8_t* p); -uint8_t ghc___aarch64_cas1_acq_rel(uint8_t old, uint8_t new, uint8_t* p) { +uint8_t ghc___aarch64_cas1_acq_rel(uint8_t old, uint8_t new, _Atomic uint8_t* p); +uint8_t ghc___aarch64_cas1_acq_rel(uint8_t old, uint8_t new, _Atomic uint8_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_acq_rel, memory_order_acquire); return old; } -uint8_t ghc___aarch64_cas1_sync(uint8_t old, uint8_t new, uint8_t* p); -uint8_t ghc___aarch64_cas1_sync(uint8_t old, uint8_t new, uint8_t* p) { +uint8_t ghc___aarch64_cas1_sync(uint8_t old, uint8_t new, _Atomic uint8_t* p); +uint8_t ghc___aarch64_cas1_sync(uint8_t old, uint8_t new, _Atomic uint8_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_seq_cst, memory_order_seq_cst); return old; } -uint16_t ghc___aarch64_cas2_relax(uint16_t old, uint16_t new, uint16_t* p); -uint16_t ghc___aarch64_cas2_relax(uint16_t old, uint16_t new, uint16_t* p) { +uint16_t ghc___aarch64_cas2_relax(uint16_t old, uint16_t new, _Atomic uint16_t* p); +uint16_t ghc___aarch64_cas2_relax(uint16_t old, uint16_t new, _Atomic uint16_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_relaxed, memory_order_relaxed); return old; } -uint16_t ghc___aarch64_cas2_acq(uint16_t old, uint16_t new, uint16_t* p); -uint16_t ghc___aarch64_cas2_acq(uint16_t old, uint16_t new, uint16_t* p) { +uint16_t ghc___aarch64_cas2_acq(uint16_t old, uint16_t new, _Atomic uint16_t* p); +uint16_t ghc___aarch64_cas2_acq(uint16_t old, uint16_t new, _Atomic uint16_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_acquire, memory_order_acquire); return old; } -uint16_t ghc___aarch64_cas2_acq_rel(uint16_t old, uint16_t new, uint16_t* p); -uint16_t ghc___aarch64_cas2_acq_rel(uint16_t old, uint16_t new, uint16_t* p) { +uint16_t ghc___aarch64_cas2_acq_rel(uint16_t old, uint16_t new, _Atomic uint16_t* p); +uint16_t ghc___aarch64_cas2_acq_rel(uint16_t old, uint16_t new, _Atomic uint16_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_acq_rel, memory_order_acquire); return old; } -uint16_t ghc___aarch64_cas2_sync(uint16_t old, uint16_t new, uint16_t* p); -uint16_t ghc___aarch64_cas2_sync(uint16_t old, uint16_t new, uint16_t* p) { +uint16_t ghc___aarch64_cas2_sync(uint16_t old, uint16_t new, _Atomic uint16_t* p); +uint16_t ghc___aarch64_cas2_sync(uint16_t old, uint16_t new, _Atomic uint16_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_seq_cst, memory_order_seq_cst); return old; } -uint32_t ghc___aarch64_cas4_relax(uint32_t old, uint32_t new, uint32_t* p); -uint32_t ghc___aarch64_cas4_relax(uint32_t old, uint32_t new, uint32_t* p) { +uint32_t ghc___aarch64_cas4_relax(uint32_t old, uint32_t new, _Atomic uint32_t* p); +uint32_t ghc___aarch64_cas4_relax(uint32_t old, uint32_t new, _Atomic uint32_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_relaxed, memory_order_relaxed); return old; } -uint32_t ghc___aarch64_cas4_acq(uint32_t old, uint32_t new, uint32_t* p); -uint32_t ghc___aarch64_cas4_acq(uint32_t old, uint32_t new, uint32_t* p) { +uint32_t ghc___aarch64_cas4_acq(uint32_t old, uint32_t new, _Atomic uint32_t* p); +uint32_t ghc___aarch64_cas4_acq(uint32_t old, uint32_t new, _Atomic uint32_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_acquire, memory_order_acquire); return old; } -uint32_t ghc___aarch64_cas4_acq_rel(uint32_t old, uint32_t new, uint32_t* p); -uint32_t ghc___aarch64_cas4_acq_rel(uint32_t old, uint32_t new, uint32_t* p) { +uint32_t ghc___aarch64_cas4_acq_rel(uint32_t old, uint32_t new, _Atomic uint32_t* p); +uint32_t ghc___aarch64_cas4_acq_rel(uint32_t old, uint32_t new, _Atomic uint32_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_acq_rel, memory_order_acquire); return old; } -uint32_t ghc___aarch64_cas4_sync(uint32_t old, uint32_t new, uint32_t* p); -uint32_t ghc___aarch64_cas4_sync(uint32_t old, uint32_t new, uint32_t* p) { +uint32_t ghc___aarch64_cas4_sync(uint32_t old, uint32_t new, _Atomic uint32_t* p); +uint32_t ghc___aarch64_cas4_sync(uint32_t old, uint32_t new, _Atomic uint32_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_seq_cst, memory_order_seq_cst); return old; } -uint64_t ghc___aarch64_cas8_relax(uint64_t old, uint64_t new, uint64_t* p); -uint64_t ghc___aarch64_cas8_relax(uint64_t old, uint64_t new, uint64_t* p) { +uint64_t ghc___aarch64_cas8_relax(uint64_t old, uint64_t new, _Atomic uint64_t* p); +uint64_t ghc___aarch64_cas8_relax(uint64_t old, uint64_t new, _Atomic uint64_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_relaxed, memory_order_relaxed); return old; } -uint64_t ghc___aarch64_cas8_acq(uint64_t old, uint64_t new, uint64_t* p); -uint64_t ghc___aarch64_cas8_acq(uint64_t old, uint64_t new, uint64_t* p) { +uint64_t ghc___aarch64_cas8_acq(uint64_t old, uint64_t new, _Atomic uint64_t* p); +uint64_t ghc___aarch64_cas8_acq(uint64_t old, uint64_t new, _Atomic uint64_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_acquire, memory_order_acquire); return old; } -uint64_t ghc___aarch64_cas8_acq_rel(uint64_t old, uint64_t new, uint64_t* p); -uint64_t ghc___aarch64_cas8_acq_rel(uint64_t old, uint64_t new, uint64_t* p) { +uint64_t ghc___aarch64_cas8_acq_rel(uint64_t old, uint64_t new, _Atomic uint64_t* p); +uint64_t ghc___aarch64_cas8_acq_rel(uint64_t old, uint64_t new, _Atomic uint64_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_acq_rel, memory_order_acquire); return old; } -uint64_t ghc___aarch64_cas8_sync(uint64_t old, uint64_t new, uint64_t* p); -uint64_t ghc___aarch64_cas8_sync(uint64_t old, uint64_t new, uint64_t* p) { +uint64_t ghc___aarch64_cas8_sync(uint64_t old, uint64_t new, _Atomic uint64_t* p); +uint64_t ghc___aarch64_cas8_sync(uint64_t old, uint64_t new, _Atomic uint64_t* p) { atomic_compare_exchange_strong_explicit(p, &old, new, memory_order_seq_cst, memory_order_seq_cst); return old; } -uint8_t ghc___aarch64_swp1_relax(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_swp1_relax(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_swp1_relax(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_swp1_relax(uint8_t v, _Atomic uint8_t* p) { return atomic_exchange_explicit(p, v, memory_order_relaxed); } -uint8_t ghc___aarch64_swp1_acq(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_swp1_acq(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_swp1_acq(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_swp1_acq(uint8_t v, _Atomic uint8_t* p) { return atomic_exchange_explicit(p, v, memory_order_acquire); } -uint8_t ghc___aarch64_swp1_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_swp1_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_swp1_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_swp1_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_exchange_explicit(p, v, memory_order_release); } -uint8_t ghc___aarch64_swp1_acq_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_swp1_acq_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_swp1_acq_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_swp1_acq_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_exchange_explicit(p, v, memory_order_acq_rel); } -uint8_t ghc___aarch64_swp1_sync(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_swp1_sync(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_swp1_sync(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_swp1_sync(uint8_t v, _Atomic uint8_t* p) { return atomic_exchange_explicit(p, v, memory_order_seq_cst); } -uint16_t ghc___aarch64_swp2_relax(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_swp2_relax(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_swp2_relax(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_swp2_relax(uint16_t v, _Atomic uint16_t* p) { return atomic_exchange_explicit(p, v, memory_order_relaxed); } -uint16_t ghc___aarch64_swp2_acq(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_swp2_acq(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_swp2_acq(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_swp2_acq(uint16_t v, _Atomic uint16_t* p) { return atomic_exchange_explicit(p, v, memory_order_acquire); } -uint16_t ghc___aarch64_swp2_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_swp2_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_swp2_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_swp2_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_exchange_explicit(p, v, memory_order_release); } -uint16_t ghc___aarch64_swp2_acq_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_swp2_acq_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_swp2_acq_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_swp2_acq_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_exchange_explicit(p, v, memory_order_acq_rel); } -uint16_t ghc___aarch64_swp2_sync(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_swp2_sync(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_swp2_sync(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_swp2_sync(uint16_t v, _Atomic uint16_t* p) { return atomic_exchange_explicit(p, v, memory_order_seq_cst); } -uint32_t ghc___aarch64_swp4_relax(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_swp4_relax(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_swp4_relax(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_swp4_relax(uint32_t v, _Atomic uint32_t* p) { return atomic_exchange_explicit(p, v, memory_order_relaxed); } -uint32_t ghc___aarch64_swp4_acq(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_swp4_acq(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_swp4_acq(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_swp4_acq(uint32_t v, _Atomic uint32_t* p) { return atomic_exchange_explicit(p, v, memory_order_acquire); } -uint32_t ghc___aarch64_swp4_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_swp4_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_swp4_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_swp4_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_exchange_explicit(p, v, memory_order_release); } -uint32_t ghc___aarch64_swp4_acq_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_swp4_acq_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_swp4_acq_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_swp4_acq_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_exchange_explicit(p, v, memory_order_acq_rel); } -uint32_t ghc___aarch64_swp4_sync(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_swp4_sync(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_swp4_sync(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_swp4_sync(uint32_t v, _Atomic uint32_t* p) { return atomic_exchange_explicit(p, v, memory_order_seq_cst); } -uint64_t ghc___aarch64_swp8_relax(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_swp8_relax(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_swp8_relax(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_swp8_relax(uint64_t v, _Atomic uint64_t* p) { return atomic_exchange_explicit(p, v, memory_order_relaxed); } -uint64_t ghc___aarch64_swp8_acq(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_swp8_acq(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_swp8_acq(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_swp8_acq(uint64_t v, _Atomic uint64_t* p) { return atomic_exchange_explicit(p, v, memory_order_acquire); } -uint64_t ghc___aarch64_swp8_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_swp8_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_swp8_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_swp8_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_exchange_explicit(p, v, memory_order_release); } -uint64_t ghc___aarch64_swp8_acq_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_swp8_acq_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_swp8_acq_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_swp8_acq_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_exchange_explicit(p, v, memory_order_acq_rel); } -uint64_t ghc___aarch64_swp8_sync(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_swp8_sync(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_swp8_sync(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_swp8_sync(uint64_t v, _Atomic uint64_t* p) { return atomic_exchange_explicit(p, v, memory_order_seq_cst); } -uint8_t ghc___aarch64_ldadd1_relax(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldadd1_relax(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldadd1_relax(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldadd1_relax(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_relaxed); } -uint8_t ghc___aarch64_ldadd1_acq(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldadd1_acq(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldadd1_acq(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldadd1_acq(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_acquire); } -uint8_t ghc___aarch64_ldadd1_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldadd1_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldadd1_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldadd1_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_release); } -uint8_t ghc___aarch64_ldadd1_acq_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldadd1_acq_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldadd1_acq_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldadd1_acq_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_acq_rel); } -uint8_t ghc___aarch64_ldadd1_sync(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldadd1_sync(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldadd1_sync(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldadd1_sync(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_seq_cst); } -uint16_t ghc___aarch64_ldadd2_relax(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldadd2_relax(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldadd2_relax(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldadd2_relax(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_relaxed); } -uint16_t ghc___aarch64_ldadd2_acq(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldadd2_acq(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldadd2_acq(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldadd2_acq(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_acquire); } -uint16_t ghc___aarch64_ldadd2_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldadd2_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldadd2_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldadd2_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_release); } -uint16_t ghc___aarch64_ldadd2_acq_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldadd2_acq_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldadd2_acq_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldadd2_acq_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_acq_rel); } -uint16_t ghc___aarch64_ldadd2_sync(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldadd2_sync(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldadd2_sync(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldadd2_sync(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_seq_cst); } -uint32_t ghc___aarch64_ldadd4_relax(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldadd4_relax(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldadd4_relax(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldadd4_relax(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_relaxed); } -uint32_t ghc___aarch64_ldadd4_acq(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldadd4_acq(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldadd4_acq(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldadd4_acq(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_acquire); } -uint32_t ghc___aarch64_ldadd4_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldadd4_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldadd4_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldadd4_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_release); } -uint32_t ghc___aarch64_ldadd4_acq_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldadd4_acq_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldadd4_acq_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldadd4_acq_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_acq_rel); } -uint32_t ghc___aarch64_ldadd4_sync(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldadd4_sync(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldadd4_sync(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldadd4_sync(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_seq_cst); } -uint64_t ghc___aarch64_ldadd8_relax(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldadd8_relax(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldadd8_relax(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldadd8_relax(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_relaxed); } -uint64_t ghc___aarch64_ldadd8_acq(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldadd8_acq(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldadd8_acq(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldadd8_acq(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_acquire); } -uint64_t ghc___aarch64_ldadd8_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldadd8_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldadd8_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldadd8_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_release); } -uint64_t ghc___aarch64_ldadd8_acq_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldadd8_acq_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldadd8_acq_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldadd8_acq_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_acq_rel); } -uint64_t ghc___aarch64_ldadd8_sync(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldadd8_sync(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldadd8_sync(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldadd8_sync(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_add_explicit(p, v, memory_order_seq_cst); } -uint8_t ghc___aarch64_ldclr1_relax(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldclr1_relax(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldclr1_relax(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldclr1_relax(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_relaxed); } -uint8_t ghc___aarch64_ldclr1_acq(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldclr1_acq(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldclr1_acq(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldclr1_acq(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_acquire); } -uint8_t ghc___aarch64_ldclr1_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldclr1_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldclr1_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldclr1_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_release); } -uint8_t ghc___aarch64_ldclr1_acq_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldclr1_acq_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldclr1_acq_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldclr1_acq_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_acq_rel); } -uint8_t ghc___aarch64_ldclr1_sync(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldclr1_sync(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldclr1_sync(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldclr1_sync(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_seq_cst); } -uint16_t ghc___aarch64_ldclr2_relax(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldclr2_relax(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldclr2_relax(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldclr2_relax(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_relaxed); } -uint16_t ghc___aarch64_ldclr2_acq(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldclr2_acq(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldclr2_acq(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldclr2_acq(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_acquire); } -uint16_t ghc___aarch64_ldclr2_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldclr2_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldclr2_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldclr2_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_release); } -uint16_t ghc___aarch64_ldclr2_acq_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldclr2_acq_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldclr2_acq_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldclr2_acq_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_acq_rel); } -uint16_t ghc___aarch64_ldclr2_sync(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldclr2_sync(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldclr2_sync(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldclr2_sync(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_seq_cst); } -uint32_t ghc___aarch64_ldclr4_relax(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldclr4_relax(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldclr4_relax(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldclr4_relax(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_relaxed); } -uint32_t ghc___aarch64_ldclr4_acq(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldclr4_acq(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldclr4_acq(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldclr4_acq(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_acquire); } -uint32_t ghc___aarch64_ldclr4_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldclr4_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldclr4_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldclr4_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_release); } -uint32_t ghc___aarch64_ldclr4_acq_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldclr4_acq_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldclr4_acq_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldclr4_acq_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_acq_rel); } -uint32_t ghc___aarch64_ldclr4_sync(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldclr4_sync(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldclr4_sync(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldclr4_sync(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_seq_cst); } -uint64_t ghc___aarch64_ldclr8_relax(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldclr8_relax(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldclr8_relax(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldclr8_relax(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_relaxed); } -uint64_t ghc___aarch64_ldclr8_acq(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldclr8_acq(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldclr8_acq(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldclr8_acq(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_acquire); } -uint64_t ghc___aarch64_ldclr8_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldclr8_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldclr8_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldclr8_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_release); } -uint64_t ghc___aarch64_ldclr8_acq_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldclr8_acq_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldclr8_acq_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldclr8_acq_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_acq_rel); } -uint64_t ghc___aarch64_ldclr8_sync(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldclr8_sync(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldclr8_sync(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldclr8_sync(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_and_explicit(p, v, memory_order_seq_cst); } -uint8_t ghc___aarch64_ldeor1_relax(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldeor1_relax(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldeor1_relax(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldeor1_relax(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_relaxed); } -uint8_t ghc___aarch64_ldeor1_acq(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldeor1_acq(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldeor1_acq(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldeor1_acq(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_acquire); } -uint8_t ghc___aarch64_ldeor1_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldeor1_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldeor1_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldeor1_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_release); } -uint8_t ghc___aarch64_ldeor1_acq_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldeor1_acq_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldeor1_acq_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldeor1_acq_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_acq_rel); } -uint8_t ghc___aarch64_ldeor1_sync(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldeor1_sync(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldeor1_sync(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldeor1_sync(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_seq_cst); } -uint16_t ghc___aarch64_ldeor2_relax(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldeor2_relax(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldeor2_relax(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldeor2_relax(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_relaxed); } -uint16_t ghc___aarch64_ldeor2_acq(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldeor2_acq(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldeor2_acq(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldeor2_acq(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_acquire); } -uint16_t ghc___aarch64_ldeor2_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldeor2_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldeor2_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldeor2_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_release); } -uint16_t ghc___aarch64_ldeor2_acq_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldeor2_acq_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldeor2_acq_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldeor2_acq_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_acq_rel); } -uint16_t ghc___aarch64_ldeor2_sync(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldeor2_sync(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldeor2_sync(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldeor2_sync(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_seq_cst); } -uint32_t ghc___aarch64_ldeor4_relax(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldeor4_relax(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldeor4_relax(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldeor4_relax(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_relaxed); } -uint32_t ghc___aarch64_ldeor4_acq(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldeor4_acq(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldeor4_acq(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldeor4_acq(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_acquire); } -uint32_t ghc___aarch64_ldeor4_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldeor4_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldeor4_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldeor4_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_release); } -uint32_t ghc___aarch64_ldeor4_acq_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldeor4_acq_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldeor4_acq_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldeor4_acq_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_acq_rel); } -uint32_t ghc___aarch64_ldeor4_sync(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldeor4_sync(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldeor4_sync(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldeor4_sync(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_seq_cst); } -uint64_t ghc___aarch64_ldeor8_relax(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldeor8_relax(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldeor8_relax(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldeor8_relax(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_relaxed); } -uint64_t ghc___aarch64_ldeor8_acq(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldeor8_acq(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldeor8_acq(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldeor8_acq(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_acquire); } -uint64_t ghc___aarch64_ldeor8_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldeor8_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldeor8_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldeor8_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_release); } -uint64_t ghc___aarch64_ldeor8_acq_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldeor8_acq_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldeor8_acq_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldeor8_acq_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_acq_rel); } -uint64_t ghc___aarch64_ldeor8_sync(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldeor8_sync(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldeor8_sync(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldeor8_sync(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_xor_explicit(p, v, memory_order_seq_cst); } -uint8_t ghc___aarch64_ldset1_relax(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldset1_relax(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldset1_relax(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldset1_relax(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_relaxed); } -uint8_t ghc___aarch64_ldset1_acq(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldset1_acq(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldset1_acq(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldset1_acq(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_acquire); } -uint8_t ghc___aarch64_ldset1_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldset1_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldset1_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldset1_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_release); } -uint8_t ghc___aarch64_ldset1_acq_rel(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldset1_acq_rel(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldset1_acq_rel(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldset1_acq_rel(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_acq_rel); } -uint8_t ghc___aarch64_ldset1_sync(uint8_t v, uint8_t* p); -uint8_t ghc___aarch64_ldset1_sync(uint8_t v, uint8_t* p) { +uint8_t ghc___aarch64_ldset1_sync(uint8_t v, _Atomic uint8_t* p); +uint8_t ghc___aarch64_ldset1_sync(uint8_t v, _Atomic uint8_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_seq_cst); } -uint16_t ghc___aarch64_ldset2_relax(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldset2_relax(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldset2_relax(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldset2_relax(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_relaxed); } -uint16_t ghc___aarch64_ldset2_acq(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldset2_acq(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldset2_acq(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldset2_acq(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_acquire); } -uint16_t ghc___aarch64_ldset2_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldset2_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldset2_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldset2_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_release); } -uint16_t ghc___aarch64_ldset2_acq_rel(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldset2_acq_rel(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldset2_acq_rel(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldset2_acq_rel(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_acq_rel); } -uint16_t ghc___aarch64_ldset2_sync(uint16_t v, uint16_t* p); -uint16_t ghc___aarch64_ldset2_sync(uint16_t v, uint16_t* p) { +uint16_t ghc___aarch64_ldset2_sync(uint16_t v, _Atomic uint16_t* p); +uint16_t ghc___aarch64_ldset2_sync(uint16_t v, _Atomic uint16_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_seq_cst); } -uint32_t ghc___aarch64_ldset4_relax(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldset4_relax(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldset4_relax(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldset4_relax(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_relaxed); } -uint32_t ghc___aarch64_ldset4_acq(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldset4_acq(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldset4_acq(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldset4_acq(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_acquire); } -uint32_t ghc___aarch64_ldset4_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldset4_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldset4_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldset4_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_release); } -uint32_t ghc___aarch64_ldset4_acq_rel(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldset4_acq_rel(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldset4_acq_rel(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldset4_acq_rel(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_acq_rel); } -uint32_t ghc___aarch64_ldset4_sync(uint32_t v, uint32_t* p); -uint32_t ghc___aarch64_ldset4_sync(uint32_t v, uint32_t* p) { +uint32_t ghc___aarch64_ldset4_sync(uint32_t v, _Atomic uint32_t* p); +uint32_t ghc___aarch64_ldset4_sync(uint32_t v, _Atomic uint32_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_seq_cst); } -uint64_t ghc___aarch64_ldset8_relax(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldset8_relax(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldset8_relax(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldset8_relax(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_relaxed); } -uint64_t ghc___aarch64_ldset8_acq(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldset8_acq(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldset8_acq(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldset8_acq(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_acquire); } -uint64_t ghc___aarch64_ldset8_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldset8_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldset8_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldset8_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_release); } -uint64_t ghc___aarch64_ldset8_acq_rel(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldset8_acq_rel(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldset8_acq_rel(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldset8_acq_rel(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_acq_rel); } -uint64_t ghc___aarch64_ldset8_sync(uint64_t v, uint64_t* p); -uint64_t ghc___aarch64_ldset8_sync(uint64_t v, uint64_t* p) { +uint64_t ghc___aarch64_ldset8_sync(uint64_t v, _Atomic uint64_t* p); +uint64_t ghc___aarch64_ldset8_sync(uint64_t v, _Atomic uint64_t* p) { return atomic_fetch_or_explicit(p, v, memory_order_seq_cst); } ===================================== rts/sm/NonMovingMark.c ===================================== @@ -2008,7 +2008,7 @@ bool nonmovingTidyWeaks (struct MarkQueue_ *queue) // See Note [Weak pointer processing and the non-moving GC] in // MarkWeak.c - bool key_in_nonmoving = Bdescr((StgPtr) w->key)->flags & BF_NONMOVING; + bool key_in_nonmoving = HEAP_ALLOCED_GC(w->key) && Bdescr((StgPtr) w->key)->flags & BF_NONMOVING; if (!key_in_nonmoving || nonmovingIsNowAlive(w->key)) { nonmovingMarkLiveWeak(queue, w); did_work = true; ===================================== rts/sm/Scav.c ===================================== @@ -1593,6 +1593,14 @@ scavenge_one(StgPtr p) #endif break; + case BCO: { + StgBCO *bco = (StgBCO *)p; + evacuate((StgClosure **)&bco->instrs); + evacuate((StgClosure **)&bco->literals); + evacuate((StgClosure **)&bco->ptrs); + break; + } + case COMPACT_NFDATA: scavenge_compact((StgCompactNFData*)p); break; ===================================== testsuite/tests/driver/multipleHomeUnits/mhu-perf/Makefile ===================================== @@ -0,0 +1,23 @@ +TOP=../../../.. +include $(TOP)/mk/boilerplate.mk +include $(TOP)/mk/test.mk + +SETUP=../Setup -v0 + +mhu-perf: clean + $(MAKE) -s --no-print-directory clean + ./genLargeHMU + '$(GHC_PKG)' init tmp.d + '$(TEST_HC)' $(TEST_HC_OPTS) -v0 --make Setup + for dir in unit-p*; do \ + cd $$dir && $(SETUP) clean && $(SETUP) configure $(CABAL_MINIMAL_BUILD) --ipid=$$dir-0.1.0.0 --with-ghc='$(TEST_HC)' --with-hc-pkg='$(GHC_PKG)' --ghc-options='$(TEST_HC_OPTS)' --package-db=../tmp.d && $(SETUP) build && $(SETUP) register --inplace && cd ..; \ + done; + + +ifeq "$(CLEANUP)" "1" + $(MAKE) -s --no-print-directory clean +endif + +clean : + $(RM) -r unitTop* unit-p* top*/ tmp*.d inst-* *.o *.hi */*.o */*.hi */Setup$(exeext) */dist Setup$(exeext) + ===================================== testsuite/tests/driver/multipleHomeUnits/mhu-perf/Setup.hs ===================================== @@ -0,0 +1,2 @@ +import Distribution.Simple +main = defaultMain ===================================== testsuite/tests/driver/multipleHomeUnits/mhu-perf/all.T ===================================== @@ -0,0 +1,10 @@ +test('mhu-perf', + [ collect_compiler_stats('bytes allocated',2), + extra_files(['genLargeHMU','Setup.hs']), + pre_cmd('$MAKE -s --no-print-directory mhu-perf'), + js_broken(22349), + when(arch('wasm32'), skip), # wasm32 doesn't like running Setup/Makefile tests + compile_timeout_multiplier(5) + ], + multiunit_compile, + [['unitTop1', 'unitTop2'], '-fhide-source-paths']) ===================================== testsuite/tests/driver/multipleHomeUnits/mhu-perf/genLargeHMU ===================================== @@ -0,0 +1,54 @@ +#!/usr/bin/env bash +# Generate $DEPTH layers of packages with $WIDTH modules on each layer +# Every package on layer N depends on all the packages on layer N-1 +# unitTop imports all the units from the last layer +DEPTH=8 +WIDTH=8 +for i in $(seq -w 1 $WIDTH); do + mkdir unit-p0M$i + echo "module DummyLevel0M$i where" > unit-p0M$i/DummyLevel0M$i.hs; + cat > unit-p0M$i/unit-p0M$i.cabal <=1.10 +library + default-language: Haskell2010 + exposed-modules: DummyLevel0M$i + build-depends: base +EOF +done +for l in $(seq 1 $DEPTH); do + for i in $(seq -w 1 $WIDTH); do + mkdir unit-p${l}M$i + cat > unit-p${l}M$i/unit-p${l}M$i.cabal <=1.10 +library + default-language: Haskell2010 + exposed-modules: DummyLevel${l}M$i + build-depends: base +EOF + echo "module DummyLevel${l}M$i where" > unit-p${l}M$i/DummyLevel${l}M$i.hs; + for j in $(seq -w 1 $WIDTH); do + echo " , unit-p$((l-1))M$j" >> unit-p${l}M$i/unit-p${l}M$i.cabal + echo "import DummyLevel$((l-1))M$j" >> unit-p${l}M$i/DummyLevel${l}M$i.hs; + done + done +done +mkdir top1 +echo "module Top1 where" > top1/Top1.hs +echo "-package-db ./tmp.d -i -itop1 Top1 -this-unit-id unit-top1 -package base" > unitTop1; +for j in $(seq -w 1 $WIDTH); do + echo "-package unit-p${DEPTH}M$j" >> unitTop1; + echo "import DummyLevel${DEPTH}M$j" >> top1/Top1.hs; +done +mkdir top2 +echo "module Top2 where" > top2/Top2.hs +echo "-package-db ./tmp.d -i -itop2 Top2 -this-unit-id unit-top2 -package base" > unitTop2; +for j in $(seq -w 2 $WIDTH); do + echo "-package unit-p${DEPTH}M$j" >> unitTop2; + echo "import DummyLevel${DEPTH}M$j" >> top2/Top2.hs; +done ===================================== testsuite/tests/driver/multipleHomeUnits/mhu-perf/mhu-perf.stderr ===================================== @@ -0,0 +1,2 @@ +[1 of 2] Compiling Top1[unit-top1] +[2 of 2] Compiling Top2[unit-top2] ===================================== testsuite/tests/primops/should_run/T24496.hs ===================================== @@ -0,0 +1,18 @@ +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE UnboxedTuples #-} +import GHC.Exts + +twoProductFloat# :: Float# -> Float# -> (# Float#, Float# #) +twoProductFloat# x y = let !r = x `timesFloat#` y + in (# r, fmsubFloat# x y r #) +{-# NOINLINE twoProductFloat# #-} + +twoProductDouble# :: Double# -> Double# -> (# Double#, Double# #) +twoProductDouble# x y = let !r = x *## y + in (# r, fmsubDouble# x y r #) +{-# NOINLINE twoProductDouble# #-} + +main :: IO () +main = do + print $ case twoProductFloat# 2.0# 3.0# of (# r, s #) -> (F# r, F# s) + print $ case twoProductDouble# 2.0## 3.0## of (# r, s #) -> (D# r, D# s) ===================================== testsuite/tests/primops/should_run/T24496.stdout ===================================== @@ -0,0 +1,2 @@ +(6.0,0.0) +(6.0,0.0) ===================================== testsuite/tests/primops/should_run/all.T ===================================== @@ -75,3 +75,10 @@ test('FMA_ConstantFold' test('T21624', normal, compile_and_run, ['']) test('T23071', ignore_stdout, compile_and_run, ['']) +test('T24496' + , [ when(have_cpu_feature('fma'), extra_hc_opts('-mfma')) + , js_skip # JS backend doesn't have an FMA implementation + , when(arch('wasm32'), skip) + , when(have_llvm(), extra_ways(["optllvm"])) + ] + , compile_and_run, ['-O']) ===================================== testsuite/tests/th/T23748.hs ===================================== @@ -0,0 +1,7 @@ +{-# LANGUAGE TemplateHaskell #-} + +module T23748 where + +$([d| a ~ b = () + a @ b = () + |]) \ No newline at end of file ===================================== testsuite/tests/th/all.T ===================================== @@ -580,3 +580,4 @@ test('T22559b', normal, compile_fail, ['']) test('T22559c', normal, compile_fail, ['']) test('T23525', normal, compile, ['']) test('T24837', normal, compile, ['-v0 -ddump-splices -dsuppress-uniques']) +test('T23748', normal, compile, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7f7df936d499a3f65dacc1cd917d61fdf9cd9add...c6a95d279e39cf2082aeff1ce91262a361d823fb -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/7f7df936d499a3f65dacc1cd917d61fdf9cd9add...c6a95d279e39cf2082aeff1ce91262a361d823fb You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 21:43:24 2024 From: gitlab at gitlab.haskell.org (Marge Bot (@marge-bot)) Date: Mon, 30 Sep 2024 17:43:24 -0400 Subject: [Git][ghc/ghc][wip/marge_bot_batch_merge_job] 3 commits: SpecConstr: Introduce a separate argument limit for forced specs. Message-ID: <66fb1b7c5bacf_161302a9d9a4796ee@gitlab.mail> Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 373a7b7a by Andreas Klebinger at 2024-09-30T17:42:47-04:00 SpecConstr: Introduce a separate argument limit for forced specs. We used to put no limit at all on specializations forced via the SPEC argument. This isn't always reasonable so we introduce a very high limit that applies to forced specializations, a flag to control it, and we now emit a warning if we fail a specialization because we exceed the warning. Fixes #25197 - - - - - 6f8b2498 by Andreas Klebinger at 2024-09-30T17:42:48-04:00 ghc-experimental: Expose primops and ghc extensions via GHC.PrimOps This will be the new place for functions that would have gone into GHC.Exts in the past but are not stable enough to do so now. Addresses #25242 - - - - - 6bdc9630 by Sylvain Henry at 2024-09-30T17:43:02-04:00 RTS: cleanup timerfd file descriptors after a fork (#25280) When we init a timerfd-based ticker, we should be careful to cleanup the old file descriptors (e.g. after a fork). - - - - - 9 changed files: - compiler/GHC/Core/Opt/SpecConstr.hs - compiler/GHC/Driver/DynFlags.hs - compiler/GHC/Driver/Session.hs - docs/users_guide/using-optimisation.rst - libraries/base/src/GHC/Exts.hs - libraries/ghc-experimental/ghc-experimental.cabal.in - + libraries/ghc-experimental/src/GHC/PrimOps.hs - libraries/ghc-internal/src/GHC/Internal/Exts.hs - rts/posix/ticker/TimerFd.c The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f678355cebdc1432e52514ecf8f27dd5066ccead...6bdc963078be47ba1807ce9d135f76918b82a362 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/f678355cebdc1432e52514ecf8f27dd5066ccead...6bdc963078be47ba1807ce9d135f76918b82a362 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gitlab at gitlab.haskell.org Mon Sep 30 23:40:45 2024 From: gitlab at gitlab.haskell.org (Simon Peyton Jones (@simonpj)) Date: Mon, 30 Sep 2024 19:40:45 -0400 Subject: [Git][ghc/ghc][wip/T25281] Now working I think Message-ID: <66fb36fdb77c7_103d464b0fc848fb@gitlab.mail> Simon Peyton Jones pushed to branch wip/T25281 at Glasgow Haskell Compiler / GHC Commits: ba79e31c by Simon Peyton Jones at 2024-10-01T00:40:23+01:00 Now working I think - - - - - 13 changed files: - compiler/GHC/Core/LateCC/OverloadedCalls.hs - compiler/GHC/HsToCore/Expr.hs - compiler/GHC/HsToCore/Types.hs - compiler/GHC/Tc/Errors.hs - compiler/GHC/Tc/Gen/App.hs - compiler/GHC/Tc/Instance/Class.hs - compiler/GHC/Tc/Solver/Monad.hs - compiler/GHC/Tc/Types/CtLoc.hs - compiler/GHC/Tc/Types/LclEnv.hs - compiler/GHC/Tc/Types/Origin.hs - compiler/GHC/Tc/Utils/Monad.hs - compiler/GHC/Tc/Validity.hs - testsuite/tests/pmcheck/should_compile/T24891.hs Changes: ===================================== compiler/GHC/Core/LateCC/OverloadedCalls.hs ===================================== @@ -26,7 +26,6 @@ import GHC.Types.Name import GHC.Types.SrcLoc import GHC.Types.Tickish import GHC.Types.Var -import GHC.Utils.Outputable type OverloadedCallsCCState = Strict.Maybe SrcSpan ===================================== compiler/GHC/HsToCore/Expr.hs ===================================== @@ -723,21 +723,16 @@ ds_app_var (L loc fun_id) hs_args core_args -- Alas, we cannot simply look at the unfolding of $dHasField below because it -- has not been set yet, so we have to reconstruct the selector Id from the types. | fun_id `hasKey` getFieldClassOpKey - , (_k : _rrep : _arep : Type x_ty : Type r_ty : _) <- core_args - -- getField :: forall {k} {r_rep} {a_rep} (x::k) (r :: TYPE r_rep) (a :: TYPE a_rep) . - -- HasField x r a => r -> a - = do { fam_inst_envs <- dsGetFamInstEnvs + = do { -- Look up the field named x/"sel" in the type r/T + fam_inst_envs <- dsGetFamInstEnvs ; rdr_env <- dsGetGlobalRdrEnv - -- Look up the field named x/"sel" in the type r/T - ; tracePm "getfield" (ppr core_args $$ ppr x_ty $$ ppr r_ty) - ; case lookupHasFieldLabel fam_inst_envs x_ty r_ty of - Just fl | isJust (lookupGRE_FieldLabel rdr_env fl) - -- isJust: Make sure the field is actually visible in this module; - -- otherwise this might not be the implicit HasField instance - -> do { sel_id <- dsLookupGlobalId (flSelector fl) - ; tracePm "getfield2" (ppr sel_id) - ; ds_app_rec_sel sel_id fun_id core_args } - _ -> ds_app_finish fun_id core_args } + ; let core_arg_tys :: [Type] = [ty | Type ty <- core_args] + ; case lookupHasFieldLabel fam_inst_envs rdr_env core_arg_tys of + Just (sel_name,_,_,_) + -> do { sel_id <- dsLookupGlobalId sel_name + ; tracePm "getfield2" (ppr sel_id) + ; ds_app_rec_sel sel_id fun_id core_args } + _ -> ds_app_finish fun_id core_args } ----------------------- -- Warn about identities for (fromInteger :: Integer -> Integer) etc ===================================== compiler/GHC/HsToCore/Types.hs ===================================== @@ -11,7 +11,7 @@ module GHC.HsToCore.Types ( DsMetaEnv, DsMetaVal(..), CompleteMatches ) where -import GHC.Prelude (Int, Bool) +import GHC.Prelude (Int) import Data.IORef import qualified Data.Set as S ===================================== compiler/GHC/Tc/Errors.hs ===================================== @@ -2313,7 +2313,7 @@ mk_dict_err ctxt (item, (matches, unifiers, unsafe_overlapped)) = case (NE.nonEm report_no_fieldnames item | Just (EvVarDest evvar) <- ei_evdest item -- we can assume that here we have a `HasField @Symbol x r a` instance - -- because of HasFieldOrigin in record_field + -- because of GetFieldOrigin in record_field , Just (_, [_symbol, x, r, a]) <- tcSplitTyConApp_maybe (varType evvar) , Just (r_tycon, _) <- tcSplitTyConApp_maybe r , Just x_name <- isStrLitTy x @@ -2328,7 +2328,7 @@ mk_dict_err ctxt (item, (matches, unifiers, unsafe_overlapped)) = case (NE.nonEm isNothing (lookupLocalRdrOcc lcl_env occ_name) record_field = case orig of - HasFieldOrigin name -> Just (mkVarOccFS name) + GetFieldOrigin name -> Just (mkVarOccFS name) _ -> Nothing cannot_resolve_msg :: ErrorItem -> [ClsInst] -> RelevantBindings ===================================== compiler/GHC/Tc/Gen/App.hs ===================================== @@ -403,18 +403,10 @@ tcApp rn_expr exp_res_ty ; (tc_fun, fun_sigma) <- tcInferAppHead fun ; let tc_head = (tc_fun, fun_ctxt) - ; let supp_incomplete_rec_sel - | XExpr (ExpandedThingRn (OrigExpr HsGetField{}) _) <- rn_expr - -- See (7) of Note [Detecting incomplete record selectors] - = setSuppressIncompleteRecSelsTc True - | otherwise - = id - -- Step 3: Instantiate the function type (taking a quick look at args) ; do_ql <- wantQuickLook rn_fun ; (inst_args, app_res_rho) - <- supp_incomplete_rec_sel $ - setQLInstLevel do_ql $ -- See (TCAPP1) and (TCAPP2) in + <- setQLInstLevel do_ql $ -- See (TCAPP1) and (TCAPP2) in -- Note [tcApp: typechecking applications] tcInstFun do_ql True tc_head fun_sigma rn_args @@ -772,9 +764,7 @@ tcInstFun do_ql inst_final (tc_fun, fun_ctxt) fun_sigma rn_args ; go (pos+1) (addArgWrap wrap acc) inst_body rest_args } go1 pos acc fun_ty (EWrap w : args) - = setSuppressIncompleteRecSelsTc False $ - -- See (7) of Note [Detecting incomplete record selectors] - go1 pos (EWrap w : acc) fun_ty args + = go1 pos (EWrap w : acc) fun_ty args go1 pos acc fun_ty (EPrag sp prag : args) = go1 pos (EPrag sp prag : acc) fun_ty args ===================================== compiler/GHC/Tc/Instance/Class.hs ===================================== @@ -22,7 +22,8 @@ import GHC.Tc.Utils.Instantiate(instDFunType, tcInstType) import GHC.Tc.Instance.Typeable import GHC.Tc.Utils.TcMType import GHC.Tc.Types.Evidence -import GHC.Tc.Types.Origin (InstanceWhat (..), SafeOverlapping) +import GHC.Tc.Types.CtLoc +import GHC.Tc.Types.Origin ( InstanceWhat (..), SafeOverlapping, CtOrigin(GetFieldOrigin) ) import GHC.Tc.Instance.Family( tcGetFamInstEnvs, tcInstNewTyCon_maybe, tcLookupDataFamInst, FamInstEnvs ) import GHC.Rename.Env( addUsedGRE, addUsedDataCons, DeprecationWarnings (..) ) @@ -65,7 +66,6 @@ import GHC.Hs.Extension import Language.Haskell.Syntax.Basic (FieldLabelString(..)) import GHC.Types.Id.Info import GHC.Tc.Errors.Types -import Control.Monad import Data.Functor import Data.Maybe @@ -138,11 +138,12 @@ instanceReturnsDictCon LocalInstance = False matchGlobalInst :: DynFlags -> Bool -- True <=> caller is the short-cut solver -- See Note [Shortcut solving: overlap] - -> Class -> [Type] -> TcM ClsInstResult + -> Class -> [Type] -> Maybe CtLoc + -> TcM ClsInstResult -- Precondition: Class does not satisfy GHC.Core.Predicate.isEqualityClass -- (That is handled by a separate code path: see GHC.Tc.Solver.Dict.solveDict, -- which calls solveEqualityDict for equality classes.) -matchGlobalInst dflags short_cut clas tys +matchGlobalInst dflags short_cut clas tys mb_loc | cls_name == knownNatClassName = matchKnownNat dflags short_cut clas tys | cls_name == knownSymbolClassName = matchKnownSymbol dflags short_cut clas tys | cls_name == knownCharClassName = matchKnownChar dflags short_cut clas tys @@ -150,7 +151,7 @@ matchGlobalInst dflags short_cut clas tys | cls_name == typeableClassName = matchTypeable clas tys | cls_name == withDictClassName = matchWithDict tys | cls_name == dataToTagClassName = matchDataToTag clas tys - | cls_name == hasFieldClassName = matchHasField dflags short_cut clas tys + | cls_name == hasFieldClassName = matchHasField dflags short_cut clas tys mb_loc | cls_name == unsatisfiableClassName = return NoInstance -- See (B) in Note [Implementation of Unsatisfiable constraints] in GHC.Tc.Errors | otherwise = matchInstEnv dflags short_cut clas tys where @@ -1246,20 +1247,14 @@ addUsedGRE extends tcg_used_gres with imported GREs only. -} -- See Note [HasField instances] -matchHasField :: DynFlags -> Bool -> Class -> [Type] -> TcM ClsInstResult -matchHasField dflags short_cut clas tys +matchHasField :: DynFlags -> Bool -> Class -> [Type] -> Maybe CtLoc + -> TcM ClsInstResult +matchHasField dflags short_cut clas tys mb_ct_loc = do { fam_inst_envs <- tcGetFamInstEnvs ; rdr_env <- getGlobalRdrEnv - ; case tys of - -- We are matching HasField {k} {r_rep} {a_rep} x r a... - [_k_ty, _r_rep, _a_rep, x_ty, r_ty, a_ty] - -- Look up the field named x in the type r - | Just fl <- lookupHasFieldLabel fam_inst_envs x_ty r_ty - -- and ensure the field selector is in scope - , Just gre <- lookupGRE_FieldLabel rdr_env fl - - -> do { let name = flSelector fl - ; sel_id <- tcLookupId name + ; case lookupHasFieldLabel fam_inst_envs rdr_env tys of + Just (sel_name, gre, r_ty, a_ty) -> + do { sel_id <- tcLookupId sel_name ; (tv_prs, preds, sel_ty) <- tcInstType newMetaTyVars sel_id -- The first new wanted constraint equates the actual @@ -1289,35 +1284,61 @@ matchHasField dflags short_cut clas tys ; if not (isNaughtyRecordSelector sel_id) && isTauTy sel_ty then do { -- See Note [Unused name reporting and HasField] addUsedGRE AllDeprecationWarnings gre - ; keepAlive name - ; let maxCons = maxUncoveredPatterns dflags - ; let fallible_cons = rsi_undef $ sel_cons $ idDetails sel_id - ; suppress <- getSuppressIncompleteRecSelsTc - -- See (7) of Note [Detecting incomplete record selectors] in GHC.HsToCore.Pmc - ; unless (null fallible_cons || suppress) $ do - addDiagnostic $ TcRnHasFieldResolvedIncomplete name fallible_cons maxCons + ; keepAlive sel_name + ; warnIncompleteRecSel dflags sel_id mb_ct_loc ; return OneInst { cir_new_theta = theta , cir_mk_ev = mk_ev , cir_canonical = EvCanonical , cir_what = BuiltinInstance } } else matchInstEnv dflags short_cut clas tys } - _ -> matchInstEnv dflags short_cut clas tys } + Nothing -> matchInstEnv dflags short_cut clas tys } + +warnIncompleteRecSel :: DynFlags -> Id -> Maybe CtLoc -> TcM () +warnIncompleteRecSel dflags sel_id mb_ct_loc + | Just ct_loc <- mb_ct_loc + , not (isGetFieldOrigin (ctLocOrigin ct_loc)) + , not (null fallible_cons) + = traceTc "tc-warn" (ppr sel_id $$ ppr (ctLocOrigin ct_loc)) >> + (setCtLocM ct_loc $ addDiagnostic $ + TcRnHasFieldResolvedIncomplete (idName sel_id) fallible_cons maxCons) -lookupHasFieldLabel :: FamInstEnvs -> Type -> Type -> Maybe FieldLabel + | otherwise + = return () + where + maxCons = maxUncoveredPatterns dflags + fallible_cons = rsi_undef $ sel_cons $ idDetails sel_id + + -- GHC.Tc.Gen.App.tcInstFun arranges that the CtOrigin of (r.x) is GetFieldOrigin, + -- despite the expansion to (getField @"x" r) + isGetFieldOrigin (GetFieldOrigin {}) = True + isGetFieldOrigin _ = False + +lookupHasFieldLabel + :: FamInstEnvs -> GlobalRdrEnv -> [Type] + -> Maybe ( Name -- Name of the record selector + , GlobalRdrElt -- GRE for the selector + , Type -- Type of the record value + , Type ) -- Type of the field of the record -- The call (lookupHasFieldLabel fam_envs (LitTy "fld") (T t1..tn)) --- returns the `FieldLabel` of field "fld" in the data type T. +-- returns the `Name` of record selector Id for field "fld" in the data type T. -- A complication is that `T` might be a data family, so we need to -- look it up in the `fam_envs` to find its representation tycon. -lookupHasFieldLabel fam_inst_envs x_ty r_ty +lookupHasFieldLabel fam_inst_envs rdr_env arg_tys + | -- We are matching HasField {k} {r_rep} {a_rep} x r a... + (_k_ty : _r_rep : _a_rep : x_ty : r_ty : a_ty : _) <- arg_tys + -- Look up the field named x in the type r -- x should be a literal string - | Just x <- isStrLitTy x_ty + , Just x <- isStrLitTy x_ty -- r should be an applied type constructor , Just (tc, args) <- tcSplitTyConApp_maybe r_ty -- use representation tycon (if data family); it has the fields , let r_tc = fstOf3 (tcLookupDataFamInst fam_inst_envs tc args) -- x should be a field of r - = lookupTyConFieldLabel (FieldLabelString x) r_tc + , Just fl <- lookupTyConFieldLabel (FieldLabelString x) r_tc + -- and ensure the field selector is in scope + , Just gre <- lookupGRE_FieldLabel rdr_env fl + = Just (flSelector fl, gre, r_ty, a_ty) | otherwise = Nothing ===================================== compiler/GHC/Tc/Solver/Monad.hs ===================================== @@ -1663,7 +1663,7 @@ matchGlobalInst :: DynFlags -- See Note [Shortcut solving: overlap] -> Class -> [Type] -> CtLoc -> TcS TcM.ClsInstResult matchGlobalInst dflags short_cut cls tys loc - = wrapTcS $ TcM.setCtLocM loc $ TcM.matchGlobalInst dflags short_cut cls tys + = wrapTcS $ TcM.matchGlobalInst dflags short_cut cls tys (Just loc) tcInstSkolTyVarsX :: SkolemInfo -> Subst -> [TyVar] -> TcS (Subst, [TcTyVar]) tcInstSkolTyVarsX skol_info subst tvs = wrapTcS $ TcM.tcInstSkolTyVarsX skol_info subst tvs ===================================== compiler/GHC/Tc/Types/CtLoc.hs ===================================== @@ -229,8 +229,7 @@ data CtLocEnv = CtLocEnv { ctl_ctxt :: ![ErrCtxt] , ctl_bndrs :: !TcBinderStack , ctl_tclvl :: !TcLevel , ctl_in_gen_code :: !Bool - , ctl_rdr :: !LocalRdrEnv - , ctl_suppress_incomplete_rec_sels :: !Bool } + , ctl_rdr :: !LocalRdrEnv } getCtLocEnvLoc :: CtLocEnv -> RealSrcSpan getCtLocEnvLoc = ctl_loc ===================================== compiler/GHC/Tc/Types/LclEnv.hs ===================================== @@ -12,7 +12,6 @@ module GHC.Tc.Types.LclEnv ( , getLclEnvRdrEnv , getLclEnvTcLevel , getLclEnvThStage - , getLclEnvSuppressIncompleteRecSels , setLclEnvTcLevel , setLclEnvLoc , setLclEnvRdrEnv @@ -20,7 +19,6 @@ module GHC.Tc.Types.LclEnv ( , setLclEnvErrCtxt , setLclEnvThStage , setLclEnvTypeEnv - , setLclEnvSuppressIncompleteRecSels , modifyLclEnvTcLevel , lclEnvInGeneratedCode @@ -119,11 +117,8 @@ data TcLclCtxt tcl_arrow_ctxt :: ArrowCtxt, -- Arrow-notation context - tcl_env :: TcTypeEnv, -- The local type environment: + tcl_env :: TcTypeEnv -- The local type environment: -- Ids and TyVars defined in this module - - tcl_suppress_incomplete_rec_sel :: Bool -- True <=> Suppress warnings about incomplete record selectors - -- See (7) of Note [Detecting incomplete record selectors] } getLclEnvThStage :: TcLclEnv -> ThStage @@ -183,12 +178,6 @@ getLclEnvRdrEnv = tcl_rdr . tcl_lcl_ctxt setLclEnvRdrEnv :: LocalRdrEnv -> TcLclEnv -> TcLclEnv setLclEnvRdrEnv rdr_env = modifyLclCtxt (\env -> env { tcl_rdr = rdr_env }) -getLclEnvSuppressIncompleteRecSels :: TcLclEnv -> Bool -getLclEnvSuppressIncompleteRecSels = tcl_suppress_incomplete_rec_sel . tcl_lcl_ctxt - -setLclEnvSuppressIncompleteRecSels :: Bool -> TcLclEnv -> TcLclEnv -setLclEnvSuppressIncompleteRecSels suppress = modifyLclCtxt (\env -> env { tcl_suppress_incomplete_rec_sel = suppress }) - modifyLclCtxt :: (TcLclCtxt -> TcLclCtxt) -> TcLclEnv -> TcLclEnv modifyLclCtxt upd env = let !res = upd (tcl_lcl_ctxt env) ===================================== compiler/GHC/Tc/Types/Origin.hs ===================================== @@ -544,7 +544,7 @@ data CtOrigin -- IMPORTANT: These constraints will never cause errors; -- See Note [Constraints to ignore] in GHC.Tc.Errors | SectionOrigin - | HasFieldOrigin FastString + | GetFieldOrigin FastString | TupleOrigin -- (..,..) | ExprSigOrigin -- e :: ty | PatSigOrigin -- p :: ty @@ -716,7 +716,7 @@ lexprCtOrigin (L _ e) = exprCtOrigin e exprCtOrigin :: HsExpr GhcRn -> CtOrigin exprCtOrigin (HsVar _ (L _ name)) = OccurrenceOf name -exprCtOrigin (HsGetField _ _ (L _ f)) = HasFieldOrigin (field_label $ unLoc $ dfoLabel f) +exprCtOrigin (HsGetField _ _ (L _ f)) = GetFieldOrigin (field_label $ unLoc $ dfoLabel f) exprCtOrigin (HsUnboundVar {}) = Shouldn'tHappenOrigin "unbound variable" exprCtOrigin (HsRecSel _ f) = OccurrenceOfRecSel (unLoc $ foLabel f) exprCtOrigin (HsOverLabel _ l) = OverLabelOrigin l @@ -915,7 +915,7 @@ pprCtO ViewPatOrigin = text "a view pattern" pprCtO (LiteralOrigin lit) = hsep [text "the literal", quotes (ppr lit)] pprCtO (ArithSeqOrigin seq) = hsep [text "the arithmetic sequence", quotes (ppr seq)] pprCtO SectionOrigin = text "an operator section" -pprCtO (HasFieldOrigin f) = hsep [text "selecting the field", quotes (ppr f)] +pprCtO (GetFieldOrigin f) = hsep [text "selecting the field", quotes (ppr f)] pprCtO AssocFamPatOrigin = text "the LHS of a family instance" pprCtO TupleOrigin = text "a tuple" pprCtO NegateOrigin = text "a use of syntactic negation" ===================================== compiler/GHC/Tc/Utils/Monad.hs ===================================== @@ -94,7 +94,6 @@ module GHC.Tc.Utils.Monad( mkTcRnMessage, reportDiagnostic, reportDiagnostics, warnIf, diagnosticTc, diagnosticTcM, addDiagnosticTc, addDiagnosticTcM, addDiagnostic, addDiagnosticAt, - getSuppressIncompleteRecSelsTc, setSuppressIncompleteRecSelsTc, -- * Type constraints newTcEvBinds, newNoTcEvBinds, cloneEvBindsVar, @@ -400,8 +399,7 @@ initTcWithGbl hsc_env gbl_env loc do_this tcl_arrow_ctxt = NoArrowCtxt, tcl_env = emptyNameEnv, tcl_bndrs = [], - tcl_tclvl = topTcLevel, - tcl_suppress_incomplete_rec_sel = False + tcl_tclvl = topTcLevel }, tcl_usage = usage_var, tcl_lie = lie_var, @@ -1276,7 +1274,6 @@ mkCtLocEnv lcl_env = , ctl_tclvl = getLclEnvTcLevel lcl_env , ctl_in_gen_code = lclEnvInGeneratedCode lcl_env , ctl_rdr = getLclEnvRdrEnv lcl_env - , ctl_suppress_incomplete_rec_sels = getLclEnvSuppressIncompleteRecSels lcl_env } setCtLocM :: CtLoc -> TcM a -> TcM a @@ -1285,7 +1282,6 @@ setCtLocM (CtLoc { ctl_env = lcl }) thing_inside = updLclEnv (\env -> setLclEnvLoc (ctl_loc lcl) $ setLclEnvErrCtxt (ctl_ctxt lcl) $ setLclEnvBinderStack (ctl_bndrs lcl) - $ setLclEnvSuppressIncompleteRecSels (ctl_suppress_incomplete_rec_sels lcl) $ env) thing_inside {- ********************************************************************* @@ -1667,13 +1663,6 @@ add_diagnostic msg ; mkTcRnMessage loc (TcRnMessageWithInfo unit_state msg) >>= reportDiagnostic } -getSuppressIncompleteRecSelsTc :: TcRn Bool -getSuppressIncompleteRecSelsTc = getLclEnvSuppressIncompleteRecSels <$> getLclEnv - -setSuppressIncompleteRecSelsTc :: Bool -> TcRn a -> TcRn a -setSuppressIncompleteRecSelsTc b = updLclEnv (setLclEnvSuppressIncompleteRecSels b) - - {- ----------------------------------- Other helper functions ===================================== compiler/GHC/Tc/Validity.hs ===================================== @@ -1400,7 +1400,7 @@ checkSimplifiableClassConstraint env dflags ctxt cls tys -- (Coercible a b) to (a ~R# b) | otherwise - = do { result <- matchGlobalInst dflags False cls tys + = do { result <- matchGlobalInst dflags False cls tys Nothing ; case result of OneInst { cir_what = what } -> addDiagnosticTc (TcRnSimplifiableConstraint pred what) ===================================== testsuite/tests/pmcheck/should_compile/T24891.hs ===================================== @@ -12,7 +12,6 @@ data T a where f :: T Bool -> Int f x = x.sel -- warn, but only once, suggesting to match on T3 -{- data Dot = No | Yes {sel2 :: Int} ldiDot :: Dot -> Int @@ -29,4 +28,4 @@ data Dot2 t = No2 | Yes2 {sel3 :: t} accessDot2 :: HasField "sel2" t Int => Dot2 t -> Int accessDot2 x = x.sel3.sel2 -- warn about x.sel3 --} + View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ba79e31c58c8f1260e081dec07478628fba82e38 -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ba79e31c58c8f1260e081dec07478628fba82e38 You're receiving this email because of your account on gitlab.haskell.org. -------------- next part -------------- An HTML attachment was scrubbed... URL: